Thermochemistry and Phase Equilibria in Calcium Zeolites

American Mineralogist, Volume 81, pages 658-667, 1996

Thermochemistry and phaseequilibria in calcium zeolites

InrN,l KrsELEvArr Ar.nxANon-l Navnorsxyrr IcoR A. Bu,rrsryr2 ANDBonrs A. FunsnNro2

rDepartment of Geological and GeophysicalSciences, Princeton University, Princeton, New Jersey08544, U.S.A. 2lnstitute of Mineralogy and Petrography,Academy of Sciences,Novosibirsk, 630090, Russia

ABSTRACT Thermodynamic properties of the natural calcium zeolites laumontite, leonhardite, dehydrated leonhardite (metaleonhardite),wairakite, and yugawaralitewere studied by calorimetry in lead borate solvenIat9T5 K. Enthalpiesof formation from the elementsat 298 K are as follows: -1251.0 + 8.5 kJlmol for laumontite,CaAlrSioO,r.4HrO; -7107.3 -r 5.6 kJ/mol for leonhardite,CaAlrSi4O,r.3.5HrO; -5964.3 + 5.1 kJ/mol for metaleonhardite,CaAlrSioO,r; -6646.7 + 6.3 kJ/mol for wairakite,CaAlrSi4O,r.2H,O; and -9051.3 + 10.4 kJ/mol for yugawaralite,CaAlrSiuO16.4H2O. The value for leonharditeis in good agreementwith early values from acid calorimetry (Barany 196l) but not with revised values from Hemingway and Robie (1977). The enthalpy of dehydration of leonhardite is 140.2 !6.7 kJlmol, and the loss of one mole of HrO is associatedwith an endothermic effectof about 40 kJ. Standardentropies, S!nr, of wairakite [400.7 J/(mol'K)] and yugawaralite [609.8 J/(mol'K)] were derived from our new enthalpydata combined with re- versedP-Zphase equilibria (Liou 197l;Zengand Liou 1982).The upperlimit of wairakite stability, the univariant curve for equilibrium of wairakite with anorthite, quartz, and fluid, was calculatedfrom thesevalues of enthalpy and entropy. Good agreementbetween thermodynamic calculations and reversed phase equilibria supports the reliability of the new thermodynamic data.

lNrnonucrroN lorimetric measurementof enthalpiesof formation of calcium zeolites and related minerals is desirable. Calcium zeolitesare common in diagenetic,sedimen- No calorimetric data exist for the enthalpiesand en- tary, hydrothermally altered, and low-grade metamor- tropies of laumontite, wairakite, and yugawaralite.The phic environments.The major zeolitesencountered are enthalpy of formation of leonhardite was obtained by laumontite (CaAlrSioO,2.4HzO), leonhardite (Ca- Barany(1961) using hydrofluoric-acid calorimetry and Alrsi4orr.3.5HrO),which is the parriallydehydrated form later revised by Hemingway and Robie (1977). Low- of laumontite, wairakite (CaAlrSioO,r.2HzO),which is temperature heat capacity and entropy of leonhardite the Ca analog of analcime, and yugawaralite (Ca- were measuredby King and Weller (1961). All pub- AlrSi6O,6).Laumontite phase equilibria are especially lished thermodynamic data for wairakite, laumontite, importantforP-Zboundariesofzeolitemetamorphicfa- and yugawaralite were estimated or calculated from cies(laumontite facies). Laumontite and wairakiteare the phaseequilibrium studies and are not consistent.For index mineralsof the zeolitefacies. instance,the standardentropy values for wairakite giv- P-TstabilityrelationsinthesystemCaAlrSirOr-SiOr- en by Helgesonet al. (1978) and by Glushko (1981) HrO have beeninvestigated both experimentallyand by differ by about 80 J/(mol'K). thermodynamic calculations(Coombs et al. 1959; Ko- Severalstudies (Kiseleva and Ogorodova 1983; Cir- izumi and Roy 1960;Liou 1970,l97l;Thompson 1970; coneand Navrotsky1992; Smelik etal. 1994;Navrotsky Zeng and Liou 1982;Senderov 1980, 1988;Ivanov and etal. 1994)have shownthat oxide melt solutioncalorim- Gurevich 1975). Despite extensiveexperimental work, etry can determinethe enthalpiesof formation of hydrous some of the zeolite phaserelations have not been satis- minerals.The presentwork usesdrop-solution calorim- factorilydeterminedbecauseoflowratesofreactionsun- etry in molten 2PbO'BrO3to measurethe enthalpiesof der low-temperature conditions and becauseof the for- formation of natural laumontite, metaleonhardite(dehy- mation of metastable phases in experiments of short drated laumontite), leonhardite,wairakite, and yugawaral- duration. Many experiments report only synthesis and ite. The enthalpy of dehydration in the laumontite-leon- not reversals;hence, estimations of free energiesare un- hardite-metaleonharditeseries was studied by transposed certain. Thermodynamic data obtained from experimen- temperature-drop calorimetry. Stability relations of wair- tal equilibria, estimated from P-T data in contemporary akite, yugawaralite, and leonhardite and the upper tem- geothermal systems, and calculated from various ther- perature limit for wairakite stability were calculated on modynamic data setsshow wide scatter.Thus, direct ca- the basisof thesethermochemical data. 0003-004x/96l0506-0658$05.00 658 KISELEVA ET AL.: THERMOCHEMISTRY OF CALCIUM ZEOLITES 659

Trau 1. Chemicalanalyses and formulasof zeolitesstudied TABLE2. Latticeparameters of zeolites

