Researchers at Columbia Engineering have published proof-of-concept results in Joule for a temperature-responsive direct lithium extraction process called switchable solvent selective extraction (S3E), achieving lithium selectivity up to 10 times higher than sodium and 12 times higher than potassium in general brines, with even stronger performance in Salton Sea simulations. The system requires no binding chemicals, regenerates its own solvent, and can be powered by low-grade waste heat or solar thermal energy, potentially opening geothermal and oilfield brines that conventional solar evaporation cannot touch.
Introduction
A research team at Columbia Engineering has demonstrated a direct lithium extraction process that outperforms existing chemical rivals on selectivity while sidestepping the energy and reagent penalties that have kept many DLE concepts locked at the laboratory bench. The technique, called switchable solvent selective extraction (S3E, pronounced "S three E"), uses a thermally responsive amine solvent to pull lithium preferentially from complex, low-concentration brines, then releases a purified lithium product stream when heated through a modest temperature swing. The solvent regenerates itself in the same step, ready for another cycle.
The work was published in Joule (Cell Press) on January 21, 2026, by a team led by Ngai Yin Yip, La Von Duddleson Krumb Associate Professor of Earth and Environmental Engineering at Columbia University, alongside co-authors Elizabeth Dach, Juliana Marston, Sara Abu-Obaid, and Allison Peng. Funding came from the U.S. Department of Energy Office of Science Basic Energy Sciences program and the National Science Foundation Graduate Research Fellowship.
The timing matters. The global DLE market reached an estimated $10.7 billion in 2025 and is projected to reach $38.4 billion by 2035. Meanwhile, conventional solar evaporation, which still accounts for nearly all brine-sourced lithium production, is running into hard physical limits: it requires two or more years of pond residence time, vast flat land in arid climates, and high initial lithium concentrations. The brines that do not fit those constraints, including geothermal reservoirs and oilfield co-produced waters, represent an enormous untapped resource that existing technology simply cannot address at scale. S3E was designed precisely for those neglected feedstocks.
How S3E Works: Temperature as the On-Off Switch
The fundamental insight behind S3E is that a class of amine solvents naturally shifts between water-attracting and water-repelling states as temperature changes. At lower temperatures, the solvent is sufficiently hydrophilic to draw lithium ions and water out of an aqueous brine and into an organic phase. At moderately elevated temperatures, that affinity drops, causing the lithium-carrying water to separate back out as a purified, lithium-enriched aqueous stream. The solvent, now returned to its hydrophobic state, is ready to be cooled and used again.
The physical arrangement takes advantage of density differences between the heavier lithium brine and the lighter low-polarity solvent. Density-driven countercurrent convection provides passive mixing without requiring mechanical agitation, which simplifies the engineering footprint. Once the two phases settle at low temperature, lithium partitions preferentially into the organic phase while competing cations, primarily sodium and potassium, remain largely excluded.
The selectivity mechanism does not depend on bespoke ligands or solid ion-exchange supports designed to recognize lithium's specific geometry. Instead, it emerges from how lithium ions interact with water molecules within the solvent environment, an interaction governed by fundamental hydration chemistry rather than engineered molecular recognition. That distinction matters for scale-up: the team found consistent lithium preference across multiple amine solvents with different chemical structures, suggesting the selectivity is a class-level property rather than a quirk of one particular compound.
Magnesium, one of the most troublesome contaminants in geothermal and salar brines, is handled by a built-in precipitation step. The basicity of the amine solvent causes magnesium to precipitate spontaneously as Mg(OH)2, removing it completely without requiring a separate reagent addition or processing unit. That integration is practically significant: in many competing DLE processes, magnesium removal demands its own dedicated treatment stage, adding cost and complexity.
The Selectivity Numbers and What They Mean
In laboratory tests using synthetic brines, S3E achieved lithium-to-sodium selectivity as high as approximately 10 and lithium-to-potassium selectivity as high as approximately 12. Those headline figures are themselves strong results in the DLE field, where maintaining high lithium selectivity across varying brine chemistries has been a persistent challenge. But the more striking performance came from the Salton Sea simulation.
When the team formulated synthetic brines to mimic the geothermal fluids under California's Salton Sea, which present an extreme case of low lithium concentration against a backdrop of abundant sodium, potassium, and magnesium, the selectivity numbers climbed significantly. Lithium-to-sodium selectivity reached approximately 13 and lithium-to-potassium selectivity reached approximately 24. Critically, the team maintained those separation ratios even when lithium concentrations in the feed were 1,000 times lower than competing sodium or potassium concentrations: an operational condition that would defeat adsorption sorbents relying on concentration-driven uptake kinetics.
