Employing piezojunction effect for resonant micro-device applications

This dissertation reports on the application of the piezojunction effect as a new mechanism for measurement of resonance frequency in silicon-based micro-systems. It has been known that mechanical stress can affect the electrical characteristics of diodes, transistors and electronic circuits. This phenomenon is recognized as piezojunction effect. To explore the piezojunction effect as an effective detection mechanism in a micro-device, a micro-resonator structure is designed to employ the inherent mechanical amplification of displacements at resonance as an enabling platform. The proposed structure is capable of mechanically amplifying the sensing signal at the resonance frequency by its quality factor, which can be in the range of tens to hundreds of thousands. This amplified displacement makes the piezojunction effect a practical sensing method for resonator applications. In this technique, the sensing current is stemmed from the dependency of electrical characteristics of an embedded p-n junction to the periodic stress profiles inside the resonating body. The p-n junction is reverse-biased, therefore, due to low sensing current, the required power for detection of resonance is rather small. To employ the piezojunction technique, the author has developed an in-house Silicon-On-Insulator (SOI) micro-machining process to fabricate the proof-of-concept micro-resonator and its embedded p-n junction. Fabricated resonators were packaged and experimentally tested to verify the feasibility of the design and to gauge the performance of the piezojunction mechanism for resonance sensing. The static and dynamic responses of the fabricated devices are experimentally verified. The extensional-mode frequency of the resonator was measured to be 7MHz with a mechanical quality factor of around 5,000. The required power consumption for this sensing mechanism was as low as 5nW. The experimental verifications demonstrate that the piezojunction effect is a promising addition to existing detection techniques in resonance-based applications, where small chip area, integration, and power consumption are key requirements.