Kubečka Jakub, Ayoubi Daniel, Tang Zeyuan, Knattrup Yosef, Engsvang Morten, Wu Haide, Elm Jonas
Department of Chemistry, Aarhus University Langelandsgade 140 8000 Aarhus C Denmark
Center for Interstellar Catalysis, Department of Physics and Astronomy, Aarhus University Ny Munkegade 120 8000 Aarhus C Denmark.
Env Sci Adv. 2024 Aug 13;3(10):1438-1451. doi: 10.1039/d4va00255e. eCollection 2024 Oct 2.
The computational cost of accurate quantum chemistry (QC) calculations of large molecular systems can often be unbearably high. Machine learning offers a lower computational cost compared to QC methods while maintaining their accuracy. In this study, we employ the polarizable atom interaction neural network (PaiNN) architecture to train and model the potential energy surface of molecular clusters relevant to atmospheric new particle formation, such as sulfuric acid-ammonia clusters. We compare the differences between PaiNN and previous kernel ridge regression modeling for the Clusteromics I-V data sets. We showcase three models capable of predicting electronic binding energies and interatomic forces with mean absolute errors of <0.3 kcal mol and <0.2 kcal mol Å, respectively. Furthermore, we demonstrate that the error of the modeled properties remains below the chemical accuracy of 1 kcal mol even for clusters vastly larger than those in the training database (up to (HSO)(NH) clusters, containing 30 molecules). Consequently, we emphasize the potential applications of these models for faster and more thorough configurational sampling and for boosting molecular dynamics studies of large atmospheric molecular clusters.
对于大分子系统进行精确的量子化学（QC）计算，其计算成本往往高得令人难以承受。与QC方法相比，机器学习在保持准确性的同时，计算成本更低。在本研究中，我们采用可极化原子相互作用神经网络（PaiNN）架构来训练与大气新粒子形成相关的分子簇（如硫酸 - 氨簇）的势能面并进行建模。我们比较了PaiNN与之前针对Clusteromics I - V数据集的核岭回归建模之间的差异。我们展示了三种能够预测电子结合能和原子间力的模型，其平均绝对误差分别小于0.3 kcal/mol和小于0.2 kcal/(mol Å)。此外，我们证明，即使对于比训练数据库中的簇大得多的簇（多达包含30个分子的(HSO)(NH)簇），建模属性的误差仍保持在1 kcal/mol的化学精度以下。因此，我们强调这些模型在更快、更全面的构型采样以及促进对大型大气分子簇的分子动力学研究方面的潜在应用。
