Magnetism

Types of magnetic systems Pauli paramagnetism in metals Landau diamagnetism in metals Larmor diamagnetism in insulators Ferromagnetism of electron gas Spin Hamiltonian Mean field approach Curie transition

Magnets
Zero external field Paramagnets Finite external field

Diamagnets

Ferromagnets Antiferromagnets Ferrimagnets

Pauli paramagnetism
Let us first look at magnetic properties of a free electron gas. Electron are spin-1/2 particles In external magnetic field B Zeeman splitting

Pauli paramagnetism
= p 2 / 2m e B / 2mc
#of majority spins: #of minority spins:

= p 2 / 2m + e B / 2mc
d3 p f ( , ) (2 )3

N , = V

= p 2 / 2m e B / 2mc

= p 2 / 2m + e B / 2mc
- minority spins Fermi level

Magnetization (magnetic moment per unit volume):

M=

e ( N N ) 2Vmc

: aligned along the field and proportional to B in low fields

- majority spins

M = B

- magnetic succeptibility

>0

- paramagnetism

Pauli succeptibility
= p 2 / 2m e B / 2mc
+ e B / 2 mc

Landau quantization
A free electron in magnetic field: Schrdinger equation:
2

= p 2 / 2m + e B / 2mc

B z
2

N N =

V ge B g ( )d V 2 e / 2mc 2 mc B

ieA + = 2m c
n, k z

Ay = Bx; Ax = Az = 0

B=1T corresponds to For any fields,

e B / mc = 1K kB

provided m is free electronss mass

Solutions: labeled by two indices

e B / mc

e g 2mc
2

nk (r ) = exp(ik y y + ikz z )n ( x ck y / eB)

n
- wave functions of a harmonic oscillator Energies:

Magnetic succeptibility:

P =

nk = 2 kz2 / 2m + (e B / mc)(n + 1/ 2)
(Landau levels)

- strongly degenerate!!

We quantized momenta transverse to the field

Landau diamagnetism
A free electron in magnetic field: moves along spiral trajectories and create magnetic field themselves. This magnetic field is directed antiparallel to the external one

Electrons in metals
We know that there are diamagnetic metals. How can we explain their existence? Different spectra of electrons in metals and free electrons. Example: renormalized electron mass

Diamagnetism Energy:

m*

E = 2s
n

dk nk f ( nk ) 2
1 e
2

This only affects orbital motion, not Zeeman splitting

Magnetization:

M = V 1E / B

* * P = P ; L = L (m / m* )2

Result for the succeptibility:

L = g = P 3 2mc 3
: paramagnetic!

We can explain paramagnetic and diamagnetic metals! But there is no way we can explain ferromagnetic and antiferromagnetic metals in non-interacting electron model. Try insulators?

Total succeptibility:

= P + L = 2 P / 3 > 0

Larmor diamagnetism
Consider an ionic insulator with filled shells. Electrons are localized at the ions. Electron Hamiltonian:

Larmor diamagnetism
Eg = e2 B2 12mc 2

H =

ieA e BS ; A = B r + + c mc 2m
2

r
i

2 g

e2 ZN 2 2 B r 12mc 2

r small: consider terms with the field as a perturbation. Correction to the energy in the ground state:

Succeptibility:

Eg = g.s H g.s

1 2 E e2 Z = cA r 2 2 V B 6mc 2

Z - ionic charge cA #of atoms per unit volume -

Total spin and total momentum of electrons in a filled shell are zero; 2 only the term with r contributes.

Diamagnetism! If the ionic shells are not filled can get paramagnetic contribution due to other terms. Can explain paramagnetic and diamagnetic insulators, but not ferromagnetic and antiferromagnetic ones!

Eg =

e2 B2 8mc 2

( x
i

2 i

+ yi2 )
g

e2 B2 12mc 2

r
i

2 g

Exchange interaction
Two electrons: antisymmetric wave function! Take an electron level

Ferromagnetism for localized electrons

Try to visualize for localized electrons (atoms; artificial atoms quantum dots; defects etc) The simplest model:

for

Wave function of two electrons:

(r1 , r2 ) =

1 [1 (r1 )2 (r2 ) 2 (r1 )1 (r2 )] 2

2 2

S = 0,1

Discrete electron states; spacing (each level is doubly degenerate)

Energy splitting due to interaction: spin-dependent!

