Amertcan Mineralogist, Volume 66, pages 298-308, l98l

Phaserelations of sometungstate under hydrothermalconditions

L. C. Hsu Nevada Bureau of Mines/Geologlt and Department of Geological Sciences, Mackay School of Mines University of Nevada, Reno, Nevada 89557

Abstract

The binary phaserelations ofseveral naturally-occurringtungstats mins14ls, ofthe general form M2*WOaew€t€ investigatedhydrothermally at fluid pressureof 1.0 kbar and temper- aturesof 400o-900oc.A completesolid solution, with tetragonalsymmetry, between scheel- ite (M2* - Ca) and (Pb) forms above725"C.The solvusis alnost symmetricalwith a crest near CaorPbrr.The observedlinear variation ofthe 116interplanar spacing, 1.687(2)A for Ca to 1.782Q)Afor Pb, may be used to determinecomposition. Sanmartinite, (ZnWOa) forms a complete solid solution of monoclinic symmetry with both (FeWOa), and huebnerite(MnWOo), down to at least400oC. The interplanar spacingof the 200reflection of these solid solutions is also observedto vary linearly with composition. Solid solution is very limited betweenmonoclinic and tetragonaltungstates. Less than 5 mole percentmiscibility of one in the other is observedat temperaturesup to 900oC. It appearsthat the solid solution among thesetungstate minerals is affectedstrongly by both the availability and environment of the structural sitesfor specific divalent cations. Even where structural limitations perrnit extensivesubstitution, physicochemical characteristics of elements such as Zn ar.d Pb mav still restrict the occurrence of solid solutions in nature.

