Although the initially formed SEI can to some extent passivate the surface of graphite, it is not static and stable during the long-term cycling process. Instead, it continuously evolves under the combined effects of electrolyte reduction, interface layer growth, and mechanical damage^[34]. As shown in Figure 2A, transport phenomena like electron tunneling and solvent diffusion further trigger deposition of insoluble species. Thermodynamically, the lithiation potential of graphite, approximately 0.1 V relative to Li⁺/Li, is substantially lower than the lower stability limit of conventional carbonate electrolytes, approximately 1.0 V relative to Li⁺/Li^[35]. The electrolyte therefore remains metastable, making reductive decomposition unavoidable during graphite lithiation. Using ab initio molecular dynamics simulations, Leung et al.^[36] investigated the decomposition pathways of ethylene carbonate (EC) during the initial formation of the SEI at the graphite/electrolyte interface. Their results identified carbon edge sites as key catalytic centers for EC decomposition and demonstrated that rapid interfacial reduction can occur under experimentally relevant conditions, yielding products including Li₂CO₃, lithium ethylene dicarbonate, C₂H₄, and CO [Figure 2B]. This indicates that SEI formation originates primarily from the highly active interfacial reduction reactions at low potentials, and the differences in reactivity at different sites on the graphite surface further determine the spatial nonuniformity of the initial SEI formation.

Although the SEI is generally regarded as electronically insulating, a small number of electrons can still migrate through the layer by quantum tunneling or along low-resistance pathways. As illustrated in Figure 2C, energy level differences and electron transport barriers exist within the SEI. When the layer is sufficiently thin or contains structural defects, electrons may overcome these barriers and reach the SEI/electrolyte interface, where they induce further reduction of solvent molecules or lithium salt anions. Wang et al.^[37] adapted a schematic energy-level diagram to illustrate the thermodynamic origin of anode SEI formation, showing that electrolyte reduction occurs when the anode redox potential (μA) lies above the electrolyte lowest occupied molecular orbital, whereas electrolyte oxidation occurs when the cathode redox potential (μC) lies below the electrolyte highest occupied molecular orbital; SEI formation is therefore expected when the electrode potentials fall outside the electrolyte electrochemical stability window (Eg). Lin et al.^[38] reported an electron tunneling model based on density functional theory (DFT) calculations to evaluate the electronic insulating abilities of SEI components, suggesting that initial irreversible capacity loss is driven by self-limiting electron tunneling. Smeu et al.^[39] further demonstrated that grain boundaries between SEI components such as LiF can enhance electron tunneling. The resulting insoluble products accumulate on the existing SEI surface, promoting further outward growth and thickening of the interphase.

Li et al.^[40] employed a single-particle model incorporating both SEI formation and stress-induced cracking, allowing for a systematic identification of the diverse mechanisms driving SEI thickening. Graphite particles experience diffusion-induced stress owing to volume changes. This stress promotes the continued growth of the initial SEI on the particle surface and induces the nucleation and propagation of internal cracks. Edge et al.^[41] presented a schematic illustration [Figure 2D] to summarize the coupling between particle fracture and SEI growth in anode electrodes, showing that cracks expose fresh electron-conductive surfaces for new SEI formation, and that progressive SEI cracking/regrowth, crack network formation, electrode pulverization, and island formation ultimately lead to loss of active material and loss of lithium inventory. These results highlight that the exposure of fresh surfaces via cracking is a main way of driving rapid SEI thickening and the associated capacity decay.

Once the SEI is disrupted, fresh and highly catalytic graphite surfaces are directly exposed to the electrolyte. Solvent molecules, including EC and ethyl methyl carbonate, and lithium salt anions, including PF6^-, can then undergo further reduction at low potentials^[42]. The products and reaction pathways associated with recurrent electrolyte reduction are highly complex. Aurbach et al.^[43] investigated SEI formation on graphite in the presence of vinyl-containing additives using in situ Fourier transform infrared spectroscopy and an electrochemical quartz crystal microbalance. They identified cathodically induced vinyl polymerization as the principal reduction pathway of vinyl-containing additives. The reduction potentials of different electrolyte components vary substantially. Additives such as vinylene carbonate are generally reduced preferentially at relatively high potentials, approximately 1.35 V, whereas the main solvent EC is reduced at approximately 0.8 V. Zhang et al.^[44] established a controlled experimental system by deliberately adjusting the negative-to-positive electrode capacity ratio. A series of tests were combined to quantify major electrolyte components. These results confirm that the electrolyte decomposition pathway is driven by lithium plating rather than following the conventional pathway associated with anode aging. Furthermore, the results showed that vinylene carbonate was rapidly consumed after controlled lithium plating occurred, whereas PF6^-, ethyl methyl carbonate, and EC exhibited time-dependent nonfaradaic consumption. Beyond the mechanisms discussed above, lithium plating and the dissolution of transition metals from the cathode can further accelerate the degradation of the SEI layer. The deposited lithium on the graphite surface will react with electrolyte components, leading to the formation of additional SEI species and continued lithium consumption. In addition, the accumulation and growth of lithium plating can also disrupt the SEI and expose the fresh graphite surfaces. Besides, at elevated temperatures, the SEI becomes less thermally stable, which accelerates the decomposition of its components and promotes further electrolyte reduction^[45-47]. Transition metal ions dissolved from the cathode can migrate to the graphite anode and catalyze electrolyte decomposition [Figure 2E]. Electrochemical mass spectrometry showed that ions such as Mn²⁺ promote the decomposition of SEI components, thereby compromising SEI integrity and triggering additional electrolyte reduction^[48]

