Thermal modulation of microfabricated cantilever-based chemical sensors for improved selectivity

Chemical sensors;Cantilevers;Gas-phase;Selectivity;MEMS;Microelectromechanical systems;Sensors;Volatile organic compounds;VOC;Resonator

Getz, Patrick Thomas ; Brand, Oliver Electrical and Computer Engineering Degertekin, Levent Sarioglu, Fatih Ansari, Azadeh Sulchek, Todd ; Brand, Oliver

University:Georgia Institute of Technology

Department:Electrical and Computer Engineering

关键词: Chemical sensors;    Cantilevers;    Gas-phase;    Selectivity;    MEMS;    Microelectromechanical systems;    Sensors;    Volatile organic compounds;    VOC;    Resonator;   

Others  :  https://smartech.gatech.edu/bitstream/1853/61127/1/GETZ-DISSERTATION-2018.pdf

美国|英语

来源: SMARTech Repository

【 摘 要 】

This work investigates the integration of heaters on a polymer-coated, microfabricated mass-sensitive resonant cantilever for gas-phase chemical detection of volatile organic compounds. Integrated diffused resistors modulate the temperature of a polymer film, and the analysis of sorption kinetics at elevated temperatures is used to discriminate between similar analytes. Multiple polymers are tested as an absorbing layer to determine suitability for analyte identification with this new technique. Heating pulses of various amplitudes, applied to the diffused resistors, raise the film temperature by up to 30 °C. As one metric for analyte identification, measuring the steady-state frequency shift of these mass sensors as a function of temperature can improve the selectivity of the polymer-coated sensors by extracting characteristic analyte properties that aid in analyte discrimination. This analysis successfully estimates the vaporization enthalpies of the tested analytes with an average error of 1% for one polymer coating. Additionally, several methods are explored to increase the selectivity of the sensor via analyzing the transient response of analyte absorption and desorption at different temperatures and heating schemes. Both the steady state and transient responses are used in supervised machine learning techniques to differentiate between similar analytes. Pursuant to these tasks, this work also demonstrates improved frequency stability, in the presence of environmental variations, through decreases in the Allan deviations, thus, improving the limits of detection for the polymer-coated mass-sensitive sensors (665 ppb for benzene, 158 ppb for toluene and 41 ppb for o-xylene).