Synthesis and evaluation of mesopore structured ZSM-5 and a CuZSM-5 catalyst for NH₃-SCR reaction: studies of simulated exhaust and engine bench testing

Abstract

A modified ZSM-5 zeolite (denoted as ZSM-5-M), which was synthesized using tetrapropylammonium hydroxide (TPAOH) and cetyltrimethylammonium bromide (CTAB) as dual templates, and the commercial ZSM-5 zeolite (denoted as ZSM-5-C), have been used to prepare the corresponding CuZSM-5 (M & C) catalysts containing 3 wt% Cu by ion exchange method. Compared to CuZSM-5-C catalyst, the CuZSM-5-M catalyst demonstrated remarkably higher catalytic activity at low temperatures (<450 °C) for selective catalytic reduction of NO_x with NH₃ both in the simulated exhaust and engine bench testing. The modified synthesis by the dual template with the product aged for a long time at room temperature, leads to the formation of the ZSM-5-M zeolite with much higher specific surface area (608 m² g⁻¹) and higher total pore volume (0.8880 cm³ g⁻¹) due to the presence of more mesoporous pores. The X-ray diffraction results showed that ZSM-5-M maintained its typical MFI structure, while its crystallinity (84.1%) was lower than that of the ZSM-5-C zeolite. The characterization results by H₂ temperature-programmed reduction and X-ray photoelectron spectra revealed that the higher redox properties of isolated Cu²⁺ ions combined with the high-dispersion CuO crystallites and Cu⁺ ions are likely the main cause for the excellent low-temperature activity of the CuZSM-5-M catalysts. The isolated active Cu²⁺ species and high-dispersion CuO crystallites had a stronger interaction with other atoms in CuZSM-5-M catalysts. The results from thermogravimetric analysis, temperature-programmed desorption of ammonia and in situ diffuse reflectance infrared Fourier transform spectroscopy demonstrated that ZSM-5-M is a strong Brønsted acid site and the CuZSM-5-M catalyst had a relatively higher exchange rate of Cu²⁺ and Cu⁺ ions and more Lewis acidic sites, giving it a high NH₃ adsorption capacity. The strong Brønsted acid site might be another cause that results in the higher NH₃-SCR performance of the CuZSM-5-M catalyst.

---

1. Introduction

Nitrogen oxides (NO_x) emissions, as the major sources of acid rain and photochemical smog, are a serious threat to the environment. It has been demonstrated that automobiles contribute more than half of NO_x emissions. At present, UWS (urea water-solution)-based SCR technologies are the most promising commercial method for eliminating NO_x emissions from diesel engine exhausts.¹ The highly efficient catalyst, which reduces NO_x from diesel engine emissions, operates over a wide temperature range of 200 to 500 °C under a high space velocity. Therefore, the biggest challenge for developing a novel SCR catalyst is ensuring its stability under the harsh practical conditions of emissions control of diesel engines.²

For an emissions control system to successfully reliminate NO_x (NO and NO₂) emissions, highly activity and durable SCR catalysts are required. Due to their toxicity, vanadium-based SCR catalysts are not considered for use in the commercial diesel market, although vanadium is a primary component in the NH₃-SCR catalyst used for NO_x removal in stationary sources.^3–5 For passenger cars in Europe and any applications in the USA and Japan (i.e., passenger cars and heavy-duty trucks), however, it is expected that stringent emission limits will be met using non-toxic zeolite-based catalyst technologies because vanadium has been listed in the California proposition 65 list as a potentially carcinogenic substance.⁶

It is well-known that CuZSM-5 and FeZSM-5 ion-exchange zeolite catalysts have excellent NO_x conversion rates and selectivity compared to traditional V₂O₅–WO₃/TiO₂ catalysts.^7–13 Meanwhile, the CuZSM-5 catalyst exhibits low-temperature catalytic activity (<350 °C), and the FeZSM-5 catalyst primarily contributes to the high temperature catalytic activity (>350 °C).^14,15 Due to its surface acidity, long-term durability, and cost effectiveness, ZSM-5 zeolite has been successfully commercialized for decades in chemical industries, including oil refining¹⁶ and the petrochemical industry,¹⁷ and has been used in other fields as a heterogeneous catalyst for a variety of different hydrocarbon reactions.^18,19

However, the sole, small-pore channels and low specific surface area of ZSM-5 zeolites sometimes confine the dispersion of active species and the diffusion of the products and reactants, especially when bulky hydrocarbon molecules are involved in the reactions. The slow mass transport rate to and away from the catalytic center increases the possibility of secondary reactions, with coke formation (carbon deposits) and catalyst deactivation as a consequence.²⁰ A promising way of overcoming these shortcomings is to introduce a secondary pore system in the zeolites to form hierarchical (or mesoporous) materials,²¹ i.e., materials that possess both micro- and meso-pores, which would facilitate the rate of molecular transport. Additionally, it was found that the mesopores promote the production of the Cu(H₂O)_n complex and hinders the adsorption of H₂O on acid sites,²² which would be helpful to promote the reduction of NO_x in diesel engine exhaust. A lot of ZSM-5 zeolites with micropores and mesopores for different catalytic reactions have been reported.^23,24 Zhang et al.²⁵ reported that a new kind of ZSM-5 (Si/Al = 50 and 19) with bimodal pore size distributions was developed, in which a part of the original microporous structure in ZSM-5 remained, but some mesopores with a narrow pore size distribution were formed in the same matrix. However, the surface area of this ZSM-5 with mesoporous and microporous structure is still lower than 320 m² g⁻¹. Jacobsen et al.²⁶ reported the synthesis of mesoporous ZSM-5 single crystals by the controlled combustion of carbon particles. Analysis of the desorption isotherm revealed a bimodal pore size distribution with micropores of ∼0.5 nm radius and in the range of 5–50 nm in length. The pore volumes of the micropores and mesopores were 0.09 and 1.01 mL g⁻¹, respectively. Seong-Su Kim et al.²⁷ reported that mesoporous ZSM-5 zeolite was obtained by using colloid-imprinted carbons as templates for the nanocasting synthesis. Although the BET surface areas of samples with the different Si/Al were lower than 350 m² g⁻¹. It proved that the nanopore structure optimizes the catalytic properties of the zeolite for the conversion of molecules that are too large to penetrate the framework micropores and are constrained to reacting only at active sites. Changsong Mei et al.²⁴ reported that different Si/Al ratios of H-ZSM-5 with mesoporosity were prepared by alkaline desilication treatment and soft template method. The surface area of the samples ranged from 329 to 367 m² g⁻¹. High propylene selectivity (42.2%) and propylene/ethylene ratio (10.1) were observed.

