自旋轨道转矩的理论与计算研究

Current-induced torque is the key element in spintronics to excite or switch magnetization and is applied for information storage and manipulation in many proposed magnetic devices. Recently, new types of torques driven by an electrical current have been found in the systems containing heavy transition metal elements and can be used to efficiently tune magnetization dynamics. These kinds of torques are believed to result from the relativistic spin-orbit interaction in heavy elements, but the different physical mechanisms that entangle with one another can hardly be separated in experiment. The present theoretical studies in terms of phenomenological models or symmetry analysis, however, can not provide comprehensive and consistent explanation for the reported experimental observations. The incomplete understanding in the current experimental and theoretical researches makes it very difficult to optimize the proposed devices. ..We propose to develop a theoretical method of calculating spin-orbit torques in real magnetic materials with high reliability, predictability and capability of dealing with large diffusive systems, based upon first-principles spin transport without tunable parameters. In this method, we are able to directly describe the real materials at the atomistic level, including alloy disorder, lattice mismatch at interfaces, temperature-induced atomic vibration and magnetization fluctuation. The method can be applicable in two typical magnetic structures, i.e., ferromagnet|nonmagnetic bilayers and magnetic domain walls, where we attempt to calculate the spin-orbit torques quantitatively and their dependence on the interface effects, temperature, impurities, etc. The purpose is to gain a deep understanding of the nature and competing mechanisms of spin-orbit torques in these structures, so that we are able to provide guidance for future experimental studies and development of devices based on the efficient spin-orbit torques.

电流诱导转矩在自旋电子学研究中被用于有效地激发和翻转磁矩。近年来实验发现在包含较重过渡金属材料中，存在新型电流转矩，能够高效率地调制磁化动力学。该转矩被普遍认为源自材料中的强自旋轨道相互作用，但由于其非单一、相互纠缠的物理机制在实验上难以分离，同时现有理论使用唯象模型或者对称性分析均难以全面、自洽地解释实验观测。目前实验和理论研究的困境也使得技术上没有成熟的依据来进行器件的优化。.我们计划发展完善一套可靠的，能够处理较大扩散系统的微观输运理论方法，使用不依赖可调参数的第一原理计算磁性材料中的转矩，对真实材料体系进行原子尺度的直观模拟；应用该方法系统地研究铁磁-非磁双层膜与磁畴壁两类典型磁性体系中自旋轨道转矩，以及对界面、温度和杂质等因素的依赖性，深入理解产生自旋轨道转矩的各种物理机制之间的竞争关系，为该领域的实验研究以及基于电流转矩的器件优化提供参考。

自旋轨道耦合产生的自旋轨道力矩是磁动力学及相关研究中的一个关键问题，所产生的自旋轨道力矩能够驱动磁矩进动，诱导磁化翻转和驱动磁畴壁移动。理论上定量描述自旋轨道力矩以及分辨其微观机制对于优化器件性能是极为重要的。本项目发展了一套基于第一性原理的自旋输运程序，能够计算出自旋霍尔，界面Rashba劈裂等产生的各类自旋轨道力矩效应，为自旋电子学的理论研究和与实验的结合提供了重要的理论方法和工具。应用这套计算程序，研究了多种块体和界面输运性质，包括5d金属的自旋扩散长度和自旋霍尔角，自旋霍尔电导率的温度依赖性，界面自旋记忆损失，界面自旋霍尔效应等实验研究中的关键材料参数，以及磁畴壁中的非共线磁化带来的自旋轨道力矩效应等。在项目进行过程中中进一步把研究内容拓展到与自旋轨道力矩研究相关的反铁磁动力学中的阻尼力矩，各向异性磁电阻的微观机制，激光诱导的超快退磁和太赫兹发射等方面，取得了一系列研究进展，得到了国内外同行的广泛关注和积极评价。累积发表论文24篇，其中物理评论快报3篇，其余物理评论系列杂志15篇。多次受邀在中国物理学会秋季会议，国际磁学会议等做邀请报告。