Catudfun Mineralogist Vol. 14, W. 322-333 (1976)

PEKOITE,GuPbBi,,S.,, A NEW MEMBEROF THE BISMUTHINITEAIKINITE MINET|ALSERIES:. ITS CRYSTALSTRUCTURE AND RELATIONSHIP WITH NATUFALLY.AND SYNTHETICALLY.FORMEDMEMBERS

W. c. MUMME exp J. A. WATTS CSIRO Diviston o! Mineral Chemistry, P.O, Box 124, Port Melbourne, Victoria 3207, Auslralia

ABSTRACT Ixtnonucrrox

The crystal structure of pekoite, ideally CuPbBirr Recent developments in the study of the (S,S€)tr, a trew min€ral from the luno mine at minerals in tbe system bismuthinite-aikinite are Tennant Creek, N.T., Australia, has been determined found in Mumme et al. (1976) and Harris & from a three-ilimensional X.ray analysis and re- Chen (1976). The history leading up to these fined to glve an R-valuo of 0.12. Pekoito has the studies is that Moore's (1967) classification space.group P2gm, wlih a 11.472, b 3X1L248, c based on Welin's (1966) studies was questioned 4.0164 for the composition Cu.ePb.raBirr.o$r.oSe..or. by Zak et al. (1975) and Synecek & Hybler The structure is a tlree-fold superstructure bared (1975) after these workers discovered a ne\tr of two slabs of bis' on bismuthinite and is made up intermediate member krupkaite, CuPbBigSi', and (BieSJ and one slab of krupkaite muthinite ribbons to con- (CuPbBisSo) ribbons alternating along the D-axis. found that its crystal structure failed The classification of the ordered mineral phases form with Moore's Z" classification. Synecek which form along tle BLSr-CbPbBiSs composition & Hybler '(L975) proposed a new nomenclature line is discussed, together with a study of the crystal which, however, had the deficiency that their structure of members of the series synthetically pre- classification symbol, Zn,.cotild be sommon to pared at 450-500'C. The results of single-crystal several intermediate members. Also, their classi- analysis of tbe synthetic compounds confirm that, fication incorrectly predicted the space group rn contrast to the ordered natural members, com- of Welin's o'new" member CusPbsBizSuto be plete sotd solution does occur between Bis\ and Pbnm, not Pbztm as was already known. CuPbBi$, with copper and lead atoms progressively filling the sites which they fully occupy in the end- Independently of. Zak et al,, the failure of member aikinite. Moore's classification was also recognized by Mumme et al. (1976) through the results of concurrent studies of the structures Sotvrvrerns separate of krupkaite (Mumme 1975), and another pre- La structure cristalline du p6koite, un nouveau viously unrecorded member of the series, min6ral CuPbBill(S,Se)16, ptovetrorlt de la mine pekoite, now presented in this paper. Both of Juno de Tennant Creek, N.T., Australie, a 6t6 d6' these minerals were discoveredin the Juno mine terminde par une analyse tridimensionnelle par dif- at Tennant Creek, Australia (Large & Mumme fraction des rayons-X et affin6e afin d'obtenir un 1975). indice R de 0.12. Sa maille 6l6mentaire est PZpm, Mumme et al. proposed that the classifica- dont a t1.472, b 3XLI.248, c 4.0t64' et de com- tion of the minerals should rest on the establish- position Cu.6sPb.?8Bi11.sSra.$See.oz.Sa structure est en fait une superstructure de l'ordre de 3, compos6e ment of distinct names for the ordered mineral princtnlement de bismuthinite: deux plaques de species.As part of this proposed nomenclature, bismuthinite (BraSs) et une plaque de krupkaite which has since been approved by the IMA, (CuPbStSJ alternent le long de I'axe-D. On drscute Welin's C\rsPbsBizSrsphase came to be called de la classification des phases min6rales ordonn6es lindstromite,* and the individual members qui so trouvent dans le systBme Bi?r$-CbPbBi$, et known as bismuthinite derivatives. d'une 6tude des membres de la structure cristalline Harris & Chen (1976) inde- de la s6rie pr6par6s par syntllse i 450-500"C. Irs Most recently, phases by r6sultats de fanalyse de cristaux uniques des com- pendently studied the six classified pos6s synthdtiques confirment qu'il y a effective- Mumme et aI. ar'd showed that all except lind- ment un€ solution solide compldte entre BigSg et stromite possess considerable solid-solution CuPbBiSs, contrairement i ce que I'on trouve pour mnges. This type of behavior for pekoite is also les exemples naturels ordonn6s. Les atomes de cxemplified in the results described in this cuiwe et de plomb remplissent progressivement les present paper. sites €t les occupent comllbtement dans le p6le extr6me, I'aikioite. (fraduit par le journal) *Proposed by Professor B. J. Wuensch. 322 PBKOITB 323

The crystal structure of pekoite wall recog- nized to be extremely important to an under- standing of the relationshipsamong the members of this mineral series, and is reported here in detail. The present investigation has also been extended to a study of possible superstructure formation in synthetic members of the series prepared by solid-state reactions.

