Fe-Modified Mesh-Structured Mn₂O₃/γ-Al₂O₃/Al Catalysts: Enriched Surface Active Oxygen and Superior Redox Properties for Enhanced NH₃-SCO Performance

Abstract

Ammonia-selective catalytic oxidation (NH₃-SCO) is an effective technology for eliminating NH₃ slip; however, the development of catalysts that simultaneously exhibit excellent low-temperature (<350 °C) activity and high N₂ selectivity remains a significant challenge. A novel structured monolithic mesh-type Fe_x-Mn₂O₃/γ-Al₂O₃/Al catalyst was developed. XPS, H₂-TPR, and O₂-TPD results demonstrate that Fe doping markedly increases the concentration of surface-adsorbed active oxygen species and enhances the redox capability. As a result, the optimally doped Fe_6.61-Mn₂O₃/γ-Al₂O₃/Al catalyst achieved complete NH₃ conversion at 210 °C with a 75% N₂ selectivity, outperforming previously reported Mn-based catalysts. Density functional theory (DFT) calculations further confirm that Fe modification enhances O₂ adsorption energy. In addition, the introduction of Fe significantly improves the catalyst’s resistance to SO₂ and H₂O. In situ FTIR results indicate that the NH₃-SCO reaction over Fe-Mn₂O₃/γ-Al₂O₃/Al proceeds predominantly via an internal selective catalytic reduction (i-SCR) pathway.

1. Introduction

Atmospheric pollution and its health risks have long been a concern in environmental science and public health. Recently, ammonia (NH₃) emissions, primarily from fertilizer use, industrial processes, and NH₃-SCR NO_x abatement [1,2], have garnered increased attention. NH₃ exposure can irritate the respiratory system, skin, and eyes, and may impair tissue functions [3]. Therefore, effective control of NH₃ emissions is crucial for improving air quality.

Among various NH₃ removal methods developed previously, ammonia-selective catalytic oxidation (NH₃-SCO) has attracted extensive attention in recent years because it can directly convert NH₃ into environmentally benign N₂ and H₂O [4] and is regarded as the most promising technology for NH₃ degradation. Noble metal catalysts such as Au and Pt exhibit the highest low-temperature activity in NH₃-SCO [5], but their low N₂ selectivity and high cost limit their large-scale application [6,7]. Therefore, great efforts have been devoted to exploring low-cost NH₃-SCO catalysts, and manganese oxides have recently been considered promising NH₃-SCO catalysts. Previous studies have shown that the crystal phase of MnO_x is crucial for the catalyst’s selectivity. Qu et al. [8] synthesized MnO₂ (UH), which achieved 100% NH₃ conversion at 170 °C, but its N₂ selectivity was only 49%. The Mn₂O₃ phase exhibits the highest N₂ selectivity in the NH₃-SCR reaction, but its low-temperature activity is relatively poor [9]. Therefore, achieving the coexistence of high conversion and high selectivity at low temperature (<350 °C [10]) remains a key challenge for Mn-based catalysts. In this context, doping modification has been recognized as an effective strategy for optimizing the catalytic performance of MnO_x. Madhuri et al. [11] demonstrated that Fe₂O₃ doping of Mn₂O₃ significantly increased the surface-adsorbed oxygen concentration, enhancing catalytic activity without affecting benzonitrile selectivity. Similarly, Yu et al. [12] reported that the Mn_1.82Fe_0.18O₃ catalyst, formed by Fe incorporation into Mn₂O₃, exhibited excellent performance in toluene oxidation, attributed to increased surface acid site density and higher active oxygen concentration due to Fe doping. These studies indicate that Fe modification of Mn₂O₃ improves its low-temperature oxidation performance. These findings provide new insights for enhancing NH₃ conversion in Mn₂O₃ catalysts through Fe doping and achieving a balance between high activity and high N₂ selectivity.

On the other hand, traditional granular or powder catalysts suffer from issues such as poor mechanical strength, high pressure drop, limited mass transfer, and particle loss [13,14], which restrict their long-term stable operation. While monolithic catalysts alleviate some of these problems, their washcoating method leads to weak interactions between the active components and the support, affecting the catalyst’s long-term stability and activity. To overcome these shortcomings, a mesh-structured γ-Al₂O₃/Al support, prepared by anodization, was proposed in our previous work and applied for the removal of pollutants such as O₃, HCHO, and CO [15,16,17,18,19]. This support features excellent mechanical properties and a three-dimensional interconnected pore structure [20], effectively enhancing mass and heat transfer [21]. However, there are few reports on the application of these anodized, in situ-grown monolithic catalysts in NH₃-SCO reactions.

In this study, a novel structured monolithic mesh Fe_x-Mn₂O₃/γ-Al₂O₃/Al catalyst was developed for low-temperature selective catalytic oxidation of NH₃ (NH₃-SCO). First, a series of characterizations were conducted to systematically reveal the effects of Fe doping on surface active adsorbed oxygen species and redox properties of the Mn₂O₃/γ-Al₂O₃/Al catalyst. Density functional theory (DFT) calculations were further performed to analyze the influence of Fe doping on the O₂ adsorption energy on the catalyst surface. In addition, water and sulfur resistance evaluations were carried out to assess the role of Fe species in the reaction system. Finally, the evolution of key surface intermediates was tracked by in situ Fourier transform infrared spectroscopy (in situ FTIR), based on which a possible NH₃-SCO reaction mechanism was proposed.

2. Results and Discussion

2.1. The Effect of Mn Crystal Phases on NH₃-SCO Performance over a γ-Al₂O₃/Al Carrier

Figure 1a–c present the SEM images of the AAO/Al and γ-Al₂O₃/Al supports obtained after anodization and subsequent treatments of the Al substrate. As shown in Figure 1a, a typical “sandwich” structure is formed after anodization, where anodic alumina layers with a thickness of approximately 90 μm are in situ grown on both sides of the Al substrate. The in situ-grown Al₂O₃ layers are tightly bonded to the metallic Al substrate, which effectively suppresses the detachment of active coatings commonly observed in conventional washcoated supports under harsh conditions such as high temperature, high humidity, or high gas velocity, thereby enhancing the structural stability of the monolithic catalyst. As shown in Figure 1b, the AAO precursor exhibits a highly ordered and uniform nanochannel structure with a relatively narrow pore size distribution and a smooth surface. After thermal treatment, the AAO precursor is successfully transformed into γ-Al₂O₃/Al, accompanied by pronounced morphological changes (Figure 1c). Compared with the AAO precursor, γ-Al₂O₃/Al shows a much rougher surface, where the ordered nanochannels evolve into a honeycomb-like porous network with wrinkled and irregular pore walls [17]. Such a porous structure provides abundant anchoring sites for the deposition of active phases, thereby strengthening the interaction between the active species and the support.

