Ultra-energy-efficient magnetic memory by controlling the shapes of atoms

Non-volatile magnetic memory using nanometer-sized magnets, MRAM (magnetoresistive random access memory), required magnetization reversal by applying voltage. Thus, ultra-energy efficient magnetization reversal in nanoseconds is preferable. However, figure-of-merit of the current technology, voltage-controlled magnetic anisotropy (VCMA), was less than one-tenth of the level necessary for application. It is important to develop the VCMA effect using the new materials.

Associate Professor Shinji MIWA at Osaka University, Dr. Motohiro SUZUKI at the Japan Synchrotron Radiation Research Institute, Assistant Professor Masahito TSUJIKAWA at Tohoku University, Dr. Takayuki NOZAKI at the National Institute of Advanced Industrial Science and Technology, and Dr. Tadakatsu OHKUBO at the National Institute for Materials Science, made a platinum monatomic layer placed on ferromagnetic iron (FePt|MgO system) that was controlled at an atomic level.

Since there is correlation between a spin-orbit interaction and a VCMA, this group focused on FePt|MgO, which contains platinum with large spin-orbit interactions. Using the FePt|MgO, this group conducted experiments in order to clarify the VCMA at X-ray beamlines in the SPring-8 synchrotron radiation facility.

From these experiments and theoretical calculations, this group discovered that the FePt|MgO system that demonstrated a VCMA of 140 fJ/Vm had two different mechanisms and potentially possesses an enormous VCMA beyond 1,000 fJ/Vm.

This group observed changes in magnetic dipole term by voltage in the experiment at SPring-8. From theoretical calculations, it was found that, in the FePt|MgO system, the VCMAs from conventionally known Mechanism A (the orbital magnetic moment inductino) and newly discovered Mechanism B (magnetic dipole term inductino) were partially cancelled out by one another, resulting in a VCMA of 140 fJ/Vm.

The use of this group's achievements in designing materials will make it possible to obtain a VCMA of 10 times larger than that of existing materials, which will allow for energy-saving non-volatile memory that can reduce heat generation.