Selective oxidation of ammonia to nitrogen on transition metal containing mixed metal oxides

Selective oxidation of ammonia to nitrogen on transition metal containing mixed metal oxides

Applied Catalysis B: Environmental 58 (2005) 235–244 www.elsevier.com/locate/apcatb Selective oxidation of ammonia to nitrogen on transition metal co...

616KB Sizes 3 Downloads 53 Views

Applied Catalysis B: Environmental 58 (2005) 235–244 www.elsevier.com/locate/apcatb

Selective oxidation of ammonia to nitrogen on transition metal containing mixed metal oxides Lucjan Chmielarz*, Piotr Kus´trowski, Alicja Rafalska-Łasocha, Roman Dziembaj Faculty of Chemistry, Jagiellonian University, Ingardena 3, 30-060 Krako´w, Poland Received 5 October 2004; received in revised form 13 December 2004; accepted 21 December 2004 Available online 21 January 2005

Abstract The selective catalytic oxidation of ammonia to nitrogen (NH3-SCO) has been studied over hydrotalcite derived mixed metal oxides containing Cu, Co, Fe or Ni. XRD, BET, NH3-TPD and TPR techniques were used for catalysts characterization. Results of NH3-SCO were compared with those of selective catalytic reduction of NO with NH3 (NO-SCR). Reaction mechanism was studied by temperatureprogrammed surface reaction (TPSR) and activity tests with a various contact time. Catalytic performance of the studied samples depends on both kind and loading of transition metals in the mixed metal oxide system. The Cu-containing samples have been found to be the most active catalysts of the NH3-SCO process. Transition metal loading strongly influences distribution of ammonia oxidation products. The highest selectivity to N2 was measured for the catalysts with the lowest transition metal content. # 2004 Elsevier B.V. All rights reserved. Keywords: Hydrotalcite derived mixed metal oxides; Transition metals; Selective oxidation of NH3

1. Introduction There are a lot of chemical processes that use ammonia as a reactant or produce ammonia as a by-product. All these processes are plagued with ammonia slip problem. Lowtemperature selective oxidation of ammonia with oxygen (NH3-SCO) to nitrogen and water is a potentially efficient method for removal of ammonia from oxygen-containing flue gases. Moreover, ammonia is used as an effective NOx (NO + NO2) reducer in power plants (4NH3 + 4NO + O2 ! 4N2 + 6H2O). The commercial catalysts of this process are based on the V2O5–TiO2 metal oxide system. In order to avoid ammonia slip the majority of DeNOx processes are carried out with an amount of ammonia, which is a bit lower to obtain the total conversion of NO (NH3/ NO = 0.9–0.95). It is possible to improve effectiveness of the DeNOx process by using stoichiometric or even excess quantity of ammonia. The unreacted ammonia could be

* Corresponding author. Tel.: +48 12 6632006; fax: +48 12 6340515. E-mail address: [email protected] (L. Chmielarz). 0926-3373/$ – see front matter # 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.apcatb.2004.12.009

oxidized by oxygen present in flue gases to nitrogen and water. The catalysts of the NH3-SCO process should operate in a relatively low-temperature range in order to reduce costs of additional heating of waste gases and should selectively convert ammonia to nitrogen. Different transition metal oxides have been intensively studied as potential catalysts of the NH3-SCO process [1–8]. There is a large number of metal oxides (e.g. CuO, Fe2O3, Co3O4, MnO2 [7–10]), which were found to be selective catalysts of ammonia oxidation. However, they are significantly less active than the noble metal containing catalysts [1,10–13]. It seems possible to enhance the activity of the transition metal containing catalysts by an increase of dispersion of transition metal cations. Magnesium–aluminum hydrotalcite-like materials containing transition metal cations are excellent precursors of high surface area mixed metal oxides. Such materials are characterized by a very high dispersion of transition metal in the Mg–Al–O oxide matrix. Thus, the calcined hydrotalcite-like materials seems to be promising catalysts for the selective oxidation of ammonia. This paper presents the studies of hydrotalcite derived

236

L. Chmielarz et al. / Applied Catalysis B: Environmental 58 (2005) 235–244

mixed metal oxides containing selected transition metals (Cu, Co, Fe and Ni) as catalysts of the NH3SCO process.