As shown in Figure 1d,e, Mn species supported on γ-Al₂O₃/Al lead to MnO₂/γ-Al₂O₃/Al and Mn₂O₃/γ-Al₂O₃/Al catalysts exhibiting polyhedral morphologies. The XRD patterns (Figure 1f) confirm the successful formation of γ-Al₂O₃, MnO₂/γ-Al₂O₃/Al. In addition, our previous study demonstrated that γ-Al₂O₃/Al possesses a high density of Lewis acid sites [22], which is favorable for the stable anchoring of Mn species as well as for NH₃ adsorption.

As shown in Figure 1g, all samples exhibit typical type IV adsorption–desorption isotherms with distinct H3 hysteresis loops, indicating the predominance of mesoporous structures. The H3-type hysteresis loop is generally associated with slit-shaped pores formed by the aggregation of plate-like particles [23]. The corresponding pore size distribution (Figure 1h) reveals that the dominant pore sizes are centered at 3–4 nm. Combined with the honeycomb-like morphology observed by SEM, these results suggest that a mesoporous network structure is formed during the thermal transformation from AAO/Al to γ-Al₂O₃/Al through pore wall stacking and channel reconstruction. As a result, the specific surface area of γ-Al₂O₃/Al increases significantly from 11 to 102 m²·g⁻¹, which contributes to improved dispersion of active species. After loading Mn species onto γ-Al₂O₃/Al, the specific surface area further increases from 89 to 140–145 m²·g⁻¹ (

Table 1. Textural properties and Mn loading of AAO, γ-Al₂O₃/Al, MnO₂/γ-Al₂O₃/Al, and Mn₂O₃/γ-Al₂O₃/Al.

Catalysts	SBET a/m2·g−1	Vpore b/cm3·g−1	Dp b/nm	Mn c/wt%
AAO	11	0.13	34.72	0
γ-Al2O3/Al	102	0.15	4.22	0
MnO2/γ-Al2O3/Al	145	0.19	3.59	17.3
Mn2O3/γ-Al2O3/Al	140	0.16	4.55	17.1

^a From the isotherm analysis in the P/P₀ range of 0.04~0.12; ^b from the desorption curve of the BJH method; ^c Mn/wt% and Fe/wt% were measured by the ICP-AES.

), which is conducive to the exposure of active sites relevant to the NH₃-SCO reaction.

In this study, the novel mesh γ-Al₂O₃/Al support was employed in the development of the optimal crystalline phase and Fe-modified Mn-based catalysts for the NH₃-SCO reaction. Figure 2 presents a comparison of the catalytic performance of MnO₂/γ-Al₂O₃/Al and Mn₂O₃/γ-Al₂O₃/Al in this reaction. In the 100–240 °C range, a higher NH₃ conversion rate was achieved by MnO₂/γ-Al₂O₃/Al compared to Mn₂O₃/γ-Al₂O₃/Al, with a T₁₀₀ (the temperature at which NH₃ conversion reaches 100%) of 200 °C, which is 240 °C lower than that of Mn₂O₃/γ-Al₂O₃/Al (Figure 2a). However, much lower N₂ selectivity was observed for MnO₂/γ-Al₂O₃/Al, with a N₂ selectivity of only 50% at T₁₀₀, while over 80% was achieved by Mn₂O₃/γ-Al₂O₃/Al (Figure 2b). As shown in Figure 2c, although the NH₃ conversion of Mn₂O₃/γ-Al₂O₃/Al was lower at low temperatures, a significantly higher N₂ yield was obtained, demonstrating superior de-ammoniation performance.

2.2. Effect of Fe Modification on NH₃-SCO Activity over Fe_x-Mn₂O₃/γ-Al₂O₃/Al Catalysts

It has been reported that Fe incorporation into Mn₂O₃ catalysts modifies the surface electronic structure and redox properties in VOC catalytic oxidation systems. In particular, Fe doping has been shown to increase the proportion of surface-adsorbed oxygen species and enhance the density of acid sites [11,12], thereby facilitating oxidation reactions. Based on these considerations, Fe was introduced into the Mn₂O₃/γ-Al₂O₃/Al catalyst in this work, and its structural and physicochemical properties were first investigated before evaluating the NH₃-SCO performance.

N₂ adsorption–desorption experiments were conducted to investigate the textural properties of the catalysts. As shown in Figure 3a, all catalysts exhibited BET Type IV isotherms [24], with an H3-type hysteresis loop in the p/p₀ = 0.4–1 range, indicating a predominant slit-like mesoporous structure. The pore size distribution (Figure 3b) reveals that the majority of the pores are concentrated in the 4–5 nm range. According to the data in Table 1 and

Table 2. Textural properties and Fe/Mn loadings of Fe_x-Mn₂O₃/γ-Al₂O₃/Al catalysts (x = 4.25, 6.61, and 8.32).

Catalysts	SBET a/m2·g−1	Vpore b/cm3·g−1	Dp b/nm	Mn c/wt%	Fe c/wt%
Fe4.25-Mn2O3/γ-Al2O3/Al	159	0.20	5.59	17.1	4.25
Fe6.61-Mn2O3/γ-Al2O3/Al	175	0.23	5.81	17.4	6.61
Fe8.32-Mn2O3/γ-Al2O3/Al	167	0.21	5.60	17.3	8.32

^a From the isotherm analysis in the P/P₀ range of 0.04~0.12; ^b from the desorption curve of the BJH method; ^c Mn/wt% and Fe/wt% were measured by ICP-AES.

