Chapter 5

Simple Applications of Statistical Mechanics

Contents
• Simple Applications of Statistical mechanics
– Partition function and their properties
– Calculations of thermodynamic quantities
– Gibbs paradox
– Validity of the classical approximation
– Proof of equipartition
– Simple applications
– Specific heat of solids
– General calculation of magnetism
– Maxwell’s velocity distribution
– Number of molecule striking a surface
– Effusion 1
 Partition function and their properties
• For a system in contact with a heat reservoir. ensemble representative the
systems are distributed over their accessible states in accordance with the
Boltzmann distribution, pr.
• The sum of the Boltzmann factor over all states, irrespective of their
energy, is called the partition function.
Z  exp( Er )
r

• Even if the system is isolated and has fixed energy, the mean values of
the macroscopic parameters of the system are still related to its
temperature T as though it were in thermal contact with a heat reservoir
of this temperature.
• Thus the calculation is again reduced to the evaluation of the partition
function Z.
• If one knows the particles which constitute the system and the
interactions between them, it is possible to find the quantum as well as
classical states of this system 2
 Calculation of thermodynamic quantities
• Let us take an ideal gas of non-interacting atoms and calculate
the partition function (classical approach)
• Consider a gas consisting of N identical monatomic molecules
of mass m enclosed in a container of volume V.
• Let us denote the position and momentum vectors of the ith
molecule by ri and pi, respectively. Then the total energy is
2
N
pi
E   U ( r1 ,..., rN )
i 1 2m
• Since the gas is ideal, there are no interatomic forces, and the
total energy is simply the sum of the individual kinetic
energies of the molecules:
N 2
pi
E 
i 1 2m 3
• If the system can be treated in the classical approximation,
then its energy E ( qI., . . , qf, pl, . . . , pf) depends on some f
generalized coordinates and f momenta.
• In this approach, we divide up phase-space into cells of equal
volume h0f.
• Here, f is the number of degrees of freedom, and h0 is a small
constant with dimensions of angular momentum
• If phase space is subdivided into cells of volume hof the
partition function can be evaluated by first summing over the
number (dql … dqf , dpl ,.. . dpf)/hof of cells of phase space
which lie in the element of volume (dql . . . dqf, dpl … dpf) at
the point (q1, . . . q f , Pl , . . . , pf ) and which have nearly the
same energy E(q1, . . . , qf, pl, . . . , pf) and then summing (or
integrating) over all such elements of volume. 4
• Each cell in phase-space corresponds to a different state.
• The partition function is the sum of the Boltzmann factor exp(−βEr)
over all possible states, where Er is the energy of state r.
• Classically, we can approximate the summation over cells in phase-
space as an integration over all 3phase-space.
3 Thus,
3 3
d r ...d r d p ...d pN
Z  ...exp  E 
/ 1 N
3N
1

h0
• where 3N is the number of degrees of freedom of a monatomic gas
containing N molecules.
 Since the position integrals for each r i can be done immediately and
yields the volume of the container, V,

∫d3ri ≡ V
 There are N such kind of integrals

∫d3r1d3r2…d3rN = VN 5
• There are such integrals, so we obtain the factor in the above
expression. Note, also, that each of the integrals over the
molecular momenta in Eq. () are identical: they differ only by
irrelevant dummy variables of integration.
VN     2 3     2 3
Z
h0
3N ...exp    p1  d p1 ... exp  
 2m   
  pN  d pN
 2m  

• It follows that the partition function of the gas is made up of

the product of identical factors: i.e.,

V     2 3
Z  ,   3 exp   
N
 p d p
h0   2m  
• This is the partition function for a single molecule.
6
• The integral in Eq. () is easily evaluated


