You are on page 1of 12

Review of Statistical Mechanics

3.1 Microcanonical, Canonical, Grand Canonical
Ensembles
In statistical mechanics, we deal with a situation in which even the quantum state
of the system is unknown. The expectation value of an observable must be averaged
over:
¸O) =

i
w
i
¸i [O[ i) (3.1)
where the states [i) form an orthonormal basis of H and w
i
is the probability of being
in state [i). The w
i
’s must satisfy

w
i
= 1. The expectation value can be written in
a basis-independent form:
¸O) = Tr ¦ρO¦ (3.2)
where ρ is the density matrix. In the above example, ρ =

i
w
i
[i)¸i[. The condition,

w
i
= 1, i.e. that the probabilities add to 1, is:
Tr ¦ρ¦ = 1 (3.3)
We usually deal with one of three ensembles: the microcanonical emsemble, the
canonical ensemble, or the grand canonical ensemble. In the microcanonical ensemble,
18
Chapter 3: Review of Statistical Mechanics 19
we assume that our system is isolated, so the energy is fixed to be E, but all states
with energy E are taken with equal probability:
ρ = C δ(H −E) (3.4)
C is a normalization constant which is determined by (3.3). The entropy is given by,
S = −ln C (3.5)
In other words,
S(E) = ln
_
# of states with energy E
_
(3.6)
Inverse temperature, β = 1/(k
B
T):
β ≡
_
∂S
∂E
_
V
(3.7)
Pressure, P:
P
k
B
T

_
∂S
∂V
_
E
(3.8)
where V is the volume.
First law of thermodynamics:
dS =
∂S
∂E
dE +
∂S
∂V
dV (3.9)
dE = k
B
T dS −P dV (3.10)
Free energy:
F = E −k
B
TS (3.11)
Differential relation:
dF = −k
B
S dT −P dV (3.12)
or,
S = −
1
k
B
_
∂F
∂T
_
V
(3.13)
Chapter 3: Review of Statistical Mechanics 20
P = −
_
∂F
∂V
_
T
(3.14)
while
E = F +k
B
TS
= F −T
_
∂F
∂T
_
V
= −T
2

∂T
F
T
(3.15)
In the canonical ensemble, we assume that our system is in contact with a heat
reservoir so that the temperature is constant. Then,
ρ = C e
−βH
(3.16)
It is useful to drop the normalization constant, C, and work with an unnormalized
density matrix so that we can define the partition function:
Z = Tr ¦ρ¦ (3.17)
or,
Z =

a
e
−βE
a
(3.18)
The average energy is:
E =
1
Z

a
E
a
e
−βE
a
= −

∂β
ln Z
= −k
B
T
2

∂T
ln Z (3.19)
Hence,
F = −k
B
T ln Z (3.20)
The chemical potential, µ, is defined by
µ =
∂F
∂N
(3.21)
Chapter 3: Review of Statistical Mechanics 21
where N is the particle number.
In the grand canonical ensemble, the system is in contact with a reservoir of heat
and particles. Thus, the temperature and chemical potential are held fixed and
ρ = C e
−β(H−µN)
(3.22)
We can again work with an unnormalized density matrix and construct the grand
canonical partition function:
Z =

N,a
e
−β(E
a
−µN)
(3.23)
The average number is:
N = −k
B
T

∂µ
ln Z (3.24)
while the average energy is:
E = −

∂β
ln Z +µk
B
T

∂µ
ln Z (3.25)
3.2 Bose-Einstein and Planck Distributions
3.2.1 Bose-Einstein Statistics
For a system of free bosons, the partition function
Z =

E
a
,N
e
−β(E
a
−µN)
(3.26)
can be rewritten in terms of the single-particle eigenstates and the single-particle
energies
i
:
E
a
= n
0

