Doctor of Philosophy

dissertationA wireless inductive sensing technology with a robust packaging scheme is proposed and demonstrated for a hydrogel-based glucose sensing system. The sensing system employs a glucose-sensitive hydrogel embedding a parylene-coated metallic sensing plate as a key sensing element. A glucose change induces a volume change of the hydrogel, thus displacing the sensing plate. The displacement can be sensed by a nearby sensing coil inductor interfaced with an oscillator for a frequency modulation (FM) scheme. Two prototype wireless and battery-less implantable glucose sensing systems are designed and implemented using the proposed sensing scheme. The first prototype employs a tunnel diode oscillator-based inductive sensor operating at approximately 40 MHz. The system is housed inside a 10 mm- , 7 mm-height 3D-printed package for in-vitro characterizations. The system is wirelessly powered over a distance of 2 cm and achieves a sensing resolution of 0.05 mM and 0.3 mM within a glucose range from 0 mM to 2 mM and 8 mM to 10 mM, respectively. The system response exhibits a temperature dependency, which can potentially limit the system sensing accuracy. To reduce the system size and enable in-situ system temperature characterization, a second prototype based on an application-specific integrated circuit (ASIC) is proposed. The system employs an integrated oscillator operating at approximately 200 MHz. The AISC is fabricated in XFab 0.6 μm CMOS process exhibiting an area of 1.1 mm x 1 mm with a power consumption of approximately 600 μW. A 6 mm x 6 mm x 25 mm 3D-printed package is employed to encapsulate the sensing iv electronics for in-vitro characterization. A rotation-insensitive wireless powering is proposed to enable a reliable wireless and battery-less operation. The system achieves a sensing resolution of 0.06 mM over a typical physiological glucose concentration. The system exhibits an exponential settling behavior with a time constant of approximately 20 minutes. The system temperature sensitivity is characterized using an on-chip temperature sensor. Various error sources that can potentially degrade the system sensing accuracy are investigated and compared. The prototype system using the 3D-printed package is tested in-vitro for over 60 days without noticeable performance degradation