A high-sulfur-loading freestanding SPANI/MWCNT electrode for high-performance lithium–sulfur batteries

Abstract

Sulfurized polyaniline (SPANI) serves as a novel advanced cathode material that can chemically immobilize the sulfur active substance via robust intramolecular C–S bonding, thereby eliminating the shuttle effect in conventional lithium–sulfur battery systems. As a result, SPANI can exhibit commendable cycling stability and is compatible with commercial carbonate electrolytes. However, the conductivity of SPANI and its discharge product, Li₂S, is exceedingly low, severely hampering the discharge capacity of SPANI at high current rates. To solve this issue, a freestanding SPANI/MWCNT (multi-walled carbon nanotube) electrode with an internal three-dimensional conductive network structure has been successfully constructed for the first time. Compared to traditional coated SPANI electrodes, which deliver only 324.9 mAh g⁻¹ at a high rate of 8C (13 376 mA g⁻¹), the freestanding design enhances the capacity by approximately 1.5 times, reaching 480.5 mAh g⁻¹. Additionally, stemming from the robust mechanical property and the interwoven winding characteristic of MWCNTs, the electrode enables stable operation even with a high sulfur loading of 13.0 mg cm⁻², substantially exceeding the feasible limits of the traditional slurry-coating technique. Furthermore, this electrode demonstrates a consistent areal capacity reaching 8.30 mAh cm⁻² at a current density of 5 mA cm⁻², maintaining 86.4% of its original capacity after 100 charge–discharge cycles. Besides, the freestanding electrode craft can eradicate the need for expensive metal current collectors, binders, and the toxic binder dispersant NMP (1-methyl-2-pyrrolidinone).

---

Introduction

In the contemporary global context, energy scarcity and environmental contamination are intertwined challenges. Consequently, the impetus for promoting greener and more environmentally benign transportation alternatives has become an imperative trend and is crucial for the sustainable progression of society. This has spurred rapid expansion in the market for zero-emission electric vehicles. However, the comparatively modest energy density of conventional lithium-ion batteries represents a substantial limitation in enabling electric vehicles to satisfy the requirements for long-distance travel. Therefore, the development of a battery system with enhanced energy density is of paramount importance.^1,2 With a theoretical energy density reaching up to 2600 Wh kg⁻¹, lithium–sulfur (Li–S) batteries stand out among advanced rechargeable storage technologies, delivering energy performance that significantly exceeds – by nearly an order of magnitude – that of typical lithium-ion cells.^3,4

In typical Li–S battery configurations, elemental sulfur serves as the cathode, while metallic lithium functions as the anode. During discharge, sulfur undergoes reduction to yield polysulfide intermediates (Li₂S₄ to Li₂S₈), which readily engage in nucleophilic reactions with carbonate-based electrolytes. These side reactions lead to the breakdown of carbonate solvents, ultimately compromising battery stability over time.⁵ To circumvent this issue, the majority of Li–S batteries utilize ether-based electrolytes, which can solubilize the polysulfide intermediates without undergoing chemical degradation.⁶ During cycling, sulfur is transformed into liquid-phase polysulfides and subsequently reduced to solid Li₂S. Nevertheless, the redox reactions involved inherently promote the diffusion of dissolved polysulfides between electrodes, causing the so-called shuttle effect. This undesirable process leads to rapid capacity fade and performance degradation.^7,8

Researchers have proposed multiple methods to tackle this problem from different perspectives. These include modifying the sulfur cathode materials to enhance their stability,^9,10 introducing interlayers within the battery configuration to physically or chemically impede the migration of polysulfides,^11,12 and designing novel binders^13,14 or electrolytes with improved functional properties.^15,16 Our research group has also been actively engaged in this endeavor, conducting investigations such as the introduction of meticulously crafted sulfur cathodes^17–19 and the modification of polypropylene (PP) films.²⁰ While these approaches have demonstrated the potential to mitigate the shuttle effect to some extent, they have not succeeded in completely eradicating it.

In 2002, Wang et al. reported the successful preparation of sulfurized polyacrylonitrile (SPAN), a polymer-based cathode material, by heating polyacrylonitrile and elemental sulfur as raw materials at 300 °C.²¹ In 2018, Tsao et al. obtained another sulfurized polymer, sulfurized polyaniline (SPANI), via a similar process of heating polyaniline and elemental sulfur at 300 °C.²² In 2023, Rao et al. reported a PANI/S composite synthesized at a lower temperature of around 160 °C. At the employed temperature, PANI and S cannot undergo a sulfurization reaction to form covalent bonds. Consequently, the electrochemically active component remains in the PANI/S is still elemental sulfur.²³ In both SPAN and SPANI, sulfur atoms are covalently bonded to the polymer backbone's carbon atoms as short chain segments (–S_2–4–), which inhibits the generation of polysulfides intermediates during discharge and instead results in the direct formation of lithium sulfide.^24,25 As a result, lithium–sulfur batteries employing SPAN or SPANI cathodes can be compatible with both ether and carbonate electrolytes.^22,24 Sulfurized polymers undergo a solid-to-solid phase reaction during discharge, effectively eliminating the shuttle effect that plagues traditional Li–S batteries. This endows the cathode material with superior cycling performance.^22,24,25

Compared to ether electrolytes, carbonate electrolytes exhibit a lower flash point, a more affordable cost, and enhanced safety.²⁶ Beyond their inherent electrochemical benefits, sulfurized polymers exhibit excellent compatibility with the carbonate electrolytes commonly used in lithium-ion batteries. This favorable interaction plays a crucial role in enhancing their industrial applicability, positioning sulfurized polymers as promising candidates for future-oriented energy storage solutions.

In spite of these benefits, the low intrinsic electronic conductivity of both sulfurized polymers and the resulting lithium sulfide restricts the battery's rate performance.^22,23,27,28 To enhance the conductivity of sulfurized polymers, the research community has undertaken extensive efforts. For example, Yin et al. and Konarov et al. ingeniously compounded high-conductivity graphene with SPAN.^29,30 In this composite structure, SPAN particles are uniformly embedded within the interlayers of graphene sheets, thereby establishing a rapid electron conduction pathway. Other researchers have also explored the use of highly conductive transition metal compounds, such as NiS₂ or CoS₂, to fabricate composite materials containing SPAN. These composites effectively improve the rate performance of the cathode.^31,32 Similarly, Zou et al. successfully enhanced the rate performance of SPANI by doping it using selenium with a relatively high conductivity of 1 × 10⁻³ S m⁻¹.³³

