| Inertial MEMS (accelerometers and gyroscopes) is a rapidly-growing billion dollar industry. At the heart of these devices is a displacement sensor. Since its commercialization in the 1980s, the core technology has not changed (viz., capacitive displacement readout of mass-on-springs), for almost all commercially-available inertial MEMS. However, recent developments in integrated optomechanics when combined with recent low-cost on-chip lasers and detectors may enable high-SNR on-chip displacement sensing. Such devices represent a new paradigm in on-chip inertial MEMS sensors, but have yet to be considered in detail in the literature.;Beyond simply investigating different sensing schemes, we find that reinventing the traditional displacement-sensing element has the effect of reinventing the entire system. For example the driving circuitry may be simpler and/or lower-power than traditional inertial MEMS driving circuitry, the noise sources are fundamentally different and are limited by different mechanisms, the footprint and cost drivers may be completely reimagined, etc.;Although we have not yet integrated the devices reported here with on-chip lasers and detectors, we show experimental results and modeling for our non-integrated devices, discuss the noise sources to be expected in an integrated device, and survey some on-chip laser/detector noise figures from the literature. Using such noise figures and the measured optomechanical sensitivities, we show that our measured devices when operated as accelerometers could easily achieve sub-microg[square root of] Hz total noise, and thus potentially exceed typical electrostatic MEMS accelerometer performance. Though we have applied these findings primarily to accelerometers, these findings may also be applied to MEMS gyroscopes, or other sensors that involve displacement sensing, such as MEMS microphones, MEMS pressure sensors, and so forth.;In this dissertation we quantitatively investigate several optomechanical displacement sensing schemes, both theoretically and experimentally, and discuss the merits of each approach. These schemes include: cavity deformation sensing, cavity evanescent field displacement sensing (both waveguide or nearby absorber moving), and two-cavity gap sensing. |
