- Email: [email protected]
TiO2 composite nanoxides synthesized by deposition-precipitation method as a superior catalyst for NO oxidation
Journal of Colloid and Interface Science 352 (2010) 143–148
Contents lists available at ScienceDirect
Journal of Colloid and Interface Science www.elsevier.com/locate/jcis
MnOx/TiO2 composite nanoxides synthesized by deposition-precipitation method as a superior catalyst for NO oxidation Zhongbiao Wu, Nian Tang, Ling Xiao, Yue Liu ⇑, Haiqiang Wang Department of Environmental Engineering, Zhejiang University, Hangzhou 310027, China Zhejiang Provincial Engineering Research Center of Industrial Boiler & Furnace Flue Gas Pollution Control, Hangzhou 310027, China
a r t i c l e
i n f o
Article history: Received 1 June 2010 Accepted 10 August 2010 Available online 13 August 2010 Keywords: NO catalytic oxidation NO2 MnOx/TiO2 composite nanoxides Deposition-precipitation method
a b s t r a c t A series of MnOx/TiO2 composite nanoxides were prepared by deposition-precipitation (DP) method, and the sample with the Mn/Ti ratio of 0.3 showed a superior activity for NO catalytic oxidation to NO2. The maximum NO conversion over MnOx(0.3)/TiO2(DP) could reach 89% at 250 °C with a GHSV of 25,000 h1, which was much higher than that over the catalyst prepared by conventional wet-impregnation (WI) method (69% at 330 °C). Characterization results including XRD, HRTEM, FTIR, XPS, H2-TPR, NO-TPD and Nitrogen adsorption–desorption implied that the higher activity of MnOx(0.3)/TiO2(DP) could be attributed to the enrichment of well-dispersed MnOx on the surface and the abundance of Mn3+ species. Furthermore, DRIFT investigations and long-time running test indicated that NO2 came from the decomposition of adsorbed nitrogen-containing species. Ó 2010 Elsevier Inc. All rights reserved.
1. Introduction Development of methods to control NOx emissions has been an active research area [1–3] because NOx are associated with acid deposition, photochemical smog and ozone depletion. NO2 is always more favored than NO for NOx conversion among several procedures. It was reported that the oxidation of NO–NO2 would definitely increase the NO conversion rate in the process of selective catalytic reduction (SCR) [4]. And this conclusion inspired Kang et al. to develop a high-performance two-stage catalyst system for SCR of NOx with the pre-oxidation of NO during the first stage [5]. For wet scrubbing method, the solubility and reactivity of NOx could also be improved significantly when NO is partially or completely converted to NO2 [6,7]. Therefore, much effort has been focused on developing eligible catalysts for NO oxidation [8–15]. The Pt-based catalysts have been most widely used for the NO oxidation [8–10] but were limited by their high cost. Metal oxides catalysts could be appropriate alternatives to the noble metal catalysts due to their potential high catalytic efficiency, low cost and persistent activity. Among them, Co-based catalysts attracted the most attention [11–13], while their applications were retarded by the toxicity of Cobalt. Nowadays, Mn-based catalysts, acting as low-temperature SCR catalysts [16–19], also showed certain ⇑ Corresponding author at: Department of Environmental Engineering, Zhejiang University, Hangzhou 310027, China. Fax: +86 571 87953088. E-mail address: [email protected] (Y. Liu). 0021-9797/$ - see front matter Ó 2010 Elsevier Inc. All rights reserved. doi:10.1016/j.jcis.2010.08.031
catalytic activity towards NO oxidation in the SCR process [16,17]. Very recently, Mn-based catalysts were directly employed as the NO oxidation catalysts and showed their potential in this field [14,15]. However, the NO conversions over these catalysts were not very satisfactory and the detail understanding was still lacking for NO oxidation over Mn-based catalysts. In this study, the main objective was to seek a Mn-based catalyst with high activity towards NO oxidation, and deposition-precipitation (DP) method was introduced to synthesize MnOx/TiO2 composite nanoxides as the catalyst. DP method involves the precipitation of a metal precursor onto a suspended support, and it could make the metal oxides highly loaded and enriched in the external surface [20,21]. The metal oxides particles could also be prevented from sintering during calcination by DP method and thus well-dispersed on the support [22]. For comparison, other catalysts with the same compositions were also prepared by conventional wet-impregnation (WI) method. 2. Materials and methods 2.1. Catalysts preparation The MnOx/TiO2 catalysts were prepared as follows: (a) DP method: 60 ml solution containing certain amount of Mn(NO3)2 and a mixed solution of NH3H2O (15 wt.%) and NH4HCO3 (15 wt.%) were added simultaneously and dropwise to a slurry solution (80 ml) containing 8 g commercial
144
Z. Wu et al. / Journal of Colloid and Interface Science 352 (2010) 143–148
TiO2 support (Degussa P25) under vigorous stirring and a constant pH value of 10 was maintained. The precipitate was aged at 50 °C for 2 h in the mother solution. After filtered and washed with deionized water for five times, the obtained sample was dried at 110 °C for 12 h and then calcined at 450 °C for 4 h in air. (b) WI method: the preparation process followed the typical WI method (not adding the mixed solution of NH3H2O and NH4HCO3 in the abovementioned procedure). The composition and calcination temperature were the same as those of DP method. 2.2. Catalytic activity tests Steady-state reaction experiments were performed in a fixedbed at 90–450 °C typically containing 1.6 g (2.8 ml) catalyst with a gas hourly space velocity (GHSV) of 25,000 h1. The reactant gas typically consisted of 600 ppm NO, 4% O2 and balance N2. The concentrations of NO and NO2 were monitored by a chemiluminescence NO–NO–NOx analyzer (Thermo 42i-HL). Since a balance of total NOx (NO + NO2) between the inlet and the outlet gases could be obtained for each experiment, the NO conversion ([Conv.]NO) was defined as:
½Conv:NO ¼ ½NO2 out =ð½NO2 out þ ½NOout Þ 100%
3. Results and discussion 3.1. Catalytic activity tests The conversion of NO–NO2 over a series of MnOx/TiO2 nanoxides with different Mn/Ti ratios prepared by DP method and WI method are shown in Fig. 1a and b, respectively. The thermodynamic equilibrium for NO–NO2 under given conditions is also shown for reference (in dashed). As expected, the reaction was kinetically limited at low temperature. Above temperatures at which the equilibrium curve was reached, the catalytic activities closely followed the equilibrium conversion and became thermodynamic control. The balance of total NOx (NO + NO2) between the inlet and the outlet gas could be obta, and no other nitrogen-containing product was detected in the outlet gas. It could be seen that the TiO2 support showed negligible catalytic activity during the entire temperature range and the introduction of MnOx improved the activities significantly. In both cases, the optimal Mn/Ti ratio appeared to be 0.3, and the corresponding samples were denoted as MnOx(0.3)/ TiO2(DP) and MnOx(0.3)/TiO2(WI), respectively. Moreover, MnOx(0.3)/TiO2(DP) exhibited much higher catalytic activity than MnOx(0.3)/TiO2(WI) towards NO oxidation, and it reached the equilibrium curve at about 250 °C, with the maximum NO conversion of 89%, which was comparable to the activities of Pt-based catalysts [10]. In contrast, the conversion of NO at the same temperature was only about 40% for MnOx(0.3)/TiO2(WI), which required a temperature of about 330 °C to reach the equilibrium, with the maximum NO conversion of only 69%. These
2.3. Characterization The crystal phase was analyzed by X-ray diffraction with Cu Ka radiation (XRD: model D/max RA, Rigaku Co., Japan). High-resolution transmission electron microscope (HRTEM: JOEL JEM-2010, Japan) was used to investigate the microstructures and morphologies of the prepared samples. Infrared absorption spectra were recorded at room temperature using a Fourier transform interferometer (FTIR: Shimadzu IRAffinity-1, Japan). Nitrogen adsorption–desorption isotherms were obtained at 77 K (Micromeritics ASAP 2020, USA). Then the specific surface areas of samples were measured using Brunauer–Emmett–Teller (BET) method, and the pore volume and pore size distributions were calculated by Barrett–Jioner–Halenda (BJH) method. X-ray photoelectron spectroscopy with Al Ka X-rays (hm = 1486.6 eV) radiation (XPS: Thermal ESCALAB 250, USA) was used to analyze the surface atomic states of the catalysts. Temperature programmed reduction (TPR) and temperature programmed desorption (TPD) tests were carried out on a custom-made TCD setup using 50 mg catalysts. Prior to H2-TPR experiments, samples were pretreated in nitrogen at 400 °C for 1 h and then cooled down to 100 °C. The reduction was conducted from 100 to 800 °C at a linear ramp of 5 °C min1 in 6% H2/N2 with a flow rate of 30 ml/min. For NO-TPD experiments, after pretreated in He at 400 °C for 1 h and cooled down to 30 °C, catalysts were saturated with 30 ml/min NO (4% in He) and 6% O2 for 20 min. After flushed with He for 1 h, desorption was carried out in 30 ml/min He from 30 to 400 °C with a heating rate of 5 °C min1. In situ diffuse reflectance infrared transform spectroscopy (DRIFT) investigations were performed with Nicolet 6700 FTIR spectrometers at 4 cm1 resolution with 64 co-added scans. In DRIFT cell, samples were pretreated at 400 °C in He for 2 h, and backgrounds were recorded after cooled to 250 °C. At this temperature co-adsorption of 5000 ppm NO and 10% O2 in balance He was performed for 60 min, after which the sample was then flushed with He for 30 min.
