4008
Ind. Eng. Chem. Res. 2002, 41, 4008-4015
Side Reactions in the Selective Catalytic Reduction of NOx with Various NO2 Fractions Giuseppe Madia, Manfred Koebel,* Martin Elsener, and Alexander Wokaun Paul Scherrer Institute, CH-5232 Villigen PSI, Switzerland
The main and side reactions of the three selective catalytic reduction (SCR) reactions with ammonia over TiO2-WO3-V2O5 catalysts have been investigated using synthetic gas mixtures matching the composition of diesel exhaust. The three SCR reactions are standard SCR with pure NO, fast SCR with an equimolar mixture of NO and NO2, and NO2 SCR with pure NO2. At high temperatures the selective catalytic oxidation (SCO) of ammonia and the formation of nitrous oxide compete with the SCR reactions. Water strongly inhibits the SCO of ammonia and the formation of nitrous oxide, thus increasing the selectivity of the SCR reactions. However, water also inhibits SCR activity, most pronounced at low temperatures. NO2 fractions exceeding 50% enhance the formation of nitrous oxide at low temperatures. If the feed of NOx consists of pure NO2, the formation of nitrous oxide may occur by two different reactions having different temperature regimes. The reaction responsible for N2O formation at low temperatures probably involves ammonium nitrate or nitroamine as an intermediate species. 1. Introduction The selective catalytic reduction (SCR) of nitrogen oxides by ammonia is a highly developed and widespread technology for the removal of NOx from stationary sources.1,2 Moreover, this technique using urea as a toxicologically and environmentally safer reducing agent is presently considered as the most promising technique to remove NOx from lean exhaust gases of mobile diesel engines.3 The reduction of nitric oxide with ammonia or other N-containing reducing agents by SCR is generally considered to be highly selective. The term “selective” refers primarily to the oxidizing educts of the reduction: NO shall be reduced in preference to O2, which is present in lean exhaust gases in much higher amounts. This means that the standard SCR reaction (1) should proceed at a much higher rate than reaction (2), usually designated as selective catalytic oxidation (SCO):
4NH3 + 4NO + O2 f 4N2 + 6H2O
(1)
4NH3 + 3O2 f 2N2 + 6H2O
(2)
in terms of electrons transferred to NO and O2, the ratio is 2:1, resulting in a selectivity of 2/3 for the reduction of NO and 1/3 for the reduction of O2. Nevertheless, SCR with N-containing reducing agents is about 2 orders of magnitude more selective than the so-called HC-SCR using hydrocarbons. Evaluating published results on HC-SCR in terms of NO selectivity yields typical values on the order of a percent. Much work on the SCR reaction and its selectivity has been done on “model” catalysts applying “pure” operating conditions (catalysts with high V content and no W, absence of water in the feed). Only a few investigations have been done with “real” catalysts and in the presence of water, e.g., by Lietti et al.5 The present investigation on the selectivity of SCR reactions is based on experimental results obtained with “real” catalysts under operating conditions matching those of diesel exhaust. The subject will be discussed not only for the “standard SCR” reaction (1) but also for the “fast SCR” reaction and NO2 SCR, i.e., for the cases where the gas contains NO2 in addition to NO. 2. Chemistry and Known Facts
However, the term “selective” also refers to the products. Elementary nitrogen (and water) is the desired product of the DeNOx process, but the formation of higher oxidized nitrogen species such as N2O, NO, and NO2 cannot be excluded. N2O is undesired because of its potential for depleting stratospheric ozone and its strong greenhouse warming effect.4 Additionally, the formation of N2O, NO, and NO2 leads to a deterioration of the DeNOx effect and to an unnecessary consumption of reducing agent. We should mention that, according to this definition, the selectivity of the SCR reaction (1) is not 100% for NO, because four NO molecules are reduced together with one oxygen molecule. When the balance is made * To whom correspondence should be addressed. Tel: +4156-310 26 04. Fax: +41-56-310 21 99. E-mail: manfred.koebel@ psi.ch.
