What is direct lithium extraction (DLE)?

Lithium occurs in economically recoverable concentrations in several types of geological resources, but most lithium produced today comes from two broad sources: hard-rock ores and lithium-bearing brines. Hard-rock mining currently supplies roughly 60% of global lithium, with brines providing much of the balance. Both routes are evolving rapidly as demand grows from electric vehicles, stationary energy storage, and other battery applications.

This overview focuses on lithium brines and the emergence of direct lithium extraction (DLE) as an alternative to conventional solar evaporation. DLE is not a single technology. It is a family of separation processes with very different recovery rates, water requirements, reagent consumption, impurity tolerance, capital costs, and operating costs. Understanding those differences is essential when evaluating a DLE project. We will explain where lithium brines occur, how DLE developed, the principal technology families, the economic trade-offs between them, and some of the challenges that emerge when a process is scaled from the laboratory to a commercial plant.

Where does lithium brine come from?

Lithium is dissolved at low concentrations in many natural waters. Seawater, for example, contains only 0.2 milligrams of lithium per liter (mg/​L) – far too little for commercial recovery. In certain geological settings, however, water remains in contact with lithium-bearing rocks for long periods and becomes much more highly enriched.

These lithium-bearing fluids occur in several forms. Continental brines are found in closed sedimentary basins, including the salars of South America’s Lithium Triangle in Argentina, Bolivia, and Chile. Lithium-rich brines also occur in deep sedimentary formations such as the Smackover Formation in the southern United States. Geothermal fluids and oilfield produced waters represent additional potential lithium resources for mining.

The first brine characteristic that matters economically is lithium concentration. In general, higher concentrations are advantageous because less brine must be pumped and processed for each tonne of lithium produced. Some of the world’s highest-grade brines occur in the Atacama region of Chile, where lithium concentrations can reach well above 1,000 mg/​L. But concentration alone does not determine whether a resource is economic for mining. Projects are also being developed around lower-concentration brines when other characteristics of the resource and process are favorable.

The second major consideration is brine chemistry. Lithium is accompanied by much larger quantities of other dissolved species, commonly including sodium, potassium, magnesium, calcium, chloride, sulfate, carbonate or bicarbonate, boron, silica and, in some resources, bromide and other trace components. Brines may also contain suspended solids, dissolved gases, hydrocarbons or other organic material.

These brine components impact mining economics because a lithium process does not need only to extract lithium from water – it must also extract lithium from everything else. Species that interfere with the extraction process, consume reagents, or precipitate in equipment, can increase both capital and operating cost. Brine temperature, pressure and flow characteristics also vary widely between resources and can strongly influence process design and economics.

Some brine resources have additional attributes that can improve their cost position. Hot geothermal brines, for example, can potentially support both lithium production and geothermal power generation. Vulcan Energy’s Lionheart project in Germany is being developed on this model and reached financial close on its €2.2 billion financing package in May 2026. Other emerging resources include surface brines such as Utah’s Great Salt Lake, where Lilac Solutions is advancing a DLE project based on ion-exchange technology.

From evaporation ponds to direct lithium extraction

North America’s first dedicated lithium-brine operation began in Nevada’s Clayton Valley. Foote Mineral Company began extracting lithium-bearing brine at Silver Peak in 1966, and lithium carbonate production followed in 1967. The operation, now owned by Albemarle, remains in production today.

Silver Peak uses the ​‘solar evaporation’ route. Brine is pumped from wells into a series of large evaporation ponds, where sunlight and dry air remove water while salts crystallize progressively from solution. This is followed by chemical purification along with conversion to lithium carbonate product. Brine at Silver Peak spends roughly a year progressing through the evaporation-pond system before reaching a concentration suitable for downstream processing.

Similar evaporation-based operations were subsequently developed at lithium resources in South America and China. The approach can be highly economic where brine chemistry and climate are favorable, but it has important limitations. Evaporation requires large areas of land, depends on suitable weather, leaves a significant portion of the lithium unrecovered, and requires months for brine to pass through the pond system.

A fundamentally different approach reached commercial production in Argentina in 1997. FMC started up their Salar del Hombre Muerto operation – today Rio Tinto’s Fénix operation – and has used a proprietary selective adsorption process to recover lithium directly from brine rather than relying primarily on solar evaporation. Fénix has produced lithium carbonate continuously since 1997 and remains one of the longest-running commercial examples of DLE.

In an adsorption-based DLE process, brine contacts a solid material engineered to preferentially capture lithium. Most of the sodium, magnesium, calcium and other dissolved species remain in the brine and pass through the process. The captured lithium is then released from the sorbent into a separate aqueous stream. That lithium-rich solution must subsequently be concentrated, purified and converted into a commercial lithium product such as lithium carbonate or lithium hydroxide.

