Synthesis and characterization
MxV2O5·nH2O (M = Mg, Ca, Sr) was prepared by a simple reducing hydrothermal method38. Using commercially available V2O5 powder as precursor, then H2O2 was used as a reducing agent to reconstruct the crystal structure of V2O5 for the intercalation of different alkaline-earth metal ions by reacting with MgCl2·6H2O, CaCl2, SrCl2·6H2O, respectively. In addition, the V2O5 nanobelt without intercalation as control sample was also as control samples prepared by same methods without adding alkaline-earth metal ions to investigate the effect of M2+ intercalation on the structure and properties of V2O5. Scanning transmission microscope (SEM) and transmission electron microscopy (TEM) images showed that V2O5 powder were irregular particle with 1–3 μm (Supplementary Fig. 1A, 1 C). The V2O5 nanobelt exhibits unique 1D structure with length of 1 to 5 μm and width of 10 to 100 nm (Supplementary Fig. 1B, D). The lengths of MxV2O5·nH2O are between 1 to 10 μm and the widths are between 200 to 450 nm (Fig. 2A, B, Supplementary Fig. 2, 3). In addition, atomic force microscope (AFM) images show that the thicknesses of most MxV2O5·nH2O materials are in the range of 20–60 nm (Supplementary Fig. 4). Due to their large aspect ratio and nanoscale thickness, the MxV2O5·nH2O materials are also defined as MxV2O5·nH2O nanobelts. By high resolution transmission electron microscopy (HRTEM) analysis, the identical lattice spacing of MgxV2O5·nH2O, CaxV2O5·nH2O and SrxV2O5·nH2O nanobelts is 0.35 nm, corresponding to (100) lattice (Fig. 2C). It is worth noting that lattice disorder and mismatching can be observed in the HRTEM images of three kinds of MxV2O5·nH2O (Fig. 2D), which may come from the surface oxygen vacancies (OV) introduced by M2+ intercalation. The concentration and formation of vacancies will be discussed in the subsequent experiments and characterization. The insets in Fig. 2C shows the selected area electron diffraction (SAED) images of MxV2O5·nH2O. The uniform and well-ordered space lattices suggest the single-crystal diffraction pattern. Energy-dispersive X-ray spectroscopy (EDX) showed the existence and uniform distribution of M, V, and O in MxV2O5·nH2O, and no element segregation was found when compared with HRTEM images (Fig. 2E). The chemical composition of MxV2O5·nH2O nanobelts were also investigated by inductively coupled plasma mass spectrometry (ICP-MS) and thermogravimetric analysis (TGA), and the results are shown in Supplementary Table 1 and Supplementary Fig. 5. The molecular formulas of the three kinds of MxV2O5·nH2O are Mg0.22V2O5·0.75H2O (MgVO), Ca0.30V2O5·0.83H2O (CaVO), Sr0.34V2O5·0.72H2O (SrVO), respectively.Fig. 2: Synthesis and characterization of MxV2O5·nH2O nanobelts.A SEM images, B TEM images, C HRTEM images (the inserts are SAED images) with clear lattice fringes and lattice disorder D, EDX-mapping (E) of MgVO (a), CaVO (b) and SrVO (c). F XRD patterns of MgVO, CaVO, SrVO, V2O5 nanobelt and V2O5 powder. G Crystal structure of α-V2O5 (a) and MxV2O5·nH2O (b). Representative images are shown from three independent experiments with similar results (A-E).The crystalline phases of V2O5 powder, V2O5 nanobelt, and MxV2O5·nH2O nanobelts were characterized by X-ray diffraction (XRD) to further verify the preparation of V2O5 nanobelt and MxV2O5·nH2O nanobelts. As shown in Fig. 2F, the XRD pattern of V2O5 powder matches the standard orthorhombic V2O5 (α-V2O5, JCPDS no. 41-1426, space group Pmmn). α-V2O5 has a typical two-dimensional (2D) layered structure, which facilitates metal ion insertion between layers (Fig. 2G). V2O5 nanobelt has similar crystal structure with V2O5 powder without crystal phase change. The XRD pattern of CaVO is well matched with the monoclinic phase Ca0.24V2O5·H2O (JCPDS no. 01-088-0579), which is a kind of layered material containing Ca2+ and coordination water. Its cell parameters are as follows: a = 11.68 Å, b = 3.65 Å, c = 10.999 Å, β = 105.41°. No standard cards matching the XRD patterns of MgVO and SrVO were found, but the diffraction peaks corresponding to (001), (003), (004) and (005) crystal planes in the XRD patterns of the three kinds of MxV2O5·nH2O were similar, indicating that they have similar crystal structures. In particular, compared with V2O5 powder and V2O5 nanobelt, sharp diffraction peaks before 2θ = 10° appeared in MxV2O5·nH2O (Fig. 2F, Supplementary Fig. 6). The diffraction peaks near 2θ = 10° are generally attributed to the (001) crystal plane, which indicate that the crystal displays lamellar ordering as dominated by the pronounced (001) reflections39. With the intercalation of M2+ and H2O in MxV2O5·nH2O, the lattice spacing of (001) increases, causing the diffraction peak of (001) shift to a smaller angle40. By substituting the above diffraction peaks into Bragg equation 2dsinθ = nλ, the crystal plane spacing of MgVO, CaVO and SrVO can be calculated as 10.54 Å, 10.66 Å, 10.23 Å, respectively. Compared with V2O5 (4.37 Å)41, the larger crystal plane spacing further indicated the successful insertion of Mg2+, Ca2+, Sr2+. The corresponding structure diagrams according to the XRD patterns and related crystal parameters are drawn in Fig. 2G. It can be seen from the schematic diagram that the α-V2O5 layer is composed of VO5 square pyramids. However, square pyramids are connected up and down by sharing apexes point. Thus, if oxygen from adjacent square pyramids is included, the vanadium coordination geometry can be described as a