This action might not be possible to undo. Are you sure you want to continue?
Professor John W. Norbury
Physics Department
University of WisconsinMilwaukee
P.O. Box 413
Milwaukee, WI 53201
November 20, 2000
2
Contents
1 Lagrangian Field Theory 7
1.1 Units . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
1.1.1 Natural Units . . . . . . . . . . . . . . . . . . . . . . . 7
1.1.2 Geometrical Units . . . . . . . . . . . . . . . . . . . . 10
1.2 Covariant and Contravariant vectors . . . . . . . . . . . . . . 11
1.3 Classical point particle mechanics . . . . . . . . . . . . . . . . 12
1.3.1 EulerLagrange equation . . . . . . . . . . . . . . . . . 12
1.3.2 Hamilton’s equations . . . . . . . . . . . . . . . . . . . 14
1.4 Classical Field Theory . . . . . . . . . . . . . . . . . . . . . . 15
1.5 Noether’s Theorem . . . . . . . . . . . . . . . . . . . . . . . . 18
1.6 Spacetime Symmetries . . . . . . . . . . . . . . . . . . . . . . 24
1.6.1 Invariance under Translation . . . . . . . . . . . . . . 24
1.6.2 Angular Momentum and Lorentz Transformations . . 25
1.7 Internal Symmetries . . . . . . . . . . . . . . . . . . . . . . . 26
1.8 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
1.8.1 Covariant and contravariant vectors . . . . . . . . . . 29
1.8.2 Classical point particle mechanics . . . . . . . . . . . . 29
1.8.3 Classical ﬁeld theory . . . . . . . . . . . . . . . . . . . 29
1.8.4 Noether’s theorem . . . . . . . . . . . . . . . . . . . . 30
1.9 References and Notes . . . . . . . . . . . . . . . . . . . . . . . 32
2 Symmetries & Group theory 33
2.1 Elements of Group Theory . . . . . . . . . . . . . . . . . . . . 33
2.2 SO(2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
2.2.1 Transformation Properties of Fields . . . . . . . . . . 34
2.3 Representations of SO(2) and U(1) . . . . . . . . . . . . . . . 35
2.4 Representations of SO(3) and SU(1) . . . . . . . . . . . . . . 35
2.5 Representations of SO(N) . . . . . . . . . . . . . . . . . . . . 36
3
4 CONTENTS
3 Free KleinGordon Field 37
3.1 KleinGordon Equation . . . . . . . . . . . . . . . . . . . . . 37
3.2 Probability and Current . . . . . . . . . . . . . . . . . . . . . 39
3.2.1 Schrodinger equation . . . . . . . . . . . . . . . . . . . 39
3.2.2 KleinGordon Equation . . . . . . . . . . . . . . . . . 40
3.3 Classical Field Theory . . . . . . . . . . . . . . . . . . . . . . 41
3.4 Fourier Expansion & Momentum Space . . . . . . . . . . . . 42
3.5 KleinGordon QFT . . . . . . . . . . . . . . . . . . . . . . . . 45
3.5.1 Indirect Derivation of a, a
†
Commutators . . . . . . . 45
3.5.2 Direct Derivation of a, a
†
Commutators . . . . . . . . 47
3.5.3 KleinGordon QFT Hamiltonian . . . . . . . . . . . . 47
3.5.4 Normal order . . . . . . . . . . . . . . . . . . . . . . . 48
3.5.5 Wave Function . . . . . . . . . . . . . . . . . . . . . . 50
3.6 Propagator Theory . . . . . . . . . . . . . . . . . . . . . . . . 51
3.7 Complex KleinGordon Field . . . . . . . . . . . . . . . . . . 66
3.7.1 Charge and Complex Scalar Field . . . . . . . . . . . 68
3.8 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
3.8.1 KG classical ﬁeld . . . . . . . . . . . . . . . . . . . . . 70
3.8.2 KleinGordon Quantum ﬁeld . . . . . . . . . . . . . . 71
3.8.3 Propagator Theory . . . . . . . . . . . . . . . . . . . . 72
3.8.4 Complex KG ﬁeld . . . . . . . . . . . . . . . . . . . . 73
3.9 References and Notes . . . . . . . . . . . . . . . . . . . . . . . 73
4 Dirac Field 75
4.1 Probability & Current . . . . . . . . . . . . . . . . . . . . . . 77
4.2 Bilinear Covariants . . . . . . . . . . . . . . . . . . . . . . . . 78
4.3 Negative Energy and Antiparticles . . . . . . . . . . . . . . . 79
4.3.1 Schrodinger Equation . . . . . . . . . . . . . . . . . . 79
4.3.2 KleinGordon Equation . . . . . . . . . . . . . . . . . 80
4.3.3 Dirac Equation . . . . . . . . . . . . . . . . . . . . . . 82
4.4 Free Particle Solutions of Dirac Equation . . . . . . . . . . . 83
4.5 Classical Dirac Field . . . . . . . . . . . . . . . . . . . . . . . 87
4.5.1 Noether spacetime current . . . . . . . . . . . . . . . . 87
4.5.2 Noether internal symmetry and charge . . . . . . . . . 87
4.5.3 Fourier expansion and momentum space . . . . . . . . 87
4.6 Dirac QFT . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
4.6.1 Derivation of b, b
†
, d, d
†
Anticommutators . . . . . . . 88
4.7 Pauli Exclusion Principle . . . . . . . . . . . . . . . . . . . . 88
4.8 Hamiltonian, Momentum and Charge in terms of creation and
annihilation operators . . . . . . . . . . . . . . . . . . . . . . 88
CONTENTS 5
4.8.1 Hamiltonian . . . . . . . . . . . . . . . . . . . . . . . . 88
4.8.2 Momentum . . . . . . . . . . . . . . . . . . . . . . . . 88
4.8.3 Angular Momentum . . . . . . . . . . . . . . . . . . . 88
4.8.4 Charge . . . . . . . . . . . . . . . . . . . . . . . . . . 88
4.9 Propagator theory . . . . . . . . . . . . . . . . . . . . . . . . 88
4.10 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
4.10.1 Dirac equation summary . . . . . . . . . . . . . . . . . 88
4.10.2 Classical Dirac ﬁeld . . . . . . . . . . . . . . . . . . . 88
4.10.3 Dirac QFT . . . . . . . . . . . . . . . . . . . . . . . . 88
4.10.4 Propagator theory . . . . . . . . . . . . . . . . . . . . 88
4.11 References and Notes . . . . . . . . . . . . . . . . . . . . . . . 88
5 Electromagnetic Field 89
5.1 Review of Classical Electrodynamics . . . . . . . . . . . . . . 89
5.1.1 Maxwell equations in tensor notation . . . . . . . . . . 89
5.1.2 Gauge theory . . . . . . . . . . . . . . . . . . . . . . . 89
5.1.3 Coulomb Gauge . . . . . . . . . . . . . . . . . . . . . 89
5.1.4 Lagrangian for EM ﬁeld . . . . . . . . . . . . . . . . . 89
5.1.5 Polarization vectors . . . . . . . . . . . . . . . . . . . 89
5.1.6 Linear polarization vectors in Coulomb gauge . . . . . 91
5.1.7 Circular polarization vectors . . . . . . . . . . . . . . 91
5.1.8 Fourier expansion . . . . . . . . . . . . . . . . . . . . . 91
5.2 Quantized Maxwell ﬁeld . . . . . . . . . . . . . . . . . . . . . 91
5.2.1 Creation & annihilation operators . . . . . . . . . . . 91
5.3 Photon propagator . . . . . . . . . . . . . . . . . . . . . . . . 91
5.4 GuptaBleuler quantization . . . . . . . . . . . . . . . . . . . 91
5.5 Proca ﬁeld . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
6 Smatrix, cross section & Wick’s theorem 93
6.1 Schrodinger Time Evolution Operator . . . . . . . . . . . . . 93
6.1.1 Time Ordered Product . . . . . . . . . . . . . . . . . . 95
6.2 Schrodinger, Heisenberg and Dirac (Interaction) Pictures . . 96
6.2.1 Heisenberg Equation . . . . . . . . . . . . . . . . . . . 97
6.2.2 Interaction Picture . . . . . . . . . . . . . . . . . . . . 97
6.3 Cross section and Smatrix . . . . . . . . . . . . . . . . . . . 99
6.4 Wick’s theorem . . . . . . . . . . . . . . . . . . . . . . . . . . 101
6.4.1 Contraction . . . . . . . . . . . . . . . . . . . . . . . . 101
6.4.2 Statement of Wick’s theorem . . . . . . . . . . . . . . 101
6 CONTENTS
7 QED 103
7.1 QED Lagrangian . . . . . . . . . . . . . . . . . . . . . . . . . 103
7.2 QED Smatrix . . . . . . . . . . . . . . . . . . . . . . . . . . 103
7.2.1 First order Smatrix . . . . . . . . . . . . . . . . . . . 103
7.2.2 Second order Smatrix . . . . . . . . . . . . . . . . . . 104
7.2.3 First order Smatrix elements . . . . . . . . . . . . . . 106
7.2.4 Second order Smatrix elements . . . . . . . . . . . . . 107
7.2.5 Invariant amplitude and lepton tensor . . . . . . . . . 107
7.3 Casimir’s trick & Trace theorems . . . . . . . . . . . . . . . . 107
7.3.1 Average over initial states / Sum over ﬁnal states . . . 107
7.3.2 Casimir’s trick . . . . . . . . . . . . . . . . . . . . . . 107
Chapter 1
Lagrangian Field Theory
1.1 Units
We start with the most basic thing of all, namely units and concentrate
on the units most widely used in particle physics and quantum ﬁeld the
ory (natural units). We also mention the units used in General Relativity,
because these days it is likely that students will study this subject as well.
