Lithium Carbonate Crystallization: Battery-Grade Route

Last updated: 2026-10-29

Lithium carbonate crystallization is the step that converts a purified lithium liquor into battery-grade Li₂CO₃ solid, and in most plants it is engineered around a carbonation–decarbonation loop rather than simple evaporation — the loop routinely delivers ≥99.5% Li₂CO₃ at 85–93% section lithium recovery. This article walks the route stage by stage: where the crystallization section sits in a brine-to-lithium-carbonate flowsheet, why the LiHCO₃ loop exists, what product specifications demand, and which lithium plant equipment belongs on the enquiry list. It sits inside our evaporation and crystallization resources cluster.

Key Takeaways

• Battery-grade Li₂CO₃ purity is set by the carbonation–decarbonation loop, not by a single evaporator.
• The route runs: brine concentration, boron and sulfate removal, LiCl intermediate, carbonation, thermal de-carbonation, separation.
• Crystal size (d50 typically 8–15 µm), bulk density and flowability are offtake specifications, not cosmetic properties.
• Feed chemistry — brine, spodumene or battery recycle — changes the front end; the crystallization section changes with it.
• Comparable quotes require a written design basis: feed tenor, product spec, crystal targets, utility limits.

Where Does Lithium Carbonate Crystallization Sit in the Flowsheet?

A train that turns brine or leach liquor into battery-grade lithium carbonate ends, always, in a crystallizer. The upstream work exists to hand the crystallization section a liquor it can convert to specification. The sequence below is the standard spine for a brine-based plant; spodumene and recycle plants swap the front end but keep the tail.

1. Brine evaporation and concentration. Salar or geothermal brine is concentrated in multi-effect or mechanical vapor recompression (MVR) evaporator trains, raising lithium tenor while rejecting sodium, potassium and much of the magnesium as salts. Our brine concentrator sizing and selection guide covers this stage; the lithium brine application page covers the upstream flowsheet.

2. Boron and sulfate removal. Boron is the classic brine poison: it follows lithium through most unit operations and is capped at very low levels in battery product. Removal is typically solvent extraction or ion exchange after concentration, with sulfate polished in dedicated steps — the loop cannot strip boron.

3. Lithium chloride or sulfate intermediate. Purified lithium is held as concentrated LiCl (brine route) or Li₂SO₄ solution (spodumene route). This is the last point where bulk impurities can be rejected cheaply.

4. Carbonation with CO₂. The purified liquor is contacted with carbon dioxide in a carbonation column, converting carbonate into soluble lithium bicarbonate: Li₂CO₃ + CO₂ + H₂O → 2 LiHCO₃. This is the heart of the purification logic explained next.

5. Thermal de-carbonation. The bicarbonate liquor is heated in agitated reactors, precipitating Li₂CO₃ while releasing CO₂: 2 LiHCO₃ → Li₂CO₃ + CO₂ + H₂O. The released CO₂ is recovered, compressed and recycled to carbonation, which is why CO₂ consumption sits far below stoichiometric demand.

6. Thickening and centrifugation. The slurry is thickened, then dewatered on centrifuges with intermediate washing. Mother liquor carries sodium, potassium, sulfate and boron, so wash efficiency translates directly into product purity; bleed strategy is treated in our mother liquor recovery resource.

7. Drying, milling and classification. The cake is dried at controlled, low temperature, then classified to hit the particle size and flowability targets of the offtake contract. Bagging runs under dry air because the product is hygroscopic.

Why the Carbonation and De-Carbonation Loop Sets Purity

Why not simply evaporate the purified liquor and crystallise Li₂CO₃ directly, the way nickel sulfate is crystallised? The answer is solubility behavior. Lithium carbonate is one of the rare technical salts whose solubility decreases with temperature, and it is only moderately soluble to begin with — roughly 13 g/L at 20 °C, falling toward 7 g/L near 100 °C (typical literature values, to be confirmed by testwork). Evaporating a concentrated LiCl liquor would co-crystallise sodium, potassium and sulfate alongside the product.