+C J /2

C = dr1dr2U (r1 r2 ) 1 (r1 ) 2 (r2 )

To put an electron into the system costs electrostatic energy U and exchange energy -J Same spin: U + J Energy to pay: Opposite spin: U

J <

Non-magnetic state

J >

Ferromagnetic state

* J = 2 dr1dr2U (r1 r2 )1 (r1 )1* (r2 )2 (r1 )2 (r2 )

H = const JS1S2

Spin Hamiltonian

Itinerant ferromagnetism
If we can not get ferromagnetism with free electrons, try interacting electrons. Hartree-Fock approximation:

Itinerant ferromagnetism
Working the interaction terms out for free electrons (see Advanced Quantum Mechanics, lecture 4) Kinetic energy:
2 3 3N 2 kF Ekin = NEF = 5 5 2m

Eint = g Vint g
2 e2 1 2 drdr ' r r ' i (r ) j (r ') i* (r ) *j (r ') i (r ') j (r ) 2 ij

Potential energy:

3 k Eint = Ne2 F 4
kF kF

NB:

3 N = VkF /(3 2 )

Try now a spin-polarized ground state:

Hartree (direct) interaction

Fock (exchange) interaction. Only exists if spin projections are the same in the states i and j

Kinetic energy loss can be compensated by the potential energy gain! For

k F aB <
2

5 1 2 21/ 3 + 1

- spin-polarized ground state (ferromagnetism!!) Never occurs in real life.

aB =

/ me2

- Bohr radius

Antiferromagnetic ordering
A different situation: a pair of magnetic atoms in an insulating matrix Consider d-electrons, 5 electrons per atom

Antiferromagnetic ordering
2nd order corrections to the ground state: virtual states Ground Virtual

E =

M i E Ei

U ionization energy; t overlap between the atoms

U
E = 25t 2 / U t,U

Unperturbed ground state: either parallel or antiparallel spins.

Ferromagnetic state: no second-order correction (forbidden by Pauli principle) Antiferromagnetic state preferential!

Spin Hamiltonian
Does not work for many atoms but still represents a good model to treat magnetism

Mean field approach

Let us single out one particular spin at i.

H = J ij Si S j
ij

i and j lattice sites

H = JSi S j + (all other spins)

j

Common approximation: only nearest neighbours interact; the same exchange integrals for all bonds Heisenberg model:

Approximation: this spin sees the average field (Weiss field)

h = J

S
j

H = hSi

H = J Si S j
ij

Now we need to calculate the average field self-consistently. Each site has

nearest neighbours. For each neighbour,

J > 0 ( J < 0)

favors ferromagnetism (antiferromagnetism)

Exact solution: only known for 1D chain (Bethe 1931) no magnetism! Can also be treated for high spins (classical) Let us see what we can do with approximate solutions.

chance to be up (parallel to the field) chance to be down (antiparallel to the field) Equation for the field:

P exp(h / 2kBT ) P exp(+ h / 2kBT ) JN h= ( P P ) = JN tanh 2kh T 2 2 B

Mean field approach

h= JN h tanh 2 2kBT

Curie transition
Magnetization close to transition temperature

tanh x x x3 / 3
Solution:
2 h = 12kBTc (Tc T )

Tc =

JN 4k B

Depending on the temperature, either one solution h=0 (no magnetism) or three solutions (ferromagnetism) Three solutions at:

Square-root singularity

JN 1 JN > 1 kBT < 2 2kBT 4

Curies temperature

Numerical solutions: give power laws, but not the square root. Obviously, these are problems of the mean-field approximation. To describe paramagnetic state: Can add the external field

h0

A phase transition between a ferromagnetic and paramagnetic state!

JN h tanh h = h0 + 2 2kBT