Quantitative Structure–Retention Relationships of Polychlorinated Biphenyls (PCBs) Gas Chromatographic Retention Times: A Quantum Similarity Approach

Abstract

Polychlorinated biphenyls (PCBs) are a series of aromatic compounds which are considered as of major concern from the environmental point of view. In this contribution, a quantitative structure–retention relationship study was performed using a set of chromatographic retention times previously reported for a group of 209 PCB compounds. The semi-empirical PM3 theory level was used for geometry optimization, and an innovative quantum similarity approach, based on quantum similarity descriptors (QSDs) estimated using the Topo-Geometrical Superposition Algorithm, provided the data matrix required for the similarity analysis. Two linear relationships were established by means of partial least square regression. Both correlate the gas-chromatography retention time and the electronic or topologic representation of molecules expressed in terms of quantum similarity. The QSD calculated with the Coulomb operator has demonstrated a significant correlation with retention time. This relationship was established quantitatively in terms of their squared coefficient of determination and cross-validated leave-segment-out coefficient values r2 = 0.987 and q2 = 0.976. The model was also consistent in terms of the theoretical basis of molecular quantum similarity descriptors, which are capable of explaining in an effective way the main forces involved in the chromatographic separation phenomena. The results were compared contrasting the Coulomb and the Dirac operators models.