Multi-Component Kinetic Analysis in Liquids with a Single Lipid-Coated Piezoelectric Crystal Sensor Using Artificial Neural Networks

Abstract

A single piezoelectric crystal sensor was used to simultaneously detect two or three components in a liquid by observing the kinetic adsorption process onto a synthetic multi-bilayer lipid membrane which was coated on both sides of the piezoelectric crystal sensor. The feed-forward artificial neural network (ANN) with back propagation (BP) training algorithm was used to do the data analysis. The correct rate of prediction for the first sample set, 1-decanol and β-naphthol, is 4.3%, for the second one, 1-lauryl amine and α-naphthylamine, it is 4.8% and for the third one, sodium lauryl sulfate, sodium dodecyl benzene sulfonate and N-cetyl pyridinium chloride, it is 5.2%.