Fig. 1. NO conversion over MnOx/TiO2 catalysts of different Mn/Ti ratios prepared by (a) DP method; (b) WI method. Reaction conditions: 600 ppm NO, 4% O2 and balance N2; GHSV = 25,000 h1.
145
Z. Wu et al. / Journal of Colloid and Interface Science 352 (2010) 143–148 Table 1 Physical properties of MnOx(0.3)/TiO2(DP) and MnOx(0.3)/TiO2(WI). Catalyst
BET surface area (m2/g)
Pore volume (102 cm3/g)
MnOx(0.3)/TiO2(DP) MnOx(0.3)/TiO2(WI)
69 53
44 38
obvious signals of MnOx were detected at low Mn loadings. When the mole ratio of Mn/Ti reached 0.4, a weak band due to a-Mn2O3 could be seen. This implied that the supported MnOx were dispersed better on MnOx/TiO2(DP) catalysts, and the sintering took place more easily for the catalysts prepared by WI method. Since it was well known that the low crystallinity of active phase was favorable for catalytic reaction, these could partially account for the relatively higher activities of MnOx/TiO2(DP) series, and it could also explain why the catalysts with the highest Mn loadings for both series did not have the highest activities. As mentioned above, activities of these two catalyst series towards NO catalytic oxidation were quite different. The difference in crystallinity was affirmatively one of the causes. At the same time, the XRD results also showed that the compositions of supported MnOx species were somewhat different between the two series, and other properties of the catalysts would also have influences on the catalytic activity. Thus, the two optimal catalysts in each series, MnOx(0.3)/TiO2(DP) and MnOx(0.3)/TiO2(WI), were selected for further characterizations in the following to investigate the effect of the preparation method. 3.3. HRTEM and nitrogen adsorption–desorption results
Fig. 2. X-ray diffractograms for MnOx/TiO2 catalysts of different Mn/Ti ratios prepared by (a) DP method; (b) WI method. b-MnO2 (), a-Mn2O3 (s), Anatase TiO2 (d), Rutile TiO2 (}).
indicated that the preparation method influenced the activities of MnOx/TiO2 nanoxides significantly, and DP method was favorable for improving the catalytic activity. 3.2. XRD results XRD patterns of the two series of MnOx/TiO2 catalysts are shown in Fig. 2. Crystalline phases of anatase and rutile associated with the TiO2 support could be clearly observed for all the samples. As for the MnOx/TiO2(WI) catalysts, apparent diffraction peaks of b-MnO2 could be observed, and their intensities increased rapidly with the Mn loading. While for the MnOx/TiO2(DP) series, no
Fig. 3 shows the HRTEM images of both MnOx(0.3)/TiO2(DP) and MnOx(0.3)/TiO2(WI) samples. Well-crystallized anatase titania (whose lattice fringe of (1 0 1) plane is 0.354 nm) and b-MnO2 (whose lattice fringe of (1 0 1) plane is 0.239 nm) could be seen in MnOx(0.3)/TiO2(WI), while no crystalline structure of MnOx was found in MnOx(0.3)/TiO2(DP). These were in good agreement with the XRD results, which implied that the dispersion of MnOx species was greatly improved by the DP method. As determined by nitrogen adsorption–desorption method, the specific surface area and pore volume for both samples are listed in Table 1. The specific surface area and pore volume of MnOx(0.3)/TiO2(DP) were higher than those of MnOx(0.3)/TiO2(WI), which was beneficial to the catalytic activity improvement. The pore size distributions of the two catalysts also showed different profiles (not shown here). The pore size distribution of MnOx(0.3)/TiO2(WI) gathered in the range of 15–40 nm, while that of MnOx(0.3)/TiO2(DP) was very broad and mainly in the range of 2– 15 nm. The smaller pore size, higher specific surface area and larger pore volume in MnOx(0.3)/TiO2(DP) could be attributed to the
Fig. 3. HRTEM images of (a) MnOx(0.3)/TiO2(DP) and (b) MnOx(0.3)/TiO2(WI).
146
Z. Wu et al. / Journal of Colloid and Interface Science 352 (2010) 143–148
Fig. 4. FTIR spectra of (a) MnOx(0.3)/TiO2(DP), (b) MnOx(0.3)/TiO2(WI) and (c) TiO2.
Fig. 5. Mn 2p XPS spectra of (a) MnOx(0.3)/TiO2(DP) and (b) MnOx(0.3)/TiO2(WI).
restraint of MnOx crystallite growth and lower degree of sintering as a result of DP method. Combining the abovementioned characterization results of XRD, HRTEM and nitrogen adsorption–desorption, it could be seen that larger MnOx particles would be formed in WI method, while their growth was well controlled during DP process. The resulting highly dispersed MnOx species could contribute to the high catalytic activity of MnOx(0.3)/TiO2(DP) in NO catalytic oxidation.
641.3–641.7 eV [26] and Mn4+ at 642.2–643 eV [27], respectively, and the relative intensities were obtained. It could be seen that there was more Mn element in trivalence in MnOx(0.3)/TiO2(DP) compared with MnOx(0.3)/TiO2(WI). This was in accordance with the FTIR results. Moreover, the Mn/Ti ratio of MnOx(0.3)/TiO2(DP) was remarkably higher than that of MnOx(0.3)/TiO2(WI), even higher than the theoretical value of 0.3. This indicated that DP method could enrich MnOx species on the surface and abundant active sites might be available for the catalytic oxidation process.
