Mechanistic study of selective catalytic reduction of NO with NH₃ over highly dispersed Fe₂O₃ loaded on Fe-ZSM-5

Abstract

ZSM-5 supported highly dispersed Fe_xO_y clusters were prepared by a sol–gel method for selective catalytic reduction (SCR) of NO with NH₃. XRD, SEM, UV-vis, H₂-temperature-programmed reduction (H₂-TPR), NH₃-temperature-programmed desorption (NH₃-TPD), and BET analyses all indicated that Fe species mainly existed as highly dispersed surface Fe_xO_y clusters with a Fe³⁺ concentration of 22 wt%. NO-temperature-programmed oxidation (NO-TPO) revealed that the Fe_xO_y clusters promoted the oxidation of NO to NO₂, which promoted the low temperature NO_X removal. NH₃ was activated above 250 °C and over-oxidation of NH₃ to NO_X was not observed, as a result, a NO_X removal efficiency of 91% was achieved at 400 °C. Moreover, the SCR reaction route was found to be temperature dependent, below 200 °C, the NO_X reduction followed the reaction between NO₂ and non-activated NH₃. Fast SCR reaction dominated the NO_X removal in the temperature window of 200–325 °C. At temperatures above 250 °C, the normal reaction between activated NH₃ and NO compensated the thermodynamic limitation induced suppression of fast SCR.

---

1. Introduction

Many industrial processes use nitrogen oxide (NO_X, mainly NO and NO₂) containing reactants or produce NO_X as by-products, especially in exhaust gases, which remain a major source for air pollution. In fact, NO_X have given rise to a variety of increasingly harmful impacts on the environment, such as photochemical smog, acid rain, ozone depletion and greenhouse effects. Due to the increasing threat of NO_X to the environment, a great many approaches have been developed to try to control its emission. Among them, selective catalytic reduction (SCR), using V₂O₅ as active catalyst and NH₃ as reductant, has been proven to be the most effective technology and has been widely applied to reduce the NO_X emission from power plants and waste incinerations.^1–4 As a widely used catalyst, the detrimental of V₂O₅-based catalyst is its high toxicity, high activity for SO₂ oxidation, formation of N₂O at high temperature, which stimulated the continuing efforts to develop new catalysts.^5–14 Recently, zeolite based catalysts have attracted much interest for the reduction of NO to N₂ by NH₃, because of their remarkable catalytic activity and nontoxic, lower activity for the oxidation of SO₂ to SO₃, and low N₂O by-product.^15–22 Among them, Fe-exchanged/loaded ZSM-5 showed high activities for the SCR reaction of NO_X by NH₃.⁹ It was reported that even under the condition of very low iron loading, a moderate activity could be achieved over Fe-ZSM-5.²³ Especially heavy/over loaded Fe-ZSM-5 showed a high NO reduction activity and stability in SO₂ and H₂O, even higher than that of commercial V₂O₅-based catalysts.^9,24,25 And the combination of the NH₃ adsorption by zeolite and oxidation of NO to NO₂ by Fe³⁺ was considered to be responsible for its high performance.^26,27

From literature, the activity was found to be enhanced by increasing the amount of the Fe³⁺ exchanged/loaded on ZSM-5; however, the preparation of heavy exchange Fe-ZSM-5 with Fe concentration above 2% is still a challenge.⁹ In this study, by conventional aqueous ion-exchange technique using FeCl₂ as reactant,²⁸ we found that highly dispersed Fe_xO_y clusters with a Fe³⁺ concentration of 22 wt% was successfully loaded on Fe-ZSM-5, which further promoted de-NO_X activity and broadened the reaction temperature window from 250 °C to 400 °C. Our results also suggest that the reaction route for SCR of NO_X is reaction temperature dependent, at least over Fe-ZSM-5 supported highly dispersed Fe_xO_y clusters.

2. Experimental

2.1 Catalyst preparation

The starting zeolite was H-ZSM-5 (Nankai University, Si/Al ≈ 25, S_BET = 425 m² g⁻¹). FeCl₂ (99 wt%, Aldrich) was used as Fe precursors. Fe-ZSM-5 was prepared using a conventional aqueous ion-exchange technique.²⁸ 10 grams of H-ZSM-5 was added to 1 L of 0.05 M FeCl₂ solution with constant stirring in air for 24 h, after stirring, the pH of the solution was measured to be 4, which facilitate the Fe ions exchange into ZSM-5 and Fe_xO_y dispersion.²⁹ For normal Fe-ZSM-5, the mixture was filtered and thoroughly washed with deionized water and then dried at 90 °C for 10 h. For Fe_xO_y clusters heavy loaded zeolite (Fe_H-ZSM-5), the mixture was dried at 90 °C directly without washing. The obtained samples were then calcined at 500 °C for 6 h in air. For comparison, pure Fe₂O₃ was prepared by the same procedure without H-ZSM-5. Details of the surface element composition and structure properties of catalysts were summarized in Table 1.

