Professional Documents
Culture Documents
PHY3106
PHY3106
DEPARTMENT OF PHYSICS
Andama Geoffrey
CONTENTS PHY3106: STATISTICAL MECHANICS
Contents
1 Introduction statistical mechanics 4
1.1 Thermal physics, thermodynamics and statistical mechanics . . . . . . . . . . . . . . . . 4
1.2 Macroscopic point of view . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.3 Microscopic point of view . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.4 Microstate, macrostate and ensemble . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
1.5 Fundamental assumption of statistical mechanics . . . . . . . . . . . . . . . . . . . . . . 6
1.6 Phase space . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
1.7 Treatment of ensembles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
1.8 Ensemble average . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.8.1 Average values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.8.2 Density distribution in the phase space . . . . . . . . . . . . . . . . . . . . . . . 9
Course Description
This course covers the fundamental postulates of statistical mechanics, which include quantum states,
systems in contact, the different particle statistics and their applications, equipartition of energy and
specific heat capacities of diatomic gases.
Course Objectives
By the end of the course, the teacher trainee should be able to:
Course outline
1. Introduction of statistical mechanics
3. Two systems in thermal contact. Two systems in diffuse contact. Two systems in thermal and
diffuse content.
5. Fermi-Dirac (F – D) statistics: Electronic specific heat capacities of metals, white dwarfs, thermionic
effect.
Mode of delivery
Blended learning, Lectures, assignments, tutorials, Peer discussions practicals/project work
Mode of Assessment
At least two take home assignments and two tests which constitute 20% of the total mark, at least two
practicals which constitute 20% of the final mark, final examinations which constitute 60% of the total
mark
Reading List
1. Statistical Mechanics by Agsesl, B.K. & Melvin E. (1988)– John Wiley & Sons.
2. Thermal Physics by Kittel, C & Kroemer, H. (2000) – Freeman & Co. NY.
3. Heat and Thermodynamics by Zeemansky M.N. & Dittman, R, (1997) – 7th Ed. McGraw Hill,
Book, Co. Singapore
• The book you are reading, the chair on which you are sitting, the air you are breathing.
• It is about things that can be hot or cold; hard or soft; solid, liquid, or gas.
• It is about machines that work for you: automobiles, heaters, refrigerators, air conditioners.
• Thermal physics is usually divided into two main topics: thermodynamics and statistical me-
chanics.
Caret-right Thermodynamics
• It provides powerful methods that connect observable quantities by equations that are
not at all obvious, but are nevertheless true for all materials.
• Particle interaction
• Basis of thermodynamics
– It provides a foundation for thermodynamics and the ultimate justification of why thermo-
dynamics works.
• Material properties.
– It also provides a way to calculate the properties of specific objects, instead of just the
universal relationships provided by thermodynamics.
• Atomic hypothesis
– Statistical mechanics poses the problem of deducing macroscopic properties of matter from
the atomic hypothesis.
– According to the hypothesis matter consists of atoms or molecules that move subject to the
laws of classical mechanics or of quantum mechanics.
– A microscopic state is described by specifying, at a given instant, the value of positions and
momenta (or, equivalently, velocities) of each of the N particles. Hence one has to specify
3N + 3N coordinates that determine a point in phase space, in the sense of mechanics.
• Probabilistic predictions
– Statistical mechanics is the subject area that deals with the analysis of the
internal motions of a many-particle system using probability theory.
– The goal of statistical mechanics is both to understand and to predict macro-
scopic behavior given the properties of the individual molecules comprising
the system, including their interactions.
• Composition, N, µ
• Volume, V
• Pressure, P
• Temperature, T
Caret-right These quantities are called macroscopic quantities because they refer to the gross or large-scale
properties of the system.
(iii) They are suggested more or less directly by our sense of perceptions.
• The molecules are assumed to interact with one another by means of collisions or by forces at a
distance.
• The system is then said to reach an equilibrium when most of the molecules are in a given energy
state.
(i) Assumptions are made concerning the structure of matter, for example, the existence of molecules
is assumed.
(ii) Many quantities must be specified. For example, each molecule has its own energy state and its
own coordinates.
(iii) The quantities specified are not suggested by our sense of perceptions.
Caret-right Note: An ensemble is a (conceptual) collection of macroscopic systems. It is not a collection of atoms
to make up a macroscopic system.
Caret-right A microstate is also called a “configuration” or a “complexion”. . . both poor terms. A macrostate
is also called a “thermodynamic state”.
Caret-right Hence we can refer to ensemble as a collection of the different ”thermodynamic states” of a macrostate
(loosely speaking, a system).
Caret-right Most of the systems you deal with in classical mechanics classes, for example, are isolated. A system
is said to be “isolated” if no energy goes in or out, and if the mechanical parameters (N , V , etc.)
are also fixed.
