1323 The Canarlian Mineralogist V o l . 3 7 .p p . 1 3 2 3 - 1 3 4t1I 9 9 9 )

NOMENCLATURE OFTHEALUNITE SUPERGROUP JOHN L. JAMBORS Depat'tment of Earth and OceanSciences, Universityof British Columbia,Vancouver, British ColumbiaV6TlZ4, Canadn

Aesrnlcr The alunite supergroupconsistsof more than 40 mineral specieswith the general formula DG3(ZO4)2(OH,H2O)6,in which D is occupiedby monovalent (e.g., K, Na, NHa, H3O), divalent (e.g., Ca, Ba, Pb), ortrivalent (e.g., Bi, REE) ions, G is typically A13* or Fe3+,and Zis 56*, Ass*, or Ps*. The current nomenclature classification is unusual in that, within the temary system defined by the SOa, AsOa, and POa apices, compositions are divided into five fields rather than the three that are conventionally recommended for such systemsby the Commission on New Minerals and Mineral Names (CNMMN) of the International Mineralogical Association. The current compositional boundaries are arbitrary, and the supergroupis examined to determine the repercussions that would ensuefrom adoption of a conventional ternary compositional system.As a result ofthe review, several inconsistencies have been revealed; for example, beaverite and osarizawaite, which are commonly formulated as Pb(Cu,Fe)3(SO+)z(OH)eand Pb(Cu,Al)dSOa)z(OH)0, respectively, not only have formula Fe > Cu and Al > Cu, but the amount of substitutional Cu also is variable. Beaverite is therefore compositionally equivalent to Cu-bearing plumbojarosite. The CNMMN system also permits the introduction of new mineral namesif a supercell is present;within the alunite supergroup,the supercell is typically manifested by a doubling of the c axis to -34 A, and the effect is evident on X-ray powder pattems by the appearanceof a diffraction line_orpeak at 11 A. In addition to the supercell, however, several other departuresfrom the standardtrigonal cell with spacegroup R3n have been observed To accommodate these structural variations, thereby minimizing the introduction of numerous potential new mineral names, the possibility of incorporating a suffix modifier is explored. To keep within CNMMN nomenclature protocols, potential solutions are offered, but none is proposed. Keywortls'. alunite supergroup,nomenclature, alunite group, beudantite group, crandallite group, compositions, sffuctures. )OMMAIRE Le supergroupe de l'alunite comprend plus de quarante espbces min6rales rdpondant h la formule gdndrale DQ(TOD2(OH,H2O)6; le site D contient des ions monovalents (e.g., K, Na, NHa, H3O), divalents (e.g.,Ca, Ba, Pb), ou trivalents (e g , Bi, tenes rares), G contient en gdndral A13+ou Fe3*,et lreprdsente 56+,As5+,ou Ps*. Le systbmede classification en vigueur actuellement est anomal: dans le contexte d'un systbme temaire d6fini par les p6les SO+, AsO+, et PO+, les compositions sont rdparties en cinq domaines plutOt que trois, comme le recommande la Commission sur les Nouveaux Mindraux et Noms de Mindraux de I'Association Min6ralogique Internationale. La ddlimitation des divers champs est arbitraire. Les r6percussionsde l'adoption d'un systbme temaire de nomenclature de ce supergroupe sont ici passdessous revue. Plusieurs cas de non concordance sont relevds; d titre d'exemple, la beaverite et 1'osarizawaite, auxquelles on attribue couramment les formules Pb(Cu,Fe)3(SOa)2(OH)6 et Pb(Cu,Al)3(SOa)2(OH)6,respectivement,ont une teneur en Fe et en Al supdrieured celle du Cu, mais cette teneur en cuivre peut aussiOtrevariable. La beaverite serait donc dquivalente en composition d une plumbojarosite cuprifBre. Le systdme en place actuellement permet I'introduction de nouveaux noms de mindraux si une supermaille est manifestde. Au sein du supergoupede 1'alunite, la pr6senced'une supermaille se voit par le d6doublagede la pdriode c jusqu'd environ 34 A, et par la prdsence d'un pic ou d'une raie d 11 A dans un spectre de diffraction. En plus de la supermaille, toutefois, on observe plusieurs autres 6cafts par rapport h la maille trigonale standarddans le groupe spatialR3m. Afin d'accommoder ces variations structurales,et ainsi de minimiser l'introduction potentielle de piusieurs nouveaux noms de mindraux, il est possible d'ajouter un suffixe au nom. Afin de satisfaire aux exigeancesde la Commission, des solutions possibles au dilemne sont prdsentdes,mais aucune n'est proposde. (Traduit par la Rddaction) Mots-clds: supergroupede l'alunite, nomenclature, groupe de l'alunite, groupe de la beudantite, groupe de la crandallite, compos1t10ns.structures.

E E-mail address:jlj @wimsey com

t324

THE CANADIAN MINERALOGIST

meaningful compositional boundary, and that mutual substitutions of SO+, PO+, and AsOa are extensive The alunite supergroup consists of three mineral within the alunite supergroup.To rationalize the previgroups that, combined, contain more than 40 mineral ously indefinite boundaries between the SOa-dominant specieswith the generalformila DGz(TOr,)z(OH,H2O)6, and other membersof the family, Scott (1987) proposed wherein D representscations with a coordination numthat the boundariesbe set at 0.5 formula (SO+)zand l 5 ber greater or equal to 9, and G and T represent sites formula (SOa)z;these values arc 25Vo and 75Voof total with octahedral and tetrahedral coordination, respec- TO4, as shown in Figure 1. The proposal was accepted tively (Smith et al.1998). The supergroup consists of by the Commission on New Minerals and Mineral the alunite group, the beudantite group, and the Names (CNMMN). The effect of Scott's classification crandallite group (Mandarino 1999). In all three, the ? is that each SO+-AsO+-POa "ternary" diagram incorin_(7Oa)2is dominated by one or more of 56*, As5*, or porates five compositional fields and mineral names, P'*. The alunite group is characterizedby (SOa)2-domi- four of which are partly governed by the POa-AsOa nant minerals, whereas in the beudantite group, one of binary dividing line (Fig. 1). Nov6k et al. (1994) prothe two SOa groups is replaced by POa or AsOa. In the posed that the SOa-POa-AsOa held be divided into six crandallite group, the (7Oa)2representsone or both of units (Fig. 1), a system that requires the redefinition of POa and AsO+. Thus the change from the alunite to the several minerals. Their system has not been submitted beudantite group, and thence to the crandallite group, to the CNMMN for a vote, but its usagehas been propacan be viewed as a progression from (SO+)2,to (SO4) gated in several publications (Nov6k & Jansa 1997, (POa) or (SO+)(AsO+),and thence to either (POa)2or Novilk e/ a|.1997,1998, Sejkoraet al.1998). (AsO+)2.This progressionfollows thatinThe Systemof The nomenclature of the alunite supergroupis comMineralogy (Palache et al. I95l), wherein the alunite plex and promisesto be increasingly so if the CNMMNgroup is classihedwith the sulfates,the beudantitegroup approved system is not modified to take account of the falls within the category of "compound phosphates, crystal-structurevariations that have beenrecognizedto etc." , andtheplumbogummite group is equivalent to the occur within the supergroup. In the following discuscurrent crandallite group. In the beudantite group, a sion, the nomenclature is evaluatedto explore not only small departure from the 1:1 ratio for (?Oa):(7Oa) was what happens in a ternary compositional system, but recognized (Palache et al. l95l), and this variation is also to explore the repercussionsof designatingthe crysevident in all of the compositions listed therein. For tal-structure variations by suffixes. In the compositional example, the two compositions chosen for beudantite system herein, the ternary series is divided into three sensustrictohaveSO3contentsof 12.30and 74.82wt%a, equal fields (Fig. 1, left), which is in accord with curwhereas the appropriate value is 11.24 wt%o for rent CNMMN recommendations for naming ternary (SOa):(AsOa)= 1:1. Under the currentrules of nomen- solid-solutions that are complete and without structural clature for such a binary system,thesecompositions fail order of the ions defining the end members (Nickel to meet the "5OVorule" in that they are sulfate-domr- 1992). nant rather than precisely at the 50:50 boundary that separatesthem from the arsenate-dominantmineral speTun, AluNrrB Gnoup cies. Compositions of someof the other minerals within the group, however, exceed the 50Vo reqtirement and The minerals of the alunite group are listed in fall within the PO+-dominantfield. Table l. Figure 2 shows that little substitution of SO+ After publication of The System of Mineralogy, it by POa and AsOa occurs for members in which D is became apparentthat the l:1 ratio of anions was not a monovalent; where D is divalent, however, TOa substiINrnooucrroN

As04 Poa

AsO4

Ftc. l. Left diagram shows the nomenclature schemefor complete tema-rysolid-solutions according to recommendationsby the CNMMN (Nickel 1992). Middle diagram is the systemin current use for the alunite supergroup(Scott 1987), and the diagram on the right shows the composition fields proposed by Novdk et al (1994).

