The selective catalytic reduction of nitrogen oxides by methane on noble metal-loaded sulfated zirconia

The selective catalytic reduction of nitrogen oxides by methane on noble metal-loaded sulfated zirconia

Applied Catalysis B: Environmental 33 (2001) 325–333 The selective catalytic reduction of nitrogen oxides by methane on noble metal-loaded sulfated z...

144KB Sizes 3 Downloads 36 Views

Applied Catalysis B: Environmental 33 (2001) 325–333

The selective catalytic reduction of nitrogen oxides by methane on noble metal-loaded sulfated zirconia Hirofumi Ohtsuka∗ Research and Development Department, Osaka Gas Co., Ltd., 6-19-9, Torishima, Konohana-ku, Osaka 554-0051, Japan Received 13 February 2001; received in revised form 19 April 2001; accepted 24 April 2001

Abstract The selective catalytic reduction of NOx by methane on noble metal-loaded sulfated zirconia (SZ) catalysts was studied. Ru, Rh, Pd, Ag, Ir, Pt, and Au-loaded sulfated zirconia catalysts were compared with the intact sulfated zirconia. For the NO–CH4 –O2 reaction, Ru, Rh, Pd, Ir, and Pt showed promotion effect on NOx reduction, while for the NO2 –CH4 –O2 reaction, only Rh and Pd showed promotion effect. Over intact and Rh, Pd, Ag, and Au-loaded sulfated zirconia, NOx conversion in NO2 –CH4 –O2 reaction was significantly higher than that in NO–CH4 –O2 reaction, while clear difference was not observed over Ru, Ir, and Pt-loaded sulfated zirconia. Comparison of [NO2 ]/([NO] + [NO2 ]) in the effluent gases in NO–O2 and NO2 –O2 reactions showed that Ru, Ir, and Pt has high activity for NO oxidation under the reaction conditions. These facts suggest that effects of these metals toward NOx reduction by methane can be categorized into the following three groups: (i) low activity for NO oxidation to NO2 , and high activity for NO2 reduction to N2 (Pd, Rh); (ii) high activity for NO oxidation to NO2 , and low activity for NO2 reduction to N2 (Ru, Ir, Pt); (iii) low activity for both reactions (Ag, Au). To confirm these suggestions, combination of these metals were investigated on binary or physically-mixed catalysts. The combination of Pd or Rh with Pt or Ru gave high activity for the selective reduction of NOx by methane. © 2001 Elsevier Science B.V. All rights reserved. Keywords: NOx reduction; Palladium; Platinum; Sulfated zirconia; Methane

1. Introduction The selective catalytic reduction of nitrogen oxides (NOx ) by hydrocarbons has received great attention since it may offer an attractive alternative to the selective catalytic reduction by ammonia (ammonia-SCR). Ammonia-SCR has been successfully applied to a number of power plants, but is not suitable for mobile or small-scale stationary sources. The reasons for this include storage and handling of the reductant and the possibility of ammonia slip which may occur when too much reductant is added. ∗ Tel.: +81-6-6462-3231; fax: +81-6-6462-3433. E-mail address: [email protected] (H. Ohtsuka).

The selective catalytic reduction of NOx by methane is especially suitable for small-scale stationary sources such as on-site power plants because methane is easily available as natural gas, which is supplied through pipelines in many countries, and many on-site power plants use natural gas as a fuel. To date, various catalysts such as Co-, Mn-, and Ni-loaded zeolites [1,2], Pd-loaded zeolites [3,4], Inand Ga-loaded zeolites [5,6], and Ag/alumina [7] have been reported for the reaction. However, most of these are severely affected by the presence of water vapor [6,8], and do not show practical activity under actual exhaust conditions, at least below 500◦ C. Uchida et al. reported that Pd/MOR showed NOx reduction activity of a practically applicable level

0926-3373/01/$ – see front matter © 2001 Elsevier Science B.V. All rights reserved. PII: S 0 9 2 6 - 3 3 7 3 ( 0 1 ) 0 0 1 9 0 - 4

