TP3-Notes Lenz

Theoretical Physics 3
PHYS3661
Quantum Mechanics
Epiphany 2015
Alexander Lenz
IPPP, Durham University
April 8, 2015
1
2
Contents
1 Scattering experiments and cross sections 8
1.1 Introduction and Motivation . . . . . . . . . . . . . . . . . . . 8
1.2 Basics in scattering . . . . . . . . . . . . . . . . . . . . . . . . 10
2 Potential scattering (general features) 15

2.1 The time-independent Schrodinger equation . . . . . . . . . . 15
2.2 The scattering amplitude . . . . . . . . . . . . . . . . . . . . . 16
3 Spherical Bessel functions 18

3.1 Separation of variables . . . . . . . . . . . . . . . . . . . . . . 18
3.2 The simplest Bessel function . . . . . . . . . . . . . . . . . . . 19
3.3 The spherical Bessel equation . . . . . . . . . . . . . . . . . . 20
3.3.1 Definition . . . . . . . . . . . . . . . . . . . . . . . . . 20
3.3.2 Deriving the solution: . . . . . . . . . . . . . . . . . . . 20
3.3.3 Properties of the solutions . . . . . . . . . . . . . . . . 22
3.3.4 Bessel functions J . . . . . . . . . . . . . . . . . . . . 23
(1,2)
3.3.5 Spherical Hankel functions h . . . . . . . . . . . . . 24
3.3.6 Orthogonality relations . . . . . . . . . . . . . . . . . . 24
3.4 Summary of Bessel functions . . . . . . . . . . . . . . . . . . . 24
3.5 Plane waves . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
4 The method of partial waves 25

4.1 Definition of partial waves . . . . . . . . . . . . . . . . . . . . 25
4.2 Structure of the solution - Phase shifts . . . . . . . . . . . . . 26
4.3 Convergence of the partial wave expansion . . . . . . . . . . . 28
4.4 Scattering by a square well . . . . . . . . . . . . . . . . . . . . 28
4.5 Resonances . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
4.6 Hard-sphere potential . . . . . . . . . . . . . . . . . . . . . . . 30
5 The integral equation of potential scattering 31

5.1 Definition of the Greens function . . . . . . . . . . . . . . . . 31
5.2 Deriving the Greens function . . . . . . . . . . . . . . . . . . 31
5.3 Lippmann-Schwinger equation . . . . . . . . . . . . . . . . . . 32
6 The Born approximation 34

6.1 The Born series . . . . . . . . . . . . . . . . . . . . . . . . . . 34
6.2 The first Born approximation . . . . . . . . . . . . . . . . . . 34
6.3 The Coulomb potential . . . . . . . . . . . . . . . . . . . . . . 35
6.4 Propagator . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
3
7 Advanced Topics in Scattering 37
7.1 Form factor . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
7.2 Electron scattering on nucleon . . . . . . . . . . . . . . . . . . 37
7.3 Collisions between identical particles . . . . . . . . . . . . . . 39
7.3.1 Two spin-less bosons . . . . . . . . . . . . . . . . . . . 40
7.3.2 Two spin 1/2 fermions . . . . . . . . . . . . . . . . . . 40
7.4 Introduction to multichannel scattering . . . . . . . . . . . . . 40
7.4.1 Elastic Scattering . . . . . . . . . . . . . . . . . . . . . 40
7.4.2 Inelastic Scattering . . . . . . . . . . . . . . . . . . . . 40
8 The density matrix 41

8.1 Definition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
8.2 Polarisation . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
8.3 Von-Neumann equation . . . . . . . . . . . . . . . . . . . . . . 45
9 Quantum mechanical ensembles 47

9.1 Micro-canonical ensemble . . . . . . . . . . . . . . . . . . . . . 47
9.2 Canonical ensemble . . . . . . . . . . . . . . . . . . . . . . . . 49
9.3 Grand canonical Ensemble . . . . . . . . . . . . . . . . . . . . 52
9.4 Spin 1/2 particle in a magnetic field . . . . . . . . . . . . . . . 53
9.5 Average of a particle in a box . . . . . . . . . . . . . . . . . . 53
10 Systems of non-interacting particles 54
11 The Klein-Gordon equation 58

11.1 Derivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
11.2 Charged particle in an electromagnetic field . . . . . . . . . . 59
11.3 Stationary solutions . . . . . . . . . . . . . . . . . . . . . . . . 60
11.4 Interpretation . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
12 The Dirac equation 63

12.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
12.2 Derivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
12.3 Continuity equation . . . . . . . . . . . . . . . . . . . . . . . . 65
12.4 Co-Variant formulation of Dirac theory . . . . . . . . . . . . . 66
13 Lorentz-invariance of the Dirac equation 69

13.1 Lorentz transformation . . . . . . . . . . . . . . . . . . . . . . 69
13.2 Lorentz invariance of the Dirac equation . . . . . . . . . . . . 71
13.3 Bilinear forms . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
4
14 Solutions of the Dirac equation 74
14.1 Plane wave solutions of the Dirac equation . . . . . . . . . . . 74
14.2 Non-relativistic limit of the Dirac-equation . . . . . . . . . . . 75
15 Solutions of the Dirac equation for a central potential 79
16 Measurements and interpretation 82

16.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
16.2 The basic question . . . . . . . . . . . . . . . . . . . . . . . . 82
16.3 Einstein-Podolsky-Rosen paradox . . . . . . . . . . . . . . . . 84
16.4 Bells equation . . . . . . . . . . . . . . . . . . . . . . . . . . 86
16.5 More fancy interpretations . . . . . . . . . . . . . . . . . . . . 88
17 Revision Lecture 89
17.1 Scattering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
17.2 Quantum statistical mechanics . . . . . . . . . . . . . . . . . . 89
17.3 Relativistic Quantum mechanics . . . . . . . . . . . . . . . . . 89
17.4 Interpretation of Quantum mechanics . . . . . . . . . . . . . . 89
5
6
Some Comments:
18 lectures (2 hours per week) + 3 example classes: (1 hour every 3

weeks)
1 revision lecture in Easter
http://www.dur.ac.uk/physics/students/problems/
38 hours of lecture + 156 hours of preparation and reading, i.e. 4.1
hours of preparation and reading per 1 hour of lecture
Prerequisites:
Foundations of Physics 2A
Theoretical Physics 2
Mathematical Methods in Physics or Analysis in many variables
Foundations of Physics 3A
Theoretical Physics 3 - Electrodynamics
The lecture course follows the textbook:

Quantum Mechanics; B.H. Bransden and C.J. Joachim; Prentice Hall,
2nd edition [1]
Further recommended books for this course are Sakurai [2], Weinberg
[3] (relatively advanced, from one of the masters of Quantum Field
Theory), the two books from Schwabl [4, 5] (my personal favourites)
and the classical book from Bjorken and Drell [6].
7
1 Scattering experiments and cross sections
Chapter [13.1] of [1]
1.1 Introduction and Motivation

The Nobel prize for physics in 2013 [8] was given to Francois Englert and
Peter Higgs for the prediction (1964) of the Higgs particle [9, 10, 11, 12, 13, 14]
which was discovered at the Large Hadron Collider (LHC) in 2012 [15, 16].
For more information about the discovery see www.nobelprize.org
Press release
Popular information
Advanced information
The LHC - is in principle a huge microscope, which tells us about the fun-
damental constituents of matter.
Atoms are built of electrons and the nuclei. The electrons are elementary ac-
cording to the current experimental precision. The nuclei consists of nucleons,
i.e. protons and neutrons. Protons and neutrons are to a first approximation
built out of three quarks, which are also elementary according to the current
8
experimental precision.
Currently we know the following fundamental particles and forces:
Matter constituents
Quarks Leptons
u c t q = +2/3 q = 0
e
d s b q = 1/3 e q = 1
Quarks and leptons have spin 1/2.

Quarks take part in all interactions, leptons do not take part in
the strong interaction and neutrinos only interact via the weak
interaction.
A proton is made out of 3 quarks: p = |uudi
Fundamental forces are carried by spin 1 (2) gauge bosons
Electromagnetic interaction: photon.

Strong interaction: 8 gluons.
Weak interaction: W , Z-bosons.
Gravity: Graviton? with spin 2.
Creation of mass
Masses of the elementary particles are in contradiction to some math-
ematical properties of the theory. The Higgs mechanism is a trick to
solve this problem, for the price of introducing a new particle: the
Higgs boson with spin 0.
This is the so-called standard model of particle physics, which has been
experimentally verified by hundreds of experiments - sometimes with an in-
credible precision.
Most of this knowledge comes from scattering-experiments, like the Ruther-
ford experiment (Geiger and Marsden 1909, Rutherford 1911) or the LHC.
The corresponding scattering events take place on a microscopic level. Thus
an understanding of the quantum mechanical foundations of scattering is the
basis for a fundamental understanding of the building blocks of the universe.
9
1.2 Basics in scattering
The basic idea of scattering is simple. We have a target and some incident
particles, that are flying towards and interacting with the target. In the case
of Rutherford scattering the target was a gold foil and the incident particles
were alpha particles (He++ ), in the case of LHC both target and incident
particles are protons.
10
The incident particle is described by its energy E and the impact parameter
b. If the incident particles and the target stay intact during the scattering
process, we speak about elastic scattering. In that case the position of the
incoming particles after scattering is given by the scattering angle (, ) and
a radial variable r.
For simplicity we assume besides elastic scattering that the target is infinitely
heavy and that there is an azimuthal symmetry of the interaction.
Now we denote by a small area of the incoming particle flux. All particle
within will be scattered into the solid angle element . If we normalise
the solid angle element to an unit element d, then the corresponding area of
the incoming particles reads d. Having an azimuthal symmetry, the problem
looks like:
11
Integrating over the full solid angle d, will then give the effective area of
the scattering target, called total cross section . Defining in addition the
differential cross-section d/d we get
Z2 Z
d d
Z
tot = d = d d sin . (1)
d d
0 0
For classical scattering of a point-like particle on a hard sphere with radius

R one gets e.g.
d R2
= , (2)
d 4
= R2 , (3)
which coincides with our expectations above. The concept of (differential)

cross sections can, however, also be applied to soft targets.
A beam of incoming particles can be described by the luminosity L, which
is defined as the number of incoming particles per unit area per time. The
number of particles coming through the element d per time is now called
dN and we have
d
dN = Ld = L d . (4)
d
Thus the differential cross section can also be defined as
d 1 dN
:= , (5)
d L d
12
which is many times used as the initial definition of the differential cross
section. An elastic scattering event at LHC looks like
p+pp+p. (6)
The total cross-section for this scattering at LHC with a proton energy of 7
TeV reads [17]
= (98.3 0.2(stat) 2.8(syst)) mb, (7)
with the unit barn for measuring the cross section:
1barn = 1b = 1028 m2 = 100f m2 . (8)
The above concepts can also be generalised to inelastic processes, like the
production of a Higgs particle
p+pX +H . (9)
The cross section for this process is given by [19] (see e.g.[18] for a part of
the state-of-the-art calculation)
20pb . (10)
The number of events for an inelastic process can be determined from the
integrated luminosity
N = L , (11)
Z
N = Ldt . (12)
13
The LHC achieved until now an integrated luminosity of 20(f b)1 per exper-
iment (ATLAS and CMS), thus the following number of Higgs particles was
produced at each experiment
pb
N = 20f b1 20pb = 400 = 4 105 . (13)
103pb
To obtain the number of a certain Higgs decay like H one has to
multiply the overall number of produced Higgses with the corresponding
branching ratio, in that case
N(H ) = 4 105 2 103 = 800 . (14)
The value of the differential cross section for two colliding particles A and B
depends on the reference frame. Two commonly used ones are the laboratory
frame (L), where particle B is in the beginning at rest and centre of mass
frame (CM). One finds
sin
tan L = , (15)
cos +
3
(1 + 2 + 2 cos ) 2 d

d
= , (16)
d L |1 + cos | d CM
with = mA /mB .
To repeat this lecture: Chapter [13.1] of [1]

- the concept of cross section -
To prepare for next lecture: Chapter [13.2] of [1]

- how to calculate a cross section within quantum mechanics -
14
2 Potential scattering (general features)
2.1 The time-independent Schrodinger equation

Now we discuss a non-relativistic particle of mass m in a fixed potential,
which is described by the 3-dimensional Schrodinger equation
~2 ~ 2

i~ (~r, t) = + V (~r) (~r, t) , (17)
t 2m
where ~r describes the distance of the particle from the origin of the potential.
The scattering of particle A with the mass mA with a particle B with the mass
mB due to a force that is described by the potential is given by
2
~ ~2
i~ (~r, t) = + V (~r) (~r, t) , (18)
t 2
where is the reduced mass and ~r = ~rA ~rB is the relative distance between
the two particles. This is e.g. the case at LHC, where two protons scatter, but
with relativistic energies. But here we concentrate first on the fixed potential
case.
If the potential does not depend on time - as in our case - we can separate
the time dependence and we look for stationary solutions:
E
(~r, t) =: (~r)ei ~ t . (19)
The wave function (~r) has to satisfy the time-independent Schrodinger-

equation
~2 ~ 2

E(~r) = + V (~r) (~r) , (20)
2m
where the energy E of the particle can be written as
p~2 ~2~k 2 1
E =: =: =: m~v 2 , (21)
2m 2m 2
with
the incident momentum of the particle ~p = ~~k = m~v ,
the incident wave vector of the particle ~k,
the incident velocity of the particle ~v .
15
Now we can rewrite Eq.(20) as
h i
~ 2 + ~k 2 U(~r) (~r) = 0 , (22)
with the redefined potential

