Statistical Mechanics

3
chapter

Classical Thermodynamics

empirical subject It describes the thermal

Classical thermodynamics is an

macroscopic systems However

it was largely developed before the
properties of
atomic picture was established and so it is largely independent
firmly
of the precise natureof the microscopic degrees of freedom This aspect
makes classical Thermodynamics a weird field whose laws are applicable

to broad including gases biological systems to black

range of systems
a

holes In a
first two cases we know what the microscopic degrees of
freedom are atoms and molecules and we can derive athough fee

issues are still unsolved the laws

of classical thermodynamics
using
statistical mechanics But in the case of blackholes we have no cln
about the nature
of microscopic degrees of freedom whichgives rise to
The lane of black hole Thermodynamics

StateVariables and statefunctions

A thermodynamic system is specified by a number of variables known as

statevariables For a
fluid the independent state variables can be taken to

be pressure p and volume V Later when we allow particles to change we

 shall also consider particle number N considered to be continuous variable

AN 1 K N 1023

There are functions which are functions of state variables These are
called statefunctions For example for a fluid like
gas liquid at fix
or

N we have internal E E p V where is pressure and is

energy p
volume

The Zeroth law of Thermodynamics

The zeroth law there exists a state functi
of says that
Thermodynamics

O P V such that when two gases are in thermal contact the two system
0 P2V2
eventually reach equilibrium when
0 p Vi T

The numerical value of this state function is called temperature and it

is denoted by T

Comment

1 The main content of the zeroth law is to state that there exists a useful
quantity called temperature

The first law of thermodynamics

The first law of thermodynamics is a statementabout the conservation of

energy
The first law states that theworkrequired to change the state of an isolated system

is independent of hone that work is done AE W

 Px
E P V2

Ellis

comments

1 Thefirst law essentially tells us that a statefunction E E pV There may

exist many ways of doing work on the system pushing or

piston stirring it
a current through it but in the
pulling passing a
the change in E will only depend on the network on the system

if the system is thermally isolated from its environment

2
If the system is not thermally isolated from its environment E p V
even
if no work is done on it This way of
may change changing

E is known as heat Thus we are

forced to generalize
ΔE G W

3 Neither heat nor work is

energy
but a transfer of energy

Quasi Static Process

The first law makes no reference to whether the process of transferring

energy was slow or sudden was violent then inbetween
If process

But the process is

processes p and will be undefined if very
 I
slow and at each step of the process
p
and are approximately we

defined then we can

plot the trajectory
of the system on its p
diagram Such processes are called quasi static processes For
a
quasi static process for an infinitesimal amount
of work done

by the system and an infinitesimal amount of heat added to

the system we write
de IQ JW
Here means an infinitesimal amount It is not an exact

meaning neither todiffe ential

IW nor can be written as If for some

f f pov Infact
for a gas in a reservoir with a piston the
the work done by the has the form
gas
Iw pdv
On the other hand mathematically we have
de
Ep p 3
It can be shown that a state function 5 5 PV known as entrop

which is defined as

5 A

where O is a reference state and A is the state whose entropy

we are assigning The infinitesimal form of the definition of entropo

is ds IQ T
Using
these definition one can write the first law as

de Tds pdV

An Application of the First Law

Joule's Free Expansion Experiment
Thermomet
Thermometer

Water Water

an
datf
IffttfadtmoFfmtg
oossaffosastorastase

y y
adiabatic Walls aliabatic Walls

before After

experiment gas is allowed to expand freely into a vac

In Joule's

chamber We assume the is an ideal It is observed

gas gas
doesn't
empirically that the temperature change Thus 19
0

Since the does no work in the AW 0 Thus we

gas expansion
see that the internal doesn't change AE 0
energy
 91 g
Of the time thermodynamic state variables p T w

can take two to be independent

Taking T we
any
see AV V but AT 0 and AE 0 This implies
the internal energy of an ideal
gas
is a function of Tonk

