# Techniques to Increase Energy Density of LFP Batteries for EVs

Topic radar · 125 documents · 2021-01-01–2026-10-12 · updated 2026-10-22

Canonical: https://slate.greyb.com/radar/techniques-to-increase-energy-density-of-lfp-batteries-for-evs/innovations

## In one minute

### Two shell chemistries do opposite jobs at the particle surface
The clearest particle design in these filings is Yenam's: a 30–70 nm core of LiFe₁₋ₓAlₓPO₄₋ᵧFᵧ inside a 1–5 nm carbon shell, made solvothermally with chelating ligands and carbonised sugars. Aluminium on the iron site widens the lattice and holds the iron oxidation state; fluorine on the oxygen site eases lithium insertion and passivates the surface against corrosion. The material is claimed at ≥170 Wh/kg and ≥160 mAh/g at 0.1C [1].

Svolt changes what the shell is for. Its coating is a lithium-ion conductor, drawn from lithium titanium phosphate and the aluminium-germanium and aluminium-titanium phosphates, ball-milled with the matrix for 1–10 hours and calcined at 300–900 °C. The stated reason is ionic conductivity and low reactivity with the electrolyte [2].

That is a real fork in the design. A carbon shell carries electrons to the surface; a phosphate glass carries lithium through it. Both documents justify their shell by rate and surface stability. Neither claims it adds capacity, which is consistent with where the capacity numbers land.

1. [KR20260033822A](/?paper=KR20260033822A&utm_source=radar&utm_medium=dashboard)
2. [US2023048254A1](/?paper=US2023048254A1&utm_source=radar&utm_medium=dashboard)

### The shipping figure sits near 205 Wh/kg, and it is sold as a manufacturing claim
When CATL launched Shenxing Plus it quoted 205 Wh/kg, presented as almost 8% above the then state of the art. The company's own account of where that came from is not a new cathode: particles graded by size and placed in the optimal position to reach ultra-high compact density, plus a 3D honeycomb anode additive to control swelling. That is the shape of the cell-level claims that have reached vehicles in this record — a number attached to a process, described in the vocabulary of packing.


### The invention has moved from the powder to the line
The filings that claim a density number increasingly claim it for a process. Hydrographenium's is a continuous line — mill, shear-condition, granulate, densify, calcine, then mechano-fuse carbon into the surface — and it labels the output Generation-4. Qinghai Taifeng does the same job earlier in the sequence, pressing and shaping the precursor before it is fired. Both are equipment-and-sequence claims. Neither says anything new about what the crystal can store.


### The volumetric gain comes from packing two particle sizes and welding carbon onto the surface
Wanrun mixes 2–4 µm particles at 10–30% with 0.2–0.4 µm particles at 70–90%, doped with vanadium, titanium and boron across the iron and phosphate sites. The coarse fraction sets a skeleton and the fines fill the gaps between it. Compacted density reaches 2.5–3.0 g/mL.


### Pack parity with nickel was demonstrated, and the company that demonstrated it has left the car business
ONE's Aries II claimed pack energy density and mass within 6% of a leading NCM pack at roughly 25% lower cost, achieved by lightweighting at the pack level rather than by changing the cell. That is the honest version of the LFP parity story: the gap closed in structure and mass, not in electrochemistry.


### The customers are writing manganese into their claims while the suppliers keep optimising iron
GM's cathode claim fixes at least 50 wt% manganese-substituted LFP with up to 50 wt% lithium manganese oxide, ≥95 wt% active material and reversible loading above 4.0 mAh/cm²; its later cell filing is written around an LMFP or NCMA cathode outright. Rivian claims the blend across almost the entire composition range. CALB has paired filings, one on the LMFP cell and one on the transport penalty it buys.


### Blending turns energy density into a function of three variables, and one filing writes the function out
Jiangsu Reliance bounds a cylindrical cell's energy density between two logarithmic functions of the ternary fraction, the charge cut-off voltage and the compaction density. Read as engineering, it says a cell's energy can be estimated from three settings before anything is built.


### The entrants are selling process, and one is selling density rather than origin
Integrals Power has been shipping LFP and LMFP cathode samples to automakers and cell makers in Europe and the US, with a manganese-rich LMFP claimed at up to 20% more energy density than LFP and externally validated on rate performance and cold-weather retention. That is a Western cathode pitch built on watt-hours, not on origin or price — the first in this coverage.


