Sie sind auf Seite 1von 26

Statistical Thermodynamics, Lecture 1 1

Alex Burin; aburin@tulane.edu; 1-504 862 3574;


Israel 508

Lecture Block 1. Thermodynamics Laws


A. Basic definitions.

Thermodynamics is the impressive area of knowledge because it has been


developed without understanding of microscopic nature of materials mostly based on
common sense reasoning and yet it is very useful practically. Reviews of the first, second
and third laws of thermodynamics are our preliminary targets. Later we will use statistical
mechanics to calculate thermodynamics parameters using microscopic models.
Thermodynamics is the sort of precise science compared to microscopic theories,
which are always approximate since they are based on approximate models.
Thermodynamics is based on definitions and common sense reasoning. Sometimes it is
valid in the same extent as the algebra relationships like 2+2=4. This is valid because it is
the immediate consequence of the definitions of 2 and 4. In the same extent we can talk
about the energy conservation law leading to the first law of thermodynamics. This is
because the energy is introduced as something that conserves.
Before getting to the laws of thermodynamics we need to define some common
sense parameters describing the macroscopic system which can be gas, liquid, solid or
some other state of matter. We know that materials are made of molecules and there are a
huge number of them (around Avogadro’s number NA=6.23⋅1023mol-1). The fully
accurate description of material requires the description of each molecule. This is usually
too complicated. Instead the macroscopic system can be described by some macroscopic
parameters including temperature, pressure, volume, internal energy, specific heat, etc.,
which was introduced without the knowledge of their precise nature. These principles are
the essence of thermodynamics and we are going to introduce all those parameters first.
The definition of each parameter contains two parts. The fundamental part is its
sense and key properties. The practical part is the way that this parameter can be
measured. Few words should be said about the units of measurements. We will use SI
(international) system of units. The seven SI base units are Kelvin (temperature), second
(time), meter (length), kilogram (mass), candela (luminous intensity), mole (amount of
substance) and Ampere (electric current). Thus our unit of volume is a cubic meter. Of
course you need to know how to switch between units. I reserve a right to give problems
in exams like how many liters are in a cubic meter or how many cubic meters are in a
cubic centimeter?
Below I define volume V, temperature T, and pressure P. The volume measures
the space occupied by the material. Its way of measurements depends on its state (solid,
gas or liquid). Gas under laboratory conditions occupies the whole volume of the
container where it is hold. So measuring the volume of the container one can find the gas
volume. Changing container we will change the gas volume. The words “under
laboratory conditions” are important here since the atmosphere is hold near the earth by
the earth’s gravity so atmospheric gases do not occupy all available volume of the
universe.
The volume of liquid does not depend on the container so we can use special
containers adjusted to measure the liquid volume like a graduated cylinder.
Statistical Thermodynamics, Lecture 1 2
Alex Burin; aburin@tulane.edu; 1-504 862 3574;
Israel 508
The solid holds its shape. Yet we can put it into in the liquid and determine the
solid volume as the change in the total volume within the graduated cylinder.
It is always important to determine where is 0? For the volume it obviously means
the absence of material so the concept of zero volume is very clear from the common
sense point of view.
The concept of temperature needs some common sense definition based on how it
behaves. If we put in touch two bodies 1 and 2 having different temperatures T1<T2 then
the hotter body 2 will share its heat with the cooler body until their temperatures get
equal. Then the system equilibrates. This common sense rule gives rise to the alternative
formulation of the second law of thermodynamics telling us that the heat can be
transferred from a hot body to a cold body and never backwards. This gives a clear
example of the irreversible process in the closed system.
The definition itself suggests the method to measure temperature. Imaging a small
device, called thermometer with the known temperature. One can bring it in a thermal
contact with a large body and wait until they will get to equilibrium. Since the
thermometer is small its temperature gets adjusted to the large body temperature, which
changes negligibly. Consistently the temperature of thermometer will tell us about the
body temperature.
The temperature of thermometers can be defined using their volumes. For the vast
majority of materials their volume increases with the temperature proportionally to the
temperature. Then measuring the volume of the thermometer we can determine the
temperature of the thermometer and, consequently, of the body. The change of the
volume of the mercury drop being in contact with the human body is used to determine
the blood temperature.
As we already noticed we use the SI system of units where the temperature unit is
1K. It is important to define what is the absolute zero temperature. This definition has
been found long time ago investigating the dependence of some gas volume on a gas
temperature. Such experiment can be made placing gas inside the air balloon and heating
it at constant atmospheric pressure. The gas volume increases with increasing the
temperature proportionally to it (this is actually true for ideal gases only but to my
knowledge all gases are nearly ideal at atmospheric pressure). This dependence can be
written in a very general form
ΔV = V2 −V1 = a(T2 − T1 ) = aΔT, (1.1)
where (T1, V1) and (T2, V2) are temperatures and volumes before and after heating and a
is some proportionality coefficient. We can resolve this equation for arbitrarily final
volume and temperature in the form
V = a(T − T0 ),
(1.2)
T0 = T1 −V1 / a.
Originally temperature was measured using metric units of degrees of Celsius bound to
the water melting (0) and boiling (100) degrees. It turns out that in spite of a broad
variation of experimental parameters for different gases (initial temperature and volume
and the proportionality constant a) the temperature T0 where the volume of all gases was
extrapolated to zero using Eqs. (1.1), (1.2) is quite universal and given by -273.150
Celsius. At this temperature the gas pressure can be also extrapolated to zero suggesting
zero motion energy of the gas. This was the basis for the definition of absolute zero
Statistical Thermodynamics, Lecture 1 3
Alex Burin; aburin@tulane.edu; 1-504 862 3574;
Israel 508
temperature and it determines the absolute (SI) temperature scale using Celsius degrees
with added number 273.15. In this new scale absolute zero (zero gas volume and
pressure) corresponds to the zero temperature -273.150C.

When temperature indeed approaches zero degrees the gas properties changes and
Eq. (1.2) is no longer valid but the measurements under normal conditions still give a
very good definition of the absolute zero temperature using extrapolation.
The pressure, P, of the
Fig. 1. Volume vs. Temperature gas is defined using the force
0.04
per the unit area acting from
P= 1 atm the gas to the walls of the
container. To hold gas at the
0.03 given volume we need to apply
the certain force to the walls to
resist to its expansion. Solids
V (m3)

