See discussions, stats, and author profiles for this publication at:
`https://www.researchgate.net/publication/222467795
`
`Catalytic Removal of NO
`
`Article in Catalysis Today · December 1998
`
`Impact Factor: 3.89 · DOI: 10.1016/S0920-5861(98)00399-X
`
`CITATIONS
`789
`
`READS
`185
`
`3 authors, including:
`
`Bernard Delmon
`Université catholique de Louvain
`
`253 PUBLICATIONS 7,528 CITATIONS
`
`SEE PROFILE
`
`Available from: Bernard Delmon
`Retrieved on: 30 April 2016
`
`BASF-2008.001
`
`

`
`Catalysis Today 46 (1998) 233–316
`
`Catalytic removal of NO
`
`V.I. Paˆrvulescua,*, P. Grangeb, B. Delmonb
`
`aDepartment of Chemical Technology and Catalysis, Faculty of Chemistry,
`University of Bucharest, B-dul Republicii 13, Bucharest 70346, Romania
`bUniversite´ Catholique de Louvain, Unite´ de catalyse et chimie des mate´riaux divise´s,
`Place Croix du Sud 2/17 1348 , Louvain-la-Neuve, Belgium
`
`Abstract
`
`The aim of this paper is to review the catalytic reactions for the removal of NO and, more particularly, to discuss the
`reduction of NO in the presence of NH3, CO, H2 or hydrocarbons as well as the decomposition of NO. The nature of the
`different active species, their formation due to dispersion and their interaction with different supports as well as the
`corresponding correlations with catalytic performance are also discussed. Another goal of this review is to explain the
`mechanism and kinetics of these reactions on different surfaces as well as the catalyst stability. # 1998 Elsevier Science B.V.
`All rights reserved.
`
`Keywords: Removal of NO; Decomposition of NO; Mechanism and kinetics
`
`1.
`
`Introduction
`
`The reduction of nitrogen oxide emissions has
`become one of the greatest challenges in environment
`protection. This is why it is being intensely studied by
`numerous groups from academic as well as industrial
`research laboratories. The interest in the subject is
`reflected in the number of papers, including a large
`number of reviews and patents, published each year.
`There is hardly an issue of a journal related to catalysis
`that does not contain contributions dealing with the
`elimination of NO.
`The nature of the catalyst is another element giving
`weight to the topic. Virtually all known categories of
`catalysts have been tested: metal and metal-supported
`catalysts, monocrystals and mixed phases, oxides and
`mixed oxides (with either acidic or basic properties),
`
`*Corresponding author.
`
`zeolites and heteropolyacids, alloys and amorphous
`alloys, membrane and monolithic catalysts, etc.
`Due to the extraordinary diversity of catalysts, this
`topic is always discussed at various conferences on
`catalysis. Each year, at least two conferences are
`entirely devoted to environmental catalysis, not to
`mention an additional periodic conference on auto-
`motive pollution control.
`The use of such a large number of catalysts to
`eliminate NO is logically associated with different
`ways of reaction. It is possible to divide these ways
`into four categories or methods:
`
`1. the selective catalytic reduction of NO with
`ammonia,
`typical of chemical
`industrial plants
`and stationary power stations;
`2. the catalytic reduction of NO in the presence of CO
`and/or hydrogen, typical of automotive pollution
`control;
`
`0920-5861/98/$ – see front matter # 1998 Elsevier Science B.V. All rights reserved.
`P I I : S 0 9 2 0 - 5 8 6 1 ( 9 8 ) 0 0 3 9 9 - X
`
`BASF-2008.002
`
`

`
`234
`
`V.I. Paˆrvulescu et al. / Catalysis Today 46 (1998) 233–316
`
`3. the selective catalytic reduction of NO in the
`presence of hydrocarbons and more particularly
`methane, a method which has not yet reached
`industrial use but can be applied both for auto-
`motive pollution control and in various industrial
`plants;
`4. the direct decomposition of NO, which is a goal
`worth striving for since it eliminates the need for
`reductants, which in turn eventually eliminates the
`additional pollution associated with the other three
`methods.
`
`All these methods have been extensively studied.
`The goal of this paper is to analyse the interaction of
`NO with different surfaces in the presence of various
`reductants or in the direct decomposition of NO. The
`paper intends to describe the effect of the different
`active species as well as the effect of their interaction
`and their dispersion with/on different supports on the
`catalytic performances of the catalysts. Explaining the
`mechanism and kinetics through which these reactions
`take place on different surfaces as well as the catalyst
`stability is another goal of this review.
`
`2. NO pollution
`
`NO is a major atmospheric pollutant. It has the
`ability to generate secondary contaminants through its
`interaction with other primary pollutants (like carbo-
`nyl corresponding molecules, alcohol radicals, etc.)
