ARBOK-Praseodymium (metal case) — praseodymium from magnet leachates, magnet-plant effluent and phosphogypsum

Overview

Praseodymium is almost never sold as praseodymium. It leaves the refinery bound to neodymium as the NdPr alloy known as didymium, because the two elements are adjacent lanthanides whose solvent-extraction separation factor is close to unity — splitting them demands hundreds to thousands of counter-current stages. That single fact, not ore and not geology, is the structure of the market: whoever holds the Nd/Pr separation train holds the permanent-magnet industry. Approximately 85% of neodymium mining and approximately 90% of world NdFeB magnet output sit in one jurisdiction. In 2026 the price record made the point without commentary: NdPr oxide fell to roughly $90/kg on 1 June and recovered to $133.02/kg on 1 July, a 21.4% monthly move and a yearly high; praseodymium metal traded at $245.40/kg, up 112.65% year on year. On 22 June 2026 ten United States entities, including MP Materials and USA Rare Earth, were added to Chinese export-control lists, and MOFCOM Announcement No. 26 of 2026, effective 1 July, created a reporting and reward mechanism for suspected export-control violations.

The case argues that the praseodymium bottleneck is a separation bottleneck sitting downstream of feedstock that is already dissolved and already in hand. Three aqueous streams carry the Nd/Pr pair today and are managed as waste: the acid leachate of end-of-life NdFeB magnets (28–32% rare earths by weight of the original magnet, with recycling of end-of-life magnets worldwide still under 1%); the wash waters and grinding effluent of magnet manufacture, carrying up to 10% of magnet mass as fine swarf and dissolved metal; and phosphogypsum, a world flow of approximately 250 Mt/year at roughly 0.3% REO that contains on the order of 30,000 t of praseodymium annually — more than total world mine output — with approximately 1.5 billion tonnes already stacked in the United States alone.

ARBOK-ZWD takes each liquor whole, returns clean water at up to 100% of intake, removes the entire dissolved load as dry separated salt fractions, separates the acid from the metal value in the same pass, and separates the rare earths into individual oxides, including the Nd/Pr split, in that same pass — the step that constitutes the actual monopoly, and the step that here needs no cascade of hundreds of stages.

Applications

Primary feedstock streams:

• Magnet leachates — end-of-life NdFeB magnets are demagnetised and dissolved, not crushed and smelted; the resulting liquor holds 28–32% rare earths by weight of the original magnet, roughly 225 kg of neodymium and 55 kg of praseodymium per tonne of scrap. Recycling of end-of-life magnets worldwide remains under 1%.
• Magnet-plant wash waters and grinding effluent — machining a hard, brittle, pyrophoric material generates fine swarf; up to 10% of magnet mass is lost this way into cutting-fluid circuits, pickling baths and plating rinse waters, conventionally treated as hazardous waste.
• Phosphogypsum — the calcium sulphate residue of wet-process phosphoric acid manufacture; 70–85% of the rare-earth content of the original phosphate rock reports to the phosphogypsum, at grades of 0.03–0.4% REO, with praseodymium accounting for roughly 4–5% of the rare-earth distribution in phosphate-derived material.

Outputs / uses: praseodymium for the Nd₂Fe₁₄B lattice of NdFeB magnets, where it substitutes for neodymium at 5–6% of magnet mass and improves retention of magnetic properties under heat; praseodymium-doped glass for welders' and glassblowers' goggles (selective absorption of sodium-flare wavelengths); praseodymium–magnesium alloys for aerospace structures needing high-temperature creep resistance; praseodymium-doped fluoride fibre for optical amplification; yttrium-praseodymium and praseodymium-doped ceramics in pigments and solid-oxide fuel-cell cathodes; individual Nd/Pr oxides from the same pass, after final polishing; clean water and dry separated salt fractions as co-products.

Industries and users: NdFeB magnet manufacturers and recyclers, EV traction-motor and wind-turbine supply chains, drone and defence-actuation programmes, optics and aerospace-alloy makers, phosphate and fertiliser producers holding phosphogypsum stacks.

Scale: standard containerised module at a nominal 200 m³/day, sited on concrete or asphalt in the open air.

Operating Principle

Conventional aqueous recovery works inside the water: solvent extraction, ion exchange, membrane processes and selective precipitation drive the stream through a multi-stage train of reagents, resins and membranes, and cost scales with stage count and consumable replacement. For the Nd/Pr pair this fails outright — the separation factor between the two neighbouring lanthanides is one of the smallest in the series, so the theoretical stage count to reach magnet-grade purity is very large, and no classical route separates the acid from the metal and the pair from each other economically at small scale.

