Synthesis and morphology characterizationsThe synthesis of Pt dispersed on hybrid CuOx/Cu catalysts (Pt1-CuOx/Cu) is schematically illustrated in Fig. 1. Typically, the porous carbon supports originated from the alkaline hydrothermal treatment of 1,3,6-trinitropyrene. The monatomic Pt dispersed on amorphous CuOx/Cu hybrid structure was synthesized by temperature-controlled pyrolysis and subsequent in situ partial oxidation. Moreover, the control samples, including Pt single atoms or Cu nanoparticles dispersed on carbon supports (PtSA-CN, CuNP-CN), bare carbon supports (CN), etc., were also synthesized through similar treatments in addition to replacing different metal sources (see the “Methods” section for more details).Fig. 1: Schematic illustration.Schematic of the synthetic process. The Pt1-CuOx/Cu dispersed on carbon nanosheets was synthesized through sequential hydrothermal and pyrolytic processes.The overall morphologies of Pt1-CuOx/Cu and control samples were characterized by transmission electron microscopy (TEM). In Supplementary Fig. 1, the Pt1-CuOx/Cu maintains the same nanosheet morphology as the carbon supports (CN). In the power X-ray diffraction (XRD) patterns (Supplementary Fig. 2a), neither PtSA-CN nor CN support exhibits metal diffraction peaks. TEM images of PtSA-CN reveal no obvious nanoparticles, and the aberration-corrected high-angle annular dark-field scanning transmission electron microscopy (AC-HAADF-STEM) image of PtSA-CN directly confirms the single-atom phase of Pt in PtSA-CN (Supplementary Figs. 2b–d). In contrast, small nanoparticles apparently existed in the Pt1-CuOx/Cu catalyst, as identified by the high-resolution transmission electron microscopy (HRTEM) images in Fig. 2a, b. It can be seen from Fig. 2a and the size statistics results that the nanoparticles are evenly dispersed on the carbon supports with an average particle size measuring 3.57 nm. The HRTEM image in Fig. 2b and Supplementary Fig. 3 reveal that the nanoparticles possess lattice stripes inside, whose interplanar spacing was about ~0.20 nm, attributing to the Cu (111) plane. The amorphous feature at the periphery of the nanoparticles may be related to the formation of a CuOx shell layer by the partial oxidation strategies. The two distinct atomic distributions of the nanoparticles are color-marked in AC-HAADF-STEM images, which suggest the existence of CuOx/Cu shell/core structures (Fig. 2c).Fig. 2: Morphology and phase characterizations.a TEM image of Pt1-CuOx/Cu; inset is the corresponding size distribution; (b) HRTEM image of Pt1-CuOx/Cu; (c–e) AC-HAADF-STEM image (c) EDS elemental mapping (d) and line-scan analysis, scale bar: 2 nm (e), of Pt1-CuOx/Cu. f XRD results of Pt1-CuOx/Cu and the control samples.The results of energy dispersive spectroscopy (EDS) elemental mapping in Fig. 2d and line-scan analysis in Fig. 2e further reveal that the Pt atoms were evenly dispersed on the surface of the CuOx/Cu hybrid nanoparticles, and Cu is the main component of the nanoparticles. The contents of Cu and Pt in Pt1-CuOx/Cu were determined by inductively coupled plasma mass spectrometry (ICP-MS) to be 3.4 wt% and 0.65 wt%, respectively. The XRD patterns of Pt1-CuOx/Cu show rather weak diffraction peaks of metallic Cu relative to the CuNP-CN samples, and no diffraction peaks of oxide species appear, which is attributed to the amorphous CuOx species in Pt1-CuOx/Cu. Moreover, the lack of metallic Pt diffraction peaks suggests that the atomic-level doping of Pt in CuOx/Cu hybrid nanoparticles (Fig. 2f). Hence, these above characterization results clearly demonstrate that the catalysts composed of the evenly dispersed Pt on hybrid CuOx/Cu nanoparticles were successfully obtained.Electronic and local structure characterizationsThe X-ray absorption near edge structure (XANES) spectra at O K-edge were obtained to evaluate the local structure of oxygen (Fig. 3a). The characteristic peaks at ~530 and ~534 eV