Ordered Macroporous MoS₂-Carbon Composite with Fast and Robust Sodium Storage Properties to Solve the Issue of Kinetics Mismatch of Sodium-Ion Capacitors

1 Introduction

The rapid development of electric vehicle and grid energy storage brings out an urgent demand for energy-storage devices delivering high energy and power density.^[1-4] Lithium-ion capacitor (LIC), as a promising intermediate between lithium-ion batteries (LIBs) and supercapacitors (SCs), has bridged performance of both LIBs and SCs.^[5] Typically, LICs are composed of prelithiated battery-type electrode as anode and a capacitor-type material as cathode in an appropriate organic electrolyte, which take advantage of both Faradaic and capacitive mechanisms to realize a higher energy density than SCs and a higher power density than LIBs.^[3-6] Nevertheless, the application of LICs will inevitably be limited by the scarcity of Li sources (20 ppm). The abundance of sodium element (23 000 ppm) in the earth’s crust and redox potential of Na/Na⁺ (−2.71 V vs standard hydrogen electrode) close to that of Li/Li⁺ (−3.04 V) enable sodium-ion capacitors (SICs) to become powerful alternatives for LICs.^{[7, 8]}

Up to now, SICs have already become one kind of the most promising candidates for next-generation energy-storage devices.^[9] However, in view of the larger ionic radius of Na⁺ (1.02 Å) compared with Li⁺ (0.76 Å), the common issue of kinetics mismatch between sluggish battery-type negative electrode and fast capacitive positive electrode is even more severe, thus limiting the energy delivery at the high-power density for SICs. Hence, it is important to develop novel high-rate anode materials to fill up the kinetics gap between two electrodes.^{[3, 10]} Currently, many novel anode materials such as insert-type,^[11-14] redox-type,^[15-17] and alloy-type materials^{[18, 19]} have been explored to promote the development of SICs. These anode materials have some progress in improving the kinetics of Na⁺ charge storage under the premise of shorting the diffusion distance and enhancing the electrical conductivity through reducing the size into nanoscale and introducing the conductive carbon and/or porous structure, even some anode materials, such as V₂O₅,^[14] MoS₂,^[20-23] MXenes,^{[24, 25]} and TiO₂,^{[16, 26, 27]} exhibit the pseudocapacitive behavior. Notably, materials whose energy-storage process is identified as pseudocapacitance undergo reversible electrochemical reactions within significantly short time, even comparable to electric double-layer capacitances (EDLCs) in some cases.^[28] In this regard, well-designed anode materials achieved through the above-state versatile strategies are highly desired to bridge the kinetics gap between anode and cathode of SICs. It is worth noting that studies on bridging the kinetics gap were based on a half-cell referred metal Na, but until recently, no studies addressed this issue on a full-cell system that incorporated anode and cathode in one cell. More importantly, once the kinetics of anode catches up to and even surpass the cathode, one question that always arises: can the energy-power performance of SICs achieve the desired results?

In this work, a MoS₂-carbon composite anode with the ordered macroporous architecture and interlayer-expanded feature was synthesized via a template method. Such favorable structure is demonstrated to have the pseudocapacitance-dominated characteristic for Na⁺ storage, thus yielding a fast kinetics comparable to capacitive double-layer activated carbon. Significantly, SICs were fabricated by employing MoS₂-carbon nanocomposite as anode and a homemade activated carbon as cathode, delivering a high energy density of 91.7 Wh kg⁻¹ at a power density of 250 W kg⁻¹. As the power density is magnified one hundred times to 25 000 kW kg⁻¹, this SICs still deliver an energy density of 40.4 Wh kg⁻¹. In addition, the kinetics gap between anode and cathode of this SIC has been filled up as confirmed by the abnormal variation law of potential swings for anode and cathode from a three-electrode Swagelok cell.

