Electronic: Spin molecular electronics
A very intriguing prospect is to generalize the thermoelectric concept to spin voltages and spin currents. This leads to the so-called spin Seebeck effect, discovered in 2008. The fact that thermoelectric effects are spin dependent has been known for a very long time but research in the area has been impeded due to experimental difficulties. Simply put and in analogy with conventional thermoelectric effects, a spin voltage can build up when the two spin channels have different scattering rates and densities, i.e., different Seebeck coefficients. Thus, the spin Seebeck effect is a manifestation of spin currents originating from a temperature gradient in a ferromagnet. There are two kinds of spin currents — spin polarized charge currents connected with electron motion, and magnon driven pure angular momentum currents. The spin Seebeck effect allows the transmission of a pure spin current over a long distance, and the spin Seebeck signal can be converted to a charge current through the inverse spin Hall effect in a nonmagnetic probe. A linear response theory of the magnon driven spin Seebeck effect in ferromagnetic insulators was recently developed.
Thermal gradients and the resulting spin Seebeck effect also has profound consequences in spin molecular electronics systems. Electronic devices are normally subjected to temperature gradients, and therefore it is vital to understand their effect on the transport properties. Furthermore, it can be linked to optics using lasers as heating source, allowing optical creation and control of spin polarization and spin currents, without the need for circularly polarized light or optical orientation. Spin caloritronic devices have recently been put forward taking advantage of the interaction between magnons and temperature gradients. In general, the interplay between molecular design and magnetism, electrode electronic structure and magnetism, and temperature gradients is virtually unexplored at this point. From a method development point of view, it can be challenged whether the usual LLG equaton is valid in the context of spin caloritronics. The temperature is usually introduced in the form of a stochastic magnetic field. It is an open question what the best way would be to perform this translation.