Floating gate MOS based olfactory sensor system
MetadataShow full item record
An artificial olfactory system, capable of detecting and discriminating different volatile organic compounds, has a vast potential to solve several sensing challenges. This work demonstrates an electronic olfaction system developed with the integration of gas sensitive conducting polymers to an array of Silicon foundry based floating gate metal oxide semiconductor (FGMOS) sensors. The FGMOS sensors were designed with a conductive extension of the floating gate terminal to a sensing surface. The sensors in the array were accessed and analysed individually using a specially designed addressing circuit. Multiple postprocessing steps were developed and practiced, creating the semiconductor chip compatibility to different sensing polymers and their electrochemical deposition environment. A sensing polymer, polypyrrole for example, was electrochemically deposited onto this sensing surface. The analyte influenced response of gas sensitive polymers was tailored by adopting different techniques at the time of electrodeposition of the polymer films. A novel integrated system having array of FGMOS sensors coupled to these chemically diverse polymers was developed. The polymers were synthesized using pyrrole and aniline monomers. Acetone, ethanol, methanol, isopropyl alcohol, petrol, toluene, ammonium hydroxide, acetic acid and water vapours were used to test the sensor system. Individual sensors coupled to distinct sensing polymers produced unique responses to the given analytes. The comparative measurements of these different sensor responses upon exposure to any vapour analyte facilitated a group signature-like response. The experiments have confirmed functional system response to different vapour analytes and their concentrations. A statistical data analysis technique, principal component analysis (PCA), was used to process the “analyte fingerprints” generated by the sensor array. In experiments involving multiple random order exposures of different vapour analytes, the PCA method produced a graphical representation having different isolated clusters of datapoints where each cluster corresponds to an analyte under test. This technique was demonstrated for its effectiveness in detection of any ‘unknown vapour analyte’ from the given set of analytes. The commercial CMOS technology used for this work enables possibility of cost-effective large-scale production of these chips. This work has potential for development of the system specific to different industrial applications.