- Email: [email protected]
CeO2 catalysts for the selective catalytic reduction of NO with NH3
Accepted Manuscript Different exposed facets VO x /CeO2 Catalysts for the selective catalytic reduction of NO with NH3
Tao Zhang, Huazhen Chang, Kezhi Li, Yue Peng, Xiang Li, Junhua Li PII: DOI: Reference:
S1385-8947(18)30838-6 https://doi.org/10.1016/j.cej.2018.05.049 CEJ 19065
To appear in:
Chemical Engineering Journal
Received Date: Revised Date: Accepted Date:
5 December 2017 18 April 2018 7 May 2018
Please cite this article as: T. Zhang, H. Chang, K. Li, Y. Peng, X. Li, J. Li, Different exposed facets VO x /CeO2
Catalysts for the selective catalytic reduction of NO with NH3, Chemical Engineering Journal (2018), doi: https:// doi.org/10.1016/j.cej.2018.05.049
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Different exposed facets VOx/CeO2 Catalysts for the selective catalytic reduction of NO with NH3 Tao Zhang, †
†,‡
†
‡
‡
‡
Huazhen Chang,* , Kezhi Li, Yue Peng,*, Xiang Li, Junhua Li*,
‡
School of Environment and Natural Resources, Renmin University of China, Beijing 100872, China ‡
State Key Joint Laboratory of Environment Simulation and Pollution Control, School of Environment, Tsinghua University, Beijing, 100084, China
* Corresponding authors: E-mail address: [email protected] Tel.: +86-10-62512572
1
Abstract VOx/CeO2 catalysts, employing CeO2 nanocubes (NCs), nanorods (NRs), and nanopolyhedrons (NPs) with predominately exposed {100}, {110}, and {111} facets as the supports, were prepared by an incipient wetness impregnation method. The catalysts were used in the selective catalytic reduction (SCR) of NOx with NH3. The SCR performance showed that V-CeO2-NPs could achieve significantly higher NO conversion than V-CeO2-NCs and V-CeO2-NRs in the entire temperature range. The characterization results confirmed that the redox and acidic properties of VOx/CeO2 catalysts were closely related to the exposed facets of the CeO2 supports. The excellent SCR activity of V-CeO2-NPs should be attributed to its appropriate redox ability and abundent surface acid sites, which are associated with the predominately exposed {111} facets of CeO2-NPs.
Keywords: CeO2 support; facets; VOx species; NH3-SCR; acidic property
1. Introduction Currently, nitrogen oxides (NOx, including NO and NO2) from stationary and mobile sources have been regarded as major air pollutants [1,2]. Selective catalytic reduction (SCR) of NOx with NH3 is one of the most effective and adopted technologies to remove NOx. V2O5-WO3(MoO3)/TiO2 is widely used as an industrial SCR catalyst at present. However, some inevitable problems remain for this catalyst, such as the narrow operating temperature window (300-400 °C) and low N2 selectivity at high temperatures [3,4]. Therefore, many researchers have been working to develop novel SCR catalysts with higher activity and longer lifetime. Ceria (CeO2) has been attracted much attentions due to its high oxygen storage capacity and excellent redox property. Recently, CeO2 nanocrystals with well-defined morphologies 2
were successfully synthesized and used as catalysts or supports for various catalytic reactions, in which the shape-dependent catalytic performances were observed clearly [5-7]. For instance, Wu et al. observed that CeO2 nanorods with exposing the {110} and {100} facets have the largest oxygen vacancy sites in quantity, followed with nanocubes exposing the {100} facets and nano-octahedra exposing the {111} facets [8]. Further, they found that CO oxidation performance on these samples follows the same trend as oxygen vacancy sites [6]. Dai et al. reported that CeO2 nanorods show a higher catalytic performance in the 1,2-dichloroethane and ethyl acetate oxidation reaction compared with nanocubes and nanooctahedrons [9]. Li et al. studied the effect of CeO2 support facets on VOx/CeO2 catalyst in oxidative dehydrogenation of methanol and proposed that nanorod catalyst with {110} and {100} facets shows higher activity than that of nanopolyhedras with dominating {111} facets and nanocube with dominating {100} facets. This might be related to the abundant oxygen vacancies appear on {110} facets. Nevertheless, the exploration of CeO2-based catalysts with different exposed facets in NH3-SCR reaction is much less reported compared with the vast investigation in above reactions. Han et al investigated the catalytic performance of two kinds of Fe2O3-supported CeO2 nanoshapes for NH3-SCR reaction [10]. They found that the Fe2O3/CeO2 nanorod catalyst shows a higher SCR activity than that of Fe2O3/CeO2 nanopolyhedra catalyst, which is accociated with the exposed facets of their supports. Herein, VOx/CeO2 catalysts with different exposed facets supports, CeO2 nanocubes (NCs), nanorods (NRs) and nanopolyhedrons (NPs), were prepared (denoted as V-CeO2-NCs, V-CeO2-NRs， V-CeO2-NPs). The SCR performances of these catalysts were different regarding to their different supports. The surface structure, redox and acidic properties of the VOx/CeO2 catalysts were investigated using various characterization techniques. The results indicated that the catalysts displayed significant support-morphology-denpendent SCR activity. 3
2. Experimental 2.2. Preparation of the catalysts. CeO2 supports with various morphologies were prepared by a hydrothermal method according to the previous report [11]. For CeO2-NCs,1.98 g of Ce(NO3)3·6H2O and 17.5 g of NaOH were dissolved in 40 and 30 mL deionized water, respectively. Then, the latter solution was added dropwise into the former solution. Next, the slurry was transferred into a 100 mL autoclave, which was heated at 180 °C for 24 h. The final product was collected by filtration, washed with deionized water thoroughly, and then dried at 60 °C for 12 h and calcined at 500 °C for 4 h. The preparation methods of CeO2-NRs and CeO2-NPs were similar to that of CeO2-NCs, except that the hydrothermal temperature was 100 °C for CeO2-NRs and the dosage of NaOH was 0.34 g for CeO2-NPs, respectively. The V-CeO2-NCs catalyst with a vanadia (VOx) loading of 5 wt% was prepared by an incipient wetness impregnation method. The CeO2-CNs support was impregnated with an aqueous solution of NH4VO3 in oxalic acid (nNH4VO3 :noxalic acid = 1:2). Then, the sample was dried at 100 °C overnight and calcined at 500 °C for 4 h. The V-CeO2-NRs and V-CeO2-NPs catalysts were prepared by a similar method.
2.3. Characterization of the catalysts. Transmission electron microscope (TEM) and high-resolution TEM (HRTEM) experiments were performed on a JEOL JEM 2100 electron microscope operated at 200 kV. X-ray diffraction (XRD) patterns were obtained on a Rigaku D/max-2200 Diffractometer using Cu Kα radiation (λ = 1.5418 Å). The Raman spectra were recorded on a Renishaw Micro-Raman System 2000 spectrometer with a 532 nm laser. X-ray photoelectron spectroscopy (XPS) was carried out on an ESCALab 220i-XL electron spectrometer with 300 W of Mg Kα radiation. 4
All the binding energies were calibrated using the C 1s peak (BE = 284.8 eV) as an internal standard. Temperature programmed reduction (TPR) of H2 was performed on a Chemisorb 2720TPx apparatus equipped with a TCD detector. The samples were pretreated in Ar at 300 °C for 1 h, and then heated from ambient temperature to 1000 °C at a rate of 10 °C min-1 under a flowing 5% H2/Ar mixture (50 mL min-1). Temperature programmed desorption (TPD) of NH3 or NO+O2 was carried out on a MultiGas 2030HS FTIR spectrometer. After saturated with NH3/N2 or NO/O2/N2, the samples were heated in the temperature range of 100-700 °C at a ramp of 10 °C min-1 in a flow of N2. In situ DRIFTS (diffuse reflectance infrared Fourier transform spectroscopy) spectra were obtained on a Nicolet Nexus 6700 spectrometer equipped with a Harrick cell and recorded by accumulating 32 scans with a resolution of 4 cm-1.
2.4. Catalyst evaluation. NH3-SCR activities were evaluated in a fixed-bed quartz flow reactor using 0.1 g catalysts of 40-60 mesh. The feed gas mixture consisted of 500 ppm NO, 500 ppm NH3, 5 % O2, 5.5 % H2O (when used), 110 ppm SO2 (when used), and N2 as balance gas. The total flow rate was 200 ml min-1 and the corresponding GHSV (gas hourly space velocity) was approximately 120,000 mL g-1 h-1. The inlet and outlet streams were detected using a MultiGas 2030HS FTIR spectrometer. The SCR kinetic parameters were calculated by the following equation:
k =−
F ln(1 − x) W
(1)
where k is the reaction rate constant (cm3 g-1 s-1), F is the total flow rate (cm3 s-1), W is the mass of catalyst (g), and x is the NO conversion. Furthermore, the apparent activation energies (Ea) was calculated using the Arrhenius equation shown as follows: 5
E k = Aexp a RT
(2)
3. Results 3.1. Sample characterization 3.1.1 TEM and HRTEM results The TEM, HRTEM images and the schematic illustrations of CeO2 supports are shown in Fig. 1. All of the CeO2 supports exhibit clear but varied shapes. The CeO2-NCs sample shows the uniform cube morphology with diameters in the range of 10-30 nm (Fig. 1a1). The HRTEM image of Fig. 1a2 shows clear (200) lattice fringes with an interplanar spacing of 0.27 nm [11], indicating that the CeO2-NCs exposes the {100} facets selectively (Fig. 1a3). Fig. 1b1 exhibits the TEM image of the CeO2-NRs with a uniform diameter (ca. 10 nm) nm and a less-uniform length in the range of 50-200 nm. The HRTEM image in Fig. 1b2 shows two interplanar spacings at 0.19 and 0.27 nm, corresponding to (220) and (200) lattice fringes, respectively [11]. The results suggest that the CeO2-NRs is enclosed by {110} and {100} facets (Fig. 1b 3). Furthermore, the shape of the CeO2-NRs indicates that it grows along [110] direction. Fig. 1c1 shows that the CeO2-NPs has a uniform particle size of 10-15 nm. The corresponding HRTEM image displays two interplanar spacings at 0.27 and 0.31 nm, corresponding to (200) and (111) lattice fringes, respectively [11]. This indicates that the CeO2-NPs is enclosed by {100} and {111} facets (Fig. 1c3). Combined the analysis above and previous literature [12], the CeO2-NCs only exposes {100} facets, while the CeO2-NRs and CeO2-NPs mainly expose {110} and {111} facets, respectively. In addition, the corresponding VOx/CeO2 catalysts show no obvious changes on the morphology compared with their supports (Fig. S1), indicating that the VOx species are highly dispersed on the surface of the supports.