2. Experimental 2.1. Catalysts preparation Mg(II)Al(III), Ni(II)Mg(II)Al(III), Mg(II)Fe(III)Al(III), Cu(II)Mg(II)Al(III), Co(II)Mg(II)Al(III) and Cu(II)Co(II)Mg(II)Al(III) hydrotalcites were prepared by coprecipitation from aqueous solutions of suitable metal nitrates. Mixture of the metal nitrates was added to an aqueous solution of Na2CO3. The pH was maintained typically at 10.0  0.2 (with exception of Fe-containing hydrotalcite, which was synthesised at pH = 8.5  0.2) by dropwise NaOH addition. Precipitates were kept in suspension at 60 8C (or 40 8C for Fe-hydrotalcite) for 30 min under stirring, then were filtered, thoroughly washed with distilled water and dried at 120 8C. Finally, the prepared hydrotalcites were calcined at 600 8C for 16 h. The samples were kept in an exiccator in order to avoid reconstruction of the hydrotalcite structure. A detailed description of the catalysts preparation was presented in our previous papers [14–16]. 2.2. Catalysts characterization The X-ray diffraction (XRD) patterns of hydrotalcites and mixed metal oxides formed by their thermal decomposition were measured with a PW3710 Philips X’pert ˚ ). diffractometer using Cu Ka radiation (l = 1.54178 A The surface areas of calcined hydrotalcites were determined by the BET method. The measurements have been performed using ASAP 2010 (Micromeritics). Prior to the nitrogen adsorption at 196 8C the samples were outgassed under vacuum at 350 8C for 12 h. Surface acidity of calcined hydrotalcites was determined by temperature-programmed desorption of ammonia (NH3TPD) in a fixed-bed continuous flow microreactor system. Before NH3-TPD measurement the sample (50 mg) was outgassed in a flow of pure helium at 600 8C for 1 h. Subsequently, calcined hydrotalcite was cooled down to 70 8C and saturated in a flow of 1% NH3/He (20 ml/min) for about 30 min. Then the sample was purged in a helium flow until a constant baseline level was attained. Ammonia desorption was carried out in the temperature range of 70– 600 8C with a linear heating rate (b = 10 8C/min) in a flow of He (25 ml/min). Traces of H2O and O2 in pure helium (grade 5) used as the eluent gas were removed by appropriate traps (Alltech). The temperature in the catalyst bed was measured by a K-type thermocouple located in a quartz capillary immersed in the catalyst bed. The molecules desorbing from the samples were monitored on-line by a quadrupole mass spectrometer (VG QUARTZ) connected to the reactor outlet

by a heated line. The NH3-TPD spectra were obtained from the m/z = 16 mass-to-charge signal ratio. Calibration of QMS with commercial mixtures allowed to recalculate the detector signal into desorption rate. The TPR (temperature-programmed reduction) of the samples was carried out in the temperature range of 80– 950 8C in a fixed-bed flow microreactor (i.d., 4.5 mm; l, 240 mm). The hydrogen consumption was controlled on line by a quadrupole mass spectrometer (VG QUARTZ) connected to the reactor outlet by a heated line. Prior to the TPR experiment the sample (50 mg) was outgassed in a flow of helium. The TPR runs were carried out with a linear heating rate (b = 10 8C/min) in a flow of 5.22% H2/Ar (20 ml/min). 2.3. NH3-TPSR (temperature-programmed surface reaction) The outgassed sample was exposed to ammonia flow according to the procedure applied in the NH3-TPD experiments. In the next step the reactor was purged in a flow of helium and then the catalyst was heated up to about 600 8C with a linear increase of temperature (b = 10 8C/ min) in a flow of 5% O2/He (20 ml/min). 2.4. Catalytic tests The catalytic performance of calcined hydrotalcites in the selective oxidation of ammonia has been studied under atmospheric pressure in a fixed-bed flow reactor (i.d., 7 mm; l, 240 mm). The reactant concentrations were continuously measured using a quadrupole mass spectrometer (VG QUARTZ) connected to the reactor via a heated line. Prior to the reaction each sample of the catalyst (50 mg, particle diameter 125–180 mm) was outgassed in a flow of pure helium at 600 8C for 1 h. The composition of the gas mixture at the reactor inlet was [NH3] = 0.5%, [O2] = 2.5% and [He] = 97%. Total flow rate of the reaction mixture was 40 ml/min, while a space velocity was about 30,000 h1. The reaction was studied at temperatures ranging from 50 to 650 8C. The intensities of the mass lines corresponding to all reactants and possible products were measured at a given temperature at least for 30 min after the reaction had reached a steady state. The signal of the helium line served as the internal standard to compensate small fluctuations of the operating pressure. The sensitivity factors of analyzed lines were calibrated using commercial mixtures of gases. The possible changes in a molar flow caused by NH3 conversion were negligible in the diluted reaction mixtures. The differences between the reactor inlet and outlet molar flows of the reactants were used to determine conversion of the reactants. Calcined hydrotalcites were also tested as catalysts of the selective reduction of NO by NH3 in oxygen presence (NO-SCR). The experiments were performed in steady-state conditions at temperatures ranging from 50 to 400 8C

