English

• Sodium-ion batteries (SIBs) emerge as promising alternatives for energy storage systems and low-speed electric vehicles, owing to their cost-effectiveness. However, the energy density and stability of the cathode are the primary challenges limiting the commercialization of SIBs [1-3]. Among various SIB cathodes, the sodium superionic conductor (NASICON)-type polyanionic materials feature a unique three-dimensional lantern-shaped framework that provides numerous interconnected interstitial sites, thus facilitating the rapid transport of sodium ions [4-6]. Furthermore, the redox potential of the NASICON cathodes can be controlled through the adjustment of the composition of transition metal (TM) elements, thereby regulating the energy density of the cathode [7,8]. So far, Na₃V₂(PO₄)₃ (NVP) has attracted extensive attention and remains extremely attractive due to its excellent structural stability and fast sodium ion kinetics. At a potential of 3.4 V, NVP can release two Na⁺ for intercalation/deintercalation, exhibiting a theoretical capacity of 117 mAh/g [9]. Unfortunately, the expensive price and toxicity of V limit the widespread application of NVP. In this context, substituting V in NVP with safer and more economical TM ions, such as Mn²⁺, Ti⁴⁺, Cr³⁺, and Fe²⁺ is an effective solution [10-12].

  The incorporation of multivalent transition metals into the NASICON structure can significantly enhance its capacity and facilitate multi-electron redox reactions. Manganese element possesses a higher redox potential, which not only enables a higher energy density but also renders it more competitive in energy storage applications [13,14]. Representatively, a series of Mn-based NASICON cathodes, such as Na₄MnV(PO₄)₃ [15], Na₃MnTi(PO₄)₃ [16], Na₃MnZr(PO₄)₃ [17] and Na₄MnCr(PO₄)₃ [10] were developed, which demonstrated the flexible tunability of NASICON and promoted the further development of Mn-based NASICON-type cathodes. Among them, the Na₃MnTi(PO₄)₃ material, characterized by three pairs of redox peaks, can achieve a theoretical capacity of 176 mAh/g [18]. Mai et al. [19,20] synthesized NMTP hollow microspheres by spray drying method, and Zhang et al. [21] prepared rGO-modified NMTP by conventional sol-gel method, both achieving a three-electron reaction. Therefore, NMTP is regarded as a promising high-energy cathode material. However, the intrinsic conductivity of NMTP is unsatisfactory, necessitating microstructural optimization, surface modification, and crystal structure tuning [22,23]. Zhao et al. [24] implemented a non-stoichiometric strategy to effectively suppress Na⁺/Mn²⁺ cation mixing, significantly alleviating the voltage hysteresis and substantially enhancing the cycling stability of NMTP. In addition, Hu et al. [25] developed a Mo-doping strategy to reduce the intrinsic anti-site defect (IASD) concentration of NMTP crystals, effectively suppressing the voltage hysteresis and capacity loss of NMTP. Cao et al. introduced Cr³⁺ into NMTP crystals to lengthen the Na2-O bond and expand the Na ion diffusion channels, thereby improving the structural stability and Na ion diffusion kinetics [26]. Furthermore, a conventional carbon coating can enhance electron transport on the material surface and buffer the volume change during sodium ion extraction and insertion. However, noncontinuous carbon coatings cannot ensure the intergranular electron transportation, thus limiting the full potential of cathode [27-29]. Additionally, ion diffusion from the cathode surface to its interior can be augmented by incorporating a superior ion conductor. Achieving a balance between ionic and electronic conductivities enables batteries to operate at high current rates while maintaining exceptional cycle performance [30,31].

  Recently, a novel class of two-dimensional transition metal carbides/nitrides (MXenes) have garnered widespread attention in electrochemical energy storage due to their unique physical and chemical properties, remarkable electrical conductivity, excellent mechanical stability and extensive specific surface area [32,33]. For cathode materials, the incorporation of MXenes could create a unique heterostructure, enhance both the electronic and ionic conductivity of the cathode by establishing a bicontinuous conductive network, and alleviate the stress of the cathode particles during the sodium ion insertion and deinsertion process, thereby preserving the integrity of the crystal structure [34,35]. In addition, the high specific surface area, high conductivity and abundant surface functional groups of MXene enhance the pseudocapacitive charge storage on the cathode surface, thereby enhancing the power density and rate performance of cathode [36,37].

  Herein, a novel strategy was developed for embedding Na₃MnTi(PO₄)₃ crystals within a electronic/ion mixed-conducting interface composed of amorphous carbon and Ti₃C₂-MXene, thereby enhancing both electronic and ionic transport kinetics at the interface. The solid-solution Ti₃C₂-MXene guides the confined growth of NMTP, and consequently shortening the diffusion distance of sodium ions in the crystal. Concurrently, the diminished diffusion barrier for sodium ions across the MXene layers further enhances the rapid ion transport. Therefore, the Na⁺ diffusion coefficient in NMTP/C@Ti₃C₂ is increased by an order of magnitude compared with NMTP/C. A reversible capacity of 158.2 mAh/g was obtained at 0.2 C (1 C = 176 mA/g), corresponding to an energy density of 466.6 Wh/kg. Meanwhile, the voltage hysteresis phenomenon of NMTP/C@Ti₃C₂ is significantly suppressed and the energy efficiency is increased. The designed mixed-conducting interphase can not only significantly improve the rate performance of NMTP, but also increase the cycle stability. After 250 cycles at 1 C, the capacity remained at 135.1 mAh/g, corresponding to 93.6% of the initial capacity. Furthermore, the energy density of 450 Wh/kg can be provided in the lab-fabricated full battery, showing outstanding competitiveness among SIB cathode materials.

