Chemical equilibria studies using multivariate analysis methods

Chemical multiequilibria systems can be monitored efficiently with the aid of spectroscopic techniques. Both hard- and soft-modeling are effective and powerful tools to extract chemical information from spectroscopic data. Recently, hybrid approaches that combine the flexibility of soft-modeling with the precise solutions provided by hard-modeling have been proposed. Here, we tested the performance of these three chemometric approaches for the analysis of several simulated data sets. In addition, experimental data recorded during the study of the acid–base equilibria of two DNA structures (G-quadruplex and i-motif) corresponding to two short sequences of the k-ras oncogene were studied. Finally, we also analyzed the interaction of the two DNA sequences with the model ligand TMPyP4. The results obtained from the analysis of these data sets may be useful to determine the most appropriate use of each approach. Whenever the presence of optically active interferences or unknown drifts can be neglected and a chemical model can easily be proposed and fitted, the hard-modeling method shows the best performance. If any of these conditions is not fulfilled, a hybrid-modeling approach may be a better option because all the contributions (chemical and unknown) can be modeled and the ambiguities inherent to soft-modeling methods show minor effects.