Interfacial Electron Distribution of Co Nanoparticles Supported on N‐Doped Mesoporous Hollow Carbon Spheres Endows Highly Efficient ORR, OER, and HER

The tailoring of the charge transfer between support material and transition metal active phase is an effective strategy for fine tuning the electronic structure of the catalyst active site, and hence improving the activity and stability of the reaction. This works presents that Co nanoparticles supported on N‐doped mesoporous hollow carbon nanospheres (Co/NMHCS) decouple the effect of electronic structure on catalytic performance. The detailed experimental and theoretical results reveal the charge distribution at the Co/NMHCS interface due to N‐doped MHCS. With tuning the electron redistribution, the interface between Co nanoparticles and NMHCS as the active site shows the strong capability to adsorb and reduce the OOH* and proton, thus leading to the optimal ORR, OER, and HER activity in Co/NMHCS. Furthermore, Co/NMHCS‐based Zn–air battery exhibits high power density of 185 mW cm−2, and high gravimetric energy density of 753 mAh gZn−1. Density functional theory (DFT) reveals the electrons accumulate directly on the NMHCS support, which originates from an interplay between Co nanoparticles and the NMHCS support. This work provides constructive guidance for precisely regulating the interface electronic structures to achieve excellent electrocatalytic performance.


Characterization of the Morphology and Structure
NMHCS was employed as a substrate due to its curved interface. The synthesis process of Co/NMHCS is schematically depicted in Figure 1a. First, the NMHCS with interconnected 3D porous structure was synthesized by carbonization of urea-doped phenolic resin (UPR) through a SiO 2 -templated strategy. NMHCS has a unique structure with a hollow cavity and flower-like surface morphology and shows high Brunauer-Emmett-Teller (BET) surface area (498.2 m 2 g −1 ; Figure S1, Supporting Information). Then, Co/NMHCS were prepared via a simple pre-precipitation method followed by thermal treatment. To investigate the influence of the pyrolysis temperature, the samples were sintered at 800, 900, or 1000 °C, which were labeled as Co/NMHCS (800), Co/NMHCS (900), and Co/NMHCS (1000), respectively. Co nanoparticles supported on N-free MHCS were also synthesized (the urea mass is 0) as reference. Scanning electron microscopy (SEM) and transmission electron microscopy (TEM) images show that the Co/NMHCS (900) takes a morphology of hollow sphere ≈260 nm with an average wall thickness of 20 nm (Figure 1b-d), as well as a graphitized edge (Figure 1e). Metal particles around 10-30 nm enveloped by carbon randomly dispersed on the surface of Co/NMHCS (900), of which the interplanar spacing was detected as 0.205 nm corresponding well to the Co (111) planes ( Figure 1f). Only the halo but not the diffraction spots can be found in the SAED result of Co/NMHCS (900), demonstrating the main feature of the amorphous structure ( Figure S2, Supporting Information). Moreover, the elemental mapping images of Co/NMHCS (900) indicate that N elements disperse uniformly on the NHMCS support surface and Co nanoparticles form on the NMHCS respectively ( Figure 1g). SEM of Co/NMHCS (800) clearly retains nanoflower morphology. When the temperature increased the 1000 °C, the Co/NMHCS (1000) clearly shows that the disappearance of nanoflower ( Figure S3, Supporting Information). In this case, Co/NMHCS (900) possesses more active sites for electrocatalytic reaction than that of Co/NMHCS (800).