  The NMTP/C@Ti₃C₂ materials were synthesized via an ultrasonic-assisted sol-gel method, with a detailed description of the preparation process available in Fig. S1 (Supporting information). Fig. 1a presents a schematic diagram of the NMTP/C@ Ti₃C₂ structure. The Rietveld refinement analysis for X-ray diffraction (XRD) were applied to study the crystal structure and phase purity of NMTP/C and NMTP/C@Ti₃C₂, as illustrated in Fig. S2 (Supporting information) and Fig. 1b, respectively. The diffraction peaks of NMTP crystals can be well indexed to the NASICON-structured Na₃V₂(PO₄)₃ phase (PDF#97-024-8140) in Fig. S3 (Supporting information). Both materials exhibit the distinctive rhombohedral NASICON structure of the R-3c space group, confirming the synthesis of pure phase NMTP crystals. Detailed Rietveld refinement results, including lattice parameters, atomic coordinates and occupancy are provided in Tables S1 and S2 (Supporting information). The lattice parameters of NMTP/C@Ti₃C₂ (a = 8.826 Å, c = 21.735 Å, V = 1466.4 ų) are slightly higher than those of NMTP/C (a = 8.821 Å, c = 21.702 Å, V = 1462.4 ų), facilitating Na⁺ extraction/insertion.

  Figure 1

  

  Figure 1.  (a) Schematic diagram and (b) the Rietveld refinements of the XRD patterns of NMTP/C@Ti₃C₂ material. (c) TGA curves and (d) Raman spectra of NMTP and NMTP/C@Ti₃C₂. (e) The BET test results of NMTP/C@Ti₃C₂. (f) Electric conductivity of the NMTP and NMTP/C@Ti₃C₂. High-resolution (g) Mn 2p and (h) Ti 2p XPS spectra of NMTP/C and NMTP/C@Ti₃C₂.

  DownLoad: Full-Size Img PowerPoint

  The carbon content in the amorphous carbon coating formed by the pyrolysis of the citric acid on the NMTP surface was determined via thermogravimetric analysis (TGA). As shown in Fig. 1c, the carbon contents of NMTP/C and NMTP/C@Ti₃C₂ are 10.26 and 12.43 wt%, respectively. Between 500 ℃ and 550 ℃, a slight increase in sample mass was observed, corresponding to the oxidation of TM ions. Raman spectroscopy was utilized to further analyze the degree of graphitization of the materials (Fig. 1d). The D and G bonds are observed at 1350 and 1594 cm⁻¹, respectively. The lower I_D/I_G ratio of NMTP/C@Ti₃C₂ (0.928) compared to NMTP/C (0.949) suggests that Ti₃C₂ modification enhances the graphitization and electronic conductivity of NMTP composites. The specific surface area and pore volume of both materials were analyzed by nitrogen adsorption/desorption method (Fig. 1e and Fig. S4 in Supporting information). NMTP/C@Ti₃C₂ exhibits a larger specific surface area (10.94 m²/g) than NMTP/C (9.13 m²/g) and a smaller pore volume. The Ti₃C₂ MXene sheets in the mixed solution can limit the aggregation of NMTP particles during sintering, and increase the surface area, facilitating the Na ion transport between NMTP particles and the electrolyte. In addition, the electric conductivity of the materials were tested by electrochemical impedance spectroscopy (EIS), and the methods and results detailed in Supplementary Information (Table S3 in Supporting information). As shown in Fig. 1f, the electric conductivity of NMTP/C@Ti₃C₂ is 3.34 × 10⁻³ S/cm, which is 2.3 times that of NMTP/C, proving that Ti₃C₂ can significantly enhance the electric conductivity of NMTP materials.

  The chemical state and bonding structure of the material surface were investigated via X-ray photoelectron spectroscopy (XPS). As depicted in Fig. S5 (Supporting information), the signals of Na, Mn, Ti, P, O, and C elements are all captured in the XPS survey spectra. As shown in Fig. 1g, the binding energy position of Mn 2p_3/2 and Mn 2p_1/2 and the ratio of Mn²⁺ to Mn³⁺ peak intensity are similar. Specifically, the binding energies for Mn²⁺ are 641.4 eV (Mn 2p_3/2) and 653.5 eV (Mn 2p_1/2), while those for Mn³⁺ are 642.9 eV and 654.5 eV, respectively. Satellite peaks at 646.7 and 657.6 eV are typical in many transition metal compounds. The Ti 2p_3/2 and Ti 2p_1/2 peaks binding energies of NMTP/C@Ti₃C₂ are the same as those of NMTP/C (Fig. 1h). The binding energies of Ti 2p_3/2 and Ti 2p_1/2 of Ti³⁺ are located at 458.9 and 465 eV, and those of Ti⁴⁺ are located at 460 eV and 465.8 eV, respectively. Notably, the peak intensity ratio of Ti³⁺ to Ti⁴⁺ has altered due to Ti₃C₂ in the mixed-conducting interphase. To confirm the thermal stability of Ti₃C₂ during synthesis, pure Ti₃C₂ powder was collected through centrifugation, drying, sintering, and the structure and chemical information were characterized. Fig. S6 (Supporting information) presents the XRD and XPS results for pure Ti₃C₂, affirming its structural integrity and chemical stability throughout NMTP/C@Ti₃C₂ production.

  The microscopic morphology and microstructure of the materials were characterized and analyzed by scanning electron microscopy (SEM) and transmission electron microscopy (TEM). SEM images verified that NMTP particles are supported on two-dimensional MXene sheets. As shown in Figs. 2a and d, both materials exhibit irregular morphological characteristics, with secondary particle sizes predominantly ranging from a few microns to over ten microns. And the particle size of the NMTP/C@Ti₃C₂ material is significantly smaller than that of NMTP/C, primarily because the Ti₃C₂-MXene guides the restricted growth and reduces the agglomeration of NMTP, thereby shortening the transmission path of sodium ions. Moreover, the sheet structure of Ti₃C₂ facilitates the formation of a cross-linked network, ultimately enabling rapid electron/ion conduction. Similarly, TEM images also exhibit irregular microstructure features (Figs. 2b and e). The obvious NMTP lattice fringe and the carbon layer formed by pyrolysis can be observed in high-resolution TEM images. Figs. 2c, f and g are partial enlarged views of HR-TEM images, respectively. It can be observed that the 0.62 nm interplanar spacings of NMTP/C correspond to the (012) plane of the NASICON structure (Fig. 2c). The lattice fringes of 0.37 and 0.62 nm can be observed in NMTP/C@Ti₃C₂, corresponding to (113) and (012) planes (Fig. 2f). The crystallization orientation of NMTP is influenced by the cross-linked network of Ti₃C₂ and amorphous carbon. As shown in Fig. 2g, lattice fringes of 1.1 nm, which correspond to the (002) crystal plane of Ti₃C₂, can be retrieved [38,39], demonstrating that Ti₃C₂ sheets are uniformly dispersed throughout amorphous carbon. High-angle annular dark field scanning TEM (HAADF-STEM) images provide additional evidence that NMTP/C@Ti₃C₂ are irregular solid particles (Fig. 2h). STEM energy dispersive spectroscopy (STEM-EDS) under a dark field proves that the elements of Na, Mn, Ti, P, O and C are uniformly distributed in the NMTP/C@Ti₃C₂.