In specific electrolyte systems, solvated Li⁺ may enter the graphite interlayers before complete desolvation, resulting in solvent cointercalation^[49,50]. Solvent molecules may undergo further decomposition within the interlayers and generate gaseous products, producing substantial internal stress that expands the graphite galleries and induces structural degradation, including layer exfoliation and particle pulverization^[51]. Meanwhile, interlayer expansion and structural disruption can rupture the preexisting SEI, exposing fresh carbon surfaces to the electrolyte. Further electrolyte reduction then occurs, accompanied by the formation of a new SEI. Repeated SEI reconstruction is therefore induced and is commonly associated with gas evolution, active lithium consumption, and rapid capacity decay. Solvent cointercalation is consequently regarded as an important cause of early graphite anode failure.

In carbonate-based electrolytes, Luchkin et al.^[52] demonstrated that interfacial enrichment of EC increases the local liquidus temperature and enhances interfacial rigidity, thereby kinetically suppressing the penetration of solvated Li⁺ species into graphite during the initial SEI formation process [Figure 3A]. In contrast, this kinetic regulation is absent in propylene carbonate (PC)-based systems, allowing more extensive solvent cointercalation, which leads to severe graphite expansion and SEI fracture along edge planes.

For ether-based electrolytes, strong Li⁺-solvent interactions facilitate the cointercalation of Li⁺-ether complexes, leading to graphite exfoliation. Chen et al.^[53] employed a topological molecular engineering strategy to regulate ether solvent structures and systematically elucidated the influence of molecular electronic properties and steric volume on graphite lithiation behavior [Figure 3B]. They found that repeated SEI rupture and reconstruction caused by Li⁺/ether cointercalation were the main contributors to capacity loss, indicating a close relationship between solvent cointercalation and interfacial instability in ether-based electrolytes. Beyond conventional ether solvents, Yan et al.^[54] further demonstrated that reversible solvent cointercalation represents a more general interfacial phenomenon occurring across various electrolyte chemistries. Functional solvents, including trimethyl phosphate, triethyl phosphate, and tetramethylene sulfone, were all found to participate in reversible graphite cointercalation. Their study identified solvent coordinating ability and Li⁺ solvation structures as the two key parameters governing the onset potential of solvent cointercalation. By introducing film-forming additives with reduction potentials higher than the cointercalation threshold, such as fluoroethylene carbonate (FEC) and vinyl ethylene carbonate, a robust SEI can be preferentially established to prevent solvent penetration and suppress subsequent electrolyte decomposition.

In addition to the preferential formation of a protective SEI, solvent polymerization provides another route to suppress cointercalation. Li et al.^[55] developed a poly(1,3-dioxolane) (PDOL)-based quasi-solid-state electrolyte for lithium-sulfur batteries with graphite anodes through the in situ ring-opening polymerization of 1,3-dioxolane (DOL), initiated by the metallic 1T-MoS₂ sulfur host. As illustrated in Figure 3C, DOL molecules in the liquid electrolyte can cointercalate with Li⁺ into graphite and cause exfoliation. In contrast, the larger molecular size of the resulting PDOL chains hinders their entry into the graphite interlayers, suppressing solvent cointercalation while maintaining reversible Li⁺ intercalation and deintercalation. XRD and Raman analyses further showed that the quasi-solid-state system preserved the graphite structure and mitigated cointercalation-induced exfoliation and disorder. Chen et al.^[56] employed ultrasound-assisted in situ atomic force microscopy to visualize the dynamic evolution of solvent cointercalation, SEI formation, and gas generation in graphite anodes at the nanoscale. The combined AFM observations and DFT calculations revealed that the cointercalation and interlayer decomposition of Li⁺(EC)₄ solvation complexes were the primary causes of gaseous cavity formation within the graphite interlayers. These gas bubbles disrupted interfacial stability during battery formation cycles [Figure 3D]. By introducing fluorobenzene, a nonsolvating diluent with a low dielectric constant, the desolvation energy and interfacial molecular configuration were optimized, resulting in the complete suppression of solvent cointercalation and interlayer decomposition.

Low temperature is a major factor contributing to lithium plating on graphite anodes. Cooling simultaneously slows Li⁺ transport through the electrolyte and across the SEI, as well as solid-state diffusion within graphite particles, while increasing the charge-transfer resistance at the electrode/electrolyte interface. These coupled kinetic limitations substantially increase anode polarization and can drive the local graphite potential toward or below the Li⁺/Li equilibrium potential. Once the local potential reaches the lithium plating regime, Li⁺ reduction to metallic lithium becomes increasingly favored over intercalation into the graphite lattice^[60]. Previous studies have shown that metallic lithium deposition is more likely to occur on graphite anodes during operation or charging at temperatures below 0 °C^[61,62].

Cheng et al.^[63] combined differential voltage analysis, scanning electron microscopy, and Raman spectroscopy to accurately determine the critical depth of discharge and critical potential for the onset of lithium plating at low temperatures [Figure 4A]. Their quantitative analysis showed that the critical potential shifted markedly toward more negative values and the critical depth of discharge decreased sharply as the temperature declined. The study further demonstrated that lithium intercalation is considerably more sensitive to low temperature than lithium deintercalation. This kinetic imbalance can trigger premature lithium deposition, resulting in active lithium loss and capacity decay. Chen et al.^[64] used three-electrode galvanostatic measurements combined with differential capacity analysis to examine the competition between lithium intercalation and plating from -40 to 30 °C. At temperatures near room temperature, intercalation occurred before the onset of lithium plating. As the temperature decreased below -20 °C, however, the two reactions increasingly overlapped and proceeded simultaneously, indicating a shift in the dominant reaction pathway under low-temperature conditions [Figure 4B].

