A series of Cu/ZSM-5, Fe/ZSM-5, and Cu-Fe/ZSM-5 catalysts (Si/Al in ZSM-5 = 25) were prepared by different metal loadings using the liquid ion-exchange method. Several characterization methods were conducted to explore the effects of metals on the physical and chemical properties of catalysts. Meanwhile, the electron paramagnetic resonance method is also used to assess the copper and/or iron elements’ coordination and valence state at intersections or in channels of ZSM-5. The metal-loading effects of all catalysts on the catalytic activities were studied for the removal of NOx in a fixed-bed flow reactor using selective catalytic reduction with ammonia (NH3-SCR). The results showed that the iron’s addition could markedly broaden the operation temperature range of the Cu/ZSM-5 catalyst for NH3-SCR between 200 and 550°C due to the presence of more isolated Cu2+ ions as well as additional oligomeric Fe3+ active sites and FexOy oligomeric species. This paper gives a facile and straightforward way to synthesize the practical-promising catalyst applied in NH3-SCR technology to control NOx emissions.

1. Introduction

Under industrialization with the industry’s rapid growth, environmental problems are also accumulating and increasing swiftly, making the community concerned. Specifically, pollutants in the atmosphere have become significant perils to human health. Therefore, air pollution control is one of the urgent issues for industrial countries and all over the world. The scientific demonstration of the polluting properties of several chemical compounds has paved the way for more rigorous regulations to control their emissions. There are lots of causes for this problem, but the contribution of motorized traffic plays an important role. Based on each chemical compound’s polluting properties, different laws are in place to control their emissions. In particular, the exhaust from the internal combustion engine contributes mainly to this problem. The vital pollutant compounds are incombustible hydrocarbon (HC), carbon monoxide (CO), particulate matter (PM), and nitrogen oxide (NOx) emissions, which are primarily due to spontaneous oxidation of NO [1]. Accordingly, one of the most notable air pollutants is NOx, which is obtained mainly from emissions in stationary and mobile sources of high-temperature combustion and from electrical discharge during thunderstorms [2]. The prominent harmful effects of NOx are acid rain formation, photochemical smog, ozone reduction, and fog.

Recently, with continuous efforts to satisfy more rigorous regulations on NOx emissions in both mobile and stationary sources, several innovative technologies have been exploited in parallel with the application of disposable methods as the three-way catalyst, NOx storage and reduction (NSR), and selective catalytic reduction (SCR) [3]. In particular, one of the most effective NOx reduction technologies is NH3-SCR, which is using NH3 or aqueous urea as a reducing agent to convert NOx to nitrogen [4]. This technology uses catalysts to assist in stimulating the reaction between NH3 and NOx to produce more quickly harmless N2 and H2O productions as the following reaction [5]:

Until now, NH3-SCR technology has been applied with a wide variety of catalysts such as precious metals, metal oxides, and zeolite-based catalysts [1]. The catalysts used in this technology are of the utmost importance with lots of particular requirements. For instance, to accommodate exhaust emissions from both light and heavy diesel engines and to increase the N2 selectivity of the NOx reduction reaction, the catalyst must have a wide operating temperature range (i.e., from 100 to 600°C). Besides, it also needs to be stabilized in water vapor and sulfur oxides generated from fuel combustion [3]. Consequently, the improvement of catalyst properties to enhance their efficiency in the SCR method has been increasingly studied. In particular, the most promising alternative to the currently common-use commercial catalyst as vanadium-based catalysts is the combination of zeolite-based catalysts and different transition metals such as Fe, Cu, Mo, and Cr [3]. The well-recognized CHA-structural zeolites for a deNOx application might be called SSZ-13 in particular, but the synthesis with a costly prototype is the key constraint on process commercialization of the SSZ-13 (N, N, N-trimethyl-l-adamantammonium hydroxide) [6, 7]. This disadvantage, on the other hand, could be overcome by using ZSM-5, whose synthesis is more economical and convenient. Specifically, the highly active iron and copper-denatured ZSM-5 catalyst has received more and more attention recently [810]. ZSM-5 is a nanometer-sized pore-structured molecular sieve material with strong acid sites that can adsorb NH3 onto the surface and converts them to NH4+. Also, the individual characteristic activity of this catalyst in the SCR reaction is shown through a miniature “nanoreactor” that exists in the internal space, providing a favorable reaction space and an electronic environment to create highly dispersed transition metal oxide clusters [11]. The advantage of forming oxide clusters and nanoparticles can be explained by the ultrasonic cavitation effect. It can support the dispersion of the active component inside micropores of ZSM-5 and diminish the relative supersaturation of solution leading to the inhibition of secondary nucleation of precursors and extra development of the crystal nucleus. Thus, the catalyst based on ZSM-5 shows outstanding catalytic activity at the low-temperature range due to the creation of more active centers through the impact of oxide clusters and nanoparticles.

