How Lithium-Ion Batteries Work (and Why They Catch Fire)

The battery in your pocket right now is doing something remarkable: it is shuttling lithium ions back and forth between two electrodes hundreds of thousands of times over its lifetime, generating electricity with no moving parts and almost no waste heat — until, occasionally, it catastrophically fails. Understanding how lithium-ion batteries work is not just academic curiosity. It is the key to understanding why your EV range fades in winter, why your phone degrades after two years, and why a single punctured cell can level a warehouse.

The chemistry is elegant, the engineering is brutal, and the failure mode is one of the most dramatic in all of consumer electronics. A thermal runaway event is not simply a fire — it is a self-accelerating exothermic cascade that can reach temperatures high enough to melt aluminum and release enough oxygen to sustain its own combustion even in a sealed enclosure.

This post covers the complete picture: what is physically inside a cell, how charge and discharge actually work at the ion level, what the SEI layer is and why researchers lose sleep over it, how degradation happens, the step-by-step chemistry of thermal runaway, and where solid-state batteries fit into the safety equation.

What Is Actually Inside a Lithium-Ion Cell

A lithium-ion cell is a surprisingly simple stack of four functional layers, repeated and wound or folded into whatever form factor the device needs. Every rechargeable lithium cell — whether cylindrical 18650, prismatic pouch, or the flat cell in your laptop — is built from the same components.

Figure 1: A lithium-ion cell is a layered stack: aluminum current collector, cathode, microporous separator, anode, and copper current collector, all soaked in liquid electrolyte.

The Cathode: Where Chemistry Gets Political

The cathode is the positive electrode, and it is the most commercially contentious component in the entire battery industry. It is a porous metal oxide that can host — “intercalate” — lithium ions within its crystal lattice without permanently changing its structure. The two dominant chemistries today are NMC (nickel-manganese-cobalt oxide, in various ratios) and LFP (lithium iron phosphate).

NMC offers higher energy density, which is why it dominates electric vehicles where range matters most. The catch is that cobalt is expensive, geopolitically concentrated, and the cathode becomes chemically unstable at high temperatures — a detail that becomes critical later when we discuss thermal runaway.

LFP uses iron and phosphate instead of nickel, manganese, and cobalt. Its energy density is lower, but its crystal structure is significantly more thermally stable. LFP cathodes do not release oxygen when they decompose — a safety advantage that is hard to overstate. Tesla’s standard-range vehicles use LFP for exactly this reason; many grid-scale storage systems do too.

The cathode is coated onto an aluminum foil current collector, which conducts electrons to the positive terminal.

The Anode: Graphite Doing Heavy Lifting

The anode is the negative electrode, and in the vast majority of commercial cells it is graphite — the same material as pencil lead. Graphite has a layered structure that lets lithium ions slide in between the carbon sheets during charging, a process called intercalation. When the cell discharges, those same ions slide back out and travel to the cathode.

Graphite is cheap, abundant, and mechanically stable through thousands of cycles — which is why it has outlasted decades of research into alternatives like silicon anodes (which offer much higher capacity but expand significantly during lithium uptake, causing mechanical stress and shorter cycle life). Some modern cells blend small amounts of silicon into the graphite for a capacity boost, but pure silicon anodes remain a work in progress.

The anode is coated onto a copper foil current collector, which conducts electrons to the negative terminal.

The Separator and Electrolyte: The Unsung Heroes

The separator is a thin sheet of microporous polymer — typically polyethylene or polypropylene — with pores small enough to block electrode particles but large enough to let lithium ions pass freely. It is electrically insulating. Its sole job is to keep the cathode and anode from touching while allowing ionic conduction. If it fails, you get an internal short circuit.

The electrolyte fills every pore in the cell: cathode, separator, and anode. Standard electrolytes are lithium salts (most commonly lithium hexafluorophosphate, LiPF₆) dissolved in a mixture of organic carbonates — flammable solvents that are excellent lithium-ion conductors but poor fire extinguishing agents. The flammability of the electrolyte is not an accident of poor design; it is a property of the only solvents that work well enough at room temperature to make practical cells. Researchers have been trying to replace them with something non-flammable for decades.

How Charge and Discharge Actually Work

Understanding how lithium-ion batteries work at the electron-and-ion level removes most of the mystery — and most of the misconceptions.

Figure 2: During charging, lithium ions leave the cathode, cross the electrolyte, and intercalate into the anode. During discharge, the reverse happens — ions return to the cathode while electrons power the external load.