distorted octahedra belonging to the saturated coordination42. MxV2O5·nH2O crystallized in the monoclinic space group C2/m. When α-V2O5 is converted to monoclinic MxV2O5·nH2O, the V-O-V bilayers are separated into two adjacent V-O-V monolayers due to the elongated V-O distances. It has been shown that this is caused by the pre-intercalation of hydrated metal ions and free water molecular chains. Each V-O-V monolayer in MxV2O5·nH2O is composed of {VaO5} tetragonal pyramids and {VbO6} octahedrons, extending infinitely in the plane (Fig. 2G). In the interlayer of V-O-V bilayers, M2+ coordinates with four water molecules to form planar square geometries, sharing edges with each other in a 1D chain. It can be seen that some V centers and M centers in MxV2O5·nH2O show unsaturated coordination, indicating that there may be abundant oxygen vacancies in MxV2O5·nH2O39,43. V2O5 nanobelt has no obvious crystal structure change compared with V2O5 powder. This suggests that metal intercalation may be able to effectively induce crystal structure change and oxygen vacancies generation, which is also consistent with related HRTEM characterization. It has been reported that transition metal center with unsaturated coordination have high catalytic activity, and abundant oxygen vacancies can provide adsorption sites for catalytic reactions44,45, indicating that metal ion intercalation may bring additional catalytic activity to MxV2O5·nH2O.To further understand the elemental composition and vacancies information of the prepared materials, X-ray photoelectron spectroscopy (XPS) analysis (Supplementary Fig. 7) was performed on V2O5 powder, V2O5 nanobelt and MxV2O5·nH2O nanobelts. Compared with V2O5 powder and V2O5 nanobelt, MxV2O5·nH2O nanobelts show new peaks at 1304.2 eV, 345.7 eV, and 351.0 eV, which corresponding to Mg 1 s, Ca 2p3/2, Ca 2p1/2 and Sr 3d3/2/3d5/2, respectively. This further confirmed the successful intercalation of M2+ in MxV2O5·nH2O nanobelts. The five materials all show two similar characteristic peaks between 512–528 eV, corresponding to the spin orbit peaks of V 2p3/2 and V 2p1/2 respectively. Compared with V2O5 powder, the V 2p3/2 and V 2p1/2 of MxV2O5·nH2O nanobelts show prominent asymmetry, indicating that V exists in a larger V4+/(V4+ + V5+) ratio in MxV2O5·nH2O nanobelts (Fig. 3A). The deconvolution of the peaks show that the proportion of V4+/(V4+ + V5+) in MxV2O5·nH2O nanobelts reach 35.7% 34.6% and 31.5%, respectively, significantly higher than 18.6% in V2O5 powder, indicating that more V5+ are reduced to V4+ during the generation of MxV2O5·nH2O nanobelts (Fig. 3C). However, the V4+/ (V4+ + V5+) ratio in V2O5 nanobelt with the same dose of reducing agent added in the preparation process is only 21.3%, meaning that the insertion of M2+ is conducive to the existence of high V4+/(V4+ + V5+) ratio. The relative content of oxygen vacancies in V2O5 nanobelt and MxV2O5·nH2O nanobelts were further analyzed based on O 1 s high resolution spectra. As shown in Fig. 3B, the O 1 s spectrum can be fitted to three peaks of 529.5 eV, 530.5 eV, and 532 eV, which belong to the oxygen vacancies and surface-adsorbed oxygen (OS) and lattice oxygen (OL), respectively. As expected, the relative content of OV in MxV2O5·nH2O nanobelts reach 22.76%, 17.96%, and 17.65% with the increase of V4+/(V4+ + V5+) ratio respectively, which higher than V2O5 powder (5.34%) and V2O5 nanobelt (11.2%) (Fig. 3C). The reconstruction of the structure and the generation of OV are further confirmed by the use of electron spin resonance (ESR) spectra. As shown in Fig. 3D, MxV2O5·nH2O nanobelts transmit an intense resonance signal at g = 1.97, corresponding to the captured electrons in OV, whereas only a weak signal is detected in V2O5 powder and V2O5 nanobelt. The results of ESR are consistent with those of XPS, and verified with HRTEM and XRD. In order to further understand the bonding structure of the prepared MxV2O5·nH2O nanobelts, Fourier transform infrared (FT-IR) spectroscopy and Raman spectroscopy were used to analyze them. The attributions of major characteristic peaks are shown in Fig. 3E, F, which is consistent with relevant reports46.Fig. 3: Characterization of MxV2O5·nH2O nanobelts.A XPS spectra of V 2p and B O 1 s of, V2O5 powder, V2O5 nanobelt, MgVO, CaVO, and SrVO. C Quantitative analysis of the V4+/(V4+ + V5+) ratio and relative content of OV of V2O5 powder, V2O5 nanobelt, MgVO, CaVO, and SrVO. D The analysis of OV of V2O5 powder, V2O5 nanobelt, MgVO, CaVO, and SrVO by ESR spectroscopy. E FTIR spectra (1007 cm−1: symmetric stretching vibration of V = O; 821 cm−1: stretching vibration of V-O-V; 1610 cm−1: vibration of O-H) and (F) Raman spectra of V2O5 powder, V2O5 nanobelt, MgVO, CaVO, and SrVO. (994 cm−1: symmetric stretching vibration of V = O; 700 and 404 cm−1: stretching vibration of V-O; 284 cm−1: vibration of V = O).Highly specific and efficient enzyme-mimicking activity of MxV2O5·nH2O nanobeltsThe MxV2O5·nH2O nanobelts prepared in this paper are expected to present excellent catalytic performance due to the unsaturated coordination of vanadium and abundant oxygen vacancies caused by M2+ intercalation. The peroxidase-mimicking activity of MxV2O5 nanobelts were studied by using 3, 3’, 5, 5’-tetramethylbenzidine (TMB), 2, 2’-azino-bis (3-ethylbenzothiazoline-6-sulfonic acid) (ABTS) and o-Phenylenediamine (OPD) as substrates for typical color reaction in the presence of H2O2. As shown in Fig. 4A, MxV2O5·nH2O nanobelts