Cunadian Mineralogist Vol. 19"pp.3l5-324 (1981)

THENEW NULLAGINITE AND ADDITIONALDATA ON THE RELATEDMINEBALS ROSASITE AND GLAUKOSPHAERITE

E.H. NICKEL Dit,'sionof Mineralogy,CSIRO, Private Bag, P'O., Wembley, WesternAustralia 6014, Australia

L.G. BERRY Departnrentof Geological Sciences,Queett's University, Kingstorl,Ontario K7L iN6

Assrnecr nullaginite, espacenouvelle. en nodules et-en veinu- les ir fibres transverses dans un assemblage riche Nullaginite. Nir(OH)rCOr. is closelv related to en nickel d6couvert dans le gisement d'Otwav (dis- rosasite, (Cu,Zn)n(OH):COr, and glaukospJraerite, trict de Nullagine, Australie occidentale). Ses para- (Cu,Ni)r(OH)2COi. It occurs as nodular grains mBtres rdticulaires, d6termin6s sur diagrammes de and as cross-fibre veinlets in a nickel-rich assem- fibre et de poudre, ainsi que par-TED' sont: d blage from the Otwav deposit, Nullagine district, s.236(3\,D 12.001(6),c 3.09r(2) A. p 90'48(7)', parameters Western Australia. Unit-cell from X-ray sroupe spatial possible P2t/m. Les huit 9ie9-l9s fibre rotation and TED patterns and refined from plus'intenses du clich6 de poudre ldhht?\(hkl)1 - powder data are a 9.236(3). D = 12.001(6).c sont: 7.30(30b, variable)( I l0), 5.04(30)( 120)' - 3.091(2) A, p = 90.48(7)'. possible space 4.62GO)(200), 3.66(40)(220,130), 2.579(100) group P2,/nr. Strongest powder-diffraction lines are O0r). 2.s57(90)(201), 1.545(30)(002). 1.541(30) 7.30(30b, variable) (ll0), 5.04(30)(120), 4.62 (600). D'un vert brillant. ce min6ral est l6g0rement (40)(200), 3.66(40)(220,130), 2.579(100) (201), pl6ochroique, avec absorption maximum perpendi 2.ss7(90)(20r), 1.545(30)(002), 1.541(30)(600). culairementir I'axe de la fibre (- Xltc = 6'). The mineral is bright green and slightly pleochroic, Les indices de r6fraction sont a 1.67, F - I 1.78. with maximum absorption normal to fibre length c. On attribue la diff6rence entre densit6 mesur6e - Indices of refraction are tr = 1.67. 9-l 1.78. 13.56(3)l et calcul6e t4.071 a la pr6sence d'eau ExtinctionX{c = 6'. S.G. is 3.56(3) (meas.) and dans le sp6cimen.La duret6 par indentation, YHNzo, 4.07 (calc.); the large discrepancyis attributed to est de 34(9). the presence of admixed water in the specimen. Pour la rosasitede Bakara (Bulgarie). les donn6es - Indentation hardnessis VHNzo 34(9). nouvelles de diffraction X et de TED ont permis New X-ray-diffraction and TED data on rosasite I'affinement des paramEtresdu r6seart: a 9.366(3)' from Bakara, Bulgaria, give refined unit-cell para- b 12.1t6(4), c 3.127(l\ A, p so.tt(a)'. groupe - - - metersa 9.366(3),b 12.116(4),c 3.127(l) spatial possib'leP2,/m. Irs rdsultats de diffraction A. 9 : 90.11(4)o; possiblespace group PZr/nt. X obtenus sur la glaucosph6rite de Kasompi X-ray crystallographic data for glaukosphaerite (Zaire) donnent des paramdtres aa' et DE semblables from Kasompi. Zaire, are consistent with two pour tous les sp6cimenset conduisent i I'un ou patterns having similar a* and 6* parameters.One I'autre de deux r6seaux: l) orthorhombiqte-A, a - has an orthorhombic-,4 lattice with a 9.354(2\, 9.354(2). b 23.954(4), c 3.128(1) A, sroupe spa- - b 23.954(4'), c : 3.128(l) A: pos-siblespace tial possibleAntmm, 2) monoclinique,a 9.35(l)' group Ammm,' the other has a monoclinic lattice h 11.97(1).c 3.13(l) A, p 96(1)". semblablei - - - with n 9.35(t). b t1.97(t),c 3.13(t) A, celui de la malachite. Il faudra d6terminer la - lt 96(l)', similar to that of malachite. Structural structure cristalline et faire plus d'analvses ir la studies and additional electron-microprobe analyses microsonde 6lectronique pour pr6ciser les relations are needed further tb elucidate the relations be- entre ces deux espdces. tween these two . (Traduit par la R6daction) Keywords: nullaginite. rosasite, glaukosphaerite, Mots-clAs: nullaginite, rosasite, glaucosph6rits, new mineral, Otway deposit, Nullagine district, espAce nouvelle, gisement Otway. district Nulla- Western Australia, Ni hydroxy carbonate, X-ray gine. Australie occidentale, hydroxy-carbonate de powder data. Ni, clich6 de poudre.

