The solid electrolyte interphase (SEI) is a chemically heterogeneous layer formed by reactions between an electrode and an electrolyte. It is best known as the passivating film on the negative electrodes of rechargeable lithium batteries. A functional SEI allows lithium ions to move between the electrolyte and the electrode while strongly restricting electron transfer to electrolyte components. This selective transport permits batteries to operate with electrode–electrolyte combinations that would otherwise undergo sustained decomposition.
The SEI is not simply a deposited impurity or a perfectly stable protective coating. Its composition, thickness, porosity, and mechanical integrity evolve during manufacture, charging, storage, and ageing. Formation consumes electrolyte and usually immobilizes some of the battery’s cyclable lithium, but an effective interphase greatly reduces subsequent losses. Its behavior influences capacity, charging rate, low-temperature performance, service life, and safety.
Although the concept originated in studies of alkali-metal electrodes, it now encompasses interphases on graphite, silicon, and other negative-electrode materials, as well as related phenomena in sodium and solid-state batteries. Understanding it requires linking electrochemical reactions, nanoscale structure, ion transport, mechanics, and the practical conditions under which a cell is manufactured and operated.
Origins of the concept and its electrochemical significance
The SEI concept developed from a central problem in electrochemistry: highly reducing electrodes can remain usable in electrolytes that are not thermodynamically stable against them. Metallic lithium, for example, reacts with many organic liquids and dissolved salts. Nevertheless, the reaction can slow substantially after a surface film forms. The film separates the electronically conducting electrode from most of the electrolyte while providing pathways for the working ion. Its protective function resembles passivation on other metals, but ion conduction through the film is essential rather than incidental.
In 1979, Emanuel Peled formulated the solid electrolyte interphase model for alkali and alkaline-earth metals in nonaqueous battery systems. The name emphasized that the reaction product could behave as a solid electrolyte between a metal and the surrounding liquid. Subsequent work extended the concept to lithium-intercalating carbon electrodes. Studies associated with Doron Aurbach and other researchers connected electrolyte composition, surface reaction products, and reversible lithium storage. Peled and collaborators later developed a mosaic description to account for the interphase’s heterogeneous constituents rather than treating it as one uniform compound.
The distinction between an interface and an interphase is important. An interface is ideally a boundary between two phases; an interphase occupies a finite volume with properties different from those of either adjoining material. The battery SEI can contain crystalline particles, amorphous inorganic material, organic compounds, trapped liquid, and pores. Its thickness is not universal. Films on well-passivated graphite commonly have dimensions ranging from several nanometres to tens of nanometres, whereas reaction layers on highly reactive or repeatedly damaged surfaces can become much thicker.
The SEI also illustrates the difference between thermodynamic stability and kinetic persistence. Electrolyte reduction may remain favorable in terms of Gibbs free energy, but the reaction becomes slow when electrons and reactants cannot readily reach a common reaction site. Passivation therefore does not imply that the electrode and electrolyte have reached chemical equilibrium. A functioning battery instead maintains a metastable configuration in which desired charge transfer proceeds much faster than parasitic decomposition.
Related terminology must be used carefully. The cathode electrolyte interphase, or CEI, usually denotes reaction-derived material at the positive electrode, often involving oxidative electrolyte decomposition. It is not simply an SEI with the opposite charge. Likewise, an intentionally applied electrode coating is not automatically an SEI: it may remain a distinct barrier, react into an interphase, or become incorporated into a subsequently formed SEI.
Formation reactions and chemical composition
SEI formation begins when an electrode reaches potentials at which electrolyte components can accept electrons. In a conventional lithium-ion cell, this occurs during the first charge as the graphite negative electrode is driven toward low potentials relative to lithium metal. Reactions involve the solvents, dissolved salt, additives, surface functional groups, and trace contaminants. Their relative importance depends on local potential, temperature, concentration, and the chemical and crystallographic character of the electrode surface.
Organic carbonate electrolytes provide the most extensively studied examples. Ethylene carbonate can undergo electron transfer followed by ring opening, radical reactions, and precipitation. One frequently used idealized net reaction is
where EC denotes ethylene carbonate and the solid organic product is lithium ethylene dicarbonate. This equation represents one proposed pathway, not a complete description of carbonate decomposition. Competing reactions can produce lithium carbonate, lithium alkoxides, other alkyl carbonates, oligomeric compounds, and gases. Subsequent chemical reactions can alter products that were initially formed electrochemically.