can be assigned to the overlapping bands between the d bands of metal atoms and the 2p orbitals of oxygen29. The appearance of the characteristic peaks in the Pt1-CuOx/Cu sample relative to carbon support suggests the M-O interactions, which are directly related to the formation of amorphous oxide layers. To elucidate the local coordination structure of Pt1-CuOx/Cu, the Fourier transformed EXAFS (FT-EXAFS) spectra at Cu K-edge and Pt L3-edge were obtained in Fig. 3b, c. As shown in Fig. 3b, the Pt1-CuOx/Cu sample displays two dominant peaks at approximately 1.5 Å and 2.30 Å, which corresponds to the Cu-O and Cu-Cu coordination of the CuO and Cu foil, respectively. It is suggested that the CuOx species are formed after the oxidation process, which results in the coexistence of CuOx and Cu nanoparticles in the Pt1-CuOx/Cu sample. In the Pt L3-edge k2-weighted FT-EXAFS spectra in Fig. 3c, the characteristic peak at the first shell of Pt1-CuOx/Cu may be attributed to Pt-O scattering according to the fact that Pt atoms are dispersed on the CuOx/Cu surface. The characteristic peak at the second coordination shell, which is distinct from the Pt-Pt characteristic peak found in Pt foil, is likely attributable to the Pt-O-M scattering. Moreover, the Pt-O-M scattering peak is different from the double characteristic peaks of Pt-O-Pt in PtO2, ruling out the presence of agglomerated Pt particles and PtO2 species. Thus, the results of FT-EXAFS spectra demonstrate that the Pt species are atomically isolated by O atoms within amorphous CuOx layers, corresponding to the phase characterization in Fig. 2. The local structure around the Pt site was also evaluated by the wavelet transforms (WT) of FT-EXAFS (Supplementary Fig. 4). The Pt1-CuOx/Cu only shows the intensity maximum at 4.2 Å−1, suggesting the monoatomic dispersion of Pt on the surface of CuOx/Cu nanoparticles30.Fig. 3: Local structure and electronic structure characterizations.a XANES spectra at O K-edge of Pt1-CuOx/Cu and CN; (b) Cu K-edge k3-weighted FT-EXAFS spectra of Pt1-CuOx/Cu, Cu foil and CuO; (c) Pt L3-edge k2-weighted FT-EXAFS spectra of Pt1-CuOx/Cu, Pt foil and PtO2; Cu 2p (d) and Pt 4 f (e) XPS spectra of Pt1-CuOx/Cu; (f) XANES spectra at Pt L3-edge of different samples.To ascertain the valence states of Cu and Pt in Pt1-CuOx/Cu, X-ray photoelectron spectroscopy (XPS) was performed. Prior to XPS spectral analysis, the binding energies of all elements in the samples were calibrated according to charge correction of the C1s spectrum (Supplementary Fig. 5). As illustrated in the Cu 2p XPS spectra results (Fig. 3d), both the peaks of Cu 2p3/2 and Cu 2p1/2 can be fitted into two peaks, representing the existence of Cu0/Cu+ and Cu2+ states, which is different from the result of CuNP-CN sample (Supplementary Fig. 6a)31,32. The associated Cu LMM spectra of Pt1-CuOx/Cu was employed to further evaluate the proportion of different valence states. In Supplementary Fig. 6b, the Cu LMM spectra of Pt1-CuOx/Cu were deconvoluted into three peaks at 564.8, 566.4 and 570.6 eV, representing the Cu0, Cu2+ and Cu+ species, respectively. It can be seen that a wide valence range (0‒2+) exists in CuOx/Cu species. Figure 3e shows the Pt 4 f XPS spectra. The binding energy of Pt 4f7/2 in PtSA/CN control sample registers at 72.6 eV, closely resembling that of Pt2+ species33. The binding energy of Pt 4f7/2 in Pt1-CuOx/Cu slightly negatively shifted by ~0.3 eV relative to that of PtSA-CN, implying that the valence state of Pt is decreased by the surrounding Cu-based supports. As shown in the XANES spectra of Pt L3-edge, the intensity of the white-line (HA) for Pt1-CuOx/Cu presents a slightly weaker than that of PtSA-CN, indicating a lower oxidation state of Pt atoms in Pt1-CuOx/Cu (Fig. 3f), which is consistent with the Pt 4 f XPS analysis. As the fact that the HA of Pt L3-edge XANES spectra corresponds to the transition from the Pt 2p3/2 core-electron to empty 5d