SovvrernE DgscRtprtoN oF NULLAGINITE La nullaginite Nir(OH)rCOs, (Cu, la rosasite Introcluction Zn ),( OH ) et la glaucosph6rite(Cu,Ni ),(OH ) CO* sont "CO36troitement apparent6es.On trouve la In the Nullagine district of Western Australia, 315

Downloaded from http://pubs.geoscienceworld.org/canmin/article-pdf/19/2/315/4005550/315_19_2cm.pdf by guest on 25 September 2021 316 THE CANADIAN MINERALOGIST

a nickel-rich mineral assemblage occurs in Nullaginite is named for the district in which shears in serpentinized peridotite, at a local.ity it was found. Both mineral and name have known as the Otway deposit. The principal been approved by the I.M.A. Commission on nickel minerals, millerite, polydymite and pe- New Minerals and Mineral Names. coraiteo occur chiefly as nodular grains. Other nickel minerals, occurring in lesser amounts, 0ccurrence are gasp6ite,otwayite, pa.rkerite,shandite, breit- hauptite and nullaginite, the new mineral de- Nullaginite occurs as nodules and as veinlets. scribed in this paper. The occurrence of nulla- The nodules, ovoid to irregular in shape, bright ginite was reported briefly in a general paper green in color, up to 2 mm in diameter, are on the geology and mineralogy of the deposit found in a matrix of chlorite and nickeloan (Nickel et al. 1979). Previous papers on the serpentine(Figs. l, 2). The nodulescommonly mineralogy of the Otway deposit describe gas- contain magnetite, which is usually distributed p6ite and pecoraite (Nickel 1973) and the re- cently discovered mineral otwayite, NL(OH)' COg.HzO (Nickel et al. 1977).

Ftc. 3. Electron-backscatter micrograph of the same nodule shown in Figure 2, but at a higher magnification, showing the felted nature of the nullaginite comprising the nodule. FIc. l. Photograph of coarsely ground surface of hand specimen. sample 10025, showing altered nullaginite nodules (white), rimmed by magnetite (black), in a matrix of chlorite and nickeloan serpentine(foliated).

Frc. 4. Electron-backscatter micrograph of a pol- ished section of sample 11052, showing zonal intensity variations in a veinlet of cross-fibre nullaginite. The white mineral on both sides of the nullaginite is pecoraite, and the dark zones Ftc. 2. Electron-backscatter micrograph of a nul- in the vein are magnesian pecoraite. The line laginite nodule in a polished section of sample scan shown in Figure 5 was made along the line 10025.The white mineral is magnetite. A-8.

Downloaded from http://pubs.geoscienceworld.org/canmin/article-pdf/19/2/315/4005550/315_19_2cm.pdf by guest on 25 September 2021 NULLAGINITE, ROSASITE AND GLAUKOSPHAERITE 317

20

%sl %Nl (App.or) (App.ox)

Dlatanco (nn)

Frc. 5. Electron-microprobeline scan acrossthe veinlet shown in Figure 4.

around the grain margins. At high magnifica- mately 0.5 x 0.05 mm) were examined by tions under the scanning electron-microscope rotation and Weissenberg X-ray-diffraction (SEM), the nodules are seen to consist of a methods. One fragment yielded a powder dia- felted intergrowth of crystallites (Fig. 3). The gram on rotation, and two gave fibre diagrams nodular nullaginite is commonly intergrown with a clear separation of zero and first levels with pecoraite, which forms nodules virtually (a separationof -2O mm with Co Ka radiation identical in appearanceto those of nullaginite. in a camera of radius 36O/2n mm), indicating In a few samplesthe nullaginite has been altered a period of c = 3.1(l) A, with all diffractions to a pale green, chalky material (Fig. l) that extended 2 to 5 mm along the powder arcs gives X-ray-diffraction patterns corresponding to either side of the median layer-line position. to nullaginite and gasp6ite.These chalky nodules No reflections show spot maxima as observed seem to be a product of decompositionof the on similar films of glaukosphaerite. On the nullaginite. zerolevel Weissenbergfilm, }&0 reflectionsex- The nullaginite occurring in veinlets consists tend almost the full length (lO0 mm) of the of tiny cross-fibre crystallites and is virtually exposed film, parallel to the rotation axis. The identical in appearanceto otwayite (Nickel er reflections recorded on the zeroJayer line of al. 1977).The nullaginiteis generallyintergrown the rotation films and on the zeroJayer Weis- with pecoraite, locally in a strikingly zoned senberg film are, for the most part, very sharp fashion; this is particularly evident in SEM arcs or lines: the reflectionscan be indexed on electron-backscatterimages (Fig. a). In such orthogonal axes with a = 9.23, b -- 11.95 A images, the intensity of backscattered radiation (average from several films), corresponding is a function of the averageelectron-backscatter to a and D of malachite and glaukosphaerite coefficient, which is, in turn, a direct function (Table 1); a few weak and diffuse reflections of composition. The vein nullaginite shown in are nearly complete powder rings. The most Figure 4 can be seen to exhibit compositional intense diffractions on the rotation films are a differences normal to the veinlet walls. Com- set of four arcs with maxima on the first layer parison with the electron-microprobeline scan line: thesearcs are much broader (ca. 0.3 mm) across the veinlet (Fig. 5) indicates that the than the zeroJevel diffractions and do not ex- differences are attributable, in part, to variations tend to the zero-levelrow. A rotation film-chart in silica content. The high silica content along overlay gives an average value for these dif- the margins of the vein is due to pecoraite, fractions of 0 : 2O.5".d - 2.56 A, which cor- which gives a relatively high electron-back- responds closely to the average of dh" and dzo, scatter intensity. However, the high content of derived from transmission electron-diffraction silica within the nullaginite veinlet itself cor- (TED) and powder-patternindexing. responds to a relatively low intensity due, at Transmission electron-diffraction patterns least in part, to the presenceof magnesium; this from six random fragments of nullaginite give - suggeststhat the low-intensity phase that forms a = 9.27(8),c = 3.13(4)A, B 90.4(3)'. the core of the veinlet in Figure 4 is probably The patterns were recorded on plates after trial- magnesianpecoraite. and-error tilting of the grain about any horizon- tal axis in the plane of the stage to improve the Crystallography pattern. The patterns display twofold rotational symmetry, with interaxial angles between 90.1 Three small needle-like fragments (approxi- and91.4o.

