Honeycomb-like N-Doped Carbon Matrix-Encapsulated Co_1−xS/Co(PO₃)₂ Heterostructures for Advanced Lithium-Ion Capacitors

Abstract

Lithium-ion capacitors (LICs) are emerging as promising hybrid energy storage devices that combine the high energy densities of lithium-ion batteries (LIBs) with high power densities of supercapacitors (SCs). Nevertheless, the development of LICs is hindered by the kinetic imbalances between battery-type anodes and capacitor-type cathodes. To address this issue, honeycomb-like N-doped carbon matrices encapsulating Co_1−xS/Co(PO₃)₂ heterostructures were prepared using a simple chemical blowing-vulcanization process followed by phosphorylation treatment (Co_1−xS/Co(PO₃)₂@NC). The Co_1−xS/Co(PO₃)₂@NC features a unique heterostructure engineered within carbon honeycomb structures, which efficiently promotes charge transfer at the interfaces, alleviates the volume expansion of Co-based materials, and accelerates reaction kinetics. The optimal Co_1−xS/Co(PO₃)₂@NC composite demonstrates a stable reversible capacity of 371.8 mAh g⁻¹ after 800 cycles at 1 A g⁻¹, and exhibits an excellent rate performance of 242.9 mAh g⁻¹ even at 8 A g⁻¹, alongside enhanced pseudocapacitive behavior. The assembled Co_1−xS/Co(PO₃)₂@NC//AC LIC delivers a high energy density of 90.47 Wh kg⁻¹ (at 26.28 W kg⁻¹), a high power density of 504.94 W kg⁻¹ (at 38.31 Wh kg⁻¹), and a remarkable cyclic stablitiy of 86.3% retention after 5000 cycles. This research is expected to provide valuable insights into the design of conversion-type electrode materials for future energy storage applications.

1. Introduction

Lithium-ion capacitors (LICs) are considered promising candidates in the field of electrochemical energy storage due to their hybrid energy storage capabilities, which combine the high energy characteristics of lithium-ion batteries (LIBs) with the high power densities of supercapacitors (SCs). However, the practical implementation of LICs remains challenging due to the distinct energy storage mechanisms: the high-energy anodes operate through a slower, repeated Faraday reaction of lithium ions, while the cathodes undergo rapid anion absorption and desorption [1,2,3].

To address this challenge, researchers have devoted significant efforts to discovering new battery-type anode materials that offer both rapid rate performance and high capacities. Among these materials, cobalt sulfides, which are typical transition metal sulfides, show promise as conversion-type anodes due to their potential application on account of their high theoretical specific capacities (590 mAh g⁻¹), weaker metal–sulfur bonds, and Co/Li₂S interfaces offering high electron transfer mobility [4,5]. However, cobalt sulfides experience significant volume expansion and contraction during repeated lithiation/delithiation cycles, leading to particles aggregation and the detrimental pulverization of electrode structures. This results in a rapid decline in rate performance, creating a mismatch with the capacitor-type cathodes in LICs. Numerous approaches have been proposed to improve reaction kinetics and structural stability, including downsizing particle size [6,7], constructing hierarchical porous structures [8,9], integrating with heteroatom-doped conductive carbonaceous materials [10,11,12], and fabricating heterogeneous architectures [13,14,15]. These approaches are expected to alleviate internal stress, increase contact with the electrolyte and shorten electronic and ionic transport pathways, enhance conductivity and maintain the integrity of the electrode materials, and regulate the electronic structure, ultimately aiming to enhance electrochemical performance, including increasing capacity, rate, and cycling stability. It is noteworthy that heterostructures composed of coupling components with large energy bandgap differences can induce a strong intrinsic electric field at the heterogeneous interfaces, thereby promoting reaction kinetics. Moreover, the as-formed heterointerfaces, characterized by abundant phase boundaries, have been shown to significantly influence rate capability and reversible capacity compared to individual components [16,17,18]. Recently, Dong and co-workers demonstrated that the CoS₂/Co₄S₃@NC heterogeneous composites exhibit high pseudocapacitive behavior and accelerate the diffusion rate of ions and electrons, thus improving the overall performance of the electrodes [19]. Wan and colleagues prepared a hierarchical binary metal sulfide (CoSn)S/C heterostructure, where the micro/nanostructures of (CoSn)S/C shorten the ion diffusion distance, facilitating the rapid transfer of Na⁺ and enhancing the mechanical strength of the electrode [20]. Although cobalt-based sulfides play a significant role in improving the electrochemical properties of LICs, unlocking the potential of heterogeneous materials still requires a great deal of work to be explored.

