Copper Solvent Extraction: Process Design & Troubleshooting

Last updated: 2026-10-25

Copper solvent extraction transfers 90–95% of dissolved copper from pregnant leach solution (PLS) into an organic phase in every well-designed extraction stage, making it the decisive separation step between heap leaching and electrowinning (EW). This pillar overview distils circuit configuration, key design parameters, extractant selection, and field-proven troubleshooting — the same ground we cover in depth across our liquid-liquid separation hub.

Key Takeaways

• A standard 2E1W1S copper SX circuit moves 90–95% of Cu per extraction stage at O/A ratios of 1:1 to 1.5:1.
• Oxime extractant choice — aldoxime vs ketoxime — sets loading strength, strip efficiency, and crud behaviour.
• Mixer-settler design is driven by mix time (1–3 min), settling rate (3–6 m³/m²·h), and entrainment targets.
• Crud, phase disengagement, and entrainment cause most recoverable copper and organic losses.
• Pilot testing before detail design routinely de-risks stage counts, O/A ratios, and organic inventories.

What Is Copper Solvent Extraction?

Copper solvent extraction is a liquid-liquid separation process that selectively transfers Cu²⁺ ions from an acidic pregnant leach solution into an organic phase, and then releases them into a strongly acidic electrolyte from which high-purity cathode is electrowon. Together with heap or dump leaching upstream and EW downstream, the circuit forms the SX-EW route, which today supplies roughly a quarter of the world’s mine copper — around 4 Mt/a — from oxide and secondary sulphide deposits in Chile, Peru, the USA, Mexico, the Democratic Republic of Congo (DRC), and Zambia.

how Solvent Extraction is decided

Mixer-Settler Settler Area and Mixer Volume Calculator in practice

The active chemistry sits in the organic phase, typically 7–12 vol% of a hydroxyoxime extractant dissolved in a kerosene-type diluent. An oxime extractant is an organic chelating reagent that exchanges its hydrogen ions for copper ions at the interface, according to the equilibrium:

2RH(org) + Cu²⁺(aq) ⇌ CuR₂(org) + 2H⁺(aq)

Because the reaction releases acid during extraction and consumes acid during stripping, the process is fully reversible: dilute acid (pH 1.5–2.0) drives extraction, and 150–180 g/L H₂SO₄ spent electrolyte drives stripping. That reversibility is what makes the organic phase a working fluid rather than a consumable — it circulates continuously between extraction and strip duty, carrying copper like a bucket carries water.

The hardware that makes this happen is almost universally the mixer-settler. A mixer-settler is a contacting vessel that disperses organic and aqueous phases with an impeller to create the interfacial area needed for fast mass transfer, then discharges them to a quiescent settler where gravity separates the phases cleanly. A single mixer-settler unit performing extraction duty is called a stage; several stages in series reach the overall recovery the flowsheet demands. Modern settlers achieve better than 99.9% phase disengagement under design conditions, but that figure erodes quickly when suspended solids, crud, or chemistry drift appear — which is why the design and troubleshooting topics later in this article matter so much to plant economics.

Two auxiliary streams close the loop. Raffinate is the copper-depleted aqueous stream leaving the extraction section; it returns to the heap or dump leach, carrying residual acid that suppresses gangue dissolution. Aqueous entrained in the stripped organic is removed in a wash stage so that chlorides and iron do not contaminate the tankhouse electrolyte. Understanding these recycles is essential before sizing any mixer-settler — see our mixer-settler design guide for the sizing fundamentals.

Key Process Parameters

Every copper solvent extraction flowsheet is ultimately defined by a short list of interacting parameters. Getting them right individually is table stakes; getting them right together is what separates plants that run at 95% overall copper recovery from those that bleed value to raffinate and electrolyte bleed.

D2EHPA vs Cyanex 272: selection rules

The most fundamental parameter is the organic-to-aqueous (O/A) ratio, the volumetric flow ratio of organic to PLS in each mixer. In extraction stages it typically sits between 1:1 and 1.5:1. A higher O/A increases copper transfer capacity per pass but enlarges organic inventory, capital cost, and entrainment losses. In stripping, the O/A ratio works in reverse: running strip O/A higher (organic-lean relative to electrolyte) builds copper tenor in the electrolyte, with loaded-organic Cu and strip O/A tuned so the stripped organic carries less than 1.5–2.5 g/L Cu.

