Mechanism of fast selective catalytic reduction of NO with NH₃ over MnO_X–CeO₂ catalysts

Abstract

MnO_X–CeO₂ composites are promising candidates as low-temperature active catalysts for selective catalytic reduction (SCR) of NO with NH₃, which is a leading technology for controlling NO emissions from non-electric flue gases. In this study, we systematically investigate the fast-SCR mechanism over MnO_X–CeO₂ through theoretical and experimental approaches. Our results reveal that fast-SCR is coupled with standard SCR through three coupled redox cycles: Mn-redox, Ce-redox, and O₂–O_v (surface oxygen vacancy in CeO₂) cycles occurring at distinct active sites. Even under O₂-rich reaction conditions, the fast-SCR reaction route still needs to overcome a higher energy barrier of 1.56 eV in the rate-determining step compared to the energy barrier of 1.44 eV via the standard SCR route. Intriguingly, fast-SCR significantly enhances the SO₂ resistance and N₂ selectivity by reducing the residence time of NH₃ adsorbed on the Mn³⁺ ions in the center of MnO_X clusters; this suppresses the reaction of NH₃ with SO_X and minimizes its deep oxidation, thereby suppressing N₂O emission.

---

1. Introduction

Low-temperature active selective catalytic reduction (LT-SCR) of NO by NH₃ is one of the most promising technologies to control the NO emissions from non-electric flue gases. MnO_X–CeO₂ composites are typically selected as some of the most important catalysts since the first report of their excellent low-temperature activity.¹ However, the LT-SCR mechanism is still not fully understood and is worth further research, especially the role of Fast-SCR reaction (F-SCR) which involves both NO and NO₂ is still controversial, though it is considered to play an important contribution under O₂ rich conditions.

Since the first discovery of the F-SCR reaction, which consumes equimolar NO and NO₂,² it is reported that the reaction rate of F-SCR could be 10 times higher than that of standard SCR (S-SCR) at temperatures below 200 °C.^3–5 In fact, the F-SCR reaction mechanism has been extensively studied. For example, some researches proposed that NO₂ formed nitrates and nitrites via a disproportionation reaction at first and the intermediate products reacted with NH₃ species to form NH₄NO₃ and NH₄NO₂; the reduction between NH₄NO₃ and NO is believed to be the rate-determining step.^6–8 The activation energy of F-SCR was reported to be higher than that of S-SCR due to the greater contribution by the reaction between NO and NH₄NO₃; although larger amounts of nitrites and NH₄NO₃ can result in a higher activity for F-SCR than for S-SCR,⁹ ammonia could block the reactivity of F-SCR.¹⁰ In another study, the crucial role of NO₂ in the F-SCR process was reported, where it participated in the hydrogen abstraction from coordinated NH₃ and promoted the production of –NH₂ active species that could directly react with NO to generate N₂ and H₂O.¹¹

An alternative mechanistic framework proposed catalyst involvement in redox cycles, particularly through reactions between NO₂ and the reduced catalyst state.^12–14 This scheme implied that all the F-SCR intermediates participated in the S-SCR cycles, with NO oxidation to bidentate nitrate by O₂ potentially serving as the S-SCR's rate-determining step.¹⁵ Experimental evidence confirmed that oxygen vacancies (O_v) enhanced the NO oxidation activity, thereby promoting F-SCR.¹⁶ Specifically, in MnO_X–CeO₂-based catalysts, Mn incorporation dramatically increases the formation of surface O_v, enabling efficient O₂ capture and generation of active oxygen species. These surface-active oxygen species effectively oxidize NO to NO₂, thereby facilitating F-SCR.¹⁷ Distinct catalytic sites for S-SCR and F-SCR have also been proposed.¹⁸ Interestingly, NO₂ may inhibit SCR reactions at low temperatures through NH₄NO₃ accumulation on active sites when its consumption rate by NO becomes rate-limiting under F-SCR conditions.¹⁹ Despite these advances, the detailed F-SCR mechanism remains unresolved.

