Springs and streams of the TarondashCeno Valleys (Northern Apennine Italy) Reactionpath modeling of waters interacting with serpentinized ultrama1047297c rocks
Tiziano Boschetti Lorenzo ToscaniEarth Sciences Department University of Parma vle GP Usberti 157A I-43100 Parma Italy
a b s t r a c ta r t i c l e i n f o
Article history
Received 6 May 2008
Received in revised form 18 August 2008Accepted 19 August 2008
In the area of the TarondashCeno Valleys (Northern Apennine Emilia-Romagna region Italy) waters of meteoricorigin interact with ophiolite rocks of the External Ligurides Fresh water springs issuing from basalts have aCandashHCO3 composition whereas freshwater springs from ultrama1047297tes vary in composition from CandashHCO3 orMgndashHCO3 to NandashOH or NandashSO4 types and in pH values from 73ndash88 up to pH 11 respectively In addition theboron content of the alkaline waters is up to 13 mgL which is unusually high for freshwaters in general andultrama1047297tes that have undergone oceanic serpentinisation in particular and gives a boric alkalinity to thewaters The springs waters show evidence of recent low-temperature continental serpentinisation and theprocess is modeled by reaction paths using an updated geochemical thermodynamic database consistentwith the local primary and secondary serpentinite paragenesis For the model bicarbonate waters evolve toalkaline waters supersaturated in Candash(Mg)-carbonate based on the assumption that the dissolution of serpentinite results in supersaturation with respect to kaolinite ferrihydrite vermiculite Fe2+ndashMg2+-saponite and poorly crystalline serpentine The alkaline composition and the chloride content of the waterssuggest a prolonged interaction with the rocks at depth that led to dissolution of albite and leaching of olivine-hosted 1047298uid inclusions A similar evolution is also proposed for the more developed springs issuingfrom the ultrama1047297c rocks of the Voltri Group (Liguria region) where solutions are supersaturated in bruciteand are in equilibrium with enstatite andor chlorite
copy 2008 Elsevier BV All rights reserved
1 Introduction
Many studies have reported the ubiquity of magnesium-bicarbo-nate and calcium-hydroxide spring waters that are associated withultrama1047297c rocks and that originated from meteoric water (eg Barneset al 1967 1972 1978 Barnes and OrsquoNeil 1969 Pfeifer 1977 Bruniet al 2002) According to Neal and Stanger (1984) these waters canpromote serpentinization at low temperature for example in theOman ophiolite where up to 10 of serpentines are estimated to besupergenic Other authors (Pfeifer 1977 Birsoy 2002 Bruni et al2002 Marini and Ottonello 2002) have downplayed the formation of serpentines during weathering of ultrama1047297c rocks theseinvestigators
consider the precipitation of sepiolite to be the dominant processLapham (1961) named deweylite the poorly crystalline serpentine thatprecipitates from colloidal suspensions at low temperature Never-theless researchers rarely study these serpentine-like mineraloids inthe 1047297eld or laboratory since poorly crystalline material is regardedwith suspicion Thus these mineraloids are not used to de1047297ne thecrystallographic chemical or structural details of serpentine mineralsIn the Northern Apennine possible protoserpentine analogues
genetically comparable with the microstructures forming mesh andbastites have been found and described (eg Mellini and Zanazzi1987 Viti and Mellini 1998) In the late 1970s hydrogen isotopescon1047297rmed the low-to-medium temperature origin (25ndash185 degC) of some chrisotile and lizardite minerals (Wenner and Taylor 1974)More recently in order to estimate the expected δ
18Ondashδ2H isotopic
composition of continental serpentine formed in equilibrium withmeteoric water at 25 degC Kyser et al (1999) and Barnes et al (2006)inferred a ldquoserpentine-H2O (25 degC)rdquo line Further the occurrence of highly deuterium depleted hydrogen gas emanating from hyperalka-line discharges also was used as an indication of ldquolow temperaturerdquo
serpentinization (Neal and Stanger 1983) However the isotopicevidence of continental-meteoric low temperature serpentinizationremains scarce This paucity of data may re1047298ect a fairly low waterrockratio during the process and consequently an oxygen isotopecomposition of serpentines largely inherited from precursor minerals(Barnes et al 1978)
Waterndashrock interaction modeling of CandashOH waters is dif 1047297cultparticularly when the detailed composition and thermodynamicparameters on the rock-forming minerals are unavailable (OrsquoHanley1996) For example in spite of the importance of vermiculite inweathering processes (Lee et al 2003) models in the literaturehave ignored their role for lack of thermodynamic data Otherfactors that play fundamental roles in controlling the serpentinite
Chemical Geology 257 (2008) 76ndash91
Corresponding author Present address via Stradone 658 47030 mdash S Mauro Pascoli(FC) Italy Tel +39 3470186527 fax +39 0521 905305
supergenic mineralogy are the dissolution-precipitation kinetics(Nesbitt and Bricker 1978 Siever and Woodford 1979) and climate(Wrucke 1996)
In this study the chemical and isotopic composition of severalspring waters issuing from different outcrops of ophiolitic rocks in theParma Province (Toscani et al 2001 Boschetti 2003) are presentedand modeled for 1047298uids circulating in and interacting with ultrama1047297crocks of the Northern Apennine The model shows the role of primary
and secondary minerals in the reaction path Solubility productsspeci1047297c for local phases were calculated and inserted in the modelingand the role of low-temperature serpentine in the chemical evolutionof groundwater from the TarondashCeno Valleys and the Voltri Group area(Bruni et al 2002 Marini and Ottonello 2002) is discussed
2 Geological and hydrogeological setting
The geology and petrology of Northern Apennine ophiolites(ultrama1047297tes basalts and cherts) have been extensively investigated(Abbate et al 1980 Beccaluva et al 1984 Ottonello et al 1984Rampone et al1995 1996) Most researchersagreethat the ophiolites(i) were generated in the PiedmontndashLigurian ocean during the Jurassicand Cretaceous periods (ii) are remnants of the ocean crust and
overlying sediments (iii) were fragmented into several nappes duringthe Upper CretaceousndashEocene stage of the orogeny and (iv) re1047298ect thefact that the entire allochthonous pile overrode the continental crustduring the Oligocene and Miocene periods
The Northern Apennine ophiolites crop out in two palaeogeo-graphic domains the Internal Ligurides (IL) and External Ligurides(EL) which are identi1047297ed on the basis of their current structuralcharacteristics and their relationship with the associated sedimen-
tary sequences (Abbate et al 1980 and references therein) IL contains strongly depleted peridotites serpentinites gabbros andbasalts (N-MORB) which are the basis of the Upper Jurassicndash
Paleocene sedimentary sequence ie deposits of Mn ore associatedwith radiolarian cherts silicic limestones Palombini shales withlimestones EL on the other hand contains less depleted lherzolitesin association with E-MORB (Ottonello et al 1984 Beccaluva et al1984 Rampone et al 1995 1996) In the TarondashCeno river valleyssprings issuing from ophiolite rocks occur in three areas (Fig 1) (A)Mt PennandashMt Aiona and Mt MaggiorascandashMt Nero where the Taroriver and the Ceno stream have their sources (B) the areas aroundthe Belforte Val Manubiola and Valmozzola along the middlecourse of the river and (C) Mt Prinzera where the Taro River entersthe Po Plain
Fig1 Simpli1047297ed geologic map of the Northern Apennine (modi1047297ed from Beccaluva et al 1984) The samples of this study are located in the TarondashCeno Valleys (sketched outline)1=Messinian to present 2 = parautochthonous and Piemontese Tertiary 3 = Tuscan ndashUmbrian sequences 4 = ophiolites and sedimentary sequences of the Liguride units 5 = high-pressure metamorphic ophiolites and calcschists of the Alpine Units 6 = Ligurides ophiolitic masses (EL = External Liguride unit IL = Internal Liguride unit) 7 = Middle Taro ldquoarea Brdquo
sampled BelfortendashVal ManubiolandashValmozzola White triangles mountains cited in the text with altitude in brackets (meters asl)
77T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
Basaltic rocks which are absent only from the area of Mt Prinzerasuffered an early low-grade oceanic metamorphism to varyingextents followed by an orogenic metamorphism (Cortesogno 1980)Albite epidote prehnite sericite chlorite actinolite titanite andhaematite association replaced the original mineralogy of plagioclaseolivine clinopyroxene ilmenite and magnetite (Cortesogno andLucchetti 1982) Spinel lherzolites re-equilibrated in the petrological1047297eld of plagioclase lherzolites and were exposed to oceanic and
orogenic metamorphisms Hydration of the early mineralogy pro-duced serpentine (lizardite and minor chrysotile) chlorite and talcand bastite after olivine and orthopyroxene respectively
The ultrama1047297tes of the Middle Taro valley and of Mt Prinzera arestrongly serpentinized (Venturelli et al 1997) whereas primaryclinopyroxene and plagioclase are largely preserved at Mt Aiona(Beccaluva et al 1973) Sedimentary rocks at the contacts with theophiolites are predominantly 1047298yschoid (marly to pelitic) and arenac-eous Typically 1047298ysch rocks are micritic marly turbidites withinterbedded pelites at MtCaio the latter are composed of illite andor illitesmectite mixed layers and chlorite (Venturelli and Frey1977)Arenaceous formations include (i) the Casanova Complex composedof quartz-feldspathic rocks with metamorphic serpentinitic andvolcanic clasts (Di Giulio and Geddo 1990) and (ii) the OstiaSandstones composed of quartz feldspar (plagioclase orthoclasemicrocline) mica (muscovite biotite) calcite and dolomite within aclayey and chlorite matrix at the base of the Mt Caio Flysch Formation(Mezzadri 1964)
Hydrogeologic studies on the TarondashCeno Valleys were focussedon their alluvial fan and plain probably because they cover adensely populated area Recently research projects consider theupstream section of the valleys and preliminary results show that77 of the springs are fed by aquifers contained in Late CretaceousLigurid 1047298ysches 11 in sandstones-pelite conglomerates andmassive sanstones (eg Ranzano Formation) and only 4 inophiolitic bodies (Segadelli et al 2006 De Nardo et al 2007)According to the authors the last value is probably underestimatedbecause the study consider only the known and gathered springs forwaterwork system whereas most of springs issuing from ophiolites
are free and unknown despite their perennial activity that ensurepreservation of the stream out1047298ow during dry periods De Nardoet al (2007) identify two types of ophiolite related springs (i)
springs with inde1047297nite permeability limit and (ii) with de1047297nitepermeability limit which are fed by aquifers contained in fractu-rated rocks and in coarse detritic covers respectively Some high1047298ow rate springs have different origin being related to gravitationalsliding (eg Mt Nero area) that increased permeability due toslackening of the pre-existing fractures (De Nardo et al 2007)
3 Analytical methods
31 Field methods
All water samples were 1047297ltered through cellulose acetate 1047297lters(045 μ m) An aliquot for cation analysis was acidi1047297ed with 65 HNO3
Suprapur Merck (1 cm3 HNO3 per 100 cm3 water) Labile parameters(temperature pH and Eh) were determined in the 1047297eld using anORION 250A instrument equipped with a Ross glass electrode formeasuring pH and a combined electrode of platinum-silversilverchloride for measuring Eh The electrode for Eh was calibrated atdifferent temperatures using ZoBells solution (Nordstrom 1977)Speci1047297c conductance at 20 degC was measured using a conductimeter(Model 85 Yellow Springs Instruments) Total alkalinity was deter-mined in the 1047297eld by acidimetric titration with 001 N HCl usingbromocresol green as an indicator and in the laboratory by Grantitration (Gran 1952) within 12 h of sampling Reduced dissolvedspecies were determined by spectrophotometry using a portablephotometer (Merck SQ300) with Spectroquant kits total ammoniumNH4= NH4
++ NH30 and total sulphide HS=H2S0+HSminus+ S2minus were deter-
mined by adding indophenol blue in basic solution (pH asymp13 method4500-NH3 F in APHA-AWWA-WEF1995) and methylene blue in acidicsolution (pHasymp05 method 4500-S2minus F in APHA-AWWA-WEF 1995)respectively
32 Laboratory methods
Clminus NO3minus and SO4
2minus were analyzed by ion chromatography usingDionex DX100 with anionic self-regenerating suppressor silica byspectrophotometry using a Merck SQ300 photometer and Na K Ca
Mg B Fe by radialsequential inductively coupled plasma opticalemission spectrometry (ICP-OES) using Philips PU7450 and Horiba
Jobin-Yvon Ultima 2 instruments Data on δ2H(H2O) and δ
18O(H2O)
Table 1
Summary of physico-chemical parameters and concentrations of chemical constituents in the spring waters coming from ultrama 1047297tes and basalts in the TarondashCeno Valleys
Type (lithology) Springs (ultrama1047297tes) Springs (basalts)
were collected using an automatic equilibration device (Finnigan GLF1086) in-line with a mass spectrometer (Finnigan Delta Plus) Theanalyzed H2 and CO2 gases were equilibrated with water samples at atemperature of 18 degC (Epstein and Mayeda 1953) using a Pt-richcatalyst Isotopic ratio are reported in per mill relative to V-SMOWand standard errors for δ
2H(H2O) and δ18O(H2O) are approximately plusmn
10permil and within plusmn01permil respectively
4 Chemical and isotopic results
The chemical compositionsof thewaters from theultrama1047297tesandbasalts are summarized in Table 1 Complete chemical and isotopedata are reported in Appendix A where all the water samples aresubdivided into four classes on the basis of the lithology present at thesource (i) ultrama1047297c (ii) basaltic (iii) sedimentary and (iv) surfacewaters
41 Chemical classi 1047297cation
All the samples investigated were freshwater (TDSb1 gL) Theternary classi1047297cation diagrams (Na+ K)ndashCandashMg and tAlkndashSO4ndashCl(Fig 2) show that waters of groups (ii) (iii) and (iv) do not differsigni1047297cantly many are bicarbonate-calcic with neutral to slightlybasic pH values (667ndash857) highly oxidized (Eh=371ndash514 mV) andcontain a small amount of chloride (Clminus up to 64 mgL) Sample UM1
from basalt is an exception since it is a ClndashCandashNa type with EC up to1043 μ S cmminus1 Clminus up to 231 mgL hardness up to 367 mgL CaCO3Waters issuing from ultrama1047297tes have more varied compositions Theyrange from Ca-bicarbonate with a pH of 73ndash88 and a hardness rangingfrommoderately hard to very hard (67ndash386mgL CaCO3 Eh129ndash581μ Scmminus1) to Mg-bicarbonate with a pH of 74ndash88 and a hardness from softto very hard (59ndash206 mgL CaCO3 92ndash330 μ S cmminus1) Among the Mg-bicarbonate waters sample UM8 shows the highest hardness values
Fig 2 MgndashSindashC ternary diagram (in molkg) and ternary classi1047297cation diagrams [Mgndash(Na+K)ndashCa plus ClndashtAlkndashSO4 in meqL] of alkaline springs in the TarondashCeno Valleys (this work)andthe VoltriGroup(Mariniand Ottonello 2002) Bottled watersfromTarovalleyare also representedRG1 andRG2Rocca Galgana 1990 and1994 respectively(chemical data fromwater bottle label) FN1 FN2 and FN3 Fontenova 1974 1990 and 1994 respectively (chemical data taken fromwater bottle label) MgndashSindashC contentsafter mineral dissolution (whitecrosses) are also represented (see text for details) Note that for dolomite and magnesite the CMg mole ratio refers to the products of the respective reactions CaMg(CO3)2+2H2O+
2CO2= Ca2+
+ Mg2+
+4HCO3
minus
and MgCO3+ H2O+CO2= Mg2+
+2HCO3
minus
79T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
(486 mgL CaCO3) conductivity (877 μ S cmminus1) and alkalinity (618 mgL HCO3) Springs PR10 and UM15 from ultrama1047297tes show an extremecomposition being sodium-hydroxide and sodium-sulfate respectivelytheir pH is high (101ndash112) their Eh is low (down to minus103 mV) theyaresoft (20ndash61 mgL CaCO3 220 to 408 μ S cmminus1) and they contain largeamounts of boron (up to 2294 μ gL and up to 13461 μ gL respectively)and moderate amounts of reduced sulfur and nitrogen species In bothPR10 and UM15 non-carbonate alkalinity predominates over carbonate
alkalinity using PHREEQCI software with the llnldat thermodynamicdatabase (Parkhurst and Appelo 1999) to calculate the composition of alkaline species under thepH Ehand T valuesmeasuredin1047297eld samplePR10 was found to contain 54 hydroxide 23 carbonate 19 borateand 4 other alkalinespecies (Si sulphide and ammonia) while sampleUM15 contained 75borate14 hydroxide10carbonate and 1 otherspecies
The SindashMgndashC diagram (Fig 2) is a useful tool for studying waterndashrockinteractions Dissolution reactions of typical magnesium minerals of ultrama1047297c rocks are represented by a line with CMg=2 (molar ratio)
Fig3 Plot of (A) Naand (B) Mgvs Cl forallwater samples from the TarondashCenoValleys including bottledwaters Rocca Galgana andFontenova(chemical data taken from labels)Alkalinesprings from the Voltri Group (smallblack diamonds Marini and Ottonello 2002) and meteoric water mean composition (MW Panettiere et al 2000 Venturelli 2003) are also plottedDottedcurvesin(A)and(B)refertomixingwith04and2wtNaCl 1047298uidinclusionsand MgCl2dissolution (1 mol)respectively Greycrosses andi csescorrespondrespectively to mixingwith04and2wtNaCl 1047298uidinclusionsfor mixing fractionsranging from 0001 to0005The solid linerefers tomixingof meteoric water(A) orMg-bicarbonate water(B) withseawater
The vertical grey line represents the water-soluble Cl content in the serpentinites from Northern Apennine (Barnes et al 2006) Symbols are as in Fig 2
80 T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
The Mg-bicarbonate watersissuingfrom ultrama1047297crocksfallnearthisline and cluster between the clinochlore and forsterite points (Fig 2)Stream waters and springs issuing from sedimentary and basaltic rocksfall between thecarbon apex and theline indicated by a CMg molar ratioof 4 which describes the dissolution of CandashMg-clinopyroxene
This is consistent with the (Ca)-bicarbonate compositions of thesestream waters and springs It is noteworthy that the Ceno streamwaters shifted toward the C apex of Fig 2 derive their chemistry fromthe interaction with Mt Caio Flysch cropping out widely in the Cenocatchment (samples from UM50 to UM59 see Map1 sheet in theannexes) whereas the Taro source spring waters plot towards theother apexes (lower C) typical of ophiolite minerals (samples fromUM41 to UM49 see Map1 sheet in the annexes) The use of the traceelements (eg Sr and Rb content Boschetti 2003) further highlightthe differences between the stream waters of the two rivers and therole of lithology Nonetheless mixing and equilibrium with atmo-sphere make the reaction-path modeling of these samples poorlyreliable so their are unsuitable for the purpose of the present work
Alkaline waters from ultrama1047297c rocks fall inside the same sub-triangle mostly along the CndashSi side probably due to the precipitationof magnesium minerals during the evolution of the water composition(Bruni et al 2002)
42 Sodiumndashchloride relationships
Chlorine and other halogens in serpentinized rocks are inheritedfrom seawater and marine sediments during oceanic metamorphism(eg Orberger et al 1990) They are transferred into serpentinite byadsorption (Anselmi et al 2000 Sharp and Barnes 2004) entrap-ment in 1047298uid inclusions (Scambelluri et al 2001ab) and substitutionin the structure of low-crystalline iron-hydroxide phases These latterare composed of the isomorphous and chemically similar iowaitehibbingite and FendashCl-oxydroxides (also named green rusts orfougerites Kohls and Rodda 1967 Saini-Eidukat et al 1994 Federet al 2005 and references therein) In the Northern Apennineserpentinite rocks the mean Cl content is sim300 ppm mean values of 221 and 375 ppm were measured respectively in the Erro ndashTobbio(Ligurian Western Alps Scambelluri et al 2004) and in the centralItaly serpentinites (Anselmi et al 2000) but thewater-soluble Cl from
serpentinites is estimated to be than 100 ppm (Barnes et al 2006) Asshown in Fig 2 waters from basalts and ultrama1047297tes of the TarondashCenoValleys show a trend towards the Cl apex whereas in Fig 3 theestimated soluble chlorine in serpentinites agrees with the highestvalue reported in spring waters from ultrama1047297tes (samples fromMarini and Ottonello analyzed in 2002) The Cl content of samplesUM1 and UM36is veryhigh(Fig 3) and is probablydue to pollutionInfact these springs are not far from the A15 highway (UM1) and theTomarlo Pass road (UM36) where salt mixtures comprising NaClMgCl2middot6H2O and probably CaCl2 are widely used for deicing in winterAlkaline water with a Mg-bicarbonate composition from ultrama1047297tesshow the highest and lowest Cl contents respectively Na-rich Mg-bicarbonate waters from basalts (UM6 UM7 Fig 3) and ultrama1047297tescan be explained by the dissolution of albite in basalts (Beccaluva
et al 1975) and in spilitized EL peridotites (Beccaluva et al 1984
Rampone et al 1995) respectively whereas the highest content of both Na and Cl in the alkaline water samples UM15 and PR10 (Fig 3)can be explained by leaching of NaCl 1047298uid inclusions in theserpentinite rocks (04ndash20 wt NaCl Scambelluri et al 2004)
43 Boron and BCl ratio
Compared with unaltered ultrama1047297c rocks serpentinites have
higher content of both boron and chlorine and most authors suggestthat both elements come from seawater (eg Sanford 1981 andreferences therein Bonatti et al 1984) Despite several studies onthese rocks the behavior of Cl and B in solution during waterndashrockinteraction processes is not well known Early investigations on 1047298uidsassociated with serpentinites in the western United States (Barnes andONeil 1969 Barnes et al 1972) revealed metamorphic waters withhighboron levels(up to 28mM) and BCl molar ratiosof 0038ndash013 aswell as Mg-bicarbonate and high-pH waters of meteoric origin withvariable BCl ratios of 0027ndash042 (median 0039) and 00006ndash0017(median 00076) respectively Bicarbonate waters issuing fromophiolitic rocks in Cyprus have BCl ratio in this range (00017ndash
0058 median 00019 Neal and Shand 2002) whereas low boron
Fig 4 Plots of (A) δ18O vs δ2H and (B) δ18O vs altitude for waters from the Taro ndashCenoValleys C-Italy and N-Italy are the meteoric water lines for Central and Northern Italy
respectively (Longinelli and Selmo 2003) Symbols are as in Fig 2
81T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
levels were found in the alkaline springs of meteoric origin in OmanNew Caledonia and ex-Yugoslavia (Barnes et al 1978)
Among the waters analyzed so far in the TarondashCeno Valleys thebicarbonate waters have BCl ratios in the range 00017ndash022 (median0073) whereas the alkaline waters show the highest ratios everreported (BCl= 034ndash33 median 169) The high boron content of upto 13 mgL of the alkaline springs could be related to metamorphic1047298uids released during serpentinite subduction in fact rocks prefer-
entially lose Cl and the BCl ratio of the released 1047298
uids increases withpressure and temperature (Scambelluri et al 2004) Nonetheless thealkaline water samples in our study differ signi1047297cantly from alkalinemetamorphic1047298uids in that they have low salinity high BCl ratios andmeteoric isotope signatures (see below) all these characteristicssuggest an origin in the interaction of meteoric water with high-pressure metamorphic serpentinite This hypothesis is furthersupported by the agreement in BCl ratio between the alkalinesamples we study here and the high BCl ratio of up to 4 previouslyreported for the residual high-pressure mineralogy of ultrama1047297tes(olivine+orthopyroxene Scambelluri et al 2004) In contrast theVoltri Groups Ca-hydroxide waters have mean values of B = 014 mgL and BCl=003 (Marini and Ottonello 2002) consistent with acontribution of boron and chlorine from 1047298uid inclusions In factassuming fractional contributions of 0001ndash0005 from NaCl 1047298uidinclusions with boron content from 4 to 8 mgL ( Fig 3a Scambelluriet al 2004) results in a B content of 0004ndash01 mgL in the solutionThese results suggest that the difference between springs from theVoltri Group and TarondashCeno Valleys is probably related to (i) differentevolution of the 1047298uids and (ii) the different types of host rocks and Pndash
T paths of the Ligurian Western Alps and EL peridotites (eg Beccaluvaet al 1984 Scambelluri et al 2004)
44 Isotopic composition of water
A diagram of δ2H vs δ18O (Fig 4A) shows that the waters sampledfrom the TarondashCeno Valleys are meteoric in origin since they areconsistent with the meteoric water lines of central and northern Italy(Longinelli and Selmo 2003) The altitude vs δ18O(H2O) (Fig 4B)
summarizes the isotopic effects due to temperature altitude andamount of rainfall (eg Criss1999) all of which help to determine theisotopic composition of the water samples The tendency is similar forbicarbonate waters from sedimentary basaltic and ultrama1047297c forma-
tions which may provide insight into the recharge areas alkalinewaters in contrast probably circulate deeper in the ground
5 Water ndashrock interaction model
51 Fundamentals
The reaction path model between meteoric water and serpentinite
was performed using PHREEQCI 2125 code (Parkhurst and Appelo1999) and the Lawrence Livermore National Laboratory thermody-namic database (llnldat as known as thermocomV8R6 ) implementedwith the equilibrium constants listed in Table 2
Computations wereperformed by reaction progress mode ( forward
modeling Plummer 1984 Zhu and Anderson 2002) ie (i) addingreacting rock at each step in the simulation (ii) activating the phase(s)precipitation only when the thermodynamic stability of the mineralparagenesis is well known and consistent with the mineralogy of thearea and (iii) plotting reaction progress on speci1047297c activity diagramsThe same thermodynamic database was used with The GeochemistsWorkbench 403 software (Bethke 2002) to generate Pourbaix andactivity diagrams (Figs 5 6 8 and 9) Modeling was performed usingan approach based on equilibrium rather than reaction transportbecause the data on reaction kinetics and reactive surfaces in thesesolid phases systems are scarce particularly with respect to theactivation energy of clay and mixed-layer minerals at low temperatureand high pH Moreover the hydrogeology and the discontinuity of theaquifers in the study area which are related to the geologicheterogeneity and structural complexity of the Northern Apenninewould make a model based on reaction transport even less credible atthis stage
The modeling in this study was performed on the basis of thefollowing knowledge (1) primary and secondary mineral assem-blages (2) chemical composition of the minerals involved (3) physicalparameters and chemical composition of the starting 1047298uid (4)thermodynamic data on pure phases
52 The solid reactant primary paragenesis
EL peridotites of the TarondashCeno Valleys (NorthernApennines Italy)are spinel lherzolites containing olivine (forsterite) orthopyroxene(enstatite) and clinopyroxene (diopside) (Giammetti 1968 and
Table 2
Thermodynamic data on hydrolysis reactions added to the llnldat database of the PHREEQCI code (Parkhurst and Appelo 1999) and thermocomV8R6dat database of TheGeochemists Workbench software (Bethke 2002)
Mineral Reaction log K (TdegC) ΔGdeg f References
0 25 60 100 (kJ mol-1)
clinochlore08-daphnite02 Mg4Fe1Al2Si3O10(OH)8 + 16H+=2Al3+ + Fe2++12H2O+4Mg2++3SiO2 7318 6384 5272 4269 minus786379 this workferr ihydrite 2-l ine Fe(OH)3+3H+= Fe3++ 3H2O 484 355 ndash ndash minus70850 Majzlan et al 2004ferr ihydrite 6-l ine Fe(OH)3+3H+= Fe3++ 3H2O 435 312 ndash ndash minus71100 Majzlan et al 2004liz09-berth01 Mg27Fe02Al02Si19O5(OH)4+64H+=02Al3++02Fe2++52H2O+27Mg2++19SiO2 3111 2786 2383 2005 minus399854 this workp-antigorite Mg
Saponite-Fe2+ Fe3175Si365Al035O10(OH)2+74H+=035Al3++3175Fe2++47H2O+365SiO2 1845 1635 1319 1001 minus452459 Wilson et al 2006Saponite-Fe2+ndashMg Fe0165Mg3Si367Al033O10(OH)2+732H+=033Al3++0165Fe2++466H2O+3Mg2++
367SiO2
3037 2729 2330 1921 ndash Marini and Ottonello(2002)
The Gibbs free energy of formation (ΔGo f ) of the two theoretical Fe3+ (Varadachari et al1994) and Fe2+ vermiculites were calculated using a previously published method (Vieillard
2002 and Vieillard pers comm)log K values extimated considering an ideal solid solution between the end-members and using data thermodynamic data from llnldat (clinochlore08-daphnite02) and Wilson
et al (2006) (liz09-berth01)
82 T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
references therein Venturelli et al 1997) Accessory phases areplagioclase picotite plusmn calciumndashsodic amphibole (Venturelli et al1997) Before modeling the crystal chemical formula of the mineralphases was calculated from data from Venturelli et al (1997) thendepending on the modal composition of the rock (Ernst and Piccardo1979) and the composition of the ideal solid solution of the minerals
the resulting stoichiometry of the solid reactant(s) was calculated andinserted in the input 1047297le of the PHREEQCI software (Table 3)Thermodynamic data of the primary mineral paragenesis are all wellknown and are included in the llnldat thermodynamic database of PHREEQCI
53 The solid reactant secondary paragenesis
Secondary minerals in the ultrama1047297tes of the TarondashCeno Valleysare serpentine (lizardite and rare chrysotile) smectite minerals (Fendash
Mg saponites) and Fe-oxi-hydroxides Other accessory minerals arechlorite mixed layer saponitendashchlorite (low-charge corrensite) talccalcite and dolomite (Beccaluva et al 1973 Dinelli et al 1997Venturelli et al 1997 Brigatti et al 1999 2000) In addition mixed
layer saponitendashtalc (aliettite) and vermiculitendashchlorite (high-charge
corrensite) are present (Alietti 1956 Settembre Blundo et al 1992Brigatti and Valdregrave 1996) Except for serpentine themodal abundanceof these phases is unknown In addition some of these minerals ndash
particularly interlayer clays ndash may have different origins (hydrothermalor weathering allogenicor authigenic)and different stabilitiesrelatedtokinetics and drainage 1047298ow rate For example chlorite weathering yields
vermiculite-bearing or saponite-bearing interlayers under good- andpoor-drainage regimes respectively (Bonifacio et al 1997 Środoń1999) In addition the thermodynamic data available for serpentineinclude information only on chrysotile which is less abundant thanlizardite in the serpentinites of the areas that we studied According toVenturelli et al (1997) the lizardite composition in the serpentinites of the Taro valley can be summarized as
which is similar to 1-T lizardite from the Monte Fico serpentinizedperidotite (Elba Island Northern Tyrrhenian Sea Italy Viti and Mellini1997)
Using the thermodynamic data of Wilson et al (2006) and
assuming an ideal solid solution between lizardite and berthierine
Fig 5 EhndashpH iron diagram of the spring waters coming from ultrama1047297c rocks of the TarondashCeno Valleys Bicarbonate and high-pH waters of the Voltri Group are also represented(dotted area andsmallblackdiamonds respectively) Thelowest Eh value of thehigh-pH waterswas calculated using theS(+VI)S(minus II)redox couple(Marini and Ottonello 2002)Thestability 1047297eld of mineral phases are in grey and those of aqueous species in white In diagrams A and B 6-line ferrihydrite-vermiculite and goethite were selected as subaerial solidphases respectively
83T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
we calculated ΔGf 0=minus399854 kJmol for a theoretical (09)(lizardite)+
(01)(berthierine) mineral (Table 2) The log K value of the stated solidsolution was calculated at different temperatures using PHREEQCIafter loading the thermodynamic data on the two end-members intothe database The log K results obtained for Al-lizardite were added tothe thermodynamic database of the solid reactant (Table 3) andcompared with data from the literature (Table 4)
54 Composition of the starting solution O 2 partial pressure and
minerals included in the model
The weighted average chemical composition of rainwater inBologna (Panettiere et al 2000) which is 95 km east of theinvestigated area was used as the starting solution in the modelFull chemical data on rainwater of the TarondashCeno Valleys are
unavailable
Total alkalinity was recalculated for log PCO2=minus35 bar and
T =11 degC which was the mean temperature of the springs sampledand for pe (or Eh) based on chemical data for the redox couple N(minusIII)N(+V) The values of pH and pe obtained were 566 and 452(Eh=025 V) respectively the former agrees with the median valuereported for rainwater in thearea (Venturelli et al1997) Because dataon Al Fe and Si in rainwater were lacking we calculated theconcentrations of these elements assuming logPCO2=minus30 bar andslight supersaturation relative to ferrihydrite-6 line and kaolinite Inthis way a hypothetical composition of water in equilibrium with thelocal soil was calculated (Table3) and adopted as the ldquoinitial waterrdquo formodeling
Evidence suggests that secondary minerals particularly Fe-bearingphases change the chemical composition of the water 1047298owingthrough the ultrama1047297tes of the TarondashCeno Valleys According to Neal
and Stanger (1983) and Fritz et al (1992) the production of CH4 and
Fig 6 Activity plotfor the MgOndashSiO2ndashAl2O3ndashFe2O3(FeO)ndashH2O systemat differentlog PO2 values Spring waters issuing from TarondashCeno Valleys (symbols as in Fig 2) andVoltri Groupultrama1047297tes are shown (grey area bicarbonate waters small black diamonds alkaline waters from Marini and Ottonello 2002) In plots A B and C the stability 1047297eld limits of vermiculite(Table 2) correspond to25 degCwhilethe amorphoussilica (am-SiO2) brucite serpentine (Sr1ndash4 in Table 4)and sepiolite(Sp1ndash2 in Table 4) phasescorrespondto 11 degCThepath calculatedfor interaction between waterand ultrama1047297c rockis also shown (dotted line)the number-letter combinations referto the appearanceand disappearance of minerals(see Table 5)
84 T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
H2 in thealkalinewaters1047298owing from theultrama1047297tes isrelatedto theconsumption of oxygen during the oxidation of iron forming thehematitendashmagnetite pair Therefore the activity of iron and theoxidation state of the water must be de1047297ned The analysis of watersamples from ultrama1047297tes were plotted in both the Pourbaix andactivity diagrams (Figs 5 and 6) where the calculated activity of Fe2+Mg2+ SiO2 SO4
2minus (Pourbaix) and log PO2 (activity diagrams) wereinvolved Iron concentration was quite variable in the TarondashCenosprings where the average content was 27 μ gL and the contentreached a maximum of 200 μ gL and in the Voltri springs where theaverage was 44 μ gL and the maximum was 948 μ gL (Marini andOttonello 2002) In both cases the variations are probably related to
the different mineralogies of the soils and the different solubilities of the Fe-rich phases like goethite and ferrihydrite Iron-bearingsaponites and montmorillonites are the prevailing clay minerals inthe area investigated probably because of the ubiquitous presence of iron oxy-hydroxide (Alietti et al 1976 G Venturelli pers comm) Forthis reason the model considers the involvement of pure Mg-smectiteto be negligible
In the Pourbaix diagram (Fig 5) the data from bicarbonate watersfall into the ferrihydrite-goethitendashvermiculite1047297elds and the data fromalkaline waters fall on the boundaries of the vermiculitesaponite andsaponitechlorite Since ferrihydrite is the 1047297rst precipitating Fe-phaseit was used rather than goethite in the modeling The diagram alsosuggests the transformation of ferrihydrite into magnetite during thechange from bicarbonate to alkaline composition but this process
requires a large amount of Fe2+ in solution (N2 mM Hansel et al2005) which is unlikely at this stage of the waterndashrock interactionThus the magnetite participating in the equilibrium is most likelyprimary in origin
The modeling and the activity diagrams (Figs 6 and 9) werecalculated using the average value PO2= 10minus26 or 10minus50 bar forbicarbonate and alkaline waters respectively These values arecompatible with those calculated by Marini and Ottonello (2002) forbicarbonate waters (PO2= 10minus40 bar) and for alkaline waters of theVoltri Group (10minus77 barbPO2b10minus42 bar) The lowest PO2 value whichwas used as a limit in the modeling was calculated from the S(+VI)S(minusII) redox couple Moreover the activity diagrams (Fig 6) indicatethat the data from the bicarbonate waters from the TarondashCeno Valleysfall between quartz whose log [SiO2] =minus433 is not shown because of
the slow precipitation kinetics under the PndashT conditions considered
here (Williams et al 1985) and amorphous silica boundary 1047297elds(Fig 6A) at the same manner bicarbonate waters with higher SiO2-activity from the Voltri Group fall on amorphous silica boundary(Fig 6B) In summary then the bicarbonate waters are supersaturatedin kaolinite ferrihydrite vermiculite and saponite Alkaline watersare also supersaturated with saponite (Fig 6AndashC) and they are alsoequilibrated with primary magnetite and under extreme conditionsprobably pyrite as well
55 Limitations of the model related to kinetics
Model calculations disregard the rates of mineral dissolution and
precipitation which becomes increasingly important as temperature
Table 3
Starting mineralogy (left side) and 1047298uid composition (right side) used in the modeling
Rock composition Fluid composition
Modal of the mineralsin the interacting rock
mol of the minerals usedin the modeling
meteoric water composition Initial 1047298uid composition used in the modeling
Spinel 003 Na 3523Eminus05 3523Eminus055 Spinel Chromite 001 S 3519Eminus05 3519Eminus05
Magnetite 0 Si ndash 1176Eminus06Kaolinite ndash 5258Eminus08
Ferrhydrite-6line ndash 5882Eminus07
The modal percentage of the minerals in a serpentinite from TarondashCeno Valleys (Ernst and Piccardo 1979) was re-calculated assuming an ideal solid solution between pure extremephasesand takingthe modal percentage tobe themass percentage Thelatter was used asthe solid reactant composition in thewaterndashrockinteraction modelPure solid phases werechosenfrom the llnldat thermodynamic database of the PHREEQCI softwareThe weighted average chemical compositionof the meteoric waterfrom Bologna (Panettiere et al 2000)was chosen as the initial 1047298uid composition in the modeling after its equilibration with kaolinite and ferrihydrite (see text for details)
Table 4
Summary of thermodynamic data for the sepiolite (Spn) and serpentine (Sr n) mineralsphases in italics do not refer to the mentioned hydrolysis reactions
ΔGf 0 log K Reference
(kJ molminus1) (at 25 degC)
Sepiolite (Spn) mdash Mg4Si6O15(OH)2middot6H2O+8H+=4Mg2++6SiO2+11H2Ominus925400 3004 Gunnarsson et al (2005)minus925133 3050andash3172 Christ et al (1973)
(Sp2) minus921986 3602andash3756 Wollast et al (1968)
Serpentine (Srn) mdash Mg3Si2O5(OH)4+ 6H
+
=3Mg
2+
+2SiO2+ 5H2ONi-lizardite minus 289324 1680 Mondesir and Decarreau (1987)
Al-lizardite2 minus411003 2544a Caruso and Chernosky (1979) Al-lizardite minus 339854 2786 this work (see text for details)
Mg-lizardite (Sr1) minus404025 3056 Wilson et al (2006)Mg-crysotile (Sr2) minus403702 3112 bthermocomV8R6Mg-lizardite minus402824 3340 Mondesir and Decarreau (1987)p-antigorite (Sr3) minus564239a 3479 Gunnarsson et al (2005)Mg-chrysotile (Sr4) minus406809 3669 Bruni et al (2002)Mg-chrysotileasbestos minus562505a 3888 Pfeifer (1977)Antigorite minus6612864 47719 bthermocomV8R6
The 1047297eld boundaries of the underlined phases shown in Fig 6 correspond to atemperature of 11 degC
a In the ΔG f deg column value calculated using the ΔG f
deg data of the elements from llnldat
and logK value in the reference in the LogK column value calculated using ΔG f deg
(mineral) in the reference and ΔG f deg data of the elements from llnldat
b Thermodynamic database of Lawrence Livermore National Laboratory included in
PHREEQCI and Geochemists Workbench softwares
85T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
decreases The waterndashrock interaction in the TarondashCeno basin iscomplicated by the presence of secondary mixed-layer clay mineralsthe origin of which is quite dif 1047297cult to model because the solidndashstatetransformations and weathering reactions are strictly related tohydrological and climatic conditions as described above An equili-
brium-based reaction path model is a valuable tool for assessing the
most probable reactions and for de1047297ning the product phases that canbe used in a kinetics-based model Since the kinetics of precipitationof Mg-clays at room temperature and pressure are poorly knownrelative reaction rate effects in the reaction path model can be mimedby omitting the stabler phase when a metastable mineral is expected
to precipitate Therefore the precipitation of MgndashFe3+
Abbreviations are as follows Fer-6line 6-line ferrihydrite kaol kaolinite verm vermiculite magn magnetite sap saponite serp serpentine clin clinochlore pyr pyrite calccalcite hydrom hydromagnesite bruc brucite enst enstatite The chromium decrease during the change in water facies from bicarbonate to alkaline was modeled withsupersaturation (precipitation) of eskolaite and thereafter with equilibrium with chromite (Boschetti 2003)- not bound dagger chlorite mineral from solution Dagger chlorite mineral added to the paragenesis
86 T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
montmorillonites vermiculite and saponites are predicted to occurrather than that of Mg-clays Moreover in order to simulate the slowprecipitation of calcite in depth precipitation was not activated atslight supersaturation (SIcalcite=001) during the initial stages of thewaterndashrock interaction in this manner saturation index of calcite wasfree to increase in the following steps and in alkaline waters
6 Discussion of the model results
61 Explanation of pre-modeling choices and post-modeling results
The results of the reaction path modeling are reported in Table 5and shown in Figs 6 and 9 The reaction paths model the dissolutionof a typical local serpentinite (mineralogically constrained as inTable 3) and the sequential precipitation of gibbsite kaoliniteferrihydrite vermiculite Fe2+ndashMg2+-saponite and a poorly-crystal-line serpentine( p-antigorite of Gunnarsson et al2005) The1047297rst stepinvolves the reaction of the ldquoinitial waterrdquo with serpentinite atconstant temperature (11 degC) PO2 (10minus26 bar) and PCO2 (10minus30 bar)this early stage corresponds to bicarbonate groundwater feeding theinvestigated spring system At the stage of primary magnetiteequilibrium and Fe2+ndashMg2+-saponite supersaturation the modelassumes the system to be closed to CO2 and to have a 1047297xed partialpressure of oxygen (PO2= 10minus40 bar) these conditions represent
when water switch from a bicarbonate to an alkaline compositionThe calculated amount of interacting serpentinite needed to obtainthe alkaline springs of the TarondashCeno Valleys (mean PCO2= 10minus7 bar)is approximately 00203 moles (see Table 5)
Bruni et al (2002) explained the origin of the alkaline waters of the Voltri Group as the result of the evolution of meteoricbicarbonate water following the precipitation of sepiolite withabundance similar to that of goethite This might be supported bythe observation that sepiolite or smectite precipitate at lowtemperature from solutions with low Al content when the pH is inthe range 75ndash95 or N95 respectively (Siffert and Wey 1962 Jonesand Galan 1988) but sepiolites form from the weathering of serpentines in arid climate regions whereas sepiolites are rare inhumid climate regions where serpentinite changes to Fe-smectite
(Singer 1989 Wrucke 1996) In the ophiolite rocks of the Northern
Apennine sepiolite is found only sporadically in veins (Giuseppetti1953 Veniale 1966) and is unknown in the ophiolites of the Tarondash
Ceno Valleys whereas it is thought that sepiolite of the Voltri Groupprecipitated from metasomatic 1047298uids during development of shearzones (Crispini and Capponi 2002)
As described above our model explains the transition frombicarbonate water to alkaline water (UM15 and PR10 samples) asthe result of low-temperature precipitation of a poor crystalline
serpentine with the solubility product of p-antigorite (see Sr3 inTable 4 and Fig 6) The model indicates that the bicarbonate waters of the Voltri Group with the highest [Mg2+][H+]2 activity ratios similarlyevolve towards alkaline waters (Fig 6B) The logK value of thesynthetic low temperature serpentine adopted in this model and its
Fig 7 Saturation indices of chrysotile and p-antigorite (Sr2 and Sr3 thermodynamic
data respectively in Table 4) for bicarbonate and alkaline waters coming fromultrama1047297c rocks of the TarondashCeno Valleys as a function of SiMg molar ratioThermodynamic data on chrysotile and p-antigorite are from the thermocomV8R6
database and Gunnarsson et al (2005) respectively
Fig 8 Activityplot forthe MgOndashSiO2ndashAl2O3ndashFe2O3(FeO)ndashH2O system analyzed forSiO2-phases and aqueous species at different log PO2 and log[aAl3+(aH+)3] values Springwaters coming from the TarondashCeno Valleys (symbols as in Fig 2) and Voltri Groupultrama1047297tes (symbols as in Fig 5) are shown The greyarea refers to bicarbonate watersfrom theVoltri Group (Mariniand Ottonello2002Bruniet al 2002) Thesepiolite 1047297eldis very narrow or hidden by saponite-Mg (1047297eld limits shown as continuous lines)saponite-Fe3+ndashMg (1047297eldlimits shown as dottedlines) or vermiculite(1047297eldlimits shownas dottedndashdashed lines) The talc stability 1047297eld and boundaries with saponite-Mg
montmorillonite-Mg and cronstedite-Mg are shown for comparison (dashed lines)
87T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
relative uncertainty of plusmn2 (see Fig 8 in Gunnarsson et al 2005) arecomparable to the logK variation reported for other low-temperatureMg-serpentines (Table 4) that is within the lines Sr1ndash4 for serpentineminerals where samples varying from bicarbonate composition toalkaline composition in the ultrama1047297c aquifer are grouped (Fig 6)Fig 7 shows a plot of calculated serpentine saturation index as a
function of the SiMg molar ratio for the water samples from theultrama1047297tes of the TarondashCeno ValleysThe SI andSiMgvalues of pureMg-serpentine (SI= 0 SiMg=067) provide a useful index forseparating bicarbonate waters (SIb0 SiMgb067) from alkalinewaters (SIN0 SiMgN067) and the distribution of data points isconsistent with the idea that bicarbonate waters approach saturationwith serpentine until they become supersaturated as alkaline waters
Based on the Mgvs SiO2(aq) activitydiagram and speci1047297cratiosoflog[aAl3+(aH+)3] Birsoy (2002) analyzed the behavior of sepiolite duringultrama1047297te weathering The diagram produces a very narrow stability1047297eldfor sepiolite(Fig 8) when it is programmedwith suitable conditionsfor the bicarbonate waters (log PO2=minus25 bar log [aAl3+(aH+)3]= 65) andthe alkaline waters (log PO2=minus50 bar log [aAl3+(aH+)3]= 45) sampled inthe present study This stability 1047297eld grows larger when the absence of
Mg-smectites and talc is assumed but this is an unrealistic conditionbecause (i) all these minerals occur in the rocks within the areainvestigated and (ii) at high [Mg2+][H+]2 activity ratios the 1047297eld of sepiolite is smaller than that of subordinate compared to those poorly-crystalline serpentine (p-antigorite) and clinochlore
Among the alkaline springs of the Voltri Group the waters withlow log [aCa2+(aH+)2] values (Fig 9) show patterns compatible withchlorite interaction (step 5c in Fig6D)The alkaline waters whose datafall near the clinochlore-14Aring 1047297eld (Fig 6D) cannot be explained bysepiolite precipitation because the sepiolite line does not cross thechlorite 1047297elds Thus the peculiar composition of some alkaline watersof the Voltri Group is probably due to interactions between water andprimary chlorites of the ultrama1047297c rocks (optional step 5c in Table 5cb) This is supported by the occurrence sometimes in large amounts
of metasomatic Mgndash(Fe)-chlorite (Scambelluri and Rampone 1999
Vissers and Strating 2001) in the Ligurian area where the relativeabundance of chlorite can reach 50 in veins in high strainserpentinite mylonites of the ErrondashTobbio ultrama1047297tes (Scambelluriet al 2001b) Otherwise in a chlorite-free primary paragenesis thesolution evolving on the Sr3 serpentine line reaches equilibrium withclinochlore but probably the latter does not precipitate due to theslow kinetics of precipitation under these conditions Thus thedissolution of additional reacting rock in the iterative weathering
process supplies Mg
2+
to the solution thereby favoring the Sr3 to Sr4poorly crystalline serpentine transformation In fact all the PR10samples and some of the alkaline springs from the Voltri Group showincreasing [Mg2+][H+]2 activity ratio and low variation in silicaactivity (Fig 6D) starting from the invariant triple point 4a (Sr3-serpentine Fe2+ndashMg-saponite chlinoclore-14 Aring saturation) andextending towards the brucite line This path can be explained asthe result of the transformation of poorly crystalline serpentine fromSr3 to Sr4 the latter is more stable at higher [Mg2+][H+]2 activityratios Finally at very low PO2 values vermiculite disappears and thestability 1047297elds of chlorite and Fe2+ndashMg-saponite become narrowerwhereas those of Fe2+-rich phases (eg Fe2+-saponite) become widerThe data for the alkaline waters with the highest log [aMg2+(aH+)2]from the Voltri Group fall near the stability 1047297eld of peridotite primaryminerals like forsterite and enstatite (Fig 6D) therefore the last stepsof theevolution involve theequilibration of thesolutionwithenstatite(steps 8a and 8c in Table 5ab)
62 Ca-rich alkaline waters
In preliminary modeling of the evolution of the TarondashCeno springwaters Boschetti (2003) concluded that meteoric water initiallysaturated in Fe3+ndashMg2+-chlorite Fe3+ndashMg2+-saponite and pyrite (andor goethite) and then interacted with the ultrama1047297te mineralassemblage eventually evolved into the alkaline waters In themodel of Boschetti (2003) chlorite saponite and calcite were added
Fig 9 Activity plot for the MgOndashCaOndash(CO2)ndashH2O system with phase limits calculated ata temperature of 11 degCSpring waterscoming from theTarondashCeno Valleys (symbols as inFig2) andVoltri Groupultrama1047297tes(symbolsas in Fig5) areshown Calculated path forwaterndashultrama1047297c rock interaction is shown (dotted line) with the number ndashlettercombinations referring to the appearance or disappearance of minerals (see Table 5)The grey area refers to bicarbonate waters from the Voltri Group (Marini and Ottonello2002)
Fig 10 Plots of (A) calcite saturation indexes and (B) PCO2 as a function of Clminus ascalculated by PHREEQCI software for waters from the TarondashCeno Valleys Alkalinesprings for the Voltri Group are also shown (small black diamonds Marini andOttonello 2002) The grey shading is deeper(A) in the saturation1047297eld(SI= 00plusmn05) and
(B) where the lowest log PCO2 value is approached
88 T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
to the solution in order to achieve a high [Ca2+][H+]2 activity ratio andsupersaturation in these phases However in light of the work of Clarket al (1992) and Bruni et al (2002) we believe that the calcitesupersaturation in alkaline waters (Fig 9 and 10) which in our modelis achieved by preventing the precipitation of calcite in the 1047297rst stepsof the reaction path (Table 5a and b) depends on the differencebetween CO2 partial pressureat thegivendepth(PCO2le10minus10bar) andat the surface (PCO2cong10minus35) Waters rising to the surface are far from
equilibrium with the atmosphere where the PCO2 is 4times10
6
timesgreater than the value at depth thus the water rapidly take up CO2which is promptly converted to CO3
2minus because of the high pH Thisleads to supersaturation with calcite through the following hydro-xylation reaction
As a result large amounts of travertine are generated in the VoltriGroup (Marini and Ottonello 2002) and carbonate concretionsknown as devilrsquo s coins are formed at Mt Prinzera in Taro valleyBelow the surface the SIcalcite of the alkaline waters is expected to belower than what the model estimates Under conditions of lowcarbonate alkalinity low CO2 and O2 partial pressure and high pHcalcite nucleation may be inhibited and the kinetics of precipitationreduced by the high concentrations of dissolved divalent cations suchas Mg2+ Fe2+ and Mn2+ (Sumner and Grotzinger 1996 Davis et al2000 Zhang and Dawe 2000 de Leeuw 2002) The fact that severalalkaline water samples fall within or near the calcitendashdolomiteequilibrium (Fig 9) may indicate magnesium calcite or protodolomitenucleation in those extreme subsurface conditions However possibleinteraction between some calcium-rich waters of the Voltri Group andCa-rich rocks such as ophicarbonates or rodingites cannot beexcluded these rocks are widespread in the Ligurian Apennineserpentinites (Treves and Harper 1994)
63 Sepiolite vs serpentine nucleation under weathering conditions
constraints on the kinetics and equilibrium-based approaches
The kinetics of precipitation of poorly-crystalline serpentine and
sepiolite cannot be compared quantitatively because rate constantsare available only for well-ordered crystalline phases (eg see Palandriand Kharaka 2004 for crystalline serpentine) Sepiolite precipitationis often invoked to explain magnesium scavenging from solutions butthe lack of kinetic data for this mineral leads to overly simplisticreaction transport models These models are forced to use kinetic datafor phyllosilicates such as kaolinite or muscovite (eg Xu et al 2001Bryan 2005) Moreover the low-temperature thermodynamic dataeven if they are fairly uncertain show that sepiolite solubilityincreases with decreasing crystallinity (Table 4 Stoessell 1988) Infact crystalline sepiolite requires a long time to reach equilibriumwith solution which imposes a slow control on thewater composition(eg Stoessell and Hay 1978) Consequently in modeling based onreaction transport that simulates freshly precipitated or poorly
crystalline sepiolite formation the logK value of Wollast et al (1968)is preferable to the lower values (Suarez and Šimůnek 1996) Thispreference agrees with the approach of Mariner (2003) and Jove Colon(2006) who adopt a logK (25 degC) value for sepiolite of 3044 It is alsosimilar to that reported by Wollast et al (1968) and it is six logK unitshigherthan the valueused in thermocomV8R6 dataset which is oftenused (perhaps improperly) to model the supergenic nucleation of sepiolite In the stability diagram in Fig 6 where the Sp1 and Sp2 linesrepresent the stability boundaries of crystalline and poorly-crystallinesepiolite respectively the evolution of TarondashCeno waters frombicarbonate to alkaline (Figs 6A and C) follows the stability 1047297eldboundary of the poorly-crystalline serpentine (between Sr2ndashSr3 lines)rather than that of the poorly-crystalline sepiolite (Sp2) as postulatedin themodel The lines for Sp2 Sr3 and amorphoussilica intersect at a
common point (Fig 6D) which is consistent with the nucleation of
sepiolite that typically occurs in solutions that are saturated orsupersaturated with amorphous silica and (Mg)-carbonate (egStoessell and Hay 1978 Birsoy 2002) Some bicarbonate watersfrom the Voltri Group have a silica activity near the amorphous silicaboundary (Fig 6B) but only a few alkaline springs of this group fall onthe poorly-crystalline sepiolite boundary line (Sp2 Fig 6D) Both of these features are inconsistent with the possibility that sepioliteprecipitates from supergenic fresh water in the Northern Apennine
Moreover these two results suggest that values for logK(sepiolite)b
3044 (see Table 4) should be used to model the dissolutionprecipitation of the well-crystallized phase that is the most commonform for the metasomatic sepiolite found in the investigated area
7 Conclusions
The waters issuing from ultrama1047297tes in the TarondashCeno Valleysoriginate from low-temperature interactions between meteoric waterand serpentinite rocks In this report the waterndashrock interaction hasbeen modeled using reaction paths in progress mode During their earlyinteraction with soil and rock minerals bicarbonate waters super-saturated in kaolinite ferrihydrite vermiculite and Fe2+ndashMg2+-saponitewhere theorigin of the secondary iron saponite as Fe2+-bearing smectiteagrees with the hypothesis of Guumlven (1988) Then while remainingsupersaturated with saponite the bicarbonate waters evolve into analkaline facies when poorly-crystalline serpentine precipitates Duringrock weathering the Al-bearing crystalline serpentine of oceanic originprobably promotes nucleation of the low-temperature serpentineamorphous phase and supplies Al3+ to the solution for saponiteprecipitation This interpretation is consistent with the existence of Al-poor supergenic serpentine (Neal and Stanger 1985) and the partiallymeteoric OH-isotopic imprint observed in other ultrama1047297c systems
Dif 1047297culties and complexities in the modeling arise fromuncertainties in the primary mineralogy of some local ophiolitesand from the presence of different lithotypes (rodingites ophi-carbonates gabbros) which can signi1047297cantly affect the composi-tion of water samples This can be seen in the example of the Candash
OH springs issuing from the Voltri Group (PO2 and PCO2 of 10minus70
and 10minus10 bar respectively Bruni et al 2002 Marini and
Ottonello 2002) These springs appear to be more evolved thanthe TarondashCeno alkaline waters as a result of their particularly longinteractions with serpentinites until equilibrium with enstatite isreached or with chlorite-bearing serpentinite The use of anupdated thermodynamic database in our model avoided problemsdue to mineralogical heterogeneity in the mixed-layer claysassociated with ophiolitic rocks this database accurately re1047298ectedthe chemical composition of the single phases Using this modelthe vermiculitendashchlorite and saponitendashchlorite associations seen inthe activity diagrams simulate respectively the presence of highand low charge corrensite observed in the 1047297eld The accuracy of ourmodel can be tested by comparing estimates of chlorine composi-tion resulting from our model and from experimental methods The
mean Cl concentration of the bicarbonate water is 41plusmn16 mgLtaking into account a Cl concentration of sim2 mgL in rainwater inthe area (cf Panettiere et al 2000 Venturelli 2003) 2 mgL of Clin spring water are estimated to come from serpentinite This valueagrees with the model which predicts that the switch frombicarbonate water to alkaline water requires the congruentdissolution of 00203 mol of serpentinite containing 100 ppm of soluble Cl thereby yielding water containing 2 ppm of Cl
In addition we think that weathering of the accessory albiteplagioclase is responsible for Na enrichment of the waters whereas Clenrichment in the alkaline waters is predicted to result from theleaching of NaCl 1047298uid inclusions Moreover the TarondashCeno alkalinesprings show the highest BCl ratios comparable with those of thehost metamorphic minerals (Scambelluri et al 2004) In contrast the
lower boron content and BCl ratio of the alkaline springs of the Voltri
89T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
Group are unexpected features and probably re1047298ect waterndashrockinteractions over a longer time period than in the TarondashCeno ValleysThe boron content in the Northern Apennine alkaline springs may becontrolled by several factors such as (i) the different histories of VoltriGroup and EL ultrama1047297tes (ii) the probable contribution of boronfrom NaCl 1047298uid inclusions (iii) the sinking of boron by travertineprecipitation (Hobbs and Reardon1999) brucite (Rhoades et al 1970)and smectite and (iv) decreases in boron release from minerals due to
increases in pH (Su and Suarez 2004) Acknowledgements
This research was carried out with the 1047297nancial support from theldquoYoung Researchers Projectrdquo (2001ndash2002) of the Italian Ministry of Universities Scienti1047297c Research and Technology (MURST) We thankGianniCortecci (Institute of Geosciences and EarthResources IGG-CNRPisa Italy) and Giampiero Venturelli (Earth Sciences Department Univer-sity of Parma Italy) for useful suggestions and critical review of themanuscript The authors are grateful to Enrico Dinelli (InterdepartmentResearch Centre for Environmental Sciences Alma Mater StudiorumUniversity of Bologna Ravenna Italy) for granting permission to use theGeochemistrsquos WorkbenchsoftwareTheconstructive reviews of theEditor
J Fein and of two anonymous Referees have signi1047297cantly improved the
manuscript
Appendix A Supplementary data
Supplementary data associated with this article can be found inthe online version at doi101016jchemgeo200808017
References
Abbate E Bortolotti V Passerini P 1980 Olistostromes and olistoliths In Sestini G(Ed) Development of the Northern Apennines geosyncline Sedimentary Geology4 pp 521ndash558
Alietti A 1956 Il minerale a strati misti saponite-talco di monte Chiaro (val di TaroAppen Emil) Rendiconti dellAccademia Nazionale dei Lincei 21 201 ndash207
Alietti A De Angelis G Montanari G Sgarlata F Tirelli G 1976 Il ferro nelle argilleappenniniche Periodico di Mineralogia 45 51ndash64
Anselmi B Mellini M Viti C 2000 Chlorine in the Elba Monti Livornesi and Murloserpentines Evidence for seandashwater interaction European Journal of Mineralogy12137ndash146
Apha-Awwa-Wef 1995 Standard Methods for the Examination of Water and Waste-water 19th Edition American Public Health Association American Water WorksAssociation Water Environment Federation
Barnes I LaMarche Jr VC Himmelberg GR 1967 Geochemical evidence of present-day serpentinization Science 156 830ndash832
Barnes I ONeil JR 1969 The relationship between 1047298uids in some fresh alpine-typeultrama1047297cs and possible modern serpentinization Western United StatesGeological Society of America Bulletin 80 1947ndash1960
Barnes I ONeil JRTrescases JJ1978 Present dayserpentinizationin New CaledoniaOman and Yugoslavia Geochimica et Cosmochimica Acta 42 144ndash145
Barnes I Rapp JB ONeil JR Sheppard RA Gude III AJ 1972 Metamorphicassemblages and the direction of 1047298ow of metamorphic 1047298uids in four instances of serpentinization Contributions to Mineralogy and Petrology 35 263ndash276
Barnes JDSelverstone J Sharp ZD 2006Chlorine isotope chemistry of serpentinitesfrom Elba Italy as an indicator of 1047298uid source and subsequent tectonic historyGeochemistry Geophysics Geosystems 7 doi1010292006GC001296
Beccaluva L Emiliani F Venturelli G Zerbi M 1973 Ca Fe Mg Mn Cr Ni Codistribution in some ultrama1047297c rocks outcropping in the Northern Apennines withsome geological remarks Ateneo Parmense Acta Naturalia 9 69ndash98
Beccaluva L Macciotta G Piccardo GB Zeda O 1984 Petrology of lherzolitic rocksfrom the Northern Apennine ophiolites Lithos 17 299 ndash316
BeccaluvaL Venturelli G ZanzucchiG 1975 Dati geologici e geochimici sui basalti diassociazione o1047297olitica dellAppennino ligure-emiliano Ateneo Parmense ActaNaturalia 11 789ndash802
BethkeCM 2002 The Geochemists Workbench A Userrsquos Guide to RxnAct2Tact ReactandGtplot Release40 HydrogeologyProgram Universityof IllinoisUrbanaIL 236pp
Birsoy R 2002Formation of sepiolitendashpalygorskite and related minerals from solutionClays and Clay Minerals 50 736ndash745
Bonatti E Lawrence JR Morandi N 1984 Serpentinization of oceanic peridotitestemperature dependence of mineralogy and boron content Earth and PlanetaryScience Letters 70 88ndash94
Bonifacio E Zanini E Boero V Franchini-Angela M 1997 Pedogenesis in a soilcatena on serpentinite in north-western Italy Geoderma 75 33ndash51
Boschetti T 2003 Studio geochimico e geochimico-isotopico di acque a composizione
estrema e termali dellrsquoAppennino Settentrionale Ph D Thesis Parma Univ
Brigatti MF Franchini G Lugli C Medici L Poppi L Turci E 2000 Interactionsbetween aqueous chromium solutions and layer silicatesApplied Geochemistry 151307ndash1316
Brigatti MF Lugli C Poppi L Venturelli G 1999 Iron-rich saponite dissolutionreactions and Cr uptake Clay Minerals 34 637ndash645
Brigatti MF Valdregrave G 1996 Mixed-layer chloritesmectite (corrensite) in ophioliticrock alteration products Neues Jahrbuch fuer Mineralogie Monatshefte 1 37 ndash47
Bruni J Canepa M Chiodini G Cioni R Cipolli F Longinelli A Marini L OttonelloG Vetuschi Zuccolini M 2002 Irreversible waterndashrock mass transfer accompany-ing the generation of the neutral MgndashHCO3 and high-pH CandashOH spring waters of the Genova province Italy Applied Geochemistry 17 455ndash474
Bryan CR 2005 Drift-Scale THC Seepage Model MDL-NBS-HS-000001 REV 04 YuccaMountain Project Las Vegas Nevada (US)Caruso LJ Chernosky JV 1979 The stability of lizardite Canadian Mineralogist 17
757ndash769Clark ID Fontes JC Fritz P 1992 Stable isotope disequilibria in travertine from high
pH waters Laboratory investigations and 1047297eld observations from Oman Geochi-mica et Cosmochimica Acta 56 2041ndash2050
Cortesogno L 1980 Il metamor1047297smo giurassico nelle o1047297oliti dellAppenninoSettentrionale due distinti cicli metamor1047297ci in ambiente oceanico O1047297oliti 5 5ndash18
Christ CLHostetler PB Siebert RM1973 Studies in the system MgO-SiO2-CO2-H2OIII American Journal of Science 273 65ndash83
Cortesogno L Lucchetti G 1982 Il metamor1047297smo oceanico nei gabbri o1047297oliticidellAppennino Ligure Aspetti mineralogici e petrogenetici Rendiconti dellaSocietagrave Mineralogica Italiana 38 561ndash579
Crispini L Capponi G 2002 Control of metasomatic alteration on the development of shear zone in ultrama1047297c rocks a case study from the Voltri Massif (Ligurian Alps)Architettura interna e cinematica delle zone di taglio Gruppo Italiano di GeologiaStrutturale mdash Riunione Annuale 2002 Pisa Italy 11ndash12 giugno 2002 XIV ndashXV
CrissRE 1999 Principlesof stableisotope distributionOxford University PressNew York
Davis KJ Dove PM De Yoreo JJ 2000 The role of Mg 2+ as an impurity in calcitegrowth Science 290 1134ndash1137
de Leeuw NH 2002 Moleculardynamics simulations of the growth inhibiting effect of Fe2+ Mg2+ Cd2+ and Sr2+ on calcite crystal growth Journal of Physical Chemistry B106 5241ndash5249
De Nardo MT Segadelli S Vescovi P 2007 Studio pilota per la caratterizzazionegeologicadellesorgentinellazona delM Nero (altaVal Cenoe alta ValNuremdash provincedi Parma e Piacenza) Il Geologo dellEmilia-Romagna 25 Nuova Serie pp 5 ndash21
Di Giulio A Geddo G 1990 Studio petrogra1047297co delle Arenarie di Casanova (alta valTrebbia Appennino settentrionale) Atti Ticinesi di Scienze della Terra 33 243ndash254
Dinelli E Lombini A Simoni A Ferrari C 1997 Heavy metals in serpentinite soils of selected outcrops of Piacenza and Parma provinces (Northern Apennines Italy)Mineralogica et Petrographica Acta 40 241ndash255
Epstein S Mayeda T 1953 Variation of 18O content of waters from natural sourcesGeochimica et Cosmochimica Acta 4 213ndash224
Ernst WG Piccardo GB 1979 Petrogenesis of some Ligurian peridotites I Mineraland bulk-rock chemistry Geochimica et Cosmochimica Acta 43 219ndash237
Feder F Trolard F Klingelhofer G Bourrieacute G 2005 In situ Moumlssbauer spectroscopy
evidence for green rust (fougerite) in a gleysol and its mineralogical transforma-tions with time and depth Geochimica et Cosmochimica Acta 69 4463ndash4483Fritz PClark IDFontes JCWhiticar MJFaberE 1992 Deuterium and
13C evidence
for low temperature production of hydrogen and methane in a highly alkalinegroundwater environment in Oman In Kharaka TK Maest AS (Eds) Proceed-ings of the 7th International Symposium on WaterndashRock Interaction BalkemaRotterdam pp 793ndash796
Giammetti F 1968 Le o1047297oliti di Roccaprebalza Costa della Guardia e Groppo di Gorro(Appennino Parmense) Ateneo Parmense Acta Naturalia 4 107ndash142
Giuseppetti G1953 Roccee minerali dellaformazioneo1047297olitica di Volpedo Rendicontidella Societagrave Mineralogica Italiana 9 85ndash134
Gran G 1952 Determination of equivalence point in the potentiometric titrationAnalyst 77 661ndash671
Gunnarsson I Arnorsson S Jakobsson S 2005 Precipitation of poorly crystalline antigoriteunder hydrothermal conditions Geochimica et Cosmochimica Acta 69 2813ndash2828
Guumlven N 1988 Smetites In Bailey SW (Ed) Reviews in Mineralogy 19 Hydrousphyllosilicates (exclusive of micas) Mineralogical Society of America MichiganBook Crafters Inc 497ndash559
Hansel CM Benner SG Fendorf S 2005 Competing Fe(II)-induced mineralization
pathways of ferrihydrite Environmental Science and Technology 39 7147ndash7153HobbsMYReardonEJ 1999 Effect of pH on boronco-precipitation by calcite further
evidence for non-equilibrium partitioning of trace elements Geochimica etCosmochimica Acta 63 1013ndash1021
Jones BF Galan E 1988 Sepiolite and palygorskite In Bailey SW (Ed) HydrousPhyllosilicates (exclusive of Micas) Reviews in Mineralogy vol 19 MineralogicalSociety of America Michigan Book Crafters Inc 631ndash674
Jove Colon CF 2006 Technical work plan for Thermodynamic database for chemicalmodelling TWP-MGR-PA-000039 REV 01 Yucca Mountain Project Las VegasNevada (US)
Kohls DW Rodda Jr JL 1967 Iowaite a new hydrous magnesium hydroxidendashferricoxychloride from the Precambrian of Iowa American Mineralogist 52 1261 ndash1271
Kyser TK OHanley DS Wicks FJ 1999 The origin of 1047298uids associated withserpentinization processes evidence from stable-isotope compositions CanadianMineralogist 37 223ndash237
Lapham DM 1961 New data on deweylite American Mineralogist 46 168ndash188Lee BD SearsSKGraham RC Amrhein CValiH 2003 Secondarymineral genesis
from chlorite and serpentine in an ultrama1047297c soil toposequence Soil Science
Society of America Journal 671309ndash1317
90 T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
Longinelli A Selmo E 2003 Isotopic composition of precipitation in Italy a 1047297rstoverall map Journal of Hydrology 270 75ndash88
Majzlan J Navrotsky A Schwertmann U 2004 Thermodynamics of iron oxides PartIIIEnthalpiesof formation and stability of ferrihydrite (simFe(OH)3) schwertmannite(simFeO(OH)34(SO4)18) and ɛ-Fe2O3 Geochimica et Cosmochimica Acta 681049ndash1059
Mariner P 2003 In-Drift PrecipitatesSalts Model ANL-EBS-MD-000045 REV 01 YuccaMountain Project Las Vegas Nevada (US)
Marini L Ottonello G 2002 Atlante degli acquiferi della Liguria Volume III Le acquedeicomplessio1047297olitici (bacini ArrestraBranega Cassinelle Cerusa ErroGorzenteLeira Lemme Lerone Orba Piota Polcevera Rumaro Sansobbia Stura Teiro
Varenna Visone) Pacini PisaMellini L Zanazzi PF 1987 Crystal structures of lizardite-l T and lizardite-2H1 fromColi Italy American Mineralogist 72 943ndash948
Mezzadri G 1964 Petrogra1047297a delle arenarie di Ostia Rendiconti Societagrave MineralogicaItaliana 20 192ndash228
Mondesir H Decarreau A 1987 Synthesegrave entre 25 et 200 degC de lizardites Ni ndashMgMesure des coef 1047297cients de partage solide-solution aqueuse pour le couple Ni ndashMgdans des lizardites Bulletin de Mineralogie 110 409ndash426
Neal C Shand P 2002 Spring and surface water quality of the Cyprus ophiolitesHydrology and Earth System Sciences 6 797ndash817
Neal C Stanger G 1983 Hydrogen generation from mantle source rocks in OmanEarth and Planetary Science Letters 66 315ndash320
Neal C Stanger G 1984 Calcium and magnesium hydroxide precipitation fromalkaline groundwaters in Oman and their signi1047297cance to the process of serpentinization Mineralogical Magazine 48 237ndash241
Neal C Stanger G 1985 Past and present serpentinisation of ultrama1047297c rocks anexample from Semail ophiolite nappe of Northern Oman In Drever JI (Ed) TheChemistry of Weathering D Reidel Publishing Company pp 249ndash275
Nesbitt HW Bricker OP 1978 Low temperature alteration processes affecting
et Cosmochimica Acta 41 1835ndash1841OrsquoHanley DS 1996 Serpentinites records of tectonic and petrological history Oxford
University PressOrberger B Friedrich G Woermann E 1990 The distribution of halogens and carbon
in Pge-bearing ultrama1047297cs of the Acoje Ophiolite Block Zambales Philippines Journal of Geochemical Exploration 37 147ndash169
Ottonello G Joron JL Piccardo GB 1984 Rare Earth and 3d transition elementsgeochemistry of peridotitic rocks II Ligurian peridotites and associated basalts Journal of Petrology 25 373ndash393
Palandri JL Kharaka YK 2004 A compilation of rate parameters of water-mineralinteraction kinetics for application to geochemical modeling Open File Report2004-1068 US Geological Survey
Panettiere P Cortecci G Dinelli E Bencini A Guidi M 2000 Chemistry and sulfurisotopic composition of precipitation at Bologna Italy Applied Geochemistry 151455ndash1467
Parkhurst DL Appelo CAJ 1999 Userrsquos guide to PHREEQC (version 2) mdash a computerprogram for speciation batch-reaction one-dimensional transport and inversegeochemical calculation US Department of the Interior US Geological SurveyWashington DC
Pfeifer HR 1977 A model for 1047298uids in metamorphosed ultrama1047297c rocks Observationsat surface and subsurface conditions (high pH spring waters) SchweizerischeMineralogische und Petrographische Mitteilungen 57 361ndash396
PlummerLN1984Geochemical modeling A comparisonof forwardand inverse methodsIn Hitchon B Wallick EI (Eds) Proceedings First CanadianAmerican Conference onHydrogeology-Practical Applications of Ground Water Geochemistry Banff AlbertaCanada mdash Worthington Ohio National Water Well Association pp 149ndash177
Rampone E Hofmann AW Piccardo GB Vannucci R Bottazzi P Ottolini L 1995Petrology mineral and isotope geochemistry of the External Ligurides peridotites(Northern Apennines Italy) Journal of Petrology 36 81ndash105
Rampone E Hofmann AW Piccardo GB Vannucci R Bottazzi P Ottolini L 1996Trace element and isotope geochemistry of depleted peridotites from an N-MORBtype ophiolite (Internal Liguride N Italy) Contributions to Mineralogy andPetrology 123 61ndash76
Rhoades JD Ingvalson RD HatcherJT1970Adsorption of boronby ferromagnesianminerals and magnesium hydroxide Soil Science Society of America Journal 34
938ndash941Saini-EidukatB KuchaH Keppler H1994Hibbingiteγ-Fe2(OH)3Clanewmineralfromthe Duluth Complex Minnesota with implications for the oxidation of Fe-bearingcompounds and the transport of metals American Mineralogist 79 555ndash561
Sanford RF 1981 Mineralogical and chemical effects of hydration reactions andapplications to serpentinization American Mineralogist 66 290ndash297
Scambelluri M Muumlntener O Ottolini L Pettke TT Vannucci R 2004 The fate of BCl and Li in the subducted oceanic mantle and in the antigorite breakdown 1047298uidsEarth and Planetary Science Letters 222 217ndash234
Scambelluri M Rampone E 1999 Mg-metasomatism of oceanic gabbros and itscontrol on Ti-clinohumite formation during eclogitization Contributions toMineralogy and Petrology 135 1ndash17
Scambelluri M Bottazzi P Trommsdorff V Vannucci R Hermann J Gogravemez-Pugnaire MT Logravepez-Sagravenchez Vizcaino V 2001a Incompatible element-rich1047298uidsreleased by antigorite breakdown in deeply subducted mantle Earth and PlanetaryScience Letters 192 457ndash470
Scambelluri M Rampone E Piccardo G 2001b Fluid and element cycling insubducted serpentinite a trace-element study of the ErrondashTobbio high-pressureultrama1047297tes (Western Alps NW Italy) Journal of Petrology 42 55ndash67
Segadelli S2006 Lageologia nelpaesaggio le rupi o1047297olitiche in Val Taro e Val Ceno IlGeologo dellrsquoEmilia-Romagna 22 15ndash29 Nuova Serie
Settembre Blundo D Martin De Vidales JL Alietti A 1992 Corrensite-type clayminerals from Taro Valley (North Italy) Mineralogica et Petrographica Acta 35215ndash222
Sharp ZD Barnes JD 2004 Water-soluble chlorides in massive sea1047298oor serpenti-nites a source of chloride in subduction zones Earth and Planetary Science Letters226 243ndash254
Siever R Woodford N 1979 Dissolution kinetics and the weathering of ma1047297cminerals Geochimica et Cosmochimica Acta 43 717ndash724Singer A 1989 Palygorskite and sepiolite group minerals In Dixon JB Weed SB
(Eds) Minerals in Soil Environments Soil Science Society of America MadisonWisconsin pp 829ndash872
Siffert B Wey R 1962 Synthese d rsquoune sepiolite a temperature ordinarie Comptesrendus hebdomadaires des seacuteances de lAcadeacutemie des Sciences de Paris 2541460ndash1463
Środoń J 1999 Nature of mixed-layer clays and mechanisms of their formation andalteration Annual Reviews of Earth and Planetary Science 27 19ndash53
Stoessell RK 1988 25 degC and 1 atm dissolution experiments of sepiolite and keroliteGeochimica et Cosmochimica Acta 52 365ndash374
Stoessell RK Hay RL 1978 The geochemical origin of Sepiolite and keorilite atAmboseli Kenya Contributions to Mineralogy and Petrology 65 255ndash267
Suarez DL Šimůnek J 1996 Solute transport modeling under variably saturatedwater 1047298ow conditions In Lichtner PC Steefel CI Oelkers EH (Eds) ReactiveTransport in Porous Media Reviews in Mineralogy vol 34 Mineralogical Society of America Michigan Book Crafters Inc 229ndash268
Sumner DY Grotzinger JP 1996 Were kinetics of Archean calcium carbonate
precipitation related to oxygen concentration Geology 24 119ndash122Su C Suarez DL 2004 Boron release from weathering of illites serpentine shales
and illiticpalygorskitic soils Soil Science Society of America Journal 68 96 ndash105Toscani L Venturelli G Boschetti T 2001 Sul1047297de-free and sul1047297de-bearing waters in
the Northern Apennines Italy general features and waterndashrock interaction AquaticGeochemistry 7 195ndash216
Treves BE Harper GD 1994 Exposure of serpentinites on the ocean 1047298oor sequenceof faulting and hydrofracturing in the Northern Apennine ophicalcites O1047297oliti 19b435ndash466
Varadachari C Kudrat M Gosh K 1994 Evaluation of standard free energies of formation of clay minerals by an improved regression method Clays and ClayMinerals 42 298ndash307
Veniale F 1966 Sepiolite in sedimenti dellrsquoAppennino Vogherese (Pavia) Periodico diMineralogia 35 343ndash386
Venturelli G 2003 Acque minerali ed ambiente mdash Fondamenti di geochimica deiprocessi di bassa temperatura Pitagora Editrice Bologna
Venturelli G Contini S Bonazzi A Mangia A 1997 Weathering of ultrama1047297c rocksand element mobility at Mt Prinzera Northern Apennines Italy MineralogicalMagazine 61 765ndash778
Venturelli G Frey M 1977 Anchizone metamorphism in sedimentary sequences of the Northern Apennines Rendiconti Societagrave Mineralogica Italiana 33 109ndash123
Vieillard P 2002 A new method for the prediction of Gibbs free energies of formationofphyllosilicates (10 Aring and14 Aring)basedon theelectronegativityscale Claysand ClayMinerals 50 352ndash363
Vissers RLM Strating EHH 2001 Structures and microstructures in a thrust-relatedgreenschist facies tectonic melange Voltri Group (NW Italy) O1047297oliti 26 33ndash46
Viti C Mellini M 1997 Contrasting chemical compositions in associated lizardite andchrysotile in veins from Elba Italy European Journal of Mineralogy 9 585 ndash596
Viti C Mellini M 1998 Mesh textures and bastites in the Elba retrogradeserpentinites European Journal of Mineralogy 10 1341ndash1359
Wenner DB Taylor Jr HP 1974 Deuteriumatomic hydrogen and oxygen-18oxygen-16 studies of serpentinization of ultrama1047297c rocks Geochimica et CosmochimicaActa 38 1255ndash1286
Williams LA Parks GA Crerar DA 1985 Silica diagenesis I Solubility controls Journal of S edimentary Petrology 55 301ndash311
Wilson J Savage D Cuadros J Shibata M RagnarsdottirKV 2006 The effect of ironon montmorillonite stability (I) Background and thermodynamic considerations
Geochimica et Cosmochimica Acta 70 306ndash322WollastR MackenzieFTBricker OP1968 Experimental precipitation and genesis of sepiolite at earth-surface conditions American Mineralogist 531 1645ndash1662
Wrucke CT1996 Serpentine and carbonated-hosted asbestos deposits InEdwardAdu Bray (Eds) Preliminary compilation of descriptive geoenvironmental mineraldeposit models 95ndash831 US Geological Survey Open-File Report pp 39 ndash46
Xu T Apps JA Pruess K 2001 Analysis of mineral trapping for CO2 disposal in deepaquifers Lawrence Berkeley National Laboratory University of California
ZhangYDaweRA2000 In1047298uence ofMg2+ on thekineticsof calcite precipitation andcalcite crystal morphology Chemical Geology 163 129ndash138
Zhu C Anderson G 2002 Environmental applications of geochemical modelingCambridge University Press
91T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
supergenic mineralogy are the dissolution-precipitation kinetics(Nesbitt and Bricker 1978 Siever and Woodford 1979) and climate(Wrucke 1996)
In this study the chemical and isotopic composition of severalspring waters issuing from different outcrops of ophiolitic rocks in theParma Province (Toscani et al 2001 Boschetti 2003) are presentedand modeled for 1047298uids circulating in and interacting with ultrama1047297crocks of the Northern Apennine The model shows the role of primary
and secondary minerals in the reaction path Solubility productsspeci1047297c for local phases were calculated and inserted in the modelingand the role of low-temperature serpentine in the chemical evolutionof groundwater from the TarondashCeno Valleys and the Voltri Group area(Bruni et al 2002 Marini and Ottonello 2002) is discussed
2 Geological and hydrogeological setting
The geology and petrology of Northern Apennine ophiolites(ultrama1047297tes basalts and cherts) have been extensively investigated(Abbate et al 1980 Beccaluva et al 1984 Ottonello et al 1984Rampone et al1995 1996) Most researchersagreethat the ophiolites(i) were generated in the PiedmontndashLigurian ocean during the Jurassicand Cretaceous periods (ii) are remnants of the ocean crust and
overlying sediments (iii) were fragmented into several nappes duringthe Upper CretaceousndashEocene stage of the orogeny and (iv) re1047298ect thefact that the entire allochthonous pile overrode the continental crustduring the Oligocene and Miocene periods
The Northern Apennine ophiolites crop out in two palaeogeo-graphic domains the Internal Ligurides (IL) and External Ligurides(EL) which are identi1047297ed on the basis of their current structuralcharacteristics and their relationship with the associated sedimen-
tary sequences (Abbate et al 1980 and references therein) IL contains strongly depleted peridotites serpentinites gabbros andbasalts (N-MORB) which are the basis of the Upper Jurassicndash
Paleocene sedimentary sequence ie deposits of Mn ore associatedwith radiolarian cherts silicic limestones Palombini shales withlimestones EL on the other hand contains less depleted lherzolitesin association with E-MORB (Ottonello et al 1984 Beccaluva et al1984 Rampone et al 1995 1996) In the TarondashCeno river valleyssprings issuing from ophiolite rocks occur in three areas (Fig 1) (A)Mt PennandashMt Aiona and Mt MaggiorascandashMt Nero where the Taroriver and the Ceno stream have their sources (B) the areas aroundthe Belforte Val Manubiola and Valmozzola along the middlecourse of the river and (C) Mt Prinzera where the Taro River entersthe Po Plain
Fig1 Simpli1047297ed geologic map of the Northern Apennine (modi1047297ed from Beccaluva et al 1984) The samples of this study are located in the TarondashCeno Valleys (sketched outline)1=Messinian to present 2 = parautochthonous and Piemontese Tertiary 3 = Tuscan ndashUmbrian sequences 4 = ophiolites and sedimentary sequences of the Liguride units 5 = high-pressure metamorphic ophiolites and calcschists of the Alpine Units 6 = Ligurides ophiolitic masses (EL = External Liguride unit IL = Internal Liguride unit) 7 = Middle Taro ldquoarea Brdquo
sampled BelfortendashVal ManubiolandashValmozzola White triangles mountains cited in the text with altitude in brackets (meters asl)
77T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
Basaltic rocks which are absent only from the area of Mt Prinzerasuffered an early low-grade oceanic metamorphism to varyingextents followed by an orogenic metamorphism (Cortesogno 1980)Albite epidote prehnite sericite chlorite actinolite titanite andhaematite association replaced the original mineralogy of plagioclaseolivine clinopyroxene ilmenite and magnetite (Cortesogno andLucchetti 1982) Spinel lherzolites re-equilibrated in the petrological1047297eld of plagioclase lherzolites and were exposed to oceanic and
orogenic metamorphisms Hydration of the early mineralogy pro-duced serpentine (lizardite and minor chrysotile) chlorite and talcand bastite after olivine and orthopyroxene respectively
The ultrama1047297tes of the Middle Taro valley and of Mt Prinzera arestrongly serpentinized (Venturelli et al 1997) whereas primaryclinopyroxene and plagioclase are largely preserved at Mt Aiona(Beccaluva et al 1973) Sedimentary rocks at the contacts with theophiolites are predominantly 1047298yschoid (marly to pelitic) and arenac-eous Typically 1047298ysch rocks are micritic marly turbidites withinterbedded pelites at MtCaio the latter are composed of illite andor illitesmectite mixed layers and chlorite (Venturelli and Frey1977)Arenaceous formations include (i) the Casanova Complex composedof quartz-feldspathic rocks with metamorphic serpentinitic andvolcanic clasts (Di Giulio and Geddo 1990) and (ii) the OstiaSandstones composed of quartz feldspar (plagioclase orthoclasemicrocline) mica (muscovite biotite) calcite and dolomite within aclayey and chlorite matrix at the base of the Mt Caio Flysch Formation(Mezzadri 1964)
Hydrogeologic studies on the TarondashCeno Valleys were focussedon their alluvial fan and plain probably because they cover adensely populated area Recently research projects consider theupstream section of the valleys and preliminary results show that77 of the springs are fed by aquifers contained in Late CretaceousLigurid 1047298ysches 11 in sandstones-pelite conglomerates andmassive sanstones (eg Ranzano Formation) and only 4 inophiolitic bodies (Segadelli et al 2006 De Nardo et al 2007)According to the authors the last value is probably underestimatedbecause the study consider only the known and gathered springs forwaterwork system whereas most of springs issuing from ophiolites
are free and unknown despite their perennial activity that ensurepreservation of the stream out1047298ow during dry periods De Nardoet al (2007) identify two types of ophiolite related springs (i)
springs with inde1047297nite permeability limit and (ii) with de1047297nitepermeability limit which are fed by aquifers contained in fractu-rated rocks and in coarse detritic covers respectively Some high1047298ow rate springs have different origin being related to gravitationalsliding (eg Mt Nero area) that increased permeability due toslackening of the pre-existing fractures (De Nardo et al 2007)
3 Analytical methods
31 Field methods
All water samples were 1047297ltered through cellulose acetate 1047297lters(045 μ m) An aliquot for cation analysis was acidi1047297ed with 65 HNO3
Suprapur Merck (1 cm3 HNO3 per 100 cm3 water) Labile parameters(temperature pH and Eh) were determined in the 1047297eld using anORION 250A instrument equipped with a Ross glass electrode formeasuring pH and a combined electrode of platinum-silversilverchloride for measuring Eh The electrode for Eh was calibrated atdifferent temperatures using ZoBells solution (Nordstrom 1977)Speci1047297c conductance at 20 degC was measured using a conductimeter(Model 85 Yellow Springs Instruments) Total alkalinity was deter-mined in the 1047297eld by acidimetric titration with 001 N HCl usingbromocresol green as an indicator and in the laboratory by Grantitration (Gran 1952) within 12 h of sampling Reduced dissolvedspecies were determined by spectrophotometry using a portablephotometer (Merck SQ300) with Spectroquant kits total ammoniumNH4= NH4
++ NH30 and total sulphide HS=H2S0+HSminus+ S2minus were deter-
mined by adding indophenol blue in basic solution (pH asymp13 method4500-NH3 F in APHA-AWWA-WEF1995) and methylene blue in acidicsolution (pHasymp05 method 4500-S2minus F in APHA-AWWA-WEF 1995)respectively
32 Laboratory methods
Clminus NO3minus and SO4
2minus were analyzed by ion chromatography usingDionex DX100 with anionic self-regenerating suppressor silica byspectrophotometry using a Merck SQ300 photometer and Na K Ca
Mg B Fe by radialsequential inductively coupled plasma opticalemission spectrometry (ICP-OES) using Philips PU7450 and Horiba
Jobin-Yvon Ultima 2 instruments Data on δ2H(H2O) and δ
18O(H2O)
Table 1
Summary of physico-chemical parameters and concentrations of chemical constituents in the spring waters coming from ultrama 1047297tes and basalts in the TarondashCeno Valleys
Type (lithology) Springs (ultrama1047297tes) Springs (basalts)
were collected using an automatic equilibration device (Finnigan GLF1086) in-line with a mass spectrometer (Finnigan Delta Plus) Theanalyzed H2 and CO2 gases were equilibrated with water samples at atemperature of 18 degC (Epstein and Mayeda 1953) using a Pt-richcatalyst Isotopic ratio are reported in per mill relative to V-SMOWand standard errors for δ
2H(H2O) and δ18O(H2O) are approximately plusmn
10permil and within plusmn01permil respectively
4 Chemical and isotopic results
The chemical compositionsof thewaters from theultrama1047297tesandbasalts are summarized in Table 1 Complete chemical and isotopedata are reported in Appendix A where all the water samples aresubdivided into four classes on the basis of the lithology present at thesource (i) ultrama1047297c (ii) basaltic (iii) sedimentary and (iv) surfacewaters
41 Chemical classi 1047297cation
All the samples investigated were freshwater (TDSb1 gL) Theternary classi1047297cation diagrams (Na+ K)ndashCandashMg and tAlkndashSO4ndashCl(Fig 2) show that waters of groups (ii) (iii) and (iv) do not differsigni1047297cantly many are bicarbonate-calcic with neutral to slightlybasic pH values (667ndash857) highly oxidized (Eh=371ndash514 mV) andcontain a small amount of chloride (Clminus up to 64 mgL) Sample UM1
from basalt is an exception since it is a ClndashCandashNa type with EC up to1043 μ S cmminus1 Clminus up to 231 mgL hardness up to 367 mgL CaCO3Waters issuing from ultrama1047297tes have more varied compositions Theyrange from Ca-bicarbonate with a pH of 73ndash88 and a hardness rangingfrommoderately hard to very hard (67ndash386mgL CaCO3 Eh129ndash581μ Scmminus1) to Mg-bicarbonate with a pH of 74ndash88 and a hardness from softto very hard (59ndash206 mgL CaCO3 92ndash330 μ S cmminus1) Among the Mg-bicarbonate waters sample UM8 shows the highest hardness values
Fig 2 MgndashSindashC ternary diagram (in molkg) and ternary classi1047297cation diagrams [Mgndash(Na+K)ndashCa plus ClndashtAlkndashSO4 in meqL] of alkaline springs in the TarondashCeno Valleys (this work)andthe VoltriGroup(Mariniand Ottonello 2002) Bottled watersfromTarovalleyare also representedRG1 andRG2Rocca Galgana 1990 and1994 respectively(chemical data fromwater bottle label) FN1 FN2 and FN3 Fontenova 1974 1990 and 1994 respectively (chemical data taken fromwater bottle label) MgndashSindashC contentsafter mineral dissolution (whitecrosses) are also represented (see text for details) Note that for dolomite and magnesite the CMg mole ratio refers to the products of the respective reactions CaMg(CO3)2+2H2O+
2CO2= Ca2+
+ Mg2+
+4HCO3
minus
and MgCO3+ H2O+CO2= Mg2+
+2HCO3
minus
79T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
(486 mgL CaCO3) conductivity (877 μ S cmminus1) and alkalinity (618 mgL HCO3) Springs PR10 and UM15 from ultrama1047297tes show an extremecomposition being sodium-hydroxide and sodium-sulfate respectivelytheir pH is high (101ndash112) their Eh is low (down to minus103 mV) theyaresoft (20ndash61 mgL CaCO3 220 to 408 μ S cmminus1) and they contain largeamounts of boron (up to 2294 μ gL and up to 13461 μ gL respectively)and moderate amounts of reduced sulfur and nitrogen species In bothPR10 and UM15 non-carbonate alkalinity predominates over carbonate
alkalinity using PHREEQCI software with the llnldat thermodynamicdatabase (Parkhurst and Appelo 1999) to calculate the composition of alkaline species under thepH Ehand T valuesmeasuredin1047297eld samplePR10 was found to contain 54 hydroxide 23 carbonate 19 borateand 4 other alkalinespecies (Si sulphide and ammonia) while sampleUM15 contained 75borate14 hydroxide10carbonate and 1 otherspecies
The SindashMgndashC diagram (Fig 2) is a useful tool for studying waterndashrockinteractions Dissolution reactions of typical magnesium minerals of ultrama1047297c rocks are represented by a line with CMg=2 (molar ratio)
Fig3 Plot of (A) Naand (B) Mgvs Cl forallwater samples from the TarondashCenoValleys including bottledwaters Rocca Galgana andFontenova(chemical data taken from labels)Alkalinesprings from the Voltri Group (smallblack diamonds Marini and Ottonello 2002) and meteoric water mean composition (MW Panettiere et al 2000 Venturelli 2003) are also plottedDottedcurvesin(A)and(B)refertomixingwith04and2wtNaCl 1047298uidinclusionsand MgCl2dissolution (1 mol)respectively Greycrosses andi csescorrespondrespectively to mixingwith04and2wtNaCl 1047298uidinclusionsfor mixing fractionsranging from 0001 to0005The solid linerefers tomixingof meteoric water(A) orMg-bicarbonate water(B) withseawater
The vertical grey line represents the water-soluble Cl content in the serpentinites from Northern Apennine (Barnes et al 2006) Symbols are as in Fig 2
80 T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
The Mg-bicarbonate watersissuingfrom ultrama1047297crocksfallnearthisline and cluster between the clinochlore and forsterite points (Fig 2)Stream waters and springs issuing from sedimentary and basaltic rocksfall between thecarbon apex and theline indicated by a CMg molar ratioof 4 which describes the dissolution of CandashMg-clinopyroxene
This is consistent with the (Ca)-bicarbonate compositions of thesestream waters and springs It is noteworthy that the Ceno streamwaters shifted toward the C apex of Fig 2 derive their chemistry fromthe interaction with Mt Caio Flysch cropping out widely in the Cenocatchment (samples from UM50 to UM59 see Map1 sheet in theannexes) whereas the Taro source spring waters plot towards theother apexes (lower C) typical of ophiolite minerals (samples fromUM41 to UM49 see Map1 sheet in the annexes) The use of the traceelements (eg Sr and Rb content Boschetti 2003) further highlightthe differences between the stream waters of the two rivers and therole of lithology Nonetheless mixing and equilibrium with atmo-sphere make the reaction-path modeling of these samples poorlyreliable so their are unsuitable for the purpose of the present work
Alkaline waters from ultrama1047297c rocks fall inside the same sub-triangle mostly along the CndashSi side probably due to the precipitationof magnesium minerals during the evolution of the water composition(Bruni et al 2002)
42 Sodiumndashchloride relationships
Chlorine and other halogens in serpentinized rocks are inheritedfrom seawater and marine sediments during oceanic metamorphism(eg Orberger et al 1990) They are transferred into serpentinite byadsorption (Anselmi et al 2000 Sharp and Barnes 2004) entrap-ment in 1047298uid inclusions (Scambelluri et al 2001ab) and substitutionin the structure of low-crystalline iron-hydroxide phases These latterare composed of the isomorphous and chemically similar iowaitehibbingite and FendashCl-oxydroxides (also named green rusts orfougerites Kohls and Rodda 1967 Saini-Eidukat et al 1994 Federet al 2005 and references therein) In the Northern Apennineserpentinite rocks the mean Cl content is sim300 ppm mean values of 221 and 375 ppm were measured respectively in the Erro ndashTobbio(Ligurian Western Alps Scambelluri et al 2004) and in the centralItaly serpentinites (Anselmi et al 2000) but thewater-soluble Cl from
serpentinites is estimated to be than 100 ppm (Barnes et al 2006) Asshown in Fig 2 waters from basalts and ultrama1047297tes of the TarondashCenoValleys show a trend towards the Cl apex whereas in Fig 3 theestimated soluble chlorine in serpentinites agrees with the highestvalue reported in spring waters from ultrama1047297tes (samples fromMarini and Ottonello analyzed in 2002) The Cl content of samplesUM1 and UM36is veryhigh(Fig 3) and is probablydue to pollutionInfact these springs are not far from the A15 highway (UM1) and theTomarlo Pass road (UM36) where salt mixtures comprising NaClMgCl2middot6H2O and probably CaCl2 are widely used for deicing in winterAlkaline water with a Mg-bicarbonate composition from ultrama1047297tesshow the highest and lowest Cl contents respectively Na-rich Mg-bicarbonate waters from basalts (UM6 UM7 Fig 3) and ultrama1047297tescan be explained by the dissolution of albite in basalts (Beccaluva
et al 1975) and in spilitized EL peridotites (Beccaluva et al 1984
Rampone et al 1995) respectively whereas the highest content of both Na and Cl in the alkaline water samples UM15 and PR10 (Fig 3)can be explained by leaching of NaCl 1047298uid inclusions in theserpentinite rocks (04ndash20 wt NaCl Scambelluri et al 2004)
43 Boron and BCl ratio
Compared with unaltered ultrama1047297c rocks serpentinites have
higher content of both boron and chlorine and most authors suggestthat both elements come from seawater (eg Sanford 1981 andreferences therein Bonatti et al 1984) Despite several studies onthese rocks the behavior of Cl and B in solution during waterndashrockinteraction processes is not well known Early investigations on 1047298uidsassociated with serpentinites in the western United States (Barnes andONeil 1969 Barnes et al 1972) revealed metamorphic waters withhighboron levels(up to 28mM) and BCl molar ratiosof 0038ndash013 aswell as Mg-bicarbonate and high-pH waters of meteoric origin withvariable BCl ratios of 0027ndash042 (median 0039) and 00006ndash0017(median 00076) respectively Bicarbonate waters issuing fromophiolitic rocks in Cyprus have BCl ratio in this range (00017ndash
0058 median 00019 Neal and Shand 2002) whereas low boron
Fig 4 Plots of (A) δ18O vs δ2H and (B) δ18O vs altitude for waters from the Taro ndashCenoValleys C-Italy and N-Italy are the meteoric water lines for Central and Northern Italy
respectively (Longinelli and Selmo 2003) Symbols are as in Fig 2
81T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
levels were found in the alkaline springs of meteoric origin in OmanNew Caledonia and ex-Yugoslavia (Barnes et al 1978)
Among the waters analyzed so far in the TarondashCeno Valleys thebicarbonate waters have BCl ratios in the range 00017ndash022 (median0073) whereas the alkaline waters show the highest ratios everreported (BCl= 034ndash33 median 169) The high boron content of upto 13 mgL of the alkaline springs could be related to metamorphic1047298uids released during serpentinite subduction in fact rocks prefer-
entially lose Cl and the BCl ratio of the released 1047298
uids increases withpressure and temperature (Scambelluri et al 2004) Nonetheless thealkaline water samples in our study differ signi1047297cantly from alkalinemetamorphic1047298uids in that they have low salinity high BCl ratios andmeteoric isotope signatures (see below) all these characteristicssuggest an origin in the interaction of meteoric water with high-pressure metamorphic serpentinite This hypothesis is furthersupported by the agreement in BCl ratio between the alkalinesamples we study here and the high BCl ratio of up to 4 previouslyreported for the residual high-pressure mineralogy of ultrama1047297tes(olivine+orthopyroxene Scambelluri et al 2004) In contrast theVoltri Groups Ca-hydroxide waters have mean values of B = 014 mgL and BCl=003 (Marini and Ottonello 2002) consistent with acontribution of boron and chlorine from 1047298uid inclusions In factassuming fractional contributions of 0001ndash0005 from NaCl 1047298uidinclusions with boron content from 4 to 8 mgL ( Fig 3a Scambelluriet al 2004) results in a B content of 0004ndash01 mgL in the solutionThese results suggest that the difference between springs from theVoltri Group and TarondashCeno Valleys is probably related to (i) differentevolution of the 1047298uids and (ii) the different types of host rocks and Pndash
T paths of the Ligurian Western Alps and EL peridotites (eg Beccaluvaet al 1984 Scambelluri et al 2004)
44 Isotopic composition of water
A diagram of δ2H vs δ18O (Fig 4A) shows that the waters sampledfrom the TarondashCeno Valleys are meteoric in origin since they areconsistent with the meteoric water lines of central and northern Italy(Longinelli and Selmo 2003) The altitude vs δ18O(H2O) (Fig 4B)