Salt chemistry is equally important. Lithium hexafluorophosphate, commonly used in carbonate electrolytes, participates in reactions that generate lithium fluoride and phosphorus-containing species. Its decomposition chemistry is sensitive to trace water, which can promote formation of acidic products, including hydrogen fluoride. Fluorinated salts and additives can also supply lithium fluoride through reduction pathways. Consequently, observing lithium fluoride does not by itself establish which precursor or reaction sequence produced it.
The electrolyte’s coordination chemistry affects which molecules approach the electrode and how they react. In relatively dilute solutions, solvent molecules often dominate the immediate environment of lithium ions. Increasing salt concentration can place more anions in the first coordination shell, changing interfacial populations and favoring anion-derived products. This principle underlies many concentrated and localized high-concentration electrolytes. The latter use a diluent intended to alter bulk handling properties without extensively disrupting the concentrated electrolyte’s local solvation structure.
Composition is strongly history-dependent. An initial precipitate can undergo further reduction, react with dissolved species, dissolve, or become buried beneath later deposits. Hydrolysis and reactions during air exposure can also change a sample after removal from a cell. For this reason, a list of detected compounds is not necessarily a record of the original formation sequence. Distinguishing primary products from later transformations is a major task in reconstructing the SEI’s reaction mechanisms.
Nanoscale structure and selective transport
The SEI is often pictured as an inorganic-rich inner region near the electrode and a more organic-rich outer region adjoining the electrolyte. This arrangement is useful because strongly reduced, relatively insoluble products may accumulate close to the electron source, while less reduced compounds and solvent-containing material occur farther away. However, it is not a universal layered architecture. Many observations instead show irregular domains, embedded crystallites, compositional gradients, and channels distributed across the film.
The mosaic model describes this lateral and three-dimensional heterogeneity. Lithium fluoride, lithium carbonate, lithium oxide, and organic material may occupy neighboring regions with different transport and mechanical properties. Boundaries between phases can matter as much as the phases themselves. Pores or cracks may admit liquid electrolyte, while compact regions impede further reaction. The structure therefore depends not only on which compounds exist, but also on their connectivity, orientation, and contact with the underlying electrode.
Effective passivation requires low electronic conductivity together with sufficient lithium-ion transport. Electrons can cross very thin regions by quantum tunneling, move through electronically active defects, or reach electrolyte exposed at cracks. As a compact insulating layer thickens, direct tunneling becomes strongly suppressed. Lithium ions, meanwhile, may migrate through point defects in inorganic compounds, disordered organic regions, grain boundaries, or electrolyte-filled pathways. The dominant route need not remain the same throughout cycling.
For a simplified homogeneous film, its ionic resistance can be represented as
where is thickness, is transport area, and is effective ionic conductivity. Real electrodes violate several assumptions behind this expression: thickness varies spatially, current paths are tortuous, and electrochemically active area is difficult to establish. Nevertheless, it captures why a thicker or less ionically conducting film can increase polarization and limit charging power.
The reputation of lithium fluoride illustrates the need to distinguish bulk properties from interphase behavior. LiF-rich interphases often accompany improved cycling in particular electrolyte systems, but bulk crystalline LiF is not an exceptionally fast lithium-ion conductor. Beneficial behavior can instead involve electronic insulation, chemical stability, nanoscale morphology, defect populations, and the distribution of LiF within a composite. A favorable correlation between LiF content and cycling performance does not demonstrate that lithium ions travel rapidly through perfect LiF crystals.
Mechanical properties add another dimension. An interphase must tolerate changes in electrode shape without developing persistent openings, yet remain attached to the substrate. Stiffness alone does not determine success: fracture resistance, adhesion, thickness, and the capacity of softer components to redistribute strain are also important. Transport and mechanics are coupled because a crack can simultaneously reduce local resistance, expose fresh reactive material, and redirect current toward a small area.
