Spin Orbit Effects for ultimately efficient spin dynamics of ultimately stable spin structures
-
30 April 2016 - updated 4 years ago -
14 comments
Fast, high density and low power microelectronic devices are crucial enablers of today’s IT technology. With semiconductor devices facing severe physical limitations for their performance in the future, spintronics technologies have been recently identified as the most likely technology for the next generation of non-volatile random access memory. This is due to the fact that magnetic technologies are inherently non-volatile and thus retain their information without power. However, current spintronic approaches based on magnetic bits made of “single domain” spin structures or “domain walls” (DWs) result in limited stability and an unacceptably high level of power consumption during operation due to the high currents and current densities required for manipulating the spins by spin transfer torque.
A radically new scientific and technological approach is necessary to tackle these key drawbacks, and obtain (i) small and stable spin structures as well as (ii) new mechanisms to efficiently manipulate these.
(i) Conventionally used single domain particles, which are stabilized by Heisenberg exchange, are intrinsically susceptible to thermal fluctuations as they can be reversed without changing the topology of the system and domain walls can be annihilated easily if they have random chiralities. More stable spin structures can be obtained based on spin-orbit interaction effects (SOI), by going beyond the commonly used Heisenberg exchange interaction, which dominates conventional spintronics. In systems without inversion symmetry, SOI leads to spin structures with topological protection that are stabilized by additional chiral exchange interactions such as the Dzyaloshinskii – Moriya interaction (DMI).These include chiral DWs and skyrmions, which are topologically distinct from the single domain state and thus topologically stabilized and can be just a few nm in diameter enabling high densities. Furthermore as displaced skyrmions do not change the overall magnetization, skyrmions are not susceptible to stray fields, enhancing the stability and their manipu¬lation is possible at ultra-low current densities due to the very weak interaction of skyrmions with defects.
(ii) So far the method of choice to manipulate magnetization is spin transfer torque, where for each electron one unit of spin angular momentum (“ħ”) is transferred when the electron turns its spin by 180° passing, for instance, across a DW. The transfer of orbital angular momentum can overcome this limit (>>1ħ/e-) and is thus potentially much more efficient. Such Spin Orbit Torques (SOTs) can then lead to fast magnetization switching at ultra-low current densities. Two mechanisms for SOTs have been identified in asymmetric systems: the Rashba-Edelstein effect (RE), where electric fields resulting from the asymmetry lead to effective magnetic fields that can manipulate the magnetization and the spin Hall effect (SHE) that converts a charge current into a spin current that can induce magnetization switching.
Together the highly stable skyrmion spin structures due to DMI and the efficient manipulation using SOTs can then be used to build novel devices, such as the skyrmion racetrack with a number of skyrmions beaded in a nanowire but these spin structures also hold prospect for logic and rf oscillators.
