1
i*f*bd
1
i
THE MECHANISM
TECHNICALNOTE4089
OF THERMAL -GWDIENT MASS TIULNSFER IN THE
SODIUM HYDROXIDE -NICKEL SYSTEM
By CharlesE.May
LewisFlightPropulsionLaboratoryCleveland,Ohio
Washington
September1957
https://ntrs.nasa.gov/search.jsp?R=19930084800 2020-07-02T23:12:40+00:00Z
TECHLIBRARYKAFB,NM
TIK!HNICALNOTE40S9
TEEMECHANISMOF TEERMAL-GRADIENTMASSTRANSFERINTHE
SODIUMEYDROXIDE- NICKELSYSTEM
By ChsrlesE.May
SUMMARY
“Thermal-gradientmasstransfer”wasinvestigatedinthemoltenso-diumhydroxide- nickelsystem.Possiblemechanis~(physical,electro-chemical,andchemical)arediscussedintermsofexperimentalandtheo-retical.evidence.Experimentaldetailsareincludedinappendixes.Thefollowingmechanismisfoundtobe mostprobable:
ColdzoneNa.#i02+ 2Na~ 2Na20+ Ni
Kineticequationssretheoreticallyderivedconsideringseparatelythefollowingthreefactorsas controllingtheover-allrateofmasstransfer:thechemicalreaction,thediffusionprocess,andtheforcedcirculation(ifpresent). Iftheabve mechanismisthetrueone,thechemicalreac-tionistherate-determiningstepindynamicsystems,andthediffusionprocessistherate-controllingstepinstaticsystems.Thebeneficialanddetrimentaleffectsofadditivesareinterpretedintermsof theprob-ablemechanism.
Evidenceisalsopresentedto showthatthermal-gradientmasstrans-ferinthemoltensodiumhydroxide- coppersystemoccursby a similarmechanism.
INTRODUCTION
Sodiumhydroxideisinmanywaysan attractivechoiceasa high-temperatureheat-transferfluid.IJtiortunatel,y,onlya fewmaterialsaresuitableforcontainingit inthemoltenconditionatthetemperaturesdesired,15M0 to 17000F. Amongthesesrenickel(refs.1 to 3),copper,silver,gold(ref.4),theLewislaboratay.“thermal-gradientmass.
-
E@ somenickel-basealloysrecentlyproducedatEventhesematerialsexhibitcorrosiontermedtransfer”,althoughtheydo notshowintergranular
2 NACATN4089
attackorrapidsolution..
Thermal-gradientmasstransferisthephenome-nonbywhichthemetalisremoved(eitherchemicallyorphysically)fromthehotterregionsofa systemanddepositedincolderregions.Withthe rpreviouslylistedmetalsthedepositisintheformof needle-likecrystals,thesizeofwhichdependsupontheexperimentalconditions.
Inordertofindwaysof inhibitingthistypeof corrosion,an in-vestigationwasconductedconcerningthepossiblemechanismsformasstransfer.Althoughmasstransferalsooccursinliquidmetalsystems(ref.5;e.g.,moltensodiumincopper),themechanisminsuchcasesisundoubtedlypurelyphysicalinnature,beingdueto solutionotthecon-tainermaterialinthemoltenmetal.In systemsusingsodiumhydroxide,transferisprobablychemicalinnature:andtheprocessisbelievedtobe similerforallthecontainermaterialspreviouslymentioned.Thepurposeofthisreportisto establishthemostprobablemechanismformasstransferinthissystem.Thisreportdiscussesmethodsofmeasure-uk?nt,choiceofmechanism,kinetics,effectofadditives,andalternatemechanisms.Manyoftheproofsrequiretheuseofpreviouslyunpublished Ae~erimentaldata,thedetailsofwhichsxeincludedintheappendixes.
*MET30~OFMEASUREMENT
At theIewislaboratorytwomethodshavebeenusedto studythephenomenonofmasstransferjtheyarethestaticcapsuletestandthedynamictoroidtest.Previousreports(refs.1 to 3)describetheseindetail.Inthestatictesta verticaltemperaturegradientexistsinnickelcapsules(crucibles)withthehotzoneatthebottom.Thistypeofgradientwasintendedtoproducetherm&1convection.Thenickeldis-solvesfromthebottomofthecapsuleandprecipitatesasa narrowringattheliquidlevelofthesodiumhydroxide.Fromradiographsof suchcapsulestherelativeamountsoftransfercanbe estimatedqualitatively.Thelossinweightofa nickelspecimenplacedinthehotzone(bottom)isfoundtobe proportionaltotheamountoftransferinthecoldzone(ref.1).
IDthecirculatingliquidrunsmadeintoroids,radiographsofthecoldsectionsalsosup@ya convenientmethodformeasuringmasstransfer.Sincethetransferrednickelis spreadoveran etiendedarea,theaveragereductionininternaldiameterofthecoldsectionof thetoroidgivesaquantitativemeasureoftransfer.
mostR 1sJ
Inthestudyofthegeneral.casecanbecou.ibinetoproduce
CEOICEOFMECHAN18M
mechanismofthermal.-~adientmasstransfer,a.
considered.InthecoldzoneallofthereactantsnickelandalloftheotherproductsPi’s:
NACATN4089 3
where nj ~d ~ srethenumberofmolesof Rj and Pi)respectively.Thereversereactionmustoccuratthehotzonetomaintaintheprocess.Therequirementsforsucha cyclicreactiontooccurarethatateachtemperaturereactionoccurstowardtheattainmentof equilibriumandthattheequilibriumconstantvariessignificantlywithtemperature.!Qms,theconcentrationsareslightlydisplacedinonedirectionfromequilibriumat thehighertemperaturesandintheotherdirectionatthelowertem-peratures.Obviously,at some“meantemperature”theconcentrationsofthematerialsinvolvedarein chemicalequilibrium.
As a consequence,threemajorprocessesareinvolvedinanyproposedmechanism:(1)thechemicalreactionsthatoccuratthehotandcoldzones,(2)themigration(diffusionor circulation)of thespeciesfromthehotto thecoldzoneandviceversa,and(3)theinitialadjustmentof concentrationsto “meantemperature”equilibrium,whichsometimesre-quiresadditionalchemicalreactions.Theevidencefororagainsta par-ticularmechanismdependsprimarilyuponthenatureofthefirstprocess,thezonereactions.
PossibleReactions
Thefirststepindeterminingthemechanismoftransferwasto enu-merateallthepossiblereactionsthatcouldbeenvisioned(table1). Asimilarlistissuggestedinreference6. Thesemechanismscanbe dividedintothreeprimarygroups:physical,electrochemical,andchemical.Thechemicalmechanismshavebeensubdividedas towhethertheoxidizing-reducingpairintheequation
ColdzoneNa2Ni02+ Reductant~= Oxidant+Ni
arisesfroman impurity,thedisproportions’tionof nickel,or thesodiumhydroxide. Ifthesodiumhydroxidesuppliestheredoxpair,theelementactuallyinvolvedintheoxidationcouldbe eitheroxygen,hydrogen,orsodium.Inthesubsequentdiscussion,theoxidizedformofnickelisgenerallyrepresentedas sodiumnickelite(N~Ni02),sincethiscompoundintheformof olive~greenneedle-likecrystalshasrecent~vteenisolatedat theLewislaboratory(appendixA) aswellasat otherlaboratories(refs.6 to 9). Itlikewiseappe=sthatthisistheonlyoxidizedform
. of nickelstableinmoltensodiumhydroxide(appendixA).
.
NAC!ATN40S9
.
Therequirementsfor
MeansofSelection
theactualmechanismarethefollowing: s
(1)Thechemicalspeciesiuvolvedinthereactionsshouldbe feasible.Their‘presenceshoUdbe supportedeitherexperimentallyortheoretically.
(2)Themechanismshouldexplaintheapparentdifferenceinthekineticsbetweenstaticanddynamictests.Sinceinstatictestshytio- tgensimultaneouslydecreasestherateoftransferandthenickelitecon- kcentration(appendixB),itappearsthattherateoftransferdependsu~n nickeliteconcentration.However,indynamictoroidtests,eventhoughthenickeliteconcentrationincreaseswithtime(appendixC),therateoftransferisindependentoftime. ThisindicatesthattherateoftransferIsindependentofnickeliteconcentration.
(3)Themechanismshouldexplaintheeffectofadditivesupontherateoftransfer.Specialattentionshouldbe devotedtothecompounds mof chromium(appentiesD andE),sincekm!zhmetallicchrormtumandchro-miumsesquioxide(Cr203]inhXbittransferinstatictests,whileonlychromiuminhtbitsitindynamictests. .
(4) Thereshouldbe noexperimentaldataindirectcontradictiontothemechanism.However,allmechanismsexceptthelast(reaction(7))arecontradictorytoon&ormoree~erimentfifacts.Th&e discre@2&iesare(th3cussedlater.
SodiumMechanism
Thefou.owhlg(table1,reaction
mechanismis(7)]:
Na2Ni02+ 2Na
consideredtobe themostprobableone
ColdZO&2Na20+ Ni
k~—
ariseexclusivelyfromdissociationofThesodiumoxideispresmedtosodiumhydroxide,whichgenerateswateratthesfietime.Theoretically,sodiumoxideshouldexistinsodiumhydroxide;thermodynamiccalculations(appendixF) showthatitsequilibriumconcentrationcanbe ashighas2.7molepercent.Itspresencehasalsobeensupportedexperimentally(ref.10). Sodiumnickelite(appendixA} andmetallicsodium(ref.7)havedefinitelybeenfoundexperimentallyinthesodiumhydroxide-nickelsystem.Metallicsodiumhaslikewisebeendetecttiinthesotiumhydroxide- coppersystem(appendixG). Sodium,aswellassodiumnick-elite,isformedby thehot-zonereaction.Since,however,theconcen-trationof sodiumnickelitedoesnotreacha constantvalue[appendixC),undoubtedlymostof itisformedby thereaction
.
.
W. + 2NaOH~ ~ + N~Ni02
.
NACATN4089
Inorderto explaintheapparentdifferencesbetweentheresultsinstaticanddynamictests,onemustfirstderivethekineticequationswhichareapplicableto themasstransferprocess.Anyoneofthefollow-ingthreefactorscanbe consideredto controltheover-allrateoftrans-fer: therateofthechemicalreaction,therateof’thediffusionproc-ess,andtherateofthecticulation(ifpresent).nextsectionsaredevelopedintermsof thegeneralforthecoldzone
ChemicalReaction
Whenthechemicalreactionisrelativel.yslow,
Theequationsinthereactionaswritten
themoltenmaterialisessentiallyhomogeneouswithrespectto thecon&entrationsof allthechemicalspecies.Therateof transferrt isobtainedby consideringthenetrateofreactionina unitareaofthecoldzone(denotedby thesubscriptc):
[1%-k 9‘t = ‘fjcy ‘j r,c~~il a (1)
where ~ istherateconstantoftheforwardreaction,~ istherateconstantofthereversereaction,a istheactivityofthenickelmetal,and [] indicatestheconcentrationofthespecies.Rearranginggives
Sinceat somevolvedsreineachotherby
‘meantemperatureti(mentionedpreviously)thespeciesin-chemicalequilibrium,theirconcentrationsererelatedtotheequilibriumconstant~ ofthat“meantemperature”:
~ [Pj‘i a%= (3)
~ [R-jl‘j
6 NACATN4089
Theequilibriumconstant~ forthecoldzoneisgivenby theequation.
