Highly Sensitive and Selective Detection of Dimethyl Methyl Phosphonate with Copolymer‐Based QCM Sensors

Abstract

Abstract In this work, the volatile organic compounds (VOCs) sensing properties of a quartz crystal microbalance (QCM) transducer coated with six different poly(3‐methylthiophene) (P3MT) copolymerized with polypyrrole (PPy) are investigated. The sensor preparation involves the electrochemical deposition of P3MT, PPy, and P3MT‐co‐PPy on Au‐coated QCM transducers by electrochemical deposition techniques with a three‐electrode cell. The structural properties of the copolymer films are characterized using scanning electron microscopy, and their oxidation/reduction behavior is investigated through cyclic voltammetry. The copolymer‐based QCM sensors exhibit high sensitivity and selectivity to dimethyl methyl phosphonate and benzonitrile, even at low concentrations (<1 ppm) at room temperature. Langmuir, Freundlich, Temkin, and Dubinin–Radushkevich adsorption isotherms are studied to understand the VOCs sensing mechanism machine learning classification algorithms including quadratic discriminant (QD), neural nets, K‐nearest neighbors, linear discriminant, and support vector machines are applied to classify the sensor responses for the 12 different analytes. With the help of machine learning algorithms, tested analytes are successfully classified into their groups. The highest accuracy of 97.34% is achieved using the QD method. The developed sensor, combined with machine learning algorithms, shows promising potential for accurate and reliable detection and classification of VOCs.

Keywords

• adsorption kinetics
• copolymers
• dimethyl methyl phosphonate
• machine learning
• poly(3‐methylthiophene)
• polypyrrole