Inside gold’s hot strip: Zadra vs. AARL, and the levers that make elution work

Miners strip gold from loaded carbon with hot caustic cyanide solutions, and two industrial methods lead the pack. In the Zadra process, carbon is treated with a hot solution of ~1–2% NaOH and 0.2–0.5% NaCN (w/v) at atmospheric pressure, traditionally ~95–100 °C, with typical elution times of 18–72 hours and simultaneous electrowinning (electrolytic gold recovery) (www.researchgate.net) (www.researchgate.net) (911metallurgist.com).

By contrast, the Anglo American Research Lab (AARL) process runs a multi-step cycle: an acid wash (≈3% HCl, 1 hour) to strip Ca/Mg salts, a preheat with ~3% NaCN + 1–2% NaOH at ≈110 °C for ~30 minutes, then a hot water flush at ≈120 °C for ~2.5 hours (≈5 bed‑volumes, where a bed-volume is the packed carbon volume), followed by electrowinning afterwards (www.researchgate.net) (911metallurgist.com).

Performance differences are practical: Zadra is prized for a simple flow-sheet but often longer cycle times and high heat input. A Ghana Goldfields plant initially ran a pressurized Zadra at 125 °C (3% NaOH, 64 m³/h) with a 16‑hour cycle (≈30 batches/month), then depressurized to 120 kPa, which halved cycle time to ~13 hours (≈2 batches/day) as operators chased better stripping stability (www.researchgate.net) (www.researchgate.net).

AARL’s decoupled electrowinning and high water usage deliver very low residual gold on carbon: after 5 bed‑volumes the carbon assay falls to ≈30 g/t (down from ~1000–1500 g/t original), i.e., roughly 98–99% of the gold removed in ~2.5 hours at 110–120 °C (www.researchgate.net) (911metallurgist.com).

Process flow and operating steps

The Zadra process treats loaded carbon with hot caustic-cyanide (≈1–2% NaOH and 0.2–0.5% NaCN w/v) at ≈95–100 °C and atmospheric pressure, with electrowinning done simultaneously during elution; typical strip times span 18–72 hours (www.researchgate.net) (www.researchgate.net).

The AARL cycle begins with acid washing (≈3% HCl, 1 hour) to remove inorganic scale, followed by rinsing and preheating with ≈3% NaCN + 1–2% NaOH at ≈110 °C for ~30 minutes, then a hot water flush (≈120 °C) for ~2.5 hours (~5 bed‑volumes). Electrowinning is performed afterwards on this eluate (www.researchgate.net) (911metallurgist.com).

Temperature and pressure as primary levers

Higher temperature greatly accelerates desorption. Empirical data show that raising preheat/elution temperature from ~25 °C to 80–90 °C can increase desorption from ~65% to ~78% per cycle (e.g., after 6 bed‑volumes) (www.researchgate.net). AARL circuits target ~110 °C “top‑of‑column” for optimal release, while Zadra typically runs near boiling (911metallurgist.com).

Excess temperature can decompose cyanide (especially >100 °C), so pressurization is often used: raising Zadra from ambient to ≈120–350 kPa allows operation at ≈125–140 °C, significantly speeding elution (www.researchgate.net) (www.researchgate.net). After pressure was reduced from 350 to 120 kPa at Ghana’s plant, elution cycles shortened from ~16 hours to ~13 hours (www.researchgate.net).

Efficiency correlates with temperature rather than pressure per se, but stable pressure and flow matter. Fluctuations in pressure/flow, such as from seals or pumps, have been observed to reduce single‑pass recovery; Ghana’s operation reported “fluctuations in flow/temperature reduced stripping efficiency” until stability improved (www.researchgate.net). The key point is maintaining the set temperature safely (via pressure control or heat exchangers).

Eluant chemistry: caustic and cyanide

NaOH keeps the gold‑cyanide complex soluble by maintaining high pH (≈12–13). Typical eluant NaOH is ~1–3% w/v. AARL preheat uses ~1–2% NaOH (and 3% NaCN), while Zadra historically uses ~1–2% as well (www.researchgate.net) (911metallurgist.com) (www.researchgate.net) (www.researchgate.net). Too little NaOH (<1%) leads to Au hydroxide precipitation; too much (>4%) can increase viscosity and carbonate formation with CO₂, potentially lowering electrowinning efficiency. In practice, target pH is ~12–12.5 after elution, and operators dose extra NaOH at the end of AARL to bring the eluate to pH 12.5 (911metallurgist.com).

Cyanide supplies free CN⁻ to reform Au(CN)₂⁻ during desorption. Zadra often adds only 0.2–0.5% NaCN, whereas AARL preheat uses ~3% NaCN (with ~1% NaOH). Zadra relies more on residual CN on carbon; AARL floods with strong cyanide. Insufficient cyanide leads to “preg‑robbing” (gold re‑adsorbing on carbon or plating onto vessels) and slows elution; excess CN (>5%) adds cost and toxicity with little gain. Cyanide is consumed by side‑reactions; higher temperatures and longer contact times accelerate HS⁻ and OCN⁻ formation, so CN must be replenished or elution time limited (www.researchgate.net) (www.researchgate.net) (911metallurgist.com) (www.researchgate.net).