0.02 P= 2 atm
and liquids usually stay at the
0
T0 = -273.15 C given volume but if we try to
P = 3 atm compress them reducing their
volume they will act with the
0.01
strong backwards pressure. In
P = 4 atm
the laboratory conditions the
pressure of material is
0 determined by the atmospheric
-300 -200 -100 0 100 200
pressure acting on them
T (0C) uniformly. The atmospheric
pressure can be measured with the help of barometer.
The SI unit for pressure is 1 N/m2 and it is also called 1Pa. The atmospheric
pressure is about 105 times higher.
Sample problem: What is the pressure equilibrating the gravity of 760 mm Hg? The
density of Hg is 13594 kg/m3.
The quantity of the matter is another important characteristics, which can be
determined by the mass of the homogeneous material of interest (not mixture). The molar
mass representing the product of Avogadro’s number to the molecular mass is used for
the unit quantity (one mole). Then the quantity of the matter is determined by the ratio of
its mass to the molar mass n=m/M. The SI unit for the quantity of matter is one mole.
Thus we introduced four characteristics of macroscopic material including
volume, temperature, pressure and quantity (V, T, P, n). There is a common believe that
knowing any of these three parameters and the state of the matter we can determine the
fourth one. For the gases under normal conditions (atmospheric pressure) the
interrelationship of parameters is given with a very high accuracy by the ideal gas
equation, which can be written in the form
PV = nRT. (1.3)
-1 -1
The universal gas constant R is defined as R=8.3144598 J mol K .
Van der Waals suggested the generalization of this equation for smaller volumes
and/or higher pressures in the form
Statistical Thermodynamics, Lecture 1 4
Alex Burin; aburin@tulane.edu; 1-504 862 3574;
Israel 508
! an 2 $
# P + 2 & (V − nb) = nRT. (1.4)
" V %
Van der Waals parameters a and b for some gases are given in the end of these notes.
This equation can be rewritten for a one mole of the gas in the form
a RT
P=− 2 + , (1.5)
V V −b
where V is the molar volume. The first term reflects the attractive interaction at long
distances between molecules. Its dependence on the volume comes from the distance
dependence of their dispersive interaction. The subtraction of the parameter b from the
volume reflects the subtraction of the minimum volume corresponding to the condensed
phase. It is practically impossible to compress the gas to smaller volume because of the
electronic repulsion.
To see how important are related corrections to the ideal gas equation one can
calculate the change in the gas mass density in Eq. (1.4) compared to Eq. (1.3) for some
gas, say CO2.
What can we learn from ideal gas laws? First of all one can express mass density
ρ in the ideal gas as
PM
ρ0 = , (1.6)
RT
where M is the molar mass of a given gas. So knowing pressure, temperature and the
chemical formula of molecules one can find the gas density. Does this really work? Let
us check that for some gas. For molecular oxygen this density is known experimentally to
be ρ=1.429 g/L=1.429 kg/m3 at standard temperature and pressure (STP) in accordance
with the Wikipedia page https://en.wikipedia.org/wiki/Oxygen. What are the standard
temperature and pressure? Standard temperature is defined as zero degrees Celsius (0 0C),
which translates to 32 degrees Fahrenheit (32 0F) or 273.15 degrees kelvin (273.15 K).
This is essentially the freezing point of pure water at sea level, in air at standard pressure.
Definition of the standard pressure is time sensitive. Before 1982 it was one atmosphere
that is 101325 Pa (N/m2), while after 1982 it becomes exactly 105Pa or 1 bar. Knowing
the gas constant R=8.314 and the molar mass of oxygen 2*15.999 g = 31.998 g =
0.031998 kg one can estimate the gaseous oxygen density as 1.4090 kg/m3. It is close to
the right number but not exactly there and differs by around 2%. If, however I use the old
definition of STP (before 1982) I would be getting ρ=1.427 kg/m3 which is about ten
times more accurate. Perhaps the numbers in Wikipedia are given for the old SPT
definitions.
The Van der Waals equation should give better approach to reality. Let us see
how can we improve the present estimate? Using the definition ρ=M/V for the one mole
one can obtain the following cubic equation for the gas mass density
ab 3 a 2
ρ − ρ + ( RT + Pb) ρ − MP = 0.
M2 M
It is hard to solve the qubic equation exactly in any reasonable form. One can construct
an approximate solution assuming that Van der Waals corrections to pressure and volume
(-a/V2 and b) are much smaller than actual pressure and volume. Then one can seek for
the solution in the form
Statistical Thermodynamics, Lecture 1 5
Alex Burin; aburin@tulane.edu; 1-504 862 3574;
Israel 508

MP
ρ = ρ0 + ρ1, ρ0 = , ρ0 >> ρ1.
RT
Substituting this form into the cubic equation and keeping only linear term in ρ1 one can
find the Van der Waals correction ρ1 in the form
⎛ ab 3 a 2 ⎞
⎜ − 2 ρ0 + ρ0 − Pbρ0 ⎟ ⎡ a b a b⎤
ρ1 = M M
⎝ ⎠
= ρ0 ⎢− 2
+ 2
− ⎥.
RT ⎣ PV0 V0 PV0 V0 ⎦

Using Van der Waals gas parameters for oxygen a=1.378 L2atm/mol2 = 0.1378
m6Pa/mol2, b=0.03183 L/mol=3.183*10-5m3/mol one can estimate ρ0+ρ1=1.4294 kg/m3.
This result matches available data within the experimental accuracy. The root of the cubic
polynomial can be found numerically and it does not deviate from the above estimate.
Sample problem: What pressure should be applied to the gas at room temperature to
have the van Der Waals correction to the density of order of 10% of the density. How big
is this pressure for a gaseous oxygen?
The ideal gas equation results in the interesting conclusion that the volume per the
mole, VM=RT/P (0.0224 m3 STP), does not depend on the specific gas properties (molar
mass, etc). This reflects the fact that the volume per the molecule in an ideal gas at given
temperature and pressure does not depend on specific molecules. As we know the
temperature determines the kinetic energy per the molecule. Independence of pressure on
the specific molecules reflects the fact that its determined by the kinetic energy of
molecules only.
The above observation helps to generalize the ideal gas equation for the gas
mixture. The number of moles n in the ideal gas equation for the gas mixture should
given by the sum of the numbers of moles of all participating gases
PV = (n1 + n2 +... + nk )RT.

B. First law of thermodynamics


The first thermodynamics law is the energy conservation law. We consider some
macroscopic body that is in contact with the external world. The important underlying
assumption is taken here that its interaction with the external world is weaker than
internal interactions. For example the bulk solid material has most of atoms located
deeply inside it and they couple only to nearest atoms so this is the internal interaction.
Only surface is in the direct contact with the external world and this is where the forces
can be applied.
Some forces act on the body and it also can exchange energy and particles with its
environment. The action of forces changes its energy and this change is called the work.
The small change can be expressed it in a differential form as dW. There are other
changes unrelated to the work of external forces, which are called a heat transfer. Assume
that our material is in contact with the other material having higher temperature. Then
this contact will bring heat to our material without any visible work of external forces.
The energy transferred to our material in this manner can be denoted ad dQ. The total
change of the energy can be expressed as the sum of two
Statistical Thermodynamics, Lecture 1 6
Alex Burin; aburin@tulane.edu; 1-504 862 3574;
Israel 508
dU = dW + dQ, (1.8)
In thermodynamics it is more convenient to use pressure instead of force and
pressure is given by the ratio of the force applied to the certain area to that area (see Fig.
1). Assume that we have some material (bulk solid, liquid or gas in container). The
system normally stays under some external pressure associated for instance with the
atmospheric pressure. The shape of the system (body) does not change though because
this external pressure is equilibrated with the internal pressure, Pin, which is equal to the
external pressure, Pex. Then assume that we applied the larger pressure Pex>Pin (see Fig. 1
below). What would happen?
The squeezing will take place. How long would it proceed? Since the material
resists to squeezing increasing the internal pressure it will take place until the internal
pressure will balance the external pressure. Until that the work of the external force can
be represented as the external force times the displacement of the medium border which
can be conveniently represented as the product of the external pressure and the volume.
Then we can express work as -PexΔV where ΔV is the change of the volume. The
negative change suggests positive work, while the positive change suggests the negative
work since the system acquires energy during squeezing and releases it during
decompression.
This gives rise to the commonly used form of the first law of thermodynamics
dU = −PdV + d 'Q, (1.9)
separating mechanical work made by external pressure and everything else called the heat
transfer. The deep sense of this law is that the internal energy can change only as the
effect of external action.

Fig. 1. Illustration of the work made on the material. It can be expressed either as
d’Wext=Fdx or as d’Wext=(F/A)Adx = -PdV