2m
U(~r) = V (~r) . (23)
~2
We will devote now several lectures to different solutions of Eq.(22)
2.2 The scattering amplitude

Now we derive a formula that directly relates the solution of the Schrodinger
equation to a measurable cross section. In this subsection we do this by
making several rough approximations in order to get fast to the desired result.
Later on in this lecture course we will directly derive the formula.
For large distances ~r we split up the wave function in an incoming part and
a scattered part

(~r) inc. (~r) + scat. (~r) . (24)
r
For large distances we typically can neglect the potential term in Eq.(22) and
we get the free equation
h i
~ 2 + ~k 2 (~r) = 0 ,
(25)
Now we describe the incoming particle with a plane wave, i.e. all the incom-
ing particles have the same momentum p~ = ~~k and travel all in the same
direction, which choose to be the z-direction.
~
inc. (~r) = eik~r = eikz . (26)
This fulfils Eq.(25). With this normalisation we have for the particle density
= |inc. (~r)|2 = 1, i.e. one particle per unit volume. Thus the incident flux
reads
p
F = v= , (27)
m
1 m 1
2
= . (28)
ms s m3
The outgoing particle we want to describe by an outgoing wave
eikr
scat. (~r) = f (k, , ) , (29)
r
16
with the scattering amplitude f (k, , ). The factor 1/r is needed for the
conservation of probability. This wave function also fulfils Eq.(25) for very
large values of r.
Thus we get for the full asymptotic wave function:
i~k~r eikr
~k e + f (k, , ) . (30)
r r
This ansatz can now be used to calculate the current density ~j
~j(~r) = ~ ()
h i
~ ~
( ) (31)
2mi
and one gets
~k |f (k, , )|2
jr = . (32)
m r2
jr represents the number of particles crossing a unit area per unit time and
the detector presents a cross-sectional area r 2 d to the scattered beam. The
number of particles entering the detector per unit time, dN is
~k
dN = jr r 2 d = |f (k, , )|2 d (33)
m
= v |f (k, , )|2 d (34)
F
= |f (k, , )|2 d . (35)

dN can also be described by the incoming flux F and the area d these
particles are crossing before the scattering:
d
dN = F d = F d . (36)
d
Equating the two expressions and using = 1 we get our master formula
d
= |f (k, , )|2 . (37)
d
The differential cross section is measured in the experiment, while the scat-
tering amplitude is calculated in theory, by solving the Schrodinger equation.
To prepare for next lecture: Chapter [7.3/4] of [1]
17
3 Spherical Bessel functions
Chapter [7.3/4] of [1]
Before continuing with the scattering problems we make some mathematics
interlude and introduce special functions that will be very helpful for the
solution of the Schrodinger equation.
3.1 Separation of variables

This time we consider a spherical symmetric potential, i.e.
V (~r) = V (r = |~r|) . (38)
To proceed we rewrite the Laplace operator in spherical coordinates
2

~ 2 1 2 1 1
= 2 r + 2 sin + 2 2 (39)
r r r r sin r sin 2
1

~2
L
= 2 r2 2 2 , (40)
r r r ~r
with the angular momentum operator L. ~ Next we separate the radial and
the angular dependence of the solution
X
(~r) = cElm Elm(~r) (41)
E,l,m
Elm (~r) = REl (r)Ylm (, ) , (42)
with the radial function REl that has to be determined and the well-known
spherical harmonics Ylm (TP2, Edynamics).
The spherical harmonics can be expressed in terms of associated Legendre
functions and these in terms of Legendre polynomials - see also the
corresponding Wikipedia articles.
Ylm (, ) = Clm eim Plm (cos ) , (43)

|m|
m
|m|
2 2 d
Pl (x) = (1 x ) Pl (x) , (44)
dx
l
1 d
Pl (x) = l (x2 1)l . (45)
2 l! dx
18
The following orthogonality relations hold
Z

Yl m Yl m
d = ll mm , (46)
Z1
2(l + m)!
Pkm (x)Plk (x)dx = ml , (47)
(2l + 1)(l m)!
1
Z1
2
Pm (x)Pn (x)dx = mn . (48)
2n + 1
1
Having a cylinder symmetry the solutions will be even more simple and there
is no dependence on m, which can then be set to 0 and thus the spherical
functions are simply replaced by Legendre polynomials:
Ylm (, ) Pl (cos ) . (49)
Coming back to the general case and using
~ 2 Ylm (, ) = l(l + 1)~2 Ylm (, ) ,
L (50)
we get for the time-independent Schrodinger equation in Eq.(22)
h i
~ 2 U(r) + ~k 2 Elm (~r) =
0 (51)
(" # )
1

L~2
r2 2 2 U(r) + ~k 2 REl (r)Ylm (, ) = 0 (52)
r 2 r r ~r

1 2 l(l + 1)
r U(r) + ~k REl (r) =
2
0 (53)
r 2 r r r2
2
2 l(l + 1)
+ U(r) + ~k REl (r) =
2
0 . (54)
r 2 r r r2
Eq.(54) is now the defining equation for the radial part of the Schrodinger
equation - the only missing part.
3.2 The simplest Bessel function

To get an idea how the solutions of Eq.(54) might look like, we start by
considering the free particle case, i.e. U(r) = 0. Defining uEl (r) := rREl (r)
we are left with
2
2 l(l + 1) ~ 2 uEl (r)
+ +k = 0 (55)
r 2 r r r2 r
2
l(l + 1) ~ 2
+ k uEl (r) = 0 . (56)
r 2 r2
19
Taking in addition l = 0 we get
2
~ 2
+ k uE0(r) = 0 , (57)
r 2
which is solved by
uE0(r) = A sin(kr) + B cos(kr) . (58)
Thus we finally get
sin(kr) cos(kr)
RE0 = A +B . (59)
r r
The cosine part gives a divergent function REl , at the origin, to avoid that
one simply can set B = 0.
3.3 The spherical Bessel equation

3.3.1 Definition
To solve Eq.(55) for general values of l we stay with the notion of the radial
function REl and make a change of variables to := kr to get
2

2 l(l + 1)
+ + 1 Rl () = 0 , (60)
2 2
where we denoted Rl () := REl (/k). Eq.(60) is called the spherical Bessel

differential equation. We know the solution for l = 0 already
sin() cos()
R0 () = Ak + Bk . (61)

3.3.2 Deriving the solution:

Now we either look into some Mathbook or we use some tricks:
Trick 1: For l 6= 0 we make the Ansatz Rl () =: l l () and Eq.(60)

turns into 2
2(l + 1)
+ + 1 l () = 0 . (62)
2
Trick 2: Next we try to derive a recursion relation.
Assume that l () is a solution of Eq.(62).
20
What equation holds for1
1 l ()
() := ? (63)

To answer that we differentiate Eq.(62) once more
3 2(l + 1) 2

2(l + 1)
+ + l () = 0 (64)
3 2 2
2
2(l + 1) 2(l + 1)
+ + 1 (()) = 0 (65)
2 2
2
2(l + 2)
+ + 1 () = 0 . (66)
2
The last line is simply the defining equation for l+1 (). Thus we have
the desired recursion relation
1 l ()
l+1 () = (67)

leading to
l
1
l () = 0 () . (68)

With 0 () given in Eq.(59) we get two fundamental solutions of Eq.(60):
the spherical Bessel functions (or Spherical Bessel functions of
the first kind) jl () and the spherical Neumann functions (or
spherical Bessel functions of the second kind) nl ().2
l
1
l l sin
Rl () = jl () := (1) , (69)

l
l l 1 cos
Rl () = nl () := (1) . (70)

The factors (1)l are just convention.
Eq.(55) is then solved by REl (r) = jl (kr) and REl (r) = nl (kr).
1
Clearly, quite some trial and error was necessary to get this idea!
2
The MATHEMATICA notation reads jl (x) = SphericalBesselJ[l, x] and nl (x) =
SphericalBesselY [l, x].
21
3.3.3 Properties of the solutions
The first few functions read
sin
j0 () = (71)

sin cos
j1 () = (72)
2

3 1 3
j2 () = 3
sin 2 cos (73)

and
cos
n0 () = (74)

cos sin
n1 () = 2 (75)

3 1 3
n2 () = 3
cos 2 sin (76)

Spherical Bessel functions j0 (x), j1 (x) and j2 (x)
1.0
0.8
0.6
0.4
0.2
5 10 15
- 0.2
- 0.4
Spherical Neumann functions n0 (x), n1 (x) and n2 (x)
22
1.0
0.5
5 10 15
- 0.5
- 1.0
For small values of these functions can be expanded as

2l l!
jl () l , (77)
(2l + 1)!
(2l)!
nl () . l (l+1) . (78)
2 l!
For large values of the functions can be approximated by

1 l
jl () sin , (79)
2

1 l
nl () cos . (80)
2
3.3.4 Bessel functions J

The spherical Bessel functions jl () and the spherical Neumann functions
nl (), that can both be expressed in terms of ordinary Bessel functions
J () of order .
21

jl () = Jl+ 1 () (81)
2 2
21

nl () = (1)l+1 Jl 1 () (82)
2 2
23
(1,2)
3.3.5 Spherical Hankel functions h
These functions are just linear combinations of the spherical Bessel and Neu-
mann functions
h(1)
() = j () + in () (83)

h(2) h(1)

() = j () in () = () (84)
3.3.6 Orthogonality relations
3.4 Summary of Bessel functions

Taking a vanishing potential in Eq.(54) we get
2
2 l(l + 1) ~ 2
+ + k REl (r) = 0 . (85)
r 2 r r r2
This equation is solved by
REl (r) = Ajl (kr) + Bnl (kr) . (86)
Thus the whole 3-dimensional Schrodinger equation is solved by
X
(~r) = cELM [Ajl (kr) + Bnl (kr)] Ylm (, ) . (87)
E,l,m
This means that spherical harmonics in combination with spherical Bessel

functions form a complete set of solutions for the free 3-dimensional Schrodinger
equation.
3.5 Plane waves

Plane waves are a different complete set of solutions for the free 3-dimensional
Schrodinger equation. Hence there is transformation between plane waves
and spherical harmonics in combination with spherical Bessel functions.
We give here the expression of a plane wave in term of spherical Bessel func-
tions and Legendre polynomials or spherical harmonics without derivation

i~k~
X
r
e = (2l + 1)il jl (kr)Pl (cos ) (88)
l=0
X
+l
~
X
= 4 il jl (kr)Ylm (k)Ylm (~r) . (89)
l=0 m=l
To repeat this lecture: Chapter [7.3/4] of [1]

24
4 The method of partial waves
Chapter [13.3/4] of [1]
4.1 Definition of partial waves
After our mathematics excursion we turn back to the ansatz for the general
scattering solution, given in Eq.(30).
~ eikr
~k = eik~r + f (k, , ) . (90)
r
If we have a central potential V = V (r), we have an azimuthal symmetry, i.e.

there is no dependence on the angle . We choose the z-axis such, that the
incoming particles fly along this axis. In that case the full wave function ~k
and the scattering amplitude f (k, ) can be expanded in terms of Legendre
polynomials

X
~k = Rl (k, r)Pl (cos ) , (91)
l=0
X
f (k, ) = fl (k)Pl (cos ) . (92)
l=0
Each term in the sum in Eq.(91) is called a partial wave and it is simultane-
ously an eigenfunction of L ~ 2 and Lz with eigenvalues l(l + 1)~2 and 0. Using
the spectroscopic notation we denote the l = 0, 1, 2, 3, , .. wave by s, p, d, f, ...
wave.
The coefficients fl (k) in the expansion of the scattering amplitude are called
partial wave amplitudes. According to our master formula for the cross
25
section, we get
d
= |f (k, )|2 (93)
d
X X
= fl (k)fl (k)Pl (cos )Pl (cos ) . (94)
l =0 l=0
Z+1
d d
Z
= d = 2 d cos (95)
d d
1
X
X Z+1
= 2 fl (k)fl (k) Pl (x)Pl (x)dx (96)
l =0 l=0 1

X 1 X
= 4 |fl (k)|2 = l . (97)
l=0
2l + 1 l=0
Thus a knowledge of the partial wave amplitudes gives immediately the total
cross section. The relation between the partial wave amplitudes and the
radial function Rl (k, r) in the partial wave can be derived as follows:
~ eikr
~k = eik~r + f (k, , ) (98)
r

X X X eikr
Rl (k, r)Pl (cos ) = (2l + 1)il jl (kr)Pl (cos ) + fl (k)Pl (cos )
l=0 l=0 l=0
r
(99)
eikr
Rl (k, r) = (2l + 1)il jl (kr) + fl (k) (100)
r
4.2 Structure of the solution - Phase shifts

Next we investigate a little more the general structure of the solution of the
radial equation Rl (k, r).
For large values of r > d we neglect the contribution of the potential and get
as a general solution
Rl (k, r) = Bl (k)jl (kr) + Cl (k)nl (kr) . (101)
The coefficients Bl and Cl will be determined by matching with the region

r < d, where the effects of the potential have to be taken into account. For
very large values of r we can use the asymptotic expression for the spherical
26
Bessel and Neumann functions

