The relentless pursuit of higher energy density in energy storage systems has positioned lithium-ion batteries at the forefront of technological innovation. Among the various components, the anode material plays a pivotal role in determining the overall performance of a lithium-ion battery. Silicon-carbon (Si/C) composites have emerged as highly promising anode materials due to silicon’s exceptionally high theoretical specific capacity of approximately 4200 mAh/g, which far surpasses that of conventional graphite anodes. This makes silicon-based anodes a key candidate for next-generation lithium-ion batteries. However, the integration of silicon into practical lithium-ion battery systems is severely hampered by its inherent physicochemical properties. During the lithiation and delithiation cycles, silicon undergoes a massive volume expansion of up to 300-400%. This repeated, severe mechanical stress leads to the pulverization of active silicon particles, loss of electrical contact within the electrode, and the continuous breakdown and reformation of the solid electrolyte interphase (SEI). These phenomena collectively result in rapid capacity fade and poor cycling stability, limiting the commercial viability of silicon anodes in lithium-ion batteries.
To mitigate these challenges, significant research efforts have been directed towards material nanostructuring, carbon coating, and the development of sophisticated composite architectures. While these strategies offer improvements, they often involve complex, costly syntheses. An alternative and complementary approach, which is both cost-effective and powerful, focuses on the engineering of the polymeric binder. The binder, though a minor component by mass, is crucial for maintaining electrode integrity. It binds active material particles together and adheres them to the current collector. Traditional binders like polyvinylidene fluoride (PVDF), which rely on weak van der Waals forces, are inadequate for silicon-based anodes in lithium-ion batteries due to their inability to accommodate large volume changes. Consequently, the search for advanced binders with robust mechanical properties, strong adhesion, and elastic recovery has become a critical research avenue for enhancing the performance of silicon anodes in lithium-ion batteries.
In this context, water-soluble binders with abundant polar functional groups have shown great promise. These binders can form strong hydrogen bonds or even covalent interactions with the silicon surface and other electrode components, providing superior adhesion. Particularly intriguing is the design of binders with a three-dimensional (3D) crosslinked network structure. Compared to linear (1D) polymer chains, a 3D network can distribute mechanical stress more effectively, restrict the sliding of polymer chains, and better encapsulate active material particles, thereby maintaining electrical connectivity throughout charge-discharge cycles in a lithium-ion battery. This work presents the design, synthesis, and comprehensive evaluation of a novel crosslinked binder system for Si/C anodes, specifically targeting the challenges faced in high-performance lithium-ion batteries.
The core innovation lies in the synthesis of a crosslinked polymer network via the formation of dynamic imine bonds (Schiff base). This network is created by reacting the amine groups of hyperbranched polyethylenimine (PEI) with the aldehyde groups of glutaraldehyde (GA). PEI is a polymer rich in primary, secondary, and tertiary amines, offering numerous sites for interaction and crosslinking. GA is a small molecule dialdehyde that acts as a crosslinker. The reaction between an amine and an aldehyde yields an imine bond (—C=N—), as shown in the chemical equation below:
$$ \text{R–NH}_2 + \text{OHC–R’–CHO} \rightarrow \text{R–N=CH–R’–CH=N–R} + 2\text{H}_2\text{O} $$
This reaction forms a 3D polymeric network, denoted as PEI-c-GA. The degree of crosslinking, and thus the mechanical properties of the binder, can be tuned by varying the ratio of GA to PEI. We hypothesize that this imine-bond-crosslinked network will exhibit high toughness, excellent adhesion, and the ability to accommodate volume changes in silicon, leading to significantly improved electrochemical performance for Si/C anodes in lithium-ion batteries.
Materials and Experimental Methods
The experimental workflow was designed to synthesize the crosslinked binder, fabricate electrodes, and perform a multi-faceted characterization to evaluate its efficacy in a lithium-ion battery context.
Binder Solution Preparation
Crosslinked binder solutions were prepared using an aqueous process. First, a 2 wt% aqueous solution of polyethylenimine (PEI, Mw ~70,000) was prepared. Separately, a 2 wt% aqueous solution of glutaraldehyde (GA, 50% in water) was prepared. These solutions were then mixed under vigorous stirring at specific mass ratios to achieve target crosslinking densities. For this study, two compositions were primarily investigated: a mixture with a PEI to GA ratio of 99.9:0.1, labeled as PEI-c-GA(0.1%), and a ratio of 99.8:0.2, labeled as PEI-c-GA(0.2%). The mixing immediately initiated the crosslinking reaction via imine bond formation. For comparative analysis, a 5 wt% solution of polyvinylidene fluoride (PVDF) in N-methyl-2-pyrrolidinone (NMP) and a 2 wt% aqueous solution of sodium carboxymethyl cellulose (CMC) were also prepared as reference binders.