According to the above cited literature, the ZSM-5 catalysts having both mesoporous and microporous structure have high activity, especially for large organic molecule reactions. However, the surface area of these bimodal pore structures of ZSM-5 zeolites is still low (<450 m² g⁻¹). Furthermore, few studies have been performed to study the influence of the ZSM-5 mesopore structure on NO_x SCR activity in application for diesel exhausts. Herein, a modified ZSM-5 was synthesized by a dual-template method. This ZSM-5, which has not only mesoporous structure, but also a high specific surface area was used to prepare the CuZSM-5 powder and honeycomb monolith catalysts by ion-exchange method. The catalytic performances of the powder and monolith catalysts were tested with simulated automobile exhaust gas and engine bench experiments. Furthermore, N₂ adsorption, X-ray diffraction (XRD), X-ray photoelectron spectroscopy (XPS), H₂ temperature-programmed reduction (H₂-TPR), NH₃ temperature-programmed desorption (NH₃-TPD), thermogravimetric analysis (TGA) and in situ diffuse reflectance infrared Fourier transform spectroscopy (DRIFTs) were performed to experimentally reveal the relationship between the specific surface area, pore structure (especially for mesopores), surface acidity and reducibility, and NH₃-SCR performance.

2. Experimental section

2.1. Preparation of ZSM-5 and CuZSM-5 catalysts

The modified ZSM-5 zeolite (Si/Al = 25) (denoted as ZSM-5-M) was synthesized by two steps: firstly, an appropriate amounts of tetraethyl orthosilicate (TEOS, 98 wt%, Sigma-Aldrich) and tetrapropylammonium hydroxide (TPAOH, 1 M aqueous solution, Sigma-Aldrich) were mixed under stirring until the TEOS was completely dissolved in the solution. Secondly, aluminum isopropoxide (AIP, 97 wt%, Sigma-Aldrich), sodium hydroxide (NaOH, 97 wt%, Sigma-Aldrich) and deionized water were added to the above solution under vigorous stirring to obtain a homogeneous mixture. The molar composition of the mixture was 1SiO₂ : 0.02Al₂O₃ : 0.01Na₂O : 0.3TPAOH : 10H₂O. The mixture was aged at room temperature for 5 days. Subsequently, cetyltrimethylammonium bromide (CTAB, 98 wt%, Sigma-Aldrich) was added to the mixture at 5% of the total weight of the mixture. The pH value of the solution was then adjusted to 11.0 using NaOH before the solution was transferred to Teflon-lined stainless steel autoclaves for hydrothermal crystallization at 95 °C for 3 days. Thereafter, the solids obtained were separated by centrifugation, washed with deionized water, and dried in air overnight at 120 °C, followed by calcination at 550 °C for 4 h in air.

Subsequently, CuZSM-5 with approximately 3 wt% of metal loading was prepared via an ordinary typical ion exchange procedure. The parent zeolites were immersed in a solution containing the desired amount of copper(II)nitrate trihydrate (Cu(NO₃)₂·3H₂O) solution. Copper ions were exchanged into the ZSM-5 zeolite under stirring at room temperature to avoid the formation of copper oxides on the zeolite surface during the course of ion exchange. Finally, the slurry was transferred into a vacuum evaporator to eliminate the solvent. The residue was dried at 120 °C for 12 h and further calcined at 550 °C in air for 4 h. One portion of the powder-form catalysts prepared according to the above procedure was then pelletized and crushed to a 40–60 mesh size prior to performance evaluation.

Another portion of the CuZSM-5 powder was used to make honeycomb monolith catalysts by washcoating onto full-length honeycomb monoliths (400 cpsi cordierite with 0.16–0.17 mm thickness). Prior to washcoating, the cordierite monoliths were pretreated with a nitric acid solution for 2 h at room temperature and were then rinsed out with hot distilled water (70 °C) for 30 min. The washcoating was carried out by dipping the monolith into a slurry consisting of 3 wt% CuZSM-5 catalysts and deionized water; single or multiple layers were made by repeating the washcoating procedures. The washcoated monoliths were dried in air at room temperature for 20 min and then dried in an oven at 120 °C for 1 h, followed by calcination at 550 °C for 4 h.

A commercial ZSM-5 zeolite (Si/Al = 25) was used as a comparison and this commercial-ZSM-5-based CuZSM-5 catalyst (denoted as CuZSM-5-C) was prepared by the same procedure described above.

2.2. Activity test of catalysts

The activities of powder CuZSM-5-M and CuZSM-5-C catalysts were tested in a fixed bed quartz reactor (i.d. 9 mm) located inside a vertical furnace. The reaction conditions were as follows: 500 ppm NO, 500 ppm NH₃, 5% O₂, 5% H₂O and N₂ as the balance gas. In all tests, the total flow rate of the feed gas was 1000 mL min⁻¹, which corresponded to a space velocity of approximately 30 000 h⁻¹. The concentrations of NO_x, N₂O, and NH₃ in the inlet and outlet gases were measured using a Thermo IS10 FTIR gas analyzer. The activity data were collected when the catalytic reaction had been substantially maintained for 30 min at each temperature, except for 150 °C, which was maintained for 60 min.

The performance of the catalysts is presented in terms of the conversion rate of NO_x (X_(NOx)) and the selectivity of N₂O (S_(N2)) as defined by the following equations:

2.3. Engine bench test of catalysts

A heavy-duty, turbocharged inter-cooler commercial diesel engine, which was manufactured by Yuchai Engineering Machinery Co., Ltd of China, was used for engine bench tests of monolith CuZSM-5-M and CuZSM-5-C catalysts. The engine featured a 17.5 : 1 compression ratio, 6 cylinders, direct injection, bore and stroke (113 × 140 mm²), and a total displacement of 8.424 L. The maximum torque was 950 N m at 1700 rpm, and the rated power was 177 kW at 2200 rpm. Fuel injection was provided by electronically controlled, cam-actuated unit pumps. The fuel used in this study was low-sulfur diesel (<50 ppm), provided by the Sinopec Group in China. An AVL 500 kW electric dynamometer was coupled to the engine and controlled by a PUMA control system.

In the engine bench experiments, the SCR system consisted of the dosing system for a urea solution and one single SCR catalytic converter, which was mounted on the engine tailpipe. The volume of the converter was 21.27 L, and it was filled with the honeycomb monolith catalyst. The exhaust temperature and space velocity were adjusted by varying the engine load and speed. The exhaust gas mass flow rate was calculated by the air flow and fuel consumption flow rate, which were measured by an air flow meter and a fuel consumption flow meter. The space velocity was calculated by multiplying the exhaust gas mass flow rate and the volume of the catalyst, and the exhaust temperature was estimated by averaging the inlet and outlet temperatures. A typical experiment began when the diesel engine was running stably. Once the test signals, such as the exhaust temperature, space velocity and gas concentration, were relatively stable, the reductant-dosing system introduced urea onto the catalyst, leading to an increase in the NO_x conversion rate. The experiment ran until a steady state was established or the NH₃ slip point (<15 ppm) was reached. For different engine operating conditions, the NO_x removal efficiency was calculated according to the following equation:

The inlet and outlet NO_x concentrations were measured using an exhaust gas analyzer (AVLAMA i60), and the NH₃ concentration was determined with an NH₃ analyzer (Siemens LDS6).

