Hydroxyl‑Mediated SO2 Promotion Enables Efficient NOx Reduction by CO over IrIn/Beta under Oxygen‑Rich Conditions Authors: Yujie Yuan, Yixi Wang, Wenqing Xu, Yanhong Wang, Yang Yang, Xiubiao Ma, Tingyu Zhu Published: 2026 ABSTRACT AbstractDeveloping high-performance Ir-based catalysts for the selective catalytic reduction of NOx with CO (CO-SCR) under oxygen-rich conditions remains a formidable challenge. Here, we identify a hydroxyl-mediated reaction pathway over IrIn/Beta catalysts, termed the ammonium sulfate mechanism. Tailoring the support’s Si/Al ratio modulates surface hydroxyls (Si–OH–Al), governing the rapid transformation of ammonium sulfate intermediates. Density functional theory calculations confirm that the formation of these intermediates is thermodynamically spontaneous and highly exothermic. These in situ-generated species rapidly decompose at 167 °C to release NH3, triggering a built-in NH3-SCR cycle. Notably, the IrIn/Beta-30 catalyst achieves 88% NOx conversion under harsh simulated industrial conditions (15% O2 and 200 ppm SO2). This work establishes SO2 as a contributing factor, providing a rational design strategy for robust denitration catalysts applicable to real-world industrial flue gas.Subject terms: Pollution remediation, Environmental impact, Heterogeneous catalysis Introduction Nitrogen oxides (NOx, mainly NO and NO2), carbon monoxide (CO), and other pollutants rich in exhaust gas from stationary sources are air pollutants that pose multiple hazards to human health, ecological environment and natural climate1–3. Selective catalytic reduction of nitrogen oxides with CO (CO-SCR) utilizes the carbon monoxide inherently present in flue gas as a reducing agent to convert NOx into harmless N2, thereby achieving the goal of treating waste with waste4–6. Compared to the commercially established NH3-SCR technology, CO-SCR offers a more sustainable and environmentally friendly alternative; it eliminates the need for expensive ammonia storage and injection systems, reduces system complexity, and avoids ammonia slip and associated secondary pollution7. However, practical industrial flue gas typically contains high concentrations of O2 (14%–16%), which subjects the CO-SCR system to complex competitive reaction mechanisms, leading to the rapid deactivation of most catalysts due to oxygen inhibition or side reactions8,9. Consequently, the development of high-performance CO-SCR catalysts capable of adapting to high-oxygen atmospheres remains a critical challenge for large-scale application. Among various active components, Ir-based catalysts have been extensively investigated for their superior antioxidant properties compared to other metals10,11. However, typically reported Ir-based catalysts perform optimally only under low-to-medium oxygen concentrations, and the prohibitive cost of precious metal Ir severely restricts its large-scale industrial application. In recent years, constructing bimetallic catalysts through the introduction of promoters has become a research focus to reduce costs and enhance performance12–14. Indium (In) exhibits distinct advantages in modulating the electronic structures of noble metals (e.g., Ir, Pt, and Pd) due to its unique electronic configuration. This exceptional electronic modification capability has been widely validated across multiple fields. In optoelectronics, In is the core component of Indium Tin Oxide (ITO), where its precise regulation of carrier density underscores its superior electron-transfer characteristics15,16. In the realm of heterogeneous catalysis, In frequently serves as an electronic promoter to form intermetallic compounds or solid solutions with active metals, significantly improving selectivity and stability in reactions such as propane dehydrogenation (PDH) and CO2 hydrogenation17–19. Furthermore, in electrochemical systems like the hydrogen evolution reaction (HER), the incorporation of In effectively optimizes the d-band center of active sites, thereby lowering reaction energy barriers20. Meanwhile, Indium offers industrial advantages such as low cost, environmental friendliness, and non-toxicity. Given its proven performance in modulating electronic states and enhancing interfacial synergy, applying Indium to the CO-SCR system to optimize NOx activation pathways at Ir sites holds great scientific value and promising application prospects. Most current studies indicate that the key to the oxygen tolerance of Ir-based catalysts lies in the promoting effect under the coexistence of SO221. For instance, Bai et al. discovered that SO2 can inhibit the direct oxidation of CO and O2, providing more CO for NO reduction at high temperatures22. Wang et al. believe that the sulfate formed on the surface of the catalyst is a key species promoting the reduction of NO and inhibiting the adsorption of O2 on the catalyst surface at low temperatures. The main generation pathways of sulfate species on Ir-based catalysts were revealed, and the reasons for the increased activity in the presence of SO2 were clarified from the perspective of reaction pathways23. Chen et al.'s research found that under oxygen-containing conditions, Ir0 was partially oxidized to Irδ+, and the presence of SO2 could promote the reduction of Irδ+ to Ir0, thereby maintaining the high denitrification performance of Ir@ZSM-524. However, the promoting effect of SO2 on the reaction is uncontrollable, and it remains unclear how to specifically regulate and enhance the promoting effect of SO2. In this study, a series of IrIn bimetallic catalysts supported on Beta zeolites with different Si/Al ratios were prepared by impregnation, and the key intermediates of SO2 in the reaction pathway were investigated. The research found that the CO-SCR reaction on IrIn/Beta follows a new reaction pathway mediated by hydroxyl groups, which we term the "ammonium sulfate pathway". By tailoring the Si/Al ratio of Beta zeolite support to directionally modulate the surface hydroxyl group ratio of the catalyst, efficient activation of NO and SO2 as well as rapid conversion of key ammonium sulfate intermediates are achieved. Among them, the IrIn/Beta catalyst with a specific Si/Al ratio of 30 can achieve a NOx conversion rate of 88% under simulated actual industrial flue gas conditions (NO 400 ppm, CO 8000 ppm, O2 15%, SO2 200 ppm). The catalytic performance is significantly superior to that of most Ir-based catalysts under the same conditions. This study significantly enhanced the feasibility of the practical application of CO-SCR catalysts. Results Performance of IrIn/Beta catalysts in CO-SCR NOx conversion and CO conversion of IrIn/Beta catalysts with different Si/Al ratios were first measured to estimate their catalytic performance in CO-SCR. Fig. 1a shows that IrIn/Beta with a