Extractant concentration, expressed in vol% of the organic phase, sets the maximum loading capacity of the organic. Most commercial plants run 7–12 vol% of oxime in diluent, which supports organic loadings of roughly 2–6 g/L Cu. Designers usually target 85–95% of the theoretical maximum loading at equilibrium, because pushing toward saturation slows kinetics and worsens strip performance.

Stage efficiency expresses how close each mixer-settler comes to the equilibrium transfer predicted by the isotherm. Well-mixed copper SX stages typically achieve 85–95% efficiency (a conservative 90% is a sound design basis). Fewer, larger stages with high efficiency usually beat many small stages once capital, pumping, and organic inventory are counted.

Mix time — the nominal residence time in the mixer — controls how much of the theoretical equilibrium is actually reached. One and a half to three minutes is the practical band; below about 90 s, mass transfer becomes kinetically limited, and above about 4 min, additional residence buys little while shear-related entrainment grows.

Settling rate, expressed as combined phase flow per settler area (m³/m²·h), determines settler area. Conventional gravity settlers handle 3–6 m³/m²·h for copper duty; conservative designs use the low end when PLS carries suspended solids.

Entrainment — organic droplets carried in aqueous and vice versa — is the silent profit leak. Targets are typically below 20–50 ppm organic-in-raffinate before cleanup and below 100 ppm aqueous-in-organic into stripping, achievable with good mixer design, picket fences, and clean chemistry.

Temperature accelerates kinetics and phase disengagement; 20–40 °C is the normal operating band, with many plants settled on 25–35 °C as a cost/kinetics compromise.

For a deeper treatment of how O/A ratio interacts with stage count and isotherm shape, see our dedicated note on O/A ratio in solvent extraction.

How Is a Copper Solvent Extraction Circuit Configured?

Circuit configuration — how many extraction, wash, and strip stages, and in what sequence — is decided by the PLS grade, acid profile, and cathode production target. Plants are described by a compact notation: 2E1W1S means two extraction stages, one wash stage, and one strip stage, and this is the most common configuration for medium-grade PLS worldwide. High-grade PLS or high recovery targets justify 3E1W1S or 2E1W2S; low-grade PLS in hot climates sometimes runs a lean 1E1S. Very large operations, notably the big Chilean and Peruvian oxide plants, often adopt series-parallel arrangements so that several leach streams can be blended without overloading a single extraction train.

The three process streams define the flowsheet spine. PLS enters the extraction section countercurrent to the loaded organic. Raffinate leaves extraction and returns to leaching. Loaded organic passes through the wash stage — where fresh or dilute acid scrubber water displaces entrained PLS and its impurities — and then into strip, where spent electrolyte picks up the copper and leaves as rich electrolyte for EW. The stripped organic returns to extraction, closing the organic loop.

Designing the circuit is a structured exercise. A defensible sequence looks like this:

1. Define the duty. Fix annual cathode production, PLS flow and grade, and required overall recovery — these bound everything downstream.
2. Characterise the PLS. Measure Cu, free acid, Fe, Mn, chloride, and suspended solids across the leach cycle, including end-of-life and start-up extremes.
3. Select the extractant system. Choose aldoxime, ketoxime, or a blend (Section below) and set vol% to meet loading capacity with margin.
4. Build the extraction isotherm and McCabe–Thiele balance. Laboratory isotherms at true PLS chemistry yield the theoretical stages; apply 85–95% stage efficiency to convert to real stages.
5. Fix O/A ratios and stream grades. Iterate extraction and strip O/A so that advance electrolyte tenor (typically 40–50 g/L Cu) and stripped-organic residual Cu both land in target.
6. Insert wash and guard stages. Add the wash stage for entrained PLS displacement, and evaluate a scrub stage where chloride or iron carryover threatens the tankhouse.
7. Size the mixer-settlers and verify by pilot test. Convert the residence-time and settling-rate basis into mixer-settler geometry, then confirm disengagement and entrainment with pilot testing on real PLS before committing to detail design.

The wash stage deserves emphasis: it is cheap insurance. Every millilitre of PLS entrained into loaded organic carries iron, manganese, and chloride directly into the electrolyte circuit, where they degrade cathode quality and current efficiency. Our copper SX-EW applications overview shows how these stages fit into the full leach-to-cathode flowsheet.

Why Does a Copper SX Plant Lose Efficiency?

