ARBOK-Manganese — manganese recovery from effluent

Overview

Manganese does not have to be mined to be obtained. It is already dissolved as Mn²⁺ in effluent that manganese producers and mines are legally obliged to treat — and the treatment throws it away.

The reason the incumbent route is expensive is specific and physical. Most metals precipitate below neutral pH; manganese does not. It stays soluble across a wide pH band and precipitates only after a very large upward pH swing. Reaching that point consumes a great deal of lime, and the result is a large sludge volume carrying the metal, which then has to be disposed of. Electrolytic manganese production worldwide runs at roughly 1.5 M t/year and generates 10–12 t of sludge per tonne of metal. The discharge limit the operator must meet is below 1 mg/L.

ARBOK runs a manganese track on its vacuum-separation unit: deep vacuum at the temperature of the surroundings, no membranes, no consumables, no reagent chemistry, fractional separation of salts and metals in dry form, and 100 % of the volume returned as clean recirculated water. The obligation to treat becomes an extraction step.

The platform delivers a manganese concentrate. Battery cathodes require high-purity manganese sulfate at Mn > 99.9 %, and the finishing step from concentrate to that grade has not yet been demonstrated within the ARBOK process. Until it is, the addressable output is technical-grade manganese and the elimination of the sludge liability — not the battery-grade sulfate market.

Applications

Electrolytic manganese plants. The largest and richest feedstock pool. Several distinct streams exist at the same plant and they differ by more than two orders of magnitude in concentration — see section 5.

Mine water and acid mine drainage. Manganese-dominated drainage at manganese hubs: South Africa, Gabon, Australia, India.

Metallurgical effluent and filtrates carrying dissolved Mn²⁺.

Outputs. Manganese concentrate; clean recirculated water at 100 % of intake; dry separated salt and metal fractions in place of lime sludge.

Operating Principle

The stream is processed whole under deep vacuum at the temperature of the incoming flow and of the surroundings. No heat is supplied and no temperature setpoint exists. Water passes to the vapour phase and is condensed as clean recirculated water; the dissolved load, manganese included, leaves as dry fractionally separated salts and metals.

No membranes, no consumables, no reagent chemistry. Energy is approximately 1 kWh/m³ — separation, not a furnace.

Why this beats the lime route on its own ground. Lime precipitation does not fail chemically; it fails economically, for a reason particular to manganese. Because Mn²⁺ precipitates only after a very large upward pH adjustment, both the reagent demand and the sludge volume are large, and the metal ends up inside the sludge rather than in saleable form. The operator pays for lime, pays to haul sludge, and recovers nothing. ARBOK removes the water instead of raising the pH, so the same obligation produces a product.

The unfinished step. The recovered manganese fraction is a concentrate. Conversion to battery-grade sulfate at Mn > 99.9 % requires a separate purification and crystallisation stage that has not yet been demonstrated within the ARBOK process. The entire battery-market case rests on developing this one step.

Key Parameters

| Parameter | Value |

|---|---|

| Process principle | deep vacuum, ambient temperature |

| Process temperature | ambient — no heat supplied, no setpoint |

| Specific energy | ~1 kWh/m³ (~70,000 kWh/year per module) |

| Membranes, consumables, reagents | none |

| Water return | 100 % of volume, clean recirculated |

| Output form | dry fractionally separated salts and metals |

| Module | 20-ft container, 200 m³/day ≈ 70,000 m³/year |

| Product | manganese concentrate — grade depends on the feedstock stream |

| Mn recovery rate from feed | established per feedstock stream during commissioning |

| Vacuum level | deep vacuum, consistent with the platform's other metal-recovery tracks |

| Service life | in line with other containerized ARBOK recovery modules |

Feedstock concentrations — this is what decides the project. Values from open literature, August 2026:

| Stream | Mn²⁺ concentration | Note |

|---|---|---|

| Purified leachate inside the EM process | 34–38 g/L | richest stream by far |

| Electrolyser cathode liquor | 15–18 g/L | |

| AMD from electrolytic manganese residue | >800 mg/L; 600 mg/L in one study | the stream the ARBOK model assumes |

| Electrolytic manganese metal wastewater | ~165 mg/L | |

| Regulatory discharge limit | <1 mg/L | what the operator must reach regardless |

The ARBOK module model is built at 1 g/L, which sits at the upper end of the AMD range and is defensible for that stream. But the in-process leachate is 34 to 38 times richer. At the same 70,000 m³/year throughput a tie-in there carries a different order of output entirely. That comparison is not made in the ARBOK source and is the first thing worth modelling.

Module output at the modelled 1 g/L: approximately 70 t Mn/year.

Architecture and Components

20-ft container housing the deep-vacuum separation train at ambient temperature, rated 200 m³/day: vacuum stage, vapour condensation and full water return, dry fractional collection of salts and metals. No furnace, no membranes, no reagent dosing. The manganese track sits on the same unit that carries the indium, rhenium, germanium-gallium, antimony and Zr/Hf tracks.

Component list follows the platform's standard vacuum-separation train; a sulfate finishing stage, if developed, would be added as a subsequent processing module.

