How Lithium-Ion Batteries Actually Work (and Degrade)

Here is the part nobody tells you: how lithium-ion batteries work has almost nothing to do with electricity being “stored” inside them like water in a tank. A charged lithium-ion cell stores energy as a position. Lithium atoms are parked in the wrong place — wedged into a graphite sponge they would rather not be in — and the entire battery is a machine for letting them crawl back to a place they prefer, harvesting electrons on the way. That single idea explains why your phone lasts a day, why it lasts noticeably less after three years, and why a punctured cell can erupt into a fire that water makes worse.

The same chemistry that makes these cells astonishingly good also guarantees they slowly poison themselves and, in rare cases, catastrophically self-destruct. Understanding both halves takes you from “battery magic” to actual mechanism.

What this covers: the intercalation mechanism, the role of electrolyte and separator, charge versus discharge, the SEI layer, why cells degrade, the chain reaction of thermal runaway, and longevity tips that follow directly from the chemistry.

Context and Background

The lithium-ion cell that dominates 2026 — in phones, laptops, electric vehicles, and grid storage — is a refinement of a design commercialized by Sony in 1991. Its inventors, John Goodenough, M. Stanley Whittingham, and Akira Yoshino, shared the 2019 Nobel Prize in Chemistry for it. The chemistry has been polished relentlessly since, but the core architecture has not changed: two host materials that can soak up and release lithium, separated by a liquid that carries lithium ions but blocks electrons.

Before lithium-ion, portable electronics ran on nickel-cadmium and nickel-metal-hydride cells. Those were heavier, lower-voltage, and plagued by “memory effect.” Lithium is the lightest metal and the most electropositive, which is exactly why it is attractive: it gives the highest possible voltage and energy per gram. The catch is that pure lithium metal is dangerously reactive, so the breakthrough was learning to use lithium ions shuttling between two stable hosts rather than plating and stripping raw lithium metal.

That trick — called intercalation — is the whole game. Two competing cathode families dominate today. Nickel-manganese-cobalt (NMC) and nickel-cobalt-aluminum (NCA) oxides chase maximum energy density for EVs and laptops. Lithium iron phosphate (LFP) trades some energy density for lower cost, longer life, and far better safety, and has taken over a large share of standard-range EVs and almost all stationary storage. If you want to see how these cells reach for their physical ceiling, the related work on how solid-state batteries work covers the next frontier. The chemistry below is also documented in depth by the U.S. Department of Energy’s battery basics.

It helps to fix the scale early. A modern cylindrical 21700 cell — 21 mm across, 70 mm tall — holds roughly 15–20 watt-hours, enough to power a laptop for an hour or two. A typical EV pack chains thousands of such cells into a 60–100 kilowatt-hour battery. The energy comes from two numbers multiplied together: how much charge each gram of active material can move (its capacity, in milliamp-hours per gram) and the voltage at which it moves it. Graphite manages about 372 mAh/g at full lithiation; an NMC cathode delivers roughly 150–200 mAh/g. Multiply the limiting capacity by the ~3.7 V cell voltage and you get the energy. Every advance in the field is a fight to raise one of those two numbers without wrecking the cell’s lifespan or safety — which, as we will see, is exactly the trade the chemistry refuses to give away for free.

To make that arithmetic concrete: imagine an electrode coating that contains 10 grams of graphite. At 372 mAh/g that anode can store about 3,720 mAh, or 3.72 amp-hours. Pair it with a cathode sized to match, and at a nominal 3.7 V the cell holds 3.72 Ah × 3.7 V ≈ 13.8 watt-hours. Notice that the cathode is almost always the bottleneck: at 150–200 mAh/g it needs roughly two to two-and-a-half times as much mass as the graphite to supply the same charge, which is why a finished cell is dominated by cathode weight and why every fraction of a percent of extra cathode capacity is fought over so hard. The anode is generous with capacity; the cathode is stingy, and it sets the price.