How to apply these concepts to the ideal gas? The molecules in ideal gas interacts
very weakly with each other, so the total energy should be expressed as the product of the
number of molecules N and typical molecular energy which is the same as the product of
molar energy multiplied by the number of moles. Consequently U=nUM(T). The energy
change due to heating can be described by the differential change
dU=ndUM(T)/dT=nCVdT where nCV is the gas heat capacity at constant volume and CV is
the molar specific heat, which is the heat capacity of one mole. If the heat capacity is
approximately temperature independent then the gas energy can be expressed as
Statistical Thermodynamics, Lecture 1 7
Alex Burin; aburin@tulane.edu; 1-504 862 3574;
Israel 508
U=c+CVT=U0+CV(T-T0), where U0 is the reference energy at the reference temperature
T0. The heat capacity increases proportionally to the quantity of matter. It is convenient to
introduce the specific heat per unit mass that is given by the ratio of the molar heat
capacity and the molar mass, cV=CV/M. The heat capacity expresses the amount of heat
which should be delivered to the gas at constant volume to raise its temperature by 1K.
The practical measurements are usually performed at constant pressure, rather
than volume. Under those conditions it is convenient to introduce the specific heat at
constant pressure. In this regime the energy (heat) transferred to the gas is used to raise
the gas internal energy (temperature) by CV(Tf-Ti) where Ti and Tf are initial and final
temperatures and to expand gas performing the work P(Vf-Vi)=R(Tf-Ti). Because of the
energy conservation the total energy transferred to the gas is given by (CV+R) (Tf-Ti).
Consequently, the heat capacity at constant pressure is given by CP=CV+R.
Molar specific heats of ideal gases at sufficiently high temperature can be defined
using the equipartition principle for the gas energy. This principle suggests the same
molar energy RT/2 or RT for each degree of freedom. For monatomic molecules (n=1) all
three degrees of freedom correspond to translational motion having each the energy
RT/2. The total energy then is 3RT/2 and for the molar specific heats we got CV=3R/2
and CP=5R/2.
For n-atomic molecules of non-linear shape the number of degrees of freedom can
be count as following. The total number is 3n because each atom has three directions to
go. Three of them correspond to the translational motion and three to rotational having
total energy 3RT. The remaining 3n-6 degrees of freedom are vibrational. Each
vibrational degree of freedom under classical conditions (high enough temperature)
possess the energy RT including RT/2 for kinetics and RT/2 for potential energies.
Consequently the total energy is U=(3n-3)RT and the specific heats are given by CV=(3n-
3)R and CP=(3n-2)R. If the molecule has a linear shape then the number of rotational
degrees of freedom is 2 so the number of vibrational degrees of freedom is 3n-5.
Consequently the energy is U=(3n-5/2)RT and the specific heats are given by CV=(3n-
5/2)R and CP=(3n-3/2)R.
Ideal and Van der Waals gas equations permits us to describe the work done on
gas. For instance if the gas is placed in container equilibrated thermally with the
environment, that is having the same temperature, the work to compress the gas from the
volume V1 to some smaller volume V2 can be evaluated integrating the first law of
thermodynamics as
V2 V2
W = − ∫ P dV = − ∫ nRT dV / V = nRT ln(V1 / V2 ),
V1 V1

V2 nRTdV V2 dV ⎛ V − nb ⎞ an 2 an 2
WVdW = − ∫ + ∫ an 2 2 = nRT ln ⎜ 1 ⎟+ − .
C.
V1 Second law
V − nb V1 ofVthermodynamics.
⎝ V2 − nb ⎠ V1 V2

B.1. Definition of entropy

Consider the fundamental concept of entropy. The irreversible evolution of the


system is characterized by the extensive parameter of the bulk systems called entropy.
One can introduce entropy S(E, V) requiring that (a) this is the extensive property (for
two different subsystems S=S1+S2) and (b) it does not change in reversible (adiabatic)
Statistical Thermodynamics, Lecture 1 8
Alex Burin; aburin@tulane.edu; 1-504 862 3574;
Israel 508
processes, while it always increases in irreversible (non-adiabatic) processes. (c) It is also
postulated that the entropy is the increasing function of energy. This information is
definitely insufficient and the last requirement that we are going to introduce during our
discussion is (d) the relation for the system in the equilibrium that (∂S/∂E)V=1/T, where T
is the temperature of the system and the derivative is taken at constant volume. Since the
temperature must be positive as we will see considering its statistical sense this also
explains why entropy must increase with energy. Probably it makes sense to add the
Nernst theorem because up to now we have only defined the change in entropy. The
Nernst theorem (The third law of thermodynamics) tells us that the entropy tends to
zero at zero temperature and then one can define entropy heating the system from zero to
the temperature of interest.
To define entropy more accurately we can consider the irreversible process
induced by the piston motion like in Fig. 1. If there is no heat transfer the work associated
with the piston motion can be expressed in the integral form as
V2
W = − ∫ Pext (V )dV . (1.10)
V1

To compress the gas the external pressure should exceed the equilibrium internal pressure
Pin(V) so the total work can be separated into two parts (V2<V1)
V2 V2
Wc = − ∫ Pin (V )dV − ∫ (Pex (V ) − Pin (V ))dV . (1.11)
V1 V1

The work for the backwards process can be represented in a similar manner
V2 V2
Wd = ∫ Pin (V )dV + ∫ (Pex (V ) − Pin (V ))dV , (1.12)
V1 V1

with the important difference that the external pressure for the decompression process
should be smaller than the internal pressure. Consequently, first terms in Eqs. (1.11),
(1.12) are equal and of the opposite sign so they compensate each other if the
compression-decompression cycle is performed, while second terms are positive in both
cases representing the heating of the system. The difference between actual external and
equilibrium pressures is responsible for the irreversibility of the system evolution in the
cycle. If the external pressure approaches the internal equilibrium pressure the
irreversible part disappears though the compression decompression takes longer and
longer in time. In the limit of infinitely slow (adiabatic) process external pressure is
getting equal to the internal pressure and the process becomes reversible. According to
the above definition of entropy it should not change in such a process.
The energy increase during the cycle represents the heat transferred to the system,
because the volume finally did not change. In the limit of adiabatically slow reversible
process there is no heat transferred to the system and mechanical works compensate each
other in forward and backwards paths.
One can represent the entropy change in the standard mathematical form of the
full differential as
⎛ ∂S ⎞ ⎛ ∂S ⎞
dS = ⎜ ⎟ dE + ⎜ ⎟ dV . (1.13)
⎝ ∂E ⎠V ⎝ ∂V ⎠ E
Statistical Thermodynamics, Lecture 1 9
Alex Burin; aburin@tulane.edu; 1-504 862 3574;
Israel 508
We assumed no change in the number of particles. Eq. (1.13) is the multivariable calculus
generalization of the standard equation df=(df/dx)dx. To prove the validity of this
substitution one can take derivatives of the function change in the left hand side and the
linear function in the right hand side, which will lead to the identities. If we need
coincidence with the accuracy to square terms one can reexpress this like
df=(df/dx)dx+(1/2)(d2f/dx2)dx2+..
Sample problem. Write expression accurate to the cubic term in dx.
Sample problem. Write expression accurate to the quadratic terms for two variable
functions.
Assume that we consider the reversible process with the external pressure equal to
the internal pressure P=Pin. In any case according to the first law of thermodynamics one
can express the entropy change using Eq. (2) as
" ∂S % " ∂S % " ∂S %
dS = $ ' d 'Q − $ ' Pin dV + $ ' dV. (1.14)
# ∂U &V # ∂U &V # ∂V &U
As we defined above the entropy should not change in the reversible process where on
the other hand d’Q=0 as we have just established and pressure must be equal to the
equilibrium pressure.
Then using the definition of entropy we got
1 # ∂S &
dS = − Pin dV + % ( dV = 0. (1.15)
T $ ∂V 'U
This results in the identity
" ∂S % P
$ ' = in , (1.16)
# ∂V &E T
which completes the definition of the entropy. The change of entropy with the change of
energy and volume can be expressed as
dU Pin dV
dS = + . (1.17)
T T
Consequently we can rewrite the first law of thermodynamics in the form
dU = −Pin dV + TdS.
(1.18)
The first term represents reversible work corresponding to the adiabatic process limit,
while the second part describes the heat transfer to the system.
Consequently, one can express the change in entropy in terms of the heat
transferred to the system in the case of constant volume as
d 'Q
dS = . (1.19)
T
The latter equation is directly associated with the second law of thermodynamics. It tells
us that for thermally isolated system the spontaneous change of entropy is always
positive or zero (see the example with piston). Another example of spontaneous increase
of entropy is associated with two bodies with different temperatures placed in thermal
contact. The heat dQ transferred to the cold body is equal to the heat removed from the
hot body (energy conservation). The system entropy change can be expressed as the
entropy raise of the first body dQ/Tl and entropy reduction of the second body –dQ/Th.
The total entropy change dQ(1/Tl-1/Th) is positive since Th>Tl.
Statistical Thermodynamics, Lecture 1 10
Alex Burin; aburin@tulane.edu; 1-504 862 3574;
Israel 508
In other words I would say that the laws of thermodynamics define two
fundamental extensive parameters of the macroscopic system. One of them must be
conserved (energy) and one must grow when the processes are irreversible (entropy).
The important property of the system is its specific heat C, which we will usually
use as a heat capacity per one mole of matter. It represents the differential measure of the
amount of heat needed to raise the temperature, which can be expressed as dQ/dT. Since
the heat can be transferred under different conditions two situations should be considered
separately of constant volume (gas in container) or constant pressure (standard laboratory
conditions with atmospheric pressure). In both cases we can define specific heats
differently as the specific heat at constant volume, CV, or at constant pressure, CP, as
" ∂Q % " ∂S %
CV = $ ' = T $ ' ,
# ∂T &V # ∂T &V
(1.20)
" ∂Q % " ∂S %
CP = $ ' = T $ ' .
# ∂T &P # ∂T &P
Guess which of two should be larger?
Sample problems. Calculate the entropy of ideal gas assuming that CV is a volume
independent constant. What would happen with the temperature and entropy of ideal gas
in case if the container size increasing twice either adiabatically or instantaneously?

