Probability Langmuir-Hinshelwood based CO2 photoreduction kinetic models

Abstract

For engineering solutions, scaling photoreactors and processes, kinetic models that describe the impact of process conditions on CO2 photoreduction are critical to driving this technology forward. Probability Langmuir-Hinshelwood based CO2 photoreduction kinetic models were developed after several criteria that included: a high purity photodifferential photoreactor with a high ratio of reagent gas volume to irradiated photocatalyst surface area and automated robust data collection and kinetic modelling using a MATLAB program. Product distribution profiles indicated the dynamic changes occurring over the photocatalyst with an initial increase in H2 product distribution, followed by an increase in CH4 and finally CO product distribution, possibly due to the photocatalytic degradation of CH2O and CH2O2 intermediates. Production of H2 increased with a decrease in CH4 when the partial pressure of H2O was increased. Using the glyoxal mechanism, this is possibly explained via the formation of CH3CO2H from H2O reacting with CH3CHO that prevents the full conversion of CH3CHO to CH4. To account for deactivation, probability Langmuir–Hinshelwood based kinetic models were used to fit CO2 photoreduction kinetic data for CH4, CO and H2 with low average standard errors of 3.44×10-4,1.54×10-4 and 1.36×10-4, respectively. The probability LH based kinetic model coefficients were estimated with low standard deviations, using a robust and repeatable numerical method using a trust-region and multi-start algorithm. The models were used to predict optimised selectivity of CH4, CO and H2.