326

H. Ohtsuka / Applied Catalysis B: Environmental 33 (2001) 325–333

even in the presence of 10% water vapor at 400◦ C [9]. However, even this catalyst is not stable under actual reaction conditions [10]. Many studies have been reported of attempts to improve Pd/zeolite catalysts. Misono et al. reported that the addition of Rh to Pd/ZSM-5 improved the activity in the presence of water vapor [11]. Ogura et al. reported that the addition of Co to Pd/ZSM-5 greatly enhanced both activity and durability in the presence of water vapor [12,13]. Suzuki et al. reported that deactivation of Pd/ZSM-5 can be suppressed by silica CVD on the catalyst [14]. In the actual exhaust, sulfur oxides (SOx ) are present in addition to water vapor. The SOx concentration is at the ppm level for gasoline exhaust and the sub-ppm level for natural gas exhaust. The author compared the activity of Pd/MOR and Pd-Pt/sulfated zirconia both in the presence and absence of SO2 [15]. Pd/MOR showed modest deactivation in the absence of SO2 but lost almost all activity in 50 h in the presence of 3 ppm SO2 . On the other hand, Pd-Pt/sulfated zirconia showed stable activity in the presence of SO2 . After further experiments, a bifunctional mechanism, Pt catalyzing NO oxidation to NO2 and Pd catalyzing the reaction of NO2 and CH4 , was suggested [16]. However, the reason for the difference in the effect of Pd and Pt on NOx reduction is still unclear, and the reactivities of other metals for the reaction are still unknown. In this paper, various noble metal-loaded sulfated zirconia are examined for the selective catalytic reduction of NOx by methane. Their reactivities are further examined by combining their reactivities on binary or physically-mixed catalysts.

N.E. Chemcat, Tokyo), AgNO3 , and HAuCl4 ·4H2 O (both from Hayashi Pure Chemicals, Osaka) were used for impregnation. Iridium was impregnated using an acetone solution of iridium acetylacetonate (Aldrich). The impregnated solid was dried and calcined in air at 500◦ C. In this paper, metal-loadings are shown in nominal wt.% calculated from the metal amount in the solution and the amount of SZ. Catalytic activity tests were carried out using a fixed-bed flow reactor. Details of the catalytic activity tests have been described elsewhere [15]. Briefly, a 4 ml catalyst sieved into 1–2 mm grains was placed in a stainless steel reactor (i.d. 14 mm) and the gas flow rate was 1 l/min, corresponding to a GHSV of 15,000 h−1 . The catalyst bed was heated in dry He flow until the temperature reached 300◦ C. After that, the gas was switched to the reaction gas. No further pretreatment of the catalyst such as oxidation or reduction was carried out. Typical reaction gas compositions were: NOx (NO or NO2 ) 150 ppm, CH4 2000 ppm, O2 10%, H2 O 9%, and the balance of He. The catalyst was kept for at least 45 min at the desired temperature in the flow of the reaction gas before sampling. The NO and NO2 concentrations were measured by a chemiluminescence NOx analyzer equipped with an NO2 converter (Yanaco Analytical Systems, Kyoto). CH4 , CO, CO2 , N2 and N2 O concentrations were measured by a gas chromatograph (Yanaco Analytical Systems). NOx conversion was defined by 1−(outlet NOx )/(inlet NOx ), and methane conversion was defined by [(outlet CO) + (outlet CO2 )]/[(outlet CO) + (outlet CO2 ) + (outlet CH4 )]. N2 O formation was not observed in any experiment in this report.

