Electroreduction of carbon monoxide on copper electrodes

Abstract

Electrochemical CO2 reduction is receiving increasing attention as a means for (i) storage of renewable electricity and (ii) recycling of CO2 into valuable chemicals and fuels. Cu is the only monometallic catalyst that can facilitate formation of >2e- products with moderate selectivity and high activity. However, a mixture of products are formed and low energy efficiency is achieved for formation of >2e- products. As a result, improvements are needed before the technology can be commercialised. One approach is electroreduction of CO, for which previous studies report high oxygenate selectivity at low overpotentials on nanostructured catalysts. Since CO is as an intermediate in >2e- product formation, any insight obtained for CO reduction will aid our understanding of CO2 reduction. In this thesis, I have studied CO reduction on nanostructured and polycrystalline Cu electrodes. Due to the mixture of products formed during CO reduction, thorough product analysis is important in order to obtain all the information possible about the reaction. We benchmarked the performance of two techniques used for liquid  product analysis; static headspace-gas chromatography (HSGC) and NMR spectroscopy. CO reduction is often carried out in alkaline electrolyte, since this leads to enhanced C2+ product selectivity and suppressed H2 evolution. The high pH can, however, also lead to undesired reactions occurring with the liquid products from CO reduction. It turns out that acetaldehyde and propionaldehyde are unstable in alkaline electrolytes, leading to polymerisation and precipitation. As a result, these aldehydes, in particular acetaldehyde, are difficult to detect using NMR spectroscopy. HS-GC, on the other hand, can quantify aldehydes with high sensitivity. Using HS-GC for liquid product analysis allowed us to detect acetaldehyde as an additional, previously overlooked product from CO reduction on oxide-derived Cu. Using a combination of experiments and DFT calculations, we determined that acetaldehyde is an intermediate in the reduction of CO to ethanol. This has been previously observed for polycrystalline Cu, confirming that ethanol formation occurs through a similar pathway on planar and nanostructured electrodes. We identied a single intermediate in the further conversion of acetaldehyde to ethanol, which represents a thermodynamically uphill step. The free energy of this intermediate is thus likely to determine whether ethanol production is favoured or not. Although nanostructured Cu electrodes are promising for CO reduction to >2e- products, a robust benchmark for the activity of Cu in this potential region is yet to be reported. Polycrystalline foils represent a robust benchmark for CO2 reduction, which led us to study CO reduction on polycrystalline Cu. We measured relatively high selectivity and activity between -0.40 and -0.59 V vs. RHE, with similar total CO reduction activity to that of nanostructured Cu, when normalised by ECSA. This suggests that nanostructuring mainly results in a change in product distribution, and might not influence the intrinsic activity of Cu to a signicant extent. Signicant deactivation could be observed for polycrystalline Cu during CO reduction. We mainly attributed this to poisoning by Si from the glass cell. However, surface restructuring has been proposed in the literature as a possible reason for activity and selectivity changes during CO2 and CO reduction. Phenomena occurring under reaction conditions are diffcult to evaluate using ex-situ characterisation. This led us to perform operando characterisation of Cu during CO reduction using synchrotron techniques.  In the results reported in this thesis, we used grazing incidence X-ray diffraction to study the average surface and near-surface structure of polycrystalline Cu lms under reaction conditions. We have obtained preliminary data this far that demonstrate the possibility to apply this technique under CO reduction conditions. I also present results where certain regions of the spectra are tracked real-time as the reaction conditions are changed.