Herein, honeycomb-like N-doped carbon matrices encapsulating Co_1−xS/Co(PO₃)₂ heterostructures are prepared using a simple chemical blowing-vulcanization process followed by phosphorylation treatment. The selection of Co_1−xS and Co(PO₃)₂ components is based on their bandgaps, which are 0 eV for Co_1−xS and approximately 2.6 eV for Co(PO₃)₂. The heterointerfaces between Co_1−xS and Co(PO₃)₂ components induce an internal electric field, increasing the number of electrochemical active sites, improving electrochemical kinetics, and stabilizing the structure during the repeated charge/discharge. Additionally, the integrated honeycomb-structured N-doped carbon matrices provide physical confinement to reduce nanograins aggregation, accommodate volume expansion during repeated long cycling, and ensure the integrity of the electrode materials, resulting in a high reversible capacity. As expected, the optimal Co_1−xS/Co(PO₃)₂@NC composite demonstrates a stable reversible capacity of 371.8 mAh g⁻¹ after 800 cycles at 1 A g⁻¹, and an excellent rate performance of 242.9 mAh g⁻¹ even at 8 A g⁻¹, accompanied with enhanced pseudocapacitive behavior. Furthermore, the assembled Co_1−xS/Co(PO₃)₂@NC//AC LIC delivers a high energy density of 90.47 Wh kg⁻¹ (at 26.28 W kg⁻¹), a high power density of 504.94 W kg⁻¹ (at 38.31 Wh kg⁻¹), and a high cyclic stability of 86.3% retention after 5000 cycles.

2. Results and Discussion

The synthetic strategy for the honeycomb-like N-doped carbon matrix decorated with Co_1−xS/Co(PO₃)₂ heterostructures is schematically illustrated in Figure 1a. Initially, polyvinylpyrrolidone (PVP), serving as the carbon source, weakly coordinates with Co²⁺ to form a sol–gel coordination complex precursor [21,22]. The precursor provides physical confinement to prevent the agglomeration of Co_1−xS nanograins during the calcination process. Subsequently, the precursor and mixed sulfur powder undergo thermal blowing-vulcanization treatment. Due to the high viscosity of molten PVP, the pyrolysis gasses (NO₂, NO, etc.) released from nitrate ions generate numerous bubbles as the temperature rises. These bubbles rapidly rupture, promoting the growth of interconnected thin carbon walls, resulting in the formation of N-doped carbon matrices decorated with cobalt sulfides that resemble a 3D honeycomb-like structure. Afterwards, Co_1−xS/Co(PO₃)₂@NC is obtained through a facile phosphorylation process of the Co_1−xS/@NC composite, yielding conformal Co_1−xS/Co(PO₃)₂ heterostructures. FESEM results from Figure 1b–d show that the as-prepared Co(PO₃)₂@NC, Co_1−xS@NC, and Co_1−xS/Co(PO₃)₂@NC composites exhibit a typical 3D interconnected honeycomb-like framework comprising numerous thin carbon walls with smooth surfaces. This indicates sulfides/phosphides are uniformly encapsulated within the carbon matrices. This structure facilitates electrolyte penetration and effectively mitigates the volume change in active species during repeated cycling, thereby maintaining structural stability. TEM was employed to further examine the morphology and chemical properties. As shown in Figure 1e–g and Figure S1–S3, numerous small nanoparticles are found within the carbon matrices of all the Co_1−xS/Co(PO₃)₂@NC, Co_1−xS@NC, and Co(PO₃)₂@NC composites. Notably, the Co_1−xS/Co(PO₃)₂@NC contains nanoparticles with an average size of less than 5 nm, demonstrating shortened electronic and ionic transport paths and effectively relieving internal strain during the electrochemical process. The lattice spacings of 0.268 and 0.254 nm in the higher magnification image (Figure 1e,f) correspond to the (130) plane of Co(PO₃)₂ and the (101) plane of Co_1−xS, respectively. Similarly, the lattice spacings of 0.194 and 0.187 nm in Figure 1g correspond to the (102) plane of Co_1−xS and the (−242) plane of Co(PO₃)₂, respectively, confirming the successful fabrication of heterostructures. The energy dispersive X-ray spectrometer (EDS) spectrum in Figure 1h confirmed the uniform dispersion of Co, S, P, N, and O in the carbon matrices.

