Investigação da adsorção de arsênio e ácidos organofosfônicos em minerais de alumínio através de uma nova abordagem baseada na teoria do funcional de densidade

Abstract

Arsenic is an ubiquitous element found in the atmosphere, surface and ground waters, soils and rocks, as well as in living beings. The mobilization of arsenic in the environment depends on various factors such as climate, weathering, and biologic activity, for example. However, anthropogenic factors such as mining and burning of fossil fuels have added an important contribution to the impact of arsenic in the environment. It is well known that As(III) shows higher mobility in the environment than As(V), which can be explained by the high reversibility of As(III) adsorption on minerals such as aluminum oxides and hydroxides. Nevertheless, EXAFS experimental results pointed out that As(III) chemically adsorbs on these minerals, which appears to be in conflict with its high mobility as it is expected for chemical adsorption complexes to be strongly attached. In this thesis theoretical calculations have been used to model the adsorption of As(III) in the gibbsite/water interface following two different mechanisms. Gibbsite is an aluminum hydroxide with expressive surface area, available as one of the main components of bauxite. Initially, density functional theory (DFT) calculations were applied to finite cluster models, resulting in evidence that, oppositely to As(V), As(III) does not adsorb following an acid-base mechanism, but a non-dissociative one, by which OH bonds do not break. The non-dissociative mechanism conciliates the high remobilization of As(III) and the experimental evidences of chemical adsorption on the gibbsite surface. Although theoretical calculations are in good agreement with EXAFS data, the relative adsorption energies of the investigated sites lead us to conclude that the adsorption energies are strongly affected by long-range interactions and, therefore, more complex models must be used. A detailed study on the geometric and electronic structure of gibbsite was carried out in order to enhance the description of As(III) adsorption. In this study the self-consistent charge density-functional based tight-binding (SCC-DFTB) method was used. In this study the intrinsic electronic states of the surface as well as the changes in the electronic structure caused by restricting the size of the models were observed. DFT calculations were also carried out and showed that, besides involving considerably less computational effort, the SCC-DFTB method gives reliable results, with minor discrepancies from DFT methods. Thereafter, the adsorption of As(III) was studied using SCC-DFTB calculations with cluster and periodic models. New evidences were gathered in favor of the non-dissociative adsorption mechanism. Also, the influence of long-range effects upon the adsorption energies and the adsorbate geometries was demonstrated, especially in the case of the non-dissociative mechanism. Consequently, a new model was proposed for the gibbsite surface. Besides the long-range interactions, the solvent effect was also included by adding a layer of water molecules over the surface. Therefore, the gibbsite/water interface was explicitly modeled by molecular dynamics simulation using the SCC-DFTB method, i.e. , following a totally quantum-mechanical approach. It was observed that the presence of water does not significantly affect the electronic structure of the surface, but the water molecules strongly interact with aquo and hydroxyl groups on gibbsite through hydrogen bonds. Thus, there are very specific interactions in the solid/liquid interface that may explain the difficulty to definitely conciliate the theoretical and experimental results obtained for the As(III) adsorbed on gibbsite. At last, the application of SCC-DFTB was extended to self-assembled systems. The adsorption of alkylphosphonic acids on corundum ( -Al2O3) was studied with SCC-DFTB. The favored adsorption site for different coordination numbers was estimated. The calculations allow to conclude that the increase of the number of bonds between the surface and the adsorbate also increase the regioselectivity. Hence, the application range for the SCC-DFTB method was broadened, allowing for the prediction, in a middle term, the application of this method to self-assembled, supra-molecular and nano-reactor systems.