3.4. FTIR spectroscopy and XPS analysis 3.5. H2-TPR and NO-TPD analysis Since MnOx species in MnOx(0.3)/TiO2(DP) were highly dispersed and could not be detected by XRD and HRTEM, FTIR investigation was employed to get more information on the supported MnOx species in MnOx(0.3)/TiO2(DP). The FTIR spectra of MnOx(0.3)/TiO2(WI) and TiO2 support were also acquired for comparison, as shown in Fig. 4. The spectra of TiO2 agreed well with that reported in the literature [23], while obvious differences could be observed within the low wavenumbers region from 400 to 800 cm1 for all the three samples. For MnOx(0.3)/TiO2(WI), a weak band at 611 cm1 appeared, which could be one of the main characteristic bands of b-MnO2 [24]. This was also supported by the XRD and HRTEM results. As for the MnOx(0.3)/TiO2(DP) sample, two bands located at 579 cm1 and 527 cm1 were clearly observed, which could be attributed to the vibrational features of a-Mn2O3 [24]. Therefore, the MnOx species in MnOx(0.3)/TiO2(DP) sample might mainly exist in the form of Mn2O3. This conclusion could be supported by the XRD results (Fig. 2a) that fine crystal of a-Mn2O3 was also detected for MnOx/TiO2(DP) at a high Mn/Ti mole ratio of 0.4. The addition of carbonate in DP method could generate MnCO3 and subsequently facilitate the formation of Mn2O3 during calcination [25]. In order to identify the oxidation states of surface species, both catalysts were examined by XPS spectroscopy. Surface atomic concentrations of Mn, Ti, O are listed in Table 2 and the photoelectron spectra of Mn 2p are displayed in Fig. 5. The Mn 2p profiles could be deconvoluted into two contributions referred to Mn3+ at
Fig. 6 presents H2-TPR profiles of the two MnOx(0.3)/TiO2 catalysts, and both samples showed two distinct reduction peaks. The two peaks for each sample could be attributed to the successive reduction steps of MnO2 or Mn2O3 ? Mn3O4 and Mn3O4 ? MnO, respectively [28]. For MnOx(0.3)/TiO2(WI), the H2 reduction started at about 230 °C, and the two overlapping peaks centered at 326 °C and 378 °C, respectively. While for MnOx(0.3)/TiO2(DP), the reduction initiated at a much lower temperature of approximate 140 °C, and the maximum positions of the two peaks located at 260 °C and 367 °C, respectively. The shift of reduction temperature to a lower temperature reflected the increase of oxygen mobility, and this might be related to the higher catalytic activity of MnOx(0.3)/ TiO2(DP) within the low temperature range. Higher dispersion of supported MnOx species in MnOx(0.3)/TiO2(DP) could contribute to the oxygen mobility improvement in comparison with MnOx(0.3)/ TiO2(WI) [12].
Table 2 XPS results of MnOx(0.3)/TiO2(DP) and MnOx(0.3)/TiO2(WI). Catalyst
MnOx(0.3)/TiO2(DP) MnOx(0.3)/TiO2(WI)
Surface atomic concentrations (%)
Surface atomic ratio
Mn
Ti
O
Mn/Ti
Mn3+/Mn
8.9 6.5
21.3 21.6
69.8 71.9
0.42 0.30
0.63 0.53
Fig. 6. H2-TPR profiles of (a) MnOx(0.3)/TiO2(DP) and (b) MnOx(0.3)/TiO2(WI).
Z. Wu et al. / Journal of Colloid and Interface Science 352 (2010) 143–148
147
Fig. 7. NO-TPD profiles of (a) MnOx(0.3)/TiO2(DP) and (b) MnOx(0.3)/TiO2(WI).
Moreover, the reduction peak areas were also calculated to estimate the H2 consumptions. Obviously, the MnOx(0.3)/TiO2(DP) sample consumed less H2 than MnOx(0.3)/TiO2(WI), and the area ratio of the first to the second reduction peak for MnOx(0.3)/ TiO2(DP) was about 0.9, which was remarkably lower than the ratio of 1.5 for MnOx(0.3)/TiO2(WI). These indicated that the average oxidation state of Mn in MnOx(0.3)/TiO2(DP) was lower than that in MnOx(0.3)/TiO2(WI) [29], and the dominant MnOx species in MnOx(0.3)/TiO2(DP) appeared to be Mn2O3, while that in MnOx(0.3)/TiO2(WI) was MnO2. These were in consistence with the FTIR results. Considering the much higher catalytic activity of MnOx(0.3)/TiO2(DP), it might be reasonable to speculate that Mn3+ was more favorable than Mn4+ for NO catalytic oxidation. Similar findings were reported by Atribak et al., and they found that pure Mn2O3 was more active than pure MnO2 in NO catalytic oxidation [30]. Cimino et al. attributed the higher activity of Mn3+ than Mn4+ in CO catalytic oxidation to the weaker Mn3+–O bond [31]. Similarly, the lower strength of Mn3+–O bonds would also favor the catalytic oxidation of NO, since the Mn3+–O bond was easy to break, which could promote the generation and release of the oxidation product NO2. Fig. 7 shows the NO-TPD curves for MnOx(0.3)/TiO2(DP) and MnOx(0.3)/TiO2(WI). One broad desorption band was spanned in the temperature range of 50–300 °C for both samples, which could be attributed to the nitrogen-containing species desorbed from active sites. In the case of MnOx(0.3)/TiO2(DP), the area of the desorption band was remarkably larger especially in the low temperature range, indicating that the NO adsorption capacity on MnOx(0.3)/ TiO2(DP) was much higher and the adsorbed species could be desorbed more easily. These also reflected the higher activity of MnOx(0.3)/TiO2(DP) towards NO catalytic oxidation at low temperature. As stated above, DP method facilitated the high dispersion and enrichment of MnOx on the surface, and thus more active sites were created in MnOx(0.3)/TiO2(DP) catalyst, which could be one of the reasons of its higher NO adsorption capacity. As for the more remarkable desorption band at low temperature, it supported the assumption raised above that the lower bond strength between oxygen and Mn3+ would facilitate the release of NO or its oxidation product NO2. 3.6. DRIFT spectra The DRIFT spectra for the co-adsorption of NO and O2 followed by He purge on both samples are shown in Fig. 8. As displayed in Fig. 8a for MnOx(0.3)/TiO2(DP), many bands at 1603, 1581, 1558, 1544, 1521, 1269 and 1244 cm1 were detected after the introduction of NO and O2 as well as some weak bands in the range of
Fig. 8. DRIFT spectra of co-adsorption of NO and O2 followed by He purge on (a) MnOx(0.3)/TiO2(DP) and (b) MnOx(0.3)/TiO2(WI).
1700–2000 cm1 (1913 and 1843 cm1). All the bands intensities increased with time and stabilized at about 60 min. After subsequent purging by He, all the bands decreased rapidly. According to the literatures [32,33], the bands at 1581, 1558 and 1544 cm1 could be assigned to bidentate nitrate; the bands at 1603 and 1244 cm1 could be attributed to bridged nitrate; the bands at 1521 and 1269 cm1 were typical of monodentate nitrate; the weak bands at 1913 and 1843 cm1 were associated with nitrosyls. Especially, the band in the 1625–1600 cm1 region shifted to higher wavenumbers with the increase of its intensity and stabilized at about 1621 cm1, and then it shifted to lower wavenumbers after He purge. As reported by literatures [33], there were two bands at 1603 and 1625 cm1 over MnOx/TiO2 sample which could be assigned to bridged nitrate and adsorbed NO2, respectively. Therefore, the band in the region of 1625–1600 cm1 might be the overlapping two bands of 1603 and 1625 cm1. The shift of the band to higher wavenumbers may indicate the fact of the more rapid growth of adsorbed NO2 compared to the nitrate species during co-adsorption of NO and O2, and its shift to lower wavenumbers during He purge suggested the easy release of adsorbed NO2 from catalyst surface. Since no other oxidation products other than NO2 were observed in the activity test, the intensities decease of all bands implied that these nitrogen-containing species may be the main intermediates of the NO catalytic oxidation reaction, which would decompose to the final oxidation product NO2. As shown in Fig. 8b, the evolution of DRIFT spectra for MnOx(0.3)/TiO2(WI) was similar to that for MnOx(0.3)/TiO2(DP). But the corresponding band in the 1625–1600 cm1 region was stabilized at about 1614 cm1, which was lower than that on
148
Z. Wu et al. / Journal of Colloid and Interface Science 352 (2010) 143–148
in MnOx(0.3)/TiO2(DP) was highly dispersed and enriched on the surface, producing more active sites. Moreover, FTIR, XPS and H2TPR results indicated that more Mn3+ species was created by DP method, which could be another reason for its higher catalytic activity. The positive effects of both higher MnOx dispersion and the abundance of Mn3+ species were further supported by NOTPD profiles. Furthermore, the DRIFT studies implied that the oxidation product NO2 may come from the decomposition of nitrogen-containing species generated during the co-adsorption NO and O2, and the decomposition seemed to be more easy over MnOx(0.3)/TiO2(DP). The high persistence of the catalyst further confirmed that the nitrogen-containing species were the main reaction intermediates. Acknowledgments Fig. 9. The effect of reaction time on the NO catalytic activity of MnOx(0.3)/ TiO2(DP). Reaction conditions: 250 °C; 600 ppm NO, 4% O2, and balance N2; flow rate = 1.2 L/min; catalyst 0.10 g.
