Batteries: chemistry, cell format, and control of the chain

Battery deployment for electric vehicles reached 1.2 terawatt-hours in 2025, an increase of almost 30 percent on 2024 and more than seven times the 2020 level. China accounted for 60 percent of EV battery deployment, the European Union almost 15 percent, and the United States 10 percent, with the US stagnating (IEA 2026). Global cell manufacturing capacity exceeded 4 terawatt-hours by the end of 2025, up roughly 30 percent, and China accounts for more than 80 percent of it, while the European Union and the United States hold 6 to 7 percent each.

Prices moved against the expectation that materials costs set them. BloombergNEF's survey found average pack prices at a record 108 dollars per kilowatt-hour in 2025, down 8 percent, despite battery metal prices increasing on supply risks at Chinese lithium assets and new cobalt export quotas in the Democratic Republic of the Congo. The drivers were continued cell manufacturing overcapacity, intense competition, and the shift to lithium iron phosphate (LFP), and BNEF expects prices to fall again in 2026 (BloombergNEF 2025). Packs in China averaged 84 dollars, while those in North America and Europe were 44 and 56 percent more expensive.

The useful frame is therefore a manufacturing system with four layers: the minerals, their processing into cathode and anode materials, the cell factory, and the pack and its integration. Chemistry decides which minerals are needed and how much energy fits in a kilogram. Scale, yield, and process know-how decide cost. In 2026 the decisive advantages lie in the second and third layers, where China dominates, more than in the laboratory.

LFP accounted for more than 55 percent of EV batteries deployed globally in 2025, up from nearly 50 percent in 2024, and LFP packs were more than 40 percent cheaper per kilowatt-hour than nickel manganese cobalt (NMC) alternatives (IEA 2026). Across all segments BNEF puts average LFP packs at 81 dollars against 128 dollars for NMC, and stationary storage, almost entirely LFP, fell 45 percent in a year to 70 dollars, the lowest-priced segment for the first time (BloombergNEF 2025).

The mechanism is cost structure. Cathode active material accounts for 40 to 50 percent of the cost of an NMC cell and 25 to 30 percent of an LFP cell, and LFP contains no cobalt or nickel (IEA 2026). Its lower energy density was once disqualifying for cars, but packaging innovations that place cells directly in the pack, longer-range targets met with larger packs, and the economics of storage, where weight matters little, removed most of the penalty. The result is a cheap, robust chemistry whose advantage compounds with scale.

LFP is also the clearest example of manufacturing concentration. LFP cathode material production is almost entirely concentrated in China (IEA 2026). The European Union's LFP demand, more than 10 percent of its EV batteries, is met almost entirely by imports from China, either as cells or packs (about 30 percent) or embedded in imported vehicles (nearly 70 percent). In the United States, LFP's share almost halved in 2025 from an already low base, as tariffs and credit rules raised the cost of the chemistry that is cheapest everywhere else.

Sodium-ion is the nearest new chemistry to volume. CATL's Naxtra cells are rated at 175 watt-hours per kilogram, a record for mass-produced sodium-ion cells, and the company said in April 2026 that mass production would begin by the end of 2026, with a sodium-ion variant of Changan's Nevo A06 the first passenger car to use them (New Atlas n.d.). CATL is investing about 5 billion yuan, roughly 735 million dollars, in 40 gigawatt-hours of sodium-ion capacity in Fujian (EHVTI n.d.). Sodium removes lithium from the cell and performs well in the cold, but it still relies on hard carbon anodes and processing capacity concentrated in China.

Solid-state batteries are moving from laboratory to small batches. CATL has advanced from 10 ampere-hour validation to 20 ampere-hour samples, targeting small-batch production in 2027 and large-scale manufacturing around 2030, with 500 watt-hours per kilogram in sight (Electrek 2026b). BYD's battery unit plans limited batches of sulfide-based all-solid-state cells in 2027 and mass production by 2030, while Toyota and Idemitsu reportedly aim for 2027 to 2028 and Samsung SDI for the second half of 2027 (Electrek 2026a). Even on these timelines, solid-state cells will serve premium segments for years, because cost and manufacturing yield, not energy density, decide adoption.

Cell format and anode materials are a parallel frontier. Samsung SDI began producing 46-series cylindrical cells in Korea in March 2025, combining a high-nickel cathode with a silicon-carbon anode (Samsung SDI 2025). Fraunhofer ISI estimates that a silicon-free 4680 cell reaches only about 240 to 250 watt-hours per kilogram, and that adding silicon requires a highly porous anode to absorb silicon's expansion (Fraunhofer ISI n.d.). Larger cells reduce the number of parts and steps per pack, which is a manufacturing gain more than a chemical one.

Every chemistry rests on a processing chain, and those chains share a geography. Cathode materials for LFP are almost entirely produced in China, and Korea, Japan, and Indonesia are only beginning to diversify (IEA 2026). Graphite anodes, natural and synthetic, are dominated by Chinese refining, and graphite exports have been licensed since December 2023, a layer that was never suspended (Battery Competence Cluster 2026). In the United States, combined anti-dumping and countervailing duties of at least 160 percent were reported on Chinese anode-grade graphite from February 2026, and some producers were reported to have rerouted material through Indonesia (ChemAnalyst 2026).