and reflects the occupation level of the Pt 5d states, the lower oxidation state of Pt in Pt1-CuOx/Cu suggests the more occupied states of Pt 5d. Besides, the intensity of the oscillation hump (HP) was significantly attenuated for Pt1-CuOx/Cu, confirming that the local structures around Pt atom in Pt1-CuOx/Cu are obviously different from those of the PtSA-CN catalyst. In addition, the valence state of Pt will increase significantly as the CuOx/Cu hetero-structure undergoes further oxidation, which is unfavorable for the adsorption and dissociation of oxygenated intermediates during the electrocatalytic process (Supplementary Fig. 7). These results clearly suggest a modified valence state of Pt 5d in the Pt1-CuOx/Cu hybrid nanoparticles.Electrochemical oxygen reduction performanceTo elucidate the electrochemical performance of Pt1-CuOx/Cu, electrocatalytic measurements were conducted on the rotating disk electrode instrument34. The linear sweep voltammetry (LSV) curves were measured for Pt1-CuOx/Cu and control samples (commercial Pt-C, PtSA-CN and CuOx/Cu). In the Fig. 4a, the Pt1-CuOx/Cu catalyst shows the best ORR performance with a half-wave potential (E1/2) of 0.92 V vs. RHE, and a diffusion limited current density (JL) of 6.1 mA cm–2 (Fig. 4a). The E1/2 of the Pt1-CuOx/Cu catalyst is 60 mV higher than that of Pt-C (inset of Fig. 4b). The kinetic current density (Jk) at 0.85 V vs. RHE of the Pt1-CuOx/Cu catalyst is three times that of Pt-C and far exceeds that of PtSA-CN and CuOx/Cu (Fig. 4b). The outstanding ORR kinetics of Pt1-CuOx/Cu is also demonstrated by the Tafel slope in Fig. 4c and Supplementary Fig. 8. The Pt1-CuOx/Cu exhibits the lowest Tafel slope (with and without iR correction) compared to that of control samples. The turnover frequency (TOF) achieved in Pt1-CuOx/Cu far exceeds the control samples in this work, implying that the intrinsic activity of Pt1-CuOx/Cu was obviously improved (Fig. 4d and Supplementary Fig. 9). Mass activity (MA) is a crucial indicator for assessing the effectiveness of electrocatalysts. The MA of Pt1-CuOx/Cu was calculated in Fig. 4d. Apparently, the Pt1-CuOx/Cu catalyst (6.1 A mgPt-1) significantly surpasses Pt-C and PtSA-CN catalysts. The E1/2 and MA of Pt1-CuOx/Cu are superior to most excellent catalysts (Supplementary Table 1). The electrochemically accessible surface area (ECSA) of Pt1-CuOx/Cu and the PtSA-CN control sample were measured by cyclic voltammetry curves at the scanning rates ranging from 10 to 50 mV s−1. The results plotted in Supplementary Fig. 10, the Pt1-CuOx/Cu catalyst possesses a high ECSA of 152 m2 g−1, which is twice that of PtSA-CN.Fig. 4: Electrochemical performance.a LSV curves of Pt1-CuOx/Cu, Pt-C, PtSA-CN and CuOx/Cu samples, the resistance measured to be 20 ± 5 Ω, no iR correction was applied in the measurements. b The comparison of kinetic current density (Jk) and the inset is the locally amplified polarization plots. c Tafel slopes for Pt1-CuOx/Cu and the reference catalysts (The Tafel slopes were obtained at the scan rate of 10 mV s–1 without iR correction). d Turnover frequency and mass activity of Pt1-CuOx/Cu and the reference catalysts at 0.9 V vs. RHE. e The selectivity parameters of ORR pathway for the catalysts involved. f Stability tests of Pt1-CuOx/Cu and Pt-C. The error bars were estimated by the standard deviations of three individual calculations.The LSV curves collected at regular rotation speed were employed for Pt1-CuOx/Cu catalyst to evaluate the selectivity of electrocatalytic ORR. As depicted in Supplementary Fig. 11, the JL of the Pt1-CuOx/Cu catalyst is increased with regular rotating speed from 400 to 2500 rpm gradually, indicating the efficient transfer of oxygen species at increased rotating speed. Based on the Levich equation, the fitting results of the LSV curves show good linear parallelism, and the electron transfer number (n) was about 3.98, implying the satisfactory 4e- selectivity