3.1.2 XRD and BET results 6
Fig. 2 shows the XRD patterns of CeO2 supports and corresponding VOx/CeO2 catalysts. All of the CeO2 supports exhibit feature peaks at 28.5°, 33.1°, 47.4°, 56.3°, 59.1°, 69.4°, 76.8°, and 79.4°, which correspond to the (111), (200), (220), (311), (222), (400), (331), and (420) of the fluorite structure of CeO2 (JCPDS 34-0394). Note that compared with the CeO2-NCs, the diffraction peaks broaden considerably for the CeO2-NRs and CeO2-NPs. This indicates a relatively smaller particle size for the latter two samples (Table 1). After the introduction of vanadium on the CeO2, the XRD patterns of the VOx/CeO2 catalysts exhibit the similar characteristics as their corresponding supports. No characteristic peak for VOx species, i.e., V2O5 with diffraction peaks at 2θ = 20.39°, 26.21°, 31.09°, 34.41°, is observed [13], suggesting that VOx species are highly dispersed on the catalysts. Furthermore, no feature peaks for CeVO4 species (2θ = 24.0° and 32.5°) are observed [14]. The N2 adsorption-desorption isotherms of CeO2 support and corresponding VOx/CeO2 catalysts are shown in Fig. S2. All the samples exhibit the type IV isotherms with H1-type hysteresis, indicating the presence of mesopores in the samples. Besides, the BET surface areas, pore volumes, and average pore diameters of CeO2 support and the corresponding catalysts were summarized in Table 1. The specific area of CeO2-NCs (47.8 m2 g-1) is much smaller than those of CeO2-NRs and CeO2-NPs (80.8 and 88.9 m2 g-1, respectively), which is consistent with the XRD results. After the addition of VOx species on the CeO2, the specific surface areas of VOx/CeO2 catalysts decreased, which may be due to the pore blockage of CeO2 by the VOx species. This could be further verified by average pore diameters of the samples shown in Table 1.
3.1.3 Raman results Fig. 3 shows Raman spectra of CeO2 supports and corresponding VOx/CeO2 catalysts. For the CeO2 supports, a strong peak at 460 cm-1 and three weak peaks at 260, 598, 1176 cm-1 are 7
observed. They can be attributed to the F2g mode, second-order transverse acoustic (2TA) mode, oxygen defect-induced (D) mode, and second-order longitudinal optical (2LO) mode of CeO2, respectively [15]. After vanadium loading, several new Raman peaks in the range of 600-1100 cm-1 are observed. The peaks at 708, 780, and 853 cm-1 can be ascribed to the bridging V-O-Ce modes, while the peak at 944 cm-1 can be assigned to the bridging modes of V-O-V bond [16,17]. The above four peaks are present over all VOx/CeO2 samples, suggesting that part of vanadium species in these samples is bonded in the form of isolated monomeric VOx. Specifically, only a weak peak at 995 cm-1 due to V=O stretch in VOx species is observed over the V-CeO2-NRs, revealing the formation of three-dimensional VOx nanoparticles (crystalline V2O5) [17]. However, the crystalline V2O5 has not been detected from XRD analysis. This may be due to that Raman could be used to probe the vibrational modes of bond configurations for even localized short-range order of materials, while XRD is applied to probe the long-range order crystallinity of materials [18]. Moreover, a peak at 1032 cm-1 attributed to the V=O modes of trimeric VOx species is observed over the V-CeO2-NCs and V-CeO2-NPs [17]. The Raman analyses indicate that at the same VOx loading, monomeric VOx, and crystalline V2O5 are formed on the V-CeO2-NRs surface, while monomeric and trimeric VOx are present on the V-CeO2-NCs and V-CeO2-NPs. Further, the peak intensity relative to monomeric VOx species follows the order: V-CeO2-NRs < V-CeO2-NCs < V-CeO2-NPs. For the V-CeO2-NPs, the peak intensity due to trimeric VOx species is much stronger than that for the V-CeO2-NCs. It is clear that the above differences are mainly related to the dominantly exposed facets of their CeO2 supports.
3.1.4 XPS results The chemical environment of Ce, O and V elements on the surface of the VOx/CeO2 catalysts were investigated by XPS. As shown in Fig. 4A, the peaks denoted as v, v’’, v’’’, u, 8
u’’, u’’’ are attributed to Ce4+ species, while two peaks denoted as v’ and u’ are assigned to Ce3+ species [19]. The ratios of Ce3+/Ce4+ are 0.31, 0.41, and 0.35 for V-CeO2-NCs, V-CeO2-NRs, and V-CeO2-NPs, respectively. The fraction of Ce3+ species on the surface of the V-CeO2-NRs catalyst is higher than that on the surface of V-CeO2-NCs and V-CeO2-NPs. This phenomenon indicates that the interaction between cerium and vanadium on the surface of the V-CeO2-NRs catalyst is strongest, which maintains cerium at the low-valence state and leads to formation of more oxygen vacancies. As shown in Fig. 4B, O 1s XPS spectra were deconvolved into two groups of sub-bands: one with binding energy (BE) of 529.1-529.9 eV is attributed to lattice oxygen (Olatt) and the other with BE of 530.8 eV is assigned to surface-adsorbed oxygen (Oads) [20]. The latter is often thought to be beneficial to oxidation reactions. The ratio of Oads/Olatt for V-CeO2-NRs is highest among the three samples, suggesting that the it has best oxidization ability. The V 2p XPS spectra of the VOx/CeO2 samples are shown in Fig. 4C. The BE of V 2p3/2 for all samples is estimated at 517.1-517.6 eV, which is ascribed to V5+ species [21].
3.1.5 H2-TPR results The redox properties of the CeO2 supports and VOx/CeO2 catalysts were investigated by H2-TPR, and the reduction profiles are shown in Fig. 5a. For the CeO2 supports, the peaks at ~300-550 °C are assigned to the reduction of surface Ce4+to Ce3+, while the peaks at ~800-900 °C are attributed to the reduction of bulk CeO2 [22,23]. It is important to note that the onset of CeO2-NRs reduction is obviously lower than CeO2-NCs and CeO2-NPs, indicating the CeO2-NRs sample was more reducible at lower temperatures. Therefore, the CeO2-NRs sample exposing {110} facets possesses a more amount of reducible sites. For the VOx/CeO2 catalysts, there are two main reduction peaks observed for the catalysts. The peaks at ~500-600 °C are predominately due to the reduction of VOx species [24] and the 9
shoulder at lower temperatures could be assigned to the reduction of surface CeO2. Meanwhile, the peaks at ~700-800 °C are attributed to the reduction of bulk CeO2 oxide [25]. Further, the initial H2 consumption rate was used to evaluate the low-temperature reducibility of VOx/CeO2 catalysts [26]. As shown in Fig. 5b, the initial H2 consumption rate increases as follows: V-CeO2-NCs < V-CeO2-NPs < V-CeO2-NRs. This trend suggests that the V-CeO2-NRs catalyst shows the best reducibility at low temperatures. In addition, the reduction peak intensities of surface CeO2 decrease and the reduction shifts to higher temperatures for the VOx/CeO2 catalysts compared with the CeO2 supports. This might be owing to that the strong interaction between VOx and CeO2 affects the redox properties of surface CeO2.
3.1.6 The results of NO+O2-TPD and DRIFTS spectra of NOx adsorption NO2 (from NO oxidation) could be easily captured and stored on the catalyst surface in the form of nitrite and nitrate species. It plays a vital role in the NH3-SCR reaction. NO+O2-TPD experiments were conducted to probe the oxidation activity for the conversion of NO to NO2 on various CeO2 supports and VOx/CeO2 catalysts and the results are summarized in Fig. 6a. Two NOx desorption peaks were observed for the CeO2 supports. According to the previous reports, the low-temperature peak is ascribed to the decomposition of nitrite species, while the high-temperature peak is due to the decomposition of nitrate species [27]. After the addition of VOx species on the CeO2 supports, we can find that the amount of desorbed NOx increase obviously over VOx/CeO2 catalysts, indicating that the existence of VOx could promote the NOx adsorption abilities of the catalysts. Furthermore, the NOx adsorption species on VOx/CeO2 catalysts at 100 °C were tested using in situ DRIFTS. As shown in Fig. 6b, the bands at 1636, 1607, 1577, and 1559 cm-1 can be attributed to adsorbed NO2, bridging nitrate, bidentate nitrate, and monodentate nitrate, 10
respectively [28,29]. Besides, the bands assigned to nitro compound (1385 cm-1) and bridging nitrite (1225 cm-1) are also observed [28,30]. Therefore, it is further proof that the NOx desorption peaks shown in Fig. 6a are attributed to the decomposition of these NOx adsorption species. In addition, the NOx desorption amounts of V-CeO2-NRs are larger than those of V-CeO2-NCs and V-CeO2-NPs (Fig. 6a), indicating that the former has much better oxidation and NOx-adsorption abilities in comparison with the latter two. This could be further verified by the NO oxidition experiments shown in Fig. S3.
3.1.7 The results of NH3-TPD and DRIFTS spectra of NH3 adsorption It is generally accepted that the catalyst acidity is another important factor for the NH3-SCR reaction. The NH3-TPD technique was used to study the strength and amount of surface acid sites on the CeO2 supports and VOx/CeO2 catalysts and the results are shown in Fig. 7a. For CeO2 supports, the adsorption amount of NH3 is in the following sequence: CeO2-NRs (0.16 µmol m-2) < CeO2-NCs (0.22 µmol m-2) < CeO2-NPs (0.40 µmol m-2). After the addition of VOx species on the CeO2 supports, the adsorption amount of NH3 increases obviously over V-CeO2-NCs (1.39 µmol m-2) and V-CeO2-NPs (2.04 µmol m-2) catalysts, while that decreases sharply over V-CeO2-NPs (0.04 µmol m-2) catalyst. This might be related to the exposed facets of their supports, which strongly affects the structure of supported VOx speices (Fig. 3). However, NH3-TPD cannot determine the different acid types (Lewis or Brønsted acid sites) on the catalyst. Therefore, the NH3 adsorption behaviors on VOx/CeO2 catalysts at 100 °C were investigated using in situ DRIFTS. As illustrated in Fig. 7b, the band at 1436 cm-1 can be attributed to vas mode of NH4+ species chemisorbed on Brønsted acid sites, whereas the band at 1127 cm-1 can be assigned to vs mode of coordinated NH3 linked to Lewis acid sites [31,32]. In addition, the bands at 3384, 3263, and 3173 cm-1 are related to the N-H stretching vibration modes of coordinated NH3 [33]. For the V-CeO2-NCs, the intensity of the band ascribed to 11
NH4 + species is higher than that assigned to coordinated NH3 on Lewis acid sites. For the V-CeO2-NPs, both the bands are attributed to NH4 + species on Brønsted acid sites and coordinated NH3 on Lewis acid sites. Their intensities are higher than those on the V-CeO2-NCs. However, only trace amount of NH4+ species adsorbed on Brønsted acid sites could be observed on the V-CeO2-NRs. Therefore, compared with the V-CeO2-NCs and V-CeO2-NRs catalysts, more NH3 will be adsorbed and activated on the V-CeO2-NPs catalyst.
3.2. NH3-SCR performance. Fig. 8a shows the NH3-SCR performance of the CeO2 supports and VOx/CeO2 catalysts in the range of 150-400 °C under a GHSV of 120, 000 mL g−1 h−1. The CeO2 supports exhibit poor SCR activity in the whole temperature range. Among them, the activity of CeO2-NPs is slightly better than those of CeO2-NCs and CeO2-NRs. With the presence of VOx species, a significant difference on the catalytic performance is observed over the VOx/CeO2 catalysts. The SCR activities of the V-CeO2-NCs and V-CeO2-NPs catalysts are improved remarkably, while that of the V-CeO2-NRs catalyst decreases to some extent. Also, the N2 selectivities for the three catalysts follow this trend (Fig. 8b). Fig. 8c shows the Arrhenius plots and the apparent activation energies (Ea). Note that the NO conversion is adjusted to less than 15% to minimize the influences of mass and heat transfer (The data are not shown here.). The Ea value of the V-CeO2-NPs (31.7 kJ mol) is much lower than those of the V-CeO2-NCs (41.2 kJ mol) and V-CeO2-NRs (47.2 kJ mol). Combined with the activity test results, it can be found that there is a clear structure sensitivity in NH3-SCR reaction: for CeO2, VOx speices deposited on {111} facets exhibit the most active, followed by VOx speices anchored to {100} and {110} facets. Finally, the best reactivity of the CeO2 {111} facets with the active VOx speices results in the highest activity of the VOx/CeO2 catalysts. In addition, we chose the best active catalyst, i.e., V-CeO2-NPs, to study the influence of 12
H2O and SO2 on its SCR activity and the results are shown in Fig. 8d. When 5.5% H2O is introduced into the gas inlet, the NO conversion decreases rapidly to ~83% and remains ~83% during 12 h test. After the removal of H2O, the NO conversion could be restored to the original level. A similar situation is found as 110 ppm SO2 added into the stream and ~86% NO conversion is maintained. The partial NO conversion could be recovered after SO2 removing. Further, 5.5% H2O and 110 ppm SO2 are introduced simultaneously, and the NO conversion decreases significantly. However, ~79% NO conversion could be also attained. These results suggest that the V-CeO2-NPs catalyst displays an excellent resistance to H2O and SO2, which is a key point of NOx emission control.