specific Si/Al of 30 exhibited a lower activation temperature compared to the IrIn/Beta-5 and IrIn/Beta-80 catalysts in the simulated actual industrial flue gas conditions (NO 400 ppm, CO 8000 ppm, O2 15%, SO2 200 ppm). As previously reported, NOx conversion follows a volcanic curve, with the peak of IrIn/Beta-30 occurring at 275 °C, and NOx conversion of approximately 88%8,10,12,24. Then, side reaction of CO oxidation gradually becomes dominant as the reaction temperature increases, and when the CO conversion reaches 100% (Fig. 1b), the NOx conversion decreases accordingly. Evaluation of the benchmark Ir/Beta-30 was performed to clarify the promotional effect of In (Fig. S1). The activity results reveal that IrIn/Beta-30 outperforms Ir/Beta-30 across the entire temperature range, regardless of the presence of SO2. This superior performance effectively confirms the critical role of In in enhancing the low-temperature DeNOx activity. Catalytic performance of catalysts in the CO-SCR reaction. a NOx conversion; b CO conversion of IrIn/Beta-5, IrIn/Beta-30, and IrIn/Beta-80 catalysts. c NOx conversion of catalysts with different Si/Al ratios at 275 °C in the absence and presence of 200 ppm SO2. d Rm (reaction rates normalized to the mass of catalysts) on IrIn/Beta-5, IrIn/Beta-30, and IrIn/Beta-80 catalysts in CO-SCR reaction. e Apparent activation energy (Ea) for the CO-SCR reaction over IrIn/Beta-5, IrIn/Beta-30, and IrIn/Beta-80 catalysts. f NOx conversion over different bimetallic catalyst systems. g 24-hour stability test results of IrIn/Beta-30 at 275 °C. h Comparison of activity among different catalysts. (The gas composition was 8000 ppm CO, 400 ppm NO, 200 ppm SO2, and 15% O2 in N2; GHSV = 24 000 h-1). To further determine the optimal Si/Al ratio for the IrIn/Beta catalyst, the CO-SCR catalytic performance of IrIn/Beta with Si/Al ratios of 20 and 40 was investigated under simulated industrial flue gas conditions and compared with IrIn/Beta-30 (Fig. S2). The NOx conversion results, shown in Fig. S2a, indicate that all three catalysts exhibit a typical volcano-shaped distribution of NOx conversion as a function of temperature. IrIn/Beta-30 maintained the highest overall NOx conversion. In contrast, when the Si/Al ratio was decreased to 20 (IrIn/Beta-20), the peak conversion temperature showed no significant shift, but the maximum NOx conversion dropped substantially to approximately 65%. Conversely, increasing the Si/Al ratio to 40 (IrIn/Beta-40) resulted in a peak conversion of only about 45%, accompanied by a slight shift of the active temperature window toward higher temperatures. Correlating these findings with CO conversion (Fig. S2b), IrIn/Beta-30 achieved 65% CO conversion at approximately 275 °C, indicating an abundant supply of the reducing agent (CO) for the selective reduction of NOx, which underpins the high NOx conversion efficiency of the catalyst. For IrIn/Beta-20, CO was completely converted at 275 °C; such excessive CO oxidation activity likely leads to the premature consumption of the reductant, thereby inhibiting the efficient reduction of NOx. Regarding IrIn/Beta-40, the concurrently low CO and NOx conversions suggest a deficiency in the density or activation capacity of intrinsic active sites at this Si/Al ratio. These results further confirm that a Si/Al ratio of 30 is optimal for supporting IrIn bimetallic active components in this research system. To elucidate the influence of SO2 on catalytic activity, the NOx conversion of catalysts with varying Si/Al ratios was evaluated at 275 °C, comparing performance in the absence and presence of SO2 (Fig. 1c). In the SO2-free atmosphere, all catalysts exhibited negligible NOx conversion, consistently remaining below 5%. However, upon introducing 200 ppm SO2 into the reactant feed, a dramatic enhancement in catalytic activity was observed across all samples. Notably, the NOx conversion over the IrIn/Beta-30 catalyst surged from a mere 1.6% to 88.0%, demonstrating the most pronounced SO2-promoted effect. Furthermore, this SO2-induced enhancement is highly sensitive to the Si/Al ratio of the zeolite support, and the catalytic performance varies significantly with the framework composition of the Beta carrier. Kinetic studies were performed (as shown in Fig. 1d) to further evaluate the impact of the Si/Al ratio on CO-SCR activity. The reaction rates normalized to the catalyst mass (Rm) over the IrIn/Beta-30 catalyst at temperatures ranging from 205–250 °C are approximately 1.5–17 times those over IrIn/Beta-5 and IrIn/Beta-80, confirming that the reaction kinetics over IrIn/Beta-30 are inherently superior to those over IrIn/Beta-5 and IrIn/Beta-80 (Table S1)25. Additionally, the apparent activation energy (Ea) of IrIn/Beta-30 (116.0 kJ·mol-1) is lower than that of IrIn/Beta-5 (181.0 kJ·mol-1) and IrIn/Beta-80 (153.1 kJ·mol-1), indicating that the IrIn/Beta-30 catalyst has the lowest CO-SCR reaction energy barrier, which is advantageous for reducing NOx at low temperatures (Fig. 1e). To further investigate the unique role of In as a promoter and the regulatory patterns of the SO2-facilitated mechanism across different bimetallic systems, analog catalysts were synthesized by introducing Ga (from the same group as In) and Sn (from the adjacent group)26,27. The CO-SCR activities of IrGa/Beta-30 and IrSn/Beta-30 were systematically evaluated for comparison. The experimental results (Fig. 1f) demonstrate that the incorporation of main-group metal promoters into the Ir system possesses a significant and universal enhancing effect, as evidenced by the varying degrees of catalytic potential stimulated by the addition of Ga and Sn. However, within the current Beta zeolite framework, the Ir–In combination exhibits the most precise electronic-transfer matching and synergistic spatial effects. This unique “tripartite” coupling constructs a more efficient reaction pathway, enabling IrIn/Beta-30 to achieve optimal catalytic performance under harsh, hyperoxic conditions. Furthermore, these findings provide a vital scientific basis for future efforts to unlock the catalytic potential of Ga, Sn, and similar systems by fine-tuning metal loading ratios or strategically modulating support acidity. From a practical application perspective, the long-term durability of the catalyst is a critical parameter for evaluating its industrial potential. The high NOx conversion of IrIn/Beta-30 at 275 °C remained nearly unchanged in the 24-hour stability test experiment (Fig. 1g), indicating the dynamic stability of the catalyst during the CO-SCR reaction process. Furthermore, to evaluate the stability of the IrIn/Beta-30 catalyst under more realistic application conditions, a dynamic H2O switching test was conducted (Fig. S3). The results demonstrate that the catalyst