Rational design of nitrogen-doping Ti₃C₂T_xmicrospheres with enhanced polysulfide catalytic activity for lithium-sulfur batteries

INTRODUCTION

With the rapid progress in energy storage technology, the demand for higher energy density is increasing. Lithium-sulfur (Li-S) batteries have shown tremendous potential because of their high theoretical energy density (2,600 Wh kg^-1) and theoretical specific capacity (1,675 mAh g^-1)^[1-3]. However, their practical applications still face several challenges^[4-6], including the poor electronic conductivity of sulfur and the discharge product lithium sulfide (Li₂S), volume expansion during discharge, sluggish redox reaction kinetics, and the notorious shuttle-effects (polysulfide intermediates (LiPSs) dissolved in electrolyte and reacted with lithium anodes). These issues lead to low discharge capacity and rapid deterioration in performance^[7,8].

Advanced cathode host has been developed to solve the aforementioned issues, including carbon and its derivate^[9-11], metal (e.g., high-entropy alloy)^[12,13], some conducting compounds or polymers^[14,15], and so on. Among them, MXenes have received considerable interest. As first proposed by Naguib et al. in 2011, they are expressed as M_n+1X_nT_x (n = 1-3), where M represents transition metal; X stands for C or/and N elements; T_xis the symbol of functional groups on the surface of MXenes^[16]. It is impressive that MXenes have the following advantages for Li-S batteries compared to traditional carbon materials: (1) inherent metal conductivity^[17,18]; (2) two-dimensional surface with high activity^[19,20]; (3) large binding energy with LiPSs based on Lewis’s interaction to inhibit shuttle effect^[21,22]; and (4) effective catalytic effect on the conversion of LiPSs achieved by abundant surface terminations^[23,24]. It is these advantages that make MXenes also have excellent performance in other metal-sulfur batteries^[25-27].

Zhao et al. selectively extracted Ti from the Ti₂SC MAX phase to form carbon/sulfur nanolaminates, suggesting that MXene material is a promising positive carrier for Li-S batteries^[28]. Liang et al. reported the Ti₂C(OH)_x/S cathode application, showing that lithium polysulfides have a strong interaction with the “acidic” Ti site and hydroxyl surface groups^[29]. The electrochemical properties of MXenes can be further improved by stripping multi-layer MXenes into nanosheets with fewer layers^[30,31]. However, severe agglomeration of MXene nanosheets happened through van der Waals forces and hydrogen bond interactions after stripping, limiting the accessibility of ions and preventing full utilization of functional surface terminations^[32]. Doping of heterogeneous atoms, such as nitrogen atoms, can enlarge the interlayer spacing and prevent the agglomeration of MXene nanosheets^[33,34]. For example, in 2017, Wen et al. obtained nitrogen-doped MXenes by heat treatment under an ammonia atmosphere^[35], effectively expanding the c-lattice parameter of MXene sheets. In addition, introducing nitrogen doping into electrodes can effectively improve the adsorption capacity of LiPSs^[36-38] and confer electrocatalytic activity on MXenes to accelerate the conversion of LiPSs^[38]. Song et al. prepared porous nitrogen-doped Ti₃C₂ MXenes (P-NTC) using melamine-formaldehyde (MF) nanospheres as hard templates^[39]. The negatively charged MXene nanosheets on the positively charged MF sphere surface through electrostatic self-assembly and the prepared P-NTC show an interconnect network with different pores, which greatly alleviate the agglomeration of the MXene nanosheets. Due to the synergistic effect of large specific surface area and nitrogen doping, P-NTC showed more effective adsorption and catalytic conversion ability for Li₂S₄ solution than MXenes (TC) without nitrogen doping.

Herein, three-dimensional nitrogen-doped porous MXene microspheres (3D N-Ti₃C₂T_x) were synthesized by spray drying and subsequent annealing as the host for sulfur. Firstly, a few layers of negatively charged MXene solution were mixed evenly with positively charged hydrochloric acid (HCl)-treated melamine, and the powder was collected by spray drying. Through sequential annealing, 3D N-Ti₃C₂T_x was obtained. The MXene nanosheets formed three-dimensional spherical nanoflower structure, effectively preventing the agglomeration of MXene nanosheets. During the heat treatment, the self-supported melamine decomposes, effectively increasing the surface area. Therefore, the synthesized 3D N-Ti₃C₂T_xshows a large surface area, which enables the sulfur to be uniformly impregnated into the MXenes layer. Meanwhile, melamine acts as a nitrogen source to introduce nitrogen doping. Remarkably, due to the synergistic effect of large surface area and nitrogen doping, the 3D N-Ti₃C₂T_x/S cathode exhibits excellent electrochemical properties.

