A novel aqueous aspartic acid modified biomass binder for high-performance Li–S batteries

Qian Wang† ^a, Shasha Liu†^a, Feifan Liang^b, Ruiqi Wang^a, Xiaoqiang Cai^b, Ya-Xiong Wang*^b and Xingxing Gu*^a
aChongqing Key Laboratory of Environmental Catalysis, College of Environment and Resources, Chongqing Technology and Business University, Chongqing 400067, China. E-mail: x.gu@ctbu.edu.cn
^bSchool of Mechanical Engineering and Automation, Fuzhou University, Fuzhou 350108, China. E-mail: yxwang@fzu.edu.cn

First published on 30th July 2025

---

---

1 Introduction

The lithium–sulfur battery (Li–S) is regarded as a highly promising energy storage device to supplant conventional lithium ion batteries with limited theoretical capacity because of its exceptionally high theoretical energy density (2600 W h kg⁻¹) and theoretical specific capacity (1675 mA h g⁻¹), as well as the advantages of the abundance of positive electrode materials and environmental friendliness.^1,2 However, the substantial volume alteration of the sulfur cathode, the shuttle impact of soluble active species (LiPSs), and the sluggish redox kinetics critically impair the performance and stability of Li–S batteries, significantly obstructing their practical advancement.^3,4

In recent years, researchers have been working to solve these problems, proposing a variety of solutions, including the use of porous conductive materials as sulfur positive carriers⁵ or membrane functionalization,⁶ regulation of electrolytes,⁷ development of new functionalized binders⁸ and so on. However, modifying the sulfur cathodes or separators introduces a large amount of inactive material, reducing the battery's specific energy density. Adjusting the electrolyte usually requires an inert atmosphere glove box, a complicated and costly procedure.⁹ As a simple and low-cost strategy, the development of new functional binders can not only avoid the introduction of inactive materials to reduce the specific energy density, but also avoid the complexity of operation, which has the application prospect of large-scale production.

At present, polyvinylidene fluoride (PVDF) is the predominant binder utilized in lithium ion batteries. However, due to the inelasticity of the semi-crystalline molecular chain structure of PVDF, although it contains rich polar-CF groups, the dipole moments in its own structure cancel each other,¹⁰ which further deteriorates its mechanical properties, and it can't adapt well to the sulfur volume change. Additionally, the weak van der Waals force makes it difficult to prevent the shuttle effect, and its application in Li–S batteries is restricted. Studies have shown that chemical covalent bonds and physical hydrogen bonds can induce self-healing repair of sulfur positive bulk expansion cracks and anchor LiPSs to alleviate the shuttle effect, so multi-functional binders designed with substances rich in polar groups show good potential in replacing PVDF.¹¹ Multifunctional binders with polar functional groups (such as –OH, –COOH, –NH₂) prepared by polymerization or crosslinking are applied to Li–S batteries. The binders can not only combine the elementary sulfur with the conductive agent, but also repair the cracks caused by volume expansion via the interaction of chemical bonds, so as to improve the electrode stability.¹² Additionally, polar groups in the binder have strong chemisorption ability towards the polysulfide, which can effectively inhibit the shuttle effect. At the same time, polar groups can easily combine with Li⁺ and accelerate its transfer, thus accelerating the redox kinetics of the polysulfide.¹³ Although multi-functional binders improve the specific capacity, rate performance and service life of Li–S batteries, there are still considerable challenges in the rational design of binders in the future commercialization process.

Therefore, this work proposes to extract aloe vera gel (AG) from waste aloe vera and cross-link it with D-aspartic acid (DAA) to prepare a sustainable aqueous multi-functional binder (AG–DAA) for the sulfur positive electrode of Li–S batteries, as shown in Fig. 1a. The main components of AG are aloin A, aloin emodin and aloin,¹⁴ all of which contain rich hydroxyl groups and can cross-link with the carboxyl branch chains in DAA. The network structure formed by crosslinking endows the binder with better mechanical strength and flexibility, and it can bond with the positive electrode material well and adapt to the volumetric changes of the electrode throughout the charge and discharge processes, so as to maintain the integrity of the electrode.¹⁵ The Density Functional Theory (DFT) calculation results combined with the characterization results of ex situ UV-vis, ex situ XPS, and in situ Raman show that the abundant polar functional groups in the AG–DAA binder can effectively enhance the chemical adsorption of polysulfides, reduce the conversion energy barrier of soluble LiPSs to insoluble Li₂S₂, and improve the redox reaction kinetics of the sulfur positive electrode. Therefore, the Li–S battery employing the AG–DAA binder demonstrates excellent electrochemical performance and cycle stability, i.e., a high initial specific capacity of 791.9 mA h g⁻¹ and a capacity decay of only 0.049% per cycle at a high rate of 4C (1C = 1675 mA g⁻¹).

2 Experimental part

2.1 Preparation of the AG–DAA binder

Fresh aloe vera was peeled, crushed, filtered, centrifuged and freeze-dried to obtain the solid aloe vera gel (AG). Then, it was mixed with D-aspartic acid (DAA) in distilled water at a mass ratio of 85:15, 75:25, and 65:35 for a cross-linking reaction at 40 °C for 4 h. The sample was freeze-dried following the cross-linking reaction, rinsed with distilled water to get rid of the unreacted DAA, and then freeze-dried once more to get the AG–DAA binders.

2.2 Electrode preparation

Carbon–sulfur composite materials were prepared by combining sulfur and carbon nanotubes (CNTs) at a mass ratio of 7:3. The composite was subsequently heated to 155 °C in an argon atmosphere for 12 hours to seal sulfur particles in the pores of carbon nanotubes. Carbon–sulfur composites, super P, and binder were combined in a 7:2:1 mass ratio to produce the electrode slurry. N-Methyl-2-pyrrolidone (NMP) was used as the solvent for the PVDF binder and distilled water was used as the solvent for AG and AG–DAA binders. The slurry was applied on an aluminium-based current collector and dried for 24 hours at 60 °C, and sectioned into circular pieces with a diameter of 12 mm to produce sulfur positive electrodes featuring an active material loading of 1.0–1.2 mg cm⁻².

