Numerical Optimization of Thermally Induced Hysteresis Effects in the Packaging of MEMS Pressure Sensors

Our research addresses time-dependent hysteresis effects in adhesive packaged MEMS pressure sensors. Typically calibrated inside a certain temperature and pressure range, they provide precise pressure measurements, given that certain settling times after temperature changes are maintained. Signal errors arise when temperature changes induce time-dependent viscoelastic relaxation in the adhesive which cannot be compensated by calibration. High-precision applications demand absolute signal accuracies below 30 Pa on chips well below 1 mm scales, while the requirement of factory calibration before soldering demands highly temperature stable adhesives. An experimentally verified, finite-element-based simulation model is used to investigate static and hysteretic stresses in the sensor, demonstrating a large potential for the reduction of temperature-dependent stresses affecting the signal, including hysteretic stresses. This is achieved by the determination of adhesive geometries that allow stress compensation, balancing opposing adverse stresses to cancel out. Using this approach, it is demonstrated that the signal hysteresis can be significantly reduced, while maintaining critical characteristics such as sensitivity and size.