2. Experimental

3. Results and discussion

Sulfated zirconia (SZ) was prepared by immersing 360 g zirconium hydroxide (Mitsuwa Chemical, Osaka) in a solution in which 54 g ammonium sulfate (Hayashi Pure Chemicals, Osaka) was dissolved. The mixture was dried and calcined in air at 550◦ C. The surface area of the resultant SZ was ca. 160 m2 /g as determined by the BET method. XRD analysis revealed that the SZ consisted solely of the tetragonal phase. Noble metals were loaded by impregnation. Aqueous solutions of Ru(NH3 )6 (NO3 )3 , Rh(NH3 )5 (H2 O)(NO3 )3 , Pd(NO3 )2 , Pt(NH3 )4 (NO3 )2 (all from

3.1. Activity of various noble metal-loaded sulfated zirconia for NOx reduction by methane Open bars in Fig. 1 show NOx and methane conversions in NO–CH4 –O2 reaction on various noble metal-loaded SZ. NOx conversion on the sulfated zirconia support showed only marginal activity. Pd/SZ showed the highest NOx conversion among the examined samples. Ru/SZ, Rh/SZ, Pt/SZ, and Ir/SZ followed in this order. NOx conversions on Ag/SZ and Au/SZ were low and almost the same as that on SZ.

H. Ohtsuka / Applied Catalysis B: Environmental 33 (2001) 325–333

327

Fig. 1. NOx and methane conversions in NO–CH4 –O2 (open bars); NO2 –CH4 –O2 (hatched bars) and CH4 –O2 reaction (solid bars) on sulfated zirconia and noble metal-loaded sulfated zirconia. Metal-loading: 0.5 wt.%. Reaction conditions: NO 150 ppm, CH4 2000 ppm (NO–CH4 –O2 ), or NO2 150 ppm, CH4 2000 ppm (NO2 –CH4 –O2 ), or CH4 2000 ppm (CH4 –O2 ) and O2 10%, H2 O 9%, GHSV 15,000 h−1 , 450◦ C.

Methane conversion on pure SZ was very low (1%). Among metal-loaded/SZ, Ir/SZ showed the highest CH4 conversion. Pd/SZ and Pt/SZ showed similar methane conversions of around 12%, and Ru/SZ followed. Ag/SZ and Au/SZ showed only slightly higher methane conversion than pure SZ. From the viewpoint of selectivity which is defined by the molar ratio of (reduced NOx )/(consumed CH4 ), Rh/SZ was the highest, and Ir/SZ was the lowest. Since Ag/SZ and Au/SZ showed NOx and methane conversions similar to those on pure SZ, Ag and Au do not seem to affect the reaction. Hatched bars in Fig. 1 show NOx and methane conversions in NO2 –CH4 –O2 reaction. Pd/SZ showed the highest NOx conversion. Rh/SZ showed comparable NOx conversion with Pd/SZ. Comparing the NOx conversions in the NO–CH4 –O2 reaction and

the NO2 –CH4 –O2 reaction, pure SZ, Rh/SZ, Pd/SZ, Ag/SZ, and Au/SZ showed significantly higher NOx conversion in the NO2 –CH4 –O2 reaction than in the NO–CH4 –O2 reaction, while Ru/SZ, Ir/SZ, and Pt/SZ showed similar NOx conversions under both conditions. The samples which showed higher NOx conversions in the NO2 –CH4 –O2 reaction than those in the NO–CH4 –O2 reaction (pure SZ, Rh/SZ, Pd/SZ, Ag/SZ, and Au/SZ) also showed higher methane conversions in the NO2 –CH4 –O2 reaction than in the NO–CH4 –O2 reaction. In contrast, Ru/SZ and Pt/SZ showed similar methane conversion in both NO–CH4 –O2 and NO2 –CH4 –O2 reactions. Methane conversions in CH4 –O2 reaction are shown by solid bars in Fig. 1. The conversion is not much different from the conversions in the NOx –CH4 –O2 reactions. Ir/SZ showed a methane conversion of 25%

