Density functional theoretical study of Au₄/In₂O₃ catalyst for CO₂ hydrogenation to methanol: The strong metal-support interaction and its effect

Highlights

• Au/In₂O₃ model catalyst is constructed by placing Au₄ cluster onto In₂O₃(110) surface.

• The strong Au-In₂O₃ interaction is confirmed.

• Gold is in Au^δ+ state for improved H₂ dissociative adsorption ability.

• Obtained Au-In₂O₃ interface is the active center for CO₂-based methanol synthesis.

• Au-In₂O₃ interface also benefits hydroxyl migration to prevent catalyst deactivation.

Abstract

Density functional theoretical (DFT) study was conducted to investigate the feasibility of CO₂ hydrogenation to methanol on an Au₄/In₂O₃ model catalyst. A strong metal-support interaction is confirmed by the binding energy between Au₄ cluster and In₂O₃ support, which is −5.31 eV. This causes the electron redistribution at the interfacial sites and leads to a positively charged Au^δ+ cluster instead of metallic Au⁰. The positive oxidation state of Au^δ+ cluster is originating from neighboring O atoms, i.e., the electron of Au cluster is transferred to adjacent O atom through AuO bond. This electron redistribution can activate Au^δ+ cluster by generating electron depleted regions, which are active for H₂ dissociation. The feasibility and reaction route of methanol synthesis from CO₂ hydrogenation at Au-In₂O₃ interface is further examined to understand the effect of Au-In₂O₃ interaction on the catalytic performance. CO₂ is activated at the interface, and formate is the key intermediate during the reaction. The rate-limiting step is hydroxyl spills over from Au cluster to In₂O₃ surface to release active Au site, with a reaction barrier of 0.95 eV. This can prevent hydroxyl from binding strongly with Au cluster and deactivate the catalyst. Compared with CH₂OH intermediate, CH₃O intermediate is more favorable in the final hydrogenation step. The DFT study suggests that the Au/In₂O₃ catalyst is a promising one for methanol synthesis from selective hydrogenation of CO₂.

Introduction

Fundamental studies on the behavior of gold catalysts are a hot topic. [[1], [2], [3], [4], [5]] Bulk gold is chemically inert for chemical processes. However, supported gold nanoparticles (NPs) have been widely reported to be highly reactive for reactions like CO oxidation, the water-gas shift reaction, semi-hydrogenation of acetylene and others [3,6]. Most of gold catalysts are active for the oxidation reaction, while only a few have been reported to be active for hydrogenation reaction [[7], [8], [9], [10], [11], [12]]. The catalytic performance of gold catalysts can be affected by electronic effects, which is usually the result of the metal-support interactions (i.e., the electron transfer between Au and the support) [13]. It has been reported that Au⁰, or electron-rich Au⁻, is the active site for CO oxidation and the water gas shift reaction. [14,15] For hydrogenation reactions, the electronic structure of gold affects the adsorption/activation of reactants, which further affects the activity and selectivity [16]. In the limited works on hydrogenation reactions, the role of Au is not sufficiently clarified.

On the other hand, CO₂ hydrogenation to valuable chemicals and fuels is promising for the utilization of carbon dioxide in a large scale, with the development of renewable hydrogen. [[17], [18], [19]] Compared with CO or methane, methanol is considered both a clean energy carrier and an essential intermediate for industrial chemical syntheses. Therefore, CO₂ hydrogenation to methanol attracts significantly increasing attention worldwide. The reaction (CO₂ + 3 H₂ → CH₃OH + H₂O) is thermodynamically exothermic. To our knowledge, only a few gold catalysts have been exploited for CO₂ hydrogenation [[7], [8], [9], [10], [11], [12]] with ZnO [7,8], ZrO₂ [8,9], CeO₂ [10,11], and TiO₂ [12] as the supports in these gold catalysts. However, the activity of these catalysts is not impressed.

Recently, In₂O₃ and In₂O₃ supported catalysts have attracted significant attention because of their high selectivity (∼100 % at low reaction temperatures and higher than 60 % at elevated temperatures) towards methanol from CO₂ hydrogenation. [[20], [21], [22], [23], [24], [25], [26], [27], [28]] Different from other regular metal oxides, like TiO₂, CeO₂, MgO, SiO₂, Al₂O₃, which are typically used as carrier to disperse the supported metal and to adsorb CO₂, indium oxide itself shows ability not only to activate carbon dioxide but also to hydrogenate the molecule selectively to methanol. [[20], [21], [22]] However, the weak H₂ dissociation ability limits its catalytic activity of In₂O₃. To further improve the CO₂ conversion, palladium [[23], [24], [25]], platinum [26], rhodium [29], nickel [27], and cobalt [28] have been applied to enhance its ability to activate hydrogen. These metals are well known as excellent H₂ splitter in various hydrogenation reactions.

The exploration of In₂O₃ based catalysts for CO₂ hydrogenation to methanol was started from a density functional theoretical (DFT) study by Ye et al. [20]. This theoretical prediction was later confirmed by the experimental study [21,22]. Since then, a significantly increasing publication has been observed on CO₂ hydrogenation to methanol over In₂O₃ catalysts. Ye et al. [23] also predicted theoretically with a Pd₄/In₂O₃(110) model catalyst that Pd/In₂O₃ is a highly active catalyst for CO₂ hydrogenation to methanol. They reported that CO₂ is activated at the interfacial site (between Pd and In₂O₃) and then further hydrogenated to methanol. [23] With Pd loading on In₂O₃, the electronic properties are strongly modified, [23,25] leading to improved H₂ dissociation. This prediction was experimentally confirmed as well [24,25].

In this work, DFT study has been conducted to theoretically investigate the feasibility of Au/In₂O₃ catalyst for CO₂ hydrogenation to methanol. Au₄/In₂O₃(110) was chosen as the model catalyst. In₂O₃(110) surface is a common theoretical model used when dealing with indium oxide. [20,23,30] It has similar surface energy but fewer atoms compared with In₂O₃(111), which is the major facet in those reported powder catalysts. [22,24,25] The reason for In₂O₃(110) was employed here is that it is quick and cheap for the modeling but with not much difference from the In₂O₃(111) for the prediction of the catalytic performance, according to the previous studies on In₂O₃(110) and In₂O₃(111). [20,23,29,30] We confirm theoretically that Au/In₂O₃ is a good catalyst for CO₂ hydrogenation to methanol with the Au-In₂O₃ interface as the active site. The strong Au-In₂O₃ interaction and its effect on the catalytic performance for CO₂ hydrogenation to methanol have been identified. The pathways of CO₂ hydrogenation have been investigated as well.