10 wt% PVDF, AG and AG–DAA as the active materials, respectively, and 90 wt% CB as the conductive agent, with NMP or distilled water as the solvent were used. After mixing the above homogeneously and pasting the slurry on carbon cloth and drying, PVDF, AG and AG–DAA binder-based electrodes were obtained.

2.3 Preparation of Li₂S₆ solution and adsorption test

To prepare a 10 M Li₂S₆ solution, sulfur and Li₂S powder with a molar ratio of 5:1 were mixed in DME/DOL (v/v = 1:1) in a glove box filled with argon. The mixture was then allowed to react for 12 hours at 70 °C. After dilution to 1.5 M, 5 mL Li₂S₆ solution was taken and adsorbed with different binders for 12 h. The color changes before and after adsorption were recorded.

2.4 Material characterization

The cross-linking of AG and DAA was assessed using X-ray photoelectron spectroscopy (XPS, Thermo Scientific ESCALAB Xi+) and Fourier transform infrared spectroscopy (FT-IR, Thermo Scientific NICOLET IS20). In addition, XPS was also used to determine the structure and component change of AG–DAA after adsorbing the Li₂S₆ solution. The absorbance of the Li₂S₆ solution was determined after adsorption using a U-2900 ultraviolet-visible (UV-vis) spectrophotometer. The 180° stripping test of the electrode was carried out on a low-voltage pulling instrument (China-SANS-UTM5305H), and the displacement speed was kept constant at 10 mm min⁻¹. In the nanoindentation test, the electrode is tested using a KLA iNano® nanoindentation mechanical tester (continuous stiffness: controlled maximum load 1 mN).

2.5 Battery assembly

CR2032 coin cells were fabricated utilizing a lithium metal anode, Celgard 2400 separator, and an electrolyte comprising 1.0 M LiTFSI in 1,3-dioxane (DOL)/1,2-dimethoxyethane (DME) (1:1 v/v) with 2 wt% LiNO₃. The electrolyte-to-sulfur (E/S) ratio was controlled at 30 μL mg⁻¹ or 11.2 μL mg⁻¹ according to the sulfur loading.

2.6 Electrochemical measurements

Constant discharge–charge assessments were conducted using the LANDdt V7 battery testing apparatus, with a voltage window of 1.5 V to 3.0 V (versus Li/Li⁺) at room temperature. An electrochemical workstation (CHI660E) was used to conduct cyclic voltammetry (CV) and electrochemical impedance spectroscopy (EIS) tests. The half-cell's CV was evaluated within a voltage range of 1.5 to 3.0 V at scan rates of 0.1 to 0.5 mV s⁻¹, whereas the symmetrical cell's CV was assessed in a voltage range of −1.0 to 1.0 V at a scan rate of 1 mV s⁻¹, utilizing a Li₂S₆ solution electrolyte. EIS was conducted using a voltage amplitude of 5 mV and a frequency range of 100 kHz to 0.01 Hz. In situ electrochemical Raman testing was conducted utilizing a HORIBA spectrometer with excitation provided by a 532 nm laser under constant discharge–charge at 0.5C.

2.7 Nucleation experiment of Li₂S

A 0.20 M Li₂S₈ solution was prepared by dissolving sulfur and Li₂S (7:1 molar ratio) in 1.0 M LiTFSI solution (solvent: tetraethylene glycol dimethyl ether) and stirring in argon for 24 hours at 70 °C. A 2032 coin cell was assembled for the Li₂S nucleation experiment, with 30 μL 0.20 M Li₂S₈ solution serving as the electrolyte, a binder-based electrode serving as the positive electrode, and lithium metal serving as the negative electrode. After being discharged to 2.06 V with a current of 0.112 mA, the battery was immediately depleted at a continuous voltage of 2.05 V until the current was less than 0.01 mA.

2.8 Density functional theory (DFT) calculation

The binding energies of Li₂S₆ on binders were calculated by using the Gaussian 16 W software package. The 6-311 (d,p) basis set was adopted for geometric optimization and frequency calculation at the B3LYP theoretical level. To remove the base group superposition error's impact on the computation accuracy of the interaction energy, the equilibrium correction method is used to calculate the energy error and obtain a more accurate interaction energy. Therefore, the calculation formulae for the binding energies between different binders and Li₂S₆ are as follows:¹²

Ebinding energies = Etotal − (Esubstrate + ELiPSs)	(1)

Among them, E_total represents the total energy of the adsorption system, E_substrate represents the energy of the binder, and E_LiPSs represents the energy of Li₂S₆.

All of the Gibbs free energy calculations were performed with the Gaussian 09 package.¹⁶ Using the 6-31G(d) basis set for C, H, O, N, F, Li, and S, geometry optimization of all the minima and transition states involved was done at the B3LYP level. The default conditions for convergence were applied. To verify whether each optimized structure is an energy minimum point, vibrational frequency calculations were carried out at the same theoretical level as geometry optimization. The strength of the interaction between the adsorbates and binders was described by the adsorption energy (E_b), and the energy change during the adsorption process was determined using the following equation:

EGibbs = Emol/slab − Eslab − Emol	(2)

where E_mol/slab, E_slab, and E_mol are the total Gibbs free energies of the adsorbed system, slab and free species.