Table 1 Fe loading, surface atom concentration and surface area of catalysts

Catalysts	Fe content^a (wt%)	Surface area BET (m^2 g^−1)	Total pore volume (mL g^−1)	Pore diameter (nm)
a Surface atomic ratio was determined by EDS.
H-ZSM-5	—	425.06	0.2273	1.16237
Fe-ZSM-5	0.97	391.38	0.2154	1.13005
Fe2O3	77.54	112.13	—	—
FeH-ZSM-5	22.12	305.65	0.1542	1.0796

---

2.2 Catalyst characterization

X-ray diffraction (XRD) patterns were recorded on a Philips XD-98 X-ray diffractometer using Cu Kα radiation (k = 0.15406 nm). Brunauer–Emmett–Teller (BET) surface areas were studied using an ASAP2000 physical absorber. The morphology was characterized by scanning electron microscopy (SEM) (JEOL S-4800); energy dispersive spectrometry (EDS) was carried out in the same facility. X-ray photoelectron spectroscopic (XPS) data were obtained using a Thermo ESCALAB 250, the X-ray source was an Al Kα radiation, and all binding energies were referenced to the 284.8 eV C 1s.

The UV-vis diffuse reflection spectra were obtained for the dry-pressed disk samples using a Scan UV-vis spectrophotometer (UV-vis DRS: TU-1901, China) equipped with an integrating sphere assembly, and BaSO₄ as reflectance sample. The spectra were recorded at room temperature in air ranged from 200 to 800 nm. All the samples were degassed at 200 °C prior to measurements.

H₂-temperature-programmed reduction (H₂-TPR) experiments were performed using 100 mg of each catalyst. The catalysts were preheated at 400 °C for 1 h in air. After they were cooled to 100 °C, the samples were heated up to 800 °C with a heating rate of 10 °C min⁻¹ under a mixed flow of 10% H₂ in helium and a flow rate of 30 mL min⁻¹.

NH₃-temperature-programmed desorption (NH₃-TPD) experiments were performed using 100 mg of each catalyst to determine their NH₃ adsorption. The sample was preheated in a N₂ stream (30 mL min⁻¹) at 400 °C for 1 h, and then cooled to 100 °C. The pretreated sample was then exposed to a mixed flow of 4% NH₃ in argon at a flow rate of 20 mL min⁻¹ for 3 h at 100 °C, and heated up to 800 °C at a heating rate of 10 °C min⁻¹. The H₂-TPR and NH₃-TPD data were recorded using an on-line gas chromatograph equipped with a thermal conductivity detector.

NO-temperature-programmed desorption (NO-TPD) experiments were performed in a fixed-bed flow reactor. 1.5 g of each catalyst was pasted on three Al plates and inserted into the fixed-bed.³⁰ The samples were exposed to a 800 ppm NO flow without O₂ for 1 h at 400 °C and cooled to 100 °C in the same gas stream, then the samples were purged by a N₂ until the NO signal returned to the baseline level. Finally, the temperature was ramped from 80 °C to 400 °C in N₂ at a heating rate of 10 °C min⁻¹. Similarly, NO₂-TPD experiments were performance in 800 ppm NO₂ diluted in N₂ without O₂. And the desorbed NO/NO₂ concentrations depended on temperature were recorded using an NO–NO₂-NO_X analyzer (Testo AG testo 340).

NO-temperature-programmed oxidation (NO-TPO) experiments were performed in the same reactor from 100 °C to 400 °C under 800 ppm of NO, 5% O₂ and a N₂ balance at a total flow rate of 1000 sccm. NH₃-temperature-programmed oxidation (NH₃-TPO) experiments were also performed in the same fixed-bed flow reactor from under from 100 °C to 400 °C in 800 ppm NH₃, 5% O₂, and a N₂ balance at a total flow rate of 1000 sccm. The inlet and outlet concentrations of NH₃ were determined by titration with 0.001 M hydrochloric acid, and methyl red was used as the indicator.

2.3 Catalyst activity characterization

The catalytic activity tests were carried out in a fixed-bed quartz reactor with 1.5 g catalyst pasted on three Al plates (4 cm × 10 cm).³⁰ The inlet and outlet concentrations of NO and NO₂ were monitored using Testo AG testo 340. The simulated gas used for these tests contained 800 ppm NO, 800 ppm NH₃ and 5 vol% O₂ balanced by N₂ at total flow rate of 1000 sccm and a GHSV of 100 000 h⁻¹.

