What is in focus: two axes, energy density and cost
High-nickel ternary (NCM) and lithium iron phosphate (LFP) give different answers even within the same lithium-ion cell. NCM leads on energy density, LFP on cost, cycle life and safety, and which one is “right” depends on which axis an application prioritizes. This article sets out the structural difference between the two chemistries, the tradeoff in numbers, and where the patent landscape around them stands today.
Starting with the industry context, LFP was once a niche chemistry but has rapidly gained ground. In its Global EV Outlook 2024, the IEA (International Energy Agency) found that in 2023 LFP supplied more than 40% of global EV battery demand by capacity. Regional differences are large, however: the IEA explains that about two-thirds of EV sales in China used LFP, while in Europe and the United States the share stayed below 10% and high-nickel chemistries dominated.
The tension is that this choice is not fixed. In its Q3 2021 update, Tesla stated that it was shifting standard range vehicles to LFP chemistry globally, and the cost gap changes year by year. This article answers three questions: where does the difference between the two chemistries come from, what is gained and lost in numerical terms, and what should be checked from a patent perspective when choosing between them?
Basic principle: a chest of drawers and a honeycomb
The difference begins with the crystal structure of the cathode active material (the cathode material that stores and releases lithium). NCM (nickel, cobalt, manganese) uses a layered structure, while LFP (lithium iron phosphate, LiFePO4) uses an olivine structure. The layered structure has wide spaces through which lithium ions can move, so it stores more energy per unit weight and volume. The olivine structure is held together by a tightly bound phosphate (PO4) framework and is thermally more stable.
Source: reconstructed by the editorial team from the original article (concept diagram)
By analogy, the layered structure is a chest with wide drawers. The drawers are wide, so many items (lithium) can be stored and taken out quickly, but if they are emptied too far the frame starts to wobble. Raising the nickel ratio widens the drawers further but makes it harder to keep the frame stable. The olivine structure is closer to a honeycomb. Each cell is small, but the framework is sturdy and does not easily collapse even when heated.
Translated into practical variables, this structural difference can be summarized as follows.
- Energy density: layered NCM has the advantage. This is the reason to choose it for vehicles where driving range is key
- Thermal stability: olivine LFP has the advantage. Its thermal runaway (a chain reaction of heat generation inside the cell) onset temperature is relatively high
- Raw-material sensitivity: NCM is sensitive to nickel and cobalt prices, while LFP uses neither cobalt nor nickel
Comparison: the tradeoff in numbers
The ranges in the table below are the approximate bands given in the original article based on publicly available general sources. They vary widely by cell design and manufacturer and should not be read as the specification of any particular product.
| Item | NCM (e.g., NCM811) | LFP |
|---|---|---|
| Energy density (cell level, Wh/kg) | About 250-280 | About 150-180 |
| Cycle life (cycles) | Around 1,000-2,000 | Around 3,000-6,000 |
| Thermal runaway onset temperature | Relatively low | Relatively high |
| Sensitivity to cobalt and nickel cost | High | Low (no cobalt) |
Source: compiled by the editorial team from the public general sources cited in the original article (includes estimates)
What this table says: on the ranges given, NCM offers about 1.5 times the energy density of LFP, while LFP offers roughly 2 to 3 times the cycle life. The numbers confirm a structure in which gaining on one axis means giving up on the other.
Multiplying the two figures gives a different picture. If the total energy a 1 kg cell can store and deliver over its life is roughly estimated as “energy density x cycle count,” the table’s ranges give about 250,000-560,000 Wh for NCM and about 450,000-1,080,000 Wh for LFP. This is a simple calculation that ignores depth of discharge and end-of-life criteria, but it puts numbers to the intuition that NCM has the advantage where a single charge must go far, while LFP has the advantage where lifetime throughput matters, as in ESS that charge and discharge several times a day. In other words, LFP’s weakness of “lower energy density” can turn into a strength of “larger lifetime throughput” depending on the application.
It is too early, however, to draw conclusions from cell figures alone. On March 29, 2020, BYD unveiled the “Blade Battery,” in which LFP cells are made in a long blade shape and placed directly in the pack, and announced that pack space utilization increased by over 50% compared with conventional LFP block batteries. In the same announcement BYD stated that in a nail penetration test the Blade Battery’s surface temperature stayed at 30-60°C, while a ternary battery exceeded 500°C. This is an attempt to close part of the cell-level energy density gap through pack structure design.
| Priority | Favored chemistry | Typical applications |
|---|---|---|
| Energy density first | NCM family | Range-focused premium EVs |
| Cost, cycle life and safety first | LFP | Entry-level EVs, ESS (energy storage systems) |
| Common variables | Raw-material prices, manufacturing yield, thermal management design, pack structure | |
Source: compiled by the editorial team from the original article, IEA Global EV Outlook 2024, Tesla Q3 2021 Update and the BYD announcement (2020-03-29)
What this table says: choosing a chemistry is a matter of priorities rather than absolute superiority, and common variables such as pack structure and raw-material prices keep shifting the boundary. The IEA found that the price premium of NMC batteries over LFP was below 25% in 2023, down from 50% in 2021.