It has been reported that Cu-exchanged to zeolites exhibited more activity below 300°C; when the temperature is above 350°C, the NO conversion is decreased due to the oxidation of ammonia. Meanwhile, Fe-doped zeolite gave better activity for NH3-SCR at high temperatures [4]. However, there is still an open discussion concerning the nuclearity and identity of Fe and Cu active sites for NH3-SCR reaction. Chronologically, the Cu-based zeolite is the first metal loading on zeolite discovered active in the SCR reaction. It is generally admitted that both the copper oxo cations and Cu2+ ions, which are active sites in this catalyst, play an important role in the deNOx process [1, 12]. In opposition to the Cu-zeolite system, the Fe-zeolite system submitted multiple active sites such as intrazeolite binuclear Fe-oxo, isolated Fe2+ and Fe3+ ions, small FexOy clusters, or oxygen bridged binuclear iron species [HO–Fe–O–Fe–OH]2+, groups of a wide range of nuclearity and large sets of oxides [1315]. Nowadays, bimetal catalysts based on zeolites have attracted much attention, especially codoping of Cu and Fe supported on zeolite, which became a promising catalyst for the NH3-SCR reaction. Recent research has indicated that such materials show higher NOx removal efficiency over a more extended temperature range than with individual Fe- or Cu-based zeolite catalysts [1618]. Iron and copper’s simultaneous appearance promotes more acid sites, an alteration of electronic properties, and augmentation of redox ability than for a single metal catalyst.

This study aims to clarify the application of copper-iron bimetallic-supported commercial ZSM-5 for the NH3-SCR catalyst. Additionally, Cu-Fe/ZSM-5 samples were compared with single metal on ZSM-5 (Cu/ZSM-5 and Fe/ZSM-5), thereby clarifying supplementary advantages, reciprocal effects of both copper and iron to catalytic properties, and the ability to widen the operating temperature range for deNOx with NH3 reductant. In this work, the aqueous ion-exchange method was applied for preparing all the metal catalysts. The Cu-Fe/ZSM-5, Cu/ZSM-5, and Fe/ZSM-5 catalysts were characterized by various techniques to determine the physicochemical properties. The main objective of this study focuses on the following contents:(i)The effects of Cu, Fe, and codoping Cu-Fe supported on ZSM-5 zeolite to their catalyst properties were clarified(ii)Also, the NH3-SCR activity and the hydrothermal durability of catalysts were conducted

2. Experimental

2.1. Catalyst Preparation

Metals (Cu and/or Fe) were added to ZSM-5 (DNH-Z, SiO2/Al2O3 = 25 of Tosoh Corporation) through a two-step liquid ion-exchange process. In the first step, ZSM-5 was changed into ammonium form by exchanging with ammonium acetate (CH3COONH4, Sigma-Aldrich) solution two times. The first exchange period was the continuous agitation of ZSM-5 zeolite with solid ammonium acetate and purified water at 80°C for 5 hours; then the substantial part was gathered by centrifugation. After that, the same process was performed one more time, but with a discrepancy in exchange time (18 hours) and stirring at room temperature. The centrifuged solid portion was subsequently dried at 120°C to accomplish NH4+/ZSM-5 powders. In the second step, the ammonium form of ZSM-5 was ion-exchanged with metals, in which Cu and/or Fe were doped using Cu(CH3COO)2.H2O (Sigma-Aldrich) and/or Fe(CH3COO)2.4H2O (Sigma-Aldrich), respectively. The solution was mixed continuously at room temperature within 24 hours and then centrifuged and washed with distilled water to reach a solid component. The collected part was then dried at 120°C for the night and calcined at 550°C for 5 hours in the flow of 20% O2/Ar. The ZSM-5 catalyst samples were denoted as xCu/ZSM-5, yFe/ZSM-5, and xCu-yFe/ZSM-5 in which x and y are the weight percentage of Cu and/or Fe used in synthesis, respectively.

