Why Cathodes Are Transition Metals
Variable oxidation state is a d-orbital property, crystal field splitting sets the voltage, and the oxygen 2p band sets a ceiling that no amount of engineering raises.
The Periodic Table of the EV · Part 2 — The Cathode · Chapter 3 · 20 min read
Part 2 — The Cathode
Why the d-block is mandatory
One of a cathode’s five jobs cannot be done by any main-group element. That single requirement decides the whole chemistry.
The anode side of this story is about one element. The cathode side is about a whole block of the periodic table, and the reason is a single requirement that main-group elements physically cannot satisfy: a cathode must accept and release electrons reversibly, thousands of times, at high potential, without falling apart.
5
Oxidation states manganese accesses
Δ_o
The splitting that sets voltage
~4.3 V
Where the O 2p band starts to matter
54.5 pm
Low-spin Co³⁺ — small and rigid
3.1 — The five jobs a cathode has to do
- •Accept and release Li⁺ reversibly — a structural requirement.
- •Accept and release electrons reversibly — a redox requirement.
- •Do the second at a high potential, because cell voltage is cathode potential minus anode potential.
- •Conduct electrons well enough to work at all.
- •Not fall apart when half its lithium has been removed.
Requirement two is where the d-block stops being one option among several and becomes mandatory.
3.2 — Variable oxidation state — the d-orbital advantage
Oxidation state is the notional charge an atom would carry if all its bonds were fully ionic. A redox couple is a pair of oxidation states an element cycles between, such as Co³⁺/Co⁴⁺.
Main-group elements have essentially fixed oxidation states. Sodium is always +1, magnesium always +2, aluminium always +3. Removing another electron means breaking into a filled shell, which is energetically prohibitive — so they cannot cycle.
Transition metals have partially filled d orbitals. The 3d and 4s levels sit close in energy, and the d orbitals are relatively contracted and poorly shielding, so successive ionisation energies rise gently rather than sharply. The result is that manganese alone accesses +2, +3, +4, +6 and +7.
Important
This is the entire reason cathodes are made of d-block metals. They can absorb an electron on discharge and release it on charge, thousands of times, without structural collapse. Nothing in the s-block or p-block can.
3.3 — Crystal field theory — where the voltage comes from
Crystal field theory models what happens when the five d orbitals of a transition metal ion — degenerate, meaning equal in energy, when isolated — are surrounded by ligands, here O²⁻ ions. Orbitals pointing at the ligands are destabilised by electrostatic repulsion; those pointing between them are not.
In octahedral coordination — six oxygens around one metal, the standard geometry in a layered oxide — the five orbitals split into two sets:
- •t₂g (d_xy, d_xz, d_yz) — pointing between the oxygens, and therefore lower in energy.
- •e_g (d_z², d_x²−y²) — pointing at the oxygens, and therefore higher in energy.
The energy gap between them is Δ_o, the octahedral crystal field splitting parameter.
Why this matters
Δ_o sets the redox potential. The energy of the d level being emptied determines how much energy is released per electron, and therefore the voltage.
Where the common couples sit
| Redox couple | Host | Approximate potential vs Li/Li⁺ |
|---|---|---|
| Co³⁺/Co⁴⁺ | LiCoO₂ layered | ~3.9 V |
| Ni³⁺/Ni⁴⁺ | LiNiO₂ layered | ~3.8 V |
| Mn³⁺/Mn⁴⁺ | LiMn₂O₄ spinel | ~4.1 V |
| Fe²⁺/Fe³⁺ | Simple oxide | ~3.0 V |
| Fe²⁺/Fe³⁺ | LiFePO₄ olivine | 3.45 V — see chapter 5 |
3.4 — High-spin, low-spin and structural stability
Δ_o also determines whether d electrons pair up in the lower set or spread out to maximise unpaired spins. If Δ_o exceeds the energy cost of pairing two electrons in one orbital, they pair — low-spin. If not, they spread — high-spin.
In plain English
Spin state is not an abstraction. It sets ionic radius, and ionic radius sets structural stability. Low-spin Co³⁺ is t₂g⁶ — all six electrons paired in the low set — which makes it compact at 54.5 pm and rigid. That is why LiCoO₂ is the best-behaved layered oxide structurally, and a large part of why cobalt has been so hard to eliminate.
The same logic runs the other way for manganese. High-spin Mn³⁺ is d⁴ with a single electron in the doubly degenerate e_g set, and that unequal occupancy triggers the Jahn–Teller distortion that is the subject of chapter 4.