`also resulting from the combustion of fossil fuels in
`stationary sources such as industrial boilers, power
`plants, waste incinerators, and gasifiers, engines, and
`gas turbines or from the decomposition of a large
`number of organic products by light or micro-organ-
`isms.
`
`NO2 contributes substantially to so-called acid rains.
`Among the reactions involving ozone, the one with
`chlorofluorocarbons is very dangerous since it has a
`determining effect on the climate. The normal average
`content of ozone in the atmosphere is about
`10(cid:255)10 vol% and its interaction with NO also contri-
`butes to its diminution. The chemical depletion of
`ozone, in an important part due to nitrogen oxide
`species, is a prolonged phenomenon [2]. Carcinogenic
`products are also formed during these reactions.
`Hydrocarbons in polluted air do not react among
`themselves under the action of sun radiations, not
`even to a very small extent, but show a high reacti-
`vity towards intermediate species such as peroxides
`RO2 [1]. Such species react with the primary pollu-
`tants, NO, NO2, O3 and HC, according to a partially
`known mechanism. The photochemical complex
`HC–NOx–Ox is formed during the HC interactions
`in the photolytic cycle of NO; the mixture of products
`generated is called ‘‘photochemical smog’’ and con-
`tains O3, CO, peroxyacetyl nitrates, alkyl nitrates,
`ketones, etc.
`The photochemical cycle of nitrogen oxides initi-
`ates under sunlight (3000–4600 A˚ ). NO2 is initially
`decomposed as follows [3]:
`NO2 (cid:135) h(cid:23) (cid:133)> 3:12 eV(cid:134) !NO (cid:135) O
`(1.1)
`O (cid:135) O2 (cid:135) M ! O3 (cid:135) M (cid:133)third body(cid:134) (cid:135) 24:2kcal
`(1.2)
`
`O3 (cid:135) NO ! NO2 (cid:135) O2 (cid:135) 48:5 kcal
`until a dynamic equilibrium is reached:
`NO2 (cid:135) O2 !h(cid:23)
`NO (cid:135) O3
`
`(1.3)
`
`(1.4)
`
`The overall dependence of photochemical air pol-
`lution on various factors may be written as [2]:
`
`photochemical air pollution (cid:136) (cid:133)NOx (cid:255) conc(cid:134)(cid:133)organic conc(cid:134)(cid:133)sunlight intensity(cid:134)(cid:133)temperature(cid:134)
`(cid:133)wind spread(cid:134)(cid:133)inversion height(cid:134)
`
`:
`
`NO plays a major role in the photochemistry of the
`troposphere and the stratosphere. It reacts with photo-
`chemical pollutants such as ozone, formaldehyde,
`organic hydroperoxides and peroxyacyl nitrates that
`all are very reactive and have a very short lifetime.
`This is a very fast reaction which generates more
`nitrogen oxides and organic nitrates. The formed
`
`In the presence of oxygen, NO is oxidized very
`quickly to NO2 which, as mentioned before, is partly
`responsible for the acid rains and the urban smog.
`They have very negative effects in agriculture, but may
`also predispose to respiratory diseases by weakening
`the ability of the bronchopulmonary structures to
`function properly [4].
`
`BASF-2008.003
`
`

`
`V.I. Paˆrvulescu et al. / Catalysis Today 46 (1998) 233–316
`
`235
`
`Table 1
`Values of the equilibrium constant, Kp, and the dissociation degree,
`(cid:11), corresponding to the thermal decomposition of NO2 into NO and
`O2 and N2 and O2, respectively
`NO2!NO(cid:135)1/2 O2
`(cid:11)
`Kp
`1.69(cid:2)10(cid:255)3
`7.23(cid:2)10(cid:255)7
`2.53(cid:2)10(cid:255)4
`3.15(cid:2)10(cid:255)2
`8.00(cid:2)10(cid:255)3
`1.70(cid:2)10(cid:255)1
`8.09(cid:2)10(cid:255)2
`4.87(cid:2)10(cid:255)1
`4.22(cid:2)10(cid:255)1
`8.77(cid:2)10(cid:255)1
