Canadian MineralogistVol. 25, pp. 393-399 (1987)

STRUCTURAL DISORDER IN LINDSTNOUTT: A BISMUTHINITE _ AIKINITE DERIVATIVE

ALLAN PRINGSouth Australian Museum, North Terrace, Adelaide, South Australia 5A00, Australia

BRUCE G. HYDEResearch School of Chemistry, Australian National University, G.P.O. Box 4, Canberra 2601, Austrolia

AssrRAcr

High-resolution electron microscopy is used to examinethe nature and extent of structural disorder in lindstriimite,Cu3Pb3Bi7Sl5, a member of the bismuthinite - aikinite ser-ies. Well-ordered regions of lindstrdmite were found to beintergrown with lamellae of krupkaite, CuPbBi3S6.Aikinite lamellae were also observed within extensiveregions of krupkaite, and homogeneous regions of krup-kaite and hammarite were also detected. An understand-ing of the mechanism of compositional variation in the end-member phases bismuthinite and aikinite will require fur-ther study.

Keywords : high-resolution electron microscopy, lindstnim-ite, aikinite, krupkaite, hammarite, bismuthinite series,compositional fields, structural disorder, type specimen.

Sora4ArRE

On a 6tudi6 la nature et la port6e du d6sordre structuraldans la lind$trdmite Cu3Pb3Bi7S15, membre de la s6rie bis-muthinite - aikinite, par microscopie dlectronique i hauter€solution. Des rdgions bien ordonndes de lindstrdmite sontjuxtaposees i des lamelles de krupkaite CuPbBi3S5. Onvoit aussi des lamelles d'aikinite dans des domaines de krup-kaite, et des rdgions homogbnes de krupkaite et de ham-marite. Une connaissance du m6canisme de variation dela composition des pdles bismuthinite et aikinite requiertune 6tude plus approfondie.

Clraduit par la Rddaction)

Mot*clfs: microscopie 6lectronique d haute rAolution, lind.str6mite, aikinite, krupkaite, hammarite, s6rie de la bis-muthinite, champs de composition, d€sordre structural,6chantillon type.

INTRoDUCTIoN

The chemical and strufiurar rature of the bis-muthinite - aikinite series has been extensivelystudied (Makovicky 1981). Six mineral phases thatare chemically intermediate between bismuthiniteBi2S3 and the fully stuffed aikinite CuPbBiS3 havebeen identified (Table 1). The bismuthinite structureis the archetype for the group, the structure of theother minerals being derived by ordered substitutionof cations, insertion of Cu and supercell ordering.It can be viewed as composed of quadruple BiaS5chains in which half the bismuth is in three-fold co-ordinatiou, and the other half is five-co-ordinate,

393

forming BiaS6 ribbons Kup&lk & Veseld-Novikov61970) (Fie. ia). Sut the structure can also be readilyderived from the NaCl type by shear of the{ 1 13 } .. . layers and periodic twinning (Hyde et al.iszsj] fi'this structure, the co-ordination polyhedraof bismuth are best visualized as capped trigonalprisms, of which there are two crystallographicallydistinct sets (Fig. 1b).

Aikinite is derived from bismuthinite by replac-ine half of the Bi3" by Pb2+ (+ interstitial Cu") inan ordered fashion (Ohmasa & Nowacki 1970,Kohatsu & Wuensch 197 1). Compositionally, krup-kaite CuPbBi3S6 lies midway between bismuthiniteand aikinite:-only one quarter of the Bi3+ arereplaced by Pb2+, and Cu+ occupies only half theavailable tetrahedral sites. This substitution alsooccurs in an ordered fashion, and results in a non-centric structure rather than the centric structures ofbismuthinite and aikinite (Mumme 195)' The struc-tures of the other intermediate members are com-posed of ordered intergtowths of bismuthinite (D),krupkaite (k) and aikinite (a) ribbons.

