1.4 Basic methods and results of statistical mechanics and the ideal gas.

The disorder number,  is defined as the number of microstates available to a macrostate, and is very
large for the equilibrium macrostate. The disorder number is related to entropy, which is maximum at
equilibrium.
To enumerate  , we need to decide whether the particles are distinguishable or indistinguishable. If
they are indistinguishable, every microstate is a macrostate.
Let us now consider independent particles, with energy levels fixed and the particles don't influence
each other's occupation. The energy levels are fixed by each particle's interaction with its surroundings
but not particle-particle interactions. Mutual interactions exist only for energy exchange (during
collisions).
Each level j has energy  j , degeneracy g j , and occupation n j .
1.4.1 Distinguishability
Small and identical particles cannot be distinguished during collisions, i.e. non-localized. However,
localized particles are distinguishable because they are fixed in space, the sites being distinguishable
e.g. in a paramagnetic solid.
For an isolated system, the volume V, energy E and total number of particles N, are fixed.
Consider the energy of states of an assembly to be 1 ,2 ,...,k and in the statistical equilibrium, the
numbers of systems assigned to these energy states are n1 , n2 ,..., nk respectively. Obtain expressions for
the total number of the systems and total energy of the assembly.
Solution
Total number of systems
E  n1  n2  ...  nk

  nk
k 1
Total energy
E  n1 1  n2 2 ...  nk k

  nk k
k 1

Fig. 1, The disorder number is proportional to thermodynamic probability, i.e. how probable for a particular
macrostate to be the one corresponding to equilibrium.
for a particular macrostate to be the one corresponding to equilibrium.

Page 6 of 34
From the foregoing, the A Priori assumption states that all microstates of a given E are equally
probable thus the equilibrium macrostate must have overwhelming  .
1.4.2 Computation of Most Probable Macrostate Distribution, n j
To find the n j at equilibrium, i.e., we must find the most probable macrostate distribution.
Considering the case for distinguishable particles:
1. Enumerate  . We distribute N particles over the energy levels: For example,
For the first level,
N!
The number of ways to assign n1 particles 
n1 ! N  n1  !
For the second level,
N  n1 !
The number of ways to assign n2 particles 
n2 ! N  n1  n2  !
N! N  n1 !
The number of ways   
n1 ! N  n1 ! n2 ! N  n1  n2 !
N!

n1 !n2 ! n j !
N!

 j nj !
Level j has degeneracy g j so any of the n j particles can enter into any of the g j levels.
n
Number of ways = g j j
For all levels,
Number of ways =  g j j
n

j
nj
g
Total = N !
j

j nj !
2. Maximize  by keeping E, N fixed. These constraints are introduced through Lagrange
multipliers.
For a fixed V, it means that the energy levels,  j are unchanged.
Maximize  subject to
 nj  N
and
 n j j  E
At maximum,
d  0
1
 d  ln    d 

0
Hence we can choose to maximize ln  instead of 
ln   ln N !  n j ln g j  ln n j ! (1.10)
j j

By Stirling's approximation for large N:

ln N !  N ln N  N
Therefore

Page 7 of 34
 ln   N ln N  N    n j ln g j  n j ln n j  n j 
j

 N ln N    n j ln g j  n j ln n j  (1.11)
j

g 
 N ln N   n j ln  j 
n
j  j 
by assuming large n j , which is not always true.
 g 
d  ln      ln j dn j  dn j   0 (1.12)
 n
j 

j 
It is subjected to the constraints:
 dn
j
j  0  for fixed N 


j
j dn j  0  for fixed E 

Note  j does not vary. Therefore

 gj 
j  n
ln  dn j  0 (1.13)
 j 
Now introduce the constraints:
  dn j  0
j

   j dn j  0
j

so we get
  gj  
  ln  n       j  dn j  0

(1.14)
j
  j 
We choose  and  such that
g 
ln  j
n       j  0
 j 
The equilibrium macrostate becomes:
 j
n j  g j e e
Ng j   j
 e