An isolated system in an equilibrium macrostate has equal probability of being in any of the
microstates corresponding to that macrostate.
Caret-right As mentioned above, we study physical systems consisting of very large number of particles N ∼ 1023 .
Caret-right From macroscopic point of view the system appears to be a continuum because of this large number
of particles.
Caret-right We assign a complete set of thermodynamic variables to the system that characterise its equilibrium
state. The complete set of thermodynamic variables we can assign are the energy E, volume V and
the number of molecules N .
Caret-right From microscopic point of view, the matter (gas) consists of discrete particles, like atoms or molecules.
Caret-right In classical mechanics, the microscopic description will specify both the momenta (or velocities) and
the positions of all the particles in the gas at a given time.
Caret-right Clearly, this is impractical, considering the very large number of particles that will require solving an
enormous number of equations of motion involving collisions.
Caret-right We can get out of this problem by relating the macroscopic descritption based on a few variables
with the microscopic description based on a large number of variables.
Caret-right This can be accomplished using (i) kinetic theory or (ii) statistical mechanics.
Caret-right At time t the position of the bead is x(t) and the velocity is vx = ẋ with momentum px = mẋ at that
instant.
Caret-right The state of the bead at any instant can be represented by a point P in a hypothetical two-dimensional
space, called the phase space, whose coordinates are x and px .
Caret-right As the bead moves on the string, the value of x changes. If the bead accelerates, px also changes.
Caret-right As a result the point P traces a trajectory in the phase space with the passage of time.
Caret-right Let us now consider a molecule of an ideal gas. This molecule has three degrees of freedom because
its motion can be described in x, y and z directions on the Cartesian coordinates.
Caret-right Hence in Cartesian coordinates, the positions can be described as x1 , x2 , x3 and the corresponding
momenta are p1 , p2 , p3 . Hence the phase space has six (6) dimensions which are paired as (x1 ,p1 ),
(x2 ,p2 ) and (x3 ,p3 ).
Caret-right For a system of N particles, the instantaneous state is represented by a set of N points in the
phase-space, one for each particle.
Caret-right The total number of translational degrees of freedom is 3 × N = 3N . Then the µ-space has 6N
dimensions for all the N particles.
Caret-right For N particles, the phase-space with 6N dimensions is called the Γ-space, which is spanned by 3N
coordinate axes and 3N momentum axes.
Caret-right Gibbs replaced this time dependent picture by a static picture. In this new picture, the entire
phase-space exists at one time, as shown in Figure 1(a).
Caret-right In this picture, each phase point represents a separate system with the same macroscopic properties
(N , V , E) as the system of interest but a different microscopic state.
Caret-right For this system, instead of taking time average, we take an average over the group existing simulta-
neously at one time.
Caret-right Gibbs called such a group of replicas an ensemble, which is a collection of similar, noninteract-
ing, independent, imagined systems, as shown in Figure 1(b).
Caret-right Each member of an ensemble is called an element. All the elements are identical in the macroscopic
features N , V , E. This means that they have the same macroscopic state.
Figure 1: (a) The ensemble (a small portion) at one time. Any region like A will appear to contain a
swarm of phase points. (b) Schematic lattice representing an ensemble of M (M ∞)
imaginary systems, at one time, each with same N , V , E.
Caret-right Although all the elements of an ensemble are in the same macroscopic state, they differ from one
another in the coordinates and momenta of the individual molecules or particles. In other words,
each element has different microscopic states, that are randomized.
There is a clear difference between the actual system of interest and an element of the ensemble.
The actual system of interest is the physical object about which we wish to make predtions.
On the other hand, the elements of the ensemble are the mental copies of the actual system.
These mental copies enable us to use probability theory to make predictions about the actual
system of interest instead of making such prediction based on the microscopic properties, which
would otherwise be an impractical task.
Caret-right An ensemble average is the average of over all the elements in an ensemble at a fixed time.
Caret-right Consider the simple case of a set of N points distributed arbitrarily along a line.
Caret-right If x(i) is the distance of the ith point from the origin, then the average distance x̄ from the origin is
given by
N
1 X
x̄ = x(i). (1)
N
i=1
Caret-right If the line is divided into cells and Nl is the number of points in the ith cell located at x(i), then we
can write
1 X
x̄ = Nl x(i). (2)
N
l
Caret-right If the distribution is known in the form of a continuous function N (x), then
Z +∞ Z +∞
1
x̄ = xN (x)dx, N= N (x)dx. (3)
N −∞ −∞
Caret-right In general, if R(x) is any arbitrary property of the points, then we have
Z +∞
R(x)N (x)dx
R̄ = −∞
Z +∞ (4)
N (x)dx
−∞
• The aim is to know only the number of systems or elements that would be found in different
states at any time. That is, in different regions of the Γ-space. Since all the elements are similar
in structure, we need not distinguish between them.
