SomerfieldModel PDF

1
Sommerfield Model for Free Electron Theory
Prof.P. Ravindran,
Department of Physics, Central University of Tamil
Nadu, India
http://folk.uio.no/ravi/CMP2013
P.Ravindran, PHY075- Condensed Matter Physics, Spring 2013 16 July: Sommerfield Model for Free Electron Theory
Quantum free electron theory
deBroglie wave concepts
The universe is made of Radiation(light) and

matter(Particles).The light exhibits the dual nature(i.e.,) it can
behave s both as a wave [interference, diffraction phenomenon]
and as a particle[Compton effect, photo-electric effect etc.,].
Since the nature loves symmetry was suggested by Louis

deBroglie. He also suggests an electron or any other material
particle must exhibit wave like properties in addition to particle
nature
In mechanics, the principle of least action states” that a
moving particle always chooses its path for which the action is a
minimum”. This is very much analogous to Fermat’s principle of
optics, which states that light always chooses a path for which the
time of transit is a minimum.
de Broglie suggested that an electron or any other material

particle must exhibit wave like properties in addition to particle
nature. The waves associated with a moving material particle are
called matter waves, pilot waves or de Broglie waves.
Wave function
A variable quantity which characterizes de-Broglie waves
is known as Wave function and is denoted by the symbol .
The value of the wave function associated with a moving
particle at a point (x, y, z) and at a time ‘t’ gives the probability of
finding the particle at that time and at that point.
de Broglie wavelength
deBroglie formulated an equation relating the momentum
(p) of the electron and the wavelength () associated with it, called
de-Broglie wave equation.
 hp
where h - is the planck’s constant.
Schrödinger Wave Equation
Schrödinger describes the wave nature of a particle in

mathematical form and is known as Schrödinger wave equation.
They are ,
1. Time dependent wave equation and
2. Time independent wave equation.
To obtain these two equations, Schrödinger connected the

expression of deBroglie wavelength into classical wave equation
for a moving particle.
The obtained equations are applicable for both

microscopic and macroscopic particles.
6
Schrödinger Time Independent Wave Equation
The Schrödinger's time independent wave equation is given by
8 2
m
 
2
2
( E  V )  0
h
For one-dimensional motion, the above equation becomes
d  2
8 2 m
2
 2 ( E  V )  0
dx h
7
Introducing,
h
 
2
In the above equation
d 2 2m
2
 2
( E  V )  0
dx 
For three dimension,
2m
   2 ( E  V )  0
2

8
Schrödinger time dependent wave equation
The Schrödinger time dependent wave equation is

2 
     V  i
2
2m t
 2  

  2
 V   i

(or)
 2 m  t
H  E
 2 2
where H =  V = Hamiltonian operator
2m

E = i = Energy operator
t
9
The salient features of quantum free electron theory
Sommerfeld proposed this theory in 1928 retaining the concept of free

electrons moving in a uniform potential within the metal as in the classical
theory, but treated the electrons as obeying the laws of quantum mechanics.
Based on the deBroglie wave concept, he assumed that a moving electron

behaves as if it were a system of waves. (called matter waves-waves associated
with a moving particle).
According to quantum mechanics, the energy of an electron in a metal is

quantized.The electrons are filled in a given energy level according to Pauli’s
exclusion principle. (i.e. No two electrons will have the same set of four
quantum numbers.)
10
Each Energy level can provide only two states namely, one with spin
up and other with spin down and hence only two electrons can be
occupied in a given energy level.
So, it is assumed that the permissible energy levels of a free electron are
determined.
It is assumed that the valance electrons travel in constant potential

inside the metal but they are prevented from escaping the crystal by
very high potential barriers at the ends of the crystal.
In this theory, though the energy levels of the electrons are discrete, the
spacing between consecutive energy levels is very less and thus the
distribution of energy levels seems to be continuous.
11
Success of quantum free electron theory
According to classical theory, which follows Maxwell-

Boltzmann statistics, all the free electrons gain energy. So it
leads to much larger predicted quantities than that is actually
observed. But according to quantum mechanics only one percent
of the free electrons can absorb energy. So the resulting specific
heat and paramagnetic susceptibility values are in much better
agreement with experimental values.
According to quantum free electron theory, both experimental

and theoretical values of Lorentz number are in good agreement
with each other.
12
Drawbacks of quantum free electron theory
It is incapable of explaining why some crystals have metallic

properties and others do not have.
It fails to explain why the atomic arrays in crystals including

metals should prefer certain structures and not others
Summerfeld’s Quantum Mechanical Model of Electron
Conduction in Metals
The Free Electron Gas: A Non-trivial Quantum Fluid
Bohr, de Broglie, Schrödinger, Heisenberg, Pauli, Fermi, Dirac….. The
development of the new theory of quantum mechanics.
A natural step was to formulate a quantum theory of electrons in metals.