L. Chmielarz et al. / Applied Catalysis B: Environmental 58 (2005) 235–244

237

(intervals every 50 8C). The reaction mixture, supplied to reactor with a total flow rate of 40 ml/min, consisted of [NH3] = [NO] = 0.25%, [O2] = 2.5% and [He] = 97%. Prior to the catalytic test the sample (200 mg, particle diameter 125–180 mm) was outgassed in a flow of helium at 600 8C for 1 h. The measurements have been carried out under atmospheric pressure in a fixed-bed flow reactor (i.d., 7 mm; l, 240 mm). The reactant concentrations were continuously measured using a quadrupole mass spectrometer (VG QUARTZ) connected directly to the reactor outlet via a heated line.

3. Results and discussion 3.1. Physicochemical characterization The chemical composition of the obtained hydrotalcites, as well as the surface area of the calcined samples are presented in Table 1, while the examples of X-ray diffractograms of the dried and calcined hydrotalcites are shown in Fig. 1. The obtained results reveal that the prepared samples have a hexagonal structure with sharp symmetric peaks for the (0 0 3), (0 0 6), and (0 0 9) planes and broad asymmetric peaks for (0 1 5) and (0 1 8) planes, which are characteristic of hydrotalcites [14]. Besides reflections that confirm the presence of hydrotalcite phase, a very weak and broad peak (2u  288) due to existence of Al2O33H2O phase was detected for the HT, Cu-10, Co-10 and Ni-10 samples. Calcination of hydrotalcites at temperature 600 8C resulted in a disappearance of peaks characteristic for the hydrotalcite structure and appearance of new reflections at 2u about 368, 438 and 628, which are related to Mg(Al)O phase. Formations of any spinel phases were not detected. Surface acidity of the catalysts was measured by temperature-programmed desorption of ammonia (NH3TPD). Fig. 2 presents the examples of ammonia desorption curves, while the surface concentrations of chemisorbed NH3 are shown in Table 1. Ammonia desorption patterns are spread in the temperature range from 110 to about 420 8C

Fig. 1. XRD patterns of dried and calcined hydrotalcite materials (X: Al2O33H2O; Y: peak attributed to sample holder).

with a maximum centred at 200–210 8C. For the Fe-10 sample, the desorption pattern consists of two maxima. The first peak is present at 175 8C, while the second one, significantly less intensive, was detected at 400 8C. Calcined hydrotalcites were studied as catalysts of ammonia oxidation. Fig. 3A presents the results obtained for the reaction carried out in the absence of catalyst. Oxidation of NH3 in the gas phase starts at temperature about 250 8C and slowly increases with an increasing temperature. However, at temperature 650 8C ammonia conversion does not exceed 50%. NO is the main product of NH3 oxidation at temperature below 450 8C. Selectivity towards N2 increases gradually with a raising of reaction temperature. Small amounts of N2O are produced at

Table 1 Composition, surface area and concentration of chemisorbed ammonia Sample code

Composition

Atomic ratio

SBET of calcined hydrotalcite (m2/g)

Concentration of chemisorbed NH3 (mmol/m2)

HT Cu-5 Cu-10 Cu-20 Co-5 Co-10 Co-20 CuCo-5 CuCo-10 CuCo-20 Fe-10 Ni-10

Mg/Al Cu/Mg/Al Cu/Mg/Al Cu/Mg/Al Co/Mg/Al Co/Mg/Al Co/Mg/Al Cu/Co/Mg/Al Cu/Co/Mg/Al Cu/Co/Mg/Al Fe/Mg/Al Ni/Mg/Al

71/29 5/66/29 10/61/29 20/51/29 5/66/29 10/61/29 20/51/29 5/5/61/29 10/10/51/29 20/20/31/29 10/61/29 10/61/29