Introduction studieson zonedcrystals of ferberitic con- Naturally-occurring tungstate minerals belong to taining inclusions from Australia. two structural groups. One, including scheelite Among the monoclinic tungstate minerals, a com- (CaWO.) and stolzite (PbWO4), crystallizes in the plete solid solution series between FeWO4 and tetragonal systemwith space grotrpl4r/a: the other, MnWOo has long been known in nature, and Hsu including ferberite (FeWOo), huebnerite (MnWO.) (1976)experimentally demonstrated that the compo- and sanmartinite(ZnWO.), crystallirss in the mono- sitional variation of this seriesin terms of FelMn ra- clinic system with space group F2/c. tio can:rot be used to evaluate temperaturesof ore The subsolidusphase relations between scheelite formation. Chang (1968)showed that the solid solu- and many other tungstateswith divalent cations,ex- bility in the system MnWOo-ZnWOo is complete cept Fe2*,were studied by Chang (1967) at temper- above 840oC,below which temperaturea broad mis- aturesfrom 550oto I150'C under anhydrousatmo- cibility gap exists.On the other hand, Chernyshevet spheric conditions. He found a complete solid al. (1976)using hydrothermal techniquesdetermined solution betweenCaWOo and PbWOo above 8l5oC that no miscibility gaps are observedin the systems with a broad miscibility gap below that temperature. FeWOo-ZnWOo and MnWOo-ZnWOo at temper- According to Chang, the system CaWOo-MnWOo ature down to 500'C at 1000atm. shows extremely low mutual solid solubility; that is, PbWO4 in nature also occurs as the monoclinic CaWOo takes up l0 mole percent MnWO. only at a , , which has a structure (spacegroup temperature above I l00oc, whereasMnWOo takes F2,/a) different from other monoclinic tungstate up only 2.5 mole percentCaWOo at the sametemper- minerals.Raspite is 6 percentdenser than stolzite,us- ature. Yet Grubb (1967) reported that an almost ing data of Fujita et al. (1977)for raspiteand thoseof complete solid solution series between wolframite, Plakhov et al. (1970)for stolzite.Raspite is reported (Fe,Mn)WO., and scheeliteexists at elevatedtemper- to transform irreversibly to stolzite at 400oC (Shaw atures, based on his chemical and X-ray diffraction and Claringbull, 1955).Raspite has so far defied lab- 0m3-ffi4X / 8| / 0304-0298$02.00 298 HSU: PHASE RELATIONS OF TUNGSTATE MINERALS oratory synthesis and its relation with stolzite re- for eachdetermination. Measurements were made to mains unsolved(e.g., Chang, 1971). 0.005o20 and resultsaveraged. The purposeof this investigationwas to resolvethe PhaseRelations above-mentionedconflicting conclusions and, in gen- eral, to determine as much as possible about tung- The subsolidusphase equilibria for nins of the ten state minerals and mineralization by using conven- possiblebinary systemswere determinedat Pr: 1.9 tional hydrothermal techniques and facilities at P-T kbar and 7 : 400-900oC. These nine systemsare ranges simulating those involved in natural pro- Ca-Pb, Fe-Zn, Mn-Zn, Ca-Fe, Ca-Mn, Fe-Pb, cesses. Mn-Pb, Ca-Zn, and Pb-Zn. The first three involve end membersof the samestructure, the rest involve Experinental end members of differing structures. Phase relations Conventional hydrothermal techniques were em- for the remaining binary system, Fe-Mn, were re- ployed throughout the work. All the experiments ported previously (Hsu, 1976). were made in cold-seal pressure vessels (Tuttle, The systemCaWO o- PbITO 1949).Details of the experimental apparatus and P- " Z calibrations are describedelsewhere (Hsu, 1976). A complete solid solution between scheelite and Three tungstate end members, CaWOo, PbWOo stolzite forms above 725"C. The solvus is alnost and ZnWOo were readily preparedby chemical pre- symmetrical with a crest near CaorPbrr.As is shown cipitation of reagentgrade NarWO4'2H2O solution, in Figure l, the position and shapeof the solvusare respectively,with reagentgrade CaCl', Pb(NO,), and quite diferent from those previously found (dotted Zn(NOr)r' 6HrO solutions in stoichiometricpropor- tine) by Chang (1967) under anhydrous, one atmo- tions. After filtering, washing, and firing these precip- sphere conditions. The difference in the results of the itates at 700'C for 24 hours, mixtures of the three bi- two studies may be attributed principally to the slow- nary systemswere prepared at l0 mole percent ness of diffusion and attainment of equilibrium in intervals by blending two respectiveend members. dry crystalline powders. Here in view of the nature of The other two end members FeWOo and MnWOo the starting material being used, hydrous conditions, cannot be readily prepared by simple chemical pre- and the sufficient duration of runs being adopted, cipitation. Mixtures of thesebinary systemswere pre- equilibrium was consideredto be attained.A reversal pared by thoroughly mixing reagent grade tungstic of reaction was demonstratedby Run #5Ca5Pbl3 in acid with respectivemetal (electrolytic and man- which a homogeneousphase formed at higher tem- ganese)powders and metal oxides (reagent grade peratureunderwent gnmiving, although a still longer CaO, PbO, and ZnO) stoichiometrically,also at l0 run duration is needed as indicated by the measured mole percent intervals. Additional mixtures at 5 mole interplanar spacings. percent intervals were prepared near two end mem- The intervals between intermediate members bers where necessary. plotted in Figure I are not l0 mole percent as in- Each run was sealedin a noble metal capsule with tended, because a formula weight for CaWOo of excessdistilled water, brought to temperature,and 207.9276instead of 287.9276was used accidentally in quenchedisobarically at a fluid pressureof 1.0 kbar. preparing mixes. This mistake was discovered while Whenevernecessary, the samerun was repeatedsev- plotting the drrcvalues of the two end membersand eral times at the same temperature for different peri- nine intermediate members composition on the bi- ods to ensureequilibrium. Somehomogeneous high- nary. Correct mole percentcompositions are plotted. temperature products were run at lower temperatures The intensity of the 116reflection remains consis- to test whether exsolution had occurred (e.9., Run tently high throughout the compositional range. The #5Ca5Pbl3). d,,u value varies linearly with compositionas shown Condensed run products were examined with a in Figure 2. Using clear natural quartz as an internal petrographic microscopeand a Norelco X-ray dif- standard,with its 112reflection calibrated against the fractometer equipped with a graphite mono- 220 rcflectionof Si(a : 4.43012L),the d,r" value was chromator and scintillation counter and using either found to vary linearly from 1.687(2)Afor scheeliteto CuKa or FeKa radiation. The interplanar spacings I.782Q)A for stolzite.The critical runs for this sytem were obtained by choosing appropriate internal stan- are listed in Table l. dards. Three oscillationsat l/4" 20 per minute scan Scheelite is the only tungstate mineral which con- speedand l/2 n. per minute chart speedwere made sistently yields a blue to whitish fluorescent color un- HSU: PHASE REI../ITIONS OF TUNGSTATE MINERALS

Pf "lKb o oo o o o o o oo o o

o o A o a o o A o o o A A. o A A

o

\ \

0 40 60 roo CoWOo Pbw04 (c) Mole "/" PbWO4 (P)

Figure l. T-X diagram at P1 : 1.9 kbar for the Ca-Pb system. Solid circles: singte solid solution; opea triangles: two solid solutions Dotted line is the solvus after Chang (1967).

A r.soo

T = 75O"C P1=LOKb

40 60 80 too Pbwo4 Molc 96 PbWO4 (P)

Figure 2. Variation of d,,6 with compositionin the Ca-Pb systcm. HSU: PHASE REI'IITIONS OF TUNGSTATE MINERALS 301