Tns CRysrar- SrnucrunE oF PEKorrE Experimental Members of the bismuthinite-aikinils ..ti.t are common in the No. 1 and No. 2 orebodies at the Juno mine, Tennant Creek (Large 1974). Most of the members studied contain oriented blebs which resemble exsolution textures (Fig. 1), and may be straight or cymoidal. Micro- probe analyses (Large & Mumme 1975) which suggested that the blebs are members of the bismuthinite-aikinite series with n higher lead and copper content than the host grains are plotted in Figure 2. T\e Pb content of the blebs was about l4Vo, which correspondedwell with tle mineral gladite, CuPbBi'Ss. The an- ftc. 1. Exsolution textures developed in members alyses for the host material (Table 2) although of the bismutbinite-aikinite series from ttre Juno mine. Host:pekoite; intergrown phae: rather more scattered, corresponded to the com- eladite. Etched with HBr for 60 sec. The field is position CuPbBilSu with a lead content of 50 pn wide. aboat 6/o. X-ray Weissenberg photo$aphs of small aligned fragments of material showing the ex- ing values of ft only. Trvvolattices aligned in a solution textures gave results similar to those parallel fashion are present, and the fragnents shown in Figure 3. Reflections on the Weissen- investigated thus consist of two submicro- berg films gradually split into two with increas- scopically intergrown minerals as already sug-

\ .\0' "f ,"$ 02&

6r- - Bir(S,Sel -Pb(S,Se) s lvt

Frc, 2. Microprobe analysesof tle intergrofih phases pekoite and gladite (foom Iarge & Mumme 1975) together with that of the pekoite specimen (R27788) studied in this paper. Compositions are plotted in the Big(S,Se)3,Pb(S,Se), Cur(S,Se) s],rtem. Exsolution pairs are joined by lines. 324 T1IE CANADIAN MINI.RALOGIST

TABLEl. CRYSIALLoGMPHICDATA FoR PEKoITE, CuPbBiil(S,Se)18 ili neral Composition Lattlce constants (i) Reference

Pekoite CuPbBil b"3x11.248(2)a"11.472(2) a.4.015(l) PZle Thls study tSl4Se4 Pekoit€ slg b.3xll.l68(2) a.11,322(2) o.3.987(l) *'r^ This study CuPbBlI I

J' ,.3xt1.182(2) a"ll,498(2) o.4.001 Pnna Thls study gested by the microprobe results. Furthetmoreo the host mineral has subcell dimensions sub- the Weissenbergresults show that the blebs and stantially larger than that of bismuthinite and a host may have parallel orientation. The two composition close to CuPbBiuSre, this mineral minerals are otrhorhombic, with identical values now being known as pekoite. This name, for of D and c, but differ slightly in a (setting which IMA approval was granted in July' 1975' bSa2c). The refined unit-cell parameters is after the Peko mine, one of the first in the derived from the Guinier powder data (cali- Tennant Creek gold field. -- brated with KCl, a. 6.29294) are given in The crystal structure of gladite had been re- present on the Table 1. Superlattice reflections ported by Kohatsu & Wuensch (1973), but the Weissenberg films show that both of the min- pekoite intimately intergrown with it repre- erals are superstructures based upon bismu- sented a previously unrecognized new member thinite, with one lattice constant equal to three of the Bi,SsCuPbBiSx series. It was initially in times that of the corresponding translation thought that the study of its crystal structure parameters bismuthinite. The blebs have lattice would have to be attempted using X-ray data gladite, nearly identical with those of whereas collected from a fragment of this material which exhibited exsolution lamellae, in a somewhat TABLE2. tllCRoPRoEEAMLYsls 0F PEKoITEIN SPECII,IENM7788 similar fashion to that used in Synecek & gladite. wt. g ilolecular ProFrtlons Hybler's (1974) studies of krupkaite and Basedon l8(srse) Ilowever, other specimensfrom Tennant Creek BI 73,6 Bt 1l,50 (Fi27798 to contain Pb (l Pb 0.78 and R27788)* were found Cu 1.3 Cu 0.65 this new member with selenium contents up to 14.3 5 I to Se S€l 7Vo (lable 2). Crystals of pekoite from R27788 TotaI lol .? ,Jol JXA-S0Ahlcroprcbe' 25Kf. Strndar{s used: 9a-lena'-crausrna- llte. Bl and Cu mials. The coretlon prcgrm' I'IAGICIV' ms de- 'rNumbering refers to specimens presently held at veloiea uy ,J.tl. colby of Bell TelephoneInc. ' A'llantm' Pennsyl- vanla. Mineral Chemistry, CSIRO. b cl

t- t- t-

I

I

=l- a - t-l - - -

Frc. 3. Tracing of Weissenbergdiagram flections, lofer fines are gladile arieflections. The reflections separatewith incroasing value of ft. No a2-reflectionsare plotted. PEKOITE 325 gave X-ray data which showed them to be sin- The final R-value for all subcell and supercell gle-phase, and confirmed the tlree-fold cell reflections was L2.LVo, after absorption correc- relationship with respect to bismuthinite. The tions were made for a cylindrical specimen of space group of pekoite was not Pnma (that of pR=1320, The weighting schemeused was that gladite) but P2nm, that is, the same spacegroup of Cruickshank et al. (1961). Final atomic para- as krupkaite and lindstromite. The other space- meters and bond lengths are given in Tables 3 group alternatives, PmaZ and Pmam, may be and 4. The structure factor table has been discounted on the basis of the substructure- deposited with the National Science Library, superstructurerelationships f61 similal reasonsas Ottawa. those already described in the accounts of the crystal structure of krupkaite (Zak et al. 1975; Description and discussion of the structure ol Mumme 1975). pekoite X-ray data for the levels I : 0, 1, 2 were The proposed ideal structure of pekoite, collected by the multiple-film technique from a CuPbBLr(S,Se)reis shown in Figure 4, As crystal of pekoite measuring.06X.06X.25 mm. distinct from krupkaite, which is composed en- The integrated equi-inclination method was tirely of slabs of krupkaite-like ribbons, and used with CuKo radiation. Intensities were gladite, which contains two krupkaite slabs and measured visually against a standard scale. Of one bismuthinite slab alternating, the structure tbe 425 reflections recorded, 51 were super- laffice reflections.

