A Nanotube Thinner Than a Virus Just Moved Lithium 31 Times Faster Than Physics Predicted

Thirty-one. That is the multiplier nobody expected.

When researchers at the University of Illinois Chicago and Rutgers University placed boron nitride nanotube membranes between two salt solutions of different concentrations, lithium ions moved through the tubes 31 times faster than standard diffusion theory predicted. Not 31 percent faster. Thirty-one times. And the lithium showed sharp selectivity, racing through the charged nanotube channels while competing ions like sodium and potassium lagged behind, as though the tubes had built a molecular express lane with a velvet rope.

The study, published June 20 in Nature Nanotechnology, describes membranes made from millions of aligned boron nitride nanotubes, each roughly five nanometers in diameter (thinner than a DNA helix), arrayed under a 40-gauss magnetic field. Boron nitride is a synthetic crystalline compound whose surfaces carry a natural charge, and when assembled into dense nanotube forests, that charge does something nobody fully understands yet: it accelerates lithium-ion transport far beyond what the Nernst-Planck equations say should be possible.

“This is a unique mechanism that transports lithium ions very quickly through nanotubes,” said Sangil Kim, associate professor of chemical engineering at UIC and a corresponding author. He compared the mechanism to electric eel electrocytes, the specialized cells that convert ion gradients into electrical bursts powerful enough to stun prey. Except here, the “eel” is a synthetic membrane smaller than a postage stamp.

What 15,300 W/m² Actually Means

To demonstrate the membrane’s potential, the researchers stacked eight one-square-centimeter membranes in series, placed them between ionic solutions, and powered a watch, a calculator, and a set of LEDs using only the energy released when concentrated salt water meets dilute. Per-pore power densities reached 15,300 watts per square meter, with energy-conversion efficiency approaching the theoretical maximum of 50 percent at pH 5.5.

Context matters here. Commercial membranes used in reverse electrodialysis, the dominant technology for harvesting so-called “blue energy” from salinity gradients, produce about 5 watts per square meter of membrane area. But 15,300 W/m² is a per-pore measurement, taken at individual nanotube openings under controlled lab conditions, and pore-level performance does not survive intact when you scale to a full membrane sheet with packing losses, boundary effects, and structural support overhead eating into the numbers.

Even so, a conservative translation reveals why this result attracted attention beyond the materials science community. Assume only 1 percent of per-pore performance survives at membrane scale. That yields 153 W/m², roughly 30 times the commercial benchmark.

Run the numbers on a hypothetical 5-megawatt blue energy plant. With today’s membranes at 5 W/m², you need one million square meters of membrane surface, which at $50 to $100 per square meter for commercial ion-exchange membranes means $50 million to $100 million in membrane costs alone, before you pour a single foundation or connect a single wire. At 30 times the power density, that same plant needs roughly 33,000 square meters of membrane. Cost: $1.65 million to $3.3 million. A 97 percent reduction in the single largest capital expenditure for blue energy infrastructure.

Blue energy is not fringe speculation. Peer-reviewed estimates place the global salinity gradient resource between 1.4 and 2.6 terawatts, equivalent to approximately 2,000 nuclear reactors running continuously, enough to supply 40 to 80 percent of current global electricity demand. Unlike solar and wind, it generates power around the clock wherever rivers meet the ocean. Nobody builds it at scale because membrane performance makes the economics prohibitive. This paper does not solve the scaling problem. It moves the theoretical ceiling by an order of magnitude.

Lithium Recovery: Where the Money Is Closer

Lithium recovery from waste batteries and brine presents a nearer-term commercial pathway, and the numbers here are equally striking. Global lithium recycling reached $3.8 billion in 2025, projected to hit $12.4 billion by 2033, driven by a supply deficit that analysts expect between 2028 and 2029. Conventional direct lithium extraction from brine carries operating costs of $4,600 to $7,850 per ton of lithium carbonate equivalent, against a spot price hovering around $8,000 to $10,000 per ton. Margins are thin.