Discharge: Chemistry Making Electricity

When you connect a load — a motor, a circuit board, a light — the following happens simultaneously:

At the anode, lithium atoms release their electrons and become lithium ions (Li⁺). The electrons cannot travel through the electrolyte (it is an electronic insulator), so they are forced through the external circuit — your device. That electron current is the electricity you are using.

The Li⁺ ions travel through the electrolyte to the cathode. At the cathode, they intercalate back into the metal oxide crystal lattice, and the electrons arriving from the external circuit neutralize them. The cathode material is reduced; the anode is oxidized. This is a classic electrochemical cell.

The voltage of the cell — roughly 3.6–3.7 V nominal for NMC/graphite, around 3.2–3.3 V for LFP/graphite — comes from the difference in electrochemical potential between the two electrode materials. You cannot change this voltage by adding more material; you change it by choosing different chemistries. To get higher voltages, you stack cells in series.

Charging: Reversing the Reaction

Apply a voltage higher than the cell’s open-circuit voltage and the reaction reverses. The charger forces electrons through the external circuit from the cathode side to the anode side. Li⁺ ions are expelled from the cathode, cross the electrolyte, and intercalate into the graphite anode. The anode is being “loaded” with lithium, like a spring being compressed.

This is why a fully charged lithium cell is, in a sense, a metastable system. The lithium is sitting in the graphite at a high electrochemical potential, ready to flow back. Under normal conditions, the SEI layer keeps it there safely. Under abnormal conditions, that energy releases very quickly.

Why Voltage Curves Matter

The voltage of a lithium cell is not constant throughout discharge. NMC cells show a relatively smooth, sloping discharge curve. LFP cells have a famously flat discharge curve — the voltage barely moves until the cell is nearly depleted. This is operationally useful (your device sees stable voltage for most of the run time) but creates a measurement challenge: because voltage does not correlate well with state of charge for most of the range, battery management systems for LFP packs rely more heavily on coulomb counting (tracking current in and out) rather than voltage alone. This is part of why LFP state-of-charge estimation has historically been harder to get right. Modern IoT-connected EV battery telematics systems address exactly this problem with real-time data pipelines.

The SEI Layer: The Most Important Thing Nobody Talks About

The SEI — solid-electrolyte interphase — is a thin, chemically complex film that forms on the surface of the graphite anode during the first charge cycle. It is not designed or deposited deliberately. It grows spontaneously from reactions between the electrolyte and the anode surface, and it is both the reason lithium-ion batteries are practical and one of the main reasons they age.

Figure 3: The SEI forms during the first charge cycle. A stable SEI passivates the anode surface, blocking further electrolyte decomposition while remaining permeable to Li⁺ ions. Over hundreds of cycles it grows thicker, consuming lithium and raising cell impedance.

Why the SEI Has to Exist

The problem is thermodynamic: the electrochemical potential of lithium in a fully charged graphite anode is low enough that the organic carbonate solvents in the electrolyte are thermodynamically unstable against the electrode. Left to equilibrium, the electrolyte would continuously react with the anode, decomposing and producing carbonate species, gases, and other byproducts.

The SEI forms precisely because of this instability. In the first cycle, electrolyte molecules are reduced at the anode surface, forming a mosaic of lithium carbonate, lithium fluoride, lithium alkyl carbonates, and other compounds. The crucial property of a good SEI is that it is ionically conductive (letting Li⁺ pass through) but electronically insulating (stopping electrons from flowing into the electrolyte and driving further decomposition). A well-formed SEI is self-limiting: it grows thick enough to block electronic contact, then stops.

The SEI formation consumes some lithium permanently — this is the “formation loss” that every battery engineer accounts for. It is also why cells are pre-cycled before shipment.

How the SEI Ages

Every charge-discharge cycle puts the graphite anode through a slight volume change as lithium intercalates and deintercalates. This mechanical stress cracks the SEI in places. Fresh electrolyte contacts the exposed graphite, re-reacts, and new SEI material fills the crack — consuming more lithium and thickening the layer.

Over hundreds or thousands of cycles, this cumulative lithium loss is one of the primary contributors to capacity fade. The thicker SEI also has higher resistance, which is why old cells charge more slowly and show higher internal resistance on measurement. Research cited in the journal Nature Energy (Peled & Menkin, 2017, widely referenced in subsequent literature) identified the SEI as one of the central unresolved challenges in battery science — understanding its exact atomic-scale structure remains an active research area.