2.1. Side Reactions of the Standard SCR Reaction (Only NO Present). The standard SCR reaction is the sole possible DeNOx reaction if the feed gas contains NOx only in the form of NO. The most evident byproduct of the standard SCR reaction (1) is nitrous oxide, N2O. On typical commercial catalysts based on TiO2-WO3-V2O5, its formation becomes perceptible at temperatures above ≈380 °C. Other possible byproducts such as NO, NO2, and N2 from the SCO reaction (2) are not detectable by the usual analytical techniques. On the other hand, a consumption of the reducing agent (NH3) exceeding the theoretical value according to (1) is often observed at temperatures above about 380 °C. Considering the formation of these byproducts at higher temperatures under standard SCR conditions, the following reactions of ammonia with oxygen are possible:
10.1021/ie020054c CCC: $22.00 © 2002 American Chemical Society Published on Web 07/03/2002
Ind. Eng. Chem. Res., Vol. 41, No. 16, 2002 4009 Table 1. Oxidation of Ammonia with Oxygen According to Reactions 2-5 reaction no. product (2) (3) (4) (5)
N2 N2O NO NO2
electrons transferred oxidation no. of N NH3 f product in the product 3 4 5 7
0 +1 +2 +4
4NH3 + 3O2 f 2N2 + 6H2O
(2)
4NH3 + 4O2 f 2N2O + 6H2O
(3)
4NH3 + 5O2 f 4NO + 6H2O
(4)
4NH3 + 7O2 f 4NO2 + 6H2O
(5)
4NH3 + 4NO + 3O2 f 4N2O + 6H2O
(6)
Considering first reactions (2)-(5), we can see that the degree of ammonia oxidation increases in a systematic way. This is evident from the increasing amount of oxygen consumed and from the increasing oxidation number of nitrogen in the product (Table 1). Il’Chenko and Golodets6 have investigated the basic phenomena of ammonia oxidation with oxygen over metal oxide catalysts. They have shown that the catalytic activity and the product selectivity depend on the bond energy of surface oxygen in a very general way. Oxides with low surface oxygen bond energy show high activity and favor the formation of products of deep oxidation (NO and NO2). The inverse is observed for oxides with high surface oxygen bond energy: low catalytic activity and the preferential formation of products of mild oxidation (N2, N2O). In addition, increasing temperature shifts the product distribution toward deep oxidation. V2O5 catalysts were reported to be very selective for N2 formation at low temperatures. On the other hand, noble metal catalysts such as Pt and Pd are very effective in promoting deep oxidation, i.e., formation of NO/NO2. Li and Armor7 investigated the SCO of ammonia on V2O5-TiO2 catalysts. Water was found to inhibit the activity and to increase the selectivity for the oxidation of ammonia to N2. Topsoe et al.8 investigated the influence of water on the reactivity of V2O5-TiO2 catalysts for the reduction of nitric oxide and observed that water inhibits the activity for SCR at temperatures lower than 663 K, while preventing the formation of N2O. Ramis et al.9 suggested that hydrazine (N2H4) is an intermediate species in the mechanism of the SCO of ammonia. Considering the formation of nitrous oxide under the conditions of standard SCR, i.e., for a gas mixture containing O2, NO, and NH3, the equations above suggest that N2O may be formed from oxidation of ammonia by O2 [reaction (3)] or from oxidation by NO + O2 [reaction (6)]. The key question is, which reaction is prevailing under typical SCR conditions? Isotopic labeling experiments over V2O5-TiO2 catalysts were performed by Okzan et al.10 and by Duffy et al.11 Both groups have shown that, under SCR conditions, one of the two N atoms originates from ammonia and the other from NO, suggesting that nitrous oxide is formed mainly according to reaction (6) over these catalysts. We will utilize this observation in the evaluation of our results below.
2.2. Side Reactions in the Presence of NO2 (Fast SCR Reaction). A major challenge in the further development of the SCR process for automotive applications is a high activity for the SCR process over a wide temperature range (180-650 °C). An improved SCR process relies on the fact that the SCR reaction is much faster with an equimolar mixture of NO and NO2 than with pure NO:12-14
4NH3 + 2NO + 2NO2 f 4N2 + 6H2O (fast SCR reaction) (7) The effect of NO2 is most pronounced at low temperatures, where an increase in the reaction rate is most essential. The required NO2 may be easily produced by introducing an oxidation catalyst (Pt-based) upstream of the SCR catalyst. However, if NOx downstream of the oxidation catalyst contains more than 50% NO2, excessive NO2 can only react according to the reaction equation of NO2 SCR:
4NH3 + 3NO2 f 3.5N2 + 6H2O
(8)
This reaction is even slower than the standard SCR reaction. This may favor the contribution of other side reactions and consequently change the order of selectivities. The formation of the byproduct N2O is expected to occur according to different mechanisms depending on the reaction temperature: At intermediate temperatures according to one of the following reactions:
6NH3 + 8NO2 f 7N2O + 9H2O
(9)
4NH3 + 4NO2 + O2 f 4N2O + 6H2O
(10)
At high temperatures (>350 °C) and low space velocities, the thermodynamic equilibrium
NO + 1/2O2 H NO2
(11)
will tend to form an increased fraction of NO, thus allowing also for N2O formation according to reaction (6). At Low Temperatures. In a recent paper Blanco et al.15 report an enhanced formation of nitrous oxide on a V2O5-TiO2 catalyst using a dry NOx feed containing NO2 in the temperature range 180-230 °C. In the present work we have also found an impressing formation of N2O at low temperatures, which we attribute to a low-temperature mechanism involving the following steps:
2NO2 H N2O4
(12)
N2O4 + 2NH3 + H2O f NH4NO3 + NH4NO2 (13) Reaction (12) is the well-known dimerization reaction of nitrogen dioxide showing a negative temperature coefficient. In the subsequent reaction, N4+ in nitrogen tetroxide disproportionates, thus allowing the formation of ammonium nitrate and ammonium nitrite. Ammonium nitrite is known to be a highly unstable compound, decomposing explosively at temperatures above 60 °C according to
