Gold recovery from a refractory pyrrhotite ore

Ž.Int.J.Miner.Process.602000247–262www.elsevier.nl r locate r ijminproGold recovery from a refractory pyrrhotite oreby biooxidation`S.Ubaldini a,),F.Veglio b,F.Beolchini b,L.Toro c,C.Abbruzzese aa CNR,Institute of Mineral Processing,Via Bologna7,00138Rome,Italyb Department of Chemistry,Chemical Engineering and Materials,UniÕersity of L’Aquila,67040Monteluco di Roio,L’Aquila,Italyc Department of Chemistry,UniÕersity of La Sapienza,00185Rome,ItalyReceived30November1998;accepted12May2000AbstractThe aim of the present investigation was to study the biooxidation of a refractory gold-bearing pyrrhotite,in order to increase the gold recovery during the subsequent conventional cyanidation.Bacterial cultures utilised in the biological test consisted predominantly of Thiobacillus genus. Tests were conducted at laboratory scale.The gold content of the ore sample,coming from Bolivia,was of10g t y1Au.Ž.After24h leaching time by direct cyanidation,low gold recovery was obtained-20%Au, with a high reagent consumption.On the other hand,a high gold recovery was achieved for the biooxidated samples:after24h cyanidation gold dissolution reached about91%Au.Experimental results have shown the technical feasibility of the biooxidative pretreatment prior to conventional leaching and a complete circuit of treatment,on laboratory scale,has been developed considering also the subsequent gold recovery by carbon adsorption r desorption and electrowinning.A gold extraction yield of about86%was determined in the whole process for gold extractionŽfrom pyrrhotite biooxidation,solid–liquid separation,cyanidation,adsorption,desorption,elec-.trowinning.q2000Elsevier Science B.V.All rights reserved.Keywords:biooxidation;pyrrhotite;gold recovery;process flowsheet;Thiobacillus)Corresponding author.0301-7516r00r$-see front matter q2000Elsevier Science B.V.All rights reserved.Ž.PII:S0301-75160000019-3()248S.Ubaldini et al.r Int.J.Miner.Process.602000247–2621.IntroductionCyanidation is a relatively simple and cheap technology for treating a wide range of Ž.gold-containing ores Young,1987.However,sulphide ores often contain very finely disseminated gold and other precious metal particles,which rise to the refractory nature of such ores to current extraction methods.As a result,these fractions are frequently lost Ž.in the tailings Abbruzzese et al.,1994;Paponetti et al.,1991.Moreover,the formation of some intermediate species,such as ferrous iron,sulphide ion,thiosulphates and arsenates,can consume vital oxygen for gold dissolution in cyanide solutions.Further-Ž.`more,these species tend to precipitate the gold already oxidised Veglio et al.,1995a, while the decomposition of pyrrhotite forms compounds such as thiocyanide or ferro-Žcyanide that can consume free cyanide during the recovery of gold Ubaldini et al., .1995.The choice of process for refractory ores or concentrate treatment is based on many technical,economic,and environmental factors.There are several ways for effective oxidative destruction of the sulphide matrix:oxidation roasting,aqueous pressureŽoxidation,bacterial oxidation and chemical oxidation Lawrence,1990;McNulty and.`Thompson,1990;Veglio et al.,1993b;Ubaldini et al.,1994.Traditional methods of pretreating refractory ores,such as roasting,are no longer satisfactory.This can be due to economic or environmental considerations,or both.With the advent of stricter environmental regulations designed to restrict the quantities and control the quality of wastes from mining,milling,and smelting operations,someŽestablished techniques no longer have a technical and r or economic advantage Bos,. 1993;Lawrence,1990;Lawrence and Poulin,1995;McNulty and Thompson,1990.Relatively new technologies for refractory ore pretreatment are being developed to overcome these shortcomings,notably including biological oxidation systems.The biological pretreatment of refractory ores containing precious metals has the obvious advantage of not requiring high-temperature or high-pressure equipment and showingŽthe lowest capital requirement Crundwell,1995;Lawrence,1990;Lawrence and Poulin,.