2.4. Characterization of the catalysts

A Quanta chrome Nova Automated Gas Sorption System was used to measure the N₂ adsorption–desorption isotherms of the samples at the temperature of liquid N₂ (−196 °C). The Brunauer–Emmett–Teller (BET) specific surface area was calculated according to the amount of N₂ adsorption in the relative pressure range of 0.001 to 0.2. The pore size distribution was calculated from the adsorption branch of the N₂ adsorption isotherm using the Barrett–Joyner–Halenda (BJH) method. The mesopore volume (V_meso) is the result of difference between the total pore volume (V_total) and the micropore (d_p < 1.7 nm) volume (V_micro). Prior to the N₂ adsorption measurement, the sample was degassed under vacuum at 300 °C for 24 h. X-ray diffraction (XRD) measurements were performed on a D/MAX-RB system equipped with a Cu Kα radiation source. The diffraction patterns were recorded in the 2θ range of 10 to 50° with a step size of 0.018° and a count time of 1 s per step.

X-ray photoelectron spectroscopy (XPS) data were obtained with an ESCALab 220i-XL electron spectrometer from VG Scientific using 300 W Mg Kα radiation. The vacuum pressure was approximately 3 × 10⁻⁹ mbar. The position of the binding energies was referenced against the line of adventitious carbon (C 1s) at 284.8 eV.

The temperature-programmed reduction with H₂ (H₂-TPR) experiments were conducted on a Micromeritics Chemisorb 2720 using approximately 0.1 g of each sample. Samples were pretreated at 300 °C for 1 h in N₂ flow. After the temperature stabilized at 50 °C, it was increased linearly from 50 to 1000 °C at a ramp of 10 °C min⁻¹, while the H₂ consumption was continuously recorded.

Temperature-programmed desorption of ammonia (NH₃-TPD) was conducted using 0.1 g of catalyst in a quartz reactor. The adsorption was performed by feeding a gas mixture flow (100 mL min⁻¹) containing 500 ppm NH₃ with N₂ as the balance gas at 25 °C for 2 h. The adsorption gas was then purged with N₂ until no NH₃ was detected in the effluent. TPD operations were performed in a N₂ flow rate of 300 mL min⁻¹ with a heating rate of 5 °C min⁻¹ to 650 °C. The concentrations of desorbed NH₃ in the effluent were continuously monitored by an FTIR spectrometer (Gasmet FTIR DX4000) equipped with a heated, low-volume multiple-path gas cell (5 m).

NH₃ adsorption and subsequent in situ desorption were also performed on a thermogravimetric analyzer (Q500, Thermal Instruments) with a total gas flow of 100 mL min⁻¹. After the sample (25–35 mg) was heated (10 °C min⁻¹) to 550 °C and held by a N₂ flow for 60 min, the system was cooled to 100 °C for NH₃ adsorption for 60 min in a 1% NH₃/N₂ flow. Prior to NH₃ release, the system was purged with a pure N₂ flow for another 60 min to remove weakly and physically adsorbed NH₃. The release of adsorbed NH₃ (NH₃ (ad)) was then carried out with a heating ramp of 10 °C min⁻¹ in a N₂ flow and analyzed by a gas FTIR analyzer (Nicolet 384).

In situ diffuse reflectance infrared Fourier transform spectroscopy (DRIFTs) was carried out using a Nicolet NEXUS 6700 FTIR spectrometer equipped with a smart collector and an MCT detector cooled with liquid N₂. The measurements were performed in situ in a high-temperature cell reactor fitted with ZnSe windows. The catalyst was finely ground, placed in a ceramic crucible, and manually pressed. Mass flow controllers and a sample temperature controller were used to simulate the real reaction conditions. Prior to each experiment, the catalyst was heated to 350 °C under N₂ with a total flow rate of 100 mL min⁻¹ for 60 min. The IR spectra were recorded by accumulating 100 scans at a resolution of 4 cm⁻¹.

3. Results and discussion

3.1. NH₃-SCR performance

Fig. 1a shows the DeNO_x catalytic activities of the CuZSM-5-M and CuZSM-5-C catalysts. Compared to the CuZSM-5-C catalyst, the CuZSM-5-M catalyst obviously showed better catalytic performance, with NO conversion above 80% from 225 to 550 °C in the simulated reaction gas containing 5% H₂O. The temperature window for the high catalytic activity of CuZSM-5-M is shifted towards a much lower temperature range compared with that of the CuZSM-5-C catalyst. The N₂ selectivity on both the CuZSM-5-M and CuZSM-5-C catalysts (shown in Fig. 1b) is similar (above 90%) throughout the temperature range below 450 °C, while the N₂ selectivity over CuZSM-5-M is slightly higher than that over CuZSM-5-C at temperatures above 450 °C. These results indicate that the CuZSM-5-M catalyst has a high DeNO_x catalytic activity and N₂ selectivity in the simulated reaction gas and therefore shows distinct advantages for potentially practical applications.

Fig. 1 Comparison of NH₃-SCR activity and N₂ selectivity of CuZSM-5-M and CuZSM-5-C catalysts over simulated reaction conditions. Reaction conditions: [NO] = [NH₃] = 500 ppm, [O₂] = 5%, 5% H₂O, GHSV = 30 000 h⁻¹.

During SCR reaction, the formation of NO₂ from NO oxidation over the catalyst promotes NO_x conversion via fast SCR reaction. Fig. S1a† shows the NO₂ concentration over both catalysts. It is obvious that the NO₂ concentration over the CuZSM-5-M catalysts is higher than that over the CuZSM-5-C catalyst in the low temperature range (<250 °C), which suggests that the fast SCR reaction easily occurs over CuZSM-5-M according to the reaction 2NH₃ + NO + NO₂ → N₂ + 3H₂O and therefore leads to the higher NO_x activity at low temperatures. On the other hand, N₂O is an ozone-layer-depleting substance and has 310 and 21 times the global warming potential (GWP) of CO₂ and CH₄, respectively.^28,29 As mentioned above, the N₂ selectivity over CuZSM-5-M is slightly higher than that over CuZSM-5-C at reaction temperatures above 450 °C (shown as in Fig. 1b). This result can be interpreted by the formation of N₂O. As shown in Fig. S1b,† the formation rate of N₂O on the CuZSM-5-C catalyst is slightly higher than that on the CuZSM-5-M catalyst when reaction temperature was above 450 °C. Five reactions could account for the formation of N₂O:

2NH₃ + 2NO₂ = N₂O + N₂ + 3H₂O (1)

3NH₃ + 4NO₂ = 3.5N₂O + 4.5H₂O (2)

2NH₃ + 2O₂ = N₂O +3H₂O (3)

4NH₃ + 4NO₂ + O₂ = 4N₂O + 6H₂O (4)

4NH₃ +4NO + 3O₂ = 4N₂O + 6H₂O (5)

3.2. N₂ adsorption and crystal structure

3.2.1. BET and pore structure analysis. N₂ adsorption–desorption was applied to determine the BET surface area, BJH pore volume, and average pore size of the ZSM-5 and the CuZSM-5 catalysts (M & C). The results are summarized in Table 1, Fig. 2 and 3. It is shown that the BET surface area of ZSM-5-M (608 m² g⁻¹) is higher than that of the commercial ZSM-5-C zeolite (415 m² g⁻¹). The total areas and microporous pore volumes were also obtained for the two samples by the t-plot method. The total pore volume of ZSM-5-M (0.8880 cm³ g⁻¹) is approximately 3.5 times greater than that of ZSM-5-C (0.2510 cm³ g⁻¹). The micropore volume of ZSM-5-M is only slightly greater than that of ZSM-5-C, but the mesopore volume (BJH adsorption cumulative volume of pores between 1.7 and 100 nm diameter) of ZSM-5-M (0.6923 cm³ g⁻¹) is nearly 7 times than that of ZSM-5-C (0.1052 cm³ g⁻¹), which indicates that mesopores improve the total pore volume of ZSM-5 material and the BET surface area. The BJH average pore size of ZSM-5-M (5.58 nm) is approximately 2 times higher than that of ZSM-5-C (2.42 nm). Therefore, a high pore volume and pore size may be the main structural factors that result in a high BET area.

