1 Introduction

As a critical sector for carbon reduction, the transportation industry holds significant responsibility in implementing the dual-carbon strategy. The replacement of traditional petroleum fuels with zero-carbon fuels plays a key role in this process. The development and popularization of ammonia-fueled engines is a good choice for the use of zero-carbon fuels. Undoubtedly, the use of ammonia as a fuel for engines has the potential to significantly decrease CO2 emissions and serves as a robust foundation for the dual-carbon approach. However, the issue of hazardous ammonia emissions that arise from engine fuel leakage and incomplete fuel combustion under special operating conditions must be resolved. Ammonia is a colorless gas with a strong odor at room temperature and pressure and is the most abundant alkaline gas in the atmosphere. It significantly contributes to haze formation by reacting with gaseous nitric acid and sulfuric acid in the air, producing ammonium sulfate, ammonium nitrate, and other secondary particulate matter. These secondary particulate matters account for 50–70% of the PM2.5 content [1, 2]. Exposure to ammonia concentrations above 500 ppm, can cause eye irritation, and lungs diseases; while levels higher than 10,000 ppm are potentially fatal [3]. Due to the significant risk ammonia poses to the environment and human beings, the management of hazardous ammonia has increasingly attracted extensive attention from society and scholars.

Several techniques have been developed for the management of hazardous ammonia emissions., mainly including absorption decomposition [4], biofiltration [5], catalytic decomposition [6], and selective catalytic oxidation [7,8,9,10]. The absorption decomposition method uses acidic solutions such as HNO3 or H2SO4 to adsorb alkaline NH3 through acid–base neutralization; the biofiltration method uses the growth and metabolism of microorganisms to decompose NH3 in a specific reaction vessel; catalytic decomposition is the direct decomposition of NH3 into N2 and H2 at high temperature, which is the reverse reaction of ammonia synthesis; selective catalytic oxidation is the use of catalysts to convert NH3 into N2 and H2O in a directional manner. Among them, NH3-SCO is considered to be the most effective treatment process for mitigating harmful ammonia emissions from the exhaust of stationary industrial and mobile sources and the main reaction equations are as follows [11]:

$$ 4{\text{NH}}_{3} + 3{\text{O}}_{2} \to 2{\text{N}}_{2} + 6{\text{H}}_{2} {\text{O}} $$
(1)
$$ 4{\text{NH}}_{3} + 5{\text{O}}_{2} \to 4{\text{NO}} + 6{\text{H}}_{2} {\text{O}} $$
(2)
$$ 4{\text{NH}}_{3} + 7{\text{O}}_{2} \to {\text{NO}}_{2} + 6{\text{H}}_{2} {\text{O}} $$
(3)
$$ 2{\text{NH}}_{3} + 2{\text{O}}_{2} \to {\text{N}}_{2} {\text{O}} + 3{\text{H}}_{2} {\text{O}} $$
(4)

The reaction pathway outlined in the process (1) is the most effective method for reducing ammonia pollution. Conversely, processes (2)–(4) involve peroxide side reactions that generate environmentally harmful NOx byproducts, which should be minimized. The catalyst, which is fundamental to the NH3-SCO process, is crucial in determining the efficiency and selectivity of the desired reaction pathway. According to the classification of the main active site, the current mainstream NH3-SCO catalysts are mainly divided into two categories: noble metal catalysts and transition metal oxide catalysts. Scholars have conducted extensive research on precious metals mainly including Pt, Pd, Ru, Au, and Ir [12,13,14,15], which have excellent catalytic activity and stability for the ammonia oxidation reaction process at low temperatures. However, the shortcomings are the high cost of precious metal catalysts and the challenge of low N2 selectivity due to peroxidation, which severely limit their wide application. Therefore, researchers have shifted their focus to transition metal oxides, including MoO3, CuO, CeO2, Fe2O3, TiO2, Co3O4, and MnO2 [16,17,18,19,20,21,22]. Transition metal oxidation catalysts are generally more N2 selective and more abundant at relatively low costs compared to noble metal-based catalysts, although they exhibit lower activity. The ideal catalyst would combine the advantages of both, with high catalytic activity and N2 selectivity for ammonia oxidation. CeO2, a rare earth oxide, is well-known for its abundant oxygen vacancies, excellent redox capacity, and effective oxygen storage and release capabilities, owing to the facile transition between Ce3⁺ and Ce4⁺ ions. Additionally, CeO2 exhibits strong interactions with other metals. These interactions enable the development of high-performance composite oxide NH3-SCO catalysts when CeO2 is combined with a second metal oxide (MOx), despite pure CeO2 having weak NH3 oxidation activity [23]. Transition metal elements such as Co, Cu, Fe, and Zr are considered cost-effective alternatives to precious metals. Among these, Co, Cu, and Fe catalysts exhibit favorable redox properties and catalytic activity, making them extensively utilized in environmental catalysis applications, including NOx selective catalytic reduction, CO oxidation, water vapor shift reaction, and VOC oxidation reaction. Zr is recognized for its exceptional thermal stability, and catalysts incorporating Zr are known for their stability and oxygen storage capabilities. Therefore, exploring the synergistic effects of these four elements in conjunction with CeO2 to develop novel catalysts for NH3-SCO holds significant research potential.

