If you collect one ton of discarded smartphone and server circuit boards, they contain roughly 1 kilogram of gold. By comparison, the ore grade of gold currently being mined is only a few grams per ton. That's an "urban mine" roughly 200 times more concentrated than natural ore, being thrown away around the world every year. In 2022, global e-waste reached 62 million tons, containing an estimated $91 billion (roughly ¥13.7 trillion) worth of metals—of which gold alone accounts for an estimated $15 billion. Yet only 22.3% of this is formally collected and recycled in an environmentally sound manner.
Technology to recover this enormous resource already exists. It's called solvent extraction (liquid-liquid extraction, SX). This technique, which selectively transfers metal ions dissolved in water into an immiscible organic solvent, was first used at industrial scale in the 1940s for uranium isotope separation in the Manhattan Project. Since its introduction for copper recovery at the Bluebird mine in Arizona in 1968, it has operated as the standard hydrometallurgical process at refineries worldwide. Copper, uranium, rare earths—this technique underpins the very foundation of the modern metal supply chain.
The "Round-Trip Correspondence" of Acids and Bases Is Holding Back Industry
The basic principle of solvent extraction is simple. An extractant in the organic phase captures metal ions dissolved in the aqueous phase, then transfers and releases them into a separate aqueous phase. The problem is that running this "capture and release" cycle requires large amounts of chemical reagents. When transferring metal into the organic phase, acids or bases are used to adjust the pH; when releasing the metal to regenerate the extractant, yet another acid or base—and sometimes a reducing agent—must be introduced. This back-and-forth of acids and bases generates most of the process's cost and waste liquid.
At copper SX plants, extractant consumption reaches 1.7 to 8.7 kilograms per ton of cathode copper. In uranium acid leaching, 10 to over 100 kilograms of sulfuric acid are consumed per ton of ore. These reagents become waste liquid after use, requiring neutralization and precipitation treatment before disposal. The environmental burden associated with manufacturing, transporting, and disposing of these reagents has relatively undermined hydrometallurgy's reputation as "more environmentally friendly than pyrometallurgy."
Electrochemical approaches have been proposed to address this structural problem. By changing electrode potential, one can control redox states through electron transfer instead of reagents—in principle, requiring neither acid nor base. However, conventional electrochemical separation has mainly relied on "electro-swing adsorption," which adsorbs metals onto solid electrode surfaces, forcing operation into intermittent batch mode. Continuous operation has been difficult to achieve.
The Final Challenge Left by "e-LLE" in 2024
The research team led by Professor Xiao Su (Department of Chemical and Biomolecular Engineering, University of Illinois Urbana-Champaign) broke through one part of this wall in 2024. In a paper published in Nature Chemical Engineering, they incorporated a redox-active extractant into the solvent extraction framework, reporting an electrochemically driven continuous-flow separation platform called "e-LLE" (electrochemically mediated liquid-liquid extraction).
This 2024 system demonstrated over 90% atom efficiency, selectivity exceeding 100:1 for practical leachates, and 16-fold enrichment of gold and platinum group metals. Economic analysis showed costs two orders of magnitude lower compared to conventional industrial processes. It enabled continuous operation and dramatically reduced external chemical inputs.
However, full electrification had not yet been achieved. Some chemical reagents were still needed to close the extraction cycle. Even though electricity replaced most of the acids and bases, reagent input remained in the final step. Whether this "last drop" could be eliminated became the next question.
Packing Three Functions into One Molecule
In July 2026, the Su research team presented their answer to this question in a paper published in ACS Energy Letters (DOI: 10.1021/acsenergylett.6c01434, selected for the supplementary cover). Postdoctoral researcher Deborah Schmitt and PhD student Aderiyike Aguda are listed as co-authors.
The molecule they designed integrates three functions into a single structure.
First, it contains ferrocene as a redox center. Ferrocene has a sandwich structure with an iron atom flanked by two cyclopentadienyl rings, exhibiting reversible one-electron redox behavior. This change in oxidation state acts as the switch controlling binding and release of metal ions.
Second, it carries a permanent charge within the molecule via a quaternary ammonium group. This positive charge ensures ionic conductivity in the organic solvent, allowing the electrochemical reaction to proceed without adding an external supporting electrolyte. As Schmitt explains: "The new molecule has a permanent charge built in, which acts like an electrolyte to carry current through the liquid. That's what allows the redox reaction to be driven by electricity instead of chemicals."
Third, an alkyl chain adjusts hydrophobicity to ensure solubility in the organic phase. The extractant needs to move back and forth between the aqueous and organic phases, but if it doesn't remain in the organic phase, extraction efficiency drops. By varying the length of the alkyl chain, this balance can be engineered at the molecular level.
