Spins are traditionally controlled by magnetic fields, which complicates the design of electronic devices exploiting the spin degree of freedom of the electrons. Through spin-transfer torque, an intense spin-polarized charge current can also act on the magnetization and thus enable an electrical control of such spintronics devices, albeit with an excessive power consumption. Unlike currents applied to conductors, electric fields applied across an insulator do not consume power, which has motivated the search for efficient means to control magnetism by an electric field.
This area of research took off with the "renaissance" of multiferroic materials, compounds in which magnetic order coexists with ferroelectricity. Both orders are usually linked through a magnetoelectric coupling, such that the magnetic properties can be acted upon by applying an external voltage. Room-temperature single-phase multiferroics are very rare - with the notable exception of BiFeO3 - but magnetoelectric coupling can be engineered in "artificial" multiferroics, for instance bilayers combining a ferroelectric and a ferromagnet. Most multiferroics and ferroelectrics are oxide materials and the field was largely seized by the oxide electronics community.
Practically, the electric-field control of magnetism can proceed through three main mechanisms: intrinsic magnetoelectric coupling, strain-mediated coupling and field-effect-driven coupling. In this talk, I will show examples on the manipulation of various magnetic properties (magnetization direction, Curie temperature, magnetic order, spin polarization, etc) by electric fields in oxide heterostructures. I will also present developing alternatives to electrically generate spin-currents and convert them into charge-currents through spin-orbit effects at oxide interfaces or at the surface of topological insulators. Finally, I will discuss new directions for the field, emerging from the interplay between correlations, spin-orbit coupling and topology.