exhibits robust performance in the presence of water vapor. Upon the introduction of 5% external H2O at 6 h, an evident transient promotion in NOx conversion was observed. Conversely, when the external water supply was abruptly removed at 18 h, the NOx conversion exhibited a sharp, transient decline. Remarkably, as the reaction proceeded under anhydrous conditions, the catalytic activity gradually recovered, ultimately returning to its initial high baseline. This capacity for activity recovery under dynamic moisture-switching conditions provides compelling evidence for the overall robustness of the catalyst. Beyond the long-term stability and moisture resistance, it is well established that the presence of O2 poses a severe challenge to the CO-SCR performance of catalysts. Compared with catalysts reported in previous literature, the IrIn/Beta-30 catalyst developed in this work demonstrates superior activity in a complex atmosphere with low metal loading (Fig. 1h and Table S2). Notably, to the best of our knowledge, reports of high-performance catalysts at O2 concentrations (14%–16%) close to actual industrial flue gas are very rare, typically below 7%5,6,14. Physical phase and morphological characteristics of catalysts To elucidate the influence of crystal structure on catalytic performance, we conducted an in-depth comparative analysis of the catalyst’s microstructure. Fig. 2a displays the X-ray diffraction (XRD) patterns of the synthesized catalysts. The characteristic diffraction peaks observed at 2θ = 7.7°, 14.9°, 22.8° and 27.3° are indexed to the Beta zeolite framework (PDF#48-0038), confirming that the crystalline integrity of the support remained intact following the loading of active components28–30. For all IrIn/Beta samples, distinct reflections at 40.7° and 47.3° are identified, corresponding to the (111) and (200) planes of metallic Ir (PDF#06-0598), respectively. This indicates a relatively high degree of crystallinity for the Ir species31,32. Notably, no characteristic diffraction peaks corresponding to metallic In or its oxide phases were detected in the XRD patterns. This not only suggests an exceptional dispersion of In species on the support surface but also implies that, driven by strong intermetallic interactions, In is highly likely to be integrated into the Ir crystalline structure, leading to the formation of structurally homogeneous bimetallic clusters18,33. Inductively Coupled Plasma Optical Emission Spectroscopy (ICP-OES) analysis (Table S3) revealed that among all the catalysts, the loading amount of Ir is approximately 0.3 wt.% and that of In is approximately 0.5 wt.%, which matches well with the molar ratios (the molar ratio of Ir to In was 1:3) used in the synthesis. The specific surface area and pore structure of the catalyst were investigated by the BET method. The N2 adsorption-desorption isotherms all exhibited type I isotherm features, indicating that the catalysts are dominated by microporous structures. After loading metals, the specific surface area and pore volume of each carrier decreased to varying degrees, further confirming that some channels were occupied by IrIn, which was consistent with the XRD results. As illustrated by the pore size distribution curve (Fig. 2b), the adsorption capacity follows the order of IrIn/Beta-80 > IrIn/Beta-30 > IrIn/Beta-5, which suggests that the specific surface area and total pore volume of the catalysts decrease gradually with the reduction of the Si/Al ratio of the support. The pore size distribution illustration of the catalyst indicates that the main micropores of the three are concentrated around 0.4 nm, enabling the reactants and products to achieve efficient diffusion and mass transfer within the rich pore structure of the catalyst34,35. Morphological characterizations of catalysts. a XRD patterns. b N2 adsorption-desorption isotherms (inset: pore size distribution of the catalysts). c–e, g–i, k–m HAADF-STEM images of IrIn/Beta-5 (c–e), IrIn/Beta-30 (g–i), and IrIn/Beta-80 (k–m) at different magnifications. f, j, n Corresponding particle size distribution histograms of the IrIn bimetallic nanoparticles for the respective catalysts. The OH groups on the surface of IrIn/Beta catalysts usually act as acidic sites, and NH3 acts as an alkaline probe molecule, which can selectively adsorb onto these acidic sites36. Therefore, the content and type of OH groups in catalysts with different Si/Al ratios were determined by in-situ DRIFTS via saturated NH3 adsorption at room temperature. As shown in Fig. S4, the negative peaks of the sample at 3630 cm-1 and 3460 cm-1 correspond to terminal Si–OH hydroxyl groups and double-bridged Si–OH–Al hydroxyl groups, respectively. Comparison of the normalized areas of different hydroxyl groups (Fig. S5) reveals that the double-bridged Si–OH–Al hydroxyl groups are dominant in all catalysts. With decreasing Beta zeolite Si/Al ratio, the difference in content between double-bridged Si–OH–Al hydroxyl groups and terminal Si–OH hydroxyl groups gradually increases. The experimental results show significant differences in the proportions of surface hydroxyl groups on IrIn/Beta catalysts with different Si/Al ratios. Numerous studies have demonstrated that the abundant surface hydroxyl groups on supports serve as effective anchoring sites for metallic species36,37. To investigate the influence of Beta zeolites with varying Si/Al ratios on the dispersion and particle size distribution of IrIn bimetallic species, aberration-corrected transmission electron microscopy (AC-TEM) was employed. As illustrated by the TEM images and corresponding particle size histograms in Fig. 2c–n, uniformly distributed metallic nanoparticles were observed across all IrIn/Beta catalysts; however, their average particle sizes exhibited significant variations. Specifically, the IrIn/Beta-5 catalyst displayed the smallest and most uniform particles, with an average size of approximately 1.3 nm. In contrast, the average particle size increased to 2.2 nm for IrIn/Beta-30 and further expanded to 3.7 nm for IrIn/Beta-80. These results indicate a clear trend: the particle size of IrIn metallic species decreases as the Si/Al ratio of the Beta support decreases (i.e., as the framework Al content increases). Combined with the NH3 DRIFTS characterization (Figs. S6, 7), it is evident that these dual-bridged hydroxyl groups (Si–OH–Al) act as the preferred anchoring sites for the active metallic species. A higher density of anchoring sites facilitates the superior dispersion of metallic species and effectively suppresses their migration and agglomeration during calcination and reduction processes. Consequently, the decrease in the Si/Al ratio leads to a higher degree of dispersion for the active components, ensuring the formation