The crystal structure of the synthesized materials was characterized using X-ray diffraction (XRD) patterns, as depicted in Figure 2a. The diffraction peaks of the Co_1−xS/Co(PO₃)₂@NC composites were consistently indexed to the monoclinic Co(PO₃)₂ phase (JCPDS No. 27-1120) and the hexagonal Co_1−xS phase (JCPDS No. 42-0826). For the Co_1−xS@NC and Co(PO₃)₂@NC composites, only single phases of Co_1−xS and Co(PO₃)₂ were observed, without any impurity peaks. Moreover, the Raman spectra for three samples (Figure 2b) exhibit two prominent peaks at ~1351 and ~1371 cm⁻¹, which represent the graphitic carbon (G-band) and disordered carbon (D-band), respectively. This confirms the presence of N-doped carbon matrices. The relatively larger intensity ratio of I_D/I_G for Co_1−xS/Co(PO₃)₂@NC (0.989) implies a higher degree of structural defects compared to Co_1−xS@NC (0.973) and Co(PO₃)₂@NC (0.969), which is conducive for fast electron transfer [23]. Thermogravimetric analysis (TGA) was conducted to determine the contents of Co_1−xS, Co(PO₃)₂, and carbon in Co_1−xS/Co(PO₃)₂@NC (Figure 2c). As reported in our previous work, Co_1−xS undergoes a multi-stage reaction involving complex oxidation processes in the air followed by carbon combustion [24,25]. The evaporation of adsorbed water accounts for the slight mass loss in the range of 50–200 °C. The oxidation of Co_1−xS to CoSO₄ and Co₃O₄ is the primary cause of a notable weight loss, as well as the combustion of NC, as the temperature increases from 200 to 650 °C. Notably, the differences between the curves of Co_1−xS/Co(PO₃)₂@NC and Co_1−xS@NC are primarily attributed to the greater weight loss from NC combustions compared to the increase from oxidation of Co_1−xS. The TGA curves indicate the percentages of Co_1−xS, Co(PO₃)₂, and NC in Co_1−xS/Co(PO₃)₂@NC, which were 40.9 wt%, 29.4 wt%, and 29.7 wt%, respectively.

