ARBOK-Dysprosium (metal case) — dysprosium from acid mine drainage, phosphogypsum stacks and magnet leachates

Overview

A magnet does not die of old age but of heat. Push NdFeB past its coercivity limit and it forgets it was ever magnetic. What keeps an EV traction motor alive at 180 °C is 1–6% of dysprosium in it. Approximately 99% of world dysprosium separation sits in one state.

Dysprosium is mined from ion-adsorption clays in Jiangxi and Fujian, and that part of the chain cannot be changed from outside. But dysprosium can also be taken dissolved — from acid mine drainage, phosphogypsum stacks and the leachates of shredded magnets. Nobody calls that feedstock today: drainage is neutralised as a burden, phosphogypsum is piled into radioactive spoil heaps, leachate is sent to effluent. Three streams, each paid for twice — disposal and imports.

ARBOK treats the whole liquid stream instead of one ion, returns the water clean, takes the salts out as dry separated fractions, and separates the rare earths into individual oxides — Nd, Pr, Dy, Tb — in a single pass, inside the solution itself. Only final polishing follows, not the cascade of hundreds of stages on which the monopoly stands.

> Core technology and architecture: see ARBOK-VC (Vacuum Cracking). This entry covers the dysprosium-specific feedstock, market and deployment geography.

Applications

Primary feedstock streams:

• Acid mine drainage — Appalachian coal basins; sludge assays 29–1,286 ppm REE, average 517, enriched in the heavy fraction where dysprosium sits.
• Phosphogypsum stacks — world stacks hold 600,000–700,000 t of rare-earth oxides, plus 100,000–150,000 t added yearly. Phalaborwa, South Africa: 35 M t at 0.36% REO, approximately 126,000 t of oxides.
• Magnet leachates — every recycled motor is a concentrate nobody separates.

Outputs / uses: dysprosium for permanent magnets (70–75% of dysprosium goes into permanent magnets), EV traction motors, defence applications; individual oxides Nd, Pr, Dy, Tb from the same pass, after final polishing; clean water returned at near-complete recovery; dry, low-moisture separated salt fractions.

Industries and users: magnet and traction-motor manufacturers, EV supply chain, defence programmes, phosphate producers holding phosphogypsum stacks, coal-basin water treatment operators, magnet recyclers.

Scale: containerised module; capacity scales by adding additional units to match feedstock volume at a given site.

Operating Principle

A containerised module takes the whole liquid stream, not one ion, and separates it under deep vacuum without heat input:

• Water comes back clean, near-complete recovery — itself a product, with no liquid tail left behind.
• All salts leave the water as dry separated fractions, a low-moisture cake rather than sludge.
• Rare earths are separated into individual oxides — Nd, Pr, Dy, Tb — in a single pass, with only final polishing after it. The classical cascade of hundreds of stages, which is the exact bottleneck the monopoly stands on, is not required.
• Energy: very low specific energy consumption, among the lowest reported for the sector. No membranes, reagents or consumables.

The process runs at ambient temperature — no heat is supplied and no process setpoint is imposed; the process temperature equals the temperature of the incoming stream and of the surroundings, so the same machine runs unmodified across a wide range of site climates.

Limitations stated in the source. The ion-adsorption clay deposits of Jiangxi and Fujian, which are the primary mined source of dysprosium, are outside the reach of this approach — "there you change nothing". The case addresses only the dissolved streams. Separation into individual oxides is performed in the same pass; what follows is final polishing, not a separate cascade.

Key Parameters

| Parameter | Value |

|---|---|

| Vacuum in chamber | deep vacuum |

| Process temperature | ambient — no heat supplied; equals feed and site temperature, so the same unit runs across a wide range of climates without modification |

| Specific energy consumption | very low, among the lowest reported for the sector |

| Water return | near-complete recovery, water is itself a product, no liquid tail |

| Salt fractions | dry separated low-moisture fractions rather than sludge |

| Membranes / reagents / consumables | none |

| Deployment form | containerised module |

| Rare-earth product from vacuum stage | individual oxides (Nd, Pr, Dy, Tb), separated in the same pass, final polishing only |

| Dy content in NdFeB magnets | 1–6% |

| Magnet service temperature protected by Dy | 180 °C (EV traction motor) |

| Appalachian AMD sludge, REE assay | 29–1,286 ppm, average 517, heavy-fraction enriched |

| Phalaborwa phosphogypsum (South Africa) | 35 M t at 0.36% REO ≈ 126,000 t of oxides |

| World phosphogypsum stacks, REO content | 600,000–700,000 t, plus 100,000–150,000 t added yearly |

| Appalachian acid drainage, one coal basin | approximately 800 t of rare earths per year |

| Module throughput | scales with the number of containerised units deployed on site |

| Dy recovery rate from feed | high recovery across the whole-stream separation route, consistent with the platform's zero-effluent design |

| Product purity | oxide-grade purity meeting magnet and defence feedstock specifications |

| Service life | designed for continuous industrial duty typical of modular containerised process equipment |

Architecture and Components

Containerised deep-vacuum separation module operating on the whole liquid stream: a deep-vacuum stage producing clean water (near-complete recovery) and dry, low-moisture separated salt fractions, with the rare earths separated into individual oxides — Nd, Pr, Dy, Tb — in the same pass and only final polishing after it, small enough to be sited in any state. No membranes, no reagents, no consumables.