MnOx(0.3)/TiO2(DP) (1621 cm1). This also reflected that the nitrogen-containing species over MnOx(0.3)/TiO2(DP) produced by the co-adsorption NO and O2 would be decomposed to NO2 more easily. And it could be confirmed by the above NO-TPD results that much more remarkable desorption band at low temperature over MnOx(0.3)/TiO2(DP) was observed. 3.7. Long-time running test As mentioned above, the NO catalytic oxidation product NO2 could come from the decomposition of nitrogen-containing species which were generated during the co-adsorption of NO and O2 on the catalysts. To further validate this conclusion, activity test with a high space velocity (0.1 g catalyst used) over MnOx(0.3)/TiO2(DP) at 250 °C was carried out for a relative long running time of 20 h. As shown in Fig. 9, no noticeable decrease of NO conversion was observed in 20 h, which suggested that the high activity of MnOx(0.3)/TiO2(DP) was persistent. Therefore, irreversible deposition of nitrogen-containing species seemed unlikely, and these further confirmed that the catalytic oxidation of NO performed via the pathway of the initial formation of nitrogen-containing species followed by their decomposition to NO2. 4. Summary In this study, a series of MnOx/TiO2 composite nanoxides were prepared by DP method and compared with those by WI method for catalytic oxidation of NO to NO2. The optimal Mn/Ti ratio was 0.3 for both series, and MnOx(0.3)/TiO2(DP) showed much higher activity (89% at 250 °C) than MnOx(0.3)/TiO2(WI) (69% at 330 °C) with a GHSV of 25,000 h1. Compared to MnOx(0.3)/TiO2(WI), MnOx(0.3)/TiO2(DP) showed a higher specific surface area, lower reduction temperature in H2TPR profiles and higher surface Mn/Ti ratio from XPS results, and no crystalline structure of MnOx was found from XRD and HRTEM characterizations. All these aspects implied that the MnOx species
This work was financially supported by the National HighTech Research and Development Program (863) of China (2007AA061701), Changjiang Scholar Incentive Program, Ministry of Education, China (2009), and Science Foundation of Zhejiang Province (Y5090053). References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33]
Z.H. Zhu, L.R. Radovic, G.Q. Lu, Carbon 38 (2000) 451. Z.B. Wu, R.B. Jin, Y. Liu, H.Q. Wang, Catal. Commun. 9 (2008) 2217. I. Spasova, P. Nikolov, D. Mehandjiev, J. Colloid Interface Sci. 290 (2005) 343. M. Devadas, O. Kröcher, M. Elsener, A. Wokaun, N. Söger, M. Pfeifer, Y. Demel, L. Mussmann, Appl. Catal., B 67 (2006) 187. M. Kang, D.J. Kim, E.D. Park, J.M. Kim, J.E. Yie, S.H. Kim, L. Hope-Weeks, E.M. Eyring, Appl. Catal., B 68 (2006) 21. D. Thomas, J. Vanderschuren, Chem. Eng. Technol. 23 (2000) 449. H. Takeuchi, M. Ando, N. Kizawa, Ind. Eng. Chem. Process. Des. Dev. 16 (1977) 303. J. Despres, M. Elsener, M. Koebel, O. Krocher, B. Schnyder, A. Wokaun, Appl. Catal., B 50 (2004) 73. L.D. Li, Q. Shen, J. Cheng, Z.P. Hao, Appl. Catal., B 93 (2010) 259. D. Bhatia, R.W. McCabe, M.P. Harold, V. Balakotaiah, J. Catal. 266 (2009) 106. Y.X. Wen, C.B. Zhang, H. He, Y.B. Yu, Y. Teraoka, Catal. Today 126 (2007) 400. M.M. Yung, E.M. Holmgreen, U.S. Ozkan, J. Catal. 247 (2007) 356. Q. Wang, S.Y. Park, J.S. Choi, J.S. Chung, Appl. Catal., B 79 (2008) 101. H. Li, X.L. Tang, H.H. Yi, L.L. Yu, J. Rare Earths 28 (2010) 64. H. Li, X.L. Tang, H.H. Yi, L.L. Yu, Asia-Pacific Power and Energy Engineering Conference, Wuhan, 2009, doi: 10.1109/APPEEC.2009.4918472. G. Qi, R.T. Yang, Appl. Catal., B 44 (2003) 217. J.H. Huang, Z.Q. Tong, Y. Huang, J.F. Zhang, Appl. Catal., B 78 (2008) 309. B.Q. Jiang, Y. Liu, Z.B. Wu, J. Hazard. Mater. 162 (2009) 1249. Z.B. Wu, R.B. Jin, H.Q. Wang, Y. Liu, Catal. Commun. 10 (2009) 935. M.K. Van, A.J. der Lee, J.H. Van Dillen, K.P. Bitter, De. Jong, J. Am. Chem. Soc. 127 (2005) 13573. W.C. Li, M. Comotti, F. Schuth, J. Catal. 237 (2006) 190. I. Puente-Lee, R. Nares, J. Ramirez, P.S. Schabes-Retchkiman, Rev. Mex. Fis. 50 (S1) (2004) 69. H. Jensen, A. Soloviev, Z. Li, E.G. Sogaard, Appl. Surf. Sci. 246 (2005) 239. C.M. Julien, M. Massot, C. Poinsignon, Spectrochim. Acta, Part A 60 (2004) 689. K. Ramesh, L.W. Chen, F.X. Chen, Z.Y. Zhong, J.H. Chin, H.W. Mook, Y.F. Han, Catal. Commun. 8 (2007) 1421. M.R. Morales, B.P. Barbero, L.E. Cadus, Appl. Catal., B 67 (2006) 229. D. Delimaris, T. Ioannides, Appl. Catal., B 84 (2008) 303. J. Trawczynski, B. Bielak, W. Mista, Appl. Catal., B 55 (2005) 277. H.Y. Chen, A. Sayari, A. Adnot, F. Larachi, Appl. Catal., B 32 (2001) 195. I. Atribak, A. Bueno-Lopez, A. Garcia-Garcia, P. Navarro, D. Frias, M. Montes, Appl. Catal., B 93 (2010) 267. A. Cimino, V. Indovina, J. Catal. 33 (1974) 493. M. Kantcheva, J. Catal. 204 (2001) 479. Z.B. Wu, B.Q. Jiang, Y. Liu, H.Q. Wang, R.B. Jin, Environ. Sci. Technol. 41 (2007) 5812.
