(86f) Dependence of Photocatalytic Degradation Pathway on Surface Planes of a Catalyst: A Case of Diuron Degradation on ZnO
In this work, the dependence of the photocatalytic degradation pathway on surface planes of the catalyst is demonstrated. Diuron [3-(3, 4-dichlorophenyl)-1, 1-dimethyl urea], which is one of the most highly persistent and highly toxic herbicides, was used as a model compound. Zinc oxide (ZnO) was chosen as the catalyst because it can be synthesized such that the dominating surface on a particle can be easily controlled. A hexagonal crystal of wurtzite ZnO is consisted of two main sets of surfaces; (i) zinc-terminated (0001) and oxygen-terminated (000-1) polar surfaces locating at the top and bottom planes of the crystal, and (ii) mixed-terminated (10-10) nonpolar surface as the side planes. Herein, two types of ZnO, i.e., conventional hexagonal particles with polar surfaces as the dominating planes, and nanorods with the mixed-terminated surfaces, were studied.
The conventional particles were prepared by the sol-gel technique at room temperature, after which the calcination at 500°C was performed. On the other hand, the nanorods were synthesized by the hydrothermal technique at 140°C. Zinc acetate was used as the precursor for both syntheses. Various characterization techniques, including X-ray diffraction (XRD), X-ray photoelectron spectroscopy (XPS), UV-Visible spectroscopy, selected area electron diffraction, and nitrogen adsorption/desorption, were performed to confirm that both catalysts are nonporous single-crystalline ZnO in wurtzite phase with similar surface compositions. Their optical bandgaps are the same at 3.1 eV. The major differences between these catalysts are their morphologies, sizes and specific surface areas. The conventional ZnO powders are micron-sized particles with (0001) or (000-1) plane, while the nanorods exposes mostly the (10-10) surface. Because of the much smaller particle size of the nanorods, their specific surface area is much higher than that of the conventional particles, i.e., 17 m2/g versus 1.4 m2/g.
The adsorption experiments were conducted in batch in the dark, using diuron aqueous solution with concentrations in the range of 0-25 mg/l. The Freundlich model was found to represent the experimental data well. The fitted parameter relating to adsorption intensity shows unambiguously that diuron adsorb more strongly on the conventional ZnO than on the ZnO nanorods. The diuron-surface interaction is suggested to be electrostatic. Interestingly, the greater amount of diuron was adsorbed on the conventional ZnO particles than on the nanorods, although the former has one-order of magnitude lower surface area than the latter.
Molecular calculations using density functional theory were applied to support the experimental findings. The adsorption configuration of diuron was found to be surface dependent. On the zinc-terminated surface, diuron adsorbs in planar configuration because both ends of diuron, which are negatively charged, are attracted to the positively charged surface zinc atoms at the same time. In case of the oxygen-terminated surface, diuron tilts its molecule almost side-way so that the electron-rich amide oxygen moves away from the negatively charged surface oxygen.
On the mixed-terminated surface of ZnO, molecule of diuron is bent so that the most negatively charged part of the molecule, i.e., amide oxygen and nitrogen attaching to the aromatic ring, could approach positively charged zinc atoms on the surface. At the same time, the aromatic end of the molecule is lifted from the surface because of the repulsion between chlorine atoms attaching to it and surface oxygen atoms. If zinc atom on the surface is defined as the adsorption site, it can be seen that only half of the surface atoms of the mixed-terminated surface are adsorption sites, unlike the polar surfaces where all atoms are able to be the adsorption sites. Furthermore, the adsorbate molecule that already adsorbed on the surface will present steric effect toward adsorption on nearby sites. This steric effect will become more severe toward the adsorption capacity on the mixed-terminated surface, where not all atoms are the adsorption sites. That explains why the adsorption capacity on the conventional ZnO particles is much higher than that on the nanorods.
The photocatalytic degradation experiments were conducted in a microreactor to eliminate mass transfer resistance for the transport of diuron from bulk liquid to the surface of the catalyst. The outlet stream from the reactor was collected and analyzed by liquid chromatography equipped with tandem mass spectroscopy (LC-MS/MS). For the degradation on the ZnO nanorods, the first set of intermediates detected using the shortest residence time (1 min), which are assumed to be the direct intermediates from the degradation of diuron, are formed by hydroxylation and demethylation at the methyl group. These reactions take place at the aliphatic moiety of diuron, which is the part of the molecule adsorbed on the mixed-terminated surface of the nanorods. On the other hand, on the conventional ZnO particles, the intermediates detected are the results from diuron being attacked by hydroxyl radicals on both aliphatic and aromatic sides simultaneously. As previously discussed, diuron adsorbs in a planar configuration on the zinc-terminated surface, while it turns perpendicularly to the oxygen-terminated surface. Both of these give rise to equal probabilities of attack by the hydroxyl radicals on either parts of diuron. The structure of the subsequent intermediates should therefore be consequently affected. Hence, we propose that the difference in pathways from using different catalysts is caused by the adsorption configurations of diuron onto the different surfaces.
Finally, the toxicity of the generated intermediates was investigated. Both the phytotoxicity and cytogenotoxicity were evaluated in comparison with diuron. The results show that the initial intermediates formed from diuron degradation on the ZnO nanorods are less toxic than those generated on the conventional ZnO particles. Nevertheless, with prolonged residence time, e.g., 15 min, the toxicity of the degradation product was dramatically reduced to a level similar to that of deionized water.