Furthermore, the surface chemical compositions of the three samples were analyzed using X-ray photoelectron spectroscopy (XPS) measurements. As shown in Figure 3a, elements such as cobalt, sulfur, phosphorus, oxygen, carbon, and nitrogen were clearly detected. The Co 2p high-resolution spectra, presented in Figure 3b, reveal the characteristic peaks at ~798.4 eV and ~782.6 eV, corresponding to Co 2p_1/2 and Co 2p_3/2 of Co²⁺, along with their satellite peaks [26,27]. An additional set of peaks in the Co_1−xS@NC composites, located at ~793.4 eV and 778.4 eV, is associated with Co⁰ and suggests that the vulcanization process of the PVP-Co²⁺ precursors is not fully completed. Figure 3c shows the S 2p spectra, where characteristic peaks at approximately 165.2 eV and 164.0 eV are associated with S 2p_1/2 and S 2p_3/2 of the C-S-C bond, while peaks at ~162.9 eV and ~161.8 eV correspond to S 2p_1/2 and S 2p_3/2 of S²⁻ [28]. Additionally, a peak at ~168.5 eV corresponds to SO_x likely due to inevitable surface oxidation. The strong chemical bonds between active species and carbon substrate are indicated by the presence of C-S-C, ensuring robust structures and excellent cycling stability during the intercalation and de-intercalation of Li⁺ [29]. The characteristic peak at ~134.4 eV (Figure 3d) corresponds to P-O bonds. The C 1s peaks can be deconvoluted into three distinct peaks at ~284.8, ~285.9, and ~289.0 eV (Figure 3e), corresponding to C-C, C-N/C-O/C-S, and C=O bonds, respectively. The successful nitrogen substitution in the carbon matrix is confirmed by the presence of the C-N bond, and this is supported by the N 1s high-resolution spectra (Figure 3f). The nitrogen dopant can not only enhance interfacial wettability, create more defects and active sites, and accelerate reaction kinetics [30].

The Li⁺ storage behavior of the Co_1−xS/Co(PO₃)₂@NC electrodes was evaluated using lithium metal as a counter electrode in an assembled 2032 coin-type half-cell. First, as shown in Figure 4a, cyclic voltammetry (CV) curves were measured at a sacn rate of 0.2 mV s⁻¹, revealing the electrochemical process over the first five cycles. In the initial cathodic scan, the sample showed significant broad peaks ~0.52 V and ~0.94 V, indicative of irreversible transformations and the formation of a solid electrolyte interface (SEI). In the subsequent scans, a set of new characteristic cathodic peaks appeared at 1.98 and 1.24 V related to the stepwise conversion reaction of Co_1−xS into Co/Li₂S, Co(PO₃)₂ to CoP₂/Li₂O intermediates, and Co/Li₃P products; additionally, notable oxidation peaks were observed at 1.32, 2.05, and 2.56 V, corresponding to the oxidation reactions of Co/Li₂S to Co_1−xS and Co/Li₃P to CoP₂ [31], which persisted in the following scans. The galvanostatic charge/discharge (GCD) profiles of the Co_1−xS/Co(PO₃)₂@NC electrode displayed initial charge/discharge capacities of 1194.0/719.1 mAh g⁻¹ with an initial Coulombic efficiency of 60.2%; in subsequent cycles, the capacities remained stable over 200 cycles with Coulombic efficiency, approaching to 99.6% (Figure 4b). The cycling performance reveals that the Co_1−xS/Co(PO₃)₂@NC electrode maintains the highest reversible capacity of 643.7 mAh g⁻¹ after 200 cycles at 0.2 A g⁻¹, significantly exceeding those of Co_1−xS@NC (336.5 mAh g⁻¹), Co(PO₃)₂@NC (274.0 mAh g⁻¹) (Figure 4c), Co_1−xS/Co(PO₃)₂@NC-1v3 (478.4 mAh g⁻¹), and Co_1−xS/Co(PO₃)₂@NC-1v8 (426.8 mAh g⁻¹) (Figure S4). The rate capability of all samples was further examined across different current densities ranging from 0.2 to 8.0 A g⁻¹. The Co_1−xS/Co(PO₃)₂@NC exhibited the best rate performance regardless of current density, with a high capacity of 642.7 mAh g⁻¹ at 0.2 A g⁻¹ and 242.9 mAh g⁻¹ at 8.0 A g⁻¹ (Figure 4d,e); the rate performance of Co_1−xS/Co(PO₃)₂@NC is superior to some previously reported cobalt sulfide-based anode materials (Figure 4f). The long-term cycling performance of Co_1−xS/Co(PO₃)₂@NC, Co_1−xS@NC, and Co(PO₃)₂@NC electrodes was further compared at a current density of 1 A g⁻¹. As shown in Figure 4g and Figure S4, the Co_1−xS/Co(PO₃)₂@NC electrode delivers an obvious higher reversible capacity and improved cycling stability (371.8 mAh g⁻¹ after 800 cycles), while Co(PO₃)₂@NC maintains a capacity of only 230.1 mAh g⁻¹, which is lower than the reversible capacity for Co_1−xS/Co(PO₃)₂@NC-1v3 and Co_1−xS/Co(PO₃)₂@NC-1v8. This difference in performance is mainly due to the synergistic effect of the heterogeneous interface and the integrated honeycomb-structured carbon, which efficiently promote charge transfer at the interfaces and accelerate the reaction kinetics, accommodate volume changes during prolonged long cycling, and ensure the structural integrity of the electrode materials.