Kinetic analysis further showed that Li⁺ intercalation deviated from conventional Arrhenius behavior. Li⁺ was first reversibly adsorbed on the graphite surface, followed by charge transfer leading to the formation of LiC₆ or metallic lithium. By contrast, lithium plating followed typical Arrhenius behavior, indicating a single-step unimolecular process. Mao et al.^[65] systematically examined the combined effects of low temperature and current rate on lithium plating through electrochemical measurements, materials characterization, and numerical analysis. As the temperature decreased and the charging rate increased, the dominant charging mechanism gradually shifted from conventional lithium intercalation to lithium plating. At -20 °C and a charging rate of 0.5 C, the cell ultimately failed completely.

High-rate charging is another major factor promoting lithium plating on graphite anodes. At high current densities, the flux of Li⁺ delivered to the graphite/electrolyte interface increases substantially. When Li⁺ transport through the electrolyte and SEI, interfacial charge transfer, and subsequent solid-state diffusion within graphite cannot accommodate the imposed ion flux, Li⁺ accumulates near the electrode surface, progressively increasing the local overpotential and favoring metallic lithium deposition over intercalation into the graphite lattice^[6]. Shkrob et al.^[66] investigated cell polarization during rapid charging through charge-pulse and voltage-relaxation experiments at different rates. As shown in Figure 4C, when 10 s charging pulses were applied to an NCM523/graphite full cell initially equilibrated at 4.2 V, increasing the rate from 0.5 to 10 C markedly increased both the voltage rise during charging and the immediate voltage drop upon current interruption, followed by gradual voltage relaxation. Separate anode-potential measurements further showed that high-rate pulses could drive the graphite anode potential below 0 V vs. Li⁺/Li, bringing the anode into a potential regime where lithium plating becomes possible. Colclasure et al.^[67] developed an electrochemical model validated with cells of different electrode loadings. The model showed that poor electrolyte transport limits the performance of high-energy-density cells, leading to salt depletion in the anode and lithium plating at the graphite separator interface. Potential analysis further indicated that restricted liquid-phase transport increases local polarization, causing this interface to reach the conditions for metallic lithium deposition earlier than other regions.

Lithium plating generally begins as graphite approaches lithium intercalation saturation, and its onset occurs at an earlier stage of charging as the charging rate rises^[68]. Huang et al.^[69] employed operando differential pressure sensing to investigate lithium plating in graphite/NMC532 pouch cells during fast charging. As shown in Figure 4D, the pressure change per unit charge (d_P/d_Q) at both 2 and 3 C exceeded the lithium-plating threshold established under low-rate lithium intercalation conditions. Compared with 2 C, the 3 C profile crossed this threshold at a lower charge capacity and exhibited a higher maximum, indicating that increasing the charging rate accelerated the onset of lithium plating and intensified metallic lithium deposition. The negative values near the end of charging were associated with a pressure decrease caused by the reintercalation of plated lithium into graphite as the current declined during the constant-voltage stage. In addition, the relationship between interfacial charge-transfer resistance and state of charge exhibits pronounced rate dependence. A separate study reported that lithium plating began at a lower state of charge as the charging rate increased. In that cell system, lithium nucleation was already observed during 3 C charging at room temperature, whereas increasing the rate from 4 to 6 C intensified metallic lithium deposition and promoted the development of dendritic morphologies^[70].

Cell design defects are another important reason for lithium loss. An excessively low N/P ratio (the ratio of the anode capacity to the cathode capacity) indicates insufficient negative electrode capacity relative to the positive electrode, causing graphite to approach full lithiation more readily during charging and thereby increasing the risk of lithium plating. Abe et al.^[71] investigated the influence of the ratio of anode and cathode electrode capacities on the charging-discharging rate capability and cycle performance of LiFePO₄/graphite LIBs. The results showed that cells with a lower N/P ratio exhibit stronger polarization phenomena, which is related to a decrease in specific capacity.

The ratio of the anode and cathode areas also affects the deposition of lithium. Kim et al.^[72] investigated the lithium deposition near the electrode edges when the N/P ratio was relatively low. They compared three electrode configurations: an anode area larger than the cathode area, a cathode area larger than the anode area, and equal anode and cathode areas. At all test rates, the battery with a cathode area larger than the anode area exhibited the highest amount of lithium metal deposition. This result indicates that insufficient anode coverage leads to the accumulation of Li⁺ flux near the electrode edges, thereby promoting lithium deposition. In contrast, when the electrode area is equal or the anode area is larger than the cathode area, the amount of lithium deposition is significantly reduced. Extending the anode beyond the cathode was more effective in suppressing lithium plating at higher charging rates. Roth et al.^[73] further investigated the influence of electrode misalignment on the local anode potential distribution using a two-dimensional electrochemical model. As shown in Figure 4E, cathode overhangs of 0.1, 0.5, and 1 mm produced decreases in the anode-edge potential exceeding 60, 125, and 175 mV relative to the center after the first cycle, indicating a substantially increased risk of edge plating. In contrast, perfectly aligned electrodes maintained a nearly zero potential difference in this model. Although anode overhang also induced local potential heterogeneity during continuous cycling, its magnitude was considerably smaller than that caused by cathode overhang. These results indicate that anode overhang helps mitigate severe edge effects arising from manufacturing-induced misalignment, while its contribution to cycling-induced lithiation heterogeneity should also be considered in cell design.

