Rolled-Up Oxide Nanosheets Crack the Out-of-Plane Proton Transport Problem

Fuel cells live or die by a simple geometric fact: the protons that carry current must cross the electrolyte membrane perpendicular to its surface. Yet in most two-dimensional materials, ion transport is spectacularly lopsided. Movement along the plane of a nanosheet can be fast, while movement across a stack of sheets can lag behind by three to five orders of magnitude, throttled by the barriers between layers. A team of Japanese researchers has now shown that this stubborn anisotropy can be tamed not by changing a material’s chemistry, but by changing its shape. Their report, published in Advanced Science, describes an HTaWO6 nanosheet membrane that spontaneously curls into nanotube-like structures and, in doing so, delivers out-of-plane proton conductivity of 1.7 millisiemens per centimeter at 80 degrees Celsius, one of the highest values ever recorded for an oxide nanosheet membrane.

The work began from a familiar frustration in electrolyte research. Proton-conducting polymers such as Nafion dominate commercial fuel cells, but they come with limits, including modest thermal stability and a hydrated architecture that is difficult to engineer at the atomic level. Metal-organic frameworks, covalent organic frameworks, graphene oxide and polyoxometalates have all been explored as alternatives, and their pore networks can indeed move protons quickly. The catch is that large, continuous pores that help protons can also let hydrogen gas sneak across the electrolyte, a phenomenon known as crossover that wastes fuel and degrades efficiency. Oxide ceramics, by contrast, are thermally robust and chemically tunable, but they are usually studied as dense, micrometer-scale bulk materials in which protons diffuse sluggishly through the grain interior, often trapped at defect sites, with correspondingly high activation energies.

The researchers’ strategy was to take an oxide and flatten it. They synthesized a family of layered compounds with the formula LiMM’O6, where M is tantalum or niobium and M’ is tungsten or molybdenum, built from corner-sharing TaO6 or NbO6 octahedra interleaved with lithium layers. Acid treatment swapped the lithium for protons, and an aqueous solution of tetrabutylammonium hydroxide then pried the charged layers apart into individual negatively charged nanosheets. Vacuum filtration of the resulting colloidal dispersion, confirmed by the telltale Tyndall light-scattering effect, produced free-standing, semi-transparent membranes in which the nanosheets are strongly oriented with their crystallographic c-axes perpendicular to the film. A final acid immersion replaced the bulky organic cations with protons, yielding the target HMM’O6 membranes. X-ray diffraction showed that the interlayer spacing in the membrane form of HTaWO6 expanded to 1.27 nanometers, substantially larger than the 0.92 nanometers of the parent powder, and thermogravimetric analysis revealed that the exfoliated membrane holds far more surface water than the bulk, a hint that hydration, and with it proton mobility, had been enhanced.

Impedance spectroscopy from 80 down to 40 degrees Celsius under fully humidified conditions told a striking story. Among the four compositions tested, the HTaWO6 membrane was the clear winner for cross-plane transport, reaching 1.7 millisiemens per centimeter at 80 degrees with an activation energy of just 0.224 electronvolts, small enough to indicate weak temperature dependence, a marked contrast to bulk oxide ceramics. When the team measured transport along the membrane plane, the picture inverted: the molybdenum-containing HTaMoO6 led with an in-plane conductivity of 0.20 siemens per centimeter at 80 degrees, actually surpassing commercial Nafion over much of the measured temperature range. Crucially, the ratio of in-plane to out-of-plane conductivity, the anisotropy, was only 28 for HTaWO6, roughly one order of magnitude, whereas HTaMoO6 and HNbMoO6 showed anisotropies of about 290-fold and 300-fold, squarely in the range typical of stacked nanosheet materials.

Before celebrating, the authors verified what was actually doing the conducting. Conductivity fell steeply as humidity dropped from 100 to 40 percent, consistent with a water-mediated proton mechanism. Switching the atmosphere from H2O to D2O raised the membrane’s resistance by a factor of 1.56 at 80 degrees, the classic H/D isotope effect that fingerprints proton transport. A water-vapor concentration cell produced an electromotive force close to the theoretical Nernst value, confirming that protons, not hydroxide ions, dominate the current. Together these tests established that the impressive cross-plane conduction in HTaWO6 is genuine proton conduction.

