Crystallization

ARBOK-Bi₂Te₃ — joint tellurium and bismuth recovery

On 4 February 2025, by Announcement No. 10 of the Ministry of Commerce and the General Administration of Customs, China imposed export controls on items related to tungsten, tellurium, bismuth, molybdenum and indium.

ARBOK-Bi₂Te₃ — joint tellurium and bismuth recovery

Technology brief

What this platform addresses

On 4 February 2025, by Announcement No. 10 of the Ministry of Commerce and the General Administration of Customs, China imposed export controls on items related to tungsten, tellurium, bismuth, molybdenum and indium.

Platform validated; the paired configuration is a feedstock-level model pending site assay

The challenge

The problem this technology addresses

Communications and data-centre optics. The thermoelectric cooler under the laser diode in an optical transceiver stabilises wavelength. Present at very large unit volume and rising with data-centre build-out.

Thermal imaging. Cooled infrared detectors and arrays.

Scientific and astronomical cameras. Sensor cooling.

Space. Instrument cooling where a compressor is mechanically and acoustically unacceptable.

Medicine. PCR thermocyclers, sample transport.

Waste heat. Thermoelectric generation.

ARBOK solution

How the ARBOK system creates value

On 4 February 2025, by Announcement No. 10 of the Ministry of Commerce and the General Administration of Customs, China imposed export controls on items related to tungsten, tellurium, bismuth, molybdenum and indium. Two of those five positions are the two elements that constitute Bi₂Te₃ — the only material that produces refrigeration at room temperature with no compressor, no working fluid and no moving parts. One notice, one day, both ends of one substance behind a licence. The point appears not to have been made publicly in the eighteen months since.

Bi₂Te₃ is 52% bismuth and 48% tellurium by mass. Neither metal is mined anywhere. Both are by-products of base-metal refining, both are collected from anode slimes and refinery liquors, and both are dissolved in the same effluent of the same plant.

> This card covers the paired case. For the single-metal cases see ARBOK-Tellurium and ARBOK-Bismuth.

Deep-vacuum phase separation at ambient temperature, drawing no external heat, below 1 kWh/m³ net with condensation heat recuperated at ~98%.

Because the process acts on the solvent rather than the solute, cost per cubic metre is set by the energy of phase change and is independent of concentration or matrix. Refinery effluent is high-salinity, low-pH and multi-metal — a solution in which selective sorbents foul and selective precipitants are consumed by the matrix rather than the target. A phase change is indifferent to that matrix; the ballast reports as a product.

Outputs: clean water at 100% of input volume, zero discharge; dry ballast fractions separated by density; and a concentrate carrying both bismuth and tellurium together with the copper, lead, selenium and silver mobilised in the same stream. No roasting, therefore no SO₂ and no CO₂ — in contrast to conventional slime treatment, a recognised emissions point in copper refining.

Why substitution is not available. Thermoelectric figure of merit ZT peaks over a narrow temperature band for any material system. Bi₂Te₃ and its alloys peak near ambient. Lead telluride, skutterudites, half-Heuslers and SiGe have their useful ranges several hundred degrees higher — above the point at which Bi₂Te₃ has already melted. They are not substitutes; they occupy a different part of the temperature axis.

Market and application

Commercial opportunity

Supply of both metals is set by the copper and lead production plan, not by demand for either. A doubling of Bi₂Te₃ demand mines nothing new, because there is nothing to mine. Supply is inelastic by construction, not by lag. The only available response is more complete extraction from material already dissolved inside a refinery.

Geographies: copper refineries in Chile, Peru, the USA, Poland and Zambia; lead refineries; thermoelectric module manufacturers outside China; anyone building data centres whose transceiver lasers now sit behind a Chinese export licence.

Basis: one module, 200 m³/day, 73,000 m³/year, on copper- or lead-refinery effluent.

| Line | Value per year |

|---|---|

| Bismuth at 0.5 g/L — 37 t | $2.5 M |

| Tellurium at 0.1 g/L — 7 t | $1.8 M |

| Copper, lead, selenium, same pass | ~$0.7 M |

| Silver, same pass | ~$0.7 M |

| Water 73,000 m³ at $1–15/m³, plus neutralisation dropped | ~$0.6–1.5 M |

| Total from one container | $6–7 M/year |

Use cases

Where the technology can be applied

Communications and data-centre optics. The thermoelectric cooler under the laser diode in an optical transceiver stabilises wavelength. Present at very large unit volume and rising with data-centre build-out.

