Modeling the Association Between Operating Conditions and Catalytic Performance in CO2-to-Methanol Conversion
Emily Cheong
07/09/2026
Converting a greenhouse gas into a useful chemical and energy carrier has been a pressing societal concern, and carbon dioxide hydrogenation to methanol is a promising pathway for it. However, catalytic performance substantially varies across experimental studies. This study focused on examining how four commonly reported operating conditions, namely; temperature, pressure, hydrogen-to-carbon dioxide ratio, and gas hourly space velocity (GHSV), were associated with methanol yield and CO2 conversion. Data were obtained from an open literature-derived dataset of thermocatalytic CO2-to-methanol experiments named TheMeCat. For methanol yield (n = 712) and CO2 conversion (n = 696), separate ordinary least squares regression models were estimated. Due to the nature of strongly right-skewed distribution of GHSV and multiplicative differences in gas flow, natural logarithm was used for GHSV. Heteroskedasticity-robust standard errors were reported. The methanol-yield model explained 29.7% of observed variation, while the CO2-conversion model explained 31.8% of observed variation. A positive association was reported between both temperature and H2/CO2 ratio with both outcomes. However, a negative association was reported between ln(GHSV) and both outcomes. Pressure behaved differently in both models. It was positively associated with methanol yield. However, it was not statistically significant in the CO2-conversion model. With these findings, the hypothesis was partially supported, suggesting that a small set of operating variables may explain a meaningful yet incomplete portion of performance differences across published experiments. The analysis also demonstrates how a simple and reproducible model may provide interpretable trends from a curated catalysis dataset without claiming causal proof.