Bioleaching Mining
Technology and Applications

Schippers and Sand described this in 1999 in Applied and Environmental Microbiology. Hansford and Vargas had laid the electrochemical groundwork a few years earlier. The energy yield is about 30 kJ per mole of Fe²⁺ oxidized, compared to about 2,870 kJ per mole from glucose, which forces Acidithiobacillus ferrooxidans to process enormous quantities of iron to grow at all and is the reason it accelerates the Fe²⁺ to Fe³⁺ conversion by five to six orders of magnitude at pH below 2.5. A second functional group, principally Acidithiobacillus thiooxidans and Acidithiobacillus caldus, oxidizes the sulfur intermediates (polysulfides, elemental sulfur, thiosulfate, tetrathionate) that form when ferric iron breaks apart sulfide lattices. Without them, elemental sulfur coats every reactive surface within weeks and the system stalls. Their product, sulfuric acid, also keeps pH at 1.5 to 2.0 where ferric iron remains soluble rather than precipitating as jarosite.

Most of what gets written about bioleaching stops here and moves on to applications. The mechanism paragraph, then the applications list. But the mechanism is where all the difficulty lives, and the difficulty is concentrated overwhelmingly in one mineral.

Chalcopyrite is about 70% of the world's copper sulfide reserves. Under mesophilic bioleaching conditions, extraction plateaus at 20 to 40% of contained copper regardless of how long you leach. Watling's 2006 review in Hydrometallurgy compiled dissolution rate data from dozens of studies and the picture was consistent. Extending a heap cycle from 12 months to 24 months gains almost nothing on chalcopyrite. The passivation layer shuts down the reaction.

The composition of this passivation layer has been debated for over forty years and remains unresolved. Dutrizac argued for elemental sulfur in the 1980s. Klauber reviewed the evidence in 2008 in International Journal of Mineral Processing and favored a metal-deficient polysulfide interpretation. Harmer used synchrotron X-ray spectroscopy and found copper-depleted sulfide phases. Others have identified jarosite overgrowths or amorphous iron oxyhydroxide films.

The disagreement has persisted for long enough that it probably reflects something about the system rather than about the quality of the researchers. The passivating material almost certainly varies with conditions. A chalcopyrite grain from a Chilean porphyry with elevated selenium in the lattice will produce different surface phases than one from a Canadian VMS deposit. The solution ferric iron concentration matters. Whether the pH is 1.2 or 1.8 matters. Temperature matters enormously, as discussed below. How long the surface has been exposed matters. Each research group works on a particular chalcopyrite source at a particular set of conditions and finds a definitive passivating phase. Another group works on different material at different conditions and finds a different definitive passivating phase. Four decades of this.

Hiroyoshi and colleagues in Sapporo published a series of papers starting in 2000 showing that chalcopyrite dissolves fastest at intermediate redox potentials, in the range of 550 to 620 mV versus the standard hydrogen electrode. Above about 650 mV, passivation intensifies sharply. This finding has been confirmed independently by Córdoba and by Third at the University of Cape Town. It is one of the more robust and more inconvenient results in bioleaching science, because Acidithiobacillus ferrooxidans and Leptospirillum ferrooxidans, doing exactly what they are supposed to do, push solution redox potential to 680 to 750 mV. Leptospirillum is more aggressive in this regard because of its higher affinity for ferrous iron at low pH.

Hackl's group at UBC explored controlled-potential leaching extensively in the late 1990s and early 2000s. In columns and stirred reactors, holding the potential in the 580 to 620 mV range roughly doubled chalcopyrite extraction compared to uncontrolled systems. The experiments are convincing.

What happens at heap scale is different, and this is where the chalcopyrite story intersects with heap hydrology in a way that most discussions of chalcopyrite passivation fail to connect.

