Elsevier

Journal of Catalysis

Volume 216, Issues 1–2, May–June 2003, Pages 425-432
Journal of Catalysis

Supported Au catalysts for low temperature CO oxidation

https://doi.org/10.1016/S0021-9517(02)00111-2Get rights and content

Abstract

Supported Au catalysts are very active for low temperature CO oxidation. However, the reported activity from different laboratories for apparently similar catalysts can differ quite substantially. Recent progress in resolving this difficulty is summarized. Residual chloride in the sample is a very effective poison of the active site. The effect of water vapor on the catalytic activity differs depending on the support and the residual chloride content. A model of the active site, which consists of an ensemble of metallic Au atoms and Au cations with hydroxyl ligands, the reaction mechanism for CO oxidation, and the mechanism for deactivation during reaction as well as regeneration are discussed with respect to the available data.

Introduction

The interest in studying supported Au catalysts has increased substantially since Haruta et al. discovered that these catalysts are exceptionally active for low temperature CO oxidation [1]. Compared to a highly dispersed supported Pt catalyst, Au catalysts can be an order of magnitude more active [2]. The attention arises not only from the commercial implication of this discovery, but also from the desire to answer the question of why the normally inert Au material can catalyze chemical reactions so effectively. A better understanding of the origin of the catalytic activity could lead to the discovery of other novel reactions. Already, supported Au catalysts have been shown to be active for selective reduction of NO by hydrocarbon in the presence of a high concentration of oxygen [3], [4], selective oxidation of CO in a hydrogen stream [5], [6], selective oxidation of hydrocarbons [7], epoxidation of propene [8], and selective hydrogenation [9], [10].

Most of the studies to date have focused on the unusual low temperature CO oxidation activity. The results suggest that the activity is highly sensitive to the details in the preparation procedure and the catalytic testing conditions. Indeed, as summarized by Bond and Thompson [11] and others (e.g., [12], [13], [14]), large variations in observed catalytic activities can be found reported in the literature over catalysts of similar compositions. This is illustrated by the data in Table 1. Among the various supported Au catalysts, Au/Al2O3 is perhaps one that shows the widest variation, ranging from being very inactive [15], [16] to being comparable or better than Au/TiO2 [12], [17]. It is important to understand the origin of the wide variations in order to eliminate them.

Ultimately, the variation in catalytic activity is due to the lack of control in the generation of the active site, either in terms of its number or its properties. Therefore, identifying the nature of the active site and the corresponding reaction mechanism would be of great value.

Significant progress has been made recently in understanding the preparation chemistry and elucidating the active site. Here, we briefly summarize the more recent results in these areas, with an emphasis on those that provide information on the underlying chemistry, and offer our view of the current state of this particular area of catalysis. Earlier work has been summarized in other reviews [11], [18].

Section snippets

Preparation of supported Au catalysts by deposition–precipitation

Active Au catalysts can be prepared using different methods. Coprecipitation, impregnation, and deposition–precipitation are the most commonly used, and they often result in quite different catalysts as shown in Table 1. Typically, coprecipitation or deposition–precipitation methods are more desirable than impregnation, especially when chloroauric acid is used as the precursor. Various processing variables have been studied, which include pH, temperature, and concentration of the preparation

Nature of the active site and reaction mechanism

The nature of the active site for CO oxidation on supported Au catalysts and the reaction mechanism are also subjects of great research interest. Although active catalysts commonly contain small, 2–5 nm Au crystallites, size alone does not seem to be a sufficient factor for high activity (e.g., [17], [19], [25]). There have been a number of proposals of the active site, which include gold-support interface [36], [37], small Au clusters that possess nonmetallic electronic properties due to a

Catalytic activity, deactivation, and effect of water vapor

As mentioned earlier, the activity of a catalyst depends not only on its method of preparation, but also on the manner that it is tested. The rate at which a catalyst attains a steady–state activity during CO oxidation depends on the catalyst. In a feed of 1% CO and 2.5% O2 in He that was purified by passing through a silica trap at dry-ice temperature, we found that a Au/TiO2 catalyst can maintain relatively stable activity. On the other hand, a Au/Al2O3 catalyst loses its activity rapidly. If

Catalytic activity of uncalcined catalysts

The dependence of catalytic activity of supported gold catalysts on the calcination temperature is a surprisingly complex issue. There are several reports that uncalcined samples or samples calcined at relatively mild temperatures are more active than those calcined at higher temperatures [17], [30], [31], [32], [33], [34], [35], [55], but there is no explanation for these observations that has been accepted unequivocally.

For Au/TiO2 and Au/Co3O4, Wolf and Schuth [17] reported that the highest

Conclusion

The specific requirements of the active site according to the active site model and the sensitivity to chloride poisoning and moisture are among the reasons there are wide variations of reported activities of catalysts of similar compositions in the literature. It is important to interpret experimental results with respect to the amount of residual chloride that may be present in the catalyst or the moisture content of the catalyst and in the reaction test system. However, it still remains to

Acknowledgements

This research was supported by the EMSI program of the NSF and Department of Energy Office of Science (CHE-9810378) at Northwestern University Institute of Environmental Catalysis.

References (56)

  • M. Haruta et al.

    J. Catal.

    (1989)
  • M. Haruta

    Stud. Surf. Sci. Catal.

    (1997)
  • M.C. Kung et al.

    Stud. Surf. Sci. Catal.

    (1996)
  • A. Ueda et al.

    Appl. Catal. B

    (1997)
  • G.K. Bethke et al.

    Appl. Catal. A

    (2000)
  • Y.A. Kalvachev et al.

    J. Catal.

    (1999)
  • T. Hayashi et al.

    J. Catal.

    (1998)
  • M. Okumura et al.

    Catal. Today

    (2002)
  • C.K. Costello et al.

    Appl. Catal. A

    (2002)
  • T.V. Choudhary et al.

    J. Catal.

    (2002)
  • S.-J. Lee et al.

    J. Catal.

    (2002)
  • A. Wolf et al.

    Appl. Catal. A

    (2002)
  • M.A. Bollinger et al.

    Appl. Catal. B

    (1996)
  • A.I. Kozlov et al.

    Appl. Catal. A

    (1999)
  • A.P. Kozlova et al.

    J. Catal.

    (1998)
  • H.-S. Oh et al.

    J. Catal.

    (2002)
  • S. Tsubota et al.
  • Y.J. Chen et al.

    J. Catal.

    (2001)
  • F. Boccuzzi et al.

    J. Catal.

    (2001)
  • E.D. Park et al.

    J. Catal.

    (1999)
  • N.A. Hodge et al.

    Catal. Today

    (2002)
  • S. Minico et al.

    Appl. Catal. B

    (2001)
  • S.-J. Lee et al.

    J. Catal.

    (2001)
  • J.-N. Lin et al.

    Appl. Catal. B

    (2002)
  • M. Haruta et al.

    J. Catal.

    (1993)
  • H.-S. Oh et al.

    Stud. Surf. Sci. Catal.

    (2001)
  • R.J.H. Grisel et al.

    Catal. Today

    (2002)
  • J.L. Margitfalvi et al.

    Catal. Today

    (2002)
  • Cited by (443)

    View all citing articles on Scopus
    View full text