Lithium — One Metal With Two Completely Unrelated Supply Chains
A brine operation and a spodumene mine produce the same lithium carbonate and share almost no equipment, no chemistry and no timescale. One takes eighteen months. The other takes hours.
Ore to Vehicle · Part 2 — The Metals · Chapter 6 · 19 min read
A brine operation and a spodumene mine produce the same lithium carbonate and share almost no equipment, no chemistry and no timescale. One takes eighteen months and evaporates a lake. The other takes hours and burns a lot of gas.
12–18 mo
Brine evaporation
1,050 °C
The spodumene phase change
~7.5–8 t
SC6 concentrate per tonne LCE
0.6–0.8 kg
LCE per kWh of cell
6.1 — Route A — continental brine
Beneath the salars of Chile, Argentina and Bolivia, groundwater has spent millions of years dissolving lithium out of volcanic rock and concentrating it by evaporation. The Salar de Atacama runs roughly 1,800–2,000 mg/L lithium — about 0.18 per cent.
- •1 · Pumping and solar evaporation, 12–18 months. Brine pumped into successive ponds; water evaporates; salts crystallise out in sequence — halite, then sylvite, then carnallite. Lithium rises from 0.18 % to about 6 %. Free of energy cost and enormously expensive in time and water rights.
- •2 · Impurity removal. Mg²⁺ is the enemy — chemically similar to Li⁺ and it ruins the product. Mg²⁺ + Ca(OH)₂ → Mg(OH)₂↓ + Ca²⁺, then Ca²⁺ + Na₂CO₃ → CaCO₃↓. The Mg:Li ratio of a brine is the single best predictor of whether it is economic.
- •3 · Precipitation to product. 2 LiCl + Na₂CO₃ → Li₂CO₃↓ + 2 NaCl. Lithium carbonate is deliberately chosen because it is one of the few lithium salts that is less soluble hot than cold, so precipitation is driven by heating.
6.2 — Direct lithium extraction
Technical framing
DLE replaces the ponds with a selective sorbent or membrane that pulls Li⁺ straight out of raw brine. Weeks instead of months, far higher recovery (70–90 per cent against 40–50), a much smaller footprint, and reinjection of spent brine.
It is the most consequential process change in the industry and is scaling now, particularly in Argentina.
6.3 — Route B — hard rock spodumene
Spodumene, LiAlSi₂O₆, is a pyroxene found in granitic pegmatites — Western Australia above all, plus Brazil, Zimbabwe, Canada and China.
- •1 · Mine and concentrate. 1–2 % Li₂O ore → dense-media separation and flotation → SC6, 6 % Li₂O concentrate.
- •2 · Decrepitation at 1,050–1,100 °C. α-spodumene → β-spodumene: a crystal structure change, not a chemical reaction. The dense monoclinic α phase converts to an open tetragonal β phase with ~30 % volume expansion. α-spodumene is essentially inert to acid; β-spodumene is not.
- •3 · Sulfation roast at 250 °C. β-LiAlSi₂O₆ + H₂SO₄ → Li₂SO₄ + residue, with lithium ion-exchanged out of the lattice by hydrogen.
- •4 · To product. Li₂SO₄ + Na₂CO₃ → Li₂CO₃↓, or Li₂SO₄ + Ca(OH)₂ → 2 LiOH + CaSO₄↓.
Important
Everything downstream depends on that one phase transition. Without the 1,050 °C step there is no acid route at all, which is why spodumene conversion is energy-intensive in a way brine is not.
6.4 — Carbonate or hydroxide is a cathode decision
Why this matters
LFP is made from lithium carbonate. High-nickel NMC and NCA need lithium hydroxide, because LiOH melts and reacts at a lower temperature — and nickel-rich cathodes degrade if fired hot enough for carbonate, leaving residual Li₂CO₃ on the particle surface that gasses in the cell.
The market is really two markets, and they can move in opposite directions.
