Pauli Paramagnetism vs Curie Paramagnetism

paultsui

As far as I am aware, Pauli Paramagnetism applies to electrons in bands while Curie Paramagnetism applies to localised electrons. Pauli Paramagnetism is usually much weaker because only the electrons near the Fermi Surface can change its spin to align with the magnetic field. On the other hand, since the electrons in a Curie Paramagnet is free, therefore all of them can change its spin - leading to a much higher magnetic susceptibility.

For example, since the conduction electrons in a metal are delocalised to from bands, I would expect metals to follow Pauli Paramagnetism. While electrons in a Mott insulator are localised, so they follow Curie Paramagnetism.

Everything seems to be fine to me until I come across the mean field theory, where one can show that Ferromagnets do behave like Curie Paramagnets in high temperature. This baffles me because if one thinks of iron, which is obviously a metal and have bands, I would expect it to behave like a Pauli paramagnet even in high temperature, but definitely not Curie Paramagnets. Could someone explains where did my logic go wrong?

PietKuip

Local electron correlation of the iron 3d electrons. There is still a Mott-Hubbard U and there are still local exchange interactions on the iron atoms, which gives local moments. The electronic structure does not really change much at the Curie temperature.

There is no good way to calculate this.