`1995;Veglio et al.,1993a.Biological leaching,which for many years has been applied for the recovery ofŽ.copper and uranium,has only been recently in the last10years considered for the treatment of refractory ores and concentrates containing precious metals,such as gold Ž.Escobar,1995.In these cases bioleaching before cyanide leaching increases the goldŽ.extraction from32%to95%Attia and El-Zeki,1989.For new technologies such as this,novelty is an additional factor which,even if technical and economic parity or superiority with other technologies can be demon-Ž. strated,can weigh against its selection as the process of choice Lawrence,1990. Bioleaching of sulphurous minerals for gold recovery is practiced commercially in South Ž.Ž.Africa Van Aswegen et al.,1991and the USA Kosich,1991,while it is applied onŽ.Ž. pilot plant scale in Australia Spencer et al.,1991,Canada Hackle and Wright,1989Ž.and the USA Moore,1989.Microbial metal leaching or bioleaching is based on the ability of a special group of acidophilic bacteria to oxidise sulphuric minerals.They convert these extremely insolu-ble compounds into water-soluble products.Their activity results in metal mobilisation()S.Ubaldini et al.r Int.J.Miner.Process.602000247–262249Ž.and sulphuric acid production Boon et al.,1995.The most used microorganisms are the chemolithotrophic,acidophilic bacteria Thiobacillus ferrooxidans as well as T. thiooxidans.They usually live at a pH range from1.5to4.5and oxidise reduced form of Ž.Ž. iron only T.ferrooxidans and sulphur both T.ferrooxidans and T.thiooxidans as Ž.energy source Grishin et al.,1988.Bacteria catalyse the oxidation of the sulphurousŽmatrix,letting to the subsequent dissolution of gold by cyanide leaching Abbruzzese et.`al.,1994;Veglio et al.,1993a.In particular,the biooxidation of pyrrhotite involvesŽseveral steps of chemical and biological oxidation Pinka,1991;Ubaldini et al.,1995;.`Veglio et al.,1995a.The process starts by a chemical step:FeS q2H q™Fe2q q H S.1Ž.2In the subsequent stage,bacteria oxidise pyrrhotite according to the following reactions:2FeS q4.5O q3H q™2Fe3q q SO2y q HSO y q H O2Ž.24422FeS q1.5O q6H q™2Fe3q q2S8q3H O.3Ž.22Sulphur is also produced by H S decomposition:2H S q2Fe3q™2Fe2q q2H q q S8.4Ž.2ŽA chemical reaction between pyrrhotite and ferric iron takes place Beolchini and.´Veglio,in press:FeS q2Fe3q™3Fe2q q S8.5Ž.Bacteria oxidise sulphur:2S8q3O q2H O™2SO y q4H q.6Ž.224The main objective of this experimental work was to test this biotechnology in order to obtain high gold extraction from pyrrhotite ores by cyanidation in a laboratory scale Ž.Attia and El-Zeki,1989;Paponetti et al.,1991.Furthermore,every unit operation of the process for gold recovery after cyanidation has been tested in a laboratory scale: gold adsorption r desorption and electrowinning,in order to evaluate the main process conditions with a high gold recovery.In this manner,an integrated process enclosing a biological and a chemical section has been proposed,in laboratory scale.Few data are reported in the literature for pyrrhotite bioleaching.Consequently,the efforts to develop new treatment technologies are justified since important deposits ofŽgold associated to sulphide minerals occur in the world Crundwell,1995;Ubaldini et .al.,1995.2.Material and methods2.1.Ore characterizationThe pyrrhotite ore sample comes from Atoroma mine in Bolivia.Representative samples were prepared and ground to y74m m for characterization.By optical()250S.Ubaldini et al.r Int.J.Miner.Process.602000247–262microscopy examination of the polished sections and scanning electron microscopy Ž.SEM analysis,it has been demonstrated that gold is mainly locked in submicronŽ.inclusions within the lattice of the pyrrhotite Abbruzzese et al.,1994.Mineralogical characterization by X-ray diffraction spectrometry evidenced pyrrhotite and quartz, while main additional minerals were pyrite and marcasite.Quantitative chemical analy-sis of the elements by emission spectroscopy technique showed that gold content is y1y1Žof10g t and silver content of9.5g t see Table1for complete chemical .composition.2.2.MicroorganismsA mixed culture of microbial strains of Thiobacillus sp.was selected from Cotilia Ž.