`9.55(cid:2)10(cid:255)1
`1.47
`
`T (K)
`
`298
`400
`500
`600
`700
`800
`
`NO2!1/2 N2(cid:135)O2
`(cid:11)
`Kp
`9.14(cid:2)108
`6.3(cid:2)10(cid:255)8
`3.32(cid:2)107
`1.7(cid:2)10(cid:255)8
`4.65(cid:2)106
`1.2(cid:2)10(cid:255)7
`1.26(cid:2)106
`4.6(cid:2)10(cid:255)7
`4.90(cid:2)105
`1.2(cid:2)10(cid:255)6
`2.50(cid:2)105
`2.3(cid:2)10(cid:255)6
`
`Table 2
`Evolution of the thermodynamic parameters: entropy (S0), enthalpy
`
`
`((cid:1)H0f ), Gibbs free energy of NO ((cid:1)G0f ), and Gibbs free energy of
`the NO decomposition
`
`T (K)
`
`NO
`
`S0
`(cal mol K)
`
`(cid:1)H0
`f
`(kcal mol(cid:255)1)
`
`(cid:1)G0
`f
`(kcal mol(cid:255)1)
`
`298
`300
`400
`500
`600
`700
`800
`900
`1000
`
`50.35
`50.40
`52.45
`54.06
`55.41
`56.57
`57.61
`58.54
`59.39
`
`21.60
`21.60
`21.61
`21.62
`21.62
`21.62
`21.63
`21.63
`21.64
`
`20.72
`20.71
`20.41
`20.11
`19.81
`19.51
`19.21
`18.91
`18.60
`
`NO reaction
`(cid:133)(cid:1)G0
`f (cid:134)r
`(kcal mol(cid:255)1)
`(cid:255)2(cid:2)20.72
`(cid:255)2(cid:2)20.71
`(cid:255)2(cid:2)20.41
`(cid:255)2(cid:2)20.11
`(cid:255)2(cid:2)19.81
`(cid:255)2(cid:2)19.51
`(cid:255)2(cid:2)19.21
`(cid:255)2(cid:2)18.91
`(cid:255)2(cid:2)18.60
`
`reactions such as NO oxidation, its decomposition to
`N2O and dimerization are. The evolution of the Gibbs
`free energy with temperature for the decomposition of
`NO to N2O is presented in Table 3.
`In the presence of a reductant, reactions lead to a
`strong decrease in the Gibbs free energy values. This
`explains that such reactions are used in practical
`applications. The Gibbs free energy values in the
`presence of the usual reductants in the catalytic reduc-
`tion of NO are given in Table 4. The introduction of
`oxygen here leads to an additional reduction in the free
`energy. This point will not be discussed.
`
`3. Selective catalytic reduction of nitric oxide
`with ammonia
`
`General aspects. The catalytic reduction of nitrogen
`oxides in effluent residual gases from various indus-
`tries, mainly nitric acid plants, can be carried out
`selectively using ammonia or urea. This is the so-
`called selective catalytic reduction (SCR) process.
`The reactions occur in a narrow temperature range;
`the main step is the reduction of NO or NO2 to N2.
`Generally, liquid ammonia is injected in the residual
`
`2.1. Thermodynamic stability of nitrogen oxide
`
`The interconversion of nitrogen oxides is relatively
`easy. The values of Kp for p(cid:136)1 atm corresponding to
`the thermal decomposition of NO2 to NO and NO2 to
`N2 and O2, respectively, are presented in Table 1.
`According to these data, NO2 is practically 100%
`decomposed in NO and O2 above 800 K. It appears
`that there is no direct thermal decomposition of NO2 to
`N2 and O2 because of the difference in Kp values.
`Even though NO is an endothermic compound, no
`decomposition is observed at 825 K. The thermody-
`namic parameters (S0, (cid:1)H0
`
`f and (cid:1)G0f ) corresponding
`to NO and the Gibbs free energy of the reaction are
`given in Table 2:
`2NO ! N2 (cid:135) O2
`for a relatively broad range of temperatures [5].
`From a thermodynamic point of view, as can been
`seen from Table 2, the NO molecule is unstable even
`at 298 K and 1 atm. This is due to the electronic
`structure of the bond in NO. The reaction is spin-
`forbidden, and NO is kinetically stable. Thus, the high
`thermal stability of nitric oxide is due to its high
`energy of dissociation (153.3 kcal mol(cid:255)1) and to cor-
`responding extremely low decomposition rates.