3. Results and discussion

3.1 Catalyst activity

Fig. 1 shows NO_X conversion as a function of temperature over different catalysts. The H-ZSM-5 catalyst presented the lowest activity for NO_X reduction in the investigated temperature window. Fe₂O₃ started to catalyze NO_X reduction at 200 °C, and the conversion increased with temperature up to 62% at 350 °C, and then fell to 38% at 400 °C. For Fe-ZSM-5, the reduction of NO_X started at 250 °C and then increased quickly and reached to 77% at 400 °C.²³ The best performance was achieved over Fe_H-ZSM-5, which showed high performance even at temperature as low as 250 °C and a NO_X conversion of 91% was achieved at 400 °C. Moreover, the activity loss at high temperature above 350 °C over normal catalysts was not observed at 400 °C.

Fig. 1 (a) NO conversion over this series of catalysts. (b) E_a calculated using the Arrhenius plot according to eqn (2) over Fe_H-ZSM-5, the inset shows the k calculated according to eqn (1). The inset is the k calculated according to eqn (1). Reaction conditions: 800 ppm NO, 800 ppm NH₃, and 5% O₂ balanced by N₂ at total flow rate of 1000 sccm and a GHSV of 100 000 h⁻¹.

In order to further clarify the temperature dependency of NO_X conversion, the kinetic parameters were calculated based on the assumption that the reaction is first-order dependent on NO and zero-order dependent on NH₃, and the kinetic parameters were calculated from NO conversion over Fe_H-ZSM-5 as:^31–33

where k is the reaction rate coefficient (mL g⁻¹ s⁻¹), V is the total gas flow rate (mL s⁻¹), M is the catalyst weight (g), x is the conversion of NO_X (%). The obtained values of k were used in the Arrhenius plot and then the activation energies (E_a) of the SCR reaction were derived.

As shown in Fig. 1b, the reaction can be divided into three stages, at temperature below 200 °C, the apparent E_a was calculated as low as 8.98 kJ mol⁻¹; the highest E_a of 27.01 kJ mol⁻¹ was calculated at temperature between 200 and 325 °C; at temperature above 325 °C, the E_a was calculated to be 10.78 kJ mol⁻¹. This suggests that the reaction route for NO reduction is temperature dependent.

3.2 XRD, SEM and BET study

The crystallographic structures of the ZSM-5 support and Fe-ZSM-5 zeolite were studied by XRD as shown in Fig. 2. All of the XRD patterns showed three diffraction peaks at 2θ = 23–25°, matching well with the standard phase of ZSM-5 zeolite,³⁴ indicating that the frame structure of ZSM-5 was not destroyed during Fe³⁺ ions exchanging and/or Fe_xO_y clusters loaded. The spectrum of H-ZSM-5 are same as that of Fe-ZSM-5, indicating Fe³⁺ occupied the H⁺ position.³⁵ Pure Fe₂O₃ existed as α-Fe₂O₃ which was supported by the diffraction peaks at 33.1°, 35.6°, 40.9°, 49.5°, and 54.1°. While for Fe_H-ZSM-5, peaks attributed to Fe₂O₃ were not detected obviously, indicates a high dispersion of Fe_xO_y clusters on Fe-ZSM-5.³⁶

Fig. 2 XRD patterns of the samples.

Fig. 3 shows typical SEM images of samples, Fe₂O₃ showed typical needle like morphology, suggesting the α-Fe₂O₃ structure,³⁷ three zeolite containing samples showed very similar morphology, again indicating the fact that frame structure was remained during Fe³⁺ ions exchanging and/or Fe_xO_y clusters loaded. The needle like Fe₂O₃ phase was not observed by SEM observation in Fe_H-ZSM-5, further proved the high dispersion of Fe_xO_y clusters. This assumption was also supported by BET measurement as shown in Table 1, after Fe³⁺ was exchanged into zeolite to form Fe-ZSM-5, the surface area decreased from 425 to 391 m² g⁻¹, in agreement with other reports.³⁸ The surface area of Fe_H-ZSM-5 reduced to 305.65 m² g⁻¹ due to the loading of Fe_xO_y clusters. From the pore size change caused by Fe³⁺ exchange and Fe_xO_y clusters loading (from 1.162 nm to 1.13 nm and 1.08 nm for Fe-ZSM-5 and Fe_H-ZSM-5), Fe³⁺ substituted the H⁺, at the same time Fe_xO_y clusters also loaded inside the micropores of zeolite, thus reduced the pore size of Fe_H-ZSM-5.

Fig. 3 Typical SEM images of (a) H-ZSM-5, (b) Fe-ZSM-5, (c) Fe₂O₃, and (d) Fe_H-ZSM-5.

