Deactivation of catalytic processes such as direct oxidation of hydrogen sulfide by coking through hydrocarbon combustion has always been a topic of concern. Using n-hexane as a model compound, we develop a kinetic model that predicts its coking rate on the basis of data obtained from a thermobalance. Considering two different processes in the coke reaction, active site coverage and coke growth, and applying a probabilistic model, we obtain an apparent activation energy of 34 and 126 kJ/mol, respectively. Thus, we estimate pore-plugging effects using two network models: a bundle of parallel cylindrical pores using a probabilistic model and a microporous network distributed in a Bethe lattice. Both models describe experimental data. Moreover, they confirm that the TiO 2 catalyst used first deactivates through active sites coverage and second through pore obstruction. As such, the model may be used for estimating deactivation of reactions taking place simultaneously on the catalyst.