Table l. Run data for the Ca-Pb system der short-wave ultraviolet Ught. With 2 mole peroent CaMoOo in solid solution, scheelite quickly changes Duration Run Product Run No. hrerial T"C (days) arrOi* its fluorescent color to yellow and remains essentially solid solution series 9Ca1Pb2 Tungsrete nlx 900 2 single solid sofn. so througftout the rest of the 9Ca1Pb5 " 75O 30 single solld soln. ' spectral d = 1.699 (Hsu and Galli, 1973), although detailed 9Ca1Pb8 r' 500 30 study enablesone to estimatewithin the accuracyof 9caub12 " 450 72 :: 9calPbr " 400 86 I to 2 mole percent in the range from 0 to l0 mole Bca2Pbi ' 850 single solid so1n. (Shoji, 1978).In the Ca-Pb series, 8ca2Pb5 " 750 30 stngle so11d soln. ' percent CaMoOo d = 1.709 changes from bright blue for 8ca2Pb9 " 650 40 single solid so1n. the fluorescent color 8Ca2Pb8 " 600 97 white with a yellowish tint 8ca2Pb11 " 550 16 two solid solns , pure scheelite, through = = -d 1.708, (q 1.752)** brightnessfor gCalPb to 7Ca3Pb,to 8Ca2Pb3 " 500 83 two solid solns,, and decreasing 4 = 1.704, (d- = 1.7s6) gray to light gray for 6Ca4Pband 5Ca5Pb. 8ca2Pb1 " 400 113 Evo solid solns., yellowish d = 1.699, (d_= | 76L) No fluorescenceis observed for the rest of the inter- 7ca3Pb2 " 900 6 s1ngle solid soln. members. Pure stolzite fluoresces rather TCa3PEi " 850 mediate 7Ca3Pb5 " 750 single solld soln , but heat- d = I.722 bright yellow when it is well crystallized, 7Ca3Pb10 " 700 12 single sol1d so1n. precipitate fails to show fluores- 7ca3Pb9 " 55p 97 two solid solns. treated chemical 7Ca3Pb8 " 600 50 tso solid solns., of stolzite evidently is 4=r.712,(d-r748) cence. The fluorescence 7ca3Pb11 " 550 76 Ewo solid solns., = = quenched with minor impurities in the struc- 4 1.707, (4 1.7s3) quickly 7Ca3Pb3 " 500 83 EUo solid solns., = = as indicated by the observation that the syn- -d 1,705, 4 1.755 ture, with 3 mole percentCaWOo (at 750oC, 6ca4Pb2 " 900 3 single solid soln. thetic stolzite 6ca4Pbi " 850 is no longer fluorescent.Thus the fluores- 6Ca4Pb5 " 750 28 slngle so1ld soln I kbar) d = L,733 behavior of natural stolzite may serve as a 6Ca4Pb10 " 700 72 Ewo so1ld solns., cence 4=I.737,4=7-123 ' good indicator for the chemistry of the mineral. A 6Ca4Pb9 650 91 cwo solid solns ' d-=L.7r7,4=I.745 fluorescencebehavior similar to stolzite is also ob- 6Ca4Pb8 " 600 30 two solid solns., :d=1.709,d=1.751 served for sanmartinite. 5ca5Pb2 " 900 single solid so1n. 5Ca5Pb7 " 850 3 MnWO 5ca5Pb1 " 800 3 Thesystems FeWO n-ZnWO o and o-ZnlV'On 5ca5Pb5 " 750 14 single solid soln. d = 1.740 From previous study on the wolframite series 5Ca5Pb10 " 700 ]0 single solld sotn. 5ca5Pb9 " 650 70 two solld solns ' (Hsu, 19?6)it was learned that the dr* spacingof the !=L.744,!=r'716 5Ca5Pb8 " 600 32 two so1ld solns. il monoclinic tungstate minerals is quite sensitive to 5ca5pbt 400 59 Ewo solid solns,, 4=t,764,d=1.700 compositional variation. By using FeKa radiation 5ca5Pbl3 5C^5Pb2 400 It3 chree solid sotns., c=] 760, q=l.739, d=1.702 the composition of the binary solid solution series 4ca6Pb2 Tur,gsEace mix 900 2 single solid soln. ,, deternined quickly and accurately through 4ca6pb5 750 30 single solid soln, can be of the separation between the 200 re- 4ca5Pbl0 " 700 10 slngle solid soln. measurement 4Ca6Pb9 " 650 27 flection of the mineral and the 102 reflection of 4ca6Pb8 " 600 30 single solid soln. ' d = 7.152 quartz. The L20 (FeKc) for FeWOo, MnWOn and 4ca5Pbl1 " 550 11 two solid soln, ' = = d 1.753, (d 1.707) respectively,in degrees 1.925(10)' 4ca6Pb3 r' 500 83 two so1ld solns, ZnWOn are, d = 1.7s5,(4 = 1.705) 2.920(lO),and 1.435(10).As shown in Figure 3, the 3calPb2 Tungstate n1x 900 2 single sol1d soIn. 3Ca7Pb7 850 4 linear relation between the A2fls and the composi- 3ca7Pb5 750 l0 single sofid soln', tions of the mineral serieshold perfeclly. A negative 3ca7Pb9 650 27 single solld so1n. 3ca7Pb8 600 30 deviation from Vegard's rule reported in the Fe-Zn 3ca7Pbr1 550 71 3Ca7Pb3 500 12 seriesby Chernyshev et al. (1976) is not apparent in 3ca7PbL2 450 83 rwo solid solns., g = 1.761'(4 " r.703) the present study. Their usageof the less sensitive, 2CagPbz 900 2 sin8le solid soln. lower order 100 reflection may partly be responsible 2CaaPbT 850 2ca8Pb5 7s0 30 single solid soln., for this discrepancy. With the dro, of quartz cali- 2Ca8Pb9 550 27 stngle solid soln, brated against the d,,o of spectrochemicallypure 2ca8Pb8 600 30 2CaBPbl 400 86 tungstenpowder, the calculatedd.,,fot the three end 1Ca9Pb2 900 2 -"-b"tt- are 2.367l(5), 2.4140(5), arrid2.3M8Q)4, lca9Pb5 150 30 slngle golld 9orn. t d = I.777 respectively. 1ca9Pb8 500 30 single solid sofn. lCa9PbI 400 15 A complete solid solution exists for both Fe-Zn Mn-Zn series(Tables 2, 3), at leastdown to 400o *CU6(A) for pureca= L587, for Pure Pb - 1 782 and **M1nor phases are in palenthesls. C at Pr: 1.0kbar, and perhapseven much lower for 302 HSU: PHASE REI..ATIONS OF TIJNGSTATE MINERALS

