- Email: [email protected]
TiO2 activity in the NO+CO reaction
Catalysis Today 93–95 (2004) 811–818
Effect of ZrO2 addition on CuO/TiO2 activity in the NO + CO reaction Xiaoyuan Jiang a,∗ , Guanghui Ding a , Liping Lou b , Yingxu Chen b , Xiaoming Zheng a b
a Faculty of Science, Institute of Catalysis, Zhejiang University, Hangzhou 310028, PR China Faculty of Environment and Resource, Institute of Environmental Engineering, Zhejiang University, Hangzhou 310029, PR China
Available online 29 July 2004
Abstract The activities of CuO/TiO2 and ZrO2 -modified CuO/TiO2 in the NO reduction by CO were examined with a microreactor-gas chromatography (GC) NO + CO reaction system, and the catalysts were characterized by physical adsorption, TG-DTA, XRD and NO-TPD techniques. The results showed that the NO conversion temperatures of 6% CuO/TiO2 were 280 ◦ C at T50% and 450 ◦ C at T99% , compared to 325 ◦ C at T99% for 6% CuO–10% ZrO2 /TiO2 . The 6% CuO–10% ZrO2 /TiO2 calcined at 500 and 750 ◦ C had similar catalytic activities (T99% at 300 ◦ C) in H2 atmosphere. The nitrogen adsorption–desorption isotherm and pore-size distribution curve of TiO2 represented type IV of the BDDT system and a typical mesoporous sample, respectively. After ZrO2 addition, the diffraction peak of ZrO2 was not detected and the diffraction peak intensity of CuO did not increase either, likely due to the strong interaction between CuO and ZrO2 that enhanced the catalytic activity and stability. In the H2 atmosphere, CuO produced dispersed Cu0 species. Four desorption species (NO, N2 O, N2 and O2 ) were detected during the thermal desorption of NO by CuO/TiO2 and CuO–ZrO2 /TiO2 treated in both air and H2 . Addition of ZrO2 onto CuO/TiO2 shifted the NO dissociation peaks towards low temperature, indicating that the activity of NO decomposition was higher by CuO–ZrO2 /TiO2 than by CuO/TiO2 . The NO + CO reaction formed intermediary product N2 O at low temperature but formed N2 at high temperature. In addition, the peak temperature of NO desorption was related to the catalytic activity under both air and H2 , and the dissociation of NO on catalysts surface was a rate-determining step in NO + CO reaction. © 2004 Elsevier B.V. All rights reserved. Keywords: CuO/TiO2 catalysts; ZrO2 ; NO + CO reaction; NO-TPD
1. Introduction Nitric oxide (NOx ) is the major toxic pollutant in urban atmosphere, and catalytic reduction of NO with CO is one of the most important reactions for NOx controlling. At present, noble metal catalysts are mainly used for DeNOx . Due to high prices and scarcity of noble metals, it is necessary to find an alternative catalytic component to reduce using or even replace the noble metal. Therefore, studies of non-noble metal catalysts have attracted much attention for the last 10 years. TiO2 is good at resisting water and trace SO2 along with high N2 selectivity in the course of DeNOx . By comparison, ZrO2 has better thermal stability, unique surface acidity and alkalescence and redox properties. Meanwhile, ZrO2 is a semiconductor, which means it can easily produce cavity and has strong interaction with active component. So it is possible to improve the perfor-
∗
Corresponding author. E-mail address: [email protected] (X. Jiang).
0920-5861/$ – see front matter © 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.cattod.2004.06.074
mance of catalysts by changing the composition of carrier. However, there have been few studies of the Ti–Zr complex oxide supported CuO for NO + CO reaction. In 1989, Iwamoto et al. [1,2] found that Cu-ZSM5 catalyst had much higher catalytic activity for NO dissociation. From then on, catalytic reduction of NO has been extensively studied, and variety of transition and noble metal catalysts were prepared [3,4]. Recently, the use of TiO2 as catalyst’s carrier has attracted much attention and TiO2 -based catalysts are widely applied in NO reduction [5–8]. The TiO2 -based catalysts overcome the shortcoming of TiO2 and SOx interaction that produces sulphate and causes structural collapse of the carrier, particle gathering of active components and reduction of active surfaces. As a result, the catalytic reduction of NO can be achieved in a sulphur-containing atmosphere, and obtain an excellent NO conversion and N2 selectivity. Various physicochemical methods are also used to investigate the state of copper ion in CuO/TiO2 catalysts, and inconsistent results have been reported. Komova et al. [7] found that TiO2 -supported CuO could exist as Cu2+ ions, bulk CuO, chain stabilized Cu2+ ions, and two different
812
X. Jiang et al. / Catalysis Today 93–95 (2004) 811–818
oxide clusters with a structure similar to CuO. In contrast, Xu et al. [8] reported two Cu2+ ions with significantly different reducibility on the anatase surface—CuO phase and highly dispersed Cu2+ , similar to CuO/␥-Al2 O3 . Larsson and co-workers [9,10] suggested that low temperature peak in Cu3–Ti of the H2 -temperature programmed reduction (TPR) profiles was monomeric copper containing Cu+ , and that in Cu12–Ti was polymeric copper containing Cu2+ . They also found that catalysts using TiO2 as carrier displayed a high catalysis in CO oxidation, and that CuO catalyst had much higher activity than MoOx , FeOx and CoOx catalysts and its activity at low temperature was even superior to Pt/␥-Al2 O3 catalyst. In this study, CuO/TiO2 and CuO–ZrO2 /TiO2 catalysts with different amount of CuO loadings were prepared by impregnating the TiO2 carrier, and these catalysts in the NO + CO reaction were examined by means of BET, TG-DTA, XRD and NO-temperature programmed desorption (TPD).
treated in 150 ml min−1 of dry pure N2 . The TG-DTA spectra were obtained with temperatures ramped at a linear heating rate of 20 ◦ C min−1 from 50 to 800 ◦ C. X-ray diffraction (XRD) data were obtained at 25 ◦ C using a horizontal Rigaku B/Max lllB powder diffractometer with Cu K␣ radiation and a power of 40 mA × 30 mA. To measure the NO-TPD, 250 mg of fresh catalysts were loaded onto a quartz reactor (Φ = 5 mm) and reduced in H2 atmosphere at 500 ◦ C for 1 h, and followed by two treatments. In treatment 1, catalysts were heated in He at 600 ◦ C for 1 h, and then cooled to 40 ◦ C in a flow of He and exposed to a 10% NO–He mixture gas. Excessive NO was removed, and the catalysts were kept in He flow until no significant amount of adsorbates could be detected. The catalysts were then ramped at 600 ◦ C at a linear heating rate of 20 ◦ C min−1 in He flow. The effluent gases were analysed with a mass spectrometer. In treatment 2, after the catalysts were heated in He at 500 ◦ C for 1 h, the procedures followed were the same as those in treatment 1.