of the ORR pathway. Moreover, the rotating ring-disk electrode (RRDE) was also employed to reveal the disk and ring currents. The analysis results for the RRDE measurements were plotted in Fig. 4e and Supplementary Fig. 12. The average n of Pt1-CuOx/Cu and PtSA-CN were 3.95 and 3.32 in the potential range of 0.2 to 0.7 V vs. RHE, respectively, corresponding to the calculated results from Levich equation. Meanwhile, the Pt1-CuOx/Cu possesses the lowest average peroxide species (H2O2) yield (below 3%), which suggests the four-electron selectivity (production of H2O or OH-) was enhanced to >97% for Pt1-CuOx/Cu, significantly larger than those of Pt-C and PtSA-CN (~70%) (Supplementary Fig. 12d). The above results evidently reveal that the atomically doped Pt on the CuOx/Cu surface has significantly enhanced catalytic performance due to the modified electronic structure. The catalytic stability of the ORR process was assessed by chronoamperometry for 100 h. The results in Fig. 4f show that the performance of commercial Pt-C is attenuated by 20%, while the Pt1-CuOx/Cu catalyst remains essentially unchanged after long-term operation. The cyclic voltammetry curves before and after 5000 cycles demonstrate the good durability of Pt1-CuOx/Cu (Supplementary Fig. 13). The ICP-MS monitoring results in Supplementary Fig. 9 reveal that only about 1.6% of Pt was leached from Pt1-CuOx/Cu after the electrochemical operation (Supplementary Fig. 14). And the XRD patterns and Pt 4 f XPS spectra of Pt1-CuOx/Cu show similar crystal phase and electronic structure information before and after the electrochemical measurement, indicating the satisfactory stability of the Pt active sites in Pt1-CuOx/Cu (Supplementary Fig. 15). Moreover, the performance of Pt1-CuOx/Cu catalyst was almost undisturbed by methanol, which is better than that of the Pt-C catalyst (Supplementary Fig. 16).Exploration of dynamic valence evolution during the reactionTo deeply uncover the dynamic changes in the valence states and valence electron configurations of the catalysts, in situ XAFS characterizations were employed for Pt1-CuOx/Cu and PtSA-CN catalysts35,36. The normalized Pt L3-edge XANES spectra, which reflect the electronic structure evolution, are depicted in Fig. 5a, b. When the ORR potential was applied (1.00 V vs. RHE), the HA of Pt1-CuOx/Cu decreases obviously in relation to the open-circuit condition (immersed in electrolyte), indicating an enhanced electron occupation of Pt 5d at the initial reaction stage. In contrast, the HA of PtSA-CN was almost unchanged from open-circuit to the ORR potential (0.80 and 0.60 V vs. RHE), suggesting that the electronic state of the atomically dispersed Pt on general CN supports is relatively stable at the initial stage. The valence state of Pt in Pt1-CuOx/Cu increases gradually during the reaction process (1.00 → 0.75 V vs. RHE), and similarly, the valence state of Pt in PtSA-CN also shows a slight tendency to increase, which may be related to the interaction with the oxygen-related species. The formal d-band electron counts and valence states of Pt were calculated to quantify the valence electron evolution during the catalytic process based on the integral area of the white-line peak in differential XANES (ΔXANES) spectra (Supplementary Fig. 17). The fitting results show that the average valence state of Pt decreases from +2.61 to +1.92 as the conditions change from open-circuit to +1.00 V vs. RHE (Supplementary Fig. 18). Correspondingly, the formal d-band occupied density was estimated using a linear relationship obtained from the standards of Pt foil (5d96s1) and PtO2 (5d66s0). Figure 5c reflects that the occupied electron counts of Pt 5d bands for the Pt1-CuOx/Cu catalyst at open-circuit was 5d7.03 and then increased to 5d7.56 under 1.00 V vs. RHE, suggesting that the d-band electron quantity dynamically increased by 0.53 units.Fig. 5: In situ XAFS measurements.In situ Pt L3-edge XANES