Department of Chemistry, James Franck Institute, and Institute for Biophysical Dynamics, The University of Chicago, Chicago, Illinois 60637, United States.
J Phys Chem A. 2022 Sep 8;126(35):6004-6019. doi: 10.1021/acs.jpca.2c04349. Epub 2022 Aug 25.
Feynman's imaginary time path integral formalism of quantum statistical mechanics and the corresponding quantum-classical isomorphism provide a tangible way of incorporating nuclear quantum effect (NQE) in the simulation of condensed matter systems using well-developed classical simulation techniques. Our previous work has presented the many-body coarse-graining of path integral (CG-PI) theory that builds an isomorphism between the quantum partition function of distinguishable particles and the classical partition function of 2 pseudoparticles. In this present work, we develop a generalized version of the many-body CG-PI theory that incorporates many-body interactions in the force field. Based on the new derivation, we provide a numerical CG-PI (n-CG-PI) modeling strategy parametrized from the underlying path integral molecular dynamics (PIMD) trajectories using force matching and Boltzmann inversion. The n-CG-PI models for two liquid systems are shown to capture well both the intramolecular and intermolecular structural correlations of the reference PIMD simulations. The generalized derivation of the many-body CG-PI theory and the n-CG-PI model presented in this work extend the scope of the CG-PI formalism by generalizing the previously limited theory to incorporate force fields of realistic molecular systems.
费曼的量子统计力学虚时路径积分形式和相应的量子-经典同构为使用成熟的经典模拟技术模拟凝聚态系统时纳入核量子效应（NQE）提供了一种切实可行的方法。我们之前的工作提出了路径积分的多体粗粒化（CG-PI）理论，该理论在可区分粒子的量子配分函数和 2 个赝粒子的经典配分函数之间建立了同构。在本工作中，我们开发了一种多体 CG-PI 理论的广义版本，该理论将多体相互作用纳入力场中。基于新的推导，我们提供了一种基于潜在路径积分分子动力学（PIMD）轨迹的数值 CG-PI（n-CG-PI）建模策略，使用力匹配和玻尔兹曼反演进行参数化。对于两个液体系统的 n-CG-PI 模型，证明能够很好地捕捉参考 PIMD 模拟的分子内和分子间结构相关性。本工作中提出的多体 CG-PI 理论的广义推导和 n-CG-PI 模型扩展了 CG-PI 形式的范围，通过将之前有限的理论推广到包含实际分子系统的力场，将其一般化。
