Please use this identifier to cite or link to this item:
|Type:||Artigo de periódico|
|Title:||QSPR study of passivation by phenolic compounds at platinum and boron-doped diamond electrodes|
|Abstract:||Blocking polycrystalline platinum (Pt) and boron-doped diamond (BDD) electrodes by 20 phenolic compounds was studied by means of chronoamperometric and theoretical methods [i.e., quantitative structure-property relationships (QSPRs)] and chemometric methods. The difference between the current densities after 15 and 90 s of oxidation time was proposed for the first time as a quantitative measurement of passivation on the electrodes. Structures of phenolic molecules and their hydrogen-bonding complexes with fluoride ion were modeled and geometry optimized with the B3LYP method and the 6-31G** basis set. Several molecular descriptors were calculated and correlated with the passivation measurements using the partial least-squares (PLS) regression method. A PLS model with one latent variable from five descriptors was built with high-level predictivity for phenolic passivation on the Pt electrode, and then was externally validated with four phenolics. The Pt model statistical parameters obtained were: Q(2) = 0.786, R(2) = 0.851, and standard error of validation (SEV) = 0.097. A BDD model with one latent variable and four descriptors was built and validated in the same way; however, the statistical parameters (Q(2) = 0.333, R(2) = 0.586, and SEV = 0.159) were of inferior quality with respect to the model for the Pt electrode. Both models were applied for prediction of 10 phenolic compounds. The Pt model showed to be suitable for predictive purposes. It was observed that passivation was much weaker on the BDD electrode than on the Pt electrode. Different interactions and reactions involving phenolics at the electrodes are the main reasons for such large differences between the models. Exploratory analyses were also performed and interpreted in terms of chemical concepts, such as phenolic reactivity, size/shape, hydrogen bonding, and electronic features. These findings can be useful to explore the possibility to predict phenolic passivation and to design electrochemical experiments involving different phenolic compounds. Furthermore, these PLS models aid in understanding electrode inactivation by phenolic compounds. (C) 2008 The Electrochemical Society.|
|Editor:||Electrochemical Soc Inc|
|Appears in Collections:||Unicamp - Artigos e Outros Documentos|
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.