Batteries

In 1986, a Japanese chemist named Akira Yoshino took his new battery to a facility built for testing explosives and dropped a large piece of iron on it. Nothing happened. Then he did the same to a battery made with pure lithium metal, and it blew up violently. He later called the quiet one "the moment when the lithium-ion battery was born". Five years later, his kind of battery went on sale. In 2025, the world installed about 1.2 terawatt-hours of batteries in electric vehicles alone: enough, at the arithmetic level, to fill some 78 billion phone batteries.

The previous piece in this series, Cars, ended with the battery as the heart of the electric car. This one opens it up. It picks up where How Electricity Actually Reaches Your Wall left off: that piece explains current, voltage and why your charger turns AC into DC. Here we look at where the electricity in a battery comes from in the first place.

Part 1 — From a pile of discs to the car battery

Frogs' legs and a pile of discs

The word "battery" is older than the battery. Benjamin Franklin used it in 1749 for a row of linked Leyden jars, glass jars that could store a jolt of static electricity. He borrowed it from the army, where a battery is a group of guns firing together. A Leyden jar could give one sharp spark. It could not give a steady current.

The steady current came out of a scientific quarrel. In the 1780s the Italian physician Luigi Galvani found that a dissected frog's leg twitched when it was touched by two different metals. He thought the electricity came from the frog: "animal electricity". His countryman Alessandro Volta, a physics professor at Pavia, suspected the opposite, that the electricity came from the two metals, and that the frog was just a wet, salty conductor. In 1794 Volta showed that two metals and a cloth soaked in salt water produce a current with no frog at all.

Then he made it stronger by stacking it. He piled up discs: copper (or silver), then zinc, then cardboard or cloth soaked in brine, then copper, zinc, brine again, over and over. Join a wire from the top of the pile to the bottom and a steady current flows. He described it in a letter dated 20 March 1800 to Sir Joseph Banks, the president of the Royal Society in London, which published it that year as "On the electricity excited by the mere contact of conducting substances of different kinds".

Volta's pile was the first device that could make electricity flow steadily, on demand. Every battery since is a variation on it.

It was a scientific earthquake. That same year, two English chemists, William Nicholson and Anthony Carlisle, used a pile to split water into hydrogen and oxygen. Within eight years Humphry Davy had used bigger piles to isolate sodium, potassium, calcium and several other elements for the first time. The unit of electrical push, the volt, is named after Volta.

There was one honest irony. Volta was right that you needed two different metals, but wrong about why. He thought simple contact between the metals was the source. In fact, as chemists later showed, the energy comes from a chemical reaction: the zinc is slowly eaten away. That is the first lesson of every battery. A battery doesn't store electricity. It stores chemicals that want to react, and it makes them react through a wire.

Volta's pile had a serious practical problem: its voltage was unsteady and it could not keep up a strong current for long. In 1836 the British chemist John Frederic Daniell built a cell that gave a much steadier voltage, and it became the workhorse that powered the electric telegraph. But all these batteries had something in common. When the chemicals were used up, the battery was finished.

Batteries you can recharge: lead and nickel

In 1859, a French physicist named Gaston Planté rolled two sheets of lead into a spiral, separated by cloth, and soaked them in sulfuric acid. When he pushed a current through it, the chemistry ran backwards and the battery could be used again. It was the first rechargeable battery. In 1881, Camille Alphonse Faure made it practical to mass-produce, by pressing a paste of lead oxide into a lead grid.

The lead-acid battery is heavy and stores little energy for its weight, but it is cheap, rugged, and can deliver a huge burst of current. That is exactly what you need to turn over a cold engine, which is why, more than 160 years later, the battery under the bonnet of a petrol car is still lead-acid. Even many electric cars keep a small one to run the lights and electronics. Lead-acid batteries were also what the electric cars of 1900 ran on, and their weight is one reason those cars lost.

Nickel came next. In 1899, the Swedish inventor Waldemar Jungner made the first nickel-cadmium battery, and Thomas Edison soon developed a nickel-iron version in the United States. Nickel-cadmium cells powered cordless phones, power tools and the first portable electronics. They were later replaced by nickel-metal hydride, which avoided toxic cadmium; consumer cells went on sale in 1989. Nickel-metal hydride batteries are still in the AA rechargeables in many drawers, and they powered the first generation of hybrid cars, including the Toyota Prius.