Structurally, lindstrdmite CurPb,BirSt, is themost complex intermediate member' being based ona ten-ribbon intergrowth^(2a + 8k).Its unit cell isq 56.11, b tI,57, c 4.00 A; thus it is a 54 supercellof bismuthinite (Fig. 2). The structure was deter-mined by Horiuchi & Wuensch OnT, but notwithout considerable difficulty. To refine the struc-ture fully, they found it necessary to assume that thecrystal studied contains 6.3 vol.9o of exsolved krup-kaite.

In contrast to the highly ordered intergrowth-structures reported for the intermediate phases,Harris & Chen (1976) reported extensive solid-solution fields for the end-members aikinite andkrupkaite. The degree of chemical segxegationrequired for the formation of the well-orderedintergrowth-structures of the intermediate phasesseems at odds with the concept of large composi-tional fields due to solid solution in the end mem-bers. The compositional fields could well be due tothe disordered intergrowth of end-member ribbons.

We undertook high-resolutron electron-microscopy studies of lindstrdmite to confirm thestructure and to determine the nature of the krup-kaite exsolution assumed by Horiuchi & Wuensch(re77).

394 THE CANADIAN MINERALOGIST

TABLE T.

M i n e r a L C o n p o s i t i o n a8 c8bR S p a c e S t r u c t u r e R e f e r e n c eG r o u p

B i B n u t h i n i t e

Peko i i re

Gladitre

Krupkai l re

L i n d s t r o m l t e

Hammari te

F r i e d r i c h i t , e

A i k i n i t e

B i " s "

C u P b B i 1 1 S 1 g

C u P b B i 5 S 9

CuPbBi 356

c u 3 P b 3 B i z s r 5

C u 2 P b 2 B i 4 S 9

t t5oo5u tTs t 8CuPbBi S3

1 r . 2 3 1 1 . 2 7 3 . 9 1

3 3 . 5 0 r r . 3 2 3 . 9 9

3 3 . 6 6 r 1 . 4 5 4 . 0 2

1 1 . 1 5 r 1 . 5 1 4 . 0 1

5 6 . 0 7 1 1 . 5 7 4 . 0 1

3 3 . 4 5 1 1 . 5 8 4 , 0 1

3 3 . 8 4 1 r . 6 5 4 . o l

1 1 . 3 2 1 r . 6 4 4 . 0 4

Pbnn

P b 2 l n

Pbno

P b 2 l m

P b 2 l d

Pbnn

P b 2 l n

Pbnn

b

4b

2 b -

8 k *

4 k *

2 k +

2 k

4 k

2 a

4 a

K u p E i k & v e s e l 5 - N o v i k o v i ( 1 9 7 0 )

M u m m e & l t / a l t s ( 1 9 7 6 )

K o h a t s u & W u e n s c h ( 1 9 7 6 )

M u @ e ( r 9 7 5 )

H o r i u c h i & w u e . s c h ( 1 9 7 7 )

H o r i u c h i & W u e n s c h ( 1 9 7 6 )

C h e n e t a I . ( 1 9 7 8 )

K o h a t 6 u & w u e n s c h ( 1 9 7 1 )

0 h m a s a & N o u a c k i ( 1 9 7 0 )

b = b i s n u t h i n i t e r i b b o n s k = k r u p k a i t e r i b b o n s

ExprruvteNtar-

Several small crystal fragments of lindstrdmitefrom the type specimen (RM24100) were obtainedfrom Dr. Eric Welin, Naturhistoriska Riksmuseet,Stockhqlm. Crystals were ground under acetone inan agatb mortzu and disper$ed on Cu grids coatedwith holey-carbon support films. The fragments wereexamined in a JEM 200CX electron microscope fit-ted with an ultrahigh-resolution top-entry goniom-eter (t l0o), objective lens with C" : 1.20 mm anda LaB6 filament. Crystal fragmentl projecting overholes in the carbon film were tilted inro the t00llzone. Lattice images, at 380,000x magnification,and electron-diffraction patterns were recorded usingexposnres of 24 s.