In a pioneering effort to augment the rate performance of SPANI, a freestanding sulfurized polyaniline/multi-walled carbon nanotube (SPANI/MWCNT) cathode material was fabricated via a straightforward filtration technique (Scheme 1). This innovative approach not only enhanced the conductivity of SPANI through the incorporation of a three-dimensional conductive network formed by highly conductive MWCNTs but also eliminated the need for conventional components. Specifically, the inherent mechanical robustness of MWCNTs obviated the necessity for costly metal current collectors, binders, and the toxic binder dispersant NMP (1-methyl-2-pyrrolidinone), which are typically employed in the construction of traditional coated sulfur electrodes.³⁴ It is well documented that metallic current collectors are susceptible to corrosion under the electrochemical conditions during battery cycling. Such corrosion gradually compromises the contact at the interface of the cathode material/current collector, thereby accelerating capacity degradation and ultimately leading to battery malfunction.³⁵ Additionally, conventional polymeric binders are electrically insulating and exhibit negligible ionic conductivity. This leads to increased interfacial resistance and longer electron/ion transport paths within the electrode, partially hindering the battery's rate performance.³⁶ Avoiding the aforementioned drawbacks, the SPANI/MWCNT cathode achieves improved structural stability and facilitates the fabrication of a binder- and current-collector-free electrode capable of facilitating elevated sulfur loadings (as high as 13 mg cm⁻²) and sustaining excellent electrochemical characteristics.

Scheme 1 Structural elucidation of SPANI and a schematic representation of the fabrication protocol for a freestanding SPANI/MWCNT composite electrode.

Experimental section

Synthesis of the SPANI cathode material

A total of 6 g of elemental sulfur and 1 g of polyaniline were initially ground together using a mortar and pestle until a uniform mixture was obtained. Subsequently, the mixture was transferred to a tube furnace and subjected to thermal annealing at 320 °C for 10 hours under a protective argon atmosphere, yielding the target sulfurized polyaniline (SPANI) cathode material.

Preparation of the SPANI/MWCNT freestanding sulfur electrode

To construct a freestanding sulfur cathode, a defined amount of sodium dodecylbenzene sulfonate (SDBS), an anionic surfactant, was initially dissolved in 10 mL of ultrapure water. Next, 0.02 g of multi-walled carbon nanotubes (MWCNTs) was added and stirred to get a suspension. Following a similar procedure, another suspension was obtained using SDBS and synthesized SPANI (0.02 g). Then, the two above suspensions were mixed. The mixture was stirred for 5–10 minutes and then subjected to 30 minutes of ultrasonication to promote uniform dispersion. This process of mechanical stirring followed by sonication was repeated several times to enhance homogeneity. The mixed suspension was then filtered and rinsed, producing a freestanding film, which was subsequently cut into 0.8 cm diameter disks. These electrode disks were dried under vacuum at 60 °C. The final sulfur loading of the electrodes was 1.0 mg cm⁻². Using the same preparation route and maintaining a mass ratio of 1 : 1 for SPANI and MWCNTs, additional electrodes with sulfur loadings of 2.2, 5.0, 9.8, and 13.0 mg cm⁻² were produced. For systematic comparison, freestanding electrodes with SPANI/MWCNT ratios of 3 : 1, 2 : 1, and 1 : 2 (total mass is 0.04 g), as well as freestanding electrodes with SPANI/acetylene black ratios of 3 : 1, 2 : 1, 1 : 1, and 1 : 2 (total mass is 0.04 g) were also prepared following the same procedure.

Preparation of the coated SPANI sulfur electrode

To prepare the coated-type electrode, a suspension was formulated by mixing 140 mg of SPANI with 40 mg of acetylene black and 20 mg of PVDF in an NMP solvent. The resulting slurry was evenly coated onto aluminum foil, which served as the conductive backing. After drying the coated substrate in air at 60 °C, circular samples measuring 0.8 cm in diameter were cut. These were subsequently placed in a vacuum oven for 12 hours at the same temperature to ensure complete solvent evaporation. The final sulfur loading of the electrode was estimated to be around 0.9 mg cm⁻². For comparative assessment, MWCNTs were substituted for acetylene black as the conductive agent in preparing coated sulfur cathodes (sulfur loading: 0.9 mg cm⁻²) while all other formulation parameters and processing conditions were held constant.

Assembly of coin cells

All CR2032 coin cells were assembled within an argon-atmosphere glovebox. The Li|electrolyte|sulfur cathode coin cells were constructed using lithium metal as the anode and either freestanding or coated SPANI-based sulfur cathodes. A Celgard 2400 membrane functioned as the separator, and the electrolyte comprised 1 mol L⁻¹ lithium hexafluorophosphate (LiPF₆) in a 1 : 1 (v/v) volumetric ratio of diethyl carbonate (DEC) and ethylene carbonate (EC). To guarantee full wetting of the electrode, 60 µL of electrolyte was dispensed into each coin cell. Besides, electrolyte conductivity was determined with a symmetric SS (stainless steel)|electrolyte|SS (stainless steel) coin cell. The Macmullin number of the freestanding electrode was obtained from a symmetric FE (freestanding electrode)|electrolyte|FE (freestanding electrode) coin cell.

Characterization of materials

The structural phases of the as-prepared materials were identified through X-ray diffractometer (XRD) analysis using the Smart Lab SE system. Thermal decomposition profiles were recorded by performing thermogravimetric analysis (TGA) on the STA 449 F5 setup. Morphological features and spatial element distribution were examined with scanning electron microscopy (SEM, GeminiSEM 300) integrated with an energy-dispersive spectroscopy (EDS) module. X-ray photoelectron spectroscopy (XPS) measurements (Thermo Scientific K-Alpha) were employed to investigate surface-level chemical environments and bonding states. Functional groups were detected via Fourier transform infrared (FTIR) spectroscopy (Nicolet iS20), and the elemental composition – specifically the contents of nitrogen, carbon, hydrogen, and sulfur – was assessed using an Elementar Unicube elemental analyzer. Brunauer–Emmett–Teller (BET) specific surface area analysis was conducted on an analyzer (Tristar 3020). Raman spectra were obtained using a Raman microscope (LabRAM HR Evolution) with a wavelength of 532 nm.

Electrochemical test

For voltammetry, a scanning potential window from 1 to 3 V was selected. Galvanostatic cycling tests were conducted at different current intensities on a CT3002AU tester, where the 1C condition was standardized as 1672 mA g⁻¹. The lithium-ion diffusion coefficient (D_Li⁺) was determined through fitting procedures using eqn (1) and (2):⁸

Z_Re = R_s + R_ct + σω^−1/2 (2)

In these equations, R denotes the molar gas constant (8.314 J mol⁻¹ K⁻¹), T refers to the system temperature, fixed at 298 K. The variable A specifies the active electrode surface area, while n indicates the number of electrons exchanged per redox reaction. The symbol F corresponds to the Faraday constant (96 485C mol⁻¹). C represents the concentration of lithium ions in mol cm⁻³, and σ characterizes the Warburg factor associated with ion diffusion resistance. Here, electrochemical impedance spectroscopy (EIS) images were collected in a frequency range, from 200 000 Hz down to 0.1 Hz, using a PARSTAT MC electrochemical station.