• The number of elements in the ensemble is so large (M → ∞) that there is a continuous change
in their number in passing from one region of the pase space to another.
Caret-right Therefore, the condition of an ensemble can be described by a density D with which the phase points
are distributed in the Γ-space.
Caret-right If an ensemble has f degrees of freedom, then then probability density D is a function of 2f position
and momentum coordinates:
Caret-right These coordinates can explicitly depend on time t, in which case the probability density will also
depend on time.
Caret-right Consider a small region A of the Γ-space such that the position coordinates lie between q1 and q1 +dq1 ,
..., qf and qf + dqf , and let the momenta lie between p1 and p1 + dp1 , ..., pf and pf + dpf , as shown
in Figure 1(a).
Caret-right By defintion, the density D is the number of systems or elements dM lying in the specified infinitesimal
Caret-right If M is the total number of elements in the phase space, then at every instant t, we have
Z
M = D(q, p, t)dΓ, (10)
Caret-right If we select a system at random from the ensemble, the probability of selecting one whose phase point
lies in the small region at the points q1 , ..., pf is ρdΓ where
Z
D D
ρ= Z = , ρdΓ = 1 (12)
M
DdΓ
Caret-right In terms of the normalised probability distribution ρ(p, q, t) the ensemble average is given by
Z
R(q, p)ρ(q, p, t)dq dp Z
R̄ = Z = RρdΓ (13)
ρ(q, p, t)dq dp
Caret-right The ensemble average in Equation (13) gives the average value of the physical quantity R for the
actual system of interest.
(i) We can proceed directly with mechanical properties such as pressure, energy, volume, num-
ber of molecules, etc. These mechanical properties can be defined using quantum or classical
mechanics.
(ii) We can proceed indirectly with non-mechanical properties such as such as temperature, entropy,
free energy (Gibbs or Helmholtz), chemical potential, etc.
Caret-right As an example let us consider pressure which is a typical mechanical variable. If we wish to calculate
pressure on the walls of the system from a molecular point of view, we have to calculate the force
exerted by each molecule, using classical or quantum mechanics. However, this force changes with
time which we have to account for in our calculation. This means we have to calculate the average
force over a period of time. The biggest problem we encounter in these calculations is that there is
tremendous number of molecules in the system which we have to deal with. Moreover, these molecules
interact with each other, which compounds the problem.
Caret-right From the above example, the calculation of pressure based on molecular properties is out of question
either in quantum or classical mechanics.
Caret-right In this case, we are forced to adopt an alternative procedure, which is called the ensemble method of
Gibbs, discussed earlier.
Caret-right The ensemble method is based on postulates that connect the desired time average of a mechanical
variable with the ensemble average of the same variable.
2.1.1 Postulate 1:
Caret-right This postulate tells us that we may replace the time average of the actual system by the instantaneous
average over a large number of systems that ”represent” the actual system.
(i) Firstly, this postulate requires that the original system that the ensemble represents must be
in equilibrium. This means the ensemble average of R in the limit as M → ∞ must be
independent of time.
(ii) Secondly, in order to actually compute an ensemble average, we need some information about
the relative probability of occurrence of different quantum states in the systems of the ensemble.
This is provided in the second postulate.
Caret-right The first postulate is applicable to the three most important thermodynamic environments we shall
consider and these include the following:
(iii) An open, isothermal system, where µ, V, T are given (grand canonical ensemble). Here
µ is the chemical potential.
2.1.2 Postulate 2
Caret-right This postulate tells us that each quantum state is represented by the same number of systems in the
ensemble.
Caret-right OR, if a system is selected at random from the ensemble, the probability that it will be found in a
particular quantum state is the same for all the possible quantum states.
Caret-right The second postulate is often called the the principle of equal a priori probabilities.
Caret-right If we combine the second postulate with the first postulate, an isolated system of actual interest will
spend equal amounts of time in each of the available quantum states, over a long period of time. This
is referred to as the quantum ergodic hypothesis.
Caret-right The value of E in the second postulate is one of the energy levels of the quantum-mechanical system
defined by N and V .
Caret-right Because N is extremely large, the energy levels of such a system will be very close, with very high
level of degeneracy. These energy levels are practically continuous.
Caret-right We shall denote the number of quantum states for such a quantum-mechanical system with N and
V associated with the energy E by Ω(N, V, E).
Hψ = Eψ (14)
Caret-right In other words, the Ω quantum states in connection with the second postulate refer to some set of
orthogonal wave functions ψ all belonging to the same E.
Caret-right The wave function representing the actual quantum-mechanical state of any system selected from the
ensemble will be some linear combination of all the wave functions.