4. Discussion As is known to all, pure CeO2 catalyst exhibits poor catalytic performance in the NH3-SCR reaction. To improve the activity of the CeO2, the introduce of active components onto CeO2 is a good choice. In this case, CeO2 not only plays a role of dispersing active species but also adjusts the electronic properties of active species. Likewise, the synergistic effect in the interfacial region could be generated, which will contribute to the excellent catalytic performance. In addtion to surface active species, the intrinsic property of CeO2 support is another important factor that influences the final catalytic performance of CeO2-based catalysts. The recent advances in the synthesis of nanoshaped CeO2 with different morphologies has enabled us to study the surface sructure dependence of NH3-SCR reaction over CeO2-based catalysts. In this work, the VOx/CeO2 catalysts with different exposing facets have been prepared. The characterization results clearly indicate that the morphology and the exposed facets of CeO2 strongly affect both the structure of supported VOx species and the chemical properties (redox and acid sites) of the VOx/CeO2 catalysts. As evidenced from TEM and HRTEM 13
images (Fig. 1), the V-CeO2-NCs catalyst is only enclosed with {100} facets, whereas the V-CeO2-NRs and V-CeO2-NPs are predominately enclosed with {110} and {111} facets, respectively. According to the results of H2-TPR, NO+O2-TPD and NO oxidition test, the redox ability follows the order of V-CeO2-NRs > V-CeO2-NPs > V-CeO2-NCs, which might be due to the percentage of oxygen defect on {110} facets [34]. This hypothesis can be further verified by the O 1s XPS results shown in Fig. 4B, in which the Oads/Olatt ratio follows the same order as the redox ability. Therefore, more NO2 from NO oxidation can occur over the V-CeO2-NRs catalyst and then form the adsorbed NOx species on catalyst surface, which is beneficial to the process of NH3-SCR reaction. However, the SCR activities of these catalysts do not follow the same tend as the redox ability. This suggests that for the VOx/CeO2 catalysts, redox property is not the only determinant of their catalytic activity during the NH3-SCR process. Our previous study has demonstrated that for the VOx/CeO2 catalysts, Brønsted acid sites could be promoted significantly by CeVO4 species, while CeO2 and polymeric VOx mainly account for Lewis acid sites [35]. The Raman results shown in Fig. 3 demonstrate that for samples with the same VOx loading, VOx species in the V-CeO2-NRs catalyst appear as monomeric VOx and crystalline V2O5, while those in the V-CeO2-NCs and V-CeO2-NPs catalysts are in the form of monomeric and trimeric VOx, respectively. Consequently, the acidic properties of the the VOx/CeO2 catalysts are associated with the quantity of the monomeric and trimeric VOx spicies formed on their surface. From the DRIFT results of NH3 adsorption (Fig. 7b), both Brønsted and Lewis acid sites are observed on the V-CeO2-NCs and V-CeO2-NPs catalysts, whereas only a very small number of Brønsted acid sites are observed on the V-CeO2-NRs catalyst. Moreover, the intensity of Raman peak at 1032 cm-1 for the V-CeO2-NPs is strongest among the three sample (Fig. 3), suggesting that it has most abundant Lewis acid sites, whcih is consistent with the DRIFT results of NH3 adsorption. 14
From the NH3-TPD results, the amount of the acid sites decreases in the order of V-CeO2-NPs > V-CeO2-NCs > V-CeO2-NRs, which is the same order as the SCR activity observed on these catalysts. Therefore, the acidic property is indispensable for the VOx/CeO2 catalysts to reach a high activity in NH3-SCR reaction. Compared with the V-CeO2-NRs catalyst, the V-CeO2-NPs catalyst shows much more excellent SCR sctivity in the entire temperature range (Fig. 8a). The H2-TPR results show that the redox property of V-CeO2-NRs is better than that of V-CeO2-NPs, whereas the acidic property of the former is much weaker than that of the latter. Moreover, the activity test indicate that the V-CeO2-NPs catalyst show better SCR sctivity than that of its support. The H2-TPR results show that the reducibility of the V-CeO2-NPs catalyst decreases after the addtion of VOx species. The NH3-TPD results demonstrate that VOx addition increases the surface acidity of the V-CeO2-NPs catalyst. For SCR reaction, the above comparison analyses indicate that the acidic property might be more important than redox property over the VOx/CeO2 catalysts. For the NH3-SCR mechanism, most researchers believe that NH3 is first adsorbed on Lewis and Brønsted acid sites in the form of coordinated NH3 and NH4+ species, and then they react with gaseous NO/NO2 through E-R mechanism or with activated nitrite/nitrate species through L-H mechanism [36]. The in situ DRIFTS results shown in Fig. S4 indicate that the NH3 species bonded to both Lewis and Brønsted acid sites could participate in the NH3-SCR reaction with various nitrate species. Therefore, the appropriate redox ability and abundant surface acid sites are responsible for the high catalytic activity of the V-CeO2-NPs catalyst. By contrast, less surface acid sites on the V-CeO2-NRs catalyst lead to its poor catalytic activity, although better redox property could be obtained on the catalyst. Finally, it is useful to compare our results to those recently reported by Han et al. [10], who applied Fe2O3/CeO2 nanorods and Fe2O3/CeO2 nanopolyhedra catalysts to NH3-SCR reaction. They found that the Fe2O3/CeO2 nanorods exhibit higher SCR activity than the Fe2O3/CeO2 15
nanopolyhedra. Conversely, our results indicate that the SCR performance of the V-CeO2-NPs catalyst is obviously better than that of the V-CeO2-NRs catalyst. This difference might be associated with the different active species (Fe2O3 vs VOx) supported on these two series of catalysts. (1) For the Fe2O3/CeO2 catalysts, the surface atomic concentration (accessible Fe3+) of the Fe2O3/CeO2 nanorods is higher than that of the Fe2O3/CeO2 nanopolyhedra. For the VOx/CeO2 catalysts, the active VOx species (monomeric and trimeric VOx) on the V-CeO2-NPs catalyst are more abundant than the V-CeO2-NRs catalyst; (2) the addition of Fe2O3 on the CeO2 suppots could enhance the redox properties of the Fe2O3/CeO2 catalysts. In contrast, the reducibility of the VOx/CeO2 catalysts decrease after the addtion of VOx species; (3) the NO and NH3 desorption on the Fe2O3/CeO2 nanorods show larger potential than the Fe2O3/CeO2 nanopolyhedra after loading Fe2O3 species on their CeO2 suppots. But for the VOx/CeO2 catalysts, it's the opposite. The V-CeO2 -NPs catalyst exhibits much better NOx and NH3 adsorption abilities than the V-CeO2-NRs catalyst when the VOx species are loaded on corresponding CeO2 suppots.
5. Conclusions In summary, the VOx/CeO2 catalysts at same VOx loading with tunable CeO2 morphology/facet were prepared, and then the catalytic properties of such catalysts were further investigated in the NH3-SCR reaction. The combination of characterization techniques and
activity
test
demonstrated
that
VOx/CeO2
catalysts
displayed
significantly
support-morphology-dependent catalytic activity. V-CeO2-NPs mainly enclosed with {111} facets showed the best SCR activity, followed by V-CeO2-NCs exposing {100} facets, and V-CeO2-NRs with dominant {110} facets showed negligible NO conversion. The results of Raman, H2-TPR, NH3-TPD, and in situ DRIFTS indicated that the redox and acidic properties of VOx/CeO2 catalysts were closely related to the exposed facets of their CeO2 supports. 16
Further, combined with activity-test results, it can be found that the appropriate redox ability and abundant surface acid sites are two critical requirements for the VOx/CeO2 catalysts. This study not only provides a facile method for the preparation of highly efficient CeO2-based SCR catalysis, but also enables a better understanding of the relationship between the support morphology and catalytic performance of supported vanadium catalysts.
Acknowledgements This work was financially supported by the National Key R&D Program of China (2016YFC0203900 and 2016YFC0203901) and the National Natural Science Foundation of China (Grants 21577173 and 51778619), the Fundamental Research Funds for the Central Universities, and the Research Funds of Renmin University of China (18XNLG09), Public Projects Foundation of China Ministry of Environmental Protection (201509021 and 201509012).