The ionic conductivity of the electrolyte (σ_el) was determined using EIS over the frequency range from 1 000 000 Hz to 1 Hz. The conductivity, σ_el, was determined using eqn (3):³⁷

where l and A are the thickness and geometric area of the polypropylene (PP) separator, respectively, and R is the bulk resistance of the symmetric SS|electrolyte|SS coin cell.

The MacMullin number (N_M) of the freestanding electrode was obtained through the EIS test from 100 000 Hz to 0.01 Hz. The N_M value was evaluated using eqn (4) and (5):^38,39

Here, τ denotes the tortuosity of the electrode, ε represents the porosity of the electrode, R_ion indicates the ionic resistance of the symmetric FE|electrolyte|FE coin cell, A signifies the area of the electrode, d denotes the thickness of the electrode, and D₀ and D_eff represent the apparent and effective Li⁺ diffusion coefficients in the freestanding electrode, respectively.

Results and discussion

Fig. 1 illustrates the morphology and dimensions of the freestanding electrode (SPANI/MWCNT = 1 : 1), revealing its regular circular shape with a diameter of 3.5 cm. By contrast, the freestanding electrode prepared with SPANI and acetylene black at a 1 : 1 mass ratio exhibits a pronounced crack (Fig. S1a, SI). This failure is ascribed to the non-fibrillar, granular morphology of acetylene black (Fig. S2, SI), which – unlike MWCNTs –cannot form an interwoven scaffold. Consequently, electrodes with other SPANI/acetylene-black ratios (3 : 1, 2 : 1, 1 : 2) underwent analogous disintegration. When MWCNTs are employed, the critical threshold for mechanical integrity lies between 25 and 33 wt%. Specifically, at a SPANI/MWCNT ratio of 3 : 1, the electrode is fragile because the MWCNT content is below the threshold (Fig. S1b, SI). Raising the MWCNT fraction to 33% (2 : 1) or 67% (1 : 2) yields crack-free, freestanding electrodes that possess certain mechanical strength (Fig. S1c–f, SI). Fig. 2 displays the optical photographs of freestanding electrodes (SPANI/MWCNT = 1 : 1) with different sulfur loadings, highlighting their benign flexibility. A clear trend is observed wherein increased sulfur content results in greater electrode thickness. The measured thicknesses corresponding to sulfur loadings of 1.0, 2.2, 5.0, 9.8, and 13.0 mg cm⁻² are 0.053, 0.174, 0.362, 0.900, and 1.400 mm, respectively.

Fig. 1 (a) Morphology and (b) size of the freestanding electrode (SPANI/MWCNT = 1 : 1) obtained by using a filtration method.

Fig. 2 Optical photographs of freestanding electrodes (SPANI/MWCNT = 1 : 1) with increasing sulfur mass loadings: (a) and (b) 1.0, (c) and (d) 2.2, (e) and (f) 5.0, (g) and (h) 9.8, and (i) and (j) 13.0 mg cm⁻².

As depicted in Fig. 3a, the peak of elemental sulfur (S) between 20° and 30° is attributed to the orthorhombic structure.⁹ Meanwhile, polyaniline (PANI) manifests a broad diffraction peak spanning from 12° to 30°. The SPANI sample exhibits a wide peak in the range of 15° to 28°, reflecting the presence of poorly ordered carbon structures rather than fully graphitized phases.²⁵ Notably, no crystalline peaks of elemental sulfur are detected in the SPANI sample, implying that sulfur exists in SPANI as short-chain sulfur in an amorphous state.^22,25 The MWCNT sample displays a sharp characteristic peak at 25.7°, indicative of the (002) crystal facet in the hexagonal graphite structure.⁴⁰ This crystal plane is also prominently visible in the diffraction peak of the freestanding electrode, thereby demonstrating the effective compounding of MWCNTs and SPANI.

Fig. 3 (a) XRD and (b) FTIR results of S, PANI, SPANI, MWCNT, and the freestanding electrode.

As illustrated in the FTIR spectrum presented in Fig. 3b, the spectrum of elemental sulfur features a characteristic peak corresponding to the S–S bond at 465.7 cm⁻¹.^41,42 For polyaniline, a series of distinct infrared absorption signals can be detected, including 3377.2 cm⁻¹ corresponding to –NH₂ stretching, 1587.1 cm⁻¹ attributed to C=N vibrations, and 1498.4 cm⁻¹ related to C=C bonds. An additional peak appears at 1304.1 cm⁻¹ is assigned to C–N stretching, while absorption at 1145 cm⁻¹ and 829.2 cm⁻¹ is associated with C–H bond vibrations. Notably, the dual C–H features at 1145 cm⁻¹ and 829.2 cm⁻¹ are nearly undetectable following the sulfurization of PANI, thereby confirming that sulfur atoms have effectively replaced the hydrogen atoms on the aromatic ring.^25,41,43

In the FTIR spectrum of SPANI, a broad absorption around 3420.6 cm⁻¹ is associated with –NH– functionalities. The band at 1548.6 cm⁻¹ corresponds to C=N stretching vibrations, while the signal near 1384.2 cm⁻¹ is linked to carbon–carbon double bonds within the aromatic ring. A distinct feature at 1267 cm⁻¹ can be ascribed to vibrations involving carbon–nitrogen linkages. Compared to PANI, these three vibrational modes exhibit a noticeable shift toward lower wavenumbers, implying that sulfur incorporation alters the molecular framework and affects electronic delocalization. Furthermore, absorption bands at 727 cm⁻¹ and 469 cm⁻¹ reflect the presence of C–S and S–S bonds in SPANI, respectively.^25,41,42 These features confirm the successful sulfurization of PANI and the covalent binding of sulfur to the carbon backbone.

The FTIR spectrum of MWCNTs displays a notable absorption band at 3435.6 cm⁻¹, assigned to –OH stretching vibrations on the nanotube surface. Another distinct peak appears at 1630.5 cm⁻¹, originating from the –C=O stretching of surface –COOH groups.^44,45 This peak is also present in the freestanding electrode spectrum. Furthermore, the freestanding electrode spectrum exhibits characteristic peaks for C=N, C=C, C–N, and C–S bonds ascribed to the SPANI, indicating the presence of these functional groups within the electrode material.