Solution and relinement of the structure As the X-ray data showed a strong subcell relationship with bismuthinite, the original atomic positions used for pekoite (in each of the three subcells) were taken as those of bis- muthinite, transposed by V+b, as required for the space group P?,tam. The three-dimensional data were then refined using a local venion of ORFLS (Busing et al. 1962).Previous experience in refinement of krupkaite (Mumme 1975), aikinite, and gladite (Kohatsu & Wuensch 1971, 1973) indicated that the Pb site in the structure would be characterizedby a significant displace- ment of the location of the Pb atom relative to that of the corresponding Bi atom in the bis- muthinite structure. However, the superstructure intensities were very weak, and the attempted refinement with the substructure reflections in- cluded in the data set resulted in these intense reflections dominating the parameter shifts and the atoms refined to the bismuthinite positions. At this stage, therefore, calculations were com- menced to determine which of the ordered models for pekoite permitted by the subcell- superstructure relationship gave the best agree- ment for the superstructure reflections. Using the 51 superstructure reflections it was demon- rtrated that for these models the best structure factor agreement (R-0.25) occurred when cop- per occupied the site in the structure shown in Figure ^4, with the adjacent lead site displaced bv 0.5A from its equivalent bismuthinite posi- tion. This, basically, is the method of key shifts which has been described by Ito (1973). The R-values obtained from similar calculations for Frc. 4. Crystal structur€ of ideal pekoite, all the other possible models varied from .36 C\rPbBirSr proiected onto (001). The shadedpor- to .52. tions correspondto the strbnite quadruple chain. 326 TTiE CANADIAN MINERALOGIST of pekoite has two bismuthinite slabs and one are composed of various combinations of bis' krupkaite slab alternating. muthinite, krupkaite and aikinite-like ribbons. The strustural relationships, and crystal Those members of the seriesbetween bismuthin- chemistry of the intermediate bismuthinite deri- ite and krupkaite are composed entirely of bis- vatives, have now been discussed by Mumme muthinite and krupkaite-like ribbons. Ham- et ql. (1976) and Harris & Chen (1976), and marite, Cu:PbzBiaSo, and lindstromite, CuaPb. there h no new information known to the BizSrs, the two other known members of the present authors which can be added here. How- series which occur between krupkaite and aiki- ever, for the sake of completeness,Figure 5 nite, are yet to be studied, but it may be pre- summarizes, in the notation of Omassa & dicted that they will incorporate both the Nowacki (1970), the structures of the four bis- aikinite-ribbon and the krupkaite-ribbon into muthinite derivatives studied to the present. All their strustures. As with gladite (Kohatsu & Wuensch 1973) (Synecek& Hybler 1975; Mumme TABLE3. ATO}IICCOORDINATES IlI PErcITE and krupkaite 1975), pekoite demonstratesthe tendency of Pb Slte occupancy r n al\)z (and Cu) to order among a$ many distinst bis- M1 Bi ,4685(4 ,9752(6) 1t2 .81(5) muthinite chains as possible. t'{2 BI .3368(7 .0710(3) 0 1.92(6) t{3 BI 0t91(5 .1408 (8) 0 1.56(7) The probe analysesgiven in Figure 1 (from a4 BI .4891(5 .3077(2) 1t2 4.00(e) LS BI .3430(5 .4060 o 5.90(9) Large & Mumme 1975, Table 8, Nos. 1,3,5,7 I't6 st . r 582(4 .2*9 t;l 1t2 1.50(7) and 8) and the composition determined here for ,d] BI .ls35(4 .9049(6) 1tz .24(4) ua Bi 0255(5 .8082(2) o 2.80(e) R27788 indicate that pekoite has a range of M9 BI .4868(5 .6411 (8) 1/2 1.74(7) Mto BI .1718(5 .5717(e) 1/2 1.e3(7) composition, and that the copper site in the lll Bi .0098(5 .474a(8) o l.20(6) strucfure may only be partially occupied with Ml2 .7Pb+.38.| .3326(5 ,7521(2) 0 3.00(9) disordering of Pb and Bi Cu .7Cu .2455(20) .6525 ) 0 5.oo(85) a necessary limited 5t . I 75e+.83S .2754(14't .0179 20\ vz I .05(61) atoms. On the other hand, analysis 6 (fable 8' s2 .lr50(13) .101821) 1/z 2.ts(lol) 53 .0500(9) .2958 l9) 0 l.65000) Large & Mumme 1975) suggestsa limited oc- s4 .0454(14) .960017) 0 2.00(ll5) the ideal (6 .4546(14) .129Izo| 1tz 1.35(el) cupancy of copper atoms in excessof 56 .3700(r6) ,937021) o 1.80(llo) formula into other sites which are available to )/ .2854(r4) .Jttt 25I ttz 2.15(e5) 58 .3705(r2) ,2@ ?'l) 0 .e5(e2) it in the structure - possibly those which are s9 .2200(lt) .1849l9) o l.06(85) gladite (Kohatsu 510 .2056(14) .5179 ) 0 1.85(lol) found to be fully occupied in sll .3795(l5) .601820) o 2.14(lo2) & Wuensch 1973). 5I2 .4549(9) .79582l| 1t2 r.44(e5) 513 .4546(r4) .4624 l9j ttz 1.85(95) The disordering of the copper in the krup- sl4 .0550(r3) .629'l 22) o r.65(87) (Mumme 515 . r 205(10) .435123) vz .85(50) kaite ribbon has beeu discussed 1975) sl6 .2146(r4) .852023) 'rtz0 l.os(eo) and may also be a feature of all those derivative Jlt .r205(r2) ,7684 24\ l.Is(eo) sI8 . r 254(9) .6849 27| 1tz 2.56(102) structures s66aining that ribbon - possibly it may even be direct evidence of transformation meihanisms which occur between members. TABLE4. BO!{DLENGru IlI PEKOIN Similar solid-solution ranges to those of Cu - (tr 2.30(r0) Mt Bt? - 55 2.55(3) pekoite have now been reported in considerable - 518 2.33 (6) x2 - 5lO 2.78 (5) ta detail by Harris & Chen (1970 for all members - 514 2.32(5) -s4 3.00 (4) x2 nl Bil -s6 2.64 (3) X2 -50 3.37 (3) X2 of the series except lindstromite (C&Pb;BirS't. -sI (41 2.64 ?r8 Bi8 - 516 ?.63(5) -s4 3.10(4) - (4) -s2 (6) )t/ 2.65 x2 3.09 -cE 3.03 (5) Cnvsrer, Srnucrunrs or SYNnrBrrc - )l 3.53(2) -s8 3.14(6) ,a 812 - 54 2.63(3) -59 3.51(?) Brsuurnnvrrn-ArrwrrB MEMBERS -sI 2.78(51 X2 's5 rE Big - Sll 2,70 (4) x2 3.r2 (5) X2 - (4) Two previous studies of the members of the -s2 (3) 516 2.74 3.40 12 - sl5 2.99 (7) bismuthinite-aikinite series (Padera 1956; rr3 Bl3 - 52 2.6414) \Z -s3 3.02 (5) X2 -s9 z.r4 (41 -s7 3.M l2) Springer 1971) were made using X-ray powder - stz 3.03(5) X2 -s6 rno 8il0 - slo 2.74(s' x2 methods. Padera studied natural minerals where- 3.13(6) 2.74(3) - sl6 3.50(2) - (5) Both 'ta sl4 3.I0 X2 as Springer studied synthetic members. ri4 Bi4 - s8 2.76 (31 _ sl.l 3.28 (3) solid solution, an ob- -s7 2.77(51 - described the system as a - r/ll Bfil slo ?.68 (1) sl7 ?.98(7) - s15 2.73 (41 X2 servation which in the case of Paderds \rork has - sl4 3.03(5) _ - )lJ 2.99 (5) x2 Al- sl8 3.41(2) - sll 2.ee(6) not been supported by more recent studies. t6 Bt5 - Sl4 2.71(41 - slo 3.50(2) though Springer's study was specifically directed - )/ 2.80 (6) 12 r1t2Pbl - slz 2.86 (4) x2 - sl3 3.05 (4) 12 -s3 towards the detection of superstructure forma- - (2) 2.e7(4) sl5 3.39 - 518 3.08(5) x2 tion in synthetic members, it was also carried !6 8i6 - Sl2 2.6r (4) - sI7 3.20 (2) X2 -s9 2.80 (5) x2 out using powder X-ray methods. To obtain -JJ 3.04 (4) X2 direct information we attempted to pre- -58 3.3t (3) 12 more pare single crystals of various intermediate mem- PEKOITE 327 r'-l l-b-l @@ Lb-8"' [-r-J (a) I'l r'---] @ Bhso o b CuPbBirs6 cureurelsu @ \