2. Experimental
2.4. Measurements of catalytic activity in NO + CO reaction
2.1. TiO2 preparation TiO2 was prepared by sol–gel method from TiCl4 with ammonia solution as precipitating agent and ethanol as dispersing agent. A 25 ml TiCl4 was dissolved in 20 ml distilled water in an ice-water bath. The titanium solution was then slowly mixed with 30 ml distilled water and 20 ml ethanol, and ammonia was added dropwise until pH = 9. During the ammonic addition, an intensive precipitation occurred. After the solvent was evaporated at 80 ◦ C for 24 h, the precipitates were dried at 300 ◦ C for 2 h to remove NH4 Cl, and then calcined in an air stream of 450 ◦ C for 4 h. 2.2. Catalyst preparation The CuO/TiO2 and CuO–ZrO2 /TiO2 catalysts were prepared by the impregnation method using Cu(NO3 )2 and Zr(NO3 )4 aqueous solutions of desired concentrations. After these catalysts being rested for 24 h, catalysts were dried at 120 ◦ C for 2 h, followed the catalysts were put into a muffle furnace by calcinations in an air stream of 500 ◦ C for 2 h (β = 15 K min−1 ). They were denoted as w% CuO/TiO2 and w% CuO–w % ZrO2 /TiO2 , where w and w are the amount of CuO and ZrO2 loading, respectively. 2.3. Measurements of catalytic characters The BET surface area of catalysts was determined by N2 adsorption at 77 K using a Coulter OMNISORP-100 instrument. The TG-DTA experiment was carried out using PE-TGA7 thermo-gravimetric analyzer and DTA/9050311 high temperature differential thermal analyzer. Ten mg of fresh catalysts were taken and measured in air atmosphere, and then
Catalytic activity was determined under the steady state in a fixed-bed quartz reactor (6 mm). The particle size of catalysts was 20–40 mesh, and 120 mg of the catalysts were used. The reaction gas (i.e. feed stream) consisted of a fixed composition of 6.0% NO, 6.0% CO and 88% He (v/v) as a dilute. The catalysts were treated using either H2 or air at 500 ◦ C for 1 h. After cooled to room temperature, they were allowed to react with the mixed gas. The reactions were operated at different temperatures with a space velocity of 5000 h−1 . Two columns and thermal conduction detectors were used for analyzing the catalytic activity. Column A was packed with 13× molecular sieve for separating N2 , NO and CO, and Column B was packed with Paropak Q for separating N2 O and CO2 . 3. Results and discussion 3.1. Textural and structural properties of TiO2 Fig. 1 shows the nitrogen adsorption–desorption isotherm and pore-size distribution curve of TiO2 . The TiO2 was prepared from TiCl4 with a N2 isotherm corresponding to type IV of the BDDT system and was typical of a mesoporous sample. The specific surface area and total pore volume of TiO2 were 64.5 m2 /g and 0.441 ml/g, and average pore diameter and stochastic pore diameter were 27.3 nm and about 10 nm, respectively. The DTA curves of TiO2 showed an absorption heat peak at around 120 ◦ C, where a weight-lose occurred in its TG curve (Fig. 2). This peak could be due to the evaporation and removal of adsorbed H2 O and organic substance. Corresponding to an obvious weight-lose step of TiO2 around
X. Jiang et al. / Catalysis Today 93–95 (2004) 811–818
813
Fig. 4. Activity of 6% CuO/TiO2 with different calcination temperature for NO + CO reaction.
Fig. 1. Adsorption–desorption isotherms (a) and pore-size distributions (b) of TiO2 .
Calcining heat also affected the catalytic activities. For example, 6% CuO/TiO2 calcined at 750 ◦ C had higher activity than those calcined at 500 or 850 ◦ C (Fig. 4). The catalyst calcined at 500 ◦ C had higher activity than that calcined at 850 ◦ C at low reaction temperatures, compared to an opposite finding at high reaction temperatures. When calcined at 850 ◦ C, the catalytic activity began to decrease at the temperatures of above 500 ◦ C, probably due to the SMSI action [11]. 3.3. Activities of CuO–ZrO2 /TiO2 in NO + CO reaction
Fig. 2. TG-DTA curves of TiO2 sample.
320 ◦ C, an absorption heat peak was also observed as a result of NH4 Cl decomposition. 3.2. Activities of CuO/TiO2 in NO+CO reaction As shown in Fig. 3, the activities of CuO/TiO2 in NO + CO reaction increased with the increase in reaction temperature, ranging from low temperatures around 150 ◦ C to high temperatures above 300 ◦ C. The catalysts calcined at 500 ◦ C had very low activity at a CuO loading of 1–3%. The NO conversion temperature (T50% ) was 450 ◦ C for 6% CuO/TiO2 and 375 ◦ C for 18% CuO/TiO2 .
Fig. 3. Activity of CuO/TiO2 (500 ◦ C) with different CuO loading for NO + CO reaction.
As shown in Fig. 5, the NO conversion temperatures of 6% CuO/TiO2 were 280 ◦ C at T50% and 450 ◦ C at T99% . By comparison, both 6% CuO–5% ZrO2 /TiO2 and 12% CuO–5% ZrO2 /TiO2 had a NO conversion temperature of 350 ◦ C at T99% . We assume that ZrO2 addition increased the catalytic activity in NO + CO reaction, whereas CuO addition made little contribution to the catalytic activity even if the loading of CuO increased to 12%. It was also found that both 6% CuO–10% ZrO2 /TiO2 and 6% CuO–15% ZrO2 /TiO2 had a NO conversion temperature of 325 ◦ C at T99% (Fig. 6), suggesting that excessive addition of ZrO2 would not further increase the catalytic activity. At low reaction temperature, 6% CuO–10% ZrO2 /TiO2 calcined at 500 ◦ C had higher activity in NO + CO reaction than that calcined at 750 or 850 ◦ C, and similar catalytic activities were observed when calcined at 750 and 850 ◦ C (Fig. 7). However, the catalysts calcined at 500, 750 or 850 ◦ C had almost the same conversion temperature (T99% ), indicating that 6% CuO–10% ZrO2 /TiO2 calcined at 750 ◦ C had better heat stability.
Fig. 5. Effect of CuO loading on activity of CuO–5% ZrO2 /TiO2 in NO + CO reaction.
814
X. Jiang et al. / Catalysis Today 93–95 (2004) 811–818
Fig. 6. Effect of ZrO2 addition on activity of CuO/TiO2 in NO + CO reaction.
Fig. 7. Effect of calcination temperature on activity of 6% CuO–10% ZrO2 /TiO2 in NO + CO reaction.
3.4. Effects of reaction temperature on NO, CO conversion and N2 selectivity Table 1 shows the effects of reaction temperature on NO, CO conversion and N2 selectivity of 6% CuO/TiO2 and 6% CuO–10% ZrO2 /TiO2 catalysts. It was found that NO and CO concentrations decreased but N2 and CO2 concentrations increased with the increase in reaction temperature. At 150 ◦ C, the amount of N2 , CO2 and N2 O and the conversions of NO and CO increased markedly. A continuous increase in N2 and CO2 but a decrease in N2 O was observed at 200 ◦ C. At 300 ◦ C, NO and CO conversions were 66.7, 70.6% by 6% CuO/TiO2 , and 98.9, 96.9% by 6% CuO–10% ZrO2 /TiO2 , respectively. At 350 ◦ C, NO and CO were almost completely converted by 6% CuO–10% ZrO2 /TiO2 compared with 91.3 and 85.6% by 6% CuO/TiO2 , respectively. These results indicate that N2 O was formed during the initial stage of the reaction at lower temperature and increased with the increase in temperature until a maximal amount of Table 1 Effects of reaction temperature on NO, CO conversion and N2 selectivities for NO + CO reaction TR (◦ C)
150 200 250 300 350 400 450 500
CNO (%)
Fig. 8. Activities of CuO/TiO2 catalysts with different CuO loadings in H2 atmosphere.