% = %@+,c
Substitutingeqwxbions(3}and(4)intoequation(2)yields
‘t =kf,c(yp,]n’)(1-~) (5),-
l?romtemperature-coefficientconsiderations,
% ,’@RTca/RTe—=Kc
(6}
awhereAH istheenthalpyofthereactionaswrittenforthecoldzone,R isthegasconstant,TC isthetemperateofthecoldzone,andTe istheequilibriumtemperature.Substitutingequation(6]intoequa- “tion(5)gives
(7)
Finally,by expandingtheexponentialandassumingTe - Tc (M ineq.(8)) tobe smll, the rate of I-S tra*sferiSobtainedfora chemical-reaction-controlledprocess:
(’m)‘j AmT‘t = -kf ilRJZ (8)
or similarly,
DiffusionProcess.
Ifthechemicalreactionsconcernedarerelativelyfast,diffusionoftheproductsandreactantsmaybecomethegoverningfactorfortherateprocess.Theresultingrateequationcanbe mosteasilyderivedfor “
NACA!IXi4089 7
.
twonickelplates,eachof unitarea,at differenttemperatures,andplacedparalleltoeachotherat a distance2 apartinmoltensodium
+ hydroxide.Thetemperaturedifferenceandthespeedofthechemicalre-actiongiveriseto concentrationdifferences~i and mj forthespeciessurroundingthetwoplates.Theseplatesaredesignatedby thesubscriptsh forhotand c forcold.
Therateoftransfer,beingequalto therateofdiffusionofeachspecies,isgivenby
(lo)
or simply
.
+‘ikimi nk “
‘t ‘~= (U)“
wherethe ki’sand kj’sarediffusionconstantsforthei’thandj’thspecies.Multiplicationofequations(10)and (U) by S/ki[Pi]andS/kj[Rj]jrespectively>andthensummationoverall i’s and j’sresultsin
‘z*)=afTw-R““ ~J i J
or
rt=(’x!%+qla)l($+rph-““Thenumeratorcanbe evaluatedfromequilibriumconsiderationsasfollaws.Sincechemicalequilibriumexistsatthehotandcoldzones,therespec-tiveequilibriumconstantsKh and ~ are
.[ 1~‘i,h%
% = ~~j,h]‘J
(14)
8 NACATN4089
Fromtemperature-coef’ficientconsiderateions,
~
%
where Th isthetemperature(15),and(16)gives
AH/RTc-@%e
Theconcentrationsatthehottionsatthecoldzoneare
AH/RTc-AE/RTh=e
of the hotzone. Combiningequations(14),
(16}
zoneexpressedintermsof’theconcentra-
~IM ~ and ~ and assum@jthatthe ARj’s and Mi’s aresmallwithrespectto thecorrespondingRJ’s and Pi’s yields
Expandingthee~onential.with AT,Th - Tc~assumedtobe
(19)
smallgives
IUC!ATN4089
.
. y=z~+x~i J
Combiningequations(13)and(20),oneobtains
9
(20)
(21}
where ka istheaverage—diffusion-rateconstant.
& Circulation
Ina systemwithforceflowtherateof transfermaybe controlled.: by therateof circulationkcti (incps). In a staticsystemwith
thermalconvectiona valuefor kcti mightbe difficultto assign.-3Nevertheless,therateoftransferina circulation-controlledsystemcanbe expressedas
‘t = ckcti%~i = ~ctinJ%(22)
where C expressesthedependenceof r+ on thevolumeofthemelt.Thefollowingequationcanthenbe obtai;ed,ina similarmannerforequation(13J,
.t=.cti&’&xE)/@+@)
J
to that
(23)
Sincechemicalequilibriumexistsatthehigherandlowertemeratures,equation(20~isagainvalid. fUponsubstitutingequation(20 intoequation(23),
‘t = -Ck#4T/RT& h
i
-+mLil (24)
10 NACATN4089
It isreadilynotedthatinallthreecases(eqs.(9],(21),and(24)) therateof transfer(tothefirstapproximation)isproportionaltothechangeinenthalpyofthereactionaswrittenforthecoldzone.Inasmuchastherate rt ispositiveaswritten,theenthalpymustbenegative.Tbthefirstapproximationthetransferispro~rtionaltothedifferenceintem~rature;thispointhasbeensupportedexperimentally(ref.l).
ApplicationofEquations
Forthesodiummechanismthethreekineticequationswouldbe:
-k+wsrrt(chemical)=
()‘T2[N~2012
(25)
4
-k&ATrt(diffusion)=
(
ka k J%(26) “
RT2j~ + m + k3~a#i0211[a21 )
-kc#!16ATrt(circulation]=
T‘T [Na~O]+ * + [NGi021
(27)
In a chemicallycontrolledprocesstherateof transferisconstantaslongaatheequilibriumconcentrationof sodiumoxideisnotdisturbedsi~ificantly-by thezonereaction
ColdzoryeN~i02 -I-ma <_..
HotzoneZNa20+-Ni
Iftheequilibriumconstantforthezonereactionhasa valuewhichkeepsthesodiumconcentrationlowwithrespecttothatof sodiumoxide,thelatterisnotappreciablyaffected.
Ina diffusion-controlledprocess,theeffectofthesodiumconcen-trateionmightbe ignoredif k2 isrelativelylarge,aswouldbe pre-dictedfromtheunchargednatureandthesmallsizeofan atomofmetallicsodium.Thus,therateoftransfershoulddependuponthenickelitecon- -centration,sinceitisofthesameorderofmagnitudeasthesodiumoxideconcentrateion. ([Na2Ni~] = 0.2molepercentat 24h, c~ctited .fromdataintableVTII;[Na20]= 0,02to 2.7molepercent,appendtiF-)
J!UCATN4089
.Therate
siderednext.
u.
of cticulation-controlledtransfer(eq.(27))canbe con-Sincethesodiumconcentrationislessthanthenickelite
q concentration,theterm 1/maJ affectstherateoftransfermorethantheterm l/[Na#i02].Therefore,since 1/[Na] isproportionalto thenickeliteconcentration,therateof masstransfermustdecreasesathenickeliteconcentrationincreases.
Onethereforeconcludesthattherateof transferindynamicsystemsischemicallycontrolled,sinceit is independentof sodiumnickeliteconcentration.However,transferinstaticsystemsmustbe diffusioncontrolled,sinceitsrateincreaseswithnickeliteconcentration(appen-dixB}. Instaticteststherateof transfershouldthenincreasewithtime. Theincreasedrateof specimenweightlosswithtimeincapsuletests(tableVIII,column9) i:an indica~ionoftransfer.
thisincreasedrateof
..! EFFECTOFADDITIVE3.:~
. Ina chemicallycontrolledmechanism(dynamicsystem)additivesmsy: affecttherateoftransferinthefollowingways:~
(1)By changingtheconcentrationofthesodiumoxide
(2)By supplyingan auxiliarytransfermechanism
{3} By changingtheactivityofthenickelmetal
Thefirsttwomethodsof sffectingtherateoftransferarealsovalidina diffusion-controlledmechanism(staticsystem).However,thethirdmethod,changingtheactivityof&echangingtheconcentrationof sodiumtransferas statedpreviously.
sodium
nickei,sh&ldhaven~effect;insteadnickeliteshouldaffectthersbeof
oxide
Onewouldpredictthattheadditionof sodiumoxideandsodiummetal(whichwouldeffectivelyincreasethesodiumoxideconcentrateion)toeithera staticor dynamicsystemshouldincreasetherateofmasstrans-f er. Thishasbeenverified-experimentally(ref.3). Addingcalcium,sodiumhydride,lithiumhydride;andcalcium-nitrid&addingmetallicsodium)shouldlikewiseincreasetheForexample,thereactiontithcalciumis
Ca+ 2NaOIi+Ca(OH)2+ 2Na
(beingequivalentrateof transfer.
to
12
ThisalsoisdemonstratedOftheSOdiUmhydrideandthusgeneratedto diffusenobeneficialeffect.
inreferences2lithiumhydriderapidlythrough
In casesinwhichmetallicadditives
MICATN4089
and3. Thefastdecomposition “shouldallowallthehydrogenthenickelandthustoproduce a
reactwithsodiumhvdroxidetoproducesotiumoxide,an increasedrateoftransfershouldag;inbeexpected.Thispointisperhapsillustratedintheincreasedrateoftransfercaused%y additionofmanganese,iron,titanium,vanadium,andmolybdenum:
MO + 6NaOH~ Na2Mo04+ 2Na@
Water
Additionsofwaterin smallamountsshouldsuppressthesodiumoxideconcentrateionandinhibittransfer.Instatictestswheregassaturatedwithwatervaporwaspassedoverthesodiumhydroxideinthenickelcruciblebeingtested,transferwasdecreased(ref.3).
Additionsofcertainoxidesmaybe equivalentto addingwater,as isevidentfromtheequation
COO+ 2NaOH~ Na2b02+ ~0
Thismaybe thecasewheninhibitionoftransferresultsfromtheuseofthefollowing:cobaltoxide,cupricoxide,cuprousoxide,stannousoxide,manganesemonoxide,andzirconiumdioxide(ref.11]. Theeffectivenessof stannousoxide,manganesemonoxide,andcuprousoxidemaybe p=tiallyduetotheirreducingaction.Inorderforbeneficialeffectsduetowatertobe observed,theadditivemusteitherbe addedaf%erpurgingorreactonlyattemperatureshigherthanthepurgingtemperature.
It ispossiblethattoomuchwatermightresultinanauxiliex’yreactionsuchas
ColdzoneNa#Ti02+ 2Na+ 2~0 ~~ Hi + 4NaOH
Thismayaccountforthecaseofunpurgedsodiumhydroxideinwhichtherateoftransferwasincreased(ref.2). Thisreactionmayalsoaccountforthefactthatauybeneficialeffectsattributableto thepresenceofwaterarealwaysslight.
NAC!ATN4089
MetallicChromium
b Metallicchromiumisan inhibitorofmass
13
transfer (ref.2). Themethodby whichchromiumworksis notby decreesingthenickel.iteconcen-tration(appendixD).