Manufacturing defects, including active material detachment near electrode edges, nonuniform calendering, and their combined occurrence, can markedly alter the internal electrochemical environment and increase the risk of lithium plating. Chen et al.^[74] developed several three-dimensional electrochemical models to systematically clarify the effects of these electrode manufacturing defects. The models separately evaluated active material detachment near the anode edge caused by coating defects, density nonuniformity caused by uneven calendering in the electrode edge region, and the combined effects of both defects. In addition to promoting lithium plating, these defects may accelerate the formation of inactive lithium, repeated SEI reformation, and interfacial side reactions, ultimately intensifying capacity decay and increasing safety risks. Each of these factors can lower the local anode potential into the lithium plating region, thereby increasing the likelihood of metallic lithium deposition.

Fracture and pulverization are the typical mechanical failure modes of active particles in graphite anodes^[77]. Both arise from volume changes and diffusion-induced stresses, which build up during repeated cycling^[78]. Because Li⁺ transport and phase changes are not uniform across the particle, the stress distribution is also heterogeneous. Local stress concentrations therefore tend to develop around grain boundaries, particle edges and corners. The maximum tensile stress increases significantly with increasing charge rate and decreasing temperature. Under these aggressive operating conditions, local stress can easily exceed graphite's tensile strength, and that's when cracks start^[79]. Once a crack appears, further cycling makes it grow and join up with nearby cracks, eventually breaking the particle apart [Figure 5C]^[80].

These processes disrupt the structural framework of the graphite electrode and can also break electrical pathways between active particles and the conductive network. Fractured material loses effective electrochemical contact, leading to irreversible capacity loss^[81]. More importantly, fracture creates fresh graphite surfaces that are no longer protected by the existing SEI. These exposed surfaces induce further electrolyte reduction and SEI formation^[82]. This triggers continuous SEI reconstruction, a parasitic process that irreversibly depletes active lithium and electrolyte. Concurrently, the accumulating interphase thickens the diffusion barrier, leading to a steady rise in interfacial resistance.

Mechanical damage of graphite particles can gradually affect the overall anode structure. Due to the expansion and cracking, the transport pathways become increasingly heterogeneous^[83]. As shown in Figure 5D, both the particle and electrode volumes vary considerably during charging, together with changes in porosity and directional tortuosity^[84]. The pore network can also influence the electrolyte access and Li⁺ transport. Although the generated voids provide additional space for electrolyte penetration, excessive cracking and particle rearrangement can simultaneously reduce the solid-phase contacts. The resulting transport heterogeneity can produce local variations in current density and increase electrode polarization^[85].

Defects in the graphite lattice can also influence the mechanical instability at the atomic scale. The typical defects include Stone-Wales defects, single vacancies, and divacancies. By changing the local sp² carbon network, these defects reduce the mechanical integrity of the lattice and lower both the Young’s modulus and ultimate tensile strength of graphite. During cycling, defect-rich regions tend to experience greater local strain and are more prone to fracture than their well-ordered counterparts^[86]. These defects can also act as hotspots for stress buildup, making them natural nucleation sites for microcracks that can subsequently propagate through the graphite particle^[87]. As cycling proceeds, these defects may become localized structural instabilities and form larger cracks and particle fragmentation. This degradation pathway directly undermines both the structural integrity of the electrode and its ability to retain capacity over extended operation^[88].

Amorphous carbon coating has been widely used to improve the interfacial stability and electrochemical kinetics. A carbon layer on the graphite surface can limit the direct exposure of highly reactive sites to the electrolyte, which is beneficial for the suppression of side reactions. The carbon coating can also accommodate the volume changes of graphite during cycling. Meanwhile, the coating modifies the subsequent formation of the SEI, favoring a more uniform interphase and lowering interfacial resistance. These effects can significantly improve the electrochemical performance. Xiao et al.^[91] prepared amorphous carbon-coated graphite (GC) by chemical vapor deposition. As shown in Figure 6A, cross-sectional SEM of GC-2, obtained after 2 h of deposition, revealed a carbon coating approximately 90 nm thick, with local thicknesses of 89.3 and 91.8 nm at the two indicated positions. This coating covered unstable structures on the graphite surface, limiting their direct contact with the electrolyte. Meanwhile, the carbon layer provided additional pathways for Li⁺ diffusion, improved interfacial reaction kinetics, and reduced electrode polarization, thereby mitigating the risk of lithium plating during fast charging. The pouch cell assembled with a GC-2 anode and a LiCoO₂ cathode retained 87% of its initial capacity after 1,200 cycles at a charge/discharge rate of 5 C/5 C, demonstrating stable cycling performance at high rates [Figure 6B]. Ma et al.^[92] developed a Zn/N co-doped amorphous carbon coating on graphite through the in situ growth of ZIF-8 followed by carbonization. It exhibited enlarged interlayer spacing and enhanced lithiophilicity, which facilitated Li⁺ transport and promoted uniform lithium deposition/intercalation. Defects introduced by edge nitrogen doping provided rapid diffusion channels for Li⁺ between the carbon layers.