In Fig. 4a, the peaks below 500 cm⁻¹ for sulfur correspond to the vibration of the S–S bond in the octatomic ring of elemental sulfur.²⁵ These peaks are absent in both SPANI and the freestanding electrode, consistent with the XRD results shown in Fig. 3. The two peaks around 1350 and 1580 cm⁻¹ reflect carbon atomic defects (D band) and the degree of graphitization (G band). In PANI and SPANI, both peaks are weak, indicating a carbon structure with low graphitization. The prominent G peak in the MWCNT spectra suggests the presence of a sp² graphitic structure and the relatively distinct D peak may result from defects introduced by abundant oxygen functional groups on the surface.^11,25 The freestanding electrode exhibits combined characteristics of both SPANI and MWCNT, confirming their coexistence within the composite. Fig. 4b shows the BET test results for the freestanding electrode. A specific surface area of 75.79 m² g⁻¹ and an average pore diameter at 16.44 nm (1.8–105 nm range) are obtained. This hierarchical porosity provides continuous pathways for rapid Li⁺ diffusion and guarantees full access of the electrolyte to the electro-active sites, thereby accelerating the redox kinetics.

Fig. 4 (a) Raman results of S, PANI, SPANI, MWCNTs, and the freestanding electrode and (b) BET result of the freestanding electrode.

Fig. S3a (SI) presents the elemental analysis of SPANI, with the following determined weight percentages: nitrogen (N) at 6.565 wt%, carbon (C) at 32.998 wt%, hydrogen (H) at 0.813 wt%, and sulfur (S) at 58.305 wt%. As depicted in Fig. S3b (SI), elemental sulfur commences sublimation and undergoes a weight loss at 200 °C. In contrast, SPANI exhibits significantly greater thermal stability, with no substantial occurrence of weight loss up to 350 °C. The improved structural stability of SPANI arises from the establishment of covalent C–S linkages between sulfur and carbon atoms, which reinforces the polymer framework.²⁵

As illustrated in Fig. 5a, the SEM image of SPANI powder indicates that most particles are in the micrometer size range. These particles are agglomerates of smaller nanoparticles, as further illustrated in Fig. S4a (SI). Fig. 5b and Fig. S4b (SI) depict SEM images of multi-walled carbon nanotube (MWCNT) powders. The one-dimensional MWCNTs are observed to be intricately intertwined, forming a porous structure characterized by numerous voids. Fig. 5c offers a planar view of the freestanding electrode sheet, where SPANI particles are uniformly dispersed within the conductive network established by the MWCNTs. The porous architecture of the intertwined MWCNTs facilitates efficient lithium-ion transport, thus contributing to improved electrochemical functionality of SPANI. Fig. 5d presents a cross-sectional view of the freestanding electrode, revealing an approximate thickness of 50 micrometers. The cross-sectional elemental distributions of the freestanding electrode, as revealed by EDS, are shown in Fig. 5e–h. The EDS maps clearly illustrate the distributional status of nitrogen (N), sulfur (S), and carbon (C) elements throughout the electrode. This uniform elemental distribution confirms the successful and effective compounding of SPANI and MWCNTs within the electrode structure.

Fig. 5 Scanning electron microscopy (SEM) characterization of materials and electrode morphology: (a) SPANI powder, (b) MWCNT powder, (c) top view and (d) cross-sectional structure of the freestanding electrode. (e)–(h) Energy-dispersive X-ray spectroscopy (EDS) of the electrode cross-section.

Fig. 6 illustrates the XPS characterization results for the freestanding electrode. As indicated in Fig. 6a, the broad-spectrum scan detects the presence of carbon, sulfur, nitrogen, and oxygen elements within the material. The detailed C 1s spectrum in Fig. 6b reveals several identifiable peaks, with the one at 284.7 eV corresponding to C–C and C=C structures, which are derived from both the SPANI component and the embedded MWCNTs.⁴⁶ The 285.2 eV signal is indicative of C–S bonds in SPANI,⁴⁷ while the 287.0 eV signal is associated with C–N and C=N linkages.⁴⁸ A higher-energy feature at 291.2 eV is assigned to π–π* interactions typical of conjugated systems in MWCNTs.^49,50 As depicted in Fig. 6c, the S 2p spectrum displays two prominent signals located at approximately 163.6 eV and 164.8 eV, corresponding to S 2p_3/2 and S 2p_1/2 levels, which are indicative of sulfur–carbon and short sulfur–sulfur linkages present in the SPANI framework.^25,46,51 A faint shoulder around 167.5 eV is linked to oxidized sulfur species (SO_x).⁴⁶Fig. 6d shows the N 1s spectrum, where three distinct binding energies are observed: a peak at 398.4 eV associated with pyridinic nitrogen (C=N–C), another at 400.0 eV attributed to pyrrolic N (C–NH–C), and a third signal near 402.6 eV assigned to nitrogen–oxygen functionalities.

Fig. 6 XPS investigations of the freestanding electrode: (a) full-spectrum scan, (b) deconvoluted C 1s peaks, (c) sulfur 2p core-level spectrum, and (d) nitrogen 1s binding energy profile.

Fig. S5 (SI) summarizes cycling performances of freestanding SPANI/MWCNT cathodes fabricated at three mass ratios. At SPANI : MWCNT = 2 : 1 (sulfur loading: 1.3 mg cm⁻²), the electrodes retain 729.5, 631.9, and 422.7 mAh g⁻¹ capacity after 50 cycles at 0.5C, 100 cycles at 1C, and 150 cycles at 2C, respectively, corresponding to capacity retentions of 93.4%, 87.4%, and 74.8% versus the second cycle. Increasing the MWCNT fraction to 1 : 1 (sulfur loading: 1.0 mg cm⁻²) raises the retained capacities to 827.5, 679.7, and 562.5 mAh g⁻¹ under the same protocols, with superior retentions of 99.0%, 93.4%, and 93.5%. Further elevating the MWCNT content to 1 : 2 (sulfur loading: 0.65 mg cm⁻²) yields intermediate values of 788.9, 658.4, and 517.1 mAh g⁻¹, accompanied by retentions of 92.6%, 83.7%, and 72.8%. Fig. S6 (SI) shows that both the charge transfer resistance (R_ct) and the Warburg factor decrease as the MWCNT content in the freestanding electrode increases (coin cells were measured after the first cycle). When the SPANI : MWCNT mass ratio is varied from 2 : 1 to 1 : 1 and 1 : 2, the R_ct value drops from 60.48 to 39.39 and 29.76 Ω, while the Li⁺ diffusion coefficient (D_Li⁺) rises from 0.19 × 10⁻¹¹ to 3.12 × 10⁻¹¹ and 5.48 × 10⁻¹¹ cm² s⁻¹, respectively. The drop of resistance is ascribed to the higher fraction of highly conductive MWCNTs, whereas the concurrent increase in the ionic diffusion coefficient may arise from the enlarged electrode porosity imparted by the interwoven MWCNT network. Nevertheless, the content of SPANI and MWCNTs must be appropriately controlled. When the mass ratio of SPANI to MWCNTs is 1 : 1, the freestanding electrode can achieve an optimal balance between high specific capacity and good cycling stability.