How Lithium-Ion Batteries Work: Intercalation

A lithium-ion battery works by shuttling lithium ions (Li+) back and forth between two electrodes while their matching electrons travel the opposite way through your device. On discharge, lithium leaves the graphite anode, drifts through the electrolyte to the layered-oxide cathode, and the freed electrons power your circuit. Charging forces everything in reverse. No lithium is consumed — it is just relocated.

Figure 1: During discharge, lithium ions move internally from anode to cathode through the electrolyte while electrons take the external circuit, delivering current to the load.

The two electrodes are not solid blocks of metal. They are crystalline host structures with atomic-scale gaps designed to accept lithium guests. The negative electrode (anode) is typically graphite — stacked sheets of carbon, like a deck of cards, with room between the sheets. The positive electrode (cathode) is a layered transition-metal oxide such as NMC, with lithium sites woven into its crystal lattice. Sliding lithium in and out of these gaps without breaking the host is lithium-ion intercalation, and it is what makes the cell rechargeable.

The word “intercalation” is precise and worth keeping. The lithium does not bond into a new chemical compound the way it would in, say, lithium oxide; it simply slips into pre-existing vacancies between layers and is held there by relatively weak electrostatic forces. Because nothing in the host is broken or rebuilt, the process is highly reversible — which is the entire reason these cells survive hundreds or thousands of cycles. A battery chemistry that consumed its electrodes by forming and breaking strong bonds, like the old non-rechargeable lithium-metal cells, can only run one way. Intercalation is what turns a one-shot reaction into a pump you can run forwards and backwards.

The anode side: lithium hiding in graphite

When the cell is charged, lithium ions burrow between graphite’s carbon sheets and pick up electrons, forming a lithium-graphite intercalation compound. Fully charged graphite reaches roughly LiC6 — one lithium for every six carbon atoms. This happens at a very low potential, between about 0 V and 0.25 V versus a lithium-metal reference. That low potential is good for voltage but dangerous: push it slightly negative and lithium stops intercalating and starts plating as metal, a failure mode we return to later.

Graphite is not the only option. Silicon can hold roughly ten times more lithium per gram — its theoretical capacity is around 3,600 mAh/g versus graphite’s 372 — which is why modern high-energy cells blend small amounts of silicon into the graphite. The problem is that silicon swells by up to ~300% when lithiated and cracks itself apart, so engineers add it sparingly — usually a few percent — to gain energy density without destroying cycle life. The swelling is not a minor nuisance: each charge inflates the silicon particle and each discharge deflates it, and that breathing motion fractures particles, severs electrical contact within the electrode, and repeatedly tears open fresh surface that has to grow a new protective layer. A pure-silicon anode would self-destruct in tens of cycles for exactly this reason, which is why the industry advances silicon content one cautious percentage point at a time.

There is also a subtlety to how lithium fills graphite, called staging. Lithium does not spread evenly through all the carbon galleries at once; it fills them in ordered steps, occupying every fourth layer, then every third, then every second, before reaching the fully packed LiC6. Electrochemists number these arrangements: a “stage 4” compound has lithium in roughly every fourth gallery, stage 3 in every third, stage 2 in alternating galleries (about LiC12), and stage 1 is the fully packed LiC6. You can actually see this staging in a charge curve as a series of small voltage plateaus, each marking a transition between stages — the potential hangs almost flat while one ordered arrangement converts to the next, then steps to a new plateau. This ordered filling is part of why graphite is such a stable, reversible host — and why pushing it too hard, too fast disrupts the orderly process and invites trouble. When current is forced through faster than lithium can rearrange itself into the next stage, the surface saturates while the interior lags, and the local potential at the particle surface dives toward the plating threshold even though the cell as a whole looks far from full.