EXPERIMENTAL

RESULTS AND DISCUSSION

Figure 1 depicts the synthesizing process of 3D N-Ti₃C₂T_x/S. The Al layer in Ti₃AlC₂ (MAX phase) was etched by HF to acquire accordion-like multi-layer Ti₃C₂T_xMXene nanosheets. Subsequently, few-layered MXenes can be synthesized by solution-phase flocculation strategy, according to our previous reports^[40]. The amino group (-NH₂) in melamine could capture protons (H⁺) from HCl solution, thus forming positively charged melamine through protonation, which can be easily dissolved in DI water^[41]. Obviously, HCl-melamine can be fully dissolved in DI water [Supplementary Figure 1].

Figure 1. Schematic illustration of the synthesis process of 3D N-Ti₃C₂T_x and 3D N-Ti₃C₂T_x/S.

The aforementioned HCl-treated melamine was mixed with the few-layered Ti₃C₂T_xsolution and stirred together. MXene nanosheets possess negative charges due to the presence of surface functional terminations (e.g., -O, -OH, and -F), allowing for the attachment of positively charged melamine onto the surface of Ti₃C₂T_x MXene nanosheets through electrostatic interactions^[37,42]. The resultant powder was then collected via spray drying in a constant stirring, forming porous microspheres. The precursor underwent calcination at 700 °C in an inert atmosphere for 2 h, yielding flower-like porous nitrogen-doped MXene microspheres. Consequently, the synthesized 3D N-Ti₃C₂T_x demonstrates remarkable thermal stability, which aligns with previous reports that the tailoring or functionalization of Ti₃C₂ MXenes through polymers proficiently improves its stability^[43,44]. As a comparative experiment, 3D Ti₃C₂T_x microspheres were obtained by directly spray drying a pure Ti₃C₂T_x solution, followed by calcination under identical conditions.

FESEM effectively captured the intricate morphology and microstructure. The precursor of Ti₃AlC₂ (MAX phase) presents an inherent structure characterized by a dense stratification, as illustrated in Supplementary Figure 2A. After HF etching, accordion-like MXenes can be obtained [Supplementary Figure 2B]. Then, few-layered MXenes can be synthesized by solution-phase flocculation strategy [Supplementary Figure 2C]^[40]. In Supplementary Figure 2D and E, 3D Ti₃C₂T_x maintains the lamellar morphology of MXenes, with its surface exhibiting nanoparticles, presumably attributable to the formation of TiO₂^[45]. Figure 2A illustrates the morphology of the HCl-treated melamine/Ti₃C₂T_x precursor. After spray drying, porous microspheres ranging from 2 to 10 μm formed. The 3D N-Ti₃C₂T_x exhibits a nano-flower-shaped morphology composed of MXene nanosheets after annealing at 700 °C [Figure 2B]. As can be seen in Supplementary Figure 2F, the surface of these Ti₃C₂T_x nanosheets is flat and smooth, with no observable TiO₂ nanoparticles. This intriguing observation suggests that the positively charged HCl-treated melamine coated itself on the surface of Ti₃C₂T_x nanosheet layers through electrostatic self-assembly interaction, serving as a physical barrier to oxygen, effectively shielding the MXenes from oxidation during high-temperature processes. Consequently, the synthesized 3D N-Ti₃C₂T_x demonstrates exceptional thermal stability.

Figure 2. SEM images of (A) HCl-treated melamine/ Ti₃C₂T_x precursor and (B) 3D N-Ti₃C₂T_x, (C-F) TEM, and (G) STEM images of 3D N-Ti₃C₂T_x with corresponding elemental mappings of C, Ti, and N. Scale ba rs: 200 nm.

The morphology and microstructure were further investigated via TEM. As shown in Figure 2C-F, TEM images provide a clearer view of 3D N-Ti₃C₂T_x. Microspheres composed of MXene nanosheets can be observed. The lattice spacing corresponding to the MXenes (002) plane is approximately 1.5 nm [Figure 2E]. It is apparent that MXene nanosheets in 3D N-Ti₃C₂T_x feature nano-sized micropores and mesopores at higher magnification, which could be attributed to the decomposition of melamine. These pores introduce defects, exposing more active sites to constrain LiPSs. Furthermore, C, Ti, and N elements are uniformly distributed within 3D N-Ti₃C₂T_x, indicating the successful nitrogen doping into Ti₃C₂T_x introduced by melamine decomposition [Figure 2G].