Fig. 7 compares the electrochemical behavior of the freestanding electrode and the coated electrode using acetylene black as a conductive agent. In Li–S batteries that utilize ether-based electrolytes, the discharge process typically exhibits two voltage steps, generally found near 2.3 V and 2.0 V.^6,28 These distinct plateaus arise from a gradual chemical change in sulfur – starting with the reduction of S₈ into dissolved long-chain intermediates (Li₂S_x, where x varies from 4 to 8), followed by their transformation into shorter-chain solids such as Li₂S₂ and Li₂S during the later stages of the reaction.^7–9 However, as illustrated in Fig. 7a, as opposed to the dual-plateau discharge behavior typically observed with elemental sulfur cathodes, both the coated electrode and the freestanding electrode display a single inclined characteristic discharge voltage platform associated with the sulfurized polymer. This unique behavior can be attributed to the distinct charge–discharge mechanism of SPANI, which is characterized by a one-step “solid–solid phase” reaction.^22,24,25 Under a discharge condition of 0.2C, the initial capacity delivered by the coated electrode is 988.2 mAh g⁻¹, whereas its freestanding equivalent achieves a higher value of 1122.6 mAh g⁻¹. Despite this fact, both electrodes experience a notable capacity fade by the second cycle, falling to 776.7 mAh g⁻¹ and 890.4 mAh g⁻¹, respectively. This early-stage reduction in performance is primarily linked to unwanted side effects at the electrode–electrolyte interface, such as electrolyte breakdown and the emergence of a solid electrolyte interphase (SEI) film on the electrode surface.^25,26 The first discharge also shows a more pronounced voltage dip than subsequent cycles, likely due to the solid-phase redox behavior of SPANI, which causes higher initial polarization.²⁵ Since the initial discharge process involves irreversible electrochemical steps, not all of the capacity is recoverable in later cycles. Moreover, the freestanding sample maintains a smaller charge–discharge voltage difference of 0.46 V, while that of the coated electrode is 0.59 V, which reflects diminished polarization behavior and more favorable electron/ion transport pathways. As indicated in Fig. 7b, over 50 cycles at 0.2C, the coated sample stabilizes at 745.1 mAh g⁻¹, while its freestanding counterpart maintains a higher value of 828.8 mAh g⁻¹, highlighting better cycling durability. Additionally, Fig. S7 (SI) shows optical images of separators after 50 cycles at 0.2C for batteries assembled with the coated electrode and the freestanding electrode. Both separators retained a pristine white surface without any visually detectable yellow staining attributable to polysulfide dissolution. This observation provides compelling evidence that SPANI can effectively suppress the polysulfide-shuttle phenomenon that typically plagues elemental-sulfur cathodes. Fig. 7c and d present the CV curves for both types of electrodes at a scan rate of 0.1 mV s⁻¹. The reduction peak in the first sweep appears at a slightly lower potential than in the subsequent scans, echoing the earlier discharge characteristics observed in Fig. 7a. Both the CV profiles exhibit a single redox couple in Fig. 7c and d. A cathodic peak at 1.8–1.9 V corresponding to the reduction of S_2–4 to Li₂S₂/Li₂S, and its anodic counterpart around 2.30 V indicating the reversible oxidization process.²⁵ The freestanding electrode shows stronger redox activity, reflected by sharper and more distinct peaks, as well as a reduced potential gap between oxidation and reduction processes, indicative of better kinetics and reversibility.

Fig. 7 (a) Charge–discharge curves and (b) capacity retention results for the coated (acetylene black as a conductive agent) electrode (sulfur loading: 0.9 mg cm⁻²) and the freestanding electrode (sulfur loading: 1.0 mg cm⁻²) under 0.2C operation. (c) and (d) Electrochemical response curves obtained via cyclic voltammetry for the coated electrode and the freestanding electrode, respectively.

As a comparison, MWCNT was used as a conductive agent instead of acetylene black to prepare the coated electrode, and the battery test results are shown in Fig. S8 (SI). As seen in Fig. S8a and b (SI), the voltage difference between the charge and discharge curve is 0.79 V and the redox peak potential gap in the cyclic voltammetry curve is 0.67 V. Fig. S8c (SI) shows that the specific capacity of the electrode after 50 cycles at 0.2C is 664.2 mAh g⁻¹, which is inferior to the freestanding electrode (SPANI/MWCNT = 1 : 1) and the coated electrode using acetylene black as a conductive agent. Fig. S8d (SI) further reveals that the charge-transfer impedance is 157.9 Ω. It is well known that MWCNTs intrinsically tend to agglomerate, disrupting the conductive network, ultimately lowering active-material utilization and degrading the battery performance. Unlike the fabrication of freestanding electrodes using SDBS as a dispersant in aqueous media, coated cathodes prepared using MWCNTs as conductive agents do not employ any dispersants. Consequently, their electrochemical performances are slightly inferior compared to that of freestanding electrodes.

As shown in Fig. 8a and b, the charge–discharge curves of both the coated (acetylene black as a conductive agent) and freestanding electrodes were tested under varying current intensities. The freestanding electrode consistently demonstrates better adaptability to rate changes than the coated one. Specifically, for the coated sample, the discharge capacities at the current rates of 0.2, 0.5, 1, 2, 4, and 8C are measured as 751.8, 719.3, 681.1, 616.2, 508.8, and 324.9 mAh g⁻¹, respectively. In comparison, the freestanding electrode delivers higher performance at the same rates, with corresponding capacities reaching 828.7, 799.5, 765.7, 711.1, 630.4, and 480.5 mAh g⁻¹. The distinction between the two electrodes becomes more pronounced with increasing current, as illustrated in Fig. 8c, where the disparity in discharge capacity continues to widen at high rates. This trend confirms that the freestanding architecture offers significantly enhanced high-rate discharge behavior. For long-term cycling analysis, Fig. 8d presents the performance of coated (acetylene black as a conductive agent) and freestanding electrodes operated at 2C over 200 continuous cycles. For the freestanding electrode, the reversible capacity begins at 601.6 mAh g⁻¹ in the second cycle and steadily decreases to 509 mAh g⁻¹ by the 200th cycle. This gradual reduction corresponds to a retention rate of 84.6%, reflecting the electrode's strong durability under prolonged high-rate cycling. In contrast, the corresponding specific capacities of the coated electrode using acetylene black as a conductive agent declined to 461.2 and 354.5 mAh g⁻¹, with an inferior capacity retention of 76.9%.