OBSERVATIoNs

A series of [@l]-zone electron-diffraction patternswas taken from various fragments of the specimen(Fig. 3). The range ofdiffraction patterns attests tothe inhomogeneity of the type material: Figure 3ashows only the strong rectangular array of subcellreflections, which corresponds to the basicQismuthinite-krupkaite-aikinite cell ( - | 1.2 x I 1.6A). Figure 3b has the same set of subcell reflectionsplus additional weaker reflections (spacing aEl5)along a*, presumably due to the ordering of krup-kaite and aikinite ribbons in the lindstrdmite struc-ture. The superlattice reflections are somewhatstreaked (cf Fie. 3d), indicating the presence of somedisorder. A highly ordered superstructure is,however, dominant. In crystal fragments where dis-order (of krupkaite and aikinite ribbons)predominates, the supercell reflections are reducedto weak streaks along o* (Fig. 3c). Regions of ham-marite, with a 3a supercell, were found intergrownwith krupkaite and lindstrdmite in the specimen (Fig.3d). Note that the superlattice reflections are verysharp, indicating that the supercell is well ordered,a conclusion confrrmed by the lattice images.

a = a i . k i n i . t e . i b b o n s

A lattice image of the well-ordered lindstr6mitecrystal (corresponding to Fig. 3b) is shown in Figure4a,Its 5a supercell dominates this low-magpificationimage, but the crystal also contains a?.00-Alamellaof krupkaite on the right. Dark patches in the imageare due to damage by the electron beam. Note thatthe image contains two types of ribbons that differin contrast: the ordering of the ribbons is clearly visi-ble. Figure 4b shows a disordered intergrowth ofkrupkaite and aikinite units, with some small regionsof the lindstrdmite structure. Large regions of per-fectly ordered lindstrdmite structure were notencountered in this study, although several regionsof well-ordered hammarite were noted (Fig. 4c).

Interpretation of the images

The observation of well-ordered lindstr6mireregions interlaced with exsolution lamellae of krup-kaite in the type material serves to confinn and clar-ify the findings of earlier workers. During their struc-ture refinement, Horiuchi & Wuensch (1977) had toassume that their crystal of lindstrdmite contains6.390 exsolved krupkaite. The electron micrographsdemonstrate the validity of this assumption. Welin(1966) observed exsolution lamellae in the hand speci-men, and demonstrated by electron-microprobe anal-ysis that these lamellae represent a sulfosalt of thebismuthinite - aikinite group but of a slightly differ-ent composition. What Welin found were clearlyexsolution lamellae of krupkaite in lindstrdmite. Thelattice images show that these lamellae extendthroughout mugh of the sample, albeit on a very finescale (100-200 A in some cases). The imaging studiesalso reveal that the type specimen consists of an inter-growth of lindstrdmite, krupkaite and hammarite.

More detailed structural and compositionalinterpretation of the lattice images is greatly ham-pered by the large thickness of the crystal fragmentsand the complex nature of the structure: image detailis highly sensitive to thickness, and to beam tilt anddefocus of the objective lens. Given that the image

STRUCTURAL DISORDER IN LINDSTROMITE 39s

pr#q%q

#^Sli

Frc. 1. Schematic diagrams of the structure of bismuthinite down [001]. (a) Bi4S5 chains in which half the number ofbismuth atoms is three-co-ordinate and the other half is five-co-ordinate (Kupdlk & Vesekl-Novrikov6 1970). (b)

Bismuth co-ordination polyhedra, two crystallographically distinct sets of capped trigonal prisms (Hyde et al. 1979)(the capping is not shown).

Frc. 2. Schematic diagram of the structure of lindstr<imite, Cu3Pb3Bi7S15, down [0011. Structure is composed of anordered intergrowth of 2 aikinite and 8 krupkaite ribbons. Copper atoms occupy tetrahedral sites.

contrast can be seen in Figure 4b; the contrastbetween the two tlpes of ribbons is reversed, so thathere the light ribbons are aikinite.

