ideide

Wa

s, CSI

ised fline 2

cathode copper. Copper production from SX-EW rose sulphides is a developing technology that has beenapplied successfully to the extraction of copper fromsecondary sulphide minerals such as chalcocite at a

(200from 0.8 to 2.0 Mt in the period 1993 to 1997 (Arbiterand Fletcher, 1994; Readett and Townson, 1997). In1. Introduction

World copper production has increased steadily in theperiod 19842005, from 9Mt to 16Mt per annum, and ispredicted by the Australian Bureau of Agricultural andResource Economics (ABARE) to reach close to 18 Mtin 2006. More than 20% of that copper is now producedvia hydrometallurgy. An indirect indicator of the notableincrease in hydrometallurgical copper production overrecent years is the increased overall capacity of solventextraction-electrowinning (SX-EW) plants producing

2001, the combined copper production of Chile and theUSA using SX-EW was about 2.1 Mt (Bartos, 2002)with additional production of about 0.16 Mt from othercountries.

While world demand for copper is growing, theminerals industry is increasingly faced with the need toprocess low grade ores, overburden and waste fromcurrent mining operations. The economic extraction ofcopper from low-grade ores requires low-cost proces-sing methods such as in situ, dump and heap leaching.Bacterially-assisted heap leaching of low-grade copperstatus of the bioprocessing of copper sulphides and evaluates promising developments.The bioleaching of sulphide minerals is reviewed with emphasis on the contribution from the microbial community, especially

attachment and biofilm formation, bioleaching mechanisms and the potential benefits to be gained by a greater understanding of themolecular genetics of the relevant microbial strains.

The leaching and bioleaching of copper sulphides is examined. The main focus is on heap bioleaching of whole ores, and thedevelopment of models to describe heap and dump processes that can be applied in the design phase as well as to optimise metalextraction. The characteristics of chalcopyrite leaching are discussed in respect of those conditions and controls that might beneeded to make a heap bioleach commercially productive. 2006 Elsevier B.V. All rights reserved.

Keywords: Bioleaching; Heap leaching; Dump leaching; Copper sulphides; Chalcopyrite; Mechanism; ModelingThis review outlines current research in heap bioleaching, pAbstract

articularly in respect of the bioleaching of chalcopyrite, assesses theThe bioleaching of sulphon copper sulph

H.R.

Parker Centre for Integrated Hydrometallurgy Solution

Received 16 January 2006; received in revAvailable on

Hydrometallurgy 84E-mail address: helen.watling@csiro.au.

0304-386X/$ - see front matter 2006 Elsevier B.V. All rights reserved.doi:10.1016/j.hydromet.2006.05.001minerals with emphasiss A review

tling

RO Minerals, PO Box 90, Bentley, WA 6982 Australia

orm 8 May 2006; accepted 12 May 20063 June 2006

6) 81108www.elsevier.com/locate/hydrometnumber of operations worldwide. However, heap bio-leaching of the refractory primary copper sulphide,

chalcopyrite, has yet to be implemented at commercialscale. Traditionally, the advantages of bacterial leachingtechnology are compatible with these requirements:

Moderate capital investment with low operatingcosts,

Appropriate recovery of metals from low-grade oresand waste materials,

Basic equipment and simple operating procedures.

The most successful copper heap leaching operationshave been those processing copper oxides and secondarycopper sulphides (Table 1). Chalcocite (Cu2S) is the maincopper sulphide mineral mined at bioleaching operations.Some of the chalcocite heap operations began as oxide(chemical) leach operations andwere converted to bioleach(oxidative) operations by heap aeration and/or inoculation,when the oxidised ore was depleted. However, even ifbacterial activity is not facilitated, microbial assisted air

oxidation of iron(II) and sulphur will contribute to copperextraction if sulphide minerals are present in a heap.

Millions of tonnes of low-grade ore and copper-richtailings await the development of an efficient and eco-nomic bioleach process for chalcopyrite (CuFeS2). Thebioleaching of chalcopyrite in situ and in dumps is apractical option only because the low and slow recoveriesare countered by the low processing costs (Schnell, 1997).Not surprisingly, the bioleaching of chalcopyrite, both themost abundant and the most refractory copper sulphide, isa key industry target. However most of the technologicaldevelopments have taken place with the bioleaching ofchalcocite and other less refractory sulphide minerals.

Straits Resources operated a test heap with chalcopy-rite ore in parallel with their copper oxide/chalcocite heapleach operation at Girilambone with promising results(D. Readett, personal communication). Titan Resourcesoperated a trial mixed nickel sulphide and chalcopyriteheap at their Radio Hill deposit with some success

Table 1Heap bioleaching of copper ores (historical and current)

Region/mine Operation reserves(t)

Ore processed(t/day)

Cu production(t/year)

Lo Aguirre, Chile 19801996 Heap bioleach 12106 at 1.5% Cu Oxides/chalcocite 16103 1415103

Cerro Colorado, Chile 1993 Heap bioleach 80106 at 1.4% Cu Chalcocite, covellite 16103 100103

Ivan Zar, Chile 1994 Heap bioleach 5106 at 2.5% Cu Oxides/sulphides 1.5103 12103

Quebrada Blanca, Chile 1994 Heap/dump bioleach 85106 at 1.4%Cu 45106 at 0.5% Cu

Chalcocite 17.3103 75103

Punta del Cobre, Chile 1994 Heap (bio)leach 10106 at 1.7% Cu Oxides/sulphides 78103

Andacollo, Chile 1996 Heap/dump bioleach 32106 at 0.58% Cu Chalcocite 15103 211033

6 at 1

Cu

0.7%

8% C

1.8

1%

5% C

.78%

82 H.R. Watling / Hydrometallurgy 84 (2006) 81108Dos Amigos, Chile 1996 Heap bioleach 2.5%Zaldivar, Chile 1998 Heap/dump bioleach 12010

Cu 115106 at 0.4% CuLomas Bayas, Chile 1998 Heap/dump 41106 at 0.4%Cerro Verde, Peru 1977 Heap bioleach at 0.7% CuEscondida, Chile Heap bioleach 1.5109 at 0.3Lince II, Chile, 1991 Heap leach 1.8% CuToquepala, Peru Heap leachMorenci, Arizona 2001 Mine for leach 3450106 0.2Equatorial Tonopah,Nevada, 20002001

Heap bioleach 0.31% Cu

Gunpowder Mammoth Mine,Australia, 1991

In situ (bio)leach 1.2106 at

Girilambone, Australia19932003

Heap bioleach at 2.4% Cu

Nifty Copper, Australia, 1998 Heap bioleach at 1.2%Whim Creek and Mons Cupri,Australia 2006

Heap bioleach 900103 at 1.Cu 6106 at 0.8% Cu

Mt Leyshon, Australia19921997

Heap bioleach 0.15%

S&K Copper, Monywa,Myanmar, 1999

Heap bioleach 126106 at 0.

Phoenix deposit, Cyprus, 1996 Heap (bio)leach 9.1106 at 0Cu 5.9106 at 0.31% CuJinchuan Copper, China 2006 240106 at 0.63% CuChalcocite 310 .4% Chalcocite 20103 150103

Oxides/sulphides 36103 60103

Oxide/sulphide 32103 54.2103

Oxides, sulphides 200103

Oxides, sulphides 27103

Oxides, sulphides 40103

u Chalcocite, pyrite 75103 380103

25103 25103

% Cu chalcocite and bornite 33103

Chalcocite/chalcopyrite 2103 14103

Oxides/chalcocite 5103 16103

Oxides/sulphides 17103

Chalcocite 1.3103 750

u Chalcocite 18103 40103

Oxide/sulphide 8103Chalcocite, covellite, enargite 10103

etallu(Hunter, 2002a). Currently, Mintek, with the NationalIranian Copper Industries Company (NICICO), areundertaking a large-scale pilot test of Mintek's heapbioleaching technology for Darehzar chalcopyrite ore atthe Sarcheshmeh Copper Complex in southern Iran (vanStaden et al., 2005).

Clearly, there remain impediments to the acceptanceand implementation of bioleaching for the processing ofchalcopyrite, not necessarily restricted to the biologicalaspects. According to Holmes and Debus (1991), bio-logical processing would need to have N20% advantageover conventional processing to interest the mining in-dustry. Poulter et al. (1999) concluded that reluctance toembrace the technology is partly a reflection of therelatively refractory nature of chalcopyrite, partly theinherent process and economic advantage of modernsmelting technologies, and partly the real and perceivedtechnical risks associated with the introduction of a novelprocessing technology.

The purposes of this review are: (i) to outline thefocus of current research, particularly in respect of thebioleaching of chalcopyrite; (ii) to assess the status of thebioprocessing of copper sulphides and, (iii) to evaluatepromising developments. The theory and practice ofheap, dump and in situ leaching are not described in anydetail because these topics have been covered in severalsubstantial reviews and numerous more specific pub-lications (e.g., Murr, 1980; Rossi, 1990; Bartlett, 1998).

2. The bioleaching of sulphide minerals

Themain research topics in the bioleaching literature are(i) fundamentals and modeling of bioleaching chemistry(and to a small extent, mineralogy), (ii) microbiology, (iii)bioleaching of selected sulphide minerals, commonly asso-ciated with specific ores or mines, and (iv) the engineeringaspects of heap and dump leaching. The fundamental che-mical and microbiological studies are usually generic,rather than copper-oriented, and conducted on sulphideconcentrates rather than ores. The engineering aspects areoutside the scope of this review.

The bioleaching of sulphide minerals occurs in anacidic medium that often contains a considerable concen-tration of iron(III). While the emphasis here is on bio-leaching, acid leaching and galvanic interactions betweensulphide minerals and iron(III) also contribute to theoverall efficiency of the leach.

2.1. Leaching chemistry

The extraction of selected metals from mineral

H.R. Watling / Hydromsulphides such as chalcopyrite can be an oxidative processin which ferric ions are the oxidant and the sulphidecomponent of the mineral is oxidized to elemental sulphurEq. (1).

CuFeS2 4Fe35Fe2 Cu2 2So 1The reaction is known to be sensitive to redox poten-

tial; and surprisingly, higher dissolution rates have oftenbeen measured at lower potentials, in the range 0.450.65 V SHE (Peters, 1976; Third et al., 2000; Hiroyoshiet al., 1997; 2001). In addition, it has been shown that aparallel, non-oxidative dissolution process that is alsosensitive to potential, contributes to the kinetics of chalco-pyrite leaching in sulphuric acid solutions Eq. (2) (Nicoland Lazaro, 2003; Lazaro and Nicol, 2003).

