In aqueous processing of silicon-graphite composite anodes, sodium alginate is introduced as a water-soluble polysaccharide binder because its β-D-mannuronate and α-L-guluronate residues adsorb onto silanol-terminated silicon surfaces and maintain interparticle contact during repeated lithiation. The dry film, however, exhibits a comparatively low strain-to-failure relative to carboxymethyl cellulose-styrene-butadiene rubber blends, and that mechanical limitation translates directly into cracking during drying, calendering, and cycling. A 2.5 wt% sodium alginate solution with a molecular weight of 100–300 kDa typically displays a low-shear viscosity of 5–30 Pa·s at 25 °C when measured on a Brookfield RVDV-II+ rotational viscometer with a small-sample adapter at 0.1 s-1; batch-to-batch variation in the M/G ratio and in residual calcium content of 0.2–1.0 wt% shifts the viscosity by as much as 40% because guluronate blocks participate in interchain junction zones. The slurry is prepared by dissolving sodium alginate in deionized water at 60–70 °C under an overhead stirrer, followed by cooling to 25 °C before the addition of conductive carbon black, silicon nanoparticles, and graphite; a rotor-stator high-shear mixer operating at 3000–6000 rpm is then used for 30–60 min to disperse agglomerates. Silicon nanoparticles with a specific surface area of 20–100 m2/g create a strong yield stress in the slurry, and the carboxylate groups of alginate undergo pH-dependent dissociation with a pKa near 3.2–3.5. When the slurry pH falls below 4.0, protonation of carboxylate groups produces alginic acid gel domains that are stiff and difficult to spread; when the pH exceeds 9.0, alkaline hydrolysis of the glycosidic linkages can reduce molecular weight and adhesive strength. The mixed slurry is then deaerated under vacuum at −0.09 MPa for 10–20 min to prevent pinhole defects in the dried coating.
The coating step on 10 µm electrolytic copper foil is usually performed by slot-die or doctor blade at a wet thickness of 50–200 µm. Aqueous alginate slurries exhibit shear-thinning behavior with a power-law index 0.4–0.7 in the shear-rate range 1–1000 s-1, but they also contain a low-shear yield stress that can be problematic in slot-die feed systems. The viscoelastic character of the slurry is measured by an Anton Paar MCR 302 rheometer with a parallel-plate geometry; the storage modulus often exceeds the loss modulus by a factor of 3–10 at 1% strain, indicating a gel-like network produced by alginate entanglements and by hydrogen bonding between carboxylate groups and oxide surfaces. This network is sensitive to shear history: excessive high-shear dispersion at 6000 rpm for more than 60 min can degrade the alginate molecular weight and reduce the slurry viscosity to a point where sedimentation of silicon particles occurs in the feed tank. Conversely, insufficient dispersion leaves carbon black agglomerates with diameters greater than 10 µm, and these agglomerates act as stress concentrators in the brittle binder after drying.
After coating, the wet film enters a convection dryer with three independent zones. The first zone is typically set to 60–80 °C, the second to 80–100 °C, and the third to 100–120 °C; belt speed and air velocity determine the drying rate. A brittle alginate film cannot relax capillary stresses that arise from water menisci between particles as the film consolidates. The critical cracking thickness is inversely related to the capillary pressure, and for a given particle packing the film will mud-crack when the tensile stress exceeds the low elongation at break of the binder. The resulting mud cracks are commonly observed as orthogonal lines in the dried electrode when the wet thickness exceeds 100 µm and the first zone temperature is above 80 °C. These cracks increase the electrical resistance of the coating and create edges where lithium plating can occur during cycling. Drying at a relative humidity above 60% reduces the skin-formation tendency but leaves residual moisture above 1000 ppm, which must be removed by a vacuum post-drying step at 110–130 °C for 12–24 h before cell assembly. The drying schedule therefore becomes a narrow processing window: too fast produces cracks, and too slow raises residual moisture and binder migration. Binder migration occurs when water evaporation carries soluble alginate to the coating surface, creating a binder-rich skin that is more brittle than the bulk and that delaminates under calendering.
