KPN Stat Mech

K. P. N.
Murthy
Statistical Mechanics
January 29, 2014
Springer
Contents
Why should we study statistical mechanics ? . . . . . . . . . . . . . . 1

1.1 Micro-Macro connections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
b
1.2 S = kB ln (E,
V, N ) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
P
1.2.1 S = k
P B i pi ln(pi ) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.2.2 Q = Pi Ei dpi . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.2.3 W = i pi dEi . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.2.4 F = kB T ln Q(T, V, N ) . . . . . . . . . . . . . . . . . . . . . . . . . . 3
2
1.2.5 E
= kB T 2 CV . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
hN i2 kB T
2
1.2.6 N
=
kT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
V
1.2.7 Micro world : Determinism and time-reversal invariance 3
1.3 Macro world : Thermodynamics . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.4 Micro-Macro synthesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.5 What am I going to teach ? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.6 Syllabus prescribed . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
1.7 Books on statistical mechanics and thermodynamics . . . . . . . . . 6
1.8 Extra reading : Books . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
1.9 Extra reading : Papers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
Experiment, outcomes, events, probabilities and ensemble .

2.1 Toss a coin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.2 Roll a die . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.3 Pick randomly a molecule from this room . . . . . . . . . . . . . . . . . . .
2.4 Sample space and events . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.5 Probabilities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.6 Rules of probability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.7 Random variable . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.8 Maxwells mischief : ensemble . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.9 Calculation of probabilities from an ensemble . . . . . . . . . . . . . . .
2.10 Construction of ensemble from probabilities . . . . . . . . . . . . . . . . .
13
13
13
13
14
14
15
15
16
17
17
VI
Contents
2.11 Counting of the number of elements in events of the sample

space : Coin tossing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.12 Gibbs ensemble . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.13 Why should a Gibbs ensemble be of large size ? . . . . . . . . . . . . .
2.13.1 Stirling and his Approximation
to Large Factorials . . . .
2.13.2 N ! = N N exp(N ) 2N . . . . . . . . . . . . . . . . . . . . . . . . .
Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3
Binomial, Poisson, and Gaussian . . . . . . . . . . . . . . . . . . . . . . . . . . .

3.1 Binomial distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.1.1 Moment generating function . . . . . . . . . . . . . . . . . . . . . . . .
3.1.2 Poisson distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.1.3 Binomial Poisson `a la Feller . . . . . . . . . . . . . . . . . . . . . .
3.1.4 Poisson process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.1.5 Easy method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.1.6 Easier method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.1.7 Characteristic function . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.1.8 Cumulant generating function . . . . . . . . . . . . . . . . . . . . . . .
3.1.9 Sum of identically distributed independent random
variable . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.1.10 Poisson Gaussian . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Isolated system: Micro canonical ensemble . . . . . . . . . . . . . . . . .
4.1 Preliminaries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.2 Configurational entropy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.3 Ideal gas law : Derivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.4 Boltzmann entropy and Clausius entropy are the same . . . . . . .
4.5 Some issues on extensitivity of entropy . . . . . . . . . . . . . . . . . . . . .
4.6 Boltzmann counting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.7 Heaviside Theta function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.8 Dirac delta function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.9 Area of a circle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.10 Volume of an N -dimensional sphere . . . . . . . . . . . . . . . . . . . . . . . .
4.11 Classical counting of micro states . . . . . . . . . . . . . . . . . . . . . . . . . .
4.11.1 Counting of the volume . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.12 Density of states : g(E) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.12.1 A sphere lives on its outer shell : Power law can be
intriguing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.13 Entropy of an isolated system . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.14 Properties of an ideal gas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.15 Quantum counting of micro states . . . . . . . . . . . . . . . . . . . . . . . . .
4.15.1 Energy eigenvalues : Integer Number of Half Wave
lengths in L . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.16 Chemical Potential . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
17
19
19
20
21
22
25
25
27
29
30
31
32
33
33
34
34
35
37
41
41
42
43
44
45
45
46
47
47
49
51
51
51
52
52
53
55
56
58
Contents
VII
4.16.1 Toy model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58

4.16.2 Chemical potential of an ideal gas . . . . . . . . . . . . . . . . . . . 59
Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
5
Closed system : Canonical ensemble . . . . . . . . . . . . . . . . . . . . . . .

5.1 What is a closed system ? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
5.2 Toy model
a la H B Callen . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
5.3 Canonical partition function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
5.4 Canonical partition function :
Method of most probable distribution . . . . . . . . . . . . . . . . . . . . . .
5.5 Lagranges method of undetermined multipliers . . . . . . . . . . . . .
5.6 Generalisation to a function of N variables . . . . . . . . . . . . . . . . .
5.7 Derivation of Boltzmann weight . . . . . . . . . . . . . . . . . . . . . . . . . . .
5.8 Canonical partition function : Transform of density of states . .
5.9 Canonical partition function and Helmholtz free energy . . . . . .
5.10 Canonical ensemble and entropy . . . . . . . . . . . . . . . . . . . . . . . . . . .
5.11 Free energy to entropy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
5.12 Energy fluctuations and heat capacity . . . . . . . . . . . . . . . . . . . . . .
5.13 Canonical partition function for an ideal gas . . . . . . . . . . . . . . . .
5.13.1 Easy method: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
5.13.2 Easier method : Transform of density of (energy) states
5.14 Microscopic interpretation of heat and
Pwork . . . . . . . . . . . . . . . .
5.15 Work in statistical mechanics : W = i pi dEi . . . . . . . . . . . . . .
5.16 Heat in statistical mechanics : . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
63
63
63
64
66
70
72
74
75
76
77
78
80
81
81
82
84
84
85
85
Grand canonical ensemble . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91

6.1 Grand canonical partition function and grand potential . . . . . . 95
6.2 Euler formula in the context of homogeneous function . . . . . . . . 97
6.3 P V = kB T ln Q . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
6.4 Gibbs-Duhem relation : d = sdT + vdP . . . . . . . . . . . . . . . . . . 98
6.5 Grand canonical ensemble : Number fluctuations . . . . . . . . . . . . 99
6.6 Number fluctuations and isothermal compressibility . . . . . . . . . . 100
2
6.7 Alternate derivation of the relation : N
/hN i2 = kB T kT /V . . 102
Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
Quantum Statistics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109

7.1 Occupation number representation of a micro state . . . . . . . . . . 109
7.2 Open system and grand canonical partition function . . . . . . . . . 110
7.3 Fermi-Dirac Statistics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
7.4 Bose-Einstein Statistics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
7.5 Classical Distinguishable Particles . . . . . . . . . . . . . . . . . . . . . . . . . 113
7.6 Maxwell-Boltzmann Statistics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
7.6.1 QMB (T, V, N ) QMB (T, V, ) . . . . . . . . . . . . . . . . . . . . 114
7.6.2 QMB (T, V, ) QMB (T, V, N ) . . . . . . . . . . . . . . . . . . . . 115
VIII
Contents
7.7 Grand canonical partition function, grand potential,

and thermodynamic properties of an open system . . . . . . . . . . . . 115
7.8 Expressions for hN i . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
7.8.1 Maxwell-Boltzmann Statistics . . . . . . . . . . . . . . . . . . . . . . . 117
7.8.2 Bose-Einstein Statistics . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
7.8.3 Fermi-Dirac Statistics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
7.8.4 Study of a system with fixed N employing grand
canonical formalism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
7.9 All the three statistics are the same at high temperature
and/or low densities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
7.9.1 Easy Method : 3 0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
7.9.2 Easier Method : 0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
b 1 , n2 , ) = 1 . . . . . . . . . . . . . . . . . . . 122
7.9.3 Easiest Method (n
7.10 Statistics of Occupation Number - Mean . . . . . . . . . . . . . . . . . . . 123
7.10.1 Ideal Fermions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
7.10.2 Ideal Bosons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124
7.10.3 Classical Indistinguishable Ideal Particles . . . . . . . . . . . . . 124
7.11 Some Remarks on hnk i from the three statistics . . . . . . . . . . . . . 125
7.11.3 Maxwell-Boltzmann Statistics . . . . . . . . . . . . . . . . . . . . . . . 126
7.11.4 At high T and/or low all statistics give the same hnk i 126
7.12 Statistics of Occupation Number - Fluctuations . . . . . . . . . . . . . 127
7.12.3 Maxwell-Boltzmann statistics . . . . . . . . . . . . . . . . . . . . . . . 132
Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133
8
Bose Einstein Condensation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135

8.1 Some Preliminaries
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
P
hn
8.1.1 hN
i
=
k
R k i . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136
P
()
()d . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
8.1.2
k
8.1.3 g3/2 () versus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140
8.1.4 Graphical inversion to determine fugacity . . . . . . . . . . . . . 141
8.1.5 Treatment of the Singular Behaviour . . . . . . . . . . . . . . . . . 141
8.1.6 Bose-Einstein Condensation Temperature . . . . . . . . . . . . . 147
8.1.7 Grand Potential for Bosons . . . . . . . . . . . . . . . . . . . . . . . . . 147
8.1.8 Average Energy of Bosons . . . . . . . . . . . . . . . . . . . . . . . . . . 148
8.1.9 Specific Heat Capacity of Bosons . . . . . . . . . . . . . . . . . . . . 150
8.1.10 Mechanism of Bose-Einstein Condensation . . . . . . . . . . . 154
Statistical Mechanics of Harmonic Oscillators . . . . . . . . . . . . . . 157

9.1 Classical Harmonic Oscillators . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157
9.2 Helmholtz Free Energy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158
9.3 Thermodynamic Properties of the Oscillator System . . . . . . . . . 159
Contents
9.4
9.5
9.6
9.7
IX
Quantum Harmonic Oscillator . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160

Specific Heat of a Crystalline Solid . . . . . . . . . . . . . . . . . . . . . . . . 163
Einstein Theory of Specific Heat of Crystals . . . . . . . . . . . . . . . . 166
Debye theory of Specific Heat . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167
9.7.1 Bernoulli Numbers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169
List of Figures
3.1
3.2
4.1
4.2
N!
pn (1 p)N n with
n!(N n)!
N = 10; B(n) versus n; depicted as sticks; (Left) p = 0.5;
(Right) p = .35 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
n
exp() with mean ;
Poisson distribution : P (n) =
n!
P (n) versus n; depicted
as sticks;
Gaussian distribution :

(x )2
1
with mean and variance
G(x) = exp
2 2
2
2 = : continuous line.
(Left) = 1.5; (Right) = 9.5. For large Poisson and
Gaussian coincide . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
Binomial distribution : B(n) =
Two ways of keeping a particle in a box divided into two equal

parts. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
Four ways of keeping two distinguishable particles in a box
divided into two equal halves. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
7.1
Average occupation number of a quantum state under

Bose-Einstein, Fermi-Dirac, and Maxwell-Boltzmann statistics . 127
8.1
8.2
8.3
g3/2 () versus . Graphical inversion to determine fugacity . . . . 140

Singular part of hN i . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143
3 versus . The singular part [3 /V ][/(1 )] (the bottom
most curve), the regular part g3/2 () (the middle curve) , and
the total (3 ) are plotted. For this plot we have taken 3 /V
as 0.05 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144
Fugacity versus 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145
Ground state occupation as a function of temperature . . . . . . . . . 146
Heat capacity in the neighbourhood of Bose - Einstein
condensation temperature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
8.4
8.5
8.6
List of Tables
3.1
Probabilities calculated from Binomial distribution :

B(n; N = 10, p = .1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
4.1
4.2
Micro states of of two particles with total energy 2 . . . . . . . . . . . 58

Micro states of three particles with total energy 2 . . . . . . . . . . . . 59
5.1
5.2
5.3
Micro states of three dice with the constraint that they add to
six . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
A micro state with occupation number representation (2, 3, 4) . . 68
A few micro states with the same occupation number
representation of (2, 3, 4) There are 1260 micro states with the
same occupation number representation . . . . . . . . . . . . . . . . . . . . . 68
7.1
Terms in the restricted sum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
1
Why should we study statistical mechanics ?
A quick answer : because it is one of the core courses, like Classical Mechanics, Quantum Mechanics, Electrodynamics, and Mathematical Physics, in your
post graduate curriculum. Fortunately, you have only one course in statistical mechanics, unlike quantum mechanics, electrodynamics and mathematical
physics !
1.1 Micro-Macro connections

On a more serious note, statistical mechanics provides a theoretical bridge
that takes you from the micro world of Newton, Schrodinger, and Maxwell,
etc. to the macro world of thermodynamics. It attempts to derive the empirical
thermodynamics, especially the Second law - the law of ever-increasing entropy
- from the microscopic laws of classical and quantum mechanics.
When do we call something, a macroscopic object ?
What do we call something, a microscopic constituent of a macroscopic
object ?
The answer depends crucially on the object and the properties under study.
For example,
if we are interested in the properties like density, pressure, temperature
etc. of a cup of coffee, then the molecules of coffee are the microscopic
constituents; the cup of coffee is the macroscopic object.
in another context, an atom can be considered a macroscopic object; the
electrons, protons and neutrons form its microscopic constituents.
A polymer is a macroscopic object; the monomers are its microscopic
constituents.
A society is a macroscopic object; men, women, children and monkeys are
its microscopic constituents.
1 Why should we study statistical mechanics ?
b
1.2 S = kB ln (E,
V, N )
This is the first and the most important link, between the microscopic and the
macroscopic worlds; it was proposed by Boltzmann1 . S stands for entropy and
b is the number
belongs to the macro world described by thermodynamics.
of micro states of a macroscopic system2 . kB is the Boltzmann constant 3
that establishes correspondence of the statistical entropy of Boltzmann to the
thermodynamic entropy of Clausius.
1.2.1 S = kB
pi ln(pi )
I would call this Boltzmann-Gibbs-Shannon entropy. The sum is over all the
micro states of a macroscopic system; the micro states are labelled by i. The
probability of a micro state i is denoted by pi .
An interesting question : We resort to probabilistic description to hide
our ignorance or to reconcile with our inability to keep track of the innumerable micro states through which an equilibrium macroscopic system would go
through, dictated by Newtons equations of motion and initial conditions.
In thermodynamics, entropy is a property of a system. However in statistical mechanics entropy is defined in terms of the probabilities of the micro
states. Does it imply that entropy is not only determined by then system but
also also by the ignorance or the inability of the observer ? Looks paradoxical
?
1.2.2 Q =
Ei dpi
This equation provides a microscopic description of heat. The sum runs over
all the micro states of the macroscopic system. Ei is the energy of the system
when it is in micro state i. The probability that the system can be found in
micro state i given by pi . We need to impose an additional constraint that
P
i dpi is zero to ensure that the total probability is unity.
1.2.3 W =
pi dEi
This equation defines work in the vocabulary of the micro world; the sum runs
over all micro states of the system.
1
engraved on the tomb of Ludwig Eduard Boltzmann (1844-1906) in Zentralfriedhof, Vienna.

For example a set of six numbers, three positions and three momenta specify a
single particle. A set of 6N numbers specify a macroscopic system of N particles.
The string labels a micro state.
kB = 1.381 1023 Joules (Kelvin)1 . We have kB = R/A where R = 8.314
Joules (Kelvin)1 is the universal gas constant and A = 6.022 1023 (mole)1 is
the Avagadro number.
1.3 Macro world : Thermodynamics
1.2.4 F = kB T ln Q(T, V, N )
Helmholtz free energy F (T, V, N ), defined in thermodynamics, is related to
the canonical partition function Q(T, V, N ) of statistical mechanics. This is
another important micro-macro connection.
2
1.2.5 E
= kB T 2 CV
This relation connects the thermodynamic specific heat at constant volume

2
(CV ) to the fluctuations (E
) of statistical energy of a closed system.
2
1.2.6 N
=
hN i2 kB T
V
kT
This relation shows that fluctuations in the number of particles in an open

system is proportional to the measurable thermodynamic property kT , called
the isothermal compressibility defined as,

1 V
kT =
V P T
We shall see of several such micro-macro connections in the course of study
of Statistical Mechanics.
We can say the aim of statistical mechanics is to synthesise the macro
world from the micro world. This is not an easy task. Why ? Let us spend a
little bit of time on this question.
1.2.7 Micro world : Determinism and time-reversal invariance
The principal character of the micro world is determinism and time-reversal
invariance.
Determinism implies that the entire past and the entire future is frozen in
the present.
The solar system is a good example. If you know the positions and momenta of all the planets now, then the Newtonian machinery is adequate to
tell you where the planets shall be a thousand years from now and where they
were some ten thousand years ago.
Microscopic laws do not distinguish the future from the past. The equations are invariant under transformation of t t.
1.3 Macro world : Thermodynamics

On the other hand the macro world obeys the laws of thermodynamics4 .
4
The laws of thermodynamics can be stated, in a lighter vein as follows.
The zeroth law that tells of thermal equilibrium, provides a basis for the
thermodynamic property, we call temperature. It is the starting point for
the game of thermodynamics.
The first law that articulates in a smart way, the law of conservation
of energy; it provides a basis for the thermodynamic property called the
internal energy. You can put in energy into a system by heat or by work.
The second law that tells, come what may, the entropy increases; it
provides a basis for the thermodynamic property called entropy. An engine
can draw energy from the surroundings by work and deliver the same
amount or energy by heat. On the other hand if the machine draws energy
from the surroundings by heat, then the energy it can deliver by work is
invariably less. The second law is a statement of this basic assymetry.
The third law that tells that entropy vanishes at absolute zero. We can
say that the third law provides the basis for absolute zero temperature on
entropy scale. The third law is also about the unattainability of absolute
zero. You can go as close as you desire but you can never reach it.
Of these the second law is tricky. It breaks the time-symmetry present
in the microscopic descriptors. Macroscopic behaviour is not time-reversal
invariant. There is a definite direction of time - the direction of increasing
entropy.
1.4 Micro-Macro synthesis

How do we comprehend the time asymmetric macroscopic behaviour emerging
from the time symmetric microscopic laws ?
Let us make life simpler by attributing two aims to statistical mechanics.
The first is to provide a machinery for calculating the thermodynamic properties of a system on the basis of the properties of its microscopic constituents
e.g. atoms and molecules, and their mutual interactions. Statistical Mechanics
has been eminently successful in this enterprise. This is precisely why we are
studying this subject.
The second aim is to derive the Second law of thermodynamics. Statistical
Mechanics has not yet had any spectacular success on this count. However,
some recent developments in non linear dynamics and chaos, have shown there
is indeed an unpredictability in some (nonlinear) deterministic system; we
now know that determinism does not necessarily imply predictability. This
Zeroth Law : You can play.
First Law : You cant win.
Second Law : You cant even win even.
Third Law : You cant quit.
The above or some variations of the above are usually attributed to Ginsberg and
C P Snow.
1.5 What am I going to teach ?
statement, perhaps, provides the raison detre for the statistics in statistical
mechanics.
In these lectures I shall not address the second issue - concerning the emergence of time asymmetry observed in macroscopic phenomena. I shall leave
this question to the philosophers and/or better equipped theoretical physicists. Instead we shall concentrate on how to derive the macroscopic properties
from the properties of its microscopic constituents and their interactions.
1.5 What am I going to teach ?

I shall tell you of the elementary principles of statistical mechanics. I shall be
as pedagogical as possible. Stop me when you do not understand.
I shall cover topics in
micro canonical ensemble that provides a description of isolated system;
canonical ensemble, useful in the study of closed system;
grand canonical ensemble that describes open system.
I shall discuss equilibrium fluctuations of
energy in canonical ensemble and relate them to heat capacity;
number of molecules in open system and relate them to isothemal compressibility.
Within the frame work of grand canonical ensemble I shall discuss BoseEinstein, Fermi-Dirac and Maxwell Boltzmann statistics.
I shall deal with ideal gas, and classical and quantum harmonic oscillators.
I shall discuss Bose Einstein condensation in some details with an emphasis
on the mechanism.
While on quantum harmonic oscillators, I shall discuss statistical mechanics of
phonons emerging due to quantization of displacement field in a crystalline
lattice, and
photons arising due to quantization of electromagnetic field.
Maxwells demon is a mischievous and interesting entity. The idea is simple. For every thermodynamic property we have, in statistical mechanics, a
corresponding statistical quantity. We take the average of the statistical quantity over a suitable ensemble - micro canonical, for isolated system, canonical
for closed system and grand canonical for open systems - and relate it to its
thermodynamic counter part. Thus entropy is a statistical entity. It is a mere
number. It corresponds to the number of micro states accessible to a macroscopic system, consistent with the constraints. If entropy is statistical then
the second law which states that entropy invariable increases in a process,
should be a statistical statement. If so, there is a non-zero probability that
the second law would be violated. Maxwell constructed a demon that violates
the second law ... a demon that extracts work from an equilibrium system. If
time permits, I shall discuss Maxwells demon and its later incarnations.
Toward the end, and again if time permits, I shall discuss some recent
developments in thermodynamics - work fluctuation theorems and second law
violations.
1.6 Syllabus prescribed

Course No. : PY454
Course Title : Statistical Mechanics
Basic Statistical ideas :
Probability concepts, states of classical and quantum systems.
Isolated systems :
Micro canonical ensemble, statistical entropy, most probable state. Systems in thermal and diffusive contact. Conditions for equilibrium. Canonical and grand canonical ensemble, and partition functions.
Thermodynamics :
Extensive and intensive variables, laws of thermodynamics, various thermodynamic potentials and their connection to partition functions.
Ideal Fermi and Bose gases :
Distribution functions, classical limit. Electron gas in a metal. Black body
radiation. Debye theory. Bose Einstein Condensation.
Phase Transitions :
Elementary ideas about phase transitions of different kinds. Examples of
some phase transitions.
Recommended books :
1. Thermal Physics, C. Kittel
2. Statistical Physics, L. D. Landau and E. M. Lifshitz
3. Problems in Thermodynamics and Statistical Physics, P. T. Landsberg
(Ed.)
4. Introduction to Statistical Mechanics, F. Reif
I shall assume you are all comfortable with calculus; I shall also assume you
have a nodding acquaintance with thermodynamics. If you have difficulty at
any time about thermodynamics, stop me. I shall explain the relevant portions.
Let me end this section by giving a list of some books and articles on
thermodynamics and statistical mechanics.
1.7 Books on statistical mechanics and thermodynamics

R K Pathria, Statistical Mechanics, Second Edition, Butterworth-Heinemann(1996)
A popular book in statistical mechanics. Starts with a beautiful historical

account of the subject. Contains a very large collection of interesting and
non-trivial problems.
Donald A McQuarrie, Statistical Mechanics, Harper & Row (1976)
A beautiful book with emphasis on applications. Contains excellent problems at the end of each chapter; suitable for self-study.
R Balescu, Equilibrium and Non-Equilibrium Statistical Mechanics, Wiley (1975)
An insightful book with emphasis on concepts. The issues on irreversibility
are explained beautifully.
David Goodstein, States of Matter
Delightful and entertraining. You are reminded of Feynmans writing when
you read this book. The discussion on dimensional analysis is excellent.
This book is a must in your bookshelf.
Debashish Chowdhury and Dietrich Stauffer, Principles of Equilibrium Statistical Mechanics, Wiley-VCH (2000)
An easy to read and enjoyable book. Contains applications to several
fields - condensed matter physics, materials science, polymers, solid state
physics, and astrophysics.
F Rief, Fundamentals of statistical and thermal physics, McGraw-Hill
(1965)
One of the best text books on statistical thermodynamics. Rief develops
thermal physics entirely in the vocabulary of statistical mechanics. As a
result after reading this book, you will get an uneasy feeling that thermodynamics has been relegated to the status of an uninteresting appendix
to statistical mechanics. My recommendation : read this book for learning
statistical - thermodynamics; then read Callen or Van Ness (probably for
a second time) for thermodynamics. Then you will certainly fall in love
with both statistical mechanics and thermodynamics, separately!
Palash B Pal, An Introductory Course of Statistical Mechanics, Narosa
(2008)
A book with a broad perspective; emphasis on relativistic systems.
D Chandler, Introduction to Modern Statistical Mechanics, Oxford University Press, New York (1987)
A book that connects neatly the traditional to modern methods of teaching statistical mechanics; gives an excellent and simple introduction to
renormalization groups. A great book for the teachers also.
Claude Garrod, Statistical Mechanics and Thermodynamics, Oxford
University Press (1995)
A good book at introductory level; neatly organized; pedagogic; nice problems and exercises.
Kerson Huang, Statistical Mechanics, Second Edition, Wiley India
(2011)
A whole generation of physicists has learnt statistical mechanics from this
book. It is perhaps one of a very few books that take kinetic theory and
Boltzmann transport equation as a starting point. Historically, statistical

mechanics originated from Boltzmann transport equation. After all, it is
Boltzmann transport - with its collision term obeying stosszahlansatz (collision number assumption or also known as molecular chaos) that establishes thermal equilibrium - an important point that is either overlooked
or not adequately emphasized in most of the text books on statistical
mechanics.
The book contains three parts : the first on thermodynamics; the second on statistical mechanics; and the third on special topics in statistical
mechanics.
Do not learn thermodynamics from this book. You will lose interest. The
other two parts are excellent - especially the third on special topics.
I would recommend, retain your copy of the first edition of this book.
Huang has omitted in his second edition, several interesting discussions
present in the first edition.
Kerson Huang, Introduction to Statistical Physics, Second Edition, CRC
Press (2009).
I think Huang has made an hurried attempt to modernize his earlier
classic : Statistical Mechanics. I do not recommend this book to students
taking their first course in statistical mechanics. However a discerning
teacher will find this book very useful.
J W Gibbs, Elementary Principles in Statistical Mechanics, Schribner,
New York (1902)
A seminal work in statistical mechanics. It feels good to read statistical
mechanics in the words of one of its masters. The book looks at canonical
formalism in a purely logical fashion and justifies it because of its simplicity and proximity to thermodynamics ! I will not recommend this book for
a first reading. Learn statistical mechanics well and then read this book.
Avijit Lahiri, Statistical Mechanics : An Elementary Outline, Revised
Edition, Universities Press (2008)
A neat and well planned book. Focuses on bridging the micro world described by quantum mechanics to the macro world of thermodynamics.
Concepts like mixed states, reduced states etc. provide the basic elements
in the development of the subject. The book contains a good collection of
worked examples and problems.
Francis W Sears, and Gerhard L Salinger, Thermodynamics, Kinetic
Theory, and Statistical Mechanics, Narosa (1974)
Provides a balanced treatment of thermodynamics, kinetic theory and
statistical mechanics. Contains nice set of problems. Verbose in several
places that tests your patience.
Joon Chang Lee, Thermal physics - Entropy and Free Energies, World
Scientific (2002)
Joon Chang Lee presents statistical thermodynamics in an unorthodox
and distinctly original style. The presentation is so simple and so beautiful
that you do not notice that the book is written in awful English and at
several places, flawed.
James P Sethna, Entropy, Order Parameters, and Complexity, Clarendon Press, Oxford (2008).
James Sethna covers an astonishingly wide range of modern applications;
a book, useful not only to physicists, but also to biologists, engineers, and
sociologists. I find exercises and footnotes very interesting; often more
interesting than the main text! However thermodynamics gets bruised.
Is entropy a property of the system or a property of the (knowledge or
ignorance) of the fellow, observing the system ?
C Kittel, and H Kr
omer, Thermal physics, W H Freeman (1980)
A good book; somewhat terse. I liked the parts dealing with entropy,
temperature, chemical potential, and Boltzmann weight; contains a good
collection of examples.
Daniel V Schrhr
oder, An Introduction to Thermal Physics, Pearson
(2000).
Schroder has excellent writing skills. The book reads well. Contains plenty
of examples. Somewhat idiosyncratic.
M Glazer, and J Wark, Statistical Mechanics : A Survival Guide, Oxford
University Press (2010) This book gives a neat introduction to statistical mechanics; well organized; contains a good collection of worked-out
problems; a thin book and hence does not threaten you !
H C Van Ness, Understanding Thermodynamics, Dover (1969).
This is an awesome book; easy to read and very insightful. In particular,
I enjoyed reading the first chapter on the first law of thermodynamics,
the second on reversibility, and the fifth and sixth on the second law. My
only complaint is that Van Ness employs British Thermal Units. Another
minor point : Van Ness takes the work done by the system as positive and
that done on the system as negative. Engineers always do this. Physicists
and chemists employ the opposite convention. For them the sign coincides
with the sign of change of internal energy caused by the work process.
When the system does work, its internal energy decreases; hence the work
done by the system is negative. When work is done on the system its
internal energy increases; hence work done on the system is positive.
H B Callen, Thermodynamics, John Wiley (1960).
A standard textbook. This book has influenced generations of teachers
and students alike, all over the world. Callen is a house hold name in the
community of physicists. The book avoids all the pitfalls in the historical
development of thermodynamics by introducing a postulational formulation.
H B Callen, Thermodynamics and an Introduction to thermostatistics,
Second Edition, Wiley, India (2005).
Another classic from H B Callen. He has introduced statistical mechanics
without undermining the inner strength of thermodynamics. In fact, the
statistical mechanics he presents, enhances the beauty of thermodynamics.
10
The simple toy problem with a red die (the closed system) and two white
dice (the heat reservoir), and the restricting sum to a fixed number (conservation of total energy) explains beautifully the canonical formalism.
The pre-gas model introduced for explaining grand canonical ensemble of
Fermions and Bosons is simply superb. I also enjoyed the discussions on
the subtle mechanism underlying Bose condensation. I can go on listing
several such examples. The book is full of beautiful insights.
A relatively inexpensive, Wiley student edition of the book is available in
the Indian market. Buy your copy now !
Gabriel Weinreich, Fundamental Thermodynamics, Addison Wesley
(1968).
Weinreichs is original; he has a distinctive style. Perhaps you will feel
uneasy when you read his book for the first time. But very soon, you will
get used to Weireichs idiosyncracy; and you would love this book 5 .
C B P Finn, Thermal Physics, Nelson Thornes (2001).
Beautiful; concise; develops thermodynamics from first principles. Finn
brings out the elegance and power of thermodynamics.
Max Planck, Treatise on Thermodynamics, Third revised edition, Dover;
first published in the year 1897. Translated from the seventh German
edition (1922).
A carefully scripted master piece; emphasises chemical equilibrium. I do
not think any body can explain irreversibility as clearly as Planck does.
If you think third law of thermodynamics is irrelevant, then read the last
chapter. You may change your opinion.
E Fermi, Thermodynamics, Dover (1936)
A great book from a great master; concise; the first four chapters (on thermodynamic systems, first law, the Second law, and entropy) are superb.
I also enjoyed the parts covering Clapeyron and van der Waal equations.
J S Dugdale, Entropy and its physical meaning, Taylor and Francis
(1998).
An amazing book. Dugdale de-mystifies entropy. This book is not just
about entropy alone, as the name would suggest. It teaches you thermodynamics and statistical mechanics. A book that cleverly avoids unnecessary
rigour.
M W Zamansky, and R H Dittman, Heat and Thermodynamics, an
intermediate textbook, Sixth edition, McGraw-Hill (1981)
A good and dependable book for a first course in thermodynamics. I am
not very excited about the problems given in the book. Most of them are
routine and requires uninteresting algebraic manipulations.
R Shanthini, Thermodynamics for the Beginners, Science Education
Unit, University of Peredeniya (2009)
Thanks to H S Mani, now at Chennai Mathematical Institute, for bringing my

attention to this book and for presenting me with a copy.
1.9 Extra reading : Papers
11
Student-friendly. Shanthini has anticipated several questions that would

arise in the minds of the students when they learn thermodynamics for
the first time. The book has a good collection of worked out examples. A
bit heavy on heat engines.
Dilip Kondepudi and Ilya Prigogine, Modern Thermodynamics : From
heat engines to Dissipative Structures, John Wiley (1998)
Classical, statistical, and non equilibrium thermodynamics are woven into
a single fabric. Usually chemists present thermodynamics in a dry fashion.
This book is an exception; it tells us learning thermodynamics can be fun.
Contains lots of interesting tit-bits on history. Deserves a better cover
design; the present cover looks cheap.
1.8 Extra reading : Books

Nicolaus Sadi Carnot, Reflexions sur la puissance motrice du feu et sur
les machines propres
a developer cette puissance, Paris (1824); for English
translation see Sadi carnot, Reflections on the motive power of fire and
on machines fitted to develop that power, in J Kestin (Ed.) The second
law of thermodynamics, Dowden, Hutchinson and Ross, Stroudsburg, PA
(1976)p.16
J Kestin (Ed.), The second law of thermodynamics, Dowden, Hutchinson
and Ross (1976)
P Atkin, The Second Law, W H Freeman and Co. (1984)
G Venkataraman, A hot story, Universities Press (1992)
Michael Guillen, An unprofitable experience : Rudolf Clausius and the
second law of thermodynamics p.165, in Five Equations that Changed the
World, Hyperion (1995)
P Atkins, Four Laws that drive the Universe, Oxford university Press
(2007).
Christopher J T Lewis, Heat and Thermodynamics : A Historical Perspective, First Indian Edition, Pentagon Press (2009)
S G Brush, Kinetic theory Vol. 1 : The nature of gases and of heat,
Pergamon (1965) Vol. 2 : Irreversible Processes, Pergamon (1966)
S G Brush, The kind of motion we call heat, Book 1 : Physics and the
Atomists Book 2 : Statistical Physics and Irreversible Processes, North
Holland Pub. (1976)
I Prigogine, From Being to Becoming, Freeman, San Francisci (1980)
K P N Murthy, Excursions in thermodynamics and statistical mechanics, Universities Press (2009)
1.9 Extra reading : Papers

K K Darrow, The concept of entropy, American Journal of Physics 12,
183 (1944).
12
M C Mackay, The dynamical origin of increasing entropy, Rev. Mod.