0
+n
1

1
+. . . (3.27)
Z =

{n
i
}
e
−β
(

i
n
i

i
−µ

i
n
i)
=

i
_

n
i
e
−β(n
i

i
−µn
i
)
_
Chapter 3: Review of Statistical Mechanics 22
=

i
1
1 −e
−β(
i
−µ)
(3.28)
¸n
i
) =
1
e
β(
i
−µ)
−1
(3.29)
The chemical potential is chosen so that
N =

i
¸n
i
)
=

i
1
e
β(
i
−µ)
−1
(3.30)
The energy is given by
E =

i
¸n
i
)
i
=

i

i
e
β(
i
−µ)
−1
(3.31)
N is increased by increasing µ (µ ≤ 0 always). Bose-Einstein condensation occurs
when
N >

i=0
¸n
i
) (3.32)
In such a case, ¸n
0
) must become large. This occurs when µ = 0.
3.2.2 The Planck Distribution
Suppose N is not fixed, but is arbitrary, e.g. the numbers of photons and neutrinos
are not fixed. Then there is no Lagrange multiplier µ and
¸n
i
) =
1
e
β
i
−1
(3.33)
Consider photons (two polarizations) in a cavity of side L with
k
= ¯ hω
k
= ¯ hck and
k =

L
(m
x
, m
y
, m
z
) (3.34)
E = 2

m
x
,m
y
,m
z
ω
m
x
,m
y
,m
z
_
n
m
x
,m
y
,m
z
_
(3.35)
Chapter 3: Review of Statistical Mechanics 23
We can take the thermodynamic limit, L → ∞, and convert the sum into an
integral. Since the allowed

k’s are

L
(m
x
, m
y
, m
z
), the

k-space volume per allowed

k
is (2π)
3
/L
3
. Hence, we can take the infinite-volume limit by making the replacement:

k
f(

k) =
1
(∆

k)
3

k
f(

k) (∆

k)
3
=
L
3
(2π)
3
_
d
3

k f(

k) (3.36)
Hence,
E = 2 V
_
k
max
0
d
3
k
(2π)
3
¯ hω
k
e
β¯hω
k
−1
= 2 V
_
k
max
0
d
3
k
(2π)
3
¯ hck
e
β¯hck
−1
=
V k
4
B
π
2
(¯ hc)
3
T
4
_
β¯hck
max
0
x
3
dx
e
x
−1
(3.37)
For β¯ hck
max
¸1,
E =
V k
4
B
π
2
(¯ hc)
3
T
4
_

0
x
3
dx
e
x
−1
(3.38)
and
C
V
=
4V k
4
B
π
2
(¯ hc)
3
T
3
_

0
x
3
dx
e
x
−1
(3.39)
For β¯ hck
max
¸1 ,
E =
V k
3
max

2
k
B
T (3.40)
and
C
V
=
V k
3
max
k
B

2
(3.41)
3.3 Fermi-Dirac Distribution
For a system of free fermions, the partition function
Z =

E
a
,N
e
−β(E
a
−µN)
(3.42)
Chapter 3: Review of Statistical Mechanics 24
can again be rewritten in terms of the single-particle eigenstates and the single-particle
energies
i
:
E
a
= n
0

0
+n
1

1
+. . . (3.43)
but now
n
i
= 0, 1 (3.44)
so that
Z =

{n
i
}
e
−β
(

i
n
i

i
−µ

i
n
i)
=

i
_
_
1

n
i
=0
e
−β(n
i

i
−µn
i
)
_
_
=

i
_
1 +e
−β(
i
−µ)
_
(3.45)
¸n
i
) =
1
e
β(
i
+µ)
+ 1
(3.46)
The chemical potential is chosen so that
N =

i
1
e
β(
i
+µ)
+ 1
(3.47)
The energy is given by
E =

i

i
e
β(
i
+µ)
+ 1
(3.48)
3.4 Thermodynamics of the Free Fermion Gas
Free electron gas in a box of side L:

k
=
¯ h
2
k
2
2m
(3.49)
with
k =

L
(m
x
, m
y
, m
z
) (3.50)
Chapter 3: Review of Statistical Mechanics 25
Then, taking into account the 2 spin states,
E = 2

m
x
,m
y
,m
z

m
x
,m
y
,m
z
_
n
m
x
,m
y
,m
z
_
= 2 V
_
k
max
0
d
3
k
(2π)
3
¯h
2
k
2
2m
e
β
_
¯ h
2
k
2
2m
−µ
_
+ 1
(3.51)
N = 2 V
_
k
max
0
d
3
k
(2π)
3
1
e
β
_
¯ h
2
k
2
2m
−µ
_
+ 1
(3.52)
At T = 0,
1
e
β
_
¯ h
2
k
2
2m
−µ
_
+ 1
= θ
_
µ −
¯ h
2
k
2
2m
_
(3.53)
All states with energies less than µ are filled; all states with higher energies are empty.
We write
k
F
=