3 Results and discussion

Firstly, as illustrated in Fig. S1, the color of the AG binder exhibits a noticeable alteration after 30 days, while the AG–DAA binder remains in its original state, indicating that DAA modification could enhance the stability of the AG binder. Then in order to demonstrate the successful modification of aloe vera gel by aspartate, a series of characterization experiments of aloe vera gel before and after modification was performed. Fig. 1b illustrates that prior to modification, aloe vera gel exhibits a pronounced and narrow infrared absorption peak at 3400 cm⁻¹, indicative of the substantial quantity of hydroxyl functional groups present in its primary constituents.¹⁷ The stretching vibration of CO and the in-plane vibration of the benzene ring are represented by the absorption peaks at 1726 and 1062 cm⁻¹, respectively.¹⁸ Simultaneously, the characteristic absorption peaks of –COO (1596 cm⁻¹), –N–H (1509 cm⁻¹), –C–O (1244 cm⁻¹) and –C–N (1420 and 1023 cm⁻¹) of DAA can be clearly observed.¹⁹ However, after the cross-linking reaction between AG and DAA, even the FT-IR spectra of AG–DAA illustrate some identical functional groups to AG; what is more notable is that the obvious –N–H (1550 cm⁻¹) bond derived from DAA appears in AG–DAA. In addition, with the esterification reaction of –OH in AG and –COOH in DAA, the corresponding absorption peak of R–OH gradually weakens,²⁰ and the hydroxyl group also experiences a blueshift, indicating the generation of hydrogen bonds during cross-linking.²¹ These changes in FT-IR spectra preliminarily indicate that aloe vera gel is successfully modified by aspartic acid.

Subsequently, in the full XPS spectra of Fig. S2a and Table S2, it is evident that following AG crosslinking with DAA, the nitrogen content augments from 1.50% in AG to 2.81% in AG–DAA. The new C–N bond is seen at 399.7 eV in the high-resolution XPS spectra of N 1s (Fig. 1c),²² and the intensity of the N–H bond rises as well, suggesting that the crosslinking reaction between AG and DAA is successful. At the same time, in the O 1s, C 1s high-resolution spectra of Fig. 1d and S2b, it is evident that the characteristic peaks of the C–OH bond in AG–DAA located at 533.1 and 286.4 eV are obviously weakened,²³ and the CO bond location and intensity show obvious changes, which are a result of the cross-linking of carboxyl groups in DAA with abundant hydroxyl groups in AG. Consequently, the FT-IR and XPS characterization demonstrate the successful chemical crosslinking of AG and DAA.

The viscosity and mechanical properties of the binder directly affect the electrode structure's integrity and are also crucial to the battery's long–term cycle stability. Thus, the viscosity and mechanical properties of the AG–DAA binder were investigated. As shown in Table S1, the viscosity of 20 mg mL⁻¹ AG is 1123.0 mpa s, and the viscosity of 20 mg mL⁻¹ AG–DAA is 426.5 mpa s, which means viscosity decreased to some extent after the cross-linking esterification reaction. However, the viscosity of 20 mg mL⁻¹ AG–DAA is still higher than the same concentration of PVDF binder solution (20.1 mpa s), indicating the good viscosity of the AG–DAA binder which is comparable with the commercial product. Simultaneously, to assess the mechanical properties of the AG–DAA binder, the electrode sheets were prepared using different binders and then subjected to the 180° stripping test (Fig. 1e). The average peel strength of the AG–DAA/S electrode is greater than that of the AG/S and PVDF/S electrodes. Among them, when the mass ratio of AG to DAA is 75:25, the average peel strength of 1.5 N is the highest. It shows that the adhesion between the AG–DAA binder, sulfur, super P and current collector is stronger, which may be because the AG–DAA binder contains richer polar groups, not only hydroxyl, but also amide group, ester group or even unreacted amino and carboxyl groups, which form stronger interaction with active materials and conductive materials.²⁴ The bond between PVDF and the active material is weak, primarily due to the feeble van der Waals forces among PVDF molecules, resulting in the active substance falling off from the collector during charging and discharging, and the battery structure is unstable.²⁵

To further verify the binder's good mechanical ability to stabilize the electrode, nanoindentation tests (Fig. 1f) were conducted on electrodes based on PVDF, AG, and AG–DAA binders with different mass ratios. The expansion/contraction process of the positive sulfur electrode during charging and discharging was simulated by loading/unloading pressure on the electrode. It can be seen that the maximum displacement of the AG–DAA (75–25)/S electrode under maximum load and the final displacement after unloading are the smallest, followed by AG–DAA (85–15)/S, AG–DAA (65–35)/S, and AG/S electrodes, and that of the PVDF/S electrode is the largest. The results show that the AG–DAA (75–25)/S electrode has the smallest deformation and the highest mechanical capacity at the same indentation depth, which helps prevent the active ingredients' volume from expanding and avoiding the structural damage of the electrode during the charging and discharging process.²⁶ Fig. 1g shows the corresponding elastic modulus. The elastic modulus of the AG–DAA/S electrode surpasses that of the AG/S and PVDF/S electrodes, particularly the AG–DAA (75–25)/S electrode, which indicates that when AG–DAA is used, the electrode has better mechanical properties and bond strength, and can efficiently mitigate the stress induced by the volumetric expansion of the active material.²⁷ And the cross-section SEM images in Fig. S3 of the electrode using various binders before and after the cycle give direct evidence for the stability of the electrode. As can be seen, the initial thicknesses of the AG–DAA/S, AG/S and PVDF/S electrodes were 17.9 μm, 18.0 μm and 19.6 μm, while after lithiation, the electrode thicknesses changed to 19.1 μm, 22.4 μm and 27.4 μm, with volume expansion of 6.7%, 24.4% and 39.8%, respectively. Additionally, after delithiation, the electrode thickness using the AG–DAA binder could recover to 18.0 μm, closest to its initial state, whereas these values for AG and PVDF binders are 21.5 and 26.5 μm. All these changes indicate that the AG–DAA binder has the best mechanical ability to stabilize the electrode structure and reduce the shedding of active materials.²⁸

