Catalytic Decomposition of an Organic Electrolyte to Methane by a Cu Complex-Derived In Situ CO2 Reduction Catalyst

Metal complexes are often transformed to metal complex-derived catalysts during electrochemical CO2 reduction, enhancing the catalytic performance of CO2 reduction or changing product selectivity. To date, it has not been investigated whether metal-complex derived catalysts also enhance the decomposition of the solvent/electrolyte components as compared to an uncoated electrode. Here, we tested the electrochemical stability of five organic solvent-based electrolytes with and without a Cu complex-derived catalyst on carbon paper in an inert atmosphere. The amount of methane and hydrogen produced was monitored using gas chromatography. Importantly, the onset potential for methane production was reduced by 300 mV in the presence of a Cu complex-derived catalyst leading to a significant amount of methane (417.7 ppm) produced at −2.17 V vs Fc/Fc+ in acetonitrile. This suggests that the Cu complex-derived catalyst accelerated not only CO2 reduction but also the reduction of the electrolyte components. This means that Faradaic efficiency (FE) measurements under CO2 in acetonitrile may significantly overestimate the amount of CH4. Only 28.8 ppm of methane was produced in dimethylformamide under an inert atmosphere, much lower than that produced under CO2 (506 ppm under CO2) at the same potential, suggesting that dimethylformamide is a more suitable solvent. Measurements in propylene carbonate produced mostly hydrogen gas while in dimethyl sulfoxide and 3-methoxypropionitrile neither methane nor hydrogen was detected. A strong linear correlation between the measured current and the amount of methane produced with and without the Cu complex-derived catalyst confirmed that the origin of methane production is solvent/electrolyte decomposition and not the decomposition of the catalyst itself. The study highlights that in a nonaqueous system, highly active catalyst in situ deposited during electrochemical testing can significantly influence background measurements as compared to uncoated electrodes, therefore the choice of solvent is paramount for reliable testing.