Mn/SAPO-34 as an efficient catalyst for the low-temperature selective catalytic reduction of NO_x with NH₃

Abstract

A series of Mn-exchanged SAPO-34 catalysts were synthesized and developed as catalysts for the low temperature selective catalytic reduction (SCR) of NO with ammonia in the presence of excess oxygen. 4Mn/SAPO-34 showed the highest SCR activity in the temperature range of 120–210 °C. The microstructure of zeolite supports, acidity, manganese species and reaction mechanism were investigated in detail by BET, NH₃-TPD, UV-Vis, H₂-TPR, XPS and in situ DRIFTS. Mn³⁺ and Mn⁴⁺ related species were proved to be the active sites for the SCR reaction. The Eley–Rideal mechanism was proved to be effective on 4Mn/SAPO-34 for the low temperature SCR reaction in which coordinated NH₄⁺ species react with gas-phase NO to form an activated transition state and subsequently decomposed to N₂ and H₂O.

---

1. Introduction

Nitrogen oxides, including NO, NO₂ and N₂O have long been considered as the major source of air pollution causing photochemical smog, acid rain, ozone depletion and the greenhouse effect. One of the most effective techniques to eliminate NO_x pollution is SCR technology with ammonia as the reductant (4NO + 4NH₃ + O₂ → 4N₂ + 6H₂O). The commercial SCR catalyst V₂O₅–MoO₃(WO₃)/TiO₂ only works efficiently under 300–400 °C. However, for many industrial applications, such as cement plants, iron and steel plants, especially gas-fired boilers, the suitable process temperature to arrange NO_x removal is usually below 200 °C, thus making it necessary to develop low temperature SCR catalysts to avoid reheating of the flue gas and reduce cost.

Many metal oxide catalysts have long been investigated for the low-temperature SCR reaction, such as MnO_x–CeO_x,^1,2 Mn–Ce–Ti mixed-oxide,³ W–Mn–Zr mixed-oxide,⁴ Ni–Mn/TiO₂.⁵ These researchers had attributed the excellent SCR activity to Mn³⁺ and Mn⁴⁺ species. The most recent SCR catalyst formulations contain copper exchanged into the chabazite (CHA) family of zeolites. Cu-SSZ-13 (ref. 6) and Cu-SAPO-34 (ref. 7 and 8) catalysts attracted much attention for their utility in NH₃–SCR reactions and had demonstrated good low-temperature SCR activity and hydrothermal stability.

Besides high surface area and good hydrothermal stability, SAPO-34 can offer many acid sites of different strength which are important to SCR activity, but Mn–M/TiO_x (M = Ce, Fe, etc.) catalysts usually lack of, therefore manganese oxides catalyst with SAPO-34 as the carrier might show good low temperature SCR activity.

In terms of the reaction mechanism, two still-under-debate mechanism have been proposed for the low-temperature SCR, some agree that the Eley–Rideal mechanism occurs, i.e. the gaseous NO react with activated NH₃ species to an activated transition state and subsequently decomposed to N₂ and H₂O;^8,9 and some claim the Langmuir–Hinshelwood mechanism occurs, i.e. adsorbed NO or NO₂ species react with adsorbed NH₃ species on the adjacent sites, followed by reaction to an activated transition state and decomposition to the reaction products.^10,11

In the present work, the treatment of exhaust from gas-fired boiler which contains high concentration of H₂O and very low concentration of SO₂ is focused. So xMn/SAPO-34 (x = 1, 2, 4, 8) was developed as a new catalyst for low temperature SCR with NH₃. With the characterization of H₂-TPR, XPS and UV-Vis, the active site could be determined. Furthermore, the SCR mechanism on Mn/SAPO-34 was investigated by in situ DRIFTS study.

2. Experimental

2.1 Catalyst preparation

The ion-exchanged xMn/SAPO-34 (x = 1, 2, 4, 8) catalyst was prepared by liquid ion-exchange method. First in order to avoid poison of catalysts from alkali metal which decrease catalytic activity for low temperature SCR reaction, commercial Na-SAPO-34 powder (Jiangsu XFNANO) was converted to NH₄/SAPO-34 by ion exchange with NH₄Cl (Aladdin, purity > 98.5%) solution. Then Mn ion-exchange was performed by mixing NH₄/SAPO-34 with different concentration of Mn(NO)₂ (Across, purity > 95%) solution at 80 °C for 6 h under vigorously stirring. Then it was dried in rotary evaporator and in an oven at 110 °C for 16 h and finally xMn/SAPO-34 (x = 1, 2, 4, 8) was obtained by calcination in the gas of 21% O₂/N₂ at 550 °C for 3 h.