The Second Law of thermodynamics

The second law of thermodynamics plays an extremely important role in

physics and also in other fields It is believed that the second law is

for the
macroscopic
respons
arrow
of ble time The Schrodingerequation
for an

isolated system is invariant under time reversal but we experience a

fixed arrow
of time in the macroscopic world This emergence of a

macroscopic arrow of time is a

very deep phenomenon
and the transition

from the microscopic realm to the realm involves

macroscopic many interesting
problems This phenomenon is ultimately related to the nature of
quantum gravity and the initial state of the universe

Although modern formulations of the second law uses entropy

it wasn't formulated
origi ally
as such

Formulations
of the second have

Kelvin No process is possible whose sole effect is to convert heat extracted from

a reservoir and convert this heat entirely into work

 I
clausius No process is possible whose sole effect is to transfer heat from a

colder body to a hotter body

It can be shown easily that violating Kelvin's law implies a violationof

Clausius law and vice Versa To derivethe modernform the second of
law we start with the Clausius inequality

011 0

where the equality holds

for reversible processes To derive this inequalito
one uses Clausius theorem which in turn uses Clausius formulation of
the second law

Now suppose there are two paths I and II connecting an initial state
is assumed to be irressible and
A to a
final state B Path I put
It is reversible Then Clausius identity implies

1 so

1 SCA
SCB
I

It is important to note that irreversible process T cannot be

for an

heat and
as
inte
of preted
the temperature
cannot
the body reservoir

ds to
that is
supplying the
so one use T
for irreversible paths
Now if the irreversible path is adiabatic i [Link] 0 then we
get

S B SCA

which means that for an isolated

system entropy can never decrease

Computing Entropy Examples

1 Consider a
large reservoir at temperature Tr and a small system

at temperature Ts They are placed in thermal contact and they both

end at the same temperature Tr The heat absorbed or given
up
by the system is
up
ΔQ C Tr Ts
compute the change in entropy of the reservoir the system and the

universe in the process

i ASR
1 C
Fr 1

TR

Ass
I
ii c c en
49 TE
Ts
iii Asuniverse ASpt AS ME 1

S ASS

Js
AS universe is always 0
universe

TS
Tr
ASR
I
2 An irreversible process The Joule Expansion

Po Vacuum
Before
No

Pf Pf
After Vo Vo

The Joule expansion is process in which a allowed to

gas is suddenly
expand in This example to double its volume The process is
and so no adiab tic
heat is allowed into the system We assume that
its an ideal
gas Although the pressure and volume of thegas
 g
its initial and final states are known it doesn't have wel

defined pressure and volume in its intermediate states

Since no heat was exchanged and no work was done the internal
is
unchanged AE 0 For an ideal the temperator
energy gas
is determined E and so
by AT 0 Thus
for an ideal gas
Pivo Pg avo Pg

Since entropy is astate variable its change only depends on

V and not the take

changes
in
P path taken Thus we can

the same and expand it isothermally in a reversible process

gas

as Sds 51
but since T is E doesnot ideal gas Thus
fixed change for an

we can write do pdV and so

200 2V
As IV
E Renz
Vo Vo
Thermodyamic Potentials
The internal E is example of potential energy Jus
an
energy
as in mechanics the potential on the
energy only depends
position of the particles in thermodynamics the potentials out
depend on the values of the independent state variables

Internal Energy E E S V
The form of the first law dE Tds pdV tells us that
the internal energy E is a natural function of the state
variables S V E E S V and

T and
Es p SE

The internal becomes

particularly useful in processes
energy
which the volume is kept constant isochovic processes In
such processes
one can define the heat capacity at constant

volume Cv as

dE Tds IQ Cudt

Then AE
f at
Ti
 E E s.V is an
example of a
thermodynamic potential
There other combinations E ST which als
are
of p
have dimensions of energy Different thermodynamic potential
areuseful for different setups Below are the other therm
dynamic potentials that are also frequently used

