Abstract
Introducing and stabilizing oxygen vacancies in the lattice of oxide support materials has been a huge challenge to boost the catalytic activity in redox reaction. Herein, we fabricate porous single-crystalline Nb2O5 monoliths at the centimeter scale, then we introduce the oxygen vacancies in lattice by reduction treatment and stabilize the oxygen vacancy defects by the ordered lattice. We then deposit Pt clusters at surface in the porous microstructures to construct an interfacial system and we show the significantly enhanced surface oxygen exchange coefficient. As a case study, we demonstrate the complete CO oxidation with air at 110 °C and 30,000 mL g-1 h-1 conditions, following the Mars-van Krevelen mechanism, with no visible performance attenuation after a continuous operation of ~500 h. At 60,000 mL g-1 h-1 and 110 °C, the CO conversion rate decreases to 83.8% and ultimately remains 76.3% after 185 h of long-term operation.
Similar content being viewed by others
Introduction
Porous single crystals are a new type of porous material at the macroscale, which combines the advantages of porosity and single crystallinity in a monolith. The properties of these porous single-crystalline (PSC) materials can be further tailored by controlling their unique structural features, making them ideal catalyst supports because of their high specific surface areas and good chemical and thermal stabilities1. The clear surface is continuous in the porous microstructures, with clear stepped atoms at the inner surface in the porous microstructures, further engineering the electronic structure of the well-defined surface structure and leading to a more controllable interaction between the adsorbed substance and the active site at the surfaces2.
Niobium oxide (Nb2O5) usually serves as a solid acid catalyst and unique Lewis acid sites and Brønsted acid sites are observed at the surface3,4. The Nb2O5 shows high acid strength and is usually utilized as a heterogeneous catalyst support in the dehydration reaction and hydrodeoxygenation reaction. Oxygen vacancies usually play a key role in enhancing the catalytic performance in the redox reaction. However, its reducibility to form oxygen vacancies in the lattice remains underexplored. To date, only a limited number of studies have leveraged the exceptional properties of Nb2O5 to create oxygen vacancies5,6. In previous studies, the ability to generate oxygen vacancies in metal oxides has been shown to be extremely beneficial for catalytic applications. Oxygen vacancies usually facilitate the gas-phase oxygen recompletion and rapid catalyst surface charge transfer in heterogeneous catalysis, and therefore they are considered to be one of the most effective ways to improve catalytic performance7,8,9,10,11. Porous single-crystalline (PSC) Nb2O5 monoliths, with well-defined surfaces capable of generating active surfaces, show great potential owing to the synergistic effect of their interconnected porous structure and single-crystalline framework, which can introduce oxygen defects and stabilize them in the lattice following proper reduction treatment. These unique structures have inherent mechanical and thermal stability, which is conducive to the rapid diffusion of reactant species and products through a three-dimensional percolation pathway12. Therefore, the PSC Nb2O5 monoliths would demonstrate high catalytic performance and thermal durability when they are used as catalyst supports in the catalytic redox reaction13,14,15.
Catalytic oxidation usually requires the activation of oxygen at the active site at the catalyst surface16,17,18,19. However, the competitive adsorption of different reactants may hinder the chemisorption of oxygen, resulting in high operating temperatures for these oxidation reactions, such as the CO oxidation with air20,21. In the process, interfacial systems show the greatest potential to achieve efficient catalysts for CO oxidation reaction by using oxide supports with oxygen non-stoichiometry to facilitate the surface reactions22. Pt clusters are known to efficiently form active metal-oxide interfaces that can effectively improve CO oxidation reactions23,24. The loading of Pt clusters on the surfaces of PSC Nb2O5 monoliths would produce well-defined interface structures, which would facilitate the identification and engineering of active sites on the active metal-oxide interfaces25,26.
In this work, we fabricate centimeter-scale PSC Nb2O5 monoliths with Pt clusters deposited on the surface to form an interfacial system for enhancing catalytic CO oxidation. Notably, we achieve the complete oxidation of CO with air at 110 °C, with no visible degradation observed even after ~500 h of continuous operation. The current work will provide a new line of thought for the preparation of novel catalysts enriched with oxygen vacancies by confining oxygen-vacancy-rich and well-defined active structures at the surface for enhancing catalytic activity and stability.
Results and discussion
Growth of porous single crystals
We grow PSC Nb4N5 monoliths from NaNbO3 single crystals using a solid-phase transformation method in an NH3 atmosphere27. We then grow PSC Nb2O5 monoliths by directly transforming the PSC Nb4N5 monoliths in oxygen atmosphere at elevated temperature. Supplementary Fig. 1a, b shows the X-ray diffraction (XRD) patterns of the NaNbO3 single crystals with the facet of (001) and (010), respectively. Supplementary Fig. 1c, d shows the c-axis and b-axis channels for removing the sodium and oxygen atoms from the lattice. Accordingly, we successfully grow PSC Nb4N5 monoliths with (002) facet from the NaNbO3 single crystals with both (100) and (010) crystal facets. In Supplementary Fig. 1e, f, the transformation of Nb4N5 to Nb2O5 is an oxidation process that involves the replacement of nitrogen with oxygen in the lattice in the phase transformation process. In Supplementary Fig. 1g, h, the polyhedral model demonstrates the lattice reconstruction process, ultimately leading to the growth of PSC Nb2O5 monoliths from PSC Nb4N5 monoliths. In this process, PSC Nb4N5 monoliths function as parent single crystals to grow porous single crystals.