Some useful quantities are [PPDB]:
¯h ≡
h
2π
= 1.055 10
−34
J sec = 6.582 10
−22
MeV sec
c = 3 10
8 m
sec
.
1 eV = 1.6 10
−19
J
¯hc = 197MeV fm
1fm = 10
−15
m
1barn = 10
−28
m
2
1mb = .1fm
2
1.1.1 Natural Units
In particle physics and quantum ﬁeld theory we are usually dealing with
particles that are moving fast and are very small, i.e. the particles are both
relativistic and quantum mechanical and therefore our formulas have lots
of factors of c (speed of light) and ¯ h (Planck’s constant). The formulas
considerably simplify if we choose a set of units, called natural units where
c and ¯ h are set equal to 1.
In CGS units (often also called Gaussian [Jackson appendix] units), the
basic quantities of length, mass and time are centimeters (cm), gram (g),
seconds (sec), or in MKS units these are meters (m), kilogram (kg), seconds.
In natural units the units of length, mass and time are all expressed in GeV.
7
8 CHAPTER 1. LAGRANGIAN FIELD THEORY
Example With c ≡ 1, show that sec = 3 10
10
cm.
Solution c = 3 10
10
cm sec
−1
. If c ≡ 1
⇒ sec = 3 10
10
cm
We can now derive the other conversion factors for natural units, in which ¯ h
is also set equal to unity. Once the units of length and time are established,
one can deduce the units of mass from E = mc
2
. These are
sec = 1.52 10
24
GeV
−1
m = 5.07 10
15
GeV
−1
kg = 5.61 10
26
GeV
(The exact values of c and ¯ h are listed in the [Particle Physics Booklet] as
c = 2.99792458 10
8
m/sec and ¯ h = 1.05457266 10
−34
Jsec= 6.5821220
10
−25
GeV sec. )
1.1. UNITS 9
Example Deduce the value of Newton’s gravitational constant
G in natural units.
Solution It is interesting to note that the value of G is one of the
least accurately known of the fundamental constants. Whereas,
say the mass of the electron is known as [Particle Physics Book
let] m
e
= 0.51099906MeV/c
2
or the ﬁne structure constant as
α = 1/137.0359895 and c and ¯ h are known to many decimal
places as mentioned above, the best known value of G is [PPDB]
G = 6.67259 10
−11
m
3
kg
−1
sec
−2
, which contains far fewer dec
imal places than the other fundamental constants.
Let’s now get to the problem. One simply substitutes the con
version factors from before, namely
G = 6.67 10
−11
m
3
kg
−1
sec
−2
=
6.67 10
−11
(5.07 10
15
GeV
−1
)
3
(5.61 10
26
GeV )(1.52 10
24
GeV
−1
)
2
= 6.7 10
−39
GeV
−2
=
1
M
2
Pl
where the Planck mass is deﬁned as M
Pl
≡ 1.22 10
19
GeV .
Natural units are also often used in cosmology and quantum gravity
[Guidry 514] with G given above as G =
1
M
2
Pl
.
10 CHAPTER 1. LAGRANGIAN FIELD THEORY
1.1.2 Geometrical Units
In classical General Relativity the constants c and G occur most often and
geometrical units are used with c and G set equal to unity. Recall that in
natural units everything was expressed in terms of GeV . In geometrical
units everything is expressed in terms of cm.
Example Evaluate G when c ≡ 1.
Solution
G = 6.67 10
−11
m
3
kg
−1
sec
−2
= 6.67 10
−8
cm
3
g
−1
sec
−2
and when c ≡ 1 we have sec = 3 10
10
cm giving
G = 6.67 10
−8
cm
3
g
−1
(3 10
10
cm)
−2
= 7.4 10
−29
cm g
−1
Now imposing G ≡ 1 gives the geometrical units
sec = 3 10
10
cm
g = 7.4 10
−29
cm
It is important to realize that geometrical and natural units are not com
patible. In natural units c = ¯ h = 1 and we deduce that G =
1
M
2
Pl
as
in a previous Example. In geometrical units c = G = 1 we deduce that
¯h = 2.6 10
−66
cm
2
. (see Problems) Note that in these units ¯ h = L
2
Pl
where
L
Pl
≡ 1.6 10
−33
cm. In particle physics, gravity becomes important when
energies (or masses) approach the Planck mass M
Pl
. In gravitation (General
Relativity), quantum eﬀects become important at length scales approaching
L
Pl
.
1.2. COVARIANT AND CONTRAVARIANT VECTORS 11
1.2 Covariant and Contravariant vectors
The subject of covariant and contravariant vectors is discussed in [Jackson],
which students should consult for a thorough introduction. In this section
we summarize the basic results.
The metric tensor that is used in this book is
η
µν
= η
µν
=
¸
¸
¸
¸
1 0 0 0
0 −1 0 0
0 0 −1 0
0 0 0 −1
¸
Contravariant vectors are written in 4dimensional form as
A
µ
= (A
o
, A
i
) = (A
o
,
A)
Covariant vectors are formed by “lowering” the indices with the metric ten
sor as in
A
µ
= g
µν
A
ν
= (A
o
, A
i
) = (A
o
, −A
i
) = (A
o
, −
A)
noting that
A
o
= A
o
Thus
A
µ
= (A
o
,
A) A
µ
= (A
o
, −
A)
Now we discuss derivative operators, denoted by the covariant symbol
∂
µ
and deﬁned via
∂
∂x
µ
≡ ∂
µ
= (∂
o
, ∂
i
) = (
∂
∂x
o
,
∂
∂x
i
) = (
∂
∂t
,
)
The contravariant operator ∂
µ
is given by
∂
∂x
µ
≡ ∂
µ
= (∂
o
, ∂
i
) = g
µν
∂
ν
= (
∂
∂x
o
,
∂
∂x
i
) = (
∂
∂x
o
, −
∂
∂x
i
) = (
∂
∂t
, −
)
Thus
∂
µ
= (
∂
∂t
,
) ∂
µ
= (
∂
∂t
, −
)
12 CHAPTER 1. LAGRANGIAN FIELD THEORY
The length squared of our 4vectors is
A
2
≡ A
µ
A
µ
= A
µ
A
µ
= A
2
o
−
A
2
and
∂
2
≡ ∂
µ
∂
µ
≡ P
2
=
∂
2
∂t
2
−
2
(1.1)
Finally, note that with our 4vector notation, the usual quantum mechanical
replacements
p
i
→i¯h∂
i
≡ −i¯h
and
p
o
→i¯h∂
o
= i¯h
∂
∂t
can be succintly written as
p
µ
→i¯h∂
µ
giving (with ¯ h = 1)
p
2
→−P
2
1.3 Classical point particle mechanics
1.3.1 EulerLagrange equation
Newton’s second law of motion is
F =
d p
dt
or in component form (for each component F
i
)
F
i
=
dp
i
dt
where p
i
= m˙ q
i
(with q
i
being the generalized position coordinate) so that
dp
i
dt
= ˙ m˙ q
i
+ m¨ q
i
. (Here and throughout this book we use the notation
˙ x ≡
dx
dt
.) If ˙ m = 0 then F
i
= m¨ q
i
= ma
i
. For conservative forces
F = −
U
where U is the scalar potential. Rewriting Newton’s law we have
−
dU
dq
i
=
d
dt
(m˙ q
i
)
1.3. CLASSICAL POINT PARTICLE MECHANICS 13
Let us deﬁne the Lagrangian L(q
i
, ˙ q
i
) ≡ T −U where T is the kinetic energy.
In freshman physics T = T( ˙ q
i
) =
1
2
m˙ q
2
i
and U = U(q
i
) such as the harmonic
oscillator U(q
i
) =
1
2
kq
2
i
. That is in freshman physics T is a function only
of velocity ˙ q
i
and U is a function only of position q
i
. Thus L(q
i
, ˙ q
i
) =
T( ˙ q
i
) − U(q
i
). It follows that
∂L
∂q
i
= −
dU
dq
i
and
∂L
∂ ˙ q
i
=
dT
d ˙ q
i
= m˙ q
i
= p
i
. Thus
Newton’s law is
F
i
=
dp
i
dt
∂L
∂q
i
=
d
dt
(
∂L
∂ ˙ q
i
)
with the canonical momentum deﬁned as
p
i
≡
∂L
∂ ˙ q
i
The next to previous equation is known as the EulerLagrange equation of
motion and serves as an alternative formulation of mechanics [Goldstein]. It
is usually written
d
dt
(
∂L
∂ ˙ q
i
) −
∂L
∂q
i
= 0
or just
˙ p
i
=
∂L
∂q
i
We have obtained the EulerLagrange equations using simple arguments.
