REACTIONS in CUBANITE and CHALCOPYRITE 173 Position Cufezssis About 1040"C

Canadian Mineralogist Vol. 14, pp. 172-t8t (1976)

REACTIONSIN CUBANITEAND CHALCOPYRITE

J. E. DUTRTZAC*

AssrR,{cr knowledge of this system is also important to ore The temperatures and enthalpies of the solidus mineralogists and geologistsin that it bears and liquidus reactions in cubanite and chalcopy- directly on the origin and history of commer- rite have been determined by DTA methods. The cially valuable ore deposits. During the past heat of the solid-state transition in cubanite was quarter of a century, many authors have studied 5.6 kcal,/mole; it occurred at 250-300.C when this co,rnplicated system and have succeeded, at the cubanite was heated at 5"C/min. This com_ least in part, in elucidating some of the reactions pound fuses inmngruently at 902"C; the liquidus which occur in it. The important contributions temperature is about l0l5"C. phase reactions in this area have been discussed recentlv were detected in CuFeS, (chalcopyrite) at 560 bv and Barton (1973) and (1973) 650"C. C\rFeg melts at 880"C and the fusion by Cabri and will not apl)€rs to occur over a narrow but finite temper_ be repeated here. Both authors cornmentd on ature range. The addition of CuFerg to CuFeS, the extreme complexity of this system, the study stghtly elevates tle incongruent melting point and of which is rendered difficult by numerous also depresses the t€tragonal-cubic ieaction in phasotransitions, by phaseswith nearly identical CuFeSz. Solidus and liquidus data are also re- chemical compositions, crystal struCtures ported and for the FeS-CuFeS2 pseudobinary system. optical properties and by irreversible reactions and unquenchable phases.Because of these ex- Sovruernp perimental difficulties, rnany details of this phase system renrain uncerlain. La temp6rature et l,enthalpie des r6actions au solidus et au liquidus de la cubanite et de la chal- Chalcopyrite, CuFeSz,and cubanite, CuFe"Ss, copynte ont 6t6 d6termin6es par les m6thodes are prominent phases in the central portion of d'analyse thermique diff6rentielli. La chaleur de the Cu-Fe-S phase system; both compounds, transition i l'6tat solide de la cubanite est 5.6 especially chalcopyrite, are important ores of kcal,/mole; cette transition s€ produit A 250.G gopper. CuFeSzis tetragonal at low temperatures 300"C lorsque la cubanite est chauff6e a 5"C,/ (chalcopyrite) but during heating min. Ce.compos6 fond incongruentement d 902.C; it reacts first la temp€rature au liquidus est approximativement to cubic intermediate solid solution (iss) and pyrite (Barton 1015'C. Les r6actions ont 6t6 d6cel6es dans le 1973) at about 557oC (pankratz CuFe$r (chalcopynte) d 560.C et 650"C. b & King 197O; Conard et al. 7975) and subse- C\.rFeS2fond i 880'C, et la fusion se produit dans quently it fuses in the temperature range g65- un domaine restreint de temp6rature. L,addition 895'C (Kullerud et al. 1965),CuFezSs is orthor- do C\rFerS, au CuFeS, 6l0ve l6gdrement le point hombic at low temperatures (cubanite) incongruente but at ds -fusion et abaisse la temp6riture 2W-210"C it undergoes do la,transformation quadratique-cubique a sluggish and seem- dans le ingly irreversible transformation jss CuFeS,. Irs donn6es pour le to cubic of solidus eC le tiquidus composition sont alssi pr6sent6es pour le systBme pseudo-binai_ CuFe:Sg (Cabri et al. 1"973).T:he re FeSCuFeSg. crystallographic changesinvolved in the CuFe$g transformations are well-doculnented, largely as a resnlt INT'nopucrtor.t of the work of Fleet (1970), of Cibri et al. (1973) and of Szymanski,s (1974) recent Phase relationships in the Cu-Fe-S system refinement of the crystal structure of cubanite 3xe 9f qrea! importance to process metallurgists and high-temperature CuFezSs. Schlegel & involved with copper sulfide ores and conien- Schuller (1952) .measuredliquidus and solidus trates. in order to predict and to optimize the temperafures along the CuFeSz-FeSr.oepseudo- flotation, leaching and thermal -processing of binary which essentially encompassestG com- these commercially significant sulfides. A position CuFerSs. A eutectic-type relationship was noted with a eutectic tempeiature of g96oi: lResegrc! Scientist, ?hysical SciencesLaboratory, C\rFeSswas observed to melf at 950.C, a tem- Canada Centre perature for Mineral aad Energy Techno- _substantiallyhigher than that noted by logy _(CANMET), Department of Eneily, Mines later workers. According to the data of Schlegel and Resources, Ottawa, Canada,K1A OGi. & Scbuller the liquidus temperature for the com- 1.72 REACTIONS IN CUBANITE AND CHALCOPYRITE 173 position CuFezSsis about 1040"C. Signiticantly, weight by difference. All loaded tubes were cen- these investigators did not indicate either the trifuged prior to heating in order to produce a intermediate solid-solution (i,rs) region or