Mineral markets remain volatile and politically shaped. The Democratic Republic of the Congo supplies almost two-thirds of the world's cobalt and has replaced an export ban with quotas that run through 2027, at 96,600 tonnes for 2026 (Benchmark 2025; IEA 2026). Lithium prices at the start of 2026 were more than twice their level a year earlier but about 70 percent below their 2022 peak, and cobalt prices roughly doubled over the year (IEA 2026).

This is the central asymmetry of the battery industry. Changing chemistry changes which minerals are needed: LFP drops cobalt and nickel, sodium drops lithium. It does not change who refines, coats, mixes, and assembles them. A Western buyer that moves from NMC to LFP to escape cobalt risk swaps a Congolese dependency for a Chinese one. A move to sodium swaps lithium exposure for hard carbon exposure. Chemistry diversifies minerals, not suppliers.

China has converted processing leadership into policy. In July 2025 it added battery cathode preparation technologies to its catalogue of restricted technologies, and on 9 October 2025 it placed lithium-ion cells with energy density of 300 watt-hours per kilogram or more, key battery production equipment, cathode materials and precursors, and artificial graphite anode materials and technology on its dual-use export list. After the US–China meeting in Busan, the measures were suspended from 7 November 2025 to 10 November 2026 (Shanghai Metals Market 2025). A Dutch industry assessment found that the suspension was reported extended to 10 January 2027 under the late-September 2026 trade-truce extension, but could not locate a separate amending notice, and notes that several cathode and graphite items remain licensed regardless (Battery Competence Cluster 2026).

The United States has answered on the demand side. The One Big Beautiful Bill Act of July 2025 introduced prohibited-foreign-entity rules that deny clean energy credits, including the manufacturing credit for battery components and the investment credit for storage, where content from Chinese-linked entities exceeds set thresholds, and Treasury's Notice 2026-15 in February 2026 provided interim guidance. Licensing arrangements in which a prohibited entity directs production can also disqualify a plant (Foley Hoag 2026). The effect is to make Chinese technology licences, the fastest route to a competitive factory, more difficult to use.

The two instruments interact. Beijing's controls restrict the export of the know-how needed to build plants abroad, and Washington's rules restrict the use of Chinese partners in plants at home. A Western manufacturer therefore faces a narrow path: develop its own processes, license from Korean or Japanese firms, or accept credit-ineligible Chinese partnerships. The dial is held on both sides, and each turn raises the cost of building outside China.

The record outside China is mixed. EU and US cell capacity grew about 50 percent in 2025 against just over 25 percent in China, but both still hold only 6 to 7 percent of the world total (IEA 2026). Europe's flagship start-up, Northvolt, filed for bankruptcy in March 2025 after struggling to reach acceptable yields, and the region's LFP supply remains almost entirely imported. Pack prices in Europe were 56 percent above China's in 2025, and in North America 44 percent above (BloombergNEF 2025).

The firms with the best chance outside China are the incumbents. Korean and Japanese manufacturers hold the premium segments, with high-nickel and silicon-rich cells such as Samsung SDI's 46-series, and are the most likely licensors for Western plants. Chinese manufacturers, meanwhile, are building in Europe, so that some of Europe's new capacity is Chinese capacity located inside the European Union. The question for policy is less whether batteries are made in Europe or America than whose processes and materials they use. A plant that assembles imported cathode and anode material into cells adds jobs and some resilience, but it leaves the most concentrated steps of the chain where they were, and it remains exposed to the same export licences and price swings as an importer of finished cells.

The strongest objection is that the concentration is self-correcting. Overcapacity depresses Chinese prices and should squeeze Chinese margins, though CATL reported an 18 percent operating margin in 2025 (IEA 2026). Solid-state batteries could reset leadership because they require new materials and processes. Sodium-ion uses abundant elements that any country can source, recycling will return materials to the chain, and controls invite substitution. Each of these is real over a decade, but each also runs through Chinese firms today: CATL and BYD lead in sodium and solid-state, and China hosts most of the processing on which recycling and new anodes depend.

Three paths follow. In a Chinese-standard world, China supplies the base layer and most materials, prices keep falling, and other regions assemble packs and vehicles on imported cells. In licensed localisation, Western and Southeast Asian plants run on Korean, Japanese, or approved Chinese processes, at costs that remain higher but are subsidised. In a technology reset, solid-state or new anodes shift advantage toward firms with new processes, though incumbents are best placed to win that race too. The signposts are observable: whether China's suspension lapses on 10 January 2027, the yield and cost of the first solid-state small batches in 2027, the utilisation of sodium-ion capacity, the price gap between Chinese and Western packs, and whether Western plants can license processes without losing credits.

The comparative record supports a selective conclusion. Battery leadership goes to those who control processing and manufacturing at scale, not to those who invent a chemistry. New chemistries change which minerals matter but leave the processing map in place, and the control regime on both sides now restricts the transfer of the know-how that would change it. Western capacity grows where subsidy, licensed technology, and secured materials are combined, and stalls where any one is missing.

For governments, the priority is processing, not mining or cell assembly alone: cathode, anode, and electrolyte capacity, with licences that survive both Chinese controls and domestic credit rules. For manufacturers, the discipline is to choose a chemistry by the supply chain it implies, not by its energy density, and to secure process technology before scale. For investors, the durable margins sit in processing know-how and low-cost cells; chemistry announcements without a manufacturing path rarely translate into market share.

The next battery breakthrough will be announced in a laboratory. Whether it matters will be decided in a factory, and most of those factories, and the materials plants behind them, are in China.