However, as shown in Supplementary Figure 3, the TiO₂ nanoparticle can be observed in TEM images, which demonstrates that Ti₃C₂T_x undergoes substantial oxidation during thermal processing. As for 3D N-Ti₃C₂T_x, melamine serves as a physical barrier to oxygen, effectively shielding the MXenes from oxidation.

After sulfur infusion, it is evident from the TEM and STEM images [Supplementary Figure 4] that sulfur has permeated uniformly into the layers of MXenes. Furthermore, the 3D N-Ti₃C₂T_x has maintained its original nano-flower spherical architecture, exhibiting remarkable structural stability.

In order to demonstrate the electrostatic self-assembly interaction between HCl-melamine and few-layers Ti₃C₂T_x, we tested the zeta potential of both [Figure 3A], and the results showed that they were positively charged and negatively charged, respectively, which was the origin of electrostatic self-assembly interaction.

Figure 3. (A) The zeta potential of HCl-treated melamine and 3D N-Ti₃C₂T_x. XRD patterns of Ti₃AlC₂, multi-layered Ti₃C₂T_x, few-layered Ti₃C₂T_x sheets and 3D N-Ti₃C₂T_x with (B) 2θ from 2° to 15°, (C) 2θ from 2° to 70°. (D) N₂ adsorption-desorption isotherms and (E) pore-size distribution curves of 3D N-Ti₃C₂T_x and 3D Ti₃C₂T_x. XPS pattern of 3D N-Ti₃C₂T_x (F) C 1s, (G) Ti 2p, (H) O 1s and (I) N 1s.

XRD techniques investigated the structure of the synthesized 3D N-Ti₃C₂T_x. As shown in Figure 3B, compared to multilayered Ti₃C₂T_xMXenes, the low-angle diffraction peak for few-layered MXenes shifted from 8.74° to 6.22°, indicative of an increased interlayer spacing of the Ti₃C₂T_xMXenes (002) crystal plane. Furthermore, the XRD spectrum of 3D N-Ti₃C₂T_x exhibited a peak at 5.98° for the (002) crystal plane, a 0.24° low-angle shift relative to few-layered Ti₃C₂ MXenes (6.22°), which is attributed to nitrogen doping, as nitrogen atoms have a larger radius than carbon atoms, expanding interlayer spacing in MXenes^[35]. Figure 3C illustrates the disappearance of the characteristic peak of the (104) crystal plane of Ti₃AlC₂ in the XRD spectrum of multilayered Ti₃C₂ MXenes, indicating the successful removal of aluminum atomic layers from Ti₃AlC₂. Without nitrogen doping, 3D Ti₃C₂T_x exhibited diffraction peaks at 25.3°, corresponding to the (101) crystal plane of anatase TiO₂ (JCPDS#21-1272), consistent with the SEM and TEM images. This phenomenon indicates that oxidation occurs during the annealing, and on the contrary, the oxidation degree of 3D N-Ti₃C₂T_x is low, indicating the protective effect of melamine on MXenes. The XRD pattern of 3D N-Ti₃C₂T_x/S [Supplementary Figure 5] demonstrates the successful infusion of orthorhombic S₈.

Nitrogen adsorption-desorption tests under isothermal conditions on 3D N-Ti₃C₂T_x, 3D Ti₃C₂T_x, 3D N-Ti₃C₂T_x/S, and the melamine/MXenes precursor were conducted. The obtained isothermal adsorption-desorption curves were analyzed to further elucidate the pore structure and surface characteristics of the samples [Figure 3D]. The N₂ adsorption-desorption curve for 3D N-Ti₃C₂T_x exhibits a typical Type IV isotherm^[46], indicating the presence of abundant mesopores in the material, as observed from the pore size distribution image [Figure 3E]. The Brunauer-Emmett-Teller (BET) specific surface areas revealed that 3D N-Ti₃C₂T_x possesses a substantial surface area of 247.4 m²g^-1 and a pore volume of 0.83 cm³g^-1, while the melamine/MXenes precursor exhibited a specific surface area of 22.2 m²g^-1 and a pore volume of 0.09 cm³g^-1 [Supplementary Figure 6A and B]. These results highlight an increased porosity in 3D N-Ti₃C₂T_x due to the melamine decomposition during annealing, which will facilitate the adsorption of LiPSs and the maintenance of discharge capacity^[47]. The porous structure not only facilitates uniform sulfur loading in 3D N-Ti₃C₂T_x but also enhances electrolyte penetration and shortens ion/electron transport pathways. The specific surface area and pore volume of 3D N-Ti₃C₂T_x/S significantly decrease compared to 3D N-Ti₃C₂T_x, only 89.7 m²g^-1 and 0.22 cm³g^-1, respectively [Supplementary Figure 6C and D]. This indicates that sulfur has embedded itself into the pores after sulfur infusion. For comparison, 3D Ti₃C₂T_x displayed a lower specific surface area of 128.5 m²g^-1 and a pore volume of 0.28 cm³g^-1.