First done by Sommerfeld.
Assumptions
Most are very similar to those of Drude. Free and independent electrons, but
no assumptions about the nature of the scattering.
Starting point: time-independent Schrödinger equation
2 2
     (7)
2m
13
Note that no other potential terms are included; hence we can
solve
for a single, independent electron and then investigate the
consequences of putting in many electrons.
To solve (7), we need appropriate boundary conditions for a metal.
Standard ‘particle in a box’: set ψ = 0 at boundaries. This is
not a good representation of a solid, however.
a) It says that the surface is important in determining the

physical properties, which is known not to be the case.
b) It implies that the surfaces of a large but not infinite sample
are perfectly reflecting for electrons, which would make it
impossible to probe the metallic state by, for example, passing a
current through it.
P.Ravindran, PHY075- Condensed Matter Physics, Spring 2013 16 July: Sommerfield Model for Free Electron Theory 14
Most appropriate boundary condition for solid state physics: the periodic
boundary condition first introduced by Born and von Karman:
x  L, y  L, z  L   x, y, z  (8)
(We consider a cube of side L for mathematical convenience; a different

choice of sample shape would have no physical consequence at the end of
the calculation.)
Solving then gives allowed wavefunctions:
2 p
 k  x, y , z  
1 i (kx xk y y kz z )
1/ 2
e , kx  , p integer, etc. (9)
V L
Here V = L3 and the V-1/2 factor ensures that normalisation is correct, i.e. that
the probability of finding the electron somewhere in the cube is 1.
15
What is the physical meaning of these eigenstates?
 2k 2
First, note energy eigenvalues: k  (10)
2m
Then, note that k is also an eigenstate of the momentum operator

 , with eigenvalue p = k.
pˆ  
i
The state k is just the de Broglie formulation of a free particle! It has a
definite momentum k.
Then we see the close analogy with a well-known classical result:
 2k 2 p 2
k   (11)
2m 2m
It thus also has a velocity v = k/m.
How does the spectrum of allowed states look?
Cubic grid of points in k-space, separated by 2/L; volume per point (2/L)3.
So, why have we come anywhere here? We have just done a quantum
calculation of a free particle spectrum, and seen close analogies with that of
classical free particles.
Answer: now we have to consider how to populate these states with a

macroscopic number of electrons, subject to the rules of quantum mechanics.
Sommerfeld’s great contribution: to apply Pauli’s exclusion principle to the

states of this system, not just to an individual atom.
Each k state can hold only two electrons (spin up and down). Make up the
ground (T = 0) state by filling the grid so as to minimise its total energy.
Result: At T = 0, get a sudden demarkation between filled and empty states,

which (for large N), has the geometry of a sphere.
ky
Fermi State volume
wavenumber (2L)3
kz . . . . . .
kF . . . . . .
kx
Filled
states State separation
2L
Empty
states
Fermi surface
We set out to do a quantum Drude model, and did not explicitly include any
direct interactions due to the Coulomb force, but we ended up with something
very different. The Pauli principle plays the role of a quantum mechanical
particle-particle interaction.
The quantum-mechanical ‘free electron gas’ is a non-trivial quantum fluid!
Is everything OK here - doesn’t kF appear to depend on the arbitrary cube

size L?
4 3 N  2 
3 1/ 3
 2 N (12)
No - k F     k F   3 
3 2 L   V
Quantities of interest depend on the carrier number per unit volume; the
sample dimensions drop out neatly.
19
How can we scale these quantum mechanical effects against something we are
more familiar with?
Calculate numerical values for the parameters. Use potassium (tutorial

question 4).
Result: kF  0.75 Å-1
vF  1 x 106 ms-1
F  2 eV
TF  25000 K ( recall kBT at room T  1/40 eV)
This is a huge effect: zero point motion so large that a Drude gas of
electrons would have to be at 25000 K for the electrons to have this
much energy!
20
A couple of much-used graphs relating to the Sommerfeld model:
a) The free electron b) The T = 0 state occupation

dispersion function.

