Development of instrumentation setup for discrimination of analytes using principal component analysis

Abstract

An indigenous instrumentation set up was designed and developed for the discrimination of analytes. The instrumentation consists of a sensor array, measurement unit, data acquisition software, and a pattern recognition tool. The conductivity changes of three metal oxide sensors (ZnO (S1), In 2 O 3 (S2) and SnO 2 (S3)) with different analytes (hydrogen (A1), formaldehyde (A2) and hydrazine (A3)) were measured using in-house developed instrumentation and is sent to PC to store the sensor data for discrimination of analytes using pattern recognition technique. The data acquisition software collects the data in PC to train the data for discrimination of analytes using pattern recognition tool in real time. The various features of sensor response were derived from the response curve and the Principal Component Analysis (PCA) algorithm was implemented in software to extract the information from sensor response data for discrimination of analytes. The instrument set up is able to distinguish among hydrogen, formaldehyde, and hydrazine in real time.