can effectively catalyze the oxidation of TMB in the presence of H2O2, showing obvious peroxidase-mimicking activity. Experiments also show that MxV2O5·nH2O nanobelts exhibit optimal peroxidase-mimicking catalytic activity under weak acidic environment (pH=5.0) and near room temperature (10–80 °C) (Supplementary Fig. 8). The MxV2O5·nH2O nanobelts maintain stable crystal structures and morphologies within 15 days at pH=5.0 (Supplementary Fig. 9, 10). Correspondingly, V2O5 powder show no detectable peroxidase-mimicking activity, while V2O5 nanobelt without M2+ intercalation show significantly lower catalytic activity than MxV2O5·nH2O nanobelts under the same conditions. These results indicate that the activated peroxidase-mimicking activity of MxV2O5·nH2O nanobelts may be related to the composition and structure changes after M2+ intercalation. The above results were also confirmed when ABTS and OPD were used as substrates (Supplementary Fig. 11). Meanwhile, cyclic voltammetry was used to verify the catalytic activity of MxV2O5·nH2O on the substrate26,47. As shown in Supplementary Fig. 12, 13, when H2O2 and TMB are added to the electrolyte, the currents corresponding to the redox of V with different valence states are significantly reduced, showing obvious catalytic responses. In the absence of H2O2, MxV2O5·nH2O nanobelts can not effectively catalyze the oxidation of TMB/ABTS/OPD, indicating that it has almost no oxidase-mimicking activity (Supplementary Fig. 14). In addition, the dissolved oxygen test and the nitrogen blue tetrazole (NBT) colorimetry showed that MxV2O5·nH2O nanobelts show no obvious catalase-mimicking activity and superoxide dismutase-mimicking activity (Supplementary Fig. 14). In addition, previous studies have shown that some vanadium-based nanozymes also demonstrate haloperoxidase and glutathione peroxidase activities48,49. Therefore, we detected the haloperoxidase-mimicking activity and glutathione peroxidase-mimicking activity of MxV2O5·nH2O. As shown in Supplementary Fig. 15 and 16, MxV2O5·nH2O show glutathione-mimicking peroxidase activity, and no significant haloperoxidase-mimicking activity was detected.Fig. 4: Detection of enzyme-mimicking activity and steady-state kinetics of MxV2O5·nH2O.A UV-vis absorption spectra of oxTMB in the presence of H2O2 with MxV2O5·nH2O, V2O5 powder, V2O5 nanobelt, respectively. B The generation of ·OH testified by ESR spectra of MxV2O5·nH2O, V2O5 powder, V2O5 nanobelt in the presence of H2O2. C Absorbance of oxTMB in the presence of H2O2 and MxV2O5·nH2O/V2O5 nanobelt after adding isopropanol. n  =  3 experimental replicates. Data are presented as mean values  ±  SD. The data were analyzed by using a one-sided unpaired t-test. ****P < 0.0001. D Time-dependent absorbance of oxTMB (a), reaction rate (b) with different concentrations of TMB and fixed concentrations of MgVO/H2O2, and the corresponding double reciprocal (Lineweaver-Burk) plots (inset). E Time-dependent absorbance of oxTMB (a), reaction rate (b) with different concentrations of H2O2 and fixed concentrations of MgVO/TMB, and the corresponding double reciprocal (Lineweaver-Burk) plots (inset). n  =  3 experimental replicates (D, E (b)). Data are presented as mean values  ±  SD (D, E (b)). Vmax and Km values of MxV2O5·nH2O/V2O5 nanobelt towards to TMB (F) and (G) H2O2. H The comparison of the TON and Vmax values of MxV2O5·nH2O with other recently reported catalysts. I The comparison of the relative catalytic activity of MxV2O5·nH2O and synthesized peroxidase mimics. n  =  3 experimental replicates. Data are presented as mean values  ±  SD. Data were analyzed by one-way ANOVA with Turkey’s multiple comparisons test, ns represented no statistical difference, ****P < 0.0001.According to the peroxidase catalytic mechanism, the ability of MxV2O5·nH2O to catalyze the decomposition of H2O2 to produce ·OH was verified by ESR spectra. Due to the short life and high chemical activity of ·OH, 5, 5-dimethyl-1-pyrrolin-n-oxide (DMPO) was used as ·OH radical catcher to evaluate the production of ·OH. As shown in Fig. 4B, in the presence of H2O2, the ESR spectra of MxV2O5·nH2O show an obvious characteristic spectroscopy of 1:2:2:1 intensity for DMPO/·OH adducts, which proved the generation of ·OH in the MxV2O5·nH2O/H2O2 system. To further verify the generation of ·OH, isopropyl alcohol as ·OH scavenger was added to the MxV2O5·nH2O/H2O2/TMB catalytic reaction system, the absorbance of oxTMB in the reaction system decreased significantly (Fig. 4C), illustrating ·OH is the main reactive species in the enzymatic reaction system of MxV2O5·nH2O.In order to evaluate the peroxidase-mimicking activity of MxV2O5·nH2O nanobelts, the steady-state kinetic analysis of MxV2O5·nH2O was carried out in proper concentration range of TMB and H2O2. Figure 4D, E and Supplementary Fig. 17–19 show the absorption spectra of catalytic oxidation of TMB by MxV2O5·nH2O at different concentrations of TMB or H2O2. The XPS analysis of MxV2O5·nH2O was performed before and after the reaction. It can be seen from Supplementary Fig. 20 that MxV2O5·nH2O remain relatively stable under catalytic conditions. According to the initial reaction rate calculated in Fig. 4D (a), 4E (a), typical Michaelis-Menten curves of TMB and H2O2 can be obtained respectively (Fig. 4D (b), 4E(b)). Next, the reciprocal of substrate concentration and initial reaction rate could be used to prepare the double reciprocal plot (Lineweaver-Burk) to obtain the maximal