Fig. 8 (a) and (b) Galvanostatic charge–discharge profiles at various current densities, (c) rate capabilities, and (d) long-term cycling behaviors under the 2C conditions for the coated electrode (acetylene black as a conductive agent) (sulfur loading: 0.9 mg cm⁻²) and the freestanding electrode (sulfur loading: 1.0 mg cm⁻²).

Fig. 9a presents the electrochemical impedance spectra (EIS) and the equivalent circuit for both the coated and freestanding electrodes after the 1st, 5th, and 10th cycles. The electrochemical impedance spectra reveal two characteristic regions: a semicircular arc at high to intermediate frequencies and a linear tail at low frequencies. As interpreted from the corresponding equivalent circuit, the semicircle observed in the impedance plot arises from the resistance to charge exchange at the interface (R_ct), while the linear portion at lower frequencies is indicative of Warburg-type impedance (Z_w), which characterizes the diffusion of ions within the system.^8,33 Within this context, Z_Re and Z_Im correspond to the real and imaginary segments of the impedance spectrum. The extracted R_ct values are summarized in Table 1, showing that the freestanding electrode consistently exhibits lower charge transfer resistance than the coated electrode throughout the cycling process. The reduction in R_ct suggests enhanced electrochemical reaction kinetics, which can be ascribed to the three-dimensional conductive network formed by the interconnected MWCNT framework within the freestanding electrode. Furthermore, Fig. 9b–d present the Warburg factor (σ) of the coated and freestanding electrodes after various cycling stages. Eqn (1) and (2) served as the basis for estimating D_Li⁺ values, which are reported in detail in Table 1. Notably, the freestanding electrode invariably manifests D_Li⁺ values markedly higher than those of the coated electrode, signifying an enormously accelerated lithium-ion diffusion kinetics. The enhanced ionic diffusion coefficient may originate from the enlarged electrode porosity created by the interwoven MWCNT network, thereby accounting for the observed improvement in diffusion.⁵²

Fig. 9 (a) EIS results (the equivalent circuit is attached) and (b), (c), and (d) Warburg factors (σ) after different cycles for the coated electrode (acetylene black as a conductive agent) (sulfur loading: 0.9 mg cm⁻²) and the freestanding electrode (sulfur loading: 1.0 mg cm⁻²).

Table 1 Fitted values of R_ct and calculated values of D_Li⁺ for batteries with coated and freestanding electrodes

Sample	Cycle	R ct (Ω)	D Li^+ (cm^2 s^−1)
Coated	1st cycle	50.46	0.022 × 10^−11
Freestanding	1st cycle	39.39	3.12 × 10^−11
Coated	5th cycle	56.55	0.017 × 10^−11
Freestanding	5th cycle	29.50	4.73 × 10^−11
Coated	10th cycle	62.01	0.019 × 10^−11
Freestanding	10th cycle	33.35	5.43 × 10^−11

---

Fig. 10a displays the cycling behavior of a freestanding sulfur electrode with a loading of 2.2 mg cm⁻². After 100 cycles operated at varying current rates – specifically 0.2, 0.5, 1, and 2C – the corresponding discharge capacities recorded are 806.5, 719.8, 619.9, and 577.6 mAh g⁻¹. Taking the second cycle as a baseline, the capacity retention is found to be 89.8%, 94.5%, 95.4%, and 96.5%, respectively. From the rate performance shown in Fig. S9 (SI), it is evident that the electrode can adapt well across a broad current density range (0.2 to 4C), with specific capacities reported at 783.2, 756.5, 724.9, 677.7, and 587.0 mAh g⁻¹. Notably, after the high-rate cycling ends and the current returns to 0.2C, the electrode recovers its capacity almost entirely, suggesting strong structural durability and efficient electrochemical reversibility. As displayed in Fig. 10b, increasing sulfur loading to 5 mg cm⁻² does not significantly impair performance, with the electrode outputting 691.1 mAh g⁻¹ at 0.2C current and 654.6 mAh g⁻¹ at 0.5C after 100 charge–discharge cycles. These figures correspond to capacity retentions of 90.4% and 94.8% when compared with their respective second-cycle values. Fig. 10c and d further demonstrate that electrodes bearing larger sulfur amounts – specifically 9.8 and 13.0 mg cm⁻² – still exhibit stable operation, providing 587.8 and 539.1 mAh g⁻¹, respectively, after 100 cycles under identical current conditions of 0.2C. The corresponding capacity retention rates, relative to the reversible discharge capacities in the second cycle, are 86.1% and 88.7%, respectively. As shown in Fig. S10 (SI), the electrode with a loading of 13.0 mg cm⁻² can exhibit a discharge specific capacity of 433.7 mAh g⁻¹ after 100 cycles at a higher rate of 0.5C, maintaining a capacity retention of 100% relative to the second cycle. Additionally, the coulombic efficiencies of freestanding electrodes with various sulfur loadings are consistently close to 100% across different current rates, indicating robust electrochemical stability and reliability.

Fig. 10 Cycling performances at various current rates of freestanding electrodes with different sulfur loadings: (a) 2.2, (b) 5.0, (c) 9.8, and (d) 13.0 mg cm⁻².

Fig. 11a depicts the cycling stability of the freestanding electrode with a sulfur loading of 13 mg cm⁻² at a current density of 5 mA cm⁻². In the second cycle, its reversible areal capacity reaches 8.30 mAh cm⁻², and after 100 cycles, it gradually declines to 7.17 mAh cm⁻², indicating a retention of 86.4%. The corresponding charge–discharge voltage profiles at various cycles are presented in Fig. 11b, confirming the electrode's consistent and robust electrochemical performance even under high loading and high-rate operating conditions. Finally, the practical applications of the battery assembled using this high sulfur loading freestanding electrode were studied. As shown in Fig. 11c–f, the battery successfully powers a lamp and a fan, thereby highlighting the excellent potential of the freestanding electrode for practical applications.

Fig. 11 The electrochemical properties and practical applications of the battery with the freestanding electrode (sulfur loading: 13.0 mg cm⁻²): (a) cycling performance and (b) charge–discharge voltage profiles at 5 mA cm⁻², a lamp (c) without and (d) with a battery, and a fan (e) without and (f) with a battery.