NH4NO2 f N2 + 2H2O
(14)
4010
Ind. Eng. Chem. Res., Vol. 41, No. 16, 2002
Summing up reactions (12)-(14), we obtain the following overall equation for the formation of ammonium nitrate at low temperatures:
2NH3 + 2NO2 f NH4NO3 + N2 + H2O
(15)
The deposition of solid ammonium nitrate on SCR catalysts according to reaction (15) at temperatures below 200 °C has recently been demonstrated in this laboratory.13,14 We have also shown that the thermal decomposition of the formed ammonium nitrate may form either HNO3 + NH3 or N2O + 2H2O. Generally, slow heating rates and the presence of water will lead to the formation of HNO3 + NH3. Fast heating rates and the absence of water will favor the formation of N2O. On the basis of these observations, we propose the following overall reaction for the formation of N2O at low temperatures in feeds containing high fractions of NO2 (NO2/NOx > 0.5):
2NH3 + 2NO2 f N2O + N2 + 3H2O
3. Selectivities The common notion of selectivity refers to the formation of different products. Usually, a species is involved in two or more reactions, leading to the formation of several products, e.g.,
νAA f νBB
(17)
νAA f νCC
(18)
The selectivity for product B is then defined as
B produced mol B produced + C produced mol
[ ]
(19)
In the case of SCR, such a definition based on the product distribution is not very useful. In this case, the primary interest is the selective reduction of NOx in the presence of oxygen. This calls for an appropriate definition referring to the selective reduction of NOx. The definition should also take into account that the desired product (N2) can be formed not only by SCR reactions but also by the SCO of ammonia, thus leading to an increased consumption of the reducing agent. Therefore, high product selectivity to nitrogen may also be observed with a catalyst showing high SCO activity for ammonia [reaction (2)]. In the evaluation of the experiments, we will use the following definition of selectivity for the standard and fast SCR reaction:
SNOx )
NOx converted mol NH3 consumed mol
[ ]
SNO2 )
(20)
According to this definition, the selectivity is 100% if all ammonia consumed in the process has reacted with NOx in the standard SCR reaction (1) and the fast SCR reaction (7). Because of the fact that the stoichiometric
4 NO2 converted mol 3 NH3 consumed mol
[ ]
(21)
In the case that side reactions take place, additional ammonia will be consumed by these reactions. This will become evident if SNOx or SNO2 is lower than 1. However, because of the fact that N2 and NO/NO2 cannot be distinguished experimentally from N2 formed in DeNOx reactions and from NO/NO2 present in the original feed, the corresponding selectivities cannot be determined. The only detectable new product is N2O, and we will use the following selectivity definition in the discussion of the results:
SN 2 O )
(16)
Ammonium nitrate or its anhydride nitroamine H2NNO2 would play the role of the intermediate in this reaction similar to the nitrosamidic intermediate species in the SCR mechanism proposed by Ramis et al.16 The observations of the present work support such a mechanism, as will be shown below in the experimental and the discussion parts.
SB )
coefficients νNH3 and νNOx are not equal for the NO2 SCR reaction (8), the corresponding equation would be
νNH3 N2O formed mol νN2O NH3 consumed mol
[ ]
(22)
where νNH3 and νN2O are the stoichiometric coefficients of ammonia and nitrous oxide of the respective reaction equation. For the special case where the feed contains no nitrogen oxides but only ammonia and oxygen, the amounts of N2O, NO, and NO2 formed on the catalyst can be measured, and N2 may be obtained by the difference. Selectivities referring to the total amount of ammonia consumed may thus be obtained as
Si )
νNH3 product i formed mol νi NH3 consumed mol
[ ]
(23)
where νNH3 and νi are again the stoichiometric coefficients of ammonia and product i of the respective reaction equation. 4. Experimental Section 4.1. Catalyst Samples and Reactor. The preparation of the catalyst has been described elsewhere.17 The ternary catalyst contains ≈3wt % V2O5 and ≈8wt % WO3 on TiO2, similar to a commercial SCR catalyst, and the Brunauer-Emmett-Teller (BET) surface amounts to ≈70 m2/g. A volume of 7.3 cm3 of a metal monolith having a cell density of 600 cells/in.2 was coated with 1.4 g of catalyst. The tests were carried out in a tubular glass reactor including a preheating zone of about 15 cm length and was heated by two separate temperaturecontrolled heating coils. The temperature of all gas lines was kept at 150 °C. 4.2. Experimental Setup. The experimental setup is shown in Figure 1. The composition of the base feed was adapted to a typical exhaust gas, containing 10% O2 and 5% H2O with balance N2. Compared to a real diesel exhaust, carbon dioxide (≈7%) and 100 ppm CO were omitted but have no influence on the behavior of the SCR catalyst. In addition, a real exhaust contains 10-200 ppm of hydrocarbons whose main effect is the inhibition of the SCR reaction at low temperatures. SO2 was omitted because of its low levels already today (≈20 ppm) and forthcoming diesel fuel standards reducing fuel sulfur still further by a factor of ≈10. NO, NO2, and NH3 were in the range of 100-1000 ppm. The gas feed was mixed from higher concentrated gas mixtures obtained from Carbagas using mass flow controllers
Ind. Eng. Chem. Res., Vol. 41, No. 16, 2002 4011
Figure 3. Overall NH3 conversion and selectivity for the ammonia oxidation reactions (2)-(5) in the presence of water. Feed: 1000 ppm NH3, 5% H2O, 10% O2, balance N2. Figure 1. Experimental setup: 1, water reservoir; 2, liquid mass flow controller; 3, mass flow controller; 4, water evaporator; 5, reactor; 6, catalyst sample; 7, filter; 8, flowmeter; 9, diaphragm pump; 10, gas cell.