In this research, M metals (M = Co, Cu, Fe, Zr) were integrated into CeO2 using the sol–gel method to achieve a composite oxide catalyst with excellent ammonia oxidation performance and high N2 selectivity. Subsequently, the simulation gas experiments were carried out to evaluate the catalytic performance of CeO2-based catalysts. Furthermore, a series of characterization methods such as BET, SEM, XRD, XPS, H2-TPR, and Raman were employed to analyze the physicochemical properties and explain the effects of the strong interactions between Ce and the metals.

2 Experimental

2.1 Catalyst Preparation

A series of M/CeO2 (M = Co, Cu, Fe, Zr) catalyst samples were prepared using the sol–gel method. Taking Co/ CeO2 as an example, the specific preparation steps were as follows: 40 g of Ce-Co mixed-oxide powders were targeted. The nitrate precursors, cobalt nitrate (Co(NO₃)₂·6H₂O) and cerium nitrate (Ce(NO₃)₃·6H₂O), were mixed in a 1:1 molar ratio. Both nitrates were dissolved in 150 ml of deionized water with magnetic stirring for 1 h. Subsequently, the complexing agent citric acid monohydrate (C₆H₈O₇·H₂O) was slowly added. The amount of C₆H₈O₇·H₂O required was determined based on the valence and quantity of the nitrate precursors, calculated as follows: the molar amount of C₆H₈O₇·H₂O = (Co2⁺ × 2 + Ce3⁺ × 3) × 1.1 (the factor 1.1 ensures complete complexation). After the citric acid addition, the solution was stirred for an additional hour, then transferred to a water bath at 80 °C and heated for 3–5 h with continuous stirring until a viscous gel was formed. The obtained gel was then dried in a constant temperature oven at 110 °C for 24 h to form a spongy substance and then calcined at 400 °C for 3 h to burn off nitrate and carbon species to obtain the finished catalyst sample powder. Finally, a portion of the powder was added with water and binder to adjust the viscosity then coated onto cordierite using a vacuum coater, and then roasted at 400 °C for 3 h to obtain a complete catalyst sample Co/CeO2 that could be tested; the rest of the catalyst powder was retained for characterization. Cu/CeO2, Fe/CeO2, and Zr/CeO2 catalyst samples were produced in the same way.

2.2 Catalysts Characterization

2.2.1 N2 Adsorption/Desorption

The structural characteristics of the catalyst samples, such as specific surface area, pore volume, and pore size, were determined on a JW-BK200B physical adsorption tester (Beijing Jingwei Gao Bo). The pretreatment conditions were: vacuum, 300 °C, 3 h; the testing conditions were: high purity N2 adsorption and desorption. The specific surface area of the samples was calculated by BET (Brunner-Emmett-Teller) equation, and the pore volume pore size, etc. were calculated by BJH (Barret–Joyner–Halenda) analysis.

2.2.2 SEM

The samples were imaged by scanning electron microscopy using a JEOL JSM-6301F scanning electron microscope with a magnification level of 500 nm.