1 l l
Rl (k, r) = Bl (k) sin kr Cl (k) cos kr . (102)
kr 2 2
By defining
p
Al (k) := Bl (k)2 + Cl (k)2 , (103)
Cl (k) Cl (k) Bl (k)
tan l (k) := , sin l (k) := , cos l (k) := , (104)
Bl (k) Al (k) Al (k)
we get

Al (k) l l
Rl (k, r) = cos l (k) sin kr + sin l (k) cos kr
kr 2 2
(105)

Al (k) l
= sin kr + l (k) . (106)
kr 2
The l (k) are called phase shifts and they denote the strength of the scat-
tering in the l-th partial wave.
If there is no potential at all acting, then the solution in Eq.(102) has to hold
everywhere and in particular for r = 0. Since nl (r) is singular at r = 0, the
corresponding coefficient has to be set to 0, Cl (k) = 0. In that case we get
also a vanishing phase shift l (k) = 0.
Combing the asymptotic form of the radial function with our result from
Eq.(100) we get
l

l sin kr 2 eikr

Al (k) l
sin kr + l (k) = (2l + 1)i + fl (k) ,
kr 2 kr r
(107)
which can be solved by
Al (k) = (2l + 1)il eil (k) , (108)
2l + 1 2il (k)
fl (k) = e 1 . (109)
2ik
With this we get some final expressions for the scattering amplitude and for
the differential cross section:

1 X
(2l + 1) e2il (k) 1 Pl (cos ) ,

f (k, ) = (110)
2ik
l=0
4
l = |fl (k)|2 , (111)
2l + 1
4(2l + 1) 2
= sin l (k) . (112)
k2
27
Therefore the knowledge of the scattering phase is sufficient to solve a scat-
tering problem.
Setting = 0 in Eq.(110) we get
il (k)
eil (k)

1X il (k) e
[f (k, 0)] = (2l + 1) e (113)
k 2i
l=0

1X
= (2l + 1) sin2 l (k) (114)
k l=0
k
= . (115)
4
This is the famous Optical Theorem: the total cross section is given by
the forward ( = 0) scattering amplitude.
4.3 Convergence of the partial wave expansion

For large values of l the term l(l+1)/r 2 will dominate over the potential
U(r) thus the latter one can be neglected and the solution is equal to
the free solution, having no phase shift and thus no contribution to the
cross section.
An estimate gives that values of l > ka can be neglected, if a being the

extension of the potential.
For k 0 only l = 0 contributes (s-wave scattering). Defining the

scattering length ls as
tan 0 (k)
ls := lim , (116)
k0 k
one finds

f k0
ls , (117)
d
k0
ls2 , (118)
d

k0
4ls2 . (119)
4.4 Scattering by a square well

-
28
4.5 Resonances
According to Eq.(112) the l-th partial cross section reads
4 1
l = (2l + 1) . (120)
k 2 1 + cot2 l (k)
The energy where cot l (k) becomes zero is called resonance energy ER .
At the resonance one can write
ER E
cot l (k) = , (121)
(ER )/2
with the resonance width (ER ). This leads to the Breit-Wigner reso-
nance formula
4 2 (ER )/4
l = (2l + 1) , (122)
k2 (E ER )2 + 2 (ER )/4
which is valid in the vicinity of a resonance. For M ER it looks like:
In practice the resonances are typically superimposed by background scatter-

ing, e.g. from other partial waves. Here we show an example for resonances
found at the LHC, when looking for the dimuon final state in proton scat-
tering: p + p + + .
29
4.6 Hard-sphere potential
-
30
5 The integral equation of potential scatter-
ing
In this lecture we start considering non-trivial potentials and we will re-

express Eq.(22) by an integral equation.
5.1 Definition of the Greens function

Eq.(22) can be written as
h i
~ 2 + ~k 2 (~r) = U(~r)(~r) . (123)
The general solution of this equation can be expressed as

Z
(~r) = 0 (~r) + G0 (k, ~r ~r )U(~r )(~r )d~r , (124)
with 0 (~r) being a solution of the homogeneous Schrodinger equation

h i
~ 2 ~ 2
+ k 0 (~r) = 0 (125)
and the Greens function G0 (k, ~r ~r ), which is defined as
h i
~ 2 + ~k 2 G0 (k, ~r ~r ) = (3) (~r ~r ) . (126)
If we choose the homogeneous solution to be a plane wave, we get
~
0,~k (~r) = eik~r , (127)
Z
~
~k (~r) = eik~r + G0 (k, ~r ~r )U(~r )~k (~r )d~r . (128)
5.2 Deriving the Greens function

In order to derive the Greens function, we consider the Fourier transforma-
tion of the definition of the Greens function, Eq.(126).
d3 q
Z
G0 (k, ~r) = G0 (k, ~q)ei~q~r , (129)
(2)3
d3 q 2 d3 q i~q~r
Z Z
2 2 i~ r !
q ~ (3)

+ k G0 (k, ~r) = (k q ) G 0 (k, ~
q )e = (~
r ) = e ,
(2)3 (2)3
(130)
1
G0 (k, ~q) = 2 . (131)
k q2
31
Now we have a solution for the Fourier transform of the Greens function, so
we are left with transforming G0 (k, ~q) back to G0 (k, ~r).
d3 q ei~q~r
Z
G0 (k, ~r) = (132)
(2)3 k 2 q 2
Z
Angular integr. 1 qeiqr
= 2 dq (133)
4 ir q2 k2

Z
eiqr eiqr

PFD 1 1
= + dq (134)
4r 2i q+k qk

ikr
Cauchy e
= . (135)
4r
This is the final result for the Greens function!
5.3 Lippmann-Schwinger equation

Thus the general solution of Eq.(22) reads
Z ik|~r~r |
i~k~ 1 e
~k (~r) = e r
U(~r )~k (~r )d~r . (136)
4 |~r ~r |
This integral equation is called the Lippmann-Schwinger equation.
For a potential with a finite extension, i.e. U(r) = 0 for r > d, we can
investigate the limit r and r < d r to get
~r
|~r ~r | r ~r + ... , (137)
r

eik|~r~r | eikr i~k ~r
e + ... , (138)
|~r ~r | r
with ~k = ~r/r k. This leads to

~ eikr
~k (~r) = eik~r + f (k, , ) , (139)
r
with
1
Z
~
f (k, , ) = eik ~r U(~r )~k (~r )d~r
4
1
= h0,~k |U|~k i . (140)
4
Remarks:
32
This is now the promised derivation of Eq.(30).
The transition matrix element T~k~k is defined as
T~k~k := h0,~k |V |~k i . (141)
The differential cross section can be expressed as
d m2 2
= 2 4 T~k~k . (142)
d 4 ~
33
6 The Born approximation
In this lecture we finally derive an approximation for the solution of the

general scattering problem.
6.1 The Born series

Eq.(128) can be also written as
Z
i~k~
r
~k (~r ) = e + G0 (k, ~r ~r )U(~r )~k (~r )d~r . (143)
Inserting Eq.(143) into Eq.(128) gives

Z Z
i~k~ i~k~
r
r
~k (~r) = e + G0 (k, ~r~r )U(~r ) e + G0 (k, ~r ~r )U(~r )~k (~r )d~r d~r .

(144)

In the next step we replace ~k (~r ) by the whole integral solution and so on.
In the end we get an expansion like
~
0 (~r) = eik~r , (145)
Z
~ ~
1 (~r) = eik~r + G0 (k, ~r ~r )U(~r )eik~r d~r , (146)
Z
i~k~ ~
2 (~r) = e r
G0 (k, ~r ~r )U(~r )eik~r d~r
+
Z Z
~
+ G0 (k, ~r ~r )U(~r ) G0 (k, ~r ~r )U(~r )eik~r d~r d~r (147)

,
... = ...
If the potential U gives rise to a small perturbation of the system, then such
an expansion might converge and the real result could be approximated by
the first terms.
6.2 The first Born approximation

We get for the first Born approximation of the scattering amplitude
1
Z
~ ~
B
f (k, , ) = eik ~r U(~r)eik~r d~r (148)
4
1
Z
~
= ei~r U(~r)d~r , (149)
4
34
~
with the momentum transfer ~.
~ := ~k ~k .
(150)
Thus in Born approximation the scattering amplitude is simply the Fourier

transform of the potential.
For elastic scattering k = k holds and thus ~k ~k = k 2 cos . Hence we get
~ 2 = 2k 2 2~k ~k
2 = ||
= 2k 2 (1 cos )

= 4k 2 sin2 ,
2

= 2k sin . (151)
2
Therefore the Born scattering amplitude for a central potential reads
Z Z Z2
B 1 2
f (k, , ) = drr U(r) d sin deir cos (152)
4
0 0 0
Z Z+1
1
= drr U(r) d cos eir cos
2
(153)
2
0 1
Z
1
= drr sin(r)U(r) . (154)

0
6.3 The Coulomb potential

Taking the Yukawa potential as an example
er
UY (r) = U0 , (155)
r
we get
U0
fYB = (156)
2 + 2
and thus
d B U02
= . (157)
d (2 + 2 )2
35
The Coulomb potential can be considered as a special case of the Yukawa
potential
1
UC (r) = U0 (158)
r
and we get
U0
fCB = , (159)
2
leading to the famous Rutherford scattering formula
d B U02
= (160)
d 4
2
qA qB 1
= . (161)
40 16E sin4
2
2
6.4 Propagator
The Greens function in
Z
i~k~ ~
n (~r) = e + G0 (k, ~r ~r )U(~r )eik~r d~r
r
Z Z
~
+ G0 (k, ~r ~r )U(~r ) G0 (k, ~r ~r )U(~r )eik~r d~rd~r

+... (162)
is also called propagator. Feynman diagrams are a graphical display

of the terms in Eq.(162). The first term corresponds to no interaction, the
second term to an interaction at ~r and a propagation from ~r to ~r. The third
term to an interaction at ~r , a propagation from ~r to ~r , an interaction at
~r and a propagation from ~r to ~r, a.s.o.
36
7 Advanced Topics in Scattering
7.1 Form factor
Assume that the potential is given by an extended distribution of electric
charge. Considering the interaction of this distribution with another charge
e we have the following potential
Ze2
Z
V (~r) = (r )dr~ , (163)
s
with ~s = ~r r~ . For the scattering amplitude we obtain

m Ze2 ~
Z Z
~ r
i~
f (k, ) = e (r )dr d~r (164)
2~2 s
mZe2 1 ~
Z Z
~ r~ +~s)
i(
= 2
e (r )dr d~r (165)
2~ s
mZe2 ~ 1
Z Z
~ r~
i ~
= 2
e (r )dr ei~s d~r (166)
2~ s
2
mZe ~ ~ ~ 1
Z
= 2
(k k ) ei~s d~s (167)
2~ s
= (~k k~ )fRutherf ord (k, ) . (168)
The Fourier transform of the charge distribution, , is called the form factor
of the extended object. This is a very general result: the cross-section of an
extended object is given by the cross-section of the corresponding point like
object times the form factor squared.
dExtended dP ointlike
= ||2 . (169)
d d
The Fourier transformation of the charge distribution can now be obtained
by comparing the measurement of the differential cross section with the cal-
culation of the corresponding point-like cross section.
7.2 Electron scattering on nucleon

Thus information about the size of a nucleon or more generally about the
charge distribution of a nucleon can be obtained by investigating the scat-
tering of electrons and nucleon. For a realistic investigation, however, more
effects have to be taken into account:
a) Scattering of an electron at a static, fixed potential
37

d d
= .
d d Rutherf ord
b) Electron has spin 1/2

d d 2 2
= 1 sin ,
d M ott d Rutherf ord 2
with = v/c.
c) Finite mass of the nucleon

E

d d
= .
d M ott d M ott E
4EE
Four momentum transfer: q 2 = (p p )2 c2
sin2 2 ; Q2 := q 2 .
d) The nucleon has a magnetic moment

d d 2
= 1 + 2 tan ,
d s=1/2P oint d M ott 2
with = Q2 /(4M 2 c2 ).
e) The nucleon has an anomalous magnetic moment (g 6= 2) and it is

extended (form factor).
gp
p = N = +2.79N = GpM (Q2 = 0) ,
2
gn
n = N = 1.91N = GnM (Q2 = 0) ,
2
eh me
nuclear magneton: N = = B .
2mp mp
38
One would expect the following structure:

d d 2 2 2
= F (Q ) 1 + 2 tan ,
d d M ott 2
with F (0) = g/2.
The electric charge and the magnetic current have, however, different
distributions, thus we get two different form factors. Due to historic
reasons one writes:
G2E (Q2 ) + G2M (Q2 )

d d
= + 2 G2M (Q2 ) tan2 .
d d M ott 1+ 2
Rosenbluth-formula (1955)
Experimental investigations have given the following results:

GpM (Q2 ) Gn (Q2 )
GpE (Q2 ) = = M = GDipole (Q2 ) ,
2.79 1.91
1
GDipole (Q2 ) = 2 ,
Q2
1+ 2
0.71GeV
GnE (Q2 ) 0 .
In the end one gets e.g. information about the structure and the size of a
proton: it turns out that a good fit for the charge density of the nucleon is
given by a Fermi distribution
1
(r) = 0 , (170)
1 + exp rR

a
with R 1f m, a 0.5f m and 0 3 1017 kg/m3 .