    2  3     2
    2m  p
exp 

d p  

exp    2m  p x  dp x
   
 
    2     2
exp 

 
 2 m


p y 

dp y exp    2m  p z  dp z
   
• Finally
 
3
    2 3  2m 
exp    2m  p  d p   
• Thus, 3
 
 2m  2
 V  2 
 h0   
3 3  2m  
ln Z  N ln   N  ln V  ln   ln  2  
 one 2can immediately
• From this partition function 2  h0 calculate
  other
physical quantities. Hence the expression for the mean pressure
yields 1  ln Z 1 N
P 
 V V
• which reduces to the ideal gas equation of state
PV  NkT RT 7
• The total mean energy of the gas is given by
 ln Z 3 N 3
E    RT  N 
 2 2
• Where ϵ =3/2kT is the mean energy per molecule.
• Note that the internal energy is a function of temperature alone, with
no dependence on volume.
• For 𝝂 is the number of moles and N, is Avogadro's number. the
molar specific heat at constant volume of a monatomic gas is
1  E  3
cV      R

  T  v 2
• where R = Nk is the gas constant. so the mean energy can be written

E cV T

8
• Let us now use the partition function to calculate a new result.
The entropy of the gas can be calculated quite simply from the
expression

S k ln Z   E 
• Thus

 
S R ln V  3 ln T   ,
2
3  2mk  3
  2 ln  2 

 ho  2

• The above expression for the entropy of an ideal gas is

certainly new.
• Unfortunately, it is also quite obviously incorrect!
9
 Gibb’s paradox

• First of all, let us be clear why is incorrect. We can see that S → −∞ as T → 0,

which contradicts the third law of thermodynamics. However, this is not a problem.
The equation was derived using classical physics, which breaks down at low
temperatures. Thus, we would not expect this equation to give a sensible answer
close to the absolute zero of temperature. The above equation is wrong because it
implies that the entropy does not behave properly as an extensive quantity.
• Thus, if we double the size of a system then we expect the entropy to double as
well. Suppose that we have a system of volume V, containing ν moles of ideal gas at
temperature T. Doubling the size of the system is like joining two identical systems
together to form a new system of volume 2 V, containing 2 ν moles of gas at
temperature T. Let

• denote the entropy of the original system, and let

• denote the entropy of the double-sized system. Clearly, if entropy is an extensive

quantity then we should have
• But, in fact, we find that
10
• So, the entropy of the double-sized system is more than double the entropy of the
original system. Where does this extra entropy come from?
• A paradox arises if we try to treat molecules as if they were distinguishable. This is
called Gibb’s paradox, after the American physicist Josiah Gibbs, who first
discussed it. The resolution of Gibb’s paradox is quite simple: treat all molecules of
the same species as if they were indistinguishable.
• In our previous calculation of the ideal gas partition function, we inadvertently
treated each of the N molecules in the gas as distinguishable.
• Because of this, we over-counted the number of states of the system. Because the
N! possible permutations of the molecules amongst themselves do not lead to
physically different situations, and, therefore, cannot be counted as separate states,
the number of actual states of the system is a factor N! less than what we initially
thought.
11
• We can easily correct our partition function by simply dividing by this factor, so that
• This gives

• Or using Stirling’s approximation.

• However, our expression for the entropy S is modified by this additive term. The new
expression is

• This gives

• Where

• It is apparent that the extra term involving avoids the difficulties of the Gibbs
paradox. The entropy S in above equation behaves properly like an extensive quantity;
i.e., it does get multiplied by a factor if both V and N are
• multiplied by .
• It is clear that the entropy behaves properly as an extensive quantity in this Equation.
In other words, it increases by a factor α when ν, V, and N each increase by the same
12
 Validity of the classical approximation
• An approximate criterion for the validity for this classical description can be
obtained by appealing to the Heisenberg uncertainty principle
q ħ
• This relates the uncertainties q and p introduced by quantum effects in any
attempt at simultaneous specification of a position q and corresponding
momentum p of a particle. Suppose that one tries to describe the motion of
the gas molecules by classical mechanics. Denote the magnitude of the
mean momentum of a molecule by and the mean separation between
molecules by . Then one would certainly expect a classical description to be
applicable if
>>ħ
when the above equation implies that quantum mechanical effects are not
important. Equivalently >>ħ expresses the condition that
>>
i.e., that the mean separation between particles is much greater than their
mean de Broglie wavelength =2=
13
• When >> is satisfied so that >> , the quantum description ought to be
equivalent to the motion of wave packets describing individual particles
which move independently in a quasi-classical manner.
• In the opposite limit, where << , a state of the whole gas will be shown in
Chapter 6 to be described by a single wave function which cannot be
decomposed in any simple way; it thus results in correlations between the
motions of the particles even if no forces exist between them.
• The mean intermolecular separation can be estimated by imagining each
molecule at the center of a little cube of side , these cubes filling the
available volume V. Then
N = V or =
• The mean momentum can be estimated from the known mean energy of a
molecule in the gas at temperature T. By = KT
= KT
thus and 14
• This shows that the classical approximation ought to be applicable if the
concentration N/V of molecules in the gas is sufficiently small, if the temperature T
is sufficiently high, and if the mass of the molecules is not too small.
• Numerical estimates Consider, for example, helium (He) gas at room temperature
and pressure. Then one has mean pressure = 760mm Hg dynes/ cm2

