Statistical Mechanics

Advanced Physical Chemistry

What is Statistical
Thermodynamics?
 Link macroscopic behavior to atomic
/molecular properties
 Calculate thermodynamic properties from
“first principles”

(Uses results for energy levels etc. obtained from

quantum mechanical calculations.)
Postulate of Statistical Mechanics
 The thermodynamic state of a large
macroscopic system is to separate into
tiny or microscopic parts. These parts are
called microsystems. The state of each
microsystem is called microstate.

 All of the microstates of the system

combine statistically to generate the
overall state called the macrostate.
Ensemble
 Ensemble = a collection of an undetermined
number of microsystems that collectively make - up
the macroscopic system

 Each microsystem in the ensemble has its own

characteristics
1.) number of particles
2.)energy
3.)volume
4.)Pressure
5.)Temperature
Canonical Ensemble
 Ensemble separated into j individual
microstates, such that the number of
particles in each microstate Nj, the
volume of the microstate Vj and the
temperature of the microstate Tj are the
same.

 N = ΣjNj = jNj
 V = Σ V = jV
j j j
T = T
j
The Most Probable Distribution
 The number of ways to distribute particles
among the energy levels of the
microstates

 Ω = (N! / Πj Nj!) ( Πgjnj)

Assumption
 No gas particles are in the energy state εo
N1 gas particles are in energy state ε1
N2 gas particles are in energy state ε2 and
so on.

Let : W = number of ways of arranging the

right number of particles in specified energy
levels.

W = N!/ N1! N2! N3!

Assumption
 If there are N particles with degeneracy gj,, the
number of ways the Nj particles can be
arranged among degenerate state is gjNj

 Wdeg = total number of arrangements due to

degeracies

 Wdeg = goNo g1N1 g2N2 gN3 = ΠjgjNj

Ω = W▪Wdeg
Ω = (N! / Πj Nj!) ( Πgjnj)
Sample Problem
Assume that you have a three – particle
system that has four possible energy. Your
system has a total 5 energy units to
distribute among particles. How many
different distinguishable distribution can
there be?

W(1,0,1,1) = 6
W(0,1,2,0) =3
W(0,2,0,1) =3
The Most Probable Distribution :
Constraints of this equation
 E = Σj Njεj

i.) the total energy of the system must be equal to

the energy of each quantized state εj times the
number of particles in that energy level Nj

N = Σj Nj

ii.) the sum of all Nj values must equal the total

number of particles in the system
Lagrance’s Method of Undetermined
Multiplier
 When the number of particles and possible arrangements
gets larger and larger the plots gets progressively narrower
  this implies that for large N , only the most probable
j
distribution needs to be considered.
 rather than maximize Ω, instead maximize
In Ω.

In Ω = In (N! / Πj Nj!) ( Πgjnj)

In Ω = In N! + Σj(In gjNj – In Nj!)
In Ω = N In N – N + Σj( Nj In gj – Nj InNj + Nj)
In Ω = N In N –N+ ΣjNjIn gj – ΣjNj InNj + ΣNj)
In Ω = N In N + ΣjNjIn gj/Nj
 3.5 25
A 3 A 6
3
20
2.5

15
2
W

W
1.5
10

1
5
0.5

0 0
0 1 2 3 0 1 2 3 4 5 6
a1 a1

1000 1.4 1014

A  12 1.2 10
14 A  50
800
1 1014

600 8 1013

W
W

13
6 10
400

4 1013
200
13
2 10

0 0
0 1 2 3 4 5 6 7 8 9 10 11 12 4 14 24 34 44
a1 a1
Lagrance’s Method of Undetermined
Multiplier
 In order to max InΩ in terms of two constraint
i.) combine the equation for In Ω,N and E into a
single linear combination
ii.) maximize the resulting three term

InΩ + αN – βE = 0
N In N + ΣjNjIn gj/Nj + α ΣjNj - β ΣjNj Ej
 Take the derivative of this expression with
respect to the Nj values

∂/∂Nj [N In N + ΣjNjIn gj/Nj + α ΣjNj - β ΣjNj Ej] = 0

Boltzmann Distribution Equation
 In gj/Nj + α – βεj = 0
In gj/Nj = - α + βεj
In Nj/gj = α – βεj
Nj/gj = e α – βεj
Nj = gjeα e- βεj
ΣjNj = N = Σj gjeα e- βεj
N = eα Σj gje- βεj
where: q = Σj gje- βεj (molecular partition function)
the fraction of the total particles in that energy;

Nj / N = gjeα e- βεj / eα Σj gje- βεj

= gj e- βεj / q
Boltzmann Distribution Equation
 The ratio of the population of the ith
energy level to the population of the kth
energy level
(Ni/N)/(Nk/N) = [(1/q)gie-βEi] / [(1/q)gke-βEk]
Ni/NK = gi/gk e-β(Ei – Ek)
Ni/NK = gi/gk e-βΔE = Boltzmann Eqn
Sample Problem
 Consider the previous problem. Assuming
that the energy levels are fourfold
degenerate ( that is gj = 4) and that the
energy levels have values of 0.00, 1.00,
2.00 and3.00 x 10-21J, what is the value
for q if the energy levels are 0.00,
1.00,2.00,3.00 x10-21J.