328

H. Ohtsuka / Applied Catalysis B: Environmental 33 (2001) 325–333

in the CH4 –O2 reaction. The high methane combustion activity may be the reason for the low selectivity of the catalyst for NOx reduction. The results in Fig. 1 show that the reactivity of Ag/SZ and Au/SZ is very similar to that of pure SZ, except that Ag/SZ shows slightly higher methane conversions than pure SZ. This suggests that except for the marginal activity of Ag/SZ for the reaction of CH4 with O2 , Ag and Au do not play any role in the reaction. There are only sparse reports on the selective NOx reduction by methane on Ag and Au-loaded catalysts. Among these, Flytzani-Stephanopoulos and coworkers recently reported that Ag/alumina prepared by a single-step co-gelation method showed high activity for NOx reduction by methane over a wide temperature range between 450 and 650◦ C [7]. However, they also showed a significant effect of water vapor and SO2 on NOx conversion below 600◦ C and pointed out that temperatures above 550◦ C are required for practical application. The temperature at which the experiments in Fig. 1 were carried out might be too low to observe the effect of these metals. Fig. 2 shows NO2 ratios in the effluent gases in NO–O2 , NO2 –O2 , NO–CH4 –O2 , and NO2 –CH4 –O2 reactions. The ratio is defined by [NO2 ]/([NO] +

[NO2 ]). A comparison of the ratio under NO–O2 and NO2 –O2 shows the extent to which the following reaction proceeded to the equilibrium: NO + 21 O2 ↔ NO2

(1)

On a blank test without any catalyst, the ratio was 0.10 and 0.82 for the NO–O2 and the NO2 –O2 reactions, respectively. Therefore, reaction (1) proceeds to some extent in an empty reactor, but is still far from reaching equilibrium. On SZ, reaction (1) is still far from equilibrium, but it is noted that the ratios in the NOx –CH4 –O2 reaction are lower than those in the NOx –O2 reaction. This means a reaction proceeds which depletes NO2 . Whether the depleted NO2 is reduced to NO or N2 will be discussed later. The NO2 ratios in the NOx –O2 reactions on Pd/SZ were similar to those on SZ, which indicates that Pd/SZ has a low activity for NO oxidation. The NO2 ratios in the NOx –CH4 –O2 reactions were almost zero, which shows that Pd has high activity for the reaction of NO2 and methane. Rh/SZ showed similar features, although its activity for the reactions of NO2 and methane seems to be lower than that of Pd/SZ. On Ru/SZ, Pt/SZ, and Ir/SZ, NO2 ratios in both NO–O2 and NO2 –O2 reactions showed similar values,

Fig. 2. NO2 ratios in the effluent gas from NO–O2 (䊏), NO2 –O2 (䊐), NO–CH4 –O2 (䉬), NO2 –CH4 –O2 (䉫) reactions. NO or NO2 150 ppm, CH4 0 or 2000 ppm, O2 10%, H2 O 9%, GHSV 15,000 h−1 , 450◦ C. Dotted line: the NO2 ratio under the equilibrium condition.

H. Ohtsuka / Applied Catalysis B: Environmental 33 (2001) 325–333

which means that these catalysts have high activity for NO oxidation. However, these three catalysts showed different activity for the reaction of NO2 and methane. On Ru/SZ, the NO2 ratios in the NOx –CH4 –O2 reactions were only slightly lower than those of NOx –O2 reactions, while on Ir/SZ, they were much lower than those of NOx –O2 reactions. The result indicates that Ir/SZ, and, to a lesser extent Pt/SZ, have activity for the reaction NO2 and methane. No difference could be observed between NO2 ratios on Ag/SZ, Au/SZ, and SZ, which again suggests that Ag and Au do not play any role in either NOx –O2 or NOx –CH4 –O2 reactions. The following four reactions can be considered in NOx –CH4 –O2 reactions: NO oxidation, NO + 21 O2 → NO2

(2)

NO2 reduction to N2 , NO2 + CH4 + O2 → (reaction intermediate) → N2 + CO2 + H2 O

(3)

NO2 reduction to NO, NO2 + CH4 + O2 → (reaction intermediate) → NO + CO2 + H2 O CH4 combustion, CH4 + O2 → CO2 + H2 O

(4) (5)