In order to assess the superior lithium storage capabilities of the Co_1−xS/Co(PO₃)₂@NC, cyclic voltammetry (CV) curves were recorded at various scanning rates, ranging from 0.1 to 5.0 mV S⁻¹ (Figure 5a). As the scan rate increased, there were no discernible changes in the number or peak voltage of the observed peaks in the CV curves, indicating that Co_1−xS/Co(PO₃)₂@NC has good reversibility. It is widely accepted that the following Equation (1) could be used to explain the relationship between peak current (i) and scan rate (v):

$$i={\mathrm{a}v}^{\mathrm{b}}$$

(1)

where both a and b are variable parameters. The value of b can be used to diagnose whether the reaction is a diffusion-controlled reaction (for b = 0.5) or surface-controlled (for b = 1.0). As shown in Figure 5b, the b values at the O1 peak and R1 peak of the Co_1−xS/Co(PO₃)₂@NC electrodes were determined to be 0.66 and 0.67, respectively, suggesting a combination of diffusion-controlled and capacitive-controlled behaviors. Furthermore, the proportion of pseudocapacitive contribution could be quantified using the following Equation (2),

$$i=k_{1}v+k_2{v}^{1/2}$$

(2)

where k₁v and k₂v^1/2 indicate the current response from the capacitive capacity and diffusion-controlled capacity, respectively. As shown in Figure 5c,d, the pseudocapacitive contributions of the Co_1−xS/Co(PO₃)₂@NC electrodes at the scanning rate of 2.0 mV s⁻¹ were evaluated to be 70.0%. With the increase in v, the pseudocapacitive contribution gradually increases from 52.3% to 74.7%, demonstrating the important effect of the heterostructure in facilitating the electrode reaction kinetics. Figure 5e shows the Nyquist plots of three samples after three cycles at 0.1 A g⁻¹. All EIS spectra display similar characteristics, including two sections: a semicircular region at a medium- to high-frequency corresponding to R_ct and a linear region at low-frequency associated with Li⁺ diffusion. Compared to Co(PO₃)₂@NC and Co_1−xS@NC, the Co_1−xS/Co(PO₃)₂@NC electrode exhibits significantly smaller R_ct in the medium- to high-frequency range (186 Ω compared to 304 Ω, 201 Ω), indicating that the unique heterostructures and effective particle refinement indeed enhance charge transfer and reaction kinetics. Moreover, based on the following equations related to diffusion coefficient of Li⁺:

$${\mathrm{Z}}^{\prime }={\mathrm{R}}_{\mathrm{s}}+{\mathrm{R}}_{\mathrm{c}\mathrm{t}}+\mathsf{\sigma }{\mathsf{\omega }}^{{\displaystyle \raisebox{1ex}{$-1$}\!\left/ \!\raisebox{-1ex}{$2$}\right.}}$$