Ž.`Italy sulphuric waters Veglio et al.,1995a.For the selection,cultivation,and maintenance of microbial strains utilised in the bioleaching process,liquid and solid Ž.media were used Visca et al.,1989;Atlas,1993.3q2q yŽThe selection was evaluated by monitoring Fe,Fe,SO and pH vs.time Bhatti,4.`1993;Ubaldini et al.,1995;Veglio et al.,1995a.The cultures were grown in a sterilisedŽ9K liquid medium in the presence of pyrrhotite as energetic substrate Paponetti et al., .1991.Table1Chemical analysis of the gold-bearing pyrrhotite from Atoroma in BoliviaCompounds%FeS46.25SiO19.722Ž.FeS pyrite11.402Ž.FeS marcasite 6.102Al O 4.5723SnO 1.352FeAsS0.85MgO0.68K O0.442TiO0.372MnO0.09CaO0.01Cu0.18Pb0.10Sb0.03Cr0.02Zn0.02y1Ž.Au g t10y1Ž.Ag g t9.5LOI 4.87LOI s loss on ignition.()S.Ubaldini et al.r Int.J.Miner.Process.602000247–262251Table2Experimental conditions of the bioleaching testsFactors ValuesŽ.Particle size m m-74Ž.Pulp content%w r v10–20Initial pH 2.0Ž.Stirring rate rpm200Ž.Temperature8C30Ž.Time of treatment days7–302.3.Bioleaching tests in shaken flasksBioleaching experiments were planned on the basis of the main factors that influenceŽthe biooxidative process Ahonen and Tuovinen,1995;Bhatti,1993;Tuovinen and.`Kelly,1973;Veglio et al.,1995a.In general,the sterilised pyrrhotite ore sample wasŽ.placed into each shaken flask volume300ml with100ml of9K culture media andŽ. inoculated with10ml of the mixed culture age of inoculum7–10days,pH was adjusted by H SO.The shaken flasks were placed in a rotary incubator.24Samples were taken periodically to monitor iron,sulphate,cell concentration in solution and pH during the bioleaching process.Table2shows the experimental conditions applied in the biological tests.After bioleaching,solid–liquid separation of the biooxidated residues were carried out by Ž.filtration Millipore,1.2m m.In the subsequent operation,the solid was neutralized by Ž.Ca OH,washed,dried,weighed and analysed for gold determination before cyanida-2Žy1.tion average of gold content was in the range9.9–10.1g t.2.4.Cyanidation testsHomogeneous biotreated samples were collected,with the aim to achieve representa-Ž.tive samples for gold recovery on laboratory scale Abbruzzese et al.,1994.Ž.Leaching tests were conducted in hemispherical glass reactor500ml volume,at atmospheric pressure.Experimental conditions of the cyanidation tests are reported in Ž.Table3Abbruzzese et al.,1994.At the start of the tests,NaCN concentration wasTable3Experimental conditions of the cyanidation testsFactors ValuesŽ.Particle size m m-74Ž.Weight of the samples g100Ž.Pulp content%50pH11Ž.Stirring rate rpm200Ž.Temperature8C25Ž.Time of treatment h24y1Ž.NaCN concentration kg t25()252S.Ubaldini et al.r Int.J.Miner.Process.602000247–262y1Ž.fixed at25kg t,after neutralization by Ca OH.During the tests,the consumption of2NaCN was determined by iodometric titration and its concentration was balanced by further addition.Samples of cyanide leached solution were taken periodically for gold and cyanide determination.At the end of each cyanidation test,the pulp was filtered.The solid was exposed to gold analysis,while the solution was stored for the subsequent adsorptionŽ.stage.pH was continuously monitored and adjusted by Ca OH.22.5.Adsorption r desorption testsThe leached gold was concentrated and purified by adsorption onto activated carbon and subsequent desorption,on laboratory scale.The gold concentration of the solution before adsorption was of10mg l y1.Ž.