The thermal gravimetric analysis (TGA) curves of the HCl-treated melamine/MXenes revealed a sharp weight drop at the temperature range of 250-350 °C [Supplementary Figure 7], which can be attributed to the melamine decomposition^[41]. During annealing, active functional groups on the Ti₃C₂T_x nanosheet surface, such as -F, -O, and -OH, are removed [Supplementary Figure 7], providing active sites suitable for nitrogen atom doping. The defects were introduced in the vicinity of doped nitrogen atoms^[34]. Nitrogen-doped Ti₃C₂T_x nanosheets exhibit significantly improved surface activities, resulting in a hydrophilic surface and enhanced Lewis acid absorption capability^[48].

Further analysis of the chemical composition and interactions was performed using XPS spectra. XPS spectra [Supplementary Figure 8A] of 3D Ti₃C₂T_x and 3D N-Ti₃C₂T_xboth reveal the presence of C, O, and Ti elements, with the appearance of O elemental peaks attributed to surface-terminated functional groups (-O, -OH) on Ti₃C₂T_x. Figure 3F and Supplementary Figure 8B compare the C 1s spectra of 3D Ti₃C₂T_x and 3D N-Ti₃C₂T_x. The peaks at 281.8, 286.3, and 284.8 eV correspond to the C-Ti, C-O, and C-C bonds, respectively^[32]. Furthermore, 3D N-Ti₃C₂T_x exhibits an additional peak at 285.7 eV corresponding to the C-N bond^[30], and a noticeable decrease in peak intensity at 281.8 eV, possibly attributed to the removal of partial C-Ti bonds during nitrogen doping. In Ti 2p spectra, all two samples exhibit peaks associated with Ti-C bonds at 455.4 and 461.3 eV^[49] and Ti-O bonds at 458.4 and 464.2 eV^[50] [Figure 3G and Supplementary Figure 8C]. The Ti 2p spectrum of 3D N-Ti₃C₂T_x shows two additional peaks at 456.6 and 462.9 eV, corresponding to Ti-N bonds^[49]. The O 1s high-resolution XPS spectra of 3D Ti₃C₂T_x and 3D N-Ti₃C₂T_x[Figure 3H and Supplementary Figure 8D] can be divided into characteristic peaks at 529.9 eV for Ti-O bonds and 531.5 eV for Ti-OH bonds^[51], corresponding to common surface-terminated functional groups (-O and -OH) on the MXene surface. The strong Ti-O bond signals indicate the formation of TiO₂ in 3D Ti₃C₂T_x.

Besides, nitrogen elemental peaks are observed in 3D N-Ti₃C₂T_x, indicating effective nitrogen doping. High-resolution N 1s XPS spectra of 3D N-Ti₃C₂T_x[Figure 3I] reveal the chemical bonding properties of nitrogen. Peaks at 398.5 and 400.3 eV correspond to pyridinic and pyrrolic nitrogen, while a peak at 396.3 eV corresponds to Ti-N, further confirming the successful incorporation of nitrogen into MXene nanosheets^[49]. Nitrogen, especially pyrrolic and pyridinic nitrogen in carbon lattice, enhances the affinity and chemical binding energy between the polar Ti atoms and sulfur and polar lithium polysulfides.

