Previous studies on the effects of pressure on the thermal-oxidative
decomposition of different sources of polycyclic aromatic hydrocarbons
(PAHs) that contain heteroatoms have shown that this process has four
main stages: oxygen chemisorption (OC), desorption/decomposition of
chemisorbed oxygen functional groups (DCO), and the first and second
stages of combustion (FC and SC, respectively). In this study, molecular
dynamics (MD) simulations were carried out to provide an atomistic
description of the reaction mechanism. Both classical force field
and reactive force field techniques were used to propose a reaction
route at conditions similar to those used in experimental evaluations.
The results obtained from the MD simulations were compared with the
experimentally derived conversion profiles. It was found that the
MD models could successfully reproduce the experimental results and
were thus a suitable description of the experimental system. In addition,
a machine learning (ML) model was proposed to predict the thermal-oxidative
profiles based on variables obtained from the experimental characterization
of the PAHs. The ML model had a coefficient of determination was close
to 95% for both the training and validation sets. Thus, the proposed
ML model can be useful in optimizing the thermal-oxidative process
by considering the structural-compositional characteristics of the
PAHs and can be potentially extended to the study of promising techniques
aimed at improving the transformation of PAHs into less hazardous
or nonhazardous compounds by thermal-oxidative processes.