A more rigorous derivation is based on the calculus of variations [HoKim47,
Huang54,Goldstein37, Bergstrom284] as follows.
In classical point particle mechanics the action is
S ≡
t
2
t
1
L(q
i
, ˙ q
i
, t)dt
where the Lagrangian L is a function of generalized coordinates q
i
, general
ized velocities ˙ q
i
and time t.
According to Hamilton’s principle, the action has a stationary value for
the correct path of the motion [Goldstein36], i.e. δS = 0 for the correct path.
To see the consequences of this, consider a variation of the path [Schwabl262,
BjRQF6]
q
i
(t) →q
i
(t) ≡ q
i
(t) +δq
i
(t)
14 CHAPTER 1. LAGRANGIAN FIELD THEORY
subject to the constraint δq
i
(t
1
) = δq
i
(t
2
) = 0. The subsequent variation in
the action is (assuming that L is not an explicit function of t)
δS =
t
2
t
1
(
∂L
∂q
i
δq
i
+
∂L
∂ ˙ q
i
δ ˙ q
i
)dt = 0
with δ ˙ q
i
=
d
dt
δq
i
and integrating the second term by parts yields
∂L
∂ ˙ q
i
δ ˙ q
i
dt =
∂L
∂ ˙ q
i
d(δq
i
) (1.2)
=
∂L
∂ ˙ q
i
δq
i
[
t
2
t
1
−
δq
i
d(
∂L
∂ ˙ q
i
)
= 0 −
δq
i
d
dt
(
∂L
∂ ˙ q
i
)dt
where the boundary term has vanished because δq
i
(t
1
) = δq
i
(t
2
) = 0. We
are left with
δS =
t
2
t
1
¸
∂L
∂q
i
−
d
dt
(
∂L
∂ ˙ q
i
)
δq
i
dt = 0
which is true for an arbitrary variation δq
i
indicating that the integral must
be zero, which yields the EulerLagrange equations.
1.3.2 Hamilton’s equations
We now introduce the Hamiltonian H deﬁned as a function of p and q as
H(p
i
, q
i
) ≡ p
i
˙ q
i
−L(q
i
, ˙ q
i
) (1.3)
For the simple case T =
1
2
m˙ q
2
i
and U = U( ˙ q
i
) we have p
i
∂L
∂ ˙ q
i
= m˙ q
i
so that
T =
p
2
i
2m
and p
i
˙ q
i
=
p
2
i
m
so that H(p
i
, q
i
) =
p
2
i
2m
+U(q
i
) = T +U which is the
total energy. Hamilton’s equations of motion immediately follow as
∂H
∂p
i
= ˙ q
i
Now L = L(p
i
) and
∂H
∂q
i
= −
∂L
∂q
i
so that our original deﬁnition of the canon
ical momentum above gives
−
∂H
∂q
i
= ˙ p
i
1.4. CLASSICAL FIELD THEORY 15
1.4 Classical Field Theory
Scalar ﬁelds are important in cosmology as they are thought to drive in
ﬂation. Such a ﬁeld is called an inﬂaton, an example of which may be the
Higgs boson. Thus the ﬁeld φ considered below can be thought of as an
inﬂaton, a Higgs boson or any other scalar boson.
In both special and general relativity we always seek covariant equations
in which space and time are given equal status. The EulerLagrange equa
tions above are clearly not covariant because special emphasis is placed on
time via the ˙ q
i
and
d
dt
(
∂L
∂ ˙ q
i
) terms.
Let us replace the q
i
by a ﬁeld φ ≡ φ(x) where x ≡ (t, x). The generalized
coordiante q has been replaced by the ﬁeld variable φ and the discrete index
i has been replaced by a continuously varying index x. In the next section
we shall show how to derive the EulerLagrange equations from the action
deﬁned as
S ≡
Ldt
which again is clearly not covariant. A covariant form of the action would
involve a Lagrangian density L via
S ≡
Ld
4
x =
Ld
3
xdt
where L = L(φ, ∂
µ
φ) and with L ≡
Ld
3
x. The term −
∂L
∂q
i
in the Euler
Lagrange equation gets replaced by the covariant term −
∂L
∂φ(x)
. Any time
derivative
d
dt
should be replaced with ∂
µ
≡
∂
∂x
µ
which contains space as well
as time derivatives. Thus one can guess that the covariant generalization of
the point particle EulerLagrange equation is
∂
µ
∂L
∂(∂
µ
φ)
−
∂L
∂φ
= 0
which is the covariant EulerLagrange equation for a ﬁeld φ. If there is more
than one scalar ﬁeld φ
i
then the EulerLagrange equations are
∂
µ
∂L
∂(∂
µ
φ
i
)
−
∂L
∂φ
i
= 0
To derive the EulerLagrange equations for a scalar ﬁeld [HoKim48, Gold
stein548], consider an arbitrary variation of the ﬁeld [Schwabl 263; Ryder
83; Mandl & Shaw 30,35,39; BjRQF13]
φ(x) →φ
(x) ≡ φ(x) +δφ(x)
16 CHAPTER 1. LAGRANGIAN FIELD THEORY
again with δφ = 0 at the end points. The variation of the action is (assuming
that L is not an explicit function of x)
δS =
X
2
X
1
¸
∂L
∂φ
δφ +
∂L
∂(∂
µ
φ)
δ(∂
µ
φ)
¸
d
4
x = 0
where X
1
and X
2
are the 4surfaces over which the integration is performed.
We need the result
δ(∂
µ
φ) = ∂
µ
δφ =
∂
∂x
µ
δφ
which comes about because δφ(x) = φ
(x) −φ(x) giving
∂
µ
δφ(x) = ∂
µ
φ
(x) −∂
µ
φ(x) = δ∂
µ
φ(x)
showing that δ commutes with diﬀerentiation ∂
µ
. Integration by parts on
the second term is a bit more complicated than before for the point particle
case, but the ﬁnal result is (see Problems)
δS =
X
2
X
1
¸
∂L
∂φ
−∂
µ
∂L
∂(∂
µ
φ)
¸
δφ d
4
x = 0
which holds for arbitrary δφ, implying that the integrand must be zero,
yielding the EulerLagrange equations.
In analogy with the canonical momentum in point particle mechanics,
we deﬁne the covariant momentum density
Π
µ
≡
∂L
∂(∂
µ
φ)
so that the EulerLagrange equations become
∂
µ
Π
µ
=
∂L
∂φ
The canonical momentum is deﬁned as
Π ≡ Π
0
=
∂L
∂
˙
φ
The energy momentum tensor is (analagous to the deﬁnition of the point
particle Hamiltonian)
T
µν
≡ Π
µ
∂
ν
φ −g
µν
L
1.4. CLASSICAL FIELD THEORY 17
with the Hamiltonian density
H ≡
Hd
3
x
H ≡ T
00
= Π
˙
φ −L
In order to illustrate the foregoing theory we shall use the example of the
classical, massive KleinGordon ﬁeld.
Example The massive KleinGordon Lagrangian density is
L
KG
=
1
2
(∂
µ
φ∂
µ
φ −m
2
φ
2
)
=
1
2
[
˙
φ
2
−(φ)
2
−m
2
φ
2
]
A) Derive expressions for the covariant momentum density and
the canonical momentum.
B) Derive the equation of motion in position space and momen
tum space.
C) Derive expressions for the energymomentum tensor and the
Hamiltonian density.
Solution A) The covariant momentum density is more easily
evaluated by rewriting L
KG
=
1
2
(g
µν
∂
µ
φ∂
ν
φ − m
2
φ
2
). Thus
Π
µ
=
∂L
∂(∂µφ)
=
1
2
g
µν
(δ
α
µ
∂
ν
φ + ∂
µ
φδ
α
ν
) =
1
2
(δ
α
µ
∂
µ
φ + ∂
ν
φδ
α
ν
) =
1
2
(∂
α
φ +∂
α
φ) = ∂
α
φ. Thus for the KleinGordon ﬁeld we have
Π
α
= ∂
α
φ
giving the canonical momentum Π = Π
0
= ∂
0
φ = ∂
0
φ =
˙
φ,
Π =
˙
φ
B) Evaluating
∂L
∂φ
= −m
2
φ, the EulerLagrange equations give
the ﬁeld equation as ∂
µ
∂
µ
φ +m
2
φ or
(P
2
+m
2
)φ = 0
¨
φ −
2
φ +m
2
φ = 0
which is the KleinGordon equation for a free, massive scalar
ﬁeld. In momentum space p
2
= −P
2
, thus
(p
2
−m
2
)φ = 0
18 CHAPTER 1. LAGRANGIAN FIELD THEORY
(Note that some authors [Muirhead] deﬁne P
2
≡
2
−
∂
2
∂t
2
dif
ferent from (1.1), so that they write the KleinGordon equation
as (P
2
−m
2
)φ = 0 or (p
2
+m
2
)φ = 0.)