Our previous work²⁰ systematically demonstrated that the MnO_X–CeO₂ synergy originates from the Mn²⁺ → Ce⁴⁺ electron transfer, which generates surface O_v on CeO₂. These O_v sites serve as the linchpin for establishing concurrent Mn-redox (NH₃ activation) and Ce-redox (O₂ activation) cycles, thereby synergistically promoting the S-SCR performance.

In the current study, we elucidate the MnO_X–CeO₂ synergy-driven F-SCR mechanism through combined density functional theory (DFT) calculations and experimental validation. Our results demonstrate that F-SCR gets initiated via NO₂ formation through reactions between NO and O_v-induced ·O radicals. However, F-SCR exhibits slower kinetics than that of S-SCR due to its higher activation energy barrier. Furthermore, while F-SCR couples with S-SCR, it only contributes moderately to the overall de-NO_X efficiency even under O₂-rich conditions over MnO_X–CeO₂. Significantly, F-SCR enhances both low-temperature SO₂ resistance^21,22 and N₂ selectivity by reducing the residence time of NH₃ adsorbed on the Mn³⁺ ions at the center of MnO_X nanoparticles.

2. Results and discussion

2.1. DFT calculations

Our previous DFT calculations revealed Mn₃O₄/CeO₂(111) as the most stable configuration for MnO_X clusters on ceria based on interfacial interactions.²⁰ The simulated F-SCR reaction pathway, as shown in Fig. 1, gets initiated with the appearance of surface oxygen vacancies (O_v, IM1), which are generated after the S-SCR cycle. When the surface O_v is filled with sufficient O₂via adsorption/activation (IM2, ΔE = −1.18 eV), an extra ·O radical is produced. The ·O radical then reacts with the CeO₂-adsorbed NO (IM3, ΔE = −1.6 eV)²⁰ to form an adsorbed NO₂via a low-energy barrier transition state (TS1, E_a = 0.27 eV). The adsorbed NO₂ subsequently migrates from CeO₂ to Mn₃O₄ (IM4) and chemisorbs on Mn²⁺ (IM5, ΔE = −1.59 eV), while NH₃ preferentially occupies the adjacent Mn³⁺ sites.²⁰ The NO₂–Mn²⁺ interaction (N⁽⁴⁺⁾O₂ + Mn²⁺ → ON⁽³⁺⁾O–Mn³⁺, E_a = 1.09 eV) facilitates NH₃ activation by lowering the N–H dissociation energy from 0.8 eV to 0.67 eV (TS2, IM6-7),²⁰ followed by H₂–N–N–O formation at Mn³⁺ (IM8) and subsequent second N–H cleavage (TS4, E_a = 1.56 eV, the rate-determining step), leading to HONNH rearrangement. Final N–H scission yields NNHOH (IM9–10) and ultimately forms N₂/H₂O (IM11-12). The residual –OH on Mn³⁺ reacts with adsorbed H (IM13) to form H₂O (IM14); the H₂O then diffuses from Mn³⁺ to the surface of CeO₂(IM15) and completes the reaction through an S-SCR cycle, which reduces Mn³⁺ to Mn^2+.20 The whole F-SCR route can be summarized as follows, which includes the essential roles of Mn²⁺/Mn³⁺ redox cycle, NH₃-activating Mn³⁺ sites, and oxygen vacancies:

·O + Ce⁴⁺–NO → Ce⁴⁺–NO₂ (R1)

Mn²⁺ + Ce⁴⁺–NO₂ → Ce⁴⁺ + Mn³⁺–ON⁽³⁺⁾O (R2)

Mn³⁺–ON⁽³⁺⁾O + NH₃–Mn³⁺ → ON–NH₂ + Mn³⁺–OH (R3)

ON–NH₂ → N₂ + H₂O (R4)

Mn³⁺–OH + Mn³⁺–OH → Mn³⁺–O + H₂O (R5)

Fig. 1 Transition states, optimal intermediate state geometry structures, and corresponding energy diagram obtained by DFT calculations in the NO reduction route over the MnO_X–CeO₂ catalyst.