2 Results and Discussion

Ordered macroporous MoS₂-carbon composite was synthesized via a facile template strategy with silica spheres as templates, polyvinylpyrrolidone as the carbon sources, and ammonium tetrathiomolybdate ((NH₄)₂MoS₄) as the MoS₂ precursors, as illustrated in Figure 1a. Typically, monodisperse silica spheres with pore diameter around 337 nm (Figure S1, Supporting Information) were first prepared according to a previous report.^[29] Then, the mixture of (NH₄)₂MoS₄ and polyvinylpyrrolidone (PVP) was coated on the silica spheres, followed by a thermal annealing treatment and HF acid etching. Subsequently, the desired product (MoS₂@PVP derived activated carbon, denoted as MoS₂@PDAC) was achieved with homogeneous honeycomb-like porous structure, which can be confirmed by the scanning electron microscopy (SEM) images (Figure 1b,c). The average macropore size of MoS₂@PDAC is about 334 nm. For comparison, pure MoS₂ and PDAC were also prepared by the same way, displaying the similar honeycomb-like morphology (Figure S2, Supporting Information). In addition, MoS₂@PDAC composites with different ratios of MoS₂ and carbon were synthesized by the change in the proportion of (NH₄)₂MoS₄/PVP precursors. The optimal product of MoS₂@PDAC (determined by the morphological and the electrochemical analysis, Figures S3 and S4, Supporting Information) was selected and discussed below.

Transmission electron microscopy (TEM) image (Figure 1d) further reveals the ordered macroporous structure with the hollow interior of MoS₂@PDAC, which is perfectly inherited from silica spheres. Similar ordered macroporous structure can also be observed in pure MoS₂ and PDAC samples (Figure S5, Supporting Information). High-resolution TEM (HRTEM) images in Figure 1e,f show that the porous skeleton of MoS₂@PDAC was mainly composed by few-layered MoS₂. The resolved MoS₂ few-layers show interlayer-expanded feature with an interlayer spacing of about 1.02 nm (Figure 1f insert). The much larger (002) interlayer distance than 2H-MoS₂ (0.62 nm) is possibly caused by the intercalation of carbon layer into MoS₂ (002) planes to form the MoS₂-carbon monolayer interoverlapped superstructure,^[30] as illustrated in Figure S6, Supporting Information. In comparison, pure MoS₂ displays an interlayer spacing of 0.64 nm (Figure S5, Supporting Information), close to 2H-MoS₂. Furthermore, the annular dark-field scanning transmission electron microscopy (ADF-STEM) and the corresponding element mapping images of MoS₂@PDAC (Figure 1g–i) reveal a uniform hybridization of MoS₂ and amorphous carbon.

X-ray diffraction (XRD) patterns (Figure 2a) of as-prepared samples show that the peaks of pure MoS₂ without the addition of PVP match with JCPDS No. 87-2416 for 2H-MoS₂, and a broad peak centered at 25° for PDAC suggests its amorphous structure. By contrast, the (002) plane peak for 2H-MoS₂ becomes weak, and a new peak for MoS₂@PDAC appears at relative low angle area of 8.46°, corresponding to the (001) reflection with d-spacing for adjacent MoS₂ monolayers of 1.04 nm. The enlarged d-spacing of adjacent MoS₂ monolayers is consistent with the above TEM observation and other reported MoS₂-carbon composites.^[31-33] Noted that the broad (001) peak for MoS₂@PDAC indicates a poorly stacked structure or reduced crystalline size of MoS₂ in carbon matrix.

Raman spectroscopy was employed to further study the structure of the as-prepared MoS₂@PDAC sample, and the result is plotted in Figure 2b. Two characteristic peaks are found around 378.1 and 400.7 cm⁻¹, corresponding to the E¹_2g (in-plane Mo-S phonon) and A_1g (out-of-plane Mo-S) mode of MoS₂.^[34-36] In contrast to pure MoS₂, the blue shift of A_1g as well as the smaller frequency difference (∆k) between E¹_2g and A_1g mode of MoS₂@PDAC can be ascribed to the enlarged interlayer spacing and decreased number of MoS₂ layers.^{[35, 36]} Besides these two peaks, the other two peaks around 1369 and 1579 cm⁻¹ of MoS₂@PDAC are attributed to the D (the defect-induced band) and G bands (the crystalline graphite band) of amorphous carbon, respectively.^[37] The intensity ratio (I_D/I_G) between D-band and G-band is indicative of the degree of graphitic ordering.^[37] I_D/I_G values of MoS₂@PDAC and PDAC are ~1.12 and ~1.06, respectively, indicating the amorphous characteristic of PDAC. The weight ratio of amorphous carbon in MoS₂@PDAC can be calculated to be about 24 wt% according to the thermos-gravimetric analysis (Figure S7, Supporting Information).