 j
j
g je
Ng j   j
nj  e (Maxwell-Boltzmann Distribution) (1.15)
Z
1
Z is called the partition function,  is equal to where k is the Boltzmann constant and T is the
kT
temperature.
For j i , we get n j  ni for g j  gi and T < 0
For T = 0, n j  0 , except for n1 , i.e. all the particles are at ground state.
For T > 0, the upper levels get piled by particles provided   kT
Otherwise, the population of upper levels remains small.
Page 8 of 34
1.4.3 Meaning and significance of the partition function
It may not be obvious why the partition function, as we have defined it above, is an important quantity.
First, let us consider what goes into it. The partition function is a function of the temperature T and the
microstate energies 1 ,2 ,3 , etc. The microstate energies are determined by other thermodynamic
variables, such as the number of particles and the volume, as well as microscopic quantities like the
mass of the constituent particles. This dependence on microscopic variables is the central point of
statistical mechanics. With a model of the microscopic constituents of a system, one can calculate the
microstate energies, and thus the partition function, which will then allow us to calculate all the other
thermodynamic properties of the system.
The partition function can be related to thermodynamic properties because it has a very important
statistical meaning. The probability Ps that the system occupies microstate s is
1
Ps  e  s (1.16)
Z
e   s is known as Boltzmann factor.
The partition function thus plays the role of a normalizing constant (note that it does not depend on s),
ensuring that the probabilities sum up to one:
1 1
s P  Z s e s  Z Z  1 (1.17)

This is the reason for calling Z the "partition function": it encodes how the probabilities are partitioned
among the different microstates, based on their individual energies. The letter Z stands for the German
word Zustandssumme, "sum over states". This notation also implies another important meaning of the
partition function of a system: it counts the (weighted) number of states a system can occupy. Hence if
all states are equally probable (equal energies) the partition function is the total number of possible
states. Often this is the practical importance of Z.

Tut. Question
The energy S is defined as S  k log C and the most probable distribution is given by
1 V
ni  i exp     i  where   and exp    3 
2 mkT  2 . Derive an expression for the
3

kT Nh
ideal gas equation given as PV  NkT

Solution
   i   1
ni  i e where    and  
kT kT
i 
 ni

S  k log 
i 1 ni !

S  k  ni  log i  log ni  1
i 1

And substituting for ni we get


S  k  ni log i  log i     i  log e  1
i 1

Page 9 of 34
  
S  k   ni    ni i   ni 
 i i i 
 k  N   E  N 
S  kN 1     k  E
Substituting the values of  and  we get
  V 3 2   1  3 
S  kN 1  log  3  2 mkT    k   kT 
  Nh   kT  2 
 V 3 2 3
 kN  kN log  3  2 mkT    kN
 Nh  2
5  V 3 2
 kN  kN log  3  2 mkT  
2  Nh 
From the general equation of thermodynamics
P  S  E ,V  
 
T  V  E

  5  V  4 mE 3 2  
  kN  kN log   3  
V  2  N  3Nh   
 5  4 mE  
  kN  kN log V  kN log  5 3 2  
V  2  3N h  
kN
 0 0
V
P kN

T V
 PV  NkT
1.4.4 Counting in statistical mechanics
Example
Given 6 distinguishable particles, 2 energy levels (one with a degeneracy of 2 and the
other degeneracy of 5). Calculate the number of macrostates and microstates in this system.
Soln
A diagramatic scheme of the problem:

NB: there are no equilibrium macrostate in this problem because all the macrostates have different
total energies.
There are 7 macrostates namely : (6,0), (5,1), (4,2), (3,3), (2,4), (1,5) and (0,6).
calculate the microstates:
For (6,0) macrostate,
  26 ways
For (5,1) macrostate,

Page 10 of 34
 N ! n1 n2
 g1 g 2
n1 !n2 !
6!
  25  5 ways
5!1!
Subsequently, calculate for the remaining macrostates, using the example above.

Problem 2:
Given 5 energy levels, each separated by E . Let the particles be weakly interacting, i.e. they
exchange energy but do not affect each other's energy levels.
Let us look at the first 3 levels:
(i) a particle moves from the 3rd level down to the 2nd level
(ii) a particle moves from st level to the 2nd level.
In this case, the change in energy is balanced by exchange of particles for the 2 processes. Hence we
can talk about a fixed macrostate. Now we try to calculate .

Soln
Consider the schematic picture of this problem:

Macrostate (5,0,1,0,0), Ω1  6C5  1C0  1C1  0C0  0C0  6 ways

Macrostate (4,2,0,0,0), Ω2  6C4  2C2  0C0  0C0  0C0  15 ways
At equilibrium, we apply the weighted average for microstates, because the number of microstates in
either macrostate is not overwhelming enough.

nj 
 n j j n j 
j
Therefore,
5  6  4 15 90
n1  
6  15 21
0  6  2 15 30
n2  
6  15 21
1 6  0 15 6
n3  
6  15 21
n4  0
n5  0
Note:
 nj  6
j

which is the same as  n j for any macrostate.