The cathode side: lithium’s preferred home

The cathode is where lithium “wants” to be, energetically. In a charged cell it has been stripped of much of its lithium; on discharge, lithium ions slot back into the oxide lattice and the transition metals (nickel, cobalt, manganese) accept the incoming electrons by changing oxidation state — nickel, for example, cycling between Ni3+ and Ni4+. This redox reaction at the cathode is what sets the cell’s voltage, typically 3.6–3.7 V nominal for NMC and a flatter 3.2 V for LFP.

The difference in lithium’s “comfort” between graphite and the oxide is the cell’s voltage. Energy density follows from voltage times the amount of lithium you can reversibly move: more nickel in the cathode means more capacity, which is why the industry keeps pushing toward high-nickel NMC811 (80% nickel) despite its tendency to crack and degrade faster.

There is a hard limit lurking here, and it is the reason you can never safely drain a cell to zero. The cathode lattice only stays intact if you leave some lithium in it. Pull out too much — by over-charging past the rated voltage — and the layered oxide structure starts to collapse, and high-nickel cathodes begin shedding oxygen. That is why a battery management system caps the upper voltage tightly, typically around 4.2 V per cell for NMC. The voltage window is not a comfort setting; it is the structural envelope inside which the crystal survives repeated cycling.

Why high-nickel cathodes are both prize and problem

It is worth dwelling on the nickel redox couple, because it captures the central tension of modern cell design. In a layered oxide, the lithium sites sit between sheets of transition-metal-and-oxygen octahedra. When you charge the cell and pull lithium out, the transition metals must give up electrons to balance the charge — nickel oxidizes from Ni3+ to Ni4+, and the more nickel you pack in, the more lithium you can extract per gram before the voltage runs away. That is the prize: NMC811 can deliver well over 200 mAh/g, where an older, cobalt-rich, low-nickel oxide delivered closer to 150.

The problem is that Ni4+ is a ferociously strong oxidizer and the highly delithiated lattice it leaves behind is unstable. Strip out too much lithium and the layered structure loses the “pillars” of lithium that hold its sheets apart; the lattice begins to contract and rearrange toward a disordered, rock-salt-like phase near the particle surface. That surface reconstruction is electrically resistive and traps lithium, raising internal resistance over hundreds of cycles. Worse, at high states of charge the lattice oxygen itself becomes liable to leave — releasing reactive oxygen that attacks the electrolyte and, in a thermal event, feeds combustion from inside the cell. This is the deep reason high-nickel chemistry is simultaneously the highest-energy and the least forgiving: the very condition that gives you more capacity (deep delithiation of a nickel-rich oxide) is the condition that destabilizes the crystal and primes the oxygen-release hazard. LFP sidesteps the whole dilemma — its iron-phosphate framework is covalently bonded and holds its oxygen tightly, which is why it barely degrades structurally and is far harder to push into oxygen release, at the cost of a lower voltage and less energy per kilogram.

A Deeper Dive: Electrolyte, Separator, and the SEI

Two ingredients make intercalation possible, and one unplanned ingredient makes it survivable. The electrolyte ferries lithium ions; the separator stops the electrodes touching; and the SEI — a layer nobody designs but every working cell grows — is the reason graphite cells last at all.

Figure 2: The internal sandwich of a lithium-ion cell. The microporous separator and liquid electrolyte sit between the two coated foils, letting ions pass while blocking electrons and physical contact.

The electrolyte: an ion highway that blocks electrons

The electrolyte is usually a lithium salt — most commonly lithium hexafluorophosphate, LiPF6 — dissolved in a mix of organic carbonate solvents like ethylene carbonate (EC) and dimethyl carbonate. Its job is precise: conduct Li+ ions freely while being an electronic insulator. If electrons could cross internally, the cell would simply short itself out and self-discharge instantly.