The surface area and pore-size characterizations of as-prepared samples were performed by the nitrogen adsorption/desorption isotherms. Figure 2c shows that all samples exhibit type H4 loops, suggesting that the presence of hollow spherical pore structure,^[38] which is consistent with the analysis of SEM and TEM images. These samples exhibit similar pore-size distribution curves (Figure 2d) in which the pores are dominated by large-sized mesopores and macropores. MoS₂@PDAC possesses a Brunauer–Emmett–Teller (BET) surface area value of ~27 m² g⁻¹ and a pore volume of 0.14 cm³ g⁻¹, higher than the value of pure MoS₂ (~13 m² g⁻¹, 0.049 cm³ g⁻¹), but lower than the value of PDAC (~279 m² g⁻¹, 0.45 cm³ g⁻¹). The increment of surface area for MoS₂@PDAC compared with pure MoS₂ is associated with the introduction of carbon.

X-ray photoelectron spectroscopy (XPS) measurement was carried out to disclose the surface characteristics of all the samples. Full-scale XPS spectra (Figure S8, Supporting Information) results confirm the presence of Mo, S, C, N, and O in the MoS₂@PDAC. In the high-resolution Mo 3d XPS spectra (Figure 2e), the binding energies of Mo 3d_5/2 and Mo 3d_3/2 are 229.58 and 232.71 eV for MoS₂@PDAC, 229.67 and 232.80 eV for pure MoS₂, respectively, both corresponding to 2H-MoS₂.^[39] The peaks located at around 226.73 and 226.83 eV for MoS₂@PDAC and pure MoS₂, respectively, are assigned to S 2s.^{[39, 40]} A distinct peak at 235.93 eV in MoS₂@PDAC exhibits a +6 oxidation state, indicating a slight oxidation during the synthesis process.^[41] The binding energies of S 2p doublet at 162.42 and 163.60 eV for MoS₂@PDAC in Figure 2f are similar to the values of pure MoS₂ (162.54 and 163.72 eV), arising from the presence of S²⁻.^[42] From high-resolution C1s spectra (Figure S9, Supporting Information), three fitted peaks of C–C, C–N, and C–O were deconvoluted from C1s for both MoS₂@PDAC and PDAC.^{[42, 43]} The N atom in carbon was generated from the decomposition of PVP during thermal treatment. The N-doped properties can be further confirmed by high-resolution N 1s spectra of MoS₂@PDAC and PDAC, as shown in Figure S10, Supporting Information. In short, the above results from physical-chemical characterizations demonstrate that MoS₂@PDAC has an ordered macroporous nanostructure in which the resolved MoS₂ few-layers with interlayer-expanded feature are confined in the amorphous carbon matrix, which may be the favorable characteristic for enhancing the Na⁺ charge storage properties in aspect of the high capacity and fast electrode kinetics.

The Na⁺ storage property of as-prepared samples was evaluated by half-cell with metal sodium as counter/reference electrode. Cyclic voltammetry (CV) and galvanostatic charge/discharge curves of anode were carried out within a potential range of 0.01–3.0 V (vs. Na⁺/Na). As is shown in Figure 3a, there are various apparent cathodic peaks at 0.92, 0.81, 0.34, and 0.09 V in the initial sodiation cycle. The peaks within 1.5–0.5 V arise from the intercalation of Na⁺ into MoS₂ interlayers and the formation of solid electrolyte interface (SEI) film.^{[20, 21]} The peaks between 0.01 and 0.4 V can be derived from the irreversible transformation from Na_xMoS₂ to Na₂S and Mo nanograins.^[21] During the first desodiation process, a broad anodic peak around 1.7 V is assigned to the oxidation of Na₂S to Na⁺ and S.^{[21, 27, 28]} In the following cycles, the CV curves are composed of a reversible redox couple at around 1.2–3.0 V and a pseudo-rectangular-shaped area at around 0.01–1.2 V, corresponding to the sulfide redox reaction (2Na⁺ + S + 2e⁻ ↔ Na₂S) and the capacitive Na⁺ storage on the boundaries or interfaces of the Mo/NaS_x mixture and Na⁺ insertion/desertion in the carbon matrix, respectively.^{[44, 45]} We also noted that the curves tend to overlap, indicating the reversible sodiation/desodiation of MoS₂@PDAC. Figure 3b shows the galvanostatic charge/discharge curves of MoS₂@PDAC at a current density of 0.05 A g⁻¹. The initial charge/discharge curve displays slight specific capacity decrease caused by the irreversible reaction and the formation of SEI on the anode surface, whereas the followed charge/discharge cycles exhibit overlapping plots, matching with the results with CV test. The charge and discharge profiles feature like a triangular in full work potential region even at the high current densities (Figure S11, Supporting Information), suggesting the fast kinetics analogous to pseudocapacitive behavior.^[1]