Downloaded from http://pubs.geoscienceworld.org/canmin/article-pdf/19/2/315/4005550/315_19_2cm.pdf by guest on 25 September 2021 318 THE CANADIAN MINERALOGIST

TABTEI. HALACHIIE,ROSASI]E ANO PELATED I,IINERALS

FonnuIa sp. cr. a[ bE cA q Reference

PZJa 9.49 l?.00 3.24 - Ramsdell& l,llolfe (1950) mal achJte Cu2(0H)2C03 I 98.42', 'PZJa 9.502 11.974 3.24 - 98'45' swansona, aL. (1960) (synthetl c) rosasite (Cu,Zn)2(0H)2C03 P?r/n 9.366 l?.1r6 3.127 - 90006' This study: Bakara, Bulgaria or P21 zincrosasite (Zn,Cu)2(0H)2C03 glaukosphaerite (lu,Ni)2(0H)2C03 9.34 rr.ss:.oz - ,o-rr": ;:;::'-tIll',,,rn, '1 91a ukosphaeri te 9.364 f.93 3.413 - 92.25' - Jambor(1976a) Kambalda, W.A. gl aukosphaefi te v. Jo6 11.99 3.387 - 92.12" Kasompi,Zaire '12.05 gl aukosphaeri te 9.52 3.21895,45o92.92o 90.78o Deliens & Plret (l 980) Kasompi, Zaire Amnm 9.354 aukosphaeri te r 23..t.1.97954 3.1 28 This study Kasompi,Zaire 9l , P?JA 9.35 3..13 - 96" : ] gl aukosphaerite 9 .363 23.896 3.138 - ThJsstudy Kambalda, ll.A. kolwdzite (Cu,Co)2(0H)2C03 ry 9.50 12.f5 3.18993.32"90.74" 91,47' Deliens & Piret (l980) Kolw6zi,Shaba nul laginite tli2(0H)2C03 P2Jn 12.001 3.091 - Thisstudy, otway deposit ot Pzt Nullagine, ltJ,A. Intennediate compoundin topo- 9.34 3.15 12.1 8 90" - Gi.inter& Oswald(1977) tactlc therma'l decompositlon of artinite l1g2(0H)2C03 9.34 12.tB 3.15 - 90" " (rosasitesetting) mcguinnessite (l.1g,Cu)2(Ott)2C03 nameapproved, no data published,oswald& crook( l 979)

From one other single grain, five patternsob- lishes the correct indexing of non-&&0 reflec- tainedat anglesof tilt of 3 .N,S.E and W and from tions in malachite and glaukosphaerite.In nul- the original position about NS and EW ar(esof laginite the distinctivedoublet at d = 2.579 and - thestage give a:9.20(6), c - 3.09(3)A, B d 2.557 provides a strong clue to the index- - 90.9(1)". In this set of patterns,the most ing. TED diagrams (h}t) have 2Ol with d : intense pair of diffractions is consistently2Ot 2.58 and 201 with d - 2.55 as the strongest and 201, lying in the obtuse angle betweend* diffractions,with llor ) /.ut.This indexingis sup- and cE. The microscopewas operatedunder the ported by a strong firstlayer diffraction on the same settingsthroughout; the patternswere ex- fibre-rotation diagram (201 and 201 unresol- ternally calibrated with graphite. No effort was ved). Refinement of the powder data of nul- made to tilt grains oriented on lr0l to other zero laginiteyields c - 9.236(3), b = 12.@L(6), c levels because it is difficult to rotate a grain : 3.091Q) A, B = 90.48(7)'; I line of .f = about a particular axis, and the normally short 10 and three of I -- 5, of uncertain hkl, are life of a grain in the beam precludesextensive presumed due to unidentified impurities. The reorienting. refinement identifies a broad line with d : The X-ray powder-diffraction pattern of nul- 2.932 as T0l and 101, establishingthe possible laginite is very similar to those of rosasiteand space-groupas PZ,/m or P2r. glaukosphaerite(Table 2). A notable feature of the pattern obtained from a high-resolution Cornltosition Guinier camera, although not observable on a normal _Debye-Scherrerpattern, is the splitting Attempts were made to concentrate sufficient of the 2Ol-2Ol doublet. The intensity of the amounts of pure material for chemical analysis, 7.3 A line varies from pattern to pattern and both by hand-picking and heavy-liquid and mag- appearsto, be due, at least in part, to different netic separation prOcedures; all concentrates amounts of an admixed serpentine mineral, were found to contain substantial amounts of probably pecoraite. impurities, especially pecoraite and gasp6ite. The similarity of the zero-level diffractions Eventtially these attempts were abandoned in in rotation and Weissenbergfilms of malachite, favor of the electron microprobe. This is not glaukosphaerite and nullaginite determines the entirely satisfactory, since hydroxyl, carbonate indexing of many of the high d-value powder and possibly water must be inferred by cal- lines: these lines are sharp and not split into culation; nevertheless,there was no reasonable doublets.The first-level Weissenbersfilm estab- alternative.