Phys. 61, 981 (1989).
T Rothman, The evolution of entropy, pp.75-108, in Science
a la mode
: physical fashions and fictions Princeton University Press (1989)
Ralph Baierlein, Entropy and the second law : A pedagogical alternative,
American Journal of Physics 62, 15 (1994)
G. Cook, and R H Dickerson, Understanding the chemical potential,
American Journal of Physics 63, 738 (1995).
K. P. N. Murthy, Ludwig Boltzmann, Transport Equation and the Second Law, arXiv: cond-mat/0601566 (1996)
Daniel F Styer, Insight into entropy, American Journal of Physics 68,
1090 (2000)
B J Cherayil, Entropy and the direction of natural change, Resonance
6, 82 (2001)
J K Bhattacharya, Entropy
a la Boltzmann, Resonance 6, 19 (2001)
J Srinivasan, Sadi Carnot and the second law of thermodynamics, Resonance 6 42 (2001)
D C Shoepf, A statistical development of entropy for introductory physics
course, American Journal of Physics 70, 128 (2002).
K P N Murthy, Josiah Willard Gibbs and his Ensembles, Resonance
12, 12 (2007).
2
Experiment, outcomes, events, probabilities
and ensemble
2.1 Toss a coin

You get either Heads or Tails. The experiment has two outcomes.
Consider tossing of two coins. Or equivalently consider tossing a coin twice.
There are four outcomes : { HH, HT, T H, T T }. An outcome is an ordered
pair. Each entry in the pair is drawn from the set {H, T }.
We can consider, in general, tossing of N coins. There are 2N outcomes.
Each outcome is an ordered string of size N with entries from the set (H, T ).
2.2 Roll a die

You get one of the six outcomes :
Consider throwing of N dice. There are then 6N outcomes. Each outcome

is an ordered string of N entries drawn from the basic set of six elements given
above.
2.3 Pick randomly a molecule from this room

In classical mechanics, a molecule is completely specified by giving its three
position and three momentum coordinates. An ordered set of six numbers
{q1 , q2 , q3 , p1 , p2 , p3 }
is an outcome of the experiment. A point in the six dimensional phase space
represents an outcome.
We impose certain constraints e.g. the molecule is always confined to this
room. Then all possible strings of six numbers, consistent with the constrains,
are the outcomes of the experiment.
14
2 Experiment, outcomes, events, probabilities and ensemble
2.4 Sample space and events

The set of all outcomes of an experiment is called the sample space. Let us
denote it by the symbol .
= {H, T } for the toss of a single coin.
= {HH, HT, T H, T T } for the toss of two coins.
A subset of is called an event. Let A denote an event. A is a subset of
underlying an experiment. When we carry out the experiment and if the
outcome belongs to A, then we say the event A has occurred.
Consider tossing of two coins. Let an event A be described by the statement
that the first toss is H. Then A consists of the following elements: {HH, HT }.
The event corresponding to the roll of an even number in the game of dice,
is the subset
2.5 Probabilities
Probability is defined for an event.
What is the probability of H in a toss of a coin ?
One-half. This would be your immediate response. The logic is simple.
There are two outcomes : Heads and Tails. We have no reason to believe
why should the coin prefer Heads over Tails or vice versa. Hence we say
both outcomes are equally probable.
What is the probability of having at least one H in a toss of two coins ?
The event corresponding this statement is {HH, HT, T H} and contains three
elements. The sample size contains four elements. The required probability is
thus 3/4. All the four outcomes are equally probable 1 .
Thus, if all the outcomes are equally probable, then the probability of an
event is the number of elements in that event divided by the total number of
elements in the sample space. For e.g., the event A of rolling an even number
in a game of dice, P (A) = 3/6 = 0.5.
The outcome can be a continuum. For example, the angle of scattering of
a neutron is a real number between zero and . We then define an interval
(1 , 2 ) where 0 i : i = 1, 2 as an event. A measurable subset of a
sample space is an event.
Physicists have a name for this. They call it the axiom (or hypothesis or assumption) of Ergodicity. Strictly ergodicity is not an assumption; it signifies absence
of an assumption.
2.7 Random variable
15
2.6 Rules of probability

The probability p that you assign to an event, should be obey the following
rules.
p0
p(A B) = p(A) + p(B) p(A B)
p() = 0
p() = 1
(2.1)
In the above denotes a null event and is a sure event.

How does one assign probability to an event ?
Actually, this question does not bother the mathematicians. It is the physicists who should worry about this2 .
2.7 Random variable

The next important concept in probability theory is random variable, x =
X() where denotes an outcome and x a real number. Random variable
is a way of stamping an outcome with a number : Real number, for a real
random variable; integer, for an integer random variable; complex number,
for a complex random variable3 . Thus the random variable x = X() is a set
function.
Consider a continuous random variable x = X(). We define a probability
density function f (x) by
f (x)dx = P (|x X() x + dx)
(2.2)
In other words f (x)dx is the probability of the event (measurable subset) that
contains all the outcomes to which we have attached a real number between
x and x + dx.
2
Maxwell and Boltzmann attached probabilities to events in some way; we got

Maxwell-Boltzmann statistics.
Fermi and Dirac had their way of assigning probabilities to Fermions e.g. electrons, occupying quantum states. We got Fermi-Dirac statistics.
Bose and Einstein came up with their scheme of assigning probabilities to
Bosons populating quantum states; we got Bose-Einstein statistics.
In fact, we stamped dots on the faces of die; this is equivalent to implementing
the idea of a random variable : attach a number between one and six to each
outcome.
For a coin, we stamped Heads on one side and Tails on the other. This is
in the spirit of defining a random variable, except that we have stamped figures;
for the random variable, however, we should stamp numbers.
16
Now consider a continuous random variable defined between a to b with

a < b. We define a quantity called the average of the random variable x
as
Z
b
dx x f (x).
is also called the mean, expectation, first moment etc.

Consider a discrete random variable n, taking values from say 0 to N .
Let P (n) define the discrete probability. We define the average of the random
variable as
N
X
=
n P (n).
n=0
But then, we are accustomed to calculating the average in a different

way. For example I am interested in knowing the average marks obtained
by the students in a class, in the last mid-semester examination. How do I
calculate it ? I take the marks obtained by each of you, sum them up and
divide by the total number of students. That is it. I do not need notions
like probabilities, probability density, sum over the product of the random
variable and the corresponding probability, integration of the product of the
continuous random variable and its probability density function etc.
Historically, before Boltzmann and Maxwell, physicists had no use for
probability theory in their work. Newtons equations are deterministic. There
is nothing chancy about a Newtonian trajectory. We do not need probabilistic
description in the study of electrodynamics described by Maxwell equations;
nor do we need probability to comprehend and work with Einsteins relativity
- special or general.
However mathematicians had developed the theory of probability as an
important and sophisticated branch of mathematics4 .
It was Ludwig Eduard Boltzmann who brought, for the first time, the
idea of probability into physical sciences; he was championing the cause of
kinetic theory of heat and of matter. Boltzmann transport equation is the
first ever equation written for describing the time evolution of a probability
distribution.
2.8 Maxwells mischief : ensemble

However, Maxwell, had a poor opinion about a physicists ability to comprehend mathematicians writings on probability theory, in general, and the
meaning of average as an integral over a probability density, in particular.
After all, if you ask a physicist to calculate the average age of a student
in the class, hell simply add the ages of all the students and divide by the
number of students.
4
perhaps for assisting the kings in their gambling.
2.11 Counting of the number of elements in events of the sample space : Coin tossing
To be consistent with this practice, Maxwell proposed the notion of an

ensemble of outcomes of an experiment (or an ensemble of realisations of a
random variable). Let us call it Maxwell ensemble 5 .
Consider a collection of a certain number of independent realisations of
the toss of a single coin. We call this collection a Maxwell ensemble if it it
obeys certain conditions, see below.
Let N denote the number of elements in the ensemble. Let nH denote
the number Heads and nT number of Tails. We have nH + nT = N .
If nH = N p, and hence nT = N q, then we say the collection of outcomes
constitutes a Maxwell ensemble.
Thus an ensemble holds information about the outcomes of the experiment
constituting the sample space; it also holds information about the probability
of each outcome. The elements of an ensemble are drawn from the sample
space; however each element occurs in an ensemble as often as to reflect correctly its probability.
2.9 Calculation of probabilities from an ensemble

Suppose we are given the following ensemble : {H, T, H, H, T }. By looking at
the ensemble, we can say the sample space contains two outcomes {H, T }. We
also find that the outcome H occurs thrice and T occurs twice. We conclude
that the probability of H is 3/5 and that of T is 2/5.
2.10 Construction of ensemble from probabilities

We can also do the reverse. Given the outcomes and their probabilities, we
can construct an ensemble. Let ni denote the number of times an outcome
i occurs in an ensemble. Let N denote the total number of elements of the
ensemble. Choose ni such that ni /N equals pi ; note that we have assumed
that pi is already known.
2.11 Counting of the number of elements in events of

the sample space : Coin tossing
Consider tossing of N identical coins or tossing of a single coin N times. Let
us say the coin is fair. In other words P (H) = P (T ) = 1/2.
Let (N ) denote the set of all possible outcomes of the experiment. An
outcome is thus a string N entries, each entry being H or T. The number
5
Later we shall generalise the notion of Maxwell ensemble and talk of ensemble
as a collection identical copies of a macroscopic system. We shall call it a Gibbs
ensemble.
17
18
b ). We have (N
b )=
of elements of this set (N ) is denoted by the symbol, (N
N
2 .
Let (n; N ) denote a subset of (N ), containing only those outcomes
with n Heads (and hence (N n) Tails). How many outcomes are there in
the set (n; N ) ?
b N ) denote the number of elements in the event (n; N ). I shall
Let (n;
tell you how to count the number of elements of this set6 .
Take one outcome belonging to (n; N ). There will be n Heads in that
outcome. Imagine for a moment that all these Heads are distinguishable. If
you like, you can label them as H1 , H2 , , Hn . Permute all the Heads
and produce n! new configurations. From each of these new configurations,
produce (N n)! configurations by carrying out the permutations of the (N
n) Tails. Thus from one outcome belonging to the set (n; N ), we have
produced n! (N n) new configurations. Repeat the above for each element
b N ) n! (N n)! configurations. A
of the set (n; N ), and produce (n;
moment of thought will tell you that this number should be the same as N !7 .
We thus have,
It follows, then,
b N ) n! (N n)! = N ! .
(n;
b N) =
(n;
N!
n!(N n)!
(2.3)
(2.4)
These are called the binomial coefficients.

Show that
N
X
n=0
b N ) = (N
b ) = 2N .
(n;
(2.5)
The binomial coefficients add to 2N :

N
X
n=1
N!
= 2N
n!(N n)!
(2.6)
I remember I have seen this method described in a book on Quantum Mechanics

by Gassiarowicz. Check it out
if you do not get the hang of it, then work it out explicitly for four coins with
two Heads and two Tails. Show explicitly that
b = 2, N = 4) 2! 2! = 4!
(n
2.13 Why should a Gibbs ensemble be of large size ?
19
2.12 Gibbs ensemble

Following Gibbs, we can think of an ensemble as consisting of large number
of identical mental copies of a macroscopic system. 8 . All the members of an
ensemble are in the same macro state9 . However they can be in different micro
states. Let the micro states of the system under consideration, be indexed
by {i = 1, 2, }. The number of elements of the ensemble in micro state j
divided by the size of the ensemble is equal to the the probability of the system
to be in micro state j. It is intuitively clear that the size of the ensemble should
be large ( ) so that it can capture exactly the probabilities of different
micro states of the system10 . Let me elaborate on this issue, see below.

b N ) is maximum11 ?
What is the value of n for which (n;
b N ) is maximum.
It is readily shown that for n = N/2 the value of (n;
b (N ). We have,
b (N ) = (n
b =
Let us denote this number by the symbol
N/2, N ).
Thus we have
b )=
(N
8
10
11
N
X
N!
= 2N
n!
(N
n)!
n=0
(2.7)
For example the given coin is a system. Let p denote the probability of Heads
and q = 1 p the probability of tails. The coin can be in a micro state Heads
or in a micro state Tails.
This means the values of p and q are the same for all the coins belonging to the
ensemble.
If you want to estimate the probability of Heads in the toss of a single coin
experimentally then you have to toss a large number of identical coins. Larger
the size of the ensemble more (statistically) accurate is your estimate .
you can find this in several ways. Just guess it. I am sure you would have guessed
the answer as N/2. We know that the binomial coefficient is largest when n = N/2
if N is even, or when n equals the two integers closest to N/2 for N odd. That is
it.
b
If you are more sophisticated, take the derivative of (n;
N ) with respect to n
and set it zero; solve the resulting equation to get the value of N for which the
function is an extremum.
b
You may find it useful to take the derivative of logarithm of (n;
N ); employ
Stirling approximation for the factorials : ln(m!) = m ln(m) m for large m.
Stirling approximation to large factorials is described in the next section.
You can also employ any other pet method of yours to show that for n = N/2
b
the function (n;
N ) is maximum.
Take the second derivative and show that the extremum is a maximum.
20
b (N ) = (n
b = N/2; N ) =
N!
(N/2)! (N/2)!
(2.8)
b for large values of N . We employ Stirling approximaLet us evaluate

12
tion :
(2.9)
N ! = N N exp(N ) 2N
We have
N N exp(N ) 2N
b (N ) = (n
b = N/2; N ) = h
i2
p
(N/2)(N/2) exp(N/2) 2(N/2)
2
=2
N
N
(2.10)
Let us evaluate the natural logarithm of both the quantities under discussion.
Let
SG
SB
b ) = N ln 2
ln (N
(2.11)
N ln 2 (1/2) ln N
(2.13)
b (N ) = N ln 2 (1/2) ln N + (1/2) ln(2/)

ln
(2.12)
Thus SB is less than SG by a negligibly small amount. SB and SG are both of

the order of N ; SG SB is of the order of logarithm of N , which is negligible
for large N . For example take a typical value for N = 1023 . We have SG =
0.69 1023 and SB = 0.69 1023 24.48.
Note that only when N is large, we have SB equals SG . It is precisely
because of this, we want the number of elements to be large, while constructing
a Gibbs ensemble. We should ensure that all the micro states of the system
are present in the ensemble in proportions, consistent with their probabilities.
For example I can simply toss N independent fair coins just once and if N
is large then I am assured that there shall be N/2 Heads and N/2
Tails, where is negligibly small : of the order of N .
2.13.1 Stirling and his Approximation to Large Factorials
N ! = N N exp(N )
Stirlings formula is an approximation to large factorials. James Stirling (16921770) was a Scottish mathematician. We have,
12
Stirling approximation is described in the next section
21
N ! = N (N 1) 3 2 1
ln N ! = ln 1 + ln 2 + ln 3 + + ln N
=
N
X
ln(k)
k=1
ln x dx
N
= (x ln x x)1
= N ln N N 1
N ln N N
(2.14)
Thus, for large N we have

n! N N exp(N ).
Stirling formula is explained well in the following references.
Daniel V. Schroeder, An Introduction to Thermal Physics, Addison Wesley (2000)
F. Rief, Fundamentals of Statistical and Thermal Physics, McGraw Hill
(1965)
2.13.2 N ! = N N exp(N )
To show this, proceed as follows.

(N + 1) = N ! =
2N
Z
dx xN exp(x),
dx exp [N ln(x) x] ,
dx exp [F (x)] ,
(2.15)
where, F (x) = N ln(x) x. Determine the value of x, say x = x , at which

dF
F (x) is extremum. Note that x is solution of the equation,
= 0. We see
dx
immediately that x = N . Thus, F (x) is extremum at x = N .
22
First check whether the extremum is maximum or minimum. I leave this

to you, as an exercise. The result : F (x) is maximum at x = x .
Carry out the Taylor expansion of F (x) around x and keep terms up to the
second derivative. Substitute it in the integral and carry out the integration
to get the Stirling approximation.
Problems
2.1. Consider a coin with probability of Heads given by 0.3. The experiment
consists of tossing the coin once. Write down a possible ensemble of realisations
of the experiment.
2.2. Consider a p-coin; i.e. a coin for which p is the probability of Heads.
Consider an experiment of tossing the p-coin independently twice. Write down
a possible ensemble of realisations, for the following cases.
(a) p = 1/2
(b) p = 1/4
2.3. Let x = X() denote a continuous random variable defined in the range
0 x 1 with a uniform probability density function. Find the mean and
variance of x.
2.4. Let x = X() denote a continuous random variable defined in the range
0 x +, with an exponential probability density : f (x) = exp(x). Let
Mn denote the n-th moment of the random variable x. It is defined as
Z
Mn =
xn exp(x)dx.
0
Show that M0 = 1; M1 = 1; and M2 = 2. Obtain an expression for Mn .

2.5. Betrands paradox
Select a chord randomly in a circle. What is the probability that the chord
length is greater than the side of an inscribed equilateral triangle?
2.6. Throw two fair independent dice. Let (n1 , n2 ) denote the result of the
experiment : n1 is the result of the first die and n2 that of the second die.
Define a random variable
n = maximum(n1 , n2 ).
What is the probability distribution of the random variable n ? Determine
the mean and variance of n.
PN
2.7. Calculate YE , the exact value of ln N ! =
k=1 ln k and the Stirling
approximation YS = N ln N N for N = 2, 3, 100 . Plot the results in the
same graph. Plot also the relative error = |YE YI |/YE as a function of N .
PN
23
2.8. Calculate YE , the exact value of ln N ! = k=1 ln k and the improved

Stirling approximation YSI = N ln N N + 21 ln(2N ) for N = 2, 3, 100
. Plot the results in the same graph. Plot also the relative error = |YE
YSI |/YE as a function of n.
3
Binomial, Poisson, and Gaussian
3.1 Binomial distribution

Consider a system consisting of one coin. It has two micro states : H and
T. The probability for the system to be in micro state H is p and that in
micro state T is q = 1 p.
Consider the case with p = 0.6 and hence q = 1 p = 0.4. A possible
ensemble of systems that contain exact information about the micro states of
the system and of their probabilities is
{T, H, H H, T, H, H T, H, T }
Notice that the ensemble contains ten systems (or mental copies of the system). Six systems are in micro state H and four are in micro state T.
As I said earlier, we usually take the size of the ensemble to be arbitrarily
large. Let N denote the size of the ensemble.
Imagine, we attempt to construct the ensemble by actually carrying out
the experiment of tossing identical coins or by tossing the same coin several
times independently. What is the probability that in the experiment there
shall be n Heads and hence (N-n) Tails ? Let us denote this by the symbol
B(n). It is readily seen
B(n) =
N!
pn q N n
n! (N n)!
(3.1)
Figure (??) depicts Binomial distribution for N = 10, p = 0.5 and 0.35. What
is average value of n ? The average is also called the mean, the first moment,
the expectation value etc.. Denote it denoted by the symbol M1 or hni. It is
given by
26
3 Binomial, Poisson, and Gaussian

0.25
0.3
0.2
B(n)
B(n)
0.25
0.15
0.1
0.2
0.15
0.1
0.05
0
0.05
0
10
10
N!
pn (1 p)Nn with N = 10;
n!(N n)!
B(n) versus n; depicted as sticks; (Left) p = 0.5; (Right) p = .35
Fig. 3.1. Binomial distribution : B(n) =
M1 = hni =
N
X
n B(n; N )
n=0
N
X
n=0
N
X
N!
pn q N n
n! (N n)!
Np
n=0
= Np
N
X
n=0
= Np
N
1
X
n=0
(N 1)!
pn1 q (N 1)(n1)
(n 1)! [(N 1) (n 1)]!
(N 1)!
pn q (N 1)n
n![(N 1) n]!
B(n; N 1)
= Np
(3.2)
Thus the first moment (or the average) of the random variable n is N p.
We can define higher moments. The k-th moment is defined as
Mk = hnk i =
N
X
n=0
nk B(n)
(3.3)
27
The next important property of the random variable is variance. It is defined

as,
n2
N
X
(n M1 )2 B(n)
N
X
n2 B(n) M12
n=0
n=0
= M2 M12
(3.4)
Let me now describe a smart way of generating the moments of a random

variable.
3.1.1 Moment generating function
Let B(n) denote the probability that n coins are in micro state Heads in
an ensemble of N coins. We have shown that
B(n) =
N!
pn q N n
n!(N n)!
(3.5)
The moment generating function is defined as
B(z)
=
N
X
z n B(n),
(3.6)
n=0
= 1) = 1. This guarantees that the probThe first thing we notice is that B(z
ability distribution B(n) is normalized. The moment generating function is
like a discrete transform of the probability distribution function. We transform
the variable n to z.
Let us now take the first derivative of the moment generating function
with respect to z. We have,
N
X
dB
(z) =
=B
n z n1 B(n)
dz
n=0
(z) =
zB
N
X
n z n B(n)
(3.7)
n=0
. Substitute in the above z = 1. We get,

(z = 1) = hni
B
(3.8)
28
evaluated at z = 1 generates the first moment.

Thus the first derivative of B
Now take the second derivative of B(z)

to get
N
X
d2 B
=
n(n 1)z n2 B(n)
dz 2
n=0
z2
N
X
d2 B
=
z n n(n 1) B(n)
dz 2
n=0
Substitute in the above z = 1 and get,

d2 B

= hn(n 1)i
dz 2 z=1
(3.9)
(3.10)
For the Binomial random variable, we can derive the moment generating
function :
B(z)
=
N
X
z n B(n)
n=0
N
X
N!
(zp)n q N n
n!
(N
n)!
n=0
= (q + zp)N
(3.11)
When N is large, it is clumsy to calculate quantities employing Binomial

distribution. Consider the following situation.
I have N molecules of air in this room of volume V . The molecules are
distributed uniformly in the room. In other words the number density, denoted
by is same at all points in the room. Consider now an imaginary small volume
v < V completely contained in the room. Consider an experiment of choosing
randomly an air molecule from this room. The probability that the molecule
shall be in the small volume is p = v/V ; the probability that it shall be out
side the small volume is q = 1 (v/V ). There are only two possibilities. We
can use Binomial distribution to calculate the probability for n molecules to
be present in v.
Consider first the problem with V = 10M 3 , v = 6M 3 and N = 10. The
value of p for the Binomial distribution is 0.6. The probability of finding n
molecules in v is then,
B(n; N = 10) =
10!
(0.1)n (0.9)10n
n!(10 n)!
(3.12)
B(n; 10)
B(n; 10)
0
1
2
3
4
5
0.0001
0.0016
0.0106
0.0425
0.1115
0.2007
6
7
8
9
10
0.2508
0.2150
0.1209
0.0403
0.0060
29
Table 3.1. Probabilities calculated from Binomial distribution : B(n; N = 10, p =

.1)
The table below gives the probabilities calculated from the Binomial distribution.
Consider the same problem with v = 103 M 3 and N = 105 . We have
p = 104 and N p = 10. Immediately we recognize that Binomial distribution
is not appropriate for this problem. Calculation of the probability of finding
n molecules in v involves evaluation of 100000!.
What is the right distribution for this problem and problems of this kind ?
To answer this question, consider what happens to the Binomial distribution
in the limit of N , p 0, and N p = , a constant1 . Note that
N p = N v/V = v = constant.
We shall show below that in this limit, Binomial goes over to Poisson
distribution.
3.1.2 Poisson distribution
We start with
B(z)
= (q + zp)N
(3.13)
We can write the above as
B(z)
= q N (1 + zp/q)N
= (1 p)N (1 + zp/q)N
1
Note that for a physicist, large is infinity and small is zero.
(3.14)
30

N
p0 N p
= B(z)
exp(N p) exp(zN p/q)
= exp() exp(z)
= P (z)
(3.15)
Thus in the limit N , p 0 and N p = , we find B(z)

P (z), given
by
P (z) = exp[(1 z)]
(3.16)
The coefficient of z n in the power series expansion of P (z) gives P (n),

P (n) =
n
exp()
n!
(3.17)
The above is called the Poisson distribution2 . Thus in the limit of N ,

p 0, N p = , the Binomial distribution goes over to Poisson distribution.
Figure (??) depicts Poisson distribution for = 1.5 and 9.5.
3.1.3 Binomial Poisson `
a la Feller
Following Feller3 , we have
B(n; N )
B(n 1; N )
p0
N ! pn q N n (n 1)! (N n + 1)!
n! (N n)!
N ! pn1 q N n+1
p (N n + 1)!
nq
N p p (n 1)
nq
N p=
(3.18)
Start with
We shall come across Poisson distribution in the context of Maxwell-Boltzmann

statistics. Let nk denote the number of indistinguishable classical particles in a
single-particle state k. The random variable nk is Poisson-distributed.
William Feller, An Introduction to PROBABILITY : Theory and its Applications,
Third Edition Volume 1, Wiley Student Edition (1968)p.153
31
B(n = 0; N ) = q N
= (1 p)N
p0
exp(N p)
= P (n = 0; ) = exp()
(3.19)
We get
P (n = 1; N ) = exp()
(3.20)
P (n = 2; N ) =
2
exp()
2!
(3.21)
P (n = 3; N ) =
3
exp()
3!
(3.22)
Finally prove by induction

P (n; N ) =
n
exp()
n!
(3.23)
3.1.4 Poisson process

Consider dots on the time axis recording the arrival time of a neutron in your
detector, the time at which a car passes by you, the time at which an atom
of radioactive substance decays etc. Let t be a small time interval. We take
t to be adequately small such that where ever you place the interval on the
time axis, there shall be either no point or only one point it it. Let
p = t,
be the probability that a point is present in t and q = 1p be the probability
that a point is not present in the interval. is a constant characteristic of the
Poisson process4 . Note that t must be chosen such that
t < < 1 .
Let P (n, t) denote the probability that there are n points between time 0 and
time t. Note that the origin of the time axis - i.e. the time at which you start
the experiment, is arbitrary. The results do not depend on time origin. Hence
4
in radioactive decay, is called the decay constant. Problem: Find how and the
half - life of the radioactive substance are related to each other?
32
we can interpret P (n, t) as the probability that your get n when you count
over a duration of t. Show that,
P (n, t) = P (n, t t)[1 t] + P (n 1, t t)t
(3.24)
The above is called a Master equation. Rewrite the Master equation as

P (n, t) P (n, t t)
= [P (n, t t) P (n 1, t t)] . (3.25)
t
Take the limit t 0 and derive an equation differential in t and difference
in n, given by
P (n, t)
= [P (n, t) P (n 1, t)]
t
(3.26)
The above equation can be solved by an easy method and an easier method.
3.1.5 Easy method
Write down the differential equation for n = 0 and solve it to get,
P (0, t) = exp(t)
where we have taken the intial condition as
0 n 6= 0
P (n, t = 0) = n,0 =
1n=0
(3.27)
(3.28)
Write down the equation for n = 1. In the resulting equation substitute for
P (0, t) and solve the resulting differential equation and get,
P (1, t) = t exp(t)
(3.29)
In the same way proceed for n = 2, 3, and show that

P (2, t) =
(t)2
exp(t)
2!
(3.30)
P (3, t) =
(t)3
exp(t)
3!
(3.31)
From the pattern that emerges we can conclude

P (n, t) =
(t)n
exp(t)
n!
(3.32)
More rigorously, employing the method of induction, we can prove that

the above is indeed the solution of the difference - differential equation.
33
3.1.6 Easier method

Employing the generating function method, show that
P (z; t)
= (1 z)P (z; t)
t
Solve the above and show
P (z, t) = P (z, t = 0) exp [ (1 z) t]
(3.33)
(3.34)
Show that
P (z, t = 0) =
z n P (n.t = 0)
n=1
z n n,0
n=1
=1
(3.35)
and hence
P (z, t) = exp [ (1 z) t]
(3.36)
The next item in the agenda is on Gaussian distribution. It is a continuous

distribution defined for x +. Before we take up the task of
obtaining Gaussian from Poisson (in the limit ), let us learn a few relevant
and important things about continuous distribution.
3.1.7 Characteristic function
Let x = X() be a continuous random variable, and f (x) its probability
density function. The Fourier transform of f (x) is called the characteristic
function of the random variable x = X():
Z +
X (k) =
dx exp(ikx) f (x)
(3.37)
Taylor expanding the exponential in the above, we get

Z
X
(ik)n
X (k) =
dx xn f (x)
n!
n=0
=
X
(ik)n
Mn
n!
n=0
Thus the characteristic function generates the moments.
(3.38)
34
3.1.8 Cumulant generating function

The logarithm of the characteristic function is called the cumulant generating
function.
X (k) = ln X (k)
(3.39)
Let us write the above as,
X
(ik)n
Mn
X (k) = ln 1 +
n!
n=1
= ln(1 + )
X
(1)n+1 n
=
n
n=1
X
(1)n+1
=
n!
n=1
X
(ik)m
Mm
m!
m=1
!n
(3.40)
We now express X (k) as a power series in k as follows

X (k) =
X
(ik)n
n
n!
n=1
(3.41)
where n is called the n-th cumulant.