N2 O was reached. Further increase in temperature, however, caused a gradual decrease in N2 O. In contrast, the amount of N2 increased continuously with increase in temperature. It can be suggested that in the process of NO reduction with CO, the reactant molecules (NO and CO) were firstly adsorbed on the adsorption sites of catalyst surface in forms of NO(a) and CO(a), and the adsorbed NO(a) then dissociated into N(a) and O(a), followed by reactions of N(a) with NO(a) and CO(a) with O(a) to produce N2 (ads) and CO2 , and N2 formation by two N(a). This study also showed that the rate of NO(a) dissociation on the catalyst surface was slower at low temperature and its reaction with N(a) formed N2 O(ads). As the reaction temperature increased, the rate of NO(a) dissociation became faster, causing a decrease in NO(a) concentration. As a result, high temperature was not beneficial to N2 O(a) formation. Therefore, the catalytic activities in the NO + CO reaction were not only related to the dispersion state of active components on TiO2 but also to the rate of NO adsorption and dissociation on the catalyst surfaces. 3.5. Activities of CuO/TiO2 catalysts in H2 atmosphere Fig. 8 shows the activities of CuO/TiO2 with different CuO loading in H2 atmosphere for NO + CO reaction. The sample treated in H2 possessed higher catalytic activity than that treated in air. The catalytic activity increased with the increase in temperature and the CuO loading from 1 to 18%. The catalysts treated in H2 had higher activity than that treated in air. For example, 6% CuO/TiO2 showed 50% NO conversation at reaction temperature of 450 ◦ C in air compared to 150 ◦ C in H2 . Calcining temperature also affected the catalytic activity in NO + CO reaction. As shown in Fig. 9, the NO conver-
SN2 (%)
CCO (%)
Cu–Ti
Cu–Zr–Ti
Cu–Ti
Cu–Zr–Ti
Cu–Ti
Cu–Zr–Ti
30.8 35.3 40.9 66.7 91.3 91.8 99.0 100
41.0 42.9 70.5 98.9 100 100 100 100
21.9 21.9 23.0 70.6 85.5 85.2 90.2 97.3
23.9 28.9 64.0 96.9 99.2 100 100 100
13.7 10.1 15.9 65.7 78.4 100 100 100
0 9.26 43.9 91.9 100 100 100 100
Fig. 9. Catalytic activity of 6% CuO/TiO2 catalysts calcined at different temperatures in H2 atmosphere.
X. Jiang et al. / Catalysis Today 93–95 (2004) 811–818
815
Fig. 10. Effect of ZrO2 addition on activity of CuO/TiO2 in H2 atmosphere.
sion of 6% CuO/TiO2 decreased with the increase in calcining temperature. The reaction temperature of 99% NO conversion was 300 ◦ C and 375 ◦ C in H2 atmosphere when the calcining temperature was 500 and 750 ◦ C, respectively. At reaction temperature of 500 ◦ C, the catalyst calcined at 850 ◦ C only had 74.84% NO conversion. Fig. 10 displays the effects of H2 treatment and ZrO2 addition on catalytic activity of the CuO/TiO2 catalysts in NO + CO reaction. After treatment in H2 , the catalytic activity increased significantly, especially for 6% CuO/TiO2 . The order of the catalytic activities was, 6% CuO/TiO2 > 12% CuO–5% ZrO2 /TiO2 > 6% CuO–5% ZrO2 /TiO2 . The activity of 6% CuO–10% ZrO2 /TiO2 calcined at 500 ◦ C was similar to that calcined at 750 ◦ C, but decreased dramatically at calcining temperature of 850 ◦ C in H2 atmosphere (Fig. 11), indicating that calcination at high temperature suppressed the catalytic activities. The NO conversion temperature (T99% ) were 300 and 350 ◦ C for the catalysts calcined at 750 and 850 ◦ C, respectively, which attributed to the different selectivities of N2 O. 3.6. XRD analysis of CuO/TiO2 and CuO–ZrO2 /TiO2 As shown in Fig. 12 (line 1), there was a diffraction peak of anatase at 2θ 25.2◦ , 37.0◦ and 48.1◦ but not a rutile phase after TiO2 was calcined at 450 ◦ C for 4 h. Two CuO diffraction peaks were obvious between 2θ 35.5◦ and 38.7◦ plus a diffraction peak of anatase phase (line 2), indicating that the CuO on TiO2 had transformed from highly dispersion CuO species to a CuO crystal phase. After addition of ZrO2 (5–15%), the diffraction peak of ZrO2 was not detected (lines 3 and 4) and the diffraction peak intensity of CuO did not increase either. Likely, it was the strong interaction between CuO and ZrO2 that enhanced the activity and stability of cat-
Fig. 11. Activity of 6% CuO–10% ZrO2 /TiO2 with different calcination temperature in H2 atmosphere.
Fig. 12. XRD patterns of catalysts: (1) TiO2 ; (2) 6% CuO/TiO2 ; (3) 6% CuO–5% ZrO2 /TiO2 ; (4) 6% CuO–10% ZrO2 /TiO2 ; (5) 6% CuO–10% ZrO2 /TiO2 (750 ◦ C); (6) 6% CuO/TiO2 (500 ◦ C, H2 atmosphere); (7) 6% CuO–10% ZrO2 /TiO2 (500 ◦ C, H2 atmosphere); (8) 6% CuO–10% ZrO2 /TiO2 (750 ◦ C, H2 atmosphere).
alyst. XRD also detected a mixture phase of anatase and rutile after the catalysts were calcined at 750 ◦ C for 2 h (line 5). After 6% CuO/TiO2 and 6% CuO–10% ZrO2 /TiO2 were pretreated at 500 ◦ C for 1 h in H2 , the CuO diffraction peaks (d = 2.52 and 2.33) disappeared but Cu diffraction peaks (d = 2.09 and 1.80) were detected, as shown in lines 6–8. By comparison, the catalysts calcined at 750 ◦ C for 2 h in both H2 and air atmospheres, the diffraction peaks of rutile (d = 3.25, 2.48 and 1.68) became much stronger additional to the conversion of CuO to Cu. The results indicate that the components changed and a strong interaction took place between the CuO and ZrO2 –TiO2 compound. Based on the results of this study and a previous study by Yasuaki and Hideki [12], we suggest that in H2 atmosphere CuO produced dispersed Cu0 species, and after one NO molecule on the Cu0 sites formed a mononitrosyl Cu0 species another NO molecule reacted with the mononitrosyl Cu0 species to produce N2 O and an unstable dinitrosyl intermediate, and then returned back to Cu0 species, i.e. a continuous redox cycle involving Cu2+ and Cu0 . It was likely that NO reacted with a mononitrosyl Cu0 species to form N2 and dispersed Cu0 species. 3.7. NO-TPD analysis After the surface of 6% CuO/TiO2 adsorbed NO, thermal desorption took place as a dissociation reaction, and four desorption species (NO, m/e = 30; N2 O, m/e = 44; N2 , m/e = 28 and O2 , m/e = 32) were detected (Figs. 13 and 14). A 6% CuO/TiO2 pretreated in air atmosphere had three NO desorption peaks, i.e. three and four NO adsorption centers, and the desorption peak temperatures were 100, 160 and 200 ◦ C. When the catalyst was pretreated in H2 , its NO desorption also had three desorption peaks, one peak due to NO adsorption at weak sites and the other two peaks due to NO adsorption at strong sites, and desorption peak temperatures were 100, 150 and 220 ◦ C. After 10% ZrO2 was added into 6% CuO/TiO2 , four NO desorption peaks occurred at 60, 100, 150 and 180 ◦ C, indicating that NO desorption by
816
X. Jiang et al. / Catalysis Today 93–95 (2004) 811–818
Fig. 13. NO-TPD spectra of 6% CuO/TiO2 catalysts (air atmosphere): NO (m/e = 30), N2 O (m/e = 44), N2 (m/e = 28), O2 (m/e = 32).
Fig. 14. NO-TPD spectra of 6% CuO–10% ZrO2 /TiO2 catalysis (air atmosphere): NO (m/e = 30), N2 O (m/e = 44), N2 (m/e = 28), O2 (m/e = 32).