the constant degree of sulfide compaction. Samples composition CuFezSgon their phase diagram. were run on a Stone Model GS-2 DTA appara- In the present study the temperatures and heats tus using either an Inconel or palladium support of the various reactions occurring in the CuFeSz- and Pt-Pt:1O7oRhthermocouples. Calcined alu- CuFezSasystem were investigatedby DTA rneth- minum oxide, also vacuum sealedin a silica tube, ods and the results are discussedwith reference served as an inert reference material. The DTA to previous work in the Cu-Fe-S system. Some was calibrated for temperature transitions using data are also presented for the solidus and the 557oC solid-solid transformation in chalco- liquidus reactions occurring along the pseudo- pyrite and the fusion points of gold, silver and binary join CuFeSzfeS. sodium chloride. The measured temperatures were those of the support and any differences between the support and cell temperatures were ExPERIMBNTAL assumedto be compensatedfor by the calibra- Chalcopyrite from Messina, Transvaal was tion procedure. Programmed heating and cooling used forthe bulk of the experi.mentalwork. This rates of 5oClmin were used for all experiments material was ground, sized and subjected to and with such programming rates all the evolved magnetic and heavy-media treatment to pro- sulfur vapors reacted with the condensedphases duce a high-grade chalcopyrite concentrate. during cooling. Condensed sulfur was not ob- Examination of the concentrate by X-ray dif- served on the walls of the silica tubes. The dif' fraction analysisusing a Guinier-de Wolff came- ferential ther,mocouple signal was amplified ra showed it to consist of chalcopyrite with a and then passed to one channel of a two-pen faint trace of some unknown impurity. Cuba" strip-chart recorder. The temperature of the nite ore from the Frood-Stobie mine, Sudbury, sample was passed to the other channel of the Ontario was the starting material for the prepara- recorder; temperatures of the various reactions tion of a cubanite concentrate. The ore was are presumed to be reliable to :t10"C. cnrshed, sized to 17O +325 mesh and then The samplesused for some of the experiments magnetically upgraded. The concentrates were were examinedmicroscopically after having been kept in sealed bottles in a desiccator to minimize mounted in plastic and polished. Various sam- oxidation during storage. Point-counting tech- ples were also subjectedto X-ray 6ilft1sfisn and niques indicated that the concentrate contained microprobe analyses to try to determine the 927o cubanrte with pyrrhotite, chalcopyrite and products formed when the sulfides were heated. pentlandite being the principal impurities. For some experiments a cubanite concentrate rnade Rrsul-ts AND DrscussroN by Dr. L. J. Cabri from the Frood-Stobie ore was also used; this material was prepared by Figure I shows the type of DTA trace obtained tle same general procedure described previous- on heating for natu,ral cubanite and for synthe- ly and also contained about 92/o clubantte. Syn' tic CuFer"Sa.Heating of natural cubanite at 5"C/ thetic bornite, CusFeSe, was used as a DTA min, yields an endothermic transformation be- calibrating material, and synthetic FeS, CuFeS, tween 250 and 300oC. The peak observed at and CuFerSiawere employed in some of the test this temperature was sharp and symmetrical. work. The synthesis and purities of these syn- Cabi et al. (1973) showed this transformation thetic phaseshave been describedby Dutrizac & to be associatedwith the transition of orthor- MacDonald (1973). hombic cubanite to cubic iss of composition For the DTA experiments, the sulfides were CuFerS". This transition is very sluggish and is, mixed in appropriate proportions and then accordingly, higbly dependent on the heating sealed under vacuum in silica tubes to prevent rate. Cabri et al. (1973) were able to detect this oxidation and the escape of sulfur vapor during reaction by holding natural cubanite at 2lOoC heating. The free volume of the cells was ap' for 99 days and, consequently, the equilibrium proxirnately four times that of the melted sul- transition temperature must be at least as low fides. The tubes were 6 mm O.D. and 4 mm as this value. It is not known if the reaction pro- I.D.; they contained a small well at the bottom ceeds at even slower rates at still lower temper- for the placement of the thermocouple. Sample atures. No additionat transformations were ob- weights varied from 10 to 10O mg depending on served in natural CuFe:S" until about 900-905oC the experiment; the exact weigbt was determined when the sulfide melted incongruently. A shoul- by crushing the tube, dissolving the sulfide in der on tle cubanite fusion peak suggests that aqua regia and then calculating the sample the incongruent melting of the concentrale 174 THE CANADIAN MINERALOGIST