Leonhardite Wairakite Yugawaralite a (A) b (A) c (A) Bf) y(A.) Ref. Chemicalanalysis (wtc/o ) Leonhardite sio, 51.72 55.20 61.78 14.747(5) 13.067(3) 7.532(4) 1119(3) 1346(1) 1 Tio, 0.01 0.00 0.01 't4.770 13.056 7.595 112.8 2 A12o3 2193 23.08 17.01 14.77(2) 13.09 (2) 7.58 (2) 11 2.0 (1) MnO 009 0.00 0.00 Wairakite Mgo 0.04 0.01 0.01 CaO 12.14 12.48 o2R 13.700(15) 13.666(9) 13.559(10) 90.44(8) 882.9(8) 1 Naro 0.05 0.01 0.02 13.692(3) 13.643(3) 13.s60(3) 90.50(10) 4 KrO 013 0.02 0.01 Yugawaralite P.o. 0.11 (s) 14.007(6) 10.049(9) 11 1 .20 (4) 882.9(8) 1 13.67 11.98 6.728 H.O- 8.44 6.729 14.008 10.050 111.18 Total 99.98 99.23 100.25 Cations and HrO molecules per unit cell Note: Fot unit cells on the basis of 48 (leonhardite),96 (wairakite),and (yugawaralite)O atoms in the framework- Values in parenthesesrep- 15.s6 (1 6)-' 32.17(32) 12.08(12) 32 resent confidenceinterval (for 95% probability).References are as follows: Ti 0.00 0.00 0.00 1 : thiswork, 2 : Pipping(1 966), 3 : Yamazakiet al.(1991), 4 : Tak6uchi AI 7 s8 (8) 1s.86(1 6) 3.92(4) et al.(1979), and 5: Eberleinet al.(1971). Mn 0.02 0.00 0.00 Mg 0.02 0.01 0.00 4.02(4) 7.79(81 1.e6(2) Na 003 0.06 0.00 K 0.05 0.01 0.00 P 0.03 Latticeparameters of samplesstudied are shownin Ta- HrO 14.07(14) 16.41(16) 7.80(8) ble 2. They agreewell with previous studies(Gottardi and 'HrO contentsalso verifiedby thermogravimetricanalysis. Galli 1985). .'Values in parenthesesrepresent ideal molar ratios. CllonrPrnrnY The enthalpiesof formation and dehydrationwere de- Znor-ttn CHARACTERrzATroN termined usinga Tian-Calvethigh-temperature heat-flux We chosewell-crystallized natural samplesof laumon- microcalorimeter describedin detail by Navrotsky (1977). tite, wairakite, and yugawaralite.The laumontite is from Drop-solution calorimetric methods (Chai and Navrot- Nidym River (left tributary of Nizhnyaya Tunguska Riv- sky 1993; Navrotsky er al. 1994) were chosento avoid er), Siberia. White to pinkish nontransparent crystals up decompositionof zeolitesat the calorimetertemperature to l0 mm in length were separatedfrom hydrothermal prior to dissolution.The sampleswere dropped from room veins in Triassicbasalts. temperatureinto molten 2PbO'BrO3at 975 K. Most ca- Wairakite from Bondai-Atami, Koriyama City, Japan lorimetric experimentswere performed using pressed pel- (MineralogicalMuseum of the RussianAcademy of Sci- lets about 3 mm in diameter,0.5-l mm in height, and ences,catalogr,o.81622), was provided by A.A. Godo- 10-25 mg in weight. A few experimentsused piecesof vikov. It had been separatedunder a microscopefrom an single crystals. The heat of drop solution was a sum of aggregateof white semitransparentcrystals with some ad- the heat of solution in the melt plus the heat content mixture of quartz and calcite.Clean, transparent, color- (H8,' - Hln)- less tabular crystals were used. Heats ofdehydration and heat contentsat 975 K were Yugawaralite from Obora Toi-cho, Shizuoko Prefec- obtained using transposedtemperature-drop calorimetry ture,Japan (Mineralogical Museum of the RussianAcad- (sampledropped into an empty platinum crucibleequil- emy of Sciences,catalog no. 87616), was provided by ibrated in the calorimeter). The heat effectcontained two A.A. Godovikov. Transparentand semitransparentcol- contributions,the enthalpyofdehydration at 975 K and orlessplaty crystalsup to l-2 mm in dimension from the heat content of the mineral (H\r' - 11!*). When the the centralparts of the small veins (10-15 mm thick) in solid decompositionproducts were droppedinto the cal- hydrothermally altered brecciated volcanic rock were orimeter, the heat effectcontained only the heat content chosen. (H8,, - .F{rr) of the dehydrated zeolite. Leonhardite was chemically analyzed by X-ray fluo- We were concernedthat when the small samplesused rescenceusing a Carl Zeiss VRA 20R instrument. Wair- in this study were dropped into the calorimeter, the heat akite and yugawaralite were analyzed by electron micro- picked up during the drop, before the sample reachedthe probe (Camebax MICRO). Table I shows the analytical region of the thermopile, might be a different fraction of daIa. the total heat effectthan seenfor more massiveplatinum- High-temperature thermal behavior was studied using encapsulatedsamples, for 30-50 mg pellets,or for plat- a TA 3000 (Mettler) thermoanalyzerconsisting of a ther- inum calibration piecesweighing about 200 mg. Rather mogravimeter(293-1213 K), a scanningcalorimeter (103- than using such large platinum pieces for calibration in 873 K), and a differentialdilatometer (173-1273 K). The the usual procedure (Navrotsky 1977), we chose to cali- thermograms are similar to those published by Gottardi brate using corundum pelletsof weight similar to that of and Galli (1985)and are discussedfurther below. the samples.The corundum (JohnsonMatthey 99.999o/o) 660 KISELEVA ET AL.: THERMOCHEMISTRY OF CALCIUM ZEOLITES

TaeLe3. Resultsof calorimetricexperiments at 975 K

Observedenthalpy Drop-- Dropsolt Mineral (J/s) (kJ/mol) (J/s) (kJ/mol)+ Laumontite CaAlrSi4Ol,.4H2O 470.44 1502.8+ 13.2(9) 707.0 + 6 2(9) Leonhardite CaAlrSi401r.3.5HrO 461.44 1450.6+ 13.1(6) 669.4 + 6.0(6) 1390.8t 6.4(9) 641.8+ 3.0(9)$ Dehydrated leonhardite CaAlrSi4OP 398.39 722.3 + 7.0(7) 287.8 + 2.8(7\ 646.9 + 4.9(6) 257.7+ 2.0(6)$ (metaleonhardite) Wairakite CaAlrSi4O,r.2HrO 434.41 1171.3+ 9.7(10) (508.8+ 4.2(10)$) 506.4+ 4.2(10)ll Yugawaralite CaAlrSi6O16.4HrO 590.62 1235.1+ 13.5(7) (729.5+ 8.0(7)$) 734.2+ 8.O(7lll Corundum A12o3 101.96 1058.6+ 9.6(8) 107.9+ 1.0(8)$ Quartz sio, 60.08 715.0+ 5.3(12) 43.0+ 0.3(12) 551.3+ 5.2(9) 39.1+ 0.3(9)$ Calcite CaCO3 100.09 1931.8+ 6.8(10) 193.4+ 0.7(10)$ . F.W. : gram formulaweight " Drop : transposed temperature-dropcalorimetry, no solvent present. f Dropsol: drop solutioninto 2PbO.BrOosolvent. + Uncertainty is two standard deviations of mean; number in parenthesesis number of experiments. $ Assumingideal stoichiometry. I Corrected for nonideal H2Ocontent; this is the preferredvalue.

was heatedat 1773 K for 15 h to ensuredryness, crys- sphereat 975 K. The calcite(Aldrich, 99.995o/o) and quartz tallinity, and completeconversion to a-AlrOr. The cali- (Fluka, >99.9o/o)were commercial products, dried at 383 bration factor obtained (using H) - HBn"from the Na- K prior to calorimetry. tional Bureau of Standards Certificate, 1982) for corundum was indeed about 2o/ohigher than that obtained for the samecalorimeter using largeplatinum piec- Cr,omvmrRrc REsLJLTS es. We believe that this modified calibration method bet- Calorimetric data are summarized in Table 3. Six to ter reflectsthe conditions usedin the presentexperiments. 12 enthalpy measurementswere made for each sub- For the calibrations,an uncertaintyof about +0.50/o(two stance.The calibration factor did not dependon the mass standard deviations of the mean) was obtained. of the alumina dropped,which rangedfrom l0 to 25 mg, Becauselaumontite is not stable at room temperature and neither the heat of drop solution nor the heat of in air, it was obtained from leonhardite by keeping the transposed temperature drop (in joules per gram or ki- sample in water-saturatedconditions for l0-16 h. Lau- lojoulesper mole) varied with samplemass. This consis- montite with a little excesswater (not more than 6 x tency, see Figure l, indicates proper equipment perfor- l0-s mole HrO per mole laumontite) was wrappedin a manceand reproduciblesample dissolution. The heatsof small platinum capsuleand dropped from room temper- drop solution did not dependon the amount of sample ature into an empty platinum crucible in the calorimeter. previously dissolved(up to about 150 mg in 30 g lead The measured heat effect contained four contributions: borate), and the calorimetric curves returned cleanly to the enthalpy of dehydration aI 975 K, the heat content the original baseline in almost all experiments. The sta- of the mineral, the platinum capsuleheat content, and tistical errors reported (two standard deviations of the the excesswater heat content. Using platinum capsules mean) are + lo/oor better, which also indicates problem- introduced the need for additional calibration by drop- free dissolution. ping empty platinum capsules.The platinum heat con- Table 4 comparesthe enthalpiesof drop solution, so- tent was l0-l2o/o of the total measuredheat effect,and lution, and transposedtemperature drop obtained for co- the excesswater heat effect was about 2o/o. rundum, quartz, and calcite in this work with valuesfrom All experiments were performed under a dry-air at- previous studies at the same temperature. In general the mosphere flowing through the calorimeter at 30-40 cm3/ agreement is good. Our value for the enthalpy of drop min. Navrotsky et al. (1994)showed that calorimetryun- solution ofcalcite is slightlyhigher than that ofChai and der flowing gas results in a negligible enthalpy of inter- Navrotsky (1993);this may reflectthe differencebetween action betweenthe evolved HrO and the solvent.Under using calibration factors based on small alumina pellets flowing gas conditions we have been able to obtain re- and large platinum pieces. However, similar differences producible results with excellent baseline stability. are not seenfor quartz and corundum. We note that both CaCOr, rather than CaO, was used as the Ca reference solution calorimetry and drop-solution calorimetry on material becausepellets of CaO did not dissolverepro- quarlz and corundum are frequently performed in our ducibly and showed evidence of partial hydration. Chai laboratory to train new users and check calorimeter op- and Navrotsky (1993) showedthat CaCO, can be used eration; this accountsfor the large amount of unpublished reliably for calorimetric purposes under flowing atmo- data in Table 4. Becausethe enthalpy of solution of quartz KISELEVA ET AL.: THERMOCHEMISTRY OF CALCIUM ZEOLITES 66r

