Overview
There is no erbium mine. Erbium is a by-product of heavy-group separation, arriving with ytterbium, thulium and lutetium. World output is ~700 t/yr of oxide at $28–38/kg; ~99% of refined supply is Chinese. Erbium is neither rare nor expensive — it is more common in the crust than tin.
The whole case rests on one processing step, not on geology. Adjacent lanthanides are chemically near-identical; the industrial answer since the 1960s is solvent extraction across hundreds of mixer-settler stages. Building such a cascade is not a secret — it is a capital commitment most jurisdictions declined to make while a cheaper supplier existed. Hence a country that mines no erbium can be totally dependent on erbium.
Demand cannot be substituted. ~40% goes into erbium-doped fibre amplifiers. The Er³⁺ 4I13/2 → 4I15/2 transition emits at 1530–1565 nm — precisely the attenuation minimum of silica fibre. Pumped at 980 or 1480 nm, the signal is amplified in place, with no optoelectronic conversion. This is what made DWDM economic and what carries essentially all intercontinental traffic. Raman and SOA fill specific roles but do not replace EDFA in the long-haul trunk; thulium amplifiers address a different band. AI infrastructure pulls demand upward: every new cluster adds optical interconnect.
Dated policy risk. October 2025 (MOFCOM Announcement No. 57) placed erbium under export control together with holmium, thulium and europium. Suspended 7 Nov 2025, suspension running to 10 Nov 2026. A suspension is not a repeal — the planning horizon is explicit.
> Core technology and architecture: see ARBOK-VC (Vacuum Cracking).
Applications
Feedstock streams: fresh phosphogypsum and stacks; acid mine drainage of coal and sulphide basins; monazite processing liquors; end-of-life optical fibre and spent catalysts.
Outputs: erbium fraction; Nd, Pr, Dy, Tb, Lu, Y and the rest of the basket from the same pass; clean water; acid returned to process; radionuclides immobilised into a small residue fraction.
End uses: EDFA amplifiers, Er:YAG lasers at 2940 nm (dentistry, dermatology, ophthalmic surgery), erbium fibre lasers, pink glass and ceramics, burnable neutron absorber in nuclear fuel.
Operating Principle
Phosphogypsum is dissolved in the acid medium the process generates itself — no purchased reagent stream — and separated by cold boiling under deep vacuum. No furnaces, membranes, consumable reagents or CO2 from the recovery step. Separation happens in a single pass, inside the solution itself, without the classical cascade of hundreds of stages.
Key Parameters
Erbium oxide: $28–38/kg. World output ~700 t/yr. China ~99% of refined supply.
Export control: MOFCOM No. 57 of 9 Oct 2025 (Er, Ho, Tm, Eu), suspended 7 Nov 2025 → 10 Nov 2026. (April's No. 18 — Sm, Gd, Tb, Dy, Lu, Sc, Y — was never suspended.)
Phosphogypsum: ~3,500 mg/kg total REE, of which erbium ~5.8 mg/kg.
Global phosphogypsum generation ~200–280 Mt/yr; US ~30 Mt/yr, concentrated in Florida.
Florida annual flow: 30 Mt/yr → ~174 t Er/yr = a quarter of world output. Accumulated stacks → ~5,800 t Er = ~8 years of world production.
Demand split: ~40% EDFA; remainder lasers, glass/ceramics, nuclear.
Essentially no erbium recycling at scale.
Architecture and Components
Arbok-SA route: dissolution + deep-vacuum cold boiling + in-solution separation. Radionuclide immobilisation into a small residue fraction. Shared platform with ARBOK-Lutetium, ARBOK-Terbium, ARBOK-Lanthanum-Cerium, ARBOK-Dysprosium, ARBOK-Yttrium, ARBOK-PHOSPHOGYPSUM.
Advantages
Technical: reaches the heavy end of the row without the cascade that makes that end expensive.
Economic: the erbium line rides on a project funded by the full REE basket; water and acid return to the process.
Strategic: erbium becomes recoverable at sites that could never justify a cascade — one stack, one drainage outfall, one acid plant.
Integrations
Phosphoric-acid plants and their stacks; mine water treatment under reclamation bonds; monazite processing; optical-fibre scrap streams. Connects to ARBOK-PHOSPHOGYPSUM, ARBOK-Lutetium, ARBOK-Scandium-REE, ARBOK-Neodymium, ARBOK-SA.
Deployment & Operation
Assay of stack/rock and of drainage with explicit Er determination → configuration → install on the existing stream. Purchase or BOOM with off-take on the separated fractions.
TRL
Platform TRL 9; Arbok-SA route per the lanthanum/phosphogypsum case. Erbium-specific field reference pending. Refining to EDFA-grade oxide purity is a separate step, not demonstrated or costed in-house.
Market Potential
Not a tonnage market — an availability market. The buyer is an operator or manufacturer whose constraint is a dated licensing regime on optical transport, not a price. Every phosphate site with a stack in a jurisdiction that wants domestic optical-grade supply is a candidate.
Typical Project Economics
Basis: site handling 1 Mt/yr of phosphogypsum.
- Erbium contained: 5.8 t/yr ≈ $190k/yr. Modest as a revenue line; almost a full percent of world output from one stack is the number that matters.
- The same pass carries ~3,000 t/yr of contained REO — Nd, Pr, Dy, Tb, Lu — which is what funds the project.
- Water and acid returned rather than consumed.
Florida scale check: 30 Mt/yr fresh = ~174 t Er/yr (a quarter of world output); stacks = ~5,800 t Er (~8 years of world production).
Risk Factors
Er content varies with phosphate source — insist on explicit Er determination, never infer from TREO. Market is ~700 t/yr: new supply must be contracted, not stockpiled. EDFA-grade purity is far above commercial oxide grade and its cost is unknown to us. The suspension may be extended, which removes the urgency argument without removing the structural one. Radionuclide-residue acceptance and conservative phosphate-industry adoption as in the other phosphogypsum cases.
Related Technologies
ARBOK-Lutetium · ARBOK-Terbium · ARBOK-Dysprosium · ARBOK-Yttrium · ARBOK-Lanthanum-Cerium · ARBOK-PHOSPHOGYPSUM · ARBOK-SA · ARBOK-VC (Vacuum Cracking)