Researchers at Columbia Engineering have published results in Joule describing Switchable Solvent Selective Extraction, or S3E, a novel direct lithium extraction method that uses a temperature-responsive amine solvent to pull lithium from brine at room temperature and release a purified stream upon heating. In laboratory tests against synthetic Salton Sea brines, S3E extracted lithium up to 10 times more selectively than sodium and 12 times more selectively than potassium, even when competing cations outnumbered lithium by 1,000 to one. The approach requires no specialized binding chemicals and can theoretically be powered by waste heat or solar collectors.
Introduction
A research team at Columbia University's School of Engineering and Applied Science has published a proof-of-concept for a direct lithium extraction method that does not look like anything currently operating at commercial scale. The technique, called Switchable Solvent Selective Extraction or S3E, uses the temperature sensitivity of certain amine solvents to toggle between lithium-capturing and lithium-releasing states, with no specialized ligands, chelating reagents, or elaborate post-processing required. The results appeared in Joule, the Cell Press journal focused on sustainable energy science, in January 2026.
Lead investigator Ngai Yin Yip, the La Von Duddleson Krumb Associate Professor of Earth and Environmental Engineering at Columbia, describes the approach with notable directness. "This is a new way to do direct lithium extraction," Yip said. "It's fast, selective, and easy to scale. And it can be powered by low-grade heat from waste sources or solar collectors."
The timing of the publication is worth noting. As explored in my coverage of the ex-China supply chain's operational milestones in October 2026, the broader critical minerals field is shifting from feasibility studies toward commissioning decisions. S3E is not yet at that stage, but the chemistry it introduces could eventually reshape how the industry thinks about brine resources that evaporation ponds and conventional direct lithium extraction cannot reach economically.
How S3E Works: Hydrophilicity as a Dial
The core principle behind S3E is thermally switchable hydrophilicity. Certain amine solvents change their affinity for water depending on temperature. Below a threshold, the solvent is hydrophilic, meaning it mixes readily with water and, crucially, pulls lithium ions along with it from a surrounding brine. Above a threshold, the solvent becomes hydrophobic, shedding the water and the dissolved lithium into a separate, concentrated aqueous stream while the solvent itself regenerates for reuse.
The operating temperature window is modest: roughly 5 degrees Celsius at the low end to 85 degrees Celsius at the high end, comfortably below boiling. This is significant because it means the energy input required to trigger the switch is small relative to processes that depend on high-temperature calcination, electrochemical potential, or chemical stripping agents.
In practical terms, the extraction step combines brine with a low-polarity amine solvent, producing a biphasic system in which an organic phase and an aqueous phase coexist. The organic phase selectively draws lithium ions across the interface. When the system is then gently heated, the organic phase expels its lithium content into a concentrated aqueous product stream, and the solvent resets to its original state. The entire cycle then repeats.
Importantly, selectivity arises from the physics of how lithium ions interact with water molecules inside the solvent phase, not from bespoke molecular cages or solid supports engineered to cradle a specific ion. This distinction matters for scalability: a mechanism rooted in bulk solvent behavior is inherently easier to produce in quantity than a mechanism dependent on the precise geometry of a synthetic ligand or ceramic sieve.
Selectivity Performance: The 1,000-to-One Challenge
The headline figures from the Joule paper are striking. Across six amine solvents tested, including triethylamine, tert-octylamine, diisopropylamine, and N,N-dimethylcyclohexylamine, S3E consistently preferred lithium over sodium by ratios as high as 10 to one, and over potassium by ratios as high as 12 to one. The team also pushed the system under stress conditions designed to mirror the worst-case chemistry of real brine resources: lithium concentrations 1,000 times lower than the competing sodium or potassium. Selectivity held.
When tested against synthetic brines calibrated to resemble the Salton Sea geothermal reservoir in Southern California, performance improved further. Lithium-to-sodium selectivity reached approximately 13, lithium-to-potassium selectivity reached approximately 24, and lithium-to-magnesium selectivity was effectively infinite. That last figure reflects a useful side effect of the amine solvent's chemistry: the basicity of the solvent causes magnesium to precipitate out as magnesium hydroxide, a solid that separates naturally rather than requiring a dedicated removal step.
