Elsevier

Applied Surface Science

Volume 321, 1 December 2014, Pages 339-347
Applied Surface Science

A cluster DFT study of NH3 and NO adsorption on the (MoO2)2+/HZSM-5 surface: Lewis versus Brønsted acid sites

https://doi.org/10.1016/j.apsusc.2014.10.048Get rights and content

Highlights

  • NH3 adsorption is found to be more favorable energetically than NO adsorption on both Lewis and Brønsted acid sites.

  • Lewis and Brønsted acid sites are competitive energetically for NH3 adsorption.

  • Reduced-state Mo5+ is suggested to play a key role in adsorption and activation of NOx together with the adsorbed NH4+.

Abstract

A systematic DFT study was carried out to investigate NH3 and NO adsorption on both Lewis and Brønsted acid sites of (MoO2)2+/HZSM-5 catalyst by using cluster models. The adsorption energy results indicate that NH3 could strongly adsorb on both Lewis and Brønsted acid sites in the form of coordinated NH3 and NH4+, respectively, whereas NO represents poorer adsorption ability. It is also found that Lewis and Brønsted acid sites are competitive energetically for NH3 adsorption. According to the difference in the proposed mechanisms for NH3 adsorption on different acid sites, particular attention has been focused on the first dissociation of coordinated NH3 for Lewis acid site and the effect of Mo site on the introduction of NO for Brønsted acid site. For the coordinated NH3 on Lewis acid site, the more electron donation from NH3 is, the greater its adsorption stability is and the higher active its H atoms are. In addition, DOS results show that stability of the H atoms is enhanced by interacting with framework oxygen and especially the H atoms chemical-bonded with framework oxygen. For the NH4+ on Brønsted acid site, reduced-state Mo5+ holds stronger reducibility and oxidizability than terminal oxygen, which is suggested to play a key role in adsorption and activation of NOx together with the adsorbed NH4+.

Introduction

Nitrogen oxides (NOx) from automobile exhaust emissions and industrial processes are the major source of air pollution and several methods have been proposed to meet the current NOx legislation emission limits. It is widely accepted that selective catalytic reduction (SCR) of NOx by NH3 to produce N2 and H2O is the most effective technique being able to reduce NOx emission at ppm levels [1]. The commercial catalysts used today for this SCR reaction are V2O5–MoO3 (WO3)/TiO2, where MoO3 behave as “chemical” promoter for SCR reaction besides playing “structural” promoter for the catalyst [2], [3]. Nevertheless, the high oxidation ability of SO2 to SO3 and the toxicity of V2O5 to the environment and human health prohibit the employment of V2O5 as active center [4], [5], [6]. Furthermore, the use of TiO2 as support is limited by its low resistance to sintering, low surface area and high cost [1]. Therefore, great efforts have been made to develop new catalysts to avoid the above defects.

Most research about SCR catalysts deal with MoOx species as promoter. Focusing growing attention on its high promotional effect, in recent years, catalysts loaded with only MoOx species acting as active component not promoter, such as Mo/HZSM-5 [7], [8], [9], [10], [11], MoO3/TiO2 [3], [12], [13] and MoO3/CeO2 [6], [14] have been applied to study the catalytic performances for SCR reaction of NOx. Wang et al. [7] investigated the C2H2-SCR reaction of NO over Mo/HZSM-5 and the results demonstrated that appropriate amount of Mo incorporation to HZSM-5 considerably enhance the title reaction, both by accelerating the intermediate formation and by strengthening the adsorption of NOx on the catalyst surface. Salgado et al. [8] found that Mo/HZSM-5 catalyst for C2H5OH-SCR of NO is less active but has a high selectivity to N2, which is consistent with NH3-SCR experiments results from Li et al. [9], [10]. Furthermore, Li et al. also pointed out that the catalytic performances of Mo/HZSM-5 for SCR of NO is strongly influenced by different synthesis methods, Si/Al ratio and the Mo content loaded on ZSM-5, which may be related with the surface structure of MoOx species.