HEATll,lGRATE= SoC/MtN, TNCONELBLOCK

U z SYNTHETIC CuFe2S! U E hJ L u l- eruoornenrurc U e

F (r U =o tu NATURAL CUBANITE

200 400 600 800 t@o TEMP("CI * Frc. 1. DTA curves for cubanite and synthetic CuFerSr. occurs oyer a narrow temperature range. The No additional peaks were observed, on heating, cubanite concentrate supplied by Cabri ihowed between90O"C and 1300'C. a more-pronounced shoulder than the other con- Figure 2 shows the type of DTA trace pro- centrate although both were of similar puritv. duced when natural CuFe,Sc was heated and

PROGRAMMINGRATE = 5oClMtN PALLADIUMSUPPORT NATURALCUBANITE

td o z. HEATING (rUJ lrl L lL d lrJ : F ar lrl = tu F

200 400 600 800 tooo t200 1400 TEMP (CC) Fro.2. - DTA curves obtained on heating and cooling for natural C\rFegS, (cubanite). REACTIONS IN CUBANITE AND CHALCOPYRITE r75 cooled at 5oC,/min.These curves were produced using a palladium support without a metallic block and, consequently,considerably more base- line drift was noted than when the Inconel block was employed. The heating cycle shows the cubanite transition at about 260"C as well as the incongruent melting reaction at about 90OoC. Further heating produced no additional peaks. During the cooling cycle, however, a small but sharp peak appeared at about 1010-1020"C; presumably this indicates the liquidus temperature. The incongruent melting reaction is also quite apparent at about 900oC but, on cooling, this peak has been nearly resolved into a doublet. A similar situation prevailed for the synthetic CuFe"S.. Further cooling revealed no additional peaks; that is, the transition at about 260"C is not reversed on cooling at 5'C/min' It is fairly Frc. 3. - Photomicrographof synthetic CuFe2S3 well establishedthat the low-temperature tran- heated to 800'C and then slowly cooled. Dark sition is, for practical pulposes, irreversible on phase is retained iss and light phase is 'ochalco- cooling and this was certainly true in the present pyrite". study. Microscopic examination of fused cubanite or synthetic CuFe:S[ revealed that both materials duced during exsolution of more iron-rich iss consisted principally of a "chalcopyrite-like are still largely unknown. This problem arises phaseo'finely exsolvedfrom the iss matrix; small because the mean exsolution-lamella size de- quantities of pyrrhotite were also evident. X-ray creases drasficaliy as the composition ap- diffraction analysis using a Guinier-de Wolff proaches CuFe:Sa. Accordingly, the exsolution camera showed that the phaseswere apparently features shown in Figure 3 have not been re- "chalcopyrite" and cubic iss of composition ported previously for heated cubanite although CuFerS". such features have been noted for less iron-rich lss by Macl-ean et al. (1,972). Figure 3 shows a photomicrograph of syn- The heat of the orthorhombic-cubic transition thetic CuFezSg which was heated above the rn cubanite is not known. Although the irrever- orthorhombic-cubic transition (800'C) and sub- sible nature of this reaction precludes its deter- sequently cooled at 0.5"C/min. The darker rna- mination by conventional drop calorimetry, the trix material is believed to be cubic is.r of com- heat effect can be estimated by DTA methods. position near CuFerSa although the particle size Ingraham & Marier (1965) have shown that the is too fine to permit this to be determined by area under a DTA peak is proportional to the microprobe analysis.The lighter exsolved phase heat exchanged during the transition and that is opically similar to chalcopyrite; the exsolu- heats of crystallographic transilions can be esti- tion laths are about two microns long and one- mated with fair accuracy by this method with half micron in width. Since the total composition the use of suitable calibration curves. Bornite, of the material shown in Figure 3 is CuFezSr chalcopyrite and cubanite are all iron-copper and because roughly equal amounts of the two sulfides with similar densities and thermal con- phases appear to be present, it follows flom a ductivities; they should, therefore, exchangeheat mass balance consideration that the dark and in a similar fashion. For these reasons, bornite light phases cannot have the compositions and chalcopyrite were selected as the calibrating CuFerSu and CuFeSr, respectively. If the iss materials in the present study. Pankratz & King has a composition near CuFezSg,then the "chal- (197O)reported a heat of 1.430 kcal,imole for copyrite" must have a similar compositron, its the bornite transition at 2I2"C and a heat of tetrigonal structure notwithstanding. If the 2.4O5 kcal/mole for the chalcopyrite reastion "chai=copyriteo'approaohes CuFeSa in composi- at 557"C; Conard et al. (1975) obtained an tion, then the iss composition must be well be- enthalpy of. 2.28 kcal/mole for the chalcopyrite yond the composition lange normally attributed reaction. io it (Cabri ilz:). ettnoagh }dacLean et aI. curve obtained (1972) and Cabri (1973) have clarified the com- Figure 4 shows the calibration the same positions of phases produced by cooling iss of for bornite and chalcopyrite; evidently, effects in both sul- iow-iron concentration, the compositions pro- curye can describe the heat 176 THE CANADIAN MINERALOGIST

o CHALCOPYRITE a BORNITE

a = 2.O z f

E = 6 E

U t e t

o.3 AH (CALI Frc. 4. _. DTA calibration curvo obtained for heating at 5.C/min with Inconel block.