ARBOK-ZWD inverts the sequence. The whole stream is processed at once under deep vacuum at ambient temperature. Water separates from the dissolved load as a phase-change process rather than as thermal evaporation, which is the origin of the low specific energy. Clean water is returned at up to 100% of intake. The entire dissolved load leaves as dry separated salt fractions, split by density. In the magnet-leachate case the acid and its dissolved metal load are removed from the water rather than titrated against a neutralising reagent, so the separation of acid from praseodymium and neodymium happens as part of the same pass, and the output is a solid concentrate rather than a loaded organic phase or resin — a refining problem, not a stripping problem.

Recovery of the mixed concentrate is necessary but not sufficient. The monopoly is built on the capacity to resolve that concentrate into individual oxides at magnet-grade purity, and within that capacity the Nd/Pr split is the specific stage that has never been replicated at scale outside one jurisdiction. In the ARBOK route that stage disappears as a separate facility: separation into individual oxides is performed in the pass itself, and only final polishing follows. The product therefore does not have to leave the country to become a product.

Key Parameters

| Parameter | Value |

|---|---|

| Vacuum | deep vacuum |

| Process temperature | ambient; no thermal input beyond the vacuum duty |

| Water return | up to 100% of intake by volume; zero discharge; no brine, no reinjection |

| Solids output | dry separated fractions |

| Consumables | none — no membranes, filters, reagents, coagulants or ion-exchange media |

| Form factor | standard containerised module, nominal 200 m³/day, sited outdoors on concrete or asphalt |

| Service life | 10–20 years; deep-vacuum operation removes the corrosion mechanism that limits conventional plant life |

| Radionuclides (phosphogypsum service) | radium-226 and associated activity report to an immobilised salt fraction, a small share of processed volume |

| Magnet leachate, REE content | 28–32% by weight of the original magnet |

| Magnet leachate, Nd/Pr yield | ≈225 kg Nd + ≈55 kg Pr per tonne of scrap |

| Wash water / grinding effluent | up to 10% of magnet mass |

| Phosphogypsum grade | 0.03–0.4% REO; Pr ≈4–5% of the rare-earth distribution |

| World phosphogypsum flow | ≈250 Mt/year |

| Contained praseodymium (world flow) | ≈30,000 t/year — above total world Pr mine output |

| US phosphogypsum stock | ≈1.5 billion t |

| Praseodymium metal price | $245.40/kg (2026), +112.65% year on year, +70.30% year to date |

| NdPr oxide benchmark | $133.02/kg, 1 July 2026, +21.4% in one month, yearly high |

| China share | ≈85% of Nd mining, ≈90% of world NdFeB magnet output |

| Recovery efficiency, product purity, module count per site | [требует уточнения из базы] |

Architecture and Components

A standard containerised deep-vacuum separation module operates on the whole liquid stream — magnet leachate, wash water/grinding effluent, or phosphogypsum liquor — producing clean water and dry separated fractions, with the rare earths resolved into individual oxides, including the Nd/Pr split, in the same pass, and only final polishing after it. In phosphogypsum service, a dedicated immobilisation line takes up the radionuclide-bearing fraction. No membranes, reagents or consumables anywhere in the train. The train is laid out to the stream it serves, so the arrangement of the separation and dissolution stages differs between a magnet-recycling cluster and a phosphogypsum stack.

Advantages

Technical: processes the whole stream rather than one target ion, so it is not dependent on selective sorption chemistry; separates acid from metal in the same pass that removes water, eliminating the stripping step that defeats classical routes; delivers dry fractions rather than a loaded organic phase, resin or sludge.

Economic: all three feedstocks are currently paid for twice — once as a disposal or compliance cost, once as an import of the metal they contain. Water and salt return as saleable co-products alongside the rare-earth concentrate.

Environmental: magnet-plant effluent and grinding sludge stop being hazardous waste; phosphogypsum stacks stop being purely a radioactive liability; recycling capture of end-of-life magnets rises above the current sub-1% rate.

Strategic: the Nd/Pr split is performed in the pass itself and can therefore be sited in any jurisdiction that already holds the feedstock, which directly targets the one step — resolution of the concentrate into individual magnet-grade oxides — where approximately 90% of world capacity sits in a single country.

Integrations

Platform basis: ARBOK-VC (Vacuum Cracking) and ARBOK Critical-Materials Recovery. Shares its Nd/Pr pair and feedstock logic directly with ARBOK-Neodymium. Phosphogypsum feedstock connects this case to ARBOK-PHOSPHOGYPSUM and to ARBOK-Lanthanum-Cerium, which draws La and Ce from the same concentrate. The rare-earth concentrate and compact oxide-separation stage relate to ARBOK-Scandium-REE and to ARBOK-Dysprosium, which shares the same three feedstock streams (magnet leachate, magnet-plant effluent, phosphogypsum) for the heavy-rare-earth fraction. Dry separated salt fractions relate to ARBOK-CRYSTALLIZER.

