γ-Ray irradiated polyacrylamide networks enable high-performance Li||S pouch cells

Abstract

Binders are essential for maintaining positive electrode integrity in Li||S batteries and significantly affect their performance. However, commercial linear binders often have disordered networks, poor binding efficiency, and insufficient mechanical strength. To address these challenges, three-dimensional covalent binders offer a promising solution. Traditional methods for producing cross-linked binders require additives and result in poorly controlled polymer networks due to the stochastic nature of liquid-phase polymerization. Moreover, the mechanisms by which reticulated binders stabilize the positive electrode remain unclear, requiring investigation under operando conditions. Herein, we present an approach to tailor cross-linked polyacrylamide networks using solid-state operando γ-ray irradiation chemistry, which eliminates additives and produces a pure, ordered network with remarkable binding capabilities. By integrating in situ high-resolution optical frequency domain reflectometry, multiscale synchrotron radiation characterization, and virtual simulations, this study reveals the role of binders in dynamically encaging and confining sulfur. Specifically, γ-ray-enabled polyacrylamide networks enhance battery performance through mechanical strengthening, optimized sulfur regeneration, and improved re-occupancy. Consequently, the well-designed composite positive electrode structure with only 5.0 wt% binder improves soft-packaged Li||S battery performance across various scenarios. Notably, a 1.2-Ah pouch cell achieves 410.1 Wh kg⁻¹ specific energy with a low electrolyte/sulfur ratio of 3.0 µL mg^–1.

Introduction

The gradual depletion of fossil fuels, coupled with increasingly severe environmental challenges, compels humanity to heighten its reliance on renewable energy sources and impose more stringent requirements on energy storage systems^1,2,3. In comparison to lithium-ion batteries, Li||S batteries are regarded as the next generation of high-energy secondary batteries with significant developmental potential^4,5,6,7. This is attributed to the advantages offered by its desirable specific energy (2600 Wh kg^–1) and sulfur, which includes natural abundance, low cost, and relative environmental friendliness^8,9,10.

The binder, an indispensable component of the sulfur electrode in Li||S batteries, plays a crucial and multifaceted role in addressing key challenges^11,12. It not only holds together critical components of the electrode assembly, but also maintains structural integrity during significant volume changes^13,14. Additionally, it effectively mitigates polysulfide shuttling through trapping and enhances the overall conductivity of the sulfur electrode by closely interfacing with conductive fillers^15,16. This role is especially vital for high sulfur-loading positive electrodes, where the rational design and application of the binder are paramount. Consequently, there has been considerable attention given to the structural design and modification of polymer binders for Li||S batteries. The development of polymer binders for Li||S batteries can be categorized into three types based on their features: linear binders, hybrid binders, and cross-linked reticulated three-dimensional (3D) binders.

The predominant binder currently employed in Li||S batteries is the linear type, encompassing a diverse range of linear polymers. Polyvinylidene fluoride (PVDF)^17,18 and LA133^19,20, which are the classic simple linear binders, are chosen for their favorable chemical and thermal stability in Li||S batteries. In recent years, new functional linear binders have emerged, including polysulfide-limiting binders such as sodium alginate²¹, chitosan²², chitosan sulfate ethylamide glycinamide²³, and tragacanth gum¹⁴, as well as Li-ion transport-promoting binders like polyethylene oxide²⁴ and polyvinyl alcohol²⁵. Despite these advancements, the limited resistance to strong tensile deformation exhibited by linear binders impedes their ability to withstand internal stress resulting from repeated volume changes in batteries. Consequently, they fail to provide satisfactory long-term cycling performance in Li||S batteries. The advantages of each component material are combined in simple hybrid binders, enabling them to impart both binding properties and specialized functions to the sulfur electrode^26,27. However, their insufficient mechanical strength poses challenges in maintaining the structural integrity of the sulfur electrode, especially under high sulfur loadings, as with linear binders. Moreover, reconciling the compatibility among complex component materials presents a formidable challenge.

Cross-linked reticulated binders form 3D network structures through covalent bonds, hydrogen bonds, and other molecular interactions, mitigating volume changes in the sulfur electrode. Recently, significant advancements have been made in the design of diverse cross-linked binders, encompassing covalently crosslinked polyacrylamide (PAM)²⁸, dynamic cross-linking zwitterionic supramolecular binder²⁹, self-healing hydrogen bond-crosslinked polyethylene pyrrolidone-polyethylene imine³⁰, and gelatin-boric acid formed through hydrogen bonding and ion coordination³¹. These binders have demonstrated the formation of robust network structures in sulfur electrodes. Notably, systems based on covalent bonds exhibit favorable mechanical strength within the resulting networks, a highly desirable characteristic for Li||S batteries. However, their synthesis processes of these materials unavoidably involve the utilization of intricate chemical additives such as initiators, crosslinking agents, catalysts, and solvents. The presence of these additives or their decomposition byproducts may induce detrimental side reactions during battery operation, potentially compromising battery performance, safety, and longevity. Furthermore, conventional polymerization and crosslinking reactions are susceptible to random molecular movements in solution. Therefore, this leads to the formation of 3D cross-linked binder networks with unpredictable structures and poor control over the networking process. Such randomness may adversely affect Li||S battery performance. Moreover, the mechanism by which reticulated binders impact the positive electrode integrity remains poorly understood and necessitates further investigation. This knowledge gap is reflected in two key areas. Firstly, there is a need for clarification regarding the precise relationship between the mechanical properties of the binder networks and their ability to inhibit volume changes during reactions. Secondly, substantive evidence and deciphering are required regarding the impact of ordered versus disordered binder networks on sulfur species evolution during reactions and subsequent battery performance. By gaining in-depth insights into these issues comprehensively, the crucial role played by binders in stabilizing sulfur electrodes can be better understood, potentially leading to enabling the commercialization of Li||S batteries.

This study presents an ‌advanced‌ technique utilizing high-energy γ-ray irradiation to enable cross-linking and create a chemically pure covalent reticulated PAM binder network for soft-packaged Li||S pouch cells. This approach capitalizes on irradiation-induced radical generation to in situ initiate cross-linking of uniformly distributed acrylamide (AM) monomers in the solid state, resulting in well-organized PAM binder networks. The whole procedure, free of additives, can be easily controlled by adjusting irradiation dosage, aiming to overcome the limitations associated with conventional liquid-phase cross-linking methods. To delve into the function of irradiated PAM (I-PAM) binder in mitigating volume variation, an implanted operando high-resolution optical frequency domain reflectometer is applied to monitor the positive electrode, enabling direct observation and authentic interpretation of how the I-PAM strengthens the sulfur electrode. Further insights are obtained through synchrotron radiation X-ray 3D nano-computed tomography (synchrotron radiation X-ray 3D nano-CT) and synchrotron radiation small-angle X-ray scattering (SAXS), elucidating the working mechanism of I-PAM binder during cycling. Virtual simulations complement these observations by providing a mechanistic understanding of the guiding effect of reticulated I-PAM binder on sulfur evolutions during discharge and charge procedures. The effectiveness of I-PAM binder networks is demonstrated in multi-layer Li||S pouch cells at Ah levels, showcasing its potential for real implementations. This solid-state, in situ irradiation-enabled cross-linking approach represents a significant advancement towards achieving practically viable Li||S batteries.