1.6 Trau 4, Comparisonof enthalpiesof dropsolution of quartz, corundum,and calcite obtained at varioustimes in thislaboratory o Leonhardite - ^ - 1.4 Transposed c temperature .9 _ Wairaklte a droP A^ (heatcontent) A Drop solution Solution E 1.2 + + a Corundum,c-AlrO. CL Yugawaralite + + o 1079 + 1.0" 33.5 0.8b 74.3 O.4. + + 74.2+ 0.8. lt 107.4 2.1. 33 0 0.8d 1.0 107.9 + 0.71 32.8 + 0.3s 75.3h o (33.6 + 1.0)i

-cL Quartz, SiO" E o.B 39.t + 0.3" -4.3 + 0.2. 43.0+ 0.3. 40.0 + 0.21 -4.3 + 0.21 42.31 C -3.9 ul Dehydrated Leonhardite 39.3 + 0.4" + 0.4s 44.01 45.1. Calcite, CaCOr 10 15 20 25 193.4+ 0.7" mg sample 189.9+ 0.7r Frcunr 1. Observedenthalpy of drop solutionvs. massof Note: a : this work, aluminacalibration; b : Brown and Navrotsky (1993' for leonhardite,wairakite, and yugawaralite. (1989,unpublished), Pt calibration;c: Tinkerand Navrotsky un- sampleused published),Pt calibration;d : Hovis and Navrotsky(1994, unpublished), Pt calibration;e : Petrovicand Navrotsky(1994, unpublished), Pt calibration;| : Gerardinet al.(1994), Pt calibration;g : calculatedas enthalpy of drop solutionminus enthalpy of transposedtemperature drop, this work; h : Robieet al. (1978),values for corundumsimilar in other tabulations; : j : only data i Ellisonand Navrotsky(1992), Pt calibration; chai and Navrotsky appearsto depend significantly on temperature, (1993),Pt calibration;k : Berman(1988), Holland and Powell(1990); | : at 973-978 K are included in Table 4. Lastly, we note Richetet al. (1982);and m : Hemingway(1987). considerable scatter in the tabulated heat contents of quafiz; this largely reflects the different ways of treating the energeticsofthe a-B transition. thalpy changeis 3.1 kJ/mol, in excellentagreement with ENrnll,ptBs oF DEHYDRATIoNoF experimental data. Thus, our experiments show that the LEONHARDITEAND LAUMONTITE loss of HrO from laumontite at room temperatureis in- The enthalpy of dehydration of leonhardite may be cal- deed associatedwith a very small endothermic effect of culated (seethermodynamic cycle A in the Appendix Ta- about 6 kJ/mol HrO lost. In contrast, the dehydration of ble). For the reaction leonhardite is associatedwith more significant endothermic effects(about 40 kJlmol HrO). CaAlrSioO,''3.5HrO(xl, 298 K) The enthalpy of dehydration of the sodium zeolite : (1,298 CaAI'SLO',(x1,298K) + 3.5H,O K), analcime was measured by acid calorimetry by Barany (l) LH: 140.2+ 6.7kJ/mol. (1961), who obtained about 30 kJ/mol HrO, similar to The enthalpy of dehydration of laumontite (see cycle our value for leonhardite. However, such comparisons B) is given by can be misleading. Dehydration is a very complicated processthat can occur continuously or stepwiseand may (xl, 298) CaAlrSioO,r'4HrO(xl, 298): CaAlrSioO', be accompaniedby changesin the aluminosilicate frame- : + (2) + 4H,O (1, 298),LH 143.4 6.8 kJlmol. work. For example,as shown by Belitsky et al- (1992' The enthalpy of partial dehydration of laumontite to 1993), dehydration can be followed by compressionof leonhardite is representedby the crystal structure owing to framework polyhedral tilt- ing. Different zeolite frameworks deform differently and : ' CaAlrSioO,,'4HrO(xl, 298 K) CaAIrSioO,u have different positions of the HrO moleculeswithin the (3) 3.5H,O (xl, 298 K) + 0.5H,O (1, 298 K). channels. Thus, the dehydration processrelated to such Its enthalpy(see cycle C) is 3.2 + 4.6 kJlmol. This value structural changesmay be accompaniedby different en- is essentiallyzero. ergy effects. Becausethis reversible transition takes place in air at Yakubovich and Simonov (1985) refined the crystal near room temperature,Crawford and Fyfe (1965) sug- structure of leonhardite, giving it the structural formula gestedthat the free-energychange of this reaction is ex- Ca(HrO)r,(AlrSioO rr)' 0.5HrO to representbetter the dif- tremely small, not exceedinga few hundredjoules. If we ferent roles of HrO in the crystal structure.Four positions assumethat AGo for the laumontite-leonhardite reaction were occupied by HrO molecules. Our differential scan- is zero, we can calculateits enthalpy changeusing entropy ning calorimetry (DSC) and differential thermogravime- values from Helgesonet al. (1978). The estimateden- try curves show three peaks,but the third peak is asym- 662 KISELEVA ET AL.: THERMOCHEMISTRY OF CALCIUM ZEOLITES