The capillary stress in a drying particulate film is governed by the solvent surface tension, the contact angle of water on the particle surface, and the effective radius of the liquid menisci between particles. For water at 25 °C, the surface tension is 72.8 mN/m; for silicon oxide surfaces with a water contact angle of 10–30°, the meniscus curvature generates a tensile stress on the order of 1–10 MPa in a packed nanoparticle film. Sodium alginate films tested under ASTM D882 at 23 °C and 50% RH typically show tensile strengths of 50–120 MPa but elongation at break values of only 2–8%. The tensile strength is therefore not the limiting property; the low strain-to-failure means that the film cannot redistribute capillary stress by plastic deformation. Cracking initiates at surface defects, carbon black agglomerates, or edge beads and propagates through the binder phase. The crack spacing in a dried electrode can be measured by optical microscopy and is typically 0.5–5 mm for coatings thicker than 100 µm, depending on the drying rate and the alginate content.
The drying rate is controlled by the first-zone air temperature, air velocity, and the solvent partial pressure. At a first-zone temperature of 90 °C and an air velocity of 6 m/s, the surface dries rapidly and forms a low-permeability skin before the interior water has migrated to the surface. The resulting moisture gradient creates a biaxial tensile stress at the coating surface that exceeds the fracture strength of the brittle alginate network. Reducing the first-zone temperature to 55 °C and increasing the relative humidity to 50% RH lowers the skin-formation tendency and permits stress relaxation, but it increases the residence time required to reach a residual moisture of 500 ppm. The drying process can be monitored by near-infrared moisture sensors, and the target residual moisture after oven exit is commonly 500–1000 ppm for water-processed anodes. Higher residual moisture can hydrolyze LiPF6 in the electrolyte and generate HF, which corrodes the copper current collector and accelerates binder degradation.
A secondary effect of drying is binder migration. The convective flux of water toward the surface carries dissolved alginate and smaller carbon particles to the coating surface, producing a binder-rich skin. This skin has a higher elastic modulus and lower elongation than the bulk, and it can be identified by attenuated total reflectance Fourier transform infrared spectroscopy as a higher carboxylate peak intensity at the surface. The migrated skin is particularly problematic for calendering because it fractures at very low tensile strain and can delaminate from the underlying particle network. The use of a lower initial drying rate, a higher slurry viscosity, or a cosolvent such as isopropanol at 5–10 wt% can reduce binder migration but may increase the overall drying energy demand. The choice of drying protocol is therefore constrained by the brittleness of the alginate film rather than by the thermal stability of the materials.
| Film condition | Tensile strength | Elongation at break | Elastic modulus |
|---|---|---|---|
| Neat sodium alginate, 23 °C, 50% RH | 50–120 MPa | 2–8% | 3–6 GPa |
| Alginate plus 5 wt% glycerol | 30–70 MPa | 10–25% | 1–3 GPa |
| Alginate plus 0.5 wt% CaCl2 | 80–140 MPa | 1–4% | 5–8 GPa |
| Alginate exposed to 80% RH for 24 h | 20–50 MPa | 15–30% | 0.5–1.5 GPa |
Values are representative ranges compiled from film-level mechanical tests under ASTM D882 at a crosshead speed of 5 mm/min. Electrode-level values differ according to particle packing, coating thickness, and adhesion to the copper foil.
The calendering step compresses the dried electrode to a target porosity of 30–40% for silicon-graphite anodes, with the purpose of improving interparticle contact and volumetric energy density. For an anode containing 10–20 wt% silicon and 80–90 wt% graphite, the coating is typically calendered through a two-roll press with a roll diameter of 400–600 mm, a linear load of 50–150 N/mm, and roll temperatures of 60–90 °C. The strain applied to the coating in the nip depends on the initial thickness and the target thickness; a reduction from 100 µm to 70 µm corresponds to a compressive strain of 30% in the thickness direction and a corresponding tensile strain in the plane of the coating. Sodium alginate films with elongation at break values below 8% cannot accommodate this in-plane strain, and microcracks appear along the length of the web. The onset of cracking is detected by optical microscopy and by an increase in the electrical resistance of the electrode as measured by a four-point probe. Crack onset is also influenced by the particle size distribution: silicon nanoparticles embedded in a graphite matrix create local stress concentrations, and the brittle binder phase between particles fractures when the local tensile strain exceeds 2–4%.
The critical compression strain for alginate-bound silicon-graphite electrodes is lower than that for electrodes using CMC-SBR or polyacrylic acid binders. Published data for this specific configuration is limited, but film-level measurements under ISO 527-3 indicate that the elongation at break of sodium alginate is 2–8% at 23 °C and 50% RH, whereas CMC-SBR films can show values above 20%. The importance of the strain-to-failure is magnified by the thickness reduction ratio: a calendering reduction of 25–35% is common, and any local thickness nonuniformity in the incoming electrode is amplified in the nip. Edge cracks are observed first because the edge beads in the coating are thicker and experience higher local strain. Roll temperature has a limited softening effect on sodium alginate without a plasticizer because the glass transition of dry alginate is well above the calendering temperature, and the roll temperature mainly affects the graphite particle reorientation and the binder creep at the particle contacts. Heated rolls at 90 °C can reduce cracking by allowing some stress relaxation, but they may also cause copper foil softening and web wrinkling if the roll gap is not precisely controlled.