SEI Thermal Sensitivity

A stable SEI that is perfectly adequate at room temperature becomes a liability at elevated temperatures. Above roughly 80–120 °C (this threshold varies by electrolyte formulation and SEI composition), the SEI begins to decompose exothermically. That exothermic reaction generates heat, which raises the temperature further. And that is where thermal runaway begins.

Degradation and Aging: Why Batteries Die

Every lithium-ion battery ages. The question is how fast, and under what conditions.

Capacity Fade: The Slow Drain

Capacity fade — the gradual loss of how much charge a cell can hold — has several concurrent mechanisms. The SEI growth described above consumes lithium. Cathode degradation (transition metal dissolution, structural changes in the oxide crystal) removes active material. Anode porosity changes and pore clogging increase the difficulty of getting lithium ions in and out at high rates.

Heat accelerates every one of these mechanisms. A cell that lasts a decade at room temperature may degrade significantly faster if stored or operated at elevated ambient temperatures. This is why laptop batteries stored at high temperature degrade faster than those kept cool, and why EV battery warranties include provisions about charging in extreme heat.

Dendrite Formation: The Safety Time Bomb

When a lithium cell is charged too fast or at too low a temperature, or when the anode surface is poorly formed, lithium can deposit on the anode surface as metallic lithium rather than intercalating cleanly into the graphite. These metallic deposits grow as branching filaments called dendrites.

Dendrites are dangerous for two reasons. First, they can break off and float in the electrolyte as “dead lithium,” permanently reducing capacity. Second, if a dendrite grows long enough to pierce the separator, it creates a direct electronic connection between the cathode and anode — an internal short circuit. Dendrites are the failure mechanism behind several high-profile EV fires involving cells that were repeatedly fast-charged in cold conditions.

The physics of precision and signal propagation in fast-switching systems parallels the dendrite problem in a curious way: in both cases, tiny spatial deviations — femtoseconds of timing error, micrometers of lithium dendrite growth — accumulate over time into system-level failure.

Thermal Runaway: The Exothermic Cascade

Thermal runaway is not simply a battery fire. It is a self-sustaining, self-accelerating exothermic chain reaction that, once initiated beyond a certain threshold, cannot be stopped by conventional means. Understanding its mechanism is essential for engineers designing battery systems and for anyone who handles lithium cells at scale.

Figure 4: Thermal runaway proceeds through a sequence of chemical thresholds. Each step releases heat that drives the next. The feedback loop (heat spreading to adjacent cells) makes it especially difficult to contain in a multi-cell pack.

Step 1: The Trigger

Thermal runaway can be initiated by several distinct triggers:

Overcharge. Forcing lithium into the anode beyond its intercalation capacity causes lithium plating on the anode surface and can oxidize the electrolyte at the cathode side. Both reactions are exothermic and generate gases. BMS systems exist specifically to prevent this.

Physical damage (crush or penetration). A nail penetration test — the standard abuse test for battery safety certification — forces the separator to fail mechanically, creating an immediate internal short. The resistance of the short determines how quickly energy dissipates and whether thermal runaway follows.

External heat. A cell that is simply heated externally — say, in a fire, or by a neighboring cell in runaway — will eventually reach the SEI decomposition temperature.

Manufacturing defects. Metallic particle contamination from the manufacturing process can create micro-shorts. This is the trigger behind several recall events — a particle that causes no problems initially grows into a short as the cell cycles.

Accelerated Rate Calorimetry (ARC) testing, used by researchers at institutions including NREL’s National Center for Photovoltaics and battery science groups at Argonne National Laboratory, characterizes these onset temperatures precisely by heating a cell in an adiabatic environment and measuring self-heating rate as a function of temperature.

Step 2: SEI Breakdown

The first chemical threshold in the cascade is SEI decomposition. The reactions are exothermic, releasing heat that raises the cell’s internal temperature. As temperature rises, the rate of SEI decomposition accelerates (Arrhenius kinetics — the reaction rate roughly doubles with every 10 °C rise in this regime). The cell is now generating heat faster than it can dissipate it.

Electrolyte decomposition begins in parallel. The organic carbonates start to react with the electrodes and with each other, generating flammable gases including ethylene, methane, carbon monoxide, and carbon dioxide. These gases build pressure inside the cell casing.