After finding that the AG–DAA binder has excellent mechanical properties and is suitable for use as a battery binder, 2032 coin batteries were manufactured and the electrochemical performance was evaluated. First, the CV curves of AG and AG–DAA (75–25) binder-based electrodes (Fig. S4) were investigated between 1.5 and 3.0 V at 0.1 mV s⁻¹. No obvious redox peaks were observed, indicating that the AG and AG–DAA binders are electrochemically inert and will not contribute to the discharge capacity.²⁹ Then, the CV curves of PVDF/S, AG/S and AG–DAA (75–25)/S cathodes were compared (Fig. 2a and S5). All cathodes exhibit two pairs of redox peaks. The peaks at approximately 2.31 V and 2.02 V correspond to the reduction of S₈ to long-chain lithium polysulfides (LiPSs) and the reduction of long-chain LiPSs to Li₂S₂/Li₂S, while the peaks at approximately 2.39 V and 2.46 V correspond to the oxidation of Li₂S₂/Li₂S to long-chain LiPSs and ultimately to S₈.³⁰ The CV curves of AG–DAA (75–25)/S electrodes show better reproducibility and lower polarization voltage than AG/S and PVDF/S electrodes, which indicates that the AG–DAA binder is beneficial for enhancing electrode stability. Then the cyclic stabilities of AG–DAA (85–18)/S, AG–DAA (75–25)/S, AG–DAA (65–35)/S, AG/S and PVDF/S electrodes were compared.³¹ As illustrated in Fig. 2b, the initial specific capacity of the AG–DAA (75–25)/S electrode can reach as high as 1130.8 mA h g⁻¹ at a rate of 0.5C, while the initial capacities of the AG–DAA (85–15)/S, AG–DAA (65–35)/S, AG/S, and PVDF/S electrodes are 989.7, 921.8, 681.7, and 483.7 mA h g⁻¹. And the AG–DAA (75–25)/S electrode can still retain a reversible capacity of 600.3 mA h g⁻¹ after 500 cycles, significantly surpassing PVDF and AG binder-based electrodes. The associated charge–discharge voltage profiles are demonstrated in Fig. 2c; as can be seen, the AG–DAA (75–25)/S electrode has the smallest polarization voltage and this trend is still satisfied at different cycles (Fig. S6a–e), consistent with the CV results in Fig. 2a. In addition, the high-voltage platform (Q_up) and low-voltage platform (Q_low) values within 500 charge–discharge cycles for these cathodes are extracted and compared as illustrated in Fig. 2d. Q_up indicates that the S₈ was reduced into soluble polysulfides and even to S₄²⁻ and Q_low indicates that S₄²⁻ was further reduced to insoluble Li₂S₂/Li₂S.³² Apparently, the AG–DAA (75–25)/S electrode demonstrates outstanding stability in maintaining the voltage plateau, especially in the high-voltage plateau region, where this advantage is more pronounced. Therefore, next, the AG–DAA (75–25)/S positive electrode was chosen as the representative to compare the electrochemical properties with AG/S and PVDF/S positive electrodes. Fig. 2e depicts the rate performance for AG–DAA (75–25)/S, AG/S and PVDF/S electrodes. As depicted, the reversible capacities of the AG–DAA (75–25)/S electrode at 0.5, 1 and 2C are 1097.9, 1000.6 and 887.1 mA h g⁻¹, respectively, and can reach 628.5 mA h g⁻¹ even at 4C, obviously higher than that of PVDF/S and AG/S electrodes. The specific capacity for the AG–DAA (75–25)/S electrode also returns to 989.4 mA h g⁻¹ after switching the current to 0.5C, also indicating the excellent reversible capability compared to other binder-based electrodes. Therefore, these electrochemical tests have proved that the DAA modified AG binder could effectively enhance the electrode stability and electrochemical performances. Finally, the long-cycling stability for the AG–DAA (75–25)/S electrode at the high rate of 4C was examined. As illustrated in Fig. 2f, the specific capacity attenuates from the original 791.9 mA h g⁻¹ to final 406.5 mA h g⁻¹ (after 1000 cycles), resulting in a capacity degradation rate of merely 0.049% per cycle. Finally, for the practical application, the cycling performance with a higher active substance loading was investigated. As demonstrated in Fig. S7, at a S loading of 3.0 mg cm⁻² and E/S ratio of 11.2 μL mg⁻¹, the AG–DAA (75–25)/S electrode can maintain a considerable areal capacity of 2.1 mA h cm⁻² after 40 cycles. Compared with previous reported studies on biomass-based binders for Li–S batteries (Fig. 2g and Table S3), the AG–DAA binder has significant comprehensive performance advantages in enhancing the cycle stability, rate performance and specific capacity of Li–S batteries.

In order to understand why Li–S batteries based on the AG–DAA binder show such excellent electrochemical performance, the enhancing mechanisms were comprehensively studied. First, the AG–DAA binder enhancing the Li⁺ diffusion was verified. The CV curves of the cathodes at various scan rates (0.1–0.5 mV s⁻¹) are displayed in Fig. 3a, S8a and b. Then the Randles–Sevcik equation (eqn (3)) is employed to calculate the Li⁺ diffusion rate.