Dixon at UBC modeled heap transport for years. His group's tracer studies showed that solution flowing through a commercial heap contacts 10 to 30% of the ore volume. The rest sits in stagnant zones exchanging solute with the mobile phase only by diffusion. Channeling develops from non-uniform particle size distribution, from fine particle migration under sustained irrigation, and from compaction under the weight of overlying lifts. When you try to impose a controlled redox potential on a system where 70% or more of the solution is either racing through preferential channels or trapped motionless in pores, the concept of "the solution redox potential" becomes an abstraction. The PLS collected at the drain is an average of conditions ranging from deeply reducing in stagnant anoxic pockets to highly oxidizing in well-aerated channels. The average has limited predictive value for chalcopyrite behavior at any specific point inside the heap.

Pilot-scale attempts at controlled-potential heap bioleaching, conducted by several copper producers and reported at conferences like ALTA and Hydroprocess rather than in journals, show a consistent pattern. For the first 60 to 90 days, controlled potential delivers chalcopyrite extraction rates roughly double the uncontrolled baseline. By six months, the extraction curves have converged. The heap's internal heterogeneity overwhelms the control strategy. This has happened on different ores, in different climates, with different operators. The reported performance gap between column test predictions and actual heap recovery on chalcopyrite runs 1.5x to 2.5x and has been remarkably stable across the literature and across internal company data presented at conferences for over fifteen years.

Agglomeration (binding fine particles to coarser ones with sulfuric acid and sometimes polyacrylamide before stacking) is the primary defense against channeling and it is practiced at virtually every heap operation. Bouffard at UBC did detailed work on agglomeration quality metrics and their correlation with long-term permeability. A 10% improvement in agglomeration quality, measured as agglomerate resistance to breakdown under irrigation, translates directly to improved solution contact and therefore to improved recovery. At many operations, money spent on better agglomeration would deliver more copper than the same money spent on microbial inoculation or chemical additives.

Jarosite precipitation inside the heap compounds the permeability problem over time. Where ferric iron encounters zones buffered to pH 2.5 or above by gangue carbonates, jarosite fills pore throats and cements the local pore network. These cemented zones are permanent dead volumes for the remainder of the leach cycle. In heaps with more than about 3% carbonate gangue, jarosite-cemented zones can account for 10 to 20% of the ore mass by mid-cycle.

Gericke at Mintek published results in Hydrometallurgy in 2001 showing 97% copper extraction from chalcopyrite using Metallosphaera sedula at 78°C in stirred reactors. The retention time was about three weeks. Several other groups have confirmed that at 70 to 80°C, chalcopyrite passivation largely ceases. The sulfur and polysulfide intermediates at the mineral surface decompose or oxidize faster than they accumulate at these temperatures.

Pyrite oxidation releases about 1,500 kJ per mole. In a heap containing a few percent pyrite by mass, this exothermic reaction generates substantial heat. The thermal mass of millions of tonnes of rock retains heat effectively. Thermocouple strings in commercial copper heaps in Chile record core temperatures of 55 to 73°C persisting for months. The hot zones are surrounded by cooler zones: the surface loses heat to the atmosphere, the base is cooled by drainage, the irrigated top is cooled by evaporation. The microbial community follows the thermal gradient. Mesophiles at the surface and base. Moderate thermophiles (Sulfobacillus, At. caldus, L. ferriphilum) through the intermediate zone. And in the hottest core, conditions that would support the extreme thermophilic archaea that Gericke worked with.

At least two large Chilean copper producers have tested managed thermal profiling in internal pilot programs: insulating the heap surface with geomembrane or soil, reducing irrigation to limit evaporative cooling, adjusting aeration to deliver oxygen without flushing heat. Results presented at ALTA and Hydroprocess conferences indicate 15 to 30% improvement in chalcopyrite extraction over mesophilic baselines. These numbers come from internal reports and conference presentations rather than peer-reviewed publications, which makes detailed evaluation difficult. The best results came from heaps with high pyrite content (providing the heat source) in arid environments with minimal rainfall (limiting cooling). No operator has published a bankable feasibility study on this basis.

The gap between a 15 to 30% improvement and a positive investment decision is partly about sustainability of the thermal management through seasonal variation and operational interruptions, and partly about whether the incremental capital for insulation and modified aeration infrastructure is justified by the incremental copper production at prevailing prices and discount rates.