fides. Considerable scatter exists and this is molar heat effect for cubanite is twice that for probably. ca5-"d by the use of sealedglass tubes, chalcopyritq the differences are less when the the precise dimensions of which are somewhat comparison is made on the basis of an equal variable. The peak areas depend on the heat number of atoms. Thus the heat effect for cu- fl-ow to the couple, and differing thicknesses of banite is O.93 kcal/atom and for chalcopyrite it glass in the thermowell or varying degrees of is 0.6 kcal/atom. Szymanski (1,974),Cabi et al. contact will affect the heat flow and hence the (1973), and Fleet (1970) all concluded thar peak the areas which were evaluated using the nro. transition in cubanite is generally similar to the cedure,of Ingraham & Marier (1965). T.n" sphalerite-wurtzite transition in ZnS. The heat tion of "qua_ the calibration curye was: of the ZnS transition at 25"C is 3.2 kcal/mole (arbitrary : *Area units) 4.36M1 (cat) (1) @arin & Knacke 1973); this heat is exchanged The error (98Vo confidence limits) associated by the realignment of the ABCABC structure of with the curve was t4Vo whrch is considered the close-packed planes of sphalerite to the average for this type of DTA measurement. ABAB structure of the close-packed planes in The heats of transition in cubanite were com_ wurtzite. The transition in cubanite requires, puted.by in measuring the area produced by a given addition to the transformation of the ABAB quantity of sulfide, by evaluating the assoiiated packing to the ABCABC packing, the movemenl heat effect, and finally by calculating the molar of L4lz moles of the three moles of metal atoms heat of transition taking into account the 92Vo acrossthe sulfur planes and the subsequentmn- purity of the cubanite concentrate used. This domization of these copper and iron procedure atoms was repeated using 20 different sam_ (Szymanski 1974). lt is felt that substantial ples of cubanite, 1O from the concentrate pro_ amounts of energy would be required to move vided by Dr. Cabri and l0 from tbe conceni^rate the atoms across the sulfur plane but that rela- prepared in the present work. The heat of tran- tively little energy would be required to effecl sition calculated for the former material was the metal randomization. Thus, although the 5.26 kcal/mole with a maximum spread of heat effect observed in cubanite is large, it ap- :t1.10 kcal/mole; the corresponding value for pears generally consistent with the value ex- the latter material was 5.8g kcal/mole with a pected for a sphalerite-wurtzite-type transition maximum spread of t1.44 kcal/mole. The m this compound. averageof the two values is about 5.6 kcal,/.mole. The heat of fusion of C\rFe,Sbis unknown but Barton (1973) remarked that the heat effect this value can also be estimated from the area of. 2.4 kcal/mole reported for the chalcopvrite t'nder the fusion peak. This procedure yielded reaction at 557oC seemedlarge, but that ii'was a heat of fusion of 5 kcal/.mole, a value, which consistentwith phase his diagram. Although the must be considered approximate since the heat REACTIONS IN CUBANITE AND CHALCOPYRITE 177 of a fusion reaction is liberated in a different at 558'C by Conard et al. (L975); the value of mann€r than that of a solid-solid transformation 56O'C found in the present study is in good of the type used to calibrate the DTA. It is pos- agreement with these previous works. sible to improve the measuredvalue to take into The inset in Figure 5 reveals a broad endo- account the differences in salibrating temper- thermic peak at about 650'C; the great width atures and in methods of heat liberation by of this peak relative to its height makes it diffi- multiplying the measured value by the ratio of cult to give an accurate temperature for this the calorimetric heat of fusion of CuFeSr (9.5 transition. Pankratz & King (1970) reported kcal/mole) to the DTA heat of fusion of CuFeSe this weak transition to occur at 657"C and at- {7.4 kcal/nole) measured in the present study tributed it to a second-order transfor,mation in and discussedlater in this paper. When this is the cubic iss of composition CuFeSz.Conard ef done, the heat of fusion of CuFesSgbecomes al. (1974) also noted this transition to occur at 6.3 kcal/mo\e. Although this too must be con- 657"C, but they attributed it to the reaction: sidered an approximate value, it is probably iss * pyrite€iss * S

HEATINGRATE=5"C/MlN INCONEL BLOCK

ENDoTHERMIc T

1[480 565 TEMP(OC)

o 200 400 600

TEMP (OC) Frc. 5. .---.DTA curves for natural chalcopyrite. 178 THE CANADIAN MINERALOGIST

