Chemical reactivity in complex media from novel QM/MM models in periodic boundary conditions
Simone Bonfrate ¹
¹ Aix Marseille Univ, CNRS, ICR, Marseille, France
Hybrid Quantum Mechanics /Molecular Mechanics (QM/MM) models have emerged as the standard approach for studying reactivity in condensed phases, particularly when dealing with proteins and other biological macromolecules.[1] These models offer the advantage of combining the accuracy of ab initio methods, which can accurately describe chemical events, with the computational efficiency of classical force fields (FF). However, accurately modeling systems in condensed phases requires careful consideration of long-range electrostatic interactions. While classical simulations commonly employ Periodic Boundary Conditions (PBC) along with highly efficient algorithms to manage long-range electrostatic interactions, relatively few QM/MM models have been fully extended to PBC.
Here we introduced a novel PBC-adapted electrostatic embedding QM/MM model, compatibles with Self-Consistent Field (SCF) based ab initio methods,[2] that takes advantage of the Electrostatic Potential Fitted (ESPF) charge operators [3] and the smooth Particle Mesh Ewald (sPME) algorithm [4] to efficiently account for long-range electrostatics, achieving near-linear scaling with the size of the MM subsystem.
Derivatives with respect to QM and MM atomic coordinates, as well as lattice parameters, have been derived, enabling the propagation of QM/MM molecular dynamics under PBC in the most common thermodynamic ensembles.[5] The model’s effectiveness, along with a related approach compatible with the semi-empirical Density-Functional Tight-Binding (DFTB) method, has been validated through extensive studies of reaction free energies for a diverse set of chemical reactions.[6]
Overall, this work demonstrates that the proposed PBC-adapted QM/MM model offers an efficient and accurate approach for simulating chemical systems in condensed phases, extending the applicability of QM/MM simulations and paving the way for more detailed studies of complex chemical reactions and biological processes.
References:
[1] Senn, H. M.; Thiel, W. Angew. Chem. Int. Ed. 2009, 48, 1198–1229.
[2] Bonfrate, S.; Ferre, N.; Huix-Rotllant, M., J. Chem. Phys. 2023, 158 (2).
[3] Ferre, N.; Angyan, J. G. Chem. Phys. Lett. 2002, 356, 331–339.
[4] Essmann, U.; Perera, L.; Berkowitz, et al. J. Chem. Phys. 1995, 103, 8577–8593.
[5] Bonfrate, S.; Ferre, N.; Huix-Rotllant,M., J. Chem. Theory Comput. 2024, 20 (10), 4338–4349.
[6] Bonfrate, S.; Park,W.; Trejo-Zamora, D.; Ferre, N.; Choi, C. H.; Huix-Rotllant, M. ChemRxiv 2024.