metric and its high-temperature shoulder may be related the first cycle requires fewer steps,it yields a value with to removal of the most tightly bonded HrO molecules. smaller uncertainty and we adopt this as our preferred Using DSC we measured the total heat effectsassoci- value. The agreementbetween the first and secondcycles ated with dehydration of leonhardite, CaAlrSioO,r. indicatesthe overall consistencyofthe data. 3.5HrO, from 298 to 863 K to be about 139 kJ/mol, Per mole HrO, the differencebetween the enthalpiesof which agreeswith the valuemeasured by transposedtem- formation obtainedfrom the first and secondcycles is 0.7 perature calorimetry (143 kJ/mol). DSC measuremenrs + 2.3 kJ/mol. Becausethe secondcycle does not useany show the disproportionate distribution ofenergy into sev- reactionsthat dissolve a hydrous phasein lead borate, eral peaks.The first peak reflectsthe loss ofweakly bond- whereas the first cycle does, their good agreementonce ed zeolitic water without changesin the Al-Si framework more validates the methodologyof drop-solution calo- and has a small heat effect correspondingto about 6-80lo rimetry under flowing atmosphere for volatile-bearing of the total heat of dehydration.The secondpeak repre- phases(Navrotsky et al. 1994). sentsremoval of HrO moleculesbonded to both frame- The standardenthalpy of formation of leonharditefrom work O atoms and Ca and comprises about 200/oof the oxidesobtained in this work is -153.3 t 3.5 kJ/mol. total heat effect.The third peakis relatedto lossofwater This agreeswithin experimental error with the value of from the Ca coordination sphereand the formation of a -154.2 + 2.7kJ/mol (Barany1961) previously obtained metastable dehydrated phase, accornpanied by changes from solution calorimetry in aqueous hydrofluoric acid in the framework structure. It accountsfor 70-800/oof the for a natural sample with slightly higher (by 0.330/o)water total energy.The loss of one mole of HrO during the first content.Later, Hemingwayand Robie (1977)recalculat- and the second peaks has an endothermic efect about ed Barany's data. They considered that the enthalpy of l0-20 kJlmol, but removing HrO moleculesfrom the Ca formation of gibbsite,A(OH)3, obtainedby Barany and coordination sphererequires about 70-80 kJlmol. Kelley (1961)was incorrectand that the heat of solution The enthalpy of dehydration of yugawaralitewas eval- value used for a-quartzrefers to very small particles.They uatedonly by DSC to be about 150 kJ/mol CaAlrSiuO,u. revised Barany's data and recommended a value of 4HrO, or about 37.5 kJlmol H,O. Kerr and Williams -169.2 + 3.5 kJlmol for the enthalpyof formation. Our (1969) found four water sites,each with 750looccupancy new determination supports the original rather than the and each linked to a Ca site. All hydrogen bonds were revisedHF value. described as long or very long. The DSC curve shows For the formation of metaleonhardite(fullv dehydrated three main peaks, but the third peak, as for leonhardite, leonhardite), has an asymmetric high-temperature shoulder. The first CaO + AlrO3 + 4SiO, : CaAIrSioO,r, peak (about l0-150/oof the total heat effect)presumably AH: - 10.7+ 2.8kJlmol (5) correspondsto loss of weakly bonded HrO moleculeswith the longesthydrogen bonds. The lossofone mole ofsuch (see cycle F). Thus, the dehydratedzeolite has only a HrO requires5-20 kJ. The secondpeak correspondsto rather modest energeticstability with respectto the ox- l5-20o/oofthe total heateffect with an endothermiceffect ides and is metastablewith respectto other anhydrous of about 10-20 kJ/mol HrO. The third peak (about 700/o phaseassemblages. For the reaction(see cycle G) of the total heat effect) probably correspondsto the re- CaAlrSioO,, (metaleonhardite) moval of HrO moleculeswith shorter hydrogenbonds. : CaAlrSirO, (anorthite)+ 2SiO, (quartz), This is associated with a largerendothermic effect of about -78.7 70 kJlmol. AH: kJ/mol. (6) On the basisof two moles of O (Ca",Aly,Siu,Or),meta- ENrg.ql,prns oF FoRMATToN oF cAr,cruM zEoLrrES leonharditeis only l3.l kJ/mol higherenergetically with respect to its stable phase assemblageof anorthite + Leonhardite and metaleonhardite quartz. Petrovic et al. (1993) found that a seriesof syn- The enthalpy of formation of leonhardite from the ox- thetic zeolitesof SiO, composition were l0-15 kJ/mol ides at 298 K was calculatedusing two thermochemical higher in energythan quartz, and Hu et al. (1995)found cycles for the reaction that severalaluminophosphate zeolites (AlorPorOr) were CaO + AlrO3 + 4SiO, + 3.5HrO only 7-10 kJlmol higher in energythan berlinite. Thus, the present data support the observation that anhydrous : CaAlrSioO,2.3.5H2O. (4) open zeolite frameworksare only slightly destabilizedwith One usesthe enthalpy ofdrop solution ofleonhardite and respect to their denseanhydrous stable assemblages. of constituentoxides (see cycle D). The secondcycle (E) The enthalpies of formation from the elements were is used to derive the enthalpy of formation of leonhardite calculated with the use of the thermodynamic values of and metaleonhardite.It involves dehydration and sub- the constituentoxides taken from Robie et al. (1978)and sequentdissolution of fully dehydratedsamples. the enthalpies of formation from the oxides obtained in The enthalpy of formation of leonhardite from the ox- rhis srudy: aH, at 298 K: -7107.3 + 5.6 kJ/mol for idesat 298 K is - 153.3+ 3.5 kJlmol usingthe first cycle leonhardite(CaAlrSioO,r.3.sHrO), and -5964.3 + 5.1 and - 150.9+ 7.2kJ/mol usingthe secondcycle. Because kJ/mol for metaleonhardite(CaAlrSioO,r). KISELEVA ET AL.:THERMOCHEMISTRY OF CALCIUM ZEOLITES 663

TABLE5. Thermodynamicproperties at 298 K of some Ca-zeolitesobtained from this study and from the literature

Mineraland LHI formula (kJ/mol) Ref. (methods) U/(molK)l Ref. (methods) Laumontite(CaAtSLO,,. 4H,O) -7251.0+ 8.5', 1,C 485.3+ 20.9 2,E -7268.9+ 6.3 2, PE 485.76 3,E -7231.3a 10.1 4, PE 499.2! 19.2 5,E -7233.6 3, PE -7265.6 6,E Leonhardite(CaAlrSi4O,, 3.5HrO) -7107.3+ 5.6 461.1+ 5.45 -7123.2+ 4.8 8, PE -7108.3+ 5.2 I,C Metaleonhardite(CaAlrSi4Olr) -5964.3+ 5.1 1,C 255.0 1,E Wairakite(CaAlrSioO," 2HrO) -6646.7+ 6.3 1,C 400.7 1, PE,C -6678.9r 5.0 2,PE 376.6+ 20.9 2,PE -6606.9+ 8.1 4, PE 439.7 2Dtr -6608.8 2Dtr 463.6! 4.2 5,E Yugawaralite(CaAlrsi6Or6 4HrO) _9051.3+ 10.4 1,C 609.8 1,C, PE -9036.19 10,PE 609.9 10,PE -9087.6 6,E

/Vote:Referencesareasfollows:1:thisstudy,2:clushko(1981),3:Helgesonetal.(1978),4:Senderov(1980),5:zen(1972)'6:Chermak and Rimstidt(1989), 7 : King and Weller(1961), 8 : Hemingwayand Robie(1977), I : Barany(1961), and 10: Zengand Liou(1982)' Methods are as follows:C : calorimetry,PE : phaseequilibrium, and E : estimation. 'Enthalpy of formationfrom elements.