Figure 1a shows the XRD patterns of PSC (002) Nb4N5 monoliths and PSC (001) Nb2O5 monoliths, confirming their single crystal characteristics even though they are porous architectures. Figure 1b, c shows scanning electron microscopy (SEM) images of the <002> oriented PSC Nb4N5 and <001> oriented PSC Nb2O5, respectively. Microscopic morphological characterization indicates that the pores in PSC Nb2O5 monoliths are larger than those in PSC Nb4N5 monoliths at the same scale, and a similar difference in internal pore density is also observed. As shown in Supplementary Fig. 2, we further conduct multiple measurements to determine the specific surface area and average pore size of PSC Nb2O5 monoliths. We observe that the PSC Nb2O5 monoliths exhibit a specific surface area of ~5.8–6.3 m2 g−1 and an average pore size of ~18–22 nm. The specific surface area of PSC Nb2O5 monoliths is slightly higher than that of PSC Nb4N5 monoliths. More pores are present within PSC Nb2O5 monoliths which is in good accordance with the SEM analysis.
a–c XRD pattern, SEM images of exhibiting the crystal structure and morphology. d–f HAADF-STEM and Cs-HRTEM images, along with the elemental mapping and atomic stacking. The blue and gray spheres represent O and Nb atoms, respectively. g–i HS-LEISS, differential charge density map of the top layer at the surface, and XPS results of PSC Nb2O5 monolith. The OI and OIII represent lattice oxygen and adsorbed oxygen, respectively.
Supplementary Fig. 3 shows the microstructure morphology of the sliced sample with insets b, c and d at the various magnification scales. These images clearly demonstrate that PSC Nb2O5 monoliths produced via the lattice reconstruction strategy that involves nitridation and re-oxidation exhibit a uniform distribution of pore channels and a good internal pore structure. The sliced samples remain intact with no fragments detached, indicating that the single-crystalline skeletons have reasonable mechanical strength. Supplementary Fig. 4 shows a magnified view of the sliced sample and we use the spherical aberration corrected transmission electron micrograph (STEM) and high-resolution transmission electron microscopy (HRTEM) images to investigate the skeleton and surface in the PSC Nb2O5 monoliths. This analysis omits some of the larger pores at several microns that are created by the curtain effect during slicing. The coexistence of twisted surfaces and pore channels can be clearly observed in the PSC Nb2O5 monoliths, further demonstrating the single-crystalline features of the skeletons in the porous microstructures. Supplementary Fig. 5 and Fig. 1d show the selected areas of electron diffraction (SAED) patterns of the sliced sample, again confirming the single crystallinity across different regions. The regular diffraction spots indicate that the overall lattice remains ordered even at different locations. The microstructure of the PSC Nb2O5 monoliths is analyzed using Cs-corrected HRTEM (Cs-HRTEM) in Fig. 1e. The results further confirm the single-crystalline nature of the interconnected Nb2O5 skeletons within the porous microstructure.
The elemental mapping characterization of the sliced samples present in Fig. 1f reveals the absence of any additional elements, indicating the complete transformation of PSC Nb4N5 monoliths to PSC Nb2O5 monoliths in the solid-solid phase transformation. The top layer at the surface in PSC Nb2O5 monoliths is confirmed to be Nb together with O atoms, as evidenced by the high sensitivity low energy ion scattering spectroscopy (HS-LEISS) results presented in Fig. 1g. Figure 1h shows the differential charge distribution of the Nb-O coordination structure, highlighting the electron-deficient active center and indicating a distinct charge transfer from the Nb atom to the O atom at the surface in PSC Nb2O5 monoliths. Charge transfer from the metal to oxygen in the metal-oxygen bond usually occurs because of the difference in the electronegativities. The PSC Nb2O5 monoliths are further characterized by X-ray photoelectron spectroscopy (XPS) analysis in Fig. 1i, revealing Nb elemental peaks corresponding to Nb5+ states, along with oxygen elemental peaks associated with the presence of lattice oxygen (OI) and adsorbed oxygen (OIII).
Supplementary Fig. 6a shows the XRD spectra of a heterogeneous junction structure where Nb4N5 and Nb2O5 co-exist in the sample which is prepared by controlling the transformation time. To better understand the transformation process, we conduct the energy dispersive spectroscopy (EDS) elemental analysis of the sample with a shorter oxidation time in Supplementary Fig. 6b. The results indicate that, in PSC Nb4N5 monoliths with a short reaction time, a small amount of oxygen enter the sample to replace nitrogen. Based on the experimental conditions explored, the oxidation time of the nitride precursors in the reaction should be at least 0.5 h to achieve the complete conversion of PSC Nb4N5 monoliths to prepare PSC Nb2O5 monoliths. We further ground PSC Nb2O5 monoliths into powder and then conduct the XRD test to determine the crystalline phase in Supplementary Fig. 6c. The result conforms with JCPDS card PDF#30-0873, indicating an orthorhombic crystal system with a space group of Pbam (55) and a dotting constant of 6.175 Å × 29.175 Å × 3.93 Å <90.0° × 90.0° × 90.0°> for T-Nb2O5. As shown in Supplementary Fig. 6d, EDS analysis of PSC Nb2O5 monoliths reveals the presence of only Nb and O, with no other elements detected, further confirming the complete conversion of PSC Nb4N5 monoliths to PSC Nb2O5 monoliths. Supplementary Fig. 7 shows the SEM images of the upper and lower surfaces of the incompletely transformed heterostructured samples. PSC Nb2O5 monolith in white and PSC Nb4N5 monoliths in black highlight the coexisting phases at the interface.