109 136 130 128 111 133 76 129 105 71 171 226

1.38 2.10 2.91 2.43 2.98 2.82 2.46 2.09 2.73 1.69 2.47 1.36

238

L. Chmielarz et al. / Applied Catalysis B: Environmental 58 (2005) 235–244

Fig. 2. NH3-TPD patterns of hydrotalcite derived mixed metal oxides.

temperature above 300 8C. According to Stephens and Pease [17] the homogenous non-catalytic oxidation of ammonia proceeds via chain radical mechanism. Significantly faster rate of ammonia oxidation was measured over the calcined Mg–Al hydrotalcite (Fig. 3B). The reaction starts at temperature 150 8C and a total NH3 conversion was measured at 650 8C. Nitrogen was found to be the main reaction product in the whole studied temperature range, however, the selectivity to nitrogen strongly depends on the reaction temperature. N2O was the second main product detected in the low-temperature range,

while at higher temperatures significant amounts of NO were formed. Fig. 4 presents the results of the catalytic test over the calcined hydrotalcites containing a different quantity of cobalt (5, 10 and 20 mol%). The cobalt content significantly influences distribution of ammonia oxidation products and only slightly affects conversion of NH3. Oxidation of ammonia over the Co-5 catalyst (Fig. 4A) starts at temperature 150 8C and at 500 8C the total conversion of NH3 was obtained. An increase in cobalt content to 10 mol% (Fig. 4B) and 20 mol% (Fig. 4C) decreases activity of the catalysts and additionally favours NO formation. Selectivity to N2O, which was found among ammonia oxidation products, decreases with temperature increase. Formation of very small amounts of NO2 was detected over the Cocontaining catalysts. In Fig. 5 are shown the results of catalytic tests performed over the Cu-containing samples. These catalysts have been found to be significantly more active in the NH3-SCO process than the Co-containing samples. Ammonia oxidation over Cu-5 (Fig. 5A) starts at about 150 8C and reaches the total conversion at temperature as low as 400 8C. An increase in copper loading to 10 mol% (Fig. 5B) and 20 mol% (Fig. 5C) only slightly influences activity of the studied catalysts. However, transition metal content strongly affects distribution of ammonia oxidation products. The highest selectivity to N2 was measured in the presence of the catalyst with the lowest copper loading (Cu-5). Significantly lower selectivity to nitrogen and higher selectivity to NO is observed for the samples Cu-10 and Cu-20. Distribution of ammonia oxidation products strongly depends on the reaction temperature. Nitrogen is a dominant product in the low-temperature region, while at higher temperatures selectivity towards NO increases. The amount of N2O

Fig. 3. Results of activity tests for reaction carried out without catalysts (A) and over calcined Mg–Al hydrotalcite (B).

L. Chmielarz et al. / Applied Catalysis B: Environmental 58 (2005) 235–244

239

Fig. 4. Results of catalytic tests for cobalt containing samples.

formed over the Cu-containing catalysts decreases with an increase in transition metal loading. Fig. 6 presents the results of the tests performed for the samples containing both copper and cobalt. Ammonia oxidation over this series of catalysts starts at 150 8C and the total NH3 conversion, depending on the transition metals loading, was achieved at about 450–500 8C. Nitrogen is a dominant product of ammonia oxidation over the CuCo-5 catalyst, however its selectivity decreases with an increase in the reaction temperature (Fig. 6A). High selectivity to N2 was observed also for the samples with higher transition metal loading at temperatures below 250 8C. However, at higher temperatures a decrease in selectivity to N2, due to formation of significant amount of NO, was found. Fig. 7 presents the results of the catalytic measurements performed over the iron (Fe-10) and nickel (Ni-10) containing catalysts. Both these samples have been found to be the significantly less active catalysts of NH3 oxidation than the Cu or/and Co-containing materials. Ammonia conversion over Fe-10 starts at about 150 8C but the total NH3 oxidation was obtained at temperature as high as 600 8C (Fig. 7A). Nitrogen was detected as the main reaction product in the whole studied temperature range, however its selectivity is not constant and strongly depends on the

reaction temperature. In Fig. 7B are shown the results of the catalytic test obtained for the Ni-10 sample. Conversion of ammonia starts at about 150 8C and increases rapidly to 500 8C reaching 95%. At higher temperatures NH3 conversion rises much slower. Nitrogen is a dominant reaction product in the whole studied temperature range. It must be however pointed out that temperature strongly influences distribution of the reaction products. The obtained results showed that both kind and content of transition metal introduced into Mg–Al–O matrix strongly influence the activity and selectivity of the catalysts of the NH3-SCO process. The activity of transition metals can be presented in the following order: Cu > Co > Ni > Fe The selectivity to N2 measured for the catalysts containing 10 mol% of transition metal can be ordered in the following way: Fe > Ni > Cu > Co It should be noticed that the order of activity is opposite to this found for selectivity towards N2 (with exception of copper). Golodets and Pyatnitskii [18] and Slavinskaya et al. [19] suggested that the strength of oxygen bonding in metal

Fig. 5. Results of catalytic tests for copper containing samples.

