Professional Documents
Culture Documents
Ludwig Boltzmann, who spent much of his life studying statistical mechanics,
died in 1906 by his own hand. Paul Ehrenfest, carrying on his work, died
similarly in 1933. Now it is our turn to study statistical mechanics.
Perhaps it will be wise to approach the subject cautiously.
2
then there was the realization that one does not
always need to know the cause (all causes), can
do statistical analyses instead
Typical problem, flipping of 100 coins,
One can try to identify all physical condition before the toss, model the
toss itself, and then predict how the coin will fall down
if all done correctly, one will be able to make a prediction on how many
heads or tails one will obtain in a series of experiments
Statistics and probabilities would just predict 50 % heads 50% tails (for
a fair coin) by ignoring all of that physics,
The more experimental trials, 100,000 coin tosses, the better this
prediction will be borne out (provided the coin is fair)
The field of statistical (macroscopic) physics that is concerned with
3
the effects of many particle at once
Path to statistical physics from classical to quantum for
bosons and fermions
Benjamin Thompson (Count Rumford) 1753 – 1814
Put forward the idea of heat as merely the kinetic energy of individual
particles in a gas or condensed matter. Ignored by vast majority for a
very long time! One really need experiments to convince people!!
James Prescott Joule 1818 – 1889
Demonstrated the mechanical equivalent of heat, so central concept of
thermodynamics becomes internal energy of systems (many many particles at
once)
4
Speed distribution of
particles in an ideal gas
in equilibrium at one
temperature, instead of
analyzing what each
individual particle is going to
do, we determine the
density of states
(degeneracy) and a
distribution function of the
property, and then calculate
the expectation value of the
physical properties of the
system (always by the same
procedures !!)
We can then predict
<KE>= <p2>/2m
what we will measure
<KE> = 3/2 kT
There is one characteristic kinetic energy (and speed) distribution as a function of T, so
we would like to have a function that gives these distribution for all temperatures !! 5
Beyond first and second year college physics
James Clark Maxwell 1831 – 1879, Josiah Willard Gibbs 1839 – 1903, Ludwig
Boltzmann 1844 – 1906 (all believed in reality of atoms, tiny minority at that time)
Gibbs invents notation for vector calculus, the form in which we use Maxwell’s
equations today
Maxwell’s electromagnetic theory succeeded his work on statistical foundation of
thermodynamics – so he was a genius twice over.
6
and then there came modern physics …
Einstein 1905
PhD thesis, the correct theory of Brownian motion, a theory that
It’s similar to the product of a wavefunction with its complex conjugate (in
3D), from that we can calculate expectation values (what is measured
on average) by an analogous integration procedure as in previous chapters
on quantum mechanics !!
8
Maxwell Velocity Distribution
Maxwell proved that the velocity distribution function is
proportional to exp(−½ mv2 / kT), special form of exp(-E/kT) –
the Maxwell-Boltzmann statistics distribution function of classical
distinguishable particles, probability that a certain state of energy
E is occupied
“… it is only slightly overstating the case to say that physics is the study of
symmetry.” Philip Anderson, Nobel Prize 1977, 1/3 for “for their fundamental theoretical
investigations of the electronic structure of magnetic and disordered systems”
Pierre Curie’s principle; physics effect due to internal symmetry breaking in crystals, 1903 Nobel
prize ½ with his wife and ½ Becquerel (radioactivity)
Anisotropy and periodic translation symmetry in crystals, conservation of structural and physical
information in the translations periodic unit cell
Anisotropy and space group symmetry in crystals, conservation of structural and physical
information in the asymmetric part of the translation periodic unit cell (also called asymmetric unit)
https://www.youtube.com/watch?v=04ERSb06dOg&list=PLsPUh22kYmNAnurqRx9JlDGukZCsYcLhF,
9:40 min, symmetry group of the sphere is incorrect 11
Expectation value for Velocity2 = 0
The mean value of vx2, also an expectation value that is a simple
function of x
13
9.3: Equipartition Theorem
For a system of classical particles (e.g. atoms or molecules in a dilute
state so that they are distinguishable) in equilibrium, a mean energy
of ½ kT per system member is associated with each independent
quadratic term in the energy of the system member.