reaction velocity (Vmax) and Michaelis−Menten constant (Km). Among them, Vmax is an important parameter to evaluate the catalytic rate of enzymatic reaction, and Km is an important indicator of enzyme affinity to substrates. As shown in Fig. 4F, when TMB was used as substrate, Vmax of MgVO, CaVO and SrVO nanobelts are 245, 220, 150 (10−8 M·s−1), Km are 0.44, 0.54, 0.27 mM, respectively; When H2O2 was used as substrate (Fig. 4G), the Vmax of MgVO, CaVO and SrVO nanobelts are 306, 177, 148 (10−8 M·s−1), and Km are 7.91, 6.46, 10.32 mM, respectively. Comparing with natural horseradish peroxidase (HRP) and classical Fe3O4 (Supplementary Table 2), the Km of MxV2O5·nH2O for the two substrates are similar to that of HRP and Fe3O4, but the Vmax of MxV2O5·nH2O are more than one order of magnitude higher than that of HRP and Fe3O4 under their respective optimal catalytic conditions. In order to systematically evaluate its catalytic performance, we compared the Vmax and TON values (the maximum number of conversing substrates via the mole concentration of metal in the whole nanomaterials50) of MxV2O5·nH2O with currently reported classical peroxidase mimics, including Zn-N-C51, Pt NPs52, CeO253, Fe3O453, and MnO253, etc (Supplementary Table 3), the results indicate MxV2O5·nH2O display the optimal catalytic performances in comparison with metal oxides-based nanozymes, even some noble metal nanozymes and single atom nanozymes (Fig. 4H). In order to eliminate data differences caused by different test conditions, four typical peroxidase-mimicking materials were synthesized by referring to the methods in the literature (Supplementary Fig. 21, 22), and catalytic oxidation of TMB was tested under the same test conditions. The results of Fig. 4I directly confirm that MxV2O5·nH2O nanobelts presented high efficiency peroxidase-mimicking activity.The synergistic effect of redox V and redox-inert M in enzyme-mimicking catalysisTheoretical calculations and some experimental studies have showed that the catalytic reaction paths of POD-like nanozymes can be classified as two types according to dissociation mode of H2O2 adsorbate (H2O2*)54,55. The path 1 is that the dissociated H2O2* generates hydroxyl adsorbate (OH*) and hydroxyl radical (·OH), then generates H2O and oxidized substrate to complete the cycle30,51,56,57,58. Because it is similar to the well-known Fenton reaction, this mechanism is also known as the Fenton-like mechanism and is widely accepted to explain the POD-like activity of many materials. The path 2 usually does not involve the formation of ·OH. H2O2* first generates two OH*, which can directly oxidize the substrate under acidic conditions59,60. Alternatively, OH* can also be converted to O* and H2O* through a hydrogen transfer reaction, with O* oxidizing the substrate53,57,61,62. This path that the adsorbed ROS directly acting on substrates has attracted more and more attention because it is similar to the catalytic mechanism of natural peroxidase (ferryl oxo species). In view of the direct detection of ·OH produced under catalysis by ESR spectroscopy and the scavenging verification ·OH by isopropyl alcohol, the possible catalytic mechanism of MxV2O5·nH2O is proposed following path 1. Therefore, the reaction mechanism was studied from the free energy, charge density difference, and band structure through density functional theory calculation. Subsequently, combined with the changes in the structure and catalytic activity of MxV2O5·nH2O after M2+ intercalation, the synergistic effect of redox-inert M2+ with redox-active V is further discussed.Four crystal models including V2O5, MgVO, CaVO and SrVO were established to simulate V2O5 nanobelt and three kinds of MxV2O5·nH2O nanobelts. In addition, to exclude the effect of crystal structure change (V2O5 and MxV2O5·nH2O) on the catalytic process, an additional V2O5·nH2O crystal model with only H2O intercalation was also established to study the effect of M2+ in catalysis comparison with MxV2O5·nH2O. At the same time, since V2O5·nH2O have unsaturated coordination structure relative to V2O5 like MxV2O5·nH2O, the contrast between V2O5·nH2O and V2O5 can be used to separately analyze the role of oxygen vacancies caused by unsaturated coordination in catalysis. The top and side views of the above five crystal models are shown in Supplementary Fig. 23. The possible catalytic mechanism of MxV2O5·nH2O catalyzing H2O2 decomposition is shown in Fig. 5A. Taking MgVO as an example, H2O2 tends to preferentially adsorb on the active site V of MgVO, and then the activated H2O2 is uniformly dissociated into two OH*. Among them, one OH* desorbs from the adsorption position to form ·OH, and the other OH* binds with a protonated hydrogen atom to form H2O molecule. MgVO returns to its initial state after H2O desorption. According to this mechanism, the free energy changes in the reaction process of the five models was shown in Fig. 5B.Fig. 5: DFT calculations and the proposed mechanism for the peroxidase-mimicking of MxV2O5·nH2O.A The possible reaction pathway on MxV2O5·nH2O, taking MgVO for example. (Purple: V atom; Red: O atom; White: H atom; Yellow: Mg atom). B Free energy diagram of the proposed reaction pathways for V2O5, V2O5·nH2O and MxV2O5·nH2O models. C Free energy change of dissociation of H2O2 (Step II) on different models. (Blue, Ca atom; Green: Sr atom). D Free energy change of OH* desorption (Step III or Step II & III) on MxV2O5·nH2O models. E The differential charge