Canadian Mineralogist Vol. 19,pp.34l-348 (1981) KAWAZULITEBi,Te,,Se, RELATED BISMUTH MINERALS AND SELENIANCOVELLITE FROMTHE NORTHWESTTERRITORIES RICHARD MILLER Departmentol Geology,University ol Alberta,Edmonton, Alberta T6G 2E3 Ansrnecr chalcopyriteassoci6s ont des propri6tdsoptiques et des compositionsnormales. Kawazulite. tellurobismuthite and selenian DIaz- (Fraduit par bleibend covellite are associated with Ditchblende la R6daction) and hematite in a breccia pipe that cuts Aphebian Mot.s-clis;kawazulite. tellurobismuthite. guanaiua- dacitic ignimbrites at Mazenod Lake. Northwesr tite, covelline,Se, Te, tsi, lac Mazenod,province Territories. Electron-microprobe analvses of kawa- de I'Ours, Territoires du Nord-Ouest.duret6. 16- zulite give Bi2.,sTe1.s6Se1.6and Bir.,rTez.nrSeo.oo,with flectivit6.microsonde. minor Cu and Fe. Three unidentified bismuth min- erals (1, B, C'), with the general formula Bi,(S, Se,Te)r, were also found. Anisotropy for kawa- INrnoouctloN zulite is distinct, with polarization colors bluish grey to brownish grey; its measured reflectivity is A small U-Cu deposit occurs at Mazenod 60.0 to 6l.6Vo (547-591 nm). Telh.rrobismuthite Lake, Northwest Territories, on the so-called and mineral A have measured reflectivities of 61.5 Dianne mineral claims of the Noranda Explora- to 62.070 and 51.1 to 5l.6Vo. and micro-indentation tion Company Limited (Long. ll6"55TV, Lat. hardnessvalues (ZIlNm) of 155 to 180 and 199 to 63o45'30"N). A breccia pipe several hundred 205, respectively. This is not only the first reported metresin diameter,defined by tourmalinevein- Canadian occurrence. but also the recorded first lets and a large aeromagneticanomaly, cuts occurr€nce of kawazulite outside Japan. The covel- Iite, which exhibits weak anisotropy. contains up dacitic ignimbrites of the Great Bear plutonic- to 15 wt. 7o Se. Associated digenite and chalco- volcanic suite. These intensely hematized host pyrite have normal compositions and optical prop- rocks are occasionally cut by small ( 1 to 2 mm erties. across) veinlets of hematite-pitchblende that also host the Bi{u-Pb-$-Se-Te mineralization. Keywords; kawazulite. tellurobismuthite. guanaiua- Primary and secondaryhematite tite, covellite, Se, Te, Bi, Mazenod Lake. Bear can be distin- Province, Northwest Territories. hardness. re- guished in the veinlets; primary hematite has a flectivity. microprobe. platy habit, whereas secondary hematite, pro- duced by the oxidation of magnetite,is present Sorulrernr, as equidimensionalgrains with relict magnetite cores. The pipe exhibits extensive superficial Kawazulite, tellurobismuthite et covelline blau- copper-rich oxidation products. bleibend s6l6nifdre sont associ6es e la Ditchblende qui et ir I'h6matite dans une briche en Dipe re- MptttopoLocy coupe des ignimbrites dacitiques d'iee Aph6bien au lac Mazenod (Territoires du Nord-Ouest). La kawa- Reflectivity measurementswere made with a zulite possddela composition Bir.rJil.soser.or€t B2.,, photometer Te2.61Sen.eeavec un peu de Cu et de Fe (d6termina- calibrated to a tungsten carbide tions par microsonde 6lectronique). Trois min6raux standard (Carl Zeissno. 47 42 53). Reflectivity de bismuth. non-identifi6s (A, B et C): de formule is 46.1Vo aI 546 nm and 45.7Vo at 589 nm, id6alis6eBir(S, Se,Te)3, sont aussipr6sents. La kawa- with an accuracy of I I .5Vo. Monochromatic zulite est grise, anisotrope de bleuitre i bru- filters were used to obtain the wavelengths re- nAtre entre nicols crois6s. On mesure une r6flec- quired. The reflectivity values given are the tivit6 de 60.0 it 6l.6Vo (de 547 !r 591 nm). Pour highest of four or more measurements.Freshly la tellurobismuthite et l'espdce l, la r6flectivit6 n polished material was always used, since mineral les valeurs 61.6-62.OVo et 5l.l-5l.6Vo et la duret6 A and tellurobismuthite both showed significant (par miqo-indentation, VHNon). 155-180 et 199- 205. respectivement. C'est la premi€re fois qu'on brown tarnishing after a period of a month. trouve la kawazulite ailleurs qu'au Japon. La coveL Micro-indentation hardness was measured Iine, I6gBrement anisorope. contient iusqu'ir 157o with a Vickers diamond indenter. T\e VHN de Se (en poids). Les cristaux de die6nite et de numbers so obtained have been corrected bv 341 342 reference to a calibration curve, constructed by Bi. A comparison with the analysis of digenite comparing the measured and actual hardnesses suggeststhat this representsabout O.057o S, al- of a range of six standard minerals. This led to though its different atomic weight makes direct a corrected value about l1Vo lower than the ap- comparison uncertain. I suggestthat a detection parent value at VHNI : 100. Iimit of O.lVo is reasonablefor S in a matrix The electron-microprobe analyses all involve of SMOVo Bi, with an analytical accuracy of the energy-dispersiontechnique and were made -+ 0.2 wt. 7o. Similar logic gives slightly higher on an ARL EMX instrument fitted with an values for Pb