14
Equipartition Theorem
In a monatomic ideal gas, each molecule has
NAvogadro kB = R
We will need often
CV = 3/2 N k - for any number of particles N the number of
For the heat capacity for 1 mole particles, often also
with energy in
interval E and E +
dE as a distribution
The ideal gas constant R = 8.31 J/K function, density of
energy states
15
As predicted, only 3
translational
degrees of freedom
2 more (rotational)
Causes degrees of freedom
global
warming
2 more (vibrational)
degrees of freedom,
which also adds two
times 1/2 kBT
Geo- discrepancies due to quantized vibrations, not due to high particle
engineering density as this is all about gasses, i.e. distinguishable particles
We get excellent agreement for the noble gasses, they are just single atom-particles
16
and far away from other particles, no overlap in wavefunctions, classical physics
“… basic difference between the care
economists bring to statistics and the
seeming carelessness of the physicists. For
example, when the Boyle law was found to
differ from the facts, the physicists simply
invented the concept of a “perfect gas,” that
is, a body that follows Boyle's law perfectly.
Naturally, perfect gas approximations are
absurd in some problems but are adequate in
many others, and they are so simple that one
must consider them first".
Benoȋt Madelbrot, New methods in statistical economics, The Journal of
Political Economy 71 (1963) 421–440.
kT
19
The Rigid Rotator Model
For diatomic gases, consider the rigid rotator model.
20
Justification for two rotational degrees of freedom for
dumbbell shaped molecules
In the quantum theory of the rigid rotator the allowed energy levels
are
22
six degrees of freedom
not that simple at high temperatures
Condensed matter rather than dilute gas
6
/2 kT ?
We will revisit this problem when we have learned about quantum distributions, concept of
phonons, which are quasi-particles that are not restricted by the Pauli exclusion principle 23
Maxwell Speed Distribution ΙvΙ derived from Velocity
Distribution as illustration of the procedure
Maxwell velocity distribution:
It is only going to be valid in the classical limit, as a few particles would have speeds in
excess of the speed of light. But there are very very few of those particles
26
Maxwell Speed Distribution
What we were looking at on slide 5
F(v) goes to zero at infinity, very very few particles have speeds higher than c 27
Maxwell Speed Distribution
Most probable speed (at the peak of the speed distribution), take the first
derivative of the speed distribution function and set it zero, derive the
consequences:
28
average (mean) of the square of the speed, will be an expectation value that we
calculate from another integral on the basis of the speed distribution function
<KE> = 3/2 kT
Slide 32
We define root mean square speed on its basis
So now we have the whole function, can make calculations for all T !!! 29
Maxwell’s speed distribution experiments
31
Number of So we recover the equipartition theorem
particles with for a mono-atomic gas, it’s what we
energy in wanted to know on side 5 so that we
interval E and could make prediction on other physical
E + dE entities about what is going to be
measure on average in equilibrium (a
stationary state)
Is temperature
dependent as it should
according to Count
Rumford, slide 4
are not subject to the Pauli exclusion principle, i.e. behave differently
There are only certain energy values allowed for bound systems in quantum
mechanics, so energy states will be a series of discrete values, rather
than a continuous function
33
Classical physics Distributions
Boltzmann showed that the statistical factor exp(−βE) is a characteristic of any
classical system in equilibrium (in agreement with Maxwell’s speed
distribution) quantities other than molecular speeds may affect the energy of a given state
(as we have already seen for rotations, vibrations)
Maxwell-Boltzmann statistics for classical system: β ≡ (kBT)−1
A is
n(E) dEa=normalization
the number offactor, problem
particles with specific
energies between E and E + dE.