DLE did not immediately displace evaporation ponds. Fénix remained unusual for many years, reflecting the site-specific nature of brine chemistry, the maturity of available separation materials, lithium prices, and the economics of competing production routes. China became the next major center of commercial development. Qinghai Lanke Lithium began production at its Qarhan operation in 2012 using adsorption technology to recover lithium from brine associated with potash production. Additional Chinese adsorption-based DLE projects followed later in the decade.

Commercial deployment outside China accelerated much more recently. Eramet commissioned its 24,000-tonne-per-year Centenario-Ratones plant in Argentina in 2024 and produced its first lithium carbonate in December of that year. The plant uses Eramet’s proprietary DLE technology and is among the largest new commercial DLE installations built outside China.

Today, DLE is moving from a small number of unusual projects toward a much broader industrial deployment. Established lithium producers, energy companies, and technology developers are pursuing projects across South America, North America, and Europe. Many large companies are betting on building multiple DLE projects ahead, including Rio Tinto, Albemarle, POSCO, Eramet, ExxonMobil, Chevron, and Equinor. The opportunity is driven by economics: DLE can achieve much higher lithium recovery and can make resources commercially accessible where evaporation ponds are technically unsuitable or economically unattractive.

How does direct lithium extraction work?

“DLE” describes a wide range of technologies rather than one standardized process. Their performance, and ultimately their economics, depends on the interaction between the separation technology, the process design, and the specific brine being treated. DLE processes all have the same overall objective: recover lithium selectively from brine and produce a concentrated lithium stream that is suitable for conversion into a saleable product. The path from raw brine to purified lithium, however, varies substantially between technologies and resources.

Depending on the brine chemistry and DLE technology, unwanted species may be removed at several points in the process. Some are removed during pre-treatment before lithium extraction. Others are rejected directly by the DLE separation step. Additional impurities may be removed during concentration, polishing, or final product conversion. Water is also removed at multiple stages, so lithium concentration and impurity removal do not always occur as separate, sequential operations.

A typical DLE flowsheet may include four broad functions:

Pre-treatment

Raw brine may contain suspended solids, organics, dissolved gases, or dissolved species that interfere with the lithium-separation process or downstream equipment. Pre-treatment is tailored to the resource and may remove selected components before the brine reaches the DLE step. The required treatment can vary widely between resources.

Selective lithium separation

The core DLE step preferentially separates lithium from the bulk brine. Depending on the technology, this may involve adsorption, ion exchange, solvent extraction, membranes, or another selective separation mechanism. Some technologies capture lithium while allowing most other dissolved species to remain in the depleted brine. Others selectively transport lithium or reject unwanted species in different ways.

Concentration and purification

The lithium-containing stream leaving the core separation step usually requires further concentration and purification. These functions may occur in several operations rather than one dedicated step. Reverse osmosis, evaporation, precipitation, selective membranes, solvent extractions, chemical treatment, or other polishing steps may remove water, reject impurities, or do both at the same time.

Product conversion

The purified lithium stream, usually lithium chloride, is converted into the desired commercial product, most commonly lithium carbonate or lithium hydroxide. Final purification may continue during this stage to meet product specifications. The depleted brine is returned to the reservoir or otherwise disposed of.

What are the main types of DLE technology?

Four families are in commercial or near-commercial use. Each trades selectivity against durability, reagent consumption, and cost.

Adsorption

A solid sorbent (typically aluminum‑, manganese- or titanium-based) captures lithium from the brine and releases it with water. Adsorption tolerates a wide range of brine chemistries and consumes little reagent compared to other DLE methods. Sorbent durability is a governing variable of the total operating cost; a plant’s sorbent inventory must be replaced on a recurring basis, so cycle life (how many load-and-release cycles the sorbent survives) and attrition rate (how fast it physically wears away) matter as much as capacity. For a sorbent that requires replacement every 5 years, this works out to be about <10% of total OpEx. Adsorption is the most widely deployed family in brine DLE today.

Ion exchange

A resin swaps hydrogen ions for lithium ions, then releases the lithium when acid is applied. Ion exchange can deliver high product purity, and it consumes acid and base continuously to do so. Reagent cost and neutralization of the resulting waste streams can drive project economics higher.

Solvent extraction

Brine is contacted with an immiscible organic liquid containing molecules designed to selectively bind lithium. Lithium transfers from the brine into the organic phase. Those two phases are separated. The lithium is then stripped into a separate water stream. Solvent-extraction systems can achieve strong selectivity in difficult brines, but their economics depend on reagent consumption, solvent losses, phase separation, organic handling, and management of any degradation products or waste streams.

Membrane separation

Membrane-based DLE uses pressure, concentration gradients, or an electric field to selectively transport lithium or reject competing ions. Depending on the membrane chemistry, separation may be based on ion size, charge, hydration behavior, or specific chemical interactions. Membrane systems can be compact and modular, but fouling, selectivity, membrane life, energy consumption, and performance in highly saline brines remain important scale-up challenges.

