The lithium iron phosphate cathode already forms a comparatively well-behaved cathode-electrolyte interphase (CEI) on its own. What LFP does not do on its own is move electrons and lithium ions quickly enough for high-rate use, and that gap is what most of the coating and doping work on this material is actually trying to close. This page looks at the specific interface and near-surface engineering used on LFP: what problem each approach targets, how it works, and what it costs.
A different starting point to nickel-rich cathodes
Most of the CEI literature is written around nickel-rich layered oxides, where the interphase keeps growing because the cathode surface sits at a high enough potential to oxidise the carbonate solvents in the electrolyte. LFP’s redox couple sits at a flat plateau around 3.4 V vs Li/Li+, comfortably inside the stability window of a standard LiPF6 carbonate electrolyte, so that particular growth driver is much weaker here. The trade-off is on the other side of the ledger. LFP’s olivine structure only allows lithium to hop along one crystallographic channel, and the material’s electronic conductivity is poor to begin with, so charge carriers struggle to reach or leave the reaction sites at anything above modest rates. Coating and doping on LFP are mostly aimed at that transport bottleneck, with interface stability as a secondary but related benefit.
What the untreated interphase looks like
Where a CEI does form on LFP, it draws on the same electrolyte breakdown products described on the main CEI page: lithium fluoride, lithium carbonate and lithium phosphates sitting alongside partially decomposed carbonate solvent, in a mix of inorganic and organic species that has a similar chemical vocabulary to the anode’s solid electrolyte interphase, just built by a different set of reactions and, thanks to the low operating voltage, forming a thinner and more stable layer than an equivalent nickel-rich cathode produces over the same number of cycles.
Coatings: closing the conductivity gap
Carbon-based coatings
A thin conductive carbon shell around each LFP particle is the default fix for the electronic conductivity problem: it gives electrons a path to and from the reaction site without requiring them to travel through the semiconducting phosphate itself. The precursor and process used to lay that shell down vary a lot in practice, from carbothermal reduction of a glucose precursor to sol-gel routes using citric acid to rapid microwave-assisted synthesis, and the resulting carbon layer typically runs a few nanometres thick at a few percent by weight, which is enough to change the rate performance considerably without eating much into the cell’s energy density.[1]
Carbon nanotubes are used the same way but for a different reason: instead of a continuous shell, a nanotube network gives the particle bed a three-dimensional conductive scaffold that keeps working even where individual particles lose direct contact with each other during cycling. Composites that combine a thin amorphous carbon coating with a nanotube framework are reported to hold up well at both high discharge rates and low temperature, which is the kind of dual demand a coating chosen for only one property would struggle with.[1]
Metal oxide coatings as HF scavengers
The other major coating family targets a chemical problem rather than a transport one. LiPF6 hydrolyses in the presence of trace moisture and releases hydrofluoric acid, which attacks the LFP surface and drives iron dissolution into the electrolyte. A thin oxide layer, typically alumina, zinc oxide or manganese dioxide applied by atomic layer deposition or chemical precipitation, reacts preferentially with that HF and forms a stable metal fluoride in its place, for example:
Al2O3 + 6HF → 2AlF3 + 3H2O
ZnO + 2HF → ZnF2 + H2O
Alumina applied by atomic layer deposition at only a few nanometres thick is reported to support useful capacity at 10C and to hold up over hundreds of cycles, with the thin, defect-rich film letting lithium through despite alumina’s own wide bandgap.[1] The limitation worth flagging is that this protection is consumed as it works: an oxide scavenger layer depletes over long-term cycling, so its benefit is front-loaded rather than permanent, and it is usually paired with a carbon coating rather than used on its own.
Engineering the CEI directly
A more recent line of work treats the CEI itself as something to be designed rather than just tolerated or scavenged around. A 2025 study from Rice University used a ten-second flash Joule heating process to lay down a carbon coating containing fluorine heteroatoms, and found that the resulting layer pushed the CEI that formed on top of it toward an inorganic-rich, more conductive composition rather than the organic-heavy film that forms on uncoated LFP. The coated cathodes delivered 96 mAh/g at 10C against 58 mAh/g for a commercial LFP reference at the same rate, a substantial jump in usable capacity specifically at fast-charge rates [2]. It is a useful example of the direction this field is heading: rather than only adding a barrier layer, the coating process is used to shape what the CEI becomes.
Engineering the LFP Cathode-Electrolyte Interface
This deserves its own look: treating the CEI itself, or the process that forms it, as the thing being engineered. Three recent papers take genuinely different routes to that same goal, and reading them side by side says more about where this field actually is than any one of them does alone.
Doping: changing the material rather than coating it
Doping works on the bulk and near-surface structure instead of adding an external layer, and shows up in two forms in the LFP literature.
Heteroatom doping of the carbon coating itself, most often with nitrogen, is used to raise carrier density in the shell and create additional active sites at the carbon-LFP boundary, lowering the energy barrier for ions crossing into the particle. Co-doping the carbon with nitrogen and manganese, or nitrogen and sulfur, is reported to go further, reducing microcracking during cycling as well as improving transport, though there is a ceiling on this: too much nitrogen produces thick, agglomerated carbon-nitrogen domains that end up hurting high-rate performance instead of helping it.[1]
Cation doping substitutes a small fraction of the iron sites in the olivine lattice itself, using ions such as magnesium, nickel, cobalt or strontium. The substituted ions expand the lattice slightly, which widens the one-dimensional lithium diffusion channel that otherwise limits how quickly ions can move through the structure. Of the dopants compared in one recent study, nickel substitution gave the highest measured electronic conductivity of the group.[1]
What this adds up to
Strip away the specific materials and the interface engineering done on LFP is addressing three things: getting electrons and lithium ions to the reaction site fast enough for the rates a modern cell needs, keeping HF and the iron dissolution it causes away from the active material for as much of the cell’s life as reasonably achievable, and, in the more recent work, actively shaping the composition of the CEI that forms rather than just building a wall next to it. None of this changes LFP’s fundamental voltage or energy density; what it buys back is rate capability, which is exactly the property LFP is otherwise weakest on. That is also why this work matters most where fast charging is a design requirement rather than a nice-to-have, which for LFP increasingly means daily-driver EVs and stationary storage doing frequent high-power cycling.
References
- Sania Ishtiaq, Abdul Majid, Abdul Qadeer, Mohammad Alkhedher and Niyazi Bulut, Recent progress in carbon coating and surface modification of LiFePO4 cathodes, 2025 RSC Advances 15, 42331-42346
- J. Chen, O. E. Onah, Y. Cheng, K. J. Silva, C. H. W. Choi, W. Chen, S. Xu, L. Eddy, Y. Han, B. I. Yakobson, Y. Zhao and J. M. Tour, Cathode-Electrolyte Interphase Engineering toward Fast-Charging LiFePO4 Cathodes by Flash Carbon Coating, 2025 Small Methods 9(1), e2400680