Zhang et al.^[93] proposed a scalable interlayer spacing regulation process that integrates mild oxidative intercalation with thermal reduction to enlarge the graphite interlayer spacing while repairing structural defects introduced during oxidation. Using this strategy, low-defect expanded graphitized needle coke was successfully synthesized. After modification, the densely stacked layers of petroleum coke-derived graphite became more open, giving rise to an expanded graphite structure. The interlayer spacing increased from 0.3375 to 0.3427 nm, while the graphite lattice was largely retained. This structural optimization enabled enhanced ion-accessible pathways without causing the severe lattice disruption typically associated with conventional strong acid oxidation treatments. A moderate increase in interlayer spacing can improve Li⁺ transport within graphite. Zha et al.^[94] prepared CO₂-derived graphite with expanded interlayer spacing (NK-M₄) by coupling molten-salt electrolysis of dilute CO₂ with in situ alkali-metal exfoliation. The volumetric expansion and shear forces associated with the in situ generated alkali metals promoted the expansion of graphite interlayers. As shown in Figure 6C, the interlayer spacing in the bulk region predominantly ranged from 0.338 to 0.341 nm, whereas local spacings near the edge increased to 0.352 and 0.364 nm. Figure 6D further reveals local edge spacings of 0.381 and 0.405 nm alongside clearly resolved graphite lattice fringes, demonstrating the coexistence of an ordered layered structure and localized interlayer expansion at the edges. Electrochemical measurements further indicated that this interlayer expansion facilitated Li⁺ diffusion and reduced charge-transfer resistance, thereby enhancing the high-rate lithium-storage performance of graphite. Rathi et al.^[95] reported a bifunctional acid treatment strategy for graphite modification. This approach simultaneously enlarged the graphite interlayer spacing and introduced nanopores, thereby shortening Li⁺ diffusion pathways and improving electrode reaction kinetics. As shown in Figure 6E, the treated expanded porous graphite (EG) exhibited a larger interlayer spacing and a more developed pore structure than pristine graphite (PG), while retaining the structural integrity of the graphite framework. The combined expansion of interlayer spacing and formation of nanopores accelerated Li⁺ intercalation and deintercalation, reduced diffusion resistance and polarization, and consequently improved rate capability and cycling stability. Based on the single-layer spacing regulation, Liu et al.^[96] further proposed a surface-embedded phase collaborative regulation strategy, and prepared ultrathin disordered carbon-coated micro-expanded graphite. The gentle intercalation and thermal expansion increased the graphite layer spacing from approximately 0.335 to 0.338 nm, forming more open edge lithium-intercalation channels; the approximately 5 nm disordered carbon layer on the surface provided defect adsorption sites and pseudocapacitive lithium storage behavior, while also improving the Li⁺ transport between particles and the interface compatibility between the electrode and the electrolyte. This combined regulation reduced the interface impedance and electrode polarization, promoted the formation of a thin and uniform inorganic-enriched SEI, and alleviated lithium shedding and particle cracking under high current density.

Elemental doping can modify the electronic structure and local chemical environment of graphite, which in turn affects Li⁺ adsorption, diffusion, and interfacial reactions. In a related two-dimensional carbon-based system, Ghorbani Shadpey et al.^[97] used first-principles calculations to investigate the effects of B, P, and S substitutional doping on the structure and lithium-storage behavior of porous C₂N monolayers. As shown in Figure 6F, substitution of carbon sites with heteroatoms produced X₂-C₂N structures (X = B, P, or S). The B- and P-doped systems remained largely planar, whereas S doping induced pronounced out-of-plane buckling, reflecting dopant-dependent changes in local bonding and lattice deformation. Further calculations showed that strong Li binding near B dopants created deep potential wells that hindered diffusion, whereas the local Li migration barrier near S dopants was approximately 0.36 eV. These results highlight the need to balance Li adsorption strength with ion-migration kinetics when designing heteroatom-doped structures, thereby avoiding excessively strong local binding that restricts Li diffusion. Huo et al.^[98] explored a bilayer graphene model and first-principles calculations to examine how nitrogen defects on graphite basal planes affect lithium nucleation and deposition. Introducing single-vacancy nitrogen defects (SV-3N) and divacancy nitrogen defects (DV-4N) changes the local electronic structure of the basal plane and increases its lithium affinity. Lithium preferentially adsorbs and deposits around the defective sites under an applied electric field. The SV-3N and DV-4N defects provide distributed nucleation sites, which helps to reduce local lithium accumulation and promotes a more even deposition process. The calculations further showed that nitrogen defects can regulate lithium adsorption energies, diffusion pathways, and lithiation behavior, providing a possible route to suppress heterogeneous lithium nucleation. Furthermore, other carbon-based electrode systems have also shown that the high electronegativity of nitrogen can redistribute electron density, while different nitrogen configurations exhibit distinct electronic structures and interfacial reactivities^[99].