Thermal imaging. Cooled infrared detectors and arrays.

Scientific and astronomical cameras. Sensor cooling.

Space. Instrument cooling where a compressor is mechanically and acoustically unacceptable.

Medicine. PCR thermocyclers, sample transport.

Waste heat. Thermoelectric generation.

Effluent assay for both metals → module count against measured flow → installation on the effluent line → commissioning → concentrate to the plant's existing refining, or to a thermoelectric-grade purification partner. Anode slime circuit added where the compound, rather than the metals, is the objective.

The tie-in is to streams the refinery already handles, so installation does not disturb the precious-metals flowsheet on which the plant's primary economics depend.

Same platform and largely the same feedstock as ARBOK-Tellurium, ARBOK-Bismuth and ARBOK-GOLD. Sits alongside ARBOK-Copper-Waters. Platform: ARBOK-VC (Vacuum Cracking).

View preserved source description

Overview

On 4 February 2025, by Announcement No. 10 of the Ministry of Commerce and the General Administration of Customs, China imposed export controls on items related to tungsten, tellurium, bismuth, molybdenum and indium. Two of those five positions are the two elements that constitute Bi₂Te₃ — the only material that produces refrigeration at room temperature with no compressor, no working fluid and no moving parts. One notice, one day, both ends of one substance behind a licence. The point appears not to have been made publicly in the eighteen months since.

Bi₂Te₃ is 52% bismuth and 48% tellurium by mass. Neither metal is mined anywhere. Both are by-products of base-metal refining, both are collected from anode slimes and refinery liquors, and both are dissolved in the same effluent of the same plant.

> This card covers the paired case. For the single-metal cases see ARBOK-Tellurium and ARBOK-Bismuth.

Applications

Communications and data-centre optics. The thermoelectric cooler under the laser diode in an optical transceiver stabilises wavelength. Present at very large unit volume and rising with data-centre build-out.

Thermal imaging. Cooled infrared detectors and arrays.

Scientific and astronomical cameras. Sensor cooling.

Space. Instrument cooling where a compressor is mechanically and acoustically unacceptable.

Medicine. PCR thermocyclers, sample transport.

Waste heat. Thermoelectric generation.

Operating Principle

Deep-vacuum phase separation at ambient temperature, drawing no external heat, below 1 kWh/m³ net with condensation heat recuperated at ~98%.

Because the process acts on the solvent rather than the solute, cost per cubic metre is set by the energy of phase change and is independent of concentration or matrix. Refinery effluent is high-salinity, low-pH and multi-metal — a solution in which selective sorbents foul and selective precipitants are consumed by the matrix rather than the target. A phase change is indifferent to that matrix; the ballast reports as a product.

Outputs: clean water at 100% of input volume, zero discharge; dry ballast fractions separated by density; and a concentrate carrying both bismuth and tellurium together with the copper, lead, selenium and silver mobilised in the same stream. No roasting, therefore no SO₂ and no CO₂ — in contrast to conventional slime treatment, a recognised emissions point in copper refining.

Why substitution is not available. Thermoelectric figure of merit ZT peaks over a narrow temperature band for any material system. Bi₂Te₃ and its alloys peak near ambient. Lead telluride, skutterudites, half-Heuslers and SiGe have their useful ranges several hundred degrees higher — above the point at which Bi₂Te₃ has already melted. They are not substitutes; they occupy a different part of the temperature axis.