in a Bi- or $rich matrix. Se, Te, ORTEC energy-dispersion analyzer. The data Fe and Cu are not subject to interference in were processedby the program EDATA2 (Smith this suite, and a detection limit of perhapsO.OSVo & Gold 1979). Count times were 400 seconds with an accuracy of * O.lVo is reasonable.The for both the standards and the samples, with situation for S deteriorates further in the Pb- full-spectrum total counts of 1.5 to 2.5 million. bearing phase, mineral A. The analytical X-ray The acceleratingvoltage was 15 kV in all cases. lines were Kcr for S, Fe and Cu;' La for Se, Ag Two sets of standards were used on differenr and Te; Mp for Pb and Bi. runs. One set comprised silver-bismuth selenide, coball telluride, chalcopyrite, pitchblende and MlNrletocv lead-silica glass. The second set comprised a gold-silver alloy, Bi, Se and Te metals, chal- The bismuth minerals found in this study all copyrite, pitchblende and lead-silica glass. As have the general formula Big(S,Se,Te)gand and Sb might be expected to be present, but a similar optical prop€rties. They were identified check by wavelength-dispersiontechniques did during routine electron-microprobe studies and not reveal them. cannot be distinguished optically except by re- EDATA2 calculates the background continu- flectivity measurements. Kawazulite was first um radiation for the average atomic number of reported and describedby Kato (1970), whose the specimen; experience to date confirms that type material came from the Kawazu mine, this procedure is entirely satisfactory (Smith & Japan. It has not apparently been reported in Gold 1979). Computer plots of the spectraafter the literature since then. removal of the calculated background showed The Canadian kawazulite occurs in trace no significant residuals.The background calcula- amounts, intimately associated with selenian tion is based upon a "normative background", covellite. The anhedral grains do not exceed derived from a diamond sample, which models 25 micrometres in diameter, so that X-ray-dif- precisely the efficiency characteristics of the fraction analysesand micro-indentation-hardness particular detector. measurements are not practicable. Trace The program is set to reject concentrations amounts of a similar phase occur nearby and of less than O.O3Voas being below detection form a very fine intergrowth with pitchblende limits. In practice, elements analyzed on high- and primary hematite. These grains are of the energy lines (i.e., 5 kV and up) under these order of one micrometre in size and may be a conditions have lower detection limits than this. different bismuth mineral. Kawazulite is also In the suite of elements sought, major overlaps found as myrmekitic, submicrometre inclusions occur htween the spectra of S, Pb and Bi, within the covellite; it has either exsolved from ranging from 26 to - lO0Vo. Overlap coeffi- covellite or coprecipitated with it. cients were calculated from the standards; the Tellurobismuthite has coprecipitated with close approach to 100% totals (with the ex- aboit SOVo pitchblende and is associatedwith ception of analyses 7-9) indicates that these some primary hematite. It is anhedral, forming coefficients are correct. However. the detect- grains generally less than 20 pm in diameter. ability of these elements is no longer of the It is also found as interstitial blebs, up to 100 order of O.O3Vo.For example, analysis l, on pm in diameter, between grains of secondary tellurobismuthite, had 301,134 counts in the hematite. There are trace amounts of chalcopy- Bi MB region of analysis.Of these counts,26Vo, rite, bornite and, rarely, covellite, but these are or 78,294, fall in the S region of analysis. The not closely associatedwith the tellurobismuthite. statistical varialion of these counts, i.e., the Mineral I was found in one veinlet as blebs square root (at the 1 a level), is 28O counts. up to l(D pm or more in diameter, interstitial At the 2.58 a or 99Vo confidence level, a peak to primary hematite. Although these blebs are of.722 counts could therefore register in the S large enough for microprobe analysis, they region, owing to Bi, which sets the limit on S nearly all contain exsolved rods of covellite, up detectability in the presence of this amount of to 10 pm long and 0.3 fr.- in diameter. Oc- KAWAZULITE FROM THE NORTHWEST TERRITORIES 343 Flc. -1..A. Mineral I (while) and minor covellite (mediunt grey) in hematite (lighr grey) and silicate (black): x8 objective, photographed under oil immersion (as B, C and D). B. Cential-grain of min- eral I in Figure lA, x40 objective; exsolution rods of covellite are barely resolved at left, and three exsolution lamellae of mineral B (white) are at top. C. Left-hand section of grain of mineral I in Figure 18, xl00 objective; exsolution rods of covellite, in long and cross sections. D. Covellite grain, sarne scale as Figure lC, 100 rr,m from the grain of mineral l, shows myrmekitic exsolution of min- eral l.