Study of the Self-Association of Alcohols by Near-Infrared Spectroscopy and Multivariate 2D Techniques

摘要

Self-association of 1-propanol, 1-butanol, 1-pentanol, 1-hexanol, 1-heptanol, 2-propanol, 2-methyl-1-propanol, and 2-methyl-2-propanol in carbon tetrachloride (concentration range 0.01-1.00 M) has for the first time been investigated by means of multivariate curve resolution in the near-infrared region of 1900-2200 nm. Rank analysis was carried out prior to resolution to determine the number of distinct species. In addition to principal component analysis (PCA) and evolving factor analysis (EFA), smooth PCA was used with good results. For all of the investigated alcohols, the rank was unambiguously assessed as three over the entire concentration range. At low concentrations, EFA detected a concentration region with only two species. Eigenstructure tracking analysis (ETA) revealed one-component regions in the spectral direction. The first loadings of the selective regions gave the concentration profiles of the two species. The spectral profiles of these two species were then resolved by using the least squares of these concentration profiles. The third profile was resolved by iterative target transformation factor analysis (ITTFA). The quality of the resolved profiles was checked by using the method of orthogonal projections (OP). The resolved profiles were subsequently used to determine the association numbers and equilibrium constants for the alcohols. The resolved spectral and concentration profiles of the linear primary alcohols and the branched alcohols separated into two distinct groups. The average sizes of the associated species increase with increasing concentration. At higher concentrations of alcohol, the cycliconpolar species exist in higher populations than do the linear species. A model that interprets the systems in terms of equilibria between a monomer, linear aggregates, and cyclic aggregates shows good agreement with the resolved profiles.