 Ans.11.524
The Boltzmann Factor
The Boltzmann factor tells that if system
has states with energies E1, E2,E3....., the
probability pj that the system will be in
the state with energy Ej depends
exponentially on the energy of that state.

pj α e- Ej/KT
pj = e-Ej/KT/Σi e-Ej/KT = e- βEj / Σi e-βEj
The summation in the denominator is
required to normalize pj so that Σpj = 1
Canonical Partition Function
 pj = e- βEj/Q or
pj (N,V,β) = e-βEj (N,V) / Q(N,V,β)

pj (N,V,T) = e-βEj (N,V) / Q(N,V,T)

where: Q = canonical partition function

Q(N,V,β) = Σj e- βEj (N,V)

The Average Ensemble Energy is
equal to the observed Energy of a
System
 <E> = Σj pj E(N,V)
(N,V,β)

= ΣjEj(N,V) e – βEj (N,V) / Q (N,V,β)

where <E> is a function of N,V, β

 Differentiate ln Q
(N,V,β) with respect to β

(∂ lnQ /∂β)N,V,β = 1/Q(N,V,β) (∂Σe-βEj (N,V)/ ∂β)

= 1/Q(N,V,β) Σj [-Ej(N,V)] e-βEj (N,V)

= - Σj Ej(N,V)e-βEj (N,V) / Q (N,V,β)

<E> = - (∂ lnQ /∂β)N,V

The Average Ensemble Energy is
equal to the observed Energy of a
System
 <E> can also be expressed as a function of a
temperature

 ∂f/∂T =(∂f/∂β)(∂β/∂T ) = (∂f/∂β)d/dT (1/KT)

= - 1/KBT2(∂f/∂β)
(∂f/∂β) = -KBT2 (∂f/∂T)

<E> = KBT2 (∂In Q/∂T)N,V

The Heat Capacity at Constant Volume
 The temperature derivative of the average energy

 Cv = (∂U/∂T)N,V = (∂ <E>/∂T)N,V

 CV (molar) = 3/2 R ( monoatomic

ideal gas )

Cv = 5/2R + R (hv/kT)2 [e-hv/kT/(1- e-hv/kT)2]

(diatomic ideal gas)
Pressure in terms of a partition
Function
 Pressure of a macroscopic system
 Pj(N,V) = - ( ∂Ej / ∂V)N

<P> = Σj pj (N,V,β) Pj(N,V)

= Σj pj (N,V,β) - ( ∂Ej / ∂V)N

= Σj -(∂Ej / ∂V)N [e – βEj (N,V) / Q ]
(N,V,β)

 Can be written in a more compact form

Q(N,V,β) = Σj e- βEj (N,V)
(∂Q/ ∂V)N, β = -βΣj (∂Ej/∂V)N e – βEj (N,V)
Pressure in terms of a partition
Function
1/β (∂Q/ ∂V)N,β = -Σj (∂Ej/∂V)N e – βEj (N,V)

<P> = 1/β [(∂Q/ ∂V)N,β / Q ]

(N,V,β)

<P> = [kT/Q(N,V,β)] (∂Q/ ∂V)N,β

<P> = kT (∂ In Q/∂V)N,β
Ideal –Gas Equation of State
 Consider a monoatomic gas
QN,V,β) = [q (V,β)]N / N!
where: q(v,β) = ( 2πm/h2β)3/2 V  to be derive
later
In Q = N In q – In N!
In Q = 3/2N In (2πm/h2β) + N InV – In N!
(∂InQ/∂V)N,β = N/V

<P> = kT (∂InQ/∂V)N,β
= NkT/V
P = RT/V
Sample Problem
 Calculate the equation of state associated
with the partition function
 Q (N,V,β) = 1/N! (2π/h2β)3/2(V –Nb)N

eβaN2/V
Partition Function of a System of
Independent, Distinguishable
molecules
 is the product of Molecular partition function (q)
 Consider the total energy of the system
 Ej(N,V) = εia(v) + εlb(v) + εkc(v)  N terms
 Q(N,V,T) = Σj e-βEj
=Σi,l,k e-β(εia + εlb + εkc + …..)