, Fe doping leads to noticeable changes in the textural properties of the catalysts: the specific surface area of Fe_6.61-Mn₂O₃/γ-Al₂O₃/Al increased from 140 m²/g to 175 m²/g, the pore volume rose from 0.16 cm³/g to 0.23 cm³/g, and the average pore size increased from 4.55 nm to 5.81 nm. Notably, when the Fe loading exceeded 8.32 wt%, both the specific surface area and pore volume slightly decreased, indicating the presence of an optimal doping amount, which is consistent with the observed catalytic activity.

Figure 3c,d display the XRD patterns of the catalysts. Diffraction peaks corresponding to the (211), (222), (400), (332), (431), and (440) planes of cubic Mn₂O₃ are observed at 2θ = 23.131°, 32.951°, 38.234°, 45.178°, 49.347°, and 55.189°, respectively. No impurity peaks associated with Fe₂O₃ or spinel structures were observed, suggesting that Fe atoms substituted Mn atoms in the lattice through isomorphous substitution, successfully forming a single-phase Mn₂O₃ [25]. Additionally, the magnified XRD pattern (Figure 3d) reveals a noticeable shift of the (222) peak towards higher diffraction angles with Fe doping. According to the Bragg equation, the lattice constants (a) for Mn₂O₃ and Fe_6.61-Mn₂O₃ are 9.401 Å and 9.380 Å, respectively, confirming a contraction of the unit cell. This is mainly attributed to the smaller ionic radius of Fe compared to Mn [26], indicating that Fe atoms are incorporated into the Mn₂O₃ lattice, forming a single-phase Fe-Mn mixed oxide.

SEM images reveal that the Fe doping amount markedly affects the particle morphology and packing of Mn₂O₃. The undoped Mn₂O₃/γ-Al₂O₃/Al (Figure 4a) mainly consists of regular polyhedral particles with a broad size distribution of 5–8 μm. The particle surfaces are smooth, the crystal planes are well-defined, and the particles are densely packed, indicative of high crystallinity. Upon Fe introduction, Fe_4.25-Mn₂O₃/γ-Al₂O₃/Al (Figure 4b) retains the polyhedral morphology, although the particle size slightly decreases, and surface features such as steps and edges emerge. The particles become more loosely packed. With a further increase in Fe doping to 6.61 wt% (Figure 4c), the particle size distribution becomes more uniform, grain size decreases further, and surface steps and crystal boundaries are more prominently exposed. This observation is consistent with the highest specific surface area observed for this sample. A larger specific surface area is beneficial for the adsorption of reactants. Wei et al. [27] analyzed unmodified 13X zeolite and Mn-Fe/13X catalysts using SEM and found that the formation of cubic structures and petal-like layered MnO₂ on the surface contributed to an increased specific surface area, thereby enhancing ozone adsorption. Furthermore, no significant large-scale agglomeration is observed for the Fe_6.61-Mn₂O₃/γ-Al₂O₃/Al catalyst, indicating good dispersion of the active species. It has been reported that good dispersion can enhance catalytic activity. For instance, Wen et al. [28] reported that the active components in Mn-Fe-Ce/Al₂O₃/CC catalysts exhibited better dispersion than those in Mn-Fe-Ce/CC samples, which was considered one of the main reasons for the improved catalytic activity. When the Fe content reaches 8.32 wt% (Figure 4d), partial particle agglomeration occurs, with small particles adhering to larger ones to form dense secondary stacking structures. Particle boundaries appear blurred, which may limit the full exposure of active surface sites. This observation is in agreement with the above discussion.

In addition, Figure 4e,f shows the EDS mapping images of the Fe_6.61-Mn₂O₃/γ-Al₂O₃/Al catalyst. The atomic percentages of Al, O, Fe, and Mn on the catalyst surface are 0.30%, 39.12%, 0.45%, and 60.13%, respectively. These results indicate that manganese oxides dominate the catalyst surface. The Fe content is relatively low. This suggests that Fe species are mainly covered by Mn species. Therefore, Mn₂O₃ serves as the main active site of the catalyst, while Fe mainly acts as a promoter.

To clarify the influence of Fe doping on catalytic behavior, the NH₃-SCO performance of the catalysts was further investigated. As shown in Figure 5a–c, the NH₃ conversion, N₂ selectivity, and N₂ yield were systematically evaluated under varying temperature conditions. The results indicated that the introduction of different Fe loadings into Mn₂O₃/γ-Al₂O₃/Al significantly enhanced NH₃ conversion in the 100–240 °C range. Among the catalysts, Fe_6.61-Mn₂O₃/γ-Al₂O₃/Al exhibited the best performance, achieving complete NH₃ conversion at 210 °C. Furthermore, Fe_6.61-Mn₂O₃/γ-Al₂O₃/Al maintained N₂ selectivity above 75% in the 200–300 °C range, indicating that Fe doping effectively improved NH₃ conversion without adversely affecting product distribution. Considering both NH₃ conversion and N₂ selectivity, the Fe_6.61-Mn₂O₃/γ-Al₂O₃/Al catalyst demonstrated superior overall NH₃-SCO performance.

In addition, XPS was used to assess the surface valence states and chemical composition of the catalysts. The Mn 2p spectrum showed peaks at 641.0 eV and 642.5 eV, corresponding to Mn³⁺ and Mn⁴⁺ [29], with the catalyst predominantly existing as Mn₂O₃ and Mn³⁺ as the major valence state, consistent with XRD. A low AOS typically indicates a higher Mn³⁺ content [30,31]. The Jahn–Teller effect of Mn³⁺ ions [32] causes distortion of the Mn-O octahedra, activating lattice oxygen and generating reactive oxygen species (O⁻, O₂²⁻, or surface-adsorbed oxygen), thereby facilitating oxidation [12,33].