Results

A high-strength sulfur electrode was successfully fabricated using an in situ synthetic approach enabled by high-energy γ-ray radiation. Sulfur spheres with a diameter of around 50 nm, conductive Super P, and AM monomers were first mixed at a specific mass ratio of 5:4.5:0.5. The resulting slurry was then coated onto an aluminum foil and kept at room temperature for 8.0 h to obtain an un-crosslinked positive electrode. The detailed preparation process is illustrated in Supplementary Fig. 1. The un-crosslinked sulfur electrode underwent planar stationary irradiation using high-energy γ-rays at room temperature to fabricate a robust sulfur electrode. High-energy γ-rays, primarily generated by the radioactive isotope ⁶⁰Co, induced polymerization and cross-linking reactions of AM monomers within the un-crosslinked sulfur electrode, enabling in situ construction of a uniform I-PAM network structure. To manipulate and tailor the I-PAM crosslinking networks within the sulfur electrode, various irradiation doses can be applied by adjusting the γ-ray irradiation time at a fixed irradiation rate, while formation and breakage of I-PAM crosslinking networks are induced by high-energy γ-rays. Insufficient γ-ray dose results in incomplete generation of cross-linked networks, as depicted in Fig. 1a. As the radiation dose increases, gradual development of more complete cross-linked I-PAM networks occurs. However, excessive irradiation doses can lead to radiation degradation and molecular chain fracture within the formed cross-linked network structure, thereby giving rise to failure of the binder’s structural stability. Therefore, determining an optimal irradiation dose is crucial for constructing sulfur electrodes with a favorable structural stability.

Fig. 1: Schematically illustrating the design of I-PAM in the positive electrode.

a The tunable in situ fabrication of I-PAM-based sulfur electrode through γ-ray irradiation. b The crosslinking mechanism of I-PAM. c The structural and functional characteristics of I-PAM, C-PAM, and commercial LA133 binders.

Compared to conventional chemical-additive enabled cross-linking, γ-ray irradiation-induced free radical polymerization has gained increasing attentions due to gentle reaction conditions, and elimination of the need for chemical additives. As illustrated in Fig. 1b, high-energy γ-ray irradiation enables in situ polymerization and covalent crosslinking of AM monomers within the sulfur electrode through a radical-mediated reaction mechanism. In detail, high-energy γ-rays trigger water molecules which stem from the ambiently dried surface of the sulfur electrode to generate hydrogen free radicals (•H) and hydroxyl free radicals (•OH). These radical species act as initiators by abstracting hydrogen atoms from AM monomers, hence producing chain propagation. These activated AM monomers undergo stepwise chain-growth polymerization, leading to the formation of PAM. •H or •OH can be released from the propagating chains, creating metastable PAM macro-radicals. Subsequently, these polymeric radicals participate in bimolecular termination reactions through radical combination, fulfilling covalent bridging between adjacent polymer chains, ultimately yielding a three-dimensional (3D) crosslinked network structure. Electron paramagnetic resonance (EPR) spectroscopy characterization of intermediate products during the irradiation polymerization process in Supplementary Fig. 2 confirms the presence of free radicals in the irradiation polymerization process. The findings indicate that the high-energy γ-ray irradiation induced polymerization and crosslinking reaction of AM monomers follows a free radical triggering mechanism. As γ-ray irradiation continues, continuous generation of free radicals occurs on the formed PAM chains, leading to the development of active free radical chains within PAM. The interaction between free radicals across different molecular chains facilitates the formation of 3D cross-linked binder networks in the positive electrode. Digital photographs presented in Supplementary Fig. 3 demonstrate that as the irradiation dose increases, there is a decrease in mobility and an increase in degree of crosslinking of the resulting I-PAM, while its degree of crosslinking increases. This observation confirms that precise control over the crosslinking degree in the resulting I-PAM.

The advancement of this measure in the preparation of reticulated I-PAM binder becomes evident when compared to PAM binder synthesized from conventional chemical-additive enabled cross-linking (C-PAM) and commercial linear LA133 binder. Figure 1c summarizes the characteristics of different synthesis routes and the resulting qualities of binder structures. The utilization of high-energy γ-ray irradiation provides a mild environment for AM polymerization and crosslinking reactions, eliminating the need for complex chemical additives such as initiators, crosslinkers, catalysts, and solvents. This approach effectively minimizes side reactions caused by these chemical additives and their decomposition products, thereby facilitating subsequent application in soft-packaged Li||S pouch cell systems. During the construction process of the polymer binder network structure in sulfur electrode under high-energy γ-ray irradiation, the un-cross-linked AM monomers, which uniformly disperse with sulfur and conductive agents, are fixed in local positions and undergo in situ uniform polymerization without the addition of any chemical additives. This eliminates the scenario observed during traditional chemical-additive enabled polymerization, which is influenced by the stochastic molecular motion in solution. Such conventional methods, unfortunately, result in a 3D cross-linked PAM network structure that exhibits randomness and lack of regulation and impurity residues, which respectively give rise to poor binder networks as well as side reactions. Furthermore, adjusting the radiation dose of γ-rays allows for strong and flexible regulation of the 3D network structure of I-PAM. This is in sharp contrast to commercial linear LA133, where only poorly structured 2D networks lacking strong and flexible structuring exist.

Fourier transform infrared (FT-IR) spectroscopy provides insights into the structural transformation from AM to crosslinked PAM. As observed in the FT-IR spectrum of AM monomers (Fig. 2a), the absorption bands located at around 3165 and 3339 cm⁻¹ are respectively assigned to the symmetric and asymmetric stretching vibrations of the −NH₂ functional group. The vibrational mode at peak at about 1668 cm⁻¹ belongs to the −C=O stretching vibration. Meanwhile, the band detected at around 1608 cm⁻¹ arises from the −C=C stretching. In addition, the signal near 1422 cm⁻¹ is assigned to the −C−N stretching, the peak at around 1348 cm⁻¹ correlates with the −C−H deformation, and the band near 1276 cm⁻¹ is indexed to the −NH₂ rocking. In comparison, the spectra of I-PAM and C-PAM exhibit significantly broadened peaks with reduced intensities, indicating a substantial decrease in crystallinity following the polymerization and cross-linking processes²⁸. Notably, I-PAM and C-PAM display remarkably similar FT-IR features. This observation is further corroborated by solid-state nuclear magnetic resonance (SSNMR) results, as shown in Supplementary Fig. 4, where both I-PAM and C-PAM exhibit closely aligned critical features, confirming the successful polymerization and crosslinking by γ-ray irradiation.

Fig. 2: Mechanical behavior analysis of binders and the corresponding sulfur electrodes.

a FT-IR spectra of AM, I-PAM, and C-PAM. b Peeling forces at 180° of I-PAM, C-PAM, and LA133 based positive electrodes. c Typical load-indentation depth curves obtained from nanoindentation tests for positive electrodes based on I-PAM, C-PAM, and LA133 binders. d Schematic diagram illustrating in situ fiber testing and structural depiction of Li||S pouch cells. The charge-discharge curves and internal strain curves of sulfur electrodes with e I-PAM, f C-PAM, and g LA133 binders (sulfur loading: 3.5–3.6 mg cm^–2, rate: 0.05 C, testing temperature: 28 ± 0.5°C, SOC denotes state of charge). h Cross-sectional SEM images of sulfur electrodes using different binders before and after 100 cycles (All batteries were in a fully charged state prior to disassembly. S loading: 1.4–1.6 mg cm^–2, rate: 0.2 C, testing temperature: 28 ± 0.5 °C, L denotes left image, R denotes right image). (1 C = 1675 mA g^–1).