Since the particles are distinguishable and

independent, we can sum over i,l,k independently
in which case Q(N,V,T) can be written as a product
of individual summations
Q(N,V,T) = Σi e-βεia Σi e-βεlb Σi e-βεkc
Molecular Partition Function
Q(N,V,T) = qa(V,T) qb(V,T) qc(V,T)
where qa,qb and qc = molecular partition functions
q(V,T) = Σj e- βεj = Σj e- εj/kT

If the atoms are distinguishable

Q(N,V,T) = [q(V,T)]N

If the atoms are indistinguishable

Q(N,V,T) = [q(V,T)]N /N!

Molecular Partition Function
 Independent Distinguishable atoms /molecules

 Example: The Einstein Model of atomic

crystals where atoms are fixed at lattice sites.

 The sublimation energy per atom at 0K.

 Q = [ e-βu ( e-βhv/2/ 1 – e- βhv)3]N

Sample Problem
In 1905, Einstein proposed a simple model for an atomic crystal that
can be used to calculate the molar heat capacity. He pictured an
atomic crystal as N atoms situated at lattice sites, with each atom
vibrating as a three –dimensional harmonic oscillator. Because all
the lattice sites are identical, he further assumed that each atom
vibrated with the same frequency. The partition function associated
with this model
Q = e-βUo ( e-βhv/2 / 1 – e- βhv)3N
where v is the frequency with which the atoms vibrate about htier
lattice positions and Uo is the sublimaiton energy at 0K. Calculate
the molar heat capacity of an atomic crystal from this partition
function.
Hint: U = - (∂ InQ/ ∂β)N,V
Cv = (∂U/∂T)N,V = -1/kT2 ((∂U/∂β)N,V

Cv(m) = 3R (hv/kT)2 ( e- hv/kT / 1 – e-hv/kT)2

The Translational Partition Function of
an atom in a Monoatomic Ideal Gas
 The energy of an atom in an ideal
monoatomic gas
 εatom = εtrans + εelec

 q(V,T) = qtrans (V,T) q elec (T)

Evaluation of the translational Partition

function in a cubic container
Enx,ny,nz = h2/8ma2 (nx2 + ny2 + nz2)
qtrans = Σnx,ny,nz =1 e -βEnxnynz
q trans (V,T)
qtrans = Σnx,ny,nz =1 e –βEnxnynz
= Σnx=1Σny=1Σnz=1 e- βh2/8ma2 (nx2+ ny2+nz2)
because: ea+b+c = ea eb ec

qtrans = Σnx =1 e –βEnx Σny=1 e –βEny Σnz=1 e –βEnz

qtrans(V,T) = [ Σn =1e- βh2n2/8ma2]3

 An excellent approximation to replace the

summation is to use an integration
q trans (V,T)
 q trans (V,T) = ∫(0 to ∞) (e-βh2n2/8ma2 dn)3

= ∫(0 to ∞)e-αn2dn = (π/4α)1/2

qtrans (V,T) = (2πmkT/h2)3/2 V

 The average translational energy of an

ideal –gas can be calculated from this
partition function
<εtrans>
 < εtrans> = kT2 ( ∂ In qtrans / ∂T)V

= kT2( ∂/∂T [3/2 InT + terms

ind. of T])V
< εtrans> = 3/2 kT
Most atoms are in the Ground
Electronic State at Room Temperature
 qelec = Σj gej e-βεej

Where: gej = degeneracy

εej = energy of the jth electronic level
 To find the expression for qelec,
i.) fix the arbitrary zero of energy,εe1 = 0
ii.) measure all electronic energies relative
to the ground electronic state
qelec
 qelec (T) = ge1 + ge2 e-βεe2 + …… (first two terms
are sufficient)

iii.) ε’s are typically of the order of tens of

thousands of wave numbers. Use 1 cm-1 = 1.986
x 10-23J

βεelec = 40,000 cm-1/ 0.6950 cm-1/K

= 104K/T
 Typically qelec is around 10-5 for most atoms at
ordinary temp., so only the first term in the
summation for qelec is significantly different from
zero.
Some Atomic Energy Levels
Atom Electron Term Degeneracy Energy
configuration Symbol cm-1

He 1s2 1
So 1 0
1s2s 3
S1 3 159850
1
So 1 166271

F 1s22s22p3 2
P3/2 4 0
2
P1/2 2 404
1s22s22p43s 4
P5/2 6 102406
4
P3/2 4 102681
2 102841
4
P1/2
qelec
 Calculate the fraction of Helium atom in the
first triplet state 3S1. The fraction is given by

 F2 = ge2 e-βεe2 / ge1 + ge2 e-βεe2 + ge3 e-βεe3

 F2 = 3 e-βεe2 / 1+ 3 e-βεe2 + 1 e-βεe3

 at 300K, βεe2 = 770, so f2 = 10-34

 at 3000K, f = 10-33
2
<εave>
 Q(N,V,T) = (qtrans qelec)N / N!