Look at the voltage column. A lithium-ion cell gives roughly three times the push of a nickel one. That, and its light weight, is why it took over. To see why, we need to look inside.

Part 2 — How a battery works, and how lithium-ion was born

The four parts of every battery

Every battery, from Volta's pile to the one in your phone, has the same four parts:

1. The anode, the negative end. When the battery is in use, it is made of something that wants to give away electrons.
2. The cathode, the positive end. It is made of something that wants to take electrons.
3. The electrolyte between them, a liquid, paste or solid that lets charged atoms (ions) move through it, but not electrons.
4. The separator, a thin porous sheet that keeps the anode and cathode from touching, because if they touch, the reaction happens inside in a rush of heat: a short circuit.

Here is the trick that makes a battery useful. The anode and cathode want to react, but the electrolyte only lets part of the reaction through. The positive ions can cross through the electrolyte. The electrons can't. So the electrons are forced to go the long way round, out of the battery, through your wire and your phone, and back in at the other end. That flow of electrons through the outside circuit is the current that does the work.

Ions inside, electrons outside. If you remember one sentence from this piece, make it that one. Inside the battery, charged atoms move. Outside the battery, electrons move. The two flows always balance, which is why a battery stops working the moment you break the circuit: the electrons can't get round, so the ions stop too.

Voltage is how strongly the anode material wants to give up electrons compared with how strongly the cathode wants them. Pick materials that are very different, and you get a bigger push. Lithium is the extreme case. It is the lightest metal, and it gives up its single outer electron more readily than almost anything else. That is why lithium batteries reach well over three volts per cell where nickel ones manage 1.2.

Charging is the reaction run backwards. A charger pushes electrons back into the anode, and the ions follow them back across. In a lead-acid or nickel battery, that means rebuilding the electrodes chemically every time, and they slowly crumble. The lithium-ion battery found a gentler trick, and it is the reason it lasts as long as it does.

The lithium-ion story

The lithium-ion battery began with an oil crisis. In the 1970s, worried that oil would run out, the oil giant Exxon hired leading energy researchers and gave them freedom to work on anything that wasn't petroleum. One of them was a young British chemist, M. Stanley Whittingham, who joined Exxon in 1972.

Whittingham's idea: houses for ions. He was studying materials made of layers with gaps between them, like a stack of paper with room between the sheets, where ions can slip in and out without destroying the structure. Chemists call it intercalation. He found that titanium disulphide could take in lithium ions this way, and used it as a cathode, with lithium metal as the anode. The battery gave just over two volts and could be recharged. After a meeting at Exxon's headquarters that lasted about fifteen minutes, management quickly decided to develop it, and in 1976 Whittingham announced it. It was first made on a small scale for a Swiss clockmaker's solar-powered watches.

Then came the fires. Each time the battery was charged, thin whiskers of lithium metal grew from the anode, called dendrites. When one reached the other side, it short-circuited the cell, which could burst into flames. According to the Nobel committee's own account, the fire brigade was called so often that it threatened to bill the laboratory for the special chemicals used to fight lithium fires. When oil prices fell in the early 1980s, Exxon gave up.

Goodenough's idea: a better cathode. John Goodenough, an American physicist then at Oxford, reasoned that a metal oxide would give a higher voltage than a sulphide. His group searched systematically and in 1980 published lithium cobalt oxide, which gave about four volts, nearly double. He also realised that a battery didn't have to be built charged: you could build it empty and charge it afterwards, which made it much safer to make. His co-author on the paper was Koichi Mizushima.

Yoshino's idea: no lithium metal at all. In Japan, where electronics companies were desperate for lightweight rechargeable batteries for video cameras, cordless phones and computers, Akira Yoshino at the Asahi Kasei company took Goodenough's cathode and paired it with an anode made of petroleum coke, a carbon material that could also house lithium ions. Now there was no lithium metal anywhere in the battery, only lithium ions shuttling between two "houses". In 1985 he had the first commercially viable lithium-ion battery, and in 1986 came the falling-iron test.

The lithium-ion battery works because nothing has to be rebuilt each cycle. Ions just move from one house to another, and back.

In 1991, Sony began selling the first lithium-ion batteries. Phones shrank, laptops became truly portable, and, eventually, cars could run on them.