LiPF6 is chosen for a frustrating reason: nothing else is as good across all the axes that matter at once — high conductivity, a wide stable voltage window, and compatibility with the aluminum cathode foil. But it carries a notorious weakness. LiPF6 is in equilibrium with phosphorus pentafluoride, and in the presence of even trace water it hydrolyzes to produce hydrofluoric acid (HF). That HF is corrosive: it attacks the cathode surface, helps dissolve transition metals out of the oxide lattice, and degrades the SEI. This is why battery manufacturing is obsessive about drying — cells are assembled in dry rooms with single-digit dewpoints, because a few hundred parts per million of moisture is enough to seed a slow internal acid attack that shortens life. The carbonate solvents matter too because they define the electrolyte’s stability window. Reduce the voltage too far at the anode and the solvents decompose (that is what builds the SEI); oxidize them too far at a high-voltage cathode and they break down on the positive side instead, which is one of the practical ceilings on how high you can push cell voltage.

This electrolyte is also the cell’s weakest link thermally. Carbonate solvents are flammable and boil well below the temperatures reached during a fault. That is the seed of thermal runaway, and it is the price paid for a liquid that conducts lithium well at room temperature.

The separator: a referee that melts

The separator is a thin microporous plastic film — typically polyethylene or polypropylene, often a multilayer of both — perhaps 10–25 micrometers thick. Its pores let electrolyte and ions through while physically keeping anode and cathode apart. A single torn or punctured separator means the electrodes touch, creating an internal short circuit that dumps the cell’s energy as heat in milliseconds.

Many separators are engineered with a “shutdown” feature: one layer melts at a lower temperature to close its pores and choke off ion flow before things get dangerous, while a higher-melting layer holds mechanical integrity a little longer. A polyethylene layer softening near 130 °C, for example, can flow into and seal the pores to stop ion transport, while a polypropylene layer that melts closer to 165 °C keeps the film from collapsing entirely and shorting. It is a clever safety valve, but a limited one — once temperatures climb high enough, the whole separator fails, and in a fast internal short the heat can arrive faster than the shutdown can respond.

The SEI: the battery’s accidental skin

Here is where it gets subtle. Remember the anode operates near 0 V versus lithium — well outside the voltage window where the carbonate electrolyte is chemically stable. On the very first charge, the electrolyte in contact with the graphite decomposes. Instead of ruining the cell, those decomposition products precipitate into a thin film on the anode surface: the solid electrolyte interphase, or SEI.

The SEI is the most important accident in battery chemistry. It is a mosaic of lithium-bearing compounds — lithium carbonate (Li2CO3), lithium ethylene dicarbonate (LEDC), lithium fluoride, and organic species — and it has one magical property: it conducts lithium ions but blocks electrons. Once it forms, it acts as a passivating skin that stops the electrolyte from continuing to react with the anode. Without an SEI, graphite would shred the electrolyte on every cycle and the cell would die in days.

The layer has a structure as well as a composition. The portion nearest the graphite tends to be dense and inorganic — rich in lithium fluoride and lithium oxide, the species that block electrons most effectively — while the outer portion is more porous and organic. A good SEI is the one that gets this gradient right: a compact inorganic inner skin that shuts down electron leakage, topped by a flexible outer layer that can accommodate the graphite’s slight expansion and contraction during cycling without cracking. When the layer cracks, fresh graphite is exposed, the electrolyte reacts again, and the SEI thickens locally — which is precisely the mechanism that ages the cell over its life.

Figure 3: SEI formation. On the first charge the electrolyte reduces at the graphite surface, consuming lithium and depositing a passivating layer that protects the anode but permanently locks away some capacity.

This protection is not free. Building the SEI consumes lithium — pulled from the cathode’s “lithium inventory” — that is then locked away forever as solid film. That is why a brand-new cell permanently loses several percent of its capacity during the manufacturer’s “formation cycling,” and why the SEI’s slow lifelong thickening is the single biggest driver of calendar aging. The accident that saves the battery is also the one that ages it.