3.3 UV-vis analysis

The UV-vis spectra were shown in Fig. 4. The UV-vis spectra of Fe-ZSM-5 and Fe_H-ZSM-5 zeolites presented in this work have been deconvoluted into subbands to facilitate the assignment of different Fe species, detailed information was shown in Table 2. For Fe-ZSM-5, two strong peaks located at 218 nm and 270 nm could be attributed to the isolated Fe³⁺ sites in tetrahedral and higher coordination (five or six oxygen ligands),^39–41 the peaks located at 300–400 nm (assigned to oligomeric clusters) and >400 nm (assigned to large Fe₂O₃ clusters) were weak.^39–41 This indicated that Fe species were mainly exchanged into zeolite to form Fe-ZSM-5. While for Fe_H-ZSM-5, besides the peaks attributed to isolated Fe³⁺, peaks for Fe_xO_y clusters were also strong, which indicated that Fe³⁺ species mainly loaded on the zeolite surface as clusters with good dispersion, in good agreement with our XRD measurements. In fact, the XRD peaks attributed to Fe₂O₃ were hardly detected. As shown in Table 2, semi-quantitative analysis revealed that more than 60% of the Fe³⁺ species were exchanged into zeolite in Fe-ZSM-5, and more than 70% of Fe³⁺ species existed as Fe_xO_y clusters in Fe_H-ZSM-5.

Fig. 4 UV-vis spectra of Fe-ZSM-5 and Fe_H-ZSM-5.

Table 2 The area of the subbands derived by deconvolution of the TPD spectra and the UV/VIS-DRS spectra (area₁ at λ < 300 nm, area₂ at 300 < λ < 400 nm, and area₃ at λ > 400 nm)

Catalysts	Amount of acid		Existing form of Fe species
	Areaw^a	Areas^b	Area1^c	Area2^d	Area3^e
a Weak surface acid sites.b Strong surface acid sites.c Isolated Fe^3+ in tetrahedral and higher coordination.d Small oligomeric FexOy clusters.e Large FexOy clusters.
H-ZSM-5	43.96	55.89	—	—	—
Fe-ZSM-5	27.14	29.95	73.6	16	24.6
Fe2O3	18.79	—	—	—	—
FeH-ZSM-5	31.16	15.90	60.1	37.9	110.3

---

3.4 XPS analysis

The XPS results were shown in Fig. 5, for Fe₂O₃, the binding energies of Fe_2p3/2 and Fe_2p1/2 in Fig. 5 are located around 711.2 and 725.5 eV, indicating that Fe is directly bonded to O.⁴² Both Fe_2p3/2 and Fe_2p1/2 peaks are accompanied by a small satellite on the high binding energy side, which is the characteristic of Fe³⁺ in Fe₂O₃, in consistent with XRD characterization.⁴³ According to Jin's report,⁴² when the calcine temperature was >500 °C, Fe³⁺ became the main products in Fe-ZSM-5. Here the Fe related peaks in Fe-ZSM-5 is very weak, only a small difference can be observed as the inset shown, indicating a low Fe exchange in Fe-ZSM-5 after water washing. While for Fe_H-ZSM-5, Fe related peaks were evident and all the iron species existed as Fe³⁺.

Fig. 5 Fe_2p XPS spectra of H-ZSM-5, Fe-ZSM-5, Fe₂O₃, and Fe_H-ZSM-5.

3.5 H₂-TPR and NH₃-TPD analysis

H₂-TPR experiments were shown in Fig. 6. H-ZSM-5 had no obvious H₂ consumption peaks during the whole temperature range, indicating that the ZSM-5 zeolite support shows no oxidation ability. Fe₂O₃ showed an obvious reduction peak around 370 °C and another large broad peak around 660 °C, which are attributed to the reduction from Fe₂O₃ to Fe₃O₄, and the overlap of Fe₃O₄ to FeO and Fe, respectively.⁴⁴ Due to the low Fe loading in Fe-ZSM-5, the Fe related reduction peaks could hardly detect, with very weak peak around 350 °C. In the case of Fe_H-ZSM-5, the peak attributed to Fe₂O₃ to Fe₃O₄ reduction disappeared, and a new peak located between 400 and 500 °C was detected, moreover, the peaks attributed to Fe₃O₄ to FeO and FeO to Fe was separated. Similar phenomena was also reported by Heinrich and Chen,^45,46 normally, the reduction peak is expected to shift to the low temperature side due to the high dispersion of Fe_xO_y clusters, however, the delayed water removal from the surrounding microporous network structure induced the peak to shifted to the high temperature side.^45,46 Also, the existence of interaction between highly dispersed Fe_xO_y clusters and zeolite could further shift the peak to higher temperature.⁴⁷

Fig. 6 H₂ temperature programmed reduction profiles of catalysts.

The NH₃-TPD results are shown in Fig. 7. All the ZSM-5 profiles display two peaks centered at about 180 °C and 340 °C, ascribed to weak and strong surface acid sites.^48,49 While for Fe₂O₃, only a strong peak around 220 °C was observed, no strong acid site in Fe₂O₃ indicating that it did not absorb NH₃ at temperature above 300 °C. The area of the subbands of the TPD spectra were calculated and shown in Table 2. The NH₃ desorption peak at 180 °C and 340 °C over Fe-ZSM-5 and Fe_H-ZSM-5 was less than that over H-ZSM-5, which suggested that part of the Fe³⁺ occupied H⁺ site and Fe_xO_y clusters further covered parts of surface H⁺ in zeolite (the surface H⁺ is considered to be the origin of the strong acid sites^50,51), thus reduced the NH₃ adsorption capacity of Fe_H-ZSM-5 even less than Fe-ZSM-5, in agreement with our TPR and BET analyses.