@ @I

(d) L:j(e) Fta. 5. Ribbon models for (a) bismuthinite, Bir$; (b) krupkaite, CuPbBigSo; (c) aikinite, c\rPbBiss; (d) pekoite, cupbBinsre; (e) eladite, cbPbBissr, io terms of the bismuthinite (BroSu), krupkaite (C\rPbBisS6) and aikinite (CurPbrBirs6) ribbons. bers of tle series,spd slamined these by single- were left for periods of up to a week at tem' crystal methods. petature, after which tley were rapidly quenched Unfortunately, single crystals of a size neces- to the temperature of ice water. The products sary for manipulation were not obtained by our obtained from, most of these teactions were methods of preparation below 450oC, higher polycrystalline, although in the slmthesis of than Springer's lower limit of 300"C for com- aikinils a macrocrystalline product was obtained pound formation. Nevertheless, the structures from which single crystals were chosen for of the crystals prepared were studied, because X-ray investigation. The preparation of the re- it was considered possible that the timitations mnining phases were carried out by iodine- of the powder method to detect weak super- transport methods in sealed silica tubes con- lattice reflections may have influenced Spring- taining 2-3 mg of Iz per ml of tube volume, with er's conclusions. temperature gradients of 500o470oC. The bulk materials used for these experiments were Pre- (l) Experimental pared as outlined above, and after grinding were sealed under vacuum (< 10' atm), with Synthetic members of the solid-solution series the carefully weighed Ig, into the transport tubes. CuPbBiSs-BirSa rilero prepared using evacuated Under the conditions of temperature gradient sealed silica tubes which were charged with the used for these experiments very fittle transport appropriate proportions of Bir$, CuaS and PbS of the material took place, with the iodine act- previously prepared from pure elements. High- ing as a vapor-phase medium for crystal growth purity Koch-Light sultur (99.9999Vo)was used of the bulk starting material which was set at together reith A.R. lead and copper foils (metal- 500oC and spread ovet l-l/z cm from the end lic impurities (30 ppm) and high-purity Mersk of the 12 cm-long tube. The reactions produced bismuth (.450 ppm total metailics). The tubes thin (<.02 mm) needle-like crystals approach- were placed into horizontal tube furnaces or ing 0.3 mm in length and very suitable for muffles controlled within t2oC of the set tem- single crystal X-ray diffraction analysis. perature with a proportional controller. They Quantitative chemical analyses of these vapor- 328 TTIE CANADIAN MINERALOGIST grown cry$tals showed that they contained less corrected for Lorentz polarization and absorp- than 5 ppm iodiue. tion using a linear absorption coeffisient of The products from the quenched reactions 1380 cm-r for C\lKc radiation. All calculations were examined by X-ray powder analysis using were made using an Elliott 803 computer with both a Philips diffractometer at a scanning rate the aid of the programs of Daly et al. (1963). o/min of I , with Si powder as an internal stand- Scattering curves for neutral Pb, Bi, Cu and S ard, and a Guinier focusing camera with KCI were taken from the International Tables of as a standard. The unit cell at each composition Crystallography, Vol. III, with appropriate ad- was determined by a least-squaresprocedure. justment to the Cu scattering curve made at The lattice paramerers of the single-crystal each composition. The initial atomic parameters phaseswere determined from Guinier data using for all atoms were taken from Kohatsu & a small number of crushed crystals.These values Wuensch's (1971) data and the least squares were found to be in good agreement \ilith para- refinement proceeded for all phases to give R meters determined with the aid of a travelling factors of. (a) LO.5OVo(b) 1,l.Zl%o,(c) 9.57Vo, microscope from unintegrated Weissenberg (d) 9.8Vo, and (e) LL.SOVofor aikinite. Isotropic films for a anld c and from rotation photo- temperature factors were used throughout the graphs for b. refinement. Single-crystal X-ray analysis was undertaken (2) at the following compositions(a) Cuo.zPbo.zBil.ss[, Discussion of Results (b) Cuo.rsPbo.r"Bir.e'Ss Gladite), (c) Cuo.sPbo.sBir.sS"Powder data. As pointed out by Springer (1971) (knrpkaite), (d) Cuo.r"Pbo.rrBit.rgs"and (e) in his analysisof synthetic BizSrCuPbBiSgmem- CuPbBiS" (aikinite). Small needle-like crystals bers and also by Welin (1966) from his study of the following dimensions were chosen for of natural minerals which fell along this com- data collections:(a) 0.14X0.03X0.02 mm; O) position line, powder patterns taken across the 0.10x0.05x0.02 mm; (c)0.12x0.04X0.03 mm; series reveal a gradual but distinct change in (d) 0.12x0.o4X0.04 mm and (e) 0.12x0.03x line positions and intensity as the Pb content 0.02 mm for aikinite. Intensity data for the levels increases. h0l-hLl were recorded on multiple films with This is clearly demonstrated in a plot of Pb 't CuKo radiation using an integrating Weissen- content versus the value or sinid value of the berg goniometer. The integrated intensities were 22O lne for bulk preparations, as shown in