Electrode Fabrication
Silicon-carbon composite (Si/C) powder with a theoretical specific capacity of 650 mAh/g was used as the active material. The electrode slurry was formulated with a mass ratio of Si/C : conductive carbon black (Super P) : binder = 8 : 1 : 1. The components were mixed in a planetary centrifugal mixer for 1 hour to ensure homogeneity. The resulting slurry was coated onto a copper foil current collector using a doctor blade. The coated electrodes were first dried in air and then transferred to a vacuum oven for overnight drying at 70°C. The dried electrodes were punched into 14 mm diameter discs. The mass loading of the active Si/C material was approximately 2.2 mg/cm². Electrodes are referred to by their binder: Si/C@PVDF, Si/C@CMC, Si/C@PEI-c-GA(0.1%), and Si/C@PEI-c-GA(0.2%). For high-mass-loading tests, electrodes with an active material loading of 3.1 mg/cm² were also prepared using the PEI-c-GA(0.1%) binder.
Cell Assembly
CR2032-type coin cells were assembled in an argon-filled glovebox with moisture and oxygen levels below 1 ppm. The prepared electrode disc was used as the working electrode. A lithium metal foil served as both the counter and reference electrode. A polypropylene separator (Celgard 2400) was used, and 70 µL of electrolyte (1 M LiPF6 in a mixture of ethylene carbonate (EC) and ethyl methyl carbonate (EMC) (1:1 by volume) with 10% fluoroethylene carbonate (FEC) additive) was added to each cell. The assembled cells were rested for 36 hours before electrochemical testing to ensure proper wetting.
Characterization and Testing Techniques
A suite of characterization techniques was employed to understand the binder’s properties and the electrode’s performance in a lithium-ion battery.
Fourier Transform Infrared Spectroscopy (FT-IR): Dried binder films were analyzed using FT-IR in the range of 400–4000 cm⁻¹ to confirm the formation of imine bonds.
180° Peel Test: The adhesive strength of electrodes was quantified using a universal testing machine according to the ASTM D903 standard. Electrode strips (35 mm x 19 mm) were peeled from the copper foil at a constant speed of 15 mm/min. The average force required for peeling was recorded.
Electrochemical Testing: All electrochemical tests were conducted using a battery cycler. The voltage window was set between 0.01 V and 2.0 V (vs. Li/Li⁺). A formation process of 3 cycles at 0.05 C (1 C = 650 mA/g) was applied to all cells. Rate capability was tested by subjecting cells to increasing current densities from 0.05 C to 4 C and back to 0.2 C. Long-term cycling stability was evaluated at a constant current density of 0.5 C for over 130 cycles. For high-loading electrodes, cycling tests at 0.2 C, 0.5 C, and 1 C were performed. Cyclic voltammetry (CV) was performed at a scan rate of 0.2 mV/s within the same voltage window.
Scanning Electron Microscopy (SEM): The surface morphology of electrodes before and after cycling was examined using SEM. Post-cycled electrodes were carefully disassembled in the glovebox, rinsed with EMC solvent, and dried under vacuum before imaging.
Results and Discussion
Chemical Confirmation of Crosslinking: FT-IR Analysis
The successful formation of the crosslinked network via imine bond formation was confirmed by FT-IR spectroscopy. The spectra for pristine PEI, PEI-c-GA(0.1%), and PEI-c-GA(0.2%) are presented and analyzed below. The pristine PEI spectrum shows characteristic peaks: a broad band around 3273 cm⁻¹ corresponding to N–H stretching vibrations, peaks between 2813–2933 cm⁻¹ for C–H stretching, a peak at 1663 cm⁻¹ for N–H bending, and a peak at 1293 cm⁻¹ for C–N stretching. Upon crosslinking with GA, significant changes are observed. The N–H bending vibration at 1663 cm⁻¹ diminishes notably. Most importantly, a new, distinct absorption band emerges at approximately 1570 cm⁻¹. This band is a clear signature of the C=N stretching vibration of the newly formed imine bonds. This spectroscopic evidence unambiguously confirms the chemical reaction between PEI’s amine groups and GA’s aldehyde groups, leading to the formation of the desired 3D crosslinked network. The intensity of this imine peak is slightly more pronounced in the PEI-c-GA(0.2%) sample, indicating a higher degree of crosslinking with increased GA content.
