Computational design and modeling of topological insulator-based heterostructures for spin-orbitronics and skyrmionics

自旋电子学探索由电子携带的自旋和电荷相互交织的现象，已经展示了一种非凡的能力，可以在为未来应用的基础研究萌芽肥沃的亚领域的方向上重新给自己充电。第一代自旋电子学导致了存储在硬盘驱动器上的数字信息量的革命性增长，也关键地依赖于新材料和异质结构的发现。下一代自旋电子学有望带来可与传统电子学集成的超低功耗存储器和逻辑器件。这项计划中的项目将探索最近发现的拓扑绝缘体(TI)材料中被称为自旋-轨道耦合(SOC)的特殊相对论效应所带来的这种进步的新来源。TiS在体积上是绝缘的，但也是具有强大SOC的金属表面的宿主。最近的实验表明，当电流被注入到异质结构中时，其中TI连接到铁磁(FM)层，FM层的磁化动力学可以被点燃，并且潜在地比目前在磁性随机存取存储器基础上可用的技术中的耗散要小得多。具有较强SOC和破缺反转对称性的TI层和FM层之间的界面也可以产生旋涡自旋织构，其特征是纳米尺寸、对缺陷和杂质的拓扑稳定性以及类似于带电粒子在磁场中的回转动力学。使用高性能计算(HPC)模拟来寻找针对这些现象的TI和FM材料的最佳组合可以显著缩短生产功能器件所需的时间。拟议研究的更广泛影响将包括对研究生进行非平衡量子统计力学方面的培训，先进的科学计算技术，以及对非平衡条件下基于TI的异质结构的基本理解。学生们将与国际合作者进行互动。利用非平衡格林函数理论(NEGF)、非共线密度泛函理论(DFT)和半经典朗之万方程技术，本研究程序将对以下问题有基本的认识：SO扭矩；自旋泵浦和自旋到电荷的转换；以及TIS带来的强界面SOC对磁化动力学的Gilbert衰减和噪声效应。建议的研究将从筛选TI/FM异质结构的第一性原理开始，以确定那些对界面周围的磁原子或界面Dzyaloshinsky-Moriya相互作用具有最大SOC邻近效应的异质结构，这些相互作用可能在室温下产生Skyrmions(所有目前已知的磁性Skyrmions的例子都发生在室温以下)。在第二阶段，将计算最有希望的异质结的扭矩。SO力矩将被用作随机Landau-Lifshitz-Gilbert(LLG)方程的输入，以研究存在快速运动(量子力学描述)电子时(经典描述的)磁化或Skyrmion自旋织构的慢动力学，包括由它们产生的阻尼和非平衡噪声效应。这项研究将为下一代基于拓扑绝缘体的超低功耗存储器和逻辑自旋电子器件奠定基础。

Spintronics, which explores phenomena intertwining spin and charge carried by an electron, has exhibited a remarkable ability to re-energize itself in directions that germinate fertile subfields for basic research aimed at future applications. The first generation spintronics, which has led to revolutionary increase in the amount of digital information stored on hard drives, has also crucially relied on the discovery of new materials and heterostructures. The next generation spintronics is expected to lead to ultralow power dissipation memory and logic devices that can be integrated with conventional electronics. The projects in this proposal will explore emerging resource for such advances brought about by the special relativistic effects, termed spin-orbit coupling (SOC), in recently discovered topological insulator (TI) materials. The TIs are insulating in their bulk, but also host metallic surfaces with strong SOC. The very recent experiments have demonstrated that when current is injected into heterostructures where TI is attached to a ferromagnetic (FM) layer, magnetization dynamics of the FM layer can be ignited with potentially much less dissipation than in presently available technologies underlying magnetic random access memory. The interface between TI and FM layers with strong SOC and broken inversion symmetry could also generate swirling spin texture characterized by nano-scale size, topological stability against defects and impurities, and gyro-dynamics analogous to that of a charged particle under magnetic field. Using high performance computing (HPC) simulations to search for optimal combination of TI and FM materials for these phenomena can significantly shorten time needed to produce functional devices. Broader impact of the proposed research will, include training for graduate students in nonequilibrium quantum statistical mechanics, advanced scientific computing techniques and fundamental understanding of TI-based heterostructures under nonequilibrium conditions. Students will interact with the international collaborators. New modeling software and computational design for high-density data-storage and nonvolatile memory with ultra-low energy cost manipulation developed under the program will be available for researchers in the field.Using combination of nonequilibrium Green function theory (NEGF), noncollinear density functional theory (DFT), and semiclassical Langevin equation techniques, this research program will develop fundamental understanding of: SO torque; spin pumping and spin-to-charge conversion; and Gilbert damping and noise effects on magnetization dynamics in the presence of strong interfacial SOC brought by TIs. The proposed research will commence with first principles screening of TI/FM heterostructures in order to identify those with the largest SOC proximity effect onto the magnetic atoms around the interface or interfacial Dzyaloshinsky-Moriya interaction that could give rise to skyrmions at room temperature (all presently known examples of magnetic skyrmions occur below room temperature). In the second stage, SO torque for most promising heterostructures will be computed. The SO torque will be used as an input for the stochastic Landau-Lifshitz-Gilbert (LLG) equation to study slow dynamics of (classically described) magnetization or skyrmion spin texture in the presence of fast moving (quantum-mechanically described) electrons, including damping and nonequilibrium noise effects generated by them. This research will form the basis for next generation of ultralow power dissipation memory and logic spintronic devices based on topological insulators.