Adhesion to the copper foil is another brittleness-related failure mode. The peel strength of the compressed coating is measured by a 180° peel test according to ASTM D903 or by a cross-cut test according to ASTM D3359. Brittle alginate films show high tensile strength but low peel strength because the adhesive layer cannot undergo large extension before fracture; the peel front propagates through the binder phase rather than along the copper interface. Peel strength values for aqueous alginate anodes are often below 10 N/m unless a thin primer layer or a secondary polymer is used. During calendering, the combination of in-plane tensile strain and weak interfacial adhesion causes delamination at the coating-foil interface, particularly at the edges. This delamination reduces the electronically active area and creates folds during subsequent slitting and cell stacking. In automated electrode stacking lines, the peeled edges are detected by vision systems as bright copper regions, and coils with edge delamination exceeding 1 mm in width are typically rejected because they cannot be reliably welded to the tab.
When the dried electrode is exposed to a carbonate electrolyte composed of ethylene carbonate and dimethyl carbonate in a 1:1 volume ratio with 1 M LiPF6, the sodium alginate binder undergoes a limited but measurable solvent uptake. Sodium alginate is insoluble in carbonate solvents, but the polar carboxylate groups and residual water in the film interact with the high-permittivity ethylene carbonate component. Swelling of the binder phase is typically reported as a mass uptake of 5–15% after 24 h immersion at 25 °C, depending on the alginate content, the film porosity, and the residual moisture. This swelling plasticizes the binder and can reduce its elastic modulus by 30–50% compared with the dry state. The plasticization effect is initially beneficial for accommodating silicon expansion, but it also reduces the cohesive strength of the binder and can promote particle detachment. Swelling is nonuniform because the electrolyte penetrates through the pore network first and then diffuses into the binder phase; local swelling gradients generate tensile stresses at the particle-binder interface.
The electrolyte uptake also changes the fracture behavior of the binder. In the dry state, failure is brittle with a sharp crack tip; in the swollen state, the crack tip becomes blunted if the solvent uptake is sufficient to enable chain mobility. However, the same solvent uptake reduces the glass transition and the yield stress, so the film may creep under the compressive stack pressure of the cell. For pouch cells assembled with a stack pressure of 0.1–0.3 MPa, creep of the swollen binder can lead to thickness loss and porosity collapse. The electrochemical consequence is an increase in ionic transport resistance because the pore volume is reduced, and the mechanical consequence is additional cracking when the compressed binder is subjected to silicon expansion. The swelling behavior is measured by gravimetric solvent uptake after immersion and by dynamic mechanical analysis of free-standing films conditioned in electrolyte vapor. The loss modulus peak shifts to lower temperature in swollen films, confirming plasticization.
The interaction between swelling and brittleness also appears in the formation of the solid electrolyte interphase. When the binder cracks or swells, fresh silicon and graphite surfaces are exposed to the electrolyte, consuming lithium inventory and forming additional SEI. The SEI layer is itself brittle and inorganic, and repeated volume changes of silicon at 300% cannot be accommodated by the stiff alginate network without some local fracture. The result is a feedback loop in which binder fracture exposes more surface, leading to more SEI growth, which in turn increases the local stiffness of the composite and promotes further binder fracture. Electrochemical impedance spectroscopy of cycled electrodes shows an increase in the interfacial resistance after 20–50 cycles when alginate alone is used, and this increase is correlated with the appearance of cracks in post-mortem scanning electron microscopy.
When trace calcium ions and residual acidity raise crosslink density in the dry film, the mechanical failure mode shifts from ductile crack blunting to brittle fracture. Sodium alginate is particularly sensitive to divalent cations because the α-L-guluronate blocks form egg-box junction zones with Ca²⁺. Even a calcium concentration of 0.1 mM in the aqueous slurry can increase the storage modulus of the film by 20–50% and reduce the elongation at break from 8% to 3% under ASTM D882. This effect is exploited intentionally to increase the cohesive strength of the binder, but it also produces a stiffer and more brittle electrode that cannot tolerate calendering strain. The source of calcium may be the water, the graphite and silicon powders, or contamination from upstream equipment. Water-hardness levels above 50 ppm as CaCO₃ are generally regarded as incompatible with sodium alginate anode slurries unless a chelating agent such as sodium citrate is added at 0.1–0.5 wt%. Chelation competes with alginate for calcium and preserves a lower modulus film.