On Pd/zeolite catalysts, it has already been reported that NOx conversion in NO2 –CH4 –O2 is higher than that in NO–CH4 –O2 , and a mechanism consisting of NO oxidation to NO2 followed by a reaction of NO2 with CHx species formed by methane activation on Pd sites has been proposed [4,17]. Similarly, on Pd/SZ, NOx conversion in NO2 –CH4 –O2 is higher than that in NO–CH4 –O2 . Therefore, we will not consider the direct reaction of NO with methane here, and this possibility will be discussed later. The activity for NO oxidation (reaction 2) can be directly evaluated by the NO2 ratios in the NOx –O2 reactions. The activity for methane combustion (reaction 5) can be evaluated by the methane conversion in the CH4 –O2 reaction. The activity for NO2 reduction to N2 (reaction 3) can be estimated by NOx conversion in the NO2 –CH4 –O2 reaction, but it should be noted that the NO2 ratio in the catalyst bed is not constant and is affected by NO oxidation activity. The activity for NO2 reduction to NO (reaction 4) can only be indirectly estimated by the NO2 ratio in the NO2 –CH4 –O2

329

reaction, but it should be noted that both NO2 reduction to N2 (reaction 3) and NO2 reduction to NO (reaction 4) deplete NO2 , while NO oxidation to NO2 (reaction 2) produces NO2 . The reduction of NO2 to N2 (reaction 3) and its reduction to NO (reaction 4) might be viewed as different routes of the decomposition of the reaction intermediates formed by the reaction of NO2 and CH4 , rather than completely different reactions. Such a mechanism has been proposed for NOx reduction by C3 H8 on Cu-based catalysts [18]. The activity of various metal-loaded SZ will be discussed in terms of the reactions (2)–(5). SZ has low activity for NO oxidation, and therefore we can neglect the NO2 formation by NO oxidation in the catalyst bed. The difference in NO2 concentration in the NO2 –O2 reaction and the NO2 –CH4 –O2 reaction was 29 ppm, which is considered to correspond to the NO2 depleted by the reaction with methane. Comparing the depleted NO2 (29 ppm) and reduced NOx (13 ppm), about one-half of the NO2 reacted with methane was reduced to N2 , although whether the second NOx molecule to give a dinitrogen molecule is NO or NO2 is not known. Both Pd/SZ and Rh/SZ have low activity for NO oxidation. They have high activity for NO2 reduction to N2 , and most of the effluent NOx is NO even when the inlet condition was NO2 –CH4 –O2 . By comparing the NO2 and NO concentrations in the NO2 –O2 and NO2 –CH4 –O2 reactions on Pd/SZ, NO2 concentration decreased by 106 ppm, NO concentration increased by 58 ppm, and the total NOx decrease was 48 ppm with the addition of methane. On Rh/SZ, the NO2 concentration decreased by 86 ppm, and the NO concentration increased by 44 ppm, and the total NOx decrease was 42 ppm. This means that NO2 reduction to NO proceeds faster than NO2 reduction to N2 . It is worth noting that the ratios of the decrease in total NOx to the decrease in NO2 were 0.44–0.49 for SZ, Pd/SZ, and Rh/SZ. It must be noted that this value is not the selectivity in the ordinary sense, which is defined by (reduced NOx )/(consumed methane), but the ratio (NOx reduced to N2 )/(reacted NO2 ). One possible explanation for the similar values on these three catalysts is that the value reflects the selectivity as to whether the reaction intermediate formed by NOx and CHx reacts to form N2 or to form NO, the former possibly involving a reaction with a second NOx molecule and the latter involving only oxygen, and

