Carwdian Mineralogist Vol. 16, pp. 651-657 (1978)

UYTENBOGAARDTITE,A NEW .GOLDSULFIDE

M. D. BARTON{' Department ol Geological Sciences,Vireinia Polytechnic lttstitute and State University, Blacksburg,Virsinia 24061, U.S-4.

C. KIEF"T NetherlandsOrganization for the Advancernentof Pure Research(ZWO)' Amsterdam' The Netherlands

E. A. J. BURKE Institute lor the Earth Sciences,Free Uniyersity,Amsterdam 11, The Netherlands

I. S. OEN GeoloeicalInstitute, University of Arnsterdam,Amsterdam, The Netherlands

ABsrRAcr genberg), Altai (J.R.S.S.). C'est un min6ral t6trago' ia, ri'zz ou.P41, a 9.76, c 9.784. (Comstock); - D(calc) Uytenbogaardtite, Aga,AuSz,occurs with , a 9.68, c 9.814 Cfambang Sawah); Z 8; (Iambang I-es raies electrum and in specimens from Tambang 8.34 (Comstock), 8.45 Sawah). plus 2.71(10) Sawah, Benkoelen district, Sumatra, Indonesia, the de diffraction-X les lntenses sont (421). compG' Comstock lode, Storey County, Nevada, U.S.A. (203), 2.60(9) (321) et 2.l2LQ) ks par proches de and Smeinogorski (Schlangenberg), Altai, U.S.S.R. sitions obtenues analyse sont quoique de Comstock The is tetragonal, P4r22 or l4r a 9.76, Ag:AuSz, l'ufienbogaardtite jusqu'i (wids) Ces donn6es c 9.784. (Comstocklode), a 9.68, c 9.81A Cfambang contienne 4Vo de cuivre. basse- Sawah), Z - 8; D(calc) - 8.34 g/cm3 (Comstock montrent que I'uytenbogaardtite et la forme En lu- lode), 8.45 g/cma Sawah). The strongest temp6rature de Ag*AuSz sont identiques. Clambang pl6ochroique, de gris- X-ray powder diffraction lines are 2.71(L0) (2O3), midre r6fl6chie, le min6ral est gris-blanc l'air, et d'un 2.60(9) (321), 2.124(7\ (421). Measured composi- blanc d teint6 de brun dans dans l'huile. I*s tions lie close to Ag"AuS:, although uytenbogaard- gris-blanc brunitre au rose brundtre (en certaines longueurs tite from the Comstock lode contatns up to 4 wt. r6flectivit6s mesur6es VoR b 33.2'34.6 (470 Vo Cu. These data indicate that uytenbogaardtite d'ondo donn6es) sont les suivantes: (589 30.5- is identical with synthetic low-temperature AgaAuSs. nm), 30.3-34.6 (546 nm)' 3l.l-35.2 nm), Pleoctrroisrn in reflected light is: grey-white to 33.3 (650 nm). La duret6 au polissage, 7IlN15, est que pour grey-white with a brownish tint (air), brownish grey- approximativement 20, soit un peu moins white to brownish pink (oil). Measured reflectivitier l'acanthite. Ce min6ral est d6di6 au Professeur (RVo at specific wavelengths) are: 33.2-34.6 Willem Uytenbogaardt, de l'Universit6 Technique (470 nm), 30.3-34.6 (546 nm), 31.1-35.2 (589 nm), do Delft (Pays Bas). 30.5-33.3 (650 nm). The polishing hardness is (Traduit Par la R6daction) sliehtb below that for acanthite; ZI[N15 is about 20. The name is in honor of Willem Uytenbogaardt, INt'noPucrtoN Professor of geology at tie Technical University, Delft. The Netherlands. Two ternary comPounds, AgsAuS, and AgAuS, are kngwn at low temperatures in the Ag-Au-S. Neither phase has been re- SorvnvrernB system ported to occur in nature, although they might gold-silver deposits. Ex- On trouve l'uytenbogaardtite, AggAuSr, accompa- be expected in certain (Ag'$ gn6e d'acanthite, 6lectrum et quartz, dans des 6chan- amination of several acanthite + elec- tillons de Tambang Sawah, disrict de Benkoelen, trum (Au,Ag) assemblages has led to the dis- Sumatra (Indon6sie), dans la veine Comstock, comt6 covery of natural AgaAuSz, uytenbogaardtite, de Storey, Nevada (E.U), et b Smeinogorsk (Schlan- from the Comstock lode, Storey Counfy, Ne- vada, U.S.A. and Smeinogorski (Schlangenberg), of *Current address: Department of the Geophysical Altai, U.S.S.R. Simultaneously, in the course Sciences, The University of Chicago, 5734 South an investigation of the gold-silver ores from Ellis Avenue, Chicago, Illinois 60637, U.S.A. the Tertiary yolcanic belt of Sumatra, Indonesia, 651 652 THE CANADIAN MINERALOGIST