To validate the adsorption efficiency of 3D N-Ti₃C₂T_x microspheres for LiPSs, visual adsorption experiment was applied to compare the adsorption capability of 3D N-Ti₃C₂T_x and 3D Ti₃C₂T_x to LiPSs. Figure 4A shows the color fading of Li₂S₆ solution with 3D N-Ti₃C₂T_x and 3D Ti₃C₂T_x, respectively. The Li₂S₆ solution containing 3D N-Ti₃C₂T_xnearly faded completely, whereas the Li₂S₆ solution containing 3D Ti₃C₂T_x remains yellow. This conclusion was further confirmed by corresponding UV-Vis spectroscopy, which showed an intensity decrease of the Li₂S₆ solution peak within 3D N-Ti₃C₂T_x compared to the others. Therefore, it can be concluded that 3D N-Ti₃C₂T_x microspheres possess strong adsorption capabilities for LiPSs. This can be attributed to the enhanced polarity due to nitrogen doping and the higher specific surface area of 3D N-Ti₃C₂T_x, which provides more adsorption sites for LiPSs. XPS analyses were applied to understand the chemical interaction between LiPSs and 3D N-Ti₃C₂T_x. In the N 1s spectrum, the peaks shift to a higher binding energy after the adsorption of Li₂S₆, indicating the interaction between N and S atoms [Figure 4B]^[52]. Besides, the proportion of configuration nitrogen had changed after the adsorption of Li₂S₆, as seen in Supplementary Table 1. Hence, it can be inferred that the pyridinic N and Ti-N are vital in the adsorption of Li₂S₆, because their relative proportions showed a greater decline^[53]. In terms of the S 2p spectrum [Figure 4C], it can be divided into S-O, S-S and S-Ti bonds. Notably, the strong peak located at 169.8 eV can be attributed to S-O bonds, indicating the robust chemical binding ability of 3D N-Ti₃C₂T_x to LiPSs^[54].

Figure 4. (A) UV-vis absorption spectra of the supernatant of Li₂S₆ solution with 3D Ti₃C₂T_x and 3D N-Ti₃C₂T_x and the inset is a digital photograph. XPS pattern: (B) N 1s and (C) S 2p of 3D N-Ti₃C₂T_x after adsorption of Li₂S₆. The CV curves of symmetric cells with electrolytes containing Li₂S₆, (D) scan rate: 0.2 mV s^-1 and (E) scan rate: 10 to 50 mV s^-1 (3D N-Ti₃C₂T_x). (F) The potentiostatic discharging profiles of Li₂S precipitation (at 2.05 V) for 3D N-Ti₃C₂T_x. (G) The potentiostatic discharging profiles of Li₂S dissolution (at 2.35 V) for 3D N-Ti₃C₂T_xand 3D Ti₃C₂T_x. (H) Tafel plot. (I) The EIS plot of 3D N-Ti₃C₂T_x symmetrical battery containing Li₂S₆ electrolytes at different temperatures.

Symmetric cells with electrolytes containing Li₂S₆ were assembled to evaluate the kinetics of LiPS conversion. The CV curves at 0.2 mV s^-1 [Figure 4D] revealed that 3D N-Ti₃C₂T_x exhibited higher current response, more positive reduction peak potential, and more negative oxidation peak potential, indicating a larger surface area and nitrogen doping, providing additional polysulfide adsorption sites and faster kinetics for Li₂S₆ redox conversion. Both 3D N-Ti₃C₂T_x and 3D Ti₃C₂T_x electrodes displayed distinct oxidation-reduction peaks in CV curves at various scan rates (10~50 mV s^-1) [Figure 4Eand Supplementary Figure 9A]. With increasing scan rate, the positions of the oxidation-reduction peaks slightly shifted, with the 3D N-Ti₃C₂T_x electrode exhibiting narrower peak separations, suggesting stronger catalytic activity in polysulfide conversion.

A potentiostatic Li₂S precipitation experiment was conducted to evaluate the catalytic effect of 3D N-Ti₃C₂T_x and 3D-Ti₃C₂T_x on LiPS conversion. The results in Figure 4F and Supplementary Figure 9B show that 3D N-Ti₃C₂T_x displayed a higher peak current, earlier current response, and higher Li₂S precipitation capacity (120.20 mAh g^-1) than 3D-Ti₃C₂T_x. These findings indicate the high catalytic activity of 3D N-Ti₃C₂T_x for Li₂S precipitation. Besides, we further investigated the impact of 3D N-Ti₃C₂T_x on the dissolution capacity of Li₂S during the charging process. As depicted in Figure 4G, 3D N-Ti₃C₂T_x exhibits a quicker current response and contributes significantly to the dissolution of Li₂S, which can be attributed to the superior electrical conductivity.