240

L. Chmielarz et al. / Applied Catalysis B: Environmental 58 (2005) 235–244

Fig. 6. Results of catalytic tests for copper and cobalt containing samples.

oxide determines distribution of ammonia oxidation products. The higher is the oxygen bonding strength, the higher is selectivity to nitrogen. The strength of oxygen bonding in transition metal oxides was determined by the temperature-programmed reduction method. The results of these measurements are presented in Fig. 8. Reduction of cobalt oxide in Mg–Al–O matrix starts at temperature 397 8C. Nickel oxide present in the Ni-10 sample is reduced from 495 8C, while reduction of iron oxide begins at 523 8C. Thus, there is a correlation between reducibility of transition metal oxide and distribution of ammonia oxidation products. Copper oxide (Fig. 8B) is reduced at significantly lower temperatures than mentioned above metal oxides (Fig. 8A). Therefore, the lowest selectivity to nitrogen could be predicted in the case of the Cu-containing catalysts. However, selectivities to N2 measured for the Cu-containing samples are higher compared to those observed for the Cocontaining catalysts. Selectivity to N2 depends on copper

loading in the catalysts. The higher is the copper content, the lower selectivity to N2 was observed. An increase in transition metal loading probably favors formation of surface metal oxide clusters, which have a lower oxygen bonding strength as compared with isolated cations. This supposition is confirmed by results of H2-TPR studies of the Cu-containing catalysts (Fig. 8B). This same effect is probably responsible for decrease of selectivity to N2 and activity of the samples with higher loading of Co as well as Cu and Co. The lack of correlation between reducibility of the Cu-containing samples and measured selectivity to N2 over these catalysts could be explained by conversion of nitrogen oxides by unreacted ammonia into nitrogen. The Cu-containing samples have been found to be considerably more effective catalysts of the DeNOx process than the other studied materials. The results of the DeNOx studies over the Cu- and Co-containing catalysts are presented in Fig. 9. It should be mentioned that the catalytic activity of the Fe- and

Fig. 7. Results of catalytic tests for iron (A) and nickel (B) containing samples.

L. Chmielarz et al. / Applied Catalysis B: Environmental 58 (2005) 235–244

241

Ni-containing samples was also significantly lower than that found for the copper catalysts (results not shown). Probably, NO formed by oxidation of ammonia over the Cu-containing catalysts is reduced by unreacted ammonia. The proposed mechanism consists of the following steps: (1) Oxidation part of ammonia to NO: 4NH3 þ 5O2 ! 4NO þ 6H2 O

(1)

(2) Reduction of NO by unreacted ammonia (NO-SCR): 4NO þ 4NH3 þ O2 ! 4N2 þ 6H2 O

(2)

4NO þ 4NH3 þ 3O2 ! 4N2 O þ 6H2 O

(3)

Such mechanism was proposed for a large number of catalytic systems (e.g. CuO/Al2O3 [20], Ag (powder) [21], Ag/ Al2O3, Ag/SiO2 [22], Pt [1,23], Pt/Al2O3, Rh/Al2O3, Pd/ Al2O3, Pt-ZSM-5, Pd-ZSM-5, Rh-ZSM-5 [24]). This mechanism was confirmed by studies of Cavani and Trifiro [25] as well as Bosch and Janssen [26], who showed high activity and selectivity of the commercial DeNOx catalyst based on the V2O5/TiO2 system in oxidation of ammonia to nitrogen. Similar correlation was found for zeolites and pillared montmorillonites modified with iron [27,28]. It should be noticed that reduction of nitrogen oxides by unreacted ammonia over the Co-, Ni- and Fe-containing samples is possible only at higher temperatures due to low activity of these catalysts in the DeNOx process in the lowtemperature region. However, apart from NO and N2O formation of N2 was detected at low temperatures. Thus, it seems possible that all these compounds are primary products of the low-temperature ammonia oxidation. There is numerous papers [19,29,30] suggesting formation of NHx

Fig. 8. Results of TPR measurements for the Fe-10, Ni-10, Co-10 (A) and the Cu-containing samples (B).