1325

NOMENCLATURE OF THE ALUNITE SUPERGROUP TABLE I MINERALS OF IIIE ALIINITE GROTJP(CURRENI USAGE) jrosite

aluite

KFq(soa),(olD5

KAls(S0rl(0rr)6

mlrojao$te

mtroaluite

N&Fe3(SOJdOH)6

NaAL(s01),(0l{)6

moniojuosite

monioaluite

(NHr)Fq(soXoH)6 hydronim jrcsite Gr.O)re(so.)doH)6 ilgentojmsite AeFq(so.),(OH)" doEllchtrite rlFq(soa),(oH). b@vqite Pb(Fe,cu[SO,),(OH,H,O)" plMbojilosit€* PbFe(SOJr(oII)o

(NlI4)Aq(SO1L(Orr)5 schlossmcherite (H'o,ca)Al'(so.)"(ol0"

oseiawaite Pb(Al,Cu)3(SO4)r(OEII2O)6

minuiite* (NaCa,K),AUsoa)a(olo' humgite* CaAl6(Sq)4(OI{)n ualthisite+ BaAL(SO)(oII)u

* Length of c unit-cell pdameter is double that of the other members of the group

tution may be substantial, and these minerals a"rediscussed separately.In the synthetic alunite-jarosite series, Fe-for-Al solid solution is complete (Brophy et al. 1962, Hartig et al. 1984); although natural members with intermediate compositions are known (Palache er al. 1951, van Tassel 1958), the degree of Fe-for-Al substitution in most occuffences is limited, and compositions are close to those of either the Fe or Al end-members. Solid solution involving K+, Na+, and (H3O)+is extensive in both alunite andjarosite (Brophy et al. 1962, Parker 1962, Brophy & Sheridan 1965, Kubisz 1960, 1970, Stoffregen & Cygan 1990,Liet al. 1992). Substitution involving K+ and Pb2* has been shown by Scott (1987) to be extensive in alunite, and by de Oliveira et al. (1996) and Roca et al. (1999) to be equally extensive in jarosite. The incorporation of NHa in ammonioalunite (Altaner et al. 1988) and ammoniojarosite (Odum er al. 1982) seemsto be mainly at the expense of K+ and (H:O)*. Apparent deficiencies in K + Na + NH+ D-site occupancy are generally attributed to

(Hso)*. unchanged: iarosite natrojarosite hydroniumjarosite ammoniojarosite argentojarosite dorallcharite

soo K:jarosite Na: natrojarosite HrO: hydroniumjarosite

Fe>Al

NHo: ammoniojarosite Ag: argentojarosite Tl: dorallcharite

AsOo

Al >Fe

geo:

minamiite Na: natroalunite,

alunite

HrO: schlossmacherite

SO,

N H o :a m m o n i o a l u n i t e

SOo

minamiite natroalunite, schlossmacherite .ttonio"lrnit"

Frc. 2. Minerals of the alunite supergroup with monovalent ions predominant as the D-site cation. Left diagram shows the current system of nomenclature, and diagram on the right shows the effects of adopting a ternary system. Minerals with Fe > Al and with Al > Fe are, respectively, jarosite - alunite, natrojarosite - natroalunite, ammoniojarosite - ammonioalunite, hydronium jarosite - schlossmacherite.Argentojarosite and dorallcharite do not have Al > Fe analogues.

r326

THE CANADIAN MINERALOGIST

Natural argentojarosite is generally near the endmember composition, although in synthetic systemsthe solid solution between Ag-Pb (t H3O) and Ag-K (+ H:O) has been shown to be complete (Ildefonse el al. 1986, Dutrizac & Jambor 1984). The Tl+ member, dorallcharite, is known only from one locality, and the mean occupancy of the D site is (TlosrKole) (Bali6 Zuntd et al. 1994).The remaining membersof the alunite group, except minamiite and schlossmacherite,contain a predominance of divalent D-site cations. As these minerals present some difficulties in nomenclature,the speciesare discussedindividually. Minamiite

regardedminamiite to be compositionally equivalent to calcian natroalunite, but with the c axis doubled in minamiite. A plot by Matsubara et al. (1998) of Na + K (alunite-natroalunite) versus Ca (huangite) does not show significant compositional gapsto be presentin the series.Becauseof small grain-size,however, X-ray data have not been obtained for all compositions. Thus, if ordering of Ca is assumedto be the sole cause for the doubling of c, the lower limit to trigger the change has not been established. B eaverite and osarizaw aite

Beaverite is typically assignedthe formula PbCuFez (SO+)z(OH)eor Pb(Cuz*,Fer+,AD3(SO4)2(OH)6 @alache The ideal formula of minamiite is (Na,Ca,K)2A16 et al. 1951, Gaines et al. 1997). Osaizawaite is the Al(SO4)4(OH)12.Chemical analysis of the type specimen dominant analogue of beaverite, i.e., Al > Fe (Taguchi gave (Nasa6K01eCa6 1atr6r+):os:Al: l1(SO4)2(OH)5 70 1961).As all compositionsof beaveriteand osarizawaite on the basis of SOa - 2, and the ideal formula of the have Fe > Cu or Al > Cu, the respectiveformulas should crystal used for the X-ray structure study (Ossakaer al. be written as Pb(Fe,Cu,Al)3(SO4)2(OH)6and Pb(Al,Cu, 1982) was determined to be (Na636K0tCaoztloz)>ogz Fe)3(SOa)2(OH)0. Most compositions of beaverite have AI3(SO4)2(OH)6.The n in the formula representsva- Cu:(Fe,Al) near 33:67, but both higher and lower valcanciesthat compensatefor the presenceof the divalent ues have been reported (van Tassel 1958, Yakhontova ion (Ca2*) in the D position. The mineral was deter- et al. 1988, Breidensteinet a\.7992). As summarized mined to-be trigonal. spacegroup R3m- a = 6.98| . c = by Yakhontova et al. (1988), the values of the relevant 33.490 A. Thus, type minamiite is compositionally ratio range from 36:64 to 25:75. equivalent to a Ca-rich natroalunite, but the c axis in As for beaverite, most analysesof the Al > Fe anaminamiite is doubled becauseof oartial order of the D logue, osarizawaite,indicate Cu:(Al,Fe) near 33:67. cations. This results in a superstructure that yields a However, Paar et al. (1980) reported an occurrence of characteristicpowder-diffraction line (or peak) at about osarizawaite for which the ratio is 25:75. [The mineral 11 A (Okadaet al.1987). described by Cortelezzi (1977) as osarizawaite seems Although type minamiite is Na-dominant (Ossakaer to be a Pb-Cu-rich alunite, requiring re-analysisl. al. 1982), the same authors subsequently synthesized the The compositional variations in beaveriteCa-dominant analogue, and they explicitly referred to osarizawaiteindicate that the ratio of divalent (Cu2+and the end-member composition of minamiite as CaosAl3 Zn2+1to trivalent (Fe3+and Al3*) ions need not be (SO4)2(OH)6(Ossaka et al. 1987, Okada et al. 1987). strictly 33:67. In synthetic beaverite-plumbojarosite, Nevertheless,although Ossakaet al. (1987) showed that substitution of Cu2* for Fe3* was found to be complete extensive substitution involving Na-K-Ca occurs rn over the rangeCu:Fe = 0:100 to Cu:Fe = 33:67 (Jambor natural minamiite. none of the comoositions indicated a & Dutrizac 1985). Up to 0.76 mol (Zn + CLr)has been predominanceof Ca as the D-site iation. reported for Pb-rich alunite (Scott 1987). For corkite, Most minerals of the alunite group have D occupied which is also in the supergroupand is describedfarther by a monovalent ion, and for these minerals the substi- below, substitution of 0.42 to 0.64 mol Cu has been retution of SOa by POa or AsOa is exrremely limited ported by de Bruiyn et al. (199O) and Tsvetanova (Fig. 2). Adoption of a ternary system SOa-POa-AsOa (1995),respectively. for theseminerals thereforewould not affect the nomenGiuseppetti & Tadini (1980) solved the structure of clature of the existing species.The three divalent ele- osarizawaite of composition Pb(Alr 62Cu6esFes3s ments predominating at the D site in the alunite-group Zn6s2)23ss(SO+)z(OH)o in space group R3nz, which minerals are Ba. Ca. and Pb. requires that the Cu and Al-Fe be disordered. X-ray powder-diffraction patterns reported for natural Huangite and wahhierite beaverite-osarizawaite have so far conformed to a basic cell of a = -7, c =-tZ A, but as Cu2* decreases, Pb The mineral with Ca predominant, ideally Caa5Al3 content also must decreaseto maintain charge balance. (SO4)2(OH)6,was named huangite, and that with Ba At some point, therefore, the unit cell must transform to predominant, ideally Ba65AI3(SO4)2(OH)6,was named the doubled c of -34 A that is commonly accepted as walthierite (Li et al. 1992). X-ray powder patterns of characteristic of plumbojarosite. In synthetic plumboboth minerals show an 11 A diffraction effect, indicar jarosite-beaverite,however, doubled cells appearedraning that their c axis is doubled to 33-34 A. Acceptance domly along the series (Jambor & Dutrizac 1983). of huangite as a new mineral indicated that the CNMMN

t327

NOMENCLATURE OF THE ALUNITE SUPERGROUP

In terms of composition, beaverite is a Cu-rich plumbojarosite. All occurrences of natural plumbojarosite have so far been reported to have a doubled c axis, but in synthetic hydronium-b-earing plumbojarosite, the c may be either 17 or 34 A, apparently depending on whether Pb is ordered or disordered. The degreeof order can be variable, and this is readily indicated by variations in the intensity ofthe 11 A diffraction line or peak in the X-ray pattern. Schlossmacherite

TABLE 2 MINERALS OF TIIE BEUDANTITE GROIJP(CIJRRENTUSAGE)

corkite

b@dutite

PbFe[(P,S)OJ,(OH,H,o)6

rbre3[(As,S)OJdOH,HTO)5 hinsdalite

hidalgoite PbAl3(As,S)O4],(OH,n kenmlitzite

PbAl3t@,s)orl,(orlE

0)6

srAl,lG,s)O-l,(oEH'o)"

(sr,ce)Al.t(nqs)oJloH,Eo)6

woodhouwite CaAl3[(P, S)O.],(OH,II'o). gallobeudmtite PbGa3(As, S)O4]2(OH,H,O)6

Either As or S my be predominaot in (A5,S)O4, ud predonimt in (P,S)O,

Schlossmacherite contains substantial amounts of both SO+and AsO+; thus the mineral has been variably assignedto the alunite group (Gaines et al. 1997) or the beudantite group (Mandarino 1999; Table 2). Recalculation of the only analytical data available (Schmetzer et al. 1980) gives [(H3O)6:zno zsCao26Na667K6.05 516s1Ba6sll;1 oo (Alz eaFe602Cu606)>302 [(SO+)rsr (AsO+)0.+ql(OH)s zo on the basis of TOa = 2 and after reassignmentof Cu from the D to the G position. The formula conforms with that generalizedby the authors (Schmetzer et al. 1980). The ratio of SOa:AsO+is 75.6:24.4,thus placing schlossmacheritein the SOaAsOa-dominant field in Figure 2. The mineral is therefore the Al-dominant analogue of hydronium jarosite. Cell dimensions ate a = 6.998, c = 16.67 A. Tue CneNoallrrE GRoUP The minerals of the crandallite group are listed in Table 3. Nine of the members have Ba. Sr. or Ca domi-

0)6

srubqgite

eiths

P or S may be

nant in D (threemembersfor eachofthese cations),four members have Pb dominant, and the remainder has REE, Ca, or Th dominant. As in the alunite group, a primary distinction rests on whether the proportion of Fe is greater than that of Al or vice versa. Ba predominance The Ba-dominant member of the alunite group is walthierite BaosAI3(SO4)2(OH)0,and its relationship to other Ba-dominant members of the alunite supergroup is shown in Figure 3. AlthoughLiet al. (1992) obtained a = 7.08, c = 1'7.18A by electron-diffraction study of walthiedte, X-ray powder pattemsof bulk samplesshow an ll A diffraction peak, requiring that the length of c be doubled. The inconsistency was attributed to disorder induced by the electron-diffraction beam.