C) The energy momentum tensor is
T
µν
≡ Π
µ
∂
ν
φ −g
µν
L
= ∂
µ
φ∂
ν
φ −g
µν
L
= ∂
µ
φ∂
ν
φ −
1
2
g
µν
(∂
α
φ∂
α
φ −m
2
φ
2
)
Therefore the Hamiltonian density is H ≡ T
00
=
˙
φ
2
−
1
2
(∂
α
φ∂
α
φ−
m
2
φ
2
) which becomes [Leon]
H =
1
2
˙
φ
2
+
1
2
(φ)
2
+
1
2
m
2
φ
2
=
1
2
[Π
2
+ (φ)
2
+m
2
φ
2
]
1.5 Noether’s Theorem
Noether’s theorem provides a general and powerful method for discussing
symmetries of the action and Lagrangian and directly relating these sym
metries to conservation laws.
Many books [Kaku] discuss Noether’s theorem in a piecemeal fashion,
for example by treating internal and spacetime symmetries separately. It
is better to develop the formalism for all types of symmetries and then to
extract out the spacetime and internal symmetries as special cases. The
best discussion of this approach is in [Goldstein, Section 127, pg. 588]
and [Greiner FQ, section 2.4, pg. 39]. Another excellent discussion of this
general approach is presented in [Schwabl, section 12.4.2, pg. 268]. However
note that the discussion presented by [Schwabl] concerns itself only with the
symmetries of the Lagrangian, although the general spacetime and internal
symmetries are properly treated together. The discussions by [Goldstein]
and [Greiner] treat the symmetries of both the Lagrangian and the action
as well.
In what follows we rely on the methods presented by [Goldstein].
The theory below closely follows [Greiner FQ 40]. We prefer to use the
notation of [Goldstein] for ﬁelds, namely η
r
(x) or η
r
(x
µ
) rather than using
1.5. NOETHER’S THEOREM 19
φ
r
(x) or ψ
r
(x) because the latter notations might make us think of scalar
or spinor ﬁelds. The notation η
r
(x) is completely general and can refer to
scalar, spinor or vector ﬁeld components.
We wish to consider how the Lagrangian and action change under a
coordinate transformation
x
µ
→x
µ
≡ x
µ
+δx
µ
Let the corresponding change in the ﬁeld (total variation) be [Ryder83,
Schwabl263]
η
r
(x
) ≡ η
r
(x) + ∆η
r
(x)
and the corresponding change in the Lagrangian
L
(x
) ≡ L(x) + ∆L(x)
with
L(x) ≡ L(η(x), ∂
µ
η(x), x)
where ∂
µ
η(x) ≡
∂η(x)
∂xµ
and
1
L
(x
) ≡ L(η
(x
), ∂
µ
η
(x
), x
)
(Note: no prime on L on right hand side) where ∂
µ
η
(x
) ≡
∂η
(x
)
∂x
µ
Notice that the variations deﬁned above involve two transformations,
namely the change in coordinates from x to x
and also the change in the
shape of the function from η to η
.
However there are other transformations (such as internal symmetries or
gauge symmetries) that change the shape of the wave function at a single
point. Thus the local variation is deﬁned as (same as before)
η
r
(x) ≡ η
r
(x) +δη
r
(x)
1
This follows from the assumption of form invariance [Goldstein 589]. In general the
Lagrangian gets changed to
L(ηr(x), ∂νηr(x), x) →L
(η
r
(x
), ∂
ν
η
r
(x
), x
)
with ∂
ν
≡
∂
∂x
ν
The assumption of form invariance [Goldstein 589] says that the Lagrangian has the same
functional form in terms of the transformed quantities as it does in the original quantities,
namely
L
(η
r
(x
), ∂
ν
η
r
(x
), x
) = L(η
r
(x
), ∂
ν
η
r
(x
), x
)
20 CHAPTER 1. LAGRANGIAN FIELD THEORY
The local and total variations are related via
δη
r
(x) = η
r
(x) −η
r
(x)
= η
r
(x) −η
r
(x
) +η
r
(x
) −η
r
(x)
= −[η
r
(x
) −η
r
(x)] + ∆η
r
(x)
Recall the Taylor series expansion
f(x) = f(a) + (x −a)f
(a) +...
= f(a) + (x −a)
∂f(x)
∂x
[
x=a
+...
or
f(x) −f(a) ≈ (x −a)
∂f
∂a
which gives
η(x
) −η(x) ≈ (x
−x)
∂η(x
)
∂x
[
x
=x
≡ (x
−x)
∂η
∂x
= δx
∂η
∂x
Thus
δη
r
(x) = ∆η
r
(x) −
∂η
r
∂x
µ
δx
µ
To lowest order η
r
≈ η
r
. We do this because the second term is second order
involving both ∂η
and δx
µ
. Thus ﬁnally we have the relation between the
total and local variations as (to ﬁrst order)
δη
r
(x) = ∆η
r
(x) −
∂η
r
∂x
µ
δx
µ
Now we ask whether the variations ∆ and δ commute with diﬀerentia
tion. (It turns out δ does commute but ∆ does not.) From the deﬁnition
δη(x) ≡ η
(x) −η(x) it is obvious that (see before)
∂
∂x
µ
δη(x) = δ
∂η(x)
∂x
µ
1.5. NOETHER’S THEOREM 21
showing that δ “commutes” with ∂
µ
≡
∂
∂xµ
. However ∆ does not commute,
but has an additional term, as in (see Problems) [Greiner FQ41]
∂
∂x
µ
∆η(x) = ∆
∂η(x)
∂x
µ
+
∂η(x)
∂x
ν
∂δx
ν
∂x
µ
Let us now study invariance of the action [Goldstein 589, Greiner FQ
41]. The assumption of scale invariance [Goldtein 589] says that the action
is invariant under the transformation
2
(i.e. transformation of an ignorable
or cyclic coordinate)
S
≡
Ω
d
4
x
L
(η
r
(x
µ
), ∂
ν
η
r
(x
µ
), x
µ
)
=
Ω
d
4
x L(η
r
(x
µ
), ∂
ν
η
r
(x
µ
), x
µ
) ≡ S
Demanding that the action is invariant, we have (in shorthand notation)
δS ≡
Ω
d
4
x
L
(x
) −
Ω
d
4
xL(x) ≡ 0
Note that this δS is deﬁned diﬀerently to the δS that we used in the deriva
tion of the EulerLagrange equations. Using L
(x
) ≡ L(x) + ∆L(x) gives
δS ≡
Ω
d
4
x
∆L(x) +
Ω
d
4
x
L(x) −
Ω
d
4
xL(x) = 0
We transform the volume element with the Jacobian
d
4
x
=
∂x
µ
∂x
ν
d
4
x
using x
µ
= x
µ
+δx
µ
which gives [Greiner FQ 41]
2
Combining both form invariance and scale invariance gives [Goldstein 589]
δS ≡ S
−S =
Ω
d
4
x
L(η
r
(x
), ∂
ν
η
r
(x
), x
) −
Ω
d
4
xL(ηr(x), ∂νηr(x), x) = 0
In the ﬁrst integral x
is just a dummy variable so that
Ω
d
4
x L(η
r
(x), ∂νη
r
(x), x) −
Ω
d
4
x L(ηr(x), ∂νηr(x), x) = 0
which [Goldstein] uses to derive current conservation.
22 CHAPTER 1. LAGRANGIAN FIELD THEORY
d
4
x
=
∂x
µ
∂x
ν
d
4
x =
1 +
∂δx
0
∂x
0
∂δx
0
∂x
1
. . .
∂δx
1
∂x
0
1 +
∂δx
1
∂x
1
.
.
.
.
.