Table 1 BET surface area (S_BET) values, pore volume and average pore size of ZSM-5 zeolite and CuZSM-5 catalysts

Sample	SBET/m^2 g^−1	V(total)/cm^3 g^−1	V(micro)/cm^3 g^−1	V(meso)^a/cm^3 g^−1	Average pore size^b/nm	Crystallinity (%)
a BJH adsorption cumulative volume of pores between 1.7 and 100 nm diameter.b 4 V A^−1 calculated by BET.
ZSM-5-M	608	0.8880	0.1957	0.6923	5.58	—
CuZSM-5-M	404	0.6237	0.1025	0.2008	6.21	84.1
ZSM-5-C	415	0.2510	0.1458	0.1052	2.42	—
CuZSM-5-C	285	0.2239	0.1121	0.0399	3.12	100

---

Fig. 2 The N₂ adsorption of (a) ZSM-5-M and (b) ZSM-5-C zeolites.

Fig. 3 The pore size distribution of (a) ZSM-5-M and (b) ZSM-5-C zeolites.

Fig. 2 shows the N₂ adsorption/desorption isotherms at 77.3 K. The isotherm over the ZSM-5-C is similar to a type-I isotherm of characteristic microporous materials³⁰ and has a narrow pore size distribution, with an average pore diameter of 2.42 nm (see Table 1). However, the ZSM-5-M catalyst shows a combination of type I and type IV isotherms, with the appearance of micropore filling at low pressure and hysteresis loops at higher pressures according to the IUPAC classification, which indicates the presence of mesopores. The BJH pore size distribution curves (see Fig. 3) of ZSM-5-C display a relatively narrow mesopore size distribution with maximum of 3 to 5 nm and an average of 3.5 nm. However, the distribution curves of ZSM-5-M not only show a pore size of 3.5 nm but also show a pore size in the range of 35 to 60 nm, which clearly indicates the presence of mesopores.

For the CuZSM-5 catalysts, the BET surface area, BJH pore volume and average pore size display the same tendency as the ZSM-5 zeolite. The incorporation of copper leads to a dramatic decrease in the BET surface area and total pore volume, from 608 m² g⁻¹ and 0.8880 cm³ g⁻¹ for ZSM-5-M, to 404 m² g⁻¹ and 0.6237 cm³ g⁻¹ for the CuZSM-5-M catalyst. The decrease is likely due to the fact that copper species cover the external surface of ZSM-5 and clog some of the micropores channels of the ZSM-5 zeolite. However, the BET surface area and total pore volume of CuZSM-5-M (404 m² g⁻¹ and 0.6237 cm³ g⁻¹) are still much higher than those of the CuZSM-5-C catalyst (285 m² g⁻¹ and 0.2239 cm³ g⁻¹). The average pore size of the CuZSM-5-M catalyst is also higher than that of the CuZSM-5-C catalyst, which favors the dispersion of active Cu species. Therefore, it is speculated that the CuZSM-5-M catalyst provides more active sites that are respnsible for the higher NO_x SCR performance.

3.2.2. XRD analysis. The XRD patterns of the CuZSM-5-M and CuZSM-5-C catalysts are shown in Fig. 4. Crystalline phases were identified by comparison with ICDD files.³¹ Fig. 4 shows that the diffraction peaks of ZSM-5 at 2θ = 7.88°, 8.76°, 13.84°, 23.04°, 26.50°, 29.79°, 36.10°, 45.3°, and 54.8°, corresponding to the (111), (220), (311), (511), and (440) crystal faces, can be observed in the CuZSM-5-M samples,³² which demonstrates that the ZSM-5-M material prepared by the dual-template method has a typical MFI structure and that this structure also remained intact in the CuZSM-5 catalysts.

Fig. 4 X-ray diffraction patterns of (a) CuZSM-5-M and (b) CuZSM-5-C catalysts.

In addition, the featured peaks of CuO (2θ = 35.5° and 38.9°) (PDF# 48-1548) were observed in CuZSM-5-C, which demonstrates that part of the Cu species exist as CuO particles in the CuZSM-5-C catalyst.³³ However, there is no noticeable CuO diffraction peak observed in the CuZSM-5-M catalyst (<4 nm). These results indicate that the copper species are finely dispersed on the surface of the ZSM-5-M support as amorphous oxides or that the copper species consist of copper oxide particles in the form of nanoparticles and Cu²⁺ ions, which may enhance NH₃-SCR activity. In addition, the MFI diffraction peak intensity of the CuZSM-5-M catalyst is lower than that of the CuZSM-5-C catalyst, and the peak width is wider on the CuZSM-5-M catalyst, which indicates that the degree of crystallinity of CuZSM-5-M is lower due to the presence of the mesopore structure. Furthermore, the wider peak width of CuZSM-5-M demonstrates that the ZSM-5-M zeolite maintains a relatively smaller particle size and therefore has higher BET area (as shown in Table 1). According to the area integral of a typical peak of the ZSM-5 zeolite, the crystallinity on CuZSM-5-M is approximately 84.1% of that of CuZSM-5-C catalyst (as shown in Table 1). These results are in excellent agreement with the higher BET surface area and pore volume observed for ZSM-5-M.

3.3. Redox properties

3.3.1. H₂-TPR analysis. To gain a better understanding of the redox properties of the catalysts, the H₂-TPR of the CuZSM-5-M and CuZSM-5-C catalysts were measured (shown in Fig. 5). The TPR results show that multiple copper species coexist in the samples. For the CuZSM-5 catalyst, the H₂ consumption signals in the temperature range of 200 to 550 °C are most likely related to the copper active species for the SCR reaction.³⁴ Peak fitting of the profiles was performed to identify the temperature of the different reduction events. Thus, by means of peak fitting, the original curves from ∼200 to 550 °C were fitted into three peaks (I, II and III); each peak corresponded with the reduction of one type of Cu species. According to the literature, the peak at approximately 255 °C (I) corresponds to the reduction of isolated Cu²⁺ ions to Cu⁺,^34–38 and the peak at approximately 356 °C (II) corresponds to the reduction of high-dispersion nano-sized CuO crystallites to Cu⁰.^{36,37,39–41} The H₂ consumption peak at approximately 524 °C (III) is attributed to the reduction of Cu⁺ to metallic Cu⁰, and Cu⁺ is formed from the reduction of isolated Cu²⁺ and original Cu⁺ ions.