Solid-Liquid Separation

Lithium carbonate crystallization, properly engineered, is a precipitation process that trades heat for selectivity. Carbonating the liquor dissolves lithium as bicarbonate while most impurities — sodium, potassium, calcium, magnesium, sulfate, boron — stay in solution or in a separate filtered-off solid phase. De-carbonation then precipitates only lithium carbonate, rejecting impurities back into the mother liquor. Each pass through the loop is an impurity rejection step, and purity is managed by controlling mother-liquor carryover and wash efficiency. The LiHCO₃ loop is a purification circuit that converts the product to a soluble intermediate and back, and it is why a plant can hold sodium below 0.025% in the final solid even when the feed carries far more.

Three engineering consequences follow:

• CO₂ recovery matters economically. The loop emits CO₂ at the de-carbonation reactor and consumes it at the carbonation column; a recovery compressor returning that gas is standard, and without it CO₂ consumption roughly doubles.
• pH and temperature are the purity levers. Under-carbonation strands product in the loop bottoms; over-carbonation drags impurities into solution that re-precipitate with the product. De-carbonation temperature sets nucleation rate, and nucleation rate sets crystal size.
• Boron discipline is won or lost here. No stage downstream of the ion-exchange boron removal step rejects boron, so any upstream excursion appears in the final solid — hence boron analysers on the loop feed, not just on the product.

For the thermal side — how de-carbonation heat demand interacts with evaporator economics — our evaporator energy consumption resource carries the specific-power framework that applies here.

What Do Battery-Grade Specifications Actually Demand?

There is no single universal standard for battery grade lithium carbonate. Every cathode and electrolyte producer publishes its own impurity schedule, versions tighten over time, and the offtake contract always overrides anything written here. The table below is the vocabulary of the negotiation — typical industry specification values to be confirmed against your own offtake contract before they harden into a design basis. In any enquiry, the schedule a bidder designs the lithium carbonate crystallization section against should be the contract schedule, verbatim.

* Typical industry specification values, to be confirmed against the customer’s offtake contract; not a standard.

Read as a flowsheet brief, three lines dictate equipment. Potassium and sodium are managed in the wash and centrifuge step plus mother-liquor bleed discipline — they are soluble, so they leave with the liquor, but every gram of entrained mother liquor in the cake carries them forward. Boron is managed entirely upstream of the loop. Sulfate is managed at the intermediate stage and by wash design, which is why spodumene-route plants typically carry one more polishing step than brine-route plants.

The commercial consequence: a vendor who quotes only on Li₂CO₃ assay is quoting half a plant. Ask every bidder to show how their flowsheet holds each line of the schedule simultaneously — the binding constraint is usually potassium or boron, not the headline assay.

How Do Crystal Size, Bulk Density and Flowability Shape the Design?

Cathode manufacturers do not buy a chemical assay; they buy a powder that must disperse, flow and react reproducibly. That is why the physical schedule — crystal size distribution, bulk density, tapped density, moisture — drives as much crystallizer engineering as the impurity table. The crystal size distribution lithium targets below are typical industry ranges, to be confirmed by testwork.

* Typical values, to be confirmed by testwork and against the offtake specification.

Crystal size distribution lithium targets are achieved in the de-carbonation step, not repaired downstream. Precipitation from hot bicarbonate liquor is fast, and uncontrolled fast precipitation makes fine, agglomerated, low-density powder. The engineering responses, in the order we usually apply them: seed the liquor with a controlled charge of product fines; hold de-carbonation temperature and CO₂ evolution rate steady rather than batch-swinging them; and give the slurry an ageing hold, which ripens fine crystals at the expense of coarse ones. A reactor designed without an ageing volume will hit assay but miss d50 — this is where lithium carbonate crystallization is won or lost as an engineering discipline.

The milling decision is a fork worth naming. Some offtakers want milled product for tight top-size control; milling, however, fractures agglomerates, can cut bulk density below the contract floor, and adds a contamination source. Plants that nail precipitation often skip the mill and classify only. Whether to run continuous or batch de-carbonation is the related structural choice — our continuous vs batch crystallizer comparison maps the trade-offs.