However, the structure of MoOx active center of Mo/HZSM-5 catalyst is still a matter of debate in the literature. The difficulties in distinguishing the MoOx active center lie in the low loading of Mo as well as the coexistence of the external surface and the inner-channel MoOx species [15]. By using Raman and X-ray absorption spectroscopy (XAS), Iglesia and co-workers [16], [17], [18], [19] proposed a structure of (Mo2O5)2+ dimer as active center associating with two neighboring Brønsted acid sites, which is consistent with some other studies [20], [21], [22], [23]. However, other groups rather suggested a monomer, either (MoO2)+ or (MoO2)2+ anchored, respectively, on one or two Brønsted acid sites [24], [25], [26], [27], [28], [29], [30], [31]. By using H/D isotope exchange and 27Al MAS NMR and NH3-TPD, Tessonnier et al. [32], [33] claimed that the anchoring mode of MoOx active center is strongly linked with the Si/Al ratio of the starting ZSM-5 zeolite: the (MoO2)2+ monomer at low Si/Al ratio and the (Mo2O5)2+ dimer at high Si/Al ratio. The results of density functional theory (DFT) calculations showed that the (MoO2)2+ monomer and (Mo2O5)2+ dimer may coexist in the actual samples and (MoO2)2+ monomer is more tightly connected with the frameworks than the (Mo2O5)2+ dimer [34]. Therefore, (MoO2)2+ monomer was chosen as the anchoring mode and (MoO2)2+/HZSM-5 was built to represent the Mo/HZSM-5 in this paper.

Adsorption performances of NH3 and NO play an important role in the SCR reaction and their adsorption properties are related to the type of active sites on the catalyst surface. Peng et al. [6] suggested that MoOx species provides MoO3–CeO2 catalysts with Brønsted acid site. Based on the FT-IR data for V2O5-MoO3/TiO2, Busca and co-workers [35] deduced that MoOx species of Mo/TiO2 act as Lewis sites in adsorbing and activating NH3 and the coordinated NH3 will be oxidized to NH2 which later reacts with NO giving adsorbed nitrosamide. Furthermore, NH3-TPD and FT-IR results from Lietti et al. indicated that NH3 could be coordinatively held over Lewis acid sites (associated with Ti, V and Mo surface cation species) and be protonated as NH4+ ions over Mo–OH or V–OH Brønsted sites of V2O5–MoO3/TiO2 catalysts [3]. Unfortunately, a detailed adsorption investigation of NH3 and NO on Mo/HZSM-5 catalyst surfaces with the DFT is absent by far.

In this paper, (MoO2)2+/HZSM-5 was built to represent the Mo/HZSM-5. Aiming to understand the initial step of the NH3-SCR reaction of NO on Mo/HZSM-5 catalyst, DFT approach was performed to investigate how NH3 and NO adsorb on both Lewis and Brønsted acid sites of (MoO2)2+/HZSM-5 catalyst models. And then the electronic properties of adsorption structures were analyzed to predict the next step of the proposed reaction mechanism. We hope that the results may be extended to understand the initial step of the SCR mechanism on other catalysts loaded with only MoOx species acting as active component.

Section snippets

Computational models and details

In the unit cell of ZSM-5, there are 12 distinct tetrahedral Si sites for Al cation substitution, denoted as T sites (T1–T12) [36]. The substitution of Si4+ by Al3+ introduces a negative charge, which is compensated by one H+ and an acidic bridging hydroxyl group (Brønsted acid site) is formed. For ZSM-5 zeolite, there is no experimental information about the preferential site for Si/Al substitution. Several theoretical investigations tried to address this issue but the results did not show

Structure and properties of the (MoO2)2+/HZSM-5 substrate model

The optimized (MoO2)2+/HZSM-5 cluster model is shown in Fig. 2 (denoted as L-model) and corresponding structural parameters are listed in Table 1. The terminal hydrogens in all models are veiled for simplicity. The Mo atom bears tetrahedral symmetry and bridges nearly symmetrically on Si3–O20′–Al12–O24–Si12–O20–Al3. By using LDA/PWC method, the bond lengths between Mo and framework oxygen (OF) atoms in bridge Si3–O20′–Al12, Al12–O24–Si12 and Si12–O20–Al3 are 2.224, 2.036 and 2.122 Å,