From the above equations we can find the relation between moments and
cumulants.
3.1.9 Sum of identically distributed independent random variable
Let us consider the sum of N independent and identically distributed
random
PN
variable, {Xi : i = 1, 2, , N }. We thus have, Y = i=1 Xi . Let us
consider scaled random variable Y = Y /N and enquire about its distribution
in the limit N . We have,
Z
Y (k) =
dy exp(iky)f (y)
dx1
exp ik
dx2
dxN
x1 + x2 + xN
N

f (x1 , x2 , xN )
(3.42)
35
The random variables are independent. Hence

f (x1 , x2 , xN ) = f (x1 )f (x2 ) f (xN ).
We have,
Y (k)
dx1 exp(ikx1 /N )f (x1 )
Z
dxN exp(ikxN /N )f (xN )
N
dx exp(ikx/N )f (x)
N
[X (k k/N )]
exp [N ln X (k k/N )]
dx2 exp(ikx2 /N )f (x2 )
"
"
X
(ik)n n
exp N
n! N n
n=1
X
(ik)n n
exp
n! N n1
n=1

k2 2
2
+ O(1/N )
exp ik
2! N

k2 2
exp ik
2! N
(3.43)
Thus the characteristic function of Y , in the limit N is exp(ik

(k 2 /2!) 2 /N ). We will show below, that this is the characteristic function of
Gaussian random variable with mean and variance 2 /N .
Thus the sum of N independent and identically distributed random variables (with finite variance) tends to have a Gaussian distribution for large N .
This is called the central limit theorem.
3.1.10 Poisson Gaussian
Start with the moment generating function of the Poisson random variable:
36
P (z; ) = exp[(1 z)].
(3.44)
In the above substitute z = exp(ik) and get,

P (k; ) = exp [ {1 exp(ik)}]
(3.45)
Substitute the power series expansion of the exponential function and get,
"
#
X (ik)n
P (k; ) = exp
(3.46)
n!
n=1
We recognise the above as the cumulant expansion of a distribution for which

all the cumulants are the same . For large value it is adequate to consider
only small values of k. Hence we retain only terms upto quadratic in k. Thus
for k small, we have,

k2
P (k) = exp ik
(3.47)
2!
The above is the Fourier transform or the characteristic function of a Gaussian
random variable with mean as and variance also as .
0.35
0.14
0.3
0.12
0.25
0.1
0.2
0.08
0.15
0.06
0.1
0.04
0.05
0.02
0
5
10
15
20
n
exp() with mean ; P (n) versus n;
n!

1
(x )2
depicted as sticks; Gaussian distribution : G(x) = exp
with mean
2 2
2
and variance 2 = : continuous line.
(Left) = 1.5; (Right) = 9.5. For large Poisson and Gaussian coincide
Fig. 3.2. Poisson distribution : P (n) =
Thus in the limit , Gaussian distribution with mean and variance

both equal to is a good approximation to Poisson distribution with mean
, see figure above.
37
Problems
3.1. Derive an expression for hn(n 1)i - called the first factorial moment.
Find the variance of the random variable n.
3.2. Consider a system to two coins each with P (H) = p = 0.6 and P (T ) =
q = 1 p = 0.4.
(a) Write down the micro states of the experiment.
(b) Write down a possible ensemble of micro states describing the probabilities
of the micro states.
3.3. Consider the moment generating function of the binomial random vari
able given by B(z)
= (q + zp)N . Let hnk i denote the k the moment of n. By
taking the derivatives with respect to z calculate the first four moments of n
. Let 2 = hn2 i hni2 denote the variance of n. Calculate 2 . Relative fluctuations of n are given by /hni. How does this quantity vary with increase
of N ?
3.4. Consider a coin for which the probability of Heads is p and the probability
of Tails is q = (1 p). The experiment consists of tossing the coin until you
get Heads for the first time. The experiment stops once you get Heads. Let n
denote the number of tosses in an experiment.
1. What is the sample space or micro-state space underlying this experiment ?
2. What is the discrete probability density function of n ?
3. From the probability density function calculate the mean and variance of
n
4. Derive an expression for the moment generating function/partition function of the random variable n
5. From the moment generating function calculate the mean and variance of
the random variale n.
3.5. Consider a random walker starting from origin of a one dimensional lattice. He tosses a coin; if Heads he takes a step toward right; if Tails he steps
to the left. Let p be the probability for Heads and q = 1p be the probability
for Tails. Let P (m, n) denote the probability that the random walker is at
m after n steps. Derive an expression for P (m, n). Let nR be the number of
right jumps and nL be the number of left jumps. We have n = nR + nL and
m = nR nL . We have P (nR , nL ; n) = [n!/(nR !nL !)]pnR q nL . etc.
3.6. Consider the problem with V = 10 M3 ; v = 103 M3 and N = 105 . The
probability of finding n molecules in v is given by the Poisson distribution
with = N p = 10. Plot the Poisson distribution.
3.7. Show that P
the mean and variance of a Poisson
variable are the
P random
2
same : M1 =
n
P
(n;
)
=
;
M
=
n
P
(n;
) =?; 2 =
2
n=0
n=0
2
M2 M1 =
38
3.8. Show that, 1 = M1 ; 2 = M2 M12 ; 3 = M3 3M2 M1 + 2M13 ; 4 =

M4 4M3 M1 3M22 + 12M2M12 6M14 .
3.9. Let x = X() be an exponential random variable. It is defined for x 0;
the probability density function is given by
f (x) = exp(x).
Show that the characteristic function is given by
(k) =
1
.
1 + ik
Show that the n-th moment is given by

Mn = (n + 1) = n!
3.10. The characteristic function of the Gaussian random variable is formally
expressed as,

Z
1
(x )2
(k) =
ikx
dx exp
2 2
2
Carry out the above integral and show that,

k2 2
(k) = exp ik
2!
3.11. Let
Y =
N
1 X
Xi .
N i=1
where {Xi : i = 2, N } are independent and identically distributed exponential random variables.
An exponential random variable is defined for x 0 and its probability
density function is given by exp(x). The n-th cumulant is (n1)!. In particular its mean is unity and its variance is also unity. Show that the characteristic
function of Y is given by

ik
Y (k) = exp N ln 1 +
N
We have,
ln(1 x) =
X
xn
.
n
n=1
Substitute the above in Y (k) and show that

Y (k) = exp
X
(ik)n (n 1)!
n!
N n1
n=1
(3.48)
39
Demonstrate that the terms with n 3 can be neglected when N . The

resulting characteristic function is that of a Gaussian with mean unity and
variance (1/N ).
3.12. There are N ideal gas molecules in a big room of volume V = 10 M 3 .
The molecules are in equilibrium. There is a box of volume v < V , contained completely inside the room. The walls of the box are permeable and
conducting. In other words the box can exchange energy and matter with the
surroundings.
Let be a random variable defined as the number of molecules in the box.
Let P (n) denote the probability that the random variable takes a value n.
Let = N v/V . Define, nL = floor0.99 , nU = ceil1.01 , where, floorx
is called the floor of x and it denotes the largest integer less than or equal to
x; e.g. floor9.8 = 9; floor = 3; etc., and ceilx is called ceil of x and it denotes
the smallest integer greater that or equal to ; e.g. ceil9.8 = 10; ceil = 4; etc..
Calculate the probability that the random variable takes a value between
nL and nU , for the following three cases.
(i) v = 6 cubic m ; N = 10
(ii) v = 10 cubic cm. ; N = 105
(iii) v = 10 cubic mm. ; N = 1014 .
3.13. Start with a Poisson distribution of mean : P (n) = n exp(/n!.
Employ Stirling approximation for n! (n! = nn exp(n)) and write P (n) formally as P (n) = exp[F (n)], where, F (n) = n ln(n) n n ln + . Solve
the equation
dF
=0
dn
and show that F is an extremum at n = . Then expand F (n) around as a
Taylor series retaining terms upto second derivatives and and show that P (n)
is a Gaussian with mean and variance . This is valid only for large n : the
Poisson distribution peaks at n = only when is large.
Plot the Poisson distribution for various values of = .1; .5; 2.0; 10. Plot
also Gaussian distribution with mean =variance =. Demonstrate that the
Poisson and Gaussian coincide for large .
4
Isolated system: Micro canonical ensemble
4.1 Preliminaries
We are going to study an isolated system of N particles confined to a volume
V . The particle do not interact with each other. We will count the number of
b of the system. This will be in general
micro states, denoted by the symbol ,
a function of energy E, volume V and the number of particles N . We shall
do the counting for both classical and quantum particles. Before we address
the full problem, we shall consider a simpler problem of counting the micro
states taking into account only the spatial coordinates neglecting completely
the momentum coordinates. Despite this simplification, we shall discover that
statistical mechanics helps you derive the ideal gas law1 .
I must tell you of a beautiful derivation of the ideal gas law by Daniel
Bernoulli(1700-1782). It goes as follows. Bernoulli imagined air to be made of
tis billiard balls all the time in motion, colliding with each other and with the
walls of the container. When a billiard ball bounces off the wall, it transmits a
certain momentum to the wall and Bernoulli imagined it as pressure. It makes
sense. First consider air contained in a cube of side one meter. There is a certain
amount of pressure felt by the wall. Now imagine the cube length to be doubled
with out changing the speeds of the molecule. In modern language this assumption is the same as keeping the temperature constant. The momentum transferred
per collision remains the same. However since each billiard ball molecule has to
travel twice the distance between two collision the force on the wall should be
smaller by an factor of two. Also pressure is force per unit area. The ares of the
side of the cube is four times more now. Hence the pressure should be less by a
further factor of four. Taking into account both these factors, we find the pressure
should be eight times less. We also find the volume of cube is now eight times
more. Bernoulli concluded that the product of pressure and volume must be a
constant when there is no change in the molecular speeds - a brilliant argument
based on simple scaling ideas.
42
4 Isolated system: Micro canonical ensemble
4.2 Configurational entropy

Consider placing a single particle in a volume V divided into two equal halves.
Let = V /2. There are two ways, see figure below.
Fig. 4.1. Two ways of keeping a particle in a box divided into two equal parts.
V
b
(V,
N = 1, = V /2) =
=2
b = kB ln(2)
S = kB ln
(4.1)
(4.2)
Now consider two distinguishable particles into these two cells each of
volume = V /2, see figure below.
Fig. 4.2. Four ways of keeping two distinguishable particles in a box divided into
two equal halves.
We then have
b
(V,
N = 2, = V /2) =
For N particles we have,
2
=4
b = 2kB ln(2)
S = kB ln
(4.3)
(4.4)
4.3 Ideal gas law : Derivation
b
(V,
N, = V /2) =
N
= 2N
b = N kB ln(2)
S = kB ln
43
(4.5)
(4.6)
Let us now divide the volume equally into V / parts and count the number
of ways or organizing N (distinguishable) particles. We find
b
(V,
N) =
N
(4.7)
b
S = kB ln
= N kB ln(V /)
= N kB ln V N kB ln
(4.8)
We will discover later that the above formula captures the volume dependence
of entropy quite accurately.
4.3 Ideal gas law : Derivation

Differentiate S given by Eq. (??), with respect V . We get,

N kB
S
=
V E,N
V
(4.9)
From thermodynamics2 we have

2
In thermodynamics we start with U U (S, V ) for a given quantity of say an

ideal gas. This internal can change in a (quasi static) reversible process either by
heat, T dS, or by work, P dV . Hence we have the first law of thermodynamics
du = T dS P dv. We have then, by definition

U
T =
S V
P =
U
V
Let us start with S S(U, V ), which is natural for statistical mechanics. We

have,
44
S
V
E,N
P
T
(4.10)
Thus we see from Eq. (??) and Eq. (??)

P V = N kB T
(4.11)
4.4 Boltzmann entropy and Clausius entropy are the

same
From Eq. (??), we have,
dS =
N kB
dV
V
(4.12)
Employing the equation of state : P V = N kB T , which we have derived, we

can rewrite the above as
dS =
P dV
T
(4.13)
Consider an isothermal process in an ideal gas. We have dU = 0. This implies

T dS = P dV . When the system absorbs a certain quantity of heat q isothermally and reversibly, we have q = T dS = P dV . Equating P dV to q in Eq.
(??), we get
dS =
q
T
(4.14)
which shows that Boltzmann entropy and thermodynamic entropy are the
same.
dS =
S
U
+
V
S
V
To express the partial derivatives in the above in terms of T and P , we rearrange

the terms in the first law equation as,
dS =
P
1
dU + dV
T
T
Equating the pre-factors of dU and dV in the above two equation, we get,

S
1
=
U V
T

S
V
=
U
P
T
4.6 Boltzmann counting
45
4.5 Some issues on extensitivity of entropy

The expression for entropy given below,
S(V, N ) = N kB ln V N kB ln
is not extensive. If I double the value of V and of N , I expect S to be doubled.
It does not. Mathematically S is extensive if it is a first order homogeneous
function V and N . In other words we should have S(V, N ) = S(V, N ).
The above expression for entropy does not satisfy this rule3 .
4.6 Boltzmann counting

To restore extensivity of entropy, Boltzmann introduced an ad-hoc notion of
indistinguishable particles. N ! permutations of the particles, should all be
counted as one micro state since they are indistinguishable4 . Hence,
1
b
(V,
N) =
N!
N
b
S(V, N ) = kB ln (V,
N)
= N kB ln
V
N
(4.15)
+ N kB N kB ln
(4.16)
S(V, N ) = S(V, N )
Time has come for us to count the micro states of an isolated system
of N non interacting point particles confined to a volume V , taking into
considerations the positions and the momenta of all the particles.
Each particle for its specification requires six numbers : three positions and
three momenta. The entire system can be specified by a string of 6N numbers.
In a 6N dimensional phase space the system is specified by a point. The phase
space point is all the time moving. We would be interested determining the
region of the phase space accessible to the system when it is in equilibrium.
3
This is called Gibbs paradox. More precisely Gibbs formulated the paradox in
terms of entropy of mixing of like and unlike gases. We shall see these in details
later when we consider closed system described by canonical ensembles.
The remedy suggested by Boltzmann is only temporary. Non extensivity of entropy points to a deeper malady in the statistical mechanics based on classical
formalism. For the correct resolution of the non-extensivity-paradox we have to
wait for the arrival of Quantum Mechanics. We shall see of these issues in details
later when we consider quantum statistics.
46
If we are able to count the phase space volume, then we can employ the first
micro-macro connection proposed by Boltzmann and get an expression for
entropy as a function of energy, volume and the number of particles.
The system is isolated. It does not transact energy or matter with the
surroundings. Hence its energy remains a constant. The potential energy is
zero since the particles do not interact with each other. The kinetic energy is
given by
E=
3N
X
p2i
2m
i=1
(4.17)
The system is thus confined to the surface of a 3N dimensional sphere. We

need a formula for the volume of an hyper-sphere in 3N dimensions. To this
end we need to know of Heaviside5 theta function and Dirac6 delta function.
4.7 Heaviside Theta function

Define a function
where > 0.
Define
0
for x

1
1
f (x; ) =
x + for x +
2
2
2
1
for + x +
2
(4.18)
lim.
(x) = 0 f (x; ).
(x) is called the step function, Heaviside step function, unit step function
or theta function. It is given by,
0 for x < 0
(4.19)
(x) =
1 for 0 < x +
5
6
Oliver Heaviside(1850-1925)
Paul Adrien Maurice Dirac(1902-1984)
4.9 Area of a circle
47
4.8 Dirac delta function

Start with the function f (x; ) defined by Eq. (??). Take the derivative of the
function. We find that the derivative is 1/, when /2 < x + /2 and zero
otherwise, Define,
0 for x < /2
df
g(x; ) =
(4.20)
= 1 for /2 < x + /2
dx
0 for +/2 < x +

The Dirac-delta function is defined as,
(x) = limit
0 g(x; )
(4.21)
Consider the following integral.

Z +
I=
dx g(x; )
(4.22)
We find that the integral is the same for all values of . This gives us an
important property of the Dirac-delta cunction:
Z +
dx (x) = 1
(4.23)
4.9 Area of a circle

Let us demonstrate how to use the theta function and delta function to derive
an expression for a circle of radius R. Let us denote the area of a circle by
the symbol V2 (R) - the volume of a two dimensional sphere of radius R. A
little thought will tell you,
!
Z +
Z +
2
X
x2i
(4.24)
V2 (R) =
dx1
dx2 R2
i=1
Let
yi = xi /R for i = 1, 2.
Then,
V2 (R) = R2
We have
dy1
dy2 R2 (1
2
X
i=1
yi2
(4.25)
48
(x) = (x) > 0.
Therefore,
V2 (R) = R
dy1
2
X
dy2 1
yi2
i=1
(4.26)
= R2 V2 (R = 1)
(4.27)
We can now write Eq. (??) as

Z
V2 (R = 1)R2 =
dx1
dx2 R2
2
X
x2i
i=1
(4.28)
Now differentiate both sides of the above equation with respect to the variable
R. We have already seen that the derivative of a Theta function is the Diracdelta function. Therefore
!
Z +
Z +
2
X
2
2
(4.29)
xi
V2 (R = 1)2R = 2R
dx1
dx2 R
i=1
Now multiply both sides of the above equation by exp(R2 )dR and integrate
over the variable R from 0 to . We get,
Z
Z
Z +
V2 (R = 1)
exp(R2 )2RdR =
exp(R2 ) 2RdR
dx1
0
V2 (R = 1)
dt exp(t) =
V2 (R = 1) 1 =
Z
V2 (R = 1) = 2
V2 (R = 1) =
dx1
Z
dx2 R
Z
(4.30)
dx2 exp(x21 x22 ) (4.31)
2
dx exp(x )
2 Z
dx exp(x ) =
i=1
x2i
Z
2
X
1/2
dx x
(4.32)
2
exp(x) (4.33)
2
2
x(1/2)1 exp(x)dx = [ (1/2)] =
(4.34)
Thus V2 (R) = V2 (R = 1) R2 = R2 , a result we are all familiar with.
4.10 Volume of an N -dimensional sphere
49
4.10 Volume of an N -dimensional sphere

The volume of an N - dimensional sphere of radius R is formally given by the
integral,
!
Z +
Z +
Z +
N
X
x2i
(4.35)
VN (R) =
dx1
dx2
dxN R2
i=1
Change the coordinate system from

{xi : i = 1, N } to {yi = xi /R : i = 1, N }.
dxi = Rdyi i = 1, N
"
N
X
yi2
i=1
#!
= 1
N
X
yi2
i=1
We have,
VN (R) = R
dy1
dy2
dyN 1
N
X
yi2
i=1
= VN (R = 1)RN
(4.36)
(4.37)
where VN (R = 1) is the volume of an N - dimensional sphere of radius unity.

To find the volume of N -dimensional sphere of radius R, we proceed as
follows.
!
Z +
Z +
Z +
N
X
2
N
2
xi(4.38)
VN (R = 1)R =
dx1
dx2
dxN R
i=1
Differentiate both sides of the above expression with respect to R and get,
Z +
Z +
dx1
N VN (R = 1)RN 1 =
dx2
dxN R
N
X
i=1
x2i
2R (4.39)
Now, multiply both sides by exp(R2 )dR and integrate over R from 0 to .
The Left Hand Side:
50
LHS = N VN (R = 1)
dR exp(R2 )RN 1
(4.40)
Let x = R2 ; then dx = 2RdR. This give

1 dx
.
2 x1/2
dR =
We get,
N
LHS = VN (R = 1)
2
= VN (R = 1)

N
+1
2
The Right Hand Side :

Z
Z
RHS =
dR exp(R2 )
N
2
VN (R = 1)
dx1
x 2 1 exp(x)dx
dxN
(4.41)
dx2
N
X
i=1
x2i
2R
(4.42)
t = R2
dt = 2RdR
RHS =
dt exp(t)
dx1
dxN
dx1
Z
= N/2
dx2
t
Z
N
X
i=1
dx2
x2i

dxN exp (x21 + x22 + x2N )
N
dx exp(x )
2
(4.43)
4.12 Density of states : g(E)
51
Thus we get
VN (R = 1) =
VN (R) =
N/2

N2 + 1
N/2
RN
N2 + 1
(4.44)
(4.45)
4.11 Classical counting of micro states

Consider an isolated system of N non-interacting point particles. Each particle requires 3 position coordinates and 3 momentum coordinates for for its
specification. A string of 6N numbers denotes a micro state of the system.
Let us first find the volume of the phase space accessible to the system. The
integral over 3N spatial coordinates gives V N . We have
E=
3N
X
p2i
2m
i=1
The volume of the phase space of the

system with energy E is the volume
of a 3N dimensional sphere of radius 2mE.
4.11.1 Counting of the volume
Let us measure the volume of the phase space in units of h3N , where h is
Planck constant. We have
xpx h.
Thus h3N is the volume of a minimum uncertainty cube. Thus we have
V N (2mE)3N/2
b

(E,
V, N ) = 3N
h
3N
2 +1
(4.46)
4.12 Density of states : g(E)

Let g(E) denote the density of (energy) states. g(E)dE gives the number of
micro states with energy between E and E + dE. In other words,
From the above, we find
b
(E)
=
E
0
g(E )dE
(4.47)
52
b
(E,
V, N )
E
g(E, V, N ) =
(4.48)
V,N
b
Let us take the partial derivative of (E,
V, N ) with respect to E and get,
g(E, V, N ) =
V N (2m)3N/2 3N (3N/2)1
E
h3N ( 3N
2
2 + 1)
(4.49)
Let us substitute N = 1 in the above and get the single particle density of
states, g(E, V ) as,
g(E, V ) =
V
(8m)3/2 E 1/2
h3 4
(4.50)
4.12.1 A sphere lives on its outer shell : Power law can be

intriguing
In the limit of N , the volume of a thin outer shell tends to the volume
of the whole sphere. This intriguing behaviour is a consequence of the power
law behaviour.
VN (R) VN (R R)
RN (R R)N
=
VN (R)
RN
(4.51)
N

R
= 1 1
R
(4.52)
= 1 for N
(4.53)
Hence in the limit of N the number of micro states with energy less
than or equal to E is nearly the same as the number of micro states with
energy between E E and E.
4.13 Entropy of an isolated system

From Eq. (??) we see that

3
3
3
E
4m
S(E, V, N ) = N kB ln V + ln
+ ln
+
2
N
2
3h2
2
We find that the above expression for entropy is not extensive. In other words
S is not an extensive function of V 7 : S(E, V, N ) 6= S(E, V, N ). We
7
Note that S is extensive in E and N .
4.14 Properties of an ideal gas
53
shall follow Boltzmanns prescription and divide (E, V, N ), see Eq. (??), by
N !.
V N 1 (2mE)3N/2
b

(E,
V, N ) = 3N
h N ! 3N
2 +1
(4.54)
The corresponding entropy is then,

V
3
3
5
E
4m
S(E, V, N ) = N kB ln
+ ln
+ ln
(4.55)
+
N
2
N
2
3h2
2
4.14 Properties of an ideal gas

The temperature of an ideal gas, as a function of E, V , and N , is given by

S
1
3N kB
=
=
(4.56)
E V,N
T
2E
T =
2E
3N kB
The energy of the system is thus given by,

1
kB T
E = 3N
2
(4.57)
(4.58)
The above is called equi-partition theorem. Each quadratic term in the Hamiltonian carries an energy of kB T /28 .
The pressure of an isolated system of ideal gas as a function of E, V , and
N , is given by,

P
N kB
S
=
=
(4.59)
V E,N
T
V
P =
N kB T
V
(4.60)
Substituting in the above T as a function of E, V , and N , see Eq. (??), we

get,
8
For an ideal gas

H=
3N
X
p2i
.
2m
i=1
There are 3N quadratic terms in the Hamiltonian.
54
P =
2E
3V
(4.61)
An expression for the chemical potential as a function of E, V , and N is

derived as follows.

S
=
(4.62)
N E,V
T
Therefore,
= N kB T ln
V
N
3
N kB T ln
2
4mE
3N h2
(4.63)
Substituting in the above the expression for T from Eq. (??), we get,

4mE
V
2E
2E ln
(4.64)
ln
=
3
N
3N h2
In the above expression for the micro canonical chemical potential, let us
substitute
E = 3N kB T /2
and express chemical potential in terms of T , V and N . We get,

3
2mkB T
V
N kB T ln
= N kB T ln
N
2
h2
= N kB T ln
V
N
+ 3N kB T ln()
(4.65)
(4.66)
where is the thermal or quantum wavelength9 given by,

h
=
2mkB T
(4.67)
Let the number density be denoted by the symbol . We have thus =

N/V . We can write the chemical potential in a compact form
= N kB T ln(3 )
(4.68)
Let me end this section by saying that the micro canonical ensemble formalism
leads to the ideal gas law: See the expression for P given in Eq.(??). We have,
P V = N kB T
we shall see about it later
(4.69)
4.15 Quantum counting of micro states
55

We have done classical counting of micro states and showed that for an isolated
system of a single particle confined to a volume V , the number of micro states
with energy less than , is given by
3/2
V (2m)
b
(E,
V) = 3
h ((3/2) + 1)
(4.70)
We have obtained the above by substituting N = 1 in Eq. (??).