The adhesion capabilities of binders play a crucial role in guaranteeing the stability as well as integrity of the positive electrode. Rheological tests, as depicted in Supplementary Fig. 5, provide critical insights into these properties. The I-PAM binder demonstrates inferior mobility compared to C-PAM and LA133, exhibiting higher shear viscosity and stress, as well as a smaller loss factor. These findings indicate stronger intermolecular interactions and enhanced pseudo-solid behavior with stronger binding and adhesion capabilities. Digital photographs of the various binders also provide visual evidence, as shown in Supplementary Fig. 6. Optical images in Supplementary Fig. 7 demonstrate that I-PAM can pledge the smooth and uniform positive electrode textures as other binders. Peeling tests of positive electrodes at 180° angle were conducted (Supplementary Figs. 8 and 9). According to the testing results in Supplementary Fig. 10, the adhesion strength for positive electrodes initially increases with irradiation dose until reaching a peak value of 3.0 kGy before declining due to potential network degradation. This optimal dose corresponds to the remarkable‌ electrochemical cycling performance of Li||S pouch cells, as shown in Supplementary Figs. 11 and 12. Comparative peeling force measurements indicate that the positive electrode, after irradiating at a dose of 3.0 kGy exhibits a higher strength of 5.85 N, the value of which is remarkably higher than that of C-PAM (3.53 N) and LA133 (0.13 N), with average peeling strengths of 2.23, 1.39, and 0.05 N cm^–1, respectively (Fig. 2b and Supplementary Fig. 13). The enhanced adhesion properties of C-PAM and I-PAM are attributed to their robust 3D network structures. Notably, despite sharing a similar reticulated network with C-PAM, I-PAM surprisingly demonstrates more favorable adhesion properties due to its more uniform and precise network control.

The nano-indentation testing results further validate the favorable mechanical properties of I-PAM@S electrode compared to C-PAM@S and LA133@S counterparts. The I-PAM@S electrode shows significantly smaller indentation depths (h_max and h_f) under peak load, indicating larger resistance to deformation. Moreover, it exhibits a higher elastic modulus of 2.6 GPa and hardness of 0.07 GPa compared to C-PAM@S (1.5 GPa and 0.05 GPa) and LA133@S electrodes (1.3 GPa and 0.03 GPa), as displayed in Fig. 2c and Supplementary Fig. 14. Furthermore, the I-PAM binder displays remarkable elasticity and suppressed plastic deformation. These enhanced mechanical properties of I-PAM can be due to its unique structural characteristics. Similar to C-PAM, I-PAM possesses a 3D covalent network structure that provides stronger bond strength and improved interactions between polymeric branches when compared to the poorly structured, loose 2D networks of the LA133 binder based on weak nonpolar intermolecular forces. This elucidates why both I-PAM and C-PAM outperform LA133 in mechanical properties. The advancement of I-PAM over C-PAM in mechanical performance can be attributed to the in situ irradiation-triggered crosslinking process used during synthesis. Such enhancement implies a more uniform, controllable, and reinforced polymer structure. The irradiation process enables more precise manipulation over the degree of crosslinking as well as distribution within the polymer networks, leading to enhanced mechanical robustness and elasticity. These improved mechanical properties of I-PAM are particularly advantageous for Li||S battery application as they enhance robustness and elasticity to restrain volume variation during cycling. This ability to withstand and accommodate volume changes contributes directly to favorable battery performance.

The implanted optical fiber sensing technology presents an innovative and effective approach for operando investigation of positive electrode volume changes in Li||S batteries^32,33. The in situ high-resolution optical frequency domain reflectometer addresses the urgent need for a straightforward, accurate, sensitive, and reliable monitoring technique that can track binder-mediated volume alteration at the positive electrode side. The advantages of this technology include its lightweight nature, high sensitivity, resistance to electromagnetic interference and chemical corrosion, as well as its slender structure, allowing non-destructive implantation into batteries. The experimental setup depicted in Supplementary Fig. 15 involves embedding a single-mode fiber optic sensor (SMF) within the positive electrode, surrounded by a mixture of nano-sulfur particles with a size of around 50 nm, binders, and conductive additives. An additional SMF with a polytetrafluoroethylene (PTFE) tube is positioned nearby to compensate for temperature-induced effects during stress monitoring. This configuration, when assembled into a pouch cell with a Li metal electrode (Supplementary Fig. 15b and Fig. 2d), enables real-time monitoring of internal strain evolution at the positive electrode.

The study utilizes optical frequency domain reflection technology for data acquisition and conversion, enabling the demodulation of volume strain signals from the sulfur electrode. This approach, depicted in Supplementary Fig. 16, provides high spatial resolution, sampling rate, and sensitivity to strain fields, facilitating stable real-time monitoring of internal volume changes at multiple sites within the sulfur electrode. The charge-discharge profiles and internal strain characteristics of positive electrodes with different binders are illustrated in Fig. 2e–g. The results reveal a consistent trend in internal volume strain that corresponds to the charge-discharge cycling periodicity. During discharging, the volume strain gradually increases and reaches its maximum at the end of the process, while it decreases during charging. This pattern aligns with the electrochemical reactions of sulfur species within the battery. Importantly, among all tested binders, I-PAM binder endows the sulfur electrode with the minimal volume strain at the sulfur electrode. This observation suggests that due to its favorable mechanical properties, I-PAM binder effectively resists deformation of the sulfur electrode during battery reactions, thereby maintaining its structural integrity. This finding is consistent with the aforementioned peeling force and nano-indentation testing results, further confirming the remarkable performance of I-PAM binder in managing volume variations and keeping positive electrode stability in Li | |S batteries. The visual post-mortem examination of cycled sulfur electrodes using SEM, compared to pre-cycling images, provides crucial evidence supporting the effectiveness of the I-PAM binder in controlling volume expansion. Figure 2h presents side views of positive electrodes prepared with various binders, where results show that the I-PAM binder causes a minimal volume expansion rate of only 3.2% after 100 cycles, which is significantly lower than C-PAM (8.8%) and LA133 binder (21.0%). The SEM results support the remarkable performance of I-PAM in maintaining structural integrity during cycling, in accordance with the observations from mechanical tests and operando optical sensing detection.

In Li||S batteries, the binder networks and sulfur dispersion in the positive electrode determine the key performance metrics such as active material utilization rate, reaction dynamics, and cycling stability. This dispersion is also heavily influenced by the binder networks, making it a critical indicator of network quality. To precisely visualize the dispersal of sulfur within the polymer network structure, synchrotron radiation X-ray 3D nano-CT was employed. This powerful imaging technique offers non-destructive 3D visualization of internal structures with decent spatial resolution, high intensity, remarkable coherence, and fast data acquisition ability³⁴. The schematic diagram illustrating the imaging procedure and subsequent 3D visualization using synchrotron radiation X-ray 3D nano-CT is presented in Fig. 3a.

Fig. 3: Synchrotron radiation X-ray 3D nano-CT analysis of binder networks in sulfur electrodes.

a The schematic diagram of the synchrotron radiation X-ray 3D nano-CT facility and the 3D visualization process. b Synchrotron radiation X-ray 3D nano-CT images of sulfur electrode materials based on different binders under various rotation angles.