The O1s spectrum (Figure 6a) showed two peaks: O_ads (531.2–531.5 eV) and O_latt (529.7–529.8 eV) [34]. O_ads serves as an indicator of oxygen vacancy content and plays a more significant role in catalytic activity due to its higher mobility and oxidative potential [35]. The O_ads/O_latt ratio follows the order: Fe_6.61-Mn₂O₃/γ-Al₂O₃/Al > Fe_8.32-Mn₂O₃/γ-Al₂O₃/Al > Fe_4.25-Mn₂O₃/γ-Al₂O₃/Al > Mn₂O₃/γ-Al₂O₃/Al, indicating that Fe_6.61-Mn₂O₃/γ-Al₂O₃/Al adsorbs and activates oxygen more efficiently [36]. The results agree with the SEM observations. The SEM images of Fe_6.61-Mn₂O₃/γ-Al₂O₃/Al (Figure 4c) show a more uniform particle size distribution and smaller grains. Mn₂O₃ particles are well exposed on the surface, which facilitates the adsorption of active oxygen species.

The Fe 2p spectrum of Fe_6.61-Mn₂O₃/γ-Al₂O₃/Al (Figure 6d) displayed peaks at 724.8 eV and 710.5 eV, corresponding to Fe 2p_1/2 and Fe 2p_3/2, and satellite peaks at 718.8 eV and 731.6 eV confirmed Fe³⁺ presence [35,37]. Fitting of the Fe 2p_3/2 main peak showed Fe²⁺ (710.5 eV) and Fe³⁺ (711.9 eV), indicating a mixed Fe³⁺/Fe²⁺ valence state on the surface [37]. The Fe³⁺/Fe²⁺ ratios for the Fe_x-Mn₂O₃/γ-Al₂O₃/Al samples were 1.11, 1.42, and 1.39 (

Table 3. XPS fitting of the O 1s, Mn 2p, Mn 3s, and Fe 2p peaks.

Catalyst	Mn4+(eV)	Mn3+(eV)	Oads(eV)	Olatt(eV)	Oads/(Oads+Olatt)	Mn3+/Mn4+	AOS	Fe3+/Fe2+
Mn2O3/γ-Al2O3/Al	642.4	641.0	531.4	529.7	0.38	0.90	2.86	/
Fe4.25-Mn2O3/γ-Al2O3/Al	642.5	641.0	531.2	529.7	0.43	1.03	2.72	1.11
Fe6.61-Mn2O3/γ-Al2O3/Al	642.5	641.1	531.5	529.7	0.50	1.60	2.76	1.42
Fe8.32-Mn2O3/γ-Al2O3/Al	642.5	641.0	531.4	529.8	0.47	1.28	2.70	1.39

). It was observed that a higher Fe³⁺/Fe²⁺ ratio was generally accompanied by improved NH₃ catalytic activity. This enhancement may be attributed to increased surface Fe³⁺ species facilitating NH₃ dehydrogenation [38,39] and thus boosting the NH₃-SCO performance.

Furthermore, although the Fe loading is relatively high (4.25–8.32 wt%), the Fe 2p signal intensity in the XPS spectra is much weaker than that of Mn. This observation suggests that Fe species are mainly located beneath the Mn layer rather than being enriched on the catalyst surface. Consequently, the catalyst surface is dominated by Mn species, which is consistent with the EDS results.

The redox properties of the catalysts were evaluated using H₂-TPR, as shown in Figure 7a. Two reduction processes were observed for all catalysts: the transformation from Mn₂O₃ to Mn₃O₄ and from Mn₃O₄ to MnO [12]. Mn₂O₃/γ-Al₂O₃/Al showed peaks at 382 °C and 478 °C, Fe_4.25-Mn₂O₃/γ-Al₂O₃/Al at 369 °C and 474 °C, Fe_6.61-Mn₂O₃/γ-Al₂O₃/Al at 354 °C and 460 °C, and Fe_8.32-Mn₂O₃/γ-Al₂O₃/Al at 358 °C and 462 °C. Compared to Mn₂O₃/γ-Al₂O₃/Al, Fe doping shifts both peaks to lower temperatures. As Fe loading increases, the reduction peak temperatures first decrease, then slightly increase. The reduction temperature correlates with catalytic performance: lower reduction temperatures typically result in higher activity [40]. The order of reduction temperatures is: Mn₂O₃/γ-Al₂O₃/Al > Fe_4.25-Mn₂O₃/γ-Al₂O₃/Al > Fe_8.32-Mn₂O₃/γ-Al₂O₃/Al > Fe_6.61-Mn₂O₃/γ-Al₂O₃/Al, indicating that Fe_6.61-Mn₂O₃/γ-Al₂O₃/Al has stronger reducibility at low temperatures. Additionally, Fe_6.61-Mn₂O₃/γ-Al₂O₃/Al exhibits a larger reduction peak area, suggesting a higher concentration of active oxygen species [41].

O₂-TPD was used to investigate the amount and binding strength of surface active oxygen species on the catalysts. It was known that oxygen species desorbed below 300 °C are typically associated with highly mobile surface-adsorbed oxygen [42,43], while peaks in the 300–600 °C range correspond to the release of lattice oxygen [43]. Since the NH₃-SCO reaction occurs in the 100–300 °C range, the low-temperature desorption peaks reflect the catalyst’s oxygen supply capability under reaction conditions. As shown in Figure 7b, Fe_6.61-Mn₂O₃/γ-Al₂O₃/Al shows the largest low-temperature peak area, indicating a higher proportion of weakly bound active oxygen. Fe doping enhances O₂ activation and dissociation, leading to increased active oxygen species and better oxygen migration. Increasing Fe loading from 0 to 6.61 wt% significantly boosts desorbed oxygen species. However, at 8.32 wt%, excess Fe may form aggregates or compete with Mn-O-Mn bonds, reducing oxygen vacancy formation and slightly lowering desorbed oxygen species compared to Fe_6.61-Mn₂O₃/γ-Al₂O₃/Al. This trend aligns with the O_ads/O_latt ratio from XPS, where Fe_6.61-Mn₂O₃/γ-Al₂O₃/Al shows the highest O_ads ratio, confirming the O₂-TPD results.

As shown in Figure 7c, the variation in O_ads/O_total closely correlates with the redox properties of the samples: catalysts with a higher proportion of surface-adsorbed oxygen exhibit stronger oxygen mobility and redox ability, as indicated by a lower T₉₀ (the temperature at which NH₃ conversion reaches 90%) and enhanced NH₃-SCO activity.