analysis of (a) V2O5, (b) MgVO, (c) CaVO, (d) SrVO and (e) V2O5·nH2O with 2OH* models (cyan and yellow represent charge depletion and accumulation, respectively). F Calculated electronic density of states with the energies of FMOs marked. G Comparison of peroxidase-mimicking activity of MxV2O5·nH2O before and after de-intercalation. n  =  3 experimental replicates. Data are presented as mean values  ±  SD. Data were analyzed by using a one-sided unpaired t-test, **P < 0.01.Combined with the free energy profile and related literature51,60, it can be inferred that there are two key steps in this catalytic path, H2O2* dissociation and OH* desorption. For V2O5 before intercalation, the energy of H2O2* dissociation is 2.03 eV, indicating that the process is endothermic. For V2O5·nH2O after H2O intercalation, the energy of H2O2* dissociation decreased to 1.09 eV, indicating that the dissociation is promoted. This may be owed to the oxygen vacancies caused by unsaturated coordination in H2O-intercalated V2O5·nH2O. In spite of this, the H2O2* dissociation for V2O5·nH2O is still an endothermic process, and the reaction is not favorable. After Mg/Ca/Sr intercalation, the H2O2 dissociations are obviously promoted, and their free energies all become negative, that is, the process changes from endothermic to exothermic (Fig. 5B, C). Since the energy of H2O2* dissociation is determined by the difference between the adsorption energy (Ead) of OH* and H2O2*, Ead (OH*) can be considered as descriptor of the process under the premise that Ead (H2O2*) are similar. Therefore, compared with V2O5 and V2O5·nH2O, M2+ intercalation significantly enhanced the adsorption of OH*, thus promoting the dissociation of H2O2*. Further comparing the effect of different intercalation ions, CaVO showed a stronger ability than MgVO and SrVO to promote H2O2* dissociation, which is inconsistent with the order of Vmax obtained by experiments. Considering that strong OH* adsorption is beneficial to the dissociation of H2O2*, but not conducive to the desorption of OH*, the catalytic activities may also be affected by OH* desorption. According to the free energy profile, the energy of OH* desorption on MgVO is only 0.196 eV, meaning desorption is relatively simple. But the energies of CaVO and SrVO during desorption are 0.503 eV and 0.484 eV, respectively (Fig. 5B, D), which is not favorable for the dissociation of OH*. On the whole, considering the energy from H2O2* to OH* (Step II & III), the order of MgVO (0.165 eV) < CaVO (0.216 eV) < SrVO (0.471 eV) is presented (Fig. 5B, D). It is worth noting that the energy change (Step II & III) is positively correlated with the experimental Vmax, and shows the consistency with the periodic change of M. Besides the above mechanism involving releasing ·OH, there are many reports on mechanism that the OH* oxidizes substrates directly under acidic conditions59,60. In this path, H2O2* first generates two OH*, and then the reductive substrates (such as TMB) continuously providing reductive hydrogen [H] to the two adsorbed OH*. Since the presence of the substrate, it is considered to be more similar in energy to the actual catalytic process. Based on this, we calculate the free energy profile under this path. As shown in Supplementary Fig. 24, the dissociation of H2O2 is not affected after the substrate is added. Compared with V2O5 and V2O5·nH2O, the energy of H2O2 dissociation decreases significantly after Mg/Ca/Sr intercalation. This is consistent with the result in path 1 (Fig. 5B). The reaction of OH* with TMB-H+ becomes completely spontaneous due to the formation of H2O. Therefore, from the perspective of free energy change, Ead (OH*), that is, the stability of OH* on the catalyst surface, is key to determine the catalytic reaction activity in both path 1 (Fig. 5B) and path 2 (Supplementary Fig. 24). The redox-inert M2+ are not considered to be catalytic centers like redox-active V. Therefore, the synergistic function of redox-inert M2+ through regulating the electronic structure of redox-active V may be the main way to improve their enzyme-like catalytic activities. The interlayer spacing of MxV2O5·nH2O indicated that the intercalated M2+ and the host was connected by non-covalent bonds43. Therefore, the effect of intercalated M2+ on the electronic structure of the crystal is mainly polarization according to their Lewis acidity, which can usually be evaluated by the ionic potential. The Ionic potential (Φ) is defined as$$\Phi={{\mbox{Z}}}_{{\mbox{eff}}}/{\mbox{r}}=({\mbox{Z}}-{{{\rm{\sigma }}}})/{\mbox{r}}$$
(1)
where Zeff is the effective nuclear charge, σ is the screening constant, and r is the ionic radius63. Among them, the effective nuclear charge is a periodic property and depends upon the electronic shell structures although it is only qualitatively conceived and measured semiempirically64. Since Mg2+, Ca2+ and Sr2+ belong to the IIA in the periodic table of the elements, they have the same ionic charge and valence shell electron configuration, so there is little difference in their Zeff. Therefore, the ionic potential of Mg2+, Ca2+, Sr2+ is inversely proportional to their ionic radius. With the increase of ionic potential, the Sr2+, Ca2+ and Mg2+ in the interlayer of MxV2O5·nH2O showed a periodic enhancement of polarization, thus polarizing the inherent V-O bond in the crystal and making V-OH* more stable.To confirm the above theory, the differential charge density calculation was used to and