CuFeS2 4HFe2 Cu2 2H2S 2The main disadvantage of (bio)leaching of sulphides

is that the process is perceived to be slow relative topyrometallurgical process routes and other high-inten-sity hydrometallurgical processes. However, the kineticsof mineral dissolution may change when two mineralsare in electrical contact with each other, as is often thecase in mineralized ores. For example, in isolation, thedissolution rate of pyrite is faster than that of sphalerite,galena or chalcopyrite, but when pyrite is in intimatecontact with one of these minerals, the situation is re-versed (Ramachandra Rao and Finch, 1988; Das et al.,1999b; Sui et al., 1995). This response is due to galvanicinteraction and could be exploited to enhance leachingrates for base metal sulphides of interest.

In the case of chalcopyrite, both sulphur- and iron-containing reaction products have been invoked as thecause of slow dissolution. The insoluble reaction pro-ducts formed on the chalcopyrite surface during lea-ching and bioleaching have been examined using X-rayphotoelectron spectroscopy (XPS), a surface sensi-tive analytical technique. Four sulphur-containing spe-cies were detected on the leached chalcopyrite surface,namely a sulphide phase (unreacted chalcopyrite), ele-mental sulphur Eq. (1), a basic ferric sulphate phase akinto jarosite, and a disulphide phase (Fig. 1) (Klauber,2003; Klauber et al., 2001; Parker et al., 2004). Similarspeciation was found for abiotic chemical leaching andbioleaching under both aerobic and anaerobic condi-tions, suggesting that there was a common leachingmechanism.

On the basis of the sulphur speciation, a mechanismfor the ferric ion oxidation of chalcopyrite has beenproposed. A key feature of the mechanism is the oxi-dation of the disulphide phase, which forms rapidly on

83rgy 84 (2006) 81108freshly fractured chalcopyrite and persists on leached

etallusurfaces. Oxidation of the disulphide phase directlyproduces thiosulphate which is then oxidized to sulphate,generating the basic ferric sulphate that then acts as atemplate for jarosite formation. The jarosite layer thenbuilds to the point that it hinders further chalcopyrite

Fig. 1. Sulphur speciation of a chalcopyrite concentrate after leachingwith ferric sulphate solution and surface analysis using X-rayphotoelectron spectroscopy (Klauber et al., 2001, redrawn).

84 H.R. Watling / Hydromoxidation (Klauber, 2003).The chemistry of microbial oxidation of sulphides is

closely related to that of ferric ion oxidation in acidicsolutions Eq. (1). The microorganisms play a catalyticrole in oxidising ferrous ion to ferric ion, thus rege-nerating the oxidant Eq. (3). They also oxidize sulphurto sulphate, generating acid Eq. (4).

2Fe2 2H 0:5O22Fe3 H2O 3

2S 3O2 2H2O2H2SO4 4

The costs associated with themaintenance of sulphidebioprocessing microorganisms are minimal becausemany gain energy from the redox reactions Eqs. (3)and (4), utilize carbon dioxide from the air as C sourceand obtain their phosphorus, nitrogen, potassium andmicronutrients, etc.), from the bioleach environment.

Because the microorganisms are particularly efficient atoxidising ferrous ions to ferric ions, the bioleaching con-ditions typically exhibit a relatively high redox potentialaround 0.650.70 V SHE, which is less conducive tochalcopyrite dissolution. One consequence of the highsolution potential is that ferric ion readily precipitates as aray fluorescence and X-ray diffraction analysis of leachresidues was combined with elemental composition tounderstand the leaching of gangue minerals in mildsulphuric acid column leaching experiments; mineralcharacterisation by microscopy together with clay phaseidentification using X-ray diffraction was correlated withacid leaching performance (Helle and Kelm, 2005; Helleet al., 2005; Kelm and Helle, 2005). When the automatedbasic sulphate, like jarosite Eq. (5) in an environmentcontaining monovalent alkali cations and sulphate ions.

3Fe3 2SO24 6H2O MMFe3SO42OH6 6H

5

where M=K+, Na+ or NH4+

Basic iron sulphate precipitates have been implicatedin hindering the complete dissolution of chalcopyriteduring bioleaching (Stott et al., 2001; Petersen et al.,2001). Kinnunen et al. (2003) proposed leaching at pH 1to hinder/delay the formation of jarosite precipitates,and used a pH 1-adapted, immobilized biomass ofLeptospirillum-like organisms to regenerate the ferricion oxidant. More recently, Tshilombo et al. (2002)showed that the passivation of chalcopyrite duringbacterial leaching could be countered by controlling thethermal (4565 C) and electrochemical (0.450.65 VSCE) conditions. The authors noted that the passivelayers formed at 25 C strongly inhibit ferric ion re-duction on polarized chalcopyrite surfaces.

2.2. Mineralogy

The impact of mineralogy on leaching and bioleachingshould not be ignored. Extensive mineralogical analysisof ore types around a deposit is required in developing aflow sheet. In heap leaching, acid consumption by gangueminerals is a key parameter and therefore sulphuric acid isusually a major processing cost. Maintenance of the pH inthe preferred range between 1 and 2 for microbial iron andsulphur oxidation is also important for ferric ion and acidregeneration by the microbial population.

Surprisingly little quantitative research on the corre-lation between ore mineralogy, reaction chemistry andleach residue mineralogy has been reported in the publicdomain. The application of quantitative mineralogicalanalysis of feed and leach residues proved valuable inunderstanding autoclave chemistry during the highpressure acid leaching of nickel laterites (Whittingtonet al., 2003a,b) and is being refined to describe coppersulphide concentrates as part of a systematic leachingstudy (Tiller-Jeffery et al., 2004). A combination of X-

rgy 84 (2006) 81108SEM techniques (Gottlieb et al., 2000; Gu, 2003) are

etallucombined with quantitative analysis of X-ray diffractiondata, SEM-microprobe data and elemental analysis ofores and residues, they can provide new insights intoleach chemistry and reaction mechanisms for difficult toprocess ores.

2.3. Microbiology and leaching

Sulphide ore dumps and heaps, with their varyingmineralogical compositions and different climatic envir-onments represent extremely complex microbiologicalhabitats. Yet only a modest number of iron- and sulphur-oxidising bacteria (Table 2) have been isolated frommineral sulphide ores, characterised physiologically andphylogenetically, and deposited in data banks (Hallbergand Johnson, 2003 and references therein). These extremeacidophiles that grow optimally at pHb3, must compriseonly a small selection of those that have mineral pro-cessing capabilities.

The limited number of bacteria that have been disco-vered is partly a consequence of the selective methods bywhich bacteria are enriched and isolated. The perceivedimportance of Acidithiobacillus ferrooxidans in bioleach-ing is a case in point. For many years,A. ferrooxidanswasthought to be the dominant bacterial strain in bioreactorsoperated at temperatures lower than 40 C because, whencultures were grown on soluble iron media in batch tests,A. ferrooxidans outgrew Leptospirillum ferriphilum(previously thought to be L. ferrooxidans). It has sincebeen shown that the reverse is true in bioreactors. This isbecause high iron(III) concentrations inhibit A. ferroox-idans growth, whereas L. ferriphilum is relatively un-affected by high ferric ion concentrations and is also moretolerant of both higher temperatures and lower pH (Norriset al., 1988; Hansford, 1997; Schrenk et al., 1998;Rawlings et al., 1999). The dominance of L. ferrooxidansand Acidithiobacillus thiooxidans in highly acidic (pH0.7) copper heap leach environments has been reported(Vasquez and Espejo, 1997).

Moderate thermophiles and hyperthermophiles alsohave a role to play in making bioleaching more efficient,because they permit the use of higher temperatures,which in turn result in faster reaction rates. At inter-mediate temperatures (4045 C), Acidithiobacillus cal-dus is most likely the dominant sulphur oxidizer inbioreactors treating arsenopyrite or copper concentrates(Rawlings et al., 1999; Okibe et al., 2003). At tempe-ratures greater than 60 C Sulfolobus metallicus andMetallosphaera spp are thought to be the most importantbioleaching strains (Hallberg and Johnson, 2001).

While A. ferrooxidans was the first iron-oxidising

H.R. Watling / Hydromacidophile to be isolated from acidic bioleachingenvironments (Colmer et al., 1950), and has sincebeen subjected to the most intensive characterisation, itis not necessarily the most important. Microbial con-sortia responsible for the solubilisation of metals fromsulphide minerals are expected to be complex mixes ofautotrophic and heterotrophic bacteria (Tuovinen et al.,1991). A. ferrooxidans and L. ferrooxidans share theirenvironment with other acidophiles that have a similarphysiology and can compete for the available nutrientand energy sources.

With new molecular microbiological methods ofenumeration and identification of organisms, it is nowpossible to follow changes in microbial consortia as afunction of time or location. This is a valuable tool indescribing biodiversity and/or understanding bioleachingprocesses (Gonzalez-Toril et al., 2003; Okibe et al., 2003).Microbiological surveys of specific environments, such ashydrothermal sites (e.g. Atkinson et al., 2000; Burton andNorris, 2000; Simmons and Norris, 2000; Plumb et al.,2002), mine sites (Robertson et al., 2002; Keeling et al.,2004; Demergasso et al., 2005) or acidic mine drainagesystems (e.g., Hallberg and Johnson, 2003; Dopson et al.,2004) are yielding more robust data on biodiversity. Agroup of acidophiles, possibly overlooked because theygrow in a mid pH range (pH 36) in an acid mine drainagesystem, are of particular interest because of their role inpromoting the oxidation and precipitation of iron (Hall-berg and Johnson, 2003). Acidophilic heterotrophs arebeing discovered and characterised with increasing fre-quency (Hallberg and Johnson, 2001). The new methodsalso facilitate elucidation of the nature of microbialinteractions in enhancing bioleaching (Johnson et al.,2001; Okibe and Johnson, 2004), a research topic worthyof more focused attention.