## Overview


LFP's remaining energy density is being won in volume, and the levers are becoming capital and specification questions as much as chemistry ones. The particle is close to spent, manganese substitution is the one route that lifts watt-hours per kilogram, and both routes are now being fenced in by export controls and a raised safety floor.


## Research

### Raising the voltage moves the constraint onto ion transport inside the crystal
CALB's LMFP cell uses both redox couples, manganese and iron, to lift the ceiling to roughly 4.1 V. The engineering sits in the mixture: high-manganese and low-manganese fractions are balanced so the cell does not cut off early and does not degrade structurally at the interface, with performance governed by manganese content, particle size and carbon coating through two named parameters.


### The anode takes the fast-charge load, and it is engineered as a separate problem
Eve Power grows carbon nanofibres on petroleum-coke graphite by chemical vapour deposition from a gas of four carbons or fewer, with an acid wash and a controlled 1–20 °C/min ramp. The composite is claimed to stop lithium plating in an LFP cell under fast charge, at more than 189.9 Wh/kg and above 89.7% retention over 2,000 cycles.


### The solvent is being attacked at the binder, and the drying step survives
Two filings build LFP electrodes without PVDF and without NMP. Fastcap adds binders in sequence — one with the carbon, one in a first solvent, one in a second — before the active material goes in, and uses long multi-wall nanotubes to trap active material in the voids. Nanoramec applies the same sequential-mixing logic to slurry rheology and homogeneity.


### Pack filings buy volume inside a fixed box, and one filing does away with the box
VE Commercial Vehicles stacks prismatic cells in two vertical layers, joins them with busbars and puts a cooling unit between the layers, all to raise volumetric density without changing the outer dimensions. Lihua fills the gaps between the main cells with a second battery set that runs a heater for the main pack and carries its own management system, so the space spent on thermal management also stores energy.


## Market

### The export line has been drawn at the density specifications themselves
China now requires a licence to transfer LFP and LMFP cathode preparation technology abroad, part of a package covering eight battery technologies, in which cathode know-how is described as what sets the ceiling on a cell's energy density.


### The chemistry patents are free; the fight is over process, and it is being fought in Europe
The founding LFP patents expired in 2022, which is why the field opened to entrants without a licence. What rose in their place is process IP: LFP-related patent activity in Europe has surged since 2020, led by CATL, Brunp, ATL and BYD, as those companies protect their position in the market they intend to sell into.


## Strategy

### The remaining cell-level numbers are coming from the anode and the electrolyte
Eve Power's fast-charge invention is an anode product, and it carries the highest cell figure in this innovation set: above 189.9 Wh/kg with lithium plating suppressed over 2,000 cycles. CATL takes the same lever by design, building the anode in two graphitisation layers and holding gram capacity per unit thickness inside fixed bands. Tesla's contribution to this record is an electrolyte — carbon-dioxide-saturated and fluorinated — aimed at stabilising a silicon anode's interphase.


### Pack-level density is being patented by the people who buy cells, not the people who make them
The two enclosure filings come from a commercial-vehicle maker and a power-technology firm, not a cell manufacturer: VE Commercial Vehicles stacks prismatic cells in two layers with an inter-layer cooling unit to raise volumetric density inside unchanged outer dimensions, and Lihua fills the gaps between main cells with a second battery set that runs a heater. The most radical pack document in this set is an individual filing — an LFP and graphite structural panel built into cellulose plies, with piezoelectric pumping to assist ion transport on high-rate discharge.


### The Western route is competing at parity on density, and on cost and origin instead
Sylvatex's precursor-free process is the clearest forward commitment in this record: DOE-backed, moving toward roughly 50 kg/hour through toll manufacturing, with ≥2.5 g/cc press density, ≥156 mAh/g, and an independent techno-economic assessment putting capital cost more than half below conventional North American production.


## Areas

- Cathode Materials & Chemistry · 57 documents
- Cell Design & Manufacturing · 47 documents
- Pack & System Integration · 3 documents
- Market, Policy & Sustainability · 16 documents