A useful way to picture the lifelong thickening is the square-root-of-time law. The SEI grows quickly at first and then ever more slowly, and to a good approximation its thickness grows in proportion to the square root of elapsed time. The reason is self-limiting kinetics: to react, electrolyte molecules or electrons have to get through the existing layer to reach fresh material, and the thicker the layer becomes, the longer that transport takes. So the very act of growing slows the growth — doubling the layer’s thickness roughly quadruples the time it took to get there. This is good news for longevity (a cell does not age at a constant rate; it settles down) and it explains the characteristic shape of a capacity-fade curve, which drops fastest in the first months and then flattens into a long, gentle decline.

Formation cycling itself is a surprisingly delicate, expensive manufacturing step. The first charge is done slowly and under careful temperature control, because the quality of that initial SEI sets the cell’s entire fate. A well-formed SEI is thin, uniform, and stable; a poorly formed one is patchy and keeps reacting, accelerating aging and even creating local hot spots. Battery makers guard their formation recipes closely, because two cells with identical materials can have very different lifespans depending purely on how that first skin was grown. The composition of the electrolyte matters here too — manufacturers add small amounts of “additives” such as vinylene carbonate (VC) and fluoroethylene carbonate (FEC) specifically to steer the SEI toward a more durable, lithium-conducting structure. These additives are chosen because they reduce at a slightly higher potential than the bulk solvents, so they decompose first and seed the SEI before the main solvents get a chance — VC tends to build a tougher, more flexible polymeric film, while FEC enriches the layer in lithium fluoride and is almost mandatory for silicon-containing anodes, whose violent volume swings demand an unusually resilient skin. The whole field of electrolyte engineering is, to a large degree, the art of growing a better accident.

Charge versus discharge: following the electrons

Putting it together, here is the full directional picture:

• Discharge (powering your device): Lithium leaves graphite, releasing an electron to the external circuit. The Li+ swims through the electrolyte to the cathode, while the electron does work in your phone and arrives at the cathode to balance the incoming ion. Lithium intercalates into the oxide. The anode is the negative terminal; current flows out of the positive cathode terminal.
• Charge (refilling): A charger forces electrons back into the anode and pulls lithium ions out of the cathode, driving them back into graphite against their energetic preference. You are paying energy to put lithium back in the “wrong,” high-energy place. That stored positional energy is what you draw on next time.

The cell never creates or destroys lithium; it only changes where the lithium sits and forces electrons to take the long way around. This is also why a “dead” battery is not empty of lithium — the lithium has simply migrated to the cathode and reached an equilibrium that no longer pushes electrons through your device.

The brain on top: the battery management system

A single cell is dumb; what makes a modern battery pack trustworthy is the battery management system (BMS) sitting on top of it. Because the chemistry is unforgiving at its edges, the BMS enforces the limits the crystals demand: it cuts charging when any cell reaches the upper voltage ceiling, stops discharge before any cell falls too low, and throttles current when temperatures stray outside a safe band — typically refusing to fast-charge below roughly 0 °C precisely to avoid lithium plating.

It also does something less obvious but essential: cell balancing. A pack of hundreds of cells in series is only as strong as its weakest cell, because charging must stop the moment any single cell hits the voltage limit. Tiny manufacturing differences mean cells drift out of step over time. The simplest and most common approach, passive balancing, bleeds off a little charge from the cells that run ahead — routing it through a small resistor as waste heat — so the laggards can catch up and the whole pack fills and empties evenly. More sophisticated active-balancing systems instead shuttle charge from the fuller cells into the emptier ones, wasting less energy but costing more in circuitry. Either way, without balancing the pack’s usable capacity would shrink to that of its worst cell, and the strongest cells would risk repeated over-voltage. The BMS also estimates two quantities you never see directly: state of charge (how full the pack is, inferred from voltage, current integration, and a model of the cell) and state of health (how much the pack has degraded relative to new). In other words, much of what we call “good battery behavior” is not the cell at all — it is the electronics keeping the chemistry inside its narrow safe envelope and quietly correcting for the differences between cells.

