TUTORIAL PROBLEMS PHYSICAL CHEMISTRY 2

STATISTICAL THERMODYNAMICS

(ST-1) STATISTICAL THERMODYNAMICS OF A THREE-LEVEL SYSTEM

At equilibrium, the probability pi that a state of energy Ei is occupied is proportional to (this is what
the symbol ∝ expresses) the Boltzmann weight for that state (β is a convenient short-hand for
1/kBT).
E
− k iT
pi ∝ e B = e −βEi

The so-called canonical partition function q is the normalization constant for the probability
distribution over states. This means it is defined as:
E
− k iT
e −βEi
∑ ∑
q= e B =
i i

where i denotes all possible states of the system. The canonical partition function plays a central
role in calculation of thermodynamic properties. In this problem, we will perform some calculations
on a simple system and see how the canonical partition function itself is calculated and how it can
be used in calculations of the internal energy U and populations of energy levels.

Consider a system that is characterized by three possible energy levels. The lowest level has energy
0 and a degeneracy of 1, the second level has energy ϵ and a degeneracy of 1, and the third level
has an energy 2ϵ and a degeneracy γ. In terms of real systems, one could think of a molecule that
can be bound to a surface in the most favorable binding site (lowest energy, 0), a molecule that can
be bound to a surface but to a less favorable binding site (intermediate energy ϵ), and a molecule
that is in free space (or solution), not bound to the surface at all (highest energy 2ϵ). Another real
system example is a substrate bound to the active site of a protein, a substrate bound to a protein,
but not in the active site, and a substrate in solution.

PC2-TP-ST-1
(a) Give an expression for the canonical partition function q of this system as a function of ϵ, γ, and
temperature T (kB may feature as a constant).

(b) Show for this system, that the canonical partition function normalizes the probability
distribution, i.e. explicitly show that for the partition function of the system described in this
problem:
e −βEi
∑ ∑ q
pi = p0 + p1 + p2 + p3 + . . . = =1
i i

∑
(c) Give an expression for the average energy U = ⟨E ⟩ = pi Ei of this system as a function of ϵ,
i

γ, and β.

(d) Show that for this system, the expression for the average energy U can be obtained directly from
d ln q
the partition function q as U = − .
dβ

(e) Calculate the occupations (populations) of each of three energy levels (P0, P1, and P2) for the
case that: ϵ /kBT = 1 and γ = 1.

(f) At which temperature T0 are the occupations (populations) of the highest and lowest energy
levels of this system equal, i.e. when is P0 = P2, given that: ϵ = 5.0 kJ ⋅ mol−1 and γ = 1,000.

(g) Calculate the occupations (populations) of the three energy levels at the temperature T0
calculated above.

(h) Give a sketch of the occupations (populations) of the three energy levels as a function of the
temperature for: ϵ = 5.0 kJ ⋅ mol−1 and γ = 1,000

PC2-TP-ST-2
(ST-2) PARTITION FUNCTION AND ENERGY OF ONE-DIMENSIONAL TRANSLATIONAL
STATES (PARTICLE IN A BOX).

Quantum mechanics shows that the energy levels of translational states in one dimension are given
by:
n2h2
En = , n = 1,2,3,...
8m L 2

L is the length of the space in which the particle moves, m its mass, h Planck's constant, and n the
quantum number of the translational state. This is the solution for the Schrödinger equation of the
particle in a box.

(a) Calculate the energy of the ground state, the first excited state, state 1,001, and state 10,001 for
the one-dimensional translational motion of a water molecule with a box-length of 1 μm at room
temperature.

(b) Calculate the Boltzmann weights of the excited states mentioned above at room temperature,
relative to the Boltzmann weight of the ground state.

(c) Show that the canonical partition function for the translational motion of a particle in one
dimension, say x, with box-length LX is to good approximation given by:

( h2 )
2π m kBT 1/2
qX = LX

Which are the approximation(s) made in deriving this expression?

(d) Use the canonical partition function to show that the average one-dimensional translational
energy to good approximation is equal to kBT /2.

(e) Show that the canonical partition function for the translational motion of a particle in three
dimensions with box-lengths LX , LY , LZ is given by:

q = qX × qY × qZ
What assumption needs to be made to arrive at this expression?
PC2-TP-ST-3
(ST-3) BUILDING ENSEMBLES

This problem is designed to give you insight into the distribution of energy over available states,
and how the Boltzmann distribution is the distribution that emerges as the result of putting
conditions on the system, in particular fixing the total energy.