Accurate dosing of caustic and cyanide underpins stability; many plants standardize this control with a dosing pump for precise chemical addition.

Hydraulics, bed‑volumes, and carbon quality

Inflow rate and contact time must match kinetics. The Granites (AARL) circuit uses ~5 bed‑volumes in 2.5 hours (~2 BV/h) to hit residual Au ≈30 g/t; slower flow (more BV) can further complete elution but raises energy cost (911metallurgist.com).

Carbon fines and channeling influence efficiency. Dense beds prevent short‑circuiting, while attrition can plug screens. Bed baffle design and screen sizes must be maintained. Ghana’s plant replaced 600→700 μm strainers to manage fine carbon and stabilize operation (www.researchgate.net). Inline protection can be supported with a strainer to keep fines from blanking screens.

Active carbon matters: modern activated carbons (≈500–1000 m²/g) desorb gold more readily than old or fouled carbons; carbon reactivation burns at ~550–750 °C if exhausted (911metallurgist.com). Plants often standardize carbon specifications, and sourcing quality activated carbon aligns with these desorption targets.

Troubleshooting common elution issues

• Low gold recovery (high residual Au on carbon): Often associated with insufficient heat or reagent concentration. Verification that the eluant reaches design temperature (e.g., 110–120 °C in AARL or ~95–100 °C in Zadra) and that NaOH/NaCN dosages are within range typically resolves the shortfall. If carbon fouling is likely (Ca, Mg, Cu), the acid wash step should be effective; a standard 3% HCl soak (1 hour, 1 BV) removes >95% of Ca carbonate (www.researchgate.net). Assays showing residual carbonate support increasing acid‑volume or contact time. Hard channels or short‑circuiting in the column can also be causal. In Ghana’s Zadra plant, uneven flows and temperature surges coincided with lower extraction; stabilizing flow via pressure changes restored efficiency (www.researchgate.net) (www.researchgate.net). If CN⁻ is exhausted early (very high T or extended contact), shortening elution or adding CN mid‑cycle addresses the deficit.
• High reagent consumption (CN loss): Excess cyanide use often signals disproportionation or thiocomplex formation. Responses can include lowering elution temperature slightly or adding oxygen sparging to prevent HCN volatilization and keep more CN⁻ in solution. Pressurizing Zadra (running at 120–140 °C) can reduce time and thus net CN decay; Ghana’s shift from 350→120 kPa improved throughput while stabilizing operation (www.researchgate.net). High pH during elution mitigates HCN formation.
• Carbon fouling/blockages: Flow drops or column pressure spikes may point to fines or precipitate accumulation. Regular cleaning or renewal of strainers to prevent <600 μm fines from blanking screens is common practice (www.researchgate.net). Effective acidification avoids Ca/Mg scale that can cake the bed and reduce desorption (www.researchgate.net). If deposits persist, higher bed‑volume washes or alternative acid chemistries (e.g., HCl+H₂SO₄) are sometimes adopted.
• Electrowinning anomalies: In Zadra, poor plating can mask elution trouble since EW is simultaneous. Correct EW pH (usually pH ≥10) is a prerequisite. If dissolved Au remains >1 g/t after 16 hours EW (an AARL target), the elution stage may have under‑delivered (911metallurgist.com). Re‑running elution with higher NaCN or longer contact is commonly used. Co‑precipitated base metals (Cu, Zn) in EW can indicate insufficient carbon wash or CN detox steps.
• Carbon degradation: Over‑heating or mechanical stress damages carbon. If capacity drops over cycles, chronic operation above design temperatures (e.g., >125 °C unless carbon is rated) may be causal. Lower agitation rates reduce attrition. High caustic can accelerate carbon oxidation, elevating carbon consumption.

Control, monitoring, and incremental gains

Stable control of temperature and flow is paramount. Regular monitoring of eluate chemistry (Au ppm, pH, CN total) and carbon assays supports incremental optimization. Small adjustments can have measurable effects: raising preheat from 25→80 °C boosted elution recovery ~20% (www.researchgate.net); lowering set pressure from 350 to 120 kPa slashed cycle time by ~20% in reported practice (www.researchgate.net).

Sources and reference values

Standard process references and recent studies of gold CIP elution underpin the data cited here. Key data: Oladele et al. 2015 (SAIMM) on temperature effects (www.researchgate.net); Beyuo & Abaka‑Wood 2016 (UMaT conf.) on a Ghana Zadra plant (www.researchgate.net) (www.researchgate.net); Faghirinejad et al. 2025 (J. Chem. Technol. Metall.) on typical Zadra/AARL conditions (www.researchgate.net) (911metallurgist.com); Forson et al. 2018 (UMaT conf.) on acid wash efficiency (www.researchgate.net); and industry practice notes (911metallurgist.com) (911metallurgist.com). These sources give concrete values (temperatures, concentrations, residual assays) to guide design and troubleshooting.