Zinc oxide (ZnO) is a versatile metal oxide compound. It can be used in many applications including being a photocatalyst for degradation of organic contaminants. Our previous work used ZnO nanoparticles as the catalyst for photo-degradation of various phenyl urea herbicides, such as diuron (3-(3, 4-dichlorophenyl)-1, 1-dimethyl urea) and linuron (3-(3, 4-dichlorophenyl)-1-methoxy-1-methyl urea), and found that the degradation intermediates depended upon how ZnO was synthesized. It was hypothesized that difference in surface structure of ZnO nanoparticles that were synthesized differently led to different characteristic adsorption of the herbicide molecules. Detailed investigation of the adsorption of the herbicides on different surface of ZnO will be presented in this work. Both experimental works and molecular simulation were conducted and reported.
For photocatalysis, the suitable crystalline structure of ZnO is wurtzite, in which atoms are arranged in hexagonal pattern. The wurtzite ZnO contains two types of surface, i.e., non-polar surfaces and polar surfaces. On the non-polar surfaces, such as mixed-terminated (10(-1)0) and (11(-2)0) surfaces, zinc and oxygen atoms are arranged alternately in equal number, resulting in no electrostatic instability. On the other hand, the polar surfaces, such as zinc-terminated (0001) surface and oxygen-terminated (000(-1)) surface are consisted of only one kind of atom. These surfaces are less stable than the non-polar surfaces because of the electrostatic instability. In this work, two types of ZnO, i.e., nanoparticle and nanorod, were synthesized. The ZnO nanoparticles were synthesized by sol-gel method at room temperature, while the nanorods were synthesized by hydrothermal technique at 140°C for 1 h. Zinc acetate was used as the precursor for both syntheses. The X-ray diffraction analysis confirmed that the products synthesized by both methods are indeed ZnO in wurtzite structure, while the scanning electron microscopy indicated the significant difference in morphology. The nanoparticles were irregular aggregates while the nanorods were about 50 nm in diameter and 100-200 nm in length. The hexagonal column of the nanorods, which is consisted of the mixed-terminated (10(-1)0) surface, was clearly defined.
The adsorption studies were carried out in batch in the dark. Aqueous solutions of diuron, linuron and 3, 4-dichloroaniline, which is a common degradation intermediate of both diuron and linuron, were prepared at various concentrations. The pH of the solution was also varied in the range from 4 to 10. It should be noted that diuron, linuron and 3, 4-dichloroaniline share the same molecular structure on the aromatic side of the molecules, while their aliphatic sides are different. ZnO was added into the solution at the content of 10 mg catalyst per 10 ml of the solution. Temperature of the system was maintained at 25±3°C, using a cooling jacket. Equilibrium concentration of the system was measured via a high performance liquid chromatography (HPLC).
For the adsorption of either diuron or linuron on both ZnO nanoparticles and nanorods, the results indicated that the adsorption capacity increases with the decrease in pH. It is therefore suggested that the adsorption is a result from electrostatic interaction between negative charge on the molecules and positive charge on the surface. When pH is lower than the point of zero charge of the catalyst (i.e., 7.6 for the synthesized ZnO), the net charge on the surface is positive, which attracts more herbicide molecules onto the surface of ZnO. The trend is reversed for the adsorption of 3, 4-dichloroaniline because the most negatively charged part of 3, 4-dichloroaniline molecule is located at –NH2 functional group, which can be easily hindered by an aromatic ring since the bond length between the aromatic ring and the –NH2 group is short. These experimental results agree with the molecular simulation using Material Studio 5.5, which is a well-known molecular surface simulation program. The simulation results showed that diuron and linuron turn their aliphatic side toward ZnO surface, while 3, 4-dichloroaniline turns chlorine atoms attaching on its aromatic ring toward the surface. These adsorption behaviors are similar on all surfaces of ZnO simulated, although the binding energy between the adsorbed molecule and the surface is varied.
The adsorption data were then fitted with adsorption models. It was found that the Freundlich isotherm can better represent the experimental data than the Langmuir isotherm. By comparing the adsorption on ZnO nanoparticles and that on nanorods, the fitted value of Kf, which is related to the adsorption capacity of the adsorbent, obtained from the adsorption on the nanorods is significantly lower than that on nanoparticles. In contrast, the n values are close, which suggests similar interaction between the herbicides and the surface of ZnO, regardless of the type of ZnO used. Considering the binding energy of adsorption obtained from the molecular simulation, it was found that the binding energy on mixed-terminated (10(-1)0) surface is lower than that on oxygen-terminated (000(-1)) surface. It confirms the results from the experimental study that the adsorption capacity of ZnO nanorods, in which majority of the surface is (10(-1)0 ) surface, is lower than that of ZnO nanoparticles.
When the pH was changed to acidic, H+ was assigned to the surface and bonded with oxygen atom. It leads to higher binding energy between the herbicide and the surface of ZnO because the interaction between the negatively charged functional group on the adsorbed molecule and the positively charged surface becomes stronger. On the other hand, OH- was assigned to the surface when the pH was changed to basic. The simulation results showed that the adsorption capacity is decreased because of weak interaction between the adsorbed molecule and the surface.