2.2 NH₃ SCR activity measurements

The catalytic activity evaluation was carried out using a flow-through powder reactor system equipped with a Fourier Transform Infrared (FT-IR) spectrometer (THERMO SCIENTIFIC IGS) with a gas-sampling cell. Fig. 1 shows the schematic diagram of experimental apparatus.

Fig. 1 Schematic diagram of experimental apparatus for activity test.

In order to prevent condensation along upstream tubing, all the gas lines were heated and maintained at 120 °C. The gas mixture for the NH₃-SCR reaction consisted of 500 ppm NO, 500 ppm NH₃, 16.3% vol. H₂O (if in the water resistance test) and 3% vol. O₂ balanced with N₂. The gas hourly space velocity (GHSV) was 65 000 h⁻¹ for the standard SCR. Prior to the activity measurements, the catalysts were pretreated at 500 °C and held for 30 min in 21% vol. O₂/N₂ flow. The catalytic activities were measured at the temperature range of 120–210 °C. The typical time to achieve steady state at each temperature was about 2.5 h. The NO conversions were calculated based on the inlet and outlet gas concentrations at steady state by the following equation:

2.3 Characterization

X-ray diffraction patterns were collected on a bruker D8 Advance X-ray diffractometer with a Ni-Filtered Cu Kα with a step size of 0.02 in the 2θ range from 5° to 40°.

The ammonia temperature programmed desorption (NH₃-TPD) experiment was carried out in quartz tube reactor equipped with a Fourier Transform Infrared (FT-IR) spectrometer (THERMO SCIENTIFIC IGS) with a gas-sampling cell. 150 mg of prepared samples were firstly pretreated in 21% vol. O₂/N₂ at 500 °C for 30 min. Then a total flow of 100 ml min⁻¹ containing 2500 ppm NH₃ in N₂ was injected into the reactor for 2 hours to achieve steady state. Once the catalyst was saturated, NH₃ was switched off and the catalyst was swept by N₂ overnight to remove any gas phase NH₃. Finally the catalyst was heated in N₂ at a temperature ramp to 700 °C with a heating rate of 10 °C min⁻¹.

The manganese contents were determined by coupled plasma (ICP) optical emission spectroscopy (Thermo IRIS Intrepid II) of dissolutions of the ground catalysts in acid solutions.

The BET surface area, pore volume and pore size of samples were measured by N₂ adsorption using a MICROMERITICS ASAP 2020 surface area and porosity analyzer.

Diffuse reflectance UV-Vis spectra were recorded in the range of 200–800 nm against a BaSO₄ reference standard on a HitachiU-300 UV-Vis-NIR spectrophotometer equipped with an integration sphere.

The H₂-TPR experiments were performed on a custom made setup using 20 mg of calcined catalyst. The catalysts were pretreated at 500 °C for 30 min in a highly pure O₂ (40 ml min⁻¹) stream to renew the samples and make sure the surface was clean. The furnace temperature was lowered to room temperature in N₂, and then the feed containing 5% vol. H₂ in N₂ was fed at a flow rate of 40 ml min⁻¹. H₂-TPR runs were performed by heating the samples from room temperature to 800 °C at a linear heating rate of 10 °C min⁻¹ and finally keeping the temperature constant for 30 min at 800 °C to ensure complete metal oxide reduction. The hydrogen consumed in the TPR was measured by a TCD.

X-ray photoelectron spectroscopy (XPS) analyses were performed on a PHI quantera SXM Scanning ESCA Microprobe (Physical Electronics) with a hemispherical detector operating at constant pass energy (PE = 55 eV) using Al Kα radiation (1486.6 eV). All binding energies were referenced to the C 1s line at 284.8 eV.

Diffuse reflectance infrared Fourier transform spectra (DRIFTS) were measured on a FT-IR spectrometer (Thermo Nicolet NEXUS870) with a MCT detector and a high temperature reaction chamber (Harrick Scientific Praying Mantis) with ZnSe windows, which was connected to a gas-dosing system. The oxidation pretreatments at 500 °C for 1 h were executed to remove the adsorbed water and clean the surface before each measurement. The DRIFTS spectra were recorded in the range of 4000–650 cm⁻¹ for 32 scans with a resolution of 4 cm⁻¹. The background spectra were collected for 32 scans with a resolution of 4 cm⁻¹ before exposed to the absorbates.

3. Results and discussion

3.1 Effect of Mn loading on NH₃ SCR activity

In low temperature SCR experiments, the NO_x and NH₃ conversions were tested as a function of steady temperature (ranging from 120 to 210 °C) over xMn/SAPO-34 catalysts. Fig. 2a shows the NO conversion profile of xMn/SAPO-34 (x = 1, 2, 4, 8) catalysts. The maximum NO conversion of NH₄/SAPO-34 is about 20%, which is consistent with the previous results on H-ZSM-5, H-BETA and H-SSZ-13, where the proton form zeolite was not very active for standard SCR reaction.^12,13 Compared with Mn/SAPO-34 catalysts, NH₄/SAPO-34 does not have much catalytic activity in the whole temperature range (120–210 °C), indicating that the Mn species could be of the great importance to the SCR activities.

