文摘
Three saccharides such as sucrose, maltose and glucose were simultaneously determined by the use of the light addressable potentiometric sensor (LAPS). One-chip integration of three saccharide sensors was performed by means of a glucose sensor using glucokinase. By the addition of oligosaccharide hydrolases such as invertase and α-d-glucosidase to the glucose sensors, the disaccharide concentrations were determined by variation of glucose concentrations. All the enzymes employed were thermostable enzymes because of their durability for repeated measurements. Clear distinctions of the response patterns were obtained among the three saccharide samples. Since the simplified LAPS system was limited in number of sensing spots, the complicated chemometrics such as partial least squares (PLS) was not always indispensable. Therefore, a simpler method such as multiple linear regression (MLR) was able to be applied to the mixture analysis under a low-concentrated sample in order to simplify the analysis. Thus, we treated the glucose as a key product in the regression equation optimized for the bienzymatic biosensensing process. The multivariate calibrations of the samples were simultaneously illustrated by using the estimated local glucose concentration. Finally, more than 0.7 of correlation coefficients were obtained between the true and the predicted sample concentrations although the linear approximation was not able to correct a few kinds of inhibition.