Spin manipulation in semiconductor spin devices

这个项目的主要目标是展示通过电操纵铁磁源极和漏极接触之间的沟道中的自旋来控制横向半导体器件的电阻。这将构成1990年设想的Datta-Das自旋场效应晶体管（sFET）概念的完全实现。在我们以前的工作中，我们已经实现了sFET的单个构建块，即我们已经显示出非常有效的电自旋注入到半导体二维电子气（2DEG）中，并且我们已经显示出在这样的设备中可以观察到高达80%的高磁阻信号。因此，我们拥有充分实现“达塔-达斯”概念的所有要素。为此，我们计划集中讨论两个主题。首先，我们将探讨准一维（1D）通道中的电自旋注入和自旋输运。这种通道虽然只有一种运输方式，但在最初的Datta-Das提案中被援引。我们的设备构成了一个完美的测试床，以调查预期的抑制Dyakonov-Perel自旋弛豫准1D通道，电注入面内自旋。这已经研究了面外自旋之前使用光学方法。其次，我们将直接解决sFET的关键功能，即，在横向局部器件中，通过自旋轨道耦合（SOC）对自旋进动进行电控制，其中自旋极化电流在源极和漏极接触之间流动。除了Datta和Das提出的原始几何形状外，在沟道顶部有一个金属栅极，我们计划在这里探索使用横向自旋轨道耦合的可能性，源于1D沟道中的横向限制，以诱导和控制自旋进动。

The main goal of this project is to demonstrate control of the resistance of a lateral semiconductor device by electrical manipulation of the spins in the channel between ferromagnetic source and drain contacts. This will constitute then a full realization of the concept of Datta-Das spin field effect transistor (sFET), envisioned in 1990. In our previous work, we have realized single building blocks of the sFET, namely we have shown very efficient electrical spin injection into a semiconductor two-dimensional electron gas (2DEG) and we have shown that high magnetoresistance signals up to 80% can be observed in such devices. We have therefore all ingredients at hand to pursue the full realization of the Datta-Das concept. To this end, we plan to focus on two topics. Firstly, we are going to explore electrical spin injection and spin transport in quasi one-dimensional (1D) channels. Such a channel, although with only one transport mode, was invoked in the original Datta-Das proposal. Our devices constitute a perfect test bed to investigate the expected suppression of Dyakonov-Perel spin relaxation in quasi 1D channels, for electrically injected in-plane spins. This has been studied before for out-of-plane spins using optical methods. Secondly, we are going to directly address the crucial functionality of the sFET, i.e., electrical control of the spin precession via spin orbit coupling (SOC), in lateral local devices, with a spin-polarized current flowing between source and drain contacts. Apart from the original geometry proposed by Datta and Das, with a metal gate on top of the channel, we plan to explore here also the possibility of using lateral spin orbit coupling, originating from the lateral confinement in 1D channels, to induce and control spin precession.