Probability
of state
occupation
1
kF k 0 , k
F or
kF
21
The specific heat of the quantum fermion gas
The T=0 occupation discussed previously is a limit of the Fermi-Dirac

distribution function for fermions:
1
f (, T )  (    ) / k BT
where the chemical potential   F. (13)
e 1
At finite T:
~ 2kBT
f() As expected, T is a minor player
when it comes to changing things.
  F 
22
The Fermi function gives us the probability of a state of energy  being
occupied. To proceed to a calculation of the specific heat, we need to know
the number of states per unit volume of a given energy  that are occupied
per unit energy range at a given T.
n(, T )  g () f (, T ) (14)
Then internal energy Etot(T) can be calculated from

Etot (T )    n( , T )d (15)
0
and the specific heat cel from dEtot/dT as before.
Our next task, then, is to derive a quantity of high and general importance, the
density of states g().
23
ky
State volume
dk (2L)3
kz . . . . . .
. . . . . .
kx
State separation
2L
Number of allowed states per unit volume per shell thickness dk:
2 Vol. of shell at k 2 4k 2 dk

g (k )dk  3  3
L Vol. per k L  2  3
 
spin  L
24
Convert to density of states per unit volume per unit  (the quantity usually
meant by the loose term ‘density of states’):
md  2m 
1/ 2
dk  2 ; k   2  (16a, b)
k   
1/ 2
2m m    2
4 2 2  
   2m 
g ()d  2 d
(2)3
 g () 
2m  
3 / 2 1/ 2
(17)
2 2  3
Very important result, but note that  dependence is different for

different dimension .
25
Evaluating integral (15) is complicated due to the slight movement of the
chemical potential  with T (see Hook and Hall and for details Ashcroft and
Mermin). However, we can ignore the subtleties and give an approximate
treatment for F >> kBT:
g(
n(,T)
Movement of
electrons in
F energy at finite T

2kBT
[Etot(T) - Etot(0)]/V  1/2g(F). kBT.2kBT = g(F). (kBT)2 (18)
26
Differentiating with respect to T gives our estimate of the specific heat
capacity:
cel = 2g(F). kB2T (19)
The exact calculation gives the important general result that
cel = (23g(F). kB2T (20)
How does this compare with the classical prediction of the Drude model?
Combining g(F) from (17) with the expression for F derived in tutorial
question 4 gives, after a little rearrangement :
2  k BT  3
cel  nkB   (21) c.f. Drude: nkB
2  F  2
27
A remarkable result: Even though our quantum mechanical interaction leads to
highly energetic states at F, it also gives a system that is easy to heat, because you
can only excite a highly restricted number of states by applying energy kBT.
The quantum fermion gas is in some senses like a rigid fluid, and its thermal
properties are defined by the behaviour of its excitations.
28
What about the response to external fields or temperature gradients?
To treat these simply, should introduce another vital and wide-ranging concept,
the Semi-Classical Effective Model.
Faced with wave-particle duality and a natural tendency to be more comfortable

thinking of particles, physicists often adopt effective models in which quantum
behaviour is conceptualised in terms of ‘classical’ particles obeying rules
modified by the true quantum situation.
In this case, the procedure is to think in terms of wave packets centred on each
k state as particles. Each particle is classified by a k label and a velocity v.
Velocity is given by the group velocity of the wave packet:

v = dw/dk = -1d/dk = k/m for free particles like those we are concerned
with at present.
29
Assumption of the above: we cannot localise our ‘particles’ to better
than about 10 lattice spacings. The uncertainty principle tells us that if we try
to do that, we would have to use states more than 10% of our full available
range (defined roughly by kF).
Not, however, a particularly heavy restriction, since it is unlikely that we would

want to apply external fields which vary on such a short length scale.
In the absence of scattering, we then use the following ‘classical’ equation of

motion in applied E and/or B fields:
mdv/dt = dk/dt= -eE - ev  B (22)
This equation would produce continuous acceleration, which we know cannot

occur in the presence of scattering.
30
Include scattering by modifying (22) to
m(dv/dt + vt  -eE - ev  B (23)
This is just the equation of motion for classical particles subject to ‘damped
acceleration’. If the fields are turned off, the velocity that they have acquired
will decay away exponentially to zero. This reveals their ‘conjuring trick’.
The physical meaning of v in (23) must therefore be the ‘extra’ or ‘drift’
velocity that the particles acquire due to the external fields, not the group
velocity that they introduced in their (3.22).
In fact, this is formally identical to the process that we discussed in deriving

