Determination of the number of light-absorbing species in the protonation equilibra of selected drugs

The determination of the number of components in a mixture is an important tool for qualitative and quantitative analysis in spectroscopy. The accuracy of nine selected indices for an estimation of the number of components that contribute to a set of spectra was critically tested on experimental data sets of protonation equilibria of four drugs using the INDICES algorithm in S-Plus. Methods are classified into two categories: precise methods based on a knowledge of the instrumental error of the sabsorbance data, s_inst(A), and approximate methods requiring no such knowledge. Indices of precise methods predict the correct number of components, even the presence of a minor one, when the quality of data is high and instrumental error is known. Improved identification of the number of species uses the second or third derivative function for some indices, namely when the number of species in the mixture is higher than four and when, due to large variations in the indicator values even at logarithmic scale, the indicator curve does not reach an obvious point where the slope changes. The number of variously protonated components and their dissociation constants for four drugs—mycophenolate, ambroxol, silybin and silydianin—at 25 °C were determined using SQUAD(84) regression and INDICES principal component analysis of the pH-spectrophotometric data. A proposed strategy of efficient experimentation in protonation constants determination, followed by a computational strategy, is presented with the goodness-of-fit tests for various regression diagnostics enabling the reliability of parameter estimates to be accessed.