Study of the spin Nernst effect: absolute moments and dynamics

我们提出研究和量化在具有有限自旋-轨道相互作用的材料中由热流（自旋能效应）产生的自旋积累。实验将使用两种互补的方法，一种超灵敏，一种超快速。在一组实验中，将使用微型SQUID技术直接测量金属棒边缘自旋能效应引起的绝对磁矩。棒材受到由电阻加热或光学加热引起的热梯度。利用这种超灵敏的技术（精确到几百个电子自旋），我们计划直接量化SNE的大小。此外，适当设计SQUID几何结构将使我们能够直接获得自旋扩散长度。在与理论的密切合作下，该技术将应用于各种金属（Cu, Au），包括杂质掺杂，以便在分离内在和外在贡献的同时最大化自旋能效应。为了研究自旋能效应的动力学，在一组互补的实验中，将使用超快泵浦探针技术产生非常大的热梯度。太赫兹领结天线结构将被100飞秒单周期太赫兹瞬变激发，峰值振幅为1 MV/cm。通过适当的设计，天线中的太赫兹电流可以限制在亚微米大小的耗散分流，允许超快速电阻加热。通过这种方式，我们期望达到超过1k /nm的热梯度。这些大的温度梯度反过来又会驱动具有自旋-轨道相互作用的金属结构中的热流，其中由于自旋能效应产生自旋电流。同时，时间分辨磁光响应将用于研究相应的自旋极化动力学（空间分辨率为200nm）。由于加热脉冲是在亚皮秒时间尺度上施加的，这些实验允许实时研究自旋能效应产生的自旋积累的时间演化和衰减。

We propose to investigate and quantify the spin accumulation that is generated by heat currents (spin Nernst effect) in a material with finite spin-orbit interaction. The experiments will use two complementary approaches, one ultra-sensitive and one ultra-fast. In one set of experiments micro SQUID technology will be used to directly measure the absolute magnetic moments that are induced by the spin Nernst effect on the edges of a metallic bar. The bar is subjected to a thermal gradient which is either induced by resistive or by optical heating. With this ultra-sensitive technique (down to a few hundred electron spins) we plan to directly quantify the magnitude of the SNE. In addition, suitable design of the SQUID geometry will allow us to directly access the spin diffusion length. In close collaboration with theory this technique will be applied to various metals (Cu, Au) including doping by impurities in order to maximize the spin Nernst effect while separating intrinsic and extrinsic contributions.In a complementary set of experiments very large thermal gradients will be generated using an ultrafast pump-probe technique in order to study the dynamics of the spin Nernst effect. THz bow-tie antenna structures will be excited by 100 femtosecond single cycle THz-transients with a peak amplitude of 1 MV/cm. By proper design the THz-current in the antenna can be confined to a sub-micron sized dissipative shunt allowing for ultra-fast resistive heating. In this way we expect to reach thermal gradients in excess of 1 K/nm. These large temperature gradients will in turn drive heat currents in metallic structures with spin-orbit interaction where due to spin Nernst effect spin currents are generated. Simultaneously the time resolved magneto optic response will be used to study the dynamics of the corresponding spin polarization (with a spatial resolution of 200 nm). Since the heating pulses are applied on a sub-picosecond time scale these experiments allow a real time study of the time evolution and the decay of the spin accumulation generated by the spin Nernst effect.