Downloaded from http://pubs.geoscienceworld.org/canmin/article-pdf/19/2/315/4005550/315_19_2cm.pdf by guest on 25 September 2021 NULLAGINITE. ROSASITE AND GLAUKOSPHAERITE 3t9

TABLE2. X-RAYPOIIDER-DIFFRACTION PATTERNS OF NULLAGINlTE,ROSASITE AND GLAUKOSPHAERITE

NULLAGINITE ROSASITE CLAUKOSPHAERI TE otway,Nul laglne,',{est Austral'ia Bakara,Bulgaria Kasompi, Zalre Kambalda,!l.A. a.s.236,b=12.001, c=3.091.4, B=90o29' a=9.366,b=12..] 16, c=3.1274, a=9.4,b=24.0, c=3.14 6=90"06' I dr"u, hkl d""1". I,^ d"* I, I dr.u. hkl d.ul.. Id hkl oca I.. I dr"., . l 'I0 30b 7.302 110 7.319 w 7.40 w l0 7,46. r10 7.410 7.398 I?0 7.372 w 20 7.413 <5b 6.018 020 6.001 vr 6.0] w 60 6.06 020 6.058 50 6.002 040 5.988 s 40 5.991 30 5.038 120 5.032 n 5.04 vs 80 5.094 120 5.087 55 5.048 140 5.043 s 50 5.048 'l 40 4.619 200 4.618 w 4.6.] m 15 4.696 200 4.683 0 4.686 200 4.677 w 20 4.693 '100 80 3.690 l?n 2 A7l 3.689 240.t 3.686 vs 4u ,.oo, t22O 3.660 s 3.67 s loo 3.708{}t3 3.133 60 3.67? w zu J. oo? 'I 20 3.097 001 3.091 40 3.025 3l0 3.029 0 3.037 260 3.036 10 3.001 040 3.00i W J.UJ 40b 2.967 T01 2.968 20 3.018 320 3.017 m 30 ; .l 40 2 .882 20 ?.994 080 ?.994 m 20 2.988 10 2,981 310 2.982 w 2.98 m tv z.bdu tl.l.l Z.ggo ,n^ , o", ,Iol 2.939 2Q 2,774 3?0 2.775 40b2.940 1II 2.944 s 70 1.J?4 '101 2.924 5 2.691 121? 2.665 031 2.912 m 60 2.931 .201 2.603 15 2.852 ]80 2.852 w ZO ?.850 5' 2.878 ni /uD z.our t2o.l 2.599 10 2,766 340 2.755 m 30 2,769

flt , eqE 'l 'lsb70 2.542 240 2.543 80vb2.5862l l 2.585 vs 00b 2 .591 20 2.854{iio i';;i 2.84 vt 2.507 ni t q17 20 2,740 320 2,739 2.73 15 2.471 330 2.470 60 2.523 280 2.522 vs 50 l0 2.726 nl vf 2.360 150? 2.346 3Ob2.484 231 2.472 s 70b 2.484 'l0b 2.653 121 2.642 50 2.34? 400 2.341 30 2.459 360 2,457 m 30 2.457 '100 2.579 201 2.579 <5 2.212 301 ?.211 30b2.340 400 2.339 s 35 2.343

9C 2.557 201 2,559 ]0 2.185 420 2.184 30b2.320 I.',I0.0 2.320 m 35b 2,315 20 2.516 240 2.516 s 2.51 s <5 2.168 311 2.172 5b 2,299 420 2,295 20 2,?98 'I 0b 2.44S 031 2.446 50 2.151 250 2.152 311 2.199 m 30 2.203 20 2.178 440 2,178 w 40 2.181 rq ) nr) f430 2.0?6 1ob 2.192 301 ?.r9l '060 ,2.10.02.132 m, 2.132 2.019 Jr z'rrrt 3'l 2.128 mJ 30b 5b2.017 460 2.018 25 2.018

10b 2,129 250 2.130 s 2.13 s 5b 1.982 ni 351 2.005 25 2.001 1.974 5 2,117 I4l? ?.099 tE r ora rz4l'l 20 1.996 0.12.0 1.996 m 20 I .989 60 'I1.974 'I5 5b 2.095 141 2.094 ]5 1.9t2 350 . 914 | .951I.12.0 I.952 20 1,947 qh 2 nnn1060 2.000 I .896 '430 2.000 m ?.00 5 1.877 T5l 1.877 20 1.9003.10.0 1.900 30 .zou t.dJ3 '1 c r '6rc | .bJ 20 1.855 260 500 I.87] l0 I.873 t44o l.goo .854 'I E)6 1 1aE .790 1n'- 1"'-"'421 1aa Jeee | . /o 20 r.?88{;tl 5b |.846 520 I.848 30 |.851 1.761 'f5b 1.789 5b I.736 251?1.750 1.746 ni 20 I.785 540 | .786 35 1.787 451 1,744 m 50 1.743 30 1.545 002 |.545 1.699 20b1.740 20 r.6ee{;3f 1.699 5b r.694 560 I .694 20 I .694 '15 'I ^* , ^orrl.14.ol.683 30 I .54t 600 I.539 I.663 t6',l .669 "" 20 I .678 'I """''3.12.01.68r 1.527 ,n^ . oro ,521 q r FoE ,441 .595 5b I.646 471 I.643 m 40b 1.649 ""'" '610 1.527 '261 1.594

1,370 'I0 '| 6h I ?^a 1640 m |.367 I.570 002 .564 5b I.576 531 |.574 25b 1.577 "-""'312 1.367 f^t 'I s 1.54rti;i .543 20b I.556 600 I .567 50 1.556 '|1.542 ]0 1.512 620 'I.51 2 15b1.527 122 | .530 50b I.531 Etl 'I .494 5b I.508 640 I.509 5 I .510 s 1.4e4 tii2 1.494 ((n 'I5 1,482 5b I .4970.16.0 I .497 ] 495 r0 r.481 r;;i 1.48? '| 17t .440 5 I.440 {;ri 'I zob1.474 5.10.0 I.474 35 1.476 'I.440 l0 1,412 541 .4t9 5b I.398 302 1.398 30b r.397 l0 I.387 r640 'rI .388 1 '611 .386 0b | .383 680 I .383 30 I .382

Ir'IntensJties estimatedfrom rotatlon patterns; In = Intensities estimatedfrom t,leissenberg patterns; I, d*uu" from Guinlerfilms, CulNi, Th02lnternal standardid"* measuredfrom rotation and 0-level !.leissenbergfilmsi ni = not indexed