analyze the electron transfer from MxV2O5·nH2O to OH* to determine the stability of OH*. As shown in Fig. 5E, there is basically no electron transfer (0.1181 and 0.0209 |e| respectively) between V and two OH* in V2O5 before intercalation, indicating that there is no stable interaction between V and OH*. After H2O intercalation (V2O5·nH2O), the electrons transferred between V and two OH* is significantly enhanced, reaching 0.3225 and 0.3176 |e| respectively. After M2+ intercalation, the number of transferred electrons increased further (0.5343/0.4325 |e | ; 0.4596/0.5301 |e | ;0.4180/0.5431 |e | ), indicating the stable bonding interaction between V and OH*. On the whole, the order of charge transfer from the V to two OH* is CaVO (0.9897 |e | ) > MgVO (0.9668 |e | ) > SrVO (0.9611 |e | ) > V2O5·nH2O (0.6401 |e | ) > V2O5 (0.1389 |e | ), where the charge transfers after M2+ intercalation are significantly greater than that of V2O5 and V2O5·nH2O, which corresponds to the energy of H2O2* dissociation. Then, the influence of Lewis acid M2+ on the intrinsic electronic structure of redox-active V is further analyzed by calculating the average bader charge of V before and after intercalation. As shown in Supplementary Fig. 25, the bader charge around V in V2O5 before intercalation is 2.1670 |e | . After H2O intercalation, the bader charge around V is 2.1730 |e | , and no significant change occurs. After M2+ intercalation, the average bader charge around V decreases to 2.1379, 2.1264 and 2.1242 |e | , respectively. Since M2+ acts directly on the coordinated O of V through electrostatic attraction, the polarization of M2+ on the V-O bond results in the decrease in electron density around V. In general, V with low electron density will be more inclined to produce stable V-OH*. Thus, the Lewis acid M2+ promotes the H2O2* homolysis by adjusting the electronic structure of V.In order to further understand the effect of Lewis acid intercalation on enzyme-mimicking catalytic activity, the band structure before and after intercalation were calculated and discussed combing with the POD-like catalytic redox potential65,66. The typical reaction catalyzed by peroxidase nanozyme is as follows: H2O2 + 2TMB + 2H+ = 2H2O + 2TMB+, where TMB+ is the oxidation state of TMB. The reaction can be divided into two half reactions as follows:$$\begin{array}{c}{{{{\rm{TMB}}}}}^{+}+{{{{\rm{e}}}}}^{-}={{{\rm{TMB}}}},\, {{{{\rm{\varphi }}}}}_{1}{{{\rm{;}}}}\\ {1/2{{{\rm{H}}}}}_{2}{{{{\rm{O}}}}}_{2}+{{{{\rm{H}}}}}^{+}+{{{{\rm{e}}}}}^{-}={{{{\rm{H}}}}}_{2}{{{{\rm{O}}}}}_{,} \, {{{{\rm{\varphi }}}}}_{2}\end{array}$$The reduction potential of TMB+/TMB (φ1) used is about 1.13 V, referring to a well-established value in the literature, and the standard reduction potential of H2O2/H2O (φ2) is 1.776 V67,68. As can be seen from Fig. 5F, the frontier molecular orbitals (FMO) of V2O5 before intercalation, including the valence band maximum (VBM) and the conduction band minimum (CBM), display more positive energy than φ2, indicating that electrons can be transferred from TMB to V2O5, but cannot be transferred from V2O5 to H2O2 to complete the catalytic reaction. After Lewis acid M2+ intercalation, the energies of the FMO of MxV2O5·nH2O move significantly toward the negative direction, indicating that the reduction abilities of MxV2O5·nH2O are significantly enhanced. Among them, the energies of VBM is close to or less negative than φ2, indicating that electrons can be passed from TMB to V2O5, and then to H2O2 to complete the catalysis. The H2O intercalated V2O5·nH2O shows a more negative FMO energy, and electrons cannot transfer from TMB to V2O5.According to the reported literatures, the redox-inert Ca2+ as cofactors in natural peroxidase with Lewis acidity play a key role in the catalytic performance, which is mainly manifested in the loss of Ca2+ leading to significant reduction of catalytic activity16,17. The structural similarities between MxV2O5·nH2O and natural peroxidase encouraged us to further investigate whether the redox-inert M2+ could also act as cofactors like Ca2+ of natural peroxidase to affect the peroxidase-mimicking activity of MxV2O5·nH2O. Thus, MxV2O5·nH2O were placed in heated alkaline liquid to remove the inserted M2+ to achieve deintercalation. The results of ICP-MS showed that the molar ratios of Mg/V, Ca/V, and Sr/V in MxV2O5·nH2O decreased to 0.08/1, 0.13/1, and 0.14/1 respectively (Supplementary Fig. 26, Supplementary Table 4), confirming partial removal of M2+ in MxV2O5·nH2O. The catalytic capacity of MxV2O5·nH2O for TMB oxidation before and after de-intercalation were then compared. From the result shown in Fig. 5G, the catalytic performance of MxV2O5·nH2O after de-intercalation show obvious decrement owing to the decreased concentration of M2+. The SEM and XRD results show that the morphology and crystal structure of MxV2O5·nH2O after partial de-intercalation have no obvious changes (Supplementary Fig. 27, 28). The above results showed that the redox-inert M2+ in MxV2O5·nH2O play similar roles to the Ca2+ in natural peroxidase in regulating the catalytic reaction. Benefiting from these, the synergism of redox-active V and redox-inert M2+ in MxV2O5·nH2O plays a key role in regulating the structure and function of peroxidase mimics, which is similar to the relationship between redox site and Lewis acid in natural peroxidase.Excellent antibacterial properties of MxV2O5·nH2OThe enzyme mimics are widely used in antibacterial