summarizes the isotopic effects due to temperature altitude andamount of rainfall (eg Criss1999) all of which help to determine theisotopic composition of the water samples The tendency is similar forbicarbonate waters from sedimentary basaltic and ultrama1047297c forma-
tions which may provide insight into the recharge areas alkalinewaters in contrast probably circulate deeper in the ground
5 Water ndashrock interaction model
51 Fundamentals
The reaction path model between meteoric water and serpentinite
was performed using PHREEQCI 2125 code (Parkhurst and Appelo1999) and the Lawrence Livermore National Laboratory thermody-namic database (llnldat as known as thermocomV8R6 ) implementedwith the equilibrium constants listed in Table 2
Computations wereperformed by reaction progress mode ( forward
modeling Plummer 1984 Zhu and Anderson 2002) ie (i) addingreacting rock at each step in the simulation (ii) activating the phase(s)precipitation only when the thermodynamic stability of the mineralparagenesis is well known and consistent with the mineralogy of thearea and (iii) plotting reaction progress on speci1047297c activity diagramsThe same thermodynamic database was used with The GeochemistsWorkbench 403 software (Bethke 2002) to generate Pourbaix andactivity diagrams (Figs 5 6 8 and 9) Modeling was performed usingan approach based on equilibrium rather than reaction transportbecause the data on reaction kinetics and reactive surfaces in thesesolid phases systems are scarce particularly with respect to theactivation energy of clay and mixed-layer minerals at low temperatureand high pH Moreover the hydrogeology and the discontinuity of theaquifers in the study area which are related to the geologicheterogeneity and structural complexity of the Northern Apenninewould make a model based on reaction transport even less credible atthis stage
The modeling in this study was performed on the basis of thefollowing knowledge (1) primary and secondary mineral assem-blages (2) chemical composition of the minerals involved (3) physicalparameters and chemical composition of the starting 1047298uid (4)thermodynamic data on pure phases
52 The solid reactant primary paragenesis
EL peridotites of the TarondashCeno Valleys (NorthernApennines Italy)are spinel lherzolites containing olivine (forsterite) orthopyroxene(enstatite) and clinopyroxene (diopside) (Giammetti 1968 and
Table 2
Thermodynamic data on hydrolysis reactions added to the llnldat database of the PHREEQCI code (Parkhurst and Appelo 1999) and thermocomV8R6dat database of TheGeochemists Workbench software (Bethke 2002)
Mineral Reaction log K (TdegC) ΔGdeg f References
0 25 60 100 (kJ mol-1)
clinochlore08-daphnite02 Mg4Fe1Al2Si3O10(OH)8 + 16H+=2Al3+ + Fe2++12H2O+4Mg2++3SiO2 7318 6384 5272 4269 minus786379 this workferr ihydrite 2-l ine Fe(OH)3+3H+= Fe3++ 3H2O 484 355 ndash ndash minus70850 Majzlan et al 2004ferr ihydrite 6-l ine Fe(OH)3+3H+= Fe3++ 3H2O 435 312 ndash ndash minus71100 Majzlan et al 2004liz09-berth01 Mg27Fe02Al02Si19O5(OH)4+64H+=02Al3++02Fe2++52H2O+27Mg2++19SiO2 3111 2786 2383 2005 minus399854 this workp-antigorite Mg
Saponite-Fe2+ Fe3175Si365Al035O10(OH)2+74H+=035Al3++3175Fe2++47H2O+365SiO2 1845 1635 1319 1001 minus452459 Wilson et al 2006Saponite-Fe2+ndashMg Fe0165Mg3Si367Al033O10(OH)2+732H+=033Al3++0165Fe2++466H2O+3Mg2++
367SiO2
3037 2729 2330 1921 ndash Marini and Ottonello(2002)
The Gibbs free energy of formation (ΔGo f ) of the two theoretical Fe3+ (Varadachari et al1994) and Fe2+ vermiculites were calculated using a previously published method (Vieillard
2002 and Vieillard pers comm)log K values extimated considering an ideal solid solution between the end-members and using data thermodynamic data from llnldat (clinochlore08-daphnite02) and Wilson
et al (2006) (liz09-berth01)
82 T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
references therein Venturelli et al 1997) Accessory phases areplagioclase picotite plusmn calciumndashsodic amphibole (Venturelli et al1997) Before modeling the crystal chemical formula of the mineralphases was calculated from data from Venturelli et al (1997) thendepending on the modal composition of the rock (Ernst and Piccardo1979) and the composition of the ideal solid solution of the minerals
the resulting stoichiometry of the solid reactant(s) was calculated andinserted in the input 1047297le of the PHREEQCI software (Table 3)Thermodynamic data of the primary mineral paragenesis are all wellknown and are included in the llnldat thermodynamic database of PHREEQCI
53 The solid reactant secondary paragenesis
Secondary minerals in the ultrama1047297tes of the TarondashCeno Valleysare serpentine (lizardite and rare chrysotile) smectite minerals (Fendash
Mg saponites) and Fe-oxi-hydroxides Other accessory minerals arechlorite mixed layer saponitendashchlorite (low-charge corrensite) talccalcite and dolomite (Beccaluva et al 1973 Dinelli et al 1997Venturelli et al 1997 Brigatti et al 1999 2000) In addition mixed
layer saponitendashtalc (aliettite) and vermiculitendashchlorite (high-charge
corrensite) are present (Alietti 1956 Settembre Blundo et al 1992Brigatti and Valdregrave 1996) Except for serpentine themodal abundanceof these phases is unknown In addition some of these minerals ndash
particularly interlayer clays ndash may have different origins (hydrothermalor weathering allogenicor authigenic)and different stabilitiesrelatedtokinetics and drainage 1047298ow rate For example chlorite weathering yields
vermiculite-bearing or saponite-bearing interlayers under good- andpoor-drainage regimes respectively (Bonifacio et al 1997 Środoń1999) In addition the thermodynamic data available for serpentineinclude information only on chrysotile which is less abundant thanlizardite in the serpentinites of the areas that we studied According toVenturelli et al (1997) the lizardite composition in the serpentinites of the Taro valley can be summarized as
which is similar to 1-T lizardite from the Monte Fico serpentinizedperidotite (Elba Island Northern Tyrrhenian Sea Italy Viti and Mellini1997)
Using the thermodynamic data of Wilson et al (2006) and
assuming an ideal solid solution between lizardite and berthierine
Fig 5 EhndashpH iron diagram of the spring waters coming from ultrama1047297c rocks of the TarondashCeno Valleys Bicarbonate and high-pH waters of the Voltri Group are also represented(dotted area andsmallblackdiamonds respectively) Thelowest Eh value of thehigh-pH waterswas calculated using theS(+VI)S(minus II)redox couple(Marini and Ottonello 2002)Thestability 1047297eld of mineral phases are in grey and those of aqueous species in white In diagrams A and B 6-line ferrihydrite-vermiculite and goethite were selected as subaerial solidphases respectively
83T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
we calculated ΔGf 0=minus399854 kJmol for a theoretical (09)(lizardite)+
(01)(berthierine) mineral (Table 2) The log K value of the stated solidsolution was calculated at different temperatures using PHREEQCIafter loading the thermodynamic data on the two end-members intothe database The log K results obtained for Al-lizardite were added tothe thermodynamic database of the solid reactant (Table 3) andcompared with data from the literature (Table 4)
54 Composition of the starting solution O 2 partial pressure and
minerals included in the model
The weighted average chemical composition of rainwater inBologna (Panettiere et al 2000) which is 95 km east of theinvestigated area was used as the starting solution in the modelFull chemical data on rainwater of the TarondashCeno Valleys are
unavailable
Total alkalinity was recalculated for log PCO2=minus35 bar and
T =11 degC which was the mean temperature of the springs sampledand for pe (or Eh) based on chemical data for the redox couple N(minusIII)N(+V) The values of pH and pe obtained were 566 and 452(Eh=025 V) respectively the former agrees with the median valuereported for rainwater in thearea (Venturelli et al1997) Because dataon Al Fe and Si in rainwater were lacking we calculated theconcentrations of these elements assuming logPCO2=minus30 bar andslight supersaturation relative to ferrihydrite-6 line and kaolinite Inthis way a hypothetical composition of water in equilibrium with thelocal soil was calculated (Table3) and adopted as the ldquoinitial waterrdquo formodeling
Evidence suggests that secondary minerals particularly Fe-bearingphases change the chemical composition of the water 1047298owingthrough the ultrama1047297tes of the TarondashCeno Valleys According to Neal
and Stanger (1983) and Fritz et al (1992) the production of CH4 and
Fig 6 Activity plotfor the MgOndashSiO2ndashAl2O3ndashFe2O3(FeO)ndashH2O systemat differentlog PO2 values Spring waters issuing from TarondashCeno Valleys (symbols as in Fig 2) andVoltri Groupultrama1047297tes are shown (grey area bicarbonate waters small black diamonds alkaline waters from Marini and Ottonello 2002) In plots A B and C the stability 1047297eld limits of vermiculite(Table 2) correspond to25 degCwhilethe amorphoussilica (am-SiO2) brucite serpentine (Sr1ndash4 in Table 4)and sepiolite(Sp1ndash2 in Table 4) phasescorrespondto 11 degCThepath calculatedfor interaction between waterand ultrama1047297c rockis also shown (dotted line)the number-letter combinations referto the appearanceand disappearance of minerals(see Table 5)
84 T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
H2 in thealkalinewaters1047298owing from theultrama1047297tes isrelatedto theconsumption of oxygen during the oxidation of iron forming thehematitendashmagnetite pair Therefore the activity of iron and theoxidation state of the water must be de1047297ned The analysis of watersamples from ultrama1047297tes were plotted in both the Pourbaix andactivity diagrams (Figs 5 and 6) where the calculated activity of Fe2+Mg2+ SiO2 SO4
2minus (Pourbaix) and log PO2 (activity diagrams) wereinvolved Iron concentration was quite variable in the TarondashCenosprings where the average content was 27 μ gL and the contentreached a maximum of 200 μ gL and in the Voltri springs where theaverage was 44 μ gL and the maximum was 948 μ gL (Marini andOttonello 2002) In both cases the variations are probably related to
the different mineralogies of the soils and the different solubilities of the Fe-rich phases like goethite and ferrihydrite Iron-bearingsaponites and montmorillonites are the prevailing clay minerals inthe area investigated probably because of the ubiquitous presence of iron oxy-hydroxide (Alietti et al 1976 G Venturelli pers comm) Forthis reason the model considers the involvement of pure Mg-smectiteto be negligible
In the Pourbaix diagram (Fig 5) the data from bicarbonate watersfall into the ferrihydrite-goethitendashvermiculite1047297elds and the data fromalkaline waters fall on the boundaries of the vermiculitesaponite andsaponitechlorite Since ferrihydrite is the 1047297rst precipitating Fe-phaseit was used rather than goethite in the modeling The diagram alsosuggests the transformation of ferrihydrite into magnetite during thechange from bicarbonate to alkaline composition but this process
requires a large amount of Fe2+ in solution (N2 mM Hansel et al2005) which is unlikely at this stage of the waterndashrock interactionThus the magnetite participating in the equilibrium is most likelyprimary in origin
The modeling and the activity diagrams (Figs 6 and 9) werecalculated using the average value PO2= 10minus26 or 10minus50 bar forbicarbonate and alkaline waters respectively These values arecompatible with those calculated by Marini and Ottonello (2002) forbicarbonate waters (PO2= 10minus40 bar) and for alkaline waters of theVoltri Group (10minus77 barbPO2b10minus42 bar) The lowest PO2 value whichwas used as a limit in the modeling was calculated from the S(+VI)S(minusII) redox couple Moreover the activity diagrams (Fig 6) indicatethat the data from the bicarbonate waters from the TarondashCeno Valleysfall between quartz whose log [SiO2] =minus433 is not shown because of
the slow precipitation kinetics under the PndashT conditions considered
here (Williams et al 1985) and amorphous silica boundary 1047297elds(Fig 6A) at the same manner bicarbonate waters with higher SiO2-activity from the Voltri Group fall on amorphous silica boundary(Fig 6B) In summary then the bicarbonate waters are supersaturatedin kaolinite ferrihydrite vermiculite and saponite Alkaline watersare also supersaturated with saponite (Fig 6AndashC) and they are alsoequilibrated with primary magnetite and under extreme conditionsprobably pyrite as well
55 Limitations of the model related to kinetics
Model calculations disregard the rates of mineral dissolution and
precipitation which becomes increasingly important as temperature
Table 3
Starting mineralogy (left side) and 1047298uid composition (right side) used in the modeling
Rock composition Fluid composition
Modal of the mineralsin the interacting rock
mol of the minerals usedin the modeling
meteoric water composition Initial 1047298uid composition used in the modeling
Spinel 003 Na 3523Eminus05 3523Eminus055 Spinel Chromite 001 S 3519Eminus05 3519Eminus05
Magnetite 0 Si ndash 1176Eminus06Kaolinite ndash 5258Eminus08
Ferrhydrite-6line ndash 5882Eminus07
The modal percentage of the minerals in a serpentinite from TarondashCeno Valleys (Ernst and Piccardo 1979) was re-calculated assuming an ideal solid solution between pure extremephasesand takingthe modal percentage tobe themass percentage Thelatter was used asthe solid reactant composition in thewaterndashrockinteraction modelPure solid phases werechosenfrom the llnldat thermodynamic database of the PHREEQCI softwareThe weighted average chemical compositionof the meteoric waterfrom Bologna (Panettiere et al 2000)was chosen as the initial 1047298uid composition in the modeling after its equilibration with kaolinite and ferrihydrite (see text for details)
Table 4
Summary of thermodynamic data for the sepiolite (Spn) and serpentine (Sr n) mineralsphases in italics do not refer to the mentioned hydrolysis reactions
ΔGf 0 log K Reference
(kJ molminus1) (at 25 degC)
Sepiolite (Spn) mdash Mg4Si6O15(OH)2middot6H2O+8H+=4Mg2++6SiO2+11H2Ominus925400 3004 Gunnarsson et al (2005)minus925133 3050andash3172 Christ et al (1973)
(Sp2) minus921986 3602andash3756 Wollast et al (1968)
Serpentine (Srn) mdash Mg3Si2O5(OH)4+ 6H
+
=3Mg
2+
+2SiO2+ 5H2ONi-lizardite minus 289324 1680 Mondesir and Decarreau (1987)
Al-lizardite2 minus411003 2544a Caruso and Chernosky (1979) Al-lizardite minus 339854 2786 this work (see text for details)
Mg-lizardite (Sr1) minus404025 3056 Wilson et al (2006)Mg-crysotile (Sr2) minus403702 3112 bthermocomV8R6Mg-lizardite minus402824 3340 Mondesir and Decarreau (1987)p-antigorite (Sr3) minus564239a 3479 Gunnarsson et al (2005)Mg-chrysotile (Sr4) minus406809 3669 Bruni et al (2002)Mg-chrysotileasbestos minus562505a 3888 Pfeifer (1977)Antigorite minus6612864 47719 bthermocomV8R6
The 1047297eld boundaries of the underlined phases shown in Fig 6 correspond to atemperature of 11 degC
a In the ΔG f deg column value calculated using the ΔG f
deg data of the elements from llnldat
and logK value in the reference in the LogK column value calculated using ΔG f deg
(mineral) in the reference and ΔG f deg data of the elements from llnldat
b Thermodynamic database of Lawrence Livermore National Laboratory included in
PHREEQCI and Geochemists Workbench softwares
85T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
decreases The waterndashrock interaction in the TarondashCeno basin iscomplicated by the presence of secondary mixed-layer clay mineralsthe origin of which is quite dif 1047297cult to model because the solidndashstatetransformations and weathering reactions are strictly related tohydrological and climatic conditions as described above An equili-
brium-based reaction path model is a valuable tool for assessing the
most probable reactions and for de1047297ning the product phases that canbe used in a kinetics-based model Since the kinetics of precipitationof Mg-clays at room temperature and pressure are poorly knownrelative reaction rate effects in the reaction path model can be mimedby omitting the stabler phase when a metastable mineral is expected
to precipitate Therefore the precipitation of MgndashFe3+
Abbreviations are as follows Fer-6line 6-line ferrihydrite kaol kaolinite verm vermiculite magn magnetite sap saponite serp serpentine clin clinochlore pyr pyrite calccalcite hydrom hydromagnesite bruc brucite enst enstatite The chromium decrease during the change in water facies from bicarbonate to alkaline was modeled withsupersaturation (precipitation) of eskolaite and thereafter with equilibrium with chromite (Boschetti 2003)- not bound dagger chlorite mineral from solution Dagger chlorite mineral added to the paragenesis
86 T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
montmorillonites vermiculite and saponites are predicted to occurrather than that of Mg-clays Moreover in order to simulate the slowprecipitation of calcite in depth precipitation was not activated atslight supersaturation (SIcalcite=001) during the initial stages of thewaterndashrock interaction in this manner saturation index of calcite wasfree to increase in the following steps and in alkaline waters
6 Discussion of the model results
61 Explanation of pre-modeling choices and post-modeling results
The results of the reaction path modeling are reported in Table 5and shown in Figs 6 and 9 The reaction paths model the dissolutionof a typical local serpentinite (mineralogically constrained as inTable 3) and the sequential precipitation of gibbsite kaoliniteferrihydrite vermiculite Fe2+ndashMg2+-saponite and a poorly-crystal-line serpentine( p-antigorite of Gunnarsson et al2005) The1047297rst stepinvolves the reaction of the ldquoinitial waterrdquo with serpentinite atconstant temperature (11 degC) PO2 (10minus26 bar) and PCO2 (10minus30 bar)this early stage corresponds to bicarbonate groundwater feeding theinvestigated spring system At the stage of primary magnetiteequilibrium and Fe2+ndashMg2+-saponite supersaturation the modelassumes the system to be closed to CO2 and to have a 1047297xed partialpressure of oxygen (PO2= 10minus40 bar) these conditions represent
when water switch from a bicarbonate to an alkaline compositionThe calculated amount of interacting serpentinite needed to obtainthe alkaline springs of the TarondashCeno Valleys (mean PCO2= 10minus7 bar)is approximately 00203 moles (see Table 5)
Bruni et al (2002) explained the origin of the alkaline waters of the Voltri Group as the result of the evolution of meteoricbicarbonate water following the precipitation of sepiolite withabundance similar to that of goethite This might be supported bythe observation that sepiolite or smectite precipitate at lowtemperature from solutions with low Al content when the pH is inthe range 75ndash95 or N95 respectively (Siffert and Wey 1962 Jonesand Galan 1988) but sepiolites form from the weathering of serpentines in arid climate regions whereas sepiolites are rare inhumid climate regions where serpentinite changes to Fe-smectite
(Singer 1989 Wrucke 1996) In the ophiolite rocks of the Northern
Apennine sepiolite is found only sporadically in veins (Giuseppetti1953 Veniale 1966) and is unknown in the ophiolites of the Tarondash
Ceno Valleys whereas it is thought that sepiolite of the Voltri Groupprecipitated from metasomatic 1047298uids during development of shearzones (Crispini and Capponi 2002)
As described above our model explains the transition frombicarbonate water to alkaline water (UM15 and PR10 samples) asthe result of low-temperature precipitation of a poor crystalline
serpentine with the solubility product of p-antigorite (see Sr3 inTable 4 and Fig 6) The model indicates that the bicarbonate waters of the Voltri Group with the highest [Mg2+][H+]2 activity ratios similarlyevolve towards alkaline waters (Fig 6B) The logK value of thesynthetic low temperature serpentine adopted in this model and its
Fig 7 Saturation indices of chrysotile and p-antigorite (Sr2 and Sr3 thermodynamic
data respectively in Table 4) for bicarbonate and alkaline waters coming fromultrama1047297c rocks of the TarondashCeno Valleys as a function of SiMg molar ratioThermodynamic data on chrysotile and p-antigorite are from the thermocomV8R6
database and Gunnarsson et al (2005) respectively
Fig 8 Activityplot forthe MgOndashSiO2ndashAl2O3ndashFe2O3(FeO)ndashH2O system analyzed forSiO2-phases and aqueous species at different log PO2 and log[aAl3+(aH+)3] values Springwaters coming from the TarondashCeno Valleys (symbols as in Fig 2) and Voltri Groupultrama1047297tes (symbols as in Fig 5) are shown The greyarea refers to bicarbonate watersfrom theVoltri Group (Mariniand Ottonello2002Bruniet al 2002) Thesepiolite 1047297eldis very narrow or hidden by saponite-Mg (1047297eld limits shown as continuous lines)saponite-Fe3+ndashMg (1047297eldlimits shown as dottedlines) or vermiculite(1047297eldlimits shownas dottedndashdashed lines) The talc stability 1047297eld and boundaries with saponite-Mg
montmorillonite-Mg and cronstedite-Mg are shown for comparison (dashed lines)
87T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
relative uncertainty of plusmn2 (see Fig 8 in Gunnarsson et al 2005) arecomparable to the logK variation reported for other low-temperatureMg-serpentines (Table 4) that is within the lines Sr1ndash4 for serpentineminerals where samples varying from bicarbonate composition toalkaline composition in the ultrama1047297c aquifer are grouped (Fig 6)Fig 7 shows a plot of calculated serpentine saturation index as a
function of the SiMg molar ratio for the water samples from theultrama1047297tes of the TarondashCeno ValleysThe SI andSiMgvalues of pureMg-serpentine (SI= 0 SiMg=067) provide a useful index forseparating bicarbonate waters (SIb0 SiMgb067) from alkalinewaters (SIN0 SiMgN067) and the distribution of data points isconsistent with the idea that bicarbonate waters approach saturationwith serpentine until they become supersaturated as alkaline waters
Based on the Mgvs SiO2(aq) activitydiagram and speci1047297cratiosoflog[aAl3+(aH+)3] Birsoy (2002) analyzed the behavior of sepiolite duringultrama1047297te weathering The diagram produces a very narrow stability1047297eldfor sepiolite(Fig 8) when it is programmedwith suitable conditionsfor the bicarbonate waters (log PO2=minus25 bar log [aAl3+(aH+)3]= 65) andthe alkaline waters (log PO2=minus50 bar log [aAl3+(aH+)3]= 45) sampled inthe present study This stability 1047297eld grows larger when the absence of
Mg-smectites and talc is assumed but this is an unrealistic conditionbecause (i) all these minerals occur in the rocks within the areainvestigated and (ii) at high [Mg2+][H+]2 activity ratios the 1047297eld of sepiolite is smaller than that of subordinate compared to those poorly-crystalline serpentine (p-antigorite) and clinochlore
Among the alkaline springs of the Voltri Group the waters withlow log [aCa2+(aH+)2] values (Fig 9) show patterns compatible withchlorite interaction (step 5c in Fig6D)The alkaline waters whose datafall near the clinochlore-14Aring 1047297eld (Fig 6D) cannot be explained bysepiolite precipitation because the sepiolite line does not cross thechlorite 1047297elds Thus the peculiar composition of some alkaline watersof the Voltri Group is probably due to interactions between water andprimary chlorites of the ultrama1047297c rocks (optional step 5c in Table 5cb) This is supported by the occurrence sometimes in large amounts
of metasomatic Mgndash(Fe)-chlorite (Scambelluri and Rampone 1999
Vissers and Strating 2001) in the Ligurian area where the relativeabundance of chlorite can reach 50 in veins in high strainserpentinite mylonites of the ErrondashTobbio ultrama1047297tes (Scambelluriet al 2001b) Otherwise in a chlorite-free primary paragenesis thesolution evolving on the Sr3 serpentine line reaches equilibrium withclinochlore but probably the latter does not precipitate due to theslow kinetics of precipitation under these conditions Thus thedissolution of additional reacting rock in the iterative weathering
process supplies Mg
2+
to the solution thereby favoring the Sr3 to Sr4poorly crystalline serpentine transformation In fact all the PR10samples and some of the alkaline springs from the Voltri Group showincreasing [Mg2+][H+]2 activity ratio and low variation in silicaactivity (Fig 6D) starting from the invariant triple point 4a (Sr3-serpentine Fe2+ndashMg-saponite chlinoclore-14 Aring saturation) andextending towards the brucite line This path can be explained asthe result of the transformation of poorly crystalline serpentine fromSr3 to Sr4 the latter is more stable at higher [Mg2+][H+]2 activityratios Finally at very low PO2 values vermiculite disappears and thestability 1047297elds of chlorite and Fe2+ndashMg-saponite become narrowerwhereas those of Fe2+-rich phases (eg Fe2+-saponite) become widerThe data for the alkaline waters with the highest log [aMg2+(aH+)2]from the Voltri Group fall near the stability 1047297eld of peridotite primaryminerals like forsterite and enstatite (Fig 6D) therefore the last stepsof theevolution involve theequilibration of thesolutionwithenstatite(steps 8a and 8c in Table 5ab)
62 Ca-rich alkaline waters
In preliminary modeling of the evolution of the TarondashCeno springwaters Boschetti (2003) concluded that meteoric water initiallysaturated in Fe3+ndashMg2+-chlorite Fe3+ndashMg2+-saponite and pyrite (andor goethite) and then interacted with the ultrama1047297te mineralassemblage eventually evolved into the alkaline waters In themodel of Boschetti (2003) chlorite saponite and calcite were added
Fig 9 Activity plot for the MgOndashCaOndash(CO2)ndashH2O system with phase limits calculated ata temperature of 11 degCSpring waterscoming from theTarondashCeno Valleys (symbols as inFig2) andVoltri Groupultrama1047297tes(symbolsas in Fig5) areshown Calculated path forwaterndashultrama1047297c rock interaction is shown (dotted line) with the number ndashlettercombinations referring to the appearance or disappearance of minerals (see Table 5)The grey area refers to bicarbonate waters from the Voltri Group (Marini and Ottonello2002)
Fig 10 Plots of (A) calcite saturation indexes and (B) PCO2 as a function of Clminus ascalculated by PHREEQCI software for waters from the TarondashCeno Valleys Alkalinesprings for the Voltri Group are also shown (small black diamonds Marini andOttonello 2002) The grey shading is deeper(A) in the saturation1047297eld(SI= 00plusmn05) and
(B) where the lowest log PCO2 value is approached
88 T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
to the solution in order to achieve a high [Ca2+][H+]2 activity ratio andsupersaturation in these phases However in light of the work of Clarket al (1992) and Bruni et al (2002) we believe that the calcitesupersaturation in alkaline waters (Fig 9 and 10) which in our modelis achieved by preventing the precipitation of calcite in the 1047297rst stepsof the reaction path (Table 5a and b) depends on the differencebetween CO2 partial pressureat thegivendepth(PCO2le10minus10bar) andat the surface (PCO2cong10minus35) Waters rising to the surface are far from
equilibrium with the atmosphere where the PCO2 is 4times10
6
timesgreater than the value at depth thus the water rapidly take up CO2which is promptly converted to CO3
2minus because of the high pH Thisleads to supersaturation with calcite through the following hydro-xylation reaction
As a result large amounts of travertine are generated in the VoltriGroup (Marini and Ottonello 2002) and carbonate concretionsknown as devilrsquo s coins are formed at Mt Prinzera in Taro valleyBelow the surface the SIcalcite of the alkaline waters is expected to belower than what the model estimates Under conditions of lowcarbonate alkalinity low CO2 and O2 partial pressure and high pHcalcite nucleation may be inhibited and the kinetics of precipitationreduced by the high concentrations of dissolved divalent cations suchas Mg2+ Fe2+ and Mn2+ (Sumner and Grotzinger 1996 Davis et al2000 Zhang and Dawe 2000 de Leeuw 2002) The fact that severalalkaline water samples fall within or near the calcitendashdolomiteequilibrium (Fig 9) may indicate magnesium calcite or protodolomitenucleation in those extreme subsurface conditions However possibleinteraction between some calcium-rich waters of the Voltri Group andCa-rich rocks such as ophicarbonates or rodingites cannot beexcluded these rocks are widespread in the Ligurian Apennineserpentinites (Treves and Harper 1994)
63 Sepiolite vs serpentine nucleation under weathering conditions
constraints on the kinetics and equilibrium-based approaches
The kinetics of precipitation of poorly-crystalline serpentine and
sepiolite cannot be compared quantitatively because rate constantsare available only for well-ordered crystalline phases (eg see Palandriand Kharaka 2004 for crystalline serpentine) Sepiolite precipitationis often invoked to explain magnesium scavenging from solutions butthe lack of kinetic data for this mineral leads to overly simplisticreaction transport models These models are forced to use kinetic datafor phyllosilicates such as kaolinite or muscovite (eg Xu et al 2001Bryan 2005) Moreover the low-temperature thermodynamic dataeven if they are fairly uncertain show that sepiolite solubilityincreases with decreasing crystallinity (Table 4 Stoessell 1988) Infact crystalline sepiolite requires a long time to reach equilibriumwith solution which imposes a slow control on thewater composition(eg Stoessell and Hay 1978) Consequently in modeling based onreaction transport that simulates freshly precipitated or poorly
crystalline sepiolite formation the logK value of Wollast et al (1968)is preferable to the lower values (Suarez and Šimůnek 1996) Thispreference agrees with the approach of Mariner (2003) and Jove Colon(2006) who adopt a logK (25 degC) value for sepiolite of 3044 It is alsosimilar to that reported by Wollast et al (1968) and it is six logK unitshigherthan the valueused in thermocomV8R6 dataset which is oftenused (perhaps improperly) to model the supergenic nucleation of sepiolite In the stability diagram in Fig 6 where the Sp1 and Sp2 linesrepresent the stability boundaries of crystalline and poorly-crystallinesepiolite respectively the evolution of TarondashCeno waters frombicarbonate to alkaline (Figs 6A and C) follows the stability 1047297eldboundary of the poorly-crystalline serpentine (between Sr2ndashSr3 lines)rather than that of the poorly-crystalline sepiolite (Sp2) as postulatedin themodel The lines for Sp2 Sr3 and amorphoussilica intersect at a
common point (Fig 6D) which is consistent with the nucleation of
sepiolite that typically occurs in solutions that are saturated orsupersaturated with amorphous silica and (Mg)-carbonate (egStoessell and Hay 1978 Birsoy 2002) Some bicarbonate watersfrom the Voltri Group have a silica activity near the amorphous silicaboundary (Fig 6B) but only a few alkaline springs of this group fall onthe poorly-crystalline sepiolite boundary line (Sp2 Fig 6D) Both of these features are inconsistent with the possibility that sepioliteprecipitates from supergenic fresh water in the Northern Apennine
Moreover these two results suggest that values for logK(sepiolite)b
3044 (see Table 4) should be used to model the dissolutionprecipitation of the well-crystallized phase that is the most commonform for the metasomatic sepiolite found in the investigated area
7 Conclusions
The waters issuing from ultrama1047297tes in the TarondashCeno Valleysoriginate from low-temperature interactions between meteoric waterand serpentinite rocks In this report the waterndashrock interaction hasbeen modeled using reaction paths in progress mode During their earlyinteraction with soil and rock minerals bicarbonate waters super-saturated in kaolinite ferrihydrite vermiculite and Fe2+ndashMg2+-saponitewhere theorigin of the secondary iron saponite as Fe2+-bearing smectiteagrees with the hypothesis of Guumlven (1988) Then while remainingsupersaturated with saponite the bicarbonate waters evolve into analkaline facies when poorly-crystalline serpentine precipitates Duringrock weathering the Al-bearing crystalline serpentine of oceanic originprobably promotes nucleation of the low-temperature serpentineamorphous phase and supplies Al3+ to the solution for saponiteprecipitation This interpretation is consistent with the existence of Al-poor supergenic serpentine (Neal and Stanger 1985) and the partiallymeteoric OH-isotopic imprint observed in other ultrama1047297c systems
Dif 1047297culties and complexities in the modeling arise fromuncertainties in the primary mineralogy of some local ophiolitesand from the presence of different lithotypes (rodingites ophi-carbonates gabbros) which can signi1047297cantly affect the composi-tion of water samples This can be seen in the example of the Candash
OH springs issuing from the Voltri Group (PO2 and PCO2 of 10minus70
and 10minus10 bar respectively Bruni et al 2002 Marini and
Ottonello 2002) These springs appear to be more evolved thanthe TarondashCeno alkaline waters as a result of their particularly longinteractions with serpentinites until equilibrium with enstatite isreached or with chlorite-bearing serpentinite The use of anupdated thermodynamic database in our model avoided problemsdue to mineralogical heterogeneity in the mixed-layer claysassociated with ophiolitic rocks this database accurately re1047298ectedthe chemical composition of the single phases Using this modelthe vermiculitendashchlorite and saponitendashchlorite associations seen inthe activity diagrams simulate respectively the presence of highand low charge corrensite observed in the 1047297eld The accuracy of ourmodel can be tested by comparing estimates of chlorine composi-tion resulting from our model and from experimental methods The
mean Cl concentration of the bicarbonate water is 41plusmn16 mgLtaking into account a Cl concentration of sim2 mgL in rainwater inthe area (cf Panettiere et al 2000 Venturelli 2003) 2 mgL of Clin spring water are estimated to come from serpentinite This valueagrees with the model which predicts that the switch frombicarbonate water to alkaline water requires the congruentdissolution of 00203 mol of serpentinite containing 100 ppm of soluble Cl thereby yielding water containing 2 ppm of Cl
In addition we think that weathering of the accessory albiteplagioclase is responsible for Na enrichment of the waters whereas Clenrichment in the alkaline waters is predicted to result from theleaching of NaCl 1047298uid inclusions Moreover the TarondashCeno alkalinesprings show the highest BCl ratios comparable with those of thehost metamorphic minerals (Scambelluri et al 2004) In contrast the
lower boron content and BCl ratio of the alkaline springs of the Voltri
89T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
Group are unexpected features and probably re1047298ect waterndashrockinteractions over a longer time period than in the TarondashCeno ValleysThe boron content in the Northern Apennine alkaline springs may becontrolled by several factors such as (i) the different histories of VoltriGroup and EL ultrama1047297tes (ii) the probable contribution of boronfrom NaCl 1047298uid inclusions (iii) the sinking of boron by travertineprecipitation (Hobbs and Reardon1999) brucite (Rhoades et al 1970)and smectite and (iv) decreases in boron release from minerals due to
increases in pH (Su and Suarez 2004) Acknowledgements
This research was carried out with the 1047297nancial support from theldquoYoung Researchers Projectrdquo (2001ndash2002) of the Italian Ministry of Universities Scienti1047297c Research and Technology (MURST) We thankGianniCortecci (Institute of Geosciences and EarthResources IGG-CNRPisa Italy) and Giampiero Venturelli (Earth Sciences Department Univer-sity of Parma Italy) for useful suggestions and critical review of themanuscript The authors are grateful to Enrico Dinelli (InterdepartmentResearch Centre for Environmental Sciences Alma Mater StudiorumUniversity of Bologna Ravenna Italy) for granting permission to use theGeochemistrsquos WorkbenchsoftwareTheconstructive reviews of theEditor
J Fein and of two anonymous Referees have signi1047297cantly improved the
manuscript
Appendix A Supplementary data
Supplementary data associated with this article can be found inthe online version at doi101016jchemgeo200808017
References
Abbate E Bortolotti V Passerini P 1980 Olistostromes and olistoliths In Sestini G(Ed) Development of the Northern Apennines geosyncline Sedimentary Geology4 pp 521ndash558
Alietti A 1956 Il minerale a strati misti saponite-talco di monte Chiaro (val di TaroAppen Emil) Rendiconti dellAccademia Nazionale dei Lincei 21 201 ndash207
Alietti A De Angelis G Montanari G Sgarlata F Tirelli G 1976 Il ferro nelle argilleappenniniche Periodico di Mineralogia 45 51ndash64
Anselmi B Mellini M Viti C 2000 Chlorine in the Elba Monti Livornesi and Murloserpentines Evidence for seandashwater interaction European Journal of Mineralogy12137ndash146
Apha-Awwa-Wef 1995 Standard Methods for the Examination of Water and Waste-water 19th Edition American Public Health Association American Water WorksAssociation Water Environment Federation
Barnes I LaMarche Jr VC Himmelberg GR 1967 Geochemical evidence of present-day serpentinization Science 156 830ndash832
Barnes I ONeil JR 1969 The relationship between 1047298uids in some fresh alpine-typeultrama1047297cs and possible modern serpentinization Western United StatesGeological Society of America Bulletin 80 1947ndash1960
Barnes I ONeil JRTrescases JJ1978 Present dayserpentinizationin New CaledoniaOman and Yugoslavia Geochimica et Cosmochimica Acta 42 144ndash145
Barnes I Rapp JB ONeil JR Sheppard RA Gude III AJ 1972 Metamorphicassemblages and the direction of 1047298ow of metamorphic 1047298uids in four instances of serpentinization Contributions to Mineralogy and Petrology 35 263ndash276
Barnes JDSelverstone J Sharp ZD 2006Chlorine isotope chemistry of serpentinitesfrom Elba Italy as an indicator of 1047298uid source and subsequent tectonic historyGeochemistry Geophysics Geosystems 7 doi1010292006GC001296
Beccaluva L Emiliani F Venturelli G Zerbi M 1973 Ca Fe Mg Mn Cr Ni Codistribution in some ultrama1047297c rocks outcropping in the Northern Apennines withsome geological remarks Ateneo Parmense Acta Naturalia 9 69ndash98
Beccaluva L Macciotta G Piccardo GB Zeda O 1984 Petrology of lherzolitic rocksfrom the Northern Apennine ophiolites Lithos 17 299 ndash316
BeccaluvaL Venturelli G ZanzucchiG 1975 Dati geologici e geochimici sui basalti diassociazione o1047297olitica dellAppennino ligure-emiliano Ateneo Parmense ActaNaturalia 11 789ndash802
BethkeCM 2002 The Geochemists Workbench A Userrsquos Guide to RxnAct2Tact ReactandGtplot Release40 HydrogeologyProgram Universityof IllinoisUrbanaIL 236pp
Birsoy R 2002Formation of sepiolitendashpalygorskite and related minerals from solutionClays and Clay Minerals 50 736ndash745
Bonatti E Lawrence JR Morandi N 1984 Serpentinization of oceanic peridotitestemperature dependence of mineralogy and boron content Earth and PlanetaryScience Letters 70 88ndash94
Bonifacio E Zanini E Boero V Franchini-Angela M 1997 Pedogenesis in a soilcatena on serpentinite in north-western Italy Geoderma 75 33ndash51
Boschetti T 2003 Studio geochimico e geochimico-isotopico di acque a composizione
estrema e termali dellrsquoAppennino Settentrionale Ph D Thesis Parma Univ
Brigatti MF Franchini G Lugli C Medici L Poppi L Turci E 2000 Interactionsbetween aqueous chromium solutions and layer silicatesApplied Geochemistry 151307ndash1316
Brigatti MF Lugli C Poppi L Venturelli G 1999 Iron-rich saponite dissolutionreactions and Cr uptake Clay Minerals 34 637ndash645
Brigatti MF Valdregrave G 1996 Mixed-layer chloritesmectite (corrensite) in ophioliticrock alteration products Neues Jahrbuch fuer Mineralogie Monatshefte 1 37 ndash47
Bruni J Canepa M Chiodini G Cioni R Cipolli F Longinelli A Marini L OttonelloG Vetuschi Zuccolini M 2002 Irreversible waterndashrock mass transfer accompany-ing the generation of the neutral MgndashHCO3 and high-pH CandashOH spring waters of the Genova province Italy Applied Geochemistry 17 455ndash474
Bryan CR 2005 Drift-Scale THC Seepage Model MDL-NBS-HS-000001 REV 04 YuccaMountain Project Las Vegas Nevada (US)Caruso LJ Chernosky JV 1979 The stability of lizardite Canadian Mineralogist 17
757ndash769Clark ID Fontes JC Fritz P 1992 Stable isotope disequilibria in travertine from high
pH waters Laboratory investigations and 1047297eld observations from Oman Geochi-mica et Cosmochimica Acta 56 2041ndash2050
Cortesogno L 1980 Il metamor1047297smo giurassico nelle o1047297oliti dellAppenninoSettentrionale due distinti cicli metamor1047297ci in ambiente oceanico O1047297oliti 5 5ndash18
Christ CLHostetler PB Siebert RM1973 Studies in the system MgO-SiO2-CO2-H2OIII American Journal of Science 273 65ndash83
Cortesogno L Lucchetti G 1982 Il metamor1047297smo oceanico nei gabbri o1047297oliticidellAppennino Ligure Aspetti mineralogici e petrogenetici Rendiconti dellaSocietagrave Mineralogica Italiana 38 561ndash579
Crispini L Capponi G 2002 Control of metasomatic alteration on the development of shear zone in ultrama1047297c rocks a case study from the Voltri Massif (Ligurian Alps)Architettura interna e cinematica delle zone di taglio Gruppo Italiano di GeologiaStrutturale mdash Riunione Annuale 2002 Pisa Italy 11ndash12 giugno 2002 XIV ndashXV
CrissRE 1999 Principlesof stableisotope distributionOxford University PressNew York
Davis KJ Dove PM De Yoreo JJ 2000 The role of Mg 2+ as an impurity in calcitegrowth Science 290 1134ndash1137
de Leeuw NH 2002 Moleculardynamics simulations of the growth inhibiting effect of Fe2+ Mg2+ Cd2+ and Sr2+ on calcite crystal growth Journal of Physical Chemistry B106 5241ndash5249
De Nardo MT Segadelli S Vescovi P 2007 Studio pilota per la caratterizzazionegeologicadellesorgentinellazona delM Nero (altaVal Cenoe alta ValNuremdash provincedi Parma e Piacenza) Il Geologo dellEmilia-Romagna 25 Nuova Serie pp 5 ndash21
Di Giulio A Geddo G 1990 Studio petrogra1047297co delle Arenarie di Casanova (alta valTrebbia Appennino settentrionale) Atti Ticinesi di Scienze della Terra 33 243ndash254
Dinelli E Lombini A Simoni A Ferrari C 1997 Heavy metals in serpentinite soils of selected outcrops of Piacenza and Parma provinces (Northern Apennines Italy)Mineralogica et Petrographica Acta 40 241ndash255
Epstein S Mayeda T 1953 Variation of 18O content of waters from natural sourcesGeochimica et Cosmochimica Acta 4 213ndash224
Ernst WG Piccardo GB 1979 Petrogenesis of some Ligurian peridotites I Mineraland bulk-rock chemistry Geochimica et Cosmochimica Acta 43 219ndash237
Feder F Trolard F Klingelhofer G Bourrieacute G 2005 In situ Moumlssbauer spectroscopy
evidence for green rust (fougerite) in a gleysol and its mineralogical transforma-tions with time and depth Geochimica et Cosmochimica Acta 69 4463ndash4483Fritz PClark IDFontes JCWhiticar MJFaberE 1992 Deuterium and
13C evidence
for low temperature production of hydrogen and methane in a highly alkalinegroundwater environment in Oman In Kharaka TK Maest AS (Eds) Proceed-ings of the 7th International Symposium on WaterndashRock Interaction BalkemaRotterdam pp 793ndash796
Giammetti F 1968 Le o1047297oliti di Roccaprebalza Costa della Guardia e Groppo di Gorro(Appennino Parmense) Ateneo Parmense Acta Naturalia 4 107ndash142
Giuseppetti G1953 Roccee minerali dellaformazioneo1047297olitica di Volpedo Rendicontidella Societagrave Mineralogica Italiana 9 85ndash134
Gran G 1952 Determination of equivalence point in the potentiometric titrationAnalyst 77 661ndash671
Gunnarsson I Arnorsson S Jakobsson S 2005 Precipitation of poorly crystalline antigoriteunder hydrothermal conditions Geochimica et Cosmochimica Acta 69 2813ndash2828
Guumlven N 1988 Smetites In Bailey SW (Ed) Reviews in Mineralogy 19 Hydrousphyllosilicates (exclusive of micas) Mineralogical Society of America MichiganBook Crafters Inc 497ndash559
Hansel CM Benner SG Fendorf S 2005 Competing Fe(II)-induced mineralization
pathways of ferrihydrite Environmental Science and Technology 39 7147ndash7153HobbsMYReardonEJ 1999 Effect of pH on boronco-precipitation by calcite further
evidence for non-equilibrium partitioning of trace elements Geochimica etCosmochimica Acta 63 1013ndash1021
Jones BF Galan E 1988 Sepiolite and palygorskite In Bailey SW (Ed) HydrousPhyllosilicates (exclusive of Micas) Reviews in Mineralogy vol 19 MineralogicalSociety of America Michigan Book Crafters Inc 631ndash674
Jove Colon CF 2006 Technical work plan for Thermodynamic database for chemicalmodelling TWP-MGR-PA-000039 REV 01 Yucca Mountain Project Las VegasNevada (US)
Kohls DW Rodda Jr JL 1967 Iowaite a new hydrous magnesium hydroxidendashferricoxychloride from the Precambrian of Iowa American Mineralogist 52 1261 ndash1271
Kyser TK OHanley DS Wicks FJ 1999 The origin of 1047298uids associated withserpentinization processes evidence from stable-isotope compositions CanadianMineralogist 37 223ndash237
Lapham DM 1961 New data on deweylite American Mineralogist 46 168ndash188Lee BD SearsSKGraham RC Amrhein CValiH 2003 Secondarymineral genesis
from chlorite and serpentine in an ultrama1047297c soil toposequence Soil Science
Society of America Journal 671309ndash1317
90 T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
Longinelli A Selmo E 2003 Isotopic composition of precipitation in Italy a 1047297rstoverall map Journal of Hydrology 270 75ndash88
Majzlan J Navrotsky A Schwertmann U 2004 Thermodynamics of iron oxides PartIIIEnthalpiesof formation and stability of ferrihydrite (simFe(OH)3) schwertmannite(simFeO(OH)34(SO4)18) and ɛ-Fe2O3 Geochimica et Cosmochimica Acta 681049ndash1059
Mariner P 2003 In-Drift PrecipitatesSalts Model ANL-EBS-MD-000045 REV 01 YuccaMountain Project Las Vegas Nevada (US)
Marini L Ottonello G 2002 Atlante degli acquiferi della Liguria Volume III Le acquedeicomplessio1047297olitici (bacini ArrestraBranega Cassinelle Cerusa ErroGorzenteLeira Lemme Lerone Orba Piota Polcevera Rumaro Sansobbia Stura Teiro
Varenna Visone) Pacini PisaMellini L Zanazzi PF 1987 Crystal structures of lizardite-l T and lizardite-2H1 fromColi Italy American Mineralogist 72 943ndash948
Mezzadri G 1964 Petrogra1047297a delle arenarie di Ostia Rendiconti Societagrave MineralogicaItaliana 20 192ndash228
Mondesir H Decarreau A 1987 Synthesegrave entre 25 et 200 degC de lizardites Ni ndashMgMesure des coef 1047297cients de partage solide-solution aqueuse pour le couple Ni ndashMgdans des lizardites Bulletin de Mineralogie 110 409ndash426
Neal C Shand P 2002 Spring and surface water quality of the Cyprus ophiolitesHydrology and Earth System Sciences 6 797ndash817
Neal C Stanger G 1983 Hydrogen generation from mantle source rocks in OmanEarth and Planetary Science Letters 66 315ndash320
Neal C Stanger G 1984 Calcium and magnesium hydroxide precipitation fromalkaline groundwaters in Oman and their signi1047297cance to the process of serpentinization Mineralogical Magazine 48 237ndash241
Neal C Stanger G 1985 Past and present serpentinisation of ultrama1047297c rocks anexample from Semail ophiolite nappe of Northern Oman In Drever JI (Ed) TheChemistry of Weathering D Reidel Publishing Company pp 249ndash275
Nesbitt HW Bricker OP 1978 Low temperature alteration processes affecting
et Cosmochimica Acta 41 1835ndash1841OrsquoHanley DS 1996 Serpentinites records of tectonic and petrological history Oxford
University PressOrberger B Friedrich G Woermann E 1990 The distribution of halogens and carbon
in Pge-bearing ultrama1047297cs of the Acoje Ophiolite Block Zambales Philippines Journal of Geochemical Exploration 37 147ndash169
Ottonello G Joron JL Piccardo GB 1984 Rare Earth and 3d transition elementsgeochemistry of peridotitic rocks II Ligurian peridotites and associated basalts Journal of Petrology 25 373ndash393
Palandri JL Kharaka YK 2004 A compilation of rate parameters of water-mineralinteraction kinetics for application to geochemical modeling Open File Report2004-1068 US Geological Survey
Panettiere P Cortecci G Dinelli E Bencini A Guidi M 2000 Chemistry and sulfurisotopic composition of precipitation at Bologna Italy Applied Geochemistry 151455ndash1467
Parkhurst DL Appelo CAJ 1999 Userrsquos guide to PHREEQC (version 2) mdash a computerprogram for speciation batch-reaction one-dimensional transport and inversegeochemical calculation US Department of the Interior US Geological SurveyWashington DC
Pfeifer HR 1977 A model for 1047298uids in metamorphosed ultrama1047297c rocks Observationsat surface and subsurface conditions (high pH spring waters) SchweizerischeMineralogische und Petrographische Mitteilungen 57 361ndash396
PlummerLN1984Geochemical modeling A comparisonof forwardand inverse methodsIn Hitchon B Wallick EI (Eds) Proceedings First CanadianAmerican Conference onHydrogeology-Practical Applications of Ground Water Geochemistry Banff AlbertaCanada mdash Worthington Ohio National Water Well Association pp 149ndash177
Rampone E Hofmann AW Piccardo GB Vannucci R Bottazzi P Ottolini L 1995Petrology mineral and isotope geochemistry of the External Ligurides peridotites(Northern Apennines Italy) Journal of Petrology 36 81ndash105
Rampone E Hofmann AW Piccardo GB Vannucci R Bottazzi P Ottolini L 1996Trace element and isotope geochemistry of depleted peridotites from an N-MORBtype ophiolite (Internal Liguride N Italy) Contributions to Mineralogy andPetrology 123 61ndash76
Rhoades JD Ingvalson RD HatcherJT1970Adsorption of boronby ferromagnesianminerals and magnesium hydroxide Soil Science Society of America Journal 34
938ndash941Saini-EidukatB KuchaH Keppler H1994Hibbingiteγ-Fe2(OH)3Clanewmineralfromthe Duluth Complex Minnesota with implications for the oxidation of Fe-bearingcompounds and the transport of metals American Mineralogist 79 555ndash561
Sanford RF 1981 Mineralogical and chemical effects of hydration reactions andapplications to serpentinization American Mineralogist 66 290ndash297
Scambelluri M Muumlntener O Ottolini L Pettke TT Vannucci R 2004 The fate of BCl and Li in the subducted oceanic mantle and in the antigorite breakdown 1047298uidsEarth and Planetary Science Letters 222 217ndash234
Scambelluri M Rampone E 1999 Mg-metasomatism of oceanic gabbros and itscontrol on Ti-clinohumite formation during eclogitization Contributions toMineralogy and Petrology 135 1ndash17
Scambelluri M Bottazzi P Trommsdorff V Vannucci R Hermann J Gogravemez-Pugnaire MT Logravepez-Sagravenchez Vizcaino V 2001a Incompatible element-rich1047298uidsreleased by antigorite breakdown in deeply subducted mantle Earth and PlanetaryScience Letters 192 457ndash470
Scambelluri M Rampone E Piccardo G 2001b Fluid and element cycling insubducted serpentinite a trace-element study of the ErrondashTobbio high-pressureultrama1047297tes (Western Alps NW Italy) Journal of Petrology 42 55ndash67
Segadelli S2006 Lageologia nelpaesaggio le rupi o1047297olitiche in Val Taro e Val Ceno IlGeologo dellrsquoEmilia-Romagna 22 15ndash29 Nuova Serie
Settembre Blundo D Martin De Vidales JL Alietti A 1992 Corrensite-type clayminerals from Taro Valley (North Italy) Mineralogica et Petrographica Acta 35215ndash222
Sharp ZD Barnes JD 2004 Water-soluble chlorides in massive sea1047298oor serpenti-nites a source of chloride in subduction zones Earth and Planetary Science Letters226 243ndash254
Siever R Woodford N 1979 Dissolution kinetics and the weathering of ma1047297cminerals Geochimica et Cosmochimica Acta 43 717ndash724Singer A 1989 Palygorskite and sepiolite group minerals In Dixon JB Weed SB
(Eds) Minerals in Soil Environments Soil Science Society of America MadisonWisconsin pp 829ndash872
Siffert B Wey R 1962 Synthese d rsquoune sepiolite a temperature ordinarie Comptesrendus hebdomadaires des seacuteances de lAcadeacutemie des Sciences de Paris 2541460ndash1463
Środoń J 1999 Nature of mixed-layer clays and mechanisms of their formation andalteration Annual Reviews of Earth and Planetary Science 27 19ndash53
Stoessell RK 1988 25 degC and 1 atm dissolution experiments of sepiolite and keroliteGeochimica et Cosmochimica Acta 52 365ndash374
Stoessell RK Hay RL 1978 The geochemical origin of Sepiolite and keorilite atAmboseli Kenya Contributions to Mineralogy and Petrology 65 255ndash267
Suarez DL Šimůnek J 1996 Solute transport modeling under variably saturatedwater 1047298ow conditions In Lichtner PC Steefel CI Oelkers EH (Eds) ReactiveTransport in Porous Media Reviews in Mineralogy vol 34 Mineralogical Society of America Michigan Book Crafters Inc 229ndash268
Sumner DY Grotzinger JP 1996 Were kinetics of Archean calcium carbonate
precipitation related to oxygen concentration Geology 24 119ndash122Su C Suarez DL 2004 Boron release from weathering of illites serpentine shales
and illiticpalygorskitic soils Soil Science Society of America Journal 68 96 ndash105Toscani L Venturelli G Boschetti T 2001 Sul1047297de-free and sul1047297de-bearing waters in
the Northern Apennines Italy general features and waterndashrock interaction AquaticGeochemistry 7 195ndash216
Treves BE Harper GD 1994 Exposure of serpentinites on the ocean 1047298oor sequenceof faulting and hydrofracturing in the Northern Apennine ophicalcites O1047297oliti 19b435ndash466
Varadachari C Kudrat M Gosh K 1994 Evaluation of standard free energies of formation of clay minerals by an improved regression method Clays and ClayMinerals 42 298ndash307
Veniale F 1966 Sepiolite in sedimenti dellrsquoAppennino Vogherese (Pavia) Periodico diMineralogia 35 343ndash386
Venturelli G 2003 Acque minerali ed ambiente mdash Fondamenti di geochimica deiprocessi di bassa temperatura Pitagora Editrice Bologna
Venturelli G Contini S Bonazzi A Mangia A 1997 Weathering of ultrama1047297c rocksand element mobility at Mt Prinzera Northern Apennines Italy MineralogicalMagazine 61 765ndash778
Venturelli G Frey M 1977 Anchizone metamorphism in sedimentary sequences of the Northern Apennines Rendiconti Societagrave Mineralogica Italiana 33 109ndash123
Vieillard P 2002 A new method for the prediction of Gibbs free energies of formationofphyllosilicates (10 Aring and14 Aring)basedon theelectronegativityscale Claysand ClayMinerals 50 352ndash363
Vissers RLM Strating EHH 2001 Structures and microstructures in a thrust-relatedgreenschist facies tectonic melange Voltri Group (NW Italy) O1047297oliti 26 33ndash46
Viti C Mellini M 1997 Contrasting chemical compositions in associated lizardite andchrysotile in veins from Elba Italy European Journal of Mineralogy 9 585 ndash596
Viti C Mellini M 1998 Mesh textures and bastites in the Elba retrogradeserpentinites European Journal of Mineralogy 10 1341ndash1359
Wenner DB Taylor Jr HP 1974 Deuteriumatomic hydrogen and oxygen-18oxygen-16 studies of serpentinization of ultrama1047297c rocks Geochimica et CosmochimicaActa 38 1255ndash1286
Williams LA Parks GA Crerar DA 1985 Silica diagenesis I Solubility controls Journal of S edimentary Petrology 55 301ndash311
Wilson J Savage D Cuadros J Shibata M RagnarsdottirKV 2006 The effect of ironon montmorillonite stability (I) Background and thermodynamic considerations
Geochimica et Cosmochimica Acta 70 306ndash322WollastR MackenzieFTBricker OP1968 Experimental precipitation and genesis of sepiolite at earth-surface conditions American Mineralogist 531 1645ndash1662
Wrucke CT1996 Serpentine and carbonated-hosted asbestos deposits InEdwardAdu Bray (Eds) Preliminary compilation of descriptive geoenvironmental mineraldeposit models 95ndash831 US Geological Survey Open-File Report pp 39 ndash46
Xu T Apps JA Pruess K 2001 Analysis of mineral trapping for CO2 disposal in deepaquifers Lawrence Berkeley National Laboratory University of California
ZhangYDaweRA2000 In1047298uence ofMg2+ on thekineticsof calcite precipitation andcalcite crystal morphology Chemical Geology 163 129ndash138
Zhu C Anderson G 2002 Environmental applications of geochemical modelingCambridge University Press
91T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
Basaltic rocks which are absent only from the area of Mt Prinzerasuffered an early low-grade oceanic metamorphism to varyingextents followed by an orogenic metamorphism (Cortesogno 1980)Albite epidote prehnite sericite chlorite actinolite titanite andhaematite association replaced the original mineralogy of plagioclaseolivine clinopyroxene ilmenite and magnetite (Cortesogno andLucchetti 1982) Spinel lherzolites re-equilibrated in the petrological1047297eld of plagioclase lherzolites and were exposed to oceanic and
orogenic metamorphisms Hydration of the early mineralogy pro-duced serpentine (lizardite and minor chrysotile) chlorite and talcand bastite after olivine and orthopyroxene respectively
The ultrama1047297tes of the Middle Taro valley and of Mt Prinzera arestrongly serpentinized (Venturelli et al 1997) whereas primaryclinopyroxene and plagioclase are largely preserved at Mt Aiona(Beccaluva et al 1973) Sedimentary rocks at the contacts with theophiolites are predominantly 1047298yschoid (marly to pelitic) and arenac-eous Typically 1047298ysch rocks are micritic marly turbidites withinterbedded pelites at MtCaio the latter are composed of illite andor illitesmectite mixed layers and chlorite (Venturelli and Frey1977)Arenaceous formations include (i) the Casanova Complex composedof quartz-feldspathic rocks with metamorphic serpentinitic andvolcanic clasts (Di Giulio and Geddo 1990) and (ii) the OstiaSandstones composed of quartz feldspar (plagioclase orthoclasemicrocline) mica (muscovite biotite) calcite and dolomite within aclayey and chlorite matrix at the base of the Mt Caio Flysch Formation(Mezzadri 1964)
Hydrogeologic studies on the TarondashCeno Valleys were focussedon their alluvial fan and plain probably because they cover adensely populated area Recently research projects consider theupstream section of the valleys and preliminary results show that77 of the springs are fed by aquifers contained in Late CretaceousLigurid 1047298ysches 11 in sandstones-pelite conglomerates andmassive sanstones (eg Ranzano Formation) and only 4 inophiolitic bodies (Segadelli et al 2006 De Nardo et al 2007)According to the authors the last value is probably underestimatedbecause the study consider only the known and gathered springs forwaterwork system whereas most of springs issuing from ophiolites
are free and unknown despite their perennial activity that ensurepreservation of the stream out1047298ow during dry periods De Nardoet al (2007) identify two types of ophiolite related springs (i)
springs with inde1047297nite permeability limit and (ii) with de1047297nitepermeability limit which are fed by aquifers contained in fractu-rated rocks and in coarse detritic covers respectively Some high1047298ow rate springs have different origin being related to gravitationalsliding (eg Mt Nero area) that increased permeability due toslackening of the pre-existing fractures (De Nardo et al 2007)
3 Analytical methods
31 Field methods
All water samples were 1047297ltered through cellulose acetate 1047297lters(045 μ m) An aliquot for cation analysis was acidi1047297ed with 65 HNO3
Suprapur Merck (1 cm3 HNO3 per 100 cm3 water) Labile parameters(temperature pH and Eh) were determined in the 1047297eld using anORION 250A instrument equipped with a Ross glass electrode formeasuring pH and a combined electrode of platinum-silversilverchloride for measuring Eh The electrode for Eh was calibrated atdifferent temperatures using ZoBells solution (Nordstrom 1977)Speci1047297c conductance at 20 degC was measured using a conductimeter(Model 85 Yellow Springs Instruments) Total alkalinity was deter-mined in the 1047297eld by acidimetric titration with 001 N HCl usingbromocresol green as an indicator and in the laboratory by Grantitration (Gran 1952) within 12 h of sampling Reduced dissolvedspecies were determined by spectrophotometry using a portablephotometer (Merck SQ300) with Spectroquant kits total ammoniumNH4= NH4
++ NH30 and total sulphide HS=H2S0+HSminus+ S2minus were deter-
mined by adding indophenol blue in basic solution (pH asymp13 method4500-NH3 F in APHA-AWWA-WEF1995) and methylene blue in acidicsolution (pHasymp05 method 4500-S2minus F in APHA-AWWA-WEF 1995)respectively
32 Laboratory methods
Clminus NO3minus and SO4
2minus were analyzed by ion chromatography usingDionex DX100 with anionic self-regenerating suppressor silica byspectrophotometry using a Merck SQ300 photometer and Na K Ca
Mg B Fe by radialsequential inductively coupled plasma opticalemission spectrometry (ICP-OES) using Philips PU7450 and Horiba
Jobin-Yvon Ultima 2 instruments Data on δ2H(H2O) and δ
18O(H2O)
Table 1
Summary of physico-chemical parameters and concentrations of chemical constituents in the spring waters coming from ultrama 1047297tes and basalts in the TarondashCeno Valleys
Type (lithology) Springs (ultrama1047297tes) Springs (basalts)
were collected using an automatic equilibration device (Finnigan GLF1086) in-line with a mass spectrometer (Finnigan Delta Plus) Theanalyzed H2 and CO2 gases were equilibrated with water samples at atemperature of 18 degC (Epstein and Mayeda 1953) using a Pt-richcatalyst Isotopic ratio are reported in per mill relative to V-SMOWand standard errors for δ
2H(H2O) and δ18O(H2O) are approximately plusmn
10permil and within plusmn01permil respectively
4 Chemical and isotopic results
The chemical compositionsof thewaters from theultrama1047297tesandbasalts are summarized in Table 1 Complete chemical and isotopedata are reported in Appendix A where all the water samples aresubdivided into four classes on the basis of the lithology present at thesource (i) ultrama1047297c (ii) basaltic (iii) sedimentary and (iv) surfacewaters
41 Chemical classi 1047297cation
All the samples investigated were freshwater (TDSb1 gL) Theternary classi1047297cation diagrams (Na+ K)ndashCandashMg and tAlkndashSO4ndashCl(Fig 2) show that waters of groups (ii) (iii) and (iv) do not differsigni1047297cantly many are bicarbonate-calcic with neutral to slightlybasic pH values (667ndash857) highly oxidized (Eh=371ndash514 mV) andcontain a small amount of chloride (Clminus up to 64 mgL) Sample UM1
from basalt is an exception since it is a ClndashCandashNa type with EC up to1043 μ S cmminus1 Clminus up to 231 mgL hardness up to 367 mgL CaCO3Waters issuing from ultrama1047297tes have more varied compositions Theyrange from Ca-bicarbonate with a pH of 73ndash88 and a hardness rangingfrommoderately hard to very hard (67ndash386mgL CaCO3 Eh129ndash581μ Scmminus1) to Mg-bicarbonate with a pH of 74ndash88 and a hardness from softto very hard (59ndash206 mgL CaCO3 92ndash330 μ S cmminus1) Among the Mg-bicarbonate waters sample UM8 shows the highest hardness values
Fig 2 MgndashSindashC ternary diagram (in molkg) and ternary classi1047297cation diagrams [Mgndash(Na+K)ndashCa plus ClndashtAlkndashSO4 in meqL] of alkaline springs in the TarondashCeno Valleys (this work)andthe VoltriGroup(Mariniand Ottonello 2002) Bottled watersfromTarovalleyare also representedRG1 andRG2Rocca Galgana 1990 and1994 respectively(chemical data fromwater bottle label) FN1 FN2 and FN3 Fontenova 1974 1990 and 1994 respectively (chemical data taken fromwater bottle label) MgndashSindashC contentsafter mineral dissolution (whitecrosses) are also represented (see text for details) Note that for dolomite and magnesite the CMg mole ratio refers to the products of the respective reactions CaMg(CO3)2+2H2O+
2CO2= Ca2+
+ Mg2+
+4HCO3
minus
and MgCO3+ H2O+CO2= Mg2+
+2HCO3
minus
79T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
(486 mgL CaCO3) conductivity (877 μ S cmminus1) and alkalinity (618 mgL HCO3) Springs PR10 and UM15 from ultrama1047297tes show an extremecomposition being sodium-hydroxide and sodium-sulfate respectivelytheir pH is high (101ndash112) their Eh is low (down to minus103 mV) theyaresoft (20ndash61 mgL CaCO3 220 to 408 μ S cmminus1) and they contain largeamounts of boron (up to 2294 μ gL and up to 13461 μ gL respectively)and moderate amounts of reduced sulfur and nitrogen species In bothPR10 and UM15 non-carbonate alkalinity predominates over carbonate
alkalinity using PHREEQCI software with the llnldat thermodynamicdatabase (Parkhurst and Appelo 1999) to calculate the composition of alkaline species under thepH Ehand T valuesmeasuredin1047297eld samplePR10 was found to contain 54 hydroxide 23 carbonate 19 borateand 4 other alkalinespecies (Si sulphide and ammonia) while sampleUM15 contained 75borate14 hydroxide10carbonate and 1 otherspecies
The SindashMgndashC diagram (Fig 2) is a useful tool for studying waterndashrockinteractions Dissolution reactions of typical magnesium minerals of ultrama1047297c rocks are represented by a line with CMg=2 (molar ratio)
Fig3 Plot of (A) Naand (B) Mgvs Cl forallwater samples from the TarondashCenoValleys including bottledwaters Rocca Galgana andFontenova(chemical data taken from labels)Alkalinesprings from the Voltri Group (smallblack diamonds Marini and Ottonello 2002) and meteoric water mean composition (MW Panettiere et al 2000 Venturelli 2003) are also plottedDottedcurvesin(A)and(B)refertomixingwith04and2wtNaCl 1047298uidinclusionsand MgCl2dissolution (1 mol)respectively Greycrosses andi csescorrespondrespectively to mixingwith04and2wtNaCl 1047298uidinclusionsfor mixing fractionsranging from 0001 to0005The solid linerefers tomixingof meteoric water(A) orMg-bicarbonate water(B) withseawater
The vertical grey line represents the water-soluble Cl content in the serpentinites from Northern Apennine (Barnes et al 2006) Symbols are as in Fig 2
80 T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
The Mg-bicarbonate watersissuingfrom ultrama1047297crocksfallnearthisline and cluster between the clinochlore and forsterite points (Fig 2)Stream waters and springs issuing from sedimentary and basaltic rocksfall between thecarbon apex and theline indicated by a CMg molar ratioof 4 which describes the dissolution of CandashMg-clinopyroxene
This is consistent with the (Ca)-bicarbonate compositions of thesestream waters and springs It is noteworthy that the Ceno streamwaters shifted toward the C apex of Fig 2 derive their chemistry fromthe interaction with Mt Caio Flysch cropping out widely in the Cenocatchment (samples from UM50 to UM59 see Map1 sheet in theannexes) whereas the Taro source spring waters plot towards theother apexes (lower C) typical of ophiolite minerals (samples fromUM41 to UM49 see Map1 sheet in the annexes) The use of the traceelements (eg Sr and Rb content Boschetti 2003) further highlightthe differences between the stream waters of the two rivers and therole of lithology Nonetheless mixing and equilibrium with atmo-sphere make the reaction-path modeling of these samples poorlyreliable so their are unsuitable for the purpose of the present work
Alkaline waters from ultrama1047297c rocks fall inside the same sub-triangle mostly along the CndashSi side probably due to the precipitationof magnesium minerals during the evolution of the water composition(Bruni et al 2002)
42 Sodiumndashchloride relationships
Chlorine and other halogens in serpentinized rocks are inheritedfrom seawater and marine sediments during oceanic metamorphism(eg Orberger et al 1990) They are transferred into serpentinite byadsorption (Anselmi et al 2000 Sharp and Barnes 2004) entrap-ment in 1047298uid inclusions (Scambelluri et al 2001ab) and substitutionin the structure of low-crystalline iron-hydroxide phases These latterare composed of the isomorphous and chemically similar iowaitehibbingite and FendashCl-oxydroxides (also named green rusts orfougerites Kohls and Rodda 1967 Saini-Eidukat et al 1994 Federet al 2005 and references therein) In the Northern Apennineserpentinite rocks the mean Cl content is sim300 ppm mean values of 221 and 375 ppm were measured respectively in the Erro ndashTobbio(Ligurian Western Alps Scambelluri et al 2004) and in the centralItaly serpentinites (Anselmi et al 2000) but thewater-soluble Cl from
serpentinites is estimated to be than 100 ppm (Barnes et al 2006) Asshown in Fig 2 waters from basalts and ultrama1047297tes of the TarondashCenoValleys show a trend towards the Cl apex whereas in Fig 3 theestimated soluble chlorine in serpentinites agrees with the highestvalue reported in spring waters from ultrama1047297tes (samples fromMarini and Ottonello analyzed in 2002) The Cl content of samplesUM1 and UM36is veryhigh(Fig 3) and is probablydue to pollutionInfact these springs are not far from the A15 highway (UM1) and theTomarlo Pass road (UM36) where salt mixtures comprising NaClMgCl2middot6H2O and probably CaCl2 are widely used for deicing in winterAlkaline water with a Mg-bicarbonate composition from ultrama1047297tesshow the highest and lowest Cl contents respectively Na-rich Mg-bicarbonate waters from basalts (UM6 UM7 Fig 3) and ultrama1047297tescan be explained by the dissolution of albite in basalts (Beccaluva
et al 1975) and in spilitized EL peridotites (Beccaluva et al 1984
Rampone et al 1995) respectively whereas the highest content of both Na and Cl in the alkaline water samples UM15 and PR10 (Fig 3)can be explained by leaching of NaCl 1047298uid inclusions in theserpentinite rocks (04ndash20 wt NaCl Scambelluri et al 2004)
43 Boron and BCl ratio
Compared with unaltered ultrama1047297c rocks serpentinites have
higher content of both boron and chlorine and most authors suggestthat both elements come from seawater (eg Sanford 1981 andreferences therein Bonatti et al 1984) Despite several studies onthese rocks the behavior of Cl and B in solution during waterndashrockinteraction processes is not well known Early investigations on 1047298uidsassociated with serpentinites in the western United States (Barnes andONeil 1969 Barnes et al 1972) revealed metamorphic waters withhighboron levels(up to 28mM) and BCl molar ratiosof 0038ndash013 aswell as Mg-bicarbonate and high-pH waters of meteoric origin withvariable BCl ratios of 0027ndash042 (median 0039) and 00006ndash0017(median 00076) respectively Bicarbonate waters issuing fromophiolitic rocks in Cyprus have BCl ratio in this range (00017ndash
0058 median 00019 Neal and Shand 2002) whereas low boron
Fig 4 Plots of (A) δ18O vs δ2H and (B) δ18O vs altitude for waters from the Taro ndashCenoValleys C-Italy and N-Italy are the meteoric water lines for Central and Northern Italy
respectively (Longinelli and Selmo 2003) Symbols are as in Fig 2
81T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
levels were found in the alkaline springs of meteoric origin in OmanNew Caledonia and ex-Yugoslavia (Barnes et al 1978)
Among the waters analyzed so far in the TarondashCeno Valleys thebicarbonate waters have BCl ratios in the range 00017ndash022 (median0073) whereas the alkaline waters show the highest ratios everreported (BCl= 034ndash33 median 169) The high boron content of upto 13 mgL of the alkaline springs could be related to metamorphic1047298uids released during serpentinite subduction in fact rocks prefer-
entially lose Cl and the BCl ratio of the released 1047298
uids increases withpressure and temperature (Scambelluri et al 2004) Nonetheless thealkaline water samples in our study differ signi1047297cantly from alkalinemetamorphic1047298uids in that they have low salinity high BCl ratios andmeteoric isotope signatures (see below) all these characteristicssuggest an origin in the interaction of meteoric water with high-pressure metamorphic serpentinite This hypothesis is furthersupported by the agreement in BCl ratio between the alkalinesamples we study here and the high BCl ratio of up to 4 previouslyreported for the residual high-pressure mineralogy of ultrama1047297tes(olivine+orthopyroxene Scambelluri et al 2004) In contrast theVoltri Groups Ca-hydroxide waters have mean values of B = 014 mgL and BCl=003 (Marini and Ottonello 2002) consistent with acontribution of boron and chlorine from 1047298uid inclusions In factassuming fractional contributions of 0001ndash0005 from NaCl 1047298uidinclusions with boron content from 4 to 8 mgL ( Fig 3a Scambelluriet al 2004) results in a B content of 0004ndash01 mgL in the solutionThese results suggest that the difference between springs from theVoltri Group and TarondashCeno Valleys is probably related to (i) differentevolution of the 1047298uids and (ii) the different types of host rocks and Pndash
T paths of the Ligurian Western Alps and EL peridotites (eg Beccaluvaet al 1984 Scambelluri et al 2004)
44 Isotopic composition of water
A diagram of δ2H vs δ18O (Fig 4A) shows that the waters sampledfrom the TarondashCeno Valleys are meteoric in origin since they areconsistent with the meteoric water lines of central and northern Italy(Longinelli and Selmo 2003) The altitude vs δ18O(H2O) (Fig 4B)
summarizes the isotopic effects due to temperature altitude andamount of rainfall (eg Criss1999) all of which help to determine theisotopic composition of the water samples The tendency is similar forbicarbonate waters from sedimentary basaltic and ultrama1047297c forma-
tions which may provide insight into the recharge areas alkalinewaters in contrast probably circulate deeper in the ground
5 Water ndashrock interaction model
51 Fundamentals
The reaction path model between meteoric water and serpentinite
was performed using PHREEQCI 2125 code (Parkhurst and Appelo1999) and the Lawrence Livermore National Laboratory thermody-namic database (llnldat as known as thermocomV8R6 ) implementedwith the equilibrium constants listed in Table 2
Computations wereperformed by reaction progress mode ( forward
modeling Plummer 1984 Zhu and Anderson 2002) ie (i) addingreacting rock at each step in the simulation (ii) activating the phase(s)precipitation only when the thermodynamic stability of the mineralparagenesis is well known and consistent with the mineralogy of thearea and (iii) plotting reaction progress on speci1047297c activity diagramsThe same thermodynamic database was used with The GeochemistsWorkbench 403 software (Bethke 2002) to generate Pourbaix andactivity diagrams (Figs 5 6 8 and 9) Modeling was performed usingan approach based on equilibrium rather than reaction transportbecause the data on reaction kinetics and reactive surfaces in thesesolid phases systems are scarce particularly with respect to theactivation energy of clay and mixed-layer minerals at low temperatureand high pH Moreover the hydrogeology and the discontinuity of theaquifers in the study area which are related to the geologicheterogeneity and structural complexity of the Northern Apenninewould make a model based on reaction transport even less credible atthis stage
The modeling in this study was performed on the basis of thefollowing knowledge (1) primary and secondary mineral assem-blages (2) chemical composition of the minerals involved (3) physicalparameters and chemical composition of the starting 1047298uid (4)thermodynamic data on pure phases
52 The solid reactant primary paragenesis
EL peridotites of the TarondashCeno Valleys (NorthernApennines Italy)are spinel lherzolites containing olivine (forsterite) orthopyroxene(enstatite) and clinopyroxene (diopside) (Giammetti 1968 and
Table 2
Thermodynamic data on hydrolysis reactions added to the llnldat database of the PHREEQCI code (Parkhurst and Appelo 1999) and thermocomV8R6dat database of TheGeochemists Workbench software (Bethke 2002)
Mineral Reaction log K (TdegC) ΔGdeg f References
0 25 60 100 (kJ mol-1)
clinochlore08-daphnite02 Mg4Fe1Al2Si3O10(OH)8 + 16H+=2Al3+ + Fe2++12H2O+4Mg2++3SiO2 7318 6384 5272 4269 minus786379 this workferr ihydrite 2-l ine Fe(OH)3+3H+= Fe3++ 3H2O 484 355 ndash ndash minus70850 Majzlan et al 2004ferr ihydrite 6-l ine Fe(OH)3+3H+= Fe3++ 3H2O 435 312 ndash ndash minus71100 Majzlan et al 2004liz09-berth01 Mg27Fe02Al02Si19O5(OH)4+64H+=02Al3++02Fe2++52H2O+27Mg2++19SiO2 3111 2786 2383 2005 minus399854 this workp-antigorite Mg
Saponite-Fe2+ Fe3175Si365Al035O10(OH)2+74H+=035Al3++3175Fe2++47H2O+365SiO2 1845 1635 1319 1001 minus452459 Wilson et al 2006Saponite-Fe2+ndashMg Fe0165Mg3Si367Al033O10(OH)2+732H+=033Al3++0165Fe2++466H2O+3Mg2++
367SiO2
3037 2729 2330 1921 ndash Marini and Ottonello(2002)
The Gibbs free energy of formation (ΔGo f ) of the two theoretical Fe3+ (Varadachari et al1994) and Fe2+ vermiculites were calculated using a previously published method (Vieillard
2002 and Vieillard pers comm)log K values extimated considering an ideal solid solution between the end-members and using data thermodynamic data from llnldat (clinochlore08-daphnite02) and Wilson
et al (2006) (liz09-berth01)
82 T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
references therein Venturelli et al 1997) Accessory phases areplagioclase picotite plusmn calciumndashsodic amphibole (Venturelli et al1997) Before modeling the crystal chemical formula of the mineralphases was calculated from data from Venturelli et al (1997) thendepending on the modal composition of the rock (Ernst and Piccardo1979) and the composition of the ideal solid solution of the minerals
the resulting stoichiometry of the solid reactant(s) was calculated andinserted in the input 1047297le of the PHREEQCI software (Table 3)Thermodynamic data of the primary mineral paragenesis are all wellknown and are included in the llnldat thermodynamic database of PHREEQCI
53 The solid reactant secondary paragenesis
Secondary minerals in the ultrama1047297tes of the TarondashCeno Valleysare serpentine (lizardite and rare chrysotile) smectite minerals (Fendash
Mg saponites) and Fe-oxi-hydroxides Other accessory minerals arechlorite mixed layer saponitendashchlorite (low-charge corrensite) talccalcite and dolomite (Beccaluva et al 1973 Dinelli et al 1997Venturelli et al 1997 Brigatti et al 1999 2000) In addition mixed
layer saponitendashtalc (aliettite) and vermiculitendashchlorite (high-charge
corrensite) are present (Alietti 1956 Settembre Blundo et al 1992Brigatti and Valdregrave 1996) Except for serpentine themodal abundanceof these phases is unknown In addition some of these minerals ndash
particularly interlayer clays ndash may have different origins (hydrothermalor weathering allogenicor authigenic)and different stabilitiesrelatedtokinetics and drainage 1047298ow rate For example chlorite weathering yields
vermiculite-bearing or saponite-bearing interlayers under good- andpoor-drainage regimes respectively (Bonifacio et al 1997 Środoń1999) In addition the thermodynamic data available for serpentineinclude information only on chrysotile which is less abundant thanlizardite in the serpentinites of the areas that we studied According toVenturelli et al (1997) the lizardite composition in the serpentinites of the Taro valley can be summarized as
which is similar to 1-T lizardite from the Monte Fico serpentinizedperidotite (Elba Island Northern Tyrrhenian Sea Italy Viti and Mellini1997)
Using the thermodynamic data of Wilson et al (2006) and
assuming an ideal solid solution between lizardite and berthierine
Fig 5 EhndashpH iron diagram of the spring waters coming from ultrama1047297c rocks of the TarondashCeno Valleys Bicarbonate and high-pH waters of the Voltri Group are also represented(dotted area andsmallblackdiamonds respectively) Thelowest Eh value of thehigh-pH waterswas calculated using theS(+VI)S(minus II)redox couple(Marini and Ottonello 2002)Thestability 1047297eld of mineral phases are in grey and those of aqueous species in white In diagrams A and B 6-line ferrihydrite-vermiculite and goethite were selected as subaerial solidphases respectively
83T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
we calculated ΔGf 0=minus399854 kJmol for a theoretical (09)(lizardite)+
(01)(berthierine) mineral (Table 2) The log K value of the stated solidsolution was calculated at different temperatures using PHREEQCIafter loading the thermodynamic data on the two end-members intothe database The log K results obtained for Al-lizardite were added tothe thermodynamic database of the solid reactant (Table 3) andcompared with data from the literature (Table 4)
54 Composition of the starting solution O 2 partial pressure and
minerals included in the model
The weighted average chemical composition of rainwater inBologna (Panettiere et al 2000) which is 95 km east of theinvestigated area was used as the starting solution in the modelFull chemical data on rainwater of the TarondashCeno Valleys are
unavailable
Total alkalinity was recalculated for log PCO2=minus35 bar and
T =11 degC which was the mean temperature of the springs sampledand for pe (or Eh) based on chemical data for the redox couple N(minusIII)N(+V) The values of pH and pe obtained were 566 and 452(Eh=025 V) respectively the former agrees with the median valuereported for rainwater in thearea (Venturelli et al1997) Because dataon Al Fe and Si in rainwater were lacking we calculated theconcentrations of these elements assuming logPCO2=minus30 bar andslight supersaturation relative to ferrihydrite-6 line and kaolinite Inthis way a hypothetical composition of water in equilibrium with thelocal soil was calculated (Table3) and adopted as the ldquoinitial waterrdquo formodeling
Evidence suggests that secondary minerals particularly Fe-bearingphases change the chemical composition of the water 1047298owingthrough the ultrama1047297tes of the TarondashCeno Valleys According to Neal
and Stanger (1983) and Fritz et al (1992) the production of CH4 and
Fig 6 Activity plotfor the MgOndashSiO2ndashAl2O3ndashFe2O3(FeO)ndashH2O systemat differentlog PO2 values Spring waters issuing from TarondashCeno Valleys (symbols as in Fig 2) andVoltri Groupultrama1047297tes are shown (grey area bicarbonate waters small black diamonds alkaline waters from Marini and Ottonello 2002) In plots A B and C the stability 1047297eld limits of vermiculite(Table 2) correspond to25 degCwhilethe amorphoussilica (am-SiO2) brucite serpentine (Sr1ndash4 in Table 4)and sepiolite(Sp1ndash2 in Table 4) phasescorrespondto 11 degCThepath calculatedfor interaction between waterand ultrama1047297c rockis also shown (dotted line)the number-letter combinations referto the appearanceand disappearance of minerals(see Table 5)
84 T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
H2 in thealkalinewaters1047298owing from theultrama1047297tes isrelatedto theconsumption of oxygen during the oxidation of iron forming thehematitendashmagnetite pair Therefore the activity of iron and theoxidation state of the water must be de1047297ned The analysis of watersamples from ultrama1047297tes were plotted in both the Pourbaix andactivity diagrams (Figs 5 and 6) where the calculated activity of Fe2+Mg2+ SiO2 SO4
2minus (Pourbaix) and log PO2 (activity diagrams) wereinvolved Iron concentration was quite variable in the TarondashCenosprings where the average content was 27 μ gL and the contentreached a maximum of 200 μ gL and in the Voltri springs where theaverage was 44 μ gL and the maximum was 948 μ gL (Marini andOttonello 2002) In both cases the variations are probably related to
the different mineralogies of the soils and the different solubilities of the Fe-rich phases like goethite and ferrihydrite Iron-bearingsaponites and montmorillonites are the prevailing clay minerals inthe area investigated probably because of the ubiquitous presence of iron oxy-hydroxide (Alietti et al 1976 G Venturelli pers comm) Forthis reason the model considers the involvement of pure Mg-smectiteto be negligible
In the Pourbaix diagram (Fig 5) the data from bicarbonate watersfall into the ferrihydrite-goethitendashvermiculite1047297elds and the data fromalkaline waters fall on the boundaries of the vermiculitesaponite andsaponitechlorite Since ferrihydrite is the 1047297rst precipitating Fe-phaseit was used rather than goethite in the modeling The diagram alsosuggests the transformation of ferrihydrite into magnetite during thechange from bicarbonate to alkaline composition but this process
requires a large amount of Fe2+ in solution (N2 mM Hansel et al2005) which is unlikely at this stage of the waterndashrock interactionThus the magnetite participating in the equilibrium is most likelyprimary in origin
The modeling and the activity diagrams (Figs 6 and 9) werecalculated using the average value PO2= 10minus26 or 10minus50 bar forbicarbonate and alkaline waters respectively These values arecompatible with those calculated by Marini and Ottonello (2002) forbicarbonate waters (PO2= 10minus40 bar) and for alkaline waters of theVoltri Group (10minus77 barbPO2b10minus42 bar) The lowest PO2 value whichwas used as a limit in the modeling was calculated from the S(+VI)S(minusII) redox couple Moreover the activity diagrams (Fig 6) indicatethat the data from the bicarbonate waters from the TarondashCeno Valleysfall between quartz whose log [SiO2] =minus433 is not shown because of
the slow precipitation kinetics under the PndashT conditions considered
here (Williams et al 1985) and amorphous silica boundary 1047297elds(Fig 6A) at the same manner bicarbonate waters with higher SiO2-activity from the Voltri Group fall on amorphous silica boundary(Fig 6B) In summary then the bicarbonate waters are supersaturatedin kaolinite ferrihydrite vermiculite and saponite Alkaline watersare also supersaturated with saponite (Fig 6AndashC) and they are alsoequilibrated with primary magnetite and under extreme conditionsprobably pyrite as well
55 Limitations of the model related to kinetics
Model calculations disregard the rates of mineral dissolution and
precipitation which becomes increasingly important as temperature
Table 3
Starting mineralogy (left side) and 1047298uid composition (right side) used in the modeling
Rock composition Fluid composition
Modal of the mineralsin the interacting rock
mol of the minerals usedin the modeling
meteoric water composition Initial 1047298uid composition used in the modeling
Spinel 003 Na 3523Eminus05 3523Eminus055 Spinel Chromite 001 S 3519Eminus05 3519Eminus05
Magnetite 0 Si ndash 1176Eminus06Kaolinite ndash 5258Eminus08
Ferrhydrite-6line ndash 5882Eminus07
The modal percentage of the minerals in a serpentinite from TarondashCeno Valleys (Ernst and Piccardo 1979) was re-calculated assuming an ideal solid solution between pure extremephasesand takingthe modal percentage tobe themass percentage Thelatter was used asthe solid reactant composition in thewaterndashrockinteraction modelPure solid phases werechosenfrom the llnldat thermodynamic database of the PHREEQCI softwareThe weighted average chemical compositionof the meteoric waterfrom Bologna (Panettiere et al 2000)was chosen as the initial 1047298uid composition in the modeling after its equilibration with kaolinite and ferrihydrite (see text for details)
Table 4
Summary of thermodynamic data for the sepiolite (Spn) and serpentine (Sr n) mineralsphases in italics do not refer to the mentioned hydrolysis reactions
ΔGf 0 log K Reference
(kJ molminus1) (at 25 degC)
Sepiolite (Spn) mdash Mg4Si6O15(OH)2middot6H2O+8H+=4Mg2++6SiO2+11H2Ominus925400 3004 Gunnarsson et al (2005)minus925133 3050andash3172 Christ et al (1973)
(Sp2) minus921986 3602andash3756 Wollast et al (1968)
Serpentine (Srn) mdash Mg3Si2O5(OH)4+ 6H
+
=3Mg
2+
+2SiO2+ 5H2ONi-lizardite minus 289324 1680 Mondesir and Decarreau (1987)
Al-lizardite2 minus411003 2544a Caruso and Chernosky (1979) Al-lizardite minus 339854 2786 this work (see text for details)
Mg-lizardite (Sr1) minus404025 3056 Wilson et al (2006)Mg-crysotile (Sr2) minus403702 3112 bthermocomV8R6Mg-lizardite minus402824 3340 Mondesir and Decarreau (1987)p-antigorite (Sr3) minus564239a 3479 Gunnarsson et al (2005)Mg-chrysotile (Sr4) minus406809 3669 Bruni et al (2002)Mg-chrysotileasbestos minus562505a 3888 Pfeifer (1977)Antigorite minus6612864 47719 bthermocomV8R6
The 1047297eld boundaries of the underlined phases shown in Fig 6 correspond to atemperature of 11 degC
a In the ΔG f deg column value calculated using the ΔG f
deg data of the elements from llnldat
and logK value in the reference in the LogK column value calculated using ΔG f deg
(mineral) in the reference and ΔG f deg data of the elements from llnldat
b Thermodynamic database of Lawrence Livermore National Laboratory included in
PHREEQCI and Geochemists Workbench softwares
85T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
decreases The waterndashrock interaction in the TarondashCeno basin iscomplicated by the presence of secondary mixed-layer clay mineralsthe origin of which is quite dif 1047297cult to model because the solidndashstatetransformations and weathering reactions are strictly related tohydrological and climatic conditions as described above An equili-
brium-based reaction path model is a valuable tool for assessing the
most probable reactions and for de1047297ning the product phases that canbe used in a kinetics-based model Since the kinetics of precipitationof Mg-clays at room temperature and pressure are poorly knownrelative reaction rate effects in the reaction path model can be mimedby omitting the stabler phase when a metastable mineral is expected
to precipitate Therefore the precipitation of MgndashFe3+
Abbreviations are as follows Fer-6line 6-line ferrihydrite kaol kaolinite verm vermiculite magn magnetite sap saponite serp serpentine clin clinochlore pyr pyrite calccalcite hydrom hydromagnesite bruc brucite enst enstatite The chromium decrease during the change in water facies from bicarbonate to alkaline was modeled withsupersaturation (precipitation) of eskolaite and thereafter with equilibrium with chromite (Boschetti 2003)- not bound dagger chlorite mineral from solution Dagger chlorite mineral added to the paragenesis
86 T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
montmorillonites vermiculite and saponites are predicted to occurrather than that of Mg-clays Moreover in order to simulate the slowprecipitation of calcite in depth precipitation was not activated atslight supersaturation (SIcalcite=001) during the initial stages of thewaterndashrock interaction in this manner saturation index of calcite wasfree to increase in the following steps and in alkaline waters
6 Discussion of the model results
61 Explanation of pre-modeling choices and post-modeling results
The results of the reaction path modeling are reported in Table 5and shown in Figs 6 and 9 The reaction paths model the dissolutionof a typical local serpentinite (mineralogically constrained as inTable 3) and the sequential precipitation of gibbsite kaoliniteferrihydrite vermiculite Fe2+ndashMg2+-saponite and a poorly-crystal-line serpentine( p-antigorite of Gunnarsson et al2005) The1047297rst stepinvolves the reaction of the ldquoinitial waterrdquo with serpentinite atconstant temperature (11 degC) PO2 (10minus26 bar) and PCO2 (10minus30 bar)this early stage corresponds to bicarbonate groundwater feeding theinvestigated spring system At the stage of primary magnetiteequilibrium and Fe2+ndashMg2+-saponite supersaturation the modelassumes the system to be closed to CO2 and to have a 1047297xed partialpressure of oxygen (PO2= 10minus40 bar) these conditions represent
when water switch from a bicarbonate to an alkaline compositionThe calculated amount of interacting serpentinite needed to obtainthe alkaline springs of the TarondashCeno Valleys (mean PCO2= 10minus7 bar)is approximately 00203 moles (see Table 5)
Bruni et al (2002) explained the origin of the alkaline waters of the Voltri Group as the result of the evolution of meteoricbicarbonate water following the precipitation of sepiolite withabundance similar to that of goethite This might be supported bythe observation that sepiolite or smectite precipitate at lowtemperature from solutions with low Al content when the pH is inthe range 75ndash95 or N95 respectively (Siffert and Wey 1962 Jonesand Galan 1988) but sepiolites form from the weathering of serpentines in arid climate regions whereas sepiolites are rare inhumid climate regions where serpentinite changes to Fe-smectite
(Singer 1989 Wrucke 1996) In the ophiolite rocks of the Northern
Apennine sepiolite is found only sporadically in veins (Giuseppetti1953 Veniale 1966) and is unknown in the ophiolites of the Tarondash
Ceno Valleys whereas it is thought that sepiolite of the Voltri Groupprecipitated from metasomatic 1047298uids during development of shearzones (Crispini and Capponi 2002)
As described above our model explains the transition frombicarbonate water to alkaline water (UM15 and PR10 samples) asthe result of low-temperature precipitation of a poor crystalline
serpentine with the solubility product of p-antigorite (see Sr3 inTable 4 and Fig 6) The model indicates that the bicarbonate waters of the Voltri Group with the highest [Mg2+][H+]2 activity ratios similarlyevolve towards alkaline waters (Fig 6B) The logK value of thesynthetic low temperature serpentine adopted in this model and its
Fig 7 Saturation indices of chrysotile and p-antigorite (Sr2 and Sr3 thermodynamic
data respectively in Table 4) for bicarbonate and alkaline waters coming fromultrama1047297c rocks of the TarondashCeno Valleys as a function of SiMg molar ratioThermodynamic data on chrysotile and p-antigorite are from the thermocomV8R6
database and Gunnarsson et al (2005) respectively
Fig 8 Activityplot forthe MgOndashSiO2ndashAl2O3ndashFe2O3(FeO)ndashH2O system analyzed forSiO2-phases and aqueous species at different log PO2 and log[aAl3+(aH+)3] values Springwaters coming from the TarondashCeno Valleys (symbols as in Fig 2) and Voltri Groupultrama1047297tes (symbols as in Fig 5) are shown The greyarea refers to bicarbonate watersfrom theVoltri Group (Mariniand Ottonello2002Bruniet al 2002) Thesepiolite 1047297eldis very narrow or hidden by saponite-Mg (1047297eld limits shown as continuous lines)saponite-Fe3+ndashMg (1047297eldlimits shown as dottedlines) or vermiculite(1047297eldlimits shownas dottedndashdashed lines) The talc stability 1047297eld and boundaries with saponite-Mg
montmorillonite-Mg and cronstedite-Mg are shown for comparison (dashed lines)
87T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
relative uncertainty of plusmn2 (see Fig 8 in Gunnarsson et al 2005) arecomparable to the logK variation reported for other low-temperatureMg-serpentines (Table 4) that is within the lines Sr1ndash4 for serpentineminerals where samples varying from bicarbonate composition toalkaline composition in the ultrama1047297c aquifer are grouped (Fig 6)Fig 7 shows a plot of calculated serpentine saturation index as a
function of the SiMg molar ratio for the water samples from theultrama1047297tes of the TarondashCeno ValleysThe SI andSiMgvalues of pureMg-serpentine (SI= 0 SiMg=067) provide a useful index forseparating bicarbonate waters (SIb0 SiMgb067) from alkalinewaters (SIN0 SiMgN067) and the distribution of data points isconsistent with the idea that bicarbonate waters approach saturationwith serpentine until they become supersaturated as alkaline waters
Based on the Mgvs SiO2(aq) activitydiagram and speci1047297cratiosoflog[aAl3+(aH+)3] Birsoy (2002) analyzed the behavior of sepiolite duringultrama1047297te weathering The diagram produces a very narrow stability1047297eldfor sepiolite(Fig 8) when it is programmedwith suitable conditionsfor the bicarbonate waters (log PO2=minus25 bar log [aAl3+(aH+)3]= 65) andthe alkaline waters (log PO2=minus50 bar log [aAl3+(aH+)3]= 45) sampled inthe present study This stability 1047297eld grows larger when the absence of
Mg-smectites and talc is assumed but this is an unrealistic conditionbecause (i) all these minerals occur in the rocks within the areainvestigated and (ii) at high [Mg2+][H+]2 activity ratios the 1047297eld of sepiolite is smaller than that of subordinate compared to those poorly-crystalline serpentine (p-antigorite) and clinochlore
Among the alkaline springs of the Voltri Group the waters withlow log [aCa2+(aH+)2] values (Fig 9) show patterns compatible withchlorite interaction (step 5c in Fig6D)The alkaline waters whose datafall near the clinochlore-14Aring 1047297eld (Fig 6D) cannot be explained bysepiolite precipitation because the sepiolite line does not cross thechlorite 1047297elds Thus the peculiar composition of some alkaline watersof the Voltri Group is probably due to interactions between water andprimary chlorites of the ultrama1047297c rocks (optional step 5c in Table 5cb) This is supported by the occurrence sometimes in large amounts
of metasomatic Mgndash(Fe)-chlorite (Scambelluri and Rampone 1999
Vissers and Strating 2001) in the Ligurian area where the relativeabundance of chlorite can reach 50 in veins in high strainserpentinite mylonites of the ErrondashTobbio ultrama1047297tes (Scambelluriet al 2001b) Otherwise in a chlorite-free primary paragenesis thesolution evolving on the Sr3 serpentine line reaches equilibrium withclinochlore but probably the latter does not precipitate due to theslow kinetics of precipitation under these conditions Thus thedissolution of additional reacting rock in the iterative weathering
process supplies Mg
2+
to the solution thereby favoring the Sr3 to Sr4poorly crystalline serpentine transformation In fact all the PR10samples and some of the alkaline springs from the Voltri Group showincreasing [Mg2+][H+]2 activity ratio and low variation in silicaactivity (Fig 6D) starting from the invariant triple point 4a (Sr3-serpentine Fe2+ndashMg-saponite chlinoclore-14 Aring saturation) andextending towards the brucite line This path can be explained asthe result of the transformation of poorly crystalline serpentine fromSr3 to Sr4 the latter is more stable at higher [Mg2+][H+]2 activityratios Finally at very low PO2 values vermiculite disappears and thestability 1047297elds of chlorite and Fe2+ndashMg-saponite become narrowerwhereas those of Fe2+-rich phases (eg Fe2+-saponite) become widerThe data for the alkaline waters with the highest log [aMg2+(aH+)2]from the Voltri Group fall near the stability 1047297eld of peridotite primaryminerals like forsterite and enstatite (Fig 6D) therefore the last stepsof theevolution involve theequilibration of thesolutionwithenstatite(steps 8a and 8c in Table 5ab)
62 Ca-rich alkaline waters
In preliminary modeling of the evolution of the TarondashCeno springwaters Boschetti (2003) concluded that meteoric water initiallysaturated in Fe3+ndashMg2+-chlorite Fe3+ndashMg2+-saponite and pyrite (andor goethite) and then interacted with the ultrama1047297te mineralassemblage eventually evolved into the alkaline waters In themodel of Boschetti (2003) chlorite saponite and calcite were added
Fig 9 Activity plot for the MgOndashCaOndash(CO2)ndashH2O system with phase limits calculated ata temperature of 11 degCSpring waterscoming from theTarondashCeno Valleys (symbols as inFig2) andVoltri Groupultrama1047297tes(symbolsas in Fig5) areshown Calculated path forwaterndashultrama1047297c rock interaction is shown (dotted line) with the number ndashlettercombinations referring to the appearance or disappearance of minerals (see Table 5)The grey area refers to bicarbonate waters from the Voltri Group (Marini and Ottonello2002)
Fig 10 Plots of (A) calcite saturation indexes and (B) PCO2 as a function of Clminus ascalculated by PHREEQCI software for waters from the TarondashCeno Valleys Alkalinesprings for the Voltri Group are also shown (small black diamonds Marini andOttonello 2002) The grey shading is deeper(A) in the saturation1047297eld(SI= 00plusmn05) and
(B) where the lowest log PCO2 value is approached
88 T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
to the solution in order to achieve a high [Ca2+][H+]2 activity ratio andsupersaturation in these phases However in light of the work of Clarket al (1992) and Bruni et al (2002) we believe that the calcitesupersaturation in alkaline waters (Fig 9 and 10) which in our modelis achieved by preventing the precipitation of calcite in the 1047297rst stepsof the reaction path (Table 5a and b) depends on the differencebetween CO2 partial pressureat thegivendepth(PCO2le10minus10bar) andat the surface (PCO2cong10minus35) Waters rising to the surface are far from
equilibrium with the atmosphere where the PCO2 is 4times10
6
timesgreater than the value at depth thus the water rapidly take up CO2which is promptly converted to CO3
2minus because of the high pH Thisleads to supersaturation with calcite through the following hydro-xylation reaction
As a result large amounts of travertine are generated in the VoltriGroup (Marini and Ottonello 2002) and carbonate concretionsknown as devilrsquo s coins are formed at Mt Prinzera in Taro valleyBelow the surface the SIcalcite of the alkaline waters is expected to belower than what the model estimates Under conditions of lowcarbonate alkalinity low CO2 and O2 partial pressure and high pHcalcite nucleation may be inhibited and the kinetics of precipitationreduced by the high concentrations of dissolved divalent cations suchas Mg2+ Fe2+ and Mn2+ (Sumner and Grotzinger 1996 Davis et al2000 Zhang and Dawe 2000 de Leeuw 2002) The fact that severalalkaline water samples fall within or near the calcitendashdolomiteequilibrium (Fig 9) may indicate magnesium calcite or protodolomitenucleation in those extreme subsurface conditions However possibleinteraction between some calcium-rich waters of the Voltri Group andCa-rich rocks such as ophicarbonates or rodingites cannot beexcluded these rocks are widespread in the Ligurian Apennineserpentinites (Treves and Harper 1994)
