Tailoring hard carbon interfaces in carbonate-based electrolytes for sodium-ion hybrid capacitors

INTRODUCTION

Nowadays, secondary batteries and supercapacitors are among the most widely used electrochemical energy storage devices^[1,2]. However, batteries often suffer from relatively low power density and limited cycle life, while supercapacitors face challenges with poor energy density^[3,4]. To bridge the performance gap between batteries and supercapacitors, hybrid capacitors that combine battery-type anodes with capacitor-type cathodes have emerged, leveraging the advantages of both technologies^[5]. Compared to lithium-ion hybrid capacitors, sodium-ion hybrid capacitors (SICs) offer a cost advantage due to the abundant reserves of sodium^[6]. Additionally, the ionic radius of the Na⁺ (102 pm) is larger than that of the Li⁺, which reduces the desolvation energy by 15%-20% for the same solvent molecule. The faster desolvation process, combined with the high abundance of sodium, makes it an attractive charge carrier for low-cost and high-kinetic hybrid capacitors^[7-9]. Hard carbon (HC), a promising material for commercialization, has numerous active sites for Na⁺ storage and a stable structure^[10]. The capacity of an HC anode typically consists of a partially slope capacity above 0.1 V (vs. Na/Na⁺) and a platform capacity below 0.1 V (vs. Na/Na⁺)^[11,12]. However, the rate performance of HC remains unsatisfactory, particularly under high currents, as the plateau capacity deteriorates rapidly. This limitation hinders the practical application of HC in hybrid capacitors.

The interaction between HC and the electrolyte is a critical factor influencing its rate performance. Electrolyte solvents for HC are primarily based on carbonate and ether solvents. Carbonate solvents are widely used in electrolyte solvents due to their excellent oxidative stability and high flash point, making them the preferred choice for high-voltage electrochemical energy storage devices. However, HC anodes for sodium-ion storage exhibit poor rate performance and low initial coulombic efficiency when matched with carbonate-based solvents. This is primarily due to the continuous growth of the solid electrolyte interphase (SEI) layer as a result of electrolyte degradation, which leads to significant kinetic challenges at the electrode/electrolyte interface, ultimately affecting the electrochemical performance^[13,14]. In contrast, when ether-based electrolytes are used in HC, capacity degradation occurs more slowly as the current increases. Yi et al. found that the co-intercalation behavior of Na⁺ and -solvent molecules in ether-based electrolytes lowers the desolvation energy, accelerating charge-transfer kinetics, particularly for Na⁺ storage behavior in the platform region^[15]. The SEI layer formed in ether-based electrolytes is thinner, facilitating faster Na⁺ storage compared to carbonate-based electrolytes^[16]. However, Yan et al. assessed the kinetic behavior of ether-based and carbonate-based electrolytes in the full cell composed of HC and sodium vanadium phosphate, and found that the rate performance in the carbonate-based electrolyte was superior, in contrast to conclusions drawn from conventional half-cell tests^[17]. Traditional half-cell tests, with rigid cutoff criteria, overlook the overpotential and impedance of metal electrodes, leading to an underestimation of the actual capabilities of HC in carbonate-based electrolytes. Additionally, carbonate-based solvents have superior oxidative stability compared to ether-based solvents, making them better suited for the long cycle life required in hybrid capacitors^[18,19]. Therefore, improving the rate performance of HC in carbonate-based solvents is crucial for meeting the practical demands of energy storage applications.

Modifying the surface chemistry of electrode materials is a widely adopted strategy to enhance the kinetics of sodium ion storage, as it directly influences the desolvation of Na⁺-(solvent)_n and the ability of Na⁺ to migrate through the SEI layer^[20-22]. For instance, increasing the surface oxygen functional groups on HC can promote electrolyte infiltration, provide more active sites for Na⁺ storage, and enhance the storage kinetics of Na^+[23]. Furthermore, differences in surface conductivity between materials, such as graphite and silicon oxide, can alter the structure of the SEI layer, which in turn affects the storage performance^[24]. Additionally, coating graphite with an ultrathin P layer that forms a crystalline Li₃P-based SEI during cycling has been shown to improve the desolvation of Li⁺ due to its high affinity for the ion, leading to ultra-fast charging performance^[25].