A state-of-the quantum field theoretical investigation of nucleon form factors
can e.g. be found in [20].
7.3 Collisions between identical particles

If we consider the scattering of two identical particles, then an event with

the angular coordinates (, ) cannot be distinguished from an event with
39
the angular coordinates ( , + ). Thus interference can occur and
the corresponding scattering amplitudes have to be added. We get for the
differential cross section (for finding a particle at (, )
d
= |f (, ) + f ( , + )|2 (171)
d
= |f (, )|2 + |f ( , + )|2 + 2 (f (, )f ( , + )) ,
(172)
which differs from the classical expectation of

dClassic
= |f (, )|2 + |f ( , + )|2 . (173)
d
7.3.1 Two spin-less bosons

7.3.2 Two spin 1/2 fermions
7.4 Introduction to multichannel scattering

7.4.1 Elastic Scattering

7.4.2 Inelastic Scattering
To prepare for next lecture: Chapter [14.1/2/3] of [1]
40
8 The density matrix
Chapter [14.1/2/3] of [1]
8.1 Definition
Quantum statistical mechanics is a combination of statistical mechanics
and quantum mechanics. In this case we have two sources for the necessity
of a statistical description:
A practical necessity because of a huge number of systems or degrees
of freedom. Having an infinite amount of computing power, we sim-
ply could overcome this problem and we would not need a statistical
description.
An unavoidable necessity because of the inherent probability character

of quantum mechanics. This problem cannot be overcome, even with
infinite computer power.
We start our discussion of quantum statistical mechanics with some defini-
tions:
A system that can be described by a single wave function (state vector) is
said to be in a pure state. One implicitly assumes that a complete set of
commuting observables has been determined, which is called a maximal mea-
surement. Basically all systems we were considering till now are in a pure
state.
If in a system only non-maximal measurements are or can be made, then the
wave function is not completely known and we speak of a mixed state. Now
a statistical mixture of wave functions must be used to describe the system
- this is one of the topics of quantum statistical mechanics.
Another way of stating this difference: for a pure state all sub-systems are
in the same state, while for a mixed state the sub-systems can be in different
states.
To make these definitions a little more clear, we consider:
An ensemble of N sub-systems = 1, 2, ..., N.
Each sub-system is in a pure state described by the state vector ()

or |i in Dirac notation.
The |i are normalised to one, but they are not necessarily orthogonal.
h|i = 1 . (174)
41
Choose a complete set of basis vectors |ni, i.e. orthogonal eigen vectors
of some complete set of operators:
hn|n i = n,n (175)
or
hn|n i = (n n ) , (176)
if n is a continuous variable.
Completeness tells us X
1= |nihn| . (177)
n
The pure state |i can be expanded in terms of the basis |ni:

X
|i = c()
n |ni , (178)
n
with
c() = hn|i , (179)

X n2
c()
n
= 1. (180)
n
Next we consider an observable represented by an operator A. The expecta-

tion value of A in the state |i is
hAi = h|A|i (181)

X X ()

= cn c()
n hn |A|ni (182)
n n
XX
= hn|ih|nihn |A|ni . (183)
n n
The average of A over the ensemble - the ensemble or statistical average

is denoted by
XN
hAi = W hAi , (184)
=1
with W being the statistical weight of the pure state |i, i.e. the proba-
bility of finding the whole system in the state |i:
0 W 1 , (185)
XN
W = 1 . (186)
=1
42
All in all we can now write
N XX
X
hAi = hn|iW h|n ihn |A|ni (187)
=1 n n
XX
= hn||n ihn |A|ni (188)
n n
X
= hn|A|ni (189)
n
= Tr(A) , (190)
with the density operator or statistical operator, being defined as

N
X
= |iWh| . (191)
=1
Remarks:
The knowledge of the density operator enables us to calculate ensemble

averages.
Taking A to be the unit operator and keeping in mind that the pure
states are normalised to one, we have
Tr() = 1 . (192)
If the states |i are not normalised to one, we get
Tr(A)
hAi = . (193)
Tr()
The density operator can also be expressed in terms of matrix elements

- the density matrix.
Definition:
nn = hn||n i (194)
XN
= hn|iWh|n i (195)
=1
N
()
X
= W cn c()
n . (196)
=1
43
Diagonal elements: the diagonal elements of the density matrix
N
X 2
nn = hn|ni = W c()
n
(197)
=1
have the following interpretation:

W is the probability of finding the whole system in the state |i
2
()
and cn is the probability of finding that |i to be in the state
|ni. Thus nn gives the probability of finding the ensemble in the
state |ni.
Moreover we have
0 nn 1 . (198)
The l.h.s. follows from Eq.(197) and the r.h.s. from Eq.(192).
Hermiticity: the density matrix is hermitian
hn||n i = hn ||ni . (199)
Thus it can be diagonalised by a unitary transformation. If we

choose a representation |ki where is diagonal, then we have
kk = kk kk . (200)
In that basis we immediately see from Eq.(198)
Tr(2 ) Tr() = 1 . (201)
This relation is also valid in any other representation, since the

trace is invariant under unitary transformations. Thus we have
XX
|nn |2 1 . (202)
n n
Now let us assume that the whole system is in a pure state |i,
meaning W = . So the density operator reads
= |ih| , (203)
which is simply a projection operator with
( )2 = . (204)
Hence Eq.(201) becomes for a pure state
Tr( )2 = Tr() = 1 . (205)
44
This is now a new criterion for deciding whether a state was pure
or not!
Tr()2 = 1 state is pure , (206)

Tr()2 < 1 state is mixed . (207)
If the system is in a pure state we have also
hAi = h|A|i . (208)
If we use a diagonal representation in that case we get in addition
kk = k k . (209)
Thus one diagonal element is one and the rest zero.

The indices n, n denote all possible quantum numbers, if we are
e.g. only interested in spin, we have a reduced density matrix.
8.2 Polarisation
8.3 Von-Neumann equation
In the Schrodinger-picture we have at the time t0
N
X
(t0 ) = W |(t0 )ih(t0 )| . (210)
=1
The time evolution of the states is given as
|(t)i = U(t, t0 )|(t0)i , (211)
with
H
U(t, t0 ) = ei ~ (tt0 ) (212)
for a time-independent Hamiltonian. Thus the density operator at time t is
given as
N
X
(t) = W |(t)ih(t)| (213)
=1
= U(t, t0 )(t0 )U (t, t0 ) (214)
i H (tt0 ) +i H (tt0 )
= e ~ (t0 )e ~ . (215)
45
This last equation is equivalent to

i~ (t) = [H, (t)] , (216)
t
which is called the von-Neumann-equation or sometimes also .Liouville-
von-Neumann-equation
Remarks:
Eq.(216) is the equivalent of the Liouville-equation for the phase

space probability in classical mechanics.
d
= {H, } . (217)
dt
Eq.(216) looks also similar to the Heisenberg equation of motion for

an observable,
d
i~ AH = [H, AH ] + i~ AH , (218)
dt t
but keep in mind that in Eq.(216) the states are given in the Schrodinger-
picture.
The time evolution of the statistical average of an operator A that is

not explicitly time-dependent reads

hAi = Tr (A) (219)
t t

= Tr A (220)
t
i
= Tr (A [H, ]) (221)
~
i
= Tr (AH AH) (222)
~
i
= Tr ([A, H] ) . (223)
~
The density operator in the Schrodinger-picture is time-dependent!
To repeat this lecture: Chapter [14.1/2/3] of [1]
46
9 Quantum mechanical ensembles
Chapter [14.4/5] of [1]
In the following we will discuss different classes of ensembles. We consider

only systems in thermal equilibrium, thus the averages of physical observables
are constant. In that case we get a time-independent density matrix

=0, (224)
t
which is equivalent to H having no explicit time dependence and
[H, ] = 0 . (225)
Now we can choose a basis |ni in which both H and are diagonal
hn|H|mi = Hnm = En nm , (226)

hn||mi = nm = nn nm . (227)
The probability that a sub-system chosen randomly from the ensemble is in

the eigenstate |ni is nn . Since can be expressed in terms of H
= (H) , (228)
we can also express nn in terms of EN .

Below we present different functional dependencies of nn from En , which
define different ensembles.
9.1 Micro-canonical ensemble

A micro-canonical ensemble describes a closed system, i.e. we have for each
sub-system:
a fixed number of particles N ;
a fixed volume V;
the energy E lies in the small interval
E0 E E0 + .
Notation:
47
: the total number of distinct pure states in which a member of the
ensemble can be found
= (N , V, E0 , ) .
H: the Hamiltonian of each sub-system.
Now we make the fundamental postulates:
1. There is an equal probability of finding the system in any of these

states, i.e.
1/ for E0 E E0 +
nn = . (229)
0 else
2. The members of the ensemble do not interact, so the corresponding

wave functions do not interfere. This can be obtained by demanding,
that the wave functions of each member |i of the ensemble are dis-
tributed at random or equally: the phases of the probability amplitudes
()
cn are distributed at random.
The entropy of the system, is defined as
S = kB log , (230)
with the Boltzmann constant kB = 1.3806488(13) 1023 J K1 .

Remarks:
The definition agrees with the usual thermodynamical one, but now
is calculated within using quantum mechanics.
If each sub-system is in the same state we have = 1 and thus S = 0

and the system is perfectly ordered.
If is greater than 1, the system is disordered. Thus S is a measure of

disorder.
The entropy has the familiar properties:
1. The entropy of a closed system is maximal, when the system is in

equilibrium.
2. The entropy is additive, i.e. if a closed system is divided into parts

with the entropy S1 and S2 we have S = S1 + S2 .
48
3. The temperature T of the system is defined as
1 S
= . (231)
T E
In practice the micro-canonical ensemble is difficult to use, thus Gibbs in-

troduced two additional concepts: the canonical ensemble and the grand
canonical ensemble.
9.2 Canonical ensemble

Definition:
In this case the investigated system S is in thermal contact with a much

larger heat reservoir R. The combined system S + R is closed and its
statistical properties are defined by the micro-canonical ensemble.
The system S is composed of a fixed number of particles NS within

a volume VS and it is in the energy eigenstate ES . The system R is
composed of a fixed number of particles NR within a volume VR and
it is in the energy eigenstate ER .
The systems S and R are in contact and can thus exchange energy -
the interaction is supposed to be so weak, that at any time S and R
are in definite energy eigenstates with energies ER and ES .
In equilibrium S and R have a common temperature T .
Conservation of energy tells us
E = ER + ES , (232)
where E is an eigenvalue of the combined system. Since the whole

system is described by the micro-canonical ensemble we have
E0 E E0 + with E0 .
Problem: Determine the probability that at a given temperature T the system

S is to be found in a state with energy ES .
Solution:
d(E): total number of distinct states of the combined system with

energies in [E, E + dE].
49
The probability of finding the combined system with energies in [E, E +
dE] is
Cd(E) E0 E E0 +
dP = , (233)
0 else
where C is a constant describing the probability of finding one state.
Denoting the corresponding numbers in the system S and R by dS
and dR , we have
d = dS dR , (234)
dP = CdS (ES )dR (ER )(E ES ER ) , (235)
where the -function expresses the energy conservation.
dPn is the probability that the system S has the energy En without
regard to the state of the reservoir (i.e. we are summing over all the
states in the reservoir that yield the same En )
Z
dPn = CdS (En ) (E En ER )dR (ER ) (236)
= CdS (En )R (E En ) , (237)
with R (E En ) being the total number of states in the reservoir,
subject to the conservation of energy.
The corresponding entropy of the reservoir is given as
SR = kB log R , (238)
SR
R = e kB
. (239)
Since the system S is supposed to be very small compared to the total

system, we have En E and thus we can Taylor expand:
SR (ER ) = SR (E En ) (240)
SR (E)
= SR (E) En + ... (241)
E
En
= SR (E) + ... . (242)
T
Putting now the pieces together we get
dPn = CdS (En )R (E En ) (243)
SR (EEn )
= CdS (En )e kB
(244)
SR (E)
kEnT
= CdS (En )e kB
e B (245)
kEnT
= Ae B dS (En ) . (246)
50
This is already the desired result: the probability of finding the system
En
S in the energy state with En is given by Ae kB T .
Thus we can write the density matrix as
nm = AeEn nm , (247)
with = 1/(kB T ).
Remarks:
1. The coefficient A can be extracted from the condition Tr() = 1.
X
1 = nn (248)
n
X
= AeEn (249)
n
X
QNS := A1 = eEn = Tr eH .

(250)
n
QNS is called the partition function.
2. The sum runs over all distinct states, which is not necessarily the same
as over all energy eigenstates.
3. The density operator can now be written as
eH
= . (251)
QNS
4. The ensemble average of an operator A is now given as
hAi = Tr(A) (252)

1
Tr eH A

= (253)
QNS
Tr eH A

= . (254)
Tr [eH ]
5. The average energy E of the system can be found by setting A = H.