The Nobel Prize, and the people it left out

On 9 October 2019, the Royal Swedish Academy of Sciences awarded the Nobel Prize in Chemistry to Goodenough, Whittingham and Yoshino "for the development of lithium-ion batteries". Goodenough was 97, the oldest person ever to receive a Nobel Prize. He died on 25 June 2023, a month short of his 101st birthday. He had kept working into his nineties, and his group also introduced lithium iron phosphate as a cathode in the 1990s, which matters later in this story.

A Nobel Prize can go to at most three people, and the lithium-ion battery had more parents than that. Fair accounts name, among others:

In 2014, Goodenough, Nishi, Yazami and Yoshino shared the US National Academy of Engineering's Draper Prize for the lithium-ion battery. Big inventions are usually like this: a relay, not a single runner.

Part 3 — Charging, and why batteries wear out

Energy and power

Two numbers describe any battery, and confusing them causes most of the muddle about electric cars.

Energy, in kilowatt-hours (kWh), is how much the battery holds: the size of the tank. Power, in kilowatts (kW), is how fast energy flows in or out: the width of the hose. If you have read Your Utility Bills, Decoded, it is the same pair: a 2 kW heater left on for three hours uses 6 kWh. A phone battery labelled 4,000 mAh at 3.85 volts holds about 15 watt-hours (4 Ah × 3.85 V). A 70 kWh car battery holds as much as about 4,500 of those phones.

Battery size (kWh) is the tank. Charger power (kW) is the hose. Time is tank ÷ hose.

The calculator below lets you try it. Pick a battery, a car and a charger, and set how full the battery starts and ends.

Three experiments are worth doing. First, keep everything the same and switch from the efficient saloon to the big pickup: the same 50 kWh of charge takes you about half as far, because the pickup uses twice the energy per kilometre. Second, compare the wall socket with the home charger for an overnight top-up from 20% to 80%: the slow one takes over a day for a big battery. Third, look at the cost, then type in your own electricity price from your bill (Your Utility Bills, Decoded shows where to find it). If you charge in public, use the price per kWh the charging company shows in its app, which is often different from what you pay at home.

Why your battery wears out

Your two-year-old phone doesn't last the day the way it used to. Nothing is broken. A lithium-ion battery loses capacity in two ways: a little every time it is cycled, and a little simply with the passing of time, especially when it sits fully charged or hot. Battery scientists call these cycle ageing and calendar ageing.

The main culprits, in plain words:

1. A crust on the anode. The first time a battery is charged, a thin film forms on the graphite anode where it meets the electrolyte, called the solid electrolyte interphase. It protects the anode, but it keeps slowly thickening, and every lithium ion trapped in it is one that can no longer carry charge. It grows faster when the battery is hot and when it sits at a high charge.
2. Lithium plating. Charge too fast, especially in the cold, and the ions can't slip into the graphite quickly enough. Some pile up on its surface as metallic lithium instead, the same problem that made Whittingham's battery catch fire. That lithium is lost, and in bad cases it can cause short circuits.
3. Cracking. Electrode particles swell and shrink slightly every time lithium moves in and out. Over thousands of cycles they crack, and bits lose contact with the rest of the electrode.

Real-world data matches the chemistry. Geotab, a vehicle-telematics company, analysed more than 22,700 electric vehicles across 21 models and reported in April 2026 that their batteries lost on average 2.3% of their capacity a year, which projects to about 82% left after eight years. Cars that used high-power fast chargers most often degraded fastest, about 3% a year; those that rarely fast-charged lost about 1.5%. Cars in hot climates lost about 0.4 percentage points a year more than those in mild ones. The US Department of Energy's Alternative Fuels Data Center notes that studies suggest an EV battery may still hold at least 70% of its original capacity at the end of the car's life.

Safety belongs here too. Most lithium-ion batteries use a liquid electrolyte made with flammable organic solvents. If a cell is crushed, overcharged or badly made, it can overheat and set off its neighbours, a chain reaction called thermal runaway, and the smoke is toxic. Such fires are hard to put out, which is one reason airlines have special rules for carrying spare batteries.

Part 4 — Why batteries got about ten times cheaper, and the chemistries racing now

Three independent measurements tell the same story.

The three use different methods, so don't splice them into one line. But together they say that a battery that would have cost well over a thousand dollars per kWh when the Tesla Roadster appeared costs around a hundred today.

