Optimization of magnetoresistive sensor current for on-chip magnetic bead detection using the sensor self-field

Abstract

We investigate the self-heating of magnetoresistive sensors used for measurements on magnetic beads in magnetic biosensors. The signal from magnetic beads magnetized by the field due to the sensor bias current is proportional to the bias current squared. Therefore, we aim to maximize the bias current while limiting the sensor self-heating. We systematically characterize and model the Joule heating of magnetoresistive sensors with different sensor geometries and stack compositions. The sensor heating is determined using the increase of the sensor resistance as function of the bias current. The measured temperature increase is in good agreement with a finite element model and a simple analytical thermal model. The heat conductance of our system is limited by the μ 1 m thick electrically insulating silicon dioxide layer between the sensor stack and the underlying silicon wafer, thus the heat conductance is proportional to the sensor area and inversely proportional to the oxide thickness. This simple heat conductance determines the relationship between bias current and sensor temperature, and we show that μ 2 5m wide sensor on a μ 1 m oxide can sustain a bias current of 30 mA for an allowed temperature increase of 5 °C. The method and models used are generally applicable for thin film sensor systems. Further, the consequences for biosensor applications of the present sensor designs and the impact on future sensor designs are discussed.