How does it compare with the competition? The 1.7 millisiemens per centimeter out-of-plane figure is 36 to 170 times higher than calcium niobate-based nanosheet membranes such as HCa2Nb3O10 and HCa2Nb2.5Ti0.5O10, and 124 to 381 times higher than strontium tantalate-based nanosheets. Part of the advantage comes from a low activation energy, comparable to the best strontium tantalates and far below the 0.38 to 0.65 electronvolts of the calcium niobates, and part comes from an unusually high pre-exponential factor in the Arrhenius relationship, which the authors attribute to the abundance of accessible conduction pathways in the curled morphology.

The morphology itself turned out to be the real surprise. Atomic force microscopy of the TBA/TaWO6 nanosheets revealed rod-like objects with heights of about 26 nanometers, far too tall for a single inorganic monolayer. High-angle annular dark-field scanning transmission electron microscopy resolved the mystery: the rods are nanotubes, formed when the flat nanosheets roll themselves up, with 63 percent of the observed sheets curled. The rolled cross-sections, about 27 nanometers tall and 39 nanometers wide, resemble flattened ellipses rather than perfect circles, which neatly explains why the anisotropy does not vanish entirely but merely shrinks to one order of magnitude. The team proposes that the tubular geometry shortens the effective diffusion path through the membrane thickness and provides continuous one-dimensional channels that bypass the interlayer barriers that cripple flat-sheet stacks. Composition matters, too: mixing tungsten and molybdenum in TaMo0.4W0.6O6 produced a mixture of flat and partially rolled sheets, showing that chemical composition governs the tendency to roll, even though the microscopic origin of the rolling remains unresolved. Meanwhile, the HNbWO6 sheets, which stayed flat but grew nanopores tens of nanometers across, also conducted well out-of-plane, echoing earlier observations in graphene oxide that nanopores offer extra cross-plane routes.

Fuel-cell tests translated the conductivity numbers into device performance. A 37-micrometer HTaWO6 membrane delivered an open-circuit voltage of 1.0 volt and, at 80 degrees, a maximum current density of 191 milliamperes per square centimeter and a power density of 41 milliwatts per square centimeter, several times better than the other three compositions. Thinning the membrane changed everything: at 18 micrometers the power density rose to 67 milliwatts per square centimeter, and at just 1 micrometer the cell reached 1.26 amperes per square centimeter and 294 milliwatts per square centimeter, with the open-circuit voltage holding at 1.00 volt, evidence of low gas crossover and good electronic insulation. Those figures exceed previously reported oxide nanosheet electrolytes by factors of dozens to hundreds. Post-mortem characterization showed the membrane’s crystal structure intact after testing, and the material resisted acid dissolution. A further boost came from treating the membrane with methanesulfonic acid, which widened the interlayer spacing, grafted sulfonic acid groups onto the sheets, improved water retention, and lowered interfacial resistance at the electrode. The membrane even operated at 100 degrees, albeit with reduced output as hydration fell away.

The broader lesson is that in two-dimensional ion conductors, geometry can matter as much as chemistry. By introducing curvature into planar nanosheets, the researchers reconciled the short diffusion lengths of nanoscale thickness with efficient cross-plane transport, a combination long thought to be out of reach for layered oxides. They argue the concept extends well beyond nanotubes to nanowires, nanospirals and nanofiber assemblies, and beyond protons to lithium, oxide-ion and hydride-ion conductors, any system where ions must move perpendicular to a stacked architecture. If morphology engineering proves general, the humble act of rolling a sheet could become a standard tool for designing the thin, stable, fast electrolytes that next-generation electrochemical devices demand.

Subject of Research: Morphology-engineered oxide nanosheet membranes for out-of-plane proton conduction in fuel-cell electrolytes

Article Title: Overcoming Out‐of‐Plane Conduction by Changing Nanosheet Morphology

Article References: Sakuda, Y., Hatakeyama, K., Yuji, A., Tsugawa, T., Ueno, K., Tomatsu, S., Utami, Z. D., & Ida, S. (2026). Overcoming Out‐of‐Plane Conduction by Changing Nanosheet Morphology. Advanced Science, Article e78017. https://doi.org/10.1002/advs.78017

Image Credits: AI Generated

DOI: 10.1002/advs.78017

Keywords: proton conduction, nanosheets, nanotubes, HTaWO6, fuel cells, oxide electrolytes, conductivity anisotropy, two-dimensional materials, exfoliation, impedance spectroscopy, morphology engineering, Advanced Science