Page 11 of 34
Tut. Question
A system consisting of 6 fermions has total energy 6 units. These particles are to be distributed in 5
energy levels ε0 = 0, ε1 = 1, ε2 = 2, ε3 = 3, ε4 = 4 units. Each energy level is triply degenerate. Find the
possible microstates.

Solution
The possible macrostates are 5. They are:{3,2,0,0,1}, {3,1,1,1,0}, {2,3,0,1,0}, {3,0,3,0,0}, {2,2,2,0,0}.
The thermodynamic probability (the number of microstates) of a macrostate is given
gi !
by   
i n ! g  n  !
i i i

The number of microstates associated with first macrostate is

gi  3 nk  {3, 2,0,0,1}
3! 3! 3! 3! 3!
I  . . . . 9
3! 3  3! 2! 3  2 ! 0! 3  0 ! 0! 3  0 ! 3! 3  1!

gi  3 nk  {3,1,1,1,0}
3! 3! 3! 3! 3!
 II  . . . .  27
3! 3  3! 1! 3  1! 1! 3  1! 1! 3  1! 0! 3  0 !

gi  3 nk  {2,3,0,1,0}
3! 3! 3! 3! 3!
 III  . . . . 9
2! 3  2 ! 3! 3  3! 0! 3  0 ! 1! 3  1! 0! 3  0 !

gi  3 nk  {3,0,3,0,0}
3! 3! 3! 3! 3!
 IV  . . . . 1
3! 3  3! 0! 3  0 ! 0! 3  0 ! 0! 3  0 ! 0! 3  0 !

gi  3 nk  {2, 2, 2,0,0}
3! 3! 3! 3! 3!
V  . . . .  27
2! 3  2 ! 2! 3  2 ! 2! 3  2 ! 0! 3  0 ! 0! 3  0 !

The thermodynamic probability of the system is Ω = Σ Ωi = 9 + 27 + 9 + 1 + 27 = 73

Tut. Ques
By Stirling's approximation for large N: ln N !  N ln N  N , Proof

Solution
N
ln N !   ln n
1

  ln ndn
N

 n ln n  n 1
N

 N ln N  N  1
For N 1, ln N !  N ln N  N

Page 12 of 34
1.4.5 The question of many energy levels
If n j 1for most levels, then we have many energy levels G N . This means that all macrostates
look roughly the same. The total number of microstates are not mainly due to the maximum term
  n j  , i.e.      n j  . Hence Stirling's approximation is invalid in this case.
Energy levels are closely spaced in groups because G is large. They are grouped into bunches of gi
with mi occupation. gi and mi can be made large in this case.

When  is counted:
i

Ngi e kT
mi   j

 i gi e kT
i

Ngi e kT
  j

e kT

take the time average value of nj over a large number of different macrostates.
m
mi  i
gi
 j

Ne kT
  j

 j
e kT
therefore
gimi
  mi   N !
i mi !
ln   n j   ln  
By grouping closely spaced levels, the time-averaged variations are small and the Maxwell-Boltzmann
distribution describes n j very well. Otherwise with n j 1 , the variation at each j level appears large.
The treatment tries to obtain the average ni over many different but similar macrostates.
Therefore all macrostates of significant probability (i.e. large  ) have roughly the same properties and
distribution close to the Maxwell- Boltzmann distribution ignoring minor changes.
The transition speed  1012 1035  1035 s-1

1.4.6 Time evolution of a system of microstates

How many microstates does the system explore within the time for the experiment?

Page 13 of 34
Consider paramagnetic system (where the atomic spins interact with the external magnetic field,
Zeeman Effect). The particles inside this system (= 1023particles) are assumed to be non-interacting,
and each electron have a spin of  12 . We also fix E and T for a fixed macrostate i.e. nj. Note in
Quantum Mechanics, the energy levels are discretized.
 k .Bk  E  f T  k-each atom
k
The change in microstates will mean that the spin energy of the electron is exchanged between
neighbouring atoms so the atoms interchange energy levels. The typical time for spin flip is 10-12s. So
for 1023 electrons:
The transition speed = 1012 1023  1035 s-1
Now, the number of microstates:
 e N ln N  e35  1035
So it takes forever to go through all the microstates. Hence, for every experiment, one can only
explore a small proportion of microstates. Thus we need a statistical assumption that what we observe
corresponds to the real average over all microstates.