In the context of bioleaching, most microbiologicalresearch is conducted utilising known iron- and sulphur-oxidising bacteria, either as single strains or in mixedcultures. However, in many instances, bacteria indigenousto the ore are not excluded and may thus contribute tobioleaching. Indeed, it is often noted that indigenous bac-teria, being acclimatised to high levels of selected metalsin their environment, are more effective as bioleachingcatalysts. The acclimatisation of bacteria to a particularmineral system by subjecting them to progressivelygreater amounts of the major elements present is commonpractice in test work. In addition, there have been manyfundamental studies on the tolerance of single strains andcultures to base metal ions (e.g., Das et al., 1997; Dopsonet al., 2003). In the same way, microbial growth can bepromoted at heap operations by the addition of nutrients toleach solutions and by creating conditions that result in

85rgy 84 (2006) 81108increased iron concentrations (Readett et al., 2003).

Table 2Iron and sulphur oxidising acidophiles

Organism Reported growth substrates Characteristics

Acidianus ambivalens S oxidation and reduction HyperthermophilesAcidianus brierleyi Sulphides pH opt 1.52.5Acidianus infernus Poor, if any, Fe oxidationAcidianus tengchongensis Acidimicrobium ferrooxidans Mixotroph Moderate thermophile pH opt 2

Fe oxidation and reductionSulphides (poor)

Acidiphilum spp Obligate heterotrophs MesophilesAcidiphilium SJH S oxidation pH opt 23

Fe(III) reductionAcidiphilium acidophilum Facultative autotroph Mesophile pH opt 23

S oxidationFe(III) reduction

Acidithiobacillus albertensis Autotrophs MesophilesAcidithiobacillus ferrooxidans S oxidation, sulphides pH range 24Acidithiobacillus thiooxidans (Af, Fe(II) oxidation; Fe(III) reduction as

a facultative anaerobe)Acidithiobacillus caldus Mixotroph Moderate thermophile

3S oxidation, sulphides pH opt 22.5Acidolobus aceticus Heterotroph Hyperthermophile

S reduction to H2S pH opt 3.8Alicyclobacillus spp S oxidation, sulphides Mesophiles moderate thermophilesAlicyclobacillus disulfidooxidans (Ad, facultative autotroph,; pH 1.52.5Alicyclobacillus tolerans At, mixotroph, Fe(III) reduction)Ferrimicrobium acidiphilium Heterotroph Mesophile

Fe(II) oxidation, sulphides pH opt 1.71.8Fe(III) reduction

Ferroglobus placidus Fe oxidation ThermophilepH neutral

Ferroplasma acidarmanus Possibly autotroph Moderate thermophilesFerroplasma cyprexacervatum Iron oxidation pH rangeb12Ferroplasma acidophilum Pyrite oxidation poorFerroplasma MT17Hydrogenobaculum acidophilus S, H oxidation to produce sulphuric acid Thermophile

pH opt 34Leptospirillum ferriphilum Fe oxidation Mesophiles, some thermo-tolerant strainsLeptospirilum thermoferrooxidans Pyrite pH range 1.61.9Leptospirillum ferrooxidans Fe oxidation, pyrite Mesophile

pH opt 1.51.7Metallosphaera sedula S oxidation ThermophilesMetallosphaera prunae Sulphides pH 14Metallosphaera hakonensisSulfobacillus acidophilus Fe(II) oxidation; Fe(III) reduction, Sulphides Moderate thermophilesSulfobacillus thermosulfidooxidans S oxidation pH 12.5Sulfolobus metallicus Strict chemolithoautotroph HyperthermophilesSulfolobus rivotincti S oxidation, sulphides Various pH in range 14.5Sulfolobus shibataeSulfolobus tokodaiiSulfolobus yangmingensisSulfolobus JP2 and JP3Sulfolobus acidocaldarius Heterotrophs HyperthermophilesSulfolobus solfataricus Not S oxidation pH 24.5Sulfurococcus yellowstonensis S and Fe oxidation HyperthermophileThiobacillus prosperus S and Fe oxidation Mesophile, halophile

sulphides pH opt 2Thiomonas cuprina S oxidation, sulphides Mesophile

pH opt 34

86 H.R. Watling / Hydrometallurgy 84 (2006) 81108

etalluRelated to this is the issue of acclimatisation to the hightotal dissolved solids (TDS) content of solutions thatdevelop during prolonged leaching, due to the dissolutionof the gangue minerals. Shiers et al. (2005) used a quan-titative batch culture method to investigate adaptation orhabituation of a mixed culture of acidophiles to growthmedia containing increased concentrations of sodium sul-phate or sodium chloride. Their results indicated relativelyrapid adaptation to sodium sulphate at levels in excess ofthose normally found in process water. However,concentrations of only 7 g/L sodium chloride inhibitedcell replication by more than 50% and no significantculture adaptation occurred during prolonged exposure.The search for acidophilic halophiles that do functionwellin high TDS water has not been particularly successful.Most iron oxidising halophiles prefer a higher pH range(e.g. Holden et al., 2001) which would promote greaterprecipitation of ferric compounds, thus diminishing theferric ion concentration available to oxidise the sulphidemineral. Acidophiles extracted from samples (pH 23.5,temperatures 3575 C) taken close to vents of Vulcano(Aeolian Islands) exhibited salt tolerance but grew betterin the absence of salt or at salinities lower than seawater(Simmons and Norris, 2000).

Recently, the possibility of utilising mineral specific (inthis case chalcopyrite-specific) bacteria has been raised(Williams et al., 1999). One [mixed] culture, in particular,apparently had a high affinity for oxidising chalcopyriteores. The bacteria were active over a wide range oftemperatures (4560 C) and solution pH values (0.82.2)that might be encountered within a heap. Evidence of thechalcopyrite specificity was obtained by leaching achalcopyrite/pentlandite and a chalcopyrite/pyrite concen-trate. Normally, in both cases, the oxidation of chalcopyritewould be slow and incomplete, in comparison with that ofthe other sulphide. However, in agitated, aerated tank tests,chalcopyritewas leached to completion in 1314 days. Therate of nickel dissolutionwas slow, and that of pyrite almostnon-existent, until most of the chalcopyrite had beenoxidised. The chalcopyrite-specific culture was tested on11 chalcopyrite ores or concentrates of diverse origins witha range of mineral phases, sulphur and copper contents.Copper extractions of 9599% were achieved consistentlyin 836 days. While the redox potentials of leach solutionswere not reported by Williams et al. (1999), it was notedmore recently (Hunter, 2002b,c) that the proprietary culturecontained sulfur-oxidising microorganisms but not iron-oxidising organisms. It may be deduced, therefore, that thereported enhanced bioleaching of chalcopyrite was aconsequence of low solution potentials (e.g. Hiroyoshiet al., 1997; Third et al., 2000) rather than chalcopyrite-

H.R. Watling / Hydromspecific microorganisms.2.3.1. Microbial attachment and biofilm formationIt is well known that bacteria attach to substrates and

form biofilms that have well-developed community struc-tures, with mechanisms for the delivery of nutrients and thedisposal of waste products. There are many observationsthat surfaces are the main sites of microbial activity innatural environments. van Loosdrecht et al. (1990) re-viewed the many influences of interfaces on microbialactivity. Those aspects relevant to bioleaching environmentshave been drawn out by Crundwell (1997) and expanded toinclude more recent studies in bioleaching systems.

In the past, attachment of bacteria to sulphide surfacesand the enhanced rates of dissolution of the sulphideswere, together, used as evidence that bioleaching pro-ceeded, in part, via a direct mechanism (enzymaticattack). The search for evidence to support this hypo-thesis has resulted in innovative studies on the role ofattachment and biofilm formation in bioleaching.

It has been reported that microorganismmineralinteractions result in changes in the surface chemistry ofthe microorganisms.

The surface charge on cells grown in media withsoluble iron (Fe2+) was different to that on cells grownon a solid substrate (sulphur, pyrite). The altered cellsurface charge was attributed to higher protein contentin the latter (Sharma et al., 2003).

The absence of a lag phase when microorganisms aregrown on a solid substrate (sulphur), compared withsoluble sulphur (tetrathionate), was attributed to thepresence of a proteinaceous cell-surface appendagethat assisted adhesion (Devasia et al., 1996). Ad-hesion is thought to be promoted both by lipopoly-saccharides as well as cell surface proteins (Amaroet al., 1993; Arredondo et al., 1994).

Organic compounds with sulpho-hydryl groups (e.g.,cysteine) might react with the sulphide surface withthe subsequent release of iron and sulphur species.Bacteria could take advantage of the biochemicalcorrosion process by uptake and oxidation of thereleased species (Rojas-Chapana and Tributsch,2000; Tributsch and Rojas-Chapana, 2000).

Many studies have been focused on the role of theextra-cellular polymeric substances (EPS) which arethought to mediate attachment.

Cells with the EPS layer removed could not attach tocovellite until the EPS layer was regenerated (Poglianiand Donati, 1999).

Bacterial cells adapted the chemical composition of

87rgy 84 (2006) 81108their exo-polymers to the substrate (Gehrke et al., 2001).

etallu The EPS might facilitate the concentration of Fe3+ bycomplexation through uronic acids or other residuesat the mineral surface, resulting in an enhanced oxi-dative attack on the sulphide (Kinzler et al., 2003).

Fe3+ reduction by attached bacteria could be arelevant process involving the EPS layer in thebioleaching of sulphides in aerobic conditions (Pronket al., 1992). The behaviour of attached bacteria isvery dependent on the Fe3+/Fe2+ ratio in the EPSlayer, which is, in turn, very dependent upon theredox potential in solution and the concentration ofsoluble iron (Hansford and Vargas, 2001).

There is also evidence that bacterial attachment ismineral and site specific, and that it results in changedproperties of the mineral surface. It is reported that:

Attachment was rapid and most leaching bacteriagrew attached to mineral sulphide surfaces (Gonzalezet al., 1999; Lizama et al., 2003). More than 80% ofan inoculum was fixed throughout the process ininoculated bioreactors not limited by surface area(DiSpirito et al., 1983; Monroy et al., 1994).

Bacteria attached to small pyrite and chalcopyriteinclusions in low grade ore, rather than to siliceousphases (Murr and Berry, 1976). Acidithiobacillusferrooxidans attached selectively to iron containingsulphides (Ohmura et al., 1993), and cell adsorptiondensity was related to mineral phase (Devasia et al.,1993; Das et al., 1999a; Santhiya et al., 2001).