Microstructure of porous single crystals
Figure 2a–c shows the HAADF-STEM images of these twisted surfaces at different scales. Analysis of the atomic distribution reveals the <200> and <0100> crystallographic orientations, with atomic spacing d of approximately 0.308 and 0.292 nm, respectively. The surface structure exhibits an identical lattice orientation around the pores, with lattice distortion present in the thickness of ~1–3 atomic layers. Figure 2d–f shows HAADF-STEM images of a flat area at various scales near another pore locations, revealing lattice orientations of <151> and <230 > , with atomic spacing d of approximately 0.288 and 0.294 nm, respectively. The overall average atomic spacing in this region is smaller than that at the upper location. In summary, the primary difference between the two atomic arrangement diagrams is the degree of lattice distortion, which can vary across different twisted surfaces and is relatively larger near the pore channel. This results in a slightly larger atomic spacing, which allows the lattice oxygen of PSC Nb2O5 monolith to be activated more easily than that of other Nb2O5 materials, thereby enhancing catalytic performance.
a–c Twisted surface near the aperture. d–f Twisted surface at a slightly distant position of the pores. g XRD diagram of PSC Nb2O5 monoliths after reduction, and the Nb2O5-H is a reduced state. h Cs-HRTEM images of surface structures around pores. i Differential charge density plot for twisted surfaces.
Despite the slight lattice distortion at the twisted surface, the single-crystalline structure maintains the typical lattice ordering characteristics. The twisted surface not only ensures the high catalytic activity but also facilitates the interactions between the adsorbed substances and the well-defined surface structure. To engineer oxygen vacancies, the PSC Nb2O5 monolith is reduced by placing it in a reaction chamber at 750 °C with a hydrogen-argon flow rate of 50 mL min−1 for 10 h. Figure 2g shows the XRD pattern, with Nb2O5-H representing the sample prepared after reduction. The intensity of the diffraction peaks in the reduced sample is lower than that of the unreduced sample, primarily because of the structural defects in the reduced sample, which not only decreases the intensity of the diffraction peaks but also shifts the positions. A spherical differential electron micrograph of the reduced PSC Nb2O5 monolith is shown in Supplementary Fig. 8a, indicating that the atomic arrangement remains largely regular; however, some areas of inhomogeneity can still be observed in the lattice structure. Analysis in the transverse and oblique directions in Supplementary Fig. 8b clearly shows that the intensity of the lattice atomic arrangement changes from strong to weak and then back to strong, which can be attributed to the presence of atomic defects in the reduced PSC Nb2O5 monolith, resulting in the formation of oxygen vacancies confined in the lattice.
To verify the presence of oxygen vacancies in the reduced PSC Nb2O5 monolith, the sample is subjected to the test of electron paramagnetic resonance (EPR) that detects unpaired electrons to determine oxygen vacancies. As shown in Supplementary Fig. 9, the pure-phase sample is measured against the reduced sample, yielding g-factor of 2.002 that indicates the presence of oxygen vacancies28. Figure 2h shows a continuously twisted surface around the pore, where the lattice orientation remains consistent when viewed from a fixed orientation. Figure 2i shows the differential charge density of the twisted surface structure, indicating that the top layer is in an electron-deficient state, which facilitates the reception of electrons during C-O activation.
Construction of interfacial system
We construct an interfacial system by depositing a Pt cluster at the skeleton surface in the PSC Nb2O5 monoliths. The content of Pt at 0.1 wt%, 0.3 wt%, 0.5 wt%, 0.7 wt%, 1 wt%, and 4 wt% is determined by using inductively coupled plasma (ICP) analysis in Supplementary Fig. 10, with the average values from three tests yielding values of 0.13 wt%, 0.45 wt%, 0.60 wt%, 0.84 wt%, 0.98 wt% and 3.35 wt%, respectively. HAADF-STEM images characterization of PSC Nb2O5 monoliths with different amounts of loaded Pt is presented in Fig. 3a–c. The HAADF-STEM images of the PSC Nb2O5 monolith with 3.35 wt% Pt content at different scales show that, owing to the excessive amount of loaded Pt atoms, clusters finally form on the surface, by stacking and arranging Pt atoms into layers and resulting in a cluster or particle state. Supplementary Fig. 11 shows the HRTEM images of the PSC Nb2O5 monolith with 0.60 wt%, 0.84 wt%, and 0.98 wt% Pt, respectively, showing an increase in the number of loaded atoms. Pt clusters form into particles with crystal spacing d = 0.23 nm, suggesting that these particles correspond to the (111) orientation of Pt, and based on the original field of view, the particles are uniformly distributed across the entire surface of the PSC Nb2O5 monolith.
a–c HAADF-STEM images of PSC Nb2O5 monolith with loaded Pt clusters on the surface. d Elemental mapping of PSC Nb2O5 monolith with loaded Pt clusters on the surface. e Differential charge density graph of the surface structures with Pt clusters on the surface. f HS-LEISS of PSC Nb2O5 monolith with loaded Pt clusters on the surface. g, h EXAFS of PSC Nb2O5 monoliths with loaded Pt clusters on the surface. i Raman spectra of PSC Nb2O5 monoliths with loaded Pt clusters on the surface.