Enthalpy H H S P

Enthalpy is defined as H E pV
Then dH de [Link]
Tds pdv dpV pdv

dH Tds V alp

with T and
Es 3
H H Sop

Enthalpy is useful in processes in which the pressure remains

constant known as isobaric processes In such processes
one can define Cp as

dH Cpdt
Legendre transformations
The transformation H E PV takes us from a function
E SV and changes its dependence on S V to a

function H S p which depends on S and P This is

a reversible transformation in that we can reconstruct

the original function E p This transformation

s V H s
from is

known as the Legendre transformation

fix

In m

se

One needs to specify intercepts notjust slopes

m is the new independent variable f matg

If for each value of m I gave a value
of g
you
the the
you
could reconstruct
f
as
envelope of
straight lines with slope m and intercept
g
g g
m
a 1m a

It shows that the Legendre transformation is a reversible

transformation
 S

Hemholtz Free Energy

Hemholtz free or is the energy that

just free energy
energy
a
system has at constant temperature that can be converted
into work In other words its the part of that is useful
energy
when entropy has been taken into account

F E TS IF TVs pdu Tds sat

dF pdv Sdt
F F V T
with

p PE
s
13

To understant its
interpretation let us consider a
system s

coupled to a reservoir at constant T and fixed total

energy
Stof Etf Sp Etp Es Sg Es
Sp Ett Es Ett Ss Es
22
 Sp Ept Ss Es
Ef
Sp Ett
7s Astop 30 AF 0

that the the second law that

Thus consequence of
we see is

free energy decreases even though total energy is conserved

From dF [Link] when T constant one

gets
V2 Fz
So pdv IF F2 F W
F

The ability
of a system to do work is given by the free
energy

Gibbs Free Energy

In
deriving the usefulness of the Hemholtz free energy we

kept both volume and pressure of the subsystem constant

If we allow volumeto vary while keeping p T constant
we
get
Stat tot E tot Sp tot V Eft E S E V
 2
Sp Vit Ent S Esv
1,5 281mV

Splutt Eta V S
E
g
E PV TS
T

So at fixed T the thing that will be minimized is

E TS PV H TS
G

when number of particles is added to the mix we get

dF Sdt partan dG SdT Vdp udN

G G PT N
of p T N
only N is extensive

And thus we have G P T N M [Link]

Maxwell's Relations

For have
any function f x
g we
22 e

One apply this identity to thermodynamic potentials and

can

obtain new relationships between physical quantities For example

from E E S V and

E
E 2 V25

T and
p Ev
one
gets
Et Es

similarly from F F T V and FE

E OVOT
p E t

s 13 v

E
 From H H SP and
yp s

T and v
24

Ep E

and
9 Itp
Finally from G G P T
2

s
3 8
Ep Et
Comment

1 The Maxwell relations useful in deriving consequences

are

about a
quantity that might be difficult to
compute or
measure from quantities which are
easy to compute or

measure
The Third Law
The second law deals with the change of entropy But
it does not us an absolute
give
measure
for it Ludwig
Boltzmann discovered that thermodynamic entropy 5 of
a
system at a is related to the
given energy
number that
of microscopic configurations
2 is
with that coenergy
sistent
and other macroscopic
quantities
S KB loger E T V N

This implies that there should be an absolute scale

for entropy The third law provides such an absolute

scale
Lim S T 0 Nernst's Law
T O

Its T 0 the system to its ground state For

goes
sy tems
with degenerate grounds states i.e more than one

configurationfor E this law can be relaxed to

given

him 5
o as N 0

T 0
where N is the number of particles or
degrees of
fre dom
From the Boltzmann formula we see that 5 0 we
for
need 2 1 Quantum mechanical systems tend to have

uniqueground states Even if one tunes the parameter of

a
quantum Hamiltonian to have a degenerate ground
state to interaction
any slight perturbation due
an

with the environment ruins the degeneracy

In contrast we consider a classical system the change

if
in its entropy is
given by

SCB SCA Cudt

A
T

For ideal we
an
gas get

Cv NKB

solid
or
for a one
gets
Cv 3N KB Dulong Petit

that not
For both go to
Thes cases we does
see entropy zero
 1
T Thus the third law sharp contrast with
as so is in

classical physics The Third law reflects our understanding

the vacuum state a out
of quantum mechanical systems

Example The ferromagnetic vacuum of a 1 D Heisenberg spin

chain I spin chain consists of N particles at N sites
where each can take the
particle value 1 or 1
spin
only to interact with its nearest neighbon
Each spin is assumed