Laumontite we obtainedAH|..: -123.5 t 4.7 kJ/mol at 298 K cycle The enthalpy of formation Calorimetric study of laumontite under excesswater using thermodynamic J. is -6646.7 + 6.3 kJlmol. allows us to calculate the enthalpy of formation of lau- from the elements,AFlfl"', water,A1{", would be -6649-2 + montite using two cycles.First (cycle H), we can use the Without correctionfor for A1{"' of wairakite calculat- enthalpy of the reaction leonhardite : laumontite, the 6.3 kJ/mol. Previous data phase or otherwise estimated differ enthalpy of formation of leonharditeobtained in this work, ed from equilibria ranging from -6606.9 + and the data for HrO from Robie et al. (1978).The en- significantly from each other, (Senderov to -6678 + 5'0 kJ/mol thalpy of formation from the elements of laumontite 5.0 kJlmol 1980) (CaAlrSioO,r.4HrO)is -7253.4 + 10.3kJlmol. The sec- (Glushko 198l). Our value, the first direct calorimetric middle of this range (see ond cycle(I) usesthe completedehydration of laumontite measurement, is close to the solid solution existsbe- to metaleonhardite, the enthalpy of formation of meta- Table 5). It is worth noting that analcime, and that leonhardite, and data for H'O. The calculated enthalpy tween wairakite and its Na analog, structure, thermody- of formationfrom theelements is -7251.0 18.5 kJlmol. this substitution influences crystal phase (Aoki and Minato This value is close to that obtained using the flrst cycle, namic properties,and equilibria but the second cycle yields a value with smaller uncer- l 980). tainty and we adopt this as our preferred value (Table 4). Our value lies betweenpreviously reported data basedon Yugawaralite phaseequilibria, -7233.6 kJ/mol (Helgesonet al. 1978) The yugawaralitestudied is nearly stoichiometric and -7268.9 + 6.3 kJ/mol (Glushko1981). (Eberleinet al. l97l Gottardi and Galli 1985),but the HrO contentis about 0. I mole lessthan theoretical.The Wairakite experimental value for the enthalpy of drop solution was The wairakite used has a nearly stoichiometric com- correctedby addingthe heatcontent of 0. I molesof HrO position, exceptfor a slightly higher water content (0.150/0), (Table 3). The thermodynamiccycle K was used to cal- as is very common for natural samplesof wairakite(Got- culatethe enthalpy of formation from the oxidesat 298 tardi and Galli 1985).The very small amountsof Na, K, K. For the reaction and Mg have a negligible effect on the enthalpy, but we made a correction to the enthalpy of drop solution for CaO + AlrO3 + 6SiO, + 4H2O: CaAlrSiuO'u'4HrO (8) 0.05 mole of excessHrO. We subtractedthe heat content of 0.41 moles of HrO from the measuredheat effectand Af{.. : -133.0 + 8.4 kJ/mol. The enthalpy of forma- obtainedthe heat of drop solutionof stoichiometricwair- tion from the elementsis -9051.3 t 10.4kJ/mol. With- akite, 506.4 + 4.2W/mol (Table3). This preferredvalue out correction for the missing HrO, AF18.rwould be was used to calculate the enthalpy of formation. For the -9046.5 + 10.4 kJ/mol. Table 5 shows that only two reactron estimatesof A119are available in the literature:- 9036.l9 CaO + AlrO3 + 4SiO, + 2H2O: CaAlrSioO,r.2HrO kJlmol (ZengandLiou 1982)calculated from phaseequi- (7\ libria involving somewhat uncertain data for wairakite oozt KISELEVA ET AL.: THERMOCHEMISTRY OF CALCIUM ZEOLITES

sition of laumontite to wairakite is the most suitablere- action for calculatingthe entropy of wairakite:

) laumontite: wairakite + 2H,O. (9) o Lau m ontlte (E 2000 a lt Becausethe breakdownproduct of laumontite may be I (b) either orderedor disorderedwairakite dependingon the o length of the experiments,we did not consider studies 5 Anorthlle o that had no proof of reactionreversal and for which the 3t, o 1000 Walrak I a products may have representedvarious intermediate o stagesof this reaction. The equilibria demonstratedby Yug awa ra llte Liou (1971)appear reliable. From thesereversals, we cal- culatedthe entropy ofwairakite using

0 - 400 500 600 700 AG,_r: LH? rA^S9+ AV?P+ Rl" ln /",o : 0 (10) Temperature(K) where AGr-., A1{, and ASf are the total free energy,en- Frcunp 2. P-Z diagram, showing the relations among zeo- thalpy, and entropy changeof the reaction,respectively, lites in the presenceofexcess quartz and HrO fluid. The curves volume phases marked a, b, and c represent the laumontite + yugawaralite, A 4 is the changeamong the solid of the yugawaralite+ wairakite, and laumontite + wairakite equilibria, reaction, atd fH,o is the fugacityof HrO (Burnham et al. respectively,calculated from thermochemical data and in agree- 1969). Becauseno experimentalheat-capacity data are ment (to +5 K) with curvescalculatedby Zengand Liou (1982). available for laumontite and wairakite we assume as a The metastableextension of the yugawaralite + wairakite reac- first approximationthat AC, for the reactionis 0 and that tion, reversedby Zeng and Liou (1982),is shown as a dashed AZof the solid phasesis constant.These assumptions are curve. Curve d representsthe breakdown of wairakite to anor- reasonableat the low-P and low-Z conditions in these thite as calculated from thermochemical data. The four solid studies.Using the new calorimetric data for the enthal- represent circles the reversalsofthat equilibriumby Liou (1970). pies of formation of laumontite and wairakite we calcu- The error in thesepoints is approximately t5 K +20 and bars. lated the enthalpy of the laumontite-wairakiteequilibri- um. We usedthe value for entropy of laumontite of 485.76 J/(mol.K) from Helgesonet al. (1978).Using the molar volumes of laumontite and wairakite from Helgesonet al. (1978)we calculatedAV": -2.07 The resulting from Helgesonet al. (1978),and a significantlymore exo- J/bar. wairakite 400.7 thermicvalue estimated by Chermakand Rimstidt (1989). standardentropy of is J/(mol'K). This is (Ta- Our value is closer to the value calculated from phase in the midrange of estimatesfrom different datasets equilibria. ble 5). The entropy ofyugawaralite at 298 K can be derived analogously.At presentthere is only one experimental ClLcuLArroN oF THE ENTRopTESoF study (Zengand Liou 1982)ofthe univariant equilibrium WAIRAKITE AND YUGAWARALITE curve for the dehydration ofyugawaralite: Low-temperature heat capacity and entropy at 298 K have been measuredonly for leonhardite (King and Wel- yugawaralite : wairakite -t 2 quarlz + 2HrO. (l l) ler l96l). There are many estimatesof the entropiesof the other mineralsin the various datasets(Table 5). Even though this reaction was studied in a metastable Becauseleonhardite is related to laumontite, the stan- field (see Fig. 2), the steep slope of the P-T curve was dard entropy of laumontite has been estimated rather establishedreliably. We obtained a value for the standard consistentlyby different authorsusing the leonharditedata entropyof yugawaraliteof 609.8J/(mol.K) (Table5). Our (see Table 5). In contrast, estimatesof the entropy of value for S0of yugawaraliteis in excellentagreement with wairakite from phaseequilibna scatterfrom 376.6 ! 20.9 the value calculatedby Zeng and Liou (1982). Ji(mol. K) (Glushko I 98 I ) to 463.6+ 4.2 J/ (mol. K) (Zen The thermochemical calculationsgive stability fields of 1972). The standard entropy of yugawaralite was calcu- laumontite, wairakite, and yugawaralite that agree with latedin only one study (Zengand Liou I 982) on the basis reversedphase equilibria (Liou 197l; Zntg and Liou 1982) of a value for the entropy of wairakite from Helgesonet (seeFig. 2). Becausethe entropiesused were based in part al. (1978). on these reversals,such agreementis expected.Never- The standard entropies of wairakite and yugawaralite theless, the simultaneous good fit to several equilibria were derived from our new, independently determined using the calorimetric results strongly suggestsinternal enthalpy data, in combination with experimentallyre- consistencyin both the calorimetric and phase-equilib- versedP-7 phaseequilibria. The equilibrium decompo- rium studies. KISELEVA ET AL.: THERMOCHEMISTRY OF CALCIUM ZEOLITES 665