Patent perspective: the basic composition patents have already expired
The basic LFP patent: US5910382A, which claims olivine LiMPO4 (M being a first-row transition metal) cathode materials, was invented by John B. Goodenough and others and filed by the University of Texas System, with a priority date of April 23, 1996, a filing date of April 21, 1997, and a grant date of June 8, 1999. The current owner is listed as Hydro-Quebec, and according to Google Patents it expired at the end of its term (Expired – Lifetime) on April 21, 2017.
The basic NCM patent: US6964828B2, which claims layered cathode materials of composition Li[M1(1-x)Mnx]O2, was filed by 3M and invented by Zhonghua Lu and Jeffrey R. Dahn, with a filing date of April 27, 2001, and a grant date of November 15, 2005. The current owner is listed as Umicore, and according to Google Patents its term expired on May 14, 2022.
The two patents also claim in contrasting ways. US5910382A broadly covers the olivine family with the general formula LiMPO4, while US6964828B2 puts into its claim, along with the range of the manganese ratio x, a structural-stability condition that the O3 layered structure does not transform to spinel even after 100 charge-discharge cycles. Even among patents that pre-empt a basic composition, the difficulty of designing around them differs depending on whether the scope is built on “formula” alone or on “formula plus performance or structural conditions.”
Source: reconstructed by the editorial team from the Google Patents bibliographic data of US5910382A and US6964828B2
Insight: the patent risk in choosing a chemistry has moved from “basic composition” to “improvement technologies.” Because representative basic composition patents for both chemistries have expired, the decision between NCM and LFP is less likely than in the past to be blocked by a basic composition patent. Instead, competition over claim scope has moved to improvement technologies such as particle design, surface coating, doping and manganese addition, and to cell and pack structures such as the Blade Battery. FTO (freedom-to-operate) reviews should likewise narrow their target from “which chemistry” to “through which process and structure it is implemented.”
Design-around points: in areas where the basic composition patents have expired, a design that stays within the prior art (technology already public and free for anyone to use) becomes the first-line avoidance route. The approach is to take the compositions and processes at the level of the expired patents’ examples as a baseline and run a separate search for each improvement element added on top.
Design-around scenario (counterargument): it is hard, however, to conclude that “basic patent expired = free to practice.” Mass-production cells almost always include improvement elements such as carbon coating, particle size control and dopant elements, and these areas are likely to be densely covered by patents still in force. Even for the same composition as an expired patent, infringement risk can remain if separate process or use patents are still alive. In the end, the avoidance potential can only be judged by comparing each process element one-to-one against the claims of patents in force.
Limits and outlook
The optimistic outlook is that the two chemistries will coexist, dividing roles by application. The view is that the pattern of NCM for premium vehicles and LFP for entry-level vehicles and ESS will hold for some time. There is also a contrary scenario. The IEA refers both to efforts to raise LFP energy density by increasing its manganese content and to efforts to cut NCM cost by raising the manganese share. If each side narrows the other’s weakness, the very dichotomy of “energy density versus cost” may blur.
The strategic alternative is a dual portfolio that does not tie every product family to one chemistry but designs chemistry separately by product line. On the IP side too, improvement patents for both chemistries must be monitored separately to reduce FTO gaps when switching chemistry due to raw-material prices or regulatory changes.
Practical action: planning and IP staff facing a chemistry decision should extract from candidate cell specifications a list of improvement elements rather than basic composition (coating material, dopant elements, particle size distribution, whether manganese is added) and prepare a comparison table with patent-in-force search results attached for each element.
Items requiring verification
- Individual primary sources and measurement conditions for the ranges in the comparison table (energy density, cycle life)
- Expiry dates of the foreign family members (Europe, Japan, etc.) of US5910382A and US6964828B2
- Status of patents in force in improvement areas such as LFP carbon coating and doping
Sources
- IEA, Global EV Outlook 2024 – Trends in electric vehicle batteries
- Tesla, Q3 2021 Quarterly Update
- BYD press release, BYD’s New Blade Battery Set to Redefine EV Safety Standards (2020-03-29)
- Google Patents, US5910382A – Cathode materials for secondary (rechargeable) lithium batteries
- Google Patents, US6964828B2 – Cathode compositions for lithium-ion batteries
Related Reading — LFP vs. NCM: Not Better or Worse, Just Different Optimization Axes · High-Nickel Cathodes: What Is the Price of Energy Density? · Electrolyte, From Liquid to Solid — The Safety-vs-Interfacial-Resistance Tradeoff