Experimentsshowthatmetallicchromiumreactsrapidlywithmoltensodiumhychmxidetoformsodiumchromite(Cr+3),whichthenreactsumreslowly(for20hr)toformsodiumhypochromate(C#) (appendixE). Thischromite-Qq@chromatereactionis inequilibriumwiththehydrogenpres-sure(appendixE). Sincethenickel-nickelitereactionisnotinequi-libriumtithhydrogen(appendixC),thefollowingreactioncannotbe at“mean-temperature’*equilibrium:
Ni+ N~Cr05c-m Na3&03+Eotzone.
Thechromiumisthereforenoteffectiveby meansnism. In somewaythechromiteandhypochromate
Na2Ni02
ofthisauxiliarymecha-forma complexinan
approximateratioof 1 to 1 at MO@ F and1 to 1.1at 17C@ F. There-fore,itisreasonablethateitherofthespeciesmightpsrtiallycomplexwithsodiumoxideandthusinhibitmasstransfer(bothinstaticanddynamictests).
ThequestionwhichnowarisesiswhythechromiteformedtrommetsJ.-licchromiumreactsdifferentlyfromthechromiteformedfromchromiumsesquioxide(appendixD). Eventhough both msy formsodiumchromite,thecoordinationoftheoxygenatomsaboutthechromiummightbe differentj
thisis quitecommonamongwater-solublesaltsof chromium.H themecha-nismby whichtransferis inhibitedusingmetallicchromiuiuvolvescoordination(complexformation),chromiumsesquioxideobviouslyneednotinhibittransferinthesamemanneraemetsllicchromium.
ReducingAgents
Sincechromiumsesquioxidereducesthesodiumnickeliteconcentra-tion(appendixD)presumedlyby thegenerationofl@rogen,it stiuldinhibitmasstransferwhenthetransferisa diffusion-controlledprocess.Indeedtransferis inhibitedinstatictestsby chromiwnsesquioxide(appendixD)butnotindynamictestsrecentlyperformedattheLewislaboratory.Anotherreducingagent,calciumhydridealsoinhibitstrans-
~ferin statictestsbutnotindynamictests(ref.3 . Hence,directly :. appliedhydrogenpressurewhichinhibitstramfer in statictestsshould
notnecessarilybe effectiveindynamictests.Publishedworkonthiseffectis inconclusive(ref.11). Eydrogencouldinhibittransferin
. dynamictestsif itspoisoningactionweresufficienttoreducethe*
14 NACATN4089
actiyityofthenickel.(Cdcim@tiide i.creasestherateoftransfer -indynamicsystems,becauseitsdecom~sitionproduces“detrimental”calciummetalaswellas ‘%eneficial”hydrogen.) ●
OxidizingAgents
Oxidizingagentsshouldhavetheoppositeeffeetofreducingagentsupontherateoftransfer,sincetheyshouldincreasethesodiumnickeliteconcentration.Thisistheeffectfoundin staticsystemsformaterialssuchas sodiumperoxide,ferricoxide,potassiumchlorate,potassiumchromate,sodiumchromate,calciumchromate,cobaltousicoxide,manganesedioxMe,sodium’bro~te,andleadsesquioxide(refs.2 and11). Thiseffectisnotexpectedindynamictests,butnoneoftheseadditiveshaveeverbeenstudiedundersuchconditions.
InsolubleCompounds r
Appsxentinhibitionoftransferresultedinstatictestsfromaddi- .ttonsofaluminum,aluminumoxide,scdiumaluminate,aluminumfluoride,magnesiumoxide,calciumchloride,andstrontiumperoxide(refs.2,3,andn). Thisk beenattributedtotheformationofprotective~nsolu-blelayersoverthespecimens,sodiumaluminateinthecompounds,andinsolubleoxidesfromtheothers.
iaseofaluminum
InertMaterials
Sincesilver,gold,andcopperdonotreactrapidlywithsodiumhydroxide,noeffectisproducedbyusingthesemetalsasadditives(ref.11)● Othermaterials,suchas saltsthatareinertto sodiumhydroxide(sodiumchloride,sodiumfluoride,andsodiumiodide]andceramicmate-rialsstableat 1500°and17M1°F (tantalumnitrMeandcolumbiumnitride),shouldnotanddonothaveanyeffectuponmasstransfer(refs.2 andn).
CONSIDERATIONOFALTERNATIVEl@X~
Physicall&chanism
In consideringthephysicalmechanism(tkble1,reaction(1))onemayadmitthatnickelis soluble
ColdzoneNi”(dissolved)= Ni”(metal)
NACATN4089
.inumltensodiumhydroxideandthatas soonas saturationisreachedtransfercouldoccur..
Althoughmasstrausfermightbe attrfbutshleto thismechanism,ex-perimentswithhydrogenatmospheresshowitscontributiontobe quitesmall,sincethepresenceofhydrogenreducestherateoftransferby alargefactor{appendixB andref.3). In nowayisit conceivablethathydrogenreducesthevolubilityof nickelinnmltensodiumhydroxide.Alsxgevarietyof othermaterialsinoftransfer(appendixesD andE andplausiblethatsuchmaterialswouldproblemofphysical.volubility.
smallquantitiesdoaffecttheraterefs=2 and3). Itdoesnotseemaffecttransferif itwereJusta
. chemical
.
ElectricalMechanism
secondpossiblemechanismfortransferhalf-cellreactions(tableT,reaction
involvestheelectro-(2))
Na@i02+ 2Na++ 2e- Coldzone— 2Na20+ Ni~
Thismechanismrequirestheexistenceof an electromotivenickelatthehotzoneandnickelatthecoldzone. Suchbeenobservede~rimentdly (appendixH andref.12);it
forcebetweena potentialhasiSprobably
attributabletoreaction(2). Moreover,thispotentialis sufficienttosustaina currentofmeasurablemagnitudefor6 hours(appendixH). Theelectrochemicalreactionrequiresthat,sincethesodiumnickelitecon-centrationbuildsupwithtime(appendixC),theelectromotiveforceaswellastheactualrateof transfershoulddecreasewithtime. Thisiscontraryto fact. Also,inthepresenceofhydrogen,themasstransfershouldoccurfromthecoldzoneto thehoczoneinaccordancewiththereversalofpolsrityobsemedwitha hydrogenatmosphere(appendixH).
Theconclusionsarethattheelectrochemical.processmaybe occur-ring,butthatitscontributionto themasstransferactuallyobservedissmallandas timeprogressesbecomeslessandless. Inthepresenceof a hydrogenatmospherethisprocesscannotoccuratall. An additionalproofisthefactthatcrystalsof nickelplateoutona ceramicmaterialIncoldregions(ref.6};inthiscasean electrochemicalprocessisimpossible.
Thelatterconclusionsareof coursevalideventhoughthehaM-cell. reactionsmightinvolvenickelwitha valenceof+3:
Coldzoe. NaNi02+ 3Na+.+3e-~_. 2Na@ + Ni
16 NACATN4089
.
Uapurity Mechanism
Of theimpuritiespresentin sodiumhydroxide(table11)andnickel -(table111),onlyiron,cobalt,andmanganesecouldcontributetotheimpuritymechanism,sincetheelementinvolvedmustexistintwooxidizedvalencestates:
ColdzoneNa.#i02+ 2Na#e02.— 2Na3Fe03+ NiHotzone
ColdzoneN@Ji02+ ~2C002 ~~ 2N@003 -I-Ni
Coldzone2Na#i02+ Na#n02 ~ Na#4104+ 2Na20+ 2Ni
HCltzone
Theprobabilitythata sulfate,a phosphste,silver,orcopperwouldexistintwoionicvalencesisexceedinglysmall.Ik?causeofthesmall.
.
concentrationsofalltheimpurities,thistypemechanism,slthoughpossible,isstillnotprobable.
Theconcentrationof impuritiesarisingfromthenickelincreasesinthesodiumhydroxidewithtime(ref.6],andthustherateoftransfershouldincrease.Therateoftransferdoesnotincreasewithtimeindynamictests.
Finally,alltheseiquritiesweredeliberatelytided(insomeforui]to thesodiumhydroxide,andalthoughsomeapparent~increasetransfer(perhapsfivefoldby qualitativeestimate)(tableIV),thisin-creaseisnotproportionaltothethousandfoldincreaseoftheimpurityconcentration.
DisproportionationMechanism
Thenextmechanismto consideristhedisproportionationreactionof nickel(table1,reaction(4))
3Na2Ni0Coldzone- 2Na3Ni03+ Ni2 Hotzone
Againreferenceismadeto thefactthatthesodiumnickeliteconcentra- -tionincreaseswithtime(appendixC). Sincethenickeliteandnickelatemustbe in “meantemperature”equilibrium,theconcentrationofthe .
NACATN4089
.nickelatemustalsoincrease.Therefore,on thenickeliteconcentrationregardless
. controlled.Thesameistrueevenifthe
17
thetransferrateshoulddependofhowtherateprocessischemicalequationis
Coldzone3Na#i02~ 2NaNi02+ 2Na20+ Ni
HotZOne
Sincethetransferrateindynamicsystemsisexperimentallyindependentof nickeliteconcentration,theactualmechanismcannotinvolvedispro-portionationof nickel.
Furthermore,thehighervalencesof nickelshouldnotexistinthepresenceofa hydrogenpressure(ref.6) suchashasbeenreportedto ex-istabovethenickel- sodiumhydroxidesystemat8000C (4mm Hg,ref.7).
w OxygenMechanism
Thepossibilitythatoxygenisthe“redOx”materialaccordingto the.~ equation
—Coldzone
Na#Vi02~ Na20 -I- 1/202 +NiHotzone
isnil,sincethe4 millimetersofurcurypressureofhydrogen(ref.7)eliminatesthepossibilitythatoxygenispresentat all.
. HydrogenMechanism
Themechanismthatfora longtimeseemedthemostlogicalinvolveshydrogenas theredoxmateridjthereactionis (table1,reaction(6))
ColdZOlleNa2Ni02+ ~ ~~ 2NaOH+Ni
Thereactionbetweennickelandsodiumhydroxideisknownto occur,sincebothhydrogenandsodiumnickelitearefoundas Produdsj the problemiswhetherthisreactioncango inreverse.Thisseemedpossible,sincenickelOXME (assumedtobe theoxidizedformof nickel)iseasilyreducedby hydrogen.Experimentsunderisothermalconditionswitha constant
- imposedhydrogenpressure(appendixC) showedthattherateofformationof sodiumnickeliteis constantevenupto timesof 6~ hours.Thus,equilibriumisnotreached,thereversereactionneverOCCmSJandt~s
. mechanismcouldneveraccountfortransfer.Evenwithnoaddedhydrogen,equilibriumisnotreached,inasmuchastheconcentrationof scdium
18 NACATN4089 ....
.
ni.ckelitedoesnotreacha constantvalue(tableVIII,column7). Thedecreaseintherateofformationof sodiumnickelitecausedby hydrogenmustbe dueto itsdirectpoisoningactionuponthenickel. *
Sincetheconcentrationof speciesof nickelotherthanNi+2(e.g.,Ni+3)whichmightbe iuequilibriumwiththehydrogenpresentisextremelylow(ref.6),thecorrespondingreactionsinvolvingthisconcentrationneednotbe considered.