Modification of electrolyte composition and lithium salt chemistry represents a straightforward approach to regulating SEI formation and interfacial transport. By introducing functional ionic species or adjusting the salt composition, the reduction pathways of electrolyte components can be redirected toward the formation of more stable interphases, while ionic conductivity and low-temperature interfacial kinetics can also be improved. Wang et al.^[100] developed a multisalt high-entropy (HE) electrolyte by combining equimolar LiPF₆, LiTFSI, LiFSI, LiDFOB, and LiNO₃ in PC, each at 0.2 M, giving a total salt concentration of 1.0 M. As illustrated in Figure 7A, increasing the diversity of lithium salts enhanced anion participation in Li⁺ solvation and weakened Li⁺-PC interactions, producing an anion-rich, disordered solvation sheath and promoting the formation of an inorganic-rich SEI dominated by salt decomposition. During the initial discharge at 0.1 C, the HE electrolyte exhibited a short plateau at approximately 1.7 V vs. Li⁺/Li, while effectively suppressing the prolonged plateau near 0.7 V associated with solvent cointercalation in the single-salt PC electrolytes. The potential subsequently decreased into the low-potential regime for lithium intercalation into graphite. These results demonstrate that varying salt composition, rather than increasing the total salt concentration, can jointly regulate solvation chemistry and interphase formation, suppressing PC cointercalation and improving compatibility with graphite anodes. Luo et al.^[101] introduced four representative SEI components, LiF, LiNO₃, Li₂O, and Li₂CO₃, into a commercial electrolyte to construct a multicomponent solute high-entropy electrolyte. The modified electrolyte maintained excellent rate performance even at low temperatures. This improvement was attributed to the combined effects of a weakly solvated Li⁺ environment and reduced interfacial activation energy, facilitating the Li⁺ transport across the electrode/electrolyte interface. The type of lithium salt also has a significant impact on the low-temperature interface dynamics. Raj et al.^[102] compared the electrochemical performance of LiPF₆, LiBF₄, and LiDFOB electrolyte salts with graphite anodes over a temperature range from -30 to 25 °C. The specific capacity of all electrolyte systems decreased as the temperature decreased, whereas the LiDFOB-based electrolyte (PEC113) exhibited the highest capacity retention and showed the lowest resistance at -30 °C, indicating more favorable ionic transport and interfacial charge-transfer kinetics.

While electrolyte composition determines the chemical environment from which the SEI forms, the local solvation structure of Li⁺ directly governs its desolvation and subsequent transfer across the electrode/electrolyte interface. In particular, excessive solvent coordination can increase the desolvation barrier and promote solvent-related side reactions, whereas weakly solvating environments can facilitate interfacial Li⁺ transport and suppress undesirable solvent participation. Solvation structure engineering has therefore emerged as an effective strategy for improving both fast-charging capability and low-temperature performance. Zheng et al.^[103] developed a weakly solvating electrolyte for fast-charging graphite anodes by introducing methyl acetate and fluorinated ethyl methyl carbonate (FEMC), both of which exhibit relatively weak Li⁺-solvent interactions. The FEMC-based electrolyte contained a larger fraction of anion-rich aggregates than the conventional EMC-based electrolyte. The modified solvation structure lowered the Li⁺ desolvation barrier and interfacial resistance, which favored faster charge transfer during high-rate charging. The corresponding SEI was also richer in inorganic components, with LiF becoming the dominant species. In the 1 M LiFSI/FEMC electrolyte, the LiF content reached about 90%, providing a more stable interface and facilitating Li⁺ transport across the interphase. Graphite electrodes in this weakly solvating electrolyte consequently delivered better rate performance and maintained relatively high discharge capacities at 2 and 4 C.

Beyond solvation structure, the size and spatial distribution of LiF within the SEI also significantly influence interfacial transport. Tao et al.^[104] investigated the influence of SEI microstructure on graphite anode kinetics using a weakly solvating electrolyte of 1 M LiPF₆ in 2-methyltetrahydrofuran (2MeTHF). Li⁺ insertion involves bulk transport, interfacial desolvation, transport across the SEI, and diffusion within graphite. The LiPF₆-2MeTHF electrolyte exhibited a contact-ion-pair-dominated solvation structure and formed an SEI with small, dispersed LiF crystallites, enabling effective passivation and favorable Li⁺ transport. In contrast, adding 2 vol.% FEC or replacing LiPF₆ with LiFSI generated larger, more densely distributed LiF crystallites, increasing SEI transport resistance. Despite similar desolvation barriers, these electrolytes showed substantially different activation energies for Li⁺ transport through the SEI, highlighting the importance of SEI microstructure. At -20 °C and 0.2 C, the LiPF₆-2MeTHF electrolyte maintained high reversible capacity. After rate testing, the cell retained approximately 91% capacity by the 200th test cycle at 2 C and 23 °C, with an average Coulombic efficiency of 99.9%. These findings demonstrate that coordinated control of solvation and SEI microstructure can improve fast-charging and low-temperature performance.

For operation over a broader temperature range, Luo et al.^[105] introduced isobutyronitrile (iBN) as a cosolvent to develop an electrolyte with a broad operating temperature range. Its low viscosity, relatively high dielectric constant, and weak coordination with Li⁺ weakened Li⁺-solvent interactions and lowered the desolvation energy barrier. The electrolyte maintained relatively high ionic conductivity as the temperature decreased, helping to reduce transport and interfacial kinetic losses. A LiCoO₂/graphite pouch cell using the iBN-60 electrolyte still delivered a relatively high discharge capacity at -70 °C. These results suggest that tuning Li⁺ solvation can be useful at both high and low temperatures, where desolvation and ion-transport kinetics become important limitations for graphite-based cells. Chen et al.^[106] introduced ethyl heptafluorobutyrate (EHFB) as a fluorinated cosolvent to regulate the Li⁺ solvation structure, shifting it from EC -dominated to diethyl carbonate-dominated coordination. This modification weakened Li⁺-solvent interactions, facilitated Li⁺ desolvation, and promoted the formation of a LiF-rich SEI on graphite. As shown in Figure 7B, the LiCoO₂/graphite pouch cell with the baseline electrolyte exhibited severe capacity decay at low temperatures, retaining only 8.6% and 2.0% of its room-temperature discharge capacity at -40 and -70 °C, respectively. In contrast, the EHFB-containing electrolyte significantly improved capacity retention to 73.9% and 58.9% at the corresponding temperatures [Figure 7C]. These results demonstrate that optimizing the solvation structure and interfacial chemistry can effectively improve the low-temperature performance of graphite-based batteries.