As the loaded Pt content increases to form particles, some visible Pt particles become encapsulated by the carrier owing to the presence of strong metal–support interactions, thereby embedding them within the single-crystalline porous oxide to form a Pt/NbOx interface. This structure endows the catalyst with high thermal stability. However, because Pt behaves similarly to absorbed CO, the catalytic activity of this catalyst is significantly compromised29. In general, increasing the amount of loaded Pt results in the formation of larger particles under reducing conditions. However, owing to its stable oxygen vacancies, PSC Nb2O5 monolith would limit the generation of large particles to some extent. Consequently, increasing the amount of loaded Pt leads to a uniform distribution across the surface in the PSC Nb2O5 monolith, which therefore would improve the catalytic activity toward CO oxidation. Figure 3d shows the elemental mapping of the PSC Nb2O5 monolith with the surface featuring Pt clusters, illustrating the distribution of nanoscale Pt clusters in the porous microstructure.
Figure 3e shows a differential charge density map of the surface structure with Pt clusters, indicating charge transfer from Pt to O, which effectively activates the lattice oxygen linked to Pt at the surface. Figure 3f shows the HS-LEISS of PSC Nb2O5 monolith with the Pt at 3.35 wt%. Notably, Pt primarily exists in a state of clusters or particles. The Ne spectrum shows the Pt signals in addition to Nb signals, indicating that Pt is stably loaded on the surface. The HS-LEISS for the PSC Nb2O5 monoliths with other loading amounts of Pt are shown in Supplementary Fig. 12, where all spectra reveal the presence of Pt on the surface of PSC Nb2O5 monolith. As can be seen in Fig. 3g, the signal of Pt-3L is weaker than that of the Pt foil, which suggests that the Pt in the PSC Nb2O5 monoliths is mainly in the reduced state (Pt0). And the PSC Nb2O5 monoliths with different Pt loading contents exhibit almost the same characteristics. With the presence of Pt in the form of nanoparticles or clusters, surface dangling bonds and low-coordinated atoms may lead to a change in the density distribution of the electronic states, thus weakening the intensity of the characteristic peaks30,31. The presence of a small amount of oxidized state Pt leads to a slight shift of the near-side energy to higher energies.
Figure 3h shows the Fourier transform (FT) of the extended X-ray absorption fine structure (EXAFS) spectra for the PSC Nb2O5 monoliths with loaded Pt. The shape of the curve of the reduced Pt/Nb2O5 sample is similar to that of Pt0. The bond length of Pt-Pt is approximately 2.7 Å, indicating the presence of a Pt-Pt bond in the sample and confirming that Pt is in the metallic state on the surface of the PSC Nb2O5 monolith. It is also observed that the Pt-O bond is present in the curve, indicating the presence of partially oxidized Pt in the form of Pt2+ in the clusters. The Raman spectra of the PSC Nb2O5 monoliths loaded with Pt clusters exhibit a high-intensity broadband centered at approximately 670–690 cm−1 in Fig. 3i, which corresponds to the characteristic vibrational peaks of the coexistence of NbO6 and NbO7. This result aligns with the literature on the T-Nb2O5 spectrum, and the absence of nitrogen characteristic peaks on the surface suggests complete transformation32. Because of its sensitivity to the structure and chemical bonding of metal oxides, Raman spectroscopy is well-suited for investigating defects caused by doping or oxygen vacancies. A partial enlargement of the Raman profile is shown in Supplementary Fig. 13, where the peak value shifts to the left from 685 to 675 cm−1 with increasing Pt loading content, indicating that the number of oxygen vacancies increases and that the lower Nb–O bond level leads to a longer Nb-O bond length in reduced states.
Catalytic performance
Figure 4a shows the results of CO oxidation experiments with the Pt loading contents, indicating that the temperature required for the complete conversion of CO to CO2 remarkably decreases at higher loading contents. At a loading of 3.35 wt% and a space velocity of 30,000 mL g−1 h−1, the complete conversion temperature finally reaches approximately 110 °C, while the complete conversion is not achieved by using PSC Nb2O5 monoliths without Pt loading even at 300 °C. Figure 4b shows that PSC Nb2O5 monoliths loaded with 3.35 wt% of Pt loading demonstrates the outstanding durability, owing to the structural stability of the single-crystalline features in the porous microstructure. We therefore observe the complete CO conversion even after a continuous operation of ~500 h under the operation condition of high space velocity. As shown in Fig. 4c, the metal-oxide interfacial system in the PSC Nb2O5 monoliths significantly increases the surface oxygen exchange coefficient, which further suggests that the fast dynamic equilibrium of surface defect reaction of OOx + 2 h•⇌VO•• + 1/2 O2 effectively facilitate the activation of lattice oxygen. This is pretty favorable for enhancing the catalytic oxidation of CO with air in the interfacial system in PSC Nb2O5 monoliths. Moreover, this relationship between the surface oxygen exchange coefficient and the surface Pt loading contents further confirms the effectiveness of the localized structures of metal-oxide interfacial system in activating lattice oxygen in the PSC Nb2O5 monoliths. Supplementary Fig. 14 shows the EXAFS spectra of the PSC Nb2O5 monoliths with 3.35 wt% of Pt loading at different temperatures under the flow of CO gas. The EXAFS indicates that the Nb K-edge and scattering distance (R) generally remain identical for the PSC Nb2O5 monoliths with 3.35 wt% of Pt loading before and after CO adsorption at different temperatures, which suggests the structural stability of PSC Nb2O5 monoliths.
a CO oxidation of PSC Nb2O5 monoliths decorated with different contents of Pt clusters. b Long-term stability experiments with PSC Nb2O5 monoliths decorated with 3.35 wt% at 110 °C. c Surface oxygen exchange coefficients of PSC Nb2O5 monoliths with different Pt loading contents. d, e In situ FTIR of CO oxidation with air using PSC Nb2O5 monoliths decorated with 3.35 wt% Pt cluster. f XPS of PSC Nb2O5 monoliths decorated with 3.35 wt% Pt cluster. g EPR of PSC Nb2O5 monoliths decorated with different loading contents of Pt cluster. h Activation energies of CO oxidation by using PSC Nb2O5 monoliths decorated with different loading contents of Pt cluster. i The potential energy diagram of CO oxidation including surface intermediates and transition states.
