Development of innovative photoreactors and photocatalytic processes for hydrogen production

Introduction The direct use of solar energy is intriguing for H2 production. The direct photocatalytic water splitting (WS) is thermodynamically limited by the high Gibbs free energy (237 kJ/mol) and very low efficiency is reported for direct WS also for kinetic reasons. Sacrificial reagents, such as methanol or EDTA, can improve hydrogen productivity, but they are non renewable. Compared to WS, the photocatalytic reforming (PR) is a valid approach to produce H2 under ambient conditions and using sunlight, the cheapest energy source available on earth. PR is also thermodynamically more feasible than WS. Thus, the attention is here focused on the use of waste organic compounds to be used as sacrificial agents [1], such as organic compounds obtained through the photoreduction of CO2 or the photoreforming of organic solutions, e.g. carbohydrate containing hydrolysed substrates. The attention was mainly focused on photoreactor design and the relative process, i.e. on the development of suitable devices that could be rather easily scaled up and that can maximize the hydrogen productivity. In particular, for the photoreduction of CO2 a fully innovative photoreactor was realized, able to operate up to 20 bar pressure. This boosted the solubility of CO2 and, thus, its conversion to regenerated fuels and hydrogen. This approach allowed us to explore high pressure and high temperature operating conditions, which are unconventional for photocatalysis. On the other hand, the optimization of a photocatalytic process based on highly non-ideal, concentrated, sugar-based solutions is non trivial. At last, calculations were done to size a full scale reactor able to harvest solar light for hydrogen production in different geographic zones. Experimental Different nanostructured materials were used. Commercial nanometric titanium dioxide P25 by Evonik was used as photocatalyst and suspended in water with a concentration of 0.25-0.75 gL-1. Alternative nanostructured materials were prepared by flame spray pyrolysis (FP) or in mesoporous form by wet template synthesis. Different metals were added as cocatalysts, e.g. Au nanoclusters (0.1-0.5 wt%), Ag and Pd (0.1 mol%). Photocatalytic testing was carried out on a bench scale reactor at ambient pressure, using different carbohydrates and organic model molecules (methanol, glucose, xylose, arabinose, formic acid, methanol, formaldehyde and levulinic acid). These were selected because they represent the basic composition of an acid-hydrolysed fraction from cellulose. On the other hand, CO2 photoreduction was carried out in a dedicated 1.2 L reactor, at high pressure (up to 20 bar) and temperature (up to 80°C), in the presence of Na2SO3 as hole scavenger. Results and discussion The photoreduction of CO2 investigated at high pressure is a fully new approach proposed by our group [2-4]. CH4 can be obtained as gas phase product (e.g. with Au/TiO2 catalysts), whereas in liquid phase formic acid, formaldehyde and methanol, can be obtained in variable amount depending on the operating conditions and catalyst used. Unexpectedly, considerable H2 amount in gas phase was also obtained, sometimes as primary product. The hypothesis that it is produced from WS was ruled out by the absence of a corresponding stoichiometric amount of oxygen. Furthermore, by exploring the products distribution as a function of time, we observed that liquid organic products accumulate until Na2SO3 is present, then, organics start to convert through photoreforming generating H2, possibly CO and CO2 [2,3]. Productivity as high as 102 mmol h−1 kgcat−1 for H2, 16537 mmol h−1 kgcat−1 for formaldehyde and 2954 mmol h−1 kgcat−1 for formic acid were achieved when operating at a 7 bar of CO2 over the aqueous solution, 80 °C with 0.5 g L−1 TiO2 by tuning reaction time and pH. The reaction time in batch mode ranged from 3 to 24 h, the longer the reaction time, the higher H2 productivity. On the other hand, different organic compounds have been tested for the photoreforming under ambient conditions in batch mode by using a 0.1 wt% suspension of the photocatalyst and an amount of 5-20 wt% of different organic compounds in water (Fig. 1 and 2). Irradiation was achieved with an UVA lamp, with maximum emission at 365 nm and measured irradiating power of 113 W/m2. The highest substrate conversion was obtained with 0.1 wt% Au/TiO2-rutile, which however gave a wider spectrum of intermediate products in liquid phase with respect to 0.1 wt% Au/TiO2-P25 and, therefore, lower H2 productivity. The best results obtained with the rutile TiO2 sample were 89 mmol kgcat-1 h-1 of H2, 7 mmol kgcat-1 h-1 of CO, 74 mmol kgcat-1 h-1 of CO2. H2 productivity increased to 276 mmol kgcat-1 h-1 by using P25, whereas it dropped to 40 mmol kgcat-1 h-1 when using anatase as polymorph for TiO2. The conceptual feasibility of a photoreactor based on these results has been investigated considering both a continuous apparatus with UV irradiation, and solar light. In both cases, the hydrogen productivity and the efficiency of solar light storage seem insufficient for a practical exploitation. However, the same study was based on one of the best hydrogen productivities reported in the literature under visible light and also in such a promising case the feasibility does not seem guaranteed. The photoreduction of CO2 has been also considered as a process for the fixation of this greenhouse gas to useful fuels by storing solar energy. Hydrogen productivity was insufficient for practical interest, being similar to the results obtained by photoreforming. On the other hand, also reduced organic products accumulate in liquid phase, among which the productivity of formaldehyde is particularly interesting. When using UV lamps 35 kg/day kgcat can be obtained, which decrease to 1.5 under solar light irradiation. This latter value corresponds to a 13% efficiency of solar light storage. Conclusions Interesting productivities of regererated fuels and H2 have been obtained by selecting proper nanostructured photocatalysts and operating conditions through photoreduction of CO2 and photoreforming of various organic substrates. The results have been used for photoreactor design and to assess the feasibility of this process. The productivity is still insufficient for the direct exploitation of solar light, but the process is feasible for the fixation of CO2 as different organic compounds to be further transformed into H2. Acknowledgements The financial contribution of MIUR through the PRIN2015 grant (20153T4REF) is gratefully acknowledged (G. Ramis and I. Rossetti). I. Rossetti and E. Bahadori are grateful to Fondazione Cariplo and Regione Lombardia for financial support through the grant 2016-0858 – Up-Unconventional Photoreactors. References [1] I. Rossetti, ISRN Chemical Engineering, Article ID 964936, (2012). doi:10.5402/2012/964936. [2] F. Galli, M. Compagnoni, D. Vitali, C. Pirola, C. Bianchi, A. Villa, L. Prati, I. Rossetti, Appl. Catal. B: Environmental, 200 (2017) 386 [3] I. Rossetti, A. Villa, M. Compagnoni, C. Pirola, L. Prati, G. Ramis , W. Wang, D. Wang, Catal. Sci & Technol., 5 (2015) 4481 [4] I. Rossetti, A. Villa, C. Pirola, L. Prati, G. Ramis, RSC Adv., 4 (2014) 28883