	(3)

As seen, the peak current (I_p) has a linear relationship with the square root of the scan rate (v^0.5) and Li⁺ diffusion rate . And the values of n (n = 2), A and C_Li represent the number of electrons transferred in the reaction, electrode area and concentration of Li⁺ in the electrolyte, respectively.¹⁸ The higher the slope (I_p/v^0.5), the larger the D_Li⁺ is and the faster the Li⁺ diffuses.²¹ The D_Li⁺ values based on reduction and oxidation peaks are calculated in Fig. 3b, c and S9. The results show that the D_Li⁺ of Ic₁, Ic₂ and Ia₁ for the AG–DAA (75–25)/S electrode are 3.21 × 10⁻⁸, 4.59 × 10⁻⁸ and 2.39 × 10⁻⁷ cm s⁻¹, respectively, which are significantly greater than those of the PVDF/S and AG/S electrodes. The quicker D_Li⁺ in AG–DAA (75–25)/S can be attributed to the abundant polar groups of the AG–DAA binder that can bind with lithium ions, thereby accelerating the kinetics rate.³³ Additionally, as illustrated in Fig. S10a and 3d, on comparing the first cycle CV curves of AG–DAA (75–25)/S, AG/S and PVDF/S, the AG–DAA (75–25)/S electrode exhibits the steepest current peak and the smallest polarization voltage (Ic₁ peak at 2.31 V, Ic₂ peak at 2.02 V and Ia₁ peak at 2.48 V), which fully indicates that this binder has a noteworthy impact on decreasing electrochemical polarization and speeding up reaction kinetics.³⁴ What's more the Tafel fitting results from the redox peaks in CV curves could also prove the excellent kinetics of the AG–DAA (75–25)/S electrode. Based on the c₁, c₂ and a₁ peaks in Fig. S10a, the Tafel fitting results are illustrated in Fig. 3e, f and S10b. The fitting slopes of Ic₁, Ic₂ and Ia₁ peaks of the AG–DAA (75–25)/S electrode are 52.7, 55.0 and 64.6 mV dec⁻¹, respectively. Compared with AG (Ic₁ = 57.7 mV dec⁻¹, Ic₂ = 58.3 mV dec⁻¹, Ia₁ = 76.1 mV dec⁻¹) and PVDF (Ic₁ = 55.2 mV dec⁻¹, Ic₂ = 64.2 mV dec⁻¹, Ia₁ = 67.2 mV dec⁻¹), the AG–DAA (75–25)/S electrode has the lowest slope, indicating excellent reaction kinetics.³⁵

The catalytic activity of the AG–DAA binder in polysulfide conversion was further studied by assembling a symmetrical battery with Li₂S₆ as the electrolyte. Compared with AG and PVDF electrodes, the AG–DAA electrode exhibits more pronounced redox peaks, higher peak current, smaller voltage gap and greater closure area under the CV curve (Fig. 3g). This indicates that the introduction of DAA brings more polar groups to effectively enhance the transformation of LiPSs.³⁶ To further validate the Li⁺ transfer more quickly, EIS experiments were performed in addition to CV characterization. The Nyquist curves of AG–DAA (75–25)/S, AG/S, and PVDF/S electrodes before and after cycling are illustrated in Fig. 3h and S11. As can be observed, the charge transfer resistance (R_ct) prior to cycling is represented by a single semicircle (Fig. 3h) in the high-frequency area.³⁷ However, after cycling, besides this semicircle, a tiny new semicircle emerges (Fig. S11) in the low-frequency range that corresponds to the solid–electrolyte interface resistance (R_s).³⁸ The fitting results derived from the equivalent circuit diagrams are presented in Table S4. Obviously, the AG–DAA (75–25)/S electrode exhibits a smaller R_ct value compared to the AG/S and PVDF/S cathodes, again indicating the quicker Li⁺ transfer.³⁹ What's more, the smaller R_s value of the AG–DAA (75–25)/S electrode indicates that less Li₂S/Li₂S₂ is accumulated during the charging and discharging process,⁴⁰ which also suggests that the redox reaction rate of LiPSs is faster.

Subsequently, the enhancement effect of the AG–DAA binder on the liquid–solid transformation kinetics of the positive electrode surface was studied through Li₂S nucleation experiment, with results illustrated in Fig. 3i, S12a and b. Compared with AG/S (109.4 s, 71.25 mA h g⁻¹) and PVDF/S (373.1 s, 64.95 mA h g⁻¹), the AG–DAA (75–25)/S electrode shows faster nucleation response time (102.8 s) and stronger nucleation peak (95.98 mA h g⁻¹) during Li₂S nucleation. These results indicate that the AG–DAA (75–25)/S electrode has excellent catalytic activity in electrochemical sulfur conversion, which can facilitate the rapid transformation of polysulfide into Li₂S₂/Li₂S. The catalytic activity enhancement could be attributed to richer polar groups in the AG–DAA binder, which can provide more sulfurophilic sites and speed up the Li⁺ diffusion thus enhancing the LiPS redox.³⁴

The above series of electrochemical experiments confirm that the AG–DAA binder shows a significant improvement in battery performance. Then, the inhibitory effect of the AG–DAA binder on the shuttle mechanism and conversion process of LiPSs during the battery cycle was verified by DFT theoretical calculation. Firstly, according to the composition of AG, the structural units of aloin A crosslinked DAA were selected to represent the AG–DAA binder for DFT theoretical calculation. The results of the DFT computation are depicted in Fig. 4a. The binding energy between Li₂S₆ and AG–DAA binder is −2.21 eV, surpassing the binding energy between the AG binder (−2.14 eV) and the PVDF binder (−0.88 eV). This indicates that there is a stronger interaction between AG–DAA and LiPSs.

Secondly, the catalytic activity of different binders was further explored by comparing the Gibbs free energy in each step of the conversion reaction. As presented in Fig. 4b and c, the AG–DAA binder exhibits the lowest Gibbs free energy at every reaction step from Li₂S₈ to Li₂S, suggesting the highest spontaneity of these reactions.⁴¹ The Gibbs free energy for the reduction of Li₂S₂ to Li₂S is the largest throughout the reduction process, indicating that this process is the rate-limiting step. Compared with using the PVDF binder (0.92 eV), using the AG–DAA binder exhibits lower Gibbs free energy (0.61 eV) in this process, which proves to be more favourable for accelerating the redox reaction of sulfur on AG–DAA.²⁸ It is shown that the modified AG–DAA binder not only exhibits better adsorption effect, but also facilitates the anchoring of LiPSs and reduces the electron transfer distance, while also enhancing the kinetics of redox conversion.