The introduction of nanomaterials on the surface of micrometer-sized electrode materials further optimizes surface chemistry while simultaneously increasing active sites, thereby enhancing specific capacity. Anode materials with nanostructures are particularly beneficial for hybrid capacitors, as nanosizing effectively improves kinetic performance by shortening the carrier transport path and accelerating surface charge transfer. Nanostructured materials such as TiO₂^[26], Co₃O₄^[27], Mo₂C^[28], and CoS^[29] have been successfully applied in lithium/sodium-ion hybrid capacitors. Therefore, selecting an appropriate nanostructured anode material is critical to improving the surface/interfacial chemistry of HC for hybrid capacitor applications.

Considering the findings from previous research, nanosized Co₃O₄ particles were introduced onto the HC surface as surface modification materials. The introduction of Co₃O₄ not only serves as an active material, providing more active sites, but also changes the surface chemistry of the HC anode. Through electrochemical testing and surface analysis techniques, it was found that the surface loading of Co₃O₄ increased the reversible capacity, promoted the reaction kinetics of Na⁺ storage in the platform region of HC, and simultaneously reduced the desolvation energy of the HC surface. This resulted in superior overall electrochemical performance compared to pure HC. The optimal composite, HC125@Co₃O₄, exhibited a higher specific capacity of 341 m Ah g^-1 compared to the specific capacity of 206 m Ah g^-1 of HC. In addition, even after 500 cycles at 0.5 A g^-1, HC125@Co₃O₄ maintained a specific capacity of 146 mAh g^-1, which is higher than that of HC (104 mAh g^-1). In A carbonate-based electrolyte, a SIC was constructed with a HC125@Co₃O₄ anode and an activated carbon (AC) cathode, demonstrating an energy density of 54.5 Wh kg^-1 and 76% capacity retention after 1,000 cycles at a high power density of 5,832 W kg^-1. Here, our work provides new insights into modulating the surface chemistry of HC anode materials in carbonate-based electrolytes to enhance its rate performance for application in hybrid capacitors.

EXPERIMENTAL

RESULTS AND DISCUSSION

The synthesis of HC125@Co₃O₄ is illustrated in Figure 1A. The defects and oxygen-containing functional groups in HC facilitate the adsorption of Co²⁺, which act as nucleation sites for the growth of Co₃O₄^[30]. Upon heating to a specific temperature, these coordinated ions decompose to form Co₃O₄ nanoparticles, which are deposited on the HC surface, resulting in the successful formation of the HC125@Co₃O₄ composite with a micro-nanostructure. HC itself exhibits an irregular block structure at the micrometer scale, while the surface of HC125@Co₃O₄ shows a more uniform distribution of Co₃O₄ particles.

Figure 1. Schematic of the preparation process and characterization of physicochemical properties. (A) Schematic of the synthesis of HC125@Co₃O₄. (B) XRD patterns of Co₃O₄ and HC75/125/175@Co₃O₄ samples. (C) TG of HC75/125/175@Co₃O₄ samples. (D) SEM, selected area (E) and the corresponding mapping (F-H) of HC125@Co₃O₄ samples. (I) nitrogen adsorption/desorption isotherms and (J) pore size distributions of HC125@Co₃O₄ powder.