Fig. 2 Catalytic activity of xMn/SAPO-34 (x = 1, 2, 4, 8) for low-temperature SCR of NO with NH₃. Conditions: (a) 500 ppm NO + 500 ppm NH₃ + 3% vol. O₂ + N₂ balanced, GHSV = 65 000 h⁻¹; (b) 500 ppm NO + 500 ppm NH₃ + 16.3% vol. H₂O + 3% vol. O₂ + N₂ balanced, GHSV = 65 000 h⁻¹.

When Mn loading increased from 1% to 4%, the catalytic activity increased significantly from 26% to 82% at 180 °C. When Mn loading increased to 8%, catalytic activity drops to 58% at 180 °C, this might be caused by the blockage of micro pores and the decrease of exposed manganese due to the aggregation of manganese oxides.

Compared with other researches, Lei Ma¹⁴ reported that for H₂O resistance test, the SCR activity of Cu-SAPO-34 catalyst at 150 °C was about 30% when GHSV = 30 000 h⁻¹ and the concentration of NO and NH₃ was 350 ppm. So comparing with Cu-SAPO-34 Lei Ma had made, the Mn/SAPO catalyst showed much better SCR activity and H₂O tolerance. Boxiong Shen¹⁵ reported that when GHSV = 5000 h⁻¹ and the concentration of NO and NH₃ was 600 ppm, the SCR activity of Mn–Ce/Ti–PILCs prepared at 120 °C was 50%. Comparing with Mn–Ce/Ti–PILCs, the Mn/SAPO catalyst showed similar SCR activity while Ce was not added to increase SCR activity of Mn/SAPO yet.

Fig. 2b shows the NO conversion profile of the xMn/SAPO-34 (x = 1, 2, 4, 8) catalysts of water resistance test. For 4Mn/SAPO-34 and 8Mn/SAPO-34 catalysts, the introduction of H₂O retard low temperature SCR reaction, while for 1Mn/SAPO and 2Mn/SAPO catalysts, the existence of H₂O in the gas phase improve the catalytic activity except for 120 °C. It might be explained that the introduction of H₂O might change the environment of Mn ions or generate acid sites that help increase SCR activity besides covering active sites as A. Lorena Picone etc.¹⁶ found the introduction of H₂O increased the SCR activity of Cu-SAPO STA-7(IE) and Raquel Martínez-Franco¹⁷ found the hydrothermal treatment could increase the SCR activity of Cu-SAPO-34.

The concentration of generated NO₂ was below 30 ppm and there was almost no N₂O generated, according to eqn (2), the selectivity was almost 100% in the temperature range of 120–210 °C, so it is beyond discussion.

3.2 XRD

Fig. 3 shows the XRD patterns of xMn/SAPO-34 with different Mn loadings. XRD is commonly used for detection of metal or metal oxides phase. As shown in Fig. 3, the diffraction peaks of xMn/SAPO-34 exhibit chabazite phase with space group of R3̄m. This observation agrees well with those of SAPO-34,^6,7 indicating the crystalline structure of SAPO-34 unchanged after catalyst preparation. As Mn loading of xMn/SAPO-34 increased, no manganese oxide phase had been detected except for 8Mn/SAPO-34 in which two peaks belonging to MnO₂ appears, indicating the aggregation of manganese oxide on 8Mn/SAPO-34. This is also supported by BET surface area results. For other samples, the manganese oxide might be amorphous or too small to be detected.

Fig. 3 XRD patterns of xMn/SAPO-34 (x = 1, 2, 4, 8).

3.3 Effect of Mn loading on surface area and pore characterization

As shown in Table 1, when Mn loading increased from 1 to 8%, BET surface area and micro pore volume decreased gradually, while micro pore diameter was almost the same except for 8Mn/SAPO-34 that it increased. According to XRD results, this might be caused by the blockage of micro pores due to the aggregation of manganese oxides.

Table 1 Physico-chemical properties of xMn/SAPO-34 (x = 1, 2, 4, 8)

Sample	BET surface area (m^2 g^−1)	Micro pore volume (cm^3 g^−1)	Micro pore diameter (nm)	Mn loading from ICP (%wt)
1Mn/SAPO-34	426	0.227	2.05	0.996
2Mn/SAPO-34	402	0.218	2.04	1.947
4Mn/SAPO-34	393	0.205	2.04	4.117
8Mn/SAPO-34	375	0.205	2.42	7.484

---

The Mn loading was almost the same as the set values for each catalyst indicating the catalyst preparing process was acceptable.