`The reaction is not sensitive to pressure variations
`(the variation of the stoichiometric number is 0) but
`
`(1.5)
`
`Table 3
`Evolution of the Gibbs free energy for the decomposition of NO to N2O
`
`T (K)
`(cid:1)G0
`f
`
`298
`33.34
`
`300
`33.24
`
`400
`28.46
`
`500
`23.66
`
`600
`18.88
`
`700
`14.10
`
`800
`(cid:255)9.36
`
`900
`(cid:255)4.62
`
`1000
`(cid:135)0.12
`
`BASF-2008.004
`
`

`
`236
`
`V.I. Paˆrvulescu et al. / Catalysis Today 46 (1998) 233–316
`
`Table 4
`Values of the Gibbs free energy for the reduction of NO in the presence of various reductants
`f;r (kcal mol(cid:255)1)
`(cid:1)G0
`(cid:135)CO
`(cid:135)H2
`(cid:255)150.92
`(cid:255)164.34
`(cid:255)150.66
`(cid:255)164.24
`(cid:255)147.86
`(cid:255)159.46
`(cid:255)144.94
`(cid:255)154.62
`(cid:255)141.94
`(cid:255)149.80
`(cid:255)138.86
`(cid:255)144.96
`(cid:255)135.72
`(cid:255)140.14
`(cid:255)132.54
`(cid:255)135.34
`(cid:255)129.28
`(cid:255)130.52
`
`(cid:135)C2H4
`(cid:255)146.21
`(cid:255)146.11
`(cid:255)145.28
`(cid:255)144.28
`(cid:255)143.67
`(cid:255)142.86
`(cid:255)142.04
`(cid:255)141.23
`(cid:255)140.38
`
`(cid:135)C2H6
`(cid:255)144.38
`(cid:255)144.21
`(cid:255)141.58
`(cid:255)139.37
`(cid:255)139.14
`(cid:255)138.81
`(cid:255)138.69
`(cid:255)138.46
`(cid:255)138.19
`
`(cid:135)C3H6
`(cid:255)144.04
`(cid:255)144.01
`(cid:255)143.52
`(cid:255)143.04
`(cid:255)142.57
`(cid:255)142.10
`(cid:255)141.61
`(cid:255)141.13
`(cid:255)140.62
`
`(cid:135)C3H8
`(cid:255)140.59
`(cid:255)140.57
`(cid:255)140.47
`(cid:255)140.39
`(cid:255)140.32
`(cid:255)140.24
`(cid:255)140.15
`(cid:255)140.07
`(cid:255)139.94
`
`(cid:135)C4H10
`(cid:255)136.83
`(cid:255)136.79
`(cid:255)135.11
`(cid:255)133.43
`(cid:255)131.75
`(cid:255)130.06
`(cid:255)128.42
`(cid:255)126.76
`(cid:255)125.11
`
`T (K)
`
`298
`300
`400
`500
`600
`700
`800
`900
`1000
`
`(cid:135)NH3
`(cid:255)87.36
`(cid:255)87.34
`(cid:255)87.62
`(cid:255)87.92
`(cid:255)88.24
`(cid:255)88.56
`(cid:255)88.86
`(cid:255)89.17
`(cid:255)89.45
`
`(cid:135)CH4
`(cid:255)134.26
`(cid:255)134.22
`(cid:255)132.19
`(cid:255)130.13
`(cid:255)128.06
`(cid:255)125.99
`(cid:255)123.93
`(cid:255)121.89
`(cid:255)119.85
`
`gas before the catalytic reaction takes place, using a
`sophisticated system of distribution.
`Bosch and Janssen [6] (and references herein)
`reviewed the catalysts used in the selective reduction
`of NO with NH3 and showed that the most investigated
`and effective systems are oxides prepared by thermal
`decomposition of an appropriate precursor or by
`impregnation of the supports. Other extensively inves-
`tigated catalysts are supported noble metals and metal
`zeolites. Recently, Janssen and Meijer [7] showed that
`more than a thousand catalyst compositions have been
`tested for application in this reaction up to now. This is
`the most used technique to control the emission of
`NOx.
`The main reactions that occur during SCR with
`ammonia are:
`4NO (cid:135) 4NH3 (cid:135) O2 ! 4N2 (cid:135) 6H2O
`6NO (cid:135) 4NH3 ! 5N2 (cid:135) 6H2O
`6NO2 (cid:135) 8NH3 ! 7N2 (cid:135) 12H2O
`2NO2 (cid:135) 4NH3 (cid:135) O2 ! 3N2 (cid:135) 6H2O
`Unwanted secondary reactions can take place as a
`result of the catalyst nature, the oxygen content, the
`temperature or the presence of acid gases:
`4NH3 (cid:135) 3O2 ! 2N2 (cid:135) 6H2O
`4NH3 (cid:135) 5O2 ! 4NO (cid:135) 6H2O
`4NH3 (cid:135) 7O2 ! 4NO2 (cid:135) 6H2O
`4NH3 (cid:135) 2O2 ! N2O (cid:135) 3H2O
`2NH3 (cid:135) 8NO ! 5N2O (cid:135) 3H2O
`6NH3 (cid:135) 8NO2 (cid:135) 3O2 ! 4N2O (cid:135) 6H2O
`
`(2.5)
`
`(2.6)
`
`(2.7)
`
`(2.8)
`
`(2.9)
`
`(2.10)
`
`(2.1)
`
`(2.2)
`
`(2.3)
`
`(2.4)
`
`(2.11)
`
`(2.12)
`
`4NH3 (cid:135) 4NO (cid:135) 3O2 ! 4N2O (cid:135) 6H2O
`16NH3 (cid:135) 12NO2 (cid:135) 7O2 ! 4N2O (cid:135) 24H2O
`The presence of SO2, CO2 or HCl determines a
`consumption of NH3 according to:
`2SO2 (cid:135) O2 ! 2SO3
`NH3 (cid:135) SO3 (cid:135) H2O ! NH4HSO4
`2NH3 (cid:135) SO3 (cid:135) H2O ! (cid:133)NH4(cid:134)2SO4
`2NH4HSO4 ! (cid:133)NH4(cid:134)2SO4 (cid:135) H2SO4
`NH4HSO4 (cid:135) NH3 ! (cid:133)NH4(cid:134)2SO4
`NH3 (cid:135) HCl ! NH4Cl
`2NH3 (cid:135) CO2 (cid:135) O2 ! (cid:133)NH4(cid:134)2CO3
`
`(2.13)
`
`(2.14)
`
`(2.15)
`
`(2.16)
`
`(2.17)
`
`(2.18)
`
`(2.19)
`
`The resulting products have corrosive properties
`and can therefore destroy the equipment.
