Why magnetite is not the only indicator of past rainfall in the … · Geophys. J. Int. (2018) 213,...

Geophys J Int (2018) 213 2128ndash2137 doi 101093gjiggy097Advance Access publication 2018 March 13GJI Geomagnetism rock magnetism and palaeomagnetism

Why magnetite is not the only indicator of past rainfall in theChinese Loess Plateau

Xuelian Guo12 Subir K Banerjee2 Ronghua Wang1 Guoyong Zhao3 Hong Song1

Bin Lu4 Qian Li1 and Xiuming Liu45

1School of Earth Sciences Key Laboratory of Western Chinarsquos Mineral Resources of Gansu Province Lanzhou University Lanzhou 730000 China E-mailxlguolzueducn2Institute for Rock Magnetism University of Minnesota Twin Cities Minneapolis MN 55414 USA3College of Urban and Environmental Science Xinyang Normal University Xinyang 464000 China4Research Centre of Global Change School of Geographical Science Fujian Normal University Fuzhou 350007 China5Department of Environment and Geography Macquarie University NSW 2109 Australia

Accepted 2018 March 10 Received 2018 February 21 in original form 2017 November 3

S U M M A R YThis study investigates the magnetic mineralogy of palaeosol S5 from Xifeng (XF) Linyou(LY) and Baoji (BJ) sections with increasing annual precipitation from north to the southon the Chinese Loess Plateau Palaeosol S5 samples from these three localities are furtherprepared as magnetic extracts and separation residues Low-temperature magnetic measure-ments including field cooled and zero field cooled (FCZFC) remanence in-phase magneticsusceptibility thermal remanent magnetization and room temperature saturation isothermalremanence magnetization (RTSIRM) with X-ray diffraction measurements are carried outfor all magnetic extracts and separation residues samples The asymmetric rounded lsquohumprsquoin cooling curves on RTSIRM and the lsquotiltedrsquo Verwey transition on ZFCFC curves suggestthat partially oxidized magnetite is the dominant magnetic contributor not pure maghemite ormagnetite Furthermore The Verwey transitions on cooling curves slightly decrease and theincreased slope of lsquotiltedrsquo Verwey transition on ZFC remanence curves show that the degree ofoxidation of magnetite between localities increases in the order XFndashLYndashBJ Hard isothermalremanent magnetization X-ray diffraction data and the difference of magnetization in warmingcurves of RTSIRM suggest that both hematite concentration in magnetic extracts and goethiteconcentration in separation residues increase from XF to BJ Frequency-dependent suscepti-bility and ZFCFC curves show that BJS5 layer formed under high palaeoprecipitation has lesssuperparamagnetic (SP) but more single domain to pseudo-single domain particles becauseSP maghemite was dissolved and transformed into goethite by temporary waterlogging Theincrease in hematite concentration is interpreted as due to SP maghemite oxidation or originalgoethite dehydration within dry soil environment Therefore transformation of maghemite togoethite in waterlogged phases of the S5 palaeosol led to the loss of magnetization

Key words Asia Environmental magnetism Rock and mineral magnetism

1 I N T RO D U C T I O N

Magnetic mineralogy of soil and its relationship with paleoclimateshave been studied over 40 yr and the magnetic enhancement inpalaeosols has proven to be very useful for paleoclimatologists torecover the paleoprecipitation (Liu et al 1995 Maher et al 2003Geiss et al 2008 Balsam et al 2011 Orgeira et al 2011 Maher ampPossolo 2013 Hyland et al 2015 Maxbauer et al 2016) Accordingto recent rock magnetic and mineralogical models the superparam-agnetic (SP) or single domain (SD) pedogenic maghemitemagnetiteare responsible for the magnetic enhancements in palaeosols (Zhou

et al 1990 Liu et al 1992 Banerjee et al 1993 Heller et al 1993Maher amp Thompson 1994 Fine et al 1995 Hunt et al 1995 Ma-her 1998 Spassov et al 2003 Liu et al 2005 Qiang et al 2005Nie et al 2010 Zhao et al 2016) Besides pedogenic goethitehematite magnetite and maghemite have also been identified frommagnetic characteristics (Hu et al 2015) and diffuse reflectancespectroscopy (Torrent et al 2007 Balsam et al 2011) the ratiosand contents of these magnetic minerals have certain relationshipwith the paleoprecipitation (Long et al 2011 Orgeira et al 2011Liu et al 2013)

2128 Ccopy The Author(s) 2018 Published by Oxford University Press on behalf of The Royal Astronomical Society

Dow

nloaded from httpsacadem

icoupcomgjiarticle-abstract213321284931740 by M

acquarie University user on 16 April 2019

Past rainfall indicators in Chinese Loess Plateau 2129

Previous study (Orgeira et al 2011) observed that relatively dryclimates (lt700 mm yminus1) provide the best conditions for pedogenicmaghemite formation the rate of pedogenic maghemite formationis proportional to rainfall and do appear to show the expected in-crease in the values of chosen magnetic parameters (susceptibilityits frequency dependence anhysteretic remanence etc) Howeveran increase of the mean annual rainfall is not accompanied by a pro-portional increase of pedogenic maghemite formation Soils becomepermanently humid when rainfall increases lacking the alternationof wet and dry conditions required for pedogenic maghemite for-mation when mean annual precipitation (MAP) increases above acertain critical value (gt700ndash1000 mm yminus1) pedogenic maghemiteis destroyed by reductive dissolution (gleization) these magneticsignals begin to decrease with further increasing rainfall They havealso pointed out that the alteration and formation of different ironoxides (hematite goethite magnetite maghemite) in soil depend onnot simply the amount of rainfall but also the retention of moistureand pedogenic gleization

Goethitehematite ratio has been proposed earlier by Schwert-mann and Kampf (1985) as a temperature proxy for soil of thetemperate zone and pointed to be as palaeoprecipitation indicatorin soil in recent rock magnetic investigations (Long et al 2011Liu et al 2013 Hyland et al 2015) The frequency dependenceof susceptibility to hard isothermal remanent magnetization ratio(χ fdHIRM) of modern soils from Shanxi are found to have a goodcorrelation with the modern rainfall where χ fd may be a proxy forSP magnetitemaghemite and HIRM for hematite (Liu et al 2013)

This paper discusses the occurrence of magnetic minerals inloessic soils along a transect over the Chinese Loess Plateau therelation of these minerals to climatic conditions and the magneticmethods used for their identification To identify the compositionalvariability of all the oxidized iron minerals and to estimate theiramounts as accurately as possible different methods including di-rect microscopic observation of thin section scanning electron mi-croscope and transmission electron microscope of the magneticextracts (Chen et al 2005 Yang et al 2013) have been applied anda variety of magnetic properties and their changes under differenttemperatures and frequencies different methods have been tried(Zhou et al 1990 Geiss amp Zanner 2006 Michal et al 2010 Huet al 2015) By comparison low-temperature measurements canlead to accurate and thermally non-destructive ways for oxidizediron minerals identification and grain-size estimates even below20 nm

