How batteries work Chapter 13 Positive electrodes

Chapter 13

Positive electrode materials: layered oxides, spinels and olivine phosphates

Why are today's cells built on NMC or LFP, and what comes next?

Figure 13.1 Three views. First, a table of nine positive electrode materials, LiCoO2, LiMn2O4, LiFePO4, LMFP, NMC622, NMC811, NCA, LNMO and lithium- and manganese-rich NMC, each marked with its structural family (layered, spinel or olivine), with a bar for its typical specific capacity, a tick for its theoretical capacity where a source gives one, and symbols rating cost, energy density, thermal stability and cycle life. Second, the four commercial families, LiCoO2, the nickel-cobalt layered oxides, LiMn2O4 and LiFePO4, compared in six numbers: practical against theoretical capacity, voltage, energy per kilogram of material, tap density, specific surface area and cycle life. Third, bars of the capacity of NMC111, NMC622 and NMC811 full cells against graphite at cycle 5 and cycle 300 for upper cut-off voltages from 4.0 to 4.6 volts, with the percentage of capacity kept.
NMC111NMC622NMC811(third view; the first two carry their own legends)
Positive electrodes side by side, and what nickel buys. Try this: in Chemistries, tap NMC811 and then LFP; in Four in numbers, find the material that lasts longest and the one that packs densest; in the NMC view, compare NMC622 at 4.4 V with NMC811 at 4.2 V.
SourceRatings (cost, energy density, thermal stability, cycle life) are the Faraday Institution’s Box 1, Marie and Gifford 2023; LMFP and LNMO are not rated there. Typical capacities: LiCoO2 135 to 140 mAh/g (Frith, Lacey and Ulissi 2023; Manthiram 2020); LiMn2O4 below 130 and LNMO about 135 (Manthiram 2020); NMC811 about 200 and LMR-NMC about 280 (Marie and Gifford 2023); NCA 200 to 210 in the 3.0 to 4.3 V window (Eshetu et al. 2021); NMC622 178 mAh/g, charged to 4.4 V in full cells at 1C (Jung et al. 2017); for LiFePO4 the dashed bar is the theoretical 170 mAh/g (chapter 4). Conditions differ between sources, so the bars are typical values, not a ranking. Structural families after Manthiram 2020. Theoretical capacities (the ticks) and the whole of the second view after Faruque et al. 2026, Table 4 (practical and theoretical capacity, voltage, tap density, specific surface area, cycle life, safety, environment and use; the review’s “ternary material” is shown as the nickel-cobalt layered oxides, NMC and NCA) and Table 5 (energy per kilogram of material); its Table 5, from another source, gives somewhat different practical capacities, 140, 160, 120 and 150 mAh/g, and average discharge potentials of 3.8, 3.9, 4.0 and 3.5 V for LiCoO2, NMC, LiMn2O4 and LiFePO4, so the two tables should be read as typical values. NMC view: Jung et al. 2017, Table I, full cells against graphite in 1 M LiPF6 in EC:EMC 3:7, 1C, 25 °C, capacity of cycle 5 (pale) and cycle 300 (solid), two cells each.

The other side: positive electrodes

So far we have changed only the negative electrode. Let us now cross to the other side, because in today’s cells it is the positive electrode that holds the cell back: its capacity lags behind that of the negative electrode, and its materials account for around half the cost of a cell.1

Almost every commercial positive electrode belongs to one of three families of oxide, identified in Goodenough’s group in the 1980s: layered oxides such as LiCoO2, spinels such as LiMn2O4, and polyanion oxides such as LiFePO4. Oxides rather than sulfides, because the top of the oxide-ion band lies lower in energy, which lets a cell reach about 4 V instead of under 2.5 V.2

The three differ in how the lithium moves and how tightly the oxygen is held. In a layered oxide, lithium moves in planes between sheets of transition-metal oxide; these materials conduct well and pack densely. The spinel gives lithium three-dimensional paths, which favours power, but in practice it is limited to LiMn2O4, whose manganese dissolves and poisons the graphite.2 In a polyanion oxide, the strongly covalent P–O or S–O bonds weaken the covalency of the metal–oxygen bonds, lower the energy of the metal’s redox couple and so raise the voltage, the inductive effect: Fe2+/Fe3+ works below 2.5 V in Fe2O3 but at 3.6 V in Fe2(SO4)3. Oxygen held by phosphorus makes these materials thermally stable and safe, and iron makes them abundant; but they conduct poorly, so their particles must be small and coated with carbon, and they pack less densely.2