  Figure 2

  

  Figure 2.  (a) SEM, (b) TEM, (c) HRTEM images and FFT pattern of NMTP/C material. (d) SEM, (e) TEM, (f, g) HRTEM images and FFT pattern. (h) HAADF-STEM image and the corresponding elemental mappings of NMTP/C@Ti₃C₂ material.

  DownLoad: Full-Size Img PowerPoint

  Before evaluating the electrochemical performance of NMTP/C@Ti₃C₂, pure Ti₃C₂ was configured into a half-cell as a cathode to assess the electrochemical activity. Its rate performance and cyclic voltammetry (CV) were tested within a voltage range of 1.5–4.3 V (Fig. S7 in Supporting information). At various discharge currents, the capacity of pure Ti₃C₂ is nearly negligible. Concurrently, the CV curves display an extremely weak redox current, further demonstrating the minimal electrochemical activity of pure Ti₃C₂. Subsequently, the NMTP/C@Ti₃C₂ was assembled into a half-cell to assess its electrochemical properties. Firstly, cyclic voltammetry (CV) was implemented to investigate the redox process of the cathode at a scan rate of 0.1 mV/s (Fig. 3a and Fig. S8 in Supporting information). The highly overlapping CV curves also show an obvious three-electron reaction. NMTP/C@Ti₃C₂ exhibits three pairs of redox peaks at 4.11/3.96 V (Mn^3+/4+), 3.69/3.44 V (Mn^2+/3+) and 2.24/1.95 V (Ti^3+/4+), showing a lower electrochemical polarization than NMTP/C. Then, galvanostatic charge/discharge (GCD) tests at 0.2 C for NMTP/C and NMTP/C@Ti₃C₂ were conducted, respectively. Fig. 3b shows the discharge capacity of NMTP/C@ Ti₃C₂ at 0.2 C is 158.2 mAh/g, which is slightly lower than that of NMTP/C (159.2 mAh/g). The GCD curve for NMTP/C reveals a distinct discharge plateau between 2.6 V and 2.2 V, corresponding to a capacity of ~25 mAh/g, which constitutes 15.7% of the total capacity. And NMTP/C also exhibits a distinct charging plateau at 3.8–4.0 V. In contrast, the plateaus around 2.5 V and 3.9 V are not obvious within the GCD curves of NMTP/C@Ti₃C₂. Moreover, the plateaus of NMTP/C@ Ti₃C₂ at 4.0 V (Mn^3+/4+) and 3.4 V (Mn^2+/3+) are significantly extended, enhancing the capacity at high-voltage region, thereby increasing the energy density. In addition, a significant gap between the charge and discharge curves of NMTP/C indicates substantial voltage hysteresis. Currently, the voltage hysteresis of Na₃MnTi(PO₄)₃ cathode is primarily attributed to Na⁺/Mn²⁺ cation mixing during the material preparation process, which increases the diffusion barrier of sodium ions around defects and inhibits the extraction and insertion processes of Na⁺[24,25]. As shown in Fig. S9 (Supporting information), for the NMTP/C cathode, an unfavorable charge/discharge plateau occurs at 3.9 V (charging-state, region Ⅰ) and ~2.5 V (discharging-state, region Ⅲ) due to voltage hysteresis, generating a low energy efficiency (84.47%, region Ⅱ). In contrast, the unfavorable plateau length of NMTP/C@Ti₃C₂ at the same position is significantly reduced, whereas the capacity of the high-voltage platform is substantially enhanced, resulting in a significant improvement in energy efficiency to about 92.13%. These results suggest that the mixed-conducting interphase can effectively regulate NMTP crystal growth and minimize defect formation. Compared to early NMTP and other NASICON-structured cathode materials [21,40-49], the synthesized NMTP/C@Ti₃C₂ exhibits a higher level of energy density (Fig. 3c and Table S4 in Supporting information).

  Figure 3

  

  Figure 3.  (a) CV curves of NMTP/C@Ti₃C₂ at 0.1 mV/s. (b) GCD curves of NMTP/C and NMTP/C@Ti₃C₂ at 0.2 C. (c) Comparison of the NMTP/C@Ti₃C₂ with other reported NASICON-type SIBs. (d) Rate performance of NMTP/C and NMTP/C@Ti₃C₂. (e) GCD curves of NMTP/C@Ti₃C₂ at different current densities. (f) The average discharge voltage of NMTP/C and NMTP/C@Ti₃C₂ at different rates. Cycling performance of NMTP/C and NMTP/C@Ti₃C₂ at (g) 1 C and (h) 5 C.