Fig. 7 NH₃ temperature programmed desorption profiles of catalysts.

3.6 NO to NO₂ oxidation and NO_X TPD

According to literature,^13,52 the existence of NO₂ can accelerate the SCR reaction. Therefore, the NO to NO₂ oxidation as a function of reaction temperature was studied and shown in Fig. 8. Both H-ZSM-5 and Fe₂O₃ showed very low activity for NO to NO₂ oxidation. NO_X conversion kept below 25% through the whole temperature range, while the Fe-ZSM-5 showed relative high activity and NO oxidation reached the maximum of 53% at 325 °C. Then the NO oxidation efficiency decreased with further increasing the temperature due to the thermodynamic limits.⁵² These results suggest that Fe-ZSM-5 has better catalytic activity to promote NO oxidation ability than Fe₂O₃, and thus accelerate the SCR reaction remarkable. The performance of Fe_H-ZSM-5 is very similar to that of Fe-ZSM-5, but it showed a higher activity at low temperature side, which could be attributed to the highly dispersed Fe_xO_y clusters in Fe_H-ZSM-5.

Fig. 8 Oxidation of NO to NO₂ by O₂ over the catalysts. Reaction conditions: [NO] = 800 ppm, [O₂] = 5%, balanced by N₂ at total flow rate of 1000 sccm and a GHSV of 100 000 h⁻¹.

According to Fig. 8, the apparent E_a for NO oxidation over Fe_H-ZSM-5 and Fe-ZSM-5 was calculated to be 13.35 and 24.98 kJ mol⁻¹, respectively. Note that the apparent E_a of NO oxidation over Fe_H-ZSM-5 was close to the E_a of NO_X reduction at temperature below 200 °C (8.98 kJ mol⁻¹), which suggests that the NO oxidation played an important role in the rate-determining step during NO reduction below 200 °C.

Adsorption of NO_X species on the catalyst surface was reported to play an important role in NO_X reduction.⁵³ Fig. 9 shows the NO_X desorption profiles of this series of catalysts. Fe₂O₃ is good adsorbent for both NO and NO₂, and H-ZSM-5 is not a good adsorbent for either NO or NO₂. Fe-ZSM-5 only adsorbed NO₂, moreover, Fe_H-ZSM-5 was also found to adsorb NO₂ only, which strongly suggested the existence of interaction between highly dispersed Fe_xO_y clusters and zeolite supporter.

Fig. 9 NO-TPD and NO₂-TPD profiles of the catalysts.

3.7 NH₃ oxidation analysis

In SCR reaction, ammonia activation is considered as one of the most important parameters to promote the NO_X conversion, especially over traditional V₂O₅-based catalysts.⁵⁴ Fig. 10 shows the NH₃ oxidation over this series of catalysts. NH₃ conversion over all the catalysts kept at almost 0% below 225 °C. Then NH₃ conversion increased with temperature. H-ZSM-5 showed low activity for NH₃ oxidation even at high temperature. The Fe-ZSM-5 catalyst only showed a limited activity promotion compared to H-ZSM-5 due to the low Fe³⁺ loading. Interestingly, the NH₃ oxidation over Fe_H-ZSM-5 was found to be higher than that over pure Fe₂O₃ and reached 56% at 400 °C, indicating that highly dispersed Fe_xO_y clusters promoted the NH₃ activation, since the Fe_H-ZSM-5 performed better than Fe-ZSM-5 which has the same amount of Fe exchanged in ZSM-5, and even better than pure Fe₂O₃. Furthermore, during NH₃ oxidation, NO or NO₂ species was not detected over all catalysts, indicating no NH₃ over oxidation, which should be beneficial for SCR NO_X removal at high temperature.

Fig. 10 NH₃ oxidized by O₂ over the catalysts. Reaction conditions: [NH₃] = 800 ppm, [O₂] = 5%, balanced by N₂ at total flow rate of 1000 sccm and a GHSV of 100 000 h⁻¹.