The 3D visualization images of the sulfur electrodes with different binders, based on various rotation angles, are shown in Fig. 3b, Supplementary Figs. 17 and 18. These images demonstrate distinct patterns of dispersion. Due to similar X-ray attenuation coefficients, carbon and the polymer matrix are indistinguishable and consequently represented as a unified phase complex. In contrast, sulfur is clearly discernible, allowing examination of its dispersibility within the composite structure. In the LA133-based sulfur electrode sample, both nano sulfur and Super P/LA133 exhibit significant local aggregation, indicating that the linear LA133 binder cannot homogeneously and effectively fix the sulfur particles. Typically, the C-PAM is first prepared and then mixed with sulfur and Super P to generate the positive electrode slurry, which negatively affects the dispersion of sulfur. As substantiated in Fig. 3b, nano sulfur particles are indeed distributed at a micrometer scale and in a uniform state within the Super P/C-PAM matrix. In stark contrast, I-PAM displays more homogeneous 3D networks that can stably fix and disperse sulfur particles, giving rise to a conspicuously interpenetrating structure formation that favors a robust sulfur electrode. The impact of binders on the Li₂S deposition behaviors is also qualitatively and quantitatively analyzed. The batteries we discharged to 1.7 V to ensure the complete deposition of Li₂S. The positive electrode materials were collected to perform the synchrotron radiation-based tests. As seen in the X-ray 3D nano-CT images in Supplementary Fig. 19, all the positive electrodes show the interpenetrating network structures. Of specific note, the I-PAM-based positive electrode achieves the most homogeneous Li₂S precipitations among the three samples, in accordance with the imaging observations of sulfur in I-PAM architecture. These findings suggest that the irradiation-triggered in situ polymerization and crosslinking confer distinct advantages in producing uniform networks with evenly distributed crosslinks, which more effectively guide the fine distributions of sulfur and Li₂S. Meanwhile, the conventional stochastic crosslinking based on the inducing effect of chemical additives manifests a limited capability to control the positive electrode networks.

Synchrotron radiation SAXS tests were used to investigate the nanoscopic architecture of binders in sulfur electrodes. This toolset utilizes a highly collimated and intense X-ray beam, enabling high angular and spatial resolution to provide accurate information about fine structures^35,36. The overall workflow diagram for SAXS analysis of cycled sulfur electrodes with various binders is illustrated in Fig. 4a, encompassing detection, feature inspection, and data analysis. A linear fitting based on the Guinier-Porod model is conducted to obtain the I(q)–q fitting curves, which are subsequently converted into one-dimensional data. The obtained 2D SAXS images and I(q)-q fitting curves are presented in Fig. 4b–g. From these fitting results, we derive the radii of gyration (R_g) of the samples that represent an ensemble weighted average of all scattering particles and provide size distribution information. Notably, prior to battery cycling, the sulfur electrode sample with I-PAM exhibits a significantly smaller R_g value of 22.03 nm compared to that with C-PAM (28.16 nm) and LA133 binder (24.32 nm). This observation suggests that a more uniform dispersion of positive electrode material with suppressed sulfur particle aggregation can be attributed to the establishment of a uniform and stable 3D I-PAM network structure. Additionally, the inhomogeneous distribution of 3D C-PAM networks resulting from the ex situ formation cannot cause the large aggregation of sulfur particles. As for the LA133 binder, its linear state is favorable for uniformly mixing with sulfur particles. Interestingly, it is found that all the R_g values for the three samples decrease after cycling, substantiating the regeneration and re-occupancy of sulfur particles in binder networks. Moreover, it can be confirmed that such reconfiguration effect closely correlates with the network structures of binders. Of specific note, the smallest R_g value of 18.79 nm is attained by the cycled I-PAM-based sample, manifesting that the in situ produced I-PAM networks effectively guide the dynamic re-occupancy of regenerated sulfur and maintain a favorable composite positive electrode structure. The cycled C-PAM-based sample achieves a smaller R_g value of 22.10 nm compared to the cycled LA133-based sample (23.49 nm), proving the significant effect of binder network structure on sulfur reconfiguration rather than distribution state in the positive electrode.

Fig. 4: Synchrotron radiation SAXS analysis of sulfur and Super P distributions in the positive electrode before and after cycling.

a The overall workflow diagram showing the detection process, feature, and data analysis of SAXS for the cycled sulfur electrodes based on various binders. Fitting curves of I(q)–q scattering signals of sulfur electrodes with b I-PAM, c C-PAM, and d LA133 before cycling, with the insets showing the SAXS matrices of scattering signals. Fitting curves of I(q)-q scattering signals of sulfur electrodes with e I-PAM, f C-PAM, and g LA133 after cycling, with the insets displaying the SAXS matrices of scattering signals. (R_g denotes Radius of Gyration). h Schematic representing the distributions of sulfur and Super P in sulfur electrodes with various binders before and after cycling operation (L denotes left image, R denotes right image).

Figure 4h schematically illustrates these mechanism findings and underscores how tunable crosslinking networks enabled by high-energy γ-ray irradiation contribute to the dynamic evolution of sulfur particles during cycling. The sulfur electrodes upon discharge to 1.7 V are also investigated to the aggregation and distribution states of Li₂S. According to the SAXS images in Supplementary Fig. 20, all the R_g values of sulfur electrodes increase to varying degrees after discharge owing to the sulfur dissolution and Li₂S nucleation onto the surface of conductive Super P. As a contrast, the I-PAM-based electrode exhibits the smallest R_g value among the samples, substantiating the more uniform dispersion of Li₂S in the I-PAM structure.

The stress and strain levels of binder materials were further detected, and their distributions in sulfur electrodes were mapped through virtual simulations. Synchrotron radiation X-ray phase contrast tomography data were used to reconstruct and simulate the sulfur electrodes with different binders. The stress and strain field on the cross section of the sulfur electrodes during charge and discharge are illustrated in Fig. 5a and Supplementary Fig. 21. The results indicate that the stress distribution on the surface of the I-PAM binder is relatively uniform during charge–discharge process, without apparent stress concentration compared to C-PAM and LA133 binders. In addition, the surface strain of I-PAM binder also exhibits favorable uniformity, with a smaller strain magnitude than C-PAM and LA133 binders. This can be mainly attributed to the adhesive network structure formed by the I-PAM binder, which disperses sulfur in a finer and more uniform state. As a result, it ensures that stress generated during sulfur deformation is more evenly distributed and reduces areas of high stress concentration. Furthermore, simulation results demonstrate that the average strain on the surface of I-PAM is −0.16, significantly smaller than C-PAM (−0.36) and LA133 (−0.64). After battery cycling, interfacial gap sizes formed inside the positive electrode due to stress and strain effects are respectively 0.04 μm for I-PAM, 0.08 μm for C-PAM, and 0.16 μm for LA133 (Fig. 5b, c, Supplementary Figs. 22 and 23). These findings substantiate that compared to C-PAM and LA133, the I-PAM binder effectively resists deformation while maintaining composite positive electrode structure integrity under battery operating conditions. The primary factor contributing to this phenomenon is due to the better elastic modulus and stronger adhesive strength presented by I-PAM, which collectively affect the stability of the composite positive electrode structure. The increased elastic modulus possessed by the binder facilitates more effective mitigation of damage caused by sulfur volume expansion within the adhesive network structure. Simultaneously, a stronger adhesive strength between the binder and sulfur restricts interface gaps resulting from sulfur volume contraction during the charge process. These two factors collectively guarantee the structural integrity of the sulfur electrode throughout the cell cycling process, as illustrated in Fig. 5d–f.

