Liquid alkaline water electrolysis is the oldest and cheapest way to split water at industrial scale, and it remains the workhorse of green hydrogen. Its engineers face an awkward trade-off, though: almost every step that makes an alkaline electrolyser more efficient also makes it less safe. A team led by Lawrence Berkeley National Laboratory has now published a way out of that trade-off in Nature Energy — and it is elegantly simple.
The trade-off: efficiency at the cost of safety
In an alkaline electrolyser, anode and cathode are kept apart by a porous separator — usually a Zirfon diaphragm around 220 or 500 µm thick. That separator is also the largest single source of ohmic resistance in the cell, so the obvious efficiency lever is to make it thinner. The second lever is a better cathode: nickel is cheap but a mediocre catalyst for the hydrogen evolution reaction, so a thin platinum coating cuts the cell voltage considerably.
Both levers have the same side effect. Dissolved hydrogen migrates through the separator into the oxygen stream — the phenomenon known as hydrogen crossover. Hydrogen becomes flammable in oxygen at around 4%, and ISO 22734 sets a stricter working threshold of 2% for industrial water electrolysis. Exceeding it forces an unplanned shutdown, which in turn causes reverse currents and electrode degradation. In practice, safety has capped how thin the separator may get and how good the cathode may be.
Why platinum cathodes leak more hydrogen
The researchers first explained the mechanism. A platinum-coated nickel cathode reduced the cell voltage by roughly 340 mV at 2 A cm⁻² compared with plain nickel foam — but pushed the hydrogen content in the anode gas up to about 4.7%, above the flammability limit.
The reason is where the hydrogen is made. Nickel is a weak enough catalyst that hydrogen evolution spreads through the whole electrode structure, towards the flow field and away from the separator. Platinum is so active that the reaction concentrates in a thin zone right at the separator interface. Hydrogen is then produced faster than bubbles can form, so the dissolved concentration locally exceeds the solubility limit — supersaturation, measured here directly for the first time using a rotating ring-disk electrode and a platinum wire probe inside the running cell. The result was one to three orders of magnitude above the solubility limit, and that steep concentration gradient is what drives hydrogen across the separator.
The team confirmed the mechanism by moving the platinum around: a platinum layer facing the separator produced dramatically more crossover than the same layer facing the flow field.
The fix: burn the leak before it becomes a hazard
Rather than fight the supersaturation, the researchers dealt with the hydrogen after it arrives. A gas recombination catalyst (GRC) converts stray H₂ and O₂ back into water directly inside the separator. The idea is established in PEM electrolysers, but had not been made to work in alkaline cells: the usual binders have poor anion conductivity and simply add resistance.
The solution avoids binders altogether. The Zirfon diaphragm is masked on one side, soaked in a platinum salt solution and then chemically reduced, growing platinum nanoparticles of about 5 nm directly on the zirconium dioxide (ZrO₂) skeleton — concentrated in the top 8 µm on the anode side, where oxygen is plentiful, and deliberately kept sparse so the particles stay electrically isolated and are not passivated by the anode potential. The platinum loading is around 0.062 mg cm⁻², an order of magnitude less than on the cathode.
What it delivers
The functionalised separator removed more than 95% of the crossing hydrogen at every current density tested. Hydrogen content in the anode gas and hydrogen flux both dropped by more than an order of magnitude. The pore structure, wettability and ionic conductivity of the diaphragm were unchanged, and the cost in hydrogen yield was below 0.1%.
Performance did not suffer — it improved slightly, by about 74 mV, because the platinum on the anode surface is also a decent oxygen evolution catalyst. Durability held for over 1,000 hours at 1 A cm⁻², and a separate 2,000-cycle test simulating start-stop operation showed no loss of recombination efficacy. That last point matters more than it might appear: electrolysers coupled to wind and solar spend their lives ramping up and down.
Why it matters for Power-to-X
Every Power-to-X pathway — eSAF, eMethanol, eAmmonia, eMethane — begins with an electrolyser, and electricity is the dominant cost in every one of them. Alkaline electrolysis is the cheapest technology available at gigawatt scale precisely because it avoids iridium and other scarce platinum-group metals. Anything that lifts its efficiency without adding materials cost feeds straight through into the price of defossilised fuels.
What this work really unlocks is not the 95% figure itself but the design freedom behind it. If crossover is no longer the binding constraint, separators can be made substantially thinner and cathodes substantially more active — both of which were previously blocked by safety limits rather than by materials science. The researchers also note that the principle should transfer to other electrochemical devices facing gas crossover problems.
Two caveats are worth keeping in view. This is a 5 cm² laboratory cell, and scaling a wet-chemical coating process to industrial diaphragms several square metres in size is a genuine engineering task. And the approach adds platinum to a technology whose main selling point is that it does not need any — although the quantity is small, and the efficiency gained plausibly pays for it. Commercial electrode materials for the durability tests were supplied by the electrolyser manufacturer Verdagy, which suggests industry is already paying attention.
Source: Haotian Liu et al., “Mitigation of hydrogen crossover in liquid alkaline water electrolysers using gas recombination catalysts”, Nature Energy, 26 June 2026 (DOI: 10.1038/s41560-026-02094-7, open access); reported by TechXplore, 12 July 2026.
🤖 AI transparency: This text and its visual were created using AI, reviewed and approved by a human.
Power-to-X Congress Switzerland 2026 — «Reality Check with Net Zero». Join us on 22 September 2026, 10–18h, at the Kursaal Bern. Co-organised by energie-cluster.ch and SPIN, with partner Réseau H2 Suisse Romande. Learn more and register →