. . . . 1 +
∂δx
3
∂x
3
d
4
x
= (1 +
∂δx
µ
∂x
µ
)d
4
x
to ﬁrst order only. Thus the variation in the action becomes
δS =
Ω
(1 +
∂δx
µ
∂x
µ
)d
4
x ∆L(x) +
Ω
(1 +
∂δx
µ
∂x
µ
)d
4
x L(x) −
Ω
d
4
x L(x)
=
Ω
d
4
x ∆L(x) +
Ω
d
4
x L(x)
∂δx
µ
∂x
µ
to ﬁrst order. The second order term
∂δx
µ
∂x
µ
∆L(x) has been discarded. Using
the relation between local and total variations gives
δS =
Ω
d
4
x (δL(x) +
∂L
∂x
µ
δx
µ
) +
Ω
d
4
x L(x)
∂δx
µ
∂x
µ
=
Ω
d
4
x ¦δL(x) +
∂
∂x
µ
[L(x)δx
µ
]¦
Recall that L(x) ≡ L(η
r
(x), ∂
µ
η
r
(x)). Now express the local variation δL in
terms of total variations of the ﬁeld as
δL =
∂L
∂η
r
δη
r
+
∂L
∂(∂
µ
η
r
)
δ(∂
µ
η
r
)
= ” +
∂L
∂(∂
µ
η
r
)
∂
µ
δη
r
because δ “commutes” with ∂
µ
≡
∂
∂xµ
. Now add zero,
δL =
∂L
∂η
r
δη
r
−
¸
∂
µ
∂L
∂(∂
µ
η
r
)
δη
r
+
¸
∂
µ
∂L
∂(∂
µ
η
r
)
δη
r
+
∂L
∂(∂
µ
η
r
)
∂
µ
δη
r
=
¸
∂L
∂η
r
−∂
µ
∂L
∂(∂
µ
η
r
)
δη
r
+∂
µ
¸
∂L
∂(∂
µ
η
r
)
δη
r
Note: the summation convention is being used for both µ and r. This ex
pression for δL is substituted back into δS = 0, but because the region of
1.5. NOETHER’S THEOREM 23
integration is abritrary, the integrand itself has to vanish. Thus the inte
grand is
¸
∂L
∂η
r
−∂
µ
∂L
∂(∂
µ
η
r
)
δη
r
+∂
µ
¸
∂L
∂(∂
µ
η
r
)
δη
r
+Lδx
µ
= 0
The ﬁrst term is just the EulerLagrage equation which vanishes. For η
r
use the relation between local and total variations, so that the second term
becomes
∂
µ
¸
∂L
∂(∂
µ
η
r
)
∆η
r
−
∂η
r
∂x
ν
δx
ν
+Lδx
µ
= 0
which is the continuity equation
∂
µ
j
µ
= 0
with [Schwabl 270]
j
µ
≡
∂L
∂(∂
µ
η
r
)
∆η
r
−
¸
∂L
∂(∂
µ
η
r
)
∂
ν
η
r
−g
µν
L
δx
ν
≡
∂L
∂(∂
µ
η
r
)
∆η
r
−T
µν
δx
ν
with the energymomentum tensor deﬁned as [Scwabl 270]
T
µν
≡
∂L
∂(∂
µ
η
r
)
∂
ν
η
r
−g
µν
L
The corresponding conserved charge is (See Problems)
Q ≡
d
3
xj
0
(x)
such that
dQ
dt
= 0
Thus we have j
0
(x) is just the charge density
j
0
(x) ≡ ρ(x)
This leads us to the statement,
Noether’s Theorem: Each continuous symmetry transformation
that leaves the Lagrangian invariant is associated with a con
served current. The spatial integral over this current’s zero com
ponent yields a conserved charge. [Mosel 16]
24 CHAPTER 1. LAGRANGIAN FIELD THEORY
1.6 Spacetime Symmetries
The symmetries we will consider are spacetime symmetries and internal sym
metries. Super symmetries relate both of these. The simplest spacetime
symmetry is 4dimensional translation invariance, involving space transla
tion and time translation.
1.6.1 Invariance under Translation
[GreinerFQ 43] Consider translation by a constant factor
µ
,
x
µ
= x
µ
+
µ
and comparing with x
µ
= x
µ
+ δx
µ
gives δx
µ
=
µ
. The shape of the ﬁeld
does not change, so that ∆η
r
= 0 (which is properly justiﬁed in Schwabl
270) giving the current as
j
µ
= −
∂L
∂(∂
µ
η
r
)
∂η
r
∂x
ν
−g
µν
L
ν
with ∂
µ
j
µ
= 0. Dropping oﬀ the constant factor
ν
lets us write down a
modiﬁed current (called the energymomentum tensor)
T
µν
≡
∂L
∂(∂
µ
η
r
)
∂
ν
η
r
−g
µν
L
with
∂
µ
T
µν
= 0
In general j
µ
has a conserved charge Q ≡
d
3
x j
o
(x). Thus T
µν
will have
4 conserved charges corresponding to T
00
, T
01
, T
02
, T
03
which are just
the energy E and momentum
P of the ﬁeld. In 4dimensional notation
[GreinerFQ 43]
P
ν
= (E,
P) =
d
3
x T
0ν
= constant.
with
dP
ν
dt
= 0
The above expression for T
µν
is the same result we obtained before where
we wrote
T
µν
= π
µ
∂
ν
φ −g
µν
L (1.4)
with
π
µ
=
∂L
∂(∂
µ
φ)
(1.5)
1.6. SPACETIME SYMMETRIES 25
1.6.2 Angular Momentum and Lorentz Transformations
NNN: below is old Kaku notes. Need to revise; Schwabl and Greiner are
best (they do J=L+S)
Instead of a simple translation δx
i
= a
i
now consider a rotation δx
i
=
a
ij
x
j
. Lorentz transformations are a generalisation of this rotation, namely
δx
µ
=
µ
ν
x
ν
. Before for spacetime translations we had δx
µ
= a
µ
and
therefore δφ =
∂φ
∂xµ
δx
µ
= δx
µ
∂
µ
φ = a
µ
∂
µ
φ. Copying this, the Lorentz
transformation is
δx
µ
=
µ
ν
x
ν
δφ =
µ
ν
x
ν
∂
µ
φ
δ∂
ρ
φ =
µ
ν
x
ν
∂
µ
∂
ρ
φ
Now repeat same step as before, and we get the conserved current
´
ρµν
= T
ρν
x
µ
−T
ρµ
x
ν
with
∂
ρ
´
ρµν
= 0
and the conserved charge
M
µν
=
d
3
x ´
0µν
with
d
dt
M
µν
= 0
For rotations in 3d space, the
d
dt
M
ij
= 0 corresponds to conservation of
angular momentum.
26 CHAPTER 1. LAGRANGIAN FIELD THEORY
1.7 Internal Symmetries
[Guidry 9192]
One of the important theorems in Lie groups is the following :
Theorem: Compact Lie groups can always be represented by
ﬁnitedimensional unitary operators. [Tung p.173,190]
Thus using the notation U(α
1
, α
2
, ...α
N
) for an element of an Nparameter
Lie group (α
i
are the group parameters), we can write any group element as
U(α
1
, α
2
, ...α
N
) = e
iα
i
X
i
≈ 1 +i
i
X
i
+...
where the latter approximation is for inﬁnitessimal group elements α
i
=
i
.
The X
i
are linearly independent Hermitian operators (there are N of them)
which satisfy the Lie algebra
[X
i
, X
j
] = i f
ijk
X
k
where f
ijk
are the structure constants of the group.
These group elements act on wave functions, as in [Schwabl 272]
η(x) →η
(x) = e
iα
i
X
i
η(x)
≈ (1 +i
i
X
i
)η(x)
giving
δη(x) = η
(x) −η(x) = i
i
X
i
η(x)
Consider the Dirac equation (i / ∂ −m)ψ = 0 with 4current, j
µ
=
¯
ψγ
µ
ψ.
This is derived from the Lagrangian L =
¯
ψ(i / ∂ − m)ψ where
¯
ψ ≡ ψ
†
γ
0
.
Now, for α
i
= constant, the Dirac Lagrangian L is invariant under the
transformation ψ → ψ
= e
iα
i
X
i
ψ. This is the signiﬁcance of group theory
in quantum mechanics. Noether’s theorem now tells us tht we can ﬁnd a
corresponding conserved charge and conserved current.
The Noether current, with δx
ν
= 0 and therefore δη(x) = ∆η(x) [Schw
abl 272], is
j
µ
=
∂L
∂(∂
µ
η
r
)
δη
r
=
∂L
∂(∂
µ
η
r
)
i
i
X
i
η
r
1.7. INTERNAL SYMMETRIES 27
and again dropping oﬀ the constant factor
i
deﬁne a new curent (and throw
in a minus sign so that we get a positive current in the example below)
j
µ
i
= −
∂L
∂(∂
µ
η
r
)
i X
i
η
r
which obeys a continuity equation
∂
µ
j
µ
i
= 0
Example Calculate j
µ
i
for the isospin transformation e
iα
i
X
i
for the Dirac
Lagrangian L =
¯
ψ(i / ∂ −m)ψ ≡
¯
ψ(iγ
µ
∂
µ
−m)ψ
Solution
j
µ
i
= −
∂L
∂(∂
µ
ψ)
iX
i
ψ
and
∂L
∂(∂
µ
ψ)
=
¯
ψ i γ
µ
giving
j
µ
i
= −i
¯
ψγ
µ
iX
i
ψ
or
j
µ
i
=
¯
ψγ
µ
X
i
ψ
where X
i
is the group generator. (Compare this to the ordinary Dirac
probabilty current j
µ
= (ρ,
j ) =
¯
ψγ
µ
ψ).
28 CHAPTER 1. LAGRANGIAN FIELD THEORY
From the previous example we can readily display conservation of charge.
Write the U(1) group elements as
U(θ) = e
iθq
where the charge q is the generator. Thus the conserved current is
j
µ
= q
¯
ψγ
µ
ψ
where J
µ
= (ρ,
j ) =
¯
ψγ
µ
ψ is just the probability current for the Dirac
equation. The conserved charge is [Mosel 17,34]
Q =
d
3
xj
0
= q
d
3
x
¯
ψ γ
0
ψ
= q
d
3
xρ
(Note that ρ =
¯
ψγ
0
ψ = ψ
†
γ
0
γ
0
ψ = ψ
†
ψ = Σ
4
i=1
[ψ
i
[
2
where ψ
i
is each
component of ψ. Thus ρ is positive deﬁnite.)
