Elsevier

Applied Catalysis B: Environmental

Volume 237, 5 December 2018, Pages 986-995
Applied Catalysis B: Environmental

CO/H2 adsorption on a Ru/Al2O3 model catalyst for Fischer Trospch: Effect of water concentration on the surface species

https://doi.org/10.1016/j.apcatb.2018.06.053Get rights and content

Highlights

  • Surface species formed after adsorption of a wet mixture (CO+H2) strongly depends on the water concentration in this mixture.

  • Water concentrations close to monolayer provoke the surface carbon gasification, increasing the amount of adsorbed CO.

  • Water in high concentrations (∼13%) influences the CO diffusion into the pores resulting in a low amount of adsorbed CO.

Abstract

Water presence and concentration strongly influence CO conversion and C5+ selectivity in the Fischer Tropsch reaction. In this work, the influence of the water concentration was investigated using a model Ru/Al2O3 (5 wt.%) catalyst. The surface species formed after CO and H2 adsorption in dry and wet (different water concentrations) conditions were analyzed by FTIR. Firstly, water adsorption was carried out up to complete filling of the pores and then CO was put in contact with the catalyst. The absence of adsorbed CO species in these conditions evidences that CO diffusion in water controls the access of the gas to the active sites and explains the negative effect of high water concentrations reported by some authors. Moreover, the adsorption of a mixture of CO+H2+H2O, being the water concentration close to that needed to have a monolayer, and a dry mixture of CO+H2 were carried out and compared. Results evidence that water in this low concentration, is able to gasify the surface carbon species formed by CO dissociation on the metallic sites. This cleaning effect is related to the positive effect of water on CO conversion detected by some authors.

Graphical abstract

Water concentration strongly influences the surface species on a model catalyst for the Fischer Tropsch reaction.

  1. Download : Download high-res image (107KB)
  2. Download : Download full-size image

Introduction

The dependence of our current energy system on fossil fuels and their harmful effects on the environment have strengthen the development of renewable energy sources. This is the case of the second generation biofuels. The production of fuels from lignocellulosic biomass and wastes often involves catalytic processes, among these Fisher-Tropsch synthesis (FTS) is particularly important [1]. Carbon life-cycle assessment of these second-generation fuels entails important reduction of CO2 emissions without extensive modification of the currently used engines, or fuel formulation and distribution, which is especially important in the automotive field. In this scenario, exploitation of different types of biomass takes special importance [2]. In a first step, biomass is gasified to syngas that further feeds a Fischer-Tropsch reactor, where a mixture of liquid hydrocarbons (C5+) that may potentially be used in the same way as oil derivatives, is obtained [3].

Although the FTS is known since 1920 and industrial plants are currently under operation, improvements of the reactors and catalysts are still needed. Microchannel reactors, that may overcome diffusion limitations and heat transport effects on the selectivity are particularly interesting [[4], [5], [6]]. Moreover, the knowledge of the reaction mechanism may help to improve the global process and to design more performant catalytic solids. It is well accepted that the mechanism of FTS proceeds according to three main steps: (i) the initiation in which the CO molecule is dissociated and CHx fragments are formed by hydrogenation, (ii) the propagation step with the formation of CHx-CHx species and (iii) chain termination and hydrocarbons desorption as a final step [[7], [8], [9]]. However, some aspects of the mechanism of this reaction are still under investigation [8,10,11].

The “carbide mechanism” was early proposed by Fischer and Tropsch. In this mechanism, CO is directly dissociated on the metal surface resulting in C* and O* adsorbed species, that are successively hydrogenated and initiate the chain growth [12]. However, according to recent studies, CO dissociation assisted by hydrogen must also be considered as an intermediate step for this reaction. A lower energy barrier for the CO dissociation in the presence of hydrogen was found on clean and H-precovered Fe (111) surface by Hou et al. [13]. These authors propose an adsorbed CHO intermediate for the H-assisted CO dissociation. Eq. (1) summarizes the proposed reaction path (X* = adsorbed species):CO*+2H*CHO*+H*CH*+O*+H*

Both pathways (direct CO dissociation and H-assisted activation) have been evidenced for CO activation on cobalt catalysts [10], although the H-assisted activation of CO was proposed as the most favorable path [14]. Loveless et al. [15] studied the Ru-catalyzed FTS trying to unravel the mechanism for CO activation, either the direct C–O bond activation on vicinal vacant sites or the H-assisted activation of CO via its reaction with coadsorbed “H” atoms. Their results allow them conclude that CO is activated on Ru(111) terraces via the H-assisted path. The CO activation on Ru terraces suggests a strong dependence of the catalytic activity on the metallic particle morphology, as previously reported in the literature [16].