Fig. 5 TPR profiles of (a) CuZSM-5-M and (b) CuZSM-5-C catalysts.

Compared with the CuZSM-5-C catalyst, the reduction peaks (I, II and III) of copper species shift to lower temperatures for the CuZSM-5-M catalyst, suggesting that the copper species consist of higher Cu²⁺ and CuO species and species with original Cu⁺ redox properties. The H₂ consumption corresponding to the reduction of Cu²⁺ (isolated Cu²⁺ ions), as represented by the total area of peak I, is higher for CuZSM-5-M than for CuZSM-5-C (Table 2). However, the reduction of CuO (CuO species), as represented by the area of peak II, is lower for CuZSM-5-M (29.51%) than it is for CuZSM-5-C (45.19%), which is consistent with the XRD results. The area of peak III (Cu⁺ ions) is higher than the area of peak I for both catalysts, especially for CuZSM-5-M (67.47%).^36,41 For Cu–zeolites, the reducibility of Cu species varied in a wide range due to their various coordination to the zeolite framework oxygens.^36,41,42 Chen et al.⁴³ proposed that the H₂ consumption peak III of Cu⁺ is contributed from the reduction of isolated Cu²⁺ ions and the reduction of original Cu⁺ ions existing upon the framework of the zeolites and that there were two types of original Cu⁺ ions coordinated with the zeolite skeleton. It had been demonstrated that the quantity of Cu⁺ ions in the CuZSM-5 catalyst is essentially entirely responsible for its low-temperature activity. Therefore, it is proposed that the lower reduction temperature of isolated Cu²⁺ ions, high dispersion CuO crystallites and the Cu⁺ ions may contributed to the excellent low-temperature activity of CuZSM-5-M catalysts.

Table 2 Chemical composition of 3% CuZSM-5-M and 3% CuZSM-5-C catalysts by XPS and TPR

Catalysts	XPS composition (mol%)				TPR composition (%)
	Al/(Al + Si)	Si/(Al + Si)	Cu	Cu/(Al + Si)	Isolated Cu^2+	CuO	Cu^+
CuZSM-5-C	12.45	87.55	1.20	1.68	1.88	45.19	52.93
CuZSM-5-M	13.99	86.00	1.86	2.48	3.02	29.51	67.47

---

3.3.2. XPS analysis. To understand the nature of the interaction between Cu/Al and other elements, CuZSM-5-M and CuZSM-5-C samples were investigated using the XPS technique, as presented in Fig. 6. Cu 2p spectra are shown in Fig. 6(I).⁴⁴

Fig. 6 XPS spectra of CuZSM-5-M (a) and CuZSM-5-C (b). (I) Cu 2p and (II) Al 2p.

Cu species in CuO show the main Cu 2p_3/2 transition peak located at approximately 933.1 eV (with a shoulder peak at 935.8 eV) and the Cu 2p_1/2 transition peak at 952.8 eV, which are associated with two shake-up satellite peaks at 943.1 eV and 963.5 eV, respectively.^45–47 Because these satellite peaks are not observed for Cu⁺ ions and Cu⁰ species, they have been used to distinguish between Cu²⁺ and Cu⁺ ions or Cu⁰ species.⁴⁸ The Cu 2p_3/2 signal could be fitted to the two peaks for the different chemical states of Cu ²⁺ ions at peak I (933–934 eV) and peak II (935–936 eV), where the former is assigned to the presence of CuO species⁶⁰ and the latter to the presence of the Cu²⁺ ion being coordinated to superficial oxygen atoms of the zeolite.^49,61 The XPS results show that the bingding energy (B.E.) of CuZSM-5-M (953.5 eV and 933.1 eV) for Cu 2p_1/2 and Cu 2p_3/2 present a slower red shift compared with that of CuZSM-5-C (953.8 eV and 933.3 eV), indicating the strong interaction between the Cu species and other atoms, such as O, Al and Si in CuZSM-5-M catalysts and the better dispersion of CuO species on the zeolite. In particular, the B.E. of the Cu²⁺ ions from the shoulder peak of Cu 2p_3/2 in the CuZSM-5-M is 0.5 eV lower, which shows that the isolated active Cu²⁺ ions have a stronger interaction with other atoms than the CuO particles. Therefore, the redox properties of active Cu species in the CuZSM-5-M catalyst are higher than they are in the CuZSM-5-C catalyst.

Al 2p XPS spectra are also shown in Fig. 6(II) with the spectra a and b for the CuZSM-5-M and CuZSM-5-C catalysts, respectively. The two Al 2p peaks (I and II) are resolved with a curve-fitting procedure. Peak I at a lower energy of 73.5 eV is attributed to Al atoms in the ZSM-5 framework, where Al atoms are connected to the bridged O atom in the –Si–OH–Al– chains.⁵⁰ Peak II at 79.7 eV is higher by approximately 0.6 eV on the CuZSM-5-M catalyst than it is on the CuZSM-5-C catalyst. These results indicate that electrons in the CuZSM-5-M catalyst are easily transferred from Al atoms to Cu ions, leading to the reduction of Cu²⁺ ions and the partial oxidation of Al atoms, which results in the chemical shift toward a higher BE of Al 2p. These results are in excellent agreement with the relatively lower binding energy of Cu 2p_3/2 on the CuZSM-5-M catalyst and the strong interaction between active Cu species and other atoms (as shown in Fig. 6(I)).

Table 2 lists the surface atomic concentrations derived from quantitative analysis of the XPS survey scan spectrum. For the CuZSM-5-M catalyst, the surface Cu/(Al + Si) ratio (1.68) is lower than in the CuZSM-5-C catalyst (2.48). It is proposed that the isolated Cu²⁺ and higher dispersion CuO are predominantly inside the ZSM-5 pores rather than on the surface in form of large CuO particles, which is consistent with the lower amount of CuO particles on the external surface shown in the XRD and TPR results (Fig. 4 and 5). The surface Si concentration of CuZSM-5-M is also higher than that of the CuZSM-5-C catalyst. The phenomenon of Si migration to the surface has been reported.⁵¹ The high concentration of Si will make the Brønsted acid sites (Si–OH–Al) stronger, which might weaken the binding of H(Si–OH–Al). Therefore, the exchange of active Cu²⁺ on the surface of ZSM-5 easily takes place and results in higher NO_x activity on the CuZSM-5-M catalyst.

3.4. Surface acidities

3.4.1. NH₃-TPD analysis. Temperature-programmed desorption of ammonia was carried out to determine the strength and amount of different acid sites. The NH₃-TPD profiles for the CuZSM-5-M and CuZSM-5-C catalysts are displayed in Fig. 7. In the entire desorption temperature range, the CuZSM-5-M and CuZSM-5-C catalysts showed three ammonia desorption peaks, A, B and C, located at 130, 271 and 580 °C, respectively. The desorption peak centered at the lower temperature (approximately 130 °C) is attributable to physically adsorbed NH₃ or ammonium species, while the peaks at 271 and 580 °C are assigned to NH₃ that is strongly adsorbed at the strong Brønsted acid sites on Si–OH–Al groups and the non-framework Lewis acid sites of CuZSM-5-M and CuZSM-5-C, respectively.^{33,52–59,62,63} Fig. 7 shows that the CuZSM-5-M catalysts have the same NH₃-desorption temperature at 271 °C as that of the CuZSM-5-C catalyst and that the area of peak B is lower, which was due to the fact that the Brønsted acid protons of ZSM-5 were substituted by Cu²⁺ and Cu⁺.⁶⁴ However, the CuZSM-5-M catalyst has stronger Lewis acidic sites originating from Si–(O)–Al groups by the integral area of the NH₃-desorption peak (C).

