SPP 1538: Spin Caloric Transport (SpinCaT)
Chemistry
Final Report Abstract
Spin polarized currents in magnetic nanostructures give rise to novel spin caloric effects. These effects modify thermal transport, magneto-resistance and possibly even magnetic states. To understand the observed effects an extension of thermodynamic laws including the spin is inevitable. In 2009 the German community has detected an internationally increased interest in this topic which combines magnetism with both, electronics and energy research. With the existing expertise, Germany was in an excellent position to proactively take the lead in this hot topic. The priority programme “Spin Caloric Transport (SpinCaT)” has pursued the investigation of the spin degree of freedom in thermoelectric transport phenomena, motivated by completely unexplored new functionalities as well as the potential to enhance performance of existing technologies. It is long known that heat and charge currents are closely linked to each other by, e.g., the Wiedemann-Franz law which connects thermal and electrical conductivities. The coupling leads to thermoelectric effects with prominent examples being the Seebeck effect and its inverse, the Peltier effect, which may be described through the Onsager relations. The effects are used in devices such as thermocouples and thermo-electric cooling devices, respectively. It became clear only recently that these long-known thermoelectric phenomena need to be re-considered by including the spin, generating completely new spin-related properties in the solid state. The hypothesis that in magnetic systems spin entropy can be created, manipulated and transported by non-equilibrium charge and heat currents is compelling. Non-equilibrium magneto-caloric transport effects such as the Nernst-Ettingshausen effects face novel spin and anomalous counterparts. In the SPP spin transport phenomena have been studied in the presence of thermal gradients. It has been the right time to address fundamental issues: Given, that spin and metal-based devices (logic gates, memory, spin-based interconnects) will be used for information processing, the deep understanding of heat-related phenomena needed for efficient heat flow management is of paramount importance. Assuming that information transport between devices will be realized using the spin channels in the future, the intimate interplay between spin-, charge- and in particular heat currents are of particular relevance. The aim of the priority programme “Spin Caloric Transport (SpinCaT)” was to develop the new research field of caloric effects in spin transport. The research programme has focussed on four priority areas: - Spin caloric effects and spin mediated heat transport in planar geometry - Thermal conductivities across interfaces in nanopatterned devices - Spin currents induced by large temperature gradients - Materials for spin caloric applications The initiative has opened a wide and, in particular, multi-disciplinary field of research bringing together groups working in the broader fields of magneto-electronics, material sciences and thermodynamics, and originating from the disciplines of solid-state physics, chemistry and, potentially, electrical engineering. The field of spin-caloric transport addresses the interplay between spin-transport phenomena and thermal properties. In particular, the separation of up- and down-spins (generating a spin voltage) via temperature gradients and, vice versa, the manipulation of temperature and entropy by spins (i.e., spin voltages) is a focus of interest. A central theme lies in the interaction between spin-caloric and spin-dynamic effects, important in particular for applications in information and communication technology. With progress in modern nanoscale technologies heat management will become more and more an issue; Peltier cooling effects are already used on the level of semiconductor devices. This will become even more a challenge in spintronic applications which are expected to complement conventional complementary metal oxide semiconductor (CMOS) technology increasingly, such as in the fields of magnetic random access memory (MRAM), magnetic logics and microwave oscillators. For example, in spin momentum transfer oscillators (STO) and spin momentum transfer memory cells (ST-MRAM), large current densities are applied to contacts of sub 100 nm dimensions. This leads to drastic heating of the material with temperatures reaching the thermal migration limit. Therefore, efficient on-spot cooling in spintronic devices is a challenging issue. This could be achieved employing spin-dependent thermoelectric effects, thus the combination of efficient spin, charge and heat transport. The envisioned improvement of Peltier cooling via heat transport through spin excitations is one paradigm of the emerging field of spin caloric transport. Thermoelectric effects are well established and used for efficient cooling or power generation, but little is known about the role of the electron spin in thermoelectric transport. Spin caloric effects are subject to Onsager reciprocity, meaning that inverse effects have to be studied as well, e.g., alteration of the magnetic state of a material by application of heat gradients is associated with thermoelectric effects caused by magnetization dynamics. In fact, a pioneering experimental example of spin dependent thermoelectric effects is the recent discovery of the spin-Seebeck effect - where temperature gradients have been used to induce spin voltages. Efficient local heating of the electronic system can be achieved experimentally by, e.g., application of strong femtosecond laser pulses. A thorough study of spin and thermal phenomena in potential spintronic materials involves the investigation of charge and spin currents as well as of entropy (heat) currents and their interplay. This avenue has so far not been followed by a larger research consortium and would pave the way for new technologically relevant spintronic devices. We have clarified inconsistent evidence from different experiments in the literature, by showing that the so-called transversal spin-Seebeck effect in metals is unobservably small, when compared to the competing (established) magnetothermoelectric effects. We carried out theoretical research on spin caloritronic phenomena with emphasis on the magnetic insulator yttrium iron garnet using a combination of analytic and numerical techniques. In collaboration with experimental groups in the Priority Program we helped discover new effects such as the spin Hall magnetoresistance. We successfully transferred central spintronic effects to highest (i.e. Terahertz) frequencies, including the spin-dependent/spin Seebeck effect to generate spin currents and the inverse spin Hall effect to convert these spin currents into charge currents. We took advantage of these effects to build a spintronic emitter of Terahertz electromagnetic pulses that cover the full range from 1 to 30 THz with an efficiency comparable or even better than standard Terahertz sources. Moreover, we revealed the elementary steps leading to the formation of the spin-Seebeck effect in archetypal YIG|Pt° bilayers whose ultrafast rise is determined by the thermalization dynamics of the optically excited electrons in Pt. We demonstrated in a joint experimental and theoretical work that the spin Seebeck effect can generate spin currents in the bulk of an insulating ferrimagnet. By varying materials and interfaces, we identified different magnon modes that contribute to the spin transport generated by a heat current experimentally. Theoretical calculations analyzed the effect of thermal spin currents on the displacement of domain walls. Topology has conquered the field of condensed matter physics with the discovery of the quantum Hall effect. Since then the zoo of topological materials is steadily increasing. In this project, we discovered how to realize different topological phases with magnons: the magnon pendants to topological insulators as well as Weyl and nodal-line semimetals are presented. Similar to the electronic case, nonzero Berry curvature causes transverse transport, that is, magnon Hall effects. We developed a method to show how these effects can be quantified by classical spin dynamics simulations.