Bacterial hydrophobicity increased as pH decreased;the degree of adhesion per unit area was greater tosulphide surfaces (hydrophobic) than to quartz surfaces(hydrophilic) (Solari et al., 1992). Sulphides can beseparated from quartz through selective flocculation/dispersion following bio-treatment (Natarajan andDas,2003; Chandraprabha et al., 2004a,b). Rojas-Chapanaet al. (1998) proposed a chemo-tactic bacterial responseto pyrite and to sulphur globules, which might haveevolved as an adaptive advantage.

Bacteria might be influenced by crystallographicorientation when selecting a site for attachment(Berry and Murr, 1977; Ndlovu and Monhemius,2005). The degree of crystallization of synthetic pyritefilms might influence bacterial behaviour (Sanhuezaet al., 1999).

Shape rather than size of a micro-topographic featurecould be the key to attachment of A. caldus grown onpyrite. Edwards and Rutenberg (2001) concluded thatsmall local surface alterations due to bacterial metab-olism could strongly affect local adhesion parameters

88 H.R. Watling / Hydromand account for bacterial adhesion on mineral surfaces.Attachment may also be influenced by solution che-mistry parameters. For example:

Ferrous ion inhibited the attachment of A. ferroox-idans to pyrite and chalcopyrite while ferric ion wasmuch less inhibiting (Ohmura et al., 1993).

Bacteria (A. ferrooxidans) attached to jarositeprecipitates, actively oxidised ferrous ion (Poglianiand Donati, 2000).

Microbial attachment and biofilm formation may pro-vide a mechanism through which the microorganism canlocate itself near an energy source. For example, in the casewhere only a low concentration of ferrous ions is present ina bioleaching solution, the most reliable source of furtherferrous ions will be the sulphide surface e.g. Eq. (1).

Bagdigian and Myerson (1986) demonstrated thatA. ferrooxidans preferentially attached at dislocationsand grain boundaries on pyrite surfaces; proximity tosuch dislocations and boundaries might confer anadvantage. Andrews (1988) suggested that sulphuratoms could diffuse along dislocations in pyrite incoal. Escobar et al. (2004) concluded that sulfuroxidation by S. metallicus grown on chalcopyrite(70 C, pH 1.8) was initiated by attached micro-organisms after a significant lag period, during whichsulphur built up on the mineral surface. Theyestimated that about 68% of the cell population wasattached to chalcopyrite particles and that attachmentoccurred within 100 h of inoculation.

2.3.2. Bioleaching mechanismsThe traditional hypothesis that bacteria oxidise sul-

phides by either a direct mechanism or an indirect mechan-ism (Silverman and Ehrlich, 1964), has evolved into acomplex chemical/electrochemical/biochemical descrip-tion of the interactions of bacteria with sulphide minerals.

Tributsch (2001) proposed that the term contactleaching be used in place of direct leaching because itdescribed the association of bacteria with a surfacerather than the means of attack. This approach has beenrefined further and summarized by Crundwell (2003),who described three mechanisms by which microorgan-isms (specifically A. ferrooxidans) might interact with asulphide mineral, as follows:

i. Bacteria oxidize ferrous ions to ferric ions in thebulk solution, and the ferric ions oxidize thesulphide phase the indirect mechanism.

ii. Bacteria attached to the mineral surface oxidize

rgy 84 (2006) 81108ferrous ions to ferric ions within a biofilm

comprised of bacteria and exo-polymeric material,and the ferric ions generated within this layeroxidize the sulphide phase the indirect contactmechanism.

iii. Bacteria attached to themineral surface oxidize thesulphide phase by biological means directly, with-out any requirement for ferric or ferrous ions the direct contact mechanism.

Importantly, Crundwell noted that the distinguishingfeature of both the direct and indirect mechanisms ofsulphide oxidation was the necessity for ferric ions toparticipate in the mineral dissolution. It is a separateissue as to whether or not bacterial attachment (contact)in itself contributes to enhanced bioleaching by eitherthe direct or indirect mechanism (Fig. 2). Crundwell'sterminology has been adopted in the summary thatfollows.

There is, to date, no evidence that direct contact lea-ching occurs, in which the bacteria break the metalsulphide bond of a mineral.

Attached microorganisms may cause an increasedindirect rate of sulphide dissolution above the

H.R. Watling / Hydrometalluchemical rate of dissolution when the dissolution ofthe sulphide is hindered by the formation of a porous

Fig. 2. Microbial roles in bioleaching of sulphide minerals. Indirectmechanism (A): bacteria oxidise soluble iron(II) to iron(III) andsulphur to sulphate. Ferric ions oxidise the sulphide minerals in anacidic environment. Indirect contact mechanism (B): bacterialattachment is important physiologically, but ferric ions oxidise thesulphide minerals. The specifics of bacterial (electro)chemical

interactions with mineral surfaces and/or their direct contact (enzy-matic) contribution to sulphide dissolution are unknown.sulphur product layer, which is removed by thebacteria.

Attached microorganisms may also cause an increasedrate of sulphide dissolution above the chemical rate ofdissolution in those cases where the dissolution of themineral increases with decreased acidity; the corrosionor mixed potential also decreases in the presence of thebacteria, relative to an abiotic system. Mineral disso-lution is achieved indirectly by ferric ion oxidation.

In both indirect leaching and indirect contact leaching,microorganisms catalyse the oxidation of ferrous ions toferric ions (Crundwell, 2003). The oxidation of a sulphidemineral by ferric ions may proceed via a thiosulphateintermediate or a polysulphide intermediate dependingupon its solubility in acid (Schippers and Sand, 1999).

Acid-insoluble sulphides (e.g. pyrite, molybdenite,tungstenite) are oxidised to metal ions and sulphatevia a thiosulphate intermediate. Bacteria catalyse thethiosulphate oxidation. For the case of pyrite, thethiosulphate is oxidised rapidly to sulphate by ferricions and, therefore, may not be available as a sub-strate for the microorganisms (Hansford and Vargas,2001).

Acid soluble sulphides (e.g. pyrrhotite, sphalerite,chalcopyrite Eq. (2) dissolve to form metal ions andH2S. In this way, the metalsulphide bond in themineral lattice is broken prior to sulphur oxidation.Bacteria catalyse the soluble sulphidepolysulphideoxidation to sulphate, and generate acid. The mecha-nism of acid dissolution of chalcopyrite and theaqueous sulphide oxidation to polysulphide havebeen described (Lazaro and Nicol, 2003; Nicol andLazaro, 2003; Steudel, 1996).

Crundwell and colleagues (Crundwell, 1999a,b;Fowler et al., 1999, 2001; Fowler and Crundwell,1998, 1999; Holmes et al., 1999; Holmes and Crundwell,2000) used a novel apparatus to determine the rates ofbiotic and abiotic leaching of both pyrite and sphaleriteunder strictly controlled solution conditions, specificallytotal soluble iron and controlled potential. Theyconcluded that the increased biotic rate of pyritedissolution was the result of attached bacteria consumingprotons near the mineral surface while oxidising ferrousions Eq. (2). In contrast, analysis of the data obtained forthe leaching of sphalerite at low potential (high ferrousion concentration) indicated that the rate limiting stepwas the diffusion of the ferrous ions away from thesulphide surface through the sulphur product layer that

89rgy 84 (2006) 81108forms on the mineral surface in abiotic leaching but is

etallusubsequently consumed in biotic leaching Eq. (3). Inbioleaching, bacteria clean the mineral surface ofsulphur, allowing the mineral to be more rapidlyoxidized by ferric ions and ferrous ions to escape thesurface into the bulk solution phase. In both cases, whilebacterial attachment results in enhanced leaching,mineral dissolution is the result of ferric ion attack onthe sulphide particles.

In their study of the bioleaching of a sphalerite con-centrate, Rodriguez et al. (2003a,b) concluded that there aretwo steps to sphalerite dissolution. In the first step, the rapidattachment ofmicroorganisms to active pyrite surfaces leadto the oxidation of the pyrite and concomitant bio-gene-ration of ferric ions and protons. This was the key toenhanced sphalerite leaching. In the second step, the keycontributions in sphalerite dissolution were the continuedregeneration of ferric ions by planktonic bacteria and theoxidation of the elemental sulphur reaction product. Theseauthors termed this cooperative bioleaching and suggestedthat there is a direct relationship between the magnitude ofcell attachment in the first step of the process; and thedissolution rate in the second step.

2.3.3. Molecular chemistry and geneticsAcclimation of bacteria to different conditions, such as

increased tolerance to highmetal levels, is a simplemeansof genetic improvement. The technique is dependent uponthe small number of errors in the DNA sequence that aremade during chromosomal replication. Most errors areharmful or neutral but some may be advantageous. Thus,when a selective pressure is applied to a population, thosebacteria that acquire an advantageous mutation will out-perform the rest and dominate the population. Forexample, by growing bacteria in a continuous flow-reactor under conditions of increasing flow rate, fastgrowing bacteria will be enriched while slow growingbacteria will be washed out.

The advantage of mutation and selection is that it canbe applied in the laboratory without requiring specialisedknowledge of bacterial physiology and biochemistry. Thedisadvantage is that it is a slow process and yields littleinformation about the bacterial population itself. Never-theless, A. ferrooxidans leach rates have been improvedten-fold, making possible the commercialisation of bio-oxidation (van Aswegen et al., 1988).

Studies on the genome and genetic engineering ofmineral processing bacteria are in their infancy andmainly relate to A. ferrooxidans (Barreto et al., 2003;Holmes et al., 2001; Leduc and Ferroni, 1994; Quatriniet al., 2004; Rawlings, 1999, 2001; Rawlings andKusano,1994; Valdes et al., 2004) This is not surprising.

90 H.R. Watling / HydromAcidithiobacillus ferrooxidans is the best characterisedand fastest growing of the iron- and sulphur-oxidisersand, fortuitously, the structures of many of its genes havebeen found to resemble those of Escherichia coli (a well-characterised bacterium of physiological importance).Genomes of other acidophiles such as Picrophilusctorridus (Futterer et al., 2004), Ferroplasma acidarma-nus (Tyson et al., 2004) and Sulfolobus acidocaldarius(Chen et al., 2005) have been sequenced more recently.

A number of studies using A. caldus (de Groot et al.,2003 and references therein) or L. ferrooxidans (Coramand Rawlings, 2002; Coram et al., 2005) have beenreported. Recent progress on the molecular genetics ofSulfolobus spp has also been reviewed (Zillig et al.,1998; Ciaramella et al., 2002). The role of Ferroplasmaacidarmanus in bioleaching is as yet unknown but thisstrain is dominant in mine water at pHb1.5 and thusimplicated in acid rock drainage (Barreto et al., 2003;Holmes et al., 2001).