63 Sepiolite vs serpentine nucleation under weathering conditions
constraints on the kinetics and equilibrium-based approaches
The kinetics of precipitation of poorly-crystalline serpentine and
sepiolite cannot be compared quantitatively because rate constantsare available only for well-ordered crystalline phases (eg see Palandriand Kharaka 2004 for crystalline serpentine) Sepiolite precipitationis often invoked to explain magnesium scavenging from solutions butthe lack of kinetic data for this mineral leads to overly simplisticreaction transport models These models are forced to use kinetic datafor phyllosilicates such as kaolinite or muscovite (eg Xu et al 2001Bryan 2005) Moreover the low-temperature thermodynamic dataeven if they are fairly uncertain show that sepiolite solubilityincreases with decreasing crystallinity (Table 4 Stoessell 1988) Infact crystalline sepiolite requires a long time to reach equilibriumwith solution which imposes a slow control on thewater composition(eg Stoessell and Hay 1978) Consequently in modeling based onreaction transport that simulates freshly precipitated or poorly
crystalline sepiolite formation the logK value of Wollast et al (1968)is preferable to the lower values (Suarez and Šimůnek 1996) Thispreference agrees with the approach of Mariner (2003) and Jove Colon(2006) who adopt a logK (25 degC) value for sepiolite of 3044 It is alsosimilar to that reported by Wollast et al (1968) and it is six logK unitshigherthan the valueused in thermocomV8R6 dataset which is oftenused (perhaps improperly) to model the supergenic nucleation of sepiolite In the stability diagram in Fig 6 where the Sp1 and Sp2 linesrepresent the stability boundaries of crystalline and poorly-crystallinesepiolite respectively the evolution of TarondashCeno waters frombicarbonate to alkaline (Figs 6A and C) follows the stability 1047297eldboundary of the poorly-crystalline serpentine (between Sr2ndashSr3 lines)rather than that of the poorly-crystalline sepiolite (Sp2) as postulatedin themodel The lines for Sp2 Sr3 and amorphoussilica intersect at a
common point (Fig 6D) which is consistent with the nucleation of
sepiolite that typically occurs in solutions that are saturated orsupersaturated with amorphous silica and (Mg)-carbonate (egStoessell and Hay 1978 Birsoy 2002) Some bicarbonate watersfrom the Voltri Group have a silica activity near the amorphous silicaboundary (Fig 6B) but only a few alkaline springs of this group fall onthe poorly-crystalline sepiolite boundary line (Sp2 Fig 6D) Both of these features are inconsistent with the possibility that sepioliteprecipitates from supergenic fresh water in the Northern Apennine
Moreover these two results suggest that values for logK(sepiolite)b
3044 (see Table 4) should be used to model the dissolutionprecipitation of the well-crystallized phase that is the most commonform for the metasomatic sepiolite found in the investigated area
7 Conclusions
The waters issuing from ultrama1047297tes in the TarondashCeno Valleysoriginate from low-temperature interactions between meteoric waterand serpentinite rocks In this report the waterndashrock interaction hasbeen modeled using reaction paths in progress mode During their earlyinteraction with soil and rock minerals bicarbonate waters super-saturated in kaolinite ferrihydrite vermiculite and Fe2+ndashMg2+-saponitewhere theorigin of the secondary iron saponite as Fe2+-bearing smectiteagrees with the hypothesis of Guumlven (1988) Then while remainingsupersaturated with saponite the bicarbonate waters evolve into analkaline facies when poorly-crystalline serpentine precipitates Duringrock weathering the Al-bearing crystalline serpentine of oceanic originprobably promotes nucleation of the low-temperature serpentineamorphous phase and supplies Al3+ to the solution for saponiteprecipitation This interpretation is consistent with the existence of Al-poor supergenic serpentine (Neal and Stanger 1985) and the partiallymeteoric OH-isotopic imprint observed in other ultrama1047297c systems
Dif 1047297culties and complexities in the modeling arise fromuncertainties in the primary mineralogy of some local ophiolitesand from the presence of different lithotypes (rodingites ophi-carbonates gabbros) which can signi1047297cantly affect the composi-tion of water samples This can be seen in the example of the Candash
OH springs issuing from the Voltri Group (PO2 and PCO2 of 10minus70
and 10minus10 bar respectively Bruni et al 2002 Marini and
Ottonello 2002) These springs appear to be more evolved thanthe TarondashCeno alkaline waters as a result of their particularly longinteractions with serpentinites until equilibrium with enstatite isreached or with chlorite-bearing serpentinite The use of anupdated thermodynamic database in our model avoided problemsdue to mineralogical heterogeneity in the mixed-layer claysassociated with ophiolitic rocks this database accurately re1047298ectedthe chemical composition of the single phases Using this modelthe vermiculitendashchlorite and saponitendashchlorite associations seen inthe activity diagrams simulate respectively the presence of highand low charge corrensite observed in the 1047297eld The accuracy of ourmodel can be tested by comparing estimates of chlorine composi-tion resulting from our model and from experimental methods The
mean Cl concentration of the bicarbonate water is 41plusmn16 mgLtaking into account a Cl concentration of sim2 mgL in rainwater inthe area (cf Panettiere et al 2000 Venturelli 2003) 2 mgL of Clin spring water are estimated to come from serpentinite This valueagrees with the model which predicts that the switch frombicarbonate water to alkaline water requires the congruentdissolution of 00203 mol of serpentinite containing 100 ppm of soluble Cl thereby yielding water containing 2 ppm of Cl
In addition we think that weathering of the accessory albiteplagioclase is responsible for Na enrichment of the waters whereas Clenrichment in the alkaline waters is predicted to result from theleaching of NaCl 1047298uid inclusions Moreover the TarondashCeno alkalinesprings show the highest BCl ratios comparable with those of thehost metamorphic minerals (Scambelluri et al 2004) In contrast the
lower boron content and BCl ratio of the alkaline springs of the Voltri
89T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
Group are unexpected features and probably re1047298ect waterndashrockinteractions over a longer time period than in the TarondashCeno ValleysThe boron content in the Northern Apennine alkaline springs may becontrolled by several factors such as (i) the different histories of VoltriGroup and EL ultrama1047297tes (ii) the probable contribution of boronfrom NaCl 1047298uid inclusions (iii) the sinking of boron by travertineprecipitation (Hobbs and Reardon1999) brucite (Rhoades et al 1970)and smectite and (iv) decreases in boron release from minerals due to
increases in pH (Su and Suarez 2004) Acknowledgements
This research was carried out with the 1047297nancial support from theldquoYoung Researchers Projectrdquo (2001ndash2002) of the Italian Ministry of Universities Scienti1047297c Research and Technology (MURST) We thankGianniCortecci (Institute of Geosciences and EarthResources IGG-CNRPisa Italy) and Giampiero Venturelli (Earth Sciences Department Univer-sity of Parma Italy) for useful suggestions and critical review of themanuscript The authors are grateful to Enrico Dinelli (InterdepartmentResearch Centre for Environmental Sciences Alma Mater StudiorumUniversity of Bologna Ravenna Italy) for granting permission to use theGeochemistrsquos WorkbenchsoftwareTheconstructive reviews of theEditor
J Fein and of two anonymous Referees have signi1047297cantly improved the
manuscript
Appendix A Supplementary data
Supplementary data associated with this article can be found inthe online version at doi101016jchemgeo200808017
References
Abbate E Bortolotti V Passerini P 1980 Olistostromes and olistoliths In Sestini G(Ed) Development of the Northern Apennines geosyncline Sedimentary Geology4 pp 521ndash558
Alietti A 1956 Il minerale a strati misti saponite-talco di monte Chiaro (val di TaroAppen Emil) Rendiconti dellAccademia Nazionale dei Lincei 21 201 ndash207
Alietti A De Angelis G Montanari G Sgarlata F Tirelli G 1976 Il ferro nelle argilleappenniniche Periodico di Mineralogia 45 51ndash64
Anselmi B Mellini M Viti C 2000 Chlorine in the Elba Monti Livornesi and Murloserpentines Evidence for seandashwater interaction European Journal of Mineralogy12137ndash146
Apha-Awwa-Wef 1995 Standard Methods for the Examination of Water and Waste-water 19th Edition American Public Health Association American Water WorksAssociation Water Environment Federation
Barnes I LaMarche Jr VC Himmelberg GR 1967 Geochemical evidence of present-day serpentinization Science 156 830ndash832
Barnes I ONeil JR 1969 The relationship between 1047298uids in some fresh alpine-typeultrama1047297cs and possible modern serpentinization Western United StatesGeological Society of America Bulletin 80 1947ndash1960
Barnes I ONeil JRTrescases JJ1978 Present dayserpentinizationin New CaledoniaOman and Yugoslavia Geochimica et Cosmochimica Acta 42 144ndash145
Barnes I Rapp JB ONeil JR Sheppard RA Gude III AJ 1972 Metamorphicassemblages and the direction of 1047298ow of metamorphic 1047298uids in four instances of serpentinization Contributions to Mineralogy and Petrology 35 263ndash276
Barnes JDSelverstone J Sharp ZD 2006Chlorine isotope chemistry of serpentinitesfrom Elba Italy as an indicator of 1047298uid source and subsequent tectonic historyGeochemistry Geophysics Geosystems 7 doi1010292006GC001296
Beccaluva L Emiliani F Venturelli G Zerbi M 1973 Ca Fe Mg Mn Cr Ni Codistribution in some ultrama1047297c rocks outcropping in the Northern Apennines withsome geological remarks Ateneo Parmense Acta Naturalia 9 69ndash98
Beccaluva L Macciotta G Piccardo GB Zeda O 1984 Petrology of lherzolitic rocksfrom the Northern Apennine ophiolites Lithos 17 299 ndash316
BeccaluvaL Venturelli G ZanzucchiG 1975 Dati geologici e geochimici sui basalti diassociazione o1047297olitica dellAppennino ligure-emiliano Ateneo Parmense ActaNaturalia 11 789ndash802
BethkeCM 2002 The Geochemists Workbench A Userrsquos Guide to RxnAct2Tact ReactandGtplot Release40 HydrogeologyProgram Universityof IllinoisUrbanaIL 236pp
Birsoy R 2002Formation of sepiolitendashpalygorskite and related minerals from solutionClays and Clay Minerals 50 736ndash745
Bonatti E Lawrence JR Morandi N 1984 Serpentinization of oceanic peridotitestemperature dependence of mineralogy and boron content Earth and PlanetaryScience Letters 70 88ndash94
Bonifacio E Zanini E Boero V Franchini-Angela M 1997 Pedogenesis in a soilcatena on serpentinite in north-western Italy Geoderma 75 33ndash51
Boschetti T 2003 Studio geochimico e geochimico-isotopico di acque a composizione
estrema e termali dellrsquoAppennino Settentrionale Ph D Thesis Parma Univ
Brigatti MF Franchini G Lugli C Medici L Poppi L Turci E 2000 Interactionsbetween aqueous chromium solutions and layer silicatesApplied Geochemistry 151307ndash1316
Brigatti MF Lugli C Poppi L Venturelli G 1999 Iron-rich saponite dissolutionreactions and Cr uptake Clay Minerals 34 637ndash645
Brigatti MF Valdregrave G 1996 Mixed-layer chloritesmectite (corrensite) in ophioliticrock alteration products Neues Jahrbuch fuer Mineralogie Monatshefte 1 37 ndash47
Bruni J Canepa M Chiodini G Cioni R Cipolli F Longinelli A Marini L OttonelloG Vetuschi Zuccolini M 2002 Irreversible waterndashrock mass transfer accompany-ing the generation of the neutral MgndashHCO3 and high-pH CandashOH spring waters of the Genova province Italy Applied Geochemistry 17 455ndash474
Bryan CR 2005 Drift-Scale THC Seepage Model MDL-NBS-HS-000001 REV 04 YuccaMountain Project Las Vegas Nevada (US)Caruso LJ Chernosky JV 1979 The stability of lizardite Canadian Mineralogist 17
757ndash769Clark ID Fontes JC Fritz P 1992 Stable isotope disequilibria in travertine from high
pH waters Laboratory investigations and 1047297eld observations from Oman Geochi-mica et Cosmochimica Acta 56 2041ndash2050
Cortesogno L 1980 Il metamor1047297smo giurassico nelle o1047297oliti dellAppenninoSettentrionale due distinti cicli metamor1047297ci in ambiente oceanico O1047297oliti 5 5ndash18
Christ CLHostetler PB Siebert RM1973 Studies in the system MgO-SiO2-CO2-H2OIII American Journal of Science 273 65ndash83
Cortesogno L Lucchetti G 1982 Il metamor1047297smo oceanico nei gabbri o1047297oliticidellAppennino Ligure Aspetti mineralogici e petrogenetici Rendiconti dellaSocietagrave Mineralogica Italiana 38 561ndash579
Crispini L Capponi G 2002 Control of metasomatic alteration on the development of shear zone in ultrama1047297c rocks a case study from the Voltri Massif (Ligurian Alps)Architettura interna e cinematica delle zone di taglio Gruppo Italiano di GeologiaStrutturale mdash Riunione Annuale 2002 Pisa Italy 11ndash12 giugno 2002 XIV ndashXV
CrissRE 1999 Principlesof stableisotope distributionOxford University PressNew York
Davis KJ Dove PM De Yoreo JJ 2000 The role of Mg 2+ as an impurity in calcitegrowth Science 290 1134ndash1137
de Leeuw NH 2002 Moleculardynamics simulations of the growth inhibiting effect of Fe2+ Mg2+ Cd2+ and Sr2+ on calcite crystal growth Journal of Physical Chemistry B106 5241ndash5249
De Nardo MT Segadelli S Vescovi P 2007 Studio pilota per la caratterizzazionegeologicadellesorgentinellazona delM Nero (altaVal Cenoe alta ValNuremdash provincedi Parma e Piacenza) Il Geologo dellEmilia-Romagna 25 Nuova Serie pp 5 ndash21
Di Giulio A Geddo G 1990 Studio petrogra1047297co delle Arenarie di Casanova (alta valTrebbia Appennino settentrionale) Atti Ticinesi di Scienze della Terra 33 243ndash254
Dinelli E Lombini A Simoni A Ferrari C 1997 Heavy metals in serpentinite soils of selected outcrops of Piacenza and Parma provinces (Northern Apennines Italy)Mineralogica et Petrographica Acta 40 241ndash255
Epstein S Mayeda T 1953 Variation of 18O content of waters from natural sourcesGeochimica et Cosmochimica Acta 4 213ndash224
Ernst WG Piccardo GB 1979 Petrogenesis of some Ligurian peridotites I Mineraland bulk-rock chemistry Geochimica et Cosmochimica Acta 43 219ndash237
Feder F Trolard F Klingelhofer G Bourrieacute G 2005 In situ Moumlssbauer spectroscopy
evidence for green rust (fougerite) in a gleysol and its mineralogical transforma-tions with time and depth Geochimica et Cosmochimica Acta 69 4463ndash4483Fritz PClark IDFontes JCWhiticar MJFaberE 1992 Deuterium and
13C evidence
for low temperature production of hydrogen and methane in a highly alkalinegroundwater environment in Oman In Kharaka TK Maest AS (Eds) Proceed-ings of the 7th International Symposium on WaterndashRock Interaction BalkemaRotterdam pp 793ndash796
Giammetti F 1968 Le o1047297oliti di Roccaprebalza Costa della Guardia e Groppo di Gorro(Appennino Parmense) Ateneo Parmense Acta Naturalia 4 107ndash142
Giuseppetti G1953 Roccee minerali dellaformazioneo1047297olitica di Volpedo Rendicontidella Societagrave Mineralogica Italiana 9 85ndash134
Gran G 1952 Determination of equivalence point in the potentiometric titrationAnalyst 77 661ndash671
Gunnarsson I Arnorsson S Jakobsson S 2005 Precipitation of poorly crystalline antigoriteunder hydrothermal conditions Geochimica et Cosmochimica Acta 69 2813ndash2828
Guumlven N 1988 Smetites In Bailey SW (Ed) Reviews in Mineralogy 19 Hydrousphyllosilicates (exclusive of micas) Mineralogical Society of America MichiganBook Crafters Inc 497ndash559
Hansel CM Benner SG Fendorf S 2005 Competing Fe(II)-induced mineralization
pathways of ferrihydrite Environmental Science and Technology 39 7147ndash7153HobbsMYReardonEJ 1999 Effect of pH on boronco-precipitation by calcite further
evidence for non-equilibrium partitioning of trace elements Geochimica etCosmochimica Acta 63 1013ndash1021
Jones BF Galan E 1988 Sepiolite and palygorskite In Bailey SW (Ed) HydrousPhyllosilicates (exclusive of Micas) Reviews in Mineralogy vol 19 MineralogicalSociety of America Michigan Book Crafters Inc 631ndash674
Jove Colon CF 2006 Technical work plan for Thermodynamic database for chemicalmodelling TWP-MGR-PA-000039 REV 01 Yucca Mountain Project Las VegasNevada (US)
Kohls DW Rodda Jr JL 1967 Iowaite a new hydrous magnesium hydroxidendashferricoxychloride from the Precambrian of Iowa American Mineralogist 52 1261 ndash1271
Kyser TK OHanley DS Wicks FJ 1999 The origin of 1047298uids associated withserpentinization processes evidence from stable-isotope compositions CanadianMineralogist 37 223ndash237
Lapham DM 1961 New data on deweylite American Mineralogist 46 168ndash188Lee BD SearsSKGraham RC Amrhein CValiH 2003 Secondarymineral genesis
from chlorite and serpentine in an ultrama1047297c soil toposequence Soil Science
Society of America Journal 671309ndash1317
90 T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
Longinelli A Selmo E 2003 Isotopic composition of precipitation in Italy a 1047297rstoverall map Journal of Hydrology 270 75ndash88
Majzlan J Navrotsky A Schwertmann U 2004 Thermodynamics of iron oxides PartIIIEnthalpiesof formation and stability of ferrihydrite (simFe(OH)3) schwertmannite(simFeO(OH)34(SO4)18) and ɛ-Fe2O3 Geochimica et Cosmochimica Acta 681049ndash1059
Mariner P 2003 In-Drift PrecipitatesSalts Model ANL-EBS-MD-000045 REV 01 YuccaMountain Project Las Vegas Nevada (US)
Marini L Ottonello G 2002 Atlante degli acquiferi della Liguria Volume III Le acquedeicomplessio1047297olitici (bacini ArrestraBranega Cassinelle Cerusa ErroGorzenteLeira Lemme Lerone Orba Piota Polcevera Rumaro Sansobbia Stura Teiro
Varenna Visone) Pacini PisaMellini L Zanazzi PF 1987 Crystal structures of lizardite-l T and lizardite-2H1 fromColi Italy American Mineralogist 72 943ndash948
Mezzadri G 1964 Petrogra1047297a delle arenarie di Ostia Rendiconti Societagrave MineralogicaItaliana 20 192ndash228
Mondesir H Decarreau A 1987 Synthesegrave entre 25 et 200 degC de lizardites Ni ndashMgMesure des coef 1047297cients de partage solide-solution aqueuse pour le couple Ni ndashMgdans des lizardites Bulletin de Mineralogie 110 409ndash426
Neal C Shand P 2002 Spring and surface water quality of the Cyprus ophiolitesHydrology and Earth System Sciences 6 797ndash817
Neal C Stanger G 1983 Hydrogen generation from mantle source rocks in OmanEarth and Planetary Science Letters 66 315ndash320
Neal C Stanger G 1984 Calcium and magnesium hydroxide precipitation fromalkaline groundwaters in Oman and their signi1047297cance to the process of serpentinization Mineralogical Magazine 48 237ndash241
Neal C Stanger G 1985 Past and present serpentinisation of ultrama1047297c rocks anexample from Semail ophiolite nappe of Northern Oman In Drever JI (Ed) TheChemistry of Weathering D Reidel Publishing Company pp 249ndash275
Nesbitt HW Bricker OP 1978 Low temperature alteration processes affecting
et Cosmochimica Acta 41 1835ndash1841OrsquoHanley DS 1996 Serpentinites records of tectonic and petrological history Oxford
University PressOrberger B Friedrich G Woermann E 1990 The distribution of halogens and carbon
in Pge-bearing ultrama1047297cs of the Acoje Ophiolite Block Zambales Philippines Journal of Geochemical Exploration 37 147ndash169
Ottonello G Joron JL Piccardo GB 1984 Rare Earth and 3d transition elementsgeochemistry of peridotitic rocks II Ligurian peridotites and associated basalts Journal of Petrology 25 373ndash393
Palandri JL Kharaka YK 2004 A compilation of rate parameters of water-mineralinteraction kinetics for application to geochemical modeling Open File Report2004-1068 US Geological Survey
Panettiere P Cortecci G Dinelli E Bencini A Guidi M 2000 Chemistry and sulfurisotopic composition of precipitation at Bologna Italy Applied Geochemistry 151455ndash1467
Parkhurst DL Appelo CAJ 1999 Userrsquos guide to PHREEQC (version 2) mdash a computerprogram for speciation batch-reaction one-dimensional transport and inversegeochemical calculation US Department of the Interior US Geological SurveyWashington DC
Pfeifer HR 1977 A model for 1047298uids in metamorphosed ultrama1047297c rocks Observationsat surface and subsurface conditions (high pH spring waters) SchweizerischeMineralogische und Petrographische Mitteilungen 57 361ndash396
PlummerLN1984Geochemical modeling A comparisonof forwardand inverse methodsIn Hitchon B Wallick EI (Eds) Proceedings First CanadianAmerican Conference onHydrogeology-Practical Applications of Ground Water Geochemistry Banff AlbertaCanada mdash Worthington Ohio National Water Well Association pp 149ndash177
Rampone E Hofmann AW Piccardo GB Vannucci R Bottazzi P Ottolini L 1995Petrology mineral and isotope geochemistry of the External Ligurides peridotites(Northern Apennines Italy) Journal of Petrology 36 81ndash105
Rampone E Hofmann AW Piccardo GB Vannucci R Bottazzi P Ottolini L 1996Trace element and isotope geochemistry of depleted peridotites from an N-MORBtype ophiolite (Internal Liguride N Italy) Contributions to Mineralogy andPetrology 123 61ndash76
Rhoades JD Ingvalson RD HatcherJT1970Adsorption of boronby ferromagnesianminerals and magnesium hydroxide Soil Science Society of America Journal 34
938ndash941Saini-EidukatB KuchaH Keppler H1994Hibbingiteγ-Fe2(OH)3Clanewmineralfromthe Duluth Complex Minnesota with implications for the oxidation of Fe-bearingcompounds and the transport of metals American Mineralogist 79 555ndash561
Sanford RF 1981 Mineralogical and chemical effects of hydration reactions andapplications to serpentinization American Mineralogist 66 290ndash297
Scambelluri M Muumlntener O Ottolini L Pettke TT Vannucci R 2004 The fate of BCl and Li in the subducted oceanic mantle and in the antigorite breakdown 1047298uidsEarth and Planetary Science Letters 222 217ndash234
Scambelluri M Rampone E 1999 Mg-metasomatism of oceanic gabbros and itscontrol on Ti-clinohumite formation during eclogitization Contributions toMineralogy and Petrology 135 1ndash17
Scambelluri M Bottazzi P Trommsdorff V Vannucci R Hermann J Gogravemez-Pugnaire MT Logravepez-Sagravenchez Vizcaino V 2001a Incompatible element-rich1047298uidsreleased by antigorite breakdown in deeply subducted mantle Earth and PlanetaryScience Letters 192 457ndash470
Scambelluri M Rampone E Piccardo G 2001b Fluid and element cycling insubducted serpentinite a trace-element study of the ErrondashTobbio high-pressureultrama1047297tes (Western Alps NW Italy) Journal of Petrology 42 55ndash67
Segadelli S2006 Lageologia nelpaesaggio le rupi o1047297olitiche in Val Taro e Val Ceno IlGeologo dellrsquoEmilia-Romagna 22 15ndash29 Nuova Serie
Settembre Blundo D Martin De Vidales JL Alietti A 1992 Corrensite-type clayminerals from Taro Valley (North Italy) Mineralogica et Petrographica Acta 35215ndash222
Sharp ZD Barnes JD 2004 Water-soluble chlorides in massive sea1047298oor serpenti-nites a source of chloride in subduction zones Earth and Planetary Science Letters226 243ndash254
Siever R Woodford N 1979 Dissolution kinetics and the weathering of ma1047297cminerals Geochimica et Cosmochimica Acta 43 717ndash724Singer A 1989 Palygorskite and sepiolite group minerals In Dixon JB Weed SB
(Eds) Minerals in Soil Environments Soil Science Society of America MadisonWisconsin pp 829ndash872
Siffert B Wey R 1962 Synthese d rsquoune sepiolite a temperature ordinarie Comptesrendus hebdomadaires des seacuteances de lAcadeacutemie des Sciences de Paris 2541460ndash1463
Środoń J 1999 Nature of mixed-layer clays and mechanisms of their formation andalteration Annual Reviews of Earth and Planetary Science 27 19ndash53
Stoessell RK 1988 25 degC and 1 atm dissolution experiments of sepiolite and keroliteGeochimica et Cosmochimica Acta 52 365ndash374
Stoessell RK Hay RL 1978 The geochemical origin of Sepiolite and keorilite atAmboseli Kenya Contributions to Mineralogy and Petrology 65 255ndash267
Suarez DL Šimůnek J 1996 Solute transport modeling under variably saturatedwater 1047298ow conditions In Lichtner PC Steefel CI Oelkers EH (Eds) ReactiveTransport in Porous Media Reviews in Mineralogy vol 34 Mineralogical Society of America Michigan Book Crafters Inc 229ndash268
Sumner DY Grotzinger JP 1996 Were kinetics of Archean calcium carbonate
precipitation related to oxygen concentration Geology 24 119ndash122Su C Suarez DL 2004 Boron release from weathering of illites serpentine shales
and illiticpalygorskitic soils Soil Science Society of America Journal 68 96 ndash105Toscani L Venturelli G Boschetti T 2001 Sul1047297de-free and sul1047297de-bearing waters in
the Northern Apennines Italy general features and waterndashrock interaction AquaticGeochemistry 7 195ndash216
Treves BE Harper GD 1994 Exposure of serpentinites on the ocean 1047298oor sequenceof faulting and hydrofracturing in the Northern Apennine ophicalcites O1047297oliti 19b435ndash466
Varadachari C Kudrat M Gosh K 1994 Evaluation of standard free energies of formation of clay minerals by an improved regression method Clays and ClayMinerals 42 298ndash307
Veniale F 1966 Sepiolite in sedimenti dellrsquoAppennino Vogherese (Pavia) Periodico diMineralogia 35 343ndash386
Venturelli G 2003 Acque minerali ed ambiente mdash Fondamenti di geochimica deiprocessi di bassa temperatura Pitagora Editrice Bologna
Venturelli G Contini S Bonazzi A Mangia A 1997 Weathering of ultrama1047297c rocksand element mobility at Mt Prinzera Northern Apennines Italy MineralogicalMagazine 61 765ndash778
Venturelli G Frey M 1977 Anchizone metamorphism in sedimentary sequences of the Northern Apennines Rendiconti Societagrave Mineralogica Italiana 33 109ndash123
Vieillard P 2002 A new method for the prediction of Gibbs free energies of formationofphyllosilicates (10 Aring and14 Aring)basedon theelectronegativityscale Claysand ClayMinerals 50 352ndash363
Vissers RLM Strating EHH 2001 Structures and microstructures in a thrust-relatedgreenschist facies tectonic melange Voltri Group (NW Italy) O1047297oliti 26 33ndash46
Viti C Mellini M 1997 Contrasting chemical compositions in associated lizardite andchrysotile in veins from Elba Italy European Journal of Mineralogy 9 585 ndash596
Viti C Mellini M 1998 Mesh textures and bastites in the Elba retrogradeserpentinites European Journal of Mineralogy 10 1341ndash1359
Wenner DB Taylor Jr HP 1974 Deuteriumatomic hydrogen and oxygen-18oxygen-16 studies of serpentinization of ultrama1047297c rocks Geochimica et CosmochimicaActa 38 1255ndash1286
Williams LA Parks GA Crerar DA 1985 Silica diagenesis I Solubility controls Journal of S edimentary Petrology 55 301ndash311
Wilson J Savage D Cuadros J Shibata M RagnarsdottirKV 2006 The effect of ironon montmorillonite stability (I) Background and thermodynamic considerations
Geochimica et Cosmochimica Acta 70 306ndash322WollastR MackenzieFTBricker OP1968 Experimental precipitation and genesis of sepiolite at earth-surface conditions American Mineralogist 531 1645ndash1662
Wrucke CT1996 Serpentine and carbonated-hosted asbestos deposits InEdwardAdu Bray (Eds) Preliminary compilation of descriptive geoenvironmental mineraldeposit models 95ndash831 US Geological Survey Open-File Report pp 39 ndash46
Xu T Apps JA Pruess K 2001 Analysis of mineral trapping for CO2 disposal in deepaquifers Lawrence Berkeley National Laboratory University of California
ZhangYDaweRA2000 In1047298uence ofMg2+ on thekineticsof calcite precipitation andcalcite crystal morphology Chemical Geology 163 129ndash138
Zhu C Anderson G 2002 Environmental applications of geochemical modelingCambridge University Press
91T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
were collected using an automatic equilibration device (Finnigan GLF1086) in-line with a mass spectrometer (Finnigan Delta Plus) Theanalyzed H2 and CO2 gases were equilibrated with water samples at atemperature of 18 degC (Epstein and Mayeda 1953) using a Pt-richcatalyst Isotopic ratio are reported in per mill relative to V-SMOWand standard errors for δ
2H(H2O) and δ18O(H2O) are approximately plusmn
10permil and within plusmn01permil respectively
4 Chemical and isotopic results
The chemical compositionsof thewaters from theultrama1047297tesandbasalts are summarized in Table 1 Complete chemical and isotopedata are reported in Appendix A where all the water samples aresubdivided into four classes on the basis of the lithology present at thesource (i) ultrama1047297c (ii) basaltic (iii) sedimentary and (iv) surfacewaters
41 Chemical classi 1047297cation
All the samples investigated were freshwater (TDSb1 gL) Theternary classi1047297cation diagrams (Na+ K)ndashCandashMg and tAlkndashSO4ndashCl(Fig 2) show that waters of groups (ii) (iii) and (iv) do not differsigni1047297cantly many are bicarbonate-calcic with neutral to slightlybasic pH values (667ndash857) highly oxidized (Eh=371ndash514 mV) andcontain a small amount of chloride (Clminus up to 64 mgL) Sample UM1
from basalt is an exception since it is a ClndashCandashNa type with EC up to1043 μ S cmminus1 Clminus up to 231 mgL hardness up to 367 mgL CaCO3Waters issuing from ultrama1047297tes have more varied compositions Theyrange from Ca-bicarbonate with a pH of 73ndash88 and a hardness rangingfrommoderately hard to very hard (67ndash386mgL CaCO3 Eh129ndash581μ Scmminus1) to Mg-bicarbonate with a pH of 74ndash88 and a hardness from softto very hard (59ndash206 mgL CaCO3 92ndash330 μ S cmminus1) Among the Mg-bicarbonate waters sample UM8 shows the highest hardness values
Fig 2 MgndashSindashC ternary diagram (in molkg) and ternary classi1047297cation diagrams [Mgndash(Na+K)ndashCa plus ClndashtAlkndashSO4 in meqL] of alkaline springs in the TarondashCeno Valleys (this work)andthe VoltriGroup(Mariniand Ottonello 2002) Bottled watersfromTarovalleyare also representedRG1 andRG2Rocca Galgana 1990 and1994 respectively(chemical data fromwater bottle label) FN1 FN2 and FN3 Fontenova 1974 1990 and 1994 respectively (chemical data taken fromwater bottle label) MgndashSindashC contentsafter mineral dissolution (whitecrosses) are also represented (see text for details) Note that for dolomite and magnesite the CMg mole ratio refers to the products of the respective reactions CaMg(CO3)2+2H2O+
2CO2= Ca2+
+ Mg2+
+4HCO3
minus
and MgCO3+ H2O+CO2= Mg2+
+2HCO3
minus
79T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
(486 mgL CaCO3) conductivity (877 μ S cmminus1) and alkalinity (618 mgL HCO3) Springs PR10 and UM15 from ultrama1047297tes show an extremecomposition being sodium-hydroxide and sodium-sulfate respectivelytheir pH is high (101ndash112) their Eh is low (down to minus103 mV) theyaresoft (20ndash61 mgL CaCO3 220 to 408 μ S cmminus1) and they contain largeamounts of boron (up to 2294 μ gL and up to 13461 μ gL respectively)and moderate amounts of reduced sulfur and nitrogen species In bothPR10 and UM15 non-carbonate alkalinity predominates over carbonate
alkalinity using PHREEQCI software with the llnldat thermodynamicdatabase (Parkhurst and Appelo 1999) to calculate the composition of alkaline species under thepH Ehand T valuesmeasuredin1047297eld samplePR10 was found to contain 54 hydroxide 23 carbonate 19 borateand 4 other alkalinespecies (Si sulphide and ammonia) while sampleUM15 contained 75borate14 hydroxide10carbonate and 1 otherspecies
The SindashMgndashC diagram (Fig 2) is a useful tool for studying waterndashrockinteractions Dissolution reactions of typical magnesium minerals of ultrama1047297c rocks are represented by a line with CMg=2 (molar ratio)
Fig3 Plot of (A) Naand (B) Mgvs Cl forallwater samples from the TarondashCenoValleys including bottledwaters Rocca Galgana andFontenova(chemical data taken from labels)Alkalinesprings from the Voltri Group (smallblack diamonds Marini and Ottonello 2002) and meteoric water mean composition (MW Panettiere et al 2000 Venturelli 2003) are also plottedDottedcurvesin(A)and(B)refertomixingwith04and2wtNaCl 1047298uidinclusionsand MgCl2dissolution (1 mol)respectively Greycrosses andi csescorrespondrespectively to mixingwith04and2wtNaCl 1047298uidinclusionsfor mixing fractionsranging from 0001 to0005The solid linerefers tomixingof meteoric water(A) orMg-bicarbonate water(B) withseawater
The vertical grey line represents the water-soluble Cl content in the serpentinites from Northern Apennine (Barnes et al 2006) Symbols are as in Fig 2
80 T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
The Mg-bicarbonate watersissuingfrom ultrama1047297crocksfallnearthisline and cluster between the clinochlore and forsterite points (Fig 2)Stream waters and springs issuing from sedimentary and basaltic rocksfall between thecarbon apex and theline indicated by a CMg molar ratioof 4 which describes the dissolution of CandashMg-clinopyroxene
This is consistent with the (Ca)-bicarbonate compositions of thesestream waters and springs It is noteworthy that the Ceno streamwaters shifted toward the C apex of Fig 2 derive their chemistry fromthe interaction with Mt Caio Flysch cropping out widely in the Cenocatchment (samples from UM50 to UM59 see Map1 sheet in theannexes) whereas the Taro source spring waters plot towards theother apexes (lower C) typical of ophiolite minerals (samples fromUM41 to UM49 see Map1 sheet in the annexes) The use of the traceelements (eg Sr and Rb content Boschetti 2003) further highlightthe differences between the stream waters of the two rivers and therole of lithology Nonetheless mixing and equilibrium with atmo-sphere make the reaction-path modeling of these samples poorlyreliable so their are unsuitable for the purpose of the present work
Alkaline waters from ultrama1047297c rocks fall inside the same sub-triangle mostly along the CndashSi side probably due to the precipitationof magnesium minerals during the evolution of the water composition(Bruni et al 2002)
42 Sodiumndashchloride relationships
Chlorine and other halogens in serpentinized rocks are inheritedfrom seawater and marine sediments during oceanic metamorphism(eg Orberger et al 1990) They are transferred into serpentinite byadsorption (Anselmi et al 2000 Sharp and Barnes 2004) entrap-ment in 1047298uid inclusions (Scambelluri et al 2001ab) and substitutionin the structure of low-crystalline iron-hydroxide phases These latterare composed of the isomorphous and chemically similar iowaitehibbingite and FendashCl-oxydroxides (also named green rusts orfougerites Kohls and Rodda 1967 Saini-Eidukat et al 1994 Federet al 2005 and references therein) In the Northern Apennineserpentinite rocks the mean Cl content is sim300 ppm mean values of 221 and 375 ppm were measured respectively in the Erro ndashTobbio(Ligurian Western Alps Scambelluri et al 2004) and in the centralItaly serpentinites (Anselmi et al 2000) but thewater-soluble Cl from
serpentinites is estimated to be than 100 ppm (Barnes et al 2006) Asshown in Fig 2 waters from basalts and ultrama1047297tes of the TarondashCenoValleys show a trend towards the Cl apex whereas in Fig 3 theestimated soluble chlorine in serpentinites agrees with the highestvalue reported in spring waters from ultrama1047297tes (samples fromMarini and Ottonello analyzed in 2002) The Cl content of samplesUM1 and UM36is veryhigh(Fig 3) and is probablydue to pollutionInfact these springs are not far from the A15 highway (UM1) and theTomarlo Pass road (UM36) where salt mixtures comprising NaClMgCl2middot6H2O and probably CaCl2 are widely used for deicing in winterAlkaline water with a Mg-bicarbonate composition from ultrama1047297tesshow the highest and lowest Cl contents respectively Na-rich Mg-bicarbonate waters from basalts (UM6 UM7 Fig 3) and ultrama1047297tescan be explained by the dissolution of albite in basalts (Beccaluva
et al 1975) and in spilitized EL peridotites (Beccaluva et al 1984
Rampone et al 1995) respectively whereas the highest content of both Na and Cl in the alkaline water samples UM15 and PR10 (Fig 3)can be explained by leaching of NaCl 1047298uid inclusions in theserpentinite rocks (04ndash20 wt NaCl Scambelluri et al 2004)
43 Boron and BCl ratio
Compared with unaltered ultrama1047297c rocks serpentinites have
higher content of both boron and chlorine and most authors suggestthat both elements come from seawater (eg Sanford 1981 andreferences therein Bonatti et al 1984) Despite several studies onthese rocks the behavior of Cl and B in solution during waterndashrockinteraction processes is not well known Early investigations on 1047298uidsassociated with serpentinites in the western United States (Barnes andONeil 1969 Barnes et al 1972) revealed metamorphic waters withhighboron levels(up to 28mM) and BCl molar ratiosof 0038ndash013 aswell as Mg-bicarbonate and high-pH waters of meteoric origin withvariable BCl ratios of 0027ndash042 (median 0039) and 00006ndash0017(median 00076) respectively Bicarbonate waters issuing fromophiolitic rocks in Cyprus have BCl ratio in this range (00017ndash
0058 median 00019 Neal and Shand 2002) whereas low boron
Fig 4 Plots of (A) δ18O vs δ2H and (B) δ18O vs altitude for waters from the Taro ndashCenoValleys C-Italy and N-Italy are the meteoric water lines for Central and Northern Italy
respectively (Longinelli and Selmo 2003) Symbols are as in Fig 2
81T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
levels were found in the alkaline springs of meteoric origin in OmanNew Caledonia and ex-Yugoslavia (Barnes et al 1978)
Among the waters analyzed so far in the TarondashCeno Valleys thebicarbonate waters have BCl ratios in the range 00017ndash022 (median0073) whereas the alkaline waters show the highest ratios everreported (BCl= 034ndash33 median 169) The high boron content of upto 13 mgL of the alkaline springs could be related to metamorphic1047298uids released during serpentinite subduction in fact rocks prefer-
entially lose Cl and the BCl ratio of the released 1047298
uids increases withpressure and temperature (Scambelluri et al 2004) Nonetheless thealkaline water samples in our study differ signi1047297cantly from alkalinemetamorphic1047298uids in that they have low salinity high BCl ratios andmeteoric isotope signatures (see below) all these characteristicssuggest an origin in the interaction of meteoric water with high-pressure metamorphic serpentinite This hypothesis is furthersupported by the agreement in BCl ratio between the alkalinesamples we study here and the high BCl ratio of up to 4 previouslyreported for the residual high-pressure mineralogy of ultrama1047297tes(olivine+orthopyroxene Scambelluri et al 2004) In contrast theVoltri Groups Ca-hydroxide waters have mean values of B = 014 mgL and BCl=003 (Marini and Ottonello 2002) consistent with acontribution of boron and chlorine from 1047298uid inclusions In factassuming fractional contributions of 0001ndash0005 from NaCl 1047298uidinclusions with boron content from 4 to 8 mgL ( Fig 3a Scambelluriet al 2004) results in a B content of 0004ndash01 mgL in the solutionThese results suggest that the difference between springs from theVoltri Group and TarondashCeno Valleys is probably related to (i) differentevolution of the 1047298uids and (ii) the different types of host rocks and Pndash
T paths of the Ligurian Western Alps and EL peridotites (eg Beccaluvaet al 1984 Scambelluri et al 2004)
44 Isotopic composition of water
A diagram of δ2H vs δ18O (Fig 4A) shows that the waters sampledfrom the TarondashCeno Valleys are meteoric in origin since they areconsistent with the meteoric water lines of central and northern Italy(Longinelli and Selmo 2003) The altitude vs δ18O(H2O) (Fig 4B)
summarizes the isotopic effects due to temperature altitude andamount of rainfall (eg Criss1999) all of which help to determine theisotopic composition of the water samples The tendency is similar forbicarbonate waters from sedimentary basaltic and ultrama1047297c forma-
tions which may provide insight into the recharge areas alkalinewaters in contrast probably circulate deeper in the ground
5 Water ndashrock interaction model
51 Fundamentals
The reaction path model between meteoric water and serpentinite
was performed using PHREEQCI 2125 code (Parkhurst and Appelo1999) and the Lawrence Livermore National Laboratory thermody-namic database (llnldat as known as thermocomV8R6 ) implementedwith the equilibrium constants listed in Table 2
Computations wereperformed by reaction progress mode ( forward
modeling Plummer 1984 Zhu and Anderson 2002) ie (i) addingreacting rock at each step in the simulation (ii) activating the phase(s)precipitation only when the thermodynamic stability of the mineralparagenesis is well known and consistent with the mineralogy of thearea and (iii) plotting reaction progress on speci1047297c activity diagramsThe same thermodynamic database was used with The GeochemistsWorkbench 403 software (Bethke 2002) to generate Pourbaix andactivity diagrams (Figs 5 6 8 and 9) Modeling was performed usingan approach based on equilibrium rather than reaction transportbecause the data on reaction kinetics and reactive surfaces in thesesolid phases systems are scarce particularly with respect to theactivation energy of clay and mixed-layer minerals at low temperatureand high pH Moreover the hydrogeology and the discontinuity of theaquifers in the study area which are related to the geologicheterogeneity and structural complexity of the Northern Apenninewould make a model based on reaction transport even less credible atthis stage
The modeling in this study was performed on the basis of thefollowing knowledge (1) primary and secondary mineral assem-blages (2) chemical composition of the minerals involved (3) physicalparameters and chemical composition of the starting 1047298uid (4)thermodynamic data on pure phases
52 The solid reactant primary paragenesis
EL peridotites of the TarondashCeno Valleys (NorthernApennines Italy)are spinel lherzolites containing olivine (forsterite) orthopyroxene(enstatite) and clinopyroxene (diopside) (Giammetti 1968 and
Table 2
Thermodynamic data on hydrolysis reactions added to the llnldat database of the PHREEQCI code (Parkhurst and Appelo 1999) and thermocomV8R6dat database of TheGeochemists Workbench software (Bethke 2002)
Mineral Reaction log K (TdegC) ΔGdeg f References
0 25 60 100 (kJ mol-1)
clinochlore08-daphnite02 Mg4Fe1Al2Si3O10(OH)8 + 16H+=2Al3+ + Fe2++12H2O+4Mg2++3SiO2 7318 6384 5272 4269 minus786379 this workferr ihydrite 2-l ine Fe(OH)3+3H+= Fe3++ 3H2O 484 355 ndash ndash minus70850 Majzlan et al 2004ferr ihydrite 6-l ine Fe(OH)3+3H+= Fe3++ 3H2O 435 312 ndash ndash minus71100 Majzlan et al 2004liz09-berth01 Mg27Fe02Al02Si19O5(OH)4+64H+=02Al3++02Fe2++52H2O+27Mg2++19SiO2 3111 2786 2383 2005 minus399854 this workp-antigorite Mg
Saponite-Fe2+ Fe3175Si365Al035O10(OH)2+74H+=035Al3++3175Fe2++47H2O+365SiO2 1845 1635 1319 1001 minus452459 Wilson et al 2006Saponite-Fe2+ndashMg Fe0165Mg3Si367Al033O10(OH)2+732H+=033Al3++0165Fe2++466H2O+3Mg2++
367SiO2
3037 2729 2330 1921 ndash Marini and Ottonello(2002)
The Gibbs free energy of formation (ΔGo f ) of the two theoretical Fe3+ (Varadachari et al1994) and Fe2+ vermiculites were calculated using a previously published method (Vieillard
2002 and Vieillard pers comm)log K values extimated considering an ideal solid solution between the end-members and using data thermodynamic data from llnldat (clinochlore08-daphnite02) and Wilson
et al (2006) (liz09-berth01)
82 T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
references therein Venturelli et al 1997) Accessory phases areplagioclase picotite plusmn calciumndashsodic amphibole (Venturelli et al1997) Before modeling the crystal chemical formula of the mineralphases was calculated from data from Venturelli et al (1997) thendepending on the modal composition of the rock (Ernst and Piccardo1979) and the composition of the ideal solid solution of the minerals
the resulting stoichiometry of the solid reactant(s) was calculated andinserted in the input 1047297le of the PHREEQCI software (Table 3)Thermodynamic data of the primary mineral paragenesis are all wellknown and are included in the llnldat thermodynamic database of PHREEQCI
53 The solid reactant secondary paragenesis
Secondary minerals in the ultrama1047297tes of the TarondashCeno Valleysare serpentine (lizardite and rare chrysotile) smectite minerals (Fendash
Mg saponites) and Fe-oxi-hydroxides Other accessory minerals arechlorite mixed layer saponitendashchlorite (low-charge corrensite) talccalcite and dolomite (Beccaluva et al 1973 Dinelli et al 1997Venturelli et al 1997 Brigatti et al 1999 2000) In addition mixed
layer saponitendashtalc (aliettite) and vermiculitendashchlorite (high-charge
corrensite) are present (Alietti 1956 Settembre Blundo et al 1992Brigatti and Valdregrave 1996) Except for serpentine themodal abundanceof these phases is unknown In addition some of these minerals ndash
particularly interlayer clays ndash may have different origins (hydrothermalor weathering allogenicor authigenic)and different stabilitiesrelatedtokinetics and drainage 1047298ow rate For example chlorite weathering yields
vermiculite-bearing or saponite-bearing interlayers under good- andpoor-drainage regimes respectively (Bonifacio et al 1997 Środoń1999) In addition the thermodynamic data available for serpentineinclude information only on chrysotile which is less abundant thanlizardite in the serpentinites of the areas that we studied According toVenturelli et al (1997) the lizardite composition in the serpentinites of the Taro valley can be summarized as
which is similar to 1-T lizardite from the Monte Fico serpentinizedperidotite (Elba Island Northern Tyrrhenian Sea Italy Viti and Mellini1997)
Using the thermodynamic data of Wilson et al (2006) and
assuming an ideal solid solution between lizardite and berthierine
Fig 5 EhndashpH iron diagram of the spring waters coming from ultrama1047297c rocks of the TarondashCeno Valleys Bicarbonate and high-pH waters of the Voltri Group are also represented(dotted area andsmallblackdiamonds respectively) Thelowest Eh value of thehigh-pH waterswas calculated using theS(+VI)S(minus II)redox couple(Marini and Ottonello 2002)Thestability 1047297eld of mineral phases are in grey and those of aqueous species in white In diagrams A and B 6-line ferrihydrite-vermiculite and goethite were selected as subaerial solidphases respectively
83T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
we calculated ΔGf 0=minus399854 kJmol for a theoretical (09)(lizardite)+
(01)(berthierine) mineral (Table 2) The log K value of the stated solidsolution was calculated at different temperatures using PHREEQCIafter loading the thermodynamic data on the two end-members intothe database The log K results obtained for Al-lizardite were added tothe thermodynamic database of the solid reactant (Table 3) andcompared with data from the literature (Table 4)
54 Composition of the starting solution O 2 partial pressure and
minerals included in the model
The weighted average chemical composition of rainwater inBologna (Panettiere et al 2000) which is 95 km east of theinvestigated area was used as the starting solution in the modelFull chemical data on rainwater of the TarondashCeno Valleys are
unavailable
Total alkalinity was recalculated for log PCO2=minus35 bar and
T =11 degC which was the mean temperature of the springs sampledand for pe (or Eh) based on chemical data for the redox couple N(minusIII)N(+V) The values of pH and pe obtained were 566 and 452(Eh=025 V) respectively the former agrees with the median valuereported for rainwater in thearea (Venturelli et al1997) Because dataon Al Fe and Si in rainwater were lacking we calculated theconcentrations of these elements assuming logPCO2=minus30 bar andslight supersaturation relative to ferrihydrite-6 line and kaolinite Inthis way a hypothetical composition of water in equilibrium with thelocal soil was calculated (Table3) and adopted as the ldquoinitial waterrdquo formodeling
Evidence suggests that secondary minerals particularly Fe-bearingphases change the chemical composition of the water 1047298owingthrough the ultrama1047297tes of the TarondashCeno Valleys According to Neal
and Stanger (1983) and Fritz et al (1992) the production of CH4 and
Fig 6 Activity plotfor the MgOndashSiO2ndashAl2O3ndashFe2O3(FeO)ndashH2O systemat differentlog PO2 values Spring waters issuing from TarondashCeno Valleys (symbols as in Fig 2) andVoltri Groupultrama1047297tes are shown (grey area bicarbonate waters small black diamonds alkaline waters from Marini and Ottonello 2002) In plots A B and C the stability 1047297eld limits of vermiculite(Table 2) correspond to25 degCwhilethe amorphoussilica (am-SiO2) brucite serpentine (Sr1ndash4 in Table 4)and sepiolite(Sp1ndash2 in Table 4) phasescorrespondto 11 degCThepath calculatedfor interaction between waterand ultrama1047297c rockis also shown (dotted line)the number-letter combinations referto the appearanceand disappearance of minerals(see Table 5)
84 T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
H2 in thealkalinewaters1047298owing from theultrama1047297tes isrelatedto theconsumption of oxygen during the oxidation of iron forming thehematitendashmagnetite pair Therefore the activity of iron and theoxidation state of the water must be de1047297ned The analysis of watersamples from ultrama1047297tes were plotted in both the Pourbaix andactivity diagrams (Figs 5 and 6) where the calculated activity of Fe2+Mg2+ SiO2 SO4
2minus (Pourbaix) and log PO2 (activity diagrams) wereinvolved Iron concentration was quite variable in the TarondashCenosprings where the average content was 27 μ gL and the contentreached a maximum of 200 μ gL and in the Voltri springs where theaverage was 44 μ gL and the maximum was 948 μ gL (Marini andOttonello 2002) In both cases the variations are probably related to
the different mineralogies of the soils and the different solubilities of the Fe-rich phases like goethite and ferrihydrite Iron-bearingsaponites and montmorillonites are the prevailing clay minerals inthe area investigated probably because of the ubiquitous presence of iron oxy-hydroxide (Alietti et al 1976 G Venturelli pers comm) Forthis reason the model considers the involvement of pure Mg-smectiteto be negligible
In the Pourbaix diagram (Fig 5) the data from bicarbonate watersfall into the ferrihydrite-goethitendashvermiculite1047297elds and the data fromalkaline waters fall on the boundaries of the vermiculitesaponite andsaponitechlorite Since ferrihydrite is the 1047297rst precipitating Fe-phaseit was used rather than goethite in the modeling The diagram alsosuggests the transformation of ferrihydrite into magnetite during thechange from bicarbonate to alkaline composition but this process
requires a large amount of Fe2+ in solution (N2 mM Hansel et al2005) which is unlikely at this stage of the waterndashrock interactionThus the magnetite participating in the equilibrium is most likelyprimary in origin
The modeling and the activity diagrams (Figs 6 and 9) werecalculated using the average value PO2= 10minus26 or 10minus50 bar forbicarbonate and alkaline waters respectively These values arecompatible with those calculated by Marini and Ottonello (2002) forbicarbonate waters (PO2= 10minus40 bar) and for alkaline waters of theVoltri Group (10minus77 barbPO2b10minus42 bar) The lowest PO2 value whichwas used as a limit in the modeling was calculated from the S(+VI)S(minusII) redox couple Moreover the activity diagrams (Fig 6) indicatethat the data from the bicarbonate waters from the TarondashCeno Valleysfall between quartz whose log [SiO2] =minus433 is not shown because of
the slow precipitation kinetics under the PndashT conditions considered
here (Williams et al 1985) and amorphous silica boundary 1047297elds(Fig 6A) at the same manner bicarbonate waters with higher SiO2-activity from the Voltri Group fall on amorphous silica boundary(Fig 6B) In summary then the bicarbonate waters are supersaturatedin kaolinite ferrihydrite vermiculite and saponite Alkaline watersare also supersaturated with saponite (Fig 6AndashC) and they are alsoequilibrated with primary magnetite and under extreme conditionsprobably pyrite as well
55 Limitations of the model related to kinetics
Model calculations disregard the rates of mineral dissolution and
precipitation which becomes increasingly important as temperature
Table 3
Starting mineralogy (left side) and 1047298uid composition (right side) used in the modeling
Rock composition Fluid composition
Modal of the mineralsin the interacting rock
mol of the minerals usedin the modeling
meteoric water composition Initial 1047298uid composition used in the modeling
Spinel 003 Na 3523Eminus05 3523Eminus055 Spinel Chromite 001 S 3519Eminus05 3519Eminus05
Magnetite 0 Si ndash 1176Eminus06Kaolinite ndash 5258Eminus08
Ferrhydrite-6line ndash 5882Eminus07
The modal percentage of the minerals in a serpentinite from TarondashCeno Valleys (Ernst and Piccardo 1979) was re-calculated assuming an ideal solid solution between pure extremephasesand takingthe modal percentage tobe themass percentage Thelatter was used asthe solid reactant composition in thewaterndashrockinteraction modelPure solid phases werechosenfrom the llnldat thermodynamic database of the PHREEQCI softwareThe weighted average chemical compositionof the meteoric waterfrom Bologna (Panettiere et al 2000)was chosen as the initial 1047298uid composition in the modeling after its equilibration with kaolinite and ferrihydrite (see text for details)
Table 4
Summary of thermodynamic data for the sepiolite (Spn) and serpentine (Sr n) mineralsphases in italics do not refer to the mentioned hydrolysis reactions
ΔGf 0 log K Reference
(kJ molminus1) (at 25 degC)
Sepiolite (Spn) mdash Mg4Si6O15(OH)2middot6H2O+8H+=4Mg2++6SiO2+11H2Ominus925400 3004 Gunnarsson et al (2005)minus925133 3050andash3172 Christ et al (1973)
(Sp2) minus921986 3602andash3756 Wollast et al (1968)
Serpentine (Srn) mdash Mg3Si2O5(OH)4+ 6H
+
=3Mg
2+
+2SiO2+ 5H2ONi-lizardite minus 289324 1680 Mondesir and Decarreau (1987)
Al-lizardite2 minus411003 2544a Caruso and Chernosky (1979) Al-lizardite minus 339854 2786 this work (see text for details)
Mg-lizardite (Sr1) minus404025 3056 Wilson et al (2006)Mg-crysotile (Sr2) minus403702 3112 bthermocomV8R6Mg-lizardite minus402824 3340 Mondesir and Decarreau (1987)p-antigorite (Sr3) minus564239a 3479 Gunnarsson et al (2005)Mg-chrysotile (Sr4) minus406809 3669 Bruni et al (2002)Mg-chrysotileasbestos minus562505a 3888 Pfeifer (1977)Antigorite minus6612864 47719 bthermocomV8R6
The 1047297eld boundaries of the underlined phases shown in Fig 6 correspond to atemperature of 11 degC
a In the ΔG f deg column value calculated using the ΔG f
deg data of the elements from llnldat
and logK value in the reference in the LogK column value calculated using ΔG f deg
(mineral) in the reference and ΔG f deg data of the elements from llnldat
b Thermodynamic database of Lawrence Livermore National Laboratory included in
PHREEQCI and Geochemists Workbench softwares
85T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
decreases The waterndashrock interaction in the TarondashCeno basin iscomplicated by the presence of secondary mixed-layer clay mineralsthe origin of which is quite dif 1047297cult to model because the solidndashstatetransformations and weathering reactions are strictly related tohydrological and climatic conditions as described above An equili-
brium-based reaction path model is a valuable tool for assessing the
most probable reactions and for de1047297ning the product phases that canbe used in a kinetics-based model Since the kinetics of precipitationof Mg-clays at room temperature and pressure are poorly knownrelative reaction rate effects in the reaction path model can be mimedby omitting the stabler phase when a metastable mineral is expected
to precipitate Therefore the precipitation of MgndashFe3+
Abbreviations are as follows Fer-6line 6-line ferrihydrite kaol kaolinite verm vermiculite magn magnetite sap saponite serp serpentine clin clinochlore pyr pyrite calccalcite hydrom hydromagnesite bruc brucite enst enstatite The chromium decrease during the change in water facies from bicarbonate to alkaline was modeled withsupersaturation (precipitation) of eskolaite and thereafter with equilibrium with chromite (Boschetti 2003)- not bound dagger chlorite mineral from solution Dagger chlorite mineral added to the paragenesis
86 T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
montmorillonites vermiculite and saponites are predicted to occurrather than that of Mg-clays Moreover in order to simulate the slowprecipitation of calcite in depth precipitation was not activated atslight supersaturation (SIcalcite=001) during the initial stages of thewaterndashrock interaction in this manner saturation index of calcite wasfree to increase in the following steps and in alkaline waters
6 Discussion of the model results
61 Explanation of pre-modeling choices and post-modeling results
The results of the reaction path modeling are reported in Table 5and shown in Figs 6 and 9 The reaction paths model the dissolutionof a typical local serpentinite (mineralogically constrained as inTable 3) and the sequential precipitation of gibbsite kaoliniteferrihydrite vermiculite Fe2+ndashMg2+-saponite and a poorly-crystal-line serpentine( p-antigorite of Gunnarsson et al2005) The1047297rst stepinvolves the reaction of the ldquoinitial waterrdquo with serpentinite atconstant temperature (11 degC) PO2 (10minus26 bar) and PCO2 (10minus30 bar)this early stage corresponds to bicarbonate groundwater feeding theinvestigated spring system At the stage of primary magnetiteequilibrium and Fe2+ndashMg2+-saponite supersaturation the modelassumes the system to be closed to CO2 and to have a 1047297xed partialpressure of oxygen (PO2= 10minus40 bar) these conditions represent
when water switch from a bicarbonate to an alkaline compositionThe calculated amount of interacting serpentinite needed to obtainthe alkaline springs of the TarondashCeno Valleys (mean PCO2= 10minus7 bar)is approximately 00203 moles (see Table 5)
Bruni et al (2002) explained the origin of the alkaline waters of the Voltri Group as the result of the evolution of meteoricbicarbonate water following the precipitation of sepiolite withabundance similar to that of goethite This might be supported bythe observation that sepiolite or smectite precipitate at lowtemperature from solutions with low Al content when the pH is inthe range 75ndash95 or N95 respectively (Siffert and Wey 1962 Jonesand Galan 1988) but sepiolites form from the weathering of serpentines in arid climate regions whereas sepiolites are rare inhumid climate regions where serpentinite changes to Fe-smectite
(Singer 1989 Wrucke 1996) In the ophiolite rocks of the Northern
Apennine sepiolite is found only sporadically in veins (Giuseppetti1953 Veniale 1966) and is unknown in the ophiolites of the Tarondash
Ceno Valleys whereas it is thought that sepiolite of the Voltri Groupprecipitated from metasomatic 1047298uids during development of shearzones (Crispini and Capponi 2002)
As described above our model explains the transition frombicarbonate water to alkaline water (UM15 and PR10 samples) asthe result of low-temperature precipitation of a poor crystalline
serpentine with the solubility product of p-antigorite (see Sr3 inTable 4 and Fig 6) The model indicates that the bicarbonate waters of the Voltri Group with the highest [Mg2+][H+]2 activity ratios similarlyevolve towards alkaline waters (Fig 6B) The logK value of thesynthetic low temperature serpentine adopted in this model and its
Fig 7 Saturation indices of chrysotile and p-antigorite (Sr2 and Sr3 thermodynamic
data respectively in Table 4) for bicarbonate and alkaline waters coming fromultrama1047297c rocks of the TarondashCeno Valleys as a function of SiMg molar ratioThermodynamic data on chrysotile and p-antigorite are from the thermocomV8R6
database and Gunnarsson et al (2005) respectively
Fig 8 Activityplot forthe MgOndashSiO2ndashAl2O3ndashFe2O3(FeO)ndashH2O system analyzed forSiO2-phases and aqueous species at different log PO2 and log[aAl3+(aH+)3] values Springwaters coming from the TarondashCeno Valleys (symbols as in Fig 2) and Voltri Groupultrama1047297tes (symbols as in Fig 5) are shown The greyarea refers to bicarbonate watersfrom theVoltri Group (Mariniand Ottonello2002Bruniet al 2002) Thesepiolite 1047297eldis very narrow or hidden by saponite-Mg (1047297eld limits shown as continuous lines)saponite-Fe3+ndashMg (1047297eldlimits shown as dottedlines) or vermiculite(1047297eldlimits shownas dottedndashdashed lines) The talc stability 1047297eld and boundaries with saponite-Mg
montmorillonite-Mg and cronstedite-Mg are shown for comparison (dashed lines)
87T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
relative uncertainty of plusmn2 (see Fig 8 in Gunnarsson et al 2005) arecomparable to the logK variation reported for other low-temperatureMg-serpentines (Table 4) that is within the lines Sr1ndash4 for serpentineminerals where samples varying from bicarbonate composition toalkaline composition in the ultrama1047297c aquifer are grouped (Fig 6)Fig 7 shows a plot of calculated serpentine saturation index as a
function of the SiMg molar ratio for the water samples from theultrama1047297tes of the TarondashCeno ValleysThe SI andSiMgvalues of pureMg-serpentine (SI= 0 SiMg=067) provide a useful index forseparating bicarbonate waters (SIb0 SiMgb067) from alkalinewaters (SIN0 SiMgN067) and the distribution of data points isconsistent with the idea that bicarbonate waters approach saturationwith serpentine until they become supersaturated as alkaline waters
Based on the Mgvs SiO2(aq) activitydiagram and speci1047297cratiosoflog[aAl3+(aH+)3] Birsoy (2002) analyzed the behavior of sepiolite duringultrama1047297te weathering The diagram produces a very narrow stability1047297eldfor sepiolite(Fig 8) when it is programmedwith suitable conditionsfor the bicarbonate waters (log PO2=minus25 bar log [aAl3+(aH+)3]= 65) andthe alkaline waters (log PO2=minus50 bar log [aAl3+(aH+)3]= 45) sampled inthe present study This stability 1047297eld grows larger when the absence of
Mg-smectites and talc is assumed but this is an unrealistic conditionbecause (i) all these minerals occur in the rocks within the areainvestigated and (ii) at high [Mg2+][H+]2 activity ratios the 1047297eld of sepiolite is smaller than that of subordinate compared to those poorly-crystalline serpentine (p-antigorite) and clinochlore
Among the alkaline springs of the Voltri Group the waters withlow log [aCa2+(aH+)2] values (Fig 9) show patterns compatible withchlorite interaction (step 5c in Fig6D)The alkaline waters whose datafall near the clinochlore-14Aring 1047297eld (Fig 6D) cannot be explained bysepiolite precipitation because the sepiolite line does not cross thechlorite 1047297elds Thus the peculiar composition of some alkaline watersof the Voltri Group is probably due to interactions between water andprimary chlorites of the ultrama1047297c rocks (optional step 5c in Table 5cb) This is supported by the occurrence sometimes in large amounts
of metasomatic Mgndash(Fe)-chlorite (Scambelluri and Rampone 1999
Vissers and Strating 2001) in the Ligurian area where the relativeabundance of chlorite can reach 50 in veins in high strainserpentinite mylonites of the ErrondashTobbio ultrama1047297tes (Scambelluriet al 2001b) Otherwise in a chlorite-free primary paragenesis thesolution evolving on the Sr3 serpentine line reaches equilibrium withclinochlore but probably the latter does not precipitate due to theslow kinetics of precipitation under these conditions Thus thedissolution of additional reacting rock in the iterative weathering
process supplies Mg
2+
to the solution thereby favoring the Sr3 to Sr4poorly crystalline serpentine transformation In fact all the PR10samples and some of the alkaline springs from the Voltri Group showincreasing [Mg2+][H+]2 activity ratio and low variation in silicaactivity (Fig 6D) starting from the invariant triple point 4a (Sr3-serpentine Fe2+ndashMg-saponite chlinoclore-14 Aring saturation) andextending towards the brucite line This path can be explained asthe result of the transformation of poorly crystalline serpentine fromSr3 to Sr4 the latter is more stable at higher [Mg2+][H+]2 activityratios Finally at very low PO2 values vermiculite disappears and thestability 1047297elds of chlorite and Fe2+ndashMg-saponite become narrowerwhereas those of Fe2+-rich phases (eg Fe2+-saponite) become widerThe data for the alkaline waters with the highest log [aMg2+(aH+)2]from the Voltri Group fall near the stability 1047297eld of peridotite primaryminerals like forsterite and enstatite (Fig 6D) therefore the last stepsof theevolution involve theequilibration of thesolutionwithenstatite(steps 8a and 8c in Table 5ab)
62 Ca-rich alkaline waters
In preliminary modeling of the evolution of the TarondashCeno springwaters Boschetti (2003) concluded that meteoric water initiallysaturated in Fe3+ndashMg2+-chlorite Fe3+ndashMg2+-saponite and pyrite (andor goethite) and then interacted with the ultrama1047297te mineralassemblage eventually evolved into the alkaline waters In themodel of Boschetti (2003) chlorite saponite and calcite were added
Fig 9 Activity plot for the MgOndashCaOndash(CO2)ndashH2O system with phase limits calculated ata temperature of 11 degCSpring waterscoming from theTarondashCeno Valleys (symbols as inFig2) andVoltri Groupultrama1047297tes(symbolsas in Fig5) areshown Calculated path forwaterndashultrama1047297c rock interaction is shown (dotted line) with the number ndashlettercombinations referring to the appearance or disappearance of minerals (see Table 5)The grey area refers to bicarbonate waters from the Voltri Group (Marini and Ottonello2002)
Fig 10 Plots of (A) calcite saturation indexes and (B) PCO2 as a function of Clminus ascalculated by PHREEQCI software for waters from the TarondashCeno Valleys Alkalinesprings for the Voltri Group are also shown (small black diamonds Marini andOttonello 2002) The grey shading is deeper(A) in the saturation1047297eld(SI= 00plusmn05) and
(B) where the lowest log PCO2 value is approached
88 T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
to the solution in order to achieve a high [Ca2+][H+]2 activity ratio andsupersaturation in these phases However in light of the work of Clarket al (1992) and Bruni et al (2002) we believe that the calcitesupersaturation in alkaline waters (Fig 9 and 10) which in our modelis achieved by preventing the precipitation of calcite in the 1047297rst stepsof the reaction path (Table 5a and b) depends on the differencebetween CO2 partial pressureat thegivendepth(PCO2le10minus10bar) andat the surface (PCO2cong10minus35) Waters rising to the surface are far from
equilibrium with the atmosphere where the PCO2 is 4times10
6
timesgreater than the value at depth thus the water rapidly take up CO2which is promptly converted to CO3
2minus because of the high pH Thisleads to supersaturation with calcite through the following hydro-xylation reaction
As a result large amounts of travertine are generated in the VoltriGroup (Marini and Ottonello 2002) and carbonate concretionsknown as devilrsquo s coins are formed at Mt Prinzera in Taro valleyBelow the surface the SIcalcite of the alkaline waters is expected to belower than what the model estimates Under conditions of lowcarbonate alkalinity low CO2 and O2 partial pressure and high pHcalcite nucleation may be inhibited and the kinetics of precipitationreduced by the high concentrations of dissolved divalent cations suchas Mg2+ Fe2+ and Mn2+ (Sumner and Grotzinger 1996 Davis et al2000 Zhang and Dawe 2000 de Leeuw 2002) The fact that severalalkaline water samples fall within or near the calcitendashdolomiteequilibrium (Fig 9) may indicate magnesium calcite or protodolomitenucleation in those extreme subsurface conditions However possibleinteraction between some calcium-rich waters of the Voltri Group andCa-rich rocks such as ophicarbonates or rodingites cannot beexcluded these rocks are widespread in the Ligurian Apennineserpentinites (Treves and Harper 1994)
63 Sepiolite vs serpentine nucleation under weathering conditions
constraints on the kinetics and equilibrium-based approaches
The kinetics of precipitation of poorly-crystalline serpentine and
sepiolite cannot be compared quantitatively because rate constantsare available only for well-ordered crystalline phases (eg see Palandriand Kharaka 2004 for crystalline serpentine) Sepiolite precipitationis often invoked to explain magnesium scavenging from solutions butthe lack of kinetic data for this mineral leads to overly simplisticreaction transport models These models are forced to use kinetic datafor phyllosilicates such as kaolinite or muscovite (eg Xu et al 2001Bryan 2005) Moreover the low-temperature thermodynamic dataeven if they are fairly uncertain show that sepiolite solubilityincreases with decreasing crystallinity (Table 4 Stoessell 1988) Infact crystalline sepiolite requires a long time to reach equilibriumwith solution which imposes a slow control on thewater composition(eg Stoessell and Hay 1978) Consequently in modeling based onreaction transport that simulates freshly precipitated or poorly
crystalline sepiolite formation the logK value of Wollast et al (1968)is preferable to the lower values (Suarez and Šimůnek 1996) Thispreference agrees with the approach of Mariner (2003) and Jove Colon(2006) who adopt a logK (25 degC) value for sepiolite of 3044 It is alsosimilar to that reported by Wollast et al (1968) and it is six logK unitshigherthan the valueused in thermocomV8R6 dataset which is oftenused (perhaps improperly) to model the supergenic nucleation of sepiolite In the stability diagram in Fig 6 where the Sp1 and Sp2 linesrepresent the stability boundaries of crystalline and poorly-crystallinesepiolite respectively the evolution of TarondashCeno waters frombicarbonate to alkaline (Figs 6A and C) follows the stability 1047297eldboundary of the poorly-crystalline serpentine (between Sr2ndashSr3 lines)rather than that of the poorly-crystalline sepiolite (Sp2) as postulatedin themodel The lines for Sp2 Sr3 and amorphoussilica intersect at a
common point (Fig 6D) which is consistent with the nucleation of
sepiolite that typically occurs in solutions that are saturated orsupersaturated with amorphous silica and (Mg)-carbonate (egStoessell and Hay 1978 Birsoy 2002) Some bicarbonate watersfrom the Voltri Group have a silica activity near the amorphous silicaboundary (Fig 6B) but only a few alkaline springs of this group fall onthe poorly-crystalline sepiolite boundary line (Sp2 Fig 6D) Both of these features are inconsistent with the possibility that sepioliteprecipitates from supergenic fresh water in the Northern Apennine
Moreover these two results suggest that values for logK(sepiolite)b
3044 (see Table 4) should be used to model the dissolutionprecipitation of the well-crystallized phase that is the most commonform for the metasomatic sepiolite found in the investigated area
7 Conclusions
The waters issuing from ultrama1047297tes in the TarondashCeno Valleysoriginate from low-temperature interactions between meteoric waterand serpentinite rocks In this report the waterndashrock interaction hasbeen modeled using reaction paths in progress mode During their earlyinteraction with soil and rock minerals bicarbonate waters super-saturated in kaolinite ferrihydrite vermiculite and Fe2+ndashMg2+-saponitewhere theorigin of the secondary iron saponite as Fe2+-bearing smectiteagrees with the hypothesis of Guumlven (1988) Then while remainingsupersaturated with saponite the bicarbonate waters evolve into analkaline facies when poorly-crystalline serpentine precipitates Duringrock weathering the Al-bearing crystalline serpentine of oceanic originprobably promotes nucleation of the low-temperature serpentineamorphous phase and supplies Al3+ to the solution for saponiteprecipitation This interpretation is consistent with the existence of Al-poor supergenic serpentine (Neal and Stanger 1985) and the partiallymeteoric OH-isotopic imprint observed in other ultrama1047297c systems
Dif 1047297culties and complexities in the modeling arise fromuncertainties in the primary mineralogy of some local ophiolitesand from the presence of different lithotypes (rodingites ophi-carbonates gabbros) which can signi1047297cantly affect the composi-tion of water samples This can be seen in the example of the Candash
OH springs issuing from the Voltri Group (PO2 and PCO2 of 10minus70
and 10minus10 bar respectively Bruni et al 2002 Marini and
Ottonello 2002) These springs appear to be more evolved thanthe TarondashCeno alkaline waters as a result of their particularly longinteractions with serpentinites until equilibrium with enstatite isreached or with chlorite-bearing serpentinite The use of anupdated thermodynamic database in our model avoided problemsdue to mineralogical heterogeneity in the mixed-layer claysassociated with ophiolitic rocks this database accurately re1047298ectedthe chemical composition of the single phases Using this modelthe vermiculitendashchlorite and saponitendashchlorite associations seen inthe activity diagrams simulate respectively the presence of highand low charge corrensite observed in the 1047297eld The accuracy of ourmodel can be tested by comparing estimates of chlorine composi-tion resulting from our model and from experimental methods The
mean Cl concentration of the bicarbonate water is 41plusmn16 mgLtaking into account a Cl concentration of sim2 mgL in rainwater inthe area (cf Panettiere et al 2000 Venturelli 2003) 2 mgL of Clin spring water are estimated to come from serpentinite This valueagrees with the model which predicts that the switch frombicarbonate water to alkaline water requires the congruentdissolution of 00203 mol of serpentinite containing 100 ppm of soluble Cl thereby yielding water containing 2 ppm of Cl
In addition we think that weathering of the accessory albiteplagioclase is responsible for Na enrichment of the waters whereas Clenrichment in the alkaline waters is predicted to result from theleaching of NaCl 1047298uid inclusions Moreover the TarondashCeno alkalinesprings show the highest BCl ratios comparable with those of thehost metamorphic minerals (Scambelluri et al 2004) In contrast the
lower boron content and BCl ratio of the alkaline springs of the Voltri
89T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
Group are unexpected features and probably re1047298ect waterndashrockinteractions over a longer time period than in the TarondashCeno ValleysThe boron content in the Northern Apennine alkaline springs may becontrolled by several factors such as (i) the different histories of VoltriGroup and EL ultrama1047297tes (ii) the probable contribution of boronfrom NaCl 1047298uid inclusions (iii) the sinking of boron by travertineprecipitation (Hobbs and Reardon1999) brucite (Rhoades et al 1970)and smectite and (iv) decreases in boron release from minerals due to
increases in pH (Su and Suarez 2004) Acknowledgements
This research was carried out with the 1047297nancial support from theldquoYoung Researchers Projectrdquo (2001ndash2002) of the Italian Ministry of Universities Scienti1047297c Research and Technology (MURST) We thankGianniCortecci (Institute of Geosciences and EarthResources IGG-CNRPisa Italy) and Giampiero Venturelli (Earth Sciences Department Univer-sity of Parma Italy) for useful suggestions and critical review of themanuscript The authors are grateful to Enrico Dinelli (InterdepartmentResearch Centre for Environmental Sciences Alma Mater StudiorumUniversity of Bologna Ravenna Italy) for granting permission to use theGeochemistrsquos WorkbenchsoftwareTheconstructive reviews of theEditor
J Fein and of two anonymous Referees have signi1047297cantly improved the
manuscript
Appendix A Supplementary data
Supplementary data associated with this article can be found inthe online version at doi101016jchemgeo200808017
References
Abbate E Bortolotti V Passerini P 1980 Olistostromes and olistoliths In Sestini G(Ed) Development of the Northern Apennines geosyncline Sedimentary Geology4 pp 521ndash558
Alietti A 1956 Il minerale a strati misti saponite-talco di monte Chiaro (val di TaroAppen Emil) Rendiconti dellAccademia Nazionale dei Lincei 21 201 ndash207
Alietti A De Angelis G Montanari G Sgarlata F Tirelli G 1976 Il ferro nelle argilleappenniniche Periodico di Mineralogia 45 51ndash64
Anselmi B Mellini M Viti C 2000 Chlorine in the Elba Monti Livornesi and Murloserpentines Evidence for seandashwater interaction European Journal of Mineralogy12137ndash146
Apha-Awwa-Wef 1995 Standard Methods for the Examination of Water and Waste-water 19th Edition American Public Health Association American Water WorksAssociation Water Environment Federation
Barnes I LaMarche Jr VC Himmelberg GR 1967 Geochemical evidence of present-day serpentinization Science 156 830ndash832
Barnes I ONeil JR 1969 The relationship between 1047298uids in some fresh alpine-typeultrama1047297cs and possible modern serpentinization Western United StatesGeological Society of America Bulletin 80 1947ndash1960
Barnes I ONeil JRTrescases JJ1978 Present dayserpentinizationin New CaledoniaOman and Yugoslavia Geochimica et Cosmochimica Acta 42 144ndash145
Barnes I Rapp JB ONeil JR Sheppard RA Gude III AJ 1972 Metamorphicassemblages and the direction of 1047298ow of metamorphic 1047298uids in four instances of serpentinization Contributions to Mineralogy and Petrology 35 263ndash276
Barnes JDSelverstone J Sharp ZD 2006Chlorine isotope chemistry of serpentinitesfrom Elba Italy as an indicator of 1047298uid source and subsequent tectonic historyGeochemistry Geophysics Geosystems 7 doi1010292006GC001296
Beccaluva L Emiliani F Venturelli G Zerbi M 1973 Ca Fe Mg Mn Cr Ni Codistribution in some ultrama1047297c rocks outcropping in the Northern Apennines withsome geological remarks Ateneo Parmense Acta Naturalia 9 69ndash98
Beccaluva L Macciotta G Piccardo GB Zeda O 1984 Petrology of lherzolitic rocksfrom the Northern Apennine ophiolites Lithos 17 299 ndash316
BeccaluvaL Venturelli G ZanzucchiG 1975 Dati geologici e geochimici sui basalti diassociazione o1047297olitica dellAppennino ligure-emiliano Ateneo Parmense ActaNaturalia 11 789ndash802
BethkeCM 2002 The Geochemists Workbench A Userrsquos Guide to RxnAct2Tact ReactandGtplot Release40 HydrogeologyProgram Universityof IllinoisUrbanaIL 236pp
Birsoy R 2002Formation of sepiolitendashpalygorskite and related minerals from solutionClays and Clay Minerals 50 736ndash745
Bonatti E Lawrence JR Morandi N 1984 Serpentinization of oceanic peridotitestemperature dependence of mineralogy and boron content Earth and PlanetaryScience Letters 70 88ndash94
Bonifacio E Zanini E Boero V Franchini-Angela M 1997 Pedogenesis in a soilcatena on serpentinite in north-western Italy Geoderma 75 33ndash51
Boschetti T 2003 Studio geochimico e geochimico-isotopico di acque a composizione
estrema e termali dellrsquoAppennino Settentrionale Ph D Thesis Parma Univ
Brigatti MF Franchini G Lugli C Medici L Poppi L Turci E 2000 Interactionsbetween aqueous chromium solutions and layer silicatesApplied Geochemistry 151307ndash1316
Brigatti MF Lugli C Poppi L Venturelli G 1999 Iron-rich saponite dissolutionreactions and Cr uptake Clay Minerals 34 637ndash645
Brigatti MF Valdregrave G 1996 Mixed-layer chloritesmectite (corrensite) in ophioliticrock alteration products Neues Jahrbuch fuer Mineralogie Monatshefte 1 37 ndash47
Bruni J Canepa M Chiodini G Cioni R Cipolli F Longinelli A Marini L OttonelloG Vetuschi Zuccolini M 2002 Irreversible waterndashrock mass transfer accompany-ing the generation of the neutral MgndashHCO3 and high-pH CandashOH spring waters of the Genova province Italy Applied Geochemistry 17 455ndash474
Bryan CR 2005 Drift-Scale THC Seepage Model MDL-NBS-HS-000001 REV 04 YuccaMountain Project Las Vegas Nevada (US)Caruso LJ Chernosky JV 1979 The stability of lizardite Canadian Mineralogist 17
757ndash769Clark ID Fontes JC Fritz P 1992 Stable isotope disequilibria in travertine from high
pH waters Laboratory investigations and 1047297eld observations from Oman Geochi-mica et Cosmochimica Acta 56 2041ndash2050
Cortesogno L 1980 Il metamor1047297smo giurassico nelle o1047297oliti dellAppenninoSettentrionale due distinti cicli metamor1047297ci in ambiente oceanico O1047297oliti 5 5ndash18
Christ CLHostetler PB Siebert RM1973 Studies in the system MgO-SiO2-CO2-H2OIII American Journal of Science 273 65ndash83
Cortesogno L Lucchetti G 1982 Il metamor1047297smo oceanico nei gabbri o1047297oliticidellAppennino Ligure Aspetti mineralogici e petrogenetici Rendiconti dellaSocietagrave Mineralogica Italiana 38 561ndash579
Crispini L Capponi G 2002 Control of metasomatic alteration on the development of shear zone in ultrama1047297c rocks a case study from the Voltri Massif (Ligurian Alps)Architettura interna e cinematica delle zone di taglio Gruppo Italiano di GeologiaStrutturale mdash Riunione Annuale 2002 Pisa Italy 11ndash12 giugno 2002 XIV ndashXV
CrissRE 1999 Principlesof stableisotope distributionOxford University PressNew York
Davis KJ Dove PM De Yoreo JJ 2000 The role of Mg 2+ as an impurity in calcitegrowth Science 290 1134ndash1137
de Leeuw NH 2002 Moleculardynamics simulations of the growth inhibiting effect of Fe2+ Mg2+ Cd2+ and Sr2+ on calcite crystal growth Journal of Physical Chemistry B106 5241ndash5249
De Nardo MT Segadelli S Vescovi P 2007 Studio pilota per la caratterizzazionegeologicadellesorgentinellazona delM Nero (altaVal Cenoe alta ValNuremdash provincedi Parma e Piacenza) Il Geologo dellEmilia-Romagna 25 Nuova Serie pp 5 ndash21
Di Giulio A Geddo G 1990 Studio petrogra1047297co delle Arenarie di Casanova (alta valTrebbia Appennino settentrionale) Atti Ticinesi di Scienze della Terra 33 243ndash254
Dinelli E Lombini A Simoni A Ferrari C 1997 Heavy metals in serpentinite soils of selected outcrops of Piacenza and Parma provinces (Northern Apennines Italy)Mineralogica et Petrographica Acta 40 241ndash255
Epstein S Mayeda T 1953 Variation of 18O content of waters from natural sourcesGeochimica et Cosmochimica Acta 4 213ndash224
Ernst WG Piccardo GB 1979 Petrogenesis of some Ligurian peridotites I Mineraland bulk-rock chemistry Geochimica et Cosmochimica Acta 43 219ndash237
Feder F Trolard F Klingelhofer G Bourrieacute G 2005 In situ Moumlssbauer spectroscopy
evidence for green rust (fougerite) in a gleysol and its mineralogical transforma-tions with time and depth Geochimica et Cosmochimica Acta 69 4463ndash4483Fritz PClark IDFontes JCWhiticar MJFaberE 1992 Deuterium and
13C evidence
for low temperature production of hydrogen and methane in a highly alkalinegroundwater environment in Oman In Kharaka TK Maest AS (Eds) Proceed-ings of the 7th International Symposium on WaterndashRock Interaction BalkemaRotterdam pp 793ndash796
Giammetti F 1968 Le o1047297oliti di Roccaprebalza Costa della Guardia e Groppo di Gorro(Appennino Parmense) Ateneo Parmense Acta Naturalia 4 107ndash142
Giuseppetti G1953 Roccee minerali dellaformazioneo1047297olitica di Volpedo Rendicontidella Societagrave Mineralogica Italiana 9 85ndash134
Gran G 1952 Determination of equivalence point in the potentiometric titrationAnalyst 77 661ndash671
Gunnarsson I Arnorsson S Jakobsson S 2005 Precipitation of poorly crystalline antigoriteunder hydrothermal conditions Geochimica et Cosmochimica Acta 69 2813ndash2828
Guumlven N 1988 Smetites In Bailey SW (Ed) Reviews in Mineralogy 19 Hydrousphyllosilicates (exclusive of micas) Mineralogical Society of America MichiganBook Crafters Inc 497ndash559
Hansel CM Benner SG Fendorf S 2005 Competing Fe(II)-induced mineralization
pathways of ferrihydrite Environmental Science and Technology 39 7147ndash7153HobbsMYReardonEJ 1999 Effect of pH on boronco-precipitation by calcite further
evidence for non-equilibrium partitioning of trace elements Geochimica etCosmochimica Acta 63 1013ndash1021
Jones BF Galan E 1988 Sepiolite and palygorskite In Bailey SW (Ed) HydrousPhyllosilicates (exclusive of Micas) Reviews in Mineralogy vol 19 MineralogicalSociety of America Michigan Book Crafters Inc 631ndash674
Jove Colon CF 2006 Technical work plan for Thermodynamic database for chemicalmodelling TWP-MGR-PA-000039 REV 01 Yucca Mountain Project Las VegasNevada (US)
Kohls DW Rodda Jr JL 1967 Iowaite a new hydrous magnesium hydroxidendashferricoxychloride from the Precambrian of Iowa American Mineralogist 52 1261 ndash1271
Kyser TK OHanley DS Wicks FJ 1999 The origin of 1047298uids associated withserpentinization processes evidence from stable-isotope compositions CanadianMineralogist 37 223ndash237
Lapham DM 1961 New data on deweylite American Mineralogist 46 168ndash188Lee BD SearsSKGraham RC Amrhein CValiH 2003 Secondarymineral genesis
from chlorite and serpentine in an ultrama1047297c soil toposequence Soil Science
Society of America Journal 671309ndash1317
90 T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
Longinelli A Selmo E 2003 Isotopic composition of precipitation in Italy a 1047297rstoverall map Journal of Hydrology 270 75ndash88
Majzlan J Navrotsky A Schwertmann U 2004 Thermodynamics of iron oxides PartIIIEnthalpiesof formation and stability of ferrihydrite (simFe(OH)3) schwertmannite(simFeO(OH)34(SO4)18) and ɛ-Fe2O3 Geochimica et Cosmochimica Acta 681049ndash1059
Mariner P 2003 In-Drift PrecipitatesSalts Model ANL-EBS-MD-000045 REV 01 YuccaMountain Project Las Vegas Nevada (US)
Marini L Ottonello G 2002 Atlante degli acquiferi della Liguria Volume III Le acquedeicomplessio1047297olitici (bacini ArrestraBranega Cassinelle Cerusa ErroGorzenteLeira Lemme Lerone Orba Piota Polcevera Rumaro Sansobbia Stura Teiro
Varenna Visone) Pacini PisaMellini L Zanazzi PF 1987 Crystal structures of lizardite-l T and lizardite-2H1 fromColi Italy American Mineralogist 72 943ndash948
Mezzadri G 1964 Petrogra1047297a delle arenarie di Ostia Rendiconti Societagrave MineralogicaItaliana 20 192ndash228
Mondesir H Decarreau A 1987 Synthesegrave entre 25 et 200 degC de lizardites Ni ndashMgMesure des coef 1047297cients de partage solide-solution aqueuse pour le couple Ni ndashMgdans des lizardites Bulletin de Mineralogie 110 409ndash426
Neal C Shand P 2002 Spring and surface water quality of the Cyprus ophiolitesHydrology and Earth System Sciences 6 797ndash817
Neal C Stanger G 1983 Hydrogen generation from mantle source rocks in OmanEarth and Planetary Science Letters 66 315ndash320
Neal C Stanger G 1984 Calcium and magnesium hydroxide precipitation fromalkaline groundwaters in Oman and their signi1047297cance to the process of serpentinization Mineralogical Magazine 48 237ndash241
Neal C Stanger G 1985 Past and present serpentinisation of ultrama1047297c rocks anexample from Semail ophiolite nappe of Northern Oman In Drever JI (Ed) TheChemistry of Weathering D Reidel Publishing Company pp 249ndash275
Nesbitt HW Bricker OP 1978 Low temperature alteration processes affecting
et Cosmochimica Acta 41 1835ndash1841OrsquoHanley DS 1996 Serpentinites records of tectonic and petrological history Oxford
University PressOrberger B Friedrich G Woermann E 1990 The distribution of halogens and carbon
in Pge-bearing ultrama1047297cs of the Acoje Ophiolite Block Zambales Philippines Journal of Geochemical Exploration 37 147ndash169
Ottonello G Joron JL Piccardo GB 1984 Rare Earth and 3d transition elementsgeochemistry of peridotitic rocks II Ligurian peridotites and associated basalts Journal of Petrology 25 373ndash393
Palandri JL Kharaka YK 2004 A compilation of rate parameters of water-mineralinteraction kinetics for application to geochemical modeling Open File Report2004-1068 US Geological Survey
Panettiere P Cortecci G Dinelli E Bencini A Guidi M 2000 Chemistry and sulfurisotopic composition of precipitation at Bologna Italy Applied Geochemistry 151455ndash1467
Parkhurst DL Appelo CAJ 1999 Userrsquos guide to PHREEQC (version 2) mdash a computerprogram for speciation batch-reaction one-dimensional transport and inversegeochemical calculation US Department of the Interior US Geological SurveyWashington DC
Pfeifer HR 1977 A model for 1047298uids in metamorphosed ultrama1047297c rocks Observationsat surface and subsurface conditions (high pH spring waters) SchweizerischeMineralogische und Petrographische Mitteilungen 57 361ndash396
PlummerLN1984Geochemical modeling A comparisonof forwardand inverse methodsIn Hitchon B Wallick EI (Eds) Proceedings First CanadianAmerican Conference onHydrogeology-Practical Applications of Ground Water Geochemistry Banff AlbertaCanada mdash Worthington Ohio National Water Well Association pp 149ndash177
Rampone E Hofmann AW Piccardo GB Vannucci R Bottazzi P Ottolini L 1995Petrology mineral and isotope geochemistry of the External Ligurides peridotites(Northern Apennines Italy) Journal of Petrology 36 81ndash105
Rampone E Hofmann AW Piccardo GB Vannucci R Bottazzi P Ottolini L 1996Trace element and isotope geochemistry of depleted peridotites from an N-MORBtype ophiolite (Internal Liguride N Italy) Contributions to Mineralogy andPetrology 123 61ndash76
Rhoades JD Ingvalson RD HatcherJT1970Adsorption of boronby ferromagnesianminerals and magnesium hydroxide Soil Science Society of America Journal 34
938ndash941Saini-EidukatB KuchaH Keppler H1994Hibbingiteγ-Fe2(OH)3Clanewmineralfromthe Duluth Complex Minnesota with implications for the oxidation of Fe-bearingcompounds and the transport of metals American Mineralogist 79 555ndash561
Sanford RF 1981 Mineralogical and chemical effects of hydration reactions andapplications to serpentinization American Mineralogist 66 290ndash297
Scambelluri M Muumlntener O Ottolini L Pettke TT Vannucci R 2004 The fate of BCl and Li in the subducted oceanic mantle and in the antigorite breakdown 1047298uidsEarth and Planetary Science Letters 222 217ndash234
Scambelluri M Rampone E 1999 Mg-metasomatism of oceanic gabbros and itscontrol on Ti-clinohumite formation during eclogitization Contributions toMineralogy and Petrology 135 1ndash17
Scambelluri M Bottazzi P Trommsdorff V Vannucci R Hermann J Gogravemez-Pugnaire MT Logravepez-Sagravenchez Vizcaino V 2001a Incompatible element-rich1047298uidsreleased by antigorite breakdown in deeply subducted mantle Earth and PlanetaryScience Letters 192 457ndash470
Scambelluri M Rampone E Piccardo G 2001b Fluid and element cycling insubducted serpentinite a trace-element study of the ErrondashTobbio high-pressureultrama1047297tes (Western Alps NW Italy) Journal of Petrology 42 55ndash67
Segadelli S2006 Lageologia nelpaesaggio le rupi o1047297olitiche in Val Taro e Val Ceno IlGeologo dellrsquoEmilia-Romagna 22 15ndash29 Nuova Serie
Settembre Blundo D Martin De Vidales JL Alietti A 1992 Corrensite-type clayminerals from Taro Valley (North Italy) Mineralogica et Petrographica Acta 35215ndash222
Sharp ZD Barnes JD 2004 Water-soluble chlorides in massive sea1047298oor serpenti-nites a source of chloride in subduction zones Earth and Planetary Science Letters226 243ndash254
Siever R Woodford N 1979 Dissolution kinetics and the weathering of ma1047297cminerals Geochimica et Cosmochimica Acta 43 717ndash724Singer A 1989 Palygorskite and sepiolite group minerals In Dixon JB Weed SB
(Eds) Minerals in Soil Environments Soil Science Society of America MadisonWisconsin pp 829ndash872
Siffert B Wey R 1962 Synthese d rsquoune sepiolite a temperature ordinarie Comptesrendus hebdomadaires des seacuteances de lAcadeacutemie des Sciences de Paris 2541460ndash1463
Środoń J 1999 Nature of mixed-layer clays and mechanisms of their formation andalteration Annual Reviews of Earth and Planetary Science 27 19ndash53
Stoessell RK 1988 25 degC and 1 atm dissolution experiments of sepiolite and keroliteGeochimica et Cosmochimica Acta 52 365ndash374
Stoessell RK Hay RL 1978 The geochemical origin of Sepiolite and keorilite atAmboseli Kenya Contributions to Mineralogy and Petrology 65 255ndash267
Suarez DL Šimůnek J 1996 Solute transport modeling under variably saturatedwater 1047298ow conditions In Lichtner PC Steefel CI Oelkers EH (Eds) ReactiveTransport in Porous Media Reviews in Mineralogy vol 34 Mineralogical Society of America Michigan Book Crafters Inc 229ndash268
Sumner DY Grotzinger JP 1996 Were kinetics of Archean calcium carbonate
precipitation related to oxygen concentration Geology 24 119ndash122Su C Suarez DL 2004 Boron release from weathering of illites serpentine shales
and illiticpalygorskitic soils Soil Science Society of America Journal 68 96 ndash105Toscani L Venturelli G Boschetti T 2001 Sul1047297de-free and sul1047297de-bearing waters in
the Northern Apennines Italy general features and waterndashrock interaction AquaticGeochemistry 7 195ndash216
Treves BE Harper GD 1994 Exposure of serpentinites on the ocean 1047298oor sequenceof faulting and hydrofracturing in the Northern Apennine ophicalcites O1047297oliti 19b435ndash466
Varadachari C Kudrat M Gosh K 1994 Evaluation of standard free energies of formation of clay minerals by an improved regression method Clays and ClayMinerals 42 298ndash307
Veniale F 1966 Sepiolite in sedimenti dellrsquoAppennino Vogherese (Pavia) Periodico diMineralogia 35 343ndash386
Venturelli G 2003 Acque minerali ed ambiente mdash Fondamenti di geochimica deiprocessi di bassa temperatura Pitagora Editrice Bologna
Venturelli G Contini S Bonazzi A Mangia A 1997 Weathering of ultrama1047297c rocksand element mobility at Mt Prinzera Northern Apennines Italy MineralogicalMagazine 61 765ndash778
Venturelli G Frey M 1977 Anchizone metamorphism in sedimentary sequences of the Northern Apennines Rendiconti Societagrave Mineralogica Italiana 33 109ndash123
Vieillard P 2002 A new method for the prediction of Gibbs free energies of formationofphyllosilicates (10 Aring and14 Aring)basedon theelectronegativityscale Claysand ClayMinerals 50 352ndash363
Vissers RLM Strating EHH 2001 Structures and microstructures in a thrust-relatedgreenschist facies tectonic melange Voltri Group (NW Italy) O1047297oliti 26 33ndash46
Viti C Mellini M 1997 Contrasting chemical compositions in associated lizardite andchrysotile in veins from Elba Italy European Journal of Mineralogy 9 585 ndash596
Viti C Mellini M 1998 Mesh textures and bastites in the Elba retrogradeserpentinites European Journal of Mineralogy 10 1341ndash1359
Wenner DB Taylor Jr HP 1974 Deuteriumatomic hydrogen and oxygen-18oxygen-16 studies of serpentinization of ultrama1047297c rocks Geochimica et CosmochimicaActa 38 1255ndash1286
Williams LA Parks GA Crerar DA 1985 Silica diagenesis I Solubility controls Journal of S edimentary Petrology 55 301ndash311
Wilson J Savage D Cuadros J Shibata M RagnarsdottirKV 2006 The effect of ironon montmorillonite stability (I) Background and thermodynamic considerations
Geochimica et Cosmochimica Acta 70 306ndash322WollastR MackenzieFTBricker OP1968 Experimental precipitation and genesis of sepiolite at earth-surface conditions American Mineralogist 531 1645ndash1662
Wrucke CT1996 Serpentine and carbonated-hosted asbestos deposits InEdwardAdu Bray (Eds) Preliminary compilation of descriptive geoenvironmental mineraldeposit models 95ndash831 US Geological Survey Open-File Report pp 39 ndash46
Xu T Apps JA Pruess K 2001 Analysis of mineral trapping for CO2 disposal in deepaquifers Lawrence Berkeley National Laboratory University of California
ZhangYDaweRA2000 In1047298uence ofMg2+ on thekineticsof calcite precipitation andcalcite crystal morphology Chemical Geology 163 129ndash138
Zhu C Anderson G 2002 Environmental applications of geochemical modelingCambridge University Press
91T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
(486 mgL CaCO3) conductivity (877 μ S cmminus1) and alkalinity (618 mgL HCO3) Springs PR10 and UM15 from ultrama1047297tes show an extremecomposition being sodium-hydroxide and sodium-sulfate respectivelytheir pH is high (101ndash112) their Eh is low (down to minus103 mV) theyaresoft (20ndash61 mgL CaCO3 220 to 408 μ S cmminus1) and they contain largeamounts of boron (up to 2294 μ gL and up to 13461 μ gL respectively)and moderate amounts of reduced sulfur and nitrogen species In bothPR10 and UM15 non-carbonate alkalinity predominates over carbonate
alkalinity using PHREEQCI software with the llnldat thermodynamicdatabase (Parkhurst and Appelo 1999) to calculate the composition of alkaline species under thepH Ehand T valuesmeasuredin1047297eld samplePR10 was found to contain 54 hydroxide 23 carbonate 19 borateand 4 other alkalinespecies (Si sulphide and ammonia) while sampleUM15 contained 75borate14 hydroxide10carbonate and 1 otherspecies
The SindashMgndashC diagram (Fig 2) is a useful tool for studying waterndashrockinteractions Dissolution reactions of typical magnesium minerals of ultrama1047297c rocks are represented by a line with CMg=2 (molar ratio)
Fig3 Plot of (A) Naand (B) Mgvs Cl forallwater samples from the TarondashCenoValleys including bottledwaters Rocca Galgana andFontenova(chemical data taken from labels)Alkalinesprings from the Voltri Group (smallblack diamonds Marini and Ottonello 2002) and meteoric water mean composition (MW Panettiere et al 2000 Venturelli 2003) are also plottedDottedcurvesin(A)and(B)refertomixingwith04and2wtNaCl 1047298uidinclusionsand MgCl2dissolution (1 mol)respectively Greycrosses andi csescorrespondrespectively to mixingwith04and2wtNaCl 1047298uidinclusionsfor mixing fractionsranging from 0001 to0005The solid linerefers tomixingof meteoric water(A) orMg-bicarbonate water(B) withseawater
The vertical grey line represents the water-soluble Cl content in the serpentinites from Northern Apennine (Barnes et al 2006) Symbols are as in Fig 2
80 T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
The Mg-bicarbonate watersissuingfrom ultrama1047297crocksfallnearthisline and cluster between the clinochlore and forsterite points (Fig 2)Stream waters and springs issuing from sedimentary and basaltic rocksfall between thecarbon apex and theline indicated by a CMg molar ratioof 4 which describes the dissolution of CandashMg-clinopyroxene
This is consistent with the (Ca)-bicarbonate compositions of thesestream waters and springs It is noteworthy that the Ceno streamwaters shifted toward the C apex of Fig 2 derive their chemistry fromthe interaction with Mt Caio Flysch cropping out widely in the Cenocatchment (samples from UM50 to UM59 see Map1 sheet in theannexes) whereas the Taro source spring waters plot towards theother apexes (lower C) typical of ophiolite minerals (samples fromUM41 to UM49 see Map1 sheet in the annexes) The use of the traceelements (eg Sr and Rb content Boschetti 2003) further highlightthe differences between the stream waters of the two rivers and therole of lithology Nonetheless mixing and equilibrium with atmo-sphere make the reaction-path modeling of these samples poorlyreliable so their are unsuitable for the purpose of the present work
Alkaline waters from ultrama1047297c rocks fall inside the same sub-triangle mostly along the CndashSi side probably due to the precipitationof magnesium minerals during the evolution of the water composition(Bruni et al 2002)
42 Sodiumndashchloride relationships
Chlorine and other halogens in serpentinized rocks are inheritedfrom seawater and marine sediments during oceanic metamorphism(eg Orberger et al 1990) They are transferred into serpentinite byadsorption (Anselmi et al 2000 Sharp and Barnes 2004) entrap-ment in 1047298uid inclusions (Scambelluri et al 2001ab) and substitutionin the structure of low-crystalline iron-hydroxide phases These latterare composed of the isomorphous and chemically similar iowaitehibbingite and FendashCl-oxydroxides (also named green rusts orfougerites Kohls and Rodda 1967 Saini-Eidukat et al 1994 Federet al 2005 and references therein) In the Northern Apennineserpentinite rocks the mean Cl content is sim300 ppm mean values of 221 and 375 ppm were measured respectively in the Erro ndashTobbio(Ligurian Western Alps Scambelluri et al 2004) and in the centralItaly serpentinites (Anselmi et al 2000) but thewater-soluble Cl from
serpentinites is estimated to be than 100 ppm (Barnes et al 2006) Asshown in Fig 2 waters from basalts and ultrama1047297tes of the TarondashCenoValleys show a trend towards the Cl apex whereas in Fig 3 theestimated soluble chlorine in serpentinites agrees with the highestvalue reported in spring waters from ultrama1047297tes (samples fromMarini and Ottonello analyzed in 2002) The Cl content of samplesUM1 and UM36is veryhigh(Fig 3) and is probablydue to pollutionInfact these springs are not far from the A15 highway (UM1) and theTomarlo Pass road (UM36) where salt mixtures comprising NaClMgCl2middot6H2O and probably CaCl2 are widely used for deicing in winterAlkaline water with a Mg-bicarbonate composition from ultrama1047297tesshow the highest and lowest Cl contents respectively Na-rich Mg-bicarbonate waters from basalts (UM6 UM7 Fig 3) and ultrama1047297tescan be explained by the dissolution of albite in basalts (Beccaluva