(1975)8 noted a fusion temperatureof 882oC. kcalmole by calorimetric techniques. This Both studies indicated a range of fusion which value is to be preferred to the heat of fusion suggests that chalcopyrite does not melt con- of 7.4 kcal/mole estimated by DTA methods in gruently but that the solidus and liquidus tem- the present study. peratures are close to each other. The present Figure 6 shows the type of DTA trace ob- work agreeswith these previous studies both in 'oaverage" tained when mixtures of natural cubanite and the fusion temperature and in the natural chalcopyrite were heated at S"C/min. fairly broad range of fusion. By contrast, Schle- As the concentration of cubanite is increased, gel & Schuller (1952) reported that CuFeSz the low-temperature transformation in cubanite melted congruently at 950oC. The reason for becomes very clear and this is indicative of this higher value is not known but recent works its large heat. By contrast, the solid-solid re- in fhis area all seem to favour a lower melting action in chalcopyrite becomesmore feeble and point for CuFeSr. appears to occur at lower temperatures with It is possible to esti.matethe heat of fusion increasing concentrations of CuFerSa.The ob- of chalcopyrite by measuring the area under the served peaks for the chalcopyrite reaction be- fusion peak and then using this area and Figure came significantly broader on the addition of 3 to determine the total heat exchanged. This CuFe$k. For concentrations of CuFexSgmuch method is only approximate since the heat of a in excessof 50 wt. Vo, the chalcopyrite reaction fusion process is liberated in a manner differ- was only faintly. discernible against the back- ent than the heat of ,the solid-solid transforma- ground noise. Significantly, there is no exother- tions which were used to calibrate the DTA mic peak for the reaction of CuFeSzand CuFe:S" apparutus. At the t;me of this there work, was to form a single homogeneoussolid solution dur- no published value for the heat of fusion of ing heating; such a heat effect might be ex- CuFeSz and this approximate technique seemed pected to occux when these two materials react. justified to produce a needed value for this heat When the mixtures were heated a second time, effect. Subsequently, (1975) Conard et al. de- the.cubanite peak vanishe4 but the chalcopyrite termined a heat of fusion of chalcopyrite of 9.5 peaks were still readily discernible. The absence of a peak for the reaction of chalcopyrite and *Personal communication from B. R. Conard. R. cubanite, together with the feeble chalcopyrite Sridhar & J, S. Warner, International Nickel Com- peak, may indicate that the reaction of the pany, J. Roy Gordon Research Laboratory. chalcopyrite to iss and pyrite and the reaction

HEATINGRATE'5oC/MlN INCONELBLOCK

T eruoornenrtlrc

60o/ocp -4O"/oCb F

50o/oCp-5Oo/o Cb

300 TEMP'("C)

Frc, 6. - DTA curyes for chalcopyrite-cubanite mixtures expressed in weight percentages. REACTIONS IN CUBANITE AND CHALCOPYRITE 179 of the CuFeSz and CuFerS, occur simultane- r300 ously with the two heat effects essentially an- nulling each other. This or some similar mech- anism needs to be postulated to explain the 1200 absence of these heat effects, especially the chalcopyrite transition which should be discernible at very low chalcopyrite concentra- iloo LtoulD tions. The cubanite transition was detec'table in mixtures