Temperature T 3000K; hence kT 4 ergs

Molecular mass m = 7 grams

• The equation of state gives = = 2.5 molecules/ cm3
• by using the above equation thus 34 cm and 0.6 cm
• Here the criterion >> is quite well satisfied, and the classical evaluation of the
partition function ought to be a very good approximation if the indistinguishability
of the particles is taken into account. Most gases have larger molecular weights and
thus smaller de Broglie wavelengths; the criterion >> is then even better satisfied.
15
• On the other hand, consider the conduction electrons in a typical metal.
• In a first approximation, interactions between these electrons can be neglected so
that they can be treated as an ideal gas. But the numerical values of the significant
parameters are then quite different. First, the mass of the electron is very small,
about g or 7000 times less than that of the He atom. This makes the de Broglie
wavelength of the electron much longer,
(0.6 ) cm
• In addition, there is about one conduction electron per atom in the metal. Since
there is roughly one atom in a cube cm on a side,
cm
• This is much smaller than for the He gas case; i.e., the electrons in a metal form a
very dense gas. Hence the criterion >> is certainly not satisfied.
• Thus there exists no justification for discussing electrons in a metal by classical
statistical mechanics; indeed, a completely quantum-mechanical treatment is
16
essential.
 The equipartition theorem
• The internal energy of a monatomic ideal gas containing N
particles is (3/2)NkT.
• This means that each particle possess, on average, (3/2) k T
units of energy.
• Monatomic particles have only three translational degrees of
freedom, corresponding to their motion in three dimensions.
• They possess no internal, rotational or vibrational degrees of
freedom.
• Thus, the mean energy per degree of freedom in a
monatomic ideal gas is (1/2) k T.
• In fact, this is a special case of a rather general result. Let us
now try to prove this.

17
 Equipartition Theorems
• Suppose that the energy of a system is determined by some f
generalized coordinates qf and corresponding f generalized momenta
pf, so that
E E q1 ,..., q f p1 ,..., p f 

• Further more, let us assume that

1. The total energy splits additively into the form
E  i pi  E q1 ,..., p f 

where εi involves only one variable pi, and the remaining part E’
does not depend on pi.
2. The function εi is quadratic in pi, so that
 i pi  bpi
2

18
• What is the mean value of ϵi in thermal equilibrium if
conditions 1 and 2 are satisfied?
• If the system is in equilibrium at the absolute temperature T =
(kB)-l, it is distributed in accordance with the canonical
distribution; the mean value

19
Equipartition Theorem

20
 Simple applications
• Mean kinetic energy of a molecule in a gas
• Consider a molecule in a gas (not necessarily an ideal gas) at
temperature T.
• If this molecule has mass m and a center-of-mass momentum p = mv, its
kinetic energy of translation is

• The eq. contains three quadratic terms, the equipartition theorem allows
one to conclude immediately that

• For an ideal monatomic gas the entire energy is kinetic, so that the mean
energy of gas is simply

• The molar specific heat, at constant volume is then CV = 3/2 Nk. 21

Simple Harmonic Oscillator

1 2 1
KE  m v x  kT
2 2

1 2 1
PE  k x  kT
2 2

E  E  KE  PE kT

22
 Quantum Mechanical Simple Harmonic Oscillator
• Quantum mechanical results for a simple harmonic oscillator
with classical frequency ω=sqrt(k/m): The energy is quantized

E  1
 nn  n    n = 0,1,2,3,..
 2
• Energy levels are E
equally spaced!





