A battery ages even when you never use it. The reason is a film a few nanometres thick

Buy a spare laptop battery, leave it sealed in its box for two years, and it will come out meaningfully weaker than it went in. It was never charged, never discharged, never warm in a pocket. It simply sat there.

This is the clearest evidence that a battery is not a bucket that wears out from being filled and emptied. It is a chemical system in which something slow and irreversible is happening all the time, at a rate set mostly by temperature and by how full the cell is being kept. That process is concentrated in a film on the anode a few nanometres to a few tens of nanometres thick — and once you know what that film is, most battery behaviour, and most of the advice that genuinely works, follows directly.

What is actually happening when a cell charges

A lithium-ion cell is often described as storing electricity. It is more accurate, and much more useful, to say it stores lithium in two different places.

On one side is the cathode, almost always a metal oxide: a layered oxide of nickel, manganese and cobalt (NMC), or lithium cobalt oxide in older consumer cells, or — increasingly — lithium iron phosphate, LFP, which is a phosphate rather than an oxide and behaves differently for it. On the other side is the anode, in the overwhelming majority of cells today, graphite. Between them sits a separator soaked in a liquid electrolyte carrying a lithium salt.

Charging pulls lithium ions out of the cathode lattice, sends them across the electrolyte, and slots them between the sheets of the graphite. Electrons take the long way round through the charger. Discharging reverses it. This slotting-in is intercalation — the lithium is a guest inside a host structure that stays largely intact, which is exactly why the process can be repeated hundreds of times. Nothing is burnt and no new bulk material is formed. That is the whole design, and its elegance is why this chemistry beat everything else.

Why particle size decides how fast you can charge

Here is the first place the nanoscale becomes decisive, and it is a piece of physics rather than a marketing claim.

A lithium ion arriving at an electrode particle does not stop at its surface — it has to diffuse into the solid. Solid-state diffusion is slow, and the time taken scales with the square of the distance travelled. Halve the particle radius and the ion needs roughly a quarter of the time to reach the centre.

This is why high-power electrodes use small particles, and why lithium iron phosphate — a material with genuinely poor intrinsic electronic and ionic conductivity — became commercially viable only when it was made as nanoparticles coated in a thin carbon layer. The nanostructuring is not an enhancement bolted on afterwards; without it the material does not work at useful rates.

The catch is exact and unavoidable. Dividing the same mass into smaller particles multiplies the surface area, and every square metre of that surface is also somewhere the electrolyte can react. The design choice that buys fast charging also buys faster ageing, and every cell on the market is a specific compromise between those two — which is why a high-power cell and a long-life cell are different products even when the chemistry on the label is the same.

The film that protects the battery by damaging it

The electrolyte in a lithium-ion cell is not thermodynamically stable at the voltage the charged graphite anode sits at. It should decompose. On the very first charge, it does.

What forms is the solid-electrolyte interphase, universally shortened to SEI: a thin, complex layer of lithium salts and organic decomposition products deposited on the anode surface. It has a very specific and rather demanding job. It must conduct lithium ions, so charging can continue, while insulating electrons, so the electrolyte stops being reduced. A good SEI is therefore self-limiting — it grows until it blocks the electron transfer that was forming it, then largely stops.

Three consequences follow, and between them they explain most of what a battery does over its life.

It costs lithium immediately. The lithium built into the SEI is chemically locked up and never shuttles again. This is the first-cycle capacity loss, typically several per cent, and manufacturers plan for it — part of a new cell's lithium inventory exists purely to be sacrificed to this layer.

It never entirely stops growing. Self-limiting is not the same as finished. The film continues to thicken slowly for the rest of the cell's life, consuming a little more lithium and adding a little more internal resistance each time. This is calendar ageing, and it is why the battery in the sealed box got weaker: the reaction does not need you to use the cell, only for the cell to exist at some temperature and some state of charge.