FeKa A20 = ZOQ,O.-2O MZzoo

20Q - 20 FZ to2

T = 9OO"C P, = l.OKb

o 4^ o- 9 x o N <40

--o2q I 40 60 too MnWOa(M) ZnWO^ FeWOa(F) Mole 7c ZnWO4 (z')

Figure 3. Variation of A2O with composition in the Fe-Zn and Mn-Zn systems.

the ie-Zn series. The Mn-Znseries takes much atures as high as llOOoC were confirmed in this longer to achieveequilibrium and to obtain adequate study. In all the systems,compositions of 5 mole per- crystallinity than the Fe-Zn series.In the Mn-Zn se- cent one end member with 95 mole percent of the ries at 400oC,the width of eachof the 200 reflections other always produced two-phase assemblagesat obtained from the products of two-month runs is on 900"C. However, within the error of measuremef,q the order of twice the width of each of the peaks ob- the interplanar spacingsof the products always show tained from the products of four-month or longer a finite but very small change in the proper direction runs; peak position remains unchanged,indicating as compared with those of the pure end members. improved crystallinity without compositionalchange. Assuming a linear relation between interplanar spac- The higher solvustemperatures reported by previous ings and composition allows the solubility limits to authors (Chang for Mn-Zn, Cherynshev et al., for be plotted as shown in Figures 4, 5, and 6. The criti- Fe-Zn and Mn-Zn) can be attributed simply to ki- cal runs are listed in Tables 4, 5, and 6. The inter- netics problems. planar spacings,type of X-ray radiation and internal standards employed are included in these tables for Thesystems CaWO FelUO n- o,CaWO r- MnWO o, each binary system investigated. Fe l4tOo- Pb llo o,M nWO o-Pb WOo, CaITO o-Zn lt O,, and Pb lV'On- Zn WO* Polynorphismof PblAOo Very wide miscibility gaps reported by Chang Chang (1971) attempted to synthesizeraspite (1967)in the Ca-Mn and Ca-Zn systemsat temper- through chemical precipitation under various pH HSI]: PHASE RELATIONS OF TUNGSTATE MINERALS 303

Table 2. Run data for the Fe-Zn svstem ments,that raspite may not have its own P-T stabil- ity field and may have grown metastably due to un- Sta! t ing ra f,aon Run Product Ruo No. Material (days) 420 (degrees)* known kinslis effects.

9EelZn2 Oxide nlx 900 J rruErr Discussionand Conclusion 420 = 1.870 9FelZnl 400 71 single solid so1n. The gradual widening of the miscibility gap with 8Fe2zr2 900 3 single solid soln. A2O = 1.830 decreasingtemperatures in the Ca-Pb series may 8Fe2ZnI 400 11 single solid soln. provide valuable geothermometry by determining 7Fe3Zo2 900 3 single solid soln. 420 = 1.780 the compositions of coexisting scheelite and stolzite. TFe3Znl 400 78 single solid so1n. L20 = 7.77O As Pb has a strong affinity for S to form sulfide (ga- 6Fe4Zo2 900 6 sirrgle solid soln. even in the presenceof L2A = I.72O lena), rather than tungstate , stolzite seldom occurs in high-temperature )te)Laz 900 10 single solld soIn. L2A = 1.67O ore depositswhere .f., is usually high. In nature, Mo 5Fe5Zn8 850 4 single solid soln. is readily incorporated into the Ca-Pb series as a 5Fe5Zi6 5Fe5Zt2 700 I2 single solid soln. 5Fe5Znl6 Oxlde nrix 650 110 substituent for W. However, this can happen only 5Fe5Zn13 550 t42 (Hsu, 5Fe5Zn10 450 180 under low /r. and/or high /". conditions 5Fe5Zn1 400 r.88 siogle solid soln. same environment where stolzite occurs. L2A = L675 1977)-16" A completesolid solution existsbetween CaWOo and 4F36zn2 900 6 single solid soln. 420 = 1.635 CaMoOo to very low temperatures(Hsu and Galli, 3F31Zn2 900 3 single solid soln. is probably also true betweenPbWOn and ^20 = 1.583 1973);this 3Fe7 znl 400 71 single solid soln. PbMoOo. Therefore, the effect of minor Mo on the A20 = 1.580 2Fe8Zn2 900 3 single solid soln. miscibility gap betweenCaWOo and PbWO. may be 420 = 1.540 negligible. Calculation basedon lJrusov's method to 2Fe82I 400 77 slngle solid so1n. 7F39za2 900 3 single solid soln. be describedbelow, however,indicates that the theo- A20 = 1.490 lFe9Zn1 400 78 single solid soln. retical cresttemperature of the solvusis about 350'C