Now let us do quantum counting of micro states 10 .
10
Some blah .... blah ... on Quantum Mechanics

A fundamental entity in quantum mechanics is the wave function (q, t), where
q is the position vector. The wave function is given a physical interpretation that
(q, t)(q, t)dq
gives the probability of finding the system in an elemental volume dq around the
point q at time t. Since the system has to be somewhere, for, otherwise we would
not be interested in it, we have the normalization,
Z
(q, t)(q, t)dq = 1,
where the integral is taken over the entire coordinate space - each of the x, y and
z coordinates extending from to +.
A central problem in quantum mechanics is the calculation of (q, t) for the
system of interest. We shall be interested in the time independent wave function
(q) describing a stationary states of the system.
How do we get (q) ?
Schr
odinger gave a prescription : Solve the equation H(q) = E(q), with appropriate boundary conditions.
We call this the time independent Schr
odinger equation.
H is the Hamiltonian operator
H=
~2
2 +U (q)
2m
The first operator on the right is kinetic energy and the second the potential
energy.
E in the Schr
odinger equation is a scalar ... a real number... called energy. It
is an eigenvalue of the Hamiltonian operator : we call it energy eigenvalue.
The Schr
odinger equation is a partial differential equation. Once we impose
boundary condition on the solution, then only certain discrete energies are permitted. We call these energy eigenvalues.
Energy Eigenvalue by Solving Schr
odinger Equation Once we specify
boundary conditions, then a knowledge of the Hamiltonian is sufficient to determine its eigenvalues and the corresponding eigenfunctions. There will be usually
several eigenvalues and corresponding eigenfunctions for a given system.
56
4.15.1 Energy eigenvalues : Integer Number of Half Wave lengths

in L
Let me tell you how to obtain the energy eigenvalues without invoking
Schrodinger equation. Consider a particle confined to a one dimensional box
of length L. We recognise the segment L must contain integral number of half
wave lengths - so that the wave function vanishes at the boundaries of the one
dimensional box. In other words,
L = n
=
2L
n
: n = 1, 2, ,
: n = 1, 2, ,
(4.71)
(4.72)
Substitute the above in the de Broglie relation

p=
h
h
=
n : n = 1, 2, .
2L
This yields
h2
p2
n2 : n = 1, 2, .
=
2m
8mL2
Consider a particle in an L L L cube - a three dimensional infinite well.
The energy of the system is given by
n =
nx ,ny ,nz =
h2
(n2 + n2y + n2z )
8mL2 x
where nx = 1, 2, , , ny = 1, 2, , and nz = 1, 2 , .
The ground state is (nx , ny , nz ) = (1, 1, 1); it is non degenerate; the energy
eigenvalue is
1,1,1 =
3h2
8mL2
(4.73)
For a single particle in a one dimensional infinite well

H=
~2 2
2m x2
Solve the one dimensional Schr

odinger equation with the boundary condition : the wave function vanishes at the boundaries. Show that the energy
eigenvalues are given by
n =
h2
n2 ;
8mL2
n = 1, 2,
57
The first excited state is three-fold degenerate. The corresponding energy

eigenvalue is
2,1,1 = 1,2,1 = 1,1,2 =
3h2
.
4mL2
(4.74)
We start with,
=
h2
(n2 + n2y + n2z )
8mL2 x
We can write the above as,

8mL2
= R2
h2
(nx , ny , nz ) represents a lattice point in the three dimensional space. The
equation n2x + n2y + n2z = R2 says we need to count the number of lattice
points that are at a distance R from the origin. It is the same as the number
of lattice points that are present on the surface of a sphere of radius R in the
positive quadrant; note that the x, y, and z coordinates of the lattice points
are all positive. It is difficult to count the number of lattice points lying on
the surface of a sphere. Instead we count the number of points contained in
a thin spherical shell. To calculate this quantity we first count the number of
points inside a sphere of radius

1/2
8mL2
R=
h2
n2x + n2y + n2z =
b
and take one-eighth of it. Let us denote this number by ().
We have,

3/2
8mL2
1 4
b
R3 =
(4.75)
()
=
8 3
6
h2
We recognize V = L3 and write the above equation as
b V ) = V (8m)3/2 = V 4 (2m)3/2
(,
h3 6
h3 3
=
V (2m)3/2
V 4 (2m)3/2
= 3
3
3/2
h 3
h (3/2)(1/2)
(2m)3/2
V (2m)3/2
V
= 3
3
h (3/2)(1/2) (1/2)
h ((3/2) + 1)
(4.76)
The above is exactly the one we obtained by classical counting, see Eq. (??)
Notice that in quantum counting of micro states, the term h3 comes naturally,
while in classical counting it is hand-put11 .
11
We wanted to count the phase space volume. We took h3 as the volume of a

six-dimensional cube. We considered the six-dimensional phase space (of a single
58
b V ) with
The density of (energy) states is obtained by differentiating (,
respect to the variable . We get
g(, V ) =
V
(8m)3/2 1/2
h3 4
(4.77)
The important point is that the density of energy states is proportional to

1/2 .
4.16 Chemical Potential

4.16.1 Toy model
Consider an isolated system of two identical, distinguishable and non-interacting
particles occupying non-degenerate energy levels { 0, , 2, 3, }, such that
b
the total energy of the system is 2. Let (E
= 2, N = 2) denote the number
of micro states of the two-particle system with total energy E = 2. For an
isolated system, since all the micro states are equally probable the micro state
space is a good candidate for the micro canonical ensemble.
We label the two particles as A and B. The micro states with total energy
2 are given below.
0
A
B
A, B
2
B
A
Table 4.1. Micro states of of two particles with total energy 2
We find that
b
(E
= 2, N = 2) = 3.
The entropy of the two-particle system with energy E = 2 is given by

b
S(E = 2, N = 2) = kB ln (E
= 2, N = 2).
Now add a particle, labelled C, such that the energy of the system does
not change. In other words, the three-particle system has energy 2 which is
b
the same as that of the two particle system. Let (E
= 2, N = 3) denote
the number of micro states of the three-particle system with a total energy of
E = 2. The table below gives the micro states.
particle) as filled with non-overlapping exhaustive set of such tiny cubes. We have
to do all these because of Boltzmann ! He told us that entropy is logarithm of
number of micro states. We need to count the number of micro states.
0
A, B
B, C
C, A
A
B
C
B, C
C, A
A, B
59
2
C
A
B
Table 4.2. Micro states of three particles with total energy 2
We find
b
(E
= 2, N = 3) = 6.
The entropy is given by
b
S(E = 2, N = 3) = kB ln (E
= 2, N = 3).

U
.
N S,V
In other words, is the change in energy of the system when one particle
is added in such a way that the entropy and volume of the system remain
unchanged. To achieve this we must remove amount of energy from the
three particle system. In other words we demand S(E = , N = 3) = S(E =
2, N = 2).
Thus = . This problem that the chemical potential of a system of
non-interacting particles is negative.
We find S(E = 2, N = 3) > S(E = 2, N = 2). Note that =
4.16.2 Chemical potential of an ideal gas

The number of micro states of a classical system of N non-interacting particles
confined to a volume V is given by,
V N 1 (2mE)3N/2
b
(E,
V, N ) = 3N
h N ! ( 3N
2 + 1)
The entropy of the system defined as
is given by,
b
S(E, V, N ) = kB ln (E,
V, N )

3
5
3
E
4m
V
+ ln
. (4.78)
+
+ ln
S(E, V, N ) = N kB ln
N
2
N
2
3h2
2
The chemical potential is given by

(E, V, N ) = T
S
N
E,V
60
Problems
4.1. Consider an isolated system of N non-interacting point particles occupying two states of energies and +. The energy of the system is E. Define
E
. Show that the entropy of the system is given by
x=
N

N kB
1+x
1x
S(x) =
(1 + x) ln
+ (1 x) ln
2
2
2
Also show that
kB
1
=
ln
T
2
1x
1+x
4.2. Sketch the function f (x; ) for = 2, 1, 1/2, 1/4; sketch also the function
in the limit of 0.
4.3. Sketch g(x; ) for = 2, 1, 1/2, 1/4. How does it look in the limit 0
?
4.4. Define the Dirac-delta function centered at x0 as (x x0 ). To this end,

start with g(x; , x0 ) and show that in the limit 0 it defines a theta
function with a step at x = x0 6= 0; the theta function is denoted by (xx0 ).
Take the derivative of g(x; , x0 ) and show that in the limit 0 we get the
Dirac-delta function centered at x0 . It is denoted by (x x0 ). Show that
Z +
dx (x x0 ) = 1
4.5. Show that (1/2) =
dx (x)(x x0 ) = (x0 )
4.6. The volume of a (three dimensional) sphere of radius R is denoted by

V3 (R). Show that
4
V3 (R) = R3
3
4.7. Substitute N = 1, 2, and 3 in the expression above for VN (R) and show
that you recover the standard results V1 (R) = 2R, V2 (R) = R2 , and V3 (R) =
4 3
R .
3
4.8. Consider a particle in a two dimensional infinite well. In other words,
consider particle in a square of area A. Carry out quantum counting and
show that the density of (energy) states is
A
2m
h2
Notice that the density of (energy) states is independent of .
g(, A) =
61
4.9. Consider a particle in a one dimensional infinite well. In other words

consider a particle confined to a line segment of length L. Carry out quantum
counting and show that the density of (energy) states is
L
1/2
(2m) 1/2
h
4.10. Carry out classical counting of the number of micro states of an isolated system of N particles confined to an area A, with energy E. Since the
particles are confined to a plane we require 2N position and 2N momentum
b
coordinates to specify a micro state. Calculate (E,
A, N ) the number of micro states with energy less than or equal to E. Substitute N = 1. Differentiate
the resulting expression with respect to energy and verify whether you get the
same answer as of Problem No. 4.7
g(, L) =
4.11. Carry out classical counting of the number of micro state of an isolated system of N particles confined to a length L with energy E. Since the
particles are confined to a line we require N position and N momentum coorb
dinates to specify a micro state. Calculate (E,
L, N ) - the number of micro
states with energy less than or equal to E. Substitute N = 1. Differentiate the
resulting expression with respect to E and verify whether you get the same
answer as of Problem No. 4.9
4.12. Show that we must remove of energy from the system to restore enb
tropy to its original value. In other words show that (E
= , N = 3) =
b
(E = 2, N = 2) = 3
4.13. Derive an expression for as a function of E, V and N . Substitute in
the expression12

1
E = 3N
kB T
2
and show that as a function of T , V and N is given by13 ,

3
2mkB T
V
kB T ln
.
(T, V, N ) = kB T ln
N
2
h2
Now add a particle to the system such that the energy change from E to E +
and the entropy changes from S(E, V, N )
to S(E + , V, N + 1).
To derive an expression for S(E + , V, N + 1) replace in Eq. (??): E by
E + and N by N + 1.
Show that for = we get S(E + , V, N + 1) = S(E, V, N ). In the
derivation, assume N to be large and the magnitude of to be small compared
to total energy i.e || << E.
12
13
Equipartition theorem : Every quadratic term in the Hamiltonian carries an energy of kB T /2.
Read G Cook and R H Dickerson, Understanding the chemical potential, American
Journal of Physics 63(8), 737 (1995)
5
Closed system : Canonical ensemble
5.1 What is a closed system ?

A closed system is one which does not exchange material with the surroundings. However, it does exchange energy. It is in thermal contact with the
surroundings. We idealize the surroundings of a closed system as a heat
bath1 .
Thus, a system in thermal equilibrium, is characterized by T , V and N .
The system is not isolated. Hence its micro states are not all equi-probable.
5.2 Toy model a la H B Callen

Let us illustrate this by considering a toy problem proposed by H B Callen2 .
Consider a fair red die representing the system and two fair white dice representing the surroundings. Let P (k) denote the probability that the red die
shows up k given the three dice together add to 6. Note if we do not impose the
condition (the three die add to six), the six micro states are equally probable
for the red die. However because of the condition imposed the micro states
are not equally probable. We find there are 10 micro states with the property
that the three dice add to 6. These are listed below.
Of these, there are four micro states for which the system die shows up
1; therefore P (1) = 0.4. Similarly we can calculate the other probabilities :
1
A heat bath is one which transacts energy with the system, but its temperature
does not change. Example: Keep a cup of hot coffee at temperature say 60 C
in a room; let the room be at temperature 30 C. The coffee cools to 30 C. The
temperature of the room does not increase. May be, I should say the temperature
of the room increases by an extremely small amount which for all practical purposes can be considered as zero. We idealize and say the room is a heat bath and
its temperature does not change when it transacts energy with the cup of coffee.
H B Callen, Thermodynamics and and Itroduction to Thermostatistics, Wiley
Student Edition (2005)
64
5 Closed system : Canonical ensemble
Table 5.1. Micro states of three dice with the constraint that they add to six
P (2) = 0.3; P (3) = 0.2; P (4) = 0.1; P (5) = P (6) = 0. The important point
is that the micro states of the system are not equi probable.
In other words, if the system interacts thermally with the surroundings
then the probability differs from one micro state to the other.
What is the probability of a micro state in a closed system ? Let us calculate the probability in the next section by a fairly straight forward procedure
involving Taylor expansion of S(E). I learnt of this first, from the book of
Balescu3 .
5.3 Canonical partition function

Consider the system, its boundary and the bath - all the three together - as
an isolated system. We know for an isolated system, all the micro states are
equally probable. Let E denote the energy of the isolated system. It remains
a constant.
Now consider a particular micro state of the system. Let us label it as C.
Let its energy be E(C). Note that E(C) < < E. When the closed system is in
b E(C)) micro
its micro state C, the surroundings can be in any one of (E
states of the isolated system4 .
3
4
R Balescu, Equilibrium and nonequilibrium statistical mechanics, Wiley (1975).

I am considering that a micro state of the isolated system can be thought of as
a simple juxtaposition of the micro state of closed system and the micro state of
5.3 Canonical partition function
65
For the isolated system, all micro states are equally probable. Thus we can
say that the probability of finding the closed system in its micro state C is
given by
P (C) =
b E(C))
(E
bt
where we have denoted the total number of micro states of the isolated system
bt .
as
b E(C)). Therefore
We have S(E E(C)) = kB ln (E

1
b
(E E(C) = exp
S(E E(C))
kB
Also since E(C) < < E, we can Taylor expand S(E E(C)) around E retaining
only the first two terms. We get,

S
S(E E(C)) = S(E) E(C)
E E=E
= S(E)
1
E(C)
T
Substituting the above in the expression for P (C), we get,

E(C)
exp [S(E)/kB ]
exp
P (C) =
ct
kB T
= exp[E(C)]
where is a constant and = 1/(kB T ). We can evaluate completely in

terms of the properties of the closed system by the normalization condition
for the probabilities, see below.
X
P (C) = 1
C
exp [E(C)] = 1
Q(T, V, N ) =
X
1
=
exp [E(C)]
the surroundings. The system and the surroundings interact at the boundaries
and hence there shall exist micro states of the isolated system which can not be
neatly viewed as the system micro state juxtaposed with the surroundings micro
state. Such micro states are so few in number we shall ignore them.
66
where Q(T, V, N ) is called the canonical partition function.

Thus we have,
X
Q(T, V, N ) =
exp [E(C)]
C
X
E
b
(E)
exp(E)
b
b
where (E)
is the degeneracy of the eigenvalue E. In other words (E)
is
the number of micro states of the equilibrium closed system with energy E.
If energy is a continuous variable the we consider g(E), the density of
(energy) states. Then g(E)dE denotes the number of micro states with energy
between E and E + dE. Canonical partition function can be expressed as an
integral
Z +
dE g(E) exp[E)
(5.1)
Q(T, V, N ) =
0
The above is just a transform of the g(E) Q(T ) where we have transformed
energy E in favour of T - the temperature5 . Physically the transform means
that We are going from a micro canonical description with independent variables E, (V and N ) to a canonical description with independent variables
T , (V , and N ). In other words we are going from an isolated system with
energy as an independent variable to a closed system with temperature as an
independent variable.
5.4 Canonical partition function :

Method of most probable distribution
Let us now derive an expression for the canonical partition function employing
an easier method - called the method of most probable distribution.
Consider an isolated system. For convenience we imagine it as a big cube.
It contains molecules moving around here and there, hitting against each other
and hitting against the wall. The isolated system is in equilibrium. Remember
that an isolated system left to itself will eventually reach a state of equilibrium
whence all its macroscopic properties are unchanging with time; also the value
of a macroscopic property shall be the same at any region in the system. Let
temperature be T . Note that the temperature is determined by the isolated
system; it is not determined by you. The system attains that temperature for
which its entropy is maximum.
5
If we take = ik, then we have one sided Fourier transform. If we take = s

then we have Laplace transform. etc.
5.4 Canonical partition function : Method of most probable distribution
67
Let us imagine that the isolated system represented by a big cube is divided
into a set of small cubes of equal volumes by means of imaginary walls. Each
cube represents a macroscopic part of the isolated system.
Each small cube is, in its own right, a macroscopic system with a volume
V . Since the the walls of a small cube permits molecules and energy to move
across, the number of molecules in a cube, is not fixed. It shall fluctuate around
some mean value; the fluctuations, however, are extremely small. The above
observations hold good for energy also. Let A denote the number of cubes
contained in the big cube.
The isolated system - the big cube, has a certain amount energy say E and
certain number of molecules N and a certain volume V and these quantities
are constants.
You can immediately see that what we have is a grand canonical ensemble
of open systems - each cube represents an open system. Each cube is a member of a grand canonical ensemble. All the members are identical as far as
their macroscopic properties are concerned. This is to say the volume V , the
temperature T and chemical potential are all the same for all the members.
Now, let us imagine that the walls are made impermeable to movement
of molecules across. A cube can not exchange matter with its neighbouring
cubes. Let us also assume that each cube contains exactly N molecules. Energy in a cube is however not fixed. Energy can flow from one cube to its
neighbouring cubes. This constitutes a canonical ensemble6 .
Aim :
To find the probability for the closed system to be in its micro state i.
First we list down all the micro states of the equilibrium closed system. Let
us denote the micro states as {1, 2, }. Note that the macroscopic properties
T , V , and N are the same for all the micro states. In fact the system switches
from one micro state to another all the time. Let Ei denote the energy of the
system when it is in micro state i. The energy can vary from one micro state
to another. However the fluctuations of energy from its mean value are very
small for an equilibrium macroscopic system.
To each cube, we can attach an index i. The index i denotes the micro
state of the closed system with fixed T , V and N . An ordered set of A indices
uniquely specifies a micro state of the isolated system.
6
In books, canonical ensemble is constructed by taking a system with a fixed value

of V and N and assembling a large number of them in such a way that each
is in thermal contact with its neighbours. Usually these are called metal copies
of the system. The system and its mental copies are then isolated. The isolated
system of a large number of mental copies of the system plus the system placed
somewhere in the middle is called a canonical ensemble. Note : all the mental
copies are identical macroscopically in the sense they all have the same value of
T , V and N . Also other macroscopic properties defined as averages of a stochastic
variable e.g. energy are also the same for all the mental copies. But they might
differ one from the other in their microscopic properties
68
Let us take an example. Let the micro states of the closed system be
denoted by the indices { 1,2,3}. There are only three micro states. Let us
represent the isolated system by a big square and construct nine small squares,
each of which represents a member of the ensemble. Each square is attached
with an index which can be 1, 2 or 3. Thus we have a micro state of the
isolated system represented by
3
2
2
1
3
3
2
3
1
Table 5.2. A micro state with occupation number representation (2, 3, 4)
In the above micro state, there are two squares with index 1, three with
index 2 and four with index 3. Let {a1 = 2, a2 = 3, a3 = 4} be the occupation
number representation of the microstate. There are several micro states having
the same occupation number representation. I haver given below a few of them.
133
212
323
233
312
321
121
232
232
112
223
333
123
123
233
Table 5.3. A few micro states with the same occupation number representation of
(2, 3, 4) There are 1260 micro states with the same occupation number representation
Notice that all the micro states given above have the same occupation
number string {2, 3, 4}. How many micro states are there with this occupation
number string ? We have
b 3, 4) =
(2,
9!
= 1260
2!3!4!
I am not going to list all the 1260 of the microstates belonging to the occupation number string {2, 3, 4}
Let me generalize and say that a string (of occupation numbers) is denoted
by the symbol a
= {a1 , a2 , }, where a1 + a2 + = A. We also have an
additional constraint namely a1 E1 + a2 E2 + = E.
b a) = (a
b 1 , a2 , ) denote the number of micro states of the
Let (
isolated system belonging to the string a
. For a given string, we can define
the probability for the closed system to be in its micro state indexed by i as
5.4 Canonical partition function : Method of most probable distribution
pi (
a) =
ai (
a)
A
69
(5.2)
Note, the string a

= {a1 , a2 } obeys the following constraints.
A
X
i=1
A
X
i=1
ai (
a) = A strings a
ai (
a)Ei = E strings a
(5.3)
(5.4)
Note that the value of pi varies from one string to another. It is reasonable
to obtain the average value of pi over all possible strings a
. We have

X ai (
a)
Pi =
P(
a)
(5.5)
A
a
where P(
a) is the number of micro states of the isolated system in the string a
divided by the total number of micro states of the isolated system : All micro
states of an isolated system are equally probable. We have,
where,
b a)
(
P(
a) = P
b a)
a
(
b a) =
(
A!
a1 !a2 !
(5.6)
(5.7)
In the above we have used the simple notation ai = ai (

a) i = 1, 2, .
b a) for various strings a
Let us take a look at (
. For large A the numb a) will be overwhelmingly large for a particular string, which we shall
ber (
denote as a
. Thus we can write
P ai (a) b
(
a)
a
A
(5.8)
Pi =
P b
a)
a
(
By taking A we can always ensure7 ai (
a) i. In this limit,
the size of the ensemble is arbitrarily large. It should be large enough so that
even a micro state of smallest probability is present in the ensemble.
70
Pi =
b a )
ai (
a ) (
b a )
A (
(5.9)
ai (
a )
A
(5.10)
ai
A
(5.11)
b a) is a maxiThus the problem reduces to finding that string a

for which (
mum. Ofcourse there are two constraints on the string. They are
X
aj (
a) = A a
(5.12)
j
X
j
aj (
a)Ej = E a
(5.13)
We need to find the maximum (or minimum) of a function of a many variable

under one or several constraints on the variables. In the above example there
are two constraints. We shall tackle this problem employing the Lagranges
method of undetermined multipliers.
To this we turn our attention, below.
5.5 Lagranges method of undetermined multipliers

Let me pose the problem through a simple example.
A mountain be described by h(x, y) where h is a function of the variable
x and y. h is the elevation of the mountain at a point (x, y) on the plane.
I want to find out (x , y ) at which h is maximum.
We write
dh =
h
h
dx +
dy = 0
x
y
(5.14)
If dx and dy are independent then dh = 0 if and only if

h
=0
x
h
=0
y
(5.15)
(5.16)
We have two equations and two unknowns. In principle we can solve the above
two equations and obtain (x , y ) at which h is maximum.
5.5 Lagranges method of undetermined multipliers
71
Now imagine there is a road on the mountain which does not necessarily
pass through the peak of the mountain. If you are travelling on the road, then
what is the highest point you will pass through ? In the equation
dh =
h
h
dx +
dy = 0
x
y
(5.17)
the infinitesimals dx and dy are not independent. You can choose only one
of them independently. The other is detemined by the constraint which says
that you have to be on the road.
Let the projection of the mountain-road on the plane be described by the
curve
g(x, y) = 0.
This gives us a constraint
g
g
dx +
dy = 0
x
y
(5.18)
From the above we get,

g
x
dy = dx
g
y
(5.19)
We then have,
dh =
h
h
dx +
dy = 0
x
y

g
x
(5.20)
h
h
dx +
dx
g
x
y
y

y
h
g
=
dx = 0

x g x
(5.21)
(5.22)
In the above dx is an arbitrary non-zero infinitesimal. Hence the above equality

holds good if and only if the terms inside the square bracket is zero. We have,
h
g
=0
x
x
(5.23)
72
where we have set,

=
h
y
g
y
(5.24)
We have a similar equation involving the partial derivative with respect to

the varible y, which follows from the definition of the Lagrange undetermined
multipler .
Thus we have two independent equations
g
h
=0
x
x
(5.25)
h
g
=0
y
y
(5.26)
We can solve and and get x x () and y = y (). The value of x and y at
which h(x, y) is maximum under constraint g(x, y) = 0 can be found in terms
of the unknown Lagrange multiplier .
Of course we can determine the value of by substituting the solution
(x (), y ()) in the constraint equation : g(x (), y ()) = 0.
5.6 Generalisation to a function of N variables

Let f (x1 , x2 , xN ) be a function of N variables. The aim is to maximize f
under one constraint g(x1 , x2 , , xN ) = 0.
We start with
df =
N
X
f
dxi = 0
xi
i=1
(5.27)
for maximum. In the set {dx1 , dx2 , dx , dxN }, not all are independent.
They are related by the constraint
N
X
g
dxi = 0
x
i
i=1
(5.28)
We pick up one of the variable, say x and write

dx =
g
xi
g
i=1,i6= x
N
X
dxi
Substitute the above in the expression for df . We get,
(5.29)
5.6 Generalisation to a function of N variables

N
X
g
h
dxi = 0
xi
xi
73
(5.30)
i=1;i6=
where

=
h
x
g
x

(5.31)
There are only N 1 values of dxi . We have eliminated dx . Instead we

have the undetermined multiplier . Since dxi : i = 1, N and i 6= are all
independent of each other we can set each term in the sum to zero. Therefore
h
g
= 0 i 6=
xi
xi
(5.32)
From the definition of we get

g
h
=0
x
x
(5.33)
Thus we have a set of N equations

h
g
= 0 i = 1, N
xi
xi
(5.34)
There are N equations and N unknowns. In principle we can solve the equation
and get
xi xi () i = 1, N,
where the function h is maximum under constraint
g(x1 , x2 , xN ) = 0.
The value of the undetermined multiplier can be obtained by substituting
the solution in the constraint equation.
If we have more than one constraints we introduce separate Lagrange
multipliers for each constraint. Let there be m N constraints. Let these
constraints be given by
gi (x1 , x2 , xN ) = 0 i = 1, m.
We introduce m number of Lagrange multipliers, i : i = 1, m and write
f
g1
g2
gm
1
2
m
= 0 i = 1, N
xi
xi
xi
xi
where the m N .
74
5.7 Derivation of Boltzmann weight

Let us return to our problem of finding Pi - the probability that a closed
equilibrium system (with macroscopic properties T, V, N ) will be found in
it micro state i with energy Ei . Employing the method of most probable
distribution, we have found that,
Pi =
ai
A
where A is the number of elements of the canonical ensemble and aj = aj (

a ).
b a) is maximum, under two constraints.
a
is that string for which (
The two constraints are
b a) =
(
X
j
X
j
A!
a1 !a2 !
aj (
a) = A
aj (
a)Ej = E
b a).
For convenience we extremize ln (
b 1 , a2 , ) = ln A!
ln (a
ln aj ln aj +
aj
We introduce two Lagrange multipliers and and write

b 1 , a2 , )
(a1 + a2 + A)
(a1 E1 + a2 E2 + E)
ln (a
=0
ai
ai
ai
Let aj denote the solution of the above equation. We get,

ln aj 1 Ej = 0 j = 1, 2,
The above can be written in a convenient form
aj = exp(Ej )
where = exp( 1).
We thus have,
Pj = exp(Ej )
where = /A.
5.8 Canonical partition function : Transform of density of states
75
Thus we get the probability that a closed system shall be found in its
micro state j in terms of the constants which can be expressed as a function
of the Lagrange multiplier and which is the Lagrange multiplier for the
constraint on the total energy of the isolated system.
The task now is to evaluate the constants and .
The
constant can be evaluated by imposing the normalization condition
P
: j Pj = 1. The closed system has to be in one of its micro state with unit
probability. Thus we have,
Pj =
Q(, V, N ) =
1
exp(Ej )
Q
X
exp(Ej )
We call Q the canonical partition function.