equation when we discussed the Drude model!
It is no surprise, then, that it leads to the same expression for the electrical
conductivity:
Set B to zero and stress that the relevant velocity is vdrift ; (23) becomes
m(dvdrift/dt + vdriftt  -eE
Steady state solution (dvdrift/dt = 0) is just
vdrift = -(et/m)E
Following the procedure from Kittel gives us the Drude expression (3):
ne 2t

m
If you give this some thought, it should concern you. What happened to our
new quantum picture?
32
To understand, consider physical meaning of the process:
ky ky
kz kz
kx kx
E=0 dk = -1mvdrift= -eEτ/
Fermi surface is shifted along the kx axis by an E field along x. The ‘quasi-
Drude’ derivation assumes that every electron state in the sphere is shifted by
dk. This is ‘mathematically correct’, but physically entirely the wrong picture.
33
ky Which states can ‘interact with the
outside world’?
kz
In the quantum model, only those
within kBT of F, i.e. those very near the
Fermi surface.
kx
Pauli principle: only those states can scatter, so
only processes involving them can relax the
Fermi surface. So how does the ‘wrong’
picture work out?
dk
Consider amount of extra velocity/momentum acquired in equilibrium:
3 3
Drude-
 2  4 3 Quantum 2
  4k 2
F dk 2
like   k F .dk picture:   . kF (24)
picture:  L  3  L  2 3
# of states mom. # of states (1/2 mom. x comp.
gain FS area) gain only
34
So the two pictures, one of which is conceptually incorrect, give the
same answer, because of a cancellation between a large number of
particles acquiring a small extra velocity and a small number of
particles acquiring a large extra velocity.
However, this is only the case for a sphere. As we shall see later,
Fermi surfaces in solids are not always spherical. In this case, the
Drude-like picture is simply wrong, and the conductivity must be
calculated using a Fermi surface integral.
P.Ravindran, PHY075- Condensed Matter Physics, Spring 2013 16 July: Sommerfield Model for Free Electron Theory 35
What about thermal conductivity?
Recall (4) from Drude model: k = 1/3vrandomlcel
Here, vrandom can clearly be identified with vF, and l = vFt.
Provided that t is the same for both electrical and thermal conduction
(basically true at low temperatures but not at high temperatures; see Hook
and Hall Ch. 3 after we have covered phonons), we can now revisit the
Wiedemann-Franz law using (21) for the specific heat:
k 1 m 1 2   k BT  
2
 kB 
2 2
 vF t nkB      (25)
T T ne t 3
2
2  F  3  e 
The approximate factor of two error from the Drude model has been
corrected (2/3 in quantum model cf. 3/2 in Drude model).
Real question - how on earth was the Drude model so close?
Answer: Because a severe overestimate of the electronic specific heat was

cancelled by a severe underestimate of the characteristic random velocity.
Thinking for the more committed (i.e. non-examinable): Would all quantum
gas models give the same result for the Wiedemann-Franz law as the
quantum fermion gas?
The modern conceptualisation of the quantum free electron gas:
Make an analogy with quantum electrodynamics (QED).
Filled Fermi sea at T = 0 is inert, so it is the vacuum. Temperature and / or

external fields excite special particle-antiparticle pairs. The role of the positron is
played by the holes (vacancies in the filled sea with an effective positive charge).
Thermal excitation: All particles Electrical excitation: All particles

with k  kF, but sum over k = 0. with k  kF, but sum over k = 2 kF/3.
ky ky
kz kz
kx kx
dk
P.Ravindran, PHY075- Condensed Matter Physics, Spring 2013 16 July: Sommerfield Model for Free Electron Theory dk 38
Scorecard so far; achievements and failures of the quantum Fermi gas
model
1. Successful prediction of basic thermal properties of metals.

2. Successful prediction of conductivity, as long as we don’t ask about the
microscopic origins of the scattering time t - why is the mean free path so
long in metals at low temperatures? What happened to electron-ion and
electron-electron scattering?
3. Failure to predict a positive Hall coefficient.

4. No understanding whatever of insulators. ‘… So insulators, which cannot
carry a current, must contain electrons too. In a metal they must be free to move,
and in an insulator they must be stuck.
Classical Drude gas Quantum Sommerfeld gas: do wave
mechanics and then think in an
‘equivalent particle’ picture
Random velocity purely Random velocity dominantly

thermal: quantum (due to Pauli principle):
3k BT / m
1/ 3
 2 N
vF  k F / m   3  /m
 V
3 2  k BT 
Specific heat cel = nkB cel  nkB  
 F 
2 2
Large number of particles Small effective number of particles

moving slowly. moving very fast, due to special
quantum mechanical constraints.
40
41
The Sommerfeld Model
Electrons are fermions.