containing only 5 wt. 7o cubanite. Figure 7 presents some limited DTA data looo obtained for the pseudobinary CuFoSb-FeS; this summarizes the temperatures as- o Figure o sociated with the tetragonal-cubic reaction, the I eoo incongruent melting temperatures of CuFeS,- CuFezSemixtures as well as the liquidus values F in the C\rFeSr-FeS system. Essentially identical fr eoo values were obtained with CuFeSr-CuFeoSumix- o- - tures as with CuFeSr-FeS components. The in- lrj congruent melting temperature increased slightly F as the bulk composition becarne richer in FeS. 700 For compositions containing ,more than 5O mole Vo FeS, the peak for incongruent melting was split into two separate peaks whose separatiott 600 increased as the composition became richer in FeS. The reason for this splitting was not determined. The liquidus temperatures dropped steadily from a value of 1170oC for a bulk composition of. IOOVo FeS to 9O0'C at about 13 mole Vo FeS. The liquidus and solidus tem' o 20 40 60 80 loo peratures became indistinguishablefor composi- FoS tions containing less than 13 rnole Vo FeS. MOLEPERCENT Schlegel & Schuller (1952) investigated the phase Frc. 7. .-.- Partial phasediagram for the quasi-bin- system C\rFeS:-FeSr.o" (4 experimental deter- ary system CuFeS2-FeS. minations) that should be quite similar to the system CuFeSr-FeSt.oo.These workers reported a eutectic-type system in contrast to the generally the apparent size of the associatedDTA peak. peritectic-type diagram produced in the present It is not known how closely these lower-temper- study. In spite of this difference, however, there ature values approximate the true equilibrium is a general agreement between the two studies temperature. No published data appear to exist for the FeS,-rich portions if the different com- for this reaction except for "pure" CuFeSr, positions of FeS, are taken into account. Higher where the transformation is fairly well docu- liquidus temperatures would be expected with mented. It is not known what happens to this FeSr.or. The eutectic temperature of 896"C reaction at CuFezSaconcentrations >5O wf. % noted by Schlegel & Schuller is in good agree- (29 mole % FeS) becausethis peak becomesso ment with the incongruent melting tenrperature feeble under these conditions it is difficult to of about 90OoC found in the present study. distinguish against the background. Schlegel & Schuller reported that CuFeSz melted Microscopic saamination of CuFeS-CuFerS" congruently at 95OoC and that it formed a eu- mixtures heated to 8O0oC and then slowly tectic with FeS", In these aspects the presont cooled revealed exsolution features similar to study is not even in general agreement. A re- those noted by Maolean et al. (1972) and by in CuFeS, at about 950"C was looked action Cabri (1973). Samples containing more than 50 for both on heating a.nd cooling but none was w| % CuFerS" (29 mole Vo Fe$-71 mole Vo detected. Conard et al. also looked for {1975) C\rFeSr) were characterized by "chalcopyritd' a reaction at this temperature and found none. exsolved from the high-temperature cubic ma' The addition of CuFegSgto CuFeSz seems to trix as noted previously and shown in Figure 3. lower the tetragonal-cubic reaction in the latter The mean size of the exsolution lamellae in- compound as well as substantially diminishing creased by at least 10 times on passing from r80 THE CANADIAN MINERALOGIST