According to the first law of thermodynamics the gas energy change can be expressed in
terms of entropy and volume changes as dU=TdS-PdV. The energy of ideal gas is
expected to be volume independent because of the absence of molecular interactions.
Accordingly, it can be expressed as dU=CVdT. Then the differential of entropy takes the
form dS=CVdT/T+PdV/T. Using the ideal gas equation one can rewrite this result as
dS=CVdln(T)+nRdln(V)=d(CVln(T)+nRln(V)). Consequently the entropy can be
approximated by S=A+ CVln(T)+nRln(V), where A is volume independent constant. As
the alternative one can use the reference entropy S0 at certain reference volume V0 and
temperature T0 in the form S(V, T)=S0+ CVln(T/T0)+nRln(V/V0).

Calculate the difference CP-CV for the ideal gas:


The term “ideal gas” suggests that all its energy is related to kinetics and or internal
energy of its molecules and, consequently, we can neglect its volume dependence. Then
using the equation dU=TdS=CVdT one can express the ideal gas energy in the form
T
U=CVT+U0 ( U = U 0 + ∫ C(t)dt ).
T0

Consequently the differential of entropy can be expressed as dS=CVdT/T+PdV/T.


In the case of constant pressure one can express the volume differential using the ideal
gas equation as gas equation as dV=RdT/P. Accordingly we got dS=CVdT/T+RdT/T.
Consequently we got CP=CV+R.
If we slowly apply a mechanical perturbation to a thermally isolated system it is
natural to expect that it undergoes a reversible change. Accordingly entropy should be the
same during all process. Such processes are called adiabatic processes. Using the
definition of the ideal gas entropy one can find the connection between different ideal gas
Statistical Thermodynamics, Lecture 1 11
Alex Burin; aburin@tulane.edu; 1-504 862 3574;
Israel 508
parameters during adiabatic processes. Setting S(V, T)=S0+ CVln(T/T0) + nRln(V/V0) =
const, we get
CV
T R V = const.
Using the equation for the ideal gas PV=nRT one can find two other relationship of
pressure and temperature and pressure and volume during the adiabatic process. They
reads
CV CP CP
CV
T RT / P = T
R R
/ P = const, PV = const.

Lecture 2. Equilibrium between interacting systems. Chemical


potential.
Here we derive certain important properties of thermodynamic parameters. They
are related to the system equilibrium and stability. Consider the small subsystem being in
contact with the large system. Let these systems be characterized by energies u and U and
volumes v and V. Assume that the total energy and volume of the whole system are
conserved suggesting v+V=V0, u+U=U0. Then one can expect that the total entropy of
both systems together
Stot = S(U 0 − u,V0 − v) + s(u, v)
(2.1)
has maximum at equilibrium values of volume and energy. This suggest zero partial
derivatives of the total entropy with respect to energy u and the volume v and maximum
requirements for the second derivative to be satisfied.
Consider zero partial derivative conditions. They can be written as
∂Stot # ∂S(U 0 − u,V0 − v) & # ∂s(u, v) & 1 1
=% ( +% ( = − + = 0 → T1 = T2 ,
∂u $ ∂u 'V $ ∂u 'v T1 T2
∂Stot # ∂S(U 0 − u,V0 − v) & # ∂s(u, v) & P P (2.2)
=% ( +% ( = − 1 + 2 = 0 → P1 = P2 .
∂v $ ∂v 'U $ ∂u 'u T1 T2
These conditions suggest that temperatures and pressures of two subsystems being in
thermal and physical contact must be equal in equilibrium in a full accord with their
definition.
It make sense to include the dependence of entropy on the number of moles of a
certain material n present in both systems in case of allowed exchange of matter. If the
total quantity of matter N0 is conserved then one more significant equilibrium
requirement can be obtained
∂Stot # ∂S(N 0 − n) & # ∂s(n) & # ∂S(N 0 − n) & # ∂s(n) &
=% ( +% ( = −% ( +% ( = 0.
∂n $ ∂n 'U,V $ ∂n 'u,v $ ∂N 'U,V $ ∂n 'u,v (2.3)
The derivative of entropy with respect to the quantity of matter defines the important
property of matter called the chemical potential as
µ " ∂s(n) %
=$ ' .
T # ∂n &u,v (2.4)
Statistical Thermodynamics, Lecture 1 12
Alex Burin; aburin@tulane.edu; 1-504 862 3574;
Israel 508
As follows from Eq. (2.3) chemical potentials of materials staying in touch with each
other should be equal each other. Thus equilibrium between two materials requires equal
temperatures, pressures and chemical potentials.
Chemical potential can be used to represent the differential energy dependence on
the system parameters in the form
dU = TdS − PdV + µ dn.
(2.5)
Consequently one can define the chemical potential as a partial derivative of the internal
energy as
" ∂U %
µ =$ ' .
# ∂n &S,V (2.6)

To evaluate the Chemical potential for ideal gas one has to express the internal energy
U=nCV(T-T0)+U0 in terms of volume V and entropy
S=nS0(V0/n,T0)+nCVln(T/T0)+nRln(V/(nV0)). The temperature can be expressed in terms
of entropy using the definition of entropy and for the internal energy we get
⎡ S − nR ln(V / (nV0 )) ⎤
U = nU 0 (S0 ,V0 ) − nCV T0 + nCV T0 exp ⎢ ⎥.
⎣ nCV ⎦ (2.7)
Using the definition of the chemical potential in Eq. (2.6) we get the definition of the
chemical potential in the form
⎡ S R ⎤ U − TS + PV
µ = U 0 (S0 ,V0 ) − CV T0 + CV T + CV T ⎢− + ⎥=
⎣ nCV CV ⎦ n . (2.8)
The numerator in this expression is called the Gibbs free energy and this is a very general
definition of the chemical potential as a molar Gibbs free energy. Using previously
obtained expressions for the internal energy and entropy of ideal gas one can express the
chemical potential in terms of any two parameters of interest.

Lecture 3. Thermodynamic potentials, their properties as extensive


functions and Maxwell relations.
The maximum entropy requirement for both large and small subsystems can be
rewritten leaving only linear dependent terms for the large subsystem as
u pv G
Stot = S(U 0 − u,V0 − v) + s(u, v) = S0 + s − − = S0 − ,
T T T
G = u + pv − Ts. (3.1)
Thus the maximum entropy requirement for the large system is equivalent to the
minimum Gibbs free energy requirement for its small subsystems at temperature and
pressure determined by the large subsystem. The Gibbs free energy is defined as
G = U + PV − TS.
(3.2)
Statistical Thermodynamics, Lecture 1 13
Alex Burin; aburin@tulane.edu; 1-504 862 3574;
Israel 508
It’s minimum requirement is often used to determine the chemical equilibrium since a
small chemical subsystem usually interacts with the large environment characterized for
example by atmospheric pressure and room temperature.
Gibbs free energy is the example of so called thermodynamic potentials
approaching minimum under certain equilibrium conditions.

