A decade of battery patenting Part 3: Contenders, challengers and the road to market

Part 3 of the Battery IP Insights series: a concise map of the major rechargeable chemistries, their trade-offs, leading players and commercial maturity, ahead of a companion analysis of where the patents are being filed.

Every battery is a compromise. No single chemistry is simultaneously the most energy-dense, the cheapest, the safest, the longest-lived and the most abundant in raw materials. Improve one of these metrics and another usually gives up ground.

The art of battery engineering, and of choosing a battery chemistry for a given product, is deciding which of those trade-offs matter most. A long-range electric vehicle (EV) and a grid-storage plant are not solving the same problem, and they increasingly do not use the same chemistry.

This article maps the modern rechargeable battery field, setting aside the legacy lead-acid and nickel chemistries, and groups it into three families: the lithium-ion incumbents that dominate today; the next-generation cell chemistries chasing higher energy or lower cost; and the stationary battery specialists built for the grid. Part 4 of our series will then examine where each of these areas sits in the patent landscape.

The lithium-ion family: One framework, many flavours

“Lithium-ion” is not one chemistry but a family, distinguished mainly by the cathode material (and, in one case, the anode). All share the same “rocking-chair” mechanism, shuttling lithium ions between electrodes. The anode is almost always graphite (carbon), increasingly blended with a small amount of silicon to raise capacity. Silicon is emerging as an anode material in its own right, and blended into or partly replacing graphite, it can push cell energy toward roughly 350 Wh/kg¹. Although, the drawback of increasing silicon content is that the anode swells substantially on charging, so for now it remains a partial rather than wholesale substitution. However, it is the cathode that predominantly varies from one chemistry to the next, and the cathode that sets battery energy density, cost, safety and lifetime. Every one of these cathodes also contains lithium, which the shorthand names below do not always make obvious.

LCO (lithium cobalt oxide), paired with a graphite anode, launched the commercial era of lithium-ion batteries with Sony in 1991. Its high energy density still suits phones and laptops, but high cobalt loading and poor thermal stability have pushed it out of applications like automotives or stationary storage due to the larger number of cells required. LMO (lithium manganese oxide), also paired with graphite, is cheaper and safer but lower in energy and lifetime, and today mostly appears blended with other cathode materials.

NMC (lithium nickel manganese cobalt oxide, LiNiMnCoO₂) and NCA (lithium nickel cobalt aluminium oxide, LiNiCoAlO₂), both paired with graphite or silicon-graphite anodes, are the high-energy workhorses of premium and long-range EVs. Raising the nickel content lifts energy density but reduces thermal stability, and both carry exposure to cobalt and nickel supply chains. These are the mainstays of LG Energy Solution, Samsung SDI, SK On and Panasonic (whose NCA cells power much of Tesla’s range), with CATL also a major NMC producer².

LFP (lithium iron phosphate), again paired with a graphite anode, gives up gravimetric energy density but offers outstanding safety (critical self-heating, the point of no return, at approximately 230 to 260°C for LFP versus approximately 170 to 190°C for NMC³), very long cycle life and the lowest cost per kWh of any lithium battery chemistry, with no cobalt or nickel. Cell-to-pack designs such as BYD’s Blade and CATL’s Qilin have clawed back much of the range gap, and LFP now dominates standard-range EVs and stationary storage. It is the core of CATL, BYD and EVE.

LTO (lithium titanate) is the exception that swaps the anode rather than the cathode: it replaces graphite with a lithium titanate anode. That sacrifices energy density but delivers extraordinary cycle life (tens of thousands of charge-discharge cycles), ultra-fast charging and wide-temperature operation, a niche fit for fast-charge buses, grid-frequency support and cold climates, led by Toshiba’s SCiB and China’s Yinlong⁴.

The manganese middle path

Between the low-cost LFP camp and the high-energy nickel camp, a family of manganese-rich cathodes is opening up a middle path, trading a little of each to reduce reliance on scarce, expensive nickel and cobalt. Two distinct approaches are emerging.

LMFP (lithium manganese iron phosphate) adds manganese to the LFP cathode, lifting cell voltage and energy density while keeping LFP’s safety, long life and freedom from nickel and cobalt, at only a small cost premium. Gotion’s Astroinno cell claims about 240 Wh/kg⁵, and CATL’s M3P variant has been validated with Tesla and is already used in Chinese-market models. The main technical hurdle is manganese dissolving out of the cathode at high temperature⁶.

