ACS PRF | ACS
All e-Annual Reports
42113-G5
Theoretical and Experimental Investigations of Propylene Epoxidation on TiO2 Supported Gold Nanoparticles
It has been shown recently that gold nano-particles dispersed on metal oxides exhibit exceptional catalytic activity in multiple reactions including low temperature CO oxidation, direct propylene epoxidation (this is the only heterogeneous propylene catalyst that utilizes O2 as the oxidizing agent), hydrogenation reactions, and reduction of NOx. These observations are surprising considering that bulk gold is the least active metal. The unique catalytic activity of Au nano-clusters supported on oxides has stimulated an intense debate about fundamental factors that govern this behavior. Possible explanations include quantum size effects, strain, charging of gold particles by interaction with defect sites of oxide supports, a cooperative action of oxide supports and metal nano-particles, the effect of metal-insulator transitions, and the effect dominated by low-coordination of Au atoms in nano-particles.
In this project we have proposed to employ a hybrid experimental/theoretical framework combining DFT calculations and well-defined surface science experiments to investigate the molecular level mechanism of oxidation reactions on oxide-supported Au nano particles. We have particularly focused on the role of Au/oxide interfaces in driving the oxidation activity.
So far, we have investigated the reactions of O2 on Au (111), Au/TiO2(110), and Au/SiO2(110) interfaces. It has been shown that the activation of O2 is the rate-controlling step in almost all oxidation reactions on Au. Density functional theory (DFT) and ab initio thermodynamic calculations have been utilized to study (i) the oxidation state of Au under reaction conditions (cationic versus anionic Au) and (ii) the role of the oxide support in O2 activation. We have focused on pressure- and temperature-dependant interactions of oxygen with Au(111) and with Au deposited on rutile TiO2(110) and SiO2(110).
To model Au/TiO2 and Au/SiO2, we have utilized a number of different Au structures adsorbed on the O-vacancy-rich, pristine stoichiometric, and oxygen-rich oxides. We have explored a few different model systems including small Au clusters adsorbed on oxides, Au nano-rods on oxides, and Au bilayers adsorbed on the oxides. The conclusions presented below were not sensitive to the choice of the model system.
In these studies we have demonstrated that the oxidation state of Au is governed by external conditions (oxygen pressure and temperature) and by the chemical interactions between oxides and Au. We find that under low oxygen chemical potentials, electron density is transferred from an oxide support (mainly from oxygen vacancies) to Au, forming Auä-. On the other hand under catalytically relevant conditions there exists a thermodynamic driving force to oxidize Au and form cationic Auä+. We also show that anionic Auä-, formed when Au is deposited on the O-vacancy-rich TiO2, interacts strongly with oxygen and is easily oxidized even at moderate oxygen chemical potentials. Similar phenomena take place on the vacancy-rich SiO2; however, the extent of charge transfer from SiO2 to Au is lower than for TiO2 and therefore the impact of this charge transfer on O2 activation is less accentuated. These observations are consistent with reported experimental observations which showed that Au supported on reducible oxides is more chemically active than Au on irreducible oxides.
Based on these calculations we have proposed a simple model grounded in first principles calculations that can account for the observed support-dependant chemical activity of nano-Au. The model consists of four major points:
1. O-vacancy-rich educible oxides such as TiO2 donate more electronic charge to Au than irreducible oxides such as SiO2. This leads to significant charging of Au when adsorbed on TiO2
2. Anionic Au, formed when Au is adsorbed on oxides can activate O2 dissociation with a lower activation barrier compared to unsupported Au
3. When O2 is activated, Au-O bonds are formed and the electronic fingerprint of Au is reversed, i.e., cationic Au+ is formed under the relevant reaction conditions.
4. We have also been able to identify special sites at the metal/oxide interfaces that exhibit dramatically higher catalytic activity than the sites that are removed from the interface.
The results outlined above have been reported in our recent publication (SIRIS LAURSEN and SULJO LINIC, "Oxidation catalysis by oxide-supported Au nanostructures: The role of supporters and the effect of external conditions", Physical Review Letters, 97, 026101, 2006). Another publication is currently in production.