ŽThe adsorption process was carried out in stirred reactors500ml volume Lu and.Bai,1992;Paponetti et al.,1991.Active coconut carbon,produced from CECA Italiana Ž.Milan,Italy was used as adsorbent:distribution of particle size was between1and5 mm,benzol index35,iodine index130,apparent density580kg m y3and specific surface1200m2g y1.Coconut shell carbon has been chosen due to its high gold loading and excellent adsorption kinetics.The carbon was dried for several days at room temperature before being weighed for adsorption tests.During the tests,solution samples were withdrawn for gold determination.After adsorption,the loaded carbon was separated from the solution by filtration.Table4 shows the main experimental conditions.Ž.Ž.Preliminary test of carbon-in-pulp CIP has been carried out Adams,1990.Stirring rate was increased to300rpm,while duration of treatment to60min.Carbon concentration was fixed at60g l y1pulp.After the test,the carbon was separated from pulp by sieve with mesh-1mm.At the end of the test,the pulp was filtered and the solid was submitted to gold analysis.ŽThe adsorbed gold was stripped from carbon by a fixed bed glass column Espiell et .al.,1988.The volume of the column was of80ml and it was completed with a thermostatic bath.Experimental determinations evidenced that each column can be loaded with10g of carbon.Such as in the case of the adsorption phase,the pH wasŽq measured by a combined glass r saturated-calomel electrode specific to measure H ionTable4Experimental conditions of the adsorption onto activated carbonFactors ValuesŽ.Temperature8C20y1Ž.Carbon concentration g l60pH11Ž.Stirring rate rpm200Ž.Duration min40()S.Ubaldini et al.r Int.J.Miner.Process.602000247–262253Table5Experimental conditions of the desorption from activated carbonFactors ValuesŽ.Temperature8C70y1Ž.Flow rate of stripping solution ml min5y1Ž.Cyanide concentration g l10Ž.Ethanol concentration%v r v10pH11Ž.Duration h24.concentration.The stripping solution passed from the bottom of the fixed bed of carbon,being fed from a1000-ml glass tank via a peristaltic pump.Table5summarises the experimental conditions.2.6.ElectrowinningThe auriferous solution was electrowon in a laboratory-scale electrolytic glass cell of 300ml of capacity,equipped with a thermostatic water jacket.An AMEL mod.555B potentiostat–galvanostat apparatus,equipped with an instrument system to automatically control process parameters,was used for the electrowinning tests.Metallic gold was recovered onto a steel cathode,while platinum wire was used as the anode.The reference electrode was a saturated calomel electrode with standard constant potential at constant temperature.The cathodic useful surface was60cm2,the external surface being made inert by using insulating plastic material.ŽThe electrolysis tests were carried out treating200ml of gold cyanide solution60 y1.mg l Au,coming from desorption stage,at constant temperature and at constantŽcathodic voltage.The current intensity was measured by a digital multimeter Gallagher .et al.,1989.Gold concentration was measured by spectrophotometry during electro-de-position.Quality of gold deposit on cathode surface was determined by X-ray diffrac-tion.The experimental conditions of the electrowinning tests are reported in Table6.Table6Experimental conditions of the electrowinning testsFactors ValuesŽ.Bath temperature8C40Ž.Cell voltage V y1.2Ž.Current intensity mA60pH11Ž.Duration min60()254S.Ubaldini et al.r Int.J.Miner.Process.602000247–2622.7.Chemical and biological determinationsFerrous and ferric iron were evaluated colorimetrically,after reaction with KSCN,and sulphate ions turbidimetrically,after reaction with BaCl,using UV–Vis spectro-2Ž.`photometer technique Veglio et al.,1995a.The evaluation of microbial growth wasŽperformed by microscopy cell counting utilising Thoma counting chambers Beolchini et .8y1 al.,in press.The initial cell concentration was of about 1.4=10cells ml ofsuspension.Solutions were analysed for gold determination by atomic absorption spectrophotom-Ž.eter Perkin-Elmer mod.460.3.Results and discussion3.1.Bioleaching of pyrrhotiteThis investigation was realised on the basis of the previous data obtained by factorial Ž.