6.4.1 — The arithmetic
Roughly 7.5 to 8 tonnes of SC6 concentrate makes one tonne of lithium carbonate equivalent. One tonne of LCE contains 188 kg of lithium metal. A kilowatt-hour of cell needs about 0.6 to 0.8 kg of LCE — so a 60 kWh pack carries something like 40 kg of LCE, which came from roughly 300 kg of concentrate and several tonnes of rock.
6.5 — Why lithium overshoots in both directions
In plain English
Time is the axis nobody plots and everybody feels. A spodumene converter responds to a price signal in months; a pond responds in a year and a half. That asymmetry is a large part of why lithium prices overshoot both ways.
Commercially, Australia mines around half the world’s lithium as spodumene, and Chile and Argentina supply most of the brine. But China converts roughly two-thirds of it — including most Australian concentrate, which is shipped raw to Chinese converters. Australia has been building domestic conversion with mixed success, which is itself a lesson in how hard the midstream is.
Pricing has no proper futures market. Lithium is sold on contracts referenced to assessed prices from Fastmarkets, Benchmark Mineral Intelligence, Argus and Platts, based on surveyed transactions. China’s GFEX launched a lithium carbonate futures contract in 2023 which has grown quickly, but most of the world still transacts against assessments — a structural difference from copper that makes hedging genuinely hard.
Quick check: test yourself
1.Why must spodumene be heated to 1,050 °C before acid will touch it?
Show answer
2.A cathode maker switches from LFP to high-nickel NMC. What changes in its lithium purchasing?
Show answer
3.Why does the Mg:Li ratio decide whether a brine is economic?
Show answer
Chapter summary
- ✓Brine and hard rock make the same product and share almost no equipment, chemistry or timescale.
- ✓Brine depends on 12 to 18 months of solar evaporation and on the Mg:Li ratio; DLE replaces the ponds with sorbents and roughly doubles recovery.
- ✓Hard rock depends entirely on the α to β phase change at 1,050 °C, without which the mineral is inert to acid.
- ✓Carbonate versus hydroxide is set by the cathode, not by preference — LFP takes carbonate, high-nickel takes hydroxide.
- ✓A converter answers a price signal in months and a pond in eighteen, which is a large part of why lithium overshoots in both directions.
Frequently asked questions
Why does spodumene have to be heated to 1,050 °C before anything else can happen?+
Because α-spodumene is essentially inert to acid. Heating converts the dense monoclinic α phase to an open tetragonal β phase with about 30 per cent volume expansion — a crystal structure change, not a chemical reaction — and β-spodumene will give up its lithium to a sulfuric acid roast at 250 °C by ion exchange with hydrogen. Everything downstream depends on that one phase transition.
Is lithium carbonate or hydroxide better?+
Neither — it is a cathode decision, not a preference. LFP is made from lithium carbonate. High-nickel NMC and NCA need lithium hydroxide, because LiOH melts and reacts at a lower temperature and nickel-rich cathodes degrade if fired hot enough for carbonate, leaving residual Li₂CO₃ on the particle surface that gasses in the cell. The market is really two markets, and they can move in opposite directions.
What is direct lithium extraction and why does it matter?+
DLE replaces the evaporation ponds with a selective sorbent or membrane that pulls Li⁺ straight out of raw brine. It takes weeks instead of months, recovers 70 to 90 per cent against a pond’s 40 to 50, has a much smaller footprint, and allows reinjection of spent brine. It is the most consequential process change in the industry and is scaling now, particularly in Argentina.
Why do lithium prices overshoot so badly in both directions?+
Because the two routes respond on completely different timescales. A spodumene converter answers a price signal in months; an evaporation pond answers in a year and a half. Add to that the absence of a proper futures market — most of the world transacts against assessed prices published by Fastmarkets, Benchmark, Argus and Platts, though China’s GFEX contract has grown quickly since 2023 — and hedging is genuinely hard for anyone building a cell plant.
Reviewed by
Ore to Vehicle is an original educational series on the battery and EV materials supply chain. Country shares, grades, prices and policy status are approximate, drawn from public reporting as of mid-2026, and move year to year — treat them as orders of magnitude rather than as a ledger, and verify before relying on them commercially.