1.4.7 Maxwell-Boltzmann distribution away from equilibrium

Suppose we change from 1 macrostate to another macrostate, keeping E, V and N fixed. We need a
transition from i to k to be accompanied by another transition from j to l.

Both the transitions from j to l or i to k requires the same amount of energy.

Therefore, the rate
Rij kl  Ni N jTijkl
where Tijkl is the transition probability.
The total rate out from level i:
Riout  Ni  N jTijkl
j , k ,l

The total rate in to level i:

Riin   N k NlTklij
j , k ,l

At equilibrium,
Riout  Riin 
Therefore it follows that:
Ni  N jTijkl   N k NlTklij
j , k ,l j , k ,l

Now we invoke the time-reversal invariance, which is true for almost all transitions,
Tijkl  Tklij
Thus for every combination i, j, k and l, we require, Ni N j  N k Nl (Principle of Detailed Balancing)

Page 14 of 34
Or ln Ni  ln N j  ln N k  ln Nl and i  j k  l is required for energy conservation.
Since Ni  N i  , both equations will lead to
ln Ni    i 
Ni  e e  
which is the Maxwell-Boltzmann Distribution.

1.4.8 Indistinguishable particles

Why are particles indistinguishable?
This is a quantum mechanical effect. The exchange of quantum particles cannot be detected. The 's
of the individual particles combine to form a total  , and they are non-localised.
We can write a many particle wavefunction  as a linear combination of  a 1 b  2  plus all the
possible permutations. For example, for three particle permutations, we get six possibilities:
 a 1 b  2  c  3
 a  2  b  3 c 1
 a 1 b  3 c  2 
etc
All particles are identical so all permutations are solutions to Schrodinger's equations, and so are all
linear combination of the permutations.

There are two physically meaningful solutions:

i. Symmetry  ,
 s   P  1, 2,..., N   (1.18)
p

e.g., for 3 particles, we get

 s   a 1 b  2  c  3  a  2  b 1 c  3  a  3 b  2  c 1
+ a 1 b  3 c  2   a  2  b  3 c 1  a  3 b 1 c  2 
ii. Antisymmetric  ,
 a    1 P  1, 2,..., N  
p
(1.19)
p

where P is the permutation operator; (-1)P means x(-1) every time we exchange two particle states.
E.g., for 3 particles, we get,
 s   a 1 b  2  c  3  a  2  b 1 c  3  a  3 b  2  c 1
- a 1 b  3 c  2   a  2  b  3 c 1  a  3 b 1 c  2 
If we exchange 1 and 2, then  a  2,1,3   a 1, 2,3
There are two kinds of quantum particles in nature:
1. Bosons: Symmetric wavefunctions with integral spins. Do not obey the Pauli‟s exclusion principle,
therefore as many particles as possible can occupy the same quantum state. They are particles
distributed according to Bose-Einsten statistics
2. Fermions: Anti-symmetric wavefunctions with half-integral spins. Obey Pauli‟s exclusion principle
according to which no two particles can occupy the same quantum state. They are particles distributed
according to Fermi-Dirac statistics

Page 15 of 34
Note when 2 fermions are in the same state, total  is zero. For example, a = b, meaning all
permutations cancel in pairs, and hence we obtain Pauli's Exclusion Principle.

1.4.9 Bose-Einstein distribution for Bosons

There is no limit to the occupation number at each level of quantum state, the more the better. Note
that  =1 for all macrostates because of the indistinguishability of particles,
(Macrostates  Microstates)

Unless g >>1 for some levels, e.g., by the clumping of levels together to make gi large. We can then
count microstates (  >> 1) because the states in i are distinguishable.
To put mi particles in gi states, imagine arranging mi particles into gi states, by arranging mi identical
particles and gi 1 identical barriers on a straight line.