Recommendedvalues for heatsand entropiesof for- Taer-e6. Recommendedvalues of thermodynamicproperties mation, and for standard entropies based on the above of zeolitesobtained in this studv in Table for zeolites considerations, are listed 6 all the Mineraland AHfl".",u S9".,u AGl., ,u studied. formula (kJ/mol) [J/(mol.K)] (kJ/mol) Laumontite Gnor,ocrc APPLrcATloNs (CaAlrSi,O,,4HrO) -7251 0 + 8.5 485 5' -6698.9 Several reactions constrain minerals in the zeolite fa- Leonhardite (CaAlrSinO,r'3.5HrO) -7107.3+ 5.6 461.1" -6582.7 cies;one such reactionthat determinesthe upper limits Metaleonhardite of wairakite stability is (CaAlrSi4O,r) -5964.3 + 5.1 255.0 -5621.6 Wairakite -bb40./ -6208.3 wairakite: anorthite * 2 quartz + 2HrO. (12) (CaAlrSioO,,2H.O) 10.J 400.7 Yugawaralite Coombs et al. (1959) reported that the equilibrium (CaALSIO16.4HrO) -9051.3+ 10.4 609.8 -8402.6 temperaturefor this reaction at 2 kbar is near 623 K. - Helgesonet al. (1978). Koizumi and Roy (1960) suggestedthat the maximum "" King and Weller(1961 ) temperaturefor this reactionis 733 K at I kbar and 143 K at 2 kbar. These studieswere synthesisexperiments rather than reversals.Seki (1968)noted that the synthetic was modifi- wairakite used in these studies a tetragonal montite-wairakite-yugawaralite phase diagram, with an cation, disorderedwairakite. The equilibrium dehydra- invariant point (in the presenceofexcess quartz and wa- tion of wairakite to anorthite, quartz, and water (Liou ter) near 500 bars and 500 K. 1970)(see Fig. 2) yieldedtemperatures of the breakdown These thermochemical data can be used to calculate curve of wairakite much lower than the earlier data. other equilibria involving these minerals. P.o, and P",o Using our new enthalpy and entropy values,we cal- may significantly affect the stability of calcium zeolites. culatedthe univariant curve of the equilibrium of wair- These minerals are unstable in the presenceof even a akite with anorthite, quartz and fluid. The Gibbs free moderate amount of CO, becausethe partial pressureof energy of reaction at P and I was calculated as above, CO, lowers the fugacity of water and becauseCa is con- and we assumedconstant AC" and AZ.. We usedvalues sumed in carbonates.Such conditions are rather common for the molar volume, standardentropy, and enthalpyof for the calcium zeolite assemblagesin uranium deposits quartz HrO from Robie et al. (1978), volume and and (Mironenko and Naumov 1982).Thus, P-Z conditions entropy of anorthite from Robie et al. (1978), the en- of calcium zeolite formation strongly depend on the geo- thalpy ofsynthetic anorthite from a recentpublication of thermal gradient, PH,o/Pro,P.o,, artd other factors, in- Zhu et al. ( I 994)and the volume of wairakitefrom Helge- cluding solution composition (Na,K). With appropriate son et al. (1978).Thermodynamic propertiesof HrO at care,the thermochemical data produced by this study are P and T were taken from Burnham et al. (1969). useful in understanding zeolite stability in complex nat- The calculated breakdown curve of wairakite (Fig. 2) ural systems. Equally important, a knowledge of ther- agreesreasonably with Liou's (1970) phaseequilibria. The modynamic stability helps distinguish equilibrium from temperatureofthe calculatedunivariant curve at 500 bars nonequilibrium conditions, a constant problem in low- is near the experimentalvalue, but at 1000-2000 bars temperatureenvironments. calculatedtemperatures are slightly lower than experimental temperatures.At a given pressure,the calculated AcxNowr,nncMENTs temperaturecan changeby as much as 50 K depending work was supportedby the U S. Department ofEnergy (grant DE- on the values of All, AT and A,V chosen within their This FG02-85ER 13437)and the RFFI ofthe AcademyofSciences, Russia. allowableranges. The assumptionsof negligibleAC, and We thank A. Godovikov for providing samples.We thank L. Topor, L. constantAV,add other uncertainties.However, these fac- Chai, E. Smelik, and K. Bose for advice and help with experimentalpro- tors probably do not vary independently,and a rigorous cedures,and S. Wasylenko and K. Rozenbergfor help with manuscript assessmentof uncertaintiesis impossible.We consider preparatlon the agreementwith experimentalvalues encouraging, es- RrrnnrNcns crrED pecially becausethe wairakite breakdown reaction was not usedto constrainthe thermochemicaldata. Thus, this Aoki, M., and Minato, H. (1980) Lattice constantsof wairakite as a func- thermochemicalcalculation of the upperthermal stability tion of chemicalcomposition. American Mineralogist,65, 1212-1216. Barany, R. (1961) Heats and free energiesofformation ofsome hydrated limit of wairakite supportsthe consistencyand reliability and anhydrous sodium- and calcium-aluminum silicates.U.S. Bureau of our new calorimetricdata. The calculatedP-Z field of of Mines.5900.l-18. wairakite agrees with field observations in low-grade Barany, R., and Kelley, K.K (1961) Heats of formation of gibbsite, ka- metamorphicand hydrothermallyaltered rocks (Coombs olinite, halloysite,and dickite. U.S. BureauofMines Report Inv. 5825, et al. 1959).The thermochemicaldata confirm that wair- l-lJ. Belitsky, I.A., Fursenko, B.A., Gabuda, S.P., Kholdeev, O.V., and Sery- akite requireshigher temperatures for its formation than otkin, Y.V. (1992) Structural transformations in natrolite and eding- laumontite. They also support the topology of the lau- tonite. Physicsand ChemistryofMinerals, 18, 497-505. 666 KISELEVA ET AL.: THERMOCHEMISTRY OF CALCIUM ZEOLITES