What is the nature of the Lagrange multiplier ? On physical ground we
identify
1
.
=
kB T
I shall refer you to
Donald A McQuairrie, Statistical Mechanics, Harper and Row (1976)pp.3544
for this.
I shall return to this issue of identifying a physical quantity for the Lagrange multiplier at a later time when I teach you a bit more of thermodynamics. For the present, assume that the Lagrange multiplier = 1/[kB T ].
5.8 Canonical partition function : Transform of density

of states
We start with
Q(, V, N ) =
exp[Ei (V, N )]
where, = 1/[kB T ], and the sum runs over all the micro states of a closed sysb
tem at temperature T , volume V and number of particles N . Let (E,
V, N )
denote the density of (energy) states. In other words
b
(E,
V, N )dE
is the number of micro states having energy between E and E + dE. The
canonical partition function can be written as an integral over energy,
76
Q(, V, N ) =
b
dE (E,
V, N ) exp [E(V, N )]
The canonical partition function is a transform of the density of states. The

variable energy is transformed to the variable temperature.
The density of states is a steeply increasing function of E. The exponential
function exp(E) decays with E for any fine value of . The decay is steeper
at higher value of or equivalently at lower temperatures. The product shall
be, in general, sharply peaked at a value of E determined by .
When is small (or temperature is large) the integrand would peak at a
large value of E. When is high (at low temperatures) it would peak at a
low value of E.
5.9 Canonical partition function and Helmholtz free

energy
The thermodynamic energy U of a closed system is the statistical energy E
averaged over a canonical ensemble. A closed system will invariably be found
with an energy U = hEi but for extremely small fluctuations around U ; these
fluctuations are proportional to the inverse of the square root of the number
of molecules.
Hence we can replace the integral over E by the value of the integrand,
evaluated at
E = hEi = U . We get,
b
Q = (E
= U ) exp(U )
b ) U
ln Q = ln (U
b
kB T ln Q = U T kB ln
= U TS
We identify the right hand side of the above as (Helmholtz) free energy :
F (T, V, N ) = U (T, V, N ) T S(T, V, N ).
Thus we get a relation between (the microscopic description enshrined in)
the canonical partition function (of statistical mechanics) and (the macroscopic description given in terms of) (Helmholtz) free energy (of thermodynamics) :
F (T, V, N ) = kB T ln Q(T, V, N )
5.10 Canonical ensemble and entropy
77
Statistical mechanics aims to connect the micro world (of say atoms and
molecules) to the macro world (of solids and liquids). In other words it helps
you calculate the macroscopic properties of a system say a solid, in terms of
the properties of its microscopic constituents (atoms and molecules) and their
interactions.
Boltzmann started the game of statistical mechanics by first proposing
a micro - macro connection for an isolated system, in the famous formula
engraved on his tomb:
b
S = kB ln
You will come across several micro-macro connection during this course on
statistical mechanics. The formula
F (T, V, N ) = kB T ln Q(T, V, N ),
provides a micro - macro connection for a closed system.
5.10 Canonical ensemble and entropy

Consider the expression
pi ln pi
in the context of canonical ensemble.

b 1 , a2 . ) which gives the number of
In an earlier lecture we talked of (a
micro states of an isolated system having a macroscopic property described by
{a1 , a2 , }. The isolated system is constructed as follows : Assemble a large
number of closed systems. Each closed system is in thermal contact with its
neighbouring closed systems. Let = {i = 1, 2, } denote the set of all micro
states of a closed system. Let A denote the number of closed systems in the
assembly. The entire assembly is isolated. We have thus an isolated system.
We describe a micro state of the isolated system by specifying the index i
of each closed system in the assembly. The index comes from the set defied
earlier. The micro states of the isolated system are all equally probable; we
group them and denote a group by specifying the string {a1 , a2 , } where ai
is the number of closed systems having the index i. We have
b 1 , a2 , ) =
(a
A!
a1 !a2 !
b For convenience
The aim is to find {ai : i = 1, 2, } that maximizes .
b Let us consider the limit ai i. Also consider the
we maximize ln .
variables pi = ai /A. Then
78
b = A ln A A
ln
=
X
i
ai ln A
= A
ai ln
ai ln ai +
ai
ai ln ai
a
i
pi ln pi
X
b
ln
pi ln pi
=
A
i
The above is the entropy of one of the
PA number of closed systems constituting
the the isolated assembly. Thus, i pi ln pi provides a natural formula for
the entropy of a system whose micro states are not equi-probable.
Physicists would prefer to measure entropy in units of Joules per Kelvin.
For, that is what they have learnt from Claussius, who defined
dS =
q
,
T
where dS is the entropy gained by a system when it receives an energy of

q Joules by heat in a reversible process at constant temperature, T Kelvin.
Hence we define,
X
pi ln pi .
S = kB
i
We call this Boltzmann-Gibbs-Shannon entropy. This expression for entropy

is natural for a closed system described by a canonical ensemble.
5.11 Free energy to entropy

We start with the thermodynamic relation
F (T, V, N ) = U T S
We recognize that F (T, V, N ) is the Legendre transform of the fundamental
eqution U (S, V, N ) where we transform the variable S in favour of T defined
as

U
T =
S V,N
5.11 Free energy to entropy
79
Read the chapter on Legendre transform in my notes on thermodynamics. If

required, I shall discuss these issues in one of the extra classes. Remind me
then.
We re-write the expression for F as,
S
= (F U )
kB
We make use of the following,

Q=
exp(Ei )
pi =
1
exp(Ei )
Q
F = kB T ln Q
U = hEi =
1 X
Ei exp(Ei )
Q i
and write,
#
"
S
1 X
Ei exp(Ei )
= kB T ln Q
kB
Q i
#
1 X
1 X
exp(Ei )
Ei exp(Ei )
= ln Q
Q i
Q i
"
X 1
i

exp(Ei )
exp(Ei ) ln
Q
Q
X 1
i

exp(Ei )
ln Q Ei
pi ln pi
from which we get the Boltzmann-Gibbs-Shannon entropy,

X
pi ln pi .
S = kB
i
80
5.12 Energy fluctuations and heat capacity

The average energy of a system is formally given by
X
Ei pi
hEi =
(5.35)
where pi is the probability of the micro state i and Ei is the energy of the
system when in micro state i. For a closed system
pi =
1
exp(Ei )
Q
(5.36)
where Q(T, V, N ) is the (canonical) partition function given by

X
exp(Ei ).
Q=
(5.37)
We have,
P
i Ei exp(Ei )
hEi = P
i exp(Ei )
=
(5.38)
1 X
Ei exp(Ei )
Q i
(5.39)
1 Q
Q
(5.40)
ln Q
(5.41)
We identify hEi with the internal energy, usually denoted by the symbol U in
thermodynamics.
We have,
U =

1 Q
Q
1 2Q
+
Q 2
(5.42)

1 Q
Q

= hE 2 i hEi2
2
= E
2
(5.43)
(5.44)
(5.45)
5.13 Canonical partition function for an ideal gas
81
Now write
U
T
U
=
= CV (kB T 2 )
(5.46)
(5.47)
We get the relation between the fluctuations of energy of an equilibrium system and the reversible heat required to raise the temperature of the system
by one degree Kelvin :
2
E
= kB T 2 CV .
(5.48)
The left hand side of the above equation represents the fluctuations of
energy when the system is in equilibrium. The right hand side is about how
the system would respond when you heat it8 . Note CV is the amount of
reversible heat you have to supply to the system at constant volume to raise
its temperature by one degree Kelvin. The equilibrium fluctuations in energy
are related to the linear response; i.e. the response of the system to small
perturbation9 .

I shall derive an expression for the canonical partition function of an ideal gas
of N molecules confined to a volume V and at temperature T . I shall do the
derivation by an easy method and an easier method.
5.13.1 Easy method:
Formally the partition function is given by,
Z +
Z +
VN 1
Q(T, V, N ) =
dp
dp2
1
N ! h3N

Z +
2
2
2
dp3N exp
p + p2 + p3N
2m 1
VN 1
N ! h3N
Z
3N

1 p2
dp exp
2 mkB T
Consider the integral

8
Notice that 2 is expressed in units of Joule2 . The quantity kB T 2 is expressed in

units of Joule-Kelvin. CV is in Joule/Kelvin. Thus kB T 2 CV has units of Joule2 .
first order perturbation.
82
I=

1 p2
dp exp
2 mkB T
since the integrand is an even function, we can write the above as

Z +
1 p2
dp exp
I=2
2 mkB T
0
Let
x=
p2
2mkB T
Therefore,
dx =
p
dp
mkB T
dp =
mkB T 1
dx
2 x1/2
The integral can now expressed as

Z
p
dx x1/2 exp(x)
I = 2mkB T
0
Z
p
2mkB T
dx x(1/2)1 exp(x)

p
1
= 2mkB T
2
=
p
2mkB T
since (1/2) =
The canonical partition function is thus given by,

=
VN 1
3N/2
(2mkB T )
N ! h3N
5.13.2 Easier method : Transform of density of (energy) states

We first derive an expression for the density of (energy) states, denoted by
g(E) from micro canonical ensemble. g(E)dE is the number of micro states
of an isolated system with energy between E and E + dE. Formally, we have
g(E) =
83
N
1 (2mE)3N/2
b= V
N ! h3N ( 3N
2 + 1)
Therefore the density of (energy) states is given by

g(E) =
V N 1 (2m)3N/2 3N 3N 1
=
E 2
E
N ! h3N ( 3N
2
2 + 1)
=
V N 1 (2m)3N/2 3N 1
E 2
N ! h3N ( 3N
2 )
where we have made use of the relation

3N
3N
3N
+1 =
2
2
2
The partition function is obtained as a transform of the density of states
where the variable E transformed to the variable .
Z
V N 1 (2m)3N/2
3N

Q(, V, N ) =
dE exp( E) E 2 1
3N
3N
N! h
2
0
Consider the integral
I=
dE exp(E)E
3N
2
Let,
x = E then dx = dE
Z
3N
1
I = 3N/2
dx x 2 1 exp(x)
0
=
( 3N
2 )
3N/2
Substituting the above in the expression for the partition function we get,
Q(T, V, N ) =
VN 1
(2mkB T )3N/2
N ! h3N
84
5.14 Microscopic interpretation of heat and work

The thermodynamic energy U is identified with statistical energy E averaged
over a suitable Gibbs ensemble of micro states. We use micro canonical ensemble for isolated system, canonical ensemble for closed system, and grand
canonical ensemble for open system. Let pi : i = 1, 2, denote formally an ensemble. pi is the probability of a micro state of the system under consideration.
For example if the system is isolated, then all micro states are equally
probable. We simply count the number of micro states of the system; let us
b micro states; then pi = 1/.
b
say there are
For a closed system, pi = (1/Q) exp(Ei ).
For an open system pi = (1/Q exp(Ei + Ni )
We can write, in general,
X
pi Ei
(5.49)
U =
i
The above equation suggests that the internal energy of a closed system can
be changed by two ways.
1. change {Ei : i = 1, 2, } keeping {pi : i = 1, 2, } the same. This
we call as work.
2. change {pi i = 1, 2, } keeping {Ei : iP= 1, 2, } the same. The changes
in pi should be done in such way that i pi = 1. This we call as heat.
Thus we have,
dU =
pi dEi +
Ei dpi
where the super script in the second sum should remind us that all dpi s are
not independent and that they should add up to zero.
In the first sum we change Ei by dEi i keeping pi i unchanged.
In the second
P sum we change pi by dpi i keeping Ei unchanged for all i
and ensuring i dpi = 0.
5.15 Work in statistical mechanics : W =

Show that dW =
We start with
X
pi dEi
pi dEi =
X
i
pi
pi dEi
P
( i pi Ei )
Ei
dV =
dV
V
V
U
hEi
dV =
dV = P dV = dW
V
V
(5.50)
5.16 Heat in statistical mechanics :
85

P
Show that dq = i Ei dpi
We start with
S = kB
dS = kB
pi ln pi
(5.51)
[1 + ln pi ] dpi
(5.52)
T dS = dq = kB T
= kB T
ln pi dpi
[Ei ln Q] dpi
Ei dpi
(5.53)
Problems
5.1. Find the probability of the micro states of the red die under the condition
that the three die add to 12.
5.2. Maximise A(x1 , x2 ) = x1 x2 under constraint x1 + x2 = 10.
We have
A
g
=0
x1
x1
g
A
=0
x2
x2
For the given problem we have,
x2 = 0
x1 = 0
The constraint x1 + x2 = 10 gives = 5. Thus for x1 = x2 = 5 the function
f is maximum under constraint x1 + x2 = 10. For a fixed perimeter, the area
of a rectangle is maximum only if the sides are same.
5.3. Maximise x3 y 5 under the constraint x + y = 8. Answer: x = 3; y = 5.
86
5.4. Let (x, y, x) be a point on the surface of a sphere x2 + y 2 + z 2 = 1. Let

(2, 1, 2) be a point. Let D(x, y, z) denote the distance between the point
(x, y, z) on the sphere and the point (2, 1, 2). Employing Lagranges method
of undetermined multiplier, find the maximum and minimum value of D.
Answer: 4 and 2
5.5. Practice Problems employing the method of Lagrange Undetermined Multipliers
P
5.6. Start with S = kB i pi ln pi , where the sum runs over all the micro
states of the closed system. Treat S as a function
variables p1 , p2 , .
P of the P
There are two constraints on the variables : i pi = 1; i pi Ei = hEi = U,
where U is constant. Employing Lagrange method of undetermined multipliers, maximize the entropy S. Show that {pi : i = 1, 2, } that maximize S
are given by
exp(Ei )
pi = P
i exp(Ei )
where is the Lagrange multiplier associated with the second constraint.

5.7. A rectangle is inscribed in an ellipse whose major and minor axes are
of lengths a and b respectively. The major axis is along the X axis and the
minor axis is along the Y axis. The centre of the ellipse is at the origin. The
centre of the inscribed rectangle is also at origin. Find the length and breath
of the rectangle with largest possible area. Employ the method of Lagrange
multiplier. If you have a circle of radius R with centre at origin instead of
ellipse what would be the length and breath of the inscribed rectangle with
largest possible area ?
5.8. Consider a right circular cylinder of volume 2 cubic meter. Employing
the method of Lagrange multiplier find out what height and radius will provide
the minimum total surface area for the cylinder ?
5.9. Let the sample space associated with the experiment of throwing a dice
be denoted by
= {1 , 2 , 3 , 4 , 5 , 6 }.
There are six outcomes. Consider throwing of N fair and independent dice. A
string of s denote an outcome of this experiment. The length of the string
is N . The number of outcomes is 6N . Let ni denote the number of times the
outcome i occurs. n
= (n1 , n2 , n3 , n4 , n5 , n6 ) denote a macro state of the
system. The number of outcomes associated with a given macro state n
is
given by multinomial distribution, see below.
b 1 , n2 , n3 , n4 , n5 , n6 ) = Q N !
(n
6
i=1
ni !
87
b n) is maximum. Employ Lagrange method of undeterFind n

for which (
mined multiplier. The constraint is
6
X
ni = N
i=1
5.10. Consider energy levels En = n, where n = 0, 1, 2, . The degeneracy

of the n-th energy level is n + 1. Noninteracting distinguishable particles are
distributed over the energy levels. The system is in equilibrium at temperature
T . Calculate
1. the canonical partition function
2. the thermodynamic energy defined as U = hEi
3. the entropy S.
We have,
X
S
pi ln pi
=
kB
i
In the above the right hand side can be interpreted as hln pi, where the angular
bracket denotes an average over {pi : i = 1, 2, }. Consider a closed system
for which pi = exp(Ei )/Q. Show that S/kB = U ln Q. From this
deduce that F = kB T ln Q.
5.11. See S. B. Cahn, G. D. Mahan, and B. E. Nadgorny, A Guide to Physics
Problems Part 2: Thermodynamics, Statistical Physics, and Quantum Mechanics, Plenum ((1997) problem No. 4.45 page 24
Consider a system composed of a very large number N of distinguishable particles at rest. The particles do not interact with each other. Each particle has
only two non-degenerate energy levels: 0 and > 0. Let E denote the total
energy of the system. Note that E is a random variable; it varies, in general,
from one micro state of the system to another. Let = E/N denote energy
per particle.
1. Assume that the system is not necessarily in thermal equilibrium. What
is the maximum possible value of ?
2. Let the system be in thermal equilibrium at temperature T . The canonical
ensemble average of E is the the thermodynamic energy, denoted by U . i.e.
U = hEi, where hi denote an average over a canonical ensemble of micro
states 10 . Let = U/N denote the (thermodynamic, equilibrium) energy
per particle. Derive an expression for as a function of temperature.
10
Note that it is meaningful to call U as thermodynamic energy only when the

average of energy is calculated for N ; only in this limit the average energy
will be unchanging with time. Fluctuations around the average value, defined as
the standard deviation (i.e. square-rootof the variance) of energy divided by the
mean energy will be of the order of 1/ N ; this goes to zero only in the limit of
N .
88
3. Find the value of in the limit T 0 and in the limit T .

4. What is the maximum possible value that can take ?
5.12. Consider a system of N distinguishable non-interacting particles each of
which can be in states designated as 1 and 2. Energy of state 1 is 1 = and
that of state 2 is 2 = +. Let the number of particles in states 1 and 2 be N1
and N2 respectively. We have N = N1 +N2 and E = N1 1 +N2 2 = (2N2 N ).
(i) Evaluate canonical partition function Q(T, V, N ). Do not forget the deb which gives the number of ways we can organize N1
generacy factor,
particles in state 1 and N2 particles in state 2.
(ii) Let q(T, V ) be the single-particle partition function. How Q(T, V, N ) and
q(t, V ) are related ?
(iii) Calculate and sketch heat capacity CV of the system.
5.13. Consider a system of two non-interacting particles in thermal equilibrium at temperature T = 1/[kB ]. Each of the particles can occupy any of
the three quantum states. The energies of the quantum states are
, 0 and + .
Obtain canonical partition function of the system for particles obeying
(i) classical statistics and are distinguishable
(ii) Maxwell-Boltzmann statistics and are indistinguishable .
For each of the above two cases calculate average energy of the system.
5.14. A zipper has N links. Each link can be in any one of the two states
(a) a closed state with zero energy
(b) an open state with energy > 0.
The zipper can be unzipped from top to bottom. A link can be open if and
only if all the links above it are also open. In other words, if we number the
links as 1, 2, , N from top to bottom, then link k can be open if and only
if all the links from 1 to k 1 are also open.
(i) Derive an expression for a canonical partition function
(ii) Let n denote the number of open links. Derive an expression for the average number, hni of open links. Employ canonical ensemble for carrying
out the averaging process.
(iii) Show that at low temperatures (kB T << 1), the average hni is independent of N .
5.15. The expression for the average energy which corresponds to thermodynamic energy is given by
hEi = U =
ln Q
89
Show that for an ideal gas the energy is given by

hEi = U = 3N
kB T
2
consistent with equi-partition theorem which says that each degree of freedom
(each quadratic term in the Hamiltonian) carries an energy of kB T /2.
5.16. Helmholtz free energy and the canonical partition function are related
:
F (T, V, N ) = kB T ln Q(T, V, N )
Derive an expression for the free energy of an ideal gas of N molecules confined
to a volume V at temperature T .
5.17. From thermodynamics we know that
dF = P dV SdT + dN.
Consider the expression for the free energy of an ideal gas, see last problem.
Take the partial derivative of F (T, V, N ), with respect to V and show that it
leads to ideal gas law P V = N kB T .
5.18. R K Pathria, Statistical Mechanics Second Edition Butterworth and
Heinemann (1996) p. 87; Problem 3.32
The quantum states available to a given physical system are
(i) a group of g1 equally likely states with a common energy 1 .
(ii) a group of g2 equally likely states with a common energy 2 =
6 1 .
Show that the entropy of the system is given by,

p2
p1
+ p2 ln
S = kB p1 ln
g1
g2
(5.54)
where p1 and p2 are, respectively, the probabilities of the system being in a

state belonging to group 1 or to group 2. Also p1 + p2 = 1. Assuming that p1
and p2 are given by canonical distribution show that,

g2
exp(x) +
(5.55)
S = kB ln g1 + ln 1 +
g1

x

(5.56)
g1
1+
exp(x)
g2
where x = (2 1 ) assumed positive.
90
5.19. Adiabatic process and Canonical ensemble

A M Glazer and J S Wark, Statistical Mechanics : A survival guide Oxford
(2001) sec. 4.4; pages : 52-53
We have earlier discussed the canonical partition function of an ideal gas.
We have
Q(T, V, N ) =
1
h3N
VN
3N/2
(2mkB T )
N!
(5.57)
From the partition function derive an expression for the entropy. The expression for entropy is called the Sackur-Tetrode equation. Consider a process in
which there is no change in entropy : S = 0. The volume and temperature
of the gas changes during such an isentropic (iso-entropic process; constant
entropy process etc.) process. Show that T V 2/3 is a constant during such a
process. Combined with the equation of state for an ideal gas : P V is a constant at constant N and T , this gives the formula for an adiabatic process :
P V 5/3 is a constant.
5.20. Adiabatic Process : A microscopic view
A M Glazer and J S Wark, Statistical Mechanics : A survival guide Oxford
(2001) sec. 4.5; pages : 53-56
Change in internal energy,
X
pi i ,
U=
i
can be brought about by two processes :

1. Work :
w=
pi di
change {i : i = 1, 2, } keeping {pi : i = 1, 2, } same

2. Heat :
X
i dpi
q=
i
Change
{pi : i = 1, 2, } keeping {i : i = 1, 2, } the same. Also
P
p
=
1.
i
i
Adiabatic process is one in which energy transacted by heat is zero. Consider

a particle of mass m in a cube of side L. The energy eigenvalues are given by
n1 ,n2 ,n3 =
h2
(n2 + n22 + n23 )
8mL2 1
ni = 1, 2, 3, for i = 1, 2, 3
We can change L keeping {pi : i = 1, 2, } the same. From these considerations show that for an adiabatic process P V 5/3 is a constant for a mono
atomic ideal gas.
6
Grand canonical ensemble
An open system is one which exchanges energy and matter with its surroundings. The surroundings act as a heat bath as well as a particle (or material)
bath.
A heat bath transacts energy with the system. The temperature of the
heat bath does not change because of the transaction of energy.
A material (or particle) bath transacts matter (or particles) with the system. The chemical potential of the material bath does not change because
of the transaction matter.
The system is in thermal equilibrium with the surroundings1 . The system
is also in diffusional equilibrium with the surroundings2 .
The system can be described by its temperature3 , T , volume, V and chemical potential4 , . Notice that the temperature, T , chemical potential, , and
volume V are independent properties of an open system.
Since the system is not isolated, its micro states are not equi-probable.
Aim : To calculate the probability of a micro state of the open system.
Let us take the open system, its boundary and surroundings and construct
an isolated system. We are interested in constructing an isolated system because, we want to start with the only assumption we make in statistical mechanics : all the micro states of an isolated system are equally probable. We
have called this the ergodic hypothesis.
1
2
equality of temperature signals thermal equilibrium.

equality of chemical potential ensures diffusional or material equilibrium. Of
course, equality of pressure shows mechanical equilibrium.

U
T =
S V,N
U
N
S,V
92
6 Grand canonical ensemble
Let E denote the total energy of the isolated system.

E
>> E,
where E is the (average or typical) energy of the open system.

N >> N,
where N is the (average or typical) number of particles in the open system. V
is the total volume of the isolated system. An isolated system is characterized
by E, N , and V and are held at constant values.
Our aim is to describe the open system in terms of its own micro states.
Let c be a micro state of the open system. Let E(c) be the energy of the open
system when it is in its micro state c. Let N (c) be the number of particles in
the open system when it is in its micro state c.
When the open system is in its micro state c the surrounding can be in
any one of the innumerable micro states (of the isolated system)5 such that
the energy of the surroundings is E E(c) and the number of particles in the
surroundings is N N (c).
Let
b E(c), V, N N (c)),
(E
denote the number of micro states of the isolated system such that the energy
of the surrounding is E E(c) and number of particles in the surroundings
is N N (c) and the open system is in its micro state c. The volume of the
surroundings is a constant V V .
Following Boltzmann we define a statistical entropy as
b E(c), V V, N N (c)).
S(E E(c), V V, N N (c)) = kB ln (E
(6.1)
Since
E(c) << E,
and
N (c) << N ,
we can Taylor-expand S retaining only the first two terms. We have
5
The picture I have is the following. I am visualizing a micro state of the isolated
system as consisting of two parts. One part holds the signature of the open system;
the other holds the signature of the surroundings. For example a string of positions
and momenta of all the particles in the isolated system defines a micro state. This
string consists of two parts. The first part contains the string of positions and
momenta of all the particle in the open system and the second part contains the
positions and momenta of all the particles in the surroundings. Since the system
is open the length system-string is a fluctuating quantity and so is the length
of bath-string. However the string of the isolated system is of fixed length. I am
neglecting those micro states of the isolated system which hold the signature of
the interaction between the system and the surroundings at the boundaries.
93

S E E(c), V V, N N (c) = S E, V, N
E(c)
N (c)
S
E
(6.2)
S
N
(6.3)
V,N E,N
(6.4)
E,V E,N
From the first law of thermodynamics we have, for a reversible process,

dE = T ds P dV + dN
dS =
(6.5)
1
P
dE + dV dN
T
T
T
(6.6)
We have
S S(E, V, N )
dS =
S
E
(6.7)
dE +
V,N
S
V
dV +
E,N
S
N
dN
Comparing the coefficients of dE, dV and dN , we get,

S
1
=
E V,N
T

S
V
S
N
E,N
E,V
(6.8)
E,V
P
T
(6.9)
(6.10)
(6.11)
Therefore,

1
S E E(c), N N (c) = S E, N E(c) + N (c)

T
T
The probability of the micro state c is given by

b E E(c), V, N N (c)
P (c) =
bTotal
(6.12)
(6.13)
94
We are able to write the above because of the postulate of ergodicity : All
micro states of an isolated system are equally probable. We have,

1
1
exp
S E E(c), N N (c)
(6.14)
P (c) =
bTotal
kB
1
bTotal
S(E, N ) E(c) N (c)
+
exp
kB
kB T
kB T
= exp[ {(E(c) N (c)}]
(6.15)
(6.16)
where the constant can be determined by the normalization condition,

X
P (c) = 1,
c
where the sum runs over all the micro states of the open system. We have,
P (c) =
1
exp[{E(c) N (c)}]
Q
where the grand canonical partition function is given by

X
Q(T, V, ) =
exp [ {E(c) N (c)}]
(6.17)
(6.18)
Let = ; then we can write the grand canonical partition function as,
X
Q(T, V, ) =
N (c) exp[E(c)]
(6.19)
c
Collect those micro states of a grand canonical ensemble with a fixed value
of N . Then these micro states constitute a canonical ensemble described the
canonical partition function, Q(T, V, N ), see also the footnote6 . Thus we can
write the grand canonical partition function as,
X
Q(T, V, ) =
N Q(T, V, N )
(6.20)
N
Grand canonical ensemble is an extremely useful ensemble. The reason

is that the constraint of constant N required for calculating the canonical
ensemble is often mathematically awkward7.
6
We can further collect all those micro states of the canonical ensemble with a
fixed energy. Then these micro states constitute a micro canonical ensemble.
We shall experience this while trying to evaluate the canonical partition function
for Fermions and Bosons. We will not be able to carry out the sum over occupation
6.1 Grand canonical partition function and grand potential
95
6.1 Grand canonical partition function and grand

potential
Entropy or energy is the thermodynamic counter part of the statistical
mechanical micro canonical ensemble. We have
b
S(E, V, N ) = kB ln (E,
V, N ).
Helmholtz free energy is the thermodynamic counter part of the statistical

mechanical canonical ensemble. We have
F (T, V, N ) = kB T ln Q(T, V, N ).
What is the thermodynamic counter part of the grand canonical ensemble
?
Let us call it, the grand potential and denote it by the symbol G. It is a
function of temperature, volume and chemical potential. G(T, V, ), obtained
from U (S, V, N ) by Legendre transform of S T and N . We have,
G(T, V, ) = U (S, V, N ) T S N
T =
U
S
U
N
(6.21)
(6.22)
V,N
(6.23)
S,V
Some authors e.g. Donald A McQuairrie, Statistical Mechanics, Harper

and Row (1976) would like to identify P V as the thermodynamic counter
part of the grand canonical ensemble. The correspondence is
G = P V = kB T ln Q.
First let us establish the above relation between the grand potential G and
grand canonical partition function Q.
numbers because of the constraint that they add up to a constant N . Hence we

shall multiply the restricted sum by N and sum over all possible values of N .
This would remove the restriction and we shall express the partition function as
sum over (micro states) of product (over occupation numbers). We shall interpret
as fugacity. The chemical potential and fugacity are related : = exp().
All these can be viewed as mathematical tricks. The language of grand canonical
ensemble gives a physical meaning to these mathematical tricks.
96
G = kB T ln Q
We follow the same method we employed for establishing the connection between Helmholtz free energy and canonical partition function. We have,
X
exp(Ei Ni )
(6.24)
Q(T, V, ) =
i
where the sum runs over all the microstates i of the open system. Ei is the
energy of the micro state i, and Ni is the number of particles in the system
when it is in its micro state i.
We replace the sum over micro states by sum over energy and number of
particles. Let g(E, N ) denote the density of states. We have then,
Z
Z
Q(T, V, ) = dE dN g(E, N ) exp[(E N )]
(6.25)
The contribution to the integrals come overwhelmingly from a single term at
hEi, and hN i. We then get,
Q(T, V, ) = g(hEi, hN i) exp[(hEi hN i)]
(6.26)
Let us denote hEi by E and hN i by N . Taking logarithm on both sides,

ln Q = ln g(E, N ) (E N )
kB T ln Q = T [kB ln g(E, N )] + E N
= E T S N
=G
(6.27)
(6.28)
(6.29)
(6.30)
Alternately, we start with

Q=
N Q(T, V, N )
(6.31)
In principle, the number of particles in an equilibrium open system is not

a fixed number. It fluctuates from one micro state to another. However the
fluctuations are very small; it can be shown that the relative fluctuations are
inversely proportional to the size of the system. In the above expression for Q,
only one term contributes overwhelmingly to the sum over N . Let the value
of N for which the N Q(T, V, N ) is maximum be N . Hence the sum over
N can be replaced by a single entry with N = N .

Q(T, V, ) = N Q(T, V, N )
ln Q(T, V, ) = N + ln Q(T, V, N ) =
N
+ ln Q(T, V, N )
kB T
(6.32)
(6.33)
6.2 Euler formula in the context of homogeneous function
kB T ln Q = N + kB T ln Q(T, V, N )
97
(6.34)
Refer to notes on Canonical ensemble. We have shown that

F (T, V, N ) = kB T ln Q(T, V, N ).
Therefore we can write the above equation as,
kB T ln Q = N F (T, V, N )
(6.35)
kB T lQ = F N
(6.36)
= U T S N
(6.37)
=G
(6.38)
Recall the discussions on Legendre Transform. We start with U U (S, V, N ).