et al 1975) and in spilitized EL peridotites (Beccaluva et al 1984
Rampone et al 1995) respectively whereas the highest content of both Na and Cl in the alkaline water samples UM15 and PR10 (Fig 3)can be explained by leaching of NaCl 1047298uid inclusions in theserpentinite rocks (04ndash20 wt NaCl Scambelluri et al 2004)
43 Boron and BCl ratio
Compared with unaltered ultrama1047297c rocks serpentinites have
higher content of both boron and chlorine and most authors suggestthat both elements come from seawater (eg Sanford 1981 andreferences therein Bonatti et al 1984) Despite several studies onthese rocks the behavior of Cl and B in solution during waterndashrockinteraction processes is not well known Early investigations on 1047298uidsassociated with serpentinites in the western United States (Barnes andONeil 1969 Barnes et al 1972) revealed metamorphic waters withhighboron levels(up to 28mM) and BCl molar ratiosof 0038ndash013 aswell as Mg-bicarbonate and high-pH waters of meteoric origin withvariable BCl ratios of 0027ndash042 (median 0039) and 00006ndash0017(median 00076) respectively Bicarbonate waters issuing fromophiolitic rocks in Cyprus have BCl ratio in this range (00017ndash
0058 median 00019 Neal and Shand 2002) whereas low boron
Fig 4 Plots of (A) δ18O vs δ2H and (B) δ18O vs altitude for waters from the Taro ndashCenoValleys C-Italy and N-Italy are the meteoric water lines for Central and Northern Italy
respectively (Longinelli and Selmo 2003) Symbols are as in Fig 2
81T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
levels were found in the alkaline springs of meteoric origin in OmanNew Caledonia and ex-Yugoslavia (Barnes et al 1978)
Among the waters analyzed so far in the TarondashCeno Valleys thebicarbonate waters have BCl ratios in the range 00017ndash022 (median0073) whereas the alkaline waters show the highest ratios everreported (BCl= 034ndash33 median 169) The high boron content of upto 13 mgL of the alkaline springs could be related to metamorphic1047298uids released during serpentinite subduction in fact rocks prefer-
entially lose Cl and the BCl ratio of the released 1047298
uids increases withpressure and temperature (Scambelluri et al 2004) Nonetheless thealkaline water samples in our study differ signi1047297cantly from alkalinemetamorphic1047298uids in that they have low salinity high BCl ratios andmeteoric isotope signatures (see below) all these characteristicssuggest an origin in the interaction of meteoric water with high-pressure metamorphic serpentinite This hypothesis is furthersupported by the agreement in BCl ratio between the alkalinesamples we study here and the high BCl ratio of up to 4 previouslyreported for the residual high-pressure mineralogy of ultrama1047297tes(olivine+orthopyroxene Scambelluri et al 2004) In contrast theVoltri Groups Ca-hydroxide waters have mean values of B = 014 mgL and BCl=003 (Marini and Ottonello 2002) consistent with acontribution of boron and chlorine from 1047298uid inclusions In factassuming fractional contributions of 0001ndash0005 from NaCl 1047298uidinclusions with boron content from 4 to 8 mgL ( Fig 3a Scambelluriet al 2004) results in a B content of 0004ndash01 mgL in the solutionThese results suggest that the difference between springs from theVoltri Group and TarondashCeno Valleys is probably related to (i) differentevolution of the 1047298uids and (ii) the different types of host rocks and Pndash
T paths of the Ligurian Western Alps and EL peridotites (eg Beccaluvaet al 1984 Scambelluri et al 2004)
44 Isotopic composition of water
A diagram of δ2H vs δ18O (Fig 4A) shows that the waters sampledfrom the TarondashCeno Valleys are meteoric in origin since they areconsistent with the meteoric water lines of central and northern Italy(Longinelli and Selmo 2003) The altitude vs δ18O(H2O) (Fig 4B)
summarizes the isotopic effects due to temperature altitude andamount of rainfall (eg Criss1999) all of which help to determine theisotopic composition of the water samples The tendency is similar forbicarbonate waters from sedimentary basaltic and ultrama1047297c forma-
tions which may provide insight into the recharge areas alkalinewaters in contrast probably circulate deeper in the ground
5 Water ndashrock interaction model
51 Fundamentals
The reaction path model between meteoric water and serpentinite
was performed using PHREEQCI 2125 code (Parkhurst and Appelo1999) and the Lawrence Livermore National Laboratory thermody-namic database (llnldat as known as thermocomV8R6 ) implementedwith the equilibrium constants listed in Table 2
Computations wereperformed by reaction progress mode ( forward
modeling Plummer 1984 Zhu and Anderson 2002) ie (i) addingreacting rock at each step in the simulation (ii) activating the phase(s)precipitation only when the thermodynamic stability of the mineralparagenesis is well known and consistent with the mineralogy of thearea and (iii) plotting reaction progress on speci1047297c activity diagramsThe same thermodynamic database was used with The GeochemistsWorkbench 403 software (Bethke 2002) to generate Pourbaix andactivity diagrams (Figs 5 6 8 and 9) Modeling was performed usingan approach based on equilibrium rather than reaction transportbecause the data on reaction kinetics and reactive surfaces in thesesolid phases systems are scarce particularly with respect to theactivation energy of clay and mixed-layer minerals at low temperatureand high pH Moreover the hydrogeology and the discontinuity of theaquifers in the study area which are related to the geologicheterogeneity and structural complexity of the Northern Apenninewould make a model based on reaction transport even less credible atthis stage
The modeling in this study was performed on the basis of thefollowing knowledge (1) primary and secondary mineral assem-blages (2) chemical composition of the minerals involved (3) physicalparameters and chemical composition of the starting 1047298uid (4)thermodynamic data on pure phases
52 The solid reactant primary paragenesis
EL peridotites of the TarondashCeno Valleys (NorthernApennines Italy)are spinel lherzolites containing olivine (forsterite) orthopyroxene(enstatite) and clinopyroxene (diopside) (Giammetti 1968 and
Table 2
Thermodynamic data on hydrolysis reactions added to the llnldat database of the PHREEQCI code (Parkhurst and Appelo 1999) and thermocomV8R6dat database of TheGeochemists Workbench software (Bethke 2002)
Mineral Reaction log K (TdegC) ΔGdeg f References
0 25 60 100 (kJ mol-1)
clinochlore08-daphnite02 Mg4Fe1Al2Si3O10(OH)8 + 16H+=2Al3+ + Fe2++12H2O+4Mg2++3SiO2 7318 6384 5272 4269 minus786379 this workferr ihydrite 2-l ine Fe(OH)3+3H+= Fe3++ 3H2O 484 355 ndash ndash minus70850 Majzlan et al 2004ferr ihydrite 6-l ine Fe(OH)3+3H+= Fe3++ 3H2O 435 312 ndash ndash minus71100 Majzlan et al 2004liz09-berth01 Mg27Fe02Al02Si19O5(OH)4+64H+=02Al3++02Fe2++52H2O+27Mg2++19SiO2 3111 2786 2383 2005 minus399854 this workp-antigorite Mg
Saponite-Fe2+ Fe3175Si365Al035O10(OH)2+74H+=035Al3++3175Fe2++47H2O+365SiO2 1845 1635 1319 1001 minus452459 Wilson et al 2006Saponite-Fe2+ndashMg Fe0165Mg3Si367Al033O10(OH)2+732H+=033Al3++0165Fe2++466H2O+3Mg2++
367SiO2
3037 2729 2330 1921 ndash Marini and Ottonello(2002)
The Gibbs free energy of formation (ΔGo f ) of the two theoretical Fe3+ (Varadachari et al1994) and Fe2+ vermiculites were calculated using a previously published method (Vieillard
2002 and Vieillard pers comm)log K values extimated considering an ideal solid solution between the end-members and using data thermodynamic data from llnldat (clinochlore08-daphnite02) and Wilson
et al (2006) (liz09-berth01)
82 T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
references therein Venturelli et al 1997) Accessory phases areplagioclase picotite plusmn calciumndashsodic amphibole (Venturelli et al1997) Before modeling the crystal chemical formula of the mineralphases was calculated from data from Venturelli et al (1997) thendepending on the modal composition of the rock (Ernst and Piccardo1979) and the composition of the ideal solid solution of the minerals
the resulting stoichiometry of the solid reactant(s) was calculated andinserted in the input 1047297le of the PHREEQCI software (Table 3)Thermodynamic data of the primary mineral paragenesis are all wellknown and are included in the llnldat thermodynamic database of PHREEQCI
53 The solid reactant secondary paragenesis
Secondary minerals in the ultrama1047297tes of the TarondashCeno Valleysare serpentine (lizardite and rare chrysotile) smectite minerals (Fendash
Mg saponites) and Fe-oxi-hydroxides Other accessory minerals arechlorite mixed layer saponitendashchlorite (low-charge corrensite) talccalcite and dolomite (Beccaluva et al 1973 Dinelli et al 1997Venturelli et al 1997 Brigatti et al 1999 2000) In addition mixed
layer saponitendashtalc (aliettite) and vermiculitendashchlorite (high-charge
corrensite) are present (Alietti 1956 Settembre Blundo et al 1992Brigatti and Valdregrave 1996) Except for serpentine themodal abundanceof these phases is unknown In addition some of these minerals ndash
particularly interlayer clays ndash may have different origins (hydrothermalor weathering allogenicor authigenic)and different stabilitiesrelatedtokinetics and drainage 1047298ow rate For example chlorite weathering yields
vermiculite-bearing or saponite-bearing interlayers under good- andpoor-drainage regimes respectively (Bonifacio et al 1997 Środoń1999) In addition the thermodynamic data available for serpentineinclude information only on chrysotile which is less abundant thanlizardite in the serpentinites of the areas that we studied According toVenturelli et al (1997) the lizardite composition in the serpentinites of the Taro valley can be summarized as
which is similar to 1-T lizardite from the Monte Fico serpentinizedperidotite (Elba Island Northern Tyrrhenian Sea Italy Viti and Mellini1997)
Using the thermodynamic data of Wilson et al (2006) and
assuming an ideal solid solution between lizardite and berthierine
Fig 5 EhndashpH iron diagram of the spring waters coming from ultrama1047297c rocks of the TarondashCeno Valleys Bicarbonate and high-pH waters of the Voltri Group are also represented(dotted area andsmallblackdiamonds respectively) Thelowest Eh value of thehigh-pH waterswas calculated using theS(+VI)S(minus II)redox couple(Marini and Ottonello 2002)Thestability 1047297eld of mineral phases are in grey and those of aqueous species in white In diagrams A and B 6-line ferrihydrite-vermiculite and goethite were selected as subaerial solidphases respectively
83T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
we calculated ΔGf 0=minus399854 kJmol for a theoretical (09)(lizardite)+
(01)(berthierine) mineral (Table 2) The log K value of the stated solidsolution was calculated at different temperatures using PHREEQCIafter loading the thermodynamic data on the two end-members intothe database The log K results obtained for Al-lizardite were added tothe thermodynamic database of the solid reactant (Table 3) andcompared with data from the literature (Table 4)
54 Composition of the starting solution O 2 partial pressure and
minerals included in the model
The weighted average chemical composition of rainwater inBologna (Panettiere et al 2000) which is 95 km east of theinvestigated area was used as the starting solution in the modelFull chemical data on rainwater of the TarondashCeno Valleys are
unavailable
Total alkalinity was recalculated for log PCO2=minus35 bar and
T =11 degC which was the mean temperature of the springs sampledand for pe (or Eh) based on chemical data for the redox couple N(minusIII)N(+V) The values of pH and pe obtained were 566 and 452(Eh=025 V) respectively the former agrees with the median valuereported for rainwater in thearea (Venturelli et al1997) Because dataon Al Fe and Si in rainwater were lacking we calculated theconcentrations of these elements assuming logPCO2=minus30 bar andslight supersaturation relative to ferrihydrite-6 line and kaolinite Inthis way a hypothetical composition of water in equilibrium with thelocal soil was calculated (Table3) and adopted as the ldquoinitial waterrdquo formodeling
Evidence suggests that secondary minerals particularly Fe-bearingphases change the chemical composition of the water 1047298owingthrough the ultrama1047297tes of the TarondashCeno Valleys According to Neal
and Stanger (1983) and Fritz et al (1992) the production of CH4 and
Fig 6 Activity plotfor the MgOndashSiO2ndashAl2O3ndashFe2O3(FeO)ndashH2O systemat differentlog PO2 values Spring waters issuing from TarondashCeno Valleys (symbols as in Fig 2) andVoltri Groupultrama1047297tes are shown (grey area bicarbonate waters small black diamonds alkaline waters from Marini and Ottonello 2002) In plots A B and C the stability 1047297eld limits of vermiculite(Table 2) correspond to25 degCwhilethe amorphoussilica (am-SiO2) brucite serpentine (Sr1ndash4 in Table 4)and sepiolite(Sp1ndash2 in Table 4) phasescorrespondto 11 degCThepath calculatedfor interaction between waterand ultrama1047297c rockis also shown (dotted line)the number-letter combinations referto the appearanceand disappearance of minerals(see Table 5)
84 T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
H2 in thealkalinewaters1047298owing from theultrama1047297tes isrelatedto theconsumption of oxygen during the oxidation of iron forming thehematitendashmagnetite pair Therefore the activity of iron and theoxidation state of the water must be de1047297ned The analysis of watersamples from ultrama1047297tes were plotted in both the Pourbaix andactivity diagrams (Figs 5 and 6) where the calculated activity of Fe2+Mg2+ SiO2 SO4
2minus (Pourbaix) and log PO2 (activity diagrams) wereinvolved Iron concentration was quite variable in the TarondashCenosprings where the average content was 27 μ gL and the contentreached a maximum of 200 μ gL and in the Voltri springs where theaverage was 44 μ gL and the maximum was 948 μ gL (Marini andOttonello 2002) In both cases the variations are probably related to
the different mineralogies of the soils and the different solubilities of the Fe-rich phases like goethite and ferrihydrite Iron-bearingsaponites and montmorillonites are the prevailing clay minerals inthe area investigated probably because of the ubiquitous presence of iron oxy-hydroxide (Alietti et al 1976 G Venturelli pers comm) Forthis reason the model considers the involvement of pure Mg-smectiteto be negligible
In the Pourbaix diagram (Fig 5) the data from bicarbonate watersfall into the ferrihydrite-goethitendashvermiculite1047297elds and the data fromalkaline waters fall on the boundaries of the vermiculitesaponite andsaponitechlorite Since ferrihydrite is the 1047297rst precipitating Fe-phaseit was used rather than goethite in the modeling The diagram alsosuggests the transformation of ferrihydrite into magnetite during thechange from bicarbonate to alkaline composition but this process
requires a large amount of Fe2+ in solution (N2 mM Hansel et al2005) which is unlikely at this stage of the waterndashrock interactionThus the magnetite participating in the equilibrium is most likelyprimary in origin
The modeling and the activity diagrams (Figs 6 and 9) werecalculated using the average value PO2= 10minus26 or 10minus50 bar forbicarbonate and alkaline waters respectively These values arecompatible with those calculated by Marini and Ottonello (2002) forbicarbonate waters (PO2= 10minus40 bar) and for alkaline waters of theVoltri Group (10minus77 barbPO2b10minus42 bar) The lowest PO2 value whichwas used as a limit in the modeling was calculated from the S(+VI)S(minusII) redox couple Moreover the activity diagrams (Fig 6) indicatethat the data from the bicarbonate waters from the TarondashCeno Valleysfall between quartz whose log [SiO2] =minus433 is not shown because of
the slow precipitation kinetics under the PndashT conditions considered
here (Williams et al 1985) and amorphous silica boundary 1047297elds(Fig 6A) at the same manner bicarbonate waters with higher SiO2-activity from the Voltri Group fall on amorphous silica boundary(Fig 6B) In summary then the bicarbonate waters are supersaturatedin kaolinite ferrihydrite vermiculite and saponite Alkaline watersare also supersaturated with saponite (Fig 6AndashC) and they are alsoequilibrated with primary magnetite and under extreme conditionsprobably pyrite as well
55 Limitations of the model related to kinetics
Model calculations disregard the rates of mineral dissolution and
precipitation which becomes increasingly important as temperature
Table 3
Starting mineralogy (left side) and 1047298uid composition (right side) used in the modeling
Rock composition Fluid composition
Modal of the mineralsin the interacting rock
mol of the minerals usedin the modeling
meteoric water composition Initial 1047298uid composition used in the modeling
Spinel 003 Na 3523Eminus05 3523Eminus055 Spinel Chromite 001 S 3519Eminus05 3519Eminus05
Magnetite 0 Si ndash 1176Eminus06Kaolinite ndash 5258Eminus08
Ferrhydrite-6line ndash 5882Eminus07
The modal percentage of the minerals in a serpentinite from TarondashCeno Valleys (Ernst and Piccardo 1979) was re-calculated assuming an ideal solid solution between pure extremephasesand takingthe modal percentage tobe themass percentage Thelatter was used asthe solid reactant composition in thewaterndashrockinteraction modelPure solid phases werechosenfrom the llnldat thermodynamic database of the PHREEQCI softwareThe weighted average chemical compositionof the meteoric waterfrom Bologna (Panettiere et al 2000)was chosen as the initial 1047298uid composition in the modeling after its equilibration with kaolinite and ferrihydrite (see text for details)
Table 4
Summary of thermodynamic data for the sepiolite (Spn) and serpentine (Sr n) mineralsphases in italics do not refer to the mentioned hydrolysis reactions
ΔGf 0 log K Reference
(kJ molminus1) (at 25 degC)
Sepiolite (Spn) mdash Mg4Si6O15(OH)2middot6H2O+8H+=4Mg2++6SiO2+11H2Ominus925400 3004 Gunnarsson et al (2005)minus925133 3050andash3172 Christ et al (1973)
(Sp2) minus921986 3602andash3756 Wollast et al (1968)
Serpentine (Srn) mdash Mg3Si2O5(OH)4+ 6H
+
=3Mg
2+
+2SiO2+ 5H2ONi-lizardite minus 289324 1680 Mondesir and Decarreau (1987)
Al-lizardite2 minus411003 2544a Caruso and Chernosky (1979) Al-lizardite minus 339854 2786 this work (see text for details)
Mg-lizardite (Sr1) minus404025 3056 Wilson et al (2006)Mg-crysotile (Sr2) minus403702 3112 bthermocomV8R6Mg-lizardite minus402824 3340 Mondesir and Decarreau (1987)p-antigorite (Sr3) minus564239a 3479 Gunnarsson et al (2005)Mg-chrysotile (Sr4) minus406809 3669 Bruni et al (2002)Mg-chrysotileasbestos minus562505a 3888 Pfeifer (1977)Antigorite minus6612864 47719 bthermocomV8R6
The 1047297eld boundaries of the underlined phases shown in Fig 6 correspond to atemperature of 11 degC
a In the ΔG f deg column value calculated using the ΔG f
deg data of the elements from llnldat
and logK value in the reference in the LogK column value calculated using ΔG f deg
(mineral) in the reference and ΔG f deg data of the elements from llnldat
b Thermodynamic database of Lawrence Livermore National Laboratory included in
PHREEQCI and Geochemists Workbench softwares
85T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
decreases The waterndashrock interaction in the TarondashCeno basin iscomplicated by the presence of secondary mixed-layer clay mineralsthe origin of which is quite dif 1047297cult to model because the solidndashstatetransformations and weathering reactions are strictly related tohydrological and climatic conditions as described above An equili-
brium-based reaction path model is a valuable tool for assessing the
most probable reactions and for de1047297ning the product phases that canbe used in a kinetics-based model Since the kinetics of precipitationof Mg-clays at room temperature and pressure are poorly knownrelative reaction rate effects in the reaction path model can be mimedby omitting the stabler phase when a metastable mineral is expected
to precipitate Therefore the precipitation of MgndashFe3+
Abbreviations are as follows Fer-6line 6-line ferrihydrite kaol kaolinite verm vermiculite magn magnetite sap saponite serp serpentine clin clinochlore pyr pyrite calccalcite hydrom hydromagnesite bruc brucite enst enstatite The chromium decrease during the change in water facies from bicarbonate to alkaline was modeled withsupersaturation (precipitation) of eskolaite and thereafter with equilibrium with chromite (Boschetti 2003)- not bound dagger chlorite mineral from solution Dagger chlorite mineral added to the paragenesis
86 T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
montmorillonites vermiculite and saponites are predicted to occurrather than that of Mg-clays Moreover in order to simulate the slowprecipitation of calcite in depth precipitation was not activated atslight supersaturation (SIcalcite=001) during the initial stages of thewaterndashrock interaction in this manner saturation index of calcite wasfree to increase in the following steps and in alkaline waters
6 Discussion of the model results
61 Explanation of pre-modeling choices and post-modeling results
The results of the reaction path modeling are reported in Table 5and shown in Figs 6 and 9 The reaction paths model the dissolutionof a typical local serpentinite (mineralogically constrained as inTable 3) and the sequential precipitation of gibbsite kaoliniteferrihydrite vermiculite Fe2+ndashMg2+-saponite and a poorly-crystal-line serpentine( p-antigorite of Gunnarsson et al2005) The1047297rst stepinvolves the reaction of the ldquoinitial waterrdquo with serpentinite atconstant temperature (11 degC) PO2 (10minus26 bar) and PCO2 (10minus30 bar)this early stage corresponds to bicarbonate groundwater feeding theinvestigated spring system At the stage of primary magnetiteequilibrium and Fe2+ndashMg2+-saponite supersaturation the modelassumes the system to be closed to CO2 and to have a 1047297xed partialpressure of oxygen (PO2= 10minus40 bar) these conditions represent
when water switch from a bicarbonate to an alkaline compositionThe calculated amount of interacting serpentinite needed to obtainthe alkaline springs of the TarondashCeno Valleys (mean PCO2= 10minus7 bar)is approximately 00203 moles (see Table 5)
Bruni et al (2002) explained the origin of the alkaline waters of the Voltri Group as the result of the evolution of meteoricbicarbonate water following the precipitation of sepiolite withabundance similar to that of goethite This might be supported bythe observation that sepiolite or smectite precipitate at lowtemperature from solutions with low Al content when the pH is inthe range 75ndash95 or N95 respectively (Siffert and Wey 1962 Jonesand Galan 1988) but sepiolites form from the weathering of serpentines in arid climate regions whereas sepiolites are rare inhumid climate regions where serpentinite changes to Fe-smectite
(Singer 1989 Wrucke 1996) In the ophiolite rocks of the Northern
Apennine sepiolite is found only sporadically in veins (Giuseppetti1953 Veniale 1966) and is unknown in the ophiolites of the Tarondash
Ceno Valleys whereas it is thought that sepiolite of the Voltri Groupprecipitated from metasomatic 1047298uids during development of shearzones (Crispini and Capponi 2002)
As described above our model explains the transition frombicarbonate water to alkaline water (UM15 and PR10 samples) asthe result of low-temperature precipitation of a poor crystalline
serpentine with the solubility product of p-antigorite (see Sr3 inTable 4 and Fig 6) The model indicates that the bicarbonate waters of the Voltri Group with the highest [Mg2+][H+]2 activity ratios similarlyevolve towards alkaline waters (Fig 6B) The logK value of thesynthetic low temperature serpentine adopted in this model and its
Fig 7 Saturation indices of chrysotile and p-antigorite (Sr2 and Sr3 thermodynamic
data respectively in Table 4) for bicarbonate and alkaline waters coming fromultrama1047297c rocks of the TarondashCeno Valleys as a function of SiMg molar ratioThermodynamic data on chrysotile and p-antigorite are from the thermocomV8R6
database and Gunnarsson et al (2005) respectively
Fig 8 Activityplot forthe MgOndashSiO2ndashAl2O3ndashFe2O3(FeO)ndashH2O system analyzed forSiO2-phases and aqueous species at different log PO2 and log[aAl3+(aH+)3] values Springwaters coming from the TarondashCeno Valleys (symbols as in Fig 2) and Voltri Groupultrama1047297tes (symbols as in Fig 5) are shown The greyarea refers to bicarbonate watersfrom theVoltri Group (Mariniand Ottonello2002Bruniet al 2002) Thesepiolite 1047297eldis very narrow or hidden by saponite-Mg (1047297eld limits shown as continuous lines)saponite-Fe3+ndashMg (1047297eldlimits shown as dottedlines) or vermiculite(1047297eldlimits shownas dottedndashdashed lines) The talc stability 1047297eld and boundaries with saponite-Mg
montmorillonite-Mg and cronstedite-Mg are shown for comparison (dashed lines)
87T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
relative uncertainty of plusmn2 (see Fig 8 in Gunnarsson et al 2005) arecomparable to the logK variation reported for other low-temperatureMg-serpentines (Table 4) that is within the lines Sr1ndash4 for serpentineminerals where samples varying from bicarbonate composition toalkaline composition in the ultrama1047297c aquifer are grouped (Fig 6)Fig 7 shows a plot of calculated serpentine saturation index as a
function of the SiMg molar ratio for the water samples from theultrama1047297tes of the TarondashCeno ValleysThe SI andSiMgvalues of pureMg-serpentine (SI= 0 SiMg=067) provide a useful index forseparating bicarbonate waters (SIb0 SiMgb067) from alkalinewaters (SIN0 SiMgN067) and the distribution of data points isconsistent with the idea that bicarbonate waters approach saturationwith serpentine until they become supersaturated as alkaline waters
Based on the Mgvs SiO2(aq) activitydiagram and speci1047297cratiosoflog[aAl3+(aH+)3] Birsoy (2002) analyzed the behavior of sepiolite duringultrama1047297te weathering The diagram produces a very narrow stability1047297eldfor sepiolite(Fig 8) when it is programmedwith suitable conditionsfor the bicarbonate waters (log PO2=minus25 bar log [aAl3+(aH+)3]= 65) andthe alkaline waters (log PO2=minus50 bar log [aAl3+(aH+)3]= 45) sampled inthe present study This stability 1047297eld grows larger when the absence of
Mg-smectites and talc is assumed but this is an unrealistic conditionbecause (i) all these minerals occur in the rocks within the areainvestigated and (ii) at high [Mg2+][H+]2 activity ratios the 1047297eld of sepiolite is smaller than that of subordinate compared to those poorly-crystalline serpentine (p-antigorite) and clinochlore
Among the alkaline springs of the Voltri Group the waters withlow log [aCa2+(aH+)2] values (Fig 9) show patterns compatible withchlorite interaction (step 5c in Fig6D)The alkaline waters whose datafall near the clinochlore-14Aring 1047297eld (Fig 6D) cannot be explained bysepiolite precipitation because the sepiolite line does not cross thechlorite 1047297elds Thus the peculiar composition of some alkaline watersof the Voltri Group is probably due to interactions between water andprimary chlorites of the ultrama1047297c rocks (optional step 5c in Table 5cb) This is supported by the occurrence sometimes in large amounts
of metasomatic Mgndash(Fe)-chlorite (Scambelluri and Rampone 1999
Vissers and Strating 2001) in the Ligurian area where the relativeabundance of chlorite can reach 50 in veins in high strainserpentinite mylonites of the ErrondashTobbio ultrama1047297tes (Scambelluriet al 2001b) Otherwise in a chlorite-free primary paragenesis thesolution evolving on the Sr3 serpentine line reaches equilibrium withclinochlore but probably the latter does not precipitate due to theslow kinetics of precipitation under these conditions Thus thedissolution of additional reacting rock in the iterative weathering
process supplies Mg
2+
to the solution thereby favoring the Sr3 to Sr4poorly crystalline serpentine transformation In fact all the PR10samples and some of the alkaline springs from the Voltri Group showincreasing [Mg2+][H+]2 activity ratio and low variation in silicaactivity (Fig 6D) starting from the invariant triple point 4a (Sr3-serpentine Fe2+ndashMg-saponite chlinoclore-14 Aring saturation) andextending towards the brucite line This path can be explained asthe result of the transformation of poorly crystalline serpentine fromSr3 to Sr4 the latter is more stable at higher [Mg2+][H+]2 activityratios Finally at very low PO2 values vermiculite disappears and thestability 1047297elds of chlorite and Fe2+ndashMg-saponite become narrowerwhereas those of Fe2+-rich phases (eg Fe2+-saponite) become widerThe data for the alkaline waters with the highest log [aMg2+(aH+)2]from the Voltri Group fall near the stability 1047297eld of peridotite primaryminerals like forsterite and enstatite (Fig 6D) therefore the last stepsof theevolution involve theequilibration of thesolutionwithenstatite(steps 8a and 8c in Table 5ab)
62 Ca-rich alkaline waters
In preliminary modeling of the evolution of the TarondashCeno springwaters Boschetti (2003) concluded that meteoric water initiallysaturated in Fe3+ndashMg2+-chlorite Fe3+ndashMg2+-saponite and pyrite (andor goethite) and then interacted with the ultrama1047297te mineralassemblage eventually evolved into the alkaline waters In themodel of Boschetti (2003) chlorite saponite and calcite were added
Fig 9 Activity plot for the MgOndashCaOndash(CO2)ndashH2O system with phase limits calculated ata temperature of 11 degCSpring waterscoming from theTarondashCeno Valleys (symbols as inFig2) andVoltri Groupultrama1047297tes(symbolsas in Fig5) areshown Calculated path forwaterndashultrama1047297c rock interaction is shown (dotted line) with the number ndashlettercombinations referring to the appearance or disappearance of minerals (see Table 5)The grey area refers to bicarbonate waters from the Voltri Group (Marini and Ottonello2002)
Fig 10 Plots of (A) calcite saturation indexes and (B) PCO2 as a function of Clminus ascalculated by PHREEQCI software for waters from the TarondashCeno Valleys Alkalinesprings for the Voltri Group are also shown (small black diamonds Marini andOttonello 2002) The grey shading is deeper(A) in the saturation1047297eld(SI= 00plusmn05) and
(B) where the lowest log PCO2 value is approached
88 T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
to the solution in order to achieve a high [Ca2+][H+]2 activity ratio andsupersaturation in these phases However in light of the work of Clarket al (1992) and Bruni et al (2002) we believe that the calcitesupersaturation in alkaline waters (Fig 9 and 10) which in our modelis achieved by preventing the precipitation of calcite in the 1047297rst stepsof the reaction path (Table 5a and b) depends on the differencebetween CO2 partial pressureat thegivendepth(PCO2le10minus10bar) andat the surface (PCO2cong10minus35) Waters rising to the surface are far from
equilibrium with the atmosphere where the PCO2 is 4times10
6
timesgreater than the value at depth thus the water rapidly take up CO2which is promptly converted to CO3
2minus because of the high pH Thisleads to supersaturation with calcite through the following hydro-xylation reaction
As a result large amounts of travertine are generated in the VoltriGroup (Marini and Ottonello 2002) and carbonate concretionsknown as devilrsquo s coins are formed at Mt Prinzera in Taro valleyBelow the surface the SIcalcite of the alkaline waters is expected to belower than what the model estimates Under conditions of lowcarbonate alkalinity low CO2 and O2 partial pressure and high pHcalcite nucleation may be inhibited and the kinetics of precipitationreduced by the high concentrations of dissolved divalent cations suchas Mg2+ Fe2+ and Mn2+ (Sumner and Grotzinger 1996 Davis et al2000 Zhang and Dawe 2000 de Leeuw 2002) The fact that severalalkaline water samples fall within or near the calcitendashdolomiteequilibrium (Fig 9) may indicate magnesium calcite or protodolomitenucleation in those extreme subsurface conditions However possibleinteraction between some calcium-rich waters of the Voltri Group andCa-rich rocks such as ophicarbonates or rodingites cannot beexcluded these rocks are widespread in the Ligurian Apennineserpentinites (Treves and Harper 1994)
63 Sepiolite vs serpentine nucleation under weathering conditions
constraints on the kinetics and equilibrium-based approaches
The kinetics of precipitation of poorly-crystalline serpentine and
sepiolite cannot be compared quantitatively because rate constantsare available only for well-ordered crystalline phases (eg see Palandriand Kharaka 2004 for crystalline serpentine) Sepiolite precipitationis often invoked to explain magnesium scavenging from solutions butthe lack of kinetic data for this mineral leads to overly simplisticreaction transport models These models are forced to use kinetic datafor phyllosilicates such as kaolinite or muscovite (eg Xu et al 2001Bryan 2005) Moreover the low-temperature thermodynamic dataeven if they are fairly uncertain show that sepiolite solubilityincreases with decreasing crystallinity (Table 4 Stoessell 1988) Infact crystalline sepiolite requires a long time to reach equilibriumwith solution which imposes a slow control on thewater composition(eg Stoessell and Hay 1978) Consequently in modeling based onreaction transport that simulates freshly precipitated or poorly
crystalline sepiolite formation the logK value of Wollast et al (1968)is preferable to the lower values (Suarez and Šimůnek 1996) Thispreference agrees with the approach of Mariner (2003) and Jove Colon(2006) who adopt a logK (25 degC) value for sepiolite of 3044 It is alsosimilar to that reported by Wollast et al (1968) and it is six logK unitshigherthan the valueused in thermocomV8R6 dataset which is oftenused (perhaps improperly) to model the supergenic nucleation of sepiolite In the stability diagram in Fig 6 where the Sp1 and Sp2 linesrepresent the stability boundaries of crystalline and poorly-crystallinesepiolite respectively the evolution of TarondashCeno waters frombicarbonate to alkaline (Figs 6A and C) follows the stability 1047297eldboundary of the poorly-crystalline serpentine (between Sr2ndashSr3 lines)rather than that of the poorly-crystalline sepiolite (Sp2) as postulatedin themodel The lines for Sp2 Sr3 and amorphoussilica intersect at a
common point (Fig 6D) which is consistent with the nucleation of
sepiolite that typically occurs in solutions that are saturated orsupersaturated with amorphous silica and (Mg)-carbonate (egStoessell and Hay 1978 Birsoy 2002) Some bicarbonate watersfrom the Voltri Group have a silica activity near the amorphous silicaboundary (Fig 6B) but only a few alkaline springs of this group fall onthe poorly-crystalline sepiolite boundary line (Sp2 Fig 6D) Both of these features are inconsistent with the possibility that sepioliteprecipitates from supergenic fresh water in the Northern Apennine
Moreover these two results suggest that values for logK(sepiolite)b
3044 (see Table 4) should be used to model the dissolutionprecipitation of the well-crystallized phase that is the most commonform for the metasomatic sepiolite found in the investigated area
7 Conclusions
The waters issuing from ultrama1047297tes in the TarondashCeno Valleysoriginate from low-temperature interactions between meteoric waterand serpentinite rocks In this report the waterndashrock interaction hasbeen modeled using reaction paths in progress mode During their earlyinteraction with soil and rock minerals bicarbonate waters super-saturated in kaolinite ferrihydrite vermiculite and Fe2+ndashMg2+-saponitewhere theorigin of the secondary iron saponite as Fe2+-bearing smectiteagrees with the hypothesis of Guumlven (1988) Then while remainingsupersaturated with saponite the bicarbonate waters evolve into analkaline facies when poorly-crystalline serpentine precipitates Duringrock weathering the Al-bearing crystalline serpentine of oceanic originprobably promotes nucleation of the low-temperature serpentineamorphous phase and supplies Al3+ to the solution for saponiteprecipitation This interpretation is consistent with the existence of Al-poor supergenic serpentine (Neal and Stanger 1985) and the partiallymeteoric OH-isotopic imprint observed in other ultrama1047297c systems
Dif 1047297culties and complexities in the modeling arise fromuncertainties in the primary mineralogy of some local ophiolitesand from the presence of different lithotypes (rodingites ophi-carbonates gabbros) which can signi1047297cantly affect the composi-tion of water samples This can be seen in the example of the Candash
OH springs issuing from the Voltri Group (PO2 and PCO2 of 10minus70
and 10minus10 bar respectively Bruni et al 2002 Marini and
Ottonello 2002) These springs appear to be more evolved thanthe TarondashCeno alkaline waters as a result of their particularly longinteractions with serpentinites until equilibrium with enstatite isreached or with chlorite-bearing serpentinite The use of anupdated thermodynamic database in our model avoided problemsdue to mineralogical heterogeneity in the mixed-layer claysassociated with ophiolitic rocks this database accurately re1047298ectedthe chemical composition of the single phases Using this modelthe vermiculitendashchlorite and saponitendashchlorite associations seen inthe activity diagrams simulate respectively the presence of highand low charge corrensite observed in the 1047297eld The accuracy of ourmodel can be tested by comparing estimates of chlorine composi-tion resulting from our model and from experimental methods The
mean Cl concentration of the bicarbonate water is 41plusmn16 mgLtaking into account a Cl concentration of sim2 mgL in rainwater inthe area (cf Panettiere et al 2000 Venturelli 2003) 2 mgL of Clin spring water are estimated to come from serpentinite This valueagrees with the model which predicts that the switch frombicarbonate water to alkaline water requires the congruentdissolution of 00203 mol of serpentinite containing 100 ppm of soluble Cl thereby yielding water containing 2 ppm of Cl
In addition we think that weathering of the accessory albiteplagioclase is responsible for Na enrichment of the waters whereas Clenrichment in the alkaline waters is predicted to result from theleaching of NaCl 1047298uid inclusions Moreover the TarondashCeno alkalinesprings show the highest BCl ratios comparable with those of thehost metamorphic minerals (Scambelluri et al 2004) In contrast the
lower boron content and BCl ratio of the alkaline springs of the Voltri
89T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
Group are unexpected features and probably re1047298ect waterndashrockinteractions over a longer time period than in the TarondashCeno ValleysThe boron content in the Northern Apennine alkaline springs may becontrolled by several factors such as (i) the different histories of VoltriGroup and EL ultrama1047297tes (ii) the probable contribution of boronfrom NaCl 1047298uid inclusions (iii) the sinking of boron by travertineprecipitation (Hobbs and Reardon1999) brucite (Rhoades et al 1970)and smectite and (iv) decreases in boron release from minerals due to
increases in pH (Su and Suarez 2004) Acknowledgements
This research was carried out with the 1047297nancial support from theldquoYoung Researchers Projectrdquo (2001ndash2002) of the Italian Ministry of Universities Scienti1047297c Research and Technology (MURST) We thankGianniCortecci (Institute of Geosciences and EarthResources IGG-CNRPisa Italy) and Giampiero Venturelli (Earth Sciences Department Univer-sity of Parma Italy) for useful suggestions and critical review of themanuscript The authors are grateful to Enrico Dinelli (InterdepartmentResearch Centre for Environmental Sciences Alma Mater StudiorumUniversity of Bologna Ravenna Italy) for granting permission to use theGeochemistrsquos WorkbenchsoftwareTheconstructive reviews of theEditor
J Fein and of two anonymous Referees have signi1047297cantly improved the
manuscript
Appendix A Supplementary data
Supplementary data associated with this article can be found inthe online version at doi101016jchemgeo200808017
References
Abbate E Bortolotti V Passerini P 1980 Olistostromes and olistoliths In Sestini G(Ed) Development of the Northern Apennines geosyncline Sedimentary Geology4 pp 521ndash558
Alietti A 1956 Il minerale a strati misti saponite-talco di monte Chiaro (val di TaroAppen Emil) Rendiconti dellAccademia Nazionale dei Lincei 21 201 ndash207
Alietti A De Angelis G Montanari G Sgarlata F Tirelli G 1976 Il ferro nelle argilleappenniniche Periodico di Mineralogia 45 51ndash64
Anselmi B Mellini M Viti C 2000 Chlorine in the Elba Monti Livornesi and Murloserpentines Evidence for seandashwater interaction European Journal of Mineralogy12137ndash146
Apha-Awwa-Wef 1995 Standard Methods for the Examination of Water and Waste-water 19th Edition American Public Health Association American Water WorksAssociation Water Environment Federation
Barnes I LaMarche Jr VC Himmelberg GR 1967 Geochemical evidence of present-day serpentinization Science 156 830ndash832
Barnes I ONeil JR 1969 The relationship between 1047298uids in some fresh alpine-typeultrama1047297cs and possible modern serpentinization Western United StatesGeological Society of America Bulletin 80 1947ndash1960
Barnes I ONeil JRTrescases JJ1978 Present dayserpentinizationin New CaledoniaOman and Yugoslavia Geochimica et Cosmochimica Acta 42 144ndash145
Barnes I Rapp JB ONeil JR Sheppard RA Gude III AJ 1972 Metamorphicassemblages and the direction of 1047298ow of metamorphic 1047298uids in four instances of serpentinization Contributions to Mineralogy and Petrology 35 263ndash276
Barnes JDSelverstone J Sharp ZD 2006Chlorine isotope chemistry of serpentinitesfrom Elba Italy as an indicator of 1047298uid source and subsequent tectonic historyGeochemistry Geophysics Geosystems 7 doi1010292006GC001296
Beccaluva L Emiliani F Venturelli G Zerbi M 1973 Ca Fe Mg Mn Cr Ni Codistribution in some ultrama1047297c rocks outcropping in the Northern Apennines withsome geological remarks Ateneo Parmense Acta Naturalia 9 69ndash98
Beccaluva L Macciotta G Piccardo GB Zeda O 1984 Petrology of lherzolitic rocksfrom the Northern Apennine ophiolites Lithos 17 299 ndash316
BeccaluvaL Venturelli G ZanzucchiG 1975 Dati geologici e geochimici sui basalti diassociazione o1047297olitica dellAppennino ligure-emiliano Ateneo Parmense ActaNaturalia 11 789ndash802
BethkeCM 2002 The Geochemists Workbench A Userrsquos Guide to RxnAct2Tact ReactandGtplot Release40 HydrogeologyProgram Universityof IllinoisUrbanaIL 236pp
Birsoy R 2002Formation of sepiolitendashpalygorskite and related minerals from solutionClays and Clay Minerals 50 736ndash745
Bonatti E Lawrence JR Morandi N 1984 Serpentinization of oceanic peridotitestemperature dependence of mineralogy and boron content Earth and PlanetaryScience Letters 70 88ndash94
Bonifacio E Zanini E Boero V Franchini-Angela M 1997 Pedogenesis in a soilcatena on serpentinite in north-western Italy Geoderma 75 33ndash51
Boschetti T 2003 Studio geochimico e geochimico-isotopico di acque a composizione
estrema e termali dellrsquoAppennino Settentrionale Ph D Thesis Parma Univ
Brigatti MF Franchini G Lugli C Medici L Poppi L Turci E 2000 Interactionsbetween aqueous chromium solutions and layer silicatesApplied Geochemistry 151307ndash1316
Brigatti MF Lugli C Poppi L Venturelli G 1999 Iron-rich saponite dissolutionreactions and Cr uptake Clay Minerals 34 637ndash645
Brigatti MF Valdregrave G 1996 Mixed-layer chloritesmectite (corrensite) in ophioliticrock alteration products Neues Jahrbuch fuer Mineralogie Monatshefte 1 37 ndash47
Bruni J Canepa M Chiodini G Cioni R Cipolli F Longinelli A Marini L OttonelloG Vetuschi Zuccolini M 2002 Irreversible waterndashrock mass transfer accompany-ing the generation of the neutral MgndashHCO3 and high-pH CandashOH spring waters of the Genova province Italy Applied Geochemistry 17 455ndash474
Bryan CR 2005 Drift-Scale THC Seepage Model MDL-NBS-HS-000001 REV 04 YuccaMountain Project Las Vegas Nevada (US)Caruso LJ Chernosky JV 1979 The stability of lizardite Canadian Mineralogist 17
757ndash769Clark ID Fontes JC Fritz P 1992 Stable isotope disequilibria in travertine from high
pH waters Laboratory investigations and 1047297eld observations from Oman Geochi-mica et Cosmochimica Acta 56 2041ndash2050
Cortesogno L 1980 Il metamor1047297smo giurassico nelle o1047297oliti dellAppenninoSettentrionale due distinti cicli metamor1047297ci in ambiente oceanico O1047297oliti 5 5ndash18
Christ CLHostetler PB Siebert RM1973 Studies in the system MgO-SiO2-CO2-H2OIII American Journal of Science 273 65ndash83
Cortesogno L Lucchetti G 1982 Il metamor1047297smo oceanico nei gabbri o1047297oliticidellAppennino Ligure Aspetti mineralogici e petrogenetici Rendiconti dellaSocietagrave Mineralogica Italiana 38 561ndash579
Crispini L Capponi G 2002 Control of metasomatic alteration on the development of shear zone in ultrama1047297c rocks a case study from the Voltri Massif (Ligurian Alps)Architettura interna e cinematica delle zone di taglio Gruppo Italiano di GeologiaStrutturale mdash Riunione Annuale 2002 Pisa Italy 11ndash12 giugno 2002 XIV ndashXV
CrissRE 1999 Principlesof stableisotope distributionOxford University PressNew York
Davis KJ Dove PM De Yoreo JJ 2000 The role of Mg 2+ as an impurity in calcitegrowth Science 290 1134ndash1137
de Leeuw NH 2002 Moleculardynamics simulations of the growth inhibiting effect of Fe2+ Mg2+ Cd2+ and Sr2+ on calcite crystal growth Journal of Physical Chemistry B106 5241ndash5249
De Nardo MT Segadelli S Vescovi P 2007 Studio pilota per la caratterizzazionegeologicadellesorgentinellazona delM Nero (altaVal Cenoe alta ValNuremdash provincedi Parma e Piacenza) Il Geologo dellEmilia-Romagna 25 Nuova Serie pp 5 ndash21
Di Giulio A Geddo G 1990 Studio petrogra1047297co delle Arenarie di Casanova (alta valTrebbia Appennino settentrionale) Atti Ticinesi di Scienze della Terra 33 243ndash254
Dinelli E Lombini A Simoni A Ferrari C 1997 Heavy metals in serpentinite soils of selected outcrops of Piacenza and Parma provinces (Northern Apennines Italy)Mineralogica et Petrographica Acta 40 241ndash255
Epstein S Mayeda T 1953 Variation of 18O content of waters from natural sourcesGeochimica et Cosmochimica Acta 4 213ndash224
Ernst WG Piccardo GB 1979 Petrogenesis of some Ligurian peridotites I Mineraland bulk-rock chemistry Geochimica et Cosmochimica Acta 43 219ndash237
Feder F Trolard F Klingelhofer G Bourrieacute G 2005 In situ Moumlssbauer spectroscopy
evidence for green rust (fougerite) in a gleysol and its mineralogical transforma-tions with time and depth Geochimica et Cosmochimica Acta 69 4463ndash4483Fritz PClark IDFontes JCWhiticar MJFaberE 1992 Deuterium and
13C evidence
for low temperature production of hydrogen and methane in a highly alkalinegroundwater environment in Oman In Kharaka TK Maest AS (Eds) Proceed-ings of the 7th International Symposium on WaterndashRock Interaction BalkemaRotterdam pp 793ndash796
Giammetti F 1968 Le o1047297oliti di Roccaprebalza Costa della Guardia e Groppo di Gorro(Appennino Parmense) Ateneo Parmense Acta Naturalia 4 107ndash142
Giuseppetti G1953 Roccee minerali dellaformazioneo1047297olitica di Volpedo Rendicontidella Societagrave Mineralogica Italiana 9 85ndash134
Gran G 1952 Determination of equivalence point in the potentiometric titrationAnalyst 77 661ndash671
Gunnarsson I Arnorsson S Jakobsson S 2005 Precipitation of poorly crystalline antigoriteunder hydrothermal conditions Geochimica et Cosmochimica Acta 69 2813ndash2828
Guumlven N 1988 Smetites In Bailey SW (Ed) Reviews in Mineralogy 19 Hydrousphyllosilicates (exclusive of micas) Mineralogical Society of America MichiganBook Crafters Inc 497ndash559
Hansel CM Benner SG Fendorf S 2005 Competing Fe(II)-induced mineralization
pathways of ferrihydrite Environmental Science and Technology 39 7147ndash7153HobbsMYReardonEJ 1999 Effect of pH on boronco-precipitation by calcite further
evidence for non-equilibrium partitioning of trace elements Geochimica etCosmochimica Acta 63 1013ndash1021
Jones BF Galan E 1988 Sepiolite and palygorskite In Bailey SW (Ed) HydrousPhyllosilicates (exclusive of Micas) Reviews in Mineralogy vol 19 MineralogicalSociety of America Michigan Book Crafters Inc 631ndash674
Jove Colon CF 2006 Technical work plan for Thermodynamic database for chemicalmodelling TWP-MGR-PA-000039 REV 01 Yucca Mountain Project Las VegasNevada (US)
Kohls DW Rodda Jr JL 1967 Iowaite a new hydrous magnesium hydroxidendashferricoxychloride from the Precambrian of Iowa American Mineralogist 52 1261 ndash1271
Kyser TK OHanley DS Wicks FJ 1999 The origin of 1047298uids associated withserpentinization processes evidence from stable-isotope compositions CanadianMineralogist 37 223ndash237
Lapham DM 1961 New data on deweylite American Mineralogist 46 168ndash188Lee BD SearsSKGraham RC Amrhein CValiH 2003 Secondarymineral genesis
from chlorite and serpentine in an ultrama1047297c soil toposequence Soil Science
Society of America Journal 671309ndash1317
90 T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
Longinelli A Selmo E 2003 Isotopic composition of precipitation in Italy a 1047297rstoverall map Journal of Hydrology 270 75ndash88
Majzlan J Navrotsky A Schwertmann U 2004 Thermodynamics of iron oxides PartIIIEnthalpiesof formation and stability of ferrihydrite (simFe(OH)3) schwertmannite(simFeO(OH)34(SO4)18) and ɛ-Fe2O3 Geochimica et Cosmochimica Acta 681049ndash1059
Mariner P 2003 In-Drift PrecipitatesSalts Model ANL-EBS-MD-000045 REV 01 YuccaMountain Project Las Vegas Nevada (US)
Marini L Ottonello G 2002 Atlante degli acquiferi della Liguria Volume III Le acquedeicomplessio1047297olitici (bacini ArrestraBranega Cassinelle Cerusa ErroGorzenteLeira Lemme Lerone Orba Piota Polcevera Rumaro Sansobbia Stura Teiro
Varenna Visone) Pacini PisaMellini L Zanazzi PF 1987 Crystal structures of lizardite-l T and lizardite-2H1 fromColi Italy American Mineralogist 72 943ndash948
Mezzadri G 1964 Petrogra1047297a delle arenarie di Ostia Rendiconti Societagrave MineralogicaItaliana 20 192ndash228
Mondesir H Decarreau A 1987 Synthesegrave entre 25 et 200 degC de lizardites Ni ndashMgMesure des coef 1047297cients de partage solide-solution aqueuse pour le couple Ni ndashMgdans des lizardites Bulletin de Mineralogie 110 409ndash426
Neal C Shand P 2002 Spring and surface water quality of the Cyprus ophiolitesHydrology and Earth System Sciences 6 797ndash817
Neal C Stanger G 1983 Hydrogen generation from mantle source rocks in OmanEarth and Planetary Science Letters 66 315ndash320
Neal C Stanger G 1984 Calcium and magnesium hydroxide precipitation fromalkaline groundwaters in Oman and their signi1047297cance to the process of serpentinization Mineralogical Magazine 48 237ndash241
Neal C Stanger G 1985 Past and present serpentinisation of ultrama1047297c rocks anexample from Semail ophiolite nappe of Northern Oman In Drever JI (Ed) TheChemistry of Weathering D Reidel Publishing Company pp 249ndash275
Nesbitt HW Bricker OP 1978 Low temperature alteration processes affecting
et Cosmochimica Acta 41 1835ndash1841OrsquoHanley DS 1996 Serpentinites records of tectonic and petrological history Oxford
University PressOrberger B Friedrich G Woermann E 1990 The distribution of halogens and carbon
in Pge-bearing ultrama1047297cs of the Acoje Ophiolite Block Zambales Philippines Journal of Geochemical Exploration 37 147ndash169
Ottonello G Joron JL Piccardo GB 1984 Rare Earth and 3d transition elementsgeochemistry of peridotitic rocks II Ligurian peridotites and associated basalts Journal of Petrology 25 373ndash393
Palandri JL Kharaka YK 2004 A compilation of rate parameters of water-mineralinteraction kinetics for application to geochemical modeling Open File Report2004-1068 US Geological Survey
Panettiere P Cortecci G Dinelli E Bencini A Guidi M 2000 Chemistry and sulfurisotopic composition of precipitation at Bologna Italy Applied Geochemistry 151455ndash1467
Parkhurst DL Appelo CAJ 1999 Userrsquos guide to PHREEQC (version 2) mdash a computerprogram for speciation batch-reaction one-dimensional transport and inversegeochemical calculation US Department of the Interior US Geological SurveyWashington DC
Pfeifer HR 1977 A model for 1047298uids in metamorphosed ultrama1047297c rocks Observationsat surface and subsurface conditions (high pH spring waters) SchweizerischeMineralogische und Petrographische Mitteilungen 57 361ndash396
PlummerLN1984Geochemical modeling A comparisonof forwardand inverse methodsIn Hitchon B Wallick EI (Eds) Proceedings First CanadianAmerican Conference onHydrogeology-Practical Applications of Ground Water Geochemistry Banff AlbertaCanada mdash Worthington Ohio National Water Well Association pp 149ndash177
Rampone E Hofmann AW Piccardo GB Vannucci R Bottazzi P Ottolini L 1995Petrology mineral and isotope geochemistry of the External Ligurides peridotites(Northern Apennines Italy) Journal of Petrology 36 81ndash105
Rampone E Hofmann AW Piccardo GB Vannucci R Bottazzi P Ottolini L 1996Trace element and isotope geochemistry of depleted peridotites from an N-MORBtype ophiolite (Internal Liguride N Italy) Contributions to Mineralogy andPetrology 123 61ndash76
Rhoades JD Ingvalson RD HatcherJT1970Adsorption of boronby ferromagnesianminerals and magnesium hydroxide Soil Science Society of America Journal 34
938ndash941Saini-EidukatB KuchaH Keppler H1994Hibbingiteγ-Fe2(OH)3Clanewmineralfromthe Duluth Complex Minnesota with implications for the oxidation of Fe-bearingcompounds and the transport of metals American Mineralogist 79 555ndash561
Sanford RF 1981 Mineralogical and chemical effects of hydration reactions andapplications to serpentinization American Mineralogist 66 290ndash297
Scambelluri M Muumlntener O Ottolini L Pettke TT Vannucci R 2004 The fate of BCl and Li in the subducted oceanic mantle and in the antigorite breakdown 1047298uidsEarth and Planetary Science Letters 222 217ndash234
Scambelluri M Rampone E 1999 Mg-metasomatism of oceanic gabbros and itscontrol on Ti-clinohumite formation during eclogitization Contributions toMineralogy and Petrology 135 1ndash17
Scambelluri M Bottazzi P Trommsdorff V Vannucci R Hermann J Gogravemez-Pugnaire MT Logravepez-Sagravenchez Vizcaino V 2001a Incompatible element-rich1047298uidsreleased by antigorite breakdown in deeply subducted mantle Earth and PlanetaryScience Letters 192 457ndash470
Scambelluri M Rampone E Piccardo G 2001b Fluid and element cycling insubducted serpentinite a trace-element study of the ErrondashTobbio high-pressureultrama1047297tes (Western Alps NW Italy) Journal of Petrology 42 55ndash67
Segadelli S2006 Lageologia nelpaesaggio le rupi o1047297olitiche in Val Taro e Val Ceno IlGeologo dellrsquoEmilia-Romagna 22 15ndash29 Nuova Serie
Settembre Blundo D Martin De Vidales JL Alietti A 1992 Corrensite-type clayminerals from Taro Valley (North Italy) Mineralogica et Petrographica Acta 35215ndash222
Sharp ZD Barnes JD 2004 Water-soluble chlorides in massive sea1047298oor serpenti-nites a source of chloride in subduction zones Earth and Planetary Science Letters226 243ndash254
Siever R Woodford N 1979 Dissolution kinetics and the weathering of ma1047297cminerals Geochimica et Cosmochimica Acta 43 717ndash724Singer A 1989 Palygorskite and sepiolite group minerals In Dixon JB Weed SB
(Eds) Minerals in Soil Environments Soil Science Society of America MadisonWisconsin pp 829ndash872
Siffert B Wey R 1962 Synthese d rsquoune sepiolite a temperature ordinarie Comptesrendus hebdomadaires des seacuteances de lAcadeacutemie des Sciences de Paris 2541460ndash1463
Środoń J 1999 Nature of mixed-layer clays and mechanisms of their formation andalteration Annual Reviews of Earth and Planetary Science 27 19ndash53
Stoessell RK 1988 25 degC and 1 atm dissolution experiments of sepiolite and keroliteGeochimica et Cosmochimica Acta 52 365ndash374
Stoessell RK Hay RL 1978 The geochemical origin of Sepiolite and keorilite atAmboseli Kenya Contributions to Mineralogy and Petrology 65 255ndash267
Suarez DL Šimůnek J 1996 Solute transport modeling under variably saturatedwater 1047298ow conditions In Lichtner PC Steefel CI Oelkers EH (Eds) ReactiveTransport in Porous Media Reviews in Mineralogy vol 34 Mineralogical Society of America Michigan Book Crafters Inc 229ndash268
Sumner DY Grotzinger JP 1996 Were kinetics of Archean calcium carbonate
precipitation related to oxygen concentration Geology 24 119ndash122Su C Suarez DL 2004 Boron release from weathering of illites serpentine shales
and illiticpalygorskitic soils Soil Science Society of America Journal 68 96 ndash105Toscani L Venturelli G Boschetti T 2001 Sul1047297de-free and sul1047297de-bearing waters in
the Northern Apennines Italy general features and waterndashrock interaction AquaticGeochemistry 7 195ndash216
Treves BE Harper GD 1994 Exposure of serpentinites on the ocean 1047298oor sequenceof faulting and hydrofracturing in the Northern Apennine ophicalcites O1047297oliti 19b435ndash466
Varadachari C Kudrat M Gosh K 1994 Evaluation of standard free energies of formation of clay minerals by an improved regression method Clays and ClayMinerals 42 298ndash307
Veniale F 1966 Sepiolite in sedimenti dellrsquoAppennino Vogherese (Pavia) Periodico diMineralogia 35 343ndash386
Venturelli G 2003 Acque minerali ed ambiente mdash Fondamenti di geochimica deiprocessi di bassa temperatura Pitagora Editrice Bologna
Venturelli G Contini S Bonazzi A Mangia A 1997 Weathering of ultrama1047297c rocksand element mobility at Mt Prinzera Northern Apennines Italy MineralogicalMagazine 61 765ndash778
Venturelli G Frey M 1977 Anchizone metamorphism in sedimentary sequences of the Northern Apennines Rendiconti Societagrave Mineralogica Italiana 33 109ndash123
Vieillard P 2002 A new method for the prediction of Gibbs free energies of formationofphyllosilicates (10 Aring and14 Aring)basedon theelectronegativityscale Claysand ClayMinerals 50 352ndash363
Vissers RLM Strating EHH 2001 Structures and microstructures in a thrust-relatedgreenschist facies tectonic melange Voltri Group (NW Italy) O1047297oliti 26 33ndash46
Viti C Mellini M 1997 Contrasting chemical compositions in associated lizardite andchrysotile in veins from Elba Italy European Journal of Mineralogy 9 585 ndash596
Viti C Mellini M 1998 Mesh textures and bastites in the Elba retrogradeserpentinites European Journal of Mineralogy 10 1341ndash1359
Wenner DB Taylor Jr HP 1974 Deuteriumatomic hydrogen and oxygen-18oxygen-16 studies of serpentinization of ultrama1047297c rocks Geochimica et CosmochimicaActa 38 1255ndash1286
Williams LA Parks GA Crerar DA 1985 Silica diagenesis I Solubility controls Journal of S edimentary Petrology 55 301ndash311
Wilson J Savage D Cuadros J Shibata M RagnarsdottirKV 2006 The effect of ironon montmorillonite stability (I) Background and thermodynamic considerations
Geochimica et Cosmochimica Acta 70 306ndash322WollastR MackenzieFTBricker OP1968 Experimental precipitation and genesis of sepiolite at earth-surface conditions American Mineralogist 531 1645ndash1662
Wrucke CT1996 Serpentine and carbonated-hosted asbestos deposits InEdwardAdu Bray (Eds) Preliminary compilation of descriptive geoenvironmental mineraldeposit models 95ndash831 US Geological Survey Open-File Report pp 39 ndash46
Xu T Apps JA Pruess K 2001 Analysis of mineral trapping for CO2 disposal in deepaquifers Lawrence Berkeley National Laboratory University of California
ZhangYDaweRA2000 In1047298uence ofMg2+ on thekineticsof calcite precipitation andcalcite crystal morphology Chemical Geology 163 129ndash138
Zhu C Anderson G 2002 Environmental applications of geochemical modelingCambridge University Press
91T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
The Mg-bicarbonate watersissuingfrom ultrama1047297crocksfallnearthisline and cluster between the clinochlore and forsterite points (Fig 2)Stream waters and springs issuing from sedimentary and basaltic rocksfall between thecarbon apex and theline indicated by a CMg molar ratioof 4 which describes the dissolution of CandashMg-clinopyroxene
This is consistent with the (Ca)-bicarbonate compositions of thesestream waters and springs It is noteworthy that the Ceno streamwaters shifted toward the C apex of Fig 2 derive their chemistry fromthe interaction with Mt Caio Flysch cropping out widely in the Cenocatchment (samples from UM50 to UM59 see Map1 sheet in theannexes) whereas the Taro source spring waters plot towards theother apexes (lower C) typical of ophiolite minerals (samples fromUM41 to UM49 see Map1 sheet in the annexes) The use of the traceelements (eg Sr and Rb content Boschetti 2003) further highlightthe differences between the stream waters of the two rivers and therole of lithology Nonetheless mixing and equilibrium with atmo-sphere make the reaction-path modeling of these samples poorlyreliable so their are unsuitable for the purpose of the present work
Alkaline waters from ultrama1047297c rocks fall inside the same sub-triangle mostly along the CndashSi side probably due to the precipitationof magnesium minerals during the evolution of the water composition(Bruni et al 2002)
42 Sodiumndashchloride relationships
Chlorine and other halogens in serpentinized rocks are inheritedfrom seawater and marine sediments during oceanic metamorphism(eg Orberger et al 1990) They are transferred into serpentinite byadsorption (Anselmi et al 2000 Sharp and Barnes 2004) entrap-ment in 1047298uid inclusions (Scambelluri et al 2001ab) and substitutionin the structure of low-crystalline iron-hydroxide phases These latterare composed of the isomorphous and chemically similar iowaitehibbingite and FendashCl-oxydroxides (also named green rusts orfougerites Kohls and Rodda 1967 Saini-Eidukat et al 1994 Federet al 2005 and references therein) In the Northern Apennineserpentinite rocks the mean Cl content is sim300 ppm mean values of 221 and 375 ppm were measured respectively in the Erro ndashTobbio(Ligurian Western Alps Scambelluri et al 2004) and in the centralItaly serpentinites (Anselmi et al 2000) but thewater-soluble Cl from
serpentinites is estimated to be than 100 ppm (Barnes et al 2006) Asshown in Fig 2 waters from basalts and ultrama1047297tes of the TarondashCenoValleys show a trend towards the Cl apex whereas in Fig 3 theestimated soluble chlorine in serpentinites agrees with the highestvalue reported in spring waters from ultrama1047297tes (samples fromMarini and Ottonello analyzed in 2002) The Cl content of samplesUM1 and UM36is veryhigh(Fig 3) and is probablydue to pollutionInfact these springs are not far from the A15 highway (UM1) and theTomarlo Pass road (UM36) where salt mixtures comprising NaClMgCl2middot6H2O and probably CaCl2 are widely used for deicing in winterAlkaline water with a Mg-bicarbonate composition from ultrama1047297tesshow the highest and lowest Cl contents respectively Na-rich Mg-bicarbonate waters from basalts (UM6 UM7 Fig 3) and ultrama1047297tescan be explained by the dissolution of albite in basalts (Beccaluva
et al 1975) and in spilitized EL peridotites (Beccaluva et al 1984
Rampone et al 1995) respectively whereas the highest content of both Na and Cl in the alkaline water samples UM15 and PR10 (Fig 3)can be explained by leaching of NaCl 1047298uid inclusions in theserpentinite rocks (04ndash20 wt NaCl Scambelluri et al 2004)
43 Boron and BCl ratio
Compared with unaltered ultrama1047297c rocks serpentinites have
higher content of both boron and chlorine and most authors suggestthat both elements come from seawater (eg Sanford 1981 andreferences therein Bonatti et al 1984) Despite several studies onthese rocks the behavior of Cl and B in solution during waterndashrockinteraction processes is not well known Early investigations on 1047298uidsassociated with serpentinites in the western United States (Barnes andONeil 1969 Barnes et al 1972) revealed metamorphic waters withhighboron levels(up to 28mM) and BCl molar ratiosof 0038ndash013 aswell as Mg-bicarbonate and high-pH waters of meteoric origin withvariable BCl ratios of 0027ndash042 (median 0039) and 00006ndash0017(median 00076) respectively Bicarbonate waters issuing fromophiolitic rocks in Cyprus have BCl ratio in this range (00017ndash
0058 median 00019 Neal and Shand 2002) whereas low boron
Fig 4 Plots of (A) δ18O vs δ2H and (B) δ18O vs altitude for waters from the Taro ndashCenoValleys C-Italy and N-Italy are the meteoric water lines for Central and Northern Italy
respectively (Longinelli and Selmo 2003) Symbols are as in Fig 2
81T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
levels were found in the alkaline springs of meteoric origin in OmanNew Caledonia and ex-Yugoslavia (Barnes et al 1978)
Among the waters analyzed so far in the TarondashCeno Valleys thebicarbonate waters have BCl ratios in the range 00017ndash022 (median0073) whereas the alkaline waters show the highest ratios everreported (BCl= 034ndash33 median 169) The high boron content of upto 13 mgL of the alkaline springs could be related to metamorphic1047298uids released during serpentinite subduction in fact rocks prefer-
entially lose Cl and the BCl ratio of the released 1047298
uids increases withpressure and temperature (Scambelluri et al 2004) Nonetheless thealkaline water samples in our study differ signi1047297cantly from alkalinemetamorphic1047298uids in that they have low salinity high BCl ratios andmeteoric isotope signatures (see below) all these characteristicssuggest an origin in the interaction of meteoric water with high-pressure metamorphic serpentinite This hypothesis is furthersupported by the agreement in BCl ratio between the alkalinesamples we study here and the high BCl ratio of up to 4 previouslyreported for the residual high-pressure mineralogy of ultrama1047297tes(olivine+orthopyroxene Scambelluri et al 2004) In contrast theVoltri Groups Ca-hydroxide waters have mean values of B = 014 mgL and BCl=003 (Marini and Ottonello 2002) consistent with acontribution of boron and chlorine from 1047298uid inclusions In factassuming fractional contributions of 0001ndash0005 from NaCl 1047298uidinclusions with boron content from 4 to 8 mgL ( Fig 3a Scambelluriet al 2004) results in a B content of 0004ndash01 mgL in the solutionThese results suggest that the difference between springs from theVoltri Group and TarondashCeno Valleys is probably related to (i) differentevolution of the 1047298uids and (ii) the different types of host rocks and Pndash
T paths of the Ligurian Western Alps and EL peridotites (eg Beccaluvaet al 1984 Scambelluri et al 2004)
44 Isotopic composition of water
A diagram of δ2H vs δ18O (Fig 4A) shows that the waters sampledfrom the TarondashCeno Valleys are meteoric in origin since they areconsistent with the meteoric water lines of central and northern Italy(Longinelli and Selmo 2003) The altitude vs δ18O(H2O) (Fig 4B)
summarizes the isotopic effects due to temperature altitude andamount of rainfall (eg Criss1999) all of which help to determine theisotopic composition of the water samples The tendency is similar forbicarbonate waters from sedimentary basaltic and ultrama1047297c forma-
tions which may provide insight into the recharge areas alkalinewaters in contrast probably circulate deeper in the ground
5 Water ndashrock interaction model
51 Fundamentals
The reaction path model between meteoric water and serpentinite
was performed using PHREEQCI 2125 code (Parkhurst and Appelo1999) and the Lawrence Livermore National Laboratory thermody-namic database (llnldat as known as thermocomV8R6 ) implementedwith the equilibrium constants listed in Table 2
Computations wereperformed by reaction progress mode ( forward
modeling Plummer 1984 Zhu and Anderson 2002) ie (i) addingreacting rock at each step in the simulation (ii) activating the phase(s)precipitation only when the thermodynamic stability of the mineralparagenesis is well known and consistent with the mineralogy of thearea and (iii) plotting reaction progress on speci1047297c activity diagramsThe same thermodynamic database was used with The GeochemistsWorkbench 403 software (Bethke 2002) to generate Pourbaix andactivity diagrams (Figs 5 6 8 and 9) Modeling was performed usingan approach based on equilibrium rather than reaction transportbecause the data on reaction kinetics and reactive surfaces in thesesolid phases systems are scarce particularly with respect to theactivation energy of clay and mixed-layer minerals at low temperatureand high pH Moreover the hydrogeology and the discontinuity of theaquifers in the study area which are related to the geologicheterogeneity and structural complexity of the Northern Apenninewould make a model based on reaction transport even less credible atthis stage
The modeling in this study was performed on the basis of thefollowing knowledge (1) primary and secondary mineral assem-blages (2) chemical composition of the minerals involved (3) physicalparameters and chemical composition of the starting 1047298uid (4)thermodynamic data on pure phases
52 The solid reactant primary paragenesis
EL peridotites of the TarondashCeno Valleys (NorthernApennines Italy)are spinel lherzolites containing olivine (forsterite) orthopyroxene(enstatite) and clinopyroxene (diopside) (Giammetti 1968 and
Table 2
Thermodynamic data on hydrolysis reactions added to the llnldat database of the PHREEQCI code (Parkhurst and Appelo 1999) and thermocomV8R6dat database of TheGeochemists Workbench software (Bethke 2002)
Mineral Reaction log K (TdegC) ΔGdeg f References
0 25 60 100 (kJ mol-1)
clinochlore08-daphnite02 Mg4Fe1Al2Si3O10(OH)8 + 16H+=2Al3+ + Fe2++12H2O+4Mg2++3SiO2 7318 6384 5272 4269 minus786379 this workferr ihydrite 2-l ine Fe(OH)3+3H+= Fe3++ 3H2O 484 355 ndash ndash minus70850 Majzlan et al 2004ferr ihydrite 6-l ine Fe(OH)3+3H+= Fe3++ 3H2O 435 312 ndash ndash minus71100 Majzlan et al 2004liz09-berth01 Mg27Fe02Al02Si19O5(OH)4+64H+=02Al3++02Fe2++52H2O+27Mg2++19SiO2 3111 2786 2383 2005 minus399854 this workp-antigorite Mg
Saponite-Fe2+ Fe3175Si365Al035O10(OH)2+74H+=035Al3++3175Fe2++47H2O+365SiO2 1845 1635 1319 1001 minus452459 Wilson et al 2006Saponite-Fe2+ndashMg Fe0165Mg3Si367Al033O10(OH)2+732H+=033Al3++0165Fe2++466H2O+3Mg2++
367SiO2
3037 2729 2330 1921 ndash Marini and Ottonello(2002)
The Gibbs free energy of formation (ΔGo f ) of the two theoretical Fe3+ (Varadachari et al1994) and Fe2+ vermiculites were calculated using a previously published method (Vieillard
2002 and Vieillard pers comm)log K values extimated considering an ideal solid solution between the end-members and using data thermodynamic data from llnldat (clinochlore08-daphnite02) and Wilson
et al (2006) (liz09-berth01)
82 T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
references therein Venturelli et al 1997) Accessory phases areplagioclase picotite plusmn calciumndashsodic amphibole (Venturelli et al1997) Before modeling the crystal chemical formula of the mineralphases was calculated from data from Venturelli et al (1997) thendepending on the modal composition of the rock (Ernst and Piccardo1979) and the composition of the ideal solid solution of the minerals
the resulting stoichiometry of the solid reactant(s) was calculated andinserted in the input 1047297le of the PHREEQCI software (Table 3)Thermodynamic data of the primary mineral paragenesis are all wellknown and are included in the llnldat thermodynamic database of PHREEQCI
53 The solid reactant secondary paragenesis
Secondary minerals in the ultrama1047297tes of the TarondashCeno Valleysare serpentine (lizardite and rare chrysotile) smectite minerals (Fendash
Mg saponites) and Fe-oxi-hydroxides Other accessory minerals arechlorite mixed layer saponitendashchlorite (low-charge corrensite) talccalcite and dolomite (Beccaluva et al 1973 Dinelli et al 1997Venturelli et al 1997 Brigatti et al 1999 2000) In addition mixed
layer saponitendashtalc (aliettite) and vermiculitendashchlorite (high-charge
corrensite) are present (Alietti 1956 Settembre Blundo et al 1992Brigatti and Valdregrave 1996) Except for serpentine themodal abundanceof these phases is unknown In addition some of these minerals ndash
particularly interlayer clays ndash may have different origins (hydrothermalor weathering allogenicor authigenic)and different stabilitiesrelatedtokinetics and drainage 1047298ow rate For example chlorite weathering yields
vermiculite-bearing or saponite-bearing interlayers under good- andpoor-drainage regimes respectively (Bonifacio et al 1997 Środoń1999) In addition the thermodynamic data available for serpentineinclude information only on chrysotile which is less abundant thanlizardite in the serpentinites of the areas that we studied According toVenturelli et al (1997) the lizardite composition in the serpentinites of the Taro valley can be summarized as
which is similar to 1-T lizardite from the Monte Fico serpentinizedperidotite (Elba Island Northern Tyrrhenian Sea Italy Viti and Mellini1997)
Using the thermodynamic data of Wilson et al (2006) and
assuming an ideal solid solution between lizardite and berthierine
Fig 5 EhndashpH iron diagram of the spring waters coming from ultrama1047297c rocks of the TarondashCeno Valleys Bicarbonate and high-pH waters of the Voltri Group are also represented(dotted area andsmallblackdiamonds respectively) Thelowest Eh value of thehigh-pH waterswas calculated using theS(+VI)S(minus II)redox couple(Marini and Ottonello 2002)Thestability 1047297eld of mineral phases are in grey and those of aqueous species in white In diagrams A and B 6-line ferrihydrite-vermiculite and goethite were selected as subaerial solidphases respectively
83T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
we calculated ΔGf 0=minus399854 kJmol for a theoretical (09)(lizardite)+
(01)(berthierine) mineral (Table 2) The log K value of the stated solidsolution was calculated at different temperatures using PHREEQCIafter loading the thermodynamic data on the two end-members intothe database The log K results obtained for Al-lizardite were added tothe thermodynamic database of the solid reactant (Table 3) andcompared with data from the literature (Table 4)
54 Composition of the starting solution O 2 partial pressure and
minerals included in the model
The weighted average chemical composition of rainwater inBologna (Panettiere et al 2000) which is 95 km east of theinvestigated area was used as the starting solution in the modelFull chemical data on rainwater of the TarondashCeno Valleys are
unavailable
Total alkalinity was recalculated for log PCO2=minus35 bar and
T =11 degC which was the mean temperature of the springs sampledand for pe (or Eh) based on chemical data for the redox couple N(minusIII)N(+V) The values of pH and pe obtained were 566 and 452(Eh=025 V) respectively the former agrees with the median valuereported for rainwater in thearea (Venturelli et al1997) Because dataon Al Fe and Si in rainwater were lacking we calculated theconcentrations of these elements assuming logPCO2=minus30 bar andslight supersaturation relative to ferrihydrite-6 line and kaolinite Inthis way a hypothetical composition of water in equilibrium with thelocal soil was calculated (Table3) and adopted as the ldquoinitial waterrdquo formodeling
Evidence suggests that secondary minerals particularly Fe-bearingphases change the chemical composition of the water 1047298owingthrough the ultrama1047297tes of the TarondashCeno Valleys According to Neal
and Stanger (1983) and Fritz et al (1992) the production of CH4 and
Fig 6 Activity plotfor the MgOndashSiO2ndashAl2O3ndashFe2O3(FeO)ndashH2O systemat differentlog PO2 values Spring waters issuing from TarondashCeno Valleys (symbols as in Fig 2) andVoltri Groupultrama1047297tes are shown (grey area bicarbonate waters small black diamonds alkaline waters from Marini and Ottonello 2002) In plots A B and C the stability 1047297eld limits of vermiculite(Table 2) correspond to25 degCwhilethe amorphoussilica (am-SiO2) brucite serpentine (Sr1ndash4 in Table 4)and sepiolite(Sp1ndash2 in Table 4) phasescorrespondto 11 degCThepath calculatedfor interaction between waterand ultrama1047297c rockis also shown (dotted line)the number-letter combinations referto the appearanceand disappearance of minerals(see Table 5)
84 T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
H2 in thealkalinewaters1047298owing from theultrama1047297tes isrelatedto theconsumption of oxygen during the oxidation of iron forming thehematitendashmagnetite pair Therefore the activity of iron and theoxidation state of the water must be de1047297ned The analysis of watersamples from ultrama1047297tes were plotted in both the Pourbaix andactivity diagrams (Figs 5 and 6) where the calculated activity of Fe2+Mg2+ SiO2 SO4
2minus (Pourbaix) and log PO2 (activity diagrams) wereinvolved Iron concentration was quite variable in the TarondashCenosprings where the average content was 27 μ gL and the contentreached a maximum of 200 μ gL and in the Voltri springs where theaverage was 44 μ gL and the maximum was 948 μ gL (Marini andOttonello 2002) In both cases the variations are probably related to
the different mineralogies of the soils and the different solubilities of the Fe-rich phases like goethite and ferrihydrite Iron-bearingsaponites and montmorillonites are the prevailing clay minerals inthe area investigated probably because of the ubiquitous presence of iron oxy-hydroxide (Alietti et al 1976 G Venturelli pers comm) Forthis reason the model considers the involvement of pure Mg-smectiteto be negligible
In the Pourbaix diagram (Fig 5) the data from bicarbonate watersfall into the ferrihydrite-goethitendashvermiculite1047297elds and the data fromalkaline waters fall on the boundaries of the vermiculitesaponite andsaponitechlorite Since ferrihydrite is the 1047297rst precipitating Fe-phaseit was used rather than goethite in the modeling The diagram alsosuggests the transformation of ferrihydrite into magnetite during thechange from bicarbonate to alkaline composition but this process
requires a large amount of Fe2+ in solution (N2 mM Hansel et al2005) which is unlikely at this stage of the waterndashrock interactionThus the magnetite participating in the equilibrium is most likelyprimary in origin
The modeling and the activity diagrams (Figs 6 and 9) werecalculated using the average value PO2= 10minus26 or 10minus50 bar forbicarbonate and alkaline waters respectively These values arecompatible with those calculated by Marini and Ottonello (2002) forbicarbonate waters (PO2= 10minus40 bar) and for alkaline waters of theVoltri Group (10minus77 barbPO2b10minus42 bar) The lowest PO2 value whichwas used as a limit in the modeling was calculated from the S(+VI)S(minusII) redox couple Moreover the activity diagrams (Fig 6) indicatethat the data from the bicarbonate waters from the TarondashCeno Valleysfall between quartz whose log [SiO2] =minus433 is not shown because of
the slow precipitation kinetics under the PndashT conditions considered
here (Williams et al 1985) and amorphous silica boundary 1047297elds(Fig 6A) at the same manner bicarbonate waters with higher SiO2-activity from the Voltri Group fall on amorphous silica boundary(Fig 6B) In summary then the bicarbonate waters are supersaturatedin kaolinite ferrihydrite vermiculite and saponite Alkaline watersare also supersaturated with saponite (Fig 6AndashC) and they are alsoequilibrated with primary magnetite and under extreme conditionsprobably pyrite as well
55 Limitations of the model related to kinetics
Model calculations disregard the rates of mineral dissolution and
precipitation which becomes increasingly important as temperature
Table 3
Starting mineralogy (left side) and 1047298uid composition (right side) used in the modeling
Rock composition Fluid composition
Modal of the mineralsin the interacting rock
mol of the minerals usedin the modeling
meteoric water composition Initial 1047298uid composition used in the modeling
Spinel 003 Na 3523Eminus05 3523Eminus055 Spinel Chromite 001 S 3519Eminus05 3519Eminus05
Magnetite 0 Si ndash 1176Eminus06Kaolinite ndash 5258Eminus08
Ferrhydrite-6line ndash 5882Eminus07
The modal percentage of the minerals in a serpentinite from TarondashCeno Valleys (Ernst and Piccardo 1979) was re-calculated assuming an ideal solid solution between pure extremephasesand takingthe modal percentage tobe themass percentage Thelatter was used asthe solid reactant composition in thewaterndashrockinteraction modelPure solid phases werechosenfrom the llnldat thermodynamic database of the PHREEQCI softwareThe weighted average chemical compositionof the meteoric waterfrom Bologna (Panettiere et al 2000)was chosen as the initial 1047298uid composition in the modeling after its equilibration with kaolinite and ferrihydrite (see text for details)
Table 4
Summary of thermodynamic data for the sepiolite (Spn) and serpentine (Sr n) mineralsphases in italics do not refer to the mentioned hydrolysis reactions
ΔGf 0 log K Reference
(kJ molminus1) (at 25 degC)
Sepiolite (Spn) mdash Mg4Si6O15(OH)2middot6H2O+8H+=4Mg2++6SiO2+11H2Ominus925400 3004 Gunnarsson et al (2005)minus925133 3050andash3172 Christ et al (1973)
(Sp2) minus921986 3602andash3756 Wollast et al (1968)
Serpentine (Srn) mdash Mg3Si2O5(OH)4+ 6H
+
=3Mg
2+
+2SiO2+ 5H2ONi-lizardite minus 289324 1680 Mondesir and Decarreau (1987)
Al-lizardite2 minus411003 2544a Caruso and Chernosky (1979) Al-lizardite minus 339854 2786 this work (see text for details)
Mg-lizardite (Sr1) minus404025 3056 Wilson et al (2006)Mg-crysotile (Sr2) minus403702 3112 bthermocomV8R6Mg-lizardite minus402824 3340 Mondesir and Decarreau (1987)p-antigorite (Sr3) minus564239a 3479 Gunnarsson et al (2005)Mg-chrysotile (Sr4) minus406809 3669 Bruni et al (2002)Mg-chrysotileasbestos minus562505a 3888 Pfeifer (1977)Antigorite minus6612864 47719 bthermocomV8R6
The 1047297eld boundaries of the underlined phases shown in Fig 6 correspond to atemperature of 11 degC
a In the ΔG f deg column value calculated using the ΔG f
deg data of the elements from llnldat
and logK value in the reference in the LogK column value calculated using ΔG f deg
(mineral) in the reference and ΔG f deg data of the elements from llnldat
b Thermodynamic database of Lawrence Livermore National Laboratory included in
PHREEQCI and Geochemists Workbench softwares
85T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
decreases The waterndashrock interaction in the TarondashCeno basin iscomplicated by the presence of secondary mixed-layer clay mineralsthe origin of which is quite dif 1047297cult to model because the solidndashstatetransformations and weathering reactions are strictly related tohydrological and climatic conditions as described above An equili-
brium-based reaction path model is a valuable tool for assessing the
most probable reactions and for de1047297ning the product phases that canbe used in a kinetics-based model Since the kinetics of precipitationof Mg-clays at room temperature and pressure are poorly knownrelative reaction rate effects in the reaction path model can be mimedby omitting the stabler phase when a metastable mineral is expected
to precipitate Therefore the precipitation of MgndashFe3+
Abbreviations are as follows Fer-6line 6-line ferrihydrite kaol kaolinite verm vermiculite magn magnetite sap saponite serp serpentine clin clinochlore pyr pyrite calccalcite hydrom hydromagnesite bruc brucite enst enstatite The chromium decrease during the change in water facies from bicarbonate to alkaline was modeled withsupersaturation (precipitation) of eskolaite and thereafter with equilibrium with chromite (Boschetti 2003)- not bound dagger chlorite mineral from solution Dagger chlorite mineral added to the paragenesis
86 T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
montmorillonites vermiculite and saponites are predicted to occurrather than that of Mg-clays Moreover in order to simulate the slowprecipitation of calcite in depth precipitation was not activated atslight supersaturation (SIcalcite=001) during the initial stages of thewaterndashrock interaction in this manner saturation index of calcite wasfree to increase in the following steps and in alkaline waters
6 Discussion of the model results
61 Explanation of pre-modeling choices and post-modeling results
The results of the reaction path modeling are reported in Table 5and shown in Figs 6 and 9 The reaction paths model the dissolutionof a typical local serpentinite (mineralogically constrained as inTable 3) and the sequential precipitation of gibbsite kaoliniteferrihydrite vermiculite Fe2+ndashMg2+-saponite and a poorly-crystal-line serpentine( p-antigorite of Gunnarsson et al2005) The1047297rst stepinvolves the reaction of the ldquoinitial waterrdquo with serpentinite atconstant temperature (11 degC) PO2 (10minus26 bar) and PCO2 (10minus30 bar)this early stage corresponds to bicarbonate groundwater feeding theinvestigated spring system At the stage of primary magnetiteequilibrium and Fe2+ndashMg2+-saponite supersaturation the modelassumes the system to be closed to CO2 and to have a 1047297xed partialpressure of oxygen (PO2= 10minus40 bar) these conditions represent
when water switch from a bicarbonate to an alkaline compositionThe calculated amount of interacting serpentinite needed to obtainthe alkaline springs of the TarondashCeno Valleys (mean PCO2= 10minus7 bar)is approximately 00203 moles (see Table 5)
Bruni et al (2002) explained the origin of the alkaline waters of the Voltri Group as the result of the evolution of meteoricbicarbonate water following the precipitation of sepiolite withabundance similar to that of goethite This might be supported bythe observation that sepiolite or smectite precipitate at lowtemperature from solutions with low Al content when the pH is inthe range 75ndash95 or N95 respectively (Siffert and Wey 1962 Jonesand Galan 1988) but sepiolites form from the weathering of serpentines in arid climate regions whereas sepiolites are rare inhumid climate regions where serpentinite changes to Fe-smectite
(Singer 1989 Wrucke 1996) In the ophiolite rocks of the Northern
Apennine sepiolite is found only sporadically in veins (Giuseppetti1953 Veniale 1966) and is unknown in the ophiolites of the Tarondash
Ceno Valleys whereas it is thought that sepiolite of the Voltri Groupprecipitated from metasomatic 1047298uids during development of shearzones (Crispini and Capponi 2002)
As described above our model explains the transition frombicarbonate water to alkaline water (UM15 and PR10 samples) asthe result of low-temperature precipitation of a poor crystalline
serpentine with the solubility product of p-antigorite (see Sr3 inTable 4 and Fig 6) The model indicates that the bicarbonate waters of the Voltri Group with the highest [Mg2+][H+]2 activity ratios similarlyevolve towards alkaline waters (Fig 6B) The logK value of thesynthetic low temperature serpentine adopted in this model and its
Fig 7 Saturation indices of chrysotile and p-antigorite (Sr2 and Sr3 thermodynamic
data respectively in Table 4) for bicarbonate and alkaline waters coming fromultrama1047297c rocks of the TarondashCeno Valleys as a function of SiMg molar ratioThermodynamic data on chrysotile and p-antigorite are from the thermocomV8R6
database and Gunnarsson et al (2005) respectively
Fig 8 Activityplot forthe MgOndashSiO2ndashAl2O3ndashFe2O3(FeO)ndashH2O system analyzed forSiO2-phases and aqueous species at different log PO2 and log[aAl3+(aH+)3] values Springwaters coming from the TarondashCeno Valleys (symbols as in Fig 2) and Voltri Groupultrama1047297tes (symbols as in Fig 5) are shown The greyarea refers to bicarbonate watersfrom theVoltri Group (Mariniand Ottonello2002Bruniet al 2002) Thesepiolite 1047297eldis very narrow or hidden by saponite-Mg (1047297eld limits shown as continuous lines)saponite-Fe3+ndashMg (1047297eldlimits shown as dottedlines) or vermiculite(1047297eldlimits shownas dottedndashdashed lines) The talc stability 1047297eld and boundaries with saponite-Mg
montmorillonite-Mg and cronstedite-Mg are shown for comparison (dashed lines)
87T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
relative uncertainty of plusmn2 (see Fig 8 in Gunnarsson et al 2005) arecomparable to the logK variation reported for other low-temperatureMg-serpentines (Table 4) that is within the lines Sr1ndash4 for serpentineminerals where samples varying from bicarbonate composition toalkaline composition in the ultrama1047297c aquifer are grouped (Fig 6)Fig 7 shows a plot of calculated serpentine saturation index as a
function of the SiMg molar ratio for the water samples from theultrama1047297tes of the TarondashCeno ValleysThe SI andSiMgvalues of pureMg-serpentine (SI= 0 SiMg=067) provide a useful index forseparating bicarbonate waters (SIb0 SiMgb067) from alkalinewaters (SIN0 SiMgN067) and the distribution of data points isconsistent with the idea that bicarbonate waters approach saturationwith serpentine until they become supersaturated as alkaline waters
Based on the Mgvs SiO2(aq) activitydiagram and speci1047297cratiosoflog[aAl3+(aH+)3] Birsoy (2002) analyzed the behavior of sepiolite duringultrama1047297te weathering The diagram produces a very narrow stability1047297eldfor sepiolite(Fig 8) when it is programmedwith suitable conditionsfor the bicarbonate waters (log PO2=minus25 bar log [aAl3+(aH+)3]= 65) andthe alkaline waters (log PO2=minus50 bar log [aAl3+(aH+)3]= 45) sampled inthe present study This stability 1047297eld grows larger when the absence of
Mg-smectites and talc is assumed but this is an unrealistic conditionbecause (i) all these minerals occur in the rocks within the areainvestigated and (ii) at high [Mg2+][H+]2 activity ratios the 1047297eld of sepiolite is smaller than that of subordinate compared to those poorly-crystalline serpentine (p-antigorite) and clinochlore
Among the alkaline springs of the Voltri Group the waters withlow log [aCa2+(aH+)2] values (Fig 9) show patterns compatible withchlorite interaction (step 5c in Fig6D)The alkaline waters whose datafall near the clinochlore-14Aring 1047297eld (Fig 6D) cannot be explained bysepiolite precipitation because the sepiolite line does not cross thechlorite 1047297elds Thus the peculiar composition of some alkaline watersof the Voltri Group is probably due to interactions between water andprimary chlorites of the ultrama1047297c rocks (optional step 5c in Table 5cb) This is supported by the occurrence sometimes in large amounts
of metasomatic Mgndash(Fe)-chlorite (Scambelluri and Rampone 1999
Vissers and Strating 2001) in the Ligurian area where the relativeabundance of chlorite can reach 50 in veins in high strainserpentinite mylonites of the ErrondashTobbio ultrama1047297tes (Scambelluriet al 2001b) Otherwise in a chlorite-free primary paragenesis thesolution evolving on the Sr3 serpentine line reaches equilibrium withclinochlore but probably the latter does not precipitate due to theslow kinetics of precipitation under these conditions Thus thedissolution of additional reacting rock in the iterative weathering
process supplies Mg
2+
to the solution thereby favoring the Sr3 to Sr4poorly crystalline serpentine transformation In fact all the PR10samples and some of the alkaline springs from the Voltri Group showincreasing [Mg2+][H+]2 activity ratio and low variation in silicaactivity (Fig 6D) starting from the invariant triple point 4a (Sr3-serpentine Fe2+ndashMg-saponite chlinoclore-14 Aring saturation) andextending towards the brucite line This path can be explained asthe result of the transformation of poorly crystalline serpentine fromSr3 to Sr4 the latter is more stable at higher [Mg2+][H+]2 activityratios Finally at very low PO2 values vermiculite disappears and thestability 1047297elds of chlorite and Fe2+ndashMg-saponite become narrowerwhereas those of Fe2+-rich phases (eg Fe2+-saponite) become widerThe data for the alkaline waters with the highest log [aMg2+(aH+)2]from the Voltri Group fall near the stability 1047297eld of peridotite primaryminerals like forsterite and enstatite (Fig 6D) therefore the last stepsof theevolution involve theequilibration of thesolutionwithenstatite(steps 8a and 8c in Table 5ab)
62 Ca-rich alkaline waters
In preliminary modeling of the evolution of the TarondashCeno springwaters Boschetti (2003) concluded that meteoric water initiallysaturated in Fe3+ndashMg2+-chlorite Fe3+ndashMg2+-saponite and pyrite (andor goethite) and then interacted with the ultrama1047297te mineralassemblage eventually evolved into the alkaline waters In themodel of Boschetti (2003) chlorite saponite and calcite were added
Fig 9 Activity plot for the MgOndashCaOndash(CO2)ndashH2O system with phase limits calculated ata temperature of 11 degCSpring waterscoming from theTarondashCeno Valleys (symbols as inFig2) andVoltri Groupultrama1047297tes(symbolsas in Fig5) areshown Calculated path forwaterndashultrama1047297c rock interaction is shown (dotted line) with the number ndashlettercombinations referring to the appearance or disappearance of minerals (see Table 5)The grey area refers to bicarbonate waters from the Voltri Group (Marini and Ottonello2002)
Fig 10 Plots of (A) calcite saturation indexes and (B) PCO2 as a function of Clminus ascalculated by PHREEQCI software for waters from the TarondashCeno Valleys Alkalinesprings for the Voltri Group are also shown (small black diamonds Marini andOttonello 2002) The grey shading is deeper(A) in the saturation1047297eld(SI= 00plusmn05) and
(B) where the lowest log PCO2 value is approached
88 T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
to the solution in order to achieve a high [Ca2+][H+]2 activity ratio andsupersaturation in these phases However in light of the work of Clarket al (1992) and Bruni et al (2002) we believe that the calcitesupersaturation in alkaline waters (Fig 9 and 10) which in our modelis achieved by preventing the precipitation of calcite in the 1047297rst stepsof the reaction path (Table 5a and b) depends on the differencebetween CO2 partial pressureat thegivendepth(PCO2le10minus10bar) andat the surface (PCO2cong10minus35) Waters rising to the surface are far from
equilibrium with the atmosphere where the PCO2 is 4times10
6
timesgreater than the value at depth thus the water rapidly take up CO2which is promptly converted to CO3
2minus because of the high pH Thisleads to supersaturation with calcite through the following hydro-xylation reaction
As a result large amounts of travertine are generated in the VoltriGroup (Marini and Ottonello 2002) and carbonate concretionsknown as devilrsquo s coins are formed at Mt Prinzera in Taro valleyBelow the surface the SIcalcite of the alkaline waters is expected to belower than what the model estimates Under conditions of lowcarbonate alkalinity low CO2 and O2 partial pressure and high pHcalcite nucleation may be inhibited and the kinetics of precipitationreduced by the high concentrations of dissolved divalent cations suchas Mg2+ Fe2+ and Mn2+ (Sumner and Grotzinger 1996 Davis et al2000 Zhang and Dawe 2000 de Leeuw 2002) The fact that severalalkaline water samples fall within or near the calcitendashdolomiteequilibrium (Fig 9) may indicate magnesium calcite or protodolomitenucleation in those extreme subsurface conditions However possibleinteraction between some calcium-rich waters of the Voltri Group andCa-rich rocks such as ophicarbonates or rodingites cannot beexcluded these rocks are widespread in the Ligurian Apennineserpentinites (Treves and Harper 1994)
63 Sepiolite vs serpentine nucleation under weathering conditions
constraints on the kinetics and equilibrium-based approaches
The kinetics of precipitation of poorly-crystalline serpentine and
sepiolite cannot be compared quantitatively because rate constantsare available only for well-ordered crystalline phases (eg see Palandriand Kharaka 2004 for crystalline serpentine) Sepiolite precipitationis often invoked to explain magnesium scavenging from solutions butthe lack of kinetic data for this mineral leads to overly simplisticreaction transport models These models are forced to use kinetic datafor phyllosilicates such as kaolinite or muscovite (eg Xu et al 2001Bryan 2005) Moreover the low-temperature thermodynamic dataeven if they are fairly uncertain show that sepiolite solubilityincreases with decreasing crystallinity (Table 4 Stoessell 1988) Infact crystalline sepiolite requires a long time to reach equilibriumwith solution which imposes a slow control on thewater composition(eg Stoessell and Hay 1978) Consequently in modeling based onreaction transport that simulates freshly precipitated or poorly
crystalline sepiolite formation the logK value of Wollast et al (1968)is preferable to the lower values (Suarez and Šimůnek 1996) Thispreference agrees with the approach of Mariner (2003) and Jove Colon(2006) who adopt a logK (25 degC) value for sepiolite of 3044 It is alsosimilar to that reported by Wollast et al (1968) and it is six logK unitshigherthan the valueused in thermocomV8R6 dataset which is oftenused (perhaps improperly) to model the supergenic nucleation of sepiolite In the stability diagram in Fig 6 where the Sp1 and Sp2 linesrepresent the stability boundaries of crystalline and poorly-crystallinesepiolite respectively the evolution of TarondashCeno waters frombicarbonate to alkaline (Figs 6A and C) follows the stability 1047297eldboundary of the poorly-crystalline serpentine (between Sr2ndashSr3 lines)rather than that of the poorly-crystalline sepiolite (Sp2) as postulatedin themodel The lines for Sp2 Sr3 and amorphoussilica intersect at a
common point (Fig 6D) which is consistent with the nucleation of
sepiolite that typically occurs in solutions that are saturated orsupersaturated with amorphous silica and (Mg)-carbonate (egStoessell and Hay 1978 Birsoy 2002) Some bicarbonate watersfrom the Voltri Group have a silica activity near the amorphous silicaboundary (Fig 6B) but only a few alkaline springs of this group fall onthe poorly-crystalline sepiolite boundary line (Sp2 Fig 6D) Both of these features are inconsistent with the possibility that sepioliteprecipitates from supergenic fresh water in the Northern Apennine
Moreover these two results suggest that values for logK(sepiolite)b
3044 (see Table 4) should be used to model the dissolutionprecipitation of the well-crystallized phase that is the most commonform for the metasomatic sepiolite found in the investigated area
7 Conclusions
The waters issuing from ultrama1047297tes in the TarondashCeno Valleysoriginate from low-temperature interactions between meteoric waterand serpentinite rocks In this report the waterndashrock interaction hasbeen modeled using reaction paths in progress mode During their earlyinteraction with soil and rock minerals bicarbonate waters super-saturated in kaolinite ferrihydrite vermiculite and Fe2+ndashMg2+-saponitewhere theorigin of the secondary iron saponite as Fe2+-bearing smectiteagrees with the hypothesis of Guumlven (1988) Then while remainingsupersaturated with saponite the bicarbonate waters evolve into analkaline facies when poorly-crystalline serpentine precipitates Duringrock weathering the Al-bearing crystalline serpentine of oceanic originprobably promotes nucleation of the low-temperature serpentineamorphous phase and supplies Al3+ to the solution for saponiteprecipitation This interpretation is consistent with the existence of Al-poor supergenic serpentine (Neal and Stanger 1985) and the partiallymeteoric OH-isotopic imprint observed in other ultrama1047297c systems
Dif 1047297culties and complexities in the modeling arise fromuncertainties in the primary mineralogy of some local ophiolitesand from the presence of different lithotypes (rodingites ophi-carbonates gabbros) which can signi1047297cantly affect the composi-tion of water samples This can be seen in the example of the Candash
OH springs issuing from the Voltri Group (PO2 and PCO2 of 10minus70
and 10minus10 bar respectively Bruni et al 2002 Marini and
Ottonello 2002) These springs appear to be more evolved thanthe TarondashCeno alkaline waters as a result of their particularly longinteractions with serpentinites until equilibrium with enstatite isreached or with chlorite-bearing serpentinite The use of anupdated thermodynamic database in our model avoided problemsdue to mineralogical heterogeneity in the mixed-layer claysassociated with ophiolitic rocks this database accurately re1047298ectedthe chemical composition of the single phases Using this modelthe vermiculitendashchlorite and saponitendashchlorite associations seen inthe activity diagrams simulate respectively the presence of highand low charge corrensite observed in the 1047297eld The accuracy of ourmodel can be tested by comparing estimates of chlorine composi-tion resulting from our model and from experimental methods The