`In the absence of catalysts, the reduction of NOx
`occurs with satisfactory conversion rates at tempera-
`tures in the range 1075–1175 K while the oxidation of
`NH3 to NOx (reactions (2.6)–(2.7)) takes place at
`temperatures above 1200 K (1325–1475 K) and the
`reduction rate decreases abruptly. The reaction rate is
`very low at temperatures below 1075–1175 K. In the
`presence of catalysts, the temperature of the reaction
`depends on the catalyst nature and can be in the 355–
`425 K range for the more active systems.
`In the process, it is vital to ensure a NH3 to NOx
`ratio close to that required by stoichiometry (accord-
`ing to reaction (2.1)) [8]. An oxygen excess reduces
`the catalyst selectivity to N2 and favours the formation
`of N2O (reactions (2.9)–(2.12)) or even the oxidation
`of ammonia to NO or NO2. However, no matter what
`
`BASF-2008.005
`
`

`
`V.I. Paˆrvulescu et al. / Catalysis Today 46 (1998) 233–316
`
`237
`
`the catalyst nature, the presence of oxygen is a very
`important factor and several authors stressed the fact
`that an increase in O2 pressure enhances the reaction
`rate [6]. Thermodynamical analyses indicate that sec-
`ondary products cannot be totally excluded. The
`degree of NOx reduction as well as the amount of
`residual NH3 in the effluent gases after the SCR could
`be controlled through an optimal NH3-to-NO ratio.
`This also depends on the volume of the catalyst.
`Generally, the NH3-to-NO ratio is about 40 mg NH3
`for 1 Nm3 NOx and corresponds to 1 vpm residual
`NH3. Good results for the reduction of NO with
`ammonia could be obtained with higher residence
`times (i.e. a high volume of catalyst), a condition
`under which the residual NH3 content is low, or with
`smaller residence times (i.e. a small volume of cata-
`lyst), but with a high residual NH3 content (that can
`even reach 8 vol%). A high volume of catalyst implies
`higher pressure drops and therefore a high energy
`consumption. The use of monolith-integrated active
`components is often reported to avoid pressure falls
`for this reaction [9].
`The SCR of NO with NH3 can also be performed in
`multilayer reactors, using two or three layers of cat-
`alyst. In this case a significant reduction of NO occurs
`on the first layer and is favoured by a high NH3
`content. The residual NH3 is then reduced on the
`second or third layer of the catalyst, where the NOx
`reduction is negligible [1]. However, an optimal pro-
`file of the reactant distribution is very difficult to
`achieve even when using the NH3 pulse technique,
`because the residual NH3 increases in the zones in
`which the NOx content is low. To have a simultaneous
`increase in NOx reduction and a low NH3 content in
`the effluent gases, some engineering systems like
`injecting NH3 between the catalysts’ layers, the use
`of a dynamic regime in compartment vibration
`chambers or the use of a periodic inversion of the
`direction of feed flow have also been considered
`[10,11–13]. The last method can achieve conversions
`above 99%.
`A new approach is the use of non-permselective
`porous membranes [14]. It allows for a high conver-
`sion in the SCR of NOx with ammonia, under fluctu-
`ating NOx flow, and ensures a very low NH3 content in
`the effluent gases. NH3 and NO effluent gases are
`separated by a porous wall which contains the cata-
`lytically active materials. The diffusion occurs as a
`
`result of the concentration gradient between the two
`gases.
`In the remainder of this section we will examine the
`main categories of catalysts used in this reaction,
`considering the influence of different additives as well
`as the nature of the support.
`
`3.1. Base oxide catalysts
`
`Bosch and Janssen [6] inventoried the base oxide
`catalysts active for the selective reduction of NOx with
`ammonia: V2O5, Fe2O3, CuO, Cr2O3, Co3O4, NiO,
`CeO2, La2O3, Pr6O11, Nd2O3, Gd2O3, Yb2O3. Cata-
`lytic tests performed with these oxides indicated
`vanadia oxide to be the most active and selective
`catalyst.