In order to explain why high annual rainfall can lead to the loss ofmagnetization and to further clarify the transformation of oxidizediron minerals in waterlogged soil environments we collect well-developed palaeosol S5 samples from three locations (XF LY andBJ) from more dry in the north to more humid environments in thesouth on the Chinese Loess Plateau Oxidized iron minerals andgrain-size of these samples are identified and estimated by usinglow-temperature magnetic measurements supplemented by X-raydiffraction Through these thermally non-destructive ways we getaccurate estimations of relative amounts of different iron oxides inS5 samples formed under different humid conditions and discussthe relation of these minerals to climatic environments

2 S A M P L E S A N D M E A S U R E M E N T S

21 Sample collection

The samples for study were collected from the S5 palaeosols atthe Xifeng (XF) Linyou (LY) and Baoji (BJ) sections on ChineseLoess Plateau (Guo et al 2015) The present climate of the areais semi-arid to mildly humid with MAP of 550 680 and 720 mm(up to 1100 mm in Qinling Mountain area) from XF to LY to BJrespectively Sixty per cent of precipitation occurs from July toSeptember The mean annual temperatures (MAT) are 87 91 and12 C respectively Precipitation which carried by the East Asiansummer monsoon decreases from south (BJ) to north (XF)

22 Magnetic extraction

Magnetic extraction of the three natural samples is carried out by us-ing the magnetic extraction apparatus aimed at separating stronglyand weakly magnetic particles for characterization The magneticextraction of the bulk sample yields two subsamples (1) mineralparticles collected with an applied magnetic field defined as themagnetic extracts (MAG) and (2) mineral particles that are non-separated fraction defined as the residual (RES)

The magnetic extraction procedure is same in design to the pumpmethod described by Strehlau et al (2014) similar to the systemdescribed in Reynolds et al (2001) A certain original S5 palaeosolsample was dispersed in 30 mL Milli-Q water using 1 g of dispersantsodium hexametaphosphate centrifuged at 5000 rpm for 3 minusing an Eppendorf 5804 centrifuge then added to the reservoirwith 200 mL Milli-Q water The suspension was circulated by aMasterflex LS (Cole Parmer) peristaltic pump at sim200 mL minminus1

and passed through Tygon LS flexible plastic tubing (Saint Gobain)oriented vertically with a joint that contained an Nd magnet coveredwith a plastic sleeve to collect the magnetic material After 90 minthe magnetic extracts were washed from the sleeve with Milli-Qwater and collected for analysis The remaining suspension in thereservoir was collected as the residues and then dried for analysis

23 Measurements

Anhysteretic remanent magnetization (ARM) was measured in Al-ternating Field Demagnetizer with pARM device the peak AF fieldused was 100 mT and the biasing field is 005 mT ARM was thennormalized by the bias field to obtain ARM susceptibility (χARM)Hysteresis loops were measured using a Princeton MeasurementsVibrating Sample Magnetometer with the maximum field of 15T We used the loopsrsquo hysteresis properties to determine saturationmagnetization (Ms) saturation remanent magnetization (Mrs) co-ercivity (Bc) remanent coercivity (Bcr) relative antiferromagneticcontent (HIRM) and relative ferrimagnetic content (S300)

A quantum design Magnetic Properties Measurement System(MPMS2) cryogenic susceptometer was used to measure coolingand warming curves of room temperature saturation isothermal re-manence magnetization (RTSIRM) field cooling (FC) and zerofield cooling (ZFC) remanence curves and also for thermal rema-nent magnetization (TRM) SIRM produced in a 25 T field at 300 Kwas measured continuously during ZFC to 20 K at 5 K steps andback to 300 K ZFC and FC is based on performing two consecutivemagnetization measurements In ZFC the sample is first cooleddown in the absence of a magnetic field and then measured in a 25T field at increasing temperature FC is performed in a 25 T fieldat decreasing temperature

Dow




2130 X Guo et al

A second set of low-temperature measurements consisted of mea-suring AC susceptibility frequency- and field-dependence over thetemperature range of 10ndash300 K at 10 K measuring steps Thein-phase susceptibility (χ acute) of the samples were measured in aconstant 03 mT field and AC frequency of 1 10 and 100 Hzχ prime = χ prime

1 Hzmdashχ prime100 Hz Goethite test using low-temperature be-

haviour of a TRM acquired by FC from 400 to 300 K in a 03mT field and of an isothermal remanence magnetization (IRM) ac-quired at 300 K in a 03 mT field after ZFC from 400 to 300 K(Lascu amp Feinberg 2011)

X-ray diffraction (XRD) patterns were obtained with a PANalyti-cal Xrsquopert Pro theta-theta diffractometer equipped with a Co anodeand an Xrsquocelerator detector and calculate out the main oxidized ironconcentrations

3 R E S U LT S

31 Hysteresis characterization

Hysteresis loops measurements give information about the magneticmineral composition and particle size (Evans amp Heller 2003) Allloops of MAG samples (Fig 1a) close at or above 300 mT indicatingthe dominant presence of softer ferrimagnetic minerals such asmagnetite andor maghemite However the loops also show lackof complete saturation (which requires near zero slope of the loopat the highest fields) above 300 mT indicating that some hardermagnetic minerals such as hematite andor goethite may be presentin MAG (Guo et al 2013) The hysteresis parameters of Mrs Bc Bcr

increase monotonically from XF to BJ (Fig 1a) pointing to increasein high-coercivity mineral such as hematite or goethite or it couldalso indicate increase in SD magnetitemaghemite

The loops of RES samples before (black curves) and after (redcurves) paramagnetic correction (Fig 1b) show the presence ofsofter ferrimagnetic minerals and hard magnetic minerals Never-theless the Mrs of RES are all below 001 Am2 kg-1 (Fig 1b)which are less than 04 per cent of the Mrs of MAG (2ndash3 Am2

kg-1 Fig 1a) meaning that most ferrimagnetic minerals exist inthe MAG samples Bc and Bcr as concentration of hard magneticminerals like hematite andor goethite are usually estimated fromincrease in HIRM (King amp Channel 1991) but hematite dominatesthe HIRM values (Liu et al 2010 Nie et al 2010) The relativeabundances of soft ferrimagnetic and hard antiferromagnetic min-erals are commonly quantified using the S300 (Bloemendal et al1992 Evans amp Heller 2003) From north to south HIRMs of MAGincrease from 0025 to 0063 and to 0107 Am2 kg-1 HIRMs of RESare around 000053ndash000060 Am2 kg-1 with no obvious variationsamong three different samples S300 decrease from 098 to 095 to093 in MAG samples and S300 decrease from 0876 to 0864 to0702 in the RES respectively Taken together comparison of pa-rameters Mrs Bc Bcr HIRM S300 data and loops of RES samplesafter paramagnetic correction of the XF LY and BJ MAG and RESsamples also confirm our preliminary conclusion that hard magneticminerals increase and relative content of soft magnetic minerals tohard magnetic minerals decrease along the transect XFndashLYndashBJ