The enhanced redox kinetics from 3D N-Ti₃C₂T_x were further validated by exchanging current density [Figure 4H]. According to the fitting results of the Tafel plot, it exhibited a high exchange current density of 3.16 × 10^-4 mA cm^-2, remarkably larger than that of 3D Ti₃C₂T_x (1 × 10^-4 mA cm^-2), indicating a significant enhancement in redox kinetics.

In order to assess the performance of 3D N-Ti₃C₂T_x at both extremely high and low temperatures, we assembled a symmetrical battery containing Li₂S₆ electrolytes and conducted EIS analysis at a temperature range from -40 to 70 °C. As shown in Figure 4I, the battery displays its highest internal resistance (R_s) and charge transfer resistance (R_ct) at the frigid temperature of -40 °C, primarily due to the slow Li⁺ transportation. As the temperature increases, the impedance progressively decreases, with the R_ct plummeting to a mere 0.13 Ω at 70 °C, signaling 3D N-Ti₃C₂T_x facilitating charge transfer to LiPSs at high temperatures. Furthermore, an Arrhenius plot was constructed based on the R_ct values of the cell at varying temperatures [Supplementary Figure 9C], unveiling a notably linear correlation. The activation energy of the R_ct between 3D N-Ti₃C₂T_x and LiPSs was 0.396 eV.

The CR2032 coin cells with 3D N-Ti₃C₂T_x/S and 3D Ti₃C₂T_x/S cathodes were assembled to evaluate their electrochemical performance. TGA was also conducted on 3D N-Ti₃C₂T_x/S, 3D Ti₃C₂T_x/S, and pure sulfur to measure the sulfur content in the composites. As shown in Figure 5A, it is evident that all three samples exhibit significant mass loss at the temperature range of 200-350 °C, corresponding to the sublimation of sulfur under a nitrogen atmosphere. Accordingly, the sulfur content in 3D N-Ti₃C₂T_x/S and 3D Ti₃C₂T_x/S composites is measured as 72 and 66.4 wt%, respectively. CV tests were conducted in the voltage range of 1.8-2.8 V at a scan rate of 0.1 mV s^-1 [Figure 5B]. Two cathodic peaks (2.28 and 2.05 V) correspond to the reduction of S₈ to high-order LiPSs (Li₂S_x, 4 ≤ x ≤ 8) and then to Li₂S₂/Li₂S, which has been described previously^[55]. The positive cathodic peaks of the 3D N-Ti₃C₂T_x/S electrode indicate the effective catalytic activity to the LiPS conversion. Besides, as evidenced by the larger peak currents, the 3D N-Ti₃C₂T_x/S electrode exhibits superior redox kinetics to LiPSs. To further assess the impact on the LiPS conversion redox kinetics and lithium-ion diffusion, CV measurements at progressively increasing scan rates ranging from 0.1 to 0.5 mV s^-1 were conducted [Figure 5C and Supplementary Figure 10A]. As the scan rate increased, the cathodic peaks shifted towards negative potentials, indicating a growing polarization. A similar phenomenon can be observed at anodic peaks. The diffusion coefficient of lithium ions (D_Li⁺) can be determined by calculating the slope of the current peaks against the square root of the scan rate, as given in the Randles-Sevcik equation^[56]: $$ I_{p}=\left(2.69 \times 10^{5}\right) n^{1.5} A D_{L i^{+}}^{0.5} C_{L i}+v^{0.5} $$, where D_Li⁺ represents the lithium-ion diffusion coefficient, which can be derived from the peak current (I_p). n is associated with the number of electrons transferred, A represents the electrode area, v denotes the sweeping rate, and C_Li⁺ signifies the concentration of lithium ions. Notably, the cell with a 3D N-Ti₃C₂T_x/S cathode demonstrated the steepest slopes in both the reduction and oxidation processes [Figure 5D and E and Supplementary Figure 10B]. This observation clearly indicates that the D_Li⁺ values of 3D N-Ti₃C₂T_x/S cathodes are significantly greater than those of 3D Ti₃C₂T_x/S cathodes.

Figure 5. (A) TGA curves in Ar atmosphere. (B) 3D N-Ti₃C₂T_x/S and 3D Ti₃C₂T_x/S cathodes at the scanning rate of 0.1 mV s^-1. (C) The CV curves of 3D N-Ti₃C₂T_x/S cathodes at different scan rates. The plot of CV peak of (D) R₁ (S₈-Li₂S₄) and (E) R₂ (Li₂S₄-Li₂S) versus the square root of scan rates. (F) The GITT plot of the cell with 3D N-Ti₃C₂T_x/S.

