Evaluating the Molecular Interaction of Organic Liquid Mixtures Using Near-Infrared Spectroscopy

The near-infrared transmission spectra of two organic liquid three-component systems of variable compositions were investigated in detail. To evaluate the interaction of the different components in the two systems the experimental spectra of the pure components were compared to mathematically constructed "pure component" spectra. Though usually the correlation coefficient (CC) and Manhattan distance (MD) are used to measure the similarity of spectra, in the present investigations principal component analysis (PCA) was found to be a more effective tool to investigate the difference between these spectra and derive parameters characterizing the interaction between the different components. Thus, PC scores for the two types of spectra established some distinct patterns which clearly expressed their differences. For a three-dimensional coordinate system of selected principal components, the Euclidean distances between the mathematically constructed and the experimental spectra of the pure components were calculated. Finally, the mean values of the distances for each component provided indices to rank the interaction of the components in the mixtures. Thus, the results offer a convenient approach that can quantitatively evaluate the molecular interactions of the individual components in organic liquid mixtures by various spectroscopies.