TABLE 3. MINERALS OF THE CRANDALLITE GROI.]P (CTJRRENTUSAGE) Fe>Al

Al>Fe gorceixite BaAlj(Po1)eq.orD(oII)6 qmdallite CaAl3[PO3(O1D(OI'rJ],(OID6 goyuite STALIPO3(On(OH)rrl,(OH)6 plMbogwite PbAl3@o4)'(oH,r{'o)6 florencit€-(Ce) CeAIr(POo)r(OHL florscite-(La) LaA[(PO1)b(OFD6 floretrcite-(Nd)

NdAteo4)'(or{)6 waylmdite (Bi, Ca)AL@Or,SiO,),(OH)e

rsnogor@ixite BaAL(AsO{XASO3.OID(OlI)6 reocmdallite CaAl3[AsO j(On(OID;)]r(OIt6 dmogoyete SrAl[AsO.(O,"(OH)'J] {0II)6 philipsbomite PbAl3(AsO4XOH,IIP)6 trsofloretrcite-(Ce) CeAl(AsO'),(OtI)c "dsoflorflcits(La)"* LaA13(As04),(OlD5 "usmoflorscite-(Nd)"* NdAl3(A504)2(0II)6 "ilsrcwylmdite"f @i,Ca)Al (AsOo),(OH,IlO).

dusrertite

BaFes(AsOJdOH)6

b@uite

Srre(PO,),(OH,H,O)6 kintorcite

pbpe@o.),(oEEo)6

wEnitite

zakite

BiFeeoXoH).

e'.lsttssite

Othes

(ThPb)ALeO4,SiO4)'(oIr)6 springcr@kite

BaV3eOaL(OH,ILO)6

* not CNMMN-approved, but included here for completeness.

PbFq(.csoaXOH,ILo)6

t328

THE CANADIAN MINERALOGIST

soo

soo

Fe > A l

dussertite AsOo

Poo sorceixite

POo

arsenogorceixiter/

arsenogorceixite

\

walthierite

AI

soo

soo

FIc. 3. Minerals of the alunite supergroupwith Ba as the predominant D-site cation. Left diagram shows the cunent nomenclature, and diagram on right shows the effects of adopting a temary system. "Weilerite" is not a valid species(see text).

The main concem in the Ba-dominant group is substitution atthe T site; for walthierite, such substitution was not detected (Li et al. 1992). Likewise, compositions of gorceixite are typically close to, or at, the POa end-member. For arsenogorceixite,however, the type material shows AsOa:POq = 74:26 (Walenta & Dunn 1993).The AsOa (arsenogorceixite)and PO4(gorceixite) end-membershave been synthesizedby Schwab et al. (1990,1991). "Weilerite" is included as a valid species in some modern compilations (Clark 1993, Gaines er al. 1997), but the mineral was discredited by the CNMMN (IMA 1968). As no quantitative chemical data had been published for the mineral (Fleischer 1962, 1967), the posrtion on the AsO+-SO+join is uncertain, and the mineral may have been identical to the more recently described arsenogorceixite.The removal of "weilerite" from the system is appropriate, thus leaving a gap between gorceixite-arsenogorceixite and walthierite (Fig. 3). Dussertite is the only known member with Fe > A1, thus theoretically allowing six additional names to be introduced in the left diagram of Figure 3. In a ternary system, the potential six are reduced to two.

Ca predominance All Ca-dominant minerals in the alunite supergroup are also Al-dominant (Fig. a). Numerous compositions of crandallite are near that of the POa end-member.but substitution of S for P is extensive, and compositions extend well into the current field for woodhouseite (Stoffregen & Alpers 1987, Spcitl 1990, Li et al. 1992), including compositions with formula SO+> PO+ (Wise 1975). For arsenocrandallite(Walenta 1981), the type material contains substantial Si in the Z position: (Aso qqPozsSioz6).The As:P ratio is 57:43, which is well within the field for arsenocrandallite. Adoption of a temary system for the Ca-Al-dominant minerals would pose the problem of whether woodhouseite or huangite is to be used to designatethe SOa end-member. Type woodhouseite has POa:SOa= 54:46 (Lemmon 1937), but it has long been accepted that either formula S or P can be predominant. The only composition given by Palache et al. (1951) has P > S, but more modern analyses,by electron microprobe, of woodhouseite from the type locality have shown both formula P > S and S > P (Wise 1975). Moreover. the

1329

NOMENCLATURE OF THE ALUNITE SUPERGROUP

Soo

AsO, POo c r a n d al l i t e

arsenocrandallite

crandallite

a r s e n o c r a n d aIli t e

woodhouseite

soo

soo

Frc. 4. Minerals of the alunite supergroupwith Ca as the predominant D-site cation. Left diagram shows the cu[ent nomenclature, and diagram on right shows the effects of adopting a ternary system.

current CNMMN-approved nomenclature extends the composition of woodhouseite well into the SOa-dominant field (Fig. a). Accordingly, woodhouseitehas been adopted to designatethe field with S > P in the ternary system (Fig. 4). Sr predominance The minerals with Sr predominant in the D site are shown in Figure 5. The two minerals with Fe > A1 are benauite and an associatedunnamed mineral richer in SOa (Walenta et al. 1996). The empirical formula for benauite is (Sre67Ba616Pbse7Ca661Koot):,0sz (Fezgo Alo o:):z s: [(PO+)r+s(SO+)o as(AsO4)0 04](OH,H2O)8 3, which was simplified by Walenta et al. (1996) as SrFe3(POa)2(OH,HzO)0. The proportion of atoms in the T site, however,is P:S:As = 74:24:2, which placesthe mineral just beyond the boundary designatedby the simplified formula, and into the field of the unnamed SOaricher mineral (Fig. 5). The rangein compositions(wt7o) reported by Walenta et al. (1996) is PzOs l'1 .98-19.93 (mean 18.53usedto calculatethe formula), As2O50.640.94 (mean0.78), SO36.24-:7.37(mean6.79),and thus

the compositions probably straddle the aforementioned boundary. The formula of the unnamed mineral rs (SrzsrBaooo)Fe:r:(PO+)r:s(SO+)oos(OH,H2O),,which places the mineral well beyond the field of benauite. If Scott's (1987) nomenclature systemis to be adheredto, benauiterequiresredefinition as ideally SrFe3[e,S)O4]2 (OH,H2O)6.The field shown Bboccupied by benauitein Figure 5 would therefore remain vaguint. For the Al-dominant group. (Fig. 5), goyazite and svanbergitehave long beenusedto indicate the POa and PO+-SO+ minerals, respectively. For svanbergite, the compositions listed in Palache et al. (1951) have formula P > S, but compositions with S > P also are known (Wise 1975); the CNMMN-approved system accepts that either S or P may predominate (Fig. 5). Type arsenogoyazite(Walenta & Dunn 1984) has AsOa:PO+ = 64:36, but the pure end-member has also been described (Zhang et al. 1987). Kemmlitzite (Hak et al. 1969) was regardedby Fleischer (1970) to be the arsenate analogue of svanbergite; analysis of type material, however, gave [(AsO+)oes@Oa)6a2(SOa)s3e(SiOa)s1e] 11e3,which places kemmlitzite in the field of arsenogoyazite(Fig. 5). Re-examinationof the holotype speci-

1330

THE CANADIAN MINERALOGIST

soo

soo

Fe

/

a"n^uit a

Poo covazite

AsOo PO,

^"tloo'o'^'n"/

arsenogoyazrte

\

AI

soo

soo

Ftc 5. Minerals of the alunite supergroupwith Sr as the predominant D-site cation. Left diagram shows the current nomenclature; unnamed mineral is that of Walenta et al ( 1996), point "a" is the composition of type arsenogoyazite,and "k" is that for t)?e kemmlitzite. Right diagram shows the effects of adopting a ternary system.

men of kemmlitzite has shown it to be zoned and inhomogeneous (Novdk et al. 1994). Kemmlitzite predates arsenogoyazite,and the latter was specifically defined as occupying the AsOa-dominant field because adoption of the current (Scott) system of nomenclature had the effect of entrenching kemmlitzite as representative of SOa-richcompositions(Fig. 5). Such a SOa-richcomposition has been reported by Nov6k et al. (1997). ln view of the reported zoning and lack of homogeneity in type kemmlitzite, retention of the name arsenogoyazite would seem,arguably, to be preferred. In a ternary system, therefore, the most appropriate names are considered to be goyazite, arsenogoyazite,and svanbergitefor the Al-dominant members, and benauite for the Fedominant mineral with POa greater than AsOa or SOa

(Fie.s).