16 sirEle crlrstol . powdcr x

x

12 (b)

fro

so E8 q)

x xa

..m4 .qr6s .@72 .0?76\ si#e o(A)rr.s tiq ttt rr€ 1i.7 Ftc. 6. (a) Sin2 d vallues vs. Pb content for tle 220 l:rte of powder and groun

Figure 6(a). {,!so included on this graph are the TASLE5. LATIICE PANAMEIEISOF III1ER€DIAIE IE$ERS TRW POI{DER observed values for the 22O line obtained from DAIA} Guinier data for powders ground from crystals Co[Fositlon a([) r([) o(I) . grown by iodine transport. It clearly shows that Bt2s3 il.294(?) 3.e7e(6) t1.143(7) the compositions of the crystals formed were crO.tPbO.tBil.gS3 n.309(8) 3.9S7(7) ll.l4e(8) consistent with those of the bulk starting ma- c'o.tqPbo.tlBll.065g 1r.341(10) 3.988(8) n.164(e) terials, and this study of the powder data was cuo.2Pbo,2Bll.g53 11.385(8) 4.018(8) ll.le5(7) primarily sarried out to obtain this confirmation. cuo.zsPbo.25Blr.t5s3 il.401(7) 4.021(6) 11.217(0) r1.412(8) 4.0r9(5) 1't.?,.ol'|) The least-squares analysis of the powder data c'0.:oPbo.:oBit.7osg cuo. 1r.425(e) 4.02r(7) n.2$(8) (Table 5) also confirms a gradual increase of all ggPbo.g:Bi l.67sg cuo.4Pbo.4Bl 1r.ru5(8) 4.022(6) 11.258(10) r . esl 'r'r.482(8) the lattice parameters with the replacement of c'o.sPbo.sBl 4.0a(5) L258(8) r.s53 'il.505(10) Pb for Bi in Bi'S' and the addition of Cul+ to c'o.6Pbo.68il.4s3 4.026(7\ 11.275(7) 'il.53s(ro) 4.026(7) 1r.276(4) maintain the charge balance. The single-crystal cuo.75Pbo.7sBll . zssa data (fable O also show this trend which gives c'o.8Pbo.gBll.zs3 il.537(lr) 4.027(8) ir.287(6) 1',r.565(ro) 4.02e(6) n.293(9) an overall increase in cell volume of 30.1A' cuo.g6Pbo.8oBlI.t45a from Bi,Sg to CuPbBiS,. Figure 6(b) indicates cul.oPbr.oBlI.osl 1t.602(e) 4.034(7) il.305(5) the linear relationship betweenthe a lattice pa.ra- resd in parentheses meter and the lead content. Careful examination of the Guinier data for all compositions showed TABLE6. LATTICEPAMI.ISIERS OF SIIIGLECRYSTALS OF ITTER}IEDIATE no indication of superlatticereflections, in agree- I.IBEERS' ment with Springer's (1971) observations. Co@osltlon a(l) D(t) o(l) Yol. ([)3