researches, because the reactive oxygen species (ROS) catalyzed by enzyme mimics can oxidize key components of bacteria, such as cell membranes/walls or intracellular compartments69,70,71. In this work, MxV2O5·nH2O are expected to acquire additional bactericidal ability in the presence of low concentration of medical H2O2 disinfectant due to its high peroxidase-mimicking activity. To verify this, we evaluated the activity of MxV2O5·nH2O against gram-positive Staphylococcus aureus (S. aureus) and gram-negative Escherichia coli (E. coli) by standard spread plate method. As shown in Fig. 6A, B, in the presence of low concentration H2O2 (200 µM), MxV2O5·nH2O nanobelts show stronger antibacterial ability toward to S. aureus and E. coli than V2O5 powder and V2O5 nanobelt. Quantitative analysis shows that the relative survival rate of E. coli decreased to 3.63%, 5.19%, and 6.79% after MxV2O5·nH2O and H2O2 treatment, while the relative survival rate of S. aureus was 4.37%, 6.65%, and 5.56%, respectively (Fig. 6C, D). In control experiments without H2O2 or with H2O2 alone, the Luria Bertani (LB) agar plate were completely covered with bacterial colonies. This indicates that the bacteria could be inhibited only in the presence of H2O2 and MxV2O5·nH2O, that is, the antibacterial activity came from the peroxidase-mimicking activity of the materials. To further visually verify the antibacterial ability of the material, SYTO9/PI was used to stain live/dead bacteria. As shown in Fig. 6E, F, compared with other control groups, the red color of PI in the presence of both H2O2 and MxV2O5·nH2O was enhanced, indicating a significant antibacterial effect. To further verify the broad-spectrum antibacterial ability of MxV2O5·nH2O, the antibacterial test for gram-positive Bacillus subtilis (B. subtilis) and gram-negative Pseudomonas aeruginosa (P. aeruginosa) by standard spread plate method were added. As shown in Supplementary Fig. 29, MxV2O5·nH2O show significant antibacterial ability against B. subtilis and P. aeruginosa in the presence of low concentration H2O2.Fig. 6: Antibacterial properties of MxV2O5·nH2O.Photographs of bacterial colonies against E. coli (A) and S. aureus (B) with different treatments. Relative survival rate of E. coli (C) and S. aureus (D) upon different treatments determined by spread plate method. n  =  3 biologically independent samples (C, D). Data are presented as mean values  ±  SD (C, D). Data were analyzed by one-way ANOVA with Turkey’s multiple comparisons test (C, D), ns represents no statistical difference, *P < 0.01. **P < 0.001. ***P < 0.0001. ****P < 0.0001. CLSM images of E. coli (E) and S. aureus (F) co-stained with Syto9 and PI after incubating with different treatments. Scale bar: 50 μm. (a) PBS, (b) V2O5 powder, (c) V2O5 nanobelt, (d) MgVO, (e) CaVO, (f) SrVO, (g) H2O2, (h) V2O5 powder + H2O2, (i) V2O5 nanobelt + H2O2, (j) MgVO + H2O2, (k) CaVO + H2O2, (l) SrVO + H2O2. Representative images are shown from three independent experiments with similar results (E, F).Composite hydrogel with MxV2O5·nH2O as additive for infected wound healingIn addition to antibacterial properties, the banded structure of MxV2O5·nH2O are widely believed to contribute to the formation of stable composites. Meanwhile, the large size reduces cytotoxicity due to the difficulty in endocytosis (Supplementary Fig. 30). Therefore, MxV2O5·nH2O nanobelts are expected to be wound dressing additives for the healing of infected wound. Hence, we then prepared an antibacterial composite hydrogel as wound dressing for wound healing. Concretely, the gelatin solution was mixed with tannic acid solution and the hydrogels (GelTA) rapidly formed via hydrogen bond crosslinking72. During the preparation of GelTA, MxV2O5·nH2O were added to form composite hydrogels for offering wound barrier and promoting the wound healing. Fourier transform infrared (FT-IR) spectrum was used to characterize the prepared hydrogel. The attribution of major characteristic peaks is consistent with relevant reports (Supplementary Fig. 31)72. SEM images showed that freeze-dried GelTA has abundant porous structure with high interconnectivity. MxV2O5·nH2O functionalized GelTA (Mg-Gel-TA/ Ca-Gel-TA/ Sr-Gel-TA) also displayed similar porous structure, which were beneficial for exchange of substances (Supplementary Fig. 32). EDX elemental quantification shows the existence of MxV2O5·nH2O in the composite hydrogel. In order to study the peroxidase-mimicking activity of three MxV2O5·nH2O functionalized GelTA, which were placed in the mixed solution of H2O2 and TMB. The results showed that the MxV2O5·nH2O functionalized GelTA could rapidly make the mixed solution turn dark blue, the absorption peak of oxTMB at 652 nm appeared in the UV-vis absorption spectrum (Supplementary Fig. 33), indicating that the MxV2O5·nH2O functionalized GelTA exhibit peroxidase-mimicking activity, which is expected to have antibacterial ability. In order to verify the antibacterial efficacy of the MxV2O5·nH2O functionalized GelTA as wound dressings, infected wound healing experiment was carried out on mice. A round whole cortical wound with a diameter of 5.5 mm was formed on the back of each mouse (Supplementary Fig. 34), which was randomly divided into 9 groups for different treatments after 24 h of S. aureus infection: (1) PBS, (2) GelTA, (3) GelTA+H2O2, (4) Mg-GelTA, (5) Ca-GelTA, (6) Sr-GelTA, (7) Mg-GelTA+H2O2, (8) Ca-GelTA+H2O2, (9) Sr-GelTA+H2O2. The wound photos of mice