The effects of electrolyte regulation are closely reflected at the electrode/electrolyte interface, where the SEI controls Li⁺ transport, electron transfer, and parasitic reactions. Artificial interphase engineering offers an alternative to the SEI formed solely through electrolyte decomposition by allowing its composition, structure, and transport properties to be tailored in advance. Such strategies can provide more uniform Li⁺ transport, suppress continuous electrolyte reduction, and mitigate the interfacial polarization associated with fast charging. Liu et al.^[107] constructed a graphite-muconic acid composite anode (Gra@MCA) through an in situ synthesis strategy, in which the MCA layer served as a rechargeable organic artificial SEI. Compared with the conventional SEI, which can exhibit heterogeneous ion-transport pathways and increased interfacial resistance under high-rate conditions, the MCA layer provided more continuous and uniform pathways for Li⁺ transport. This regulated interfacial environment promoted more homogeneous lithiation, reduced anode polarization, and lowered interfacial resistance. Consequently, the Gra@MCA electrode exhibited substantially improved rate capability and cycling stability compared with PG. Weng et al.^[108] introduced 10 vol.% FEC into a LiPF₆-EC/DMC electrolyte to regulate SEI composition and interfacial transport kinetics on graphite anodes. As shown in Figure 7D, the SEI formed in the conventional electrolyte consisted mainly of amorphous organic species and Li₂CO₃ nanograins, with a thickness of approximately 3-7 nm. In contrast, FEC addition produced an SEI containing LiF nanoparticles, Li₂O nanograins, and amorphous organic components, resulting in a distinct interfacial microstructure. Despite its increased thickness of 44-56 nm, the modified SEI exhibited a reduced ion-transport resistance from 13.77 to 4.67 Ω, indicating improved Li⁺ transport across the interphase. Consequently, the reversible capacity of graphite at 4 C increased from 62.0 to 276.2 mAh g^-1, demonstrating enhanced high-rate lithium-storage performance. Studies of other carbon-based electrode systems have shown that the higher electronegativity of nitrogen atoms can redistribute the electron density within the carbon framework, while different types of nitrogen sites may exhibit distinct interfacial reactivities^[109]. More sophisticated interfacial architectures can further integrate the functions of electronic conduction, Li⁺ transport, and mechanical protection.

Pore architecture plays a central role in determining electrolyte penetration, Li⁺ transport, and the distribution of local current within graphite electrodes. Therefore, rational adjustment of pore size, connectivity, tortuosity, and spatial distribution is an important way to improve ion transport and reduce local Li accumulation. Ahmadi et al.^[110] investigated the effects of electrode microstructure on Li⁺ transport and cycling stability by constructing particle-size-gradient (PSG) and PG graphite anodes. As illustrated in Figure 8A, unlike conventional electrodes with a uniform microstructure, the PSG electrodes featured progressively larger graphite particles from the current collector toward the separator, whereas the PG electrode exhibited a gradual increase in porosity from 24% near the current collector to 46% near the separator. These gradient structures improved Li⁺ transport across the electrode thickness and reduced local concentration gradients and charge-transfer resistance, thereby alleviating electrode polarization. As shown in Figure 8B, after 200 cycles at room temperature with lithiation at 2 C and delithiation at C/3, the PG electrode retained approximately 80% of its initial capacity, outperforming PSG1 (67%), PSG2 (60%), and the conventional electrode (47%). The gradient electrodes also maintained Coulombic efficiencies above 99%, demonstrating that optimizing particle size and pore distribution can improve graphite-anode reaction kinetics and mitigate capacity degradation.

In conventional graphite electrodes, the pressing process causes the particles to be densely packed and compresses the pore network, thereby confining Li⁺ transport pathways and promoting local polarization and surface lithium plating during high-rate charging. By contrast, introducing through-hole carbon spheres (THCS) can create a gradient pore structure and interconnected transmission network within the electrode. As a result, Li⁺ can penetrate more effectively from the surface into deeper regions of the electrode, thereby reducing local concentration gradients and lowering the likelihood of lithium deposition. Electrochemical measurements further confirmed the effectiveness of this structural design. Under different charging rates, the THCS electrode delivered higher capacity and exhibited lower polarization, demonstrating significantly improved ion transport kinetics. However, merely optimizing the pore structure may still result in local current concentration in the channel region. Therefore, it is necessary to combine the design of low-curvature channels with interface chemical regulation^[111].

Binders play a critical role in maintaining particle-particle and particle-current-collector contact and thus preserving the mechanical and electronic integrity of graphite electrodes during repeated volume changes^[112]. Conventional aqueous CMC/SBR binder systems provide favorable dispersibility, flexibility, and adhesion, but their primarily mechanical function limits their ability to regulate interfacial chemistry. Functional binder design has therefore emerged as a strategy for simultaneously strengthening electrode cohesion and modifying the local electrode/electrolyte interface. Wang et al.^[113] introduced Nafion as a functional additive into a commercial CMC/SBR binder system to regulate the interfacial structure of graphite anodes. Nafion contains sulfonic acid groups and fluorinated chain segments that can interact with Li⁺ and participate in the formation of a more uniform and stable interfacial layer on the graphite surface. This structure helps suppress direct side reactions between solvated Li⁺ and graphite and promotes the formation of a thin and uniform SEI with favorable Li⁺ transport properties.

