Hydrogen is increasingly seen as a fuel of the future because its use produces no exhaust gases, only water. Currently, however, most hydrogen is produced from fossil fuels, primarily natural gas, generating carbon dioxide emissions.

In a paper published in the Journal of the American Chemical Society (https://dx.doi.org/10.1021/jacs.5c11004), the researchers describe a route toward artificial photosynthesis: producing hydrogen directly from water using light, without chemical additives—so-called sacrificial reagents—that many existing methods require to improve reaction efficiency.

“These compounds must be purchased, dosed, and then dealt with as a byproduct,” the researchers note. “This complicates the installation and reduces profitability on an industrial scale. Moreover, because these reagents are oxidized during the process, the reaction is no longer pure water splitting. In practice, a stream of byproducts, often oxidized derivatives of organic compounds, and in extreme cases CO2, can be produced.”

Photocatalysis uses a light-activated material, called a catalyst, to provide reacting molecules with an alternative, less energy-intensive route to a reaction. The catalyst is not consumed and can be reactivated after the reaction.

One of the main challenges in this type of photocatalysis is that light excites electrons, leaving holes behind—spaces where electrons are missing. If the electrons and holes quickly recombine, the energy is lost as heat and no useful reaction occurs. Effective catalysts must separate these charges.

Seawater presents an additional challenge. It is abundant, but salts and impurities can block the catalyst surface and disrupt its structure, causing many materials to perform poorly or degrade quickly.

The scientists focused on graphitic carbon nitride (g-C3N4), a carbon and nitrogen compound with a graphitic structure that can be imagined as a very thin, ordered “sheet” of atoms. The material is stable and capable of absorbing light.

The researchers added cyano groups (–C≡N) embedded in the sheet structure near the aromatic ring. These fragments extend the material’s conjugated bond system, improving light absorption and charge separation—akin to providing many small pathways for electrons.

The most important ingredient is nickel, used not as ordinary nanoparticles but as single atoms scattered across the catalyst surface. This arrangement, called a single-atom catalyst, ensures that each metal atom is utilized to its full potential while allowing precise control of its action.

The team confirmed the “atomic solitude” of nickel using X-ray techniques, including X-ray absorption spectroscopy (XAS), which examines the local environment of the atoms.

Splitting water into hydrogen and oxygen is difficult because the step producing oxygen requires multiple electrons and the formation of an O–O bond in a controlled manner. The researchers bypassed this challenge by dividing the reaction into more manageable steps. In the first step, hydrogen and hydrogen peroxide (H2O2) are simultaneously produced.

The H2O2 then spontaneously converts into oxygen and water. The researchers observed that H2O2 initially appears and then decreases, while hydrogen continues to be released, marking the trace of this two-step mechanism.

The best-performing material achieved a hydrogen release rate of up to 270 micromoles per gram per hour (μmol/g/h) in pure water under violet light (390 nm). Importantly, it also worked with seawater, reaching 144 μmol/g/h in North Sea water. Under direct sunlight, efficiency was lower but still measurable.

The catalyst demonstrated excellent stability in long-term tests totaling over 720 hours, including 140 hours of illumination and 580 hours in the dark.

“The point of this research is not simply the number of molecules produced,” the scientists say. “If hydrogen is to truly alleviate the climate burden, it must be produced without chemical ‘helpers’ that require purchase and disposal, and without using precious freshwater where seawater is available.” (PAP)

kmp/ zan/

tr. RL