Answers are grounded in this paper and cite their sources.
As now practiced in experimental nanomagnetic spintronics, spin-transfer torque acting on a free metallic moment is driven by electric current flowing serially through it and a metallic reference magnet. I propose driving spin-transfer torque by flow of heat serially through the free magnet and an insulating reference ferrite. The needed spin current initiates from magnons present in the ferrite. A quantum yield of heat-driven in-plane spin-transfer torque can be substantially greater, in principle, than that achievable using electric current in a magnetic tunnel junction. A Bloch-type dynamical equation for the conduction-electron-spin polarization excited by a paramagnetic-monolayer model of the ferrite/metal interface predicts the dependence of this yield on material parameters. In practice, achieving a high yield beneficial to applications will require strong exchange coupling of the local $3d$-electron atomic spins in this monolayer to both to the ferrite moment with ferromagnetic sign and also with either sign to the conduction s electrons in a normal metallic spacer. Advantageous will be suppression of the interfacial heat flow diverted to phonons within the ferrite. If a nonmagnetic electrically insulating layer additionally adjoins the free magnet, the theory also predicts a perpendicular spin-transfer torque component whose angular dependence mimics conventional uniaxial magnetic anisotropy.
No takes yet. Share an insight, caveat, or question.
J. C. Slonczewski (2010) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: