Atmospheric contaminant elimination has become an important research topic in recent years. NOx emissions from mobile and stationary sources are major air pollutants and pose serious threats to the environment because they can cause acid rain, photochemical smog, and ozone depletion [1]. Selective catalytic reduction (SCR) with NH3 is one of the most promising techniques for reducing NOx emissions from stationary sources. The main commercial industrial catalysts currently available for this process are V2O5/TiO2 modified with WO3 or MoO3 [2, 3]. However, they have drawbacks such as low N2 selectivity, a narrow temperature window (300-400 ℃), and the toxicity of V species to humans [4]. The development of novel SCR catalysts to replace conventional V-based catalysts has therefore been widely investigated.
In recent years, the use of Fe-based catalysts, which have the advantages of low cost, high chemical stability, and non-toxicity, in SCR reactions has attracted much attention [5-8]. In particular, supported Fe/TiO2 catalysts have been well studied because of their excellent catalytic activities at mid-high temperatures (200-400 ℃) in SCR reactions [9, 10]. It has been suggested that the multiple valence states of iron oxides can provide active redox components. Anatase TiO2 is regarded as one of the best supports for SCR catalysts because (1) active components can be well dispersed on the TiO2 surface [11-14], (2) electron excitation and charge transfer from TiO2 to the active components can occur [13, 15, 16], and (3) the anti-SO2 performance of TiO2 is better than those of other supports [12, 17, 18].
Several methods have been developed for synthesizing supported Fe/TiO2 catalysts. Alves et al. [19] reported that Fe/TiO2 prepared by solution impregnation showed high activity at temperatures lower than 450 ℃ and pointed out that interactions occurred between iron(Ⅲ) oxide and TiO2. Coprecipitation is another common technique for synthesizing supported Fe/TiO2 catalysts [9, 20, 21]. Some studies have shown that this method can provide large specific surface areas, and this is beneficial to SCR reactions [12]. However, Wachs et al. [22] argued that the higher activities of coprecipitated catalysts were the result of new sites associated with surface defects on the TiO2 support rather than the higher surface area. In addition, a sol-gel method using butyl titanate as the Ti source was used to synthesize an Fe/TiO2 catalyst [23]. In this method, doping ions were incorporated into the TiO2 lattice, resulting in lattice defects and improved catalytic activity [24].
Recently, methods for the preparation of mesoporous, high-surface-area materials using structure-directing agents have been reported. Organic molecules such as surfactants (e.g. Pluronic P123 and F127) are used as templating agents to adjust the pore structures of the catalysts [25-27]. The majority of SCR catalysts are oxides of transition metals, e.g. Ti, Fe, Mn, and Ce. During precursor drying and calcination, the tension in the pores increases, causing pores to collapse. A surfactant can reduce the interfacial energy and decrease capillary stress, ultimately enabling the production of mesoporous catalysts with high surface areas and stable pore structures. Surfactants should therefore be good promoters for the preparation of SCR catalysts. In this work, Fe/TiO2 catalysts were synthesized using a sol-gel process assisted by an anionic surfactant, namely Pluronic F127. For comparison, catalysts were also synthesized using impregnation and coprecipitation. The effects of the preparation process on the catalyst properties were studied using the Brunauer-Emmett-Teller (BET) method, transmission electron microscopy (TEM), ultraviolet-visible (UV-vis) spectroscopy, X-ray photoelectron spectroscopy (XPS), and H2 temperature-programmed reduction (H2-TPR). The effects of the synthetic method on the SCR reaction mechanism were also investigated using in situ diffuse reflectance infrared Fourier-transform spectroscopy (DRIFTS).
The triblock copolymer surfactant Pluronic F127 (poly(ethylene oxide)-(polypropylene oxide)-poly(ethylene oxide)), tetraethyl titanate (TBOT), Fe(NO3)3·9H2O, and TiOSO4· 2H2O were purchased from Sigma-Aldrich (USA). Analytical-grade anhydrous ethanol (>99.7%), CH3COOH (36%-38%), and NH3·H2O (25%-28%) were used.