Magnesium is one of the persistent headaches of lithium brine chemistry. It shares a charge and a similar ionic radius neighborhood with lithium, and high magnesium-to-lithium ratios have made many otherwise promising brine resources economically unattractive for conventional processing. S3E's ability to exclude magnesium through an inherent precipitation mechanism, rather than a separately engineered purification stage, removes one of the field's more stubborn technical barriers.
Yield across four repeated cycles using the same solvent batch reached approximately 40 percent lithium recovery. The team describes this as practical rather than optimal, and the paper is explicit that S3E remains a proof-of-concept requiring further optimization. The 40 percent figure is not the performance ceiling; it is the floor established under early semibatch laboratory conditions.
The Salton Sea as Test Case: Why That Choice Matters
The Columbia team's decision to validate S3E against Salton Sea brine chemistry is not arbitrary. The Salton Sea Geothermal Resource Area in California's Imperial Valley sits above one of the densest lithium brine deposits identified in North America. A Lawrence Berkeley National Laboratory assessment estimated the recoverable lithium beneath the reservoir at roughly 4.1 million metric tons of lithium carbonate equivalent, with higher-end estimates reaching 18 million metric tons if assumptions about reservoir porosity and total extent are expanded. The DOE has cited that resource as potentially sufficient to supply more than 375 million electric vehicle battery packs.
Yip is direct about the inadequacy of conventional methods for this resource. "There's no way solar evaporation alone can match future demand," he said. "And there are promising lithium-rich brines, like those in California's Salton Sea, where this method simply can't be used at all."
The Salton Sea's geothermal brines are hot, high in dissolved solids, and rich in competing cations. Traditional evaporation ponds require flat, arid land, a dry climate, and 12 to 24 months of processing time, conditions that have no relevance to an underground geothermal reservoir accessed through wellheads. Sorbent-based direct lithium extraction systems, which use ceramic ion sieves or ion-exchange beads, are promising but can be sensitive to high temperatures and high suspended-solids loads typical of geothermal fluids. S3E's temperature-driven mechanism and tolerance for competing cations were tested precisely because the Salton Sea represents the kind of difficult, high-value resource that the industry most needs to unlock.
Commercial interest in the Salton Sea is already moving. Controlled Thermal Resources, one of the leading developers in the area, has indicated plans to begin geothermal power generation by the end of 2026, with lithium production to follow. The research context S3E provides is therefore timely even if S3E itself is years from deployment.
Where S3E Sits in the DLE Technology Landscape
Direct lithium extraction is a category, not a single technology. The field encompasses sorbent systems using lithium manganese oxide or lithium iron phosphate ion sieves, membrane-based electrochemical methods, and conventional solvent extraction using chelating agents such as organophosphorus compounds. Each approach has a different selectivity profile, energy demand, capital cost structure, and sensitivity to brine chemistry.
S3E's separation mechanism is explicitly described in the Joule paper as fundamentally different from all of these. It does not depend on coordination chemistry, does not use chelating or stripping agents, and does not require a solid support material that can foul, degrade, or require periodic replacement. The selectivity mechanism is thermodynamic rather than geometric: it emerges from how ions of different sizes and hydration energies partition between the solvent phase and the aqueous brine phase as the solvent's polarity shifts.
This distinction matters for the cost profile. Sorbent-based systems require the manufacture of precision ceramic or polymeric materials, the management of solid-liquid contact, and often multiple washing and elution stages to produce a clean lithium product. Membrane systems require engineered membranes that can be expensive and sensitive to scaling. S3E, in its laboratory form, requires only the amine solvent, temperature control, and phase separation equipment. If the chemistry translates to engineering practice, the capital and operating cost structure could be substantially different from existing DLE approaches.