330

H. Ohtsuka / Applied Catalysis B: Environmental 33 (2001) 325–333

the loaded metals do not affect this reaction step. The selectivity defined by (reduced NOx )/(consumed methane) was 0.15, 0.6 and 0.5 for Pd/SZ, Rh/SZ and pure SZ, respectively. Rh/SZ and pure SZ showed similar values, but Pd/SZ showed a significantly lower value. This is because methane combustion is significant over Pd/SZ, as shown by the high methane conversion in the CH4 –O2 reaction (Fig. 1). On Ru/SZ, Ir/SZ, and Pt/SZ, the NO–CH4 –O2 and NO2 –CH4 –O2 reactions yield similar NOx conversions because they have high activity for NO oxidation and the NO2 ratio rapidly converges regardless of whether NO or NO2 is used. The NOx conversions on these catalysts are similar to that on SZ in the NO2 – CH4 –O2 reaction, which suggests that SZ plays a key role in the NO2 reduction to N2 on these catalysts. For zeolite-based catalysts, it is reported that H-ZSM-5 shows moderate activity for NO reduction by methane, although the activity is significantly lower than its Pd-loaded forms, and direct involvement of acid sites in the reaction as well as methane activation on Pd sites are proposed [11]. Similar roles of acid sites and Pd can be considered for SZ and Pd/SZ. The possibility that NO directly reacts with methane should be mentioned. On SZ and Rh/SZ, NO reaction with methane should be a minor reaction because the NO–CH4 –O2 reaction showed a much lower NOx conversion than the NO2 –CH4 –O2 reaction. On Pd/SZ, NO reaction with methane should be slower than NO2 reaction with methane considering the lower NOx conversion in the NO–CH4 –O2 reaction than that of NO2 –CH4 –O2 reaction, but the fact that 28% NOx conversion can be obtained in the NO–CH4 –O2 reaction where NO2 ratio is below 13% (NO2 ratio under NO–O2 reaction) may suggest the involvement of a direct reaction of NO and methane. It is not possible to distinguish NO reaction with methane and NO2 reaction with methane on Ru/SZ, Ir/SZ and Pt/SZ, because they have high activity for NO oxidation and similar NO2 ratios are attained regardless of whether NO or NO2 was used. 3.2. The effect of various noble metal additions on Pd and Rh/sulfated zirconia on the NOx reduction by methane Fig. 3 shows the activities of various bimetallic catalysts and physically-mixed catalysts prepared

from Pd/SZ or Rh/SZ and other noble metal-loaded SZ in the NO–CH4 –O2 reaction. Methane combustion activity of the catalysts in the CH4 –O2 reaction are also shown in Fig. 3. Sulfated zirconia loaded with both Pd and Pt (Pd-Pt/SZ) and a physical mixture of Pd/SZ and Pt/SZ showed much higher NOx conversion than Pd/SZ. On both catalysts the NO2 ratio under both NO–O2 and NO2 –O2 showed similar value (0.34–0.38), indicating that these catalysts have high activity for NO oxidation. The NO2 ratio in both the NO–CH4 –O2 and NO2 –CH4 –O2 reactions were below 0.1. It is worth noting that methane conversion in NO–CH4 –O2 was greatly enhanced by the presence of Pt, while methane conversion in CH4 –O2 was less affected by Pt. These facts indicate that while on Pd/SZ, NO2 is rapidly consumed and the reaction with methane cannot proceed, in the presence of Pt, NO2 is continuously supplied by NO oxidation on Pt, and the reaction proceeds to yield a high NOx conversion. This agrees with the fact that the reaction of NO2 with methane proceeds much faster than the reaction of NO with methane on Pd/SZ. Comparing the reactivity of the Pd-Pt/SZ and the physical mixture of Pd/SZ and Pt/SZ, they showed similar NOx conversions, but Pd-Pt/SZ showed a slightly higher methane conversion than the physicallymixed catalyst. In addition, comparing methane conversions in the CH4 –O2 condition, Pd-Pt/SZ showed a significantly higher conversion than that of Pd/SZ, while physically-mixed catalysts showed a conversion similar to Pd/SZ. Pd-Pt bimetallic catalysts are known to have high activity for methane combustion [19], and to obtain high selectivity for the NOx reduction, a physical mixture of Pd/SZ and Pt/SZ may be preferable to bimetallic Pd-Pt/SZ. The addition of Ru similarly promoted NOx reduction, and the addition of Rh was not effective, which is in agreement with their NO oxidation activity. Pd-Ru/SZ showed lower activity than the physical mixture of Pd/SZ and Ru/SZ. In addition, the former yielded a much lower methane conversion than the latter. Ru may affect the state of Pd on SZ when loaded on the same support. However, the methane conversion on Pd-Ru/SZ is low, which means it has high selectivity for NOx reduction. The physical mixture of Pd/SZ and Ir/SZ showed a NOx conversion similar to Pd/SZ, and no promotion effect was found, although it showed promoted NO

