Elsevier

Applied Catalysis A: General

Volume 319, 1 March 2007, Pages 202-209
Applied Catalysis A: General

The use of hydrogen chemisorption for the determination of Ru dispersion in Ru/γ-alumina catalysts

https://doi.org/10.1016/j.apcata.2006.12.005Get rights and content

Abstract

Ru nanoparticles supported on γ-Al2O3, prepared by reduction of RuCl3 in ethylene glycol with using microwave irradiation, were thoroughly characterized by ICP-AES, BET, XRD, TEM, XPS and H2 chemisorption. Structure and chemisorptive properties of the supported colloidal Ru nanoparticles were compared with those of the Ru/γ-Al2O3 catalyst prepared by the incipient wetness impregnation method, using the same RuCl3 precursor. Results obtained by volumetric H2 chemisorption performed at different temperatures 20–200 °C, showed that the irreversible H2 adsorption onto colloidal Ru/γ-Al2O3 catalyst is a nonactivated process. For this catalyst saturation of the ruthenium surface with hydrogen was achieved already at room temperature. Ruthenium dispersion and particle sizes obtained from H2 measurements agreed well with the results of TEM and XRD methods. In contrast, activated chemisorption behavior has been observed on traditionally prepared Ru/γ-Al2O3 catalyst. Even at 100 °C, irreversible H2 uptake was lower then expected as evidenced by the large discrepancies between the mean particle sizes obtained from H2 chemisorption and TEM. Superior chemisorptive properties of the Ru nanoparticles supported on γ-Al2O3 are assigned to the higher dispersion of the smaller ruthenium particles and also their weaker interaction with the support. The colloidal catalyst, in contrast to traditionally prepared one, is free of chlorine contamination. Also, contamination of Ru by aluminum ions, possible during the catalyst preparation by conventional impregnation with acidic solution, was avoided.

Graphical abstract

Ru nanoparticles on γ-Al2O3 were prepared by reduction of RuCl3 in ethylene glycol with using microwave irradiation. Chemisorptive properties of this system were compared with those of the Ru/γ-Al2O3 catalyst prepared by impregnation. Superior chemisorptive properties of the former system are assigned to the higher dispersion of the smaller ruthenium particles and weaker interaction of Ru nanoparticles with the support.

Introduction

There is a growing interest in understanding the adsorptive and catalytic properties of ruthenium catalysts due to their very high activity for ammonia synthesis [1], [2]. Supported Ru catalysts can also be used in other industrial processes, such as the F–T synthesis of paraffins, methanation of CO or in the partial hydrogenation of benzene to cycloheksene [3], [4]. The conventional ruthenium-based catalysts are usually prepared by impregnation of the supports with an aqueous solution of RuCl3·3H2O [1], [2], [3], [4], [5], [6], [7], [8], [9], [10], [11], or other precursors, if Cl retention in the catalyst is to be avoided [2], [4], [12], [13], [14], [15], [16], [17]. The impregnation method is simple but has severe drawbacks, most important being not uniform distribution of metal particles in size and shape [18]. For Ru/Al2O3 catalysts prepared by impregnation metal dispersion is generally low, and additionally, if chlorine precursor is used, contamination of the surface of metal with chlorine occurs [5], [6], [9], [11]. Another source of metal contamination in such system may be alumina support dissolved in highly acidic solution used for impregnation [19].

An alternative method for obtaining supported catalysts with well-defined metal particles is that using metal colloids. Polyol reduction method has been widely applied as an effective way to synthesize the highly dispersed metal colloids [20], but only few studies were devoted to ruthenium [20], [21], [22], [23], [24]. There are also only few studies concerning the deposition of ruthenium nanoparticles on the supports [21], [25], [26], [27]. Miyazaki et al. [21], [25] prepared Ru/Al2O3 catalysts by deposition colloidal particles obtained by reduction of RuCl3 in ethylene glycol. It was found that almost all of the ruthenium ions could be reduced to metallic state by the polyol agent, thus, the chlorine ions remained in solution could be easily removed.