Recovery across repeated cycles points toward a viable continuous process architecture. Using a single batch of solvent across four semibatch extraction cycles, the team recovered close to 40% of the lithium present in the feed. Yip's group was candid about the current state: this is a proof-of-concept study and the system has not been optimized for maximum yield. But 40% recovery with demonstrated solvent regenerability and preserved selectivity across all four cycles establishes the technical prerequisites for continuous operation. No binding site degradation, no accumulating impurity penalty, and no external reagent consumption between cycles.
For context, the DLE field has competitors across several technology platforms: adsorption using aluminum-based or manganese oxide sorbents (the current commercial leader, particularly in Argentina and China); membrane-based ion-selective separation; and electrochemical extraction methods. Each carries trade-offs. Adsorption systems require periodic sorbent regeneration with acid washes and can lose selectivity when brine chemistry shifts. Electrochemical approaches are energy-intensive and face materials durability challenges. Membrane separation achieves high selectivity but struggles with flux at very low lithium concentrations. S3E's distinguishing claims are that it requires no binding chemicals, no acid regeneration, and no post-processing beyond the temperature swing itself.
Energy Requirements: Waste Heat as the Economic Unlock
Perhaps the most commercially consequential aspect of S3E is its energy profile. The temperature swing required to toggle the solvent between its hydrophilic and hydrophobic states is modest by industrial standards, well within the range that can be supplied by low-grade waste heat at temperatures that power plants, industrial facilities, and geothermal installations routinely discard. Solar thermal collectors of the type already deployed at modest scale in the American Southwest can also provide the necessary heat input.
"This is a new way to do direct lithium extraction," said Yip. "It's fast, selective, and easy to scale. And it can be powered by low-grade heat from waste sources or solar collectors."
The energy economics become particularly attractive in the Salton Sea context. The 11 geothermal power plants operating in the Known Geothermal Resource Area already produce 400 megawatts of electricity and generate waste heat as a byproduct. A lithium extraction process sited at those facilities could treat the same brine that powers the turbines, using thermal energy that would otherwise be rejected to the environment. The Lawrence Berkeley National Laboratory's 2023 characterization of the Salton Sea geothermal resource estimated that existing plant capacity alone is co-producing roughly 21,500 tons of lithium-in-brine annually, none of which is currently being recovered.
Yip framed the broader stakes in terms that connect materials science to the credibility of the clean energy transition itself: "We talk about green energy all the time. But we rarely talk about how dirty some of the supply chains are. If we want a truly sustainable sustainable transition, we need cleaner ways to get the materials it depends on. This is one step in that direction." For comparison, industry analyses of DLE broadly estimate that the process class requires roughly 99% less water per ton of lithium than conventional evaporation-based methods and emits substantially less carbon dioxide, though those figures apply to the DLE category as a whole rather than to S3E's specific configuration.
The Salton Sea and the Strategic Resource It Represents
The Salton Sea sits at the center of U.S. domestic lithium ambitions for reasons that go well beyond its geographic novelty. A 2023 assessment by DOE's Lawrence Berkeley National Laboratory estimated the geothermal reservoir beneath the lake contains approximately 18 million metric tons of lithium in solution, with at least 4 million tons confirmed through drilling data. That resource, if developed, would be sufficient to supply batteries for more than 375 million electric vehicles. It would also materially shift the United States away from its current posture of near-total import dependence for lithium supply.
The challenge is the brine chemistry. Salton Sea geothermal fluids are hypersaline, hot, and contain high concentrations of competing cations. Conventional solar evaporation is not merely inefficient in this environment; it is non-functional. The brine cannot be spread in open ponds because the climate, topography, and chemistry of the region do not support the multi-year evaporation process that works in the Atacama Desert or parts of Nevada. As Yip put it: "There's no way solar evaporation alone can match future demand. And there are promising lithium-rich brines, like those in California's Salton Sea, where this method simply can't be used at all."
Several companies are already investing at the Salton Sea ahead of a viable large-scale extraction process. Berkshire Hathaway Energy Renewables operates the majority of the existing geothermal capacity in the area through its CalEnergy subsidiary. Controlled Thermal Resources has secured hundreds of millions in investment, including participation from General Motors and Stellantis. EnergySource Minerals is advancing its own DLE pilot with DOE support. The common thread across these projects is that none of them has a fully proven extraction chemistry at commercial scale for the Salton Sea's specific brine profile. S3E's demonstrated Salton Sea simulation results, particularly the 13x sodium selectivity and 24x potassium selectivity with complete magnesium precipitation, position it as a candidate chemistry for exactly that gap.