2.2.3 XRD

The crystalline structure analysis of the samples was carried out on a Smart lab SE-type X-ray diffractometer, using Cu Ka rays as the ray source (λ = 0.15406). Test conditions: operating voltage 40 kV, operating current 40 mA; signal acquisition resolution 0.02°/step, scanning speed 5°/min, 2θ range 10°-90°.

2.2.4 XPS

X-ray photoelectron spectroscopy was performed on a Kratos AXIS-Ultra DLD under ultra-high vacuum at 10–9 Pa with Al-Kα radiation as the source (voltage: 15 kV; power: 250 W) and an analytical range of 0–5000 eV. All data were calibrated to a C 1 s binding energy of 284.8 eV. Standard for calibration.

2.2.5 H2-TPR

H2-temperature programmed reduction (H2-TPR) experiments were carried out on a fully automated chemisorption instrument, Auto Chem II 2920, equipped with a thermal conductivity detector (TCD). The catalyst sample powder (50 mg) was pretreated in a pure He gas stream at a gas flow rate of 50 ml/min and a temperature of 350 °C for 30 min. After that, the temperature was lowered to 40 °C, and the powder was heated up to 800 °C in a 10 vol% H2/He gas stream at a temperature rise rate of 10 °C/min, and the change of the signal was detected and recorded by the TCD.

2.2.6 Raman

The instrument used for Raman testing was a Thermo Scientific DXR micro-spectrometer. A He–Cd laser with a wavelength of 532 nm was used for excitation and the power was 3 mW.

2.3 Catalytic Activity Tests

The NH3 oxidation performance of the M/CeO2 catalyst was evaluated using a fixed-bed flow reactor. The fixed-bed flow reactor mainly consists of three parts: an inlet air system, catalytic reaction system, and data acquisition system, as shown in Fig. 1.

Fig. 1
figure 1

Experimental platform of fixed-bed reactor

Full size image

Experimental procedures involved wrapping the catalyst-coated cordierite with quartz cotton and placing it in a heating furnace, forming a cylinder (50 cm length, 24 cm diameter). Gas flow rates were regulated before entering the reaction gas path to stabilize components. A temperature sensor, positioned 0.5–1 cm ahead of the catalyst, monitored real-time gas temperature. Inlet gas contained 500 ppm NH3 and 8% O2, with N2 as the equilibrium gas. Gas hourly space velocity (GHSV) was maintained at 40,000 h−1, with reactions occurring at temperatures ranging from 100 to 600 °C. NOx, N2O, and NH3 concentrations in exhaust gas were analyzed using an FTIR gas analyzer (MKS). In this study, NH3 conversion and N2 selectivity were calculated by the following formulas:

$$ {\text{NH}}_{3} \;{\text{conversion}}\left( \% \right) = \left( {\frac{{\begin{array}{*{20}c} {\left[ {\begin{array}{*{20}c} {{\text{NH}}_{3} } \\ \end{array} } \right]_{{{\text{in}}}} - \left[ {\begin{array}{*{20}c} {{\text{NH}}_{{3}} } \\ \end{array} } \right]_{{{\text{out}}}} } \\ \end{array} }}{{\left[ {\begin{array}{*{20}c} {{\text{NH}}_{3} } \\ \end{array} } \right]_{{{\text{in}}}} }}} \right) \times 100 $$
(5)
$$ {\text{N}}_{2} \;{\text{selectivity}}\left( \% \right) = \left( {\frac{{\left[ {\begin{array}{*{20}c} {{\text{NH}}_{3} } \\ \end{array} } \right]_{{{\text{in}}}} - \left[ {\begin{array}{*{20}c} {{\text{NH}}_{3} } \\ \end{array} } \right]_{{{\text{out}}}} - \left[ {\begin{array}{*{20}c} {{\text{NO}}} \\ \end{array} } \right]_{{{\text{out}}}} - \left[ {\begin{array}{*{20}c} {{\text{NO}}_{{2}} } \\ \end{array} } \right]_{{{\text{out}}}} - 2\left[ {\begin{array}{*{20}c} {{\text{N}}_{2} {\text{O}}} \\ \end{array} } \right]_{{{\text{out}}}} }}{{\left[ {\begin{array}{*{20}c} {{\text{NH}}_{{3}} } \\ \end{array} } \right]_{{{\text{in}}}} - \left[ {\begin{array}{*{20}c} {{\text{NH}}_{3} } \\ \end{array} } \right]_{{{\text{out}}}} }}} \right) \times 100 $$
(6)