Why? Partly better chemistry: more energy packed into each cell, and cheaper materials, especially the shift to iron phosphate cathodes described below. Mostly, though, it is the ordinary magic of making something in enormous and growing quantities. Factories got bigger, processes got better, scrap rates fell, and suppliers competed. Economists call this a learning curve: costs fall by a roughly steady fraction every time cumulative production doubles. Solar panels followed the same path.

The battery didn't get cheap because of one breakthrough. It got cheap because the world built a lot of them, and kept building more.

There is a catch worth knowing. The International Energy Agency's Global EV Outlook 2026 warns that today's record-low prices are "partly sustained by losses" further up the supply chain: many makers of cathode materials, the most important ingredient in a cell's cost, are operating at a loss while still expanding. Lithium prices in early 2026 were more than double their level a year earlier, and cobalt prices doubled. Prices might not keep falling in a straight line.

LFP, NMC, sodium-ion and solid-state

"Lithium-ion" is a family, usually named after the cathode. Two members dominate cars today.

NMC (nickel manganese cobalt oxide) is the descendant of Goodenough's cobalt oxide. It packs the most energy into the least weight, which is why it suits long-range and premium cars. Modern versions use much more nickel and less cobalt than earlier ones.

LFP (lithium iron phosphate) holds less energy for its weight but uses cheap, abundant iron and phosphate, contains no nickel or cobalt, is more resistant to overheating and lasts many cycles. For years it was seen as second-best. Then Chinese manufacturers improved how cells are packed together, and LFP became good enough for most cars. The IEA says LFP made up over 55% of the batteries put into electric vehicles worldwide in 2025, and over 90% of the batteries installed for grid storage. BloombergNEF's 2025 survey found LFP packs averaging $81 per kWh against $128 for NMC.

Sodium-ion swaps lithium for sodium, which is in ordinary salt and found almost everywhere. It stores less energy, so the IEA puts the range of an average SUV with a sodium-ion battery at up to about 350 km, against 400–600 km for lithium-ion. But it works much better in the cold: the latest cells keep about 90% of their capacity at −40 °C. The IEA expects it to suit short-range city cars, urban delivery vans, two- and three-wheelers, forklifts and grid storage. In 2026, sodium-ion factories had just over 1% of the capacity of lithium-ion ones.

Solid-state batteries replace the flammable liquid electrolyte with a solid. The promise is more range and better safety, and possibly a lithium-metal anode at last, the dream Whittingham's fires ended. The IEA's verdict in 2026 is blunt: those advantages "have not yet been demonstrated in real-world applications." Some simpler semi-solid designs are already on sale. Toyota has said it plans its first all-solid-state car by 2028; BYD plans a first car from 2027 and mass production from 2030. Treat any company's date as a plan, not a fact: battery announcements have a long history of slipping.

Part 5 — Digging it up: lithium, cobalt and the supply chain

A battery is a mining product before it is a chemistry product.

Lithium. According to the US Geological Survey, the world (outside the United States, whose figures are confidential) mined about 290,000 tonnes of lithium in 2025, 31% more than the year before. Australia mined the most, about 92,000 tonnes, mainly from hard rock; then China (62,000), Chile (56,000), Zimbabwe (28,000) and Argentina (23,000). Batteries took an estimated 88% of all lithium. In Chile and Argentina, lithium comes from brine pumped up from under salt flats and concentrated in huge evaporation ponds in some of the driest places on Earth, and the use of water there has brought disputes with local and Indigenous communities, as well as negotiated benefit-sharing deals. Lithium's price is famously wild: the USGS's average US contract price for battery-grade lithium carbonate went from $63,700 a tonne in 2022 to $9,000 in 2025.

Cobalt. The Democratic Republic of the Congo (DRC) produced an estimated 73% of the world's mined cobalt in 2025, the USGS says, followed by Indonesia with 14%. Much of it comes from large industrial copper-cobalt mines; some is dug by hand by independent "artisanal" miners. In 2016, Amnesty International and the Congolese group Afrewatch documented children as young as seven working in such mines, citing a UNICEF estimate that about 40,000 children worked in mines across southern DRC in 2014, and traced that cobalt through traders and a Chinese processor to battery-component makers that supplied well-known electronics and car brands. Of 16 multinationals Amnesty contacted, none gave enough detail to verify where their cobalt came from. Companies have since adopted auditing and sourcing programmes, and campaigners continue to argue about how much has changed on the ground.