CuO–ZrO2 /TiO2 was easier than that by CuO/TiO2 . In addition, desorption peak temperature of NO by 6% CuO–10% ZrO2 /TiO2 was lower than that by 6% CuO/TiO2 , i.e. the former catalyst had higher NO dissociation activity than the latter catalyst. Both catalysts also showed N2 O desorption peak at low temperatures but N2 desorption peak at high temperatures, indicating the production of N2 O and N2 in NO + CO reaction. The NO-TPD analysis of both catalysts also detected NO but not O2 desorption peak. It was likely that NO dissociation produced atomic oxygen (O) that dispersed or entered into the bulk of Cu [13,14], and the desorption of atomic oxygen could only occur at high temperatures. As all the NO-TPD spectra have shown that NO is the major desorption species, we compared the NO-TPD spectra of NO after NO adsorption by 6% CuO/TiO2 with that by 6% CuO–10% ZrO2 /TiO2 in order to explain the higher catalytic activity in H2 atmosphere than in air atmosphere. The adsorption–desorption process of NO absorbed on the different active sites of catalyst surface is very complex because of the complexity of catalyst structure. As shown in Fig. 15, while the desorption temperature decreased markedly and the number of adsorption sites increased in H2 atmosphere, the total amount of NO desorption in H2 atmosphere was almost the same as that in air atmosphere, indicating that
catalytic activity increased with the decrease in NO desorption temperature and the increase in active cites but was not related to the amount of NO desorption. In addition, since CuO became Cu in the process of H2 reduction, we suggest that the desorption peak at 100 ◦ C corresponded to the Cu adsorption site and played some role in the increment of catalytic activity. The NO desorption peaks of 6% CuO–10% ZrO2 /TiO2 (750 ◦ C) showed little change in shapes before and after H2
Fig. 15. NO-TPD spectra of 6% CuO/TiO2 (500 ◦ C) in air and H2 atmosphere.
X. Jiang et al. / Catalysis Today 93–95 (2004) 811–818
817
the formation of N2 O. Moreover, N2 O dissociation easily at high temperature also decreased the amount of N2 O, and N2 was the main product at high temperature. The Cu species also have great effect in the process of NO(a) dissociation beside temperature. The catalytic activity in H2 atmosphere was higher than in air, so we think that the process of dissociation occurred at the Cu0 active sites, forming more N(a) and O(a) but remaining less undissociated NO(a). Comparing to step (7), the reaction rate of step (9) was very fast, so at the same temperature the undissociated NO(a) would be less on the catalysts having Cu0 , and more N2 and less N2 O were formed. Fig. 16. NO-TPD spectra of 6% CuO–10% ZrO2 /TiO2 (750 ◦ C) in air and H2 atmosphere.
pretreatment (Fig. 16). However, the amount of NO desorption increased after H2 pretreatment, followed by increment in dissociation activity and N2 desorption, and as a result, the catalytic activity increased. 3.8. The mechanisms of NO + CO reaction by CuO/TiO2 and CuO–ZrO2 /TiO2 catalysts It is widely known that the mechanisms of NO + CO reaction are as the followings [12,15,16]: CO + s ↔ CO(a)
(1)
NO + s ↔ NO(a)
(2)
NO(a) + s → N(a) + O(a)
(3)
CO(a) + O(a) → CO2 + 2s
(4)
NO(a) + N(a) → N2 O(a) + s
(5)
NO(a) + N(a) → N2 + O(a) + s
(6)
N2 O(a) → N2 O + s
(7)
N2 O(a) → N2 + O(a)
(8)
N(a) + N(a) → N2 + 2s
(9)
where ‘s’ represents the adsorption site and ‘a’ the adsorption status. We think the above mechanisms also apply to the NO + CO reaction by CuO/TiO2 and CuO–ZrO2 /TiO2 and the dissociation of NO(a) might be the rate-determining step of NO + CO reaction. NO(a) hardly dissociated at low temperature and the rate of step (9) was also very slow, and thus more NO(a) would exist on the catalyst surface and easily reacted with the N(a) to form N2 O (step (5)). As a result, N2 O was detected as main intermediate product at low temperatures. The formation of N2 may mainly attribute to the reaction between absorbed N atom and absorbed N molecule (step (6)). By comparison, the rate of NO(a) dissociation was very fast at high temperature (step (3)) and the amount of NO(a) on the catalyst surface was very little, suppressing
4. Conclusions (1) The activities of CuO/TiO2 catalysts pretreated in H2 were much higher than those pretreated in air at 500 ◦ C for 1 h. The NO conversion temperatures of 6% CuO/TiO2 were 280 ◦ C at T50% and 450 ◦ C at T99% , compared to 325 ◦ C at T99% for 6% CuO–10% ZrO2 /TiO2 . In H2 atmosphere, the NO conversion temperature of 6% CuO–10% ZrO2 /TiO2 decreased to 300 ◦ C at T99% . (2) An addition of 5% ZrO2 into CuO/TiO2 calcined at 750 and 850 ◦ C for 2 h caused small changes in anti-high temperature property, whereas 10% ZrO2 addition significantly increased the catalytic activities, thermal stability and N2 selectivity in the NO + CO reaction. (3) The catalysts pretreated in air displayed CuO diffraction peaks, whereas those pretreated in H2 displayed Cu diffraction peaks. (4) Four desorption species (NO, N2 O, N2 and O2 ) were produced during the thermal desorption of NO adsorbed on CuO/TiO2 and CuO–ZrO2 /TiO2 . ZrO2 addition into CuO/TiO2 shifted the NO desorption peaks towards low temperature, indicating that the NO dissociation activity of CuO–ZrO2 /TiO2 was higher than that of CuO/TiO2 .
References [1] M. Iwamoto, H. Yahiro, Y. Mine, J. Chem. Lett. (1989) 213. [2] M. Iwamoto, H. Yahiro, T. Torikai, J. Chem. Lett. (1990) 1967. [3] D. Stoyanova, M. Christova, P. Dimitrova, Appl. Catal. B: Environ. 17 (1998) 233. [4] G.D. Mariadasson, F. Fajardie, J. Francois, J. Mol. Catal. A: Chem. 161 (2000) 179. [5] Y.J. Li, J.N. Armor, Appl. Catal. B: Environ. 4 (1994) L1. [6] G. Cordaba, M. Viniegra, J. Fierro, J. Solid Chem. 138 (1998) 1. [7] O.V. Komova, A.V. Simakov, V.A. Rogov, J. Mol. Catal. A: Chem. 161 (2000) 19l. [8] B. Xu, D. Li, Y. Chen, Chem. Soc., Faraday Trans. 94 (1998) 1905. [9] P.O. Lnrsson, A. Andersson, L.R. Wallenberg, J. Catal. 163 (1996) 279. [10] P.O. Larsson, A. Anderson, J. Catal. 179 (1998) 72.
818
X. Jiang et al. / Catalysis Today 93–95 (2004) 811–818
[11] Y. Okamoto, H. Gotoh, K. Hishida, et al., Appl. Surf. Sci. 121/122 (1997) 509. [12] O. Yasuaki, G. Hideki, Catal. Today 36 (1997) 71. [13] K.Y. Simon Ng, D.N. Belton, S.J. Schmieg, J. Catal. 146 (1994) 394.
[14] J. Kaspar, C. Leitenburg, P. Fomasiero, J. Catal. 146 (1994) 136. [15] M. Shelef, G.W. Graham, Catal. Rev. Sci. Eng. 36 (1994) 433. [16] F. Boccuzzi, E. Gulieminotti, G. Martra, et al., J. Catal. 146 (1994) 449.