For instance internal energy approaches minimum at fixed entropy and


temperature. Assume that the system can be found in two states with the same entropy
(volume and number of particles) and two different energies
U(S,V, N ), (3.3)
U '(S,V, N ). (3.4)
Assume that U<U’. Then the state with the energy U can be transferred to the state with
energy U’ without changing its volume and number of particles adding the heat Q=U’-U.
This will raise the entropy of the system. The resulting state will be more stable than the
second state with the energy U’ because it possesses higher entropy. Therefore the second
state is thermodynamically unstable and the correct state possesses the energy minimum
at the given entropy. Thus amongst two system realizations at given entropy, volume and
number of particles we should choose the one possessing the minimum of energy.
The system energy is very inconvenient variable for the practical use. It is defined
as the function of entropy, which is very difficult to measure and one can prefer to study
the system at given temperature and volume (or very often, particularly in chemistry,
pressure). To do that one can often modify the definition of energy introducing new
thermodynamics potential, which must possess a minimum at equilibrium at other
desirable set of parameters. This modification is done using Legendre transformation.
Assume that we wish to describe the system at fixed volume and temperature.
Then one can use the Helmholtz free energy (I can call it sometimes simply free energy)
denoted by the letter A and defined as
A = U − TS. (3.5)
A is treated as a function of volume and temperature (also of the number of particles
which is not very important for us in this consideration) and the same should be done for
energy and entropy. Indeed one always have the so called state equation relating volume,
pressure and temperature or perhaps energy and entropy so these values are
interreplacible through each other and we can consider entropy as the function of
temperature and volume. Then Eq. (3.5) can be rewritten as
A(T,V ) = U(S(T,V ),V ) − TS(T,V ). (3.6)
Under fixed temperature and volume free energy takes the absolute possible
minimum value. Let us prove that. Consider the energy as the function of the additional
parameter a (say the entropy of the one of two subsystems in Eq. (3.1)) and assume that it
takes a minimum with respect to this parameter at fixed entropy and volume. Let us prove
that A(T, V) realizes the minimum with respect to this parameter at fixed temperature and
volume. Eq. (3.6) can be rewritten as
A(T,V, a) = U(S(T,V, a),V, a) − TS(T,V, a). (3.7)
The derivative of the free energy with respect to the parameter a at fixed temperature and
volume can be expressed as
Statistical Thermodynamics, Lecture 1 14
Alex Burin; aburin@tulane.edu; 1-504 862 3574;
Israel 508
⎛ ∂A(T , V , a) ⎞
⎜ ⎟
⎝ ∂a ⎠ T ,V
⎛ ∂E ( S (T , V , a),V , a) ⎞ ⎛ ∂E ( S (T ,V , a),V , a) ⎞ ⎛ ∂S (T ,V , a) ⎞
=⎜ ⎟ +⎜ ⎟ ⎜ ⎟ (3.8)
⎝ ∂a ⎠ S ,V ⎝ ∂S ⎠V ,a ⎝ ∂a ⎠ T ,V
⎛ ∂S (T ,V , a) ⎞ ⎛ ∂E ( S (T ,V , a),V , a) ⎞
−T⎜ ⎟ =⎜ ⎟ = 0.
⎝ ∂a ⎠ T ,V ⎝ ∂a ⎠ S ,V
This proves the minimum of free energy so we can use it instead of energy at
given volume and temperature. Similarly one can define other thermodynamics potential
possessing the minima under the properly chosen parameters. This includes the Gibbs
free energy
G(T, P) = U − TS + PV, (3.9)
having minimum at fixed pressure and temperature and enthalpy
H (S, P) = U + PV, (3.10)
having minimum at fixed entropy and temperature.
The differentials of these thermodynamics potentials can be expressed as
dU = TdS − PdV + µ dn,
dH = TdS +VdP + µ dn,
(3.11)
dA = −SdT − PdV + µ dn,
dG = −SdT +VdP + µ dn.
It is interesting that using the extensive properties of thermodynamic potentials
one can express them in terms of the combination of products of thermodynamic
parameters. Consider free energy A(T, V, N). Because of its extensive nature it is
proportional to the system volume A(T, V, N)= Va(T, N/V), where a(T, N/V) is the free
energy of the unit volume having the number of particles n=N/V which is the number
density of particles. Accordingly one can express the pressure as
⎛ ∂A ⎞ ⎛ ∂a ⎞ A
P(n, T ) = −⎜ ⎟ = −a(T, n) + n⎜ ⎟ = - + µn . (3.12)
⎝ ∂V ⎠ T,N ⎝ ∂n ⎠ T V
Then we can express the free energy as
A = µ n-PV . (3.13)
Using definitions of other thermodynamic potentials one can express them as
U = TS − PV + µ n,
H = TS + µ n,
A = µ dn − PV,
G = µ n. (3.14)
Thermodynamics potentials can be used to derive the Maxwell relations between
the derivatives of thermodynamics parameters. These relations are using the calculus
theorem that the second derivative of the function taking with respect to two different
variables does not depend on order of taking derivatives
∂2 f ∂2 f
= . (3.15)
∂x∂y ∂y∂x
Statistical Thermodynamics, Lecture 1 15
Alex Burin; aburin@tulane.edu; 1-504 862 3574;
Israel 508
Consider for example the derivative
∂2 A ∂2 A
= .
∂T∂V ∂V∂T
Evaluation of the left-hand-side yields (∂P/∂T)V while for the right-hand-side we got
(∂S/∂V)T so we got the important identity
" ∂P % " ∂S %
$ ' =$ ' . (3.16)
# ∂T &V ,n # ∂V &T ,n
Of course in both cases the number of particles (moles) is treated as constant. Another
example can be obtained using the Gibbs free energy derivatives with respect to pressure
and temperature. It yields
⎛ ∂V ⎞ ⎛ ∂S ⎞
⎜ ⎟ = −⎜ ⎟ . (3.17)
⎝ ∂T ⎠ P ⎝ ∂P ⎠ T
Combining Eqs. (3.11) and (3.14) one can obtain the differential definition of the
chemical potential in the form
d µ = −sdT + vdP. (3.18)
where s and v are molar volumes and temperature.
Sample problem. Calculate all thermodynamic potentials for ideal gas.

One can use the expression of the chemical potential in terms of the derivative of
entropy with respect to the quantity of matter at fixed internal energy and volume.
Employing the ideal gas equation and assuming U=U0+CVT we get
! dU nRdV $ ! U − nu0 $ !V $
dS = # + & → S = nCV ln # & + nR ln # & + ns0 = ns =
" T V %n " n % "n%
! RT $
= nCV ln (CV T ) + nR ln # & + ns0 .
" P %

Consequently one can express the chemical potential in the form


µ = u0 + CPT − sT = u0 + CPT − Ts 0 (P0 ,T ) + RT ln(P / P0 ). (3.19)
Eq. (3.22) will be used later when considering chemical equilibrium. It is important that
chemical potential depends logarithmically of the pressure of the gas.
Chemical potential is somewhat similar to the electric potential. Indeed the
difference of chemical potentials leads to the current of material similarly to the electric
current due to the difference in electric potentials.

Calculate internal energy and entropy for Van der Waals gas at given temperature
and volume.

This tricky calculation can be performed using the fact the second derivative of
internal energy of Van der Waals gas with respect to volume and temperature is equal to
zero, which justifies separation of two those dependencies. To prove that, consider the
partial derivative of energy with respect to the volume at constant temperature. This
derivative can be evaluated using the first thermodynamic law and entropy differential
definition as
Statistical Thermodynamics, Lecture 1 16
Alex Burin; aburin@tulane.edu; 1-504 862 3574;
Israel 508
⎛ ∂S ⎞ ⎛ ∂S ⎞
dU = −PdV + TdS = −PdV + T ⎜ ⎟ dV + T ⎜ ⎟ dT.
⎝ ∂V ⎠T ⎝ ∂T ⎠V
Consequently, one can express the derivative of interest as
⎛ ∂U ⎞ ⎛ ∂S ⎞
⎜ ⎟ = −P + T ⎜ ⎟ .
⎝ ∂V ⎠T ⎝ ∂V ⎠T
Using the Maxwell identity
⎛ ∂P ⎞ ⎛ ∂S ⎞
⎜ ⎟ =⎜ ⎟ .
⎝ ∂T ⎠V ⎝ ∂V ⎠T
and Van der Waals equation one gets
⎛ ∂U ⎞ TR a
⎜ ⎟ = −P + = 2.
⎝ ∂V ⎠T V −b V
Consequently, the internal energy can be expressed as U=-a/V+U(T), where the second
term does not depend on the volume. Its dependence can be expressed as U=U0+CV(T-
T0)-a/V. Using the definition of entropy derivative with respect to temperature
⎛ ∂S ⎞ CV
⎜ ⎟ =
⎝ ∂T ⎠V T
one can then obtained the expression for entropy in the form S(V, T)=S0+
CVln(T/T0)+nRln((V-b)/(V0-b)).