The Salton Sea is not the only relevant resource. Alberta's deep subsurface brines, associated with oil, gas, and geothermal systems, represent an analogous opportunity: lithium-bearing saline waters already produced through existing energy infrastructure, awaiting an extraction process that can handle their chemistry economically. S3E's demonstrated robustness across multiple amine solvent types suggests adaptability to different brine compositions, which would matter enormously for a technology seeking to serve geographically and geochemically diverse feedstocks.
Where This Sits in the Broader Supply Chain Effort
It is worth placing S3E precisely in the current supply chain landscape. Building on my coverage of the DOE's $162 million critical minerals recovery program in September 2026, which targeted recovery of rare earths, scandium, and antimony from industrial byproduct streams, the Columbia research represents a complementary strategic thread: basic energy sciences investment aimed at unlocking entirely new lithium resource categories rather than incrementally improving existing extraction from established deposits. The DOE Office of Science BES funding model that supported S3E prioritizes fundamental discovery, meaning the timeline to commercial deployment is measured in years, not months. But the selectivity data and solvent regenerability results are strong enough that they should attract serious interest from the commercial DLE developers already active at the Salton Sea and in similar geothermal settings.
The broader DLE competitive landscape is crowded and still sorting itself out. Adsorption-based systems have the most commercial validation, but their performance in very low-grade or contaminated brines is variable. Electrochemical extraction can achieve high purity but faces energy cost headwinds that become significant at industrial throughputs. Membrane approaches are progressing but struggle with concentration polarization at the low lithium levels characteristic of geothermal and oilfield brines. S3E enters this landscape with a different value proposition: it does not optimize around one of those trade-offs, but instead sidesteps the binding-chemistry paradigm entirely, relying on bulk thermodynamic properties of a solvent class rather than the engineered specificity of a sorbent or membrane.
Yip's prior work on temperature swing solvent extraction for desalination, and his co-founding of Trident Desalination to commercialize that platform, provides a meaningful commercialization precedent. The S3E platform is a direct extension of the same switchable hydrophilicity principle applied to ion separation rather than water-salt separation. Columbia Engineering has indicated openness to collaboration and licensing discussions with industry partners, which is an early signal that the team is thinking about a commercialization pathway even at this proof-of-concept stage.
Conclusion: Proof of Concept, but the Right Proof at the Right Time
S3E is not a commercial product. Its authors are clear that yield optimization, process integration engineering, and pilot-scale validation remain ahead. A 40% lithium recovery across four cycles with a single solvent batch is a demonstration of regenerability and selectivity persistence, not a production efficiency target. The gap between a Joule paper and a functioning brine processing facility is wide, and the DLE field has seen more than a few promising laboratory results fail to translate.
That said, the combination of properties that S3E brings to the table is unusual in the DLE literature. Selectivity that exceeds existing benchmarks, maintained across extreme concentration ratios and across multiple solvent chemistries. Magnesium exclusion built into the process chemistry rather than added as an external step. An energy requirement modest enough to be met by thermal sources that cost nothing at co-located geothermal facilities. And a platform with at least one prior commercial translation in the desalination domain, giving Yip's group a cleaner line of sight to scale-up than most academic DLE research teams can claim.
The strategic context amplifies the significance of the timing. As covered throughout this publication's reporting in recent months, the combination of China's tightening export licensing architecture for critical minerals, surging investment in domestic processing infrastructure, and accelerating demand from EV and battery manufacturers has created a genuinely urgent need for new domestic supply pathways. Lithium is not subject to the same export control regime as rare earth elements, but it faces its own supply concentration risk: the United States imports nearly all of the lithium it consumes, and that dependence runs through a small number of producing countries and refining nodes.
The Salton Sea alone, if developed using a process capable of handling its brine chemistry, could in principle satisfy U.S. lithium demand for decades. Whether S3E, or a refined descendant of it, becomes part of that development will depend on engineering work that has not yet been done. What the Columbia team has established is that the thermodynamic basis for solvent-based lithium selectivity in hypersaline, low-concentration brines is real, reproducible, and potentially generalizable across a range of amine chemistries. In a field defined by the gap between laboratory promise and industrial reality, that foundation is not nothing. It is, precisely, the right kind of proof of concept at the right moment in the supply chain conversation.