Akademik Veri Yönetim Sistemi
International Workshop on Impedance Spectroscopy (IWIS), Chemnitz, Almanya, 27 - 30 Eylül 2022, ss.61-65
Impedance measurement of quartz crystal microbalance (QCM) to quantify mass adhered on the surface and/or viscoelastic properties thereof in solution is the most informative methodology among all available techniques. State of the art impedance measurement requires advanced and costly equipment. However, recent advances in digital technology lead the way for the development of battery powered antenna analyzers capable of measuring impedance at a fraction of the cost of an expensive, benchtop impedance analyzer. Machine learning (ML), or the ability of self-training computer algorithms to autonomously interpret the structure of the data enables us to learn specific patterns or trends from data in order to perform complex tasks such as prediction, classification, clustering and regression. In this study a machine learning (ML) approach allowing optimization of impedance parameters to reduce limit of detection (LOD) is used to compare the performance of 2 different battery powered antenna analyzers with a benchtop impedance analyzer, by determining the LOD of glycerol concentration in distilled water. Monitoring of the fundamental resonance peak (10 MHz) of an AT cut QCM with the ML optimized parameters determined for each instrument, revealed a approximate to 22% difference in the lowest LODs of impedance analyzer and antenna analyzers, respectively.