Its growth rate is set by things you control. Chemical reaction rates rise steeply with temperature, so a hot cell ages far faster than a cool one — heat is, by a wide margin, the single most damaging thing in ordinary use. And a fully charged anode sits at the lowest potential it ever reaches, which is the strongest possible driving force for reducing the electrolyte. Storing a cell at 100% and warm is therefore the worst combination available, and storing it around half full and cool is the best.

Almost every piece of battery advice that actually works is a way of saying the same thing twice: keep it cool, and do not keep it full. The rest is detail.

Lithium plating, and why your phone refuses to fast-charge in the cold

There is a second, faster failure that is worth understanding separately, because it is the one behind several visible behaviours.

When lithium ions arrive at the graphite faster than they can intercalate into it, they have nowhere to go and take the alternative path: depositing as metallic lithium on the anode surface. This is lithium plating, and it is bad in two distinct ways. The plated lithium largely does not come back — it reacts with electrolyte and forms more SEI, so capacity is permanently lost. And it deposits unevenly, in needle-like structures that can in the worst case grow far enough to reach the other electrode.

Plating is driven by anything that outpaces intercalation: very high charging currents, charging a cell that is already nearly full, and above all low temperature, because diffusion into graphite slows dramatically as the cell gets cold. Charging a lithium-ion battery below freezing is genuinely damaging.

This is why a well-designed device will refuse to fast-charge when cold, throttle charging when hot, and slow down markedly as it approaches full. Those are not defects and not manufacturers being cautious for its own sake; they are the battery management system avoiding a failure mode with permanent consequences. The taper near 100% is the same story — the last stretch is charged at reducing current specifically because pushing hard there is what plates lithium.

Cycle counts are not what most people think

A cell rated for "500 cycles" is not rated for 500 plug-ins. Cycles are counted as full equivalent cycles: discharging 50% and recharging twice counts as one, not two.

More importantly, not all cycles are equally damaging. A shallow cycle in the middle of the range — say between 40% and 70% — puts far less strain on the electrode structures than a full 0% to 100% swing, because the electrode materials expand and contract as lithium moves in and out, and repeated large excursions crack particles, break electrical contact and expose fresh surface for yet more SEI to form on.

This is the real basis for the charge-limiting features now common on phones and laptops. Capping charge at 80% is not a placebo. It avoids the highest-potential region where electrolyte oxidation and SEI growth are fastest, and it reduces the depth of every cycle — two independent benefits at the cost of some usable capacity today for meaningfully more capacity in two years.

The cathode ages too, more quietly: particles crack under repeated volume change, transition metals slowly dissolve into the electrolyte and end up poisoning the anode's SEI, and the surface reconstructs into phases that hold lithium less well. High voltages and high temperatures accelerate all of it, which is the same advice arriving from a different direction.

Why silicon anodes are so hard, and so nano

Graphite stores roughly one lithium for every six carbon atoms. Silicon stores far more — around ten times the capacity by weight — which is why every battery roadmap for a decade has had silicon on it.

The obstacle is mechanical. Silicon expands by roughly three hundred per cent in volume when fully lithiated. A solid silicon particle cracks apart within a few cycles. Worse, every new crack exposes fresh silicon surface, the electrolyte immediately reacts with it to form more SEI, and each round of that consumes more lithium — so the cell dies not from the cracking directly but from the lithium inventory being eaten by an SEI that keeps having to re-form.

Every workable answer is a nanostructuring answer: silicon nanoparticles small enough to accommodate the strain without fracturing, porous silicon with room to swell inwards, silicon embedded in a carbon matrix, and conductive networks — often carbon nanotubes — flexible enough to keep electrical contact while the electrode breathes. This is the clearest current example of a nanomaterial being not an improvement to a working product but the precondition for the product existing at all.

Which is also why commercial cells today mostly use silicon as a small percentage blended into graphite rather than as a wholesale replacement. The engineering is real, the progress is real, and the announcements are usually well ahead of the shipping product.