Raman spectroscopy is employed to study the defects in NMHCS. [23,24] As shown in Figure 2b and Figure S5 (Supporting Information), the intensity ratio of the D-band and G-band (I D /I G) in Co/NMHCS (900) increased to 1.14 compared with that in Co/MHCS (0.91), due to the doped N-species. [25][26][27] The large intensity ratio of I D /I G showed an increased defect site in carbon matrix for Co/NMHCS (900), in line with the BET results and further indicating more available active sites for electrochemical reactions. [28] Furthermore, the distinct peak at 676 cm −1 is attributed to the doped Co-species. The typical X-ray photoelectron spectroscopy (XPS) survey spectrum confirms that the coexistence of Co (2.50 at.%), N (3.2 at.%), and C (95.2 at.%) in the spectrum Co/NMHCS (900) ( Figure S6, Supporting Information). The peaks at 779.7, 780.8, and 782.2 eV are assigned to metallic Co, Co-O, and Co-N/C peaks from the Co 2p3/2 of Co/NMHCS, respectively, in Figure 2c. [28][29][30][31] In contract, there are metallic Co and Co-O peaks in Co/MHCS(900). This result indicated that Co atoms could partially have some interaction with nitrogen. Furthermore, the high-resolution XPS spectrum for N 1s is shown in Figure 2d, namely, pyridinic N (390.8 eV), Co-N (399.1 eV), and graphitic N (400.3 eV), further demonstrating the formation of CoNC in Co/NMHCS (900). [32][33][34] There is no any N peak for Co/MHCS. [35][36][37] The C1s spectrum reveals the existence of CC (284.5 eV), CN (285.0 eV), and CO (285.8 eV) bonds in Co-NMHCS ( Figure S7, Supporting Information). [38,39] The presence of CN demonstrates that N atom has been successfully incorporated into the MHCS. The effect of the reaction temperature on the content of N element is investigated. Co/NMHCS (900) show the higher total atomic ratio of pyridinic-N and graphitic-N (4.43%) than Co/NMHCS (800)(3.87%) and Co/NMHCS (1000) (3.56%) in Figure S8 (Supporting Information). The higher N contents in Co/NMHCS(900) lead to stronger metal-support interactions, thus accelerating the charge transfer process for ORR, OER, and HER. [40][41][42] The coordination environment of Co species is further investigated by X-ray absorption spectra. It is founded that X-ray absorption near edge structure (XANES) spectra at Co K-edge of Co/NMHCS (900) are similar with that of Co foil ( Figure 2e and Table S1, Supporting Information), indicating the oxidation state of Co in Co/NMHCS (900) is ≈0. Notably, in the Fouriertransformed (FT) k3-weighted EXAFS in Figure 2f, the peaks at 2.1 Å of Co/NMHCS (900) are assigned the CoCo coordination, respectively. [43,44] The stronger CoCo bond proves the existence of metallic Co. [45][46][47]

Electrocatalytic Performance
The ORR activity of the Co/NMHCS samples was investigated in 0.1 m KOH solution ( Figure S9, Supporting Information). As shown in Figure 3a, the onset potential of Co/NMHCS (900) (0.956 V) is comparable with that of Pt/C (0.943 V), and outperforms that of Co/NMHCS(800) (0.876 V), Co/NMHCS(1000) (0.851 V), NMHCS (0.835 V), and Co/MHCS(900) (0.847 V) ( Figure S10, Supporting Information). The half-wave potential (E 1/2 ) of Co/NMHCS (900) is 0.850 V ( Figure S11, Supporting Information), higher than that of Pt/C (0.816 V) and most advanced non-precious metal-based ORR catalysts (Table S2, Supporting Information). The small Tafel slope (58.7 mV dec −1 ) further confirms the superior ORR dynamic of Co/NMHCS (900) (Figure 3b). The electron transfer number (n) of Co/ NMHCS (900) is measured as 3.9, indicating an approximate four-electron ORR pathway ( Figure S12  measurements were also determined by the capacitance of the double layer (C dl ) from CV measurements in 0.1 m KOH solutions (FIGURE S13, Supporting Information). The C dl of Co/ NMHCS (900) (35.6 mF cm −2 ) is ≈5.2 times higher than that of NMHCS (6.8 mF cm −2 ), reflecting more exposed active sites for promoting the catalytic reaction. [48] Moreover, the electrochemical impedance spectroscopy (EIS) of Co/NMHCS (900) reveals the smallest Nyquist circle compared with the other Co/ NMHCS samples, indicating its quickest charge transfer ability ( Figure S14, Supporting Information). The robust durability of Co/NMHCS (900) is demonstrated by the accelerated durability tests, of which just slight potential variation is observed after 10 000 cycles ( Figure 3c). The above results indicate the superior ORR performance of Co/NMHCS (900). The main reasons are as follows: 1) the NMHCS increases the active sites; 2) the high-percentage pyridinic N (56.9%) establishes more ORR active sites by carbon atoms with Lewis basicity next to pyridinic N. [49,50] and 3) the interface effect between Co and NMHCS thus leads to excellent ORR activity. [51,52] To demonstrate the versatility of Co/NMHCS (900), the electrocatalytic performance for OER and HER was further estimated in 1 M KOH solution. As expected, the OER performance of Co/NMHCS (900) exhibits superior overpotentials of 335 mV to drive a current density of 10 mA cm −2 , which is much lower than that of Co/NMHCS samples and pristine NMHCS and inferior to those of IrO 2 (320 mV) (Figure 3d). It outperforms non-precious metal-based OER catalysts reported recently (Table S3, Supporting Information). The Tafel slopes of the Co/NMHCS (900) (51.4 mV dec −1 ) are smaller than other samples and RuO 2 (79.9 mV dec −1 ), possessing the fast OER kinetics ( Figure 3e). [53] The Co/NMHCS(900) also exhibits a long-term cycling stability of almost unchanged LSV curves after 1000 CV cycles ( Figure 3f). The difference between OER and ORR metrics (ΔE = E j = 10 −E 1/2 ) is usually used to evaluate the performance of bifunctional oxygen electrocatalysts. [54][55][56] As shown in Figure S15 (Supporting Information), remarkably, a low ΔE value of 0.79 V is calculated over Co/NMHCS(900), confirming its bifunctionality. The Co/NMHCS was also investigated as an electrocatalyst for the HER. The Co/NMHCS(900) catalyst delivers a current density of 10 mA cm −2 at the low overpotential of 148 mV ( Figure 3g) and a small Tafel slope of 53.4 mV dec −1 (Figure 3h), outperforming most other stateof-the-art electrocatalysts. (Table S4, Supporting Information). The stability of the Co/NMHCS (900) electrocatalysts is also confirmed in Figure 3i. Furthermore, the TOF values of Co/NMHCS (900), Co/NMHCS (800), Co/NMHCS (1000), and NMHCS for HER were calculated as 0.05, 0.04, 0.03, and 0.02 s −1 at the overpotential of −200 mV, respectively.
Encouraged by the impressively multifunctionally catalytic activity of Co/NMHCS (900) for ORR and OER, the Zn-air battery was constructed using Co/NMHCS (900) coated on carbon fiber paper as an air-cathode, Zn foil as the anode and zinc acetate as the electrolyte. [57,58] The loading of Co/NMHCS (900) electrocatalyst was 0.6 mg cm −2 . As anticipated, Co/NMHCS Adv. Mater. Interfaces 2023, 10, 2202394 (900) (Figure 4a) displays a maximized power density of 185 mW cm −2 , which is significantly higher than commercial Pt/C (143.5 mW cm −2 ) and previously reported bifunctional catalysts (Table S5, Supporting Information). Moreover, Figure 4b reveals a much lower charge-discharge voltage gap of Co/ NMHCS (900) than Pt/C+RuO 2 . No visible change in chargingdischarging voltage gap was observed after long-term Zn-air battery operation, suggesting the prominent battery durability ( Figure 4c). Moreover, Co/NMHCS (900) based battery delivers a large specific discharge capacity of 753 mAh g Zn −1 at 10 mA cm −2 (Figure 4d), outperforming Pt/C+RuO 2 based battery (717 mAh g Zn  (Figure 4f). Inspired by the excellent OER and HER performance of Co/NMHCS(900), a water-splitting device is assembled using Co/NMHCS (900) as anode and cathode in 1.0 M KOH (Figure 4g). The overall water splitting displayed a low cell voltage of 1.61 V at a current density of 10 mA cm −2 (Figure 4h), which is superior to Pt/C||RuO 2 based electrolyzer (1.63 V) and recently reported water electrolyzers (Table S6, Supporting Information). At the same time, the Co/NMHCS (900) based electrolyzer also shows robust durability for 10 h (Figure 4i). Furthermore, we carried out the post-reaction characterization of TEM ( Figure S16, Supporting Information) and XPS testing ( Figure S17, Supporting Information).