3 Results and Discussion

3.1 Catalytic Activity

A comprehensive evaluation of the catalytic performance of M/CeO2 (M = Co, Cu, Fe, Zr) catalysts in the selective catalytic oxidation of ammonia as a function of temperature was carried out, and the results are shown in Fig. 2. Notably all four catalyst samples, Co/CeO2, Cu/CeO2, Fe/CeO2, and Zr/CeO2, exhibited good ammonia oxidation activities, with T50 of 196.8, 229.5, 283.7, and 313.5 °C, respectively, and T90 of 239.2, 292.1, 373.7, and 416.3 °C, respectively. Over a wide range of temperatures, different catalysts converted ammonia to nitrogen at different rates and produced different reaction by-products. Relative to other catalyst types, Zr/CeO2 has lower catalytic activity, shows no ammonia reaction below 250 °C, and has greater T50 and T90 values as depicted in Fig. 2a. At 250 °C, however, Co/CeO2 had exceptional catalytic activity, almost completely converting ammonia. A general order of total ammonia conversion over M/CeO2 catalysts was found to be as follows: Co/CeO2 > Cu/CeO2 > Fe/CeO2 > Zr/CeO2.

Fig. 2
figure 2

Ammonia oxidation activity of M/CeO2 (M = Co, Cu, Fe, Zr) catalysts: a NH3 conversion, b N2 selectivity

Full size image

As can be seen from Fig. 2b, there is a trend of decreasing N2 selectivity with increasing reaction temperature for all catalyst samples. The N2 selectivity of ammonia oxidation over M/CeO2 catalysts is ranked as Co/CeO2 < Cu/CeO2 < Fe/CeO2 < Zr/CeO2. The results revealed an inverse relationship between ammonia conversion activity and N2 selectivity in the catalyst samples. Higher ammonia conversion activity resulted in lower N2 selectivity, while lower ammonia conversion activity enhanced N2 selectivity. Among the four catalysts studied, Cu/CeO2 exhibited the most balanced performance, achieving both high ammonia conversion and favorable N2 selectivity. This optimal performance was attributed to the distinct synergistic effects between the copper and cerium oxide components.

3.2 Catalyst Characterization

3.2.1 BET Analysis

Figure 3a shows the N2 adsorption/desorption isotherms for the M/CeO2 (M = Co, Cu, Fe, Zr) given in Fig. 3a, used to determine their specific surface area and pore structure. From the adsorption and desorption curves of nitrogen, all the samples exhibit similar isothermal curves i.e. type IV isotherms. The hysteresis loops are all in the relative pressure (P/P0) range of 0.4–1.0 and are classified as H3 hysteresis by IUPAC classification, indicating a slit-like pore structure due to the aggregation of lamellar particles. The pore size distribution of each sample was determined by the BJH method, as shown in Fig. 3b, ranging from 3 to 16 nm, within the mesopore size (2–50 nm). This confirms that all the samples exhibit typical mesoporous structures. Among the catalysts, Fe/CeO2 catalyst demonstrated broad pore size distribution, while Co/CeO2, Cu/CeO2, and Zr/CeO2, had narrower, similar pore size distribution ranges. The corresponding specific surface area, pore size, and pore volume of each catalyst sample are detailed in Table 1. The superior ammonia oxidation capabilities of Co/CeO2 could be attributed to its optimized surface areas, while the good performance of Cu/CeO2 may be linked to its minimal aperture. In contrast, the reduced performance of Zr/CeO2 could be ascribed to its limited surface area and high pore size.

