Elsevier

Catalysis Today

Volume 186, Issue 1, 1 June 2012, Pages 54-62
Catalysis Today

Effect of Ag on the control of Ni-catalyzed carbon formation: A density functional theory study

https://doi.org/10.1016/j.cattod.2011.08.041Get rights and content

Abstract

First-principles calculations have been performed to examine the effect of doped Ag on the kinetics of Ni-catalyzed methane dissociation and coke formation. The close-packed Ag/Ni(1 1 1) and stepped Ag/Ni(2 1 1) surfaces as well as the defect facets with step sites blocked by Ag or C atoms are constructed to investigate the role of the coordinatively unsaturated sites in the catalytic performance of Ni nanoparticles. The most stable CHx (x = 0–4) adsorption configurations and transition states for methane dissociation have been identified on both Ni and Ag/Ni surfaces. The calculated results indicate that the activation energy for methane dissociation is increased with the Ag coverage on Ni(1 1 1), and the C atoms deposited on the catalyst surface can be readily separated into small islands by Ag. On Ni(2 1 1) Ag atoms are predicted to bind preferentially to the middle-step sites which act as the nucleation center for the growth of filamentous carbon and therefore have the potential to prevent catalyst particles from being destroyed. Meanwhile, as the energy barrier for methane dissociation on the Ag-blocked Ni(2 1 1) surface is even higher than that on pure Ni(1 1 1), the active center is transferred from the stepped surface to the close-packed surface. These findings provide a rational interpretation of the experimental observations that Ag/Ni catalyst exhibits lower catalytic activity towards steam methane reforming but high resistance to coke deposition.

Graphical abstract

DFT calculations have been performed to examine the effect of doped Ag on Ni-catalyzed methane dissociation and coke formation. The doped Ag is likely to suppress the formation of filamentous carbon. The active center is transferred from the stepped surface to the close-packed surface once Ag is introduced into Ni catalyst.

  1. Download : Download high-res image (153KB)
  2. Download : Download full-size image

Highlights

► The effect of Ag on Ni-catalyzed coke formation is investigated. ► Methane activation is hindered with the increase of the Ag coverage. ► Step sites dominate methane activation and coke formation. ► The active center is transferred from Ni(2 1 1) to Ni(1 1 1) with the Ag introduction.

Introduction

It has long been recognized that there are three main types of coke deposition obstructing the proceeding of steam reforming of hydrocarbons: polymeric, filamentous carbon (often referred to as whisker-like carbon or carbon nanofiber), and graphitic carbon [1]. The polymeric form derives from the gas-phase decomposition of hydrocarbons, whereas the filamentous and graphitic forms require the participation of metal surfaces [2]. In general, the polymeric form can be removed under oxygen atmosphere at relatively low temperatures, while the filamentous and graphitic types are more refractory and higher oxidation temperatures are required.

Filamentous carbon, is the primary product of carbon formation in transition metal-catalyzed steam methane reforming [3]. The fibers have the capability of modifying the catalyst structure on a microscopic scale and further destroying the catalyst pellets completely [4]. Moreover, they were also found to attack and rupture the reactor walls in the industrial practice. As the formation of filamentous carbon cannot be tolerated in a tubular reformer, there is a strong motivation for developing a new catalyst with less coke deposition [5].

Ni is among the most attractive catalysts both for the activation of C–H bond and for the synthesis of new carbon materials such as carbon nanotube and carbon nanofiber, because of its relatively high catalytic activity and low cost. Thus, the mechanism for the catalyst deactivation in Ni-catalyzed steam methane reforming is closely related to the growth of filamentous carbon [6]. It was recently reported that the Ni step sites act as the active center for the nucleation of graphene sheet [7]. To suppress the growth of filamentous carbon, one approach is connected with the modification of the Ni superficial structure through the formation of Ni-based alloys. Several combinations, such as Au/Ni [8], Ag/Ni [9], Cu/Ni [10], B/Ni [11], K/Ni [12], Sn/Ni [13], and La/Ni [14], have been screened by both experimental and theoretical means, and all the secondary elements were expected to promote the long-term stability of Ni catalyst. It was suggested that the principal reason for the observed suppression of coke formation is the preferential blockage of the step sites by the additives [7], [12], [15]. In particular, Ag was found to exhibit distinctive properties to eliminate the growth of filamentous carbon [9].

Experimentally, Parizotto et al. [9] investigated the effect of Ag loading on the catalytic behavior of Al2O3-supported Ni catalyst in steam methane reforming. The results showed that the doped Ag strongly modifies the surface properties and suppresses the graphitic carbon deposition. A large amount of surface and filamentous carbon is formed on unpromoted Ni catalyst while just a thin layer of amorphous carbon is deposited on the 0.3 wt.% Ag-promoted Ni catalyst without the formation of filamentous carbon. Then, Gavrielatos et al. [16] studied carbon deposition on the Ag-doped Ni/YSZ electrode under internal steam methane reforming by thermogravimetric, catalytic, and electrocatalytic experiments. The activity of Ni/YSZ towards carbon deposition is lowered substantially by adding 1–5 at.% Ag on Ni, indicating that the hydrogenolysis and coking reactions are significantly suppressed by the co-impregnation of Ag and Ni. More recently, Rovik et al. [17] investigated the structure sensitivity of different X/Ag (X = Ni, Ru, Rh, and Pd) alloys in ethane conversion. On the Ag–Ni/spinel catalyst, graphene sheet and a small amount of carbon nanotube were observed, but no filamentous carbon was produced. The lack of filamentous carbon and the reduced amount of graphene sheet were considered as the reason for the enhanced stability of Ni catalyst. Jeong and Kang [18] studied steam reforming of butane over the Ag(1)/Ni(9)/MgAlO3 catalyst and found that the Ag-loaded catalyst can reduce the degree of carbon deposition. They attributed the improved coke resistance to the Ag-induced depressions in the NiAlO3 spinel structure.