of smaller, highly active bimetallic clusters. To provide a more detailed characterization of the microscopic distribution of active species across catalysts with varying Si/Al ratios, high-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM) and corresponding energy-dispersive X-ray spectroscopy (EDS) elemental mapping were conducted (Fig. S8). The results reveal an exceptionally high degree of spatial correlation between Ir (red) and In (green) elements across all three catalysts. As demonstrated in the elemental mapping images (Fig. S8d, h, l), Ir and In are highly co-localized at the nanoparticle scale, with no discernible phase segregation or individual component enrichment. Further analysis using EDS line scans (Fig. S9) confirms that the signal intensity profiles of the Ir M-shell and In L-shell exhibit highly synchronized fluctuations, with their peak intensities precisely centered within the geometric core of the nanoparticles. To further investigate the bimetallic interactions at the atomic scale, high resolution characterization was performed on the representative IrIn/Beta 30 sample. As illustrated in Fig. S10a, b, the catalyst exhibits well resolved atomic lattices, indicating the formation of well crystallized nanostructures on the support. The corresponding atomic-level EDS mapping (Fig. S10c, d) reveals a high degree of spatial co-location of Ir and In signals across the atomic sites. This exceptional uniformity, spanning from the nanoparticle to the atomic scale, provides compelling evidence for the robust interaction between Ir and In. Concurrently, the close lattice spacing between the In (002) and Ir (111) planes facilitates the formation of stable Ir–In bonds, which maintains the structural integrity of the bimetallic entities and suppresses lattice distortion. Collectively, these findings confirm the successful construction of structurally stable bimetallic active centers on the Beta zeolite surface. Electronic structure and coordination environment of IrIn Using CO as a probe molecule, diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS) was employed to characterize the electronic structure changes of Ir before and after the formation of the IrIn intermetallic compound catalyst. As illustrated in Fig. 3a, the characteristic peak at 2063 cm-1 for Ir/Beta-30 is assigned to the stretching vibration of CO molecules linearly adsorbed on Ir sites. Upon the introduction of In, this adsorption band for the IrIn/Beta-30 system exhibits a significant red-shift to 2055 cm-1. This shift is attributed to the σ-donation from In to the Ir sites, which enhances the π-back-donation from Ir d-orbitals into the anti-bonding 2π* orbitals of CO18,38,39. This electronic modification weakens the C–O bond, thereby lowering its vibrational frequency. Characterizations of Ir and In interactions. a CO DRIFTS results of Ir/Beta-30 and IrIn/Beta-30. Normalized Ir L3-edge XANES b and EXAFS c spectra of Ir/Beta-30 and IrIn/Beta-30. d Normalized In K-edge EXAFS spectra of IrIn/Beta-30. The XPS Ir 4f (e) and In 3d (f) results of IrIn/Beta-5, IrIn/Beta-30, and IrIn/Beta-80 after H2 pretreatment. Ir L3-edge WT-EXAFS contour plots of IrO2 (g), Ir foil (h), Ir/Beta-30 (i)and IrIn/Beta-30 (j), respectively. To gain deeper insights into the electronic states and local coordination environments of Ir species in Ir/Beta-30 and IrIn/Beta-30 catalysts, synchrotron X-ray absorption spectroscopy (XAS) measurements were conducted (Figs. S11–13). Fig. 3b presents the normalized Ir L3-edge XANES spectra. Given that the white line intensity at the Ir L3-edge reflects the dipole-allowed transition from 2p3/2 to unoccupied 5d5/2 orbitals, the results show that the white line intensities for all samples lie between those of Ir foil and IrO2, while remaining closer to the Ir foil, confirming that Ir predominantly exists in a metallic state18,20. Notably, the white line intensity of IrIn/Beta-30 is significantly lower than that of Ir/Beta-30 (Fig. 3b inset), indicating that the introduction of In increases the electron cloud density of the Ir orbitals. The EXAFS fitting results (Fig. 3c and Table S4) further reveal the evolution of the coordination environment. Compared to Ir/Beta-30, the coordination number (CN) of the Ir–O path in IrIn/Beta-30 decreases from 0.8 to 0.3, suggesting that the presence of In facilitates the reduction of Ir species. Furthermore, a well-defined Ir–In coordination contribution (CN = 1.2) appears at 2.817 Å in IrIn/Beta-30, directly confirming the formation of Ir–In metallic bonds. Simultaneously, the In K-edge EXAFS fitting (Fig. 3d and Fig. S14) yields an In–O CN of 5.9, indicating that In species primarily exist in a stable, highly coordinated oxidized state (In3+)17,40. The wavelet transform (WT-EXAFS) analysis (Fig. 3g–j) clearly illustrates this structural evolution: while Ir/Beta-30 exhibits only a single Ir–Ir signal at approximately (12.5 Å⁻¹, 2.7 Å), the WT maximum for IrIn/Beta-30 shifts toward higher k-space, displaying characteristic overlapping signals from both Ir–Ir and Ir–In paths. These collective results substantiate that a stable Ir–In active center is formed through strong intermetallic interactions between Ir and In31,41. X-ray photoelectron spectroscopy (XPS) was employed to characterize the chemical states of Ir and In in Ir/Beta-30 and IrIn/Beta catalysts with varying Si/Al ratios. As illustrated in Fig. 3e and Fig. S15, the Ir 4f spectra exhibit two peaks at 61.0 eV and 64.0 eV, which are assigned to metallic Ir0 species, while the peaks at 62.4 eV and 64.9 eV correspond to the characteristic signals of oxidized Irδ⁺ species24,42. Quantitative analysis (Table S5) reveals that the proportion of metallic Ir0 in the bimetallic IrIn/Beta-30 is significantly higher than that in the monometallic Ir/Beta-30 baseline. This result clearly indicates that the incorporation of In facilitates the reduction of Ir active components and effectively maintains the reduced state distribution of Ir sites on the surface through the electron transfer effect from In to Ir. This observation is highly consistent with the aforementioned CO-DRIFTS and XAS findings. Meanwhile, the content of Ir0 species increases progressively as the Si/Al ratio of the IrIn/Beta support decreases. This trend further confirms the pivotal regulatory role of the Al site density within the zeolite framework on the chemical state of Ir: specifically, the Al-rich Beta zeolite induces the formation of more active Ir0 species by providing a higher density of anchoring sites, thereby offering abundant active centers for the CO-SCR reaction. The In 3d XPS results are shown in Fig. 3f. All catalysts exhibit a characteristic peak at 446.2 eV, primarily attributed to the dominant In3+ species. Trace amounts of In0 