Silver ions at 5 to 50 mg/L accelerate chalcopyrite dissolution by 5x to 20x. The silver reacts with nascent polysulfide on the mineral surface, forming Ag₂S, which ferric iron rapidly oxidizes to regenerate Ag⁺. The silver catalytically scavenges the specific sulfur species responsible for passivation. Dreisinger's group at UBC demonstrated copper extraction above 95% from chalcopyrite concentrate at 65°C with silver catalysis in ferric sulfate media.

This connects to the Galvanox process that Dreisinger also developed. Galvanox uses fine-ground chalcopyrite mixed with pyrite in ferric sulfate at elevated temperature. The galvanic couple between pyrite (high rest potential, cathode) and chalcopyrite (low rest potential, anode) accelerates dissolution. The elevated temperature suppresses passivation. No biological participation. Copper recovery above 95%.

At heap scale, silver fails because it adsorbs irreversibly onto the enormous surface area of gangue minerals and precipitates inside the heap. The silver losses over a leach cycle would be measured in tonnes. There has been some exploration of whether other catalytic ions might substitute for silver at lower cost and with less adsorptive loss. Bismuth and mercury have been tried. Mercury has obvious environmental problems. No substitute has matched silver's catalytic efficiency for this specific reaction.

Adding ground pyrite concentrate to chalcopyrite heaps to exploit galvanic enhancement has been tested at several operations. The reported improvements, 20 to 50% in laboratory studies and less in heaps, correlate with how well the added pyrite achieved physical contact with chalcopyrite surfaces. In a heap, contact depends on particle size, agglomeration practice, stacking geometry, and the stochastic arrangement of particles. Sometimes the pyrite ends up in the right place. Sometimes it ends up in a high-permeability channel where solution flows past without prolonged contact with chalcopyrite.

Pyrite and molybdenite dissolve through the thiosulfate pathway described by Schippers and Sand. Ferric iron extracts electrons from the crystal lattice. Sulfur goes thiosulfate to sulfate without depositing elemental sulfur. No passivation. The reaction runs to completion. Biooxidation of pyritic refractory gold concentrates has been commercial since the BIOX plant at Fairview mine in Barberton, South Africa started in 1986. Plants at Wiluna in Western Australia, Sao Bento in Brazil, Obuasi in Ghana, and Suzdal in Kazakhstan followed over the next two decades. The technology is mature and the technical risk is well-bounded by commercial experience on this mineral class.

Sphalerite dissolves through the polysulfide pathway, like chalcopyrite, but with passivation that is moderate enough not to prevent commercially viable recovery. Zinc bioleaching has been practiced commercially and the kinetics, while slower than acid leaching of zinc oxides, are adequate for heap and stirred-tank operations.

BIOX reactors process finely ground refractory gold concentrates at 40 to 45°C with 4 to 6 day retention times. Leptospirillum ferriphilum dominates iron oxidation. Acidithiobacillus caldus handles sulfur. Gold recoveries from concentrates yielding below 50% by direct cyanidation exceed 95% after biooxidation. Fairview has operated since 1986. The combined operating history across all BIOX plants spans decades.

The community in BIOX reactors is monitored by qPCR and adjusted through process control.

The amount of research effort directed at phage ecology in industrial biooxidation is small relative to its operational importance. Maybe a dozen researchers worldwide work on it actively. There is no standard protocol for predicting or managing phage outbreaks. Operators handle them by experience.

CO₂ supply is a second constraint in stirred tanks processing low-carbonate concentrates at high cell density. The organisms are autotrophs fixing CO₂, and at high biomass concentrations the CO₂ dissolving from sparged air may not keep up with demand. Some plants add 5 to 15 kg of limestone per tonne of concentrate to supplement CO₂, accepting the acid consumption penalty in exchange for faster microbial growth.

Coram and Rawlings at Stellenbosch published community analyses from South African biooxidation plants in 2002 showing that microbial diversity in these systems exceeded expectations. Subsequent metagenomic work, enabled by cheaper high-throughput sequencing, has found 200 to 500 species in typical commercial heaps. The named autotrophic iron and sulfur oxidizers account for a few percent of total species count. Heterotrophic bacteria, archaea of unknown function, fungi, and phages make up the rest.