Trade-offs, Gotchas, and What Goes Wrong

Every mechanism above has a dark twin. The same low anode potential that gives high voltage invites lithium plating; the same flammable electrolyte that conducts ions well feeds fires; the same crystal that hosts lithium cracks when you cycle it hard.

Calendar versus cycle aging. Batteries die two ways at once. Calendar aging happens just sitting on a shelf, dominated by the SEI slowly thickening — faster when the cell is stored full and hot. Cycle aging is wear from charging and discharging: each cycle slightly expands and contracts the electrodes, cracking particles and exposing fresh surface that grows more SEI, consuming more lithium. The two combine, which is why a hot, always-full battery ages fastest of all.

The numbers are worth internalizing. A consumer lithium-ion cell is typically rated to retain 80% of its original capacity after 300–800 full cycles, while an LFP cell can manage 2,000–5,000. “End of life” by convention is that 80% mark — the cell still works, it just holds less. Crucially, SEI growth follows roughly a square-root-of-time law: it grows fast at first and then slows, because the thicker the layer gets, the harder it is for new electrolyte to reach the surface and react. Temperature, though, is multiplicative on top of that — a rough rule of thumb from reaction kinetics is that aging reactions roughly double in rate for every 10 °C of extra temperature. That single fact is why heat is the villain in nearly every longevity story.

It is worth seeing where that “doubles every 10 °C” rule comes from, because it is not arbitrary. The side reactions that consume lithium are chemical reactions, and chemical rates follow the Arrhenius relationship, in which rate rises exponentially with temperature. For the activation energies typical of SEI growth, that exponential works out to roughly a 2× speed-up per 10 °C over the relevant range. The practical consequence is dramatic: a cell aging gently at 25 °C will, all else equal, age something like four times as fast at 45 °C — two doublings — and that is before you account for the extra strain of being held at a high state of charge. This is why a phone left charging on a sunny dashboard, or an EV fast-charged repeatedly without active cooling, loses capacity so much faster than the spec sheet’s cycle count would suggest. The cycle count is quoted at a benign temperature; real life is rarely benign.

C-rate — the speed dial. Charge and discharge speed is measured in “C.” A 1C rate fully charges or discharges the cell in one hour; 2C in half an hour; 0.5C in two. High C-rates strain everything: ions have to move faster than the host structures comfortably accept, internal resistance turns the rushed current into heat, and the anode potential dips closer to the lithium-plating threshold. This is the underlying reason “fast charging is hard on the battery” is true rather than folklore — a high C-rate pushes the cell toward every failure mode at once.

The heat is not a vague hand-wave; it follows the I²R law, where a cell’s internal resistance R turns current I into heat at a rate proportional to the square of the current. That squaring is the crux. Take a cell with an internal resistance of 30 milliohms. At a modest 1 A it dissipates I²R = 1² × 0.03 = 0.03 watts — negligible. Push it to 10 A (a high C-rate for a small cell) and the heating leaps to 10² × 0.03 = 3 watts inside the same small volume, a hundredfold increase for a tenfold rise in current. That self-heating raises the cell’s temperature, which (via Arrhenius) accelerates aging, and the warmth also lowers internal resistance slightly, encouraging still more current — a feedback loop that designers must actively manage with cooling. The squared dependence is exactly why fast charging is disproportionately punishing rather than merely faster.

Lithium plating. If you charge a cell too fast, or while it is cold, lithium ions arrive at the graphite faster than they can intercalate. The anode potential dips below 0 V versus lithium and metallic lithium plates onto the surface instead of slotting inside. The thermodynamics are unforgiving: intercalation into graphite happens at a potential only a fraction of a volt above the lithium-metal plating potential, so there is very little margin. Anything that pushes the local anode potential down — high current, low temperature (which slows the lithium’s ability to move into the host), or an already-full electrode — can cross that line. This is doubly bad: plated lithium is “lost” capacity, much of it sealed off behind fresh SEI as “dead lithium,” and it can grow into needle-like dendrites that pierce the separator and trigger an internal short. Cold-weather fast charging is the classic trigger, because in the cold the solid-state diffusion of lithium into graphite slows dramatically while the charger still tries to push current in — the lithium has nowhere fast to go but onto the surface as metal.