2mµ
T=0
¯ h
,
F
= µ
T=0
(3.54)
N
V
= 2
_
k
F
0
d
3
k
(2π)
3
=
k
3
F

2
(3.55)
E
V
= 2
_
k
F
0
d
3
k
(2π)
3
¯ h
2
k
2
2m
=
1
π
2
¯ h
2
k
5
F
10m
=
3
5
N
V

F
(3.56)
2
_
d
3
k
(2π)
3
=
m
3
2
2
1
2
π
2
¯ h
3
_
d
1
2
(3.57)
For k
B
T ¸e
F
,
N
V
=
m
3
2
2
1
2
π
2
¯ h
3
_

0
d
1
2
1
e
β(−µ)
+ 1
=
m
3
2
2
1
2
π
2
¯ h
3
_
µ
0
d
1
2
+
m
3
2
2
1
2
π
2
¯ h
3
_
µ
0
d
1
2
_
1
e
β(−µ)
+ 1
− 1
_
+
m
3
2
2
1
2
π
2
¯ h
3
_

µ
d
1
2
1
e
β(−µ)
+ 1
Chapter 3: Review of Statistical Mechanics 26
=
(2m)
3
2

2
¯ h
3
µ
3
2

m
3
2
2
1
2
π
2
¯ h
3
_
µ
0
d
1
2
1
e
−β(−µ)
+ 1
+
m
3
2
2
1
2
π
2
¯ h
3
_

µ
d
1
2
1
e
β(−µ)
+ 1
=
(2m)
3
2

2
¯ h
3
µ
3
2
+
m
3
2
2
1
2
π
2
¯ h
3
_

0
k
B
T dx
e
x
+ 1
_
(µ +k
B
Tx)
1
2
−(µ −k
B
Tx)
1
2
_
+O
_
e
−βµ
_
=
(2m)
3
2

2
¯ h
3
µ
3
2
+
(2m)
3
2
π
2
¯ h
3

n=1
(k
B
T)
2n
_
_
µ
3
2
−2n
1
(2n −1)!
Γ
_
3
2
_
Γ
_
5
2
−2n
_
_

0
dx
x
2n−1
e
x
+ 1
_
_
=
(2m)
3
2

2
¯ h
3
µ
3
2
_
_
1 +
3
2
_
k
B
T
µ
_
2
I
1
+O
_
T
4
_
_
_
(3.58)
with
I
k
=
_

0
dx
x
k
e
x
+ 1
(3.59)
We will only need
I
1
=
π
2
12
(3.60)
Hence,
(
F
)
3
2
= µ
3
2
_
_
1 +
3
2
_
k
B
T
µ
_
2
I
1
+O
_
T
4
_
_
_
(3.61)
To lowest order in T, this gives:
µ =
F
_
_
1 −
_
k
B
T

F
_
2
I
1
+O(T
4
)
_
_
=
F
_
_
1 −
π
2
12
_
k
B
T

F
_
2
+O(T
4
)
_
_
(3.62)
E
V
=
m
3
2
2
1
2
π
2
¯ h
3
_

0
d
3
2
1
e
β(−µ)
+ 1
=
m
3
2
2
1
2
π
2
¯ h
3
_
µ
0
d
3
2
+
m
3
2
2
1
2
π
2
¯ h
3
_
µ
0
d
3
2
_
1
e
β(−µ)
+ 1
− 1
_
+
m
3
2
2
1
2
π
2
¯ h
3
_