Table 2 Results of catalytic tests performed with various contact times, t T (8C)

t = 0.12 s NH3 conversion

t = 0.06 s Selectivity (%)

NH3 conversion

N2

NO

N2O

NO2

t = 0.03 s Selectivity (%)

NH3 conversion

N2

NO

N2O

NO2

Cu-10 200 250 300 350 400 450 500 550

24.0 65.1 88.4 93.8 96.9 100.0 100.0 100.0

92.0 85.1 81.4 77.5 68.8 63.5 57.5 54.5

9.0 12.7 15.2 18.0 27.5 33.6 41.4 45.0

1.0 2.2 3.4 4.5 3.7 2.9 1.1 0.5

0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0

11.6 39.2 60.3 84.7 91.1 95.6 98.5 100.0

100.0 95.1 92.0 91.1 90.3 88.2 83.8 74.4

0.0 4.9 8.0 8.3 5.8 6.8 10.7 20.7

0.0 0.0 0.0 0.0 3.5 4.5 5.3 5.1

0.0 0.0 0.0 0.6 0.4 0.5 0.2 0.2

Co-10 200 250 300 350 400 450 500 550

10.7 22.9 33.6 48.9 70.7 89.1 95.5 100.0

39.4 40.0 41.3 42.0 43.1 45.2 46.0 47.9.

49.5 50.0 51.5 53.1 51.7 50.1 49.1 46.8

10.9 10.0 6.3 3.9 4.2 3.6 3.5 3.7

0.0 0.0 0.9 1.0 1.0 1.1 1.4 1.6

5.2 10.5 18.1 30.6 48.8 70.5 88.4 97.3

41.1 42.2 58.0 67.9 67.6 71.5 67.7 57.5

50.7 47.2 37.0 27.7 29.1 25.4 29.4 40.1

8.2 6.6 5.0 4.4 3.3 3.1 2.9 2.4

0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0

Selectivity (%) N2

NO

N2O

NO2

5.2 20.1 33.3 58.3 81.0 89.3 95.1 99.0

100.0 100.0 99.3 98.4 93.9 93.9 88.2 76.5

0.0 0.0 0.7 1.0 1.4 3.8 8.7 19.7

0.0 0.0 0.0 0.0 0.4 1.9 2.9 3.8

0.0 0.0 0.0 0.6 0.7 0.4 0.2 0.0

2.9 5.9 10.4 16.2 26.6 42.4 70.2 86.6

42.1 43.2 56.7 60.0 69.8 74.5 73.6 61.9

49.1 47.9 36.3 34.0 23.8 19.7 22.3 31.8

8.8 7.9 7.0 6.0 6.4 5.8 4.1 6.3

0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0

242

L. Chmielarz et al. / Applied Catalysis B: Environmental 58 (2005) 235–244

Fig. 9. Results of NH3-SCR measurements over the Cu- (A) and Co-containing (B) catalysts.

species through oxydehydrogenation of ammonia as the first reaction step. In the next stage, successive abstraction of hydrogen atoms results in the formation of Nads species, which interact with Oads with production of NO molecules. Formation of N2 molecules is possible by interaction of two Nads species. On the other hand, Il’chenko and Golodets [7] suggested transformation of chemisorbed ammonia into surface imide (NH) and nitrosyl (HNO) species. Interaction of NH with HNO results in formation of nitrogen and water molecules, while N2O is produced by interaction of two surface nitrosyl species. It should be noticed that according to both proposed mechanisms lower strength of oxygen bounding in transition metal oxides should favor formation of nitrogen oxides, while stronger bonding of oxygen in oxides should lead to higher selectivity to nitrogen. In Table 2 are presented the results of the catalytic tests for the Cu-10 and Co-10 samples performed with a various contact time. In the case of the Cu-containing catalyst a decrease in a contact time resulted in an increase of selectivity to N2 and a decrease of NO and N2O selectivities. Therefore, it seems that nitrogen is a primary product of ammonia oxidation, which in a subsequent step is oxidize to NO. However, oxidation of nitrogen to nitrogen oxides in the studied temperature range is thermodynamically forbidden. Thus, this hypothesis has to be rejected. The other possible explanation of these results is reduction of NO to N2 by