This integration of three functions eliminates the need for both external redox reagents and supporting electrolytes. The molecule itself carries electricity, captures the metal, transports it into the organic phase, and releases it via electricity. As Professor Su states: "This is the first time we've been able to run the electrochemical solvent extraction we've been dreaming of. You charge the molecule, it binds to the metal, carries it into the organic phase, and then electricity releases the metal again."
Selective 89% Gold Recovery from E-Waste Leachate
The research team conducted a demonstration experiment using this molecule to recover gold from e-waste leachate (a solution made by dissolving waste circuit boards, etc.). The results showed 89% gold recovery with a separation factor of 35. The separation factor indicates the ratio of extraction efficiency between gold and other metals (such as copper and nickel); a value of 35 means gold can be preferentially extracted even in an environment where other metals coexist.
According to the research team's report, chemical reagent consumption was reduced by one to two orders of magnitude (10 to 100-fold) compared to conventional methods (the absolute consumption values for the compared method and the new method have not been disclosed in currently published materials). Fewer reagents mean less waste liquid and lower costs for neutralization and disposal. The techno-economic analysis accompanying the paper showed significant reductions in both capital expenditure (CAPEX) and operating expenditure (OPEX).
| Metric | 2024 e-LLE (Nature Chemical Engineering) | 2026 New Molecule (ACS Energy Letters) |
|---|---|---|
| External chemical reagents | Partially required (used to close cycle) | Not required (closed via built-in molecular charge) |
| Supporting electrolyte | Required | Not required |
| Gold recovery rate | High (demonstrated with continuous flow) | 89% (e-waste leachate) |
| Separation factor | Over 100:1 (selectivity metric) | 35 (Au vs. coexisting metals) |
| Atom efficiency | Over 90% | Not disclosed |
| Enrichment factor | 16x (Au, PGM) | Not disclosed |
| Operating mode | Continuous flow | Continuous flow (in principle, same platform) |
| Chemical reagent reduction | Substantial reduction (vs. conventional) | 1–2 orders of magnitude vs. conventional (absolute values not disclosed) |
One caveat: the "selectivity" metrics reported in the two papers differ. The "over 100:1" figure in the 2024 paper is reported as selectivity against practical leachates (a ratio between specific metal pairs), while the "separation factor of 35" in the 2026 paper is defined as the ratio of extraction efficiency between gold and coexisting metals. Since the measurement conditions and definitions differ, a direct numerical comparison isn't possible.
The "Universality" of the Molecular Design Is the Real Contribution
The significance of this research doesn't lie merely in recovering gold using electricity. As Professor Su states, "What this research reveals is the underlying principle—that is, how to think about it"—the real contribution of this work lies in the molecular design framework itself.
Ferrocene's redox potential, the position and number of ammonium groups, the length of the alkyl chain—by varying these, one can design extractants tailored to the target metal. The electrochemical platform (electrodes, cell structure, flow system) can remain unchanged; only the molecule needs to be swapped out.
Co-author Aguda explains: "This system can be adapted for selective recovery of platinum group metals from waste automotive catalysts, or diverse critical elements from mine tailings and other complex sources. The electrochemical platform stays largely the same—you just need to redesign the extractant chemistry for the application."
This extensibility relates directly to current problems in critical mineral supply chains. Platinum group metals are essential for catalytic converters and fuel cells, while rare earths are essential for permanent magnets and batteries—yet their supply is geographically concentrated, and recovery processes have depended on environmentally burdensome chemical reagents. If extractants can be redesigned at the molecular level, there's potential to build electrified processes optimized for each target metal.
Three Remaining Questions
This research has been published as a peer-reviewed paper and received high enough acclaim to be selected for a supplementary cover. However, unverified challenges remain on the path to practical application.
First, there is long-term stability. Ferrocene is a relatively redox-stable molecule, but how much molecular decomposition or loss into the organic phase would occur over thousands to tens of thousands of cycles of continuous operation in an industrial plant remains unverified. The atom efficiency figure of over 90% reported in the 2024 paper is also based on short-term operation at laboratory scale.
Second, there is expansion to other target metals. Only gold recovery was demonstrated in the paper. Application to platinum group metals and rare earths is described as "possible in principle based on the design principle," but data on selectivity and recovery rates in actual leachates has not yet been produced.
Third, there is scale-up. Between laboratory cells and industrial plants lies a gap that must be closed in fluid dynamics, electrode surface area, power supply, and economics—all at once. The research team is working on accelerating molecular exploration using computational modeling and AI, but this remains at the planning stage.
Professor Su states: "It's a step toward showing that electrochemistry can actually provide scalable, low-waste separation. Amid growing interest in critical minerals and supply chains, it's also a step toward rethinking how we recover metals in a way that's cleaner and fully electrified."
The grade of urban ore is hundreds of times higher than natural ore. The technology to mine it already exists. The question is whether we keep piling up mountains of acid and base every time we mine it, or whether we can get by with nothing but a flow of electrons. Whether this single molecule becomes the starting point for changing that choice depends on the success of the next scale-up.