µ
d
3
2
1
e
β(−µ)
+ 1
=
(2m)
3
2

2
¯ h
3
µ
5
2

m
3
2
2
1
2
π
2
¯ h
3
_
µ
0
d
3
2
1
e
−β(−µ)
+ 1
+
m
3
2
2
1
2
π
2
¯ h
3
_

µ
d
3
2
1
e
β(−µ)
+ 1
=
(2m)
3
2

2
¯ h
3
µ
5
2
+
m
3
2
2
1
2
π
2
¯ h
3
_

0
k
B
T dx
e
x
+ 1
_
(µ +k
B
Tx)
3
2
−(µ −k
B
Tx)
3
2
_
+O
_
e
−βµ
_
=
(2m)
3
2

2
¯ h
3
µ
5
2
+
(2m)
3
2
π
2
¯ h
3

n=1
(k
B
T)
2n
_
_
µ
5
2
−2n
1
(2n −1)!
Γ
_
5
2
_
Γ
_
7
2
−2n
_
_

0
dx
x
2n−1
e
x
+ 1
_
_
Chapter 3: Review of Statistical Mechanics 27
=
(2m)
3
2

2
¯ h
3
µ
5
2
_
_
1 +
15
2
_
k
B
T
µ
_
2
I
1
+O(T
4
)
_
_
=
3
5
N
V

F
_
_
1 +

2
12
_
k
B
T

F
_
2
+O(T
4
)
_
_
(3.63)
Hence, the specific heat of a gas of free fermions is:
C
V
=
π
2
2
Nk
B
k
B
T

F
(3.64)
Note that this can be written in the more general form:
C
V
= (const.) k
B
g (
F
) k
B
T (3.65)
The number of electrons which are thermally excited above the ground state is ∼
g (
F
) k
B
T; each such electron contributes energy ∼ k
B
T and, hence, gives a specific
heat contribution of k
B
. Electrons give such a contribution to the specific heat of a
metal.
3.5 Ising Model, Mean Field Theory, Phases
Consider a model of spins on a lattice in a magnetic field:
H = −gµ
B
B

i
S
z
i
≡ 2h

i
S
z
i
(3.66)
with S
z
i
= ±1/2. The partition function for such a system is:
Z =
_
2 cosh
h
k
B
T
_
N
(3.67)
The average magnetization is:
S
z
i
=
1
2
tanh
h
k
B
T
(3.68)
The susceptibility, χ, is defined by
χ =
_

∂h

i
S
z
i
_
h=0
(3.69)
Chapter 3: Review of Statistical Mechanics 28
For free spins on a lattice,
χ =
1
2
N
1
k
B
T
(3.70)
A susceptibility which is inversely proportional to temperature is called a Curie suc-
septibility. In problem set 3, you will show that the susceptibility is much smaller for
a system of electrons.
Now consider a model of spins on a lattice such that each spin interacts with its
neighbors according to:
H = −
1
2

i,j
JS
z
i
S
z
j
(3.71)
This Hamiltonian has a symmetry
S
z
i
→−S
z
i
(3.72)
For k
B
T ¸ J, the interaction between the spins will not be important and the
susceptibility will be of the Curie form. For k
B
T < J, however, the behavior will be
much different. We can understand this qualitatively using mean field theory.
Let us approximate the interaction of each spin with its neighbors by an interaction
with a mean-field, h:
H = −

i
hS
z
i
(3.73)
with h given by
h =

i
J¸S
z
i
) = Jz¸S
z
i
) (3.74)
where z is the coordination number. In this field, the partition function is just
2 cosh
h
k
B
T
and
¸S
z
) = tanh
h
k
B
T
(3.75)
Using the self-consistency condition, this is:
¸S
z
) = tanh
Jz¸S
z
)
k
B
T
(3.76)
Chapter 3: Review of Statistical Mechanics 29
For k
B
T < Jz, this has non-zero solutions, S
z
,= 0 which break the symmetry
S
z
i
→ −S
z
i
. In this phase, there is a spontaneous magnetization. For k
B
T > Jz,
there is only the solution S
z
= 0. In this phase the symmetry is unbroken and the
is no spontaneous magnetization. At k
B
T = Jz, there is a critical point at which a
phase transition occurs.