Table 2. Selectivities for a Feed Containing Pure NOa T [°C]
NOOUT [ppm]
N2OOUT [ppm]
NH3,OUT [ppm]
SCR selectivity [%]
N2O selectivity [%]
200 250 300 350 400 450
173 46 10 4 16 81
0 0 0 1 5 16
180 44 13 7 3 2
100 100 100 100 96 73
0 0 0 0.3 1.7 5.4
a
Feed: 300 ppm NO, 300 ppm NH3, 5% H2O, 10% O2, balance
N2.
Figure 2. Overall NH3 conversion and selectivity for the ammonia oxidation reactions (2)-(5) in the absence of water. Feed: 1000 ppm NH3, 10% O2, balance N2.
(Brooks 5850S). Water was dosed into an electrically heated evaporator by means of a liquid mass flow controller (Brooks 5881). The gas flow was adjusted to 0.38 mN3/h, resulting in a GHSV of 52 000 h-1. 4.3. Analysis of Gaseous Components. The gases at the reactor outlet were continuously analyzed by means of an FTIR spectrometer (Nicolet Magna IR 560, OMNIC QuantPad software) equipped with a heated multiple-pass gas cell (Graseby Specac G-2-4-BA-AU, path length 2 m) and a mercury cadmium telluride detector cooled by liquid nitrogen. The method developed allows one to quantify the concentrations of NO, NO2, N2O, NH3, and H2O. The detection limits are 1 ppm for NO2, N2O, and NH3, 5 ppm for NO, and 500 ppm for H2O. 5. Results 5.1. Direct Oxidation of Ammonia. The overall conversion of NH3 in 10% oxygen over a TiO2-WO3V2O5 catalyst may be seen in Figure 2 for a dry feed and in Figure 3 for a humid feed. The conversion of ammonia increases with temperature, and it is strongly inhibited by the presence of water. For example, at 350 °C, the conversion of ammonia is 85% with a dry feed, but it is below 2% when water is present.
The amount of ammonia consumed in reactions (3)(5) is determined indirectly from the amounts of N2O, NO, and NO2 measured at the catalyst outlet. The remaining amount of ammonia consumed in the process must have been oxidized to nitrogen according to the SCO reaction (2). The selectivities for reactions (2)-(5) were evaluated according to (23), and the results are plotted in Figures 2 and 3 as a function of temperature for dry and humid feeds. At low temperatures the selectivity for the SCO of ammonia is high but decreases with rising temperature. In a dry feed the oxidation of ammonia forms a relevant amount of N2O above 250 °C. Moreover, at 450 °C, significant amounts of NO and NO2 are produced (Figure 2). According to Figure 3, the presence of water strongly inhibits all reactions but enhances considerably the selectivity for the SCO reaction (2). Only at temperatures above 350 °C does N2O formation according to reaction (3) become noticeable, and this leads to a loss in the selectivity for the SCO reaction. The amounts of NO and NO2 formed by direct oxidation of ammonia in a humid feed remain negligible up to 450 °C. 5.2. Selectivity and DeNOx under SCR Conditions. NOx Consisting of Pure NO. Under conditions prevailing in a typical exhaust gas, NO is consumed either in the standard SCR reaction (1) or in reaction (6) forming N2O. As shown in the previous paragraph, the formation of NO by direct oxidation of ammonia may be neglected in a humid feed. However, SCO of ammonia may occur at very high temperatures. Table 2 reports the gas concentrations at the reactor outlet for a feed containing 300 ppm NO and 300 ppm NH3 in the humid base feed as well as the selectivities for SCR and N2O formation. The selectivity for the standard SCR reaction was evaluated according to (20). The selectivity for the formation of N2O was evaluated
4012
Ind. Eng. Chem. Res., Vol. 41, No. 16, 2002
Table 3. Selectivities for a Feed Containing an Equimolar Mixture of NO and NO2a
Table 5. Selectivities for the Standard, the Fast, and the NO2 SCR Reaction in Dry and Humid Feedsa
SCR N2O T NOOUT NO2,OUT N2OOUT NH3,OUT selectivity selectivity [%] [°C] [ppm] [ppm] [ppm] [ppm] [%] 200 250 300 350 400 450
4 2 0 2 14 83
4 0 0 0 2 10
0 0 0 0 4 16
8 0 0 0 0 0
100 100 100 100 95 69
0 0 0 0 1.3 5.3
a Feed: 150 ppm NO, 150 ppm NO , 300 ppm NH , 5% H O, 2 3 2 10% O2, balance N2.
