All chapters
Magnetocrystalline anisotropySpin–orbit couplingBrown’s paradoxSingle-domain grainsGrain boundary diffusion

Iron has a large moment and a high Curie temperature, and it is useless as a permanent magnet. Its magnetisation turns whenever a field asks it to.

What separates a magnet from a lump of ferromagnet is not how strong it is. It is how firmly its magnetisation is anchored to a crystal direction.

4.1The property that makes a magnet permanent

  • Magnetocrystalline anisotropy (K₁, J/m³) — the energy difference between magnetising a crystal along its easy axis and along a hard axis. It arises from spin–orbit coupling: the electron spin is coupled to its orbital motion, the orbital shape is fixed to the lattice by the crystal field, and so the spin is indirectly bolted to the crystal.
  • Anisotropy field (μ₀H_A, tesla) — the field required to rotate the magnetisation coherently away from its easy axis, and the theoretical upper limit on coercivity.

μ₀H_A = 2K₁ / J_s

J_s is the saturation polarisation.

4.2Why the answer is always a rare earth

Spin–orbit coupling scales steeply with atomic number, and the 4f charge cloud is strongly aspherical — Nd³⁺ and Dy³⁺ are shaped like flattened or elongated ellipsoids rather than spheres. An aspherical charge cloud sitting in a crystal field has a strongly preferred orientation, and that preference is transmitted straight to the spin.

The 3d metals have the opposite problem: their orbital moment is largely quenched by the crystal field, because 3d orbitals stick out and interact strongly with their neighbours. Little orbital moment means little spin–orbit anchoring means little anisotropy.

Important

Iron’s K₁ is 0.048 MJ/m³. Nd₂Fe₁₄B’s is 4.9 — a hundred times higher. That ratio, not the moment, is what makes one a magnet and the other a transformer core.

4.3Brown’s paradox — nobody gets more than a third

0.00010.0010.010.1110μ₀H / tesla (log)403SmCo₅8 %6.52.5Sm₂Co₁₇38 %7.61.6Nd₂Fe₁₄B21 %1.80.44Sr ferrite24 %0.30.07AlNiCo 523 %0.060.0001Fe (pure)0 %theoretical anisotropy field μ₀H_Aachieved coercivity μ₀H_cJNobody gets more than about a third. The gap is microstructure — defects, misalignment, grain surfaces.
Figure 4.1Brown's paradox: every real magnet achieves only 15–30 % of its theoretical anisotropy field. Magnetisation reversal does not happen by coherent rotation of a whole grain — it nucleates at a defect, a grain surface, or a misaligned region, and then sweeps through. Coercivity is therefore a microstructure property, not just a chemistry one, which is why two magnets of identical composition can differ by a factor of two.

Technical framing

Every real magnet achieves only 15 to 30 per cent of its theoretical anisotropy field. Magnetisation reversal does not happen by coherent rotation of a whole grain — it nucleates at a defect, a grain surface or a misaligned region, and then sweeps through.

Coercivity is therefore a microstructure property, not just a chemistry one, which is why two magnets of identical composition can differ by a factor of two.

4.4Domains, and why a magnet demagnetises itself

A uniformly magnetised block stores a great deal of energy in the external field it creates. It can lower that energy by splitting into domains pointing different ways, at the cost of the energy needed to build domain walls between them. The equilibrium is a balance between the two.

In plain English

Below a critical grain size the wall energy exceeds the saving, and splitting stops being worthwhile. The grain stays single-domain — and a single-domain grain can only reverse by rotating its whole magnetisation against the full anisotropy, which is hard. For Nd₂Fe₁₄B that critical size is about 0.2 to 0.3 µm.

Why this matters

This is why sintered magnet processing is what it is. The powder is milled to a few microns, aligned in a field, pressed and sintered — with the whole process aimed at small, well-isolated, well-aligned grains. Grain size and grain boundary chemistry are the coercivity levers, and both are metallurgy rather than composition.

4.5Grain boundary diffusion — using dysprosium only where it does work

Reversal nucleates at grain surfaces. So the high-anisotropy heavy rare earth is only needed at grain surfaces — not distributed through the bulk, which is what conventional alloying does.

4.5.1The process

Coat a sintered magnet in a Dy or Tb compound and heat-treat it, letting the heavy element migrate along grain boundaries and form a thin high-anisotropy shell on each grain.