Downloaded from http://pubs.geoscienceworld.org/canmin/article-pdf/19/2/315/4005550/315_19_2cm.pdf by guest on 25 September 2021 320 THE CANADIAN MINERALOGIST

TABLE3. ELECTRON-I.IICROPROBEANALYSIS OF in intensity presumably attributable differ- NODULARNULLAGINITE. SAI4PLE IOOZ5 to ences in atomic number. These variations were l,ltg Pecalc. to l00X Atomtc ili,(0H),C0?-(wrZ) - A B props. too subtle to be picked up by microprobe line- 'r scans; their interpretation therefore remains un- 49.4 46.76 53.96 .93 55.53 l'lg 0.56 0.51 0.59 0.0s clear. Perhaps they represent minor variations Cr 0.23 0,25 0.29 0.0r in the amount of water held by the nullaginite. te 0.15 0.rg 0.l7 <0.01 Cu '1.880.07 0.08 0.09 <0.10 Infrared absorptionanalyses were undertaken st '14.44 0H (calc. ) 14.07 16.24 2'.OO t6.09 in an attempt to obtain independent confirma- C03 (calc.) 22.70 24.84 28.66 t.00 28.38 pres- H20 (?) - 13.34 tion for assumptionsmade on the anions ent in nullaginite.Absorption peakscorrespond- 89.43 100.00 100.00 ing to H:O, OH and COg were recorded, but Eoth A and B are corrected for 8.6U pecora.ite. A assunesthat becauseof impurities of pecoraiteand gasp6ite ihe shortfall ffom 100i1Js due to H20i ln B no excessls assuned, andthe analysls ls recalculatedto-l0O?. Theatomic DroDortlons in the sample, the data did not provide signifi- are taken fron the I recalculatJon. cant additional information.

Optical and physical properties Both nodular and veinlet varieties of nul- laginite were analyzed in five different samples. Nullaginite is bright green. In its nodular The results of all analysesare similar, showing form it has a dnll, clay-like ; in cross- appreciable but variable amounts of silicon. The fibre veinletsit has a silky sheen.In transmitted resultsof one analysis(of the nullaginitenodule light it is pale green and weakly pleochroic, with shown in Fig. 2) are given in Table 3. The absorption slightly greater normal to the fibre analytical results are an averageobtained from axis (c) than parallel to it. Indices of refraction five different spots on the same nodule. are 6g = 1.67, l) - y : 1.78. Extinction The Si is assumedto be due to an impurity; XA,c : 6". Becauseof the extremely fine size as X-ray-diffraction patterns show a line of of crystallites, it was not possible to measure variable intensity at 73 A, one of the strongest the optic angle or to uniquely determine the lines in the X-ray-diffraction pattern of ser- optical orientation.However, as X is closeto c, pentine, it is concluded that the observed Si the 6 crystallographicaxis should correspondto is due to the presenceof a serpentine mineral, YorZ. Electron-microprobeline-scans across the ser- The indices of refraction are appreciably pentine mineral associatedwith nullaginite show lower than those reported for glaukosphaerite that it generallyhas a high Ni content (Fie. 5); (Pryce & Just 1974, Deliens 1975), which is for the purpose of the calculations, it was there- contrary to expectations,since NiO has a higher fore assumed that the serpentine mineral re- specific refractivity than CuO (Mandarino sponsible for the Si is pecoraite, with a compo- 1976), This tends to confirm that nullaginite sition correspondingto that reported by Nickel contains some admixed water or other submi- (1973). On this assumption,the 1.887oSi shown croscopic intergrown impurity. in Table 2 represents 8.6Vo pecoraite, The specific gravity of tiny particles was After the pecoraite correction was made, measured by suspension in Clerici solution. there was still a shortfall of 10.617o. ln one Values obtained ranged between 3.528 and recalculation (A, Table 3), this shortfall was 3.606, with an averageof 3.56. The calculated attributed to admixed water; as is demonstrated specific gravity, using a unit-cell volume of below, optical and physical properties tend to 342.6 Au and the recalculatedcomposition (8, support this assumption. Analysis B assumes Table 3), and assumingZ = 4, is 4.07. How- the mineral to be anhydrous,by analogy with ever, if the specific gravity is corrected for the rosasite and glaukosphaerite. The atomic pro- 13.347o water suggestedby recalculation A, portions are calculatedon a cation sum of 2; then a value of 3.660 is obtained, which is the proportions of OH and COa slightly exceed in reasonable agreement with the measured 2 and 1, respectively,because of the trivalent value. If the excess water were incorporated charge of Cr. into the structural formula, the calculated spe- An electron-microprobeline-scan across the cific gravity would increase, widening the dis- veinlet illustrated in Figure 4 is reproduced in crepancy between measured and. calculated Figure 5. This shows that appreciable Si occurs values. in even the purest part of the nullaginite veinlet. A Gladstone-Dale calculation (Mandarino A high-resolution electron-backscatterimage of 1976) of the mean index of refraction, using the analyzed nodule (Fig. 3) shows variations the measnred specific gravity and the empirical