Single crystal analysis.None of the Weissenberg 't1.282 3.971 ll. n2 491.&6 cuo.oPb6.pBl2sJ films revealed superlattice formation at any of 4.014 1r,2A(5) 513.17. CuO.aPbo. ZBl.t .gs: r.392(6) the intermediate compositions examined. All of 1r.ir22(3) 4.018(5) 11.224(6) 515.r0 cuo.rsPbo. srBlr .ofr the crystals had lattice parameters intermediate cuo.sPbo.sBlt.isg il.4e5(6) 4.M2(41 il.264(5) 520.81 between bismuthinite and aikinite, and all had Guo.7sho.75Bll.z5sln.535(lo) 4.(n6(7) ll.2t6(4) 523.65 4.03r(6) lr.3ot(5) 5a,52 space group Pnma. Crl.OPbl.oBlt .dt 1t.603(7) A complete description and comparison of the clPbBlsSr lr.6$3(10) 4.0279(3)tr,z7il(17) 527.20 'esd structures of aikinite and bismuthinite has been ln parenthgsas. given by Kohatsu & Wuensch (1971) and a com- +Kumlk i vosola-l{ovakova(1970). ttt{aiunlly occurrlng alklnlte (Kohltsu I luansh l97l). parison of their lattice parameters,bond lengths, and atomic coordinates with our svnthetic aikinite is given in Tables 6-8. These clearly TABLE7. INIEMIO}IICDISTNCES indicate that aikinite prepared from pure Cu,Bi, (a) cu and Pb sulfides at 500"C is in all respects Composltjon sl s3lr2 52 identical with the natural mineral. Aikinite is Cuo.2Pbo.zBll.8S3 2.22(8) 2.38(4) 2.29(21 an expanded version of bismuthinite; lead sub- cuo.lgPbo.::Blr.ef3 2.23(6) 2.38(3) 2.42(1, stitutes in an ordered manner for bismuth in Cuo.sPbo.5Blt.5Sg 2.25(6, 2.36(3) 2.38(2') cu'.usPbo.zsllr.zssg 2.31(3) 2.37(1't 2.43(l) the MQ) site of bismuthinite and copper atoms Crl.oPbl.oBit.oS3 2.31(3) 2.38(l) 2.43(2', occupy the tetrahedral sites. Cu.r.oPb.l.oBit.oS:fi 2.31(l) 2.37(l) 2.43(l) A comparison of bond distancesfor the syn- (b) thetic intermediates is shown in Table 7, which l'(2)"lPb,Bll Conpositlon Sl s:lxz 5llx2 szlxz refers to the drawing of the unit cell in Figure Bi2s3r 2.54 2.74 2.97 3.33 7. There is a small change in bond lengths for Cuo.2Pbo.zBll.805g 2.64(3) 2.82(2) 3.0r(r) 3.34(2) Cu-S(l), even when the large e.s.d.'s of the cuO.::Pbo.llPbt.Of3 2.69(3) z.6flf/ ..t.ltl J.Jf(Z, Cuo.zphase are considered.The variation in the CuO.5Pbo.5Bll.sS3 2.67(2) 2.85(r) 3.00(r) 3.35(t) Cu-S(3) and Cu-S(2) distance is very small, but CrO.7sPb0.75Bll.255s2.83(l) 2.8e(r) 2.e8(t) 3.28(t) for Cu-S(2) there is an apparent sharp increase Crt.OPb.l.OBlt.OS: 2.85(l) 2,e2(1) 2.ee(r) 3.27(l) Crt.OPb.t.OBl.l.0Sg 2.84(l) 2.e5(r) 2.e8(1) 3.28(l) of 0.13A from the Cuo.z composition to the (c) u(1)"[Bi] Cur.rs Phase. 'IUIQ) ConposIt i on 53 SZlx2 511llx2 52ll S3Il The site has seven nearest neighbor Bi253t 2.63 2.62 3.05 3.05 3.47 sulfur atoms and their bond distancesare shown cuo.zPbo.zBlr.a53 2,69(3) 2.70(2)3.00(2) 3.0e(t) 3.41(2) cuo, in Table 7O). The change in lengtls f.or M(2)- slPbo.:lBlt.of: 2,5s(2) 2.70(2'13.06(1) 3.02(l) 3.47(2) cuo.5Pbo.sBlr S(1) bonds of 2.58A to 2.85A from BisS, to .553 2.72(11 2.70(2\3.05(1) 3.10(l) 3.44(2) cuo.75Pbo.75Bli CuPbBiSs confirms that lead progressively re- .25s: 2.67(rt 2.74(1)2.ee(l) 3.0s(l) 3.52(l) Cut places bismuth in the M(2) sites, not in the . OPb.l. oBl.l . OS3 2.68(l) 2.76(r)2.e7(l) 3.07(2) 3.53(l) Cur-oPbr 2.66(1') 2.73(t')2.e7(l 3.12(l 3.53(t M(1) sites. The M(2)-"S(3)r distances show a nBlt-nsrft ) ) ) correspondingincrease as lead replacesbismuth. tKupcik & vesela-Novakova(1970) tiKohatsu& $uensch(1973) 330 TIIE CANADIAN MINERALOGIST

7, Fig. 7) forms a severely distorted octahedron l with a seventh sulfur 0.5A distant from tb c next-nearest neighbor. These bond distances show relatively smnll changes from Bir$ through to CuFbBiSo, the only sipificant change being for the M(1)-S(2)1 distance which increases from 2.62A in Bi,$ to 2.764 in aikinits, $s1 more than half this increase takes place between BizSs and Cuo*Pbo.rBit..oSa. There is a small non-linear increase at Cuo.o for the M(l)-Sg distance. The increase ftom 2.684 at the Cuo.escomposition to a maximum ot 2.724 at the Cuo.sphase and back to 2.66A at pure aikinite may be related to the desrea$ of 0.02A in the M(2)-S(1) distance at the same composition, suggesting some stght disorder of the Pb and Bi n the M(I) sites.As no evidence of superstructure formation was observed, tle results of this investigation indicate tlat, under the conditions that the phases were formed, the Frc. 7. The crystal structure proiected of aikinite copper to onto (010). Labelling of atoms is that referred atoms tend fill the tetrahedral holes to in the text. progressively, and at the same time there is an ordered replacement of bismuth by lead in the M(2) sites of the structure. Ftgure 8 shows a plot of these two distances vefirus composition. There appearsto be a linear Fontvre'uoN or BtsuurutNrrs DnRwetrvEs relationship for tle M(2)-S(3) distance, with a maximum value of 2.974 at the limiting com- This study of the synthetic bismuthinite- position. The M(2)-S(L) distances exhibit a aikinite system has confirmed Springer's (197I, similar relationship. The M(2)-S(l)t alrd MQ)- claim that, at least at 450-500'C, complete S(2)1 distances vary only slightly, by 0.07A, solid solution exists between the end members. over the whole composition range. It should be noted that Springer's results could The M(IFS coordination polyhedron (Iable be interpreted to indicate tle formation eitler