DFT calculation was used to explore the interfacial electron distribution between Co nanoparticles and NMHCS. The  optimized structure of Co cluster loaded on the NMHCS substrate is shown in Figure 5a. The NMHCS substrate deforms and the plane becomes warped, reflecting the strong interaction between the Co cluster with the substrate. The charge density difference plot is shown in Figure 5b, and the interaction between NMHCS substrate and the Co cluster is clearly viewed. The strong accumulation and depletion of electrons at the interface suggest that the interface should be the most active area for catalytic reaction. The planar averaged charge density difference is shown in Figure 5c. As observed, the electrons accumulate directly above the NMHCS substrate, which originates from both the Co cluster as well as the NMHCS itself. The reaction-free energy diagrams for ORR, OER, and HER are further calculated. The chemisorption models of H 2 O*, O*, OH*, and OOH* intermediates on the interfaces of Co nanoparticles and NMHCS substrate, the top of Co nanoparticles and NMHCS are indicated in Figures S18-S20 (Supporting Information), respectively. The free energy diagrams of ORR at different active sites are shown in Figure 5d and Table S7 (Supporting Information). As one can observe, the free energy diagrams with U = 0 V for ORR at the NMHCS substrate and at the interface of Co nanoparticles and NMHCS substrate are downward, implying that all the reactions can proceed without additional energy required. We also show the free energy diagrams with an applied potential of 1.23 V. It was observed that the free energy diagrams shift with applied potential. For NMHCS substrate and the interface of Co nanoparticles and NMHCS substrate, the rate-determining steps are predicted to be the OOH*→O* and O*→OH*, respectively. Namely, the elementary steps of (2) and (3)   performance at the interface outperforms the NMHCS substrate and the Co nanoparticles. This reason may be ascribed to the charge redistribution and the synergistic effects of two parts. The free energy diagrams of OER at the three different sites are also illustrated in Figure 5e and Table S8 (Supporting  Information). To clarify the rate-determining step for OER, the free energy diagrams are conducted. The OER at the NMHCS substrate and at the interface between Co nanoparticles and NMHCS substrate proceed upward, implying that additional energy is required to initiate the OER, which is also consistent with the above-mentioned ORR. Similarly, the free energy diagram for OER with an applied potential of 1.23 V. The OER at the interface between Co nanoparticles and NMHCS substrate is limited by the last step of dissociation of OOH*, which requires an energy barrier of 0.63 eV. The smallest energy barrier in the interface between Co nanoparticles and NMHCS substrate thus favors the OER, facilitating the water splitting with a high performance. It may conclude that the ORR and OER are likely to take place at the interface of Co nanoparticles and NMHCS substrate with the lowest energy barriers. It also calculated the HER mechanisms on different sites to elucidate the hydrogen evolution on the constructed substrate. The results are shown in Figure 5f. The interface between Co nanoparticles and NMHCS substrate favors the HER with the lowest energy required to desorb the intermediate H* (0.25 eV). Hence the interface between Co nanoparticles and NMHCS substrate provides a favorable site for ORR, OER, and HER, while the NMHCS substrate only favors the ORR and OER.

Conclusion
In summary, we have developed a facile SiO 2 -templated strategy to prepare Co nanoparticles dispersed on NMHCS, the resultant Co/NMHCS(900) exhibits excellent trifunctionality toward ORR, OER, and HER. Meanwhile, Co/NMHCS(900) electrocatalyst presents superior activity and stability for Zn-air battery. Our theoretical calculations demonstrate the interface between Co nanoparticles and NMHCS is the active site in Co/ NMHCS(900) for the formation/deposition of O* and OOH* intermediates, thanks to the interfacial electron redistribution caused by the modulation of N doping between the interface of Co nanoparticles and NMHCS. Furthermore, the present study would shed light on the development of low-cost and scalable non-noble metal multifunctional catalysts for wide applications.