`The deposition of vanadia as well as other oxides on
`supports leads to an increase in the catalytic activity.
`The nature of the support is also a very important
`factor [15]. Bauerle et al. [16] reported a very good
`activity for V2O5 supported on TiO2 and Al2O3.
`Shikada et al. [17,18] were among the first to inves-
`tigate more systematically the effect of the support,
`evidencing the promotional effect of TiO2. They
`reported that the order of activity for supported vana-
`dia is TiO2–SiO2>g-Al2O3>SiO2. At the same time,
`Pearson et al. [19] revealed that the anatase form is
`more active than the rutile-based catalyst.
`A first explanation for the superior activity of
`titania-supported vanadia was given by Murakami et
`al. [20] who studied different supports: TiO2, ZrO2,
`SiO2 and MgO, and suggested that this activity could
`be associated with a crystallographic concordance
`between the structures of the two components. In
`addition, Wachs et al. [21] conducted studies using
`18O. They indicated that the stability of the terminal
`V=O during the SCR reaction suggests that the brid-
`ging V–O support is involved in the rate-determining
`step. Other studies by Bond and Bruckman [22] and
`Wachs et al. [23] revealed that the optimal activity was
`achieved when approximately one monolayer of vana-
`dia was dispersed on the anatase surface. Much atten-
`tion has been given to vanadia–titania catalysts in
`recent years, because of their effectiveness in selec-
`tively reducing NO in the presence of ammonia. The
`conversion of NO and NH3 on vanadia–titania cata-
`lysts with different vanadia contents is presented in
`Fig. 1.
`
`BASF-2008.006
`
`

`
`238
`
`V.I. Paˆrvulescu et al. / Catalysis Today 46 (1998) 233–316
`
`Fig. 1. NO (a) and NH3 (b) conversion on vanadia–titania catalysts with different content of vanadia (50 mg catalyst; NH3-to-NO ratio(cid:136)1.38;
`O2(cid:136)3500 ppm).
`
`Different techniques were used to characterize the
`structure of highly dispersed VOx species on the
`anatase surface. Based on IR spectroscopy measure-
`ments, Busca et al. [24,25] showed that mainly mono-
`low V2O5
`meric vanadyl species are formed at
`concentration on the anatase surface. The presence
`of vanadyl species was confirmed by Bell et al. [26]
`and Wachs et al. [27], using Raman spectroscopy. If
`the vanadia loading increases, the monomeric species
`predominant at low vanadia loading react to form
`polymeric vanadates [28]. When the amount of
`V2O5 exceeds that corresponding to a full coverage
`of the TiO2 support, V2O5 crystallites are formed [29].
`
`After reduction with hydrogen, Raman spectroscopy
`reveals that part of the oxygen atoms from terminal
`V=O groups associated with monomeric and poly-
`meric species are preferentially removed to make V–
`O–V bridging oxygens. Structural forms of vanadia-
`supported titania catalysts are presented in Scheme 1.
`51V NMR studies performed by Eckert and Wachs
`[30] offered additional information. They indicated
`the presence of monomeric species containing a four-
`fold coordinated vanadium with a symmetry greater
`than twofold. An increase of the vanadia loading leads
`to an increase in the vanadium coordination from
`fourfold to sixfold.
`
`BASF-2008.007
`
`

`
`V.I. Paˆrvulescu et al. / Catalysis Today 46 (1998) 233–316
`
`239
`
`Scheme 1. Structural forms of vanadia-supported titania catalysts.
`
`To improve the performance of this catalyst, dif-
`ferent methods of preparation of TiO2 were investi-
`gated. Generally,
`these methods correspond to
`different techniques for synthesizing the titania sup-
`ports or to the deposition of precursors of a different
`chemical nature. Concerning the synthesis of the
`titania supports,
`the most
`investigated techniques
`involve the hydrolysis of TiCl4 or titanium alkoxides
`or the precipitation of Ti(SO4)2 [6,31–35]. The sol–gel
`method was also investigated [36,37–39]. Recently,
`Ciambelli et al. [40] suggested a CO2 laser pyrolysis of
`titanium alkoxides, indicating that monocrystalline
`TiO2 powders with a uniform size could be obtained
`that way.
`Different methods were also used for the deposition
`of vanadia. At the beginning, the usual method was the
`impregnation of titania using ammonium metavana-
`date in the presence of oxalic acid acting as a compe-
`titive, or regulative agent [6] or directly with vanadium
`oxalate, as Grange et al. [34] proposed. Watanabe et al.
`[41] suggested a sequential precipitation. Later, a
`selective grafting from non-aqueous solutions was
`proposed by Delmon et al.
`[35], Baiker et al.
