Uncovering the Missing Physics in the Metrology of Spin-Orbit Torques

Cornell University, Ithaca NY

Investigators

Abstract

Non-Technical Abstract Magnetic devices offer a combination of virtues for computer memory that no other technology can match – they can retain information with no applied power, they can withstand an unlimited number of writing and reading operations without wearing out, and they can be made fast and high-density. However, widespread applications in electronics will require finding a way to write information to magnetic memories with lower power. Recently, a promising new mechanism has been discovered for controlling magnetic memories very efficiently, known as “spin-orbit torque,” but there is a problem that different experimental methods used to measure the strength of this mechanism often give inconsistent values. This project is investigating what is the missing science that has not been properly taken into account, causing these conflicting results. The practical aim of the research is to enable trustworthy measurements of spin-orbit torques. This will provide the scientific foundation to optimize the next generation of magnetic memory technologies, with the goal that they will enable improved performance and lower energy consumption for applications ranging from machine learning to low-power internet-of-things networks. This project trains graduate and undergraduate students in advanced device fabrication, measurement techniques, and computer modeling along with science communication and other professional skills. Graduates typically find employment in research laboratories of electronics hardware companies. Participants in the project are also active in public outreach programs, in particular a partnership between the Cornell Nanofabrication Facility and 4-H clubs. Technical Abstract Recent advances in understanding the interactions between charge currents, spin currents, and magnets have led to the development of magnetic-memory technologies in which the orientation of magnets is efficiently controlled by torques exerted from spin currents. However, this field faces a fundamental-science puzzle because different experimental techniques used to measure spin-orbit torques (the most-efficient known mechanism for current-driven magnetic manipulation) often give contradictory results. This indicates that the intellectual framework used to analyze these measurements is missing essential physics. This project is performing experiments to test whether the excitation of short-wavelength magnons, heating, nonlinear transport effects, spin currents emitted by ferromagnets, or other yet-to-be recognized effects might explain this missing physics. The ultimate project goal is to establish trustworthy measurement techniques for use in the development of a new generation of magnetic memory devices with improved performance and lower energy consumption, for applications ranging from machine learning to low-power internet-of-things networks. This project trains graduate and undergraduate students in advanced device fabrication, measurement techniques, and computer modeling along with science communication and other professional skills. Participants in the project are also active in public outreach programs, in particular a partnership between the Cornell Nanofabrication Facility and 4-H clubs. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.