Effect of ZrO2 addition on CuO/TiO2 activity in the NO + CO reaction Xiaoyuan Jiang a,∗ , Guanghui Ding a , Liping Lou b , Yingxu Chen b , Xiaoming Zheng a b
a Faculty of Science, Institute of Catalysis, Zhejiang University, Hangzhou 310028, PR China Faculty of Environment and Resource, Institute of Environmental Engineering, Zhejiang University, Hangzhou 310029, PR China
Available online 29 July 2004
Abstract The activities of CuO/TiO2 and ZrO2 -modified CuO/TiO2 in the NO reduction by CO were examined with a microreactor-gas chromatography (GC) NO + CO reaction system, and the catalysts were characterized by physical adsorption, TG-DTA, XRD and NO-TPD techniques. The results showed that the NO conversion temperatures of 6% CuO/TiO2 were 280 ◦ C at T50% and 450 ◦ C at T99% , compared to 325 ◦ C at T99% for 6% CuO–10% ZrO2 /TiO2 . The 6% CuO–10% ZrO2 /TiO2 calcined at 500 and 750 ◦ C had similar catalytic activities (T99% at 300 ◦ C) in H2 atmosphere. The nitrogen adsorption–desorption isotherm and pore-size distribution curve of TiO2 represented type IV of the BDDT system and a typical mesoporous sample, respectively. After ZrO2 addition, the diffraction peak of ZrO2 was not detected and the diffraction peak intensity of CuO did not increase either, likely due to the strong interaction between CuO and ZrO2 that enhanced the catalytic activity and stability. In the H2 atmosphere, CuO produced dispersed Cu0 species. Four desorption species (NO, N2 O, N2 and O2 ) were detected during the thermal desorption of NO by CuO/TiO2 and CuO–ZrO2 /TiO2 treated in both air and H2 . Addition of ZrO2 onto CuO/TiO2 shifted the NO dissociation peaks towards low temperature, indicating that the activity of NO decomposition was higher by CuO–ZrO2 /TiO2 than by CuO/TiO2 . The NO + CO reaction formed intermediary product N2 O at low temperature but formed N2 at high temperature. In addition, the peak temperature of NO desorption was related to the catalytic activity under both air and H2 , and the dissociation of NO on catalysts surface was a rate-determining step in NO + CO reaction. © 2004 Elsevier B.V. All rights reserved. Keywords: CuO/TiO2 catalysts; ZrO2 ; NO + CO reaction; NO-TPD
1. Introduction Nitric oxide (NOx ) is the major toxic pollutant in urban atmosphere, and catalytic reduction of NO with CO is one of the most important reactions for NOx controlling. At present, noble metal catalysts are mainly used for DeNOx . Due to high prices and scarcity of noble metals, it is necessary to find an alternative catalytic component to reduce using or even replace the noble metal. Therefore, studies of non-noble metal catalysts have attracted much attention for the last 10 years. TiO2 is good at resisting water and trace SO2 along with high N2 selectivity in the course of DeNOx . By comparison, ZrO2 has better thermal stability, unique surface acidity and alkalescence and redox properties. Meanwhile, ZrO2 is a semiconductor, which means it can easily produce cavity and has strong interaction with active component. So it is possible to improve the perfor-
∗
Corresponding author. E-mail address: [email protected] (X. Jiang).
0920-5861/$ – see front matter © 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.cattod.2004.06.074
mance of catalysts by changing the composition of carrier. However, there have been few studies of the Ti–Zr complex oxide supported CuO for NO + CO reaction. In 1989, Iwamoto et al. [1,2] found that Cu-ZSM5 catalyst had much higher catalytic activity for NO dissociation. From then on, catalytic reduction of NO has been extensively studied, and variety of transition and noble metal catalysts were prepared [3,4]. Recently, the use of TiO2 as catalyst’s carrier has attracted much attention and TiO2 -based catalysts are widely applied in NO reduction [5–8]. The TiO2 -based catalysts overcome the shortcoming of TiO2 and SOx interaction that produces sulphate and causes structural collapse of the carrier, particle gathering of active components and reduction of active surfaces. As a result, the catalytic reduction of NO can be achieved in a sulphur-containing atmosphere, and obtain an excellent NO conversion and N2 selectivity. Various physicochemical methods are also used to investigate the state of copper ion in CuO/TiO2 catalysts, and inconsistent results have been reported. Komova et al. [7] found that TiO2 -supported CuO could exist as Cu2+ ions, bulk CuO, chain stabilized Cu2+ ions, and two different
812
X. Jiang et al. / Catalysis Today 93–95 (2004) 811–818
oxide clusters with a structure similar to CuO. In contrast, Xu et al. [8] reported two Cu2+ ions with significantly different reducibility on the anatase surface—CuO phase and highly dispersed Cu2+ , similar to CuO/␥-Al2 O3 . Larsson and co-workers [9,10] suggested that low temperature peak in Cu3–Ti of the H2 -temperature programmed reduction (TPR) profiles was monomeric copper containing Cu+ , and that in Cu12–Ti was polymeric copper containing Cu2+ . They also found that catalysts using TiO2 as carrier displayed a high catalysis in CO oxidation, and that CuO catalyst had much higher activity than MoOx , FeOx and CoOx catalysts and its activity at low temperature was even superior to Pt/␥-Al2 O3 catalyst. In this study, CuO/TiO2 and CuO–ZrO2 /TiO2 catalysts with different amount of CuO loadings were prepared by impregnating the TiO2 carrier, and these catalysts in the NO + CO reaction were examined by means of BET, TG-DTA, XRD and NO-temperature programmed desorption (TPD).
treated in 150 ml min−1 of dry pure N2 . The TG-DTA spectra were obtained with temperatures ramped at a linear heating rate of 20 ◦ C min−1 from 50 to 800 ◦ C. X-ray diffraction (XRD) data were obtained at 25 ◦ C using a horizontal Rigaku B/Max lllB powder diffractometer with Cu K␣ radiation and a power of 40 mA × 30 mA. To measure the NO-TPD, 250 mg of fresh catalysts were loaded onto a quartz reactor (Φ = 5 mm) and reduced in H2 atmosphere at 500 ◦ C for 1 h, and followed by two treatments. In treatment 1, catalysts were heated in He at 600 ◦ C for 1 h, and then cooled to 40 ◦ C in a flow of He and exposed to a 10% NO–He mixture gas. Excessive NO was removed, and the catalysts were kept in He flow until no significant amount of adsorbates could be detected. The catalysts were then ramped at 600 ◦ C at a linear heating rate of 20 ◦ C min−1 in He flow. The effluent gases were analysed with a mass spectrometer. In treatment 2, after the catalysts were heated in He at 500 ◦ C for 1 h, the procedures followed were the same as those in treatment 1.