The XRD patterns clearly confirm the successful formation of HC and Co₃O₄ composite during the solvothermal process [Figure 1B]. The diffraction peaks located at 19°, 31.3°, 36.8°, 44.8°, 59.3°, and 65.2° correspond to the (110), (220), (311), (400), (511), and (440) crystal planes of Co₃O₄ (PDF#42-1467), indicating the successful synthesis of the Co₃O₄ with a face-centered cubic structure. The intensities of the (311) peaks of HC75@Co₃O₄, HC125@Co₃O₄ and HC175@Co₃O₄ are higher than that of pure Co₃O₄, indicating that the addition of HC enhances the crystallinity of Co₃O₄. The HC content in the prepared anode materials (HC75@Co₃O₄, HC125@Co₃O₄, HC175@Co₃O₄) was determined from the TG curves [Figure 1C]. The slight weight loss between 25-250 °C corresponds to the release of absorbed gases or moisture, while the decomposition of HC begins at 250 °C. The weight percentages of HC in HC75@Co₃O₄, HC125@Co₃O₄, and HC175@Co₃O₄ are 66%, 72%, and 87%, respectively, indicating the dominant presence of HC in the composite materials. Energy Dispersive X-ray Spectroscopy (EDS) elemental mapping analysis [Figure 1D-H] clearly demonstrates the uniform distribution of Co and O atoms on the surface of HC. The addition of ammonia solution does not significantly increase the N content in HC125@Co₃O₄[Supplementary Figure 2]. From nitrogen adsorption/desorption isotherms, the N₂ isotherms for HC125@Co₃O₄ resemble type IV and H4-type hysteresis loops, indicating the presence of mesopores [Figure 1I]. Distinct peaks at 2.7, 5.4, 8.1, 12.8, 20.1, and 28.7 nm are observed in the pore size distribution for HC125@Co₃O₄[Figure 1J]. In contrast, the N₂ isotherm of HC shows type I and H3-type hysteresis loops, indicating micropores. The pore size distribution reveals that HC125@Co₃O₄ contains a significant amount of mesoporous and microporous structures, while the SSA of HC is mainly attributed to micropores. HC125@Co₃O₄ exhibits a higher SSA compared to HC, with the increase primarily due to the successful loading of mesoporous Co₃O₄ on the surface of HC. It is important to note that micropores store Na⁺, contributing to capacity, while mesopores enhance the kinetics of Na⁺ storage.

To evaluate the effect of Co₃O₄ incorporation on enhancing the electrochemical performance of the HC anode, different composite ratio materials were tested in half-cells paired with metallic sodium foils. Supplementary Figure 3 displays the CV results of HC, HC75@Co₃O₄, HC125@Co₃O₄, HC175@Co₃O₄, and Co₃O₄ at a scanning rate of 0.1 mV s^-1 for the initial three cycles. During the first cathodic scan, two distinct irreversible peaks were observed at approximately 1.1 and 0.5 V, corresponding to the decomposition of NaPF₆ and solvent, respectively^[31-34]. Compared to HC, the composite materials containing Co₃O₄ exhibit these two irreversible peaks at higher potentials, suggesting that solvent molecules and salt anions preferentially decompose at the Co₃O₄ layer. This early SEI formation helps prevent solvent molecules from co-intercalating with Na⁺ into the HC, further indicating the successful loading of Co₃O₄ onto the HC surface. During the first anodic scan, the removal of sodium ions from the HC is observed in two steps. The first step is the disappearance of Na metal clusters in the nanopores and Na⁺ de-insertion (0.22 V), while the second step involves further de-adsorption of Na⁺ from the surfaces and defects (0.67 V). Multiple oxidation peaks, corresponding to the transformation of Co₃O₄ into CoO and Co₃O₄, are observed between 0.49 and 2.1 V. The CV peaks of the composite material (HC75@Co₃O₄, HC125@Co₃O₄, HC175@Co₃O₄) show a combination of the characteristic peaks from both HC and Co₃O₄. As depicted in Figure 2A, peaks A1 and A2 primarily correspond to the anodic peaks of Na⁺ on HC, while peaks A3, A4, and A5 correspond to the anodic peaks of Na⁺ on Co₃O₄. Furthermore, as the scanning rate increases, the CV profiles show minimal changes, indicating excellent electrochemical reversibility and robust structural stability for the composite electrode HC125@Co₃O₄. The CV profiles of the other composite electrodes also remain favorable, as shown in Supplementary Figure 4.