32 Low-temperature susceptibility

Examination of magnetic susceptibility as a function of temperatureand field frequency has been shown to be a useful tool in distin-guishing composition and grain size controls on low-temperaturemagnetic behaviour (Moskowitz et al 1998 Brachfeld amp Banerjee

2000) The in-phase magnetic susceptibility (χ prime) components of theMAG fractions exhibit similar features (Fig 2a) The MAG frac-tions contain magnetite as displayed by the Verwey transition (Tvasymp 130 K Banerjee et al 1993) but Verwey transition in Fig 2(a)is not sharp as for pure magnetite the tilted straight line indicatesa slight oxidation of magnetite or lsquomaghemitersquo composition eventhough pure maghemite would not have displayed Verwey transi-tion Above Tv most particles are SP due to magnetic susceptibilityof iron oxide minerals (such as magnetite andor maghemite) andfield frequency is inverse so with the increasing temperature thefrequency susceptibility rises pointing to the SP fractions increaseBJS5 samples show no frequency dependence of magnetic suscep-tibility at any temperature (Fig 2a) indicating minimal contents ofSP grains (Brachfeld amp Banerjee 2000) By contrast SP fractionsincrease from XF to LY then decline to BJ in the MAG minerals

The residues of XF and LY show SP presence (Fig 2b) BJ quitedifferent because SP contents are very low The sharp increasedχ prime components at T lt 50 K could be due to paramagnetic clay orbecause of cooling below the Neel temperature of an antiferromag-netic iron silicate minerals with very little Fe

The low-temperature difference (χ prime) of χ prime is measured in twospecific frequencies (1 and 100 Hz Fig 2c) The increasing χ prime ofMAG fractions at T gt 50 K point to SP fraction presence BJ showslittle SP behaviour in agreement with Figs 2(a) and (b) The χ prime

of three samples all drop sharply below 50 K when paramagneticor antiferromagnetic behaviour becomes important But comparingto MAG the χ prime of three RES samples (Fig 2d) are near to zeropoint to very low SP fractions in RES

Clear frequency dependence susceptibility in Fig 2 suggest thatSP fractions increase from XF to LY and decline to BJ with veryless SP fractions and most of χ is carried by the MAG

33 Mineral compositional information from cooling ofRTSIRM

The room temperature remanence curves can provide estimates ofoxidation from the shapes of cooling curves and also can determinethe types of magnetic minerals (Ozdemir amp Dunlop 2010) The Ver-wey transitions (Tv = sim130 K) on cooling curves (Fig 3a) indicatethe existence of magnetite in MAG samples When temperature ap-proaching Tv the asymmetric rounded lsquohumprsquo in cooling curves forall three sites confirm earlier suspicion that it is oxidized magnetite(Ozdemir amp Dunlop 2010 Guo et al 2015) not pure magnetite orpure maghemite that is the dominant magnetic carrier (Ozdemir ampDunlop 2010) From the normalized comparison of warming curves(Fig 3c) we see that BJS5-Mag shows the highest oxidation whileXFS5-MAG is the lowest Tv shows a slight decrease from 132 131to 126 K indicating the oxidation degree of magnetite enhance fromXFndashLYndashBJ (Ozdemir et al 1993) The presence of Morin transitionof hematite at sim220 K in BJS5-MAG (Fig 3a) shows appreciablyhigh amount of hematite (Ozdemir amp Dunlop 2002) or single largehematite particles (Ozdemir et al 2008) It is lsquoappreciably highrsquobecause Morin transition is seen over and above a background ofoxidized magnetite signal even though hematite magnetization is200 times weaker (05 Am2 kg-1 versus 93 Am2 kg-1) in saturationmagnetization than magnetite There is a subtle but recognizableMorin transition at sim220 K in BJS5-RES (Fig 3b) show a smallhematite presence

Finally the warming curves of RES samples from all three sites(Figs 3b and d) show strong decrease between 20 and 300 K usuallyshown by goethite (Carter-Stiglitz et al 2006) The difference of

Dow





Figure 1 Hysteresis loops of MAG and the RES samples for the S5 transect from XF LY and BJ The hysteresis loops were measured in fields up to plusmn15T to saturate ferrimagnetic magnetite and maghemite (a) Corresponds to MAG (b) hysteresis loops of RES before (black curves) and after (red curves)paramagnetic correction (gt500 mT) The values of saturation remanence (Mrs) saturation magnetization (Ms) coercivity (Bc) and coercivity of remanence(Bcr) are listed inside the hysteresis loop plots all values are prior to high-field slope correction

magnetization (M = M20K minus M300K) in the warming curves is 652484 and 378 times 10minus4 Am2 kg-1 respectively (Fig 3d) indicatingthe goethite fractions increase from XFndashLYndashBJ

In all the partially oxidized magnetite is the dominant magneticcarrier and the degree of oxidation enhance from XFndashLYndashBJ BJsample has appreciably high amount of hematite or single largehematite particles the RES fractions show strong likelihood ofgoethite presence in all three sites with the highest amount beingin BJ

34 Magnetic mineral composition and grain sizesconfirmation from low temperature

Low-temperature ZFCFC measurements were performed to fur-ther reveal the magnetic assemblage in the samples (Brachfeld ampBanerjee 2000) The tilted Verwey transition at sim120 K on ZFCFCcurves (Fig 4a) point to the presence of partially oxidized mag-netite not pure magnetite in XF LY and BJ MAG samples Fromthe normalized comparison of ZFC remanence curves (Fig 4c)we see that the slope of lsquotiltedrsquo Verwey transition increase fromXFndashLYndashBJ indicating the oxidation degree of magnetite enhancedThese further confirm RTSIRM results that it is partially oxidized

magnetite not pure magnetite or pure maghemite that is the dom-inant magnetic carrier Moreover FC curve has higher values thanZFC curve (Fig 4a) indicating SD andor PSD behaviour not largeparticles (gt10 μm) in magnetic separates SDPSD are the mainlymagnetic particles In contrast ZFC curves of three RES samples(Fig 4b) without obvious Verwey transition at sim120 K show thatthe particles are oxidized and finer (more SP) grains the steeptemperature-dependence curves suggest the presence of high coer-civity magnetically unsaturated goethite in the RES (Guyodo et al2003 Carter-Stiglitz et al 2006) The remanence loss is approx-imately 67 per cent 66 per cent and 60 per cent of the initialremanence for XF LY and BJ RES samples (Fig 4d) suggestingSP fractions are more in XF and LY than in BJ and show a slightlydecrease upon going from XFndashLYndashBJ

