Sensitivity optimization of micro-machined thermo-resistive flow-rate sensors on silicon substrates

Abstract: We report on an optimized micro-machined thermal flow-rate sensor as part of an autonomous multi-parameter sensing device for water network monitoring. The sensor has been optimized under the following constraints: low power consumption and high sensitivity, while employing a large thermal conductivity substrate, namely silicon. The resulting device consists of a platinum resistive heater deposited on a thin silicon pillar ~ 100 $\mu$m high and 5 $\mu$m wide in the middle of a nearly 100 $\mu$m wide cavity. Operated under the anemometric scheme, the reported sensor shows a larger sensitivity in the velocity range up to 1 m/s compared to different sensors based on similar high conductivity substrates such as bulk silicon or silicon membrane with a power consumption of 44 mW. Obtained performances are assessed with both CFD simulation and experimental characterization.