at day 0, 3, 5 and 7 were respectively shown in Fig. 7B. It could be seen that the wound had crusted and no obvious infection occurred on day 3 after treatment with Mg-GelTA+ H2O2, Ca-GelTA+ H2O2, Sr-GelTA+ H2O2. On day 5, the relative wound area of mice decreased to 34.1%, 47.8% and 39.9%, which are much smaller than other treatment groups (Fig. 7C). Besides, there were no ulceration or suppuration occurred during the treatment. In contrast, the other 6 groups show slower wound healing and varying degrees of inflammation. The wound healing traces of mice treated with different treatments within 7 days were analyzed (Fig. 7D), which further indicated that the wound healing rate of group (7) (8) (9) are significantly faster than other treatment group. In addition, the bacteria from wound tissues of mice in each group were cultured on LB agar plates, and the results showed that the number of bacteria significantly decreased at the wound sites in the treatment groups (Fig. 7E, Supplementary Fig. 35). Further, Hematoxylin and eosin (H&E) staining was analyzed for the wound tissue of different groups of mice after seven days of treatment (Fig. 7F). The results show that the epidermal layers at the wound sites are more intact in the treatment groups, and only a few inflammatory cells are distributed. In the control groups, the epidermal layers at the wound sites are fragmented and many inflammatory cells gather in the wound areas. Masson staining shows that the density of collagen fibers at the wound site are higher in the treatment groups than in the control groups (Fig. 7G). To ensure its biosafety in vivo, weight changes in mice were recorded during the treatment. As shown in Supplementary Fig. 36, there are no significant changes in body weight the body weight of mice during the treatment, indicating that the mice lived well. Further, H&E staining analysis of the major organs of the mice in all experimental groups show that no obvious organ damage, abnormality or inflammation occurred in the mice (Supplementary Fig. 37). In conclusion, MxV2O5·nH2O functionalized GelTA can effectively inhibit bacterial growth and promote wound healing as a biosafe antibacterial wound dressing.Fig. 7: MxV2O5·nH2O functionalized hydrogels used as antibacterial dressings to promote wound healing.A Schematic illustration of the in vivo wound disinfection treatment with wound dressing containing MxV2O5·nH2O in a mice model. B Photographs of the infected wound on mice and (C) relative wound area of different groups at day 0, 3, 5, 7 during antibacterial treatment process. n  =  3 biologically independent samples. Data are presented as mean values  ±  SD. D Wound healing traces of mice in each group during 7 days. E Photographs of bacterial colonies separated from wound tissue with different treatments at day 5. F H&E staining and G Masson’s trichrome staining of the excised wound tissues treated by different groups. (1) PBS, (2) GelTA, (3) GelTA + H2O2, (4) Mg-GelTA, (5) Ca-GelTA, (6) Sr-GelTA, (7) Mg-GelTA + H2O2, (8) Ca-GelTA + H2O2, (9) Sr-GelTA + H2O2. Scale bar, 100 μm. Representative images are shown from three independent experiments with similar results (F, G).In summary, we have developed a series of MxV2O5·nH2O peroxidase mimics with synergistic structure of redox site and Lewis acid, which have more significant peroxidase-mimicking activity compared with other typical peroxidase mimics according to the Vmax and TON values comparison. The experimental results and theoretical calculation indicate that the prominent enzyme-mimicking catalytic activity of MxV2O5·nH2O are caused by the change of geometric and electronic structure after redox-inert M2+ intercalation. Specifically, the M2+ intercalation lead to the central redox-active V in V2O5 crystal transform into a low coordination number structure, and the resulting high concentration of oxygen vacancies can initially reduce the energy of H2O2* dissociation. More importantly, the charge redistribution around V caused by M2+ intercalation makes V more incline to connect with OH*, thus promoting the H2O2* dissociation to become completely spontaneous. This significantly enhanced the POD-mimicking activity of MxV2O5·nH2O. Interestingly, we found that the catalytic activity of MxV2O5·nH2O are periodically related to the Lewis acidity of intercalated M2+, it should be caused by the non-covalent polarization of intercalated metal ions. The redox-inert M2+ regulation of the geometrical and electronic structure of the V2O5 host highly mimics the synergistic effect of the redox site and Lewis acid in natural peroxidase. On this basis, MxV2O5·nH2O were applied to antibacterial therapy as an antibacterial additive in wound dressing. In vivo and in vitro experiments showed that MxV2O5·nH2O functionalized wound dressings displayed significant antibacterial activity in the presence of low concentration of medical hydrogen peroxide. This study not only reported a series of highly efficient peroxidase mimics with practical application potential, but also proposed and explained a strategy to activate enzyme-mimicking activity through synergetic structure of redox site and Lewis acid at the experimental and theoretical levels. At the same time, it also provides a general idea for the regulation of enzyme-mimicking catalytic activity of many other metal oxides layered compounds.
Biomimetic synergistic effect of redox site and Lewis acid for construction of efficient artificial enzyme