2.2. Characterizations

All characterization methods have been effectuated for as-calcined samples. Powder X-ray diffractometer Bruker D8 with Cu-Kα radiation (λ = 1.5418 Å) at a scanning speed of 1°/min within 2θ = 5–50° was used to achieve the X-ray diffraction (XRD) patterns of catalysts. The field emission scanning electron microscopy (FE-SEM) was performed on Hitachi S-4800 (Japan) to explore the samples’ different morphology. Simultaneously, the chemical compositions of the catalysts were also determined by the Hitachi S-4800 (Japan) machine equipped with an energy-dispersive X-ray spectrometer (EDS). Measurements of nitrogen adsorption-desorption were performed at −196°C using a Micromeritics Analyzers ASAP 2020. The samples were degassed under a vacuum at 300°C before the examination. The overall surface area was evaluated based on the Brunauer–Emmette–Teller (BET) theory and t-plot for external surface and surface and volume. UV-Vis diffuse reflectance spectra (UV-Vis DRS) were performed on an Avantes Avaspec-ULS2048XL-EVO spectrometer equipped with an integrating sphere coated with BaSO4, from 200 to 1000 nm, and the program AvaSoft 8 was used. The different copper and iron species were quantified by the area ratios of the respective subbands relative to each other [19]. Temperature-programmed desorption of ammonia (NH3-TPD) experiments for the surface acidity determination were performed on a Micromeritics Auto Chem 2920 instrument. The TPD measurements were conducted in He flow (30 mL/min), and the temperature was increased from 100 to 600°C with a heating rate of 10°C/min. Besides, the temperature-programmed reduction with H2 (H2-TPR) was also carried out on Micromeritics Auto Chem 2920 instrument. In the initial, 0.1 g of the catalyst was placed into the quartz U-shaped microreactor and pretreated in He at 300°C for 1 hour, followed by cooling down to 50°C before the acquisition. Afterward, the ramping program was processed by the flow of 10 vol% H2/He from 50 to 900°C (30 mL/min), with a heating rate of 10°C/min. A detector continuously collected the signals during this process. Additionally, the reactor’s cooling trap was positioned next to the experiment to prevent the byproducts of experiments (i.e., H2O, CO2) from seeping into the detector. Electron paramagnetic resonance (EPR) spectra were performed on the Bruker EMX-Micro-EPR spectrometer, with a continuous wave X-band, using a microwave power of 6.3 mW, a modulation frequency of 100 kHz, an amplitude of 0.5 mT, and a magnetic field of 1000–5000 G.

2.3. Catalytic Performance

The catalytic activity test of all catalysts for the NH3-SCR reaction was carried out in a fixed-bed quartz reactor working under a steady flow mode with ABB Gas Analyzer AO2020 Limas21 (UV sensor) spectrometer in the condition of excess oxygen (Figure 1). For each activity test, 100 mg catalyst (40–60 mesh) and the total flow rate of 200 mL/min were used, leading to the gaseous hourly space velocity of 120 000 h−1. Moreover, to minimize condensation in downstream tubing, all the gas lines were heated and kept at approximately 100°C. In this reaction system, the used reactant gas has a composition of 1000 ppm NO, 1000 ppm NH3, 8 vol% O2, and Ar as balance. The temperature range of the reaction was from 200°C to 550°C. Firstly, the catalyst was activated in a flow of 20% O2/Ar at 550°C for 1 hour. At each temperature, the expected time to obtain a steady state was about 1 hour. A similar condition was followed for the NH3 oxidation reaction, where NO was not added to the gas mixture. In the investigation of hydrothermal stability, the catalysts were aged in a quartz tube reactor at 650°C or 750°C in 10% H2O/air under a total flow rate of 1 L/min for 24 hours before the activity test. When the system was already stable, calculation of the NO and NH3 conversions was performed through the inlet and outlet gas contents by the following equations:

3. Results and Discussion

3.1. Structure and Texture of Catalysts

The catalysts’ structure investigated by the XRD method is presented in Figure 2. In JCPDS data cards (JCPDS 42-0024), which show peak diffractions at 7.92°, 8.80°, 14.78°, 23.10°, 23.90°, and 24.40°, the prevalent diffraction peaks for commercial ZSM-5 can be simple to index into the standard parameters for MFI topology. The XRD patterns of all catalysts have included all characteristic peaks corresponding to ZSM-5, suggesting that ZSM-5 does not destroy its original form during the process of aqueous ion-exchange and calcination [20]. However, in contrast to those of the parent sample, the intensities of Fe/ZSM-5, Cu/ZSM-5, and Cu-Fe/ZSM-5 decrease due to the higher absorption coefficient of metal compounds for X-ray radiation or treatment with neutral aqueous solutions [21].

