Møller–Plesset 2 and density functional theory studies of the interaction between aromatic compounds and Zn-porphyrins Rafael Añez a,⇑ , Anibal Sierraalta a , David Coll b , Olga Castellanos c , Humberto Soscun d a Laboratorio de Química Física y Catálisis Computacional, Centro de Química, Instituto Venezolano de Investigaciones Científicas, Apartado 21827, Caracas, Venezuela b Laboratorio de Físico Química Teórica de Materiales, Centro de Química, Instituto Venezolano de Investigaciones Científicas, Apartado 21827, Caracas, Venezuela c Gerencia Departamental de Investigaciones RIIF, Gerencia Departamental de Refinación e Industrialización, Gerencia General de Refinación, Petróleos de Venezuela (PDVSA)-Intevep, Los Teques, Estado Miranda, Apartado 76343, Caracas 1070-A, Venezuela d Laboratorio de Química Inorgánica Teórica (LQIT), Departamento de Química, Facultad Experimental de Ciencias, La Universidad del Zulia, Apartado 526, Maracaibo 4001-A, Venezuela article info Article history: Received 17 February 2016 Received in revised form 16 March 2016 Accepted 17 March 2016 Available online 19 March 2016 abstract The M05 and M06 family of functionals were tested in the study of the interaction of five aromatic com- pounds with a Zn-porphyrin model and their results were compared with those obtained with dispersive functionals (APFD and B97D3), B3LYP and the MP2 methodology. Results obtained with the MP2 method- ology and functionals showed that interaction between aromatic compounds and Zn-porphyrin molecule is more like to a metal–ethylene interaction than a common cation-p cloud of an aromatic ring. Among all hybrid meta functionals analyzed, M06HF displayed the lowest mean absolute errors for the interaction energy and the structural parameters with respect to the MP2/6-311+G(d) methodology. Hartree–Fock exchange seems to play an important role in the performance of the M05 and M06 functionals for the description of the structural interaction of aromatic compounds with metal–porphyrin molecules. Ó 2016 Elsevier B.V. All rights reserved. 1. Introduction From a theoretical point of view, specifically from the density functional theory (DFT) perspective, there is an increased interest in the development of new methodologies that can correctly char- acterize non-covalent interactions. Hybrid functionals, called hybrid DFT functionals (HDFT), were the first intent of designing functionals able to perform well not only for the covalent interac- tion but also the weak interactions as the dispersion forces. Thereby, exchange functional as mPW and mPW1PW improved the long-range behavior retaining the essential characteristic of the PW functional [1]. The X3LYP extended density functional [2], derived from an exact exchange energy density for a density decaying with Gaussian-like behavior at long range, showed a significant improvement of bonding properties of van der Waals complexes with respect to the functionals of these times. New methodologies based on the mPW and Becke ´ s 1995 meta correla- tion functionals led to the methodologies MPW1B95 and MPWB1 K called hybrid meta DFT functionals (HMDFT) which showed the best performance respecting to non-covalent and weak interac- tions [3,4]. From a more complete treatment of the kinetic energy density, arise the M05 and M052X functionals [5,6]. The major difference between these functionals consists in the amount of nonlocal Hartree–Fock (HF) exchange which is represented with 2X and that is parametrized only for non-metals. Then M05 functional should be used for systems with metal–metal or metal–ligand interaction. Comparing 28 functionals, M05 and M052X showed the best performance for non-covalent interactions both in single and multi-reference systems. More recently, combining M05 and VXSC exchange functionals [7,8], born the M06 functionals which slightly improved the performance of the M05 functionals. Based in the evolution of the DFT functionals, seems to be quite obvious that to study systems where non-covalent interaction are important, the M05 or M06 family of functionals must be considered the best option. However, there are many types of non-covalent interactions which imply that it is necessary to be careful choosing the func- tional to describe the corresponding surface potential energy. Non-covalent interactions do not comprise merely interactions between aromatic organic compounds [9–12], the binding force between a cation and a p cloud is a non-covalent interaction involves in processes such as receptor-ligand interactions, enzyme-substrate binding and antigen–antibody recognition and this is not taken into account in the most of the training sets used to test these functionals. Then, a suitable functional for these interactions must include metals and p cloud interactions in the training sets for its parametrization. Based on the training sets http://dx.doi.org/10.1016/j.comptc.2016.03.024 2210-271X/Ó 2016 Elsevier B.V. All rights reserved. ⇑ Corresponding author. E-mail address: ranez@ivic.gob.ve (R. Añez). Computational and Theoretical Chemistry 1084 (2016) 133–139 Contents lists available at ScienceDirect Computational and Theoretical Chemistry journal homepage: www.elsevier.com/locate/comptc