qtrans (V,T) = (2πmkT/h2)3/2 V

qelec(T) = ge1 + ge2 e-βεe2

Uave = kT2(∂ InQ/∂T)N,V

= NkT2((∂ In q/∂T)V
= 3/2 NkT + Nge2εe2e / qelec
–βεe2
The energy of a Diatomic Molecule can
be approximated as a sum of separate
terms
 The total energy of the molecule
εtotal = εtrans + εrot + εvib + εelec

q(V,T) = qtransqrotqvibqelec
Q(N,V,T) = (qtransqrotqvibqelec)N/ N!

Diatomic Molecule = a rigid rotator –harmonic oscillator

approximation was used.

 Allows to separate the rotational motion from the vibrational

motion of the molecules so we can threat each separately
The energy of the ground Electronic
State
For diatomic molecule

qtrans (V,T) = [ 2π(m1 + m2)kT/h2]3/2 V

qelec = ge1 eDe/kT + ge2 e-εe2/kT

 The ground and first electronic states as a function

of the inter nuclear separation.

 εe1 = - De
Most Molecules are in the Ground
Vibrational State at room Temperature
 The vibrational energy levels relative to the
bottom of the internuclear potential well.
 εv = (v + ½) hv with v = (k/μ)1/2 /2ħ

 qvib (T) = Σv e- βEv = Σv=o e-β(v+1/2)hv

= e-βhv/2 Σv=0 e-βhv2
Σn=0 xn = 1/1 – x
with x = e-βhv < 1
Most Molecules are in the Ground
Vibrational State at room Temperature
 Σv =0 e-βhv2 = Σv=0 ( e-βhv)v = 1/1 – e-βhv
qvib(T) = e-βhv/2 / 1 – e-βhv

 Partitionfunction for the rigid rotator –

harmonic oscillator model of an ideal
diatomic gas.
 where : Θvib = hv/k  vibrational energy

qvib (T) = e-Θvib/2T/ 1 – e-Θvib/T

 Average Vibrational Energy from qvib(T)
 <Evib> = NkT2 ( dIn qvib / dT)

= Nk( Θvib/2 + Θvib / eΘvib/T -1)

and for vibrational contribution to the molar
heat capacity

Cvib(molar) = d <Evib>/dT
= R(Θvib/T)2 (e-Θvib/T/ 1 – e-Θvib/T)2
Sample Problem
The rigid rotator –harmonic oscillator model of an
ideal diatomic gas, has a partion function given
by;
Q(N,V,β) = [q(v,β)]N /N!
Where: q(v,β) = (2πm/h2β)3/2 V(8π2I/h2β)
(e-βhv/2 / 1 – e-βhv)
Use this partition function to calculate the average
energy of one mole of a diatomic molecule, given
by the expression below. Derive this expression
Um = 3/2RT + RT + NAhv/2 + NAhv e-βhv/ 1 –e-βhv
Sample Problem
 Use the partition function of a diatomic
molecule to calculate <εvib>

 Hint: use qvib(T) = e-βhv/ 1 –e-βhv

<εvib> = hv/2 + hv e-βhv /1 –e-βhv

Most Molecules are in Excited Rotational
States at Ordinary Temperature
 The energy levels of a rigid rotator
εj = ħ2 J(J+1)/ 2I
Each energy level has a degeneracy of

gj = 2J + 1
qrot (T) = Σj=0 (2J+1)e-βħ2J(J+1)/2I

introduce a quantity called rotational temp Θrot

Θrot = ħ2/ 2Ik = hB/k where B = h/8π2I

qrot (T) = Σj=0 (2J+1)e-Θrot J(J+1)/T

can be written as
qrot (T) = ∫ (0 to ∞) (2J+1)e-Θrot J(J+1)/T
dj
Most Molecules are in Excited
Rotational States at Ordinary
Temperature
 To evaluate this integral
Let x = J(J+1)
dx = (2J +1) dj
qrot(T) = ∫(0 to ∞) e-Θrot x/T dx
= T/ Θrot Θrot << T
= 8π2IkT/h2
 Partition function for the rigid rotator –
harmonic oscillator model of an ideal
diatomic gas
<εrot>
 <εrot> = NkT2 ( d In qrot / dT)
= NkT
 The rotational contribution to the molar
heat capacity

Cv,(rot) (molar) = R
Rotational Partition Function contains a
Symmetry Number for a Homonuclear diatomic
Molecule
 qrot(T) = T / 2Θrot

 Where: 2 = factor that comes from the additional symmetry

of the homonuclear diatomic molecule ( there is a two –fold
axis of symmetry perpendicular to the internuclear axis).
 qrot (T) = T / σ Θrot

 σ = symmetry number of the molecule

(the number of indistiguishable orientations of the
molecule.)