Statistical thermodynamics is based on knowledge of the DISTRIBUTION of particles over the

possible states. Once the distribution is known, all properties of the system can be calculated, see
for example the calculation of the energy as a function of temperature in TP-ST-1 and TP-ST-2. This
calculation was based on the partition function, which reflects the Boltzmann distribution.

Any distribution can in principle be discovered by COUNTING possibilities, and extracting

PROBABILITIES that a certain situation occurs. We will do this now for a simple system.

Consider a system consisting of 3 particles. Each particle can at any time be found in one of 4
molecular states, with energies of 0, 1, 2, and 3 quanta, respectively. It is in principle possible to
measure which energy levels (or even molecular states) are occupied by the particles, for example
when a particle is added to or removed from the system, the change in energy of the system is
measurable. However, it may not always be possible to know exactly which particle occupies which
state.

A microstate is a distinguishable way to distribute the particles over molecular states. A distribution
over energy levels is specified by giving a list of the number of particles in each level, e.g.,
{0,1,2,0} describes the microstate: there are 0 particles in the lowest level, 1 particle in the next
level, 2 in the next level, and 0 particles in the highest level. If there are three particles, they can in
general satisfy this distribution in different ways, e.g. by swapping distinguishable particles
between molecular states of the same energy. Whether or not two distributions are distinguishable,
and therefore constitute one or two microstates, depends on the character of the particles and on the
nature of the system. In general, molecules are indistinguishable in the gas and liquid phases, but
distinguishable in the solid state; in principle, you can tell which molecule is which in a solid, but
not in a liquid or in a gas. In macroscopic molecular systems, molecules of the same species
(methane, ethanol, etc.) are indistinguishable, but distinguishable from other species. Finally, an
ensemble is the collection of all possible and allowed microstates of a system.
PC2-TP-ST-4
Restrictions (constraints, conditions) may limit the number of microstates because only allowed
microstates contribute to the ensemble. Restrictions may be, for example the total number of
particles, or the total energy.

(a) We will study four ensembles (I-IV) for the system described of three particles with four
possible molecular states. The ensembles differ in the conditions that are imposed:

I. The particles are distinguishable, the energy states are distinguishable, and there is no restriction
on the occupation of a level.
II. The particles are indistinguishable, the energy states are distinguishable, and there is no
restriction on the occupation of a level.
III. The particles are indistinguishable, the energy states are distinguishable, there is no restriction
on the occupation of a level, but the total energy must be exactly 3 quanta.
IV. The particles are indistinguishable, the energy states are distinguishable, there is no restriction
on the occupation of a level, but the total energy must be exactly 2 quanta.

For each of the ensembles described above, characterize/describe:

(i) the allowed distributions, or microstates, of the particles over the energy levels,
(ii) the number of realizations of each allowed distribution or microstate,
(iii) the total number of allowed realizations,
(iv) the average occupation of the energy levels in the ensemble.

This can be done by assigning the particles one by one to an energy level, in which the order in
which the particles are placed in the levels matters, i.e. you must place the first particle in column 1,
the second in column 2, etc... This implements the presumption that we can measure in which state
a particle enters the system, or in other words, what is the energy and the molecular state of the
particle as it enters the system. Therefore, choose a particle and place it in a level. e.g. level 3. Put
the first particle in column 1.The first particle is then in state 3-1. Choose a second particle and
place it in a level, e.g. level 2. The second particle is then in state 2-2. When all particles are placed,
count the number of different ways in which the same distribution over the energy states is realized.

PC2-TP-ST-5
A distribution over energy levels is specified by giving a list of the number of particles in each
level, e.g., {0,1,2,0} describes the microstate: there are 0 particles are in the lowest level, 1 particle
in the next level, 2 in the next level, and 0 particles in the highest level. In this distribution, it does
not matter whether particles are in a state labeled 1, 2, or 3.

(b) Can you now come up with a general expression for the number of microstates for N particles,
with a distribution over energy levels{n 0, n1, n 2, . . . }, under the conditions that the energy states are
distinguishable and the particles are indistinguishable?

PC2-TP-ST-6
(ST-4) DERIVING THE BOLTZMANN DISTRIBUTION FOR MOLECULAR SYSTEMS FROM
STATISTICS

The Boltzmann distribution can be derived by studying the number of ways the particles in a large
system can be distributed over the available energy levels.