CuFesSsto 50 wt. % CuFuS:-5} wt. Vo CuFeS,. grains and the DTA detects only a minor heat {:nly diffraction analysis using a Guinier-de effect. Wolff camera showed the above mixtures to consist of a chalcopyrite-like phase and cubic iss;_the chalcopyrite-like phase was present in a CoNcrusroNs slight excess at 50 wt. % CuFe,Sr-5O vt. % CuFeSs and this could explain the weak DTA The heat of the orthorhombic-cubic 19ak a!_this composition. Samples containing trans- 25 wt. Vo CuFe,S"-75vtt. Vo CuFeS, (i5.5 mole formation in CuFegSgwas found to be 5.6 - Vo FeS 84.5 mole Zo CuFeS:) consisted of a kcal/mole at 2@"C; the heat of fusion was predominantly chalcopyrite matrix containing estimated at 6 kcal/ nole at the inconeruent some grains which showed chalcopyrite exsolved melting point of %OZoC.The liquidus teinper- from a cubic is,r matrix. X-ray difiraction anal_ ature for this composition was about 1015bC. Solid-solid ysis confirmed the cubic phaje to be present in reactions in CuFeSbwere identified at 560oC only trace amounts at this composition. Mix_ and 650oC. This compound fused tures containing less than 25 wt. Vo CuFetS" at 880'C and the fusion appearld to occur ovel consisted only of chalcopyrite with no residual a narrow temperature range. The addi_ tion iss detectable. This was the basis for the con- of CuFesSsto CuFeSs elevates the incongruent struction of the chalcopyrite (cp) field in Figure melting temperature slightly and de- 7. presses the tetragonal-cubic transformation. The_ CuFeSrFeS pseudobinary Chalcopyrite-cubanite mixtures containing system is gen- _ erally of the peritectic type less than 25 trr4.Vo CuFezSg(15.5 mole Vo FeS _ with an incongruent melting point of about 84.5 mole % CuFeSlg)behave essentiallv like 900oC. chalcopyrite once they have been heated. An qol-rich tetragonal chalcopyrite predominates although some minor exsolution features are AcKNowlEDGEMENTs evident at 25 wI. Vo CuFezSa;the most iron_ rich chalc_opyriteproduced had the approximate The author would like to acknowledge tormula Cuo.nrFer.orSr.oo. the Most of the-srains ao_ assistance o{ Dr. L. J. Cabri throughout this pear to consist only of chalcopyritJ with rio yor!. X-ray diffraction analysis evrdence of .!H: was done retained iss. This iron_rich chalco_ by S. Kaiman and the photograpLs pyrite undergoes were taken the tetragonal-cubic reaction by P. CarriAre. Some of the with the DTA experiments exchange of large amounts of heat. were done by C. Leblanc. Karpenkov et al. (1974) reported a natural iron- g:q_crylc9e1r'ite from ihe Norit,sk regiono USSR; it had the formula Cuo.noFer.rnl.{io.olrS,.oo and produced a chalcopyrite X-ray pattern- This Rnr.BnrNcss compound differs from other mineials reported tl" central portion 1o of the Cu-Fe-S s^ystem BARTN,I. & KNacrc, O. (1973): Thermochemical by having both a one-to-one metal to ;dfur lyo7ertiel of Inorganic Substances. Springer_ ratio as well as a substantial excess of iron Verlag, Berlin. o_ver collper. The composition determined by BenroN, P. B. (1973): Karpenkov et al. (1974) can Solid solutions in the svs_ be considerej to tem Cu-Fe-S. Part consist of approximately g2 I: The Cu_S and CuFe_S mole 7o CuFeSr joins. Econ. Geol. 68, 455-465. 18 mole Vo FeS and must be, according to the pleselt study, near the iron-rich limit per_ CAlRr,- L._J. (7973): New data on phase relations - in the Cu-Fe-S mitted-for chalcopyrite. presumably tne afAi_ system. Econ. Geil. 68, 4B-4t4. tion of more iron would =-_-, causethe compound Harr,, S. R., SzvrvraNsrcr, J, T. & !o lreak down into iron-rich chalcopyrite and Sr:nwanr, J. M. (1973): On the transformation "cubanite". As the composition is' increased of cubanite. Can. Mineral. !2, 33_3g, beyond 25 wt. lo CuFezSa,tle exsolutio; i;"_ Coxenn, B. R., Snmnen, R. & WenNr4 tures predominate and the free grains of J. S. iron_ (l-974): Paper presented at the Confertce rich diminish in nimber. jO ot -chalcopyrire ai Metallurgists of the CIM, Toronto, Olrtario. wt.,Vo AtFerSt only small amounts of o.free,, August. cnarcopyflte are present, and the DTA peak Durnrzeq rs corespondingly weak. The addition J. E. & MecDoNer.o, R. J. C. (1973): of still The synthesis more CuFeuSr results in a of some copper sulfides aud coD- material consisting per sulfosalts in almost entfuely of 500-700 gram quantiti"r, Uit. chalcopyrite_r* Res. Bull, 8, 961-972. "*r"f*O REACTIONS IN CUBANITE AND CHALCOPYRITE 181