Table 4. Selectivities for a Feed Containing Pure NO2a SCR N2O T NOOUT NO2,OUT N2OOUT NH3,OUT selectivity selectivity [%] [°C] [ppm] [ppm] [ppm] [ppm] [%] 200 250 300 350 400 450 a
0 0 1 4 6 21
235 257 187 87 64 54
0 4 7 9 11 13
233 252 149 16 3 3
129 119 100 100 106 109
0 7.1 4 2.7 3.2 3.8
Feed: 300 ppm NO2, 300 ppm NH3, 5% H2O, 10% O2, balance
N2.
according to (22), using the stoichiometric coefficients of reaction (6), νNH3/νN2O ) 1. At temperatures below 400 °C, the amount of N2O produced is negligible, resulting in a very high selectivity for the standard SCR reaction. At temperatures above 350 °C, the selectivity for the standard SCR reaction decreases with increasing temperature. At 450 °C the sum of the selectivities for the standard SCR reaction and for N2O formation is lower than 100%. The missing difference must be attributed to SCO of ammonia [reaction (2)]. NOx Consisting of an Equimolar Mixture of NO and NO2. The selectivity of the fast SCR reaction was investigated with 150 ppm NO, 150 ppm NO2, and 300 ppm NH3 added to the humid base feed. Table 3 reports the concentrations of the gases at the reactor outlet and the selectivities for SCR and N2O formation. The selectivity for the fast SCR reaction was evaluated according to (20) and the selectivity for the formation of N2O according to (22) using the stoichiometric coefficients of reaction (6), νNH3/νN2O ) 1. As in the case of the standard SCR, the selectivity for the fast SCR reaction is very high at temperatures below 400 °C, then it decreases with increasing temperature. Vice versa, the selectivity for the formation of N2O increases with increasing temperature. Again, at 450 °C, the sum of the two selectivities is lower than 100%. The missing difference must be attributed to SCO of ammonia [reaction (2)]. NOx Consisting of Pure NO2. The SCR reaction with NO2 according to (8) is much slower than the fast SCR reaction and still slower than the standard SCR reaction.14 Table 4 resumes the results of experiments with a humid base feed containing 300 ppm NO2 and 300 ppm NH3. The SCR selectivity was calculated using (21). The N2O selectivity was evaluated according to (22), using the stoichiometric coefficients of reaction (9), νNH3/ νN2O ) 6/7. Because of the fact that at low temperatures N2O is probably formed according to reaction (16) with νNH3/νN2O ) 2, the real N2O selectivities at low temperatures would be higher. However, the contribution of reactions (9) and (16) to N2O formation at various
selectivity standard SCR
selectivity fast SCR
selectivity NO2 SCR
T [°C]
dry feed [%]
humid feed [%]
dry feed [%]
humid feed [%]
dry feed [%]
humid feed [%]
200 250 300 350 400 450
100 100 98 95 70 29
100 100 100 100 96 73
100 100 100 96 69 28
100 100 100 100 95 69
139 119 103 105 111 128
129 119 100 100 106 109
a Feed: 300 ppm NO , 300 ppm NH , 5% H O (humid feed), 10% x 3 2 O2, balance N2.
temperatures cannot be quantified, which justifies the use of νNH3/νN2O of the single reaction (9). It can be seen that the SCR selectivity is 100% only at medium temperatures of 300-350 °C but may easily reach values above 100% at lower and higher temperatures. This means that the NO2 SCR reaction (8) is observed as the sole reaction only at medium temperatures. At low and at high temperatures, the ratio νNH3/ νNO2 is no longer 4/3, as assumed in (21), but converges toward 1, thus resulting in SCR selectivities > 100%. At high temperatures, this is due to conversion of NO2 to NO according to the thermodynamic equilibrium NO + 1/2O2 H NO2, thus enabling the fast SCR reaction (7) with νNH3/νNO2 ) 1. At low temperatures, reactions (15) and (16) compete with the NO2 SCR reaction (8). Both of them have νNH3/νNO2 ) 1, thus explaining the apparent higher SCR selectivities at low temperatures. When the N2O selectivity is looked at, it can be seen that the trend is completely different from the first two cases, where N2O formation was restricted to the highest temperatures. In the case of pure NO2 in the feed a moderate N2O selectivity is observed over the whole temperature range of 250-450 °C. Influence of Water on the Selectivities. Table 5 compares calculated SCR selectivities for humid and dry feeds at various temperatures. It can be seen that the presence of water increases the selectivity of both the standard and the fast SCR reactions. The beneficial effect of water leads to a widening of the temperature window toward higher temperatures at which high SCR selectivities are observed. For the case of NO2 in the feed, it appears that the selectivities are higher in the case of the dry feed. Selectivities above 100% must again be interpreted as a stoichiometry νNH3/νNO2 differing from the reaction for pure NO2 SCR (νNH3/νNO2 ) 4/3), i.e., a contribution from standard and fast SCR reactions (νNH3/νNO2 ) 1). This means that the apparent higher selectivities in the case of dry feed are due to a higher formation of intermediate NO from NO2, thus leading to an increased contribution of fast SCR to the total NOx conversion. The higher redox activity of the catalyst with dry feed thus speeds up the reaction of the thermal equilibrium NO + 1/ O H NO . 2 2 2 Dependence of DeNOx on Temperature. Figure 4 reports the NOx conversion (“DeNOx”) as a function of temperature for a humid feed with 300 ppm of NOx for the three cases: pure NO, equimolar mixture of NO and NO2, and pure NO2. The rest of the feed consists of 300 ppm NH3, 5% H2O, and 10% O2 with balance N2. In the case of pure NO, the activity of the catalyst at 200 °C is low, resulting in low values of DeNOx. With
Ind. Eng. Chem. Res., Vol. 41, No. 16, 2002 4013
Figure 4. DeNOx as a function of temperature for various humid feeds: (a) 150 ppm NO + 150 ppm NO2; (b) 300 ppm NO; (c) 300 ppm NO2. The rest of the feed consists of all cases of 300 ppm NH3, 10% O2, 5% H2O, balance N2.
