Michigan Nanowire Device Makes Hydrogen From Salty Water in Lab

A University of Michigan-led team has kept a light-driven seawater hydrogen device running for more than 50 hours in the lab, using chromium and platinum on a forest of gallium nitride nanowires. The work, published in Nature Synthesis on 30 October, targets one of the stubborn problems in making hydrogen from the sea: salt water is neutral, and in neutral water the reaction that frees hydrogen tends to be slow.

The photocathode reached a current density of around 100 milliamps per square centimetre under concentrated sunlight, the paper says. These are lab results from small test electrodes in simulated seawater, not from a working plant or open water.

How the seawater hydrogen photocathode works

The device is a photocathode, an electrode that absorbs light and uses the energy to split water. It is built from gallium nitride nanowires grown on a silicon junction, decorated with platinum and with clusters of chromium hydroxide, each just a few atoms across.

The team describes a “triple-interface architecture” in which three neighbouring surfaces share the work. One activates water molecules, the next passes hydrogen along, and the platinum binds it before it is released as gas. The paper calls this “a fast hydrogen pumping process”. Co-authors at Yale, the University of Pennsylvania and Argonne National Laboratory, whose Advanced Photon Source was used, also contributed.

Neutral water matters because seawater is close to neutral. Commercial electrolysers generally run on purified water, and purifying seawater first adds cost and equipment, which is why direct seawater hydrogen has drawn so much research. The approach complements AI-guided searches for cheaper hydrogen catalysts, which have mostly worked in conventional electrolytes.

Simulated seawater, concentrated light and hydrazine

The university’s own data record for the study, held in the University of Michigan Library’s Deep Blue repository, sets out the test conditions behind the headline figures. The salt water was either a sodium chloride solution or simulated seawater made from commercial sea salts at 35 grams per litre. The exposed electrode areas were between 0.05 and 0.15 square centimetres, smaller than a fingernail.

Light came from a solar simulator, and the source data describe measurements under 10-sun illumination, ten times the intensity of ordinary sunlight. During the long-term test, the record says, “Light switching and electrolyte refreshment were adopted … to mitigate heat accumulation.”

The bias-free figure of about 70 mA cm-2 came from a separate set-up in which hydrazine was added to the salt water at 0.1 molar. Hydrazine is oxidised at the other electrode in place of water, which lowers the energy needed, so the system needs a steady supply of a chemical that has to be made elsewhere.

A materials advance, with a long road to the sea

The strongest part of the result is the chemistry. Neutral-pH hydrogen evolution has been a bottleneck, and showing that a carefully placed cluster of chromium hydroxide can speed it up on a light-absorbing surface is a design idea others can test on different materials. The current density is also high for a photoelectrode, although concentrated light makes that easier.

The weaker part is everything a seawater hydrogen system would need around it. Fifty hours is a short run when equipment is expected to last for years, and the team had to switch the light and refresh the electrolyte to manage heat. Real seawater carries microbes, sediment and organic matter that can foul electrodes, which a lab-made sea-salt solution does not. Electrodes a few millimetres across say little about how a panel-sized device would behave, and the bias-free result depends on hydrazine, a consumable reagent rather than free sunlight alone.

There is also a commercial interest to declare. The paper states that some intellectual property from the work “has been licensed to NX Fuels, Inc., which was co-founded by Z.M.”, meaning Zetian Mi, the Michigan professor who led the group, and that the university and Mi “have a financial interest in the company.” That does not undermine a peer-reviewed result, but it means the most likely next announcement will come from a company with a stake in it.

The work fits a broader run of lab materials results TSN has covered, from new rare-earth MXenes to ultra-thin carbon membranes, where the hard part is surviving outside the lab. The heat-stable tandem solar cells reported this week cleared 1,000 hours; this seawater hydrogen device has some way to go to match that.

The research was funded by the US Army Research Office and the National Science Foundation. The test to watch is a run of hundreds of hours in natural seawater, on a larger electrode, without hydrazine.

Sources

• Nature Synthesis: Enhanced neutral seawater photoelectrolysis by fast hydrogen pumping at subnanometre triple interfaces (30 Oct 2026)
• University of Michigan Deep Blue Data: dataset for the study

Related stories

• AI-Guided Lab Finds Hydrogen Catalysts It Calls Comparable to Platinum
• Tandem Solar Cells Keep 90% Efficiency After 1,000 Hours at 65C