We assume periodic boundary condition The interaction

is
energy function Hamiltonian given by

Hamiltonian H Jois JSO

I for

where Ti It represents the spin of the site i

For J 70 we see that lowest

energy
state consists
of either
all spin up 1 or all spin down 1

or
This is an example of quantum system for which
a

5 KB log 2 at T O but as T O and N 0

o
I
Consequences of the Third Law

1 Heat capacities to zero as T 0 For

example if we
go
take Cp T 2 we can rewrite this as
Cp
Its T o
logT
0 Then according to the

third law him T 0 him As

5 0
gives 4 0
To

Thermal expansion stops Let β 2 how the volume

2
f measure of a

changes with changing temperature at constant This called

gas pressure

the isobaric expansivity Now since o as s o and T o we

sp
get from using one of the Maxwell relations that
f o as T 0

so we see that thermal expansion stops at low temperature

3 We have already seen that the ideal law breaks down

gas
 4 It take an infinite number of steps to reach T 0

We shall not prove this rigorously Let us consider a hypothetical situation in which

go to zero Let two

the entropy doesn't as T 0 us consider curves in the

ST plane for two different values X X2 of some parameter Then

it is possible to get to T 0 in a
finite number of steps involving

decrease in via adiabatic steps and increase in via isotherm

steps 5

On the other hand if all the curves forfixed values of X converge as T 0

the steps getting us closer and closer to T 0 become smaller and smaller

and as a result we need an infinite number of steps to get there

Entropy of Mixing

Consider two different types

of gasses contained in two chamber with

xV and 1 V vol mes

where 04 1 which are connected
by a stopp

er We assume that both gasses are at the same temperature T and

pressure p

Now if we the stop the two gasseswill mix freely but neither the pressure nor
open

the temperature will change We are

assuming our
container is thermally insulated

But the volume of both gasses will increase And so the change in
will be given by
ent opy n
of
AS 5 an edva molecules

V nx fraction of
gast
a 1
[Link] n HX
nKBx
v gas
nkB xlux 1 x lu 1 x
AS A

1
0 I
Comments

1 Entropy of mixing increases due to the distinguishability of the two gasses

Reveals informational aspect

of entropy

2 AS is maximum when AS n KB ln2 KB lu 2N

where
Interpretation S KBlog r r is the number of possible

With x configuratof on
for every position in one the chambers none there is

a corresponding position in the other chamber So the number possible

change conf
by for gurations
2 N molecules Hence AS KB log 2

Microscopic Viewof Entropy and Maxwell's Demon

two thermally insulated chambers

Consider
of equal volume with stop inbetween

Now one of the chambers is filled with gas and the other is empty If we n'one

open
the stop the
gas will expand into
the new chamber irreversibly and the

pressure will dropto half of its initial value As we save in the previous

lecture the entropy will also increase

Now imagine a small demon or a computer located at the stop which controls

the stop If it sees a moleculefrom chamber 2 hurtling towards chamber 1

then it opens the stop and lets it through otherwise it keeps the stop closed

Thus we see that the existence ofsuch a demon can turn an irreversible process

into a reversible one

But allowing molecules with higher KE to seepinto one of the chambers the demon

transfer heat without doing any work on the

can
gas and thus violate
the second law
of thermodynamics
As should be obvious the demon can also decrease the entropy
of mixing
without work
any
So what is the solution to this paradox

To solve This we needto realize that the demon has to store information about
the molecules inits
memory or in some kind of information storage device When

the demon gets new information it then has to erase some of the old

This act of erasing info mation

generates heat and causes the entropy of the total
system gas demon to increase And so the laws of thermodynamics are

saved