“ SUMMARYOFRESULTS
Thefollowingzonereactionsaretheprobablethermal-gradientmasstransferof nickelinmlten
ColdzoneNa#$i02+ 2Na~ 2Na20+
Hotzone
Thediffusionofthemolecularspeciesinthemelt
onesoccurringinsodiumhydroxide:
NiP
allowstheprocesstocontinueindefinitely.Thesodiumoxidearisesfromthedissociationof - -sodiumhydroxideathightemperatures.Thesodiumnickelite(N~Ni02)isformedprimarilyfromthefollowingreactionandat a constantratewithanygivenpressureofhydrogen:
2NaOH+ Ni-E2 + Na2Ni02
Indynamictests,therateoftransferiscontrolledby therateofthezonereactionsandthusisconstantwithtime. Instatictests,therateoftransferiscontrolledby therateofdiffusionandthusincreaseswithtime. .
Theeffectsofadditivesupontherateoftransfercanbe explainedintermsofthismechanism.Additiveswhicheffectivelyincreasedthesodiumoxideconcentrationincreasetherateoftransfer;thosewhichdecreasethisconcentrationdecreasetherateoftransfer.Reducingagentsdecreasetherateoftransferin statictestsandhaveno effectind~mictests.Oxidizingagentshavetheoppositeeffectin statictests. .
LewisFlightPropulsionLaboratoryNationalAdvisoryCommitteeforAeronautics
Cleveland,Ohio,June20,1957
NACATN 4089
APPENDIXA
.
EXISTENCEOFSODIUMNX3CEUTEINSODIUMHYDROXIDE- NICKELSYSTEM
An investigationwasmadeto demonstratetheexistenceof sodiumnickeliteinthesodiumhydroxide- nickelsystem.Theexistenceofthisor a similarcompoundisnecessaryforsixofthesevenpossi.hlemechanismsofmasstransfer(tableI}. Thenatureofthehydrolysisproductofsodiumnickelitewasalsostudiedinorderto stistantiatethetheorythatnickelwaspresentinthe+2 valencestate.Thereactionbetweennickeloxideandsodiumhydroxidewasinvestigatedto showtherelativeinstabilityofthiscompoundwithrespectto sodiumnickellte.
. MaterialsandFrocedure
. The‘L”nickelcruciblesaswellastheprocedureusedinthisprepar-; ationwereidenticaltothoseusedinthestemdsrdstattccapsuletest: usedattheIewtslaboratory(fig.1 andrefs.2 and3). Thecrucibles.? were5 incheslongandhadan outsidediameterof 5/8inchandl/l6-inch-j thickwalls.Allthecruciblecomponentsexcepttheventtubewerefab-
ricatedfromthesamelengthoftubing.
Eeforethecrucibleswereassenibled,thecomponentswerecleanedina modifiedaquaregiasolutionfor2 minutes.Eelisrcweldingwithanatmospherecontaining15percenthydrogeninheliumwasusedinfabrica-tion. A flowof argonwasmaintainedinthetubeduringwelding.Sodiumhydroxide([email protected])wasweighedintoeachcrucibleinthedrybox.Thetopandventt&e werethenweldedintoplace.Themoistureandairwereremovedby evacuationto 5 micronsofmercuryat 5800F (in16hr)andthento 5 micronsofmercuryat72@ F (in4 hr). Afterevacuationandcooling,a covergasofheliumatabou~2 poundspersqusreinchgagewasadmitted.Crimpingandweldingatthecrimpservedto sealthecrucible.Thermocoupleswerespotweldedto thecrucibles1/8inchfromthebottomandalsoattheliquidlevel.A specialfurnace(ref.2)wasused.
IsolationandIdentificationofSodiumNickelite
Fourcrucibleswereheatedto 170@ F for24hours.Aftercooling,theywerecutintosectionsabout1/2inchlongina dryheliumatmosphere.Oliveto dark-greenneedle-likecrystalsabout1/8inchlongwerevisiblethroughoutthewhitesodiumhydroxidematrix.Similarcrystalsarere-.portedinreference6. Thesodiumhydroxidewasdissolvedawayfromthecrystalswithabsolutemethylalcohol.TheX-raypawderdiffraction
● pattern(tableV) determinedforthecrystelsmatchedtheoneswhich
20 NACATN 4089
havebeenreportedby otherlaboratories(refs.7 and9)andassignedtheformulaNa2Ni02.No Hues ofnickeloxideor nickelhydroxidewerefound.Theutmostcarewasusedto keepairandmoistureawayfromthecrystalsatalltimes.Thechemicalanalysisreportedforthiscompound(residue) inreference9 roughlycorrespondstotheformul.aNa2Ni02.Reference7 reportsanX-raypatternfroma residueobtainedby heatingsodiumhydroxidein nickeluntilallthehydrogenpresentwasevolvedasa gas;thiswouldlikewisecorrespondto,.theformulaNa2Ni02.Thefactthatthepatternofthisworkmatchestheothersindicatesthatsodiumnickelitedoesnotsignificantlyreactwithabsolutemethylalcohol.
Nowelldefinedcrystalswereobtainedby heatingsodiumhydroxideinnickelat 1500°F, sincethenickeliteconcentrationwastoosmall.
.
PreparationofHydrolysisProducts
Inorderto investigatethehydrolysisproducts,werepreparedinthewaydescribedpreviously.After
othercruciblesthe24-hourtreat-
mentat170@ F, thecontentsofthecrucibleswerewashedoutwithdis-tilledwaterandthecrystalspresentwereallowedtoremaininthewaterabout3 days.Brightgrass-greenneedle-likecrystalswereformed.Therewasalsopresenta slightamountoffineblackprecipitate,whichwasprobablyformedfromthehydrolysisof sodiumnickelitedissolved“in‘“thesodiumhydroxide.Afterremovaloftheblackprecipitateby recanta-tion,thegreencrystatiwerewashedandthendriedina desiccatorcontainingDrierite.Theywereshownspectroscopicallyto containnosodiumandtheirX-raypatternmatchedthatfornickelhydroxide(Ni(OEj2].Theindividualcrystalswerehexagonalinshapeas seenundera microscope.Thecrystalsturnedblackupon&r@ng at 120°C for24hoursbutstillretainedthesamecrystallineformandX-raypattern.Thecrystalslost11.8percentoftheirweightduringtheheatingprocess.Thechemicalanalysisoftheblackcrystalsshowedthemtobe nickelhydroxide(Ni(OH)2).AlthoughthegreencrystalscorrespondtothecompositionNi(OH)2.H20,theyprobablycontainmerelyphysicallytrappedwater(inasmuchastheX-raypatternwasunchanged).
At 1500°F thereactionbetweensodiumhydroxideandnickeldidnotyielda crystallinehydrolysisproduct(exceptinoneisolstedinstance).Theblacktobrownpowderregularlyobtainedby filteringthewatersolu-tionofthemeltneverthelessgavetheX-raypatternfornickelhydroxide.Thispowderwasprobablythesameasthes@@t atiouitofblackpowderobtainedfromthecruciblerunat 1700°F.
.
NACATN 4089 21
Hydrol~isIYoductsfromReactionof SodiumHydroxide
withNickelOxide
A nickelcruciblecontaining0.13gramof nickeloxideandtheusual13gramsof sodimnhydroxidewaspurged,sealed,andheatedto 1700°Ffor15minutes.Whenthecontentswerewashedout,largegrass-greenneedle-likecrystalswerefo~d. Onlya traceofthenickeloxideX-raypatternwasdetectableinthenickelhytroxidepatternfoundforthesegreencrystals.X-raypatternsof individualcrystalsprovedthemtobeactuallypolycrystalline.
Conclusions
Thereactionbetweennickelandmoltensodiumhydroxideproducessodiumnickelite(Na2Ni02). Thesodiumnickelitecanbe Isolatedas
●
needle-Me crysta~ifthereactionoccursat 17000F butnotif itoccursat 15000F or lower;thisis attributedtothegreaterconcentrationof
. thenickel.iteat 170@ F (tableVIII)andto a probablylargerateofchangeof volubilitywithtemperatebetween1500°and1700°F.
Thereactionbetweennickeloxideandsodiumhydroxidea; 17000F(andlikelybelowthattemperature)alsoproducessodiumnickelite(in-dicatedby thepresenceof itshydrolysisproductnickelhydroxide):
NiO+ 2NaOH+Na2Ni02+H20
Sincethisreactiongoesessentiallyto completion,verylittlepossibilityexiststhatthespeciesMO canbe presentinmoltensodiumhydroxide- .
nickelsystems.
Thereactionbetweensodiumnickeliteandwater
Na2Ni02+ 2H20+ 2NaOE+Ni(OH)2
“resultsintheremovaloftheN~O componentby leachingwithoutdisturbingtheexternalshapeofthecrystals.Wobablyduringthisprocesssingle
. crystalsofsodiumnickeliteconverttopolycrystal.linenickelhydroxide(containingphysicallytrappedwater).
.
.
22 ?MCATN4089
APPENDIXB.
.
THEEFFECTW HYIMXWNUPONMASSTRANSFERN!ZONICKEUTECONCENTRATION
Expertientswereconductedto showthesimultaneousdependenceofsodiumnickeliteconcentrationandrateofmasstransferuponthehydrogenpressureinnickelcapsulescontainingsodiumhydroxide.A morecompletestudyoftherateofformationof sodiumnickeliteispresentedinappendixC. tf-#
Method1
Fourstandsrd“L”nickelcruciblescontainingweighednickelspeci-mens(1/16in.thickand5/16in.indiam.)andsodiumhydroxide(13.W. 1 g)werepreparedandsealedintheusualmannerforstanda@capsuletests(appendixA andref.2). TheywereIncasedin Inconel’” Achambers(fig.2). Thechamberswereevacuatedto a pressureof 5 micronsofmercuryandthenconnectedto a supplyofhydrogen.ata constantpres-sureof 17inchesofmercury.Thetemperatureofthecrucibleswasraised -to 1500°F by meansofa specialGlobarfurnace(ref.2). Thisfurnace
..
by supplyi~,heatfromthebottommaintainedan 8@ F differencebetween1
thebottomoftheInconelchambersanda position~ inchesfromthebottom(thesodiumhydroxideliquidlevel).Thepressureofhydrogeninsidethenickelcrucibleswasalso17inchesofmercurybecauseofthepermeabilityofnickeltowardhydrogen.&_ter80hoursattemperaturethecrucibleswerecooledandwereradiographerto detecttheamountoftransfer.Theamountof sodiumnickelitewasdeterminedasnickelplus
-.
two,andthespecimenswerereweighed(tableVI). Theeqerimentwas—
repeatedwithcrucibleshavingnoaddedhydrogeninthechambers.Even “-
thoughnohydrogenwasaddedto thesecontrolcrucibles,a steady-state..
hydrogenpressureofabout4 millimetersofmercurywaspresent(ref.7).Thesteady-statepressurearisesin a sealedcontainerfromthefactthathydrogenisgeneratedbythereactionof u$ckelwithsodiunhydroxideandthathydrogenis lostby diffusionthroughthewallsofthenickelcon-tainer.Theratesofthesetwoprocessesbecomeequalwhena certainsteady-statehydrogenpressureisreached.