3.4 NH₃-TPD

The NH₃-TPD can be used to study the acidity property of each sample in different temperature. Fig. 4 shows the effluent NH₃ profiles during NH₃-TPD process on each sample. As Mn loading increased, the desorption amount of NH₃ decreased, this could be explained by Mn ion exchange process in which Mn ions might replace H⁺, leading to the decreasing of acid sites on catalysts. In Fig. 4, there were four NH₃ desorption peaks: a low temperature peak at 220 °C, two middle temperature peaks at 320 and 400 °C, and a high temperature peak at 500 °C. Different desorption peaks represent different NH₃ adsorption strength. Before 220 °C, the shapes of all the profiles were almost the same, while according to Fig. 2 the activities of these catalysts differ a lot, suggesting that these related NH₃ adsorption sites may not be involved in catalytic cycle or in another word, these adsorbed NH₃ were not well activated. This was supported by D. Wang et al.¹⁸ that the appearance of the low temperature peak at about 200 °C was attributed to weakly adsorbed NH₃.

Fig. 4 NH₃-TPD of xMn/SAPO-34 (x = 1, 2, 4, 8).

Whichever mechanism might dominate, NH₃ needs to be adsorbed and well activated, therefore, the interactions between NH₃ and catalyst surface cannot be too weak like those desorbed at 200 °C or too strong. According to Fig. 1, the catalysis activity increased as Mn loading increased from 1% to 4%, while it dropped when Mn loading further increased to 8%. Since all the NH₃-TPD profiles were similar except that of 8Mn/SAPO-34, the different activity between 1Mn/SAPO-34, 2Mn/SAPO-34 and 4Mn/SAPO-34 could be ascribed to different Mn loading, while the decreased activity of 8Mn/SAPO might be ascribed to the decreased acid sites from which NH₃ desorbed at 400 and/or 500 °C. This indicated that the acid sites responsible for the NH₃ desorption at 400 °C or/and 500 °C were important to low temperature SCR reaction.

3.5 DR UV-Vis

Diffuse reflectance UV-Vis spectroscopy is the common characterization method for the determination of oxidation and coordination state of metal complexes on the out surface and in the zeolite channels. Fig. 5 shows the UV-Vis spectra of all catalysts. The absorption bands of all samples at 240 nm were related to the charge transfer processes between framework aluminum and oxygen atoms of aluminophosphate and water molecules.¹⁹ The broad band in the range 320–380 nm was attributed to Mn³⁺ ← O²⁻ charge transfer transition superimposed on ⁵B_1g → ⁵B_2g crystal field d–d transition.²⁰ The band at 332 nm was tentatively assigned to Mn³⁺ ← O²⁻ charge transfer in Mn₃O₄ in which manganese was octahedrally coordinated with oxygen.^21–23 The band at 255–276 nm could be assigned to the Mn²⁺ ← O²⁻ charge transfer transition in tetrahedral oxygen coordination.^21,22 In the α-Mn₂O₃ structure, Mn³⁺ ions occupy octahedral sites, if highly symmetric, a single spin-allowed absorption band in the d–d transition region was expected similarly to [Mn(H₂O)₆]³⁺ at 500 nm.

Fig. 5 DR UV-Vis spectra recorded under ambient atmosphere of xMn/SAPO-34 (x = 1, 2, 4, 8).

In Fig. 5 it showed there was a distinct difference between the spectra of 2Mn/SAPO-34 and 4Mn/SAPO-34, while the spectra of 1Mn/SAPO-34 and 2Mn/SAPO-34 were similar as well as that of 4Mn/SAPO-34 and 8Mn/SAPO-34. Specifically, when Mn loading increased from 2% to 4%, the band at 320–380 nm increased greatly, indicating the percentage of Mn³⁺ increased. Meanwhile the negative band at 255–276 nm emerged, indicating the percentage of Mn²⁺ decreased, which might be caused by partial oxidation of Mn²⁺ to Mn³⁺ or Mn⁴⁺ when Mn loading increased.

3.6 H₂-TPR

Fig. 6 presented the H₂-TPR profiles of four catalysts. H₂-TPR is commonly used for information on the oxidation state of metal oxides. For 4Mn/SAPO-34 and 8Mn/SAPO-34 catalysts, there were two sharp peaks respectively at about 350 °C and 440 °C, the former one indicated the reduction of Mn⁴⁺ to Mn²⁺, and the latter one indicated the reduction of Mn³⁺ to Mn²⁺.^23,24 For 1Mn/SAPO-34 and 2Mn/SAPO-34 catalysts, there was only a broad peak between 300–500 °C, indicating that most of the Mn species on these two catalysts were Mn²⁺ which could not be reduced below 700 °C. As indicated, there was a significant change when Mn loading increased from 2% to 4% that the Mn²⁺ species changed to Mn³⁺ and/or Mn⁴⁺ species. This result was consistent with UV-Vis results.