  DownLoad: Full-Size Img PowerPoint

  Figs. 3d and e and Fig. S10 (Supporting information) exhibit the rate performance and charge/discharge curves at different current densities, revealing a distinct three-electron reversible reaction. The NMTP/C@Ti₃C₂ displays the capacities of 158.2, 151.3, 147.5, 144.3, 139.3, 131.5 and 105.9 mAh/g from 0.2 C to 20 C, respectively. And NMTP/C@ Ti₃C₂ shows an energy density of 466.6 Wh/kg at 0.2 C. In contrast, the discharge capacity of NMTP/C shows obvious decreases with increasing current density: 159.2, 144.8, 135.2, 124.5, 113.3, 94.7 and 49.2 mAh/g from 0.2 C to 20 C, proving that the former has better rate performance. Remarkably, NMTP/C@Ti₃C₂ recovers 98.7% of its initial capacity when the current returns to 0.2 C, also higher than NMTP/C (95%), demonstrating that NMTP/C@Ti₃C₂ has higher stability. NMTP/C@Ti₃C₂ exhibits a longer high-voltage plateau at different current densities, which boosts the redox kinetics of Mn^3+/4+ and Mn^2+/3+. Both materials demonstrate an increasing ~2.5 V plateau with the increase of current density, accompanied by more pronounced voltage hysteresis. Moreover, Fig. 3f illustrates the average discharge voltages at different current densities. NMTP/C@Ti₃C₂ has a higher average discharge voltage, indicating that the artificial mixed-conducting interphase can reduce the polarization of NMTP during the charge/discharge process. Furthermore, NMTP/C@Ti₃C₂ demonstrates exceptional cycling stability (Figs. 3g and h). After 250 cycles at 1 C, it retains a capacity of 135.1 mAh/g with a retention rate of 93.6%, much higher than that of NMTP/C (76.1 mAh/g and 52.3%). After 500 cycles at 5 C, NMTP/C@Ti₃C₂ maintains a capacity of 120.9 mAh/g (88.5%), whereas NMTP/C retains only 63.4 mAh/g (44.69%). The average discharge voltage of NMTP/C@Ti₃C₂ is significantly higher than that of NMTP/C by approximately 0.5 V, and it has higher stability (Fig. S11 in Supporting information). The mixed-conducting interphase effectively mitigates polarization in NMTP during charge and discharge cycles. Fig. S12 (Supporting information) depicts the charge/discharge curve evolution during 1 C and 5 C cycles, respectively, with NMTP/C@Ti₃C₂ showing less variation, thus confirming the enhanced cycling stability provided by the C@Ti₃C₂ layer.

  Electrochemical impedance spectroscopy (EIS) was utilized to evaluate the electron/ion diffusion kinetics of the cathode during cycling. Figs. 4a and b show the EIS spectra of the two materials at different temperatures. The diffusion activation energies of the materials are fitted according to the Arrhenius equation, calculated as shown in Supporting information. It is observed that the diffusion activation energy of NMTP/C@Ti₃C₂ is lower than that of NMTP/C (0.42 < 0.39 eV), indicating superior sodium ion diffusion behavior. Fig. 4c and Fig. S13 (Supporting information) show the Nyquist plots of different cycle numbers. The impedance of NMTP/C@ Ti₃C₂ remains invariably lower than that of NMTP/C. After 200 cycles, NMTP/C@Ti₃C₂ shows only a marginal increase in impedance, significantly less than that observed in NMTP/C. The results of the EIS fitting analysis for different cycle times are shown in Fig. 4d and Table S5 (Supporting information). After 200 cycles, the electrode/electrolyte interphase impedance (R_CEI) of NMTP/C increased by 278.5%, and the charge transfer resistance (R_ct) increased by 89.2%. Correspondingly, the NMTP/C@Ti₃C₂ increased by 16.3% and 41.5%, respectively, indicating that C@Ti₃C₂ can significantly suppress the impedance growth on the NMTP surface. Consequently, NMTP/C@Ti₃C₂ has greater charge transfer efficiency and higher Na⁺ diffusion rate, leading to enhanced electrochemical performance. Additionally, Fig. S14 (Supporting information) show the Nyquist plots at different discharge states, indicating that NMTP/C@Ti₃C₂ demonstrates consistently lower electrochemical reaction impedance than NMTP/C at different state of charge.

  Figure 4

  

  Figure 4.  Nyquist plots at different temperatures and Arrhenius plots of (a) NMTP/C and (b) NMTP/C@Ti₃C₂. (c) The Nyquist plots and (d) the impedance change of NMTP/C@Ti₃C₂ at different cycling numbers. (e) GITT curves and Na⁺ diffusion coefficient and (f) overpotential of NMTP/C and NMTP/C@Ti₃C₂ cathodes. (g) CV curves of NMTP/C@Ti₃C₂ obtained at different scan rates of 0.1, 0.2, 0.3, 0.4, and 0.5 mV/s. (h) The log(i) versus log(v) plots of NMTP/C@Ti₃C₂ at different redox states. (i) Capacitance contribution of NMTP/C and NMTP/C@Ti₃C₂.

  DownLoad: Full-Size Img PowerPoint

  Also, the galvanostatic intermittent titration technique (GITT) was implemented to investigate the Na⁺ diffusion characteristics during the charge/discharge process. Fig. 4e illustrates the GITT curves and the apparent Na⁺ diffusion coefficient (D_Na) of NMTP/C and NMTP/C@Ti₃C₂ respectively. The methodology for calculating D_Na is detailed in Supporting information. It can be found that the D_Na of NMTP/C@Ti₃C₂ is superior to that of NMTP/C during the charge/discharge process. The D_Na of NMTP/C@Ti₃C₂ is stable around 10⁻⁹ cm²/s in the Ti^3+/4+ redox region (1.5–2.5 V), which is almost one order of magnitude higher than that of NMTP/C (10⁻¹⁰ cm²/s). Furthermore, the D_Na of NMTP/C@Ti₃C₂ fluctuates in the range of 10⁻⁹~10⁻¹² cm²/s in the redox region of Mn^2+/3+/4+, which is also much higher than that of NMTP/C (10⁻¹⁰–10⁻¹³ cm²/s), however, the kinetic process in this region is lower than that of Ti. And the overpotential of Ti³⁺/Ti⁴⁺ redox is smaller than that of Mn²⁺/Mn⁴⁺ redox (Fig. 4f). Meanwhile, the polarization of NMTP/C@Ti₃C₂ exhibits reduced polarization compared to NMTP/C in each GITT titration step.

  Since amorphous carbon and Ti₃C₂ can provide larger active surfaces and defects for NMTP, the pseudocapacitive effect of NMTP electrodes cannot be ignored [50,51]. The CV curves at different scan speeds from 0.1 mV/s to 0.5 mV/s are recorded in Fig. 4g and Fig. S15 (Suporting information). The curves have a similar shape, however, exhibit slight peak shifts with increasing scan rates. The identified redox peaks align with those presented in Fig. 3a, both in position and intensity. Notably, Fig. 4h reveals that the b values obtained by all peak fittings of NMTP/C@Ti₃C₂ consistently exceed 0.5, which proves that there is an obvious pseudocapacitive effect. Analysis of the pseudocapacitive contribution rates at various scan speeds, presented in Fig. 4i, indicates that both materials demonstrate an increased contribution with higher scanning speeds. Compared with NMTP/C, the pseudocapacitance contribution rate of NMTP/C@Ti₃C₂ is significantly higher, and it is higher than 90% at different scan rates, which can effectively improve the electrochemical performance of materials.