uytenbogaardtite was found in a sample from TABLE2. REPRESENTATIVEmntVSfS Or UYTET'|gOOAARDTIru* Tambang Sawah, Benkoelen district. At all three localities, uytenbogaardtite occurs as small blebs Cu 5 ToTAL up to 100 pm across and as rims intimately (Jdeal associatedwith acanthite, electrgm, quartz and Aq-AuS^'JZ ) 55.34 10.97 100.00 other . Comstocklode 56.l 20R 2.2 11.2 99.3 (pronounced y0feNnO.Cen- 55.8 27.3 3.0 12.3 98.4 The name 56.5 27.4 J-b 12.4 99.9 DAIT) is for Willem Uytenbogaardt, Professor Smeinogorski 56.0 Ja.+ I 0.4 98.8 n t7 I 32.7 10.3 r00.1 of mineralogy and petrology at the Free Uni- . 54.6 35.3 I0.8 100.7 versity in Amsterdam TambangSawah 54.5 33.4 I I .3 99.2 from 1960 to 1976, dedi- 53.2 35.1 lI.7 100.0 cated promoter of guantitative ore microscopy 56.7 32.6 ll.0 100.3 in his function as chairman of the I.M.A. Com- in pef mission on Ore Microscopyfrom 1962 to 1970, weight cent and now Professor of geology at the Technical lode material develops a completely different University in Delft, The Netherlands. surface in several tens of secondsunder moder- Both the mineral and the name have been ate illumination. This surface suggestsphoto- approved by the Commission on New Minerals induced decomposition;it requires a fresh polish and Mineral Names of the International Min- for its removal. Tambang Sawah and Smeino- eralogical Association. Type material will be gorski uytenbogaardtites, in contrast, seem to preserved in the collections of the Free Univer- be relativelv stable. sity in Amsterdam, the University of Amsterdam Microindentation hardness(VHN), measured and the United States National Museum with a Leitz Durimet-Pol hardnesstester apply- (Comstock IN.M.N.H. 105328 lode) and ing the minimum load of 15 e and an indenta- N.M.N.H. B239 (Smeinogorski)1. tion time of 15 seconds,is about 20, which corresponds to a Mohs hardness of about 2. Becauseof the small grain-sizeand the low hard- Prtystcer- eNn Oprrcel PRopERTIEs ness only one indentation could be made. The polishins hardness is lower than that for acan- In polished section uytenbogaardtite shows thite. Uytenbogaardtite is considerably more weak to distinct reflection . In air the color varies from grey-white to grey-white with a brownish tint. In oil imrnersion the pleochroism is more distincl brownish grey- white to brownish pint; in contrast, acanthite is distinctly blue. The anisotropy is strong in O AgsAuSa both air and in oil, but without distinct coior.. The reflectance is slightly higher than that for A Comstocklodo acanthite and was measured with a Leitz MpV microscope photometer at four standard wave- I TombongSoroh lengths. The instrument was equipped with a Knott 9592A photomultiplier tube, a Veril I Snelnogonhl 8-200 continuous-band interference filter and a l6:L objective of O.40 N.A. A tungsten car- bide, WC-6, provided by Messrs. Carl Zeiss (Oberdochen), was used as a standard. The reflectance values, measured on a grain with maximum bireflectance, are given in Table 1. / 40 Light (i.e., heat) etching is extremely strong, even stronger than for acanthite.The Comstock- 'I. TABLE REFLECTANCEVALUES FOR UYTENBOGAARDTITE* moloZ S Ennx xmin 470 nm 34.6% 33.2/" 546 34.6 30.3 mob % Au 589 35.2 3t.l 650 33.3 30.5 Frc. 1. Selectedanalyses from the Comstock lode, * Tambang Sawah, and Smeinogorski specimens measuredon material from TambangSawah plotted in tle triangular diagram Ag-Au-S. UYTENBOGAARDTITE. A SILVER-GOLD SULFIDE 653

