Elsevier

Journal of Catalysis

Volume 250, Issue 1, 15 August 2007, Pages 85-93
Journal of Catalysis

Promotion of the long-term stability of reforming Ni catalysts by surface alloying

https://doi.org/10.1016/j.jcat.2007.04.020Get rights and content

Abstract

Carbon-induced catalyst deactivation is one of the main problems associated with the electrocatalytic and catalytic reforming of hydrocarbons over supported Ni catalysts. We have used DFT calculations to study various aspects of carbon chemistry on Ni surfaces. We demonstrate that the carbon tolerance of Ni can be improved by synthesizing Ni-containing surface alloys that, compared to monometallic Ni, preferentially oxidize C atoms rather than form C–C bonds and have a lower thermodynamic driving force, associated with the nucleation of carbon atoms on low-coordinated Ni sites. Using the molecular insights obtained in the DFT calculations, we have identified Sn/Ni surface alloy as a potential carbon-tolerant reforming catalyst. The predictions of the DFT calculations were supported by our reactor and catalyst characterization studies, which showed that Sn/Ni is much more resistant to carbon poisoning than monometallic Ni in the steam reforming of methane, propane, and isooctane at moderate steam-to-carbon ratios.

Introduction

Steam reforming (rxn1) is an endothermic catalytic process involving the conversion of hydrocarbons and water into hydrogen and oxidized carbon. The reaction is usually accompanied by a slightly exothermic water-gas shift process (rxn2),CnHm+nH2OnCO+(n+m/2)H2(rxn1),CO+H2OCO2+H2(rxn2). Besides being a crucial process for the catalytic hydrogen production, steam reforming is also important in direct electrochemical hydrocarbon reforming over solid oxide fuel cells (SOFCs) [1], [2]. SOFCs are devices that generate electricity by electrochemically oxidizing various fuels, such as hydrogen, CO, and hydrocarbons. One of the main issues associated with the electrocatalytic and catalytic reforming of hydrocarbons is that commonly used catalysts, such as Ni supported on oxides, deactivate due to the formation of carbon deposits [3], [4], [5], [6]. The carbon-induced catalyst deactivation can be suppressed by increasing the feed steam concentration [3]. High steam concentration is not desirable, because additional heat is required to heat and vaporize water. This approach is also not optimal for SOFCs, because higher inlet steam concentration results in lower energy density [7]. The need for the steam introduction also requires additional system components, complicating fuel cell system integration.

Several attempts have been made to identify carbon-tolerant steam-reforming catalysts [5], [6], [8], [9], [10], [11]. For example, it has been suggested that Ru and Rh do not facilitate the formation of carbon deposits due to poor carbon solubility in these metals [12], [13]. However, Ru and Rh are prohibitively expensive. It has also been shown that small amounts of sulfur can suppress carbon poisoning by blocking carbon nucleation sites [14], [15]. Similarly, quantum chemical density functional theory (DFT) calculations have predicted that Au/Ni surface alloys should exhibit better carbon tolerance than monometallic Ni [16], [17]. These predications have been verified experimentally [6], [16], [17]. In addition, it has also been shown that Cu supported on ceria is a stable electrooxidation catalyst for the internal reforming of hydrocarbons [18], [19], [20]. These materials operate at lower temperatures compared to the Ni-based electrocatalysts due to the inferior thermal stability of Cu, thermal incompatibility of ceria and YSZ, and poor thermal stability of ceria [21], [22], [23], [24].

In this contribution, we have utilized DFT calculations to study the factors that govern the stability of Ni-based reforming catalysts. We demonstrate that monometallic Ni deactivates due to the high rates of C–C bond formation and the high thermodynamic driving force associated with the nucleation and growth of extended carbon structures. DFT calculations also show that Sn/Ni surface alloys, with small amount of Sn alloyed into the Ni surface layers, should be more carbon-tolerant than monometallic Ni. Furthermore, our DFT calculations illustrate that in the limit of low Sn concentrations, the formation energy of the Sn/Ni surface alloy is lower than the formation energies associated with the bulk mixing of Sn and Ni and the separate Sn and Ni phases. The predictions of the DFT calculations are supported by reactor tests performed under the conditions corresponding to the internal on-cell reforming over SOFC anodes. We have also performed various catalyst characterization studies, including scanning electron microscopy (SEM), transmission electron microscopy (TEM), scanning transmission electron microscopy (STEM), energy-dispersive X-ray spectroscopy (EDS), X-ray diffraction (XRD), and X-ray photoelectron spectroscopy (XPS).