- Ground state: Fermi sphere, k F  (3 2 n)1 / 3
- Distribution function
m mv 2
f MB (v)  n( ) 3 / 2 exp(  )
2 k B T 2k B T
(m / ) 3 1
f FD (v) 
4 3
1
exp[( mv 2  E F ) / k B T ]  1
2
Modification of the Drude model

3k B T 1 / 2 3  2 k BT
v : vT  ( )  v F , cv  n k B  ( )nk B
m 2 2 EF
- the mean free path l  vT t  v Ft

k 3 k  2 kB 2
 ( B )2 
- the Wiedemann-Frantz law T 2 e
( )
3 e
- the thermopower kB  2 k B k BT
Q  ( )( )
2e 6 e EF
43
The Sommerfeld theory of metals
3/ 2
 m   mv 2 
the Drude model: electronic velocity distribution f MB ( v )  n   exp   
is given by the classical  2 k B 
T  2 k B 
T
Maxwell-Boltzmann distribution
m / 
3
1
the Sommerfeld model: electronic velocity distributionf ( v) 
4 3 1 2 
FD
is given by the quantum mv  k T
Fermi-Dirac distribution  B 0 
exp  2  1
 k BT 
 
Pauli exclusion principle: at most one electron
can occupy any single electron level
normalization
n   dvf ( v ) condition T0
consider noninteracting electrons 44
electron wave function

associated with a level of energy E 2   2 2 2 
satisfies the Schrodinger equation   2  2  2  ( r )  E ( r )
2m  x y z 
  x, y , z  L     x, y , z 
periodic
boundary
conditions
  x, y  L, z     x, y, z 
  x  L, y, z     x, y, z  3D:
a solution neglecting 1 ikr

the boundary conditions  k (r )  e 1D: L
V
normalization constant: probability 2
of finding the electron somewhere
in the whole volume V is unity 1   dr (r )
 2k 2
energy E (k ) 
2m
momentum
p  k
velocity
k p2 1 2
wave vector v E  mv
de Broglie m 2m 2
wavelength
k
2

k
45
  x, y , z  L     x, y , z 
1 ikr   x, y  L, z     x, y, z 
 2 
2
 k (r )  e the area
V   x  L, y, z     x, y, z  per point  
 L 
the volume  2 3 2 3
 e  e 1
ik x L ik y L ik z L
apply the boundary conditions
e per point   
 L  V
2 2 2
kx  nx , k y  ny , kz  nz
components of k must be L L L
nx, ny, nz integers
a region of k-space of  V states


volume  contains 2 / L 3 2 3 i.e. allowed
V values of k
the number of states
per unit volume of k-space, 2 3
k-space density of states
k-space
consider T=0 46
the Pauli exclusion principle postulates that only one electron can occupy a single state
therefore, as electrons are added to a system, they will fill the states in a system
like water fills a bucket – first the lower energy states and then the higher energy states
the ground state of the N-electron system is formed by occupying all single-particle levels with k < kF
volume density
state of the lowest energy of states
the number of allowed values of  4 k F 3  V kF 3
 3   2V
  2  6
k within the sphere of radius kF 3

kF 3 ky
to accommodate N electrons
N 2 2V
2 electrons per k-level due to spin 6 Fermi sphere kF
kF 3
n 2
3 Fermi surface
kx
Fermi wave vector k F ~108 cm-1 at energy EF
E F   2 k F / 2m
k F  3 2n 
2
Fermi energy ~1-10 eV 13
Fermi temperature TF  EF k B ~104-105 K

Fermi momentum pF  k F EF 
2
2m
3 2n 
23
vF  k F / m
vF  3 2n 
Fermi velocity ~108 cm/s  13
vthermal  3k BT / m  ~ 107 cm/s at T=300K

1/ 2
compare to the m
classical
thermal velocity
P.Ravindran, PHY075- Condensed Matter Physics, Spring 2013 16 July: 0 at T=0
Sommerfield Model for Free Electron Theory
47
Density of states
V 3
Total number of states with wave vector < k N k k 2 2
3 2 E
32 2m
V  2mE 
Total number of states with energy < E N 2 2 
3   
32
The density of states – number of states per unit energyD ( E )  dN  V  2m  E
dE 2 2   2 
32
dn 1  2m 
The density of states per unit volume or the density of D( E )   2 2  E
states dE 2   
V
k-space density of states – the number of states per unit volume of k-space
2 3
48
Ground state energy of N electrons
2 2
E  2 k
Add up the energies of all electron k  k F 2m
states inside the Fermi sphere
k  8 3 V
volume of k-space per state V V
 F (k )  3 
8 k
F ( k )k k 0i . e.V 
    