Belitsky, I.A., Fursenko, B.A , Seryotkin Yu A., Goryainov, S.V., and Koizumi, M., and Roy, R. (1960) Zeolite studies: l. Synthesisand sta- Drebushchak, V.A. (l 993) High-pressure,high-temperature behavior bility ofthe calciumzeolites. Joumal ofGeology,68, 4l-53. ofwairakite. In Zeolite 1993 4th International Conferenceon the Oc- Liou, J.G. (1970) Synthesis and stability relations of wairakite Ca- currence,Properties, and Utilization ofNatural Zeolites,Boise, Idaho, Al,SioO,, .2HrO. Contributions to Mineralogy and Petrology,27 , 259- June20-28, 1993,Program and Abstracts,45-48. 282. Berman, R G (1988) Internally-consistentthermodynamic data for min- - (1971) P-?nstabilities oflaumontite, wairakite, lawsonite, and re- eralsin NarO-KrO-CaO-MgO-FeO-FerO,-AlrOr-SiOr-TiOr-HrO-CO2. lated minerals in the system CaAlrSirOE-SiOz-HrO.Journal of Petrol- Journal of Petrology, 29, 445-522 ogy,12,379-4ll Bumham, C.W., Holloway, J R., and Davis, N.F. (1969)Thermodynamic Mironenko, M.V., and Naumov, G.B. (1982) Physicaland chemical con- properties of water to 1000 'C and 10,000 bars. Geological Society ditions of laumontite formation in COr-hydrothermal systems.Geo- America. 132. l-96. khimia, 4, 18-22 (in Russian) Chai, L., and Navrotsky, A. (1993) Thermochemistry of carbonate-py- Navrotsky, 4,. (1977) Progressand new directions in high temperature roxeneequilibria. Contributionsto Mineralogyand Petrology,ll4, 139- calorimetry. Physicsand Chemistry of Minerals, 2, 89-104. 147. Nawotsky, A., Rapp, R.P., Smelik, E., Burnley, P., Circone, S., Chai, L., Chermak, J.A , and Rimstidt, J.D. (1989) Esrimating the thermodynamic Bose,K., and Westrich, H.R. (1994) The behavior of HrO and CO, in properties (AGPand Alf) of silicate minerals at 298 K from the sum high-temperaturelead borate solution calorimetry of volatile-bearing of polyhedral contributions. American Mineralogist, 74, 1023-l}3l. phases.American Mineralogist, 79, 1099- I 109. Circone, S , and Navrotsky, A. (1992) Substitution of t6,4rAlin phologo- Petrovic,I, Navrotsky, A, Davis, M.E., and Zones,S.I. (1993) Ther- pite: High+emperature solution calorimetry, heat capacities,and thermochemical study of the stability of frameworks in high silica zeolites. modynamic properties of the phologopite-eastonitejoin. American Chemistryof Materials,5, 1805-1813. Mineralogrst,77, 1191-1205. Pipping, F. (1966) The dehydration and chemical composition of lau- Coombs,D.S., Ellis, A.J., Fyfe, W S.,and Taylor,A.M. (1959)The zeolite montite. Mineral Society,India IMA, 1, 159-166. facieswith comments on the interpretation of hydrothermal syntheses. Richet, P., Bottinga, Y., Denielou, L., Pettet, J.P., and Tequi, C. (1982) Geochimicaet CosmochimicaActa, 17, 53-107. Thermodynamic propertiesof quartz, cristobaliteand amorphousSiOr: Crawford, WA., and Fyfe, W.S. (1965) Lawsonite equilibria. American Drop calorimetry measurementsbetween 1000 and 1800 K and a re- ioumal of Science.263. 262-270. view from 0 to 2000 K. Geochimica et Cosmochimica Acta. 46.2639- Eberlein,G.D., Erd, R.C., Weber,F., and Beatty,L.B. (1971)New oc- 2658. currence of yugawaralite from the Chena Hot Springs area, Alaska. Robie, R.A , Hemingway,B.S., and Fisher,J.R. (1978)Thermodynamic American Mineralogist, 56, 1699-17 17. properties of minerals and related substancesat 298.15 K and I bar Ellison, A J.G., and Navrotsky, 4,. (1992) Enthalpy of formation of zir- ( 105 Pascals)pressure and at higher temperatures.U.S. GeologicalSur- con. Joumal ofAmerican Ceramic Society, 75(6), 1430-1433. vey Bulletin 1452,456 p. Gerardin, C., Sundaresan,S., Benziger, J., and Navrotsky, A. (1994) Seki, Y ( I 968) Synthesizedwairakites their differencefrom natural wair- Structural investigation and energeticsof mullite formation from sol- akites. Journal of Geological Society,Japan, 74, 457-458. gel precursorsChemistry ofMaterials, 6, 160-170 Senderov,E.E. (1980) Estimation ofgibbs energyfor laumontite and wair- Glushko, VP (1981) Thermodynamic propertiesof individual sub- akite from conditions of their formation in geothermalareas. Proceed- stances,vol. 3,251 p Nauka,Moscow ings of The Fifth International Conferenceon Zeolites, Naples, Italy, Gottardi, G., and Galli, E (l 985) Natural zeolites,409 p. Springer,Berlin. June2-6. Helgeson,H C , Delany,J.M., Nesbitt,H.W., and Bird, D.K. (1978)Sum- - (1988) Physical-chemical aspects of zeolite formation in na- mary and critique of the thermodynamic properties of rock-forming ture. In Occurrence,properties and utilization of natural zeolites, I I l, minerals. Arnerican Journal of Science.278A. l-229. 147 p. Hemingway, B S. (1987) Quartz: Heat capacitiesfrom 340 to 1000 K and Smelik, E.A., Jenkins, D.M., and Navrotsky, A. (1994) A calorimetric revised values for the thermodynamic properties. American Mineral- study of synthetic amphiboles along the tremolite-tschermakitejoin ogist,72,273-279. and the heats offormation of magnesiohornblendeand tschermakite. Hemingway, B.S.,and Robie, R.A (1977) Enthalpiesof formation of low American Mineralogist, 79, | | l0-l 122. albite NaAlSirOr, gibbsite AIOH, and NaAlOr; revised values for Tak6uchi, Y.,Mazzi, F., Haga, N., and Galli, E (1979) The crystal struc- AI{rr. and AG!r* of some aluminosilicate minerals. Journal of Re- ture of wairakite American Mineralogist, 64, 993-1001 searchU S. Geological Survey, 5, 413-429. Thompson, A.B. (1970) Laumontite equilibria and the zeolite facies. Holland, T.J B, and Powell, R. (1990) An enlargedand updated inter- American Journal of Science,269, 267-27 5 nally consistentthermodynamic dataset with uncertaintiesand corre- Yakubovich, O.V., and Simonov, M A. (1985) Refrned crystal structure lations: The system KrO-NarO-CaO-MgO-MnO-FeO-Fe.O,-AlrOr- of the zeolite laumontite Ca(H'O)r.[Al,SioO'r] 0 5HrO. Society of TiO,-SiO,-C-Hr-Or. Journal of Metamorphic Geology, 8, 89-124 Physicsand Crystals,30,624-626. Hu, Y., Navrotsky,A., Chen, C.-Y., and Davis, ME (1995)Thermo- Yamazaki, A., Shiraki, T, Mishido, H., and Otsuka,R. (1991) Phase chemical study of the relative stability of denseand microporous alu- changeof laumontite under relatively humidity-controlled conditions. minophosphateframeworks. Chemistry of Materials,7, l8l6-1823. Clay Science,8, 79-86 Ivanov, P.P., and Gurevich, L.P. (1975) Experimentalstudy of :t-X (COr) Zen, E-an (1972) Gibbs free energy,enthalpy, and entropy of ten rock- boundariesof metamorphic zeolite facies.Contributions to Mineralogy forming minerals: Calculations,discrepancies, implications. American and Petrology,53, 55-60. Mineralogist, 57, 524-553. Kerr, I.S., and Williams, D.J. ( 1969)The crystal structureof yugawaralite Znng,Y., and Liou, J G. (1982)Experimental investigation ofyugawar- Acta Crystallography,825, I I 83- I I 90. alite-wairakite equilibrium, American Mineralogist, 67, 937-943 King, E.G., and Weller, WW. (1961)Low-temperature heat capacities Zhu, H., NeMon, R.C., and Kleppa, O.J. (1994)Enthalpy of formation and entropiesat 298.15 K of some sodium- and calcium-aluminum of wollastonite (CaSiO') and anorthite (CaAlrSi'Or) by experimental silicates.U.S. Bureauof Mines ReportInv. 5855, l-8. phase equilibrium measurementsand high-temperaturesolution calo- Kiseleva, I.A., and Ogorodova, L.P. (1983) On using of the high-temper- rimetry. American Mineralogist, 79, 134- 144. ature calorimetry of dissolution for determination of enthalpiesof formation of hydroxyl bearing minerals (using the examples of talc and Mar.n;scnrsr RECETvEDMnncu 9, 1995 tremolite). Geokhimia, 12, 1745-1756 (in Russian). MeNuscnrsr AccEprEDJexunnv 20, 1996 KISELEVA ET AL.: THERMOCHEMISTRY OF CALCIUM ZEOLITES 667