Bi predominance Minerals with Bi predominant in D are shown in Figure 6. Analyses of waylandite (Clark et al. 1986i)and

zairite (van Wambeke 1975) are within the designated compositional fields. "Arsenowaylandite" is an unofficial name (Scharm et al.1994) that has not been submitted to the CNMMN for approval. The anhydrous composition for ZOa = 2 vaies from (Bi68esr005 Caoos)>os7 (Al2 e6Fe610)[(AsOa)1or(SO+)o zz@O+)o rr] to (Bi6 eeSrs65)116a(A12 e6Fe6 ea)[(AsO+)r sr(SO+)ors,. Pb predominance Figure 7 is illustrative of the profusion of mineral names possible as the compositional fields within the alunite supergroup are frlled. For Fe-dominant minerals, plumbojarosite, corkite, and beudantite have been in use for many years, and corkite and beudantite occupy the bulk ofthe field. Segnitite (Birch et al. 1992) and kintoreite (Pring et al. 1995) occupy the SOa-poor fields that opened only as a consequenceof the adoption of the current CNMMN-approved system of nomenclature.For Al-dominant minerals,plumbogummite and hinsdalitehave beenknown for many years(Palache

r33r

NOMENCLATURE OF THE ATUNITE SUPERGROUP

soo

SO,

AsOo POo waylandite

"arsenowayl

SO, FIc. 6. Minerals of the alunite supergroup with Bi as the predominant D-site cation Left diagram shows current nomenclature, and that on the right shows the effects of adopting a temary system. "Arsenowaylandite" (Scharm et aI. 1994) is not an approved name.

et al. l95l). Hidalgoite has the formula ratio SOa:AsOa = 59:47, and the mineral was specifically named as the arsenate analogue of hinsdalite (Smith er al. 1953). Philipsbornite (Walenta et al. 1982) is somewhat unusual in its high Cr content in TO+, for which = 76:19 5. The small freld at the SO+ AsOa:CrOa:SO+ apex (Figs. 7, 8) has beenreferred to as "plumboalunite" (Novrik et al. 1994) and "plumbian alunite" (Sejkora er al. 1998). However, if it is acceptedthat (Cu + Zn) substitution is non-essential,then the Pb-Al sulfate in this field is osarizawaite. Adoption of a ternary system for the Pb-dominant members would pose some difficulties. In the system with Fe predominant in G, corkite and beudantite have priority, thereby requiring the abandonment of the recently named kintoreite and segnitite.In the systemwith Al predominant in G, plumbogummite and hinsdalite have priority. Osarizawaite was named as the Al analogue ofbeaverite, but the requirement, as will be discussed, is that neither be retained as a mineral name. Plumbogummite and hinsdalite thus occupy the PO+-

SO+join as shown in Figure 7. The consequenceis that the field for hidalgoite, which includes the composition oftype hidalgoite, would be overlappedby the new field for hinsdalite. Therefore, as was done for arsenogoyazite,philipsbornite would be retained to designate the As-dominant member. REE predominance The REE-bearing minerals are characterized by a predominanceof trivalent cations in D, and it has been suggested(Scott 1987) that such minerals be classified as the florencite group. Accepted minerals within the group (Fig. 9) are florencite-(Ce), florencite-(La), fl orencite-(Nd), and arsenoflorencite-(Ce). Compositions corresponding to the La and Nd analogues of arsenoflorencite-(Ce)have been reported by Scharm er al. (1991,1994), but the new nameshave not been submitted to the CNMMN for approval. Adoption of a ternary system for the REE-dominant members would not affect the current nomenclature.

1332

THE CANADIAN MINERALOGIST

soo

soo

rmbojarosite

plumbojarosite

/

kintor"it"

segnitite

beudantite

soo Po.

Poo ohilivsbornite \lumbogummite

AI

,/

plumbogummite philipsbornite

arizawaite

soo

soo

Ftc. 7. Minerals of the alunite supergroupwith Pb predominant as the D-site cation. Left diagram shows current nomenclature. The mineral at the A1-SO4 apex is osarizawaite if it is acceptedthat Cu is non-essential.Compositions are shown by Sejkora et al. (1998) as extending into this field, which has been named 'plumboalunite" by Nov6k e/ a/. (1994) Diagram on right shows the results of adopting a temary system.

Other minerals

Other analogues

Eylettersite is considefedto be a Th-dominant member of the crandallite group, with P as the main cation in 70+ (Table 3). Hqryever, the calculated formulas also incorporate substantiai(H+O+)in ZOa,and show excessive Al (>3.5 mol) for G (Van Wambeke 1972). Such a G-cation excess is untenable for an alunite-type mineral, and eylettersite is considered to need restudy. Gallobeudantitewas defined ashaving subequalvalues of AsOa and SO+, and a predominance of Ga in G (Jambor et al. 1996). However, ZO4 compositions extend into the fields encompassedby Ga analogues of segnitite, corkite, and kintoreite, thus permitting the introduction of three additional new names for the Gadominant minerals. The distribution of 7Oa values in the Ga minerals is such that adoption of a ternary system would not reduce the number of potential new names.Also presentwith gallobeudantiteis the Ga analogue of arsenocrandallite.

Numerous members of the alunite supergroup have been synthesized.Among those not known as minerals are the Rb* and Hg2+ analogues of jarosite; the Hgdominant and Pb-dominant phases described by Dutizac & Kaiman (1976) and Dutrizac et al. (1980) are indexed with c = 33 A, but no diffraction line at 11 A was reported. Mumme & Scott (1966) also noted that the 11 A line is absent from their synthetic plumbojarosite. Tananaevet al. (1961) preparedanalogueswith Na, K, Rb, H3O, and NHa in D, and Ga in G. Only small amounts of various divalent metals other than Cu2+and Znz* have been reported to be taken up by the alunite supergroup. The In3+, V3*, and Cr3* analogues have been synthesized (Dutrizac 1982, Dt;trizac & Dinardo 1983,T,engauer et al. 1994):up to 0.6 mol V3* and 0.18 mol Crr* have beenreportedin natural gorceixite (Johan et al. 1995), and Vr+ predominates at the G site in

NOMENCLATIJRE OF THE ALUNITE SUPERGROUP

r333

Stnucrunel Asprcrs The classification of the alunite supergroup discussedto this point hasbeenmainly chemical, and structural aspects have been largely ignored. Numerous single-crystal X-ray structure studiesof the supergroup have been done, and the findings are summarized in Table 4. Most of these studies have revealed that the minerals have trigonal symmetry, with a x 7, c = 17 A, space group R3m, but several exceptions have been documented. The structures of alunite, jarosite, and plumbojarosite were first determinedby Hendricks (1937), who a -^ n)\-/4 concluded that becausealunite andjarosite show strong pyroelectric properties, their spacegroup must be R3m plumbojarosite was deterrather than R3m. In co_ntrast, mined to belong to R3m, and the Pb atoms showed an ordered arrangement that had the effect of doubling the c axis, i.e., the typical cell of a = 7, c = 17 A increased loa=7,cx34A. In addition to the data given in Table 4, severalstudies of synthetic analogues have been done, either by single-crystal or Rietveld methods. All of the studies, both on minerals and their synthetic equivalents,and on those not known as minerals, confirm that the basic topology of the structure remains the same regardlessof SO, chemical composition. Even where a monoclinic or triclinic system has been established, the structures are pseudotrigonal. Frc. 8 Compositionsof mineralswithin the Pb-dominant strongly The basic structural motif of the supergroupconsists field, asin Figure7, illustratingtheextensiverangeof ZOa solid solution Solidcircles:abridgeddatafrom Sejkoraer of TOa tetrahedra and variably distorted R-cation octafrom Rattrayel hedra, the latter corner-shared to form sheets perpenal. (1998);opensquaresarecompositions al. (1996\. dicular to the c axis. Substitutionsinvolving G therefore mainly affect the a dimension, and a increasesas Fefor-Al substitutionincreases.The ZO+tetrahedra,which are aligned along [001], occur as two crystallographispringcreekite (Table 3 ; Kolitsch et al. 1999a). Substan- cally independent sets within a layer; one set of TOa tial Sbs*-for-Fe3+ substitution in dussertite has been points upward along c, and this set altemates with another pointing downward. The oxygen and hydroxyl documentedby Kolitsch et al. (1999b). The vanadate(Tudo et al. 1973), chromate (Powers form an icosahedron,amidst which is the D cation. For et al. 1976, Cudennec et al. 1980), and selenate compositions with identical T tn TOq,the length of c is (Dutrizac et al. 1981, Breitinger et al.1997) analogues mainly influenced by the size of the D cation. Symmetry changes arise principally becauseof orhave been synthesized,and substantial Cr has been reder-disorder relationships among the 7Oa tetrahedra,or ported in the Z site in philipsbornite (Walenta et al. 1982). Some analysesreveal Si, which has been as- because of order-disorder or distortion involving D signedeither as SiO+(as in kemmlitzite and eylettersite) sites. To maintain the space grotp R3m, the SO+-PO+AsO+ tetrahedra must be disordered; ordering, as has or as amorphous SiO2. Other substitutions, such as Nb (Lottermoser 1990) and CO3, are possible but generally been found in corkite and gallobeudantite, reduces the not significant; analysesshowing percentagequantities symmetry to R3m. A second principal effect is related to the presence of CO2 (e.9., Frirtsch 1967) predatethe widespreaduse ofthe electron microprobe, and the high valuesreported of a divalent cation in D. In such a case, the standard jarosite-typeformula D*G3*3(SOa)2(OH)6 becomeseither in a few older analysesmay be the result of inclusions -7|+162+ of carbonateminerals. 3(Soa)2 t(soo)r(og)ul, (ordered) or, D2* 112G3* Substitution of Cl for OH is rarely reported and, (OH)o (disordered). Thus in plumbojarosite, for example, where detected,amounts are small. Accounts of F sub- the D-site cation is Pb2*,and to maintain electroneutralstitution are more cofirmon, and up to 4.'l wtVo F has ity, half of the D sites are vacant. Ordering of the Pb been determined to be present in gorceixite (Taylor et and thesevacanciesproducesa supercell,which is manifested as a doubling of the c axis to 33-34 A. al. 1984\.

r334

THE CANADIAN MINERALOGIST

soo

Soo

fe>Al

Fe)Al

AsOu

AsOo PO, arsenoflorencite {Ce)

florencite-(Ce) arsenoflorencite-(Ce) _(Nd)* -(La)*

Al >Fe

Al )Fe

sou

500 Soo

Frc 9. Minerals of the alunite-jarosite supergroup with REE as the predominant D-site cations. "Arsenoflorencite-(La)*" un6 "arsenoflorencite-(Nd;*" *" names (Scharm et al. 1991, 1994) that have not been submitted to the CNMMN for a vote. Diagram on right shows the effects of adopting a temiuy system.