Figure 2. Electrochemical performance tests. (A) CV plots of HC125@Co₃O₄, (B) rate capability, and (C) cycling performance of HC, HC125@Co₃O₄ and Co₃O₄ samples. (D) pseudocapacitive contribution. (E) CV plots for the first cycle, (F) the relationship of lgi vs. lgv (A1), and (G) cycling stability at 2 A g^-1 of HC and HC125@Co₃O₄ electrode.

Galvanostatic charge/discharge (GCD) tests were performed within the potential range of 0.01-3 V (vs. Na/Na⁺) to evaluate the electrochemical performance of the electrodes before and after composite formation, as well as for various composite ratio electrodes (HC, Co₃O₄, HC75@Co₃O₄, HC125@Co₃O₄, HC175@Co₃O₄) [Supplementary Figures 5 and 6]. Figure 2B illustrates the specific charge capacity-current density curves for these electrodes at 0.05-2 A g^-1. Among these, HC exhibits the lowest initial capacity, with a specific initial charge capacity of 206 mAh g^-1 at 0.05 A g^-1. In contrast, Co₃O₄ demonstrates the highest initial specific charge capacity, reaching 597 mAh g^-1. For the composite electrodes HC75@Co₃O₄, HC125@Co₃O₄, and HC175@Co₃O₄, the initial specific charge capacity increases sequentially with the Co₃O₄ loading [Supplementary Figure 7]. After cycling at different currents, HC125@Co₃O₄ exhibits higher reversible capacity when restored to 0.05 A g^-1. Additionally, the reversible capacity of HC125@Co₃O₄ at 0.05-2 A g^-1 (341-101 mAh g^-1) is significantly higher than that of HC (206-53 mAh g^-1). As observed in Figure 2C, Co₃O₄ exhibits the poorest cycling stability, with only 31 mAh g^-1 remaining after 500 cycles at 0.5 A g^-1. The rapid capacity decay of Co₃O₄ is attributed to its large volume change during charge and discharge cycles. HC125@Co₃O₄ demonstrates the best cycling performance, retaining a reversible capacity of 146 mAh g^-1 after 500 cycles, compared to 104 mAh g^-1 for the commercial HC after the same number of cycles. Considering that Co₃O₄ comprises 28 wt% of HC125@Co₃O₄, the capacity contribution ratios of Co₃O₄ in HC125@Co₃O₄ are 60%, 26%, 14%, and 5%, respectively, at the 1st, 100th, 200th, and 500th cycles. These results indicate that HC125@Co₃O₄ is the best anode among all materials, with the capacity advantage of Co₃O₄ in HC125@Co₃O₄ being most pronounced during the first 200 charge-discharge cycles.

As shown in Figure 2D, the pseudocapacitive contribution of HC125@Co₃O₄ at scan rates of 0.1-1.0 mV s^-1 is calculated to range from 38% to 66%, indicating its pseudocapacitive-dominant kinetic behavior in sodium-based energy storage devices. Although the introduction of the battery-type material Co₃O₄ reduces the pseudocapacitance contribution ratio of the HC (55%-79%) electrode material, pseudocapacitive-dominant kinetic behavior still prevails overall [Supplementary Figure 8]. Furthermore, the kinetic b-values of A3, A4 and A5 oxidation peaks associated with Co₃O₄ are 0.62, 0.66 and 0.70, respectively, confirming that the nanosized Co₃O₄ contributes to both pseudocapacitive and diffusive behaviors, enhancing the overall performance of the electrodes [Supplementary Figure 9].

Additionally, for HC electrode materials, the disappearance of quasi-metallic Na⁺ clusters in nanopores and their de-insertion from the interlayers, corresponding to the plateau capacity (A1), are the rate-limiting steps that influence the kinetic dynamics of Na⁺ storage [Figure 2E]. Based on the relationship between the measured peak current (i) and scan rate (v): i = av^b, where a and b are constants. The b-value of peak A1 for HC125@Co₃O₄ is 0.71, while the b-value of HC (A1) is 0.56 [Figure 2F]. This suggests that the introduction of Co₃O₄ promotes the kinetics of Na⁺ in the platform region, thereby enhancing the reversible capacity of HC. As shown in Figure 2C and G, HC125@Co₃O₄ maintains a higher reversible capacity than HC even after 500 cycles.

