Comprehensive noise characterisation of magnetic tunnel junction sensors for optimising sensor performance and temperature detection

Abstract Noise performance of magnetic tunnel junction (MTJ) sensors is impacted by various factors including junction structure, post-deposition treatment, and operating parameters. The optimisation of these factors can lead to a better MTJ sensor design with minimised noise level and enhanced detectivity for functioning as a magnetometer. In this paper, the authors studied the influence of several parameters (bias voltage, temperature, magnetic field, and junction area) on the noise performance of MTJ sensors. Relatively high bias voltage and low ambient temperature were suggested to be helpful in reducing the electronic 1/f noise. A mechanism of utilising MTJ as a temperature sensor by making use of the mid-frequency noise (from 10.0 kHz to 22.8 kHz) was proposed. The relation between temperature and noise power was obtained by numerically fitting the measured noise power with an equation composing of three components representing background noise, intertwined thermal and shot noise, and non-linear noise source, respectively. Temperature of the junction could be determined by measuring the mid-frequency noise power at certain bias voltage and substituting it into the equation. This provides a possible route of using a MTJ as a multifunctional sensor for sensing both magnetic field and temperature.