AppENDrxTrele. Thermodynamiccycles APPENDTXT ABLE. - Conti n ued

A. Leonha?dite dehydration 2SiO,(sol, 975 K) : 2SiO.(xl, 298 K) (21f = CaAl,SioO,23.5H,O (xl, 298 K) : CaAl,SilOr2(xl, 975 K) + CaAl,SioO,"(xl, 298 K) CaALSizOe(xl, 298 K) + 2SiO,(xl, 3.5H,o(9,97s K) (1)" 298 K) (221 CaAl,Si4O12(xl, 975 K) : CaAl2SilO,,(xl, 298 K) (21' AH22: -AHs + A4e + AH2o+ LH21 3.5H,O(9, 975 K) : 3.5H,O(1, 298 K) (3)" H. Laumontite formation (fi?st cycle) CaAl,Si4O1r.3.5HrO(xl, 298 K) : CaAlrsiool,(xl, 298 K) + : 3.5H,O(r, 298 K) (4) CaAI,SLO,,.3.5H"O(xl, 298 K) + 0.5HrO(1, 298 K) CaAl,Si4O1,.4HrO(xl, 298 K) -(sr L,Hn: AH. + LH2+ LHs cao (xl, 298 K) + Al,O3 (xl, 298 K) + 4SiO, (xl' 298 K) + B. Laumontite dehydration 3.5H,O (1,298 K) : CaAl.Si.O,,.3.5H,O(xl, 298 K) (16f CaAl2Si4Olr'4HrO(xl,298 K): CaAlrsinol2(xl, 975 K) + CaO (xl, 298 K) + Al,O3(xl, 298 K) + 4SiO, (xl,298 K) + : 4H,O(9,298 K) (5f" 4H,O (1,298 K) CaAlzSirO'"4HrO (xl, 298 K) (23) CaAl,Si4O,,(xl 975 K) : CaAlzSirO,z(xl, 298 K) (2)"D aHB: -LHs + AH16 4H,O(9,975 K):4H,O (1,298 K) (6) l. Laumontite fo.mation (second cycle) CaAlrSioO,r.4HrO(xl, 298 K): CaAl2SioO,,(xl, 298 K) + 4H,O (r,298 K) (7) CaAl,SioO,,(xl, 298 K) + 4H,O (1,298 K): CaAl2Si.O,,' 4H.O (xl, 298 K) (24)' L,Hr: A,Hu+ AH, + AH6 CaO (xl, 298 K) + Al,O3(xl, 298 K) + 4SiO, (xl, 298 K) : C, Laumontite-leonhardite reaction CaAl.SioO,.(xl, 298 K) (18f GaAl.SioO,.4HrO (xl, 298 K) : CaAlzSi.O,.(xl, 975 K) + cao (xt, 298 K) + Al,O3 (xl, 298 K) + 4SiO' (xl' 298 K) + 4H,O (9, 298 K) (sr 4H,O 0, 298 K): CaAlzSLOrz.4H.O(xl, 298 K) (25) CaAl,SioO,"(xl, 975 K) + 3.5H,O(9, 975 K): CaAlzSirO,z' LH25: L'H2aI AH16 3 sH,O (xl, 298 K) -(1)' J. Wairakite formation 0.5H,O(9, 975 K) : 0.5H,O(1, 298 K) (8)" CaAl,Si4O1,'4H,O(xl, 298 K): CaAl,SioO,,3.5H.O (xl, 298 CaO (sol, 975 K) + Al,Oo (sol, 975 K) + 4SiO, (sol, 975 K) : K) + 0.5H,O(r, 298 K) (s) + 2H,O (9, 975 K) CaAl,SioO,.'2H.O(xl,298 K) (26f CaCO.(x1,298 K): CaO(so|,975 K) + CO, (9,975 K) (10)' aHe: AH" - LH + AH6 CaO (xl, 298 K) + CO, (9, 298 K) : CaCOs(xl, 298 K) (11)' D. Formation of leonhardite (using drop-solution calorimetry ot co, (9, 975 K) = CO. (9, 298 K) (12)' leonhardite) Al,O" (xl, 298 K): Al,O3(so|,975 K) (13)" CaCO.(xl, 298 K): CaO(sol, 975 K) + CO,(9,975 K) (10)' 4SiO, (xl, 298 K): 4SiO, (sol,975 K) (14)" cao (xl,298 K) + CO,(9, 298 K) : CaGOs(xl, 298 K) (11)" 2H,O (t,298 K) : 2H,O (9, 97s K) (271o CO,(9, 975 K) : CO,(9, 298 K) (12)' cao (xl, 298 K) + AlrO3(xl,298 K) + 4SiO, (xl,298 K) + AlrO3(xl, 298 K) : Al,O3(sol, 975 K) (13r 2H,O (1,298 K) : CaAlzSi.O'"2H,O (xl, 298 K) (28) 4SiO,(xl, 298 K): 4SiO"(sol, 975 K) (14r LH26: AH"5+ A4o + LHr + AHp + AHls + LHn + LH27 (1, : (9, (3)' 3.5H,O 298K) 3 sH,O 975K) K. Yugawaralitefotmation CaO (sol, 975 K) + Al,O3 (sol, 975 K) + 4SiO, (sol, 975 K) + 3.5H,O(9, 975 K): CaAlzSi+O'z'3.5HzO (15f CaO(so|,975 K) + Al,Oo(sol, 975 K) + 6SiO,(sol, 975 K) CaO (xl, 298 K) + Al,Os(xl, 298 K) + 4SiO, (xl, 298 K) + + 4HrO(9, 975 K): CaAl,SLO,u'4H.O(xl, 298 K) (2sr (xl, = (sol, K) + CO, (9, 975 K) (10) 3.5H,O0,298 K): CaAl,Si.O,,3.5H,O (xl' 298 K) (16) CaCO. 298 K) CaO 975 CaO(xl, 298 K) + CO,(9, 298 K): CaGOs(xl, 298 K) (11)P AH16: L,H6 + A41 + LH12+ LHB + LHn + LHs + LH$ CO, (9, 975 K) : CO, (9, 298 K) (12)P E. Formation of leonhardite (using dehydration and drop-golution Al,O3(xl, 298 K) : AlrO,(sol, 975 K) (13r calolimetry of metaleonhardite) 6SiO,(xl, 298 K): 6SiO,(sol, 975 K) (30r = -(6) CaCO. (xl, 298 K): CaO (sol,975 K) + CO, (9, 975 K) (10)" 4H,O0' 298K) 4H.O(9, 975 K) cao (xr,298 K) + CO, (9, 298 K) : CaCOs(xl, 298 K) (11)" CaO(xl, 298 K) + Al,Os(xl, 298 K) + 6SiO,(xl, 298 K) + : CO, (9, 975 K): CO, (9, 298 K) (12)" 4H,O (1,298 K) CaAl"Si.O'. 4H,O (xl, 298 K) (31) Al,Os(xl, 298 K) : Al,Os(sol, 975 K) (13r LHsl: LHn + AH,o + AHr + AHe + AHrs + AHs - AH6 4SiO, (xl, 298 K) : 4SiO, (sol,975 K) (14r : : : produced CaO (sol,975 K) + Al,O3(sol, 975 K) + 4SiO. (sol,975 K) : Note: a this work; b Robieet al. (1978); c the materials (xl, 298 K) (17)" by dehydrating laumontite and leonharditeare essentiallythe same and CaAlrSioO,2 : GaAlrSioO,,(xl, 298 K) : CaAlrSioO',(xl, 975 K) -(21" have the same heat contents, calledAH, in the cycles above; and d Zhu CaAl,SioO,"(xl, 975 K) + 3.5H,O(9, 975 K): et al. (1994). CaAlrSi401r.3.5HrO(xl, 298 K) 3.5H,O(1, 298 K): 3.5H,O(9, 975 K) -(1r" CaO (xl, 298 K) + Al,O3(xl, 298 K) + 4SiO, (xl, 298 K) + 3.5H,O(1, 298 K) : CaAl.Si.O,..3.5H,O(xl, 298 K) (16) L,H,u: A,H,s+ AH'1 + LHp + LHn + A44 + LHn - LH2' AH, F. Formation of metaleonhardite (tully dehydrated leonhardite) CaCO. (xl, 298 K): CaO (sol,975 K) + CO, (9,975 K) (10f CaO (xl, 298 K) : CO, (g, 298 K) : CaCOs(xl, 298 K) (11)' CO, (9, 975 K) : CO, (9, 298 K) (12)P Al,Oo(xl, 298 K): Al,Os(sol, 975 K) (13r 4SiO, (xl, 298 K): 4SiO, (sol,975 K) (14)" CaO (sol,975 K) + Al,Os(sol, 975 K) + 4SiO, (sol,975 K): CaAl2SinOl,(xl, 298 K) (17f CaO + AlrOs+ 4SiOr: CaAl.SioO," (18) AHe : A4o + A4, + AHp + LHn + LHn + AHi G. Metaleonhardite : anorthiie + quartz CaAl,SioO,,(xl, 298): CaO (sol,975 K) + Al,O3(sol, 975 K) + 4SiO, (sol,975 K) -(17f CaO (sol,975 K) + Al,Os(sol, 975 K) + 2SiO, (sol,975 K) : CaAl,Si,O"(xl,975 K) (19f CaAl,Si2Os(xl, 975 K) : CaAlzSizOa(xl, 298 K) (20)"