`experiments Ubaldini et al.,1995;Veglio et al.,1993b.It was demonstrated that 40–45%iron dissolution occurs in the presence of10–20%of mineral concentration, after7days of biooxidation time,at initial pH2.In the present study,biological tests have been carried out extending the time of treatment to30days,checking the microbial growth.The aim was to verify the behaviour of biotreatment during a longer period and its effect on gold extraction yields by cyanidation.The cell concentration in solution,pH evolution,total iron,ferric ironŽ.and sulphate were monitored vs.time Figs.1–4.Fig.1shows the behaviour of microbial growth during bioleaching experiments.It is evident that at10%pulp content, the cell concentration in solution is larger than the one at20%pulp content.This factŽmay be due to the relation between cell adsorption and mineral concentration Konishi et .al.,1994.In fact,the amount of adsorbed cells increases with the solution ore content. Consequently,cell concentration results are lower in the case of the highest pulp content.Furthermore,the adsorption of cells onto the mineral surface has been demon-Ž´strated elsewhere to follow Langmuir law Beolchini et al.,in press;Veglio et al.,in .press.Fig.1also shows that cell concentration was maximum during the first day of bioprocess,as in the case of20%of ore concentration.This was probably associated toŽ.the pH vs.time profile Fig.2.In fact,it presents a maximum too,corresponding to about4.3.This value is far from the optimum pH for microbial activity,that is about2,Ž.so that cells might be desorbed from the mineral Touvinen and Kelly,1973.ThisŽ. initial increasing of pH might be due to the chemical step described by the reaction1.Ž.In fact,it has also been observed in the blank test not shown here and is proportional to pyrrhotite concentration.The subsequent decreasing of pH after7days of bioleaching is similar for both ore concentrations,which is related to the bacterial activity.In fact,Ž.this behaviour was not observed in the blank test not reported here and was related toŽ.3q the oxidation of sulphur as described by the reaction4and the precipitation of Fe Ž.Ž.Fig.3Toro et al.,1988.Fig.3shows that ferric iron concentrations are relatively low()S.Ubaldini et al.r Int.J.Miner.Process.602000247–262255Fig.1.Microbial growth during the biological tests at two different pulp densities.Ž.Fig.2.Trend of pH during biological tests10%and20%pulp densities.()S.Ubaldini et al.r Int.J.Miner.Process.602000247–2622563q Ž.Fig.3.Fe total and Fe vs.time in bioleaching test 10%and 20%pulp densities .Ž.for both leaching conditions 10%and 20%pulp concentration .Consequently,most of the dissolved iron is in the form of ferrous iron.This might be due to the limiting oxygen conditions.Oxygen plays an important role on the biooxidation in shaken flasks:ŽŽ.Ž.Ž.this is suggested by Eqs.2,3,and 6,where oxygen appears as a reagent.Therefore,this process cannot take place in abiotic conditions.Theoretical stoichiomet-ric considerations demonstrate the limiting oxygen conditions during the bioleaching Ž.tests.Gassing-out tests Veglio et al.,1995b,c showed that the overall oxygen mass `transfer in shaken flasks is about 0.09min y 1,in the investigated experimental condition,y 1Ž.with an oxygen supply of about 0.06mg min of O Veglio et al.,1995b,c .On the`2other hand,an oxygen requirement of about 0.08mg min y 1was estimated during thebioleaching tests when mineral concentration is 10%,considering the reactions involved Ž.in the bioleaching process Veglio et al.,1995a .A further confirmation of limiting `Ž.oxygen conditions is given by sulphate profiles Fig.4.In fact,they are similar and the result does not depend on the ore concentration.Consequently,oxygen-limiting condi-tions seem to play an important role during bioleaching in shaken flasks.For this reason,lower ore concentrations have been used in the bioleaching tests,in order to develop a Ž.kinetic model for the bioleaching process Veglio et al.,in press in the absence of ´oxygen-limiting conditions.In the present case,the main aim of the bioleaching tests was the production of significant quantities of biotreated ores to