*A20 FeKo(102 of quartz and 200 of Ie-Zn) for pure Fe = 1.925"; IOr pute zn = f.4J) Table 3. Run data for the Mn-Zn system

Material T'C (days) 420 (degrees)* conditions.Stolzite was produced in all cases.When 9Mn1Zn2 Oxide mix 900 4 single solid so1n. subjectedto high pressuretreatment using opposed- L2A = 2.760 anvil apparatus, these precipitates transformed to a 9MnlZnf 400 78 single so1ld soln. 8Mt\22\2 900 4 single solid so1n. high-pressureform of PbWO4.A singlecrystal of this ^2a= 2.620 8l4n2znL 400 748 s1ng1e solid soln. form was prepared and its structure analyzedlater by L2a = 2-615 7Mn3Z\2 900 4 single solid soln. Richter et al. (1976). L2A = 2.470 7Mn3Zn7 400 148 single solid soln. In this study, oxide mixes or tungstatesfrom chem- L20 = 2.465 @In4Znz 900 6 siogle sofid so1n. ical precipitation with seedsof natural raspite or mix- A2A= 2,310 tures of natural raspit€ and stolzite were used as 6Ma4Zn7 400 7a single solid soln. 5mn)znr 900 10 single solid soln. starting material. The natural specimenwas from the L20 = 2.165 km 5Mn5Zn6 5Mn5Z^2 700 12 single solid soln Cordillera mine, 600 east of Broken Hill New 5Mn52n16 oxide n1x 650 110 (McColl, provided sun5ZDz 600 35 South Wales, Australia 1974) 5Mn5Zn13 oxide nix 550 742 through the generosity of D. H. McColl, Museum 5Mn5Zn10 450 180 5h5Znl 400 188 singfe solid soln. Curator, Bureau of Mineral Resources,Canberra, A2A - 2 775 Australia. The charges were sealed in Ag-Pd cap- th6zaz 900 6 single solid soln. A20 = 2.020 sules with fluids of various nature, insluding pure 4Mn6Znf 400 78 single so11d soln. 3Vn7zn2 900 4 single solid soln. ^20 = I.855 distilled water, 0.5 m NaCl solution, and lD%oHF ,, 3M\7 Zn7 400 r60 single solid so1n. acid. The experiments were conducted at temper- 420 = 1.860 2Mt8Za2 " 900 single solld soln. aturesas low as l50oC under Pr: 0.5 to 3.5kbar for ^2A= r.705 2h8Zn1 " 4oo 160 single solid soln. a duration as long as four months. In all cases,stol- \Mngzt2 " goo 4 single solid soln. 420 = 1.570 zite grew at the expenseof raspite. Even for the IMn9Znl singl-e solid so]n. = charge with oxide mix and raspite seeds,stolzite was *A2O FeKo(102 of quartz and 200 of Mn-Zn) for pure Mn 2.920"; foi Pure still the final product. It appears from these experi- Z\ = I.435' 3M HSU: PHASE RELATIONS OF TUNGSTATE MINERALS

Pf= rKb

o roo CoWO4 MnWO4 CoWOa FeWO4 (c) Mole "/" MnWOa (M) (c) Mole 7o FeWOe (F)

Figure 4. T-X diagramsat Pr: 1.0kbar for the Ca-Fe and Ca-Mn systems.Symbols as for Figure l.

Pf=lKb Pf=lKb

ip- FeWO. a MnWOo (F) Mole 7o PbWO4 (M) Mole % PbWOq

Figure 5. T-X diagramsat Pr = 1.0kbar for the Fe-Pb and Mn-Pb systems.Symbols as for Figure L HSU: PHASE RELATIONS OF TUNGSTATE MINERALS 305