References [1] Z. Liu, S.I. Woo, Catal. Rev.-Sci. Eng. 48 (2006) 43-89. [2] A. Boyano, M.J. Lázaro, C. Cristiani, F.J. Maldonado-Hodar, P. Forzatti, R. Moliner, Chem. Eng. J. 149 (2009) 173-182. [3] G. Busca, L. Lietti, G. Ramis, F. Berti, Appl. Catal. B 18 (1998) 1-36. [4] X. Wang, X. Lia, Q, Zhao, W. Sun, M. Tade, S. Liu, Chem. Eng. J. 288 (2016) 216-222. [5] W.-Q. Han, L.Wu, Y. Zhu, J. Am. Chem. Soc. 127 (2005) 12814-12815. [6] Z. Wu, M. Li, S.H. Overbury, J. Catal. 285 (2012) 61-73. [7] D. Kang, X. Yu, M. Ge, Chem. Eng. J. 330 (2017) 36-43. 17
[8] Z. Wu, M. Li, J. Howe, H.M. Meyer, S.H. Overbury, Langmuir 26 (2010) 16595-16606. [9] Q. Dai, H. Huang, Y. Zhu, W. Deng, S. Bai, X. Wang, G. Lu, Appl. Catal. B 117-118 (2012) 360-368. [10] J. Han, J. Meeprasert, P. Maitarad, S. Nammuangruk, L. Shi, D. Zhang, J. Phys. Chem. C 120 (2016) 1523-1533. [11] H. Mai, L. Sun, Y. Zhang, R. Si, W. Feng, H. Zhang, H. Liu, C.-H. Yan, J. Phys. Chem. B 109 (2005) 24380-24385. [12] M. Hirano, E. Kato, J. Am. Ceram. Soc. 82 (1999) 786-788. [13] A. Pan, J. Zhang, Z. Nie, G. Cao, B.W. Arey, G. Li, S. Liang, J. Liu, J. Mater. Chem. 20 (2010) 9193-9199. [14] Z. Wu, M. Li, S.H. Overbury, ChemCatChem 4 (2012) 1653-1661. [15] W.H. Weber, K.C. Hass, J.R. McBride, Phys. Rev. B 48 (1993) 178-185. [16] M.V. Martínez-Huerta, J.M. Coronado, M. Fernández-García, A. Iglesias-Juez, G. Deo, J.L.G. Fierro, M.A. Bañares, J. Catal. 225 (2004) 240-248. [17] Z. Wu, A.J.Rondinone, I.N. Ivanov, S.H. Overbury, J. Phys. Chem. C 115 (2011) 25368-25378. [18] D. Vernardou, E. Spanakis, G. Kenanakis, E. Koudoumas, N. Katsarakisa, Mater. Chem. Phys. 124 (2010) 319-322. [19] E. Bêche, P. Charvin, D. Perarnau, S. Abanades, G. Flamant, Surf. Interface Anal.40 (2008) 264-267. [20] H.C. Yao, Y.F.Yu Yao, J. Catal. 86 (1984) 254-265. [21] L. Liu, Z. Yao, Y. Deng, F. Gao, B. Liu, L. Dong, ChemCatChem 3 (2011) 978-989. [22] L. Zhang, L. Li, Y. Cao, X. Yao, C. Ge, F. Gao, Y. Deng, C. Tang, L. Dong, Appl.Catal. B 165 (2015) 589-598. [23] S.M. Lee, S.C. Hong, Appl. Catal. B 163 (2015) 30-39. 18
[24] Y. Li, Z. Wei, F. Gao, L. Kovarik, R.A.L. Baylon, C.H.F. Peden, Y. Wang, ACS Catal. 5 (2015) 3006-3012. [25] W. Daniell, A. Ponchel, S. Kuba, F. Anderle, T. Weingand, D.H. Gregory, H. Knozinger, Top. Catal. 20 (2002) 65-74. [26] K. Chen, S. Xie, A.T. Bell, E. Iglesia, J. Catal. 198 (2001) 232-242. [27] L. Chen, J. Li, M. Ge, R. Zhu, Catalysis Today 153 (2010) 77-83. [28] K.I. Hadjiivanov, Catal. Rev.-Sci. Eng. 42 (2000) 71-144. [29] L. Chen, J. Li, M. Ge, Environ. Sci. Technol. 44 (2010) 9590-9596. [30] P. Li, Y. Xin, Q. Li, Z. Wang, Z. Zhang, L. Zheng, Environ. Sci. Technol. 46 (2012) 9600-9605. [31] M.A. Larrubia, G. Ramis, G. Busca, Appl. Catal. B 27 (2000) 145-151. [32] W. Shan, F. Liu, H. He, X. Shi, C. Zhang, Appl. Catal. B 115 (2012) 100-106. [33] A.-S. Mamede, E. Payen, P. Grange, G. Poncelet, A. Ion, M. Alifanti, V.I. Pârvulescu, J. Catal. 223 (2004) 1-12. [34] Y. Li, Z. Wei, F. Gao, L. Kovik, C.H.F. Peden, Y. Wang, J. Catal. 315 (2014) 15-24. [35] Y. Peng, C. Wang, J. Li, Appl. Catal. B 144 (2014) 538-546. [36] J. Li, H. Chang, L. Ma, J. Hao, R.T. Yang, Catal. Today, 175 (2011) 147-156.
19
Captions for Tables Table 1. The structural and physical parameters of CeO2 supports and corresponding VOx/CeO2 catalysts.
20
Table 1. The structural and physical parameters of CeO2 supports and corresponding VOx/CeO2 catalysts. Total pore volume
Average pore
3
(cm g )
diameter (nm)
47.8
0.31
12.9
18.6
CeO2-NRs
80.8
0.75
18.7
12.5
CeO2-NPs
88.9
0.26
4.6
12.4
V-CeO2-NCs
46.0
0.18
7.9
19.8
V-CeO2-NRs
77.6
0.52
11.5
12.7
V-CeO2-NPs
63.3
0.16
5.1
12.8
Sample
SBETa (m2 g-1)
CeO2-NCs
-1
Rb (nm)
a
Calculated by the BET method.
b
Crystallite size estimated by the Debye-Scherrer equation, applied to the (111) reflection of fluorite
CeO2.
21
Figure captions Fig. 1. TEM images, HRTEM images, and schematic illustrations for CeO2-NCs (a1, a2, a3), CeO2-NRs (b1, b2, b3), CeO2-NPs (c1, c2, c3) Fig. 2. XRD patterns of CeO2-NCs (a), CeO2-NRs (b), CeO2-NPs (c), V-CeO2-NCs (d), V-CeO2-NRs (e), and V-CeO2-NPs (f). Fig. 3. Raman spectra of CeO2-NCs (a), CeO2-NRs (b), CeO2-NPs (c), V-CeO2-NCs (d), V-CeO2-NRs (e), and V-CeO2-NPs (f). Fig. 4. Ce 3d (A), O 1s (B), V 2p (C) XPS spectra of V-CeO2-NCs (a), V-CeO2-NRs (b), and V-CeO2-NPs (c). Fig. 5. H2-TPR profiles (a) of the CeO2 supports and VOx/CeO2 catalysts. Arrhenius plots of the rate of H2 consumption during TPR versus inverse temperature (b). Fig. 6. NO+O2-TPD curves (a) of the CeO2 supports and VOx/CeO2 catalysts. DRIFTS spectra of NO+O2 (b) on the VOx/CeO2 catalysts. Fig. 7. NH3-TPD curves (a) of the CeO2 supports and VOx/CeO2 catalysts. DRIFTS spectra of NH3 (b) on the VOx/CeO2 catalysts. Fig. 8. NO conversion (a) and N2 selectivity (b) as a function of temperature in the NH3-SCR reaction. Arrehenius plots of the intrinsic reaction rate constants (c). NH3-SCR activity over V-CeO2-NPs catalyst in the presence of H2O/SO2 at 250 °C (d). Reaction conditions: [NO] = [NH3] = 500 ppm, [O2] = 5%, [H2O]= 5.5% (when used), [SO2]=110 ppm (when used), N2 balance, total flow rate 200 mL min-1 and GHSV = 120,000 mL g-1 h-1.
22
Fig. 1. TEM images, HRTEM images, and schematic illustrations for CeO2-NCs (a1, a2, a3), CeO2-NRs (b1, b2, b3), CeO2-NPs (c1, c2, c3).
23
(220)
(200)
(331) (222)
Intensity / a.u.
f
(400)
(331) (420)
(111)
e d c b a
20
30
40
50
2θ /
60
70
80
ο
Fig. 2. XRD patterns of CeO2-NCs (a), CeO2-NRs (b), CeO2-NPs (c), V-CeO2-NCs (d), V-CeO2-NRs (e), and V-CeO2-NPs (f).
24
e d
1176
f
598 708 780 853 944 995 1032
Intensity / a.u.
260
460
c b a
200 400 600 800 1000 1200 900 1050 -1
Raman Shift / cm
Fig. 3. Raman spectra of CeO2-NCs (a), CeO2-NRs (b), CeO2-NPs (c), V-CeO2-NCs (d), V-CeO2-NRs (e), and V-CeO2-NPs (f).
25
(A) Ce /Ce 0.35
u v'''
u'''
v''
v'
c 0.41
b
O 1s
Oads
Oads/Olatt
v
u'' u'
Olatt
(A)
Ce 3d 4+
Intensity / a.u.
Intensity / a.u.
3+
0.38
c
0.51
b
0.32
a
0.31
a 930
920
910
900
890
535
880
534
533
532
531
530
529
528
527
Binding Energy / eV
Binding Energy / eV
(C)
V 2p
V 2p3/2
Intensity / a.u.
c b
a 526
524
522
520
518
516
514
512
Binding Energy / eV
Fig. 4. Ce 3d (A), O 1s (B), V 2p (C) XPS spectra of V-CeO2-NCs (a), V-CeO2-NRs (b), and V-CeO2-NPs (c).
26
534
V-CeO2-NRs
588
V-CeO2-NCs
CeO2-NPs CeO2-NRs CeO2-NCs
200
300
40 V-CeO2-NCs
-4
V-CeO2-NPs
Intensity / a.u.
Initial H2 consumption rate ( 10 mol )
(a)
529
386
491 500
351 399
400
503
500
600
o
700
800
900
V-CeO2-NRs
30
V-CeO2-NPs
20
10
(b) 0 1.35
1.40
1.45
1.50
1.55
1.60
-1
1000 / T (K )
Temperature / C
Fig. 5. H2-TPR profiles (a) of the CeO2 supports and VOx/CeO2 catalysts. Arrhenius plots of the rate of H2 consumption during TPR versus inverse temperature (b).
27
200 V-CeO2-NRs CeO2-NRs
150 100
V-CeO2-NCs CeO2-NCs
50 0 100
200
300
400
500
600
V-CeO2-NPs
1225
1385
(b)
V-CeO2-NPs CeO2-NPs
Absorbance / a.u.
NOx Concentration / ppm
(a) 250
1636 1607 1577 1559
300
V-CeO2-NRs
V-CeO2-NCs
700
2000
o
1800
1600
1400 -1
1200
Wavenumber / cm
Temperature / C
Fig. 6. NO+O2-TPD curves (a) of the CeO2 supports and VOx/CeO2 catalysts. DRIFTS spectra of NO+O2 (b) on the VOx/CeO2 catalysts.
28
(b)
-2
CeO2-NPs: 0.40 µmol m
0.8
Absorbance / a.u.
150
-2
100
V-CeO2-NRs: 0.04 µmol m -2
CeO2-NRs: 0.16 µmol m
-2
50
0 100
V-CeO2-NCs: 1.39 µmol m
200
300
400
V-CeO2-NRs
0.4
0.0 4000
500
o
V-CeO2-NPs
0.6
V-CeO2-NCs
0.2
-2
CeO2-NCs: 0.22 µmol m
1181
-2
V-CeO2-NPs: 2.04 µmol m
1427
1.0
(a)
3384 3263 3173
NH3 Concentration / ppm
200
3500
2000
1800
1600
1400
1200
-1
Temperature / C
Wavenumber / cm
Fig. 7. NH3-TPD curves (a) of the CeO2 supports and VOx/CeO2 catalysts. DRIFTS spectra of NH3 (b) on the VOx/CeO2 catalysts.
29
100
100
(a) 60
CeO2-NCs
V-CeO2-NCs
CeO2-NRs
V-CeO2-NRs
CeO2-NPs
V-CeO2-NPs
80
N2 Selectivity / %
NO Conversion / %
80
40 20
60 V-CeO2-NCs
40
V-CeO2-NRs
20
CeO2-NRs
V-CeO2-NPs CeO2-NCs
(b)
0
0 150
200
250
300
350
o
400
150
CeO2-NPs
200
250
300
350
o
400
Temperature / C
Temperature / C
100 2
V-CeO2-NCs
(c)
V-CeO2-NRs
1
-1
-1
NO Conversion / %
V-CeO2-NPs
lnk / ( cm g s )
100
(d)
Ea~31.7 kJ/mol
3
Ea~47.2 kJ/mol Ea~41.2 kJ/mol 0
90
90
80
80
70
5.5% H2O
110 ppm SO2
70
5.5% H2O + 110 ppm SO2
60
60
-1 1.6
1.8
2.0
2.2
-1
2.4
2.6
50
2.8
1000/T ( K )
0
10
20
30
40
50
Reaction Time / h
Fig. 8. NO conversion (a) and N2 selectivity (b) as a function of temperature in the NH3-SCR reaction. Arrehenius plots of the intrinsic reaction rate constants (c). NH3-SCR activity over V-CeO2-NPs catalyst in the presence of H2O/SO2 at 250 °C (d). Reaction conditions: [NO] = [NH3] = 500 ppm, [O2] = 5%, [H2O]= 5.5% (when used), [SO2]=110 ppm (when used), N2 balance, total flow rate 200 mL min-1 and GHSV = 120,000 mL g-1 h-1.
30
Graphical Abstract
31
Research Highlights ●
VOx/CeO2 catalysts with different exposed facets were first applied into NH3-SCR reaction.
●
V-CeO2-NPs catalyst with the predominately exposed {111} facets showed the best SCR performance.
