Titlebar

Export bibliographic data
Literature by the same author
plus on the publication server
plus at Google Scholar

 

Microwave acoustic sensors : From chemical to electrical signals

Title data

Fischerauer, Gerhard:
Microwave acoustic sensors : From chemical to electrical signals.
2005
Event: SSD 2005 , 21.-25.03.2005 , Sousse, Tunisia.
(Conference item: Conference , Speech )

Abstract in another language

Mass-sensitive chemical sensors generate electrical signals such as frequency changes in response to the selective sorption of molecules. The sensors tend to respond slowly owing to the fact that the chemical species to be bound in the selective sensor film diffuses into the film with characteristic diffusion coefficients on the order of only 10^(-17) m²/s. This contribution presents a model-based sensor signal processing strategy capable of predicting the steady-state response soon after an analyte concentration change, thereby reducing the effective response time by as much as an order of magnitude. To be more specific, the sensor is characterized by an analytic model which is valid for all mass-sensitive sensors the dynamic behavior of which is dominated by diffusion effects. The impulse response turns out to be proportional to the Jacobian theta function theta2(0,exp(-pi²*t/t0)) where t0 is a characteristic time. Based on this result, we proceed to derive a formula that expresses the sensor input (the time-dependent analyte concentration) in terms of the sensor output (the time-dependent frequency shift) in closed form. The model is validated by way of measured sensor characteristics and by comparing its predictions to known numerical results. A signal processing algorithm based on the model is presented and discussed. Finally, the closed-form solution is applied to the steady-state prediction problem. We present an algorithm which effectively reduces the sensor response time to concentration steps by as much as one order of magnitude. The improvement upon the raw sensor response is the better, the less noisy the measured data are. This is because one has to smooth the data before looking for signal features indicating a concentration step, an operation which introduces a time lag relative to the raw sensor data. An example serves to demonstrate the effectiveness of the prediction algorithm together with either a simple moving-point averaging or a Savitzky-Golay (polynomial least-squares) filter.

Further data

Item Type: Conference item (Speech)
Refereed: No
Keywords: Mass-sensitive chemical sensors; SAW; QMB; dynamic response; analytic model; diffusion
Institutions of the University: Faculties > Faculty of Engineering Science > Chair Measurement and Control Technology > Chair Measurement and Control Technology - Univ.-Prof. Dr.-Ing. Gerhard Fischerauer
Faculties
Faculties > Faculty of Engineering Science
Faculties > Faculty of Engineering Science > Chair Measurement and Control Technology
Result of work at the UBT: Yes
DDC Subjects: 600 Technology, medicine, applied sciences > 620 Engineering
Date Deposited: 30 Nov 2021 13:57
Last Modified: 30 Nov 2021 13:57
URI: https://eref.uni-bayreuth.de/id/eprint/68047