Contents lists available at ScienceDirect
Journal of Colloid and Interface Science www.elsevier.com/locate/jcis
MnOx/TiO2 composite nanoxides synthesized by deposition-precipitation method as a superior catalyst for NO oxidation Zhongbiao Wu, Nian Tang, Ling Xiao, Yue Liu ⇑, Haiqiang Wang Department of Environmental Engineering, Zhejiang University, Hangzhou 310027, China Zhejiang Provincial Engineering Research Center of Industrial Boiler & Furnace Flue Gas Pollution Control, Hangzhou 310027, China
a r t i c l e
i n f o
Article history: Received 1 June 2010 Accepted 10 August 2010 Available online 13 August 2010 Keywords: NO catalytic oxidation NO2 MnOx/TiO2 composite nanoxides Deposition-precipitation method
a b s t r a c t A series of MnOx/TiO2 composite nanoxides were prepared by deposition-precipitation (DP) method, and the sample with the Mn/Ti ratio of 0.3 showed a superior activity for NO catalytic oxidation to NO2. The maximum NO conversion over MnOx(0.3)/TiO2(DP) could reach 89% at 250 °C with a GHSV of 25,000 h1, which was much higher than that over the catalyst prepared by conventional wet-impregnation (WI) method (69% at 330 °C). Characterization results including XRD, HRTEM, FTIR, XPS, H2-TPR, NO-TPD and Nitrogen adsorption–desorption implied that the higher activity of MnOx(0.3)/TiO2(DP) could be attributed to the enrichment of well-dispersed MnOx on the surface and the abundance of Mn3+ species. Furthermore, DRIFT investigations and long-time running test indicated that NO2 came from the decomposition of adsorbed nitrogen-containing species. Ó 2010 Elsevier Inc. All rights reserved.
1. Introduction Development of methods to control NOx emissions has been an active research area [1–3] because NOx are associated with acid deposition, photochemical smog and ozone depletion. NO2 is always more favored than NO for NOx conversion among several procedures. It was reported that the oxidation of NO–NO2 would definitely increase the NO conversion rate in the process of selective catalytic reduction (SCR) [4]. And this conclusion inspired Kang et al. to develop a high-performance two-stage catalyst system for SCR of NOx with the pre-oxidation of NO during the first stage [5]. For wet scrubbing method, the solubility and reactivity of NOx could also be improved significantly when NO is partially or completely converted to NO2 [6,7]. Therefore, much effort has been focused on developing eligible catalysts for NO oxidation [8–15]. The Pt-based catalysts have been most widely used for the NO oxidation [8–10] but were limited by their high cost. Metal oxides catalysts could be appropriate alternatives to the noble metal catalysts due to their potential high catalytic efficiency, low cost and persistent activity. Among them, Co-based catalysts attracted the most attention [11–13], while their applications were retarded by the toxicity of Cobalt. Nowadays, Mn-based catalysts, acting as low-temperature SCR catalysts [16–19], also showed certain ⇑ Corresponding author at: Department of Environmental Engineering, Zhejiang University, Hangzhou 310027, China. Fax: +86 571 87953088. E-mail address: [email protected] (Y. Liu). 0021-9797/$ - see front matter Ó 2010 Elsevier Inc. All rights reserved. doi:10.1016/j.jcis.2010.08.031
catalytic activity towards NO oxidation in the SCR process [16,17]. Very recently, Mn-based catalysts were directly employed as the NO oxidation catalysts and showed their potential in this field [14,15]. However, the NO conversions over these catalysts were not very satisfactory and the detail understanding was still lacking for NO oxidation over Mn-based catalysts. In this study, the main objective was to seek a Mn-based catalyst with high activity towards NO oxidation, and deposition-precipitation (DP) method was introduced to synthesize MnOx/TiO2 composite nanoxides as the catalyst. DP method involves the precipitation of a metal precursor onto a suspended support, and it could make the metal oxides highly loaded and enriched in the external surface [20,21]. The metal oxides particles could also be prevented from sintering during calcination by DP method and thus well-dispersed on the support [22]. For comparison, other catalysts with the same compositions were also prepared by conventional wet-impregnation (WI) method. 2. Materials and methods 2.1. Catalysts preparation The MnOx/TiO2 catalysts were prepared as follows: (a) DP method: 60 ml solution containing certain amount of Mn(NO3)2 and a mixed solution of NH3H2O (15 wt.%) and NH4HCO3 (15 wt.%) were added simultaneously and dropwise to a slurry solution (80 ml) containing 8 g commercial
144
Z. Wu et al. / Journal of Colloid and Interface Science 352 (2010) 143–148
TiO2 support (Degussa P25) under vigorous stirring and a constant pH value of 10 was maintained. The precipitate was aged at 50 °C for 2 h in the mother solution. After filtered and washed with deionized water for five times, the obtained sample was dried at 110 °C for 12 h and then calcined at 450 °C for 4 h in air. (b) WI method: the preparation process followed the typical WI method (not adding the mixed solution of NH3H2O and NH4HCO3 in the abovementioned procedure). The composition and calcination temperature were the same as those of DP method. 2.2. Catalytic activity tests Steady-state reaction experiments were performed in a fixedbed at 90–450 °C typically containing 1.6 g (2.8 ml) catalyst with a gas hourly space velocity (GHSV) of 25,000 h1. The reactant gas typically consisted of 600 ppm NO, 4% O2 and balance N2. The concentrations of NO and NO2 were monitored by a chemiluminescence NO–NO–NOx analyzer (Thermo 42i-HL). Since a balance of total NOx (NO + NO2) between the inlet and the outlet gases could be obtained for each experiment, the NO conversion ([Conv.]NO) was defined as:
½Conv:NO ¼ ½NO2 out =ð½NO2 out þ ½NOout Þ 100%
3. Results and discussion 3.1. Catalytic activity tests The conversion of NO–NO2 over a series of MnOx/TiO2 nanoxides with different Mn/Ti ratios prepared by DP method and WI method are shown in Fig. 1a and b, respectively. The thermodynamic equilibrium for NO–NO2 under given conditions is also shown for reference (in dashed). As expected, the reaction was kinetically limited at low temperature. Above temperatures at which the equilibrium curve was reached, the catalytic activities closely followed the equilibrium conversion and became thermodynamic control. The balance of total NOx (NO + NO2) between the inlet and the outlet gas could be obta, and no other nitrogen-containing product was detected in the outlet gas. It could be seen that the TiO2 support showed negligible catalytic activity during the entire temperature range and the introduction of MnOx improved the activities significantly. In both cases, the optimal Mn/Ti ratio appeared to be 0.3, and the corresponding samples were denoted as MnOx(0.3)/ TiO2(DP) and MnOx(0.3)/TiO2(WI), respectively. Moreover, MnOx(0.3)/TiO2(DP) exhibited much higher catalytic activity than MnOx(0.3)/TiO2(WI) towards NO oxidation, and it reached the equilibrium curve at about 250 °C, with the maximum NO conversion of 89%, which was comparable to the activities of Pt-based catalysts [10]. In contrast, the conversion of NO at the same temperature was only about 40% for MnOx(0.3)/TiO2(WI), which required a temperature of about 330 °C to reach the equilibrium, with the maximum NO conversion of only 69%. These
2.3. Characterization The crystal phase was analyzed by X-ray diffraction with Cu Ka radiation (XRD: model D/max RA, Rigaku Co., Japan). High-resolution transmission electron microscope (HRTEM: JOEL JEM-2010, Japan) was used to investigate the microstructures and morphologies of the prepared samples. Infrared absorption spectra were recorded at room temperature using a Fourier transform interferometer (FTIR: Shimadzu IRAffinity-1, Japan). Nitrogen adsorption–desorption isotherms were obtained at 77 K (Micromeritics ASAP 2020, USA). Then the specific surface areas of samples were measured using Brunauer–Emmett–Teller (BET) method, and the pore volume and pore size distributions were calculated by Barrett–Jioner–Halenda (BJH) method. X-ray photoelectron spectroscopy with Al Ka X-rays (hm = 1486.6 eV) radiation (XPS: Thermal ESCALAB 250, USA) was used to analyze the surface atomic states of the catalysts. Temperature programmed reduction (TPR) and temperature programmed desorption (TPD) tests were carried out on a custom-made TCD setup using 50 mg catalysts. Prior to H2-TPR experiments, samples were pretreated in nitrogen at 400 °C for 1 h and then cooled down to 100 °C. The reduction was conducted from 100 to 800 °C at a linear ramp of 5 °C min1 in 6% H2/N2 with a flow rate of 30 ml/min. For NO-TPD experiments, after pretreated in He at 400 °C for 1 h and cooled down to 30 °C, catalysts were saturated with 30 ml/min NO (4% in He) and 6% O2 for 20 min. After flushed with He for 1 h, desorption was carried out in 30 ml/min He from 30 to 400 °C with a heating rate of 5 °C min1. In situ diffuse reflectance infrared transform spectroscopy (DRIFT) investigations were performed with Nicolet 6700 FTIR spectrometers at 4 cm1 resolution with 64 co-added scans. In DRIFT cell, samples were pretreated at 400 °C in He for 2 h, and backgrounds were recorded after cooled to 250 °C. At this temperature co-adsorption of 5000 ppm NO and 10% O2 in balance He was performed for 60 min, after which the sample was then flushed with He for 30 min.