Overall the oxidized magnetite is the main magnetic carrier inMAG and the oxidation degree enhance from XFndashLYndashBJ In addi-tion more goethite present in RES By contrast BJS5 sample hasminimal SP fractions This consistent with the RTSIRM characters(Fig 3) and in-phase susceptibility (χ prime) results (Fig 2)

Dow




2132 X Guo et al

Figure 2 Low-temperature variations of in-phase magnetic susceptibility (χ prime) for field frequencies of 1 10 and 100 Hz for MAG samples (a) and RES samples(b) from XF LY and BJ Low-temperature difference (χ prime) of frequency dependence χ prime in two specific frequencies (1 and 100 Hz) from MAG samples (c)and RES samples (d) from XF LY and BJ

35 Oxidized iron concentrations in magnetic extracts

In order to quantify the amount of oxidized iron minerals wemeasured MAG samples using XRD the results are shown inFig 5 XFS5-MAG has the least hematite and the highest mag-netite concentrations LYS5-MAG has the same magnetite andmaghemite concentrations as BJS5-MAG and intermediate amountsof hematite BJS5-MAG has the highest hematite concentrationwhich is about 27 times higher than magnetite or maghemite con-centration The total concentration of magnetite + maghemite (576per cent 470 per cent 417 per cent respectively) decline from XFndashLYndashBJ Hematite concentration increases from XFndashLYndashBJ whichquantitatively confirms the above results of RTSIRM FCZFC andhysteresis data Hematite have maximum contents in MAG sam-ples by contrast goethite concentration is very low in all MAGsamples (Fig 5) a reasonable explanation could be fine hematiteparticles often appeared on the edge and surface of ferrimagneticminerals and extracted easily (Hu et al 2013) the most of goethiteconcentration still left in the residues

36 Goethite test in separation residues

Goethite was identified using a rock magnetic test to target thismineral (Lascu amp Feinberg 2011) and this approach is similar to thatof demagnetizing the low coercivity minerals employed by Guyodoet al (2006) Curve a of Fig 6 is the ZFC from 300 to 20 K themagnetization increase and Verwey transition at sim120 K confirmthe presence of magnetite Above Tv the increasing magnetizationshow existence of high coercivity goethite Curve b linearly dropfrom 20 to 400 K shows the presence of goethite Both IRM coolingand warming curves (c d) in a 03 T field display low coercivitymagnetite character because goethite has very little remanence in 03T field The overlap of e and f curves confirm the presence of goethitein BJS5-RES sample the difference at room temperature betweenthe total magnetization and remainder was taken to represent theconcentration of goethite in the BJS5-RES sample

4 D I S C U S S I O N

We have studied S5-1 palaeosol from the Chinese Loess Plateauand suggested that pedogenesis and chemical weathering of thecoeval S5-1 palaeosol layers increased from north to south from

Dow





Figure 3 RTSIRM produced in a 25 T field at 300 K was measured continuously during zero field cooling to 20 K at 5 K steps and back to 300 K Panel (a)displays cooling and warming back of RTSIRM of MAG of S5 palaeosols from XF LY and BJ Panel (b) shows the same type of data for the RES from S5palaeosol from XF LY and BJ Panel (c) displays normalized RTSIRM on warming of MAG and panel (d) shows normalized RTSIRM on warming of RES ofS5 palaeosols from XF LY and BJ Imparting a high field SIRM to a sample containing magnetite or oxidized magnetite at room temperature and then cyclingthe remanence in zero fields from 300 to 20 to 300 K can be a very effective and non-destructive technique for identifying the compositions

localities XF to BJ Some fine-grained strongly magnetic mineralswere converted into weakly magnetic minerals (mainly hematiteand goethite) by pedogenesis which resulted in a decline in SP and

stable single domain (SSD) ferrimagnetic minerals and decreas-ing susceptibility of S5-1 palaeosol from north to south(Guo et al

Dow




2134 X Guo et al

Figure 4 Low-temperature field cooled (FC) and zero field cooled (ZFC) remanent magnetization acquired at 20 K in 25 T field from 300 to 20 K in MAG(a) and RES (b) from palaeosol S5 from XF LY and BJ (c) and (d) show normalized ZFC remanent magnetization in MAG and RES from palaeosol S5 fromXF LY and BJ

2015) We used low-temperature magnetism and XRD to quantita-tively examine how high annual rainfall in Chinese Loess Plateauleads to loss of magnetization and further clarify the transformationof oxidized iron in waterlogged soil environments from XF to BJ

41 The compositional variability of the oxidized ironminerals from XFndashLYndashBJ in Chinese Loess Plateau

According to above low-temperature magnetic behaviour (RT-SIRM χ prime FCZFC) we find that magnetic minerals in S5 palaeosol

Dow





Figure 5 Magnetic oxidized iron concentration of MAG acquired by XRDMght maghemite Ht hematite Mt magnetite Gt geothite

Figure 6 Low-temperature behaviour of a TRM (a and b triangles) ac-quired by field cooling (FC) from 400 to 300 K in a 03 T field and ofan IRM (c and d circles) acquired at 300 K in a 03 T field after zero-field cooling (ZFC) from 400 to 300 K the separation of the curves abovesim120 K (Verwey transition Tv) is diagnostic of magnetite The difference(squares) between the TRM and IRM warming (e square grey) and cooling(f square red) curves respectively is a measure of the presence of goethitewhich acquires remanence during the FC pre-treatment and is demagnetizedduring the ZFC pre-treatment (Guyodo et al 2006 Lascu amp Feinberg 2011)

are oxidized-magnetite maghemite hematite and goethite Hystere-sis parameters Bc Bcr and HIRM increase monotonically from XFndashLYndashBJ (Fig 1) but S300 and the total concentration of magnetite +maghemite (Fig 5) decline over the same environmental transectimplying to hard magnetic mineral concentrations increase and rel-ative concentrations of soft magnetic minerals descend along thetransect XFndashLYndashBJ The asymmetric rounded lsquohumprsquo in coolingcurves on RTSIRM (Figs 3a and c) and the lsquotiltedrsquo Verwey tran-sition on ZFCFC curves (Fig 4) suggest that partially oxidized

magnetite neither pure magnetite nor pure maghemite is the dom-inant magnetic carrier Low-temperature magnetic properties are inagreement with magnetic hysteresis parameters the partially oxi-dized magnetite in SDndashPSD ranges can reliably record paleomag-netic signals (Ge et al 2014) From the normalized comparison ofcooling curves on RTSIRM we see that Tv slightly decrease from132 131 to 126 K (Fig 3c) and the increased slope of lsquotiltedrsquo Ver-wey transition of ZFC remanence curves (Fig 4c) from XFndashLYndashBJshow that the oxidation degree of magnetite enhance with increas-ing MAT (87 C 91 C to 12 C) from north to south The ratiosof FeDFeT elementsrsquo concentrations and redness values also con-firm that pedogenic degree enhances from north to south in ChineseLoess Plateau (Hao amp Guo 2005 Guo et al 2015)