See also [Mosel 17,34; BjRQM 9]. Note that if ψ is normalized so that
[Strange 123]
d
3
xρ =
d
3
xψ
†
ψ = 1
then we have Q = q as required. This is explained very clearly in [Gross
122124]. However often diﬀerent normalizations are used for the Dirac
wave functions [Muirhead 72, Halzen & Martin 110]. For example [Halzen
& Martin 110] have
d
3
xρ =
d
3
xψ
†
ψ = u
†
u = 2E
See also [Griﬃths 223].
1.8. SUMMARY 29
1.8 Summary
1.8.1 Covariant and contravariant vectors
Contravariant and covariant vectors and operators are
A
µ
= (A
o
,
A) A
µ
= (A
o
, −
A)
and
∂
µ
= (
∂
∂t
,
) ∂
µ
= (
∂
∂t
, −
)
1.8.2 Classical point particle mechanics
The point particle canonical momentum is
p
i
=
∂L
∂ ˙ q
i
and the EL equations are
d
dt
(
∂L
∂ ˙ q
i
) −
∂L
∂q
i
= 0
The point particle Hamiltonian is
H(p
i
, q
i
) ≡ p
i
˙ q
i
−L(q
i
, ˙ q
i
) (1.6)
giving Hamilton’s equations
∂H
∂p
i
= ˙ q
i
−
∂H
∂q
i
= ˙ p
i
1.8.3 Classical ﬁeld theory
For classical ﬁelds φ
i
, the EL equations are
∂
µ
∂L
∂(∂
µ
φ
i
)
−
∂L
∂φ
i
= 0
The covariant momentum density is
Π
µ
≡
∂L
∂(∂
µ
φ)
and the canonical momentum is
Π ≡ Π
0
=
∂L
∂
˙
φ
30 CHAPTER 1. LAGRANGIAN FIELD THEORY
The energy momentum tensor is (analagous to point particle Hamiltonian)
T
µν
≡ Π
µ
∂
ν
φ −g
µν
L
with the Hamiltonian density
H ≡ T
00
= Π
˙
φ −L
1.8.4 Noether’s theorem
The conserved (∂
µ
j
µ
= 0) Noether current is
j
µ
≡
∂L
∂(∂
µ
η
r
)
∆η
r
−T
µν
δx
ν
with
T
µν
≡
∂L
∂(∂
µ
η
r
)
∂
ν
η
r
−g
µν
L
The conserved (
dQ
dt
= 0) charge is
Q ≡
d
3
xj
0
(x)
which is just the charge density,
j
0
(x) ≡ ρ(x)
If we consider the spacetime symmetry involving invariance under transla
tion then we can derive T
µν
from j
µ
. The result for T
µν
agrees with that
given above. For the KleinGordon Lagrangian this becomes
T
µν
= ∂
µ
φ∂
ν
φ −g
µν
L (1.7)
The momentum is (with E ≡ H)
P
ν
= (H,
P) =
d
3
x T
0ν
= constant (1.8)
where
dP
ν
dt
= 0.
For internal symmetries the group elements can be written
U(α
1
, α
2
, ...α
N
) = e
iα
i
X
i
≈ 1 +i
i
X
i
+...
1.8. SUMMARY 31
where
[X
i
, X
j
] = i f
ijk
X
k
The group elements act on wave functions
η →η
= e
iα
i
X
i
η ≈ (1 +i
i
X
i
)η
giving
δη = η
−η = i
i
X
i
η
The Noether current, for an internal symmetry (δx
ν
= 0 and therefore
δη = ∆η) becomes
j
µ
i
= −
∂L
∂(∂
µ
η
r
)
i X
i
η
r
For the Dirac Lagrangian, invariant under e
iα
i
X
i
, with α
i
= constant,
this becomes
j
µ
i
=
¯
ψγ
µ
X
i
ψ
32 CHAPTER 1. LAGRANGIAN FIELD THEORY
1.9 References and Notes
General references for units (Section 1.1) are [Aitchison and Hey, pg.
526531; Halzen and Martin, pg. 1213; Guidry, pg.511514; Mandl
and Shaw, pg. 9697; Griﬃths, pg. 345; Jackson, pg. 811821; Misner,
Thorne and Wheeler, pg. 3536]. References for Natural units (Section
1.1.1) are [Guidry, pg. 511514; Mandl and Shaw, pg. 9697; Griﬃths, pg.
345; Aitchison and Hey, pg. 526531; Halzen and Martin, pg. 1213] and
references for Geometrical units (Section 1.1.2) are [Guidry, pg. 514; MTW,
pg. 3536].
For a complete introduction to covariant and contravariant vectors see
[Jackson]
The most commonly used metric is
η
µν
=
¸
¸
¸
¸
1 0 0 0
0 −1 0 0
0 0 −1 0
0 0 0 −1
¸
This metric is used throughout the present book and is also used by the
following authors: [Greiner, Aitchison & Hey, Kaku, Peskin & Schroeder,
Ryder, Bjorken and Drell, Mandl & Shaw, Itzykson & Zuber, Sterman,
Chang, Guidry, Griﬃths, Halzen & Martin and Gross].
Another less commonly used metric is
η
µν
=
¸
¸
¸
¸
1 0 0 0
0 −1 0 0
0 0 −1 0
0 0 0 −1
¸
This metric is not used in the present book, but it is used by [Weinberg and
Muirhead].
The best references for the classical ﬁeld ELE and Noether’s theorem
are [Schwabl, Ryder]
Chapter 2
Symmetries & Group theory
2.1 Elements of Group Theory
SUSY nontrivially combines both spacetime and internal symmetries.
2.2 SO(2)
In SO(2) the invariant is x
2
+y
2
. We write
x
y
=
cos θ sin θ
sin θ cosθ
x
y
or
x
i
= O
ij
x
j
For small angles this is reduced to
δx = θy and δy = −θx
or
δx
i
= θ
ij
x
j
where
ij
is antisymmetric and
12
= −
21
= 1
Read Kaku p. 3638
33
34 CHAPTER 2. SYMMETRIES & GROUP THEORY
2.2.1 Transformation Properties of Fields
References: [Kaku 38; Greiner FQ 95, 96; Elbaz 192; HoKim 28; Mosel 21]
Consider a transformation U which transforms a quantum state [Elbaz
l92]
α
>≡ U
α >
< α
[=< α[ U
†
To conserve the norm, we impose
< α
β
>=< α
β >
which means that U is unitary, i.e. UU
†
= 1 implying U
†
= U
−1
.
Now consider transformation of an operator O, with expectation value
< O >≡< α[O[α >=< α
[O
[α
>
This gives
< O >=< α[O[α >=< α[U
†
O
U[α >
giving the transformation rule for the operator as
O = U
†
O
U
or
O
= UOU
†
To summarize, if a quantum state transforms as [Elbaz 192]
ψ
= Uψ
or
[α
>= U[α >
then an operator transforms as
O
= UOU
†
Fields can be grouped into diﬀerent categories [Mosel 21] according to their
behavior under general Lorentz transformations, which include spatial rota
tion, Lorentz boost transformations and also the discrete transformations of
space reﬂection, time reversal and spacetime reﬂection. The general Lorentz
transformation is written [Mosel 21, Kaku 50]
x
µ
= Λ
µν
x
ν
(2.1)
2.3. REPRESENTATIONS OF SO(2) AND U(1) 35
but let’s write it more generally (in case we consider other transformations)
x
µ
= a
µν
x
ν
(2.2)
Λ
µν
or a
µν
for rotation, boost, space inversion is very nicely discussed in
[HoKim 1922]. A scalar ﬁeld transforms under (2.1) or (2.2) as
φ
(x
) = φ(x)
[Mosel 21; HoKim 26; Greiner 95, 96]. Now if φ(x) is an operator then its
transformation is also written as [Greiner 96, Kaku 38]
φ
(x
) = Uφ(x
)U
†
Thus a scalar ﬁeld transforms under (2.2) as
φ
(x
) = Uφ(x
)U
†
= φ(x)
A vector ﬁeld transforms as [HoKim 30; Kaku 39]
φ
µ
(x
) = a
ν
µ
φ
ν
(x) ≡ Uφ
µ
(x
)U
†
i.e. it just transforms in the same way as an ordinary 4vector.
2.3 Representations of SO(2) and U(1)
Read Kaku 3942, 741748
SO(2) can be deﬁned as the set of transformations that leave x
2
+y
2
invari
ant. There is a homomorphism between SO(2) and U(1). A U(1) transfor
mation can be written ψ
= e
iθ
ψ = Uψ. This will leave the inner product
ψ
∗
φ invariant. Thus the group U(1) can be deﬁned as the set of transfor
mations that leave ψ
∗
φ invariant [Kaku 742, Peskin 496]
2.4 Representations of SO(3) and SU(1)
SO(3) leaves x
2
+x
2
+z
2
invariant.