mean Cl concentration of the bicarbonate water is 41plusmn16 mgLtaking into account a Cl concentration of sim2 mgL in rainwater inthe area (cf Panettiere et al 2000 Venturelli 2003) 2 mgL of Clin spring water are estimated to come from serpentinite This valueagrees with the model which predicts that the switch frombicarbonate water to alkaline water requires the congruentdissolution of 00203 mol of serpentinite containing 100 ppm of soluble Cl thereby yielding water containing 2 ppm of Cl
In addition we think that weathering of the accessory albiteplagioclase is responsible for Na enrichment of the waters whereas Clenrichment in the alkaline waters is predicted to result from theleaching of NaCl 1047298uid inclusions Moreover the TarondashCeno alkalinesprings show the highest BCl ratios comparable with those of thehost metamorphic minerals (Scambelluri et al 2004) In contrast the
lower boron content and BCl ratio of the alkaline springs of the Voltri
89T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
Group are unexpected features and probably re1047298ect waterndashrockinteractions over a longer time period than in the TarondashCeno ValleysThe boron content in the Northern Apennine alkaline springs may becontrolled by several factors such as (i) the different histories of VoltriGroup and EL ultrama1047297tes (ii) the probable contribution of boronfrom NaCl 1047298uid inclusions (iii) the sinking of boron by travertineprecipitation (Hobbs and Reardon1999) brucite (Rhoades et al 1970)and smectite and (iv) decreases in boron release from minerals due to
increases in pH (Su and Suarez 2004) Acknowledgements
This research was carried out with the 1047297nancial support from theldquoYoung Researchers Projectrdquo (2001ndash2002) of the Italian Ministry of Universities Scienti1047297c Research and Technology (MURST) We thankGianniCortecci (Institute of Geosciences and EarthResources IGG-CNRPisa Italy) and Giampiero Venturelli (Earth Sciences Department Univer-sity of Parma Italy) for useful suggestions and critical review of themanuscript The authors are grateful to Enrico Dinelli (InterdepartmentResearch Centre for Environmental Sciences Alma Mater StudiorumUniversity of Bologna Ravenna Italy) for granting permission to use theGeochemistrsquos WorkbenchsoftwareTheconstructive reviews of theEditor
J Fein and of two anonymous Referees have signi1047297cantly improved the
manuscript
Appendix A Supplementary data
Supplementary data associated with this article can be found inthe online version at doi101016jchemgeo200808017
References
Abbate E Bortolotti V Passerini P 1980 Olistostromes and olistoliths In Sestini G(Ed) Development of the Northern Apennines geosyncline Sedimentary Geology4 pp 521ndash558
Alietti A 1956 Il minerale a strati misti saponite-talco di monte Chiaro (val di TaroAppen Emil) Rendiconti dellAccademia Nazionale dei Lincei 21 201 ndash207
Alietti A De Angelis G Montanari G Sgarlata F Tirelli G 1976 Il ferro nelle argilleappenniniche Periodico di Mineralogia 45 51ndash64
Anselmi B Mellini M Viti C 2000 Chlorine in the Elba Monti Livornesi and Murloserpentines Evidence for seandashwater interaction European Journal of Mineralogy12137ndash146
Apha-Awwa-Wef 1995 Standard Methods for the Examination of Water and Waste-water 19th Edition American Public Health Association American Water WorksAssociation Water Environment Federation
Barnes I LaMarche Jr VC Himmelberg GR 1967 Geochemical evidence of present-day serpentinization Science 156 830ndash832
Barnes I ONeil JR 1969 The relationship between 1047298uids in some fresh alpine-typeultrama1047297cs and possible modern serpentinization Western United StatesGeological Society of America Bulletin 80 1947ndash1960
Barnes I ONeil JRTrescases JJ1978 Present dayserpentinizationin New CaledoniaOman and Yugoslavia Geochimica et Cosmochimica Acta 42 144ndash145
Barnes I Rapp JB ONeil JR Sheppard RA Gude III AJ 1972 Metamorphicassemblages and the direction of 1047298ow of metamorphic 1047298uids in four instances of serpentinization Contributions to Mineralogy and Petrology 35 263ndash276
Barnes JDSelverstone J Sharp ZD 2006Chlorine isotope chemistry of serpentinitesfrom Elba Italy as an indicator of 1047298uid source and subsequent tectonic historyGeochemistry Geophysics Geosystems 7 doi1010292006GC001296
Beccaluva L Emiliani F Venturelli G Zerbi M 1973 Ca Fe Mg Mn Cr Ni Codistribution in some ultrama1047297c rocks outcropping in the Northern Apennines withsome geological remarks Ateneo Parmense Acta Naturalia 9 69ndash98
Beccaluva L Macciotta G Piccardo GB Zeda O 1984 Petrology of lherzolitic rocksfrom the Northern Apennine ophiolites Lithos 17 299 ndash316
BeccaluvaL Venturelli G ZanzucchiG 1975 Dati geologici e geochimici sui basalti diassociazione o1047297olitica dellAppennino ligure-emiliano Ateneo Parmense ActaNaturalia 11 789ndash802
BethkeCM 2002 The Geochemists Workbench A Userrsquos Guide to RxnAct2Tact ReactandGtplot Release40 HydrogeologyProgram Universityof IllinoisUrbanaIL 236pp
Birsoy R 2002Formation of sepiolitendashpalygorskite and related minerals from solutionClays and Clay Minerals 50 736ndash745
Bonatti E Lawrence JR Morandi N 1984 Serpentinization of oceanic peridotitestemperature dependence of mineralogy and boron content Earth and PlanetaryScience Letters 70 88ndash94
Bonifacio E Zanini E Boero V Franchini-Angela M 1997 Pedogenesis in a soilcatena on serpentinite in north-western Italy Geoderma 75 33ndash51
Boschetti T 2003 Studio geochimico e geochimico-isotopico di acque a composizione
estrema e termali dellrsquoAppennino Settentrionale Ph D Thesis Parma Univ
Brigatti MF Franchini G Lugli C Medici L Poppi L Turci E 2000 Interactionsbetween aqueous chromium solutions and layer silicatesApplied Geochemistry 151307ndash1316
Brigatti MF Lugli C Poppi L Venturelli G 1999 Iron-rich saponite dissolutionreactions and Cr uptake Clay Minerals 34 637ndash645
Brigatti MF Valdregrave G 1996 Mixed-layer chloritesmectite (corrensite) in ophioliticrock alteration products Neues Jahrbuch fuer Mineralogie Monatshefte 1 37 ndash47
Bruni J Canepa M Chiodini G Cioni R Cipolli F Longinelli A Marini L OttonelloG Vetuschi Zuccolini M 2002 Irreversible waterndashrock mass transfer accompany-ing the generation of the neutral MgndashHCO3 and high-pH CandashOH spring waters of the Genova province Italy Applied Geochemistry 17 455ndash474
Bryan CR 2005 Drift-Scale THC Seepage Model MDL-NBS-HS-000001 REV 04 YuccaMountain Project Las Vegas Nevada (US)Caruso LJ Chernosky JV 1979 The stability of lizardite Canadian Mineralogist 17
757ndash769Clark ID Fontes JC Fritz P 1992 Stable isotope disequilibria in travertine from high
pH waters Laboratory investigations and 1047297eld observations from Oman Geochi-mica et Cosmochimica Acta 56 2041ndash2050
Cortesogno L 1980 Il metamor1047297smo giurassico nelle o1047297oliti dellAppenninoSettentrionale due distinti cicli metamor1047297ci in ambiente oceanico O1047297oliti 5 5ndash18
Christ CLHostetler PB Siebert RM1973 Studies in the system MgO-SiO2-CO2-H2OIII American Journal of Science 273 65ndash83
Cortesogno L Lucchetti G 1982 Il metamor1047297smo oceanico nei gabbri o1047297oliticidellAppennino Ligure Aspetti mineralogici e petrogenetici Rendiconti dellaSocietagrave Mineralogica Italiana 38 561ndash579
Crispini L Capponi G 2002 Control of metasomatic alteration on the development of shear zone in ultrama1047297c rocks a case study from the Voltri Massif (Ligurian Alps)Architettura interna e cinematica delle zone di taglio Gruppo Italiano di GeologiaStrutturale mdash Riunione Annuale 2002 Pisa Italy 11ndash12 giugno 2002 XIV ndashXV
CrissRE 1999 Principlesof stableisotope distributionOxford University PressNew York
Davis KJ Dove PM De Yoreo JJ 2000 The role of Mg 2+ as an impurity in calcitegrowth Science 290 1134ndash1137
de Leeuw NH 2002 Moleculardynamics simulations of the growth inhibiting effect of Fe2+ Mg2+ Cd2+ and Sr2+ on calcite crystal growth Journal of Physical Chemistry B106 5241ndash5249
De Nardo MT Segadelli S Vescovi P 2007 Studio pilota per la caratterizzazionegeologicadellesorgentinellazona delM Nero (altaVal Cenoe alta ValNuremdash provincedi Parma e Piacenza) Il Geologo dellEmilia-Romagna 25 Nuova Serie pp 5 ndash21
Di Giulio A Geddo G 1990 Studio petrogra1047297co delle Arenarie di Casanova (alta valTrebbia Appennino settentrionale) Atti Ticinesi di Scienze della Terra 33 243ndash254
Dinelli E Lombini A Simoni A Ferrari C 1997 Heavy metals in serpentinite soils of selected outcrops of Piacenza and Parma provinces (Northern Apennines Italy)Mineralogica et Petrographica Acta 40 241ndash255
Epstein S Mayeda T 1953 Variation of 18O content of waters from natural sourcesGeochimica et Cosmochimica Acta 4 213ndash224
Ernst WG Piccardo GB 1979 Petrogenesis of some Ligurian peridotites I Mineraland bulk-rock chemistry Geochimica et Cosmochimica Acta 43 219ndash237
Feder F Trolard F Klingelhofer G Bourrieacute G 2005 In situ Moumlssbauer spectroscopy
evidence for green rust (fougerite) in a gleysol and its mineralogical transforma-tions with time and depth Geochimica et Cosmochimica Acta 69 4463ndash4483Fritz PClark IDFontes JCWhiticar MJFaberE 1992 Deuterium and
13C evidence
for low temperature production of hydrogen and methane in a highly alkalinegroundwater environment in Oman In Kharaka TK Maest AS (Eds) Proceed-ings of the 7th International Symposium on WaterndashRock Interaction BalkemaRotterdam pp 793ndash796
Giammetti F 1968 Le o1047297oliti di Roccaprebalza Costa della Guardia e Groppo di Gorro(Appennino Parmense) Ateneo Parmense Acta Naturalia 4 107ndash142
Giuseppetti G1953 Roccee minerali dellaformazioneo1047297olitica di Volpedo Rendicontidella Societagrave Mineralogica Italiana 9 85ndash134
Gran G 1952 Determination of equivalence point in the potentiometric titrationAnalyst 77 661ndash671
Gunnarsson I Arnorsson S Jakobsson S 2005 Precipitation of poorly crystalline antigoriteunder hydrothermal conditions Geochimica et Cosmochimica Acta 69 2813ndash2828
Guumlven N 1988 Smetites In Bailey SW (Ed) Reviews in Mineralogy 19 Hydrousphyllosilicates (exclusive of micas) Mineralogical Society of America MichiganBook Crafters Inc 497ndash559
Hansel CM Benner SG Fendorf S 2005 Competing Fe(II)-induced mineralization
pathways of ferrihydrite Environmental Science and Technology 39 7147ndash7153HobbsMYReardonEJ 1999 Effect of pH on boronco-precipitation by calcite further
evidence for non-equilibrium partitioning of trace elements Geochimica etCosmochimica Acta 63 1013ndash1021
Jones BF Galan E 1988 Sepiolite and palygorskite In Bailey SW (Ed) HydrousPhyllosilicates (exclusive of Micas) Reviews in Mineralogy vol 19 MineralogicalSociety of America Michigan Book Crafters Inc 631ndash674
Jove Colon CF 2006 Technical work plan for Thermodynamic database for chemicalmodelling TWP-MGR-PA-000039 REV 01 Yucca Mountain Project Las VegasNevada (US)
Kohls DW Rodda Jr JL 1967 Iowaite a new hydrous magnesium hydroxidendashferricoxychloride from the Precambrian of Iowa American Mineralogist 52 1261 ndash1271
Kyser TK OHanley DS Wicks FJ 1999 The origin of 1047298uids associated withserpentinization processes evidence from stable-isotope compositions CanadianMineralogist 37 223ndash237
Lapham DM 1961 New data on deweylite American Mineralogist 46 168ndash188Lee BD SearsSKGraham RC Amrhein CValiH 2003 Secondarymineral genesis
from chlorite and serpentine in an ultrama1047297c soil toposequence Soil Science
Society of America Journal 671309ndash1317
90 T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
Longinelli A Selmo E 2003 Isotopic composition of precipitation in Italy a 1047297rstoverall map Journal of Hydrology 270 75ndash88
Majzlan J Navrotsky A Schwertmann U 2004 Thermodynamics of iron oxides PartIIIEnthalpiesof formation and stability of ferrihydrite (simFe(OH)3) schwertmannite(simFeO(OH)34(SO4)18) and ɛ-Fe2O3 Geochimica et Cosmochimica Acta 681049ndash1059
Mariner P 2003 In-Drift PrecipitatesSalts Model ANL-EBS-MD-000045 REV 01 YuccaMountain Project Las Vegas Nevada (US)
Marini L Ottonello G 2002 Atlante degli acquiferi della Liguria Volume III Le acquedeicomplessio1047297olitici (bacini ArrestraBranega Cassinelle Cerusa ErroGorzenteLeira Lemme Lerone Orba Piota Polcevera Rumaro Sansobbia Stura Teiro
Varenna Visone) Pacini PisaMellini L Zanazzi PF 1987 Crystal structures of lizardite-l T and lizardite-2H1 fromColi Italy American Mineralogist 72 943ndash948
Mezzadri G 1964 Petrogra1047297a delle arenarie di Ostia Rendiconti Societagrave MineralogicaItaliana 20 192ndash228
Mondesir H Decarreau A 1987 Synthesegrave entre 25 et 200 degC de lizardites Ni ndashMgMesure des coef 1047297cients de partage solide-solution aqueuse pour le couple Ni ndashMgdans des lizardites Bulletin de Mineralogie 110 409ndash426
Neal C Shand P 2002 Spring and surface water quality of the Cyprus ophiolitesHydrology and Earth System Sciences 6 797ndash817
Neal C Stanger G 1983 Hydrogen generation from mantle source rocks in OmanEarth and Planetary Science Letters 66 315ndash320
Neal C Stanger G 1984 Calcium and magnesium hydroxide precipitation fromalkaline groundwaters in Oman and their signi1047297cance to the process of serpentinization Mineralogical Magazine 48 237ndash241
Neal C Stanger G 1985 Past and present serpentinisation of ultrama1047297c rocks anexample from Semail ophiolite nappe of Northern Oman In Drever JI (Ed) TheChemistry of Weathering D Reidel Publishing Company pp 249ndash275
Nesbitt HW Bricker OP 1978 Low temperature alteration processes affecting
et Cosmochimica Acta 41 1835ndash1841OrsquoHanley DS 1996 Serpentinites records of tectonic and petrological history Oxford
University PressOrberger B Friedrich G Woermann E 1990 The distribution of halogens and carbon
in Pge-bearing ultrama1047297cs of the Acoje Ophiolite Block Zambales Philippines Journal of Geochemical Exploration 37 147ndash169
Ottonello G Joron JL Piccardo GB 1984 Rare Earth and 3d transition elementsgeochemistry of peridotitic rocks II Ligurian peridotites and associated basalts Journal of Petrology 25 373ndash393
Palandri JL Kharaka YK 2004 A compilation of rate parameters of water-mineralinteraction kinetics for application to geochemical modeling Open File Report2004-1068 US Geological Survey
Panettiere P Cortecci G Dinelli E Bencini A Guidi M 2000 Chemistry and sulfurisotopic composition of precipitation at Bologna Italy Applied Geochemistry 151455ndash1467
Parkhurst DL Appelo CAJ 1999 Userrsquos guide to PHREEQC (version 2) mdash a computerprogram for speciation batch-reaction one-dimensional transport and inversegeochemical calculation US Department of the Interior US Geological SurveyWashington DC
Pfeifer HR 1977 A model for 1047298uids in metamorphosed ultrama1047297c rocks Observationsat surface and subsurface conditions (high pH spring waters) SchweizerischeMineralogische und Petrographische Mitteilungen 57 361ndash396
PlummerLN1984Geochemical modeling A comparisonof forwardand inverse methodsIn Hitchon B Wallick EI (Eds) Proceedings First CanadianAmerican Conference onHydrogeology-Practical Applications of Ground Water Geochemistry Banff AlbertaCanada mdash Worthington Ohio National Water Well Association pp 149ndash177
Rampone E Hofmann AW Piccardo GB Vannucci R Bottazzi P Ottolini L 1995Petrology mineral and isotope geochemistry of the External Ligurides peridotites(Northern Apennines Italy) Journal of Petrology 36 81ndash105
Rampone E Hofmann AW Piccardo GB Vannucci R Bottazzi P Ottolini L 1996Trace element and isotope geochemistry of depleted peridotites from an N-MORBtype ophiolite (Internal Liguride N Italy) Contributions to Mineralogy andPetrology 123 61ndash76
Rhoades JD Ingvalson RD HatcherJT1970Adsorption of boronby ferromagnesianminerals and magnesium hydroxide Soil Science Society of America Journal 34
938ndash941Saini-EidukatB KuchaH Keppler H1994Hibbingiteγ-Fe2(OH)3Clanewmineralfromthe Duluth Complex Minnesota with implications for the oxidation of Fe-bearingcompounds and the transport of metals American Mineralogist 79 555ndash561
Sanford RF 1981 Mineralogical and chemical effects of hydration reactions andapplications to serpentinization American Mineralogist 66 290ndash297
Scambelluri M Muumlntener O Ottolini L Pettke TT Vannucci R 2004 The fate of BCl and Li in the subducted oceanic mantle and in the antigorite breakdown 1047298uidsEarth and Planetary Science Letters 222 217ndash234
Scambelluri M Rampone E 1999 Mg-metasomatism of oceanic gabbros and itscontrol on Ti-clinohumite formation during eclogitization Contributions toMineralogy and Petrology 135 1ndash17
Scambelluri M Bottazzi P Trommsdorff V Vannucci R Hermann J Gogravemez-Pugnaire MT Logravepez-Sagravenchez Vizcaino V 2001a Incompatible element-rich1047298uidsreleased by antigorite breakdown in deeply subducted mantle Earth and PlanetaryScience Letters 192 457ndash470
Scambelluri M Rampone E Piccardo G 2001b Fluid and element cycling insubducted serpentinite a trace-element study of the ErrondashTobbio high-pressureultrama1047297tes (Western Alps NW Italy) Journal of Petrology 42 55ndash67
Segadelli S2006 Lageologia nelpaesaggio le rupi o1047297olitiche in Val Taro e Val Ceno IlGeologo dellrsquoEmilia-Romagna 22 15ndash29 Nuova Serie
Settembre Blundo D Martin De Vidales JL Alietti A 1992 Corrensite-type clayminerals from Taro Valley (North Italy) Mineralogica et Petrographica Acta 35215ndash222
Sharp ZD Barnes JD 2004 Water-soluble chlorides in massive sea1047298oor serpenti-nites a source of chloride in subduction zones Earth and Planetary Science Letters226 243ndash254
Siever R Woodford N 1979 Dissolution kinetics and the weathering of ma1047297cminerals Geochimica et Cosmochimica Acta 43 717ndash724Singer A 1989 Palygorskite and sepiolite group minerals In Dixon JB Weed SB
(Eds) Minerals in Soil Environments Soil Science Society of America MadisonWisconsin pp 829ndash872
Siffert B Wey R 1962 Synthese d rsquoune sepiolite a temperature ordinarie Comptesrendus hebdomadaires des seacuteances de lAcadeacutemie des Sciences de Paris 2541460ndash1463
Środoń J 1999 Nature of mixed-layer clays and mechanisms of their formation andalteration Annual Reviews of Earth and Planetary Science 27 19ndash53
Stoessell RK 1988 25 degC and 1 atm dissolution experiments of sepiolite and keroliteGeochimica et Cosmochimica Acta 52 365ndash374
Stoessell RK Hay RL 1978 The geochemical origin of Sepiolite and keorilite atAmboseli Kenya Contributions to Mineralogy and Petrology 65 255ndash267
Suarez DL Šimůnek J 1996 Solute transport modeling under variably saturatedwater 1047298ow conditions In Lichtner PC Steefel CI Oelkers EH (Eds) ReactiveTransport in Porous Media Reviews in Mineralogy vol 34 Mineralogical Society of America Michigan Book Crafters Inc 229ndash268
Sumner DY Grotzinger JP 1996 Were kinetics of Archean calcium carbonate
precipitation related to oxygen concentration Geology 24 119ndash122Su C Suarez DL 2004 Boron release from weathering of illites serpentine shales
and illiticpalygorskitic soils Soil Science Society of America Journal 68 96 ndash105Toscani L Venturelli G Boschetti T 2001 Sul1047297de-free and sul1047297de-bearing waters in
the Northern Apennines Italy general features and waterndashrock interaction AquaticGeochemistry 7 195ndash216
Treves BE Harper GD 1994 Exposure of serpentinites on the ocean 1047298oor sequenceof faulting and hydrofracturing in the Northern Apennine ophicalcites O1047297oliti 19b435ndash466
Varadachari C Kudrat M Gosh K 1994 Evaluation of standard free energies of formation of clay minerals by an improved regression method Clays and ClayMinerals 42 298ndash307
Veniale F 1966 Sepiolite in sedimenti dellrsquoAppennino Vogherese (Pavia) Periodico diMineralogia 35 343ndash386
Venturelli G 2003 Acque minerali ed ambiente mdash Fondamenti di geochimica deiprocessi di bassa temperatura Pitagora Editrice Bologna
Venturelli G Contini S Bonazzi A Mangia A 1997 Weathering of ultrama1047297c rocksand element mobility at Mt Prinzera Northern Apennines Italy MineralogicalMagazine 61 765ndash778
Venturelli G Frey M 1977 Anchizone metamorphism in sedimentary sequences of the Northern Apennines Rendiconti Societagrave Mineralogica Italiana 33 109ndash123
Vieillard P 2002 A new method for the prediction of Gibbs free energies of formationofphyllosilicates (10 Aring and14 Aring)basedon theelectronegativityscale Claysand ClayMinerals 50 352ndash363
Vissers RLM Strating EHH 2001 Structures and microstructures in a thrust-relatedgreenschist facies tectonic melange Voltri Group (NW Italy) O1047297oliti 26 33ndash46
Viti C Mellini M 1997 Contrasting chemical compositions in associated lizardite andchrysotile in veins from Elba Italy European Journal of Mineralogy 9 585 ndash596
Viti C Mellini M 1998 Mesh textures and bastites in the Elba retrogradeserpentinites European Journal of Mineralogy 10 1341ndash1359
Wenner DB Taylor Jr HP 1974 Deuteriumatomic hydrogen and oxygen-18oxygen-16 studies of serpentinization of ultrama1047297c rocks Geochimica et CosmochimicaActa 38 1255ndash1286
Williams LA Parks GA Crerar DA 1985 Silica diagenesis I Solubility controls Journal of S edimentary Petrology 55 301ndash311
Wilson J Savage D Cuadros J Shibata M RagnarsdottirKV 2006 The effect of ironon montmorillonite stability (I) Background and thermodynamic considerations
Geochimica et Cosmochimica Acta 70 306ndash322WollastR MackenzieFTBricker OP1968 Experimental precipitation and genesis of sepiolite at earth-surface conditions American Mineralogist 531 1645ndash1662
Wrucke CT1996 Serpentine and carbonated-hosted asbestos deposits InEdwardAdu Bray (Eds) Preliminary compilation of descriptive geoenvironmental mineraldeposit models 95ndash831 US Geological Survey Open-File Report pp 39 ndash46
Xu T Apps JA Pruess K 2001 Analysis of mineral trapping for CO2 disposal in deepaquifers Lawrence Berkeley National Laboratory University of California
ZhangYDaweRA2000 In1047298uence ofMg2+ on thekineticsof calcite precipitation andcalcite crystal morphology Chemical Geology 163 129ndash138
Zhu C Anderson G 2002 Environmental applications of geochemical modelingCambridge University Press
91T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
levels were found in the alkaline springs of meteoric origin in OmanNew Caledonia and ex-Yugoslavia (Barnes et al 1978)
Among the waters analyzed so far in the TarondashCeno Valleys thebicarbonate waters have BCl ratios in the range 00017ndash022 (median0073) whereas the alkaline waters show the highest ratios everreported (BCl= 034ndash33 median 169) The high boron content of upto 13 mgL of the alkaline springs could be related to metamorphic1047298uids released during serpentinite subduction in fact rocks prefer-
entially lose Cl and the BCl ratio of the released 1047298
uids increases withpressure and temperature (Scambelluri et al 2004) Nonetheless thealkaline water samples in our study differ signi1047297cantly from alkalinemetamorphic1047298uids in that they have low salinity high BCl ratios andmeteoric isotope signatures (see below) all these characteristicssuggest an origin in the interaction of meteoric water with high-pressure metamorphic serpentinite This hypothesis is furthersupported by the agreement in BCl ratio between the alkalinesamples we study here and the high BCl ratio of up to 4 previouslyreported for the residual high-pressure mineralogy of ultrama1047297tes(olivine+orthopyroxene Scambelluri et al 2004) In contrast theVoltri Groups Ca-hydroxide waters have mean values of B = 014 mgL and BCl=003 (Marini and Ottonello 2002) consistent with acontribution of boron and chlorine from 1047298uid inclusions In factassuming fractional contributions of 0001ndash0005 from NaCl 1047298uidinclusions with boron content from 4 to 8 mgL ( Fig 3a Scambelluriet al 2004) results in a B content of 0004ndash01 mgL in the solutionThese results suggest that the difference between springs from theVoltri Group and TarondashCeno Valleys is probably related to (i) differentevolution of the 1047298uids and (ii) the different types of host rocks and Pndash
T paths of the Ligurian Western Alps and EL peridotites (eg Beccaluvaet al 1984 Scambelluri et al 2004)
44 Isotopic composition of water
A diagram of δ2H vs δ18O (Fig 4A) shows that the waters sampledfrom the TarondashCeno Valleys are meteoric in origin since they areconsistent with the meteoric water lines of central and northern Italy(Longinelli and Selmo 2003) The altitude vs δ18O(H2O) (Fig 4B)
summarizes the isotopic effects due to temperature altitude andamount of rainfall (eg Criss1999) all of which help to determine theisotopic composition of the water samples The tendency is similar forbicarbonate waters from sedimentary basaltic and ultrama1047297c forma-
tions which may provide insight into the recharge areas alkalinewaters in contrast probably circulate deeper in the ground
5 Water ndashrock interaction model
51 Fundamentals
The reaction path model between meteoric water and serpentinite
was performed using PHREEQCI 2125 code (Parkhurst and Appelo1999) and the Lawrence Livermore National Laboratory thermody-namic database (llnldat as known as thermocomV8R6 ) implementedwith the equilibrium constants listed in Table 2
Computations wereperformed by reaction progress mode ( forward
modeling Plummer 1984 Zhu and Anderson 2002) ie (i) addingreacting rock at each step in the simulation (ii) activating the phase(s)precipitation only when the thermodynamic stability of the mineralparagenesis is well known and consistent with the mineralogy of thearea and (iii) plotting reaction progress on speci1047297c activity diagramsThe same thermodynamic database was used with The GeochemistsWorkbench 403 software (Bethke 2002) to generate Pourbaix andactivity diagrams (Figs 5 6 8 and 9) Modeling was performed usingan approach based on equilibrium rather than reaction transportbecause the data on reaction kinetics and reactive surfaces in thesesolid phases systems are scarce particularly with respect to theactivation energy of clay and mixed-layer minerals at low temperatureand high pH Moreover the hydrogeology and the discontinuity of theaquifers in the study area which are related to the geologicheterogeneity and structural complexity of the Northern Apenninewould make a model based on reaction transport even less credible atthis stage
The modeling in this study was performed on the basis of thefollowing knowledge (1) primary and secondary mineral assem-blages (2) chemical composition of the minerals involved (3) physicalparameters and chemical composition of the starting 1047298uid (4)thermodynamic data on pure phases
52 The solid reactant primary paragenesis
EL peridotites of the TarondashCeno Valleys (NorthernApennines Italy)are spinel lherzolites containing olivine (forsterite) orthopyroxene(enstatite) and clinopyroxene (diopside) (Giammetti 1968 and
Table 2
Thermodynamic data on hydrolysis reactions added to the llnldat database of the PHREEQCI code (Parkhurst and Appelo 1999) and thermocomV8R6dat database of TheGeochemists Workbench software (Bethke 2002)
Mineral Reaction log K (TdegC) ΔGdeg f References
0 25 60 100 (kJ mol-1)
clinochlore08-daphnite02 Mg4Fe1Al2Si3O10(OH)8 + 16H+=2Al3+ + Fe2++12H2O+4Mg2++3SiO2 7318 6384 5272 4269 minus786379 this workferr ihydrite 2-l ine Fe(OH)3+3H+= Fe3++ 3H2O 484 355 ndash ndash minus70850 Majzlan et al 2004ferr ihydrite 6-l ine Fe(OH)3+3H+= Fe3++ 3H2O 435 312 ndash ndash minus71100 Majzlan et al 2004liz09-berth01 Mg27Fe02Al02Si19O5(OH)4+64H+=02Al3++02Fe2++52H2O+27Mg2++19SiO2 3111 2786 2383 2005 minus399854 this workp-antigorite Mg
Saponite-Fe2+ Fe3175Si365Al035O10(OH)2+74H+=035Al3++3175Fe2++47H2O+365SiO2 1845 1635 1319 1001 minus452459 Wilson et al 2006Saponite-Fe2+ndashMg Fe0165Mg3Si367Al033O10(OH)2+732H+=033Al3++0165Fe2++466H2O+3Mg2++
367SiO2
3037 2729 2330 1921 ndash Marini and Ottonello(2002)
The Gibbs free energy of formation (ΔGo f ) of the two theoretical Fe3+ (Varadachari et al1994) and Fe2+ vermiculites were calculated using a previously published method (Vieillard
2002 and Vieillard pers comm)log K values extimated considering an ideal solid solution between the end-members and using data thermodynamic data from llnldat (clinochlore08-daphnite02) and Wilson
et al (2006) (liz09-berth01)
82 T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
references therein Venturelli et al 1997) Accessory phases areplagioclase picotite plusmn calciumndashsodic amphibole (Venturelli et al1997) Before modeling the crystal chemical formula of the mineralphases was calculated from data from Venturelli et al (1997) thendepending on the modal composition of the rock (Ernst and Piccardo1979) and the composition of the ideal solid solution of the minerals
the resulting stoichiometry of the solid reactant(s) was calculated andinserted in the input 1047297le of the PHREEQCI software (Table 3)Thermodynamic data of the primary mineral paragenesis are all wellknown and are included in the llnldat thermodynamic database of PHREEQCI
53 The solid reactant secondary paragenesis
Secondary minerals in the ultrama1047297tes of the TarondashCeno Valleysare serpentine (lizardite and rare chrysotile) smectite minerals (Fendash
Mg saponites) and Fe-oxi-hydroxides Other accessory minerals arechlorite mixed layer saponitendashchlorite (low-charge corrensite) talccalcite and dolomite (Beccaluva et al 1973 Dinelli et al 1997Venturelli et al 1997 Brigatti et al 1999 2000) In addition mixed
layer saponitendashtalc (aliettite) and vermiculitendashchlorite (high-charge
corrensite) are present (Alietti 1956 Settembre Blundo et al 1992Brigatti and Valdregrave 1996) Except for serpentine themodal abundanceof these phases is unknown In addition some of these minerals ndash
particularly interlayer clays ndash may have different origins (hydrothermalor weathering allogenicor authigenic)and different stabilitiesrelatedtokinetics and drainage 1047298ow rate For example chlorite weathering yields
vermiculite-bearing or saponite-bearing interlayers under good- andpoor-drainage regimes respectively (Bonifacio et al 1997 Środoń1999) In addition the thermodynamic data available for serpentineinclude information only on chrysotile which is less abundant thanlizardite in the serpentinites of the areas that we studied According toVenturelli et al (1997) the lizardite composition in the serpentinites of the Taro valley can be summarized as
which is similar to 1-T lizardite from the Monte Fico serpentinizedperidotite (Elba Island Northern Tyrrhenian Sea Italy Viti and Mellini1997)
Using the thermodynamic data of Wilson et al (2006) and
assuming an ideal solid solution between lizardite and berthierine
Fig 5 EhndashpH iron diagram of the spring waters coming from ultrama1047297c rocks of the TarondashCeno Valleys Bicarbonate and high-pH waters of the Voltri Group are also represented(dotted area andsmallblackdiamonds respectively) Thelowest Eh value of thehigh-pH waterswas calculated using theS(+VI)S(minus II)redox couple(Marini and Ottonello 2002)Thestability 1047297eld of mineral phases are in grey and those of aqueous species in white In diagrams A and B 6-line ferrihydrite-vermiculite and goethite were selected as subaerial solidphases respectively
83T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
we calculated ΔGf 0=minus399854 kJmol for a theoretical (09)(lizardite)+
(01)(berthierine) mineral (Table 2) The log K value of the stated solidsolution was calculated at different temperatures using PHREEQCIafter loading the thermodynamic data on the two end-members intothe database The log K results obtained for Al-lizardite were added tothe thermodynamic database of the solid reactant (Table 3) andcompared with data from the literature (Table 4)
54 Composition of the starting solution O 2 partial pressure and
minerals included in the model
The weighted average chemical composition of rainwater inBologna (Panettiere et al 2000) which is 95 km east of theinvestigated area was used as the starting solution in the modelFull chemical data on rainwater of the TarondashCeno Valleys are
unavailable
Total alkalinity was recalculated for log PCO2=minus35 bar and
T =11 degC which was the mean temperature of the springs sampledand for pe (or Eh) based on chemical data for the redox couple N(minusIII)N(+V) The values of pH and pe obtained were 566 and 452(Eh=025 V) respectively the former agrees with the median valuereported for rainwater in thearea (Venturelli et al1997) Because dataon Al Fe and Si in rainwater were lacking we calculated theconcentrations of these elements assuming logPCO2=minus30 bar andslight supersaturation relative to ferrihydrite-6 line and kaolinite Inthis way a hypothetical composition of water in equilibrium with thelocal soil was calculated (Table3) and adopted as the ldquoinitial waterrdquo formodeling
Evidence suggests that secondary minerals particularly Fe-bearingphases change the chemical composition of the water 1047298owingthrough the ultrama1047297tes of the TarondashCeno Valleys According to Neal
and Stanger (1983) and Fritz et al (1992) the production of CH4 and
Fig 6 Activity plotfor the MgOndashSiO2ndashAl2O3ndashFe2O3(FeO)ndashH2O systemat differentlog PO2 values Spring waters issuing from TarondashCeno Valleys (symbols as in Fig 2) andVoltri Groupultrama1047297tes are shown (grey area bicarbonate waters small black diamonds alkaline waters from Marini and Ottonello 2002) In plots A B and C the stability 1047297eld limits of vermiculite(Table 2) correspond to25 degCwhilethe amorphoussilica (am-SiO2) brucite serpentine (Sr1ndash4 in Table 4)and sepiolite(Sp1ndash2 in Table 4) phasescorrespondto 11 degCThepath calculatedfor interaction between waterand ultrama1047297c rockis also shown (dotted line)the number-letter combinations referto the appearanceand disappearance of minerals(see Table 5)
84 T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
H2 in thealkalinewaters1047298owing from theultrama1047297tes isrelatedto theconsumption of oxygen during the oxidation of iron forming thehematitendashmagnetite pair Therefore the activity of iron and theoxidation state of the water must be de1047297ned The analysis of watersamples from ultrama1047297tes were plotted in both the Pourbaix andactivity diagrams (Figs 5 and 6) where the calculated activity of Fe2+Mg2+ SiO2 SO4
2minus (Pourbaix) and log PO2 (activity diagrams) wereinvolved Iron concentration was quite variable in the TarondashCenosprings where the average content was 27 μ gL and the contentreached a maximum of 200 μ gL and in the Voltri springs where theaverage was 44 μ gL and the maximum was 948 μ gL (Marini andOttonello 2002) In both cases the variations are probably related to
the different mineralogies of the soils and the different solubilities of the Fe-rich phases like goethite and ferrihydrite Iron-bearingsaponites and montmorillonites are the prevailing clay minerals inthe area investigated probably because of the ubiquitous presence of iron oxy-hydroxide (Alietti et al 1976 G Venturelli pers comm) Forthis reason the model considers the involvement of pure Mg-smectiteto be negligible
In the Pourbaix diagram (Fig 5) the data from bicarbonate watersfall into the ferrihydrite-goethitendashvermiculite1047297elds and the data fromalkaline waters fall on the boundaries of the vermiculitesaponite andsaponitechlorite Since ferrihydrite is the 1047297rst precipitating Fe-phaseit was used rather than goethite in the modeling The diagram alsosuggests the transformation of ferrihydrite into magnetite during thechange from bicarbonate to alkaline composition but this process
requires a large amount of Fe2+ in solution (N2 mM Hansel et al2005) which is unlikely at this stage of the waterndashrock interactionThus the magnetite participating in the equilibrium is most likelyprimary in origin
The modeling and the activity diagrams (Figs 6 and 9) werecalculated using the average value PO2= 10minus26 or 10minus50 bar forbicarbonate and alkaline waters respectively These values arecompatible with those calculated by Marini and Ottonello (2002) forbicarbonate waters (PO2= 10minus40 bar) and for alkaline waters of theVoltri Group (10minus77 barbPO2b10minus42 bar) The lowest PO2 value whichwas used as a limit in the modeling was calculated from the S(+VI)S(minusII) redox couple Moreover the activity diagrams (Fig 6) indicatethat the data from the bicarbonate waters from the TarondashCeno Valleysfall between quartz whose log [SiO2] =minus433 is not shown because of
the slow precipitation kinetics under the PndashT conditions considered
here (Williams et al 1985) and amorphous silica boundary 1047297elds(Fig 6A) at the same manner bicarbonate waters with higher SiO2-activity from the Voltri Group fall on amorphous silica boundary(Fig 6B) In summary then the bicarbonate waters are supersaturatedin kaolinite ferrihydrite vermiculite and saponite Alkaline watersare also supersaturated with saponite (Fig 6AndashC) and they are alsoequilibrated with primary magnetite and under extreme conditionsprobably pyrite as well
55 Limitations of the model related to kinetics
Model calculations disregard the rates of mineral dissolution and
precipitation which becomes increasingly important as temperature
Table 3
Starting mineralogy (left side) and 1047298uid composition (right side) used in the modeling
Rock composition Fluid composition
Modal of the mineralsin the interacting rock
mol of the minerals usedin the modeling
meteoric water composition Initial 1047298uid composition used in the modeling
Spinel 003 Na 3523Eminus05 3523Eminus055 Spinel Chromite 001 S 3519Eminus05 3519Eminus05
Magnetite 0 Si ndash 1176Eminus06Kaolinite ndash 5258Eminus08
Ferrhydrite-6line ndash 5882Eminus07
The modal percentage of the minerals in a serpentinite from TarondashCeno Valleys (Ernst and Piccardo 1979) was re-calculated assuming an ideal solid solution between pure extremephasesand takingthe modal percentage tobe themass percentage Thelatter was used asthe solid reactant composition in thewaterndashrockinteraction modelPure solid phases werechosenfrom the llnldat thermodynamic database of the PHREEQCI softwareThe weighted average chemical compositionof the meteoric waterfrom Bologna (Panettiere et al 2000)was chosen as the initial 1047298uid composition in the modeling after its equilibration with kaolinite and ferrihydrite (see text for details)
Table 4
Summary of thermodynamic data for the sepiolite (Spn) and serpentine (Sr n) mineralsphases in italics do not refer to the mentioned hydrolysis reactions
ΔGf 0 log K Reference
(kJ molminus1) (at 25 degC)
Sepiolite (Spn) mdash Mg4Si6O15(OH)2middot6H2O+8H+=4Mg2++6SiO2+11H2Ominus925400 3004 Gunnarsson et al (2005)minus925133 3050andash3172 Christ et al (1973)
(Sp2) minus921986 3602andash3756 Wollast et al (1968)
Serpentine (Srn) mdash Mg3Si2O5(OH)4+ 6H
+
=3Mg
2+
+2SiO2+ 5H2ONi-lizardite minus 289324 1680 Mondesir and Decarreau (1987)
Al-lizardite2 minus411003 2544a Caruso and Chernosky (1979) Al-lizardite minus 339854 2786 this work (see text for details)
Mg-lizardite (Sr1) minus404025 3056 Wilson et al (2006)Mg-crysotile (Sr2) minus403702 3112 bthermocomV8R6Mg-lizardite minus402824 3340 Mondesir and Decarreau (1987)p-antigorite (Sr3) minus564239a 3479 Gunnarsson et al (2005)Mg-chrysotile (Sr4) minus406809 3669 Bruni et al (2002)Mg-chrysotileasbestos minus562505a 3888 Pfeifer (1977)Antigorite minus6612864 47719 bthermocomV8R6
The 1047297eld boundaries of the underlined phases shown in Fig 6 correspond to atemperature of 11 degC
a In the ΔG f deg column value calculated using the ΔG f
deg data of the elements from llnldat
and logK value in the reference in the LogK column value calculated using ΔG f deg
(mineral) in the reference and ΔG f deg data of the elements from llnldat
b Thermodynamic database of Lawrence Livermore National Laboratory included in
PHREEQCI and Geochemists Workbench softwares
85T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
decreases The waterndashrock interaction in the TarondashCeno basin iscomplicated by the presence of secondary mixed-layer clay mineralsthe origin of which is quite dif 1047297cult to model because the solidndashstatetransformations and weathering reactions are strictly related tohydrological and climatic conditions as described above An equili-
brium-based reaction path model is a valuable tool for assessing the
most probable reactions and for de1047297ning the product phases that canbe used in a kinetics-based model Since the kinetics of precipitationof Mg-clays at room temperature and pressure are poorly knownrelative reaction rate effects in the reaction path model can be mimedby omitting the stabler phase when a metastable mineral is expected
to precipitate Therefore the precipitation of MgndashFe3+
Abbreviations are as follows Fer-6line 6-line ferrihydrite kaol kaolinite verm vermiculite magn magnetite sap saponite serp serpentine clin clinochlore pyr pyrite calccalcite hydrom hydromagnesite bruc brucite enst enstatite The chromium decrease during the change in water facies from bicarbonate to alkaline was modeled withsupersaturation (precipitation) of eskolaite and thereafter with equilibrium with chromite (Boschetti 2003)- not bound dagger chlorite mineral from solution Dagger chlorite mineral added to the paragenesis
86 T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
montmorillonites vermiculite and saponites are predicted to occurrather than that of Mg-clays Moreover in order to simulate the slowprecipitation of calcite in depth precipitation was not activated atslight supersaturation (SIcalcite=001) during the initial stages of thewaterndashrock interaction in this manner saturation index of calcite wasfree to increase in the following steps and in alkaline waters
6 Discussion of the model results
61 Explanation of pre-modeling choices and post-modeling results
The results of the reaction path modeling are reported in Table 5and shown in Figs 6 and 9 The reaction paths model the dissolutionof a typical local serpentinite (mineralogically constrained as inTable 3) and the sequential precipitation of gibbsite kaoliniteferrihydrite vermiculite Fe2+ndashMg2+-saponite and a poorly-crystal-line serpentine( p-antigorite of Gunnarsson et al2005) The1047297rst stepinvolves the reaction of the ldquoinitial waterrdquo with serpentinite atconstant temperature (11 degC) PO2 (10minus26 bar) and PCO2 (10minus30 bar)this early stage corresponds to bicarbonate groundwater feeding theinvestigated spring system At the stage of primary magnetiteequilibrium and Fe2+ndashMg2+-saponite supersaturation the modelassumes the system to be closed to CO2 and to have a 1047297xed partialpressure of oxygen (PO2= 10minus40 bar) these conditions represent
when water switch from a bicarbonate to an alkaline compositionThe calculated amount of interacting serpentinite needed to obtainthe alkaline springs of the TarondashCeno Valleys (mean PCO2= 10minus7 bar)is approximately 00203 moles (see Table 5)
Bruni et al (2002) explained the origin of the alkaline waters of the Voltri Group as the result of the evolution of meteoricbicarbonate water following the precipitation of sepiolite withabundance similar to that of goethite This might be supported bythe observation that sepiolite or smectite precipitate at lowtemperature from solutions with low Al content when the pH is inthe range 75ndash95 or N95 respectively (Siffert and Wey 1962 Jonesand Galan 1988) but sepiolites form from the weathering of serpentines in arid climate regions whereas sepiolites are rare inhumid climate regions where serpentinite changes to Fe-smectite
(Singer 1989 Wrucke 1996) In the ophiolite rocks of the Northern
Apennine sepiolite is found only sporadically in veins (Giuseppetti1953 Veniale 1966) and is unknown in the ophiolites of the Tarondash
Ceno Valleys whereas it is thought that sepiolite of the Voltri Groupprecipitated from metasomatic 1047298uids during development of shearzones (Crispini and Capponi 2002)
As described above our model explains the transition frombicarbonate water to alkaline water (UM15 and PR10 samples) asthe result of low-temperature precipitation of a poor crystalline
serpentine with the solubility product of p-antigorite (see Sr3 inTable 4 and Fig 6) The model indicates that the bicarbonate waters of the Voltri Group with the highest [Mg2+][H+]2 activity ratios similarlyevolve towards alkaline waters (Fig 6B) The logK value of thesynthetic low temperature serpentine adopted in this model and its
Fig 7 Saturation indices of chrysotile and p-antigorite (Sr2 and Sr3 thermodynamic
data respectively in Table 4) for bicarbonate and alkaline waters coming fromultrama1047297c rocks of the TarondashCeno Valleys as a function of SiMg molar ratioThermodynamic data on chrysotile and p-antigorite are from the thermocomV8R6
database and Gunnarsson et al (2005) respectively
Fig 8 Activityplot forthe MgOndashSiO2ndashAl2O3ndashFe2O3(FeO)ndashH2O system analyzed forSiO2-phases and aqueous species at different log PO2 and log[aAl3+(aH+)3] values Springwaters coming from the TarondashCeno Valleys (symbols as in Fig 2) and Voltri Groupultrama1047297tes (symbols as in Fig 5) are shown The greyarea refers to bicarbonate watersfrom theVoltri Group (Mariniand Ottonello2002Bruniet al 2002) Thesepiolite 1047297eldis very narrow or hidden by saponite-Mg (1047297eld limits shown as continuous lines)saponite-Fe3+ndashMg (1047297eldlimits shown as dottedlines) or vermiculite(1047297eldlimits shownas dottedndashdashed lines) The talc stability 1047297eld and boundaries with saponite-Mg
montmorillonite-Mg and cronstedite-Mg are shown for comparison (dashed lines)
87T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
relative uncertainty of plusmn2 (see Fig 8 in Gunnarsson et al 2005) arecomparable to the logK variation reported for other low-temperatureMg-serpentines (Table 4) that is within the lines Sr1ndash4 for serpentineminerals where samples varying from bicarbonate composition toalkaline composition in the ultrama1047297c aquifer are grouped (Fig 6)Fig 7 shows a plot of calculated serpentine saturation index as a
function of the SiMg molar ratio for the water samples from theultrama1047297tes of the TarondashCeno ValleysThe SI andSiMgvalues of pureMg-serpentine (SI= 0 SiMg=067) provide a useful index forseparating bicarbonate waters (SIb0 SiMgb067) from alkalinewaters (SIN0 SiMgN067) and the distribution of data points isconsistent with the idea that bicarbonate waters approach saturationwith serpentine until they become supersaturated as alkaline waters
Based on the Mgvs SiO2(aq) activitydiagram and speci1047297cratiosoflog[aAl3+(aH+)3] Birsoy (2002) analyzed the behavior of sepiolite duringultrama1047297te weathering The diagram produces a very narrow stability1047297eldfor sepiolite(Fig 8) when it is programmedwith suitable conditionsfor the bicarbonate waters (log PO2=minus25 bar log [aAl3+(aH+)3]= 65) andthe alkaline waters (log PO2=minus50 bar log [aAl3+(aH+)3]= 45) sampled inthe present study This stability 1047297eld grows larger when the absence of
Mg-smectites and talc is assumed but this is an unrealistic conditionbecause (i) all these minerals occur in the rocks within the areainvestigated and (ii) at high [Mg2+][H+]2 activity ratios the 1047297eld of sepiolite is smaller than that of subordinate compared to those poorly-crystalline serpentine (p-antigorite) and clinochlore
Among the alkaline springs of the Voltri Group the waters withlow log [aCa2+(aH+)2] values (Fig 9) show patterns compatible withchlorite interaction (step 5c in Fig6D)The alkaline waters whose datafall near the clinochlore-14Aring 1047297eld (Fig 6D) cannot be explained bysepiolite precipitation because the sepiolite line does not cross thechlorite 1047297elds Thus the peculiar composition of some alkaline watersof the Voltri Group is probably due to interactions between water andprimary chlorites of the ultrama1047297c rocks (optional step 5c in Table 5cb) This is supported by the occurrence sometimes in large amounts
of metasomatic Mgndash(Fe)-chlorite (Scambelluri and Rampone 1999
Vissers and Strating 2001) in the Ligurian area where the relativeabundance of chlorite can reach 50 in veins in high strainserpentinite mylonites of the ErrondashTobbio ultrama1047297tes (Scambelluriet al 2001b) Otherwise in a chlorite-free primary paragenesis thesolution evolving on the Sr3 serpentine line reaches equilibrium withclinochlore but probably the latter does not precipitate due to theslow kinetics of precipitation under these conditions Thus thedissolution of additional reacting rock in the iterative weathering
process supplies Mg
2+
to the solution thereby favoring the Sr3 to Sr4poorly crystalline serpentine transformation In fact all the PR10samples and some of the alkaline springs from the Voltri Group showincreasing [Mg2+][H+]2 activity ratio and low variation in silicaactivity (Fig 6D) starting from the invariant triple point 4a (Sr3-serpentine Fe2+ndashMg-saponite chlinoclore-14 Aring saturation) andextending towards the brucite line This path can be explained asthe result of the transformation of poorly crystalline serpentine fromSr3 to Sr4 the latter is more stable at higher [Mg2+][H+]2 activityratios Finally at very low PO2 values vermiculite disappears and thestability 1047297elds of chlorite and Fe2+ndashMg-saponite become narrowerwhereas those of Fe2+-rich phases (eg Fe2+-saponite) become widerThe data for the alkaline waters with the highest log [aMg2+(aH+)2]from the Voltri Group fall near the stability 1047297eld of peridotite primaryminerals like forsterite and enstatite (Fig 6D) therefore the last stepsof theevolution involve theequilibration of thesolutionwithenstatite(steps 8a and 8c in Table 5ab)
62 Ca-rich alkaline waters
In preliminary modeling of the evolution of the TarondashCeno springwaters Boschetti (2003) concluded that meteoric water initiallysaturated in Fe3+ndashMg2+-chlorite Fe3+ndashMg2+-saponite and pyrite (andor goethite) and then interacted with the ultrama1047297te mineralassemblage eventually evolved into the alkaline waters In themodel of Boschetti (2003) chlorite saponite and calcite were added
Fig 9 Activity plot for the MgOndashCaOndash(CO2)ndashH2O system with phase limits calculated ata temperature of 11 degCSpring waterscoming from theTarondashCeno Valleys (symbols as inFig2) andVoltri Groupultrama1047297tes(symbolsas in Fig5) areshown Calculated path forwaterndashultrama1047297c rock interaction is shown (dotted line) with the number ndashlettercombinations referring to the appearance or disappearance of minerals (see Table 5)The grey area refers to bicarbonate waters from the Voltri Group (Marini and Ottonello2002)
Fig 10 Plots of (A) calcite saturation indexes and (B) PCO2 as a function of Clminus ascalculated by PHREEQCI software for waters from the TarondashCeno Valleys Alkalinesprings for the Voltri Group are also shown (small black diamonds Marini andOttonello 2002) The grey shading is deeper(A) in the saturation1047297eld(SI= 00plusmn05) and
(B) where the lowest log PCO2 value is approached
88 T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
to the solution in order to achieve a high [Ca2+][H+]2 activity ratio andsupersaturation in these phases However in light of the work of Clarket al (1992) and Bruni et al (2002) we believe that the calcitesupersaturation in alkaline waters (Fig 9 and 10) which in our modelis achieved by preventing the precipitation of calcite in the 1047297rst stepsof the reaction path (Table 5a and b) depends on the differencebetween CO2 partial pressureat thegivendepth(PCO2le10minus10bar) andat the surface (PCO2cong10minus35) Waters rising to the surface are far from
equilibrium with the atmosphere where the PCO2 is 4times10
6
timesgreater than the value at depth thus the water rapidly take up CO2which is promptly converted to CO3
2minus because of the high pH Thisleads to supersaturation with calcite through the following hydro-xylation reaction
As a result large amounts of travertine are generated in the VoltriGroup (Marini and Ottonello 2002) and carbonate concretionsknown as devilrsquo s coins are formed at Mt Prinzera in Taro valleyBelow the surface the SIcalcite of the alkaline waters is expected to belower than what the model estimates Under conditions of lowcarbonate alkalinity low CO2 and O2 partial pressure and high pHcalcite nucleation may be inhibited and the kinetics of precipitationreduced by the high concentrations of dissolved divalent cations suchas Mg2+ Fe2+ and Mn2+ (Sumner and Grotzinger 1996 Davis et al2000 Zhang and Dawe 2000 de Leeuw 2002) The fact that severalalkaline water samples fall within or near the calcitendashdolomiteequilibrium (Fig 9) may indicate magnesium calcite or protodolomitenucleation in those extreme subsurface conditions However possibleinteraction between some calcium-rich waters of the Voltri Group andCa-rich rocks such as ophicarbonates or rodingites cannot beexcluded these rocks are widespread in the Ligurian Apennineserpentinites (Treves and Harper 1994)
63 Sepiolite vs serpentine nucleation under weathering conditions
constraints on the kinetics and equilibrium-based approaches
The kinetics of precipitation of poorly-crystalline serpentine and
sepiolite cannot be compared quantitatively because rate constantsare available only for well-ordered crystalline phases (eg see Palandriand Kharaka 2004 for crystalline serpentine) Sepiolite precipitationis often invoked to explain magnesium scavenging from solutions butthe lack of kinetic data for this mineral leads to overly simplisticreaction transport models These models are forced to use kinetic datafor phyllosilicates such as kaolinite or muscovite (eg Xu et al 2001Bryan 2005) Moreover the low-temperature thermodynamic dataeven if they are fairly uncertain show that sepiolite solubilityincreases with decreasing crystallinity (Table 4 Stoessell 1988) Infact crystalline sepiolite requires a long time to reach equilibriumwith solution which imposes a slow control on thewater composition(eg Stoessell and Hay 1978) Consequently in modeling based onreaction transport that simulates freshly precipitated or poorly
crystalline sepiolite formation the logK value of Wollast et al (1968)is preferable to the lower values (Suarez and Šimůnek 1996) Thispreference agrees with the approach of Mariner (2003) and Jove Colon(2006) who adopt a logK (25 degC) value for sepiolite of 3044 It is alsosimilar to that reported by Wollast et al (1968) and it is six logK unitshigherthan the valueused in thermocomV8R6 dataset which is oftenused (perhaps improperly) to model the supergenic nucleation of sepiolite In the stability diagram in Fig 6 where the Sp1 and Sp2 linesrepresent the stability boundaries of crystalline and poorly-crystallinesepiolite respectively the evolution of TarondashCeno waters frombicarbonate to alkaline (Figs 6A and C) follows the stability 1047297eldboundary of the poorly-crystalline serpentine (between Sr2ndashSr3 lines)rather than that of the poorly-crystalline sepiolite (Sp2) as postulatedin themodel The lines for Sp2 Sr3 and amorphoussilica intersect at a
common point (Fig 6D) which is consistent with the nucleation of
sepiolite that typically occurs in solutions that are saturated orsupersaturated with amorphous silica and (Mg)-carbonate (egStoessell and Hay 1978 Birsoy 2002) Some bicarbonate watersfrom the Voltri Group have a silica activity near the amorphous silicaboundary (Fig 6B) but only a few alkaline springs of this group fall onthe poorly-crystalline sepiolite boundary line (Sp2 Fig 6D) Both of these features are inconsistent with the possibility that sepioliteprecipitates from supergenic fresh water in the Northern Apennine
Moreover these two results suggest that values for logK(sepiolite)b
3044 (see Table 4) should be used to model the dissolutionprecipitation of the well-crystallized phase that is the most commonform for the metasomatic sepiolite found in the investigated area
7 Conclusions
The waters issuing from ultrama1047297tes in the TarondashCeno Valleysoriginate from low-temperature interactions between meteoric waterand serpentinite rocks In this report the waterndashrock interaction hasbeen modeled using reaction paths in progress mode During their earlyinteraction with soil and rock minerals bicarbonate waters super-saturated in kaolinite ferrihydrite vermiculite and Fe2+ndashMg2+-saponitewhere theorigin of the secondary iron saponite as Fe2+-bearing smectiteagrees with the hypothesis of Guumlven (1988) Then while remainingsupersaturated with saponite the bicarbonate waters evolve into analkaline facies when poorly-crystalline serpentine precipitates Duringrock weathering the Al-bearing crystalline serpentine of oceanic originprobably promotes nucleation of the low-temperature serpentineamorphous phase and supplies Al3+ to the solution for saponiteprecipitation This interpretation is consistent with the existence of Al-poor supergenic serpentine (Neal and Stanger 1985) and the partiallymeteoric OH-isotopic imprint observed in other ultrama1047297c systems
Dif 1047297culties and complexities in the modeling arise fromuncertainties in the primary mineralogy of some local ophiolitesand from the presence of different lithotypes (rodingites ophi-carbonates gabbros) which can signi1047297cantly affect the composi-tion of water samples This can be seen in the example of the Candash
OH springs issuing from the Voltri Group (PO2 and PCO2 of 10minus70
and 10minus10 bar respectively Bruni et al 2002 Marini and
Ottonello 2002) These springs appear to be more evolved thanthe TarondashCeno alkaline waters as a result of their particularly longinteractions with serpentinites until equilibrium with enstatite isreached or with chlorite-bearing serpentinite The use of anupdated thermodynamic database in our model avoided problemsdue to mineralogical heterogeneity in the mixed-layer claysassociated with ophiolitic rocks this database accurately re1047298ectedthe chemical composition of the single phases Using this modelthe vermiculitendashchlorite and saponitendashchlorite associations seen inthe activity diagrams simulate respectively the presence of highand low charge corrensite observed in the 1047297eld The accuracy of ourmodel can be tested by comparing estimates of chlorine composi-tion resulting from our model and from experimental methods The
mean Cl concentration of the bicarbonate water is 41plusmn16 mgLtaking into account a Cl concentration of sim2 mgL in rainwater inthe area (cf Panettiere et al 2000 Venturelli 2003) 2 mgL of Clin spring water are estimated to come from serpentinite This valueagrees with the model which predicts that the switch frombicarbonate water to alkaline water requires the congruentdissolution of 00203 mol of serpentinite containing 100 ppm of soluble Cl thereby yielding water containing 2 ppm of Cl
In addition we think that weathering of the accessory albiteplagioclase is responsible for Na enrichment of the waters whereas Clenrichment in the alkaline waters is predicted to result from theleaching of NaCl 1047298uid inclusions Moreover the TarondashCeno alkalinesprings show the highest BCl ratios comparable with those of thehost metamorphic minerals (Scambelluri et al 2004) In contrast the
lower boron content and BCl ratio of the alkaline springs of the Voltri
89T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
Group are unexpected features and probably re1047298ect waterndashrockinteractions over a longer time period than in the TarondashCeno ValleysThe boron content in the Northern Apennine alkaline springs may becontrolled by several factors such as (i) the different histories of VoltriGroup and EL ultrama1047297tes (ii) the probable contribution of boronfrom NaCl 1047298uid inclusions (iii) the sinking of boron by travertineprecipitation (Hobbs and Reardon1999) brucite (Rhoades et al 1970)and smectite and (iv) decreases in boron release from minerals due to
increases in pH (Su and Suarez 2004) Acknowledgements
This research was carried out with the 1047297nancial support from theldquoYoung Researchers Projectrdquo (2001ndash2002) of the Italian Ministry of Universities Scienti1047297c Research and Technology (MURST) We thankGianniCortecci (Institute of Geosciences and EarthResources IGG-CNRPisa Italy) and Giampiero Venturelli (Earth Sciences Department Univer-sity of Parma Italy) for useful suggestions and critical review of themanuscript The authors are grateful to Enrico Dinelli (InterdepartmentResearch Centre for Environmental Sciences Alma Mater StudiorumUniversity of Bologna Ravenna Italy) for granting permission to use theGeochemistrsquos WorkbenchsoftwareTheconstructive reviews of theEditor
J Fein and of two anonymous Referees have signi1047297cantly improved the
manuscript
Appendix A Supplementary data
Supplementary data associated with this article can be found inthe online version at doi101016jchemgeo200808017
References
Abbate E Bortolotti V Passerini P 1980 Olistostromes and olistoliths In Sestini G(Ed) Development of the Northern Apennines geosyncline Sedimentary Geology4 pp 521ndash558
Alietti A 1956 Il minerale a strati misti saponite-talco di monte Chiaro (val di TaroAppen Emil) Rendiconti dellAccademia Nazionale dei Lincei 21 201 ndash207
Alietti A De Angelis G Montanari G Sgarlata F Tirelli G 1976 Il ferro nelle argilleappenniniche Periodico di Mineralogia 45 51ndash64
Anselmi B Mellini M Viti C 2000 Chlorine in the Elba Monti Livornesi and Murloserpentines Evidence for seandashwater interaction European Journal of Mineralogy12137ndash146
Apha-Awwa-Wef 1995 Standard Methods for the Examination of Water and Waste-water 19th Edition American Public Health Association American Water WorksAssociation Water Environment Federation
Barnes I LaMarche Jr VC Himmelberg GR 1967 Geochemical evidence of present-day serpentinization Science 156 830ndash832
Barnes I ONeil JR 1969 The relationship between 1047298uids in some fresh alpine-typeultrama1047297cs and possible modern serpentinization Western United StatesGeological Society of America Bulletin 80 1947ndash1960
Barnes I ONeil JRTrescases JJ1978 Present dayserpentinizationin New CaledoniaOman and Yugoslavia Geochimica et Cosmochimica Acta 42 144ndash145
Barnes I Rapp JB ONeil JR Sheppard RA Gude III AJ 1972 Metamorphicassemblages and the direction of 1047298ow of metamorphic 1047298uids in four instances of serpentinization Contributions to Mineralogy and Petrology 35 263ndash276
Barnes JDSelverstone J Sharp ZD 2006Chlorine isotope chemistry of serpentinitesfrom Elba Italy as an indicator of 1047298uid source and subsequent tectonic historyGeochemistry Geophysics Geosystems 7 doi1010292006GC001296
Beccaluva L Emiliani F Venturelli G Zerbi M 1973 Ca Fe Mg Mn Cr Ni Codistribution in some ultrama1047297c rocks outcropping in the Northern Apennines withsome geological remarks Ateneo Parmense Acta Naturalia 9 69ndash98
Beccaluva L Macciotta G Piccardo GB Zeda O 1984 Petrology of lherzolitic rocksfrom the Northern Apennine ophiolites Lithos 17 299 ndash316
BeccaluvaL Venturelli G ZanzucchiG 1975 Dati geologici e geochimici sui basalti diassociazione o1047297olitica dellAppennino ligure-emiliano Ateneo Parmense ActaNaturalia 11 789ndash802
BethkeCM 2002 The Geochemists Workbench A Userrsquos Guide to RxnAct2Tact ReactandGtplot Release40 HydrogeologyProgram Universityof IllinoisUrbanaIL 236pp
Birsoy R 2002Formation of sepiolitendashpalygorskite and related minerals from solutionClays and Clay Minerals 50 736ndash745
Bonatti E Lawrence JR Morandi N 1984 Serpentinization of oceanic peridotitestemperature dependence of mineralogy and boron content Earth and PlanetaryScience Letters 70 88ndash94
Bonifacio E Zanini E Boero V Franchini-Angela M 1997 Pedogenesis in a soilcatena on serpentinite in north-western Italy Geoderma 75 33ndash51
Boschetti T 2003 Studio geochimico e geochimico-isotopico di acque a composizione
estrema e termali dellrsquoAppennino Settentrionale Ph D Thesis Parma Univ
Brigatti MF Franchini G Lugli C Medici L Poppi L Turci E 2000 Interactionsbetween aqueous chromium solutions and layer silicatesApplied Geochemistry 151307ndash1316
Brigatti MF Lugli C Poppi L Venturelli G 1999 Iron-rich saponite dissolutionreactions and Cr uptake Clay Minerals 34 637ndash645
Brigatti MF Valdregrave G 1996 Mixed-layer chloritesmectite (corrensite) in ophioliticrock alteration products Neues Jahrbuch fuer Mineralogie Monatshefte 1 37 ndash47
Bruni J Canepa M Chiodini G Cioni R Cipolli F Longinelli A Marini L OttonelloG Vetuschi Zuccolini M 2002 Irreversible waterndashrock mass transfer accompany-ing the generation of the neutral MgndashHCO3 and high-pH CandashOH spring waters of the Genova province Italy Applied Geochemistry 17 455ndash474
Bryan CR 2005 Drift-Scale THC Seepage Model MDL-NBS-HS-000001 REV 04 YuccaMountain Project Las Vegas Nevada (US)Caruso LJ Chernosky JV 1979 The stability of lizardite Canadian Mineralogist 17
757ndash769Clark ID Fontes JC Fritz P 1992 Stable isotope disequilibria in travertine from high
pH waters Laboratory investigations and 1047297eld observations from Oman Geochi-mica et Cosmochimica Acta 56 2041ndash2050
Cortesogno L 1980 Il metamor1047297smo giurassico nelle o1047297oliti dellAppenninoSettentrionale due distinti cicli metamor1047297ci in ambiente oceanico O1047297oliti 5 5ndash18
Christ CLHostetler PB Siebert RM1973 Studies in the system MgO-SiO2-CO2-H2OIII American Journal of Science 273 65ndash83
Cortesogno L Lucchetti G 1982 Il metamor1047297smo oceanico nei gabbri o1047297oliticidellAppennino Ligure Aspetti mineralogici e petrogenetici Rendiconti dellaSocietagrave Mineralogica Italiana 38 561ndash579
Crispini L Capponi G 2002 Control of metasomatic alteration on the development of shear zone in ultrama1047297c rocks a case study from the Voltri Massif (Ligurian Alps)Architettura interna e cinematica delle zone di taglio Gruppo Italiano di GeologiaStrutturale mdash Riunione Annuale 2002 Pisa Italy 11ndash12 giugno 2002 XIV ndashXV
CrissRE 1999 Principlesof stableisotope distributionOxford University PressNew York
Davis KJ Dove PM De Yoreo JJ 2000 The role of Mg 2+ as an impurity in calcitegrowth Science 290 1134ndash1137
de Leeuw NH 2002 Moleculardynamics simulations of the growth inhibiting effect of Fe2+ Mg2+ Cd2+ and Sr2+ on calcite crystal growth Journal of Physical Chemistry B106 5241ndash5249
De Nardo MT Segadelli S Vescovi P 2007 Studio pilota per la caratterizzazionegeologicadellesorgentinellazona delM Nero (altaVal Cenoe alta ValNuremdash provincedi Parma e Piacenza) Il Geologo dellEmilia-Romagna 25 Nuova Serie pp 5 ndash21
Di Giulio A Geddo G 1990 Studio petrogra1047297co delle Arenarie di Casanova (alta valTrebbia Appennino settentrionale) Atti Ticinesi di Scienze della Terra 33 243ndash254
Dinelli E Lombini A Simoni A Ferrari C 1997 Heavy metals in serpentinite soils of selected outcrops of Piacenza and Parma provinces (Northern Apennines Italy)Mineralogica et Petrographica Acta 40 241ndash255
Epstein S Mayeda T 1953 Variation of 18O content of waters from natural sourcesGeochimica et Cosmochimica Acta 4 213ndash224
Ernst WG Piccardo GB 1979 Petrogenesis of some Ligurian peridotites I Mineraland bulk-rock chemistry Geochimica et Cosmochimica Acta 43 219ndash237
Feder F Trolard F Klingelhofer G Bourrieacute G 2005 In situ Moumlssbauer spectroscopy
evidence for green rust (fougerite) in a gleysol and its mineralogical transforma-tions with time and depth Geochimica et Cosmochimica Acta 69 4463ndash4483Fritz PClark IDFontes JCWhiticar MJFaberE 1992 Deuterium and
13C evidence
for low temperature production of hydrogen and methane in a highly alkalinegroundwater environment in Oman In Kharaka TK Maest AS (Eds) Proceed-ings of the 7th International Symposium on WaterndashRock Interaction BalkemaRotterdam pp 793ndash796
Giammetti F 1968 Le o1047297oliti di Roccaprebalza Costa della Guardia e Groppo di Gorro(Appennino Parmense) Ateneo Parmense Acta Naturalia 4 107ndash142
Giuseppetti G1953 Roccee minerali dellaformazioneo1047297olitica di Volpedo Rendicontidella Societagrave Mineralogica Italiana 9 85ndash134
Gran G 1952 Determination of equivalence point in the potentiometric titrationAnalyst 77 661ndash671
Gunnarsson I Arnorsson S Jakobsson S 2005 Precipitation of poorly crystalline antigoriteunder hydrothermal conditions Geochimica et Cosmochimica Acta 69 2813ndash2828
Guumlven N 1988 Smetites In Bailey SW (Ed) Reviews in Mineralogy 19 Hydrousphyllosilicates (exclusive of micas) Mineralogical Society of America MichiganBook Crafters Inc 497ndash559
Hansel CM Benner SG Fendorf S 2005 Competing Fe(II)-induced mineralization
pathways of ferrihydrite Environmental Science and Technology 39 7147ndash7153HobbsMYReardonEJ 1999 Effect of pH on boronco-precipitation by calcite further
evidence for non-equilibrium partitioning of trace elements Geochimica etCosmochimica Acta 63 1013ndash1021
Jones BF Galan E 1988 Sepiolite and palygorskite In Bailey SW (Ed) HydrousPhyllosilicates (exclusive of Micas) Reviews in Mineralogy vol 19 MineralogicalSociety of America Michigan Book Crafters Inc 631ndash674
Jove Colon CF 2006 Technical work plan for Thermodynamic database for chemicalmodelling TWP-MGR-PA-000039 REV 01 Yucca Mountain Project Las VegasNevada (US)
Kohls DW Rodda Jr JL 1967 Iowaite a new hydrous magnesium hydroxidendashferricoxychloride from the Precambrian of Iowa American Mineralogist 52 1261 ndash1271
Kyser TK OHanley DS Wicks FJ 1999 The origin of 1047298uids associated withserpentinization processes evidence from stable-isotope compositions CanadianMineralogist 37 223ndash237
Lapham DM 1961 New data on deweylite American Mineralogist 46 168ndash188Lee BD SearsSKGraham RC Amrhein CValiH 2003 Secondarymineral genesis
from chlorite and serpentine in an ultrama1047297c soil toposequence Soil Science
Society of America Journal 671309ndash1317
90 T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
Longinelli A Selmo E 2003 Isotopic composition of precipitation in Italy a 1047297rstoverall map Journal of Hydrology 270 75ndash88
Majzlan J Navrotsky A Schwertmann U 2004 Thermodynamics of iron oxides PartIIIEnthalpiesof formation and stability of ferrihydrite (simFe(OH)3) schwertmannite(simFeO(OH)34(SO4)18) and ɛ-Fe2O3 Geochimica et Cosmochimica Acta 681049ndash1059
Mariner P 2003 In-Drift PrecipitatesSalts Model ANL-EBS-MD-000045 REV 01 YuccaMountain Project Las Vegas Nevada (US)
Marini L Ottonello G 2002 Atlante degli acquiferi della Liguria Volume III Le acquedeicomplessio1047297olitici (bacini ArrestraBranega Cassinelle Cerusa ErroGorzenteLeira Lemme Lerone Orba Piota Polcevera Rumaro Sansobbia Stura Teiro
Varenna Visone) Pacini PisaMellini L Zanazzi PF 1987 Crystal structures of lizardite-l T and lizardite-2H1 fromColi Italy American Mineralogist 72 943ndash948
Mezzadri G 1964 Petrogra1047297a delle arenarie di Ostia Rendiconti Societagrave MineralogicaItaliana 20 192ndash228
Mondesir H Decarreau A 1987 Synthesegrave entre 25 et 200 degC de lizardites Ni ndashMgMesure des coef 1047297cients de partage solide-solution aqueuse pour le couple Ni ndashMgdans des lizardites Bulletin de Mineralogie 110 409ndash426
Neal C Shand P 2002 Spring and surface water quality of the Cyprus ophiolitesHydrology and Earth System Sciences 6 797ndash817
Neal C Stanger G 1983 Hydrogen generation from mantle source rocks in OmanEarth and Planetary Science Letters 66 315ndash320
Neal C Stanger G 1984 Calcium and magnesium hydroxide precipitation fromalkaline groundwaters in Oman and their signi1047297cance to the process of serpentinization Mineralogical Magazine 48 237ndash241
Neal C Stanger G 1985 Past and present serpentinisation of ultrama1047297c rocks anexample from Semail ophiolite nappe of Northern Oman In Drever JI (Ed) TheChemistry of Weathering D Reidel Publishing Company pp 249ndash275
Nesbitt HW Bricker OP 1978 Low temperature alteration processes affecting
et Cosmochimica Acta 41 1835ndash1841OrsquoHanley DS 1996 Serpentinites records of tectonic and petrological history Oxford
University PressOrberger B Friedrich G Woermann E 1990 The distribution of halogens and carbon
in Pge-bearing ultrama1047297cs of the Acoje Ophiolite Block Zambales Philippines Journal of Geochemical Exploration 37 147ndash169
Ottonello G Joron JL Piccardo GB 1984 Rare Earth and 3d transition elementsgeochemistry of peridotitic rocks II Ligurian peridotites and associated basalts Journal of Petrology 25 373ndash393
Palandri JL Kharaka YK 2004 A compilation of rate parameters of water-mineralinteraction kinetics for application to geochemical modeling Open File Report2004-1068 US Geological Survey
Panettiere P Cortecci G Dinelli E Bencini A Guidi M 2000 Chemistry and sulfurisotopic composition of precipitation at Bologna Italy Applied Geochemistry 151455ndash1467
Parkhurst DL Appelo CAJ 1999 Userrsquos guide to PHREEQC (version 2) mdash a computerprogram for speciation batch-reaction one-dimensional transport and inversegeochemical calculation US Department of the Interior US Geological SurveyWashington DC
Pfeifer HR 1977 A model for 1047298uids in metamorphosed ultrama1047297c rocks Observationsat surface and subsurface conditions (high pH spring waters) SchweizerischeMineralogische und Petrographische Mitteilungen 57 361ndash396
PlummerLN1984Geochemical modeling A comparisonof forwardand inverse methodsIn Hitchon B Wallick EI (Eds) Proceedings First CanadianAmerican Conference onHydrogeology-Practical Applications of Ground Water Geochemistry Banff AlbertaCanada mdash Worthington Ohio National Water Well Association pp 149ndash177
Rampone E Hofmann AW Piccardo GB Vannucci R Bottazzi P Ottolini L 1995Petrology mineral and isotope geochemistry of the External Ligurides peridotites(Northern Apennines Italy) Journal of Petrology 36 81ndash105
Rampone E Hofmann AW Piccardo GB Vannucci R Bottazzi P Ottolini L 1996Trace element and isotope geochemistry of depleted peridotites from an N-MORBtype ophiolite (Internal Liguride N Italy) Contributions to Mineralogy andPetrology 123 61ndash76
Rhoades JD Ingvalson RD HatcherJT1970Adsorption of boronby ferromagnesianminerals and magnesium hydroxide Soil Science Society of America Journal 34
938ndash941Saini-EidukatB KuchaH Keppler H1994Hibbingiteγ-Fe2(OH)3Clanewmineralfromthe Duluth Complex Minnesota with implications for the oxidation of Fe-bearingcompounds and the transport of metals American Mineralogist 79 555ndash561
Sanford RF 1981 Mineralogical and chemical effects of hydration reactions andapplications to serpentinization American Mineralogist 66 290ndash297
Scambelluri M Muumlntener O Ottolini L Pettke TT Vannucci R 2004 The fate of BCl and Li in the subducted oceanic mantle and in the antigorite breakdown 1047298uidsEarth and Planetary Science Letters 222 217ndash234
Scambelluri M Rampone E 1999 Mg-metasomatism of oceanic gabbros and itscontrol on Ti-clinohumite formation during eclogitization Contributions toMineralogy and Petrology 135 1ndash17
Scambelluri M Bottazzi P Trommsdorff V Vannucci R Hermann J Gogravemez-Pugnaire MT Logravepez-Sagravenchez Vizcaino V 2001a Incompatible element-rich1047298uidsreleased by antigorite breakdown in deeply subducted mantle Earth and PlanetaryScience Letters 192 457ndash470
Scambelluri M Rampone E Piccardo G 2001b Fluid and element cycling insubducted serpentinite a trace-element study of the ErrondashTobbio high-pressureultrama1047297tes (Western Alps NW Italy) Journal of Petrology 42 55ndash67
Segadelli S2006 Lageologia nelpaesaggio le rupi o1047297olitiche in Val Taro e Val Ceno IlGeologo dellrsquoEmilia-Romagna 22 15ndash29 Nuova Serie
Settembre Blundo D Martin De Vidales JL Alietti A 1992 Corrensite-type clayminerals from Taro Valley (North Italy) Mineralogica et Petrographica Acta 35215ndash222
Sharp ZD Barnes JD 2004 Water-soluble chlorides in massive sea1047298oor serpenti-nites a source of chloride in subduction zones Earth and Planetary Science Letters226 243ndash254
Siever R Woodford N 1979 Dissolution kinetics and the weathering of ma1047297cminerals Geochimica et Cosmochimica Acta 43 717ndash724Singer A 1989 Palygorskite and sepiolite group minerals In Dixon JB Weed SB
(Eds) Minerals in Soil Environments Soil Science Society of America MadisonWisconsin pp 829ndash872
Siffert B Wey R 1962 Synthese d rsquoune sepiolite a temperature ordinarie Comptesrendus hebdomadaires des seacuteances de lAcadeacutemie des Sciences de Paris 2541460ndash1463
Środoń J 1999 Nature of mixed-layer clays and mechanisms of their formation andalteration Annual Reviews of Earth and Planetary Science 27 19ndash53
Stoessell RK 1988 25 degC and 1 atm dissolution experiments of sepiolite and keroliteGeochimica et Cosmochimica Acta 52 365ndash374
Stoessell RK Hay RL 1978 The geochemical origin of Sepiolite and keorilite atAmboseli Kenya Contributions to Mineralogy and Petrology 65 255ndash267
Suarez DL Šimůnek J 1996 Solute transport modeling under variably saturatedwater 1047298ow conditions In Lichtner PC Steefel CI Oelkers EH (Eds) ReactiveTransport in Porous Media Reviews in Mineralogy vol 34 Mineralogical Society of America Michigan Book Crafters Inc 229ndash268
Sumner DY Grotzinger JP 1996 Were kinetics of Archean calcium carbonate
precipitation related to oxygen concentration Geology 24 119ndash122Su C Suarez DL 2004 Boron release from weathering of illites serpentine shales
and illiticpalygorskitic soils Soil Science Society of America Journal 68 96 ndash105Toscani L Venturelli G Boschetti T 2001 Sul1047297de-free and sul1047297de-bearing waters in
the Northern Apennines Italy general features and waterndashrock interaction AquaticGeochemistry 7 195ndash216
Treves BE Harper GD 1994 Exposure of serpentinites on the ocean 1047298oor sequenceof faulting and hydrofracturing in the Northern Apennine ophicalcites O1047297oliti 19b435ndash466
Varadachari C Kudrat M Gosh K 1994 Evaluation of standard free energies of formation of clay minerals by an improved regression method Clays and ClayMinerals 42 298ndash307
Veniale F 1966 Sepiolite in sedimenti dellrsquoAppennino Vogherese (Pavia) Periodico diMineralogia 35 343ndash386
Venturelli G 2003 Acque minerali ed ambiente mdash Fondamenti di geochimica deiprocessi di bassa temperatura Pitagora Editrice Bologna
Venturelli G Contini S Bonazzi A Mangia A 1997 Weathering of ultrama1047297c rocksand element mobility at Mt Prinzera Northern Apennines Italy MineralogicalMagazine 61 765ndash778
Venturelli G Frey M 1977 Anchizone metamorphism in sedimentary sequences of the Northern Apennines Rendiconti Societagrave Mineralogica Italiana 33 109ndash123
Vieillard P 2002 A new method for the prediction of Gibbs free energies of formationofphyllosilicates (10 Aring and14 Aring)basedon theelectronegativityscale Claysand ClayMinerals 50 352ndash363
Vissers RLM Strating EHH 2001 Structures and microstructures in a thrust-relatedgreenschist facies tectonic melange Voltri Group (NW Italy) O1047297oliti 26 33ndash46
Viti C Mellini M 1997 Contrasting chemical compositions in associated lizardite andchrysotile in veins from Elba Italy European Journal of Mineralogy 9 585 ndash596
Viti C Mellini M 1998 Mesh textures and bastites in the Elba retrogradeserpentinites European Journal of Mineralogy 10 1341ndash1359
Wenner DB Taylor Jr HP 1974 Deuteriumatomic hydrogen and oxygen-18oxygen-16 studies of serpentinization of ultrama1047297c rocks Geochimica et CosmochimicaActa 38 1255ndash1286
Williams LA Parks GA Crerar DA 1985 Silica diagenesis I Solubility controls Journal of S edimentary Petrology 55 301ndash311
Wilson J Savage D Cuadros J Shibata M RagnarsdottirKV 2006 The effect of ironon montmorillonite stability (I) Background and thermodynamic considerations
Geochimica et Cosmochimica Acta 70 306ndash322WollastR MackenzieFTBricker OP1968 Experimental precipitation and genesis of sepiolite at earth-surface conditions American Mineralogist 531 1645ndash1662
Wrucke CT1996 Serpentine and carbonated-hosted asbestos deposits InEdwardAdu Bray (Eds) Preliminary compilation of descriptive geoenvironmental mineraldeposit models 95ndash831 US Geological Survey Open-File Report pp 39 ndash46
Xu T Apps JA Pruess K 2001 Analysis of mineral trapping for CO2 disposal in deepaquifers Lawrence Berkeley National Laboratory University of California
ZhangYDaweRA2000 In1047298uence ofMg2+ on thekineticsof calcite precipitation andcalcite crystal morphology Chemical Geology 163 129ndash138
Zhu C Anderson G 2002 Environmental applications of geochemical modelingCambridge University Press
91T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
references therein Venturelli et al 1997) Accessory phases areplagioclase picotite plusmn calciumndashsodic amphibole (Venturelli et al1997) Before modeling the crystal chemical formula of the mineralphases was calculated from data from Venturelli et al (1997) thendepending on the modal composition of the rock (Ernst and Piccardo1979) and the composition of the ideal solid solution of the minerals
the resulting stoichiometry of the solid reactant(s) was calculated andinserted in the input 1047297le of the PHREEQCI software (Table 3)Thermodynamic data of the primary mineral paragenesis are all wellknown and are included in the llnldat thermodynamic database of PHREEQCI
53 The solid reactant secondary paragenesis
Secondary minerals in the ultrama1047297tes of the TarondashCeno Valleysare serpentine (lizardite and rare chrysotile) smectite minerals (Fendash
Mg saponites) and Fe-oxi-hydroxides Other accessory minerals arechlorite mixed layer saponitendashchlorite (low-charge corrensite) talccalcite and dolomite (Beccaluva et al 1973 Dinelli et al 1997Venturelli et al 1997 Brigatti et al 1999 2000) In addition mixed
layer saponitendashtalc (aliettite) and vermiculitendashchlorite (high-charge
corrensite) are present (Alietti 1956 Settembre Blundo et al 1992Brigatti and Valdregrave 1996) Except for serpentine themodal abundanceof these phases is unknown In addition some of these minerals ndash
particularly interlayer clays ndash may have different origins (hydrothermalor weathering allogenicor authigenic)and different stabilitiesrelatedtokinetics and drainage 1047298ow rate For example chlorite weathering yields
vermiculite-bearing or saponite-bearing interlayers under good- andpoor-drainage regimes respectively (Bonifacio et al 1997 Środoń1999) In addition the thermodynamic data available for serpentineinclude information only on chrysotile which is less abundant thanlizardite in the serpentinites of the areas that we studied According toVenturelli et al (1997) the lizardite composition in the serpentinites of the Taro valley can be summarized as
which is similar to 1-T lizardite from the Monte Fico serpentinizedperidotite (Elba Island Northern Tyrrhenian Sea Italy Viti and Mellini1997)
Using the thermodynamic data of Wilson et al (2006) and
assuming an ideal solid solution between lizardite and berthierine
Fig 5 EhndashpH iron diagram of the spring waters coming from ultrama1047297c rocks of the TarondashCeno Valleys Bicarbonate and high-pH waters of the Voltri Group are also represented(dotted area andsmallblackdiamonds respectively) Thelowest Eh value of thehigh-pH waterswas calculated using theS(+VI)S(minus II)redox couple(Marini and Ottonello 2002)Thestability 1047297eld of mineral phases are in grey and those of aqueous species in white In diagrams A and B 6-line ferrihydrite-vermiculite and goethite were selected as subaerial solidphases respectively
83T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
we calculated ΔGf 0=minus399854 kJmol for a theoretical (09)(lizardite)+
(01)(berthierine) mineral (Table 2) The log K value of the stated solidsolution was calculated at different temperatures using PHREEQCIafter loading the thermodynamic data on the two end-members intothe database The log K results obtained for Al-lizardite were added tothe thermodynamic database of the solid reactant (Table 3) andcompared with data from the literature (Table 4)
54 Composition of the starting solution O 2 partial pressure and
minerals included in the model
The weighted average chemical composition of rainwater inBologna (Panettiere et al 2000) which is 95 km east of theinvestigated area was used as the starting solution in the modelFull chemical data on rainwater of the TarondashCeno Valleys are
unavailable
Total alkalinity was recalculated for log PCO2=minus35 bar and
T =11 degC which was the mean temperature of the springs sampledand for pe (or Eh) based on chemical data for the redox couple N(minusIII)N(+V) The values of pH and pe obtained were 566 and 452(Eh=025 V) respectively the former agrees with the median valuereported for rainwater in thearea (Venturelli et al1997) Because dataon Al Fe and Si in rainwater were lacking we calculated theconcentrations of these elements assuming logPCO2=minus30 bar andslight supersaturation relative to ferrihydrite-6 line and kaolinite Inthis way a hypothetical composition of water in equilibrium with thelocal soil was calculated (Table3) and adopted as the ldquoinitial waterrdquo formodeling
Evidence suggests that secondary minerals particularly Fe-bearingphases change the chemical composition of the water 1047298owingthrough the ultrama1047297tes of the TarondashCeno Valleys According to Neal
and Stanger (1983) and Fritz et al (1992) the production of CH4 and
Fig 6 Activity plotfor the MgOndashSiO2ndashAl2O3ndashFe2O3(FeO)ndashH2O systemat differentlog PO2 values Spring waters issuing from TarondashCeno Valleys (symbols as in Fig 2) andVoltri Groupultrama1047297tes are shown (grey area bicarbonate waters small black diamonds alkaline waters from Marini and Ottonello 2002) In plots A B and C the stability 1047297eld limits of vermiculite(Table 2) correspond to25 degCwhilethe amorphoussilica (am-SiO2) brucite serpentine (Sr1ndash4 in Table 4)and sepiolite(Sp1ndash2 in Table 4) phasescorrespondto 11 degCThepath calculatedfor interaction between waterand ultrama1047297c rockis also shown (dotted line)the number-letter combinations referto the appearanceand disappearance of minerals(see Table 5)
84 T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
H2 in thealkalinewaters1047298owing from theultrama1047297tes isrelatedto theconsumption of oxygen during the oxidation of iron forming thehematitendashmagnetite pair Therefore the activity of iron and theoxidation state of the water must be de1047297ned The analysis of watersamples from ultrama1047297tes were plotted in both the Pourbaix andactivity diagrams (Figs 5 and 6) where the calculated activity of Fe2+Mg2+ SiO2 SO4
2minus (Pourbaix) and log PO2 (activity diagrams) wereinvolved Iron concentration was quite variable in the TarondashCenosprings where the average content was 27 μ gL and the contentreached a maximum of 200 μ gL and in the Voltri springs where theaverage was 44 μ gL and the maximum was 948 μ gL (Marini andOttonello 2002) In both cases the variations are probably related to
the different mineralogies of the soils and the different solubilities of the Fe-rich phases like goethite and ferrihydrite Iron-bearingsaponites and montmorillonites are the prevailing clay minerals inthe area investigated probably because of the ubiquitous presence of iron oxy-hydroxide (Alietti et al 1976 G Venturelli pers comm) Forthis reason the model considers the involvement of pure Mg-smectiteto be negligible
In the Pourbaix diagram (Fig 5) the data from bicarbonate watersfall into the ferrihydrite-goethitendashvermiculite1047297elds and the data fromalkaline waters fall on the boundaries of the vermiculitesaponite andsaponitechlorite Since ferrihydrite is the 1047297rst precipitating Fe-phaseit was used rather than goethite in the modeling The diagram alsosuggests the transformation of ferrihydrite into magnetite during thechange from bicarbonate to alkaline composition but this process
requires a large amount of Fe2+ in solution (N2 mM Hansel et al2005) which is unlikely at this stage of the waterndashrock interactionThus the magnetite participating in the equilibrium is most likelyprimary in origin
The modeling and the activity diagrams (Figs 6 and 9) werecalculated using the average value PO2= 10minus26 or 10minus50 bar forbicarbonate and alkaline waters respectively These values arecompatible with those calculated by Marini and Ottonello (2002) forbicarbonate waters (PO2= 10minus40 bar) and for alkaline waters of theVoltri Group (10minus77 barbPO2b10minus42 bar) The lowest PO2 value whichwas used as a limit in the modeling was calculated from the S(+VI)S(minusII) redox couple Moreover the activity diagrams (Fig 6) indicatethat the data from the bicarbonate waters from the TarondashCeno Valleysfall between quartz whose log [SiO2] =minus433 is not shown because of
the slow precipitation kinetics under the PndashT conditions considered
here (Williams et al 1985) and amorphous silica boundary 1047297elds(Fig 6A) at the same manner bicarbonate waters with higher SiO2-activity from the Voltri Group fall on amorphous silica boundary(Fig 6B) In summary then the bicarbonate waters are supersaturatedin kaolinite ferrihydrite vermiculite and saponite Alkaline watersare also supersaturated with saponite (Fig 6AndashC) and they are alsoequilibrated with primary magnetite and under extreme conditionsprobably pyrite as well
55 Limitations of the model related to kinetics
Model calculations disregard the rates of mineral dissolution and
precipitation which becomes increasingly important as temperature
Table 3
Starting mineralogy (left side) and 1047298uid composition (right side) used in the modeling
Rock composition Fluid composition
Modal of the mineralsin the interacting rock
mol of the minerals usedin the modeling
meteoric water composition Initial 1047298uid composition used in the modeling
Spinel 003 Na 3523Eminus05 3523Eminus055 Spinel Chromite 001 S 3519Eminus05 3519Eminus05
Magnetite 0 Si ndash 1176Eminus06Kaolinite ndash 5258Eminus08
Ferrhydrite-6line ndash 5882Eminus07
The modal percentage of the minerals in a serpentinite from TarondashCeno Valleys (Ernst and Piccardo 1979) was re-calculated assuming an ideal solid solution between pure extremephasesand takingthe modal percentage tobe themass percentage Thelatter was used asthe solid reactant composition in thewaterndashrockinteraction modelPure solid phases werechosenfrom the llnldat thermodynamic database of the PHREEQCI softwareThe weighted average chemical compositionof the meteoric waterfrom Bologna (Panettiere et al 2000)was chosen as the initial 1047298uid composition in the modeling after its equilibration with kaolinite and ferrihydrite (see text for details)
Table 4
Summary of thermodynamic data for the sepiolite (Spn) and serpentine (Sr n) mineralsphases in italics do not refer to the mentioned hydrolysis reactions
ΔGf 0 log K Reference
(kJ molminus1) (at 25 degC)
Sepiolite (Spn) mdash Mg4Si6O15(OH)2middot6H2O+8H+=4Mg2++6SiO2+11H2Ominus925400 3004 Gunnarsson et al (2005)minus925133 3050andash3172 Christ et al (1973)
(Sp2) minus921986 3602andash3756 Wollast et al (1968)
Serpentine (Srn) mdash Mg3Si2O5(OH)4+ 6H
+
=3Mg
2+
+2SiO2+ 5H2ONi-lizardite minus 289324 1680 Mondesir and Decarreau (1987)
Al-lizardite2 minus411003 2544a Caruso and Chernosky (1979) Al-lizardite minus 339854 2786 this work (see text for details)
Mg-lizardite (Sr1) minus404025 3056 Wilson et al (2006)Mg-crysotile (Sr2) minus403702 3112 bthermocomV8R6Mg-lizardite minus402824 3340 Mondesir and Decarreau (1987)p-antigorite (Sr3) minus564239a 3479 Gunnarsson et al (2005)Mg-chrysotile (Sr4) minus406809 3669 Bruni et al (2002)Mg-chrysotileasbestos minus562505a 3888 Pfeifer (1977)Antigorite minus6612864 47719 bthermocomV8R6
The 1047297eld boundaries of the underlined phases shown in Fig 6 correspond to atemperature of 11 degC
a In the ΔG f deg column value calculated using the ΔG f
deg data of the elements from llnldat
and logK value in the reference in the LogK column value calculated using ΔG f deg
(mineral) in the reference and ΔG f deg data of the elements from llnldat
b Thermodynamic database of Lawrence Livermore National Laboratory included in
PHREEQCI and Geochemists Workbench softwares
85T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
decreases The waterndashrock interaction in the TarondashCeno basin iscomplicated by the presence of secondary mixed-layer clay mineralsthe origin of which is quite dif 1047297cult to model because the solidndashstatetransformations and weathering reactions are strictly related tohydrological and climatic conditions as described above An equili-
brium-based reaction path model is a valuable tool for assessing the
most probable reactions and for de1047297ning the product phases that canbe used in a kinetics-based model Since the kinetics of precipitationof Mg-clays at room temperature and pressure are poorly knownrelative reaction rate effects in the reaction path model can be mimedby omitting the stabler phase when a metastable mineral is expected
to precipitate Therefore the precipitation of MgndashFe3+
Abbreviations are as follows Fer-6line 6-line ferrihydrite kaol kaolinite verm vermiculite magn magnetite sap saponite serp serpentine clin clinochlore pyr pyrite calccalcite hydrom hydromagnesite bruc brucite enst enstatite The chromium decrease during the change in water facies from bicarbonate to alkaline was modeled withsupersaturation (precipitation) of eskolaite and thereafter with equilibrium with chromite (Boschetti 2003)- not bound dagger chlorite mineral from solution Dagger chlorite mineral added to the paragenesis
86 T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
montmorillonites vermiculite and saponites are predicted to occurrather than that of Mg-clays Moreover in order to simulate the slowprecipitation of calcite in depth precipitation was not activated atslight supersaturation (SIcalcite=001) during the initial stages of thewaterndashrock interaction in this manner saturation index of calcite wasfree to increase in the following steps and in alkaline waters
6 Discussion of the model results
61 Explanation of pre-modeling choices and post-modeling results
The results of the reaction path modeling are reported in Table 5and shown in Figs 6 and 9 The reaction paths model the dissolutionof a typical local serpentinite (mineralogically constrained as inTable 3) and the sequential precipitation of gibbsite kaoliniteferrihydrite vermiculite Fe2+ndashMg2+-saponite and a poorly-crystal-line serpentine( p-antigorite of Gunnarsson et al2005) The1047297rst stepinvolves the reaction of the ldquoinitial waterrdquo with serpentinite atconstant temperature (11 degC) PO2 (10minus26 bar) and PCO2 (10minus30 bar)this early stage corresponds to bicarbonate groundwater feeding theinvestigated spring system At the stage of primary magnetiteequilibrium and Fe2+ndashMg2+-saponite supersaturation the modelassumes the system to be closed to CO2 and to have a 1047297xed partialpressure of oxygen (PO2= 10minus40 bar) these conditions represent
when water switch from a bicarbonate to an alkaline compositionThe calculated amount of interacting serpentinite needed to obtainthe alkaline springs of the TarondashCeno Valleys (mean PCO2= 10minus7 bar)is approximately 00203 moles (see Table 5)
Bruni et al (2002) explained the origin of the alkaline waters of the Voltri Group as the result of the evolution of meteoricbicarbonate water following the precipitation of sepiolite withabundance similar to that of goethite This might be supported bythe observation that sepiolite or smectite precipitate at lowtemperature from solutions with low Al content when the pH is inthe range 75ndash95 or N95 respectively (Siffert and Wey 1962 Jonesand Galan 1988) but sepiolites form from the weathering of serpentines in arid climate regions whereas sepiolites are rare inhumid climate regions where serpentinite changes to Fe-smectite
(Singer 1989 Wrucke 1996) In the ophiolite rocks of the Northern
Apennine sepiolite is found only sporadically in veins (Giuseppetti1953 Veniale 1966) and is unknown in the ophiolites of the Tarondash
Ceno Valleys whereas it is thought that sepiolite of the Voltri Groupprecipitated from metasomatic 1047298uids during development of shearzones (Crispini and Capponi 2002)
As described above our model explains the transition frombicarbonate water to alkaline water (UM15 and PR10 samples) asthe result of low-temperature precipitation of a poor crystalline
serpentine with the solubility product of p-antigorite (see Sr3 inTable 4 and Fig 6) The model indicates that the bicarbonate waters of the Voltri Group with the highest [Mg2+][H+]2 activity ratios similarlyevolve towards alkaline waters (Fig 6B) The logK value of thesynthetic low temperature serpentine adopted in this model and its
Fig 7 Saturation indices of chrysotile and p-antigorite (Sr2 and Sr3 thermodynamic
data respectively in Table 4) for bicarbonate and alkaline waters coming fromultrama1047297c rocks of the TarondashCeno Valleys as a function of SiMg molar ratioThermodynamic data on chrysotile and p-antigorite are from the thermocomV8R6
database and Gunnarsson et al (2005) respectively
Fig 8 Activityplot forthe MgOndashSiO2ndashAl2O3ndashFe2O3(FeO)ndashH2O system analyzed forSiO2-phases and aqueous species at different log PO2 and log[aAl3+(aH+)3] values Springwaters coming from the TarondashCeno Valleys (symbols as in Fig 2) and Voltri Groupultrama1047297tes (symbols as in Fig 5) are shown The greyarea refers to bicarbonate watersfrom theVoltri Group (Mariniand Ottonello2002Bruniet al 2002) Thesepiolite 1047297eldis very narrow or hidden by saponite-Mg (1047297eld limits shown as continuous lines)saponite-Fe3+ndashMg (1047297eldlimits shown as dottedlines) or vermiculite(1047297eldlimits shownas dottedndashdashed lines) The talc stability 1047297eld and boundaries with saponite-Mg
montmorillonite-Mg and cronstedite-Mg are shown for comparison (dashed lines)
87T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
relative uncertainty of plusmn2 (see Fig 8 in Gunnarsson et al 2005) arecomparable to the logK variation reported for other low-temperatureMg-serpentines (Table 4) that is within the lines Sr1ndash4 for serpentineminerals where samples varying from bicarbonate composition toalkaline composition in the ultrama1047297c aquifer are grouped (Fig 6)Fig 7 shows a plot of calculated serpentine saturation index as a
function of the SiMg molar ratio for the water samples from theultrama1047297tes of the TarondashCeno ValleysThe SI andSiMgvalues of pureMg-serpentine (SI= 0 SiMg=067) provide a useful index forseparating bicarbonate waters (SIb0 SiMgb067) from alkalinewaters (SIN0 SiMgN067) and the distribution of data points isconsistent with the idea that bicarbonate waters approach saturationwith serpentine until they become supersaturated as alkaline waters
Based on the Mgvs SiO2(aq) activitydiagram and speci1047297cratiosoflog[aAl3+(aH+)3] Birsoy (2002) analyzed the behavior of sepiolite duringultrama1047297te weathering The diagram produces a very narrow stability1047297eldfor sepiolite(Fig 8) when it is programmedwith suitable conditionsfor the bicarbonate waters (log PO2=minus25 bar log [aAl3+(aH+)3]= 65) andthe alkaline waters (log PO2=minus50 bar log [aAl3+(aH+)3]= 45) sampled inthe present study This stability 1047297eld grows larger when the absence of
Mg-smectites and talc is assumed but this is an unrealistic conditionbecause (i) all these minerals occur in the rocks within the areainvestigated and (ii) at high [Mg2+][H+]2 activity ratios the 1047297eld of sepiolite is smaller than that of subordinate compared to those poorly-crystalline serpentine (p-antigorite) and clinochlore
Among the alkaline springs of the Voltri Group the waters withlow log [aCa2+(aH+)2] values (Fig 9) show patterns compatible withchlorite interaction (step 5c in Fig6D)The alkaline waters whose datafall near the clinochlore-14Aring 1047297eld (Fig 6D) cannot be explained bysepiolite precipitation because the sepiolite line does not cross thechlorite 1047297elds Thus the peculiar composition of some alkaline watersof the Voltri Group is probably due to interactions between water andprimary chlorites of the ultrama1047297c rocks (optional step 5c in Table 5cb) This is supported by the occurrence sometimes in large amounts
of metasomatic Mgndash(Fe)-chlorite (Scambelluri and Rampone 1999
Vissers and Strating 2001) in the Ligurian area where the relativeabundance of chlorite can reach 50 in veins in high strainserpentinite mylonites of the ErrondashTobbio ultrama1047297tes (Scambelluriet al 2001b) Otherwise in a chlorite-free primary paragenesis thesolution evolving on the Sr3 serpentine line reaches equilibrium withclinochlore but probably the latter does not precipitate due to theslow kinetics of precipitation under these conditions Thus thedissolution of additional reacting rock in the iterative weathering
process supplies Mg
2+