Fig. 1. NO conversion over MnOx/TiO2 catalysts of different Mn/Ti ratios prepared by (a) DP method; (b) WI method. Reaction conditions: 600 ppm NO, 4% O2 and balance N2; GHSV = 25,000 h1.
145
Z. Wu et al. / Journal of Colloid and Interface Science 352 (2010) 143–148 Table 1 Physical properties of MnOx(0.3)/TiO2(DP) and MnOx(0.3)/TiO2(WI). Catalyst
BET surface area (m2/g)
Pore volume (102 cm3/g)
MnOx(0.3)/TiO2(DP) MnOx(0.3)/TiO2(WI)
69 53
44 38
obvious signals of MnOx were detected at low Mn loadings. When the mole ratio of Mn/Ti reached 0.4, a weak band due to a-Mn2O3 could be seen. This implied that the supported MnOx were dispersed better on MnOx/TiO2(DP) catalysts, and the sintering took place more easily for the catalysts prepared by WI method. Since it was well known that the low crystallinity of active phase was favorable for catalytic reaction, these could partially account for the relatively higher activities of MnOx/TiO2(DP) series, and it could also explain why the catalysts with the highest Mn loadings for both series did not have the highest activities. As mentioned above, activities of these two catalyst series towards NO catalytic oxidation were quite different. The difference in crystallinity was affirmatively one of the causes. At the same time, the XRD results also showed that the compositions of supported MnOx species were somewhat different between the two series, and other properties of the catalysts would also have influences on the catalytic activity. Thus, the two optimal catalysts in each series, MnOx(0.3)/TiO2(DP) and MnOx(0.3)/TiO2(WI), were selected for further characterizations in the following to investigate the effect of the preparation method. 3.3. HRTEM and nitrogen adsorption–desorption results
Fig. 2. X-ray diffractograms for MnOx/TiO2 catalysts of different Mn/Ti ratios prepared by (a) DP method; (b) WI method. b-MnO2 (), a-Mn2O3 (s), Anatase TiO2 (d), Rutile TiO2 (}).
indicated that the preparation method influenced the activities of MnOx/TiO2 nanoxides significantly, and DP method was favorable for improving the catalytic activity. 3.2. XRD results XRD patterns of the two series of MnOx/TiO2 catalysts are shown in Fig. 2. Crystalline phases of anatase and rutile associated with the TiO2 support could be clearly observed for all the samples. As for the MnOx/TiO2(WI) catalysts, apparent diffraction peaks of b-MnO2 could be observed, and their intensities increased rapidly with the Mn loading. While for the MnOx/TiO2(DP) series, no
Fig. 3 shows the HRTEM images of both MnOx(0.3)/TiO2(DP) and MnOx(0.3)/TiO2(WI) samples. Well-crystallized anatase titania (whose lattice fringe of (1 0 1) plane is 0.354 nm) and b-MnO2 (whose lattice fringe of (1 0 1) plane is 0.239 nm) could be seen in MnOx(0.3)/TiO2(WI), while no crystalline structure of MnOx was found in MnOx(0.3)/TiO2(DP). These were in good agreement with the XRD results, which implied that the dispersion of MnOx species was greatly improved by the DP method. As determined by nitrogen adsorption–desorption method, the specific surface area and pore volume for both samples are listed in Table 1. The specific surface area and pore volume of MnOx(0.3)/TiO2(DP) were higher than those of MnOx(0.3)/TiO2(WI), which was beneficial to the catalytic activity improvement. The pore size distributions of the two catalysts also showed different profiles (not shown here). The pore size distribution of MnOx(0.3)/TiO2(WI) gathered in the range of 15–40 nm, while that of MnOx(0.3)/TiO2(DP) was very broad and mainly in the range of 2– 15 nm. The smaller pore size, higher specific surface area and larger pore volume in MnOx(0.3)/TiO2(DP) could be attributed to the
Fig. 3. HRTEM images of (a) MnOx(0.3)/TiO2(DP) and (b) MnOx(0.3)/TiO2(WI).
146
Z. Wu et al. / Journal of Colloid and Interface Science 352 (2010) 143–148
Fig. 4. FTIR spectra of (a) MnOx(0.3)/TiO2(DP), (b) MnOx(0.3)/TiO2(WI) and (c) TiO2.
Fig. 5. Mn 2p XPS spectra of (a) MnOx(0.3)/TiO2(DP) and (b) MnOx(0.3)/TiO2(WI).
restraint of MnOx crystallite growth and lower degree of sintering as a result of DP method. Combining the abovementioned characterization results of XRD, HRTEM and nitrogen adsorption–desorption, it could be seen that larger MnOx particles would be formed in WI method, while their growth was well controlled during DP process. The resulting highly dispersed MnOx species could contribute to the high catalytic activity of MnOx(0.3)/TiO2(DP) in NO catalytic oxidation.
641.3–641.7 eV [26] and Mn4+ at 642.2–643 eV [27], respectively, and the relative intensities were obtained. It could be seen that there was more Mn element in trivalence in MnOx(0.3)/TiO2(DP) compared with MnOx(0.3)/TiO2(WI). This was in accordance with the FTIR results. Moreover, the Mn/Ti ratio of MnOx(0.3)/TiO2(DP) was remarkably higher than that of MnOx(0.3)/TiO2(WI), even higher than the theoretical value of 0.3. This indicated that DP method could enrich MnOx species on the surface and abundant active sites might be available for the catalytic oxidation process.