As isobviousfromthedataoftableVI,theaddedhydrogensimul-taneouslyinhibitedtheformationofthenickeliteandtherateoffistransfer.
Method2..
Testswerealsoruninwhichhydrogenwasintroduced&&ectlyintothehotzone(bottomofthecrucible)(fig.3). Thesetestsagainshowed “
NACATN 4089 23
11:1
.
thataddedhydrogensimultaneouslyinhibitedtheformationof sodiumnickeliteandtherateofmasstransfer.
.
Wthod 3
Thethirdmethodwasdesignedtodeterminewhetherinhibitingthelossofhydrogennormallyformedina cruciblewouldlikewiseinhibittransfer.Thiswasaccomplishedby placinga shellof Inconelaroundastandsrdnickelcrucible(fig.1(b)],inasmuchastherateof cliffusionofhydrogenthroughInconelismuchlessthanthroughnickel.
Theinternalandexternaldimensionsofthemodifiedcrucibleswereascloseaspracticableto thoseof a standardone (fig.1(a)).Twostandardcrucibleswereusedforcomparison(appendixA andref.2). TwoInconel-coveredcrucibIeswerepreparedby firsttreatingtheinnernickelcruciblesinessentiallynormalfashion.Thetopplugswerehsm-
$ meredintoplacedin a dryboxandthenweldedto sealthecrucibles.TheInconelshellswerethenassembledsroundthenickelandwelded
. together.Thesefom crucibleswereheatedat 156@ F (intheGlobsrfurnace)for24hours;afterthisthecrucibleswereradiographerandthecontentswerechemicallyanalyzed(tableVI).
TheInconelsheIlinhibitedtransferase~ected. Reductionoftherate offormationofthenickelitewasalsodetectableandwasattributedto an increasedhydrogenpressurecausedby thedecreasedlossofhydrogenby diffusion.
.
..
NACATN4089
RATEOFFORM4TIONOFSODIUMNICKELITE
APPENDIXC
A studywasmadetodeterminetherateofformaticmof sodiumnickeliteaccordingto theequation
Ni + 2NaOH+Na2Ni02+ H2
Ifthereactiondoesnotreachequilibriumina reasonabletime,themechanismofmasstransferinvolvinghydrogenas the“redox”material
●
(tableI,reaction(6))isimpossible.‘ .-
Procedure
Standardtestnickelcrucibles(fig.l(a))containing13.OW.1 b
gramofsodiumhydroxidewereprepared(purgedandcrimped)in theusualmanner(appendixA). Theywerein turnsuspendedinan Inconeltubeby .meansofa longNichromeloopspotwelded<o thecrimpofthenickelcrucible(fig.4). Thetopoftheloopwasplacedovera nickelplunger, ~whichcouldbepulledoutfromundertheloopby meansofa magnetout-sidethevacuumsystem.Brasscapsweresoftsolderedto thetopandbottcmoftheInconeltubesothatthesystemcouldbe evacuated.After - “-evacuation,hydrogengaswasadmittedand_controlledat thedesiredpres-sure.Thetemperaturewasthenraisedas quicklyaspossibleto 1500°F(44°F). ThetempemturemeasurementsalongtheoutsideoftheInconeltubeshowedthata temperaturedifferenceofatmost3°F existedbetween - “.thetopandthebottcmof themoltensodiumhydroxide.At theendofaruna magneticfieldwasappliedtotheplungersothatthenickelcruciblewasallowedtofalltothebottomofthebconeltube. Thenickelcrucible@s therebyrapidlycooled.Theexperimentwasrepeatedforvarioustimesandvarioushydrogenpressures(tableVII).Aftereachrunthecruciblewascutopenanditscontentswereanalyzedforoxidized
—
nickel(fig.5 aidtableVII).
Kinetics
Thesodiumnickeliteconcentrationa~s tobe a linearfunctionoftime(fig.5)andtobe independentofhydrogenpressuresabove26inchesofmercury.Theresults,althoughscattered,indicatetherateofformationtobe greaterforlowerpressures.Thedataforsteady-statehydrogenpressureconditions(37.5mg/24hr (appendixD)}confirm .
.
NAC!ATN4089 25
.
thisfact. Toexpressthiseffectanalytically,onemustusea kineticequationforreacticmsinwhichsurfacesareeasilypoisoned(inthis. caseby hydrogen):
rfom=kl +k2/(p#
where rform istherateofformationof sodiumnickelite,kl,k2,andn areconstants,and ‘H2 isthehydrogenpressure.‘ince ‘form ‘sisequslto kl athighpressure,kl = ti (mg/dm2)/yr(area= 0.267dm2).Thevsluesfortheotherconstantsare n= 1.4and kz = 360 [g(mmofHg)1”4/dm2]/yr.Thekineticsareobviouslyindependentofthereversereaction.
Conclusions.
Sincetherateofthereactionunderdiscussionislinearwithtime,thereactionitselfneverreachesequilibriumundertheexperimentalccm-
;- ditions(at1503°F, fortimesup to 600hr,andwithhydrogenpressuresbetween0.16and81 in.Hg). Itfollowsthatthemechanismofmasstrans-fercannotinvolvehydrogenas the“redox”material.
.
.
26 HACATN4089
.
APPENDIXD
REACTIONSOFCHROMIUMANDITSCOMFOUN12S
.
WITHMOLTENSODIUMEYRROXIDE
An Investigationwasmde todeterminethechemicaldifferencesbe-tweenthesodiumchromites(orcbmmates)producedfromthefollowingthreereactions:
1500°FCr +NaOH~ *
E1500°F P
Cr203+ NaOH~
1500°FNa2Cr04+ NaOH~
Thisworkwasinitiatedbecausechromiummetalwasreportedtobe an in-hibitorofmasstransfer,whilechromiumsesquioxideandsodiumchrmate “wereoriginallyinterpretedtobe detrimentalasadditives(ref.2).
.
Procedure Y
Standardstaticcruciblecorrosion(masstrsusfer)tests(appendixesA andB andref.2)wereperformedusingchromium(325mesh),chromiumsesquioxide(Cr203),andsodiumchromate(Na2Cr04)as additivesin thesodiumhydroxide- nickelsystem Theseadditiveswereaddedbeforepurg-ing. Inordertoaccanplisha morethoroughinvestigation,certainvari-ationsintheprocedurewereusedforeachgroupofcrucibles,as indi-catedin
The
thef~otnotesoftableVIII.
ChromiumMetal
ratioofchromateconcentration[Cr*]tochrmiteconcentra-tion[Cr-)as foundb watersolutionsisapproximately1 to 2 (tableWII, group1,column6),beingindependentofthepercentageofchrcmiummetalused(column3)andindependentoftimeforrelativelylongperiods(column5). Thecompositionis essentiallyuniformthroughoutthecruci-bles(tableVIII,footnote(b)).Forperiodsoftimelessthan1 hour,Cr+3aloneexisted(groupU). Group111resultsindicatethatthisapparentl-to-2mtio for[Cr*] to [Cr+3]doesnotdependuponthesteady-statehydrogenpressurethatispresenth closed(sealed)systems(ref.7). Thesefactsperhapsindicatethe,presenceof~* fi tieso~~ . =hydroxide.
.
NACA‘IN4089 27
.Since,however,this[cr*]to [Cr+3]ratiodependsupontemperature
(tableVIII,grtipIV),thel-to-2ratiomustbe onlyamroximate,and# is ~~ doubtful.. thustheexistenceof Cr Forfurtherdiscussion
seeappendixE.
ThereactionformingthehypotheticalCrw isfastenoughsothatanyhydrogensoproducedisquicklylost(bydiffusion),andis thuspre-ventedfrominhibitingsignificantlytheproductionof sodiumnickelite(tableVHI, column7}.
ChromiumSesquioxide
Theresultswithchromiumsesquioxide(tableVIII,groupV) showthatit isan inhibitoroftransfer.Thisis indicatedby ccanparison@thespecimenweightchanges(column8)andwassubstantiatedby thede-creasein therateoftransfera~rent inradiographsofthecrucibles.
~ (notshown). Chromiumsesquiaxideisnotbeneficalin dymmictests.
~Euringtheprocessthechromiumconvertsslowlyfrm Cr+3to theCr* as”foundin thewatersolution,butthereactionapparentlystopswhenit
~- is two-thirdscomplete.l%isperhaysindicatesthepresenceofC@.~ Thehydrogenproducedfromtheoxidationof thechramiumsesquioxide
keepsthesodiumnickeliteconcentrationlow(tableTCCII,column7].Themagnitudeofthisphencwnenondependsupontheactualpercentof chro-miumsesquioxideused. It seemspossiblethattheCr203inhibitstrans-ferby thispropertyofkeepingthenickeliteconcentrationlow(thusworkingina differentmannerthanchromiummetal).
GroupVI results(tableVIII)ccmrparethethreeadditivesused.Theseresultsconfirmthepreviousdata(groupsI toV). Thetestswithsodiumchrcmateindicatethatthechromiumconvertstoessentiallythesameoxidationstateas it doesin thechromiumsesquioxide(tableVICI,column7). Thisresultsintheoxidationofmetallicnickel(tableVTIX,footnote(f)).
summary
.
!llbefollowingequationssummarizetheinformationobtained:
(1)Cr+3NaOH ~Na3Cr03 +~E2;
Moderatelyfast
Na3Cr03+ NaOE~ cr+4,+5,or+8+ H2
*
28 NACATN4089
“
(2)Cr203+ 2NaOH+2NaCr02+~O; NaCr02+ 2NaOH~Na3Cr04 +H2;.
~t + ~ao= Inhibitedby hydrogen> Na2Ni02+ H2pressures
(3) 2Na2Cr04+Ni + 2NaOE+2Na3Cr04+H20 +NiO
SeeappendixE forfurtherdiscussionofchramiummetal.
.
.
.
.
NhcATN4089 29
.
APPENDIXE
C!HROMATE-CEROMITEEQUILIBRIUMINM3LTENSODIUMHYDROXIDE
A studywasmadeconcerningtheequilibriumbetweenthevalencesofchromiumformedwhenthemetalreactswithmoltensodiumhydroxideat1500°F. Itwashopedthatsuchan investigationmightuniquelydefinethevalencesofchromium.