Fig. 7 NH₃-TPD and DTG-TG profiles of (a) CuZSM-5-M and CuZSM-5-C catalysts and (b) ZSM-5-M and ZSM-5-C zeolites.

For comparison, NH₃-adsorption TG-DTG was used to investigate the surface acidities on the ZSM-5 zeolite. Fig. 7b shows the TG-DTG spectra of the ZSM-5 zeolite after NH₃ adsorption pretreatment. As shown in Fig. 8a, the first weight loss peaks appear in the DTG spectra at approximately 156 °C; these could be assigned to weakly and physically adsorbed NH₃ species, whereas the weight loss peaks at approximately 338 and 580 °C could be attributed to the desorption of NH₃ on strong Brønsted acidic sites and Lewis acid sites, respectively.⁶⁵ Compared to the result from the ZSM-5 zeolite, the temperature of desorption of NH₃ on Brønsted acidic sites on the CuZSM-5 catalyst shifted to lower temperatures, which shows that active Cu species facilitate the desorption of NH₃. We found that the area of peak B on ZSM-5-M is higher than it is on the ZSM-5-C zeolite and that the area of peak B on CuZSM-5-M is lower than it is on CuZSM-5-C.

Fig. 8 DRIFT spectra of (a) CuZSM-5-M and (b) CuZSM-5-C treated in flowing 500 ppm NH₃ at room temperature for 1 h and then purged by N₂ at 50 °C.

In the TG spectra results, the two weight loss steps can be observed on ZSM-5-M and ZSM-5-C at approximately 494 and 630 °C, respectively. The weight loss from Brønsted acidic sites and Lewis acid for ZSM-C are 0.48 and 0.32%, respectively, but the weight losses from these sites for ZSM-M are 1.43 and 0.09%, respectively, which represents the quality of the acid on different acid sites. Comparing the ZSM-5 zeolite and CuZSM-5 catalyst results, the ZSM-5-M has more strong acid sites than ZSM-5-C zeolite and more Brønsted acid protons have been substituted by active Cu²⁺ on the CuZSM-5-M catalyst and reduced the strength of Brønsted acid sites.

3.4.2. In situ DRIFT_S. Ammonia has been reported to be adsorbed on Brønsted or Lewis acid sites to form NH₄⁺ or coordinated NH₃ in the SCR reaction; gaseous or adsorbed nitric oxides then react with NH₄⁺ or coordinated NH₃ to form N₂ and H₂O.⁴⁴ Therefore, the surface acidity of the catalyst is critical for the SCR reaction of NO_x by NH₃. Fig. 8 shows the DRIFT spectra of NH₃ adsorption at 50 °C on CuZSM-5-M and CuZSM-5-C catalysts. Several bands in the range of 1000 to 1950 and 3100–3400 cm⁻¹ were detected. The bands at 1180, 1283, and 1627 cm⁻¹ can be assigned to asymmetric and symmetric bending vibrations of the N–H bonds in NH₃ that are coordinately linked to Lewis acid sites,⁴⁵ while the bands at 1470 cm⁻¹ and in the range of 1750 to 1627 cm⁻¹ could be attributed to asymmetric and symmetric bending vibrations of NH₄⁺ species at Brønsted acid sites.⁴⁶ In the NH stretching region, bands were found at 3360, 3260, and 3160 cm⁻¹. Some negative bands at approximately 3660 cm⁻¹ were also observed, which could be assigned to surface O–H stretching.

Brønsted acid was necessary to bind and disperse the metal ions, which might also prevent the aggregation of exchanged metal ions. Both the weak and strong Lewis acid sites in the zeolite framework might act as adsorption sites for NH₃ and therefore act as a reservoir of the reductant.⁶⁶ From the NH₃ adsorption results, the intensities of the 1470 and 1750 cm⁻¹ bands due to the Brønsted acid sites on the CuZSM-5-M catalyst are noticeably weaker than they are on CuZSM-5-C, which shows that more copper species would substitute the H proton of bridging hydroxyls (Brønsted acid sites) on the CuZSM-5-M catalyst and decrease the number of Brønsted acid sites. Meanwhile, these exchanged copper species seem to generate additional Lewis acid sites (1627 cm⁻¹ bands). Therefore, the intensity of the 1627 cm⁻¹ band due to Lewis acid sites remains stronger, which is consistent with the NH₃-TPD and TG-DTG results. All of the above results show that the CuZSM-5-M catalyst has a relatively higher number of isolated Cu²⁺ ions, which decreases the intensity of Brønsted acid sites and increases the intensity of Lewis acid sites, thus improving both NH₃ adsorption on Lewis acid sites and NO_x activity.

3.5. Engine bench testing

Steady-state engine-bench testing was performed to determine NO_x reduction across the SCR catalyst at the engine bench alone under several engine mode conditions, as shown in Fig. S2 and Table S1.† Different space velocities and inlet temperatures of the catalyst could be obtained by adjusting the engine operating mode. The target urea dosing ratio was adjusted manually and produced different NH₃/NO ratios (i.e., NSR = 0.6, 0.8 and 1.2).⁶⁷ The measurements indicate that the SCR process is dependent on both the space velocity and exhaust temperature, and their effects on NO_x conversion are shown in Fig. 9. Fig. 9 shows that the conversion of NO_x at 200 °C by the CuZSM-5-M catalyst at 20 000 h⁻¹, 30 000 h⁻¹ and 40 000 h⁻¹ is approximately 69.26%, 61.5% and 45.1%, respectively, which is essential for the diesel urea-SCR process. The NO_x activity of the CuZSM-5-M is significantly higher than that of CuZSM-5-C at different space velocities, especially at low temperatures, which is consistent with the results of the powder sample shown in Fig. 1. We also obtained a high NO_x conversion rate (>80%) from 250 °C to 400 °C at 40 000 h⁻¹ with the CuZSM-5-M catalyst, which shows that CuZSM-5-M has a higher NO_x conversion rate at a high space velocity than CuZSM-5-C on engine bench testing. Interestingly, the difference of NO_x activity between the CuZSM-5-M and CuZSM-5-C catalysts is larger at 20 000 h⁻¹ than it is at 40 000 h⁻¹. We speculate that the high NH₃ storage stability may be the primary reason for the higher NO_x activity with the CuZSM-5-M catalyst.

Fig. 9 The conversion of NO_x on CuZSM-5-M and CuZSM-5-C at different temperatures (NH₃/NO = 1.2).