o A 3'0

2.9 (Bi,Pb)'s3 r

2'S I (BiEg I I l---- 2.6

v1 0.1 cf203cE-0506W00 0€ 1.0 Bizss PbCuBiS,

Frc. 8. Graphs showing tle variation of M(2)-fi3)l and Jtt(2)-S(l) bond distancesin composition BisSs-ClPbBiSs.Estimated staudard deviadons are indicated by line bars. PEKorrB 331 of a complete solid-solution series, or of a series greater than 450oC are currently being under- of compositions based oa.a disordered stracture taken. type in which lead and bismuth atoms statis- Unless slaminltisn sf singls crystals pre- tically occupy the atomic positions available to pared at various temperatures below 450" them in the bismuthinite structure. As less dis- shows otherwise, Springer's limit of 300oC must ordering would be expected at lower tempera- be accepted as approximately representing the tures, the partial ordering behavior observed in maximum temperature at which natural pekoite this study must be considered to continue to etc, have been formed. Either there has oc- lower temperatures where it is known that the curred a slow gssling process with unmixing fully ordered phases are stable. and segregation from an original solid solution been direct Exsolution textures have been reported in the or 309-'c, or there has 1t -p1loy process' minerals of this series by Ramdohr (1969), b;; formation by some other ore-forming evi- Welin (1966) was the ftst to roppott his ob- Supported by _the now well-established servations of these phenomenawidiquantitative dence of the pekoite/gtuditq exsolutions, one results. In RM24099 he reported a mimber with might expect that all individual solvi in the SVo Ca,2TVo pb,56/o Bi'andLgVo S (probably system..would.be as is represented diagram- krupkaite) containing exsolved bismutninite. tn matically in Figure 9, which is drawn in the anoither two crystals-of RM24100, he observed style of Hqis & Chen (L976). The pekoite/ similar exsolution textures; probe analysis made gladite exsolution proves that these two minerals (or as a on the host mineral in oni lave results close to have been deposited at least existed) the composition of gladite. Correia Neves e/ a/., solid_solution, and Harris & Chen (1976) give gg70 ind, Harris & Chen (1976) have also good evidencethat the final quenching of these reporied quantitative exsolution studies which, pay.lave taken place at different lemperatures tolether viitn tn" results of the present work, in -different cases' fifoly establish an imrni.scibifity^ between pe- Ifowever, the, coexistence of krupkaite and bismutlinite (Welin 1966) as exsolutions and koite'and gladite. ^":"ffi'#r:Jiffi#ri"* rntergrowrhsamols fts mineralsortre series t#';"!;"S;il and described h toT: :1? have also been examined ;A;^ililLT-riiipr"'immiscibiliry gaps,'and detail. These are intergrowths of bismuthinite/ ;.;rd-;;;i;,ff -;r; co-ptiiuild i,n^" less fre- krupkaite; -bismuthinite/el{i!9; and ,"f"-ti""rnlir:b;A;.th;;n*a, i[" coexistence quently, bismuthinite/gladite/lrupkaitg : ;f ;;;;;i;;r"f. -uV t" good evidence that all (L975) - * reported bv zak et al. and **l; ;; f;r-.d-'iii"tti at l6w temperarures, per- M0ller (I972'l hx described a bismuthinitr ..rindstromite,; different stages in the mineralization lprotarty r.*pt"it"vpu-ooJi" intergrowth. ilj|.*: Attempts to prepare single crystals of gladite, pekoite or tru:ptiite telow 400"C weie not Cr,esslrrcetroN or SupBnsrRUcrtJREs successful. Very 1imi1s6 crystal growth was obtained and the powder patterns corresponded Previous attempts to classify the ordered to the solid-solution members. No superlattice mineral phases in terms of a general formula reflections were observed on any of the Guinier have proved unsuccessful in terms of predicting photographs. Attempts to anneal at 300-350"C the structures of all the known members. How- irystaG that have bein prepared at temperatures ever, based on present knowledge of the series,

TABLE8. ATOMICCOORDINATES*

Blrs"t cu.2Pb.2Bit,ss3cu.::Pb^::Blr.estSc cu-sPb.5Bit.5s3cu.7sPb.75Bll.2ss3 cuPbBis3 CuPbBiS,ti Bi(r) 0,0168 0.0165(3) 0.0157 (3) 0.0167 (4) o.ol 69 (3) 0.0175(3) o.or85(2) 0.6741 0.6762(4) 0.677114) 0.6767(3) 0.6780(3) 0.6802(4) 0.6812(21 Pb(ei 0.3407 0.3398(4) 0.3386(4) 0.3398(4) 0.333s(4) 0.3342(4) 0.3332(2) ) 0.4660 0.4701(5) 0.4726(4) 0,4720(4) 0.4839(4) 0.4864(3) 0.4S80(3) Cu - 0.2324(41) 0.2308(3e) 0.22e7(15) 0.2355(r 0) 0.2347(22t 0.2320(8) - o.2o1r (40) 0.2075(38) 0.2062(2o) 0.2089(11) 0.2076(24,0,2081(9) (r2) -l 0.0486 0.0488(21) 0.0516(20) 0.0481('l 1 ) 0.0487(r o) 0.0471 0,0454(9) 0.1300 0.I 345(22) 0.r291(te) 0.131 3(1 0) 0.r 363(20) 0.r389(lr) 0.1373(1r) 0.3785 0.3770(24) 0.3813(20) 0.3772(12) 0.3807(r 0) 0.3784(r2) 0.3795(9) 0.0576 0.061s{24) 0.0548(t7 ) 0.0585(17) 0.0522(7) o.o5r4(13) o.os53(ro) 0.2169 0.2164(20) 0.2136(2r) 0.2r85(r5) 0.2123(8) 0.2r28(14) 0.2146(e) 0.8062 0.802i(19) 0.8056(l9) 0,8032(l 7) 0.8055(19) 0.8054(12)0.8036(e)