As the introduction of Ag into Ni catalyst has a positive effect on the suppression of coke formation, a comprehensive investigation of the role of Ag/Ni alloy in steam methane reforming may shed light on the interaction between the promoters and step sites and help to design new catalysts with higher coke resistance. In this contribution, density functional theory (DFT) calculations have been carried out to investigate the effect of Ag on the kinetics of methane dissociation and the mechanism for coke formation over the flat Ag/Ni(1 1 1) and stepped Ag/Ni(2 1 1) surfaces. In Section 2, the computational method is detailed. The calculated adsorption energies of intermediates and activation energies for each elementary step are given in Section 3 and are compared with previous experimental and theoretical data. In Section 4, we conclude by discussing the implication of our results for understanding the effect of Ag on the catalytic activity of Ni towards CH4 dissociation and the coking mechanism.

Section snippets

Computational details

Self-consistent DFT calculations were performed using the VASP code [19], [20], [21], [22], in which the interactions between valence electrons and ion cores are described by pseudopotentials and the electronic wave functions at each k-point are expanded in terms of a discrete plane-wave basis set. Here we used the Blöchl's all-electron-like projector augmented wave (PAW) method [23], [24], which is superior to the ultrasoft pseudopotentials (USPs) for accurate calculations of certain

CHx adsorption on Ag/Ni(1 1 1) and Ag/Ni(2 1 1)

The adsorption of CHx (x = 4–0) on the flat (1 1 1) and stepped (2 1 1) surfaces with different Ag coverages (θAg) was first examined. The geometries of the most stable adsorption configurations on the flat and stepped surfaces are shown in Fig. 3, Fig. 4, respectively, and the corresponding adsorption energies are listed in Table 2.

According to our previous calculations on Ni(1 1 1), the reaction intermediates are predominantly adsorbed at the threefold hollow sites [29]. Therefore, the hcp and fcc

Conclusions

Spin-polarized DFT–GGA calculations have been performed to investigate methane dissociation and coke formation on alloyed Ag/Ni catalyst. At low Ag coverages, the introduction of Ag into Ni(1 1 1) has a minor effect on the binding strength of the reaction intermediates involved in methane dissociation. As the Ag coverage is increased to 1/4 ML, the adsorption energies of CH3 and CH2 are lowered while the interaction between CH (C) and the metal surface is strengthened. The Ag–Ni–Ni threefold

Acknowledgements

This research is supported by Natural Science Foundation of China (No. 21003046), 973 project of Ministry of Science and Technology of China (No. 2012CB720500), and Fundamental Research Funds for Central Universities (No. WA1014027). The computational time provided by the Notur project is highly acknowledged (nn4685k).

References (43)

  • J.R. Rostrup-Nielsen

    Surface science and catalysis

  • R.L. Mieville

    J. Catal.

    (1986)
  • J. Rostrup-Nielsen

    J. Catal.

    (1984)
  • J.R. Rostrup-Nielsen et al.

    Catalyst deactivation 2001

    Proceedings

    (2001)
  • J. Xu et al.

    J. Catal.

    (2006)
  • N.C. Triantafyllopoulos et al.

    J. Catal.

    (2006)
  • N.V. Parizotto et al.

    Appl. Catal. A: Gen.

    (2007)
  • J. Xu et al.

    J. Catal.

    (2009)
  • H.S. Bengaard et al.

    J. Catal.

    (2002)
  • E. Nikolla et al.

    J. Catal.

    (2009)
  • J.W.C. Liberatori et al.

    Appl. Catal. A: Gen.

    (2007)
  • A.K. Rovik et al.

    Appl. Catal. A: Gen.

    (2009)
  • H. Jeong et al.

    Appl. Catal. B: Environ.

    (2010)
  • G. Kresse et al.

    Comput. Mater. Sci.

    (1996)
  • Y.A. Zhu et al.

    Catal. Today

    (2009)
  • A. Michaelides et al.

    Surf. Sci.

    (1999)
  • F. Abild-Pedersen et al.

    Surf. Sci.

    (2005)
  • S.D. Robertson

    Nature

    (1969)
  • J. Rostrup-Nielsen et al.

    Top. Catal.

    (2006)
  • W. An et al.

    J. Chem. Phys.

    (2009)
  • R.T. Vang et al.

    Nat. Mater.

    (2005)
  • Cited by (0)

    View full text