Using Eq.(250) it can also be expressed as
1 QNS
E = hHi = = log QNS . (255)
QNS
51
9.3 Grand canonical Ensemble
This is similar the canonical ensemble, but now we allow in addition the
exchange of particles between the system S and the reservoir R. The total
number of particles
N = NR + NS (256)
is constant.
All derivations are similar up to difference in the Taylor expansion of the
entropy, where now the change of particle numbers is taken into account.
SR (E En , N NS ) = SR (E, N )
SR (E, N ) SR (E, N )
En (NS ) NS + ... .
E N
En (NS )
= SR (E, N ) + NS + ... .
T T
(257)
The density matrix gets now the form
nm = Be[En (NS )NS ] nm , (258)
with B being a constant and = (SR /N /kB ). = / is called
chemical potential.
In order to give an operator form of the density operator, we introduce the
number operator S , that is counting the number of particles in the system
S. Its eigenvalues are N = 0, 1, 2, . . . . Thus we get for the density operator
= BeH(S )S . (259)
The constant B can again be determined from Tr() = 1 and we get
B1 := Z(, T ) = Tr eH(S )S .

(260)
Z is called the grand partition function and we can write
XX
Z(, T ) = eEn (N )N (261)
N n
" #
X X
= eEn (N ) eN (262)
N n
" #
X X N
eEn (N ) e

= (263)
N n
X
= QN (T )z N , (264)
N
where z = exp() is called fugacity and QN (T ) is the partition function

of a system of N particles.
52
9.4 Spin 1/2 particle in a magnetic field
9.5 Average of a particle in a box
53
10 Systems of non-interacting particles
We discuss now systems of large numbers of non-interacting objects (parti-

cles) which are equivalent and possess the same energy levels.
Ej (j = 1, 2, 3, ..) denotes an energy level of one of the particles. The total
energy E of the system can be written as
X
ES = E = nj Ej , (265)
j
where the sum runs over all energy eigenstates and nj is the number of
particles having the same energy Ej . The total number of particles in the
system is X
NS = N = nj . (266)
j
The individual particle numbers can have the following values:

0, 1, 2, 3, ... Maxwell-Boltzmann particles distinguishable
nj = 0, 1, 2, 3, ... Bose-Einstein particles indistinguishable - bosons .
0, 1 Fermi-Dirac particles indistinguishable - fermions

(267)
For obtaining the partition functions we have to sum over all distinct states
and not only over all energy eigenstates. Thus we need to know the number
of distinct states gE having the same energy E.
To find that we start by looking at the total wave function of a system and
we find

Ej (1)Ek (2) El (N) Maxwell-Boltzmann
= total symmetric Bose-Einstein . (268)
total antisymmetric Fermi-Dirac

This leads to
N!
MB
gE = j (nj !) Maxwell-Boltzmann
gE = g BE = 1 Bose-Einstein . (269)
EF D
gE = 0, 1 Fermi-Dirac
Remarks:
ad MB: If all energies Ej , Ek , ..., El are different, then each energy level
is N!-fold degenerate. If, however, nj particles with nj > 1 have the
same energy, then an interchange among them does not change the
wave function.
54
ad BE: there is only one total symmetric wave function.
ad FD: there is only one total anti-symmetric wave function. If nj > 1

then gE becomes zero.
Now the partition function reads for all the three cases:
" #
X X
QN (T ) = gE exp nj Ej , (270)
{nj } j
P
where denotes a sum over every possible set of numbers nj that is com-
{nj }
patible with the total number N.
The average value of the energy hEi of the system can be calculated if the
average value hni i of each occupation number ni is known.
X
hEi = Ei hni i . (271)
i
hni i can be determined from the canonical ensemble

" #
1 X X
hni i = gE exp nj Ej ni , (272)
QN (T ) j
{nj }
which is equivalent to
log QN (T )
hni i = . (273)
(Ei )
For the case of MB the sum can be performed explicitly to obtain
" #N
X
QN (T ) = exp (Ej ) , (274)
j
leading to
hni i = exp [ (Ei + )] , (275)
with
P
exp (Ej )

j
= log . (276)
N
55
For the case of BE and FD it is easier to start with the grand partition
function
X
( )
X X X
Z = . . . gE exp [nj Ej (N)] nj (277)
n1 =0 n2 =0 j j
X X Y
= . . . gE exp {nj [Ej (N) + ]} (278)
n1 =0 n2 =0 j
YX
= gE exp {nj [Ej (N) + ]} . (279)
j nj =0
In the case of BE we have gE = 1 and nj has values from 0 to . Thus we

have a simple geometric series and get
Y 1
Z BE = . (280)
j
1 exp { [Ej (N) + ]}
In the case of FD we have gE = 0, 1 and nj has only the values 0 and 1. So

we simply can write down the two terms and get
Y
ZF D = [1 + exp { [Ej (N) + ]}] . (281)
j
The average number of states is now given as

( )
1 X X X X
hni i = . . . gE exp [nj Ej (N)] nj ni (282)
Z n =0 n =0 j j
1 2
log Z
= . (283)
[Ei (N)]
Now all three cases (MB, BE and FD) can be expressed with one formula:
1
hni i = (284)
exp [Ei (N) + ] +
with
0 MB
= 1 BE . (285)
+1 FD

The function
1
hn(E)i = (286)
exp [E + ] +
is called the distribution function. Plotted as a function of E + one
gets
56
4
-4 -2 2 4
Remarks:
For large values of E + all three cases become indistinguishable.
The FD-distribution cannot exceed one.
The BE-distribution diverges for E + 0. This is the case where

Bose-Einstein condensation occurs.
Finally can be determined from

X X 1
hNi = hnj i = . (287)
j j
exp [Ej (N) + ] +
57
11 The Klein-Gordon equation
11.1 Derivation
Using the correspondence principle

E i~ , (288)
t
~~
~p , (289)
i
the Schrodinger equation can be derived from the non-relativistic energy
momentum relation:
p~2
E = , (290)
2m
~2
~2
i~ (t, ~x) = (t, ~x) . (291)
t 2m
It is a priori clear, that the Schrodinger equation cannot be applied to rela-
tivistic energies.
If we want to discuss highly energetic phenomena, like particle accelerators,
early universe, black holes,.. we have to start from the relativistic energy mo-
mentum relation. This will immediately lead to the Klein-Gordon equa-
tion, which was derived by Schrodinger (1925), Gordon (1926) and Klein
(1927).
E 2 = p~2 c2 + m2 c4 (292)
2
~ 2 + m2 c4 )(t, ~x) .
~2 2
(t, ~x) = (~2 c2 (293)
t
The last equation can be rewritten as
1 2
mc 2
~ +
2
=0 . (294)
c2 t2 ~
Remarks:
The differential operator appears also e.g. in the wave equation in
electro dynamics. It is called DAlembert operator and it is Lorentz
invariant. This can be seen by writing
1 2 ~2,
= = = (295)
x x c2 t2
58
~ and
with x = (ct, ~x), x = (ct, ~x), = /x = (1/c /t, )
~
= /x = (1/c /t, ).
Thus (m can be written as p p and is assumed to be a scalar) the
whole Klein-Gordon equation is Lorentz invariant.
The Klein-Gordon-equation describes scalar particles like the Higgs-

boson.
The Klein-Gordon-equation, Eq.(294), is a second order equation with

respect to time, while the Schrodinger-equation was a first-order equa-
tion.
~/(mc) is the Compton wave length.
11.2 Charged particle in an electromagnetic field

In order to describe a scalar particle with charge q in an electro-magnetic
~ r , t) and a scalar potential (~r, t) we
field described by a vector potential A(~
use the principle of minimal coupling. This means we make the following
replacements3
E E q , (296)
q~
p~ ~p A . (297)
c
Hence the energy-momentum relating changes to
q~ 2 2
(E q)2 = (~p A) c + m2 c4 , (298)
c
leading to
2 " 2 #
~~ q~
i~ q (t, ~x) = A c2 + m2 c4 (t, ~x) . (299)
t i c
By making the ansatz

mc2
(t, ~x) = (t, ~x)ei ~
t
(300)
we split of the rest energy of the particle. If the particle is moving non-
relativistic then this is the dominant part of the energy and one can investi-
gate the non-relativistic limit. We will do this in the next section in a slightly
modified way.
3 ~
In 4-vector notation it reads: p p qc A , with A = (, A).
59
11.3 Stationary solutions
Now we look for stationary solutions
E
(~x, t) =: (~x)ei ~ t , (301)
of the Klein-Gordon equation, given in Eq.(299).
Assuming that A~ and are time independent, we get
" 2 #
~ q
~ A ~ c2 + m2 c4 (~x) .
(E q)2 (~x) = (302)
i c
~ = 0 and to be spherical symmetric, we get

If we choose in addition A
(E q)2 m2 c4 (~x) = ~2 c2 ~ 2 (~x) .

(303)
Now we can separate the solution in spherical coordinates
(~x) = REl (r)Ylm (, ) , (304)
leading to
2
2 2 d 2 d l(l + 1) 2 2 4

~ c + REl (r) = (E q(r)) m c REl (r)
dr 2 r dr r2
" ( 2 )#
d2

2 d l(l + 1) E q(r) mc 2
2
+ + REl (r) = 0 .
dr r dr r2 ~c ~
(305)
Remarks:
This is the same structure as in the case of the Schrodinger-equation,
given in Eq.(54). We get the old result, if we replace the curly bracket
by k 2 U(r).
For the non-relativistic limit we write the total energy as
E = E + mc2 (306)
and we assume E mc2 and also |q| mc2 . We then have
2
(E q)2 m2 c4 = E + mc2 q m2 c4 (307)
" 2 #

E q
= m2 c4 1+ 1 (308)
mc2
2mc2 (E q) . (309)
60
This gives
2
d 2 d l(l + 1) m
+ + 2 2 (E q(r)) REl (r) = 0 ,
dr 2 r dr r2 ~
(310)
which gives Eq.(54), when we equate

2mE
k2 = , (311)
~2
2m
U(r) = q(r) . (312)
~2
So we have proven that the non-relativistic limit of the Klein-Gordon
equation gives the Schrodinger equation.
11.4 Interpretation
Looking at the Klein-Gordon equation a little more in detail one immediately
finds some seemingly inconsistencies.
1. In the same way as it was done for the Schrodinger-equation one can
also derive a continuity equation for the case of the Klein-Gordon equa-
tion.
~ ~
+j =0. (313)
t
In this case the probability density (~r, t) and the probability current
density ~j(~r, t) reads

i~
(~r, t) = , (314)
2mc2 t t
~j(~r, t) = ~ ()

~ ~ ) .
( (315)
2mi
Here one easily sees that the probability density can also have negative
values, prohibiting a simple interpretation of this quantity.
2. The free Klein-Gordon equation in Eq.(294) can be solved by plane
waves
~
(~r, t) = Aei(k~rt) . (316)
Inserting this ansatz into Eq.(294) one gets
~2 2 = m2 c4 + ~2 c2 k 2 (317)

E = m2 c4 + ~2 c2 k 2 . (318)
61
Now we have in principle a positive and a negative energy value; a
result that has to be understood somehow.
The source of all these problems lies actually in the fact, that we tried to
derive a relativistic single-particle wave function. Such an endeavour has,
however, to fail. Having access to very large energies, we can always have
pair creation. If we want to describe single electrons |e i interacting with
photons, we always will have the possibility of e.g. an electron sending out
a photon, which again splits up into an electron-positron pair. Thus the 3
particle state |e e+ e i has also to be taken into account, as well as states with
even more particle. This can only be done in the framework of Quantum
Field Theory, the topic of one of the Level 4 courses.
We will here , however, follow the historic path, where first another single
particle wave function was introduced.
62
12 The Dirac equation
12.1 Introduction
Dirac saw the root of some of the problems of the Klein-Gordon-equation

in the fact that this is a second order equation in time. Thus he tried hard
to get a first order equation from the relativistic energy-momentum relation
and finally succeeded.
Today we know that the correct language for relativistic processes is quan-
tum field theory and within this framework the mentioned problems of
the Klein-Gordon-equation disappear. Thus we do not have to care a priori
whether the wave equations are first or second order. Moreover a system of
n second order differential equations can always be written as a system of 2n
first order equations.
From a modern point of view, we start at looking at the possible represen-
tations of the Lorentz group, to get the Klein-Gordon-equation for s=0
particles, the Dirac-equation for s=1/2 particle, the Maxwell-equations
for s = 1 particles (Proca-equation for massive s= 1) and the Rarita-
Schwinger-equation for s = 3/2 particles.
For pedagogical reasons we will follow here, however, the historical route,
which led Dirac in 1928 [21] - to some extent by chance - to the right result
for the electron, by demanding the wave equation to be of first order in time.
12.2 Derivation
Dirac started from the following first order Ansatz

~c k 2
i~ = k + mc = H . (319)
t i
63
This equation should of course fulfil the relativistic energy-momentum rela-
tion. To test this requirement we look at H 2 :
2

2 ~c i 2 ~c j 2
~ = i + mc j + mc (320)
t2 i i
~mc3 i

2 2 i j i 2 4 2
= ~ c i j + ( + )i + m c (321)
i
i j j i
~mc3 i

2 2 + i 2 4 2
= ~ c i j + ( + )i + m c
2 i
(322)
!
2 2 ij 2 4

=
~ c i j + m c . (323)
Thus i and have to fulfil the following conditions
i j + j i = 2 ij , (324)
i + i = 0 , (325)
2 = 1 . (326)
This clearly can not be fulfilled by simple numbers, but matrices might do
the job. i and are called Dirac-matrices.
The following form of the Dirac- matrices is called Dirac representation;
it fulfils Eq.(324-326):
0 k

k I2 0
= , = , (327)
k 0 0 I2
with the Pauli matrices

1 0 1 2 0 i 3 1 0
= , = , = . (328)
1 0 i 0 0 1
The Pauli matrices have the following properties

i j
, = 2iijk k , (329)
i j
, = 2 ij I2 . (330)
From this we immediately see
( i )2 = I2 , (331)
i j = iijk k for i 6= j , (332)
Tr( i ) = 0, (333)
det( i ) = 1 . (334)
64
With this definitions Eq.(324-326) can be easily checked.
Remarks:
I is obvious that the matrices in Eq.(327) are hermitian, i.e.
= , k = k . (335)
The form of the Dirac matrices is not unique! Every non-singular trans-
formation of Eq.(327) Mi M 1 , MM 1 fulfils also Eq.(324-326).
Mi M 1 Mj M 1 + Mj M 1 Mi M 1 = 2 ij , (336)
Mi M 1 MM 1 + MM 1 Mi M 1 = 0 , (337)
MM 1 MM 1 = 1 . (338)
Every unitary transformation of Eq.(327) gives again hermitian Dirac

matrices.

