How Does the Carbon Layer “Grow” onto the Surface of LFP? (Part 1)

Lithium iron phosphate (LiFePO4) is an insulator. To improve its electronic conductivity, scientists coat each LFP particle with a carbon shell only a few nanometers thick. This carbon layer is formed by mixing an organic carbon source into the precursor; in the sintering furnace, it simultaneously pyrolyzes, migrates, and deposits, ultimately “growing” into a conductive carbon shell.

So, from raw material to shell, what steps does the carbon layer onto LFP actually undergo?

At each step of this roadmap, carbon changes its form of existence.

Step	Temperature	What carbon is doing	Form change
Step 1 · Placement	Room temperature	Mixed into the precursor and finds its position	Dissolved/dispersed organic matter
Step 2 · Melting/Spreading	~200°C	Melts, spreads, and wets the particles	Solid → liquid film
Step 3 · Decomposition	~500°C	Pyrolyzes; deoxygenation and dehydrogenation leave a carbon skeleton	Liquid film → amorphous carbon
Step 4 · Assembly	~600°C	Graphene microcrystals nucleate, grow, and coalesce	Amorphous carbon → continuous carbon shell
Step 5 · Consolidation	~800°C	sp³ → sp² rearrangement; begins to conduct	Insulating carbon → conductive carbon
Step 6 · Cooling	Cooling	Core–shell structure + carbon-bridge interconnection	Finished LFP/C product

Step 1 · Placement: First, Put the Carbon Source Next to the Particles

Core of this step: where the carbon layer ultimately grows depends on where the organic matter has positioned itself before heating. Mixing directly determines whether the carbon layer can be continuous.

The carbon source is not added afterward; it is dispersed at the molecular level around the LFP particles during the precursor stage. A common industrial practice is to dissolve the carbon source (glucose, sucrose, citric acid, PVA, polyethylene glycol, etc.) together with the precursor into a slurry, mix uniformly by ball milling or the sol–gel method, and then dry into powder or granulate. After drying, the organic matter coats the outside of the precursor as a solid film or fine particles—equivalent to “pre-sugaring” every future LFP particle in advance.

One easily overlooked detail: if the precursor contains water of crystallization, it will clash with carbon-source pyrolysis during heating.

Taking the FePO4⋅2H2OFePO4​⋅2H2​O system as an example, the TG-DTA curve (Figure 2) shows that the precursor undergoes major weight loss at 160–400°C, with a strong endothermic peak near 250°C in DTA. This weight-loss region simultaneously includes the release of water of crystallization, decomposition of oxalic acid/glucose, and reduction of Fe3+Fe3+ to Fe2+Fe2+. Water and carbon are released in the same temperature window; gas evolution can sweep away part of the carbon and disrupt the uniformity of the carbon layer.

Therefore, precursor selection and dehydration pretreatment are the first checkpoint for carbon-layer continuity.

Fig. 1. The TG/DTA curve for the precursor recorded from room temperature to 700°C at a heating rate of 5°C min⁻¹ in Ar atmosphere. WeChat Official Account: Tie Li Yan Xi Lu.

(Data source: Peng et al., Journal of Power Sources 196 (2011) 2841–2847, Nankai University, No. 00029)

Figure 2 / TG-DTA curve of the precursor. 160–400°C is the main weight-loss region, indicating that water removal, carbon-source decomposition, and Fe reduction overlap in the same temperature window. Source: Peng et al., J. Power Sources 2011.

Step 2 · Melting/Spreading: The Carbon Source First “Flows” Over the Particle Surface

Core of this step: whether the carbon layer can conform depends on whether the carbon source can “flow” over the entire particle surface at this step. The places where carbon ultimately cannot cover are often those the liquid film failed to reach.

As the temperature rises to 150–200°C, most small-molecule carbon sources begin to melt—glucose at about 160°C and sucrose at about 197°C. The viscosity of the molten organic matter drops sharply; it capillary-infiltrates along particle gaps and wets the surfaces, wrapping the precursor particles in a continuous liquid film.

The more continuously the liquid film spreads, the more continuous the carbon shell formed upon subsequent carbonization. This is also why mixing and drying processes are so critical—they determine how much area the liquid carbon source can cover.

In the 150–350°C range, the molten carbon source also undergoes caramelization and dehydration condensation. Taking glucose as an example, its decomposition pathway is roughly:

These viscous intermediates glue the particles together, forming a plastic encapsulating layer that has not yet carbonized but has already set up the carbon skeleton.

(Mechanistic interpretation based on: Wang & Sun 2011 review, No. 00008)

Step 3 · Decomposition: The Organic Matter “Burns Itself Off,” Leaving Only the Carbon Skeleton

Core of this step: the carbon source is disassembling itself—driving off oxygen and hydrogen as small molecules and leaving carbon atoms to rebuild a framework. But the carbon at this stage is still insulating; it is only a “carbon precursor.”

At 200–500°C, the carbon source undergoes intense pyrolysis: C–C, C–H, and C–O bonds break; oxygen and hydrogen escape as small molecules such as CO2, CO, H2O, and CH4; the remaining carbon atoms undergo aromatization and rearrangement to form an amorphous carbon skeleton linked by aromatic rings.

At this point, the main carbon structure is predominantly sp³-coordinated, and its electronic conductivity remains poor. However, it has already “replicated” the particle shape, preparing the material for subsequent shell formation.

During pyrolysis, one side effect is worth noting: the released water vapor can blister the viscous carbon layer, leaving pores in the carbon shell. This is precisely why in situ carbon coating increases the specific surface area of the material—the pores are not added externally; they are “blown” out during carbonization.

Different carbon sources decompose at different rates (Figure 3). The three TG curves show that glucose, PEG, and phenolic resin differ significantly in weight-loss temperature and carbon residue yield. Here only one conclusion is drawn: the carbon source not only provides carbon; it determines when, how fast, and in what form carbon is released—different release rhythms lead to different deposition processes.

(Data source: Liu et al., Journal of Alloys and Compounds 2025, 1022, No. 00172)

Fig. 4. (a) TG curves of glucose, PEG and PF. (b) XRD patterns of LFP-3 and LFP-4 samples. (c,d) SEM images of LFP-3 and LFP-4 samples. (e,f) Galvanostatic charge/discharge profiles of LFP-3 and LFP-4 samples for the first three cycles.

Figure 3 | Comparison of TG curves and residual carbon morphology for different carbon sources. The difference among carbon sources illustrates only one point: the rhythm of carbon release differs, and the deposition process changes accordingly. Source: Liu et al., J. Alloys Compd. 2025.

Step 4 · Assembly: The Carbon Layer onto LFP Is Built from Individual Graphene Nanocrystallites

The key point: The carbon layer is not formed by carbon atoms depositing one by one. Instead, graphene nanocrystallites measuring 1–5 nm precipitate from pyrolysis products, undergo nucleation and growth, and randomly stack together to form the carbon layer. Furthermore, the thickness of the carbon layer varies across different crystal planes.

Stay tuned for Part 2 in our next article.