Sample problems. Derive all Maxwell identities. Prove that CV ≤CP.

Additional requirements for the true entropy maximum should include the certain
conditions for its second derivatives. Indeed, all first derivatives should be equal zero in
the maximum. However this is necessary but insufficient condition for true maximum.
For the single variable function the second derivative in the maximum must be non-
positive. Otherwise this is minimum rather than maximum.
The second order expansion for the entropy change can be evaluated in terms of
thermodynamic variables as
Statistical Thermodynamics, Lecture 1 17
Alex Burin; aburin@tulane.edu; 1-504 862 3574;
Israel 508
1 # ∂(S) & 1 # ∂2 S & 1 # ∂2 S &
Δ 2 S = % 2 ( ΔU 2 + % 2 ( ΔV 2 + % 2 ( Δn 2 +
2 $ ∂U ' 2 $ ∂V ' 2 $ ∂n '
1# ∂ S &2
1# ∂ S & 2
1 # ∂2 S & 1 # ∂2 S &
+ % ( ΔUΔV + % ( ΔVΔU + % ( ΔUΔn + % ( ΔUΔn +
2 $ ∂U∂V ' 2 $ ∂U∂V ' 2 $ ∂U∂n ' 2 $ ∂n∂U '
1 # ∂2 S & 1 # ∂2 S &
+ % ( ΔVΔn + % ( ΔVΔn =
2 $ ∂V∂n ' 2 $ ∂n∂V '
1 # ∂(1 / T ) & 2 1 # ∂(P / T ) & 2 1 # ∂(µ / T ) & 2
= % ( ΔU + % ( ΔV − % ( Δn +
2 $ ∂U ' 2 $ ∂V ' 2 $ ∂n '
1 # ∂(1 / T ) & 1 # ∂(P / T ) & 1 # ∂(1 / T ) & 1 # ∂(µ / T ) &
+ % ( ΔUΔV + % ( ΔVΔU + % ( ΔUΔn − % ( ΔUΔn +
2 $ ∂V ' 2 $ ∂U ' 2 $ ∂n ' 2 $ ∂U '
1 # ∂(P / T ) & 1 # ∂(µ / T ) &
+ % ( ΔVΔn − % ( ΔVΔn =
2 $ ∂n ' 2 $ ∂V '
ΔU *# ∂(1 / T ) & # ∂(1 / T ) & # ∂(1 / T ) & -
= ,% ( ΔU + % ( ΔV + % ( Δn/ +
2 +$ ∂U ' $ ∂V ' $ ∂n ' .
ΔV *# ∂(P / T ) & # ∂(P / T ) & # ∂(P / T ) & -
+ ,% ( ΔV + % ( ΔU + % ( Δn/ +
2 +$ ∂V ' $ ∂U ' $ ∂n ' .
Δn *# ∂(µ / T ) & # ∂(µ / T ) & # ∂(µ / T ) & -
− ,% ( Δn + % ( ΔU + % ( ΔV / =
2 +$ ∂n ' $ ∂U ' $ ∂V ' .
1
= [ ΔUΔ(1 / T ) + ΔVΔ(P / T ) − ΔnΔ(µ / T )].
2

It is convenient to express it in the form


1" "1% "P% " µ %%
Δ 2 S = $ ΔUΔ $ ' + ΔVΔ $ ' − ΔnΔ $ '' =
2# #T & #T & # T &&
1 " ΔT " ΔPT − PΔT % " TΔµ − µΔT %%
= $ − 2 (TΔS − PΔV + µΔn ) + ΔV $ 2 ' − Δn $ '' = (3.20)
2# T # T & # T2 &&
1
= − ( ΔTΔS − ΔPΔV + ΔµΔn ).
2T

To investigate the system stability with respect to its volume and temperature
changes at fixed quantity of matter (Δn=0, something should be fixed because there is
always instability with respect to proportional changes of volume, energy and number of
particles) one can represent second order term in Eq. (3.19) in terms of temperature and
volume changes as
Statistical Thermodynamics, Lecture 1 18
Alex Burin; aburin@tulane.edu; 1-504 862 3574;
Israel 508
1
Δ 2 S = − ( ΔTΔS − ΔPΔV ) =
2T
(3.21)
ΔT 2 ΔTΔV ⎛⎛ ∂S ⎞ ⎛ ∂P ⎞ ⎞ ⎛ ∂P ⎞ ΔV 2
= −CV − ⎜⎜ ⎟ − ⎜ ⎟ ⎟ + ⎜ ⎟ .
2T 2 2 ⎝⎝ ∂V ⎠T ,n ⎝ ∂T ⎠V ,n ⎠ ⎝ ∂V ⎠T ,n 2T
The expression has to be non-positive for true entropy maximum. This leads to three
basic thermodynamics inequalities for diagonal terms
⎛ ∂P ⎞
CV ≥ 0, ⎜ ⎟ ≤ 0. (3.22)
⎝ ∂V ⎠T ,n
Off diagonal term can be evaluated to zero later Maxwell identities.
The common sense treatment of these inequalities is straightforward. Indeed, in
case of a negative heat capacity the small heat transfer from the system of interest to the
outside world will raise the temperature of the system and should stimulate more heat
transfer outside of the system making the present situation unstable. In case of positive
derivative of pressure vs. volume the small increase of volume leads to an increase in the
internal pressure, which will lead to further expansion.
Alternatively one can express Eq. (3.21) in terms of entropy and changes as
ΔS 2 ⎛ ∂V ⎞ ΔP 2
Δ2S = − +⎜ ⎟ (3.23)
2CP ⎝ ∂P ⎠S,µ 2T
Here the new thermodynamic inequalities can be derived including the positiveness of the
heat capacity at constant pressure CP>0.
Sample problem: obtain thermodynamic inequalities for the ideal gas assuming that it
can be characterized by temperature dependent specific heat CV(T) taken at constant
volume.

Table 1. Van der Waals parameters

Boiling temp. (K) a (L2atm/mol2) b (L/mol)


Acetic acid 17.71 0.1065
Acetic anhydride 20.158 0.1263
Acetone 16.02 0.1124
Acetonitrile 17.81 0.1168
Acetylene 4.516 0.0522
Ammonia 4.170 0.0371
Argon (1) 87.302 1.355 0.0320
Benzene (2) 353.24 18.24 0.1154
Bromobenzene 28.94 0.1539
Butane (3) 273 14.66 0.1226
Carbon dioxide (4) 216 3.640 0.04267
Carbon disulfide (5) 319.45 11.77 0.07685
Carbon monoxide 1.505 0.03985
Statistical Thermodynamics, Lecture 1 19
Alex Burin; aburin@tulane.edu; 1-504 862 3574;
Israel 508
Boiling temp. (K) a (L2atm/mol2) b (L/mol)
Carbon tetrachloride (6) 349.87 19.7483 0.1281
Chlorine (7) 239.11 6.579 0.05622
Chlorobenzene 25.77 0.1453
Chloroethane 11.05 0.08651
Chloromethane 7.570 0.06483
Cyanogen 7.769 0.06901
Cyclohexane 23.11 0.1424
Diethyl ether (8) 307.75 17.61 0.1344
Diethyl sulfide 19.00 0.1214
Dimethyl ether 8.180 0.07246
Dimethyl sulfide 13.04 0.09213
Ethane (9) 184 5.562 0.0638
Ethanethiol 11.39 0.08098
Ethanol (10) 351.52 12.18 0.08407
Ethyl acetate 20.72 0.1412
Ethylamine 10.74 0.08409
Fluorobenzene 20.19 0.1286
Fluoromethane 4.692 0.05264
Freon (11) 232.35 10.78 0.0998
Germanium tetrachloride 22.90 0.1485
Helium (12) 4.55 0.03457 0.0237
Hexane (13) 341 24.71 0.1735
Hydrogen (14) 20.25 0.2476 0.02661
Hydrogen bromide (15) 207 4.510 0.04431
Hydrogen chloride (16) 3.716 0.04081
Hydrogen selenide (17) 5.338 0.04637
Hydrogen sulfide (18) 4.490 0.04287
Iodobenzene 33.52 0.1656
Krypton (19) 2.349 0.03978
Mercury (20) 8.200 0.01696
Methane (21) 2.283 0.04278
Methanol (22) 9.649 0.06702
Neon (23) 0.2135 0.01709
Nitric oxide 1.358 0.02789
Nitrogen (24) 1.408 0.03913
Nitrogen dioxide (25) 5.354 0.04424
Nitrous oxide 3.832 0.04415
Oxygen (26) 1.378 0.03183
Statistical Thermodynamics, Lecture 1 20
Alex Burin; aburin@tulane.edu; 1-504 862 3574;
Israel 508
Boiling temp. (K) a (L2atm/mol2) b (L/mol)
Pentane 19.26 0.146
Phosphine 4.692 0.05156
Propane (27) 8.779 0.08445
Silane 4.377 0.05786
Silicon tetrafluoride 4.251 0.05571
Sulfur dioxide (28) 6.803 0.05636
Tin tetrachloride 27.27 0.1642
Toluene (29) 24.38 0.1463
Water (30) 5.536 0.03049
Xenon (31) 4.250 0.05105