Fig. 6 H₂-TPR of xMn/SAPO-34 (x = 1, 2, 4, 8).

3.7 XPS

To investigate the surface chemical states of these catalysts, the XPS spectra of the Mn 2P_3/2 were obtained, as shown in Fig. 7. Two main peaks respectively assigned to Mn 2P_3/2 at 642.5 ev and Mn 2P_1/2 at 654 ev were observed. Comparing with those reported for MnO, Mn₂O₃ and MnO₂,²⁵ the measured binding energy of Mn 2P_3/2 in xMn/SAPO-34 (x = 1, 2, 4, 8) catalysts was slightly higher, it could be strong evidence that strong interactions existed between Mn species and SAPO-34 carrier.

Fig. 7 XPS spectra for Mn 2P of the xMn/SAPO-34 (x = 1, 2, 4, 8) catalysts.

By performing a peak-fitting deconvolution, Mn species could be separated into three peaks: 640.8–640.9 ev, 642.1–642.3 ev and 643.8–644.1 ev, which correspond to Mn²⁺, Mn³⁺ and Mn⁴⁺ respectively. Table 2 had listed the atom percentage of Mn species in different valence state determined by XPS. As Mn loading increased, the percentage of Mn³⁺ increased while the percentage of Mn⁴⁺ decreased. This observation is quite different from that obtained from UV-Vis and H₂-TPR spectra. The could be explained that the information acquired from XPS spectra is about the element on the outer surface of the samples, since for XPS spectra, only the information of the element about 3λ depth below the surface (λ is the depth of several atoms) can be obtained, while H₂-TPR spectra reflect the characteristics of bulk phase and UV-Vis spectra reflect the characteristics of metal oxides species on the outer and inner surface especially in zeolite channels of catalyst.^26,27

Table 2 Binding energies (ev) of core electrons of the xMn/SAPO-34 (x = 1, 2, 4, 8) catalysts and the percentage of different valence state

Compound	Mn^2+		Mn^3+		Mn^4+
	Peak (ev)	%	Peak (ev)	%	Peak (ev)	%
1Mn/SAPO-34	641.0	0.9	642.12	46.6	643.88	52.5
2Mn/SAPO-34	640.68	9.5	641.82	46.1	643.34	44.4
4Mn/SAPO-34	640.58	11.0	641.77	54.2	643.11	34.8
8Mn/SAPO-34	639.92	0.8	641.57	62.6	643.19	36.6

---

According to UV-Vis and H₂-TPR results, most of the Mn species on the surface of 1Mn/SAPO-34 and 2Mn/SAPO-34 were Mn²⁺, while that of 4Mn/SAPO-34 and 8Mn/SAPO-34 were Mn³⁺. The XPS spectra showed that, on outer surface of the samples, most of the Mn species were Mn³⁺ and Mn⁴⁺. These results indicated that for 1Mn/SAPO-34 and 2Mn/SAPO-34, most of the Mn species are Mn²⁺, the Mn species in the zeolite channels were Mn²⁺, while that on the outer surface were Mn³⁺ and Mn⁴⁺; for 4Mn/SAPO-34 and 8Mn/SAPO-34, most of the Mn species were Mn³⁺ and Mn⁴⁺.

3.8 DRIFTS

3.8.1 NO adsorption and subsequent reaction with NH₃. In order to identify the adsorbed NO species, in situ DRIFTS was performed on catalysts exposed to 500 ppm NO with 3% vol. O₂ balanced by N₂ (100 ml min⁻¹). In Fig. 8a, there was only one positive band at about 1360 cm⁻¹, which belonged to nitrite groups.^28,29 There was no other detectable positive bands between 1200–1600 cm⁻¹, suggesting that the nitrite groups might be the only NO species. The negative bands at 3300 cm⁻¹ and 2700 cm⁻¹ might originate from the depletion of Brønsted base sites, i.e. (Si–OH–Al), (P–OH) and OH groups associated with the extra-framework Al by nitrite groups. The bands at 3677 cm⁻¹, 3620 cm⁻¹ and 3600 cm⁻¹ might be assigned to the generation of O–H bond in the nitrite groups (–N–O–H) in different chemical environments.