Mechanical Integrity: Adhesion Peel Strength
The adhesive strength between the electrode layer and the copper current collector is a critical parameter, as poor adhesion leads to active material detachment and rapid failure in a lithium-ion battery. The 180° peel test results provide a direct measure of this property. The average peel forces for the different electrodes are summarized in the table below.
| Electrode Binder System | Average Peel Force (N) | Standard Deviation |
|---|---|---|
| Si/C@PVDF | 1.43 | ±0.08 |
| Si/C@CMC | 2.69 | ±0.11 |
| Si/C@PEI-c-GA(0.1%) | 2.82 | ±0.13 |
| Si/C@PEI-c-GA(0.2%) | 1.66 | ±0.09 |
The data reveals a clear trend. The conventional PVDF binder exhibits the lowest adhesion strength (1.43 N), consistent with its weak van der Waals interactions. The linear polysaccharide binder, CMC, shows significantly better adhesion (2.69 N) due to its ability to form hydrogen bonds with the active material and current collector. Remarkably, the PEI-c-GA(0.1%) crosslinked binder demonstrates the highest average peel force of 2.82 N, surpassing even CMC. This superior adhesion can be attributed to the synergistic combination of multiple interaction mechanisms: (i) the numerous amine groups in PEI can form strong hydrogen bonds and possibly coordinate with surface oxides on silicon; (ii) the crosslinked 3D network creates a cohesive, tangled structure that mechanically interlocks with the electrode particles. However, when the crosslinker content is doubled to 0.2% (PEI-c-GA(0.2%)), the peel strength decreases to 1.66 N. This suggests that excessive crosslinking may reduce the number of free, adhesive functional groups (amines) available for interfacial bonding and could make the binder film too brittle. Therefore, an optimal crosslinking density exists, and for this system, it is achieved with the 0.1% GA formulation. This optimal mechanical adhesion is foundational for maintaining electrode integrity in a lithium-ion battery undergoing repeated volume changes.
Electrochemical Stability and Reaction Kinetics
Cyclic voltammetry (CV) was employed to investigate the electrochemical stability of the binder itself and the redox behavior of the Si/C composite within the electrode. First, a control electrode containing only the PEI-c-GA(0.1%) binder and conductive carbon (no active Si/C) was tested. Its CV curves over five cycles showed no distinct redox peaks within the 0.01–2.0 V window, confirming the electrochemical inertness of the crosslinked binder in the operational voltage range of a lithium-ion battery anode. This is a crucial requirement, as an electrochemically active binder would participate in side reactions, consuming lithium ions and degrading performance.
The CV profiles for the Si/C@PEI-c-GA(0.1%) electrode are shown in the analysis. During the first cathodic (lithiation) scan, a broad reduction peak below 0.2 V is observed, corresponding to the alloying reaction of lithium with silicon to form LixSi phases and the lithiation of the carbon component. In the subsequent anodic (delithiation) scans, two prominent oxidation peaks at around 0.30 V and 0.37 V appear, which are characteristic of the de-alloying process of the LixSi phases. The intensities of these peaks grow from the first to the fifth cycle, indicating the progressive activation of the silicon material. The excellent overlap of the CV curves from the second cycle onward, with minimal shift in peak positions, signifies highly reversible lithium insertion/extraction processes. This reversibility is a direct consequence of the robust electrode structure maintained by the crosslinked binder, which mitigates particle isolation and preserves electrical pathways—a key factor for stable operation in a lithium-ion battery.