4. Phase equilibriums.
At the first order phase transition point several phases of the same matter can
coexist with each other. The stability conditions for the whole material does not change
requiring maximum entropy/minimum energy in the stable state so for two phases in
touch (say ice and water) temperatures, pressures and chemical potentials must be the
same. This results in the Clausius-Clapeyron equation for the line separating phases in the
P-T plane. In terms of chemical potentials one can define equilibrium using equivalence
of the chemical potential differentials along the phase border, which yields

dµ1 = v1dP − s1dT = dµ 2 = v2 dP − s2 dT , (4.1)


where v1, v2 and s1, s2 are molar volumes and entropies of phases 1 and 2, respectively.
The differential of chemical
V potential can be obtained using
differentials and definition of the
Va Gibbs free energy.
One can resolve Eq. (4.1)
for dP/dT which results in the
famous Clausius-Clapeyron
equation
Vb
dP ΔS
= . (4.2)
dT ΔV
There is one more known Clausius-
Clapeyron equation for the vapor
P pressure. This equation expresses
the saturated vapor pressure at
Fig. 1. liquid vapor border as the function
of the temperature. It can be
expressed as
Statistical Thermodynamics, Lecture 1 21
Alex Burin; aburin@tulane.edu; 1-504 862 3574;
Israel 508
# ΔH &
P = P0 exp % − ( (4.3)
$ RT '
assuming that the enthalpy difference of gaseous and liquid matters is approximately
temperature independent. This enthalpy difference is called molar vaporization heat or
enthalpy.

Sample problem. Derive Eq. (4.3).


Sample problem. Is it possible to skate on the dry ice (solid CO2)?
The transition liquid gas can be illustrated by Van der Waals gas phase diagram.
Consider the pressure dependence on the molar volume at fixed temperature, which can
be expressed by the equation
a RT
P=− 2 + . (4.4)
v v−b

The graph shows the pressure volume dependence for CO2. At large volume
RT/v>>a/v2àv>>a/(RT) the gas behave like ideal gas obeying the dependence P=RT/v.
Then if temperature is sufficiently low meaning a/RT >>b the pressure decreases with
decreasing the volume, which conflicts with the thermodynamics inequality (2.11).
Therefore is case of three possible values of volume corresponding to one pressure (the
intersections with the horizontal straight line in the graph) only solutions one the
solutions at the edges are stable, while the solution in the middle is unstable. Increase of
pressure towards this domain leads to the jump from the gas phase (most right solution)
to the liquid phase (most left solution).
Let us first determine the regime where discontinuous transition exists.
Depending on the temperature the graph can have or not have domains where the
horizontal line crosses it more than once. The second line from the top represents the
crossover regime between the two where there is always only one intersection and more
than pone intersections are possible. This crossover line has the inflection point where
both first and second derivatives of pressure over volume are equal zero.
One can determine the crossover line using the inflection point equations
Statistical Thermodynamics, Lecture 1 22
Alex Burin; aburin@tulane.edu; 1-504 862 3574;
Israel 508
2a RT
P' = 3 − = 0,
v (v − b)2
(4.5)
6a 2RT
P '' = − 4 + = 0.
v (v − b)3
Solving two equations together we got v=3b and P=a/(27b2) at the inflection point
realized at the temperature Tc=8a/(27bR). This inflection point is called critical point
where the transition liquid gas is continuous (second order like transition) but still exists.
This temperature is higher than the typical laboratory temperatures. For instance
for water it is given by 656 K, which is much higher than the boiling temperature at the
atmospheric pressure. The pressure of 220Atm should be in the related critical point.
Actual temperature and pressure at the critical point for water T=647K and P=218atm are
not far from Van der Waals theory predictions.
The continuous transition occurring between liquid and gas at the critical point is
the example of the second order phase transition, while the discontinuous transition at
lower temperature represents the first order transition. Consider the discontinuous
transition within the Van der Waals model.
In this case the pressure derivative with the respect to volume in Eq. (4.5)
approaches zero at two positive values of volume approaching first local maximum Pmax
at v=vmax and then local minimum Pmin at v=vmin. At pressures P>Pmax the only liquid
phase is stable and can exist, while at P<Pmin only gas phase is stable and can exist. At the
intermediate pressures Pmin<P<Pmax Eq. (4.4) for the volume has three solution and only
two of them vmin(P) and vmax(P) are stable with respect to the thermodynamic inequality,
while the middle solution is unstable and cannot exist. Considering two stable solutions
one should choose more stable which possesses the minimum Gibbs free energy. The
difference in Gibbs free energies is convenient to express using Helmholtz free energies
as

vmax
ΔG = ΔA + P(vmax (P) − vmin (P)) = − ∫ P(v)dv + P(v
max (P) − vmin (P)). (4.6)
vmin

This difference in Gibbs free energies can be represented as the difference between two
areas formed by intersections of the horizontal line of constant pressure with P(v)
dependence in the graph (B-A). If B>A a liquid phase is more stable because it has
smaller Gibbs free energy. Otherwise the gas phase is preferable. The second phase
possessing larger Gibbs free energy is metastable (undercooled gas or overheated liquid).
It turns out though that overheated liquid boils to form the gas. The pressure where
A+B=0 corresponds to the equilibrium point between liquid and gas phases which can
exist at only one pressure at given temperature.
Equations describing phase transition point for Van der Waals gas can be
formulated algebraically evaluating analytically integrals in Eq. (4.6) and using the
pressure volume relationship as
Statistical Thermodynamics, Lecture 1 23
Alex Burin; aburin@tulane.edu; 1-504 862 3574;
Israel 508
a a #v −b&
ΔG = − − RT ln % max ( + P(vmax − vmin ) = 0;
vmin vmax $ vmin − b '
(4.7)
a RT a RT
P=− 2 + =− 2 + .
vmin vmin − b vmax vmax − b
The solution of these equations for water boiling temperature using matlab code
(T=258.1677865506905;x = roots([P -P*b-Rgas*T a -a*b]); a/min(x)-a/max(x)-
Rgas*T*log((max(x)-b)/(min(x)-b))+P*(max(x)-min(x))) yields water boiling
temperature as low as 258K which disagrees remarkably with its actual value of 373.15
K. For Kr Van der Waals model gives reasonable prediction of 100K boiling temperature
compared to 120K actual temperature at 1 Atm. Why does it fail for water?