State 1 State 2
AB 1/4
A B
B A
AB
There are four possible states the system can be in, each of them is
equally likely.
35
Quantum Distributions
If the two particles are indistinguishable:
State 1 State 2
A and B in XX
previous X X
1/3
slide
XX
There are only three possible states of the system.
36
6 particles distributed over 9 energy states
37
Quantum Distributions
Fermi-Dirac distribution: different energy states have different degeneracies
Different version 38
Classical and Quantum Distributions
For
photons in
cavity,
Planck’s
black
body
radiation,
B2 = 1
E is quantized in units of hf
if part of a bound system
e0 = 1
39
Quantum Distributions If all three
normalization
has to do with factors = 1,
specific just for
normalization comparison
factor
The exact forms of normalization factors for the distributions depend on the
physical problem being considered.
Because bosons do not obey the Pauli exclusion principle, more bosons can
fill lower energy states (are actually attracted to do so per minimization of total E)
All three graphs coincide at high energies – the classical limit.
Maxwell-Boltzmann statistics may be used in the classical limit when
particles are so far apart that they are distinguishable, can be tracked by their
paths, even when densely packed, at high T we recover classical physics 40
When there are so many
states that there is a very low
probability of occupation, high
temperatures
https://www.youtube.com/watch?v=2Z6UJbwxBZI, 2 min 43
degeneracy of
the first exited
state in H atom
n l ml ms up ms down
2 0 0 +1/2 -1/2
2 1 1 +1/2 -1/2
2 1 0 +1/2 -1/2
g(E) is the statistical weight of the state with energy E 2 1 -1 +1/2 -1/2
g functions here: how many states there are per unit energy value, in other words:
the degeneracy if we talk about many hydrogen atoms
(moderate to high T)
hetero-polar bond in diamond much stronger than metallic bond in lead and
aluminum, so much larger Einstein Temperature for diamond (1,320 K) >> 50 to
100 K for typical metals 46
Peter Debye lifted in 1912
the assumption that atoms
vibrate independently,
similar statistics, Debye
temperature TD
47
https://
www.youtube.com/
watch?v=le_ORQZzkmE
quasiparticles in
semiconductors,
phonons, Cooper pairs,
14 min
collective motions called phonons (Public Domain; Sean Kelley via NIST)
48
Blackbody Radiation
Blackbody Radiation
Power density of the emitted radiation
49
Phase space again, this time is how many “momentum states” are
there in the whole volume
1/8 comes from the restriction to positive values of ni and 2 comes from the
fact that there are two possible photon polarizations.
Rewrite Energy equation in terms of r,
For photons, the normalization factor is 1, they are created and destroyed as needed
51
Bose-Einstein Statistics
Convert from an energy “number” distribution to an energy
“density” distribution u(E).
3
L3 cancels, so
the shape of
For all photons in the range E to E + dE the cavity does
not matter
and worldwide fame for Satyendra Nath Bose 1894 – 1974 but he didn’t get a Nobel
52
Liquid Helium
Has the lowest boiling point of any element (4.2 K at 1 atmosphere
pressure) and has no solid phase at normal pressure.
The density of liquid helium as a function of temperature.
54
Liquid Helium
where σ is the surface tension (SI units: N/m), g is gravitational acceleration (m/s 2), ρ is
density (kg/m3), β is inclination angle (rad), and ν is kinematic viscosity (m 2/s).
59
Fermi-Dirac Statistics
T=0 T>0 ½
T = TF T >> TF
When T >> TF, FFD approaches a decaying exponential of the Maxwell Boltzmann
statistics. At room temperature, only tiny
= Fermi energy / kB
amount of fermions are in the region
around EF,i.e. can contribute to
electric current, heat capacity, …60
In metals, we have
a free quantum
electron gas of
fermions
(electrons)
A mono-atomic classical
gas has a molar heat
capacity of 3/2 kT NA
and very low density
61
Classical Theory of Electrical Conduction
Paul Drude (1900) showed on the basis of the idea of a free
electron gas inside a metal that the current in a conductor should
be linearly proportional to the applied electric field, that would be
consistent with Ohm’s law, an empirical observation
His prediction for electrical conductivity:
62
Classical Theory of Electrical Conduction
Based on Maxwell’s speed distribution, as derived on
slide 26
63
Quantum Theory of Electrical Conduction
At T = 0
65
Quantum Theory of Electrical Conduction
In general,
At T > 0
Where α is a constant π2/4.