The following reaction follows an S-SCR route:²⁰

Mn³⁺–O–HNH³⁺ + Mn²⁺–NH₂ + H₂O (R6)

Mn²⁺–NH₂ + Ce⁴⁺–ONO → Mn²⁺ + H₂O + N₂ (R7)

In summary, the reaction of Mn²⁺ + Ce⁴⁺ → Mn³⁺ + Ce³⁺ constitutes the key synergistic mechanism between CeO₂ and MnO_X, driving the formation of surface O_v near the MnO_X–CeO₂ interface/perimeter.²⁰ These O_vs readily activate the gaseous O₂, with the resulting ·O radicals reacting with adsorbed NO to generate NO₂, thereby initiating the F-SCR pathway. The critical F-SCR reaction involves Mn²⁺ + Ce⁴⁺–NO₂ → Ce⁴⁺ + Mn³⁺–ON⁽³⁺⁾O, which couples with the S-SCR reaction to establish a comprehensive SCR mechanism over the MnO_X–CeO₂ catalysts. This integrated reaction network encompasses four interconnected reaction cycles: (i) an Mn²⁺/Mn³⁺ redox cycle, (ii) a Ce³⁺/Ce⁴⁺ redox cycle, (iii) an O₂/O_v cycle, and (iv) an acid–base cycle involving NH₃ adsorption–desorption on the Mn³⁺ sites. Importantly, the rate-determining step of the F-SCR route exhibits a higher energy barrier (1.56 eV) than that of the S-SCR cycle (1.44 eV),²⁰ resulting in its limited contribution to the overall NO_X removal efficiency in low-temperature SCR systems, where the S-SCR route remains dominant.

2.2. Experimental verification

To experimentally investigate the F-SCR reaction route over the MnO_X–CeO₂ catalysts, we prepared a series of catalysts with different Ce⁴⁺ concentrations through precise annealing protocols. XRD analysis confirmed the phase purity of all the catalysts, with patterns perfectly matching those of the standard CeO₂ (PDF#78-0694), while the absence of the characteristic peaks of MnO_X indicated the high dispersion of Mn species on the CeO₂ support (Fig. 3a). XPS quantification analysis revealed a positive correlation between Ce⁴⁺ concentration and annealing temperature, with optimal conditions identified as the annealing of Ce-MOF precursors at 700 °C followed by MnO_X loading by annealing at 300 °C (Fig. 3b, S4 and S5). This thermal treatment protocol achieved Ce⁴⁺ concentrations up to 91.74%, establishing an ideal platform for probing the MnO_X–CeO₂ interactions.

The catalytic performance evaluation revealed a consistent positive correlation between the NO conversion efficiency and Ce⁴⁺ ion concentration across all the tested catalysts (Fig. 4a and b). This trend strongly corroborates with our previous findings regarding the enhanced synergistic effects between CeO₂ and MnO_X at elevated Ce⁴⁺ concentrations.²⁰ Particularly noteworthy is the M₃C₇ catalyst, which exhibited excellent performance with high NO conversion, exceeding 90% throughout a broad temperature range of 125–275 °C. Fig. 4c displays the temperature-dependent NO to NO₂ oxidation performance of the MnO_X–CeO₂ catalysts. The results reveal that NO oxidation efficiency depends critically on both Ce⁴⁺ concentration and reaction temperature. Below 250 °C, NO to NO₂ conversion remains minimal, reaching less than 5% at 125 °C, while total NO conversion attains 60% primarily through the standard SCR pathways; evidently, the contribution of F-SCR to the overall NO conversion is low. Above 250 °C, NO₂ production increases substantially with Ce⁴⁺ concentration in the catalysts. When the Ce⁴⁺ concentration increases from 90.43% to 91.22% and then to 91.74%, the NO to NO₂ oxidation yield reaches 40%, 44%, and 58%, respectively, at 300 °C. However, the corresponding total NO conversions of 57%, 75%, and 85% indicate that not all the generated NO₂ participates in the SCR reactions. These findings clearly demonstrate the limited contribution of fast-SCR to the overall NO_X removal under our experimental conditions.

