Quantitative structure-property relationships for predicting group IIB metal binding by organic ligands.

摘要

Mercury (Hg), cadmium (Cd), and zinc (Zn) in the environment are all of toxicological and environmental concern, and the pollution of natural waters by any of these three elements is most serious. Mercury is the most environmentally concerning of the three because of the neurotoxin species monomethylmercury produced in aquatic systems through the methylation of Hg2+ by aquatic microorganisms. An important chemical process in natural waters that limits the availability of mercury for methylation is the binding of Hg(II) by natural organic matter (NOM). These associations are exceptionally strong, and as NOM is ubiquitous in aquatic environments, estimating equilibrium constants for Hg(II) binding to NOM in natural waters is important. Cadmium is moderately toxic to all organisms, and skeletal damage caused by exposure to cadmium-contaminated water has been reported. Also high concentrations of zinc that are toxic or even lethal to organisms have been observed in natural waters. As the free ion forms of cadmium and zinc in natural waters are thought to be most toxic, Cd(II) and Zn(II) complexation by NOM and estimating the complexation equilibrium constants are, similarly to Hg(II), of interest.;With experimental determination of M(II)-NOM (M = Hg, Cd, Zn) binding constants being costly and time consuming, it is desirable to estimate those constants without the benefit of additional experimental data. This work uses QSPRs (Quantitative Structure-Property Relationships) to predict binding constants from hypothetical structures of NOM molecules. For the first time, to our knowledge, a QSPR for predicting Hg(II) complexation by organic ligands has been developed. Also two QSPRs for predicting Cd(II) and Zn(II) complexation by organic ligands, that had been developed earlier, have been improved to be capable of predicting the binding of Cd(II) and Zn(II) to thiol-containing molecules.;Most of the compounds used in the calibration data sets of the three QSPRs contained some or all of carboxylate, amine, and thiol ligand groups. The Hg(II), Cd(II), and Zn(II) QSPRs respectively have standard error of prediction (Spred) values of 1.60, 0.935, and 0.984 log units and describe 96.5%, 93.1%, and 93.4% of the variability in data. The most noteworthy observation in the developed QSPRs was the exceptionally high affinity Hg(II) had for thiols. Although thiols form a very small fraction of NOM, this binding is considerably important because of its strength. This work also presents certain potential applications of the developed QSPRs in predicting M(II)-NOM binding as well as predicting M(II) binding to organic molecules which would be synthesized for M(II) remediation and chelation therapy.