POX vs Biooxidation: Refractory Gold Route Compared

Last updated: 2026-10-24

In the POX vs biooxidation decision, kinetics is the first divide: pressure oxidation (POX) converts sulfide-hosted gold concentrates in 1–2 hours at 190–230 °C (374–446 °F), while bacterial oxidation needs 4–6 days at 40–80 °C (104–176 °F). Both routes liberate refractory gold for downstream carbon-in-leach (CIL) recovery, but they differ sharply in throughput capacity, capital intensity, and tolerance to feed disturbance.

Neither route recovers gold by itself. Both are pretreatment stages ahead of cyanide leaching — the fundamentals of that downstream step are covered in our leaching pillar guide — so the real question is which oxidation chemistry fits your concentrate, your site, and your balance sheet.

Key Takeaways

• POX oxidizes sulfide in 1–2 h at 190–230 °C; biooxidation needs 4–6 days at 40–80 °C.
• Typical overall gold recovery: POX 92–96%; biooxidation 85–93% (illustrative, mineralogy-dependent).
• POX suits concentrates above roughly 500 t/d; biooxidation fits 50–500 t/d duties.
• POX carries higher CAPEX (autoclave plus oxygen plant); biooxidation carries larger tankage.
• Arsenic exits POX as ferric arsenate in residue; biooxidation needs separate neutralization.

What Are POX and Biooxidation?

Pressure oxidation (POX) is a hydrometallurgical pretreatment process that oxidizes pyrite (FeS₂) and arsenopyrite (FeAsS) in an acidic, oxygen-sparged slurry inside a horizontal autoclave, typically at 190–230 °C and 300–500 kPa oxygen overpressure. In refractory ores, much of the gold occurs as sub-micron particles locked within the sulfide lattice — invisible-gold mineralogy that direct cyanidation cannot touch. POX destroys the host mineral, converting sulfide sulfur to sulfate and ferrous iron to hematite or basic ferric sulfate, and the washed residue then reports to carbon-in-leach (CIL), a cyanidation circuit that leaches the liberated gold and adsorbs it onto activated carbon. For autoclave hardware, materials of construction, and compartment design, see our pressure oxidation autoclave guide rather than re-covering that ground here.

Leaching

Biooxidation is a bacterial leaching technique that achieves the same mineral destruction using acidophilic microorganisms — principally Acidithiobacillus ferrooxidans, Leptospirillum ferrooxidans, and Sulfobacillus species — which catalyze the oxidation of Fe²⁺ to Fe³⁺ and reduced sulfur to sulfuric acid at atmospheric pressure. Commercially proven configurations such as the BIOX® process use a series of agitated, aerated tanks; mesophile cultures operate near 40 °C (104 °F), while thermophile cultures extend the window to roughly 45–80 °C (113–176 °F). The mechanism, culture management, and tank configurations are detailed in our bioleaching article.

The governing reactions make the comparison concrete. In POX, pyrite oxidizes approximately as FeS₂ + 15/4 O₂ + 2 H₂O → 2 FeOOH + 2 H₂SO₄ (or to hematite at higher temperature), while arsenopyrite follows FeAsS + 7/2 O₂ + H₂O → FeAsO₄(s) + H₂SO₄ — equations that show both the oxygen demand and the acid generation driving each circuit’s design. In biooxidation, the same stoichiometry applies, but bacteria mediate each half-reaction at rates a thousand times slower; the microbes are catalysts, not reagents.

Scorodite is a crystalline ferric arsenate (FeAsO₄·2H₂O) that precipitates directly under POX conditions and fixes arsenic in a compact, stable residue suitable for long-term disposal. In biooxidation, arsenic dissolves as As(V) and must be precipitated afterward in a dedicated ferric precipitation and neutralization stage — a structural difference in arsenic management, not merely an operating detail.

Both routes are oxidative pretreatment steps, not gold recovery circuits. Their output is a liberated residue whose cyanide-extractable gold responds at 92–96% (POX) or 85–93% (biooxidation), illustrative ranges that depend on liberation quality, preg-robbing carbon, and cyanide-soluble copper. The choice between them therefore hinges on throughput, capital, energy, and risk — which is exactly what the rest of this refractory gold pretreatment comparison quantifies.