to the solution thereby favoring the Sr3 to Sr4poorly crystalline serpentine transformation In fact all the PR10samples and some of the alkaline springs from the Voltri Group showincreasing [Mg2+][H+]2 activity ratio and low variation in silicaactivity (Fig 6D) starting from the invariant triple point 4a (Sr3-serpentine Fe2+ndashMg-saponite chlinoclore-14 Aring saturation) andextending towards the brucite line This path can be explained asthe result of the transformation of poorly crystalline serpentine fromSr3 to Sr4 the latter is more stable at higher [Mg2+][H+]2 activityratios Finally at very low PO2 values vermiculite disappears and thestability 1047297elds of chlorite and Fe2+ndashMg-saponite become narrowerwhereas those of Fe2+-rich phases (eg Fe2+-saponite) become widerThe data for the alkaline waters with the highest log [aMg2+(aH+)2]from the Voltri Group fall near the stability 1047297eld of peridotite primaryminerals like forsterite and enstatite (Fig 6D) therefore the last stepsof theevolution involve theequilibration of thesolutionwithenstatite(steps 8a and 8c in Table 5ab)
62 Ca-rich alkaline waters
In preliminary modeling of the evolution of the TarondashCeno springwaters Boschetti (2003) concluded that meteoric water initiallysaturated in Fe3+ndashMg2+-chlorite Fe3+ndashMg2+-saponite and pyrite (andor goethite) and then interacted with the ultrama1047297te mineralassemblage eventually evolved into the alkaline waters In themodel of Boschetti (2003) chlorite saponite and calcite were added
Fig 9 Activity plot for the MgOndashCaOndash(CO2)ndashH2O system with phase limits calculated ata temperature of 11 degCSpring waterscoming from theTarondashCeno Valleys (symbols as inFig2) andVoltri Groupultrama1047297tes(symbolsas in Fig5) areshown Calculated path forwaterndashultrama1047297c rock interaction is shown (dotted line) with the number ndashlettercombinations referring to the appearance or disappearance of minerals (see Table 5)The grey area refers to bicarbonate waters from the Voltri Group (Marini and Ottonello2002)
Fig 10 Plots of (A) calcite saturation indexes and (B) PCO2 as a function of Clminus ascalculated by PHREEQCI software for waters from the TarondashCeno Valleys Alkalinesprings for the Voltri Group are also shown (small black diamonds Marini andOttonello 2002) The grey shading is deeper(A) in the saturation1047297eld(SI= 00plusmn05) and
(B) where the lowest log PCO2 value is approached
88 T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
to the solution in order to achieve a high [Ca2+][H+]2 activity ratio andsupersaturation in these phases However in light of the work of Clarket al (1992) and Bruni et al (2002) we believe that the calcitesupersaturation in alkaline waters (Fig 9 and 10) which in our modelis achieved by preventing the precipitation of calcite in the 1047297rst stepsof the reaction path (Table 5a and b) depends on the differencebetween CO2 partial pressureat thegivendepth(PCO2le10minus10bar) andat the surface (PCO2cong10minus35) Waters rising to the surface are far from
equilibrium with the atmosphere where the PCO2 is 4times10
6
timesgreater than the value at depth thus the water rapidly take up CO2which is promptly converted to CO3
2minus because of the high pH Thisleads to supersaturation with calcite through the following hydro-xylation reaction
As a result large amounts of travertine are generated in the VoltriGroup (Marini and Ottonello 2002) and carbonate concretionsknown as devilrsquo s coins are formed at Mt Prinzera in Taro valleyBelow the surface the SIcalcite of the alkaline waters is expected to belower than what the model estimates Under conditions of lowcarbonate alkalinity low CO2 and O2 partial pressure and high pHcalcite nucleation may be inhibited and the kinetics of precipitationreduced by the high concentrations of dissolved divalent cations suchas Mg2+ Fe2+ and Mn2+ (Sumner and Grotzinger 1996 Davis et al2000 Zhang and Dawe 2000 de Leeuw 2002) The fact that severalalkaline water samples fall within or near the calcitendashdolomiteequilibrium (Fig 9) may indicate magnesium calcite or protodolomitenucleation in those extreme subsurface conditions However possibleinteraction between some calcium-rich waters of the Voltri Group andCa-rich rocks such as ophicarbonates or rodingites cannot beexcluded these rocks are widespread in the Ligurian Apennineserpentinites (Treves and Harper 1994)
63 Sepiolite vs serpentine nucleation under weathering conditions
constraints on the kinetics and equilibrium-based approaches
The kinetics of precipitation of poorly-crystalline serpentine and
sepiolite cannot be compared quantitatively because rate constantsare available only for well-ordered crystalline phases (eg see Palandriand Kharaka 2004 for crystalline serpentine) Sepiolite precipitationis often invoked to explain magnesium scavenging from solutions butthe lack of kinetic data for this mineral leads to overly simplisticreaction transport models These models are forced to use kinetic datafor phyllosilicates such as kaolinite or muscovite (eg Xu et al 2001Bryan 2005) Moreover the low-temperature thermodynamic dataeven if they are fairly uncertain show that sepiolite solubilityincreases with decreasing crystallinity (Table 4 Stoessell 1988) Infact crystalline sepiolite requires a long time to reach equilibriumwith solution which imposes a slow control on thewater composition(eg Stoessell and Hay 1978) Consequently in modeling based onreaction transport that simulates freshly precipitated or poorly
crystalline sepiolite formation the logK value of Wollast et al (1968)is preferable to the lower values (Suarez and Šimůnek 1996) Thispreference agrees with the approach of Mariner (2003) and Jove Colon(2006) who adopt a logK (25 degC) value for sepiolite of 3044 It is alsosimilar to that reported by Wollast et al (1968) and it is six logK unitshigherthan the valueused in thermocomV8R6 dataset which is oftenused (perhaps improperly) to model the supergenic nucleation of sepiolite In the stability diagram in Fig 6 where the Sp1 and Sp2 linesrepresent the stability boundaries of crystalline and poorly-crystallinesepiolite respectively the evolution of TarondashCeno waters frombicarbonate to alkaline (Figs 6A and C) follows the stability 1047297eldboundary of the poorly-crystalline serpentine (between Sr2ndashSr3 lines)rather than that of the poorly-crystalline sepiolite (Sp2) as postulatedin themodel The lines for Sp2 Sr3 and amorphoussilica intersect at a
common point (Fig 6D) which is consistent with the nucleation of
sepiolite that typically occurs in solutions that are saturated orsupersaturated with amorphous silica and (Mg)-carbonate (egStoessell and Hay 1978 Birsoy 2002) Some bicarbonate watersfrom the Voltri Group have a silica activity near the amorphous silicaboundary (Fig 6B) but only a few alkaline springs of this group fall onthe poorly-crystalline sepiolite boundary line (Sp2 Fig 6D) Both of these features are inconsistent with the possibility that sepioliteprecipitates from supergenic fresh water in the Northern Apennine
Moreover these two results suggest that values for logK(sepiolite)b
3044 (see Table 4) should be used to model the dissolutionprecipitation of the well-crystallized phase that is the most commonform for the metasomatic sepiolite found in the investigated area
7 Conclusions
The waters issuing from ultrama1047297tes in the TarondashCeno Valleysoriginate from low-temperature interactions between meteoric waterand serpentinite rocks In this report the waterndashrock interaction hasbeen modeled using reaction paths in progress mode During their earlyinteraction with soil and rock minerals bicarbonate waters super-saturated in kaolinite ferrihydrite vermiculite and Fe2+ndashMg2+-saponitewhere theorigin of the secondary iron saponite as Fe2+-bearing smectiteagrees with the hypothesis of Guumlven (1988) Then while remainingsupersaturated with saponite the bicarbonate waters evolve into analkaline facies when poorly-crystalline serpentine precipitates Duringrock weathering the Al-bearing crystalline serpentine of oceanic originprobably promotes nucleation of the low-temperature serpentineamorphous phase and supplies Al3+ to the solution for saponiteprecipitation This interpretation is consistent with the existence of Al-poor supergenic serpentine (Neal and Stanger 1985) and the partiallymeteoric OH-isotopic imprint observed in other ultrama1047297c systems
Dif 1047297culties and complexities in the modeling arise fromuncertainties in the primary mineralogy of some local ophiolitesand from the presence of different lithotypes (rodingites ophi-carbonates gabbros) which can signi1047297cantly affect the composi-tion of water samples This can be seen in the example of the Candash
OH springs issuing from the Voltri Group (PO2 and PCO2 of 10minus70
and 10minus10 bar respectively Bruni et al 2002 Marini and
Ottonello 2002) These springs appear to be more evolved thanthe TarondashCeno alkaline waters as a result of their particularly longinteractions with serpentinites until equilibrium with enstatite isreached or with chlorite-bearing serpentinite The use of anupdated thermodynamic database in our model avoided problemsdue to mineralogical heterogeneity in the mixed-layer claysassociated with ophiolitic rocks this database accurately re1047298ectedthe chemical composition of the single phases Using this modelthe vermiculitendashchlorite and saponitendashchlorite associations seen inthe activity diagrams simulate respectively the presence of highand low charge corrensite observed in the 1047297eld The accuracy of ourmodel can be tested by comparing estimates of chlorine composi-tion resulting from our model and from experimental methods The
mean Cl concentration of the bicarbonate water is 41plusmn16 mgLtaking into account a Cl concentration of sim2 mgL in rainwater inthe area (cf Panettiere et al 2000 Venturelli 2003) 2 mgL of Clin spring water are estimated to come from serpentinite This valueagrees with the model which predicts that the switch frombicarbonate water to alkaline water requires the congruentdissolution of 00203 mol of serpentinite containing 100 ppm of soluble Cl thereby yielding water containing 2 ppm of Cl
In addition we think that weathering of the accessory albiteplagioclase is responsible for Na enrichment of the waters whereas Clenrichment in the alkaline waters is predicted to result from theleaching of NaCl 1047298uid inclusions Moreover the TarondashCeno alkalinesprings show the highest BCl ratios comparable with those of thehost metamorphic minerals (Scambelluri et al 2004) In contrast the
lower boron content and BCl ratio of the alkaline springs of the Voltri
89T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
Group are unexpected features and probably re1047298ect waterndashrockinteractions over a longer time period than in the TarondashCeno ValleysThe boron content in the Northern Apennine alkaline springs may becontrolled by several factors such as (i) the different histories of VoltriGroup and EL ultrama1047297tes (ii) the probable contribution of boronfrom NaCl 1047298uid inclusions (iii) the sinking of boron by travertineprecipitation (Hobbs and Reardon1999) brucite (Rhoades et al 1970)and smectite and (iv) decreases in boron release from minerals due to
increases in pH (Su and Suarez 2004) Acknowledgements
This research was carried out with the 1047297nancial support from theldquoYoung Researchers Projectrdquo (2001ndash2002) of the Italian Ministry of Universities Scienti1047297c Research and Technology (MURST) We thankGianniCortecci (Institute of Geosciences and EarthResources IGG-CNRPisa Italy) and Giampiero Venturelli (Earth Sciences Department Univer-sity of Parma Italy) for useful suggestions and critical review of themanuscript The authors are grateful to Enrico Dinelli (InterdepartmentResearch Centre for Environmental Sciences Alma Mater StudiorumUniversity of Bologna Ravenna Italy) for granting permission to use theGeochemistrsquos WorkbenchsoftwareTheconstructive reviews of theEditor
J Fein and of two anonymous Referees have signi1047297cantly improved the
manuscript
Appendix A Supplementary data
Supplementary data associated with this article can be found inthe online version at doi101016jchemgeo200808017
References
Abbate E Bortolotti V Passerini P 1980 Olistostromes and olistoliths In Sestini G(Ed) Development of the Northern Apennines geosyncline Sedimentary Geology4 pp 521ndash558
Alietti A 1956 Il minerale a strati misti saponite-talco di monte Chiaro (val di TaroAppen Emil) Rendiconti dellAccademia Nazionale dei Lincei 21 201 ndash207
Alietti A De Angelis G Montanari G Sgarlata F Tirelli G 1976 Il ferro nelle argilleappenniniche Periodico di Mineralogia 45 51ndash64
Anselmi B Mellini M Viti C 2000 Chlorine in the Elba Monti Livornesi and Murloserpentines Evidence for seandashwater interaction European Journal of Mineralogy12137ndash146
Apha-Awwa-Wef 1995 Standard Methods for the Examination of Water and Waste-water 19th Edition American Public Health Association American Water WorksAssociation Water Environment Federation
Barnes I LaMarche Jr VC Himmelberg GR 1967 Geochemical evidence of present-day serpentinization Science 156 830ndash832
Barnes I ONeil JR 1969 The relationship between 1047298uids in some fresh alpine-typeultrama1047297cs and possible modern serpentinization Western United StatesGeological Society of America Bulletin 80 1947ndash1960
Barnes I ONeil JRTrescases JJ1978 Present dayserpentinizationin New CaledoniaOman and Yugoslavia Geochimica et Cosmochimica Acta 42 144ndash145
Barnes I Rapp JB ONeil JR Sheppard RA Gude III AJ 1972 Metamorphicassemblages and the direction of 1047298ow of metamorphic 1047298uids in four instances of serpentinization Contributions to Mineralogy and Petrology 35 263ndash276
Barnes JDSelverstone J Sharp ZD 2006Chlorine isotope chemistry of serpentinitesfrom Elba Italy as an indicator of 1047298uid source and subsequent tectonic historyGeochemistry Geophysics Geosystems 7 doi1010292006GC001296
Beccaluva L Emiliani F Venturelli G Zerbi M 1973 Ca Fe Mg Mn Cr Ni Codistribution in some ultrama1047297c rocks outcropping in the Northern Apennines withsome geological remarks Ateneo Parmense Acta Naturalia 9 69ndash98
Beccaluva L Macciotta G Piccardo GB Zeda O 1984 Petrology of lherzolitic rocksfrom the Northern Apennine ophiolites Lithos 17 299 ndash316
BeccaluvaL Venturelli G ZanzucchiG 1975 Dati geologici e geochimici sui basalti diassociazione o1047297olitica dellAppennino ligure-emiliano Ateneo Parmense ActaNaturalia 11 789ndash802
BethkeCM 2002 The Geochemists Workbench A Userrsquos Guide to RxnAct2Tact ReactandGtplot Release40 HydrogeologyProgram Universityof IllinoisUrbanaIL 236pp
Birsoy R 2002Formation of sepiolitendashpalygorskite and related minerals from solutionClays and Clay Minerals 50 736ndash745
Bonatti E Lawrence JR Morandi N 1984 Serpentinization of oceanic peridotitestemperature dependence of mineralogy and boron content Earth and PlanetaryScience Letters 70 88ndash94
Bonifacio E Zanini E Boero V Franchini-Angela M 1997 Pedogenesis in a soilcatena on serpentinite in north-western Italy Geoderma 75 33ndash51
Boschetti T 2003 Studio geochimico e geochimico-isotopico di acque a composizione
estrema e termali dellrsquoAppennino Settentrionale Ph D Thesis Parma Univ
Brigatti MF Franchini G Lugli C Medici L Poppi L Turci E 2000 Interactionsbetween aqueous chromium solutions and layer silicatesApplied Geochemistry 151307ndash1316
Brigatti MF Lugli C Poppi L Venturelli G 1999 Iron-rich saponite dissolutionreactions and Cr uptake Clay Minerals 34 637ndash645
Brigatti MF Valdregrave G 1996 Mixed-layer chloritesmectite (corrensite) in ophioliticrock alteration products Neues Jahrbuch fuer Mineralogie Monatshefte 1 37 ndash47
Bruni J Canepa M Chiodini G Cioni R Cipolli F Longinelli A Marini L OttonelloG Vetuschi Zuccolini M 2002 Irreversible waterndashrock mass transfer accompany-ing the generation of the neutral MgndashHCO3 and high-pH CandashOH spring waters of the Genova province Italy Applied Geochemistry 17 455ndash474
Bryan CR 2005 Drift-Scale THC Seepage Model MDL-NBS-HS-000001 REV 04 YuccaMountain Project Las Vegas Nevada (US)Caruso LJ Chernosky JV 1979 The stability of lizardite Canadian Mineralogist 17
757ndash769Clark ID Fontes JC Fritz P 1992 Stable isotope disequilibria in travertine from high
pH waters Laboratory investigations and 1047297eld observations from Oman Geochi-mica et Cosmochimica Acta 56 2041ndash2050
Cortesogno L 1980 Il metamor1047297smo giurassico nelle o1047297oliti dellAppenninoSettentrionale due distinti cicli metamor1047297ci in ambiente oceanico O1047297oliti 5 5ndash18
Christ CLHostetler PB Siebert RM1973 Studies in the system MgO-SiO2-CO2-H2OIII American Journal of Science 273 65ndash83
Cortesogno L Lucchetti G 1982 Il metamor1047297smo oceanico nei gabbri o1047297oliticidellAppennino Ligure Aspetti mineralogici e petrogenetici Rendiconti dellaSocietagrave Mineralogica Italiana 38 561ndash579
Crispini L Capponi G 2002 Control of metasomatic alteration on the development of shear zone in ultrama1047297c rocks a case study from the Voltri Massif (Ligurian Alps)Architettura interna e cinematica delle zone di taglio Gruppo Italiano di GeologiaStrutturale mdash Riunione Annuale 2002 Pisa Italy 11ndash12 giugno 2002 XIV ndashXV
CrissRE 1999 Principlesof stableisotope distributionOxford University PressNew York
Davis KJ Dove PM De Yoreo JJ 2000 The role of Mg 2+ as an impurity in calcitegrowth Science 290 1134ndash1137
de Leeuw NH 2002 Moleculardynamics simulations of the growth inhibiting effect of Fe2+ Mg2+ Cd2+ and Sr2+ on calcite crystal growth Journal of Physical Chemistry B106 5241ndash5249
De Nardo MT Segadelli S Vescovi P 2007 Studio pilota per la caratterizzazionegeologicadellesorgentinellazona delM Nero (altaVal Cenoe alta ValNuremdash provincedi Parma e Piacenza) Il Geologo dellEmilia-Romagna 25 Nuova Serie pp 5 ndash21
Di Giulio A Geddo G 1990 Studio petrogra1047297co delle Arenarie di Casanova (alta valTrebbia Appennino settentrionale) Atti Ticinesi di Scienze della Terra 33 243ndash254
Dinelli E Lombini A Simoni A Ferrari C 1997 Heavy metals in serpentinite soils of selected outcrops of Piacenza and Parma provinces (Northern Apennines Italy)Mineralogica et Petrographica Acta 40 241ndash255
Epstein S Mayeda T 1953 Variation of 18O content of waters from natural sourcesGeochimica et Cosmochimica Acta 4 213ndash224
Ernst WG Piccardo GB 1979 Petrogenesis of some Ligurian peridotites I Mineraland bulk-rock chemistry Geochimica et Cosmochimica Acta 43 219ndash237
Feder F Trolard F Klingelhofer G Bourrieacute G 2005 In situ Moumlssbauer spectroscopy
evidence for green rust (fougerite) in a gleysol and its mineralogical transforma-tions with time and depth Geochimica et Cosmochimica Acta 69 4463ndash4483Fritz PClark IDFontes JCWhiticar MJFaberE 1992 Deuterium and
13C evidence
for low temperature production of hydrogen and methane in a highly alkalinegroundwater environment in Oman In Kharaka TK Maest AS (Eds) Proceed-ings of the 7th International Symposium on WaterndashRock Interaction BalkemaRotterdam pp 793ndash796
Giammetti F 1968 Le o1047297oliti di Roccaprebalza Costa della Guardia e Groppo di Gorro(Appennino Parmense) Ateneo Parmense Acta Naturalia 4 107ndash142
Giuseppetti G1953 Roccee minerali dellaformazioneo1047297olitica di Volpedo Rendicontidella Societagrave Mineralogica Italiana 9 85ndash134
Gran G 1952 Determination of equivalence point in the potentiometric titrationAnalyst 77 661ndash671
Gunnarsson I Arnorsson S Jakobsson S 2005 Precipitation of poorly crystalline antigoriteunder hydrothermal conditions Geochimica et Cosmochimica Acta 69 2813ndash2828
Guumlven N 1988 Smetites In Bailey SW (Ed) Reviews in Mineralogy 19 Hydrousphyllosilicates (exclusive of micas) Mineralogical Society of America MichiganBook Crafters Inc 497ndash559
Hansel CM Benner SG Fendorf S 2005 Competing Fe(II)-induced mineralization
pathways of ferrihydrite Environmental Science and Technology 39 7147ndash7153HobbsMYReardonEJ 1999 Effect of pH on boronco-precipitation by calcite further
evidence for non-equilibrium partitioning of trace elements Geochimica etCosmochimica Acta 63 1013ndash1021
Jones BF Galan E 1988 Sepiolite and palygorskite In Bailey SW (Ed) HydrousPhyllosilicates (exclusive of Micas) Reviews in Mineralogy vol 19 MineralogicalSociety of America Michigan Book Crafters Inc 631ndash674
Jove Colon CF 2006 Technical work plan for Thermodynamic database for chemicalmodelling TWP-MGR-PA-000039 REV 01 Yucca Mountain Project Las VegasNevada (US)
Kohls DW Rodda Jr JL 1967 Iowaite a new hydrous magnesium hydroxidendashferricoxychloride from the Precambrian of Iowa American Mineralogist 52 1261 ndash1271
Kyser TK OHanley DS Wicks FJ 1999 The origin of 1047298uids associated withserpentinization processes evidence from stable-isotope compositions CanadianMineralogist 37 223ndash237
Lapham DM 1961 New data on deweylite American Mineralogist 46 168ndash188Lee BD SearsSKGraham RC Amrhein CValiH 2003 Secondarymineral genesis
from chlorite and serpentine in an ultrama1047297c soil toposequence Soil Science
Society of America Journal 671309ndash1317
90 T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
Longinelli A Selmo E 2003 Isotopic composition of precipitation in Italy a 1047297rstoverall map Journal of Hydrology 270 75ndash88
Majzlan J Navrotsky A Schwertmann U 2004 Thermodynamics of iron oxides PartIIIEnthalpiesof formation and stability of ferrihydrite (simFe(OH)3) schwertmannite(simFeO(OH)34(SO4)18) and ɛ-Fe2O3 Geochimica et Cosmochimica Acta 681049ndash1059
Mariner P 2003 In-Drift PrecipitatesSalts Model ANL-EBS-MD-000045 REV 01 YuccaMountain Project Las Vegas Nevada (US)
Marini L Ottonello G 2002 Atlante degli acquiferi della Liguria Volume III Le acquedeicomplessio1047297olitici (bacini ArrestraBranega Cassinelle Cerusa ErroGorzenteLeira Lemme Lerone Orba Piota Polcevera Rumaro Sansobbia Stura Teiro
Varenna Visone) Pacini PisaMellini L Zanazzi PF 1987 Crystal structures of lizardite-l T and lizardite-2H1 fromColi Italy American Mineralogist 72 943ndash948
Mezzadri G 1964 Petrogra1047297a delle arenarie di Ostia Rendiconti Societagrave MineralogicaItaliana 20 192ndash228
Mondesir H Decarreau A 1987 Synthesegrave entre 25 et 200 degC de lizardites Ni ndashMgMesure des coef 1047297cients de partage solide-solution aqueuse pour le couple Ni ndashMgdans des lizardites Bulletin de Mineralogie 110 409ndash426
Neal C Shand P 2002 Spring and surface water quality of the Cyprus ophiolitesHydrology and Earth System Sciences 6 797ndash817
Neal C Stanger G 1983 Hydrogen generation from mantle source rocks in OmanEarth and Planetary Science Letters 66 315ndash320
Neal C Stanger G 1984 Calcium and magnesium hydroxide precipitation fromalkaline groundwaters in Oman and their signi1047297cance to the process of serpentinization Mineralogical Magazine 48 237ndash241
Neal C Stanger G 1985 Past and present serpentinisation of ultrama1047297c rocks anexample from Semail ophiolite nappe of Northern Oman In Drever JI (Ed) TheChemistry of Weathering D Reidel Publishing Company pp 249ndash275
Nesbitt HW Bricker OP 1978 Low temperature alteration processes affecting
et Cosmochimica Acta 41 1835ndash1841OrsquoHanley DS 1996 Serpentinites records of tectonic and petrological history Oxford
University PressOrberger B Friedrich G Woermann E 1990 The distribution of halogens and carbon
in Pge-bearing ultrama1047297cs of the Acoje Ophiolite Block Zambales Philippines Journal of Geochemical Exploration 37 147ndash169
Ottonello G Joron JL Piccardo GB 1984 Rare Earth and 3d transition elementsgeochemistry of peridotitic rocks II Ligurian peridotites and associated basalts Journal of Petrology 25 373ndash393
Palandri JL Kharaka YK 2004 A compilation of rate parameters of water-mineralinteraction kinetics for application to geochemical modeling Open File Report2004-1068 US Geological Survey
Panettiere P Cortecci G Dinelli E Bencini A Guidi M 2000 Chemistry and sulfurisotopic composition of precipitation at Bologna Italy Applied Geochemistry 151455ndash1467
Parkhurst DL Appelo CAJ 1999 Userrsquos guide to PHREEQC (version 2) mdash a computerprogram for speciation batch-reaction one-dimensional transport and inversegeochemical calculation US Department of the Interior US Geological SurveyWashington DC
Pfeifer HR 1977 A model for 1047298uids in metamorphosed ultrama1047297c rocks Observationsat surface and subsurface conditions (high pH spring waters) SchweizerischeMineralogische und Petrographische Mitteilungen 57 361ndash396
PlummerLN1984Geochemical modeling A comparisonof forwardand inverse methodsIn Hitchon B Wallick EI (Eds) Proceedings First CanadianAmerican Conference onHydrogeology-Practical Applications of Ground Water Geochemistry Banff AlbertaCanada mdash Worthington Ohio National Water Well Association pp 149ndash177
Rampone E Hofmann AW Piccardo GB Vannucci R Bottazzi P Ottolini L 1995Petrology mineral and isotope geochemistry of the External Ligurides peridotites(Northern Apennines Italy) Journal of Petrology 36 81ndash105
Rampone E Hofmann AW Piccardo GB Vannucci R Bottazzi P Ottolini L 1996Trace element and isotope geochemistry of depleted peridotites from an N-MORBtype ophiolite (Internal Liguride N Italy) Contributions to Mineralogy andPetrology 123 61ndash76
Rhoades JD Ingvalson RD HatcherJT1970Adsorption of boronby ferromagnesianminerals and magnesium hydroxide Soil Science Society of America Journal 34
938ndash941Saini-EidukatB KuchaH Keppler H1994Hibbingiteγ-Fe2(OH)3Clanewmineralfromthe Duluth Complex Minnesota with implications for the oxidation of Fe-bearingcompounds and the transport of metals American Mineralogist 79 555ndash561
Sanford RF 1981 Mineralogical and chemical effects of hydration reactions andapplications to serpentinization American Mineralogist 66 290ndash297
Scambelluri M Muumlntener O Ottolini L Pettke TT Vannucci R 2004 The fate of BCl and Li in the subducted oceanic mantle and in the antigorite breakdown 1047298uidsEarth and Planetary Science Letters 222 217ndash234
Scambelluri M Rampone E 1999 Mg-metasomatism of oceanic gabbros and itscontrol on Ti-clinohumite formation during eclogitization Contributions toMineralogy and Petrology 135 1ndash17
Scambelluri M Bottazzi P Trommsdorff V Vannucci R Hermann J Gogravemez-Pugnaire MT Logravepez-Sagravenchez Vizcaino V 2001a Incompatible element-rich1047298uidsreleased by antigorite breakdown in deeply subducted mantle Earth and PlanetaryScience Letters 192 457ndash470
Scambelluri M Rampone E Piccardo G 2001b Fluid and element cycling insubducted serpentinite a trace-element study of the ErrondashTobbio high-pressureultrama1047297tes (Western Alps NW Italy) Journal of Petrology 42 55ndash67
Segadelli S2006 Lageologia nelpaesaggio le rupi o1047297olitiche in Val Taro e Val Ceno IlGeologo dellrsquoEmilia-Romagna 22 15ndash29 Nuova Serie
Settembre Blundo D Martin De Vidales JL Alietti A 1992 Corrensite-type clayminerals from Taro Valley (North Italy) Mineralogica et Petrographica Acta 35215ndash222
Sharp ZD Barnes JD 2004 Water-soluble chlorides in massive sea1047298oor serpenti-nites a source of chloride in subduction zones Earth and Planetary Science Letters226 243ndash254
Siever R Woodford N 1979 Dissolution kinetics and the weathering of ma1047297cminerals Geochimica et Cosmochimica Acta 43 717ndash724Singer A 1989 Palygorskite and sepiolite group minerals In Dixon JB Weed SB
(Eds) Minerals in Soil Environments Soil Science Society of America MadisonWisconsin pp 829ndash872
Siffert B Wey R 1962 Synthese d rsquoune sepiolite a temperature ordinarie Comptesrendus hebdomadaires des seacuteances de lAcadeacutemie des Sciences de Paris 2541460ndash1463
Środoń J 1999 Nature of mixed-layer clays and mechanisms of their formation andalteration Annual Reviews of Earth and Planetary Science 27 19ndash53
Stoessell RK 1988 25 degC and 1 atm dissolution experiments of sepiolite and keroliteGeochimica et Cosmochimica Acta 52 365ndash374
Stoessell RK Hay RL 1978 The geochemical origin of Sepiolite and keorilite atAmboseli Kenya Contributions to Mineralogy and Petrology 65 255ndash267
Suarez DL Šimůnek J 1996 Solute transport modeling under variably saturatedwater 1047298ow conditions In Lichtner PC Steefel CI Oelkers EH (Eds) ReactiveTransport in Porous Media Reviews in Mineralogy vol 34 Mineralogical Society of America Michigan Book Crafters Inc 229ndash268
Sumner DY Grotzinger JP 1996 Were kinetics of Archean calcium carbonate
precipitation related to oxygen concentration Geology 24 119ndash122Su C Suarez DL 2004 Boron release from weathering of illites serpentine shales
and illiticpalygorskitic soils Soil Science Society of America Journal 68 96 ndash105Toscani L Venturelli G Boschetti T 2001 Sul1047297de-free and sul1047297de-bearing waters in
the Northern Apennines Italy general features and waterndashrock interaction AquaticGeochemistry 7 195ndash216
Treves BE Harper GD 1994 Exposure of serpentinites on the ocean 1047298oor sequenceof faulting and hydrofracturing in the Northern Apennine ophicalcites O1047297oliti 19b435ndash466
Varadachari C Kudrat M Gosh K 1994 Evaluation of standard free energies of formation of clay minerals by an improved regression method Clays and ClayMinerals 42 298ndash307
Veniale F 1966 Sepiolite in sedimenti dellrsquoAppennino Vogherese (Pavia) Periodico diMineralogia 35 343ndash386
Venturelli G 2003 Acque minerali ed ambiente mdash Fondamenti di geochimica deiprocessi di bassa temperatura Pitagora Editrice Bologna
Venturelli G Contini S Bonazzi A Mangia A 1997 Weathering of ultrama1047297c rocksand element mobility at Mt Prinzera Northern Apennines Italy MineralogicalMagazine 61 765ndash778
Venturelli G Frey M 1977 Anchizone metamorphism in sedimentary sequences of the Northern Apennines Rendiconti Societagrave Mineralogica Italiana 33 109ndash123
Vieillard P 2002 A new method for the prediction of Gibbs free energies of formationofphyllosilicates (10 Aring and14 Aring)basedon theelectronegativityscale Claysand ClayMinerals 50 352ndash363
Vissers RLM Strating EHH 2001 Structures and microstructures in a thrust-relatedgreenschist facies tectonic melange Voltri Group (NW Italy) O1047297oliti 26 33ndash46
Viti C Mellini M 1997 Contrasting chemical compositions in associated lizardite andchrysotile in veins from Elba Italy European Journal of Mineralogy 9 585 ndash596
Viti C Mellini M 1998 Mesh textures and bastites in the Elba retrogradeserpentinites European Journal of Mineralogy 10 1341ndash1359
Wenner DB Taylor Jr HP 1974 Deuteriumatomic hydrogen and oxygen-18oxygen-16 studies of serpentinization of ultrama1047297c rocks Geochimica et CosmochimicaActa 38 1255ndash1286
Williams LA Parks GA Crerar DA 1985 Silica diagenesis I Solubility controls Journal of S edimentary Petrology 55 301ndash311
Wilson J Savage D Cuadros J Shibata M RagnarsdottirKV 2006 The effect of ironon montmorillonite stability (I) Background and thermodynamic considerations
Geochimica et Cosmochimica Acta 70 306ndash322WollastR MackenzieFTBricker OP1968 Experimental precipitation and genesis of sepiolite at earth-surface conditions American Mineralogist 531 1645ndash1662
Wrucke CT1996 Serpentine and carbonated-hosted asbestos deposits InEdwardAdu Bray (Eds) Preliminary compilation of descriptive geoenvironmental mineraldeposit models 95ndash831 US Geological Survey Open-File Report pp 39 ndash46
Xu T Apps JA Pruess K 2001 Analysis of mineral trapping for CO2 disposal in deepaquifers Lawrence Berkeley National Laboratory University of California
ZhangYDaweRA2000 In1047298uence ofMg2+ on thekineticsof calcite precipitation andcalcite crystal morphology Chemical Geology 163 129ndash138
Zhu C Anderson G 2002 Environmental applications of geochemical modelingCambridge University Press
91T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
we calculated ΔGf 0=minus399854 kJmol for a theoretical (09)(lizardite)+
(01)(berthierine) mineral (Table 2) The log K value of the stated solidsolution was calculated at different temperatures using PHREEQCIafter loading the thermodynamic data on the two end-members intothe database The log K results obtained for Al-lizardite were added tothe thermodynamic database of the solid reactant (Table 3) andcompared with data from the literature (Table 4)
54 Composition of the starting solution O 2 partial pressure and
minerals included in the model
The weighted average chemical composition of rainwater inBologna (Panettiere et al 2000) which is 95 km east of theinvestigated area was used as the starting solution in the modelFull chemical data on rainwater of the TarondashCeno Valleys are
unavailable
Total alkalinity was recalculated for log PCO2=minus35 bar and
T =11 degC which was the mean temperature of the springs sampledand for pe (or Eh) based on chemical data for the redox couple N(minusIII)N(+V) The values of pH and pe obtained were 566 and 452(Eh=025 V) respectively the former agrees with the median valuereported for rainwater in thearea (Venturelli et al1997) Because dataon Al Fe and Si in rainwater were lacking we calculated theconcentrations of these elements assuming logPCO2=minus30 bar andslight supersaturation relative to ferrihydrite-6 line and kaolinite Inthis way a hypothetical composition of water in equilibrium with thelocal soil was calculated (Table3) and adopted as the ldquoinitial waterrdquo formodeling
Evidence suggests that secondary minerals particularly Fe-bearingphases change the chemical composition of the water 1047298owingthrough the ultrama1047297tes of the TarondashCeno Valleys According to Neal
and Stanger (1983) and Fritz et al (1992) the production of CH4 and
Fig 6 Activity plotfor the MgOndashSiO2ndashAl2O3ndashFe2O3(FeO)ndashH2O systemat differentlog PO2 values Spring waters issuing from TarondashCeno Valleys (symbols as in Fig 2) andVoltri Groupultrama1047297tes are shown (grey area bicarbonate waters small black diamonds alkaline waters from Marini and Ottonello 2002) In plots A B and C the stability 1047297eld limits of vermiculite(Table 2) correspond to25 degCwhilethe amorphoussilica (am-SiO2) brucite serpentine (Sr1ndash4 in Table 4)and sepiolite(Sp1ndash2 in Table 4) phasescorrespondto 11 degCThepath calculatedfor interaction between waterand ultrama1047297c rockis also shown (dotted line)the number-letter combinations referto the appearanceand disappearance of minerals(see Table 5)
84 T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
H2 in thealkalinewaters1047298owing from theultrama1047297tes isrelatedto theconsumption of oxygen during the oxidation of iron forming thehematitendashmagnetite pair Therefore the activity of iron and theoxidation state of the water must be de1047297ned The analysis of watersamples from ultrama1047297tes were plotted in both the Pourbaix andactivity diagrams (Figs 5 and 6) where the calculated activity of Fe2+Mg2+ SiO2 SO4
2minus (Pourbaix) and log PO2 (activity diagrams) wereinvolved Iron concentration was quite variable in the TarondashCenosprings where the average content was 27 μ gL and the contentreached a maximum of 200 μ gL and in the Voltri springs where theaverage was 44 μ gL and the maximum was 948 μ gL (Marini andOttonello 2002) In both cases the variations are probably related to
the different mineralogies of the soils and the different solubilities of the Fe-rich phases like goethite and ferrihydrite Iron-bearingsaponites and montmorillonites are the prevailing clay minerals inthe area investigated probably because of the ubiquitous presence of iron oxy-hydroxide (Alietti et al 1976 G Venturelli pers comm) Forthis reason the model considers the involvement of pure Mg-smectiteto be negligible
In the Pourbaix diagram (Fig 5) the data from bicarbonate watersfall into the ferrihydrite-goethitendashvermiculite1047297elds and the data fromalkaline waters fall on the boundaries of the vermiculitesaponite andsaponitechlorite Since ferrihydrite is the 1047297rst precipitating Fe-phaseit was used rather than goethite in the modeling The diagram alsosuggests the transformation of ferrihydrite into magnetite during thechange from bicarbonate to alkaline composition but this process
requires a large amount of Fe2+ in solution (N2 mM Hansel et al2005) which is unlikely at this stage of the waterndashrock interactionThus the magnetite participating in the equilibrium is most likelyprimary in origin
The modeling and the activity diagrams (Figs 6 and 9) werecalculated using the average value PO2= 10minus26 or 10minus50 bar forbicarbonate and alkaline waters respectively These values arecompatible with those calculated by Marini and Ottonello (2002) forbicarbonate waters (PO2= 10minus40 bar) and for alkaline waters of theVoltri Group (10minus77 barbPO2b10minus42 bar) The lowest PO2 value whichwas used as a limit in the modeling was calculated from the S(+VI)S(minusII) redox couple Moreover the activity diagrams (Fig 6) indicatethat the data from the bicarbonate waters from the TarondashCeno Valleysfall between quartz whose log [SiO2] =minus433 is not shown because of
the slow precipitation kinetics under the PndashT conditions considered
here (Williams et al 1985) and amorphous silica boundary 1047297elds(Fig 6A) at the same manner bicarbonate waters with higher SiO2-activity from the Voltri Group fall on amorphous silica boundary(Fig 6B) In summary then the bicarbonate waters are supersaturatedin kaolinite ferrihydrite vermiculite and saponite Alkaline watersare also supersaturated with saponite (Fig 6AndashC) and they are alsoequilibrated with primary magnetite and under extreme conditionsprobably pyrite as well
55 Limitations of the model related to kinetics
Model calculations disregard the rates of mineral dissolution and
precipitation which becomes increasingly important as temperature
Table 3
Starting mineralogy (left side) and 1047298uid composition (right side) used in the modeling
Rock composition Fluid composition
Modal of the mineralsin the interacting rock
mol of the minerals usedin the modeling
meteoric water composition Initial 1047298uid composition used in the modeling
Spinel 003 Na 3523Eminus05 3523Eminus055 Spinel Chromite 001 S 3519Eminus05 3519Eminus05
Magnetite 0 Si ndash 1176Eminus06Kaolinite ndash 5258Eminus08
Ferrhydrite-6line ndash 5882Eminus07
The modal percentage of the minerals in a serpentinite from TarondashCeno Valleys (Ernst and Piccardo 1979) was re-calculated assuming an ideal solid solution between pure extremephasesand takingthe modal percentage tobe themass percentage Thelatter was used asthe solid reactant composition in thewaterndashrockinteraction modelPure solid phases werechosenfrom the llnldat thermodynamic database of the PHREEQCI softwareThe weighted average chemical compositionof the meteoric waterfrom Bologna (Panettiere et al 2000)was chosen as the initial 1047298uid composition in the modeling after its equilibration with kaolinite and ferrihydrite (see text for details)
Table 4
Summary of thermodynamic data for the sepiolite (Spn) and serpentine (Sr n) mineralsphases in italics do not refer to the mentioned hydrolysis reactions
ΔGf 0 log K Reference
(kJ molminus1) (at 25 degC)
Sepiolite (Spn) mdash Mg4Si6O15(OH)2middot6H2O+8H+=4Mg2++6SiO2+11H2Ominus925400 3004 Gunnarsson et al (2005)minus925133 3050andash3172 Christ et al (1973)
(Sp2) minus921986 3602andash3756 Wollast et al (1968)
Serpentine (Srn) mdash Mg3Si2O5(OH)4+ 6H
+
=3Mg
2+
+2SiO2+ 5H2ONi-lizardite minus 289324 1680 Mondesir and Decarreau (1987)
Al-lizardite2 minus411003 2544a Caruso and Chernosky (1979) Al-lizardite minus 339854 2786 this work (see text for details)
Mg-lizardite (Sr1) minus404025 3056 Wilson et al (2006)Mg-crysotile (Sr2) minus403702 3112 bthermocomV8R6Mg-lizardite minus402824 3340 Mondesir and Decarreau (1987)p-antigorite (Sr3) minus564239a 3479 Gunnarsson et al (2005)Mg-chrysotile (Sr4) minus406809 3669 Bruni et al (2002)Mg-chrysotileasbestos minus562505a 3888 Pfeifer (1977)Antigorite minus6612864 47719 bthermocomV8R6
The 1047297eld boundaries of the underlined phases shown in Fig 6 correspond to atemperature of 11 degC
a In the ΔG f deg column value calculated using the ΔG f
deg data of the elements from llnldat
and logK value in the reference in the LogK column value calculated using ΔG f deg
(mineral) in the reference and ΔG f deg data of the elements from llnldat
b Thermodynamic database of Lawrence Livermore National Laboratory included in
PHREEQCI and Geochemists Workbench softwares
85T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
decreases The waterndashrock interaction in the TarondashCeno basin iscomplicated by the presence of secondary mixed-layer clay mineralsthe origin of which is quite dif 1047297cult to model because the solidndashstatetransformations and weathering reactions are strictly related tohydrological and climatic conditions as described above An equili-
brium-based reaction path model is a valuable tool for assessing the
most probable reactions and for de1047297ning the product phases that canbe used in a kinetics-based model Since the kinetics of precipitationof Mg-clays at room temperature and pressure are poorly knownrelative reaction rate effects in the reaction path model can be mimedby omitting the stabler phase when a metastable mineral is expected
to precipitate Therefore the precipitation of MgndashFe3+
Abbreviations are as follows Fer-6line 6-line ferrihydrite kaol kaolinite verm vermiculite magn magnetite sap saponite serp serpentine clin clinochlore pyr pyrite calccalcite hydrom hydromagnesite bruc brucite enst enstatite The chromium decrease during the change in water facies from bicarbonate to alkaline was modeled withsupersaturation (precipitation) of eskolaite and thereafter with equilibrium with chromite (Boschetti 2003)- not bound dagger chlorite mineral from solution Dagger chlorite mineral added to the paragenesis
86 T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
montmorillonites vermiculite and saponites are predicted to occurrather than that of Mg-clays Moreover in order to simulate the slowprecipitation of calcite in depth precipitation was not activated atslight supersaturation (SIcalcite=001) during the initial stages of thewaterndashrock interaction in this manner saturation index of calcite wasfree to increase in the following steps and in alkaline waters
6 Discussion of the model results
61 Explanation of pre-modeling choices and post-modeling results
The results of the reaction path modeling are reported in Table 5and shown in Figs 6 and 9 The reaction paths model the dissolutionof a typical local serpentinite (mineralogically constrained as inTable 3) and the sequential precipitation of gibbsite kaoliniteferrihydrite vermiculite Fe2+ndashMg2+-saponite and a poorly-crystal-line serpentine( p-antigorite of Gunnarsson et al2005) The1047297rst stepinvolves the reaction of the ldquoinitial waterrdquo with serpentinite atconstant temperature (11 degC) PO2 (10minus26 bar) and PCO2 (10minus30 bar)this early stage corresponds to bicarbonate groundwater feeding theinvestigated spring system At the stage of primary magnetiteequilibrium and Fe2+ndashMg2+-saponite supersaturation the modelassumes the system to be closed to CO2 and to have a 1047297xed partialpressure of oxygen (PO2= 10minus40 bar) these conditions represent
when water switch from a bicarbonate to an alkaline compositionThe calculated amount of interacting serpentinite needed to obtainthe alkaline springs of the TarondashCeno Valleys (mean PCO2= 10minus7 bar)is approximately 00203 moles (see Table 5)
Bruni et al (2002) explained the origin of the alkaline waters of the Voltri Group as the result of the evolution of meteoricbicarbonate water following the precipitation of sepiolite withabundance similar to that of goethite This might be supported bythe observation that sepiolite or smectite precipitate at lowtemperature from solutions with low Al content when the pH is inthe range 75ndash95 or N95 respectively (Siffert and Wey 1962 Jonesand Galan 1988) but sepiolites form from the weathering of serpentines in arid climate regions whereas sepiolites are rare inhumid climate regions where serpentinite changes to Fe-smectite
(Singer 1989 Wrucke 1996) In the ophiolite rocks of the Northern
Apennine sepiolite is found only sporadically in veins (Giuseppetti1953 Veniale 1966) and is unknown in the ophiolites of the Tarondash
Ceno Valleys whereas it is thought that sepiolite of the Voltri Groupprecipitated from metasomatic 1047298uids during development of shearzones (Crispini and Capponi 2002)
As described above our model explains the transition frombicarbonate water to alkaline water (UM15 and PR10 samples) asthe result of low-temperature precipitation of a poor crystalline
serpentine with the solubility product of p-antigorite (see Sr3 inTable 4 and Fig 6) The model indicates that the bicarbonate waters of the Voltri Group with the highest [Mg2+][H+]2 activity ratios similarlyevolve towards alkaline waters (Fig 6B) The logK value of thesynthetic low temperature serpentine adopted in this model and its
Fig 7 Saturation indices of chrysotile and p-antigorite (Sr2 and Sr3 thermodynamic
data respectively in Table 4) for bicarbonate and alkaline waters coming fromultrama1047297c rocks of the TarondashCeno Valleys as a function of SiMg molar ratioThermodynamic data on chrysotile and p-antigorite are from the thermocomV8R6
database and Gunnarsson et al (2005) respectively
Fig 8 Activityplot forthe MgOndashSiO2ndashAl2O3ndashFe2O3(FeO)ndashH2O system analyzed forSiO2-phases and aqueous species at different log PO2 and log[aAl3+(aH+)3] values Springwaters coming from the TarondashCeno Valleys (symbols as in Fig 2) and Voltri Groupultrama1047297tes (symbols as in Fig 5) are shown The greyarea refers to bicarbonate watersfrom theVoltri Group (Mariniand Ottonello2002Bruniet al 2002) Thesepiolite 1047297eldis very narrow or hidden by saponite-Mg (1047297eld limits shown as continuous lines)saponite-Fe3+ndashMg (1047297eldlimits shown as dottedlines) or vermiculite(1047297eldlimits shownas dottedndashdashed lines) The talc stability 1047297eld and boundaries with saponite-Mg
montmorillonite-Mg and cronstedite-Mg are shown for comparison (dashed lines)
87T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
relative uncertainty of plusmn2 (see Fig 8 in Gunnarsson et al 2005) arecomparable to the logK variation reported for other low-temperatureMg-serpentines (Table 4) that is within the lines Sr1ndash4 for serpentineminerals where samples varying from bicarbonate composition toalkaline composition in the ultrama1047297c aquifer are grouped (Fig 6)Fig 7 shows a plot of calculated serpentine saturation index as a
function of the SiMg molar ratio for the water samples from theultrama1047297tes of the TarondashCeno ValleysThe SI andSiMgvalues of pureMg-serpentine (SI= 0 SiMg=067) provide a useful index forseparating bicarbonate waters (SIb0 SiMgb067) from alkalinewaters (SIN0 SiMgN067) and the distribution of data points isconsistent with the idea that bicarbonate waters approach saturationwith serpentine until they become supersaturated as alkaline waters
Based on the Mgvs SiO2(aq) activitydiagram and speci1047297cratiosoflog[aAl3+(aH+)3] Birsoy (2002) analyzed the behavior of sepiolite duringultrama1047297te weathering The diagram produces a very narrow stability1047297eldfor sepiolite(Fig 8) when it is programmedwith suitable conditionsfor the bicarbonate waters (log PO2=minus25 bar log [aAl3+(aH+)3]= 65) andthe alkaline waters (log PO2=minus50 bar log [aAl3+(aH+)3]= 45) sampled inthe present study This stability 1047297eld grows larger when the absence of
Mg-smectites and talc is assumed but this is an unrealistic conditionbecause (i) all these minerals occur in the rocks within the areainvestigated and (ii) at high [Mg2+][H+]2 activity ratios the 1047297eld of sepiolite is smaller than that of subordinate compared to those poorly-crystalline serpentine (p-antigorite) and clinochlore
Among the alkaline springs of the Voltri Group the waters withlow log [aCa2+(aH+)2] values (Fig 9) show patterns compatible withchlorite interaction (step 5c in Fig6D)The alkaline waters whose datafall near the clinochlore-14Aring 1047297eld (Fig 6D) cannot be explained bysepiolite precipitation because the sepiolite line does not cross thechlorite 1047297elds Thus the peculiar composition of some alkaline watersof the Voltri Group is probably due to interactions between water andprimary chlorites of the ultrama1047297c rocks (optional step 5c in Table 5cb) This is supported by the occurrence sometimes in large amounts
of metasomatic Mgndash(Fe)-chlorite (Scambelluri and Rampone 1999
Vissers and Strating 2001) in the Ligurian area where the relativeabundance of chlorite can reach 50 in veins in high strainserpentinite mylonites of the ErrondashTobbio ultrama1047297tes (Scambelluriet al 2001b) Otherwise in a chlorite-free primary paragenesis thesolution evolving on the Sr3 serpentine line reaches equilibrium withclinochlore but probably the latter does not precipitate due to theslow kinetics of precipitation under these conditions Thus thedissolution of additional reacting rock in the iterative weathering
process supplies Mg
2+
to the solution thereby favoring the Sr3 to Sr4poorly crystalline serpentine transformation In fact all the PR10samples and some of the alkaline springs from the Voltri Group showincreasing [Mg2+][H+]2 activity ratio and low variation in silicaactivity (Fig 6D) starting from the invariant triple point 4a (Sr3-serpentine Fe2+ndashMg-saponite chlinoclore-14 Aring saturation) andextending towards the brucite line This path can be explained asthe result of the transformation of poorly crystalline serpentine fromSr3 to Sr4 the latter is more stable at higher [Mg2+][H+]2 activityratios Finally at very low PO2 values vermiculite disappears and thestability 1047297elds of chlorite and Fe2+ndashMg-saponite become narrowerwhereas those of Fe2+-rich phases (eg Fe2+-saponite) become widerThe data for the alkaline waters with the highest log [aMg2+(aH+)2]from the Voltri Group fall near the stability 1047297eld of peridotite primaryminerals like forsterite and enstatite (Fig 6D) therefore the last stepsof theevolution involve theequilibration of thesolutionwithenstatite(steps 8a and 8c in Table 5ab)
62 Ca-rich alkaline waters
In preliminary modeling of the evolution of the TarondashCeno springwaters Boschetti (2003) concluded that meteoric water initiallysaturated in Fe3+ndashMg2+-chlorite Fe3+ndashMg2+-saponite and pyrite (andor goethite) and then interacted with the ultrama1047297te mineralassemblage eventually evolved into the alkaline waters In themodel of Boschetti (2003) chlorite saponite and calcite were added
Fig 9 Activity plot for the MgOndashCaOndash(CO2)ndashH2O system with phase limits calculated ata temperature of 11 degCSpring waterscoming from theTarondashCeno Valleys (symbols as inFig2) andVoltri Groupultrama1047297tes(symbolsas in Fig5) areshown Calculated path forwaterndashultrama1047297c rock interaction is shown (dotted line) with the number ndashlettercombinations referring to the appearance or disappearance of minerals (see Table 5)The grey area refers to bicarbonate waters from the Voltri Group (Marini and Ottonello2002)
Fig 10 Plots of (A) calcite saturation indexes and (B) PCO2 as a function of Clminus ascalculated by PHREEQCI software for waters from the TarondashCeno Valleys Alkalinesprings for the Voltri Group are also shown (small black diamonds Marini andOttonello 2002) The grey shading is deeper(A) in the saturation1047297eld(SI= 00plusmn05) and
(B) where the lowest log PCO2 value is approached
88 T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
to the solution in order to achieve a high [Ca2+][H+]2 activity ratio andsupersaturation in these phases However in light of the work of Clarket al (1992) and Bruni et al (2002) we believe that the calcitesupersaturation in alkaline waters (Fig 9 and 10) which in our modelis achieved by preventing the precipitation of calcite in the 1047297rst stepsof the reaction path (Table 5a and b) depends on the differencebetween CO2 partial pressureat thegivendepth(PCO2le10minus10bar) andat the surface (PCO2cong10minus35) Waters rising to the surface are far from
equilibrium with the atmosphere where the PCO2 is 4times10
6
timesgreater than the value at depth thus the water rapidly take up CO2which is promptly converted to CO3
2minus because of the high pH Thisleads to supersaturation with calcite through the following hydro-xylation reaction
As a result large amounts of travertine are generated in the VoltriGroup (Marini and Ottonello 2002) and carbonate concretionsknown as devilrsquo s coins are formed at Mt Prinzera in Taro valleyBelow the surface the SIcalcite of the alkaline waters is expected to belower than what the model estimates Under conditions of lowcarbonate alkalinity low CO2 and O2 partial pressure and high pHcalcite nucleation may be inhibited and the kinetics of precipitationreduced by the high concentrations of dissolved divalent cations suchas Mg2+ Fe2+ and Mn2+ (Sumner and Grotzinger 1996 Davis et al2000 Zhang and Dawe 2000 de Leeuw 2002) The fact that severalalkaline water samples fall within or near the calcitendashdolomiteequilibrium (Fig 9) may indicate magnesium calcite or protodolomitenucleation in those extreme subsurface conditions However possibleinteraction between some calcium-rich waters of the Voltri Group andCa-rich rocks such as ophicarbonates or rodingites cannot beexcluded these rocks are widespread in the Ligurian Apennineserpentinites (Treves and Harper 1994)
63 Sepiolite vs serpentine nucleation under weathering conditions
constraints on the kinetics and equilibrium-based approaches
The kinetics of precipitation of poorly-crystalline serpentine and
sepiolite cannot be compared quantitatively because rate constantsare available only for well-ordered crystalline phases (eg see Palandriand Kharaka 2004 for crystalline serpentine) Sepiolite precipitationis often invoked to explain magnesium scavenging from solutions butthe lack of kinetic data for this mineral leads to overly simplisticreaction transport models These models are forced to use kinetic datafor phyllosilicates such as kaolinite or muscovite (eg Xu et al 2001Bryan 2005) Moreover the low-temperature thermodynamic dataeven if they are fairly uncertain show that sepiolite solubilityincreases with decreasing crystallinity (Table 4 Stoessell 1988) Infact crystalline sepiolite requires a long time to reach equilibriumwith solution which imposes a slow control on thewater composition(eg Stoessell and Hay 1978) Consequently in modeling based onreaction transport that simulates freshly precipitated or poorly
crystalline sepiolite formation the logK value of Wollast et al (1968)is preferable to the lower values (Suarez and Šimůnek 1996) Thispreference agrees with the approach of Mariner (2003) and Jove Colon(2006) who adopt a logK (25 degC) value for sepiolite of 3044 It is alsosimilar to that reported by Wollast et al (1968) and it is six logK unitshigherthan the valueused in thermocomV8R6 dataset which is oftenused (perhaps improperly) to model the supergenic nucleation of sepiolite In the stability diagram in Fig 6 where the Sp1 and Sp2 linesrepresent the stability boundaries of crystalline and poorly-crystallinesepiolite respectively the evolution of TarondashCeno waters frombicarbonate to alkaline (Figs 6A and C) follows the stability 1047297eldboundary of the poorly-crystalline serpentine (between Sr2ndashSr3 lines)rather than that of the poorly-crystalline sepiolite (Sp2) as postulatedin themodel The lines for Sp2 Sr3 and amorphoussilica intersect at a
common point (Fig 6D) which is consistent with the nucleation of
sepiolite that typically occurs in solutions that are saturated orsupersaturated with amorphous silica and (Mg)-carbonate (egStoessell and Hay 1978 Birsoy 2002) Some bicarbonate watersfrom the Voltri Group have a silica activity near the amorphous silicaboundary (Fig 6B) but only a few alkaline springs of this group fall onthe poorly-crystalline sepiolite boundary line (Sp2 Fig 6D) Both of these features are inconsistent with the possibility that sepioliteprecipitates from supergenic fresh water in the Northern Apennine
Moreover these two results suggest that values for logK(sepiolite)b
3044 (see Table 4) should be used to model the dissolutionprecipitation of the well-crystallized phase that is the most commonform for the metasomatic sepiolite found in the investigated area
7 Conclusions
The waters issuing from ultrama1047297tes in the TarondashCeno Valleysoriginate from low-temperature interactions between meteoric waterand serpentinite rocks In this report the waterndashrock interaction hasbeen modeled using reaction paths in progress mode During their earlyinteraction with soil and rock minerals bicarbonate waters super-saturated in kaolinite ferrihydrite vermiculite and Fe2+ndashMg2+-saponitewhere theorigin of the secondary iron saponite as Fe2+-bearing smectiteagrees with the hypothesis of Guumlven (1988) Then while remainingsupersaturated with saponite the bicarbonate waters evolve into analkaline facies when poorly-crystalline serpentine precipitates Duringrock weathering the Al-bearing crystalline serpentine of oceanic originprobably promotes nucleation of the low-temperature serpentineamorphous phase and supplies Al3+ to the solution for saponiteprecipitation This interpretation is consistent with the existence of Al-poor supergenic serpentine (Neal and Stanger 1985) and the partiallymeteoric OH-isotopic imprint observed in other ultrama1047297c systems
Dif 1047297culties and complexities in the modeling arise fromuncertainties in the primary mineralogy of some local ophiolitesand from the presence of different lithotypes (rodingites ophi-carbonates gabbros) which can signi1047297cantly affect the composi-tion of water samples This can be seen in the example of the Candash
OH springs issuing from the Voltri Group (PO2 and PCO2 of 10minus70
and 10minus10 bar respectively Bruni et al 2002 Marini and
Ottonello 2002) These springs appear to be more evolved thanthe TarondashCeno alkaline waters as a result of their particularly longinteractions with serpentinites until equilibrium with enstatite isreached or with chlorite-bearing serpentinite The use of anupdated thermodynamic database in our model avoided problemsdue to mineralogical heterogeneity in the mixed-layer claysassociated with ophiolitic rocks this database accurately re1047298ectedthe chemical composition of the single phases Using this modelthe vermiculitendashchlorite and saponitendashchlorite associations seen inthe activity diagrams simulate respectively the presence of highand low charge corrensite observed in the 1047297eld The accuracy of ourmodel can be tested by comparing estimates of chlorine composi-tion resulting from our model and from experimental methods The
mean Cl concentration of the bicarbonate water is 41plusmn16 mgLtaking into account a Cl concentration of sim2 mgL in rainwater inthe area (cf Panettiere et al 2000 Venturelli 2003) 2 mgL of Clin spring water are estimated to come from serpentinite This valueagrees with the model which predicts that the switch frombicarbonate water to alkaline water requires the congruentdissolution of 00203 mol of serpentinite containing 100 ppm of soluble Cl thereby yielding water containing 2 ppm of Cl
In addition we think that weathering of the accessory albiteplagioclase is responsible for Na enrichment of the waters whereas Clenrichment in the alkaline waters is predicted to result from theleaching of NaCl 1047298uid inclusions Moreover the TarondashCeno alkalinesprings show the highest BCl ratios comparable with those of thehost metamorphic minerals (Scambelluri et al 2004) In contrast the
lower boron content and BCl ratio of the alkaline springs of the Voltri
89T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
Group are unexpected features and probably re1047298ect waterndashrockinteractions over a longer time period than in the TarondashCeno ValleysThe boron content in the Northern Apennine alkaline springs may becontrolled by several factors such as (i) the different histories of VoltriGroup and EL ultrama1047297tes (ii) the probable contribution of boronfrom NaCl 1047298uid inclusions (iii) the sinking of boron by travertineprecipitation (Hobbs and Reardon1999) brucite (Rhoades et al 1970)and smectite and (iv) decreases in boron release from minerals due to
increases in pH (Su and Suarez 2004) Acknowledgements
This research was carried out with the 1047297nancial support from theldquoYoung Researchers Projectrdquo (2001ndash2002) of the Italian Ministry of Universities Scienti1047297c Research and Technology (MURST) We thankGianniCortecci (Institute of Geosciences and EarthResources IGG-CNRPisa Italy) and Giampiero Venturelli (Earth Sciences Department Univer-sity of Parma Italy) for useful suggestions and critical review of themanuscript The authors are grateful to Enrico Dinelli (InterdepartmentResearch Centre for Environmental Sciences Alma Mater StudiorumUniversity of Bologna Ravenna Italy) for granting permission to use theGeochemistrsquos WorkbenchsoftwareTheconstructive reviews of theEditor
J Fein and of two anonymous Referees have signi1047297cantly improved the
manuscript
Appendix A Supplementary data
Supplementary data associated with this article can be found inthe online version at doi101016jchemgeo200808017
References
Abbate E Bortolotti V Passerini P 1980 Olistostromes and olistoliths In Sestini G(Ed) Development of the Northern Apennines geosyncline Sedimentary Geology4 pp 521ndash558
Alietti A 1956 Il minerale a strati misti saponite-talco di monte Chiaro (val di TaroAppen Emil) Rendiconti dellAccademia Nazionale dei Lincei 21 201 ndash207
Alietti A De Angelis G Montanari G Sgarlata F Tirelli G 1976 Il ferro nelle argilleappenniniche Periodico di Mineralogia 45 51ndash64
Anselmi B Mellini M Viti C 2000 Chlorine in the Elba Monti Livornesi and Murloserpentines Evidence for seandashwater interaction European Journal of Mineralogy12137ndash146
Apha-Awwa-Wef 1995 Standard Methods for the Examination of Water and Waste-water 19th Edition American Public Health Association American Water WorksAssociation Water Environment Federation
Barnes I LaMarche Jr VC Himmelberg GR 1967 Geochemical evidence of present-day serpentinization Science 156 830ndash832
Barnes I ONeil JR 1969 The relationship between 1047298uids in some fresh alpine-typeultrama1047297cs and possible modern serpentinization Western United StatesGeological Society of America Bulletin 80 1947ndash1960
Barnes I ONeil JRTrescases JJ1978 Present dayserpentinizationin New CaledoniaOman and Yugoslavia Geochimica et Cosmochimica Acta 42 144ndash145
Barnes I Rapp JB ONeil JR Sheppard RA Gude III AJ 1972 Metamorphicassemblages and the direction of 1047298ow of metamorphic 1047298uids in four instances of serpentinization Contributions to Mineralogy and Petrology 35 263ndash276
Barnes JDSelverstone J Sharp ZD 2006Chlorine isotope chemistry of serpentinitesfrom Elba Italy as an indicator of 1047298uid source and subsequent tectonic historyGeochemistry Geophysics Geosystems 7 doi1010292006GC001296
Beccaluva L Emiliani F Venturelli G Zerbi M 1973 Ca Fe Mg Mn Cr Ni Codistribution in some ultrama1047297c rocks outcropping in the Northern Apennines withsome geological remarks Ateneo Parmense Acta Naturalia 9 69ndash98
Beccaluva L Macciotta G Piccardo GB Zeda O 1984 Petrology of lherzolitic rocksfrom the Northern Apennine ophiolites Lithos 17 299 ndash316
BeccaluvaL Venturelli G ZanzucchiG 1975 Dati geologici e geochimici sui basalti diassociazione o1047297olitica dellAppennino ligure-emiliano Ateneo Parmense ActaNaturalia 11 789ndash802
BethkeCM 2002 The Geochemists Workbench A Userrsquos Guide to RxnAct2Tact ReactandGtplot Release40 HydrogeologyProgram Universityof IllinoisUrbanaIL 236pp
Birsoy R 2002Formation of sepiolitendashpalygorskite and related minerals from solutionClays and Clay Minerals 50 736ndash745
Bonatti E Lawrence JR Morandi N 1984 Serpentinization of oceanic peridotitestemperature dependence of mineralogy and boron content Earth and PlanetaryScience Letters 70 88ndash94
Bonifacio E Zanini E Boero V Franchini-Angela M 1997 Pedogenesis in a soilcatena on serpentinite in north-western Italy Geoderma 75 33ndash51
Boschetti T 2003 Studio geochimico e geochimico-isotopico di acque a composizione
estrema e termali dellrsquoAppennino Settentrionale Ph D Thesis Parma Univ
Brigatti MF Franchini G Lugli C Medici L Poppi L Turci E 2000 Interactionsbetween aqueous chromium solutions and layer silicatesApplied Geochemistry 151307ndash1316
Brigatti MF Lugli C Poppi L Venturelli G 1999 Iron-rich saponite dissolutionreactions and Cr uptake Clay Minerals 34 637ndash645
Brigatti MF Valdregrave G 1996 Mixed-layer chloritesmectite (corrensite) in ophioliticrock alteration products Neues Jahrbuch fuer Mineralogie Monatshefte 1 37 ndash47
Bruni J Canepa M Chiodini G Cioni R Cipolli F Longinelli A Marini L OttonelloG Vetuschi Zuccolini M 2002 Irreversible waterndashrock mass transfer accompany-ing the generation of the neutral MgndashHCO3 and high-pH CandashOH spring waters of the Genova province Italy Applied Geochemistry 17 455ndash474
Bryan CR 2005 Drift-Scale THC Seepage Model MDL-NBS-HS-000001 REV 04 YuccaMountain Project Las Vegas Nevada (US)Caruso LJ Chernosky JV 1979 The stability of lizardite Canadian Mineralogist 17
757ndash769Clark ID Fontes JC Fritz P 1992 Stable isotope disequilibria in travertine from high
pH waters Laboratory investigations and 1047297eld observations from Oman Geochi-mica et Cosmochimica Acta 56 2041ndash2050
Cortesogno L 1980 Il metamor1047297smo giurassico nelle o1047297oliti dellAppenninoSettentrionale due distinti cicli metamor1047297ci in ambiente oceanico O1047297oliti 5 5ndash18
Christ CLHostetler PB Siebert RM1973 Studies in the system MgO-SiO2-CO2-H2OIII American Journal of Science 273 65ndash83
Cortesogno L Lucchetti G 1982 Il metamor1047297smo oceanico nei gabbri o1047297oliticidellAppennino Ligure Aspetti mineralogici e petrogenetici Rendiconti dellaSocietagrave Mineralogica Italiana 38 561ndash579
Crispini L Capponi G 2002 Control of metasomatic alteration on the development of shear zone in ultrama1047297c rocks a case study from the Voltri Massif (Ligurian Alps)Architettura interna e cinematica delle zone di taglio Gruppo Italiano di GeologiaStrutturale mdash Riunione Annuale 2002 Pisa Italy 11ndash12 giugno 2002 XIV ndashXV
CrissRE 1999 Principlesof stableisotope distributionOxford University PressNew York
Davis KJ Dove PM De Yoreo JJ 2000 The role of Mg 2+ as an impurity in calcitegrowth Science 290 1134ndash1137
de Leeuw NH 2002 Moleculardynamics simulations of the growth inhibiting effect of Fe2+ Mg2+ Cd2+ and Sr2+ on calcite crystal growth Journal of Physical Chemistry B106 5241ndash5249
De Nardo MT Segadelli S Vescovi P 2007 Studio pilota per la caratterizzazionegeologicadellesorgentinellazona delM Nero (altaVal Cenoe alta ValNuremdash provincedi Parma e Piacenza) Il Geologo dellEmilia-Romagna 25 Nuova Serie pp 5 ndash21
Di Giulio A Geddo G 1990 Studio petrogra1047297co delle Arenarie di Casanova (alta valTrebbia Appennino settentrionale) Atti Ticinesi di Scienze della Terra 33 243ndash254
Dinelli E Lombini A Simoni A Ferrari C 1997 Heavy metals in serpentinite soils of selected outcrops of Piacenza and Parma provinces (Northern Apennines Italy)Mineralogica et Petrographica Acta 40 241ndash255
Epstein S Mayeda T 1953 Variation of 18O content of waters from natural sourcesGeochimica et Cosmochimica Acta 4 213ndash224
Ernst WG Piccardo GB 1979 Petrogenesis of some Ligurian peridotites I Mineraland bulk-rock chemistry Geochimica et Cosmochimica Acta 43 219ndash237
Feder F Trolard F Klingelhofer G Bourrieacute G 2005 In situ Moumlssbauer spectroscopy
evidence for green rust (fougerite) in a gleysol and its mineralogical transforma-tions with time and depth Geochimica et Cosmochimica Acta 69 4463ndash4483Fritz PClark IDFontes JCWhiticar MJFaberE 1992 Deuterium and
13C evidence
for low temperature production of hydrogen and methane in a highly alkalinegroundwater environment in Oman In Kharaka TK Maest AS (Eds) Proceed-ings of the 7th International Symposium on WaterndashRock Interaction BalkemaRotterdam pp 793ndash796
Giammetti F 1968 Le o1047297oliti di Roccaprebalza Costa della Guardia e Groppo di Gorro(Appennino Parmense) Ateneo Parmense Acta Naturalia 4 107ndash142
Giuseppetti G1953 Roccee minerali dellaformazioneo1047297olitica di Volpedo Rendicontidella Societagrave Mineralogica Italiana 9 85ndash134
Gran G 1952 Determination of equivalence point in the potentiometric titrationAnalyst 77 661ndash671
Gunnarsson I Arnorsson S Jakobsson S 2005 Precipitation of poorly crystalline antigoriteunder hydrothermal conditions Geochimica et Cosmochimica Acta 69 2813ndash2828
Guumlven N 1988 Smetites In Bailey SW (Ed) Reviews in Mineralogy 19 Hydrousphyllosilicates (exclusive of micas) Mineralogical Society of America MichiganBook Crafters Inc 497ndash559
Hansel CM Benner SG Fendorf S 2005 Competing Fe(II)-induced mineralization
pathways of ferrihydrite Environmental Science and Technology 39 7147ndash7153HobbsMYReardonEJ 1999 Effect of pH on boronco-precipitation by calcite further
evidence for non-equilibrium partitioning of trace elements Geochimica etCosmochimica Acta 63 1013ndash1021
Jones BF Galan E 1988 Sepiolite and palygorskite In Bailey SW (Ed) HydrousPhyllosilicates (exclusive of Micas) Reviews in Mineralogy vol 19 MineralogicalSociety of America Michigan Book Crafters Inc 631ndash674
Jove Colon CF 2006 Technical work plan for Thermodynamic database for chemicalmodelling TWP-MGR-PA-000039 REV 01 Yucca Mountain Project Las VegasNevada (US)
Kohls DW Rodda Jr JL 1967 Iowaite a new hydrous magnesium hydroxidendashferricoxychloride from the Precambrian of Iowa American Mineralogist 52 1261 ndash1271
Kyser TK OHanley DS Wicks FJ 1999 The origin of 1047298uids associated withserpentinization processes evidence from stable-isotope compositions CanadianMineralogist 37 223ndash237
Lapham DM 1961 New data on deweylite American Mineralogist 46 168ndash188Lee BD SearsSKGraham RC Amrhein CValiH 2003 Secondarymineral genesis
from chlorite and serpentine in an ultrama1047297c soil toposequence Soil Science
Society of America Journal 671309ndash1317
90 T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
Longinelli A Selmo E 2003 Isotopic composition of precipitation in Italy a 1047297rstoverall map Journal of Hydrology 270 75ndash88
Majzlan J Navrotsky A Schwertmann U 2004 Thermodynamics of iron oxides PartIIIEnthalpiesof formation and stability of ferrihydrite (simFe(OH)3) schwertmannite(simFeO(OH)34(SO4)18) and ɛ-Fe2O3 Geochimica et Cosmochimica Acta 681049ndash1059
Mariner P 2003 In-Drift PrecipitatesSalts Model ANL-EBS-MD-000045 REV 01 YuccaMountain Project Las Vegas Nevada (US)
Marini L Ottonello G 2002 Atlante degli acquiferi della Liguria Volume III Le acquedeicomplessio1047297olitici (bacini ArrestraBranega Cassinelle Cerusa ErroGorzenteLeira Lemme Lerone Orba Piota Polcevera Rumaro Sansobbia Stura Teiro
Varenna Visone) Pacini PisaMellini L Zanazzi PF 1987 Crystal structures of lizardite-l T and lizardite-2H1 fromColi Italy American Mineralogist 72 943ndash948
Mezzadri G 1964 Petrogra1047297a delle arenarie di Ostia Rendiconti Societagrave MineralogicaItaliana 20 192ndash228
Mondesir H Decarreau A 1987 Synthesegrave entre 25 et 200 degC de lizardites Ni ndashMgMesure des coef 1047297cients de partage solide-solution aqueuse pour le couple Ni ndashMgdans des lizardites Bulletin de Mineralogie 110 409ndash426
Neal C Shand P 2002 Spring and surface water quality of the Cyprus ophiolitesHydrology and Earth System Sciences 6 797ndash817
Neal C Stanger G 1983 Hydrogen generation from mantle source rocks in OmanEarth and Planetary Science Letters 66 315ndash320
Neal C Stanger G 1984 Calcium and magnesium hydroxide precipitation fromalkaline groundwaters in Oman and their signi1047297cance to the process of serpentinization Mineralogical Magazine 48 237ndash241
Neal C Stanger G 1985 Past and present serpentinisation of ultrama1047297c rocks anexample from Semail ophiolite nappe of Northern Oman In Drever JI (Ed) TheChemistry of Weathering D Reidel Publishing Company pp 249ndash275
Nesbitt HW Bricker OP 1978 Low temperature alteration processes affecting
et Cosmochimica Acta 41 1835ndash1841OrsquoHanley DS 1996 Serpentinites records of tectonic and petrological history Oxford
University PressOrberger B Friedrich G Woermann E 1990 The distribution of halogens and carbon
in Pge-bearing ultrama1047297cs of the Acoje Ophiolite Block Zambales Philippines Journal of Geochemical Exploration 37 147ndash169
Ottonello G Joron JL Piccardo GB 1984 Rare Earth and 3d transition elementsgeochemistry of peridotitic rocks II Ligurian peridotites and associated basalts Journal of Petrology 25 373ndash393
Palandri JL Kharaka YK 2004 A compilation of rate parameters of water-mineralinteraction kinetics for application to geochemical modeling Open File Report2004-1068 US Geological Survey
Panettiere P Cortecci G Dinelli E Bencini A Guidi M 2000 Chemistry and sulfurisotopic composition of precipitation at Bologna Italy Applied Geochemistry 151455ndash1467
Parkhurst DL Appelo CAJ 1999 Userrsquos guide to PHREEQC (version 2) mdash a computerprogram for speciation batch-reaction one-dimensional transport and inversegeochemical calculation US Department of the Interior US Geological SurveyWashington DC
Pfeifer HR 1977 A model for 1047298uids in metamorphosed ultrama1047297c rocks Observationsat surface and subsurface conditions (high pH spring waters) SchweizerischeMineralogische und Petrographische Mitteilungen 57 361ndash396
PlummerLN1984Geochemical modeling A comparisonof forwardand inverse methodsIn Hitchon B Wallick EI (Eds) Proceedings First CanadianAmerican Conference onHydrogeology-Practical Applications of Ground Water Geochemistry Banff AlbertaCanada mdash Worthington Ohio National Water Well Association pp 149ndash177
Rampone E Hofmann AW Piccardo GB Vannucci R Bottazzi P Ottolini L 1995Petrology mineral and isotope geochemistry of the External Ligurides peridotites(Northern Apennines Italy) Journal of Petrology 36 81ndash105
Rampone E Hofmann AW Piccardo GB Vannucci R Bottazzi P Ottolini L 1996Trace element and isotope geochemistry of depleted peridotites from an N-MORBtype ophiolite (Internal Liguride N Italy) Contributions to Mineralogy andPetrology 123 61ndash76
Rhoades JD Ingvalson RD HatcherJT1970Adsorption of boronby ferromagnesianminerals and magnesium hydroxide Soil Science Society of America Journal 34
938ndash941Saini-EidukatB KuchaH Keppler H1994Hibbingiteγ-Fe2(OH)3Clanewmineralfromthe Duluth Complex Minnesota with implications for the oxidation of Fe-bearingcompounds and the transport of metals American Mineralogist 79 555ndash561
Sanford RF 1981 Mineralogical and chemical effects of hydration reactions andapplications to serpentinization American Mineralogist 66 290ndash297
Scambelluri M Muumlntener O Ottolini L Pettke TT Vannucci R 2004 The fate of BCl and Li in the subducted oceanic mantle and in the antigorite breakdown 1047298uidsEarth and Planetary Science Letters 222 217ndash234
Scambelluri M Rampone E 1999 Mg-metasomatism of oceanic gabbros and itscontrol on Ti-clinohumite formation during eclogitization Contributions toMineralogy and Petrology 135 1ndash17
Scambelluri M Bottazzi P Trommsdorff V Vannucci R Hermann J Gogravemez-Pugnaire MT Logravepez-Sagravenchez Vizcaino V 2001a Incompatible element-rich1047298uidsreleased by antigorite breakdown in deeply subducted mantle Earth and PlanetaryScience Letters 192 457ndash470
Scambelluri M Rampone E Piccardo G 2001b Fluid and element cycling insubducted serpentinite a trace-element study of the ErrondashTobbio high-pressureultrama1047297tes (Western Alps NW Italy) Journal of Petrology 42 55ndash67
Segadelli S2006 Lageologia nelpaesaggio le rupi o1047297olitiche in Val Taro e Val Ceno IlGeologo dellrsquoEmilia-Romagna 22 15ndash29 Nuova Serie
Settembre Blundo D Martin De Vidales JL Alietti A 1992 Corrensite-type clayminerals from Taro Valley (North Italy) Mineralogica et Petrographica Acta 35215ndash222
Sharp ZD Barnes JD 2004 Water-soluble chlorides in massive sea1047298oor serpenti-nites a source of chloride in subduction zones Earth and Planetary Science Letters226 243ndash254
Siever R Woodford N 1979 Dissolution kinetics and the weathering of ma1047297cminerals Geochimica et Cosmochimica Acta 43 717ndash724Singer A 1989 Palygorskite and sepiolite group minerals In Dixon JB Weed SB
(Eds) Minerals in Soil Environments Soil Science Society of America MadisonWisconsin pp 829ndash872
Siffert B Wey R 1962 Synthese d rsquoune sepiolite a temperature ordinarie Comptesrendus hebdomadaires des seacuteances de lAcadeacutemie des Sciences de Paris 2541460ndash1463
Środoń J 1999 Nature of mixed-layer clays and mechanisms of their formation andalteration Annual Reviews of Earth and Planetary Science 27 19ndash53
Stoessell RK 1988 25 degC and 1 atm dissolution experiments of sepiolite and keroliteGeochimica et Cosmochimica Acta 52 365ndash374
Stoessell RK Hay RL 1978 The geochemical origin of Sepiolite and keorilite atAmboseli Kenya Contributions to Mineralogy and Petrology 65 255ndash267
Suarez DL Šimůnek J 1996 Solute transport modeling under variably saturatedwater 1047298ow conditions In Lichtner PC Steefel CI Oelkers EH (Eds) ReactiveTransport in Porous Media Reviews in Mineralogy vol 34 Mineralogical Society of America Michigan Book Crafters Inc 229ndash268
Sumner DY Grotzinger JP 1996 Were kinetics of Archean calcium carbonate
precipitation related to oxygen concentration Geology 24 119ndash122Su C Suarez DL 2004 Boron release from weathering of illites serpentine shales
and illiticpalygorskitic soils Soil Science Society of America Journal 68 96 ndash105Toscani L Venturelli G Boschetti T 2001 Sul1047297de-free and sul1047297de-bearing waters in
the Northern Apennines Italy general features and waterndashrock interaction AquaticGeochemistry 7 195ndash216
Treves BE Harper GD 1994 Exposure of serpentinites on the ocean 1047298oor sequenceof faulting and hydrofracturing in the Northern Apennine ophicalcites O1047297oliti 19b435ndash466
Varadachari C Kudrat M Gosh K 1994 Evaluation of standard free energies of formation of clay minerals by an improved regression method Clays and ClayMinerals 42 298ndash307
Veniale F 1966 Sepiolite in sedimenti dellrsquoAppennino Vogherese (Pavia) Periodico diMineralogia 35 343ndash386
Venturelli G 2003 Acque minerali ed ambiente mdash Fondamenti di geochimica deiprocessi di bassa temperatura Pitagora Editrice Bologna
Venturelli G Contini S Bonazzi A Mangia A 1997 Weathering of ultrama1047297c rocksand element mobility at Mt Prinzera Northern Apennines Italy MineralogicalMagazine 61 765ndash778
Venturelli G Frey M 1977 Anchizone metamorphism in sedimentary sequences of the Northern Apennines Rendiconti Societagrave Mineralogica Italiana 33 109ndash123
Vieillard P 2002 A new method for the prediction of Gibbs free energies of formationofphyllosilicates (10 Aring and14 Aring)basedon theelectronegativityscale Claysand ClayMinerals 50 352ndash363
Vissers RLM Strating EHH 2001 Structures and microstructures in a thrust-relatedgreenschist facies tectonic melange Voltri Group (NW Italy) O1047297oliti 26 33ndash46
Viti C Mellini M 1997 Contrasting chemical compositions in associated lizardite andchrysotile in veins from Elba Italy European Journal of Mineralogy 9 585 ndash596
Viti C Mellini M 1998 Mesh textures and bastites in the Elba retrogradeserpentinites European Journal of Mineralogy 10 1341ndash1359
Wenner DB Taylor Jr HP 1974 Deuteriumatomic hydrogen and oxygen-18oxygen-16 studies of serpentinization of ultrama1047297c rocks Geochimica et CosmochimicaActa 38 1255ndash1286
Williams LA Parks GA Crerar DA 1985 Silica diagenesis I Solubility controls Journal of S edimentary Petrology 55 301ndash311
Wilson J Savage D Cuadros J Shibata M RagnarsdottirKV 2006 The effect of ironon montmorillonite stability (I) Background and thermodynamic considerations
Geochimica et Cosmochimica Acta 70 306ndash322WollastR MackenzieFTBricker OP1968 Experimental precipitation and genesis of sepiolite at earth-surface conditions American Mineralogist 531 1645ndash1662
Wrucke CT1996 Serpentine and carbonated-hosted asbestos deposits InEdwardAdu Bray (Eds) Preliminary compilation of descriptive geoenvironmental mineraldeposit models 95ndash831 US Geological Survey Open-File Report pp 39 ndash46
Xu T Apps JA Pruess K 2001 Analysis of mineral trapping for CO2 disposal in deepaquifers Lawrence Berkeley National Laboratory University of California
ZhangYDaweRA2000 In1047298uence ofMg2+ on thekineticsof calcite precipitation andcalcite crystal morphology Chemical Geology 163 129ndash138
Zhu C Anderson G 2002 Environmental applications of geochemical modelingCambridge University Press
91T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
H2 in thealkalinewaters1047298owing from theultrama1047297tes isrelatedto theconsumption of oxygen during the oxidation of iron forming thehematitendashmagnetite pair Therefore the activity of iron and theoxidation state of the water must be de1047297ned The analysis of watersamples from ultrama1047297tes were plotted in both the Pourbaix andactivity diagrams (Figs 5 and 6) where the calculated activity of Fe2+Mg2+ SiO2 SO4
2minus (Pourbaix) and log PO2 (activity diagrams) wereinvolved Iron concentration was quite variable in the TarondashCenosprings where the average content was 27 μ gL and the contentreached a maximum of 200 μ gL and in the Voltri springs where theaverage was 44 μ gL and the maximum was 948 μ gL (Marini andOttonello 2002) In both cases the variations are probably related to
the different mineralogies of the soils and the different solubilities of the Fe-rich phases like goethite and ferrihydrite Iron-bearingsaponites and montmorillonites are the prevailing clay minerals inthe area investigated probably because of the ubiquitous presence of iron oxy-hydroxide (Alietti et al 1976 G Venturelli pers comm) Forthis reason the model considers the involvement of pure Mg-smectiteto be negligible
In the Pourbaix diagram (Fig 5) the data from bicarbonate watersfall into the ferrihydrite-goethitendashvermiculite1047297elds and the data fromalkaline waters fall on the boundaries of the vermiculitesaponite andsaponitechlorite Since ferrihydrite is the 1047297rst precipitating Fe-phaseit was used rather than goethite in the modeling The diagram alsosuggests the transformation of ferrihydrite into magnetite during thechange from bicarbonate to alkaline composition but this process
requires a large amount of Fe2+ in solution (N2 mM Hansel et al2005) which is unlikely at this stage of the waterndashrock interactionThus the magnetite participating in the equilibrium is most likelyprimary in origin
The modeling and the activity diagrams (Figs 6 and 9) werecalculated using the average value PO2= 10minus26 or 10minus50 bar forbicarbonate and alkaline waters respectively These values arecompatible with those calculated by Marini and Ottonello (2002) forbicarbonate waters (PO2= 10minus40 bar) and for alkaline waters of theVoltri Group (10minus77 barbPO2b10minus42 bar) The lowest PO2 value whichwas used as a limit in the modeling was calculated from the S(+VI)S(minusII) redox couple Moreover the activity diagrams (Fig 6) indicatethat the data from the bicarbonate waters from the TarondashCeno Valleysfall between quartz whose log [SiO2] =minus433 is not shown because of
the slow precipitation kinetics under the PndashT conditions considered
here (Williams et al 1985) and amorphous silica boundary 1047297elds(Fig 6A) at the same manner bicarbonate waters with higher SiO2-activity from the Voltri Group fall on amorphous silica boundary(Fig 6B) In summary then the bicarbonate waters are supersaturatedin kaolinite ferrihydrite vermiculite and saponite Alkaline watersare also supersaturated with saponite (Fig 6AndashC) and they are alsoequilibrated with primary magnetite and under extreme conditionsprobably pyrite as well
55 Limitations of the model related to kinetics
Model calculations disregard the rates of mineral dissolution and
precipitation which becomes increasingly important as temperature
Table 3
Starting mineralogy (left side) and 1047298uid composition (right side) used in the modeling
Rock composition Fluid composition
Modal of the mineralsin the interacting rock
mol of the minerals usedin the modeling
meteoric water composition Initial 1047298uid composition used in the modeling
Spinel 003 Na 3523Eminus05 3523Eminus055 Spinel Chromite 001 S 3519Eminus05 3519Eminus05
Magnetite 0 Si ndash 1176Eminus06Kaolinite ndash 5258Eminus08
Ferrhydrite-6line ndash 5882Eminus07
The modal percentage of the minerals in a serpentinite from TarondashCeno Valleys (Ernst and Piccardo 1979) was re-calculated assuming an ideal solid solution between pure extremephasesand takingthe modal percentage tobe themass percentage Thelatter was used asthe solid reactant composition in thewaterndashrockinteraction modelPure solid phases werechosenfrom the llnldat thermodynamic database of the PHREEQCI softwareThe weighted average chemical compositionof the meteoric waterfrom Bologna (Panettiere et al 2000)was chosen as the initial 1047298uid composition in the modeling after its equilibration with kaolinite and ferrihydrite (see text for details)
Table 4
Summary of thermodynamic data for the sepiolite (Spn) and serpentine (Sr n) mineralsphases in italics do not refer to the mentioned hydrolysis reactions
ΔGf 0 log K Reference
(kJ molminus1) (at 25 degC)
Sepiolite (Spn) mdash Mg4Si6O15(OH)2middot6H2O+8H+=4Mg2++6SiO2+11H2Ominus925400 3004 Gunnarsson et al (2005)minus925133 3050andash3172 Christ et al (1973)
(Sp2) minus921986 3602andash3756 Wollast et al (1968)
Serpentine (Srn) mdash Mg3Si2O5(OH)4+ 6H
+
=3Mg
2+
+2SiO2+ 5H2ONi-lizardite minus 289324 1680 Mondesir and Decarreau (1987)
Al-lizardite2 minus411003 2544a Caruso and Chernosky (1979) Al-lizardite minus 339854 2786 this work (see text for details)
Mg-lizardite (Sr1) minus404025 3056 Wilson et al (2006)Mg-crysotile (Sr2) minus403702 3112 bthermocomV8R6Mg-lizardite minus402824 3340 Mondesir and Decarreau (1987)p-antigorite (Sr3) minus564239a 3479 Gunnarsson et al (2005)Mg-chrysotile (Sr4) minus406809 3669 Bruni et al (2002)Mg-chrysotileasbestos minus562505a 3888 Pfeifer (1977)Antigorite minus6612864 47719 bthermocomV8R6
The 1047297eld boundaries of the underlined phases shown in Fig 6 correspond to atemperature of 11 degC
a In the ΔG f deg column value calculated using the ΔG f
deg data of the elements from llnldat
and logK value in the reference in the LogK column value calculated using ΔG f deg
(mineral) in the reference and ΔG f deg data of the elements from llnldat
b Thermodynamic database of Lawrence Livermore National Laboratory included in
PHREEQCI and Geochemists Workbench softwares
85T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
decreases The waterndashrock interaction in the TarondashCeno basin iscomplicated by the presence of secondary mixed-layer clay mineralsthe origin of which is quite dif 1047297cult to model because the solidndashstatetransformations and weathering reactions are strictly related tohydrological and climatic conditions as described above An equili-
brium-based reaction path model is a valuable tool for assessing the
most probable reactions and for de1047297ning the product phases that canbe used in a kinetics-based model Since the kinetics of precipitationof Mg-clays at room temperature and pressure are poorly knownrelative reaction rate effects in the reaction path model can be mimedby omitting the stabler phase when a metastable mineral is expected
to precipitate Therefore the precipitation of MgndashFe3+
Abbreviations are as follows Fer-6line 6-line ferrihydrite kaol kaolinite verm vermiculite magn magnetite sap saponite serp serpentine clin clinochlore pyr pyrite calccalcite hydrom hydromagnesite bruc brucite enst enstatite The chromium decrease during the change in water facies from bicarbonate to alkaline was modeled withsupersaturation (precipitation) of eskolaite and thereafter with equilibrium with chromite (Boschetti 2003)- not bound dagger chlorite mineral from solution Dagger chlorite mineral added to the paragenesis
86 T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
montmorillonites vermiculite and saponites are predicted to occurrather than that of Mg-clays Moreover in order to simulate the slowprecipitation of calcite in depth precipitation was not activated atslight supersaturation (SIcalcite=001) during the initial stages of thewaterndashrock interaction in this manner saturation index of calcite wasfree to increase in the following steps and in alkaline waters
6 Discussion of the model results
61 Explanation of pre-modeling choices and post-modeling results
The results of the reaction path modeling are reported in Table 5and shown in Figs 6 and 9 The reaction paths model the dissolutionof a typical local serpentinite (mineralogically constrained as inTable 3) and the sequential precipitation of gibbsite kaoliniteferrihydrite vermiculite Fe2+ndashMg2+-saponite and a poorly-crystal-line serpentine( p-antigorite of Gunnarsson et al2005) The1047297rst stepinvolves the reaction of the ldquoinitial waterrdquo with serpentinite atconstant temperature (11 degC) PO2 (10minus26 bar) and PCO2 (10minus30 bar)this early stage corresponds to bicarbonate groundwater feeding theinvestigated spring system At the stage of primary magnetiteequilibrium and Fe2+ndashMg2+-saponite supersaturation the modelassumes the system to be closed to CO2 and to have a 1047297xed partialpressure of oxygen (PO2= 10minus40 bar) these conditions represent
when water switch from a bicarbonate to an alkaline compositionThe calculated amount of interacting serpentinite needed to obtainthe alkaline springs of the TarondashCeno Valleys (mean PCO2= 10minus7 bar)is approximately 00203 moles (see Table 5)
Bruni et al (2002) explained the origin of the alkaline waters of the Voltri Group as the result of the evolution of meteoricbicarbonate water following the precipitation of sepiolite withabundance similar to that of goethite This might be supported bythe observation that sepiolite or smectite precipitate at lowtemperature from solutions with low Al content when the pH is inthe range 75ndash95 or N95 respectively (Siffert and Wey 1962 Jonesand Galan 1988) but sepiolites form from the weathering of serpentines in arid climate regions whereas sepiolites are rare inhumid climate regions where serpentinite changes to Fe-smectite
(Singer 1989 Wrucke 1996) In the ophiolite rocks of the Northern
Apennine sepiolite is found only sporadically in veins (Giuseppetti1953 Veniale 1966) and is unknown in the ophiolites of the Tarondash
Ceno Valleys whereas it is thought that sepiolite of the Voltri Groupprecipitated from metasomatic 1047298uids during development of shearzones (Crispini and Capponi 2002)
As described above our model explains the transition frombicarbonate water to alkaline water (UM15 and PR10 samples) asthe result of low-temperature precipitation of a poor crystalline
serpentine with the solubility product of p-antigorite (see Sr3 inTable 4 and Fig 6) The model indicates that the bicarbonate waters of the Voltri Group with the highest [Mg2+][H+]2 activity ratios similarlyevolve towards alkaline waters (Fig 6B) The logK value of thesynthetic low temperature serpentine adopted in this model and its
Fig 7 Saturation indices of chrysotile and p-antigorite (Sr2 and Sr3 thermodynamic
data respectively in Table 4) for bicarbonate and alkaline waters coming fromultrama1047297c rocks of the TarondashCeno Valleys as a function of SiMg molar ratioThermodynamic data on chrysotile and p-antigorite are from the thermocomV8R6
database and Gunnarsson et al (2005) respectively
Fig 8 Activityplot forthe MgOndashSiO2ndashAl2O3ndashFe2O3(FeO)ndashH2O system analyzed forSiO2-phases and aqueous species at different log PO2 and log[aAl3+(aH+)3] values Springwaters coming from the TarondashCeno Valleys (symbols as in Fig 2) and Voltri Groupultrama1047297tes (symbols as in Fig 5) are shown The greyarea refers to bicarbonate watersfrom theVoltri Group (Mariniand Ottonello2002Bruniet al 2002) Thesepiolite 1047297eldis very narrow or hidden by saponite-Mg (1047297eld limits shown as continuous lines)saponite-Fe3+ndashMg (1047297eldlimits shown as dottedlines) or vermiculite(1047297eldlimits shownas dottedndashdashed lines) The talc stability 1047297eld and boundaries with saponite-Mg
montmorillonite-Mg and cronstedite-Mg are shown for comparison (dashed lines)
87T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
relative uncertainty of plusmn2 (see Fig 8 in Gunnarsson et al 2005) arecomparable to the logK variation reported for other low-temperatureMg-serpentines (Table 4) that is within the lines Sr1ndash4 for serpentineminerals where samples varying from bicarbonate composition toalkaline composition in the ultrama1047297c aquifer are grouped (Fig 6)Fig 7 shows a plot of calculated serpentine saturation index as a
function of the SiMg molar ratio for the water samples from theultrama1047297tes of the TarondashCeno ValleysThe SI andSiMgvalues of pureMg-serpentine (SI= 0 SiMg=067) provide a useful index forseparating bicarbonate waters (SIb0 SiMgb067) from alkalinewaters (SIN0 SiMgN067) and the distribution of data points isconsistent with the idea that bicarbonate waters approach saturationwith serpentine until they become supersaturated as alkaline waters
Based on the Mgvs SiO2(aq) activitydiagram and speci1047297cratiosoflog[aAl3+(aH+)3] Birsoy (2002) analyzed the behavior of sepiolite duringultrama1047297te weathering The diagram produces a very narrow stability1047297eldfor sepiolite(Fig 8) when it is programmedwith suitable conditionsfor the bicarbonate waters (log PO2=minus25 bar log [aAl3+(aH+)3]= 65) andthe alkaline waters (log PO2=minus50 bar log [aAl3+(aH+)3]= 45) sampled inthe present study This stability 1047297eld grows larger when the absence of
Mg-smectites and talc is assumed but this is an unrealistic conditionbecause (i) all these minerals occur in the rocks within the areainvestigated and (ii) at high [Mg2+][H+]2 activity ratios the 1047297eld of sepiolite is smaller than that of subordinate compared to those poorly-crystalline serpentine (p-antigorite) and clinochlore
Among the alkaline springs of the Voltri Group the waters withlow log [aCa2+(aH+)2] values (Fig 9) show patterns compatible withchlorite interaction (step 5c in Fig6D)The alkaline waters whose datafall near the clinochlore-14Aring 1047297eld (Fig 6D) cannot be explained bysepiolite precipitation because the sepiolite line does not cross thechlorite 1047297elds Thus the peculiar composition of some alkaline watersof the Voltri Group is probably due to interactions between water andprimary chlorites of the ultrama1047297c rocks (optional step 5c in Table 5cb) This is supported by the occurrence sometimes in large amounts
of metasomatic Mgndash(Fe)-chlorite (Scambelluri and Rampone 1999
Vissers and Strating 2001) in the Ligurian area where the relativeabundance of chlorite can reach 50 in veins in high strainserpentinite mylonites of the ErrondashTobbio ultrama1047297tes (Scambelluriet al 2001b) Otherwise in a chlorite-free primary paragenesis thesolution evolving on the Sr3 serpentine line reaches equilibrium withclinochlore but probably the latter does not precipitate due to theslow kinetics of precipitation under these conditions Thus thedissolution of additional reacting rock in the iterative weathering
process supplies Mg
2+
to the solution thereby favoring the Sr3 to Sr4poorly crystalline serpentine transformation In fact all the PR10samples and some of the alkaline springs from the Voltri Group showincreasing [Mg2+][H+]2 activity ratio and low variation in silicaactivity (Fig 6D) starting from the invariant triple point 4a (Sr3-serpentine Fe2+ndashMg-saponite chlinoclore-14 Aring saturation) andextending towards the brucite line This path can be explained asthe result of the transformation of poorly crystalline serpentine fromSr3 to Sr4 the latter is more stable at higher [Mg2+][H+]2 activityratios Finally at very low PO2 values vermiculite disappears and thestability 1047297elds of chlorite and Fe2+ndashMg-saponite become narrowerwhereas those of Fe2+-rich phases (eg Fe2+-saponite) become widerThe data for the alkaline waters with the highest log [aMg2+(aH+)2]from the Voltri Group fall near the stability 1047297eld of peridotite primaryminerals like forsterite and enstatite (Fig 6D) therefore the last stepsof theevolution involve theequilibration of thesolutionwithenstatite(steps 8a and 8c in Table 5ab)
62 Ca-rich alkaline waters
In preliminary modeling of the evolution of the TarondashCeno springwaters Boschetti (2003) concluded that meteoric water initiallysaturated in Fe3+ndashMg2+-chlorite Fe3+ndashMg2+-saponite and pyrite (andor goethite) and then interacted with the ultrama1047297te mineralassemblage eventually evolved into the alkaline waters In themodel of Boschetti (2003) chlorite saponite and calcite were added
Fig 9 Activity plot for the MgOndashCaOndash(CO2)ndashH2O system with phase limits calculated ata temperature of 11 degCSpring waterscoming from theTarondashCeno Valleys (symbols as inFig2) andVoltri Groupultrama1047297tes(symbolsas in Fig5) areshown Calculated path forwaterndashultrama1047297c rock interaction is shown (dotted line) with the number ndashlettercombinations referring to the appearance or disappearance of minerals (see Table 5)The grey area refers to bicarbonate waters from the Voltri Group (Marini and Ottonello2002)
Fig 10 Plots of (A) calcite saturation indexes and (B) PCO2 as a function of Clminus ascalculated by PHREEQCI software for waters from the TarondashCeno Valleys Alkalinesprings for the Voltri Group are also shown (small black diamonds Marini andOttonello 2002) The grey shading is deeper(A) in the saturation1047297eld(SI= 00plusmn05) and
(B) where the lowest log PCO2 value is approached
88 T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
to the solution in order to achieve a high [Ca2+][H+]2 activity ratio andsupersaturation in these phases However in light of the work of Clarket al (1992) and Bruni et al (2002) we believe that the calcitesupersaturation in alkaline waters (Fig 9 and 10) which in our modelis achieved by preventing the precipitation of calcite in the 1047297rst stepsof the reaction path (Table 5a and b) depends on the differencebetween CO2 partial pressureat thegivendepth(PCO2le10minus10bar) andat the surface (PCO2cong10minus35) Waters rising to the surface are far from
equilibrium with the atmosphere where the PCO2 is 4times10
6
timesgreater than the value at depth thus the water rapidly take up CO2which is promptly converted to CO3
2minus because of the high pH Thisleads to supersaturation with calcite through the following hydro-xylation reaction
As a result large amounts of travertine are generated in the VoltriGroup (Marini and Ottonello 2002) and carbonate concretionsknown as devilrsquo s coins are formed at Mt Prinzera in Taro valleyBelow the surface the SIcalcite of the alkaline waters is expected to belower than what the model estimates Under conditions of lowcarbonate alkalinity low CO2 and O2 partial pressure and high pHcalcite nucleation may be inhibited and the kinetics of precipitationreduced by the high concentrations of dissolved divalent cations suchas Mg2+ Fe2+ and Mn2+ (Sumner and Grotzinger 1996 Davis et al2000 Zhang and Dawe 2000 de Leeuw 2002) The fact that severalalkaline water samples fall within or near the calcitendashdolomiteequilibrium (Fig 9) may indicate magnesium calcite or protodolomitenucleation in those extreme subsurface conditions However possibleinteraction between some calcium-rich waters of the Voltri Group andCa-rich rocks such as ophicarbonates or rodingites cannot beexcluded these rocks are widespread in the Ligurian Apennineserpentinites (Treves and Harper 1994)
63 Sepiolite vs serpentine nucleation under weathering conditions
constraints on the kinetics and equilibrium-based approaches
The kinetics of precipitation of poorly-crystalline serpentine and
sepiolite cannot be compared quantitatively because rate constantsare available only for well-ordered crystalline phases (eg see Palandriand Kharaka 2004 for crystalline serpentine) Sepiolite precipitationis often invoked to explain magnesium scavenging from solutions butthe lack of kinetic data for this mineral leads to overly simplisticreaction transport models These models are forced to use kinetic datafor phyllosilicates such as kaolinite or muscovite (eg Xu et al 2001Bryan 2005) Moreover the low-temperature thermodynamic dataeven if they are fairly uncertain show that sepiolite solubilityincreases with decreasing crystallinity (Table 4 Stoessell 1988) Infact crystalline sepiolite requires a long time to reach equilibriumwith solution which imposes a slow control on thewater composition(eg Stoessell and Hay 1978) Consequently in modeling based onreaction transport that simulates freshly precipitated or poorly