`[42,43] and Bond and Tahir [44]. In the past, Baiker
`et al. [37,38,45] had stressed the performance of
`catalysts prepared using the sol–gel-solution method
`and subsequently removal of the solvent by super-
`critical drying.
`SIMS and Raman spectroscopy results [38] showed
`that vanadia–titania aerogels contain active vanadium
`clusters in a relatively high degree of dispersion that
`differentiate them from traditional catalysts in which
`VOx clusters coexist with V2O5 crystallites. The SCR
`activity of these catalysts is very high and comparable
`to that of multiple grafted vanadia species.
`It is now clear that polymeric species are 10 times
`more active than monomeric species [28,46]. Addi-
`tional arguments were given by Lietti and Forzatti
`[47], based on temperature-programmed desorption,
`temperature-programmed
`surface
`reaction
`of
`
`adsorbed ammonia with gas phase NO and tempera-
`ture-programmed reaction and steady-state reaction
`experiments in the presence and in the absence of
`oxygen on a series of V2O5/TiO2 catalysts. These
`authors reported a higher reactivity of polymeric
`metavanadate species compared to isolated vanadyls,
`as well as a faster reduction by NH3 and a faster
`reoxidation by gaseous oxygen of the polymeric
`metavanadate groups.
`Vanadia/alumina and vanadia/silica were pre-
`viously thought to have a lower catalytic activity than
`vanadia/titania. Shikada et al. [17,18] first observed
`the beneficial effects of adding titania to silica. Grange
`et al. [34,35,48] reported an improvement
`in the
`catalytic activity of vanadia supported on titania
`grafted on silica or alumina using selective grafting
`from non-aqueous solutions. Fig. 2 compares the NO
`conversion on mixed V2O5–TiO2–SiO2 (a) and V2O5–
`TiO2–Al2O3 (b) with different percentages of TiO2
`and V2O5.
`Lapina et al. [49–52] used the 51V NMR technique
`to characterize supported vanadium catalysts. In the
`case of vanadia supported on titania grafted on silica
`surfaces [52], they suggested that V2O5 interacted
`both with SiO2 and TiO2. The structure of the V
`complexes depended on the sequence of the V and
`Ti deposition. The formation of several surface tetra-
`hedral and octahedral complexes as well as two types
`of mixed triple V–Ti–Si complexes on the SiO2 sur-
`face was mentioned.
`Another way to promote the activity of vanadia–
`titania catalysts is by adding an other oxide species to
`vanadium oxide. The positive effect of WO3 was
`reported by many groups [53–56]. The same effect
`was also evidenced for vanadia supported on alumina
`[57] or zirconia-promoted alumina supports [58].
`Good results were also reported by Ross et al. [59]
`for zirconia-supported vanadia catalysts. It was shown
`that WO3 increases the surface acidity by generating
`both Brønsted and Lewis sites. This was accompanied
`by an increased resistance towards deactivation by
`SO2 or basic compounds like alkali metal or arsenious
`oxide. The catalytic activity concomitantly increased.
`Other oxides like MoO3, CeO2, SnO2 or ZrO2 have
`a similar effect on V2O5/TiO2 catalysts [60–63]. The
`corresponding catalysts are called ‘‘low temperature
`conversion catalysts’’, as the reaction can occur even
`at 393 K. However, this results in a relatively high
`
`BASF-2008.008
`
`

`
`240
`
`V.I. Paˆrvulescu et al. / Catalysis Today 46 (1998) 233–316
`
`Fig. 2. NO conversion with NH3 over V2O5–TiO2–SiO2 (a); V2O5–TiO2–Al2O3 (b); V2O5–TiO2–Nb2O3 (c) catalysts. (a) V2O5–TiO2–SiO2:
`NO(cid:136)970 ppm; NH3(cid:136)12.40 ppm; O2(cid:136)3 vol%; 100 mg catalyst;
`flow rate(cid:136)50 ml min(cid:255)1;
`(b) V2O5–TiO2–Al2O3; NO(cid:136)0.1 vol%;
`flow rate(cid:136)50 ml min(cid:255)1;
`NH3(cid:136)0.106 vol%; O2(cid:136)3 vol%; 100 mg catalyst;
`(c) V2O5–TiO2–Nb2O5: NO(cid:136)0.4 vol%; NH3(cid:136)0.42 vol%;
`O2(cid:136)2 vol%; space velocity(cid:136)58 000 h(cid:255)1.
`
`BASF-2008.009
`
`

`
`V.I. Paˆrvulescu et al. / Catalysis Today 46 (1998) 233–316
`
`241
`
`NH3 content in the effluent gases. Nb2O5 was reported
`to be another promising promoter for V2O5/TiO2
`catalysts [64,65] (Fig. 2(c)). Vikulov et al. [65] sug-
`gested that the promoting effect of niobia might be
`associated with the stabilization of the surface area of
`the catalyst. Another opinion is that of Wachs et al.