23
• The mean energy of the oscillator is then given by

• The partition function for such an oscillator is given by

  
exp     

     2 
Z  exp  n  exp     exp n   
0  2  0 1  exp   

geometric series

1  exp    exp 2    ...


ln Z    ln 1  exp   
2
24
 Harmonic oscillators

• The average energy of the oscillator:

  exp    1 1 
E  ln Z      
 2 1  exp     2 exp   1
• In the limit of high T
 1 1  1
   1 exp   1    E       k BT
k BT  2    
• In agreement with the classical result
• In the limit of low T

 
   1 E
k BT 2

25
 Specific heats of solids
• Consider any simple solid with Avogadro's number N, of
atoms per mole.
• Assume that the heat supplied to a solid is transformed into the
vibrational kinetic & potential energies of the lattice
3N Pi 2 1 2
E    k i qi 
i 1  2m 2 
 Thus the total energy is the same as that of 3N independent one dimensional
harmonic oscillators.
 If the temperature T is high enough so that classical mechanics is applicable (and
room temperature is usually sufficiently high for that), application of the
equipartition theorem allows one to conclude immediately that the total mean energy
per mole is
E 3Nk  T
 E 
 Thus the molar specific heat at constant volume becomes C v   3R

 T v
 The result asserts that at sufficiently high temperatures all simple solids have the
same molar specific heat equal to 3R 26
 Con’d
• The validity of this result was first discovered empirically and is known as the law
of Dulong and Petit.
• One can obtain an approximate idea of the behavior of cv at all temperatures by
making the crude assumption (first introduced by Einstein) that all atoms in the
solid vibrate with the same angular frequency ω.
• Then ki= mω2 for all terms i in eq(), and the mole of solid is equivalent to an
assembly of 3N, independent one dimensional harmonic oscillators.
• These can be treated by quantum mechanics SO that their total mean energy is just 3N times
that of the single Oscillator discussed previously 1 1 
E 3 N   
 2 exp    1
• Hence the molar specific heat of the solid on the basis of this simple Einstein model is given by

 E    E      E 
  
1 2

C v           e k BT

  
 T  v    v  T  k T 2    Cv k B
e 
 v
 B 
2 2
k T 
k BT
1

 2 
Let Einstein   e T

 Cv k B  
e 
2
temperature k T  
T
1
27
If the temperature is so high that kT >> ℏω or T >> θ, then θ/T < < 1 and

for T>> θ
expansion of the exponentials in eq() yields again the classical result. Simply

However, if the temperature is so low that kT << ℏω or T<< θ θ/T >>1, and
Cv = 3R

the exponential factor becomes very large compared to unity. The specific heat
then becomes
•The specific heat at constant volume Cv depends
qualitatively on temperature T as shown in the
figure below.
For high temperatures, Cv (per mole) is
close to 3R (R = universal gas constant. R  2 cal/K- mole).
So, at high temperatures Cv  6 cal/K-mole
The figure shows that
3R Cv = 3R
at high temperatures for all substances. This
is the classical Dulong-Petit law.
Cv
This states that specific heat of a given number
of atoms of any solid is independent of
T, K temperature & is the same for all materials!28
 General calculation of magnetization
• Consider system consisting of N noninteracting atoms in a substance at absolute
temperature T and placed in an external magnetic field H pointing along the z
direction. Then the magnetic energy of an atom be written as

Here μ is thc magnetic moment of the atom. It is proportional to the total

angular momentum and is conventionally written in the form

where is a standard unit of magnet moment( usually the Bohr magneton ,m being
the electron mass) and where g is n number of the order of unity, the so-called g factor of the
atom. By combining one obtain

Since H points in the z direction. In a quantum-mechanical description the values which can
assume are discrete and are given by =m
Where m can take on all values between –J and +J in integral steps; i.e.
m = -J, -J+1, -J+2, …, J-1, J
Thus there are 2J+1 possible values of m corresponding to that many possible projections of the
angular momentum vector along the z axis. By virtue of the possible magnetic energies of the
atom are then