Mi M 1 = (M 1 ) (i ) M = Mi M 1 (339)
M 1 = M . (340)
This derivation also implies that the wave function is now a four
component object. It is called the Dirac-spinor.
12.3 Continuity equation

One of the motivations for deriving the Dirac-equation was to cure the prob-
lem of negative probabilities. Hence we derive here a continuity equation and
check whether the probability density is positive definite.
The Dirac spinor and its adjoint (complex conjugated and transpose)
spinor are defined as

1
2
= 3 ,
= (1 , 2 , 3 , 4 ) . (341)
4
We multiply the Dirac-equation from Eq.(319)b with from the left and
get
~c
i~ = k k + mc2 . (342)
t i
65
The adjoint relation to that reads
~c
i~ = k k + mc2 . (343)
t i
The difference of these two equations, Eq.(342) - Eq.(343) - gives
~c k
= k + k k + mc2 . (344)

i~
t i
Taking into account that k and are hermitian, i.e. k = k and =
we get
~c
= k k

i~ (345)
t i

k

+ ck = 0 (346)
t
~ ~j = 0 .
+ (347)
t
In the last line we defined the probability density to be
= = |1 |2 + + |4 |2 , (348)
which is obviously positive definite and the current density to be
jk = c k . (349)
The continuity equation can also be written in a co-variant form
j = 0 , (350)
with
j 0 := c , (351)
j := (j 0 , j k ) . (352)
12.4 Co-Variant formulation of Dirac theory

In order to write the Dirac equation in Eq.(319) in a form, where the Lorentz
invariance will become more obvious we multiply Eq.(319) with /(~c) from
the right-hand side:

1 k mc
i + k + = 0. (353)
ct i ~
66
With 0 = /(ct) and the definitions for the -matrices
0 := , (354)
k := k , (355)
this becomes
mc
i + = 0. (356)
~
Remarks:
Hermiticity: 0 is hermitian and k is anti-hermitian.

0 = = = 0 , (357)

k = k = (k ) = k = k = k . (358)
Defining equation for the -matrices:
( 0 )2 = ()2 = 12 , (359)
2
k = k k = k k = 12 , (360)
0 k + k 0 = k + k = 0 , (361)
k l + l k = k l + l k = 0 (l 6= k) . (362)
This can be written compactly as
{ , } = + = 2g 14 . (363)
Co-Variant -matrices:
= g . (364)
Different representations of the -matrices. Every non-singular trans-

formation of fulfils also Eq.(363). Examples are the Majorana
representation or the chiral representation.
Using the so-called Feynman-slash, that abbreviates the contraction of an
arbitrary four-vector B with the -matrices
B ~ ~ ,
/ := B = B = B 0 0 B (365)
the Dirac equation becomes now

mc
/
i + = 0. (366)
~
67
Using in addition natural units the Dirac equation becomes very simple

i/ + m = 0 . (367)
68
13 Lorentz-invariance of the Dirac equation
13.1 Lorentz transformation

In order to discuss the Lorentz-invariance of the Dirac equation we start by
setting our notation. The space-time vectors are given as a contra-variant
vector 0
x ct
x1 x
x =
x2 = y
(368)
x3 z
or as a co-variant vector

x0 ct
x1 x
x2 = y
x = .
(369)
x3 z
The metric tensor is given by

1 0 0 0
0 1 0 0
g 0 0 1 0 .
= (370)
0 0 0 1
This gives the following relations
g = g , (371)
x = g x , (372)
x = g x , (373)
g = g g = . (374)
A Lorentz-transformation4 is a transformation between two systems, that
move relatively to each other with a constant velocity. The most general form
is given by the inhomogeneous Lorentz-transformation or Poincare-
transformation; it reads

x = x + a . (375)
4
Named after Hendrik Antoon Lorentz (1853-1928, NL, NP 1902) - not to confuse
with the Lorenz gauge from Ludvig Lorenz (1829 -1891, DK). There is also a Lorentz-
Lorenz equation in the literature...
69
These transformations consist of translations (a ) and the homogeneous
Lorentz-transformations ( ), which can again be split up in Rotations
(in space), Lorentz-boosts and some discrete transformations like parity
(space-reflection) or time-reflection5
Remarks:
The differential operator transforms as

x
= = (376)
x x x

= = . (377)
x
It has the same transformation property as x .
Every 4-component object that transforms as given in Eq.(375), is

called a 4-vector. Examples are x and , the four-momentum p
and the vector potential A . Scalar products of four-momenta have the
same value in all systems!
The DAlembert operator has to be invariant under Lorentz-transformations
= = g (378)
= g (379)
!
=
g , (380)
which gives the relation
g = g gt = g . (381)
The last equation actually defines the homogeneous Lorentz group and
it gives
det = 1 , (382)
0 0 1 or 0 0 1 . (383)
The proper Lorentz-transformations are defined by det = +1

and 0 0 1. They consist of rotations of the three-dimensional space
5
An investigation of such a symmetry might sound to be a very academic exercise,
but it actually turns out that the study of this symmetry and its violations is crucial to
understand the existence of matter in the universe, see e.g. [22].
70
and of Lorentz-boosts.
An example for a space rotation is e.g.

1 0 0 0
0 cos sin 0
R12 = 0 sin cos 0
,
(384)
0 0 0 1
an example for a Lorentz-boost is e.g.

1 0 0
cosh sinh 0 0 1 2 1 2
sinh cosh 0 0 1 0 0
L01 = = 1 2 1 2 ,
0 0 1 0
0 0 1 0
0 0 0 1
0 0 0 1
(385)
with = v/c and tanh = . The representation of the full Lorentz-
group is described in detail in [5].
13.2 Lorentz invariance of the Dirac equation

Now we want to see how the Dirac equation transforms when performing a
Lorentz-transformation. We consider two systems:
System I : Coordinate x spinor

(386)
System I : Coordinate x spinor
The transformations from I to I read
x = x + a , (387)
(x ) = S()(x) , (388)
with an unknown transformation S().

Starting from the Dirac equation in System I
mc
i + (x) = 0 , (389)
~
we transform and (x) into the system I
= , (390)
(x) = S()1 (x ) , (391)
71
to get
mc
i + S()1 (x ) = 0 (392)
~
mc
iS() S()1 + (x ) = 0 . (393)
~
In the second line we have multiplied everything with S() from the r.h.s. .
The Dirac equation is now form invariant, if
S() S()1 = (394)
= S()1 S() . (395)
Remarks:
Any four component that transforms as given by Eq.(388) is called a
Dirac-spinor.
The transformation matrix S is defined by Eq.(395).
A detailed construction of S is described in detail e.g. in [5]. Here two
examples.
1. A space rotation around the z-axis:

1 0 0 0
0 cos sin 0
R12 = 0 sin cos 0

0 0 0 1
cos 2 + i sin 2

0 0 0
0
cos 2 i sin 2 0 0
SR12 = .
0 0 cos 2 + i sin 2 0
0 0 0 cos 2
i sin 2
(396)
Note that the matrix R12 has a periodicity of 2, while SR12 has
a periodicity of 4.
2. A Lorentz-boost along the x-axis

cosh sinh 0 0
sinh cosh 0 0
L01 =
0 0 1 0
0 0 0 1

sinh 2

cosh 2 0
0 cosh 2 sinh 2 0
SL01 = 0
. (397)
sinh 2 cosh 2 0
sinh 2 0 0 cosh 2
72
13.3 Bilinear forms
The adjoint spinor is defined as
= 0 . (398)
It transforms as
= S = S (399)
= S 0 . (400)
One can show (e.g. [5]) that
S 0 = b 0 S 1 , (401)
with
+1 for 00 +1

b= . (402)
1 for 00 1
Thus we get for the transformation of the adjoint spinor
= S 1 , (403)
if we only consider transformations that do not change the direction of time.
Now we can derive the transformation properties of the following bilinears
= ,

= , (404)
= ,
with
i
:=
[ , ] . (405)
2
Thus we get the following transformation properties
1. transforms as a Lorentz-scalar.
2. transforms as a Lorentz-vector. alone is not a Lorentz-
vector! It is always given by the Definition in Eq.(363).
3. transforms as a Lorentz-tensor.

73
14 Solutions of the Dirac equation
Chapter [15.4/6] of [1]
14.1 Plane wave solutions of the Dirac equation
We are looking now for solutions of the free Dirac equation of the form
~~
p r Et p x
(~r, t) = A u ei ~ = A u ei ~ . (406)
A is a constant, u is a four component spinor being independent from the
space-time coordinate x . Inserting this ansatz into Eq.(319) we get
Eu = ck pk + mc2 u .

(407)
To get a first idea about the structure of this solution we consider the case
of a particle at rest, i.e. ~p = 0. We get
mc2 I2

2 0
Eu(0) = mc u(0) = u(0) . (408)
0 mc2 I2
This equation has two different eigenvalues E+ = +mc2 and E = mc2 ,
that occur each twice and the following four eigenvectors

1 0 0 0
0 1 , u(3) (0) = 0 , u(4) (0) = 0 .
u(1) (0) = (2)

0 , u (0) = 0 1 0
0 0 0 1
(409)
The first two spinors describe a spin 1/2 particle and the second two a spin
1/2 anti-particle.
The positron, the anti-particle of the electron was discovered in 1932, about
4 years afterwards the Dirac-equation was postulated. This created also a
whole new field of science-fiction literature, see e.g. [23].
If we have a solution of the form c1 u(1) (0) + c2 u(2) (0) + c3 u(3) (0) + c4 u(4) (0)
in the system, where the particle is at rest ~p = 0, then we get in the system
that has a finite momentum ~p in x-direction

~
r E t p x
i p~ i
(~r , t ) = A u e ~ = Au e ~ , (410)
with
cosh 2 sinh 2

0 0

0 cosh +c3 sinh 2 +c4 0 .
u = Su = c1

+c2 2
0 sinh cosh 0
2 2
sinh 2 0 0 cosh 2
(411)
74
14.2 Non-relativistic limit of the Dirac-equation
To discuss the non-relativistic limit of the Dirac equation we start from the
free Dirac-equation and couple the electro-magnetic field via minimal cou-
pling.
h k e i
i~ = c pk Ak + mc2 + e . (412)
t c
Next we decompose the four-component Dirac-spinor into two two-component
spinors and .