To investigate the enhancement of oxygen activation in Mn₂O₃ by Fe doping in NH₃-SCO, an analysis from an electronic structure perspective was performed. DFT calculations revealed that the O₂ adsorption energy of Fe-Mn₂O₃ is −2.10 eV, whereas for undoped Mn₂O₃(222) it is only −0.68 eV (Figure 7d). The more negative adsorption energy suggests that the Fe-doped system forms stronger chemisorption with O₂, thereby promoting oxygen activation and dissociation on the catalyst surface.

Therefore, it is reasonable to conclude that a higher proportion of surface-adsorbed oxygen species, in conjunction with its superior redox properties, plays a key role in endowing the Fe_6.61-Mn₂O₃/γ-Al₂O₃/Al catalyst with excellent NH₃-SCO activity.

It is well established that surface acidity plays a crucial role in NH₃ adsorption and activation during the NH₃-SCO reaction. Therefore, NH₃-TPD was conducted to investigate the acidic properties of the Fe_x-Mn₂O₃/γ-Al₂O₃/Al catalysts. As shown in Figure 8, Fe incorporation significantly altered the acidic structure. All Fe-doped samples exhibited NH₃ desorption peaks around 200 °C, which are assigned to weak acid sites [44]. The undoped Mn₂O₃/γ-Al₂O₃/Al showed much weaker intensity in this region. In addition, Fe-doped catalysts displayed clear desorption peaks at approximately 350 °C, corresponding to medium-strength acid sites [45]. These peaks were absent in the undoped sample, indicating that Fe doping promotes the formation of medium-strength acid sites. As the Fe loading increased, the intensity of the medium-temperature peaks gradually increased. Among the samples, Fe_6.61-Mn₂O₃/γ-Al₂O₃/Al exhibited the strongest medium-strength acidity, which may account for its superior NH₃-SCO performance.

To further clarify the intrinsic effect of Fe doping on reaction kinetics, kinetic analysis was carried out over Mn₂O₃/γ-Al₂O₃/Al and Fe_6.61-Mn₂O₃/γ-Al₂O₃/Al catalysts. To eliminate diffusion effects, 1 g of catalyst was mixed and diluted with 5 g of quartz sand of the same particle size, and the catalytic tests were conducted at a space velocity of 165,000 mL·g_cat⁻¹·h⁻¹. Under these conditions, the NH₃ conversion was maintained below 15%, satisfying the criteria for excluding internal and external diffusion limitations. The reaction rate (r) and apparent activation energy (Eₐ) were calculated according to Equations (1)–(3):

$$\begin{array}{c}{\mathrm{f}}_{{\mathrm{N}\mathrm{H}}_3}(\mathrm{m}\mathrm{o}\mathrm{l}·{\mathrm{h}}^{-1}) = {\displaystyle \frac{{\mathrm{Q}}_{{\mathrm{N}\mathrm{H}}_3}({\mathrm{m}}^3·{\mathrm{h}}^{-1})}{{\mathrm{V}}_{\mathrm{m}}({\mathrm{m}}^3·{\mathrm{mol}}^{-1})}}\end{array}$$

(1)

$$\begin{array}{c}\mathrm{r} (\mathrm{m}\mathrm{o}\mathrm{l}·{\mathrm{g}}^{-1}·{\mathrm{h}}^{-1})={\displaystyle \frac{{\mathrm{X}}_{{\mathrm{N}\mathrm{H}}_3} (\mathrm{vol}\%) \times {\mathrm{f}}_{{\mathrm{N}\mathrm{H}}_3}(\mathrm{mol}·{\mathrm{h}}^{-1})}{{\mathrm{m}}_{\mathrm{cat}} (\mathrm{g})}}\end{array}$$

(2)

$$\begin{array}{c}\mathrm{l}\mathrm{n}\mathrm{r}=-{\displaystyle \frac{{\mathrm{E}}_{\mathrm{a}}}{\mathrm{R} \times \mathrm{T}}}+\mathrm{l}\mathrm{n}\mathrm{A}\end{array}$$

(3)

As shown in Figure 9, the apparent activation energy (Ea) for the Mn₂O₃/γ-Al₂O₃/Al catalyst was 58.03 kJ·mol⁻¹, while for the Fe-modified Fe_6.61-Mn₂O₃/γ-Al₂O₃/Al catalyst, Ea decreased to 42.11 kJ·mol⁻¹. The lower activation energy accelerates the NH₃ oxidation reaction at lower temperatures, further confirming that Fe modification enhances reaction efficiency by reducing the reaction energy barrier.

2.3. Effect of Fe Modification on the H₂O and SO₂ Resistance Behavior of the Catalyst

Industrial flue gas typically contains complex components, among which H₂O and SO₂ are two major impurities. Therefore, evaluating the tolerance of catalysts toward H₂O and SO₂ is essential. Firstly, the water resistance of Mn₂O₃/γ-Al₂O₃/Al and Fe_6.61-Mn₂O₃/γ-Al₂O₃/Al was evaluated at their respective temperatures for complete NH₃ conversion. As shown in Figure 10a, after introducing 10 vol.% H₂O, the NH₃ conversion gradually decreased and then stabilized. The steady-state conversion was 65% for Mn₂O₃/γ-Al₂O₃/Al and 80% for Fe_6.61-Mn₂O₃/γ-Al₂O₃/Al. This decline is mainly attributed to the competitive adsorption of H₂O and NH₃ on active sites [46]. After H₂O was removed from the feed, both catalysts recovered their activity. The NH₃ conversion over Mn₂O₃/γ-Al₂O₃/Al returned to 100% within 40 min. In contrast, Fe_6.61-Mn₂O₃/γ-Al₂O₃/Al recovered to full conversion within 20 min.

