Elsevier

Journal of Catalysis

Volume 237, Issue 2, 25 January 2006, Pages 431-434
Journal of Catalysis

Research note
Quantitative determination of gold active sites by chemisorption and by infrared measurements of adsorbed CO

https://doi.org/10.1016/j.jcat.2005.11.026Get rights and content

Abstract

Quantitative measurements of CO chemisorption in the range 140–180 K, supported by FTIR data on adsorbed CO, were performed on Au/TiO2, Au/Fe2O3, and Au/CeO2 catalysts. On the first two samples, which had similar particle size distributions, an average Au/CO chemisorption stoichiometry of about 3, referred to step-edge Au atoms, was found. On Au/CeO2, where very small clusters and quite large particles are present, the CO-chemisorbed volume was much higher than expected, due to the prevailing contribution of very small Au clusters. On the same sample, a change in the IR absorption coefficient was observed and was reasonably explained.

Introduction

The catalytic properties of finely dispersed gold particles have attracted much attention of late [1]. Bulk gold itself is an inert material, and the reason for small gold particle activity is still a matter of debate. It has been shown that the gold's catalytic activity depends critically on the preparation method, the support type, and the pretreatment procedure. The most popular explanation for the variability of the catalytic properties of gold catalysts focuses on the size of the gold particles and on the amount of low-coordinated gold sites. Other factors have also been considered in the recent literature, including the metal–support interface, charge transfer from the support, and metal cationic sites [1]. In an Au/TiO2 model system, Goodman [1] showed that active Au spread on titania nucleating on reduced Ti defects, forming bilayer structures and electron-rich gold sites. However, the real catalysts are less well-defined than the model ones discussed in Goodman's work: therefore, well-defined and widely accessible characterization methods are needed to improve our understanding of the origin of the different activities observed for different samples and to enable us to discriminate between conflicting explanations. As already mentioned, the most popular explanation for the variability of the catalytic properties of gold catalysts is the amount of low-coordinated gold sites. The concentration of low-coordinated sites is usually related to the size of gold metallic particles, which in most work is determined by transmission electron microscopy (TEM). When performed accurately, TEM provides a particle size distribution from which a mean size can be derived, but because very small metallic particles (<1 nm) are barely detectable, the method is not ideally applicable for samples with highly dispersed gold. In these cases, other techniques, such as EXAFS and XANES, can be used to obtain information on the coordination number of gold. But these alternative measurement techniques are not yet in wide use and thus cannot be proposed as routine tests.

Easier, faster, and less expensive than TEM, chemisorption is a widely used technique, especially in industrial laboratories, for measuring metal particle size in supported catalysts. But despite the increasing attention given to gold-based catalysts over the past 15 years, characterization of supported gold by chemisorption methods has not been as widely investigated as common applications of H2, O2, CO, and N2O chemisorption for characterizing other metals have been. In fact, gold surfaces do not readily chemisorb many molecules. Only few papers have reported O2 and CO chemisorption on supported gold [2], [3], [4]. Recently, the static oxygen adsorption and hydrogen pulse titration of O2 chemisorbed on Au/Al2O3 have been reported [5] to determine surface area and average particle size, although at an unusual temperature (323–473 K). Margitfalvi et al. [6] attempted to measure CO chemisorption on supported gold catalysts using volumetric equipment. Due to the low sticking coefficient of CO on gold, these authors used an unusually high pressure for the chemisorption measurements. Apart from the long time required for the test, they were faced with problems of physical adsorption. Both inconveniences can be overcome by using the pulse-flow technique, provided that chemisorption is fast enough at the temperature of chemisorption.

In the present work, CO chemisorption by a pulse-flow technique and Fourier transform infrared measurements of adsorbed CO in well-defined and well-controlled conditions of temperature and pressure are proposed as widely accessible and reproducible methods that can be used in both academic and industrial laboratories to determine the concentration of gold active sites and to compare different samples.

Section snippets

Experimental

Reference catalysts 1.51 wt% Au/TiO2 and 4.48 wt% Au/Fe2O3, provided by the World Gold Council, were examined. As reported in the catalyst data sheets of World Gold Council, these catalysts expose gold particles with mean diameter and standard deviation (as determined by TEM) of 3.8±1.5nm for Au/TiO2 and of 3.7±0.9nm for Au/Fe2O3. Moreover, a 3.0 wt% Au/CeO2 [7], prepared by deposition–precipitation of gold hydroxide, was studied. High-resolution TEM analysis and EDS measurements were

Results and discussion

The CO pulse chemisorption and spectroscopic experiments were performed on mildly reduced and hydrated samples to avoid CO chemisorption on uncoordinated support ions. CO chemisorption was preliminary investigated in a wide temperature range (77–273 K), and the 140–180 K range was found to be suitable and proper. The results are reported in Table 1. For the two reference samples (Au/TiO2 and Au/Fe2O3), the same value of 0.03 mol of CO per mol of total Au was found. On the other hand,

Acknowledgements

Financial support was provided by MIUR (grants 2004038984_003 and 2004038984_005).

References (20)

  • T. Fukushima et al.

    J. Catal.

    (1979)
  • A.G. Shastri et al.

    J. Catal.

    (1984)
  • Y. Iizuka et al.

    Catal. Today

    (1997)
  • H. Berndt et al.

    Appl. Catal. A: Gen.

    (2003)
  • J.L. Margitfalvi et al.

    Catal. Today

    (2002)
  • T. Tabakova et al.

    Appl. Catal. A

    (2003)
  • P. Hollins

    Surf. Sci. Rep.

    (1992)
  • R.G. Greenler et al.

    Surf. Sci.

    (1985)
  • D.W. Goodman

    Catal. Lett.

    (2005)
  • F. Boccuzzi et al.

    J. Phys. Chem.

    (1996)
There are more references available in the full text version of this article.

Cited by (85)

  • CO oxidation over gold catalysts supported on CuO/Cu<inf>2</inf>O both in O<inf>2</inf>-rich and H<inf>2</inf>-rich streams: Necessity of copper oxide

    2019, Applied Catalysis B: Environmental
    Citation Excerpt :

    Most of gold particles should locate on or close to the surface of Cu2O support; however, it is hard to get a clearer image because these delicate bushes are easily melted in TEM measurement. F. Menegazzo et al. have demonstrated that CO chemisorption performed by a pulse flow system at −116 °C can be taken as a method for the quantitative determination of the gold dispersion on mild reduced and hydrated Au/Fe2O3, Au/TiO2 and Au/ZrO2 catalysts [42,43]. In the present work we therefore alternatively performed pulse flow CO chemisorption measurements at the temperature of −116 °C over the mild reduced samples of Au/CuOx together with the corresponding CuOx supports as references to the TEM observations.

View all citing articles on Scopus
View full text