Based on the results of electrochemical characterization and theoretical calculation, a series of ex situ and in situ material characterization studies were further conducted to verify the positive role of AG–DAA in enhancing the sulfur cathode performances. The visual absorption experiment in Fig. 5a shows that after adding different binders to the Li₂S₆ solution, the solution with AG–DAA changed from the original dark yellow to almost colorless, while the solution with AG became lighter yellow, and the solution with PVDF was almost unchanged. The corresponding UV-vis curves also reveal the same conclusion. According to the absorption peak in the 350–400 nm region, the absorption strength of the AG–DAA/Li₂S₆ solution is greatly reduced and shows minimal absorption strength, also indicating that AG–DAA has an effective adsorption effect on soluble polysulfide, thereby mitigating the shuttle effect throughout the cycle of Li–S batteries.⁴² The mechanism of chemisorption of LiPSs by the AG–DAA binder was further studied by XPS analysis. In comparison to the high-resolution XPS spectra of N 1s (Fig. 5b) and O 1s (Fig. 5c) for AG–DAA prior to absorption, it is found that the C–OH (533.2 eV) and CO (531.9 eV) bond redshift to 533.9 and 532.6 eV, respectively, and Li–N and Li–O bonds also appear at 402.4 and 535.2 eV, respectively.²² These changes prove the existence of chemical interaction between the AG–DAA binder and sulfur species. At the same time, it can be observed that Li–O (55.4 eV), Li–S (54.8 eV), Li–N (54.2 eV) and sulfonate (170.1 and 169.0 eV) bonds for the AG–DAA also appear in the Li 1s spectrum and S 2p spectrum in Fig. S13a and b,⁴³ which further proves that there is a chemical interaction between the AG–DAA binder and the LiPSs.

Finally, in situ Raman spectroscopy was employed to demonstrate the role of the AG–DAA (75–25) binder in alleviating the shuttle effect of lithium polysulfides. The in situ Raman contour maps and selected Raman signals of the commercial PVDF binder and AG–DAA binder assembled batteries during discharge were compared. As shown in Fig. 5d–g, at the OCV stage, signal peaks appear at 150, 214, and 466 cm⁻¹ to represent S₈²⁻.⁴⁴ Following the initiation of discharge in the PVDF/S battery, S₈²⁻ undergoes oxidation and progressively diminishes, while the signals at 150 and 214 cm⁻¹ persist but exhibit gradually weakening peak intensities. When the discharge voltage reaches the second plateau (≈1.95 V), the signal peak of S₆²⁻ and S₄²⁻appears at 400 cm⁻¹, which is attributed to the poor adsorption properties of PVDF.⁴⁵ Even at the end of the discharge process of 1.69 V, the S₆²⁻ and S₄²⁻ peaks can still be observed, indicating that the reduction process was slow and incomplete.⁴⁶ No discernible S₈²⁻ peaks are observed until charge termination, indicating pronounced shuttle effects in cells employing the PVDF binder that prevent the complete reoxidation of LiPSs to the S₈²⁻ state.⁴⁵ In contrast, for the AG–DAA/S battery, only a weak characteristic peak signal of S₄²⁻ is observed at 457 cm⁻¹ during the whole charging and discharging process,⁴⁷ demonstrating that AG–DAA significantly facilitates the fast transformation of LiPSs and mitigates the onset of the shuttle effect.

According to the above in situ and ex situ characterization, the anchoring mechanisms are demonstrated in Fig. 5h. Throughout the entire charge–discharge process, using the AG–DAA binder, S₈ can be effectively reduced to LiPSs and then re-oxidized to S₈²⁻ or even S₈ again, simultaneously the shuttle phenomenon of high-order LiPSs is effectively inhibited, whereas, while the PVDF binder is used, the LiPS redox conversion process is limited and causes serious loss of active material. This difference in the transformation mechanism could be attributed to the polar groups (oxygen-containing and nitrogen-containing groups) in the AG–DAA binder that can chemically bind the soluble polysulfides (Li₂S_n, n = 4–8) and inhibit the active LiPS shuttle to the lithium anode which causes corrosion and large amounts of lithium dendrite growth.¹⁵

4 Conclusions

A green, aqueous, multi-functional binder (AG–DAA) is successfully developed by cross-linking aloe vera gel and D-aspartic acid. The abundant hydroxyl functional groups in AG cross-link with the amino and carboxyl groups in DAA to form a stable network structure, which efficiently improves the influence of the volume change of the positive sulfur electrode during the charge and discharge process and maintained the stability of the electrode structure. In addition, its rich polar functional groups have strong adsorption capacity for polysulfide, which can accelerate the kinetics of the polysulfide conversion reaction, and help to achieve good cycle performance of Li–S batteries. Therefore, Li–S batteries based on the AG–DAA (75–25) binder exhibit superior rate performance and long cycle performance at high current density. At a low current of 0.5C, the initial specific capacity can reach 1130.8 mA h g⁻¹, and the reversible capacity can reach 600.3 mA h g⁻¹ after 500 cycles. Even at 4C high current, it also shows excellent cycle stability. After 1000 cycles, the specific capacity attenuates from the original 791.9 mA h g⁻¹ to 406.5 mA h g⁻¹, and the capacity attenuation rate per cycle is only 0.049%. This strategy based on the cross-linking modification of natural viscous biomass provides valuable guidance for the exploration of new binders that can meet the requirements of next-generation Li–S batteries.

Author contributions

All authors revised and edited the manuscript. All authors have read and agreed to the published version of the manuscript.

Conflicts of interest

The authors declare no competing financial interests.