The number of ways =

 mi  gi  1!
mi ! gi  1 !
Therefore,

 mi  gi  1!
mi ! gi  1 !
To maximise subject to the constraints:
m i
i N

m   E
i
i i

therefore
ln    ln  mi  gi  1! ln mi !  gi  1!
1

   mi  gi  1 ln  mi  gi  1   mi  gi  1  mi ln mi  mi   gi  1 ln  gi  1   gi  1 
i

   mi  gi  ln  mi  gi   mi ln mi  gi ln gi 
i

For gi , n i 1
d  ln     dmi ln  mi  gi   dmi  dmi ln mi  dmi  0
i

 mi  gi 
  ln  dmi  0
i  mi 
Add in:
  dmi  0
i

   i dmi  0
i
so
  mi  gi  
 ln       i  dmi  0 (1.20)
i   mi  

Page 16 of 34
then
 mi  gi    i
 e e
 mi 
therefore
gi
mi 
e e  i  1


Averaging over the gi levels in each clump,

1
ni  (1.21)
e e  i  1



which is the Bose-Einstein Distribution.   , where  is the chemical potential.
kT
Two systems at thermodynamic equilibrium have the same temperature T. The systems at chemical
(particle number) equilibrium have the same  .

1.4.10 Fermi-Dirac distribution of Fermions

In this case, we allow 0 or 1 particle to each quantum state. Either there is a particle or no particle in
each quantum state.
Now we clump the energy levels to produce gi states and mi (<<gi).

gi
Number of ways to fill mi boxes out of gi =
mi ! gi  mi  !

Therefore
gi

i mi ! gi  mi  !
To maximise subject to the constraints:
m i
i N

m   E
i
i i

Therefore,
ln    gi ln gi  mi ln mi   gi  mi  ln  gi  mi 
i

where gi mi 1
d ln    dmi ln mi  dmi  dmi ln  gi  mi   dmi
i

=  ln
 gi  mi  dm
i
i mi
=0
add in:
  dmi  0
i

   i dmi  0
i

Page 17 of 34
So
  gi  mi  
 ln       i  dmi  0 (1.22)
i   mi  
Then,
 gi  mi    i
 e e
 mi 
Therefore,
gi
mi 
e e  i  1


Averaging over the gi levels in each clump,

gi
ni  (1.23)
e e  i  1


which is the Fermi-Dirac Distribution.

Note:
i. The positive sign in Fermi-Dirac distribution implies that n j  1 which is Pauli Exclusion
principle (which states that no 2 fermions can share the same quantum state).
ii. If  n j  N is not a constraint,  is zero and this is true for open systems. For isolated systems
with fixed N,  has a value.

For a system of identical fermions, the average number of fermions in a single-particle state is given by
the Fermi–Dirac (F–D) distribution,
1
ni      / kT
e i
1

where k is Boltzmann's constant, T is the absolute temperature, i is the energy of the single-particle
state i , and  is the chemical potential. At T = 0 K, the chemical potential is equal to the Fermi
energy. For the case of electrons in a semiconductor,  is also called the Fermi level.
The F–D distribution is only valid if the number of fermions in the system is large enough so that
adding one more fermion to the system has negligible effect on  . Since the F–D distribution was
derived using the Pauli exclusion principle, which allows at most one electron to occupy each possible
state, a result is that 0  ni  1 .
Fermi–Dirac distribution

Page 18 of 34
Energy dependence
More gradual at higher T. = 0.5 when   . Not shown is that  decreases for higher T.

Temperature dependence for   .

Distribution of particles over energy

Fermi function F  vs. energy  , with  = 0.55 eV and for various temperatures in the range
50K ≤ T ≤ 375K.
The above Fermi–Dirac distribution gives the distribution of identical fermions over single-particle
energy states, where no more than one fermion can occupy a state. Using the F–D distribution, one can
find the distribution of identical fermions over energy, where more than one fermion can have the
same energy.
The average number of fermions with energy i can be found by multiplying the F–D distribution ni
by the degeneracy gi (i.e. the number of states with energy i ),
ni i   gi ni
gi
 i    / kT
e 1

Page 19 of 34
When gi  2 , it is possible that n i   1 since there is more than one state that can be occupied by
fermions with the same energy i .
When a quasi-continuum of energies  has an associated density of states g  (i.e. the number of
states per unit energy range per unit volume) the average number of fermions per unit energy range per
unit volume is,
N   g  F 
where F  is called the Fermi function and is the same function that is used for the F–D distribution
ni ,
1
F   i    / kT
e 1
so that,
g 
N  
e
i    / kT
1

