Understanding the Inconsistencies in the Literature for the CO2 Reduction Reaction on Polycrystalline Copper

Abstract

As the desire to for carbon neutral processes increases with the prominence of global warming, the electrochemical reduction of carbon dioxide has the ability to not only reduce atmospheric CO2 levels, but also produce value added chemicals such as CO, C2H4, and CHOOH using renewable energy. Polycrystalline copper is a popular electrocatalyst for the CO2 reduction reaction (CO2RR) as it is the only pure metal catalyst to reduce CO2 to products that require more than two electrons, at reasonable faradaic efficiencies [1]. This has led to the development of many novel copper-based catalysts, such as oxide derived copper, copper nanoneedles and copper nanospheres to enhance the activity and selectivity of the CO2RR [2-5]. Often when developing these unique catalysts, polycrystalline copper will be used as a reference material to show how the new material is more active and has better selectivity [2-5]. However, the literature shows substantial inconsistencies for the selectivity of the CO2RR on polycrystalline copper (under identical conditions), which brings into question the reliability of the results on the more complicated materials. For example, on polycrystalline copper, Mistry et al. reported a faradaic efficiency for methane and ethylene of 43% and 31%, respectively, at -1 V vs RHE in CO2 saturated 0.1 M KHCO3 solution [6], whereas Hori et al. reported a faradaic efficiency of 0% for both methane and ethylene under identical conditions [7].Here, we show that the discrepancy is likely not due to a single factor alone, but it is made up of several contributing factors. These include surface pre-treatments of the polycrystalline copper [7, 8] (such as sanding, mechanical polishing, electropolishing and etching), natural variation in the crystal orientation on the copper surface [9-11], and the way that iR compensation is applied and subsequently corrected for post-experiment. It was found that the faradaic efficiency for methane and ethylene could increase by up to 140% and 124%, respectively, over a 0.035 V range, which if iR compensation is not correctly adjusted for post experiment, could lead to significant errors in the reported results. We also show that the solution resistance can change by 51% over a 6 hour period of electrolysis, requiring the solution resistance to be regularly measured throughout the experiment to be able to correctly adjust the potential post experiment. Single crystal CO2RR experiments have also shown differences between the mechanism for reducing CO on a Cu(111) surface, compared to a Cu(100) surface, which leads to different activities and selectivities [12]. Interestingly, we show electron backscatter diffraction (EBSD) measurements for different pieces of the same polycrystalline copper that have different crystal orientation distributions.