3.4. FTIR spectroscopy and XPS analysis 3.5. H2-TPR and NO-TPD analysis Since MnOx species in MnOx(0.3)/TiO2(DP) were highly dispersed and could not be detected by XRD and HRTEM, FTIR investigation was employed to get more information on the supported MnOx species in MnOx(0.3)/TiO2(DP). The FTIR spectra of MnOx(0.3)/TiO2(WI) and TiO2 support were also acquired for comparison, as shown in Fig. 4. The spectra of TiO2 agreed well with that reported in the literature [23], while obvious differences could be observed within the low wavenumbers region from 400 to 800 cm1 for all the three samples. For MnOx(0.3)/TiO2(WI), a weak band at 611 cm1 appeared, which could be one of the main characteristic bands of b-MnO2 [24]. This was also supported by the XRD and HRTEM results. As for the MnOx(0.3)/TiO2(DP) sample, two bands located at 579 cm1 and 527 cm1 were clearly observed, which could be attributed to the vibrational features of a-Mn2O3 [24]. Therefore, the MnOx species in MnOx(0.3)/TiO2(DP) sample might mainly exist in the form of Mn2O3. This conclusion could be supported by the XRD results (Fig. 2a) that fine crystal of a-Mn2O3 was also detected for MnOx/TiO2(DP) at a high Mn/Ti mole ratio of 0.4. The addition of carbonate in DP method could generate MnCO3 and subsequently facilitate the formation of Mn2O3 during calcination [25]. In order to identify the oxidation states of surface species, both catalysts were examined by XPS spectroscopy. Surface atomic concentrations of Mn, Ti, O are listed in Table 2 and the photoelectron spectra of Mn 2p are displayed in Fig. 5. The Mn 2p profiles could be deconvoluted into two contributions referred to Mn3+ at
Fig. 6 presents H2-TPR profiles of the two MnOx(0.3)/TiO2 catalysts, and both samples showed two distinct reduction peaks. The two peaks for each sample could be attributed to the successive reduction steps of MnO2 or Mn2O3 ? Mn3O4 and Mn3O4 ? MnO, respectively [28]. For MnOx(0.3)/TiO2(WI), the H2 reduction started at about 230 °C, and the two overlapping peaks centered at 326 °C and 378 °C, respectively. While for MnOx(0.3)/TiO2(DP), the reduction initiated at a much lower temperature of approximate 140 °C, and the maximum positions of the two peaks located at 260 °C and 367 °C, respectively. The shift of reduction temperature to a lower temperature reflected the increase of oxygen mobility, and this might be related to the higher catalytic activity of MnOx(0.3)/ TiO2(DP) within the low temperature range. Higher dispersion of supported MnOx species in MnOx(0.3)/TiO2(DP) could contribute to the oxygen mobility improvement in comparison with MnOx(0.3)/ TiO2(WI) [12].
Table 2 XPS results of MnOx(0.3)/TiO2(DP) and MnOx(0.3)/TiO2(WI). Catalyst
MnOx(0.3)/TiO2(DP) MnOx(0.3)/TiO2(WI)
Surface atomic concentrations (%)
Surface atomic ratio
Mn
Ti
O
Mn/Ti
Mn3+/Mn
8.9 6.5
21.3 21.6
69.8 71.9
0.42 0.30
0.63 0.53
Fig. 6. H2-TPR profiles of (a) MnOx(0.3)/TiO2(DP) and (b) MnOx(0.3)/TiO2(WI).
Z. Wu et al. / Journal of Colloid and Interface Science 352 (2010) 143–148
147
Fig. 7. NO-TPD profiles of (a) MnOx(0.3)/TiO2(DP) and (b) MnOx(0.3)/TiO2(WI).
Moreover, the reduction peak areas were also calculated to estimate the H2 consumptions. Obviously, the MnOx(0.3)/TiO2(DP) sample consumed less H2 than MnOx(0.3)/TiO2(WI), and the area ratio of the first to the second reduction peak for MnOx(0.3)/ TiO2(DP) was about 0.9, which was remarkably lower than the ratio of 1.5 for MnOx(0.3)/TiO2(WI). These indicated that the average oxidation state of Mn in MnOx(0.3)/TiO2(DP) was lower than that in MnOx(0.3)/TiO2(WI) [29], and the dominant MnOx species in MnOx(0.3)/TiO2(DP) appeared to be Mn2O3, while that in MnOx(0.3)/TiO2(WI) was MnO2. These were in consistence with the FTIR results. Considering the much higher catalytic activity of MnOx(0.3)/TiO2(DP), it might be reasonable to speculate that Mn3+ was more favorable than Mn4+ for NO catalytic oxidation. Similar findings were reported by Atribak et al., and they found that pure Mn2O3 was more active than pure MnO2 in NO catalytic oxidation [30]. Cimino et al. attributed the higher activity of Mn3+ than Mn4+ in CO catalytic oxidation to the weaker Mn3+–O bond [31]. Similarly, the lower strength of Mn3+–O bonds would also favor the catalytic oxidation of NO, since the Mn3+–O bond was easy to break, which could promote the generation and release of the oxidation product NO2. Fig. 7 shows the NO-TPD curves for MnOx(0.3)/TiO2(DP) and MnOx(0.3)/TiO2(WI). One broad desorption band was spanned in the temperature range of 50–300 °C for both samples, which could be attributed to the nitrogen-containing species desorbed from active sites. In the case of MnOx(0.3)/TiO2(DP), the area of the desorption band was remarkably larger especially in the low temperature range, indicating that the NO adsorption capacity on MnOx(0.3)/ TiO2(DP) was much higher and the adsorbed species could be desorbed more easily. These also reflected the higher activity of MnOx(0.3)/TiO2(DP) towards NO catalytic oxidation at low temperature. As stated above, DP method facilitated the high dispersion and enrichment of MnOx on the surface, and thus more active sites were created in MnOx(0.3)/TiO2(DP) catalyst, which could be one of the reasons of its higher NO adsorption capacity. As for the more remarkable desorption band at low temperature, it supported the assumption raised above that the lower bond strength between oxygen and Mn3+ would facilitate the release of NO or its oxidation product NO2. 3.6. DRIFT spectra The DRIFT spectra for the co-adsorption of NO and O2 followed by He purge on both samples are shown in Fig. 8. As displayed in Fig. 8a for MnOx(0.3)/TiO2(DP), many bands at 1603, 1581, 1558, 1544, 1521, 1269 and 1244 cm1 were detected after the introduction of NO and O2 as well as some weak bands in the range of
Fig. 8. DRIFT spectra of co-adsorption of NO and O2 followed by He purge on (a) MnOx(0.3)/TiO2(DP) and (b) MnOx(0.3)/TiO2(WI).
1700–2000 cm1 (1913 and 1843 cm1). All the bands intensities increased with time and stabilized at about 60 min. After subsequent purging by He, all the bands decreased rapidly. According to the literatures [32,33], the bands at 1581, 1558 and 1544 cm1 could be assigned to bidentate nitrate; the bands at 1603 and 1244 cm1 could be attributed to bridged nitrate; the bands at 1521 and 1269 cm1 were typical of monodentate nitrate; the weak bands at 1913 and 1843 cm1 were associated with nitrosyls. Especially, the band in the 1625–1600 cm1 region shifted to higher wavenumbers with the increase of its intensity and stabilized at about 1621 cm1, and then it shifted to lower wavenumbers after He purge. As reported by literatures [33], there were two bands at 1603 and 1625 cm1 over MnOx/TiO2 sample which could be assigned to bridged nitrate and adsorbed NO2, respectively. Therefore, the band in the region of 1625–1600 cm1 might be the overlapping two bands of 1603 and 1625 cm1. The shift of the band to higher wavenumbers may indicate the fact of the more rapid growth of adsorbed NO2 compared to the nitrate species during co-adsorption of NO and O2, and its shift to lower wavenumbers during He purge suggested the easy release of adsorbed NO2 from catalyst surface. Since no other oxidation products other than NO2 were observed in the activity test, the intensities decease of all bands implied that these nitrogen-containing species may be the main intermediates of the NO catalytic oxidation reaction, which would decompose to the final oxidation product NO2. As shown in Fig. 8b, the evolution of DRIFT spectra for MnOx(0.3)/TiO2(WI) was similar to that for MnOx(0.3)/TiO2(DP). But the corresponding band in the 1625–1600 cm1 region was stabilized at about 1614 cm1, which was lower than that on
148
Z. Wu et al. / Journal of Colloid and Interface Science 352 (2010) 143–148
in MnOx(0.3)/TiO2(DP) was highly dispersed and enriched on the surface, producing more active sites. Moreover, FTIR, XPS and H2TPR results indicated that more Mn3+ species was created by DP method, which could be another reason for its higher catalytic activity. The positive effects of both higher MnOx dispersion and the abundance of Mn3+ species were further supported by NOTPD profiles. Furthermore, the DRIFT studies implied that the oxidation product NO2 may come from the decomposition of nitrogen-containing species generated during the co-adsorption NO and O2, and the decomposition seemed to be more easy over MnOx(0.3)/TiO2(DP). The high persistence of the catalyst further confirmed that the nitrogen-containing species were the main reaction intermediates. Acknowledgments Fig. 9. The effect of reaction time on the NO catalytic activity of MnOx(0.3)/ TiO2(DP). Reaction conditions: 250 °C; 600 ppm NO, 4% O2, and balance N2; flow rate = 1.2 L/min; catalyst 0.10 g.