Commercial maturity varies significantly within each DLE family. Adsorption has the longest operating history at commercial scale, while ion-exchange, solvent extraction, and membrane technologies remain in pilot and demonstration phases.

These DLE categories are useful distinctions, but they do not determine project performance. Two technologies within the same family can differ substantially in recovery, selectivity, water use, reagent consumption, durability, and lithium concentration. Performance also depends strongly on the brine being treated and on how the separation step is integrated with pretreatment, concentration, purification, and product conversion.

Why is DLE gaining ground now?

Four things have changed over the last five to ten years.

Demand is outpacing supply. Analysts disagree on timing, but most now show lithium moving into structural deficit in the late 2020s, with cumulative shortfalls measured in millions of tonnes of lithium carbonate equivalent (LCE) by 2050. Grid-scale energy storage and data center growth are adding demand on top of what electric-vehicle forecasts predicted. Closing that gap requires responsive lithium supply from resources that existing methods can’t develop.

The pond-friendly brines are taken. Salars that suit evaporation ponds are largely developed. What remains falls into three groups. Some brines are uneconomic for ponds, such as lower-grade and higher-magnesium brines. Some sit in regions too cold, wet, or cloudy for evaporation. And geothermal fluid and oilfield produced water flow continuously from geothermal and oil and gas wells, in volumes that usually have to be returned underground, which rules out ponds entirely. DLE can process all three.

Expectations from regulators and communities have risen. Water use in arid basins is now a permitting issue, and land footprint and brine management carry more weight in project approvals. A process that returns brine to the reservoir and occupies a fraction of the land is better placed to meet those expectations than a field of evaporation ponds.

The technology has matured. Sorbent-based DLE has moved from pilot claims to active deployment programs across the industry, including at Albemarle, Eramet, Equinor/​Standard Lithium, SQM, and Rio Tinto. Producers can now see the technology running at scale, which makes it easier to finance and permit the next project.

What costs contribute most to DLE economics?

Capital intensity for brine DLE generally falls between hard rock and evaporation ponds. The number that decides a project’s viability is the levelized cost per tonne of lithium carbonate equivalent across the asset’s life: total lifetime cost divided by total lifetime production. Five things drive this number:

• Recovery. The share of the lithium in the brine that the plant captures. A plant that recovers 90 percent produces 29 percent more lithium than one recovering 70 percent from the same brine. 
• Consumables. Sorbent, resin and solvent inventories degrade and have to be replaced. Media with a three-year service life and media with an eight-year life can have identical initial costs and very different ongoing costs.
• Lithium-stream purity and concentration. How much magnesium, calcium, sodium, boron and other impurities accompany the lithium through the process. Each one has to be removed before the product meets battery-grade specifications, and every purification step adds reagents, equipment and some lithium loss. A cleaner stream from the sorbent section of the process means a shorter, cheaper path to a saleable product.
• Water and energy. Removing water from a dilute lithium stream can be a major consumer of both water and energy. The more concentrated the lithium stream leaving the separation process, the less downstream concentration duty is generally required.
• Uptime. A plant that runs 95 percent of the year and one that runs 80 percent cost the same to build, but the first produces 19 percent more.

Published DLE cost figures are hard to compare because they come from different brine chemistries, plant scales, assumptions about what the estimate includes (such as whether pre-treatment or conversion is counted), and assumptions about how long the separation media lasts. A production cost per tonne of lithium quoted without the brine chemistry, recovery rate, eluate purity, and media life behind it can’t be used to compare the lifetime cost of DLE technologies.

Where does DLE still struggle?

Several prominent DLE programs have underperformed at field scale after strong bench-scale results. The failures tend to fall into six areas.

• Scale-up. A single column in a laboratory behaves nothing like a multi-column plant with continuous flow, cycling valves, and real brine variability. How brine moves through and between columns (column sequencing, pressure drop, channeling) and the control logic that runs it are where DLE projects often fail, and these issues are expensive to solve after a plant is built.
• Brine specificity. Performance depends on the technology and the resource together. A process validated on one brine can lose selectivity, foul or shed media on another. Magnesium-to-lithium ratio, boron, silica, organics and temperature all matter.
• Media degradation. Accelerated testing produces a projection of service life rather than a direct measurement of it. Few DLE systems have multi-year operating records on real brine, so most lifespan figures in circulation are extrapolations.
• Impurity handling. Raw brine may contain suspended solids, organics, dissolved gases, or dissolved species that interfere with the lithium-separation process or downstream equipment. That front end is specific to each resource, can be a large share of capital cost, and is often underestimated in early studies. 
• Energy and reagents. DLE substitutes electricity, reagents, and process control for evaporation’s solar energy. On a grid with high carbon intensity or high energy prices, that substitution changes both the environmental and the economic case.
• Reinjection. Returning depleted brine to the reservoir is the optimal way to manage it, and permitting it depends on the technology and reagents used in the process, the hydrogeology of a given salar, and regulator/​community consent. 