The steep temperature-dependent ZFCFC curves (Fig 4b) andRTSIRM on warming curves of Res samples (Figs 3b and d) indi-cate the presence of goethite we also measured BJS5-RES by low-temperature behaviour of TRM and IRM (Fig 6) The overlap ofthe difference between the TRM and IRM warming (Fig 6e squaregrey) and cooling (Fig 6f square red) curves further confirm thepresence of goethite (Lascu amp Feinberg 2011) Simultaneously theM (M = M20K minus M300K) in warming curves of RTSIRM (Fig 3d)show the goethite fractions increase from XF to BJ These are con-sistent with the field observations no FendashMn coatings were seen inXF S5 palaeosol layer and only a little amount of FendashMn coatingsin LY S5 palaeosol layer and abundant FendashMn coatings in BJ S5palaeosol layer (Guo et al 2015)

The Morin transition of hematite at sim220 K in BJS5-MAG andBJS5-RES (Figs 3a and b) show appreciably high hematite con-centration or single large hematite particles (Ozdemir et al 2008)But XRD data quantitatively suggest hematite concentration are 27times higher than magnetite maghematite and geothite concentra-tions in MAG samples for all three sections and rapidly increasewith increasing MAP from XF to BJ (Fig 5) HIRM data (from0025 and 0063 to 0107 Am2 kg-1) of MAG also confirm this in-terpretation HIRM of RES are 000055 000053 and 000060 Am2

kg-1 with no obvious variations among three different samples thisconfirm that hematite dominates the HIRM value (Liu et al 2010Nie et al 2010) All suggest that hematite concentration of BJS5-MAG is indeed high a reasonable explanation could be its fineparticles often appeared on the edge and surface of ferrimagneticminerals and extracted easily (Hu et al 2013)

From XF to LY with low MAP (which was 550 and680 mm) magnetite was oxidized to maghemite during pedoge-nesis maghemite continued to be oxidized to hematite under dryconditions (Liu et al 2008) Thus magnetite concentration declinesand maghemite and hematite concentrations go up from XF to LYFrom LY to BJ (modern MAP from 680 to 720 mm) pedogenesisoccurred intermittently between wet and dry conditions In water-logged soil environments fine-grain maghemite dissolved releasingFe3+ and goethite was precipitated (Schwertmann amp Murad 1983)The increase of hematite concentration with increasing precipita-tion from XF to BJ is due to dehyrdation of original goethite oroxidation of maghemite While maghemite concentration displaysa little reduction due to pedogenic maghemite is destroyed underreducing conditions during the wetting phase (Orgeira et al 2011)In humid climates where MAP exceeds sim1000 mm yminus1 modernsoil shows that negative correlations between MAP and magneticenhancement parameters (Balsam et al 2011 Long et al 2011)This is attributed to the increased dissolution of iron oxides andleaching that persists in water-saturated soil with only limited dryperiods (Maher 2011 Orgeira et al 2011)

Dow




2136 X Guo et al

42 The compositional variability of pedogenic magneticparticles from oxidizing to weakly reducing environments

Frequency dependent susceptibility χ prime curves (Fig 2) show that SPparticles are more abundant in XF and LY samples than in BJ GreatFC than ZFC magnetization below the Verwey transition (Fig 4) isindicative of an SD to PSD dominated magnetite grain size distribu-tion Moreover the low-temperature frequency susceptibility χ primendashTcurves (Fig 2) show SP composition slightly increases from XF toLY and then rapidly decline to BJ with very low SP compositionsin BJ S5 sample Likewise as an extremely sensitive indicator forSSD particles (King amp Channell 1991) the χARM of MAG are 195193 and 175 times 10minus4 m3 kg-1 among XFS5-MAG LYS5-MAG andBJS5-MAG samples the χARM of BJS5-RES (93 times 10minus5 m3 kg-1)is higher than XFS5-RES (65 times 10minus6 m3 kg-1) and LYS5-RES(55 times 10minus6 m3 kg-1) in RES samples In all the SP compositionincreases and SDPSD composition decreases from XF to LY be-cause with increasing pedogenesis the magnetite was oxidized tomaghemite and hematite while BJS5 has much more SDPSD par-ticles very less SP particles This may be because SP maghemitewas dissolved and recrystallized into goethite under temporary wa-terlogging caused by abundant rainfall in BJS5 palaeosol This iscompatible with the observation of Smirnov amp Tarduno (2000) whosuspected selective elimination of small grains first The dissolu-tion of magnetic minerals commonly occurs in weakly reducing orgleyed environments (Liu et al 2008)

5 C O N C LU S I O N S

Low-temperature magnetic measurements and XRD study of MAGand RES from XF LY and BJ S5 palaeosols show that

(1) The oxidized magnetite not pure maghemite or pure mag-netite is the main magnetic carrier in S5 palaeosols and the oxida-tion degree of magnetite enhances along section from XFndashLYndashBJ

(2) Both hematite concentration of MAG and goethite concen-tration of RES increase with increasing MAP from XF to BJ Therapid increase of hematite concentration is interpreted as previouslyformed goethite dehydration or SP maghemite oxidized within drysoil environment

(3) The SP concentration increases and SDPSD concentrationdecreases from XF to LY because with increasing pedogenesis themagnetite was oxidized to maghemite and hematite while BJS5has much more SDPSD particles very less SP particles due toSP maghemite was dissolved and transformed into goethite undertemporary waterlogging caused by abundant rainfall which resultedin goethite concentration increasing

S U P P O RT I N G I N F O R M AT I O N

Supplementary data are available at GJI onlineTable Hysteresis parameters of MAG and RES samples before

high-field slope correctionPlease note Oxford University Press is not responsible for the

content or functionality of any supporting materials supplied bythe authors Any queries (other than missing material) should bedirected to the corresponding author for the paper