Downloaded from http://pubs.geoscienceworld.org/canmin/article-pdf/19/2/315/4005550/315_19_2cm.pdf by guest on 25 September 2021 NULLAGINITE, ROSASITE AND GLAUKOSPHAERITE 321

composition corrected for the pecoraite impurity All evidence on nullaginite indicates that the (recalculation A, Table 3), gives fl = L.768. mineral contains nonstructural water not reD- This agrees reasonably well with the mean of resented in the formula Nir(OH)zCOr; the the measuredindices of refraction: [2(1.78) + question ariseswhether the formula should, per- 1.671/3: 1.74. haps. be given as Ni,(OH)rCOs.nHzO. A crys- The hardness of nullaginite was measured tal-structure determination is needed to give the with a Vickers indentor using a 20 g load. The answer. average of five determinations gave VHN Regarding the origin of the nullaginite, in 34.4(9\. its veinlet form it is late in the paragenesisand may even product (Nickel Discttssiott be a supergene et c/. 1979\. It is more difficult to determine the The close similarity of the X-ray-diffraction origin of the nodular variety, but the weight of powder pattern of nullaginite, rosasite, - evidence regarding the other nodular minerals rosasite and glaukosphaerite indicates that they in the deposit suggeststhat they were formed are all members of the same mineral group, as the result of hydrothermal activity; they may which shor.rldbe termed the rosasite group, after have been precipitated from a gel medium the first mineral that was described(Table l). (Nickel et al. 1979). Crystallization in a gel According to presentusage, the formula of the would explain the intergrown silica and the group can be expressedas Mr(OH)rCOr, with water that is assumedto be in excessof the M comprisingCu and Zn, Cu > Zn in rosasite, formula requirements. Zn > Cu in zincrosasite,Cu and Ni in glauko- sphaeriteand Ni alone in nullaginite.The crystal Nrw X-Rev Dnte oN Rosestrn euo structures of minerals of the group have not Gleurospueenrrr (LGB) been worked out, but a structure has been deduced (Giinter & Oswald 1977) for IVIgu Rosasite (OH)COI, the decompositionproduct of arti- New Guinier powder data for rosasitefrom nite, which has a unit cell similar in dimensions, Bakara, Bulgaria, are presented in Table I bt"ttwhich differs in the orientation of symmetry for comparison with nullaginite and glauko- elementsand in the systematicextinctions from sphaerite. TED diagrams from two grains are those of the rosasite group of minerals. [Note similar to those of nullaginite and^are index- that Oswald & Crook (1979)' introduced the ableas &0/with a:9.3. c - 3.15A, B : 90" name "mcguinessite" (without data) for (Mg, 55'l T01. 201 and 301 appear consistently as Cu):(OH)rCOr. The mineral and name had the strongest diffractions. Mirror symmeFy is receivedprior I.M.A. approval'!.1According to distinctly lacking, although TED intensitiesare Giinter & Oswald (1977\, the structure con- unreliable.This information leads to refinement sists of "tMgttt(OH);it(COr) sheets of edge- of the powder data with cell dimensions d : sharing octahedraparallel to 102, corresponding 9.366(3)"b : 12.116(4),c = 3.127(ll A, B to 120 in malachiteand rosasite;adjacent layers : 90.I l(4)'; a few weak diffractions with un- are connected by bridging carbonate groups. certain indicesare presumablydue to impurities. The structure allows only one kind of octa- The refinement identifies a broad line at d : hedral composition, whereas the malachite 2.967 as T0l and l0l, confirming the possible structure, with its two different octahedral sites" space-groupsas P2r/ m or P2r. Rosasitespeci- is incompatible with the topotactic relation of mens from Bakara (Bulgaria), Durango (Mexi- Mgr(OH)rCOr with artinite. If this sructure co) and Rosas (Sardinia) give virtually identi- proves to be the correct one for the rosasite cal powder-diffraction patterns. group of minerals. there would appear to be New electron-microprobe analyses of rosasite little justification to regard glaukosphaeriteas from Bakara, given in Table 4, are compared a separate species.If the Cu end-member is with unpublished analyses obtained by Jiri Just called rosasite"and the Ni end-member,nulla- (pers. comm.) at CSIRO in 1973 on specimens ginite" then the mineral originally describedas from Durango, Mexico (#1789 CSIRO) and rosasite,(Cu,Zn)g(OH)sCOr, would be more from Rosas,Sardinia (British Museum, Natural correctly termed zincian rosasite.Glaukosphae- History, l9Z4 167). There is no distinctive rite, (Cu"Ni),(OH),CO". is simply a nickeloan trend in the ratio of Zn to total metal from the rosasite. centre of a rosetteto the border' Glaukosphaerite *Note acldeclin proof; a full description of mcguin- nessitehas been publishedby Erd et al. (1981). Previotts work: Glaukosphaerite, (Cu,Ni),

Downloaded from http://pubs.geoscienceworld.org/canmin/article-pdf/19/2/315/4005550/315_19_2cm.pdf by guest on 25 September 2021 322 THE CANADIAN MINERALOGIST