CO conversion in FTS strongly decreases with particle size (lower activity for smaller metallic particles) when metallic clusters are smaller than 10 nm [17,18]. For larger particles (>10 nm), the CO conversion is not affected by the particle size. In the case of Co-based catalysts, the lower activity of smaller particles has been attributed to the formation of partially oxidized species. However, when Ru is used as active phase, the behavior of particles smaller than 10 nm is explained in terms of CO adsorption. The low coordination atoms, prevalent in small particles, are less active or that for these atoms the CO adsorption energy is too high. Moreover, large metallic particles improve the selectivity since step sites (numerous in small particles) inhibit chain growth [19]. Water, an unavoidable byproduct in the FTS, has a significant effect on the catalyst activity and selectivity. This effect, however, depends on the metal and support nature, the catalyst composition, and/or the preparation method used [20]. The effect of water on the CO conversion is not very clear from literature, but authors always agree on the improvement in C5+ selectivity by the presence of water either formed during reaction or added to the reactive mixture [21,22]. Most studies show that CO conversion increases in the presence of water on Co- [23,24] or Ru-catalyzed [25] FTS. However, catalyst oxidation in the presence of water has been proposed as responsible for a reduction of CO conversion on Fe [25] or Co catalysts [27,28]. In this latter case, the effect is particularly important for metal particle sizes below 4 nm [29]. Water may also affect the support: Storsæter et al. [30] report the effect of the presence of water on the support, which is especially important in the case of alumina, where water irreversibly deactivates the solid. Moreover, it has been reported that the formation of mixed oxide phases induced by the presence of water in the reaction stream affects the reducibility of the active phase and induces changes on the activity and selectivity of the metal catalysts [[31], [32], [33], [34]]. Claeys et al. [35] tested a Ru/SiO2 catalyst and found a positive effect of water addition. These authors affirm that water participates in the reaction mechanism, increases the catalyst activity and modifies the product distribution. The formation of methane is suppressed and chain growth favored, probably because water is an additional hydrogen source to the formation of monomers that participate in the chain growth. Other authors explain the water effects in terms of diffusivity of the products into the pores of the solid and to conclude that catalytic activity and products distribution depends on the pore size [36]. For Co-based catalysts, some works found a strong dependence between the water concentration and the pore size [25,37] and attributed this effect to diffusive aspects, because no modification of the surface species was detected after water addition [25]. However, in the presence of water Co-based catalysts deactivate. It has been observed for Co/SiO2 catalysts that water addition to the inlet flow increases the selectivity to C5+, but simultaneously deactivates the process by forming cobalt oxide species [37]. Bertole et al. [38] propose monomeric carbon as the active form responsible for the increase in the selectivity to C5+, this through an increase of the active carbon concentration as a result of the decrease in the activation barrier for CO dissociation in the presence of water [21]. Ruthenium, however, is highly resistant to oxidation and therefore is very well suited for studying the effect of water on the FTS [20].

In this work, the effect of water has been investigated using a Ru/Al2O3 (5 wt.% Ru) working at 150 °C, the lowest temperature at which Ru is active in FTS. A Ru-based catalyst has been selected to carry out this study due to the high activity of this metal even at low temperature and pressure and that our experimental device doesn’t permit to work at high pressure or very high temperatures. Moreover, the proved resistance of these solids to the oxidation by water is an interesting aspect to consider if water effect on the FTS is being investigated [39]. Initially, an isotherm of water adsorption was carried out in order to assure the complete filling of the pores. The CO adsorption was carried out both on the water saturated surface and on the dry surface in order to detect the effect of water at high concentration. Moreover, to observe possible differences in the surface species formed during the process, hydrogenation of CO saturated surface was compared with the surface species formed after adsorption of CO+H2 and CO+H2+H2O mixtures. In this latter case, the water concentration was close to that needed to form a monolayer on the catalyst surface.

Section snippets

Experimental

The catalyst was prepared by wet impregnation. Commercial γ-Al2O3 (Sasol) was impregnated with 5 wt.% Ru, using Ru(NO)(NO3)3 solution (Johnson Matthey, CAS Number 34513-98-9) as precursor. The excess of solvent was eliminated at 100 °C and reduced pressure to dryness. The resulting solid was further air-calcined for 3 h at 400 °C. The calcination temperature was reached by linear heating at 10 °C min−1. The prepared solid was characterized by different techniques. The calcination conditions

Results and Discussion

The prepared catalyst was characterized by N2 adsorption, XRD and TPR. The measured specific surface area of the catalyst is 173 m2/g, the pore volume 0.4 cm3/g and the average pore size 7 nm.