Figure 5. DeNOx as a function of temperature for various dry feeds: (a) 150 ppm NO + 150 ppm NO2; (b) 300 ppm NO; (c) 300 ppm NO2. The rest of the feed consists of all cases of 300 ppm NH3, 10% O2, balance N2.
increasing temperature, the catalyst becomes more active, resulting in a rising DeNOx. Similar to the values shown in Table 2, ammonia is almost fully consumed at temperatures of 350 °C and higher. However, a decreasing DeNOx is observed above 350 °C. Therefore, at high temperatures, part of the ammonia must be consumed in side reactions, leaving less ammonia to react in the standard SCR reaction. This leads to the maximum of DeNOx at about 350 °C in Figure 4. Similar results were obtained for feeds with NOx consisting of pure NO2 or an equimolar mixture of NO and NO2. In case of a feed with pure NO2, the negative temperature coefficient of reaction (15) leads to the minimum in the curve of DeNOx vs temperature at about 250 °C.13 Influence of Water on DeNOx. In a similar way Figure 5 compares the temperature dependence of DeNOx for a dry feed with that of NOx consisting of pure NO, an equimolar mixture of NO and NO2, and pure NO2. Comparing Figures 4 and 5 leads to the conclusion that the presence of water influences both the activity and selectivity of the three SCR reactions. Generally, the presence of water inhibits all SCR reactions. However, for a feed containing the equimolar mixture of NO and NO2, the influence of water is very small because of the high rate of the fast SCR reaction even in the presence of water.
Figure 6. N2O formation as a function of temperature in the absence and in the presence of water. Feed: 1000 ppm NOx, 1000 ppm NH3, 10% O2, 5% H2O (humid feed), balance N2.
Figure 7. N2O formation as a function of NO2,IN/NOx,IN in the presence of water at T ) 300 °C. Feed: 1000 ppm NOx, 1000 ppm NH3, 10% O2, 5% H2O, balance N2.
5.3. Formation of N2O. Figure 6 reports the formation of N2O for dry and humid feeds with different ratios of NO2/NOx. It can be seen that the formation of N2O for dry feeds containing either pure NO or equimolar NO/NO2 increases with temperature. N2O can be formed by direct oxidation of ammonia and by the reaction between NH3 and NOx. Comparing experiments carried out with and without NOx in the dry feed have shown that most of the N2O is formed in the reactions between NH3 and NOx. If the dry feed contains pure NO2, the amount of nitrous oxide formed exhibits a maximum at low temperatures (≈250 °C). On the other hand, this feed leads to the lowest formation of N2O at high temperatures. If the feeds contains water, the N2O concentrations are much lower, indicating that formation of N2O is strongly inhibited by water. This is true for all three feeds. For feeds containing either pure NO or equimolar NO/NO2, the amount of N2O formed at temperatures below 400 °C is low (some ppm). In the case of a feed containing pure NO2, a considerable formation of N2O is observed already at 250 °C with a flat maximum at ≈350 °C. In another series of experiments, the influence of the ratio NO2/NOx on the formation of nitrous oxide for humid feeds was tested. Figure 7 shows that the formation of N2O increases with the fraction of NO2 in the feed if the ratio of NO2/NOx is higher than 50%. The influence of oxygen on the formation of N2O was investigated for both dry and humid feeds. These
4014
Ind. Eng. Chem. Res., Vol. 41, No. 16, 2002
experiments have shown that the formation of N2O is not influenced by the presence of oxygen for feeds containing pure NO2. This proves that reaction (10) can be neglected. 6. Discussion The selectivity of the standard and fast SCR reactions over “real” SCR catalysts under “real” operating conditions ()presence of water) is very high at temperatures below 400 °C (Table 5). However, a decrease in SCR selectivity is observed at temperatures above 350 °C. The experiments have shown that the decrease in SCR selectivity of the standard and fast SCR reactions is mainly due to the SCO of ammonia and to the formation of nitrous oxide. Other hypothetical reactions do not significantly contribute to the loss of SCR selectivity under real exhaust gas conditions. The presence of water inhibits the SCR activity of all three reactions and increases the selectivity in the case of the standard and fast SCR reactions. In the following, we propose an explanation for the influence of water on the SCR reaction considering some published reaction mechanisms. Many authors11,16,18 agree that the presence of water converts part of the Lewis acid sites of the catalyst into Brønsted acid sites. Ramis et al.16 proposed that the activation of ammonia on a Lewis acid site is the fundamental step in the mechanism of the standard SCR reaction. Therefore, the inhibiting