Fig. 8 DRIFTS spectra of 4Mn/SAPO-34 at 150 °C exposed to (a) 500 ppm NO + 3% vol. O₂ + N₂ (balance) (b) 500 ppm NH₃ + 3% vol. O₂ + N₂ (balance) after saturated with NO (c) 500 ppm NH₃ + 3% vol. O₂ + N₂ (balance) (d) 500 ppm NO + 3% vol. O₂ + N₂ (balance) after saturated with NH₃ (e) 500 ppm NO + 500 ppm NH₃ + 3% vol. O₂ + N₂ (balance).

As shown in Fig. 8b, when NH₃ introduced into the system, the bands associated with nitrite groups decreased rapidly. Before 5 minutes when there was an obvious decrease for bands assigned to nitrite groups, there was no bands belonging to adsorbed NH₃, this might suggest that NH₃ moved fast on surface that they reacted with the nitrite groups and did not accumulate, and some of the groups on the surface of the samples might take the role of transferring ammonia groups to active sites.

3.8.2 NH₃ adsorption and subsequent reaction with NO. As shown in Fig. 8c, after NH₃ introduced into the system, the band at 1456 cm⁻¹ increased rapidly due to the generation of NH₃ species on BrØnsted acid sites,^8,30 and the growing band at 1210 cm⁻¹ (not shown) was assigned to the generation of NH₃ species on Lewis acid though it was very weak. This indicates that most of the acid sites were from BrØnsted acid sites. The appearance of negative bands at 3592 cm⁻¹, 3629 cm⁻¹ and 3674 cm⁻¹ indicated the reducing of BrØnsted acid sites and the forming of NH₄⁺ which was also supported by the band at 3283 nm⁻¹ assigned to NH₄⁺ groups.^7,9 Specifically, the negative band at 3674 cm⁻¹ was assigned to the occupation of P–OH site by NH₃, which consequently reduced the corresponding OH stretching vibrations. The changes of other two bands were related to the depletion of Si–OH–Al acid site, which was also supported by the existing negative band at 2600 cm⁻¹ attributed to the Si–OH–Al acid sites. The band at 3320–3550 cm⁻¹ was attributed to N–H stretching vibrations of NH₄⁺.³⁰ As shown in Fig. 8d, when NO introduced into the system, bands related to adsorbed NH₃ species decreased gradually, this could be the evidence that BrØnsted acid sites and the NH₄⁺ species were important to the low temperature SCR activity.

3.8.3 Reactions between NO and NH₃ under SCR reaction condition. Fig. 8e was similar to Fig. 8c that only NH₃ related bands were visible, any evidence of NO related bands did not appear especially the negative bands at 2700 and 3000–3400 cm⁻¹. It could be deduced that there is a great opportunity that the SCR reaction mechanism on the 4Mn/SAPO-34 at 150 °C was Eley–Rideal mechanism, since if Langmuir–Hinshelwood mechanism were effective, many NO would have been adsorbed on the surface due to the accumulation of NO species of which only the well activated parts had the appropriate energy to react with activated NH₃. In this case the negative bands at 2700 and 3000–3400 nm⁻¹ would appear, as Fig. 8a and b suggested. As these peaks were not shown in Fig. 8e, all these observations lead to the exclusion of Langmuir–Hinshelwood mechanism and the assurance of Eley–Rideal mechanism.

From the evidence above, a mechanism for low temperature SCR reaction might be figured out. Fig. 9 showed the proposed low temperature SCR mechanism on 4Mn/SAPO-34. Firstly, NH₃ was adsorbed on the BrØnsted acid sites of SAPO-34, and then interacted with the adjacent manganese oxides. After one of the hydrogen of NH₃ captured by Mn^α+–O²⁻-like species, the remaining NH₂⁻¹ species was transferred to Mn³⁺ ion and then oxidized to NH₂, i.e. the electron from NH₂⁻¹ was captured by Mn³⁺ ion. The NO molecule from gas phase attacked the NH₂ on Mn²⁺ ion and formed NH₂NO-like transition state and subsequently decomposed to N₂ and H₂O. The Mn²⁺ ion left was then oxidized by O₂ to Mn³⁺ and Mn³⁺–O²⁻-like species which further reacted with Mn^α+–O²⁻-H-like species to Mn³⁺.

Fig. 9 The proposed low temperature SCR mechanism on 4Mn/SAPO-34.

4. Conclusions

For Mn/SAPO-34 catalyst, as Mn loading increased from 1% to 8%, the UV-Vis and H₂-TPR results indicated that the main Mn species in the zeolite channels changed from Mn²⁺ to Mn³⁺ or/and Mn⁴⁺, the XPS showed that the main Mn species on the outer surface were Mn³⁺ and Mn⁴⁺ as the results of MnO_x aggregation on the extra-framework of SAPO-34.