Discussion

The adsorption energy results clearly reveal that NH3 adsorption is much more favorable energetically than NO adsorption on both Lewis and Brønsted acid sites. This means that the first step of the SCR reaction on (MoO2)2+/HZSM-5 catalyst is adsorption of NH3 on acid site rather than NO, which is consistent with the researches on other catalysts such as V2O5 [3], [35], [61], [65], [66], [67], Fe/HZSM-5 [68], CuO/γ-Al2O3 [59] and MnOx [69] catalysts. Consequently, it is not further discussed

Conclusions

NH3 and NO adsorption on Lewis and Brønsted acid sites of (MoO2)2+/HZSM-5 have been investigated based on DFT method by using cluster models. NH3 adsorption is found to be more favorable energetically than NO adsorption on both Lewis and Brønsted acid sites. Our results confirm that the SCR reaction on (MoO2)2+/HZSM-5 catalyst begins with NH3 adsorption on acid site, which is consistent with the researches on other catalysts. Furthermore, the adsorption energy results also reveal that Lewis and

Acknowledgements

The authors gratefully acknowledge the financial support from the key program of National Natural Science Foundation of China (No. 21336006) and National Natural Science Foundation of China (No. 21073131).

References (71)

  • M. Mhamdi et al.

    Catal. Today

    (2009)
  • L. Casagrande et al.

    Appl. Catal. B: Environ.

    (1999)
  • L. Lietti et al.

    J. Catal.

    (1999)
  • J.P. Dunn et al.

    Appl. Catal. B: Environ.

    (1998)
  • P. Balle et al.

    Appl. Catal. B: Environ.

    (2009)
  • X.P. Wang et al.

    Appl. Catal. B: Environ.

    (2007)
  • A.L.S.M. Salgado et al.

    Catal. Today

    (2003)
  • Z. Li et al.

    Molybdenum loaded on HZSM-5: a catalyst for selective catalytic reduction of nitrogen oxides

  • K. Bourikas et al.

    Appl. Catal. B: Environ.

    (2004)
  • C. Fountzoula et al.

    Appl. Catal. B: Environ.

    (2002)
  • W. Yu et al.

    J. Colloid Interf. Sci.

    (2011)
  • W. Ding et al.

    J. Catal.

    (2002)
  • W. Li et al.

    J. Catal.

    (2000)
  • Y.-H. Kim et al.

    Micropor. Mesopor. Mater.

    (2000)
  • B. Li et al.

    Micropor. Mesopor. Mater.

    (2006)
  • H. Minming et al.

    J. Catal.

    (1987)
  • D. Ma et al.

    J. Catal.

    (2000)
  • L.A. Pine et al.

    J. Catal.

    (1984)
  • D. Zhou et al.

    Chem. Phys. Lett.

    (2003)
  • J.-P. Tessonnier et al.

    Appl. Catal. A: Gen.

    (2008)
  • G. Busca et al.

    Catal. Today

    (2005)
  • H. van Koningsveld et al.

    Zeolites

    (1990)
  • I. Mayer

    Chem. Phys. Lett.

    (1983)
  • T. Tsumuraya et al.

    J. Alloy. Compd.

    (2007)
  • F. Cao et al.

    Appl. Surf. Sci.

    (2012)
  • S. Soyer et al.

    Catal. Today

    (2006)
  • H. Yao et al.

    Surf. Sci.

    (2012)
  • C. Busco et al.

    Thermochim. Acta

    (2004)
  • H. Yao et al.

    J. Catal.

    (2013)
  • G. Busca et al.

    Appl. Catal. B: Environ.

    (1998)
  • M. Calatayud et al.

    Surf. Sci. Rep.

    (2004)
  • R.Q. Long et al.

    J. Catal.

    (2002)
  • D. Fang et al.

    Appl. Surf. Sci. B

    (2013)
  • J. Li et al.

    Catal. Today

    (2011)
  • Y. Peng et al.

    Chem. Commun.

    (2013)
  • Cited by (0)

    View full text