To analyze the catalytic effect of PSC Nb2O5 monoliths with different Pt loading contents, in situ infrared measurements have been conducted to assess the adsorption and activation for CO at different temperatures. As shown in Fig. 4d, the main adsorption peak of PSC Nb2O5 monoliths loaded with 3.35 wt% Pt is located at approximately 2090 cm−1, which indicates the linear adsorption at the interfacial system. The adsorption peaks shift relatively to blue region when the temperature ranges from low to high temperatures, indicating that Pt is positively charged, while the intensity of the adsorption of CO gradually increases with increasing temperature. The peaks at approximately 2175 cm−1 can be primarily attributed to CO gas. Figure 4e shows the position of the adsorption peaks for CO2 generation from PSC Nb2O5 monoliths with 3.35 wt% of Pt loading content. The intensities of the peaks increase with increasing temperature, indicating that the interfacial system exhibits higher activity in catalytic oxidation of CO at higher temperatures. A higher operation temperature promotes faster oxidation of CO by effectively activating the lattice oxygen linked to the Pt clusters in the interfacial system at the surface. As shown in Supplementary Fig. 15, the linear adsorption is also observed because the main adsorption peaks locate at approximately 2070–2090 cm−1 for the Pt-loaded PSC Nb2O5 monoliths, which is also validated by the relative blue shift of the adsorption peaks at high temperatures. Another set of plots shows the position of the adsorption peaks during the CO2 production for the PSC Nb2O5 monoliths with 3.35 wt% of Pt loading content. The peak intensity increases with the rise in temperature, indicating that the interaction between CO molecules and the active lattice oxygen atoms present on the surface becomes stronger at higher temperatures, and the desorption amount of CO2 also increases accordingly.
In Fig. 4f and Supplementary Fig. 16, XPS results of the Pt-loaded PSC Nb2O5 monoliths after reduction treatment reveal that the elemental peaks of Nb are nearly identical. Each of the O peaks consists of lattice oxygen (OI), oxygen vacancies (OII) and adsorbed oxygen (OIII) with binding energies of approximately 529.8, 531.0 and 532.3 eV, respectively. This further provides the evidence of the presence of oxygen vacancies in the lattice. Pt has peaks in two states, Pt0 and Pt2+, and although some of the Pt is oxidized, but most of the Pt remains in the metallic state, with the Pt0 peak dominating the curve. Based on the EXAFS, XPS, in situ FTIR results, we consider the IR bands centered at 2070–2090 cm−1 to represent CO molecules adsorbed on Pt clusters. Furthermore, we investigate the active species in CO oxidation catalyzed with different Pt loading contents. Electron Paramagnetic Resonance (EPR) analysis has been performed to detect the unpaired electron originating from the oxygen vacancies in the PSC Nb2O5 monoliths. The number of oxygen vacancies present in PSC Nb2O5 monoliths can be quantitatively compared based on the variations in the intensity of the EPR spectrum. As shown in Fig. 4g, the spectra of PSC Nb2O5 monoliths loaded with different Pt loading contents exhibit a sharp peak with a g-factor of 2.002, corresponding to the paramagnetic signal of surface oxygen vacancies associated with the superoxide ion (O2−) generated by the adsorption of atmospheric oxygen onto the surface. Notably, the peak intensity increases with increasing the Pt loading content and finally reaches the highest peak intensity for PSC Nb2O5 monoliths with the 3.35 wt% of Pt loading content. In addition, the EPR data provide direct spectroscopic evidence for the presence of oxygen vacancies in Pt-loaded PSC Nb2O5 monoliths, aligning with the H2-temperature-programmed reduction (H2-TPR) results as shown in Supplementary Fig. 17. The activation energy for CO oxidation is approximately 46–56 kJ mol−1 for Pt-loaded PSC Nb2O5 monoliths as shown in Fig. 4h. The HAADF-STEM image in Supplementary Fig. 18 shows the similar sizes of Pt clusters, further validating the thermal stability of interfacial system in the PSC Nb2O5 monoliths.
