构建基于低级数据集的激发状态动力学模拟的高精度机器学习潜在能量表面
Shuai Li1, Bin-Bin Xie2, Bo-Wen Yin2
1Key Laboratory of Theoretical and Computational Photochemistry of Ministry of Education, College of Chemistry, Beijing Normal University, Beijing 100875, P. R. China.
The journal of physical chemistry. A
|July 2, 2024
概括
机器学习准确地预测了分子潜在能量,加速了光化学反应模拟. 神经网络擅长激发状态动态,尽管对Z-azobenzene等特定异构体的准确性仍然是一个挑战.
科学领域:
- 计算化学计算化学
- 摄影化学的使用.
- 机器学习 机器学习
背景情况:
- 机器学习 (ML) 可以预测分子潜在能量,帮助模拟光化学反应.
- 高质量的数据集至关重要,但在计算上昂贵,用于激发状态计算.
- 这种瓶阻碍了非adiabatic分子动力学模拟的加速.
研究的目的:
- 使用ML构建以乙烯为桥梁的亚博的地面和兴奋状态潜在能量表面 (PES).
- 评估不同ML模型在预测动态模拟PES中的性能.
- 执行激发状态动力学模拟并将结果与高级计算进行比较.
主要方法:
- 生成了两个数据集:一个来自OM2/MRCI表面跳跃,另一个来自之前的CASSCF轨迹.
- 在CASSCF级别使用OM2/MRCI数据集构建了地面和激发状态的PES.
- 训练并测试了各种ML模型,包括神经网络 (BPNN,SchNet) 与Δ-ML和力项.
主要成果:
- 只有神经网络模型 (BPNN,SchNet) 在外部测试数据上提供了可靠的预测.
- 在ML-PES上的激发状态动态模拟显示了与E异构体S1生命周期的高水平估计的良好一致.
- 对Z异构体寿命的高估仍然存在,这表明训练准确性不能保证高水平的预测准确性.
结论:
- 在动态模拟中,建议用于ML-PES的神经网络,特别是具有特定方案的BPNN和SchNet.
- ML加速了对光化学反应的研究,例如阿佐烯光异构化.
- 为了在所有异构体和动态情景中准确预测,需要进一步改进.
更多相关视频
08:04Excitonic Hamiltonians for Calculating Optical Absorption Spectra and Optoelectronic Properties of Molecular Aggregates and Solids
Published on: May 27, 2020
8.4K
08:54Vibrational Spectra of a N719-Chromophore/Titania Interface from Empirical-Potential Molecular-Dynamics Simulation, Solvated by a Room Temperature Ionic Liquid
Published on: January 25, 2020
5.7K
相关概念视频
02:34Arrhenius Plots
39.2K
The Arrhenius equation relates the activation energy and the rate constant, k, for chemical reactions. In the Arrhenius equation, k = Ae−Ea/RT, R is the ideal gas constant, which has a value of 8.314 J/mol·K, T is the temperature on the kelvin scale, Ea is the activation energy in J/mole, e is the constant 2.7183, and A is a constant called the frequency factor, which is related to the frequency of collisions and the orientation of the reacting molecules.
The Arrhenius equation can be used...
The Arrhenius equation can be used...
39.2K
02:45The Quantum-Mechanical Model of an Atom
42.2K
Shortly after de Broglie published his ideas that the electron in a hydrogen atom could be better thought of as being a circular standing wave instead of a particle moving in quantized circular orbits, Erwin Schrödinger extended de Broglie’s work by deriving what is now known as the Schrödinger equation. When Schrödinger applied his equation to hydrogen-like atoms, he was able to reproduce Bohr’s expression for the energy and, thus, the Rydberg formula governing hydrogen spectra.
42.2K