ammonia. In the case of measurements carried out with a shorter contact time the amount of unreacted ammonia is higher in comparison to the experiments performed with a longer contact time. Therefore, the excess of unreacted ammonia favors conversion of NO to N2. In the case of experiment performed with t = 0.12 s formation of nitrous oxide starts at low temperatures. Moreover, an increase in a contact time results in a disappearance of N2O evolution in the low-temperature region. It seems that nitrous oxide is formed mainly as a side product of DeNOx process. An increase in a contact time resulted in an appearance of small amount of NO2 over the Cu-10 catalyst. Olofsson et al. [31] proposed for the Pt/CuO/Al2O3 catalyst formation of NO2 as a product of ammonia oxidation, which interact with the surface NHx species with production of N2. The very low selectivity to NO2 measured over the Cu-containing catalysts (Table 2) suggests that mechanism proposed by Olofsson is not a main route of N2 formation. The distribution of ammonia oxidation products over the Co-10 catalyst is nearly independent on a contact time used at temperatures below 300 8C (Table 2). Thus, is seems that N2, NO and N2O are primary products of ammonia oxidation, which are formed parallel. At higher temperatures a decrease in a contact time caused an increase of the N2 selectivity and a drop of selectivity to NO. Therefore, the most probably NO is partially reduced by unreacted ammonia in that temperature region.

L. Chmielarz et al. / Applied Catalysis B: Environmental 58 (2005) 235–244

243

Fig. 10. Results of temperature-programmed surface reaction (TPSR) for Cu-10 (A), Co-10 (B) and CuCo-10 (C).

Reactivity of ammonia chemisorbed on the catalyst with oxygen was studied by temperature-programmed surface reaction (TPSR). The results of these studies performed for the Cu-10, Co-10 and CuCo-10 catalysts are presented in Fig. 10. Apart from desorbing ammonia also products of its oxidation (N2, NO and N2O) were detected. The evolution of N2 takes place in the temperature range, which is very similar to that measured for ammonia desorption. However, it should be noticed that amount of formed nitrogen over the Co-10 catalysts is higher than that measured for the Cucontaining samples. Thus, the distribution of ammonia oxidation products is different from that found in the catalytic test, however, proves the hypothesis that higher oxygen bonding strength in metal oxide favors formation of nitrogen as a product of ammonia oxidation. Differences in product distribution observed in catalytic tests and TPSR measurements could be explained by the various ratios of NH3/O2 reactants in both these experiments. In the case of TPSR amount of ammonia is limited to quantity chemisorbed on the catalyst surface. Therefore, the reduction of NO by NH3 is limited due to ammonia deficit. Evolution of NO was detected at temperatures significantly higher than ammonia desorption. The results of NH3-TPD experiments have shown that ammonia cannot be chemisorbed on the catalysts surface at temperature higher than 450 8C (Fig. 2). Thus, it seems possible that NO was oxidized to thermally stable nitrates. The high temperature peaks present in the spectra of NO evolution are probably related to the thermal decomposition of these nitrates. Formation of surface NO3 anions during NO sorption over calcined hydrotalcites was proved by FT-IR and NO-TPD methods in our previous studies [14]. Such thermally stable surface nitrates can block surface sites and therefore decrease the activity of the catalyst. It should be noticed that amounts of N2O produced in TPRS measurements are significantly higher than it could

be predicted from the results of the catalytic tests. It seems possible that under conditions of the catalytic tests (higher concentration of NH3) nitrous oxide is reduced by ammonia: 3N2 O þ 2NH3 ! 4N2 þ 3H2 O

(4)

4. Conclusions Metal mixed oxides, obtained by calcination of hydrotalcite materials, have been found to be active catalysts of ammonia oxidation. Their activity strongly depends on the kind of transition metal introduced to the mixed oxides system. The highest activity was observed for the copper containing catalysts. The samples modified with cobalt were less active, while the catalysts modified with nickel or iron showed poor catalytic performance. Oxidation of ammonia is possible also in the absence of catalyst, however, in this case ammonia conversion did not exceed 50% at temperatures below 650 8C. Distribution of ammonia oxidation products strongly depends on transition metal loading. The highest selectivities towards nitrogen were found for the catalysts with the lowest transition metal content. An increase in transition metal loading resulted in a decrease of the selectivity to nitrogen and an increase of the selectivity to nitric oxide. The reaction temperature is also a very important factor affecting distribution of ammonia oxidation products. Nitrogen is a dominant reaction product in the lower temperatures, while at higher temperatures the selectivity to NO increases. The study of reaction mechanism showed that distribution of ammonia oxidation products is dependent on reducibility of transition metal oxide dispersed in Mg–Al–O matrix. In the low-temperature range parallel scheme of product formation is proposed, while at higher temperatures NO is reduced by ammonia