2.3. Effects of F-SCR

The effect of F-SCR is evaluated by the correlation between NO₂ concentration and catalyst resistance to SO₂. As demonstrated in Fig. 5a, for the reaction over the M₃C₇ catalyst in the presence of 200 ppm SO₂, two distinct performance regimes emerge. Under low NO to NO₂ oxidation yield (3.8% NO₂ concentration measured in NO-TPO) at 150 °C and an O₂ concentration of 5%, the catalyst suffered from rapid SO₂ deactivation, with its NO_X conversion dropping from 82% to 41% in 30 hours and further declining to 25% after 80 hours of reaction. By contrast, when the NO₂ yield increased to 8.6% at the same temperature under an O₂ concentration of 15%, the catalyst maintained a stable performance with 63% NO conversion after 80 hours of reaction. More impressively, even under aggressive conditions of 200 °C and only 5% O₂ concentration (NO to NO₂ oxidation yield of 12%), the catalyst sustained a high NO_X conversion of more than80%. Extended 80-hour testing at 200 °C with an NO to NO₂ oxidation yield of 23% revealed no obvious SO₂-induced deactivation. Notably, under the conditions of high O₂ concentration, the gas phase reaction between NO and O₂ producing additional NO₂ is inevitable; fortunately, during the SCR reaction, the gas phase NO₂ is also preferentially adsorbed on the Mn²⁺ ions, and thus, its effects are included in Fig. 5.

The F-SCR route also significantly improves the N₂ selectivity, as evidenced in Fig. 5b. When the Ce⁴⁺ concentration increased by 1.31%, from 90.43% to 91.74% at 225 °C, N₂ selectivity rose correspondingly from 79% to 85%. This enhancement is particularly valuable because NH₃ deep-oxidation constitutes a major competing reaction for SCR in the MnO_X–CeO₂ systems, which consumes the valuable NH₃ and accelerates the N₂O emission at elevated temperatures.

2.4. Discussion

The F-SCR reaction is crucial for maintaining high activity in low-temperature SCR systems.^3–5 Nevertheless, the mechanistic details of the F-SCR pathway remain elusive.^6–19 In our investigation on the MnO_X–CeO₂ catalysts, we observed that NO₂ enhances the SCR performance through reaction (R2), and F-SCR is synergistically coupled with S-SCR via interconnected catalyst redox cycles (R1–R7).¹⁵ Notably, the rate-determining step in the F-SCR pathway exhibits a higher energy barrier (1.56 eV) than that of S-SCR (1.44 eV), explaining its relatively limited contribution to the overall NO reduction while confirming S-SCR as the predominant reaction pathway.⁹ These theoretical predictions align precisely with our experimental observations (Fig. 4).

The principal significance of the F-SCR pathway in the MnO_X–CeO₂ catalysts primarily relates to its substantial impact on the SO₂ resistance. As clearly demonstrated in Fig. 2, F-SCR and S-SCR reactions occur at distinct active sites. The S-SCR reactions predominantly take place at the CeO₂–MnO_X interface (marked by blue dotted circles in Fig. 2), where the Mn³⁺ ions facilitate NH₃ adsorption and activation, while Ce⁴⁺ sites activate the NO molecules.²⁰ The adsorbed NO can subsequently undergo oxidation to NO₂ by ·O radicals generated through the reaction between surface O_v and O₂. The resulting NO₂ molecules (including the gas phase NO₂) then migrate to the Mn²⁺ at the center of MnO_X clusters, where they react with neighbouring NH₃ to form H₂O and N₂. Consequently, F-SCR reactions primarily occur around the center of MnO_X clusters (indicated by orange dotted circles in Fig. 2), while S-SCR reactions occur around the CeO₂/MnO_X perimeter.

Fig. 2 Schematic of the synergy mechanism over the MnO_X–CeO₂ catalysts based on microstructure characterization (Fig. S1–S3). S-SCR occurred near the MnO_X–CeO₂ perimeter as the blue dotted cycle indicates. F-SCR occurred near the center of MnO_X cluster as the orange dotted cycle shows.

Fig. 3 Structural and compositional characterization of the MnO_X–CeO₂ catalysts: (a) XRD patterns of the M_XC_Y series catalysts (X and Y are the annealing temperatures in 100 °C units); (b) Ce⁴⁺ ion concentration in catalysts prepared under different annealing temperatures; the blue line shows the samples prepared by annealing Ce-MOF at 300–700 °C to obtain the CeO₂ support, followed by loading MnO_X by annealing at 300 °C, and the red line shows the samples prepared by annealing Ce-MOF at 700 °C to obtain the CeO₂ support, followed by loading MnO_X by annealing at 300–700 °C.