The shift of the two irreversible peaks to higher potentials indicates that the loading of Co₃O₄ onto HC promotes the early formation of the SEI film. The alteration may influence the ability of Na⁺ to de-solvate and pass through the SEI. To further investigate this, the impedance curves of HC//Na and HC125@Co₃O₄//Na batteries after five cycles were fitted using the Arrhenius equation at different temperatures, and the energy barriers for overcoming the SEI film and the desolvation process were analyzed^[35-37]:

Which can also be transformed into:

where k is the rate constant, R_SEI/ct represents the SEI film resistance (R_SEI) or charge transfer resistance (R_ct), A is the frequency coefficient, R is the ideal gas constant, T is the absolute temperature, and E_a is the reaction activation energy. In the Nyquist plots of Figure 3A and B, the R_SEI values of HC at temperatures 263, 273, 283, 293, and 303 K are 288.5, 151.6, 90.3, 61.1, and 38.9 Ω, respectively, while for HC125@Co₃O₄, the R_SEI values are 103.5, 54.8, 37.1, 29.8, and 18.7 Ω. The equivalent circuit diagram used for fitting is shown in Figure 3C. Based on the calculation of Eq. (2), the activation energy for Na⁺ passing through SEI in HC125@Co₃O₄ is 27(2) kJ mol^-1, which is lower than that of the HC (32.7(1.2) kJ mol^-1) [Figure 3D]. This indicates that the SEI film formed at the interface of Co₃O₄-loaded HC can effectively enhance the diffusion rate of Na⁺.

Figure 3. Electrochemical impedance and surface characterization of HC and HC125@ Co₃O₄. EIS plots at different temperatures of (A) HC and (B) HC125@Co₃O₄, and (C) the corresponding equivalent circuits. Arrhenius fitted lines for (D) R_SEI and (E) R_ct of HC and HC125@Co₃O₄. (F) Raman fitted plots of NaPF₆, HC, Co₃O₄, HC125@Co₃O₄. XPS spectra of (G) F1s, (H) O1s and (I) Na1s for HC and HC125@Co₃O₄ in sodium intercalated state after five cycles.

Furthermore, the R_ct values of HC and HC125@Co₃O₄ at different temperatures are shown in Figure 3A and B. For HC, the R_ct values are 5,048.2, 1,575.1, 558.4, 301.9, and 106.3 Ω, respectively; for the HC125@Co₃O₄, the values are 3011.0, 842.5, 386.5, 227.1, and 85.8 Ω, respectively. The activation energy, derived from the fitting R_ct, represents the desolvation energy of sodium ions. As shown in Figure 3E, the surface loading of Co₃O₄ reduces the desolvation energy of HC from 62(3) to 55(3) kJ mol^-1. At various temperatures, both the R_SEI and R_ct of HC125@Co₃O₄ are lower than those of commercial HC. Compared with HC, the surface loading of Co₃O₄ reduces the energy required for Na⁺ to pass through the SEI film by 17%(10%) and the energy of the Na⁺ desolvation by 11%(10%), which significantly contributes to the enhanced electrochemical performance of HC.

The change in desolvation energy is linked to the solvation structure of Na⁺ on the electrode surface, and Raman spectroscopy is a common method for studying the solvation structure. To simulate the state of the electrode sheets in the battery, the HC and HC125@Co₃O₄ sheets were soaked in the electrolyte overnight before performing Raman testing. Figure 3F illustrates the changes in the participation of anionic PF₆^- in the solvation structure at the interface of HC and HC125@Co₃O₄.