●
The excellent SCR activity of V-CeO2-NPs was attributed to its appropriate redox ability and abundent surface acid sites
32
Tao Zhang, Huazhen Chang, Kezhi Li, Yue Peng, Xiang Li, Junhua Li PII: DOI: Reference:
S1385-8947(18)30838-6 https://doi.org/10.1016/j.cej.2018.05.049 CEJ 19065
To appear in:
Chemical Engineering Journal
Received Date: Revised Date: Accepted Date:
5 December 2017 18 April 2018 7 May 2018
Please cite this article as: T. Zhang, H. Chang, K. Li, Y. Peng, X. Li, J. Li, Different exposed facets VO x /CeO2
Catalysts for the selective catalytic reduction of NO with NH3, Chemical Engineering Journal (2018), doi: https:// doi.org/10.1016/j.cej.2018.05.049
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Different exposed facets VOx/CeO2 Catalysts for the selective catalytic reduction of NO with NH3 Tao Zhang, †
†,‡
†
‡
‡
‡
Huazhen Chang,* , Kezhi Li, Yue Peng,*, Xiang Li, Junhua Li*,
‡
School of Environment and Natural Resources, Renmin University of China, Beijing 100872, China ‡
State Key Joint Laboratory of Environment Simulation and Pollution Control, School of Environment, Tsinghua University, Beijing, 100084, China
* Corresponding authors: E-mail address: [email protected] Tel.: +86-10-62512572
1
Abstract VOx/CeO2 catalysts, employing CeO2 nanocubes (NCs), nanorods (NRs), and nanopolyhedrons (NPs) with predominately exposed {100}, {110}, and {111} facets as the supports, were prepared by an incipient wetness impregnation method. The catalysts were used in the selective catalytic reduction (SCR) of NOx with NH3. The SCR performance showed that V-CeO2-NPs could achieve significantly higher NO conversion than V-CeO2-NCs and V-CeO2-NRs in the entire temperature range. The characterization results confirmed that the redox and acidic properties of VOx/CeO2 catalysts were closely related to the exposed facets of the CeO2 supports. The excellent SCR activity of V-CeO2-NPs should be attributed to its appropriate redox ability and abundent surface acid sites, which are associated with the predominately exposed {111} facets of CeO2-NPs.
Keywords: CeO2 support; facets; VOx species; NH3-SCR; acidic property
1. Introduction Currently, nitrogen oxides (NOx, including NO and NO2) from stationary and mobile sources have been regarded as major air pollutants [1,2]. Selective catalytic reduction (SCR) of NOx with NH3 is one of the most effective and adopted technologies to remove NOx. V2O5-WO3(MoO3)/TiO2 is widely used as an industrial SCR catalyst at present. However, some inevitable problems remain for this catalyst, such as the narrow operating temperature window (300-400 °C) and low N2 selectivity at high temperatures [3,4]. Therefore, many researchers have been working to develop novel SCR catalysts with higher activity and longer lifetime. Ceria (CeO2) has been attracted much attentions due to its high oxygen storage capacity and excellent redox property. Recently, CeO2 nanocrystals with well-defined morphologies 2
were successfully synthesized and used as catalysts or supports for various catalytic reactions, in which the shape-dependent catalytic performances were observed clearly [5-7]. For instance, Wu et al. observed that CeO2 nanorods with exposing the {110} and {100} facets have the largest oxygen vacancy sites in quantity, followed with nanocubes exposing the {100} facets and nano-octahedra exposing the {111} facets [8]. Further, they found that CO oxidation performance on these samples follows the same trend as oxygen vacancy sites [6]. Dai et al. reported that CeO2 nanorods show a higher catalytic performance in the 1,2-dichloroethane and ethyl acetate oxidation reaction compared with nanocubes and nanooctahedrons [9]. Li et al. studied the effect of CeO2 support facets on VOx/CeO2 catalyst in oxidative dehydrogenation of methanol and proposed that nanorod catalyst with {110} and {100} facets shows higher activity than that of nanopolyhedras with dominating {111} facets and nanocube with dominating {100} facets. This might be related to the abundant oxygen vacancies appear on {110} facets. Nevertheless, the exploration of CeO2-based catalysts with different exposed facets in NH3-SCR reaction is much less reported compared with the vast investigation in above reactions. Han et al investigated the catalytic performance of two kinds of Fe2O3-supported CeO2 nanoshapes for NH3-SCR reaction [10]. They found that the Fe2O3/CeO2 nanorod catalyst shows a higher SCR activity than that of Fe2O3/CeO2 nanopolyhedra catalyst, which is accociated with the exposed facets of their supports. Herein, VOx/CeO2 catalysts with different exposed facets supports, CeO2 nanocubes (NCs), nanorods (NRs) and nanopolyhedrons (NPs), were prepared (denoted as V-CeO2-NCs, V-CeO2-NRs， V-CeO2-NPs). The SCR performances of these catalysts were different regarding to their different supports. The surface structure, redox and acidic properties of the VOx/CeO2 catalysts were investigated using various characterization techniques. The results indicated that the catalysts displayed significant support-morphology-denpendent SCR activity. 3
2. Experimental 2.2. Preparation of the catalysts. CeO2 supports with various morphologies were prepared by a hydrothermal method according to the previous report [11]. For CeO2-NCs,1.98 g of Ce(NO3)3·6H2O and 17.5 g of NaOH were dissolved in 40 and 30 mL deionized water, respectively. Then, the latter solution was added dropwise into the former solution. Next, the slurry was transferred into a 100 mL autoclave, which was heated at 180 °C for 24 h. The final product was collected by filtration, washed with deionized water thoroughly, and then dried at 60 °C for 12 h and calcined at 500 °C for 4 h. The preparation methods of CeO2-NRs and CeO2-NPs were similar to that of CeO2-NCs, except that the hydrothermal temperature was 100 °C for CeO2-NRs and the dosage of NaOH was 0.34 g for CeO2-NPs, respectively. The V-CeO2-NCs catalyst with a vanadia (VOx) loading of 5 wt% was prepared by an incipient wetness impregnation method. The CeO2-CNs support was impregnated with an aqueous solution of NH4VO3 in oxalic acid (nNH4VO3 :noxalic acid = 1:2). Then, the sample was dried at 100 °C overnight and calcined at 500 °C for 4 h. The V-CeO2-NRs and V-CeO2-NPs catalysts were prepared by a similar method.
2.3. Characterization of the catalysts. Transmission electron microscope (TEM) and high-resolution TEM (HRTEM) experiments were performed on a JEOL JEM 2100 electron microscope operated at 200 kV. X-ray diffraction (XRD) patterns were obtained on a Rigaku D/max-2200 Diffractometer using Cu Kα radiation (λ = 1.5418 Å). The Raman spectra were recorded on a Renishaw Micro-Raman System 2000 spectrometer with a 532 nm laser. X-ray photoelectron spectroscopy (XPS) was carried out on an ESCALab 220i-XL electron spectrometer with 300 W of Mg Kα radiation. 4
All the binding energies were calibrated using the C 1s peak (BE = 284.8 eV) as an internal standard. Temperature programmed reduction (TPR) of H2 was performed on a Chemisorb 2720TPx apparatus equipped with a TCD detector. The samples were pretreated in Ar at 300 °C for 1 h, and then heated from ambient temperature to 1000 °C at a rate of 10 °C min-1 under a flowing 5% H2/Ar mixture (50 mL min-1). Temperature programmed desorption (TPD) of NH3 or NO+O2 was carried out on a MultiGas 2030HS FTIR spectrometer. After saturated with NH3/N2 or NO/O2/N2, the samples were heated in the temperature range of 100-700 °C at a ramp of 10 °C min-1 in a flow of N2. In situ DRIFTS (diffuse reflectance infrared Fourier transform spectroscopy) spectra were obtained on a Nicolet Nexus 6700 spectrometer equipped with a Harrick cell and recorded by accumulating 32 scans with a resolution of 4 cm-1.
2.4. Catalyst evaluation. NH3-SCR activities were evaluated in a fixed-bed quartz flow reactor using 0.1 g catalysts of 40-60 mesh. The feed gas mixture consisted of 500 ppm NO, 500 ppm NH3, 5 % O2, 5.5 % H2O (when used), 110 ppm SO2 (when used), and N2 as balance gas. The total flow rate was 200 ml min-1 and the corresponding GHSV (gas hourly space velocity) was approximately 120,000 mL g-1 h-1. The inlet and outlet streams were detected using a MultiGas 2030HS FTIR spectrometer. The SCR kinetic parameters were calculated by the following equation:
k =−
F ln(1 − x) W
(1)
where k is the reaction rate constant (cm3 g-1 s-1), F is the total flow rate (cm3 s-1), W is the mass of catalyst (g), and x is the NO conversion. Furthermore, the apparent activation energies (Ea) was calculated using the Arrhenius equation shown as follows: 5
E k = Aexp a RT
(2)
3. Results 3.1. Sample characterization 3.1.1 TEM and HRTEM results The TEM, HRTEM images and the schematic illustrations of CeO2 supports are shown in Fig. 1. All of the CeO2 supports exhibit clear but varied shapes. The CeO2-NCs sample shows the uniform cube morphology with diameters in the range of 10-30 nm (Fig. 1a1). The HRTEM image of Fig. 1a2 shows clear (200) lattice fringes with an interplanar spacing of 0.27 nm [11], indicating that the CeO2-NCs exposes the {100} facets selectively (Fig. 1a3). Fig. 1b1 exhibits the TEM image of the CeO2-NRs with a uniform diameter (ca. 10 nm) nm and a less-uniform length in the range of 50-200 nm. The HRTEM image in Fig. 1b2 shows two interplanar spacings at 0.19 and 0.27 nm, corresponding to (220) and (200) lattice fringes, respectively [11]. The results suggest that the CeO2-NRs is enclosed by {110} and {100} facets (Fig. 1b 3). Furthermore, the shape of the CeO2-NRs indicates that it grows along [110] direction. Fig. 1c1 shows that the CeO2-NPs has a uniform particle size of 10-15 nm. The corresponding HRTEM image displays two interplanar spacings at 0.27 and 0.31 nm, corresponding to (200) and (111) lattice fringes, respectively [11]. This indicates that the CeO2-NPs is enclosed by {100} and {111} facets (Fig. 1c3). Combined the analysis above and previous literature [12], the CeO2-NCs only exposes {100} facets, while the CeO2-NRs and CeO2-NPs mainly expose {110} and {111} facets, respectively. In addition, the corresponding VOx/CeO2 catalysts show no obvious changes on the morphology compared with their supports (Fig. S1), indicating that the VOx species are highly dispersed on the surface of the supports.
3.1.2 XRD and BET results 6
Fig. 2 shows the XRD patterns of CeO2 supports and corresponding VOx/CeO2 catalysts. All of the CeO2 supports exhibit feature peaks at 28.5°, 33.1°, 47.4°, 56.3°, 59.1°, 69.4°, 76.8°, and 79.4°, which correspond to the (111), (200), (220), (311), (222), (400), (331), and (420) of the fluorite structure of CeO2 (JCPDS 34-0394). Note that compared with the CeO2-NCs, the diffraction peaks broaden considerably for the CeO2-NRs and CeO2-NPs. This indicates a relatively smaller particle size for the latter two samples (Table 1). After the introduction of vanadium on the CeO2, the XRD patterns of the VOx/CeO2 catalysts exhibit the similar characteristics as their corresponding supports. No characteristic peak for VOx species, i.e., V2O5 with diffraction peaks at 2θ = 20.39°, 26.21°, 31.09°, 34.41°, is observed [13], suggesting that VOx species are highly dispersed on the catalysts. Furthermore, no feature peaks for CeVO4 species (2θ = 24.0° and 32.5°) are observed [14]. The N2 adsorption-desorption isotherms of CeO2 support and corresponding VOx/CeO2 catalysts are shown in Fig. S2. All the samples exhibit the type IV isotherms with H1-type hysteresis, indicating the presence of mesopores in the samples. Besides, the BET surface areas, pore volumes, and average pore diameters of CeO2 support and the corresponding catalysts were summarized in Table 1. The specific area of CeO2-NCs (47.8 m2 g-1) is much smaller than those of CeO2-NRs and CeO2-NPs (80.8 and 88.9 m2 g-1, respectively), which is consistent with the XRD results. After the addition of VOx species on the CeO2, the specific surface areas of VOx/CeO2 catalysts decreased, which may be due to the pore blockage of CeO2 by the VOx species. This could be further verified by average pore diameters of the samples shown in Table 1.