Transform S in favour of the slope T (partial derivative of U with respect to
S). We get the intercept F (T, V, N ) as U T S. Let us do one more transform : N . The partial derivative of U with respect to N is the chemical
potential . We get the intercept G(T, V, ) - the grand potential. We have
G(T, V, ) = U T S N . Thus we have,
G(T, V, ) = kB T ln Q(T, V, )
(6.39)
Our next task is to show that G(T, V, ) = P V . To this end, let me tell
you of a beautiful formula proposed by Euler, in the context of homogeneous
function.
6.2 Euler formula in the context of homogeneous

function
U is a homogeneous function of S, V and N . U is an extensive property; so are
S, V and N . In the words of mathematicians, U is a first order homogeneous
function of S, V , and N . This means,
U (S, V, N ) = U (S, V, N )
(6.40)
where is a constant. Eulers trick consists of differentiating both sides of the

above equation with respect to . We get,
U S
U V
U N
+
+
= U (S, V, N )
(S)
(V )
(N )
(6.41)
U
U
U
+V
+N
= U (S, V, N )
(S)
(V )
(N )
(6.42)
98
The above is true for any value of . In particular it is true for = 1.

Substitute in the above = 1 and get,
S
U
U
U
+V
+N
= U (S, V, N )
S
V
N
T S P V + N = U
(6.43)
(6.44)
6.3 P V = kB T ln Q
We proceed as follows. From Eq. (??) we have
P V = U T S N
(6.45)
The RHS of the above is grand potential. Hence,

P V = G(T, V, )
(6.46)
= kB T ln Q(T, V, )
P V = kB T ln Q(T, V, )
(6.47)
(6.48)
6.4 Gibbs-Duhem relation : d = sdT + vdP

Now that we are on the Eulers formula, let me digress a little bit and see if
the equations we have derived, can be used to establish a relation amongst
the intensive properties T , P and of the system.
Derivation from U (S, V, N )
To this end we proceed as follows.
U = T S P V + N
(6.49)
dU = T dS P dV + dN + SdT V dP + N d
(6.50)
From the first law of thermodynamics, we have dU = T dS P dV + dN .

Hence,
N d + SdT V dP = 0
d =
(6.51)
V
S
dT +
dP
N
N
= sdT + vdP
(6.52)
(6.53)
6.5 Grand canonical ensemble : Number fluctuations
99
where s is the specific entropy - entropy per particle and v is specific volume
- volume per particle.
6.5 Grand canonical ensemble : Number fluctuations

In an open system, the energy E and the number of molecules N are random
variables. Energy fluctuates when the system goes from micro state to another.
The number of molecules fluctuates from one micro state to the other. Let us
now derive an expression for the variance of N :
2 = hN 2 i hN i2 .
To this end, we start with
Q(T, V, ) =
X
c

exp {E(c) N (c)}
(6.54)
In the above
c denotes a micro state of the open system
E(c) denotes the energy of the open system when in micro state c
N (c) denotes the number of particles of the open when in micro state c
Let us now take the partial derivative of all the terms in the above equation,
with respect to the variable , keeping the temperature and volume constant.
We have,

X
Q
=
N (c) exp {E(c) N (c)} = hN iQ(T, V, (6.55)
)
T,V
c
"

2

#
hN i
Q
Q
(6.56)
+Q
= hN i
2 T,V
T,V
T,V
The left hand side of the above equation equals
Substituting this in the above, we get,
2 hN 2 iQ.

Q
2

N (c) exp {E(c) N (c)}

[N (c)] exp {E(c) N (c)}
T,V
T,V
= 2 hN 2 iQ.
100
2 hN 2 iQ = 2 hN i2 Q + Q
= hN i hN i = kB T
hN i
hN i
(6.57)
T,V
(6.58)
T,V
In the above, we have an example of fluctuation-dissipation theorem. We have

the Gibbs free energy. G(T, P, N ) = N .
6.6 Number fluctuations and isothermal compressibility

Let us express the number fluctuations in terms of experimentally measurable
properties 9 of the open system. Let us define
v=
V
.
hN i
It is called specific volume. It is the volume per particle. We have,

hN i
(V /v)
=
T,V
T,V

hN i2
=
V
V
v2
(6.59)
(6.60)
T,V
(6.61)
T,V
In the above we can express,

9
We shall show that,

hN i
=
T,V
hN i2
kT
V
where kT denotes isothermal compressibility - an experimentally measurable property. Isothermal compressibility is defined as

1 V
kT =
V P T
6.6 Number fluctuations and isothermal compressibility
=
T,V
v
P
T,V
101
(6.62)
T,V
Employing Gibbs-Duhem relation, we find 10 ,

hN i
P
=
T
V
(6.64)
Therefore,

T,V
hN i
=
V
v
P
T,V
1
=
v
v
P
T,V
= kT
(6.65)
Finally we get,

hN i
T,V
hN i2
kT
V
= kB T
= kB T
hN i
hN i2
kT
V
2
kB T
=
kT
2
hN i
V
(6.66)

(6.67)
T,V
(6.68)
(6.69)
Thus the fluctuations in the number of molecules of an open system is directly

proportional to the isothermal compressibility. Hence we expect isothermal
compressibility to be positive 11 .
The number fluctuations are small in the thermodynamic limit; they are
of the order of inverse of the square-root of the number of particles in the
system.
Thus equilibrium fluctuations are related to an appropriate susceptibility
which measures the response of the system to a small external perturbation. When heated, the system responds by raising its temperature. Energy
10
Gibbs - Duhem relation reads as hN id = V dP SdT. At constant temperature,

we have,
hN id = V dP,

hN i
P
=
(6.63)
T
V
11
The relative fluctuations of energy in a canonical ensemble is proportional to heat

capacity at constant volume. Hence we expect heat capacity to be positive .
102
absorbed by heat divided by the increase in temperature is called the heat

capacity. The volume does not change during this process. We saw that the
equilibrium energy fluctuations are proportional heat capacity.
Thus, the susceptibility is heat capacity for energy fluctuations; it is
isothermal compressibility for the fluctuations of number density. These are
special cases of a more general principle called Fluctuation-Dissipation theorem enunciated by Albert Einstein in the context of Brownian motion. If time
permits, I shall speak on Brownian motion in one of the extra classes.
However, close to first order phase transition, the isothermal compressibility diverges. The fluctuations in the number of particles are large; they are of
the order of the system size. The pressure-volume phase diagram has a flat
region very near first order phase transition temperature, see discussions on
van der Waal gas and Maxwells construction.
6.7 Alternate derivation of the relation :

2
N
/hN i2 = kB T kT /V
In my last lecture I had derived a relation is between the number fluctuation
and isothermal compressibility. I had followed closely R K Pathria, Statistical
Mechanics, Second Edition, Butterworth and Henemann (1996), I told you
that I was not very excited about Pathrias derivation and that I would try
and produce a better one; or at least an alternate derivation. Here it is.
Start with a fluctuation-dissipation relation derived in the last lecture12 ,

hN i
2
= kB T
(6.70)
V,T
We consider the reciprocal,

hN i
(6.71)
V,T
We observe that is a function13 of T and P . We are keeping T a constant.

Hence can change only when P changes.
Gibbs-Duhem relation tells us,
hN id = V dP SdT.
(6.72)
When T is a constant, we have dT = 0. This gives us

12
13
relating number fluctuations to response to small changes in the chemical potential.

Gibbs and Duhem told us that the three intensive properties T , P , and are not
all independent. Only two of them are independent. The third is automatically
fixed by the other two.
2
/hN i2 = kB T kT /V
V
dP
hN i
d =

hN i
V
hN i
T,V
(6.73)
P
hN i
(6.74)
T,V
V
hN i
T,V
103
(6.75)
Let us now define
hN i
,
V
which denotes the particle density : number of particles per unit volume. We
have,

P
P
=
(6.76)
hN i V,T
(V ) V,T
=
1
V
(6.77)
V,T
The density can be changed either by changing hN i and/or V . Here we change

infinitesimally, keeping V constant. As a result the pressure changes infinitesimally. Let me repeat : both these changes happen at constant V and
T.
As far as P is concerned, it does not care whether has changed by change
of hN i or of V . Hence it is legitimate to write

P
P
=
(6.78)
V,T
(hN i/V ) hN i,T
V2
=
hN i
Thus we get,
hN i
V,T
P
V
V
hN i
P
hN i
1
hN i
V2
hN i2
(6.79)
hN i,T
(6.80)
V,T
(6.81)
V,T
P
V
(6.82)
hN i,T
104
Take a reciprocal of the above and get,

hN i2 V
hN i
= 2
V,T
V
P hN i,T
(6.83)
Then we get,

hN i2
V
2
N
= kB T 2
V
P T,hN i
= kB T
hN i2
kT
V
(6.84)
(6.85)
Problems
6.1. Start with Helmholtz free energy : F F (T, V, N ). F is an extensive
thermodynamic variable. F is a first order homogeneous function of the extensive thermodynamic variables V, N . Note that F also depends on the intensive variable T . Therefore we can write, F (T, V N ) = F (T, V, N ). Employ
Eulers formula and derive Gibbs-Duhem relation.
6.2. Start with enthalpy: H H(S, P, N ). H is an extensive thermodynamic
variable. H is a first order homogeneous function of the extensive thermodynamic variables S, N . Note that H also depends on the intensive variable P .
Therefore we can write H(S, P, N ) = H(S, P, N ). Employ Eulers formula
and show that H = T S + N . Also derive Gibbs-Duhem relation.
6.3. Start with Gibbs free energy G G(T, P, N ). Employ Eulers formula
and show that G = N. Derive Gibbs-Duhem relation.
6.4. Let f be a second order homogeneous function of the variable x1 , x2 ,
xN . By this we mean
f (x1 , x2 , xN ) = 2 f (x1 , x2 , xN ).
Show that
2 f (x1 , x2 . xN ) =
N
X
xi
i=1
f
xi
6.5. Let f be an n-th order homogeneous function of the variables x1 , x2 , xN .

:
f (x1 , x2 , , xN ) = n f (x1 , x2 , xN ).
Derive the Euler relation
nf (x1 , x2 , xN ) =
N
X
i=1
xi
f
xi
2
/hN i2 = kB T kT /V
105
6.6. For a grand canonical ensemble show that,

ln Q
S = kB ln Q + kB T
T
V,

ln Q
N = kB T
V,T

lnQ
P = kB T
V
,T
= kB T
1
ln Q
V
6.7. Start with the Boltzmann-Gibbs-Shannon entropy

X
pi ln pi ,
S(p1 , p2 , ) = kB
i
where pj is the probability of the micro state i of the open system.

Show that the {pj : j = 1, 2, } for which the entropy is maximum
under the three constraints,
X
pi = 1;
i
X
i
X
i
pi i = hEi = U ;
pi Ni = hN i
are given by,

pi =
1
exp[Ei Ni )]
Q
where the and are the Lagrange multipliers for the second and third
constraints respectively. The first constraint gives rise to the grand canonical
partition function Q.
The energy of the system when in micro state i is Ei ; the number of
particles in the system when in micro state i is Ni .
6.8. Start with
Q(, V, ) =
X
c
exp [ {E(c) N (c)}]
where the sum runs over all the micro states of an open system. The average
number of particles is given by
hN i =
1 X
N (c) exp [ {E(c) N (c)}]
Q c
106
Show that the variance of N is given by

2
= hN i hN i = kB T
hN i
T,V
6.9. Donald A McQuarrie, Statistical Mechanics, Harper and Row (1976) page
: 65; Problem : 3-4.
Show that the pressure of an open system, in a canonical ensemble, is given
by
P = kB T
ln Q
V
(6.86)
,V
Use Eulers theorem for first order homogeneous functions and show that
P =
kB T
ln Q(T, V, ).
V
6.10. R K Pathria, Statistical Mechanics, Second Edition, Butterworth and

Henemann (1996) page : 102; Problem : 4.4
The probability that an open system, has N particles is given by,
P (N ) =
exp(N )Q(T, V, N )
Q(T, V, )
Verify the above statement. Consider a system of classical, indistinguishable,

ideal gas. Show that N has a Poisson distribution. Calculate the variance of
N employing the general formula

hN i
2 = kB T
T,V
Show that the result agrees with the one obtained from the Poisson distribution.
6.11. Donald A McQuarrie, Statistical Mechanics, Harper and Row (1976)
page : 67; Problem : 3-22.
Show that the fluctuations of energy in a grand canonical ensemble is

hEi
2
E
= kB T 2 CV +
2
hN i T,V N
2
/hN i2 = kB T kT /V
107
6.12. Donald A McQuarrie, Statistical Mechanics, Harper and Row (1976)

page : 67; Problem : 3-26.
Show that

hN i
V,T
V2
=
hN i2
P
V
hN i,T
Hint : Use Gibbs-Duhem relation :

hN id V dP + SdT = 0.
Use the chain rule:

T,V
hN i
T,V
hN i
P
T,V
Use the cyclic product rule of partial derivatives with variables hN i, P, V .

hN i
P
V
= 1
P T,V V T,hN i hN i T,P
7
Quantum Statistics
7.1 Occupation number representation of a micro state

Let
1, 2,
denote the single-particle quantum states. Let

{i : i = 1, 2, }
denote the corresponding energies. Notice that there can be several quantum
states with the same energy.
We have N non-interacting particles occupying the single particle quantum
states. A micro state of the macroscopic system of N particles in a volume V
is uniquely specified by a string of numbers
{n1 , n2 , },
where ni is the number of particles in the single- particle quantum state i.
Thus, a string of occupation numbers uniquely describes a micro state of
the quantum system. Note that such a string should obey the constraint :
n1 + n2 + = N.
The energy of a micro state (specified by a string of occupation numbers)
is
n1 1 + n2 2 + .
The canonical partition function can now be written as

Q(T, V, N ) =
{n1 ,n2 , }
exp [(n1 1 + n2 2 + )]
(7.1)
where the sum runs over all possible strings of occupation numbers (i.e. micro
states) obeying the constraint
110
7 Quantum Statistics
ni = N.
To remind us of this constraint I have put a star over the summation sign.
7.2 Open system and grand canonical partition function

The presence of the constraint renders evaluation of the sum a difficult task.
A way out is to transform the variable N in favour of . Do not confuse this
here with the one I had mentioned while discussing quantum or thermal or
de Broglie wave length. here is the fugacity and is related to the chemical
potential ; the relation is = exp(). The transform defines the grand
canonical partition function 1 see below.
Q(, V, ) =
N Q(, V, N )
(7.2)
N =0
This provides a description of an open system with its natural variable :

temperature, volume and chemical potential. The system can exchange energy
as well as matter with the environment. Thus energy E, and the number of
particles N , of the system would be fluctuating around the means hEi and
hN i respectively, where the angular bracket denotes averaging over a grand
canonical ensemble of the open system.
Thus we have,
Q(T, V, ) =
=
N =0
{n1 ,n2 , }
N =0 {n1 ,n2 , }
N =0 {n1 ,n2 , }
exp [(n1 1 + n2 2 + )]
(7.3)
n1 +n2 + exp [(n1 1 + n2 2 + )] (7.4)

exp [(1 )n1 (2 )n2 + )] (7.5)
we have already seen that a partition function is a transform. We start with

b
density of states (E)
and transform the variable E in favour of = 1/[kB T ] to
get canonical partition function:
Z
b
Q(, V, N ) =
dE (E,
V, N ) exp(E(V, N )).
0
I shall discuss formally the grand canonical ensemble in full glory in some later
lectures. For the present it is sufficient to consider the grand canonical partition
function as a transform of the canonical partition function with the variable N
transformed to fugacity or chemical potential = kB T ln().
7.2 Open system and grand canonical partition function
[exp { (1 )}]n1
xn1 1 xn2 2
N =0 {n1 ,n2 , }
[exp { (2 )}]n2
N =0 {n1 ,n2 , }
111
(7.6)
(7.7)
where we have introduced in the last line a short hand notation

xi = exp(i ) = exp[(i )].
(7.8)
We have a restricted sum over strings of occupation numbers. The restriction is that the occupation numbers constituting a string should add to N .
We then take a sum over N from 0 to . which removes the restriction. To
appreciate this, see Donald A McQuairrie, Statistical Mechanics, Harper and
Row (1976)p.77;Problem:4-6. I have worked out this problem below.
Consider first the sum over restricted sums :
I1 =
X
X
xn1 1 xn2 2
(7.9)
N =0 {n1 ,n2 }
where the star over the summation sign reminds us of the restriction n1 +n2 =
N . Also let us assume ni can be 0, 1 or 2, for i = 1, 2. Thus we have We can
write down I1 as,
I1 = 1 + x1 + x2 + x21 + x22 + x1 x2 + x21 x2 + x1 x22 + x21 x22
(7.10)
Now consider the product over the unrestricted sums.

I2 =
2 X
2
Y
xni i
(7.11)
(1 + xi + x2i )
(7.12)
i=1 ni =0
2
Y
i=1
= (1 + x1 + x21 ) (1 + x2 + x22 )
= 1 + x2 + x22 + x1 + x1 x2 + x1 x22 + x21 + x21 x2 + x21 x22
(7.13)
(7.14)
= I1
(7.15)
We can now write the grand canonical partition function of N quantum particles occupying single-particle quantum states determined by volume V .
Q(T, V, ) =
all
YX
xni i =
i ni =0
all
YX
[ exp(i )]
ni
(7.16)
i ni =0
all
YX

i ni =0
exp { (i )}
ni
(7.17)
112
1 n2
xn
1 x2
n1
n2
x1
x2
x21
x22
x1 x2
x21 x2
x1 x22
x21 x22
Table 7.1. Terms in the restricted sum
7.3 Fermi-Dirac Statistics

We have ni = 0, 1 i. This is a consequence of Pauli exclusion principle. No
single quantum state can have more than one Fermion.
QF D (T, V, ) =
Y
i

1 + exp[(i )
(7.18)
7.4 Bose-Einstein Statistics

We have ni = 0, 1, i. Let exp[(i )] < 1 i. This condition is
met if i > i. Then we have
7.6 Maxwell-Boltzmann Statistics
QBE =
Y
i
1
1 exp[(i )]
(i > i)
113
(7.19)
7.5 Classical Distinguishable Particles

If the particles are classical and distinguishable, then the string {n1 , n2 , }
will not uniquely specify a micro state. The string will have a set of classical
micro states associated with it. The number of micro states associated with
the string is given by,
b 1 , n2 , ) =
(n
We then have
QCS (T, V, ) =
N =0
N =0
{n1 ,n2 m }
"
X
i
N!
n1 !n2 !
(7.20)
N!
[exp(1 )]n1
n1 !n2 !
#N
exp(i )
[exp(2 )]n2
N [Q1 (T, V )]N
(7.21)
N =0
where Q1 (T, V ) is the single-particle partition function.

We have already seen that the partition function for classical distinguishable non-interacting point particles leads to an entropy which is not extensive. This is called Gibbs paradox. To take care of this Boltzmann asked
us to divide the number of micro states by N !, saying that the particles are
indistinguishable.
The non-extensitivity of entropy indicates a deep flaw in classical formulation of statistical mechanics. Classical particles do not simply exist in nature.
We have only quantum particles : Bosons and Fermions. It is quantum mechanics that would set the things right eventually. But quantum mechanics
had not arrived yet, in the times of Boltzmann.
Boltzmann corrected the (non-extensive entropy) flaw by introducing a
strange notion of indistinguishability of particles. Genius he was, Boltzmann
was almost on the mark. Division by N ! arises naturally in quantum mechanical formulation, because of the symmetry of the wave function
In the derivation of grand canonical partition function, we shall divide the
classical degeneracy by N !, as recommended by Boltzmann, see below, and
call the resulting statistics as Maxwell-Boltzmann statistics.
7.6 Maxwell-Boltzmann Statistics

Formally we have,
114
QMB =
=
N =0
{n1 ,n2 , }
N =0 {n1 ,n2 , }
N =0 {n1 ,n2 , }
1
exp[(n1 1 + n2 2 + )]
n1 !n2 !
[ exp(2 )]n2
[ exp(1 )]n1

n1 !
n2 !
[exp((1 )]n1
[exp((2 )]n2

n1 !
n2 !
xn1 1 xn2 2
n1 ! n2 !
N =0 {n1 ,n2 , }
!
!
X
X
xn2 2
xn1 1
=
n !
n !
n =0 2
n =0 1
= exp(x1 ) exp(x2 )
=
exp(xi )
exp[ exp(i )] =
Y
i
exp[exp{(i )}]
(7.22)
We can also express the grand canonical partition function for classical indistinguishable ideal gas as,
QMB = exp(x1 ) exp(x2 )
= exp
xi
X
exp[(i )]
= exp
X
= exp[
exp(i )]
= exp[Q1 (T, V )]
7.6.1 QM B (T, V, N ) QM B (T, V, )
We could have obtained the above in a simple manner, by recognizing that
QMB (T, V, N ) =
[QMB (T, V, N = 1)]N

N!
7.7 Grand canonical partition function, grand potential,and thermodynamic properties of an open system
Then we have,
=
N =0
[QMB (T, V, N = 1)]N

N!
= exp[QMB (T, V, N = 1)]
(7.23)
(7.24)
7.6.2 QM B (T, V, ) QM B (T, V, N )

We start with QMB (T, V, ) = exp[QMB (T, V, N = 1)]. Taylor expanding
the exponential,
QMB (T, V, ) =
N =0
QN
MB (T, V, N = 1)
N!
The coefficient of N is the canonical partition function2 and hence is given

by
QMB (T, V, N = 1)N
QMB (T, V, N ) =
N!
7.7 Grand canonical partition function, grand potential,

and thermodynamic properties of an open system
We can write the grand canonical partition function for the MB,BE, and FD
statistics as
exp[exp{(i )}] Maxwell Boltzmann
1
Y
BoseEinstein
Q=
(7.25)
1
exp
[(
i )]
1 + exp {(i )} Fermi Dirac
The connection to thermodynamics is established by the expression for grand

potential denoted by the symbol G(T, V, ). We have,
G(T, V, ) = kB T ln Q(T, V, )
2
Recall that
QM B (T, V, ) =
N=0
N QM B (T, V, N )
(7.26)
115
116
Recall from thermodynamics that the G is obtained as a Legendre transform of U (S, V, N ) : S T ; N ; and U G.
G(T, V, ) = U T S N
T =
U
S
U
N
(7.27)
(7.28)
V,N
(7.29)
S,V
From the above, we get,

dG = P dV SdT N d
(7.30)
It follows,
P (T, V, ) =
G
V
(7.31)
S(T, V, ) =
G
T
(7.32)
N (T, V, ) =
(7.33)
T,
V,
T,
If we have an open system of particles obeying Maxwell-Boltzmann, BoseEinstein, or Fermi-Dirac statistics at temperature T and chemical potential
, in a volume V , then the above forrmulae help us calculate the pressure,
entropy and average number of particles in the system. In fact, in the last
section on grand ganonical ensemble, we have derived formal expressions for
the mean and fluctuations of the number of particles in an open system; we
have related the fluctuations to isothermal compressibility - an experimentally
measurable property.
The grand potential for the three statistics is given by,
G(T, V, ) = kB T ln Q
(7.34)
P
Maxwell Boltzmann
kB T i exp[(i )]
P
= kB T i ln [1 exp {(i )}] Bose Einstein (7.35)
P
kB T i ln [1 + exp {(i )}] Fermi Dirac
7.8 Expressions for hN i
117
7.8 Expressions for hN i

7.8.1 Maxwell-Boltzmann Statistics
Y
exp [exp { (i )}}
(7.36)
X
exp [(i )]
= kB T
(7.37)
Q(T, V, ) =
T,V
hN i =
X
i
exp[(i )
(7.38)
exp[(i )]
exp(i )
= Q1 (T, V, )
(7.39)
In the above Q1 is the single-particle canonical partition function. The above

result on hN i is consistent with formal result from grand canonical formalism,
see below.
Q(T, V, ) =
=
=
N =0
N =0
N =0
Q(T, V, N )
(7.40)
QN
1
N!
(7.41)
N ,
exp(N )
QN
1
N!
= Q1
G(T, V, ) = kB T exp()Q1

G
hN i =
= Q1 (T, V )
T,V
(7.42)
(7.43)
(7.44)
(7.45)
7.8.2 Bose-Einstein Statistics

Q(T, V, ) =
Y
i
1
1 exp[(i )]
(7.46)
118
ln Q =
hN i =
T,V
ln [1 exp {(i )}]
exp[(i )]
1 exp[(i )]
(7.47)
(7.48)
7.8.3 Fermi-Dirac Statistics

Q=
Y
i
1 + exp[(i )]
G = kB T ln Q = kB T
hN i =
T,V
X
i
X
i
ln [1 + exp{(i )}]
exp[(i )]
1 + exp[(i )]
(7.49)
(7.50)
(7.51)
7.8.4 Study of a system with fixed N employing grand canonical

formalism
In my next lecture, I shall derive an expression for the mean of occupation
number,
nk , of a single particle quantum state. Then we can express hN i =
P
hn
i.
These results on hN i shall be the same as the one derived in the last
k
k
three subsections.
Grand canonical formalism is not the best suited for studying ideal gas
- classical or quantum. Recall, we introduced an adhoc-indistinquishabilityfactor of N !, suggested by Boltzmann, to restore the extensivity of entropy
while studying closed system. In an open system the number of particles
fluctuate and the issue of non-extensitivity becomes more awkward.
We shall adopt the following strategy. We shall employ grand canonical
formalism, but with a fixed N . The price we have to pay is that in such
an approach the chemical potential is no longer an independent property. It
becomes a function of temperature. Let me explain.
Let us say we keep fixed and change the temperature3 . The statistics
of the number of particles, in particular the mean number of particles in the
system changes when temperature changes. To restore the value of hN i to
the fixed value N , we change the chemical potential. In what follows, we
shall investigate the behaviour of particles under Maxwell-Boltzmann, BoseEinstein, and Fermi-Dirac statistics employing grand canonical formalism but
3
This is permitted in grand canonical formalism since T and are independent

properties of the open system. T is determined by the heat bath and is determined by particle bath.
7.9 All the three statistics are the same at high temperature and/or low densities
with a fixed N ; This implies is not any more an independent property; it is

a function of T .
I must say there is nothing unphysical about this strategy. We are studying
a physical system enclosed by a non-permeable wall - a wall that does not
permit particle exchange. The chemical potential , is a well defined property
of the system. It is just that is not any more under our control4 . The system
automatically selects the value of depending on the temperature.
7.9 All the three statistics are the same at high

temperature and/or low densities
I shall show below that three statistics are the same at high temperatures
and / or low densities by an easy method, an easier method and perhaps the
easiest method.
7.9.1 Easy Method : 3 0

For ease of notation let,
i =
i
kB T
(7.52)
and write the grand canonical partition function under Maxwell-Boltzmann,

Bose-Einstein, and Fermi-Dirac statistics as
exp[exp(i )] Maxwell Boltzmann
1
Y
Bose Einstein
Q=
(7.53)
1
exp(
i)
1 + exp(i ) Fermi Dirac
Let us make an estimate of i in the limit of high temperatures and low

densities.
We had shown earlier that the canonical partition function for an ideal gas
can be written as, see notes,
Q(T, V, N ) =
VN 1
N ! 3N
where the thermal wave length is given by

4
is not determined by us externally by adjusting the particle bath.
(7.54)
119
120
h
=
2mkB T
(7.55)
The free energy is given by

F (T, V, N ) = kB T ln Q(T, V, N )
(7.56)
= kB T [3N ln + N ln V N ln N + N ]
(7.57)
When you take the partial derivative of the free energy with respect to N
keeping temperature and volume constant, you get the chemical potential5 .
Therefore,

F
= kB T [3 ln ln V + 1 + ln N 1]
(7.62)
=
N T,V

= kB T ln 3 + ln(N/V )
= kB T ln(3 )
= ln(3 )
kB T
(7.63)
(7.64)
(7.65)
where = N/V is the number density, i.e. number of particles per unit
volume. Thus we get,
i =
i
ln(3 )
kB T
(7.66)
We have earlier shown that classical limit obtains when 3 0. Thus in

this limit, i or exp(i ) 0. Let xi = exp(i ). Let us express the
grand canonical partition function for the three statistics in terms of the small
parameter xi .
5
In thermodynamics we have
F = U TS
(7.58)
dF = dU T dS SdT
(7.59)
From the first law of thermodynamics we have dU = T dS P dV + dN . Substituting this in the expression for dF above we get,
dF = P dV + dN SdT
(7.60)
Therefore,
=
F
N
(7.61)
T,V
7.9 All the three statistics are the same at high temperature and/or low densities
exp(xi ) Maxwell Boltzmann
Y 1
Bose Einstein
Q=
1 xi
1 + xi Fermi Dirac
(7.67)
In the limit xi 0 we have exp(xi ) = 1 xi and (1 xi )1 = 1 + xi .