On probing with a sharp- TABLE3. X-MY P0l,,lDERDIFFRACTION DATA FOR brittle than acanthite. UYTENBoGMRDTITEANDSYNTHETIC AglAus2 ened steel needle uytenbogaardtite broke up into small pieces, whereas associatedacanthite Synthetic Ag3AuS2* Uytenbogaardtite tended to smear. Synthetic AgsAuS, is reason- dca.lc I AAT -calc AdT -calc ' ably easy to grind to a powder. hkl dobs "obs "obs I I 0 6.98 6.894 B 6.94 6.845 4 210 4.38 4.360 6 4.33 4.329 3 4.40 4.365 220 3.46 3.447 2 3.47 3.422 | CnsMlcel CovtPosrrroN 300 3.2+ 3.250 3 3-.24 3.253 310 3.09 3.083 6 :.0: 3.061 ; 3.09 3.086 222 2.809 2.825 6 2.802 2.807 3 2.82 2.8r9 203 2.731 2.724 10 2,712 2.710 10 2.71 2.711t0 Electron-microprobe analyses were obtained 321 2.609 2.609 9 2.591 2.590 9 2.60 2..609 331 2.236 2.238 3 2.226 2.222 I independently in three laboratories. Despite tle 421 2.124 2.129 8 2.112 2.114 4 2.13 2.r30; use of different operating conditions, different 422 1.992 I .994 5 1.980 1.980 2 : 520 t .807 1.811 5 r 7aa l 798 l r.ar 1.812 standards and different correction schemes,the 521 1.780 1.781 4 1.768 s.7s g.oa r results obtained are comparable. " lil from the three localities a (A) 9.85 9.81 9.78 Analyses of'that material demonstrate uytenbogaardtite is close to densities 8.29 glw3 B.45 g/cn3 8.34g/m3 ideal AggAuSa. Neither the Ag/Au nor the 'Powder data and cell edgesfrcn Graf (1968)r only those data (Ag + Au)/S ratios (atomic) show any sys- that correspondto observedlines ln the natural material are tematic departure from the expected values of given here. Graf's intensities are ten tires those given here. 3/1 and 2/1. Analysesselected from the more cone grease.In both cases'owing to small grain than 70 done are given in Table 2 and plotted size and intimate intergrowth with acanthite or in Figure 1. The only additional elements de- electrum (or both), a1l patterns contained lines tected occur in the Comstocklode material, of acanthite or electrum. The small amounts of where up to 4 wt. Eo coppet is present along material used vielded rather diffuse diffraction with traces of selenium and tellurium. lines. For dnxr, Eteartetthan 3A, background In spite of inaccuracies introduced by the from the rubber solution was significant. In small grain-size and instability under the elec- neither case could a correction for film shrink- tron beam, possibly due to diffusion processes age be made because of the absence of back similar to those described by Rucklidge & reflections. Stumpfl (1968) for petzite, AgrAuTer, analyses Messien et al. (1966) reported low-tempera- were reproducible. During long exposures of ture Ag'AuSr to be cubic, a 9.724, but Graf a single spot to the beam a detectable shift of (1968) later demonstrated that the symmetry is the sold to silver ratio occurs in which the ex- actually tetragonal. Observed reflections show cited area of sulfide becomesmore silver-rich. that the space group is primitive and that the only possible non-unit translational symmetry elements are a 4r- or a 4raxis. As there are no CnvsrenocnePnY observed reflections of the tvpe 00/ it is not possible to ascertain if a four-fold screw axis is Owing to the small grain-size and the multi- present; however, as Messien et al. (1966) in' granular nature of uytenbogaardtite, neither dicate the presence of a 4t axis, the probable powder smear-mountsnor single-crystalmethods spacegroup is either P4r22 or P4r. could be used.X-ray powder data obtainedusing a Straumanis-typeDebye-Scherrer camera (Tam- bang Sawah specimen; Mn-filtered FeKcu radia- OccURRENcEANo PenecsNssts tion) and a Gandolfi camera (Comstock lode specimen; Ni-filtered CuKa radiation) indicate Uvtenbogaardtite occurs with acanthite, elec- that the new mineral is identical with the tetra- trum and q:uafiz at the three localities, all hy- gonal low-temperature form of synthetic drothermal precious-metal deposits. The Tam- As,AuS, (Graf 1968; PDF 20-461). The X-ray bang Sawah and Comstock lode deposits closely data and cell edges(calculated by least-squares resemble each other and the late-stage uyten- methods) are given in Table 3. bogaardtite paragenesesare similar. The Smeino- For the Debye-Scherrer camera exposure, gorski deposit is of a different type; there, uytenbogaardtitewas powderedunder the micro- uytenbogaardtitemay be supergene.All three scope with a tungsten carbide microdrill and assemblages,however, contain typical low-tem- taken into droplets of rubber solution. For the perature gold-silver parageneses. Gandolfi camera exposure, it was gouged from Hydrothermally-altered Tertiarv andesitic a polished section with a sharpenedsteel needle rocks are the hosts for the gold- and silver- and picked up on a glass fibre coated with sili- bearing quartz veins of Tambang Sawah and 654 THE CANADIAN MINERALOGIST the Comstocklode. Hypogeneminerals common bearing sulfosalts (besides ) are re- to both deposits include quartz, chalcedony, ported only from the Comstock lode. The oxida- kaolinite, , rhodochrosite,adularia, sphal- tion zone at Tambang Sawah consistsof a mass erite,galena, acanthite, pyrargyrite and electrum; of quartz and manganeseoxides, whereasthe opal, inesite and rhodonite occur only at Tam- cementation zone is enriched in acanthite. A bang Sawah. and pyrite, chalcopyrite and Ag- variety of supergenesulfides and sulfosalts are