 σ = 1 for a heteronuclear diatomic molecule

 σ = 2 for a homonuclear diatomic molecule
Rotational Partition Function contains a
Symmetry Number for a Homonuclear diatomic
Molecule
 The molecular partition function of a
diatomic molecule for the rigid rotator –
harmonic oscillator approximation.

 q(V,T) = qtransqrotqvibqelec
= (2πMkT/h2)3/2 V ( T/σΘrot)
( e-Θvib/2T / 1 –e- Θvib/T) ( ge1e De/kT)
The Vibrational Partition Function of a
Polyatomic Molecule
 is a product of Harmonic Oscillator Partition
function for each normal coordinate.

 For Polyatomic Molecules

 Q(N,V,T) = (qtransqrotqvibqelec)N/ N!

where qtrans(V,T) = [2πMkT/h2]3/2 V

M = total mass of the molecule
The Vibrational motion of a Polyatomic
Molecule can be expressed in terms of normal
Coordinates.
 εvib = Σjα =1 ( vj + ½)hvj vj = 0,1,2 …..

 where vj = the vibrational frequency

associated with the jth normal mode

 α = number of vibrational mode degrees of

freedom
 Linear molecule ( 3n – 5)
 Non – linear molecule (3n -6)
 Normal modes
 qvib = Πj=1α e –Θvib/2T / (1 – e- Θvib/T)

 Evib = Nk Σαj =1 (Θvibj /2) +

Θvibj e- Θvibj/T / 1 – eΘvibj/T

Cv(vib)= Nk Σαj =1 [(Θvibj /2)2 +

Θvibj e- Θvibj/T / (1 – eΘvibj/T)2

where Θvib = characteristic vibrational temperature

The form of the Rotational Partition
function of a Polyatomic Molecule
 Depends upon the shape of the molecule

 Linear Polyatomic Molecule

 The energies and degeneracies of a linear
polyatomic molecule are the same for a diatomic
molecule
 εj = J(J+1)h2/ 8π2I with J = 0,1,2,3
gj = 2J +1
I = Σnj =1 mjdj2  dj is the distance of the jth
nucleus from the center of
mass of the molecule
Rotational Function of a Polyatomic
Molecule
 qrot = 8π2IkT/ σh2 = T / σΘrot

 The rotational properties of nonlinear

polyatomic molecules depend upon the
magnitude of their principal moments of
inertia

 Θrot,j = ħ2/ 2Ijk where j = A,B,C

Types of Rotation
 Θrot,A = Θrot,B = Θrot,C Spherical Top
 Θrot,A = Θrot,B ≠ Θrot,C Symmetric Top
 Θrot,A ≠ Θrot,B ≠ Θrot,C Asymmetric Top

The quantum – mechanical problem of a

Spherical top  can be solved exactly
 εj = J(J+1)h2/ 2I

 gj = (2J +1)2 where : J = 0,1,2

Quantum –mechanical Problem of a
Spherical Top
 qrot(T)= ΣJ= 0 (2J+1)2 e-h2 J(J+1)/2IkT

 For almost spherical top molecule

Θrot << T at ordinary T.
 qrot(T)= 1/σ ∫(0 to ∞)(2J+1)2e- Θrot J(J+1)/T dJ

 Let Θrot / T = a

= 4/σ ∫(0 to ∞) x2 e-ax2dx

= 4/σ ( 1/4a ( π/a)1/2
q(rot)
 qrot(T) = π1/2/σ ( T/ Θrot)3/2  Spherical Top

 qrot(T) = π1/2/σ ( T/ ΘrotA) ( T/ ΘrotC)1/2

 Symmetric Top

 qrot(T) = π1/2/σ ( T3 / ΘrotA ΘrotB ΘrotC)1/2

 Asymmetric Top
Calculated Molar heat Capacities are in
very Good Agreement with Exp.Data
 For a linear Molecule

 q(V,T) = (2πMkT/h2)3/2 V (T/ σΘrot)

{Π3n-5j=1[ e –Θvibj/2T / (1 – e- Θvibj/T)]}
ge1eDe/kT

The energy is
U/NkT = 3/2 + 2/2 + Σ3n-5j=1 {Θvibj/2T + [(Θvibj/T) /
eΘvib/T-1]} – De/kT

Cv/Nk = 3/2 + 2/2 + Σ3n-5j=1 {(Θvibj/T)2

(e-Θvibj/T) /(1–e -Θvib/T)2}
 Calculated Molar heat Capacities are in
very Good Agreement with Exp.Data
 For a Non -linear Molecule

 q(V,T) = (2πMkT/h2)3/2 V (π1/2/σ)