Currently, the genetic engineering of mineral proces-sing bacteria tends not to have a high priority amongmineral bioleaching researchers. There are a number ofreasons for this:

i there is a great deal of uncertainty about regulatoryissues concerning the release of genetically engi-neered strains into the environment;

ii there is some doubt that engineered strains would besufficiently robust to survive and compete effectivelyin the complex environment of a mineral processingoperation;

iii before strain improvement by genetic engineeringcan be contemplated, it will be necessary to establishthose attributes that provide the organisms with theircompetitive edge and, from that, determine whichgenes to target;

iv there is the generally held view that indigenousbacteria growing in a particular environment arelikely to be those best adapted to that environment.

Natural, but sometimes extreme environments, havebeen the source of all bacterial strains used in mineralprocessing thus far (Rawlings, 1997; Norris, 1997;Hallberg and Johnson, 2001).

3. Copper sulphide bioleaching

The majority of sulphide bioleaching research studieshave been carried out with one of two main goals. Thefirst has been to assess and/or compare the capability ofbacterial strains to oxidise different minerals. Related tothese studies, are those concerned with quantifying

rgy 84 (2006) 81108bacterial tolerance to particular leaching environments

etallu(e.g. Franzmann et al., 2005) and/or with adaptingbacteria to changed conditions. Increasing bacterialtolerance to high concentrations of heavy metals is atopic that has also received close attention (e.g. Daset al., 1997). The mechanisms of metal resistance inacidophilic microorganisms have been reviewed byDopson et al. (2003).

The second goal has been to test the possibility ofutilising bioleaching/bio-oxidation to process differentores and concentrates, particularly those associated witha particular ore body or mine. While the results of suchstudies are seldom directly comparable, because mostores contain a variety of copper minerals, some gene-ralisations can be made.

In those locations where leaching is a secondaryoperation with the purpose of processing tailings, low-grade ore or material unsuited to concentration, then thepreferred approach for processing, based on the numberof operations are ranked:

Minerals: copper oxidesN secondary sulphidesNchalcopyrite

Methods: dumpsNheapsN in situ

The efficiency of leaching depends strongly upon theminerals that make up the ore (Table 3). For example,whereas tenorite [CuO], cuprite [Cu2O], malachite [Cu2(CO3)(OH)2] and chrysocolla [CuSiO32H2O] mightrequire only hours of leaching, chalcocite [Cu2S] andcovellite [CuS] require months of heap leaching andchalcopyrite would require years of dump leaching.Chalcopyrite leaches at about one fifth the rate ofchalcocite (Rear et al., 1994). While grinding of the oreto a smaller particle size very often increases copperrecovery in a given time (Rhodes et al., 1998), thebenefit conferred by the treatment will be offset byincreased acid and energy consumption. In addition, fineparticles in a heap can decrease the permeability of theheap to both air and solution.

The bioprocessing of whole ores presents differentchallenges to those experienced in stirred tank technol-ogies. For example, during the dump and heap leaching ofwhole ores, leaching progress can be monitored by chem-ical and physical means, but only minimal control can beexercised over the conditions within the ore in situ and thereactions occurring there. The most widely practicedexample is the heap leaching of chalcocite ores (Table 1).

3.1. Whole ore heap (and dump) bioleaching

Thin-layer leaching exemplifies heap leaching in that

H.R. Watling / Hydromit is the foundation of so many of the current heap leachoperations. Originally thin layer leaching was developedto process copper oxides, but it was subsequentlymodified to accommodate an increased contributionfrom sulphide minerals in the ore (Montealegre et al.,1995). The key features are (i) curing, in whichconcentrated acid is added to the crushed ore and reactswith acid soluble and gangue minerals, and (ii)consolidation of fines with larger particles in a rotatingdrum, which ensures even acidification and wetting ofthe ore prior to stacking. In those cases where organicfree raffinate is used for curing/acidification, bacteriamay also be distributed through the ore before it isstacked. The ore is stacked in carefully designed heapsso as to create a bed that exhibits good permeability forboth solutions and gases. The bed is irrigated fromsprinklers or drippers arranged in a grid across the top ofthe heap. Typically, a short-duration acid leach (18 days)recovers 80% of copper from oxides, and between 30and 40% of copper from bornite and chalcocite. Thetransition from a heap leach to a bioleach is accom-plished by constructing heaps of the acid-leached,crushed ore (100% less than 6 mm particle size) stackedto 56 m high. Solution pH is controlled to 1.8 tooptimize bacterial activity. Leach effluents typicallycontains 23 g/L total soluble iron, 2030 g/L sulphateand 106 bacterial cells/mL. Temperatures inside theheap are in the range 1227 C. Leach duration ismarkedly longer than that required for the oxides, andtypically ranges from 150 to 210 days for 7580%recovery of copper from bornite and chalcocite.

The success of many recent bioleaching operations,mainly chalcocite (Table 1) is testament to this robusttechnology, which has been successfully modified toaccommodate ore characteristics peculiar to differentmineral deposits. Some of those modifications havebeen directed towards optimizing heaps as bioreactors toincrease copper production (Fig. 3).

Sulphur- and iron-oxidising microorganisms colonisesulphide ores naturally, and are well known contributors tothe process of acid rock drainage (Colmer et al., 1950;Temple and Colmer, 1952). It is therefore surprising thatforced aeration of heaps to optimize bacterial activity bymeeting their physiological requirements for O2 and CO2was only implemented in the mid 1990s, concurrently inChile and Australia (Readett and Sylwestrzak, 2002). InChile, the calculated required amount of air was deliveredto the heap, whereas in Australia excess air was appliedby means of low-pressure blowers. The significantlyimproved recovery of copper from an aerated chalcocitetest heap at Girilambone, NSWprompted the strategy of re-mining and aeration of under-performing heaps (Lancaster

91rgy 84 (2006) 81108and Walsh, 1997). This was accomplished by using an

Table 3Dissolution and oxidation reactions of some copper-containing minerals in h

Mineral Leaching and oxidation reactions

Atacamite Cu2ClOH3 3H2Cu2 Cl 3H2OChrysocolla CuSiO3:2H2O 2HCu2 SiO2:3H2ONeotocite Cu;Mn2H2Si2O5OH4:nH2O 4HCu2 Tenorite CuO H2SO4CuSO4 H2OMalachite Cu2CO3OH2 2H2SO42CuSO4 CO2 3Azurite Cu3CO32OH2 3H2SO43CuSO4 2CO2 Brochantite Cu4SO4OH6 6HCuSO4 3Cu2 6H2

H2O

O4

inera

2CuS

92 H.R. Watling / HydrometalluNative copper Cu 1=2O2 H2SO4CuSO4 H2OCuprite Cu2O 1=2O2 2H2SO42CuSO4 2H2OChalcocite Cu2S 1=2O2 H2SO4CuS CuSO4

Cu2S Fe2SO43CuS CuSO4 2FeS CuS is a reaction product rather than the m

Bornite Cu5FeS4 2Fe2SO432CuS CuFeS2 Covelliteexcavator to turn the heapwhilst aeration pipingwas placedin the slot where the ore was turned. The heaps werebrought back on line with drippers and air was introducedto the heap from both sides via a vent bag. The additionalcopper extracted in the short term by this strategy amplycompensated the cost of re-mining.

Whenheaps are stacked and irrigated, there can be a lagtime before bacterial growth and metabolism contributes

CuS 2O2CuSO4CuS Fe2SO43CuSO4 2FeSO4 So

Enargite Cu3AsS4 41=2Fe2SO43 2H2O3CuSO4 Chalcopyrite CuFeS2 O2 2H2SO4CuSO4 FeSO4 2S

Fig. 3. Simplified heap leach schematic showing solution managementand heap aeration to promote bioleaching of sulphide minerals. Solu-tion management is aimed at minimizing water use and maximizingPLS to the SX circuit.eaps and dumps

Duration

Hours to days

Mn2 4SiO2 6 nH2O

H2O

4H2O

O

Days to months

l covellite

O4 4FeSO4 Months to years

rgy 84 (2006) 81108to sulphide oxidation. Lag times can be shortened byutilising heap recycle solutions (e.g. raffinate from SX)that contain indigenous microbial populations alreadyacclimated to the leaching conditions, for the acid con-ditioning step, thus ensuring that an active bacterial popu-lation is distributed throughout the ore. Heap irrigationutilizing similarly populated recycle solutions maintainsthe population for the duration of leaching and serves toinoculate subsequent heaps. The patented BIOPROprocess, which involves heap inoculation during acidconditioning and agglomeration, (Brierley, 1997) and hasbeen implemented for the biooxidation of refractory goldores prior to gold extraction, claims rights over an ino-culation strategy that has been widely practiced becausewater recycle is cost effective, if not critical, in heapleaching operations.

Dump leaching and in situ leaching can be thought ofas sub-sets of heap leaching in that solutions percolatethrough a bed of ore and copper is recovered from theleachate. Typically, dump leaching, with minimal orepreparation, is used to extract copper from low-grade(run-of-mine) ore (0.10.5% Cu). Dump leaching isrelatively inefficient because large sized particles

9FeSO4 4So HAsO2 11=2H2SO4o 2H2O Years

etallupresent small surface areas to lixiviants whilst smallparticles block solution flows and impede aeration.Furthermore, haulage vehicles and time cause the ore tocompact, making it even less permeable. Modern dumppractices that minimise compaction and maximiseaeration and solution permeability, can improve copperrecoveries. Dump material may be pre-conditioned withhigh acid levels, but the value of this strategy dependsupon the mineralisation and on the acid consumption ofgangue minerals. Variable irrigation programs may beemployed as the permeability of the dump decreaseswith age. The enormous tonnages and the fact that thecost of removal is already borne by the mine make iteconomical to scavenge copper from low-grade (0.10.5% Cu) dumps (Schnell, 1997). Studies on dumpbiodiversity and biodynamics are sparse (e.g. Edwardset al., 1999) but it can be assumed that bacteria canreadily colonise exposed mineral sulphide surfaces andengage in iron and sulphur oxidation.