3.8 Discussions

Fe-ZSM-5 has been widely studied as catalyst for the ammonia SCR reaction. The Fe species in Fe-ZSM-5 were partly exchanged into zeolite, and partly loaded on the zeolite surface as clusters.^39–41 On the one hand, Fe in Fe-ZSM-5 was reported to promote the oxidation of NO to NO₂. On the other hand, the loaded Fe ions replaced the H⁺, which caused the slightly decrease of NH₃ adsorption.^26,55 Fortunately, the reaction between NO and O₂, but not the NH₃ adsorption, was considered as the rate-determining step.⁵⁶ As a result, Fe-ZSM-5 promoted the NO_X reduction. This is supported by another report which found that the SCR activity of Fe-ZSM-5 is tremendously improved by the addition of NO₂ to the feed gas, especially at low reaction temperature.⁵² Further study suggested that the role of NO₂ in the fast SCR reaction was to form surface nitrites and nitrates, while the role of NO is to reduce nitrates to nitrites, and then nitrates further reacted with NO to produce more nitrides, finally nitrites were removed from the surface by NH₃ to produce N₂, the reduction of nitrates by NO is the rate-limiting step in this fast SCR at 200 °C.⁵⁷ While at high temperature above 300 °C, the NO_X conversion in SCR reaction is mainly governed by ammonia activation since the NO oxidation is suppressed due to the thermodynamic limitation. However the undesirable over oxidation of NH₃ to NO_X occurred easily at high temperature.^54,58,59

In the present study, according to NO_X TPD and NO oxidation results, H-ZSM-5 showed no ability for either NO to NO₂ oxidation or NO + NO₂ adsorption, so NO_X conversion could not happen over H-ZSM-5 at low temperature. At temperature above 350 °C, H-ZSM-5 started to activate NH₃ as shown in Fig. 7 and 10, and this promoted NO_X reduction and a NO_X conversion up to 15% was obtained. For, Fe-ZSM-5, it started to reduce NO at 150 °C and the reduction increased quickly at temperature above 250 °C, which is in good agreement with the NO oxidation and adsorption as shown in Fig. 8 and 9.^13,52 Note that the Fe³⁺ concentration is low in Fe-ZSM-5, which indicated that the isolated Fe³⁺ sites in tetrahedral and higher coordination are excellent active species for NO_X reduction. At temperature above 325 °C, the NO to NO₂ oxidation were reduced due to the thermodynamic limit,⁵² and NO₂ adsorption also reduced, however, the NO_X reduction still increased, which could be attributed to the NH₃ activation at high temperature as shown in Fig. 7 and 10.

For Fe₂O₃, the oxidation of NO to NO₂ was poor, even poorer than that of H-ZSM-5, but it adsorbed NO and NO₂ as shown in Fig. 9, and it also activated the NH₃ above 250 °C, which enhanced the NO_X reduction up to 62% at 350 °C, at temperature above 350 °C, NO and NO₂ adsorptions were suppressed, and NH₃ was also totally desorbed (Fig. 7), which finally decreased the NO_X reduction.

The detrimental of the Fe-ZSM-5 is its low Fe³⁺ concentration (less than 2%)⁹, fortunately the surface loaded Fe_xO_y clusters are also active species, and thus in the case of Fe_H-ZSM-5, the catalyst showed 20% of NO oxidation even at 100 °C (Fig. 8), this promoted its activity and a NO_X conversion of 20% was achieved at 100 °C. In fact, as shown in Fig. 4, more than 70% of Fe species existed as surface Fe_xO_y clusters. At temperature above 350 °C, though the NO₂ adsorption and NO to NO₂ oxidation was suppressed, but this was compensated by the promotion of NH₃ activation benefited from Fe_xO_y clusters as shown in Fig. 10, which still promoted the NO_X reduction, and a NO_X conversion of 91% was achieved at 400 °C.

From the view of SCR reactions, NO₂ is involved in the following reactions:

6NO₂ + 8NH₃ → 7N₂ + 12H₂O (3)

2NO₂ + 4NH₃ + O₂ → 3N₂ + 6H₂O (4)

2NO + 2NO₂ + 4NH₃ → 4N₂ + 6H₂O (5)

Among them, reaction (5) is the well-known fast SCR, while in reactions (3) and (4), NO does not react with NH₃ directly, thus the E_a would be less than that of the reaction (5) in which NO was involved. Taken the catalyst Fe_H-ZSM-5 as an example, from Fig. 1a and 8, we can find that the NO_X removal efficiency is almost the same as the NO to NO₂ oxidation below 200 °C, which suggested that NO_X removal mainly follows reactions (3) and/or (4). At temperature between 200 and 325 °C, the NO_X removal efficiency is approximately twice as that of the NO to NO₂ oxidation, indicating that reaction (5) becomes the main route. At temperature above 250 °C, NH₃ activation become notable. Therefore, the SCR reaction can be divided into three stages here: In stage-1 (<200 °C), the main reaction is between NO₂ and non-activated NH₃; in stage-2 (200–325 °C), the main reaction is NO + NO₂ and non-activated NH₃; while in stage-3 (>300 °C), the reaction between NO + NO₂ and activated NH₃ becomes the main route, and the reaction between NO and activated NH₃ also occurred considering from the difference between NO to NO₂ oxidation and the NO removal efficiency. Accordingly, the lowest E_a was needed in stage-1, and the highest E_a was needed in stage-2, in good agreement with our E_a calculation in Fig. 1b.