spectra of Pt1-CuOx/Cu (a) and PtSA-CN (b). c The d-band electrons at different conditions of Pt1-CuOx/Cu and PtSA-CN. d In situ XANES spectra at Cu K-edge of Pt1-CuOx/Cu. e In situ Pt L3-edge EXAFS spectra of Pt1-CuOx/Cu. The error bars were estimated by the standard deviations of three individual calculations.Furthermore, the dynamic evolution of Cu in Pt1-CuOx/Cu was analyzed by the in situ Cu K-edge XAFS to correlate with the changes of valence electrons at the Pt sites. In the Cu K-edge XANES spectra of Fig. 5d, as the ORR potential of 1.00 V was applied, the shift of the absorption edge towards higher-energy by approximately 0.8 eV compared to the open-circuit condition, which suggests an obviously increased Cu oxidation state at the initial reaction stages. To determine whether the increased oxidation state of Cu is influenced by the adsorption of oxygen-related species, the in situ XAFS measurements of the CuOx/Cu control sample were conducted (Supplementary Fig. 19). In contrast to those of Pt1-CuOx/Cu, the in situ Cu K-edge XANES spectra of CuOx/Cu show essentially no changes in the absorption edge at the initial reaction stage (before 0.80 V vs. RHE), and the slightly positive shift during the reaction process may be related to the adsorption of oxygen species on the surface of CuOx/Cu. The above results prove that the changes in the valence electron quantity for Pt and Cu in Pt1-CuOx/Cu are originated from the electron interaction between the Pt sites and CuOx/Cu supports at the initial reaction stage. The increased oxidation state of Cu corresponds to a decrease in the valence electron quantity, suggesting that the valence electron transformation is characterized by the CuOx/Cu support serving as a VER to rapidly increase the quantity of Pt 5d valence electrons at the initial reaction stage, which is beneficial for the orbital hybridization with O 2p and increasing the adsorption strength of oxygen species at the Pt sites during the ORR process.The facilitated flow of valence electrons from the surrounding Cu atoms to single Pt sites must be driven by the dynamic interaction between the Pt and the valence-flexible CuOx/Cu supports. Thus, the corresponding local structure evolution of the CuOx/Cu hybrid nanoparticles and single-Pt sites were further analyzed by in situ FT-EXAFS. The typical FT-EXAFS spectra of Cu K-edge at different reaction stages are displayed in Supplementary Fig. 20. As the working conditions changed from open-circuit to the potential of 1.00 V vs. RHE, the intensity of the dominant peak (about 2.43 Å), which corresponds to the Cu-Cu coordination, is notably decreased, along with the increase of the Cu-O coordination peak at 1.50 Å. This phenomenon implies that the oxidation degree on the surface of core-Cu particles is deepened at the initial reaction stage, which well coincides with the in situ XANES results at Cu K-edge, demonstrating the possible electron exchange between the surrounding CuOx species and single Pt sites through the Cu-O-Pt coordination. The local structure evolution around Pt sites was further identified by Pt L3-edge FT-EXAFS. In the FT-EXAFS spectra of Fig. 5e and Supplementary Fig. 21, all curves at different conditions present a dominant peak attributed to the Pt-O coordination at about 1.60 Å. Notably, the intensity of this dominant peak shows a damping by 20% as the potential changed from open-circuit to 1.00 V vs. RHE, which suggests distinct variations in the local coordination structures of Pt sites at initial reaction stage. The peak displays an increase in intensity as the applied potential drops to 0.90 and 0.75 V vs. RHE, likely due to the interaction with oxygen-related species during the electrocatalytic oxygen reduction process. Quantitatively, the FT-EXAFS fitting results exhibit that the coordination number of Pt-O is four under open-circuit condition (Fig.5e, Supplementary Fig. 22, and Supplementary Table 2). Notably, the