Cathode cracking and transition-metal dissolution. High-nickel cathodes like NMC811 squeeze out more energy but are mechanically fragile. Repeated lithiation strains the oxide lattice until particles micro-crack, exposing new surface, raising internal resistance, and releasing transition-metal ions that drift to the anode and poison the SEI. Manganese and nickel ions that dissolve out of the cathode — a process the HF from electrolyte hydrolysis actively encourages — migrate across the cell and deposit on the anode, where they catalyze further electrolyte decomposition and force the SEI to keep regrowing. Fast charging accelerates all of it.

Thermal runaway — the catastrophic failure. When a cell overheats — from an internal short, overcharge, crush, or external fire — it can enter a self-sustaining exothermic cascade. The chain proceeds in stages: around 80–120 °C the SEI breaks down, exposing graphite to electrolyte; the anode-electrolyte reaction releases heat that pushes the cell hotter; the separator melts (typically ~130–160 °C) and causes a large internal short; then the cathode itself decomposes and releases oxygen, which feeds combustion of the flammable electrolyte from the inside. Each step raises the temperature and triggers the next — a runaway. Engineers track this with the concept of a self-heating onset temperature: the point above which the cell generates heat faster than it can shed it, so that even with the external trigger removed, the temperature keeps climbing on its own. Once a cell crosses that threshold, the rate of self-heating accelerates by orders of magnitude as each stage ignites the next, and peak internal temperatures can exceed several hundred degrees in seconds. This is why lithium fires are so hard to extinguish: the cathode supplies its own oxygen, so smothering does little, and adjacent cells cascade. Water cools but cannot stop the internal oxygen-fed reaction, which is why responders use lots of it mainly to cool neighboring cells and prevent propagation.

Figure 4: The thermal runaway cascade. Each stage releases heat that triggers the next, ending in cathode oxygen release that feeds an internal fire smothering cannot stop.

The peer-reviewed literature on these mechanisms, including a state-of-the-art review of lithium battery degradation and failure modes, confirms that SEI growth, lithium plating, and particle cracking are the dominant capacity-fade pathways, while SEI decomposition is the recognized trigger for thermal runaway. The same literature explains why LFP cells are so much harder to ignite: their phosphate framework holds oxygen far more tightly than a layered nickel oxide, so the final, fire-feeding oxygen-release stage either does not occur or occurs only at much higher temperatures, blunting the cascade.

Practical Recommendations

The chemistry hands you a clear playbook for making batteries last. Almost every longevity tip is really a way to slow SEI growth, avoid lithium plating, or reduce mechanical strain on the electrodes — the three mechanisms above.

The biggest lever is state of charge combined with heat. SEI growth accelerates sharply when a cell is stored full and warm, because a high anode potential and high temperature both speed the side reactions that consume lithium. Keeping a cell near the middle of its range and cool does more for its lifespan than any other habit. Avoiding cold fast-charging prevents lithium plating; gentle charging reduces particle cracking. These are not folk remedies — they map directly onto the failure modes.

Practical checklist, grounded in the mechanisms:

• Keep daily charge between roughly 20% and 80%. The top and bottom of the range stress the electrodes most. Use the “optimized” or “80% limit” charging settings most phones and EVs now ship.
• Avoid heat. Don’t leave devices in hot cars or in direct sun while charging. Heat is the single biggest accelerator of calendar aging, and because of the Arrhenius doubling, even a 10 °C difference matters.
• Don’t fast-charge in the cold. Let a cold battery warm toward room temperature first to avoid lithium plating.
• Don’t store a battery full or empty long-term. For months of storage, leave it around 40–60% in a cool place; a fully depleted cell can drop low enough to be permanently damaged.
• Prefer slower charging when you have time. Fast charging is convenient but mechanically and thermally harder on the cell — remember the I²R heat scales with the square of the current.
• Take physical damage seriously. A swollen, punctured, or crushed cell has a compromised separator — stop using it and dispose of it properly. Never charge a damaged cell.

