Computational Study and QSPR Approach on the Relationship between Corrosion Inhibition Efficiency and Molecular Electronic Properties of Some Benzodiazepine Derivatives on C-steel Surface

The density functional theory (dft) study was used to investigate the corrosion inhibition performance of four inhibitors namely: 2,3-dihydro-1h-pyrrolo [2,1c][1,4] benzodiazepine-5,11 (10h,11ah)-dione (bzd1); 2,3-dihydro-1h-pyrrolo [2,1c][1,4] benzodiazepine-5,11(10h,11ah)- dithione (bzd2); 10-benzyl-pyrrolo [2,1-c][1,4] benzodiazepine-5,11-dione (bzd=2o) and 10- benzyl-pyrrolo [2,1-c][1,4] benzodiazepine-5,11-dithione (bzd=2s) on carbon steel using the b3lyp/6-311g(d, p) level of theory. the geometry optimization was conducted only on the neutral form of all benzodiazepine derivatives, without any protonated form as testified by the marvin software. the most relevant quantum chemical parameters according to their potential action as corrosion inhibitors were calculated. the descriptors considered were: the ehomo (highest occupied molecular orbital energy), the elumo (lowest unoccupied molecular orbital energy), the energy gap (δe), the dipole moment (μ), the hardness (η), the softness (σ), the absolute electronegativity (χ), the total energy (e_tot), the ionization potential (ip), the electron affinity (ea) and the fraction of electrons transferred (δn). besides, the local reactivity was analyzed through the fukui function in order to compare the possible sites for nucleophilic and electrophilic attacks. the electronic properties of these inhibitors obtained by dft were correlated with their experimental efficiencies using two mathematical models, based-qspr approaches; the multiple linear regressions (mlr) and the multiple polynomial regressions (mpr). the quantum chemical study showed that the theoretical and experimental results were in good agreement, and the statistical results revealed that the mpr was the most relevant and predictive model in comparison with the mlr model, with a very high determination coefficient (r^2 = 0.99), adjusted determination coefficient (r^2_adj = 0.99) and predicted determination coefficient (r^2_pred = 0.97)