Lithium supply is produced by three primary industrial routes — spodumene hard-rock mining, salar brine evaporation, and adsorption-based Direct Lithium Extraction (DLE) — yielding battery-grade lithium carbonate and lithium hydroxide at recoveries, water intensities and cost-per-tonne-LCE that vary by an order of magnitude between the methods [S4][S5].
For process engineers, the lithium question is no longer "is the chemistry feasible" but "which route matches the resource, the water budget, and the downstream cathode line." The feedstock decision cascades into every downstream spec, from sulfate-rich brines that favour LiOH production for NMC811 cathodes, to low-magnesium brines that Li2CO3 lines can consume without extensive purification [S4][S5].
Spodumene Hard-Rock Concentrate to Li2CO3: α→γ Decrepitation and Sulfation Roast
The dominant hard-rock feedstock is spodumene (LiAlSi2O6, theoretical 8.03% Li2O), and the unit operation that defines the route is decrepitation — heating α-spodumene past the 1050 °C polymorphic transition to the acid-leachable γ-phase, with a 30% volume change that physically crumbles the ore [S4][S5].
After γ-phase conversion, the standard flowsheet is a 250 °C sulfation roast with concentrated H2SO4, water leach, impurity removal (Fe, Al, Mg via staged pH adjustment), soda-ash (Na2CO3) precipitation of technical-grade Li2CO3, and a bicarbonation or CO2-reslurry polish to reach battery-grade 99.5% Li2CO3. CLPC's Sichuan operation, established March 1997 and supplied by a 6 million-tonne lithium-mine reserve, runs this route at a nameplate of 39 000 t/yr lithium carbonate plus lithium hydroxide capacity [S4].
Solar Evaporation Brine Routes: Salar Ponds, Mg/Li Ratio and 18-Month Cycle Time
Brine routes — solar evaporation in staged ponds followed by selective precipitation — are the second pillar of global supply and remain the lowest unit-cost option where land and climate cooperate. The spec gate for a brine project is the Mg:Li ratio; brines above roughly Mg:Li 6:1 require an extra boron/magnesium-removal stage that drives capex up and recovery down [S4][S5].
Typical industrial cycles concentrate 200–400 ppm Li brines up to 6000 ppm over 12–18 months of staged pond evaporation, precipitate Mg, Ca and B impurities along the path, and finish with a carbonate precipitation step to yield technical-grade Li2CO3 (typically 99.0–99.5%) that is then bicarbonated and ion-exchanged to battery-grade 99.5%+ [S5]. Water intensity is the route's defining weakness: conventional pond evaporation consumes on the order of 2 million litres per tonne LCE in arid basins, which is one of the main technical drivers for DLE adoption in water-stressed jurisdictions [S5].
Direct Lithium Extraction (DLE): Adsorbent Columns, 95% Recovery, Hours Instead of Months
Direct Lithium Extraction covers a family of processes — lithium-selective alumina/aluminosilicate or manganese-oxide sorbents, solvent extraction with tributyl phosphate/kerosene, and membrane/ion-exchange hybrids — that strip lithium from brine in contactors rather than open ponds, with reported single-pass recoveries above 95% and cycle times measured in hours [S5].
Summit Lithium Technologies' "denaLi" DLE stack claims a 1 000–2 000 USD per tonne LCE cost reduction, 8× lower water intensity than pond evaporation, and 95%+ validated uptime, and as of June 2026 the company reported completion of basic engineering for its first commercial DLE plant after its 2025 rebrand from Summit Nanotech [S5]. For process engineers, the spec frontier is now the sorbent cycle life (typically thousands of Li-strip/Li-load cycles before acid wash recovery), the brine pre-treatment tolerance (Fe, Mn, organics foul most alumina sorbents within weeks), and the post-DLE polishing train needed to reach battery-grade 99.5% Li2CO3 or LiOH·H2O [S4][S5].
Battery-Grade LiOH·H2O: Causticisation, Sulfate Removal and NMC811 Pull
Nickel-rich NMC811 and single-crystal NMC cathodes require LiOH·H2O rather than Li2CO3 as the lithiation agent, because residual carbonate on the cathode surface degrades the cathode's tap density and cycle life above 4.2 V operation [S2]. LiOH is normally produced by causticising technical-grade Li2CO3 with Ca(OH)2 (the "lime route") or, more directly, by electrolysis of Li2SO4 to LiOH in membrane cells [S4][S5].
