Ab initio Study of the Interactions between CO2 and N‐Containing Organic Heterocycles

Abstract In the garden of dispersion: High‐accuracy ab initio calculations are performed to determine the nature of the interactions and the most favorable geometries between CO 2 and heteroaromatic molecules containing nitrogen (see figure). Dispersion forces play a key role in the stabilization of the dimer, because correlation effects represent about 50 % of the total interaction energy. magnified image The interactions between carbon dioxide and organic heterocyclic molecules containing nitrogen are studied by using high‐accuracy ab initio methods. Various adsorption positions are examined for pyridine. The preferred configuration is an in‐plane configuration. An electron donor–electron acceptor (EDA) mechanism between the carbon of CO 2 and the nitrogen of the heterocycle and weak hydrogen bonds stabilize the complex, with important contributions from dispersion and induction forces. Quantitative results of the binding energy of CO 2 to pyridine (C 5 H 5 N), pyrimidine, pyridazine, and pyrazine (C 4 H 4 N 2 ), triazine (C 3 H 3 N 3 ), imidazole (C 3 H 4 N 2 ), tetrazole (CH 2 N 4 ), purine (C 5 H 4 N 4 ), imidazopyridine (C 6 H 5 N 3 ), adenine (C 5 H 5 N 5 ), and imidazopyridamine (C 6 H 6 N 4 ) for the in‐plane configuration are presented. For purine, three different binding sites are examined. An approximate coupled‐cluster model including single and double excitations with a perturbative estimation of triple excitations (CCSD(T)) is used for benchmark calculations. The CCSD(T) basis‐set limit is approximated from explicitly correlated second‐order Møller–Plesset (MP2‐F12) calculations in the aug‐cc‐pVTZ basis in conjunction with contributions from single, double, and triple excitations calculated at the CCSD(T)/6‐311++G** level of theory. Extrapolations to the MP2 basis‐set limit coincide with the MP2‐F12 calculations. The results are interpreted in terms of electrostatic potential maps and electron density redistribution plots. The effectiveness of density functional theory with the empirical dispersion correction of Grimme (DFT‐D) is also examined.