3.1.3 Raman results Fig. 3 shows Raman spectra of CeO2 supports and corresponding VOx/CeO2 catalysts. For the CeO2 supports, a strong peak at 460 cm-1 and three weak peaks at 260, 598, 1176 cm-1 are 7
observed. They can be attributed to the F2g mode, second-order transverse acoustic (2TA) mode, oxygen defect-induced (D) mode, and second-order longitudinal optical (2LO) mode of CeO2, respectively [15]. After vanadium loading, several new Raman peaks in the range of 600-1100 cm-1 are observed. The peaks at 708, 780, and 853 cm-1 can be ascribed to the bridging V-O-Ce modes, while the peak at 944 cm-1 can be assigned to the bridging modes of V-O-V bond [16,17]. The above four peaks are present over all VOx/CeO2 samples, suggesting that part of vanadium species in these samples is bonded in the form of isolated monomeric VOx. Specifically, only a weak peak at 995 cm-1 due to V=O stretch in VOx species is observed over the V-CeO2-NRs, revealing the formation of three-dimensional VOx nanoparticles (crystalline V2O5) [17]. However, the crystalline V2O5 has not been detected from XRD analysis. This may be due to that Raman could be used to probe the vibrational modes of bond configurations for even localized short-range order of materials, while XRD is applied to probe the long-range order crystallinity of materials [18]. Moreover, a peak at 1032 cm-1 attributed to the V=O modes of trimeric VOx species is observed over the V-CeO2-NCs and V-CeO2-NPs [17]. The Raman analyses indicate that at the same VOx loading, monomeric VOx, and crystalline V2O5 are formed on the V-CeO2-NRs surface, while monomeric and trimeric VOx are present on the V-CeO2-NCs and V-CeO2-NPs. Further, the peak intensity relative to monomeric VOx species follows the order: V-CeO2-NRs < V-CeO2-NCs < V-CeO2-NPs. For the V-CeO2-NPs, the peak intensity due to trimeric VOx species is much stronger than that for the V-CeO2-NCs. It is clear that the above differences are mainly related to the dominantly exposed facets of their CeO2 supports.
3.1.4 XPS results The chemical environment of Ce, O and V elements on the surface of the VOx/CeO2 catalysts were investigated by XPS. As shown in Fig. 4A, the peaks denoted as v, v’’, v’’’, u, 8
u’’, u’’’ are attributed to Ce4+ species, while two peaks denoted as v’ and u’ are assigned to Ce3+ species [19]. The ratios of Ce3+/Ce4+ are 0.31, 0.41, and 0.35 for V-CeO2-NCs, V-CeO2-NRs, and V-CeO2-NPs, respectively. The fraction of Ce3+ species on the surface of the V-CeO2-NRs catalyst is higher than that on the surface of V-CeO2-NCs and V-CeO2-NPs. This phenomenon indicates that the interaction between cerium and vanadium on the surface of the V-CeO2-NRs catalyst is strongest, which maintains cerium at the low-valence state and leads to formation of more oxygen vacancies. As shown in Fig. 4B, O 1s XPS spectra were deconvolved into two groups of sub-bands: one with binding energy (BE) of 529.1-529.9 eV is attributed to lattice oxygen (Olatt) and the other with BE of 530.8 eV is assigned to surface-adsorbed oxygen (Oads) [20]. The latter is often thought to be beneficial to oxidation reactions. The ratio of Oads/Olatt for V-CeO2-NRs is highest among the three samples, suggesting that the it has best oxidization ability. The V 2p XPS spectra of the VOx/CeO2 samples are shown in Fig. 4C. The BE of V 2p3/2 for all samples is estimated at 517.1-517.6 eV, which is ascribed to V5+ species [21].
3.1.5 H2-TPR results The redox properties of the CeO2 supports and VOx/CeO2 catalysts were investigated by H2-TPR, and the reduction profiles are shown in Fig. 5a. For the CeO2 supports, the peaks at ~300-550 °C are assigned to the reduction of surface Ce4+to Ce3+, while the peaks at ~800-900 °C are attributed to the reduction of bulk CeO2 [22,23]. It is important to note that the onset of CeO2-NRs reduction is obviously lower than CeO2-NCs and CeO2-NPs, indicating the CeO2-NRs sample was more reducible at lower temperatures. Therefore, the CeO2-NRs sample exposing {110} facets possesses a more amount of reducible sites. For the VOx/CeO2 catalysts, there are two main reduction peaks observed for the catalysts. The peaks at ~500-600 °C are predominately due to the reduction of VOx species [24] and the 9
shoulder at lower temperatures could be assigned to the reduction of surface CeO2. Meanwhile, the peaks at ~700-800 °C are attributed to the reduction of bulk CeO2 oxide [25]. Further, the initial H2 consumption rate was used to evaluate the low-temperature reducibility of VOx/CeO2 catalysts [26]. As shown in Fig. 5b, the initial H2 consumption rate increases as follows: V-CeO2-NCs < V-CeO2-NPs < V-CeO2-NRs. This trend suggests that the V-CeO2-NRs catalyst shows the best reducibility at low temperatures. In addition, the reduction peak intensities of surface CeO2 decrease and the reduction shifts to higher temperatures for the VOx/CeO2 catalysts compared with the CeO2 supports. This might be owing to that the strong interaction between VOx and CeO2 affects the redox properties of surface CeO2.
3.1.6 The results of NO+O2-TPD and DRIFTS spectra of NOx adsorption NO2 (from NO oxidation) could be easily captured and stored on the catalyst surface in the form of nitrite and nitrate species. It plays a vital role in the NH3-SCR reaction. NO+O2-TPD experiments were conducted to probe the oxidation activity for the conversion of NO to NO2 on various CeO2 supports and VOx/CeO2 catalysts and the results are summarized in Fig. 6a. Two NOx desorption peaks were observed for the CeO2 supports. According to the previous reports, the low-temperature peak is ascribed to the decomposition of nitrite species, while the high-temperature peak is due to the decomposition of nitrate species [27]. After the addition of VOx species on the CeO2 supports, we can find that the amount of desorbed NOx increase obviously over VOx/CeO2 catalysts, indicating that the existence of VOx could promote the NOx adsorption abilities of the catalysts. Furthermore, the NOx adsorption species on VOx/CeO2 catalysts at 100 °C were tested using in situ DRIFTS. As shown in Fig. 6b, the bands at 1636, 1607, 1577, and 1559 cm-1 can be attributed to adsorbed NO2, bridging nitrate, bidentate nitrate, and monodentate nitrate, 10
respectively [28,29]. Besides, the bands assigned to nitro compound (1385 cm-1) and bridging nitrite (1225 cm-1) are also observed [28,30]. Therefore, it is further proof that the NOx desorption peaks shown in Fig. 6a are attributed to the decomposition of these NOx adsorption species. In addition, the NOx desorption amounts of V-CeO2-NRs are larger than those of V-CeO2-NCs and V-CeO2-NPs (Fig. 6a), indicating that the former has much better oxidation and NOx-adsorption abilities in comparison with the latter two. This could be further verified by the NO oxidition experiments shown in Fig. S3.
3.1.7 The results of NH3-TPD and DRIFTS spectra of NH3 adsorption It is generally accepted that the catalyst acidity is another important factor for the NH3-SCR reaction. The NH3-TPD technique was used to study the strength and amount of surface acid sites on the CeO2 supports and VOx/CeO2 catalysts and the results are shown in Fig. 7a. For CeO2 supports, the adsorption amount of NH3 is in the following sequence: CeO2-NRs (0.16 µmol m-2) < CeO2-NCs (0.22 µmol m-2) < CeO2-NPs (0.40 µmol m-2). After the addition of VOx species on the CeO2 supports, the adsorption amount of NH3 increases obviously over V-CeO2-NCs (1.39 µmol m-2) and V-CeO2-NPs (2.04 µmol m-2) catalysts, while that decreases sharply over V-CeO2-NPs (0.04 µmol m-2) catalyst. This might be related to the exposed facets of their supports, which strongly affects the structure of supported VOx speices (Fig. 3). However, NH3-TPD cannot determine the different acid types (Lewis or Brønsted acid sites) on the catalyst. Therefore, the NH3 adsorption behaviors on VOx/CeO2 catalysts at 100 °C were investigated using in situ DRIFTS. As illustrated in Fig. 7b, the band at 1436 cm-1 can be attributed to vas mode of NH4+ species chemisorbed on Brønsted acid sites, whereas the band at 1127 cm-1 can be assigned to vs mode of coordinated NH3 linked to Lewis acid sites [31,32]. In addition, the bands at 3384, 3263, and 3173 cm-1 are related to the N-H stretching vibration modes of coordinated NH3 [33]. For the V-CeO2-NCs, the intensity of the band ascribed to 11
NH4 + species is higher than that assigned to coordinated NH3 on Lewis acid sites. For the V-CeO2-NPs, both the bands are attributed to NH4 + species on Brønsted acid sites and coordinated NH3 on Lewis acid sites. Their intensities are higher than those on the V-CeO2-NCs. However, only trace amount of NH4+ species adsorbed on Brønsted acid sites could be observed on the V-CeO2-NRs. Therefore, compared with the V-CeO2-NCs and V-CeO2-NRs catalysts, more NH3 will be adsorbed and activated on the V-CeO2-NPs catalyst.
3.2. NH3-SCR performance. Fig. 8a shows the NH3-SCR performance of the CeO2 supports and VOx/CeO2 catalysts in the range of 150-400 °C under a GHSV of 120, 000 mL g−1 h−1. The CeO2 supports exhibit poor SCR activity in the whole temperature range. Among them, the activity of CeO2-NPs is slightly better than those of CeO2-NCs and CeO2-NRs. With the presence of VOx species, a significant difference on the catalytic performance is observed over the VOx/CeO2 catalysts. The SCR activities of the V-CeO2-NCs and V-CeO2-NPs catalysts are improved remarkably, while that of the V-CeO2-NRs catalyst decreases to some extent. Also, the N2 selectivities for the three catalysts follow this trend (Fig. 8b). Fig. 8c shows the Arrhenius plots and the apparent activation energies (Ea). Note that the NO conversion is adjusted to less than 15% to minimize the influences of mass and heat transfer (The data are not shown here.). The Ea value of the V-CeO2-NPs (31.7 kJ mol) is much lower than those of the V-CeO2-NCs (41.2 kJ mol) and V-CeO2-NRs (47.2 kJ mol). Combined with the activity test results, it can be found that there is a clear structure sensitivity in NH3-SCR reaction: for CeO2, VOx speices deposited on {111} facets exhibit the most active, followed by VOx speices anchored to {100} and {110} facets. Finally, the best reactivity of the CeO2 {111} facets with the active VOx speices results in the highest activity of the VOx/CeO2 catalysts. In addition, we chose the best active catalyst, i.e., V-CeO2-NPs, to study the influence of 12
H2O and SO2 on its SCR activity and the results are shown in Fig. 8d. When 5.5% H2O is introduced into the gas inlet, the NO conversion decreases rapidly to ~83% and remains ~83% during 12 h test. After the removal of H2O, the NO conversion could be restored to the original level. A similar situation is found as 110 ppm SO2 added into the stream and ~86% NO conversion is maintained. The partial NO conversion could be recovered after SO2 removing. Further, 5.5% H2O and 110 ppm SO2 are introduced simultaneously, and the NO conversion decreases significantly. However, ~79% NO conversion could be also attained. These results suggest that the V-CeO2-NPs catalyst displays an excellent resistance to H2O and SO2, which is a key point of NOx emission control.