Fig. 3
figure 3

a N2 adsorption/desorption isotherms and b the corresponding pore size distribution curves of M/CeO2 (M = Co, Cu, Zr, Fe)

Full size image
Table 1 Pore structure parameters of prepared samples
Full size table

3.2.2 SEM Analysis

Figure 4a-d shows the scanning electron micrographs of M/CeO2 (M = Co, Cu, Fe, Zr) catalysts synthesized by the citric acid sol–gel method. Due to the high viscosity of the sol–gel, the surface of all the samples was not regular and showed aggregates with disordered shapes, which somewhat reduced the dispersion of the particles [24]. Figure 4a illustrates that the surface of the Co/CeO2 is notably rough, characterized by numerous protrusions, depressions, and a substantial number of aggregates. This observation is consistent with the data presented in Table 1, which indicates that Co/CeO2 possesses the largest specific surface area. A larger specific surface area can expose more active sites, potentially enhancing the adsorption and conversion of NH3, thereby improving the efficiency of ammonia oxidation. [25], this also corresponds to the best ammonia oxidation activity of the Co/CeO2 sample in Fig. 2a. In Fig. 4c, the Cu/CeO2 sample shows a disseminated shape with a high number of pores, but not many particles are distributed on the surface, possibly due to the mosaic inside a large number of pores. The Fe/CeO2 and Zr/CeO2 samples in Fig. 4b, d are relatively smooth, with more rounded and sparse pores. This indicates that these two samples exhibit easier sintering during the preparation process, consistent with the observed lower ammonia oxidation activity presented in Fig. 2a.

Fig. 4
figure 4

Scanning electron micrographs of the four catalyst samples: a Co/CeO2, b Fe/CeO2, c Cu/CeO2, d Zr/CeO2

Full size image

3.2.3 XRD Analysis

To investigate the physical phases of M/CeO2 (M = Co, Cu, Fe, Zr), wide-angle XRD tests were conducted. Figure 5 shows the XRD spectra for the Co/CeO2, Cu/CeO2, Fe/CeO2, and Zr/CeO2 produced after roasting at 400 °C. From the figure, it can be seen that the above catalysts show XRD diffraction peaks around 2θ = 28.83°, 33.41°, 47.97°, 56.93°, 59.71°, and their diffraction peak positions are in agreement with the standard card of CeO2 (PDF#97–019-2225). The study reveals CeO2 as the primary phase among the catalysts examined, maintaining its cubic fluorite structure i.e., face-centered cubic structure, even after incorporating transition metals (M = Co, Cu, Fe, Zr) via the citric acid sol–gel method. Only minor characteristic peaks corresponding to Fe2O3 and ZrO2 were observed in the XRD spectra of the Fe/CeO2 and Zr/CeO2 samples, suggesting a homogeneous distribution of Fe and Zr species on the surface of the catalyst. Analysis of the Cu/CeO2 spectrum revealed distinct characteristic diffraction peaks observed at 2θ = 35.6° and 38.8°, which could be attributed to CuO (PDF#00-005-0661), suggesting that aggregation and crystallization of certain CuO species taking place on the catalyst surface. For the Co/CeO2 catalyst sample, characteristic diffraction peaks were observed at 2θ = 31.7° and 36.8°, corresponding to Co3O4 (PDF#04-005-4386), suggesting that the Co species were not uniformly dispersed on the catalyst surface.

Fig. 5
figure 5

XRD spectra of four catalyst samples of M/CeO2 (M = Co, Cu, Fe, Zr)

Full size image

3.2.4 XPS Analysis

The XPS results of M/CeO2 (M = Co, Cu, Fe, Zr) are shown in Fig. 6, while Table 2 shows the elemental and atomic ratios of each catalyst. XPS analysis effectively characterizes the valence states of active substances present on the surface of the catalyst and quantifies the atomics distribution. The XPS profiles of M (M = Co, Cu, Fe, Zr) are shown in Fig. 6a. In the case of Co, the high-resolution Co 2p spectrum reveals two prominent peaks at roughly 780.0 and 795.5 eV which corresponds to Co 2p3/2 and Co 2p1/2, respectively. In addition, the characteristic satellite peaks of Co oxides such as Co3O4, Co2O3, and CoO can be observed at the position around 786 eV [26]. It is clear that the specific valence state of the cobalt ion cannot be determined from the main peak alone and needs to be confirmed in combination with the satellite peaks. A strong oscillatory satellite peak belonging to paramagnetic Co2+ was observed around 786 eV, which is 6 eV higher than the binding energy of Co 2p3/2, the antimagnetic low-spin Co3+ lacks such an oscillatory peak [27]. By comparing the ratio of peak intensities between the oscillatory peak and the main peak, an approximate assessment of the Co2+ content can be inferred.