Fig. 5: Virtual simulations deciphering the stress and strain evolution in sulfur electrodes with various binders during the discharge–charge procedure.

a The stress and strain distribution mapping on different binders in the positive electrode during battery reaction process (for the stress map, a darker shade of blue indicates higher stress levels. For the strain diagram, a deeper red color in the color bar corresponds to greater strain values). b Average volume strain during discharge and c average hollow gap during charge of various sulfur electrodes with different binders. Schematic of the mechanism by d I-PAM, e C-PAM, and f LA133 inhibiting the volume expansion effect of sulfur electrodes (the sphere in these charts represents a simplified illustration of a carbon/sulfur composite material). g SEM surface images of sulfur electrodes with various binders before and after 100 cycles (All batteries were in a fully charged state prior to disassembly. S loading: 1.4–1.6 mg cm^–2, rate: 0.2 C, testing temperature: 28 ± 0.5°C, L denotes left image, R denotes right image).

The scanning electron microscopy (SEM) images of positive electrodes with various binders, captured before and after cell cycling, reveal distinct alterations in surface morphology. Prior to cycling, these positive electrodes exhibit relatively smooth surfaces, as depicted in Fig. 5g. However, significant disparities become apparent after 100 cycles. Sulfur electrodes utilizing the I-PAM binder maintain a consistently even surface without notable holes or cracks, thereby preserving a relatively intact electrode structure. In contrast, sulfur electrode with C-PAM binder displays noticeable perforations, while those with the LA133 binder exhibit not only large holes but also substantial cracks, indicating severe damage to their electrode structure. These visualized evidences underscore the critical role of the I-PAM binder in maintaining the structural integrity of the positive electrode. The ability of I-PAM to prevent surface degradation and structural compromise highlights its effectiveness in enhancing both durability and performance of sulfur-positive electrodes during the long-term cycling procedure.

Linear sweep voltammetry (LSV), electrochemical impedance spectroscopy (EIS), and shuttle current characterizations were conducted on Li||S batteries with different binders to assess the influence of I-PAM networks on battery electrochemistry. In Supplementary Fig. 24a, the LSV curves of batteries with I-PAM reveal that both reduction peaks (i and ii) shift towards the lower potentials in contrast with the control samples, implying the optimized sulfur conversion reaction kinetics by the I-PAM networks. The Tafel slopes were attained from the LSV curves to further identify the impact of I-PAM binder networks on sulfur evolution kinetics during cycling. In the first conversion step of S₈ to Li₂S₄, the Tafel slope value of S/I-PAM is 47.9 mV dec⁻¹, which is lower than those of S/C-PAM (58.1 mV dec⁻¹), S/LA133 (102.2 mV dec⁻¹) (Supplementary Fig. 24b). In the second conversion step of Li₂S₄ to Li₂S₂/Li₂S, the lower Tafel slope value of 53.7 mV dec⁻¹ is obtained by the S/I-PAM in contrast with that of S/C-PAM (61.1 mV dec⁻¹), S/LA133 (88.2 mV dec⁻¹) (Supplementary Fig. 24c). These results confirm that the optimized I-PAM networks within the positive electrode modulate the electron conductivity and sulfur distribution, facilitating the interconversions of sulfur species. In addition, EIS profiles reveal that the fresh S/I-PAM based battery present lower charge transfer resistance value of 33.3 Ω compared with the other two samples owing to the homogeneous and stable fixation of sulfur and Super P in I-PAM binder networks, as shown in Supplementary Fig. 25. The inhibitory effect of I-PAM binder on the polysulfide shuttle, real-time shuttle current tests were performed on the batteries assembled with electrolytes without LiNO₃. The shuttle current curves are shown in Supplementary Fig. 26. It is evident that the I-PAM and C-PAM binders display a lower shuttle current than the LA133 binder, which can be attributed to the presence of ample polar amide groups in both I-PAM and C-PAM, which facilitates the shuttle effect alleviation in practically operated Li||S pouch cells.

We conducted a series of electrochemical performance tests on soft-packaged Li||S pouch cells with I-PAM, C-PAM, and LA133 binders. As shown in Fig. 6a and Supplementary Figs. 27 and 28, the S/I-PAM electrode harvests a higher initial capacity of 1157.8 mA h g⁻¹ and a remarkable capacity retention of 86.2% at 0.2 C after 100 cycles compared to S/C-PAM (999.9 mA h g⁻¹ and 80.0% capacity retention) and S/LA133 (905.8 mA h g⁻¹ and 61.6% capacity retention). The first discharge–charge profiles of the positive electrodes at 0.2 C are displayed in Fig. 6b. It can be observed that the I-PAM binder endows the pouch cell with prolonged low-voltage plateaus than the other two binders, substantiating its remarkable ability to increase sulfur utilization efficiency. The achieved capacity and operation stability of the I-PAM-based pouch cell at 0.2 C surpass those of previously reported Li||S batteries (Supplementary Table 1). The rate capacities of the pouch cells with various binders at varying rates from 0.2 to 2 C (1 C = 1675 mA g⁻¹) are exhibited in Fig. 6c and Supplementary Figs. 29 and 30. The pouch cell based on sulfur/I-PAM positive electrode delivers capacities of 1121.1, 912.6, 825.1, and 726.9 mA h g⁻¹ at 0.2, 0.5, 1, and 2 C, respectively, the values of which are significantly higher than those exhibited by other samples under the same testing rates. Notably, when the rate shifts back to 0.2 C, the capacity of the I-PAM enabled pouch cell is up to 983.2 mA h g⁻¹, implying its more favorable capacity recovery capability. The long-term cycling performance of the Li||S pouch cell with I-PAM at an elevated rate of 1 C was evaluated. In Fig. 6d, the cell with I-PAM-based positive electrode displays a remarkable initial capacity of 800.7 mA h g⁻¹ at 1 C and maintains substantial stability with a capacity retention of 74.4% after 200 cycles.

Fig. 6: Electrochemical performance of soft-packaged Li||S pouch cells.

a Cycling performance at 0.2 C. b Typical charge-discharge curves at 0.2 C. c Rate capabilities. d Cycling performance at 1 C. e Cycling stability of cells with I-PAM binder at a sulfur mass loading of 6.4 mg cm⁻² at 0.1 C. f Cyclic performance and g charge–discharge curves of the cell with I-PAM binder under bending 90° at 0.2 C. h Cycling performance of 0.21-Ah cell. i Digital photographs of a powered electric toy car by our 0.21-Ah Li||S pouch cell with I-PAM binder. j Discharge curve of a 1.0 g-sulfur pouch cell providing high cell gravimetric specific energy of 410.1 Wh kg⁻¹. k Cycling operation of a 1.0 g-sulfur pouch cell with I-PAM binder. (1 C = 1675 mA g^–1).