`[66] who showed that, in the frame of the general
`agreement that a dual site mechanism operates, the
`role of the promoter is to create a non-reducible oxide
`site adjacent to the surface vanadia site. Such a dual
`site created by tungsten, niobium or even sulphur
`generates a 5–10 fold increase in the SCR TOF.
`FT-IR and laser Raman studies [67] indicate that
`the promoter generates species similar to those gen-
`erated by vanadia,
`i.e. monomeric and polymeric
`wolframyls and molybdyls, onto the titania surface.
`The effect of mixed deposition was also reported by
`Kasaoka et al. [68]. They showed that vanadium oxide
`supported on either active carbon or active carbon
`covered with titania is also a very active catalyst.
`The presence of vitreous P2O5 was reported as
`another factor that could enhance the catalytic activity
`of vanadia/titania or V2O5–MO3/TiO2 catalysts [69].
`Some VOPO4 species are also formed in such cases.
`Very recently, Wachs et al. [70] reported that titania-
`supported rhenium catalysts exhibit the same SCR
`activity as titania-supported vanadium catalysts but
`with a lower selectivity to N2. They showed that the
`reaction on these catalysts requires the same dual site
`mechanism as vanadia.
`Supported chromia catalyst systems are also much
`investigated in the SCR of NO. Niyama et al. [71,72]
`studied the behaviour of chromia–alumina catalysts
`with different compositions and reported good activ-
`ities for lower chromia concentrations and for pure
`Cr2O3 oxide. The conversions were insignificant for
`intermediary amounts of deposited chromium. How-
`ever, the conversions obtained with the active formu-
`lations were inferior to those obtained when using
`vanadia, and important quantities of N2O were formed
`during the reaction. Later, Baiker et al. [73,74–82]
`showed that amorphous chromia exhibited interesting
`properties for low temperature SCR and selectively
`reduced NO to N2. They showed that the selectivity
`and initial activity of titania-supported chromia cata-
`lysts depended markedly on the pretreatment. In a
`series of
`reductive pretreatments at
`temperatures
`below 720 K an improved selectivity to N2 was
`
`reported at reaction temperatures up to 470 K. N2 is
`therefore the major product below 470 K. The
`decrease in activity at higher temperatures was attrib-
`uted to a partial crystallization of chromia, while the
`improved selectivity was attributed to the reduction of
`Cr(VI) and Cr(V) surface species. In fact, hydroxy-
`lated Cr(III) species formed during the reductive
`pretreatment constitute Brønsted sites, active for
`ammonia chemisorption. However, SCR tests indicate
`that a totally reduced surface is inactive and that the
`surface must be maintained in a partly oxidized state
`by the presence of oxygen in the feed in order to
`balance the partial reduction by NH3. In the absence of
`oxygen the reaction between NO and NH3 is much
`slower, but at higher temperatures (T>500 K) and in
`the production of N2O
`the presence of oxygen,
`becomes significant [83]. Some oxidation of the
`ammonia to N2 occurs in parallel to the SCR of
`NO. The variation of TOF with the chromia content
`on Cr2O3–TiO2 is presented in Fig. 3.
`Other oxide-supported catalysts were also studied.
`Experiments carried out by Grange et al. [84,85]
`indicated that molybdena–titania catalysts obtained
`by grafting from non-aqueous solutions were selective
`for the reduction of NO to N2 with ammonia.
`Fe2O3–Al2O3 catalysts exhibit a behaviour similar
`to that of chromia–alumina catalysts [71,72,86].
`
`Fig. 3. Variation of TOF with the chromia content on Cr2O3–TiO2
`catalysts (NH3-to-NO ratio(cid:136)1, O2(cid:136)1.8 vol%, 100 mg catalyst,
`T(cid:136)450 K).
`
`BASF-2008.010
`
`

`
`242
`
`V.I. Paˆrvulescu et al. / Catalysis Today 46 (1998) 233–316
`
`Bosch and Janssen [6] made an inventory of a large
`number of catalysts, Fe2O3/TiO2, Fe2O3–WO3/TiO2,
`Fe2O3/TiO2–ZrO2 (or SiO2), and indicated that these
`catalysts were active in the 470–770 K range. Gryzbek
`and Papp [87,88] reported that carbon is also an
`effective support for iron oxides. In addition, these
`authors showed that low iron contents supported on
`active carbon are very active in the reduction of NO
`with ammonia at low temperatures. In that case, a
`good distribution of Fe3(cid:135) is only achieved after a
`controlled oxidation of the support surface.
`Copper compounds are also effective catalysts in
`SCR with ammonia. Bosch and Janssen [6] reported
`the activity of copper oxide supported on titania,
`alumina and substituted alumina. However, many
`studies used CuSO4 because copper oxide is trans-
`formed in copper sulphate in the presenc