Goldencell specifies LiFePO4 (LFP) cell-grade LiFePO4 cathode material with independent-patented high-temperature-stable structure, wide operating-temperature performance, and long cycle life — the chemistry where Li2CO3 is acceptable and the cost pressure on the upstream lithium route is the most aggressive in the market [S2]. For plant engineers evaluating LiOH lines, the critical specs are SO4²⁻ < 100 ppm, Na < 50 ppm and Fe < 5 ppm in the finished LiOH·H2O; exceeding any of these biases the cathode powder toward agglomeration and off-stoichiometry [S2][S4].
Route Comparison: Cost, Recovery, Water, Time-to-Market
For a 20 000 t/yr LCE plant, the three routes line up against four decision criteria: capital intensity, lithium recovery, water consumption, and time-to-first-LCE. Spodumene hard-rock sits at moderate capex (the flow meters and pressure transmitters on the sulfation train are the dominant instrumentation cost), 70–80% recovery, water-light at roughly 50 000 L/t LCE, and 24–30 months to first product. Brine evaporation is capex-light but land-intensive, recovers 40–60% of the lithium in the original brine, consumes on the order of 2 000 000 L/t LCE, and needs 18–24 months before first Li2CO3. DLE is capex-heavy, recovers more than 95%, uses roughly 250 000 L/t LCE, and can reach first-LCE inside 12 months once the sorbent train is qualified [S4][S5].
The lithium-material flow on the cathode side also locks in the route: LFP and LMO cathodes accept Li2CO3 directly and are indifferent to the lithium source, while high-nickel NMC811 lines need low-sulfate LiOH·H2O and are the natural pull on a sulfuric-acid-flavoured hard-rock or DLE flowsheet [S2]. For a new entrant in 2026, the decision tree is therefore: resource dictates the route, route dictates the product (Li2CO3 or LiOH·H2O), and product dictates the cathode customer.
Recycling and Mechanochemical Recovery: Closing the Loop on Cell Scrap
Primary lithium production is being complemented by cell-scrap recycling, and mechanochemistry is one of the more widely cited research-stage approaches: ball-milling of cathode black mass with reagents extracts lithium selectively and avoids the high-temperature pyrometallurgical step [S1]. A 2023 study in Communications Chemistry reported a universal, ecologically benign mechanochemical route for extracting lithium from mixed Li-ion battery waste, with selective lithium recovery that leaves transition metals in the solid phase for downstream hydrometallurgical refinement [S1].
For a 2026 plant engineer evaluating end-of-life streams, the practical spec gate is the scrap-feed impurity mix: LFP scrap contains Fe and P that require different leaching chemistry than NMC scrap, and a single recycling line that handles both chemistries is now the spec demand from cell makers [S1][S2]. Industrial recycling capacity in 2026 remains a fraction of primary supply, but mechanochemical and hydrometallurgical trains are the two most active research-to-commercial pipelines, and they will determine the long-run supply elasticity when scrap-to-cathode ratios rise above 30% by the early 2030s [S1].
Process Control Specs: Where the Li Plant Meets the Instrument Engineer
On a modern Li2CO3 or LiOH plant, the high-wear instrumentation points are the acid handling and brine pre-treatment skids: pressure transmitters on the H2SO4 train, flow meters on the brine feed and on the mother-liquor recycle, industrial valves in alloy-20 or PTFE-lined bodies on the chloride and sulfate streams, and the PLC / SCADA layer that sequences the staged-precipitation and ion-exchange polishers [S4][S5].
On a DLE line, the most instrument-intensive skid is the sorbent regeneration loop, where the acid-wash industrial valves and the pH/conductivity probes define the column cycle time and therefore the plant's lithium throughput; under-speccing this loop is the single most common cause of DLE capacity de-rating in commercial operation [S5]. Adjacent plant systems — the servo motor-driven crystalliser stirrers on the Li2CO3 precipitation train, the reagent dosing skids, and the bagging/dispatch line — are off-the-shelf industrial specs and rarely the bottleneck. As a result, lithium-plant engineering in 2026 is increasingly a process-control procurement problem rather than a chemistry problem, and the spec sheets that win are the ones that match the DLE column cycle, not the sorbent chemistry [S5].
Trackable signals into the second half of 2026: Summit Lithium Technologies' first commercial DLE plant commissioning after basic-engineering completion in June [S5], CLPC's continued 39 000 t/yr Li2CO3/LiOH output from its Sichuan operation [S4], and the next data point from mechanochemical recycling pilots scaling from gram-lab to kilogram-continuous reactors [S1].
For related coverage, see V-Ribbed Belt vs Timing Pulley: Spec Cut for Power Transmission Buyers.