Therefore in the limit xi 0, we have
exp(xi ) xi
0 1 + xi Maxwell Boltzmann
1
Y
xi 0 1 + xi Bose Einstein
(7.68)
Q=
1
x
i
1 + xi
= 1 + xi Fermi Dirac
For all the three statistics, the grand canonical partition function take the
same expression. Bosons, Fermions and classical indistinguishable particles
behave the same way when 3 0.
When do we get 3 0 ?
Note that is inversely proportional to square-root of the temperature.
h
=
2mkB T
Hence 0, when T . For a fixed temperature ( is constant), 3 0
when 0. For a fixed , when T ( the same as 0), then 3 0.
Classical behaviour obtains at low densities and/or high temperatures.
Quantum effects manifest only at low temperatures and/or high densities.
7.9.2 Easier Method : 0
Another simple way to show that the three statistics are identical in the limit
of high temperatures and low densities is to recognise (see below) that 3 0
implies 0. Here = exp(), is the fugacity. Let us show this first.
We have shown that
= ln(3 )
kB T
Therefore the fugacity is given by
(7.69)
121
122
= exp
kB T
= exp(ln[3 ]) = 3
Thus 3 0 implies 0.
In the limit 0 we have, for Maxwell-Boltzmann statistics,
Y
exp[ exp(i )]
QMB =
(7.70)
(7.71)
(1 + exp(i ))
(7.72)
In the limit 0, for Bose-Einstein statistics, we have,

QBE =
1
1 exp(i )
(7.73)
[1 + exp(i )]
(7.74)
For Fermi-Dirac statistics, we have exatly,

Y
[1 + exp(i )]
QF D =
(7.75)
Thus in the limit of hight temperatures and low densities Maxwell Boltzmann
statistics and Bose Einstein statistics go over to Fermi - Dirac statistics.
b 1 , n2 , ) = 1
7.9.3 Easiest Method (n
We could have shown easily that in the limit of high temperature and low
density, the three statistics are identical by considering the degeneracy factor
1
Bose Einstein and Fermi Dirac statistics
b=
(7.76)
1
Maxwell Boltzmann statistics

n1 !n2 !
When the temperature is high the number of quantum states that become
available for occupation is very large; When the density is low the number
of particles in the system is low. Thus we have a very few number of particles occupying a very large of quantum states. In other words the number of
quantum states is very large compared to the number of particles.
7.10 Statistics of Occupation Number - Mean
123
Hence the number of micro states with ni = 0 or 1 i are overwhelmingly

large compared to the those with ni 2. In any case, in Fermi-Dirac statistics
ni is always 0 or 1.
In other words, when particles are a few in number(number density is
low) and the accessible quantum levels are large in number (temperature is
high), then micro states with two or more particles in one or more quantum
states are very rare. In other words bunching of particles is very rare at low
densities and high temperatures. Almost always every quantum state is either
unoccupied or occupied by one particle. Very rarely will you find two or more
particles in a single quantum state. Hence the degeneracy factor is unity at
high temperatures and/or low densities. Hence all the three statistics are
identical in the limit of high temperatures and/or low densities.
7.10 Statistics of Occupation Number - Mean

Let us consider the single particle quantum state k with energy k . Let nk
be the number of particles occupying the state k. nk is called the occupation
number. It is a random variable. Here we shall calculate the statistics of the
random variable nk .
Let hnk i denote the average occupation number; the averaging is done over
a grand canonical ensemble of micro states. Formally we have for Bosons and
Fermions,
all
all
all
YX
X
X
n
xi
nxnk
nxnk
i6=k n=0
hnk i =
n=0
all
X
i6=k n=0
xni
all
X
n=0
xnk
n=0
all
X
n=0
7.10.1 Ideal Fermions

For Fermions, n = 0, 1. Hence
hnk iF D =
=
xk
1 + xk
1
+1
x1
k
1
exp[(k )] + 1
xnk
124
7.10.2 Ideal Bosons

For Bosons n = 0, 1, 2, , . In other words, a single particle quantum state
can hold any number of particles. In fact, we shall see later, that at low enough
temperature all the particles would condense into the lowest energy state, i.e.
the ground state. We call this Bose-Einstein condensation.
We carry out the summation6 , in the numerator and and the denominator
of the expression for hnk i, see above, analytically and get,
hnk iBE =
=
xk
(1 xk )
(1 xk )2
xk
1
1
=
= 1
1 xk
exp[(
xk 1
k )] 1
7.10.3 Classical Indistinguishable Ideal Particles
n
n
X
YX
x
x
i
k
n
n!
n!
n=0
i6=k n=0
=
!
n
n
X
YX
x
x
i
k
n!
n!
n=0
n=0
hnk iMB
i6=k
X
xn
n k
n!
n=0
n
X
x
k
n=0
n!
In the above the summation in the numerator and the denominator are evaluated analytically as follows. We start with the definition,
6
Consider
S(x) = 1 + x + x2 + =
1
1x
1
dS
= 1 + 2x + 3x2 + 4x3 + =
dx
(1 x)2
x
dS
x
= x + 2x2 + 3x3 + =
dx
(1 x)2
7.11 Some Remarks on hnk i from the three statistics
125
X
xn
= exp(x)
n!
n=0
Differentiate both sides of the above equation with respect to x. You get
X
nxn1
= exp(x)
n!
n=0
1 X nxn
= exp(x)
x n=0 n!
X
nxn
= x exp(x)
n!
n=0
Therefore,
hnk iMB =
xk exp(xk )
1
= xk = exp[(k )] =
exp(xk )
exp[(k )]
7.11 Some Remarks on hnk i from the three statistics

We can write for all the three statistics,
+1
Fermi Dirac
1
hnk i =
0 Maxwell Boltzmann
where a =
exp[(k )] + a
1
Bose Einstein
Variation of hnk i with energy is shown in the figure, see next page. Note that
k
the x axis is
and the y axis is hnk i.
kB T
We see that for Fermi-Dirac statistics the occupation number never exceeds
unity. When k is negative and |k | is large the value of hnk i tends to
unity. For = and T 6= 0, we have hnk i = 1/2.
126
It is also clear that at high temperature the chemical potential of a Fermion

must be negative, because its behaviour coincides with that of Bosons 7 and
classical indistinguishable particles8 .
For Bose-Einstein statistics we must have
k > k.
In particular the lowest value of energy, say 0 corresponding to the ground
state of the macroscopic system, must be greater than . Let us take 0 = 0
without loss of generality. Hence for Bosons must be negative. Also is a
function of temperature. As we lower T the chemical potential increases and
at T = TBE , becomes zero. TBE is called the Bose-Einstein temperature. At
this temperature the occupancy of the ground state becomes infinitely high.
This leads to the phenomenon of Bose-Einstein Condensation.
7.11.3 Maxwell-Boltzmann Statistics
For the indistinguishable classical particles hnk i takes the familiar exponential
decay form,
hnk i = exp[(k )].
7.11.4 At high T and/or low all statistics give the same hnk i
When
k
,
kB T
all the three statistics coincide. We have already seen that at high temperatures classical behaviour obtains. Then, the only way
k
kB T
can become large at high temperature (note in the expression T is in the
denominator) is when is negative and its magnitude also should increase with
increase of temperature. Thus for all the statistics at high temperature the
chemical potential is negative and its magnitude must be large. Essentially
kB T must be negative and its magnitude must be large at high temperature,

when classical behaviour obtains. But the we know
7
for Bosons the chemical potential is negative at all temperature, and zero at zero
temperature and at temperatures less than a critical temperature called BoseEinstein condensation temperature.
for classical indistinguishable the chemical potential is negative at high temperature, positive at low temperatures and zero at zero temperature.
7.12 Statistics of Occupation Number - Fluctuations
127
1.5
M axwell Boltzmann
hnk i
Bose E instein
0.5
F ermi Dirac
0
0.5
1
3
k
kB T
Fig. 7.1. Average occupation number of a quantum state under Bose-Einstein,
Fermi-Dirac, and Maxwell-Boltzmann statistics
= ln(3 ).
kB T
This means that 3 << 1 for classical behaviour to emerge9 . This is in
complete agreement with our earlier surmise that classical behaviour obtains
at low and/or high T .
Hence all the approaches are consistent with each other and all the issues
fall in place.
In this lecture we saw of the first order statistics - the mean/average/expectation/first
moment - of the occupation number in Fermi-Dirac, Bose-Einstein, and
Maxwell-Boltzmann statistics. In the next lecture I shall tell you of the second
order statistics - the fluctuations of the occupation number.

In the last lecture, I introduced a random variable nk , which denotes the
number of particles in the quantum state k of energy k . We call nk a random
9
Note that ln(x) = 0 for x = 1 and is negative for x < 1. As x goes from 1 to 0,
the quantity ln(x) goes from 0 to .
128
variable because it takes values that are, in general, different for different micro
states. For Bose-Einstein and Fermi-Dirac statistics, a string of occupation
numbers specifies completely a micro state. We found that for Bosons and
Fermions, the average value of nk can be expressed as,
hnk i =
all
X
nxnk
n=0
all
X
xnk
n=0
where,
xk = exp[(k )]
Formally
hnk i =
all
X
nP (n)
n=0
In the above P (n) P (nk = n) is the probability that the random variable
nk takes a value n. Comparing the above with the first equation, we find
P (n) P (nk = n) =
1
all
X
xnk
xm
k
m=0
where the denominator ensures that the total probability is normalized to

unity.
Here, I shall work out explicitly P (n) for Fermi-Dirac, Bose-Einstein, and
Maxwell-Boltzmann statistics. We shall calculate the variance of the random
variable nk ; we shall derive the relative fluctuations : the standard deviation
divided by the mean, for the three statistics.
In Fermi-Dirac statistics the random variable nk can take only two values :
n = 0 and n = 1. Thus,
1
for n = 0
1 + xk
P (n) =
xk for n = 1
1 + xk
We have
hnk i =
1
X
nP (n) =
n=0
129
xk
,
1 + xk
consistent with the result obtained earlier.

For convenience of notation let us denote the mean of the random variable
nk by the symbol . We have,
= hnk i =
xk
.
1 + xk
We thus have,
P (n) =
1 for n = 0
for n = 1
Thus Fermi-Dirac statistics defines a quantum coin with and 1 as the

probabilities of Heads (n = 0) and Tails (n = 1) respectively.
Formally,
hnk i =
hn2k i =
1
X
nP (n) =
n=0
1
X
n2 P (n) =
n=0
2 = hn2k i hnk i2 = (1 )
The relative fluctuations of the random variable nk is defined as the standard
deviation divided by the mean . Let us denote the relative fluctuation by
the symbol . For Fermi-Dirac statistics we have,
r
1
F D = =
1

For Bosons,
P (nk = n) = P (n) = xnk (1 xk )
from which it follows,
=
n=0
nP (n) =
xk
1 xk
130
consistent with the result obtained earlier. Inverting the above, we get,
xk =
1+
Then the probability distribution of the random variable nk can be written

in a convenient form
P (n) =
n
(1 + )n+1
The distribution is geometric, with a constant common ratio /( + 1). We

have come across geometric distribution earlier10 .
Calculation of variance is best done by first deriving an expression for the
moment generating function given by,
P (z) =
z n P (n)
n=0
n

1 X n
=
z
1 + n=0
1+
1
1

z
1+
1
1+
1
1 + (1 z)
Let us now differentiate P (z) with respect to z and in the resulting expression
set z = 1. We shall get hnk i, see below.
P
=
z
(1 + (1 z))2
10

P
=
z z=1
see Problem No. 7 (Problem set : 2, page 4. Let me recollect : The simplest
problem in which geometric distribution arises is in coin tossing. Take a p-coin;
.e. a coin for which the probability of Heads is p and that of Tails is q =
1 p. Toss the coin until the side Heads appears. The number of tosses is a
random variable with a geometric distribution P (n) = q n1 p. We can write this
distribution in terms of = hni = 1/p and get P (n) = ( 1)n1 / n .
131
Differentiating twice with respect to z and setting z = 1 in the resulting

expression shall yield the factorial moment hnk (nk 1)i, see below.
2P
2 2
=
3
2
z
[1 + (1 z)]

2 P
= hnk (nk 1)i = 2 2
z 2 z=1
hn2k i = 2 2 +
2 = hn2k i hnk i2 = 2 +
BE
= =
1
+1
For doing the problem , You will need the following tricks.
S(x) =
n=0
xn = 1 + x + x2 + x3 + =
1
1x
X
1
dS
=
nxn1 = 1 + 2x + 3x2 + 4x3 + =
dx
(1
x)2
n=1
X
x
dS
=
nxn = x + 2x2 + 3x3 + 4x4 + =
dx
(1
x)2
n=1
d
dx

X
dS
2x
1
x
=
n2 xn1 = 1 + 22 x + 32 x2 + 42 x3 + =
+
3
dx
(1
x)
(1
x)2
n=1
d
dx

X
dS
2x2
x
x
=
n2 xn = x + 22 x2 + 32 x3 + 42 x4 + =
+
3
dx
(1 x)
(1 x)2
n=1
You can employ the above trick to derive power series for ln(1 x), see
below.
Z
x2
x3
x4
dx
= ln(1 x) = x +
+
+
1x
2
3
4
132
7.12.3 Maxwell-Boltzmann statistics

For Maxwell-Boltzmann statistics we have,
hnk i = xk
P (nk = n) P (n) =
1
xnk
n! exp(xk )
n
exp()
n!
The random variable nk has Poisson distribution. The variance equals the
mean. Thus the relative standard deviation is given by
MB =
1
=
We can now write the relative fluctuations for the three statistics in one
single formula as,
+1 for
Fermi Dirac Statistics
1
0 for Maxwell Boltzmann Statistics
=
a with a =
1 for
Bose Einstein Statistics
Let us look at the ratio,
r=
P (n)
.
P (n 1)
For the Maxwell-Boltzmann statistics, r = /n. The ratio r is inversely proportional to n. This is the normal behaviour; inverse dependence of r on n is
what we should expect, when the events are uncorrelated. Recall the discussions we had on Poisson process : Problem 22, Assignment 6.
On the other hand, for Bose-Einstein statistics, the ratio is given by
r=
P (n)
=
P (n 1)
+1
r is independent of n. This means, a new particle will get into any of the
quantum states, with equal probability irrespective of how abundantly or
how sparsely that particular quantum state is already populated. An empty
quantum state has the same probability of acquiring an extra particle as an
abundantly populated quantum state.
133
Thus, compared to classical particles obeying Maxwell-Boltzmann statistics, Bosons exhibit a tendency to bunch together. By nature, Bosons like
to be together. Note that this bunching-tendency is not due to interaction
between Bosons. We are considering ideal Bosons. This bunching is purely a
quantum mechanical effect; it arises due to symmetry property of the wave
function.
For Fermions, the situation is quite the opposite. There is what we may
call an aversion to bunching; call it anti-bunching if you like. No Fermion
would like to have another Fermion in its quantum state. A Fermion behaves
like a dog in the manger.
Problems
7.1. Consider Fermi-Dirac statistics at T = 0. Show how the graph of
hnk i versus
k
kB T
will look like. In particular investigate the region around k = .

7.2. See R K Pathria, Statistical Mechanics - Second Edition, Buttteworth
and Heinemann (1996)p.152;Problem:6.1
Show for ideal Bosons in thermal equilibrium the entropy is given by
X
X
hni i lnhni i
hni + 1i lnhni + 1i kB
SF D = kB
i

Show that for ideal Fermions in thermal equilibrium, the entropy is given by
X
X
hni i lnhni i
h1 ni i lnh1 ni i kB
SMB = kB
i

The occupation number nk is random variable. nk is the number of particles
in quantum state k.
In Fermi-Dirac statistics nk can either be zero or one. In Bose-Einstein
statistics nk can take any value from zero to infinity.
Now imagine an intermediate statistics in which nk can take any value
between zero and say L. Let us call the particles that obey such a statistics
as anyons. Show that
hnk i =
1
L+1
exp[(k )] 1 exp[(k )(L + 1)] 1
134
7.5. Start with

P (n) =
n
(1 + )n+1
for Bose-Einstein statistics and show that

hnk i =
nP (n) =
n=0
7.6. For Bose-Einstein statistics, calculate the second moment, hn2k i, the hard
way
hn2k i =
n

1 X 2
n
1 + n=0
1+
Calculate the relative fluctuations and show that your results agree with the
ones obtained employing generating function technique
7.7. Show that the chemical potential of Bosons is negative at all temperatures. At best it can become zero at very low temperatures. Consider the
expression
hnk i =
1
exp[(k )] 1
Take the lowest energy quantum state to be of energy zero. i.e. 0 = 0. Show
positive is unphysical.
8
Bose Einstein Condensation

For Bosons we found that the grand canonical partition funtion is given by,
Q(T, V, ) =
Y
i
1
1 exp[(i )]
(8.1)
The correspondence with thermodynamics is established by the expression for

grand potential denoted by the symbol G(T, V, ). We have,
G(T, V, ) = kB T ln Q(T, V, )
= kB T ln [1 + exp {(i )}]
(8.2)
(8.3)
Recall, from thermodynamics, that the G is obtained by Legendre transform of U (S, V, N ) : S T ; N ; and U G.
G(T, V, ) = U T S N
(8.4)
U
S
(8.5)
U
N
T =
V,N
(8.6)
S,V
From the above, we get,

dG = P dV SdT N d
It follows,
(8.7)
136
8 Bose Einstein Condensation
P (T, V, ) =
G
V
(8.8)
S(T, V, ) =
G
T
(8.9)
N (T, V, ) =
(8.10)
T,
V,
T,
If we have an open system of Bosons at temperature T and chemical potential

, in a volume V , then the above forrmulae help us calculate the pressure,
entropy and number of Bosons in the system. In fact we have calculated the
fluctuations in the number of particles in the open system and related it to
isothermal compressibility - an experimentally measurable property.
The grand potential of the Bosonic system is given by,
(8.11)
(8.12)
8.1.1 hN i =
k hnk i
For Bosons, we found that the average occupancy of a (single-particle) quantum state k, is given by,
hnk i =
exp(k )
1 exp(k )
where is fugacity. We have

= exp().
In the above is the chemical potential and equals the energy change due to
addition of a single particle under constant entropy and volume :

U
=
.
N S,V
The average number of particles in the open system is then
X
hN i =
hnk i
k
X
k
exp(k )
1 exp(k )
(8.13)
137
We shall study Bosonic system with a fixed number of Bosons. We shall

employ grand canonical ensemble formalism in the study. This means is
not an independent property of the system which we can control through a
particle bath. The system chooses the appropriate value of and not we. In
other words is a function of temperature.
8.1.2
k ()
R
()d
Let us now convert the sum over quantum states to an integral over energy.
To this end we need an expression for the number of quantum states in infinitesimal intervals d centered at . Let us denote this quantity by g()d.
We call g() the density of (energy) states. Thus we have,
Z
exp()
N =
g()d
(8.14)
1
exp()
0
We need an expression for the density of states. We have done this exercise
earlier. In fact we have carried out classical counting and quantum counting
and found both lead to the same result. The density of states is given by,
g() = V 2
2m
h2
3/2
1/2
(8.15)
We then have,
N = V 2
2m
h2
3/2 Z
exp() 1/2
d
1 exp()
(8.16)
We note that 0 < 1. This suggests that the integrand in the above can
be expanded in powers of . To this end we write
X
1
=
k exp(k)
1 exp()
(8.17)
k=0
This gives us
X
exp()
=
k+1 exp[(k + 1)]
1 exp()
(8.18)
k=0
k exp[k]
k=1
Substituting the above in the integral we get,
(8.19)
138
2m
h2
3/2 X
2m
h2
3/2 X
2mkB T
h2
3/2 X
= V 2
2mkB T
h2
3/2 X
= V 2
2mkB T
h2
3/2
2mkB T
h2
3/2 X
2
k 3/2
k=1
N = V 2
= V 2
= V 2
= V 2
=V
exp(k)1/2 d
(8.20)
exp(k)(k)1/2 d(k)
3/2 k 3/2
(8.21)
k=1
2mkB T
h2
k=1
k
k 3/2
k=1
exp(x)x1/2 dx
k
(3/2)
k 3/2
k=1
k=1
(8.23)
X k
1
(1/2)
2
k 3/2
3/2 X
(8.22)
(8.24)
k=1
k
k 3/2
(8.25)
(8.26)
(8.27)
In an earlier lecture, we defined a thermal wave length denoted by the symbol
. This is the de Broglie wave length associated with a particle having thermal
energy of kB T . It is also called quantum wavelength. It is given by, see earlier
notes,
h
=
2mkB T
(8.28)
The sum over k, in the expression for hN i, is usually denoted by the symbol
g3/2 ():
g3/2 () =
X
k
k 3/2
k=1
3
2
= + + +
2 2 3 3
(8.29)
(8.30)
139
Thus we get,
V
g3/2 ()
3
(8.31)
N 3
= g3/2 ()
V
(8.32)
N =
We can write the above as,
It is easily verified that at high temperature we get results consistent with

Maxwell Boltzmann statistics1
1
The fugacity is small at high temperature. For small we can replace g3/2 ()
by . We get
N =
(8.33)
This result is consistent with Maxwell-Boltzmann statistics, see below.

For Maxwell-Boltzmann statistics
hnk i = exp(k )
N=
X
k
hnk i =
(8.34)
2m
h2
3/2 Z
2m
h2
3/2 Z
= 2V
2mkB T
h2
3/2
(3/2)
(8.37)
= 2V
2mkB T
h2
3/2
1
(1/2)
2
(8.38)
= 2V
2mkB T
h2
3/2
(8.39)
exp(k ) = 2V
= 2V
=V
2mkB T
h2
d 1/2 exp()(8.35)
0
3/2
()1/2 exp() d()

(8.36)
3/2
(8.40)
(8.41)
140
8.1.3 g3/2 () versus

I have plotted g3/2 () as a function of in the figure below. Note that the
4
3.5
g3/2 ( = 1) = (1) = 2.612
3
2.5
2
1.5
g3/2 ()
1
0.5
0.2
0.4
0.6
0.8
Fig. 8.1. g3/2 () versus . Graphical inversion to determine fugacity
infinite series converges only for 0 1. We have

g3/2 ( = 1) =
X
1
k 3/2
(8.42)
k=1
= (3/2)
(8.43)
= 2.612
(8.44)
where (n) is the Riemann zeta function, defined as,

(n) =
X
1
.
kn
k=1
(8.45)
141
We note that the value of . g3/2 does not exceed 2.612.

8.1.4 Graphical inversion to determine fugacity
Let us say we know the values of N , V and temperature T of a system. Then
we can find the value of fugacity by graphical inversion : Draw a line parallel
to the x axis at y = N 3 /V and read off the value of at which the line cuts
the curve g3/2 ().
Once we get the fugacity, we can determine all other thermodynamic properties of the open system employing the formalism of grand canonical ensemble.
So far so good.
But then we realise that the above graphical inversion scheme does not
permit evaluation of the fugacity of a system with N 3 /V greater than 2.612.
This is absurd.
There must be something wrong with what we have done.
8.1.5 Treatment of the Singular Behaviour
We realise that when N 3 /V approaches 2.612, the fugacity approaches
unity; the chemical potential approaches zero2 . We have already seen that
at = 0 the occupancy of the ground state diverges. The singular behaviour
of the ground state occupancy was completely lost when we replaced the sum
over quantum states by an integral over energy : the weighting function is the
density of states, given by, g() 1/2 . The density of states vanishes at zero
energy. Hence we must take care of the singular behaviour separately.
We have,
X exp(k )
N =
(8.46)
1 exp(k )
k
X exp(k
+
1
1 exp(k )
(8.47)
In the above, we have separated the ground state occupancy and the occupancy of all the excited states. Let hN0 i denote the ground state occupancy.
It is given by the first term,
N0 =
(8.48)
The occupancy of all the excited states is given by the second term, where
the sum is taken only over the indices k representing the excited states. Let
Ne denote the occupancy of excited states. It is given by,
2
The chemical potential approaches the energy of the ground state. With out loss
of generality, we can set the ground state at zero energy.
142
Ne =
X
k
exp(k
1 exp(k )
(8.49)
In the above, the sum over k can be replaced by an integral over energy. In
the integral over energy, we can still keep the lower limit of integration as
zero, since the density of states giving weight factors for occupancy of states
is zero at zero energy. Accordingly we write
N = hN0 i + hNe i
=
V
g3/2 ()
+
1 3
(8.50)
(8.51)
We thus have,
N 3
3
=
+ g3/2 ()
V
V 1
(8.52)
Let us define the number of density - number of particles per unit volume,
denoted by the symbol . It is given by
=
N
V
(8.53)
The function /(1 ) diverges at = 1, as you can see from the figure
below.
Hence the relevant curve for carrying out graphical inversion should be the
one that depicts the sum of the singular part (that takes care of the occupancy
of the ground state) and the regular part (that takes care of the ocupancy
of the excited states). For a value of 3 /V = .05 we have plotted both the
curves and their sum in the figure below. Thus for any value of 3 we can
now determine the fugacity by graphical inversion.
We carry out such an exercise and obtain the values of for various values
of 3 and the figure below depicts the results.
It is clear that when 3 > 2.612, the fugacity is close unity. How close
can it approach unity ?
Let us postulate3
a
= 1 .
N
where a is a number. To determine a we proceed as follows.
3
a
We have reasons to postulate = 1 . This is related to the mechanism underlyN
ing Bose-Einstein condensation we shall discuss the details later. In fact, following
Donald A McQuarrie, Statistical Mechanics, Harper and Row (1976)p.173 we can
a
make a postulate = 1 . This should also lead to the same conclusions.
V
143
100
90
80
70
60
50
40
30
20
10
0
0.2
0.4
0.6
0.8
Fig. 8.2. Singular part of hN i
We have,
N
=
1
1
a
N
if N >> a
a
(8.54)
(8.55)
We start with,
3 =
3
+ g3/2 ()
V 1
(8.56)
Sunstitute = 1 a/N in the above and get4 ,

3 =
4
g3/2 (1 a/N ) g3/2 (1),
3
+ g3/2 (1)
a
(8.57)
144
8
7
6
5
4
g3/2 ( = 1) = (1) = 2.612
3
+ g3/2 ()
V 1
2
1
0
0.2
0.4
0.6
0.8
Fig. 8.3. 3 versus . The singular part [3 /V ][/(1)] (the bottom most curve),
the regular part g3/2 () (the middle curve) , and the total (3 ) are plotted. For this
plot we have taken 3 /V as 0.05
Thus we get,
a=
3
3 g3/2 (1)
(8.58)
The point 3 = g3/2 (1) = 2.612 is a special point indeed. What is the
physical significance of this point ? To answer this question, consider the
quantity 3 as a function of temperature with kept at a constant value.
The temperature dependence of this quantity is shown below.
3
2mkB T
3
(8.59)
At high temperature for which 3 < g3/2 (1) = 2.612, we can determine
the value of from the equation g3/2 () = 3 by graphical or numerical
inversion.
145
1
0.9
0.8
0.7
0.6
0.5
0.4
3
+ g3/2 ()
=
V 1
3
= .05
V
3
0.3
0.2
0.1
0
2.612
0
0.5
1.5
2.5
3.5
3
Fig. 8.4. Fugacity versus 3
At low temperatures for which 3 > 2.612, we have = 1 a/hN i

where
a=
3
g3/2 (1)
(8.60)
The quantity /(1 ) is the average number of particles in the ground

state. At temperatures for which 3 > 2.612, we have,
(8.61)
N
a
(8.62)
N0
1
=
N
a
(8.63)
N0 =
= 1
1
g3/2 (1)
3
(8.64)
We can write the above in a more suggestive form by defining a temperature

TBEC by
146
3BEC = g3/2 (1)
(8.65)
Therefore,
N0
1
=
N
a
(8.66)
= 1
3BEC
3
= 1
3
(8.68)
!3
TBEC
= 1
= 1
BEC
(8.67)
T
TBEC
3/2
(8.69)
fer T < TBEC
(8.70)
We have depicted the above behaviour of the fractional number of particles

in the ground state as a function of temperature in the figure below.
1.5
N0
=1
N
T
TBEC
43/2
N0
N
0.5
0.5
1.5
T
TBEC
Fig. 8.5. Ground state occupation as a function of temperature
147
8.1.6 Bose-Einstein Condensation Temperature

Thus we can define the temperature at which Bose-Einstein condensation
takes place as,
3

h
N
= 2.612
(8.71)
V
2mkB T
kB TBEC =
h2
2m
N
2.612 V
2/3
(8.72)
At T = TBEC , Bose- Einstein condensation sets in and the ground state

occupancy becomes anomalously larger and larger as temperature decreases
further.
8.1.7 Grand Potential for Bosons
The grand potential for Bosons is given by
G(T, V, ) = kB T ln Q(T, V, )
= kB T ln
= kB T
X
k
"
Y
k
(8.73)
1
(1 exp(k )
ln[1 exp(k )]
(8.74)
(8.75)
Now we shall be careful and separate the singular part and regular part to
get,
X
ln[1 exp(k )]
(8.76)
G = kB T ln(1 ) + kB T
k
We have
ln[1 exp(k )] =
k exp(kk )
(8.77)
k=1
Substitute the above in the expression for G. Convert the sum of k by an

integral over by the prescription below :
X
k
() V 2
We proceed as follows:
2m
h2
3/2 Z
() 1/2 d,
(8.78)
148
G = kB T ln(1 ) + kB T
X
k
ln[1 exp(k )]
= kB T ln(1 ) kB T V 2
2m
h2
3/2 X
2m
h2
Z
3/2 X
k=1
= kB T ln(1 ) kB T V
= kB T ln(1 ) kB T
d 1/2 exp(k)
(8.80)
(k)1/2 exp(k)
d(k)(8.81)
k 3/2 3/2
2mkB T
h2
2mkB T
h2
k=1
(8.79)
3/2
X
k
(3/2)
k 3/2
3/2 X
k=1
(8.82)
k=1
k
k 3/2
V
g3/2 ()
3
(8.83)
(8.84)
Thus we have,
G(T, V, ) = kB T ln(1 )
V
kB T g3/2 ()
3
(8.85)
8.1.8 Average Energy of Bosons

An open system is described by a grand canonical partition function. It is
formally given by,
X
exp[(Ei Ni )]
(8.86)
Q(, V, ) =
i
In the above Ei is the energy of the open system when in micro state i; Ni is
the number of particles in the open system when in micro state i. Let = .
Then we get,
X
exp(Ei ) exp(+Ni )
(8.87)
Q(, V, ) =
i
We differentiate Q with respect to the variable , keeping constant. We get
X
Q
i exp[i + Ni )
=
149
(8.88)
1 Q
= hEi = U
Q
(8.89)
(8.90)
For Bosons, we have,