There is only limited evidence that bacteria play a rolein the in situ processing of undisturbed copper ores. Forexample, solutions circulating through a chalcocite/chalcopyrite/pyrite ore at La Hermosa (part of theAndacollo mining district) were found to contain106 cells/mL (Concha et al., 1991). Leptospirillumferrooxidans is thought to play a role in the leaching of achalcocite and bornite ore body below an old open-pitmine, relatively near the surface (Concha et al., 1991).However, while solutions injected into deep in situ orebodies may contain bacteria, it is unlikely that thebacteria will be particularly active at depth under con-ditions of oxygen limitation. That is not to say that an insitu leach of the future cannot be engineered to meetbacterial physiological requirements.

3.2. Modelling applied to heaps or dumps

Sulphide heaps are complex but poorly understoodbioreactors. To a degree, the lack of understanding hasarisen because many investigators have focused onspecific sub processes in isolation, such as the che-mistry, the microbiology or the hydrodynamics, andhave failed to account for the interactions between thoseprocesses. In order to realize fully the potential of heapbioleaching, a holistic model is required that accountsfor as many as possible of the complex micro- andmacro-scale processes and their interactions.

There is currently a considerable effort being made todevelop predictivemodels of a generic nature to facilitateheap design and control. Much of the early research hasbeen summarized by Ritchie (1997). Recent studies in

H.R. Watling / Hydromrespect of both heap leaching and acid rock drainage(Casas et al., 1998; Bouffard and Dixon, 2001; Lefebvreand colleagues, 2001a,b) are testament to the growinginterest in models as tools to improve heap design andmanagement.

A critical assessment of the significant, public-domain,heap leach models has been undertaken by Dixon (2003).The review cites 47 references describing developmentsover the last 30 years. Fifteen models including heap anddump leaching (five models), heap and dump bioleaching(eight models) and acid rock drainage (two models) arecompared in terms of the extent to which each representsthe most important macro-scale transport, chemical andkinetic, and biological phenomena involved in heapleaching. Dixon uses an extensive matrix of heap sub-processes in his collation. Table 4 presents a simplifiedmatrix summarising the capabilities of the heap bioleach-ing models reviewed by Dixon (2003), augmented withsome more recent work. As was the case in Dixon'sreview, distinction ismade between those phenomena thatare accounted for explicitly, and are well documented, andthose that are accounted for only implicitly or in a cursoryfashion or are poorly documented.

Ritchie and colleagues (Pantelis et al., 2002 andreferences therein) describe the development of a modelinitially focused on dump leaching and sulphidic minewaste stockpiles. In respect of heap leach phenomena,there seem to be limitations in respect of boundary phe-nomena, reaction networks, gangue mineral reactions andinteractions, andmicrobial growth and oxidation, as is alsothe case for the copper dump model (Casas et al., 1998).

The point of departure for the development of theHeapSim model (Dixon and Petersen, 2003, 2004) hasbeen a systematic conceptual and mathematical descrip-tion of many of the processes relevant in heap leachingdrawn together within a comprehensive simulationengine. A relatively large set of parameters are requiredby the model. Many can be taken from the literature butsome must be measured, usually through targeted labo-ratory-scale column tests. The dual experimental andmodelling approach has been applied in a number ofheap bioleach contexts, including high-grade zinc sul-fide (marmatite) ores, low grade chalcocite ores andsupported copper concentrates. HeapSim is currentlybeing revised and refined with the focus on the heapbioleaching of chalcopyrite.

The key differentiator of the multi-dimensional mod-els Phelps Dodge (PD) copper stockpile model(Bennett et al., 2003a,b) and CSIRO Heap model(Leahy et al., 2005, submitted for publication) is theability to simulate spatial structures in heaps. The PDmodel can simulate different conditions within individ-

93rgy 84 (2006) 81108ual lifts at various locations within the canyon

etallutopography. The CSIRO model has been used toinvestigate the local structure related to individual airspargers and acid drips. These two models also have thecapability of simulating two phases (gas and liquid)simultaneously, conferring greater predictive modelpower.

Bennett et al. (2003a,b, 2006) described the designand development of a comprehensive model for a copperstockpile leach process based on the computational fluiddynamic software framework PHYSICA+. The authorsreport that the model accounts for transport phenomenathrough the stockpile, reaction kinetics for the importantmineral species and bacterial affects on the leach reac-tions plus heat, energy and acid balances for the overallleach process. However, in comparison with others(Table 4), boundary phenomena, ganguemineral reactionsand interactions, and microbial growth and oxidation areeither dealt with in cursory fashion or poorly documented.

The CSIRO model is built on the computational fluiddynamics CFX platform (Leahy et al., 2005 andreferences therein). Like HeapSim, it incorporates amathematical description of many of the important heapphenomena. It is perhaps less robust in accounting forboundary phenomena and, like all its predecessors, thusfar fails to address gangue mineral reactions and inter-actions explicitly.

4. Chalcopyrite bioleaching

Not surprisingly nearly all the research on chalcopyriteleaching and bioleaching has been undertaken withconcentrates and, therefore, not all of the examples givenbelow are directly relevant to the heap leaching of lowgrade ores. Nevertheless, it is worth briefly summarizingsome of the peculiarities of chalcopyrite leaching as ameans of determining what controls may be desirable in aheap leach environment. Some of the detail is specific toagitated tank leaching processes, for example BioCOP(Clark et al., 2005) or the BacTech-Mintek process(ANON, 2001). The key points that do carry across toheaps are:

Chalcopyrite (or the target sulphide) surface must beexposed to the leach solution, particularly the oxidant(iron(III)). Thus appropriate ore preparation and ironchemistry is important.

Microbial catalysts are needed to regenerate theoxidant and acid and to enhance leaching rates. Thusconditions conducive to growth are important.

Chalcopyrite leachesmore rapidly at higher temperatures. Chalcopyrite may leach more rapidly at lower redox

94 H.R. Watling / Hydrompotentials. Silver catalyses chalcopyrite dissolution, but is not aneconomic additive. No other effective catalyst hasbeen identified.

Chloride enhances chalcopyrite leaching.

4.1. Fine grinding

The slow and incomplete extraction of copper fromchalcopyrite concentrates and ores can be overcomewhen the ore/concentrate is finely ground (p80b15 m)(Rhodes et al., 1998; van Staden, 1998). Small volumeparticles combined with a high surface area promotedrapid copper extraction. By the time the inhibitinglayer had developed to the point of hindering furtherreaction, copper extraction was all but complete.However, care must be taken not to damage cellsphysically during agitation (e.g., Deveci, 2002, 2004).

Moderate leach temperatures and fine grinding, madeeconomic by a very favourable power cost, combined tomake the recent pilot trial at Mt Lyell a success (Rhodeset al., 1998). This process utilised moderate thermo-philes (4547 C) and pH controlled between pH 0.82.0; the residence time was about 56 days for 96%copper extraction. When operating in the thermophilictemperature range, fine grinding may not be necessary.Greater than 95% copper extraction from a chalcopyrite/pyrite concentrate was achieved in a three-stage pilotplant (Gericke and Pinches, 1999; Gericke et al., 2001).The key parameters were delivery of sufficient oxygenand carbon dioxide for bacterial growth and oxidation,and accommodation of an increased microbial sensitiv-ity to solids concentration and particle size distribution.The process has been operated at pilot and demonstra-tion scales with a view to commercialization (Rhodeset al., 1998; ANON, 2005).

4.2. Heap inoculation

Copper bioleaching operations typically rely uponnatural colonisation by indigenous microbial strains well-acclimatised to the ore. However, with the continued pushto achieve the desired metal extractions in ever shortertimes, the need arises to not only promote microbialcolonisation throughout the heap efficiently and in a shorttime period but also to maintain a population appropriateto the conditions that become established in the heap as afunction of time. Further, for chalcopyrite, heaps mayneed to be operated at increased temperature to achievethe necessary copper extraction rate, in which caseinoculation with a consortium of microorganisms thatgrow well at different temperatures would be required to

rgy 84 (2006) 81108overcome the delay in natural colonisation. In the case of a

Table 4Comparison of factors considered in various bioleach models in terms of heap leach phenomena

Model A B C D E F G H I

Hydrology and solute transportHeap wetting Advection Diffusion Multidimensional solute transport

Gas transportAdvection (forced aeration) Convection (natural aeration) Diffusion

Heat transportAdvection and conduction In situ evaporation/condensation

Boundary phenomenaNon-uniform solution application Surface evap'n, radiation, convection Heap base evaporative cooling

Operator-induced phenomenaSolution stacking/change Rest/rinse cycles

Leaching kineticsLeaching reaction kinetics Pore diffusion kinetics Liquidsolid mass transfer

Reaction networksLeaching in series Homogeneous reaction kinetics Gasliquid mass transfer

Multiple minerals leaching in parallel

Component sources/sinksGangue interactions Precipitation/adsorption/solution Solution speciation

Non-linear effectsChanges in pellet porosity/liberation

Microbial growth and oxidationMicrobial speciation by T and function Growth and oxidation kinetics Temperature factors Substrate limitations/inhibitory factors Cell yield Maintenance/decay/death

Microbial transportMicrobial diffusion/mobility Microbial attachment/adsorption Inoculation (continuous, periodic)

Adapted from Dixon (2003) and augmented.

95H.R. Watling / Hydrometallurgy 84 (2006) 81108

about 55 C (Fig. 4). Greater copper extraction is achievedbefore leaching rates are slowed by the formation ofinhibitory layers on the mineral surface.

The exploitation of thermophiles in the bioleachingof base metal concentrates represents a breakthroughdevelopment for the bioleaching of chalcopyrite. Inbatch bioleaching tests using thermophiles and tem-peratures 68 and 78 C, a relatively-long residence timeof 30 days yielded 9597% copper extraction fromchalcopyrite (Dew et al., 1999). Process improvementshave been achieved by utilising oxygen-enriched air tosparge the bioreactors, thus overcoming oxygen limita-tion due to the low solubility of oxygen in aqueoussolutions at high temperatures. The BioCOP processcontinues to be improved and tested with a view tocommercialization, for example the Alliance Copperdemonstration plant (Clark et al., 2005).

There is no doubt that heaps of quite low gradesulfidic material can be self-sustaining at above ambienttemperatures, but this may not be achieved in practice.Norton and Crundwell (2004) cite insufficient and/orinefficient aeration and the implementation of constantirrigation rates, carried over from heap leaching of

etallurgy 84 (2006) 81108chalcocite (minor chalcopyrite) test heap that was notspecifically inoculated, cell numbers indicative of afunctioning population, includingmoderate thermophiles,were only observed at about three months (Readett et al.,2003). In the same test heap, thermophiles were not de-tected, even though the heap temperatures rose and re-mained quite high for a significant period.