*esd in parenthesls tKupcik & Vesela-Novakova(1970) r"Kohatsu & [uensch (1973), 332 TTIE CANADIAN MINERALOGISf

s 'F o !, Toc -c s E,5t = o c c E = 5 E ts. o .- J E E 0 |n = i 59 (9 o_ b

30 40 50 MOLE% BrzEs FIa. 9. Compositional ranges (see Harris & Chen 1976) in terms of mole Vo BirS" and suggested sub- solidus phase relationships in the system Bi:Ss-CuPbBig. The liquidus and solidus are from Springer (r97r). it is possible to represent the ideal composition ConnsrA Nnveg J. M., Lorrs NuNes, f. E., of all of the ordered members between bis- Seneve, TH. G., LnnrnvrN, M. & Kxonnnc, muthinite and krupkaite by the formula O.Y. (1974): Bismuth and antimony minerals in CuJb;Biu"Sre, where r = 0 for bismuthinite, the granite pegmatites of Northern Mosambique. A, x = | for pekoite, x = 2 for gladite, and Rev. Cienc. Geol., Lourenco Marques 7, Ser. t-37. x = 3 for krupkaite. Two of the other known members between krupkaite and aikinite may CRurcKsHANK,D. W. J., P[LrNc, P. E., BuJose, (1961): also be represented by this formula. These are A., Lovrr.l, F. M. & Tnurrn, M. R. Computing methods and the phase problem. In hammarite, for which x = 4, and aikinite, for - X-roy Crystal Analysis. Pergamon Press, Ox- which.r 6. Welin's lindstromite, CusPbsBizSrr, ford. with a subcell 5 times that of bismuthinite, is DAI,Y, J. J., Srrrnnrcs, F. S. & Wm,ATLBY, P. J. the only member of the series which does not (1963): Monsanto Research, S.{., Final Report satisfy this simFle classification and apparently No. 52. does not have an extensive solid-solution ranee (1976): (Harris Hennrs, D. C. & CnrN, T. T. Crystal chem- & Chen 1976). istry and re-examination of nomenclature of Can. The reason for this apparent exception must sulfosalts in the aikinite-bismuthinite series. Mineral. L4, 194-205. be sought in the crystal structure of lindstromite, and classification of the system can be further Iro, T. (1973): On the application of a minimrrm residual clarified only as the structure of it and any method to the structure determination of super structures. Z. Krist. L37, 399-411. additional members are determined - thoueh judging by the solid-solution ranges presentid Kenur-Mgr-r-rn, S. (1972): New data on pavonite, gustavite some related sulphosalt by Harris & Chen (1976) there is little scope and minerals. Neues Jahrb. Mineral. Abh. 177, 19-37. for this. The use of distinct names for the or- dered members of the sequence is to be en- KupcrK, V. & VBsne-NovAKoLA, L. (1970): The crystal structure redetermined. couraged, as it emphasizesthe real nature of of bismuthinite Tschermaks Mineral. Petrog. Mttt, 14, 55-58. the composition series. Kouersu, I. & WurNscn, B. J. (1971): The crystal structure of aikinils PbCuBi$. Acta Cryst,B.27, RBFERENCES 1245-1252. & .-- (L973): The crystal srructure Busllc, W. R., MARrr\, K. O. & Lew, H. A. of gladite, PbclrBisso. Amer. Minerur. 58, 1098 (1962): ORELS, a Fortran crystallographic least- (Abstr.). squares program. U.S, Nat. Tech. Inf. $erv. Lencn, R. R. (1974): Gold, Bismuth, Coppe4 ORNL.TM.3O,. Selenium Mineralization in the Tennant Creek PBKOITE 333

District, Cen*al Australia. Ph.D. thesis, Univ. eralien aus der Gruppe von Wismutglanz und New England, New South Wales. Aikinit. Chem. Erde 18, 1+18. & Muuur, W. G. (1975): Junoite, RAMDoHR,P. (1969): The Ore Minerals and Thelr *wittite" and related seleniferous bismuth sul- Intergrowths, 3rd ed., Pergamon, Oxford. fosalts from Juno Mine, Northern Territory, SyNEcF<, V. & Hvor-nn, J. (7975): The crystal Australia. Econ. Geol. 70, 369-383. structures of krupkaite, CuPbBLSo, antl of gladite, Moonr, P. B. (1967): A classificationof sulfosalt CuPbBisSg, and the classification of $uperstruc- structures derived from the structure of aikinite. tures in the bismuthinite-aik;ni1s grotp. Neues Amer. Mineral. 52, L874-1876. Jahrb. Mineral. Monatsh., 541-560. Mururvrn, W. G. (1975): The c:rystal structure of SpRnIcER, G. (1971): The synthetic solid-solution krupkaite, CuPbBisSo,from the funo mine at series BirSs-BiCuPbSs (bismuthinite-aikinite). Tennant Creek, Nortiern Territory, Australia. Neues Jahrb. Mineral. Monatsh,, 19-24. Amer. Mineral. 60, 300-308. Weltry Eruc (1966): Notes on the mineralogy of , WnrN, E. & WrrENscn,B. f. G976): Sweden. 5. Bismuth-bearing sulphosalts from Crystal chemistry and proposed nomenclafire Gladhammer, a revision. Ark, Mineral. Geol. 4, for sulfosalts interrnediate in the system bis- 377-386. muthinite-aikinite (BirSr-CuPbBiS) Amer. Mln- Zer, L,, SvNEcrx, V. & Hvsr-sn, J. (1975): KruIr eral,6L, t5-20. kaite, CuPbBrrS., a new mineral of the bis- OuMesA,M. & Nowecrr, W. (1970): Note on the muthinite-aikinite group. Neaes lahrb. Mineral space gxoup and on the structure of aikinite Monatsh,, 533-547, derivatives. Neues Jahrb. Mineral. Monatsh. 158-162. Manuscript received lanuary 1976, emended March PADERA,K. (1956): Beitrag Zur Revisionder Min- 1976.