Zhang Changwei, Zhong Yang, Tao Zhi-Guo, Qin Xinming, Shang Honghui, Lan Zhenggang, Prezhdo Oleg V, Gong Xin-Gao, Chu Weibin, Xiang Hongjun
Key Laboratory of Computational Physical Sciences (Ministry of Education), Institute of Computational Physical Sciences, State Key Laboratory of Surface Physics, and Department of Physics, Fudan University, Shanghai, 200433, China.
Key Laboratory of Precision and Intelligent Chemistry, Hefei National Research Center for Physical Sciences at the Microscale, University of Science and Technology of China, Hefei, Anhui, 230026, China.
Nat Commun. 2025 Feb 27;16(1):2033. doi: 10.1038/s41467-025-57328-1.
Non-adiabatic molecular dynamics (NAMD) simulations have become an indispensable tool for investigating excited-state dynamics in solids. In this work, we propose a general framework, NAMD (Neural-Network Non-Adiabatic Molecular Dynamics), which employs an E(3)-equivariant deep neural Hamiltonian to boost the accuracy and efficiency of NAMD simulations. Distinct from conventional machine learning methods that predict key quantities in NAMD, NAMD computes these quantities directly with a deep neural Hamiltonian, ensuring excellent accuracy, efficiency, and consistency. NAMD not only achieves impressive efficiency in performing NAMD simulations at the hybrid functional level within the framework of the classical path approximation (CPA), but also demonstrates great potential in predicting non-adiabatic coupling vectors and suggests a method to go beyond CPA. Furthermore, NAMD demonstrates excellent generalizability and enables seamless integration with advanced NAMD techniques and infrastructures. Taking several extensively investigated semiconductors as the prototypical system, we successfully simulate carrier recombination in both pristine and defective systems at large scales where conventional NAMD often significantly underestimates or even qualitatively incorrectly predicts lifetimes. This framework offers a reliable and efficient approach for conducting accurate NAMD simulations across various condensed materials.
非绝热分子动力学（NAMD）模拟已成为研究固体中激发态动力学不可或缺的工具。在这项工作中，我们提出了一个通用框架，即NAMD（神经网络非绝热分子动力学），它采用E(3)等变深度神经哈密顿量来提高NAMD模拟的准确性和效率。与预测NAMD中关键量的传统机器学习方法不同，NAMD通过深度神经哈密顿量直接计算这些量，确保了出色的准确性、效率和一致性。NAMD不仅在经典路径近似（CPA）框架内的混合泛函水平上执行NAMD模拟时实现了令人印象深刻的效率，还在预测非绝热耦合矢量方面展现出巨大潜力，并提出了一种超越CPA的方法。此外，NAMD具有出色的通用性，能够与先进的NAMD技术和基础设施无缝集成。以几种经过广泛研究的半导体作为典型系统，我们成功地在大尺度上模拟了原始和缺陷系统中的载流子复合，而传统的NAMD通常会显著低估甚至定性地错误预测寿命。该框架为在各种凝聚态材料中进行准确的NAMD模拟提供了一种可靠且高效的方法。