Fig. 4 Catalytic performance as a function of Ce⁴⁺ concentration in the MnO_X–CeO₂ catalysts: (a) NO conversion profiles for the M₃C_3–9 catalysts at Ce⁴⁺ concentrations of 88.87%, 89.65%, 90.26%, and 91.74%. (b) NO conversion profiles for the M_3–7C₇ catalysts at Ce⁴⁺ concentrations of 90.43%, 91.22%, and 91.74%. (c) Temperature-dependent NO oxidation to NO₂ for the M_3–7C₇ catalysts (Ce⁴⁺% = 90.43–91.74%).

Fig. 5 SO₂ resistance and N₂ selectivity characteristics:(a) SO₂ resistance of the M₃C₇ catalyst under different NO₂ concentrations at 150 °C and 200 °C in the presence of 200 ppm of SO₂; (b) temperature-dependent N₂ selectivity of the M₃C₇ (Ce⁴⁺% of 91.74%) versus M₇C₇ (Ce⁴⁺% of 90.43%) catalysts, demonstrating enhanced performance through the MnO_X–CeO₂ synergy in F-SCR.

Critically, NH₃ molecules adsorbed at the center of MnO_X clusters are spatially separated from NO adsorbed on the CeO₂ surfaces. Without the F-SCR pathway involving NO₂, these NH₃ molecules would be adsorbed on the surface Mn³⁺ for a long time, potentially forming NH₄HSO₄ deposits that could partially cover and poison the catalyst. The incorporation of these centrally adsorbed NH₃ molecules into the F-SCR pathway significantly reduces their surface residence time, thereby effectively suppressing the NH₄HSO₄ formation and mitigating the SO₂-induced deactivation. This interpretation is fully consistent with our experimental results (Fig. 5a) and is supported by previous studies.^22,27

At elevated temperatures exceeding 200 °C, the enhanced oxidative capacity of the catalyst promotes undesirable NH₃ deep oxidation, generating ·NH_X (X ≤ 1) radicals, particularly at the center of MnO_X clusters. This side reaction leads to N₂O formation and consequent deterioration of the N₂ selectivity.^28–30 Remarkably, the F-SCR mechanism serves to minimize the residence time of NH₃ adsorbed on the central Mn³⁺ ions, thereby substantially improving the overall N₂ selectivity.

3. Conclusions

Based on combined DFT simulations and experimental studies, we established a detailed reaction mechanism for LT-SCR over the MnO_X–CeO₂ catalysts. The MnO_X–CeO₂ synergy facilitates electron transfer from Mn²⁺ to Ce⁴⁺, thereby inducing surface O_v formation on CeO₂ and promoting dual NO-removal pathways through both the S-SCR and F-SCR routes. Nevertheless, the higher energy barrier of F-SCR (vs. S-SCR) maintains S-SCR as the dominant NO-removal pathway even under O₂-rich conditions. Notably, the coexistence of F-SCR substantially enhances the SO₂ resistance while suppressing the NH₃ deep oxidation side reactions, consequently improving the N₂ selectivity.

4. Experimental section

4.1. Preparation and characterization of the catalysts

Catalyst samples were prepared by an impregnation method using CeO₂ as the support. Typically, 0.13 g of Mn(CH₃COO)₂·4H₂O powder was dissolved in 5 mL of C₂H₆O/C₂H₄O₂ solution. Then, CeO₂ (1.30 g) powder was added into the mixture. After being fully mixed, the grey slurry was dried at 80 °C for 12 h and then calcined at 300–700 °C for 6 h at a heating rate of 5 °C min⁻¹. Finally, dark gray MnO_X–CeO₂ catalysts were obtained.

The CeO₂ supports were obtained by annealing the Ce-MOFs precursor at 300–900 °C for 2 h at a heating rate of 5 °C min⁻¹.²³ The catalysts were named as M_XC_Y, where M and C are MnO_X and CeO₂, and digital numbers X and Y are annealing temperature (100 °C unit), respectively.

