Abstract
Electronic noses (e-noses) are increasingly being used as vapour sensors in a range of application areas. E-noses made up of arrays of organic field-effect transistors (OFETs) are particularly valuable due the range and diversity of the information which they provide concerning analyte binding. This study demonstrates that arrays of OFETs, when combined with a data analysis technique using Genetic Programming (GP), can selectively detect airborne analytes in real time. The use of multiple parameters – on resistance, off current and mobility – collected from multiple transistors coated with different semiconducting polymers gives dramatic improvements in the sensitivity (true positive rate), specificity (true negative rate) and speed of sensing. Computer-controlled data collection allows the identification of analytes in real-time, with a time-lag between exposure and detection of the order of 4
s.