Two developments have changed the picture since then. First, chemistry: LFP batteries use no cobalt at all, and the nickel-rich NMC types use much less than older ones. Second, politics: in February 2025 the DRC temporarily banned cobalt exports to push prices up, then from 16 October replaced the ban with export quotas, and cobalt prices roughly doubled. The USGS reports that the quotas allow up to 96,600 tonnes a year in 2026 and 2027, far below the roughly 230,000 tonnes the country mined in 2025. Batteries are now the leading use of cobalt, so these decisions in Kinshasa reach every battery maker.

Processing is the real choke point. Mining is spread across several continents; refining and manufacturing are not. The IEA reports that China accounted for over 80% of battery-cell production in 2025 and even higher shares of the active materials inside them, including almost all LFP cathode material. Chinese, Korean and Japanese companies supply nearly all the battery cells in the world. In October 2025 China announced broader export controls on battery materials, including cathode and anode materials, then paused them for one year. Europe and the United States are building their own factories; the EU and the US each hold about 6–7% of global manufacturing capacity, the IEA says, against China's more than 80%.

An honest comparison runs both ways. A petrol car burns its fuel once and needs more every week, from an oil industry with its own history of spills, wars and pollution. A battery's metals stay in the battery and can, in principle, be recovered. That doesn't excuse a child in a mine, but it does mean the long-run answer is better mining and much more recycling, not pretending either car is free of extraction.

Part 6 — Recycling, grid storage, and the state of affairs in late 2026

A second life, then recycling

A battery too worn for a car is not necessarily finished. As Part 3 noted, the US Department of Energy says an EV battery may still hold at least 70% of its capacity when the car is retired, and it suggests such batteries could go on to serve for ten years or more in less demanding jobs, such as storing energy for a building or the grid. The IEA says reuse remains challenging in practice: second-life systems need safety checks and warranties, nobody knows exactly how long a used battery will last, taking packs apart safely costs money, and new batteries keep getting cheaper. Selling, repairing and refurbishing used electric cars, by contrast, is a fast-growing business.

Battery recycling starts by shredding cells into "black mass", a powder of the valuable metals. The metals are then recovered by smelting (high heat, large scale, energy-hungry), by dissolving them in chemicals (hydrometallurgy, cheaper to build), or by "direct recycling", a newer approach that tries to recover the cathode material intact instead of breaking it down into separate metals.

The big constraint is time. Nearly all the EV batteries ever made are still in cars. The IEA estimates a lag of roughly 15 years between a surge in battery sales and a matching surge in worn-out batteries, so for now most recycling feedstock is factory scrap, and end-of-life batteries won't dominate until the mid-2030s. China hosts over 85% of global recycling capacity. There's also an economic twist: LFP and sodium-ion batteries contain fewer valuable metals, so recycling them may need rules that make makers responsible for their old batteries, as the EU's battery regulation now does, rather than relying on metal prices alone.

Batteries on the grid

The same cells that move cars can steady power grids, soaking up cheap solar power at midday and releasing it in the evening. BloombergNEF found that battery packs for stationary storage fell 45% in a year to $70 per kWh in 2025, making them the cheapest battery segment for the first time. In the United States, the Energy Information Administration reported that a record 15 gigawatts of utility-scale battery storage was added in 2025, with 24 gigawatts planned for 2026, about 28% of all the new generating capacity planned for the year, and more than half of the batteries in Texas. (For how that power then reaches your home, see How Electricity Actually Reaches Your Wall.)

The state of affairs, late 2026

Open questions

Will costs keep falling? Optimists point to the learning curve and new chemistries. Sceptics point to loss-making suppliers, volatile metals and tariffs. Both have evidence.

Can supply chains be diversified without making batteries dearer? Building factories outside China improves security but, for now, costs more; BloombergNEF found pack prices in North America and Europe 44% and 56% above China's.

Can mining be made decent? Better sourcing rules, cobalt-free chemistries and recycling all help. None of them is yet enough on its own.

Is the next breakthrough sodium, solid-state, or something else? History suggests caution: lithium-ion took about twenty years to go from Whittingham's lab to Sony's shelves, and decades more to become cheap.