Leaky Surface-Acoustic-Wave resonators on lithium niobate: analysis, design, and applications to wireless temperature sensing​

Micro-electro-mechanical systems (MEMS) are playing an increasingly pivotal role in manifold applications, fostering advancements in key technologies such as 5G and the Internet of Things (IoT). In particular, piezoelectric resonators are extensively adopted as building blocks of both sensor and RF filters due to their high quality factor and small footprint. In this thesis, we illustrate the design, fabrication, and measurement of Leaky Surface Acoustic Wave (LSAW) resonators that leverage the shear-horizontal acoustic mode present in the YZ 0° X-cut lithium niobate substrate. Through numerical simulations, we highlight the features that make them attractive especially for sensing applications, and optimize their electrode structure in order to maximize their main figures of merit: the electromechanical coupling and the quality factor. We then show how they can be fabricated through a simple and cost-effective process involving just one mask. Finally, we characterize the devices experimentally by measuring their scattering parameters by using a probe station equipped with a Vector Network Analyzer, and compare the measured response with simulations to validate our design. The variation of the resonance frequency as a function of temperature, encapsulated in the temperature coefficient of frequency (TCF), is also evaluated experimentally. Piezoelectric resonators are usually modeled through the Butterworth-Van Dyke (BVD) equivalent circuit, whose parameters are retrieved by fitting the analytical expression of the BVD admittance to measured or simulated data; however, the presence of spurious acoustic modes makes the simple BVD model inaccurate and complicates the fitting algorithm. To address this issue, well-established system identification techniques implemented through Matlab are adopted, allowing the numerical evaluation of the frequency response of a resonator from its measured S-parameters. The resulting black-box model of the resonator can then be easily implemented in a circuit or system level simulator. We conclude this work by illustrating potential architectures for the realization of wireless and passive temperature sensors, envisioning the monolithic integration of LSAW resonators and antennas on the same substrate to reduce size and cost compared to standard PCB solutions.