Oregon State’s Zinc MOF Makes Hydrogen From Sunlight Alone

Oregon State chemists built a zinc-and-cystine crystal that makes hydrogen from water in sunlight, with no extra metal catalyst and no grid power.

The university posted the work on Sept. 28, 2026. The crystal is a photocatalyst for hydrogen gas, a chemical that industry still buys mostly from methane.

Oregon State’s Zinc Framework Splits Water in Sunlight

Kyriakos Stylianou, professor of chemistry at Oregon State University who directs the Materials Discovery Laboratory, led a team that reported a new family of materials that use light to produce hydrogen from water. The lab calls the series BVR-19-M. The zinc member, BVR-19-Zn, is the one that works best.

A photocatalyst absorbs light, rises in energy, and spends that energy on a chemical reaction without being used up. Here the reaction is water giving up hydrogen, the gas already used in ammonia plants, metal refining, plastics, and fuel cells.

The findings ran in the Journal of the American Chemical Society. The paper was received on June 27, 2026, accepted on Sept. 1, published online on Sept. 14, and placed in the Sept. 23 issue, volume 148, starting at page 40255. Emmanuel Nyela Musa, a MaD Lab researcher, is the first author.

The powder forms in water at room temperature, and it forms on its own. Stylianou said that route carries a strong energy advantage over frameworks that need heat, solvents, or extra steps. No second, costly metal is added to finish the hydrogen step.

The Murdock Charitable Trust, the National Science Foundation, and the OSU College of Science funded the study. Stylianou’s group has published other light-driven hydrogen materials, including a 2024 oxide mix that leaned on ruthenium. This paper turns that habit around. The organic linker is supposed to do the photochemistry, and the metal is supposed to stay out of the way.

The Disulfide Bond Does the Work the Metal Usually Does

Most photocatalyst recipes treat a metal atom, often a precious one, as the place where light becomes chemistry. BVR-19 is built from divalent metal ions and L-cystine, an amino-acid linker that already carries a sulfur-to-sulfur bond. That disulfide is the feature the Corvallis team went after.

The organic component does the important work. Instead of relying primarily on the metal atoms, our material uses its sulfur-containing organic building blocks to capture light energy and move electrons where they are needed to produce hydrogen. This represents a different way of thinking about how these materials should be designed.

Kyriakos Stylianou, professor of chemistry, Oregon State University

The JACS paper describes intraligand charge transfer in zinc BVR-19 as the path that makes that claim concrete. Under light, the disulfide bridges in BVR-19-Zn break for a moment by homolytic cleavage. The split produces a thiyl radical and a thiolate anion, a pair that separates charge and drives hydrogen evolution with no cocatalyst on the side.

Zinc’s Filled Orbitals Leave the Linker in Charge

The team kept the cystine linker and swapped the metal. Magnesium, manganese, copper, zinc, and cadmium all went into the same BVR-19-M scaffold. Zinc won.

Zinc’s d-orbitals are filled, so the metal does not offer an easy place for the light-generated charge to sit. The disulfide ligand can run the redox chemistry itself. That is the design bet: pick a quiet metal, and let an organic S-S bond do the job platinum usually gets.

Open-Shell Metals Bleed off the Light Energy

Open-shell ions in the same family hurt the result. The paper says they either open a mid-gap state that acts as a recombination center, or they pull the charge onto the metal. Either way, less of the absorbed light ends up as hydrogen.

THE FIVE METALS IN BVR-19-M

• Zinc: Highest hydrogen activity, filled d-orbitals, and the linker-centered charge path the authors want.
• Magnesium: Tested as a divalent ion in the same cystine framework, with zinc still the clear lead.
• Manganese: An open-shell ion, in the group the paper ties to mid-gap recombination.
• Copper: An open-shell ion, in the group tied to metal-centered charge transfer that cuts output.
• Cadmium: Tested as another divalent ion in the BVR-19-M series, behind zinc.

L-cystine is not an exotic ligand. It is two cysteine molecules joined by a disulfide, the same motif enzymes already use for redox. The MaD Lab’s move is to park that motif inside a porous crystal and shine light on it until hydrogen comes off.

Green Hydrogen Still Costs About $5 a Kilogram

Stylianou put the commercial problem in dollar terms. Methane-steam reforming, the usual way to pull hydrogen out of natural gas, makes a kilogram for about $1.50. The green route he compared, hydrogen from water with renewable power, runs about $5 a kilogram.

Those figures are his, from the Oregon State release. They describe current practice, not a price the new powder has posted. BVR-19-Zn has no published plant cost. It is a laboratory solid with a new charge path.

Water splitting that uses a catalyst and electricity, electrocatalysis, is only as clean as the current that feeds it. Stylianou noted that the method needs cheap renewable power before it can compete. The International Renewable Energy Agency makes the same point in plainer accounting: the electricity price is the largest cost driver for renewable hydrogen, and the electrolyzer’s capital cost is the next hurdle.

A photocatalyst tries to skip the electrolyzer bill by putting the light absorber and the water-splitting chemistry in one powder. That is the hope. The price floor it still has to beat is the methane kilogram, not a rooftop watt.

WHAT A KILOGRAM OF HYDROGEN COSTS

The International Energy Agency’s Global Hydrogen Review 2026 puts the maximum acceptable hydrogen cost below $2 a kilogram for most sector and region pairs if there is no policy support. Steel is harsher. In that analysis the hydrogen cost would have to be negative, meaning a subsidy, to match the usual iron route. A $5 green kilogram and a sub-$2 buyer cap do not overlap.

In the near term, the IEA says fossil-based hydrogen stays cheaper than renewable hydrogen in most of the world. China is the named exception, where renewable hydrogen could become competitive by 2030. Everywhere else, the agency says, low-emissions hydrogen still carries a premium, and support policy remains part of the sale.