The increased catalytic activity when Mn loading increased from 1% to 2% might be ascribed to the increase of Mn³⁺ or/and Mn⁴⁺ oxides of the outer surface since Mn³⁺ and Mn⁴⁺ species were mainly responsible for the low-temperature selective catalytic reduction activity. The higher SCR reaction activity of 4Mn/SAPO-34 and 8Mn/SAPO-34 was the result of the higher content of Mn³⁺ and Mn⁴⁺ species as compared with 2Mn/SAPO-34 and 1Mn/SAPO-34. The lack of exposed Mn oxides caused by aggregation and the BrØnsted acid sites on 8Mn/SAPO-34 as indicated by XRD and NH₃-TPD profile led to the lower catalytic activity compared with 4Mn/SAPO-34.

From the DRIFTS results, it could be deduced that NH₃ was mainly adsorbed on the BrØnsted acid, and the main mechanism on 4Mn/SAPO-34 at low temperature was Eley–Rideal mechanism, i.e. NO molecule from gas phase directly reacted with the well activated NH₃ species adsorbed on the surface and leaf as products to gas phase with the active site left.

Acknowledgements

This research is sponsored by China National Key Research and Development Program via Project No. 2016YFB0600603.

References

1. G. Qi and R. T. Yang, Performance and kinetics study for low-temperature SCR of NO with NH₃ over MnO_x–CeO₂ catalyst, J. Catal., 2003, 217, 434–441.
2. Z. Liua, Y. Yi, S. Zhang, T. Zhu, J. Zhu and J. Wang, Selective catalytic reduction of NO_x with NH₃ over Mn–Ce mixed oxide catalyst at low temperatures, Catal. Today, 2013, 216, 76–81.
3. Z. Liu, J. Zhu, J. Li, L. Ma and S. I. Woo, Novel Mn–Ce–Ti Mixed-Oxide Catalyst for the Selective Catalytic Reduction of NO_x with NH₃, ACS Appl. Mater. Interfaces, 2014, 6, 14500–14508.
4. Z. Liu, Y. Liu, Y. Li, H. Su and L. Ma, WO₃ promoted Mn–Zr mixed oxide catalyst for the selective catalytic reduction of NO_x with NH₃, Chem. Eng. J., 2016, 283, 1044–1050.
5. B. Thirupathi and P. G. Smirniotis, Nickel-doped Mn/TiO₂ as an efficient catalyst for the low-temperature SCR of NO with NH₃: Catalytic evaluation and characterizations, J. Catal., 2012, 288, 74–83.
6. J. J. Xue, X. Q. Wang, G. S. Qi, J. Wang, M. Q. Shen and W. Li, TiO₂-supported metal oxide catalysts for low-temperature selective catalytic reduction of NO with NH₃: I. Evaluation and characterization of first row transition metals, J. Catal., 2013, 297, 56–64.
7. L. Wang, W. Li, S. J. Schmieg and D. Weng, Role of Brønsted acidity in NH₃ selective catalytic reduction reaction on Cu/SAPO-34 catalysts, J. Catal., 2015, 324, 98–106.
8. D. Wang, L. Zhang, K. Kamasamudram and W. S. Epling, In situ-DRIFTS Study of Selective Catalytic Reduction of NO_x by NH₃ over Cu-Exchanged SAPO-34, ACS Catal., 2013, 3, 871–881.
9. M. Devadas, O. Kröcher, M. Elsener, A. Wokaun, N. Söger, M. Pfeifer, Y. Demel and L. Mussmann, Influence of NO₂ on the selective catalytic reduction of NO with ammonia over Fe-ZSM₅, Appl. Catal., B, 2006, 67, 187–196.
10. A. Grossale, I. Nova, E. Tronconi, D. Chatterjee and M. Weibel, The chemistry of the NO/NO₂–NH₃ “fast” SCR reaction over Fe-ZSM₅ investigated by transient reaction analysis, J. Catal., 2008, 256, 312–322.
11. M. Iwasaki and H. Shinjoh, A comparative study of “standard”, “fast” and “NO₂” SCR reactions over Fe/zeolite catalyst, Appl. Catal., A, 2010, 390, 71–77.
12. S. Alexander, C. Per-Anders, H. Hanna and S. Magnus, Effect of preparation procedure on the catalytic properties of Fe-ZSM-5 as SCR catalyst, Top. Catal., 2013, 56(9–10), 567–575.
13. F. Gao, E. D. Walter, E. M. Karpl, J. Y. Luo, R. G. Tonkyn, J. H. Kwak, J. Szanyi and C. H. F. Peden, Structure-activity relationships in NH₃-SCR over Cu-SSZ-13 as probed by reaction kinetics and EPR studies, J. Catal., 2013, 300, 20–29.