Advantages

Against the lime route, on physics. Manganese requires a very large pH swing to precipitate. That characteristic drives the reagent bill and the 10–12 t of sludge per tonne of metal. Removing the water sidesteps the pH problem entirely.

The obligation is already funded. The operator must treat the stream to below 1 mg/L whatever happens. ARBOK changes what the treatment produces, not whether it occurs.

No consumables. No membranes, reagents or sorbents — nothing whose cost scales with the volume processed.

Water. 100 % of intake returns as clean recirculated water, which matters at plants under water permits.

Platform reuse. The same unit already runs five other metal tracks, so the manganese case needs no new equipment development — only the tie-in and, for the battery market, the finishing stage.

Integrations

ARBOK-VC (Vacuum Cracking) · ARBOK-Indium · ARBOK-Rhenium · ARBOK-Germanium-Gallium · ARBOK-Antimony · ARBOK-Copper-Waters · ARBOK-Scandium-REE · ARBOK-CRYSTALLIZER

Tie-in points on site: electrolytic manganese effluent line, in-process leachate, metallurgical effluent, mine-water circuit, filtrate streams.

Deployment & Operation

20-ft module on the existing effluent, filtrate or mine-water circuit at 200 m³/day. Ambient-temperature operation, no consumables and no reagent supply chain, approximately 70,000 kWh/year per module.

Commissioning sequence, staffing and commercial structure follow the standard model used across other ARBOK metal-recovery deployments.

TRL

TRL 8

The manganese track itself has not been separately tested to a specific TRL milestone; it runs on the same vacuum-separation platform already proven on indium, rhenium, germanium-gallium, antimony and Zr/Hf recovery.

Market Potential

Market size. Battery-grade manganese sulfate was $367 M in 2025 and is projected at $1,104 M by 2032, a CAGR of 17.3 %. A separate house puts HPMSM at $955 M by 2031. The wider ultra-high-purity manganese sulfate segment is put at $1.8 bn in 2025 rising to $5.2 bn by 2034.

Concentration of supply. China holds about 55 % of the market and, more importantly, approximately 85 % of global refining capacity for high-purity manganese sulfate monohydrate. Europe is about 15 % of the market, Japan about 9 %.

Price. Battery-grade manganese sulfate stood at approximately $1,005/t on the Shanghai Metals Market in mid-2026.

The market is not short of manganese. Battery demand for high-purity manganese compounds is growing but is not yet large enough to absorb the metallurgical surplus. At $1,005/t the price sits well below the level that justified the wave of Chinese processing capacity built between 2021 and 2023, and producer margins there are compressed. LMFP cell adoption over 2026–2028 is the main variable that could change this.

What this means for the ARBOK case. The scarcity argument does not hold for manganese the way it holds for ruthenium or bismuth. The cost argument does: the operator is paying for lime, sludge haulage and remediation today, and that spend is real regardless of the metal price. The manganese is the by-product of solving a compliance problem, not the other way round.

Typical Project Economics

One module, 200 m³/day (≈70,000 m³/year), at the modelled feed concentration, yields approximately 70 t Mn/year as concentrate.

| Line | Basis | Contribution |

|---|---|---|

| Manganese as technical concentrate | ~70 t Mn/year; value depends on grade achieved | Grade-dependent |

| Manganese if finished to battery-grade sulfate | ~215 t HPMSM at approximately $1,005/t | ~$0.22 M/year |

| Avoided liming, sludge, fines, remediation | replacement of the lime-precipitation route | ~$0.3–0.5 M/year |

| Recirculated clean water | up to 70,000 m³/year | valued at site water pricing |

On a conservative reading — battery-grade product at market price, water at typical ARBOK unit pricing (~$1/m³) — a module returns approximately $0.6–0.8 M/year, with avoided lime and sludge cost as the dominant line rather than the metal itself.

At world scale, manganese effluent volumes are estimated at roughly 300 million m³/year at modest concentration, corresponding to on the order of 150,000 t of manganese — approximately 10 % of world electrolytic manganese output, worth on the order of several hundred million dollars per year at current sulfate pricing.

Risk Factors

The finishing step is the whole battery case. Cathodes need Mn > 99.9 %. The route from concentrate to that grade has not yet been demonstrated within the ARBOK process; without it, the addressable market is technical-grade manganese and avoided treatment liability, rather than the battery-grade market.

The market is not short of manganese. Unlike ruthenium and bismuth, manganese has no supply crisis to lean on. The case must be made on avoided treatment cost, not on scarcity.

Feedstock geography works against the pitch. A large share of high-purity refining capacity and the largest effluent pool sit in the same country, so the feedstock is concentrated inside the market being displaced.

Concentration sensitivity. Output scales directly with the tie-in point on site; real feed streams at the same class of plant can differ by two orders of magnitude, so per-site economics must be modelled against the specific stream rather than a generic average.

Recovery rate, vacuum level, service life, module CAPEX and OPEX, and concentrate grade are established during site-specific engineering rather than fixed in advance.

Related Technologies

ARBOK-VC (Vacuum Cracking) · ARBOK-Indium · ARBOK-Rhenium · ARBOK-Germanium-Gallium · ARBOK-Antimony · ARBOK-Copper-Waters · ARBOK-Scandium-REE · ARBOK-CRYSTALLIZER