\ -i. ,.;d'.it' t, ffil:

ffi- UYTENBOGAARDTITE. A SILVER-GOI.D SULFIDE 655 reported from the Comstock lode (Bastin 1922, Massive acanthite fills the interstices between van Bemmelen 1949, Bonham 1969). euhedral quartz crystals (about 1 mm) at one The uytenbogaardtite-bearing samples from end of the Comstocklode specimen(N M.N'H' Tambang Sawah and the Comstosk lode are 105328). The remainder of the specimen is similar in texture and paragenesis.The sample composed of an ejquigranular aggregate of from Tambang Sawah is composedof an inequi- quartz with minor acanthite. granular quartz aggregateenclosing small vugs The textural relationshipsindiqate that quartz, partly filled either with manganeseoxides and the earliest mineral, is followed by sphalerite, a chloritic mineral or with acanthite and a chalcopyrite, pearceite, electrum and acanthite. chloritic mineral. Locally there is brecciation The acanthite enclosesirregular massesof chal- and recementation with chloritic-sericitic ma- copyrite, sphalerite, pearceite and electrum- terial. The sample shows the usual black coating Covellite replaces some of the chalcopyrite, of manganese-bearingrocks exposed at the whereas the electrum blebs are almost univer- surface, but not the expectedveining with man- sally surrounded by 5-10 g,m rims of uytenbo- ganese oxides or acanthite. gaardtite (Fig. 3). No other phasesare observed Textural relations in the acanthite-filled Yugs in contact with the uytenbogaardtite apart from suggest an overlapping order of deposition of acanthite and electrum. The mole fraction of quattz, galena, sphalerite, acanthite, electrum silver in the electrum is about 0.45 adjacent to and uvtenbogaardtite. There is a gradation in- the uvtenbogaardtite and is up to 0.7 in the ward from anhedral quartz to prismatic qtattz centres of the larger grains. The acanthite con- crystals that protrude into the vugs. Inclusions tains about 0.5 wt. Vo as well as traces of acanthite in the prismatic quartz crystals of selenium. demonstrate the overlapping deposition of The depositsof the Smeinogorski area consist acanthite and vug-filling quartz. The acanthite of veins in hydrothermally altered, sfiongly aggregatesin the vugs encloseand partly replace folded slates, sandstones and carbonates of a few corroded and veined grains of galena and Devonian and Carboniferous age that have been sphalerite. Electrum generally occurs as very intruded by.Hercynian-age granites. Most of the fine grains, partly interstitial to and partly en- rich precious-metal ore is thought to be super- closed in the acanthite grains. Occasionally, ag- sene (Emmons 1937). gregates of electrum form the central filling of The uytenbogaardtitespecimen from Smeino' a vug and enclose corroded and rounded grains gorski (N.M.N.H.8239) is a yellow-stainedag- of acanthite. The acanthite adjoining such aggregatesof late electrurn shows replacement by irregular, locally vein-like spots of uvtenbo- gaardtite less than 100 pm across (Fig. 2a, b). In places the acanthite has apparently recrys- tallized into a fine-grained aggregate of acan- thite with interstitial uytenbogaardtite grains (Fig. 2c). The dispersed fine-grained electrum is locally associatedwith a few small grains of uytenbogaardtite in the surrounding acanthite. Microprobe analyses indicate that the silver content of the electrum (about 43 wL 7o) is similar to that in the hypogene electrum from the Salida and Mangani deposits in the same Au-Ag ore province (Kieft & Oen 1973,1977). Manganese oxides appear only where acan- thite is almost completely replaced. Other vugs are filled with chloritic material or aggregates of severalmanganese oxides, or both, commonly with concentric banded textures. Electron micro- probe analvses indicate that the manganese oxides enclosing electrum contain between 4 Frc. 3. Uytenbogaardtite rims (grey with black and 20 wt. 7o silver. The dissolution of acan- spots) around electrum (white) and enclosed by thite and replacementby a chloritic mineral and acanthite (uniform grey). The spotting of the manganese oxides is presumably related to a uytenbogaardtite is due to the extremely rapid late change in the physico-chemical environ- and pronounced photo-etching. The bar is 20 ment. pm long. 656 THE CANADIAN MINERALOGIST