The peak of Na⁺-PF₆^- in pure NaPF₆ is present at 761 cm^-1. The peak located at 717 cm^-1 represents the presence of EC molecules (O_o-c-o) with ring respiration^[38]. The peak around 739 cm^-1 represents the stretching vibration of PF₆^- (V_P-F). PF₆^- at the interface can generally be categorized into three types: free ions (SSIP), contact ion pairs (CIP, where one PF₆^- coordinated with one Na⁺), and aggregated clusters (AGG, where one PF₆^- coordinated with two or more Na⁺)^[39]. The V_P-F of PF₆^- at 739.1 cm^-1 in the electrolyte at the HC125@Co₃O₄ interface exhibits the smallest redshift, which is higher than those observed at the HC (737.7 cm^-1) and Co₃O₄ (738.4 cm^-1) interfaces. This suggests a higher content of CIP and AGG in HC125@Co₃O₄. Compared to HC, the introduction of Co₃O₄ alters the solvent structure of the HC interface, with more PF₆^- participating in the interfacial solvent structure. As a result, during the desolvation process at the HC125@Co₃O₄ interface, more salt anions decompose to form an SEI film enriched with NaF^[40-42]. The NaF-rich SEI film enhances the desolvation ability of Na⁺, mainly by weakening the interaction between Na⁺ and solvents^[43]. Furthermore, the high content of NaF reduces the activation energy of the SEI formation and interfacial charge transfer resistance, thereby improving the reversible capacity of HC at high currents.

To confirm that the introduction of Co₃O₄ nanocrystals alters the composition of the SEI, XPS analysis was conducted on the electrodes of HC and HC125@Co₃O₄ after five cycles, respectively. As shown in Figure 3G, the peaks in the F1s spectrum primarily correspond to the decomposition products of the anions in NaPF₆. Compared to HC, the SEI film on the surface of HC125@Co₃O₄ contains higher amounts of Na_xPO_yF_z (686.6 eV) and NaF (684 eV), indicating that the introduction of Co₃O₄ promotes the decomposition of the PF₆^- anion. The products of the reaction between NaPF₆ and Na₂CO₃ are Na_xPO_yF_z and NaF^[44]. The organic polymer Na_xPO_yF_z can effectively inhibit electrolyte decomposition, while NaF is recognized for its ability to enhance the rapid transport of sodium ions. The combination of these two components improves both the sodium ion conductivity of the SEI and the cycling stability of the electrode material.

The fitting of O1s shown in Figure 3H reveals peaks located at 530.1, 530.8, 531.5, and 532.7 eV, corresponding to Na₂O, C=O, Na₂CO₃, and C-O, respectively^[45]. The C-O peak is generally associated with organic components in the SEI. The lower C-O content in HC125@Co₃O₄ compared to HC indicates that the presence of Co₃O₄ reduces the generation of soluble organic components, such as NaCO₂R, within the SEI. From the Na1s [Figure 3I], the NaF content in HC125@Co₃O₄ is higher than that in HC, consistent with the conclusions obtained from Figure 3G.