Quantitative Performance Metrics: Initial Efficiency and Rate Capability
The initial coulombic efficiency (ICE) is a vital metric for any new anode material or system, as it reflects the irreversible lithium loss during the first cycle, primarily due to SEI formation. A higher ICE is desirable for practical lithium-ion batteries. The average ICE values for cells with different binders are compiled below.
| Electrode | Initial Coulombic Efficiency (ICE) – Average (%) |
|---|---|
| Si/C@PVDF | 78.29 |
| Si/C@CMC | 81.43 |
| Si/C@PEI-c-GA(0.1%) | 86.69 |
The Si/C@PEI-c-GA(0.1%) electrode achieves the highest ICE of 86.69%, significantly outperforming the PVDF (78.29%) and CMC (81.43%) based electrodes. This enhancement suggests that the crosslinked binder promotes the formation of a more compact and stable SEI layer, possibly by suppressing excessive electrolyte decomposition and by maintaining a stable electrode-electrolyte interface from the very first cycle. The strong adhesion and conformal coating provided by the binder may limit direct, continuous exposure of fresh silicon surfaces to the electrolyte.
Rate capability, which assesses the battery’s performance under high discharge/charge currents, is critical for applications like electric vehicles. The rate performance of the electrodes was tested from 0.05 C to 4 C. The specific discharge capacities at various C-rates are summarized in the following table, and the trend can be modeled by a simplified empirical equation for capacity retention under rate stress:
$$ C_{rate} = C_{0.1C} \cdot \left(1 – \alpha \cdot I^{\beta}\right) $$
where \( C_{rate} \) is the capacity at current \( I \) (in C-rate), \( C_{0.1C} \) is the baseline capacity, and \( \alpha \), \( \beta \) are constants related to polarization and kinetic limitations.
| C-rate | Si/C@PVDF (mAh/g) | Si/C@CMC (mAh/g) | Si/C@PEI-c-GA(0.1%) (mAh/g) |
|---|---|---|---|
| 0.1 C | ~625 | ~640 | ~655 |
| 0.5 C | ~520 | ~580 | ~620 |
| 1 C | ~350 | ~450 | ~550 |
| 2 C | ~180 | ~280 | ~420 |
| 4 C | ~100 | ~162 | ~305 |
| Return to 0.2 C | ~546 | ~590 | ~612 |
The Si/C@PEI-c-GA(0.1%) electrode exhibits outstanding rate capability. Even at a high current of 4 C, it retains a capacity of 305 mAh/g, which is approximately three times higher than the PVDF-based electrode and nearly double that of the CMC-based electrode. Furthermore, when the current is returned to 0.2 C, the PEI-c-GA(0.1%) electrode recovers 93% of its initial low-rate capacity, indicating minimal structural degradation. This superior rate performance stems from the crosslinked binder’s ability to maintain excellent electrical contact and ionic pathways within the electrode despite rapid lithium-ion flux, a common challenge in high-power lithium-ion batteries.
Long-Term Cycling Stability
The ultimate test for a silicon-based anode in a lithium-ion battery is its long-term cycling stability. Cells were cycled at a constant current of 0.5 C (325 mA/g) for over 130 cycles after the formation process. The evolution of charge specific capacity and coulombic efficiency is a key dataset. The capacity retention can be described by a semi-empirical decay model often used for lithium-ion battery analysis:
$$ Q_n = Q_0 – k_1 \cdot n – k_2 \cdot \sqrt{n} $$
Here, \( Q_n \) is the capacity at cycle \( n \), \( Q_0 \) is the initial capacity, and \( k_1 \), \( k_2 \) are constants representing linear loss (e.g., from SEI growth) and square-root related loss (e.g., from diffusion limitations or particle isolation), respectively.
| Electrode Binder | Charge Capacity at Cycle 10 (mAh/g) | Charge Capacity at Cycle 130 (mAh/g) | Capacity Retention (Cycle 130/Cycle 10) | Average Coulombic Efficiency (Cycles 5-130) |
|---|---|---|---|---|
| Si/C@PVDF | ~520 | 387.3 | 74.5% | 98.7% |
| Si/C@CMC | ~560 | 425.0 | 75.9% | 99.1% |
| Si/C@PEI-c-GA(0.1%) | ~610 | 517.2 | 84.8% | 99.5% |
The Si/C@PEI-c-GA(0.1%) electrode delivers the highest specific capacity throughout the cycling test. After 130 cycles, it retains a charge capacity of 517.2 mAh/g, which is substantially higher than 387.3 mAh/g for PVDF and 425.0 mAh/g for CMC. Its capacity retention from the 10th to the 130th cycle is 84.8%, indicating slower degradation. Moreover, its average coulombic efficiency after the first few cycles stabilizes at 99.5%, higher than the other two systems. This superior cycling stability is a direct manifestation of the crosslinked binder’s efficacy in accommodating volume strain, preventing electrode disintegration, and fostering a stable SEI. The 3D network acts as a flexible yet resilient matrix that holds the Si/C particles together, allowing them to expand and contract without losing electrical contact—a fundamental advancement for durable silicon anodes in lithium-ion batteries.