A third principal effect is related to the charge of the ZOa group. Where ZOa is POa or AsOa rather than SOa, an extra proton is required to maintain charge balance. In crandallite, Blount (1974) concluded that the formula takes the form CaAl3[PO30 1p(OH) 1p]2(OH)6, thereby achieving compensation by having partial substitution of hydroxyl for oxygen of the PO+ tetrahedron. Radoslovich (1982), however, determined that such a configuration would not be permitted in gorceixite because of the larger size of Ba. He therefore concluded that the formula of gorceixite is best represented by BaAl3eO4)(PO3.OHXOH)6. It is possible, therefore, that order may exist without the necessity of having the 7 site occupied by different elements (As, P, S). Furthermore, the crystal-structure determination (R = 0.0195) of unnamed PbFe3GO4)2 (OH,H2O)6 by J.T. Szymanski (Table 4) revealed that distortion at the Pb sites leads to two indeoendent Pb atoms in the structure. Thus, a ruperstru.ture is formed in this mineral without the necessiw of havine D-site vacancres.

Szymafiski (1985) has pointed out additional aspects relevant to the crystal chemistry of the alunite supergroup. Among these is the report by Loiacono et al. (1982) that the space group of their natural alunite is R3m. For hydronium jarosite - schlossmacherite, Szymafiski(1985) pointed out that "... the hydronium ion, whether it remains in its original H3O+ form or whether it is reduced to water, cannot have a centre of symmetry and cannot be located at a centre of symmetry. Any hydronium-for-cation substitution will destroy the symmetry, and reduce the space group to R3m or lower. This is not to say that the hydrogen atoms of the hydronium ion or the water molecule cannot be statistically distributed or dynamically disordered, and hence give the structure the overall appearanceof having a centre of symmetry." As with hydronium, the NHa ion cannot be accommodated within R3m (Arkhipenko et al. 1985, Sema et al. 1986).However, as discussedby Szymahski (1985), in the absenceofincontrovertible sfuctural data,the space group of alunite minerals should be consideredas R3m.

I 335

NOMENCLATURE OF THE ALUNITE SUPERGROUP TABLE 4 SINGLB-CRYSTALX-RAY STRUCTURESTTIDIESOF MINERALS IN TIIE ALUMTE SUPERGROTJP a,L

Compositio4oments

alunite natroalsite mimiite JilOSrte

@mpo$uoDnor grvm (Nao,slqrJAls(SO),(OH)6 (Nao*Ig rCa".rrtrnr)Alr(Sq)r(O$. not givq; synthetic not giv€n;mtural

gorccixite qmdallite woodhouseite goyuite svmbergite os@awilte plmbojaosite

bfldatite

corkite hin$datite galobildantite plmbogwite kirio!eite du$ertite nnnnrBed florencite-(Ce)

[email protected](OID6 Cao"At*P,*; CaAlrlPO3(O@(OIDdlr(OID6 not givm; CaAI,(PO*)(SO4XOID5 not givm; SrAI3[PO3(O,"(OID,J]2(OI06 rot givs; SrAL(POa)(SO4XOID6 Pb(Al, .rCuo,rFeo,lq d(SO1)r(OH)5 trot givd; structurefomula u below PboJq *So*; Pb[Fe(sOJr(o]D51, Pboe(FqsCuo@zrboilSO).lAsOJ.'(0II)6d Pb,*(Fe,"Alo*)(AsOJr odsO4te(OEHp)6 Pb,.(FerorAlorr)(AsO,)rB(SO)0'(OH,ILO)6 Pbo*(Fer"Cuo *)(PO4)I.6(SOJ0*(Asq)o ml(OH)6q PbALleo,), 3r(so{Lcl(orr,H,o)5 Pbl u(Gal fAli laFeo szno ilGei 6XAsO{)r&(SQI *(OIO" " PbALI(PoJ, s(Aso)o,J(oI!H,o). PbFq(POJ',(AsOn)ur(SO4)03(OII,Hp)6 Ba(Fd.' s4sb*o,b)3(Aso"),(ollHp). PbooFq n,@Oo), ,,(SOn)o,(OI1I4O)' ,, (CeoalrorNdo,rSmo*Car*)A1r(PO,l(OlD.

c, A spaegroup Ref.

17 27 6970 16095 6 990 33 490 6 981 l7 l9O 1 305 17268 7304 monoclinic monoclinic* 16192 7 005 6 993 16386 ? 015 16558 16567 6992 70'15 17 I 33 60 720 7 3055 33 675 1 339 17 034 7 3l5l I7 0355 triclinic** 7 ZaO 16821 16789 7 029 1703 7 225 16761 7 039 't 331,0 16885 l'l 484 7 410 7 28A5 33 680 6912 16-261

Rim Bm R3m Rlm Rim Cm Rlm Bm Rlm Rim Rlm Rim Em R1m Rlm R3m Rim R3m Bm Bm Rlm R3m Bm

1 2 3 4 5 6 7 8 9 10 5 1l 12 13 14 15 15 16 f'l 18 17 '19 20 21 22

* a 12 195,b 7 o4o,c ? 055A 0 125 l0' S@slroBlEnchdd(1989) * * q 7 3r9, b 7 309,c 17032A, a 90 004,p 90 022,y 119974" Referenes:7 Wngetdl.(1965),2 Okedaetal0982), 3 O$ska"t4l. (1982),4 Mf,chetti&Sabeli(1976), 5 Kato E Blout(1974),9 Kato097l, 1977)' l0 Ksto ? Radostovich(1982), &M{m(1977),6 Radoslovich&Slade(1980), (1971,198?),1I Giuseppetti (1937),who alsogavereflrltsfor aluite edjeositq 13 Szf& Tadini(1980),12 Hendricks & Tadini(1987), 17Kolitsh miski (1985), 14 Giuseppelti & Tadini(1989), 15 SzJmiski (1988), 16 Giuseppetti etdl.(1999c), 78 lanboretal (1996), 19 Khdirueta/. (1997), 20 Kolitsch€tdl (1999b),2l IMANo 93-039, 22 Kato(1990)

Formula calculations

Significance of the supercell

Various methods have been used to calculate the formulas of minerals in the alunite supergroup.Among the most common are normalization to 14 oxygen atoms, or to TO4 = 2. Some authors (e.g., Scharm et al. 1991, Novrik & Jansa1997, Sejkoraet al. 1998)have used G3(IOa)2 = 5. Use of TOa - 2 is preferred here becausenonstoichiometry in both D and G is common, particularly in synthetic samples. To paraphrase Szymafrski(1985), use of ZO4 = 2 has a sound structural basisbecause"... it is inconceivableto visualizea stablejarosite structure with vacanciesin the [TOa] layers as well." A second difficulty in formula calculations is that hydronium cannot be determineddirectly. This problem is not significant for most of the minerals in the supergroup, but it doesemphasizethe precarious statusof the sole composition available for schlossmacherite, for which D is [(H:O)o :zno z8ca026].Ia0 07Kos5Sr6slBaos1l.

Regardless of symmetry variations (Table 4), the topology of the unit cell of all minerals in the family can be related to a rhombohedral cell that has hexagonal dimensions of a x 7, c N l7 A. The development of a supercell with c = 2 X l7 A can arise becauseof ordering ofD-site cations, regardlessof whether D contains vacanciesor is filled completely. Moreover, order does not require the presence of a combination of monovalent and divalent cationsin D, but can take place solely with monovalent ions (Dutrizac & Jambor 1984). Many rapidly crystallized syntheticproducts show signs of an ordered structure,and it is inconceivablethat some of the commonly more slowly crystallized natural minerals would not be ordered.On the other hand, synthetic Pb-H3O systemsshow various degreesof disorder, and synthetic plumbojarosite without detectablediffraction effects due to a superstructureis well known. At the opposite extreme, in their X-ray structure studies both

r336

THE CANADIAN MINERALOGIST

Hendricks (1937) and Szymairski (1985) used plumbojarosite from the Tintic Standardmine, Utah, for which an exceptionally high degreeof order was found; in the case of the latter author at least, the material was selected specifically becauseX-ray powder patterns had previously indicated an exceptionally high degree of order to be present. If minerals in the alunite supergroup are to be named simply on the presenceor absenceof the c-axis superstructure, the potential for the introduction of "trivial" names will increase enormously. I believe that detection of the superstructurewill become more common once there is an increasedawarenessof the sisnificance of the I I A powder-diffraction line or peak. Recol,rlr,rewoauoNsoN Noprexcr-erunE If the proliferation of mineral names in the alunite supergroup is to be avoided, two properties or factors must be addressed:(a) crystallographic, and (b) chemical. In some casesthese factors are independent,as for example minamiite and as-yet-unnamedmineral IMA No. 93-039, both of which are distinguished from previously named minerals solely on the presence of standard cell versus supercell relationships. The fundamental topology of the alunite structure remains the same regardlessof space group or symmetry changes. The fundamental cell is rhombohedral, space group R3m, with a x 7 and c = 17 A as expressedwitli hexagonal parameters.Slight distortion of this cell can lead to orthorhombic (Jambor & Dutrizac 1983), monoclinic (Radoslovich 1982),or triclinic (Szymairski 1988)polymorphs, and patterns of order of various kinds can lead to a doubling of c. Crystallographic aspects Various systems have been used to accommodate structural changeswithin a mineral group while maintaining a comprehensible nomenclature. For example, Pring et al. (1990) used the CNMMN-approved name baumhauerite-2a to designate a silver-bearing mineral having a superstructurederived from a baunihauerite-