(3)

According to Equation (3), the Warburg factor (σ) of the two electrodes was calculated based on the linear relationship between Z’ and ω^−1/2 (Figure 5f). The value of σ for the Co_1−xS/Co(PO₃)₂@NC electrode (41.0) is much lower than the Co_1−xS@NC (57.6) and Co(PO₃)₂@NC electrodes (168.9). As the D_Li+ diffusion coefficient is inversely proportional to the square of the Warburg factor (σ) [3], this lower σ value indicates a faster Li⁺ diffusion coefficient for the Co_1−xS/Co(PO₃)₂@NC electrode.

Considering the exceptional electrochemical performance of the Co_1−xS/Co(PO₃)₂@NC in half-cells, a prelithiated Co_1−xS/Co(PO₃)₂@NC anode and an AC cathode were combined in an optimal mass ratio of 2:1 to assemble a Co_1−xS/Co(PO₃)₂@NC@NC//AC LIC device. This device was subsequently evaluated for its electrochemical energy storage properties within a voltage window of 0–4.0 V. The working principle of the assembled LIC is illustrated in Figure 6a. During charging, Li⁺ ions either intercalate into Co_1−xS/Co(PO₃)₂@NC or adsorb onto the anode surface, while PF₆^- anions are absorbed onto the AC surface. The discharge process reverses these actions. Figure 6b presents the CV curves of the LICs at scan rates of 5, 10, 20, 30, and 50 mV s⁻¹. Even at 50 mV s⁻¹, a rectangular profile is observed, indicating the capacitive behavior of the LICs [37,38]. The rate performance of the LICs was assessed under various current densities, revealing reversible capacities of 19.15 mAh g⁻¹ even at 2 A g⁻¹, respectively (Figure 6c,d). The assembled optimal Co_1−xS/Co(PO₃)₂@NC//AC LIC delivers a high energy density of 90.47 Wh kg⁻¹ (at 26.28 W kg⁻¹) and a high power density of 504.94 W kg⁻¹ (at 38.31 Wh kg⁻¹). Furthermore, Figure 6e illustrates the long-term cycling stability of the LICs, showing a high cyclic stability of 86.3% retention after 5000 cycles at a current of 2 A g⁻¹, accompanied by a Coulombic efficiency close to 100%. The superiorities of the as-prepared Co_1−xS/Co(PO₃)₂ heterostructures were compared with the reported literature in terms of synthetic methods and electrochemical performance (Table S1). In practical applications, the LICs successfully power a light-emitting diode (LED) to illuminate a “wang” character, further demonstrating the potential of the Co_1−xS/Co(PO₃)₂ heterostructures as an electrode material for Li-ion capacitors.

3. Conclusions

In summary, we have constructed high-performance LICs using Co_1−xS/Co(PO₃)₂@NC as the anode material. This material is synthesized through the chemical blowing-vulcanization of a Co²⁺-PVP precursor followed by phosphorylation treatment. The characteristic heterointerface and 3D honeycomb-like N-doped carbon matrix efficiently boost interfacial charge dynamics, provide spatial confinement for refined particles, and prevent structural degradation during repeated charge/discharge cycles. As a result, the optimal Co_1−xS/Co(PO₃)₂@NC composite demonstrates a stable reversible capacity of 371.8 mAh g⁻¹ after 800 cycles at 1 A g⁻¹, showcasing significant pseudocapacitive contributions. The assembled Co_1−xS/Co(PO₃)₂@NC//AC LIC delivers a power density of 504.94 W kg⁻¹ and an energy density of 90.47 Wh kg⁻¹ within the voltage range of 0–4.0 V. Moreover, they exhibit a high cyclic stability of 86.3% retention after 5000 cycles at 2 A g⁻¹. The facile synthetic method opens new avenues for the fabrication of cost-effective, advanced energy storage devices utilizing Co-based sulfides heterostructures with balanced performance.