Data availability

All the relevant data are within the manuscript and the SI.

Supplementary information is available. See DOI: https://doi.org/10.1039/d5ta04685h.

Acknowledgements

This work was supported by the Science and Technology Major Project of Fujian Province of China (No. 2022HZ028018), the National Natural Science Foundation of China (Grant No. 51902036, 51907030), the Science and Technology Major Project of Fuzhou of China (No. 2024-ZD-001), the Chongqing Bayu Scholars Support Program (No. YS2022050), the Key Science and Technology Research Program of Chongqing Education Commission (No. KJZD-K202200807), and the Research Project of Innovative Talent Training Engineering Program of Chongqing Primary and Secondary School (No. CY240806).

References

1. C. Ye, H. Li, Y. J. Chen, J. N. Hao, J. H. Liu, J. Q. Shan and S. Z. Qiao, Nat. Commun., 2024, 15, 4797 CrossRef CAS PubMed .
2. J. A. Feng, C. Shi, X. X. Zhao, Y. Zhang, S. Q. Chen, X. B. Cheng and J. J. Song, Adv. Mater., 2024, 36, 2414047 CrossRef CAS .
3. Y. J. Li, C. Wang, W. Y. Wang, A. Y. S. Eng, M. T. Wan, L. Fu, E. Y. Mao, G. C. Li, J. Tang and Z. W. Seh, ACS Nano, 2019, 14, 1148–1157 CrossRef .
4. Y. J. Li, J. B. Wu, B. Zhang, W. Y. Wang, G. Q. Zhang, Z. W. Seh, N. Zhang, J. Sun, L. Huang and J. J. Jiang, Energy Storage Mater., 2020, 30, 250–259 CrossRef .
5. S. Awasthi, S. Moharana, V. Kumar, N. Wang, E. Chmanehpour, A. D. Sharma, S. K. Tiwari, V. Kumar and Y. K. Mishra, Nano Mater. Sci., 2024, 6, 504–535 CrossRef CAS .
6. Y. J. Li, E. Y. Mao, Z. W. Min, Z. Cai, Z. H. Chen, L. Fu, X. R. Duan, L. Y. Wang, C. Zhang and Z. H. Lu, ACS Nano, 2023, 17, 19459–19469 CrossRef CAS PubMed .
7. Z. X. Shi, S. Thomas, Z. N. Tian, D. Guo, Z. M. Zhao, Y. Z. Wang, S. Li, N. Wehbe, A.-H. Emwas and O. M. Bakr, Nano Res. Energy, 2024, 3, e9120126 CrossRef .
8. G. T. Pace, M. L. Le, R. J. Clément and R. A. Segalman, ACS Energy Lett., 2023, 8, 2781–2788 CrossRef CAS .
9. G. M. Zhou, H. Chen and Y. Cui, Nat. Energy, 2022, 7, 312–319 CrossRef CAS .
10. Z. H. Li, Z. Y. Wu, S. X. Wu, W. T. Tang, J. C. Qiu, T. F. Liu, Z. Lin and J. Lu, Adv. Funct. Mater., 2024, 34, 2307261 CrossRef CAS .
11. F. X. Zhou, Y. H. Mei, Q. P. Wu, J. Q. Liu, X. Chen, H. Li, J. Xu and H. Y. Chen, Chem. Eng. J., 2025, 160009 CrossRef CAS .
12. J. Chen, X. Geng, C. Y. Wang, X. Hou, H. L. Wang, Q. L. Rong, N. Sun, W. Liu, L. Y. Hu and X. W. Fu, J. Mater. Chem. A, 2024, 12, 11038–11048 RSC .
13. X. He, Z. M. Liu, G. P. Gao, X. T. Liu, M. Swietoslawski, J. Feng, G. Liu, L. W. Wang and R. Kostecki, J. Energy Chem., 2021, 59, 1–8 CrossRef CAS .
14. İ. Kahramanoğlu, C. Y. Chen, J. Y. Chen and C. P. Wan, Agronomy, 2019, 9, 831 CrossRef .
15. Z. Z. Chen, M. J. Lu, Y. Qian, Y. Yang, J. Liu, Z. Lin, D. J. Yang, J. Lu and X. Q. Qiu, Adv. Energy Mater., 2023, 13, 2300092 CrossRef CAS .
16. A. D. Becke, J. Chem. Phys., 1993, 98, 5648–5652 CrossRef CAS .
17. S. Z. Li, Z. Z. Chen, J. Chen, X. T. Luo, X. Q. Qiu and Y. Qian, Adv. Energy Mater., 2025, 202405461 Search PubMed .
18. F. X. Zhou, Y. H. Mei, Q. P. Wu, H. Li, J. Xu and H. Y. Chen, Energy Storage Mater., 2024, 67, 103315 CrossRef .
19. J. W. Dong, L. P. Fang, J. Li, X. J. Gao, D. X. Li and S. J. Wang, J. Mol. Struct., 2023, 1291, 136045 CrossRef CAS .
20. W. H. Wang, W. Y. Li, S. Y. Jing, H. P. Yang, H. Q. Wang, L. Huang, Y. X. Mao, X. K. Pang, Y. D. Huang and L. Zhang, J. Energy Chem., 2024, 97, 352–360 CrossRef CAS .
21. S. Z. Li, W. S. Xiao, H. Do, H. Q. Yang, X. Q. Xu and C. Peng, Small, 2022, 18, 2107109 CrossRef CAS .
22. W. Chen, T. Y. Lei, T. Qian, W. Q. Lv, W. D. He, C. Y. Wu, X. J. Liu, J. Liu, B. Chen and C. L. Yan, Adv. Energy Mater., 2018, 8, 1702889 CrossRef .