  To investigate the structural evolution and sodium storage mechanisms of NMTP/C and NMTP/C@Ti₃C₂ during electrochemical reactions, the in-situ XRD characterization technique was employed for the charge/discharge process, which operated within a voltage range of 1.5–4.3 V (Fig. 5a and Fig. S16 in Supporting information). Both NMTP/C and NMTP/C@Ti₃C₂ cathodes preserve the distinct crystalline characteristics of the NASICON structure and exhibit a series of reversible peak shifts, demonstrating a fully reversible Na+ extraction and insertion process. Specifically, peaks such as (104), (113), (204), (211) and (300) gradually shift to higher angles during the charging process and vanish at approximately 3.8 V. The (104), (113), (204) and (300) crystal planes reappear at ~4.0 V and disappear again at ~4.15 V, and then reemerge when the voltage exceeds 4.2 V. In particular, the (211) crystal plane vanishes at 3.8 V and does not recover. The (116) plane consistently shifts to a higher angle with the charging process. These changes in peak positions are reversible throughout the charging and discharging process, suggesting that the Mn²⁺/Mn³⁺/Mn^/4+ reactions are two-phase. Conversely, no new diffraction peaks emerge between 1.5–2.5 V, pointing out that the transformation of Na₄MnTi(PO₄)₃ to Na₃MnTi(PO₄)₃ is a solid-solution reaction. Overall, NMTP/C exhibits a charge-discharge mechanism similar to that of NMTP/C@Ti₃C₂. However, upon charging, the (113) diffraction peak of NMTP/C gradually weakens and broadens from its initially sharp and distinct form, becoming nearly indistinguishable. In contrast, a more pronounced evolution of the peak is observed in NMTP/C@Ti₃C₂. Lattice defects and crystal structure attenuation result in the broadening of diffraction peaks, demonstrating that NMTP/C@Ti₃C₂ exhibits superior crystal structure stability. In addition, the calculated total unit cell change of NMTP/C@Ti₃C₂ is 12.3%, which is smaller than that of NMTP/C (13.7%) with similar structural evolutions (Fig. 5b). The difference indicates that Ti₃C₂ helps to alleviate the stress and strain on the cathode during sodium ion insertion/deinsertion process, thereby maintaining the integrity of the crystal structure.

  Figure 5

  

  Figure 5.  (a) In-situ XRD patterns of NMTP/C@Ti₃C₂ cathode during the charge/discharge process. (b) The volume variation of the unit cell of NMTP/C and NMTP/C@Ti₃C₂ cathodes during the charge/discharge process. (c) Mn 2p and Ti 2p XPS spectrum of NMTP/C@Ti₃C₂ under different voltages. (d) C 1s, O 1s and F 1s XPS spectrum of NMTP/C@Ti₃C₂ cathode before and after cycling. (e) Density of states of Ti₃C₂-MXene. (f) Schematic diagram of sodium ion diffusion between Ti₃C₂-MXene layers. (g) Energy barrier of sodium ion diffusion between Ti₃C₂-MXene layers.

  DownLoad: Full-Size Img PowerPoint

  As shown in Fig. 5c, the changes in the chemical states of TM ions in NMTP/C@Ti₃C₂ during the discharge process were collected by ex-situ XPS. Between 4.3 V and 3.9 V, the valence state of Mn shifts from Mn⁴⁺ (643.4 eV) to Mn³⁺ (642.9 eV), and subsequently to Mn²⁺ (641.4 eV) after further discharge to 2.5 V. The valence state of Ti remains 4+ (~460 eV) between 4.3 V and 2.5 V, and converted to 3+ (~458.9 eV) when the voltage drops to 1.5 V. These results confirm the reversible three-electron redox reactions of NMTP/C@Ti₃C₂. Additionally, a stable and uniform CEI significantly influences the electrochemical performance of the cathode [52-55]. To analyze the composition of the CEI, XPS was performed on the NMTP/C@Ti₃C₂ cathode after cycling (Fig. 5d). In the C 1s spectrum, C–C/C–H (~284.8 eV), C–O (~286.5 eV), C═O (~289.0 eV) and C–F (291 eV) are detected on the cathode surface, and the content of C–O and C═O increased significantly after long-time cycling. The O 1s spectrum further proves the existence of C–O and C═O in CEI. In addition, it can be found in F 1s spectrum that the Na-F content in CEI increases greatly after cycling. Therefore, it can be inferred that the composition of CEI is mainly composed of organic components (RO–C–, RO═C–, etc.) and inorganic components (NaF), which originate from the decomposition of the electrolyte, and the thickness of CEI increases with the number of cycles.

  To further understand the reasons for the enhanced sodium storage performance, the electronic structure and the sodium ion diffusion capability of Ti₃C₂-MXene were investigated using density functional theory (DFT). As shown in Fig. 5e, the density of states (DOS) of MXene were calculated, which exhibits a metallic feature with continuous electronic states near the Fermi level, thereby demonstrating high electronic conductivity. Subsequently, climbing image nudged elastic band (CI-NEB) method was employed to investigate the diffusion barrier of sodium-ion between the interlayers of Ti₃C₂ [56,57]. As shown in Fig. 5f, sodium ions diffuse above along the Ti-C-Ti atoms. As shown in Fig. 5g, the energy barrier for Na⁺ between the interlayers of MXene is only about 0.305 eV. Therefore, the lower migration energy barrier of MXene facilitates sodium ion diffusion more effectively. Benefit by the high conductivity and low energy barrier of Na⁺ on MXene, charge transfer and sodium ion transport during electrochemical reactions are significantly enhanced, thereby ensuring superior kinetics and rate performance [58,59].