Experimental Section
Preparation of NHMCS: All chemicals were used as received without further purification. The N-doped mesoporous hollow carbon nanospheres (NMHCS) were prepared through a high-temperature carbonization strategy under N 2 atmosphere followed by HF etching treatment. [59] First, cetyltrimethylammonium chloride (CTAC; 2.0 g), Adv. Mater. Interfaces 2023, 10, 2202394 Figure 5. a) The structure of Co cluster loaded on the NMHCS sheet. The brown, cyan, and blue spheres represent C, N, and Co atoms, and graphene sheet with randomly placed pyridinic N and graphitic N atoms, respectively. b) The structure of Co cluster loaded on the NMHCS sheet. Yellow and green spheres represent the accumulation and depletion of electrons, respectively. The yellow sphere is positive (+), which means that this part gets electrons. The green one is negative (-), indicating the loss of this part. c) The charge density difference plot of optimized structure. d) Free energy diagrams of ORR at different active sites on interface of Co NP and NMHCS substrate, Co NP, and NMHCS substrate. e) Free energy diagrams of OER at different active sites interface of Co NP and NMHCS substrate, Co NP, and NMHCS substrate. f) Free energy diagrams of HER at different active sites. deionized water (120 mL), ethanol (20 mL), ammonia solution (0.5 mL), and resorcinol (0.55 g) were strongly stirred for 30 min at 70 °C. After that, tetraethyl orthosilicate (TEOS; 3 mL), formaldehyde (37 wt.%, 0.65 mL), and urea (0.6 g) were introduced into the solution with stirring for another 24 h, the SiO 2 @N-HMCS mesoporous sphere was vacuumdried at 70 °C under air-dried overnight, followed by heat treatment at 800 °C to form SiO 2 template. The SiO 2 template was etched using 20 wt.% HF solution and the N-doped hollow mesoporous carbon spheres were denoted as NMHCS for further use. MHCS was also similarly synthesized but without adding urea.
Synthesis of Co/NMHCS and Co/MHCS: Typically, a predetermined amount of Co(NO 3 ) 2 ·6H 2 O was dissolved in 5 mL of ethanol, and added dropwise to a suspension containing 10 mg of NHMCSs. After stirring for 12 h under ambient condition, Co(NO 3 ) 2 ·6H 2 O/NMCSs composite was collected by evaporating solvent at room temperature overnight. The catalysts were then annealed at different temperatures (800-1000 °C) in Ar/H 2 (95/5) atmosphere to obtain the Co/NHMCS. Co/MHCS sample was synthesized following the same process as preparing Co/NMHCS at 900 °C.
Physical Characterization: Powder X-ray diffraction (XRD) patterns of samples directly were performed using a Rigaku D/max-RC diffractometer. Scanning electron microscope was characterized by a Hitachi S4800 SEM. Transmission electron microscope with energy dispersive X-ray analysis (EDX) was carried out on a JEOL 2010F TEM/ STEM operated at 200 kV. The HAADF-STEM images were conducted on transmission electron microscope (JEM-ARM200F TEM/STEM). The Brumauer-Emmett-Teller (BET) specific surface area was measured with a Micromeritics ASAP 2050 system at 77 K. The composition of Co content was detected using inductively coupled plasma atomic emission spectroscopy (ICP-AES) techniques. Raman spectra were recorded on a Raman spectrometer (LabRAM HR800, λ = 514 nm). X-ray photoelectron spectroscopy (XPS) analysis was performed using Thermo ESCALAB 250XI. X-ray absorption fine structure (XAFS) spectra at the Co K-edge were recorded and performed in fluorescence mode at the beamline BL14W1.
Electrochemical Measurements: The electrocatalytic ORR, OER, and HER performance were evaluated at 25 °C on a CHI760E electrochemical workstation with a standard three-electrode system. [60] For ORR, platinum wire was used as the counter electrode, a Co catalyst-modified glass carbon was used as the working electrode (diameter: 5mm, area: 0.196 cm 2 ) from Pine Instruments, and Ag/AgCl was used as the reference electrode, respectively. All potentials were converted to values with reference to the reversible hydrogen electrode (RHE) according to the equation (E RHE = E Ag/AgCl +0.197+0.059×pH). For the preparation of the working electrode, 2.5 mg catalyst was added into 450 µL isopropyl alcohol and 50 µL 5 wt.% Nafion and then ultrasonicated for 1 h to form the homogeneous catalyst ink. Then, 12 µL of this catalyst was pipetted on the glassy carbon disk and dried at room temperature to obtain a catalyst loading of 0.2 µg cm −2 . Linear sweep voltammetry (LSV) measurements were carried out in N 2 or O 2 saturated 0.1 m KOH solutions at a scan rate of 5 mV s −1 at a rotation rate of 1600 rpm.
