Smart Platform for Low-Cost MEMS Sensors – Pressure, Flow and Thermal Conductivity

In a technological world that is trending towards smart and autonomous engineering, the collection of quality data is of unrivalled importance. This has led to a huge market demand for the development of low-cost, small and accurate sensors and thus has resulted in significant research into sensors, with the aim of advancing the price/performance ratio in commercial solutions. Micro Electro Mechanical Systems (MEMS) have recently offered an attractive solution to miniaturise and drastically improve the performance of sensors. In this thesis, MEMS technology is exploited to create a multi-sensor technology platform that is used to fabricate several sensing technologies. Piezo-resistive and piezo-electronic pressure sensors are designed, fabricated and tested. Different doping profiles, stress-engineered structures and electronic devices for pressure transduction are investigated, with focus on their sensitivity and non-linearity. A ring is fabricated in the metal layer around the circumference of the membrane that alleviates the effects of over/under etching. This is achieved by creating a new rigid edge of the membrane in the metal layer, which has tighter fabrication tolerances. A piezo-MOSFET is developed and shown to have greater sensitivity than similar state-of-the-art devices. Flow sensors based on a heated tungsten wire are designed, fabricated, tested and substantiated with numerical modelling. Calorimetric and anemometric driving modes are optimised with regards to device structure. Thermodiodes are also used as the temperature transduction devices and are compared to the traditional resistor method and showed to be preferable when further miniaturising the sensor. Thermal conductivity gas sensors based on a heated tungsten resistor are designed, tested and substantiated with numerical modelling. Holes through the membrane are used to improve the sensitivity to measuring carbon dioxide by 270%. Asymmetric holes are utilised to prove a novel method of measuring thermal conductivity in a calorimetric method. Designs improving this new concept are outlined and substantiated with analytical and numerical models. Linear statistical methods and artificial neural networks are used to differentiate flow rate and gas concentration using three on-membrane resistors. With membrane holes, the discrimination between gases in the presence of flow is improved. Neural networks provide a viable solution and show an increase in the accuracy of both flow rate and gas concentration. The main objective of the work in this thesis was to develop low-cost, low-power, small devices capable of high-volume production and monolithic integration using a single smart technology platform for fabrication. The smart technology platform was used to create pressure sensors, flow sensors and thermal conductivity gas sensors. Within each sensing technology, proof-of-concepts and optimisations have been carried out in order to maximise performance whilst using the low-cost, high-volume fabrication process, ultimately helping towards smart and autonomous engineering solutions driven by data