= (413)

Inserting this into Eq.(412) we get

~ ~ 2
i~ =c + e + mc , (414)
t ~ ~
with the operator of the canonical momentum

e~
~ = p~ A . (415)
c
In order to investigate the non-relativistic limit, we first split of the dominant
energy dependence, which is given by mc2 . We define

imc
2
t
=: e ~ . (416)

In the non-relativistic limit the remaining time-dependence of and will

be much weaker than the explicit exponential term. Then we get

~ ~ 2 0
i~ =c + e 2mc . (417)
t ~
~

Next we make in the second line of the equation above the following assump-
tion
i~t , e 2mc2 (418)
to get
~ ~ v
= =O . (419)
2mc c
Thus is denoted the large component of the Dirac-spinor and as the small
component. Next we insert Eq.(419) into the first line of Eq.(417) and we
get
1
i~ = (~ ~ )(~ ~ ) + e . (420)
t 2m
75
Using the identity
(~ ~a)(~ ~b) = ~a ~b + i~ (~a ~b) , (421)
one gets
(~ ~ )(~ ~ ) = 2 + i~ (~ ~ ) (422)
e~ ~ .
= 2 ~ B (423)
c
So we get for the large component

1 e ~ 2 e~ ~
i~ = ~p A ~ B + e . (424)
t 2m c 2mc
This is the Pauli-equation that describes non-relativistic electrons with
spin 1/2. Moreover the Dirac equation gives automatically the correct Lande
factor g = 2 of the electron, which had to be introduced by hand to the Pauli-
equation in order to be compatible with experiment. We can introduce the
Lande factor in the above equation via ~ g~ /2.
~
Looking at the special case of a constant B-field
~ = 1B
A ~ ~x , (425)
2
~ =
B ~ A ~, (426)
we arrive at
2
e2 ~ 2

p~ e ~ ~ ~
i~ = (L + 2S) B + A + e , (427)
t 2m 2mc 2mc2
with S~ = ~~ /2 and L
~ = ~x p~. The factor 2 in front of S
~ in Eq.(427) repre-
sents the Lande factor g.
More precise measurements have shown, however, that the Lande-factor de-
viates slightly from 2, thus the anomalous magnetic moment of the electron
is introduced as
g2
a= . (428)
2
This quantity is measured very precisely by now. Theoretical deviations from
the prediction of the Dirac equation (a = 0) arise as soon as one is considering
the effects of Quantum electro dynamics (QED). The most precise values are
currently [24, 25]
aExp. = 0.00115965218073(28) , (429)

aT heo. = 0.00115965218178(77) . (430)
76
This is an amazing agreement! This is the most precise tested prediction in
science.
To reach this, accuracy 12672 5-loop diagrams have to be calculated [25] in
QED.
I(a) I(b) I(c) I(d) I(e)
I(f) I(g) I(h) I(i) I(j)
II(a) II(b) II(c) II(d) II(e)
II(f) III(a) III(b) III(c) IV
V VI(a) VI(b) VI(c) VI(d) VI(e)
VI(f) VI(g) VI(h) VI(i) VI(j) VI(k)
Nowadays the above comparison is used to fix the numerical value of the
electro-magnetic coupling, e = e2 /(4).
Short excursion: Quantum electro dynamics (QED)

Classical mechanics can be described by Newtons law
2
d
F~ = m 2 ~x . (431)
dt
Some problems can, however, be solved much easier if one starts from
the Lagrange function L, defined as the difference between kinetic and
potential energy
L= T V . (432)
77
Newtons law can be derived from the Lagrange function via the Euler-
Lagrange equations.
Similarly free relativistic electrons can be described by the Dirac equa-

tion
i/ m = 0 , (433)
while it is many times advantageous to use instead a Lagrangian density
given by
L = i/ m . (434)
Here again the Dirac equation can be derived from the Lagrange density
via generalisations of the Euler-Lagrange equations.
Finally the free Maxwell-equations
F = 0 (435)
are equivalent to the following Lagrange density

1
L = F F . (436)
4
Combining free electrons with free photons one gets thus

1
L = i/ m F F .

(437)
4
Interaction between electrons and photons can again be switched on

via minimal coupling, this is equivalent to the following replacement
D = ieA (438)
to obtain the complete Lagrangian of QED

1
/ m F F .

L = iD (439)
4
Remarks:
All terms in the QED Lagrangian are Lorentz-scalars!

/
The interaction term looks like eA.
78
15 Solutions of the Dirac equation for a cen-
tral potential
For the case of a central potential the Dirac-equation reads

(~r, t) k
= c pk + mc2 + e (~r, t) .

i~ (440)
t
E
With e = V (r) and the ansatz (~r, t) = ei ~ t (~r) this becomes
E(~r) = ck pk + mc2 + V (r) (~r) .

(441)
Next we would like to separate the variables as in the case of the Schrodinger-
equation. There we had L ~ 2 and Lz as conserved quantum numbers and we
found solutions that are simultaneous eigenvectors of H, L ~ 2 and Lz . Since
the Dirac-equation always contains spin, we expect now that not the angular
momentum L ~ is conserved, but the total angular momentum J~ = L ~ + S.
~ An
explicit calculation gives
~ = i~c(~
[H, L] ~p) 6= 0 , (442)
proofing that the angular momentum is not conserved.

The spin operator is given by
~ = ~
S ~ , (443)
2
with
k 0
k = . (444)
0 k
Remarks: this definition gives
The angular momentum commutation relation is fulfilled
[Si , Sj ] = i~ijk Sk . (445)
~ 2 read
The eigenvalues of S
~ 2 = 3 ~2 = s(s + 1)~2 ,
S (446)
4
with s = 1/2.
79
The commutator of the spin operator with the Hamiltonian reads
~ = +i~c(~
[H, S] p~) 6= 0 . (447)
Comparing Eq.(442) with Eq.(447) we get the conservation of the total an-
gular momentum
~ =0.
[H, J] (448)
One can also show that
[H, J~2 ] = 0 . (449)
Thus we can find wave functions that are simultaneous eigenfunctions of H,
J~2 and Jz , with the eigenvalues E, j(j + 1)~2 and mj ~.
We make the following ansatz for the Dirac-equation with a spherical poten-
tial !
iGlj (r) l
(, )
ljmj (r, , ) = r
iFlj (r) ~
x
jml
j
(450)
r
~
x
jmj
(, )
with the 2-component spinors
( (+) 1
jmj (, ) for j = l +
ljmj (, ) = ()
2
1
(451)
jmj (, ) for j = l 2
and q
l+mj + 12
Y 1 (, )
(+)
jmj (, ) = q 2l+1 1 l,mj 2 , (452)
lmj + 2
2l+1
Yl,mj + 1 (, )
2
q
lmj + 12
Y 1 (, )
()
jmj (, ) = q 2l+1 1 l,mj 2 . (453)
l+mj + 2
2l+1
Yl,mj + 1 (, )
2
Y are the spherical harmonics and Glj (r) and Flj (r) are some unknown func-
tions that describe the radial dependence. They obey the following differen-
tial equation

E m V (r) dFlj (r) 1 Flj (r)
Glj (r) = j+ (454)
~c dr 2 r

E + m V (r) dGlj (r) 1 Glj (r)
Flj (r) = + j+ (455)
~c dr 2 r
for j = l 1/2. The radial equation are solved by the ansatz
F (r) = f (r)er =: r s [a0 + a1 r + . . . ] er , (456)

G(r) = g(r)er =: r s [b0 + b1 r + . . . ] er . (457)
80
In the case a hydrogen-like problem we have V (r) = Z/r, with =
e2 /(40 ~c) = 1/137 and we get then the following energy eigenvalues
mc2
En,j = s 2 , (458)
1+ qZ
2
n(j+ 21 )+ (j+ 21 ) (Z)2
with = e2 /(40 ~c). Expanding this in (Z) we get
Z 2 2 (Z)4

2 1 3 6 6

En,j = mc 1 +O Z . (459)
2n2 2n3 j + 12 4n
The first term is just the rest mass, the second term is the result of the
Schrodinger theory for the hydrogen atom and the third term is the first
relativistic correction, that gives a fine structure splitting.
More details of this calculation can be found in [5].
To prepare for next lecture: Chapter [17.1-4] of [1]
81
16 Measurements and interpretation
Chapter [17.1-4] of [1]
16.1 Introduction
Quantum mechanics is one of the most successful concepts of physics. It is
the foundation of many technical applications like Computer, Laser,... The
theory is thoroughly tested by experiments and all the predictions agree
perfectly with the measurements. In that sense there is no doubt that
quantum mechanics is the correct language to describe microscopic
effects, at least in the energy range, that is currently accessible to us.
The inherent statistical nature of quantum mechanical predictions as well
as the quantum mechanical concept of measurements has, however, always
raised fundamental questions about the interpretation of the theory. To
some extend this is a philosophical question, depending also strongly on the
definition of the meaning of the word understanding. Nevertheless, we
would like to conclude this lecture course with a brief overview of different
interpretations of quantum mechanics.
Some reviews of this topic can also be found e.g. in [26, 27, 28].
16.2 The basic question

According to what we learnt in the course, the predictions of quantum me-
chanics have an inherent probabilistic nature. One arising question is now,
if this probabilistic nature is a real feature of nature or just a failure of our
current mathematical formulation. A related question is: what is the nature
of the measurement process in quantum mechanics?
Let us start by recapitulating the basics of quantum mechanics:
1. Observables are described by hermitian operators.
An observable A is related to a hermitian operator A.

There is a complete set of eigenstates |ni with
A|ni = an |ni . (460)
The real numbers an are the possible outcomes of a measurement

of the observable A.
82
2. States, particles,... are described by wave functions (states vectors).
A general wave function can be written as a linear combination of
the |ni: X
= cn |ni . (461)
n
3. The time evolution of the system is given by the Schrodinger-equation,

the Dirac-equation, the Klein-Gordon-equation...
4. The probability of finding in the state |ni is given by |cn |2 . The

average value of the observable A is given by
X
hAi = h|A|i = |cn |2 an . (462)
n
5. If we measure the observable A in a system described by and we get

the result an then the wave function collapses to the state |ni.
Here two concepts arise, that are not obvious to understand: the inherent
probabilistic nature of quantum mechanics (point 4) and the nature of the
measurements (point 5), or in other words: what is actually the collapse of
the wave function?
To become more concrete, let us assume that our concern is to measure the
position of a particle.
Quantum mechanics gives us only the probability for finding the parti-
cle at a certain location. The probability of finding the particle at time
t0 at the location ~x0 is given by |(~x = ~x0 , t = t0 )|2 .
To get a definite answer, we have to measure the position. Let us

assume now, that the measurement of the position at time t0 gave the
result ~x0 . This means that the original wave function collapsed into a
position eigenstate with the eigenvalue ~x0 .
Now we know for sure that the particle was at the time t0 at the position
~x0 and we can ask ourselves next: where was the particle slightly before
the measurement?6 This question is not governed by our general rules of
quantum mechanics, some possible answers are [7]:
1. at point ~x0 (realistic position);

6
The question, where the particle is slightly afterwards, is trivial. Because of the
collapse it is still at ~x0 . With time the collapsed wave function will, however, smear out
again.
83
2. nowhere (orthodox position);
3. there is no answer to this question (agnostic position),
Remarks:
1. If the particle was actually at the position ~x0 and we just cannot de-
scribe this with quantum mechanics, then this is a kind a failure of
the current formulation. In that case there will be a more fundamental
theory, that contains all the results from quantum mechanics and in
addition knows, where the particle was before the measurement.
This was e.g. the position of Albert Einstein.
Since such a new theory is expected to contain more variables, they are
also called hidden variable theories. They were e.g. investigated by
Bohm [29, 30].
How does the fundamental theory look like?
2. This is somehow the basis of the famous and widespread Copenhagen

interpretation, that was started by Max Born and worked out e.g. by
Niels Bohr. The particle simply did not have the property of having a
position before the actual measurement. The measurement itself forced
the particle to go into a certain position, by the collapse of the wave
function
What is actually the measurement process? What is the collapse of the
wave function?
3. This is a practical attitude. We can not answer this hypothetical ques-

tion, because the only way to answer it, would be to perform a mea-
surement at time t0 t and then we still do not know what happened
before that measurement. We have no access to information without
doing a measurement.
Dont worry, be happy.
16.3 Einstein-Podolsky-Rosen paradox

The aim of the paper of Einstein, Podolsky and Rosen [31] from 1935, was
to prove that the realistic position is the correct one.
To see their argument, consider the decay of a neutral pion (spin = 0) at
time t0 into an electron and a positron (each spin 1/2) 7
0 e+ + e . (463)
7
This example actually stems from [32].
84
If the pion is at rest then the electron-positron pair is described by the
following wave function (spin = 0)
1
(| i | i) . (464)
2
Next we assume that in some distance xe of the former position of the pion,
the spin of the electron will be measured at the time te > t0 .
Let us assume that the outcome is spin-up. Thus it is clear that any future
measurement of the spin of the positron will give spin-down.
How can this be interpreted?
1. According to the realistic interpretation, the spin of the electron, was

always (starting from t0 on) up and the spin of the positron was always
down. This fact was just unknown to us, because we do not know the
full theory.
2. According to the orthodox interpretation we have:
t0 < t < te The electron-positron state is completely described by the super-

position given in Eq.(464).
t = te The wave function collapses.
t > te The electron and positron are in pure spin states.
Remarks:
Somehow it seems that the information about the collapse of the wave
function is transported at the time te instantly from the location xe to
the position of the positron. Einstein and friends considered this to be
obviously impossible, because this would violate locality (spukhafte
Fernwirkung ghost-like long-distance action.
Thus people were seriously looking for extensions of quantum mechan-

ics that include hidden variables, see e.g. the models of Bohm [29, 30]
from 1951.
A very fundamental and crucial point here is the entangled stated

given in Eq.(464). Schrodinger coined the name verschrankte Zustande
= entangled states.
But actually there is no real information transported, so the issue is

still not settled.
85
16.4 Bells equation
In 1964 Bell [33] was generalising the above experiment and has shown that
there is a measurable difference between position 1 and 2 (this excludes by
definition position 3). In addition he pointed out severe problems of theories
with hidden variables, giving thus evidence for the orthodox position, albeit
locality has to be given up to some extent and we are still left with a lack of
understanding of the measurement process.
We take the same experimental setup as in the EPR case, with the gen-
eralisation that now the spin-detectors can be moved in arbitrary directions.
In the EPR case both were measuring the z-component of the spin, while
the electron-spin component is measured relative to the vector ~a and the
positron-spin component is measured relative to the vector ~b. In the end the
relevant observable will be the quantum mechanical average of the product
of the two spins P (~a, ~b). The case ~a = ~b gives the result of the EPR-case. In
the EPR case quantum mechanics told us that the product will always give
1 1. Thus we have
P (~a, ~a) = 1 . (465)
If the vectors are anti-parallel ~a = ~b, we will obviously get
P (~a, ~a) = +1 . (466)
For a general direction of the vectors quantum mechanics will give
P (~a, ~b) = ~a ~b . (467)
It will turn out, however, that this simple result is incompatible with theo-
ries with hidden variables. To show that, we consider very general models
with hidden variables, assuming only that reality is given by the quantum
mechanical wave function and some hidden variable .
A(~a, ) gives the result of the measurement of the electron spin relative
to the vector ~a. It can take the values 1.
B(~b, ) gives the result of the measurement of the positron spin relative
to the vector ~b. It can take the values 1.
If ~a = ~b, we have
A(~a, ) = B(~a, ) . (468)
The average of the product of the two measurements is given by
Z
P (~a, b) = ()A(~a, )B(~b, )d ,
~ (469)
86
with () being the unknown probability distribution of the hidden variable.
We only demand
0 () 1 , (470)
Z
()d = 1 . (471)
Because of B(~b, ) = A(~b, ) we get

Z
P (~a, ~b) = ()A(~a, )A(~b, )d . (472)
Next we take an additional general unit vector ~c to get

Z h i
~ ~
P (~a, b) P (~a, ~c) = () A(~a, )A(b, ) A(~a, )A(~c, ) d (473)
Z h i
= () |A(~b, )|2 A(~a, )A(~b, ) |A(~b, )|2 A(~a, )A(~c, ) d
(474)
Z h i
= () 1 A(~b, )A(~c, ) A(~a, )A(~b, )d . (475)
Here we used |A(~b, )|2 = 1. Next we keep in mind

h i
() 1 A(~b, )A(~c, ) 0 , (476)
1 A(~a, )A(~b, ) +1 , (477)
to get
Z h i
|P (~a, ~b) P (~a, ~c)| () 1 A(~b, )A(~c, ) d (478)
= 1 + P (~b, ~c) . (479)
This is the famous unequality of Bell.