Section snippets

Density functional theory

The Dacapo pseudopotentials plane wave code (http://www.camp.dtu.dk) was used for all calculations. Our approach was to find a set of parameters that ensured the relative convergence, that is, the convergence of the respective energies of various structures. With this set of parameters, we have been able to reproduce various properties, such as carbon atom adsorption energies, oxygen adsorption energies, graphene sheet adsorption energy on Ni(111), and an activation barrier for the attachment

DFT studies

The elementary-step mechanism of hydrocarbon steam reforming on Ni has been studied by many investigators [3], [26], [35]. It is generally accepted that a hydrocarbon decomposes over Ni, forming carbon and hydrogen adsorbates [14], [35], [36]. Carbon is removed from the catalyst surface in oxidation reactions forming CO and CO2. Oxidizing agents (O and OH) are formed on Ni in the process of steam activation. In addition to reacting to form CO and CO2, the C atoms and fragments also react with

Catalyst synthesis and characterization

To test the predictions of the DFT calculations, we have synthesized and characterized Ni/YSZ and Sn/Ni/YSZ catalysts. We tested these catalysts in the steam reforming of methane, propane, and isooctane at different steam-to-carbon ratios. Fig. 5a shows a STEM image of an Sn/Ni particle. Elemental mapping of the Sn/Ni particles supported on YSZ shows that at low Sn loadings [1 wt% with respect to Ni (1 wt% Sn/Ni)], the Sn concentration is highest at the boundaries of the particle. This is most

Reactor studies

The synthesized catalysts, monometallic Ni supported on YSZ and 1 wt% Sn/Ni alloy supported on YSZ, were tested in the steam reforming of methane, propane, and isooctane. Fig. 7 shows the normalized conversion of methane (measured conversion divided by the highest measured conversion) over the 1 wt% Sn/Ni/YSZ and Ni/YSZ catalysts at a steam-to-carbon ratio of 0.5 and an operating temperature of 1073 K. Fig. 7 shows that the Ni/YSZ catalyst lost ∼45% of its activity after 2 h of operations,

Conclusion

We have used DFT to obtain molecular insights associated with the formation of extended carbon structures on Ni. We have demonstrated that the carbon tolerance of Ni can be improved by formulating Ni-containing surface alloys that, compared with Ni, preferentially oxidize C atoms rather than form C–C bonds and/or have lower a thermodynamic driving force for carbon nucleation on the low-coordinated sites. Using these molecular insights, we have identified Sn/Ni surface alloy as a potential

Acknowledgements

This work was supported by the DOE-BES, Division of Chemical Sciences (Grant FG-02-05ER15686), the DOE-NETL (Grant FC26-05-NT-42516), and the National Automotive Center. The authors thank SDSC for providing supercomputer time.

References (48)

  • E. Achenbach et al.

    J. Power Sources

    (1994)
  • C.M. Finnerty et al.

    Catal. Today

    (1998)
  • D.L. Trimm

    Catal. Today

    (1999)
  • N.C. Triantafyllopoulos et al.

    J. Catal.

    (2003)
  • D.L. Trimm

    Catal. Today

    (1997)
  • J. Xu et al.

    J. Catal.

    (2006)
  • J.R. Rostrup-Nielsen et al.

    Appl. Catal. A

    (1995)
  • T. Takeguchi et al.

    J. Power Sources

    (2002)
  • J.R. Rostrup-Nielsen et al.

    Recent Adv. Basic Appl. Asp. Ind. Catal.

    (1998)
  • J.J. Strohm et al.

    J. Catal.

    (2006)
  • F. Abild-Pedersen et al.

    Surf. Sci.

    (2005)
  • J.R. Rostrup-Nielsen

    J. Catal.

    (1984)
  • R.J. Gorte et al.

    J. Catal.

    (2003)
  • H. Kim et al.

    J. Electrochem. Soc. A

    (2002)
  • M. Mogensen et al.

    Solid State Ionics

    (2000)
  • A. Atkinson

    Solid State Ionics

    (1997)
  • R.M. Watwe et al.

    J. Catal.

    (2000)
  • S. Linic et al.

    J. Catal.

    (2003)
  • G. Mills et al.

    Surf. Sci.

    (1995)
  • H.S. Bengaard et al.

    J. Catal.

    (2002)
  • J.M. Wei et al.

    J. Catal.

    (2004)
  • J.W. Shabaker et al.

    J. Catal.

    (2005)
  • J.W. Shabaker et al.

    J. Catal.

    (2004)
  • Z.Y. Hou et al.

    Appl. Surf. Sci.

    (2004)
  • Cited by (218)

    • A review on bi/polymetallic catalysts for steam methane reforming

      2023, International Journal of Hydrogen Energy
    View all citing articles on Scopus
    View full text