Interphases on graphite and silicon electrodes
Graphite stores lithium by intercalation between carbon layers. Its potential during much of charging is only a few tenths of a volt or less above the lithium-metal reference, well below the stability range of many organic electrolytes. The SEI is therefore essential to conventional graphite-based lithium-ion batteries. It limits continuing electrolyte reduction while allowing lithium to enter and leave the graphite structure.
Graphite surface orientation affects both storage and interphase growth. Edge regions provide important access routes for lithium insertion and have different reactivity from basal planes. Defects, particle coatings, surface oxygen groups, and manufacturing treatments further influence where reactions begin. The SEI is consequently not necessarily uniform over a single particle, much less throughout a porous electrode. Local differences in reaction rate can translate into uneven film thickness and current distribution.
Solvent selection was crucial to making graphite practical. Propylene carbonate can undergo co-intercalation with lithium into graphite under many conventional electrolyte conditions, followed by decomposition and structural damage. Ethylene-carbonate-containing formulations helped establish protective interphases compatible with reversible graphite cycling. This distinction is conditional rather than absolute: salt concentration, additives, graphite properties, and solvation chemistry can modify co-intercalation behavior. The relevant issue is the full electrolyte–electrode combination, not a solvent name considered in isolation.
Silicon presents a more severe mechanical challenge. It stores lithium through alloying rather than graphite-like intercalation and can undergo volume expansion of roughly 300% on deep lithiation, depending on the phases and reference conditions involved. The resulting deformation can fracture the SEI, exposing new silicon to electrolyte. Fresh film then forms, consuming additional lithium and electrolyte. Repetition produces an expanding, compositionally complex interphase and can contribute to particle isolation or loss of electrode cohesion.
Fluoroethylene carbonate, commonly abbreviated FEC, is widely studied as an additive or cosolvent for silicon-containing electrodes. Its reduction can promote formation of fluorine-containing inorganic products together with organic components that differ from those formed in additive-free carbonate electrolytes. Performance depends on concentration and operating conditions; beneficial initial formation does not mean that the additive is inexhaustible or that its decomposition has no adverse consequences. Gas evolution, additive consumption, and changes during extended cycling remain relevant.
Practical silicon–graphite electrodes combine these behaviors. Silicon content, particle size, carbon coatings, binder chemistry, porosity, and the voltage window all affect interphase demands. Restricting silicon’s lithiation depth can reduce mechanical strain, while electrode architecture can preserve contact as particles move. The SEI cannot be optimized independently of these variables: a film that performs well on a flat silicon surface may behave differently in a thick composite electrode with nonuniform electrolyte access and local mechanical constraints.
Lithium metal, sodium systems, and solid electrolytes
On lithium metal, the SEI forms on an electrode whose surface is repeatedly created and removed during deposition and dissolution. Unlike an intercalation host that largely retains its framework, the metal changes morphology directly as charge passes. Local variations in interphase thickness, ion transport, and substrate contact can concentrate current and promote uneven deposits. These deposits may have high surface area, increasing the amount of electrolyte exposed to reactive lithium.
During stripping, lithium can dissolve from beneath its surface film. The residual interphase may collapse, remain suspended, or lose contact with the current collector. Some metallic lithium then becomes electronically isolated and cannot be recovered during ordinary cycling. This material is often called “dead lithium,” although inactive lithium measured in an electrode can include both isolated metal and lithium bound in SEI compounds. Distinguishing these categories is necessary when assigning the causes of capacity loss.
An SEI can influence lithium dendrite formation, but no single descriptor such as stiffness or fluoride content guarantees smooth deposition. Current density, areal capacity, electrolyte supply, temperature, substrate geometry, and stack pressure also matter. A protective film may fail because it fractures, conducts ions nonuniformly, or loses contact during stripping. Conversely, favorable deposition can arise from coordinated control of interphase chemistry, electrolyte transport, and electrode architecture rather than from the film alone.
Related interphases form in sodium-ion batteries, particularly on hard-carbon negative electrodes. Sodium-containing decomposition products differ from their lithium analogues in solubility, crystal chemistry, and transport properties. Sodium alkyl carbonates and other products can dissolve or reorganize differently, affecting passivation. In ether-based systems, solvated-ion co-intercalation into graphite can also support storage under suitable conditions, showing that behavior considered destructive in one electrolyte system can become an intended mechanism in another.