log fr" -to.-

Ag T"c 200 300

tooolr Fro. 5. Log fugacity of versas1000/T ("K) plot (Barton 1978) showing stability of the metal- saturated phases in the system Ag-Au-S. Ab- breviations: low-temperature AgAuS m, uytei- bogaardtite a, (Ag,Au):S solid solution I, acan- thite a, solid sulfur S, liquid sulfur S, silver Ag. The dashedlines are isoplethsshowing mole frac- tion Ag in coexistingelectrum. Frc. 4. Intergrowth of acanthiteand electrum having a bulk Ag:Au ratio of 3:l (expanded-beammi- croprobeanalysis). Enclosing material is electrum eguilibrium during cooling of a high-tempera- (white).The bar is 10 pm long. Oil immersion. ture, gold-bearing argentite * electrum assem- blage. On cooling the argentite would have ex- gregate of anhedral to euhedral quartz crystals, solved uytenbogaardtite leaving the original sil- partly coated on one Surfaceby asanthite, elec- ver-rich electrum with acanthite * uytenbo- trum, chlorargyrite and minor naumannite. The saardtite. acanthite, naumannite and chlorargyrite sur- The variation in AglAu ratio in the electrum round massesof electrum up to 1 cm in max- from the Comstock lode suggeststhat the uyten- imum dimension. Euhedral quartz crystals are bogaardtite is breaking down to acanthite * interspersed.Uvtenbogaardtite occurs in contact gold-rich electrum or that it is reacting with with the outer edge of the electrum and as in- silver-rich electrum to yield assemblagesmore clusions in the electrum. Extremely fine-grained compatible with the phase diagram. Although intergrowths of acanthite and electrum having the occurrence of uytenbogaardtite as rims a bulk AglAu ratio of 3/1 (determined by ex- around the electrum blebs in the Comstock-lode panded-beam electron-microprobe analysis) are specimenmight suggestthat acanthite and elec- interpreted to be breakdown products of uyten- trum reacted to give uytenbogaardtite, this is bogaardtite @ig. ). The presenceof chloragyrite unlikely given the electrum composition(s) and supports a supergeneorigin for at least part of the experimentally determined phase relations. the observed assemblage. Barton (1978) has estimatedthe standard free energy of formation of low-temperatureAgsAuSz as a linear funstion of temperature: AG" = oK). Stenrr-ttY 4O75O + 22.OT (calories;T - The es- timated error is -f 1500 calories. Experimental work on the system AgaAuSz- Ag,S by Graf (1968) iudicates that only below 113'C can AgaAuSa (uytenbogaardtite) coexist AcrNowLeocEMENTs with AgzS (acanthite). Low AggAuS, is stable with gold-bearing argentite solid solution from The authors are greatly indebted to Mr. H. A. 113" to 185oCwhere it invertsto high AgaAuSz, van Egmond of the Free University of Amster- part of the solid solution. Figure 5 (Barton dam for the painstaking and inventive polishing 1978) shows