To challenge the binder further, electrodes with higher active mass loading (3.1 mg/cm²) were fabricated using PEI-c-GA(0.1%) and cycled at different rates. The results are summarized below, demonstrating the binder’s scalability and effectiveness under more realistic conditions for a lithium-ion battery.
| Current Density (C-rate) | Initial Capacity (mAh/g) | Capacity after 250 cycles (mAh/g) | Capacity Retention after 250 cycles |
|---|---|---|---|
| 0.2 C | ~620 | 454.7 | 73.3% |
| 0.5 C | ~600 | 411.8 | 68.6% |
| 1 C | ~580 | 378.6 | 65.3% |
Even with this higher mass loading, which amplifies mechanical stresses, the electrodes maintain good cycling performance over 250 cycles, underscoring the robustness of the crosslinked binder architecture in a lithium-ion battery configuration.
Post-Mortem Morphological Analysis
Scanning electron microscopy (SEM) provides visual evidence of the electrode’s structural evolution. Fresh electrodes for all binder systems show relatively smooth and crack-free surfaces. However, dramatic differences emerge after 130 cycles at 0.5 C. The Si/C@PVDF electrode surface appears severely cracked and granular, with many isolated Si/C particles visibly detached, indicating catastrophic failure of the binder network. The Si/C@CMC electrode shows large cracks and a wrinkled, uneven morphology, suggesting that while better than PVDF, the linear CMC binder cannot fully restrain the volume changes. In stark contrast, the Si/C@PEI-c-GA(0.1%) electrode surface remains remarkably intact and uniform. Only fine, hairline cracks are observable, and the overall electrode coating maintains its adhesion to the substrate. This morphological stability directly correlates with the superior electrochemical performance and confirms the central hypothesis: the imine-bond-crosslinked 3D network effectively constrains the silicon particles, buffers the mechanical stress, and preserves the electrode’s structural integrity throughout the demanding cycles of a lithium-ion battery.
Conclusion and Perspectives
In this comprehensive study, a novel water-soluble binder with a three-dimensional crosslinked network was successfully designed and synthesized for silicon-carbon composite anodes in lithium-ion batteries. The binder was created through a simple yet effective Schiff base reaction between polyethylenimine (PEI) and glutaraldehyde (GA), forming dynamic imine linkages. Systematic characterization confirmed the chemical crosslinking and revealed its profound impact on mechanical and electrochemical properties.
The optimized binder, PEI-c-GA(0.1%), demonstrated exceptional adhesive strength (2.82 N peel force), surpassing both conventional PVDF and widely studied CMC binders. When deployed in Si/C anodes for lithium-ion batteries, it enabled a high initial coulombic efficiency of 86.69%, outstanding rate capability (305 mAh/g at 4 C), and remarkable long-term cycling stability (517.2 mAh/g after 130 cycles at 0.5 C). These performance metrics significantly outperformed those of electrodes employing PVDF or CMC binders. Post-cycling morphological analysis provided direct visual evidence that the crosslinked network effectively maintains electrode integrity by accommodating volume changes and preventing active material detachment.
The success of this binder strategy can be attributed to multiple synergistic factors: (i) the strong adhesion from PEI’s amine groups, (ii) the stress-distributing capability of the 3D crosslinked network, and (iii) the potential dynamic nature of the imine bonds, which may allow for some self-healing or stress relaxation. This work underscores the critical importance of binder engineering as a powerful and relatively simple approach to unlock the full potential of high-capacity anode materials like silicon. The principles demonstrated here—using dynamic covalent chemistry to create resilient polymeric networks—can be extended to other electrode systems facing similar mechanical degradation challenges. Future work could explore the tunability of the network (e.g., using different crosslinkers or PEI molecular weights), investigate the binder’s behavior in full-cell configurations with various cathodes, and delve deeper into the SEI stabilization mechanism. This research contributes a significant step toward the realization of durable, high-energy-density lithium-ion batteries capable of meeting the growing demands of portable electronics, electric transportation, and grid-scale energy storage.