Bulk density is where drying and handling meet crystal quality. A low-density, high-fines powder floods dust collection, bridges in silos and quantises badly into big bags; a properly ripened crystal at 0.9–1.3 g/cm³ packages and flows like the industrial chemical it is. If a bidder’s pilot data stops at assay and moisture, ask for the size distribution and tapped density.

Feedstock Variants: Brine, Spodumene, Recycled Battery Leach and Geothermal

The crystallization tail is similar for every feed; the front of the loop is not. The table summarises what changes in and around the lithium carbonate crystallization section by feedstock — all entries are design-typical and subject to testwork confirmation.

Two patterns matter when comparing bids. First, the loop itself is feed-agnostic — carbonation, de-carbonation, separation and drying hardware barely changes — but the materials-of-construction and wash-water story changes completely: a chloride-bearing plant buys alloys a sulfate plant does not. Second, lithium recovery differs by feed mainly through loop yield and wash losses: plants with heavy impurity loads bleed more liquor, and the lithium in that bleed is recovered or lost depending on whether a dedicated recovery step exists. Our mother liquor recovery treatment covers the bleed economics, and the battery recycling application page covers the black-mass front end that increasingly feeds small lithium loops.

Two decision sentences follow from the table. For a salar brine carrying boron and magnesium, the carbonation–de-carbonation loop is preferred because it rejects both to the mother liquor in a single pass instead of requiring a dedicated downstream polish. For a spodumene leach liquor already loaded with sulfate, sulfate polishing ahead of carbonation is preferred because the loop cannot remove sulfate and will carry it straight into the product specification.

Which Equipment Makes Up the Lithium Carbonate Crystallization Section?

Lithium carbonate crystallization equipment is a train of pressure vessels, columns and separation machines spanning roughly the second half of the flowsheet, and it is the package a buyer will typically tender as one lot. The list below is what a complete inquiry should cover.

• Carbonation column or reactor. An agitated, often packed column contacting the purified lithium liquor with CO₂, sized on gas-liquid mass transfer, with pH and pressure control. A draft-tube baffle crystallizer is a vessel that suspends growing crystals in a controlled circulation loop while separating fines — where a plant runs continuous de-carbonation, a DTB or similar controlled precipitator can carry that duty directly.
• De-carbonation reactors. Agitated, steam-heated vessels where bicarbonate breaks down and Li₂CO₃ precipitates; residence time, temperature profile and seeding hardware are the specification points.
• Thickener. Settles the slurry and raises solids concentration ahead of filtration, with clarified overflow returned toward the loop.
• Centrifuge. Peeler or pusher type with cake-wash capability; the most purity-critical machine in the section, because wash efficiency is Na/K/B policy made mechanical.
• Dryer. Low-temperature, indirect or controlled-convection duty; moisture ≤0.2% without degrading crystal habit or redissolving the mildly soluble product.
• Mill and classifier (optional). Present only where the offtake demands milled product; otherwise a classification loop alone.
• CO₂ recovery system. Compressor and scrubbing returning de-carbonation off-gas to the carbonation column, plus vent treatment where the loop bleeds.
• Mother-liquor handling and bleed. Tanks, filters and a bleed-recovery path; decides where loop lithium losses go.

Vendors differ most in the loop internals — column packing, reactor agitation, seeding hardware — while separation and drying machines are comparatively commodity. The proprietary risk sits in the loop, and the performance guarantee should attach there. Equipment detail on the thermal machines is in our evaporator and crystallizer resource, and mechanical design in the crystallizer design guide.

What Does the Section Consume — Energy, Steam and CO₂?

Utilities decide operating cost, and they are where an equipment comparison becomes numeric. The figures below are indicative industry ranges, to be confirmed by testwork and the vendor’s heat balance for your liquor. In practice, the consumption profile of the lithium carbonate crystallization section is a design choice as much as a chemistry fact.

ASME VIII Div 1 vs Div 2 in practice

* Indicative values only; confirm against offtake spec and vendor heat balance.