  To further demonstrate the competitiveness of NMTP/C@Ti₃C₂, it is assembled with a hard carbon (HC) anode as a full-coin cell (Fig. 6a) [60]. The GCD curves of the HC anode are provided in Fig. S17 (Supporting information). Fig. 6b exhibits the rate performance of the full-coin cells. Consistent with the test results of the half-cell, the capacity of NMTP/C is slightly higher than that of NMTP/C@Ti₃C₂ at a small current density. As the current density increases, the discharge capacity of NMTP/C decreases more obviously, and NMTP/C@Ti₃C₂ maintains a better rate performance. NMTP/C@Ti₃C₂//HC full-coin can obtain the discharge capacities of 158.79, 151.7, 158.79, 147.44, 143.11, and 126.39 mAh/g from 0.2 C to 5 C, respectively. NMTP/C@Ti₃C₂//HC full-coin cell also exhibits an energy density of 450 Wh/kg at 0.2 C. In addition, it can be seen from the charge-discharge curves of different rates in Fig. 6c and Fig. S18 (Supporting information) that NMTP/C@Ti₃C₂ also has a better ability to resist voltage hysteresis in full-coin cells. The cycling stability of the full cell under 1 C is shown in Fig. 6d. The NMTP/C@Ti₃C₂//HC full-coin cell delivers a high capacity of 144.7 mAh/g at 1 C and maintains 129.5 mAh/g after 150 cycles with a capacity retention of 89.5%. In contrast, the capacity of NMTP/C//HC full-coin cell after 150 cycles is only 113.5 mAh/g, corresponding to a retention rate of 78.9%, which is 10.6% lower than the former. Compared to the previously reported NASICON-type SIB full cell [21,22,40,41,43,44,46-49], NMTP/C@Ti₃C₂ demonstrated better performance in terms of discharge capacity and energy density, proving its potential for practical applications (Fig. 6e and Table S6 in Supporting information). The half-cell and full-cell results demonstrate that the NMTP/C@Ti₃C₂ is a promising cathode material for SIBs. The anode material and other battery components need to be further optimized to better match this material for commercialization.

  Figure 6

  

  Figure 6.  (a) Schematic diagram of the full cell. (b) Rate performance of NMTP/C//HC and NMTP/C@Ti₃C₂//HC. (c) The GCD curves of NMTP/C@Ti₃C₂//HC at different current density. (d) Cycling performance of NMTP/C//HC and NMTP/C@Ti₃C₂//HC at 1 C. (e) Comparison of the NMTP/C@Ti₃C₂//HC full cell with other reported NASICON-type SIBs.

  DownLoad: Full-Size Img PowerPoint

  In summary, the amorphous carbon and MXene artificial mixed-conducting interphase modified NMTP/C@Ti₃C₂ material is rationally designed through an ultrasonic-assisted sol-gel method and applied in high-performance SIBs. The incorporation of Ti₃C₂-MXene not only enhances electronic conductivity but also facilitates the ion diffusion at the interfaces. The rationally designed microstructure enables the formation of an interconnected ion/electron transport channel at the cathode surface, leading to a fast ion/electron kinetic process. Benefiting from the fast ion diffusion coefficient (10⁻⁹–10⁻¹² cm²/s), the pseudocapacitive effect (> 90%), and the ability to resist voltage hysteresis, the optimized NMTP/C@Ti₃C₂ exhibits a fully reversible three electronic redox reaction and prominent electrochemical performance. A reversible capacity of 158.2 mAh/g is obtained at 0.2 C, corresponding to the energy density of 466.6 Wh/kg. Furthermore, NMTP/C@Ti₃C₂ also exhibits good cycling stability, the capacity remained at 135.1 mAh/g after 250 cycles at 1 C, with a capacity retention of 93.6%. Similarly, the capacity retention is 88.5% (120.9 mAh/g) after 500 cycles at 5 C. Further, the NMTP/C@Ti₃C₂//HC full-coin cell also showed a gratifying energy density of 450 Wh/kg at 0.2 C, proving its excellent competitiveness in SIB cathode materials. This work offers valuable insights into the design of NASICON-type cathodes to meet the high energy density requirements of the commercial market.

  Declaration of competing interest

  The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

  CRediT authorship contribution statement

  Nan Zhang: Writing – original draft, Methodology, Investigation, Conceptualization. Qian Yan: Methodology, Investigation. Xiaorui Dong: Methodology, Investigation. Jingyang Wang: Writing – review & editing, Software. Fan Jin: Methodology, Investigation. Jiaxuan Liu: Investigation. Dianlong Wang: Supervision, Resources. Huakun Liu: Writing – review & editing. Bo Wang: Writing – review & editing, Supervision, Resources, Funding acquisition. Shixue Dou: Writing – review & editing.

  Acknowledgments

  This work was supported by the National Natural Science Foundation of China (Nos. 51604089, 51874110 and 22479035), Natural Science Foundation of Heilongjiang Province (No. YQ2021B004). The computational analysis was performed using computational resources from e-Science Center of Collaborative Innovation Center of Advanced Microstructure.

  Supplementary materials

  Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2024.110328.

  
  
• 1. [1]

     S.W. Kim, D.H. Seo, X. Ma, et al., Adv. Energy Mater. 2 (2012) 710–721. doi: 10.1002/aenm.201200026

  2. [2]

     H. Pan, Y.S. Hu, L. Chen, Energy Environ. Sci. 6 (2013) 2338–2360. doi: 10.1039/c3ee40847g

  3. [3]

     Z. Lun, B. Ouyang, D.H. Kwon, et al., Nat. Mater. 20 (2021) 214–221. doi: 10.1038/s41563-020-00816-0

  4. [4]

     N. Yabuuchi, K. Kubota, M. Dahbi, et al., Chem. Rev. 114 (2014) 11636–11682. doi: 10.1021/cr500192f

  5. [5]

     M. Chen, W. Hua, J. Xiao, et al., Angew. Chem. Int. Ed. 59 (2020) 2449–2456. doi: 10.1002/anie.201912964

  6. [6]

     M. Chen, W. Hua, J. Xiao, et al., Nat. Commun. 10 (2019) 1480. doi: 10.1038/s41467-019-09170-5

  7. [7]

     Y. Tian, G. Zeng, A. Rutt, et al., Chem. Rev. 121 (2021) 1623–1669. doi: 10.1021/acs.chemrev.0c00767

  8. [8]

     J. Wang, T. He, X. Yang, et al., Nat. Commun. 14 (2023) 5210. doi: 10.1038/s41467-023-40669-0

  9. [9]

     Z.L. Jian, W.Z. Han, X. Lu, et al., Adv. Energy Mater. 3 (2013) 156–160. doi: 10.1002/aenm.201200558

  10. [10]

