机器学习绘图方法用于计算旋转放松动力学
Mohammad Shakiba1, Adam B Philips1, Jochen Autschbach1
1Department of Chemistry, University at Buffalo, The State University of New York, Buffalo, New York 14260, United States.
本研究介绍了一种机器学习方法,用于预测原子系统属性. 它使用比传统计算更少的数据准确预测电场梯度和旋转放松率.
科学领域:
- 计算化学和材料科学.
- 机器学习在量子力学中的应用.
- 开发用于预测分子属性的方法.
背景情况:
- 预测原子系统的特性往往需要计算上昂贵的高级理论计算.
- 电场梯度 (EFG) 对于理解核特性和旋转放松至关重要.
- 计算EFG和旋转放松率的现有方法可能耗时.
研究的目的:
- 开发和验证一种机器学习 (ML) 方法来预测电场梯度 (EFG) 张量.
- 使用ML预测的EFG来计算水溶液中离子的自旋放松率.
- 在数据要求方面证明ML方法的效率.
主要方法:
- 采用机器学习映射方法,使用原子轨道重叠,密度或Kohn-Sham (KS) Fock矩阵元素从扩展紧密结合作为输入特征.
- 这些特征被用来预测EFG张量,这些张量通常在较高的理论水平 (例如混合函数) 得到.
- 然后,预测的EFG张量被用来计算各种离子的四极异极旋转放松率.
主要成果:
- 该ML方法成功地预测了高准确度的EFG张量.
- 预测的EFG张量器能够准确计算水溶液中的几个离子的自旋放松率.
- 该方法在使用直接计算所需数据的仅一小部分的情况下,对于旋转放松率实现了良好的准确性 (2-8%的相对误差).
结论:
- 机器学习为预测EFG张量和旋转放松率提供了一个高效而准确的替代方案.
- 与传统方法相比,这种方法显著降低了计算成本和数据需求.
- 开发的ML模型有望加速材料发现和理解分子动力学.
更多相关视频
09:25Author Spotlight: Exploring Intrinsically Disordered Protein Dynamics Through NMR Relaxation Experiments
Published on: November 1, 2024
07:24Paramagnetic Relaxation Enhancement for Detecting and Characterizing Self-Associations of Intrinsically Disordered Proteins
Published on: September 23, 2021
相关概念视频
Atomic Nuclei: Types of Nuclear Relaxation
In spin–lattice or longitudinal relaxation, the excited spins exchange energy with the surrounding lattice as they return to the lower energy level. Among several mechanisms that contribute to spin–lattice relaxation, magnetic dipolar interactions are significant. Here, the excited nucleus transfers...
Atomic Nuclei: Nuclear Relaxation Processes
Atomic Nuclei: Nuclear Spin State Overview
Spin–Spin Coupling Constant: Overview
Qualitatively, any spin plus-half nucleus polarizes the spins of its electrons to the minus-half state. Consequently, the paired electron in the hydrogen–carbon bond must...
¹H NMR: Interpreting Distorted and Overlapping Signals
As Δν decreases and the signals move closer, the doublets appear increasingly distorted. The intensities of the inner lines increase at the cost of those of the outer lines as the signals are...
Atomic Nuclei: Nuclear Spin State Population Distribution
![Measuring the Spin-Lattice Relaxation Magnetic Field Dependence of Hyperpolarized [1-13C]pyruvate](/_next/image?url=https%3A%2F%2Fcloudfront.jove.com%2FCDNSource%2Fteasers%2F59399.jpg&w=3840&q=50)