(T3/ΘrotAΘrotB ΘrotC)1/2
{Π3n-6j=1[ e –Θvibj/2T/(1 – e- Θvibj/T)]} ge1eDe/kT

The energy is
U/NkT = 3/2 + 2/2 + Σ3n-6j=1 {Θvibj/2T + [(Θvibj/T)/
eΘvibj/T-1]} – De/kT

Cv/Nk = 3/2 + 2/2 + Σ3n-6j=1 {(Θvibj/T)2

(e-Θvibj/T) /(1–e -Θvib/T)2}
First Law of Thermodynamics
Work and Heat have a Simple Molecular Interpretation
 The average energy of a macroscopic
system

 U = Σj pj(N,V,β) Ej (N,V)

with pj(N,V,β) = e-βEj(N,V)/Q(N,V,β)

Differentiate this equation

dU = Σj pjdEj + ΣjEjdpj
Work and Heat have a Simple Molecular
Interpretation
 Bec.Ej = Ej(N,V), dEj is the change of E due to
a small change in the volume, dV.

 dEj = ( ∂Ej/ ∂V)NdV

dU = Σjpj (∂Ej/∂V)NdV + Σj Ej dpj

dU = Σjpj(N,V,β) (∂Ej/∂V)NdV + Σj Ej(N,V) dpj
Work and Heat have a Simple
Molecular Interpretation
 Comparing this equation with the macroscopic
equation:
 dU = δwrev + δqrev

 δwrev = Σjpj(N,V,β) (∂Ej/∂V)N dV

 δqrev = ΣjEj(N,V)dpj

 δwrev = -PdV
= - Σjpj(N,V,β) (∂Ej/∂V)N

P = - <(∂E/∂V)N>
Entropy can be expressed in terms of
Partition Function
 Recall: U = kT2(∂ InQ/∂T)N,V
= - (∂ InQ/∂β)N,V

P = kT (∂ InQ/∂V)N,T

S = k In W = Boltzmann Equation for S

Sensemble = k In (A!/ πjaj!)
= k InA! - kΣInaj!
= k AInA – kA –kΣjajIn aj + k Σjaj
= k A In A – kΣjajIn aj
Sample Problem
Use the fact that
Ω(E) = c(N) f (E)VN
For an ideal gas to show that the change in
entropy (per mole) when two gases are mixed
isothermally is given by
ΔmixSm/R = -y1Iny1 – y2In y2
where : ΔmixSm = Smix – S1 – S2
= k In Ωmix/ Ω1 Ω2
the quantities Ω1 and Ω2 are given by
Ω1 = c(N1) f (E1)V1 N1
Ω2 = c(N2) f (E2 )V2 N2
Entropy can be expressed in terms of
Partition Function
 Where: Σaj = A and Ssystem= Sensemble /A

 The probability of finding a system in the jth q system

 aj = Apj
 Sensemble = k A In A – kΣj pjA In pjA
= k AInA – kΣj pjA In pj - kΣj pjA InA

where: Σj pjA InA = A InAΣjpj = A In A

and Σjpj =1

Ssystem = -kΣj pjA In pj /A

= -kΣj pj In pj
Entropy can be expressed in terms of
Partition Function
 To derive an expression for S in terms of Q(N,V,T)
 Recall: Pj (N,V,T) = e-βEj (N,V) / Q(N,V,T)
Ssystem= -kΣj pj In pj
S = -kΣj e-βEj / Q (-βEj – In Q)
= βk Σj Eje-βEj / Q + kInQ/Q Σj e-βEj

S = U/T + k InQ
Since : U = kT2 ( ∂ InQ/∂T)N,V

S = kT ( ∂ InQ/∂T)N,V + k In Q
Entropy can be expressed in terms of
Partition Function
 Recall :for ideal gas where all atoms are in
their ground state electronic state

 Q(N,V,T) = 1/N!(2πmkT/h2)3/2VN ge1

 The molar entropy of one mole of a
monoatomic ideal gas
 S(molar) = 3/2R + R In [(2πmkT/h2)3/2
Vge1] –k In NA !
Entropy can be expressed in terms of
Partition Function
 Applying Stirling Approximation:
- kIn NA! = -k NA In NA + kNA

 = - R In NA + R

S(molar)= 5/2R + R In [(2πmkT/h2)3/2 Vm ge1/NA]