Fig. 6
figure 6

XPS profiles of M: a Co 2p, b Fe 2p, c Cu 2p, d Zr 3d

Full size image
Table 2 XPS data for different M/CeO2 (M = Co, Cu, Fe, Zr) catalysts: surface atomic concentration of M, Ce, O, and valence distribution of elements
Full size table

The spectral results of Fe 2p of the Fe/CeO2 catalyst samples are shown in Fig. 6b. The primary peaks of Fe 2p3/2 and Fe 2p1/2, with binding energies approximately at 711.3 and 724.4 eV, respectively, can be deconvoluted into two distinct peaks corresponding to the characteristic signals of Fe2+ and Fe3+. This observation indicates that Fe2+ and Fe3+ coexist on the surface of the synthesized Fe/CeO2 catalyst samples. The relative concentrations of Fe2⁺ and Fe3⁺ can be determined from the characteristic peak areas. Previous studies [28] have noted that the presence of Ce on the catalyst surface can increase the Fe2⁺ content. This is likely due to the synergistic interaction between Fe and Ce, which facilitates the redox equilibrium of Fe3⁺ + Ce3⁺ ↔ Fe2⁺ + Ce4⁺. Figure 6c depicts the Cu 2p spectra of the Cu/CeO2 catalyst sample. The main peak of Cu 2p3/2 appears at approximately 934.0 eV, with a satellite peak in the range of 938.1–946.2 eV, both of which are indicative of CuO [29]. In contrast, Cu⁺ generally displays lower binding energy, and its distinctive peaks are not visible in the figure. This observation implies that Cu⁺ species are minimally present, suggesting that copper predominantly exists as CuO on the catalyst surface. Figure 6d shows the Zr 3d spectra of the Zr/CeO2 catalyst sample. The main peaks of Zr 3d5/2 are observed at 182.2 and 184.6 eV, with a binding energy difference of 2.4 eV. This is consistent with the standard spectrum for Zr 3d, a typical feature of ZrO2, indicating that zirconium on the catalyst surface predominantly exists as ZrO2, with few or no other zirconium oxides present.

The Ce 3d photoelectron spectra of all catalyst samples are shown in Fig. 7a, which all have similar peak patterns, and the two sets of spin–orbit coupling curves, Ce 3d5/2 and Ce 3d3/2, are denoted by the v-series and u-series, respectively, for ease of differentiation. Each M/CeO2 catalyst sample Ce 3d spectral profile is characterized by eight distinct peaks arranged from low to high binding energy as follows: v(882.0 eV), v′(884.2 eV), v″(888.6 eV), v‴(898.2 eV), u(900.5 eV), u′(902.2 eV), u″(907.4 eV) and u‴(916.6 eV). Among them, two characteristic peaks, u′ and v′, are attributed to the characteristic peaks of Ce3+, while the other six characteristic peaks are attributed to the characteristic peaks of Ce4+. The molar ratio of Ce3+/Ce4+ on the surface of the samples as determined from Ce 3d mapping of every catalyst is represented in Table 2. These ratios are derived from the area under the deconvolution integral peaks. In CeO2 crystals, when Ce4+ is converted to Ce3+, CeO2 undergoes deoxygenation and generates oxygen vacancies to balance the charge, and this process can be expressed as 4Ce4+  + 2O lattice → 4Ce3+  + 2OV + O2 (O lattice is lattice oxygen)[30]. Therefore, the amount of Ce3+ is directly related to oxygen vacancies, and more Ce3+ represents more lattice oxygen converted to oxygen vacancies, which is conducive to O2 adsorption and activation, and thus improves the NH3 oxidation capacity of the catalysts. The Ce3+/Ce4+ molar ratios of each catalyst were ranked as follows: Co/CeO2 > Cu/CeO2 > Fe/CeO2 > Zr/CeO2, this is in perfect agreement with the ranking of ammonia-oxidizing activity above.

