Influence of NO2 on the selective catalytic reduction of NO with ammonia over Fe-ZSM5

https://doi.org/10.1016/j.apcatb.2006.04.015Get rights and content

Abstract

The influence of NO2 on the selective catalytic reduction (SCR) of NO with ammonia was studied over Fe-ZSM5 coated on cordierite monolith. NO2 in the feed drastically enhanced the NOx removal efficiency (DeNOx) up to 600 °C, whereas the promoting effect was most pronounced at the low temperature end. The maximum activity was found for NO2/NOx = 50%, which is explained by the stoichiometry of the actual SCR reaction over Fe-ZSM5, requiring a NH3:NO:NO2 ratio of 2:1:1. In this context, it is a special feature of Fe-ZSM5 to keep this activity level almost up to NO2/NOx = 100%. The addition of NO2 to the feed gas was always accompanied by the production of N2O at lower and intermediate temperatures. The absence of N2O at the high temperature end is explained by the N2O decomposition and N2O-SCR reaction. Water and oxygen influence the SCR reaction indirectly. Oxygen enhances the oxidation of NO to NO2 and water suppresses the oxidation of NO to NO2, which is an essential preceding step of the actual SCR reaction for NO2/NOx < 50%. DRIFT spectra of the catalyst under different pre-treatment and operating conditions suggest a common intermediate, from which the main product N2 is formed with NO and the side-product N2O by reaction with gas phase NO2.

Introduction

During the last years, Fe-ZSM5 has been discovered to be an efficient catalyst for the selective catalytic reduction of nitrogen oxides in the exhaust of diesel vehicles by urea (urea-SCR) besides the established V2O5–WO3/TiO2 catalysts [1], [2], [3]. For both types of catalysts the reaction obeys the “standard-SCR” stoichiometry, involving equal amounts of NO and ammonia—the latter being the actual reducing agent, which is released from urea under the hydrothermal conditions in the exhaust pipe:4NH3 + 4NO + O2  4N2 + 6H2O

The main challenge for the application of Fe-ZSM5 in the urea-SCR process on board of diesel vehicles is to achieve high SCR activity and selectivity over a broad temperature range (150–700 °C). One possibility is the addition of NO2 to the feed, whose promoting effect on the SCR reaction was already observed in the 1980s [4], [5]. Koebel et al., who investigated the influence of NO2 on the SCR reaction over V2O5–WiO3/TiO2 in detail, pointed out the importance of having a 1:1 mixture of NO2 and NO for reaching the maximum SCR activity [6], [7]. The high rates at this NO2 to NO ratio are explained by SCR reaction (2), which is much faster than reaction (1), therefore called “fast-SCR”:4NH3 + 2NO + 2NO2  4N2 + 6H2O

If the NO2:NO ratio is larger than 1:1 NO2 reacts also via the “NO2-SCR” route (3), which is even slower than the “standard-SCR” reaction over vanadia-based catalysts:4NH3 + 3NO2  3.5N2 + 6H2O

Very recently, the promoting effect of NO2 was also reported for zeolite-based SCR catalysts by Tolonen et al., who performed a screening of metal-exchanged zeolites [8]. They added NO2 and NO to the feed in the ratio 0.4:0.6, which is nearly ideal for the “fast-SCR” stoichiometry, and found an increase in NOx removal efficiency at lower temperatures compared to 100% NO in feed.

The addition of NO2 to the feed implies also some disadvantages. The formation of N2O was observed over Fe-ZSM5 in the presence of NO2 [8] and over Cu-ZSM5 N2O is formed even without NO2 in the feed. On the basis of first performance tests the formation of N2O over Fe-ZSM5 was generally attributed to ammonia oxidation [8]. Based on a more detailed investigation, Long and Yang tried to explain the elevated formation of N2O by the reaction of NO2 with the postulated intermediate species NO2(NH4+)2 [9]. Despite this plausible explanation, many more reactions are conceivable to produce N2O that have mainly been discussed for vanadia-based SCR catalysts [10], but cannot be ruled out to take part on Fe-ZSM5 and other metal-exchanged catalysts. In the following, these reactions are summarized as a basis on which the observations in the present work are made.