TABLE4. ROSASITE,PARTIAL ANALYSES doublet at d : 2.95 and,2.93. This doublet is Bakara,Bulgarla not resolvedin the films of the Kasompi material wt.g electron mJcroprobe chen. lnner 0uter anal.l and appearsas a single broad line. Cu 33,2 32.6 33.6 30.6 31.1 37.50 Weissenbergstudy; lniti^lly, a thin prismatic 7n 24.5 25.8 23.6 ?6.6 25.2 3l .09 crystal of glaukosphaerite from Kasompi was Ni 0.28 0.24 0.20 0.21 0.22 nd mounted for Weisbenbergstudy by M.W. Pryce Co 0.02 0.00 0.00 0.00 0.00 nd of GovernmentChemical Laboratoriesat Perth. FE 0.01 0.00 0.0r 0.00 0.00 nd On a rotation film (Co Ka 57.3 mm diameter) zi matal; 0,42 0.43 0.44 0.45 about the needle axis, the zeroJayer line shows sharp reflections with streaking along powder Durango,llexico, J. Just (!npubljshed) arcs for 2 to 4 mm on either side of the layer wt.3 l3 analyses2 wt.%7 analyses line. The first-layer line, about 20 mm from range mean range nean Cu 32.2 to 35.1 33.7 30.3 to 32.9 31.6 the zero line, shows, within powder arcs, spot 7n 20.A h 22.3 21.6 20.1 to 24.3 22.2 maxima that are more diffr.rsethan zero-layer Zn 0.39 0.40 diffractions; not one of the powder arcs is con- net!l; tinous between the zero- and first-layer pattern, Rosas,Sardinia, J. Just (unpublished) lines. On a zero-levelWeissenberg the wt.U l3 analyses2 reflections are sharp and perpendicularto the rangeo nean axis, btrt drawn out 2-4 mm parallel to the Cu 40.0 to 43.0 41.5 axis, with weaker powder-diffractionsconnect- Zn ll.l to 14.4 12-7 ing the stronger /rkO diffractions for the fr.rll Zn meGl; 0.24 length of the film. The zeroJayer pattern is almost identical to the ftk0level pattern of It'llncheva-Stefanova(1961) 2no apparenttrend in ratjo from first-level Weissenbergfilm dis- centre t0 rJm of rosette malachite.The plays a full pattern of hkl diffractions, all as distinct spots but more diffuse than those of the zero level. A reciprocal-latticeplot of the firstlevel Weissenberg film displays two pat- (OH),CO,, with Cu/Ni - 1.4-1.5. was first terns with similar a{' and b'r parameters.The describedby Pryce & Just (1974) from speci- stronger, more complete pattern displays mm2 mens found at several localities in Western symmetry. but with b'! halved, k odd in hkl Australia from 1967 to 1973. The Australian and even in hkjt ft01 reflections are absent. material did not yield complete single-crystal This leads to an orthorhombic - I lattice with X-ray data but the powder pattern, with the a : 9.5, b = 24.0, c = 3.7 A. ttre second. exceptionof three lines, could be indexed on a weaker patternodisplaying t? symmetry, offset cell similar to that of malachite, except for a along ft01, is readily indexable as two patterns p angle close to 90o. in twin relationship,with (100) as a twin plane. A similar mineralo(Cu,Ni)z(OH)rCOr. with This pattern is consistentwith a coexistingsec- Cu/Ni : 1.73 from Kasompi (southernShaba. ond crystal, having a monoclinic lattice with Zaire) was describedby Deliens (1975). The a = 9.35,b= 11.97,c - 3.13A, B:96" O'. powder data for the minerals from Kambalda i.e.. similar to that of malachite.The distinctive (W.A.) and from Kasompi are very similar. observedreflections, all in the first level, are2Ol Jambor (1976a) suggesteda variation on the s" 2l I w. 2Ol w, 211 vw, 11I w, 021 vw and indexing that still left one line (d - 2.928) 03 I vw: 2ot. with d = 2.851. would coincide unindexed, resulting in a cell closer to that of with d - 2.852 in the powder-diffraction pat- malachite. tern of glaukosphaerite. No other reflections The opportunity to re-examine this problem appear to be sufficiently intense to appear in arose during a sabbatical-leave visit by the the powder pattern. author to CSIRO (Perth, W.A.) in 1976. Ad- A second crystal confirmed the results ob- ditional Australian specimenshad becomeavail- tained from the first crystal and enabled the able, and a specimen of the Kasompi material accurate location of a* and D8 relative to the had kindly been supplied by Michel Deliens at dial aris of the instrument. On a precession the request of Jiri Just. Gandolfi patterns on a camera, zero- and firstlevel films about a and 57.3 mm powder camera with ColFe radia- zero-level films about b. although weak, con- tion, and Guinier focusing-camerafilms (Cu/Ni) firmed the Weissenbergindexing, especially the with ThO, as an internal standard, confirm the extinction conditions ft&/, present only with i * previously publishedpowder data, including the I - 2n and characteristic of the I lattice. These

Downloaded from http://pubs.geoscienceworld.org/canmin/article-pdf/19/2/315/4005550/315_19_2cm.pdf by guest on 25 September 2021 NULLAGINITE, ROSASITE AND GLAUKOSPHAERITE 323