From LiCoO2 to nickel-rich NMC

LiCoO2 gives up only about half its lithium, because beyond that the Co3+/Co4+ band overlaps the top of the oxide band and oxygen leaves the lattice.2 Replacing cobalt with nickel and manganese gives NMC, LiNi1−y−zMnyCozO2, and each metal brings something. Nickel brings the capacity: it can be oxidized from Ni2+ all the way to Ni4+ without much loss of oxygen. Manganese, as Mn4+, stabilizes the structure without taking part in the redox. Cobalt keeps the layers ordered and the kinetics fast; in NCA, aluminium in place of manganese adds thermal stability.2,3

More nickel means that more lithium can be taken out within the same voltage window,4 and NMC811 gives about 200 mAh/g.1 The price is stability. Ni4+ is unstable against the electrolyte; nickel and manganese dissolve, migrate to the graphite and thicken its SEI, and the phase changes at high charge crack the particles and open new surface.2 In one comparison of full cells against graphite, NMC111 and NMC622 cycled stably when charged to 4.4 V and NMC811 only to 4.0 V; above those limits the impedance of the positive electrode grew, from the point where each began to release oxygen from its surface, about 4.7 V vs Li for NMC111 and NMC622 and about 4.3 V for NMC811. The best stable result came from NMC622 charged to 4.4 V, about 650 mWh per gram of NMC (figure 13.1).4 Nickel-rich materials are also less thermally stable and more sensitive to air and moisture, which makes them harder and costlier to manufacture.1,3

Worked example. What does a higher cut-off buy, and what does it cost? Take NMC622 in the full cells of figure 13.1. to 4.2 V: 155.4 mAh/g at cycle 5, 147.3 at cycle 300: kept 147.3/155.4= 95 % to 4.4 V: 177.8 → 166.8 mAh/g: 166.8/155.4=1.07, still 7 % more than the 4.2 V cell gave at cycle 5 to 4.6 V: 203.1 → 79.9 mAh/g: kept 79.9/203.1= 39 % Each extra 0.2 V first buys capacity; past the oxygen-release point it destroys it.

Which material wins depends on the use. Nickel-rich NMC and NCA serve long-range and high-performance cars; LFP, cheaper and longer-lived, has come back strongly, its uptake in electric vehicles doubling since 2020, and it serves grid storage and shorter-range cars.1 Cell and pack design have narrowed its energy gap: state-of-the-art LFP cells rose from 90 Wh/kg in 2010 to 160 Wh/kg in 2022, against about 300 Wh/kg for cells with NMC-type positive electrodes.1

What comes next

Several routes lead on. Single-crystal NMC particles avoid the cracks between primary grains that shorten the life of nickel-rich materials.1 Manganese-rich chemistries balance cost against range: LMFP, LiFePO4 with manganese added, reaches cells of 210 to 240 Wh/kg; LNMO works at about 4.7 V but waits for an electrolyte that survives there; lithium- and manganese-rich NMC offers about 280 mAh/g but loses voltage as it cycles.1,2 In 2023, none of these high-manganese cells had clearly passed technology readiness level 5.5 Conversion materials such as FeF3 and FeS2 promise 700 to 900 mAh/g at 1 to 2.5 V vs Li, at the price of large hysteresis, sluggish kinetics and strain.5 And molecules can take the transition metal out altogether, which is where we end this module.