POX vs Biooxidation: Side-by-Side Technical Comparison

The table below consolidates the operating envelope of each route. All figures are typical ranges for sulfide concentrates (20–30% S basis) drawn from published industry practice; they are illustrative and every project must be confirmed by testwork.

ASME VIII Div 1 vs Div 2, step by step

Three consequences of this table deserve emphasis in the pressure oxidation vs biooxidation debate.

Speed compounds into capital efficiency. A 1–2 hour residence time means one autoclave train replaces the volume of dozens of biooxidation tanks. At 1,000 t/d concentrate, the tank farm required for six days of residence would be impractically large; this single geometric fact, not chemistry, is why POX owns the large-tonnage end of the market.

Speed also compounds into oxygen demand. POX consumes its stoichiometric oxygen — typically 200–300 kg O₂ per tonne of concentrate for a 25% S feed (typical basis) — in a compressed window, which is why a dedicated cryogenic or VPSA oxygen plant is mandatory. Biooxidation spreads the same demand over days and can often use air blown through spargers, sometimes with oxygen enrichment at the lead tanks. The oxygen bill per tonne is similar; the infrastructure and its cost profile are not.

Recovery differences are real but modest. The 3–5 percentage-point gap in overall gold recovery (illustrative ranges) reflects POX’s more complete sulfide destruction and its ability to burn off carbonaceous preg-robbing material. Where feeds are clean pyrite concentrates with moderate arsenic, well-run biooxidation plants close much of that gap; where arsenopyrite dominance and organic carbon coexist, the gap widens and POX vs biooxidation tilts decisively toward the autoclave.

Which Route Should You Choose for Your Scenario?

Route selection rewards disciplined cut-off thinking. The framework below uses concentrate tonnage, mineralogy, and site conditions — in that order — because tonnage constrains everything else.

Scenario 1: Concentrate throughput. Below about 50 t/d of concentrate, both routes struggle to earn their keep; flotation concentrate shipping or ultra-fine grinding plus intensive cyanidation often beats either. For throughputs below ~50 t/d, biooxidation is preferred because modular stirred tanks scale down gracefully where an autoclave and oxygen plant cannot. Between roughly 500 and 800 t/d sits the crossover band: biooxidation tankage grows linearly while autoclave economics improve with scale. For concentrate duties above ~500 t/d, POX is preferred because autoclave throughput and oxygen-plant economies of scale drive unit costs down faster than biooxidation tankage can. For mid-band duties of 50–500 t/d, biooxidation is preferred because its lower CAPEX releases capital for mining rather than static pressure vessels.

Scenario 2: Mineralogy. Arsenopyrite-dominant concentrates with 3–10% As favor POX strongly: the ferric arsenate forms and stabilizes in one step, whereas biooxidation would dissolve tonnes of As(V) daily into solution before precipitating it. For arsenic grades above ~3% As in the concentrate, POX is preferred because in-autoclave scorodite formation eliminates a large, lime-hungry liquid effluent stream. Carbonaceous ores with significant organic carbon (preg-robbing behavior) also favor POX, which oxidizes carbon at 190–230 °C; biooxidation leaves much of it intact, forcing partial-oxidation compromises or heavier CIL carbon inventories. For preg-robbing feeds, POX is preferred because it destroys the carbonaceous thief rather than working around it. Clean pyrite concentrates with low As and no carbon are the natural biooxidation feed.

Scenario 3: Site conditions. Where grid power is unreliable or diesel-generated, biooxidation’s low power draw (blowers and agitators only, no oxygen plant) is a structural advantage. For sites with power supply below ~20 MW spare capacity, biooxidation is preferred because it avoids a dedicated oxygen plant and high-temperature heat balance. At high altitude or in extreme cold, oxygen-plant derating and bio-tank heat losses cut in opposite directions: cold climates favor POX because the autoclave generates its own surplus heat, while biooxidation tanks may need supplemental heating below 15 °C ambient. Where lime or limestone supply is distant, remember that biooxidation’s dissolved-arsenic load multiplies neutralization reagent demand.