To investigate the internal resistances of Li-S batteries with 3D N-Ti₃C₂T_x/S cathodes during the charging and discharging, the galvanostatic intermittent titration technique (GITT) was employed [Figure 5F]. The internal resistance associated with Li₂S nucleation and Li₂S activation was discernible through the depth of the dips observed in the charging and discharging profiles. Notably, the cell with a 3D N-Ti₃C₂T_x/S cathode shows a low discharge/charge polarization voltage plateau. The low potential variation between the points of Li₂S nucleation and Li₂S activation suggests the reduced polarization and enhanced kinetic efficiency.

Figure 6A presents the cycling performance of 3D N-Ti₃C₂T_x/S and 3D Ti₃C₂T_x/S electrodes at 0.2 C. The 3D N-Ti₃C₂T_x/S electrode exhibits an initial discharge capacity of 1,115 mAh g^-1, maintained at 812.1 mAh g^-1 even after 150 cycles, significantly higher than that of the 3D Ti₃C₂T_x/S. The long-term cycling performance at 0.5 and 1 C is illustrated in Figure 6B and C, respectively. The 3D N-Ti₃C₂T_x/S electrode delivered a high initial capacity of 937.0 and 853.7 mAh g^-1, and maintained 578.5 mAh g^-1 (500 cycles) and 462.5 mAh g^-1 (1,000 cycles) at 0.5 and 1 C, corresponding to a low decay rate of 0.076% and 0.046% per cycle, respectively. Both are much better than that of 3D Ti₃C₂T_x/S electrodes. The 3D N-Ti₃C₂T_x/S electrode outperformed the 3D Ti₃C₂T_x/S electrode in terms of reversible capacity and cycling stability. These results collectively demonstrate that nitrogen doping improves the adsorption and sulfur conversion capabilities of 3D N-Ti₃C₂T_x, significantly enhancing the capacity and cycling stability of Li-S batteries.

Figure 6. Cycle performance of 3D Ti₃C₂T_x/S and 3D N-Ti₃C₂T_x/S cathodes at (A) 0.2 C, (B) 0.5 C, and (C) 1 C. (D) The galvanostatic charge-discharge (GCD) profiles based on 3D N-Ti₃C₂T_x/S and 3D Ti₃C₂T_x/S cathode at 0.1 C. (E) Rate performance of 3D Ti₃C₂T_x/S and 3D N-Ti₃C₂T_x/S cathodes. (F) The GCD profiles of 3D N-Ti₃C₂T_x/S at 0.1, 0.2, 0.5, 1, 2, and 3 C. Nyquist plots of 3D N-Ti₃C₂T_x/S and 3D Ti₃C₂T_x/S cathodes (G) before and (H) after cycles.

At 0.1 C, the battery demonstrates a distinctive charging plateau and two discharging plateaus [Figure 6D]. The charging plateau corresponds to the transformation of Li₂S₂/Li₂S into S₈. At the discharge process, the higher potential plateau is linked to the conversion of S₈ into soluble high-order LiPSs, while the lower plateau is associated with the reduction of soluble LiPSs to Li₂S₂/Li₂S, which is consistent with the discussion of CV curves. Due to enhanced affinity of 3D N-Ti₃C₂T_x to LiPSs and its catalytic performance, the kinetics of LiPS transformation have been markedly improved. Consequently, the polarization voltage (∆E) measures 125.3 mV, noteworthy smaller than that of 3D Ti₃C₂T_x.

The rate performance tests were conducted at current densities ranging from 0.1 to 3 C [Figure 6E]. The 3D N-Ti₃C₂T_x/S electrode achieved capacities of 1,274.1, 1,128.9, 876.8, 772.5, 694.3, and 651.2 mAh g^-1 at 0.1, 0.2, 0.5, 1, 2, and 3 C, respectively. When the current density is back to 0.2 C, the discharge capacity of 1,070.6 mAh g^-1 was achieved. In contrast, the 3D Ti₃C₂T_x/S electrode shows lower discharge specific capacities of 966.9, 833.5, 738.6, 675.8, 568.2, and 192.1 mAh g^-1 at the same rates, which means lower utilization of active material at high-rate conditions.