Four families in numbers

A 2026 review puts numbers on the four commercial families (the second view of figure 13.1). In practice LiCoO2 gives 135 to 140 mAh/g of a theoretical 274, the nickel-cobalt layered oxides 155 to 165 of 270 to 285, LiMn2O4 100 to 115 of 148 and LiFePO4 130 to 140 of 170; the cycle lives it lists run from at least 300 for LiCoO2 to at least 2000 for LiFePO4.6 The powders differ as well. LiCoO2 packs to a tap density of 2.8 to 3.0 g/cm3 and LiFePO4 to only 1.0 to 1.4, while the specific surface area of LiFePO4, 12 to 20 m2/g, is some thirty times that of the others,6 as expected of the small, carbon-coated particles it needs. A loose powder costs volume: per litre, LiFePO4 loses more of its ground than per kilogram.

The review adds detail on each family. The Mn3+ of LiMn2O4 becomes unstable above 55 °C, and doping with Ni, Cr, Fe, Co or Mg and acid-resistant coatings are used to stabilize it. Layered LiNiO2 gives about 200 mAh/g, some 40 % more than LiMn2O4, and Panasonic, adding aluminium or cobalt to stabilize it, reached cells of about 3.1 Ah, 248 Wh/kg and 660 Wh/L. LiFePO4 has a lower global warming potential, toxicity and resource depletion than LiCoO2 and NMC, and leaves fewer by-products when recycled.6 On the review’s own overall score for safety, cost, sustainability and cycle life, LiFePO4 comes first, then LiMn2O4, the nickel-cobalt oxides, and LiCoO2 last.6

A positive electrode is a compromise between capacity, voltage, stability and cost; more nickel buys capacity and spends margin.

What a whole cell delivers

The trade-offs carry through to whole cells. A 2026 review tabulates graphite-LiFePO4 cells at 120 to 160 Wh/kg (90 to 130 Wh/kg as packs), 2000 to more than 5000 cycles to 80 % capacity and 90 to 140 US$ per kWh, and graphite-NMC or NCA cells at 180 to 250 Wh/kg (140 to 200 Wh/kg as packs), 1000 to more than 2000 cycles and 110 to 180 US$ per kWh.6 The same review quotes about 145 US$ per kWh for producing cells and 190 US$ per kWh for packs, against an industry target of 125 US$ per kWh for packs, and a study in which coating the electrodes thicker, from 50 to 100 μm, lowered the cost of a cell by about a quarter.6 Thicker coatings bring the transport limits of chapter 11, so that saving has a price in power.

References for this chapter

Numbers in the text refer to this list. All references of the course, with the chapters that use each one.

  1. 1 J.-J. Marie, S. Gifford, "Developments in lithium-ion battery cathodes", Faraday Insights, Issue 18 (The Faraday Institution), 2023, September 2023. link
  2. 2 A. Manthiram, "A reflection on lithium-ion battery cathode chemistry", Nature Communications 11, 2020, 1550. doi:10.1038/s41467-020-15355-0
  3. 3 G. G. Eshetu, H. Zhang, X. Judez, H. Adenusi, M. Armand, S. Passerini, E. Figgemeier, "Production of high-energy Li-ion batteries comprising silicon-containing anodes and insertion-type cathodes", Nature Communications 12, 2021, 5459. doi:10.1038/s41467-021-25334-8
  4. 4 R. Jung, M. Metzger, F. Maglia, C. Stinner, H. A. Gasteiger, "Oxygen Release and Its Effect on the Cycling Stability of LiNixMnyCozO2 (NMC) Cathode Materials for Li-Ion Batteries", Journal of The Electrochemical Society 164, 2017, A1361–A1377. doi:10.1149/2.0021707jes
  5. 5 J. T. Frith, M. J. Lacey, U. Ulissi, "A non-academic perspective on the future of lithium-based batteries", Nature Communications 14, 2023, 420. doi:10.1038/s41467-023-35933-2
  6. 6 O. Faruque, M. J. Isalm, O. Faruq, M. M. Alam, "A comprehensive review of past and present developments of Li-ion batteries", Discover Sustainability 7, 2026, 881. doi:10.1007/s43621-026-02920-8

Cite this page

Using this page in teaching, a thesis or an article? Cite it as follows; the date is the last update of the course.

Shyma Sajeevan, A. (2026). Positive electrode materials: layered oxides, spinels and olivine phosphates. How batteries work. https://asajeevan.github.io/batteries/positive-electrodes/

↑