8 3 
F ( k )dk
 smooth F(k) k
The energy density E 1  2k 2 1  2k F

5
 3  dk  2
The energy per electron V 4 k k F 2m  10m
in the ground state  2k 2
F (k ) 
2 2m
E 3  2k F 3
  EF dk  4k 2 dk
N 10 m 5
kF 3
N  2V
3
Remarks on statistics I 49
In quantum mechanics particles are indistinguishable

systems where particles are exchanged are identical
exchange of identical particles can lead to changing
of the system wave function by a phase factor only   ,
1 2   eia
 2 , 1  system of N=2 particles
1, 2 - coordinates and
 1, 2    2 , 1 
spins for each of the
repeated particle exchange → e2ia  1 particles
Antisymmetric wavefunction with respect symmetric wavefunction with respect

to the exchange of particles to the exchange of particles
 1, 2  
1

 p1 1  p2 2   p1 2  p2 1    1, 2  
1

 p1 1  p2 2   p1 2  p2 1 
2 2
p1, p2 – single particle states
Fermions are particles which have half-integer spin Bosons are particles which have integer spin
the wavefunction which describes a collection the wavefunction which describes a collection
of Fermions must be antisymmetric with respect of bosons must be symmetric with respect
to the exchange of identical particles to the exchange of identical particles
Fermions: electron, proton, neutron Bosons: photon, Cooper pair, H atom, exciton
if p1 = p2   0 Unlimited number of bosons can occupy

→ at most one fermion can occupy a single particle state
any single particle state – Pauli principle
obey Fermi-Dirac statistics Obey Bose-Einstein statistics

50
Distribution function f(E) → probability that a state at energy E
will be occupied at thermal equilibrium
fermions Fermi-Dirac 1 degenerate

f FD ( E ) 
particles with distribution E Fermi gas
half-integer spins function exp   1 fFD(k) < 1
 k BT 
bosons Bose-Einstein 1 degenerate
f BE ( E ) 
particles with distribution E Bose gas
integer spins function exp   1 fBE(k) can be any
 B 
k T
both fermions and Maxwell-Boltzmann E classical
bosons at high T distribution f MB ( E )  exp   gas
when E    k BT function  k BT  fMB(k) << 1
n   dEn( E )   dED( E ) f ( E )
=(n,T) – chemical potential
remarks on statistics II m v 51
x
BE and FD distributions differ from the classical MB distribution
because the particles they describe are indistinguishable. (x)
vg=v
Particles are considered to be indistinguishable if their wave packets
x
overlap significantly.
Two particles can be considered to be distinguishable
x
if their separation is large compared to their de Broglie wavelength.
g(k’) k
12 k’
Thermal de Broglie  2  h2 k0
  k '2  
wavelength dB    ~  (r, t )   g (k ') exp i  k ' r  t 
 
 mk BT  
p k' 2 m
A particle is represented by a
Particles become
dB ~ d  n 1 3
wave group or wave packets
indistinguishable when of limited spatial extent,
2 2 2 3 which is a superposition of many matter
i.e. at temperatures below TdB  n waves with a spread of wavelengths
mk B
centered on 0=h/p
At T < TdB fBE and fFD are strongly different from fMB The wave group moves
At T >> TdB fBE ≈ fFD ≈ fMB with a speed vg – the group speed,
Electron gas in metals: which is identical to the classical
particle speed
n = 1022 cm-3, m = me → TdB ~ 3×104 K
Heisenberg uncertainty principle, 1927:
Gas of Rb atoms: If a measurement of position is made with
precision x and a simultaneous
n = 1015 cm-3, matom = 105me → TdB ~ 5×10-6 K measurement of momentum in the x
direction is made with precision px,
Excitons in GaAs QW then
n = 1010 cm-2, mexciton= 0.2 me → TdB ~ 1 K p x 
x Theory 2
P.Ravindran, PHY075- Condensed Matter Physics, Spring 2013 16 July: Sommerfield Model for Free Electron
T≠0 the Fermi-Dirac distribution 52
3D
32
dn 1  2m 
Density of states D( E )   2 2  E
dE 2  
1
Distribution function f ( E )  lim f ( E )  1, E  
E T 0
exp   1 0 E
 B 
k T
n   dED ( E ) f ( E )   lim   EF
T 0
1
D( E )dE   [the number of states in the energy range from E to E + dE]
V
1
D( E ) f ( E )dE   [the number of filled states in the energy range from E to E + dE]
V
Density of
states
D(E)
per unit volume