1.4.11 Derivation of the B-E and F-D statistics away from equilibrium
For equilibrium consideration, Tij kl is the transition rate from i  k accompanied by j  l to
maintain energy conservation.
Suppose the rate depends on the occupation at k and l,
Ri ,out  Ni  N jTij kl 1  N k 1  N l 
j , k ,l

Ri ,in   N k NlTkl ij 1  Ni  1  N j 

j , k ,l

= 1  N i   N k N lTkl ij 1  N j 
j , k ,l

At equilibrium,
Ni  N jTij kl 1  N k 1  Nl   1  Ni   N k NlTkl ij 1  N j 
j , k ,l j , k ,l

Applying time-reversal invariance,

Tij kl  Tkl ij
Then for every combination i, j, k, l, we require that
N i N j 1  N k 1  N l   N k N l 1  N i  1  N j 
So
Ni Nj Nk Nl

1  Ni 1  N j 1  N k 1  Nl
Ni Nj Nk Nl
ln  ln  ln  ln
1  Ni 1 N j 1  Nk 1  Nl
With i  j k  l
Therefore
Ni
ln     i
1  Ni
then,

Page 20 of 34
 Ni
 e e  i
1  Ni
Ni Ni e e  i  e e   i
1
Ni 
e e  i 1


Two cases:
1. positive (+) sign, we obtain
1
Ni  (Fermi-Dirac distribution) (1.24)
e e  i  1


2. negative (-) sign, we obtain

1
Ni  (Bose-Einstein distribution) (1.25)
e e  i  1


For fermions, the rate Ri ,out 1  N k 1  Nl  which demonstrates Pauli's Exclusion Principle, and for
bosons, the rate Ri ,out 1  N k 1  Nl  , i.e. the more particles in k and l, the more likely the transition.
(The more you have, the more you receive).

NB: Both the Fermi-Dirac and Bose-Einstein become Maxwell-Boltzmann statistics at high
temperature and low concentration.

2.0 Ensembles in statistical mechanics

An ensemble is a collection of many configurations of a system, all describing the same
thermodynamic state. Ensembles are introduced in statistical mechanics since the particles constituting
a system may interact, but the systems constituting the ensemble are assumed to be non-interacting.

2.1 Types of ensembles

2.1.1 Microcanonical ensemble (isolated systems)
A microcanonical ensemble represents a collection of configurations of isolated systems that have
reached thermal equilibrium.
A system is isolated from its environment if it exchanges neither particles nor energy with its
surrounding. The volume, internal energy and number of particles of such a system are constant and
are the same for all configurations that are part of the same microcanonical ensemble.
A Microcanonical ensemble is a theoretical tool used to analyze an isolated thermodynamic system. In
such a system, the macrostate of the system has a fixed given energy and there is one and the same
probability for every accessible microstate.
The microcanonical ensemble is also called the NVE ensemble because it describes a system with a
fixed number of particles  N  , a fixed volume V  , and a fixed energy  E  .
2.1.2 The canonical Ensemble (system at fixed temperature)
A canonical ensemble is a collection of closed systems at constant temperature. Closed systems have
fixed number of particles and volume but can exchange energy with their surroundings. The energy of
a closed system is therefore not constant.
As an example, imagine a gas of particles (air molecules, e.g.) in a closed container at constant
temperature. Consider now an isolated system composed of a small closed (sub) system s with N
particles in a volume V and of a “heat bath” (b) at temperature T.
Page 21 of 34
Let the two subsystems be in contact so that they can exchange energy but no particles. In equilibrium,
the system s has equal temperature as the heat bath. Now, the energy of the small system, Es, is no
longer fixed and the system s will be in any energy state Es, in each one with a different probability ps .
The heat bath must be large ( Nb  N s ), so that the change in the energy of the bath, Eb, during energy
exchange between the small system and the bath, is negligible. Evidently, since the compound system
(= system + bath) is isolated, the total energy Et = Eb+Es is constant and the compound system belongs
to a microcanonical ensemble for which the number of configurations in the energy interval between Et
and Et  E is:
t  Et   s  Es  b  Et  Es 

2.1.3 Grand canonical ensemble (Open systems, nothing is fixed)

Consider now a more general case with the system of volume Vs being not only in contact with a heat
bath at temperature T, but also exchanging particles with the bath.