some of the electrum compositions of the new mineral, and for devising the tung- coexisting with uytenbogaardtite and acanthite. sten carbide microdrill. The help of Dr. A. H. Observed electrum compositions are not com- van der Veen of Billiton Research Co. of Arn- patible with these experimentally determined hem" and of Mr. R. H. Verschure of the Labo- tielines. One possible explanation is that the ratory of Isotope Geology at Amsterdam with uytenbogaardtite-bearingassemblages reflect dis- the X-ray work is acknowledged. M.D.B. ac- UYTENBOGAARDTITE. A SILYER-GOLD SULFIDE 67 knowledgesthe assistanceof Mr. L. B. Wiggins veN BeMrvrsLE\ R. W. (1949): The Geology ol of the U.S. Geological Survey and Mr. S. H. Indonesia. fr. Economic Geology, Government Loh of V.P.I. & S.U. with the microprobe Printing Office, The Hague. analyses,and the helpful comments of Dr. J. R. BoNHAM, H. F. (1969): Geology and mineral de' Craig of V.P.I. & S.U., Dr. J. R. Goldsmith of posits of Washoe and Storey Counties, Nevada. the University of Chicago and Drs. P. B. Barton, Nev. Bur. Mines Bull, 20. J. R. Clark and G. K. Czamanske of the U.S. EMMoNs, W. H. (1937): Gold Deposits ol the Geological Survey. Mr. J. S. White of the U.S. World. McGraw-Hill, New York. National Museum provided the material from Gnen, R. B. (1968): The system AggAuSg-AgaS. the Comstock lode and Smeinogorski.Dr. A. T. Amer. Mineral. 53, 496-500. Anderson of the University of Chicago assisted Krrrt, C. & OsN, I. S. (1973): Ore minerals in the with some of the photography. Microprobe fa- telluride-bearing gold-silver ores of Salida, In- cilities were provided by the U.S. Geological donesia. with special reference to the distribution Survey (Reston, Virginia), V.PJ. & S.U., and of selenium. Mineral. Deposita 8,312-320. WACOM, a working group for analfiical geo- & _ (L977): Ore mineral parageneses chemistry subsidized by the Netherlands Or- in Mn-Sn-Ag-Au-Se-bearing veins of Mangani, ganization for the Advancement of Pure Re- Sumatra. Indonesia. /a Problems of Ore Depos! (Varna 2, 295'302. search. M.D.B. was partially supported by an tion,4th IAGOD Symp. 1974) NSF graduate fellowship and by a McCormick MsssrEN, P., Barwrn, M. & Tevrr.mBn, B. (1966): fellowship ([Jniversity of Chicago). Structure cristalline du sulfure mixte d'argent et d'or. Soc. Roy. Sci. Lidee Bull. 56,727-733, (1968): RennnnNcss RucKLTDGE,I. & Srtnvrpn", E. F. Changes in the composition of petzite (AgsAuTez) during BARroN, M. D. (1978): The Ag-Au-S System. analysis by electron probe. Naaes lahrb, Mineral. M.S. thesis,Virginia PolytechnieInst. and State Monatsh., 6I-68, Univ., Blacksburg,Va. BAsrtr{, E. S. (1922): Bonanza ores of the Com- stock lode, Virginia City, Nevada. U.S. Geol. Received May 1978; revised manuscript accepted Surv, Bull. 785,41-63. tuly 1978,