Density of
filled states
D(E)f(E,T)
shaded area – filled
states at T=0
EF E
53
Specific heat of the degenerate electron gas, estimate
T ~ 300 K for typical metallic densities

1  U   u  T=0
Specific heat cv 
    
V  T V  T V
U – thermal U
u
kinetic energy V f(E) at T ≠ 0 differs from f(E) at T=0
1 3 only in a region of order kBT about 
Classical gas u  n mv 2  nkBT because electrons just below
2 2 EF have been excited to levels just above EF
3
cv  nkB The observed electronic
2
contribution at room T is
usually 0.01 of this value
Classical gas: with increasing T all electron gain an energy ~ kBT

Fermi gas: with increasing T only those electrons in states within
an energy range kBT of the Fermi level gain an energy ~ kBT
k T
Number of electrons which gain energy with increasing temperature ~ N B
EF
 k BT 
The total electronic thermal kinetic energy U ~  N  k BT
 E F 
EF/kB ~ 104 – 105 K
1  U  k BT
The electronic specific heat cv    ~ nk kBTroom / EF ~ 0.01
V  T V Model for Free
B
P.Ravindran, PHY075- Condensed Matter Physics, Spring 2013 16 July: Sommerfield
EF Electron Theory
Specific heat of the degenerate electron gas 54

u   3 E (k ) f E (k )    dED( E ) Ef ( E )
dk
4 0  u 
  and u  c   
 T V
v
n   3 f E ( k )    dED( E ) f ( E )
dk
4 0

The way in which integrals of the form  H ( E ) f ( E )dE differ from their zero T values H ( E )dE
EF
 
is determined by the form of H(E) near E= 
( E   )n d n
Replace H(E) by its Taylor expansion about E= H ( E )   n
H ( E ) E 
n 0 n! dE
  
d 2 n 1
The Sommerfeld expansion  H ( E ) f ( E )dE   H ( E )dE   k T 
2n
B an 2 n 1
H ( E ) E 
  n 1 dE
 6
2 7 4
  H (E )E  k BT  H (  )  kBT 4 H (  )  O kBT 
2

6 360   

Successive terms are 2
smaller by O(kBT/)2 u   ED ( E )dE  k BT 2 D(  )  D(  )  O (T 4 )
0
6

2
For kBT/ << 1 n   D ( E )dE  k BT 2 D(  )  O (T 4 ) Replace 
6
0
by T0 = EF
 EF
 H ( E )dE   H ( E )dE  (   E
0 0
F ) H ( EF )
correctly to order T2
55
Specific heat of the degenerate electron gas
 1   k T 2 
  EF 1   B  
 3  2 E F  
2
u  u0   k BT 
2
D( EF )
6
u  2 2
cv   k B TD ( E F )
1  2m 
32 T 3
D( E )  2  2  E
2   3 n
2
D( EF ) 
 3 n  2 EF
23
EF  2
2m
 2 k BT
cv  nk B
2 EF
3
(1) cclassical  nk B
2
FD statistics depress  k BT
2
cv by a factor of 3 EF
(2) cv  T
56
Thermal conductivity
thermal current density jq – a vector parallel to the direction of heat flow
whose magnitude gives the thermal energy per unit time
jq  kT crossing a unite area perpendicular to the flow
1 1
k  v 2tcv  lvcv
3 3
k cv mv 2 3  k B 
2
ne2t
    T Wiedemann-Franz law (1853)
m  3ne2 2  e  Lorenz number ~ 2×10-8 watt-ohm/K2
Drude: 3
cv  nkB success of the Drude model is due to the cancellation
application of 2 of two errors: at room T the actual electronic cv is 100
classical ideal 1 2 3 times smaller than the classical prediction, but v is 100
gas laws mv  k BT times larger
2 2
For  2 k BT
the correct cv  nkB cv cv classical ~ k BT / EF ~ 0.01 at room T
degenerate 2 EF
Fermi gas of 2 2
the correct estimate of v2 is vF2 vF vclassical ~ EF / k BT ~ 100 at room T
electrons
k  2  kB 
2
  