We can claim safely from Table 1, in conjunction with the chemical composition, that the synthesized materials are Fe- and Cu-containing ZSM-5-zeolites. No diffraction peaks were found at either metal and/or to metal oxide clusters, indicating that the copper and/or iron formed species are well distributed and within the range of nanometers [22]. Interestingly, the Cu-Fe/ZSM-5 synthesized does not have a peak of diffraction like α-Fe2O3 and CuOx. In contrast, its peak amplitude and corresponding crystallinity are slightly more significant than those of Cu/ZSM-5, as shown in Figure 2. This result indicates that incorporating Cu and Fe in ZSM-5 simultaneously could decrease the effects of a single metal onto ZSM-5. Furthermore, in all samples, the Na concentrations are below 0.09 wt.% (Table 1), which does not impact catalyst NH3-SCR performance of the catalysts [23].

FE-SEM observed the morphological properties of the synthesized zeolites and the relevant photographs in Figure 3. It is apparent that the ZSM-5 parent primarily consists of bright polycrystalline aggregates, irregularly localized, and distinct edged in the typical geometry [24]. Moreover, after doping metal(s), the primary morphology for all catalysts remains almost unchanged. However, depending on the metal loading, small amounts of aggregates composed of nanosized particles, particularly in the high amount of metal loading that can be created during calcination, emerge from the primary morphology of Fe/ZSM-5, Cu/ZSM-5, and Cu-Fe/ZSM-5. Or Doan et al. note that the combined CuxO or α–Fe2O3 nanoparticles can be ascribed to particular layers of suspended matter [21]. Alongside the EDS results (Table 1) and the XRD patterns (Figure 2), this can be due to the fact that some large metal comprising agglomerations are well-distributed crystal metal oxide particles onto ZSM-5 composition. For Cu-Fe bimetal catalysts based on ZSM-5, which have a metal content of about three percent, the catalysts are smoother than the other two individual metal-fueled zeolites with a metal load of over two percent.

Nitrogen adsorption-desorption isotherms for all catalysts are shown in Figure 4. The ZSM-5, Fe/ZSM-5, Cu/ZSM-5, and Cu-Fe/ZSM-5 isotherms are categorized as category I, characteristic of the micropores materials according to IUPAC classification. The nitrogen intake at about P/P0 = 0.01 of all catalysts is close to that of the parent ZSM-5, indicating that the microporous is not greatly blocked during the preparation. Table 2 displays the significant textural properties of all nanocatalysts. The calculation of N2 physisorption is a function of zeolites’ topology, including SBET, Vpore, and average pore diameter. It suggests that the ZSM-5 carrier has large surface areas and the active components could scatter well on the carrier through the liquid ion-exchange technique. As seen in Table 2, metal species introduction had little effect on the textural properties of ZSM-5. Overall, a slight decrease in the BET surface area typically depends on the content and form of metal. However, Vpore is a little higher for all nanocatalysts than the ZSM-5. This is why urea alkaline environment-induced ZSM-5 particles split into smaller particles and thus increased pore volume [25]. Furthermore, a minor increase in pore diameter of ZSM-5 also resulted from metal loading. The explanation may be due to the dissociation of some boundaries in ZSM-5 because of the urea alkaline environment, and consequently, the pore diameter of all catalysts increased [25].

3.2. Chemisorption Results

The catalyst’s surface acidity always makes a significant contribution to the NH3-SCR reaction for the removal of NOx [26]. This research also used the NH3-TPD method to measure the strength and the number of acid sites in the prepared catalysts. The corresponding profiles are shown in Figure 5 and Table 3. Typically, NH3 adsorption on weak, medium, and strong acid sites answers to the three distinct peaks at low (120–220°C), medium-high (250–380°C), and high (380–500°C) temperatures, respectively. Ammonia desorption at a lower temperature is related to weakly physisorbed NH3 at Lewis acid sites. Oppositely, the high-temperature desorption peak is attributed to the adsorbed strongly NH3 molecules on Brønsted acid sites [27, 28]. Although there are still many controversies, the assignment of the medium-high desorption peak is somehow attributed to NH3 weakly adsorbed on Brønsted acid sites or associated with extraframework Al [28]. These Brønsted and Lewis sites can be attributed to acidic protons and to extraframework Al atoms, which compensate framework Al atoms charge, respectively [29]. Consequently, it appears that the presence of Al atoms in ZSM-5 zeolite is in charge of the formation of centers for ammonia chemisorption. According to Table 3 and Figure 5, the doping of Cu and/or Fe into ZSM-5 decreases the acidity of ZSM-5. In comparison with ZSM-5, sites for NH3 chemisorption are less generated over metal introduction into the catalyst due to the replacement of Brønsted acid protons by metal [30]. Compared to individual Fe-doped ZSM-5, both strength and acid amounts of the strong acid sites are slightly decreased when codoping Cu and Fe together. In this case, it is assigned to the incorporation of Fe into the framework of Cu-Fe/ZSM-5 [31]. It was reported that the strength and acid amount of the strong acid sites is an essential factor for the adsorption of NH3. Therefore, it influences the NH3-SCR activity significantly [32]. With a suitable number of acid sites on Cu-Fe/ZSM-5, the catalytic activity may be further enhanced, and thus, it is more beneficial for the SCR reaction [33].