2. Experimental
2.4. Measurements of catalytic activity in NO + CO reaction
2.1. TiO2 preparation TiO2 was prepared by sol–gel method from TiCl4 with ammonia solution as precipitating agent and ethanol as dispersing agent. A 25 ml TiCl4 was dissolved in 20 ml distilled water in an ice-water bath. The titanium solution was then slowly mixed with 30 ml distilled water and 20 ml ethanol, and ammonia was added dropwise until pH = 9. During the ammonic addition, an intensive precipitation occurred. After the solvent was evaporated at 80 ◦ C for 24 h, the precipitates were dried at 300 ◦ C for 2 h to remove NH4 Cl, and then calcined in an air stream of 450 ◦ C for 4 h. 2.2. Catalyst preparation The CuO/TiO2 and CuO–ZrO2 /TiO2 catalysts were prepared by the impregnation method using Cu(NO3 )2 and Zr(NO3 )4 aqueous solutions of desired concentrations. After these catalysts being rested for 24 h, catalysts were dried at 120 ◦ C for 2 h, followed the catalysts were put into a muffle furnace by calcinations in an air stream of 500 ◦ C for 2 h (β = 15 K min−1 ). They were denoted as w% CuO/TiO2 and w% CuO–w % ZrO2 /TiO2 , where w and w are the amount of CuO and ZrO2 loading, respectively. 2.3. Measurements of catalytic characters The BET surface area of catalysts was determined by N2 adsorption at 77 K using a Coulter OMNISORP-100 instrument. The TG-DTA experiment was carried out using PE-TGA7 thermo-gravimetric analyzer and DTA/9050311 high temperature differential thermal analyzer. Ten mg of fresh catalysts were taken and measured in air atmosphere, and then
Catalytic activity was determined under the steady state in a fixed-bed quartz reactor (6 mm). The particle size of catalysts was 20–40 mesh, and 120 mg of the catalysts were used. The reaction gas (i.e. feed stream) consisted of a fixed composition of 6.0% NO, 6.0% CO and 88% He (v/v) as a dilute. The catalysts were treated using either H2 or air at 500 ◦ C for 1 h. After cooled to room temperature, they were allowed to react with the mixed gas. The reactions were operated at different temperatures with a space velocity of 5000 h−1 . Two columns and thermal conduction detectors were used for analyzing the catalytic activity. Column A was packed with 13× molecular sieve for separating N2 , NO and CO, and Column B was packed with Paropak Q for separating N2 O and CO2 . 3. Results and discussion 3.1. Textural and structural properties of TiO2 Fig. 1 shows the nitrogen adsorption–desorption isotherm and pore-size distribution curve of TiO2 . The TiO2 was prepared from TiCl4 with a N2 isotherm corresponding to type IV of the BDDT system and was typical of a mesoporous sample. The specific surface area and total pore volume of TiO2 were 64.5 m2 /g and 0.441 ml/g, and average pore diameter and stochastic pore diameter were 27.3 nm and about 10 nm, respectively. The DTA curves of TiO2 showed an absorption heat peak at around 120 ◦ C, where a weight-lose occurred in its TG curve (Fig. 2). This peak could be due to the evaporation and removal of adsorbed H2 O and organic substance. Corresponding to an obvious weight-lose step of TiO2 around
X. Jiang et al. / Catalysis Today 93–95 (2004) 811–818
813
Fig. 4. Activity of 6% CuO/TiO2 with different calcination temperature for NO + CO reaction.
Fig. 1. Adsorption–desorption isotherms (a) and pore-size distributions (b) of TiO2 .
Calcining heat also affected the catalytic activities. For example, 6% CuO/TiO2 calcined at 750 ◦ C had higher activity than those calcined at 500 or 850 ◦ C (Fig. 4). The catalyst calcined at 500 ◦ C had higher activity than that calcined at 850 ◦ C at low reaction temperatures, compared to an opposite finding at high reaction temperatures. When calcined at 850 ◦ C, the catalytic activity began to decrease at the temperatures of above 500 ◦ C, probably due to the SMSI action [11]. 3.3. Activities of CuO–ZrO2 /TiO2 in NO + CO reaction
Fig. 2. TG-DTA curves of TiO2 sample.
320 ◦ C, an absorption heat peak was also observed as a result of NH4 Cl decomposition. 3.2. Activities of CuO/TiO2 in NO+CO reaction As shown in Fig. 3, the activities of CuO/TiO2 in NO + CO reaction increased with the increase in reaction temperature, ranging from low temperatures around 150 ◦ C to high temperatures above 300 ◦ C. The catalysts calcined at 500 ◦ C had very low activity at a CuO loading of 1–3%. The NO conversion temperature (T50% ) was 450 ◦ C for 6% CuO/TiO2 and 375 ◦ C for 18% CuO/TiO2 .
Fig. 3. Activity of CuO/TiO2 (500 ◦ C) with different CuO loading for NO + CO reaction.
As shown in Fig. 5, the NO conversion temperatures of 6% CuO/TiO2 were 280 ◦ C at T50% and 450 ◦ C at T99% . By comparison, both 6% CuO–5% ZrO2 /TiO2 and 12% CuO–5% ZrO2 /TiO2 had a NO conversion temperature of 350 ◦ C at T99% . We assume that ZrO2 addition increased the catalytic activity in NO + CO reaction, whereas CuO addition made little contribution to the catalytic activity even if the loading of CuO increased to 12%. It was also found that both 6% CuO–10% ZrO2 /TiO2 and 6% CuO–15% ZrO2 /TiO2 had a NO conversion temperature of 325 ◦ C at T99% (Fig. 6), suggesting that excessive addition of ZrO2 would not further increase the catalytic activity. At low reaction temperature, 6% CuO–10% ZrO2 /TiO2 calcined at 500 ◦ C had higher activity in NO + CO reaction than that calcined at 750 or 850 ◦ C, and similar catalytic activities were observed when calcined at 750 and 850 ◦ C (Fig. 7). However, the catalysts calcined at 500, 750 or 850 ◦ C had almost the same conversion temperature (T99% ), indicating that 6% CuO–10% ZrO2 /TiO2 calcined at 750 ◦ C had better heat stability.
Fig. 5. Effect of CuO loading on activity of CuO–5% ZrO2 /TiO2 in NO + CO reaction.
814
X. Jiang et al. / Catalysis Today 93–95 (2004) 811–818
Fig. 6. Effect of ZrO2 addition on activity of CuO/TiO2 in NO + CO reaction.
Fig. 7. Effect of calcination temperature on activity of 6% CuO–10% ZrO2 /TiO2 in NO + CO reaction.
3.4. Effects of reaction temperature on NO, CO conversion and N2 selectivity Table 1 shows the effects of reaction temperature on NO, CO conversion and N2 selectivity of 6% CuO/TiO2 and 6% CuO–10% ZrO2 /TiO2 catalysts. It was found that NO and CO concentrations decreased but N2 and CO2 concentrations increased with the increase in reaction temperature. At 150 ◦ C, the amount of N2 , CO2 and N2 O and the conversions of NO and CO increased markedly. A continuous increase in N2 and CO2 but a decrease in N2 O was observed at 200 ◦ C. At 300 ◦ C, NO and CO conversions were 66.7, 70.6% by 6% CuO/TiO2 , and 98.9, 96.9% by 6% CuO–10% ZrO2 /TiO2 , respectively. At 350 ◦ C, NO and CO were almost completely converted by 6% CuO–10% ZrO2 /TiO2 compared with 91.3 and 85.6% by 6% CuO/TiO2 , respectively. These results indicate that N2 O was formed during the initial stage of the reaction at lower temperature and increased with the increase in temperature until a maximal amount of Table 1 Effects of reaction temperature on NO, CO conversion and N2 selectivities for NO + CO reaction TR (◦ C)
150 200 250 300 350 400 450 500
CNO (%)
Fig. 8. Activities of CuO/TiO2 catalysts with different CuO loadings in H2 atmosphere.
