*********************************
There is now a CONTENT FREEZE for Mercury while we switch to a new platform. It began on Friday, March 10 at 6pm and will end on Wednesday, March 15 at noon. No new content can be created during this time, but all material in the system as of the beginning of the freeze will be migrated to the new platform, including users and groups. Functionally the new site is identical to the old one. webteam@gatech.edu
*********************************
Title: Development of A MEMS-Based Resonant and Impedimetric Multivariate Sensor for the Speciation and Quantification of Volatile Organic Compounds
Committee:
Dr. Brand, Advisor
Dr. Hao, Chair
Dr. Ansari
Abstract: The objective of the proposed reserach is to further develop a novel MEMS-based micro gas sensor which attempts to overcome the issues facing real-world application of gas sensing systems. To properly address this challenge, it is important to enumerate the drawbacks of modern gas sensor systems which typically leverage univariate sensor arrays to ultimately improve sensor sensitivity, selectivity, and stability issues. The challenges faced by such univariate arrays include uncorrelated sensor-to-sensor drift, inability to quantify gas concentrations given the presence of mixtures, and the degradation of prediction accuracy due to system drift or interferents without a means for stable correction. This work attempts to solve these challenges at the sensor element level using a newly designed resonant and impedimetric MEMS-based multisensor. This multivariate sensor leverages three simultaneous sensor outputs produced by two collocated transduction mechanisms of different physical domains (i.e. mechanical and electrical) to drastically enhance sensor selectivity, while preserving the high sensitivity of resonant-based gravimetric sensors. The multivariate sensor does not require the use of principal component analysis, thus reducing instability in sensor performance due to uncorrelated sensor-to-sensor drift if this sensor were to be arrayed. Additional efforts to improve sensor stability were undertaken at the device-level through the application of silicon-based the thermal compensation techniques yielding a x15 reduction in the thermal coefficient of frequency. The gas classification ability of the sensor was demonstrated with five volatile organic compounds of the BEXT and alcohol subclasses when steady state signals were used. Suggested future work includes the use of machine learning to utilize non-equilibrium sensor data to improve gas classification and concentration quantification capabilities of both unary and binary mixtures while improving sensor response times. Additionally, a device limitation study further characterizing effects of pressure, humidity, and temperature on device operation would enable improved performance in future designs.