The H2-TPR profiles of Fe/ZSM-5, Cu/ZSM-5, and Cu-Fe/ZSM-5 catalysts are shown in Figure 6 to understand the redox proprieties of the catalysts better. There can be three H2 consumption zones observed in the TPR pattern of the Cu/ZSM-5 catalyst: the first peak at 180°C is corresponding to the reduction of external surface nanosized CuO to Cu+ [34], the second sharp peak at about 250°C is related to the reduction of isolated Cu2+ ions to Cu+ in ion-exchange positions of zeolites, and another one located at 370°C as a trace, can be assigned to the reduction of CuO particles to Cu0 and Cu + to Cu0 [35]. According to the high copper content, the excess amount of copper is detected as CuO when the exchange sites have been involved. One primary peak of the Fe/ZSM-5 catalyst at around 470°C ascribes to the reduction of Fe3+ to Fe2+ at zeolite ion-exchange positions [36, 37] as well as the reduction of Fe2O3 to Fe3O4 [38]. Also, a small and broad peak attended at approximately 630°C, which can refer to the reduction of Fe3O4 to FeO [39] and a residual fraction of framework Fe3+ that is hardly reduced [40]. Krishna et al. [41] reported that if there have been extended FeOx phases, they commonly point to reduction peaks around 600°C. Ariel Guzmán-Vargas et al. [39] based on different temperature-programmed reduction studies of Fe/ZSM-5 arranged that the peak for hydrogen consumption met at around 430°C can be associated with the reduction of Fe3+ to Fe2+ in Fe species by the framework-charge compensation. When Cu and Fe are doped simultaneously into ZSM-5, the results showed a transition of the first reduction peak to a higher temperature than Cu/ZSM-5. This change is due to the strong interaction between copper and iron metallic components and the integration in Cu-Fe/ZSM-5, making them harder to reduce. Moreover, the peak centered at 420°C relates to the reduction of Fe species [17], leading to the appearance of the iron reduction peaks at high temperatures which become more self-evident, and the copper peaks become weaker. The broader peaks present after Fe loading demonstrates that Fe species had various forms. Thus, improving the redox proficiency of such Fe species can enhance the NOx reduction effectiveness at high temperatures.

3.3. Cu and Fe Species onto ZSM-5

The UV-Vis DRS was also used to analyze the nature and coordination of copper and iron species in the catalysts and the respective spectra are compared in Figure 7. The spectra of Fe/ZSM-5, Cu/ZSM-5, and Cu-Fe/ZSM-5 are strikingly different, showing that the Fe and Cu species exist in distinct forms. The absorption peaks of around 270 nm are due to ZSM-5 zeolite [16]. In the Cu/ZSM-5 sample’s spectra, the strong band at about 230 nm is assigned to the d − d transition of the charge-transfer band, which relates to the transition O ⟶ Cu from lattice oxygen to isolated Cu2+ ions [6]. In addition, the broad but less intense band between 650 and 850 nm in Cu/ZSM-5 and Cu-Fe/ZSM-5 relates to the d − d transition of Cu2+ ions in an octahedral environment due to the distributed CuO particles [42]. With the Fe species, the band below 300 nm confirms the existence of isolated iron ions in the structure of tetrahedrons or octahedrons. Besides, the signal between 300 and 400 nm refers to the Fe-oxo oligomers, and the peak higher than 400 nm is due to iron oxide aggregates [43]. In this research, the Fe/ZSM-5 pattern shows a band at 214 nm which is ascribed to the Fe intrusion instead of Fe3+ ions coordinated with the tetrahedra within the framework of the zeolites [44], and a band located at 390 nm relative to the oligo nuclear FexOy cluster [45, 46]. In the meantime, both Fe/ZSM-5 and Cu-Fe/ZSM-5 display an aggressive absorption band at 250 nm due to the charge transfer of oxygen to iron, explaining the presence of isolated Fe species entered into the zeolite framework in the tetrahedral coordination [47]. Moreover, the d-d transition of the quite-inactive factor of the NH3-SCR as α-Fe2O3 also gives a band centered at 500 nm [48].