What to actually do, and what is myth

Heat is the main enemy. A phone left on a car dashboard, a laptop charging on a bed with blocked vents, a power bank in direct sun — these do more damage in an afternoon than months of ordinary cycling. If a device is hot while charging, that is the moment worth intervening in.

Do not store at 100%, and do not store at 0%. For long storage, roughly half charge in a cool place is the target. A cell left completely flat for months can self-discharge to a voltage low enough that charging it again is genuinely unsafe, which is why chargers refuse to revive deeply discharged packs.

Partial charging is fine, and better than fine. There is no memory effect in lithium-ion — that was nickel-cadmium, and it does not apply here. Topping up whenever convenient is exactly right, and shallow cycles are gentler than deep ones.

Occasional full discharges do not "calibrate the battery", they calibrate the fuel gauge — the software estimating remaining charge. That is occasionally useful when the percentage reading has drifted, and it does nothing for the cell's health.

A swollen battery is finished. Gas generation means the electrolyte is decomposing badly. Stop using it, do not puncture or compress it, and dispose of it through proper e-waste channels rather than household rubbish.

Why it matters for students and researchers

The SEI has a reasonable claim to being the most consequential few nanometres in modern technology. It determines the life, safety, low-temperature behaviour and fast-charge capability of every lithium-ion cell on Earth, it forms spontaneously rather than by design, and after three decades of intensive study its structure and evolution are still not fully characterised.

Part of the reason is a measurement problem, and it connects directly to why nanoscale characterisation is so difficult. The SEI is air-sensitive, moisture-sensitive, beam-sensitive and buried inside a sealed cell. Removing it from the cell to look at it changes it; imaging it with an electron beam damages it. Much of the progress in this field has come from developing methods to observe the layer without destroying it, and that instrumentation work is as much the science as the chemistry is.

The open problems are large and unusually well defined. Artificial SEI layers engineered deliberately rather than grown accidentally. Electrolyte additives that build a better film — currently one of the highest-leverage and most empirical areas in the whole field. Solid electrolytes, which change the interface problem rather than removing it. Silicon at high loading. And sodium-ion chemistry, where the resource argument is compelling and where India's own materials position is considerably better than it is for lithium.

Frequently asked questions

Should I let my phone discharge fully before charging?

No. That advice is a survival from nickel-cadmium batteries, which genuinely had a memory effect. Lithium-ion cells prefer shallow cycles in the middle of their range, and deep discharges are mildly harmful rather than helpful. Charge whenever it suits you, and if your device offers an 80% charge limit, using it is a real and measurable benefit.

Does fast charging damage the battery?

It contributes, mainly through heat and through the risk of lithium plating at high current — but a modern battery management system limits current based on temperature and state of charge specifically to keep this in hand. Occasional fast charging on a cool device is not something to worry about. Repeated fast charging while the device is already hot is the case actually worth avoiding.

Why does my two-year-old phone die at 30%?

Two things are usually happening together. The cell has genuinely lost capacity, and its internal resistance has risen, so under a heavy load the voltage sags enough that the device shuts down while the gauge still reads charge remaining. This is why sudden shutdowns at moderate percentages, especially in cold weather, are a classic sign of an aged cell rather than of faulty software.

Is it bad to leave a laptop plugged in all the time?

It is not the plugging in that hurts, it is what it implies: a cell held near 100% and often warm, which is precisely the combination that maximises SEI growth. Most modern laptops mitigate this with a charge limit setting that holds the battery near 60–80% when it detects mains use. Turning that on is the single most useful thing most laptop owners can do for battery life.

Are LFP batteries really longer-lived?

Generally yes, and for structural reasons rather than marketing. Lithium iron phosphate operates at a lower voltage, which is gentler on the electrolyte, and its olivine structure changes volume very little during cycling, so particles crack less. The trade-offs are lower energy density, so more weight and volume for the same capacity, and weaker cold-temperature performance. That is why LFP dominates in stationary storage and in vehicles where range is adequate, while high-nickel oxides persist where energy per kilogram is what matters.