The result: coercivity gains equivalent to 2 to 6 wt % bulk dysprosium, using well under 1 wt %, with almost no loss of remanence because the grain interiors stay pure Nd₂Fe₁₄B.

Important

It is the single most consequential process innovation in magnets in twenty years, and it exists because of a microstructural insight rather than a new material. Chapter 10 covers what it did to heavy rare earth demand.

Quick check: test yourself

1.Iron has a bigger moment than neodymium and a higher Curie temperature. Why is it not the magnet?

Show answer
Because its magnetisation is not anchored to a crystal direction. Anisotropy, not moment, makes a magnet permanent, and iron’s K₁ of 0.048 MJ/m³ is a hundred times below Nd₂Fe₁₄B’s 4.9. Its 3d orbital moment is largely quenched by the crystal field, so there is almost no spin–orbit anchoring.

2.Two magnets have identical composition and differ in coercivity by a factor of two. How?

Show answer
Microstructure. Reversal nucleates at defects, grain surfaces and misaligned regions rather than by coherent rotation, so grain size, grain alignment and boundary chemistry dominate. That is Brown’s paradox — nobody achieves more than about a third of the theoretical anisotropy field, and how close you get is a metallurgy question.

3.Why does grain boundary diffusion save so much dysprosium?

Show answer
Because reversal starts at grain surfaces, so that is the only place high anisotropy is needed. Coating and heat-treating puts a thin Dy or Tb shell on each grain instead of distributing it through the bulk, giving coercivity equivalent to 2–6 wt % using under 1 wt % — and leaving remanence almost intact because grain interiors stay pure Nd₂Fe₁₄B.

Chapter summary

Frequently asked questions

Why is iron useless as a permanent magnet despite being strongly ferromagnetic?+

Because its magnetisation turns whenever a field asks it to. Being ferromagnetic means having a large moment and a high Curie temperature; being a permanent magnet means having the magnetisation anchored to a crystal direction so it resists turning. Iron’s magnetocrystalline anisotropy constant K₁ is 0.048 MJ/m³ against Nd₂Fe₁₄B’s 4.9 — a hundred times higher. Anisotropy, not strength, is what makes a magnet permanent.

Why does high anisotropy always come from a rare earth?+

Two reasons that compound. Spin–orbit coupling scales steeply with atomic number, and the 4f charge cloud is strongly aspherical — Nd³⁺ and Dy³⁺ are shaped like flattened or elongated ellipsoids rather than spheres. An aspherical charge cloud in a crystal field has a strongly preferred orientation, and that preference is transmitted straight to the spin. The 3d metals have the opposite problem: their orbital moment is largely quenched by the crystal field because 3d orbitals stick out and interact strongly with neighbours.

What is Brown’s paradox?+

Every real magnet achieves only about 15 to 30 per cent of its theoretical anisotropy field. The reason is that magnetisation reversal does not happen by coherent rotation of a whole grain — it nucleates at a defect, a grain surface or a misaligned region, and then sweeps through. Coercivity is therefore a microstructure property, not just a chemistry one, which is why two magnets of identical composition can differ by a factor of two.

How does grain boundary diffusion work, and why does it matter so much?+

Because reversal nucleates at grain surfaces, the high-anisotropy heavy rare earth is only needed at grain surfaces — not distributed through the bulk, which is what conventional alloying does. Grain boundary diffusion coats a sintered magnet in a Dy or Tb compound and heat-treats it so the heavy element migrates along grain boundaries and forms a thin high-anisotropy shell on each grain. The result is coercivity equivalent to 2 to 6 wt % bulk dysprosium using well under 1 wt %, with almost no loss of remanence because the grain interiors stay pure Nd₂Fe₁₄B.

Reviewed by

SG

Sahil Goyal

Co-founder, Wingzman

LinkedIn
SG

Sourabh Goyal

Co-founder, Wingzman

LinkedIn

Magnets and Motors is an original educational series on permanent magnets and electric machines. Datasheet ranges, temperature coefficients and efficiency figures are representative standard-condition literature values, not measurements of a specific product, and sources differ on several. Always verify against the specific magnet and lamination datasheets in use before making design, procurement or certification decisions.