Non-Covalent Molecular Interaction Rules to Define Internal Dimer Coordinates for Quantum Mechanical Potential Energy Scans

Abstract

Non-covalent interactions (NCI) dominate the properties of condensed phase systems. Towards a detailed understanding of NCI, quantum mechanical (QM) methods allow for accurate estimates of interaction energies and geometries, allowing for the contributions of different types of NCI to condensed phase properties to be understood. In addition, such information can be used for the optimization of empirical force fields, including the specific contribution of electrostatic versus van der Waals interactions. However, to date, the relative orientation of monomers defining molecular interactions of dimers is often based on full geometry optimizations of all degrees of freedom or extracted from known experimental structures of biological molecules. In such cases, the spatial relationship of the monomers often leads to multiple atoms in each monomer making significant contributions to the interactions occurring in the dimer, confounding understanding of the contributions of specific atoms or functional groups. To overcome this, a workflow is presented that allows for systematic control of the interaction orientation between monomers to be performed through the use of molecular interaction rules (MIR) in an extendable tool that can be applied to a broad range of chemical space. Using the “MIR workflow” allows a user to perform automation of the determination of well-defined monomer interaction orientations in dimers using Z-matrices, allowing for potential energy scans (PES) to be performed on combinatorial pairs of the monomers. In addition, compiled monomer and dimer geometries and PES data are stored in an extendable database. Illustration of the utility of the workflow is performed based on a collection of 89 monomers encompassing a variety of functional group classes from which 10,616 interaction dimers can be automatically generated. PES between all dimers were calculated at the QM HF/6-31G*, MP2/6-31G*, and ωb97x-d3/6-31G* model chemistries. In addition, analysis of the benzene dimer in three interaction orientations, a hydrogen bond interaction between azetidinone and N-methylacetamide, and the interaction of pyridine with acetone in the Burgi–Dunitz orientation are presented including results with the aug-cc-pVDZ basis set. Results show the impact of different QM model chemistries on minimum interaction energies and distances over a large ensemble of intermolecular interactions with emphasis on the contributions of dispersion.