(a) Explain that the number of realizations W in which N indistinguishable particles can occupy N
distinguishable energy states such that n1 particles are in energy level 1, n 2 particles are in energy
level 2, etc.. is given by:
N
N!
∏i=1 ni ! ∏
W= N
; ni ! = n1 ! × n 2 ! × n3 ! × . . . × nN !
i=1

(b) In statistical mechanics, the distribution {n1, n 2, n3, . . . } of particles over energy levels for
which the number of realizations W is largest is of special interest. Explain why this so-called most
probable distribution is so important and use a simple example to argue why other distributions are
usually not that relevant.

(c) The most probable distribution dominates the properties of a macroscopic system. The most
probable distribution is the distribution with the maximum number of realizations, W. Show that
maximizing the number of realizations of a distribution can also be written in terms of occupations
(or probabilities) {pi} of the energy levels as maximizing:

∑
− pi ln pi
i

∑
In other words, show that max W is equivalent to max − pi ln pi.
i

Hints: Use the Stirling approximation for large numbers: ln N ! = N ln N − N, and that pi = ni /N.

PC2-TP-ST-7
(d) Calculate the occupations (probabilities) of the energy levels for the most probable distribution
for placing N particles in N energy levels, under the conditions (restrictions) that the average energy

∑ ∑
U = ⟨E ⟩ = pi Ei is equal to Eand that the distribution is normalized, i.e. pi = 1.

Hint: use the method of the undetermined Lagrange multipliers.

(e) The average energy of a system, U, can be changed in two ways: by changing the distribution
over energy levels (not changing the energy of the levels) and by changing the levels (not changing
the distribution over the levels). Explain how these two ways of changing the energy as described
by statistical mechanics are related to the Fundamental Equation of thermodynamics:

dU = Td S − pd V

PC2-TP-ST-8
(ST-5) PARTITION FUNCTION AND ENERGY OF A RIGID LINEAR ROTOR.

Rotational motions of molecules are often modeled using the rigid rotor approximation. In this
approximation, the molecular geometry is fixed (there is no vibrational motion). Quantum
mechanics shows that the rotational energy levels with quantum number J for a rigid di-atomic
molecule with moment of inertia I, are given by:

ℏ2 J(J + 1)
EJ = , J = 0,1,2,..., where the degeneracy of level J is gJ = 2J + 1.
2I

(a) Calculate the rotational contribution to the molecular canonical partition function of a rigid di-
atomic molecule under the assumption that the separation between the energy levels is much
smaller than kBT.

(b) Make a sketch of the canonical rotational partition function as a function of T, for
ℏ2 /2I kB = 0.1, 1.0, and 10.0.

(c) Make a sketch of the average energy of rotation as a function of T, for

ℏ2 /2I kB = 0.1, 1.0, and 10.0.

(d) Calculate the rotational contribution to the CV for the rigid di-atomic molecules studied here
under the assumption that the separation between the energy levels is much smaller than kBT.

PC2-TP-ST-9
(ST-6) PARTITION FUNCTION AND ENERGY OF A HARMONIC OSCILLATOR (CHEMICAL
BOND).

Vibrational energy levels of molecules are often approximated using the simple harmonic oscillator
model. The quantum mechanical simple harmonic oscillator is characterized by equidistant energy
levels:

( 2)
1
En = n+ h ν, n = 0,1,2,...,∞

∞
−1
x n = 1 + x + x 2 + x 3 + . . . by performing a
∑
(a) Show that the function f (x) = (1 − x) =
n=0

Taylor series of the function around x0 = 0.

(b) Calculate the canonical partition function for a quantum mechanical simple harmonic oscillator
with respect to the minimum of the potential energy curve, E = 0.

(c) Make a sketch of the vibrational partition function of the quantum mechanical simple harmonic
oscillator model as a function of T for h ν/kB = 0.1, 1.0, and 10.0.

(d) Make a sketch of the average energy as a function of T for h ν/kB = 0.1, 1.0, and 10.0.

(e) Calculate the vibrational contribution to the CV for the molecules studied here.

(f) The equipartition principle states that at thermal equilibrium, each degree of freedom contributes
on average kBT /2 to the kinetic energy. At which temperature is the classical description of
vibrational motion a good enough approximation for the three quantum oscillators studied here?

PC2-TP-ST-10
(ST-7) ENTROPY CHANGES FROM PARTITION FUNCTIONS

In this question, the entropy change in the solvation or evaporation process is calculated.

(a) Combine two expressions for the Helmholtz free energy to show that the entropy change in the
vapor-liquid transition is given by:

ΔU + kBT ln Ql − kBT ln Qg ΔU Q
ΔS = = + kB ln l
T T Qg

(b) Calculate the contribution of the translation to the entropy change when 1 mol of water
condenses at 373 K and a pressure of 1 bar. (HINT: find the densities of liquid water and water
vapor at 373 K and 1 bar and assume the water molecules occupy a cubic volume. Use only the
translational partition functions in your calculation of the entropy.)