N2 O was reached. Further increase in temperature, however, caused a gradual decrease in N2 O. In contrast, the amount of N2 increased continuously with increase in temperature. It can be suggested that in the process of NO reduction with CO, the reactant molecules (NO and CO) were firstly adsorbed on the adsorption sites of catalyst surface in forms of NO(a) and CO(a), and the adsorbed NO(a) then dissociated into N(a) and O(a), followed by reactions of N(a) with NO(a) and CO(a) with O(a) to produce N2 (ads) and CO2 , and N2 formation by two N(a). This study also showed that the rate of NO(a) dissociation on the catalyst surface was slower at low temperature and its reaction with N(a) formed N2 O(ads). As the reaction temperature increased, the rate of NO(a) dissociation became faster, causing a decrease in NO(a) concentration. As a result, high temperature was not beneficial to N2 O(a) formation. Therefore, the catalytic activities in the NO + CO reaction were not only related to the dispersion state of active components on TiO2 but also to the rate of NO adsorption and dissociation on the catalyst surfaces. 3.5. Activities of CuO/TiO2 catalysts in H2 atmosphere Fig. 8 shows the activities of CuO/TiO2 with different CuO loading in H2 atmosphere for NO + CO reaction. The sample treated in H2 possessed higher catalytic activity than that treated in air. The catalytic activity increased with the increase in temperature and the CuO loading from 1 to 18%. The catalysts treated in H2 had higher activity than that treated in air. For example, 6% CuO/TiO2 showed 50% NO conversation at reaction temperature of 450 ◦ C in air compared to 150 ◦ C in H2 . Calcining temperature also affected the catalytic activity in NO + CO reaction. As shown in Fig. 9, the NO conver-
SN2 (%)
CCO (%)
Cu–Ti
Cu–Zr–Ti
Cu–Ti
Cu–Zr–Ti
Cu–Ti
Cu–Zr–Ti
30.8 35.3 40.9 66.7 91.3 91.8 99.0 100
41.0 42.9 70.5 98.9 100 100 100 100
21.9 21.9 23.0 70.6 85.5 85.2 90.2 97.3
23.9 28.9 64.0 96.9 99.2 100 100 100
13.7 10.1 15.9 65.7 78.4 100 100 100
0 9.26 43.9 91.9 100 100 100 100
Fig. 9. Catalytic activity of 6% CuO/TiO2 catalysts calcined at different temperatures in H2 atmosphere.
X. Jiang et al. / Catalysis Today 93–95 (2004) 811–818
815
Fig. 10. Effect of ZrO2 addition on activity of CuO/TiO2 in H2 atmosphere.
sion of 6% CuO/TiO2 decreased with the increase in calcining temperature. The reaction temperature of 99% NO conversion was 300 ◦ C and 375 ◦ C in H2 atmosphere when the calcining temperature was 500 and 750 ◦ C, respectively. At reaction temperature of 500 ◦ C, the catalyst calcined at 850 ◦ C only had 74.84% NO conversion. Fig. 10 displays the effects of H2 treatment and ZrO2 addition on catalytic activity of the CuO/TiO2 catalysts in NO + CO reaction. After treatment in H2 , the catalytic activity increased significantly, especially for 6% CuO/TiO2 . The order of the catalytic activities was, 6% CuO/TiO2 > 12% CuO–5% ZrO2 /TiO2 > 6% CuO–5% ZrO2 /TiO2 . The activity of 6% CuO–10% ZrO2 /TiO2 calcined at 500 ◦ C was similar to that calcined at 750 ◦ C, but decreased dramatically at calcining temperature of 850 ◦ C in H2 atmosphere (Fig. 11), indicating that calcination at high temperature suppressed the catalytic activities. The NO conversion temperature (T99% ) were 300 and 350 ◦ C for the catalysts calcined at 750 and 850 ◦ C, respectively, which attributed to the different selectivities of N2 O. 3.6. XRD analysis of CuO/TiO2 and CuO–ZrO2 /TiO2 As shown in Fig. 12 (line 1), there was a diffraction peak of anatase at 2θ 25.2◦ , 37.0◦ and 48.1◦ but not a rutile phase after TiO2 was calcined at 450 ◦ C for 4 h. Two CuO diffraction peaks were obvious between 2θ 35.5◦ and 38.7◦ plus a diffraction peak of anatase phase (line 2), indicating that the CuO on TiO2 had transformed from highly dispersion CuO species to a CuO crystal phase. After addition of ZrO2 (5–15%), the diffraction peak of ZrO2 was not detected (lines 3 and 4) and the diffraction peak intensity of CuO did not increase either. Likely, it was the strong interaction between CuO and ZrO2 that enhanced the activity and stability of cat-
Fig. 11. Activity of 6% CuO–10% ZrO2 /TiO2 with different calcination temperature in H2 atmosphere.
Fig. 12. XRD patterns of catalysts: (1) TiO2 ; (2) 6% CuO/TiO2 ; (3) 6% CuO–5% ZrO2 /TiO2 ; (4) 6% CuO–10% ZrO2 /TiO2 ; (5) 6% CuO–10% ZrO2 /TiO2 (750 ◦ C); (6) 6% CuO/TiO2 (500 ◦ C, H2 atmosphere); (7) 6% CuO–10% ZrO2 /TiO2 (500 ◦ C, H2 atmosphere); (8) 6% CuO–10% ZrO2 /TiO2 (750 ◦ C, H2 atmosphere).
alyst. XRD also detected a mixture phase of anatase and rutile after the catalysts were calcined at 750 ◦ C for 2 h (line 5). After 6% CuO/TiO2 and 6% CuO–10% ZrO2 /TiO2 were pretreated at 500 ◦ C for 1 h in H2 , the CuO diffraction peaks (d = 2.52 and 2.33) disappeared but Cu diffraction peaks (d = 2.09 and 1.80) were detected, as shown in lines 6–8. By comparison, the catalysts calcined at 750 ◦ C for 2 h in both H2 and air atmospheres, the diffraction peaks of rutile (d = 3.25, 2.48 and 1.68) became much stronger additional to the conversion of CuO to Cu. The results indicate that the components changed and a strong interaction took place between the CuO and ZrO2 –TiO2 compound. Based on the results of this study and a previous study by Yasuaki and Hideki [12], we suggest that in H2 atmosphere CuO produced dispersed Cu0 species, and after one NO molecule on the Cu0 sites formed a mononitrosyl Cu0 species another NO molecule reacted with the mononitrosyl Cu0 species to produce N2 O and an unstable dinitrosyl intermediate, and then returned back to Cu0 species, i.e. a continuous redox cycle involving Cu2+ and Cu0 . It was likely that NO reacted with a mononitrosyl Cu0 species to form N2 and dispersed Cu0 species. 3.7. NO-TPD analysis After the surface of 6% CuO/TiO2 adsorbed NO, thermal desorption took place as a dissociation reaction, and four desorption species (NO, m/e = 30; N2 O, m/e = 44; N2 , m/e = 28 and O2 , m/e = 32) were detected (Figs. 13 and 14). A 6% CuO/TiO2 pretreated in air atmosphere had three NO desorption peaks, i.e. three and four NO adsorption centers, and the desorption peak temperatures were 100, 160 and 200 ◦ C. When the catalyst was pretreated in H2 , its NO desorption also had three desorption peaks, one peak due to NO adsorption at weak sites and the other two peaks due to NO adsorption at strong sites, and desorption peak temperatures were 100, 150 and 220 ◦ C. After 10% ZrO2 was added into 6% CuO/TiO2 , four NO desorption peaks occurred at 60, 100, 150 and 180 ◦ C, indicating that NO desorption by
816
X. Jiang et al. / Catalysis Today 93–95 (2004) 811–818
Fig. 13. NO-TPD spectra of 6% CuO/TiO2 catalysts (air atmosphere): NO (m/e = 30), N2 O (m/e = 44), N2 (m/e = 28), O2 (m/e = 32).
Fig. 14. NO-TPD spectra of 6% CuO–10% ZrO2 /TiO2 catalysis (air atmosphere): NO (m/e = 30), N2 O (m/e = 44), N2 (m/e = 28), O2 (m/e = 32).