A C K N OW L E D G E M E N T S

The low-temperature magnetic measurements were made at the In-stitute for Rock Magnetism (IRM) University of Minnesota XRD

was measured at Department of Chemistry University of Min-nesota We thank Mike Jackson Dario Bilardello and Peat Soslashlheidof IRM for their help with the experiments and thank Prof R LeePenn and PhD Alex Henrique Pinto of Department of ChemistryUniversity of Minnesota for their help with the XRD measure-ments The IRM is supported by US National Foundations EARIFdivision and the University of Minnesota This is IRM contribu-tion no1605 This research was supported by the National Natu-ral Science Foundation of China (grant nos 41772168 4177218041402147 41402149 and 41602187) XG was further supported byScientific Research Foundation for the Returned Overseas ChineseScholars Gansu Province

R E F E R E N C E SBalsam WL Ellwood BB Ji JF Williams ER Long XY amp Hassani

AE 2011 Magnetic susceptibility as a proxy for rainfall worldwidedata from tropical and temperate climate Quat Sci Rev 30 2732ndash2744

Banerjee SK Hunt CP amp Liu XM 1993 Separation of local signalsfrom the regional paleomonsoon record of the Chinese Loess Plateau arock-magnetic approach Geophys Res Lett 20(9) 843ndash846

Bloemendal J King JW Hall FR amp Doh SJ 1992 Rock magnetismof Late Neogene and Pleistocene deep-sea sediments relationship tosediment source diagenetic processes and sediment lithology J geophysRes 97 4361ndash4375

Brachfeld S A amp Banerjee SK 2000 Rock-magnetic carriers of century-scale susceptibility cycles in glacial-marine sediments from the PalmerDeep Antarctic Peninsula Earth planet Sci Lett 176 443ndash455

Carter-Stiglitz B Moskowitz B Solheid P Berquo TS Jackson M ampKosterov A 2006 Low-temperature magnetic behavior of multi domaintitanomagnetites TM0 TM16 and TM35 J geophys Res 111(B12)

Chen TH Xu HF Xie QQ Chen J Ji JF amp Lu HY 2005 Char-acteristics and genesis of maghemite in Chinese loess and paleosolsmechanism for magnetic susceptibility enhancement in paleosols Earthplanet Sci Lett 240 790ndash802

Evans ME amp Heller F 2003 Environmental Magnetism Principles andApplications of Enviromagnetics Academic Press pp 1ndash299

Fine P Verosub KL amp Singer MJ 1995 Pedogenic and lithogenic contri-butions to the magnetic susceptibility record of the Chinese loesspaleosolsequence Geophys J Int 122 97ndash107

Ge KP Williams W Liu QS amp Yu YJ 2014 Effects of the core-shell structure on the magnetic properties of partially oxidized magnetitegrains experimental and micromagnetic investigations Geochem Geo-phys Geosyst 15 2021ndash2038

Geiss CE amp Zanner CW 2006 How abundant is pedogenic magnetiteAbundance and grain size estimates for loessic soils based on rock mag-netic analyses J geophys Res 111 B12S21

Geiss CE Egli R amp Zanner CW 2008 Direct estimates of pedogenic-magnetite as a tool to reconstruct past climates from buried soils Jgeophys Res 113 B11102

Guo XL Liu XM Li PY Lu B Guo H Chen Q amp Ma MM2013 The magnetic mechanism of paleosol S5 in the Baoji section of thesouthern Chinese Loess Plateau Quat Int 306 129ndash136

Guo XL Liu XM Miao S J Zhao GY amp Liu YX 2015 Variabilityof magnetic character of S5-1 paleosol (age sim 470 Ka) along a rainfalltransect explains why susceptibility decreased with high rainfall AeolianRes 19 55ndash63

Guyodo Y Mostrom A Lee PR amp Banerjee SK 2003 From nanodotsto nanorods Oriented aggregation and magnetic evolution of nanocrys-talline goethite Geophys Res Lett 30 19ndash11

Guyodo Y Banerjee SK Lee PR Burleson D Berquo TS Seda Tamp Solheid P 2006 Magnetic properties of synthetic six-line ferrihydritenanoparticles Phys Earth planet Inter 154 222ndash233

Hao QZ amp Guo ZT 2005 Spatial variations of magnetic susceptibilityof Chinese loess for the last 600 kyr implications for monsoon evolutionJ geophys Res 110 B12101

Dow





Heller F Shen CD Beer J Liu XM Liu TS Bronger A Suter Mamp Bonani G 1993 Quantitative estimates of pedogenic ferromagneticmineral formation in Chinese loess and palaeoclimatic implications Earthplanet Sci Lett 114 385ndash390

Hu PX Liu QS Torrent J Barron V amp Jin CS 2013 Characterizingand quantifying iron oxides in Chinese loesspaleosols implications forpedogenesis Earth planet Sci Lett 369ndash370 271ndash283

Hu PX Liu QS Heslop D Roberts A P amp Jin CS 2015 Soil moisturebalance and magnetic enhancement in loessndashpaleosol sequences from theTibetan Plateau and Chinese Loess Plateau Earth planet Sci Lett 409120ndash132

Hunt CP Banerjee SK Han JM Solheid PA Oches E Sun WWamp Liu TS 1995 Rock magnetic proxies of climate change in the loess-paleosol sequences of the western Loess Plateau of China Geophys JInt 123 232ndash244

Hyland E Sheldon ND Van der Voo R Badgley C amp Abrajevitch A2015 A new paleoprecipitation proxy based on soil magnetic propertiesimplications for expanding paleoclimate reconstructions Bull geol SocAm 127(7) 975ndash981

King J amp Channell J 1991 Sedimentary magnetism environmental mag-netism and magneto-stratigraphy 1987ndash1990 Rev Geophys 39 358ndash370

Lascu I amp Feinberg J M 2011 Speleothem magnetism Quat Sci Rev30 3306ndash3320

Liu QS Torrent J Maher BA Yu YJ Deng CL Zhu RX amp ZhaoXX 2005 Quantifying grain size distribution of pedogenic magneticparticles in Chinese loess and its significance for pedogenesis J geophysRes 110 B11102

Liu QS Barron V Torrent J Eeckhout SG amp Deng CL 2008 Mag-netism of intermediate hydromaghemite in the transformation of 2-lineferrihydrite into hematite and its paleoenvironmental implications J geo-phys Res 113 B01103

Liu QS Hu PX Torrent J Barron V Zhao XY Jiang ZX amp SuYL 2010 Environmental magnetic study of a Xeralf chronosequence innorthwestern Spain indications for pedogenesis Palaeogeogr Palaeocli-matol Palaeoecol 293144ndash156

Liu XM Shaw J Liu TS Heller F amp Yuan BY 1992 Magneticmineralogy of Chinese loess and its significance Geophys J Int 108301ndash308