Read Kaku 4245.
36 CHAPTER 2. SYMMETRIES & GROUP THEORY
2.5 Representations of SO(N)
Students should read rest of chapter in Kaku.
NNN NOW do a.m.
Chapter 3
Free KleinGordon Field
NNN write general introduction
3.1 KleinGordon Equation
Relativistic Quantum Mechanics (RQM) is the subject of studying relativis
tic wave equations to replace the nonrelativistic Schrodinger equation. The
two prime relativistic wave equations are the KleinGordon equation (KGE)
and the Dirac equation. (However these are only valid for 1particle prob
lems whereas the Schrodinger equation can be written for many particles.)
Our quantum wave equation (both relativistic and nonrelativistic) is
written
ˆ
Hψ =
ˆ
Eψ
with
ˆ
H ≡
ˆ
T +
ˆ
U
and
ˆ
E = i¯h
∂
∂t
Nonrelativistically we have
ˆ
T =
ˆ p
2
2m
with ˆ p = −i¯h
∇ giving
−
¯h
2
2m
∇
2
+U
ψ = i¯h
∂
∂t
ψ
The timeindependent Schrodinger equation simply has E
b
instead of
ˆ
E
where E
b
is the binding energy, i.e.
−
¯h
2
2m
∇
2
+U
ψ = E
b
ψ
37
38 CHAPTER 3. FREE KLEINGORDON FIELD
However in special relativity we have (with c = 1)
E = T +m
and
E
2
= p
2
+m
2
giving
T =
p
2
+m
2
−m
With the replacement p →−i¯h
∇, the relativistic version of the free particle
(U = 0) Schrodinger equation would be (
ˆ
Tψ = E
b
ψ)
−¯h
2
∇
2
+m
2
−m
ψ = E
b
ψ
In the nonrelativistic case E = E
b
, but relativistically E = E
b
+m, giving
−¯h
2
∇
2
+m
2
ψ = Eψ
which is called the Spinless Salpeter equation. There are two problems with
this equation; ﬁrstly the operator
−¯h
2
∇
2
+m
2
is nonlocal [Landau 221]
making it very diﬃcult to work with in coordinate space (but actually it’s
easy in momentum space) and secondly the equation is not manifestly co
variant [Gross, pg. 92]. Squaring the Spinless Salpeter operator gives the
KleinGordon equation (KGE)
(−¯h
2
∇
2
+m
2
)φ = E
2
φ
= −¯h
2
∂
2
∂t
2
φ
or (with ¯h = c = 1, see Halzen & Martin, pg. 12, 13; Aitchison & Hey, pg.
526528)
¨
φ −∇
2
φ +m
2
φ = 0
Recall the wave equation (with y
≡ ∂
2
y/∂x
2
, ¨ y ≡ ∂
2
y/∂t
2
)
y
−
1
c
2
¨ y = 0
where c is the wave velocity. Thus the KGE (c = 1) is (with ∇
2
= ∂
2
/∂x
2
)
φ
−
1
c
2
¨
φ = m
2
φ
3.2. PROBABILITY AND CURRENT 39
which is like a massive (inhomogeneous) wave equation. The KGE is written
in manifestly covariant form as
(P
2
+m
2
)φ = 0
which in momentum space is (using p
2
→−P
2
)
(p
2
−m
2
)φ = 0
A quick route to the KGE is with the relativistic formula p
2
≡ p
µ
p
µ
= m
2
(= p
2
) giving p
2
− m
2
= 0 and (p
2
− m
2
)φ = 0 and p
2
→ −P
2
giving
(P
2
+m
2
)φ = 0. The KGE can be written in terms of 4vectors, (p
2
−m
2
)φ =
0 and is therefore manifestly covariant. Finally, note that the KGE is a 1
particle equation!
(Note that some authors [Muirhead] use g
µν
=
¸
¸
¸
¸
−1
1
1
1
¸
and
so have P
2
≡ ∇
2
−
∂
2
∂t
2
and (P
2
− m
2
)φ = 0 or (p
2
+ m
2
)φ = 0 for the
KleinGordon equation.)
3.2 Probability and Current
The KleinGordon equation was historically rejected because it predicted a
negative probability density. In order to see this let’s ﬁrst review probability
and current for the Schrodinger equation. Then the KG example will be
easier to understand.
3.2.1 Schrodinger equation
The free particle Schrodinger equation (SE) is
−
¯h
2
2m
∇
2
ψ = i¯h
∂ψ
∂t
The complex conjugate equation is (SE
∗
)
−
¯h
2
2m
∇
2
ψ
∗
= −i¯h
∂ψ
∗
∂t
Multiply SE by ψ
∗
and SE
∗
by ψ
−
¯h
2
2m
ψ
∗
∇
2
ψ = i¯hψ
∗
∂ψ
∂t
40 CHAPTER 3. FREE KLEINGORDON FIELD
−
¯h
2
2m
ψ∇
2
ψ
∗
= −i¯hψ
∂ψ
∗
∂t
and subtract these equations to give
−
¯h
2
2m
(ψ
∗
∇
2
ψ −ψ∇
2
ψ
∗
) = i¯h
ψ
∗
∂ψ
∂t
+ψ
∂ψ
∗
∂t
= i¯h
∂
∂t
(ψ
∗
ψ)
= −
¯h
2
2m
∇ [ψ
∗
∇ψ −(
∇ψ
∗
)ψ]
or
∂
∂t
(ψ
∗
ψ) +
¯h
2mi
∇ [ψ
∗
∇ψ −(
∇ψ
∗
)ψ] = 0
which is just the continuity equation
∂ρ
∂t
+
∇
j = 0 if
ρ ≡ ψ
∗
ψ
j ≡
¯h
2mi
[ψ
∗
∇ψ −(
∇ψ
∗
)ψ]
which are the probability density and current for the Schrodinger equation.
3.2.2 KleinGordon Equation
The free particle KGE is (P
2
+m
2
)φ = 0 or (using P
2
=
∂
2
∂t
2
−∇
2
)
∂
2
φ
∂t
2
−∇
2
φ +m
2
φ = 0
and the complex conjugate equaiton is (KGE
∗
)
∂
2
φ
∗
∂t
2
−∇
2
φ
∗
+m
2
φ
∗
= 0
Multiplying KGE by φ
∗
and KGE
∗
by φ gives
φ
∗
∂
2
φ
∂t
2
−φ
∗
∇
2
φ +m
2
φ
∗
φ = 0
φ
∂
2
φ
∗
∂t
2
−φ∇
2
φ
∗
+m
2
φφ
∗
= 0
and subtract these equations to give
φ
∗
∂
2
φ
∂t
2
−φ
∂
2
φ
∗
∂t
2
−φ
∗
∇
2
φ +φ∇
2
φ
∗
= 0
3.3. CLASSICAL FIELD THEORY 41
=
∂
∂t
φ
∗
∂φ
∂t
−φ
∂φ
∗
∂t
−
∇
φ
∗
∇φ −φ
∇φ
∗
which is the continuity equation ∂ρ/∂t +
∇
j = 0 if
ρ ≡ φ
∗
∂φ
∂t
−φ
∂φ
∗
∂t
j ≡ φ
∇φ
∗
−φ
∗
∇φ
or j
µ
= φ
∗
∂
µ
φ −φ∂
µ
φ
∗
but to get this to match the SE wave function we should deﬁne
j in the
same way, i.e.
j ≡
¯h
2mi
φ
∗
∇φ −φ
∇φ
∗
=
−i¯h
2m
φ
∗
∇φ −φ
∇φ
∗
which is
−¯h
2mi
times
j above. Thus for
∂ρ
∂t
+
∇
j = 0 to hold we must have
ρ ≡
i¯h
2m
φ
∗
∂φ
∂t
−φ
∂φ
∗
∂t
The problem with ρ (in both expressions above) is that it is not positive
deﬁnite and therefore cannot be interpreted as a probability density. This
is one reason why the KGE was discarded. (Note: because ρ can be either
positive or negative it can be interpreted as a charge density. See [Landau,
pg. 227]
Notice how this problematic ρ comes about because the KGE is 2nd order
in time. We have
∂ρ
∂t
and ρ itself constains
∂
∂t
; this does not happen with
the SE or DE.
By the wave, note that we can form a 4vector KG current
j
µ
=
i¯h
2m
(φ
∗
∂
µ
φ −φ∂
µ
φ
∗
)
3.3 Classical Field Theory
Reference: [Schwabl, Chapter 13; Kaku, Chapter 3]
Some of the key results for the free and real KleinGordon ﬁeld were
worked out in an Example in Chapter 1. Let’s remind outselves of these
results.