In order to analyze the state of iron and copper ions in our catalysts, EPR was conducted due to its powerful technique [49]. The obtained EPR spectra at −173°C are shown in Figure 8. In the EPR spectrum of copper samples, isolated Cu2+ presents a parallel signal with four adsorption peaks and a sharp peak in the vertical region [50]. On the other hand, Cu2+ species in dimeric species and Cu + ions do not interact with this technique [50, 51]. This phenomenon is due to the fact that isolated Cu2+ ions owned a nuclear spin (I = 3/2) and the 3d unpaired electrons coupled with each other. The EPR spectra of the samples were simulated using the software package EasySpin implemented in MATLAB [52]. In the results shown in Figure 8(a), all Cu/ZSM-5 samples exhibit an axial EPR spectrum of isolated Cu2+ ions with resolved hyperfine structure with || = 2.389, A|| = 400 MHz,  = 2.095, and A = 57 MHz. This has been assigned to copper species close to the channel intersection in several zeolites, for example, hydrated Cu-BEA and Cu-MFI [53]. Increasing the amount of doped Cu from 1 wt.% to 2 wt.% leads to a significant increase in the axial signal’s intensity, indicating the formation of more isolated Cu2+ ions (25.5% in the sample 2Cu/ZSM-5 compared to 21.5% in the sample 1Cu/ZSM-5). However, with higher Cu content of 3 wt.%, the amount of isolated Cu2+ species decreases to 17.5%. Besides, the isotropic signal at = 2.17 in the sample 3Cu/ZSM-5 is also broader. This broad signal is related to the magnetically interacting Cu2+ species in dimeric species and such Cu sites located in close vicinity or CuO clusters.

Figure 8(b) illustrates the EPR spectra of Fe/ZSM-5 samples recorded at −173°C. It is clear to see that iron species give three different EPR signals, containing a low-field shoulder at  = 4.23 and two broader signals at  = 2.29 and  = 2.05. To be more specific, the signal at  = 4.23 is attributed to strong rhombic, highly distorted isolated Fe3+ species and higher coordination numbers. This could be due to the isolated iron monomers located paramagnetically at ion-exchange positions inside the zeolite micropores [54]. The weak low-field line  = 2.29 and 2.05 signal might be an axial stack of distorted Fe3+ monomers and/or for interacting octahedrally with residual Fe2O3 phase [54]. Another observation made from the spectrum is that the intensity of EPR signals was raised significantly when increasing the content of Fe doping, especially with samples containing more than 1 wt.% of Fe loading. In other words, the higher intensity of the signals at  = 2.29 and  = 2.05 in these samples suggests the presence of more clustered Fe oxides.

As shown in Figure 8(c), compared to the monometal catalysts, EPR signals of bimetallic catalysts exhibit dominant axial signals from isolated Cu2+ sites. Compared to the same amount of copper loading in monocopper catalysts, the intensity of isolated Cu2+ in 2Cu-1Fe/ZSM-5 was increased while that of 2Cu-2Fe/ZSM-5 was lower. This behavior suggests that a suitable content of iron could lead to the formation of a higher amount of isolated Cu2+ ions. Moreover, increasing the amount of Fe loading in bimetallic catalysts could reveal the signals of isolated Fe3+ in tetrahedral coordination at  = 4.23. Meanwhile, the appearance of FexOy clusters cannot be shown given that their signals at  = 2.29 and 2.05 might be covered by the isolated active Cu2+ signal.