The functional contribution of the unnamed majority is uncharacterized. Potential roles include consuming organic excretions that would otherwise inhibit autotrophs, producing biosurfactants that alter solution wetting behavior on mineral surfaces, competing with autotrophs for attachment sites, and lysing cells through phage predation. The net effect on bioleaching kinetics has not been quantified in any heap. Process models treat the system as if it runs on the activity of five or six named species plus bulk solution chemistry, and these models consistently overpredict heap recovery by 30 to 50%. Part of that overprediction comes from hydrological problems. Some of it may come from ecological interactions that the models do not represent.

Rawlings at Stellenbosch tracked community development in South African bioleaching systems for over a decade. Side-by-side column comparisons between inoculated and uninoculated systems, conducted by his group and independently by others on different continents with different ores, converge on the same result: after 60 to 120 days the communities are indistinguishable. The indigenous organisms, adapted to the local ore chemistry, water quality, and climate, displace the laboratory strains.

Inoculation reduces the lag phase, the period before autotrophic activity shows up in PLS chemistry. For short cycles (under six months, which covers some zinc heap operations), lag reduction matters. For 12 to 18 month copper cycles, it does not measurably affect total extraction.

Molecular community monitoring (qPCR ratios of key functional groups every two weeks) produces more operational value for long-cycle heaps than inoculation does. A declining sulfur oxidizer population shows up in the community data before passivation becomes visible in copper tenors. A shift toward Leptospirillum dominance indicates rising redox potential, which is fine for secondary sulfides and counterproductive for chalcopyrite. Catching these signals early allows adjustment of acid addition, aeration, or irrigation before production is affected.

Sand's group in Hamburg, particularly work by Gehrke published in Applied and Environmental Microbiology in 1998, showed that the EPS layer secreted by Acidithiobacillus ferrooxidans on mineral surfaces concentrates ferric iron far above bulk solution levels through complexation with glucuronic acid residues. The EPS functions as a localized high-oxidant-concentration zone at the cell-mineral interface.

This means ore surface properties that control cell attachment (charge, roughness, hydrophobicity, weathering films) directly influence bioleaching kinetics through their effect on biofilm formation and EPS-mediated ferric iron concentration. Two ores with identical mineralogical composition and particle size can produce different recoveries because of surface characteristics that standard assays do not measure. Column testing on representative ore is non-negotiable for bioleaching feasibility, and even column tests have limited scale-up predictive power because biofilm dynamics change between a 10 cm diameter column and a 10-meter heap lift.

Porphyry copper deposits have a layered structure: oxide cap near the surface, secondary sulfide enrichment zone below, primary chalcopyrite extending to depth. Hydrometallurgical copper production for the past three decades has fed primarily on the upper two layers. Acid leach the oxides, bioleach the secondary sulfides, run PLS through SX-EW. This is how operations at Escondida, Cerro Verde, Radomiro Tomic, Zaldivar, Spence, and others produce cathode copper.

The oxide caps and secondary enrichment zones at most of these deposits are approaching exhaustion. Codelco's annual reports document progressive reduction in leachable ore categories at their northern Chilean properties. BHP's Escondida has shifted increasingly to concentrator feed. Anglo American's Los Bronces faces the same trajectory.

Below the leachable material sits primary chalcopyrite extending hundreds of meters deeper. The contained copper is enormous. The conventional path for primary sulfide is a flotation concentrator followed by smelting and refining, at $5 to $10 billion for a world-class greenfield complex. The alternative is bioleaching the chalcopyrite and continuing to use the existing SX-EW infrastructure, already built and depreciated. For secondary sulfides, this transition is technically straightforward and already underway. For primary chalcopyrite, it depends on solving the passivation problem at heap scale.

Copper smelting emits 2.5 to 4.0 tonnes CO₂ equivalent per tonne of cathode copper depending on energy mix. The bioleaching-SX-EW route emits 1.0 to 2.0 tonnes, dominated by electrowinning at 2,000 to 2,500 kWh per tonne of copper deposited. With solar-powered electrowinning in the Atacama, where solar PV is now the lowest-cost power source, the carbon intensity drops below 0.5 tonnes CO₂ per tonne of Cu. No sulfur dioxide emissions. Lower water consumption in closed-circuit operation. No conventional tailings dam for the leached residue.