Rrocedure
Standardstatic-testnickelcruciblescontainingsodiumhydroxide(13.H. 1 gram)andvariousamountsof chromiummetal(325mesh)werepreparedintheusualmanner(appendixA andref.2]. Theywerethenheatedina hydrogenatmosphereat 1500°F andquenchedinthemannerdescribedinappendixC. Aftereachcruciblewasopened,itscontentwasimmediatelyacidifiedwithdilutesulfuricacid. Thenthewaterphasewas
. analyzedforCr+3andCr* (tableIX,columns4 and5). Theinsolublematerialwasassumedto containonlyCr+3.
Equilibrium
Proofthatequilibriumwasactuallyreachedisthefactthatthe~fequilibriumconstants”subsequentlycalculatedareindependentof testduration(tableIX,column3).
Theequilibriumbetweenthelowerandhigheroxidationstatesofchromiumas itexistsinmoltensodiumhydroxide(notnecessarily,as itexistsinwater)isexpressibleby thegeneralequation
[Cr+?l“P=,=‘1= [Cr+x]
where K1 istheequilibriumconstant.SinceinfactthevalueofPnO[Cr+61/[totalCr]isessentiallyconstant(tableIX,column6),one
m~t concludethat+3and+5mustbe thevalencespresent([Cr+5]beingproportionalto [cr+6];[Cr+3]beingessentiallyequalto [totalCr];and
. thedependenceupon‘%
beingthefirstpower).Thisselectionisnot
●
30
unambiguousifchromiumcanconstantforthiscase K2
NACA’I!N4089
existas a dimer.Sincetheequilibriumcanbe expressed
l!!r~2 pH2Y-xK2 =
[1+2XC!r2
.
thevalenceswould.be +-3and+4. Thevalences+3 and+5are,however,thesimplerexplanation,beinginagreementwiththeworkinwhichchromiumsesquioxideisoxidizedtoC!r+5(appendixD). Furthermore, ‘“ “-sincechromiumlsapparentvalenceat 1700°F is+4.1(appendixD),somepentavalentchromiummustbe present.
ComplexFormation
Whena compoundcontainingCr‘5 isdissolvedinwaterandacidified,thefollowingdisproportionationreactionshouldoccur:
6Na5Cr05+ 29E2S04+Cr2(S04)3+ 2Na2C!~207+ 26NaHS04+ 16E20..
Ifalltheresultinchromiumcompounds~ watersoluble,onewouldexpect8[solubleCr+3]/[Cr+]to eqpal0.5,not3,3(tableIX,column7). A-~alue
of3.5(almost3.3)to 1 forthisratioi~understandable,iftheCr+5isassumedtobe coordinatedwithCr+3accordingto thefirstformulaintheequation
—
-—
._
6Na5Cr05(Na3Cr03)2+ 83H2S04~7Cr2(S04)3~2Na2Cr207+ 62NaHS04+ 52H20—
Underconditionsof lowandvaryinghydrogenpressure(unsealedsystems)onechromite(Cr+3)probablycoordinateswithonehypochromate,aa in-dicatedbytheapparent+4 valence(appendixD andtableVIII,grouyIII).Thiscoordinationchangeswithtemperature,as indicatedby theapparentvalenceof+4.1at 1700°F (appendixD).
—..—
—
NACATN4089
.
31
.
EQUILIBRIUMCONCENTRATION
APPENDIXF
OF SODIUMOXIDEINSODIUMHYDROXIDE
Theequilibriumconstantforthereaction
2NaOH(liquid)#N~O(solid)+ H20(gas)
wascalculatedfromthermodynamicdata(tableX andrefs.13and14). It
wasfoundtobe 10‘3”96atmosphereat 822°C. Sincethepartialpressureofwateroversodiumhydroxideisabout3 millimetersofmercuryat822°C(ref.7),themolarconcentrationofsodiumoxidemustbe about2.7percent.Thisis ingeneralagreementwithrecentexperimentalwork(ref.10).Thisconcentrationseemsmorethansufficientfortheoxideto takepartinthechemicalreactionsinvolvedinmasstransfer..
Themolarconcentrationofwatermustalsobe 2.7percent,sinceverylittlewateris inthevaporstate(becauseofthelowvaporpressure)..Thislowpressureisindicatedindirectlyinstaticcrucibletestsusingflowingatmospheres(appendixD andref.3). Ifpressureslargerthanseveralmillimetersofmercuryexisted,largequantitiesofwaterwouldbe removed,andthesodiumoxideconcentrationandmasstransferwouldthusbe increased.
Ifonedisregardstheexperimentalevidenceandassumesallthewaterformedfromthedissociationtobe containedinthefreespace(10mlina standardcrucible),themolarconcentrationof sodiumoxidewouldstillbe sufficient(0.02percent)to takepartinmasstransfer.Theresultingwaterpressure(400mm Hg)wouldbe,of course,ingrossdis-agreementwiththefacts.
.
.
32 NACATN40s9
APPENDIXG.
.
CORROSIONANDMASSTRANSFERINSODIUMHYDROXIDE- COPPERSYSTEM
Inordertocomparethemechanismof transferinntckelwiththatinanothermetal,thesodiumhydroxide- coppersystemwasinvestigated.
Procedure
Staticcorrosiontestsinwhichcoppercrucibleswereexposedtoairfor24hoursresultedinexcessoxidationandleaking.Therefore,static-capsulecorrosiontestswereperformedwiththecoppercruciblesprotectedfromoxidationbybeingincasedinIncortelchambers(fig.2).
Fourcrucibleswerefabricatedfromcoppertubinginthesamemanneras the“LWnickelcrucibles(appendixA andref.2). Inthreeoutof .fourcasesduringthe1500°F test,leaksoccurredat thelowerweld(belowthesodiumhydroxidelevel).Subsequentcapsuleswerethereforefabricatedfroma singlebarof coppertothesamedimensionsas the
.
standardnickelcrucibles.Thespecimens.~ndtopsweremachinedfromthe- ‘-samebarstock,whiletheventtubesweremadefromtubingintheusualmanner.Theloadingandpurgingprocedurewasidenticaltothatusedfornickelcrucibles.Theconditionsofthetestwerethesameas thoseformethod1 inappendixB exceptthattheInconelchatierswerecontinuallyevacuated.Nickelcrucibleswereusedforco?qparison.Afterthecrucibles
—
wereradiographer,themeltswereanalyzedchemically,andtheweightlossesof thespecimensweredetermined(tableXI).
PresenceofSodium
Severaldaysafterthetest,a whitealkalinecoatingwasdetectedaroundthecrimpon someunopenedcoppercrucibles(tableXI,column5).Thisphenomenonwasinterpretedasbeingduetotheformationofmetallicsodiuminthemeltanditsdiffusionthroughthecoppermetalathightemperaturefollowedatroomtemperatureby itsreactionwiththecom-ponentsof theairtoformsodiumcarbonate.Photomicrographsof thecrimpedsectionsindicatedthatsomeforeignmaterialhadvaporizedfromthesurfaceofthecopper.
——
Furthermore,theextremelylargeweightlossesofthecoppercrucibles(tableXI,column6)canbeattributedonlytothelossofmetallicsodium.Radiographsof thecoppercruciblesshowthislossof sodiumhydroxide.Unfortunatelythecontentswerenotanalyzedforsodiumhydroxide.All ●
thisevidenceindicatesa reactionforcopper
Cu + 2Na20+Na2Cu02+ 2Na .
whichissimilarto thatfornickel.
NACATN4089
EffectofChromium
33
& Radiographs(notshown]andthespecimenweightchanges(tableXI,column7] indicatethatchromiumisan inhibitorof transferinthesodiumhydroxide- coppersystem,as couldbepredictedfromtheproposedmechan-ism. However,anotherfactorhastobe consideredhere. Thepressureofhydrogengeneratedby thechromiumreactionisnoteasilyeliminated,sincecopperislesspermeabletohydrogenthanisnickel.Theresultinghydro-genpressureinhibitsthechromite-to-chromatereaction(column8)aswellas thecopper-to-cqpratereaction(column9). Thisaccountsforsomeoftheeffectivenessof thechromiuminc~er butnotforthatinnickel.
ComparisonofTransferofCopperandNickel
Radiographsof cruciblesshowlittledifferencein theactualamuntof transferforcopperandnickel.Thefactthatthespecimenshave
. relativelylargeweightlossesisattributableto thelargeamountofcopper(fromthespecimensaswellas thecrucibles)whichreactswithsodiumoxidetoformmetallicsodiumandthesodiumcuprate.Itisthe.
y sidereactionsof copper(oxidation,sodiumdiffusion,andfastrateofH reactionwithsodiumhydroxide)whichmakeita farlessfavorablecon-V tainermaterialformoltencausticthannickel.
34 NACATN4089
AFTENDIXH
*
TIDIRMOEUMTRIC!ALPOTENTIALOFNICKELINMOLTENSODIUMHYDROXDI!3
Thissectiondescribesthethermoelectricalpotentialandcurrentfoundtoexistinthesodiumhydroxide- nickelsystemandpresentsatentativeexplanationforthem.
1+P
Apparatus *
Althoughmanytypesofapparatuswereused(mostofwhichhadaboutthesamedegreeof successandreproducibility),onlythefinalapparatusisdescribedhere(fig.6). TheoutsidecontainerwasfabricatedfromInconelandmadevacuumtightby meansofa water-cooledO-ringsealbetweenthetwobrassplatesontop. Thecupinside(containingNaOH)andtheelectrodeswerefabricatedfrom“L”nickeltubing.Theelectrodes *werevacuumsealedtothesystemby meansofrubbertubingandelectricallyinsulatedfromtheInconelsu~ortsby meansof ceramictubes.Heatingwasaccomplishedbymeansof ceramic-coveredNichromeheatingwirewound
.
aroundtheInconelcontainer.Severallayersofasbestossuppliedsuf-—
ficientheatinsulation.Thetemperaturewasmanuallycontrolledthrougha Variacandmeasuredby thermocouplescontactingthetiPsof theelectrodes.
...
ThermoelectricalPotential
Whilethetemperatureof thevesselandoneelectrodewasheldconstant,thesecondelectrodewascooledbypassinga streamofairthroughit. Thetemperaturedifferencebetweenthetwoelectrodeswasmeasuredaswellastheelectricalpotentialexistingbetweenthem. Gen-erally,theresultswereveryscattered,althoughthehotterelectrodewasalwaysnegativewithrespecttothecoolerelectrode.OccasionallY~_thescatterwaslow(fig.7).
Valuesforelectromotiveforcedividedby thetemperaturedifference(emf/AT)formanytrialsrangedfrom10 to120microvoltperdegree.Thesevalueswereessentiallyindependentofwhethera vacuumora heliumatmospherewaspresentabovethesodiumhydroxide.However,whena hydro-genpressurewasused,thesignof emf/ATwasreversed(i.e.,thecoolerelectrodebecamenegativewithrespecttothehotterone).Uponremovalofthehydrogenpressuretheoriginalpolaritywasrestored.