like structure.For minerals of the alunite supergroup,a similarly simplified system of nomenclature system could be adopted such that: (a) no modifier accompany the mineral name if the unit cell has not been determined, or if only one structure type is known; (b) if there is a need to distinguish between the conventional cell and the supercell, the former should be designated lc, and the latter 2c, each followed by the standard (CNMMN-approved) abbreviation for the unircell type [1: hexagonal,rt: rhombohedral, M: monoclinic, A: triclinic (anorthic), erc.l. Note that it is important to retain the c to avoid confusion with the nomenclature designations for polytypes. Thus, for example, rhombohedral plumbojarositelacking a superstructurewould be named plumbojarosite-lcR, and that with a superstructure would be plumbojarosite-2cR. Monoclinic gorceixite remains as gorceixite, but if a doubled c axis for the monoclinic cell were to be found, the nomenclature would distinguish gorceixite-7cM and gorceixite-2cM. It has not yet been proved that gorceixite with a conventional rhombohedral cell exists, but such a mineral would simply be designatedas gorceixite-lcR. Adoption of such a system would involve the following nomenclature changesor revisionsin: 1. Minamiite, which is fundamentally different from natroalunite only in order-disorder relationships, is natroalunite-2cR. 2. Monoclinic gorceixite, assuming that the rhombohedral form also exists, would not be entitled to a trivial name. 3. Triclinic crandallite, reported by Cowgill et al. (1963), ifverified, would be designatedcrandallite-lcA. 4. Triclinic beudantite reported by Szymafiski (1988) would be designatedwith the suffix lcA. 5. Unnamed rhombohedral mineral IMA No. 93039 would likewise adopt the appropriate already established name, together with the suffix 2cR. 6. Beaverite is cuprian plumbojarosite without superstructure reflections. Beaverite is therefore cuprian plumbojarosite-1cR. 7. Orthorhombic jarosite with the doubled c and composition K6e6Fe2qo(SO+)z(OH)6 16 (Jambor & Dfiizac 1983) is jarosite-2co.

TABLE 5 SI]MMARY OF POTENTIALLY CHANGED NOMENCLATURE FOR A TERNARY COMPOSITIONAL SYSTEM

Previous nue

Prwiously wmed

mitmiite

nahoalunil€-2cI

triclinic qedallite

crmd.llite

1cl

huugite

woodhoureite-2cR

triclinic beudmtite

b@ddtite-

lc,{

kef,nrlitzite

menogoyuite

rhombohedral kiatoreite

b@vqite

cupriu

corkite2cR jtosite-2co

kintoreite

@rkite

plufi$ojdositelcR

segilwe

bodmtite

hidalgoite

hinsdalite

osrizawaite

hinsdalite

orthorhombic juosite

r337

NOMENCLATURE OF THE ALUNITE SUPERGROUP

Compositional aspects Two approachesare used, namely, (a) an evaluation in terms of the existing (Scott 1987) system of nomenclature, and (b) an evaluation on the basis of a ternary compositional system.Neither systemtakesinto account possible miscibility gaps, which have not been proved to exist within the alunite supergroup, although such a possibility is highly likely. For example, Stoffregen & Cygan (1990) have argued that a miscibility gap may exist on the simple alunite-natroalunite binary join if theseminerals crystallize under equilibrium conditions. Under nonequilibrium conditions, however, a gap is not evident.Numerous proposalsfor miscibility gapsamong minerals other than the alunite supergroupcan be found in the literature, but in many casesthe purported gaps simply reflect plots of existing chemical data, without good evidencethat the gapscannot be breached.For the alunite supergroup,it is likely that it will be many years before miscibility gaps can be incontrovertibly demonstrated to exist. This aspect,therefore, is not taken into considerationin this review of nomenclature;nevertheless, it is evident that, over the long term, the multicompaftment systemin current use for alunite nomenclature would fare less well in terms of avoiding complexity than would a ternary system. If a ternary system of nomenclature,combined with the superstructure-symmetrynotation, were adoptedfor the alunite supergroup,the following would result: 1. Alunite group with monovalent ions in D: removal of minamiite, beaverite, and osarizawaite, and expansion of the compositional fields as shown in Figure 2 (right). 2. Ba predominant in D (Fig. 3): the compositional field available for a "weilerite"-t1pe mineral disappears. 3. Ca predominant in D (Fig. 4): huangite is woodhouseite-2cR. 4. Sr predominant in D (Fig. 5): kemmlitzite is no longer retained. 5. Bi predominant in D (Fig. 6): no change other than expansion of the compositional fields. 6. Pb predominant in D (Fig. 7): for Fe > A1, corkite and beudantite have priority, so that kintoreite and segnititeare no longer retained.Similarly, hinsdalite has priority over hidalgoite, and the latter is no longer retained. 7. REE predominant in D (Fig. 8): no change other than expansion of the compositional fields. The nomenclature changes are summarized rn Table 5.

CoNcrusroNs Independentof any new nomenclatureproposals,the following minerals and namesrequire reappraisal: (a) Beaverite: Cu:(Fe,Al) is not 1:2, and Fe is predominant.

(b) Osarizawaite:Cu:(Al,Fe) is not 1:2, and Al rs predominant. (c) Minamiite is compositionally equivalent to natroalunite. but has c ! 33 A. Mineral IMA No. 93039 (Table 4) could be given a trivial name on identical grounds. (d) Benauite: the compositional field doesnot coincide with that of the ideal formula. (e) Eylettersite requiresre-examination becausethe formula(s) exceed acceptablelimits for minerals of the alunite supergroup. (f) The possible triclinic analogue of crandallite (Cowgill et al. 1963, Blount 1974) requires re-examination both with respectto composition and symmetry; no single-crystal X-ray study has been done, and the formula deviates significantly from that of the alunite supergroup. (g) Orpheite, a Pb-Al phosphate-sulfate, is variously classified as in (Gaines et al. 1997) or out (Mandarino 1999) of the alunite supergroup.Fleischer (in Fleischer et al. 7976) concluded that orpheite is hinsdalite, but orpheite has retained speciesstatus. AcrNowrsocrvsrvrs The initial version of the manuscript was examined by, among others,J.E. Dutrizac, J.D. Grice, U. Kolitsch, J.A. Mandarino, E.H. Nickel, and A.C. Roberts. I am grateful for their comments and, in some cases, their strong encouragementto continue to pursue the nomenclature issues.I am also thankful to W.D. Birch for his appreciable input during the formative stages of this review, to P. Bayliss for incisive and useful refereecomments, and to R.F. Martin both for editorial comments and for acceptingthat this review of nomenclaturemerited airing. Acknowledgement is not intended to imply agreementwith the views that have been expressed. RrpBnsNcBs Ar-raNnn,S P.,FrrzpArRrcK, J.J.,KnonN,M.D.,Bnrrt B,P.M., HAvBA,D.O.,coss, J.H.& Bnown,Z.A. (1988):Ammonium in alunitesAm Mineral.73, I45-I52. AnKureeuro,D.K., DBvv.crrrne,E.T. & PAL'cHrK,N.A. (1985):Local symmetryof ammoniumions in the structureof aluniteandjarosite.lnCrystalChemistryandStructural Typomorphism of Minerals.Nauka,Leningrad,Russia(151-155; in Russ). Ber-leZuNr6,T., MoELo,Y , LoNean,Z. & MrcrsH-sEN,H. (1994):Dorallcharite,Tls 3K62Fe3(SOa)2(OH)6, a new memberof thejarosite-alunitefamily. Eur. J. Mineral.6, 255-263. BrncH,W.D.,pnrNc,A. & GlrEsousB,B.M. (1gg2):Segnitite, pbFe3H(AsOa)2(OH)6, a newmineralin thelusungitegroup from BrokenHill, New SouthWales. A11stralia. Am. Min_ era1.77.656-659.

13 3 8

THE CANADIAN MINERALOGIST

B L , q N c H e n o ,F . N . ( 1 9 8 9 ) : N e w X - r a y p o w d e r d a t a f o r gorceixite, BaA1:(PO+)z(OH)5.H2O, an evaluation of dspacingsand intensities, pseudosymmetryand its influence on the f,lgure of merit. P owder Diffrac tion 4, 22'7-230 Br-ouNr, A.M. (1974): The crystal structure of crandallite. Arn. Mineral. 59,4l-47. BnBrlBNsrBru, B., ScHLUTER,J. & Gssuenl, G. (1992): On beaverite:new occurrence,chemical data, and crystal structrne. N eues Jahrb. M ineral., M onatsh., 213-220. BnlrrrNcBn, D.K., KnlnclsrElN, R., BocNER, A., ScHwAB, R.G., Pnapr, T.H., Mosn, J. & Scturow, H. (1997): Vibrational spectra of synthetic minerals of the alunite and crandallite type. J. Molecular Structure 4081409,28'7-290. Bnopsv, G.P., Scorr, E.S & SNu-r-cnovr, R.A (1962): Sulfate studies. II. Solid solution between alunite and jarosite 4m. Mineral 47,112-126.

Selenateanalogues & -(1981): of jarosite-type compounds. Hy drometall 6, 327-337 & JAMBoR.J.L (1984): Formation and characterization of argentojarosite and plumbojarosite and their relevance to metallurgical processing. In Proc. Second Int. Congress on Applied Mineralogy in the Minerals Industry (W.C. Park, D M. Hausen & R.D. Hagni, eds.). AIME, New York. N.Y. (507-530). & KAIMAN, S (1976): Synthesis and properties of jarosite-typecompounds Can. Mineral. 14, 151-158. Fr-rtscHsn. M 0962): New mineral names.An. Mineral- 47 , 414-420. -(1967): 1589.

New mineral names Am. Mineral. 52,1579-

(1970): New mineral names.Am. Mineral. 55,317' 323.

& SnBnoeN, M.F. (1965): Sulfate studies. IV. The jarosite - natrojarosite - hydronium jarosite solid solution series.Am. Mineral. 50, 1595-1607

PABST,A., Menoannro, J.A., Ctuo, G.Y. & Cernr, L.J |1976i: New mineral names.Am. Mineral.6l,174-186

Cr-enr, A.M. (1993): Hey's Mineral Index. Chapman & Hall, New York. N.Y

FORrscH,E.B. (1967): "Plumbogummite" from Roughten Gill, Cumberland. Mineral. Mag 36,530- 538.