Examination of the literature shows that a numberof heap inoculation methods have been proposed,including:

Heap irrigation from ponds in which microbialgrowth has been enhanced through the addition ofnutrients (P, K, N) to the iron-rich solution. Ponds aresometimes covered and aerated.

Formation of agglomerates from ore particles usingan inoculum containing iron and sulphur oxidisingorganisms and then stacking the ore (and variations)(Brierley and Hill, 1993).

Inoculating a biological contactor with iron oxidisingstrains and feeding the ferric-rich stream (containingsomemicroorganismswashed out) to the heap (Hunter,2001).

Creating a culture of microorganisms on part of theore and mixing this ore with the bulk ore beingstacked a means of distributing a viable culturealready attached to the ore throughout the heap(King, 2001). In an apparently similar invention, themicroorganisms are first adapted to an ore and thenscaled up to the required volume for heap inoculation(Hunter and Williams, 2002).

Inoculatingwith a tailored bacterial culture that growsparticularly well on a specific mineral (e.g. chalco-pyrite) and feeding the culture from a pond to the heap(Hunter, 2002c). A similar application could be heapinoculation with tailored microbial cultures that growwell under the targeted heap conditions such as aboveambient temperatures, increased acidity, or thepresence of organic compounds originating from theore or the downstream processing.

Microbial inoculation with strains that have been ren-dered temporarily non-adhesive so that they permeate togreater depths in the heap when delivered via irrigationto the heap surface (StickiBugs) (Gericke et al., 2005).

Delivery of ultra-small microorganisms via a gaseoussuspension into the heap using the aeration lines (duPlessis, 2003).

4.3. Heap leach at above ambient temperatures

It is well known that the bioleaching of chalcopyrite

96 H.R. Watling / Hydromproceeds more rapidly if the temperature is raised aboveoxides, as key factors why heaps seldom heat to thetemperatures required to recover copper from chalcopy-rite. A number of methods have been described in respectof generating and then maintaining heat in a heap.

Fig. 4. Thermophilic sulphur- and iron-oxidising microorganisms arerequired in order to take advantage of increased chalcopyrite leachingrates. Tests used a chalcopyrite concentrate of composition 64%chalcopyrite, 6.6% pyrite, 3.3% pyrrhotite and 25% quartz, mean

2particle size 27 um, P80-66 um and surface area 0.36 m /g (5-pointBET analysis).

etallu Acid conditioning of ore using hot water to addheat to the heap and promote more rapid oxidation(Schnell, 1997).

Heating the heap (Kohr et al., 2000). Increasing the heat-generating capacity of the heapwith, for example, additional sulfides (Kohr et al.,2002).

The use of covers to limit evaporation and/or insulatethe heap (Schnell, 1997).

Heat transfer to heap irrigation solution via a flexible,heat-absorbing distribution mat with parallel spacedemitter tubes (solar heating) (Lane, 2000).

Heat transfer to heap leach solution from anotherprocess (Batty and Norton, 2003; Schnell, 1997).

Aeration with humidity and temperature control tomaintain heap environment suitable for thermophiles(Winby and Miller, 2000).

Managed aeration and irrigation to maximise heatconservation in the heap (Norton and Crundwell,2004).

Addition of carbon (carbonate, carbon dioxide,organic carbon) to the heap to enhance microbialgrowth and activity at temperatures above 60 C (duPlessis and de Kock, 2005).

Combinations of these management and controlmeasures should deliver and maintain heap temperaturesin a range that favours the necessary microbial oxida-tion and promotes more rapid release of copper fromchalcopyrite.

4.4. Silver catalysis

The acceleration of chalcopyrite dissolution by thepresence of some catalytic ions, particularly silver, iswell known (Hu et al., 2002). Silver ions react with thechalcopyrite surface to form Ag2S which modifies theanodic dissolution of chalcopyrite and inhibits passiv-ation. The problem has been to develop a means ofrecovering the silver added to a chalcopyrite dissolutionprocess from the solids with which the silver reports. Forexample, silver is readily locked up as a silver jarositeand does not report to the solution. The IBES and BRISAtwo-stage processes for the treatment of a chalcopyriteconcentrate (Romero et al., 2003; Carranza et al., 2004)involve the optimisation of both the biooxidation offerrous ions and the chemical oxidation of the sulphidesin separate reactors the latter at higher temperaturesconducive to chalcopyrite leaching. Silver, added as acatalyst to the chemical leach reactor, significantly en-hanced copper extraction from the chalcopyrite. Silver

H.R. Watling / Hydromion toxicity to the microbial population was avoided bythe use of a two-stage process. The efficient recovery ofadded silver, by brine treatment of the leach residues tosolubilise silver as a chloro-complex, is an important partof the process economics. No doubt, this would alsorequire transformation of jarosite into other iron oxides.

Canfell et al. (1998) propose a process for the bio-leaching of chalcopyrite ores that comprises treating theore with silver during acid conditioning to distribute thesilver through the ore, and bioleaching the ore to oxidiseferrous to ferric ion.

4.5. Controlled Eh

The effect of redox potential on the leaching ofchalcopyrite has been investigated extensively. RecentlyHiroyoshi et al. (1997, 2000, 2001) reported that thedissolution rate of chalcopyrite is higher at redox poten-tials of leach solutions below a critical value. They havesince proposed a practical method to control the redoxpotential during the leaching of low grade copper ore(Okamoto et al., 2003).

Earlier, Bruynesteyn et al. (1986) had taken advan-tage of the increased reaction rates afforded by a silvercatalysed bioleach of chalcopyrite. They achievedfurther enhancement of the copper extraction by control-ling the oxidation potential of the slurry between 0.54and 0.66 V SHE with the addition of thiosulphate andsoluble copper added as copper sulphate. The processwas applied to the leaching of a chalcopyrite concentrateas a means of oxidising sulphide stoichiometrically toelemental sulphur at ambient temperature and pressure(Lawrence et al., 1985). While this particular processwas not developed to commercial scale, redox control isan inherent part of some more modern processes (e.g.Dixon and Tshilombo, 2005), and can be achieved byprescribing the total iron concentration in bioleachsolutions and/or controlling the ferric/ferrous ion ratio orpotential. The two-stage bioprocesses such as IBES(Carranza et al., 1997) are particularly amenable to thiskind of control. Redox-controlled bioleaching of chalco-pyrite, either as concentrate in agitated tank reactors oras ore in heaps, is a key claim of recent patents (van derMerwe et al., 1998; Gericke and Pinches, 1999; Pincheset al., 2001).

4.6. Iron chemistry

The acidic dissolution of sulphide minerals requires anoxidant that, in heap leaching, is typically ferric ion Eq.(1). Once a sufficient iron concentration has built up in theleachate (Readett et al., 2003), the natural bacterial

97rgy 84 (2006) 81108population regenerates the ferric ions reduced by reaction

phatechloride media and found that the presence ofchloride enhanced chalcopyrite oxidation. Examinationof leached chalcopyrite surfaces showed that the sulphurformed in the presence of chloride was crystalline andporous, while that deposited in the absence of chlorideformed an amorphous or cryptocrystalline film of sul-phur (Fig. 6). Given that ground- or bore-water oftencontains significant concentrations of chloride, and thatsome copper ores contain atacamite, it is possible thatthe chloride effect could be exploited in a heap leachsituation, if the microbial catalysts continued to growwell.

The impact of chloride on the growth of a number ofbioleaching strains has been reported, for example,A. ferrooxidans (Alexander et al., 1987; Lawson et al.,1995), L. ferriphilum (Kinnunen and Puhakka, 2004), S.metallicus (Huber and Stetter, 1991), S. rivotincti (Gomezet al., 1999) and a mixed mesophilic culture (Shiers et al.,2005). In summary, the results indicated that the presence

etalluwith the sulphide, thus sustaining dissolution over longperiods of time. At the same time, the acidic dissolution ofgangue minerals will result in increased solution pHwithin the heap. Under the conditions of high pH and highferric ion concentrations, iron hydroxy compounds, mostcommonly jarosites [MFe3(SO4)2(OH)6], will precipitateEq. (5). The immediate consequence is to diminish theconcentration of ferric ions available for further sulphideoxidation, but this is unlikely to be a limiting factor for alow-grade sulphide ore. A second consequence is that ironprecipitation could contribute to the production of fineswithin the heap, possibly impacting heap permeability.Anecdotal evidence suggests that the contribution fromiron precipitation is a minor contributor. However, ironprecipitation preferentially on the target sulphide surfaceshas the potential to impact significantly on leaching rates.

The results of a systematic study of the acidic oxi-dative dissolution of chalcopyrite surfaces indicatedthat, while a rapidly formed copper-deficient sulphide ordisulphide played a key role in the leaching mechanism,subsequent reactions of a likely thiosulphate intermedi-ate and ferric ion occurred in the immediate vicinity ofthe chalcopyrite surface, leading to the formation of aferric sulphate phase (Parker et al., 2004). This sulphateis thought to be the precursor for the eventual massdeposition of jarosite on the sulphide surface. Theprecipitate forms a coherent layer that adheres stronglyto the chalcopyrite surface and cannot be removedcompletely by bio-reduction (Fig. 5). It constitutes aphysical barrier that prevents microbial access and slowsthe diffusion of ferric ions to the sulphide surface and ofreaction products away from the sulphide surface (Boonand Heijnen, 1993; Stott et al., 2000).

Kohr et al. (1997), recognising the importance of ironcontrol, proposed a method of precipitating some ironfrom bioleach solutions prior to recycle to the heap.Marsden et al. (2002) described a method of seededprecipitation for iron control. However, such reactionsare slow and difficult to control.

4.7. Chloride effect

The leaching of chalcopyrite in chloride media hasbeen reviewed and advanced by Dutrizac (1990, 1992).Early processes were focused on the leaching ofconcentrates under aggressive conditions, for examplethe CLEAR process and the INTEC process (Lunt et al.,1997). In general, leaching rates are more rapid inchloride media than in sulphate media. In their work onthe dissolution of chalcopyrite under dump leachingconditions, Dutrizac and MacDonald (1971) investigat-

98 H.R. Watling / Hydromed conditions under which modest salt addition (6 g/L)would accelerate leaching in low-grade ores. The oresthey tested gave variable responses to leaching, with orwithout added sodium chloride. They concluded thatsome ores did respond to chloride addition, but that asignificant acceleration was only achieved at tempera-tures above 50 C. At lower temperatures (25 C) thepresence of chloride might inhibit chalcopyrite dissolu-tion. Kinnunen and Puhakka (2004) reported similarresults; the presence of 0.25 g /L of Cl concentrateenhanced copper yield at temperatures between 67 and90 C but decreased copper yield at 50 C.