Pf = lKb

3oo3 CoWOo ZnWO4 PbWO4 ZnWO" (c) Mole 7" ZnWOo (z) (P) Mole "/" ZnWOo e) Figure 6. T-X diagrans at P, : 1.9 kbar for tJre C-a-Znand Pb-Zn systems.Symbols as for Figure 1. below the experimentally-determinedvalue of e'z.n-'zx'n'm'(Lr/dr)2/2R;wherecisanempiri- 7Z5oC. cal parameterand equals 25,000cal/mol for oxides; According to Urusov (1975) the absolute maxi- z* arrd z* are formal charges(valences) of M and X, mum temperature of exsolution, T, in a binary sys- respectively;m is the number of atoms in the formula tem M,.X,, may be determinedby the equation 7: of compound (m : k + l); n is the coordination number of the cation; Ar is the difference of radii of Table4. Rundata for theCa-Fe and Ca-Mn systems the substituents;d, is the smaller averageinteratomic srartins Duration Runproducr distance; and R is the universal gas constant. Using Run No. Maleria1 T"C (days) A20 (degrees)* the structure data of monoclinic tungstates listed in 95CaO5Fe2 oxlde nix 900 C*A2a = 0.945 Table 3 of Weitzel (1976) and those of tetragonal lrace Fe and WO3 95ca05lel " 400 Ca, trace Ie tungstatesby Plakhov et al. (1970), and the ionic 9Ca1Fe2 " 900 Ca: 420 = 0.945 ninor Fe, Erace WO3 ,radius data of Whittaker and Muntus (1970),the ab- 9Ca1Fe1 " 400 7I Ca, ninor Fe 5Ca5Fe2 " 900 6 Ca: A2O= 0.955 solute maximum temperaturesof unmixing of binary Ie: ^2o = 0.640 5Ca5Fe3 500 115 Ca,Fe solid solutionsare calculatedas following: fCa9Ie2 900 2 Fe: 420 = 0.665 minor Ca 05ca95Ie2 900 3 I'e: 420 = 0,665 For Fe-Mn (monoclinic) trace Cd 05Ca95Fe1 400 Ie, 7r trace Ca T : 25,000'2'2'2'6' (0.05/3.973)2/2' 1.987 95ca95Hn2 oxide nix 900 3 Ca: A2O= 0.995 trace Mn, trace WO3 :48 K : -225"C 95ca05Mn1 400 78 9ca1h2 900 2 Ca: 420 = 0.990 ninor I, trace WO3 For Fe-Zn (monoclinic) 9CaIMnf 400 78 5ca5Mn2 900 6 Ca: 420 = 0.995 Mn: A2O = 1.690 T : 25,000'2'2'2' 6' (0.03/3.947)2/2' 1.987 5ca5Mn3 500 115 Ca, Mn 1Ca9Mn2 900 2 Mn: A2O = 1.695 :17 : -256"C nlnor Ca K 1ca9Mn1 400 78 05Ca95Mn2 900 3 h: 420 = 1.705 (monoclinic) lrace Ca For Mn-Zn 05ca95Mn1 400 7a Mn, trace C T : 25,00O'2'2'2' 6' (0.08/3.947)2/2'1.987 *420 FeKo(lIl of A1 and 2I1 of Ca,111 of AI and 200 of Fe or lln) for pure Ca - 0.953'r for Dute Fe = 0.555': for oure Mn = 1.655o : 124K: -149oC 306 HSU: PHASE REI.A,TIONS OF TUNGSTATE MINERALS

Table 5. Run data for the Fe-Pb and Mn-Pb svstems 400'C or below is to be expectedon the basisof this

rrino DIr.t Run ?roduct theoretical calculation. Yet, in nature, Fe-Mn is the Run No Material T'C (days) A2o (degrees)* only system which is found to form a complete solid 95Ie05Pb2 oxide n1x 900 4 Fe:420=0790 pres- trace Pb solution (wolframite). In spite of the constant 9Fe1Pb2 900 6 ence of zinc in most wolframite deposits, ZnWOo is 9FelPb3 500 40 Ie: 420 = 0,795 mlnor ?b not only absent from them in pure form but also sel- 9Fe1P1 400 65 Fe, ninor Pb 5Fe5Pb2 900 7 Pb: 420 = 0.960 dom present as an isomorphous component in wol- Fet L2A = 0.795 5Fe5Pbl 400 r33 ?b: ^20 = 0.950 framite, although some analyses of sanmartinite in- Fe: A2O = 0.805 1Fe9Pb2 900 6 ?b: 420 = 0.960 dicate that up to 46 mole percent FeWOo may be ninor Ie tFe9Pbl " 400 65 Pb, ninor Fe incorporatedin this mineral (Dunn, 1978).This may 05Fe95Pb2 "900A Pb: ^2o = 0.965 lrace Ie be attributed to two propertiesof zinc: (l) Basedon