MnOx(0.3)/TiO2(DP) (1621 cm1). This also reflected that the nitrogen-containing species over MnOx(0.3)/TiO2(DP) produced by the co-adsorption NO and O2 would be decomposed to NO2 more easily. And it could be confirmed by the above NO-TPD results that much more remarkable desorption band at low temperature over MnOx(0.3)/TiO2(DP) was observed. 3.7. Long-time running test As mentioned above, the NO catalytic oxidation product NO2 could come from the decomposition of nitrogen-containing species which were generated during the co-adsorption of NO and O2 on the catalysts. To further validate this conclusion, activity test with a high space velocity (0.1 g catalyst used) over MnOx(0.3)/TiO2(DP) at 250 °C was carried out for a relative long running time of 20 h. As shown in Fig. 9, no noticeable decrease of NO conversion was observed in 20 h, which suggested that the high activity of MnOx(0.3)/TiO2(DP) was persistent. Therefore, irreversible deposition of nitrogen-containing species seemed unlikely, and these further confirmed that the catalytic oxidation of NO performed via the pathway of the initial formation of nitrogen-containing species followed by their decomposition to NO2. 4. Summary In this study, a series of MnOx/TiO2 composite nanoxides were prepared by DP method and compared with those by WI method for catalytic oxidation of NO to NO2. The optimal Mn/Ti ratio was 0.3 for both series, and MnOx(0.3)/TiO2(DP) showed much higher activity (89% at 250 °C) than MnOx(0.3)/TiO2(WI) (69% at 330 °C) with a GHSV of 25,000 h1. Compared to MnOx(0.3)/TiO2(WI), MnOx(0.3)/TiO2(DP) showed a higher specific surface area, lower reduction temperature in H2TPR profiles and higher surface Mn/Ti ratio from XPS results, and no crystalline structure of MnOx was found from XRD and HRTEM characterizations. All these aspects implied that the MnOx species
This work was financially supported by the National HighTech Research and Development Program (863) of China (2007AA061701), Changjiang Scholar Incentive Program, Ministry of Education, China (2009), and Science Foundation of Zhejiang Province (Y5090053). References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33]
Z.H. Zhu, L.R. Radovic, G.Q. Lu, Carbon 38 (2000) 451. Z.B. Wu, R.B. Jin, Y. Liu, H.Q. Wang, Catal. Commun. 9 (2008) 2217. I. Spasova, P. Nikolov, D. Mehandjiev, J. Colloid Interface Sci. 290 (2005) 343. M. Devadas, O. Kröcher, M. Elsener, A. Wokaun, N. Söger, M. Pfeifer, Y. Demel, L. Mussmann, Appl. Catal., B 67 (2006) 187. M. Kang, D.J. Kim, E.D. Park, J.M. Kim, J.E. Yie, S.H. Kim, L. Hope-Weeks, E.M. Eyring, Appl. Catal., B 68 (2006) 21. D. Thomas, J. Vanderschuren, Chem. Eng. Technol. 23 (2000) 449. H. Takeuchi, M. Ando, N. Kizawa, Ind. Eng. Chem. Process. Des. Dev. 16 (1977) 303. J. Despres, M. Elsener, M. Koebel, O. Krocher, B. Schnyder, A. Wokaun, Appl. Catal., B 50 (2004) 73. L.D. Li, Q. Shen, J. Cheng, Z.P. Hao, Appl. Catal., B 93 (2010) 259. D. Bhatia, R.W. McCabe, M.P. Harold, V. Balakotaiah, J. Catal. 266 (2009) 106. Y.X. Wen, C.B. Zhang, H. He, Y.B. Yu, Y. Teraoka, Catal. Today 126 (2007) 400. M.M. Yung, E.M. Holmgreen, U.S. Ozkan, J. Catal. 247 (2007) 356. Q. Wang, S.Y. Park, J.S. Choi, J.S. Chung, Appl. Catal., B 79 (2008) 101. H. Li, X.L. Tang, H.H. Yi, L.L. Yu, J. Rare Earths 28 (2010) 64. H. Li, X.L. Tang, H.H. Yi, L.L. Yu, Asia-Pacific Power and Energy Engineering Conference, Wuhan, 2009, doi: 10.1109/APPEEC.2009.4918472. G. Qi, R.T. Yang, Appl. Catal., B 44 (2003) 217. J.H. Huang, Z.Q. Tong, Y. Huang, J.F. Zhang, Appl. Catal., B 78 (2008) 309. B.Q. Jiang, Y. Liu, Z.B. Wu, J. Hazard. Mater. 162 (2009) 1249. Z.B. Wu, R.B. Jin, H.Q. Wang, Y. Liu, Catal. Commun. 10 (2009) 935. M.K. Van, A.J. der Lee, J.H. Van Dillen, K.P. Bitter, De. Jong, J. Am. Chem. Soc. 127 (2005) 13573. W.C. Li, M. Comotti, F. Schuth, J. Catal. 237 (2006) 190. I. Puente-Lee, R. Nares, J. Ramirez, P.S. Schabes-Retchkiman, Rev. Mex. Fis. 50 (S1) (2004) 69. H. Jensen, A. Soloviev, Z. Li, E.G. Sogaard, Appl. Surf. Sci. 246 (2005) 239. C.M. Julien, M. Massot, C. Poinsignon, Spectrochim. Acta, Part A 60 (2004) 689. K. Ramesh, L.W. Chen, F.X. Chen, Z.Y. Zhong, J.H. Chin, H.W. Mook, Y.F. Han, Catal. Commun. 8 (2007) 1421. M.R. Morales, B.P. Barbero, L.E. Cadus, Appl. Catal., B 67 (2006) 229. D. Delimaris, T. Ioannides, Appl. Catal., B 84 (2008) 303. J. Trawczynski, B. Bielak, W. Mista, Appl. Catal., B 55 (2005) 277. H.Y. Chen, A. Sayari, A. Adnot, F. Larachi, Appl. Catal., B 32 (2001) 195. I. Atribak, A. Bueno-Lopez, A. Garcia-Garcia, P. Navarro, D. Frias, M. Montes, Appl. Catal., B 93 (2010) 267. A. Cimino, V. Indovina, J. Catal. 33 (1974) 493. M. Kantcheva, J. Catal. 204 (2001) 479. Z.B. Wu, B.Q. Jiang, Y. Liu, H.Q. Wang, R.B. Jin, Environ. Sci. Technol. 41 (2007) 5812.