For the freestanding cathode with a sulfur loading of 13.0 mg cm⁻², the resultant thickness (≈1.4 mm) exceeds the limit at which the semi-infinite diffusion assumption remains valid. Consequently, the apparent Li⁺ diffusion coefficient extracted from the EIS is no longer representative. To correct for the geometric constraints, the porosity (ε) and tortuosity (τ) of the electrode are quantitatively determined to yield the effective diffusion coefficient (D_eff).^38,39 According to eqn (3), the electrolyte ionic conductivity (σ_el) is first calculated to be 1.1 mS cm⁻¹ from the bulk impedance shown in Fig. 12a. Subsequently, the ionic resistance (R_ion) of the porous electrode is resolved as 14.1 Ω (Fig. 12b), giving a MacMullin number (N_M) of 0.0558 viaeqn (4). Combining this with the apparent Li⁺ diffusion coefficient (D₀ = 3.8 × 10⁻¹¹ cm² s⁻¹) obtained from the results in Fig. 12c and d, the effective diffusion coefficient (D_eff) is finally calculated to be 6.81 × 10⁻¹⁰ cm² s⁻¹ according to eqn (5).

Fig. 12 Relevant measurements on D_eff of the freestanding electrode (sulfur loading: 13.0 mg cm⁻²): EIS results of (a) the SS|electrolyte|SS coin cell, (b) the FE|electrolyte|FE coin cell, and (c) the Li|electrolyte|FE coin cell. (d) Warburg factors (δ) of the Li|electrolyte|FE coin cell.

Conclusions

In this work, a freestanding SPANI/MWCNT composite electrode featuring an internal three-dimensional conductive architecture was fabricated for the first time using a simple filtration technique, and it was applied as the cathode within lithium–sulfur batteries. This binder- and current-collector-free structure effectively combines the robust cycling stability of SPANI with the admirable electrical conductivity of MWCNT, yielding a highly integrated and efficient electrode system. As a result, it exhibits significantly enhanced specific capacity and rate performance compared with the conventional coated SPANI electrode. Additionally, due to the strong structural integrity and entangled fiber morphology of MWCNTs, it becomes possible to construct an electrode capable of accommodating up to 13.0 mg cm⁻² of sulfur loading. When tested under a current of 5 mA cm⁻², this electrode achieves an areal capacity of 8.30 mAh cm⁻², with 86.4% of the output retained after 100 charge–discharge cycles. These findings highlight the outstanding high-rate discharge performance and superior cycling durability of the freestanding electrode. This research not only aligns with the current trend of cost reduction in manufacturing but also meets the stringent requirements for green and environmentally friendly production processes.

Author contributions

Caiyu Yang and Zhongxiang Zheng: investigation, data curation, formal analysis, and writing original draft. Longlong Ma, Yayang Tian, and Qiyun Pan: resources and writing – review & editing. Peiyue Yang, Wenfei Wu, Ziyan Yang, and Yanting Ye: visualization and data validation. Dabei Wu, Yi Cao, Jinnan Xuan, and Nanfeng Xu: methodology, resources, and data curation. Lun Yang: resources and supervision. Zhong Li: conceptualization, resources, methodology, investigation, validation, data curation, and writing – review & editing.

Conflicts of interest

There are no conflicts to declare.

Data availability

The data (images, characterizations, and electrochemical measurements for the reference samples and freestanding electrodes) supporting this article have been included as part of the supplementary information (SI). Supplementary information is available. See DOI: https://doi.org/10.1039/d5nj03309h.

Acknowledgements

The authors gratefully acknowledge financial support from the National Natural Science Foundation of China (No. 52007054 and 51901074), the Hubei Provincial Natural Science Foundation (No. 2020CFB472), the Natural Science Foundation of Shanxi Province (No. 202303021212275), the Key Project of the Joint Fund for Natural Science of Hubei Province (No. 2025AFD002), and the Innovation and Development Joint Fund of Hubei Province (No. 2022CFD002).