If BNNT membranes deliver even a fraction of their 31× throughput advantage at production scale, those margins widen considerably. A standard DLE membrane module processes 10 to 50 cubic meters of brine per hour; applying the 31× transport multiplier to identical membrane area yields 310 to 1,550 cubic meters per hour per module. At typical Smackover Formation brine concentrations of roughly 200 parts per million lithium, that translates to 62 to 310 kilograms of lithium per hour per module. Spread capital amortization across 31 times the throughput and the effective extraction cost drops toward $2,000 to $3,000 per ton LCE, well below current spot prices and below every competing DLE technology’s published operating cost.

Strongest Counterargument

The strongest case against reading too much into this result is that the gap between a pore-level measurement and a deployable industrial membrane has swallowed most promising materials breakthroughs of the past two decades, and boron nitride nanotubes face a particularly brutal version of that gap. Per-pore power density was measured at individual nanotube openings using freshly prepared membranes totaling eight square centimeters of area. Commercial deployment would require membrane sheets measured in thousands of square meters, manufactured consistently, surviving months or years of continuous operation in feedstock far dirtier than lab-grade salt solutions. No durability data appears in the paper. No fouling resistance tests. No cost projections for membrane fabrication at scale.

BNNT production cost is the other constraint that does not yield to clever engineering alone. Commercial boron nitride nanotubes currently cost more than $100 per gram from suppliers like BNNT LLC and NAiEEL Technology. Building membrane-scale quantities requires cost reductions of at least two orders of magnitude, and no public roadmap from any manufacturer addresses that trajectory.

Limitations

Our calculations above apply scaling factors that are illustrative, not experimentally validated. The 1 percent pore-to-membrane translation factor is an assumption without empirical basis in this paper; actual membrane-area power density could be higher or lower depending on packing density and fabrication methods that do not yet exist at scale. The 31× throughput multiplier applied to DLE modules assumes the transport advantage measured in clean ionic solutions survives contact with complex brine chemistry containing calcium, magnesium, and organic contaminants that could foul or block BNNT channels. The paper does not report quantified Li+/Na+ or Li+/K+ selectivity ratios, so precise separation efficiency calculations are impossible from public data. Blue energy LCOE projections use membrane cost figures ($50–$100/m²) from commercial RED systems using entirely different membrane chemistries; BNNT membranes will almost certainly cost more, at least initially. No independent replication of the 31× result has been published.

What You Can Do

If you work in membrane science or electrochemistry, the paper’s DOI is 10.1038/s41565-026-02182-5; the anomalous transport mechanism begs for independent replication using different nanotube diameters and electrolyte compositions. Investors in lithium mining or battery recycling should watch for follow-up publications from Kim’s UIC lab addressing durability and selectivity under real brine conditions, because those experiments will determine whether this has a commercial path or stays in the lab. Battery recyclers evaluating membrane-based extraction technology should track BNNT production costs; if prices fall below $1 per gram, roughly a 100× reduction from today, membranes built with this architecture become economically plausible for industrial use. For the rest of us: this is a fundamental science result, not a product launch. No company sells BNNT lithium membranes or BNNT blue energy systems today. The honest timeline, if everything works, is measured in decades.

The Bottom Line

Researchers at UIC, Rutgers, and Argonne discovered that lithium ions move through boron nitride nanotube membranes 31 times faster than diffusion theory says they should. The membranes are eight square centimeters of lab material, production costs exceed $100 per gram, and nobody has tested whether they survive a week of continuous operation. But the numbers hint at something worth tracking: a membrane architecture that, if even modestly scalable, could cut blue energy infrastructure costs by 97 percent and compress lithium extraction economics below any competing technology. The operative word is “if.” Between a demonstration that powered a calculator and an industrial system that processes brine at 1,550 cubic meters per hour sits the widest gap in materials science, the gap between a measurement and a product. This paper, published in Nature Nanotechnology on June 20, earned the right to be watched carefully. Believed? Not yet.