Ultrafast spin-dependent and spin Seebeck effect: beyond diffusive spin transport, toward a spin-caloritronic terahertz emitter

纵向自旋型塞贝克效应，即自旋电流与温度梯度反平行的驱动效应，是自旋热电子学中的一个关键现象。在铁磁性金属中，它被称为自旋相关的塞贝克效应，因为它是由多数自旋电子和少数自旋电子的不同电导率引起的。在铁磁绝缘体中，它被称为自旋塞贝克效应，是由磁振子引起的。到目前为止，这两种效应大多是在准静态条件下或扩散状态下研究的。这样的条件模糊了基本过程，对需要非常快的自旋电流变化的应用提供了很少的见解。在这个提议中，我们将研究超快时间尺度上的自旋依赖效应和自旋塞贝克效应。为此，用飞秒激光脉冲照射铁磁体（如金属铁、钴、镍或绝缘的Y3Fe5O12）将产生强烈的热梯度。我们的光学方法允许简单的样品几何形状，即平面薄膜或多层，而不需要电接触或微结构。在产生热密度梯度之后，将通过(i)利用磁光和热光效应的飞秒探针脉冲和（ii）检测由时变自旋和电荷电流发出的太赫兹电磁脉冲来监测由此产生的自旋、电荷和热输运。由于我们的时间分辨率与携带自旋角动量的电子（在d型金属中为~10fs）和磁振子（在绝缘体中为~10ps至~100ns）的速度弛豫时间相当，我们将直接了解基本输运步骤。例如，在绝缘体中，声子热和磁振子自旋输运之间的相关性有多强？当我们驱动光学声子而不是电子时会发生什么？在金属中，当我们逐渐将泵浦光子能量从~50meV增加到~1.5eV，从而将初始局部电子分布从几乎费米-狄拉克型调整为高度非平衡态时，自旋电流是如何演变的？能否用自旋阀控制超快自旋输运？最后，产生具有大瞬态峰值的自旋电流的最佳样品结构是什么？通过使用这种强烈的自旋电流爆发与逆自旋霍尔效应相结合，我们将建立一个有效的自旋热电子太赫兹电磁辐射源，特别是覆盖到目前为止仍然难以捉摸的5到10THz的频率间隙。

The longitudinal spin-type Seebeck effect, the driving of a spin current antiparallel to a temperature gradient, is a key phenomenon in spin caloritronics. In ferromagnetic metals, it is termed the spin-dependent Seebeck effect as it arises from the different conductivity of majority- and minority-spin electrons. In ferromagnetic insulators, it is called the spin Seebeck effect and is due to magnons. So far, both effects have mostly been investigated under quasi-static conditions or in the diffusive regime. Such conditions blur elementary processes and provide little insight into applications where very fast spin-current variations are required.In this proposal, we will study both the spin-dependent and the spin Seebeck effect on ultrafast time scales. For this purpose, strong heat gradients will be induced by illuminating a ferromagnet (such as metallic Fe, Co, Ni or insulating Y3Fe5O12) with femtosecond laser pulses. Our optical approach permits simple sample geometries, that is, plane thin films or multilayers, without the need for electrical contacts or micro-structuring. Following generation of the heat-density gradient, the resulting spin, charge, and heat transport will be monitored by (i) a femtosecond probe pulse taking advantage of magneto-optic and thermo-optic effects and by (ii) detecting the terahertz electromagnetic pulse emitted by the time-varying spin and charge currents. As our time resolution is comparable to the velocity relaxation time of the electrons (~10fs in d-type metals) and magnons (~10ps to ~100ns in insulators) carrying the spin angular momentum, we will gain direct insight into elementary transport steps. For example, in insulators, how strongly are phonon heat and magnon spin transport correlated with each other? What happens when we drive optical phonons instead of electrons? In metals, how does the spin current evolve when we gradually increase the pump photon energy from ~50meV to ~1.5eV, thereby tuning the initial local electron distribution from almost Fermi-Dirac-type to highly non-equilibrium-like? Can we control the ultrafast spin transport with spin valves? Finally, what is the optimum sample structure to generate spin currents with large transient peak magnitudes? By using such intense spin-current bursts in conjunction with the inverse spin Hall effect, we will build an efficient spin-caloritronic source of terahertz electromagnetic radiation that in particular covers the so far still elusive frequency gap from 5 to 10THz.