The comparison to conventional evaporation ponds is even starker. Evaporation ponds recover 40 to 60 percent of lithium over 12 to 24 months, require vast flat land in arid climates, and produce a product that still requires substantial downstream processing. DLE in general, including S3E in principle, uses 99 percent less water per ton of lithium than conventional mining processes and produces no significant direct carbon dioxide emissions during operation. S3E adds to those general DLE advantages the specific potential to operate on low-grade waste heat, removing the need for a dedicated high-temperature energy source.
It is worth noting that Yip's laboratory at Columbia previously developed Temperature Swing Solvent Extraction, or TSSE, a related approach that applied switchable-hydrophilicity amine solvents to the desalination of hypersaline brines rather than to selective ion extraction. S3E represents a deliberate extension of that underlying chemistry into the selectivity domain, targeting the separation of lithium from sodium, potassium, and magnesium rather than simply removing salt from water.
What Remains to Be Solved Before S3E Leaves the Laboratory
The Columbia team is transparent about the gap between proof-of-concept and commercial viability. A 40 percent recovery yield after four cycles in a semibatch laboratory setup is meaningful as a demonstration that the solvent regenerates and that selectivity is preserved across repeated use. It is not, however, sufficient for a commercial operation. Lithium extraction economics depend on recovery yields well above 40 percent, and continuous-flow engineering poses challenges that semibatch experiments do not.
Solvent losses over many thousands of cycles will need to be characterized. The amine solvents used in S3E are organic compounds, and any losses to the aqueous phase or through volatilization at elevated temperatures would impose operating costs and raise environmental questions. The team acknowledges that optimization of both yield and solvent retention is required before the technology can be assessed for scale-up.
Brine chemistry variability is another open question. The Salton Sea brines that served as the primary validation case are geochemically unusual in ways that may favor or disfavor S3E relative to other brine types. Oilfield produced waters, which represent another large potential resource for lithium extraction, contain organic compounds, suspended solids, and microbial material that could interact with amine solvents in ways that laboratory-scale synthetic brines do not capture.
The paper's authors note that the work demonstrates the principle and establishes the selectivity benchmark. Engineering development from this point forward involves flow design, materials compatibility assessment, heat integration, and eventually pilot-scale validation, a sequence that typically takes years under the best circumstances. Given the compressed timelines facing Western lithium supply chains, particularly as defense procurement requirements and EV battery supply agreements accelerate domestic sourcing pressure, the question of how quickly S3E can move from bench to field is not academic.
The broader demand context provides the urgency. Global lithium demand is projected to grow more than 40 times by 2040, and brine resources, which constitute approximately 60 percent of global lithium reserves, currently contribute only 35 percent of production, precisely because conventional evaporation methods are unsuited to so many of them. Technologies that can access low-grade or geochemically challenging brines without massive infrastructure footprints are not merely interesting research directions; they are structural necessities for the energy transition.
Conclusion: A Mechanism Worth Watching
The Columbia Engineering S3E paper does not claim to have solved lithium extraction. It claims to have demonstrated a selectivity mechanism that works at the laboratory scale, that survives repeated cycling, and that operates on a thermodynamic principle different from anything currently in the DLE toolkit. Those claims are supported by peer-reviewed data published in one of the most selective journals in the energy field, and they were validated against the specific brine chemistry of a resource that the domestic lithium industry considers strategically essential.
The thermally switchable solvent concept has genuine advantages that are not merely incremental improvements on existing approaches. The absence of specialized binding chemicals, the potential to run on waste heat, the inherent magnesium exclusion, and the generalizability of the mechanism across different amine structures all suggest a technology with degrees of freedom that sorbent and membrane systems lack. Whether those advantages survive engineering scale-up is the question that laboratory results cannot answer.
For industry professionals and policymakers watching the build-out of domestic lithium supply infrastructure, S3E is best understood as a technology to track at the pilot stage rather than one to underwrite at commercial scale today. The current moment in critical minerals supply chain development, which has seen a wave of operational commissioning events and federal investment decisions over recent months, is generating urgency that could accelerate the transition from published result to funded demonstration if the right development partners engage. The chemistry published in Joule gives them a solid starting point.