Benchmarking of Density Functional Theories for Combustion Reactions in Alkanes by Evaluating Thermodynamic Properties

Computational calculations are playing vital role in this era of artificial intelligence and machine learning which is capable to lessen the burden of experimental expenses. Inevitably, to benchmark different theoretical methods is essential to produce accurate results while keeping in mind the computational cost. In this study, different functionals of Density Functional Theory (DFT) with 2 different basis sets were benchmarked to compute thermodynamic properties, enthalpy, Gibbs free energy, and entropy, of alkane combustion reactions. Results indicate a linear relationship between the number of carbon atoms and reaction parameters, with deviations arising from method-dependent approximations. LSDA and dispersion-corrected methods demonstrated closer agreement with experimental values with correlation-consistent basis set, while higher-rung functionals like PBE and TPSS exhibited significant errors with split-valence basis set, especially for larger chain lengths. Notably, convergence issues were observed for n-hexane with PBE and TPSS, attributed to near-degenerate states and SCF instability. These findings highlight the importance of functional choice along with the basis set to define the orbitals to accurately predict thermodynamic properties.