Two observations from plant experience. First, the CO₂ number is a design choice, not a fact of chemistry: stoichiometric demand is fixed, but a recovery compressor turns most of the emitted gas back into reagent, which is why we treat CO₂ recovery as standard scope, not an option. Second, steam-versus-MVR is a tariff question: where steam is cheap and grid power expensive, multi-effect wins; where power is cheap or decarbonization targets bind, MVR shifts load to electricity — our MVR evaporator selection resource walks that evaluation. Both routes hit the same product spec; the difference shows up in the utility contract.

One honest caveat on recovery: 85–93% is lithium recovery across the purification-and-crystallization section for a well-run plant. Route recovery from raw brine is lower once upstream concentration losses are counted, and the number that matters commercially is the section recovery your flowsheet model can defend with a mass balance.

How Is Demand for Battery-Grade Lithium Carbonate Evolving?

Demand for battery grade lithium carbonate is driven by two end markets with different characters. Electric vehicles — particularly in China, where LFP cathode chemistry consumes lithium carbonate directly rather than via the hydroxide route — set the volume trend, and grid-scale storage adds a second, policy-driven demand leg. Authoritative production and trade statistics are published annually in the USGS Mineral Commodity Summaries, the source we recommend for any market-sizing claim. This article deliberately does not quote prices or market volumes: they move too fast to embed in a technical reference and must come from current market sources at the time you read them.

What a project manager can reason about without market data is the qualification structure. Cathode makers do not switch carbonate suppliers quickly: a new producer typically passes a multi-month qualification cycle in which samples from several production runs are tested against the offtake schedule — assay, each impurity, particle size, bulk density, moisture — before volumes are committed. Two consequences follow. First, consistency outranks peak performance: a plant that holds d50 and impurity levels in a narrow band across campaigns is worth more than one that averages well but swings. Second, physical-spec capability is a market-access question: a producer who cannot meet a 0.9–1.3 g/cm³ bulk-density window is locked out of contracts requiring it.

For equipment buyers the practical reading is that the crystallization section is where qualification is won: upstream chemistry determines whether you can make battery grade at all, and the loop plus the separation-drying train determine whether you can make it the same way every day. Peer-reviewed work in Hydrometallurgy continues to refine loop chemistry and crystallization control, and we track it because vendors’ design margins should reflect the current state of the art.

What Should You Specify in an Equipment Inquiry?

A comparable quotation is a written design basis, and the checklist below is what we ask for when we are on the buyer’s side — and what we ask our own clients when we are on the vendor’s.

1. Feed definition. Intermediate type (LiCl or Li₂SO₄), design tenor and range, full impurity schedule including boron, and the annual production basis — dry or wet, stated.
2. Product schedule. The offtake contract’s impurity table verbatim, plus physical properties: d50 and distribution target, bulk and tapped density, moisture, flowability method.
3. Loop philosophy. Continuous or batch de-carbonation, seeding policy, ageing volume expectation, and CO₂ recovery in base scope.
4. Separation and drying spec. Cake moisture basis, wash-water quality at site, dryer heat source and temperature limits, and the milling-or-not decision.
5. Utilities envelope. Steam pressure and cost, power price and reliability, cooling water, and any emissions constraint on CO₂ venting — these choose multi-effect versus MVR.
6. Materials and standards. Chloride and fluoride exposure points, preferred codes for pressure vessels, and site environment (arid salar, coastal recycle hub, geothermal field).
7. Guarantees and testwork. Which numbers the vendor will guarantee (assay, d50, recovery, specific consumption), on what testwork basis, and what pilot work precedes the guarantee.

Attach a liquor analysis, however preliminary, to the enquiry — quotations built on assumed tenors get renegotiated after the order, never in the buyer’s favor. When you are ready to put a package like this to the market, request a quote with the checklist attached, and the response you get back will be an engineering document rather than a price list.

Conclusion

Lithium carbonate crystallization earns its place as the decisive section of a battery-grade plant: the carbonation–decarbonation loop is what converts a purified lithium liquor into a product that simultaneously holds ≥99.5% assay, a tight impurity schedule and an 8–15 µm crystal size, campaign after campaign, at 85–93% section lithium recovery. A project manager who specifies the loop deliberately — CO₂ recovery as standard scope, seeding and ageing as engineered features, wash efficiency as the purity policy — will buy a plant that qualifies with offtakers, not merely one that makes a white powder. Confirm every figure here against your offtake contract and your testwork, then tender the section as the engineered package it is.

