The production of hydrogen through photoreforming of aqueous solutions of organic compounds is considered as a way to exploit solar energy storage in the form of hydrogen. The photocatalytic reforming can be promoted by a photocatalyst, optimally a solid semiconductor able to absorb the major quota of solar energy. In this work, glucose was used as a model substrate for photoreforming, possibly derived from the hydrolysis of waste biomass. The selected photocatalysts were based on TiO2. The materials were prepared by different methods, e.g. flame spray pyrolysis, precipitation and in mesoporous form through soft template synthesis, and compared with commercial samples of nanostructured TiO2 P25 by Evonik. 0.1 mol% Pt was also added as co-catalyst. The role of the metal was that of electron sink, to inhibit the recombination of the electron/hole couple obtained through light absorption.
The photoreforming reaction was carried out in different prototypes of photoreactors, specifically developed and optimised. In one reactor configuration, an external 200 W lamp was used, with emission wavelengths centred around 365 nm. A first photoreactor was developed with internal capacity ca. 0.3 L, with big head space for gas collection and very efficient mixing of the suspension thanks to an optimized length/diameter ratio (L/D) ca. 2. A drawback was the poor irradiation efficiency of the suspension, which limited the overall productivity, irrespectively or the substrate. A different home-designed photoreactor was equipped with an immersion lamp, coaxial with the reactor. Two reactor sizes were tested, 0.3 L or 1.5 L. A significant amount of H2 was obtained with these very simple catalyst formulations, up to 14.7 mol kgcat-1 h-1 over the Pt/TiO2 sample and using glucose as substrate. This result is very remarkable with respect to conventional photoreactors.