crystalline sepiolite formation the logK value of Wollast et al (1968)is preferable to the lower values (Suarez and Šimůnek 1996) Thispreference agrees with the approach of Mariner (2003) and Jove Colon(2006) who adopt a logK (25 degC) value for sepiolite of 3044 It is alsosimilar to that reported by Wollast et al (1968) and it is six logK unitshigherthan the valueused in thermocomV8R6 dataset which is oftenused (perhaps improperly) to model the supergenic nucleation of sepiolite In the stability diagram in Fig 6 where the Sp1 and Sp2 linesrepresent the stability boundaries of crystalline and poorly-crystallinesepiolite respectively the evolution of TarondashCeno waters frombicarbonate to alkaline (Figs 6A and C) follows the stability 1047297eldboundary of the poorly-crystalline serpentine (between Sr2ndashSr3 lines)rather than that of the poorly-crystalline sepiolite (Sp2) as postulatedin themodel The lines for Sp2 Sr3 and amorphoussilica intersect at a
common point (Fig 6D) which is consistent with the nucleation of
sepiolite that typically occurs in solutions that are saturated orsupersaturated with amorphous silica and (Mg)-carbonate (egStoessell and Hay 1978 Birsoy 2002) Some bicarbonate watersfrom the Voltri Group have a silica activity near the amorphous silicaboundary (Fig 6B) but only a few alkaline springs of this group fall onthe poorly-crystalline sepiolite boundary line (Sp2 Fig 6D) Both of these features are inconsistent with the possibility that sepioliteprecipitates from supergenic fresh water in the Northern Apennine
Moreover these two results suggest that values for logK(sepiolite)b
3044 (see Table 4) should be used to model the dissolutionprecipitation of the well-crystallized phase that is the most commonform for the metasomatic sepiolite found in the investigated area
7 Conclusions
The waters issuing from ultrama1047297tes in the TarondashCeno Valleysoriginate from low-temperature interactions between meteoric waterand serpentinite rocks In this report the waterndashrock interaction hasbeen modeled using reaction paths in progress mode During their earlyinteraction with soil and rock minerals bicarbonate waters super-saturated in kaolinite ferrihydrite vermiculite and Fe2+ndashMg2+-saponitewhere theorigin of the secondary iron saponite as Fe2+-bearing smectiteagrees with the hypothesis of Guumlven (1988) Then while remainingsupersaturated with saponite the bicarbonate waters evolve into analkaline facies when poorly-crystalline serpentine precipitates Duringrock weathering the Al-bearing crystalline serpentine of oceanic originprobably promotes nucleation of the low-temperature serpentineamorphous phase and supplies Al3+ to the solution for saponiteprecipitation This interpretation is consistent with the existence of Al-poor supergenic serpentine (Neal and Stanger 1985) and the partiallymeteoric OH-isotopic imprint observed in other ultrama1047297c systems
Dif 1047297culties and complexities in the modeling arise fromuncertainties in the primary mineralogy of some local ophiolitesand from the presence of different lithotypes (rodingites ophi-carbonates gabbros) which can signi1047297cantly affect the composi-tion of water samples This can be seen in the example of the Candash
OH springs issuing from the Voltri Group (PO2 and PCO2 of 10minus70
and 10minus10 bar respectively Bruni et al 2002 Marini and
Ottonello 2002) These springs appear to be more evolved thanthe TarondashCeno alkaline waters as a result of their particularly longinteractions with serpentinites until equilibrium with enstatite isreached or with chlorite-bearing serpentinite The use of anupdated thermodynamic database in our model avoided problemsdue to mineralogical heterogeneity in the mixed-layer claysassociated with ophiolitic rocks this database accurately re1047298ectedthe chemical composition of the single phases Using this modelthe vermiculitendashchlorite and saponitendashchlorite associations seen inthe activity diagrams simulate respectively the presence of highand low charge corrensite observed in the 1047297eld The accuracy of ourmodel can be tested by comparing estimates of chlorine composi-tion resulting from our model and from experimental methods The
mean Cl concentration of the bicarbonate water is 41plusmn16 mgLtaking into account a Cl concentration of sim2 mgL in rainwater inthe area (cf Panettiere et al 2000 Venturelli 2003) 2 mgL of Clin spring water are estimated to come from serpentinite This valueagrees with the model which predicts that the switch frombicarbonate water to alkaline water requires the congruentdissolution of 00203 mol of serpentinite containing 100 ppm of soluble Cl thereby yielding water containing 2 ppm of Cl
In addition we think that weathering of the accessory albiteplagioclase is responsible for Na enrichment of the waters whereas Clenrichment in the alkaline waters is predicted to result from theleaching of NaCl 1047298uid inclusions Moreover the TarondashCeno alkalinesprings show the highest BCl ratios comparable with those of thehost metamorphic minerals (Scambelluri et al 2004) In contrast the
lower boron content and BCl ratio of the alkaline springs of the Voltri
89T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
Group are unexpected features and probably re1047298ect waterndashrockinteractions over a longer time period than in the TarondashCeno ValleysThe boron content in the Northern Apennine alkaline springs may becontrolled by several factors such as (i) the different histories of VoltriGroup and EL ultrama1047297tes (ii) the probable contribution of boronfrom NaCl 1047298uid inclusions (iii) the sinking of boron by travertineprecipitation (Hobbs and Reardon1999) brucite (Rhoades et al 1970)and smectite and (iv) decreases in boron release from minerals due to
increases in pH (Su and Suarez 2004) Acknowledgements
This research was carried out with the 1047297nancial support from theldquoYoung Researchers Projectrdquo (2001ndash2002) of the Italian Ministry of Universities Scienti1047297c Research and Technology (MURST) We thankGianniCortecci (Institute of Geosciences and EarthResources IGG-CNRPisa Italy) and Giampiero Venturelli (Earth Sciences Department Univer-sity of Parma Italy) for useful suggestions and critical review of themanuscript The authors are grateful to Enrico Dinelli (InterdepartmentResearch Centre for Environmental Sciences Alma Mater StudiorumUniversity of Bologna Ravenna Italy) for granting permission to use theGeochemistrsquos WorkbenchsoftwareTheconstructive reviews of theEditor
J Fein and of two anonymous Referees have signi1047297cantly improved the
manuscript
Appendix A Supplementary data
Supplementary data associated with this article can be found inthe online version at doi101016jchemgeo200808017
References
Abbate E Bortolotti V Passerini P 1980 Olistostromes and olistoliths In Sestini G(Ed) Development of the Northern Apennines geosyncline Sedimentary Geology4 pp 521ndash558
Alietti A 1956 Il minerale a strati misti saponite-talco di monte Chiaro (val di TaroAppen Emil) Rendiconti dellAccademia Nazionale dei Lincei 21 201 ndash207
Alietti A De Angelis G Montanari G Sgarlata F Tirelli G 1976 Il ferro nelle argilleappenniniche Periodico di Mineralogia 45 51ndash64
Anselmi B Mellini M Viti C 2000 Chlorine in the Elba Monti Livornesi and Murloserpentines Evidence for seandashwater interaction European Journal of Mineralogy12137ndash146
Apha-Awwa-Wef 1995 Standard Methods for the Examination of Water and Waste-water 19th Edition American Public Health Association American Water WorksAssociation Water Environment Federation
Barnes I LaMarche Jr VC Himmelberg GR 1967 Geochemical evidence of present-day serpentinization Science 156 830ndash832
Barnes I ONeil JR 1969 The relationship between 1047298uids in some fresh alpine-typeultrama1047297cs and possible modern serpentinization Western United StatesGeological Society of America Bulletin 80 1947ndash1960
Barnes I ONeil JRTrescases JJ1978 Present dayserpentinizationin New CaledoniaOman and Yugoslavia Geochimica et Cosmochimica Acta 42 144ndash145
Barnes I Rapp JB ONeil JR Sheppard RA Gude III AJ 1972 Metamorphicassemblages and the direction of 1047298ow of metamorphic 1047298uids in four instances of serpentinization Contributions to Mineralogy and Petrology 35 263ndash276
Barnes JDSelverstone J Sharp ZD 2006Chlorine isotope chemistry of serpentinitesfrom Elba Italy as an indicator of 1047298uid source and subsequent tectonic historyGeochemistry Geophysics Geosystems 7 doi1010292006GC001296
Beccaluva L Emiliani F Venturelli G Zerbi M 1973 Ca Fe Mg Mn Cr Ni Codistribution in some ultrama1047297c rocks outcropping in the Northern Apennines withsome geological remarks Ateneo Parmense Acta Naturalia 9 69ndash98
Beccaluva L Macciotta G Piccardo GB Zeda O 1984 Petrology of lherzolitic rocksfrom the Northern Apennine ophiolites Lithos 17 299 ndash316
BeccaluvaL Venturelli G ZanzucchiG 1975 Dati geologici e geochimici sui basalti diassociazione o1047297olitica dellAppennino ligure-emiliano Ateneo Parmense ActaNaturalia 11 789ndash802
BethkeCM 2002 The Geochemists Workbench A Userrsquos Guide to RxnAct2Tact ReactandGtplot Release40 HydrogeologyProgram Universityof IllinoisUrbanaIL 236pp
Birsoy R 2002Formation of sepiolitendashpalygorskite and related minerals from solutionClays and Clay Minerals 50 736ndash745
Bonatti E Lawrence JR Morandi N 1984 Serpentinization of oceanic peridotitestemperature dependence of mineralogy and boron content Earth and PlanetaryScience Letters 70 88ndash94
Bonifacio E Zanini E Boero V Franchini-Angela M 1997 Pedogenesis in a soilcatena on serpentinite in north-western Italy Geoderma 75 33ndash51
Boschetti T 2003 Studio geochimico e geochimico-isotopico di acque a composizione
estrema e termali dellrsquoAppennino Settentrionale Ph D Thesis Parma Univ
Brigatti MF Franchini G Lugli C Medici L Poppi L Turci E 2000 Interactionsbetween aqueous chromium solutions and layer silicatesApplied Geochemistry 151307ndash1316
Brigatti MF Lugli C Poppi L Venturelli G 1999 Iron-rich saponite dissolutionreactions and Cr uptake Clay Minerals 34 637ndash645
Brigatti MF Valdregrave G 1996 Mixed-layer chloritesmectite (corrensite) in ophioliticrock alteration products Neues Jahrbuch fuer Mineralogie Monatshefte 1 37 ndash47
Bruni J Canepa M Chiodini G Cioni R Cipolli F Longinelli A Marini L OttonelloG Vetuschi Zuccolini M 2002 Irreversible waterndashrock mass transfer accompany-ing the generation of the neutral MgndashHCO3 and high-pH CandashOH spring waters of the Genova province Italy Applied Geochemistry 17 455ndash474
Bryan CR 2005 Drift-Scale THC Seepage Model MDL-NBS-HS-000001 REV 04 YuccaMountain Project Las Vegas Nevada (US)Caruso LJ Chernosky JV 1979 The stability of lizardite Canadian Mineralogist 17
757ndash769Clark ID Fontes JC Fritz P 1992 Stable isotope disequilibria in travertine from high
pH waters Laboratory investigations and 1047297eld observations from Oman Geochi-mica et Cosmochimica Acta 56 2041ndash2050
Cortesogno L 1980 Il metamor1047297smo giurassico nelle o1047297oliti dellAppenninoSettentrionale due distinti cicli metamor1047297ci in ambiente oceanico O1047297oliti 5 5ndash18
Christ CLHostetler PB Siebert RM1973 Studies in the system MgO-SiO2-CO2-H2OIII American Journal of Science 273 65ndash83
Cortesogno L Lucchetti G 1982 Il metamor1047297smo oceanico nei gabbri o1047297oliticidellAppennino Ligure Aspetti mineralogici e petrogenetici Rendiconti dellaSocietagrave Mineralogica Italiana 38 561ndash579
Crispini L Capponi G 2002 Control of metasomatic alteration on the development of shear zone in ultrama1047297c rocks a case study from the Voltri Massif (Ligurian Alps)Architettura interna e cinematica delle zone di taglio Gruppo Italiano di GeologiaStrutturale mdash Riunione Annuale 2002 Pisa Italy 11ndash12 giugno 2002 XIV ndashXV
CrissRE 1999 Principlesof stableisotope distributionOxford University PressNew York
Davis KJ Dove PM De Yoreo JJ 2000 The role of Mg 2+ as an impurity in calcitegrowth Science 290 1134ndash1137
de Leeuw NH 2002 Moleculardynamics simulations of the growth inhibiting effect of Fe2+ Mg2+ Cd2+ and Sr2+ on calcite crystal growth Journal of Physical Chemistry B106 5241ndash5249
De Nardo MT Segadelli S Vescovi P 2007 Studio pilota per la caratterizzazionegeologicadellesorgentinellazona delM Nero (altaVal Cenoe alta ValNuremdash provincedi Parma e Piacenza) Il Geologo dellEmilia-Romagna 25 Nuova Serie pp 5 ndash21
Di Giulio A Geddo G 1990 Studio petrogra1047297co delle Arenarie di Casanova (alta valTrebbia Appennino settentrionale) Atti Ticinesi di Scienze della Terra 33 243ndash254
Dinelli E Lombini A Simoni A Ferrari C 1997 Heavy metals in serpentinite soils of selected outcrops of Piacenza and Parma provinces (Northern Apennines Italy)Mineralogica et Petrographica Acta 40 241ndash255
Epstein S Mayeda T 1953 Variation of 18O content of waters from natural sourcesGeochimica et Cosmochimica Acta 4 213ndash224
Ernst WG Piccardo GB 1979 Petrogenesis of some Ligurian peridotites I Mineraland bulk-rock chemistry Geochimica et Cosmochimica Acta 43 219ndash237
Feder F Trolard F Klingelhofer G Bourrieacute G 2005 In situ Moumlssbauer spectroscopy
evidence for green rust (fougerite) in a gleysol and its mineralogical transforma-tions with time and depth Geochimica et Cosmochimica Acta 69 4463ndash4483Fritz PClark IDFontes JCWhiticar MJFaberE 1992 Deuterium and
13C evidence
for low temperature production of hydrogen and methane in a highly alkalinegroundwater environment in Oman In Kharaka TK Maest AS (Eds) Proceed-ings of the 7th International Symposium on WaterndashRock Interaction BalkemaRotterdam pp 793ndash796
Giammetti F 1968 Le o1047297oliti di Roccaprebalza Costa della Guardia e Groppo di Gorro(Appennino Parmense) Ateneo Parmense Acta Naturalia 4 107ndash142
Giuseppetti G1953 Roccee minerali dellaformazioneo1047297olitica di Volpedo Rendicontidella Societagrave Mineralogica Italiana 9 85ndash134
Gran G 1952 Determination of equivalence point in the potentiometric titrationAnalyst 77 661ndash671
Gunnarsson I Arnorsson S Jakobsson S 2005 Precipitation of poorly crystalline antigoriteunder hydrothermal conditions Geochimica et Cosmochimica Acta 69 2813ndash2828
Guumlven N 1988 Smetites In Bailey SW (Ed) Reviews in Mineralogy 19 Hydrousphyllosilicates (exclusive of micas) Mineralogical Society of America MichiganBook Crafters Inc 497ndash559
Hansel CM Benner SG Fendorf S 2005 Competing Fe(II)-induced mineralization
pathways of ferrihydrite Environmental Science and Technology 39 7147ndash7153HobbsMYReardonEJ 1999 Effect of pH on boronco-precipitation by calcite further
evidence for non-equilibrium partitioning of trace elements Geochimica etCosmochimica Acta 63 1013ndash1021
Jones BF Galan E 1988 Sepiolite and palygorskite In Bailey SW (Ed) HydrousPhyllosilicates (exclusive of Micas) Reviews in Mineralogy vol 19 MineralogicalSociety of America Michigan Book Crafters Inc 631ndash674
Jove Colon CF 2006 Technical work plan for Thermodynamic database for chemicalmodelling TWP-MGR-PA-000039 REV 01 Yucca Mountain Project Las VegasNevada (US)
Kohls DW Rodda Jr JL 1967 Iowaite a new hydrous magnesium hydroxidendashferricoxychloride from the Precambrian of Iowa American Mineralogist 52 1261 ndash1271
Kyser TK OHanley DS Wicks FJ 1999 The origin of 1047298uids associated withserpentinization processes evidence from stable-isotope compositions CanadianMineralogist 37 223ndash237
Lapham DM 1961 New data on deweylite American Mineralogist 46 168ndash188Lee BD SearsSKGraham RC Amrhein CValiH 2003 Secondarymineral genesis
from chlorite and serpentine in an ultrama1047297c soil toposequence Soil Science
Society of America Journal 671309ndash1317
90 T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
Longinelli A Selmo E 2003 Isotopic composition of precipitation in Italy a 1047297rstoverall map Journal of Hydrology 270 75ndash88
Majzlan J Navrotsky A Schwertmann U 2004 Thermodynamics of iron oxides PartIIIEnthalpiesof formation and stability of ferrihydrite (simFe(OH)3) schwertmannite(simFeO(OH)34(SO4)18) and ɛ-Fe2O3 Geochimica et Cosmochimica Acta 681049ndash1059
Mariner P 2003 In-Drift PrecipitatesSalts Model ANL-EBS-MD-000045 REV 01 YuccaMountain Project Las Vegas Nevada (US)
Marini L Ottonello G 2002 Atlante degli acquiferi della Liguria Volume III Le acquedeicomplessio1047297olitici (bacini ArrestraBranega Cassinelle Cerusa ErroGorzenteLeira Lemme Lerone Orba Piota Polcevera Rumaro Sansobbia Stura Teiro
Varenna Visone) Pacini PisaMellini L Zanazzi PF 1987 Crystal structures of lizardite-l T and lizardite-2H1 fromColi Italy American Mineralogist 72 943ndash948
Mezzadri G 1964 Petrogra1047297a delle arenarie di Ostia Rendiconti Societagrave MineralogicaItaliana 20 192ndash228
Mondesir H Decarreau A 1987 Synthesegrave entre 25 et 200 degC de lizardites Ni ndashMgMesure des coef 1047297cients de partage solide-solution aqueuse pour le couple Ni ndashMgdans des lizardites Bulletin de Mineralogie 110 409ndash426
Neal C Shand P 2002 Spring and surface water quality of the Cyprus ophiolitesHydrology and Earth System Sciences 6 797ndash817
Neal C Stanger G 1983 Hydrogen generation from mantle source rocks in OmanEarth and Planetary Science Letters 66 315ndash320
Neal C Stanger G 1984 Calcium and magnesium hydroxide precipitation fromalkaline groundwaters in Oman and their signi1047297cance to the process of serpentinization Mineralogical Magazine 48 237ndash241
Neal C Stanger G 1985 Past and present serpentinisation of ultrama1047297c rocks anexample from Semail ophiolite nappe of Northern Oman In Drever JI (Ed) TheChemistry of Weathering D Reidel Publishing Company pp 249ndash275
Nesbitt HW Bricker OP 1978 Low temperature alteration processes affecting
et Cosmochimica Acta 41 1835ndash1841OrsquoHanley DS 1996 Serpentinites records of tectonic and petrological history Oxford
University PressOrberger B Friedrich G Woermann E 1990 The distribution of halogens and carbon
in Pge-bearing ultrama1047297cs of the Acoje Ophiolite Block Zambales Philippines Journal of Geochemical Exploration 37 147ndash169
Ottonello G Joron JL Piccardo GB 1984 Rare Earth and 3d transition elementsgeochemistry of peridotitic rocks II Ligurian peridotites and associated basalts Journal of Petrology 25 373ndash393
Palandri JL Kharaka YK 2004 A compilation of rate parameters of water-mineralinteraction kinetics for application to geochemical modeling Open File Report2004-1068 US Geological Survey
Panettiere P Cortecci G Dinelli E Bencini A Guidi M 2000 Chemistry and sulfurisotopic composition of precipitation at Bologna Italy Applied Geochemistry 151455ndash1467
Parkhurst DL Appelo CAJ 1999 Userrsquos guide to PHREEQC (version 2) mdash a computerprogram for speciation batch-reaction one-dimensional transport and inversegeochemical calculation US Department of the Interior US Geological SurveyWashington DC
Pfeifer HR 1977 A model for 1047298uids in metamorphosed ultrama1047297c rocks Observationsat surface and subsurface conditions (high pH spring waters) SchweizerischeMineralogische und Petrographische Mitteilungen 57 361ndash396
PlummerLN1984Geochemical modeling A comparisonof forwardand inverse methodsIn Hitchon B Wallick EI (Eds) Proceedings First CanadianAmerican Conference onHydrogeology-Practical Applications of Ground Water Geochemistry Banff AlbertaCanada mdash Worthington Ohio National Water Well Association pp 149ndash177
Rampone E Hofmann AW Piccardo GB Vannucci R Bottazzi P Ottolini L 1995Petrology mineral and isotope geochemistry of the External Ligurides peridotites(Northern Apennines Italy) Journal of Petrology 36 81ndash105
Rampone E Hofmann AW Piccardo GB Vannucci R Bottazzi P Ottolini L 1996Trace element and isotope geochemistry of depleted peridotites from an N-MORBtype ophiolite (Internal Liguride N Italy) Contributions to Mineralogy andPetrology 123 61ndash76
Rhoades JD Ingvalson RD HatcherJT1970Adsorption of boronby ferromagnesianminerals and magnesium hydroxide Soil Science Society of America Journal 34
938ndash941Saini-EidukatB KuchaH Keppler H1994Hibbingiteγ-Fe2(OH)3Clanewmineralfromthe Duluth Complex Minnesota with implications for the oxidation of Fe-bearingcompounds and the transport of metals American Mineralogist 79 555ndash561
Sanford RF 1981 Mineralogical and chemical effects of hydration reactions andapplications to serpentinization American Mineralogist 66 290ndash297
Scambelluri M Muumlntener O Ottolini L Pettke TT Vannucci R 2004 The fate of BCl and Li in the subducted oceanic mantle and in the antigorite breakdown 1047298uidsEarth and Planetary Science Letters 222 217ndash234
Scambelluri M Rampone E 1999 Mg-metasomatism of oceanic gabbros and itscontrol on Ti-clinohumite formation during eclogitization Contributions toMineralogy and Petrology 135 1ndash17
Scambelluri M Bottazzi P Trommsdorff V Vannucci R Hermann J Gogravemez-Pugnaire MT Logravepez-Sagravenchez Vizcaino V 2001a Incompatible element-rich1047298uidsreleased by antigorite breakdown in deeply subducted mantle Earth and PlanetaryScience Letters 192 457ndash470
Scambelluri M Rampone E Piccardo G 2001b Fluid and element cycling insubducted serpentinite a trace-element study of the ErrondashTobbio high-pressureultrama1047297tes (Western Alps NW Italy) Journal of Petrology 42 55ndash67
Segadelli S2006 Lageologia nelpaesaggio le rupi o1047297olitiche in Val Taro e Val Ceno IlGeologo dellrsquoEmilia-Romagna 22 15ndash29 Nuova Serie
Settembre Blundo D Martin De Vidales JL Alietti A 1992 Corrensite-type clayminerals from Taro Valley (North Italy) Mineralogica et Petrographica Acta 35215ndash222
Sharp ZD Barnes JD 2004 Water-soluble chlorides in massive sea1047298oor serpenti-nites a source of chloride in subduction zones Earth and Planetary Science Letters226 243ndash254
Siever R Woodford N 1979 Dissolution kinetics and the weathering of ma1047297cminerals Geochimica et Cosmochimica Acta 43 717ndash724Singer A 1989 Palygorskite and sepiolite group minerals In Dixon JB Weed SB
(Eds) Minerals in Soil Environments Soil Science Society of America MadisonWisconsin pp 829ndash872
Siffert B Wey R 1962 Synthese d rsquoune sepiolite a temperature ordinarie Comptesrendus hebdomadaires des seacuteances de lAcadeacutemie des Sciences de Paris 2541460ndash1463
Środoń J 1999 Nature of mixed-layer clays and mechanisms of their formation andalteration Annual Reviews of Earth and Planetary Science 27 19ndash53
Stoessell RK 1988 25 degC and 1 atm dissolution experiments of sepiolite and keroliteGeochimica et Cosmochimica Acta 52 365ndash374
Stoessell RK Hay RL 1978 The geochemical origin of Sepiolite and keorilite atAmboseli Kenya Contributions to Mineralogy and Petrology 65 255ndash267
Suarez DL Šimůnek J 1996 Solute transport modeling under variably saturatedwater 1047298ow conditions In Lichtner PC Steefel CI Oelkers EH (Eds) ReactiveTransport in Porous Media Reviews in Mineralogy vol 34 Mineralogical Society of America Michigan Book Crafters Inc 229ndash268
Sumner DY Grotzinger JP 1996 Were kinetics of Archean calcium carbonate
precipitation related to oxygen concentration Geology 24 119ndash122Su C Suarez DL 2004 Boron release from weathering of illites serpentine shales
and illiticpalygorskitic soils Soil Science Society of America Journal 68 96 ndash105Toscani L Venturelli G Boschetti T 2001 Sul1047297de-free and sul1047297de-bearing waters in
the Northern Apennines Italy general features and waterndashrock interaction AquaticGeochemistry 7 195ndash216
Treves BE Harper GD 1994 Exposure of serpentinites on the ocean 1047298oor sequenceof faulting and hydrofracturing in the Northern Apennine ophicalcites O1047297oliti 19b435ndash466
Varadachari C Kudrat M Gosh K 1994 Evaluation of standard free energies of formation of clay minerals by an improved regression method Clays and ClayMinerals 42 298ndash307
Veniale F 1966 Sepiolite in sedimenti dellrsquoAppennino Vogherese (Pavia) Periodico diMineralogia 35 343ndash386
Venturelli G 2003 Acque minerali ed ambiente mdash Fondamenti di geochimica deiprocessi di bassa temperatura Pitagora Editrice Bologna
Venturelli G Contini S Bonazzi A Mangia A 1997 Weathering of ultrama1047297c rocksand element mobility at Mt Prinzera Northern Apennines Italy MineralogicalMagazine 61 765ndash778
Venturelli G Frey M 1977 Anchizone metamorphism in sedimentary sequences of the Northern Apennines Rendiconti Societagrave Mineralogica Italiana 33 109ndash123
Vieillard P 2002 A new method for the prediction of Gibbs free energies of formationofphyllosilicates (10 Aring and14 Aring)basedon theelectronegativityscale Claysand ClayMinerals 50 352ndash363
Vissers RLM Strating EHH 2001 Structures and microstructures in a thrust-relatedgreenschist facies tectonic melange Voltri Group (NW Italy) O1047297oliti 26 33ndash46
Viti C Mellini M 1997 Contrasting chemical compositions in associated lizardite andchrysotile in veins from Elba Italy European Journal of Mineralogy 9 585 ndash596
Viti C Mellini M 1998 Mesh textures and bastites in the Elba retrogradeserpentinites European Journal of Mineralogy 10 1341ndash1359
Wenner DB Taylor Jr HP 1974 Deuteriumatomic hydrogen and oxygen-18oxygen-16 studies of serpentinization of ultrama1047297c rocks Geochimica et CosmochimicaActa 38 1255ndash1286
Williams LA Parks GA Crerar DA 1985 Silica diagenesis I Solubility controls Journal of S edimentary Petrology 55 301ndash311
Wilson J Savage D Cuadros J Shibata M RagnarsdottirKV 2006 The effect of ironon montmorillonite stability (I) Background and thermodynamic considerations
Geochimica et Cosmochimica Acta 70 306ndash322WollastR MackenzieFTBricker OP1968 Experimental precipitation and genesis of sepiolite at earth-surface conditions American Mineralogist 531 1645ndash1662
Wrucke CT1996 Serpentine and carbonated-hosted asbestos deposits InEdwardAdu Bray (Eds) Preliminary compilation of descriptive geoenvironmental mineraldeposit models 95ndash831 US Geological Survey Open-File Report pp 39 ndash46
Xu T Apps JA Pruess K 2001 Analysis of mineral trapping for CO2 disposal in deepaquifers Lawrence Berkeley National Laboratory University of California
ZhangYDaweRA2000 In1047298uence ofMg2+ on thekineticsof calcite precipitation andcalcite crystal morphology Chemical Geology 163 129ndash138
Zhu C Anderson G 2002 Environmental applications of geochemical modelingCambridge University Press
91T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
decreases The waterndashrock interaction in the TarondashCeno basin iscomplicated by the presence of secondary mixed-layer clay mineralsthe origin of which is quite dif 1047297cult to model because the solidndashstatetransformations and weathering reactions are strictly related tohydrological and climatic conditions as described above An equili-
brium-based reaction path model is a valuable tool for assessing the
most probable reactions and for de1047297ning the product phases that canbe used in a kinetics-based model Since the kinetics of precipitationof Mg-clays at room temperature and pressure are poorly knownrelative reaction rate effects in the reaction path model can be mimedby omitting the stabler phase when a metastable mineral is expected
to precipitate Therefore the precipitation of MgndashFe3+
Abbreviations are as follows Fer-6line 6-line ferrihydrite kaol kaolinite verm vermiculite magn magnetite sap saponite serp serpentine clin clinochlore pyr pyrite calccalcite hydrom hydromagnesite bruc brucite enst enstatite The chromium decrease during the change in water facies from bicarbonate to alkaline was modeled withsupersaturation (precipitation) of eskolaite and thereafter with equilibrium with chromite (Boschetti 2003)- not bound dagger chlorite mineral from solution Dagger chlorite mineral added to the paragenesis
86 T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
montmorillonites vermiculite and saponites are predicted to occurrather than that of Mg-clays Moreover in order to simulate the slowprecipitation of calcite in depth precipitation was not activated atslight supersaturation (SIcalcite=001) during the initial stages of thewaterndashrock interaction in this manner saturation index of calcite wasfree to increase in the following steps and in alkaline waters
6 Discussion of the model results
61 Explanation of pre-modeling choices and post-modeling results
The results of the reaction path modeling are reported in Table 5and shown in Figs 6 and 9 The reaction paths model the dissolutionof a typical local serpentinite (mineralogically constrained as inTable 3) and the sequential precipitation of gibbsite kaoliniteferrihydrite vermiculite Fe2+ndashMg2+-saponite and a poorly-crystal-line serpentine( p-antigorite of Gunnarsson et al2005) The1047297rst stepinvolves the reaction of the ldquoinitial waterrdquo with serpentinite atconstant temperature (11 degC) PO2 (10minus26 bar) and PCO2 (10minus30 bar)this early stage corresponds to bicarbonate groundwater feeding theinvestigated spring system At the stage of primary magnetiteequilibrium and Fe2+ndashMg2+-saponite supersaturation the modelassumes the system to be closed to CO2 and to have a 1047297xed partialpressure of oxygen (PO2= 10minus40 bar) these conditions represent
when water switch from a bicarbonate to an alkaline compositionThe calculated amount of interacting serpentinite needed to obtainthe alkaline springs of the TarondashCeno Valleys (mean PCO2= 10minus7 bar)is approximately 00203 moles (see Table 5)
Bruni et al (2002) explained the origin of the alkaline waters of the Voltri Group as the result of the evolution of meteoricbicarbonate water following the precipitation of sepiolite withabundance similar to that of goethite This might be supported bythe observation that sepiolite or smectite precipitate at lowtemperature from solutions with low Al content when the pH is inthe range 75ndash95 or N95 respectively (Siffert and Wey 1962 Jonesand Galan 1988) but sepiolites form from the weathering of serpentines in arid climate regions whereas sepiolites are rare inhumid climate regions where serpentinite changes to Fe-smectite
(Singer 1989 Wrucke 1996) In the ophiolite rocks of the Northern
Apennine sepiolite is found only sporadically in veins (Giuseppetti1953 Veniale 1966) and is unknown in the ophiolites of the Tarondash
Ceno Valleys whereas it is thought that sepiolite of the Voltri Groupprecipitated from metasomatic 1047298uids during development of shearzones (Crispini and Capponi 2002)
As described above our model explains the transition frombicarbonate water to alkaline water (UM15 and PR10 samples) asthe result of low-temperature precipitation of a poor crystalline
serpentine with the solubility product of p-antigorite (see Sr3 inTable 4 and Fig 6) The model indicates that the bicarbonate waters of the Voltri Group with the highest [Mg2+][H+]2 activity ratios similarlyevolve towards alkaline waters (Fig 6B) The logK value of thesynthetic low temperature serpentine adopted in this model and its
Fig 7 Saturation indices of chrysotile and p-antigorite (Sr2 and Sr3 thermodynamic
data respectively in Table 4) for bicarbonate and alkaline waters coming fromultrama1047297c rocks of the TarondashCeno Valleys as a function of SiMg molar ratioThermodynamic data on chrysotile and p-antigorite are from the thermocomV8R6
database and Gunnarsson et al (2005) respectively
Fig 8 Activityplot forthe MgOndashSiO2ndashAl2O3ndashFe2O3(FeO)ndashH2O system analyzed forSiO2-phases and aqueous species at different log PO2 and log[aAl3+(aH+)3] values Springwaters coming from the TarondashCeno Valleys (symbols as in Fig 2) and Voltri Groupultrama1047297tes (symbols as in Fig 5) are shown The greyarea refers to bicarbonate watersfrom theVoltri Group (Mariniand Ottonello2002Bruniet al 2002) Thesepiolite 1047297eldis very narrow or hidden by saponite-Mg (1047297eld limits shown as continuous lines)saponite-Fe3+ndashMg (1047297eldlimits shown as dottedlines) or vermiculite(1047297eldlimits shownas dottedndashdashed lines) The talc stability 1047297eld and boundaries with saponite-Mg
montmorillonite-Mg and cronstedite-Mg are shown for comparison (dashed lines)
87T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
relative uncertainty of plusmn2 (see Fig 8 in Gunnarsson et al 2005) arecomparable to the logK variation reported for other low-temperatureMg-serpentines (Table 4) that is within the lines Sr1ndash4 for serpentineminerals where samples varying from bicarbonate composition toalkaline composition in the ultrama1047297c aquifer are grouped (Fig 6)Fig 7 shows a plot of calculated serpentine saturation index as a
function of the SiMg molar ratio for the water samples from theultrama1047297tes of the TarondashCeno ValleysThe SI andSiMgvalues of pureMg-serpentine (SI= 0 SiMg=067) provide a useful index forseparating bicarbonate waters (SIb0 SiMgb067) from alkalinewaters (SIN0 SiMgN067) and the distribution of data points isconsistent with the idea that bicarbonate waters approach saturationwith serpentine until they become supersaturated as alkaline waters
Based on the Mgvs SiO2(aq) activitydiagram and speci1047297cratiosoflog[aAl3+(aH+)3] Birsoy (2002) analyzed the behavior of sepiolite duringultrama1047297te weathering The diagram produces a very narrow stability1047297eldfor sepiolite(Fig 8) when it is programmedwith suitable conditionsfor the bicarbonate waters (log PO2=minus25 bar log [aAl3+(aH+)3]= 65) andthe alkaline waters (log PO2=minus50 bar log [aAl3+(aH+)3]= 45) sampled inthe present study This stability 1047297eld grows larger when the absence of
Mg-smectites and talc is assumed but this is an unrealistic conditionbecause (i) all these minerals occur in the rocks within the areainvestigated and (ii) at high [Mg2+][H+]2 activity ratios the 1047297eld of sepiolite is smaller than that of subordinate compared to those poorly-crystalline serpentine (p-antigorite) and clinochlore
Among the alkaline springs of the Voltri Group the waters withlow log [aCa2+(aH+)2] values (Fig 9) show patterns compatible withchlorite interaction (step 5c in Fig6D)The alkaline waters whose datafall near the clinochlore-14Aring 1047297eld (Fig 6D) cannot be explained bysepiolite precipitation because the sepiolite line does not cross thechlorite 1047297elds Thus the peculiar composition of some alkaline watersof the Voltri Group is probably due to interactions between water andprimary chlorites of the ultrama1047297c rocks (optional step 5c in Table 5cb) This is supported by the occurrence sometimes in large amounts
of metasomatic Mgndash(Fe)-chlorite (Scambelluri and Rampone 1999
Vissers and Strating 2001) in the Ligurian area where the relativeabundance of chlorite can reach 50 in veins in high strainserpentinite mylonites of the ErrondashTobbio ultrama1047297tes (Scambelluriet al 2001b) Otherwise in a chlorite-free primary paragenesis thesolution evolving on the Sr3 serpentine line reaches equilibrium withclinochlore but probably the latter does not precipitate due to theslow kinetics of precipitation under these conditions Thus thedissolution of additional reacting rock in the iterative weathering
process supplies Mg
2+
to the solution thereby favoring the Sr3 to Sr4poorly crystalline serpentine transformation In fact all the PR10samples and some of the alkaline springs from the Voltri Group showincreasing [Mg2+][H+]2 activity ratio and low variation in silicaactivity (Fig 6D) starting from the invariant triple point 4a (Sr3-serpentine Fe2+ndashMg-saponite chlinoclore-14 Aring saturation) andextending towards the brucite line This path can be explained asthe result of the transformation of poorly crystalline serpentine fromSr3 to Sr4 the latter is more stable at higher [Mg2+][H+]2 activityratios Finally at very low PO2 values vermiculite disappears and thestability 1047297elds of chlorite and Fe2+ndashMg-saponite become narrowerwhereas those of Fe2+-rich phases (eg Fe2+-saponite) become widerThe data for the alkaline waters with the highest log [aMg2+(aH+)2]from the Voltri Group fall near the stability 1047297eld of peridotite primaryminerals like forsterite and enstatite (Fig 6D) therefore the last stepsof theevolution involve theequilibration of thesolutionwithenstatite(steps 8a and 8c in Table 5ab)
62 Ca-rich alkaline waters
In preliminary modeling of the evolution of the TarondashCeno springwaters Boschetti (2003) concluded that meteoric water initiallysaturated in Fe3+ndashMg2+-chlorite Fe3+ndashMg2+-saponite and pyrite (andor goethite) and then interacted with the ultrama1047297te mineralassemblage eventually evolved into the alkaline waters In themodel of Boschetti (2003) chlorite saponite and calcite were added
Fig 9 Activity plot for the MgOndashCaOndash(CO2)ndashH2O system with phase limits calculated ata temperature of 11 degCSpring waterscoming from theTarondashCeno Valleys (symbols as inFig2) andVoltri Groupultrama1047297tes(symbolsas in Fig5) areshown Calculated path forwaterndashultrama1047297c rock interaction is shown (dotted line) with the number ndashlettercombinations referring to the appearance or disappearance of minerals (see Table 5)The grey area refers to bicarbonate waters from the Voltri Group (Marini and Ottonello2002)
Fig 10 Plots of (A) calcite saturation indexes and (B) PCO2 as a function of Clminus ascalculated by PHREEQCI software for waters from the TarondashCeno Valleys Alkalinesprings for the Voltri Group are also shown (small black diamonds Marini andOttonello 2002) The grey shading is deeper(A) in the saturation1047297eld(SI= 00plusmn05) and
(B) where the lowest log PCO2 value is approached
88 T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
to the solution in order to achieve a high [Ca2+][H+]2 activity ratio andsupersaturation in these phases However in light of the work of Clarket al (1992) and Bruni et al (2002) we believe that the calcitesupersaturation in alkaline waters (Fig 9 and 10) which in our modelis achieved by preventing the precipitation of calcite in the 1047297rst stepsof the reaction path (Table 5a and b) depends on the differencebetween CO2 partial pressureat thegivendepth(PCO2le10minus10bar) andat the surface (PCO2cong10minus35) Waters rising to the surface are far from
equilibrium with the atmosphere where the PCO2 is 4times10
6
timesgreater than the value at depth thus the water rapidly take up CO2which is promptly converted to CO3
2minus because of the high pH Thisleads to supersaturation with calcite through the following hydro-xylation reaction
As a result large amounts of travertine are generated in the VoltriGroup (Marini and Ottonello 2002) and carbonate concretionsknown as devilrsquo s coins are formed at Mt Prinzera in Taro valleyBelow the surface the SIcalcite of the alkaline waters is expected to belower than what the model estimates Under conditions of lowcarbonate alkalinity low CO2 and O2 partial pressure and high pHcalcite nucleation may be inhibited and the kinetics of precipitationreduced by the high concentrations of dissolved divalent cations suchas Mg2+ Fe2+ and Mn2+ (Sumner and Grotzinger 1996 Davis et al2000 Zhang and Dawe 2000 de Leeuw 2002) The fact that severalalkaline water samples fall within or near the calcitendashdolomiteequilibrium (Fig 9) may indicate magnesium calcite or protodolomitenucleation in those extreme subsurface conditions However possibleinteraction between some calcium-rich waters of the Voltri Group andCa-rich rocks such as ophicarbonates or rodingites cannot beexcluded these rocks are widespread in the Ligurian Apennineserpentinites (Treves and Harper 1994)
63 Sepiolite vs serpentine nucleation under weathering conditions
constraints on the kinetics and equilibrium-based approaches
The kinetics of precipitation of poorly-crystalline serpentine and
sepiolite cannot be compared quantitatively because rate constantsare available only for well-ordered crystalline phases (eg see Palandriand Kharaka 2004 for crystalline serpentine) Sepiolite precipitationis often invoked to explain magnesium scavenging from solutions butthe lack of kinetic data for this mineral leads to overly simplisticreaction transport models These models are forced to use kinetic datafor phyllosilicates such as kaolinite or muscovite (eg Xu et al 2001Bryan 2005) Moreover the low-temperature thermodynamic dataeven if they are fairly uncertain show that sepiolite solubilityincreases with decreasing crystallinity (Table 4 Stoessell 1988) Infact crystalline sepiolite requires a long time to reach equilibriumwith solution which imposes a slow control on thewater composition(eg Stoessell and Hay 1978) Consequently in modeling based onreaction transport that simulates freshly precipitated or poorly
crystalline sepiolite formation the logK value of Wollast et al (1968)is preferable to the lower values (Suarez and Šimůnek 1996) Thispreference agrees with the approach of Mariner (2003) and Jove Colon(2006) who adopt a logK (25 degC) value for sepiolite of 3044 It is alsosimilar to that reported by Wollast et al (1968) and it is six logK unitshigherthan the valueused in thermocomV8R6 dataset which is oftenused (perhaps improperly) to model the supergenic nucleation of sepiolite In the stability diagram in Fig 6 where the Sp1 and Sp2 linesrepresent the stability boundaries of crystalline and poorly-crystallinesepiolite respectively the evolution of TarondashCeno waters frombicarbonate to alkaline (Figs 6A and C) follows the stability 1047297eldboundary of the poorly-crystalline serpentine (between Sr2ndashSr3 lines)rather than that of the poorly-crystalline sepiolite (Sp2) as postulatedin themodel The lines for Sp2 Sr3 and amorphoussilica intersect at a
common point (Fig 6D) which is consistent with the nucleation of
sepiolite that typically occurs in solutions that are saturated orsupersaturated with amorphous silica and (Mg)-carbonate (egStoessell and Hay 1978 Birsoy 2002) Some bicarbonate watersfrom the Voltri Group have a silica activity near the amorphous silicaboundary (Fig 6B) but only a few alkaline springs of this group fall onthe poorly-crystalline sepiolite boundary line (Sp2 Fig 6D) Both of these features are inconsistent with the possibility that sepioliteprecipitates from supergenic fresh water in the Northern Apennine
Moreover these two results suggest that values for logK(sepiolite)b
3044 (see Table 4) should be used to model the dissolutionprecipitation of the well-crystallized phase that is the most commonform for the metasomatic sepiolite found in the investigated area
7 Conclusions
The waters issuing from ultrama1047297tes in the TarondashCeno Valleysoriginate from low-temperature interactions between meteoric waterand serpentinite rocks In this report the waterndashrock interaction hasbeen modeled using reaction paths in progress mode During their earlyinteraction with soil and rock minerals bicarbonate waters super-saturated in kaolinite ferrihydrite vermiculite and Fe2+ndashMg2+-saponitewhere theorigin of the secondary iron saponite as Fe2+-bearing smectiteagrees with the hypothesis of Guumlven (1988) Then while remainingsupersaturated with saponite the bicarbonate waters evolve into analkaline facies when poorly-crystalline serpentine precipitates Duringrock weathering the Al-bearing crystalline serpentine of oceanic originprobably promotes nucleation of the low-temperature serpentineamorphous phase and supplies Al3+ to the solution for saponiteprecipitation This interpretation is consistent with the existence of Al-poor supergenic serpentine (Neal and Stanger 1985) and the partiallymeteoric OH-isotopic imprint observed in other ultrama1047297c systems
Dif 1047297culties and complexities in the modeling arise fromuncertainties in the primary mineralogy of some local ophiolitesand from the presence of different lithotypes (rodingites ophi-carbonates gabbros) which can signi1047297cantly affect the composi-tion of water samples This can be seen in the example of the Candash
OH springs issuing from the Voltri Group (PO2 and PCO2 of 10minus70
and 10minus10 bar respectively Bruni et al 2002 Marini and
Ottonello 2002) These springs appear to be more evolved thanthe TarondashCeno alkaline waters as a result of their particularly longinteractions with serpentinites until equilibrium with enstatite isreached or with chlorite-bearing serpentinite The use of anupdated thermodynamic database in our model avoided problemsdue to mineralogical heterogeneity in the mixed-layer claysassociated with ophiolitic rocks this database accurately re1047298ectedthe chemical composition of the single phases Using this modelthe vermiculitendashchlorite and saponitendashchlorite associations seen inthe activity diagrams simulate respectively the presence of highand low charge corrensite observed in the 1047297eld The accuracy of ourmodel can be tested by comparing estimates of chlorine composi-tion resulting from our model and from experimental methods The
mean Cl concentration of the bicarbonate water is 41plusmn16 mgLtaking into account a Cl concentration of sim2 mgL in rainwater inthe area (cf Panettiere et al 2000 Venturelli 2003) 2 mgL of Clin spring water are estimated to come from serpentinite This valueagrees with the model which predicts that the switch frombicarbonate water to alkaline water requires the congruentdissolution of 00203 mol of serpentinite containing 100 ppm of soluble Cl thereby yielding water containing 2 ppm of Cl
In addition we think that weathering of the accessory albiteplagioclase is responsible for Na enrichment of the waters whereas Clenrichment in the alkaline waters is predicted to result from theleaching of NaCl 1047298uid inclusions Moreover the TarondashCeno alkalinesprings show the highest BCl ratios comparable with those of thehost metamorphic minerals (Scambelluri et al 2004) In contrast the
lower boron content and BCl ratio of the alkaline springs of the Voltri
89T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
Group are unexpected features and probably re1047298ect waterndashrockinteractions over a longer time period than in the TarondashCeno ValleysThe boron content in the Northern Apennine alkaline springs may becontrolled by several factors such as (i) the different histories of VoltriGroup and EL ultrama1047297tes (ii) the probable contribution of boronfrom NaCl 1047298uid inclusions (iii) the sinking of boron by travertineprecipitation (Hobbs and Reardon1999) brucite (Rhoades et al 1970)and smectite and (iv) decreases in boron release from minerals due to
increases in pH (Su and Suarez 2004) Acknowledgements
This research was carried out with the 1047297nancial support from theldquoYoung Researchers Projectrdquo (2001ndash2002) of the Italian Ministry of Universities Scienti1047297c Research and Technology (MURST) We thankGianniCortecci (Institute of Geosciences and EarthResources IGG-CNRPisa Italy) and Giampiero Venturelli (Earth Sciences Department Univer-sity of Parma Italy) for useful suggestions and critical review of themanuscript The authors are grateful to Enrico Dinelli (InterdepartmentResearch Centre for Environmental Sciences Alma Mater StudiorumUniversity of Bologna Ravenna Italy) for granting permission to use theGeochemistrsquos WorkbenchsoftwareTheconstructive reviews of theEditor
J Fein and of two anonymous Referees have signi1047297cantly improved the
manuscript
Appendix A Supplementary data
Supplementary data associated with this article can be found inthe online version at doi101016jchemgeo200808017
References
Abbate E Bortolotti V Passerini P 1980 Olistostromes and olistoliths In Sestini G(Ed) Development of the Northern Apennines geosyncline Sedimentary Geology4 pp 521ndash558
Alietti A 1956 Il minerale a strati misti saponite-talco di monte Chiaro (val di TaroAppen Emil) Rendiconti dellAccademia Nazionale dei Lincei 21 201 ndash207
Alietti A De Angelis G Montanari G Sgarlata F Tirelli G 1976 Il ferro nelle argilleappenniniche Periodico di Mineralogia 45 51ndash64
Anselmi B Mellini M Viti C 2000 Chlorine in the Elba Monti Livornesi and Murloserpentines Evidence for seandashwater interaction European Journal of Mineralogy12137ndash146
Apha-Awwa-Wef 1995 Standard Methods for the Examination of Water and Waste-water 19th Edition American Public Health Association American Water WorksAssociation Water Environment Federation
Barnes I LaMarche Jr VC Himmelberg GR 1967 Geochemical evidence of present-day serpentinization Science 156 830ndash832
Barnes I ONeil JR 1969 The relationship between 1047298uids in some fresh alpine-typeultrama1047297cs and possible modern serpentinization Western United StatesGeological Society of America Bulletin 80 1947ndash1960
Barnes I ONeil JRTrescases JJ1978 Present dayserpentinizationin New CaledoniaOman and Yugoslavia Geochimica et Cosmochimica Acta 42 144ndash145
Barnes I Rapp JB ONeil JR Sheppard RA Gude III AJ 1972 Metamorphicassemblages and the direction of 1047298ow of metamorphic 1047298uids in four instances of serpentinization Contributions to Mineralogy and Petrology 35 263ndash276
Barnes JDSelverstone J Sharp ZD 2006Chlorine isotope chemistry of serpentinitesfrom Elba Italy as an indicator of 1047298uid source and subsequent tectonic historyGeochemistry Geophysics Geosystems 7 doi1010292006GC001296
Beccaluva L Emiliani F Venturelli G Zerbi M 1973 Ca Fe Mg Mn Cr Ni Codistribution in some ultrama1047297c rocks outcropping in the Northern Apennines withsome geological remarks Ateneo Parmense Acta Naturalia 9 69ndash98
Beccaluva L Macciotta G Piccardo GB Zeda O 1984 Petrology of lherzolitic rocksfrom the Northern Apennine ophiolites Lithos 17 299 ndash316
BeccaluvaL Venturelli G ZanzucchiG 1975 Dati geologici e geochimici sui basalti diassociazione o1047297olitica dellAppennino ligure-emiliano Ateneo Parmense ActaNaturalia 11 789ndash802
BethkeCM 2002 The Geochemists Workbench A Userrsquos Guide to RxnAct2Tact ReactandGtplot Release40 HydrogeologyProgram Universityof IllinoisUrbanaIL 236pp
Birsoy R 2002Formation of sepiolitendashpalygorskite and related minerals from solutionClays and Clay Minerals 50 736ndash745
Bonatti E Lawrence JR Morandi N 1984 Serpentinization of oceanic peridotitestemperature dependence of mineralogy and boron content Earth and PlanetaryScience Letters 70 88ndash94
Bonifacio E Zanini E Boero V Franchini-Angela M 1997 Pedogenesis in a soilcatena on serpentinite in north-western Italy Geoderma 75 33ndash51
Boschetti T 2003 Studio geochimico e geochimico-isotopico di acque a composizione
estrema e termali dellrsquoAppennino Settentrionale Ph D Thesis Parma Univ
Brigatti MF Franchini G Lugli C Medici L Poppi L Turci E 2000 Interactionsbetween aqueous chromium solutions and layer silicatesApplied Geochemistry 151307ndash1316
Brigatti MF Lugli C Poppi L Venturelli G 1999 Iron-rich saponite dissolutionreactions and Cr uptake Clay Minerals 34 637ndash645
Brigatti MF Valdregrave G 1996 Mixed-layer chloritesmectite (corrensite) in ophioliticrock alteration products Neues Jahrbuch fuer Mineralogie Monatshefte 1 37 ndash47
Bruni J Canepa M Chiodini G Cioni R Cipolli F Longinelli A Marini L OttonelloG Vetuschi Zuccolini M 2002 Irreversible waterndashrock mass transfer accompany-ing the generation of the neutral MgndashHCO3 and high-pH CandashOH spring waters of the Genova province Italy Applied Geochemistry 17 455ndash474
Bryan CR 2005 Drift-Scale THC Seepage Model MDL-NBS-HS-000001 REV 04 YuccaMountain Project Las Vegas Nevada (US)Caruso LJ Chernosky JV 1979 The stability of lizardite Canadian Mineralogist 17
757ndash769Clark ID Fontes JC Fritz P 1992 Stable isotope disequilibria in travertine from high
pH waters Laboratory investigations and 1047297eld observations from Oman Geochi-mica et Cosmochimica Acta 56 2041ndash2050
Cortesogno L 1980 Il metamor1047297smo giurassico nelle o1047297oliti dellAppenninoSettentrionale due distinti cicli metamor1047297ci in ambiente oceanico O1047297oliti 5 5ndash18
Christ CLHostetler PB Siebert RM1973 Studies in the system MgO-SiO2-CO2-H2OIII American Journal of Science 273 65ndash83
Cortesogno L Lucchetti G 1982 Il metamor1047297smo oceanico nei gabbri o1047297oliticidellAppennino Ligure Aspetti mineralogici e petrogenetici Rendiconti dellaSocietagrave Mineralogica Italiana 38 561ndash579
Crispini L Capponi G 2002 Control of metasomatic alteration on the development of shear zone in ultrama1047297c rocks a case study from the Voltri Massif (Ligurian Alps)Architettura interna e cinematica delle zone di taglio Gruppo Italiano di GeologiaStrutturale mdash Riunione Annuale 2002 Pisa Italy 11ndash12 giugno 2002 XIV ndashXV
CrissRE 1999 Principlesof stableisotope distributionOxford University PressNew York
Davis KJ Dove PM De Yoreo JJ 2000 The role of Mg 2+ as an impurity in calcitegrowth Science 290 1134ndash1137
de Leeuw NH 2002 Moleculardynamics simulations of the growth inhibiting effect of Fe2+ Mg2+ Cd2+ and Sr2+ on calcite crystal growth Journal of Physical Chemistry B106 5241ndash5249
De Nardo MT Segadelli S Vescovi P 2007 Studio pilota per la caratterizzazionegeologicadellesorgentinellazona delM Nero (altaVal Cenoe alta ValNuremdash provincedi Parma e Piacenza) Il Geologo dellEmilia-Romagna 25 Nuova Serie pp 5 ndash21
Di Giulio A Geddo G 1990 Studio petrogra1047297co delle Arenarie di Casanova (alta valTrebbia Appennino settentrionale) Atti Ticinesi di Scienze della Terra 33 243ndash254
Dinelli E Lombini A Simoni A Ferrari C 1997 Heavy metals in serpentinite soils of selected outcrops of Piacenza and Parma provinces (Northern Apennines Italy)Mineralogica et Petrographica Acta 40 241ndash255
Epstein S Mayeda T 1953 Variation of 18O content of waters from natural sourcesGeochimica et Cosmochimica Acta 4 213ndash224
Ernst WG Piccardo GB 1979 Petrogenesis of some Ligurian peridotites I Mineraland bulk-rock chemistry Geochimica et Cosmochimica Acta 43 219ndash237
Feder F Trolard F Klingelhofer G Bourrieacute G 2005 In situ Moumlssbauer spectroscopy
evidence for green rust (fougerite) in a gleysol and its mineralogical transforma-tions with time and depth Geochimica et Cosmochimica Acta 69 4463ndash4483Fritz PClark IDFontes JCWhiticar MJFaberE 1992 Deuterium and
13C evidence
for low temperature production of hydrogen and methane in a highly alkalinegroundwater environment in Oman In Kharaka TK Maest AS (Eds) Proceed-ings of the 7th International Symposium on WaterndashRock Interaction BalkemaRotterdam pp 793ndash796
Giammetti F 1968 Le o1047297oliti di Roccaprebalza Costa della Guardia e Groppo di Gorro(Appennino Parmense) Ateneo Parmense Acta Naturalia 4 107ndash142
Giuseppetti G1953 Roccee minerali dellaformazioneo1047297olitica di Volpedo Rendicontidella Societagrave Mineralogica Italiana 9 85ndash134
Gran G 1952 Determination of equivalence point in the potentiometric titrationAnalyst 77 661ndash671
Gunnarsson I Arnorsson S Jakobsson S 2005 Precipitation of poorly crystalline antigoriteunder hydrothermal conditions Geochimica et Cosmochimica Acta 69 2813ndash2828
Guumlven N 1988 Smetites In Bailey SW (Ed) Reviews in Mineralogy 19 Hydrousphyllosilicates (exclusive of micas) Mineralogical Society of America MichiganBook Crafters Inc 497ndash559
Hansel CM Benner SG Fendorf S 2005 Competing Fe(II)-induced mineralization
pathways of ferrihydrite Environmental Science and Technology 39 7147ndash7153HobbsMYReardonEJ 1999 Effect of pH on boronco-precipitation by calcite further
evidence for non-equilibrium partitioning of trace elements Geochimica etCosmochimica Acta 63 1013ndash1021
Jones BF Galan E 1988 Sepiolite and palygorskite In Bailey SW (Ed) HydrousPhyllosilicates (exclusive of Micas) Reviews in Mineralogy vol 19 MineralogicalSociety of America Michigan Book Crafters Inc 631ndash674
Jove Colon CF 2006 Technical work plan for Thermodynamic database for chemicalmodelling TWP-MGR-PA-000039 REV 01 Yucca Mountain Project Las VegasNevada (US)
Kohls DW Rodda Jr JL 1967 Iowaite a new hydrous magnesium hydroxidendashferricoxychloride from the Precambrian of Iowa American Mineralogist 52 1261 ndash1271
Kyser TK OHanley DS Wicks FJ 1999 The origin of 1047298uids associated withserpentinization processes evidence from stable-isotope compositions CanadianMineralogist 37 223ndash237
Lapham DM 1961 New data on deweylite American Mineralogist 46 168ndash188Lee BD SearsSKGraham RC Amrhein CValiH 2003 Secondarymineral genesis
from chlorite and serpentine in an ultrama1047297c soil toposequence Soil Science
Society of America Journal 671309ndash1317
90 T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
Longinelli A Selmo E 2003 Isotopic composition of precipitation in Italy a 1047297rstoverall map Journal of Hydrology 270 75ndash88
Majzlan J Navrotsky A Schwertmann U 2004 Thermodynamics of iron oxides PartIIIEnthalpiesof formation and stability of ferrihydrite (simFe(OH)3) schwertmannite(simFeO(OH)34(SO4)18) and ɛ-Fe2O3 Geochimica et Cosmochimica Acta 681049ndash1059
Mariner P 2003 In-Drift PrecipitatesSalts Model ANL-EBS-MD-000045 REV 01 YuccaMountain Project Las Vegas Nevada (US)
Marini L Ottonello G 2002 Atlante degli acquiferi della Liguria Volume III Le acquedeicomplessio1047297olitici (bacini ArrestraBranega Cassinelle Cerusa ErroGorzenteLeira Lemme Lerone Orba Piota Polcevera Rumaro Sansobbia Stura Teiro
Varenna Visone) Pacini PisaMellini L Zanazzi PF 1987 Crystal structures of lizardite-l T and lizardite-2H1 fromColi Italy American Mineralogist 72 943ndash948
Mezzadri G 1964 Petrogra1047297a delle arenarie di Ostia Rendiconti Societagrave MineralogicaItaliana 20 192ndash228
Mondesir H Decarreau A 1987 Synthesegrave entre 25 et 200 degC de lizardites Ni ndashMgMesure des coef 1047297cients de partage solide-solution aqueuse pour le couple Ni ndashMgdans des lizardites Bulletin de Mineralogie 110 409ndash426
Neal C Shand P 2002 Spring and surface water quality of the Cyprus ophiolitesHydrology and Earth System Sciences 6 797ndash817
Neal C Stanger G 1983 Hydrogen generation from mantle source rocks in OmanEarth and Planetary Science Letters 66 315ndash320
Neal C Stanger G 1984 Calcium and magnesium hydroxide precipitation fromalkaline groundwaters in Oman and their signi1047297cance to the process of serpentinization Mineralogical Magazine 48 237ndash241
Neal C Stanger G 1985 Past and present serpentinisation of ultrama1047297c rocks anexample from Semail ophiolite nappe of Northern Oman In Drever JI (Ed) TheChemistry of Weathering D Reidel Publishing Company pp 249ndash275
Nesbitt HW Bricker OP 1978 Low temperature alteration processes affecting
et Cosmochimica Acta 41 1835ndash1841OrsquoHanley DS 1996 Serpentinites records of tectonic and petrological history Oxford
University PressOrberger B Friedrich G Woermann E 1990 The distribution of halogens and carbon
in Pge-bearing ultrama1047297cs of the Acoje Ophiolite Block Zambales Philippines Journal of Geochemical Exploration 37 147ndash169
Ottonello G Joron JL Piccardo GB 1984 Rare Earth and 3d transition elementsgeochemistry of peridotitic rocks II Ligurian peridotites and associated basalts Journal of Petrology 25 373ndash393
Palandri JL Kharaka YK 2004 A compilation of rate parameters of water-mineralinteraction kinetics for application to geochemical modeling Open File Report2004-1068 US Geological Survey
Panettiere P Cortecci G Dinelli E Bencini A Guidi M 2000 Chemistry and sulfurisotopic composition of precipitation at Bologna Italy Applied Geochemistry 151455ndash1467
Parkhurst DL Appelo CAJ 1999 Userrsquos guide to PHREEQC (version 2) mdash a computerprogram for speciation batch-reaction one-dimensional transport and inversegeochemical calculation US Department of the Interior US Geological SurveyWashington DC
Pfeifer HR 1977 A model for 1047298uids in metamorphosed ultrama1047297c rocks Observationsat surface and subsurface conditions (high pH spring waters) SchweizerischeMineralogische und Petrographische Mitteilungen 57 361ndash396
PlummerLN1984Geochemical modeling A comparisonof forwardand inverse methodsIn Hitchon B Wallick EI (Eds) Proceedings First CanadianAmerican Conference onHydrogeology-Practical Applications of Ground Water Geochemistry Banff AlbertaCanada mdash Worthington Ohio National Water Well Association pp 149ndash177
Rampone E Hofmann AW Piccardo GB Vannucci R Bottazzi P Ottolini L 1995Petrology mineral and isotope geochemistry of the External Ligurides peridotites(Northern Apennines Italy) Journal of Petrology 36 81ndash105
Rampone E Hofmann AW Piccardo GB Vannucci R Bottazzi P Ottolini L 1996Trace element and isotope geochemistry of depleted peridotites from an N-MORBtype ophiolite (Internal Liguride N Italy) Contributions to Mineralogy andPetrology 123 61ndash76
Rhoades JD Ingvalson RD HatcherJT1970Adsorption of boronby ferromagnesianminerals and magnesium hydroxide Soil Science Society of America Journal 34
938ndash941Saini-EidukatB KuchaH Keppler H1994Hibbingiteγ-Fe2(OH)3Clanewmineralfromthe Duluth Complex Minnesota with implications for the oxidation of Fe-bearingcompounds and the transport of metals American Mineralogist 79 555ndash561
Sanford RF 1981 Mineralogical and chemical effects of hydration reactions andapplications to serpentinization American Mineralogist 66 290ndash297
Scambelluri M Muumlntener O Ottolini L Pettke TT Vannucci R 2004 The fate of BCl and Li in the subducted oceanic mantle and in the antigorite breakdown 1047298uidsEarth and Planetary Science Letters 222 217ndash234
Scambelluri M Rampone E 1999 Mg-metasomatism of oceanic gabbros and itscontrol on Ti-clinohumite formation during eclogitization Contributions toMineralogy and Petrology 135 1ndash17
Scambelluri M Bottazzi P Trommsdorff V Vannucci R Hermann J Gogravemez-Pugnaire MT Logravepez-Sagravenchez Vizcaino V 2001a Incompatible element-rich1047298uidsreleased by antigorite breakdown in deeply subducted mantle Earth and PlanetaryScience Letters 192 457ndash470
Scambelluri M Rampone E Piccardo G 2001b Fluid and element cycling insubducted serpentinite a trace-element study of the ErrondashTobbio high-pressureultrama1047297tes (Western Alps NW Italy) Journal of Petrology 42 55ndash67
Segadelli S2006 Lageologia nelpaesaggio le rupi o1047297olitiche in Val Taro e Val Ceno IlGeologo dellrsquoEmilia-Romagna 22 15ndash29 Nuova Serie
Settembre Blundo D Martin De Vidales JL Alietti A 1992 Corrensite-type clayminerals from Taro Valley (North Italy) Mineralogica et Petrographica Acta 35215ndash222
Sharp ZD Barnes JD 2004 Water-soluble chlorides in massive sea1047298oor serpenti-nites a source of chloride in subduction zones Earth and Planetary Science Letters226 243ndash254
Siever R Woodford N 1979 Dissolution kinetics and the weathering of ma1047297cminerals Geochimica et Cosmochimica Acta 43 717ndash724Singer A 1989 Palygorskite and sepiolite group minerals In Dixon JB Weed SB
(Eds) Minerals in Soil Environments Soil Science Society of America MadisonWisconsin pp 829ndash872
Siffert B Wey R 1962 Synthese d rsquoune sepiolite a temperature ordinarie Comptesrendus hebdomadaires des seacuteances de lAcadeacutemie des Sciences de Paris 2541460ndash1463
Środoń J 1999 Nature of mixed-layer clays and mechanisms of their formation andalteration Annual Reviews of Earth and Planetary Science 27 19ndash53
Stoessell RK 1988 25 degC and 1 atm dissolution experiments of sepiolite and keroliteGeochimica et Cosmochimica Acta 52 365ndash374
Stoessell RK Hay RL 1978 The geochemical origin of Sepiolite and keorilite atAmboseli Kenya Contributions to Mineralogy and Petrology 65 255ndash267
Suarez DL Šimůnek J 1996 Solute transport modeling under variably saturatedwater 1047298ow conditions In Lichtner PC Steefel CI Oelkers EH (Eds) ReactiveTransport in Porous Media Reviews in Mineralogy vol 34 Mineralogical Society of America Michigan Book Crafters Inc 229ndash268
Sumner DY Grotzinger JP 1996 Were kinetics of Archean calcium carbonate
precipitation related to oxygen concentration Geology 24 119ndash122Su C Suarez DL 2004 Boron release from weathering of illites serpentine shales
and illiticpalygorskitic soils Soil Science Society of America Journal 68 96 ndash105Toscani L Venturelli G Boschetti T 2001 Sul1047297de-free and sul1047297de-bearing waters in
the Northern Apennines Italy general features and waterndashrock interaction AquaticGeochemistry 7 195ndash216
Treves BE Harper GD 1994 Exposure of serpentinites on the ocean 1047298oor sequenceof faulting and hydrofracturing in the Northern Apennine ophicalcites O1047297oliti 19b435ndash466
Varadachari C Kudrat M Gosh K 1994 Evaluation of standard free energies of formation of clay minerals by an improved regression method Clays and ClayMinerals 42 298ndash307
Veniale F 1966 Sepiolite in sedimenti dellrsquoAppennino Vogherese (Pavia) Periodico diMineralogia 35 343ndash386
Venturelli G 2003 Acque minerali ed ambiente mdash Fondamenti di geochimica deiprocessi di bassa temperatura Pitagora Editrice Bologna
Venturelli G Contini S Bonazzi A Mangia A 1997 Weathering of ultrama1047297c rocksand element mobility at Mt Prinzera Northern Apennines Italy MineralogicalMagazine 61 765ndash778
Venturelli G Frey M 1977 Anchizone metamorphism in sedimentary sequences of the Northern Apennines Rendiconti Societagrave Mineralogica Italiana 33 109ndash123
Vieillard P 2002 A new method for the prediction of Gibbs free energies of formationofphyllosilicates (10 Aring and14 Aring)basedon theelectronegativityscale Claysand ClayMinerals 50 352ndash363
Vissers RLM Strating EHH 2001 Structures and microstructures in a thrust-relatedgreenschist facies tectonic melange Voltri Group (NW Italy) O1047297oliti 26 33ndash46
Viti C Mellini M 1997 Contrasting chemical compositions in associated lizardite andchrysotile in veins from Elba Italy European Journal of Mineralogy 9 585 ndash596
Viti C Mellini M 1998 Mesh textures and bastites in the Elba retrogradeserpentinites European Journal of Mineralogy 10 1341ndash1359
Wenner DB Taylor Jr HP 1974 Deuteriumatomic hydrogen and oxygen-18oxygen-16 studies of serpentinization of ultrama1047297c rocks Geochimica et CosmochimicaActa 38 1255ndash1286
Williams LA Parks GA Crerar DA 1985 Silica diagenesis I Solubility controls Journal of S edimentary Petrology 55 301ndash311
Wilson J Savage D Cuadros J Shibata M RagnarsdottirKV 2006 The effect of ironon montmorillonite stability (I) Background and thermodynamic considerations
Geochimica et Cosmochimica Acta 70 306ndash322WollastR MackenzieFTBricker OP1968 Experimental precipitation and genesis of sepiolite at earth-surface conditions American Mineralogist 531 1645ndash1662
Wrucke CT1996 Serpentine and carbonated-hosted asbestos deposits InEdwardAdu Bray (Eds) Preliminary compilation of descriptive geoenvironmental mineraldeposit models 95ndash831 US Geological Survey Open-File Report pp 39 ndash46
Xu T Apps JA Pruess K 2001 Analysis of mineral trapping for CO2 disposal in deepaquifers Lawrence Berkeley National Laboratory University of California
ZhangYDaweRA2000 In1047298uence ofMg2+ on thekineticsof calcite precipitation andcalcite crystal morphology Chemical Geology 163 129ndash138
Zhu C Anderson G 2002 Environmental applications of geochemical modelingCambridge University Press
91T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
montmorillonites vermiculite and saponites are predicted to occurrather than that of Mg-clays Moreover in order to simulate the slowprecipitation of calcite in depth precipitation was not activated atslight supersaturation (SIcalcite=001) during the initial stages of thewaterndashrock interaction in this manner saturation index of calcite wasfree to increase in the following steps and in alkaline waters
6 Discussion of the model results
61 Explanation of pre-modeling choices and post-modeling results
The results of the reaction path modeling are reported in Table 5and shown in Figs 6 and 9 The reaction paths model the dissolutionof a typical local serpentinite (mineralogically constrained as inTable 3) and the sequential precipitation of gibbsite kaoliniteferrihydrite vermiculite Fe2+ndashMg2+-saponite and a poorly-crystal-line serpentine( p-antigorite of Gunnarsson et al2005) The1047297rst stepinvolves the reaction of the ldquoinitial waterrdquo with serpentinite atconstant temperature (11 degC) PO2 (10minus26 bar) and PCO2 (10minus30 bar)this early stage corresponds to bicarbonate groundwater feeding theinvestigated spring system At the stage of primary magnetiteequilibrium and Fe2+ndashMg2+-saponite supersaturation the modelassumes the system to be closed to CO2 and to have a 1047297xed partialpressure of oxygen (PO2= 10minus40 bar) these conditions represent
when water switch from a bicarbonate to an alkaline compositionThe calculated amount of interacting serpentinite needed to obtainthe alkaline springs of the TarondashCeno Valleys (mean PCO2= 10minus7 bar)is approximately 00203 moles (see Table 5)
Bruni et al (2002) explained the origin of the alkaline waters of the Voltri Group as the result of the evolution of meteoricbicarbonate water following the precipitation of sepiolite withabundance similar to that of goethite This might be supported bythe observation that sepiolite or smectite precipitate at lowtemperature from solutions with low Al content when the pH is inthe range 75ndash95 or N95 respectively (Siffert and Wey 1962 Jonesand Galan 1988) but sepiolites form from the weathering of serpentines in arid climate regions whereas sepiolites are rare inhumid climate regions where serpentinite changes to Fe-smectite
(Singer 1989 Wrucke 1996) In the ophiolite rocks of the Northern
Apennine sepiolite is found only sporadically in veins (Giuseppetti1953 Veniale 1966) and is unknown in the ophiolites of the Tarondash
Ceno Valleys whereas it is thought that sepiolite of the Voltri Groupprecipitated from metasomatic 1047298uids during development of shearzones (Crispini and Capponi 2002)
As described above our model explains the transition frombicarbonate water to alkaline water (UM15 and PR10 samples) asthe result of low-temperature precipitation of a poor crystalline
serpentine with the solubility product of p-antigorite (see Sr3 inTable 4 and Fig 6) The model indicates that the bicarbonate waters of the Voltri Group with the highest [Mg2+][H+]2 activity ratios similarlyevolve towards alkaline waters (Fig 6B) The logK value of thesynthetic low temperature serpentine adopted in this model and its
Fig 7 Saturation indices of chrysotile and p-antigorite (Sr2 and Sr3 thermodynamic
data respectively in Table 4) for bicarbonate and alkaline waters coming fromultrama1047297c rocks of the TarondashCeno Valleys as a function of SiMg molar ratioThermodynamic data on chrysotile and p-antigorite are from the thermocomV8R6
database and Gunnarsson et al (2005) respectively
Fig 8 Activityplot forthe MgOndashSiO2ndashAl2O3ndashFe2O3(FeO)ndashH2O system analyzed forSiO2-phases and aqueous species at different log PO2 and log[aAl3+(aH+)3] values Springwaters coming from the TarondashCeno Valleys (symbols as in Fig 2) and Voltri Groupultrama1047297tes (symbols as in Fig 5) are shown The greyarea refers to bicarbonate watersfrom theVoltri Group (Mariniand Ottonello2002Bruniet al 2002) Thesepiolite 1047297eldis very narrow or hidden by saponite-Mg (1047297eld limits shown as continuous lines)saponite-Fe3+ndashMg (1047297eldlimits shown as dottedlines) or vermiculite(1047297eldlimits shownas dottedndashdashed lines) The talc stability 1047297eld and boundaries with saponite-Mg
montmorillonite-Mg and cronstedite-Mg are shown for comparison (dashed lines)
87T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
relative uncertainty of plusmn2 (see Fig 8 in Gunnarsson et al 2005) arecomparable to the logK variation reported for other low-temperatureMg-serpentines (Table 4) that is within the lines Sr1ndash4 for serpentineminerals where samples varying from bicarbonate composition toalkaline composition in the ultrama1047297c aquifer are grouped (Fig 6)Fig 7 shows a plot of calculated serpentine saturation index as a
function of the SiMg molar ratio for the water samples from theultrama1047297tes of the TarondashCeno ValleysThe SI andSiMgvalues of pureMg-serpentine (SI= 0 SiMg=067) provide a useful index forseparating bicarbonate waters (SIb0 SiMgb067) from alkalinewaters (SIN0 SiMgN067) and the distribution of data points isconsistent with the idea that bicarbonate waters approach saturationwith serpentine until they become supersaturated as alkaline waters
Based on the Mgvs SiO2(aq) activitydiagram and speci1047297cratiosoflog[aAl3+(aH+)3] Birsoy (2002) analyzed the behavior of sepiolite duringultrama1047297te weathering The diagram produces a very narrow stability1047297eldfor sepiolite(Fig 8) when it is programmedwith suitable conditionsfor the bicarbonate waters (log PO2=minus25 bar log [aAl3+(aH+)3]= 65) andthe alkaline waters (log PO2=minus50 bar log [aAl3+(aH+)3]= 45) sampled inthe present study This stability 1047297eld grows larger when the absence of
Mg-smectites and talc is assumed but this is an unrealistic conditionbecause (i) all these minerals occur in the rocks within the areainvestigated and (ii) at high [Mg2+][H+]2 activity ratios the 1047297eld of sepiolite is smaller than that of subordinate compared to those poorly-crystalline serpentine (p-antigorite) and clinochlore
Among the alkaline springs of the Voltri Group the waters withlow log [aCa2+(aH+)2] values (Fig 9) show patterns compatible withchlorite interaction (step 5c in Fig6D)The alkaline waters whose datafall near the clinochlore-14Aring 1047297eld (Fig 6D) cannot be explained bysepiolite precipitation because the sepiolite line does not cross thechlorite 1047297elds Thus the peculiar composition of some alkaline watersof the Voltri Group is probably due to interactions between water andprimary chlorites of the ultrama1047297c rocks (optional step 5c in Table 5cb) This is supported by the occurrence sometimes in large amounts
of metasomatic Mgndash(Fe)-chlorite (Scambelluri and Rampone 1999
Vissers and Strating 2001) in the Ligurian area where the relativeabundance of chlorite can reach 50 in veins in high strainserpentinite mylonites of the ErrondashTobbio ultrama1047297tes (Scambelluriet al 2001b) Otherwise in a chlorite-free primary paragenesis thesolution evolving on the Sr3 serpentine line reaches equilibrium withclinochlore but probably the latter does not precipitate due to theslow kinetics of precipitation under these conditions Thus thedissolution of additional reacting rock in the iterative weathering
process supplies Mg
2+
to the solution thereby favoring the Sr3 to Sr4poorly crystalline serpentine transformation In fact all the PR10samples and some of the alkaline springs from the Voltri Group showincreasing [Mg2+][H+]2 activity ratio and low variation in silicaactivity (Fig 6D) starting from the invariant triple point 4a (Sr3-serpentine Fe2+ndashMg-saponite chlinoclore-14 Aring saturation) andextending towards the brucite line This path can be explained asthe result of the transformation of poorly crystalline serpentine fromSr3 to Sr4 the latter is more stable at higher [Mg2+][H+]2 activityratios Finally at very low PO2 values vermiculite disappears and thestability 1047297elds of chlorite and Fe2+ndashMg-saponite become narrowerwhereas those of Fe2+-rich phases (eg Fe2+-saponite) become widerThe data for the alkaline waters with the highest log [aMg2+(aH+)2]from the Voltri Group fall near the stability 1047297eld of peridotite primaryminerals like forsterite and enstatite (Fig 6D) therefore the last stepsof theevolution involve theequilibration of thesolutionwithenstatite(steps 8a and 8c in Table 5ab)
62 Ca-rich alkaline waters
In preliminary modeling of the evolution of the TarondashCeno springwaters Boschetti (2003) concluded that meteoric water initiallysaturated in Fe3+ndashMg2+-chlorite Fe3+ndashMg2+-saponite and pyrite (andor goethite) and then interacted with the ultrama1047297te mineralassemblage eventually evolved into the alkaline waters In themodel of Boschetti (2003) chlorite saponite and calcite were added
Fig 9 Activity plot for the MgOndashCaOndash(CO2)ndashH2O system with phase limits calculated ata temperature of 11 degCSpring waterscoming from theTarondashCeno Valleys (symbols as inFig2) andVoltri Groupultrama1047297tes(symbolsas in Fig5) areshown Calculated path forwaterndashultrama1047297c rock interaction is shown (dotted line) with the number ndashlettercombinations referring to the appearance or disappearance of minerals (see Table 5)The grey area refers to bicarbonate waters from the Voltri Group (Marini and Ottonello2002)
Fig 10 Plots of (A) calcite saturation indexes and (B) PCO2 as a function of Clminus ascalculated by PHREEQCI software for waters from the TarondashCeno Valleys Alkalinesprings for the Voltri Group are also shown (small black diamonds Marini andOttonello 2002) The grey shading is deeper(A) in the saturation1047297eld(SI= 00plusmn05) and
(B) where the lowest log PCO2 value is approached
88 T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
to the solution in order to achieve a high [Ca2+][H+]2 activity ratio andsupersaturation in these phases However in light of the work of Clarket al (1992) and Bruni et al (2002) we believe that the calcitesupersaturation in alkaline waters (Fig 9 and 10) which in our modelis achieved by preventing the precipitation of calcite in the 1047297rst stepsof the reaction path (Table 5a and b) depends on the differencebetween CO2 partial pressureat thegivendepth(PCO2le10minus10bar) andat the surface (PCO2cong10minus35) Waters rising to the surface are far from
equilibrium with the atmosphere where the PCO2 is 4times10
6
timesgreater than the value at depth thus the water rapidly take up CO2which is promptly converted to CO3
2minus because of the high pH Thisleads to supersaturation with calcite through the following hydro-xylation reaction
As a result large amounts of travertine are generated in the VoltriGroup (Marini and Ottonello 2002) and carbonate concretionsknown as devilrsquo s coins are formed at Mt Prinzera in Taro valleyBelow the surface the SIcalcite of the alkaline waters is expected to belower than what the model estimates Under conditions of lowcarbonate alkalinity low CO2 and O2 partial pressure and high pHcalcite nucleation may be inhibited and the kinetics of precipitationreduced by the high concentrations of dissolved divalent cations suchas Mg2+ Fe2+ and Mn2+ (Sumner and Grotzinger 1996 Davis et al2000 Zhang and Dawe 2000 de Leeuw 2002) The fact that severalalkaline water samples fall within or near the calcitendashdolomiteequilibrium (Fig 9) may indicate magnesium calcite or protodolomitenucleation in those extreme subsurface conditions However possibleinteraction between some calcium-rich waters of the Voltri Group andCa-rich rocks such as ophicarbonates or rodingites cannot beexcluded these rocks are widespread in the Ligurian Apennineserpentinites (Treves and Harper 1994)
63 Sepiolite vs serpentine nucleation under weathering conditions
constraints on the kinetics and equilibrium-based approaches
The kinetics of precipitation of poorly-crystalline serpentine and
sepiolite cannot be compared quantitatively because rate constantsare available only for well-ordered crystalline phases (eg see Palandriand Kharaka 2004 for crystalline serpentine) Sepiolite precipitationis often invoked to explain magnesium scavenging from solutions butthe lack of kinetic data for this mineral leads to overly simplisticreaction transport models These models are forced to use kinetic datafor phyllosilicates such as kaolinite or muscovite (eg Xu et al 2001Bryan 2005) Moreover the low-temperature thermodynamic dataeven if they are fairly uncertain show that sepiolite solubilityincreases with decreasing crystallinity (Table 4 Stoessell 1988) Infact crystalline sepiolite requires a long time to reach equilibriumwith solution which imposes a slow control on thewater composition(eg Stoessell and Hay 1978) Consequently in modeling based onreaction transport that simulates freshly precipitated or poorly
crystalline sepiolite formation the logK value of Wollast et al (1968)is preferable to the lower values (Suarez and Šimůnek 1996) Thispreference agrees with the approach of Mariner (2003) and Jove Colon(2006) who adopt a logK (25 degC) value for sepiolite of 3044 It is alsosimilar to that reported by Wollast et al (1968) and it is six logK unitshigherthan the valueused in thermocomV8R6 dataset which is oftenused (perhaps improperly) to model the supergenic nucleation of sepiolite In the stability diagram in Fig 6 where the Sp1 and Sp2 linesrepresent the stability boundaries of crystalline and poorly-crystallinesepiolite respectively the evolution of TarondashCeno waters frombicarbonate to alkaline (Figs 6A and C) follows the stability 1047297eldboundary of the poorly-crystalline serpentine (between Sr2ndashSr3 lines)rather than that of the poorly-crystalline sepiolite (Sp2) as postulatedin themodel The lines for Sp2 Sr3 and amorphoussilica intersect at a
common point (Fig 6D) which is consistent with the nucleation of
sepiolite that typically occurs in solutions that are saturated orsupersaturated with amorphous silica and (Mg)-carbonate (egStoessell and Hay 1978 Birsoy 2002) Some bicarbonate watersfrom the Voltri Group have a silica activity near the amorphous silicaboundary (Fig 6B) but only a few alkaline springs of this group fall onthe poorly-crystalline sepiolite boundary line (Sp2 Fig 6D) Both of these features are inconsistent with the possibility that sepioliteprecipitates from supergenic fresh water in the Northern Apennine
Moreover these two results suggest that values for logK(sepiolite)b
3044 (see Table 4) should be used to model the dissolutionprecipitation of the well-crystallized phase that is the most commonform for the metasomatic sepiolite found in the investigated area
7 Conclusions
The waters issuing from ultrama1047297tes in the TarondashCeno Valleysoriginate from low-temperature interactions between meteoric waterand serpentinite rocks In this report the waterndashrock interaction hasbeen modeled using reaction paths in progress mode During their earlyinteraction with soil and rock minerals bicarbonate waters super-saturated in kaolinite ferrihydrite vermiculite and Fe2+ndashMg2+-saponitewhere theorigin of the secondary iron saponite as Fe2+-bearing smectiteagrees with the hypothesis of Guumlven (1988) Then while remainingsupersaturated with saponite the bicarbonate waters evolve into analkaline facies when poorly-crystalline serpentine precipitates Duringrock weathering the Al-bearing crystalline serpentine of oceanic originprobably promotes nucleation of the low-temperature serpentineamorphous phase and supplies Al3+ to the solution for saponiteprecipitation This interpretation is consistent with the existence of Al-poor supergenic serpentine (Neal and Stanger 1985) and the partiallymeteoric OH-isotopic imprint observed in other ultrama1047297c systems
Dif 1047297culties and complexities in the modeling arise fromuncertainties in the primary mineralogy of some local ophiolitesand from the presence of different lithotypes (rodingites ophi-carbonates gabbros) which can signi1047297cantly affect the composi-tion of water samples This can be seen in the example of the Candash
OH springs issuing from the Voltri Group (PO2 and PCO2 of 10minus70
and 10minus10 bar respectively Bruni et al 2002 Marini and
Ottonello 2002) These springs appear to be more evolved thanthe TarondashCeno alkaline waters as a result of their particularly longinteractions with serpentinites until equilibrium with enstatite isreached or with chlorite-bearing serpentinite The use of anupdated thermodynamic database in our model avoided problemsdue to mineralogical heterogeneity in the mixed-layer claysassociated with ophiolitic rocks this database accurately re1047298ectedthe chemical composition of the single phases Using this modelthe vermiculitendashchlorite and saponitendashchlorite associations seen inthe activity diagrams simulate respectively the presence of highand low charge corrensite observed in the 1047297eld The accuracy of ourmodel can be tested by comparing estimates of chlorine composi-tion resulting from our model and from experimental methods The
mean Cl concentration of the bicarbonate water is 41plusmn16 mgLtaking into account a Cl concentration of sim2 mgL in rainwater inthe area (cf Panettiere et al 2000 Venturelli 2003) 2 mgL of Clin spring water are estimated to come from serpentinite This valueagrees with the model which predicts that the switch frombicarbonate water to alkaline water requires the congruentdissolution of 00203 mol of serpentinite containing 100 ppm of soluble Cl thereby yielding water containing 2 ppm of Cl
In addition we think that weathering of the accessory albiteplagioclase is responsible for Na enrichment of the waters whereas Clenrichment in the alkaline waters is predicted to result from theleaching of NaCl 1047298uid inclusions Moreover the TarondashCeno alkalinesprings show the highest BCl ratios comparable with those of thehost metamorphic minerals (Scambelluri et al 2004) In contrast the
lower boron content and BCl ratio of the alkaline springs of the Voltri
89T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
Group are unexpected features and probably re1047298ect waterndashrockinteractions over a longer time period than in the TarondashCeno ValleysThe boron content in the Northern Apennine alkaline springs may becontrolled by several factors such as (i) the different histories of VoltriGroup and EL ultrama1047297tes (ii) the probable contribution of boronfrom NaCl 1047298uid inclusions (iii) the sinking of boron by travertineprecipitation (Hobbs and Reardon1999) brucite (Rhoades et al 1970)and smectite and (iv) decreases in boron release from minerals due to
increases in pH (Su and Suarez 2004) Acknowledgements
This research was carried out with the 1047297nancial support from theldquoYoung Researchers Projectrdquo (2001ndash2002) of the Italian Ministry of Universities Scienti1047297c Research and Technology (MURST) We thankGianniCortecci (Institute of Geosciences and EarthResources IGG-CNRPisa Italy) and Giampiero Venturelli (Earth Sciences Department Univer-sity of Parma Italy) for useful suggestions and critical review of themanuscript The authors are grateful to Enrico Dinelli (InterdepartmentResearch Centre for Environmental Sciences Alma Mater StudiorumUniversity of Bologna Ravenna Italy) for granting permission to use theGeochemistrsquos WorkbenchsoftwareTheconstructive reviews of theEditor
J Fein and of two anonymous Referees have signi1047297cantly improved the
manuscript
Appendix A Supplementary data
Supplementary data associated with this article can be found inthe online version at doi101016jchemgeo200808017
References
Abbate E Bortolotti V Passerini P 1980 Olistostromes and olistoliths In Sestini G(Ed) Development of the Northern Apennines geosyncline Sedimentary Geology4 pp 521ndash558
Alietti A 1956 Il minerale a strati misti saponite-talco di monte Chiaro (val di TaroAppen Emil) Rendiconti dellAccademia Nazionale dei Lincei 21 201 ndash207
Alietti A De Angelis G Montanari G Sgarlata F Tirelli G 1976 Il ferro nelle argilleappenniniche Periodico di Mineralogia 45 51ndash64
Anselmi B Mellini M Viti C 2000 Chlorine in the Elba Monti Livornesi and Murloserpentines Evidence for seandashwater interaction European Journal of Mineralogy12137ndash146
Apha-Awwa-Wef 1995 Standard Methods for the Examination of Water and Waste-water 19th Edition American Public Health Association American Water WorksAssociation Water Environment Federation
Barnes I LaMarche Jr VC Himmelberg GR 1967 Geochemical evidence of present-day serpentinization Science 156 830ndash832
Barnes I ONeil JR 1969 The relationship between 1047298uids in some fresh alpine-typeultrama1047297cs and possible modern serpentinization Western United StatesGeological Society of America Bulletin 80 1947ndash1960
Barnes I ONeil JRTrescases JJ1978 Present dayserpentinizationin New CaledoniaOman and Yugoslavia Geochimica et Cosmochimica Acta 42 144ndash145
Barnes I Rapp JB ONeil JR Sheppard RA Gude III AJ 1972 Metamorphicassemblages and the direction of 1047298ow of metamorphic 1047298uids in four instances of serpentinization Contributions to Mineralogy and Petrology 35 263ndash276
Barnes JDSelverstone J Sharp ZD 2006Chlorine isotope chemistry of serpentinitesfrom Elba Italy as an indicator of 1047298uid source and subsequent tectonic historyGeochemistry Geophysics Geosystems 7 doi1010292006GC001296
Beccaluva L Emiliani F Venturelli G Zerbi M 1973 Ca Fe Mg Mn Cr Ni Codistribution in some ultrama1047297c rocks outcropping in the Northern Apennines withsome geological remarks Ateneo Parmense Acta Naturalia 9 69ndash98
Beccaluva L Macciotta G Piccardo GB Zeda O 1984 Petrology of lherzolitic rocksfrom the Northern Apennine ophiolites Lithos 17 299 ndash316
BeccaluvaL Venturelli G ZanzucchiG 1975 Dati geologici e geochimici sui basalti diassociazione o1047297olitica dellAppennino ligure-emiliano Ateneo Parmense ActaNaturalia 11 789ndash802
BethkeCM 2002 The Geochemists Workbench A Userrsquos Guide to RxnAct2Tact ReactandGtplot Release40 HydrogeologyProgram Universityof IllinoisUrbanaIL 236pp
Birsoy R 2002Formation of sepiolitendashpalygorskite and related minerals from solutionClays and Clay Minerals 50 736ndash745
Bonatti E Lawrence JR Morandi N 1984 Serpentinization of oceanic peridotitestemperature dependence of mineralogy and boron content Earth and PlanetaryScience Letters 70 88ndash94
Bonifacio E Zanini E Boero V Franchini-Angela M 1997 Pedogenesis in a soilcatena on serpentinite in north-western Italy Geoderma 75 33ndash51
Boschetti T 2003 Studio geochimico e geochimico-isotopico di acque a composizione
estrema e termali dellrsquoAppennino Settentrionale Ph D Thesis Parma Univ
Brigatti MF Franchini G Lugli C Medici L Poppi L Turci E 2000 Interactionsbetween aqueous chromium solutions and layer silicatesApplied Geochemistry 151307ndash1316
Brigatti MF Lugli C Poppi L Venturelli G 1999 Iron-rich saponite dissolutionreactions and Cr uptake Clay Minerals 34 637ndash645
Brigatti MF Valdregrave G 1996 Mixed-layer chloritesmectite (corrensite) in ophioliticrock alteration products Neues Jahrbuch fuer Mineralogie Monatshefte 1 37 ndash47
Bruni J Canepa M Chiodini G Cioni R Cipolli F Longinelli A Marini L OttonelloG Vetuschi Zuccolini M 2002 Irreversible waterndashrock mass transfer accompany-ing the generation of the neutral MgndashHCO3 and high-pH CandashOH spring waters of the Genova province Italy Applied Geochemistry 17 455ndash474
Bryan CR 2005 Drift-Scale THC Seepage Model MDL-NBS-HS-000001 REV 04 YuccaMountain Project Las Vegas Nevada (US)Caruso LJ Chernosky JV 1979 The stability of lizardite Canadian Mineralogist 17
757ndash769Clark ID Fontes JC Fritz P 1992 Stable isotope disequilibria in travertine from high
pH waters Laboratory investigations and 1047297eld observations from Oman Geochi-mica et Cosmochimica Acta 56 2041ndash2050
Cortesogno L 1980 Il metamor1047297smo giurassico nelle o1047297oliti dellAppenninoSettentrionale due distinti cicli metamor1047297ci in ambiente oceanico O1047297oliti 5 5ndash18
Christ CLHostetler PB Siebert RM1973 Studies in the system MgO-SiO2-CO2-H2OIII American Journal of Science 273 65ndash83
Cortesogno L Lucchetti G 1982 Il metamor1047297smo oceanico nei gabbri o1047297oliticidellAppennino Ligure Aspetti mineralogici e petrogenetici Rendiconti dellaSocietagrave Mineralogica Italiana 38 561ndash579
Crispini L Capponi G 2002 Control of metasomatic alteration on the development of shear zone in ultrama1047297c rocks a case study from the Voltri Massif (Ligurian Alps)Architettura interna e cinematica delle zone di taglio Gruppo Italiano di GeologiaStrutturale mdash Riunione Annuale 2002 Pisa Italy 11ndash12 giugno 2002 XIV ndashXV
CrissRE 1999 Principlesof stableisotope distributionOxford University PressNew York
Davis KJ Dove PM De Yoreo JJ 2000 The role of Mg 2+ as an impurity in calcitegrowth Science 290 1134ndash1137
de Leeuw NH 2002 Moleculardynamics simulations of the growth inhibiting effect of Fe2+ Mg2+ Cd2+ and Sr2+ on calcite crystal growth Journal of Physical Chemistry B106 5241ndash5249
De Nardo MT Segadelli S Vescovi P 2007 Studio pilota per la caratterizzazionegeologicadellesorgentinellazona delM Nero (altaVal Cenoe alta ValNuremdash provincedi Parma e Piacenza) Il Geologo dellEmilia-Romagna 25 Nuova Serie pp 5 ndash21
Di Giulio A Geddo G 1990 Studio petrogra1047297co delle Arenarie di Casanova (alta valTrebbia Appennino settentrionale) Atti Ticinesi di Scienze della Terra 33 243ndash254
Dinelli E Lombini A Simoni A Ferrari C 1997 Heavy metals in serpentinite soils of selected outcrops of Piacenza and Parma provinces (Northern Apennines Italy)Mineralogica et Petrographica Acta 40 241ndash255
Epstein S Mayeda T 1953 Variation of 18O content of waters from natural sourcesGeochimica et Cosmochimica Acta 4 213ndash224
Ernst WG Piccardo GB 1979 Petrogenesis of some Ligurian peridotites I Mineraland bulk-rock chemistry Geochimica et Cosmochimica Acta 43 219ndash237
Feder F Trolard F Klingelhofer G Bourrieacute G 2005 In situ Moumlssbauer spectroscopy
evidence for green rust (fougerite) in a gleysol and its mineralogical transforma-tions with time and depth Geochimica et Cosmochimica Acta 69 4463ndash4483Fritz PClark IDFontes JCWhiticar MJFaberE 1992 Deuterium and
13C evidence
for low temperature production of hydrogen and methane in a highly alkalinegroundwater environment in Oman In Kharaka TK Maest AS (Eds) Proceed-ings of the 7th International Symposium on WaterndashRock Interaction BalkemaRotterdam pp 793ndash796
Giammetti F 1968 Le o1047297oliti di Roccaprebalza Costa della Guardia e Groppo di Gorro(Appennino Parmense) Ateneo Parmense Acta Naturalia 4 107ndash142
Giuseppetti G1953 Roccee minerali dellaformazioneo1047297olitica di Volpedo Rendicontidella Societagrave Mineralogica Italiana 9 85ndash134
Gran G 1952 Determination of equivalence point in the potentiometric titrationAnalyst 77 661ndash671
Gunnarsson I Arnorsson S Jakobsson S 2005 Precipitation of poorly crystalline antigoriteunder hydrothermal conditions Geochimica et Cosmochimica Acta 69 2813ndash2828
Guumlven N 1988 Smetites In Bailey SW (Ed) Reviews in Mineralogy 19 Hydrousphyllosilicates (exclusive of micas) Mineralogical Society of America MichiganBook Crafters Inc 497ndash559
Hansel CM Benner SG Fendorf S 2005 Competing Fe(II)-induced mineralization
pathways of ferrihydrite Environmental Science and Technology 39 7147ndash7153HobbsMYReardonEJ 1999 Effect of pH on boronco-precipitation by calcite further
evidence for non-equilibrium partitioning of trace elements Geochimica etCosmochimica Acta 63 1013ndash1021
Jones BF Galan E 1988 Sepiolite and palygorskite In Bailey SW (Ed) HydrousPhyllosilicates (exclusive of Micas) Reviews in Mineralogy vol 19 MineralogicalSociety of America Michigan Book Crafters Inc 631ndash674
Jove Colon CF 2006 Technical work plan for Thermodynamic database for chemicalmodelling TWP-MGR-PA-000039 REV 01 Yucca Mountain Project Las VegasNevada (US)
Kohls DW Rodda Jr JL 1967 Iowaite a new hydrous magnesium hydroxidendashferricoxychloride from the Precambrian of Iowa American Mineralogist 52 1261 ndash1271
Kyser TK OHanley DS Wicks FJ 1999 The origin of 1047298uids associated withserpentinization processes evidence from stable-isotope compositions CanadianMineralogist 37 223ndash237
Lapham DM 1961 New data on deweylite American Mineralogist 46 168ndash188Lee BD SearsSKGraham RC Amrhein CValiH 2003 Secondarymineral genesis
from chlorite and serpentine in an ultrama1047297c soil toposequence Soil Science
Society of America Journal 671309ndash1317
90 T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
Longinelli A Selmo E 2003 Isotopic composition of precipitation in Italy a 1047297rstoverall map Journal of Hydrology 270 75ndash88
Majzlan J Navrotsky A Schwertmann U 2004 Thermodynamics of iron oxides PartIIIEnthalpiesof formation and stability of ferrihydrite (simFe(OH)3) schwertmannite(simFeO(OH)34(SO4)18) and ɛ-Fe2O3 Geochimica et Cosmochimica Acta 681049ndash1059
Mariner P 2003 In-Drift PrecipitatesSalts Model ANL-EBS-MD-000045 REV 01 YuccaMountain Project Las Vegas Nevada (US)
Marini L Ottonello G 2002 Atlante degli acquiferi della Liguria Volume III Le acquedeicomplessio1047297olitici (bacini ArrestraBranega Cassinelle Cerusa ErroGorzenteLeira Lemme Lerone Orba Piota Polcevera Rumaro Sansobbia Stura Teiro
Varenna Visone) Pacini PisaMellini L Zanazzi PF 1987 Crystal structures of lizardite-l T and lizardite-2H1 fromColi Italy American Mineralogist 72 943ndash948
Mezzadri G 1964 Petrogra1047297a delle arenarie di Ostia Rendiconti Societagrave MineralogicaItaliana 20 192ndash228
Mondesir H Decarreau A 1987 Synthesegrave entre 25 et 200 degC de lizardites Ni ndashMgMesure des coef 1047297cients de partage solide-solution aqueuse pour le couple Ni ndashMgdans des lizardites Bulletin de Mineralogie 110 409ndash426
Neal C Shand P 2002 Spring and surface water quality of the Cyprus ophiolitesHydrology and Earth System Sciences 6 797ndash817
Neal C Stanger G 1983 Hydrogen generation from mantle source rocks in OmanEarth and Planetary Science Letters 66 315ndash320
Neal C Stanger G 1984 Calcium and magnesium hydroxide precipitation fromalkaline groundwaters in Oman and their signi1047297cance to the process of serpentinization Mineralogical Magazine 48 237ndash241
Neal C Stanger G 1985 Past and present serpentinisation of ultrama1047297c rocks anexample from Semail ophiolite nappe of Northern Oman In Drever JI (Ed) TheChemistry of Weathering D Reidel Publishing Company pp 249ndash275
Nesbitt HW Bricker OP 1978 Low temperature alteration processes affecting
et Cosmochimica Acta 41 1835ndash1841OrsquoHanley DS 1996 Serpentinites records of tectonic and petrological history Oxford
University PressOrberger B Friedrich G Woermann E 1990 The distribution of halogens and carbon
in Pge-bearing ultrama1047297cs of the Acoje Ophiolite Block Zambales Philippines Journal of Geochemical Exploration 37 147ndash169
Ottonello G Joron JL Piccardo GB 1984 Rare Earth and 3d transition elementsgeochemistry of peridotitic rocks II Ligurian peridotites and associated basalts Journal of Petrology 25 373ndash393
Palandri JL Kharaka YK 2004 A compilation of rate parameters of water-mineralinteraction kinetics for application to geochemical modeling Open File Report2004-1068 US Geological Survey
Panettiere P Cortecci G Dinelli E Bencini A Guidi M 2000 Chemistry and sulfurisotopic composition of precipitation at Bologna Italy Applied Geochemistry 151455ndash1467
Parkhurst DL Appelo CAJ 1999 Userrsquos guide to PHREEQC (version 2) mdash a computerprogram for speciation batch-reaction one-dimensional transport and inversegeochemical calculation US Department of the Interior US Geological SurveyWashington DC
Pfeifer HR 1977 A model for 1047298uids in metamorphosed ultrama1047297c rocks Observationsat surface and subsurface conditions (high pH spring waters) SchweizerischeMineralogische und Petrographische Mitteilungen 57 361ndash396
PlummerLN1984Geochemical modeling A comparisonof forwardand inverse methodsIn Hitchon B Wallick EI (Eds) Proceedings First CanadianAmerican Conference onHydrogeology-Practical Applications of Ground Water Geochemistry Banff AlbertaCanada mdash Worthington Ohio National Water Well Association pp 149ndash177
Rampone E Hofmann AW Piccardo GB Vannucci R Bottazzi P Ottolini L 1995Petrology mineral and isotope geochemistry of the External Ligurides peridotites(Northern Apennines Italy) Journal of Petrology 36 81ndash105
Rampone E Hofmann AW Piccardo GB Vannucci R Bottazzi P Ottolini L 1996Trace element and isotope geochemistry of depleted peridotites from an N-MORBtype ophiolite (Internal Liguride N Italy) Contributions to Mineralogy andPetrology 123 61ndash76
Rhoades JD Ingvalson RD HatcherJT1970Adsorption of boronby ferromagnesianminerals and magnesium hydroxide Soil Science Society of America Journal 34
938ndash941Saini-EidukatB KuchaH Keppler H1994Hibbingiteγ-Fe2(OH)3Clanewmineralfromthe Duluth Complex Minnesota with implications for the oxidation of Fe-bearingcompounds and the transport of metals American Mineralogist 79 555ndash561
Sanford RF 1981 Mineralogical and chemical effects of hydration reactions andapplications to serpentinization American Mineralogist 66 290ndash297
Scambelluri M Muumlntener O Ottolini L Pettke TT Vannucci R 2004 The fate of BCl and Li in the subducted oceanic mantle and in the antigorite breakdown 1047298uidsEarth and Planetary Science Letters 222 217ndash234
Scambelluri M Rampone E 1999 Mg-metasomatism of oceanic gabbros and itscontrol on Ti-clinohumite formation during eclogitization Contributions toMineralogy and Petrology 135 1ndash17
Scambelluri M Bottazzi P Trommsdorff V Vannucci R Hermann J Gogravemez-Pugnaire MT Logravepez-Sagravenchez Vizcaino V 2001a Incompatible element-rich1047298uidsreleased by antigorite breakdown in deeply subducted mantle Earth and PlanetaryScience Letters 192 457ndash470
Scambelluri M Rampone E Piccardo G 2001b Fluid and element cycling insubducted serpentinite a trace-element study of the ErrondashTobbio high-pressureultrama1047297tes (Western Alps NW Italy) Journal of Petrology 42 55ndash67
Segadelli S2006 Lageologia nelpaesaggio le rupi o1047297olitiche in Val Taro e Val Ceno IlGeologo dellrsquoEmilia-Romagna 22 15ndash29 Nuova Serie