(c) Calculate the standard entropy change of vaporization of methane at the standard boiling point
temperature based on the data given in TP-PD-2(a), and based on the translational partition
functions only and comment on the observed difference.

(ST-8) HELMHOLTZ FREE ENERGY DIFFERENCES

In general, show that the Helmholtz free energy difference between two states can be written in
terms as a ratio of probabilities of finding a molecule in a particular state.
HINT: think about the meaning of the partition function.

PC2-TP-ST-11
(ST-E1) VAPOR PRESSURE IN A STATISTICAL THERMODYNAMICS TREATMENT

The vapor pressure of pure compounds can be predicted from statistical mechanical considerations
using a simple model. Assume that the pressure of the vapor is given by the perfect gas law and that
the volume per particle of the liquid, vl, is independent of temperature. Furthermore, assume that the
energy levels of the translational states in the vapor start at an energy ΔvapU (this is the

vaporization energy of the compound) above the start of the translational energy levels of the liquid.
Finally, treat the molecules as indistinguishable and do not consider rotational and vibrational
contributions to the energy.

In this question, we will show that the pressure and temperature at which liquid and vapor are at
equilibrium, under the assumptions stated, are given by:

−ΔvapU
[ kBT ]
kBT
p= × exp
e × vl

This expression is an equation for the liquid-vapor co-existence curve in a p-T diagram (vapor
pressure as a function of temperature). In subsequent questions, we will calculate a number of
thermodynamic quantities for the liquid-vapor transition.

(a) Explain that the total Gibbs free energy of a system containing only vapor and liquid molecules
of a single substance and under the conditions described is given by:

N
qlNl qg g
Gtot = − kBT ln Q + pV = − kBT ln × + p (Nl vl + Ng vg)
Nl ! Ng !

In this expression, the subscripts g and l denote quantities pertaining the the vapor and liquid,
respectively. State which assumptions and/or approximations are made in writing down this
expression.

PC2-TP-ST-12
(b) Show that for a particular temperature and pressure on the co-existence curve, i.e. at liquid-
vapor equilibrium, the chemical potentials of the liquid and the vapor phase must be the same.

HINT: write down a general expression for the (change in the) total Gibbs free energy in terms of
the chemical potentials in each phase and the (change in the) number of molecules in each phase;
the only allowed change in the system is the number of molecules in the liquid and vapor phase, but
the total number of molecules must remain the same.

(c) Show that for a particular temperature and pressure, at equilibrium:

qg Nl p
kBT ( g
ln = v − vl)
ql Ng

HINTS: calculate the chemical potentials in the form of ∂Gtot /∂Nα, α = l, g. Use the Stirling
approximation for large numbers. Then equate the chemical potentials and simplify.

(d) By writing q = qtr × qrot × qvib × qel × e −βΔE0, and using the perfect gas law for the vapor
phase, show that, at equilibrium, for the system described, the relation between pressure and
temperature is given by:
−ΔvapU
[ kBT ]
kBT
p= × exp
e × vl

Do you need to make any further assumptions in arriving at this expression?

PC2-TP-ST-13
(ST-E3) GIBBS ENERGY OF SOLVATION

The Gibbs energy of solvation Δsolv g is a measure of the difference in (effective) interaction of a
molecule in the vapor and the liquid phase, respectively. To that end, the chemical potential is
written as a part that contains quantities that depend on the environment, and a part that does not
depend on the environment. Taking the expression for μ developed in TP-ST-E1(c), the separation is
made explicit as follows:

[ Λ N ]
qtr × qrot × qvib × qel × qint V
μ = − kBT ln = − kBT ln 3 + ln (qrot × qvib × qel × qint)
N
V
= − kBT ln + μ*
Λ3 N

μ* is the quantity that depends on the environment.

(a) Explain why the terms collected in μ* depend on the environment and the other term does not.

(b) The Gibbs energy of solvation Δsolv g is defined as μ*

l
− μ*
g , i.e. the difference in the part of the

chemical potentials that depends on the environment. It can be related to the number densities of the
vapor and liquid at equilibrium, and can therefore be observed experimentally! The number density
ρ is the number of (moles of) particles per unit volume. Show that the Gibbs energy of solvation
can be expressed as:

ρl
Δsolv g = − kBT ln
ρg

PC2-TP-ST-14