Dispersion of ruthenium in traditionally prepared supported catalysts has been studied repeatedly during the past 30 years. However, there is no universally accepted method to measure the exposed ruthenium surface area, but selective gas chemisorption (H2, O2 and CO) is mostly used for this purpose. H2 chemisorption has been extensively applied [1], [3], [4], [5], [6], [7], [8], [9], [10], [11], [12], [13], [14], [15], [16], because the chemisorption stoichiometry H/Rus = 1/1 is well established [28], [29], [30]. In the case of oxygen adsorption the surface stoichiometry is more complex and O/Rus values of 1/2 [31], 1/1 [32], [33], [34] or 2/1 [8], [29], [30], [35] were proposed. Moreover, it was shown in few papers that the stoichiometry for O2 adsorption varies with Ru particle size [33], [36]. The formation of oxygen subsurface and a progress of the chemisorption to a bulk oxidation can lead to overestimation, especially in the case of small particles [33], [36]. CO chemisorption has also been frequently used to determine the dispersion of supported ruthenium catalyst [19], [28], [32], [37], however, its application has been questioned because of its tendency to form multiple adsorption bonds, the number of which depends on surface morphology [33]. Moreover, there is a suggestion that CO oxidizes Ru0, i.e. corrosive adsorption of CO takes place [38]. It was reported also that the adsorption of CO on the supported ruthenium catalyst causes disruption of some Ru–Ru bonds in the metal clusters increasing the dispersion of ruthenium [38], [39].

In contrast to the selective chemisorption of O2 and CO, it was reported that the stoichiometry of H2 chemisorption on Ru is not particle size dependent [30], provided equilibrium is established and only the quantity of irreversibly chemisorbed gas is considered. However, there are still problems and inconsistencies concerning the conditions for saturation of a hydrogen monolayer corresponding to H/Rus = 1/1. According to many authors the time needed to reach the equilibration is 1–2 h, though a period of 4–24 h was required at lower pressures and at room temperature [5], [6], [28], [30]. More recent studies have shown that the chemisorption equilibrium can be reached much quicker, in just 10 or even 5 min at temperature of 60–90 or 100 °C [11], [13], [16]. Several effects complicating the utilization of H2 chemisorption to determine Ru dispersion, such as: hydrogen spillover from Ru to the support at elevated temperature [5], [13], formation of subsurface hydrogen species [40], and presence of a “hydrogen fog” over the ruthenium particles [12] have been also reported.

In this work a new method of preparation of the colloidal Ru/γ-Al2O3 catalyst, based on a solvothermal process with using microwave irradiation, has been applied. Thanks to the microwave irradiation the heating time was much shorter than in traditional polyol methods, such as an oil bath used by Miyazaki et al. [21], [25] and thus smaller, more uniform Ru particles could be obtained. The catalyst, in the “as prepared state”, as well as after thermal heating in hydrogen, was characterized by various physical methods including BET, XRD, XPS and TEM. Such well-characterized, model catalyst was used as an object of thorough studies of the H2 chemisorption. The main goal was re-evaluation of existing data and determination optimal procedures for using H2 chemisorption as reliable tool for proper measurement the fraction of surface Ru atoms in highly dispersed catalysts. In this study, static volumetric H2 chemisorption method was applied. In order to evaluate the effect of the preparation method and thus possible contaminations on H2 chemisorption a conventional catalyst with similar loading of the metal, prepared by incipient wetness impregnation with RuCl3 solution was also studied.

Section snippets

Experimental

RuCl3·3H2O (Koch & Ligth) and γ-Al2O3 (K, Fe, Mg and Si < 10−3%) with BET surface area of 245.9 m2/g and pore volume of 0.21 cm3/g, after previous calcination in air at 550 °C for 20 h, were used for catalysts preparation. The support was prepared in another laboratory by hydrolysis of aluminum isopropoxide and Cl ions were not detected by XPS and chemical analysis.