The RDE measurement was performed at a sweep rate of 10 mV s −1 with different rotating speeds (900-2025 rpm). The electron transfer number per oxygen molecule for oxygen reduction was calculated by Koutecky-Levich (K-L) equation [61] where J is the measured current density and is the electrode rotating rate (rad s −1 ), J K and J L are the kinetic-and diffusion-limiting current densities. ωis the angular velocity.
where C s represents the specific capacitance. The value of specific capacitance is 0.04 mF cm −2 in this calculation. [62] The turnover frequency (TOF, s −1 ) for OER can be calculated from the following equation: where J represents the current during LSV measurement, F is the Faraday constant (C mol −1 ), n represents the number of active sites (mol), and the factor 1/4 represents the corresponding electron transfer number. LSV for OER and HER was conducted at a scan rate of 5 mV s −1 in 1.0 m KOH solution, where graphite rod was employed as the counter. The iR compensation was performed by electrochemical impedance spectroscopy to correct the compensation potential.
Water Splitting Measurement: The air electrode was prepared by uniformly coating the as-prepared catalyst ink onto carbon paper and then drying under an infrared lamp. Two-electrode electrolysis devices, employing Co/NHMCS loaded on carbon paper are used as anode and cathodes, guaranteeing a mass loading of 0.6 mg cm −2 . LSV measurements were conducted in 0.1 m KOH with a scan rate of 5 mV s −1 .
Zn-Air Battery Measurements: The homemade aqueous Zn-air battery was assembled. The carbon cloth substrate coating with the catalyst layer (0.6 mg cm −2 ) was used as the air cathode. A polished Zn plate was applied as the anode (thickness: 1.0 mm). The mixed solution of 6 m KOH solution containing 0.2 m Zn (OAc) 2 was used as the electrolyte. LSV measurements were performed on a CHI-660 electrochemical workstation with a scan rate of 10 mV s -1 at room temperature. The cycling test was conducted by using the testing system for Zn-air battery was tested using recurrent galvanostatic pulses for 10 min of discharge (Land-CT2100A testing system). The LED screen (>3 V) was commercially available. The energy density was calculated with the following equation [63] = × ×∆ − Energy density(Whkg ) I V t Wzn 1 (4) where I represents the applied current (A), V represents average discharge voltage (V), t is service time, w Zn represents weight of zinc consumed (w Zn ) as in the following equation: Computational Methods: All the spin-polarized first-principle calculations were carried out by using the Vienna ab initio simulation package (VASP). [64] The energy cutoff was set as 480 eV. The projected augmented wave (PAW) pseudopotentials with valence-electron configurations of 1s 1 , 2s 2 2p 2 , 2s 2 2p 3 , 2s 2 2p 4 , and 4s 1 3d 8 were employed for H, C, N, O, and Co, respectively. [65] The exchange-correlation interaction was described by generalized gradient approximation (GGA) with the revised Perdew-Burke-Ernzerhof (PBE) functional. [66] The NMHCS substrate was modeled initially using the graphene sheet, in which pyridinic N and graphitic N atoms are randomly placed in the 6×4 orthogonal unit cell of the graphene sheet, the Co cluster is constructed from the fcc Co bulk with (111) direction along the c axis. In such a way, the distance between (111) lattice planes is 2.03 Å.
A vacuum space of 15.8 Å along the c direction was added for all substrates to avoid strong interactions between neighboring substrates. The graphene sheet with defect N atoms was first optimized with both atomic positions and lattice parameters. After that, the Co cluster was placed on the optimized NMHCS substrate and only atomic positions were optimized. Monkhorst-Pack special k-point meshes of 2×2×1 were proposed to carry out until the energy and force converged within 10 −5 eV and 0.05 eV Å −1 , respectively. The van der Waals interactions were corrected by zero damping DFT-D3 method of Grimme [67] The energy cut-off was chosen to be 400 eV. The charge density difference was calculated as ρ diff = ρ sys -ρ NMHCS-substrate -ρ Co-nanoparticles , where ρ sys presents the charge density of the system with Co nanoparticles loaded on the NMHCS substrate, ρ NMHCS-substrate and ρ Co-nanoparticles represent the charge densities of the pure NMHCS substrate and Co nanoparticles, respectively.

Supporting Information
Supporting Information is available from the Wiley Online Library or from the author.