How Summit approaches DLE

Summit Lithium Technologies builds sorbent-based DLE systems for brine resources in Chile, Argentina and the United States. Summit’s technology is denaLi™, an integrated system combining proprietary column sequencing, eLivate™ sorbent, and a predictive modeling platform used to design and optimize each plant against its specific brine. 

Scale-up

About 80 percent of Summit’s commercial design is standardized, and that core is proven before it reaches a customer’s site. Column sequencing and control logic are developed on a multi-column bench platform, then tested in an in-house system with commercial-height columns. Pressure drop, flow distribution and sequencing are validated at the dimensions they will have in a commercial plant, so a new project doesn’t have to solve them again.

Brine specificity

The remaining 20 percent of the design is set by the resource. Lithium producers and asset developers can ship 30 cubic metres of brine to Summit’s rapid validation denaLi™ demonstration plant in Santiago, which runs it continuously. Within two to three months of a defined test scope, the plant returns an asset-specific commercial design report. The report covers expected recovery and impurity rejection, water and energy use, engineering design factors, and capital and operating cost estimates. It also supports a fixed-price proposal with contractually defined performance commitments. Depending on scope and asset, this can substantially shorten the timeline and cut the cost of validation compared with a traditional field demonstration.

The results feed Summit’s predictive modeling platform, which configures the plant for that brine’s chemistry. The platform runs roughly 10,000 multi-column DLE simulations a day and is refined continuously against bench, pilot, and demonstration data.

Media life

eLivate™ is engineered for balanced performance across capacity, kinetics, selectivity, and durability, because media life can drive operating cost as much as recovery does. Summit’s reliability engineering systems put the sorbent through accelerated chemical and mechanical stress; that testing projects more than 10,000 cycles with minor degradation, retention of more than 98 percent of lithium recovery and more than 95 percent of impurity rejection, and a service life of five to eight years or more. 

Impurities, water, and energy

Impurity rejection above 95 percent reduces the load on downstream purification. Because the design report measures rejection on the customer’s own brine, downstream conversion is sized from data instead of assumptions. The same report predicts water and energy use for the specific asset. 

Aggregated across all of Summit’s systems, denaLi™ has delivered DLE-specific water use of 5 to 17 cubic metres per tonne LCE and uptime of 95 percent or better.

Testing a brine resource? Summit’s testing facilities, in both North and South America, can run your brine and return an asset-specific commercial design report. Contact sales@summitlithium.tech to learn more.

Frequently asked questions

What does DLE stand for?

Direct lithium extraction. ​“Direct” means the lithium is recovered straight from the brine. The conventional route first concentrates the brine in solar evaporation ponds and recovers the lithium afterward.

Is DLE commercially proven?

Sorbent-based DLE runs commercially today at a growing number of projects across the Americas, Europe, and Asia. Most ion-exchange, solvent extraction and membrane processes currently operating are at pilot or demonstration scale.

How much water does DLE use?

The standard measure is cubic metres of water per tonne of lithium carbonate equivalent (m³/​t LCE). Published figures for brine DLE range from single digits to several tens of cubic metres per tonne. The number depends on the process and on the brine at each project. Published figures may vary depending on whether they cover only DLE-specific water use, or end-to-end water use through conversion to a battery-grade material.

What is a good lithium recovery rate for DLE?

Most DLE processes report recoveries above 90 percent at the extraction step, compared with roughly 40 to 60 percent for evaporation ponds. Whole-plant recovery is usually somewhat lower once concentration and conversion losses are counted. Recovery is also only one measure; a process with slightly lower recovery can still be the better choice if it uses less water, energy or reagent, or rejects impurities more cleanly and therefore lowers downstream costs.

Does DLE work on oilfield produced water and geothermal brine?

Yes, with more pre-treatment. These resources usually carry more dissolved solids, more organics and a less consistent composition than salar brine. Their lithium concentrations also tend to be lower, so more volume has to be processed for each tonne of product.

How long does a DLE project take to reach production?

Engineering and construction take about as long as for any process plant of similar scale, typically several years. The bigger variable is validation: proving the technology works on a specific brine and producing the data needed to size a commercial plant. Historically that meant years of field demonstration. In-house validation facilities and modular plant designs have shortened it considerably.

Is DLE better than evaporation ponds?

For most resources now in development, DLE recovers more lithium, on less land, and works where ponds can’t. Ponds are still cheaper to operate where the climate and brine chemistry suit them. The right choice depends on the resource.