effect of water at low temperatures could be due to a decrease of the surface concentration of Lewis acid sites. In a later paper Ramis et al.9 proposed hydrazine (N2H4) as an intermediate in the SCO of ammonia. They suggested that N2H4 is formed by dimerization of two NH2 species originating from the dissociative adsorption of ammonia on two adjacent Lewis acid sites. At high temperatures the adsorption of water is weakened and this should lead to a higher amount of (adjacent) Lewis acid sites. This should favor the SCO of ammonia and consequently cause a loss of selectivity for the SCR reactions. This effect is even more pronounced in dry conditions, resulting in the very low SCR selectivities observed at high temperatures (Table 5). The mechanism of N2O formation by the reaction between NH3 and NO has not been fully explained. Topsoe et al.8 proposed that the active sites for the N2O formation and the SCR reactions are the same and suggested a nitrosamidic species as intermediate in the formation of nitrous oxide. The presence of water influences the decomposition of this intermediate either to nitrogen or to nitrous oxide according to dry: dehydrogenation
wet: dehydration
N2O 79 NHxNO 98 N2 (24) Topsoe et al.8 proposed that the surface hydroxylation by water promotes dehydration, leading to the formation of nitrogen. This could explain the strong inhibiting effect of water on the formation of N2O. For a feed with high amounts of NO2, i.e., NO2/NOx > 1, a considerable formation of N2O is also seen at low temperatures (Figures 6 and 7) and this amount increases with increasing NO2 fraction. The formation of N2O is much higher in a dry feed than in a humid feed, and a pronounced maximum is found around 250 °C. We attribute this formation of N2O at low temperatures to a second reaction mechanism depending on a reaction
step with negative temperature dependence. The dimerization equilibrium of NO2 [reaction (12)] shows the required negative temperature dependence. A similar behavior, i.e., a maximum in the NOx conversion for humid feeds containing only NO2, has recently been observed, and this was attributed to the competition of the ammonium nitrate reaction (15) with the NO2 SCR reaction (8).13 The formation of solid ammonium nitrate by reactions (12)-(15) could also be shown at temperatures below ≈200 °C. Although the maximum observed for N2O formation around 250 °C seems somewhat high for the formation of the intermediate ammonium nitrate, we propose a reaction pathway for N2O formation based on the disproportionation of N2O4. Ammonium nitrate would form according to the overall reaction (15) and subsequently decompose into water and nitrous oxide. It is well-known that rapid heating of bulk ammonium nitrate will yield nitrous oxide and water at temperatures up to 260 °C.19 Another possible mechanism would involve the formation of the anhydride of ammonium nitrate according to
2NH3 + N2O4 f H2N-NO2 + N2 + 2H2O (25) The intermediate nitroamine is the analogue to nitrosamine postulated in the SCR mechanism proposed by Ramis et al.16 The fact that water shows an inhibiting effect on the N2O formation at low temperatures suggests that this species may be a more probable intermediate than ammonium nitrate. An ammonium nitrate intermediate species has also been proposed by Centi et al. in the formation of N2O by the SCR process on Cu/Al2O3 catalysts.20-22 At intermediate temperatures, reaction (9) is responsible for the formation of N2O. At high temperatures, the thermodynamic equilibrium NO2 H NO + 1/2O2 will provide increasing fractions of NO, thus allowing also for N2O formation according to reaction (6). 7. Conclusions Under typical exhaust gas conditions, the selectivity for both the standard and fast SCR reactions is high at temperatures below 400 °C. The presence of water favors the SCR selectivity, whereas it inhibits the SCR activity. SCO of ammonia and the formation of nitrous oxide are the main side reactions, which lower the SCR selectivity above 350 °C. For feeds with NOx consisting of pure NO or equimolar NO and NO2, the formation of N2O is negligible at temperatures below 400 °C. An additional strong formation of N2O at low temperatures is observed for feeds with a ratio of NO2/NOx above 50%. The reaction responsible for N2O formation at low temperatures probably involves ammonium nitrate or nitroamine as an intermediate species. Acknowledgment The financial support of the Swiss Federal Office of Energy (BFE) is gratefully acknowledged. Literature Cited (1) Bosch, H.; Janssen, F. Catalytic Reduction of Nitrogen Oxides. Catal. Today 1998, 2, 369. (2) Lietti, L.; Forzatti, P. Recent advances in DeNO(x)ing catalysis for stationary applications. Hetercycl. Chem. Rev. 1996, 3, 33.