T 3  e 
57
Thermopower
Seebeck effect: a T gradient in a long, thin bar should be accompanied by an electric field
directed opposite to the T gradient
high T low T E  QT thermoelectric field
cv
gradT Thermopower Q
3ne
E
Drude:
application of 3 kB
cv  nkB Q
classical ideal 2 2e
gas laws
For  2 k BT
degenerate the correct cv  nkB cv cv classical ~ k BT / EF ~ 0.01 at room T
2 EF
Fermi gas of
electrons Q/Qclassical ~ 0.01 at room T
 2 k B  k BT 
Q  
6 e  EF 
58
Electrical conductivity and Ohm’s law
Equation of motion dv dk
Newton’s law m    eE dp ( t ) p( t )
dt dt   f (t )  0
In the absence of collisions the Fermi sphere in dt t
k-space is displaced as a whole at a uniform rate k (t )  k (0)   eE t p  ft   eEt
by a constant applied electric field
Because of collisions the displaced Fermi sphere eE
is maintained in a steady state in an electric field k avg   t
ky Ohm’s law
F k eE
v avg  avg
 t  ne2t 
Fermi sphere m m j   E
 m 
j  nev avg
kx ne2t

kavg m
1 m
  2
the mean free path l = vFt  ne t
because all collisions involve only electrons near the Fermi surface
vF ~ 108 cm/s for pure Cu:
at T=300 K t ~ 10-14 s l ~ 10-6 cm = 100 Å
at T=4 K t ~ 10-9 s l ~ 0.1 cm
kavg << kF for n = 1022 cm-3 and j = 1 A/mm2 vavg = j/ne ~ 0.1 cm/s << vF ~ 108 cm/s

SomerfieldModel PDF

Uploaded by

Copyright:

Available Formats

SomerfieldModel PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

SomerfieldModel PDF

Uploaded by

Copyright:

Available Formats

1

Sommerfield Model for Free Electron Theory

deBroglie wave concepts

The universe is made of Radiation(light) and

Since the nature loves symmetry was suggested by Louis

de Broglie suggested that an electron or any other material

Schrödinger describes the wave nature of a particle in

To obtain these two equations, Schrödinger connected the

The obtained equations are applicable for both

Schrödinger Time Independent Wave Equation

The Schrödinger's time independent wave equation is given by

Schrödinger time dependent wave equation

The Schrödinger time dependent wave equation is

The salient features of quantum free electron theory

Sommerfeld proposed this theory in 1928 retaining the concept of free

Based on the deBroglie wave concept, he assumed that a moving electron

According to quantum mechanics, the energy of an electron in a metal is

It is assumed that the valance electrons travel in constant potential

Success of quantum free electron theory

According to classical theory, which follows Maxwell-

According to quantum free electron theory, both experimental

Drawbacks of quantum free electron theory

It is incapable of explaining why some crystals have metallic

It fails to explain why the atomic arrays in crystals including

A natural step was to formulate a quantum theory of electrons in metals.

Starting point: time-independent Schrödinger equation

a) It says that the surface is important in determining the

x  L, y  L, z  L   x, y, z  (8)

(We consider a cube of side L for mathematical convenience; a different

Solving then gives allowed wavefunctions:

Then, note that k is also an eigenstate of the momentum operator

Then we see the close analogy with a well-known classical result:

How does the spectrum of allowed states look?

Answer: now we have to consider how to populate these states with a

Sommerfeld’s great contribution: to apply Pauli’s exclusion principle to the

Result: At T = 0, get a sudden demarkation between filled and empty states,

The quantum-mechanical ‘free electron gas’ is a non-trivial quantum fluid!

Is everything OK here - doesn’t kF appear to depend on the arbitrary cube

Calculate numerical values for the parameters. Use potassium (tutorial

Result: kF  0.75 Å-1

TF  25000 K ( recall kBT at room T  1/40 eV)

a) The free electron b) The T = 0 state occupation

The T=0 occupation discussed previously is a limit of the Fermi-Dirac

n(, T )  g () f (, T ) (14)

Then internal energy Etot(T) can be calculated from

and the specific heat cel from dEtot/dT as before.

2 Vol. of shell at k 2 4k 2 dk

Very important result, but note that  dependence is different for

cel = 2g(F). kB2T (19)

The exact calculation gives the important general result that

cel = (23g(F). kB2T (20)

Faced with wave-particle duality and a natural tendency to be more comfortable

Velocity is given by the group velocity of the wave packet:

Not, however, a particularly heavy restriction, since it is unlikely that we would

In the absence of scattering, we then use the following ‘classical’ equation of

mdv/dt = dk/dt= -eE - ev  B (22)

This equation would produce continuous acceleration, which we know cannot

m(dv/dt + vt  -eE - ev  B (23)

In fact, this is formally identical to the process that we discussed in deriving