—
.
.
l!WCATN4089 35
.
ThermoelectricalCurrent
. An experimentwasalsocarriedouttoseewhetherthisthermoelectri-calpotentialcouldsustainan electriccurrent.Thehotandcoldelectrodeswereconnectedthrougha lowresistance(2ohms),andthepotentialdropacrossitwasmeasuredovera periodof 6 hours.Thecur-rentwasquiteunsteady.
Interpretationn
Themeasuredpotentialcouldbedueto theSeebeckeffector tooneof thefollowinghalf-cellreactions:
ColdzoneNa2Ni02+ 2Na++ 2e- — 2N~0 +Ni
E~Coldzone
2H20+ 2e-~ 20H-+ H2HotzoneColdzone
NaOH+ e- ,~_ OH-+ NaHotzone
ColdzoneNa++NaNi02+ e- - Na2Ni02
Hotzone
TheSeebeck(thermocouple)effectisnotlikelyforthefollowingreasons:Sodiumhydroxideisprobablya poorelectronicconductor,andtheeffectofhydrogenisfartoogreatforthepotentialtobe dueto siqplethermo-coupleaction.Theeffectof thehydrogenuponthethermoelectricalpotentialisevendifficulttoexplainintermsof anyof thethermocellreactionslisted.Thepresenceofhydrogenshouldinallcases(byreducingtheratioofproductstoreactants)increaseemf/AT(notdecreaseitpastzeroas isfoundexperimentally}inasmuchas thevalueof emf/ATisdeterminedby theequation(ref.15)
where n isthenumberof electronsshowninthehalf-cellreaction,~istheFaradayconversionfactor,and AS” isthestandardentropychangeforthereactionaswrittenforthehotzone.Theonlyreasonablewaytoexplaintheeffectistoassumethatthethermoelectricpotentialisproducedby thenickel-nickelitereactionwiththehydrogenpoisoningthe
. nickel.InappendixC anothercaseispresentedinwhichhydrogenappar-entlypoisonsnickelin thesodiumhydroxidesystem.
36
1.
2.
3.
4.
5.
6.
7.
8.
9.
NACATN4089
.REFERENCES
Mosher,DonR.,andLad,RobertA.: KineticStudyofMassTransferby ●
SodiumHydroxideinNickelUnderFree-ConvectionConditions.NACARME53K24,1954.
Forestieri,AmericoF.: EffectsofAdditivesonCorrosionandMassTransferinSodiumHydroxide- NickelSystemsUnderFree-ConvectIonConditions.NACARME54E19,1954.
Lad,RobertA.,andSimon,SidneyL.: A Studyof CorrosionandMassTransferofNickelby ~lten SodiumHydroxide.Corrosion,vol.10,no.12,Dec.1954,~. 435-439.
Le Blanc,M.,andBergmannjL. (A.Pingell,trans.):TheEffectofMetalsonMoltenSodiumHydroxide.Trans.393,NavalRes.Lab.,1952.
Manly,W. D.: FundamentalsofLiquid-MetalCorrosion.ORNL-2055,MetallurgyDiv.,OakRidgeNat.Lab. (ContractW-7405-eng-26.)
Smith,G.P.: CorrosionofMaterialsinFusedHydroxides.ORNL-2048,MetallurgyDiv.,OakRidgeNat.Lab.,Mar.27,1956. (ContractW-7405-eng-26.)
Williams,DaleD.,andMiller,R. R.: ThermalandRelatedPhysicalPropertiesofMoltenMaterials.Pt.II- HighTemperatureReactionsofSodiumHydroxide.WADCTech.Rep.54-185,WrightAirDev.Center,Wright-PattersonAirForceBase,Feb.1955. (MIPRNo.33(616)-54-102,Proj,No.1252.)
Peoples,RobertS.,Miller,PaulD.,andHanqan,H.Dale: ReactionofNickelinMoltenSodiumHydroxide.Rep.No.BMI-1041,BattelleMemorialInst.,Sept.27,1955. (ContractW-7405-eng-92.)
Anon.:High-TemperatureChemistryofFusedSubstances.Prog.Rep.,CollegeofArtsandSci.,Univ.-Ark.,Nov.1,1954. (Subcontract-501,W-7405-eng-26.)
10.Smith,G.P.: ProblemsPertainingto theDehydrationofSodiumHydrox-idebyVolatizationofWater.ORNL-2130,MetallurgyDiv.,OakRidgeNat.Lab.,Sept.10,1956. (ContractW-7405-eng-26.)
11.Simons,EugeneM.,Miller,NeilE.,Stang,JohnH.,andWeaver,C.Vernon:CorrosionandComponentStudiesonSystemsContainingFusedNaOH. Rep.No.BMI-1118,BattelleMemorialInst.,July30,1956. (ContractW-7405-eng-92.)
.
.
NACATN4069 37
.12.Smothers,W. J.: ProgressReportJan.1,1953-Mar.31,1953. Inst.
Sci.andTech.~Univ.Ark.,Apr.20,1953. (Subcontract501-Under
a W-7405-eng-26.
13.Kubaschewski,O.,andEvans,E. L.: MetallurgicalThermochemistry.vol.1. AcademicPress,Inc.,1951,pp.274-275;.291.
14. Powers,W. D.,andBlalock,G. C.: EnthalpiesandSpecificHeatsofAlkaliandAlkalineEarthHydroxidesatHighTemperatures.ORNL-1653,ReactorE@. Eng.Div.,OakRidgeNat.Lab.,Jan.20,1954.(ContractW-7405-eng-26.)
15.Eastman,E. I).:ThermodynamicsofNon-IsothermalSystems.Jour.Am.Chem.Sot.,vol.46,June1926,pp.1482-1493.
TAKE I. - RIOFOSBD~
me of reaction Rsaction M@atlon~ocemes A&lltlonalreactionsr~quireaforhltlatiou
coldzonebysical (1)~lO(dissolved)— Ni”(raetal)
~NIO(hot)- (CO~) Iiooe
coldZO*leda’ochmical(Z)Ha@Oz+me++2.-~ ma20+Hi Hs#i~(hot)~ (col@ ZN~ ~H~O+ ~0
Hotzone MS+(hot)~ (COl@e (hot)~ (cold)Ma20(cold)+ (hot)
coldzonehmlcd lV~~+● RedUctant
~“Oximt+Hi
ColdZowImpority (3)Ha#i02 + ZNa#eOp. “m~e~ +Hi Na2Nio2(hot)+ (cold) None
Hot zone Na2F~ (hot)+ (CO~)?l~Fe03(cold)+ (hot)
cold zon~Dlsproportionatiou(4)5Ra$i02~ “~a5Hi03~NI Nz#iOz(hot)~ (ColdHi● ~~ + N&#iOZ~Hz,. ,, Hotzone H~N105(COld)~ (hot)
IElementpresentinsodiumhydroxide Cold zoneO~gen (5)Ha.-j’?i02= H~O + ~ 02 + Hi Na2Ni02(hot)+ (cold)
NwO(COM) + (hot) Hi + 2NaOfl+ lk#i~ + ~02(cold)+ (hot)
Cold zcq~n (6)Iiu#02 + ~ & 2NaOH+ ill NS2NI%(hot)+ (cold)
H2(hot)+ (COhi) HoneNam(cola) + (hot)
cold zoneSedlum (7)Na#i02+ ma _— 2N~0 + Ni ria#f02(hot)+ (cold)
Hot zons Na(hot) + (cold) mam + N%o + F$OIh#(cOld) + (hot)
,I
NACATN 40S9
.
39
TABLEII.- ANALYSISCTSODIUMEYDROXIDE
PELLErs
[Ref.2~
I Material
Sodiumhydroxide
Chloride(Cl)
Iron(Fe)
Otherhea’vymetals
Carbonate(Na2C03)
Phosphate
(asAg]
Silicaandsmmoniumhydroxideprecipitate
Totalnitrogen(asNH3,N02)
Sulfate(S04)
Percent
97.6
.005
.001
0
.32
0
0
.001
0
TABLEIII.- COMPOSITIONCl?“L”
NICKELTUBING
~ef, 2.]
Material Percent
Manganese
Iron
Copper
Carbon
silicon
Sulflu
Cobalt
Nickel
0.149*0.015
.059k.006
.02$&-.006
.02~.005
.02c&.cQ2
.Ow&.ool
.119*.013
Balance
40 HACATN 4089
TABLEIV.- EFFECTUPONTRANSFEROFINCREASINGIMPURITYCONTENT
[Refs.2,3,and112
Specimenweightchangeforvariouspercentsofaddedimpurities,mgo 1 3 5
percentpercentpercentpercent
fickelimpurities
Manganese -4.7 -20.0 ----- -----
Iron -5.1 -25.7 -28.4 -28.1
Copper -5.1 -8.4 -6.7 -5.8
Carbon -7.9 -78.6 -8.5 -7.8
Silicon(seesodiumsilicatebelow)---- ----- ----- -----
Sulfur(seesodiumsulYatebelow) ---- ----- ----- -----
Cobalt -5.1 +0.6 +7.2 +8.2
3odiumhydroxideimpurities
Sodiumchloride -7.9 -6.5 -6.6 -----
Ironoxide -5.1 -9.2 -11.1 -4.8
Silvermetal -5*1 -6.0 -4.8 -5.1
Sodiumcarbonate -7.9 -17.5 -19.3 -22.0
Sodiumphosphate -5.1 -5.4 -5.2 -8*6
Sodiumsilicate a-7 a-n ----- -----
Nickelnitrate a-7 a-n ------. -----
Sodiumsulfate a-7 a-n ----- -----
aApproximate(ref.U).
NACATN 4089
.
●
A--
.
TKBLEv. - X-RAYDIFFRACTIONDATAFORSODIUMNICXELITE
hteratomitlistances,
d,
:
8.48----4.354.17----3.22----2.55----2.432.232.172.101.921.87----1.65----1.601.57--------1.44----1.38------------1.27--------
Intensityof line
Medium------StrongStrong------Weak------Medium------StrongWeakWeakWeakWeakWeak------Weak------WeakWeak------------Medium------We&------------------Weak------------
d values‘remref.9OakRidgej.
:
--------
4.32--------3.212.712.572.532.422.272.142.091.931.871.701.641.611.601.561.531.491.441.411.381.341.311.291.271.231.22
d valuesfromref.9(University)fArkansas)
;--------
, 4.37--------3.19----2.55----2.442.282.162.091.92----------------------------1.491.451.43------------------------1.22
d values‘remref.7,
:
----4.874.334.173.67--------2.55--------------------1.91--------1.65------------------------1.41----------------------------
MethodTotalnumberof crucibles
a4
4
ai?