No single criterion decides a real project; the cut-offs above are sequencing heuristics. But they convert an open-ended comparison into a shortlist that testwork can settle — and they end the “it depends” conversation with numbers.

What Are the Failure Modes and Risks of Each Route?

Every route has a characteristic failure signature. Knowing the symptom–cause–mitigation chain in advance is cheaper than learning it during commissioning.

1. Bio-oxidation culture crash after a toxic flush. Symptom: overnight collapse in sulfide oxidation, Fe²⁺/Fe³⁺ ratio rising, tank temperature sagging. → Cause: a pulse of flotation reagents (xanthates, frothers), residual cyanide, soluble copper, or mercury exceeding microbial tolerance entering the lead tank. → Mitigation: feed hold-up and blending capacity, activated-carbon polishing of incoming solution, online monitoring of ORP and CO₂ uptake, and a preserved backup culture to re-inoculate within days rather than weeks.

2. POX autoclave scaling. Symptom: rising steam-to-slurry pressure drop, declining heat-transfer coefficients, unplanned washout frequency. → Cause: silica, hematite, and basic ferric sulfate depositing on titanium coils and compartment walls at high temperature and supersaturation. → Mitigation: controlled feed water chemistry, oxidation-potential management, and scheduled online descaling; the full diagnostic tree is covered in our autoclave scaling resource.

3. Arsenic pentoxide and unstable residue. Symptom: elevated arsenic in off-gas scrubber liquor or leachate-testing failures on the residue pond. → Cause: insufficient ferric iron relative to arsenic (Fe:As ratio below ~3:1, typical guideline) or low-pH pockets allowing soluble As(V) species rather than ferric arsenate to persist. → Mitigation: strict Fe:As ratio control, adequate oxidation potential to force As(V), and routine residue stability testing before disposal.

4. Elemental sulfur passivation in incomplete oxidation. Symptom: gold recovery drops even though headline sulfide conversion looks adequate. → Cause: at temperatures straddling sulfur’s melting point (~119 °C), unoxidized sulfur becomes molten and coats fine gold particles, blocking cyanide access — a known hazard in low-temperature POX and sulfur-accumulating biooxidation circuits. → Mitigation: operate well above the molten-sulfur window, hold conversion targets above ~99% sulfide-to-sulfate, or use a two-stage bio configuration where sulfur-oxidizers polish the first stage.

5. Acid imbalance from feed drift. Symptom: free acid drifting out of control band, neutralization reagent costs climbing month over month. → Cause: changing pyrite-to-pyrrhotite ratios in the flotation concentrate, since pyrrhotite oxidation generates substantially more net acid. → Mitigation: frequent mineralogical monitoring and a maintained acid-balance model; our acid consumption calculation resource walks through the stoichiometry.

6. Neutralization thickener overload. Symptom: gypsum and ferric hydroxide sludge volumes overwhelming the tailings circuit. → Cause: sulfate and dissolved-iron loads — far larger in biooxidation, where all the arsenic and iron report to solution — exceeding the high-density sludge (HDS) design basis. → Mitigation: two-stage HDS with sludge recycle, and conservative solids-loading margins sized on the worst-case feed.

7. Cold-climate heat collapse in biooxidation. Symptom: tank temperatures falling below ~30 °C (86 °F) in winter, oxidation rates dropping with the thermometer. → Cause: biological heat generation, which depends on sulfide oxidation rate, losing the battle against ambient heat loss through the tank farm. → Mitigation: tank insulation, feed preheating using process streams, and designing heat balance for the coldest recorded month, not the annual average.

The pattern is consistent: POX failures are mechanical and thermal, biooxidation failures are biological and chemical. Which family your operations team is better equipped to handle is a legitimate selection criterion in the POX vs biooxidation decision, and it should be scored explicitly during feasibility.

POX vs Biooxidation: Cost and Energy Comparison

Cost structures differ more in shape than in total. POX concentrates its spending in pressure vessels and oxygen; biooxidation spreads it across tankage, neutralization reagents, and a longer time-to-revenue because residence time ties up working capital in inventory. The figures below are illustrative planning ranges for a 200 t/d sulfide concentrate duty (2026 US dollars, 90% utilization, common CIL circuit excluded from both columns). They are typical industry-magnitude estimates for orientation only — not quotations.