Practical scenarios were designed to evaluate the electrochemical performance of pouch cells under possible harsh conditions, including high sulfur mass loadings, high/low temperatures, and flexible powering. When the sulfur mass loading reaches 6.4 mg cm⁻² with an electrolyte dosage of 5.0 μL mg_S⁻¹, the pouch cell incorporating I-PAM binder shows a remarkable initial capacity of 853.9 mA h g⁻¹ at 0.1 C, and demonstrates remarkable capacity retention of 93.4% after 50 cycles, indicating stable cycling operation. The corresponding charge–discharge curves based on different cycling numbers are also displayed in Supplementary Fig. 31. The areal capacities of the pouch cells with I-PAM, C-PAM, and LA133 at 0.1 C are presented in Supplementary Fig. 32, where the I-PAM more effectively boosts the cycling operational performance of high-sulfur-load pouch cells than the other two binders. Increasing the sulfur content in the positive electrode to 70 wt% and 80 wt%, high capacities of approximately 943.7 and 831.7 mA h g⁻¹ for up to 100 cycles are respectively maintained (Supplementary Figs. 33 and 34). Even under extreme temperatures such as 278 and 328 K, the Li||S pouch cell respectively displays impressive capacities of 812.5 and 845.6 mA h g⁻¹ at 0.2 C, along with favorable cycling stability over 50 cycles (Supplementary Fig. 35), corroborating the promising implementation potential for I-PAM binder enabled pouch cells across a wide temperature range. Additionally, we investigated the electrochemical performance of the I-PAM-enabled pouch cell at a bending angle of 90° (Fig. 6f). The results show that the cell delivers a remarkable initial capacity of 1026.3 mA h g⁻¹ at 0.2 C and presents a negligible capacity decay of only 0.13% per cycle over 100 cycles. The charge-discharge curves of the cell with the cycle number ranging from 1st to 100th are displayed in Fig. 6g, where a narrow curve distribution can be clearly observed. SEM examinations were conducted on the sulfur electrodes from disassembled cells. The results indicate the absence of significant cracks or structural collapses on the positive electrode surface, both in the bending and smooth regions. This observation implies that even under bending condition, the integrity of the sulfur electrode can be well kept while achieving stable battery operation and retaining a high specific capacity, as depicted in Supplementary Fig. 36. The cell is also capable of powering a light-emitting diode (LED), further highlighting the potential integration of I-PAM based cells in flexible and wearable devices, as shown in Supplementary Fig. 37.

The effectiveness of the I-PAM binder was further confirmed in high-sulfur, as well as lean electrolyte operation, by assembling Ah-level Li||S pouch cells. The interdigitated and winded pouch cell configuration is illustrated in Supplementary Fig. 38. Initially, a pouch cell with 0.24 g sulfur loading was fabiricated and tested at a specific current of 83.8 mA g⁻¹, and the weight details are revealed in Supplementary Fig. 39. This pouch cell exhibits an initial capacity of 0.21 Ah, and achieves a cell gravimetric specific energy of 354.9 Wh kg⁻¹, along with stable cycling operation over 80 cycles (Fig. 6h and Supplementary Fig. 40). Furthermore, the I-PAM enabled pouch cell succefully powers an electric toy car which reaches approximately 1.0 km mileage within 80 min (Fig. 6i), thus demonstrating its suitability as apower source for electronic devices. We further elevated the sulfur mass loading to 1.0 g and lowered the electrolyte dosage to 3 μL mg_S⁻¹. The resulting pouch cell holds a high cell gravimetric specific energy of 410.1 Wh kg⁻¹, as shown in Fig. 6j, k. The detailed information is also listed in Supplementary Fig. 41. These findings confirm the feasibility of optimizing pouch cells through the I-PAM binder strategy. In terms of cell energy, our soft-packaged Li||S pouch cell incorporating I-PAM binder stands for a significant advancement in contrast to previously reported Li||S pouch cells (Supplementary Fig. 42a). Further, in contrast to the recently reported results, our optimized sulfur electrode with I-PAM can provide the pouch cell with overwhelming electrochemical performance even with reduced binder usage (5 wt.%) and lean electrolyte dosage (3 μL mg_S^–1) (Supplementary Fig. 42b). By continuously tuning technique parameters during cell fabrication, and incorporating active catalysts or electrolyte electrolytes, we expect further enhancement of performance in our multi-layer Li||S pouch cells.

Discussion

In summary, we propose a versatile approach for fabricating PAM crosslinked networks in sulfur electrode through in situ irradiation of high-energy γ-rays. This method effectively overcomes the limitations of traditional chemical additive-crosslinked binder networks, and remarkably enhances the controllability, cleanliness, homogeneity, and stability of the PAM binder. Of specific note, a relatively low PAM content of 5 wt% in positive electrode realizes the effective stress manipulation and efficient deformation resistance, and guides the rational sulfur regeneration as well as re-occupancy in the 3D binder networks along with the cycling procedure according to the combined verification results of synchrotron radiation X-ray 3D nano-CT, SAXS and virtual simulations. These delicate structural manipulations enable the sulfur electrode with I-PAM to mitigate the volume change effect and maintain the structural integrity. As a result, the assembled Li||S pouch cell demonstrates favorable discharge capacities and cycling stability even under harsh conditions such as folding and high/low temperatures. Impressively, a 1.2 A-h-level Li||S pouch cell harvests a distinguished cell-specific energy of 410.1 Wh kg⁻¹ using a high sulfur mass (1.0 g) and lean electrolyte dosage (3.0 µL mg_S^–1). This work presents valuable academic insights into the advanced fabrication of durable and high-energy Li||S pouch cells.

Methods

Preparation of sulfur electrode with I-PAM binder

To prepare the sulfur electrode, 1.0 g of nano-sulfur powder (particle size: 50 nm, purity: 99.9%, Beijing Deke Daojin Science and Technology Co., Ltd.), 0.9 g of Super P Li carbon black (purity: ≥99.5%, particle size: 40–50 nm, electrical conductivity: 10–15 S/cm, Guangdong Canrd New Energy Technology Co., Ltd.), and 0.1 g of AM monomers (purity: 99.9%, Macklin) were dispersed in 1.9 mL of deionized water and stirred using a twin-shaft mixer (THINKY SR-500, Shin-Ki Co., Ltd. Japan) for 30 min. The obtained slurry was rapidly cast onto a piece of aluminum foil (thickness: 12 μm, lateral dimension: 200 mm, Shenzhen Kejingstar Technology Co., Ltd.) by using an automatic coating machine (SR-500, Thinky Corporation, Japan), and allowed to naturally evaporate the water in ambient air. The γ-ray irradiation experiments were conducted at the ⁶⁰Co irradiation facility of the Institute of Nuclear Physics and Chemistry, China Academy of Engineering Physics (CAEP). The un-crosslinked sulfur electrodes were subjected to ⁶⁰Co γ-ray radiation field in ambient atmospheric conditions at a dose rate of 10 Gy min^–1. The total doses for γ-ray irradiation experiments were set to 2–5 kGy to induce the controllable crosslinking in the sulfur electrodes. An alanine dosimeter was utilized to precisely measure the total irradiation doses. Finally, the resultant sulfur electrode was cut into squares (3 × 3 cm) by using a cutter, and thoroughly dried in a vacuum oven at 60 °C for 12 h prior to subsequent assembly into pouch cells. During the synthesis process, the environmental temperature was strictly maintained within the range of 25 ± 2 °C. The prepared sulfur electrode material was subsequently stored in an argon-filled glove box (H₂O < 0.01 ppm, O₂ < 0.01 ppm) to ensure its stability.

Synthesis of C-PAM hydrogel

The synthesis of C-PAM hydrogel was conducted by following the previous report²⁸. 5 g of AM monomers were dissolved in 94 mL of deionized water under magnetic stirring, followed by the addition of ammonium persulfate (APS, purity: 99.99%, Macklin) as the free radical initiator for AM polymerization. Upon the formation of a viscous PAM, N,N’-methylenebis (acrylamide) (MBAA, purity: 99%, Aladdin) was introduced as a cross-linker at a usage of 0.005 g while maintaining vigorous stirring. After degassing the solution in a vacuum oven for 5 min, 25 μL of N,N,N’,N’-tetramethylethylenediamine (TEMED, purity: 99%, Aladdin) was added as a cross-linking enabler. It was observed that the initially viscous polymer gradually lost its mobility and ultimately converted into a stretchable C-PAM hydrogel.