species are difficult to detect via XPS. To investigate the effect of Ir valence states on catalytic performance, we constructed a series of catalysts with varying Ir0/(Ir0 + Irδ⁺) ratios by adjusting the pretreatment temperature (as-prepared–800 °C) of the IrIn/Beta-30 catalyst (Fig. S16 and Table S6). Activity evaluation results reveal (Fig. S17) that the unreduced catalyst (as-prepared) exhibits almost no NOx conversion capability. However, as the pretreatment temperature increases, the proportion of Ir0 significantly rises from 2.63% at 200 °C to 63.96% at 800 °C, accompanied by a sharp increase in NOx conversion efficiency. This strongly proves that metallic Ir0 is the core active center driving the CO-SCR reaction. Additionally, the byproduct distribution curves indicate that catalysts with higher reduction degrees generate significantly less NO2 and N2O within the active window, further confirming that Ir0 sites effectively lower the energy barrier for N–O bond cleavage, promoting complete dissociation into N2. Consequently, among the various forms of Ir, metallic Ir0 has the greatest impact on the reaction and is the key species governing de-NOx activity. In the constructed IrIn bimetallic nanoparticles, the introduction of In as a critical electronic promoter significantly facilitates the reduction of Ir species and stabilizes a higher density of active Ir0 sites through electron transfer effects, thereby playing a vital promotional role in enhancing NO activation and conversion. Reaction gas adsorption testing The effects of Si/Al ratio changes on the adsorption-desorption behaviors of NO and CO were investigated using NO-TPD and CO-TPD. The NO-TPD results of IrIn/Beta catalysts and Beta carriers with different Si/Al ratios are shown in Fig. 4a and Fig. S18. Both NO and N2O desorption products were present on the Beta zeolite, among which N2O originated from the thermal decomposition of the unstable ONNO intermediate formed by the adsorption of NO43. On Beta with Si/Al of 30 and 80, NO desorption peaks only occurred at 114 °C, which was attributed to the weak physical adsorption of NO. However, desorption peaks at 300 °C and 693 °C were observed on Beta with Si/Al of 5, indicating that some NO could form a relatively stable structure on Beta-5. Comparing IrIn/Beta catalysts with varying Si/Al ratios, we found that the onset desorption temperature of N2O shifted significantly toward lower temperatures as the Si/Al ratio decreased (Fig. 4a). This provides strong evidence that high-density acidic sites can strongly polarize the adsorbed NO molecules via electrostatic field effects, thereby creating a favorable local microenvironment for NO adsorption and its subsequent dimerization into N2O intermediates. The area of the desorption peak usually represents the number of adsorption sites. The desorption peak area of the carrier is shown in Fig. S19. Compared with Beta-30 and Beta-80, Beta-5 has more adsorption sites and the largest N2O desorption capacity at the same time. As illustrated in Fig. 4c, the desorption behaviors of NO and N2O undergo significant evolution upon the introduction of metallic species. Primarily, the total desorption capacity of NO increases progressively as the Si/Al ratio decreases. This trend suggests that the abundant Si–OH–Al hydroxyl sites provided by the low-silica carrier, together with the highly dispersed metal active components, collectively offer a higher density of effective adsorption centers for NO. Simultaneously, a substantial downward shift in the onset desorption temperature of N2O is observed for all catalysts compared to the pristine supports as the Si/Al ratio decreases. This remarkable low-temperature promotional effect is fundamentally attributed to the electronic back-donation between the Ir0 active sites and NO molecules. Specifically, the d-orbitals of Ir sites donate electrons into the anti-bonding π* orbitals of NO, which induces a pronounced weakening of the N–O bond and significantly lowers the activation energy barrier for its cleavage8,44. Consequently, these results demonstrate that the Si–OH–Al hydroxyl sites are predominantly responsible for the initial adsorption and polar activation of NO, while the highly dispersed metal sites serve as the pivotal active centers for the complete dissociation of the N–O bond at reduced temperatures. The adsorption properties of catalysts toward reactants. The results of NO-TPD a and CO-TPD b of IrIn/Beta-5, IrIn/Beta-30, and IrIn/Beta-80; c The relative areas of N2O and NO desorption peaks in IrIn/Beta-5, IrIn/Beta-30, and IrIn/Beta-80 of the result of NO-TPD. d The relative areas of CO2 desorption peaks in IrIn/Beta-5, IrIn/Beta-30, and IrIn/Beta-80 of the result of CO-TPD. The CO-TPD results of IrIn/Beta catalysts with varying Si/Al ratios are shown in Fig. 4b. CO2 is the only desorption product on each catalyst, and its desorption peaks mainly occur at 270 °C, 430 °C and 700 °C. The desorption of CO2 is generally considered to be the product of the reaction between CO and the lattice oxygen on the catalyst surface45–47. Fig. 4d compares the relative areas of the CO2 desorption peaks of each catalyst. As the Si/Al ratio of the catalyst decreases, the amount of CO2 desorption gradually increases, which is attributed to more Ir0 sites in catalysts with a low Si/Al ratio. Studies have shown that CO exhibits a stronger binding strength on Ir0 and will preferentially adsorb onto Ir024,48. Therefore, the results of NO-TPD and CO-TPD jointly indicate that the catalyst with a low Si/Al ratio, due to the presence of more Ir0 sites and abundant Si–OH–Al sites, significantly improves the adsorption performance of the catalyst for NO and CO, which is conducive to the CO-SCR reaction. Identification of key intermediate substances To investigate the mechanism of SO2 adsorption on the IrIn/Beta catalyst, TPD tests were conducted on IrIn/Beta-30 with SO2 and various reaction gases (Fig. 5a). The SO2-TPD results revealed an SO2 desorption peak only at 127 °C, indicating that the sulfur on the catalyst surface was predominantly weakly adsorbed and showed minimal interaction with active surface species49,50. Compared to SO2-TPD, the SO2 + CO-TPD results showed a slightly reduced SO2 desorption peak at 127 °C without additional peaks, indicating that CO only marginally affected SO2 adsorption without reacting with it. The SO2 + O2-TPD results reveal a weak desorption peak for sulfur species at 764 °C, in addition to the SO2 desorption peak at 127 °C. This is attributed to the formation of strongly adsorbed sulfate species resulting from the reaction of trace amounts of O2 with SO2 on the catalyst surface. Further analysis of SO2 + NO-TPD revealed the near-complete disappearance of the SO2 desorption peak at 127 °C, accompanied by a new desorption peak at 646 °C. This indicates a significant interaction between NO and SO2. This reaction significantly suppresses the weak adsorption of SO2 and forms a new species on the catalyst surface. Therefore, we speculate that the formation pathway of this new species represents a novel mechanism where SO2 plays a key catalytic role in the reaction. SO2 Promotion Mechanism. a The results of SO2-TPD, SO2 + CO-TPD, SO2 + O2-TPD, SO2 + NO-TPD of IrIn/Beta-30. b The results of SO2 + NO-TPD, SO2 + NO + O2-TPD of IrIn/Beta-30. The XPS S 2p c and N 1s d results of IrIn/Beta-5, IrIn/Beta-30, and IrIn/Beta-80 after adsorption of NO, SO2, and O2 at 50 °C. e The results of NO + SO2 + O2-TPD of IrIn/Beta-5, IrIn/Beta-30, and IrIn/Beta-80. f The relative areas of NH3 and SO2 desorption peaks in IrIn/Beta-5, IrIn/Beta-30, and IrIn/Beta-80 of the result of NO + SO2 + O2-TPD. In situ DRIFTS results of successive adsorption of NO, CO, O2, and SO2 over IrIn/Beta-5 (g), IrIn/Beta-30 (h), and IrIn/Beta-80 (i) at 275 °C. Considering the influence of the simultaneous presence of NO, O2, and SO2 in the atmosphere on this species, we further tested SO2 + NO + O2-TPD and SO2 + NO-TPD on IrIn/Beta-30 for comparison (Fig. 5b). The SO2 + NO + O2-TPD results revealed that the SO2 desorption peak disappeared at 127 °C, with the species desorption peak appearing only at 646 °C. Quantitative comparison, based on the area of the SO2 desorption peak, revealed that the desorption peak for this new species increased when NO, O2, and SO2 coexisted in the atmosphere. Experimental results further confirm that the desorption species at 646 °C is the key intermediate species responsible for the significant promotion effect of SO2 on the IrIn/Beta catalyst. To further reveal the specific composition of this substance, we performed saturated adsorption on catalysts with different Si/Al ratios under an atmosphere of NO, O2, and SO2 at 50 °C, followed by X-ray photoelectron spectroscopy (XPS) to characterize the valence states of surface S and N elements. As shown in Fig. 5c, d, the characteristic peak at 169.5 eV in the S 2p spectrum is assigned to surface sulfate species, while the peak near 402.0 eV in the N 1s spectrum is attributed to ammonium salts51,52. Experimental results confirm that the key intermediate formed on the catalyst surface is ammonium sulfate. Concurrently, as the Si/Al ratio decreases (i.e., increasing Al content), the characteristic peaks for both sulfate and ammonium salts significantly intensify, indicating a marked increase in the formation of ammonium sulfate species on the catalyst surface. Subsequently, NO + O2 + SO2-TPD was employed to test catalysts with different Si/Al ratios, while monitoring the signal intensities for NH3 (m/z = 17) and SO2 (m/z = 48). As shown in Fig. 5e, all samples exhibited NH3 desorption peaks at 167 °C and SO2 desorption peaks at 646 °C. The XPS and TPD results collectively confirm an interaction mechanism among SO2, NO, and O2 on the IrIn/Beta surface, which significantly promotes catalytic activity by generating key reaction substances such as ammonium sulfate salts. Quantitative analysis of the NH3 and SO2 desorption peak areas in the TPD results (Fig. 5f) reveals that both peak values increase significantly as the Si/Al ratio of the support decreases, showing high consistency with the XPS results. Further comparison of the SO2 signal values reveals that the initial SO2 desorption temperature on IrIn/Beta-30 is approximately 270 °C, which is significantly lower than those on IrIn/Beta-5 and IrIn/Beta-80. This matches the optimal operating temperature window of IrIn/Beta-30, indicating that the generation and evolution of key ammonium sulfate intermediates can maintain a favorable dynamic balance within this system, thereby exerting a significant promotional effect on catalytic activity. To further elucidate the respective roles of bridging hydroxyl groups (Si–OH–Al) versus terminal hydroxyl groups (Si–OH) in the ammonium sulfate pathway, an aluminum-free control catalyst, IrIn/Si-Beta, was synthesized and evaluated via NO + CO + O2 + SO2-TPD and CO-SCR activity tests. The TPD profiles (Fig. S20) reveal that the all-silica catalyst, which possesses only terminal Si–OH groups, exhibited the lowest SO2 desorption intensity, indicating negligible formation of ammonium sulfate species. Concurrently, activity evaluations (Fig. S21) demonstrate that conversion over this all-silica catalyst is extremely low under simulated industrial flue gas conditions. This comparative result provides compelling evidence that the accumulation of ammonium sulfate species and the subsequent enhancement of this reaction pathway are positively correlated with the framework Al content (i.e., the density of Si–OH–Al sites). Consequently, it can be concluded that the ammonium sulfate mechanism over IrIn/Beta catalysts is mediated by surface hydroxyl species, among which the bridging hydroxyl groups (Si–OH–Al) play a predominant and governing role by significantly facilitating the formation of key ammonium sulfate intermediates. To investigate the influence of the support itself on the SO2 promotion mechanism, SO2 + NO + O2-TPD tests were conducted on Beta zeolites with different Si/Al ratios (Fig. S22). All zeolite samples exhibited desorption peaks for NH3 and SO2, indicating that surface hydroxyl groups on Beta zeolite participate in the reaction and play a crucial role in this mechanism. Specifically, Beta-5 showed distinct NH3 desorption peaks at 170 °C, 240 °C, and 400 °C, with SO2 desorption peaks appearing at 640 °C and 870 °C. Beta-30 exhibited desorption peaks at 165 °C, 290 °C, and 530 °C, with an SO2 desorption peak only at 640 °C. Beta-80 showed pronounced NH3 desorption peaks at 163 °C, 360 °C, and 710 °C, along with SO2 desorption peaks at 663 °C and 836 °C. Experimental results indicate that the desorption temperature of NH3 generally shifts toward lower temperatures as the Si/Al ratio of Beta zeolite decreases, while the desorption temperature of SO2 exhibits a volcano-shaped profile, with Beta-30 showing the lowest complete desorption temperature for SO2. Further quantitative plot analysis of NH3 and SO2 desorption peak areas in TPD results (Fig. S23) revealed that both desorption amounts gradually increased with decreasing Si/Al ratio of Beta zeolite. Experimental results indicate that an increase in Si−OH−Al bridging hydroxyl sites on the carrier surface promotes ammonium sulfate formation. However, excessive ammonium sulfate generation shifts the complete decomposition temperature toward higher temperatures. The promotion mechanism of SO2 on catalysts with different Si/Al ratios under actual reaction conditions was investigated using in situ DRIFTS (Fig. 