4. Discussion As is known to all, pure CeO2 catalyst exhibits poor catalytic performance in the NH3-SCR reaction. To improve the activity of the CeO2, the introduce of active components onto CeO2 is a good choice. In this case, CeO2 not only plays a role of dispersing active species but also adjusts the electronic properties of active species. Likewise, the synergistic effect in the interfacial region could be generated, which will contribute to the excellent catalytic performance. In addtion to surface active species, the intrinsic property of CeO2 support is another important factor that influences the final catalytic performance of CeO2-based catalysts. The recent advances in the synthesis of nanoshaped CeO2 with different morphologies has enabled us to study the surface sructure dependence of NH3-SCR reaction over CeO2-based catalysts. In this work, the VOx/CeO2 catalysts with different exposing facets have been prepared. The characterization results clearly indicate that the morphology and the exposed facets of CeO2 strongly affect both the structure of supported VOx species and the chemical properties (redox and acid sites) of the VOx/CeO2 catalysts. As evidenced from TEM and HRTEM 13
images (Fig. 1), the V-CeO2-NCs catalyst is only enclosed with {100} facets, whereas the V-CeO2-NRs and V-CeO2-NPs are predominately enclosed with {110} and {111} facets, respectively. According to the results of H2-TPR, NO+O2-TPD and NO oxidition test, the redox ability follows the order of V-CeO2-NRs > V-CeO2-NPs > V-CeO2-NCs, which might be due to the percentage of oxygen defect on {110} facets [34]. This hypothesis can be further verified by the O 1s XPS results shown in Fig. 4B, in which the Oads/Olatt ratio follows the same order as the redox ability. Therefore, more NO2 from NO oxidation can occur over the V-CeO2-NRs catalyst and then form the adsorbed NOx species on catalyst surface, which is beneficial to the process of NH3-SCR reaction. However, the SCR activities of these catalysts do not follow the same tend as the redox ability. This suggests that for the VOx/CeO2 catalysts, redox property is not the only determinant of their catalytic activity during the NH3-SCR process. Our previous study has demonstrated that for the VOx/CeO2 catalysts, Brønsted acid sites could be promoted significantly by CeVO4 species, while CeO2 and polymeric VOx mainly account for Lewis acid sites [35]. The Raman results shown in Fig. 3 demonstrate that for samples with the same VOx loading, VOx species in the V-CeO2-NRs catalyst appear as monomeric VOx and crystalline V2O5, while those in the V-CeO2-NCs and V-CeO2-NPs catalysts are in the form of monomeric and trimeric VOx, respectively. Consequently, the acidic properties of the the VOx/CeO2 catalysts are associated with the quantity of the monomeric and trimeric VOx spicies formed on their surface. From the DRIFT results of NH3 adsorption (Fig. 7b), both Brønsted and Lewis acid sites are observed on the V-CeO2-NCs and V-CeO2-NPs catalysts, whereas only a very small number of Brønsted acid sites are observed on the V-CeO2-NRs catalyst. Moreover, the intensity of Raman peak at 1032 cm-1 for the V-CeO2-NPs is strongest among the three sample (Fig. 3), suggesting that it has most abundant Lewis acid sites, whcih is consistent with the DRIFT results of NH3 adsorption. 14
From the NH3-TPD results, the amount of the acid sites decreases in the order of V-CeO2-NPs > V-CeO2-NCs > V-CeO2-NRs, which is the same order as the SCR activity observed on these catalysts. Therefore, the acidic property is indispensable for the VOx/CeO2 catalysts to reach a high activity in NH3-SCR reaction. Compared with the V-CeO2-NRs catalyst, the V-CeO2-NPs catalyst shows much more excellent SCR sctivity in the entire temperature range (Fig. 8a). The H2-TPR results show that the redox property of V-CeO2-NRs is better than that of V-CeO2-NPs, whereas the acidic property of the former is much weaker than that of the latter. Moreover, the activity test indicate that the V-CeO2-NPs catalyst show better SCR sctivity than that of its support. The H2-TPR results show that the reducibility of the V-CeO2-NPs catalyst decreases after the addtion of VOx species. The NH3-TPD results demonstrate that VOx addition increases the surface acidity of the V-CeO2-NPs catalyst. For SCR reaction, the above comparison analyses indicate that the acidic property might be more important than redox property over the VOx/CeO2 catalysts. For the NH3-SCR mechanism, most researchers believe that NH3 is first adsorbed on Lewis and Brønsted acid sites in the form of coordinated NH3 and NH4+ species, and then they react with gaseous NO/NO2 through E-R mechanism or with activated nitrite/nitrate species through L-H mechanism [36]. The in situ DRIFTS results shown in Fig. S4 indicate that the NH3 species bonded to both Lewis and Brønsted acid sites could participate in the NH3-SCR reaction with various nitrate species. Therefore, the appropriate redox ability and abundant surface acid sites are responsible for the high catalytic activity of the V-CeO2-NPs catalyst. By contrast, less surface acid sites on the V-CeO2-NRs catalyst lead to its poor catalytic activity, although better redox property could be obtained on the catalyst. Finally, it is useful to compare our results to those recently reported by Han et al. [10], who applied Fe2O3/CeO2 nanorods and Fe2O3/CeO2 nanopolyhedra catalysts to NH3-SCR reaction. They found that the Fe2O3/CeO2 nanorods exhibit higher SCR activity than the Fe2O3/CeO2 15
nanopolyhedra. Conversely, our results indicate that the SCR performance of the V-CeO2-NPs catalyst is obviously better than that of the V-CeO2-NRs catalyst. This difference might be associated with the different active species (Fe2O3 vs VOx) supported on these two series of catalysts. (1) For the Fe2O3/CeO2 catalysts, the surface atomic concentration (accessible Fe3+) of the Fe2O3/CeO2 nanorods is higher than that of the Fe2O3/CeO2 nanopolyhedra. For the VOx/CeO2 catalysts, the active VOx species (monomeric and trimeric VOx) on the V-CeO2-NPs catalyst are more abundant than the V-CeO2-NRs catalyst; (2) the addition of Fe2O3 on the CeO2 suppots could enhance the redox properties of the Fe2O3/CeO2 catalysts. In contrast, the reducibility of the VOx/CeO2 catalysts decrease after the addtion of VOx species; (3) the NO and NH3 desorption on the Fe2O3/CeO2 nanorods show larger potential than the Fe2O3/CeO2 nanopolyhedra after loading Fe2O3 species on their CeO2 suppots. But for the VOx/CeO2 catalysts, it's the opposite. The V-CeO2 -NPs catalyst exhibits much better NOx and NH3 adsorption abilities than the V-CeO2-NRs catalyst when the VOx species are loaded on corresponding CeO2 suppots.
5. Conclusions In summary, the VOx/CeO2 catalysts at same VOx loading with tunable CeO2 morphology/facet were prepared, and then the catalytic properties of such catalysts were further investigated in the NH3-SCR reaction. The combination of characterization techniques and
activity
test
demonstrated
that
VOx/CeO2
catalysts
displayed
significantly
support-morphology-dependent catalytic activity. V-CeO2-NPs mainly enclosed with {111} facets showed the best SCR activity, followed by V-CeO2-NCs exposing {100} facets, and V-CeO2-NRs with dominant {110} facets showed negligible NO conversion. The results of Raman, H2-TPR, NH3-TPD, and in situ DRIFTS indicated that the redox and acidic properties of VOx/CeO2 catalysts were closely related to the exposed facets of their CeO2 supports. 16
Further, combined with activity-test results, it can be found that the appropriate redox ability and abundant surface acid sites are two critical requirements for the VOx/CeO2 catalysts. This study not only provides a facile method for the preparation of highly efficient CeO2-based SCR catalysis, but also enables a better understanding of the relationship between the support morphology and catalytic performance of supported vanadium catalysts.
Acknowledgements This work was financially supported by the National Key R&D Program of China (2016YFC0203900 and 2016YFC0203901) and the National Natural Science Foundation of China (Grants 21577173 and 51778619), the Fundamental Research Funds for the Central Universities, and the Research Funds of Renmin University of China (18XNLG09), Public Projects Foundation of China Ministry of Environmental Protection (201509021 and 201509012).
References [1] Z. Liu, S.I. Woo, Catal. Rev.-Sci. Eng. 48 (2006) 43-89. [2] A. Boyano, M.J. Lázaro, C. Cristiani, F.J. Maldonado-Hodar, P. Forzatti, R. Moliner, Chem. Eng. J. 149 (2009) 173-182. [3] G. Busca, L. Lietti, G. Ramis, F. Berti, Appl. Catal. B 18 (1998) 1-36. [4] X. Wang, X. Lia, Q, Zhao, W. Sun, M. Tade, S. Liu, Chem. Eng. J. 288 (2016) 216-222. [5] W.-Q. Han, L.Wu, Y. Zhu, J. Am. Chem. Soc. 127 (2005) 12814-12815. [6] Z. Wu, M. Li, S.H. Overbury, J. Catal. 285 (2012) 61-73. [7] D. Kang, X. Yu, M. Ge, Chem. Eng. J. 330 (2017) 36-43. 17
[8] Z. Wu, M. Li, J. Howe, H.M. Meyer, S.H. Overbury, Langmuir 26 (2010) 16595-16606. [9] Q. Dai, H. Huang, Y. Zhu, W. Deng, S. Bai, X. Wang, G. Lu, Appl. Catal. B 117-118 (2012) 360-368. [10] J. Han, J. Meeprasert, P. Maitarad, S. Nammuangruk, L. Shi, D. Zhang, J. Phys. Chem. C 120 (2016) 1523-1533. [11] H. Mai, L. Sun, Y. Zhang, R. Si, W. Feng, H. Zhang, H. Liu, C.-H. Yan, J. Phys. Chem. B 109 (2005) 24380-24385. [12] M. Hirano, E. Kato, J. Am. Ceram. Soc. 82 (1999) 786-788. [13] A. Pan, J. Zhang, Z. Nie, G. Cao, B.W. Arey, G. Li, S. Liang, J. Liu, J. Mater. Chem. 20 (2010) 9193-9199. [14] Z. Wu, M. Li, S.H. Overbury, ChemCatChem 4 (2012) 1653-1661. [15] W.H. Weber, K.C. Hass, J.R. McBride, Phys. Rev. B 48 (1993) 178-185. [16] M.V. Martínez-Huerta, J.M. Coronado, M. Fernández-García, A. Iglesias-Juez, G. Deo, J.L.G. Fierro, M.A. Bañares, J. Catal. 225 (2004) 240-248. [17] Z. Wu, A.J.Rondinone, I.N. Ivanov, S.H. Overbury, J. Phys. Chem. C 115 (2011) 25368-25378. [18] D. Vernardou, E. Spanakis, G. Kenanakis, E. Koudoumas, N. Katsarakisa, Mater. Chem. Phys. 124 (2010) 319-322. [19] E. Bêche, P. Charvin, D. Perarnau, S. Abanades, G. Flamant, Surf. Interface Anal.40 (2008) 264-267. [20] H.C. Yao, Y.F.Yu Yao, J. Catal. 86 (1984) 254-265. [21] L. Liu, Z. Yao, Y. Deng, F. Gao, B. Liu, L. Dong, ChemCatChem 3 (2011) 978-989. [22] L. Zhang, L. Li, Y. Cao, X. Yao, C. Ge, F. Gao, Y. Deng, C. Tang, L. Dong, Appl.Catal. B 165 (2015) 589-598. [23] S.M. Lee, S.C. Hong, Appl. Catal. B 163 (2015) 30-39. 18
[24] Y. Li, Z. Wei, F. Gao, L. Kovarik, R.A.L. Baylon, C.H.F. Peden, Y. Wang, ACS Catal. 5 (2015) 3006-3012. [25] W. Daniell, A. Ponchel, S. Kuba, F. Anderle, T. Weingand, D.H. Gregory, H. Knozinger, Top. Catal. 20 (2002) 65-74. [26] K. Chen, S. Xie, A.T. Bell, E. Iglesia, J. Catal. 198 (2001) 232-242. [27] L. Chen, J. Li, M. Ge, R. Zhu, Catalysis Today 153 (2010) 77-83. [28] K.I. Hadjiivanov, Catal. Rev.-Sci. Eng. 42 (2000) 71-144. [29] L. Chen, J. Li, M. Ge, Environ. Sci. Technol. 44 (2010) 9590-9596. [30] P. Li, Y. Xin, Q. Li, Z. Wang, Z. Zhang, L. Zheng, Environ. Sci. Technol. 46 (2012) 9600-9605. [31] M.A. Larrubia, G. Ramis, G. Busca, Appl. Catal. B 27 (2000) 145-151. [32] W. Shan, F. Liu, H. He, X. Shi, C. Zhang, Appl. Catal. B 115 (2012) 100-106. [33] A.-S. Mamede, E. Payen, P. Grange, G. Poncelet, A. Ion, M. Alifanti, V.I. Pârvulescu, J. Catal. 223 (2004) 1-12. [34] Y. Li, Z. Wei, F. Gao, L. Kovik, C.H.F. Peden, Y. Wang, J. Catal. 315 (2014) 15-24. [35] Y. Peng, C. Wang, J. Li, Appl. Catal. B 144 (2014) 538-546. [36] J. Li, H. Chang, L. Ma, J. Hao, R.T. Yang, Catal. Today, 175 (2011) 147-156.