Product of near
step function for FFD
with density of state
function g(E)
66
Quantum Theory of Electrical Conduction
With the value TF ≈ 80,000 K for copper, we obtain cV ≈ 0.02 R,
which is consistent with the experimental value! Quantum theory
has proved to be a success. For cV of crystalline metals, we
can ignore contribution of the free electron gas to the molar
heat capacity, Dulong-Petit law cV_crystal ≈ 3 R at moderate to
higher temperature
Replace mean electron speed on slide 61 by Fermi speed uF
defined from EF = ½ m uF2.
Conducting electrons are loosely bound to their atoms.
these electrons must be at the highest energy levels
at room temperature the highest energy level is close to
the Fermi energy.
Fermi speed to be used instead of
68
Quantum Theory of Electrical Conduction
Drude thought that the mean free path could be no more than a
few tenths of a nanometer, but it is much longer than his simple
estimation, because electrons have wave-particle duality
Exact quantum calculations will get it exactly right, the T-1 dependency of σ is approx.
69
) Like an atom
) 70
Fermi Problems
The classic Fermi problem, generally attributed to Enrico Fermi, is "How many
piano tuners are there in Chicago?"
4. Pianos that are tuned regularly are tuned on average about once per year.
5. It takes a piano tuner about two hours to tune a piano, including travel time.
6. Each piano tuner works eight hours in a day, five days in a week, and 50 weeks
in a year.
71
From these assumptions we can compute that the number of piano
tunings in a single year in Chicago is: (5,000,000 persons in Chicago) / (2
persons/household) × (1 piano/20 households) × (1 piano tuning per
piano per year) = 125,000 piano tunings per year in Chicago.
And we can similarly calculate that the average piano tuner performs: (50
weeks/year)×(5 days/week)×(8 hours/day)×(1 piano tuning per 2 hours
per piano tuner) = 1,000 piano tunings per year per piano tuner.
Dividing gives: (125,000 piano tuning per year in Chicago) / (1,000 piano
tunings per year per piano tuner) = 125 piano tuners in Chicago.
73
N = 1.5 × 10−5 × 10−9 × 0.2 × 304 = 9.1 × 10−13
If communication capable civilizations are short
lived, a few hundred years, or don’t make an effort
announcing their presence
N = 3 × 1 × 0.2 × 0.13 × 1 × 0.2 × 109 =
15,600,000
If communication capable civilizations live for a
very long time and are actually broadcasting all the
time, wasting an enormous amount of energy
Take your pick ½ Physics Nobel prize 2019 jointly to Michel
Mayor and Didier Queloz "for the discovery
of an exoplanet orbiting a solar-type star”
Such planets are some 50 or 100 light years away from us.
74
An “earth-like” planet around Proxima Centauri in 2016, only about half as hot as sun,
Stefan-Bolzmann law predicts 6.25 % of the total radiant energy, red dwarf, about 1 m
peak of radiation (from Wien’s displacement law), so not a good place for us to live 75on
76
https://physicsworld.com/a/spin-3-2-superconductor-is-a-first-say-physicists/
Spin
3/2
Ideal gas
78
eα problem specific
>>
Typically at high T
e.g. laser
79
Change from Maxwell speed distribution to kinetic energy distribution
where
80
statistical
All condensed matter (liquids and solids) problems are statistical quantum
mechanics problems !!
Ideal gasses can be modeled classically, because they have very low matter
densities
81
After a true story
https://www.youtube.com/watch?v=TaSz18osf5w
82