3.5 — The oxygen 2p band and the voltage ceiling
In a solid, discrete metal d levels broaden into bands. The Fermi level is the energy up to which states are filled.
Charging a cathode means lowering its Fermi level, because you are removing electrons. Lower it far enough and it drops below the top of the oxygen 2p band — and at that point you begin pulling electrons off oxygen rather than off the metal.
Important
Oxidised O²⁻ is unstable. It can dimerise into peroxide-like species and it can leave the lattice as O₂ gas. That released oxygen is what makes an overcharged NMC cell burn so violently: the cell contains its own oxidiser, so excluding air does not stop it.
Technical framing
This is the fundamental voltage ceiling of oxide cathodes, and it is a periodic-table property — it is set by where the metal d band sits relative to the O 2p band.
Nickel-rich cathodes have their nickel d band closer to the O 2p band, which is exactly why higher nickel content delivers higher capacity and lower thermal stability. It is one trade-off, not two independent facts, and no amount of pack engineering separates them.
Quick check: test yourself
1.Why can’t aluminium be a cathode redox centre even though it is cheap, light and abundant?
Show answer
2.Why does more nickel mean both more capacity and more danger?
Show answer
3.What does low-spin Co³⁺ actually buy a layered oxide?
Show answer
Chapter summary
- ✓A cathode must cycle electrons reversibly at high potential — and only partially filled d orbitals allow that, because main-group elements would have to break into a filled shell.
- ✓Crystal field splitting (Δ_o) separates the d orbitals into t₂g and e_g sets, and the energy of the level being emptied is what sets the voltage.
- ✓Spin state follows from Δ_o and decides ionic radius, which decides structural stability. Low-spin Co³⁺ is the reason LiCoO₂ behaves so well.
- ✓Lower the Fermi level below the oxygen 2p band and the cathode starts oxidising its own oxygen, releasing O₂. That is the hard voltage ceiling for oxide cathodes.
- ✓Nickel-rich cathodes trade thermal stability for capacity because both properties trace to the same band alignment.
Frequently asked questions
Why are lithium-ion cathodes made from transition metals?+
Because a cathode must accept and release electrons reversibly at high potential, thousands of times, without structural collapse — and only partially filled d orbitals allow that. Main-group elements have essentially fixed oxidation states: removing another electron means breaking into a filled shell, which is energetically prohibitive. In transition metals the 3d and 4s levels are close in energy and the d orbitals are contracted and poorly shielding, so successive ionisation energies rise gently rather than sharply. Manganese alone accesses +2, +3, +4, +6 and +7.
What is crystal field splitting and how does it set battery voltage?+
In an isolated ion the five d orbitals have equal energy. Surround the metal with six oxygens in octahedral coordination and the orbitals pointing directly at the oxygens (the e_g set) are pushed higher in energy by electrostatic repulsion, while those pointing between them (the t₂g set) sit lower. The gap is Δ_o. The energy of the d level being emptied on charge determines how much energy is released per electron, and therefore the voltage — which is why Co³⁺/Co⁴⁺ sits near 3.9 V, Ni³⁺/Ni⁴⁺ near 3.8 V and Mn³⁺/Mn⁴⁺ in spinel near 4.1 V.
Why do nickel-rich cathodes release oxygen?+
Charging a cathode lowers its Fermi level by removing electrons. Lower it far enough and it drops below the top of the oxygen 2p band — at which point you begin pulling electrons off oxygen rather than off the metal. Oxidised O²⁻ is unstable, can dimerise to peroxide-like species and can leave the lattice as O₂ gas. Nickel-rich cathodes have their nickel d band closer to the O 2p band, which is exactly why more nickel means both higher capacity and lower thermal stability. It is one trade-off, not two independent facts, and it is why an overcharged NMC cell burns so violently: it contains its own oxidiser.
What does high-spin versus low-spin mean for a cathode?+
Whether d electrons pair up in the lower t₂g set or spread out to maximise unpaired spins, decided by whether Δ_o exceeds the electron pairing energy. It matters because spin state sets ionic radius, which sets structural stability. Low-spin Co³⁺ is t₂g⁶ — all six electrons paired in the low set — making it compact at 54.5 pm and rigid. That is why LiCoO₂ is the best-behaved layered oxide structurally, and a large part of why cobalt has been so hard to eliminate.
Reviewed by
The Periodic Table of the EV is an original educational series on the materials science of electric vehicles. All values are standard-condition literature figures for representative materials, not measured data from a specific product, and sources differ on several of them. Always verify against the specific material datasheet in use before making design, purchasing or certification decisions.