Two asymmetries dominate the POX vs biooxidation economics. First, CAPEX: the autoclave route typically costs two to three times more in pretreatment capital, which is tolerable only when throughput is large enough to amortize it — the root of the ~500 t/d cut-off cited earlier. Second, OPEX composition: POX spends on oxygen and fuel (predictable, energy-price-linked), while biooxidation spends on lime and sludge handling (volume-linked and feed-sensitive). When arsenic is high, biooxidation’s lime line can erode much of its OPEX advantage; when power tariffs are high, POX’s energy line does the same in reverse.

Per tonne of gold rather than per tonne of concentrate, the gap narrows further: applying the illustrative recovery ranges (92–96% vs 85–93%) to a 40 g/t concentrate, the recovered-gold difference is roughly 25–40 g per tonne of concentrate processed, worth far more than the $20–30/t OPEX spread at any realistic gold price. Recovery, not operating cost, is usually the decisive economic variable — which is why leaching recovery calculation work belongs in every route-selection model before CAPEX arguments begin.

Composite Case: 250 t/d Refractory Concentrate

Consider a composite case assembled from publicly reported project parameters (anonymised, illustrative): a Central Asian underground mine producing a 250 t/d flotation concentrate grading 38 g/t Au, 24% S, and 3.8% As, with 0.9% organic carbon causing measurable preg-robbing in diagnostic cyanidation (direct CIL recovery 58%).

The mineralogy — arsenopyrite-dominant sulfides plus preg-robbing carbon — pushed the decision framework hard toward POX despite the moderate tonnage: biooxidation would dissolve roughly 9 t/d of arsenic into solution while leaving the carbon thief largely intact, and the operator had industrial gas infrastructure within 40 km. A single-train POX autoclave with a VPSA oxygen plant was selected. Tested outcomes at pilot scale: overall gold recovery improved from 58% (direct CIL) to 94.5%, residue ferric arsenate passed stability leach testing with Fe:As held above 3.5:1, and lime demand of 0.3 t CaO/t concentrate stayed within the design envelope.

The economics closed the argument: illustrative pretreatment CAPEX of $105 million against a recovered-gold uplift of roughly 14 g/t concentrate, payback of under four years at then-current gold prices, and pretreatment OPEX of $92/t concentrate. A biooxidation option priced at $48 million CAPEX was rejected because modeled recovery of 86–89% — depressed by residual preg-robbing and incomplete arsenopyrite oxidation — left approximately 20 g/t of gold per tonne of concentrate unrecovered, an annual revenue sacrifice that dwarfed the CAPEX saving. Subsequent trade-off review confirmed the robustness of the choice: sensitivity analysis showed the POX case remained positive down to a gold price roughly 35% below the design basis, whereas the biooxidation option’s economics broke even only with optimistic (≥91%) recovery assumptions that diagnostic testwork could not support. The case illustrates the framework’s central lesson: match the route to mineralogy first, then check the arithmetic on throughput and capital.

Frequently Asked Questions

In the POX vs biooxidation comparison, which route delivers higher gold recovery?
POX typically delivers 92–96% overall gold recovery versus 85–93% for biooxidation (illustrative ranges for sulfide concentrates). The gap comes from POX’s near-complete sulfide destruction and its ability to oxidize carbonaceous preg-robbing material at 190–230 °C. With clean pyrite feeds and good culture management, biooxidation plants can close much of the difference; with arsenopyrite-dominant or carbonaceous feeds, the gap widens materially.

Why is POX faster than bacterial oxidation?
POX is a thermochemical reaction driven by 190–230 °C temperatures and 300–500 kPa oxygen overpressure, so sulfide oxidation completes in 1–2 hours. Biooxidation relies on microbes operating at 40–80 °C whose metabolic rate — not chemistry — sets the pace, requiring 4–6 days of residence. Each route’s kinetics are intrinsic to its mechanism; neither can be “tuned” into the other’s timescale.