For example if J =1/2 as would be the case for an atom with a single net electron spin, there are
only two possible energies corresponding to m =±1/2 29
• The probability that an atom is in a state labeled m is given by

• The z component of its magnetic moment in the state is equal to

• The mean z component of the magnetic moment of an atom is therefore

Here the numerator can conveniently be written as a derivative with respect to the external
parameter H, i.e.,

is the partition function of one atom. Hence the above equation becomes

30
• To calculate introduce the abbreviation

• which is a dimensionless parameter which measures the ratio of the magnetic

energy gH, which tends to align the magnetic moment, to the thermal energy kT,
which tends to keep it randomly oriented. Thus the equation becomes

• which is simply a finite geometric series where each term is obtained from the
preceding one as a result of multiplication by This can immediately be summed to
give

• This can be brought to more symmetrical form by multiplying both numerator and
denominator by Then

• where we have used the definition of the hyperbolic sine

31
 • The function () thus defined is sometimes called the "Brillouin function.“
• Let us investigate its limiting behavior for large and small values of the
parameter . The hyperbolic cotangent is defined as

Conversely, for y << 1, both and can be expanded in power series. Retaining all terms quadratic
in y, the result is

32
• Applying these results to the function () defined in above equation yields

In the opposite limit where 𝜂 << 1,

Figure below shows how () depends on for various values of J.

Fig. Dependence of the Brillouin

junction () on its argument for
various values of J.
33
• If there are atoms per unit volume, the mean magnetic moment per unit volume
(or magnetization) becomes by

If 𝜂 << 1, the above equation implies that One can write this relation in
the following form:

where the constant' of proportionality, i.e., the susceptibility 𝜒 , is given by

Thus 𝜒 , a result known as Curie's law. In the other limiting case

when

One gets then saturation behavior where each atom has the maximum z component
of magnetic moment, gJ, that it can possibly have.

34
 Maxwell Velocity Distribution
• Laplace claimed that it is possible to know everything about an
ideal gas by knowing the position and velocity precisely
• There are six parameters—the position (x, y, z) and the velocity
(vx, vy, vz)—per molecule to know the position and instantaneous
velocity of an ideal gas.
• These parameters make up 6D phase space
• The velocity components of the molecules are more important
than positions, because the energy of a gas should depend only
on the velocities.
• Define a velocity distribution function = the probability of
finding a particle with velocity between
where

35
 Maxwell Velocity Distribution
• Maxwell proved that the probability distribution function is
proportional to
Therefore .
where C is a proportionality constant and β ≡ (kT)−1.
• Because v2 = vx2 + vy2 + vz2,

• Rewrite this as the product of three factors (i.e. probability

density).

36
• Since the probability is 1 when integrated over entire space, we obtain

Solve for C’

• Thus

• The average velocity in x direction is

• The average of the square of the velocity in x direction is

• Where T is the absolute temperature (temp in C+273), m is the molecular

mass and k is the Boltzman constant
37
 Maxwell Velocity Distribution
 The results for the x, y, and z velocity components are
identical.
 The mean translational kinetic energy of a molecule:

 Purely statistical considerations is good evidence of the

validity of this statistical approach to thermodynamics.
 Note no dependence of the formula to the mass!!

38
 The Maxwell Speed Distribution
• The probability that a molecule has an energy between  and +d is:
1    1   
P d  g  exp   d  g  exp   d
Z1  k BT  nQ  k BT 
3/ 2
1  2m  2 2   
g    2  2 
3D
 1/ 2   1/ 2  3 / 2 nQ P    3 / 2 1/ 2 exp  
4      k T
 B 

   g() g   C  P()
exp  
 k BT  =


0  0  0 
1
Now let’s look at the speed distribution for these particles:   mv 2
2
The probability to find a particle with the speed between v and v+dv, irrespective of the
direction of its velocity, is the same as that finding it between  and +d where d =
mvdv: Pv dv P d P mvdv
1/ 2 3/ 2
 2  1 2  m   mv 2  Maxwell
P v    m  v exp   mv  
3/ 2 2
 4v exp 
2
 distribution
   2   2 k BT   2 k BT 