      J. Wang, Y. Wang, D.H. Seo, et al., Adv. Energy Mater. 10 (2020) 1903968. doi: 10.1002/aenm.201903968

  11. [11]

      C. Xu, J. Zhao, E. Wang, et al., Adv. Energy Mater. 11 (2021) 2100729. doi: 10.1002/aenm.202100729

  12. [12]

      Q. Ni, Y. Bai, F. Wu, et al., Adv. Sci. 4 (2017) 1600275. doi: 10.1002/advs.201600275

  13. [13]

      H. Li, W. Zhang, K. Sun, et al., Adv. Energy Mater. 11 (2021) 2100867. doi: 10.1002/aenm.202100867

  14. [14]

      Y. Liu, C. Sun, Y. Li, et al., Energy Stor. Mater. 57 (2023) 69–80. doi: 10.1117/12.2655822

  15. [15]

      W. Zhou, L. Xue, X. Lu, et al., Nano Lett. 16 (2016) 7836–7841. doi: 10.1021/acs.nanolett.6b04044

  16. [16]

      H. Gao, J.B. Goodenough, Angew. Chem. Int. Ed. 55 (2016) 12768–12772. doi: 10.1002/anie.201606508

  17. [17]

      H. Gao, I.D. Seymour, S. Xin, et al., J. Am. Chem. Soc. 140 (2018) 18192–18199. doi: 10.1021/jacs.8b11388

  18. [18]

      H. Gao, Y. Li, K. Park, et al., Chem. Mater. 28 (2016) 6553–6559. doi: 10.1021/acs.chemmater.6b02096

  19. [19]

      T. Zhu, P. Hu, X.P. Wang, et al., Adv. Energy Mater. 9 (2019) 1803436. doi: 10.1002/aenm.201803436

  20. [20]

      P. Hu, T. Zhu, C. Cai, et al., Angew. Chem. Int. Ed. 62 (2023) e202219304. doi: 10.1002/anie.202219304

  21. [21]

      H. Li, M. Xu, C. Gao, et al., Energy Stor. Mater. 26 (2020) 325–333.

  22. [22]

      C. Xu, W. Hua, Q. Zhang, et al., Adv. Funct. Mater. 33 (2023) 2302810. doi: 10.1002/adfm.202302810

  23. [23]

      T. Zhu, P. Hu, C. Cai, et al., Nano Energy 70 (2020) 104548. doi: 10.1016/j.nanoen.2020.104548

  24. [24]

      J. Zhang, C. Lin, Q. Xia, et al., ACS Energy Lett. 6 (2021) 2081–2089. doi: 10.1021/acsenergylett.1c00426

  25. [25]

      Y. Liu, X. Rong, R. Bai, et al., Nat. Energy 8 (2023) 1088–1096. doi: 10.1038/s41560-023-01301-z

  26. [26]

      F. Ji, K. Chen, F. Li, et al., Energy Storage Mater. 68 (2024) 103371. doi: 10.1016/j.ensm.2024.103371

  27. [27]

      X. Rui, W. Sun, C. Wu, et al., Adv. Mater. 27 (2015) 6670–6676. doi: 10.1002/adma.201502864

  28. [28]

      T. Long, P. Chen, B. Feng, et al., Chin. Chem. Lett. 35 (2024) 109267. doi: 10.1016/j.cclet.2023.109267

  29. [29]

      Y. Katsuyama, A. Kudo, H. Kobayashi, et al., Small 18 (2022) 2202277. doi: 10.1002/smll.202202277

  30. [30]

      B. Kang, G. Ceder, Nature 458 (2009) 190–193. doi: 10.1038/nature07853

  31. [31]

      H. Zhang, L. Shen, X. Geng, et al., Chem. Eng. J. 466 (2023) 143132. doi: 10.1016/j.cej.2023.143132

  32. [32]

      Q. Yan, R. Ding, H. Zheng, et al., Adv. Funct. Mater. 33 (2023) 2301982. doi: 10.1002/adfm.202301982

  33. [33]

      M. Han, C.E. Shuck, R. Rakhmanov, et al., ACS Nano 14 (2020) 5008–5016. doi: 10.1021/acsnano.0c01312

  34. [34]

      B. Anasori, M.R. Lukatskaya, Y. Gogotsi, Nat. Rev. Mater. 2 (2017) 16098. doi: 10.1038/natrevmats.2016.98

  35. [35]

      X. Li, Z. Huang, C.E. Shuck, et al., Nat. Rev. Chem. 6 (2022) 389–404. doi: 10.1038/s41570-022-00384-8

  36. [36]

      Y. Gogotsi, R.M. Penner, ACS Nano 12 (2018) 2081–2083. doi: 10.1021/acsnano.8b01914

  37. [37]

      M.R. Lukatskaya, S. Kota, Z. Lin, et al., Nat. Energy 2 (2017) 17105. doi: 10.1038/nenergy.2017.105

  38. [38]

      Y. Xu, S. Wang, J. Yang, et al., Nano Energy 51 (2018) 442–450. doi: 10.1016/j.nanoen.2018.06.086

  39. [39]

      Y. Xia, L. Que, F. Yu, et al., Nanomicro. Lett. 14 (2022) 143.

  40. [40]

      Q.M. Yin, Z.Y. Gu, Y. Liu, et al., Adv. Funct. Mater. 33 (2023) 2304046. doi: 10.1002/adfm.202304046

  41. [41]

      N. Jiang, C. Yang, Y. Wang, et al., Energy Stor. Mater. 63 (2023) 102978.

  42. [42]

      Q. Wei, X. Chang, J. Wang, et al., Adv. Mater. 34 (2021) e2108304.

  43. [43]

      C. Xu, J. Zhao, Y.-A. Wang, et al., Adv. Energy Mater. 12 (2022) 2200966. doi: 10.1002/aenm.202200966

  44. [44]

      X.X. Zhao, W. Fu, H.X. Zhang, et al., Adv. Sci. 10 (2023) e2301308. doi: 10.1002/advs.202301308

  45. [45]

      Z.Y. Gu, J.Z. Guo, J.M. Cao, et al., Adv. Mater. 34 (2022) e2110108. doi: 10.1002/adma.202110108

  46. [46]

      C. Xu, R. Xiao, J. Zhao, et al., ACS Energy Lett. 7 (2021) 97–107.

  47. [47]