3.4. Catalyst Performance

The catalytic performance of the prepared catalysts has been evaluated in SCR of NO with NH3 and is shown in Figure 9 in a microcatalytic flow reactor. The conversion of NOx with Cu/ZSM-5 samples in Figure 9(a) dramatically improves from 250°C to 300°C and stays above 80% between 300°C and 450°C, particularly in the 300–400°C range when the yield achieves over 90%. Later, the conversion of NOx decreases steadily above 400°C and only exceeds 58% at 550°C. Among copper-loaded samples, the sample with 2 wt.% Cu shows the highest activity. Fe/ZSM-5 samples’ catalytic efficiency reveals, as seen in Figure 9(b), that the addition of iron in ZSM-5 increases the SCR reaction slightly over the whole temperature range. However, a decent range of temperatures from 400 to 500°C could only be obtained, which is typical of the Fe-zeolite mechanism with maximal operation at high temperatures [2]. Higher Fe content (up to 3 wt.%) NOx conversion tends to a slight decrease at high temperatures. These findings suggest that higher metal loading could have a detrimental impact on conversions of NOx. Therefore, selection for the content of the doped metal is necessary to enhance the removal of NOx. In summary, the results of Figures 9(a) and 9(b) show that sufficient metal content might support the conversion of NOx while excessive Cu or Fe loads may block the zeolite channel. This could be investigated in Figure 4 and Table 2 by the result of N2 adsorption-desorption.

When incorporating both Cu and Fe into ZSM-5, 2Cu-1Fe/ZSM-5 catalyst significantly enhances the NOx conversion in the broader temperature range of 350–550°C, while 2Cu-2Fe/ZSM-5 sample only presents a temperature window of 450–550°C as seen in Figure 9(c) (defined as NO conversion more than 2Cu/ZSM-5 sample). Compared to both of these experiments, the findings in Figure 9(c) reveal that the 2Cu-1Fe/ZSM-5 sample shows the most potent catalyst. In specific, relative to 2Cu/ZSM-5, the conversion of NOx over 2Cu-1Fe/ZSM-5 catalyst decreases marginally at low temperatures (below 250°C), accompanied by a rise in the range of 350–550°C. Notably, the NOx removal yields still reached 75% at 550°C, suggesting that Cu-Fe/ZSM-5 catalysts have expanded the operating temperature window.

It is well known that NOx conversion decrease with the reduction in N2 selectivity at high temperatures is due to oxidation in NH3 to unwanted N2O formation. The NH3 oxidation activity was also carried out for chosen catalysts. As seen in Figure 10, 2Cu/ZSM-5 and 2Cu-1Fe/ZSM-5 samples give the same curve shapes of NH3 conversion. The reaction to ammonia oxidation begins at 225°C but happens at a low rate in the temperature range below 300°C. The NH3 conversion is rising dramatically above this temperature. A singular copper catalyst shows the highest activity of NH3 oxidation, while bimetal samples demonstrate the lower action. The Cu/ZSM-5 conversion was almost 100% at 550°C, while it was about 90% in the codoping Cu-Fe. Figure 9 also shows the catalytic behavior of Fe/ZSM-5 decomposing NH3 catalysts that display an initial reaction temperature of 400°C and achieve just 35% with 550°C.

The reactions on the catalyst involve mainly the standard SCR and NH3 oxidation reactions expressing

As shown in Figures 9 and 10, the sample efficiency of the NH3-SCR reaction has been slightly enhanced over the NH3 oxidation reaction below 300°C, indicating a slight effect on catalytic activity in this temperature range due to the NH3 oxidation reaction. A strong ammonia inhibition effect of iron heads to a limited NH3-SCR performance below 350°C for the 2Cu-1Fe/ZSM-5 samples [13]. However, this catalyst’s catalytic activity displayed an improvement in the range of 250–450°C, explained by the rise in the quantity of isolated Cu2+ ions identified as active sites for SCR reactions (Figures 7 and 8) [50]. Compared to Cu/ZSM-5, the 2Cu-1Fe/ZSM-5 sample could not increase its NH3 oxidation activity. This may be explained by the decrease in the number of bulk CuO species that are expected to be extremely active for NH3 oxidation in the bimetal study [55]. A reduction in the number of bulk CuO species on the surface with Fe and Cu prevents the NH3 oxidation reaction. Consequently, the decrease in the NH3 oxidation potential and increase in the sample catalytic activity of 2Cu-1Fe/ZSM-5 contributes by the small amount of oligomeric FexOy species and isolated Cu2+.

In the 2Fe-2Cu/ZSM-5 sample, as discussed in EPR results, the oligomeric Fe3+ ions have replaced a small number of isolated Cu2+ ions, whereas Cu2+ isolated ions are readily combined with CuOx species [6]. This means that the decrease in the isolated Cu2+ count (Figure 8(c)), surface area decreases, and ZSM5 pores volume (Table 2) by the aggregated Fe2O3 particles decreases 2Fe-2Cu/ZSM-5 sample SCR operation at lower and medium temperatures at below 400°C [56]. In the meantime, the increase in SCR performance above 400°C might affect an increased amount of oligomeric FexOy and Fe2O3 clusters in the sample.