Frequently Asked Questions

Why do lithium-ion batteries lose capacity over time?

Capacity fade is mostly lithium being permanently consumed and electrodes wearing out. The SEI layer on the anode keeps slowly thickening throughout the battery’s life, locking away lithium that can no longer shuttle. On top of that, repeated cycling cracks electrode particles, exposing fresh surface that grows still more SEI, and high-voltage storage and heat speed the whole process. The lithium isn’t “leaking” — it is being chemically trapped, mostly as solid film and as isolated “dead lithium” cut off from the circuit.

Is it bad to charge my phone to 100% every night?

It is not catastrophic, but it accelerates aging. A cell held at 100% sits at maximum anode potential, which speeds the side reactions that grow the SEI and consume lithium, especially if the device is also warm. Keeping the daily ceiling near 80% measurably extends cycle life, which is exactly why most phones and EVs now offer an optimized-charging or charge-limit setting that holds the battery lower overnight and tops it up just before you wake.

What actually causes a lithium battery to catch fire?

Fires come from thermal runaway: a self-feeding chain of heat-releasing reactions. A trigger — an internal short from a dendrite or manufacturing defect, a crush, overcharge, or external heat — pushes the cell hot enough to break down the SEI. That exposes graphite to electrolyte, melts the separator into a bigger short, and finally decomposes the cathode, which releases oxygen and ignites the flammable electrolyte from inside. Each stage drives the next hotter, and once the cell passes its self-heating onset temperature the cascade sustains itself without any further outside help.

Why is water bad for putting out lithium battery fires?

It is not exactly that water is bad — it is that water can’t stop the core reaction. Once the cathode decomposes it supplies its own oxygen internally, so the fire keeps burning even when smothered or doused. Firefighters use large volumes of water mainly to cool the burning cell and, crucially, the neighboring cells to stop the runaway propagating from one cell to the next.

What is the difference between NMC and LFP batteries?

NMC (nickel-manganese-cobalt) cathodes pack more energy per kilogram, so they dominate high-range EVs and laptops, but they are more prone to cracking and thermal runaway and cost more. LFP (lithium iron phosphate) holds less energy per kilogram but is cheaper, far more thermally stable, and lasts more cycles — often 2,000–5,000 versus a few hundred — which is why it has taken over standard-range EVs and most stationary storage. The deep reason for the safety gap is that LFP’s phosphate framework holds its oxygen tightly, so it resists the oxygen-release stage that drives NMC thermal runaway. It is an energy-density-versus-safety-and-cost trade.

Can a lithium-ion battery be overcharged?

Yes, and it is dangerous, which is why every cell ships with protection electronics. Forcing charge past full keeps pulling lithium out of the cathode and pushing it onto the anode beyond what the structures can hold, driving lithium plating and over-stressing the cathode toward lattice collapse and oxygen release. This generates heat and can tip the cell into thermal runaway. The battery management system exists precisely to cut charging at the voltage limit and prevent this.

Further Reading

• How Solid-State Batteries Work — the next-generation chemistry trying to replace the flammable liquid electrolyte described here.
• How Noise-Cancelling Headphones Work — another “how it actually works” deep dive into everyday hardware physics.
• How GPS Actually Works — the relativity and signal physics behind the chip in your pocket.
• External: U.S. DOE — How does a lithium-ion battery work? and Lithium Battery Degradation and Failure Mechanisms: A State-of-the-Art Review (Energies, MDPI).

By Riju — about