Fe/TiO2 was prepared using an F127-assisted sol-gel process. In a typical preparation, F127 (0.4 mmol) was added to sufficient anhydrous ethanol and the solution was heated gently until the F127 was completely dissolved. Then certain amounts of TBOT and CH3COOH were dissolved in the F127 solution under vigorous stirring for 30 min, giving a transparent yellow solution. Fe(NO3)3·9H2O solution was then added drop-wise to the solution under continuous stirring. The molar ratios of the components in the mixture were TBOT:F127: CH3COOH:Fe(NO3)3·9H2O = 1:0.004:0.7:0.14. The mixture was stirred for 2 h and then kept at room temperature for about 3 d to form a gel. The gel was dried in an oven at 110 ℃ for 12 h, and then calcined in air at 450 ℃ for 3 h. The obtained sample was denoted by Fe/TiO2(t). The nominal loading of Fe in the catalyst was 10 wt% (Fe/TiO2 = 10 wt%, the same below).
For comparison, supported Fe/TiO2 catalysts were also prepared using wet impregnation and coprecipitation methods. In the impregnation method, anatase TiO2 (Hehai Technology Company, China) was impregnated with an appropriate amount of Fe(NO3)3·9H2O solution. After stirring for 30 min, the mixture was dried overnight at 110 ℃ and then calcined at 450 ℃ in air for 3 h. In the coprecipitation method, TiOSO4· 2H2O was used as a TiO2 precursor. Certain amounts of Fe(NO3)3·9H2O and TiOSO4·2H2O were completely dissolved in deionized water (80 mL). Aqueous NH3 was added to the solution under vigorous stirring until it was completely deposited. The obtained precipitate was removed by filtration, washed several times with deionized water, and dried at 110 ℃ overnight. The catalyst was obtained by calcination at 450 ℃ in air for 3 h. The Fe/TiO2 catalysts prepared by impregnation and coprecipitation were denoted by Fe/TiO2(i) and Fe/TiO2(c), respectively.
The samples activities were evaluated using a fixed-bed reactor under a simulated flue gas containing 200 ppm NO, 200 ppm NH3, 80 ppm SO2 (when used), 5 vol% H2O (when used), 5 vol% O2, and balance N2. The quartz tube diameter was 8 mm. The total gas flow rate was maintained at 400 cm3/min over 2.0 g of 40-60 mesh catalyst, corresponding to a gas hourly space velocity (GHSV) of 12 000 h-1. The reaction temperature was controlled using a programmable temperature controller with a K-type thermocouple inserted into the catalyst bed. The NOx concentrations at the reactor inlet and outlet were determined continuously using a flue gas analyser (KM9106 Quintox, Kane International Limited).
The specific surface area was determined by the BET method using N2 adsorption at -196 ℃ (Micromeritics ASAP 2020). The pore size distribution was calculated using the Barrett-Joyner-Halenda (BJH) method. Samples were degassed at 200 ℃ for 2 h prior to measurements.
TEM images were obtained using a JEM-2010 instrument operated at 200 kV.
UV-vis diffuse reflectance spectroscopy was performed using a Shimadzu UV-3600 spectrophotometer. BaSO4 was used as the reference material and spectra were recorded in the range 200-900 nm.
Raman spectroscopy was performed using a Horiba Labram instrument at a laser wavelength of 514.5 nm. The spectra were recorded from 200 to 1000 cm-1 with a resolution of 1 cm-1.
XPS was performed using a Kratos XSAM 800 electron spectrometer fitted with an Al Kα source (1486.6 eV). All binding energies were calibrated using the C 1s core level at 284.8 eV as an internal standard.
NH3 temperature-programmed desorption (NH3-TPD) was performed at a total flow rate of 50 mL/min. Before the experiments, the samples were pretreated in N2 at 300 ℃ for 1 h. NH3 adsorption was performed at 180 ℃. Desorption was achieved by heating the samples from 120 to 600 ℃; NH3 was continuously monitored using a portable FT-IR gas analyser
H2-TPR was performed (Micromeritics Autochem 2920) using 100 mg of catalyst in a quartz U-tube under a mixture of 5 vol% H2 and N2. The total flow rate was 50 mL/min and the heating rate was 10 ℃ /min. Prior to the experiments, each sample was pretreated in pure N2 at 300 ℃ for 1 h.
In situ DRIFT spectra were recorded from 800 to 4000 cm-1 at a resolution of 8 cm-1 with 100 scans in Kubelka-Munk mode, using a Bruker Tensor 27 FT-IR spectrometer equipped with a mercury cadmium telluride detector. Prior to each experiment, the sample was pretreated in pure N2 at 300 ℃ for 1 h to eliminate physically adsorbed water and other impurities. The sample background spectrum at each target temperature was recorded during cooling. After cooling to room temperature, the sample was exposed to a controlled stream of 200 ppm NH3 or 200 ppm NO + O2 and balance N2 at 100 mL/min for 1 h. After purging with purified N2 for 30 min, desorption spectra were obtained at various target temperatures by subtraction of the corresponding background reference.
Fig. 1 shows the NH3-SCR performance of the samples synthesized using different methods as a function of temperature from 90 to 300 ℃. NO conversion increased significantly across the entire temperature range, in the order Fe/TiO2(t) > Fe/TiO2(i) > Fe/TiO2(c). Although nearly 100% NO conversion was achieved on all catalysts at 300 ℃, the activities in the temperature window 180-250 ℃ (medium temperature zone) greatly differed. Fe/TiO2(t) gave the best performance and nearly 90% NO was removed at 180 ℃, whereas only 46% and 18% NO were removed using Fe/TiO2(i) and Fe/TiO2(c), respectively. Fig. 2 shows the N2 selectivities of the samples. In the mid-temperature range, the N2 selectivities of the three samples were greater than 90% and the N2 selectivity of Fe/TiO2(t) was slightly higher than those of the other two samples. All the results indicate that the method used to synthesize Fe/TiO2 greatly influenced the mid-temperature activity and N2 selectivity. This is investigated in detail in the following sections.
The structural properties of the prepared catalysts were investigated using N2 adsorption-desorption isotherms. The results are shown in Fig. 3 and the BET surface areas, pore volumes, and average pore diameters are listed in Table 1. Fe/TiO2(t) had the largest specific surface areas (112.40 m2/g) and gave a type-Ⅳ isotherm with an H1 hysteresis loop, indicating that mesopores with a narrow size distribution were formed during synthesis. The corresponding pore size distribution curves (insets in Fig. 3) show that Fe/TiO2(t) consisted of uniform mesopores (5.3 nm). Fe/TiO2(c) also gave a typical type-Ⅳ isotherm, but the quantity adsorbed was much lower than in the case of Fe/TiO2(t) because of the smaller pore volume (0.13 cm3/g), which significantly reduced the specific surface area. Fe/TiO2(i) gave a type-Ⅱ isotherm, suggesting the presence of both mesopores and some macropores, therefore the average pore size of Fe/TiO2(i) was much larger than those of the other catalysts, leading to a small specific surface area. These results show that the synthetic method had important effects on the structural properties of Fe/TiO2.
The samples were also investigated using high-resolution TEM. The images in Fig. 4 show that the primary particle sizes of Fe/TiO2(t), Fe/TiO2(i), and Fe/TiO2(c) were about 10, 11, and 15 nm, respectively. For Fe/TiO2(t), clear lattice fringes (0.352 nm) were observed on the catalyst surface, corresponding to the (101) crystal phase of anatase TiO2. However, unlike Fe/TiO2(i) and Fe/TiO2(c), Fe2O3 was not directly detected on the Fe/TiO2(t) surface, although the presence of Fe was confirmed using inductively coupled plasma atomic emission spectroscopy (ICP-AES; Table 2). These results show that the particles of Fe species were better dispersed on the Fe/TiO2(t) surface, and this promotes the SCR reaction.
The characteristic absorption peaks in the UV-vis spectra (Fig. 5) can be used to confirm the presence and the charges of transition-metal ions. Anatase TiO2 showed a clear O2- → Ti4+ charge-transfer band centred at 256 nm. Generally, Fe2O3 shows a broad absorption peak at around 533 nm, attributed to d-d absorption of Fe(Ⅲ) [28]. The TiO2 absorption band broadened when Fe was introduced, possibly because of interactions between Fe and Ti resulting from incorporation of Fe into the TiO2 lattice [24]. The absorption band shifted by about 40 nm in the Fe/TiO2(t) and Fe/TiO2(c) spectra, indicating stronger interactions in these two samples. Bond energy changes or charge transfer caused by interactions generate defects, producing more active sites. Interactions between Fe and Ti were also identified using Raman spectroscopy (Fig. 6). The Raman spectrum of anatase TiO2 has three major bands, at around 403, 519, and 640 cm-1 [29]. Among the three samples, the Fe/TiO2(t) catalyst showed the largest red shift to a lower wavenumber of the peak at 393 cm-1, suggesting strong interactions between Fe and Ti; this is consistent with the UV-vis results.
The elementary oxidation states and surface compositions of the different catalysts were investigated using XPS; the Fe 2p and O 1s spectra are shown in Fig. 7. The surface atonic concentrations and concentrations obtained using ICP-AES are shown in Table 2. Fig. 7(a) shows that the Fe 2p spectra of the three samples were approximately the same. Clear peaks corresponding to Fe3+ were observed at 710.8 eV (Fe 2p3/2), 725.1 eV (Fe 2p3/2), and 710.8 eV (Fe 2p3/2 satellite peak); however, no peaks at 715 eV, attributable to the Fe2+ shakeup satellite, were detected, implying that the predominant Fe species in all the catalysts was Fe3+ [20, 30]. Table 2 show that Fe/TiO2(i) had the highest surface atomic Fe concentration among the samples. This could be the result of aggregation of Fe atoms on the catalyst surface layer, which would severely disrupt dispersion of the active components, and reduce the catalytic activity. The ICP-AES results (Table 2) show that the Fe contents on all the catalysts were almost the same, implying that the impact of Fe loading on the SCR activity could be ignored.
The O 1s XPS spectrum was fitted with two peaks and the results after deconvolution are shown in Fig. 7(b). The peak located at 529.2-529.7 eV corresponds to lattice oxygen (denoted by Oβ). The peak centred at 531.1-531.5 eV is ascribed to surface adsorbed oxygen (denoted by Oα). The relative centration ratios, i.e. Oα/(Oα + Oβ), were 27.4%, 21.1%, and 16.8% for Fe/TiO2(t), Fe/TiO2(i), and Fe/TiO2(c), respectively. The surface active adsorbed oxygen could enhance the oxidation of NO to NO2, resulting in a fast SCR reaction (NO + NO2 + 2NH3 → 2N2 + 3H2O) [31]. The UV-vis and Raman spectra and catalytic performance suggest that a high concentration of surface adsorbed oxygen derived from defects could partly explain the high catalytic activity of Fe/TiO2(t).
The adsorption of NH3 on the catalyst surfaces at 180 ℃ was investigated using NH3-TPD; the results are shown in Fig. 8. One broad desorption peak spanning the temperature range 200-400 ℃ was observed for all three samples. This peak is attributed to NH3 desorption from weak and medium acidic sites. For the Fe/TiO2(t) catalyst, the peak intensity increased sharply with increasing desorption temperature; the peak was most intense at 305 ℃, and the intensity was twice that of the peak for Fe/TiO2(c). In addition, the area under the peak, which represents the quantity of chemisorbed NH3 molecules, showed that the NH3 adsorption capacities followed the order Fe/TiO2(t) > Fe/TiO2(i) > Fe/TiO2(c). A combination of the BET and NH3-TPD results shows that the large specific surface area and optimal pore structure of Fe/TiO2(t) helped to increase the number of acidic sites for NH3 adsorption, and the improved NH3 adsorption capacity greatly benefited the SCR reaction.
The reducibility, which is an important factor in NH3-SCR, can be determined using H2-TPR. Fig. 9 shows that all the three samples showed clear H2 consumption below 500 ℃, namely at 388 ℃ for Fe/TiO2(t), 431 ℃ for Fe/TiO2(i), and 447 ℃ for Fe/TiO2(c). This is ascribed to the reduction of Fe2O3 to Fe3O4 [19]. It is worth noting that the Fe2O3 to Fe3O4 reduction peak in the case of Fe/TiO2(t) was at a lower temperature (388 ℃) than those for the other samples, indicating that the interactions between ions and Ti species in the template method enhanced the redox properties of the catalyst. Two reduction peaks, at around 524/587 ℃ and 710/761 ℃, assigned to reduction of Fe3O4 to FeO and FeO to Fe, respectively, were also observed for the samples. The initial reduction temperatures for the samples differed from each other. In general, the initial reduction temperature reflects the redox capability of the catalyst, and a lower temperature indicates a stronger redox capability. It can therefore be concluded that the order of the catalyst reducibilities was Fe/TiO2(t) > Fe/TiO2(c) > Fe/TiO2(i). The enhanced redox properties of the catalysts are one possible reason for their excellent activities [32].
The adsorption of NO + O2 over the catalysts at various temperatures was investigated using in situ DRIFTS; the spectra are shown in Fig. 10. The results show that the methods used to synthesize the catalysts significantly affected NO adsorption and states. DRIFT spectra for NO adsorption on the three samples were recorded to clarify the NO species adsorbed on the different samples; the spectra are shown in Fig. 10(a). Fe/TiO2(i) showed the fewest bands for NO adsorption, with only four peaks, because it gave the weakest interactions between the active component and the support. Strong peaks were observed at 1184 (NO- species) and 1461 cm-1 (monodentate nitrate), and weak peaks were observed at 1603 (gaseous NO2 molecules) and 1668 cm-1 (bridged nitrate) [25, 33, 34]. It is worth noting that these peaks, except the one corresponding to NO- species (1184 cm-1), gradually disappeared with increasing temperature from 50 to 150 ℃ (Fig. 10(c)), suggesting low thermal stability of the adsorption bands. More nitrate species were formed on Fe/TiO2(c) than on Fe/TiO2(i). Fig. 10(d) shows that the peak at 1642 cm-1, attributed to the asymmetric frequency of gaseous NO2 molecules, disappeared at 200 ℃ [33, 35]. The peak at 1251 cm-1 was assigned to unstable bridged nitrate, which was only present at low temperatures. After this peak vanished, M-NO2 compounds (1346 cm-1) appeared with increasing temperature, implying that the stable bridged nitrates were transformed into M-NO2 compounds on the surface of the coprecipitated catalyst [34]. The band at 1471 cm-1 can be attributed to monodentate nitrate because it disappeared at around 150 ℃ [34, 36]. The two bands at 1187 and 1133 cm-1 are assigned to NO species and NO2- species, respectively [33]. It is worth noting that as the temperature increased the bands at 1346 and 1187 cm-1 red shifted, indicating that the corresponding species were unstable. For Fe/TiO2(t), unlike the other two catalysts, high-intensity bands for adsorption of bridged nitrate (1241 cm-1), bidentate nitrate (1578 cm-1), and gaseous NO2 molecules (1603 cm-1) were observed at 50 ℃ (Fig. 10(a)). Fig. 10(b) shows that as the temperature increased, the bands for monodentate nitrate at 1289 and 1479 cm-1 vanished at 150 ℃ and the intensities of the other peaks decreased slowly, implying that the nitrate species on Fe/TiO2(t) were more stable than those on Fe/TiO2(i) and Fe/TiO2(c). Chen et al. [25] reported that bridged and bidentate nitrates could react with preadsorbed NH4+ or NH3 to produce more reactive intermediates, thereby accelerating the standard SCR reaction (2NO + 2NH3 + 1/2O2 → 2N2 + 3H2O). It was also reported that NO2 improves the SCR activity via the fast SCR reaction (NO + NO2(a) + 2NH3(a) → 2N2 + 3H2O) [37]. The oxidation of NO to NO2 is therefore generally accepted as being an important step in enhancing the SCR reaction. The addition of F127 promoted NO oxidation to NO2, possibly because high dispersion of the active component enables combination of more NO2 species, producing more monodentate nitrate species (Fe-O-NO2). The results show that a larger number of interactions assisted NO oxidation and further improved the catalytic activity.
Chemisorption of NH3 can be used to investigate the surface acidity of a supported Fe/TiO2 catalyst; this is another important factor in its catalytic performance in NH3-SCR [38, 39]. The DRIFT spectra of NH3 adsorption over the three catalysts are shown in Fig. 11. The three samples showed various bands in the ranges 1000-1700 and 3000-3500 cm-1. The bands at 1435-1450 and 1658-1670 cm-1 correspond to the symmetric and asymmetric bending vibrations, respectively, of NH4+ chemisorbed on Brönsted acidic sites [5, 40]. The peaks at around 1180 and 1610 cm-1 are assigned to the symmetric and asymmetric bending vibrations, respectively, of surface NH3 chemisorbed on Lewis acidic sites [40, 41]. The band at 1340 cm-1 is attributed to bending vibrations of NH3 coordinated to one type of Lewis acidic site [42]. The N-H stretching modes of NH3 coordinated to Lewis acidic sites are observed at 3157, 3230-3250, and 3372-3392 cm-1 [2, 12, 43], and the N-H stretching modes of NH4+ bound to Brönsted acidic sites vibrate at 3083 cm-1 [40].
Only weak Lewis acidic sites are present on TiO2 and the addition of metal oxides introduces Brönsted acidic sites [44]. Fig. 11(a) suggests that both Brönsted and Lewis acidic sites were present on the catalyst surfaces. It should be noted that the intensities of the bands for NH4+ chemisorbed on Brönsted acidic sites (1435-1450 and 1658-1670 cm-1) decreased significantly with increasing temperature over the range 50-150 ℃ as a result of decomposition and desorption. When the temperature exceeded 150 ℃, the Brönsted acid signals almost disappeared for all the samples, but the intensities of the bands for coordinated NH3 on Lewis acidic sites (1188/1184/1181, 1340, and 1610 cm-1) were little changed. These results imply that the Lewis acidic sites were more stable than the Brönsted acidic sites. Fig. 11(a) shows that the adsorption of NH3 species on the sample surfaces differed greatly at 50 ℃. There were more Lewis acidic sites on the surface of Fe/TiO2(t) than on Fe/TiO2(i) and Fe/TiO2(c), shown by the band intensities for coordinated NH3 (1188 cm-1) and the bending vibration of N-H bonds (3242 cm-1). The reason could be related to the synthetic process. (1) In the template method, the active component was encapsulated by the template, i.e. F127, and the number of Lewis acidic sites increased during template calcination [45]. (2) In the coprecipitation method, the active component was primarily covered with hydroxyl groups [TiO(OH)2/Fe(OH)3], therefore more Brönsted acidic sites were generated from the residual hydroxyl group after calcination, identified by the strong band near 1658 cm-1. In the NH3-SCR reaction, Lewis acidic sites are more important than Brönsted acidic sites [46], resulting in a better catalytic performance by Fe/TiO2(t). In addition, it is worth noting that a band at 1553 cm-1, attributed to amide species (-NH2), was observed for Fe/TiO2(i). As the temperature increased, the intensity of -NH2 band decreased rapidly. A new peak at 1590 cm-1 was detected, which could be attributed to intermediates (-NH2, -NH, and -N) formed during NH3 oxidation, which mechanistic studies have been shown to be unnecessary for, but have a positive effect on, the NH3-SCR reaction [40]. In addition, for the coprecipitated catalyst, a clear peak was observed at 1341 cm-1; it could be related to sulfur species remaining in the catalyst.
Flue gas usually contains some SO2 and H2O, therefore the resistances to SO2 and H2O of the catalysts were examined at 210 ℃. Fig. 12 shows that the NO conversions over Fe/TiO2(i) and Fe/TiO2(c) declined sharply when SO2 and H2O were introduced. The NO conversions of these two samples decreased by about 40% in the first 5 h, and only achieved 38% and 5%, respectively. In comparison, the NO conversion over Fe/TiO2(t) decreased slightly, and was still higher than 70% after the first 5 h. Moreover, in the following 7 h, the NO conversion over Fe/TiO2(t) remained stable and changed little. We suggest that there are two main reasons for deactivation of the catalyst by H2O and SO2: (1) deposition of sulfate species on the catalyst surface, blocking the active sites; and (2) sulfation of the active component by SO2, making it inactive in the SCR reaction. The large specific surface area and well-developed pore structure of Fe/TiO2(t) could slow the deposition of sulfate species, facilitating reactant transportation and improving its resistance against H2O and SO2.
The effects of the methods used to prepare Fe/TiO2 catalysts on the morphological structure, surface chemical properties, and catalytic performances in the SCR of NO by NH3 were compared. The results show that the catalyst prepared using a template method, i.e. Fe/TiO2(t), had the highest specific surface area and largest concentration of pore size distribution, and showed the best SCR activity. The characterization results suggest that the Fe/TiO2(t) catalyst had the highest concentration of surface adsorbed oxygen (Oα) and the strongest interactions between the active component and the support; this enhanced the oxidation of NO to NO2 and improved the fast SCR reaction (NO + NO2 + 2NH3 → 2N2 + 3H2O). Moreover, in situ DRIFTS showed that NO + O2 adsorption was most stable on Fe/TiO2(t), because the intensities of the peaks at 1241 cm-1 (bridged nitrate), 1578 cm-1 (bidentate nitrate), and 1603 cm-1 (gaseous NO2 molecules) were still high at 200 ℃, whereas the peaks in the spectra for Fe/TiO2(i) and Fe/TiO2(c) almost disappeared. The Fe/TiO2(t) catalyst therefore had excellent SCR activity. In addition, the large NH3 adsorption capacity, shown by the intensity of the peak at 1188 cm-1 (coordinated NH3 species), and the presence of Lewis acidic sites, shown by the strong N-H stretching modes of coordinated NH3 (3242 and 3388 cm-1), played a crucial role in the NH3-SCR reaction.
2017,
Vol. 38 Issue (11): 1831-1841 
PDF
2017,
Vol. 38 Issue (11): 1831-1841 PDF