H. Ohtsuka / Applied Catalysis B: Environmental 33 (2001) 325–333

331

Fig. 3. NOx and methane conversions in NO–CH4 –O2 (open bars) and CH4 –O2 reaction (solid bars) on noble metal-loaded sulfated zirconia. Reaction conditions: same as in Fig. 1. Pd: 0.5% Pd/SZ, Pd-M (M = Pt, Ru, Rh): 0.38% Pd-0.12% M/SZ, Pd/M (M = Pt, Ru, Rh, Ir): 3:1 mixture of 0.5% Pd/SZ and 0.5% M/SZ, Rh: 0.5% Rh/SZ, Rh/Pt: 3:1 mixture of 0.5% Rh/SZ and 0.5% Pt/SZ.

oxidation activity (NO2 ratio: 0.32 under NO–O2 ; 0.43 under NO2 –O2 ). As shown in Fig. 2, Ir/SZ has high activity for NO oxidation, but also has high activity for NO2 reduction to NO and does not increase the NO2 ratio under NOx –CH4 –O2 conditions, which explains the result that the addition of Ir does not promote NOx reduction. Since Rh/SZ showed a higher NOx conversion in NO2 –CH4 –O2 than in NO–CH4 –O2 , an effect of Pt addition similar to that on Pd/SZ can be expected. Indeed, a physical mixture of Rh/SZ and Pt/SZ showed a NOx conversion of 31%, which is higher than the 14% of Rh/SZ. However, the conversion on the physical mixture of Rh/SZ and Pt/SZ was much lower than the physical mixture of Pd/SZ and Pt/SZ. The difference between Pd/SZ and Rh/SZ will be further examined in the following section. Considering that the selective catalytic reduction of NOx by methane proceeds in the bifunctional mechanism of NO oxidation to NO2 and NO2 reaction with

methane, Ir/SZ should have high activity, because Ir is active for both reactions. The reactivity of Ir/SZ requires further studies including the state of loaded Ir and dispersion. 3.3. Comparison of Pd/SZ and Rh/SZ Fig. 4 shows the temperature dependence of the activity of Pd/SZ, Rh/SZ, and their physical mixtures with Pt/SZ. Obviously, Pd/SZ and Rh/SZ show similar activities and similar effects from Pt addition, except that the conversion curves of Rh/SZ are ca. 50◦ C higher than those of Pd/SZ. Pd/SZ showed a maximum NOx conversion of 29% at 450◦ C, while Rh/SZ showed a maximum NOx conversion of 32% at 500◦ C. When mixed with Pt/SZ, Pd/SZ showed a maximum NOx conversion of 55% at 450◦ C, while Rh/SZ showed 56% at 500◦ C. Pd/SZ showed a significant difference in NOx conversion in the NO–CH4 –O2 and NO2 –CH4 –O2 reactions

332

H. Ohtsuka / Applied Catalysis B: Environmental 33 (2001) 325–333

Fig. 4. Temperature dependence of the activity for NOx reduction by methane on: (a) 0.5% Pd/SZ; (b) 0.5% Rh/SZ; (c) 3:1 mixture of 0.5% Pd/SZ and 0.5% Pt/SZ, and (d) 3:1 mixture of 0.5% Rh/SZ and 0.5% Pt/SZ. (䊉, 䊊) NOx conversion; (䉱, 䉭) methane conversion. Reaction conditions: NO (filled symbol) or NO2 (open symbol) 150 ppm, CH4 2000 ppm, O2 10%, H2 O 9%, GHSV 15,000 h−1 .

Fig. 5. NO2 ratio in the effluent gas from NO–O2 (䊏), NO2 –O2 (䊐), NO–CH4 –O2 (䉬), NO2 –CH4 –O2 (䉫) reactions over: (a) 0.5% Pd/SZ and (b) 0.5% Rh/SZ. NO or NO2 150 ppm, CH4 0 or 2000 ppm, O2 10%, H2 O 9%, GHSV 15,000 h−1 .