Considering the charge storage mechanism of the Co₃O₄ conversion reaction, it is essential to discuss the components formed after the conversion reaction and how each component influences the Na⁺ solvation structure. To explore this, a series of ex-situ XRD and XPS analyses were conducted to investigate the mechanism of HC125@Co₃O₄ as the anode and the composition of the electrode surface. Figure 4A shows the changes in the pristine electrode and the electrode at different potentials during the second GCD cycle [Figure 4B]. In the XRD spectrum of the pristine HC125@Co₃O₄, strong diffraction peaks corresponding to the spinel structure of Co₃O₄, such as (111), (220), and (311), are clearly visible. Since the XRD test was performed on HC125@Co₃O₄ coated on copper foil, the diffraction peaks from Co₃O₄ are relatively weak due to the strong diffraction signal from the collector (Cu foil). However, the peaks from crystal surfaces with high intensity are still distinguishable. In the magnified region between 35°-40°, it is evident that the (311) crystal plane disappears during the second charging/discharging process, indicating that Co₃O₄ has transformed into Co-based species with an amorphous state after sodiation. Additionally, NaF peaks are clearly observed, suggesting the presence of an SEI layer rich in NaF on the surface of HC125@Co₃O₄. This NaF-rich SEI layer provides excellent electronic isolation capability and reduces sodium diffusion resistance, thereby accelerating the kinetics of sodium ion transfer^[46]. Regarding the Co-based species formed after the transformation of Co₃O₄, ex-situ XPS analysis was performed [Figure 4C]. It can be seen that the main peak of Co2p_3/2 shifts at different stages of the second GCD cycle: at the onset of discharge (3V-D), fully-discharged (0.01V-C), and fully-charged (3V-C). During discharging, the binding energy of the Co2p_3/2 peak decreased (from 780.4 to 780.1 eV), and increased again during recharge (from 780.1 to 780.3 eV), indicating a change in the relative proportion of Co-based species on the electrode surface. The Co2p_3/2 spectrum can be convoluted into multiple distinguishable peaks: the peak at 779.3 eV corresponds to metallic Co, while peaks at 780.5 and 782.1 eV correspond to Co₃O₄ and CoO, respectively. During discharge, the content of Co species increases, while the content of CoO species remains nearly unchanged. This suggests that CoO transforms into Co, and part of Co₃O₄ also converts into CoO and Co species during discharge. Conversely, during charging, the content of Co decreases while the content of CoO increases, indicating the transformation from Co to CoO. These results demonstrate that three cobalt-based species (Co, CoO and Co₃O₄) are present during the charge/discharge process, and their relative proportions change accordingly.

Figure 4. Structural and surface evolution of HC125@ Co₃O₄ during cycling. (A) Ex situ XRD of the pristine electrode and HC125@Co₃O₄ electrode at different potentials during the second cycle at 0.05 A g^-1 (C/D mean charge/discharge), and (B) the corresponding potential and GCD curves. Ex situ XPS spectra of (C) Co_2p3/2 and (D) C1s for HC125@Co₃O₄ electrode in pristine, fully discharged-0.01 V and charged-3 V states during the second discharging/charging cycles at 0.05 A g⁻¹. (E) RDF and coordination number of Na⁺-PF₆^- in the presence of HC, Co, CoO and Co₃O₄.

The storage characteristics of Na⁺ in HC and the dynamic changes at the SEI interface were further explored through C1s XPS analysis, as shown in Figure 4D. The peak at 283.2 eV, corresponding to an extremely low binding energy, is attributed to Na_x-HC^[47]. The peak at 287.5 eV corresponds to the C=O bond. The intensity of this C=O peak decreases as Na⁺ is absorbed by the HC (during discharge) and increases again during Na⁺ extraction (during charge), indicating the reversible reaction between C=O and Na^+[48]. The peak at 288.7 eV corresponds to -CO₃^-, representing SEI film components such as Na₂CO₃ and NaOCO₂R. The intensity of this peak increases during discharge and decreases during charge, suggesting dynamic changes in the SEI film during the early cycles and indicating potential for SEI regulation.

To confirm the influence of three species formed after Co₃O₄ conversion on the Na⁺ solvation sheath, classic MD simulations were performed. The Raman tests show different Na⁺ solvation structures on the overall electrode surface, while MD calculations further present the effects of various species in the electrode material on the Na⁺ solvation, complementing the Raman results. In the initial stages of sodium ion storage, Co₃O₄ mainly exists as three species: Co₃O₄, CoO, and Co. To better understand the influence of Co-based species on the coordination between Na⁺ and PF₆^-, four models were constructed [Supplementary Figure 10].Figure 4E shows the radial distribution function (RDF) of Na⁺-PF₆^- in the electrolyte in the presence of four materials (Co₃O₄, CoO, Co, HC). The peak at 2.23 Å represents the first solvation sheath of Na⁺, corresponding to the Na⁺-PF₆^- coordination structure. In the presence of Co-based species, this peak value is higher compared to HC, with Co₃O₄ showing the highest peak value. The intensity of this peak in the presence of Co-based species is greater than that of HC in all cases, with Co₃O₄ exhibiting the largest peak. This change in peak intensity indicates that the presence of Co-based species enhances the coordination ability of Na⁺-PF₆^-, which is consistent with the findings from Raman spectroscopy. Figure 4F shows the coordination numbers derived from RDF. In the primary solvation shell of Na⁺, the coordination number of PF₆^- in the presence of HC is 1.53. In contrast, in the presence of Co-based species, the average coordination numbers of PF₆^- are all greater than 1.53, with values of 1.94 (Co), 2.19 (CoO), and 2.49 (Co₃O₄), respectively. The increase in coordination numbers further supports the idea that the surface loading of Co₃O₄ regulates the solvation structure of Na⁺, promoting more PF₆^- to participate in coordination and thereby contributing to the formation of an SEI film enriched with NaF.