From the rate capability test (Figure 3c), MoS₂@PDAC delivers specific capacities of 460.5, 444.2, 422.6, 404.2, 373.4, 346.7, 328.2, 277.5, 227.4, and 213.6 mAh g⁻¹ at current densities of 0.05, 0.1, 0.2, 0.5, 0.8, 1.0, 2.0, 5.0, 8.0, and 10.0 A g⁻¹, respectively. As the current density goes back to 0.5 A g⁻¹, MoS₂@PDAC still preserves a high specific capacity of 382.8 mAh g⁻¹. Obviously, MoS₂@PDAC displays a highest specific capacity retention of 46.4% at a high current density of 10.0 A g⁻¹ compared to 26.6% of pure MoS₂ (26.6%) and 14.1% of PDAC. Most importantly, the rate capability of MoS₂@PDAC is better than that of capacitive activated carbon cathode (36.7% for homemade polyaniline derived porous carbon (PDPC) cathode (Figure S12, Supporting Information), and other typical anode materials for Na⁺ storage (Figure 3d and Figure S13, Supporting Information), such as insert-type materials,^{[11, 13, 46]} alloy-type materials,^{[47, 48]} and redox-type materials.^{[20, 49-52]}

Cycling stability test on MoS₂@PDAC electrode at a low current density of 0.2 A g⁻¹ yields a capacity retention value of 84.8% after 300 cycles, demonstrating outstanding cycling stability of the MoS₂@PDAC electrode (Figure S14, Supporting Information). MoS₂@PDAC also exhibits the remarkable cycling stability at a high current density of 1 A g⁻¹ (Figure 3e). MoS₂@PDAC still maintains a capacity of 300.2 mAh g⁻¹ (corresponding to capacity retention value of 82.5%) after 1000 cycles with nearly 100% columbic efficiency. Despite the initial specific capacities at the low current densities for pure MoS₂ are higher than that of MoS₂@PDAC, the specific capacity of pure MoS₂ drops rapidly to 115.6 mAh g⁻¹ after 200 cycles. For comparison, PDAC exhibits stable cycling stability, but much lower capacity around 150 mAh g⁻¹ is achieved. Based on the above discussion, MoS₂@PDAC exhibits an outstanding electrochemical performance in terms of specific capacity, rate capability, and cycling stability.

Based on this methodology, the capacitive contribution for current response is depict as shaded area through evaluating the pseudocapacitive currents of electrodes. From CV curves of MoS₂@PDAC, pure MoS₂, and PDAC at 0.5, 1.0, 2.0, 5.0, and 10.0 mV s⁻¹, capacitive contribution is marked as shaded area (Figure S17, Supporting Information) and the corresponding capacitive contribution ratios are summarized in Figure 3g. The capacitive contribution of all samples has a unified tendency to grow gradually with the increased scan rate. Compared with MoS₂ and PDAC, MoS₂@PDAC anode exhibits the highest capacitive contribution ratio with an eventually value as 67.78% at 10.0 mV s⁻¹, indicating its fast kinetics in nature.^{[23, 53]}

To provide a greater insight into the Na⁺ diffusion kinetics of MoS₂@PDAC, MoS₂, and PDAC anodes, galvanostatic intermittent titration technique (GITT) measurement at 0.2 A g⁻¹ was performed to quantify the Na⁺ diffusion coefficient (D_k, cm² s⁻¹) in the electrodes according to Fick’s second law (Figures S18 and S19, Supporting Information).^{[55, 56]} According to Figure 3h, D_k in MoS₂@PDAC has a general tendency to decrease during the sodiated process, while a reverse trend is shown for the desodiated process. The variation of D_k during the sodiated and desodiated process is associated with different bonding energy of redox sites. Regardless of the similar trend during the charging/discharging process, D_k of MoS₂@PDAC hybrid (0.18–2.6 × 10⁻⁹ cm² s⁻¹) is apparently higher than MoS₂ (0.012–0.88 × 10⁻⁹ cm² s⁻¹), which may be associated with the synergistic effect between PDAC with much higher D_k value (1.03–6.61 × 10⁻⁹ cm² s⁻¹) and MoS₂ with relatively low D_k value. The fast Na⁺ diffusion kinetics of MoS₂@PDAC further confirmed by electrochemical impedance spectroscopy (EIS) plotted in Figure 3i with a high-phase angle (exceeding 45°) at low-frequency area compared with MoS₂.