LMR (lithium manganese-rich) takes a layered-oxide route instead, using a cathode of roughly two-thirds manganese and one-third nickel with virtually no cobalt. General Motors and LG Energy Solution plan to be first to market, with US production of LMR prismatic cells targeted for 2028. They claim LMR cells achieve around a third more energy density than LFP at comparable cost, enough for a 400-mile electric truck, and GM has built a $900 million cell-development centre in Michigan to manufacture them⁷.

The next-generation cell chemistries

Sodium-ion batteries are touted as the near-term disruptor to the current lithium-ion technologies. Sodium-ion batteries typically rely on the same mechanism as lithium-ion batteries, but using abundant, cheap materials, with no lithium, cobalt, nickel or copper. They also offer better cold-weather performance and safety, at the cost of lower energy density. 2025 to 2026 appears to be a commercial inflection point for sodium-ion batteries: CATL’s Naxtra cells reach about 175 Wh/kg⁸, the first mass-production sodium-ion passenger car (a CATL and Changan collaboration) is arriving in 2026, with mass cell production slated for the end of the year, and the first gigawatt-hours are already shipping into stationary storage⁹. HiNa, BYD and Faradion (now part of Reliance) are also scaling. Sodium-ion is aimed at entry-level EVs, two-wheelers and grid storage, not long-range vehicles. Two economic caveats temper the cost story. Sodium-ion needs a hard-carbon anode rather than graphite, and hard carbon is currently more expensive than the graphite used in lithium cells¹; but because sodium does not alloy with aluminium, sodium cells can use cheap, light aluminium current collectors on both electrodes¹, whereas lithium needs heavier, more expensive copper on the anode side¹⁰.

Solid-state batteries are the long-promised leap in energy density. Replacing the flammable liquid electrolyte with a solid conductor removes the main fire risk and, crucially, enables a lithium-metal anode, roughly doubling energy density (claims of 400 to 500 Wh/kg) with faster charging. The barrier is manufacturing rather than the chemistry: the challenges are production yield, an estimated 5 to 10 times cost premium as of mid-2026, and durable solid-solid interfaces. Toyota and Samsung SDI are running pilot lines; QuantumScape (with Volkswagen’s PowerCo) has reported multi-layer cells passing 1,000 cycles in 2024 testing¹¹. More recent disclosures through 2026 have focused on energy density and manufacturing scale-up rather than cycle life, with no updated figure published since. Realistic volume is projected to sit in premium vehicles around 2027 to 2028, with manufacture and deployment scaling toward 2030^{12, 13, 14}.

In practice, the best measured cells today sit closer to 300 Wh/kg (QuantumScape’s most recent disclosed cell, QSE-5, reaches 301 Wh/kg, 844 Wh/L)¹⁵ than to the 400 to 500 Wh/kg the technology ultimately promises. A subtler point is that a solid electrolyte suppresses lithium dendrites (a common problem with lithium-ion batteries that ultimately cause the battery to cease functioning) through interface stability, not merely mechanical stiffness. Dendrites can still creep along grain boundaries, so a hard electrolyte alone is no guarantee against them, but suppression is still valued to improve the battery lifetime. Developers are pursuing several solid-electrolyte families, chiefly sulfides, oxides and, more recently, halides, each trading off conductivity, stability and manufacturability.

A second US developer, Factorial Energy, is pursuing a polymer-based electrolyte rather than QuantumScape’s ceramic one. Its FEST cells, validated with Stellantis and tested by Mercedes-Benz, reach about 375 Wh/kg, and in mid-2026 became the first solid-state cells to be road-tested in a Stellantis vehicle¹⁶. Encouragingly for manufacturing, real production progress is now visible as QuantumScape integrated its scalable Cobra separator process into baseline production in 2025 and inaugurated a pilot production line in early 2026¹⁷. A note of caution is nonetheless warranted. The field has a long history of overclaiming, from cells later found to be conventional lithium-ion to repeatedly slipping timelines, so independently validated results deserve more weight than headline specifications¹⁸.