Two preparation methods of the catalysts were applied. The first – the polyol reduction method – consisted of the reduction of RuCl3

Overall characteristics of the catalysts

Table 1 summarizes the physicochemical and textural properties of the catalysts. The Ru content of 5.1 and 4.6 wt.% for colloidal and impregnated Ru/γ-Al2O3 catalyst, respectively, did not change upon heat treatment procedures. Since the catalysts were prepared from RuCl3·3H2O precursor we first determined an overall and surface content of residual chlorine ions. Fig. 1 shows X-ray photoelectron spectra (Cl 2p line) of the as-prepared colloidal Ru/Al2O3 catalyst, the impregnated catalyst reduced

Conclusions

Good correspondence between H2 chemisorption and XRD and TEM data, obtained in this work for the model, colloidal system, indicates that traditional volumetric H2 chemisorption at room temperature or at 100 °C, can be used to measure accurately the dispersion of Ru in Ru/γ-Al2O3 catalysts. The model, highly dispersed Ru catalyst, prepared from RuCl3 precursor by the microwave-polyol method, is free of chlorine contamination and thus the interfering effect of chlorine on the gas chemisorption

Acknowledgements

This work was financially supported by Polish Committee for Scientific Research (Grant No. 1 T09B 084 30). The authors thanks A. Cielecka for chemisorption measurements.

References (50)

  • W. Raróg-Pilecka et al.

    J. Catal.

    (2005)
  • M.C.J. Bradford et al.

    J. Catal.

    (1999)
  • C. Milone et al.

    J. Catal.

    (1996)
  • K. Lu et al.

    J. Catal.

    (1987)
  • K. Lu et al.

    J. Catal.

    (1987)
  • S. Murata et al.

    Appl. Catal. A

    (1992)
  • S.K. Masthan et al.

    J. Catal.

    (1990)
  • M.C.J. Bradford et al.

    J. Catal.

    (1997)
  • P. Betancourt et al.

    Appl. Catal. A

    (1998)
  • V. Mazzieri et al.

    Appl. Surf. Sci.

    (2003)
  • D.O. Uner et al.

    J. Catal.

    (1995)
  • B. Chen et al.

    J. Catal.

    (1996)
  • C. Zupanc et al.

    J. Catal.

    (2002)
  • S.E. Siporin et al.

    J. Catal.

    (2004)
  • M. Nawdali et al.

    Appl. Catal. A

    (2002)
  • J.A. Mieth et al.

    J. Catal.

    (1989)
  • A. Miyazaki et al.

    J. Catal.

    (2001)
  • I. Balint et al.

    J. Catal.

    (2002)
  • J. Zhang et al.

    Catal. Commun.

    (2006)
  • J.G. Goodwin

    J. Catal.

    (1981)
  • K.C. Taylor

    J. Catal.

    (1975)
  • I. Rossetti et al.

    Appl. Catal. A

    (2003)
  • M.A. Vannice

    J. Catal.

    (1975)
  • J.T. Yates et al.

    Surf. Sci.

    (1985)
  • J.R. Regalbuto et al.

    J. Catal.

    (1999)
  • Cited by (92)

    • Regulating the nanoscale intimacy of metal and acidic sites in Ru/γ-Al<inf>2</inf>O<inf>3</inf> for the selective conversions of lignin-derived phenols to jet fuels

      2022, Journal of Energy Chemistry
      Citation Excerpt :

      Then, the FT-IR spectra were collected after pyridine adsorption for 1 h. The amounts of Brønsted and/or Lewis acidic sites were determined from pyridine adsorption FT-IR using the molar extinction coefficients from Emeis [36]. where CB is the amounts of Brønsted acidic sites and CL is the amounts of Lewis acidic sites in mmol g−1, IMECB and IMECL are molar extinction coefficient of 1.67 and 2.22 cmμmol−1, respectively, IA(B) is integrated area of band at 1540 cm−1 and IA(L) is integrated area of band at 1450 cm−1, R and W are the radius (cm) and weight (mg) of wafer, respectively.

    • Utilizing Cl coordination to facilitate Ru-Ag self-assembling into alloy and recover thermally-inactivated catalyst for propane combustion

      2021, Applied Catalysis B: Environmental
      Citation Excerpt :

      Then, the adsorbed oxygen can be consumed by H2 pulse tiration as reaction: Ags-Oa + H2 → Ags + H2O [38]. The H2 chemisorption on Ru is an activated process demanding elevated temperature and equilibrium time according to literature [39,40]. In order to measure chemical dispersion of Ag, the temperature for H2 pulse titration is set at 50 °C for inhibition of H2 chemisorption on Ru.

    View all citing articles on Scopus
    View full text