Liu XM Rolph T Bloemendal J Shaw J amp Liu TS 1995 Quantitativeestimates of paleoprecipitation at Xifeng in the loess plateau of ChinaPalaeogeogr Palaeoclimatol Palaeoecol 113 243ndash248

Liu ZF Liu Q S Torrent J Barronc V amp Hu PX 2013 Testingthe magnetic proxy χFDHIRM for quantifying paleorainfall in modernsoil profiles from Shaanxi Province China Glob Planet Change 110368ndash378

Long X Ji J amp Balsam W 2011 Rainfall-dependent transformations ofiron oxides in a tropical saprolite transect of Hainan Island South Chinaspectral and magnetic measurements J geophys Res 116 F03015

Maher BA 1998 Magnetic properties of modern soils and quaternary loes-sic paleosols paleo-climatic implications Palaeogeogr PalaeoclimatolPalaeoecol 137 25ndash54

Maher BA 2011 The magnetic properties of Quaternary aeolian dusts andsediments and their palaeoclimatic significance Aeolian Res 3 87ndash144

Maher BA amp Possolo A 2013 Statistical models for use of palaeosolmagnetic properties as proxies of palaeorainfall Glob Planet Change111 280ndash287

Maher BA amp Thompson R 1994 Comments on pedogenesis and pale-oclimate interpretation of the magnetic susceptibility record of Chineseloess-paleosol sequences Geology 23 857ndash858

Maher BA Alekseev A amp Alekseeva T 2003 Variation of soil mag-netism across the Russian steppe its significance for use of soil magnetismas a palaeorainfall proxy Quat Sci Rev 21 1571ndash1576

Maxbauer DP Feinberg JM amp Fox DL 2016 Magnetic mineral assem-blages in soils and paleosols as the basis for paleoprecipitation proxies areview of magnetic methods and challenges Earth-Sci Rev 155 28ndash48

Michel FM Barron V Torrent J Morales MP Serna CJ Boily JFLiu QS Ambrosini A Cismasu AC amp Brown GE 2010 Orderedferrimagnetic form of ferrihydrite reveals links among structure compo-sition and magnetism Proc Natl Acad Sci USA 107 2787ndash2792

Moskowitz BM Jackson M amp Kissel C 1998 Low-temperature mag-netic behavior of titanomagnetites Earth planet Sci Lett 157 141ndash149

Nie JS Song YG King JW Fang XM amp Heil C 2010 HIRMvariations in the Chinese red-clay sequence insights into pedogenesis inthe dust source area J Asian Earth Sci 38 96ndash104

Orgeira MJ Egli R amp Compagnucci RH 2011 A quantitative model ofmagnetic enhancement in loessic soils in The Earthrsquos Magnetic Interiorpp 361ndash397 eds Petrovsky E Ivers D Harinarayana T amp Herrero-Bervera E Springer

Ozdemir O amp Dunlop DJ 2002 Thermoremanence and stable memoryof single-domain hematites Geophys Res Lett 29(18) 24ndash21

Ozdemir O amp Dunlop DJ 2010 Hallmarks of maghemitization in low-temperature remanence cycling of partially oxidized magnetite nanopar-ticles J geophys Res 115 B02101

Ozdemir O Dunlop DJ amp Moskowitz BM 1993 The effect of theVerwey transition in magnetite Geophys Res Lett 20 1671ndash1674

Ozdemir O Dunlop DJ amp Berquo TS 2008 Morin transition inhematite Size dependence and thermal hysteresis Geochem GeophysGeosyst 9

Qiang XK An ZS Li HM Chang H amp Song YG 2005 Magneticproperties of Jiaxian red clay sequences from northern Chinese LoessPlateau and its paleoclimatic significance Sci China Earth Sci 48 1234ndash1245

Reynolds RL Sweetkind DS amp Axford Y 2001 An inexpensive mag-netic mineral separator for fine-grained sediment US Geological SurveyOpen-File Report 1ndash281 7 p

Schwertmann U amp Kampf N 1985 Properties of goethite and hematitein kaolinitic soils of southern and central Brazil Soil Sci 139 344ndash350

Schwertmann U amp Murad E 1983 Effect of pH on the formation ofgoethite and hematite from ferrihydrite Clays Clay Miner 31 277ndash284

Smirnov AV amp Tarduao JA 2000 Low-temperature magnetic propertiesof pelagic sediments (Ocean Drilling Program site 805C) tracers of mag-nemitization and magnetic mineral reduction J geophys Res 105 16457ndash16 471

Spassov S Heller F Kretzschmar R Evans ME Yue LP amp Nour-galiev DK 2003 Detrital and pedogenic magnetic mineral phases inthe loesspalaeosol sequence at Lingtai (central Chinese Loess Plateau)Phys Earth planet Inter 140 255ndash275

Strehlau JH Hegner LA Strauss BE Feinberg JM amp PennRL 2014 Simple and efficient separation of magnetic minerals fromspeleothems and other carbonates J Sediment Res 84 1096ndash1106

Torrent J Liu QS Bloemendal J amp Barron V 2007 Magnetic enhance-ment and iron oxides in the upper Luochuan loessndashpaleosol sequenceChinese Loess Plateau Soil Sci Soc Am J 71 1570ndash1578

Yang TS Hyodo M Zhang SH Maeda M Yang ZY Wu HCamp Li HY 2013 New insights into magnetic enhancement mechanismin chinese paleosols Palaeogeogr Palaeoclimatol Palaeoecol 369(1)493ndash500

Zhao GY Han Y Liu XM Chang L Lu B Chen Q Guo XL ampYan JH 2016 Can the magnetic susceptibility record of Chinese RedClay sequence be used for palaeomonsoon reconstructions Geophy JInt 204 1421ndash1429

Zhou LP Oldfield F Wintle AG Robinson SG amp Wang JT 1990Partly pedogenic origin of magnetic variations in Chinese loess Nature346 737ndash739

Dow





Previous study (Orgeira et al 2011) observed that relatively dryclimates (lt700 mm yminus1) provide the best conditions for pedogenicmaghemite formation the rate of pedogenic maghemite formationis proportional to rainfall and do appear to show the expected in-crease in the values of chosen magnetic parameters (susceptibilityits frequency dependence anhysteretic remanence etc) Howeveran increase of the mean annual rainfall is not accompanied by a pro-portional increase of pedogenic maghemite formation Soils becomepermanently humid when rainfall increases lacking the alternationof wet and dry conditions required for pedogenic maghemite for-mation when mean annual precipitation (MAP) increases above acertain critical value (gt700ndash1000 mm yminus1) pedogenic maghemiteis destroyed by reductive dissolution (gleization) these magneticsignals begin to decrease with further increasing rainfall They havealso pointed out that the alteration and formation of different ironoxides (hematite goethite magnetite maghemite) in soil depend onnot simply the amount of rainfall but also the retention of moistureand pedogenic gleization