H. Ohtsuka / Applied Catalysis B: Environmental 33 (2001) 325–333

below 400◦ C, while the same was true for Rh/SZ below 450◦ C. To elucidate this difference, the temperature dependence of the outlet NO2 ratio was compared over Pd/SZ and Rh/SZ. The results are shown in Fig. 5. On Pd/SZ, almost all the outlet NOx was NO in NO2 – CH4 –O2 even at 350◦ C. This means that Pd/SZ can catalyze the reaction of NO2 with methane at this temperature. Over Rh/SZ, no clear difference was observed between the outlet NO2 ratio under NO2 –CH4 – O2 and that under NO2 –O2 (not shown) at 350◦ C and even at 400◦ C, the outlet NO2 ratio was as high as 0.5. This indicates that Rh requires a temperature of 400◦ C or higher to catalyze the reaction on NO2 and methane, while the reaction occurs even at 350◦ C on Pd/SZ. This is the reason for the 50◦ C difference in the conversion curves of Pd/SZ and Rh/SZ. On the other hand, while Pd/SZ showed a methane conversion of 69% in the CH4 –O2 reaction at 500◦ C, Rh/SZ showed only 16% at 500◦ C and 40% even at 550◦ C, which leads to higher NOx conversion on Rh/SZ than on Pd/SZ at higher temperatures.

333

(Ag, Au). A combination of Pd or Rh with Pt or Ru gave high activity for the selective reduction of NOx by methane. References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14]

4. Conclusions

[15]

The selective catalytic reduction of NOx by methane on noble metal-loaded sulfated zirconia can be categorized into the following three groups: (i) low activity for NO oxidation to NO2 , and high activity for NO2 reduction to N2 (Pd, Rh); (ii) high activity for NO oxidation to NO2 , and low activity for NO2 reduction to N2 (Ru, Ir, Pt); (iii) low activity for both reactions

[16] [17] [18] [19]

Y. Li, J.N. Armor, Appl. Catal. B 1 (1992) L31. Y. Li, J.N. Armor, Appl. Catal. B 2 (1993) 239. Y. Nishizaka, M. Misono, Chem. Lett. (1993) 1295. Y. Nishizaka, M. Misono, Chem. Lett. (1994) 2237. E. Kikuchi, K. Yogo, Catal. Lett. 22 (1994) 73. Y. Li, J.N. Armor, J. Catal. 145 (1994) 1. A. Keshavaraja, X. She, M. Flytzani-Stephanopoulous, Appl. Catal. B 27 (2000) L1. J.N. Armor, Catal. Today 26 (1995) 147. H. Uchida, K. Yamaseki, I. Takahashi, Catal. Today 29 (1996) 99. F. Hoshi, K. Yamaseki, S. Satokawa, H. Uchida, in: Proceedings of the 80th CATSJ Meeting Abstract 3A10, 1997. M. Misono, Y. Nishizaka, M. Kawamoto, H. Kato, Stud. Surf. Sci. Catal. 105 (1997) 1501. M. Ogura, Y. Sugiura, M. Hayashi, E. Kikuchi, Catal. Lett. 42 (1996) 185. M. Ogura, S. Kage, M. Hayashi, H. Matsukata, E. Kikuchi, Appl. Catal. B 27 (2000) L213. M. Suzuki, J. Amano, M. Niwa, Microporous Mesoporous Mater. 21 (1998) 541. H. Ohtsuka, T. Tabata, T. Hirano, Appl. Catal. B 28 (2000) L73. H. Ohtsuka, T. Tabata, Appl. Catal. B 29 (2001) 177. H. Kato, C. Yokoyama, M. Misono, Catal. Today 45 (1998) 93. K.A. Bethke, C. Li, M.C. Kung, B. Yang, H.H. Kung, Catal. Lett. 31 (1995) 287. K. Nomura, K. Noro, Y. Nakamura, Y. Yazawa, H. Yoshida, A. Satsuma, T. Hattori, Catal. Lett. 53 (1998) 167.