Combing this with the quantum mechanical result for P (~a, ~b) we get
|~a ~b ~a ~c| 1 ~b ~c . (480)
This relation can easily be violated, assume e.g. ~a, ~b and ~c lie in a plane, ~a
and ~b are perpendicular and ~c lies in between them (with an angle of 45 ).
Then we get

|0 1/ 2| = 0.707 1 1/ 2 = 0.29 , (481)
87
which is obviously not correct.
Combing this result - hidden variable are not compatible with quantum me-
chanics8 - with the result from EPR have two possible resolutions:
1. Quantum mechanics is wrong.
2. Locality does not hold in general.
All experimental tests have shown that Quantum mechanics is correct and
that Bells inequality is violated. It seems that locality does not hold and
that there is an instantaneous collapse of the wave function.
We are still left with lack of understanding of the collapse of the wave func-
tion. Scientist tried to establish the collapse of the wave-function experimen-
tally; this comes under the name of quantum Zeno effect, but there are
currently no convincing hints.
16.5 More fancy interpretations

Many world theories from Hugh Everett in 1957 [34]. Everett was a PhD
student of John Archibald Wheeler. The idea was later on worked out by
Bruce de Witt and more recently by David Deutsch and Wojciech H. Zurek.
This theory tries to explain the collapse of the wave function.
Wikipedia states Many-worlds implies that all possible alternative histories
and futures are real, each representing an actual world (or universe). In
lay terms, the hypothesis states there is a very large, perhaps infinite number
of universes, and everything that could possibly have happened in our past,
but did not, has occurred in the past of some other universe or universes.
To repeat this lecture: Chapter [17.1-4] of [1]
To prepare for next lecture: All of [1]
8
One can, however, construct some fancy hidden variable models for which Bells in-
equality is not applicable.
88
17 Revision Lecture
The course covered four topics
1. Scattering
2. Quantum statistical mechanics
3. Relativistic Quantum mechanics
4. Interpretation of Quantum mechanics
17.1 Scattering
17.2 Quantum statistical mechanics
17.3 Relativistic Quantum mechanics
17.4 Interpretation of Quantum mechanics
89
90
References
- Reference text -
[1] Quantum Mechanics
Brian Harold Bransden, C. J. Joachain
Prentice Hall (28th January 2000)
ISBN-10: 0582356911
ISBN-13: 978-0582356917
- Recommended further reading -
[2] Modern Quantum Mechanics

J.J.Sakurai, Jim Napolitano
Prentice Hall (12th October 2007)
ISBN-10: 0321503368
ISBN-13: 978-0321503367
[3] Lectures on Quantum Mechanics

S. Weinberg
Cambridge University Press (22nd November 2012)
ISBN-10: 1107028728
ISBN-13: 978-1107028722
[4] Quantum Mechanics

F. Schwabl
Springer: (19th September 2007)
ISBN-10: 3540719326
ISBN-13: 978-3540719328
[5] Advanced Quantum Mechanics

F. Schwabl
Springer (12th August 2008) ISBN-10: 1107028728
ISBN-13: 978-1107028722
[6] Relativistic Quantum Mechanics

J. Bjorken and S. Drell
Mcgraw Hill Book Co (5th December 2008)
ISBN-10: 0072320028
ISBN-13: 978-0072320022
[7] Introduction to Quantum Mechanics

David J. Griffiths
Prentice Hall; (1st May 2003)
91
ISBN-10: 0131911759
ISBN-13: 978-0131911758

[8] http://www.nobelprize.org/nobel prizes/physics/
[9] F. Englert and R. Brout

Broken Symmetry and the Mass of Gauge Vector Mesons
Phys. Rev. Lett. 13 (1964) 321
2775 citations counted in INSPIRE as of 20 Jan 2015
[10] P. W. Higgs
Broken symmetries, massless particles and gauge fields
Phys. Lett. 12 (1964) 132
[11] P. W. Higgs
Broken Symmetries and the Masses of Gauge Bosons
Phys. Rev. Lett. 13 (1964) 508
[12] G. S. Guralnik, C. R. Hagen and T. W. B. Kibble

Global Conservation Laws and Massless Particles
Phys. Rev. Lett. 13 (1964) 585
[13] P. W. Higgs
Spontaneous Symmetry Breakdown without Massless Bosons
Phys. Rev. 145 (1966) 1156
[14] T. W. B. Kibble
Symmetry breaking in non abelian gauge theories
Phys. Rev. 155 (1967) 1554
[15] G. Aad et al. [ATLAS Collaboration]

Observation of a new particle in the search for the Standard Model Higgs
boson with the ATLAS detector at the LHC
Phys. Lett. B 716 (2012) 1 [arXiv:1207.7214 [hep-ex]]
92
[16] S. Chatrchyan et al. [CMS Collaboration]
Observation of a new boson at a mass of 125 GeV with the CMS exper-
iment at the LHC
Phys. Lett. B 716 (2012) 30 [arXiv:1207.7235 [hep-ex]]
[17] T. Csorgo et al. [TOTEM Collaboration],

Elastic Scattering and Total Cross-Section
in p + p reactions measured
by the LHC Experiment TOTEM at s = 7 TeV,
Prog. Theor. Phys. Suppl. 193 (2012) 180 [arXiv:1204.5689 [hep-ex]].
[18] C. Anastasiou, C. Duhr, F. Dulat, F. Herzog and B. Mistlberger,

Real-virtual contributions to the inclusive Higgs cross-section at N 3 LO,
JHEP 1312 (2013) 088 [arXiv:1311.1425 [hep-ph]].
[19] https : //twiki.cern.ch/twiki/bin/view/

LHCP hysics/CrossSections#Higgs cross sections at 7 8 and
[20] V. M. Braun, A. Lenz and M. Wittmann,

Nucleon Form Factors in QCD,
Phys. Rev. D 73 (2006) 094019 [hep-ph/0604050].
[21] P. A. M. Dirac,
The Quantum theory of electron,
Proc. Roy. Soc. Lond. A 117 (1928) 610.
[22] New York Times

http://www.nytimes.com/2010/05/18/science/space/18cosmos.html? r=0
based on the masurement in
V. M. Abazov et al. [D0 Collaboration],

Evidence for an anomalous like-sign dimuon charge asymmetry,
Phys. Rev. D 82 (2010) 032001 [arXiv:1005.2757 [hep-ex]].
which deviates significantly from the standard model prediction
A. Lenz and U. Nierste,
93
Theoretical update of Bs Bs mixing,
JHEP 0706 (2007) 072 [hep-ph/0612167].
[23] http://angelsanddemons.web.cern.ch/antimatter
[24] D. Hanneke, S. F. Hoogerheide and G. Gabrielse,

Cavity Control of a Single-Electron Quantum Cyclotron: Measuring the
Electron Magnetic Moment,
arXiv:1009.4831 [physics.atom-ph].
[25] T. Aoyama, M. Hayakawa, T. Kinoshita and M. Nio,

Tenth-Order QED Contribution to the Electron g-2 and an Improved
Value of the Fine Structure Constant,
Phys. Rev. Lett. 109 (2012) 111807 [arXiv:1205.5368 [hep-ph]].
[26] S. Weinberg,
Testing Quantum Mechanics,
Annals Phys. 194 (1989) 336.
[27] R. Omnes,
Consistent interpretations of quantum mechanics,
Rev. Mod. Phys. 64 (1992) 339.
[28] http://en.wikipedia.org/wiki/Interpretations of quantum mechanics
[29] D. Bohm,
A Suggested interpretation of the quantum theory in terms of hidden
variables. 1.,
Phys. Rev. 85 (1952) 166.
[30] D. Bohm,
A Suggested interpretation of the quantum theory in terms of hidden
variables. 2.,
Phys. Rev. 85 (1952) 180.
[31] A. Einstein, B. Podolsky and N. Rosen,

Can quantum mechanical description of physical reality be considered
94
complete?,
Phys. Rev. 47 (1935) 777.
[32] D. Bohm and Y. Aharonov,

Discussion of Experimental Proof for the Paradox of Einstein, Rosen,
and Podolsky,
Phys. Rev. 108 (1957) 1070.
[33] J. S. Bell,
On the Einstein-Podolsky-Rosen paradox,
Physics 1 (1964) 195.
[34] H. Everett,
Relative state formulation of quantum mechanics,
Rev. Mod. Phys. 29 (1957) 454.
95

TP3-Notes Lenz

Uploaded by

Copyright:

Available Formats

TP3-Notes Lenz

Uploaded by

Document Information

Original Description:

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

TP3-Notes Lenz

Uploaded by

Copyright:

Available Formats

Theoretical Physics 3

2 Potential scattering (general features) 15

3 Spherical Bessel functions 18

4 The method of partial waves 25

5 The integral equation of potential scattering 31

6 The Born approximation 34

8 The density matrix 41

9 Quantum mechanical ensembles 47

10 Systems of non-interacting particles 54

11 The Klein-Gordon equation 58

12 The Dirac equation 63

13 Lorentz-invariance of the Dirac equation 69

15 Solutions of the Dirac equation for a central potential 79

16 Measurements and interpretation 82

18 lectures (2 hours per week) + 3 example classes: (1 hour every 3

The lecture course follows the textbook:

1.1 Introduction and Motivation

Quarks and leptons have spin 1/2.

Fundamental forces are carried by spin 1 (2) gauge bosons

Electromagnetic interaction: photon.

For classical scattering of a point-like particle on a hard sphere with radius

which coincides with our expectations above. The concept of (differential)

= (98.3 0.2(stat) 2.8(syst)) mb, (7)

with the unit barn for measuring the cross section:

1barn = 1b = 1028 m2 = 100f m2 . (8)

N(H ) = 4 105 2 103 = 800 . (14)

To repeat this lecture: Chapter [13.1] of [1]

To prepare for next lecture: Chapter [13.2] of [1]

2.1 The time-independent Schrodinger equation

The wave function (~r) has to satisfy the time-independent Schrodinger-

the incident momentum of the particle ~p = ~~k = m~v ,

the incident wave vector of the particle ~k,

the incident velocity of the particle ~v .

with the redefined potential

2.2 The scattering amplitude

To repeat this lecture: Chapter [13.2] of [1]

To prepare for next lecture: Chapter [7.3/4] of [1]

3.1 Separation of variables

V (~r) = V (r = |~r|) . (38)

To proceed we rewrite the Laplace operator in spherical coordinates

Ylm (, ) = Clm eim Plm (cos ) , (43)

3.2 The simplest Bessel function

uE0(r) = A sin(kr) + B cos(kr) . (58)

Thus we finally get

3.3 The spherical Bessel equation

where we denoted Rl () := REl (/k). Eq.(60) is called the spherical Bessel

3.3.2 Deriving the solution:

Trick 1: For l 6= 0 we make the Ansatz Rl () =: l l () and Eq.(60)

Assume that l () is a solution of Eq.(62).

To answer that we differentiate Eq.(62) once more

The factors (1)l are just convention.

The first few functions read

Spherical Bessel functions j0 (x), j1 (x) and j2 (x)

Spherical Neumann functions n0 (x), n1 (x) and n2 (x)

For small values of these functions can be expanded as

3.3.4 Bessel functions J

3.3.6 Orthogonality relations

3.4 Summary of Bessel functions

This means that spherical harmonics in combination with spherical Bessel