95Mn05Pb2 oxide mix 900 4 Mn: 420 = 1.320 geometrical considerations the ionic radius of zinc trace Pb 9MntPb2 " 900 6 Mn: 420 = 1.320 dictates that six-fold coordination is the stable one Pb: 420 = 0.980 9Mn1PbI 400 65 Mn: 420 = 1.315 fot Zn'* ion. Yet zinc has a pronouncedtendency to Pb: 420 = 0.955 5Mn5Pb2 900 7 ?b: 420 = 0.978 form covalent, tetrahedrally-directed bonds even Mn 420 = 1.320 with oxygen.Indeed, smithsonite,ZnCOr, is the only 5Mn5Pb1 400 65 Pb: 420 = 0.970 Mn: A2o = 1.315 common zinc mineral with six-fold coordination of lMn9Pb 2 900 4 Pb: 420 = 0.970 Einor Mn Zn'* (Neumann, 1960).(2) Znc has a strongeraffin- l-Mn9Pb1 400 65 Pb: A2o = 0.960 trace Mn ity for sulfur than iron and (e.g., Maraku- 05Mf,95Pb2 900 ?b: 420 = 0.980 trace Mn shev and Bezmen, 1969)and forms sulfide minerals *A2O CuKd(211 of rutile and 3f2 of Pb, 211 of rutife atrd 202 of Fe or !tn) such as sphalerite and wurtzite. Hence the occur- for pure Fe = pure = pure = 0.810'; for Mn 1.315"; for ?b 0.955' rence of ZnWOo either in its pure form or as an iso- morphous component in wolframite requires a rather For Ca-Pb (tetragonal) unusual geologic environment; that is, in the pres- enceof tungstenhigh /", and low /s,, or sufficient to T : 25,000'2'2'2. 8 . (0.17 1.987 /4.263)2/2. say,low /",, in the presenceof water (e.g.,Burl, 1977, : 640K :367"C 1980). Experimental results, both dry and hydrothermal, Thus, even if a similar degree of discrepancy be- indicate that mutual solid solubilities are extremely tween theoretical and experimental values of unrnix- low even at high temperatures for binary tungstates ing temperatureis assumedi1 msaqslinic tungstates with different crystal structures as end members. as in tetragonaltungstates, the formation of solid so- Theseresults appear to be in contradictionto the ob- lutions among Fe, Mn, and Zn tungstatesdown to servationmade by Grubb (1967)on the zoned ferbe- ritic wolframite with scheelite exsolution. If the Table 6. Run data for the Ca-Zn and PFZn systems scheelitelamellae within the zoned wolframite are in-

Starting hration deed of exsolution rather than replacementorigin, Run No. Material T'C 420 (degrees)* three possibilitiesmay be considered:(l) an unrea-

95caO5zn2 Oxide nix 900 ca: A2o = 0.480 sonably long time is required to attain any degree of lrace Zn 9Calzn2 " 9oo Ca: 420 = 0.480 solid solution in the laboratory due to unknown ki- Erace Z\ 900 4 Ca:420=0480 netic problems, (2) exsolution takes place so fast in Zni L2O = l.ru O5Ca95z^2 900 Zn: L2A = I.72O the laboratory that any solid solution attained at high trace Ca temperaturescannot be quenched to room temper- 95Pbo5zt1 Tungs tate 850 4 Pb: 420 = 3.390 trace Zn afixe, and/or (3) a metastable solid solution forms in 9Pb72n2 900 4 Pb: A2o = 3.385 possibly ninor zn nature (due to very rapid crystal growth) 5Pb5zn2 900 ?b: 420 = 3.385 with later exsolution observed Prolonged zil A2A = L.r20 by Grubb. sPb5Zi6 700 10 Pb: A2o = 3.395 run durations or high temperature X-ray apparatus is Zn: A2O= I.Il5 5Pb5Zn1 400 133 ?b: A2o = 3.400 neededto test thesepossibilities. Zn: 420 = I.lI5 lPb9zD2 900 4 Znt A2O= I.L2O The ionic radii and electronegativitiesof the cat- ninor Pb 05Pb95Zn7 850 4 Zn: 420 = I.150 ions involved in the study are listed in Table 7. The Erace Pb extensive solid solutions among the three monoclinic *A2O CuRo(211 of quarEz and 224 of Ca and Pb) for pure Ca = 0.490'; for pure Pb = 3,400. 420 CuKd(102 of quartz and 200 of zn) for pure Zn = tungstatesand between the two tetragonal tungstates 1.110" are explained in terms of similarities in these two HSU: PHASE RELATIONS OF T,UNGSTATE MINERALS

Table 7. Ionic radii and electronegativities of some divalent (A) and monoclinic (B) tungstates.In the tetragonal cauons tungstate structure, the tungsten cations (W6*) are

-t+ coordinated by four oxygen ions and the divalent Ca2* Mn2a cations (Ca'* or Pb'z*),by eight. Each oxygen ion, in

,Vr (A) r.26 r.08 0.91 0. 86 0.83 turn, is surrounded by one W'* and two divalent cat- rvrrr (&) 1--37 1.20 1.01 ions. In the monoclinic tungstate structure, on the , Kcal, 137 170 200 208 L -g.aEoh' other hand, both the hexavalent and divalent cations (Mn'*, Fe'* or Zn2*) arc coordinatedby six oxygen Rooan numerals represent cootdination numbers. Ionic radius (r) data from whittaker and MunEus (1970). Electronegativity (E) ions. There are, however, two di-fferent oxygens in (1964). data from Povarenykh this structure;one oxygen is surroundedby two W'* and one divalent cation, the other, by one W'* and two divalent cations. These differences in environ- quantities. However, if ionic radius and electro- ment, not only for the divalent cations but also for negativity are the only controlling factors, extensive the other ions, may serve to prevent any extensive solid solution between Ca- and Mn-tungstates, at solid solution. least as extensiveas that betweenCa- and Pb-tung- states,should be expected.Indeed, complete solid so- Acknowledgments lutions betweenCa- and Mn-end membersare often Critical reviews of the manuscript by D. M. Burt of Arizona observedin silicate minerals, e.9.,between grossular State University and W. A. Dollase of University of California, and spessartine(Hsu, 1980).Yet, as experinentally Los Angeles are gratefully acknowledged. This work was sup- ported Research Advisory Board, University of Nevada, mutual by the demonstrated,the substitution betweenCa2* Reno. and Mn2*, not to mention betweenCa2* and Fe2*,is extremely limited in the tungstates.Here the host References appearsto play an important role. Burt, D. M. (1977)Chalcophile-lithophile tendencies in the helvite Figure 7 shows the crystal structures of tetragonal group: genthelvite stability in the system ZnO-BeO-Al2O3-

b

A B Figure 7. The structuresofthe tetragonal(A, after Plakhov et al,1970) and monoclinic (B, after Cid-Dresdnerand Escobar, 1968) tungstates. 308 HSII: PHASE RELAI:IONS OF TUNGSTATE MINERALS

