Automated Parameterization of Quantum Mechanically Derived Force Fields for Soft Materials and Complex Fluids: Development and Validation

• DOI: 10.1021/acs.jctc.1c00213
• Publication Date: 2021-06-29T20:43:53Z
• Authors: J. G. Vilhena; Leandro Greff da ...; Paolo Roberto Liv...; Ivo Cacelli; Giacomo Prampolini
• Keywords: procedure; Chemical Sciences not elsewhere classified; Information Systems not elsewhere classified; Biophysics; macroscopic properties; 01 natural sciences; Mathematical Sciences not elsewhere classified; Lengthy atomistic MD simulations; Space Science; 5 CB systems; QMD; phase; molecular dynamics (MD) simulations; fluid; Quantum Mechanically Derived Force .; QM; QMD-FF; ab initio quantum mechanical (QM) data; FF parameterization protocol; fragmentation reconstruction method; 541; 0104 chemical sciences; ab initio quantum; intramolecular FF term; Medicine; adopted force field (FF); FRM; Physical Sciences not elsewhere classified; Biotechnology; Biological Sciences not elsewhere classified

The reliability of molecular dynamics (MD) simulations in predicting macroscopic properties complex fluids and soft materials, such as liquid crystals, colloidal suspensions, or polymers, relies on the accuracy adopted force field (FF). We present an automated protocol to derive specific accurate FFs, fully based ab initio quantum mechanical (QM) data. integration Joyce Picky procedures, recently proposed by our group provide description simple liquids, is here extended larger molecules, capable exhibiting more fluid phases. While standard employed parameterize intramolecular FF term, a new procedure handle computational cost QM calculations required for parameterization intermolecular term. latter thus obtained integrating old with fragmentation reconstruction method (FRM) that allows reliable, yet computationally feasible sampling energy surface at level. whole tested benchmark crystal, performances resulting mechanically derived (QMD) were compared those delivered general-purpose, transferable one, third, "hybrid" FF, where only bonded terms refined against Lengthy atomistic MD are carried out each 5CB systems both isotropic nematic phases, eventually validating comparing other models experiments. QMD-FF yields best performances, reproducing phases correct range temperatures well describing their structure, dynamics, thermodynamic properties, providing clear may be explored predict materials.