244

L. Chmielarz et al. / Applied Catalysis B: Environmental 58 (2005) 235–244

with production of N2 and N2O. In the case of the Cucontaining samples, which have been found to be very active catalyst of DeNOx process, NO is reduced by unreacted ammonia also in the low-temperature region. References [1] N.I. Il’chenko, Russ. Chem. Rev. 45 (1976) 1119. [2] P. Chen, R.T. Yang, Appl. Catal. 80 (1992) 135. [3] E.T.C. Vogt, A. Boot, J.W. Geus, F.J.J.G. Janssen, J. Catal. 114 (1988) 313. [4] J.J.P. Biermann, F.J.J.G. Janssen, J.W. Geus, J. Mol. Catal. 60 (1990) 229. [5] M. de Boer, A.J. van Dillen, J.W. Geus, Catal. Lett. 11 (1991) 227. [6] G. Ramis, L. Yi, G. Busca, Catal. Today 28 (1996) 373. [7] N. Il’chenko, G.I. Golodets, J. Catal. 39 (1975) 57. [8] O.N. Sil’chenkova, V.N. Korchak, V.A. Matyshak, Kinet. Catal. 43 (2002) 363. [9] G.I. Golodets, Studies in Surface Science and Catalysis, Elsevier, Amsterdam, 1983. [10] A. Wollner, E. Lange, H. Knozinger, Appl. Catal. A 94 (1993) 181. [11] P.D. Sobczyk, E.J.M. Hensen, A.M. de Jong, A.S. van Santen, Topics Catal. 23 (2003) 109. [12] R.Q. Long, R.T. Yang, Catal. Lett. 78 (2002) 353. [13] J.J. Ostermaier, J.R. Katzer, W.H. Manoque, J. Catal. 41 (1976) 227.

[14] L. Chmielarz, P. Kus´trowski, A. Rafalska-Łasocha, D. Majda, R. Dziembaj, Appl. Catal. B 35 (2002) 195. [15] L. Chmielarz, P. Kus´trowski, A. Rafalska-Łasocha, R. Dziembaj, Thermochim. Acta 395 (2003) 225. [16] P. Kus´trowski, A. Rafalska-Łasocha, D. Tomaszewska, R. Dziembaj, Solid State Ionics 141–142 (2001) 237. [17] E.R. Stephens, R.N. Pease, J. Am. Chem. Soc. 72 (1933) 1187. [18] G.J. Golodets, Yu.I. Pyatnitskii, Katal. Katal. 4 (1968) 25. [19] E.M. Slavinskaya, S.A. Veniaminov, P. Notte, S. Ivanova, A.I. Borowin, Yu.A. Cheslakov, I.A. Polukhina, A.S. Noskom, J. Catal. 222 (2004) 129. [20] T. Curtin, F. O’Regan, C. Deconinck, N. Knutte, B.H. Hodnett, Catal. Today 55 (2000) 189. [21] L. Gang, J. van Grondelle, B.G. Anderson, R.A. van Santen, J. Catal. 199 (2001) 107. [22] L. Gang, B.G. Anderson, J. van Grondelle, R.A. van Santen, Appl. Catal. B 40 (2003) 101. [23] N. Il’chenko, G.I. Golodets, I.M. Avilova, Kinet. Catal. 16 (1975) 1455. [24] Y. Li, J.N. Armor, Appl. Catal. B 13 (1997) 131. [25] F. Cavani, F. Trifiro, Catal. Today 4 (1989) 253. [26] H. Bosch, F. Janssen, Catal. Today 2 (1988) 369. [27] R.Q. Long, R.T. Yang, J. Catal. 201 (2001) 145. [28] R.Q. Long, M.T. Chang, R.T. Yang, Appl. Catal. B 33 (2001) 97. [29] G.S. Selwyn, G.T. Fujimoto, M.C. Lin, J. Phys. Chem. 86 (1982) 760. [30] W.D. Mieher, W. Ho, Surf. Sci. 322 (1995) 151. [31] G. Olofsson, A. Hinz, A. Andersson, Chem. Eng. Sci. 59 (2004) 4113.