4. Conclusions

In summary, highly dispersed Fe_xO_y clusters with a Fe³⁺ concentration of 22 wt% was successfully loaded on ZSM-5, the Fe_xO_y clusters played a key role for the promotion of de-NO_X activity, and a good NO_X removal efficiency of 91% was achieved at 400 °C. Our results discovered that the NO to NO₂ oxidation is essential for NO_X removal at temperature below 200 °C. At temperature between 200 °C and 325 °C, fast SCR reaction became the main route for NO_X reduction. The NH₃ activation at temperature above 300 °C promoted the reaction between NO and NH₃, which compensated thermodynamic limitation induced suppression of fast SCR. And the suppression of NH₃ to NO over oxidation promised a high NO reduction efficiency at 400 °C.

Acknowledgements

This work was supported by the Environmentally Sustainable Management of Medical Wastes in China (Contract No. C/V/S/10/251), and the National Natural Foundation of Zhejiang Province, China (Grant No. Z4080070).

References

1. F. C. Meunier, J. P. Breen, V. Zuzaniuk, M. Olsson and J. R. H. Ross, J. Catal., 1999, 187, 493–505.
2. G. Busca, L. Lietti, G. Ramis and F. Berti, Appl. Catal., B, 1998, 18, 1–36.
3. G. Busca, M. A. Larrubia, L. Arrighi and G. Ramis, Catal. Today, 2005, 107–108, 139–148.
4. G. Ramis, L. Yi and G. Busca, Catal. Today, 1996, 28, 373–380.
5. Z. H. Lian, F. D. Liu and H. He, Catal. Sci. Technol., 2015, 5, 389–396.
6. S. W. Choi, S. K. Choi and H. K. Bae, J. Air Waste Manage. Assoc., 2015, 65, 485–491.
7. A. Yamamoto, Y. Mizuno, K. Teramura, S. Hosokawa, T. Shishido and T. Tanaka, Catal. Sci. Technol., 2015, 5, 556–561.
8. X. Y. Fan, F. M. Qiu, H. S. Yang, W. Tian, T. F. Hou and X. B. Zhang, Catal. Commun., 2011, 12, 1298–1301.
9. R. Q. Long and R. T. Yang, J. Catal., 1999, 188, 332–339.
10. Z. X. Ma, H. S. Yang, F. Liu and X. B. Zhang, Appl. Catal., A, 2013, 467, 450–455.
11. Z. X. Ma, H. S. Yang, Q. Li, J. W. Zheng and X. B. Zhang, Appl. Catal., A, 2012, 427–428, 43–48.
12. Q. Li, X. X. Hou, H. S. Yang, Z. X. Ma, J. W. Zheng, F. Liu, X. B. Zhang and Z. Y. Yuan, J. Mol. Catal. A: Chem., 2012, 356, 121–127.
13. Q. Li, H. S. Yang, F. M. Qiu and X. B. Zhang, J. Hazard. Mater., 2011, 192, 915–921.
14. W. Tian, H. S. Yang, X. Y. Fan and X. B. Zhang, J. Hazard. Mater., 2011, 188, 105–109.
15. R. Q. Long, R. T. Yang and R. Chang, Chem. Commun., 2002, 5, 452–453.
16. R. Q. Long and R. T. Yang, J. Catal., 2001, 201, 145–152.
17. K. Krishna and M. Makkee, Catal. Today, 2006, 114, 23–30.
18. S. Brandenberger, O. Kröcher, A. Tissler and R. Althoff, Catal. Rev., 2008, 50, 492–531.
19. P. S. Metkar, V. Balakotaiah and M. P. Harold, Catal. Today, 2012, 184, 115–128.
20. E. Kolobova, A. Pestryakov, A. Shemeryankina, Y. Kotolevich, O. Martynyuk, H. J. T. Vazquez and N. Bogdanchikova, Fuel, 2014, 138, 65–71.
21. A. Sultana, T. Nanba, M. Haneda and H. Hamada, Catal. Commun., 2009, 10, 1859–1863.
22. H. Sjövall, R. J. Blint and L. Olsson, Appl. Catal., B, 2009, 92, 138–153.
23. A. Uddin, T. Komatsu and T. Yashima, J. Chem. Soc., Faraday Trans., 1995, 91, 3275–3279.
24. A. Z. Ma and W. Grünert, Chem. Commun., 1999, 71–72.
25. H. Y. Chen, T. Voskoboinikov and W. M. H. Sachtler, J. Catal., 1998, 180, 171–183.
26. R. Q. Long and R. T. Yang, J. Catal., 2002, 207, 224–231.