Step 3: Separator Failure

At temperatures in the range of roughly 130–150 °C for common polyethylene/polypropylene separators, the polymer melts and the pore structure collapses. The separator may also shrink physically, exposing areas where cathode and anode are separated only by electrolyte. A polyolefin separator with a ceramic coating (aluminum oxide or boehmite) can raise this threshold and buy time — but not eliminate the risk.

Once the separator fails mechanically, the cell has an internal short circuit. The internal resistance of this short is typically very low, so a large current flows. Joule heating from the short adds yet more heat to the system.

Step 4: Cathode Decomposition and Oxygen Release

This is the step that makes NMC cells particularly dangerous compared to LFP. At temperatures above roughly 200–300 °C (varying by NMC stoichiometry — higher-nickel variants decompose at lower temperatures), the cathode oxide lattice becomes unstable. The metal oxide breaks down, releasing oxygen gas within the cell.

Now you have three things present simultaneously inside the cell: high temperature, flammable gases from electrolyte decomposition, and a supply of gaseous oxygen. You have all three elements of the fire triangle, generated internally. This is why a lithium cell fire is so difficult to extinguish: the oxidizer is being produced by the cell itself, not drawn from the surrounding air.

LFP cathodes do not undergo this oxygen-releasing decomposition. The iron-phosphate crystal structure is far more thermally stable, and iron in its reduced form does not release oxygen in the same way. This is the fundamental safety advantage of LFP.

Step 5: Fire, Venting, and Propagation

The cell vents — either through a designed pressure relief vent or catastrophically through the casing — releasing a jet of hot, burning gas. The event is sometimes accompanied by an audible pop and a characteristic sharp chemical odor. Ejected electrolyte vapors and electrode particles can ignite.

The worst outcome in a multi-cell pack is propagation: the heat from the failing cell raises the temperature of adjacent cells past their SEI decomposition threshold, initiating their own runaway. This cell-to-cell propagation is the mechanism behind the large-scale fires seen in EV battery packs and stationary storage systems. Pack design and thermal management are specifically engineered to slow or stop this propagation — but arresting it entirely once it has started remains extremely difficult. Architectures for connected monitoring in critical physical systems like maglev train control face analogous cascading failure design problems.

Trade-offs, Gotchas, and What Goes Wrong

The Energy-Density / Safety Trade-off Is Real, Not Marketing

Higher nickel content in NMC cathodes raises energy density but lowers the thermal decomposition onset temperature. NMC 811 (80% nickel) is denser than NMC 532 but decompose at a lower temperature. There is no chemistry today that gives you both maximum energy density and the thermal stability of LFP. Anyone claiming otherwise is selling something.

BMS Limitations Under Fast-Charging Conditions

Battery management systems monitor cell voltage, temperature, and current to prevent overcharge and over-discharge. The challenge is measurement latency and spatial resolution. A BMS measures temperature at a small number of probe points in a large pack. Internal temperature can rise significantly faster than the probe readings indicate, especially during rapid charging. By the time a probe registers a dangerous temperature, the cell may already be past the SEI decomposition threshold. This is not a software problem — it is physics and sensor placement.

Solid-State Batteries: Not a Magic Bullet Yet

Solid-state batteries replace the liquid electrolyte with a solid ionic conductor — ceramics like LLZO (lithium lanthanum zirconium oxide) or sulfide-based conductors. The appeal is clear: eliminate the flammable solvent, and you eliminate one of the key ingredients in thermal runaway. Solid electrolytes also suppress dendrite growth more effectively in some configurations.

The gotchas: solid electrolytes have lower ionic conductivity than liquids at room temperature, leading to worse rate capability. The mechanical interface between a solid electrolyte and an electrode that changes volume during cycling is hard to maintain. Manufacturing at scale remains expensive. And critically, while a solid-state battery cannot burn due to flammable liquid electrolyte, cathode oxygen release is still a mechanism at high temperatures — so the risk is reduced, not eliminated. Most industry timelines for high-volume solid-state EV cells have repeatedly slipped.

Capacity Numbers Are Best-Case

Rated capacity is measured at room temperature, at slow charge/discharge rates, when the cell is new. Cold temperature dramatically reduces available capacity (lithium-ion kinetics slow down). High discharge rates reduce effective capacity. Aging reduces it further. A battery rated at a given kilowatt-hour at cell level delivers meaningfully less usable energy at the system level after accounting for pack overhead, thermal management, and aging margins built into the state-of-charge window.