observations clearly indicate the presence of a analyses might confirm this. The higher nickel parallel intergrowth of two phases in the Ka- content of Kambalda glaukosphaerite suggests sompi material. that glaukosphaerite at Kasompi is intimately Electron diffraction: Transmission electfon- intergrown, in a regular fashion, with nickeloan diffraction plates of glaukosphaerite give the malachite. following average cell-dimensionsfrom hOl pat- Clearly, a further study of these materials terns:Kambalda, W,A.: a :9.39, c : 3.08A, by electron microprobe is essential to relate 90' 3Y (8 samples); Kasompi, Zaire: variations in chemistry to the phases present. - a 9.42, c = 3,L2 A, "8" = 90o 5V' 0A samples). These patterns are not all of equal Kolwdzite quality; becauseof poor orientation, some pat- terns are less complete than others. Some of Cobaltiferor.rsmalachite from Shaba (Zaire) the above patterns show distinct 2mm sym- was describedby Deliens et al. (1973). Anal- o'B" metry and angles very close to 9Oo (within yses show a cobalt content ranging from O.25 15'). The "average B" is the average divergence to 18.46 wt. Vo. The X-ray powder patternsare from 90o, since there is no distinction between generally similar to malachite for low cobalt ftOl and E'Ol in an ft01 pattern with 2mrn sym- content (0.25 to 4.22 wt. Vo) and to rosasite metry. for the higher cobalt content (16.05 and 18.46 TED diagrams of two crystal fragments wt. 7o) (Jambor 1976b, p. 1O2). Recently the from Kambalda and three from Kasompi are new name "kolw6zite" was proposed by Deliens zero-level patterns, iStZ) rittr diffractions 26O, & Piret (1980) for the mineral representedby 1ll, T5l, 371,1.11.1, Usz, 2zz ana2.1A.2. the latter two analyses with cobalt contents of These diffractions are compatible with the l- 16.05 and 18.46%. These authors have de- lattice extinction condition, but not all are pres- duced a triclinic cell for indexing the powder ent on X-ray-diffraction films. An average of pattern of kolw6zite. They also apply a similar the measurements from these patterns gave a triclinic indexing to the X-ray pattern of glau- - - 9.3, b = 24.2, c 3.2 A. f.tre orthorhombic kosphaerite from Kasompi. If this interpreta- A lattice, indicated unambiguously by [t0 and tion of riclinic character is correct, there may fttl Weissenbergdiffractions and confirmed by be three phases in the Kasompi glaukosphae- indexing of the zero-level (312), is not con- rite. firmed by h0l diffractograms, since the system- atic extinction of hOl is not observed. In TED AcruownocEMENTs diffractograms, l0l may well have l1l superim- posed on it, especially in this case where d* The authors acknowledge the assistance of [010] is very short (b - 24 A and b'k = 1.33 severalmembers of staff at CSIRO (Perth)' the mm, on the scale of the TED diffractograms cooperation of M.W. Pryce of Government obtained here). Chemical Laboratories (Perth) and of Dr. Syd- Discussion: Single-crystal Weissenberg films ney Hall of the University of Western Australia, of the first layer hkl of the reciprocal lattice and the kindness of Dr. Mincheva-Stefanova of of glaukosphaeritefrom Kasompi, Zaire, clearly the Mineralogical Museum, Sofia, who supplied indicate that two phasesare present, one ortho- a fragment of the original rosasite sample from rhombic A and the other monoclinic (twinned Bakara, Bulgaria. The authors are also grateful on 1O0). This relation is not evidencedin zero- for the valuable comments of the reviewers. layer films, TED plates or powder-diffraction patterns. One powder line with d = 2.852 cor- RsreneNcEs responds to 201 of the monoclinic phase (cal- culated 2.851), but it also correspondsto 180 DalrrNs, M. (I975): La glaucosphaeritede Kasom- in the orthorhombic -l phase. The powder-dif- pi (Shabam6ridional, Zaire). Soc. Frang. Min6- fraction patterns of Kasompi and Kambalda ral. Crist. Bull. 98, 175-178. glaukosphaerite are quite similar. The latter OosrEnnosctt,R. & VrnseEr, T. (1973): shows better resolution, especially at d : 2.940 Les malachitescobaltifbres du Shabam6ridional (111) in Kasompi,2.952(Lll),2.931 (031) Qdire). Soc. Frang. Mindral. Crist. Bull' 96' in Kambalda; three other weak lines fail to ap- 37t-377 . pear in the Kasompi pattern. The cell dimen- & Pnrr, P. ( 1980):La kolw6zite,un hydro- sions of the monoclinic phase, not refined from xycarbonatede cuivre et de cobalt analoguei powder spacings, indicate a similarity to mala- la glaukosphaeriteet i la rosasite.Bull. Mineral. chite, possibly a nickeloan variety; microprobe 103,179-184.

Downloaded from http://pubs.geoscienceworld.org/canmin/article-pdf/19/2/315/4005550/315_19_2cm.pdf by guest on 25 September 2021 324 THE CANADIAN MINERALO6IST

Eno, R.C., Gorr, F,E., Cr,snnoN, F.P. & Cr,em, RoBINsoN, 8.W., DAvIs, C.E.S. & Mec' J.R. ( 1981): Mcguinnessite, a new carbonate DoNALD, R.D. (1977): Otwayite, a new nickel from California, Mineral. Record L2, 143-147, mineral from Western Australia. Amer. Mineral. 62. 999-1002. GiiNran, J.R. & OswALD, H.R. (1977): Crystal (1979): structure of Mgg(OH)zCOs, deduced from the Oswerp, S.G. & Cnoor, W.W., lll Cupro- topotactic thermal decomposition of artinite. .I. hydromagnesiteand cuproartinite, two new,min- Solid State Chem. 21, 2ll-215. eials from Gabbs. Nevada, Amer. Mineral, 64, 886-889. JAMBoR, J.L. (1976a): A possible unit cell for glaukosphaerite, Can. Mineral. t4, 574-576, Pnvcs, M.W. & Juvr, J. (1974): Glaukosphaerite, a new nickel analogue of rosasite. Mineral. Mag. (1976b): Studies of basic copper and zinc 39. 737-743. carbonates. 3. Powder X-ray data for zincian malachite, rosasite and cobalt analogues. Geol. RAMSDELL,L.S. & Worrr, C.W. (1950): The unit Surv. Can. Pap. 76-1C,97-105. cell of malachite. Amer. Mineral. t5, ll9'121. MANDARINo,J.A. (1976): The Cladstone-Dalerela- SrnuNz, H. ( 1959): Tsumeb, seine Erze und tionship. I. Derivation of new constants. Car?. Sekunddrmineralien. insbesondere der neu auf- Mineral. 14, 498-542. geschlossenen zweiten Oxydationzone. Fortsch. Mineral. 37. 87-90. MtNcHeva-SrsFANovA, I. ( 1961): Cuprozincite, rosasite and from the Vracha ore- SwnNsoN, H.E., Coor, M.I., EveNs, E.H. & de' district. Inst. Geol. Bulgarian Acad. Sci. Bull. 9, Gnoor, J.H. (1960): Standard X-ray diffraction 197-207. powder patterns. 10. Data for 40 substances. Nat. Bur. Standards Circ. 539. NrcKEL, E.H. (1973): An occurrence of gasp6ite and pecoraite in the Nullagine region of Western Australia. Mineral Mag. ?9, ll3-115. HALLBERc,J.A. & H.llrtcAN, R. (1979): Unusual nickel mineralisation at Nullagine, West- Received September 1979, revised ntanuscript ac' ern Australia. Ceol. Soc. Aust. l. 26. 6l-71, cepted January 198/,,

Downloaded from http://pubs.geoscienceworld.org/canmin/article-pdf/19/2/315/4005550/315_19_2cm.pdf by guest on 25 September 2021