The effect of the NH₃/NO_x ratio on DeNO_x activity was studied at different temperatures and space velocities, as shown in Fig. 10. The NO_x conversion rate dropped as the NH₃/NO_x ratio decreased. The NO_x conversion of the CuZSM-5-M catalyst is significantly higher than that of the CuZSM-5-M catalyst at NH₃/NO_x = 0.6, 0.8 and 1.2. The NO_x conversion rate is an approximation to the theoretical ratio from the standard SCR reaction at approximately 225 °C. However, the NO_x conversion rate of the CuZSM-5 catalyst is only 42.37% and 64.02% at 200 °C and 225 °C, respectively. It is noted that only at 20 000 h⁻¹ the NO_x conversion approached the theoretical SCR reaction ratio (100%) above 300 °C, and the NO_x conversion reaches approximately 60% of the maximum at 30 000 h⁻¹ and 40 000 h⁻¹. According to the NH₃-TPD and DRIFTs results, the weak surface acid makes the CuZSM-5-C have weaker NH₃ adsorption, especially at a high space velocity. All of the above results show that the NO_x activity of the CuZSM-5-M catalyst is higher than that of the CuZSM-5-C at different space velocities below 300 °C, and the DeNO_x performance of the CuZSM-5-M catalyst is also higher at different NH₃/NO ratios due to its higher BET area and pore volume and strong surface acidity.

Fig. 10 The influence of NH₃/NO (NSR) ratio of NO_x activity on CuZSM-5-M and CuZSM-5-C catalysts at 20 000 h⁻¹ and at different temperatures (NH₃/NO = 0.6, 0.8 and 1.2).

In automotive applications, the NH₃ adsorption rate and consequently the NO_x conversion rate are strongly affected by the dynamic conditions of the engine loading and speed because the exhaust mass flow and exhaust temperature vary widely.⁶⁸ Therefore, it is imperative to increase the NH₃ adsorption and storage capacity of the SCR catalyst to improve the NO_x SCR activity, especially in the transient condition. It is possible to obtain an even higher NO_x conversion efficiency to meet stringent emission standards (Euro 5 or 6) by slightly over-dosing without exceeding NH₃ slip targets on the CuZSM-5-M catalysts.

4. Conclusion

A modified ZSM-5 zeolite (denoted as ZSM-5-M) was synthesized with tetrapropylammonium hydroxide (TPAOH) and cetyltrimethylammonium bromide (CTAB) as dual templates. A commercial ZSM-5 zeolite (denoted as ZSM-5-C) was used as a comparison. The corresponding CuZSM-5 catalysts containing 3 wt% Cu were prepared by ion exchange method. The experimental data in present work show that the CuZSM-5-M catalyst has higher NH₃-SCR activity than the counterpart prepared based on the commercial ZSM-5 zeolite for the NH₃-SCR reaction over both simulated atmosphere and engine bench testing, and the activity is not sensitive to the NH₃/NO ratio and space velocity.

The characterization results showed that the dual-templates method, especially with long term aging at room temperature, results in the formation of ZSM-5-M zeolite having mesopores, which significantly increase the specific surface area and total pore volume compared to those of the commercial ZSM-5-C zeolite. Additionally, the mesoporous ZSM-5-M retains a typical MFI structure and has a lower crystallinity (84.1%) than the ZSM-5-C zeolite. The isolated Cu²⁺ ions, the high dispersion CuO crystallites and the Cu⁺ ions in the CuZSM-5-M catalysts have higher redox properties, which favor excellent low temperature activity. The XPS results showed that the isolated active Cu²⁺ species and high dispersion CuO crystallites have a stronger interaction with other atoms, such as O, Al and Si, in the CuZSM-5-M catalysts. The TG-TGA, NH₃-TPD and in situ DRIFTs results showed that the ZSM-5-M has a strong Brønsted acid and thus the CuZSM-5-M catalyst had a relatively higher rate of Cu²⁺ and Cu⁺ ion exchange and more Lewis acidic sites, which leads to a high NH₃ adsorption capacity. The strong Brønsted acid site might be another reason for the high NH₃-SCR performance of the CuZSM-5-M catalyst.

Acknowledgements

This work was financially supported by National Natural Science Foundation of China (21325731, 21221004 and 21477057) and National High-Tech Research and Development (863) Program of China (2013AA065304).

