Energy-efficient phase-locked arrays of spin torque nano-oscillators based on current-induced torques in magnetic metals

能源效率和计算速度的进一步发展需要新的革命性方法，如脑启发（神经形态）信号处理。基于传统半导体组件的神经形态设备是耗电的。因此，神经形态信号处理需要新类型的组件。其中一个很有前途的元件是自旋力矩纳米振荡器，它由磁性和非磁性金属层组成，电流在非磁性层中流动，在磁性层中产生微波电信号。该研究项目的目标是通过演示一种通过在磁性层而不是非磁性层中流动的电流产生微波信号的新方法来显着提高自旋扭矩纳米振荡器的能量效率。预计所提出的方法将使自旋扭矩纳米振荡器的功耗降低90%以上，从而使计算机技术的进步成为可能。研究生和本科生，包括来自代表性不足的群体的人，将接受设备纳米制造和表征技术的培训。这将有助于培训重新确立美国在半导体芯片开发和制造方面的领导地位所需的合格劳动力。PI将参加每年两次的中学外展活动。作为活动的一部分，中学生将参观PI的实验室，并参与磁性、超导、真空技术和等离子体物理的实践实验。自旋-轨道力矩是几种有前途的自旋电子技术的核心，包括非易失性磁存储器、用于非布尔信号处理的自旋力矩纳米振荡器和超灵敏微波探测器。对自旋轨道力矩的基本理解对于这些器件的发展至关重要。本项目将研究一类新的自旋-轨道力矩-铁磁金属中的普适自生自旋-轨道力矩。这些转矩是自生的，因为它们被施加到产生它们的铁磁体的磁化，并且是普遍的，因为它们的存在预计在所有类型的铁磁金属中，而与它们的晶体结构无关。的扭矩将被测量的自旋扭矩铁磁共振纳米线和纳米十字为基础的结构，使用电子束光刻从磁控溅射沉积的外延和多晶磁性多层膜。 两种主要类型的自生转矩，反常霍尔转矩和平面霍尔转矩，将进行研究，并将确定最大限度地提高这些转矩的大小为技术应用的铁磁系统。为了说明新扭矩的变革潜力，将展示高能效的自旋扭矩纳米振荡器锁相阵列。将评估这些设备用于神经形态信号处理的潜力。这种自旋力矩振荡器阵列在零磁场中的操作将在具有界面和应变诱导的磁各向异性的结构中演示。由通用自生自旋轨道扭矩驱动的自旋扭矩纳米振荡器的能效将与由拓扑绝缘体和Weyl半金属的自旋轨道扭矩驱动的自旋扭矩纳米振荡器的能效进行比较。该奖项反映了NSF的法定使命，并通过使用基金会的智力价值和更广泛的影响审查标准进行评估，被认为值得支持。

Further advances in energy-efficiency and speed of computing require new revolutionary approaches such as brain-inspired (neuromorphic) signal processing. Neuromorphic devices based on conventional semiconductor components are power-hungry. Therefore, new types of components are needed for neuromorphic signal processing. One such promising component is a spin torque nano-oscillator that consists of a magnetic and a non-magnetic metallic layer, electric current flowing in the non-magnetic layer generates microwave electric signal in the magnetic layer. The goal of this research project is to significantly improve energy efficiency of spin torque nano-oscillator via demonstration of a new method of microwave signal generation by electric current that flows in the magnetic layer rather than the non-magnetic layer. It is expected that the proposed method will reduce power consumption of spin torque nano-oscillators by over 90 percent thereby enabling the advancement in computer technology. Graduate and undergraduate students, including persons from underrepresented groups, will be trained in device nanofabrication and characterization techniques. This will contribute to training of the qualified workforce needed for reestablishing the United States leadership in semiconductor chip development and manufacturing. The PI will participate in middle school outreach activity held twice a year. As a part of this activity, middle school students will visit the PI’s lab and participate in hands-on experiments on magnetism, superconductivity, vacuum technology and plasma physics.Spin-orbit torques are at the core of several promising spintronic technologies, including non-volatile magnetic memory, spin torque nano-oscillators for non-Boolean signal processing and ultrasensitive microwave detectors. Fundamental understanding of spin-orbit torques is crucial for the development of these devices. In this project, a new class of spin-orbit torques – universal self-generated spin-orbit torques in ferromagnetic metals will be studied. These torques are self-generated because they are applied to magnetization of the ferromagnet in which they are produced, and are universal because their presence is expected in all types of ferromagnetic metals independent of their crystal structure. The torques will be measured by spin-torque ferromagnetic resonance in nanowire- and nanocross-based structures prepared using e-beam lithography from epitaxial and polycrystalline magnetic multilayers deposited by magnetron sputtering. Two major types of self-generated torques, anomalous Hall torque and planar Hall torque, will be studied, and ferromagnetic systems maximizing the magnitude of these torques for technological applications will be determined. To illustrate the transformative potential of the novel torques, highly energy-efficient phase-locked arrays of spin torque nano-oscillators will be demonstrated. Potential of these devices for neuromorphic signal processing will be evaluated. Operation of such spin torque oscillator arrays in zero magnetic field will be demonstrated in structures with interfacial and strain-induced magnetic anisotropy. Energy efficiency of spin torque nano-oscillators driven by universal self-generated spin-orbit torques will be compared to that of spin torque nano-oscillators driven by spin-orbit torques from topological insulators and Weyl semimetals.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.