14. L. Ma, Y. Cheng, G. Cavataio, R. W. McCabe, L. Fu and J. Li, Characterization of commercial Cu-SSZ-13 and Cu-SAPO-34 catalysts with hydrothermal treatment for NH₃-SCR of NO_x in diesel exhaust, Chem. Eng. J., 2013, 225, 323–330.
15. B. Shen, H. Ma and Y. Yao, Mn–CeO_x/Ti–PILCs for selective catalytic reduction of NO with NH₃ at low temperature, J. Environ. Sci., 2014, 24, 499–506.
16. A. L. Picone, S. J. Warrender, A. M. Z. Slawin, D. M. Dawson, S. E. Ashbrook and P. A. Wright, etc., A co-templating route to the synthesis of Cu SAPO STA-7, giving an active catalyst for the selective catalytic reduction of NO, Microporous Mesoporous Mater., 2011, 146, 36–47.
17. R. Martínez-Franco, M. Moliner, P. Concepcion, J. R. Thogersen and A. Corma, Synthesis, characterization and reactivity of high hydrothermally stable Cu-SAPO-34 materials prepared by ‘‘one-pot’’ processes, J. Catal., 2014, 314, 73–82.
18. D. Wang, L. Zhang, J. H. Li, K. Kamasamudram and W. S. Epling, NH₃-SCR over Cu/SAPO-34 – Zeolite acidity and Cu structure changes as a function of Cu loading, Catal. Today, 2014, 231, 64–74.
19. M. A. Zanjanchi and M. K. Rashidi, Structural charge transfer in the aluminophosphate molecular sieves by diffuse reflectance spectroscopy, Spectrochim. Acta, Part A, 1999, 55, 947–954.
20. T. Kharlamova, G. Mamontov, M. Salaev, V. Zaikovskii, G. Popova, V. Sobolev, A. Knyazev and O. Vodyankina, Structural charge transfer in the aluminophosphate molecular sieves by diffuse reflectance spectroscopy, Appl. Catal., A, 2013, 467, 519–529.
21. Q. Tang, S. Hu, Y. Chen, Z. Guo, Y. Hu, Y. Chen and Y. Yang, Highly dispersed manganese oxide catalysts grafted on SBA-15: Synthesis, characterization and catalytic application in trans-stilbene epoxidation, Microporous Mesoporous Mater., 2010, 132, 501–509.
22. I. Spassova, T. Tsontcheva, N. Velichkova, M. Khristova and D. Nihtianova, Catalytic reduction of NO with decomposed methanol on alumina-supported Mn–Ce catalysts, J. Colloid Interface Sci., 2012, 374, 267–277.
23. Z. Qu, Y. Bu, Y. Qin, Y. Wang and Q. Fu, The improved reactivity of manganese catalysts by Ag in catalytic oxidation of toluene, Appl. Catal., B, 2013, 132–133, 353–362.
24. D. A. Peña, B. S. Uphade and P. G. Smirniotis, TiO₂-supported metal oxide catalysts for low-temperature selective catalytic reduction of NO with NH₃: I. Evaluation and characterization of first row transition metals, J. Catal., 2004, 221, 421–431.
25. C. Norsic, J.-M. Tatibouët, C. Batiot-Dupeyrat and E. Fourré, Non thermal plasma assisted catalysis of methanol oxidation on Mn, Ce and Cu oxides supported on γ-Al₂O₃, Chem. Eng. J., 2016, 304, 563–572.
26. B. R. Strohmeier and D. M. Hercules, Surface spectroscopic characterization of manganese/aluminum oxide catalysts, J. Phys. Chem., 1984, 88, 4922–4929.
27. M. H. Groothaert, K. Lievens, H. Leeman, B. M. Weckhuysen and R. A. Schoonheydt, An operando optical fiber UV–vis spectroscopic study of the catalytic decomposition of NO and N₂O over Cu-ZSM-5, J. Catal., 2003, 220, 500–512.
28. C. Lamberti, S. Bordiga, M. Salvalaggio, G. Spoto and A. Zecchina, XAFS, IR, and UV-Vis Study of the Cu^I Environment in Cu^I-ZSM-5, J. Phys. Chem. B, 1997, 101, 344–360.
29. M. Iwasaki and H. Shinjoh, NO evolution reaction with NO₂ adsorption over Fe/ZSM-5: In situ FT-IR observation and relationships with Fe sites, J. Catal., 2010, 273, 29–38.
30. J. Y. Luo, H. Oh, C. Henry and W. Epling, Effect of C₃H₆ on selective catalytic reduction of NO_x by NH₃ over a Cu/zeolite catalyst: A mechanistic study, Appl. Catal., B, 2012, 123–124, 296–305.

---