References

1. F. Birkhold, U. Meingast, P. Wassermann and O. Deutschmann, Appl. Catal., B, 2007, 70, 119–127.
2. S. Brandenberger, O. Krocher, A. Tissler and R. Althoff, Catal. Rev., 2008, 50, 492–531.
3. J. H. Goo, M. F. Irfan, S. D. Kim and S. C. Hong, Chemosphere, 2007, 67, 718–723.
4. J. Park, H. J. Park, J. H. Baik, I.-S. Nam, C. Shin, J. Lee, B. K. Cho and S. H. Oh, J. Catal., 2006, 240, 47–57.
5. H. Yahiro and M. Iwamoto, Appl. Catal., A, 2001, 222, 163–181.
6. R. Aneja, K. Flathmann, C. Savonen and T. Tindall, in Proceedings of the Detroit Diesel Corporation 10th Annual DEER Conference, 2 September, 2004.
7. H. Sjövall, L. Olsson, E. Fridell and R. J. Blintet, Appl. Catal., B, 2006, 64, 180–188.
8. N. Wilken, K. Kamasamudram, N. W. Currier, J. Li, A. Yezerets and L. Olsson, Catal. Today, 2010, 151, 237–243.
9. A. Grossale, I. Nova and E. Tronconiet, Catal. Today, 2008, 136, 18–27.
10. J. H. Kwak, D. Tran, S. D. Burton, J. Szanyi, J. H. Lee and C. H. F. Peden, J. Catal., 2012, 287, 203–209.
11. J. H. Kwak and D. Tran, Catal. Lett., 2012, 142, 295–301.
12. F. Göltl, R. E. Bulo, J. Hafner, P. Sautet and J. Phys, Chem. Lett., 2013, 4, 2244–2249.
13. D. W. Fickel, E. D'Addioa, J. A. Lauterbacha and R. F. Loboa, Appl. Catal., B, 2011, 102, 441–448.
14. G. Carja, Y. Kameshima, K. Okada and C. D. Madhusoodana, Appl. Catal., B, 2007, 73, 60–64.
15. H. Sjövall, L. Olsson, E. Fridell and R. J. Blint, Appl. Catal., B, 2006, 64, 180–188.
16. M. A. den Hollander, M. Wissink, M. Makkee and J. A. Moulijn, Appl. Catal., A, 2002, 223, 85–102.
17. M. J. B. Souza, F. A. N. Fernandes, A. M. G. Pedrosa and A. S. Araujo, Fuel Process. Technol., 2008, 89, 819–827.
18. V. Vishwanathan, K. W. Jun, J. W. Kim and H. S. Roh, Appl. Catal., A, 2004, 276, 251–255.
19. K. Fajerwerg and H. Debellefontaine, Appl. Catal., B, 1996, 10, 229–235.
20. H. G. Karge, Stud. Surf. Sci. Catal., 2001, 137, 707–746.
21. K. Egeblad, C. H. Christensen, M. Kustova and C. H. Christensen, Chem. Mater., 2008, 20, 946–960.
22. H. Zhao, J. Ma, Q. Zhang, Z. Liu and R. Li, Ind. Eng. Chem. Res., 2014, 53, 13810–13819.
23. Y. Hsin Chou, C. S. Cundy and A. A. Garforth, et al., Microporous Mesoporous Mater., 2006, 89, 78–87.
24. C. Mei, P. Wen and Z. Liu, et al., J. Catal., 2008, 258, 243–249.
25. C. Zhang, Q. Liu and Z. Xu, et al., Microporous Mesoporous Mater., 2003, 62, 157–163.
26. C. J. Jacobsen, C. Madsen and J. Houzvicka, et al., J. Am. Chem. Soc., 2000, 122, 7116–7117.
27. S.-S. Kim, J. Shah and T. J. Pinnavaia, Chem. Mater., 2003, 15, 1664–1668.
28. Climate Change 1992, The supplementary report to the IPCC scientific assessment, ed. T. Houghton, B. A. Callander and S. K. Verney, Intergovernmental Panel on Climate Change, Cambridge, UK, 1992.
29. M. H. Thiemens and W. C. Trogler, Science, 1991, 251, 932–934.
30. F. Bin, C. L. Song, G. Lv, J. N. Song, X. F. Cao, H. T. Pang and K. P. Wang, J. Phys. Chem. C, 2012, 116, 26262–26274.
31. H. van Koningsveld, J. C. Jansen and H. van Bekkum, Zeolites, 1990, 10, 235–242.
32. Z. Zhu, G. Lu, Z. Zhang, Y. Guo, Y. Guo and Y. Wang, ACS Catal., 2013, 3, 1154–1164.
33. T. Zhang, J. Liu, D. Wang, Z. Zhao, Y. Wei, K. Cheng, G. Jiang and A. Duan, Appl. Catal., B, 2014, 148–149, 520–531.
34. H. Praliaud, S. Mikhailenko, Z. Chajar and M. Primet, Appl. Catal., B, 1998, 16, 359–374.
35. J. Y. Yan, G. D. Lei, W. M. H. Sachtler and H. H. Kung, J. Catal., 1996, 161(1), 43–54.
36. C. Torre-Abreu, M. F. Ribeiro, C. Henriques and G. Delahay, Appl. Catal., B, 1997, 14(3–4), 261–272.
37. M. Richter, M. J. G. Fait, R. Eckelt, E. Schreier, M. Schneider, M. M. Pohl and R. Fricke, Appl. Catal., B, 2007, 73(3–4), 269–281.
38. A. Sultana, T. Nanba, M. Haneda, M. Sasaki and H. Hamada, Appl. Catal., B, 2010, 101(1–2), 61–67.
39. T. Nanba, S. Masukawa, A. Ogata, J. Uchiawa and A. Obuchi, Appl. Catal., B, 2005, 61, 288–296.
40. M. Richter, M. J. G. Fait, R. Eckelt, M. Schneider, J. Radnik, D. Heidemann and R. Fricke, J. Catal., 2007, 245, 11–24.
41. R. Kefirov, A. Penkova, K. Hadjiivanov, S. Dzwigaj and M. Che, Microporous Mesoporous Mater., 2008, 116(1–3), 180–187.
42. R. Bulánek, B. Wichterlová, Z. Sobalík and J. Tichý, Appl. Catal., B, 2001, 31(1), 13–25.
43. B. Chen, R. Xu, R. Zhang and N. Liu, Environ. Sci. Technol., 2014, 48(23), 13909–13916.
44. D. D. Sarma and C. N. R. J. Rao, J. Electron Spectrosc. Relat. Phenom., 1980, 18, 25–45.
45. T. Liese and W. Grünert, J. Catal., 1997, 172, 34–45.
46. A. Corma, A. Palomares and F. Márquez, J. Catal., 1997, 170, 132–139.
47. M. Navlani-García, B. Puértolas, D. Lozano-Castelló, D. Cazorla-Amorós, M. V. Navarro and T. García, Environ. Sci. Technol., 2013, 47, 5851–5857.
48. L. Wang, J. R. Gaudet, W. Li and D. Weng, J. Catal., 2013, 306, 68–77.
49. S. Contarini and L. Kevan, J. Phys. Chem., 1986, 90, 1630–1632.
50. H. Y. Chen, L. Chen, J. Lin, K. L. Tan and J. Li, Inorg. Chem., 1997, 36, 1417–1423.
51. R. F. Howe, R. N. Lamb and K. Wandelt, Surface Science: Principles and Application, Springer-Verlag, Berlin, 1993, p. 242.
52. R. Q. Long and R. T. Yang, J. Catal., 2001, 198, 20–28.
53. N. Katada, H. Igi, J. Kim and M. Niwa, J. Phys. Chem. B, 1997, 101, 5969.
54. M. Sawa, M. Niwa and Y. Murakami, Zeolites, 1990, 10, 532.
55. M. Niwa, M. Iwamoto and K. Segawa, Bull. Chem. Soc. Jpn., 1986, 59, 3735.
56. L. Ma, Y. Cheng, G. Cavataio, R. W. McCabe, L. Fu and J. Li, Chem. Eng. J., 2013, 225, 323–330.
57. W. L. Fanchiang and Y. C. Lin, Appl. Catal., A, 2012, 419–420, 102–110.
58. J. A. Botas, D. P. Serrano, A. García, J. Vicente and R. Ramos, Catal. Today, 2012, 195, 59–70.
59. L. Rodríguez-González, F. Hermes, M. Bertmer, E. Rodríguez-Castellón, A. Jiménez-López and U. Simon, Appl. Catal., A, 2007, 328, 174–182.
60. X. Long, Q. J. Zhang, Z. T. Liu, P. Qi, J. Lu and Z. W. Liu, Appl. Catal., B, 2013, 134–135, 381–388.
61. K.-Y. Lee, S.-W. Lee and S.-K. Ihm, Ind. Eng. Chem. Res., 2014, 53, 10072–10079.
62. H. Feng, X. Chen, H. Shan and J. W. Schwank, Catal. Commun., 2010, 11, 700–704.
63. X. Chen, W. Li and J. W. Schwank, Catal. Today, 2011, 175, 2–11.
64. G. H. Kuehl and H. K. C. Timken, Microporous Mesoporous Mater., 2000, 35–36, 521–532.
65. H. Chang, X. Chen, J. Li, L. Ma, C. Wang, C. Liu, J. W. Schwank and J. Hao, Environ. Sci. Technol., 2013, 47(10), 5294–5301.
66. S. Brandenberger, O. Kröcher, A. Wokaun, A. Tissler and R. Althoff, J. Catal., 2009, 268, 297–306.
67. W. Tang, D. Youngren, M. Santa Maria and S. Kumar, SAE International Journal of Engine, 2012, 6, 862–872,  DOI:10.4271/2013-01-1066.
68. D. Rauch, D. Kubinski and U. Simonc, Sens. Actuators, B, 2014, 205, 88–93.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra20237c

---