Settembre Blundo D Martin De Vidales JL Alietti A 1992 Corrensite-type clayminerals from Taro Valley (North Italy) Mineralogica et Petrographica Acta 35215ndash222
Sharp ZD Barnes JD 2004 Water-soluble chlorides in massive sea1047298oor serpenti-nites a source of chloride in subduction zones Earth and Planetary Science Letters226 243ndash254
Siever R Woodford N 1979 Dissolution kinetics and the weathering of ma1047297cminerals Geochimica et Cosmochimica Acta 43 717ndash724Singer A 1989 Palygorskite and sepiolite group minerals In Dixon JB Weed SB
(Eds) Minerals in Soil Environments Soil Science Society of America MadisonWisconsin pp 829ndash872
Siffert B Wey R 1962 Synthese d rsquoune sepiolite a temperature ordinarie Comptesrendus hebdomadaires des seacuteances de lAcadeacutemie des Sciences de Paris 2541460ndash1463
Środoń J 1999 Nature of mixed-layer clays and mechanisms of their formation andalteration Annual Reviews of Earth and Planetary Science 27 19ndash53
Stoessell RK 1988 25 degC and 1 atm dissolution experiments of sepiolite and keroliteGeochimica et Cosmochimica Acta 52 365ndash374
Stoessell RK Hay RL 1978 The geochemical origin of Sepiolite and keorilite atAmboseli Kenya Contributions to Mineralogy and Petrology 65 255ndash267
Suarez DL Šimůnek J 1996 Solute transport modeling under variably saturatedwater 1047298ow conditions In Lichtner PC Steefel CI Oelkers EH (Eds) ReactiveTransport in Porous Media Reviews in Mineralogy vol 34 Mineralogical Society of America Michigan Book Crafters Inc 229ndash268
Sumner DY Grotzinger JP 1996 Were kinetics of Archean calcium carbonate
precipitation related to oxygen concentration Geology 24 119ndash122Su C Suarez DL 2004 Boron release from weathering of illites serpentine shales
and illiticpalygorskitic soils Soil Science Society of America Journal 68 96 ndash105Toscani L Venturelli G Boschetti T 2001 Sul1047297de-free and sul1047297de-bearing waters in
the Northern Apennines Italy general features and waterndashrock interaction AquaticGeochemistry 7 195ndash216
Treves BE Harper GD 1994 Exposure of serpentinites on the ocean 1047298oor sequenceof faulting and hydrofracturing in the Northern Apennine ophicalcites O1047297oliti 19b435ndash466
Varadachari C Kudrat M Gosh K 1994 Evaluation of standard free energies of formation of clay minerals by an improved regression method Clays and ClayMinerals 42 298ndash307
Veniale F 1966 Sepiolite in sedimenti dellrsquoAppennino Vogherese (Pavia) Periodico diMineralogia 35 343ndash386
Venturelli G 2003 Acque minerali ed ambiente mdash Fondamenti di geochimica deiprocessi di bassa temperatura Pitagora Editrice Bologna
Venturelli G Contini S Bonazzi A Mangia A 1997 Weathering of ultrama1047297c rocksand element mobility at Mt Prinzera Northern Apennines Italy MineralogicalMagazine 61 765ndash778
Venturelli G Frey M 1977 Anchizone metamorphism in sedimentary sequences of the Northern Apennines Rendiconti Societagrave Mineralogica Italiana 33 109ndash123
Vieillard P 2002 A new method for the prediction of Gibbs free energies of formationofphyllosilicates (10 Aring and14 Aring)basedon theelectronegativityscale Claysand ClayMinerals 50 352ndash363
Vissers RLM Strating EHH 2001 Structures and microstructures in a thrust-relatedgreenschist facies tectonic melange Voltri Group (NW Italy) O1047297oliti 26 33ndash46
Viti C Mellini M 1997 Contrasting chemical compositions in associated lizardite andchrysotile in veins from Elba Italy European Journal of Mineralogy 9 585 ndash596
Viti C Mellini M 1998 Mesh textures and bastites in the Elba retrogradeserpentinites European Journal of Mineralogy 10 1341ndash1359
Wenner DB Taylor Jr HP 1974 Deuteriumatomic hydrogen and oxygen-18oxygen-16 studies of serpentinization of ultrama1047297c rocks Geochimica et CosmochimicaActa 38 1255ndash1286
Williams LA Parks GA Crerar DA 1985 Silica diagenesis I Solubility controls Journal of S edimentary Petrology 55 301ndash311
Wilson J Savage D Cuadros J Shibata M RagnarsdottirKV 2006 The effect of ironon montmorillonite stability (I) Background and thermodynamic considerations
Geochimica et Cosmochimica Acta 70 306ndash322WollastR MackenzieFTBricker OP1968 Experimental precipitation and genesis of sepiolite at earth-surface conditions American Mineralogist 531 1645ndash1662
Wrucke CT1996 Serpentine and carbonated-hosted asbestos deposits InEdwardAdu Bray (Eds) Preliminary compilation of descriptive geoenvironmental mineraldeposit models 95ndash831 US Geological Survey Open-File Report pp 39 ndash46
Xu T Apps JA Pruess K 2001 Analysis of mineral trapping for CO2 disposal in deepaquifers Lawrence Berkeley National Laboratory University of California
ZhangYDaweRA2000 In1047298uence ofMg2+ on thekineticsof calcite precipitation andcalcite crystal morphology Chemical Geology 163 129ndash138
Zhu C Anderson G 2002 Environmental applications of geochemical modelingCambridge University Press
91T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
relative uncertainty of plusmn2 (see Fig 8 in Gunnarsson et al 2005) arecomparable to the logK variation reported for other low-temperatureMg-serpentines (Table 4) that is within the lines Sr1ndash4 for serpentineminerals where samples varying from bicarbonate composition toalkaline composition in the ultrama1047297c aquifer are grouped (Fig 6)Fig 7 shows a plot of calculated serpentine saturation index as a
function of the SiMg molar ratio for the water samples from theultrama1047297tes of the TarondashCeno ValleysThe SI andSiMgvalues of pureMg-serpentine (SI= 0 SiMg=067) provide a useful index forseparating bicarbonate waters (SIb0 SiMgb067) from alkalinewaters (SIN0 SiMgN067) and the distribution of data points isconsistent with the idea that bicarbonate waters approach saturationwith serpentine until they become supersaturated as alkaline waters
Based on the Mgvs SiO2(aq) activitydiagram and speci1047297cratiosoflog[aAl3+(aH+)3] Birsoy (2002) analyzed the behavior of sepiolite duringultrama1047297te weathering The diagram produces a very narrow stability1047297eldfor sepiolite(Fig 8) when it is programmedwith suitable conditionsfor the bicarbonate waters (log PO2=minus25 bar log [aAl3+(aH+)3]= 65) andthe alkaline waters (log PO2=minus50 bar log [aAl3+(aH+)3]= 45) sampled inthe present study This stability 1047297eld grows larger when the absence of
Mg-smectites and talc is assumed but this is an unrealistic conditionbecause (i) all these minerals occur in the rocks within the areainvestigated and (ii) at high [Mg2+][H+]2 activity ratios the 1047297eld of sepiolite is smaller than that of subordinate compared to those poorly-crystalline serpentine (p-antigorite) and clinochlore
Among the alkaline springs of the Voltri Group the waters withlow log [aCa2+(aH+)2] values (Fig 9) show patterns compatible withchlorite interaction (step 5c in Fig6D)The alkaline waters whose datafall near the clinochlore-14Aring 1047297eld (Fig 6D) cannot be explained bysepiolite precipitation because the sepiolite line does not cross thechlorite 1047297elds Thus the peculiar composition of some alkaline watersof the Voltri Group is probably due to interactions between water andprimary chlorites of the ultrama1047297c rocks (optional step 5c in Table 5cb) This is supported by the occurrence sometimes in large amounts
of metasomatic Mgndash(Fe)-chlorite (Scambelluri and Rampone 1999
Vissers and Strating 2001) in the Ligurian area where the relativeabundance of chlorite can reach 50 in veins in high strainserpentinite mylonites of the ErrondashTobbio ultrama1047297tes (Scambelluriet al 2001b) Otherwise in a chlorite-free primary paragenesis thesolution evolving on the Sr3 serpentine line reaches equilibrium withclinochlore but probably the latter does not precipitate due to theslow kinetics of precipitation under these conditions Thus thedissolution of additional reacting rock in the iterative weathering
process supplies Mg
2+
to the solution thereby favoring the Sr3 to Sr4poorly crystalline serpentine transformation In fact all the PR10samples and some of the alkaline springs from the Voltri Group showincreasing [Mg2+][H+]2 activity ratio and low variation in silicaactivity (Fig 6D) starting from the invariant triple point 4a (Sr3-serpentine Fe2+ndashMg-saponite chlinoclore-14 Aring saturation) andextending towards the brucite line This path can be explained asthe result of the transformation of poorly crystalline serpentine fromSr3 to Sr4 the latter is more stable at higher [Mg2+][H+]2 activityratios Finally at very low PO2 values vermiculite disappears and thestability 1047297elds of chlorite and Fe2+ndashMg-saponite become narrowerwhereas those of Fe2+-rich phases (eg Fe2+-saponite) become widerThe data for the alkaline waters with the highest log [aMg2+(aH+)2]from the Voltri Group fall near the stability 1047297eld of peridotite primaryminerals like forsterite and enstatite (Fig 6D) therefore the last stepsof theevolution involve theequilibration of thesolutionwithenstatite(steps 8a and 8c in Table 5ab)
62 Ca-rich alkaline waters
In preliminary modeling of the evolution of the TarondashCeno springwaters Boschetti (2003) concluded that meteoric water initiallysaturated in Fe3+ndashMg2+-chlorite Fe3+ndashMg2+-saponite and pyrite (andor goethite) and then interacted with the ultrama1047297te mineralassemblage eventually evolved into the alkaline waters In themodel of Boschetti (2003) chlorite saponite and calcite were added
Fig 9 Activity plot for the MgOndashCaOndash(CO2)ndashH2O system with phase limits calculated ata temperature of 11 degCSpring waterscoming from theTarondashCeno Valleys (symbols as inFig2) andVoltri Groupultrama1047297tes(symbolsas in Fig5) areshown Calculated path forwaterndashultrama1047297c rock interaction is shown (dotted line) with the number ndashlettercombinations referring to the appearance or disappearance of minerals (see Table 5)The grey area refers to bicarbonate waters from the Voltri Group (Marini and Ottonello2002)
Fig 10 Plots of (A) calcite saturation indexes and (B) PCO2 as a function of Clminus ascalculated by PHREEQCI software for waters from the TarondashCeno Valleys Alkalinesprings for the Voltri Group are also shown (small black diamonds Marini andOttonello 2002) The grey shading is deeper(A) in the saturation1047297eld(SI= 00plusmn05) and
(B) where the lowest log PCO2 value is approached
88 T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
to the solution in order to achieve a high [Ca2+][H+]2 activity ratio andsupersaturation in these phases However in light of the work of Clarket al (1992) and Bruni et al (2002) we believe that the calcitesupersaturation in alkaline waters (Fig 9 and 10) which in our modelis achieved by preventing the precipitation of calcite in the 1047297rst stepsof the reaction path (Table 5a and b) depends on the differencebetween CO2 partial pressureat thegivendepth(PCO2le10minus10bar) andat the surface (PCO2cong10minus35) Waters rising to the surface are far from
equilibrium with the atmosphere where the PCO2 is 4times10
6
timesgreater than the value at depth thus the water rapidly take up CO2which is promptly converted to CO3
2minus because of the high pH Thisleads to supersaturation with calcite through the following hydro-xylation reaction
As a result large amounts of travertine are generated in the VoltriGroup (Marini and Ottonello 2002) and carbonate concretionsknown as devilrsquo s coins are formed at Mt Prinzera in Taro valleyBelow the surface the SIcalcite of the alkaline waters is expected to belower than what the model estimates Under conditions of lowcarbonate alkalinity low CO2 and O2 partial pressure and high pHcalcite nucleation may be inhibited and the kinetics of precipitationreduced by the high concentrations of dissolved divalent cations suchas Mg2+ Fe2+ and Mn2+ (Sumner and Grotzinger 1996 Davis et al2000 Zhang and Dawe 2000 de Leeuw 2002) The fact that severalalkaline water samples fall within or near the calcitendashdolomiteequilibrium (Fig 9) may indicate magnesium calcite or protodolomitenucleation in those extreme subsurface conditions However possibleinteraction between some calcium-rich waters of the Voltri Group andCa-rich rocks such as ophicarbonates or rodingites cannot beexcluded these rocks are widespread in the Ligurian Apennineserpentinites (Treves and Harper 1994)
63 Sepiolite vs serpentine nucleation under weathering conditions
constraints on the kinetics and equilibrium-based approaches
The kinetics of precipitation of poorly-crystalline serpentine and
sepiolite cannot be compared quantitatively because rate constantsare available only for well-ordered crystalline phases (eg see Palandriand Kharaka 2004 for crystalline serpentine) Sepiolite precipitationis often invoked to explain magnesium scavenging from solutions butthe lack of kinetic data for this mineral leads to overly simplisticreaction transport models These models are forced to use kinetic datafor phyllosilicates such as kaolinite or muscovite (eg Xu et al 2001Bryan 2005) Moreover the low-temperature thermodynamic dataeven if they are fairly uncertain show that sepiolite solubilityincreases with decreasing crystallinity (Table 4 Stoessell 1988) Infact crystalline sepiolite requires a long time to reach equilibriumwith solution which imposes a slow control on thewater composition(eg Stoessell and Hay 1978) Consequently in modeling based onreaction transport that simulates freshly precipitated or poorly
crystalline sepiolite formation the logK value of Wollast et al (1968)is preferable to the lower values (Suarez and Šimůnek 1996) Thispreference agrees with the approach of Mariner (2003) and Jove Colon(2006) who adopt a logK (25 degC) value for sepiolite of 3044 It is alsosimilar to that reported by Wollast et al (1968) and it is six logK unitshigherthan the valueused in thermocomV8R6 dataset which is oftenused (perhaps improperly) to model the supergenic nucleation of sepiolite In the stability diagram in Fig 6 where the Sp1 and Sp2 linesrepresent the stability boundaries of crystalline and poorly-crystallinesepiolite respectively the evolution of TarondashCeno waters frombicarbonate to alkaline (Figs 6A and C) follows the stability 1047297eldboundary of the poorly-crystalline serpentine (between Sr2ndashSr3 lines)rather than that of the poorly-crystalline sepiolite (Sp2) as postulatedin themodel The lines for Sp2 Sr3 and amorphoussilica intersect at a
common point (Fig 6D) which is consistent with the nucleation of
sepiolite that typically occurs in solutions that are saturated orsupersaturated with amorphous silica and (Mg)-carbonate (egStoessell and Hay 1978 Birsoy 2002) Some bicarbonate watersfrom the Voltri Group have a silica activity near the amorphous silicaboundary (Fig 6B) but only a few alkaline springs of this group fall onthe poorly-crystalline sepiolite boundary line (Sp2 Fig 6D) Both of these features are inconsistent with the possibility that sepioliteprecipitates from supergenic fresh water in the Northern Apennine
Moreover these two results suggest that values for logK(sepiolite)b
3044 (see Table 4) should be used to model the dissolutionprecipitation of the well-crystallized phase that is the most commonform for the metasomatic sepiolite found in the investigated area
7 Conclusions
The waters issuing from ultrama1047297tes in the TarondashCeno Valleysoriginate from low-temperature interactions between meteoric waterand serpentinite rocks In this report the waterndashrock interaction hasbeen modeled using reaction paths in progress mode During their earlyinteraction with soil and rock minerals bicarbonate waters super-saturated in kaolinite ferrihydrite vermiculite and Fe2+ndashMg2+-saponitewhere theorigin of the secondary iron saponite as Fe2+-bearing smectiteagrees with the hypothesis of Guumlven (1988) Then while remainingsupersaturated with saponite the bicarbonate waters evolve into analkaline facies when poorly-crystalline serpentine precipitates Duringrock weathering the Al-bearing crystalline serpentine of oceanic originprobably promotes nucleation of the low-temperature serpentineamorphous phase and supplies Al3+ to the solution for saponiteprecipitation This interpretation is consistent with the existence of Al-poor supergenic serpentine (Neal and Stanger 1985) and the partiallymeteoric OH-isotopic imprint observed in other ultrama1047297c systems
Dif 1047297culties and complexities in the modeling arise fromuncertainties in the primary mineralogy of some local ophiolitesand from the presence of different lithotypes (rodingites ophi-carbonates gabbros) which can signi1047297cantly affect the composi-tion of water samples This can be seen in the example of the Candash
OH springs issuing from the Voltri Group (PO2 and PCO2 of 10minus70
and 10minus10 bar respectively Bruni et al 2002 Marini and
Ottonello 2002) These springs appear to be more evolved thanthe TarondashCeno alkaline waters as a result of their particularly longinteractions with serpentinites until equilibrium with enstatite isreached or with chlorite-bearing serpentinite The use of anupdated thermodynamic database in our model avoided problemsdue to mineralogical heterogeneity in the mixed-layer claysassociated with ophiolitic rocks this database accurately re1047298ectedthe chemical composition of the single phases Using this modelthe vermiculitendashchlorite and saponitendashchlorite associations seen inthe activity diagrams simulate respectively the presence of highand low charge corrensite observed in the 1047297eld The accuracy of ourmodel can be tested by comparing estimates of chlorine composi-tion resulting from our model and from experimental methods The
mean Cl concentration of the bicarbonate water is 41plusmn16 mgLtaking into account a Cl concentration of sim2 mgL in rainwater inthe area (cf Panettiere et al 2000 Venturelli 2003) 2 mgL of Clin spring water are estimated to come from serpentinite This valueagrees with the model which predicts that the switch frombicarbonate water to alkaline water requires the congruentdissolution of 00203 mol of serpentinite containing 100 ppm of soluble Cl thereby yielding water containing 2 ppm of Cl
In addition we think that weathering of the accessory albiteplagioclase is responsible for Na enrichment of the waters whereas Clenrichment in the alkaline waters is predicted to result from theleaching of NaCl 1047298uid inclusions Moreover the TarondashCeno alkalinesprings show the highest BCl ratios comparable with those of thehost metamorphic minerals (Scambelluri et al 2004) In contrast the
lower boron content and BCl ratio of the alkaline springs of the Voltri
89T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
Group are unexpected features and probably re1047298ect waterndashrockinteractions over a longer time period than in the TarondashCeno ValleysThe boron content in the Northern Apennine alkaline springs may becontrolled by several factors such as (i) the different histories of VoltriGroup and EL ultrama1047297tes (ii) the probable contribution of boronfrom NaCl 1047298uid inclusions (iii) the sinking of boron by travertineprecipitation (Hobbs and Reardon1999) brucite (Rhoades et al 1970)and smectite and (iv) decreases in boron release from minerals due to
increases in pH (Su and Suarez 2004) Acknowledgements
This research was carried out with the 1047297nancial support from theldquoYoung Researchers Projectrdquo (2001ndash2002) of the Italian Ministry of Universities Scienti1047297c Research and Technology (MURST) We thankGianniCortecci (Institute of Geosciences and EarthResources IGG-CNRPisa Italy) and Giampiero Venturelli (Earth Sciences Department Univer-sity of Parma Italy) for useful suggestions and critical review of themanuscript The authors are grateful to Enrico Dinelli (InterdepartmentResearch Centre for Environmental Sciences Alma Mater StudiorumUniversity of Bologna Ravenna Italy) for granting permission to use theGeochemistrsquos WorkbenchsoftwareTheconstructive reviews of theEditor
J Fein and of two anonymous Referees have signi1047297cantly improved the
manuscript
Appendix A Supplementary data
Supplementary data associated with this article can be found inthe online version at doi101016jchemgeo200808017
References
Abbate E Bortolotti V Passerini P 1980 Olistostromes and olistoliths In Sestini G(Ed) Development of the Northern Apennines geosyncline Sedimentary Geology4 pp 521ndash558
Alietti A 1956 Il minerale a strati misti saponite-talco di monte Chiaro (val di TaroAppen Emil) Rendiconti dellAccademia Nazionale dei Lincei 21 201 ndash207
Alietti A De Angelis G Montanari G Sgarlata F Tirelli G 1976 Il ferro nelle argilleappenniniche Periodico di Mineralogia 45 51ndash64
Anselmi B Mellini M Viti C 2000 Chlorine in the Elba Monti Livornesi and Murloserpentines Evidence for seandashwater interaction European Journal of Mineralogy12137ndash146
Apha-Awwa-Wef 1995 Standard Methods for the Examination of Water and Waste-water 19th Edition American Public Health Association American Water WorksAssociation Water Environment Federation
Barnes I LaMarche Jr VC Himmelberg GR 1967 Geochemical evidence of present-day serpentinization Science 156 830ndash832
Barnes I ONeil JR 1969 The relationship between 1047298uids in some fresh alpine-typeultrama1047297cs and possible modern serpentinization Western United StatesGeological Society of America Bulletin 80 1947ndash1960
Barnes I ONeil JRTrescases JJ1978 Present dayserpentinizationin New CaledoniaOman and Yugoslavia Geochimica et Cosmochimica Acta 42 144ndash145
Barnes I Rapp JB ONeil JR Sheppard RA Gude III AJ 1972 Metamorphicassemblages and the direction of 1047298ow of metamorphic 1047298uids in four instances of serpentinization Contributions to Mineralogy and Petrology 35 263ndash276
Barnes JDSelverstone J Sharp ZD 2006Chlorine isotope chemistry of serpentinitesfrom Elba Italy as an indicator of 1047298uid source and subsequent tectonic historyGeochemistry Geophysics Geosystems 7 doi1010292006GC001296
Beccaluva L Emiliani F Venturelli G Zerbi M 1973 Ca Fe Mg Mn Cr Ni Codistribution in some ultrama1047297c rocks outcropping in the Northern Apennines withsome geological remarks Ateneo Parmense Acta Naturalia 9 69ndash98
Beccaluva L Macciotta G Piccardo GB Zeda O 1984 Petrology of lherzolitic rocksfrom the Northern Apennine ophiolites Lithos 17 299 ndash316
BeccaluvaL Venturelli G ZanzucchiG 1975 Dati geologici e geochimici sui basalti diassociazione o1047297olitica dellAppennino ligure-emiliano Ateneo Parmense ActaNaturalia 11 789ndash802
BethkeCM 2002 The Geochemists Workbench A Userrsquos Guide to RxnAct2Tact ReactandGtplot Release40 HydrogeologyProgram Universityof IllinoisUrbanaIL 236pp
Birsoy R 2002Formation of sepiolitendashpalygorskite and related minerals from solutionClays and Clay Minerals 50 736ndash745
Bonatti E Lawrence JR Morandi N 1984 Serpentinization of oceanic peridotitestemperature dependence of mineralogy and boron content Earth and PlanetaryScience Letters 70 88ndash94
Bonifacio E Zanini E Boero V Franchini-Angela M 1997 Pedogenesis in a soilcatena on serpentinite in north-western Italy Geoderma 75 33ndash51
Boschetti T 2003 Studio geochimico e geochimico-isotopico di acque a composizione
estrema e termali dellrsquoAppennino Settentrionale Ph D Thesis Parma Univ
Brigatti MF Franchini G Lugli C Medici L Poppi L Turci E 2000 Interactionsbetween aqueous chromium solutions and layer silicatesApplied Geochemistry 151307ndash1316
Brigatti MF Lugli C Poppi L Venturelli G 1999 Iron-rich saponite dissolutionreactions and Cr uptake Clay Minerals 34 637ndash645
Brigatti MF Valdregrave G 1996 Mixed-layer chloritesmectite (corrensite) in ophioliticrock alteration products Neues Jahrbuch fuer Mineralogie Monatshefte 1 37 ndash47
Bruni J Canepa M Chiodini G Cioni R Cipolli F Longinelli A Marini L OttonelloG Vetuschi Zuccolini M 2002 Irreversible waterndashrock mass transfer accompany-ing the generation of the neutral MgndashHCO3 and high-pH CandashOH spring waters of the Genova province Italy Applied Geochemistry 17 455ndash474
Bryan CR 2005 Drift-Scale THC Seepage Model MDL-NBS-HS-000001 REV 04 YuccaMountain Project Las Vegas Nevada (US)Caruso LJ Chernosky JV 1979 The stability of lizardite Canadian Mineralogist 17
757ndash769Clark ID Fontes JC Fritz P 1992 Stable isotope disequilibria in travertine from high
pH waters Laboratory investigations and 1047297eld observations from Oman Geochi-mica et Cosmochimica Acta 56 2041ndash2050
Cortesogno L 1980 Il metamor1047297smo giurassico nelle o1047297oliti dellAppenninoSettentrionale due distinti cicli metamor1047297ci in ambiente oceanico O1047297oliti 5 5ndash18
Christ CLHostetler PB Siebert RM1973 Studies in the system MgO-SiO2-CO2-H2OIII American Journal of Science 273 65ndash83
Cortesogno L Lucchetti G 1982 Il metamor1047297smo oceanico nei gabbri o1047297oliticidellAppennino Ligure Aspetti mineralogici e petrogenetici Rendiconti dellaSocietagrave Mineralogica Italiana 38 561ndash579
Crispini L Capponi G 2002 Control of metasomatic alteration on the development of shear zone in ultrama1047297c rocks a case study from the Voltri Massif (Ligurian Alps)Architettura interna e cinematica delle zone di taglio Gruppo Italiano di GeologiaStrutturale mdash Riunione Annuale 2002 Pisa Italy 11ndash12 giugno 2002 XIV ndashXV
CrissRE 1999 Principlesof stableisotope distributionOxford University PressNew York
Davis KJ Dove PM De Yoreo JJ 2000 The role of Mg 2+ as an impurity in calcitegrowth Science 290 1134ndash1137
de Leeuw NH 2002 Moleculardynamics simulations of the growth inhibiting effect of Fe2+ Mg2+ Cd2+ and Sr2+ on calcite crystal growth Journal of Physical Chemistry B106 5241ndash5249
De Nardo MT Segadelli S Vescovi P 2007 Studio pilota per la caratterizzazionegeologicadellesorgentinellazona delM Nero (altaVal Cenoe alta ValNuremdash provincedi Parma e Piacenza) Il Geologo dellEmilia-Romagna 25 Nuova Serie pp 5 ndash21
Di Giulio A Geddo G 1990 Studio petrogra1047297co delle Arenarie di Casanova (alta valTrebbia Appennino settentrionale) Atti Ticinesi di Scienze della Terra 33 243ndash254
Dinelli E Lombini A Simoni A Ferrari C 1997 Heavy metals in serpentinite soils of selected outcrops of Piacenza and Parma provinces (Northern Apennines Italy)Mineralogica et Petrographica Acta 40 241ndash255
Epstein S Mayeda T 1953 Variation of 18O content of waters from natural sourcesGeochimica et Cosmochimica Acta 4 213ndash224
Ernst WG Piccardo GB 1979 Petrogenesis of some Ligurian peridotites I Mineraland bulk-rock chemistry Geochimica et Cosmochimica Acta 43 219ndash237
Feder F Trolard F Klingelhofer G Bourrieacute G 2005 In situ Moumlssbauer spectroscopy
evidence for green rust (fougerite) in a gleysol and its mineralogical transforma-tions with time and depth Geochimica et Cosmochimica Acta 69 4463ndash4483Fritz PClark IDFontes JCWhiticar MJFaberE 1992 Deuterium and
13C evidence
for low temperature production of hydrogen and methane in a highly alkalinegroundwater environment in Oman In Kharaka TK Maest AS (Eds) Proceed-ings of the 7th International Symposium on WaterndashRock Interaction BalkemaRotterdam pp 793ndash796
Giammetti F 1968 Le o1047297oliti di Roccaprebalza Costa della Guardia e Groppo di Gorro(Appennino Parmense) Ateneo Parmense Acta Naturalia 4 107ndash142
Giuseppetti G1953 Roccee minerali dellaformazioneo1047297olitica di Volpedo Rendicontidella Societagrave Mineralogica Italiana 9 85ndash134
Gran G 1952 Determination of equivalence point in the potentiometric titrationAnalyst 77 661ndash671
Gunnarsson I Arnorsson S Jakobsson S 2005 Precipitation of poorly crystalline antigoriteunder hydrothermal conditions Geochimica et Cosmochimica Acta 69 2813ndash2828
Guumlven N 1988 Smetites In Bailey SW (Ed) Reviews in Mineralogy 19 Hydrousphyllosilicates (exclusive of micas) Mineralogical Society of America MichiganBook Crafters Inc 497ndash559
Hansel CM Benner SG Fendorf S 2005 Competing Fe(II)-induced mineralization
pathways of ferrihydrite Environmental Science and Technology 39 7147ndash7153HobbsMYReardonEJ 1999 Effect of pH on boronco-precipitation by calcite further
evidence for non-equilibrium partitioning of trace elements Geochimica etCosmochimica Acta 63 1013ndash1021
Jones BF Galan E 1988 Sepiolite and palygorskite In Bailey SW (Ed) HydrousPhyllosilicates (exclusive of Micas) Reviews in Mineralogy vol 19 MineralogicalSociety of America Michigan Book Crafters Inc 631ndash674
Jove Colon CF 2006 Technical work plan for Thermodynamic database for chemicalmodelling TWP-MGR-PA-000039 REV 01 Yucca Mountain Project Las VegasNevada (US)
Kohls DW Rodda Jr JL 1967 Iowaite a new hydrous magnesium hydroxidendashferricoxychloride from the Precambrian of Iowa American Mineralogist 52 1261 ndash1271
Kyser TK OHanley DS Wicks FJ 1999 The origin of 1047298uids associated withserpentinization processes evidence from stable-isotope compositions CanadianMineralogist 37 223ndash237
Lapham DM 1961 New data on deweylite American Mineralogist 46 168ndash188Lee BD SearsSKGraham RC Amrhein CValiH 2003 Secondarymineral genesis
from chlorite and serpentine in an ultrama1047297c soil toposequence Soil Science
Society of America Journal 671309ndash1317
90 T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
Longinelli A Selmo E 2003 Isotopic composition of precipitation in Italy a 1047297rstoverall map Journal of Hydrology 270 75ndash88
Majzlan J Navrotsky A Schwertmann U 2004 Thermodynamics of iron oxides PartIIIEnthalpiesof formation and stability of ferrihydrite (simFe(OH)3) schwertmannite(simFeO(OH)34(SO4)18) and ɛ-Fe2O3 Geochimica et Cosmochimica Acta 681049ndash1059
Mariner P 2003 In-Drift PrecipitatesSalts Model ANL-EBS-MD-000045 REV 01 YuccaMountain Project Las Vegas Nevada (US)
Marini L Ottonello G 2002 Atlante degli acquiferi della Liguria Volume III Le acquedeicomplessio1047297olitici (bacini ArrestraBranega Cassinelle Cerusa ErroGorzenteLeira Lemme Lerone Orba Piota Polcevera Rumaro Sansobbia Stura Teiro
Varenna Visone) Pacini PisaMellini L Zanazzi PF 1987 Crystal structures of lizardite-l T and lizardite-2H1 fromColi Italy American Mineralogist 72 943ndash948
Mezzadri G 1964 Petrogra1047297a delle arenarie di Ostia Rendiconti Societagrave MineralogicaItaliana 20 192ndash228
Mondesir H Decarreau A 1987 Synthesegrave entre 25 et 200 degC de lizardites Ni ndashMgMesure des coef 1047297cients de partage solide-solution aqueuse pour le couple Ni ndashMgdans des lizardites Bulletin de Mineralogie 110 409ndash426
Neal C Shand P 2002 Spring and surface water quality of the Cyprus ophiolitesHydrology and Earth System Sciences 6 797ndash817
Neal C Stanger G 1983 Hydrogen generation from mantle source rocks in OmanEarth and Planetary Science Letters 66 315ndash320
Neal C Stanger G 1984 Calcium and magnesium hydroxide precipitation fromalkaline groundwaters in Oman and their signi1047297cance to the process of serpentinization Mineralogical Magazine 48 237ndash241
Neal C Stanger G 1985 Past and present serpentinisation of ultrama1047297c rocks anexample from Semail ophiolite nappe of Northern Oman In Drever JI (Ed) TheChemistry of Weathering D Reidel Publishing Company pp 249ndash275
Nesbitt HW Bricker OP 1978 Low temperature alteration processes affecting
et Cosmochimica Acta 41 1835ndash1841OrsquoHanley DS 1996 Serpentinites records of tectonic and petrological history Oxford
University PressOrberger B Friedrich G Woermann E 1990 The distribution of halogens and carbon
in Pge-bearing ultrama1047297cs of the Acoje Ophiolite Block Zambales Philippines Journal of Geochemical Exploration 37 147ndash169
Ottonello G Joron JL Piccardo GB 1984 Rare Earth and 3d transition elementsgeochemistry of peridotitic rocks II Ligurian peridotites and associated basalts Journal of Petrology 25 373ndash393
Palandri JL Kharaka YK 2004 A compilation of rate parameters of water-mineralinteraction kinetics for application to geochemical modeling Open File Report2004-1068 US Geological Survey
Panettiere P Cortecci G Dinelli E Bencini A Guidi M 2000 Chemistry and sulfurisotopic composition of precipitation at Bologna Italy Applied Geochemistry 151455ndash1467
Parkhurst DL Appelo CAJ 1999 Userrsquos guide to PHREEQC (version 2) mdash a computerprogram for speciation batch-reaction one-dimensional transport and inversegeochemical calculation US Department of the Interior US Geological SurveyWashington DC
Pfeifer HR 1977 A model for 1047298uids in metamorphosed ultrama1047297c rocks Observationsat surface and subsurface conditions (high pH spring waters) SchweizerischeMineralogische und Petrographische Mitteilungen 57 361ndash396
PlummerLN1984Geochemical modeling A comparisonof forwardand inverse methodsIn Hitchon B Wallick EI (Eds) Proceedings First CanadianAmerican Conference onHydrogeology-Practical Applications of Ground Water Geochemistry Banff AlbertaCanada mdash Worthington Ohio National Water Well Association pp 149ndash177
Rampone E Hofmann AW Piccardo GB Vannucci R Bottazzi P Ottolini L 1995Petrology mineral and isotope geochemistry of the External Ligurides peridotites(Northern Apennines Italy) Journal of Petrology 36 81ndash105
Rampone E Hofmann AW Piccardo GB Vannucci R Bottazzi P Ottolini L 1996Trace element and isotope geochemistry of depleted peridotites from an N-MORBtype ophiolite (Internal Liguride N Italy) Contributions to Mineralogy andPetrology 123 61ndash76
Rhoades JD Ingvalson RD HatcherJT1970Adsorption of boronby ferromagnesianminerals and magnesium hydroxide Soil Science Society of America Journal 34
938ndash941Saini-EidukatB KuchaH Keppler H1994Hibbingiteγ-Fe2(OH)3Clanewmineralfromthe Duluth Complex Minnesota with implications for the oxidation of Fe-bearingcompounds and the transport of metals American Mineralogist 79 555ndash561
Sanford RF 1981 Mineralogical and chemical effects of hydration reactions andapplications to serpentinization American Mineralogist 66 290ndash297
Scambelluri M Muumlntener O Ottolini L Pettke TT Vannucci R 2004 The fate of BCl and Li in the subducted oceanic mantle and in the antigorite breakdown 1047298uidsEarth and Planetary Science Letters 222 217ndash234
Scambelluri M Rampone E 1999 Mg-metasomatism of oceanic gabbros and itscontrol on Ti-clinohumite formation during eclogitization Contributions toMineralogy and Petrology 135 1ndash17
Scambelluri M Bottazzi P Trommsdorff V Vannucci R Hermann J Gogravemez-Pugnaire MT Logravepez-Sagravenchez Vizcaino V 2001a Incompatible element-rich1047298uidsreleased by antigorite breakdown in deeply subducted mantle Earth and PlanetaryScience Letters 192 457ndash470
Scambelluri M Rampone E Piccardo G 2001b Fluid and element cycling insubducted serpentinite a trace-element study of the ErrondashTobbio high-pressureultrama1047297tes (Western Alps NW Italy) Journal of Petrology 42 55ndash67
Segadelli S2006 Lageologia nelpaesaggio le rupi o1047297olitiche in Val Taro e Val Ceno IlGeologo dellrsquoEmilia-Romagna 22 15ndash29 Nuova Serie
Settembre Blundo D Martin De Vidales JL Alietti A 1992 Corrensite-type clayminerals from Taro Valley (North Italy) Mineralogica et Petrographica Acta 35215ndash222
Sharp ZD Barnes JD 2004 Water-soluble chlorides in massive sea1047298oor serpenti-nites a source of chloride in subduction zones Earth and Planetary Science Letters226 243ndash254
Siever R Woodford N 1979 Dissolution kinetics and the weathering of ma1047297cminerals Geochimica et Cosmochimica Acta 43 717ndash724Singer A 1989 Palygorskite and sepiolite group minerals In Dixon JB Weed SB
(Eds) Minerals in Soil Environments Soil Science Society of America MadisonWisconsin pp 829ndash872
Siffert B Wey R 1962 Synthese d rsquoune sepiolite a temperature ordinarie Comptesrendus hebdomadaires des seacuteances de lAcadeacutemie des Sciences de Paris 2541460ndash1463
Środoń J 1999 Nature of mixed-layer clays and mechanisms of their formation andalteration Annual Reviews of Earth and Planetary Science 27 19ndash53
Stoessell RK 1988 25 degC and 1 atm dissolution experiments of sepiolite and keroliteGeochimica et Cosmochimica Acta 52 365ndash374
Stoessell RK Hay RL 1978 The geochemical origin of Sepiolite and keorilite atAmboseli Kenya Contributions to Mineralogy and Petrology 65 255ndash267
Suarez DL Šimůnek J 1996 Solute transport modeling under variably saturatedwater 1047298ow conditions In Lichtner PC Steefel CI Oelkers EH (Eds) ReactiveTransport in Porous Media Reviews in Mineralogy vol 34 Mineralogical Society of America Michigan Book Crafters Inc 229ndash268
Sumner DY Grotzinger JP 1996 Were kinetics of Archean calcium carbonate
precipitation related to oxygen concentration Geology 24 119ndash122Su C Suarez DL 2004 Boron release from weathering of illites serpentine shales
and illiticpalygorskitic soils Soil Science Society of America Journal 68 96 ndash105Toscani L Venturelli G Boschetti T 2001 Sul1047297de-free and sul1047297de-bearing waters in
the Northern Apennines Italy general features and waterndashrock interaction AquaticGeochemistry 7 195ndash216
Treves BE Harper GD 1994 Exposure of serpentinites on the ocean 1047298oor sequenceof faulting and hydrofracturing in the Northern Apennine ophicalcites O1047297oliti 19b435ndash466
Varadachari C Kudrat M Gosh K 1994 Evaluation of standard free energies of formation of clay minerals by an improved regression method Clays and ClayMinerals 42 298ndash307
Veniale F 1966 Sepiolite in sedimenti dellrsquoAppennino Vogherese (Pavia) Periodico diMineralogia 35 343ndash386
Venturelli G 2003 Acque minerali ed ambiente mdash Fondamenti di geochimica deiprocessi di bassa temperatura Pitagora Editrice Bologna
Venturelli G Contini S Bonazzi A Mangia A 1997 Weathering of ultrama1047297c rocksand element mobility at Mt Prinzera Northern Apennines Italy MineralogicalMagazine 61 765ndash778
Venturelli G Frey M 1977 Anchizone metamorphism in sedimentary sequences of the Northern Apennines Rendiconti Societagrave Mineralogica Italiana 33 109ndash123
Vieillard P 2002 A new method for the prediction of Gibbs free energies of formationofphyllosilicates (10 Aring and14 Aring)basedon theelectronegativityscale Claysand ClayMinerals 50 352ndash363
Vissers RLM Strating EHH 2001 Structures and microstructures in a thrust-relatedgreenschist facies tectonic melange Voltri Group (NW Italy) O1047297oliti 26 33ndash46
Viti C Mellini M 1997 Contrasting chemical compositions in associated lizardite andchrysotile in veins from Elba Italy European Journal of Mineralogy 9 585 ndash596
Viti C Mellini M 1998 Mesh textures and bastites in the Elba retrogradeserpentinites European Journal of Mineralogy 10 1341ndash1359
Wenner DB Taylor Jr HP 1974 Deuteriumatomic hydrogen and oxygen-18oxygen-16 studies of serpentinization of ultrama1047297c rocks Geochimica et CosmochimicaActa 38 1255ndash1286
Williams LA Parks GA Crerar DA 1985 Silica diagenesis I Solubility controls Journal of S edimentary Petrology 55 301ndash311
Wilson J Savage D Cuadros J Shibata M RagnarsdottirKV 2006 The effect of ironon montmorillonite stability (I) Background and thermodynamic considerations
Geochimica et Cosmochimica Acta 70 306ndash322WollastR MackenzieFTBricker OP1968 Experimental precipitation and genesis of sepiolite at earth-surface conditions American Mineralogist 531 1645ndash1662
Wrucke CT1996 Serpentine and carbonated-hosted asbestos deposits InEdwardAdu Bray (Eds) Preliminary compilation of descriptive geoenvironmental mineraldeposit models 95ndash831 US Geological Survey Open-File Report pp 39 ndash46
Xu T Apps JA Pruess K 2001 Analysis of mineral trapping for CO2 disposal in deepaquifers Lawrence Berkeley National Laboratory University of California
ZhangYDaweRA2000 In1047298uence ofMg2+ on thekineticsof calcite precipitation andcalcite crystal morphology Chemical Geology 163 129ndash138
Zhu C Anderson G 2002 Environmental applications of geochemical modelingCambridge University Press
91T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
to the solution in order to achieve a high [Ca2+][H+]2 activity ratio andsupersaturation in these phases However in light of the work of Clarket al (1992) and Bruni et al (2002) we believe that the calcitesupersaturation in alkaline waters (Fig 9 and 10) which in our modelis achieved by preventing the precipitation of calcite in the 1047297rst stepsof the reaction path (Table 5a and b) depends on the differencebetween CO2 partial pressureat thegivendepth(PCO2le10minus10bar) andat the surface (PCO2cong10minus35) Waters rising to the surface are far from
equilibrium with the atmosphere where the PCO2 is 4times10
6
timesgreater than the value at depth thus the water rapidly take up CO2which is promptly converted to CO3
2minus because of the high pH Thisleads to supersaturation with calcite through the following hydro-xylation reaction
As a result large amounts of travertine are generated in the VoltriGroup (Marini and Ottonello 2002) and carbonate concretionsknown as devilrsquo s coins are formed at Mt Prinzera in Taro valleyBelow the surface the SIcalcite of the alkaline waters is expected to belower than what the model estimates Under conditions of lowcarbonate alkalinity low CO2 and O2 partial pressure and high pHcalcite nucleation may be inhibited and the kinetics of precipitationreduced by the high concentrations of dissolved divalent cations suchas Mg2+ Fe2+ and Mn2+ (Sumner and Grotzinger 1996 Davis et al2000 Zhang and Dawe 2000 de Leeuw 2002) The fact that severalalkaline water samples fall within or near the calcitendashdolomiteequilibrium (Fig 9) may indicate magnesium calcite or protodolomitenucleation in those extreme subsurface conditions However possibleinteraction between some calcium-rich waters of the Voltri Group andCa-rich rocks such as ophicarbonates or rodingites cannot beexcluded these rocks are widespread in the Ligurian Apennineserpentinites (Treves and Harper 1994)
63 Sepiolite vs serpentine nucleation under weathering conditions
constraints on the kinetics and equilibrium-based approaches
The kinetics of precipitation of poorly-crystalline serpentine and
sepiolite cannot be compared quantitatively because rate constantsare available only for well-ordered crystalline phases (eg see Palandriand Kharaka 2004 for crystalline serpentine) Sepiolite precipitationis often invoked to explain magnesium scavenging from solutions butthe lack of kinetic data for this mineral leads to overly simplisticreaction transport models These models are forced to use kinetic datafor phyllosilicates such as kaolinite or muscovite (eg Xu et al 2001Bryan 2005) Moreover the low-temperature thermodynamic dataeven if they are fairly uncertain show that sepiolite solubilityincreases with decreasing crystallinity (Table 4 Stoessell 1988) Infact crystalline sepiolite requires a long time to reach equilibriumwith solution which imposes a slow control on thewater composition(eg Stoessell and Hay 1978) Consequently in modeling based onreaction transport that simulates freshly precipitated or poorly
crystalline sepiolite formation the logK value of Wollast et al (1968)is preferable to the lower values (Suarez and Šimůnek 1996) Thispreference agrees with the approach of Mariner (2003) and Jove Colon(2006) who adopt a logK (25 degC) value for sepiolite of 3044 It is alsosimilar to that reported by Wollast et al (1968) and it is six logK unitshigherthan the valueused in thermocomV8R6 dataset which is oftenused (perhaps improperly) to model the supergenic nucleation of sepiolite In the stability diagram in Fig 6 where the Sp1 and Sp2 linesrepresent the stability boundaries of crystalline and poorly-crystallinesepiolite respectively the evolution of TarondashCeno waters frombicarbonate to alkaline (Figs 6A and C) follows the stability 1047297eldboundary of the poorly-crystalline serpentine (between Sr2ndashSr3 lines)rather than that of the poorly-crystalline sepiolite (Sp2) as postulatedin themodel The lines for Sp2 Sr3 and amorphoussilica intersect at a
common point (Fig 6D) which is consistent with the nucleation of
sepiolite that typically occurs in solutions that are saturated orsupersaturated with amorphous silica and (Mg)-carbonate (egStoessell and Hay 1978 Birsoy 2002) Some bicarbonate watersfrom the Voltri Group have a silica activity near the amorphous silicaboundary (Fig 6B) but only a few alkaline springs of this group fall onthe poorly-crystalline sepiolite boundary line (Sp2 Fig 6D) Both of these features are inconsistent with the possibility that sepioliteprecipitates from supergenic fresh water in the Northern Apennine
Moreover these two results suggest that values for logK(sepiolite)b
3044 (see Table 4) should be used to model the dissolutionprecipitation of the well-crystallized phase that is the most commonform for the metasomatic sepiolite found in the investigated area
7 Conclusions
The waters issuing from ultrama1047297tes in the TarondashCeno Valleysoriginate from low-temperature interactions between meteoric waterand serpentinite rocks In this report the waterndashrock interaction hasbeen modeled using reaction paths in progress mode During their earlyinteraction with soil and rock minerals bicarbonate waters super-saturated in kaolinite ferrihydrite vermiculite and Fe2+ndashMg2+-saponitewhere theorigin of the secondary iron saponite as Fe2+-bearing smectiteagrees with the hypothesis of Guumlven (1988) Then while remainingsupersaturated with saponite the bicarbonate waters evolve into analkaline facies when poorly-crystalline serpentine precipitates Duringrock weathering the Al-bearing crystalline serpentine of oceanic originprobably promotes nucleation of the low-temperature serpentineamorphous phase and supplies Al3+ to the solution for saponiteprecipitation This interpretation is consistent with the existence of Al-poor supergenic serpentine (Neal and Stanger 1985) and the partiallymeteoric OH-isotopic imprint observed in other ultrama1047297c systems
Dif 1047297culties and complexities in the modeling arise fromuncertainties in the primary mineralogy of some local ophiolitesand from the presence of different lithotypes (rodingites ophi-carbonates gabbros) which can signi1047297cantly affect the composi-tion of water samples This can be seen in the example of the Candash
OH springs issuing from the Voltri Group (PO2 and PCO2 of 10minus70
and 10minus10 bar respectively Bruni et al 2002 Marini and
Ottonello 2002) These springs appear to be more evolved thanthe TarondashCeno alkaline waters as a result of their particularly longinteractions with serpentinites until equilibrium with enstatite isreached or with chlorite-bearing serpentinite The use of anupdated thermodynamic database in our model avoided problemsdue to mineralogical heterogeneity in the mixed-layer claysassociated with ophiolitic rocks this database accurately re1047298ectedthe chemical composition of the single phases Using this modelthe vermiculitendashchlorite and saponitendashchlorite associations seen inthe activity diagrams simulate respectively the presence of highand low charge corrensite observed in the 1047297eld The accuracy of ourmodel can be tested by comparing estimates of chlorine composi-tion resulting from our model and from experimental methods The
mean Cl concentration of the bicarbonate water is 41plusmn16 mgLtaking into account a Cl concentration of sim2 mgL in rainwater inthe area (cf Panettiere et al 2000 Venturelli 2003) 2 mgL of Clin spring water are estimated to come from serpentinite This valueagrees with the model which predicts that the switch frombicarbonate water to alkaline water requires the congruentdissolution of 00203 mol of serpentinite containing 100 ppm of soluble Cl thereby yielding water containing 2 ppm of Cl
In addition we think that weathering of the accessory albiteplagioclase is responsible for Na enrichment of the waters whereas Clenrichment in the alkaline waters is predicted to result from theleaching of NaCl 1047298uid inclusions Moreover the TarondashCeno alkalinesprings show the highest BCl ratios comparable with those of thehost metamorphic minerals (Scambelluri et al 2004) In contrast the
lower boron content and BCl ratio of the alkaline springs of the Voltri
89T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
Group are unexpected features and probably re1047298ect waterndashrockinteractions over a longer time period than in the TarondashCeno ValleysThe boron content in the Northern Apennine alkaline springs may becontrolled by several factors such as (i) the different histories of VoltriGroup and EL ultrama1047297tes (ii) the probable contribution of boronfrom NaCl 1047298uid inclusions (iii) the sinking of boron by travertineprecipitation (Hobbs and Reardon1999) brucite (Rhoades et al 1970)and smectite and (iv) decreases in boron release from minerals due to
increases in pH (Su and Suarez 2004) Acknowledgements
This research was carried out with the 1047297nancial support from theldquoYoung Researchers Projectrdquo (2001ndash2002) of the Italian Ministry of Universities Scienti1047297c Research and Technology (MURST) We thankGianniCortecci (Institute of Geosciences and EarthResources IGG-CNRPisa Italy) and Giampiero Venturelli (Earth Sciences Department Univer-sity of Parma Italy) for useful suggestions and critical review of themanuscript The authors are grateful to Enrico Dinelli (InterdepartmentResearch Centre for Environmental Sciences Alma Mater StudiorumUniversity of Bologna Ravenna Italy) for granting permission to use theGeochemistrsquos WorkbenchsoftwareTheconstructive reviews of theEditor
J Fein and of two anonymous Referees have signi1047297cantly improved the
manuscript
Appendix A Supplementary data
Supplementary data associated with this article can be found inthe online version at doi101016jchemgeo200808017
References
Abbate E Bortolotti V Passerini P 1980 Olistostromes and olistoliths In Sestini G(Ed) Development of the Northern Apennines geosyncline Sedimentary Geology4 pp 521ndash558
Alietti A 1956 Il minerale a strati misti saponite-talco di monte Chiaro (val di TaroAppen Emil) Rendiconti dellAccademia Nazionale dei Lincei 21 201 ndash207
Alietti A De Angelis G Montanari G Sgarlata F Tirelli G 1976 Il ferro nelle argilleappenniniche Periodico di Mineralogia 45 51ndash64
Anselmi B Mellini M Viti C 2000 Chlorine in the Elba Monti Livornesi and Murloserpentines Evidence for seandashwater interaction European Journal of Mineralogy12137ndash146
Apha-Awwa-Wef 1995 Standard Methods for the Examination of Water and Waste-water 19th Edition American Public Health Association American Water WorksAssociation Water Environment Federation
Barnes I LaMarche Jr VC Himmelberg GR 1967 Geochemical evidence of present-day serpentinization Science 156 830ndash832
Barnes I ONeil JR 1969 The relationship between 1047298uids in some fresh alpine-typeultrama1047297cs and possible modern serpentinization Western United StatesGeological Society of America Bulletin 80 1947ndash1960
Barnes I ONeil JRTrescases JJ1978 Present dayserpentinizationin New CaledoniaOman and Yugoslavia Geochimica et Cosmochimica Acta 42 144ndash145
Barnes I Rapp JB ONeil JR Sheppard RA Gude III AJ 1972 Metamorphicassemblages and the direction of 1047298ow of metamorphic 1047298uids in four instances of serpentinization Contributions to Mineralogy and Petrology 35 263ndash276
Barnes JDSelverstone J Sharp ZD 2006Chlorine isotope chemistry of serpentinitesfrom Elba Italy as an indicator of 1047298uid source and subsequent tectonic historyGeochemistry Geophysics Geosystems 7 doi1010292006GC001296
Beccaluva L Emiliani F Venturelli G Zerbi M 1973 Ca Fe Mg Mn Cr Ni Codistribution in some ultrama1047297c rocks outcropping in the Northern Apennines withsome geological remarks Ateneo Parmense Acta Naturalia 9 69ndash98
Beccaluva L Macciotta G Piccardo GB Zeda O 1984 Petrology of lherzolitic rocksfrom the Northern Apennine ophiolites Lithos 17 299 ndash316
BeccaluvaL Venturelli G ZanzucchiG 1975 Dati geologici e geochimici sui basalti diassociazione o1047297olitica dellAppennino ligure-emiliano Ateneo Parmense ActaNaturalia 11 789ndash802
BethkeCM 2002 The Geochemists Workbench A Userrsquos Guide to RxnAct2Tact ReactandGtplot Release40 HydrogeologyProgram Universityof IllinoisUrbanaIL 236pp
Birsoy R 2002Formation of sepiolitendashpalygorskite and related minerals from solutionClays and Clay Minerals 50 736ndash745
Bonatti E Lawrence JR Morandi N 1984 Serpentinization of oceanic peridotitestemperature dependence of mineralogy and boron content Earth and PlanetaryScience Letters 70 88ndash94
Bonifacio E Zanini E Boero V Franchini-Angela M 1997 Pedogenesis in a soilcatena on serpentinite in north-western Italy Geoderma 75 33ndash51
Boschetti T 2003 Studio geochimico e geochimico-isotopico di acque a composizione
estrema e termali dellrsquoAppennino Settentrionale Ph D Thesis Parma Univ
Brigatti MF Franchini G Lugli C Medici L Poppi L Turci E 2000 Interactionsbetween aqueous chromium solutions and layer silicatesApplied Geochemistry 151307ndash1316
Brigatti MF Lugli C Poppi L Venturelli G 1999 Iron-rich saponite dissolutionreactions and Cr uptake Clay Minerals 34 637ndash645
Brigatti MF Valdregrave G 1996 Mixed-layer chloritesmectite (corrensite) in ophioliticrock alteration products Neues Jahrbuch fuer Mineralogie Monatshefte 1 37 ndash47
Bruni J Canepa M Chiodini G Cioni R Cipolli F Longinelli A Marini L OttonelloG Vetuschi Zuccolini M 2002 Irreversible waterndashrock mass transfer accompany-ing the generation of the neutral MgndashHCO3 and high-pH CandashOH spring waters of the Genova province Italy Applied Geochemistry 17 455ndash474
Bryan CR 2005 Drift-Scale THC Seepage Model MDL-NBS-HS-000001 REV 04 YuccaMountain Project Las Vegas Nevada (US)Caruso LJ Chernosky JV 1979 The stability of lizardite Canadian Mineralogist 17
757ndash769Clark ID Fontes JC Fritz P 1992 Stable isotope disequilibria in travertine from high
pH waters Laboratory investigations and 1047297eld observations from Oman Geochi-mica et Cosmochimica Acta 56 2041ndash2050
Cortesogno L 1980 Il metamor1047297smo giurassico nelle o1047297oliti dellAppenninoSettentrionale due distinti cicli metamor1047297ci in ambiente oceanico O1047297oliti 5 5ndash18
Christ CLHostetler PB Siebert RM1973 Studies in the system MgO-SiO2-CO2-H2OIII American Journal of Science 273 65ndash83
Cortesogno L Lucchetti G 1982 Il metamor1047297smo oceanico nei gabbri o1047297oliticidellAppennino Ligure Aspetti mineralogici e petrogenetici Rendiconti dellaSocietagrave Mineralogica Italiana 38 561ndash579
Crispini L Capponi G 2002 Control of metasomatic alteration on the development of shear zone in ultrama1047297c rocks a case study from the Voltri Massif (Ligurian Alps)Architettura interna e cinematica delle zone di taglio Gruppo Italiano di GeologiaStrutturale mdash Riunione Annuale 2002 Pisa Italy 11ndash12 giugno 2002 XIV ndashXV
CrissRE 1999 Principlesof stableisotope distributionOxford University PressNew York
Davis KJ Dove PM De Yoreo JJ 2000 The role of Mg 2+ as an impurity in calcitegrowth Science 290 1134ndash1137
de Leeuw NH 2002 Moleculardynamics simulations of the growth inhibiting effect of Fe2+ Mg2+ Cd2+ and Sr2+ on calcite crystal growth Journal of Physical Chemistry B106 5241ndash5249
De Nardo MT Segadelli S Vescovi P 2007 Studio pilota per la caratterizzazionegeologicadellesorgentinellazona delM Nero (altaVal Cenoe alta ValNuremdash provincedi Parma e Piacenza) Il Geologo dellEmilia-Romagna 25 Nuova Serie pp 5 ndash21
Di Giulio A Geddo G 1990 Studio petrogra1047297co delle Arenarie di Casanova (alta valTrebbia Appennino settentrionale) Atti Ticinesi di Scienze della Terra 33 243ndash254
Dinelli E Lombini A Simoni A Ferrari C 1997 Heavy metals in serpentinite soils of selected outcrops of Piacenza and Parma provinces (Northern Apennines Italy)Mineralogica et Petrographica Acta 40 241ndash255
Epstein S Mayeda T 1953 Variation of 18O content of waters from natural sourcesGeochimica et Cosmochimica Acta 4 213ndash224
Ernst WG Piccardo GB 1979 Petrogenesis of some Ligurian peridotites I Mineraland bulk-rock chemistry Geochimica et Cosmochimica Acta 43 219ndash237
Feder F Trolard F Klingelhofer G Bourrieacute G 2005 In situ Moumlssbauer spectroscopy
evidence for green rust (fougerite) in a gleysol and its mineralogical transforma-tions with time and depth Geochimica et Cosmochimica Acta 69 4463ndash4483Fritz PClark IDFontes JCWhiticar MJFaberE 1992 Deuterium and
13C evidence
for low temperature production of hydrogen and methane in a highly alkalinegroundwater environment in Oman In Kharaka TK Maest AS (Eds) Proceed-ings of the 7th International Symposium on WaterndashRock Interaction BalkemaRotterdam pp 793ndash796
Giammetti F 1968 Le o1047297oliti di Roccaprebalza Costa della Guardia e Groppo di Gorro(Appennino Parmense) Ateneo Parmense Acta Naturalia 4 107ndash142
Giuseppetti G1953 Roccee minerali dellaformazioneo1047297olitica di Volpedo Rendicontidella Societagrave Mineralogica Italiana 9 85ndash134
Gran G 1952 Determination of equivalence point in the potentiometric titrationAnalyst 77 661ndash671
Gunnarsson I Arnorsson S Jakobsson S 2005 Precipitation of poorly crystalline antigoriteunder hydrothermal conditions Geochimica et Cosmochimica Acta 69 2813ndash2828
Guumlven N 1988 Smetites In Bailey SW (Ed) Reviews in Mineralogy 19 Hydrousphyllosilicates (exclusive of micas) Mineralogical Society of America MichiganBook Crafters Inc 497ndash559
Hansel CM Benner SG Fendorf S 2005 Competing Fe(II)-induced mineralization
pathways of ferrihydrite Environmental Science and Technology 39 7147ndash7153HobbsMYReardonEJ 1999 Effect of pH on boronco-precipitation by calcite further
evidence for non-equilibrium partitioning of trace elements Geochimica etCosmochimica Acta 63 1013ndash1021
Jones BF Galan E 1988 Sepiolite and palygorskite In Bailey SW (Ed) HydrousPhyllosilicates (exclusive of Micas) Reviews in Mineralogy vol 19 MineralogicalSociety of America Michigan Book Crafters Inc 631ndash674
Jove Colon CF 2006 Technical work plan for Thermodynamic database for chemicalmodelling TWP-MGR-PA-000039 REV 01 Yucca Mountain Project Las VegasNevada (US)
Kohls DW Rodda Jr JL 1967 Iowaite a new hydrous magnesium hydroxidendashferricoxychloride from the Precambrian of Iowa American Mineralogist 52 1261 ndash1271
Kyser TK OHanley DS Wicks FJ 1999 The origin of 1047298uids associated withserpentinization processes evidence from stable-isotope compositions CanadianMineralogist 37 223ndash237
Lapham DM 1961 New data on deweylite American Mineralogist 46 168ndash188Lee BD SearsSKGraham RC Amrhein CValiH 2003 Secondarymineral genesis
from chlorite and serpentine in an ultrama1047297c soil toposequence Soil Science
Society of America Journal 671309ndash1317
90 T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
Longinelli A Selmo E 2003 Isotopic composition of precipitation in Italy a 1047297rstoverall map Journal of Hydrology 270 75ndash88
Majzlan J Navrotsky A Schwertmann U 2004 Thermodynamics of iron oxides PartIIIEnthalpiesof formation and stability of ferrihydrite (simFe(OH)3) schwertmannite(simFeO(OH)34(SO4)18) and ɛ-Fe2O3 Geochimica et Cosmochimica Acta 681049ndash1059
Mariner P 2003 In-Drift PrecipitatesSalts Model ANL-EBS-MD-000045 REV 01 YuccaMountain Project Las Vegas Nevada (US)
Marini L Ottonello G 2002 Atlante degli acquiferi della Liguria Volume III Le acquedeicomplessio1047297olitici (bacini ArrestraBranega Cassinelle Cerusa ErroGorzenteLeira Lemme Lerone Orba Piota Polcevera Rumaro Sansobbia Stura Teiro
Varenna Visone) Pacini PisaMellini L Zanazzi PF 1987 Crystal structures of lizardite-l T and lizardite-2H1 fromColi Italy American Mineralogist 72 943ndash948
Mezzadri G 1964 Petrogra1047297a delle arenarie di Ostia Rendiconti Societagrave MineralogicaItaliana 20 192ndash228
Mondesir H Decarreau A 1987 Synthesegrave entre 25 et 200 degC de lizardites Ni ndashMgMesure des coef 1047297cients de partage solide-solution aqueuse pour le couple Ni ndashMgdans des lizardites Bulletin de Mineralogie 110 409ndash426
Neal C Shand P 2002 Spring and surface water quality of the Cyprus ophiolitesHydrology and Earth System Sciences 6 797ndash817
Neal C Stanger G 1983 Hydrogen generation from mantle source rocks in OmanEarth and Planetary Science Letters 66 315ndash320
Neal C Stanger G 1984 Calcium and magnesium hydroxide precipitation fromalkaline groundwaters in Oman and their signi1047297cance to the process of serpentinization Mineralogical Magazine 48 237ndash241
Neal C Stanger G 1985 Past and present serpentinisation of ultrama1047297c rocks anexample from Semail ophiolite nappe of Northern Oman In Drever JI (Ed) TheChemistry of Weathering D Reidel Publishing Company pp 249ndash275
Nesbitt HW Bricker OP 1978 Low temperature alteration processes affecting
et Cosmochimica Acta 41 1835ndash1841OrsquoHanley DS 1996 Serpentinites records of tectonic and petrological history Oxford
University PressOrberger B Friedrich G Woermann E 1990 The distribution of halogens and carbon
in Pge-bearing ultrama1047297cs of the Acoje Ophiolite Block Zambales Philippines Journal of Geochemical Exploration 37 147ndash169
Ottonello G Joron JL Piccardo GB 1984 Rare Earth and 3d transition elementsgeochemistry of peridotitic rocks II Ligurian peridotites and associated basalts Journal of Petrology 25 373ndash393
Palandri JL Kharaka YK 2004 A compilation of rate parameters of water-mineralinteraction kinetics for application to geochemical modeling Open File Report2004-1068 US Geological Survey
Panettiere P Cortecci G Dinelli E Bencini A Guidi M 2000 Chemistry and sulfurisotopic composition of precipitation at Bologna Italy Applied Geochemistry 151455ndash1467
Parkhurst DL Appelo CAJ 1999 Userrsquos guide to PHREEQC (version 2) mdash a computerprogram for speciation batch-reaction one-dimensional transport and inversegeochemical calculation US Department of the Interior US Geological SurveyWashington DC
Pfeifer HR 1977 A model for 1047298uids in metamorphosed ultrama1047297c rocks Observationsat surface and subsurface conditions (high pH spring waters) SchweizerischeMineralogische und Petrographische Mitteilungen 57 361ndash396
PlummerLN1984Geochemical modeling A comparisonof forwardand inverse methodsIn Hitchon B Wallick EI (Eds) Proceedings First CanadianAmerican Conference onHydrogeology-Practical Applications of Ground Water Geochemistry Banff AlbertaCanada mdash Worthington Ohio National Water Well Association pp 149ndash177
Rampone E Hofmann AW Piccardo GB Vannucci R Bottazzi P Ottolini L 1995Petrology mineral and isotope geochemistry of the External Ligurides peridotites(Northern Apennines Italy) Journal of Petrology 36 81ndash105
Rampone E Hofmann AW Piccardo GB Vannucci R Bottazzi P Ottolini L 1996Trace element and isotope geochemistry of depleted peridotites from an N-MORBtype ophiolite (Internal Liguride N Italy) Contributions to Mineralogy andPetrology 123 61ndash76
Rhoades JD Ingvalson RD HatcherJT1970Adsorption of boronby ferromagnesianminerals and magnesium hydroxide Soil Science Society of America Journal 34
938ndash941Saini-EidukatB KuchaH Keppler H1994Hibbingiteγ-Fe2(OH)3Clanewmineralfromthe Duluth Complex Minnesota with implications for the oxidation of Fe-bearingcompounds and the transport of metals American Mineralogist 79 555ndash561
Sanford RF 1981 Mineralogical and chemical effects of hydration reactions andapplications to serpentinization American Mineralogist 66 290ndash297
Scambelluri M Muumlntener O Ottolini L Pettke TT Vannucci R 2004 The fate of BCl and Li in the subducted oceanic mantle and in the antigorite breakdown 1047298uidsEarth and Planetary Science Letters 222 217ndash234
Scambelluri M Rampone E 1999 Mg-metasomatism of oceanic gabbros and itscontrol on Ti-clinohumite formation during eclogitization Contributions toMineralogy and Petrology 135 1ndash17
Scambelluri M Bottazzi P Trommsdorff V Vannucci R Hermann J Gogravemez-Pugnaire MT Logravepez-Sagravenchez Vizcaino V 2001a Incompatible element-rich1047298uidsreleased by antigorite breakdown in deeply subducted mantle Earth and PlanetaryScience Letters 192 457ndash470
Scambelluri M Rampone E Piccardo G 2001b Fluid and element cycling insubducted serpentinite a trace-element study of the ErrondashTobbio high-pressureultrama1047297tes (Western Alps NW Italy) Journal of Petrology 42 55ndash67
Segadelli S2006 Lageologia nelpaesaggio le rupi o1047297olitiche in Val Taro e Val Ceno IlGeologo dellrsquoEmilia-Romagna 22 15ndash29 Nuova Serie
Settembre Blundo D Martin De Vidales JL Alietti A 1992 Corrensite-type clayminerals from Taro Valley (North Italy) Mineralogica et Petrographica Acta 35215ndash222
Sharp ZD Barnes JD 2004 Water-soluble chlorides in massive sea1047298oor serpenti-nites a source of chloride in subduction zones Earth and Planetary Science Letters226 243ndash254
Siever R Woodford N 1979 Dissolution kinetics and the weathering of ma1047297cminerals Geochimica et Cosmochimica Acta 43 717ndash724Singer A 1989 Palygorskite and sepiolite group minerals In Dixon JB Weed SB
(Eds) Minerals in Soil Environments Soil Science Society of America MadisonWisconsin pp 829ndash872
Siffert B Wey R 1962 Synthese d rsquoune sepiolite a temperature ordinarie Comptesrendus hebdomadaires des seacuteances de lAcadeacutemie des Sciences de Paris 2541460ndash1463
Środoń J 1999 Nature of mixed-layer clays and mechanisms of their formation andalteration Annual Reviews of Earth and Planetary Science 27 19ndash53
Stoessell RK 1988 25 degC and 1 atm dissolution experiments of sepiolite and keroliteGeochimica et Cosmochimica Acta 52 365ndash374
Stoessell RK Hay RL 1978 The geochemical origin of Sepiolite and keorilite atAmboseli Kenya Contributions to Mineralogy and Petrology 65 255ndash267
Suarez DL Šimůnek J 1996 Solute transport modeling under variably saturatedwater 1047298ow conditions In Lichtner PC Steefel CI Oelkers EH (Eds) ReactiveTransport in Porous Media Reviews in Mineralogy vol 34 Mineralogical Society of America Michigan Book Crafters Inc 229ndash268
Sumner DY Grotzinger JP 1996 Were kinetics of Archean calcium carbonate
precipitation related to oxygen concentration Geology 24 119ndash122Su C Suarez DL 2004 Boron release from weathering of illites serpentine shales
and illiticpalygorskitic soils Soil Science Society of America Journal 68 96 ndash105Toscani L Venturelli G Boschetti T 2001 Sul1047297de-free and sul1047297de-bearing waters in
the Northern Apennines Italy general features and waterndashrock interaction AquaticGeochemistry 7 195ndash216
Treves BE Harper GD 1994 Exposure of serpentinites on the ocean 1047298oor sequenceof faulting and hydrofracturing in the Northern Apennine ophicalcites O1047297oliti 19b435ndash466
Varadachari C Kudrat M Gosh K 1994 Evaluation of standard free energies of formation of clay minerals by an improved regression method Clays and ClayMinerals 42 298ndash307
Veniale F 1966 Sepiolite in sedimenti dellrsquoAppennino Vogherese (Pavia) Periodico diMineralogia 35 343ndash386
Venturelli G 2003 Acque minerali ed ambiente mdash Fondamenti di geochimica deiprocessi di bassa temperatura Pitagora Editrice Bologna
Venturelli G Contini S Bonazzi A Mangia A 1997 Weathering of ultrama1047297c rocksand element mobility at Mt Prinzera Northern Apennines Italy MineralogicalMagazine 61 765ndash778
Venturelli G Frey M 1977 Anchizone metamorphism in sedimentary sequences of the Northern Apennines Rendiconti Societagrave Mineralogica Italiana 33 109ndash123
Vieillard P 2002 A new method for the prediction of Gibbs free energies of formationofphyllosilicates (10 Aring and14 Aring)basedon theelectronegativityscale Claysand ClayMinerals 50 352ndash363
Vissers RLM Strating EHH 2001 Structures and microstructures in a thrust-relatedgreenschist facies tectonic melange Voltri Group (NW Italy) O1047297oliti 26 33ndash46
Viti C Mellini M 1997 Contrasting chemical compositions in associated lizardite andchrysotile in veins from Elba Italy European Journal of Mineralogy 9 585 ndash596
Viti C Mellini M 1998 Mesh textures and bastites in the Elba retrogradeserpentinites European Journal of Mineralogy 10 1341ndash1359
Wenner DB Taylor Jr HP 1974 Deuteriumatomic hydrogen and oxygen-18oxygen-16 studies of serpentinization of ultrama1047297c rocks Geochimica et CosmochimicaActa 38 1255ndash1286
Williams LA Parks GA Crerar DA 1985 Silica diagenesis I Solubility controls Journal of S edimentary Petrology 55 301ndash311
Wilson J Savage D Cuadros J Shibata M RagnarsdottirKV 2006 The effect of ironon montmorillonite stability (I) Background and thermodynamic considerations
Geochimica et Cosmochimica Acta 70 306ndash322WollastR MackenzieFTBricker OP1968 Experimental precipitation and genesis of sepiolite at earth-surface conditions American Mineralogist 531 1645ndash1662
Wrucke CT1996 Serpentine and carbonated-hosted asbestos deposits InEdwardAdu Bray (Eds) Preliminary compilation of descriptive geoenvironmental mineraldeposit models 95ndash831 US Geological Survey Open-File Report pp 39 ndash46
Xu T Apps JA Pruess K 2001 Analysis of mineral trapping for CO2 disposal in deepaquifers Lawrence Berkeley National Laboratory University of California
ZhangYDaweRA2000 In1047298uence ofMg2+ on thekineticsof calcite precipitation andcalcite crystal morphology Chemical Geology 163 129ndash138
Zhu C Anderson G 2002 Environmental applications of geochemical modelingCambridge University Press
91T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
Group are unexpected features and probably re1047298ect waterndashrockinteractions over a longer time period than in the TarondashCeno ValleysThe boron content in the Northern Apennine alkaline springs may becontrolled by several factors such as (i) the different histories of VoltriGroup and EL ultrama1047297tes (ii) the probable contribution of boronfrom NaCl 1047298uid inclusions (iii) the sinking of boron by travertineprecipitation (Hobbs and Reardon1999) brucite (Rhoades et al 1970)and smectite and (iv) decreases in boron release from minerals due to
increases in pH (Su and Suarez 2004) Acknowledgements
This research was carried out with the 1047297nancial support from theldquoYoung Researchers Projectrdquo (2001ndash2002) of the Italian Ministry of Universities Scienti1047297c Research and Technology (MURST) We thankGianniCortecci (Institute of Geosciences and EarthResources IGG-CNRPisa Italy) and Giampiero Venturelli (Earth Sciences Department Univer-sity of Parma Italy) for useful suggestions and critical review of themanuscript The authors are grateful to Enrico Dinelli (InterdepartmentResearch Centre for Environmental Sciences Alma Mater StudiorumUniversity of Bologna Ravenna Italy) for granting permission to use theGeochemistrsquos WorkbenchsoftwareTheconstructive reviews of theEditor
J Fein and of two anonymous Referees have signi1047297cantly improved the
manuscript
Appendix A Supplementary data
Supplementary data associated with this article can be found inthe online version at doi101016jchemgeo200808017
References
Abbate E Bortolotti V Passerini P 1980 Olistostromes and olistoliths In Sestini G(Ed) Development of the Northern Apennines geosyncline Sedimentary Geology4 pp 521ndash558
Alietti A 1956 Il minerale a strati misti saponite-talco di monte Chiaro (val di TaroAppen Emil) Rendiconti dellAccademia Nazionale dei Lincei 21 201 ndash207
Alietti A De Angelis G Montanari G Sgarlata F Tirelli G 1976 Il ferro nelle argilleappenniniche Periodico di Mineralogia 45 51ndash64
Anselmi B Mellini M Viti C 2000 Chlorine in the Elba Monti Livornesi and Murloserpentines Evidence for seandashwater interaction European Journal of Mineralogy12137ndash146
Apha-Awwa-Wef 1995 Standard Methods for the Examination of Water and Waste-water 19th Edition American Public Health Association American Water WorksAssociation Water Environment Federation
Barnes I LaMarche Jr VC Himmelberg GR 1967 Geochemical evidence of present-day serpentinization Science 156 830ndash832
Barnes I ONeil JR 1969 The relationship between 1047298uids in some fresh alpine-typeultrama1047297cs and possible modern serpentinization Western United StatesGeological Society of America Bulletin 80 1947ndash1960
Barnes I ONeil JRTrescases JJ1978 Present dayserpentinizationin New CaledoniaOman and Yugoslavia Geochimica et Cosmochimica Acta 42 144ndash145
Barnes I Rapp JB ONeil JR Sheppard RA Gude III AJ 1972 Metamorphicassemblages and the direction of 1047298ow of metamorphic 1047298uids in four instances of serpentinization Contributions to Mineralogy and Petrology 35 263ndash276
Barnes JDSelverstone J Sharp ZD 2006Chlorine isotope chemistry of serpentinitesfrom Elba Italy as an indicator of 1047298uid source and subsequent tectonic historyGeochemistry Geophysics Geosystems 7 doi1010292006GC001296
Beccaluva L Emiliani F Venturelli G Zerbi M 1973 Ca Fe Mg Mn Cr Ni Codistribution in some ultrama1047297c rocks outcropping in the Northern Apennines withsome geological remarks Ateneo Parmense Acta Naturalia 9 69ndash98
Beccaluva L Macciotta G Piccardo GB Zeda O 1984 Petrology of lherzolitic rocksfrom the Northern Apennine ophiolites Lithos 17 299 ndash316
BeccaluvaL Venturelli G ZanzucchiG 1975 Dati geologici e geochimici sui basalti diassociazione o1047297olitica dellAppennino ligure-emiliano Ateneo Parmense ActaNaturalia 11 789ndash802
BethkeCM 2002 The Geochemists Workbench A Userrsquos Guide to RxnAct2Tact ReactandGtplot Release40 HydrogeologyProgram Universityof IllinoisUrbanaIL 236pp
Birsoy R 2002Formation of sepiolitendashpalygorskite and related minerals from solutionClays and Clay Minerals 50 736ndash745
Bonatti E Lawrence JR Morandi N 1984 Serpentinization of oceanic peridotitestemperature dependence of mineralogy and boron content Earth and PlanetaryScience Letters 70 88ndash94
Bonifacio E Zanini E Boero V Franchini-Angela M 1997 Pedogenesis in a soilcatena on serpentinite in north-western Italy Geoderma 75 33ndash51
Boschetti T 2003 Studio geochimico e geochimico-isotopico di acque a composizione
estrema e termali dellrsquoAppennino Settentrionale Ph D Thesis Parma Univ
Brigatti MF Franchini G Lugli C Medici L Poppi L Turci E 2000 Interactionsbetween aqueous chromium solutions and layer silicatesApplied Geochemistry 151307ndash1316
Brigatti MF Lugli C Poppi L Venturelli G 1999 Iron-rich saponite dissolutionreactions and Cr uptake Clay Minerals 34 637ndash645
Brigatti MF Valdregrave G 1996 Mixed-layer chloritesmectite (corrensite) in ophioliticrock alteration products Neues Jahrbuch fuer Mineralogie Monatshefte 1 37 ndash47
Bruni J Canepa M Chiodini G Cioni R Cipolli F Longinelli A Marini L OttonelloG Vetuschi Zuccolini M 2002 Irreversible waterndashrock mass transfer accompany-ing the generation of the neutral MgndashHCO3 and high-pH CandashOH spring waters of the Genova province Italy Applied Geochemistry 17 455ndash474
Bryan CR 2005 Drift-Scale THC Seepage Model MDL-NBS-HS-000001 REV 04 YuccaMountain Project Las Vegas Nevada (US)Caruso LJ Chernosky JV 1979 The stability of lizardite Canadian Mineralogist 17
757ndash769Clark ID Fontes JC Fritz P 1992 Stable isotope disequilibria in travertine from high
pH waters Laboratory investigations and 1047297eld observations from Oman Geochi-mica et Cosmochimica Acta 56 2041ndash2050
Cortesogno L 1980 Il metamor1047297smo giurassico nelle o1047297oliti dellAppenninoSettentrionale due distinti cicli metamor1047297ci in ambiente oceanico O1047297oliti 5 5ndash18
Christ CLHostetler PB Siebert RM1973 Studies in the system MgO-SiO2-CO2-H2OIII American Journal of Science 273 65ndash83
Cortesogno L Lucchetti G 1982 Il metamor1047297smo oceanico nei gabbri o1047297oliticidellAppennino Ligure Aspetti mineralogici e petrogenetici Rendiconti dellaSocietagrave Mineralogica Italiana 38 561ndash579
Crispini L Capponi G 2002 Control of metasomatic alteration on the development of shear zone in ultrama1047297c rocks a case study from the Voltri Massif (Ligurian Alps)Architettura interna e cinematica delle zone di taglio Gruppo Italiano di GeologiaStrutturale mdash Riunione Annuale 2002 Pisa Italy 11ndash12 giugno 2002 XIV ndashXV
CrissRE 1999 Principlesof stableisotope distributionOxford University PressNew York
Davis KJ Dove PM De Yoreo JJ 2000 The role of Mg 2+ as an impurity in calcitegrowth Science 290 1134ndash1137
de Leeuw NH 2002 Moleculardynamics simulations of the growth inhibiting effect of Fe2+ Mg2+ Cd2+ and Sr2+ on calcite crystal growth Journal of Physical Chemistry B106 5241ndash5249
De Nardo MT Segadelli S Vescovi P 2007 Studio pilota per la caratterizzazionegeologicadellesorgentinellazona delM Nero (altaVal Cenoe alta ValNuremdash provincedi Parma e Piacenza) Il Geologo dellEmilia-Romagna 25 Nuova Serie pp 5 ndash21
Di Giulio A Geddo G 1990 Studio petrogra1047297co delle Arenarie di Casanova (alta valTrebbia Appennino settentrionale) Atti Ticinesi di Scienze della Terra 33 243ndash254
Dinelli E Lombini A Simoni A Ferrari C 1997 Heavy metals in serpentinite soils of selected outcrops of Piacenza and Parma provinces (Northern Apennines Italy)Mineralogica et Petrographica Acta 40 241ndash255
Epstein S Mayeda T 1953 Variation of 18O content of waters from natural sourcesGeochimica et Cosmochimica Acta 4 213ndash224
Ernst WG Piccardo GB 1979 Petrogenesis of some Ligurian peridotites I Mineraland bulk-rock chemistry Geochimica et Cosmochimica Acta 43 219ndash237
Feder F Trolard F Klingelhofer G Bourrieacute G 2005 In situ Moumlssbauer spectroscopy
evidence for green rust (fougerite) in a gleysol and its mineralogical transforma-tions with time and depth Geochimica et Cosmochimica Acta 69 4463ndash4483Fritz PClark IDFontes JCWhiticar MJFaberE 1992 Deuterium and
13C evidence
for low temperature production of hydrogen and methane in a highly alkalinegroundwater environment in Oman In Kharaka TK Maest AS (Eds) Proceed-ings of the 7th International Symposium on WaterndashRock Interaction BalkemaRotterdam pp 793ndash796
Giammetti F 1968 Le o1047297oliti di Roccaprebalza Costa della Guardia e Groppo di Gorro(Appennino Parmense) Ateneo Parmense Acta Naturalia 4 107ndash142
Giuseppetti G1953 Roccee minerali dellaformazioneo1047297olitica di Volpedo Rendicontidella Societagrave Mineralogica Italiana 9 85ndash134
Gran G 1952 Determination of equivalence point in the potentiometric titrationAnalyst 77 661ndash671
Gunnarsson I Arnorsson S Jakobsson S 2005 Precipitation of poorly crystalline antigoriteunder hydrothermal conditions Geochimica et Cosmochimica Acta 69 2813ndash2828
Guumlven N 1988 Smetites In Bailey SW (Ed) Reviews in Mineralogy 19 Hydrousphyllosilicates (exclusive of micas) Mineralogical Society of America MichiganBook Crafters Inc 497ndash559
Hansel CM Benner SG Fendorf S 2005 Competing Fe(II)-induced mineralization
pathways of ferrihydrite Environmental Science and Technology 39 7147ndash7153HobbsMYReardonEJ 1999 Effect of pH on boronco-precipitation by calcite further
evidence for non-equilibrium partitioning of trace elements Geochimica etCosmochimica Acta 63 1013ndash1021
Jones BF Galan E 1988 Sepiolite and palygorskite In Bailey SW (Ed) HydrousPhyllosilicates (exclusive of Micas) Reviews in Mineralogy vol 19 MineralogicalSociety of America Michigan Book Crafters Inc 631ndash674
Jove Colon CF 2006 Technical work plan for Thermodynamic database for chemicalmodelling TWP-MGR-PA-000039 REV 01 Yucca Mountain Project Las VegasNevada (US)
Kohls DW Rodda Jr JL 1967 Iowaite a new hydrous magnesium hydroxidendashferricoxychloride from the Precambrian of Iowa American Mineralogist 52 1261 ndash1271
Kyser TK OHanley DS Wicks FJ 1999 The origin of 1047298uids associated withserpentinization processes evidence from stable-isotope compositions CanadianMineralogist 37 223ndash237
Lapham DM 1961 New data on deweylite American Mineralogist 46 168ndash188Lee BD SearsSKGraham RC Amrhein CValiH 2003 Secondarymineral genesis
from chlorite and serpentine in an ultrama1047297c soil toposequence Soil Science
Society of America Journal 671309ndash1317
90 T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
Longinelli A Selmo E 2003 Isotopic composition of precipitation in Italy a 1047297rstoverall map Journal of Hydrology 270 75ndash88
Majzlan J Navrotsky A Schwertmann U 2004 Thermodynamics of iron oxides PartIIIEnthalpiesof formation and stability of ferrihydrite (simFe(OH)3) schwertmannite(simFeO(OH)34(SO4)18) and ɛ-Fe2O3 Geochimica et Cosmochimica Acta 681049ndash1059
Mariner P 2003 In-Drift PrecipitatesSalts Model ANL-EBS-MD-000045 REV 01 YuccaMountain Project Las Vegas Nevada (US)
Marini L Ottonello G 2002 Atlante degli acquiferi della Liguria Volume III Le acquedeicomplessio1047297olitici (bacini ArrestraBranega Cassinelle Cerusa ErroGorzenteLeira Lemme Lerone Orba Piota Polcevera Rumaro Sansobbia Stura Teiro
Varenna Visone) Pacini PisaMellini L Zanazzi PF 1987 Crystal structures of lizardite-l T and lizardite-2H1 fromColi Italy American Mineralogist 72 943ndash948
Mezzadri G 1964 Petrogra1047297a delle arenarie di Ostia Rendiconti Societagrave MineralogicaItaliana 20 192ndash228
Mondesir H Decarreau A 1987 Synthesegrave entre 25 et 200 degC de lizardites Ni ndashMgMesure des coef 1047297cients de partage solide-solution aqueuse pour le couple Ni ndashMgdans des lizardites Bulletin de Mineralogie 110 409ndash426
Neal C Shand P 2002 Spring and surface water quality of the Cyprus ophiolitesHydrology and Earth System Sciences 6 797ndash817
Neal C Stanger G 1983 Hydrogen generation from mantle source rocks in OmanEarth and Planetary Science Letters 66 315ndash320
Neal C Stanger G 1984 Calcium and magnesium hydroxide precipitation fromalkaline groundwaters in Oman and their signi1047297cance to the process of serpentinization Mineralogical Magazine 48 237ndash241
Neal C Stanger G 1985 Past and present serpentinisation of ultrama1047297c rocks anexample from Semail ophiolite nappe of Northern Oman In Drever JI (Ed) TheChemistry of Weathering D Reidel Publishing Company pp 249ndash275
Nesbitt HW Bricker OP 1978 Low temperature alteration processes affecting
et Cosmochimica Acta 41 1835ndash1841OrsquoHanley DS 1996 Serpentinites records of tectonic and petrological history Oxford
University PressOrberger B Friedrich G Woermann E 1990 The distribution of halogens and carbon
in Pge-bearing ultrama1047297cs of the Acoje Ophiolite Block Zambales Philippines Journal of Geochemical Exploration 37 147ndash169
Ottonello G Joron JL Piccardo GB 1984 Rare Earth and 3d transition elementsgeochemistry of peridotitic rocks II Ligurian peridotites and associated basalts Journal of Petrology 25 373ndash393
Palandri JL Kharaka YK 2004 A compilation of rate parameters of water-mineralinteraction kinetics for application to geochemical modeling Open File Report2004-1068 US Geological Survey
Panettiere P Cortecci G Dinelli E Bencini A Guidi M 2000 Chemistry and sulfurisotopic composition of precipitation at Bologna Italy Applied Geochemistry 151455ndash1467
Parkhurst DL Appelo CAJ 1999 Userrsquos guide to PHREEQC (version 2) mdash a computerprogram for speciation batch-reaction one-dimensional transport and inversegeochemical calculation US Department of the Interior US Geological SurveyWashington DC
Pfeifer HR 1977 A model for 1047298uids in metamorphosed ultrama1047297c rocks Observationsat surface and subsurface conditions (high pH spring waters) SchweizerischeMineralogische und Petrographische Mitteilungen 57 361ndash396
PlummerLN1984Geochemical modeling A comparisonof forwardand inverse methodsIn Hitchon B Wallick EI (Eds) Proceedings First CanadianAmerican Conference onHydrogeology-Practical Applications of Ground Water Geochemistry Banff AlbertaCanada mdash Worthington Ohio National Water Well Association pp 149ndash177
Rampone E Hofmann AW Piccardo GB Vannucci R Bottazzi P Ottolini L 1995Petrology mineral and isotope geochemistry of the External Ligurides peridotites(Northern Apennines Italy) Journal of Petrology 36 81ndash105
Rampone E Hofmann AW Piccardo GB Vannucci R Bottazzi P Ottolini L 1996Trace element and isotope geochemistry of depleted peridotites from an N-MORBtype ophiolite (Internal Liguride N Italy) Contributions to Mineralogy andPetrology 123 61ndash76
Rhoades JD Ingvalson RD HatcherJT1970Adsorption of boronby ferromagnesianminerals and magnesium hydroxide Soil Science Society of America Journal 34
938ndash941Saini-EidukatB KuchaH Keppler H1994Hibbingiteγ-Fe2(OH)3Clanewmineralfromthe Duluth Complex Minnesota with implications for the oxidation of Fe-bearingcompounds and the transport of metals American Mineralogist 79 555ndash561
Sanford RF 1981 Mineralogical and chemical effects of hydration reactions andapplications to serpentinization American Mineralogist 66 290ndash297
Scambelluri M Muumlntener O Ottolini L Pettke TT Vannucci R 2004 The fate of BCl and Li in the subducted oceanic mantle and in the antigorite breakdown 1047298uidsEarth and Planetary Science Letters 222 217ndash234
Scambelluri M Rampone E 1999 Mg-metasomatism of oceanic gabbros and itscontrol on Ti-clinohumite formation during eclogitization Contributions toMineralogy and Petrology 135 1ndash17
Scambelluri M Bottazzi P Trommsdorff V Vannucci R Hermann J Gogravemez-Pugnaire MT Logravepez-Sagravenchez Vizcaino V 2001a Incompatible element-rich1047298uidsreleased by antigorite breakdown in deeply subducted mantle Earth and PlanetaryScience Letters 192 457ndash470
Scambelluri M Rampone E Piccardo G 2001b Fluid and element cycling insubducted serpentinite a trace-element study of the ErrondashTobbio high-pressureultrama1047297tes (Western Alps NW Italy) Journal of Petrology 42 55ndash67
Segadelli S2006 Lageologia nelpaesaggio le rupi o1047297olitiche in Val Taro e Val Ceno IlGeologo dellrsquoEmilia-Romagna 22 15ndash29 Nuova Serie
Settembre Blundo D Martin De Vidales JL Alietti A 1992 Corrensite-type clayminerals from Taro Valley (North Italy) Mineralogica et Petrographica Acta 35215ndash222
Sharp ZD Barnes JD 2004 Water-soluble chlorides in massive sea1047298oor serpenti-nites a source of chloride in subduction zones Earth and Planetary Science Letters226 243ndash254
Siever R Woodford N 1979 Dissolution kinetics and the weathering of ma1047297cminerals Geochimica et Cosmochimica Acta 43 717ndash724Singer A 1989 Palygorskite and sepiolite group minerals In Dixon JB Weed SB
(Eds) Minerals in Soil Environments Soil Science Society of America MadisonWisconsin pp 829ndash872
Siffert B Wey R 1962 Synthese d rsquoune sepiolite a temperature ordinarie Comptesrendus hebdomadaires des seacuteances de lAcadeacutemie des Sciences de Paris 2541460ndash1463
Środoń J 1999 Nature of mixed-layer clays and mechanisms of their formation andalteration Annual Reviews of Earth and Planetary Science 27 19ndash53
Stoessell RK 1988 25 degC and 1 atm dissolution experiments of sepiolite and keroliteGeochimica et Cosmochimica Acta 52 365ndash374
Stoessell RK Hay RL 1978 The geochemical origin of Sepiolite and keorilite atAmboseli Kenya Contributions to Mineralogy and Petrology 65 255ndash267
Suarez DL Šimůnek J 1996 Solute transport modeling under variably saturatedwater 1047298ow conditions In Lichtner PC Steefel CI Oelkers EH (Eds) ReactiveTransport in Porous Media Reviews in Mineralogy vol 34 Mineralogical Society of America Michigan Book Crafters Inc 229ndash268
Sumner DY Grotzinger JP 1996 Were kinetics of Archean calcium carbonate
precipitation related to oxygen concentration Geology 24 119ndash122Su C Suarez DL 2004 Boron release from weathering of illites serpentine shales
and illiticpalygorskitic soils Soil Science Society of America Journal 68 96 ndash105Toscani L Venturelli G Boschetti T 2001 Sul1047297de-free and sul1047297de-bearing waters in
the Northern Apennines Italy general features and waterndashrock interaction AquaticGeochemistry 7 195ndash216
Treves BE Harper GD 1994 Exposure of serpentinites on the ocean 1047298oor sequenceof faulting and hydrofracturing in the Northern Apennine ophicalcites O1047297oliti 19b435ndash466
Varadachari C Kudrat M Gosh K 1994 Evaluation of standard free energies of formation of clay minerals by an improved regression method Clays and ClayMinerals 42 298ndash307
Veniale F 1966 Sepiolite in sedimenti dellrsquoAppennino Vogherese (Pavia) Periodico diMineralogia 35 343ndash386
Venturelli G 2003 Acque minerali ed ambiente mdash Fondamenti di geochimica deiprocessi di bassa temperatura Pitagora Editrice Bologna
Venturelli G Contini S Bonazzi A Mangia A 1997 Weathering of ultrama1047297c rocksand element mobility at Mt Prinzera Northern Apennines Italy MineralogicalMagazine 61 765ndash778
Venturelli G Frey M 1977 Anchizone metamorphism in sedimentary sequences of the Northern Apennines Rendiconti Societagrave Mineralogica Italiana 33 109ndash123
Vieillard P 2002 A new method for the prediction of Gibbs free energies of formationofphyllosilicates (10 Aring and14 Aring)basedon theelectronegativityscale Claysand ClayMinerals 50 352ndash363
Vissers RLM Strating EHH 2001 Structures and microstructures in a thrust-relatedgreenschist facies tectonic melange Voltri Group (NW Italy) O1047297oliti 26 33ndash46
Viti C Mellini M 1997 Contrasting chemical compositions in associated lizardite andchrysotile in veins from Elba Italy European Journal of Mineralogy 9 585 ndash596
Viti C Mellini M 1998 Mesh textures and bastites in the Elba retrogradeserpentinites European Journal of Mineralogy 10 1341ndash1359
Wenner DB Taylor Jr HP 1974 Deuteriumatomic hydrogen and oxygen-18oxygen-16 studies of serpentinization of ultrama1047297c rocks Geochimica et CosmochimicaActa 38 1255ndash1286
Williams LA Parks GA Crerar DA 1985 Silica diagenesis I Solubility controls Journal of S edimentary Petrology 55 301ndash311
Wilson J Savage D Cuadros J Shibata M RagnarsdottirKV 2006 The effect of ironon montmorillonite stability (I) Background and thermodynamic considerations
Geochimica et Cosmochimica Acta 70 306ndash322WollastR MackenzieFTBricker OP1968 Experimental precipitation and genesis of sepiolite at earth-surface conditions American Mineralogist 531 1645ndash1662
Wrucke CT1996 Serpentine and carbonated-hosted asbestos deposits InEdwardAdu Bray (Eds) Preliminary compilation of descriptive geoenvironmental mineraldeposit models 95ndash831 US Geological Survey Open-File Report pp 39 ndash46
Xu T Apps JA Pruess K 2001 Analysis of mineral trapping for CO2 disposal in deepaquifers Lawrence Berkeley National Laboratory University of California
ZhangYDaweRA2000 In1047298uence ofMg2+ on thekineticsof calcite precipitation andcalcite crystal morphology Chemical Geology 163 129ndash138
Zhu C Anderson G 2002 Environmental applications of geochemical modelingCambridge University Press
91T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
Longinelli A Selmo E 2003 Isotopic composition of precipitation in Italy a 1047297rstoverall map Journal of Hydrology 270 75ndash88
Majzlan J Navrotsky A Schwertmann U 2004 Thermodynamics of iron oxides PartIIIEnthalpiesof formation and stability of ferrihydrite (simFe(OH)3) schwertmannite(simFeO(OH)34(SO4)18) and ɛ-Fe2O3 Geochimica et Cosmochimica Acta 681049ndash1059
Mariner P 2003 In-Drift PrecipitatesSalts Model ANL-EBS-MD-000045 REV 01 YuccaMountain Project Las Vegas Nevada (US)
Marini L Ottonello G 2002 Atlante degli acquiferi della Liguria Volume III Le acquedeicomplessio1047297olitici (bacini ArrestraBranega Cassinelle Cerusa ErroGorzenteLeira Lemme Lerone Orba Piota Polcevera Rumaro Sansobbia Stura Teiro
Varenna Visone) Pacini PisaMellini L Zanazzi PF 1987 Crystal structures of lizardite-l T and lizardite-2H1 fromColi Italy American Mineralogist 72 943ndash948
Mezzadri G 1964 Petrogra1047297a delle arenarie di Ostia Rendiconti Societagrave MineralogicaItaliana 20 192ndash228
Mondesir H Decarreau A 1987 Synthesegrave entre 25 et 200 degC de lizardites Ni ndashMgMesure des coef 1047297cients de partage solide-solution aqueuse pour le couple Ni ndashMgdans des lizardites Bulletin de Mineralogie 110 409ndash426
Neal C Shand P 2002 Spring and surface water quality of the Cyprus ophiolitesHydrology and Earth System Sciences 6 797ndash817
Neal C Stanger G 1983 Hydrogen generation from mantle source rocks in OmanEarth and Planetary Science Letters 66 315ndash320
Neal C Stanger G 1984 Calcium and magnesium hydroxide precipitation fromalkaline groundwaters in Oman and their signi1047297cance to the process of serpentinization Mineralogical Magazine 48 237ndash241
Neal C Stanger G 1985 Past and present serpentinisation of ultrama1047297c rocks anexample from Semail ophiolite nappe of Northern Oman In Drever JI (Ed) TheChemistry of Weathering D Reidel Publishing Company pp 249ndash275
Nesbitt HW Bricker OP 1978 Low temperature alteration processes affecting
et Cosmochimica Acta 41 1835ndash1841OrsquoHanley DS 1996 Serpentinites records of tectonic and petrological history Oxford
University PressOrberger B Friedrich G Woermann E 1990 The distribution of halogens and carbon
in Pge-bearing ultrama1047297cs of the Acoje Ophiolite Block Zambales Philippines Journal of Geochemical Exploration 37 147ndash169
Ottonello G Joron JL Piccardo GB 1984 Rare Earth and 3d transition elementsgeochemistry of peridotitic rocks II Ligurian peridotites and associated basalts Journal of Petrology 25 373ndash393
Palandri JL Kharaka YK 2004 A compilation of rate parameters of water-mineralinteraction kinetics for application to geochemical modeling Open File Report2004-1068 US Geological Survey
Panettiere P Cortecci G Dinelli E Bencini A Guidi M 2000 Chemistry and sulfurisotopic composition of precipitation at Bologna Italy Applied Geochemistry 151455ndash1467
Parkhurst DL Appelo CAJ 1999 Userrsquos guide to PHREEQC (version 2) mdash a computerprogram for speciation batch-reaction one-dimensional transport and inversegeochemical calculation US Department of the Interior US Geological SurveyWashington DC
Pfeifer HR 1977 A model for 1047298uids in metamorphosed ultrama1047297c rocks Observationsat surface and subsurface conditions (high pH spring waters) SchweizerischeMineralogische und Petrographische Mitteilungen 57 361ndash396
PlummerLN1984Geochemical modeling A comparisonof forwardand inverse methodsIn Hitchon B Wallick EI (Eds) Proceedings First CanadianAmerican Conference onHydrogeology-Practical Applications of Ground Water Geochemistry Banff AlbertaCanada mdash Worthington Ohio National Water Well Association pp 149ndash177
Rampone E Hofmann AW Piccardo GB Vannucci R Bottazzi P Ottolini L 1995Petrology mineral and isotope geochemistry of the External Ligurides peridotites(Northern Apennines Italy) Journal of Petrology 36 81ndash105
Rampone E Hofmann AW Piccardo GB Vannucci R Bottazzi P Ottolini L 1996Trace element and isotope geochemistry of depleted peridotites from an N-MORBtype ophiolite (Internal Liguride N Italy) Contributions to Mineralogy andPetrology 123 61ndash76
Rhoades JD Ingvalson RD HatcherJT1970Adsorption of boronby ferromagnesianminerals and magnesium hydroxide Soil Science Society of America Journal 34
938ndash941Saini-EidukatB KuchaH Keppler H1994Hibbingiteγ-Fe2(OH)3Clanewmineralfromthe Duluth Complex Minnesota with implications for the oxidation of Fe-bearingcompounds and the transport of metals American Mineralogist 79 555ndash561
Sanford RF 1981 Mineralogical and chemical effects of hydration reactions andapplications to serpentinization American Mineralogist 66 290ndash297
Scambelluri M Muumlntener O Ottolini L Pettke TT Vannucci R 2004 The fate of BCl and Li in the subducted oceanic mantle and in the antigorite breakdown 1047298uidsEarth and Planetary Science Letters 222 217ndash234
Scambelluri M Rampone E 1999 Mg-metasomatism of oceanic gabbros and itscontrol on Ti-clinohumite formation during eclogitization Contributions toMineralogy and Petrology 135 1ndash17
Scambelluri M Bottazzi P Trommsdorff V Vannucci R Hermann J Gogravemez-Pugnaire MT Logravepez-Sagravenchez Vizcaino V 2001a Incompatible element-rich1047298uidsreleased by antigorite breakdown in deeply subducted mantle Earth and PlanetaryScience Letters 192 457ndash470
Scambelluri M Rampone E Piccardo G 2001b Fluid and element cycling insubducted serpentinite a trace-element study of the ErrondashTobbio high-pressureultrama1047297tes (Western Alps NW Italy) Journal of Petrology 42 55ndash67
Segadelli S2006 Lageologia nelpaesaggio le rupi o1047297olitiche in Val Taro e Val Ceno IlGeologo dellrsquoEmilia-Romagna 22 15ndash29 Nuova Serie
Settembre Blundo D Martin De Vidales JL Alietti A 1992 Corrensite-type clayminerals from Taro Valley (North Italy) Mineralogica et Petrographica Acta 35215ndash222
Sharp ZD Barnes JD 2004 Water-soluble chlorides in massive sea1047298oor serpenti-nites a source of chloride in subduction zones Earth and Planetary Science Letters226 243ndash254
Siever R Woodford N 1979 Dissolution kinetics and the weathering of ma1047297cminerals Geochimica et Cosmochimica Acta 43 717ndash724Singer A 1989 Palygorskite and sepiolite group minerals In Dixon JB Weed SB
(Eds) Minerals in Soil Environments Soil Science Society of America MadisonWisconsin pp 829ndash872
Siffert B Wey R 1962 Synthese d rsquoune sepiolite a temperature ordinarie Comptesrendus hebdomadaires des seacuteances de lAcadeacutemie des Sciences de Paris 2541460ndash1463
Środoń J 1999 Nature of mixed-layer clays and mechanisms of their formation andalteration Annual Reviews of Earth and Planetary Science 27 19ndash53
Stoessell RK 1988 25 degC and 1 atm dissolution experiments of sepiolite and keroliteGeochimica et Cosmochimica Acta 52 365ndash374
Stoessell RK Hay RL 1978 The geochemical origin of Sepiolite and keorilite atAmboseli Kenya Contributions to Mineralogy and Petrology 65 255ndash267
Suarez DL Šimůnek J 1996 Solute transport modeling under variably saturatedwater 1047298ow conditions In Lichtner PC Steefel CI Oelkers EH (Eds) ReactiveTransport in Porous Media Reviews in Mineralogy vol 34 Mineralogical Society of America Michigan Book Crafters Inc 229ndash268
Sumner DY Grotzinger JP 1996 Were kinetics of Archean calcium carbonate
precipitation related to oxygen concentration Geology 24 119ndash122Su C Suarez DL 2004 Boron release from weathering of illites serpentine shales
and illiticpalygorskitic soils Soil Science Society of America Journal 68 96 ndash105Toscani L Venturelli G Boschetti T 2001 Sul1047297de-free and sul1047297de-bearing waters in
the Northern Apennines Italy general features and waterndashrock interaction AquaticGeochemistry 7 195ndash216
Treves BE Harper GD 1994 Exposure of serpentinites on the ocean 1047298oor sequenceof faulting and hydrofracturing in the Northern Apennine ophicalcites O1047297oliti 19b435ndash466
Varadachari C Kudrat M Gosh K 1994 Evaluation of standard free energies of formation of clay minerals by an improved regression method Clays and ClayMinerals 42 298ndash307
Veniale F 1966 Sepiolite in sedimenti dellrsquoAppennino Vogherese (Pavia) Periodico diMineralogia 35 343ndash386
Venturelli G 2003 Acque minerali ed ambiente mdash Fondamenti di geochimica deiprocessi di bassa temperatura Pitagora Editrice Bologna
Venturelli G Contini S Bonazzi A Mangia A 1997 Weathering of ultrama1047297c rocksand element mobility at Mt Prinzera Northern Apennines Italy MineralogicalMagazine 61 765ndash778
Venturelli G Frey M 1977 Anchizone metamorphism in sedimentary sequences of the Northern Apennines Rendiconti Societagrave Mineralogica Italiana 33 109ndash123
Vieillard P 2002 A new method for the prediction of Gibbs free energies of formationofphyllosilicates (10 Aring and14 Aring)basedon theelectronegativityscale Claysand ClayMinerals 50 352ndash363
Vissers RLM Strating EHH 2001 Structures and microstructures in a thrust-relatedgreenschist facies tectonic melange Voltri Group (NW Italy) O1047297oliti 26 33ndash46
Viti C Mellini M 1997 Contrasting chemical compositions in associated lizardite andchrysotile in veins from Elba Italy European Journal of Mineralogy 9 585 ndash596
Viti C Mellini M 1998 Mesh textures and bastites in the Elba retrogradeserpentinites European Journal of Mineralogy 10 1341ndash1359
Wenner DB Taylor Jr HP 1974 Deuteriumatomic hydrogen and oxygen-18oxygen-16 studies of serpentinization of ultrama1047297c rocks Geochimica et CosmochimicaActa 38 1255ndash1286
Williams LA Parks GA Crerar DA 1985 Silica diagenesis I Solubility controls Journal of S edimentary Petrology 55 301ndash311
Wilson J Savage D Cuadros J Shibata M RagnarsdottirKV 2006 The effect of ironon montmorillonite stability (I) Background and thermodynamic considerations
Geochimica et Cosmochimica Acta 70 306ndash322WollastR MackenzieFTBricker OP1968 Experimental precipitation and genesis of sepiolite at earth-surface conditions American Mineralogist 531 1645ndash1662
Wrucke CT1996 Serpentine and carbonated-hosted asbestos deposits InEdwardAdu Bray (Eds) Preliminary compilation of descriptive geoenvironmental mineraldeposit models 95ndash831 US Geological Survey Open-File Report pp 39 ndash46
Xu T Apps JA Pruess K 2001 Analysis of mineral trapping for CO2 disposal in deepaquifers Lawrence Berkeley National Laboratory University of California
ZhangYDaweRA2000 In1047298uence ofMg2+ on thekineticsof calcite precipitation andcalcite crystal morphology Chemical Geology 163 129ndash138
Zhu C Anderson G 2002 Environmental applications of geochemical modelingCambridge University Press
91T Boschetti L Toscani Chemical Geology 257 (2008) 76 ndash91