When a copper SX plant underperforms, the cause is usually one of six recurring problems. Each presents a recognisable symptom, and each has a field-proven remedy.

1. Crud. Symptom: a stable grey-to-black emulsion band accumulating at the organic-aqueous interface, growing settler rag layers, and rising organic make-up. Cause: fine solids (silica, clays, precipitated iron and alumina), leach residues, and degradation products that stabilise droplets at the interface; crud is agitated into the organic during mixing. Remedy: remove solids upstream (clarification or ponding), withdraw and process crud through dedicated treatment — centrifuges, acid/clay washing, or clay-mat filtration (see our crud treatment guide) — and re-agent the recovered organic. Prevention is always cheaper than cure.

2. Slow phase disengagement. Symptom: settler times rising beyond 60–90 s, hazy interfaces, organic dragging into raffinate. Cause: fine solids, high mixing intensity, degraded or overly strong organic, low temperature, or chemistry drift (excess extractant vol%). Remedy: verify chemistry in the lab, cut impeller tip speed, raise temperature toward 30–40 °C, and rebuild settler capacity with picket fences or additional area.

3. Entrainment losses. Symptom: organic sheens on raffinate ponds, organic inventory falling faster than mechanical losses explain, or copper reporting to EW as dark, streaky cathode. Cause: excessive mixer shear, flooded settlers, coalescer failure, and crud-stabilised emulsions. Remedy: audit entrainment at each stream (typical audit points: raffinate, advance electrolyte, tankhouse bleed), fit or rebuild coalescers and flotation cleanup on raffinate, and correct the root mixer/settler hydraulics. Organic losses at 50–100 ppm organic-in-raffinate can cost hundreds of thousands of dollars per year at large plants — often more than the extractant itself.

4. Silica and suspended solids. Symptom: progressive crud layer growth, interfacial “gunk” that resists normal withdrawal. Cause: reactive silica from ores with high acid-soluble silicates — a well-known challenge in parts of Chile, Peru, Indonesia, and the Zambian Copperbelt — plus unclarified PLS. Remedy: quantify reactive silica in PLS during testwork, add upstream solid/liquid separation, and design settlers and crud handling for the measured load rather than an assumed one.

5. Manganese and oxidative degradation. Symptom: extractant performance decaying over months, rising reagent consumption, unexplained disengagement deterioration. Cause: Mn²⁺ carried in PLS is oxidised (by MnO₂ or microbiology in the leach) to permanganate, which attacks oxime molecules; this is a documented problem in Mn-bearing ores, notably in DRC Katanga operations and some Chilean deposits. Remedy: minimise Mn carryover in the leach, avoid strong oxidants in PLS contact, monitor extractant strength routinely, and consider the more oxidation-tolerant end of the extractant spectrum.

6. Iron carryover to the tankhouse. Symptom: falling current efficiency (each 1 g/L Fe in electrolyte typically costs several points of current efficiency), elevated cell voltage, and acid mist aggravation. Cause: aqueous entrained in loaded organic bypassing an effective wash, or wash-stage water rate too low. Remedy: raise wash-stage O/A and water availability, audit aqueous-in-organic entrainment, and control electrolyte bleed so Fe stabilises below roughly 2–4 g/L.

Most of these failure modes interact — crud worsens entrainment, entrainment worsens iron carryover — so a systematic audit that traces copper, organic, and impurity mass balances across the whole circuit is usually faster than attacking symptoms one by one. Our centrifugal extractor overview also covers how intensified contactors sidestep several of these problems in compact duties.

Extractant Selection: Aldoxime vs Ketoxime

The commercial extractant decision for copper solvent extraction comes down to two hydroxyoxime families: aldoximes (e.g., LIX 973N-class and Acorga M5640/M5910 products, frequently ester-modified) and ketoximes (e.g., LIX 84-I-class reagents). Both chelate Cu²⁺ through the same oxime chemistry, but their substituent structure changes strength, kinetics, and stripping behaviour in ways that matter at plant scale.

Aldoximes are the strong reagents: they load more copper per unit concentration, extract faster, and reject iron exceptionally well (Cu/Fe selectivity above 1000:1, often quoted 2000:1+). The trade-off is stripping — heavily loaded aldoximes need the full 150–180 g/L acid spent electrolyte to release copper, and some straight aldoximes require modifiers (esters or alcohols) to strip cleanly and to suppress the third-phase and crud tendencies that very strong reagents exhibit.