Does This Photocatalyst Replace Solar Panels?

Solar panels sell electricity to a grid. BVR-19-Zn is a laboratory powder meant to make hydrogen gas for chemicals and fuel cells, a market still priced by methane. A home that wants kilowatt-hours does not buy a cystine crystal, and an ammonia plant that wants hydrogen does not buy a rooftop array and call the job done.

The Debrief-style claim that this is a new way to convert sunlight into energy, as if panels were the thing being retired, mixes those products. Oregon State’s own notice was more careful. It described a photocatalyst for hydrogen used in cars’ fuel cells and in making ammonia, metals, and plastics. It compared the chemistry with electrocatalysis and with methane-steam reforming. It did not present a module, an inverter, or a rooftop kit.

Light-driven water splitting as a field has spent a decade on small panels and outdoor pilots, while electrolysis has gone into factories. The IEA counted installed electrolyzer capacity that doubled in 2025 to exceed 4 GW, with China behind nearly three-quarters of the new units, and more than 2.5 GW already under construction for 2026. That is the industrial sunlight-to-hydrogen path that already has a supply chain: cheap power in, hydrogen out, through a box you can order in megawatts.

A room-temperature MOF does not erase that stack. If BVR-19-Zn ever leaves the bench, it would compete with those boxes on cost per kilogram, uptime, and how you collect a mixed gas over a slurry. It would not compete with a silicon panel on the price of household current.

A Design Rule for Half a Million Unmade Frameworks

Stylianou likes to remind people that metal-organic frameworks are a combinatorial sport. Positively charged metal ions and organic linker molecules can be combined in millions of ways. Almost 100,000 MOFs have been made. The properties of another half a million have been predicted. Most of that library has been scored by pore size, metal identity, and gas uptake, not by whether a linker can throw a radical when the sun hits it.

DESIGN RULES FROM THE MAD LAB

• Linker first: Put light absorption and the redox step on the organic disulfide, not on a sprinkled platinum particle.
• Quiet metal: Zinc’s filled d-orbitals keep the charge on the cystine instead of trapping it.
• No extra catalyst: Hydrogen comes off without a second expensive metal in the mix.
• Cold water synthesis: The crystal forms in aqueous solution at room temperature, on its own.

Those rules are the part of the paper that can move without a factory. Anyone screening predicted MOFs for solar fuels now has a reason to keep closed-shell metals and redox-active sulfur ligands near the top of the list, and to stop assuming that a better photocatalyst means a rarer metal.

“By changing the metal while keeping the rest of the material essentially the same, we discovered why some versions of the MOF work much better than others,” Stylianou said. “These findings provide new design rules for creating more effective materials for solar fuel production.” He also called the work a blueprint for materials that can bring down the cost of green hydrogen. The blueprint is chemical. The cost drop is still a claim about the future, not a measured plant gate price.

That is also why the finding did not land as an energy-market event. It reads as a mechanism paper with a zinc winner and a sulfur trick, which is exactly what the journal of record posted. The solar-panel headline is what happens when a charge-transfer study is asked to carry a whole fuel industry on its back.

Ammonia, Refineries and a 100-Million-Tonne Market

The buyers Stylianou named are not rooftop customers. They are the plants that already run on hydrogen. The IEA’s 2026 review found that global hydrogen demand surpassed 100 million tonnes in 2025, up almost 3%, with industry and refining taking almost all of it. New uses are growing faster in percentage terms and still make up a thin slice of the total.

HYDROGEN IN 2025, PER THE IEA

• Total demand: Surpassed 100 million tonnes, up almost 3%, almost all in industry and refining.
• Low-emissions output: Grew 20% to almost 1 million tonnes, and is due to pass 1% of supply in 2026.
• Electrolysers: Installed capacity doubled to exceed 4 GW, with China behind nearly three-quarters of new units.
• Fuel-cell vehicles: Stock grew 20% to almost 130,000, a separate count from the 20% rise in low-emissions hydrogen output.

Low-emissions supply is still a rounding error on a methane business. The IEA says 2026 should be the first year it accounts for more than 1% of global production, helped by policy in China and Europe and by supply-chain work in Japan. Offtake agreements for low-emissions hydrogen were about 1.7 million tonnes in 2025, matching 2024, and only about 20% of the new volumes were firm contracts. Demand, the agency says, is still the missing piece.

Projects that have reached a final investment decision point to 2.5 million tonnes of low-emissions hydrogen in refineries and industrial plants by 2030, about 60% of committed production. The broader 2030 pipeline of committed projects and those with a strong chance of running shrank from 10 million tonnes in the 2025 review to just above 6 million tonnes. Announced projects that might still make 2030 sit at 27 million tonnes, with 22 million tonnes at risk if money is not committed by early 2027.

Fuel-cell cars are in that mix, which is why OSU mentioned automotive cells. Almost 130,000 fuel-cell vehicles were on the road after a 20% rise in 2025, driven by trucks in China and a rebound in car sales in Korea. Trucks and buses are projected to take about 60% and 30% of hydrogen use in that vehicle stock by 2030. Even that fleet is a rounding error next to ammonia and refining.

Those plants already know what they will pay. The IEA’s buyer cap below $2 a kilogram, Stylianou’s $1.50 methane kilogram, and his $5 green kilogram describe the same squeeze from three sides. A cystine powder that forms in cold water, needs no platinum speck, and lets zinc sit quiet is a clever answer to the chemistry. It has not yet answered the invoice.

Stylianou said the findings give new design rules for solar fuel materials. In the same reporting season, the IEA put the price most users will pay without policy below $2 a kilogram, and counted a hydrogen market that still runs on methane.