In the catalytic oxidation of CO with air, the reaction mechanism follows the Mars-van Krevelen (MvK) mechanism33,34,35. In the process, the CO adsorbs onto the Pt clusters in interfacial system with the presence of oxygen vacancies in the PSC Nb2O5 monoliths, and the oxygen from the reactive gas enters the oxygen vacancies to form lattice oxygen. As the temperature increases, the lattice oxygen linked to Pt in interfacial system becomes effectively activated. Subsequently, the CO reacts with the lattice oxygen to form CO2 which is then desorbed, allowing oxygen from the gas to recharge the lattice to form lattice oxygen. Further, the transition state for the oxidation of CO at the Pt/Nb2O5 interface is determined by the calculation of density functional theory (DFT). The potential energy diagram is shown in Fig. 4i. The adsorption energy of CO on the Pt/Nb2O5 interface structure is approximately 1.59 eV, while the reaction energy barrier for the formation of CO2 by the combination of the lattice oxygen connected to the Pt clusters and CO is approximately 0.72 eV. This indicates that the reaction between CO and the lattice oxygen at this interface has a high possibility. We further conduct the XPS and Raman tests after the long-term reaction to validate the stability of the Pt-loaded PSC Nb2O5 monoliths as shown in Supplementary Fig. 19, 20. We observe that the Pt-loaded PSC Nb2O5 monoliths generally remain unchanged before and after the reaction test as seen in Supplementary Fig. 19, thus indicating the outstanding stability originating from the structural feature of single crystallinity and porosity in a monolith. In Supplementary Fig. 20, the peaks at 244 and 321 cm−1 in the Raman spectra belong to the bending mode of the Nb-O-Nb bond in the orthorhombic T-Nb2O5 structure, whereas the broad band at 675 cm−1 is due to the coexistence of two different types of niobium coordination (NbO6 and NbO7) in the crystal structure36. In Supplementary Fig. 21, We evaluate the long-term stability assessment under high-temperature conditions of 110 °C combined with higher WHSV. First, we evaluated the CO conversion rate of the catalyst at different WHSV values by increasing the WHSV at 110 °C. As shown in Supplementary Fig. 21a, the CO conversion rate decreased from 100% to 83.83% as the WHSV increased from 30,000 mL g-1 h-1 to 60,000 mL g-1 h-1 at 110 °C. Subsequently, in Supplementary Fig. 21b, the long-term stability of Pt/PSC Nb2O5 was tested for 185 h under conditions of 110 °C and 60,000 mL g-1 h-1. The conversion rate decreased by 9%, ultimately reaching 76.3%. Pt/PSC Nb₂O₅ catalysts exhibit excellent durability.
In conclusion, we grow PSC Nb2O5 monoliths at the centimeter scale while we introduce oxygen vacancies in lattice by reduction and then stabilize the oxygen vacancies in lattice by the ordered structure. We then deposit Pt clusters at the surface in the porous microstructures to construct an interfacial system and we show the significantly enhanced surface oxygen exchange coefficient that contributes to the effective activation of lattice oxygen. We show the remarkably improved catalytic activity in the Pt cluster-decorated interfacial system in the PSC Nb2O5 monoliths. We achieve the complete conversion of CO at 110 °C and 30,000 mL g-1 h-1 conditions, with no visible performance degradation being observed even after a continuous operation for ~500 h. We believe that the current work will provide a new line of thought for the preparation of novel catalysts enriched with oxygen vacancies at the surface for enhancing the catalytic activity as well as the stability.
Methods
Synthesis
By using molten salt method, the appropriate amount of Nb2O5 and Na2CO3 and excess WO3 or B2O3 are mixed in a certain ratio in a platinum crucible, and finally placed in a furnace with the temperature set above 1260 °C. During the growth process, the temperature is raised and kept at a constant temperature for 10 h. The powder in the crucible is being observed until it forms a molten state, and the control temperature is lowered to 960 °C for 2 h. At this time, a transparent crystal surface could be clearly seen in the crucible, and then it starts to crystallize at a slow rate of 2 °C h−1. Then the temperature is cooled to room temperature at 10 °C/min to obtain NaNbO3 crystals. NaNbO3 crystals are used as precursors for the growth of PSC Nb4N5 monoliths using a lattice reconstruction strategy. The NaNbO3 crystals oriented with (001) and (010) planes are placed in ammonia gas at a flow rate of 200–350 ml min−1 at a pressure of 200–760 Torr at 800–950 °C for 50–100 h to obtain PSC Nb4N5 monoliths. We then grow PSC Nb2O5 monoliths by converting PSC Nb4N5 monoliths using a lattice reconstruction strategy. The temperature of 600 °C and a pressure of 200–760 Torr are required for the gradual oxidation in an O2/Ar mixture at a flow rate of 60–500 ml min−1 for 0.5–6 h to obtain grow PSC Nb2O5 monoliths. We use impregnation method to load Pt in PSC Nb2O5 monoliths. We use H2PtCl6 as the Pt precursor and prepare solutions with different concentrations of H2PtCl6 for Pt loadings of 0.1 wt%, 0.3 wt%, 0.5 wt%, 0.7 wt%, 1 wt% and 4 wt%, respectively.
Characterization
We use a field emission scanning electron microscope (FE-SEM) (SU-8010) to observe the microstructure of the samples. The facet orientation is characterized by X-ray diffractometry (XRD, Cu-Kα, Mniflex 600). The elemental content is analyzed using ICP-AES with utlima-2 (France. JY). Raman spectra are obtained by Labram HR evolution Raman spectroscopy (Horiba Jobin Yvon). The lattice structure is analyzed using Cs-TEM (FEI Titan3 G2 60–300). The chemical state of the elements is determined by XPS (ESCALAB 250Xi). XAFS measurements are carried out on the 1W1B beam line of the Beijing Synchrotron Radiation Facility (BSRF). In situ EXAFS is tested at the Shanghai Synchrotron Radiation Facility (SSRF) BL05U beam line. High sensitivity low energy ion scattering spectroscopy (HS-LEISS, Qtac100, ION-TOF) is used to examine the atomic termination layers of the samples. Diffuse Reflectance Infrared Fourier Transform Spectroscopy (DRIFTS) measurements are performed on an infrared spectrometer (Bruker VERTEX 70). Surface oxygen exchange coefficient tests are performed using the electrical conductivity relaxation (ECR) method. EPR spectra are acquired on a spectrometer (Bruker Biospin GMBH E500 10/12) to identify oxygen vacancies in the interfacial system.