Secondly, as shown in Figure 10b, the sulfur resistance of Mn₂O₃/γ-Al₂O₃/Al and Fe_6.61-Mn₂O₃/γ-Al₂O₃/Al was evaluated at their respective temperatures for complete NH₃ conversion. For Mn₂O₃/γ-Al₂O₃/Al, NH₃-SCO activity was stable at 100% conversion with 200 ppm SO₂, but after 260 min of SO₂ exposure, deactivation occurred, and the conversion dropped to 67% after 360 min. No recovery was observed after SO₂ removal, mainly due to sulfation of active sites and ammonium sulfate deposition [47]. In contrast, Fe_6.61-Mn₂O₃/γ-Al₂O₃/Al showed enhanced SO₂ tolerance, maintaining complete NH₃ conversion for 320 min under 200 ppm SO₂. After 360 min, the conversion remained above 98%. After SO₂ removal, conversion decreased to 92% and stabilized. Figure 10c shows H₂-TPR profiles of both catalysts before and after SO₂ exposure. For Mn₂O₃/γ-Al₂O₃/Al, after the SO₂ reaction, low-temperature reduction peaks (364 °C and 464 °C) decreased significantly, and new peaks appeared at 685 °C and 717 °C, indicating the formation of inert sulfate species [48], In contrast, Fe_6.61-Mn₂O₃/γ-Al₂O₃/Al maintained the low-temperature reduction peak (367 °C), with weak high-temperature sulfate reduction peaks, demonstrating that Fe doping effectively suppresses sulfation poisoning.

2.4. Reaction Mechanism of NH₃-SCO over Fe-Mn₂O₃

To reveal the reaction mechanism, in situ FTIR spectroscopy was systematically employed to observe NH₃ adsorption behavior and the dynamic evolution of surface intermediates on Fe-Mn₂O₃ catalysts between 50 and 300 °C. As shown in Figure 11a, absorption peaks at 1087, 1162, 1210–1212, and 1235–1245 cm⁻¹ correspond to NH₃ adsorbed on Lewis acid sites (LA) [49,50,51], while the characteristic peaks at 1357 and 1367 cm⁻¹ can be attributed to NH₄⁺ on Brønsted acid (BA) sites [52]. As the temperature increases, the intensity of the NH₃/NH₄⁺-related peaks gradually strengthens, indicating that the Fe-Mn₂O₃ surface possesses both LA and BA sites and has good NH₃ adsorption and activation abilities. Meanwhile, during the heating process, the formation of nitrate intermediates was clearly observed, including bridging nitrates (1619 cm⁻¹), bidentate nitrates (1557, 1545 cm⁻¹), and monodentate nitrates (1536, 1507, 1453 cm⁻¹) [52,53,54]. It is widely accepted that nitrate species are key intermediates in the i-SCR pathway [55,56]. Therefore, the above results indicate that the NH₃-SCO reaction on Fe-Mn₂O₃ mainly follows the i-SCR mechanism. Furthermore, the broad peak at 3200–3500 cm⁻¹ can be assigned to N-H stretching vibrations (NH₃, NH₄⁺, -NH₂, and -NH) [52,53], and its intensity significantly increases with temperature, suggesting that the adsorbed NH₃ and NH₄⁺ species are gradually converted to -NH_x species and nitrate intermediates with the involvement of the catalyst’s lattice oxygen [55].

To further elucidate the role of oxygen during the reaction, Fe-Mn₂O₃ was pre-adsorbed with NH₃ for 30 min at 300 °C, followed by the introduction of 10 vol.% O₂, and the evolution of surface species is shown in Figure 11b. It can be observed that the peak intensity of bridging nitrates (1619 cm⁻¹) and bidentate and monodentate nitrates (1557–1507 cm⁻¹) remained stable during the entire O₂ introduction process, without significant accumulation. Meanwhile, the LA-NH₃ (1235, 1086 cm⁻¹), BA-NH₄⁺ (1367 cm⁻¹), monodentate nitrate (1453 cm⁻¹), and N-H vibration peaks at 3100–3500 cm⁻¹ gradually weakened with the introduction of O₂. Notably, the characteristic peak of -NH species (1463 cm⁻¹) [50], which appeared during NH₃ pre-adsorption, also gradually diminished and eventually disappeared as the reaction progressed.

These results indicate that the -NH_x species on the Fe-Mn₂O₃ surface continue to participate in the reaction and are gradually consumed in the O₂ atmosphere. However, their consumption does not lead to significant accumulation of nitrate intermediates. This suggests that once nitrate species are formed, they can rapidly react with the adjacent -NH_x species and be converted into N₂, following the i-SCR reaction pathway (-NH_x + NO_x⁻ → N₂). This process forms a relatively balanced and mild redox cycle on Fe-Mn₂O₃, effectively preventing excessive accumulation of nitrate and high-valent nitrogen oxides, thus explaining its high N₂ selectivity and suppression of N₂O by-products.

Based on the above in situ FTIR results, the following reaction pathway can be reasonably deduced for the NH₃-SCO reaction on Mn-based catalysts (Figure 12): First, NH₃ molecules are adsorbed on the Lewis acid and Brønsted acid sites as NH₃ and NH₄⁺, respectively. Then, with the participation of surface active oxygen species, the adsorbed NH₃/NH₄⁺ species are gradually dehydrogenated and oxidized to form -NH₂ and -NH intermediates (Equations (4)–(6)). During this process, some -NH₂ species are further oxidized by gaseous O₂ or highly active surface oxygen species, generating surface NO_x species (mainly in the form of NO₂⁻/NO₃⁻) (Equation (7)). The formed surface NO_x species can couple with adjacent -NH species and be reduced through the i-SCR pathway, ultimately forming N₂ (Equation (8)). However, when the supply of surface -NH species is insufficient or overly oxidized, some NO_x species will not be fully reduced, and their reduction process will remain in the intermediate oxidation state, leading to the formation of N₂O by-products through incomplete reduction (Equation (9)).

$${\mathrm{N}\mathrm{H}}_3+\left[\mathrm{O}\right]\to -{\mathrm{N}\mathrm{H}}_2$$

(4)

$${\mathrm{N}\mathrm{H}}_4^{+}+\left[\mathrm{O}\right]\to -{\mathrm{N}\mathrm{H}}_2$$

(5)

$$-{\mathrm{N}\mathrm{H}}_2+\left[\mathrm{O}\right]\to -\mathrm{N}\mathrm{H}$$

(6)

$$-\mathrm{N}\mathrm{H}+\left[\mathrm{O}\right]\to -{\mathrm{N}\mathrm{O}}_{\mathrm{X}}^{-}$$

