Heterogeneous Pair Approximation of Methanol Oxidation on TiO2 Reveals Two Reaction Pathways

18 November 2021, Version 1
This content is a preprint and has not undergone peer review at the time of posting.


We propose a novel method to simulate the chemical kinetics of methanol oxidation on the rutile TiO2(110) surface. This method must be able to capture the effects of static disorder (site-to-site variations in the rate constants), as well as dynamic correlation (interdependent probabilities of finding reactants and products next to each other). Combining the intuitions of the mean-field steady state (MFSS) method and the pair approximation (PA), we consider representative pairs of sites in a self-consistent bath of the average pairwise correlation. Pre-averaging over the static disorder in one site of each pair makes this half heterogeneous pair approximation (HHPA) efficient enough to simulate systems of several species and calibrate rate constants. According to the simulated kinetics, a static disorder in the hole transfer steps suffices to reproduce the stretched exponentials in the observed kinetics. The identity of the dominant hole scavenger is found to be temperature-dependent -- the methoxy anion at 80 K and the methanol molecule at 180 K. Moreover, two distinct groups of 5-coordinate titanium (Ti5c) sites emerge -- a high-activity group and a low-activity group -- even though no such division exists in the rate constants. Since the division is quite insensitive to the type of static disorder, the emergence of the two groups might play a significant role in a variety of photocatalytic processes on TiO2.


microkinetic modelling
methanol oxidation
titnaium dioxide
pair approximation
static disorder
dynamic correlation
reaction mechanism

Supplementary materials

Supplementary Information: Heterogeneous Pair Approximation of Methanol Oxidation on Rutile TiO2(110) Reveals Two Reaction Pathways
Demonstration of HMF and HHPA on a number of simple model systems; cross-examination of calibrated parameters in the MF and the PA methods; simulated kinetics in the modified model; sensitivity analysis of the calibrated parameters in the original model and the modified model; construction and outcomes of the compact model; extended discussion on the origin of the bimodal distributions in the coverages and the rates; distributions of the coverages and the rates in the original, modified, and compact models; and normal, Poisson (k = 2), and hyperbolic secant distributions of trap energies.

Supplementary weblinks


Comments are not moderated before they are posted, but they can be removed by the site moderators if they are found to be in contravention of our Commenting Policy [opens in a new tab] - please read this policy before you post. Comments should be used for scholarly discussion of the content in question. You can find more information about how to use the commenting feature here [opens in a new tab] .
This site is protected by reCAPTCHA and the Google Privacy Policy [opens in a new tab] and Terms of Service [opens in a new tab] apply.