27. V. V. Lissianski, P. M. Maly and V. M. Zamansky, Ind. Eng. Chem. Res., 2001, 40, 3287–3293.
28. G. Qi and R. T. Yang, Catal. Lett., 2005, 100, 243–246.
29. J. A. Z. Pieterse, S. Booneveld and R. W. van den Brink, Appl. Catal., B, 2004, 51, 215–228.
30. Q. Li, H. S. Yang, A. M. Nie, X. Y. Fan and X. B. Zhang, Catal. Lett., 2011, 141, 1237–1242.
31. J. H. Goo, M. F. Irfan, S. D. Kim and S. C. Hong, Chemosphere, 2007, 67, 718–723.
32. X. Y. Guo, C. Bartholomew, W. Hecker and L. L. Baxter, Appl. Catal., B, 2009, 92, 30–40.
33. D. S. Zhou, Z. Y. Ren, B. Li, Z. X. Ma, X. B. Zhang and H. S. Yang, RSC Adv., 2015, 40, 31708–31715.
34. M. M. J. Treacy, J. B. Higgins and R. von Ballmoos, Zeolites, 1996, 16, 330–802.
35. A. Cihanoglu, G. Gündüz and M. Dükkancı, Appl. Catal., B, 2015, 165, 687–699.
36. M. Rauscher, K. Kesore, R. Monnig, W. Schwieger, A. Tissler and T. Turek, Appl. Catal., A, 1999, 184, 249–256.
37. G. Huo, X. G. Lu, Y. Huang, W. Y. Li and G. Y. Liang, J. Electrochem. Soc., 2014, 161, 1144–1148.
38. B. M. Reddy, K. N. Rao, G. K. Reddy, A. Khan and S. E. Park, J. Phys. Chem. C, 2007, 111, 18751–18758.
39. S. Bordiga, R. Buzzoni, F. Geobaldo, C. Lamberti, E. Giamello, A. Zecchina, G. Leofanti, G. Petrini, G. Tozzola and G. Vlaic, J. Catal., 1996, 158, 486–501.
40. G. Lehmann, Z. Phys. Chem., 1970, 72, 279.
41. M. Santhosh Kumar, M. Schwidder, W. Grünert and A. Brückner, J. Catal., 2004, 227, 384–397.
42. M. M. Jin, R. G. Yang, M. F. Zhao, G. Y. Li and C. W. Hu, Ind. Eng. Chem. Res., 2014, 53, 2932–2939.
43. J. A. Botas, J. A. Melero, F. Martınez and M. I. Pariente, Catal. Today, 2010, 149, 334–340.
44. F. D. Liu, H. He, C. B. Zhang, Z. C. Feng, L. R. Zheng, Y. N. Xie and T. D. Hu, Appl. Catal., B, 2010, 96, 408–420.
45. F. Heinrich, C. Schmidt, E. Löffler, M. Menzel and W. Grünert, J. Catal., 2002, 212, 157–172.
46. H. Y. Chen and W. M. H. Sachtler, Catal. Today, 1998, 42, 73–83.
47. H. L. Huang, Y. Lan, W. P. Shan, F. H. Qi, S. C. Xiong, Y. Liao, Y. W. Fu and S. J. Yang, Catal. Lett., 2014, 144, 578–584.
48. T. Miyamoto, N. Katada, J. H. Kim and M. Niwa, J. Phys. Chem. B, 1998, 102, 6738–6745.
49. Y. M. Ni, A. M. Sun, X. L. Wu, G. L. Hai, J. L. Hu, T. Li and G. X. Li, Microporous Mesoporous Mater., 2011, 143, 435–442.
50. Y. H. Seo, E. A. Prasetyanto, N. Z. Jiang, S. M. Oh and S. E. Park, Microporous Mesoporous Mater., 2010, 128, 108–114.
51. G. Qi and R. T. Yang, Appl. Catal., B, 2005, 60, 13–22.
52. M. Devadas, O. Kröcher, M. Elsener, A. Wokaun, N. Söger, M. Pfeifer, Y. Demel and L. Mussmann, Appl. Catal., B, 2006, 67, 187–196.
53. M. Koebel, G. Madia and M. Elsener, Catal. Today, 2002, 73, 239–247.
54. S. Roy, B. Viswanath, M. S. Hegde and G. Madras, J. Phys. Chem. C, 2008, 112, 6002–6012.
55. R. Q. Long and R. T. Yang, J. Catal., 2001, 198, 20–28.
56. H. Y. Huang, R. Q. Long and R. T. Yang, Appl. Catal., A, 2002, 235, 241–251.
57. A. Grossale, I. Nova, E. Tronconi, D. Chatterjee and M. Weibel, J. Catal., 2008, 256, 312–322.
58. J. H. Kwak, R. Tonkyn, D. Tran, D. H. Mei, S. J. Cho, L. Kovarik, J. H. Lee, C. H. F. Peden and J. Szanyi, ACS Catal., 2012, 2, 1432–1440.
59. S. Suárez, S. M. Jung, P. Avila, P. Grange and J. Blanco, Catal. Today, 2002, 75, 331–338.

---