The catalysts were characterized by X-ray diffraction using Cu Kα radiation (XRD, Philips, XD-98), Cs-corrected transmission electron microscopy (TEM, G2, Titan, FEI), X-ray photoelectron spectrometry (XPS, Escalab 250Xi, Thermo Fisher) using Al Kα radiation, temperature programmed desorption of ammonia (NH₃-TPD) and temperature programmed reduction with hydrogen (H₂-TPR) (AutoChem II 2920 chemisorption analyzer), and N₂ adsorption–desorption isotherms (ASAP2010C, Macmillan). For detailed information, please refer to ref. 20.

The NH₃-SCR and NO oxidation performances were measured using a plate catalyst test system similar to the industrial one. Briefly, 0.5 g of catalyst powder (40–60 mesh) was uniformly spread onto five aluminium plates (4 cm × 10 cm), and the space between the plates was 5.0 mm. The total flue gas flow for the reaction was 600 mL min⁻¹, containing 500 ppm NO, 500 ppm NH₃, 5% O₂, N₂ as the balance gas, and the weight hourly space velocity (WHSV) was 72 000 mL g⁻¹ h⁻¹. The reaction temperature was raised from 100 to 300 °C. At each target temperature, the system was stabilized for 10 minutes before data recording. The inlet and outlet concentrations of NO, NO₂, and N₂O were measured by Testo-340 and Medi-Gas G200 gas analyzers, respectively. The NO_X conversion was calculated as:²⁰ NO_X conversion = 100% × ([NO_X]_in − [NO_X]_out)/[NO_X]_in.

4.2. DFT calculations

Density functional theory (DFT) calculations were carried out using the Vienna ab initio simulation package (VASP) code. The generalized gradient approximation (GGA) of the Perdew–Burke–Ernzerhof (PBE) functional and the projector augmented-wave (PAW) potential were employed.^24–26 Detailed information is reported in ref. 20.

Author contributions

Hongmei Zheng: investigation, methodology, validation, writing – review & editing. Zhihao Zhao: investigation, methodology, validation, writing – review. Kai Zhang: software, resources. Fei Wang: investigation, methodology. Songda Li: software, resources. Zhongkang Han: software, resources funding acquisition, supervision. Yong Wang: conceptualization, formal analysis, project administration, funding acquisition, supervision. Ze Zhang: conceptualization, project administration, funding acquisition, supervision. Hangsheng Yang: conceptualization, formal analysis, funding acquisition, supervision.

Conflicts of interest

There are no conflicts to declare.

Data availability

The data supporting this article have been included as part of the supplementary information (SI).

Supplementary information is available. See DOI: https://doi.org/10.1039/d5cy01103e.

Acknowledgements

This work was supported by the Key Research and Development Program of Zhejiang Province (2021C01003), the National Natural Science Foundation of China (51872260, 52025011, 51971202 and 52171019), the National Key Research and Development Program of China (2017YFB0310400, 2022YFA1505500), the Zhejiang Provincial Natural Science Foundation of China (LD19B030001, Z4080070 and LR23B030004), and the Fundamental Research Funds for the Central Universities.