Note that Planck’s constant has vanished from the equation – it is a classical result.
39
 The Characteristic Values of Speed
 m 
3/ 2
 mv 2 
Because Maxwell distribution is skewed (not symmetric in
Pv    4v exp 
2

 2 k BT   2 k BT  v), the root mean square speed is not equal to the most
probable speed:
The root-mean-square speed is proportional to the
P(v) square root of the average energy:

1 2E 3k BT
E  mvrms 
2
vrms  
2 m m
vmax v vrms v

 dP v  2 k BT
The most probable speed:  dv  0  vmax 
  v vmax m


 m   mv 2  8k B T
The average speed: v v Pv dv   4v exp 
3
 dv 
0  2 k BT  0  2 k BT  m

vmax v vrms  2  8 /   3 1 1.13 1.22

40
 005, MB speed distribution)
• Consider an ideal gas of atoms with mass m at temperature T.
(a) Using the Maxwell-Boltzmann distribution for the speed v, find the corresponding
distribution for the kinetic energy  (don’t forget to transform dv into d).
(b) Find the most probable value of the kinetic energy.
(c) Does this value of energy correspond to the most probable value of speed? Explain.
3/ 2
• (a)  m   mv 2 
dN v   NPv  dv  N  2
 4v exp   dv
 2 k BT   2 k BT 
3/ 2
dv  2 1 2   m  8   1 2
dN    NP  d  v  dv  d   N   exp   d
d Problemm (final 2 2005,
m MB speed
 2 k T
B  m k
 B  T 2 m
 
3/ 2
distribution)
 m  4 2  1/ 2   
P    exp   
 2 k BT 
3/ 2
m  k BT 
P  1  1/ 2    1/ 2     1  1  max 1
(b) 0  exp     exp      0   max  k BT
 2  k BT   k BT   k BT  2 k BT 2

(c) the most probable value of speed  dPv  2 k BT

 dv  0  vmax 
v vmax m doesn’t
correspond
the kin. energy that corresponds to the most probable value of speed  k BT
41
 Number of molecule striking a surface
• Consider an element of area dA of the wall of the container. Choose the z axis so as
to point along the outward normal of this element of area (see Fig.). Consider first
those molecules (in the immediate vicinity of the wall) whose velocity is such that it
lies between υ and υ + d υ. (That is, the velocity is such that its magnitude lies in the
range between υ and υ + d υ; its direction, specified by its polar angle θ (with
respect to the normal, or z axis) and its azimuthal angle ϕ, is such that these angles
lie between θ and θ + d θ and between ϕ and + d ϕ, respectively.)

Fig. Molecules, with velocity between v

υ and υ + d υ, colliding with an element
of area of a wall. (Note that the height bf
the cylinder shrinks to zero as dt → 0.)

• Molecules of this type suffer a displacement υ dt in the infinitesimal time interval

dt. Hence all molecules of this type which lie within the infinitesimal cylinder of
cross-sectional area dA and of a length υ dt making an angle θ with the z axis will
strike the wall within the time interval dt; the molecules lying outside this cylinder
will not.
42
• The volume of this cylinder is dA υ dt cos θ, while the number of molecules per unit
volume in this velocity range is f (υ) . Hence [the number of molecules of this type
which strike the area dA of the wall in time dt]=
• Dividing this by the area dA and the time interval dt, we get for Ф (υ) = the number
of molecules, with velocity between υ and υ + d υ, which strike a unit area of the
wall per unit Ф (υ) = f (υ) cos θ
• Let = the total number of molecules which strike a unit area of the wall per unit
time.
• This is simply given by summing over all possible velocities which. A molecule
can have and which will cause it to collide with the element of area.
• This means that we have to sum over all possible speeds 0 < υ < , all possible
azimuthal angles 0 < ϕ < 2ᴨ, and all angles θ in the range 0 < θ < ᴨ/2. (Those
molecules for which ᴨ/2 < θ < ᴨ have their velocity directed away from the wall
and hence will not collide with it.) In other words, we have to sum over all possible
velocities υ subject to the restriction that the velocity
• component = υcosθ > 0 (since molecules with < 0 will not collide with the element
of area). Thus we have