Key Parameters

| Parameter | Value |

|---|---|

| Bi₂Te₃ composition | 52% Bi, 48% Te by mass |

| Raw metal content of the compound | ~$151/kg at August 2026 prices |

| Tellurium price | $243/kg (Aug 2026), +67% in a year |

| Bismuth price | $67/kg (Aug 2026), +25% in a year, +216% in two — 17-year high |

| Tellurium world output | 500–750 t/year — less than gold |

| Bismuth world output | ~16,000 t/year — less than one day of copper mining |

| Chinese share | Te ~75% of refining; Bi ~80% of output and refining |

| Trade status | Both under Chinese export licence since 4 February 2025, Announcement No. 10 |

| Module | standard containerized unit, 200 m³/day = 73,000 m³/year |

| Energy | < 1 kWh/m³ net, ambient temperature, deep vacuum |

Architecture and Components

A standard containerized module on the refinery effluent line: vacuum separation stage, condensate return with full water recycle, density-separated dry fractions, combined Bi/Te concentrate take-off. Installed in parallel to match stream volume. No membrane cleaning cycle, no resin regeneration downtime, no reagent inventory. A second circuit on the anode slime line is required for any project aimed at the compound (see section 7).

Integrations

Same platform and largely the same feedstock as ARBOK-Tellurium, ARBOK-Bismuth and ARBOK-GOLD. Sits alongside ARBOK-Copper-Waters. Platform: ARBOK-VC (Vacuum Cracking).

Deployment & Operation

Effluent assay for both metals → module count against measured flow → installation on the effluent line → commissioning → concentrate to the plant's existing refining, or to a thermoelectric-grade purification partner. Anode slime circuit added where the compound, rather than the metals, is the objective.

The tie-in is to streams the refinery already handles, so installation does not disturb the precious-metals flowsheet on which the plant's primary economics depend.

TRL

Platform validated. No TRL is recorded for the paired configuration; the figures here are a feedstock-level model built on platform parameters, with project-level output established from site assay.

Market Potential

Supply of both metals is set by the copper and lead production plan, not by demand for either. A doubling of Bi₂Te₃ demand mines nothing new, because there is nothing to mine. Supply is inelastic by construction, not by lag. The only available response is more complete extraction from material already dissolved inside a refinery.

Geographies: copper refineries in Chile, Peru, the USA, Poland and Zambia; lead refineries; thermoelectric module manufacturers outside China; anyone building data centres whose transceiver lasers now sit behind a Chinese export licence.

Typical Project Economics

Basis: one module, 200 m³/day, 73,000 m³/year, on copper- or lead-refinery effluent.

| Line | Value per year |

|---|---|

| Bismuth at 0.5 g/L — 37 t | $2.5 M |

| Tellurium at 0.1 g/L — 7 t | $1.8 M |

| Copper, lead, selenium, same pass | ~$0.7 M |

| Silver, same pass | ~$0.7 M |

| Water 73,000 m³ at $1–15/m³, plus neutralisation dropped | ~$0.6–1.5 M |

| Total from one container | $6–7 M/year |

Risk Factors

The tellurium grade in effluent is an assumption, not a measurement — 0.1 g/L is a working figure and is the single number in this case most in need of site assay. It is also precisely the number no refinery currently collects, because tellurium in effluent is not a product and is therefore not assayed.

The bismuth grade at 0.5 g/L is taken from the existing ARBOK-Bismuth model and is itself a range.

Water tariff at $1–15/m³ is a wide band that swings the total by nearly $1 M; it is jurisdiction-specific.

Thin markets. Tellurium at 500–750 t/year and bismuth at ~16,000 t/year are small enough that volume released without regard to price would move it. Deployment should target supply security for identified offtake rather than maximum tonnage.

Operator priorities. Gold and silver govern the refinery business case. Both tellurium and bismuth have historically been recovered only where plant-specific economics justified an extra circuit, and the same logic applies to adopting a new one.

Arsenic. Bismuth-bearing refinery streams frequently carry arsenic; it is immobilised in the same pass, but handling carries permitting exposure regardless.

Related Technologies

ARBOK-Tellurium · ARBOK-Bismuth · ARBOK-GOLD · ARBOK-Copper-Waters · ARBOK-VC (Vacuum Cracking)

Related technologies

Explore adjacent ARBOK systems

TEGFIL (TEG Filter)
CrystallizationTRL 8–9: Deployment-ready

TEGFIL (TEG Filter)

Sludge from mine water and landfill leachate is normally paid for twice: once to generate it through neutralization or filtration, and again to haul it…

Partnership pathway

Evaluate ARBOK-Bi₂Te₃ — joint tellurium and bismuth recovery for your application or pilot site.