coordination number of Pt-O bonds is reduced to three as the potential of 1.00 V vs. RHE was applied, indicating that the lower-saturated coordination active site was formed under ORR conditions. The local structure with lower coordination (Pt-O3) highly mitigates the electron grabbing by surrounding O atoms and favors the valence electron transfer from CuOx/Cu support to Pt centers at the initial stage. In the whole process of dynamic regulation, the CuOx/Cu support with elevated valence state acts as an effective VER and completes an in situ electron transportation to Pt 5d states. As the potential decreases to 0.90 V vs. RHE, an additional Pt-O coordination emerges, which corresponds to the adsorption of oxygen-related intermediates (*O, *OOH) on Pt sites. Meanwhile, the coordination of Cu-O in Cu K-edge FT-EXAFS spectra changed weakly as the ORR potential decreased from 1.00 to 0.90 V vs. RHE, which implies that the evolution of oxygen-related species primarily occurred at Pt sites after the rapid supplementation of valence electrons in the initial stage (Supplementary Fig. 20). These results indicate that because of the hybrid CuOx/Cu as an effective VER, highly occupied Pt 5d states are formed during the ORR, which is highly favorable for the hybridization of O 2p with Pt 5d orbitals. Moreover, as shown in the Pt L3-edge XAFS spectra, the oxidation state and the local coordination structure of Pt sites return to the initial state when the condition reversed to the open-circuit condition, indicating that the dynamically valence transformation between Pt atom and CuOx/Cu support is reversible (Supplementary Fig. 23), which can prevent the dissolution of the supports and leaching of active sites in the process of ORR. Hence, the dynamic valence evolution during the catalytic reactions promotes the evolution of reactive oxygen species and maintains the structural integrity for boosting activity and durability of the catalysts.Understanding of the catalytic enhancement mechanismTo reveal the nature of the catalytic performance enhanced by dynamic valence evolution, in situ SR-FTIR and electrochemical impedance spectroscopy (EIS) measurements were performed37. The results of Pt1-CuOx/Cu in Fig. 6a reflect that an obvious IR absorption band located at 890 cm−1 are observed during the ORR at various applied potentials. The absorption bands within the range of 800‒900 cm−1 are associated with the stretching vibration of oxygen species (*O)38,39. Therefore, the phenomenon of oxygen species (*O) accumulation can be attributed to the rapid dissociation of *OOH at the CuOx-isolated Pt sites to form *O intermediates, which is the key step of four-electron ORR40. Moreover, the absorption bands around 1200 cm–1 are usually associated with the stretching vibration of *OOH. The relatively weak vibration peak at 1180 cm–1 in Fig. 6a also demonstrates the rapid dissociation of *OOH intermediates by Pt1-CuOx/Cu. For comparison, in situ SRIR measurement was performed for PtSA-CN within the range of the corresponding catalytic ORR potentials. The results in Fig. 6b are obviously different from those for Pt1-CuOx/Cu. The absence of *O vibration peaks at 800‒900 cm−1 and the accumulation of *OOH at 923 cm-1 indicate that the dissociation of *OOH intermediates at the Pt-N4 active site is restricted. Meanwhile, the absorption bands at 1405 cm-1, which can be attributed to the surface-adsorbed *HOOH, are observed in Fig. 6b, implying the production of the H2O2 by-product. The variation of transmission intensity with ORR potential is plotted in Fig. 6c. It is clear that the *HOOH and *OOH intermediates accumulate continuously on the surface of PtSA-CN catalyst as the oxygen reduction progresses, indicating the occurrence of two-electron side reactions due to the difficulty of *OOH dissociation, which is in consistence with the results of selectivity testing41,42,43.Fig. 6: Intermediate species identification