Catalytic test
The catalytic tests are carried out at atmospheric pressure in a tubular reaction system with a flow-through quartz micro-reactor with an inner diameter of 5 mm. PSC Nb2O5 monoliths with a diameter of less than 5 mm and a thickness of less than 1 mm is loaded repeatedly as a catalyst and then preheated at 300 °C for 3 h in a 5% H2/Ar air stream. The reaction products are analyzed using an on-line gas chromatograph (GC) (Shimadzu GC-2014) equipped with FID and TCD detectors and a 30 m long CP-poraplot Q packed column. The reaction gas, consisting of 1% CO and 99% air, is introduced into the reactor at a flow rate of 50 sccm. In situ EXAFS experiments use the same gas as the catalytic reaction, and the gas flow rate is 20 sccm.
Theoretical calculations
Density Functional Theory (DFT) calculations are performed using the Vienna Ab Initio Simulation Software Package (VASP 6.3.2) code37,38. The generalized gradient approach (GGA) and Perdew–Burke–Ernzerhof (PBE) functional are used to describe exchange and correlation. Projected Augmented Wave (PAW) method is used to describe the interaction between nuclear and valence electrons. The transition states are searched using the climbing image nudged elastic band method (CI-NEB) and confirmed by the vibrational frequency analysis. The plane wave cut-off is set to 520 eV. The energies and residual forces are converged to 10−5 eV and 0.02 eV Å−1 during electron and geometry optimization. The optimized lattice parameters of Nb2O5 crystal are a = 6.175 Å, b = 29.175 Å, c = 3.93 Å with a 5 × 1 × 8 k-point grid. A p (2 × 1) superstructure of surface is used for the slab models. A supercell surface model of Nb2O5 decorated with a tetrahedral Pt4 cluster was constructed to investigate the mechanism of CO oxidation. A 20 Å vacuum area is used to avoid interaction between the plate and its repeated images for these models. A 2 × 1 × 1 k-point grid is used for the models for sampling in the Brillouin zone. The Gibbs free energy change (ΔG) at T = 383 K is calculated as ΔG = ΔE + ΔZPE-TΔS, where ΔE is the total energy difference obtained from DFT calculations, ΔZPE represents the zero-point energy correction derived from vibrational frequency analysis, and ΔS is the entropy change.
Data availability
The data that support the findings of this study are available from the corresponding author upon request. All reported data are included in the paper and supplementary materials. The Source Data can be downloaded from: https://pan.baidu.com/s/1HpEju49RJZMBPVG-xrR5sQ(Code: uexa).
References
Li, W. & Xie, K. Porous single crystals at the macroscale: From growth to application. Acc. Chem. Res 56, 374–384 (2023).
Ye, L., Li, H. & Xie, K. Sustainable ammonia production enabled by membrane reactor. Nat. Sustain. 5, 787–794 (2022).
Su, K. et al. Nb2O5-based photocatalysts. Adv. Sci. 8, 2003156 (2021).
Du, Y. et al. Post-synthesis deposition of mesoporous niobic acid with improved thermal stability by kinetically controlled sol–gel overcoating. J. Mater. Chem. A 7, 23803–23811 (2019).
Xie, Y. et al. Nanocrystalline high-dimensional Nb2O5 for efficient electroreductive dicarboxylation of CO2 with cycloalkane. ACS Catal. 14, 18624–18632 (2024).
Qi, L. et al. Ultra-efficient electrooxidation of ethylene glycol enable by Pd-loaded Fe-doped Nb2O5 with abundant oxygen vacancies. Chem. Eng. J. 475 (2023).
Choi, H. et al. Influence of lattice oxygen on the catalytic activity of blue titania supported Pt catalyst for CO oxidation. Catal. Sci. Technol. 11, 1698–1708 (2021).
Li, Y. et al. Engineering cobalt oxide with coexisting cobalt defects and oxygen vacancies for enhanced catalytic oxidation of toluene. ACS Catal. 12, 4906–4917 (2022).
Zheng, Y. et al. Oxygen vacancies in a catalyst for VOCs oxidation: synthesis, characterization, and catalytic effects. J. Mater. Chem. A 10, 14171–14186 (2022).
Qin, H. et al. Efficient catalyzed oxidation of CO by highly dispersed Co3O4 anchored in oxygen vacancies of Ta2O5. Mol. Catal. 546 (2023).
Tran, S. et al. Defective Nb2O5-supported Pt catalysts for CO oxidation: promoting catalytic activity via oxygen vacancy engineering. J. Catal. 375, 124–134 (2019).
Sankar, M. et al. Role of the support in gold-containing nanoparticles as heterogeneous catalysts. Chem. Rev. 120, 3890–3938 (2020).
Cheng, F., Zhang, J. & Xie, K. In situ observation of porosity formation in porous single-crystalline TiO2 monolith for enhanced and stable catalytic CO oxidation. Angew. Chem. Int Ed. 62, e202300480 (2023).
Xiao, Y. & Xie, K. Active exsolved metal-oxide interfaces in porous single-crystalline ceria monoliths for efficient and durable CH4/CO2 reforming. Angew. Chem. Int. Ed. 61, e202113079 (2022).
Li, W. & Xie, K. Nanoporous single-crystalline oxide catalysts for preferential oxidation of CO in H2 with an ultra-wide temperature window. Angew. Chem. Int. Ed. 61, e202212489 (2022).
Lin, G., Li, H. & Xie, K. Identifying and engineering active sites at the surface of porous single-crystalline oxide monoliths to enhance catalytic activity and stability. CCS Chem. 4, 1441–1451 (2022).
Zhu, Y. et al. Boosting oxygen evolution reaction by creating both metal ion and lattice-oxygen active sites in a complex oxide. Adv. Mater. 32, e1905025 (2020).