Sample Problem
Calculate the molar entropy of Argon at
298.15K and 1 bar and compare your
result with the expt value of 154.8J/Kmol
 S = kT ( ∂ InQ/∂T)N,V + k In Q
Or S = Nk [T(∂ Inq/∂T)V + In q]
S= Nk [T(∂ Inq/∂T)V + In q/N +1]
Further simplified
S = Nk {[ In (2πmkT/h2) ▪kT/P] +5/2}
 Sackur –Tetrode Equation
Sample Problem
 What is the entropy of 1 mole of He at
25oC and 1.00 atm pressure using Sackur
–Tetrode equation. Compare this with the
exptal value = 126.04 J/molK.
The Molecular formula S = k InW is
analogous to the dS = δqrev / T

Recall : Ssystem= -kΣj pj In pj

(dS/dpj) = -kΣj (dPj + Inpjdpj)
Σdpj = 0 and Σpj =1
dS = -k Σj In pj dpj
dS = -k Σj [ -βEj – In Q] dpj
ΣIn Q dpj = In Q Σjdpj = 0
dS = βk Σj Ej dpj
Since: δqrev = ΣjEj(N,V)dpj
dS = βk δqrev and βk =1/T
dS = δqrev / T
Practical Absolute Entropies of Gases
from Partition Function
 Recall:S = kT ( ∂ InQ/∂T) N,V + k In Q

where: Q(N,V,T) = Σj e- Ej (N,V)/kT

S = k In Σj e- Ej/kT +(1/T)Σj Ej e-Ej/kT/Σe-Ej/kT

to study behavior of this equation as T  0.

i.) assume the first n states have the same energy, E 1 =

E2 =E3 (the ground state in n – fold degeneracy
ii.) the next m states have the same energy, E n+1 = En+2 =
En+m (the first excited state in m- fold degeneracy)
Absolute Entropies
 Σj e-Ej/kT = ne-E1/kT + me-En+1/kT + ……

By factoring out e-E1/kT

Σj e-Ej/kT = e-E1/kT [n + me-En+1 – E1/kT + ……]
But En+1 – E1 > 0
e-En+1 – E1/kT  0 as T  0
Therefore as T 0
Σj e-Ej/kT  ne-E1/kT
Absolute Entropies
 Applying the entropy equation

 S = kInΣj e- Ej/kT+(1/T)Σj Eje-Ej/kT/Σe-Ej/kT

= k In n – E1/T + E1/T
= k In n
As T  0, S is proportional to the logarithm of
the degeneracy of the ground state
Recall: Q(N,V,T) = [q(V,T)] N / N!
Absolute Entropies
Monoatomic ideal gas equation
q(V,T) = (2πMkT/h2)3/2 V ge1

Diatomic ideal gas

q(V,T) = (2πMkT/h2)3/2 V ( T/σΘrot)
( e-Θvib/2T / 1 –e- Θvib/T) ( ge1e De/kT)

Linear Polyatomic ideal gas

q(V,T) = (2πMkT/h2)3/2 V (T/ σΘrot)
{Π3n-5j=1[ e –Θvibj/2T / (1 – e- Θvibj/T)]} ge1eDe/kT

Non -linear Molecule

q(V,T) = (2πMkT/h2)3/2 V (π1/2/σ)

(T3/ΘrotAΘrotB ΘrotC)1/2
{Π3n-6j=1[ e –Θvibj/2T/(1 – e- Θvibj/T)]} ge1eDe/kT
Absolute Entropies
 Recall:S = kT ( ∂ InQ/∂T)N,V + k In Q
Q(N,V,T) = [q(V,T)] N / N!

S = NK Inq – k InN! + NkT (∂ In q /∂T)V

Using Stirling Approximation:

S = Nk Inq - Nk InN + NK + NkT(∂In q/∂T)V

S = Nk + Nk In[q /N]+ NkT(∂In q/∂T)V
(V,T)
Standard Molar entropy
 Calculate the standard molar entropy of
N2(g) at 298.15 K

S = Nk + Nk In[q (V,T) /N]+ NkT(∂In q/∂T)V

So/R = In [(2πMkT/h2)3/2 ] + In Te /2Θrot

- In (1- e-Θvib/T)+[(Θvib/T) /eΘvib/T -1]
+ In ge1
Given: Θrot = 2.88; Θvib = 3374 and ge1 =1
Solution
 (2πMkT/h2)3/2 = [2π(4.653x1026)(1.3807x 10-
23
)(298.15)/ (6.626x10-34) ]3/2
= 1.436 x 1032/m3
Vm/NA = RT/NAP
=(0.08314)(298.15)/(6.022x1023)(1bar)
= 4.117 x10-23 L = 4.117x 10-26/m3
Te/2Θrot = (298.15) (2.718)/ 2(2.88) = 140.7
1- e-Θviv/T = 1 – e-11.31 = 1.000
(Θvib/T)/ eΘvib /T -1 = 1.380 x 10-4

