Optimizing effective anisotropy in magnetic tunnel junctions for operation at cryogenic temperatures
Phys. Rev. Applied 24, 024016 – Published 7 August, 2025
DOI: https://doi.org/10.1103/62tn-vnzb
Abstract
The development of working cryogenic memories is very important for the scalability and feasibility of quantum computing. In this context, /-based spin-transfer-torque magnetic random-access memories emerge as promising candidates to address this challenge. At low temperatures, the reduction in thermal noise allows for a lower energy barrier while retaining sufficient memory performance. Since the effective anisotropy of the system is directly proportional to the energy barrier, understanding how it varies with material parameters and temperature is crucial. In this work, we present an experimental method for extrapolating the thin-film magnetic properties measured at room temperature to the electrical properties of nanofabricated devices at various targeted temperatures. We employ magneto-optical Kerr-effect (MOKE) magnetometry to calculate the effective anisotropy energy density across the entire 100-mm wafer surface at room temperature. This method is entirely adapted to probe wafers of various sizes, and it is found to be faster and equivalent to conventional methods, such as vibrating sample magnetometry and superconducting quantum interference device (SQUID) magnetometry. The temperature evolution of magnetic anisotropy is further investigated using SQUID magnetometry from 300 to 10 K. Finally, the critical switching voltage from 300 to 40 K is measured for various nanofabricated devices with different storage-layer thicknesses. A model correlating the switching voltage of these devices with the effective anisotropy at the thin-film level is used to explain their temperature and thickness dependencies. These results confirm the validity of the MOKE method presented in this work and its potential in selecting devices across the wafer to function at a given operational target temperature.