      J. Zhang, Y. Liu, X. Zhao, et al., Adv. Mater. 32 (2020) e1906348. doi: 10.1002/adma.201906348

  48. [48]

      X.J. Pu, H.M. Wang, T.C. Yuan, et al., Energy Storage Mater. 22 (2019) 330–336. doi: 10.1016/j.ensm.2019.02.017

  49. [49]

      X. Ge, H. Li, J. Li, et al., Small 19 (2023) e2302609. doi: 10.1002/smll.202302609

  50. [50]

      V. Augustyn, P. Simon, B. Dunn, Energy Environ. Sci. 7 (2014) 1597. doi: 10.1039/c3ee44164d

  51. [51]

      V. Augustyn, J. Come, M.A. Lowe, et al., Nat. Mater. 12 (2013) 518–522. doi: 10.1038/nmat3601

  52. [52]

      Q. Wang, Z. Yao, C. Zhao, et al., Nat. Commun. 11 (2020) 4188. doi: 10.1038/s41467-020-17976-x

  53. [53]

      S. Jiao, X. Ren, R. Cao, et al., Nat. Energy 3 (2018) 739–746. doi: 10.1038/s41560-018-0199-8

  54. [54]

      B. Wen, Z. Deng, P.C. Tsai, et al., Nat. Energy 5 (2020) 578–586. doi: 10.1038/s41560-020-0647-0

  55. [55]

      N. Zhang, B. Wang, F. Jin, et al., Cell Rep. Phys. Sci. 3 (2022) 101197. doi: 10.1016/j.xcrp.2022.101197

  56. [56]

      G. Henkelman, B.P. Uberuaga, H. Jónsson, J. Chem. Phys. 113 (2000) 9901–9904. doi: 10.1063/1.1329672

  57. [57]

      T. Ruan, B. Wang, Y. Yang, et al., Advanced Materials 32 (2020) 2000151. doi: 10.1002/adma.202000151

  58. [58]

      B. Wang, W. Al Abdulla, D. Wang, et al., Energy Environ. Sci. 8 (2015) 869–875. doi: 10.1039/C4EE03825H

  59. [59]

      B. Wang, T. Ruan, Y. Chen, et al., Energy Storage Mater. 24 (2020) 22–51. doi: 10.1016/j.ensm.2019.08.004

  60. [60]

      H. Ren, S. Li, B. Wang, et al., Adv. Mater. 35 (2023) 2208237. doi: 10.1002/adma.202208237

• 

  Figure 1  (a) Schematic diagram and (b) the Rietveld refinements of the XRD patterns of NMTP/C@Ti₃C₂ material. (c) TGA curves and (d) Raman spectra of NMTP and NMTP/C@Ti₃C₂. (e) The BET test results of NMTP/C@Ti₃C₂. (f) Electric conductivity of the NMTP and NMTP/C@Ti₃C₂. High-resolution (g) Mn 2p and (h) Ti 2p XPS spectra of NMTP/C and NMTP/C@Ti₃C₂.

  下载: 全尺寸图片 幻灯片
  

  Figure 2  (a) SEM, (b) TEM, (c) HRTEM images and FFT pattern of NMTP/C material. (d) SEM, (e) TEM, (f, g) HRTEM images and FFT pattern. (h) HAADF-STEM image and the corresponding elemental mappings of NMTP/C@Ti₃C₂ material.

  下载: 全尺寸图片 幻灯片
  

  Figure 3  (a) CV curves of NMTP/C@Ti₃C₂ at 0.1 mV/s. (b) GCD curves of NMTP/C and NMTP/C@Ti₃C₂ at 0.2 C. (c) Comparison of the NMTP/C@Ti₃C₂ with other reported NASICON-type SIBs. (d) Rate performance of NMTP/C and NMTP/C@Ti₃C₂. (e) GCD curves of NMTP/C@Ti₃C₂ at different current densities. (f) The average discharge voltage of NMTP/C and NMTP/C@Ti₃C₂ at different rates. Cycling performance of NMTP/C and NMTP/C@Ti₃C₂ at (g) 1 C and (h) 5 C.

  下载: 全尺寸图片 幻灯片
  

  Figure 4  Nyquist plots at different temperatures and Arrhenius plots of (a) NMTP/C and (b) NMTP/C@Ti₃C₂. (c) The Nyquist plots and (d) the impedance change of NMTP/C@Ti₃C₂ at different cycling numbers. (e) GITT curves and Na⁺ diffusion coefficient and (f) overpotential of NMTP/C and NMTP/C@Ti₃C₂ cathodes. (g) CV curves of NMTP/C@Ti₃C₂ obtained at different scan rates of 0.1, 0.2, 0.3, 0.4, and 0.5 mV/s. (h) The log(i) versus log(v) plots of NMTP/C@Ti₃C₂ at different redox states. (i) Capacitance contribution of NMTP/C and NMTP/C@Ti₃C₂.

  下载: 全尺寸图片 幻灯片
  

  Figure 5  (a) In-situ XRD patterns of NMTP/C@Ti₃C₂ cathode during the charge/discharge process. (b) The volume variation of the unit cell of NMTP/C and NMTP/C@Ti₃C₂ cathodes during the charge/discharge process. (c) Mn 2p and Ti 2p XPS spectrum of NMTP/C@Ti₃C₂ under different voltages. (d) C 1s, O 1s and F 1s XPS spectrum of NMTP/C@Ti₃C₂ cathode before and after cycling. (e) Density of states of Ti₃C₂-MXene. (f) Schematic diagram of sodium ion diffusion between Ti₃C₂-MXene layers. (g) Energy barrier of sodium ion diffusion between Ti₃C₂-MXene layers.

  下载: 全尺寸图片 幻灯片
  

  Figure 6  (a) Schematic diagram of the full cell. (b) Rate performance of NMTP/C//HC and NMTP/C@Ti₃C₂//HC. (c) The GCD curves of NMTP/C@Ti₃C₂//HC at different current density. (d) Cycling performance of NMTP/C//HC and NMTP/C@Ti₃C₂//HC at 1 C. (e) Comparison of the NMTP/C@Ti₃C₂//HC full cell with other reported NASICON-type SIBs.

  下载: 全尺寸图片 幻灯片
•