Lithium-sulfur batteries offer a very high theoretical energy density using cheap, abundant sulfur and no nickel, cobalt or graphite, attractive for weight-critical and applications requiring secure supply chains. Its Achilles heel is the “polysulfide shuttle” that limits cycle life. Cells are already shipping into niches such as drones, defence and space (Lyten, which has absorbed former Northvolt assets) ¹⁹, with Zeta Energy and Theion also active. However, EV-grade cycle life remains some years out²⁰.

Lithium-air batteries have the highest theoretical energy density of all these chemistries, approaching that of gasoline, but is the least mature commercially. Laboratory prototypes (for example at Argonne National Laboratory) have reached around 1,200 Wh/kg over roughly 1,000 cycles²¹, and CATL has publicly flagged it as a long-horizon direction, but practical deployment, if it comes, is a post-2030 prospect²².

The stationary specialists: Built for the grid

For multi-hour and long-duration grid storage, the priorities invert. Gravimetric energy density barely matters when the battery sits in a field. In these applications, what counts is cost per cycle over decades, safety at scale and the ability to decouple power from energy. Three non-lithium families are built around exactly this.

Vanadium redox flow batteries (VRFB) store energy in liquid electrolyte held in tanks, so power (stack size) and energy (tank volume) scale independently, ideal for long durations. The electrolyte does not degrade, giving 20-to-25-year lifetimes and well over 10,000 cycles with a non-flammable, aqueous chemistry²³. Upfront capital cost is high, but lifecycle economics are strong for long-duration use, and the technology has just crossed the gigawatt-hour threshold: Rongke Power’s 200 MW / 1,000 MWh Jimusar project in Xinjiang, the world’s first GWh-scale flow battery, entered operation at the end of 2025²⁴, alongside players such as Invinity and Sumitomo Electric²⁵.

Zinc-based battery chemistries, led by zinc-bromine, target the same long-duration niche with abundant materials, an aqueous non-flammable design and no need for active cooling. The US manufacturer Eos Energy is the flag-bearer, scaling zinc-bromine systems for three to 11-hour applications with federal backing²⁶.

Iron-air batteries push furthest into multi-day storage. It works by “reversible rusting”, oxidising and reducing cheap iron in an aqueous electrolyte, using three of the most abundant materials on Earth: iron, water and air. Round-trip efficiency is low (around 50%)²⁷, and it is far too heavy and slow for vehicles, but the materials are so cheap that it targets roughly 100-hour durations at a fraction of lithium-ion cost. The US company Form Energy is the clear leader, with its first commercial projects and a West Virginia factory now coming online²⁸ and iron-air is the youngest of the grid chemistries but potentially the cheapest for very long durations.

The materials question: Chemistry is also geopolitics

Choosing a battery chemistry is not only an engineering decision, but also a supply-chain and geopolitical one. Each chemistry relies on a different basket of raw materials, and those baskets are unevenly distributed. Nickel and cobalt concentrate the risk in a handful of (often unstable) countries; LFP and sodium-ion largely sidestep it; and manganese-rich cathodes are attractive partly because manganese is cheap and widely available. But one material sits under almost every lithium-ion cell regardless of cathode: graphite, the anode, and the single largest component by weight. China mines roughly three-quarters of the world’s natural graphite and processes an estimated 85 to 90% of the battery-grade material²⁹.

That concentration became a live issue in late 2025, when China placed lithium batteries above 300 Wh/kg, cathode materials and artificial graphite anode materials under export-licensing controls, before suspending the measure the day before it was due to take effect, for one year through November 2026, as part of a wider trade understanding³⁰. This episode underlines how tightly the chemistry map and the minerals map are now intertwined, a theme this series has explored in the context of recycling, and one that helps explain why so much of the battery industry is racing toward chemistries that need less of the contested materials in the first place.

How they compare

The table below summarises the major metrics across the fifteen core chemistries. The figures are indicative, cell-level industry data for 2024 to 2026. The real numbers vary by manufacturer and by pack design, and the emerging chemistries in particular are moving quickly. The Relative cost column tracks broadly with underlying pack-level pricing. LFP and NMC packs averaged $81/kWh and $128/kWh respectively in late 2025, a volume-weighted average of $108/kWh³¹.