P.G & FBnn,E.E.(1986): GerNes,R.V., SrrNNsn,H.C.W., Foono, E.E, MASON,B & CoupBn, A.G.,Erraenrv, Waylandite: new data from an occurrencein Cornwall, with a note on "agnesite".Mineral. Mag.50,731-733

RosENZwEIG,A. (1997): Dana's New Mineralogy. lohn Wiley & Sons, New York, N Y.

ConrsI-pzzI. C R. (1977): Occurrence of osarizawaite in Argenlina.N eues J ahrb. M ineral., M onatsh., 39- 44.

GrusreeEtr, G. & Teorxr, C (1980): The crystal structure of osarizawaite.Neues Jahrb. Mineral, Monatsh.,401-40'7'

Cowcnr-, U.M., HurcsINsoN, G E & JoENsuu,O. (1963): An apparently triclinic dimorph of crandallite from a tropical swamp sediment in El Pat6n, Gtatemala. Am. Mineral.48, lt44-1153. CurBNNnc,Y., Rrou, A, Bor.{NIN,A. & CaILLer, P. (1980): Etudes cnstallographiques et infrarouges d'hydroxychromates de fer et d'aluminium de structure de I'alunite. Rev. Chimie Mindrale l7, 158-167. W A., BBurBs, G.J. & DE BRUryN,H., VAN DERWESTHUTzEN, Mrvnn, T.Q (1990): Corkite from Aggeneys, Bushmanland, South Africa. Mineral Mag. 54,603-608. R.A.L. & MAGAT, Ds OLrvBrRA,S.M.B., BLor, A., INTnERNIoN, P. (1996): Jarositae plumbojarosita nos gossansdo distrito mineiro de Canoas (PR). Rev Bras GeociAnc.26,3-I2 Durnrzac, J.E. (1982): The behavior of impurities during jarosite precipitalion. In Hydrometallurgical ProcessFundamentals (R.G Bautista, ed.). Plenum Press, New York, N Y. (125-169). & DrNanoo,O. (1983):The co-precipitationofcopper and zinc with lead jarosite. Hydrometall l1., 6l-78. & KArNrAr'r,S. (1980): Factors affecting lead j arosite formation. Hy dr ometall. 5, 3O5-324.

(1987): Corkite, PbFe:(SO+)(PO+) & (OH)0, its crystal structue and ordered arrangement of the tetrahedralcations.N eues Jahrb. M iner al., M omt sh., 7 | -81( 1989):Beudantite: PbFe3(So4XAso4) &(OH)6, its crystal structure, tetrahedral site disordering and scatteredPb distribttion. NeuesJahrb. Mineral, Monatsh., z t-JJ.

Har, J., JourN, 2., Kvabsr, M. & LrssscHsn, W. (1969): Kemmlitzite, a new mineral of the woodhouseite group. Neues Jahrb. M ineral., M onatsh., 2O1-212. HARTrc. C.. BneNo, P. & BoHMHAMunr, K. (1984): Fe-Alisomorphie und Strukturwasser in Kristallen vom JarositAlunite-Typ. Z. Anorg. AIlg. Chem.508, 159-164. S.B. (1937): The crystal structure of alunite and HENDRTcKS, the jarosites.Am. Mineral. 22,773- 784. IlnsroNsn, J.P , LB Toulr-Bc, C. & PBnnorel, V. (1986): About the synthetic lead-silverjarosite solid solutions Int. Symp on Experimental Mineralogy and Geochemistry (Nancy), Abstr.Vol..12-'73. IMA ( I 968): International Mineralogical Association: Commission on New Minerals and Mineral Names. Mineral' Mag.36,131-136.

NOMENCLATURE OF THE ALUNITE SUPERGROUP

Jlrvreon,J.L & Durnrzlc, J.E. (1983): Beaverite-plumbojarosite solid solutions.Can. Mineral.21, 101-113. (1985): The synthesisof beaverite & _ Can. Mineral.23,47-5L OwtNs, D.R , Gntce, J D. & FsrNcr-os, M.D. (1996): Gallobeudantite,PbGa:[(AsO+),(SO+)]z(OH)e, a new mineral species from Tsumeb, Namibia, and assocrated new gallium analoguesof the alunite-jarosite family. Can. Mineral 34, 1305-1315. JonnN, 2., JoHAN, V., Scuenu, H. & Pouea, Z. (1995): Mindralogie et gdochimie des tenes rareset du chrome dans les cherts prot6rozoiques de Kok5in, Rdpublique tchbque. C.R. Acad. Sci. Paris 321, Sdr. IIa, lI27-1138. K,rro, T (1971): The crystal structures of goyazite and woodhouseite N eues J ahrb. M ineral., M onatsh.. 241-247 (1977): Further refinement of the woodhouseite structure. Neues Jahrb. Mineral., Monatsh, 54-58 _

(1987): Further refinement of the goyazite structure. Mineral. J. 13,390-396 (1990): The crystal structure of fTorcncite.Neues Jahrb. Mine ral., M onatsh.. 227-23 L & Mniu, Y (1977): The crystal structures of jarosite and svanbergite.Mineral. J. 8,419- 430.

1339

and structure refinement by the Rietveld technique and a compilation of alunite-type compounds. Powder Dffiaction 9.265-217 LI, GEJrNc,Psacon, D.R, EssrNe,E.J., BnosNeruN, D R. & Bs,Alre, R E. (1992): Walthierite, Ba65n65A13(SOa)2 (OH)6, and huangite, CaosEosAl:(SO+)z(OH)0, two new minerals of the alunite group from the Coquimbo region, Ch1le.Am. Mineral. TT, l2'15-1284 LorecoNo, GM., KosrEcKy, G. & Wmrp, J.S., Jn. (1982): Resolution of space group ambiguities in minerals. An Mineral. 67. 846-847. Lorrnnrraospn,B.G. (1990): Rare-earthelement mineralisation within the Mt. Weld carbonatitelaterite, Westem Australia Lithos 24, 15I-167. MeNrenrNo, J.A. (1999): Fleischer's Glossary of Mineral Speclas Mineralogical Record, Tucson, Arizona. Marsuslnl, S , Kero, A., MarsuyAlral, F & Krvorn, K. (1998): Huangite ftom Okumanza, Gumma Prefecture,Japan. Mineral J. 20, 1-8 (in Japanese). MENcHErl, S. & SansLLr,C. (1976): Crystal chemistry of the alunite series: crystal structure refinement of alunite and synthetic jarosite. Neues Jahrb. Mineral , Monatsh.,4O64t7. Munrr.tn,W G. & Scorr, T R. (1966): The relationship between basic ferric sulfate and plumbojarosite. Am. Mineral. 51, 443-453.

KnenrsuN, Teyr-oR,M.R., BEvAN,D.J.M. & PnrNc,A. (1997): The crystal structureof kintoreire, PbFe3(POa)2(OH,H2O)". Mineral Mag. 61, 123-129. NICKEL,E.H. (1992): Solid solutions in mineral nomenclature Can. M ineral. 30, 231-234. Kot-rrscH, U , PrrNc, A., TeyloR, M.R. & Far-r-oN,G (1999a): Springcreekite,B a(V3*,Fe)3(pO4)2(OH,H2O)6, a new memNovAr, F. & JeNsA.,J. (1997): Minerals of the crandallite and ber of the crandallite group from the Spring Creek mine, kemmlitzite group from Upper Carboniferous sedimentsat South Australia: the first natural V3+ member of the alunite Bbl6 and Libbtdt in the Krkonobe piedmont basin, northem family and its crystal structure. Neues Jahrb Mineral., Bohemia. Vbstnik aesk6ho geol. ristava 72,361-3'77 (in Monatsh.. 529-544. Czech).

Suor, P.G, Trcrrxr, E.R.& PnrNc,A. (1999b): The structue of antimonian dussertite and the role of antimony in oxysalt minerals. Mineral. Mag 63, 17-26. TTEKTNK, E.R.T., Srnor, P.G., TAyLoR, M.R_ & PnrNc, A. (1999c): Hinsdalite and plumbogummite, their atomic arrangementsand disorderedlead sites.Ear. J. Mineral. ll,513-520. KuBISz,J. (1960): Hydroniumjarosite- (H3O)Fe3(SOa)2(OH)6. Bull. Acad. Polon. Sci.,Sir. Sci. Gdol. G6ogr. E, 95-99. (1970): Studies of synthetic alkali-hydronium jarosites. I. Synthesesof jarosite and natrojarosite.Mine ral. Polonica l,47-57. LBunoN, D.M. (1937): Woodhouseite, a new mineral of the beudantite grotp. Am Mineral 22, 939-948. LsNcausn, C.L., GrBsrrn, G & Inn.lN, E. (1994): KCr3(SOa)2 (OH)6: synthesis,characterization,powder diffraction data,

& PnacHe[. | (199$: Classification and nomenclature of alunite-iarosite and related mineral groups V?srnikteskdho geol ilsrava 69.51-57. P a u r - r S ,P . & J A N S A ,J ( 1 9 9 8 ) : C r a n d a l l i t e , gorceixite, goyazite and kemmlitzite from pyrope-gravels of the Cesk6 stiedohoff Mts. VEstnik Ceskiho seol. rtstuva

73.r01-trl. & MoRAVEc.B. (1997): Minerals of the goyazite - svanbergite series and kemmlitzite from the Vestiev pyrope deposit near Hastinn6, northern Bohemia. VEstnik Ceskdhogeol. itstava 72, 3'13-380(in Czech). Oouu, J.K., Heupp, P.L. & Fernow, R.A. (1982): A new occunence of ammoniojarosite in Buffalo, Wyoming Can. Mineral. 20,9l-95. Or.toa, K., Hrnaney,rsHr, J & Oss,q.r,q,,J (1982): Crystal structureofnatroalunite and crystal chemistry of the alunite group Neues Jahrb. Mineral., Monatsh,534-540