be exposed to further treatments for gold recovery.Consequently,it was not effective to perform trials at low ore concentration.Obviously,higher concentrations should be employed for an indus-Fig.4.Behaviour of dissolution of SO2y during bioleaching tests.4trial application,even higher than20%.In that case,mechanically aerated conditions might be used,in order to avoid oxygen-limiting conditions.Fig.5shows iron extraction yields vs.time profiles during bioleaching tests.It is evident that similar iron extraction yields have been achieved in both experimental Ž.conditions10%and20%ore concentration:about50%iron extraction yield are obtained after15days of treatment.On the other hand,no significant iron extraction Ž.yields-10%have been observed in the blank test,in the absence of Thiobacillus Ž.data not shown.3.2.Cyanidation of the bioleached residuesŽ.Cyanidation tests have been performed both on non-treated pyrrhotite as is samples and on bioleached samples.An average experimental error of about2.5%was deter-Ž.mined Davies,1979,from three replications of the experimental tests.Considering the non-treated samples,only6.2%Au was recovered after2h,while19.7%Au recovery was obtained after24h leaching,with high NaCN and lime concentrations.In light of this result,it is possible to affirm that the ore sample is refractory under the processing aspect,because the response to conventional treatment for gold extraction is negative.The bioleaching residues were then collected,obtaining homogeneous samples asŽy1.concerns the gold content about10g t Au,in order to evaluate the effect of theŽbiotreatment on the subsequent cyanidation tests for gold recovery loss of weight by.biooxidation was about50%.The obtained results evidenced gold recoveries signifi-Fig.5.Curves of iron extraction yield IEY at different pulp densities.cantly higher than those obtained from direct cyanidation.In fact,42.2%Au was obtained just after2h,while67.3%Au and84.1%Au were dissolved after10and24h,Ž.respectively Table7,from samples7days bioleached.Increasing the time of biotreat-ment to30days,gold dissolution increased to51.2%after2h,to72.5%Au after10h,Ž.and to91%Au after24h of cyanidation Table7.Lime consumption was10.1kg t y1for as is samples,while it was about18and19 y1Žkg t for the pretreated samples after7and30days of bioleaching,respectively Table .8.As for pretreated samples,the higher consumption is mainly due to the operation of neutralization.y1Ž.Cyanide consumption increases from16kg t without bioleaching to about21kg y1Ž.y1Ž.Ž.t7days bioleaching and to about22kg t30days bioleaching Table8.This increase can be explained considering that pyrrhotite is a common refractory mineralTable7Gold recovery by cyanidation,without biological pretreatment or after7and30days of biooxidationTime Without bioleaching After7days bioleaching After30days bioleachingŽ.Ž.Ž.Ž.h%Au%Au%Au2 6.242.251.21015.367.372.52419.784.191.0Table8Cyanidation process:consumption of reagentsy1y1Ž.Ž.Ž.Process route Ca OH kg t NaCN kg t2Without bioleaching10.116.07days bioleaching18.221.230days bioleaching19.122.1and its decomposition forms ferrocyanide which removes free cyanide from solution Ž.`Veglio et al.,1995a.3.3.Purification and precipitation of the goldIn this section,the results obtained in the unit operations aimed at the recovery ofŽpure gold after cyanidation are presented.As for the purification cycle adsorption r de-.y1 sorption,98.1%Au was adsorbed onto activated carbon,obtaining0.2g kg Au of Žcarbon data confirm the high gold loading and excellent adsorption kinetics of the .carbon utilised,while98.9%Au was stripped during desorption phase.Gold wasy1Ž.y1concentrated from10mg l Au leached solution to60mg l Au.The results are very interesting;in fact,this cycle consists of innovative operations aimed at purification and concentration of gold.Furthermore,regeneration studies have demonstrated that it isŽpossible to utilise the unloaded carbon four times after chemical regeneration with.