Ranges in comPosition

The composition of lindstrdmite was establisledby Welin (1966) as Cu6Pb6Bi1aS3s [= Cu+PbaBirzSz+ Cu2Pb2Bi2S6, i.e.,4 krupkaite to 1 aikinitel; thiswas later confirmed by results of a crystal-structureanalysis (Horiuchi & Wuensch L977).The currentlyaccepted formula differs slightly from the composi-tion given in the original description by Johannson(IgZ), CuPbBi3S5, which is the currently accepted

S o O.25 .O.75 Bi o O 25 a O.75

C a o a

B i O a

P b o a

roAI ' r r r l r r r r l

in Figure 4a was recorded with the value of objec-tive lens defocus near the Scherzer defocus (-800 A),limited direct-image interpretation can be made byconsidering charge density alone. Image detail isdominated by two types of ribbons, the darker rib-bons being in the minority. The lindstrdmite struc-ture contains aikinite and krupkaite ribbons in theratio 1:4; this is approximately the ratio of dark andlight ribbons in the lattice images, suggesting thatthe dark strips are aikinite-type ribbons. Pb and Biare indistinguishable, but the darker aikinite cor-responds to the larger concentration of interstitialcopper atoms [krupkaite: Cu@bBi)S6, aikinite:Cur(PbrBi)S61. The effects of thickness on image

TFIE CANADIAN MINERALOGIST

W*.

Ftc. 3. Electron-diffraction patterns taken down the [001] axis (lindstrdmite setting). (a) Iftupkaite. (b) Lindstrdmite(note the supercell reflections corresponding to a 5a supercell). (c) Disordered kruptaiie - ait<inite intergrowth (notestreaking along a'). (d) Hammarite (showing well-developed ia supercell).

formula for krupkaite. The difference between thetwo compositions is small and most probably dueto Johannson's chance selection of a portion ofthetype specimen in which krupkaite predominates.Large domains of pure krupkaite were found in thetype material during this study @ig. 5).

In addition to the fragments of lindstrdmite, ham-marite and krupkaite found by electron microscopy,many regions of disorder were encountered. The dis-ordered crystals fall into two broad categories: l)lindstromite with krupkaite disorder, and 2) krup-kaite with aikinite disorder. The widespread inter-growth of ribbons of aikinite and krupkaite withina matrix of krupkaite or lindstrdmite provides a

ready mechanism for significant variation from theideal composition. Extensive solid solution has beenreported in natural materials for both aikinite,Cu1-fb1_piSr-, (0 < x s 0,29), and krupkaiteCuPbBil,S6-r.s: (0 s x < 0.4) (Harris & Chen1976). Even though Harris & Chen (1976) foundstrealing in some of their X-ray-diffraction patterris,they do not appear to have considered that composi-tional ranges could be due to the disordered inter-growth ofvarious types ofribbons. They attributedthe streaks to 'stacking' disorder of aikinite units.The streaking can equally well be interpreted as dueto the disordered intergrowth of a number of krup-kaite ribbons within an aikinite matrix. Such an inter-

STRUCTURAL DISORDER IN LINDSTROMITE 397

gxowth would also result in a significant variationfrom the ideal composition, which indeed theyreported.

Springer (197 l) reported complete solid-solutionbetween bismuthinite and aikinite above 300oC insynlh{ic studies. He was lnable to prepare theordered)ntermediate phases, bven after annealing forup to 3 mo\ths, and concluded that the process ofchemical segregation leading to formation of super-structures can only occur bdlow 300"C. Clearly in

nature both mechanisms, solid solution and the dis'ordered intergrowth of ordered structure units, mayoperate.

But the nature of compositional variation in thebismuthinite-aikinite series clearly warrants furtherwork. With the new generation of high-resolutionanalytical electron-microscopes' and better imagesimulation, it should be possible to determine themechanism of compositional variation in theseminerals.

398 THE CANADIAN MINERALOGIST

FIc. 5. Lattice image of krupkaite taken down the [001] zone.