Solid electrolytes do not eliminate interphase formation. In a solid-state battery, contact between lithium and a sulfide or oxide electrolyte can generate a chemically distinct reaction layer. A simplified reduction of the sulfide electrolyte may be written as
Actual products and reaction extent depend on composition and conditions. If the products conduct lithium ions but block electrons, decomposition can become self-limiting. If they provide connected electronic pathways, reaction may continue into the solid electrolyte. Solid–solid contact adds mechanical complications: void formation, interfacial stress, and imperfect physical contact can limit performance even when the reaction products themselves appear chemically favorable.
Growth, ageing, and links to cell safety
SEI formation consumes charge that is not recovered as useful discharge capacity. In a full lithium-ion cell, lithium immobilized in interphase compounds is generally removed from the finite inventory available for reversible shuttling between electrodes. This contributes to first-cycle irreversible capacity and, when growth continues, to long-term capacity loss. Electrolyte consumption can also reduce wetting and ion transport in porous electrodes, particularly in cells designed with little excess liquid.
A simplified measure of charge reversibility is coulombic efficiency,
for a specified cycling protocol. Small persistent inefficiencies can accumulate over many cycles, but a measured inefficiency cannot automatically be assigned entirely to SEI growth. Lithium plating, positive-electrode reactions, redox shuttles, and changes in the accessible electrode inventory may also contribute. Accurate attribution requires complementary measurements rather than charge accounting alone.
During storage, SEI growth can continue even without external current because internal parasitic reactions remain possible. Elevated temperature generally accelerates many of these reactions, while electrode potential affects their driving force and accessible reaction pathways. Dissolved transition-metal species from the positive electrode may migrate to the negative electrode and alter interphase chemistry. This “cross-talk” means that negative-electrode degradation can originate partly in reactions elsewhere in the cell.
Models of gradual SEI growth often produce an approximately square-root-of-time dependence when transport through an increasingly thick layer controls the rate:
Here is initial thickness and is an effective growth parameter under specified conditions. Such behavior can arise from more than one microscopic mechanism, so agreement with the time law does not identify the transported species. Cracking, dissolution, changing temperature, and cycling-induced surface creation can produce different growth histories, including renewed rapid reaction.
The SEI also interacts with lithium plating during fast charging or low-temperature operation. Increased interphase resistance contributes to polarization, and local negative-electrode potentials can become favorable for metallic lithium deposition. Plated lithium creates additional reactive surface and can produce further interphase growth. This feedback links transport limitations, lithium inventory loss, and morphological damage rather than constituting one isolated failure process.
At elevated temperatures, SEI components may decompose or lose their protective function, enabling stronger reactions between lithiated electrode material and electrolyte. These processes can contribute heat and gas during the sequence leading to thermal runaway. Their onset and magnitude depend on electrode state of charge, electrolyte formulation, ageing history, and cell construction. No single SEI decomposition temperature or composition provides a universal measure of whole-cell safety.
Experimental characterization and computational models
Characterizing the SEI is difficult because it is thin, heterogeneous, chemically reactive, and often altered by the act of measurement. Removing an electrode from a cell changes its environment; washing can dissolve or redistribute components, while drying can collapse solvent-containing structures. Exposure to air introduces water and oxygen. Consequently, sample preparation and transfer conditions are part of the experiment rather than merely preliminary handling details.
X-ray photoelectron spectroscopy is widely used to identify elemental composition and chemical environments near the surface. It can distinguish classes of carbon-, oxygen-, fluorine-, and phosphorus-containing species, but overlapping spectral features make some assignments nonunique. Depth profiling by ion sputtering can expose buried material while also decomposing organics or changing oxidation states. Time-of-flight secondary-ion mass spectrometry provides sensitive fragment and spatial information, although fragmentation and matrix effects complicate quantitative interpretation.
Microscopy reveals morphology and local structure. Atomic force microscopy can track changes in surface roughness and measure local mechanical response, including in liquid environments. Electron microscopy offers high spatial resolution but can damage beam-sensitive materials. Adaptations of cryo-electron microscopy to battery materials, demonstrated prominently in 2017, enabled observations of lithium metal and associated interphases under conditions that better preserve sensitive structures. Freezing reduces some artifacts but does not eliminate preparation, sampling, or beam-induced changes.
