Prediction of dye concentrations in a three‐component dye mixture solution by a PCA‐derivative spectrophotometry technique

Derivative spectrophotometry is one of the most important techniques that can be used to determine the dye concentration. In addition, principal component analysis (PCA) is a linear method to condense the dimensionality of large numbers of absorbance spectra. In this work, PCA and derivative spectrophotometry techniques are used to improve the accuracy of Beer's law prediction of the concentrations in three‐component dye mixtures. The performance of the new method is compared with the normal Beer's law by calculation absolute error, relative error, and ternary relative error of prediction. As obtained results indicate, the prediction accuracy of dye concentration prediction in PCA‐derivative spectrophotometry method is higher than normal Beer's law method. © 2009 Wiley Periodicals, Inc., Col Res Appl, 2010.