2
PressureDf hydrogen,
in. Hg
0.16(Steadystate)
17
0.16(Steadyi3tate)
>0.16(slightlyshovesteadystate)
aC!rucfblesnaedforcomparison.
● ✎
Temperature,
‘%?
1500
1500
1560
1360
Time,h
80
80
24
24
Specimenweight10ss,
ma
4.3s.?
2.8M.6
7.&o.8
4..6M.3
Amount
~f+2
in melt,
w
11.4A-.9
0.4M0.05
79.&l.8
8.3MI.1
● ✌
Appearanceoftransfer inradiograph
Clearlyvisible
Hardlyvisible
Clearly visible
Nexrowerbandthanincruci-bleswithsteady-statehydrogenpressure
NACAm 4089
.
●
43
II!ABLEVII.- KPKETICDATAFOR
80DIUMNICKELITEFORMATION
rime,hr
615303303284228184I-388989888782634543A--4026222119181400
3!u+2inmelt,W
1.40.66
1.11.64
1.11.437.343.285
1.03.21.42.25.638.167.093. Xl-5.1.30.080.062.218.150● 175.090.100● 070
Hydrogenpressure,
in.m
81.026.5-16.016.5-15.028.328.329.13*538.420.738.712.770.673.237.686.560.086.120.738.428.968.4--------
44 NACATN 4089
TASLSVXII. - CORR@IONDATAWITHCHROMIUM,CHR~IUMsESQOIOXIDE, ANDSODIUM
LDDITIVI?96-
Peroentcd:+sk esCrremaind6r_
L332:5:.5
L
30.133.s_0.7:5.0.533.730.8-Ti30.4‘-
{32.0““33.s:-32.6
o_01.5:--
------.—33.s33.9----35.936.1..--_&_35.336.6z
r----.43165.3----60 “-
Y2 ,.-—- ~5661 ~3437 “-57 =
----32.0‘“.S7.O32..3:.:63.0-24.215.5-”.66.1--—42.837.z”-33.2.4s.524.613.k63.?----39.s:38.2-46.526.0
9-i- 3 7 I 82lroupoflrucib19B
~
‘b’cI
‘eII]
munt 0dditlvepercent
+
mountof Specimen~i+z weightn melt, ahange,
s ❑s
Averagerateor specimenielghtchange,
mtipT1500 24
24
2496
2%
245
c-r 2.5
7.5
55.8
;:
10
Crc??Cr
NoneNoneCrCr .
None
ENonec??(E-
Ncmecr203CI’2U3cr203NoneCT203Cr203Cr203None2ACr203Cr20~”b$s
Nonezti&$3Cr2U~Cr20tiazC’&NoneCrCTCrcr*03Cr203CrzosNa2CrO&Ncmecl’CrCr203cr*03
/j
-----—.33 -t-
-12.9-10.5+4.5+4.s
+
24
‘r 1
1700 24
H
..IV 0.251 -32.3.255 +7.0.195 +9.0.232 -30.5.235 -7.0J?C7 +7.0
----
2----13
T~15CQ 24
I96
i18
1240240
.-0.2s
-.?tl
-.48
-.56
0.0375 -5.9.01065 +2.s.0L602 +2.6.0369 +2.1.aem. -xl.5.05112-24.1.0286 +3.6.0311 -5.5
----
:.25----
:6.3—-.
;4.916.2
----
:513.53----
:5.
:53----33351
-1135 -81.2.0945 -79.6.0465 -11.4.0668 +.6
[.0914 -139.4.0931 -’74.6
T0.0834.1053.0669.0W2.0376.0227.0285(f)
.0748
.07s0
.0772
.07s4
.0504
.0244
.021.5(f)
I
1 ) 67
[71
I
1111700 8.5 _l-.1591.1640.1659.0396.0423
..“lheaec.ruclble~wereoutinto6everalsections; differencesin an.alysiebetween.?.eotione
werealight(column6).bNoanalg’sii.WR~.madqforNi+2.%0 epecirnens.d10Percent325meshplus13 percentle,rgermesh.‘Flowingheliumatmosphereoversodiumhydroxide(ref..3)..fCoatingof oxidewasfoundon lrmidesurfaceof crucible.
. -
1
Totalmount ofLu?omimj
g
0.1300.1310.129i3.1296.1304.1307.39CCI.3903.3899.3900.3901.3900
2
Iydrogen)ressure,
‘H2?iu. Hg
79.268.428.322.414.810.669.263.860.140.640.217.5
.. -.4314 -
r , * ,
TABm Ix. - CKFKMITE-CHRCMATE liQUILUIHJXMIM’M
3
mm,hr
141616Is3215184065148814
4
Cr+6inmelt,B
o.mx15.0016.0026.0039.0070.0088.0046.0049.0055.0103.0094.0163
5
Acid-soluble
Cre Inmelt,i3
o.CY353.0050.00s3.0140.0177(a).0135.0149.0175(a).0246.0447
6
P [Cr+q/[total Cr]H2
in. Hg
0.91.84.56.67.80.71.81.80.85
1.07.97.73
I Average~
[Solub~e Cr+3 ]
[Cr+6]
3.53.13.22.83.9--”3.43.33.1---3.82.7
Average =
.
TAmE x. - TH2BM2DMC DATA FOR SODIUMHYDR~E, SODIUM~E, MD WATER
[Refs.X5 and 14.]
Heatof forwtlon, Entropy,S, at Specificheat at conetant Heat of Heat of fusion,
-Orm’pressure, transformation
%u3ion’at 25° C,
cal/~;l~\(%) c? from alphacal/mcle
calfmnle cal/(m2;e)(%)to beta,
‘txans ‘cal/mole
Sacllumhydrotide -102,000 14.2 17.6 1000(CL)(eolid)
----
BodluaIhydrofide -------- ---- 17.6 ---- 1900(B)(eolid)
Sodium hydroxide -------- ----(liquid)
19.6 ---” ----
Sodiwnoxide -102,900 17.4(solid)
M .9+5.4X1O-3T ---- ----
Water (gae) -57,8CXI 45.11 1.19+2.63x10-%-0.16x10-%C ---- ----
● ✌
Cruciblematerial
rNickel tubingNickel tubingCopper tubingCopper baCopper bar
CopperbarCopper bar
l--Nickel tubingCopperbarCopperbar
43,1.4e , .,
TABLEXI.- COBR0510NDATAFORSODIUMHYDROXIDE- COPPERSYSTZM
lime 1500 24None 15C0 24None 15Cm 24None 1500 24None. 1500 24
3 Em 243 1.5oo 24
None Mm soNone 1500 soNone I-5oo so
None -4.4 -2.1None -4.9 -3.560 ------ -3.40.5 -45.9 -9.4.5 -25.6 -9.6
----- ------ +5.2----- -943.3 +6.2
None +0.7 -8.115.8 -604.1 -28.54.8 -788.2 -33.7
--- I --------- -------.. 0,0569--- .0S52--- .0S46
48 NACATN 4089
.
.
—
1
—
—
————
—
—
<ITlckeltube
/J(0.D.,58”-wall,1 16”
~l%rgedsodium
/(C%rcmel-Alumelthermocouples
—
&———
\\“L” nickelspecimens
(Diam.,5/16”;1/16”thlck)~
——
——
———————
r
_Inconeltube(0.D.,5/8”;wall,1/32”)
.
-“L” nickeltube(0.D.,1/2”;wall,1/32”) .
(a)Standardorucible. (b)Modifiedcrucible.
Figure1.- CcmpmiaonofstandardcrucibleandcruciblewithIncoR.1shell.
9
*
NAC!ATN 4089 49
To hydrogencylinder
To otherInconelchanibers
.Bourdon-
er(l%”long)
forthreeaddit1 chambers
To @ ,,,,A,,,,:.:,:,:..,,.:;..........~.,....~..,=,...,,.,.,...,.,-. Firebrick
Figure2.- VacuumsystemandInconelchamber.
.
\
50 NACATN4089
T“+Helium
gas
I
I5“
rWeld F’u&To vacuum\
ff
+-
CrimpsVacuumvalve
i
Sodiumhydroxide
Surfacetbl?OU@ whiohhydrogendiffusesintocructble-
—
——————
————————————
is
+To hydrogencylinder
●
✎
- “L”nickeltube(0.D.,5/8”;wall,1/16”)
~“L” nickeltube(0.D.,1/4”;wall,1/32”)
.
F Chromel-Alumelthermocouples
-.
Figure3.- Crucibleforintroductionofhydrogenathotzone..
.
NACATN 4089 51
●
✎
BraB8capBourdon- J.typegage 0’/
d~
Electromagnet
Tow30UWRlckelrod
PwPt %+ -.
-/”vacuum 1 “. IValve .-!,
“’f------k-.D.,,nTo h@rO@ncylinder
-I. D.,0.630”
12’
Chromel-Almelthermocouples1“ Epart
_ Inooneltube(wall,1./l6°)
_ “L”nickelmucible
- Thermocoupleatmiddleof acdiumhydroxide
Waterforcmling
Brase cap
Figure4. - Apparatusforkineticandequilibriumetudies.
37
37I 1A 14tozl 1 [1111
--H ; ‘2LJ-H-I-H0.16 (4 m Eg)
o w Km Eio m 2S3 m 3!s0 400 430 Sal 350 Ka 650Tim, h
MBLUC 5. - Rate of formationof nndiwmnickelite_ nickl arkl.atim hydroxideat Moo” F at v!nlcawh@rc4en preamlrea.
,,
NACATN 4089“ 53
.
.To mmpreasedair
Rubber
Bressplates -Diem,5“7 I Mlr---- ‘T”-”’“’—”=
l\ ILp3/8” –
II!
Chr0m91-Alumelthermcouples~
Sodlmhydroxide-
for
“?
:
‘L”nickelelectrode
/(0.D.,3 16”;wall,1 32’;1“betweenL?entera)
“L”tiCkelcup](0.D.,2$; walll/1#’;3“hi@~
HInoonelcontainer(0.D.~; wall,l/16”~
J-1T3/4”
-
Figure6.- Apperatuaformeammmnt ofthermeleotricpotentiah.. .
. .
16
14 /0
o/
12 0
B1/ d
o10
/~$
kja
i J 0
J“ ; /
6 0
d/
4 u
{)
*o 50 100 150 ?.Cm 250 300 350 Km
!Cem@raturedifference,%,Ij.
Fi&Ure-1.- !F&rmeelectricpotentiallw3tveentw nlc.)mlelectrodesat differenttemperaturesIn aadim hydrmdde. Highertemqezature,67@ C.
1
;1 iI,,, , ,,
,,I,, ,i,, d : i,l ,,. “ ~ ‘ ‘“ > •~!LdIT;ti))I;- ‘~ l’;; 1: ‘iI .I%ll .1. I I $$$ .1,1Iul.,1..1,!.1..,> !,, .I. JL;,
I
ul*
,.
I