Figure 6F and Supplementary Figure 11 depict the GCD curves of the 3D N-Ti₃C₂T_x/S and 3D Ti₃C₂T_x/S electrodes at different rates, respectively. The 3D N-Ti₃C₂T_x/S electrode displayed two typical discharge plateaus, corresponding to the conversion of S₈ to long-chain LiPSs and the subsequent conversion to Li₂S₂/Li₂S, respectively. All discharge curves of 3D N-Ti₃C₂T_x/S exhibited two characteristic plateaus even at a high current of 3 C. However, the discharge curves of 3D Ti₃C₂T_x/S show vague plateaus, indicating lower sulfur utilization than 3D N-Ti₃C₂T_x/S.

EIS tests further validated the improvement in electrochemical performance before and after 100 cycles at 0.5 C [Figure 6G and H]. Compared to the 3D Ti₃C₂T_x/S electrode, the pristine EIS curve of the 3D N-Ti₃C₂T_x/S electrode exhibits a smaller semicircle shape in the high-frequency region, indicating that nitrogen doping enhances the conductivity of Ti₃C₂T_xMXenes. After 100 cycles, the R_ct of the 3D N-Ti₃C₂T_x/S electrode is significantly lower than that of 3D Ti₃C₂T_x/S, which is attributed to the large surface area of 3D N-Ti₃C₂T_x for electrolyte infiltration and lithium-ion transport. Additionally, nitrogen doping promotes the formation and conversion of Li₂S. This further underscores that the porous structure of 3D N-Ti₃C₂T_x can provide ample diffusion pathways, and nitrogen doping can effectively immobilize polysulfides, reducing irreversible consumption of sulfur and lithium passivation caused by shuttle effects, enabling the battery to maintain high electron/ion transfer rates, thereby enhancing the electrochemical performance. To explore the potential for practical application, the cells with high sulfur loadings (2.4 and 4.2 mg_s cm^-2) were assembled. As depicted in Supplementary Figure 12, the cell with 2.4 mg_s cm^-2 sulfur loading exhibits an areal capacity close to 1.8 mAh cm^-2 at 2 C, and the cell with 4.2 mg_s cm^-2 sulfur loading can maintain a capacity of 3 mAh cm^-2 at 1 C, experiencing a capacity reduction only at 2 C. This observation underscores the electrochemical performance of 3D N-Ti₃C₂T_x /S with a high sulfur loading.

The deterioration of lithium anodes is related to the severe shuttle effects^[57]. Additionally, the degree of polysulfide dissolution can be evaluated by the simple optimal photograph of the separators after cycling. Supplementary Figure 13A depicts the scanning images of the disassembled lithium anode. It can be observed that the lithium anode surface is exceptionally smooth before cycling. After cycling, the lithium anode surface with 3D Ti₃C₂T_x/S exhibits pulverization and cracks, along with a significant accumulation of dead lithium [Supplementary Figure 13B], while the battery with 3D N-Ti₃C₂T_x/S is relatively smoother, and only a small amount of lithium dendrites can be observed [Supplementary Figure 13C]. As shown in Supplementary Figure 13D, the separator of the 3D Ti₃C₂T_x/S cell (left) is covered with a thick layer of dark yellow substance, corresponding to the LiPSs. However, only a small amount of yellow precipitate remains on the separator with the 3D N-Ti₃C₂T_x/S cathode (right). All the above results show that 3D N-Ti₃C₂T_x can effectively alleviate the shuttle effect in Li-S batteries, and excellent electrochemical performance can be achieved.

CONCLUSIONS

In summary, few-layered Ti₃C₂T_xwas obtained, and 3D N-Ti₃C₂T_x was successfully synthesized by spray drying. Positively charged HCl-melamine is electrostatically attached to the surface of negatively charged MXenes, preventing the aggregation of MXene nanosheets and enhancing its thermal stability. Spray drying facilitates the transformation of unique nanoflower-shaped microspheres, endowing 3D N-Ti₃C₂T_x with a large specific surface area and pore volume. Due to the synergistic effect of the large specific surface area and nitrogen doping, MXenes exhibit increased adsorption capacity and catalytic conversion ability for LiPSs, effectively preventing the shuttle effect. Therefore, high-capacity retention (578.5 mAh g^-1 after 500 cycles at 0.5 C, and 462.5 mAh g^-1 after 1,000 cycles at 1 C remained), superior rate performance (651.2 mAh g^-1 at 3 C), and excellent long-term cycling performance (capacity fading rate of 0.076% per cycle at 0.5 C and 0.046% per cycle at 1 C) have been achieved. This work broadens the application of MXenes and provides new insight into the practical use of Li-S batteries.