43
• The results of the above equation are generally valid even if the gas is not in
equilibrium (although f might then also be a function of r and t). But if we consider
a gas in thermal equilibrium, f() = f() is only a function of l l. The element of volume
in velocity space can be expressed in spherical coordinate

• Where is just the element of solid angle. Hence the

ave equation becomes

The integration over φ gives 2ᴨ, while the integral over θ yields the value 1/2.
Hence

This can be expressed in terms of the mean speed already computed in. Thus

44
• since the integration over the angles θ andϕ is just the total solid angle 4ᴨ about a
point. Hence the above equation can also be written

• This rigorous result can be compared with our previous crude estimate in the
equation. We see that the latter was off by a factor of only 2/3.
• The mean speed was already computed from the Maxwell distribution. Combining
this with the equation of state, one obtains for

45
 Effusion
• It is the escape of a fluid from anatomical vessels by rupture or exudation
• It is the flow of a gas through an aperture whose diameter is small as compared with
the distance between the molecules of the gas.
• In physics and chemistry, effusion is the process in which a gas escapes from a
container through a hole of diameter considerably smaller than the mean free path
of the molecules. Such a hole is often described as a pinhole and the escape of the
gas is due to the pressure difference between the container and the exterior. Under
these conditions, essentially all molecules which arrive at the hole continue and pass
through the hole, since collisions between molecules in the region of the hole are
negligible. Conversely, when the diameter is larger than the mean free path of the
gas, flow obeys the Sampson flow law.

46
• If a sufficiently small hole (or slit) is made in the wall of the container, the equilibrium
of the gas inside the container is disturbed to a negligible extent.
• In that case the number of molecules which emerge through the small hole is the same
as the number of molecules which would strike the area occupied by the hole if the
latter were closed off.
• The process whereby molecules emerge through such a small hole is called "effusion.“
• Let us consider the situation when the hole is sufficiently small so that molecules
emerge through the hole by effusion.
• If a vacuum is maintained outside the container, the effusing molecules can be
collimated further by additional slits so that one is left with a well-defined "molecular
beam." Such molecular beams have been extensively used in experimental physics
investigations because they provide one with the possibility of studying individual
molecules under circumstances where interactions between them are negligible.

47
 The number of molecules which have speed in the range between υ and υ + d υ and which
emerge per second from a small hole of area A into a solid angle range dΩ in the forward
direction θ = 0 is given by

Note that this expression involves the factor rather than the factor which occurs in the
Maxwellian speed distribution

Fig. Formation of a molecular beam by effusing molecules.

• The equation ∝ shows that the rate of effusion of a molecule depends on the mass of the
molecule, lighter molecules effusing more rapidly than heavier ones. This suggests the
application of effusion as a method for the separation of isotopes. Suppose that a
container is closed off by a membrane which has very many small holes through which
molecules can effuse.
48
 If this container is surrounded by a vacuum on the outside and is filled with a gas mixture of
two isotopes at. some initial time, then the relative concentration of the isotope of larger
molecular weight will increase in the container as time goes on. Similarly, the gas pumped off
from the surrounding vacuum will be more concentrated in the lighter isotope.
Another example of interest is illustrated in Figure below. Here a container is divided into two
parts by a partition containing a small hole. The container is filled with gas, but one part of the
container is maintained at temperature the other part a t temperature .

Fig. A container divided into two parts by a partition containing a small hole. The gas in the two
parts is at different temperatures and pressures.
• One may ask the following question: What is the relation between the mean gas
pressures and in the two parts when the system is in equilibrium, i.e., when a
situation is reached where neither or , nor the amount of gas in either part, changes
with time?
49
•
• If the linear dimension D of the hole is large (D >> l ) , then the conditions simply
= ; for otherwise the pressure difference would give rise to mass motion of the gas
from one side to the other until the pressures on both sides reach equality.
• But if D << l, one deals wi1.h effusion through the hole rather than with
hydrodynamic flow.
• In this case the equilibrium condition requires that the mass of gas on each side
remain constant, i.e., that [the number of molecules which pass per second through
the hole from left to right] equals [the number of molecules which pass per second
through the hole from right to left].
• By = n this leads to the simple equality
=
By ∝ this condition becomes
=
• Thus the pressures are then not at all equal, but higher gas pressure prevails in the
part of the container at higher temperature.

