Impact of defect-driven surface inhomogeneity and minimal island formation on temperature-programmed desorption spectra: A kinetic Monte Carlo study

Inhomogeneity often manifests itself on surfaces with pre-adsorbed adsorbates in which islands have widely-disparate sizes due to the presence of distinct morphological features or perhaps even impurities, or defects. In this kinetic Monte Carlo study, we developed a model in which these defects are in the form of vacancy clusters or chains and grow progressively longer with each repeated adsorption-desorption cycle owing to repeated stress that a substrate is subjected to with each cycle. We then explored different simulation-imposed conditions vacancy or defect concentration as well as maximum vacancy chain length - and noted their impact on the resulting temperature-programmed desorption spectra. We found that growing chains can induce single desorption peak to progressively shift, then split into two closely-overlapping ones over a course of several such cycles. Interestingly, such evolving peaks may hide the true desorption kinetics since calculation of desorption order may yield unconventional, non-integer values - neither zero, first or second-order. We also observed that for the same vacancy concentration, peak-splitting becomes more pronounced with increasing maximum chain length. Analysis of simulation data also revealed that unlike bigger and more compact islands, small ones follow first-order desorption kinetics rather than half-order kinetics.