(7)

$$-{\mathrm{N}\mathrm{O}}_{\mathrm{X}}^{-}+-\mathrm{N}\mathrm{H}\to {\mathrm{N}}_2+{\mathrm{H}}_2\mathrm{O}$$

(8)

$${-\mathrm{N}\mathrm{O}}_{\mathrm{X}}^{-}\to {\mathrm{N}}_2\mathrm{O}$$

(9)

3. Experimental Section

3.1. Catalyst Preparation

Commercial aluminum mesh (1060-type) was used as the substrate. The aluminum mesh was first sequentially treated in 10 wt% NaOH and 10 wt% HNO₃ solutions for several minutes to remove surface grease and impurities. After rinsing with deionized water and drying, the mesh was anodized in an oxalic acid electrolyte under a constant current density of 30 A·m⁻² for 12 h, resulting in the formation of an ordered anodic aluminum oxide (AAO) film with vertically aligned pore channels. The obtained AAO was calcined at 350 °C for 1 h to remove residual oxalic acid, followed by hydration in deionized water at 80 °C for 1 h [19]. Subsequently, the sample was calcined at 500 °C for 4 h, yielding the mesh-structured γ-Al₂O₃/Al support.

Fe-modified γ-Al₂O₃/Al supports were prepared via an excess impregnation method. Briefly, γ-Al₂O₃/Al was immersed in an aqueous Fe(NO₃)₃·9H₂O solution (0.05 mol·L⁻¹) for 4–18 h. After impregnation, the samples were thoroughly washed with deionized water, dried under ambient conditions, and calcined at 500 °C for 4 h. The resulting samples were denoted as Fe_x/γ-Al₂O₃/Al, where x represents the Fe loading (wt%), with x = 4.25, 6.61, and 8.32. MnO₂/γ-Al₂O₃/Al and Mn₂O₃/γ-Al₂O₃/Al catalysts were synthesized by a homogeneous precipitation method. Specifically, γ-Al₂O₃/Al was immersed in an aqueous solution containing manganese acetate ((CH₃COO)₂Mn·4H₂O) and urea (CH₄N₂O) with a molar ratio of 1:10 and aged for 20 h. After washing with deionized water and drying under ambient conditions, the samples were calcined at 400 °C or 500 °C for 4 h to obtain MnO₂/γ-Al₂O₃/Al and Mn₂O₃/γ-Al₂O₃/Al, respectively. The Fe_x-Mn₂O₃/γ-Al₂O₃/Al catalysts were prepared following the same procedure, except that the γ-Al₂O₃/Al support was replaced by the corresponding Fe_x/γ-Al₂O₃/Al.

3.2. Catalyst Characterization

The morphology of the samples was characterized by field-emission scanning electron microscopy (FESEM, TESCAN MIRA LMS, Brno, Czech Republic). The textural properties were determined by N₂ adsorption–desorption at 77 K using a Micromeritics ASAP 2020-M analyzer (Norcross, GA, USA). Prior to the measurements, all samples were degassed at 350 °C for 12 h, and the specific surface area was calculated by the Brunauer–Emmett–Teller (BET) method. X-ray diffraction (XRD) patterns were recorded on a Bruker (Billerica, MA, USA) D8 Advance diffractometer with Cu Kα radiation, and the data were collected in the 2θ range of 10–80°. The surface chemical states of the elements were analyzed by X-ray photoelectron spectroscopy (XPS, Thermo Scientific K-Alpha, Waltham, MA, USA), and the binding energies were calibrated using the C 1s peak at 284.8 eV. The reducibility, oxygen mobility, and surface acidity of the catalysts were investigated by H₂ temperature-programmed reduction (H₂-TPR), O₂ temperature-programmed desorption (O₂-TPD), and NH₃ temperature-programmed desorption (NH₃-TPD), respectively, using an AutoChem (Singapore) 2720 chemisorption analyzer. For H₂-TPR, approximately 100 mg of sample was pretreated in a He flow (20 mL·min⁻¹) at 200 °C for 60 min and then cooled to 40 °C. The reduction was carried out from 40 to 600 °C at a heating rate of 10 °C min⁻¹ under a 5 vol% H₂/He flow (20 mL·min⁻¹). For O₂-TPD, about 100 mg of sample was pretreated at 80 °C in He (20 mL·min⁻¹) for 60 min and cooled to 40 °C, followed by oxidation under 5 vol% O₂/He (20 mL·min⁻¹) for 30 min. After purging with He (30 mL·min⁻¹) for 1 h, desorption was performed from 40 to 600 °C at a heating rate of 10 °C min⁻¹. The NH₃-TPD measurements were conducted following a similar procedure, except that NH₃/Ar was used as the adsorbing gas. In situ Fourier transform infrared (FTIR) spectroscopy with modulation excitation spectroscopy (MES) was performed on a Bruker Tensor 27 spectrometer equipped with a mercury–cadmium–telluride (MCT) detector to identify reaction intermediates. The spectra were recorded at a resolution of 4 cm⁻¹ with 32 scans.

3.3. Catalytic Activity Evaluation

The NH₃ selective catalytic oxidation (NH₃-SCO) performance of the prepared monolithic catalysts (2.5 g) was evaluated in a continuous-flow fixed-bed reactor. As shown in Figure 13, the catalytic activity evaluation setup consisted of a continuous-flow fixed-bed reaction system. The concentrations of NH₃ in the inlet stream and NH₃, NO, NO₂, and N₂O in the outlet stream were analyzed by an online gas chromatograph (GC6600). The feed gas consisted of 3000 ppm NH₃, air, with Ar as the balance gas. The total flow rate was maintained at 340 mL·min⁻¹, corresponding to a weight hourly space velocity (WHSV) of approximately 25,000 mL·g_cat⁻¹·h⁻¹. The NH₃ conversion, N₂ selectivity, and N₂ yield were calculated according to the following equations:

$${\mathrm{N}\mathrm{H}}_3 \mathrm{C}\mathrm{o}\mathrm{n}\mathrm{v}\mathrm{e}\mathrm{r}\mathrm{s}\mathrm{i}\mathrm{o}\mathrm{n}(\%)={\displaystyle \frac{{\mathrm{N}\mathrm{H}}_{3 \mathrm{i}\mathrm{n}}-{\mathrm{N}\mathrm{H}}_{3 \mathrm{o}\mathrm{u}\mathrm{t}}}{{\mathrm{N}\mathrm{H}}_{3 \mathrm{i}\mathrm{n}}}}\times 100\%$$

