Machine learning-driven smart web app interface for effective urine pH detection using polyaniline functionalized paper sensor

A paper-based disposable impedimetric pH sensor has been developed to measure urine pH. In situ polymerization of aniline over paper substrate by chemical oxidative method is utilized to fabricate the sensing device. The acidic pH of the analyte solution maintains the emeraldine salt form of polyaniline, while the alkaline pH converts the emeraldine salt into emeraldine base. This sensing principle is employed to detect the urine pH with a sensitivity of 1.6 kΩ/pH for acidic region and 6.2 kΩ/pH for alkaline region, at a frequency of 1kHz. The sensing mechanism is also verified by employing a density functional theory-based first principle calculation. The formation of emeraldine salt and emeraldine base at acidic and alkaline pH can be validated by experimental Raman spectra. The observed spectra corroborate with the theoretical Raman spectra computed from the first principle calculations. The structural changes after exposure to acidic and alkaline solutions were experimentally verified by impedance spectroscopic method and correlated with the theoretical energy band gap study using density functional theory. Furthermore, the impedimetric responses were used to develop a machine learning-based smart and interactive web application, which predicts urine pH with an average accuracy of 98%.