As an example we can imagine a room at constant temperature with an open window. In this case,
neither the energy nor the number of particles in the system (air molecules in the room, e.g.) are fixed.
Such a system is described by the grand canonical ensemble.
the total system being part of a microcanonical ensemble, therefore the number of configurations of the
total system is:
  Et , Nt   s  Ei , N s  b  Et  Es , Nt  N s 

2.2 Distribution laws

The statistical distribution laws can be used to describe may of the observed properties of different
kinds of physical systems having different properties. There are four kinds of systems which together
provide a description of the many diverse phenomena in physics:
i. identical but distinguishable particles
ii. identical indistinguishable particles
iii. identical, indistinguishable particles of half-integral spins
iv. identical, indistinguishable particles of integral spin
Assume that we have an assembly of N particles and with a total energy E. We assume that the
assembly is in statistical equilibrium at this total energy E. The problem now is to find out in how
many different ways (P), all of which is equally probable, these N particles can be arranged distributed
among the various stationary energy states of the assembly and to find out what is statistically the most
probable such distribution.
Solution

Page 22 of 34
 Energy state
ϵk nk
.
..
.
.

ϵ2 n2

ϵ1 n1
Total number of particles is the sum total of particles assigned to each energy state
N  n1  n2  ...  nk

  nk
k 1
While the total energy
E  n1 1  n2 2 ...  nk k

  nk k
k 1

a. For identical and distinguishable particles

-1 arrangement is N ways, 2nd is (N-1), 3rd is (N-2) etc to N! ways.
st

-Since the particles are identical, the sequence of arrangement is irrelevant, can start at ϵ1, ϵ2 …. ϵk.
-The total number of ways in which energy levels can be filled up with N systems is given by
N!
P
n1 !, n2 !,..., nk !
For the whole assembly the probability is given by the product over k of all such expressions as
given by
1
P  N !
k nk !

-Consider the degeneracy in a state k as k . It implies that

*there is no restriction as to how particles permute, the 1st does not affect the 2nd, 3rd , …nkth .
*1st can be assigned k , 2nd k , …. and last  nk  assigned in k ways, hence the total number of ways
in which nk systems can be assigned to k states of the kth level = k  k .
n

*Since this replicates in other states, the total number of ways in which N systems can be assigned to
stationary energy levels of the assembly is
N!
1  1 2  2 ....k  k
n n n

n1 !n2 !...nk !
This is the Maxwell-Boltzmann distribution count
k 
nk

CM-B  N !
k nk !

Page 23 of 34
 b. For identical indistinguishable particles
*Since the systems are indistinguishable, their permutation is 1(must be excluded)
*The rest of the permutations remains as above leading to the Classical distribution
N !  k 
nk

CClassical  
N ! k nk !
 k 
nk


nk ! k

c. For identical indistinguishable systems with integral spins

*One of the levels k is fixed (partitioned)
* The remaining k  1 are then permuted
*There are nk particles and k  1 partitions to be permuted. This can be permuted in  nk  k  1!
ways
* nk (particles) and number of partitions k  1 are alike, such permutation must be excluded, hence

P
 nk  k  1!
nk !k  1 !
For the whole assembly the probability is given by the product over k of all such expressions as given
by the Bose-Einstein distribution:
CB-E   k
 n  k  1!
k nk !k  1 !
d. For identical indistinguishable systems with half-integral spins
Constraint is made by the fact that such particles obey the Pauli exclusion principle, hence no more
than one particle can occupy an energy state.
*Ones the 1st energy has been filled, the 2nd will fill, followed by the next one in that order.
*Since there are k states in the kth level, the 1st is filled in k ways, 2nd in , next in k  2  etc. The
permutations in the k states is in k ! ways.
*For Pauli exclusion principle to be fully obeyed then k nk
*The unfilled states k  nk  are identical and must be excluded, similarly, the permutations of
nk identical particles must be excluded.
* For the whole assembly the probability is given by the product over k of all such expressions as
given by the Fermi-Dirac distribution:
k !
CF-D  
k nk ! k  nk  !

3.0 The Laws of Thermodynamics and Statistical Mechanics

The laws of thermodynamics, in principle, describe the specifics for the transport of heat and work in
thermodynamic processes.

3.1.1 Zeroth law

If two thermodynamic systems are in thermal equilibrium with a third, they are also in thermal
equilibrium with each other, similar situation seen in the microcanonical ensemble.

Page 24 of 34