Lu et al. (2000a,b) focused their electrochemicalstudies on the leaching of chalcopyrite in mixed sul-

Fig. 5. Massive chalcopyrite cleaved after 24 h exposure to ferricsulphate solution shows a thick, coherent layer of jarosite closelyadhered to the exposed surface (reprinted with permission from Stottet al., 2000).

rgy 84 (2006) 81108of chloride is detrimental tomicrobial growth, but that the

etalluH.R. Watling / Hydromdegree of inhibition varies between strains. In addition,Shiers et al. (2005) showed that concentrations of 7 g/LNaCl reduced cell replication by 50% and that nosignificant culture adaptation or habituation occurredwith prolonged exposure to that concentration.

The question remains as to whether the benefits ofchloride enhanced chalcopyrite dissolution outweigh theinhibition of the microbial population in a heap of low-grade sulphide ore. The downstream impact of a mixedsulphatechloride solution on solvent extraction mustalso be considered.

5. Enargite bioleaching

Enargite is a copper arsenic sulphide (Cu3AsS4) thatoccurs frequently with chalcocite. It does not easily

Fig. 6. SEMmicrographs showing a residues particle leached (a) in thepresence of NaCl and (b) in its absence. The concentrate in theseexperiments had a d50 of 15.1 m (Lu et al., 2000b, reprinted withpermission).separate from the other sulphide minerals during flota-tion concentration, and thus the resulting high-arsenicconcentrates cannot be roasted. Heap leaching of en-argite-rich copper ores may be a process option.

Arsenic toxicity to bioleaching microorganisms is welldocumented, often in respect of the biooxidation ofpyritearsenopyrite refractory gold concentrates (e.g.,Breed et al., 1996). Barrett et al. (1989) reported that anarsenic-acclimated culture of moderate thermophilesgrew vigorously at up to 15 g/L of arsenic in a bio-oxidation process but that if As(III) production Eq. (6)exceededAs(III) oxidation toAs(V) by ferric ions Eq. (7),then microbial growth rates diminished. In addition theynoted that the onset of biotoxicity could be sudden andwas accompanied by an increase in pH of the medium.Arsenic in either oxidation state is toxic to bacteria, but As(III) is more toxic than As(V) (Silver et al., 1981). Themechanisms of toxicity are not the same, as evidenced bymeasurements of the maximum oxygen utilization rate,and of soluble arsenic concentrations and solution spe-ciation; relative levels of toxicitymay be influenced by theenergy source (Breed et al., 1996).

4FeAsS 11O2 2H2O4Fe2 4SO24 4HAsO2 6

Fe2SO43 HAsO2 2H2O2FeSO4 H3AsO4 H2SO4 7

In a biooxidation reactor, arsenic(III) oxidation byferric ions occurs in the presence of excess ferric ions.Under such conditions, the precipitation of ferric arsenateis favoured Eq. (8), thus lowering soluble arsenic levels tobelow the bacterial toxicity threshold (Escobar et al.,1997a; Wiertz et al., 1997). The same condition of excessferric ions and consequent control of soluble As(III)may also prevail in a heap leach system.

H3AsO4 Fe3 nH2OFeAsO4d H2O 3H 8

Enargite is known to be refractory, but data on itsleaching/bioleaching are sparse (Dutrizac and MacDonald,1972; Ehrlich, 1964; Hao et al., 1972; Wiertz et al., 1997;1998; Escobar et al., 1997a,b). Only 24% copper recoverywas achieved during the bioleaching of an enargiteconcentrate using mesophilic organisms. Some improve-ment was achieved when the air used to sparge the reactorswas supplemented with carbon dioxide (Acevedo et al.,1997, 1998). Biooxidation tests in a continuous reactorshowed that bacterial attack was directed towards the pyritein an enargite rich concentrate (Canales et al., 2002).

99rgy 84 (2006) 81108Neither enargite nor luzonite (its dimorph) were oxidized

etalluduring the bioleaching at 25 C and 50 C of a copperconcentrate that also contained chalcocite, covellite andpyrite (Olson and Clark, 2001).

In contrast, Inoue et al. (2001) reported that silvercatalysis of a mesophilic bioleach of a concentrate con-taining chalcopyrite, tennantite and enargite and pyriteresulted in preferential biooxidation of the copperminerals compared with pyrite. Pyrite was oxidizedpreferentially in the absence of silver, leading the authorsto propose that silver-catalysed biooxidation could be thekey to bioleaching enargite. Dew et al. (1999) noted thatenargite dissolution required a high redox potential.They achieved this by pre-leaching the mixed sulphideconcentrate with ferric sulphate to remove easilyoxidised minerals such as chalcocite before inoculatingwith bacteria. Redox potentials remained high during thesubsequent bioleach stage, promoting the dissolution ofenargite, but not of chalcopyrite.

6. Future directions

The last two decades have witnessed exponentialgrowth in the technological advancement and commer-cialisation of heap bioleaching of copper ores, partic-ularly acid soluble and secondary sulphides such aschalcocite.

Heap (bio)leaching of copper ores is a stand alonetechnology that is

robust and proven under different climatic conditionsfor oxides and secondary sulphides;

flexible heap engineering and management canaccommodate site peculiarities in remote localities;suited to small deposits;

simple a technology that can be communicated tonon-scientific personnel;

low cost stacking, irrigation, aeration, solutioncollection are all basic;

The technology has also benefited from concomitantimprovements in SX-EW for the production of LMEgrade cathode copper from sulphate solutions and in thecontrol of solvent extraction reagent losses.

However, heap leaching technology has yet to bedemonstrated on a large scale for the bioleaching ofchalcopyrite ore. In heap and dump leaching of run-of-mine chalcopyrite ores, relatively rapid leaching ratesdecrease as a function of time and overall copper reco-veries seldom exceed 50%. Nevertheless, there are pro-cessing and economic reasons why the development ofwhole ore bioleaching is the preferred option for low-

100 H.R. Watling / Hydromgrade chalcopyrite ores.6.1. Microbiology

Among the many groups researching biodiversity, afew are focused on the discovery and characterization ofacidophiles across a range of temperatures. The knowl-edge gained by identifying and characterising morebioleaching strains can only enhance our understandingof bioleaching and acid mine drainage processes. Thesymbiotic relationships between mining microbial strainsin mixed culture are not well understood and should be acontinued focus for researchers.

In heaps, the reduction in microbial growth rates inhigh ionic strength process water might become the rate-controlling step for the bioleaching of copper sulphide.Acidophiles that oxidize iron and sulphur in high ionicstrength solutions should be targeted. Further quantita-tive data on microbial growth rates obtained fromsystematic targeted experimental programs under variedbut controlled conditions, relevant to heap leaching,would be valuable for understanding heap behaviourand for model development.

The variety of organisms that abound in acidic en-vironments, many of which are associated with sulphidemineralization, necessitates the use of novel enrichmenttechniques and modern molecular methods of character-ization that are beyond the scope ofmostminerals industryin-house research laboratories. It may be argued that themicrobial capability to enhance sulphide dissolution canbe assessed without knowing what organisms are presentin a mixed culture and without determining their roles inbioleaching. However, the information that is arising fromthis research, and indeed from the related moleculargenetics research, will undoubtedly benefit the improve-ment of mineral processing technologies in the longerterm. Continued research on the biodiversity of sulphidicenvironments and on microbial characterization andgrowth kinetics should be supported.

The ability to quantify microbial population compo-sition and strain-specific activities and to monitorchanges in populations and their activity as a result ofchanges in different environments would confer twobenefits. First, knowledge of the changes in populationsthat occur as a response to changes in the microbialenvironment would facilitate the design and operation ofbioreactors that optimize the desired activity. For exam-ple, it may be beneficial to promote sulphur oxidationbut reduce iron oxidation in a bioreactor to maximize theextraction of a selected metal. Secondly, by monitoringchanges in the microbial mix that is found in leachates,for example from a heap operation, it should be possibleto infer the prevailing conditions in the heap having

rgy 84 (2006) 81108identified and quantified those leach parameters that

Biomine '97 (Sydney). Australian Mineral Foundation, Glenside,pp. M3.2.1M3.2.9.

Ingledew,parameters

etalluimpact on microbial population dynamics. If they werenot as required, measures could then be taken to restorethe appropriate conditions.

There are a variety of microbiological and biochem-ical methods that can be used to identify and quantifymicrobial groups or strains in environmental samples,for example 16S rRNA gene sequencing, analysis of cellmembrane phospholipids, fluorescent in situ hybridisa-tion, and denaturing gradient gel electrophoresis. How-ever, most were developed for use in cleaner systemsand problems arise in their application to acidic, mineralleaching environments. Currently, no one method givesreliable results for all microorganisms of interest. Thedevelopment of relatively rapid, robust methods for theidentification and quantification of microbial strains,both attached and planktonic, in bioreactor sampleswould greatly assist research into microbial roles inmineral bioprocesses (e.g., Lehman et al., 2001; John-son et al., 2005; Coram-Uliana et al., 2005).

Finally, with the need to operate heaps at aboveambient temperatures to increase leaching chemistryrates, there is a need to improve methods of generatinginocula to suit the selected leach conditions and todeliver them effectively to the heap. This is particularlyimportant in the heap temperature range. Inocula couldalso be tailored to specific ores or minerals.

6.2. Leaching chemistry and mineralogy

Even at high temperatures, faster leaching rateswould be beneficial to processing. Ferrous/ferric ionratios in solution impact on dissolution rates and totalsoluble iron concentrations increase to a point wheremassive precipitation of iron oxides, hydroxides andbasic sulphates can occur. Iron-rich reaction productsreport with the leach residues and must be disposed of inan environmentally responsible manner. Focused re-search on the iron chemistry of bioleaching as part of anoptimization study should be well rewarded.

The mineralogy of ores and concentrates impactsdirectly on the efficiency of leaching and bioleaching,and on the nature of reaction products. Yet, to date, therehas been no (published) systematic quantitative study ofmineralogical effects in respect of leach chemistr