Furthermore, to probe the excellent electrochemical properties of MoS₂@PDAC anode during the Na⁺ storage, ex situ measurements including XRD, Raman, XPS, and TEM were carried out. The ex situ XRD patterns (Figure S20, Supporting Information) of MoS₂@PDAC anode shows the irreversible phase transformation of MoS₂ in the initial sodiated state and the sole redox couple (S + 2Na ↔ Na₂S) in the followed sodiated and desodiated process.^[22] The irreversible phase transformation of MoS₂ is also confirmed by ex situ Raman patterns (Figure S21, Supporting Information) where the characteristic peaks E¹_2g and A¹_g of MoS₂ completely disappear in the initial sodiated process and no longer show up in the following desodiated process. Apparently, D and G peaks display no obvious difference before and after long cycling process, suggesting the structural stability of carbon matrix of MoS₂@PDAC anode.^{[57, 58]}

Ex situ XPS spectra of MoS₂@PDAC anode provide more details on the redox process during the different sodiated/desodiated state (Figure S22, Supporting Information). The high-resolution Mo 3d spectra (Figure S23a, Supporting Information) shows that Mo 3d peaks are blue shifting after sodiation, suggesting the formation of metallic Mo caused by decomposed of MoS₂. After the desodiation, the Mo 3d peak does not return original position, and two new peaks around 232.4 and 235.5 eV appear, which are characteristic for Mo 3d_5/2 and Mo 3d_3/2 of Mo⁶⁺, respectively. The emerging peaks for metallic Mo and Mo⁶⁺ suggest the irreversible decomposition of MoS₂, which suggests that the followed reversible conversion reaction between Na₂S and S is feasible. After 10th and long cycles, the peaks for Mo⁶⁺ still preserved. Mo⁶⁺ is mainly stemmed from MoO₃ caused by of metallic Mo in air, which is also observed in other studies.^[59] The irreversible decomposition of MoS₂ during initial sodiation process is also confirmed by blue shifting of S 2p (Figure S23b, Supporting Information), where S²⁻ from Na₂S is detected. The S⁴⁺ peak appears after desodiation and long-term cycles, which can be attributed to adsorbed SO₃²⁻ functional groups generated by the desodiation treated sample exposed in air.^[60] High-resolution XPS spectra for C 1s, O 1s, F 1s, and Na 1s of MoS₂@PDAC anode (Figure S23c–f, Supporting Information) further show that the SEI layer for MoS₂@PDAC is mainly composed by organic (RO-Na and RO-COONa) and inorganic (Na₂CO₃ and NaF) ingredients, which is consistent with other reports.^{[61, 62]} After desodiation and long-term cycles, C1s, O1s, Na1s, and F1s spectra show little change, suggesting the structure stability of SEI layer.

TEM image (Figure S24, Supporting Information) shows that MoS₂@PDAC anode still preserves the ordered macroporous structure after the long cycles and is uniformly covered by SEI, suggesting the stable pristine architecture of MoS₂@PDAC. Therefore, it can be concluded that even initial irreversible decomposition of MoS₂ and followed repetitive chemical conversion reaction, the pristine porous architecture of MoS₂@PDAC electrode remains unchanged, which is responsible for excellent cycling performance of MoS₂@PDAC electrode. Furthermore, electrical impedance spectroscopy (EIS) measurements were employed to monitor the impendence change in MoS₂@PDAC electrode during the cycling. Figure S25, Supporting Information shows that the gradually increased charge transfer/SEI resistances at the high-frequency area are the main reason for the capacity decay of MoS₂@PDAC electrode by during the long cycles. Overall, such extraordinarily Na⁺ charge-storage properties of MoS₂@PDAC electrode may bridge the kinetics gap between capacitive porous carbon cathode in full-cell type SICs.