Frc. 4. High-resolution electron-microscope images of lindstrdmite (type material) down [001]. (a) Well-ordered regionof lindstrdmite. Note the perfect oqdering of the aikinite and krupkaite units to a 5a supercell. On the right of theimage is a lamella of krupkaite 200 A wide. (b) Disordered intergxowth of krupkaite and aikinite units. Note reversalof contrast from (a) due to thickness effects. (c) Well-ordered crystal of hammarite showing 3a repeat.

AcKNowLEDGEMENTS

We thank Dr. Eric Welin for providing the sam-ple of lindstrdmite, and Dr. T. White for his prelimi-

nary examination of this sample. Thanks are also dueto Dr. Robert F. Martin, Dr. Alan Clark and theanonymous referee who suggested improvements tothe text. One of us (A.P.) wishes to acknowledge the

STRUCTURAL DISORDER IN LINDSTROMITE 399

financial support of the Australian Research GrantsScheme.

REFEFGNCES

CurN, T.T., KIncnNen, E. & Pean, W. (1978):Friedrichite. Cu.Pb.Bi"S,", a new member of theaikinite - bismuitrinite's6iies. Can. Minerol. 16,ln-t30.

Henms, D.C. & CneN, T.T. (1976): Crystal chemistryand re-examination of nomenclature of sulfosalts inthe aikinite - bismuthinite series. Can. Mineral,74,r94-20s,

Homucsr, H. & WusNscn, B.J. (1976): The orderingscheme for metal atoms in the crystal structure ofhammarite, Cu2Pb2BiaSe. Can. Mineral. 14,536-539.

-& - (1977): Lindstr6mite, Cu3Pb3Bi7S15:its space group and ordering scheme for metalatomsin the crystal structure. Can. Mineral. l5' 527-535.

HvoB, B.G., Arppnssolq, S., Berrsn, M., Pt-uc, C.M.& O'KSBFFS, M.A. (1979): The (twin) compositionplane as an extended defect and structure-buildingentity in crystals. Prog. Solid State Chem. 12,2n-3n.

JoHerssoN, K. (1924): Bidrag till Gladhammar -gruvornas mineralogi. Arkiv Kemi Mineral. Geol.9, r-22.

Konarsu, I. & WureNscs, B.J. (1971): The crystal struc-ture of aikinite, PbCuBiS3. Acto Cryst. B.27,las-r2s2.

-& -(1976): The crystal structure of glad-

ite, PbCuBi{Sq, a superstructure intermediate in theseries Bi2S3---PbCuBiS3 Oismuthinite - aikinite).Acta Cryst. 832, UOI-24O9.

Kudir, V. & VeserA-NovAxovA, L. (1910): ZwKristallstruktur des Bismuthinits, Bi2S3. Tscher-maks Mineral. Petrog. Mitt. t4,55-58.

Maxovtcry, E. (1981): The building principles andclassification of bismuth - lead sulphosahs andrelated compo vnds. Fortschr. Mineral. 59' I 37- I 90.

MurvttvtB, W.G. (1975): The crystal structure of krup-kaite CuPbBi3S5, from the Juno Mine at TennantCreek, Northein Territory, Austtalia, Amer.Mineral.60, 300-308.

- & Werrs, J.A. (1976): Pekoite, CuPbBiqlSls'a new member of the bismuthinite - aikinite mineralseries: its crystal structure and relationship withnaturally- and synthetically-formed members, Can.Mineral. 74, 322-333.

Ornrrese, M. & Nowecrr, W. (1970): A redeterminationof the crystal structure of arkinite, [BiS2[S]Cu''Pb l. Z. Krist. 132,71-86.

SpnrNcsn, G. (1971): The synthetic solid-solution ser-ies Bi,S: - BiCuPbSr (bismuthinite - aikinite).Neuesiahrb. Mineral. Monatsh., 19-24-

Wrrnr, E. (1966): Notes on the mineralogy of Sweden.5. Bismuth-bearing sulphosalts from Gladhammar,a revision. Arkiv Mineral. Geol. 4' 377-386.

Received July 30, t986, revised manuscript acceptedNovember 7, 1986.