and their adsorption kinetic measurement.In situ SR-FTIR spectra in the range of 700–1500 cm–1 of Pt1-CuOx/Cu (a) and PtSA-CN (b). c The corresponding infrared signal fluctuations. In situ EIS measurements of Pt1-CuOx/Cu (d) and PtSA-CN (e). f The corresponding adsorption resistance (Rct) of oxygen species.To demonstrate the facilitating effect of Pt1-CuOx/Cu on the dissociation of *OOH intermediates, in situ SR-FTIR was also measured for the Pt-C control sample by the same three-electrode system (Supplementary Fig. 24a). The results in Supplementary Fig. 24b are quite different from those of Pt1-CuOx/Cu catalyst. The presence of only absorption bands located at 1128 cm-1 indicates the formation of *OOH on the Pt nanoparticle sites within Pt-C, and the continuous accumulation of *OOH at low potentials was attributed to the difficulty of further cleavage of *OOH for the Pt-C catalyst. In comparison with PtSA-CN, Fig. 6c shows that *OOH intermediates possess imperceptible fluctuations on the surface of Pt1-CuOx/Cu, implying the rapid decomposition. Correspondingly, the *O species accumulated rapidly as the potential decreased from 1.00 V vs. RHE. Combined with the in situ XAFS results, the mechanism diagram of the catalytic process is shown in Supplementary Fig. 25, and it can be concluded that the valence electron transport from the VER (CuOx/Cu) to Pt active sites at the initial reaction stage (open circuit → 1.00 V vs. RHE) promotes the proton coupling and electron transfer process of intermediate evolution, thereby accelerating the kinetics of the four-electron ORR.To further clarify the kinetics process of the species evolution over the electrode surface, in situ EIS was performed for the Pt1-CuOx/Cu and PtSA-CN catalysts under different ORR potentials44,45. In the Nyquist plots of Fig. 6d, e, the Pt1-CuOx/Cu possesses the distinct smaller semicircle size than that of PtSA-CN at a series of ORR potentials, which suggests a faster adsorption and evolution rate of reactants on the single Pt active sites. Furthermore, the Nyquist plots were simulated based on the equivalent circuit model in Supplementary Fig. 26. The fitted EIS parameter are displayed in Supplementary Tables S3, S4. Generally, the parameters of Rct represent the charge transfer resistance, and Cd are the capacitance. As shown in Fig. 6f, the Rct values of Pt1-CuOx/Cu are smaller than that of PtSA-CN in the whole reaction process and decrease rapidly at the kinetic control region, suggesting significantly accelerated adsorption dynamics of oxygen-related intermediates at the electrode/electrolyte interfaces. Moreover, the higher Cd values at various ORR potentials of Pt1-CuOx/Cu further demonstrates that the optimum adsorption of oxygen species on valence-reduced Pt sites dominates a highly selective four-electron ORR pathway. Therefore, the above in situ SR-FTIR and EIS results reveal that the dynamic valence transformation between Pt and CuOx/Cu at the reaction stage facilitates the formation and conversion of oxygen intermediates.In summary, a strategy of using valence-adjustable metal oxide/metal (CuOx/Cu) hybrid nanoparticles as an VER was proposed to tune the Pt 5d valence states for efficient ORR. Benefiting from the flexible valence electron exchange between the dynamically valence-reconstructed CuOx/Cu supports and single Pt center, the obtained Pt1-CuOx/Cu catalyst delivers improved four-electron selectivity and long-term stability of ORR. Using combined in situ synchrotron radiation characterization technologies, we uncovered that the increase of oxidation state in CuOx/Cu nanoparticles can effectively elevate the Pt 5d electron quantity by 0.53 units at the initial reaction stage, which effectively tunes the adsorption strength of the *OOH intermediates thereby improving the selectivity of ORR. Our work provides a universal wide-range valence-regulatory strategy for the development of advanced electrocatalysts.