A SIC was then assembled with MoS₂@PDAC as anode and PDPC as cathode in a NaClO₄ based organic electrolyte (Figure 4a). More structure information on PDPC can be found in our previous work.^[63] The optimized mass ratio between PDPC and MoS₂@PDAC is 1:1 within an operating potential window of 1.0–4.0 V, displaying best rate capability and cycling life (Figure S26, Supporting Information) among different anode/cathode mass ratio. CV curves of MoS₂@PDAC//PDPC SIC under different scan rates were recorded in Figure 4b, showing the quasi-rectangular shape. Even the sweep rate increases to 200 mV s⁻¹, the near-rectangular CV profile is still preserved, indicating the good capacitive behavior of as-fabricated SIC. The rectangular feature of CVs is related to the coupling effect between pseudocapacitance-like MoS₂@PDAC and double-layer capacitive PDPC. Galvanostatic charge/discharge (GCD) curves at different current densities in Figures 4c and S27, Supporting Information exhibit quasi-triangular shape. The specific capacities of as-fabricated SIC are 44.0, 42.5, 39.8, 37.8, 36.7, 33.2, and 26.3 mAh g⁻¹ at 0.1, 0.2, 0.5, 0.8, 1.0, 2.0, and 5.0 A g⁻¹, respectively (Figure 4d). Even the charge/discharge current density has a magnification of 100 (up to 10 A g⁻¹), this SIC still delivers a specific capacity of 19.4 mAh g⁻¹ with a high-capacity retention ratio of 44.1%, indicating the good rate capability of MoS₂@PDAC//PDPC SIC. Furthermore, as-fabricated MoS₂@PDAC//PDPC SIC displays the excellent long-term cycling stability (Figure 4e). After 10 000 cycles at a current density of 2 A g⁻¹, this SIC still preserves the 88.4% of its initial capacity. During the whole cycling process, the columbic efficiency is nearly 100%. Figure 4f illustrates the Ragone plot (energy density vs power density) of as-fabricated SIC. At a power density of 250 W kg⁻¹, as-fabricated MoS₂@PDAC//PDPC SIC delivers a high energy density of 91.7 Wh kg⁻¹. Even at an ultra-high-power density of 25 000 kg⁻¹, MoS₂@PDAC//PDPC SIC still delivers a high energy density of 40.4 Wh kg⁻¹. Ragone plots also show that the energy and power densities of this SIC are highly comparable to a variety of representative literature reports of SICs (Table S1, Supporting Information and Figure S28, Supporting Information), including Na₂Ti₃O₇-CT//GF,^[11] hard carbon//graphene,^[13] V₂O₅-CNT//AC,^[14] MoS₂-C//C-HNT,^[64] N-Ti₃C₂T_x//AC,^[65] NaTi₂(PO₄)₃-rGO//AC,^[66] and TiO₂-C//ZDPC.^[67]

To further understanding the excellent electrochemical behaviors of MoS₂@PDAC//PDPC SIC with respect to high-rate and long-term cycling stability, a three-electrode Swagelok cell was employed with metal Na as the auxiliary electrode (Figure S29, Supporting Information). Figure 5a depicts the CV results for a 1.00–4.00 V voltage window plotted for the full devices (red) as well as the individual electrodes (positive electrode in blue and negative electrode in orange) at a sweep rate of 2 mV s⁻¹. The MoS₂@PDAC anode operates within 0.14–1.19 V versus Na/Na⁺, while the PDPC cathode works within 2.19–4.14 V versus Na/Na⁺. As the scan rate increased to 100 mV s⁻¹, MoS₂@PDAC//PDPC SIC still preserved the rectangular-shaped CV (Figure S30, Supporting Information). In this case, the potential swings for PDPC cathode and MoS₂@PDAC are 0.40–1.19 V versus Na/Na⁺ and 2.19–4.40 V versus Na/Na⁺. No sign of Na plating and apparent electrolyte decomposition occurs, as demonstrated by the working potential ranges for anode and cathode within stable area.

