Macro-spin Modeling and Experimental Study of Spin-orbit Torque Biased Magnetic Sensors

TL;DR

This paper combines macro-spin modeling and experimental validation to develop spin-orbit torque biased magnetic sensors with tunable sensitivity, low power, and simple structure, suitable for on-chip applications.

Contribution

It introduces a systematic approach to design and optimize NiFe/Pt bilayer magnetic sensors using modeling and experiments, demonstrating high sensitivity and linearity.

Findings

Achieved a sensitivity up to 0.548 Ohm/Oe

Demonstrated linearity error below 6%

Validated sensor performance in on-chip current detection

Abstract

We reported a systematic study of spin-orbit torque biased magnetic sensors based on NiFe/Pt bilayers through both macro-spin modeling and experiments. The simulation results show that it is possible to achieve a linear sensor with a dynamic range of 0.1 - 10 Oe, power consumption of 1uW - 1 mW, and sensitivity of 0.1-0.5 Ohm/Oe. These characteristics can be controlled by varying the sensor dimension and current density in the Pt layer. The latter is in the range of 1 x 10^5 - 10^7 A/cm^2. Experimental results of fabricated sensors with selected sizes agree well with the simulation results. For a Wheatstone bridge sensor comprising of four sensing elements, a sensitivity up to 0.548 Ohm/Oe, linearity error below 6%, and detectivity of about 2.8 nT/Sqrt(Hz) were obtained. The simple structure and ultrathin thickness greatly facilitate the integration of these sensors for on-chip applications. As a proof-of-concept experiment, we demonstrate its application in detection of current flowing in an on-chip Cu wire.