products in heated chalco- Fu,Br, M. E. (1970): Refinement of the crystal (1972): Exsolution 1305-I3L7' structure of cubanite and polymorphism of pyite: Cdn. J. Earth Sci. 9, CuFez$. Z. Krist. 132, 267-287. Par.rrnerz, L. B. & KrNc, E. G. (1970): High- temperaturo enthalpies and entropies of cha-lco- INcRAHAM,T. R. & Menrnn, P. (1965): Heats of pyrite and bornite. U.S. Bur. Mines RI-7435' some polymorphic metal sulphate transitions estimated by semi-quantitative differential ther- Scnr,rcsr,, H. & Scnur,r-rn, A. (1952): The equi- mal analysii. Can. Met. Quart. 4(3), 169'176. libria of melting anct crystallization in the Cu- Fe.S system, and their importance !or!e -n1o- G. A., Ruoesnrv' KAr,pBrrrov, A. M., Mrrr,rcov, duction of copper. Freiberger Forsch' 82, 1-18' sKn, N. S., Soror-ove, N. G. & SrusnrrN, N. N, of the (1974): An iron-enriched variety of chalco- SzyMANsK, I. T. (L974a): A refinement Kri"st' l4O' pyite. Zap. Vses. Mineral. Obschchest. ln9' structure of cubanite, C\rFezSs. Z' 601-605. 218-239. (1974b): The crystal structure q! KuLLERuD, G., Brr-r-, P. M. & ENcr,eNo, J. L. -niqn- temperature C\lFezSg. Z, Kri,st. 140, 240-248' (1965): High-pressure differential thermal analysis. Carnegi.eInst. Wash. Yearb. 64' 197'199. Manuscript received May 1975, emended Novent- MecLreN, W. H., CesRr, L. J. & Gu.r-, J. E. ber 1975.