Residual acidity is another processing variable that increases brittleness. Silicon nanoparticles may carry surface silanol and silicic acid species that lower the slurry pH to 5–6. If the pH is adjusted below 4.0 during handling or by absorption of carbon dioxide, the carboxylate groups become partially protonated, and the polymer precipitates as alginic acid gel. The resulting film is more brittle than sodium alginate because the hydrogen-bonded acid form restricts chain mobility. In practice, the slurry pH is maintained between 6.5 and 8.0 by the addition of dilute sodium hydroxide or sodium carbonate; pH values above 9.0 are avoided because alkaline degradation reduces the degree of polymerization and weakens the binder. The pH of the slurry is measured by a calibrated electrode before coating, and the pH drift during storage is monitored because alginate solutions can support microbial growth that produces acidic byproducts. A preservative such as sodium benzoate at 0.05–0.2 wt% is sometimes used to prevent viscosity loss and acidification during extended storage.
The processing window for calcium and pH control is therefore narrow. A slight increase in calcium concentration may improve the initial adhesion of the electrode, but it also shifts the failure mode from cohesive yielding to brittle cracking. The correct balance depends on the silicon content and the calendering reduction. For silicon contents above 15 wt%, the binder must accommodate larger volume changes, so lower crosslink density and higher elongation are preferred; for graphite-rich anodes with silicon contents below 5 wt%, a slightly higher crosslink density may be tolerated because the mechanical strain from silicon expansion is lower. The addition of plasticizers such as glycerol or sorbitol at 1–5 wt% relative to alginate can partially offset the embrittlement caused by calcium, but plasticizers also increase the hydrophilicity of the dry film and raise the equilibrium moisture content. This trade-off must be managed by storing finished electrodes in a dry room at −40 °C dew point before cell assembly.
| Parameter | Test method or instrument | Target range | Failure mode outside range |
|---|---|---|---|
| Slurry viscosity at 25 °C, 10 s-1 | Brookfield RVDV-II+ / ISO 3219 | 1.5–4 Pa·s for 2 wt% alginate | Coating streaks or particle sedimentation |
| Slurry pH | Calibrated pH electrode | 6.5–8.0 | Below 4.0 forms alginic acid gel; above 9.0 degrades binder |
| Film tensile elongation | ASTM D882 | > 5% at 50% RH | Increased drying cracks and calendering fractures |
| Coating peel strength | ASTM D903 | > 10 N/m | Delamination from copper foil |
| Residual moisture | Karl Fischer / ISO 15512 | <1000 ppm | LiPF6 hydrolysis and HF generation |
| Calcium content | ICP-OES | <50 ppm as CaCO₃ | Excess crosslinking and embrittlement |
| Coating porosity after calendering | Mercury intrusion or pycnometry | 30–40% | Below range restricts ion transport; above range lowers density |
Electrochemical cycling of silicon-graphite anodes introduces a dynamic mechanical load that is fundamentally different from the static stresses of drying and calendering. During lithiation, silicon expands by approximately 300% by volume, while graphite expands by approximately 10% at full lithiation. The sodium alginate binder at the particle boundaries must accommodate this strain at C-rates that range from 0.1C to 5C in production cells. A brittle binder with an elongation at break below 8% cannot follow the expansion of the silicon domains, and it fractures locally at the particle surface. The fracture of the binder does not necessarily lead to immediate capacity loss, because the carbon black and graphite network can maintain electronic contact over short distances. However, the exposed silicon surface then reacts with the electrolyte to form an additional SEI layer, and the repeated fracture and reformation of SEI consume active lithium and increase the cell impedance. The capacity fade accelerates after 20–50 cycles for alginate-only silicon-graphite electrodes, whereas formulations with a more ductile binder or with a plasticizer tend to show a more gradual fade.