SiO2-SO-r-F2O-,. (abstr.) American Geophysical Union denii, 1969,No. 4, 8-23 (transl. International Geology Review, Transactions, 58, 1242. 13, t78r-r794, l97l). Burt, D. M. (1980) The stability of danalite, FeoBe3(SiO.)rS. McColl, D. H. (1974) A new occurrenceof raspite at the Cordille- American Mineralogist,65, 355-360. ran mine. New South Wales. Australia. Australia Gems and Chang, L. L. Y. (1967) Solid solutions of scheelitewith other Crafts Magazine,9-l l. RrrWOa-type tungstates.American Mineralogist, 52, 427435. Neumann, H. (1960)Not€s on the and geochemistry Chang, L. L. Y. (1968) Subsolidusphase relationsin thc system of zinc. Mineralogical Magazine,28, 575-581. ZnWOa-ZnMoOo-MnWOn-MnMoOa. Mineralogical Maga- Plakhov, G. F., Pobedimskaya,E. A., Simonov,M. A., and Belov, zine, 36, 992-996. N. V. (1970) The crystal structur€ of PbWOa. Kristallografiya, Chang, L. L. Y. (1971)Phase relations in the systemPbO*WO3. 15, 1067-1068 (transl. Soviet Physics-Crystallography,15, 928- Journal of American Ceramic Society,54,357-358. 929, r97t). Chernyshev,L. V., Belykh, L. A., and Pastushkova,T. M. (1976) Povarennykh,A. S. (1964)Some fundamental problems of crystal An investigation of solid solutions in the system FeWOa- chemistry in relation to mineralogy, In M. H. Battey and S. I. MnWOo-ZnWOa. Geokhimiya, 5, 7 16-727 (transl. Geochemis- Tomkeie$ Eds., Aspects of Theoretical Mineralogy in the try Intemational, 3, 6M9, 1976). U.S.S.R.p. 135-169.MacMillan, New York. Cid-Dresdner,H. and Escobar,C. (1968)The crystal structure of Richter, P. W., Krugger, C. J. and Pistorius, W. F. T. (1976) ferberite,FeWOa. Zeitschrift ftr Kristallographie, 127,6l-72. PbWO4-III (a high-pressure form). Acta Crystallographica, Dunn, P. J. (1978) Sanmartinite:new data. Mineralogical Maga- 832,928-929. zine,42,281. Shaw, R. and Clarhgbull, G. F. (1955) X-ray study of raspite Fujita, T., Kawada, I. and Kato, K. (1977)Raspite from Broken (monoclinic PbWO4).(abstr.) American Mineralogist,40, 933. Hill. Acta Crystallographica, 833, 162-16/.. Shoji, T. (1978)Fluorescent color and x-ray powder data of syn- Grubb, P. L. C. (1967)Solid solution relationshipsbetween wol- thpsized scheelite-powellite series as guides to d€termine its franite and scheclitc. American Mineralogist, 52, 418426. composition.Mining Geology, 28, 39740/'. Hsu, L. C. (1976)The stability relations of the wolframite series. Tuttle, O. F. (1949)Two pressurevessels for silicate-waterstudies. American Mincralogist, 61, 9M-955. Geological Societyof America Bulletin, 60, 1727-1729. Hsu, L. C. (1971) Etrectsof oxygen and sulfur fugacitieson the Urusov, V. S. (1975)Energetic theory of miscibility gaps in min- scheelite-tungstenite and powellite-molyMenite stability rela- eral solid solutions.Fortschritte der Mineralogie,52, l4l-150. tions. Economic Geology, 72, 664-670. Weitzel, H. ( 1976) Kristallstrukturverfeinerung von Wolframiten Hsu, L. C. (1980)Hydration and phaserelations ofgrossular-spes- und Columbiten. Zeitschritf fiir Kristallographie, 144,238-258. sartinegarnets al P:H.s:2 kb. Contributionsto Mineralogy and Whittaker, E. J. W. and Muntus, R. (1970)Ionic radii for use in Petrology,71,407415. geochemistry.Geochimica et Cosmochimica Acta, 34, 945-956. Hsu, L. C. and Galli, P. E. (1973)Origin of the scheelite-powellite seriesof minerals.Economic Geology, 68, 681-696. Marakushev,A. A. and Bezmen,N. I. (1969)Chemical affinity of Manuscript received,fuly 28, 1980; metals for oxygen and sulfur. Geologiya Rudnykh Mestorozh- acceptedforpublication, November 10, 1980.