Notes and references

1. G. Qi, R. Yang and R. Chang, Appl. Catal., B, 2004, 51, 93–106.
2. A. Kato, S. Matsuda, T. Kamo, F. Nakajima, H. Kuroda and T. Narita, J. Phys. Chem. C, 1981, 85, 4099–4102.
3. H. Curry-Hyde, H. Musch and A. Baiker, Appl. Catal., 1990, 65, 211–223.
4. C. Cristian, N. Isabella, T. Enrico, C. Daniel, B. Brigitte, W. Michel and K. Bernd, Appl. Catal., B, 2007, 70, 80–90.
5. M. Koebel, M. Elsener and G. Madia, Ind. Eng. Chem. Res., 2001, 40, 52–59.
6. C. Ciardelli, E. Tronconi, D. Chatterjee and B. Bandl-Konrad, Chem. Commun., 2004, 2718–2719.
7. E. Tronconi, I. Nova, C. Ciardelli, D. Chatterjee and M. Weibel, J. Catal., 2007, 245, 1–10.
8. I. Nova, C. Ciardelli, E. Tronconi, D. Chatterjee and B. Bandl-Konrad, Catal. Today, 2006, 114, 3–12.
9. K. Liu, H. He and B. Chu, Chem. Eng. J., 2021, 423, 130128.
10. A. Grossale, I. Nova and E. Tronconi, J. Catal., 2009, 265, 141–147.
11. Z. Liu, C. Chen, J. Zhao, L. Yang, K. Sun, L. Zeng, Y. Pan, Y. Liu and C. Liu, Chem. Eng. J., 2020, 379, 122288.
12. L. Arnarson, H. Falsig, S. Rasmussen, J. Lauritsen and P. Moses, J. Catal., 2017, 346, 188–197.
13. M. P. Ruggeri, A. Grossale, I. Nova, E. Tronconi, H. Jirglova and Z. Sobalik, Catal. Today, 2012, 184, 107–114.
14. M. Koebel, G. Madia, F. Raimondi and A. Wokaun, J. Catal., 2002, 209, 159–165.
15. T. V. W. Janssens, H. Falsig, L. F. Lundegaard, P. N. R. Vennestrøm, S. B. Rasmussen, P. G. Moses, F. Giordanino, E. Borfecchia, K. A. Lomachenko, C. Lamberti, S. Bordiga, A. Godiksen, S. Mossin and P. Beato, ACS Catal., 2015, 5, 2832–2845.
16. Y. Zhou, S. Ren, J. Yang, W. Liu, Z. Su, Z. Chen, M. Wang and L. Chen, J. Environ. Chem. Eng., 2021, 9, 106218.
17. M. Li, M. Gao, G. He, Y. Yu and H. He, Environ. Sci. Technol., 2023, 57, 3875–3882.
18. M. Schwidder, S. Heikens, A. Toni, S. Geisler, M. Berndt, A. Brückner and W. Grünert, J. Catal., 2008, 259, 96–103.
19. L. Xie, F. Liu, K. Liu, X. Shi and H. He, Catal. Sci. Technol., 2014, 4, 1104–1110.
20. F. Wang, S. Li, R. You, Z. Han, W. Yuan, B. Zhu, Y. Gao, H. Yang and Y. Wang, Appl. Surf. Sci., 2023, 638, 158124.
21. X. Wang, X. Du, L. Zhang, Y. Chen, G. Yang and J. Ran, Appl. Catal., A, 2018, 559, 112–121.
22. B. Li, Z. Ren, Z. Ma, X. Huang, F. Liu, X. Zhang and H. Yang, Catal. Sci. Technol., 2016, 6, 1719–1725.
23. F. Wang, K. Deng, G. Wu, H. Liao, H. Liao, L. Zhang, S. Lan, J. Zhang, X. Song and L. Wen, J. Inorg. Organomet. Polym., 2012, 22, 680–685.
24. G. Henkelman, B. Uberuaga and H. Jónsson, J. Chem. Phys., 2000, 113, 9901–9904.
25. Z. Han, Y. Yang, B. Zhu, M. Ganduglia-Pirovano and Y. Gao, Phys. Rev. Mater., 2018, 2, 035802.
26. B. Tian, S. Ma, J. Guo, Y. Zhao, T. Gao and X. Jiang, Appl. Surf. Sci., 2022, 601, 154162.
27. Z. Ma, L. Sheng, X. Wang, W. Yuan, S. Chen, W. Xue, G. Han, Z. Zhang, H. Yang, Y. Lu and Y. Wang, Adv. Mater., 2019, 31, e1903719.
28. L. Sheng, Z. Ma, S. Chen, J. Lou, C. Li, S. Li, Z. Zhang, Y. Wang and H. Yang, Chin. J. Catal., 2019, 40, 1070–1077.
29. L. Sheng, S. Li, Z. Ma, F. Wang, H. He, Y. Gao, B. Zhu, Y. Wang and H. Yang, Catal. Sci. Technol., 2021, 11, 1709–1716.
30. X. Fan, F. Qiu, H. Yang, W. Tian, T. Hou and X. Zhang, Catal. Commun., 2011, 12, 1298–1301.

---

Footnote

† First authors.

---