Frequently Asked Questions

Why is lithium carbonate produced via a carbonation–decarbonation loop instead of direct evaporation?

Because lithium carbonate’s solubility decreases with temperature and is only moderate to begin with, direct evaporation would co-precipitate sodium, potassium and sulfate salts with the product. The loop dissolves lithium as soluble bicarbonate on CO₂ absorption, then precipitates only Li₂CO₃ on heating, rejecting impurities to the mother liquor. Each pass is a purification step, and the released CO₂ is recovered and reused.

What crystal size should battery-grade lithium carbonate target?

Typical offtake schedules ask for a d50 of about 8–15 µm with a narrow distribution, bulk density around 0.9–1.3 g/cm³ and moisture at or below 0.2% — typical values to be confirmed against your offtake contract. Crystal size is set in the de-carbonation step through temperature control, seeding and ageing hold time. Milling can enforce top size but risks lowering bulk density, so plants prefer to hit size in precipitation.

How much CO₂ does the loop consume per tonne of product?

With a CO₂ recovery system returning de-carbonation off-gas to the carbonation column, indicative consumption is 0.5–0.8 t CO₂ per tonne of Li₂CO₃ product; without recovery the figure is substantially higher, approaching stoichiometric demand. Treat the number as indicative and confirm it with the vendor’s gas balance for your liquor, since absorption efficiency and loop bleed policy move it materially.

Where does boron go in a lithium carbonate plant?

Boron is removed upstream of the crystallization loop, typically by solvent extraction or ion exchange on the concentrated brine, because no stage inside the loop rejects it. Battery-grade schedules cap boron around 0.001% (typical value, contract-dependent), so plants carry boron analysers on the loop feed and build redundancy into the removal step.

What overall lithium recovery should the section achieve?

A well-run purification-and-crystallization section typically recovers 85–93% of the lithium presented to it — indicative figures to be confirmed by your mass balance and testwork. Losses concentrate in three places: lithium retained in bleeding mother liquor, wash losses at the centrifuge, and residual bicarbonate in loop bottoms. Bleed-recovery hardware, wash-water quality and seeding discipline decide where a plant lands.

Which materials of construction suit the carbonation and de-carbonation loop?

Stainless steels serve most sulfate-route loops, but chloride-bearing brine plants usually step up to higher-alloy grades or lined equipment at hot, wet chloride exposure points such as de-carbonation off-gas handling. Recycle-fed loops add fluoride and organic exposure, pushing toward lined reactors and carbon treatment ahead of the column. Confirm the selection against your liquor analysis — materials are among the largest cost swings between quotations.

Can one plant serve both brine and recycled-battery feed?

Yes, and many newer plants plan for it: the loop, separation and drying equipment is essentially feed-agnostic, so a dual-feed plant shares the back half and runs separate front ends for brine concentration and black-mass leach purification. The design consequences sit upstream — extra polishing for fluoride, organics and metals from recycle, and scale management from salar or geothermal brines. Validate the loop on both liquors in pilot.

References

1. U.S. Geological Survey. “Lithium Statistics and Information.” National Minerals Information Center. https://www.usgs.gov/centers/national-minerals-information-center/lithium-statistics-and-information
2. U.S. Geological Survey. Mineral Commodity Summaries 2025 — Lithium. https://pubs.usgs.gov/periodicals/mcs2025/mcs2025-lithium.pdf
3. Hydrometallurgy (Elsevier journal) — peer-reviewed literature on lithium purification and crystallization circuits. https://www.sciencedirect.com/journal/hydrometallurgy
4. Society for Mining, Metallurgy & Exploration (SME) — reference literature on mineral processing and extractive metallurgy.
5. Lexin-tech. “Evaporator and Crystallizer Resources.” https://www.lexin-hydromet.com/resources/evaporator-crystallizer/

Changelog

• 2026-10-29: Initial publication — flowsheet, carbonation–decarbonation loop, specifications, equipment and inquiry guidance for battery-grade lithium carbonate crystallization.