Complementary chemical methods are essential. Solid-state nuclear magnetic resonance can identify environments of lithium and organic nuclei, while solution measurements can detect soluble products. Infrared spectroscopy and Raman spectroscopy probe molecular vibrations. Gas analysis identifies volatile decomposition products, and neutron- or X-ray-based methods can reveal averaged structural features. Operando measurements, performed while a cell functions, help connect chemical changes to potential and time, although specialized cell designs may differ from commercial configurations.
Electrical impedance spectroscopy measures the cell’s response to small alternating perturbations over a range of frequencies. It is frequently used to estimate interphase-related resistance, but assigning each feature to a unique physical process is difficult. Charge transfer, pore transport, contact resistance, and distributed capacitance can overlap. Equivalent circuits are useful representations only when constrained by physical understanding and independent evidence; a fitted resistance is not automatically a direct measurement of SEI thickness.
Computational approaches range from quantum-chemical calculations of initial electron-transfer reactions to molecular dynamics, kinetic reaction networks, and continuum growth models. Each accesses different scales. Short simulations can reveal plausible elementary steps without establishing their importance over months of cycling; continuum models can reproduce long-term trends without uniquely identifying chemistry. Productive comparisons therefore connect predicted reactions, experimentally detected species, transport measurements, and observed ageing under matched conditions.
Electrolyte engineering and industrial formation
SEI engineering aims to create a persistent, ion-permeable barrier with limited irreversible consumption of active materials. Electrolyte additives are one route. Vinylene carbonate is commonly investigated and used because it can react preferentially and contribute oligomeric or polymeric products. FEC is especially important in silicon research. Other additives target salt decomposition, acidity, gas generation, or particular inorganic components. Their effects depend on the complete formulation: combining individually beneficial additives does not necessarily produce a beneficial mixture.
Concentrated electrolytes provide another approach by changing solvation and interfacial reaction pathways. Greater anion participation can favor salt-derived interphases, but high viscosity, cost, wetting, and low-temperature transport may offset advantages. Localized high-concentration formulations seek to retain favorable coordination environments while reducing some bulk-property penalties. Their behavior still depends on whether the diluent remains sufficiently nonreactive and on how the solution reorganizes near a polarized electrode.
Artificial interphases can be introduced before cell assembly by chemical treatment, polymer coating, atomic-layer deposition, or other surface-processing methods. Such layers may reduce direct electrolyte contact, distribute ion flux, or reinforce the natural reaction film. The relevant question is not simply whether a coating is present, but whether it remains adherent and ionically accessible after electrolyte exposure and cycling. On expanding silicon or stripping lithium, even a chemically stable coating can fail mechanically or lose contact with the underlying active material.
Manufacturing uses a controlled formation process to establish the initial interphases. After electrolyte filling and wetting, cells undergo prescribed charging, resting, and sometimes discharging sequences. Current, voltage limits, temperature, and rest periods influence how quickly reactive components are consumed and how uniformly films develop across porous electrodes. Gas generated during formation may require removal in cell formats that permit a degassing step. Formation is therefore both an electrochemical conditioning operation and a substantial production-time and quality-control consideration.
Compensating for irreversible lithium loss is another strategy. Prelithiation introduces additional lithium into an electrode or cell before normal use, helping offset losses associated with SEI formation and other irreversible processes. It does not by itself create a stable interphase: continuing parasitic reactions can consume the added inventory as well. Its practical implementation must also account for lithium distribution, compatibility with manufacturing, and control of highly reactive materials.
Assessment under realistic cell conditions is crucial. An electrolyte may perform well with excess liquid, a thick lithium-metal counterelectrode, or a very thin active-material coating yet fail when electrolyte quantity and lithium inventory are restricted. Relevant comparisons specify electrode loading, areal capacity, temperature, pressure, charging protocol, and test duration. Manufacturing-compatible SEI design consequently couples surface chemistry to cell architecture, material utilization, and the reproducibility of the formation process.