Li, X. et al. Engineering active surface oxygen sites of cubic perovskite cobalt oxides toward catalytic oxidation reactions. ACS Catal. 13, 6338–6350 (2023).
Luo, X. et al. Fe-S dually modulated adsorbate evolution and lattice oxygen compatible mechanism for water oxidation. Nat. Commun. 15, 8293 (2024).
Song, S. et al. A facile way to improve Pt atom efficiency for CO oxidation at low temperature: modification by transition metal oxides. ACS Catal. 9, 6177–6187 (2019).
Kersell, H. et al. CO oxidation mechanisms on CoOx-Pt thin films. J. Am. Chem. Soc. 142, 8312–8322 (2020).
Vincent, J. L. & Crozier, P. A. Atomic level fluxional behavior and activity of CeO2-supported Pt catalysts for CO oxidation. Nat. Commun. 12, 5789 (2021).
Wang, B. et al. Heterostructure-strengthened metal-support interaction of single-atom Pd catalysts enabling efficient oxygen activation for CO and VOC oxidation. Appl. Catal. B 332 (2023).
Wei, D. et al. In situ Raman observation of oxygen activation and reaction at platinum-ceria interfaces during CO oxidation. J. Am. Chem. Soc. 143, 15635–15643 (2021).
Li, Z. et al. Reactive metal–support interactions at moderate temperature in two-dimensional niobium-carbide-supported platinum catalysts. Nat. Catal. 1, 349–355 (2018).
Parastaev, A. et al. Breaking structure sensitivity in CO2 hydrogenation by tuning metal–oxide interfaces in supported cobalt nanoparticles. Nat. Catal. 5, 1051–1060 (2022).
Lin, G., Li, H. & Xie, K. Twisted surfaces in porous single crystals to deliver enhanced catalytic activity and stability. Angew. Chem. Int. Ed. 59, 16440–16444 (2020).
Zhang, Y. et al. Oxygen-vacancy-boosted visible light driven photocatalytic oxidative dehydrogenation of saturated N-heterocycles over Nb2O5 nanorods. Appl. Catal. B 316, 121622 (2022).
Lou, Y. et al. Metal-support interaction for heterogeneous catalysis: from nanoparticles to single atoms. Mater. Today Nano 12, 100093 (2020).
Nan, B. et al. Unique structure of active platinum-bismuth site for oxidation of carbon monoxide. Nat. Commun. 12, 3342 (2021).
Tang, X. et al. Significance of epitaxial growth of PtO2 on rutile TiO2 for Pt/TiO2 catalysts. J. Am. Chem. Soc. 146, 3764–3772 (2024).
Skrodczky, K. et al. Niobium pentoxide nanomaterials with distorted structures as efficient acid catalysts. Commun. Chem. 2, 129 (2019).
Zhang, J. et al. Advances in CO catalytic oxidation on typical noble metal catalysts: Mechanism, performance and optimization. Chem. Eng. J. 495, 153523 (2024).
Kim, H. Y., Lee, H. M. & Henkelman, G. CO oxidation mechanism on CeO2-supported Au nanoparticles. J. Am. Chem. Soc. 134, 1560–1570 (2012).
Widmann, D. et al. How temperature affects the mechanism of CO oxidation on Au/TiO2: A combined EPR and TAP reactor study of the reactive removal of TiO2 surface lattice oxygen in Au/TiO2 by CO. ACS Catal. 6, 5005–5011 (2016).
Li, T. et al. A niobium oxide with a shear structure and planar defects for high-power lithium ion batteries. Energy Environ. Sci. 15, 254–264 (2022).
Kresse, G. & Furthmiiller, J. Efficiency of ab-initio total energy calculations for metals and semiconductors using a plane-wave basis set. Comp. Mater. Sci. 6, 15–50 (1996).
Perdew, J. P., Burke, K. & Ernzerhof, M. Generalized gradient approximation made simple. Phys. Rev. Lett. 77, 3865–3868 (1996).
Acknowledgements
The authors acknowledge the Natural Science Foundation of China (22325506 to K.X. and 22379147 to L.Y.), the Shanghai Science and Technology Innovation Action Plan (24TS1414400 to K.X.) and the Shanghai Jiao Tong University 2030 Initiative (2030B24 to K.X.), the Natural Science Foundation of Fujian Province (2024J09052 to L.Y.) and the start-up funding from Fujian Normal University (Y072R048K13 to L.Y.) for funding this work. We thank beamline BL11B, BL05U at SSRF, as well as 1W1B at BSRF for the beam time allocation and technical assistance.
Author information
Authors and Affiliations
Contributions
L. Ye and K. Xie designed and supervised the project. X. Gao and G. Ma conducted the experimental work. C. Luo carried out theoretical calculations. All authors contributed to data analysis and gave approval to the final version of the manuscript. Correspondence and requests for materials should be addressed to L. Ye or K. Xie (ltye@fjnu.edu.cn; xiekui@sjtu.edu.cn).
Corresponding authors
Ethics declarations
Competing interests
The authors declare no competing interests.
Peer review
Peer review information
Nature Communications thanks Yun Guo and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. A peer review file is available.
Additional information
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary information
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.
About this article
Cite this article
Gao, X., Ma, G., Luo, C. et al. Boosted catalytic activity by metal-oxide interface in porous single-crystalline niobium oxide monoliths. Nat Commun 17, 1365 (2026). https://doi.org/10.1038/s41467-025-68117-1
Received:
Accepted:
Published:
Version of record:
DOI: https://doi.org/10.1038/s41467-025-68117-1