Q=
ln Q =
U =
1
1 exp(i )
X
i
ln[1 exp(i )]
X i exp(i )
1 exp(i )
i
(8.91)
(8.92)
Let us now go to continuum limit by converting the sum over micro states by
an integral over energy and get,
U =
3
1
V kB T 3 g5/2 ()
2
(8.93)
Let us now investigate the energy of the system at T > TBEC . When
temperature is high, the number of Bosons in the ground state is negiligibly
small. Hence the total energy of the system is the same as the one given above.
For temperatures less that TBEC , the ground state gets populated anomalously. The Bosons in the ground state do not contribute to the energy. Hence
for T < TBEC , we have

N0
1
3
(8.94)
U = V kB T 3 g5/2 (1) 1
2
N
1
3
= V kB T 3 g5/2 (1)
2
T
TBEC
3/2
(8.95)
Thus we have,
U =
3
V
kB T 3 g5/2 ()
for T > TBEC
3/2

V
T
3
for T < TBEC

kB T 3 g5/2 (1)
2
TBEC
We can cast the above in a suggestive form, see below.
(8.96)
150
We have,
N = N0 + Ne
(8.97)
N0 =
(8.98)
Ne =
V
g3/2 ()
3
(8.99)
For T > TBEC , we have Ne = N . Therefore,

V
N
=
3
g3/2 ()
(8.100)
substituting the above in the expression for U we get

g5/2 ()
3
hN
ik
T
for T > TBEC
B
2
g3/2 ()
U =

3/2

g5/2 (1)
3
T
hN ikB T
for T < TBEC
2
g3/2 (1) TBEC
(8.101)
8.1.9 Specific Heat Capacity of Bosons

CV
for T > TBEC
N kB
Let us consider first the case with T > TBEC . We have

U
3 g5/2 ()
= T
N kB
2 g3/2 ()
1 U
CV
=
=
N kB T
N kB
T
3T g5/2 ()
2 g3/2 ()
(8.102)

To carry out the derivative in the above, we need the following :

3
g3/2 () =
g3/2 ()
T
2T
First Relation:

3
g3/2 () =
g3/2 ()
T
2T
(8.103)
151
Proof : We start with 3 = g3/2(). Therefore,

g3/2 () = 3
(8.104)
[g ()] = 3 2
T 3/2
T
= 3 3
(8.105)

h
2mkB T
(8.106)
3
1
h
2
2
2mkB T 3/2
(8.107)
h
3
2
2T
2mkB T
(8.108)
3
3
2T
(8.109)
3
g3/2 ()
2T
(8.110)

Q.E.D
[gn/2 ()] =
g(n/2)1 ()
Second Relation
1
[gn/2 ()] = g(n/2)1 ()
Proof :
X
k
. Therefore,
We have by definition, gn/2 () =
k n/2
k=1
152
#
"
X k
[g ()] =
n/2
k n/2
k=1
=
X
kk1
k=1
k n/2
(8.111)
(8.112)
1X
k
k (n/2)1
(8.113)
1
g(n/2)1 ()
(8.114)
k=1
1 d
dT
Q.E.D
g3/2 ()
2T g1/2 ()
Third Relation
3 g3/2 ()
1 d
=
dT
2T g1/2 ()
Proof :
We proceed as follows :

[g3/2 ()] =
[g3/2 ()]
T
dT
3
1
d
g3/2 () = g1/2 ()
2T
dT
(8.115)
(8.116)
From the above we get,

3
1 d
=
dT
2T
Q.E.D
We have,
g3/2 ()
g1/2 ()
(8.117)
CV
=
N kB
T
3T g5/2 ()
2 g3/2 ()
153
3 g5/2 () 3T
+
2 g3/2 ()
2 T
g5/2 ()
g3/2 ()
3 g5/2 () 3T
=
2 g3/2 ()
2
"
g5/2 () d
g5/2 () g3/2 ()
1
2 ()
g3/2
T
g3/2 ()
dT
3 g5/2 () 3T
=
2 g3/2 ()
2
"
g5/2 ()
2 ()
g3/2
#

3
1
1
d
g3/2 ()
g3/2 ()
2T
g3/2 ()
dT
3 g5/2 () 3T
=
2 g3/2 ()
2
"
g5/2 ()
2 ()
g3/2
#

3
1 d
g3/2 ()
2T
dT
3 g5/2 () 3T
=
2 g3/2 ()
2
"
g5/2 ()
2 ()
g3/2
#

3
3 g3/2 ()
g3/2 () +
2T
2T g1/2 ()
3 g5/2 () 9 g5/2 () 9 g3/2 ()

+
2 g3/2 () 4 g3/2 () 4 g1/2 ()
15 g5/2 () 9 g3/2 ()
4 g3/2 () 4 g1/2 ()
(8.118)
CV
for T < TBEC
N kB
Now, let us consider the case with T < TBEC . We have,
Thus we have,
g5/2 (1)
U
3
= T
N kB
2
g3/2 (1)
3 g5/2 (1) 5
1
CV =
N kB
2 g3/2 (1) 2
T
TBEC
T
TBEC
3/2
(8.119)
3/2
(8.120)
154
15 g5/2 () 9 g3/2 ()
for T > TBEC
4 g3/2 () 4 g1/2 ()
1
CV =
N kB

3/2
15 g5/2 (1)
T
for T < TBEC
4 g3/2 (1) TBEC
(8.121)
The specific heat is plotted against temperature in the figure below. The
2
1.8
1.6
Classical : 3NkB /2
1.4
CV
Nk B
1.2
1
0.8
0.6
0.4
0.2
0
0.2
0.4
0.6
0.8
1.2
1.4
1.6
1.8
T
TBEC
Fig. 8.6. Heat capacity in the neighbourhood of Bose - Einstein condensation temperature
cusp in the heat capacity at T = TBEC is the signature of the Bose-Einstein

condensation. Asymptotically T , the heat capacity tends to the classical
value consistent with equi-partition.
8.1.10 Mechanism of Bose-Einstein Condensation
Let the ground state be of energy 0 0. For example consider particle in a
three dimensional box of length L. The ground state is
(nx , ny , nx ) = (1, 1, 1).
The ground state energy is
1,1,1 = 0 =
.
3h2
8mL2
155
The chemical potential is always less than or equal to 0 . As temperature

decreases, the chemical potential increases and comes closer and closer to the
ground state energy 0 0. Let us estimate how close can get to 0 . In other
words, we want to estimate the smallest possible value of (0 )/[kB T ]. To
this end, consider the expression for the average number of Bosons in the
ground state. Let us denote this by N0 . It is given by,
N0 =
1

0
1
exp
kB T

(8.122)
As temperature goes to zero, the chemical potential goes toward the

0
ground state energy. For a non-zero value of T , when
is small, we
kB T
can write

0
0
exp
=1+
(8.123)
kB T
kB T
Substituting this in the expression for N0 , given above, we get,
N0 =
kB T
0 (T )
(8.124)
N0 goes to zero5 as T . At high temperature, the ground state occupancy

is extremely small, as indeed it should.
Therefore we have,
1
0
=
kB T
N0
(8.125)
The largest value that N0 can take is N , i.e. when all the particles condense
into the ground state. In other words, the smallest value that 1/N0 can take is
1/N . Hence (0 )/[kB T ] can not be smaller than 1/N . The smallest possible
value it can take is 1/N - inverse of the average number of particles in the
entire system.
0
1
kB T
N
(8.126)
Thus, measured in units of kB T , the chemical potential shall always be less

that the ground state energy at any non-zero temperature. At best, the quantity ( ), expressed in units of thermal energy (kB T ), can only be of the
order of 1/N .
Therefore the chemical potential can never take a value close to any of
the excited states, since all of them invariably lie above the ground state. In
5
For large T , the numerator is large; but the denominator is also large. Note that
(T ) is negative and large for large T . In fact the denominator goes to infinity
faster than the numerator.
156
a sense, the ground state forbids the chemical potential to come close to any
energy level other than the ground state energy. It sort of guards all the excited
states from a close visit of . As T 0, the number of Bosons in the ground
state increases.
This precisely is the subtle mechanism underlying Bose-Einstein
condensation.
9
Statistical Mechanics of Harmonic Oscillators
9.1 Classical Harmonic Oscillators

Consider a closed system of 3N harmonic oscillators at temperature T . The
oscillators do not interact with each other and are distinguishable. Let us
derive an expression for the single-oscillator partition function.
The energy of an harmonic oscillator is given by
E=
1
p2
+ m 2 q 2
2m 2
(9.1)
where q is the distance between the current position of the harmonic oscillator
and its mean position and p its momentum. is the characteristic frequency
of the oscillator and m its mass.
We have,

2
Z
Z +
1
p
1 +
(9.2)
+ m 2 q 2
dq
dp exp
Q1 (T ) =
h
2m 2
We can write the above in a convenient way as a product of two Gaussian

integrals, one over dq and the other over dp, as
1
Q1 (T ) =
h
dq exp
2
"
!2
r
kB T
m 2
q2
p2
1
dp exp
3
2
mkB T
(9.3)
(9.4)
Let 1 and 2 denote the standard deviations of of the two zero-mean Gaussian
distributions. These are given by,
158
9 Statistical Mechanics of Harmonic Oscillators
1 =
kB T
m 2
(9.5)
2 =
p
mkB T
(9.6)
1 2 =
kB T
We have normalization identity for a Gaussian

Z +
1 x2
dx exp 2 = 2
2
(9.7)
(9.8)
Therefore,
Q1 (T ) =
2
kB T
1 2 =
h
~
(9.9)
If all the oscillators are identical i.e. they all have the same characteristic
frequency of oscillations, then
3N

kB T
(9.10)
Q3N (T ) =
~
See footnote1 where we have considered 3N harmonic oscillators with 3N
characteristic frequencies.
9.2 Helmholtz Free Energy

The free energy of a system of 3N non-interacting, identical classical harmonic
oscillators is given by

kB T
~
F (T, V, N ) = 3N kB T ln
= 3N kB T ln
(9.12)
~
kB T
See footnote2 where I have expressed the free energy as an integral over the
distribution of frequencies of the harmonic oscillator for N large.
Once we know of free energy, we can employ the machinery of thermodynamics and get expressions for all other thermodynamic properties of the
system, see below.
1
On the other hand if the oscillators have distinct characteristic frequency {i :

i = 1, 2, , 3N }, then
Q3N (T ) =
3N
Y
kB T
~i
i=1
If the oscillators have different frequencies then
(9.11)
9.3 Thermodynamic Properties of the Oscillator System
159
9.3 Thermodynamic Properties of the Oscillator System

F (T, V, N ) = U T S
(9.16)
dF = dU T dS SdT
(9.17)
= SdT P dV + dN
(9.18)
Thus for a system of identical, non-interacting classical harmonic oscillators

F
P =
(9.19)
V T,N
=0
why?
F
N
S=
(9.21)
T,V
= kB T ln

(9.20)
F
T

~
kB T
= N kB ln
(9.22)
(9.23)
V,N
kB T
~
+1
(9.24)
We also have,
F (T, V, N ) = kB T
3N
X
ln
i=1
kB T
~i
(9.13)
If N is large we can define g()d as the number of harmonic oscillators with

frequencies in an interval d around . The sum can be replaced by an integral,

Z
kB T
F (T ) = kB T
ln
g()d
(9.14)
~
0
We have the normalization
Z
g()d = 3N
0
(9.15)
160
U =
ln Q
(9.25)
V,N
= 3N kB T,
(9.26)
consistent with equipartition theorem which says each quadratic term in the
Hamiltonian carries kB T /2 of energy. The Hamiltonian of a single harmonic
oscillator has two quadratic terms - one in position q and the other in momentum p.
We also find that the results are consistent with the Dulong and Petits
law which says that the heat capacity at constant volume is independent of
temperature:

U
(9.27)
CV =
T V
= 3N kB = 3nR
(9.28)
CV
= 3R = 6 calories (mole)1 (Kelvin)1
n
(9.29)
More importantly, the heat capacity is the same for all the materials; it depends only on the number of molecules or the number of moles of the substance
and not on what the substance is. The heat capacity per mole is approximately
6 calories per Kelvin.
9.4 Quantum Harmonic Oscillator

Now let us consider quantum harmonic oscillators. The energy eigenvalues of
a single one dimensional harmonic oscillator is given by

1
~ : n = 0, 1, 2,
(9.30)
n = n +
2
The canonical partition function for a single (quantum) harmonic oscillator is
then,
Q1 () = exp(~/2)
[exp(~)]
(9.31)
n=0
exp(~/2)
1 exp(~)
(9.32)
The partition function of a collection of 3N non-interacting quantum harmonic

oscillators is then given by
9.4 Quantum Harmonic Oscillator
exp(3N ~/2)
QN (T ) =
[1 exp(~)]
161
(9.33)
3N
See footnote3 for an expression of the partition function for 3N independent

harmonic oscillators with different frequencies.
The free energy is given by,
F (T, V, N ) = kB T ln Q3N (T )
(9.35)
1
= 3N
~ + kB T ln {1 exp(~)}
2
(9.36)
See footnote4 for an expression for free energy for 3N independent harmonic
oscillators with different frequencies.
We can obtain the thermodynamic properties of the system from the free
energy. We get,
If the harmonic oscillators are all of different frequencies, the partition function
is given by
Q(T ) =
3N
Y
exp(~i /2)
1
exp(~i )
i=1
(9.34)
For 3N independent harmonic oscillators with different frequencies we have

F =
3N
X
~i
i=1
=
Z
d
0
+ kB T ln {1 exp(~i )}

~
+ kB T ln {1 exp(~)} g()
2
(9.37)
(9.38)
d g() = 3N
(9.39)
162
F
N
(9.40)
T,V
1
~ + kB T ln [1 exp(~)]
2
P =
F
V
(9.41)
(9.42)
T,N
=0
S=
(9.43)

F
T
= 3N kB
U =
(9.44)
V,N
~
ln {1 exp(~)}
exp(~) 1
ln Q
= 3N
(9.45)
(9.46)
~
~
+
2
exp(~) 1
(9.47)
The expression for U tells that the equipartition theorem is the first victim of
quantum mechanics : Quantum harmonic oscillators do not obey equipartition
theorem. The average energy per oscillator is higher than the classical value
of kB T . Only for T , we have kB T >> ~, the quantum results
coincide with the classical results.
The heat capacity at constant volume is given by

U
(9.48)
CV =
T V,N
=
3N
kB
~
T
2
exp[~]
2
(exp[~] 1)
(9.49)
The second victim of quantum mechanics is the law of Dulong and Petit. The
heat capacity depends on temperature and on the oscillator frequency. The
heat capacity per mole will change from substance to substance because of
its dependence on the oscillator frequency. Only in the limit of T (the
same as 0), do we get the classical results.
9.5 Specific Heat of a Crystalline Solid
163
The temperature dependence of heat capacity is an important quantum

outcome. In fact we find that the heat capacity goes to zero exponentially as
T 0. However experiments suggest that the fall is algebraic rather than
exponential. The heat capacity goes to zero as T 3 . This is called T 3 law.

In the above we studied the behaviour of a collection of independent identical
harmonic oscillators in a canonical ensemble. We shall see below how such
a study is helpful toward understanding of the behaviour of specific heat of
crystalline solid as a function of temperature.
A crystal is a collection of say N atoms organized in a regular lattice. Let
{x1 , x2 , , x3N } specify the 3N positions of these atoms. For example we
can consider a periodic array of atoms arranged at regular intervals along the
three mutually perpendicular directions, constituting a three dimensional cubic structure. We can think of other structures like face-centred cubic (FCC),
body centred cubic (BCC) lattices.
An atom vibrates around its lattice location say (
xi , xi+1 , xi+2 ). It does
not make large excursions away from its lattice location. We must bear in
mind that the atoms are not independently bound to their lattice position.
They are mutually bound5 .
Consider a total of N atoms organized in a three dimensional lattice. Each
atom executes small oscillations about its mean position. In the process of
oscillations each atom pulls or pushes its neighbours; these neighbours in turn
pull or push their neighbours and so on. The disturbance propagates in the
crystal. We can set up equations of motion for the three coordinates of each
of the atoms. We shall have 3N coupled equations. Consider the Hamiltonian
of a solid of N atoms with position coordinates are {x1 , x2 , x3N }. When
the system of atoms is in its lowest energy, the coordinates are x
1 , x
2 , x
3N .
Let V (x1 , x2 , x3N ) denote the potential energy. We express the potential
energy under harmonic approximation, as
5
To appreciate the above statement, consider a class room wherein the chairs
are already arranged with constant spacing along the length and breadth of the
class room. The students occupy these chairs and form a regular structure. This
corresponds to a situation wherein each student is bound independently to his
chair.
Now consider a situation wherein the students are mutually bound to each
other. Let us say that the students interact with each other in the following way :
Each is required to keep an arms length from his four neighbours. If the distance
between two neighbouring students is less, they are pushed outward; if more,
they are pulled inward. Such mutual interactions lead to the student organizing
themselves in a two dimensional regular array
I shall leave it to you to visualize how such mutual nearest neighbour interactions can give rise to three dimensional arrays.
164
V (x1 , x2 , x3N ) = V (
x1 , x
2 , x
3N ) +

3N
X
V
i=1
xi
(9.50)
x
1 ,
x2 , ,
x3N
(xi xi ) +
(9.51)
2

3N X
3N
X
1
V
(xi x
i )(xj x
(9.52)
j)
2 xi xj x1 ,x2 ,x3N
i=1 j=1
The first term gives the minimum energy of the solid when all its atoms are
in their equilibrium positions. We can denote this energy by V0 .
The second set of terms involving the first order partial derivatives of
the potential are all identically zero by definition : V has a minimum at
{xi = xi i = 1, 3N }
The third set of terms involving second order partial derivatives describe
harmonic vibrations. We neglect the terms involving higher order derivatives
and this is justified if only small oscillations are present in the crystalline .
Thus under harmonic approximations we can write the Hamiltonian as,
H = V0 +

2
3N
X
1 di
i=1
dt
3N
3N X
X
i,j i j
(9.53)
i=1 j=1
where
= xi xi
i,j
1
=
2
(9.54)
2V
xi xj
(9.55)
x
1 ,
x2 , ,
x3N
We shall now introduce a linear transformation from the coordinates {i :

i = 1, 3N } to the normal coordinates {qi : i = 1, 3N }. We choose the linear
transformation matrix such that the Hamiltonian does not contain any cross
terms in the q coordinates.
H = V0 +
3N
X
1
i=1
m q2 + i2 qi2
(9.56)
where {i : i = 1, 3N } are the characteristic frequencies of the normal modes

of the system. These frequencies are determined by the nature of the potential
energy function V (x1 , x2 , x3N . Thus the energy of the solid can be considered as arising out of a set of 3N one dimensional, non interacting, harmonic
oscillators whose characteristic frequencies are determined by the nature of
the atoms of the crystalline solid, the nature of their mutual interaction, the
nature of the lattice structure etc..
165
Thus we can describe the system in terms of independent harmonic oscillators by defining a normal coordinate system, in which the equations of motion
are decoupled. If there are N atoms in the crystals there are 3N degrees of
freedom. Three of the degrees of freedom are associated with the translation
of the whole crystal; and three with rotation. Thus, there are strictly 3N 6
normal mode oscillations. If N is of the order of 1025 or so, it doesnt matter
if the number of normal modes is 3N 6 and not 3N .
We can write the canonical partition function as
Q=
3N
Y
exp(~i /2)
1
exp(~i )
i=1
(9.57)
There are 3N normal frequencies. We can imagine them to be continuously

distributed. Let g()d denote the number of normal frequencies between
and + d. The function g() obeys the normalization
Z
g()d = 3N
(9.58)
0
We have,
ln Q =
3N
X
~i
i=1

+ ln {1 exp(~i )}
~
+ ln {1 exp(~)}
2
(9.59)
g()d
(9.60)
The problem reduces to finding the function g(). Once we know g(), we
can calculate the thermodynamic properties of the crystal. In particular we
can calculate the internal energy U and heat capacity, see below.

Z
~ exp(~)
~
+
g()d
(9.61)
U =
2
1 exp(~)
0
CV = kB
~
~
+
2
exp(~) 1
g()d
(~)2 exp(~)
g()d
[exp(~) 1]2
(9.62)
(9.63)
166
The problem of determining the function g() is a non-trivial task. It is precisely here that the difficulties lie. However, there are two well known approximations to g(). One of them is due to Einstein and the other due to
Debye.
9.6 Einstein Theory of Specific Heat of Crystals

Einstein assumed all the 3N harmonic oscillators to have the same frequency.
In other words,
g() = 3N ( E )
(9.64)
where E is the Einstein frequency or the frequency of the Einstein oscillator.

The Einstein formula for the heat capacity is then given by

2
~
exp(~E /[kB T ])
(9.65)
CV = 3N kB
kB T
(exp(~E /[kB T ]) 1)2
Let us define
E =
~E
kB
(9.66)
and call E as Einstein temperature. Verify that this quantity has the unit
of temperature. In terms of Einstein temperature we have,

2
E
exp(E /T )
CV = 3N kB
(9.67)
2
T
[exp(E /T ) 1]
9.1 Show that in the limit of T , the heat capacity

of the Einstein solid tends to the value 3N kB = 3R =
6 cal (mole)1 K 1 predicted by Dulong and Petit.
9.2 Show that in the low temperature limit,
CV
T 0
3N kB
E
T
2
exp(E /T )
(9.68)
Experiments suggest T 3 decay of CV with temperature. In the next class I

shall discuss Debyes theory of heat capacity. We will find that Debyes theory
gives the T 3 law.
9.7 Debye theory of Specific Heat
167

Debye assumed a continuous spectrum of frequencies, cut off at an upper limit
D . Let us call it Debye frequency. Debye assumed based on an earlier work
of Rayleigh, that g() = 2 , where the proportionality constant depends on
the speed of propagation of the normal mode, its nature6 , its degeneracy7 .
From the normalization condition,
Z D
2 d = 3N
(9.69)
3
we get, = 9N/D
. Thus we have
g() =
9N
2 for D
3
D
(9.70)
for > D
Let us now calculate CV under Debyes theory. We start with

Z
(~)2 exp(~)
CV (T ) = kB
g()d
[exp(~) 1]2
0
(9.71)
Let
x = ~
(9.72)
~D
kB
(9.73)
D =
D is called the Debye temperature. Then we have,

CV = (3N kB ) 3
T
D
3 Z
D /T
x4 exp(x)
dx
[exp(x) 1]2
(9.74)
Let us consider the integral in the above expression and write,

I=
/T
x4 exp(x)
dx
(exp(x) 1)2
(9.75)
Integrating by parts8 we get,

6
7
transverse or longitudinal
transverse mode is doubly degenerate and longitudinal mode is non-degenerate,
etc..
Take u(x) = x4 and dv(x) = exp(x)dx/[exp(x) 1]2
168
1
I=
exp(D /T ) 1
D
T
4
+4
D /T
x3
dx
exp(x) 1
(9.76)
The expression for heat capacity is then,

"
#

3 Z D /T

D
1
T
x3
CV = (3N kB ) 3
+ 12
(9.77)
dx
T
exp(D /T ) 1
D
exp(x) 1
0
Let us now consider the behaviour CV in the limit of T . we have
T >> D . We can set exp(D /T ) 1 + (D /T ); also in the integral we
can set exp(x) = 1 + x. Then we get,
"
#

3 Z D /T
T
CV = 3N kB 3 + 12
x2 dx
(9.78)
D
0
= 3N kB (3 + 4)
(9.79)
= 3N kB
(9.80)
In the low temperature limit we have T << D . We start with,

#
"

3 Z D /T

1
T
x3
D
+ 12
(9.81)
dx
CV = (3N kB ) 3
T
exp(D /T ) 1
D
exp(x) 1
0
In the limit T 0, the first term inside the square bracket goes to zero like
exp(D /T ). The upper limit of the integral in the second term inside the
square bracket can be set to . From standard integral tables9 , we have,
Z
4
x3
dx =
(9.82)
exp(x) 1
15
0
Thus we have in the low temperature limit,
CV T 0
12 4
N kB
5
T
D
3
(9.83)
Riemann Zeta Function

In an earlier class, we came across an integral,
Z
x3
dx
exp(x) 1
0
9
(9.84)
The integral equals (4)(4), where () is the gamma function and () is the
Riemann zeta function. (4) = 3! = 6 and (4) = 4 /90. See e.g. G B Arfken
and H J Weber, Mathematical Methods for Physicists, Fourth Edition, Academic
Press, INC, Prism Books PVT LTD (1995).
169
The value of the integral is 4 /15.

This is a particular case of a more general result based on Riemann zeta
function,
Z
xp
dx = (p + 1)(p + 1)
(9.85)
exp(x) 1
0
where () is the usual gamma function defined as
Z
(z) =
xz1 exp(x)dx for Real(z) > 0,
(9.86)
and () is the Riemann zeta function, see below. Note (2) = 2 /6 and
(4) = 4 /90, etc..
Riemann zeta function is defined as
(p) =
np
(9.87)
n=1
Take f (x) = xp and then
xp dx =
p+1

x
for p 6= 1
p + 1 1
ln x
(9.88)
for p = 1
The integral and hence the series is divergent for p 1 and convergent for
p > 1.
9.7.1 Bernoulli Numbers
Bernoulli numbers Bn are defined by the series
X
x
xn
=
Bn
exp(x) 1 n=0 n!
(9.89)
which converges for |x| < 2.

By differentiating the power series repeatedly and then setting x = 0, we
obtain

n
x
d
.
(9.90)
Bn =
dxn exp(x) 1 x=0
170
9.3 Show that B0 = 1; B1 = 1/2; B2 = 1/6; B4 = 1/30;

B6 = 1/42; B2n+1 = 0 n 1.
Euler showed that,

B2n = (1)n1
= (1)n1
2(2n)! X 2n
p
,
(2)2n p=1
2(2n)!
(2n),
(2)2n
n = 1, 2, 3,
n = 1, 2, 3,
(9.92)
9.4 Employing the above relation between Bernoulli numbers

and Riemann zeta function, show that
2
6
(4) =
4
90
(9.93)
6
945
(8) =
8
9450
(9.94)
(2) =
(6) =
(9.91)

KPN Stat Mech

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

KPN Stat Mech

Uploaded by

Copyright:

Available Formats

K. P. N.

Why should we study statistical mechanics ? . . . . . . . . . . . . . . 1

Experiment, outcomes, events, probabilities and ensemble .

2.11 Counting of the number of elements in events of the sample

Binomial, Poisson, and Gaussian . . . . . . . . . . . . . . . . . . . . . . . . . . .

4.16.1 Toy model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58

Closed system : Canonical ensemble . . . . . . . . . . . . . . . . . . . . . . .

Grand canonical ensemble . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91

Quantum Statistics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109

7.7 Grand canonical partition function, grand potential,

Bose Einstein Condensation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135

Statistical Mechanics of Harmonic Oscillators . . . . . . . . . . . . . . 157

Quantum Harmonic Oscillator . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160

Binomial distribution : B(n) =

Two ways of keeping a particle in a box divided into two equal

Average occupation number of a quantum state under

g3/2 () versus . Graphical inversion to determine fugacity . . . . 140

Probabilities calculated from Binomial distribution :

Micro states of of two particles with total energy 2 . . . . . . . . . . . 58

Terms in the restricted sum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112

1.1 Micro-Macro connections

1 Why should we study statistical mechanics ?

engraved on the tomb of Ludwig Eduard Boltzmann (1844-1906) in Zentralfriedhof, Vienna.

1.3 Macro world : Thermodynamics

This relation connects the thermodynamic specific heat at constant volume

This relation shows that fluctuations in the number of particles in an open

1.3 Macro world : Thermodynamics

The laws of thermodynamics can be stated, in a lighter vein as follows.

1 Why should we study statistical mechanics ?

1.4 Micro-Macro synthesis

1.5 What am I going to teach ?

1.5 What am I going to teach ?

1 Why should we study statistical mechanics ?

1.6 Syllabus prescribed

1.7 Books on statistical mechanics and thermodynamics

1.7 Books on statistical mechanics and thermodynamics

A popular book in statistical mechanics. Starts with a beautiful historical

1 Why should we study statistical mechanics ?

Boltzmann transport equation as a starting point. Historically, statistical

1.7 Books on statistical mechanics and thermodynamics

1 Why should we study statistical mechanics ?

Thanks to H S Mani, now at Chennai Mathematical Institute, for bringing my

1.9 Extra reading : Papers

Student-friendly. Shanthini has anticipated several questions that would

1.8 Extra reading : Books

1.9 Extra reading : Papers

1 Why should we study statistical mechanics ?

M C Mackay, The dynamical origin of increasing entropy, Rev. Mod.

2.1 Toss a coin

2.2 Roll a die

Consider throwing of N dice. There are then 6N outcomes. Each outcome

2.3 Pick randomly a molecule from this room

2 Experiment, outcomes, events, probabilities and ensemble

2.4 Sample space and events

2.7 Random variable

2.6 Rules of probability

In the above denotes a null event and is a sure event.

2.7 Random variable

Maxwell and Boltzmann attached probabilities to events in some way; we got

2 Experiment, outcomes, events, probabilities and ensemble

Now consider a continuous random variable defined between a to b with

is also called the mean, expectation, first moment etc.

But then, we are accustomed to calculating the average in a different