5g−i). Two peaks at 2175 cm-1 and 2115 cm-1 were observed in the spectra of all samples, attributed to the adsorption of gaseous CO53,54. The peak near 2350 cm-1 corresponds to the adsorption of gaseous CO2, and its intensity gradually decreased as the Si/Al ratio of the catalyst decreased, indicating an enhanced suppression of the direct oxidation of CO by O2. The peak at 1839 cm-1 corresponds to the stretching vibration characteristic peak of NO molecules linearly adsorbed on Ir0 sites. The Ir0–NO characteristic peak in the IrIn/Beta-30 spectrum is significantly stronger than those in IrIn/Beta-5 and IrIn/Beta-80. Compared with IrIn/Beta-80, this is mainly attributed to the fact that the surface of IrIn/Beta-30 catalyst has more Ir0 sites, which is consistent with the XPS results. For the IrIn/Beta-5 catalyst, excessive ammonium sulfate species rapidly formed on the surface. Undecomposed ammonium sulfate deposits formed on the catalyst surface, subsequently covering surface Ir0 sites. The peaks located in the range of 1420–1630 cm-1 are assigned to the nitrate species formed on the catalyst surface55,56. As the Si/Al ratio of the catalyst support decreases, nitrate species significantly diminish, indicating that the enhanced ammonium sulfate pathway substantially inhibits the progression of the nitrate pathway. In summary, the hydroxyl-mediated ammonium sulfate salt mechanism follows an appropriate dynamic reaction equilibrium. When the Si/Al ratio increases from 30 to 80, the aluminum content in the zeolite framework decreases significantly, leading to a substantial reduction in the density of surface bridging hydroxyl groups. Consequently, the formation rate of ammonium sulfate intermediates is severely restricted, thereby inhibiting the subsequent built-in NH3-SCR reaction pathway. Thus, the extremely low hydroxyl density of IrIn/Beta-80 prevents it from effectively leveraging the SO2 promotion effect, thereby limiting its catalytic activity. Although a lower Si/Al ratio (such as Beta-5) provides a higher density of hydroxyl groups, the excessive generation of ammonium sulfate species is difficult to decompose completely at the reaction temperature. Instead, these species tend to cover active metal sites, thereby creating mass-transfer resistance. Therefore, the superior performance exhibited by IrIn/Beta-30 essentially stems from reaching an optimal balance between a sufficient supply of active sites and a moderate rate of intermediate evolution. Key DFT calculation results and reaction mechanisms To gain a profound insight into the reaction mechanism and the ammonium sulfate pathway over the IrIn/Beta catalyst, systematic density functional theory (DFT) calculations were performed. Based on the aforementioned comprehensive characterization results, a representative IrIn bimetallic site model was established (Fig. S24). As illustrated in Fig. 6a, the charge density difference plot intuitively reveals a robust electronic interaction between Ir and In, characterized by a significant transfer of electron density from the In sites to the neighboring Ir atoms. This pronounced electron-rich character of the Ir sites aligns perfectly with observations from CO-DRIFTS, XAS, and XPS, which consistently indicate the high electron density of the active Iridium centers. Based on this model, we focused on calculating the reaction energy barriers for several potential pathways of NO conversion under complex atmospheres, including the dimerization of NO to ONNO, the oxidation of NO to NO2, direct N−O bond dissociation, and the stepwise hydrogenation of NO to NH3. First, regarding the activation and oxidation of NO (Fig. 6b and Figs. S25, 26), the calculations reveal that the direct oxidation of NO* to NO2* must overcome a substantial energy barrier of 1.44 eV, indicating that this pathway is kinetically sluggish. In contrast, the ONNO pathway, characterized by bimolecular coupling, exhibits a significantly lower activation energy of 0.86 eV for the critical ONN−O (TS) step. This strongly demonstrates that on the Ir−In dual-active sites, NO transformation preferentially proceeds through the dimerization pathway rather than direct oxidation57,58. Second, focusing on the core ammonium sulfate pathway (Fig. 6c and Figs. S27–29), the full energy evolution of NO hydrogenation to NH3 was simulated. The results indicate that the proton spillover pathway mediated by surface hydroxyl groups on the support is thermodynamically more favorable than direct N−O bond cleavage (Ea = 0.47 eV). Specifically, the migration of protons from the zeolite framework to the metal sites to facilitate the formation of the NOH* intermediate (via the NO−H TS transition state) requires a barrier of only 0.48 eV. The subsequent dehydration of NOH* and stepwise hydrogenation steps are highly exothermic, with energy levels dropping significantly (from –1.07 eV to –4.61 eV). This confirms that protons provided by the surface hydroxyl groups effectively mediate the efficient reduction of NO to NH3*, laying the essential foundation for the formation of ammonium sulfate species. Furthermore, to elucidate the regulatory role of SO2 in the system (Fig. 6d and Fig. S30), the competitive oxidation of CO and SO2 was investigated. Thermodynamic analysis shows that the free energy change for SO2 oxidation to SO3* (ΔG = –0.33 eV) is significantly more favorable than that of CO oxidation (ΔG = –0.13 eV). This thermodynamic advantage ensures that SO2 preferentially consumes surface-active oxygen species, thereby inhibiting the side reaction of CO oxidation. Subsequently, the in-situ generated NH3* rapidly combines with SO3* to form transient ammonium sulfate intermediates. These species dynamically decompose under reaction conditions to release active NH3, triggering the highly efficient in-situ NH3-SCR reaction pathway. Simultaneously, under reaction conditions, the generated water molecules can undergo dissociative adsorption on the catalyst surface, effectively replenishing the hydroxyl groups and protons (H*) consumed from the zeolite framework. This water-assisted proton regeneration pathway ensures the continuous restoration of active sites, endowing the ammonium sulfate mechanism with exceptional sustainability and stability throughout the reaction process. Key DFT calculation results and reaction mechanisms. a Charge density difference of the IrIn bimetallic dual-site model. b The reaction energy barriers of the ONNO pathway and the nitrate pathway on IrIn. c Calculated potential energy profile for the ammonium sulfate pathway on the catalyst surface. d Free energy profiles were calculated for the combination of CO and SO2 with the active oxygen species on the catalyst surface. e The reaction mechanism of CO-SCR on IrIn/Beta. Based on the above