To assess the practicality of HC125@Co₃O₄ in SICs, the SIC comprising HC125@Co₃O₄ anode and AC cathode (HC125@Co₃O₄//AC) was assembled, based on the operating principle shown in Figure 5A. During charge and discharge processes, the HC125@Co₃O₄ anode undergoes the sodiation/desodiation, while the AC cathode facilitates the adsorption/desorption of PF₆^- on its surface. The operating voltage window was optimized to be 0.01-3.8 V to avoid side reactions [Figure 5B]. The CV curves of SIC devices, tested at scan rates of 2 to 50 mV s^-1, show minimal deviation from ideal rectangle shapes, reflecting the combined energy storage mechanism of the anode and cathode. Figure 5C displays the GCD curves, which exhibit quasi-triangular shapes, further indicating multiple energy storage mechanisms in the device. The Ragone plot of the HC125@Co₃O₄//AC SIC [Figure 5D] was calculated based on the total mass of the cathode and anode, achieving an energy density of 129.5 Wh kg^-1 at a power density of 583 W kg^-1. Notably, even at an ultra-high power density of 11,650 W kg^-1, the SIC device maintains an energy density of 26.5 Wh kg^-1, which offers a higher power density compared to the latest literature, as shown in Supplementary Table 1. The HC125@Co₃O₄//AC also demonstrates excellent cycling stability [Figure 5E], with a capacity retention rate of 76% after 1,000 cycles at 2.0 A g^-1. The above SIC highlights the promising potential of HC125@Co₃O₄ as an anode material for advanced sodium-based energy storage devices and also underscores the application prospects of commercial HC in hybrid capacitors.

Figure 5. Electrochemical performance of HC125@Co₃O₄//AC sodium-ion hybrid capacitor (SIC) (A) The schematics of work principles during the charging process, (B) CV plots, (C) GCD curves, (D) Ragone plots and (E) cycling performance of HC125@Co₃O₄//AC SIC.

CONCLUSIONS

To enhance the rate performance and Na⁺ reaction kinetics of HC in carbonate-based electrolytes, Co₃O₄ nanoparticles were uniformly loaded onto the surface of HC via a simple one-step solvothermal method. As an active material, Co₃O₄ not only contributes to an increase in reversible specific capacity but also regulates the Na⁺ solvation structure on the electrode surface. After 200 cycles, the NaF-rich SEI plays a dominant role in improving the capacity of HC. The presence of Co-based species facilitates greater participation of PF₆^- in Na⁺ coordination, promotes Na⁺ desolvation, and optimizes the composition of the SEI film. In the SEI of HC125@Co₃O₄, the NaF content is higher than in HC, leading to faster charge transfer kinetics. As a result, HC125@Co₃O₄ exhibits an excellent reversible capacity of 341 mAh g^-1 at 0.05 A g^-1, and retains a high capacity of 87 mAh g^-1 after 500 cycles at 2 A g^-1. The assembled SIC achieves a maximum energy density of 129.5 Wh kg^-1 and a maximum power density of 11,650 W kg^-1, with a capacity retention rate of 76% after 1,000 cycles at 2 A g^-1. This work provides a novel strategy for regulating the solvation environment of sodium ions and promotes the practical application progress of battery-type HC materials in hybrid capacitors.