Practical Recommendations

For engineers, product managers, and informed users who work with lithium-ion systems:

Choose chemistry for the application. LFP for stationary storage and applications where weight is not the constraint. NMC or NCA where energy density is paramount and the BMS and thermal management are engineered accordingly.
Never charge at low temperatures without a pre-warming step. Lithium plating onset is real and cumulative. EVs that allow fast charging below ~10 °C without first warming the pack are trading long-term health for short-term convenience.
Storage state of charge matters. Storing cells at 100% state of charge accelerates cathode degradation. Storing at 0% risks copper dissolution at the anode. Long-term storage at 40–60% SoC is the established recommendation.
Monitor internal resistance, not just capacity. Rising internal resistance is often the earlier signal of aging and the precursor to thermal problems under load. Many BMS implementations report this; use the data.
Design thermal runaway containment, not just prevention. In any multi-cell pack, assume that a cell will eventually fail. The pack design must contain the event — venting paths, thermal barriers between cells, fire-suppressing materials — so that one cell failure does not become a pack-level event.
For IoT and connected systems, instrument the pack. Cell-level temperature and voltage monitoring with cloud-connected telemetry can catch anomalous behavior — a cell diverging from its neighbors in voltage or temperature — long before it becomes dangerous.

Frequently Asked Questions

Why do lithium-ion batteries catch fire?

Lithium-ion batteries catch fire through a process called thermal runaway. A trigger — overcharge, physical damage, external heat, or a manufacturing defect — starts an exothermic cascade: the SEI layer decomposes, electrolyte breaks down and releases flammable gases, the separator fails, and in NMC-chemistry cells the cathode oxide releases oxygen. With heat, fuel, and oxygen all generated internally, the cell sustains its own combustion even in a sealed environment.

Is LFP safer than NMC?

Yes, significantly so in terms of thermal runaway risk. LFP (lithium iron phosphate) cathodes do not release oxygen when they thermally decompose, and their decomposition onset temperature is substantially higher than NMC’s. This eliminates one of the key steps in the thermal runaway cascade. LFP cells can still vent and burn if severely abused, but the risk threshold is considerably higher, and the severity of failure tends to be lower.

What is the SEI layer and why does it matter?

The SEI (solid-electrolyte interphase) is a nanometer-scale film that forms spontaneously on the graphite anode during the first charge cycle. It forms because the anode’s electrochemical potential is low enough to reduce the electrolyte solvents. A well-formed SEI passivates the anode, blocking continuous electrolyte decomposition while letting lithium ions pass. Without it, the electrolyte would continuously react with the anode. Over time, the SEI grows thicker with each cycle, consuming lithium and raising impedance — two of the main contributors to battery aging.

Why does fast charging degrade batteries?

Fast charging pushes lithium ions into the anode at a higher rate than they can comfortably intercalate into the graphite crystal structure. If the rate is too high — especially at low temperatures where ion mobility is reduced — lithium deposits as metallic dendrites on the anode surface rather than intercalating cleanly. Each fast-charging session that causes lithium plating permanently reduces capacity and increases the risk of a dendrite eventually piercing the separator. BMS systems on quality chargers adjust current based on temperature and state of charge to minimize this, but they cannot eliminate the physics.

How do solid-state batteries improve safety?

Solid-state batteries replace the flammable liquid electrolyte with a solid ionic conductor (ceramic or polymer-based). Eliminating the organic carbonate solvents removes the main source of flammable gas in thermal runaway and also suppresses dendrite growth more effectively in many formulations. However, cathode oxygen release remains a potential failure mechanism at extreme temperatures, solid electrolytes have lower ionic conductivity than liquids at room temperature, and the manufacturing challenges are substantial. They are safer in important respects, but not unconditionally safe.

How long do lithium-ion batteries last?

Under normal use conditions, most modern lithium-ion cells are designed for several hundred to over a thousand full charge-discharge cycles before capacity drops to roughly 80% of original capacity — the common end-of-useful-life threshold. The actual number varies enormously with chemistry (LFP typically shows better cycle life than NMC at deep discharge), depth of discharge (shallower cycles are much easier on the cell), temperature (heat is the primary accelerant of aging), and charge rate. EV battery packs are engineered with wider safety margins and more sophisticated thermal management than consumer electronics, which is why their packs typically outlast the vehicles they are in.