Fig. 7
figure 7

XPS profiles of M/CeO2: a Ce 3d, b O 1s

Full size image

In the study of metal oxide catalysts, understanding how oxygen binds to the metal is crucial for assessing catalytic performance. O 1s spectroscopy is commonly used to identify oxygen types on catalyst surfaces. Figure 7b displays O 1s spectra for all samples, revealing binding energies of 529.7 eV for Oα, 531.6 eV for Oβ, and 532.4 eV for Oγ. Among them, Oα is attributed to lattice oxygen (O2−), and both Oβ and Oγare uniformly attributed to surface adsorbed oxygen OA, but Oβ contains more molecularly adsorbed oxygen and O− species, while Oγ refers more to the oxygen in the hydroxyl group attached to the metal M (M-OH) [31, 32]. In general, surface-adsorbed oxygen OA has better mobility [33], it has higher NH3 oxidizing activity than lattice oxygen, and more OA means better oxidizing activity. However, among the four M/CeO2 prepared catalysts, as shown in Table 2, the relative content of surface adsorbed oxygen OA (Oβ + Oγ) is the same for each catalyst, i.e., the difference in oxidizing activity cannot be explained by OA. However, it can also be found that the order of the size of the percentage of Oγ in the total O content of each catalyst corresponds to the order of the Ce3+/Ce4+ concentration, indicating that for all the catalysts prepared in this paper, there is a synergistic effect between Oγ and Ce3+/Ce4+, with larger values corresponding to better NH3 oxidation activity.

3.2.5 H2-TPR Analysis

It is widely recognized by many scholars that the catalytic performance of catalysts is closely related to their reducing properties, and the H2-TPR technique is usually employed to detect the reducing properties of catalysts. The H2-TPR curves of different catalysts M/CeO2 (M = Co, Cu, Fe, Zr) are shown in Fig. 8. In previous studies, the presence of a major reduction peak near 500 °C in pure CeO2 was attributed to the reduction of surface oxygen [34,35,36]. The catalysts Co/CeO2, Zr/CeO2, and Fe/CeO2 exhibited distinct reduction peaks of surface oxygen at temperatures of 472.2, 519.2, and 595.8 °C, respectively. These variations in peak temperatures, characterized by both decreasing and increasing trends around the peak centers, are indicative of the interactions occurring between the CeO2 support and the Co, Zr, and Fe metals. For the Co/CeO2 sample, the peaks observed at 264.5, 312.2, and 472.2 °C are associated with the successive reduction of Co3+ to Co2+ and then to Co0. The peak at a higher temperature (472.2 °C) suggests the presence of more stable cobalt oxides or interactions between cobalt and ceria [37]. In the Fe/CeO2 sample, reduction peaks were identified at 286.6 and 349.9 °C, corresponding to the reduction of Fe3+ to Fe2+ and Fe2+ to Fe0, respectively [38]. Additional peaks at 427.3 and 595.8 °C are likely attributed to the formation of complex iron oxides and strong interactions with ceria [38]. The Zr/CeO2 sample exhibited a peak at 519.2 °C, which is indicative of ceria reduction facilitated by the presence of zirconium, highlighting significant metal-support interactions [39]. For Cu/CeO2, the reduction peak of surface oxygen typically observed around 500 °C is absent, which can be attributed to the strong metal-oxide supported interaction (SMSI) between CuO and CeO2 [36]. Below 300 °C, three distinct reduction peaks were observed at around 139.8, 161.0 and 209.3 °C indicating the presence of three distinct Cu species on the Cu/CeO2 catalyst surface. These species are identified, in decreasing order, as highly dispersed CuOX, the Cu–O–Ce structure, and the crystalline state of the CuO [40]. For the Fe/CeO2 catalyst, three additional reduction peaks at temperatures 286.6, 349.9, and 427.3 °C correspond to the sequential reduction of FeOX species. For Co/CeO2, two distinct peaks observed around 264.5 and 312.2 °C were associated with the reduction of CoOX.