50
 Pressure and momentun transfer
• It is of interest to consider from a detialed kinetic point of view how a gas exerts a pressure.
• The mean force exerted on a wall of the container is due to the many collision of molecules
with the wall.
• Suppose that we wish to calculate the mean force F exerted by the gas on a small element of
area dA of the container wall. (See Fig., where we have chosen the z axis to be normal to the
element of area.) Then we must calculate the mean rate of change of momentum of this element
of wall, i.e., the mean net momentum delivered to this wall element per unit time by the
impinging molecules.
• If we focus attention on an element of area dA lying inside the gas an infinitesimal distance in
front of the wall, then the above calculation is equivalent to finding the mean net molecular
momentum which is transported per unit time across this surface from left to right as the
molecules cross this surface from both directions.
• Let us denote by G(+) the mean molecular momentum crossing this surface dA per unit time
from left to right, and by G(-) the mean molecular momentum crossing this surface dA per unit
51
time from right to left. Then one has simply
 • To calculate G(+), consider the element of surface dA in the gas and focus first
attention on those molecules with velocity between υ and υ + d υ. (See Fig.1, which
is similar to Fig.,2.)
• The mean number of such molecules which cross this area in an infinitesimal time
dt is again the mean number of such molecules contained in the cylinder of
volume ; i.e., it is equal to f (υ) .

Fig.1 An element of area dA of the Fig.2 Molecules crossing a surface dA in a gas

container wall and a surface of area from left to right (+) and from right to left (-).
dA lying inside the gas just in front of (Note that the height of the cylinders shrinks
the ball. to zero as dt→ 0.) 52
• By multiplying this number by the momentum mυ of each such molecule and
dividing by the time dt, one obtains the mean momentum transported across the area
dA per unit time by molecules with velocity between υ and υ + d υ. By summing
over all molecules which cross this area from left to right, i.e., over all molecular
velocities with > 0, one then gets for the total mean molecular momentum G(+)
transported across this area from left to right the result

• where we have put = υcosθ and where the integration is over all velocities for
which > 0. A similar expression gives the total mean molecular momentum
G(-)transported across this area from right to left, except that the integration must
now be over all molecules for which < 0. Thus

53
• The force from the equation is then given by the net mean momentum transported
across the surface, i.e., by subtracting the second equation from first equation . But
in first equation , where the integration is only over positive values of , one can
simply put = in the integrand. In second equation , where the integration is only
over negative values of , one can put = - in the integrand. Hence first equation
gives simply

where the two integrals have been combined into a single integral over all possible
velocities. This equation is a very general expression and would be valid even if the gas
is not in equilibrium, i.e., even if f is completely arbitrary.
If the gas is in equilibrium, then f(υ) is only a function of υ = l υ l. Note first that

54
• since the integrand is odd, having opposite signs for + and -. This equation
expresses the obvious fact that there can be no mean tangential force on the wall in
an equilibrium situation. The mean normal force does, of course, not vanish.
Measured per unit area it gives the mean pressure, which is thus, by the above
equation, equal to

Here we have used the definition

55
• Hence the above equation can be written equivalently as

• This agrees substantially with our crudely derived result with the first equation
(except that the averaging is now done carefully so that what appears is rather
than ).
• Since is related to the mean kinetic energy K of a molecule, the above equation
implies the general relation

• i.e., the mean pressure is just equal to 2/3 the mean kinetic energy per unit volume
of the gas.
• Up to now we have not yet made use of the fact that the mean number density of
molecules f (υ) is given by the Maxwell velocity distribution.
• This information allows us to calculate explicitly and is equivalent to using the
equipartition theorem result that = kT. Then the above equation becomes

= nKT
so that one regains the equation of state of a classical ideal gas.

56