CuO–ZrO2 /TiO2 was easier than that by CuO/TiO2 . In addition, desorption peak temperature of NO by 6% CuO–10% ZrO2 /TiO2 was lower than that by 6% CuO/TiO2 , i.e. the former catalyst had higher NO dissociation activity than the latter catalyst. Both catalysts also showed N2 O desorption peak at low temperatures but N2 desorption peak at high temperatures, indicating the production of N2 O and N2 in NO + CO reaction. The NO-TPD analysis of both catalysts also detected NO but not O2 desorption peak. It was likely that NO dissociation produced atomic oxygen (O) that dispersed or entered into the bulk of Cu [13,14], and the desorption of atomic oxygen could only occur at high temperatures. As all the NO-TPD spectra have shown that NO is the major desorption species, we compared the NO-TPD spectra of NO after NO adsorption by 6% CuO/TiO2 with that by 6% CuO–10% ZrO2 /TiO2 in order to explain the higher catalytic activity in H2 atmosphere than in air atmosphere. The adsorption–desorption process of NO absorbed on the different active sites of catalyst surface is very complex because of the complexity of catalyst structure. As shown in Fig. 15, while the desorption temperature decreased markedly and the number of adsorption sites increased in H2 atmosphere, the total amount of NO desorption in H2 atmosphere was almost the same as that in air atmosphere, indicating that
catalytic activity increased with the decrease in NO desorption temperature and the increase in active cites but was not related to the amount of NO desorption. In addition, since CuO became Cu in the process of H2 reduction, we suggest that the desorption peak at 100 ◦ C corresponded to the Cu adsorption site and played some role in the increment of catalytic activity. The NO desorption peaks of 6% CuO–10% ZrO2 /TiO2 (750 ◦ C) showed little change in shapes before and after H2
Fig. 15. NO-TPD spectra of 6% CuO/TiO2 (500 ◦ C) in air and H2 atmosphere.
X. Jiang et al. / Catalysis Today 93–95 (2004) 811–818
817
the formation of N2 O. Moreover, N2 O dissociation easily at high temperature also decreased the amount of N2 O, and N2 was the main product at high temperature. The Cu species also have great effect in the process of NO(a) dissociation beside temperature. The catalytic activity in H2 atmosphere was higher than in air, so we think that the process of dissociation occurred at the Cu0 active sites, forming more N(a) and O(a) but remaining less undissociated NO(a). Comparing to step (7), the reaction rate of step (9) was very fast, so at the same temperature the undissociated NO(a) would be less on the catalysts having Cu0 , and more N2 and less N2 O were formed. Fig. 16. NO-TPD spectra of 6% CuO–10% ZrO2 /TiO2 (750 ◦ C) in air and H2 atmosphere.
pretreatment (Fig. 16). However, the amount of NO desorption increased after H2 pretreatment, followed by increment in dissociation activity and N2 desorption, and as a result, the catalytic activity increased. 3.8. The mechanisms of NO + CO reaction by CuO/TiO2 and CuO–ZrO2 /TiO2 catalysts It is widely known that the mechanisms of NO + CO reaction are as the followings [12,15,16]: CO + s ↔ CO(a)
(1)
NO + s ↔ NO(a)
(2)
NO(a) + s → N(a) + O(a)
(3)
CO(a) + O(a) → CO2 + 2s
(4)
NO(a) + N(a) → N2 O(a) + s
(5)
NO(a) + N(a) → N2 + O(a) + s
(6)
N2 O(a) → N2 O + s
(7)
N2 O(a) → N2 + O(a)
(8)
N(a) + N(a) → N2 + 2s
(9)
where ‘s’ represents the adsorption site and ‘a’ the adsorption status. We think the above mechanisms also apply to the NO + CO reaction by CuO/TiO2 and CuO–ZrO2 /TiO2 and the dissociation of NO(a) might be the rate-determining step of NO + CO reaction. NO(a) hardly dissociated at low temperature and the rate of step (9) was also very slow, and thus more NO(a) would exist on the catalyst surface and easily reacted with the N(a) to form N2 O (step (5)). As a result, N2 O was detected as main intermediate product at low temperatures. The formation of N2 may mainly attribute to the reaction between absorbed N atom and absorbed N molecule (step (6)). By comparison, the rate of NO(a) dissociation was very fast at high temperature (step (3)) and the amount of NO(a) on the catalyst surface was very little, suppressing
4. Conclusions (1) The activities of CuO/TiO2 catalysts pretreated in H2 were much higher than those pretreated in air at 500 ◦ C for 1 h. The NO conversion temperatures of 6% CuO/TiO2 were 280 ◦ C at T50% and 450 ◦ C at T99% , compared to 325 ◦ C at T99% for 6% CuO–10% ZrO2 /TiO2 . In H2 atmosphere, the NO conversion temperature of 6% CuO–10% ZrO2 /TiO2 decreased to 300 ◦ C at T99% . (2) An addition of 5% ZrO2 into CuO/TiO2 calcined at 750 and 850 ◦ C for 2 h caused small changes in anti-high temperature property, whereas 10% ZrO2 addition significantly increased the catalytic activities, thermal stability and N2 selectivity in the NO + CO reaction. (3) The catalysts pretreated in air displayed CuO diffraction peaks, whereas those pretreated in H2 displayed Cu diffraction peaks. (4) Four desorption species (NO, N2 O, N2 and O2 ) were produced during the thermal desorption of NO adsorbed on CuO/TiO2 and CuO–ZrO2 /TiO2 . ZrO2 addition into CuO/TiO2 shifted the NO desorption peaks towards low temperature, indicating that the NO dissociation activity of CuO–ZrO2 /TiO2 was higher than that of CuO/TiO2 .
References [1] M. Iwamoto, H. Yahiro, Y. Mine, J. Chem. Lett. (1989) 213. [2] M. Iwamoto, H. Yahiro, T. Torikai, J. Chem. Lett. (1990) 1967. [3] D. Stoyanova, M. Christova, P. Dimitrova, Appl. Catal. B: Environ. 17 (1998) 233. [4] G.D. Mariadasson, F. Fajardie, J. Francois, J. Mol. Catal. A: Chem. 161 (2000) 179. [5] Y.J. Li, J.N. Armor, Appl. Catal. B: Environ. 4 (1994) L1. [6] G. Cordaba, M. Viniegra, J. Fierro, J. Solid Chem. 138 (1998) 1. [7] O.V. Komova, A.V. Simakov, V.A. Rogov, J. Mol. Catal. A: Chem. 161 (2000) 19l. [8] B. Xu, D. Li, Y. Chen, Chem. Soc., Faraday Trans. 94 (1998) 1905. [9] P.O. Lnrsson, A. Andersson, L.R. Wallenberg, J. Catal. 163 (1996) 279. [10] P.O. Larsson, A. Anderson, J. Catal. 179 (1998) 72.
818
X. Jiang et al. / Catalysis Today 93–95 (2004) 811–818
[11] Y. Okamoto, H. Gotoh, K. Hishida, et al., Appl. Surf. Sci. 121/122 (1997) 509. [12] O. Yasuaki, G. Hideki, Catal. Today 36 (1997) 71. [13] K.Y. Simon Ng, D.N. Belton, S.J. Schmieg, J. Catal. 146 (1994) 394.
[14] J. Kaspar, C. Leitenburg, P. Fomasiero, J. Catal. 146 (1994) 136. [15] M. Shelef, G.W. Graham, Catal. Rev. Sci. Eng. 36 (1994) 433. [16] F. Boccuzzi, E. Gulieminotti, G. Martra, et al., J. Catal. 146 (1994) 449.