Photoreduction of CO2 to formic acid in a high pressure photoreactor

Introduction The photoreduction of CO2 is one of the most challenging uphill reactions, whose performance is limited by intrinsic catalyst efficiency features and by physical phenomena, such as CO2 solubility in water. Nevertheless, it represents an intriguing mean to produce fuels and chemicals such as formic acid, formaldehyde, methanol and methane. In this work an innovative high pressure photoreactor is tested, operating up to 20 bar and reaching unprecedented productivity of formic acid (ca. 40 mol/h kgcat) over a litre scale size. Thanks to 3D printing and in-house developed coating techniques up scaling options were studied and the possibility to mode to continuous operation was investigated. Materials and Methods Photocatalysts testing was carried out in a pilot scale photoreactor in stainless steel, equipped with an immersion UV lamp (365 nm maximum emission, with measured irradiance of 75 W/m2) on 1.2 L of catalyst suspension or solution with immobilised catalytic tiles (31 mg/L) and using sodium sulphite as hole scavenger. The products were analysed both in liquid and gas phase through GC and HPLC. Testing was carried out up to 20 bar and 80°C In this work the main photocatalyst tested is graphitic carbon nitride (g-C3N4). This substance has a lower bandgap than titanium dioxide (2.7 eV of the g-C3N4 with respect to 3.23 eV of TiO2), allowing a theoretical better sunlight harvesting. To further increase the productivity of the g-C3N4 a chemical treatment employing sulphuric acid was optimized and compared with exfoliation under UltraSound (US) treatment at modulated power (0-120 W). The surface area increase deriving from this treatment induced an increase of the catalyst productivity. Different strategies for the functionalization of graphitic carbon nitride have been employed with the aim to obtain direct Z-Scheme photocatalysts. The functionalization of the g-C3N4 exfoliated was performed using different types of metal oxides with various loadings. In particular, the chosen co-semiconductors for the modifications were iron oxide, zinc oxide and tin oxide. All the catalysts have been characterized by XRD, BET and DRS analysis. Results and Discussion Exfoliation of graphitic carbon nitride g-C3N4 by means of US treatment using water as a solvent was demonstrated at varying input power at constant frequency, constant amplitude and time of effective sonication. This positively contributed to the properties of the final material without critical handling or environmental issues. Exfoliation of g-C3N4 in water displays a strong dependence of US input power, with a slightly enhanced bandgap (2.8 eV), but most of all increased lifetime of photogenerated electrons, as observed through Diffuse Reflectance Spectroscopy (DRS) and Spectrofluorimetry data. Among all applied power (varied between 30W and 120W), 120W sufficiently exfoliated and tuned physicochemical properties of g-C3N4. Compared to bulk as prepared sample, exfoliated g-C3N4 exhibited improvement in photoinduced charge carrier transfer and separation, resulting in higher photocatalytic efficiencies. FE-SEM and TEM images of both bulk and exfoliated g-C3N4 show the effect of the exfoliation power on the nanosheet, pseudo-lamellar structure. Accordingly, the bandgap and charges lifetime of the materials correlate well with change in input power and, as well, the catalytic performance determined through an innovative high-pressure reactor in solid-liquid-gas phase. The catalytic results demonstrate this metal free material as an efficient photocatalyst to obtain high yield of formic acid with productivities ranging from ~5100 to ~8200 mmol/kgcath at 80°C in water, which is among the highest reported in the literature. Another effective method to improve the surface area of the material is its synthesis in H2SO4, which improves by 5 times the surface area and activity with respect to the bulk sample obtained by calcination of melamine, though the activity of the sample exfoliated at 120 W US power was the most active of the series. The graphitic carbon nitride with a loading of hematite equal to 8% in weight showed the best performances among this series, with an increase of the productivity of formic acid (the main product of the photoreduction process) of 26.1% respect to the bare graphitic carbon nitride. The addition of sodium sulphite a hole scavenger (HS) which is oxidized into sulphate was fundamental to achieve practically relevant productivity. It was found out that in presence of an UV radiation with a wavelength shorter than 265 nm, sodium sulphite undergoes a radicalic reaction, producing sulphite radical and a solvated electron, having a high reducing power. This contributes a most significant reaction of the photoreduction process, enhancing the effective conversion that can be addressed to the photocatalyst. Significance Significant productivity of formic acid was achieved through a high pressure photoreactor operating up to 20 bar and 80°C. Different samples of graphitic carbon nitride were compared and the best results were achieved upon exfoliation of the material, thanks to increased photocharges lifetime as determined by spectrofluorimetry. Acknowledgements The authors gratefully acknowledge the financial contribution of Fondazione Cariplo through the grant 2021-0855 – “SCORE - Solar Energy for Circular CO2 Photoconversion and Chemicals Regeneration”, funded in the frame of the Circular Economy call 2021. I. Rossetti acknowledges Università degli Studi di Milano for support through the grant PSR 2021 - GSA - Linea 6 “One Health Action Hub: University Task Force for the resilience of territorial ecosystems”. This study was carried out within the Agritech National Research Center and received funding from the European Union Next-Generation EU (PIANO NAZIONALE DI RIPRESA E RESILIENZA (PNRR) – MISSIONE 4 COMPONENTE 2, INVESTIMENTO 1.4 – D.D. 1032 17/06/2022, CN00000022). This manuscript reflects only the authors’ views and opinions, neither the European Union nor the European Commission can be considered responsible for them. I. Rossetti and M. Tommasi acknowledge specifically the participation and funding of Tasks 8.2.3, 8.3.2 and 8.4.1.