Ind. Eng. Chem. Res., Vol. 41, No. 16, 2002 4015 (3) Koebel, M.; Elsener, M.; Kleemann, M. Urea-SCR: a promising technique to reduce NOx emissions from automotive diesel engines. Catal. Today 2000, 59, 335. (4) Weiss, R. F. The temporal and spatial distribution of tropospheric nitrous oxide. J. Geophys. Res. 1981, 86, 7185. (5) Lietti, L.; Nova, I.; Ramis, G.; Dall’ Acqua, L.; Busca, G.; Giamello, E.; Forzatti, P.; Bregani, F. Characterization and reactivity of V2O5-MoO3/TiO2 DeNOx SCR catalysts. J. Catal. 1999, 187, 419. (6) Il’Chenko, N. I.; Golodets, G. I. Catalytic Oxidation of Ammonia. J. Catal. 1975, 39, 57. (7) Li, Y.; Armor, J. N. Selective NH3 oxidation in a wet stream. Appl. Catal. B 1997, 13, 131. (8) Topsoe, N. Y.; Slabiak, T.; Clausen, B. S.; Srnak, T. Z.; Dumesic, J. A. Influence of water on the reactivity of vanadia/ titania for catalytic reduction of NOx. J. Catal. 1992, 134, 742. (9) Ramis, G.; Yi, L.; Busca, G. Ammonia activation over catalysts for the selective catalytic reduction of NOx and the selective catalytic oxidation of NH3. An FT-IR study. Catal. Today 1996, 28, 373. (10) Okzan, U. S.; Cai, Y.; Kumthekar, M. W. Investigation of the Reaction Pathways in the Selective Catalytic Reduction of NO with NH3 over V2O5 Catalysts: Isotopic Labeling Studies using 18O , 15NH , 15NO and 15N18O. J. Catal. 1994, 149, 390. 2 3 (11) Duffy, B. L. D.; Curry Hyde, H. E.; Cant, N. W.; Nelson, P. F. Isotopic labeling studies of the effects of temperature, water and vanadia loading on the selective catalytic reduction of NO with NH3 over vanadia-titania catalysts. J. Phys. Chem. 1994, 98, 7153. (12) Jacob, E.; Emmerling, G.; Do¨ring, A.; Graf, U.; Harris, M.; van den Tillaart, J. A. A.; Hupfeld, B. NOx-Verminderung fu¨r Nutzfahrzeugmotoren mit Harnstoff-SCR Kompaktsystemen. 19th Internationale Wiener Motorensymposium, Vienna, May 7-8, 1998. (13) Koebel, M.; Elsener, M.; Madia, G. Reaction Pathways in the Selective Catalytic Reduction Process with NO and NO2 at Low Temperatures. Ind. Chem. Eng. Res. 2001, 40, 52.
(14) Koebel, M.; Madia, G.; Elsener, M. Selective catalytic reduction of NO and NO2 at low temperatures. Catal. Today 2002, 2660, 1. (15) Blanco, J.; Avila, P.; Suarez, S.; Martin, J. A.; Knapp, C. Alumina- and titania-based monolithic catalysts for low temperature selective catalytic reduction of nitric oxides. Appl. Catal. B 2000, 28, 235. (16) Ramis, G.; Busca, G.; Bregani, F.; Forzatti, P. Fourier Transform Infrared Study of the Adsorption and Co-adsorption of Nitric Oxide, Nitrogen Dioxide and Ammonia on Vanadia-Titania and Mechanism of Selective Catalytic Reduction. Appl. Catal. B 1990, 64, 259. (17) Kleemann, M. Beschichtung von Cordierit-Wabenko¨rpern fu¨r die selektive katalytische Reduktion von Stickoxiden. Ph.D. Dissertation Nr. 13401, ETH Zurich, Zurich, Switzerland, 1999. (18) Topsoe, N. Y. Characterization of the nature of surface sites on vanadia titania catalysts by FTIR. J. Catal. 1991, 128, 499. (19) Gmelin. Handbuch der anorganischen Chemie, Ammonium, 8th ed.; Verlag Chemie: Leipzig, Germany, 1936/69; p 109. (20) Centi, G.; Perathoner, S. Nature of active species in copperbased catalysts and their chemistry of transformation of nitrogen oxides. Appl. Catal. A 1995, 132, 179. (21) Centi, G.; Perathoner, S.; Biglino, P.; Giamiello, E. Adsorption and reactivity of NO on copper on alumina catalysts. 1. Formation of nitrate species and their influence on reactivity in NO and NH3 conversion. J. Catal. 1995, 152, 75. (22) Centi, G.; Perathoner, S. Adsorption and reactivity of NO on copper on alumina catalysts. 2. Adsorbed species and competitive pathways in the reaction of NO with NH3 and O2. J. Catal. 1995, 152, 93.
Received for review January 17, 2002 Revised manuscript received May 30, 2002 Accepted May 9, 2002 IE020054C