19
Captions for Tables Table 1. The structural and physical parameters of CeO2 supports and corresponding VOx/CeO2 catalysts.
20
Table 1. The structural and physical parameters of CeO2 supports and corresponding VOx/CeO2 catalysts. Total pore volume
Average pore
3
(cm g )
diameter (nm)
47.8
0.31
12.9
18.6
CeO2-NRs
80.8
0.75
18.7
12.5
CeO2-NPs
88.9
0.26
4.6
12.4
V-CeO2-NCs
46.0
0.18
7.9
19.8
V-CeO2-NRs
77.6
0.52
11.5
12.7
V-CeO2-NPs
63.3
0.16
5.1
12.8
Sample
SBETa (m2 g-1)
CeO2-NCs
-1
Rb (nm)
a
Calculated by the BET method.
b
Crystallite size estimated by the Debye-Scherrer equation, applied to the (111) reflection of fluorite
CeO2.
21
Figure captions Fig. 1. TEM images, HRTEM images, and schematic illustrations for CeO2-NCs (a1, a2, a3), CeO2-NRs (b1, b2, b3), CeO2-NPs (c1, c2, c3) Fig. 2. XRD patterns of CeO2-NCs (a), CeO2-NRs (b), CeO2-NPs (c), V-CeO2-NCs (d), V-CeO2-NRs (e), and V-CeO2-NPs (f). Fig. 3. Raman spectra of CeO2-NCs (a), CeO2-NRs (b), CeO2-NPs (c), V-CeO2-NCs (d), V-CeO2-NRs (e), and V-CeO2-NPs (f). Fig. 4. Ce 3d (A), O 1s (B), V 2p (C) XPS spectra of V-CeO2-NCs (a), V-CeO2-NRs (b), and V-CeO2-NPs (c). Fig. 5. H2-TPR profiles (a) of the CeO2 supports and VOx/CeO2 catalysts. Arrhenius plots of the rate of H2 consumption during TPR versus inverse temperature (b). Fig. 6. NO+O2-TPD curves (a) of the CeO2 supports and VOx/CeO2 catalysts. DRIFTS spectra of NO+O2 (b) on the VOx/CeO2 catalysts. Fig. 7. NH3-TPD curves (a) of the CeO2 supports and VOx/CeO2 catalysts. DRIFTS spectra of NH3 (b) on the VOx/CeO2 catalysts. Fig. 8. NO conversion (a) and N2 selectivity (b) as a function of temperature in the NH3-SCR reaction. Arrehenius plots of the intrinsic reaction rate constants (c). NH3-SCR activity over V-CeO2-NPs catalyst in the presence of H2O/SO2 at 250 °C (d). Reaction conditions: [NO] = [NH3] = 500 ppm, [O2] = 5%, [H2O]= 5.5% (when used), [SO2]=110 ppm (when used), N2 balance, total flow rate 200 mL min-1 and GHSV = 120,000 mL g-1 h-1.
22
Fig. 1. TEM images, HRTEM images, and schematic illustrations for CeO2-NCs (a1, a2, a3), CeO2-NRs (b1, b2, b3), CeO2-NPs (c1, c2, c3).
23
(220)
(200)
(331) (222)
Intensity / a.u.
f
(400)
(331) (420)
(111)
e d c b a
20
30
40
50
2θ /
60
70
80
ο
Fig. 2. XRD patterns of CeO2-NCs (a), CeO2-NRs (b), CeO2-NPs (c), V-CeO2-NCs (d), V-CeO2-NRs (e), and V-CeO2-NPs (f).
24
e d
1176
f
598 708 780 853 944 995 1032
Intensity / a.u.
260
460
c b a
200 400 600 800 1000 1200 900 1050 -1
Raman Shift / cm
Fig. 3. Raman spectra of CeO2-NCs (a), CeO2-NRs (b), CeO2-NPs (c), V-CeO2-NCs (d), V-CeO2-NRs (e), and V-CeO2-NPs (f).
25
(A) Ce /Ce 0.35
u v'''
u'''
v''
v'
c 0.41
b
O 1s
Oads
Oads/Olatt
v
u'' u'
Olatt
(A)
Ce 3d 4+
Intensity / a.u.
Intensity / a.u.
3+
0.38
c
0.51
b
0.32
a
0.31
a 930
920
910
900
890
535
880
534
533
532
531
530
529
528
527
Binding Energy / eV
Binding Energy / eV
(C)
V 2p
V 2p3/2
Intensity / a.u.
c b
a 526
524
522
520
518
516
514
512
Binding Energy / eV
Fig. 4. Ce 3d (A), O 1s (B), V 2p (C) XPS spectra of V-CeO2-NCs (a), V-CeO2-NRs (b), and V-CeO2-NPs (c).
26
534
V-CeO2-NRs
588
V-CeO2-NCs
CeO2-NPs CeO2-NRs CeO2-NCs
200
300
40 V-CeO2-NCs
-4
V-CeO2-NPs
Intensity / a.u.
Initial H2 consumption rate ( 10 mol )
(a)
529
386
491 500
351 399
400
503
500
600
o
700
800
900
V-CeO2-NRs
30
V-CeO2-NPs
20
10
(b) 0 1.35
1.40
1.45
1.50
1.55
1.60
-1
1000 / T (K )
Temperature / C
Fig. 5. H2-TPR profiles (a) of the CeO2 supports and VOx/CeO2 catalysts. Arrhenius plots of the rate of H2 consumption during TPR versus inverse temperature (b).
27
200 V-CeO2-NRs CeO2-NRs
150 100
V-CeO2-NCs CeO2-NCs
50 0 100
200
300
400
500
600
V-CeO2-NPs
1225
1385
(b)
V-CeO2-NPs CeO2-NPs
Absorbance / a.u.
NOx Concentration / ppm
(a) 250
1636 1607 1577 1559
300
V-CeO2-NRs
V-CeO2-NCs
700
2000
o
1800
1600
1400 -1
1200
Wavenumber / cm
Temperature / C
Fig. 6. NO+O2-TPD curves (a) of the CeO2 supports and VOx/CeO2 catalysts. DRIFTS spectra of NO+O2 (b) on the VOx/CeO2 catalysts.
28
(b)
-2
CeO2-NPs: 0.40 µmol m
0.8
Absorbance / a.u.
150
-2
100
V-CeO2-NRs: 0.04 µmol m -2
CeO2-NRs: 0.16 µmol m
-2
50
0 100
V-CeO2-NCs: 1.39 µmol m
200
300
400
V-CeO2-NRs
0.4
0.0 4000
500
o
V-CeO2-NPs
0.6
V-CeO2-NCs
0.2
-2
CeO2-NCs: 0.22 µmol m
1181
-2
V-CeO2-NPs: 2.04 µmol m
1427
1.0
(a)
3384 3263 3173
NH3 Concentration / ppm
200
3500
2000
1800
1600
1400
1200
-1
Temperature / C
Wavenumber / cm
Fig. 7. NH3-TPD curves (a) of the CeO2 supports and VOx/CeO2 catalysts. DRIFTS spectra of NH3 (b) on the VOx/CeO2 catalysts.
29
100
100
(a) 60
CeO2-NCs
V-CeO2-NCs
CeO2-NRs
V-CeO2-NRs
CeO2-NPs
V-CeO2-NPs
80
N2 Selectivity / %
NO Conversion / %
80
40 20
60 V-CeO2-NCs
40
V-CeO2-NRs
20
CeO2-NRs
V-CeO2-NPs CeO2-NCs
(b)
0
0 150
200
250
300
350
o
400
150
CeO2-NPs
200
250
300
350
o
400
Temperature / C
Temperature / C
100 2
V-CeO2-NCs
(c)
V-CeO2-NRs
1
-1
-1
NO Conversion / %
V-CeO2-NPs
lnk / ( cm g s )
100
(d)
Ea~31.7 kJ/mol
3
Ea~47.2 kJ/mol Ea~41.2 kJ/mol 0
90
90
80
80
70
5.5% H2O
110 ppm SO2
70
5.5% H2O + 110 ppm SO2
60
60
-1 1.6
1.8
2.0
2.2
-1
2.4
2.6
50
2.8
1000/T ( K )
0
10
20
30
40
50
Reaction Time / h
Fig. 8. NO conversion (a) and N2 selectivity (b) as a function of temperature in the NH3-SCR reaction. Arrehenius plots of the intrinsic reaction rate constants (c). NH3-SCR activity over V-CeO2-NPs catalyst in the presence of H2O/SO2 at 250 °C (d). Reaction conditions: [NO] = [NH3] = 500 ppm, [O2] = 5%, [H2O]= 5.5% (when used), [SO2]=110 ppm (when used), N2 balance, total flow rate 200 mL min-1 and GHSV = 120,000 mL g-1 h-1.
30
Graphical Abstract
31
Research Highlights ●
VOx/CeO2 catalysts with different exposed facets were first applied into NH3-SCR reaction.
●
V-CeO2-NPs catalyst with the predominately exposed {111} facets showed the best SCR performance.
●
The excellent SCR activity of V-CeO2-NPs was attributed to its appropriate redox ability and abundent surface acid sites
32