So = 150.4 + 41.13 + 1.15x10-3 + 0 = 191.5 J/Kmol

Expt Value : 191.6 J/Kmol
Sample Problem
Calculate the standard molar entropy of CO2
at 298.1K and compare the expt. Result of So
= 213.8 j/Kmol

Hint: CO2 is a symmetric linear molecule

Θrot = 0.561 K
Θvib = 3360, 9542, 1890
Do (kj/mol) = 1596
σ=2
Chemical Potential from a Partition
Function
 Recall:S = kT (∂ InQ/∂T)N,V + k In Q
U = kT2 (∂ InQ/∂T)N,V

Since: A = U – TS
A = kT2 (∂ InQ/∂T)N,V – [TkT
(∂ InQ/∂T)N,V + kT In Q
A = -kT InQ
Helmholtz Energy
 A = f (N,V,T)
 dA = (∂A/∂T)N,V dT + (∂A/∂V)N,T dV
+ (∂A/∂N)T,V dN
dA = -SdT – PdV + (∂A/∂N)T,V dN

 More convenient to express this quantity in terms

of n
(∂A/∂N)T,V dN = (∂A/∂n)T,V dn
dA = -SdT – PdV + (∂A/∂n)T,V dn

(∂A/∂n)T,V = μ
Gibbs Free Energy
 Definition: G = A + PV
dG = dA + d(PV)
=- SdT + VdP + (∂A/∂n)T,V dn
Total Derivative of G = G (T,P,n)
dG = (∂G/∂T)P,n dT + (∂G/∂P)n,T dP
+ (∂G/∂n)T,P dn
= =- SdT + VdP + μ dn
μ = (∂G/∂n)T,P = (∂A/∂n)T,V
Chemical Potential
 Since A = -kT In Q
and μ = (∂A/∂n)T,V
μ = [∂ (-kT InQ)/∂n]T,V
= -kT(∂ InQ/∂n)V,T
μ = -RT(∂ InQ/∂N)V,T

Another simple form for an ideal gas:

Recall :Q(N,V,T) = [q(V,T)] N / N!
In Q = N In q – N InN + N
where In N! = N In N –N
Standard Chemical Potential
 μ = -RT ∂/∂N ( N In q – N In N + N)
= -RT (In q – In N – 1 + 1)
Since q(V,T) α V

μ = -RT In [ (q/V) (V/N)]

μ = -RT In [ (q/V) (kT/P)]
μ = - RT In [(q/v) kT] + RT InP

Comparing this equation with

μ(T,P) = μo(T) + RT InP
μo(T) = - RT In [(q/v) kT]
Standard Chemical Potential
 Recall: q/V is a function of T only for ideal gas

 To calculate μo(T) remember that P is expressed

relative to the standard state pressure Po

 μ(T,P) = μo(T) + RT In(P/Po)

μo(T) = - RT In [(q/V) kT] + RT In Po

or = - RT In [(q/V) (kT/Po)]
Calculation of Standard Chemical
Potential
 For Ar(g) at 298.15K

q/V = (2πmkT/h2)3/2
= [(2π)(0.03995)(1.3806 x10-23)(298.15)/ (6.022 x
1023)(6.626 x10-34)2 ]3/2
= 2.444 x 1032/m3
kT/Po =RT/NAPo = (8.314)(298.15)/ (6.022 x
1023)(1.00x105Pa)
= 4.116 x10-26m3
RT = (8.314) (298.15) = 2479 J/mol
μo = -39.97 Kj/mol
Expt Value : -39.97 Kj/mol
Standard Chemical Potential for
Diatomic Molecules
 Take the ground state energy of the
molecule

 q(V,T) = Σj e-Ej/kT = e-Eo/kT + e-E1/kT

 Factoring out : e-Eo/kT

 q(V,T) = e-Eo/kT [1 + e-E1 – E0 /kT +…….]

= e-Eo/kT (qo(V,T))
Standard Chemical Potential for
Diatomic Molecules
 μo(T) - Eo = - RT In [(qo/V) kT] + RT In Po

=- RT In [(qo/V) (kT/Po)]
or =- RT In [(qo/V) (RT/NAPo)]

where: Eo = NAεo and Po =1 bar or 105Pa

Calculation of Standard Chemical
Potential for HI
qo/V = (2πMkT/h2)3/2 V ( T/σΘrot)
( 1 / 1 –e- Θvib/T) ge1
Given: Θrot = 9.25 and Θvib =3266
qo/V = (2πMkT/h2)3/2
= [(2π)(0.127)(1.3806x10-23)(298.15)/ (6.022
x1023)( 6.626 x10-34)2]3/2
= 4.51 x 1034/m3
RT/NAPo = (8.314) (298.15)/6.022x1023(105)
= 4.116 x 10-26m3
μo(T) – Eo = -52.90 Kj/mol
Expt. Value = -52.94 kj/mol