Deployment & Operation

A containerised module is installed on an existing liquid stream — magnet-recycling leachate, magnet-plant wash water, or phosphogypsum stack liquor — with final polishing where individual oxides are required on site. The process is additive to existing operations: dissolution of scrap magnets and phosphoric-acid manufacture already occur, and the module intercepts the resulting liquor rather than replacing an upstream step.

Target geographies stated in the source:

• European Union — magnets are collected under end-of-life vehicle and WEEE obligations, so scrap is already gathered while separation capacity does not exist; the Critical Raw Materials Act sets recycling and domestic-processing targets, and the Water Framework Directive drives the discharge side.
• United States — magnet manufacturing capacity is under construction and the bottleneck for those plants is oxide, not assembly; Florida and five further states hold approximately 1.5 billion tonnes of phosphogypsum.
• Japan and Korea — mature collection systems, large magnet-consuming industries, long-standing exposure to a single supplier of separated oxide.
• Morocco, Turkey, Jordan, Mexico — the largest phosphate complexes on the planet, in several cases co-located with utilities and workforce already in place.
• Any jurisdiction with its own motor industry — the requirement is a separation train and feedstock already in the country, not a new mine.

TRL

Not stated as a number in either source. The magnet-leachate, magnet-plant-effluent and phosphogypsum separation described here is presented as an application of the ARBOK-ZWD platform, which the base records at TRL 8–9 on comparable streams elsewhere (see ARBOK-Dysprosium, ARBOK-Scandium-REE, ARBOK-Yttrium). The Nd/Pr split is performed in the same pass; purity per pass is commercial information and is not recorded here.

Market Potential

NdPr oxide and praseodymium metal moved sharply through 2026: NdPr oxide from a trough near $90/kg on 1 June to $133.02/kg on 1 July, a 21.4% monthly rise and a yearly high; praseodymium metal at $245.40/kg, up 112.65% year on year and 70.30% year to date. These are read in the source as the signature of an administered market rather than a geological shortage: on 22 June 2026 ten United States entities, including MP Materials and USA Rare Earth, were added to Chinese export-control lists, and MOFCOM Announcement No. 26 of 2026, effective 1 July, created a formal reporting and reward mechanism for suspected export-control violations.

Western separation capacity is beginning to appear but remains small relative to demand: Lynas produced its first terbium oxide in Malaysia in June 2026, following dysprosium earlier in the quarter, on a circuit rated at roughly 1,500 t/year. Approximately 90% of neodymium and praseodymium consumption is absorbed by permanent magnets — one application class embedded in the electrification of transport, wind generation, robotics and defence actuation, not a diversified demand base that can be re-routed.

Typical Project Economics

Per tonne of magnet scrap entering the leach, at prices taken 20% under prevailing 2026 levels (on the assumption that a genuine alternative supply softens the market):

• Praseodymium at 5–6% of magnet mass: approximately 55 kg.
• Neodymium at 22–23%: approximately 225 kg.
• The Nd/Pr pair together: approximately 280 kg at $106/kg, or approximately $30,000.
• Dysprosium and terbium at 1–4%: approximately $7,000.
• Total: approximately $37,000 per tonne of scrap now remelted as iron. At 1,000 t of scrap, $37 million.

The phosphogypsum route is a different arithmetic: grade is low and volume is enormous, and the operator carries a perpetual custodial obligation on the stack. The source frames the value there as the sum of three things arriving together — the contained praseodymium and its companion rare earths, the water returned to circulation, and the avoided cost of managing a radioactive liability that currently produces no revenue at all — without stating a per-tonne dollar figure for this route specifically.

Risk Factors

Grade in the phosphogypsum route. Rare-earth concentrations there are low in absolute terms. The source treats this as not decisive on its own, because the relevant comparison is cost per kilogram of contained oxide delivered to the separation train, not percentage in the feed: in phosphogypsum the mining, grinding and acid attack have already been paid for by the fertiliser business, while in magnet leachate the concentration already runs two orders of magnitude above any ore.

Praseodymium's fraction is a distribution statement, not a per-site assay. Site-specific characterisation is a precondition for any commercial commitment. The source reports higher confidence in magnet compositions, which are well constrained by manufacture, than in phosphogypsum distributions, which vary by rock source.

Market absorption. NdPr is large by rare-earth standards but thin by commodity standards, and volume released without regard to price would move it. The source frames the objective as holding supply not subject to another state's licensing regime rather than flooding the market — a national requirement of hundreds to low thousands of tonnes per year is a small fraction of the resource identified here.

Recovery efficiency, product purity and site-level module economics. [требует уточнения из базы]

Related Technologies

ARBOK-Neodymium · ARBOK-VC (Vacuum Cracking) · ARBOK Critical-Materials Recovery · ARBOK-PHOSPHOGYPSUM · ARBOK-Lanthanum-Cerium · ARBOK-Scandium-REE · ARBOK-Dysprosium · ARBOK-CRYSTALLIZER