Ketoximes are the forgiving reagents: they strip easily even at moderate acidities, disengage quickly, and tolerate poorer feed cleanliness, but they load less copper per stage and extract more slowly, so a ketoxime-only circuit usually needs an extra extraction stage or higher vol% to hit the same recovery.

The decision is conditional, not absolute. For high-grade PLS (above roughly 3–4 g/L Cu) at large throughput, a modified aldoxime is preferred because its loading strength delivers more copper per stage and shrinks mixer and organic inventory for the same cathode output. For low-grade, high-acid PLS — or feeds with high silica and solids, as met in parts of Zambia and the DRC — a ketoxime or aldoxime/ketoxime blend (such as the widely used 1:1 LIX 984N-type blends) is preferred because easier stripping and better disengagement outweigh the modest strength penalty. For manganese-rich feeds, the more oxidation-tolerant option and a rigorous Mn management plan are preferred because reagent degradation, not loading capacity, becomes the controlling cost. In practice, blends have become the pragmatic default: they capture most of the aldoxime’s strength while keeping ketoxime-like strip and disengagement behaviour, and most reagent suppliers will tune the blend ratio against a client’s actual PLS during testwork.

What Do Plant Data Show?

The figures below are illustrative composites drawn from anonymised plant audits and commissioning reports across three continents; they are representative of the industry rather than attributable to any named client.

Throughput and recovery. A large Atacama oxide operation in Chile running a 3E1W2S circuit on 5 g/L Cu PLS holds 92–94% single-stage efficiency and above 96% overall Cu recovery to rich electrolyte, transferring 4.1–4.4 g/L Cu into organic at 11 vol% extractant. A Zambian Copperbelt plant on lower-grade, higher-silica PLS runs a 2E1W1S circuit at 88–91% stage efficiency — the gap is almost entirely attributable to solids and crud load, not reagent choice.

The cost of entrainment. An audit at a Sonora, Mexico plant measured 68 ppm organic-in-raffinate against a 25 ppm target. Correcting mixer shear and settler loading cut this to 22 ppm, saving an estimated 380 m³ of organic per year — worth several hundred thousand dollars annually at prevailing diluent and extractant prices, with no change to copper recovery.

The value of one more stage. A Mongolian project evaluating 2E1W1S versus 2E1W2S found the extra strip stage raised advance electrolyte tenor from 38 to 45 g/L Cu, shrinking EW tankhouse area by roughly 12% for the same cathode tonnage — the strip stage paid for itself in tankhouse capex alone.

Crud economics. A Katanga, DRC operation that installed dedicated crud centrifugation recovered about 70% of the organic held in crud, cutting organic make-up from 6.5 to 3.8 L per tonne of cathode within six months. Comparable results have been reported from Indonesian and Kazakh hydrometallurgical plants where solids management, not chemistry, was the binding constraint.

Taken together, the pattern is consistent: plants that measure entrainment and interfacial behaviour frequently, and that budget real resources for crud handling, run within 1–2 points of design recovery. Plants that treat these as housekeeping afterthoughts pay for it continuously in organic make-up and cathode quality.

Frequently Asked Questions

What is the difference between solvent extraction and electrowinning in copper processing?
Solvent extraction is a purification and concentration step: it moves Cu²⁺ from dilute, impure PLS into a strong, clean electrolyte using an organic carrier. Electrowinning is the recovery step: insoluble anodes and stainless or copper cathodes in that electrolyte plate LME-grade cathode directly. SX feeds EW; EW could not economically plate cathode from raw PLS because iron, chloride, and manganese would ruin efficiency and quality.

practical guide: Extractant Degradation

How many extraction stages does a copper SX circuit need?
Most circuits use two extraction stages against a 2E1W1S configuration, which achieves 96%+ overall copper transfer at 90% stage efficiency. The true answer comes from the extraction isotherm and McCabe–Thiele construction: high-grade PLS may justify three stages, while low-grade feeds sometimes run one. Stage efficiency (85–95%) converts theoretical to real stages — always design with real stages, not equilibrium ones.

What O/A ratio should I use in copper solvent extraction?
Extraction stages typically run 1:1 to 1.5:1 (organic:aqueous). The optimum balances copper transfer per pass against organic inventory and entrainment risk. Strip O/A is set by the electrolyte tenor you need — usually 40–50 g/L Cu advance. Pilot isotherms on your actual PLS beat any rule of thumb; a 0.1 shift in O/A can move recoveries by several points.