5. Chemical equilibrium
The chemical potential is important for considering the chemical reactions in the mixtures
of different materials. The analysis of mixtures is particularly needed to describe
solutions and chemical reactions there. Then one can define an energy as a function of the
quantities ni (moles) of materials U(S, T, N1, N2, … Nk). One can define the chemical
potential for each component similarly to the previous definitions as
⎛ ∂E ⎞
µ i = ⎜⎜ ⎟⎟ . (5.1)
∂N
⎝ i ⎠ S ,V , N k k ≠i
Since for the bulk material under consideration the energy should scale proportionally to
other extensive parameters, U(λ S, λV, λ N1,..λ N k ) = λU(S,V, N1,..N k ). we can similarly
express the energy as
k
U(S,V, n1,..nk ) = TS − PV + ∑ µi ni . (5.2)
i=1
One can similarly define the Gibbs free energy as
k
G(T, P, n1,..nk ) = U − TS + PV = ∑ µi ni . (5.3)
i=1

As was derived earlier at constant temperature and pressure the direction of spontaneous
system evolution is determined by the reduction of the Gibbs free energy. This is what we
need to know to use the Gibbs free energy Eq. (5.3) for understanding reaction
equilibriums.
Consider some reaction occurring at fixed pressure P and temperature T of the
environment. The very general reaction equation can be written in the form
Nr Np

∑ r Rei i → ∑ pi Pri . (5.4)


i=1 i=1
where r, p and Re, Pr stand for stoichiometric coefficients and compound names,
respectively.
The change in the Gibbs free energy during the chemical reaction occurring at
constant pressure and temperature can be expressed in terms of the changes in the
numbers of moles composing the mixture as
Statistical Thermodynamics, Lecture 1 24
Alex Burin; aburin@tulane.edu; 1-504 862 3574;
Israel 508
dG = ∑ µi dni . (5.5)

In the course of chemical reaction the changes in quantities of products and reactants are
proportional to their stoichiometric coefficients with positive signs for products and
negative signs for reactants. Accordingly the change in the Gibbs free energy can be
expressed as
dG ∝ −∑ µi pi + ∑ µ j rj . (5.6)
p p

The reaction proceeds spontaneously if the sum of the product chemical potentials
weighted by stoichiometric coefficients is smaller than that for reactants. In the opposite
case the reaction goes backwards. Chemical equilibrium between products and reactants
takes place if the following equation is satisfied

∑µ p = ∑µ r .
i i i i
(5.7)
p r
This is the condition for the chemical equilibrium.
To apply the concept of the chemical equilibrium to the practical reactions we
begin the consideration with the mixture of ideal gases. The chemical potentials of ideal
gases within the mixture do not depend on each other because of the absence of
interaction between them. As it was derived earlier one can express the chemical
potential of the ideal gas as
µ (T,V ) = µ (T,V0 ) − RT ln(V / n). (5.8)
We did not use the pressure dependence here because the total pressure is not quite
relevant for the individual component. However it is convenient to use partial pressure,
defined for each component i as
n
Pi = P i , n = n1 + n2 +...nk . (5.9)
n
Then the chemical potential of each component can be conveniently expressed as
!P $
µi (T, P) = µi (T, P0 ) + RT ln # i &. (5.10)
" P0 %
If we use the atmospheric pressure for P0 then the first term in Eq. (5.10) is defined at
normal pressure while the second term is expressed in terms of activity ai, which can be
defined for the ideal gas as
P
ai = i . (5.11)
Patm
Consequently the chemical potential can be expressed in terms of activity as

µ (T , P) = µ 0 (T ) + RT ln(ai ). (5.12)
It is important to notice that if activity of certain material goes to zero the chemical
potential approaches negative infinity. This is a consequence of very high entropy for
small density of material, which gives to each individual molecule a large volume to
occupy. Therefore the reaction always equilibrates at finite quantities of participating.
Statistical Thermodynamics, Lecture 1 25
Alex Burin; aburin@tulane.edu; 1-504 862 3574;
Israel 508
If we substitute Eq. (5.12) into the chemical equilibrium equation (5.7) one can express it
in the form
⎛ a pp11 ⋅ a pp22 ... ⎞
− ΔG0 = RT ln⎜ r1 r2 ⎟.
⎜ a ⋅ a ... ⎟
⎝ r1 r 2 ⎠ (5.13)
ΔG0 = ∑ µ i 0 pi − ∑ µ i 0 ri .
p r

This equation expresses the law of mass action. Activities are equal pressures expressed
in atmospheric units. One can introduce the equilibrium constant of the reaction at the
atmospheric pressure as
a pp11 ⋅ a pp22 ... Ppp11 ⋅ Ppp22 ...
K p (T ) = r1 r2 = r1 r2 . (84)
a r1 ⋅ a r 2 ... Pr1 ⋅ Pr 2 ...
Eq. (83) can then be expressed as
(
− ΔG0 = RT ln K p (T ) ..) (5.15)
Sample problem. Write the expression for the equilibrium constant of the reaction
2CO(g)+O2(g)=2CO2(g). What is its absolute value? If we had initially only CO2 at the
atmospheric pressure what would be the equilibrium pressures of gases? What if we had
initially 2 moles of CO and one of O2? What if one mole of each?

One can express the equilibrium constant of the reaction using the concentrations (mole
densities) of reactants and products c=P/(RT), expressed in some standard units of mol/L.
The concentration dependent equilibrium reaction constant can be expressed as
c pp11 ⋅ c pp22 ...
K c (T ) = r1 r2 . (5.16)
c r1 ⋅ c r 2 ...
Sample problem. What is the relationship of Kp and Kc for the reaction given above at
room temperature?

Out of equilibrium we can introduce the reaction quotient approaching the


reaction constant in the equilibrium. It can be written as
Ppp11 ⋅ Ppp22 ...
Q p (T ) = r1 r2 . (5.17)
Pr1 ⋅ Pr 2 ...
at given, non-equilibrium pressures. The current difference of the Gibbs free energies for
reactants and products associated with the reaction evolution from the reactants to the
products can be expressed as
ΔGr = ΔG0 + RT ln(QP ). (5.18)
Remember that ΔGeq = ΔG0 + RT ln(K P ) = 0. If the reaction quotient QP is smaller than
the reaction equilibrium constant KP then the change in Gibbs free energy is negative
with the reaction occurrence and it will proceed toward products until the equilibration. If
QP<KP the reaction will take place backwards. Thus the relationship of the reaction
quotient and equilibrium constant determines its direction.
Statistical Thermodynamics, Lecture 1 26
Alex Burin; aburin@tulane.edu; 1-504 862 3574;
Israel 508
Sample problem. The reaction quotient for the reaction 2SO2(g)+O2(g)à2SO3(g) is 10
at T=960K. Calculate the reaction Gibbs free energy and predict the direction of
spontaneous changes for initial partial pressures of 10-3atm for SO2, 0.2atm. for O2 and
10-4atm for SO3. What would be final partial pressures?
The equilibrium constant of reaction and its molar Gibbs free energy difference
depend on the temperature. The main parameter of interest can be expressed as the ratio
of Gibbs free energy and temperature. As usually we assume the constant pressure. Then
the parameter of interest is characterized by the derivative
" ∂G / T %
$ ' (5.19)
# ∂T &P
which can be evaluated as
⎛ ∂G / T ⎞ 1 ⎛ ∂G ⎞ G ST − G H
⎜ ⎟ = ⎜ ⎟ − 2 = 2
=− 2. (5.20)
⎝ ∂T ⎠ P T ⎝ ∂T ⎠ P T T T
Accordingly the temperature dependent equilibrium reaction constant can be expressed as
⎛ ΔG T ΔH (t) ⎞
K P (T ) = exp ⎜⎜ − 0 − ∫ 2
dt ⎟⎟. (5.21)
⎝ RT 0 T0 Rt ⎠
In case of enthalpy difference weakly dependent on the temperature one can simplify this
result to the form
⎛ ΔG (T ) − ΔH 0 ΔH 0 ⎞
K P (T ) = exp ⎜ − 0 0 − ⎟. (5.22)
⎝ RT0 RT ⎠
Sample problem. Given that ΔH0 has an average value of -69.8kJ/mol over the
temperature range 500-700K for the reaction PCl3(g)+Cl2(g)=PCl5(g) estimate KP at
700K given that KP=0.0408 at 500K.

Das könnte Ihnen auch gefallen