experimental characterization and DFT calculation results, the possible CO-SCR reaction pathways on the IrIn/Beta catalyst are shown in Fig. 6e. Our previous research has established that the low-temperature CO-SCR reaction predominantly follows the Langmuir−Hinshelwood (L−H) mechanism, with the NO reduction process primarily involving the ONNO pathway and the nitrate pathway8,10,24. Under anaerobic conditions, the ONNO pathway is dominant: two NO molecules adsorb on the IrIn active sites to form an ONNO intermediate, which then reacts with adsorbed CO to release N2 and CO2. Under oxygen-rich conditions, however, some NO and CO are directly oxidized by O2, leading to competitive oxidation side reactions. The remaining NO adsorbs on the IrIn active sites, forming surface nitrates that then react with adsorbed CO species. Unfortunately, these nitrate species easily undergo further decomposition, releasing NO2 and NO by-products, which inevitably reduces the NOx conversion and N2 selectivity. This is thus characterized as an "ineffective" reaction pathway. However, this study reveals that in the presence of SO2, the reaction shifts toward a thermodynamically more favorable ammonium sulfate pathway mediated by surface hydroxyl groups. Specifically, the hydroxyl groups on the support induce the hydrogenation of NO via a proton spillover effect, sequentially forming NOH* and NH3* intermediates. This pathway is energetically superior to direct N−O dissociation. Simultaneously, since SO2 is more readily oxidized than CO, it preferentially consumes active surface oxygen species, thereby effectively blocking the CO oxidation side reactions and mitigating the ineffective nitrate pathway. The resulting NH3* subsequently combines with surface sulfur oxides to form transient ammonium sulfate species, which rapidly decompose under the reaction conditions to release active NH3, further triggering a built-in NH3-SCR cycle. Notably, trace water molecules generated as a reaction byproduct undergo dissociative adsorption to facilitate water-assisted proton regeneration, continuously replenishing the consumed hydroxyl groups and protons on the zeolite framework. This spontaneous hydroxyl cycle, relying entirely on internal byproduct water, exhibits exceptionally strong kinetic resilience. The macroscopic dynamic water resistance test perfectly corroborates this microscopic process: even after the severe perturbation of abruptly removing the external water supply, the catalyst can rapidly re-establish the microscopic generation-consumption dynamic equilibrium of hydroxyls, autonomously recovering to its excellent steady-state activity. Such rapid self-repair at the microscopic level ensures the inherent sustainability of this mechanism in the absence of external water, ultimately achieving a significant enhancement in both the NOx conversion and N2 selectivity of the IrIn/Beta catalyst under high-oxygen, complex industrial atmospheres. Discussion In summary, this study systematically compared a series of IrIn bimetallic catalysts supported on Beta zeolites with different Si/Al ratios and found that the hydroxyl-mediated CO-SCR reaction over IrIn/Beta follows a novel reaction pathway, which we term the ammonium sulfate mechanism. The in-situ generated ammonium sulfate salt species can rapidly decompose and produce NH3 at 167 °C, thereby triggering the built-in NH3-SCR reaction. This unique mechanism can effectively inhibit the oxidation side reactions between NO and CO. It is notable that, compared with the IrIn/Beta catalysts with Si/Al ratios of 5 and 80, the specific catalyst with Si/Al ratio of 30 achieved an excellent NOx conversion efficiency of 88% under the simulated actual industrial flue gas conditions (NO 400 ppm, CO 8000 ppm, O2 15%, SO2 200 ppm). Its catalytic activity is significantly superior to that of most Ir-based catalysts reported under the same conditions, thereby greatly enhancing the feasibility of practical industrial applications of CO-SCR catalysts. Methods Catalyst preparations At first, 1.98 g of Beta-30 was added into 100 mL of deionized water under stirring. Subsequently, added 1.54 mL of hexachloroiridium acid hydrate solution (containing 0.006 g of Ir) while simultaneously adding 0.036 g of indium nitrate hydrate (the molar ratio of Ir to In was 1:3). After stirring for 3 hours, evaporate the mixture under reduced pressure at 60 °C using a rotary evaporator at 80 r min-1. The solid residue was dried overnight in an oven at 100 °C. The resulting powder was calcined at 500 °C with a heating rate of 10 °C min-1 for 4 hours. The calcined powder is denoted as IrIn/Beta-30. Prior to catalytic activity evaluation, IrIn/Beta-30 samples were pretreated at 800 °C under a 5% H2/N2 atmosphere for 3 hours to yield the final IrIn/Beta-30 catalyst. Catalysts with different Si/Al ratios were prepared using the same method as IrIn/Beta-30, employing Beta-5 and Beta-80 as supports, respectively. The final catalysts were labeled as IrIn/Beta-5 and IrIn/Beta-80. Catalyst characterizations Catalysts were characterized by X-Ray Diffraction (XRD), Fully Automated BET Surface area and Porosity Analyzers (BET), Transmission Electron Microscopy (TEM), High-Angle Annular Dark-Field Scanning Transmission Electron Microscopy (HAADF-STEM), Energy Dispersive X-Ray Spectroscopy (EDS) elemental mapping, Inductively Coupled Plasma Optical Emission Spectroscopy (ICP-OES), in situ Diffuse Reflectance Infrared Fourier Transform Spectroscopy (in situ DRIFTS), X-Ray Absorption Fine Structure (XAFS), X-Ray Photoelectron Spectroscopy (XPS), H2 Pulse Desorption (H2-TPD), NO Temperature-Programmed Desorption (NO-TPD), CO Temperature-Programmed Desorption (CO-TPD), NO, O2, SO2 Temperature-Programmed Desorption (NO + O2 + SO2-TPD), and Density Functional Theory calculations (DFT). More information about the catalyst characterizations is given in the supplementary information. Catalytic activity evaluation SCR activity tests were carried out in a fixed-bed quartz reactor with an inner diameter of 6 mm. The reaction conditions were as follows: 400 ppm NO, 8000 ppm CO, 15 % O2, and 200 ppm SO2 in nitrogen. Approximately 100 mg of catalyst (40−60 mesh, 0.25 cm3) was used for each test. The total flow rate was 100 sccm under ambient conditions, with a GHSV of 24,000 h-1. The composition of the inlet and outlet gases was continuously monitored online using an infrared gas analyzer (Bruker Tensor Ⅱ). Before all tests, the catalyst was placed in a vertical fixed-bed quartz tube and pretreated at 800 °C with 5% H2/N2. The NOx conversion (ηNOX), CO conversion (ηCO), and N2 selectivity (SN2) were calculated using the following equations: Supplementary information Supplementary Information Transparent Peer Review file Source data Source Data