The mechanical degradation is aggravated by the constraint imposed by the cell stack. In a pouch cell, the electrode stack is held under a pressure of 0.1–0.3 MPa to maintain contact and reduce gas accumulation. Under this constraint, the thickness change of the anode during cycling is restricted, and the volumetric expansion of silicon is converted into lateral tensile strain in the binder phase. Sodium alginate films under constrained cyclic loading fail by low-cycle fatigue because the cyclic strain exceeds the elastic limit. The fatigue life of brittle alginate films under 1% tensile strain may be below 1000 cycles, as determined by dynamic mechanical analysis in strain-controlled mode; by comparison, CMC-SBR films can withstand higher cyclic strains before failure. The actual strain amplitude in a silicon-graphite electrode depends on the silicon loading, the degree of lithiation, the particle size distribution, and the local porosity. Because published data for this specific configuration is limited, the fatigue life of alginate-bound silicon-graphite anodes must be assessed by post-mortem cross-sectional scanning electron microscopy and by in situ dilatometry.
Post-mortem analysis of cycled electrodes typically shows three distinct crack populations: mud cracks inherited from drying, calendering cracks parallel to the web direction, and fatigue cracks that radiate from expanding silicon particles. The inherited mud cracks act as initiation sites for fatigue crack growth, so the drying defects are not eliminated by subsequent calendering or cycling. The fatigue cracks create pathways for electrolyte ingress, which then swells the binder and reduces its strength further. In addition, the interaction between sodium alginate and the acidic species produced by LiPF6 hydrolysis can lead to acid-catalyzed degradation of the glycosidic linkages. Trace amounts of HF, formed from residual moisture in the electrolyte, hydrolyze the polysaccharide backbone and reduce the molecular weight of the binder, causing a transition from brittle fracture to weak cohesion failure. This degradation mode is particularly severe if the electrode residual moisture exceeds 1000 ppm or if the cell is cycled at elevated temperatures above 45 °C. The operational boundary for aqueous sodium alginate silicon-graphite anodes therefore includes both dry-room storage and cell formation protocols that minimize HF generation.
The brittleness of sodium alginate can be reduced by blending with a secondary polymer or by adding a low-volatility plasticizer. Polyacrylic acid and carboxymethyl cellulose are miscible with sodium alginate in aqueous solution, and their carboxylate groups contribute to silicon adhesion while reducing the crosslink density of the dry film. A blend of sodium alginate and polyacrylic acid at a 1:1 mass ratio typically shows an elongation at break of 10–20% under ASTM D882, compared with 2–8% for neat alginate, while maintaining a peel strength above 10 N/m on copper foil. The improvement arises from the lower molecular rigidity of the polyacrylic acid segments and from the disruption of alginate-adjacent hydrogen bonding. However, polyacrylic acid is more acidic than sodium alginate, and the slurry pH may fall below 5.0 unless adjusted. Carboxymethyl cellulose blends improve slurry stability and reduce binder migration but may increase the viscosity beyond the slot-die limit if the degree of substitution is high.
Low-volatility plasticizers such as glycerol, sorbitol, and polyethylene glycol with a molecular weight of 200–400 g/mol are added at 1–5 wt% relative to alginate to increase the strain-to-failure. Glycerol at 5 wt% reduces the elastic modulus of the dry film by approximately 50% and increases the elongation at break from 5% to 15–25% under 50% RH. The plasticizer also lowers the drying stress by increasing the mobility of the alginate chains during film consolidation. The trade-off is a lower tensile strength and a higher equilibrium moisture content; glycerol-containing films may absorb more than 10% moisture at 80% RH, which is unacceptable for lithium-ion cell assembly. Plasticizer migration into the electrolyte can also occur during cycling, leaving behind a more brittle film at the anode surface. This migration is measured by aging free-standing films in electrolyte and monitoring mass loss and mechanical properties; glycerol and low-molecular-weight polyethylene glycol are known to leach out slowly, while sorbitol is more resistant to extraction.
Crosslinker control is an alternative strategy when high cohesive strength is required for calendering. A low concentration of calcium ions can increase the tensile strength without an excessive loss of elongation, but the processing window is narrow. Calcium chloride at 0.1–0.3 wt% relative to alginate can increase the peel strength from below 10 N/m to 15–25 N/m while keeping the elongation at break above 5%. Higher calcium concentrations produce a brittle film that fails under calendering. The use of a chelating agent such as sodium citrate at 0.1–0.5 wt% can buffer the free calcium concentration and prevent uncontrolled crosslinking. In production, the calcium content of the water, the powders, and the storage tanks is monitored by inductively coupled plasma optical emission spectroscopy, and the slurry is not released for coating unless the calcium concentration is below 50 ppm as CaCO₃. This control strategy reduces batch-to-batch variation in the mechanical properties of the anode and avoids the sudden appearance of calendering cracks that are otherwise difficult to diagnose.