References

1. A. Mahmood, Z. Bai, T. Wang, Y. Lei, S. Wang, B. Sun, H. Khan, K. Khan, K. Sun and G. Wang, Chem. Soc. Rev., 2025, 54, 2369–2435.
2. L. Wang, Z. Su, R. Wang, H. Liang, B. Fang and R. Mo, J. Mater. Chem. A, 2025, 13, 21116–21171.
3. Z. Yu, B. Singh, Y. Yu and L. F. Nazar, Nat. Mater., 2025, 24, 1082–1090.
4. S. Yari, A. Conde Reis, Q. Pang and M. Safari, Nat. Commun., 2025, 16, 1–15.
5. T. Yim, M.-S. Park, J.-S. Yu, K. J. Kim, K. Y. Im, J.-H. Kim, G. Jeong, Y. N. Jo, S.-G. Woo, K. S. Kang, I. Lee and Y.-J. Kim, Electrochim. Acta, 2013, 107, 454–460.
6. M. R. Kaiser, S. Chou, H. K. Liu, S. X. Dou, C. Wang and J. Wang, Adv. Mater., 2017, 29, 1700449.
7. X. Ren, H. Wu, Y. Guo, H. Wei, H. Wu, H. Wang, Z. Lin, C. Xiong, H. Liu, L. Zhang and Z. Li, Small, 2024, 20, 2312256.
8. J. Chen, W. Feng and W. Zhao, Appl. Surf. Sci., 2022, 584, 152613.
9. N. S. Reddy, C. Bakli, V. Arya and D. Ghosh, Chem. – Eur. J., 2025, 31, e202501030.
10. Q. Wang, B. Deng, X. Zhang, L. Cao, K. Wang, W. Yao, C. Chen, H. Zhao and J. Xu, J. Materiomics, 2025, 11, 100920.
11. Y. Li, H. Wu, D. Wu, H. Wei, Y. Guo, H. Chen, Z. Li, L. Wang, C. Xiong, Q. Meng, H. Liu and C. K. Chan, Adv. Sci., 2022, 9, 2200840.
12. S. Yang, D. Jiang, Q. Su, S. Yuan, Y. Guo, K. Duan, M. Xiang, J. Guo, W. Bai and S. Chou, Adv. Energy Mater., 2024, 14, 2400648.
13. Y. Wu, M. Yang, Y. Zou, S. Hou, B. Hu, S. Wang, Y. Tao and C. Yang, New J. Chem., 2022, 46, 20129–20137.
14. S. Ma, D. Qiu, L. Zhang, X. Liu, X. Ren, J. Dai, X. Gu and Y. Hou, Adv. Funct. Mater., 2025, 2503333.
15. J. Wu, R. Liu, H. Cheng, C. Tian, Y. Zhang, X. Wang, Y. Liu, W. Liu, X. Lin, C. Zhao, L. Shi, X. Feng and Y. Ma, New J. Chem., 2025, 49, 9157–9165.
16. E. Zhang, W. Han, J. Hou, S. Chen, L. Zhang, Y. Zhang, P. Dong, B. Zhang and Y. Zhang, Small Methods, 2025, 2401961.
17. Z. Li, S. Deng, H. Li, H. Ke, D. Zeng, Y. Zhang, Y. Sun and H. Cheng, J. Power Sources, 2017, 347, 238–246.
18. Z. Li, N. Zhang, Y. Sun, H. Ke and H. Cheng, J. Energy Chem., 2017, 26, 1267–1275.
19. Z. Li, P. Yang, Q. Pan, S. Jiang, L. Jiang, Y. Liu, G. Xu and Y. Chen, Energy Technol., 2021, 9, 2100418.
20. Z. Li, Q. Pan, P. Yang, S. Jiang, Z. Zheng, W. Wu, J. Xia, S. Tang, D. Wu, Y. Cao, J. Xuan, L. Yang, L. Ma and Y. Tian, Chem. – Eur. J., 2023, 29, e202302334.
21. J. Wang, J. Yang, J. Xie and N. Xu, Adv. Mater., 2002, 14, 13–14.
22. C.-H. Tsao, C.-H. Hsu, J.-D. Zhou, C.-W. Chin, P.-L. Kuo and C.-H. Chang, Electrochim. Acta, 2018, 276, 111–117.
23. K. S. Rao, D. D. Pathak, B. P. Mandal, S. Kumar and A. K. Tyagi, Mater. Today Commun., 2023, 36, 106708.
24. X. Wang, Y. Qian, L. Wang, H. Yang, H. Li, Y. Zhao and T. Liu, Adv. Funct. Mater., 2019, 29, 1902929.
25. J. Ma, G. Xu, Y. Li, C. Ge and X. Li, Chem. Commun., 2018, 54, 14093–14096.
26. L. Hu, Y. Lu, X. Li, J. Liang, T. Huang, Y. Zhu and Y. Qian, Small, 2017, 13, 1603533.
27. Y. Zhang, Y. Zhao, Z. Bakenov, M.-R. Babaa, A. Konarov, C. Ding and P. Chen, J. Electrochem. Soc., 2013, 160, A1194–A1198.
28. Y. Yang, G. Zheng and Y. Cui, Chem. Soc. Rev., 2013, 42, 3018.
29. L. Yin, J. Wang, F. Lin, J. Yang and Y. Nuli, Energy Environ. Sci., 2012, 5, 6966–6972.
30. A. Konarov, Z. Bakenov, H. Yashiro, Y.-K. Sun and S.-T. Myung, J. Power Sources, 2017, 355, 140–146.
31. Y. Liu, W. Wang, A. Wang, Z. Jin, H. Zhao and Y. Yang, J. Mater. Chem. A, 2017, 5, 22120–22124.
32. A. Abdul Razzaq, X. Yuan, Y. Chen, J. Hu, Q. Mu, Y. Ma, X. Zhao, L. Miao, J.-H. Ahn, Y. Peng and Z. Deng, J. Mater. Chem. A, 2020, 8, 1298–1306.
33. R. Zou, W. Liu and F. Ran, J. Energy Chem., 2023, 79, 148–157.
34. Y. He, S. Bi, C. Jiang and J. Song, Mater. Today Sustain., 2022, 19, 100181.
35. A. Gabryelczyk, S. Ivanov, A. Bund and G. Lota, J. Energy Storage, 2021, 43, 103226.
36. H. Yang, X. Zhang, J. Guo, W. Zhu, S. Zhao, F. Wang, Q. Fan, H. Xiao and F. Zhang, J. Alloys Compd., 2018, 768, 495–502.
37. C. Li, Y. Chen, Z. Li, Y. Zhang, Z. Fang, J. Xu, Y. Sun, H. Bao and H. Cheng, J. Power Sources, 2021, 495, 229765.
38. H. Jiang, Z. Sun, P. Liu, N. Yao, T. Jin, Q. Wang and L. Jiao, Adv. Mater., 2025, e09953.
39. J. Landesfeind, D. Pritzl and H. A. Gasteiger, J. Electrochem. Soc., 2017, 164, A1773–A1783.
40. A. Mentbayeva, A. Belgibayeva, N. Umirov, Y. Zhang, I. Taniguchi, I. Kurmanbayeva and Z. Bakenov, Electrochim. Acta, 2016, 217, 242–248.
41. Q. Wang, X. Zhang, J. Wang, J. Chu, M. Li, L. Huang, Z. Hu, Y. Wang and Z. Song, Chem. Eng. J., 2023, 471, 144402.
42. S. Wei, L. Ma, K. E. Hendrickson, Z. Tu and L. A. Archer, J. Am. Chem. Soc., 2015, 137, 12143–12152.
43. A. Kumar, V. Kumar, M. Kumar and K. Awasthi, Polym. Compos., 2017, 39, 3858–3868.
44. F. Shahamatifard, D. Rodrigue, K. W. Park, S. Frikha and F. Mighri, Can. J. Chem. Eng., 2023, 101, 1881–1896.
45. A. Neelesh, S. Vidhyashree and B. Meera, J. Appl. Polym. Sci., 2020, 137, e49271.
46. T. Wang, Q. Zhang, J. Zhong, M. Chen, H. Deng, J. Cao, L. Wang, L. Peng, J. Zhu and B. Lu, Adv. Energy Mater., 2021, 11, 2100448.
47. A. Abdul Razzaq, Y. Yao, R. Shah, P. Qi, L. Miao, M. Chen, X. Zhao, Y. Peng and Z. Deng, Energy Storage Mater., 2019, 16, 194–202.
48. J. Xiang, Z. Guo, Z. Yi, Y. Zhang, L. Yuan, Z. Cheng, Y. Shen and Y. Huang, J. Energy Chem., 2020, 49, 161–165.
49. K. A. Wepasnick, B. A. Smith, K. E. Schrote, H. K. Wilson, S. R. Diegelmann and D. H. Fairbrother, Carbon, 2011, 49, 24–36.
50. H. Gong, S.-T. Kim, J. D. Lee and S. Yim, Appl. Surf. Sci., 2013, 266, 219–224.
51. S. Kalybekkyzy, A. Mentbayeva, M. V. Kahraman, Y. Zhang and Z. Bakenov, J. Electrochem. Soc., 2019, 166, A5396–A5402.
52. D. Wang, Q. Cao, B. Jing, X. Wang, T. Huang, P. Zeng, S. Jiang, Q. Zhang and J. Sun, Chem. Eng. J., 2020, 399, 125723.

---

Footnote

† These authors contributed equally to this work.

---