Technological breakthroughs

A research team at the University of Cambridge has reported outdoor test results for a 1 m² photocatalytic reactor system capable of converting organic waste - including cellulose and PET plastic - into hydrogen gas using sunlight. The study, led by Ariffin Bin Mohamad Annuar and Yongpeng Liu under the supervision of Erwin Reisner, was published in Nature Chemical Engineering.

Why this approach

Global hydrogen demand has nearly doubled over the past two decades, yet most of it is still produced from fossil fuels. Splitting water with sunlight is a long-known alternative, but it's energy-intensive because water oxidation is slow and must overcome a large energy barrier. The team instead used organic waste - such as cellulose from biomass or PET plastic - as the electron donor, since these materials oxidize far more easily than water and also generate useful byproducts for the chemical industry along the way.

How the reactor is built

The reactor consists of two layers coated onto a glass base: a light-absorbing layer made of aluminum-doped strontium titanate (Al:SrTiO3) powder - a cheap, stable material that's easy to manufacture at scale - topped with a thin cobalt-zirconium catalyst layer that drives hydrogen production.

The key innovation lies in how the cobalt catalyst layer is made. Previously, attaching this type of catalyst to a reactor surface usually required high temperatures or a polymer binder, both of which are difficult to scale up. Instead, the team synthesized a single compound containing both cobalt and zirconium in a fixed ratio (known as a single-source precursor). Once dissolved in solvent, this compound can be sprayed directly onto a glass surface and left to dry at room temperature, forming a thin, transparent coating that doesn't block light from reaching the absorbing layer below - unlike many precious-metal catalysts.

Results across different scales

The team tested the reactor at multiple scales: from 1 cm² in the lab, up to 20.25 cm², and finally four 0.25 m² panels combined into a 1 m² outdoor system. At small scale under simulated light, the reactor performed best with glucose as the feedstock, and still worked with cellulose or pretreated PET plastic (shredded and dissolved in hot alkaline solution).

In the outdoor test at Cambridge - where natural sunlight intensity was only about half that of the lab simulator - the 1 m² system ran continuously for six hours, producing roughly 5.24 mmol of hydrogen per square meter from glucose and 1.51 mmol per square meter from pretreated cellulose, while also generating valuable byproducts such as formate and acetate. The reactor surface stayed above 30°C despite the relatively cool outdoor weather, and bubble formation was observed throughout the run.

Catalyst leaching and reusability

After about 22 hours of operation, the team found that roughly 60% of the cobalt in the catalyst layer had dissolved away. However, spraying a fresh layer of cobalt catalyst onto the same used panel fully restored hydrogen output from glucose, showing that the underlying light-absorbing layer can be reused multiple times. With cellulose, performance was only partially restored, since leftover molecular fragments from the cellulose pretreatment process stick to the catalyst surface and reduce its activity.

Techno-economic analysis

Using real data from the 1 m² system, the team calculated the cost of producing hydrogen this way. The result came out higher than both conventional fossil-fuel-based hydrogen and previous theoretical estimates for this kind of technology - since earlier estimates were typically based on assumed ideal performance well above what's actually been measured. The team says costs could drop significantly with larger-scale deployment, repeated reuse of the reactor panels, and by accounting for the value of the byproducts generated during the reaction, none of which were factored into the initial cost estimate.

Limitations and next steps

The current semiconductor layer only absorbs ultraviolet light, which makes up just about 5% of the solar spectrum, meaning most of the sunlight's energy goes unused. Going forward, the team aims to develop a material that can also absorb visible light, build a continuous-flow system instead of one that needs the solution reloaded after each run, and improve the catalyst to produce fewer types of byproducts, simplifying downstream separation.

Overall, the study suggests that combining single-source precursor chemistry with spray coating could offer a path toward building large-scale photocatalytic reactors without relying on precious metals or high-temperature processing - a step toward real-world deployment of this technology.

Source: Annuar, A.B.M., Liu, Y., Bhattacharjee, S. et al. Photoreforming of solid waste on 1 m² scale using single-source precursor-derived co-catalyst films. Nature Chemical Engineering 3, 351–362 (2026). https://doi.org/10.1038/s44286-026-00406-y