Consistent with this mechanism, TEM characterization after cycling revealed a relatively uniform and stable SEI for the CMC/SBR/Nafion electrode, whereas the conventional CMC/SBR system produced a thicker, cracked, and heterogeneous interphase. The incorporation of Nafion enabled the SEI to maintain greater structural stability during prolonged cycling. Binder engineering can also strengthen interparticle adhesion. Lim et al.^[114] introduced polydopamine (PD) and polyacrylic acid (PAA) coatings onto graphite particles to enable in situ chemical crosslinking during dry electrode fabrication, thereby enhancing interparticle adhesion and mechanical stability. As illustrated in Figure 8C, unmodified dry-processed graphite electrodes (DP-bGr) are susceptible to contact loss, cracking, and delamination during cycling, whereas PD coating partially improves interparticle bonding. Further incorporation of PAA enables chemical crosslinking with PD, forming an interconnected three-dimensional network that reinforces electrode structural integrity. As shown in Figure 8D, graphite/NCM622 full cells using crosslinked graphite electrodes (DP-xPDAA@Gr) retained 87.1% of their initial capacity after 500 cycles at 1 C, outperforming PD-coated electrodes (84.0%) and unmodified dry-processed electrodes (77.9%). These results demonstrate that strengthening interparticle mechanical bonding can mitigate structural degradation and electronic contact loss during cycling, thereby improving the long-term stability of graphite anodes. Taken together, these studies show that the role of the binder extends well beyond simply holding the electrode together. Functional groups and crosslinkable structures introduced into the binder can interact with electrode components and electrolyte, while also improving particle-to-particle contact and accommodating the deformation of graphite during cycling. These functions become increasingly important as cycling proceeds, particularly under high-rate conditions where mechanical and interfacial degradation are more pronounced.

For thick and high-loading graphite electrodes, simply increasing the amount of conductive additive or reducing electrode thickness is insufficient to overcome the coupled limitations of ionic and electronic transport. Three-dimensional conductive networks and spatially heterogeneous electrode architectures provide alternative approaches by independently tailoring transport pathways and local material composition across the electrode thickness. In principle, these architectures can simultaneously improve electronic conductivity, electrolyte accessibility, mechanical support, and current distribution.

Gradient electrodes are designed according to the distinct functional requirements of different regions within the electrode. Gradients in active material content, conductive additive fraction, or porosity can effectively alleviate volume expansion and stress concentration during cycling and thereby improve mechanical integrity. A bottom layer with a higher conductive additive content and a lower active material content is positioned near the current collector to enhance electron transport and interfacial adhesion to the current collector. A top layer with a higher active material content and greater porosity is positioned near the electrolyte to improve electrolyte wettability and ion transport while providing sufficient space to accommodate volume expansion^[115]. Wang et al.^[116] developed a three-dimensional carbon nanotube (CNT)-coated copper mesh composite current collector to improve interfacial adhesion and structural stability in graphite anodes. As illustrated in Figure 8E, graphite electrodes supported on conventional copper foil are susceptible to cracking and delamination during repeated cycling, while bare copper mesh provides insufficient mechanical support within its pores, resulting in active material cracking and detachment. In contrast, the CNT conductive network provides mechanical support for graphite within the copper mesh, enhances adhesion between the active material and current collector, and mitigates structural damage caused by repeated volume changes. As shown in Figure 8F, the graphite electrode using the composite current collector retained a specific capacity of 297.9 mAh g^-1 after 200 cycles at 0.5 C, corresponding to 94.3% capacity retention, outperforming the conventional copper foil electrode (238.3 mAh g^-1 and 81.2% retention). These results demonstrate that three-dimensional conductive current collectors can mitigate active material detachment and electronic contact loss, thereby improving the long-term cycling stability of graphite anodes.

In addition to constructing macroscopic gradients in porosity and active material content, establishing a three-dimensional network at the particle scale that provides both electrical conduction and mechanical support can also improve the cycling stability of composite anodes^[117]. Li et al.^[118] introduced CNTs into an artificial graphite matrix to construct a composite anode with a hierarchical conductive skeleton and multidirectional graphite microcrystalline structure. The CNTs formed continuous electronic transport networks between adjacent graphite domains, while the multidirectional graphite microcrystallites shortened the Li⁺ diffusion pathways, thereby synergistically improving electronic and ionic transport within the electrode. The optimized graphite/CNT anode delivered a reversible capacity of 256 mAh g^-1 at 5 C, substantially higher than the 126 mAh g^-1 obtained for PG, and retained a capacity of 268 mAh g^-1 after 800 cycles at 2 C. These results demonstrate that constructing a continuous three-dimensional conductive framework within graphite electrodes can improve electronic contact and Li⁺ transport, thereby enhancing rate capability and cycling stability. Three-dimensional conductive additives provide another method to separate electron and ion transport within the electrode. He et al.^[119] reported a reduced graphene oxide/Super P binary conductive additive. The reduced graphene oxide provides efficient two-dimensional electron conduction, and the Super P particles help to create additional pathways for ion transport. With this conductive architecture, LiFePO₄ and LiCoO₂ cells delivered specific capacities of 131.4 and 150.1 mAh g^-1 at 5 C.