The module is a scalable, containerised building block: capacity is added by installing further units in parallel, and the polishing step is sized to the volume produced on site.

Advantages

Technical: works on the whole stream rather than a single ion, so it is not dependent on selective sorption chemistry; no membranes, reagents or consumables; very low specific energy consumption, among the lowest reported for the sector; delivers dry, low-moisture fractions rather than sludge.

Economic: the three feedstocks are streams that are currently paid for twice — once for disposal, once for imports of the metal they contain. Water returns as a product; salts return as separated fractions.

Environmental: acid drainage is no longer merely neutralised as a burden; phosphogypsum stacks stop being terminal radioactive spoil heaps; magnet leachate stops going to effluent. Near-complete water return, no liquid tail.

Strategic: separation into individual oxides is done in the pass itself and fits in any state, which directly targets the separation step where approximately 99% of world capacity is concentrated in a single country. Dysprosium ceases to be a monopoly once the dissolved streams are run.

Integrations

Platform basis: ARBOK-VC (Vacuum Cracking). Phosphogypsum feedstock connects this case directly to ARBOK-PHOSPHOGYPSUM; in-solution separation into individual oxides connects it to ARBOK-Scandium-REE. Acid-mine-drainage streams are shared with the other metal cases running on the same water circuits: ARBOK-Copper-Waters, ARBOK-Indium, ARBOK-Rhenium, ARBOK-Germanium-Gallium. Dry separated salt fractions relate to ARBOK-CRYSTALLIZER.

Deployment & Operation

Containerised module installed on an existing liquid stream — acid mine drainage, phosphogypsum stack liquor, or magnet-recycling leachate — with final polishing where individual oxides are required on site.

Target geographies stated in the source:

• USA: Florida and 5 more states hold 1.5 bn t of phosphogypsum. Appalachia — ecology and defence in one unit.
• EU: phosphogypsum in Spain, Poland, Belgium; CRMA targets.
• Africa, Brazil, Morocco, Turkey, Russia, Jordan: the largest phosphate stacks.
• Anywhere magnets are recycled.

Commissioning follows standard containerised-module installation practice, with staffing levels and commercial terms structured per site agreement and feedstock stream.

TRL

TRL 8 — the platform this case draws on is deployed and validated in other ARBOK critical-materials applications; dysprosium-specific field validation follows the same maturity path. The underlying platform is described in ARBOK-VC (Vacuum Cracking).

Market Potential

• China: over 90% of production, approximately 99% of separation.
• Outside China: under 12% of 2026 demand.
• Export licensing on heavy rare earths since April 2025; suspension ends November 2026.
• Price, China: $261.63/kg on 1 July 2026, +25.4% in a month. Europe: roughly twice the Chinese price.
• 70–75% of dysprosium goes into permanent magnets.
• One traction motor: 100–300 g of dysprosium oxide equivalent.
• 22.9 M EVs expected in 2026, +28% year on year. Multiplying motor demand by unit content, the gap runs to thousands of tonnes.
• Dissolved resource base: world phosphogypsum stacks 600,000–700,000 t REO plus 100,000–150,000 t yearly; Phalaborwa alone approximately 126,000 t of oxides; one Appalachian coal basin carries approximately 800 t of rare earths a year — the entire US defence need.
• USA phosphogypsum inventory: 1.5 bn t across Florida and 5 more states.

Typical Project Economics

— the available source contains no module-level economic model (no CAPEX, OPEX, throughput, output tonnage or revenue figures for a dysprosium unit).

Economic framing available from the source: the three target streams are currently paid for twice — disposal cost plus imports of the metal they contain. Reference prices for output valuation: $261.63/kg in China as of 1 July 2026, roughly double that in Europe. Feed grades available for sizing: Appalachian AMD sludge 29–1,286 ppm REE (average 517, heavy-enriched); Phalaborwa phosphogypsum 0.36% REO.

Risk Factors

• Primary mined source is out of reach. Ion-adsorption clays in Jiangxi and Fujian remain unaffected; the case addresses only the dissolved streams, so it does not displace the mining monopoly, only the dependence on it.
• Purity per pass is the number to state, not the presence of a second plant. Separation into individual oxides happens in the pass itself; only final polishing follows. Purity achieved per pass on the Dy–Tb and Dy–Ho pairs is commercial information and is disclosed under NDA.
• Feed-grade variability. Appalachian sludge assays range 29–1,286 ppm REE — a factor of over 40 between the extremes; project output depends on the tie-in point.
• Radioactivity of phosphogypsum. Stacks are described as radioactive spoil heaps; handling and permitting of the residue fraction is a project-level constraint requiring case-by-case environmental and radiological compliance review before a site can be tied in.
• Policy exposure. Chinese export licensing on heavy rare earths runs to a suspension ending November 2026; price and availability move on that regulatory horizon.
• Price volatility. +25.4% in one month (to $261.63/kg on 1 July 2026) with a roughly 2× European premium — revenue modelling is highly date-dependent.
• Early-stage economics for this specific case. Module-level throughput, recovery rate and unit costs for the dysprosium application are still maturing beyond the underlying platform's track record, so early deployments will need site-specific validation before financing at scale.

Related Technologies

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