Fig. 8
figure 8

H2-TPR profiles of M/CeO2 catalysts

Full size image

3.2.6 Raman Analysis

To study the structural information of M/CeO2 (M = Co, Cu, Fe, Zr) catalysts in more detail, Raman spectroscopy was used in this study for each catalyst sample. In general, it provides specific information about oxygen vacancies and oxygen sublattice distortions [41]. Figure 9 demonstrates the Raman spectra of each catalyst sample in this study. They all have a distinct Raman characteristic peak near 465 cm−1, which is attributed to the F2g peak of fluorite-structured CeO2 and is caused by the symmetric stretching of the Ce–O vibrational unit in the octet of coordination [42]. In pure CeO2, the F2g peak is usually a sharp peak located at 465 cm−1. As shown in Fig. 9, the F2g peaks of each M/CeO2 (M = Co, Cu, Fe, Zr) catalyst sample showed some degree of positional shift, which was attributed to the slight deformation of the CeO2 lattice induced by the participation of Mx+ (M = Co, Cu, Fe, Zr) ions [43]. In addition, the different degrees of broadening and enlargement of the F2g peaks can be assumed to be related to the increase in the concentration of oxygen vacancies since the F2g peaks are very sensitive to the disorder of the oxygen sublattice. Therefore, the changes in Raman F2g peak shape and peak position of the Raman F2g peaks of each catalyst sample compared to the Raman F2g peaks of pure CeO2 indicate the existence of strong interactions between each M metal and Ce. For the Co/CeO2 sample, additional Raman peaks at 520, 615, and 685 cm−1, respectively, are attributed to Co3O4 [44]. For the Cu/CeO2 sample, three additional Raman peaks at 290, 347, and 625 cm−1 were attributed to CuO. For the Fe/CeO2 sample, the Raman peak at 662 cm−1 is attributed to Fe2O3. For the Zr/CeO2 sample, the Raman peaks appearing at 310 and 638 cm−1 are attributed to ZrO2. The appearance of new Raman peaks in all the above catalyst samples suggests that some MOx material is still present on the surface of the prepared catalyst samples despite the incorporation of M-metals (M = Co, Cu, Fe, Zr) into the CeO2-based catalysts to form a Ce–O-M solid solution.

Fig. 9
figure 9

Raman profiles of M/CeO2 catalysts

Full size image

4 Conclusions

In this study, a series of CeO2-based catalysts denoted as M/CeO2 (M = Co, Cu, Fe, Zr) catalysts, were synthesized using the citric acid sol–gel method. These catalysts were evaluated for NH3-SCO performance on a simulated gas experimental platform. To elucidate the relationship between the physicochemical properties and catalytic performance, comprehensive characterization techniques including BET, XRD, and XPS were employed. The key findings are summarized as follows:

  1. 1.

    In terms of ammonia-oxidizing activity, the Co/CeO2 catalyst demonstrated superior performance, with a T90 as low as 239.2 °C. The catalysts were ranked based on their ammonia oxidation activities as follows Co/CeO2 > Cu/CeO2 > Fe/CeO2 > Zr/CeO2. Conversely, N2 selectivity followed an inverse trend: Co/CeO2 < Cu/CeO2 < Fe/CeO2 < Zr/CeO2, indicating a trade-off between ammonia oxidation activity and N2 selectivity. Considering both metrics, the Cu/CeO2 catalyst emerged as the best catalytic performance, which offers an optimal balance between high ammonia oxidation activity and favorable N2 selectivity.

  2. 2.

    Structural characterization via SEM, BET, and XRD confirmed that CeO2 was the predominant phase in each catalyst, exhibiting a cubic fluorite structure, consistent with the synthesis approach. Among the four catalysts with the same mesoporous structure, Co/CeO2 and Cu/CeO2 had the largest specific surface area (77.37 m2g−1) and the smallest pore volume (0.13 cm3g−1), respectively, which may be associated to their possessed better ammonia oxidation activity.

  3. 3.

    Further insights were gained from XPS, H2-TPR, and Raman spectroscopy. XPS analysis showed that the Oγ content in the surface adsorbed oxygen OA had a certain synergistic effect with the Ce3+/Ce4+ value, with high ratios correlating, with the improved ammonia oxidation activity. H2-TPR results showed that the incorporation of Co, Zr, Fe, and CeO2 shifted the surface oxygen reduction peak of pure CeO2 near 500 °C. The Co/CeO2 peak position shifted toward the low temperature indicating enhanced surface oxygen activity and improved reduction performance. The Cu/CeO2 catalyst displayed a complete disappearance of the reduction peak suggesting a strong interaction between Cu and Ce. Raman results confirmed the formation of Ce–O–M solid solution in all M/CeO2 catalysts, though some MOX material remained on the surface of the samples.