As evidenced by the TPR profile of the prepared catalyst (Fig. 1), 94.5% of ruthenium is reduced below 200 °C. Fig. 1 also presents the XRD patterns of the calcined and reduced catalyst. Together with the diffraction peaks of the alumina support (JCPDS 002-1420), diffraction lines corresponding to RuO2 are

Conclusions

It has been concluded that CO interaction with the surface of a Ru/Al2O3 at 150 °C produces CO2 and adsorbed carbon on the surface by the WGS and Boudouard reactions. Water in low concentrations (close to the monolayer) participates in the carbon gasification, cleaning the surface sites, and may explain the positive effect observed by some authors on the CO conversion in Fischer Trospch conditions.

However, water concentrations high enough to fill the pores of the solid restrict the CO

Acknowledgement

The authors gratefully acknowledge the financial support from the Spanish Ministerio de Economía y Competitividad –MINECO (ENE2015-66975-C3-2R).

References (74)

  • G. Arzamendi et al.

    Chem. Eng. J.

    (2010)
  • V. Sage et al.

    Catal. Today

    (2011)
  • M. Ojeda et al.

    J. Catal.

    (2010)
  • C.F. Huo et al.

    J. Catal.

    (2007)
  • G. Prieto et al.

    J. Catal.

    (2009)
  • C.J. Bertole et al.

    J. Catal.

    (2002)
  • S. Krishnamoorthy et al.

    J. Catal.

    (2002)
  • R.A. Van Santen et al.

    Adv. Catal.

    (2011)
  • E. Iglesia

    Appl. Catal. A Gen

    (1997)
  • S. Storsæter et al.

    J. Catal.

    (2005)
  • Y. Zhang et al.

    J. Catal.

    (1999)
  • W. Ma et al.

    Catal. Today

    (2014)
  • M. Lualdi et al.

    Catal. Today

    (2013)
  • M. Claeys et al.

    Catal. Today

    (2002)
  • A.K. Dalai et al.

    Appl. Catal. A Gen.

    (2008)
  • C.J. Bertole et al.

    J. Catal.

    (2004)
  • L. Liu et al.

    Catal. Today

    (2012)
  • K. Hadjiivanov et al.

    J. Catal.

    (1998)
  • O. Dulaurent et al.

    Appl. Catal. A

    (2000)
  • H. Wise et al.

    Surf. Sci.

    (1983)
  • G.H. Yokomizo et al.

    J. Catal.

    (1989)
  • G. Held et al.

    Surf. Sci.

    (1994)
  • M. Nakamura et al.

    Chem. Phys. Lett.

    (2000)
  • Z. Mei et al.

    Int. J. Hydrogen Energy

    (2014)
  • W. Xu et al.

    J. Catal.

    (2012)
  • C. Morterra et al.

    Catal. Today

    (1996)
  • E. Guglielminotti et al.

    J. Catal.

    (1994)
  • J.M.G. Carballo et al.

    Catal. Today

    (2013)
  • D.L. King

    J. Catal.

    (1978)
  • A.Y. Khodakov et al.

    Chem. Rev.

    (2007)
  • M.L. Savaliya et al.

    Res. Chem. Int.

    (2015)
  • F.E.M. Farias et al.

    Lat. Am. Appl. Res.

    (2007)
  • G.C. Visconti et al.

    Chem. Eng. J.

    (2011)
  • L.C. Almeida et al.

    Washcoating of metallic monoliths and microchannel reactors

    Elsevier Masson SAS

    (2010)
  • O.O. James et al.

    RSC Adv.

    (2012)
  • S. Shetty et al.

    J. Am. Chem. Soc.

    (2009)
  • R.A. van Santen et al.

    ChemCatChem

    (2013)
  • Cited by (24)

    • Hydrogen production from ammonia decomposition over Ru-rich surface on La<inf>2</inf>O<inf>2</inf>CO<inf>3</inf>-Al<inf>2</inf>O<inf>3</inf> catalyst beads

      2023, Catalysis Today
      Citation Excerpt :

      These types of experiments are useful to estimate the extent of the metal-support interaction by considering shifts in the reduction temperature [9]. As shown in Fig. 8, the reduction of Ru/Al2O3 occurred as a single broad peak at 118 °C assigned to the reduction of RuOx to metallic Ru [41,61] indicating that the support was composed of a single component of alumina. The Ru/La2O3-Al2O3 catalyst exhibited two major reduction peaks at 131 and 209 °C.

    View all citing articles on Scopus
    View full text