What causes crud in copper SX plants, and how is it treated?
Crud is a stabilised interfacial emulsion built from fine solids — silica, clays, precipitated iron — plus degradation products and sometimes biological matter. Treatment combines upstream solids removal, dedicated crud withdrawal, and processing through centrifuges, acid washing, or clay-mat filtration to recover trapped organic. Prevention (clean PLS, moderate shear, correct chemistry) always costs less than treating accumulated crud.

How much copper is lost to entrainment in a well-run plant?
Direct copper loss is small — entrained organic in raffinate carries perhaps 4–5 g/L Cu, so 50 ppm entrainment loses only grams per cubic metre — but the organic inventory loss dominates economics. Plants target below 20–50 ppm organic-in-raffinate. At large throughputs, halving entrainment typically saves more money than a full percentage point of copper recovery, because diluent and extractant are expensive consumables.

Why is iron rejection important in copper SX?
Iron transfers to electrolyte mainly via aqueous entrained in loaded organic. Each gram per litre of Fe in the tankhouse electrolyte costs current efficiency, because Fe³⁺/Fe²⁺ redox cycling consumes current that should plate copper. Effective wash stages, low aqueous-in-organic entrainment, and controlled electrolyte bleed keep Fe manageable; aldoxime extractants additionally reject iron at the chemistry level with selectivity above 2000:1.

Should I choose an aldoxime, a ketoxime, or a blend?
Let the PLS decide. High-grade, clean PLS favours modified aldoximes for maximum loading per stage. Low-grade, high-acid, or solids-laden PLS favours ketoximes or blends for easier stripping and better disengagement. Manganese-rich feeds demand oxidation-tolerant systems. Blends like the common 1:1 aldoxime/ketoxime products capture most aldoxime strength with ketoxime handling, and are the pragmatic default for most new circuits.

Conclusion

Copper solvent extraction rewards disciplined engineering more than almost any other hydrometallurgical unit operation. The chemistry is mature and forgiving; the economics are made or lost in the details — the O/A ratios that set organic inventory, the stage counts proven on the isotherm, the wash stage that protects the tankhouse, and the crud and entrainment programmes that keep losses near zero. Aldoxime versus ketoxime is not a fashion choice but a conditional decision driven by your PLS grade, impurity suite, and solids load, and the difference between the right and wrong extractant system can amount to a full extraction stage of capital.

If you are designing a new circuit, debottlenecking an existing one, or selecting an extractant for a feed that has never run through SX, the highest-return next step is structured testwork on your actual solution. Lexin-tech supplies mixer-settler and complete liquid-liquid separation equipment for copper SX-EW plants worldwide, and our engineering team supports you from isotherm to commissioning. Contact us to review your PLS chemistry and circuit options — or start with our mixer-settler design resources and request a pilot-test proposal tailored to your project.

References

1. Kordosky, G.A. (2002). “Copper recovery using leach/solvent extraction/electrowinning technology: Forty years of innovation, 1968–2002.” Journal of the Southern African Institute of Mining and Metallurgy, 102(8), 445–450.
2. Davenport, W.G., King, M.J., Schlesinger, M.E. & Biswas, A.K. (2011). Extractive Metallurgy of Copper, 5th ed. Oxford: Elsevier/PERGAMON.
3. Ritcey, G.M. (2006). Solvent Extraction: Principles and Applications to Process Metallurgy, revised 2nd ed. Ottawa: G.M. Ritcey & Associates.
4. Sole, K.C. (2016). “Copper solvent extraction: Status, operating practices and challenges in the Zambian Copperbelt.” Journal of the Southern African Institute of Mining and Metallurgy, 116(6).
5. Bartos, P.J. (2002). “SX-EW copper and the technology cycle.” Resources Policy, 28(3–4), 85–94.
6. U.S. Geological Survey (2025). Mineral Commodity Summaries 2025 — Copper. Reston, VA: USGS.
7. Proceedings of the Hydrocopper conference series (2005–2019). Santiago, Chile: GECAMIN.
8. Flett, D.S. (1974). “Solvent extraction in copper hydrometallurgy: A review.” Transactions of the Institution of Mining and Metallurgy, Section C, 83, C30–C38.

Changelog

• 2026-10-25 — First published.