1340

THE CANADIAN MINERALOGIST

Soce,H., Ossere,J. & Orsur-q,N. (1987):Synthesis of minamiite type compounds, Mo sAl:(SO+)z(OH)o with M = Sr2*, Pb2* and Ba2+ Neues Jahrb. Mineral., Monatsh.,64-70. Ossere, J., HrnasavasHr, J.-I., Or.qo.A,K. & KosavesHr, R. (1982): Crystal structure of minamiite, a new mineral of the alunite group. Am. Mineral. 67, ll4-119. Orsure, N, HrnAslvesHr, J., Orala, K. & SocA, H. (1987): Synthesisof minamiite, CaosAl:(SO+)z(OH)0. Neues Jahrb Mineral., Monatsh., 49-63. P A A R , W H . , B u R G s r A L r - B n ,J . & C H S N , T . T . ( 1 9 8 0 ) : Osarizawaite-beaverite intergrowths from Sierra Gorda, Chlle. Mineral. Rec.ll. 101-104. Per-ecnn, C., BBru.raN,H. & Fnor.,nBI-,C. (1951): The System of Mineralogy (seventh ed.) 2. John Wiley & Sons, New York, N Y. PARKER,R.L (1962): Isomorphous substitution in natural and synthetic al:unrte.Am. Mineral 47,127-136. Pownns, D.A., Rossrr,raN,G.R., Scuucln, H J & Gnav, H.B. (1976): Magnetic behaviour and infrared specffaofjarosite, basic iron sulphate, and their chromate analogs. l. Solid S t a t eC h e m . 1 3 ,1 - 1 3 .

(Nd) from the uranium district in northern Bohemia, Czechoslovakia. easopis mineral. geol. 36, 103-113. Scruprzrn. K., OrrsN{eNN, J & BeNr, H (1980): Schlossmacherite, (H:O,Ca)At:[(OH)o | ((S,As)O+)2],ein neues Mineral der Alunit-Jarosit-R eihe. N eues J ahrb. M ineral., Monatsh.,2l5-222. Scrwes, R.G., Gorz, C., Honolo, H. & hNro DEOLIvsrRA,N. (1991): Compounds of the crandallite type: synthesis and properties of pure (Ca,Sr,Ba,Pb,La,Ceto Eu)-arsenocrandalIites.Neues Jahrb. Mineral , Monatsh.,97-172. & PrNro ne Orrvsrnl, N. (1990): Compounds of the crandallite type: synthesis and properties of pure goyazite, gorceixite, and plumbogummite. N eues J ahrb. M ineral., M ondtsh., 113- 126. Scorr. KM. (1987): Solid solution in, and classification of, gossan-derived members of the alunite-jarosite family, northwestQueensland,Australia.A m. Mineral. 72, 1'18-187. J.,6rrxa, J.,SnBN,V., NovolNA, M. & EDERovA,J. SEJKoRA, (1998): Minerals of the plumbogummite-philipsbomite series from Moldava deposit, Kru5nd Hory Mts., Czech Republic. Neues Jahrb. Mineral., Monatsh ,145-163. SenNa. C.J.. Pauol ConrINl, C. & Gencn Reuos, J.V.G. (1986): Infrared and Raman study of alunite-jarosite compowds. Spectrochim Acta 42I^, 129-734.

PnrNc,A., BrncH, W.D., DAwE, J., T.r.vlon, M., DuENs, M. & WALeNrn, K. (1995): Kintoreite, PbFe3(PO4)2(OH,H2O)6, SlnrH, D.K., RoBERrs,A.C., BAvrrss,P. & Lmseu, F. (1998): a new mineral of the jarosite-alunite family, and lusungite A systematic approach to general and structure-type forMag 59, 143-148. discredited. Mineral. mulas for minerals and other inorganic phases.Am. Mm' eral. 83, 126-132. S., ErsNHAnrrn, SBwnrr, D., GRAESER, A. & Cnroors, A. (1990): Baumhauerite-2a: a silver-bearF.S. & Vlrsnrs, A C. (1953):Hidalgoite, SvnH, R.L., Snr,roNS, ing mineral with a baumhauerite-like supercell from a new mineral. Am. Mineral 38, 1218-1224. Lengenbach, Switzerland Am. Mineral. 75, 915-922. Reooslovrcn, E.W (1982): Refinement of the gorceixite structure in Cm. Neues Jahrb. Mineral., Monatsh., 446464.

Sporl, C (1990): Authigenic aluminium phosphate-sulphates in sandstonesof the Mitterberg Formation, northern CalcareousAlps, Atstrra. SedimentoIogy 37, 837-845.

& Shos, P.G. (1980): Pseudo-trigonal symmetry and the structure of gorceixite. Neues Jahrb. Mineral., Monatsh.,l5'7-170.

SroprnscsN, R.E & ALPERS,C.N. (1987): Woodhouseite and svanbergitein hydrothermal ore deposits: products of apatite destruction during advanced arglllic allerutton. Can Mineral. 25,201- 211.

Rerrn-Av, K.J., Tevlon, M.R., BBvlN, D.J.M. & PRTNG,A. (1996): Compositional segregationand solid solution in the lead-dominant alunite-type minerals from Broken Hill, N.S.W. Mineral. Mag. 60, 779-785.

& Cvc.lN, G.L. (1990): An experimental study of Na-K exchangebetween alunite and aqueoussulfate solutions. Am. Mineral 75,209-220.

RocA, A., VNar-s, J., Annmz, M. & Car-Bno,J. (1999): Characterization and alkaline decompositiorVcyanidation of beudantite -jarosite materials from Rio Tinto gossanores. Can. Metall. Quart. 38, 93-103.

Szwe(rsxr, J.T. (1985): The crystal structureof plumbojarosite PblFe:(SO+)z(OH)alz.Can. Mineral. 23, 659-668. (1988): The crystal structure of beudantite, Pb(Fe,Al):[(As,S)O4]2(OH)6. Can. Mineral. 26, 923-932'

S c n a n r u , B . , S c H a n r r a o v AM , . & K U N D R A T ,M . ( 1 9 9 4 ) : Crandallite group minerals in the uranium ore district of northem Bohemia (CzechRepublicl. Vb srnik a eskdho geol. fistava69,'79-85.

TAGUCHI,Y. (1961): On osarizawaite, a new mineral of the alunite group, from the Osarizawa mine,Japan. Mineral. J. 3, 181-194

Surovsri, P. & KUuN,P. (1991)

TANANAEV,I.V., KuzNErsov, V.G & Bol'sHerove, N.K. (1967): Basic gallium salts of the alunite type. Russian J. Inorg. Chem. 12,28-30.

Philipsbornite, arsenoflorencite-(La),and arsenoflorencite-

NOMENCLATURE OF THE ALUNITE SUPERGROUP

134l

TAyLoR, M , SunH, R W. & Anr-pn, B.A. (1984): Gorceixite in topaz greisen assemblages,Silvermine area, Missourr. Am. Minerol. 69, 984-986.

& Duuu, P.J. (1984): Arsenogoyazit, ein neuesMineral der Crandallitgruppe aus dem Schwarzwald. Schweiz. Mineral. PetroBr. Mitt. 64, 7l-19

TsvnteNova, I. (1995): Corkite from Brussevtzi deposit, East Rhodope massif, Bulgaria. DokI Bulg. Akad. Nauk 48,4750.

& _(1993): Arsenogorceixit von der Grube Clara im mittleren Schwarzwald. A ufschluss 44, 250-254.

TuDo, J., LAILACE,G, Tecsez, M. & THEosaLo,F. (19731: Sur l'hydroxysulfate VOHSO4. C.R. Acad. Sci.Paris 277, Sdr. C,767-'17O. VAN TASSEL,R (1958): Jarosite, natrojarosite, beaverite, leonhardtite et hexahydrite du Congo belge. BuIl Inst. royal Sci. naturelles Belge 34, l-12. VeN Wnvstrr, L (1972): Eylettersite, un nouveau phosphate de thorium appartenant i la sdrie de la crandallite. Billl. Soc fr. Mindral. Cristallogr. 95, 98-105. (1975):La za:rite, un nouveau mindral appartenant d la sdrie de la crandallite. BulI. Soc. fr. Mindral. Cristallogr. 98, 35 1-353. Wer-BNrlr, K. (1981): Mineralien der Beudantit-Crandallitgruppe aus dem Schwarzwald: Arsenocrandallit und sulfatfreier Weilerit. Schweiz. Mineral Petrogr. Mitt. 61,

ZwTENER, M. & DuNr,r,P.J. (1982): Philipsbornit, ein neues Mineral der Crandallitreihe von Dundas auf Tasmanien.Neues Jahrb. Mineral.. Monatsh.. 1-5. WaNc, R., BRADLEv,W.F & SrrrNpnrr, H. (1965): The crystal stfucture of alunite. Acta Crystallogr. 18, 249-252 Wrsr, W.S. (1975): Solid solution between alunite, woodhouseite,and crandallite mineral series.NeaesJahrb. Mtneral., Monatsh., 540-545. Y e r H o N r o v l , L K . , D v U R E c H E N S K A y AS, . S . , S l N n o rr,rrnsxav,t,N.E., SBncBrva, N.E. & PALCrilK,N.A. (1988): Sulfates from the cryogenic hypergenesiszone. New findings Nomenclature problems. Mineral. Zh 10,3-15 (rn Russ.).

23-35.

ZneNc Ruso, Y,qNc Nr,lNznpN, YI SHUANGTTNc & Du CnoNcI-IlNc (1987): Arsenogoyazite discovered in Xinjiang, China. Acta Mineral Sinica 7, 313-316 (in Chrnese).

BIRCH,W.D. & Dt,.l+N,P.J (1996): Benauite, a new mineral of the crandallite group from the Clara mine in the central Black Forest, Germany. Chem. Erde 56, l7I-1i6.

ReceivedJune I 1, 1999, reyised mnnuscript acceptedNovember 12,1999.