What concentrate tonnage justifies an autoclave?
Industry experience places the practical POX threshold at roughly 500 t/d of concentrate (typical guideline), below which autoclave and oxygen-plant CAPEX per annual tonne of gold becomes hard to justify. Between 50–500 t/d, biooxidation’s modular stirred tanks dominate; below ~50 t/d, neither route is usually economic and alternatives such as ultra-fine grinding should be screened first.

How is arsenic disposed of in each route?
POX precipitates arsenic inside the autoclave as crystalline ferric arsenate (scorodite-like) within the residue, which can be disposed of directly after testing. Biooxidation dissolves arsenic as As(V) into solution, requiring a separate ferric precipitation and neutralization stage producing large hydroxide sludge volumes. POX therefore offers intrinsically simpler and more compact arsenic management, especially above ~3% As in concentrate.

Can biooxidation handle carbonaceous, preg-robbing concentrates?
Only partially. Biooxidation targets sulfide minerals and leaves most organic carbon intact, so preg-robbing losses persist unless the circuit adds chlorination, CIL with high carbon inventories, or other workarounds. POX oxidizes carbonaceous matter at operating temperature, neutralizing preg-robbing directly. For concentrates with significant organic carbon, this is frequently the single decisive point in route selection.

Does biooxidation work at remote sites with weak infrastructure?
Yes, and it is often selected for exactly that reason. Biooxidation needs no oxygen plant, no high-pressure equipment, and modest power (30–50 kWh/t concentrate, illustrative), so it suits sites with limited grid capacity. The trade-offs are larger tankage, a longer commissioning ramp while cultures acclimate, and disciplined feed-quality control to protect the microbial population from toxic excursions.

How long does commissioning take for each route?
POX plants typically reach nameplate in 3–6 months after mechanical completion, dominated by heat-up, pressure-system commissioning, and scale-management tuning. Biooxidation requires 6–12 months (typical) because the microbial culture must acclimate to the specific concentrate, and recovery builds gradually as sulfide-oxidizing populations establish across the tank train. Project schedules should carry these ramps explicitly.

Conclusion

The POX vs biooxidation decision reduces to four numbers: throughput, arsenic grade, carbon content, and capital. Above ~500 t/d of concentrate, POX is preferred because autoclave economics and 92–96% recovery (illustrative) dominate; below that band, biooxidation’s lower CAPEX and modest infrastructure demands win, provided the concentrate is pyrite-dominant, low in arsenic, and free of preg-robbing carbon. High arsenic or carbonaceous mineralogy overrides tonnage and pulls the decision back toward the autoclave, because in-autoclave ferric arsenate formation and carbon destruction have no biological equivalent. Neither route is universally superior; each is provably better inside identifiable boundaries, and a competent testwork program will place your ore within them within months, not years.

The most expensive mistake in refractory gold projects is selecting a route from analogy rather than from your own mineralogy. Lexin-tech supports route selection and de-risking through supervised pilot campaigns — bench-scale biooxidation culture acclimation, batch and continuous POX testwork, residue stability testing, and full mass-and-energy balance modeling. Before committing autoclave steel or a tank farm, validate the recovery, acid balance, and reagent consumption on your actual concentrate: see our pilot testing services or browse the wider resource library for the underlying calculation methods. The chemistry is settled; the only open variable is your ore.

References

1. Marsden, J.O. and House, C.I. (2006). The Chemistry of Gold Extraction, 2nd edition. Society for Mining, Metallurgy and Exploration (SME), Littleton, Colorado.
2. Habashi, F. (1993). A Textbook of Hydrometallurgy. Métallurgie Extractive Québec, Québec City.
3. Free, M.L. (2013). Hydrometallurgy: Fundamentals and Applications. John Wiley & Sons, Hoboken, New Jersey.
4. Rawlings, D.E. and Johnson, D.B. (eds.) (2007). Biomining. Springer, Berlin/Heidelberg.
5. U.S. Geological Survey (2024). Mineral Commodity Summaries 2024 — Gold. U.S. Department of the Interior. Available at: https://pubs.usgs.gov/periodicals/mcs2024/mcs2024-gold.pdf

Changelog

2026-10-24 — First published.

Related reading

• What Is Hydrometallurgy? Definition, Process, Applications