(10)

$${\mathrm{N}}_2 \mathrm{S}\mathrm{e}\mathrm{l}\mathrm{e}\mathrm{c}\mathrm{t}\mathrm{i}\mathrm{v}\mathrm{i}\mathrm{t}\mathrm{y}(\%)={\displaystyle \frac{{\mathrm{N}\mathrm{H}}_{3 \mathrm{i}\mathrm{n}}-{\mathrm{N}\mathrm{H}}_{3 \mathrm{o}\mathrm{u}\mathrm{t}}-{\mathrm{N}\mathrm{O}}_{\mathrm{o}\mathrm{u}\mathrm{t}}-{\mathrm{N}\mathrm{O}}_{2 \mathrm{o}\mathrm{u}\mathrm{t}}-2\times {\mathrm{N}}_2{\mathrm{O}}_{\mathrm{o}\mathrm{u}\mathrm{t}}}{{\mathrm{N}\mathrm{H}}_{3 \mathrm{i}\mathrm{n}}-{\mathrm{N}\mathrm{H}}_{3 \mathrm{o}\mathrm{u}\mathrm{t}}}}\times 100\%$$

(11)

$${\mathrm{N}}_2 \mathrm{y}\mathrm{i}\mathrm{e}\mathrm{l}\mathrm{d}(\%)={\displaystyle \frac{{\mathrm{N}\mathrm{H}}_3 \mathrm{C}\mathrm{o}\mathrm{n}\mathrm{v}\mathrm{e}\mathrm{r}\mathrm{s}\mathrm{i}\mathrm{o}\mathrm{n}\times {\mathrm{N}}_2 \mathrm{S}\mathrm{e}\mathrm{l}\mathrm{e}\mathrm{c}\mathrm{t}\mathrm{i}\mathrm{v}\mathrm{i}\mathrm{t}\mathrm{y}}{100}}$$

(12)

3.4. DFT Method

First-principles calculations were performed using the Vienna Ab initio Simulation Package (VASP 6.3.2). The exchange–correlation interactions were described by the Perdew–Burke–Ernzerhof (PBE) functional within the generalized gradient approximation (GGA). The projector augmented-wave (PAW) method was employed with a plane-wave cutoff energy of 450 eV. All structures were fully optimized until the residual forces on each atom were less than 0.02 eV Å⁻¹ and the total energy convergence criterion reached 1 × 10⁻⁵ eV. The Brillouin zone was sampled using a 1 × 1 × 1 Monkhorst-Pack k-point mesh for geometry optimizations. A vacuum layer of 15 Å was introduced to avoid interactions between periodic slabs. To properly describe the strong on-site Coulomb interactions of transition-metal 3d electrons, the DFT+U approach based on the Dudarev scheme was adopted, with effective U values of 3.9 eV for Mn and 4.0 eV for Fe.

The adsorption energy of O₂ (E_ads) can be defined as:

E_ads = E_total − E_substrate − E_(O-O)

(13)

where E_total denotes the total energy of the O₂-adsorbed substrate, and E_substrate and E_(O-O) are the energies of the substrate and free O₂, respectively.

4. Conclusions

In this work, a novel structured monolithic Fe-Mn₂O₃/γ-Al₂O₃/Al catalyst was developed to achieve simultaneous high NH₃ conversion and N₂ selectivity at relatively low temperature. The results demonstrate that Fe doping increases the proportion of reactive surface oxygen species (O_ads) and strengthens the redox properties of the catalyst, thereby facilitating oxygen activation and effectively improving the catalytic activity toward NH₃ oxidation. DFT calculations further reveal that the introduction of Fe promotes O₂ adsorption on the catalyst surface. Among the catalysts investigated, Fe_6.61-Mn₂O₃/γ-Al₂O₃/Al achieves complete NH₃ conversion at 210 °C with a high N₂ selectivity of 75%. As shown in

Table 4. Comparison of Mn-based catalysts for NH₃-SCO.

Catalysts	Reaction Conditions	Reaction Temperature/°C	NH3 Conversion/% (N2 Selectivity/%)	Reference
MnO2	NH3 = 1000 ppm; GHSV = 30 L/g·h−1	170	100 (49)	[8]
Mn/Ce-Ti	NH3 = 50 ppm; GHSV = 40,000 h−1	300	77 (93)	[57]
MnOx/TiO2	NH3 = 500 ppm; 200 mL·min−1	350	93 (80)	[58]
MnMoOx	NH3 = 500 ppm; 150 mL·min−1	340	90 (80)	[59]
CoMnAl	NH3 = 5000 ppm; WHSV = 24,000 mL·h−1·g−1	250	100 (40)	[60]
SmMn2O5	NH3 = 500 ppm; 200 mL·min−1	175	100 (45)	[61]
Fe6.61-Mn2O3/γ-Al2O3/Al	NH3 = 3000 ppm; WHSV = 25,000 mL·gcat−1·h−1	210	100 (75)	This work

, this performance represents a more favorable balance between low-temperature activity and N₂ selectivity compared with reported Mn-based catalysts. For example, SmMn₂O₅ reaches 100% NH₃ conversion at 175 °C but exhibits low N₂ selectivity (45%). MnMoO_x shows relatively high N₂ selectivity (80%) but requires a much higher temperature (340 °C) to achieve 90% NH₃ conversion. In contrast, the present Fe-modified structured catalyst achieves full NH₃ conversion at a relatively low temperature while maintaining improved N₂ selectivity, demonstrating superior overall performance. In addition, the catalyst maintains 92% and 80% of its activity in the presence of 200 ppm SO₂ and 10 vol.% H₂O, respectively. In situ FTIR analysis suggests that the NH₃-SCO reaction over Fe-Mn₂O₃/γ-Al₂O₃/Al predominantly proceeds via the internal selective catalytic reduction (i-SCR) pathway.