Fabrication of sulfur electrode with C-PAM and LA133 binders

Sulfur electrode incorporating with a C-PAM binder was prepared using a widely recognized methodology that involves casting a slurry composed of commercial nano-sulfur powders (particle size: 50 nm, purity: 99.9%, Beijing Deke Daojin Science and Technology Co., Ltd.), Super P Li carbon black (purity: ≥99.5%, particle size: 40–50 nm, electrical conductivity: 10 ~ 15 S/cm, Guangdong Canrd New Energy Technology Co., Ltd.) and C-PAM binder in a mass ratio of 5:4.5:0.5 onto an aluminum foil by using an automatic coating machine (SR-500, Thinky Corporation, Japan). The resulting laminates were subsequently cut into squares by using a cutter and dried at 60 °C under vacuum for 12 h prior to further use. Sulfur electrode utilizing the LA133 (solid content: 15 ± 0.2%, molecular weight: 4×10⁵–5×10⁵g mol^–1; Guangdong Canrd New Energy Technology Co., Ltd.) binder was also prepared as reference following an analogous procedure.

Assembly of pouch cells implanted in an optical frequency domain reflectometer

Positive electrode slurry comprises nano-sulfur powder, Super P, and I-PAM binder at a mass ratio of 5:4.5:0.5. Following this, the positive electrode was dried in a 60 °C vacuum oven for 12 h. The pouch cells were assembled by using a positive electrode and a copper/lithium electrode prior to the test. The sensing signals of optical fiber demonstrate inherent susceptibility to dual interference from thermal fluctuations and positive electrode material volume variations. Building upon this fundamental understanding, our experimental design strategically applied an embedded fiber-optic sensor to establish real-time thermal compensation, hence enabling precise decoupling of thermally induced strain influences from the electrochemical-mechanical response characteristics. Therefore, the T-fiber and Ɛ-fiber, which are respectively utilized for thermal compensation and monitoring the volume variations of positive electrode materials, were positioned at a standoff distance. In this context, a polytetrafluoroethylene (PTFE) sleeve was applied to protect the T-fiber from the interference caused by positive electrode material volume variations during temperature alteration.

Material characterizations

The morphologies of sulfur electrodes were examined by Scanning Electron Microscope (SEM, ZEISS, Sigma 500). Structural analysis was carried out via Fourier Transform Infrared Spectroscopy (FTIR, Bruker Optics, TENSOR 27) and Solid State Nuclear Magnetic Resonance (SSNMR, Bruker BioSpin GmbH, Bruker 400 M). Electron paramagnetic resonance (EPR) was employed to track and detect the presence of free radicals generated during γ-ray-induced polymerization and crosslinking processes. UV–vis spectra were measured on a UV–vis spectrometer (Shimadzu, UV1900). The rheological properties of binders, including viscosity, shear stress, shear strain, storage modulus, and loss modulus, were measured on a rotational rheometer (TA, HR10).

180° peeling force tests

The adhesive properties of the binder in sulfur electrodes were characterized through a 180° peeling force test. Initially, the positive electrode with a sulfur mass loading of 1.4–1.6 mg cm^–2 was cut into elongated strips with a size of 3 cm × 7 cm, and the aluminum foil side of the positive electrode was fixed to a stainless-steel plate slide using 3 M double-sided tape of the same dimensions. Following this, a uniform application of peeling strength test tape was applied to one side of the positive electrode material, starting from the top end. After securing both the steel plate and peeling strength test tape onto a testing fixture, the test tape was pulled at a constant speed of 100 mm min⁻¹. Gradually, a layer of the positive electrode material was peeled off and adhered to the test tape. The corresponding peeling force was recorded as an aggregate measure of interactions between binder and other components in the positive electrode, reflecting its anti-deformation capability.

Nano-indentation tests

The nano-indentation tests were performed using a continuous stiffness measurement indentation method on a Keysight UTM150 nanoindentation system, equipped with a Berkovich indenter (a three-sided pyramidal tip featuring a radius of 150 nm and a total included angle of 142.3°), to evaluate the mechanical properties of the binders.

In situ optical frequency domain reflectometer monitoring

The high-resolution optical frequency domain reflectometry technique is founded on the principle of heterodyne interferometry, wherein a laser emits linearly chirped light to maintain a constant frequency difference between the echo signal at a fixed distance and the reference optical path. This system facilitates the comprehensive acquisition of the Rayleigh scattering spectrum along the entire fiber link. By performing a Fourier transform on the acquired time-domain signals, it generates distance-domain signals for precise localization. The battery was laid aside under a constant temperature for a specific duration, after which a set of fiber optic sensing signals was collected as reference group data. During the battery cycling process, real-time sampling occurred at 1-min intervals to synchronously collect fiber optic sensing signals, which served as sensing group data. The Fourier transform was applied to both reference and sensing group signals to derive distance-domain data. Distance-domain signals within identical range were extracted and subjected to inverse Fourier transform to attain local Rayleigh scattering spectra. Cross-correlation between the spectra of both groups determined wavelength shifts at each position, enabling calculation of wavelength shifts along the entire tested fiber optic cable. By integrating this with the strain sensitivity coefficient of the tested cable, distribution curves for distance-wavelength shift/strain were generated. Furthermore, by selecting positions coated with negative electrode material and demodulating signals collected at various time points, it became possible to ascertain the temporal evolution law of strain. By correlating these findings with cell cycling performance data, insights into the volume variation effect during battery cell operation were analyzed.

Synchrotron radiation X-ray 3D nano-CT tests of pristine sulfur electrodes

Synchrotron radiation X-ray 3D nano-CT was conducted at the beamline 4W1A of Beijing Synchrotron Radiation Facility (BSRF), Beijing, China. The samples for the imaging measurements were obtained from positive electrodes disassembled from pouch cells after working for 100 cycles at 0.2 C (S loading: 1.4-1.6 mg cm^–2). The sulfur electrode underwent repeated immersion in 1,2-Dimethoxyethane (DME, purity: ≥99.5%, Aladdin) to thoroughly eliminate any traces of polysulfides. Subsequently, the treated sulfur electrode was placed inside a glovebox for a minimum duration of 24 h. Following this, the electrode material was meticulously detached from the current collector. Samples with a size lower than 15 µm were selected by virtue of an optical microscope, and subsequently transferred onto a pin tip with the assistance of a pre-collimation toolset. The sample-loaded pin tip was put into the vacuum chamber of the synchrotron radiation X-ray imaging system. To achieve 2D computed tomographic images, an exposure time of 15 s was applied. The rotation was performed from −70° to +70° with steps of 0.5° per increment. Ultimately, the obtained 2D images were reconstructed and 3D visualized based on Avizo Fire VSG software (Visualization Sciences Group, Bordeaux, France). Due to nearly identical X-ray attenuation coefficients, Super P and binders were indistinguishable and consequently visualized as a unified phase complex in the images. Sulfur shows a different attenuation coefficient in contrast with Super P and binders, enabling phase segmentation based on threshold. The final images allow a clear identification of the dispersal states of sulfur particles within the Super P/binder networks.