The potential swings of anode and cathode under the different charge/discharge rates were further monitored by three-electrode Swagelok cell. Figure 5b,c shows that the down potential for anode gradually increases from 0.18 to 0.60 V versus Na/Na⁺ as the current density increase from 0.1 to 10 A g⁻¹. The down potentials of MoS₂@PDAC anode are higher than the plating potential of metal Na, indicating the no safety concern of this SIC from Na plating. We also noted that the up potential of PDPC cathode also gradually increases with the increased current density. As the current density increases over 5 A g⁻¹, the up potential exceeds 4.5 V versus Na/Na⁺, possibly leading the decomposition of electrolyte, which may be alleviated by the reducing the working potential of full cell or using the electrolytes with high working voltage. Most interestingly, the potential swing window of MoS₂@PDAC anode becomes narrower from 1.17 to 0.60 V as the current density amplified from 0.1 to 10 A g⁻¹, while the potential swing window of cathode becomes broader from 1.83 to 2.40 V (Figure 5d). The variation law of potential swings of this SIC is different from other’s works on lithium-ion capacitor (graphite//AC LIC, where the kinetics of graphite anode is noticeably more sluggish than AC cathode) in which gradually broaden working potential range of graphite anode and negligible potential change in AC cathode were observed.^[68] This can be attributed to that the kinetic mismatch issues between anode and cathode in our SIC are solved (Figure 5e). In this SIC, the faradaic sodiated/desodiated process of MoS₂@PDAC anode shows the pseudocapacitive behavior, which is not inferior to the non-faradaic capacitive PDPC cathode, shown in Figure 3d. During the high charged/discharged current densities, the rate capability of SIC depends on the PDPC cathode instead of MoS₂@PDAC anode, leading to a high overpotential of PDPC cathode. In view of the relatively low specific capacity of capacitive PDPC, overpotential issue becomes serious. As shown in Figure 5d, the PDPC cathode potential swing window is widened by 30.7%; however, the MoS₂@PDAC anode potential window is reduced by 33.4% at a current density of 10 A g⁻¹. The above results suggest that as the kinetics of anode is improved sufficiently to match the kinetics of cathode, the issues from capacity mismatch between anode and cathode should not be ignored. The potential range of cathode side will broaden to provide additional capacity to match anode side. On the contrary, the specific capacity for anode largely outweighs cathode side, and the potential range of anode will be no change or even shrinkage to match cathode side. In this case, the overpotential issues of cathode will arise and even become more serious due to the relatively low specific capacity of cathode, presenting new challenge on developing high-performance SICs.

Furthermore, the potential swings of MoS₂@PDAC anode and PDPC cathode in as-assembled SIC were recorded by a three-electrode Swagelok cell during long-term cycling (Figure S31, Supporting Information). Figure 5f,g exhibits that the up and down voltages for MoS₂@PDAC anode and PDPC cathode are in stable area, indicating no sign of Na metal deposition and apparent electrolyte decomposition. The potential swing windows for both anode and cathode are little changed during the constant charging/discharging process, thus leading to the long-term cycle stability of as-fabricated MoS₂@PDAC//PDPC SIC. EIS results (Figure S32, Supporting Information) further demonstrated that the capacity decay of as-assembled SIC is mainly caused by the increased interface resistance and the step-down Na⁺ diffusion after the long cycles.

3 Conclusion

In summary, we used a facile sacrificing template method to successfully synthesize ordered macroporous MoS₂-carbon composite that the few-layered MoS₂ featured interlayer-expanded properties are confined in the amorphous carbon matrix. This rationally designed architecture of MoS₂@PDAC increases atomic interface contact/interaction with Na⁺, facilitating ion diffusion and charge transfer. Specifically, MoS₂@PDAC anode delivers a high-rate capability, which is not inferior to double-layer capacitive carbon cathode. Three-electrode Swagelok cell confirms that the kinetics gap between anode and cathode has been filled up via the MoS₂@PDAC anode with fast kinetics. SICs were fabricated by employing MoS₂@PDAC as anode and homemade active carbon as cathode. The assembled SIC delivers a high energy density of 40.4 Wh kg⁻¹ at an extremely high-power density of 25 000 W kg⁻¹ and excellent cycling performance exceeding 10 000 cycles, indicating its potential for practical applications. Our work demonstrates that it is true that accelerating the Na⁺ kinetics of anode via engineering the structure in nanoscale matches the capacitive carbon cathode, showing its significant potential for practical application in SICs and other metal-ion capacitors.

4 Experimental Section/Methods

Conflict of Interest

The authors declare no conflict of interest.