Ruthenium Cannot Be Mined Into Existence: Recovery from PGM Refinery, Semiconductor and Catalyst Effluent by Deep-Vacuum Phase Separation

Abstract

Ruthenium is the platinum-group metal that the semiconductor industry has quietly made indispensable, and it is the one metal whose supply cannot respond to price. World production is approximately 30 tonnes per year, obtained exclusively as a by-product of platinum and nickel refining, with roughly 90% originating in South Africa, where platinum-group metal output is in decline. Because the mining decision is taken on platinum, ruthenium supply is close to perfectly inelastic: a demand shock passes entirely into price. That shock has arrived. The price reached USD 1,750 per ounce in March 2026 against USD 560 a year earlier, and stood at USD 62.59 per gram — USD 62,590/kg — in 2026, a rise of 609.73% since 2020. The forecast 2026 deficit of 203,000 ounces is 6.3 tonnes, or roughly one fifth of world output. The entire world ruthenium market, at 30 tonnes and current prices, is under USD 2 billion a year — less than the cost of a single mid-sized data centre that cannot be built without it. Demand is driven by two applications that admit no substitute at scale. In advanced-node interconnects, ruthenium displaces copper and cobalt because its short electron mean free path leaves resistivity far less degraded at nanometre cross-sections, and because it requires no diffusion barrier, allowing the trench to be filled entirely. In hard disk media, a sub-nanometre ruthenium spacer sets the coupling between magnetic layers and underpins the areal density on which high-capacity data-centre drives depend. Ruthenium oxide additionally serves as the active component of dimensionally stable anodes in chlor-alkali production, as a catalyst in ammonia and acetic acid synthesis, and as the resistive element of thick-film resistors manufactured in the tens of billions. This paper argues that the only expandable source of ruthenium is the fraction currently discharged. Three aqueous streams carry dissolved ruthenium today and are managed as hazardous waste: platinum-group metal refinery effluent, including electrolysis bleed and wash waters; semiconductor fabrication effluent, comprising rinses, chemical-mechanical polishing slurries and spent etch and plating baths; and catalyst plant effluent from spent solutions and regeneration rinses. In all three the dissolution step that dominates the cost and environmental burden of primary production has already been performed and paid for by another process. Classical recovery fails on these streams not for chemical reasons but for economic ones: sorption, solvent extraction and precipitation are configured for concentrate, and at single-digit milligrams per litre the cost of reagents and media exceeds the value of the recovered metal even at USD 62,590/kg. We describe the application of ARBOK Zero Waste Discharge (ZWD) — deep-vacuum phase separation at approximately 1 kPa and ambient temperature — to these streams. The process takes the stream whole, returns clean water at up to 100% of intake by volume, and removes the entire dissolved load as dry separated fractions, delivering a concentrate to a compact refining stage that yields ruthenium together with its companion platinum-group metals. Specific energy is under 1 kWh/m³, with no membranes, no reagents and no consumables. Because the process removes the water rather than targeting a solute, its cost is independent of feed concentration — the property that makes dilute streams viable. We set out the process rationale, the throughput economics of a single 20-ft containerised module at 73,000 m³/year, the siting logic, and the strategic consequence for jurisdictions building semiconductor capacity without access to a platinum-group metal mine.