Ž. dilute solution of HCl at pH5and thermal treatment at high temperature5008C.Moreover,by preliminary test of CIP about95%Au was adsorbed onto activated carbon;the activated carbon showed a good attrition resistance:the loss of carbon was of0.2g kg y1ore for each cycle adsorption r desorption.As for the precipitation step,97.8%Au was recovered by electrowinning the concentrated and purified solution.The final concentration of gold in the barren Žy1.solutions1mg l resulted after60min of electrolysis.The gold deposit on cathodeŽ.surface was of good quality high compactness and purity.The current efficiency was Ž.low 6.5%after60min due to the concurrent parasitic reactions at the electrodes,while the energetic consume was of18kW h y1kg y1Au recovered.The implementation of the electrowinning in industrial scale,together with the application of cathode with a high surface,could improve the efficiency and the economy of the process.Table9summarises gold recovery obtained in each unit operation of the process.It is possible to observe that86.3%Au recovery was achieved by the complete circuit of treatment,on laboratory scale.Table9Gold recovery for each section of the complete flowsheetBioleaching–cyanidation Purification Precipitation Total recoveryŽ.Ž.Ž.Ž.%Au%Au%Au%Au91.097.097.886.34.ConclusionsIn this work,an integrated process for gold recovery from the auriferous pyrrhotiteŽ.coming from Atoroma mine Bolivia has been investigated at a laboratory scale.The biological pretreatment prior to the conventional leaching has been demonstrated to lead to high gold recovery.In fact,only19.7%Au recovery was obtained,by direct cyanidation of Atoroma ore,with high reagent concentration.On the other hand,bench scale bacterial oxidation experiments permitted achievement of higher gold recovery: about84%Au was obtained from samples bioleached7days,while91%Au was achieved from samples bioleached30days,in batch conditions.ŽThe gold recovery of the whole process bioleaching,solid–liquid separation,.cyanidation,purification and precipitation was about86%Au.Moreover,the recovered gold was demonstrated to have good characteristics.Further work is in progress,aimed at the process scaling up.In particular,a technical and economical comparison of two gold extraction processes,cyanidation with and without biotreatment,will be carried out,considering that there are many possibilities for technology transfer in the East-Europe,that could benefit from this biotechnology.AcknowledgementsThe authors are grateful to Dr.A.Nardini,Mr.M.Centofanti,Mr.P.Fornari,Mr.R. Massidda and Ms.Emanuela Tempesta for their helpful collaboration in the experimen-tal work;moreover,the authors thank Mr.Giampaolo Marruzzo for the X-ray analysis.References`Abbruzzese, C.,Ubaldini,S.,Veglio, F.,Toro,L.,1994.Preparatory bioleaching to the conventionalŽ.cyanidation of arsenical gold ores.Miner.Eng.71,49–60.Adams,M.D.,1990.The mechanism of adsorption of aurocyanide onto activated carbon:1.Relation between the effects of oxygen and ionic strength.Hydrometallurgy25,171–184.Ahonen,L.,Tuovinen,O.H.,1995.Bacterial leaching of complex sulfide ore samples in bench-scale column reactors.Hydrometallurgy37,1–21.Ž.Atlas,R.M.,1993.Thiobacillus medium B.In:Parks,L.C.Ed.,Handbook of Microbiological Media.894 pp.Attia,Y.A.,El-Zeki,M.,1989.Bioleaching of gold pyrite tailings with adapted bacteria.Hydrometallurgy22, 291–300.`Beolchini,F.,Nardini,A.,Veglio,F.,Development of a direct count method for cells determination and application in sorption of Thiobacilli on pyrrhotite ore particles,Res.Environ.Biotechnol.,in press.`Beolchini,F.,Veglio,F.,Kinetic modelling of pyrrhotite ore leaching by ferric iron and related statistical analysis,Ind.Eng.Chem.,in press.Bhatti,T.M.,1993.Mineral products of pyrrhotite oxidation by Thiobacillus ferrooxidans.Appl.Environ.Ž.Microbiol.596,1984–1990.Boon,M.,Hansford,G.S.,Heijnen,J.J.,1995.The role of bacterial ferrous iron oxidation in the biooxidationŽ.of pyrite.In:Jerez,C.A.,Vargas,T.,Toledo,H.,Wiertz,J.V.Eds.,Biohydrometallurgical Processing, vol.I,IBS,Chile,pp.153–163.。