Synchrotron radiation X-ray 3D nano-CT tests of discharged sulfur electrodes

The synchrotron radiation X-ray 3D nano-CT tests of discharged sulfur electrodes were performed on the BL07W beamline at the National Synchrotron Radiation Laboratory (NSRL) in Hefei, China. This equipment featured a photon flux of 2 × 10¹⁰ Phs s⁻¹ and a spatial resolution of 30 nm. The test samples were obtained from batteries that had been disassembled after being discharged to 1.7 V. During the experiment, the samples were maintained under a pure argon atmosphere and mounted on a carbon-free Formvar film (100 mesh). Subsequently, the Formvar films with the samples were transferred to a vacuum chamber for the tests. Within the tilt angle range of −60° to 60° and with a step size of 1°, 2D tomographic images were acquired with an exposure time of 2 s for each image. 3D visualized images were achieved from the collected 2D tomographic data using the X Radia XMR reconstructor software. Finally, the reconstructed images underwent further 3D visualization and segmentation analysis using Avizo Fire VSG software (Visualization Sciences Group, Bordeaux).

Synchrotron radiation SAXS tests

Synchrotron radiation SAXS measurements were conducted at beamline BL16B1 of the Shanghai Synchrotron Radiation Facility (SSRF) in Shanghai, China. The X-ray radiation wavelength was set to 0.124 nm. Scattered X-ray intensities were recorded using a Mar165 CCD detector with a resolution of 2048 × 2048 pixels and a pixel size of 80 μm, while the exposure time was set to 0.1 s. The sample-to-detector distance was calibrated to be 4710 mm. Two-dimensional scattering images were transformed into one-dimensional SAXS (I(q) vs q) curves using Fit2D software from the European Synchrotron Radiation Laboratory, with background scattering corrected by subtracting air contribution. The resulting one-dimensional scattering profile varied with q = 4π(sin θ)/λ, where q represents the modulus of the scattering vector, 2θ denotes the scattering angle, and λ stands for the X-ray wavelength. The Guinier–Porod³⁴ model equation for data analysis was defined as: ${{\mbox{I}}}\left({{\bf{q}}}\right)=\frac{{{\mbox{G}}}}{{{{\bf{q}}}}^{{{\mbox{s}}}}}\exp \left(\frac{-{{{\bf{q}}}}^{2}{{{\mbox{R}}}}_{{{\mbox{g}}}}^{2}}{3-{{\mbox{s}}}}\right)\,{\mbox{for}}\,{{\bf{q}}}\,\le \,{{\bf{q}}}{{\bf{1}}}$; ${{\mbox{I}}}\left({{\bf{q}}}\right)=\frac{{{\mbox{D}}}}{{{{\bf{q}}}}^{{{\mbox{d}}}}}\,{\mbox{for}}\,{{\bf{q}}}\,\ge \,{{\bf{q}}}{{\bf{1}}}$, where I(q) represent the scattered intensity as a function of q; q denotes the scattering vector; R_g is the radius of gyration; s is the form factor; d is the fractal dimension-related parameter; G and D are the scale factors for the Guinier and Porod regions, respectively.

Virtual simulations

To detect the influence of Young’s modulus and interfacial adhesion of binders on the structural stability of positive electrodes, we conducted a series of simulations to model the stress/strain evolution of sulfur electrodes during discharge and charge procedures. Utilizing synchrotron radiation X-ray phase contrast tomography data obtained from 3D nano-CT characterizations, we reconstructed and simulated the sulfur electrodes with various binders. To achieve an optimal balance between computational cost and structural representativeness, three cross-sections were selected as representative surface elements for each sulfur electrode. The software Avizo was employed to convert the 3D tomography data into standard tessellation language (STL) format, which could then be imported into COMSOL for selecting the cross-section surfaces. We implemented the solid mechanics module in COMSOL Multiphysics (Version 5.5) The built-in stationary solver was used to solve the weak form of the displacement field. Material properties such as the Young’s modulus and interfacial adhesion of different binders were derived from prior nano-indentation and peeling force testing results. Strain/stress responses were calculated under the assumption that sulfur experienced constant strain during expansion and contraction. To characterize the interfacial debonding between the sulfur and binder, a zero-thickness spring layer was prescribed at their interface. The stiffness of the spring represented the interfacial adhesion of binder.

Shuttle current measurements

During the testing of shuttle current, an electrolyte devoid of LiNO₃ additives was selected to prevent passivation of the lithium electrode surface. Typically, cells were galvanostatically charged to 2.8 V under a specific current of 0.2 C (1 C = 1675 mA g^–1) until they reached this voltage and were then held in this state for 1000 s to ensure stability. Subsequently, to maximize the shuttle current, the cells were discharged to 2.38 V under the same specific current and maintained in this condition for another 1000 s to achieve stable voltage conditions. Afterwards, they were switched to potentiostatic mode, and the i–t curves were recorded until the potentiostatic current reached a steady-state value, which could be interpreted as shuttle current.

Electrochemical evaluations

All electrochemical tests were performed on Li||S pouch cells at a testing temperature of 28 ± 0.5 °C. The sulfur electrodes, as-prepared with various binders (with a sulfur loading of 1.4–1.6 mg cm^–2), were cut into appropriate sizes. A fresh lithium foil (with a thickness of 50 μm, Tianjin Zhongneng Lithium Industry Co., Ltd.) was then affixed to the Cu foil (with a thickness of 6 μm, Hefei KJ Materials Technology Co., Ltd.), serving as the negative electrode. The sulfur electrodes and lithium foil were stacked up and down, separated by a PP separator (thickness: 16 μm, lateral dimension: 88 mm, porosity: 38–40%, average pore size: 0.064 μm, Guangdong Canrd New Energy Technology Co., Ltd.). The aluminum-plastic film was utilized as the shell material for the soft-package battery, while the tab film was securely affixed to the heat-sealing position of the tab in order to ensure electrolyte containment and prevent leakage. The electrolyte (1.0 mol L^–1 LiTFSI and 2 wt% LiNO₃ in a mixture of DME/DOL with a volume ratio of 1:1, Guangdong Canrd New Energy Technology Co., Ltd.) is evenly added to both sides of the separator using a pipette at a fixed electrolyte/sulfur ratio (9 μL mg^–1 for low sulfur loading, 5 μL mg^–1 for high sulfur loading). After undergoing vacuum degassing at a vacuum level of –54 kPa and high-temperature packaging using a vacuum sealing machine (MSK-115A, Hefei Kejing Materials Technology Co., Ltd.), followed by a 12-h resting period, the pouch cell was prepared for testing (with an external pressure of approximately 2.94 kN applied during cycling). All of these procedures were conducted within an argon-filled glove box (H₂O < 0.01 ppm, O₂ < 0.01 ppm) (Universal Series, Shanghai Mikrouna Mech. Tech. Co., Limited.). It is important to note that the transportation of the lithium electrode and the Li||S electrolyte must be conducted in an argon-protected environment while ensuring the temperature ≤24 °C, and the materials should be stored in a glove box (H₂O < 0.01 ppm, O₂ < 0.01 ppm) to maintain their stability. Galvanostatic discharge/charge behaviors, rate capacities, and cycling performances of batteries were measured by using a Neware CT4008 battery testing system with a voltage window of 1.7–2.8 V. Linear scanning voltammetry (LSV) and electrochemical impedance spectroscopy (EIS) curves were recorded on a Multi Autolab M204 at a temperature of 28 ± 0.5 °C. The standard EIS test was performed under potentiostatic conditions, with a frequency range of 0.01–100,000 Hz, an amplitude of 0.005 V, an open-circuit voltage stabilization time of 60 s, and 10 points per frequency decade.