23. C. Senthil, S.-S. Kim and H. Y. Jung, Nat. Commun., 2022, 13, 145 CrossRef CAS .
24. W. Q. Wang, L. Hua, Y. F. Zhang, G. C. Wang and C. Z. Li, Angew. Chem., Int. Ed., 2024, 63, e202405920 CrossRef CAS PubMed .
25. J. Liu, J. S. Yang, X. J. Shi, J. J. Shen, J. W. Mai, Y. Yang and Z. H. Yang, J. Electroanal. Chem., 2024, 968, 118502 CrossRef CAS .
26. S. J. Liang, J. Y. Zhang, C. E. Jia, Z. T. Luo and L. L. Zhang, Carbon, 2024, 222, 118807 CrossRef CAS .
27. Y. Y. Yu, C. Yang, J. D. Zhu, B. L. Xue, J. H. Zhang and M. J. Jiang, Angew. Chem., Int. Ed., 2024, e202418794 Search PubMed .
28. X. J. Lin, X. X. Liu, Y. Q. Tong, X. Y. Zhou, J. Li, J. Song, X. M. Feng, R. Q. Liu, L. Shi and A. S. Yu, Adv. Funct. Mater., 2024, 34, 2406985 CrossRef CAS .
29. G. W. Yu, G. F. Ye, C. Wang, C. Y. Wang, Z. Y. Wang, P. Hu, Y. Li, X. X. Feng, S. J. Tan and M. Yan, Sci. China Chem., 2024, 67, 1028–1036 CrossRef CAS .
30. Y. He, Q. L. Ruan, H. Kang, L. Ye, Y. Wang, T. F. Liu, H. R. Cheng and X. X. Gu, Nano Mater. Sci., 2025 DOI:10.1016/j.nanoms.2025.06.003 .
31. M. J. Lu, K. Chen, Z. Y. Jia, J. G. Ren, P. He, S. Y. Yang, R. Bagherzadeh, F. L. Lai, Y. E. Miao and T. X. Liu, Energy Storage Mater., 2024, 73, 103870 CrossRef .
32. S. B. Tu, X. Chen, X. X. Zhao, M. R. Cheng, P. X. Xiong, Y. W. He, Q. Zhang and Y. H. Xu, Adv. Mater., 2018, 30, 1804581 CrossRef .
33. Z. B. Cheng, Y. Y. Chen, J. Lian, X. L. Chen, S. C. Xiang, B. L. Chen and Z. J. Zhang, Angew. Chem., 2024, e202421726 Search PubMed .
34. W. Y. Jiang, T. P. Zhang, R. Y. Mao, Z. H. Song, S. Y. Liu, C. Song, X. G. Jian and F. Y. Hu, eScience, 2024, 4, 100203 CrossRef .
35. Y. Wen, X. Y. Lin, X. S. Sun, S. S. Wang, J. Wang, H. Liu and X. Xu, J. Colloid Interface Sci., 2024, 660, 647–656 CrossRef CAS .
36. Z. H. Shen, P. F. Song, W. Xie, L. Tannesia, K. Tang, Y. M. Sun, S. B. Xi and Z. c. Xu, Adv. Mater., 2025, 24, 18090 Search PubMed .
37. Q. He, W. K. Chen, B. Fan, Q. Y. Wei and Y. P. Zou, Chem. Eng. J., 2024, 496, 153813 CrossRef CAS .
38. C. A. Jenkins, S. R. Coles and M. J. Loveridge, ACS Appl. Energy Mater., 2020, 3, 12494–12505 CrossRef CAS .
39. J. C. Qiu, S. X. Wu, Y. J. Yang, H. Y. Xiao, X. J. Wei, B. K. Zhang, K. N. Hui and Z. Lin, ACS Appl. Mater. Interfaces, 2021, 13, 55092–55101 CrossRef CAS PubMed .
40. J. B. Liu, X. Y. Zhang, H. Y. Li, S. F. Jia, J. H. Li, Q. Li, Y. G. Zhang and G. R. Li, InfoMat, 2024, e12649 Search PubMed .
41. B. R. Li, T. P. Zhang, Z. H. Song, W. Y. Jiang, J. P. Yang, M. F. Pei, R. Y. Mao, X. G. Jian and F. Y. Hu, Adv. Funct. Mater., 2023, 33, 2306990 CrossRef CAS .
42. C. Li, Q. F. Sun, Q. Zhang, C. R. Xu, S. Wang, Y. Ma, X. X. Shi, H. Z. Zhang, D. W. Song and L. Q. Zhang, Chem. Eng. J., 2023, 455, 140706 CrossRef CAS .
43. W. Bi, C. H. Li, D. W. Yang, Y. Z. Zhang, L. Hu, Q. H. Gong, J. Zhang, Y. C. Zhang, M. Y. Li and J. S. Wei, Energy Environ. Sci., 2025, 18, 1929–1940 RSC .
44. Q. Gong, L. Hou, T. Y. Li, Y. C. Jiao and P. Y. Wu, ACS Nano, 2022, 16, 8449–8460 CrossRef CAS PubMed .
45. Z. H. Luo, M. Zheng, M. X. Zhou, X. T. Sheng, X. L. Chen, J. J. Shao, T. S. Wang and G. M. Zhou, Adv. Mater., 2025, 2417321 CrossRef CAS .
46. M. J. Shi, X. Han, W. Qu, M. H. Jiang, Q. Li, F. Jiang, X. Xu, S. Ifuku, C. L. Zhang and C. Wang, Adv. Mater., 2025, 2419918 CrossRef CAS .
47. Z. M. Huang, Y. H. Liang, Z. Z. Wu, Y. Kong, M. H. Bai, M. Li, B. Hong, T. Y. Huang, S. C. Huang and H. Chen, Adv. Mater., 2025, 2410318 CrossRef CAS .

---

Footnote

† These authors contributed equally to this work.

This journal is © The Royal Society of Chemistry 2025