The hydrothermal stability of catalysts is an essential problem for SCRs since the generation of heat during the process of regeneration of DPF must be dealt with [57]. To investigate this property, the 2Cu/ZSM-5 and 2Cu-1Fe/ZSM-5 catalysts were treated in 10% H2O/air under a total flow rate of 1 L/min at 650°C and 750°C for 24 hours. Figure 11(a) depicts the SCR activity differences between fresh and hydrothermal aged samples, observing that the NOx conversion of hydrothermally processed samples reduces the whole temperature range. In particular, both 2Cu-1Fe/ZSM-5_650°C and 2Cu-1Fe/ZSM-5_750°C achieved more than 60% NOx conversion from 300 to 500°C. Both 2Cu-1Fe/ZSM-5 samples with two hydrothermal aging conditions have considerably higher NOx conversion than 2Cu/ZSM-5_650°C, meaning that Cu-Fe/ZSM-5 is more stable and resistant to relatively harsh hydrothermal treatment than Cu/ZSM-5. To be more specific, XRD patterns of the fresh and hydrothermal aging catalysts are shown in Figure 11(b). The result deconvoluted that all the other catalysts have the MFI zeolite structure apart from the 2Cu/ZSM-5_750°C sample. This singular copper exchanged ZSM-5 catalyst appeared to be more sensitive to hydrothermal treatment since the collapse of the MFI structure occurred at 750°C. This framework damage could be probably due to the aggregation of copper species, which could affect negatively the stability of the structure and lead to the deterioration in catalytic activity on NH3-SCR [18]. On the opposite, the hydrothermal treatment even at harsh conditions just had a negligible effect on the structure of 2Cu-1Fe/ZSM-5 samples. This insignificant effect on the bimetallic catalyst suggests that the addition of iron can help to stabilize the copper dispersion during aging; therefore, the increase in dispersing copper attends to be the primary explanation for the fact that the Cu-Fe/ZSM-5 catalysts improve hydrothermal stability.

4. Conclusion

In conclusion, hybrid Cu-Fe/ZSM-5 zeolites have been successfully synthesized using a liquid ion-exchange technique. Both catalysts systematically characterized the physicochemical properties by different techniques and were compared with mono Fe/ZSM-5, Cu/ZSM-5 zeolites developed using the same process. The XRD, FE-SEM, and N2 adsorption-desorption also demonstrated that most of the physical structure and crystal of ZSM-5 remained while Cu and Fe are codoped. The H2-TPR findings show that Cu-Fe/ZSM-5, owing to high-temperature redox capacities and abundance types of iron materials, has strong redox capability even over a broader temperature spectrum than Cu/ZSM-5. The UV-Vis DRS detected multiple active locations after Fe loading, namely, Fe3+ and FexOy, and a close relationship between Fe and Cu. The EPR results also confirm this. Compared to the Cu/ZSM-5 catalyst having the same amount of copper loading, the quantity of Cu2+ isolated was increased when the required amount of Fe was included in the Cu-Fe/ZSM-5 catalyst. Besides, the increased number of FexOy oligomeric in this experiment profits from the NH3-SCR reaction at high temperatures. Due to these characteristics, the Cu-Fe/ZSM-5 synthesized hybrid exhibits superior catalytic efficiency in the NH3-SCR reaction., i.e., high NO conversion in a wide temperature window (350–550°C) and increase in N2 selectivity. Our research offers an alternate NH3-SCR catalyst for potential applications shortly.

Data Availability

X-ray diffraction, field emission scanning electron microscopy, energy-dispersive X-ray spectroscopy, nitrogen adsorption/desorption, UV-Vis diffuse reflectance spectroscopy, temperature-programmed desorption of ammonia, temperature-programmed reduction of hydrogen, and electron paramagnetic resonance methods used to support the findings of this study are included within the article.

Conflicts of Interest

The authors declare that they have no conflicts of interest.


This research was funded by the Vietnam National Foundation for Science and Technology Development (NAFOSTED) under grant no. 104.05–2018.306 and Vingroup Innovation Foundation (VINIF) under grant no. VINIF.2020.TS.124. This work has also been supported by the RoHan Project funded by the German Academic Exchange Service (DAAD, No. 57315854) and the Federal Ministry for Economic Cooperation and Development (BMZ) inside the framework “SDG Bilateral Graduate School Programme”.