Goethitehematite ratio has been proposed earlier by Schwert-mann and Kampf (1985) as a temperature proxy for soil of thetemperate zone and pointed to be as palaeoprecipitation indicatorin soil in recent rock magnetic investigations (Long et al 2011Liu et al 2013 Hyland et al 2015) The frequency dependenceof susceptibility to hard isothermal remanent magnetization ratio(χ fdHIRM) of modern soils from Shanxi are found to have a goodcorrelation with the modern rainfall where χ fd may be a proxy forSP magnetitemaghemite and HIRM for hematite (Liu et al 2013)

This paper discusses the occurrence of magnetic minerals inloessic soils along a transect over the Chinese Loess Plateau therelation of these minerals to climatic conditions and the magneticmethods used for their identification To identify the compositionalvariability of all the oxidized iron minerals and to estimate theiramounts as accurately as possible different methods including di-rect microscopic observation of thin section scanning electron mi-croscope and transmission electron microscope of the magneticextracts (Chen et al 2005 Yang et al 2013) have been applied anda variety of magnetic properties and their changes under differenttemperatures and frequencies different methods have been tried(Zhou et al 1990 Geiss amp Zanner 2006 Michal et al 2010 Huet al 2015) By comparison low-temperature measurements canlead to accurate and thermally non-destructive ways for oxidizediron minerals identification and grain-size estimates even below20 nm

In order to explain why high annual rainfall can lead to the loss ofmagnetization and to further clarify the transformation of oxidizediron minerals in waterlogged soil environments we collect well-developed palaeosol S5 samples from three locations (XF LY andBJ) from more dry in the north to more humid environments in thesouth on the Chinese Loess Plateau Oxidized iron minerals andgrain-size of these samples are identified and estimated by usinglow-temperature magnetic measurements supplemented by X-raydiffraction Through these thermally non-destructive ways we getaccurate estimations of relative amounts of different iron oxides inS5 samples formed under different humid conditions and discussthe relation of these minerals to climatic environments

2 S A M P L E S A N D M E A S U R E M E N T S

21 Sample collection

The samples for study were collected from the S5 palaeosols atthe Xifeng (XF) Linyou (LY) and Baoji (BJ) sections on ChineseLoess Plateau (Guo et al 2015) The present climate of the areais semi-arid to mildly humid with MAP of 550 680 and 720 mm(up to 1100 mm in Qinling Mountain area) from XF to LY to BJrespectively Sixty per cent of precipitation occurs from July toSeptember The mean annual temperatures (MAT) are 87 91 and12 C respectively Precipitation which carried by the East Asiansummer monsoon decreases from south (BJ) to north (XF)

22 Magnetic extraction

Magnetic extraction of the three natural samples is carried out by us-ing the magnetic extraction apparatus aimed at separating stronglyand weakly magnetic particles for characterization The magneticextraction of the bulk sample yields two subsamples (1) mineralparticles collected with an applied magnetic field defined as themagnetic extracts (MAG) and (2) mineral particles that are non-separated fraction defined as the residual (RES)

The magnetic extraction procedure is same in design to the pumpmethod described by Strehlau et al (2014) similar to the systemdescribed in Reynolds et al (2001) A certain original S5 palaeosolsample was dispersed in 30 mL Milli-Q water using 1 g of dispersantsodium hexametaphosphate centrifuged at 5000 rpm for 3 minusing an Eppendorf 5804 centrifuge then added to the reservoirwith 200 mL Milli-Q water The suspension was circulated by aMasterflex LS (Cole Parmer) peristaltic pump at sim200 mL minminus1

and passed through Tygon LS flexible plastic tubing (Saint Gobain)oriented vertically with a joint that contained an Nd magnet coveredwith a plastic sleeve to collect the magnetic material After 90 minthe magnetic extracts were washed from the sleeve with Milli-Qwater and collected for analysis The remaining suspension in thereservoir was collected as the residues and then dried for analysis

23 Measurements

Anhysteretic remanent magnetization (ARM) was measured in Al-ternating Field Demagnetizer with pARM device the peak AF fieldused was 100 mT and the biasing field is 005 mT ARM was thennormalized by the bias field to obtain ARM susceptibility (χARM)Hysteresis loops were measured using a Princeton MeasurementsVibrating Sample Magnetometer with the maximum field of 15T We used the loopsrsquo hysteresis properties to determine saturationmagnetization (Ms) saturation remanent magnetization (Mrs) co-ercivity (Bc) remanent coercivity (Bcr) relative antiferromagneticcontent (HIRM) and relative ferrimagnetic content (S300)

A quantum design Magnetic Properties Measurement System(MPMS2) cryogenic susceptometer was used to measure coolingand warming curves of room temperature saturation isothermal re-manence magnetization (RTSIRM) field cooling (FC) and zerofield cooling (ZFC) remanence curves and also for thermal rema-nent magnetization (TRM) SIRM produced in a 25 T field at 300 Kwas measured continuously during ZFC to 20 K at 5 K steps andback to 300 K ZFC and FC is based on performing two consecutivemagnetization measurements In ZFC the sample is first cooleddown in the absence of a magnetic field and then measured in a 25T field at increasing temperature FC is performed in a 25 T fieldat decreasing temperature

Dow




2130 X Guo et al





3 R E S U LT S
















Dow












Dow




2132 X Guo et al








Dow








Dow




2134 X Guo et al





Dow













Dow




2136 X Guo et al
































Dow












































Dow




2130 X Guo et al





3 R E S U LT S
















Dow












Dow




2132 X Guo et al








Dow








Dow




2134 X Guo et al





Dow













Dow




2136 X Guo et al
































Dow












































Dow












Dow




2132 X Guo et al








Dow








Dow




2134 X Guo et al





Dow













Dow




2136 X Guo et al
































Dow












































Dow




2132 X Guo et al








Dow








Dow




2134 X Guo et al





Dow













Dow




2136 X Guo et al
































Dow












































Dow








Dow




2134 X Guo et al





Dow













Dow




2136 X Guo et al
































Dow












































Dow




2134 X Guo et al





Dow













Dow




2136 X Guo et al
































Dow












































Dow













Dow




2136 X Guo et al
































Dow












































Dow




2136 X Guo et al
































Dow












































Dow












































Dow




Date post:	03-Aug-2020
Category:	Documents
Upload:	others
View:	1 times
Download:	0 times

Why magnetite is not the only indicator of past rainfall in the … · Geophys. J. Int. (2018) 213,...

Documents