Uranium bioleaching dates to the 1960s. Organisms generate ferric iron and sulfuric acid from pyrite oxidation, and these reagents dissolve uranium from uraninite. The biology is a reagent factory. Established practice.

Terrafame in Finland heap-bioleaches a polymetallic black schist for nickel, zinc, cobalt, and copper at grades uneconomic for conventional processing. The operation has had environmental incidents (a major tailings leak in 2012 when it was operating as Talvivaara), went through bankruptcy and restructuring, and has required more capital than projected. It remains the largest commercial demonstration of bioleaching for battery minerals.

Electronic waste bioleaching is at pilot scale. Circuit boards at 10 to 25% Cu and 200 to 400 g/t Au. Two-step process: biogenic ferric sulfate generated in one reactor, contacted with shredded e-waste in another to avoid exposing organisms to brominated flame retardants. Copper recovery above 95% at pilot scale. The metallurgy works. The constraint is collection and preprocessing logistics.

Rare earth leaching from phosphogypsum and fly ash using Aspergillus niger producing organic acids operates by a completely different mechanism (chelation, not iron-mediated oxidation), with slow kinetics, poor selectivity between individual rare earth elements, and an expensive back-end separation step that bioleaching does not address. It has attracted research funding. It has not demonstrated commercial viability.

Mine tailings reprocessing combines metal recovery with acid mine drainage reduction. Historical tailings from mid-twentieth-century operations contain residual sulfide-hosted metals that old flotation technology left behind. Bioleaching extracts the residual metal while oxidizing the sulfides that generate AMD. The economics improve when material handling costs are shared with dam remediation programs. Demonstrated commercially in Chile, South Africa, and Australia.

CRISPR-Cas editing has been demonstrated in Acidithiobacillus ferrooxidans at laboratory scale. Valdés and Quatrini's groups in Chile have sequenced key bioleaching species and developed genetic tools. Modification targets include the rus operon (rusticyanin-based electron transfer chain for iron oxidation), copper efflux transporters, and EPS biosynthesis genes.

The distance between a gene knockout in a flask and a stable engineered strain in a heap is not a matter of scale-up optimization. A heap contains hundreds of competing species, bacteriophages, fluctuating conditions, and no sterile barrier. Maintaining an engineered genotype as the dominant community member in this environment over a multi-year leach cycle would require a level of control that heaps do not provide. No credible pathway across this gap has been described.

Adaptive laboratory evolution (hundreds of generations of selection under controlled conditions without foreign DNA) produces natural variants with enhanced traits. Zammit at CSIRO worked on thermotolerant strains this way. The strains face no regulatory barriers. Whether they persist against indigenous communities at field scale is unknown because the field trials have not been done.

Managing the heap environment to shift selective pressure on the resident community (temperature control through aeration and irrigation, ferrous/ferric ratio manipulation, pH adjustment) can favor organisms with desired phenotypes using existing infrastructure. The limitation is obvious: this can only select for traits present in the indigenous gene pool.

Heap cycles run 300 to 500 days for copper. Smelting processes concentrate in hours.

Carbonate gangue above about 3% consumes acid faster than sulfide oxidation generates it. Clay content above about 10% causes compaction that agglomeration cannot fully prevent. Both are identifiable through standard characterization and both have killed projects where the characterization was inadequate.

SX plant capital and reagent replacement costs are frequently underestimated in early feasibility work. In circuits processing iron-rich PLS, organic reagent degradation and entrainment losses are higher and the operating cost of the SX circuit can exceed all microbial and chemical management costs in the heap.

Chalcopyrite passivation under mesophilic conditions limits primary sulfide copper recovery to levels that rarely support a positive investment case at prevailing copper prices. The entire preceding discussion of redox control, temperature management, silver catalysis, and galvanic enhancement is an extended description of the various attempted solutions to this one problem. None of them has reached commercial deployment at heap scale.