The most important source of N2O over vanadia-based SCR catalyst is the selective catalytic oxidation (SCO) of ammonia by oxygen (reaction (4)):4NH3 + 4O2  2N2O + 6H2O

At lower temperatures up to 350 °C, NO2 which is a much stronger oxidant than O2 is expected to contribute to the oxidation of ammonia to N2O also by the following reactions:6NH3 + 8NO2  7N2O + 9H2O4NH3 + 4NO2 + O2  4N2O + 6H2O

Due to the thermodynamic equilibrium:NO + 1/2O2  NO2considerable amounts of NO are formed at T > 350 °C and low space velocities, facilitating the formation of N2O by reaction (8), even if only NO2 is present in the feed gas:4NH3 + 4NO + 3O2  4N2O + 6H2O

In the very low temperature region below 200 °C the formation of N2O is explained by the decomposition of NH4NO3 as an intermediate complex:NH4NO3  N2O + 2H2O

NH4NO3 itself is formed from NH3 and N2O4, which is the stable dimerization product of NO2:N2O4 + 2NH3 + H2O  NH4NO3 + NH4NO2

The simultaneously formed ammonium nitrite does not play an important role, as it is readily decomposed to nitrogen and water at T > 60 °C (reaction (11)):NH4NO2  N2 + 2H2O

The aim of the present paper is to describe in detail the influence of NO2 on the SCR reaction over Fe-ZSM5 coated on a cordierite monolith, also comprising investigations on the formation of N2O at lower and intermediate temperatures and its depletion at higher temperatures due to the N2O decomposition and N2O-SCR reaction.

Section snippets

Experimental

Fe-ZSM5 coated on a cordierite monolith was provided by Umicore automotive catalysts, Germany. The catalytic tests were carried out in the laboratory test apparatus described in [10]. The flow rates were controlled by mass flow controllers (Brooks 5850S). Liquid water was dosed by means of a liquid mass flow controller (Brooks 5881) through a capillary tube into an electrically heated evaporator. All lines of the experimental apparatus were heated to 150 °C by heating tapes. The concentrations

NH3-SCR

Fig. 1a shows the NOx removal efficiency (DeNOx) of the Fe-ZSM5 monolith catalyst for varying NO:NO2 ratios in the feed at 10 ppm ammonia slip. Details about this way to represent the data are given in [11], [12]. DeNOx increased drastically with the addition of NO2 to the feed over the entire temperature range, most pronounced at lower and intermediate temperatures. Regarding the entire temperature range, NO2/NOx = 50% resulted in the best DeNOx performance. However, it has to be noticed that the

NH3-SCR

From the DeNOx measurements it is obvious that NO2 is involved in the mechanism of the SCR reaction. In the case of “standard-SCR” (reaction (1)), where no NO2 is dosed, the reaction is able to proceed due to the capability of Fe-ZSM5 to oxidize NO to NO2 [12]. When NO was dosed over Fe-ZSM5, first the NO2 fraction increased with temperature but decreased again above ∼350 °C due to the thermodynamic equilibrium between NO and NO2, which shifts to the NO side for higher temperatures [12]. The

Conclusions

By the present study it was shown that the SCR activity of Fe-ZSM5 is tremendously improved by the addition of NO2 to the feed gas. The maximum DeNOx was found for NO2/NOx = 50%, but, surprisingly, higher NO2 fractions resulted in similar high DeNOx values. This behaviour is explained by the assumption of [NH4+]2NO2 as intermediate complex species, which is formed fast from NH3, NO and NO2 in the optimum ratio 2:1:1 and reacts to the final product nitrogen. Higher NO2 fractions are also easily

References (30)

  • M. Koebel et al.

    Catal. Today

    (2000)
  • R.Q. Long et al.

    J. Catal.

    (1999)
  • M. Koebel et al.

    Catal. Today

    (2002)
  • H.Y. Huang et al.

    App. Catal. A

    (2002)
  • S.A. Stevenson et al.

    J. Catal.

    (2000)
  • T. Komatsu et al.

    J. Catal.

    (1994)
  • T. Turek

    Appl. Catal. B

    (1996)
  • B. Coq et al.

    Appl. Catal. B

    (2000)
  • J. Despres et al.

    Microporous Mesoporous Mater.

    (2003)
  • J. Eng et al.

    J. Catal.

    (1997)
  • A. Waclaw et al.

    Catal. Today

    (2004)
  • Y.H. Yeom et al.

    J. Catal.

    (2005)
  • R.Q. Long et al.

    J. Catal.

    (2001)
  • F. Kapteijn et al.

    Appl. Catal. B

    (1996)
  • F. Kapteijn et al.
  • Cited by (302)

    View all citing articles on Scopus
    View full text