42 CHAPTER 3. FREE KLEINGORDON FIELD
The massive KleinGordon Lagrangian was
L
KG
=
1
2
(∂
µ
φ∂
µ
φ −m
2
φ
2
)
which gave the equation of motion in position and momentum space as
(P
2
+m
2
)φ = 0
(p
2
−m
2
)φ = 0
The covariant momentum density was
Π
µ
= ∂
µ
φ
giving the canonical momentum
Π ≡ Π
o
=
˙
φ(x)
The Hamiltonian density was (with H ≡
d
3
xH)
H =
1
2
[Π
2
+ (
∇φ)
2
+m
2
φ
2
]
Finally, the momentum operator of the KleinGordon ﬁeld is (see Problems)
P = −
d
3
x
˙
φ(x)
∇φ(x)
3.4 Fourier Expansion & Momentum Space
[See Greiner FQ 76 , Jose and Saletin 589]
As with the nonrelativistic case we expand plane wave states as
φ(x, t) =
d
˜
k a(
k, t)e
i
k·x
but now with
d
˜
k = N
k
d
3
k
where N
k
is a normalization constant, to be determined later.
Substitute φ(r, t) into the KGE
P
2
+m
2
φ = 0
∂
2
t
−∇
2
+m
2
φ = 0
3.4. FOURIER EXPANSION & MOMENTUM SPACE 43
giving
d
˜
k
¨ a +k
2
a +m
2
a
e
i
k·x
= 0
Deﬁning
ω ≡ ω(
k) =
√
k
2
+m
2
and requiring the integrand to be zero gives
¨ a +ω
2
a = 0
which is a 2nd order diﬀerential equation, with Auxilliary equation
r
2
+ω
2
= 0
or
r = ±−iω
The two solutions ± are crucial ! They can be interpreted as positive and
negative energy, or as particle and antiparticle. In the nonrelativistic (NR)
case we only got one solution. Thus [Teller, pg. 67]
a(
k, t) = c(
k)e
iωt
+a(
k)e
−iωt
giving our original expansion as
φ(x, t) =
d
˜
k
a(
k)e
i(
k·x−ωt)
+c(
k)e
i(
k·x+ωt)
(Remember that ω ≡ ω(
k) =
√
k
2
+m
2
) Now the Schrodinger wave function
ψ(x, t) is complex but the KG wave function φ(x, t) is real. (The Schrodinger
equation has an i in it, but the KGE does not.) Thus [Greiner, pg. 77]
φ(x, t) = φ
∗
(x, t) = φ
†
(x, t)
giving
d
˜
k
a(
k)e
i(
k·x−ωt)
+c(
k)e
i(
k·x+ωt)
=
d
˜
k
a
†
(
k)e
−i(
k·x−ωt)
+c
†
(
k)e
−i(
k·x+ωt)
Rewrite this as
d
˜
k
a(
k)e
i(
k·x−ωt)
+c(−
k)e
−i(
k·x−ωt)
=
d
˜
k
a
†
(
k)e
−i(
k·x−ωt)
+c
†
(−
k)e
i(
k·x−ωt)
44 CHAPTER 3. FREE KLEINGORDON FIELD
where we have made the substitution
k →−
k in two terms.
(This does not mean k → −k; it cannot. k is the magnitude of
k. We
are simply reversing the direction (magnitude) of
k. This does not aﬀect the
volume element d
3
k.)
Obviously then
c(−
k) = a
†
(
k)
c
†
(−
k) = a(
k)
Rewriting our expansion as
φ(x, t) =
d
˜
k
a(
k)e
i(
k·x−ωt)
+c(−
k)e
−i(
k·x−ωt)
gives [same conventions as Kaku 152, Mandl and Shaw 44, Schwabl 278]
φ(x, t) ≡ φ
+
(x) +φ
−
(x)
=
d
˜
k
a(
k)e
i(
k·x−ωt)
+a
†
(
k)e
−i(
k·x−ωt)
=
d
˜
k
a(
k)e
−ik·x
+a
†
(
k)e
ik·x
=
¸
k
1
√
2ωV
a
k
e
−ik·x
+a
†
k
e
ik·x
(3.1)
[see Greiner FQ 79 for box vs. continuum normalization] where k x ≡
k
µ
x
µ
= k
o
x
o
−
k x = ωt −
k x.
The conjugate momentum is
Π(x, t) =
˙
φ(x, t) = −i
d
˜
k ω
a(
k)e
i(
k·x−ωt)
−a
†
(
k)e
−i(
k·x−ωt)
= −i
d
˜
k ω
a(
k)e
−ik·x
−a
†
(
k)e
ik·x
≡ Π
+
(x) + Π
−
(x)
(3.2)
Note that we are still studying classical ﬁeld theory when we Fourier expand
these classical ﬁelds [Goldstein 568, Jose & Saletan 588].
3.5. KLEINGORDON QFT 45
3.5 KleinGordon QFT
Recall the commutation relations from nonrelativistic quantum mechanics,
namely
[x, p] = i¯h
and
[a, a
†
] = 1
To develop the QFT we impose similar relations, but this time for ﬁelds.
The equal time commutation relations
[φ(x, t), Π(x
, t)] = iδ(x −x
) (3.3)
and
[φ(x, t), φ(x
, t)] = [Π(x, t), Π(x
, t)] = 0 (3.4)
where we had
Π =
˙
φ
We now need to ﬁnd the commutation relations for a(
k) and a
†
(
k). We
expect the usual results
[a(
k), a
†
(
k
)] = δ(
k −
k
) (3.5)
and
[a(
k), a(
k
)] = [a
†
(
k), a
†
(
k
)] = 0 (3.6)
3.5.1 Indirect Derivation of a, a
†
Commutators
[Greiner, pg. 77]
To check whether our expection above is correct evaluate
[φ(x, t), Π(x
, t)]
= [φ
+
(x, t) +φ
−
(x, t), Π
+
(x
, t) + Π
−
(x
, t)]
= [φ
+
(x, t), Π
+
(x
, t)] + [φ
+
(x, t), Π
−
(x
, t)]
+[φ
−
(x, t), Π
+
(x
, t)] + [φ
−
(x, t), Π
−
(x
, t)]
= −i
d
˜
k
d
˜
k
ω
[a(
k), a(
k
)]e
−i(k·x+k
·x
)
−[a(
k), a
†
(
k
)]e
−i(k·x−k
·x
)
+ [a
†
(
k), a(
k
)]e
i(k·x−k
·x
)
−[a
†
(
k), a
†
(
k
)]e
i(k·x+k
·x
)
]
¸
where k x = ωt −
k x and t
≡ t and ω
≡ ω(
k
) =
k
2
+m
2
.
46 CHAPTER 3. FREE KLEINGORDON FIELD
Inserting (3.5) and (3.6) gives
[φ(x, t), Π(x
, t)] = −i
d
˜
k
d
˜
k
ω
−δ(
k −
k
)e
−i(k·x−k
·x
)
−δ(
k
−
k)e
i(k·x−k
·x
)
= i
d
3
k N
2
k
ω
e
i
k·(x−x
)
+e
−i
k·(x−x
)
Now if [Greiner, pg. 77]
N
k
=
1
2ω(2π)
3
then we obtain
[φ(x, t), Π(x
, t)] = i
d
3
k
1
2ω(2π)
3
ω
e
i
k·(x−x
)
+e
−i
k·(x−x
)
= iδ(x −x
)
as required. Here we have used the result,
1
(2π)
3
d
3
k e
±i
k·(x−x
)
= δ(x −x
)
With the above normalization we have
d
˜
k ≡
d
3
k
2ω(2π)
3
(which is diﬀerent to IZ114, but same as Kaku). Using the result [IZ114,
Kaku 6465]
d
3
k
2ω
= d
4
k δ(k
2
−m
2
) θ(k
0
)
gives
d
˜
k =
2ω
(2π)
3
d
4
kδ(k
2
−m
2
) θ(k
o
)
(which is diﬀerent from IZ 114, but same as Kaku)
Note: Some authors [IZ114] use the normalization N
k
=
1
(2π)
3
2ω
in which
case [a(
k), a
†
(
k
)] = (2π)
3
2ω
k
δ
3
(
k −
k
) [Greiner, pg. 77, footnote].
3.5. KLEINGORDON QFT 47
3.5.2 Direct Derivation of a, a
†
Commutators
The best way to obtain the commutators directly is to invert the Fourier
expansions to obtain a and a
†
in terms of φ and Π. The commutators are
then derived directly.
Inverting (3.1) and (3.2) gives [see Problems]
⇒
a
k
=
1
2ω(2π)
3
d
3
x e
ik·x
[ωφ(x) +iΠ(x)]
a
†
k
=
1
2ω(2π)
3
d
3
x e
−ik·x
[ωφ
†
(x) −iΠ
†
(x)]
where a
k
≡ a(
k) and φ(x) ≡ φ(x, t) and k x ≡ ωt −
k x and by direct
evaluation of the commutators we arrive at (3.5) and (3.6) [do Problem 5.7]
But remember for KGE φ is real and therefore φ = φ
†
and Π = Π
†
giving
a
†
k
=
1
2ω(2π)
3
d
3
x e
−ik·x
[ωφ(x) −iΠ(x)]
3.5.3 KleinGordon QFT Hamiltonian
The 2nd quantized ﬁeld Hamiltonian is (see Problems)
H =
d
3
k
ω
2
(a
†
k
a
k
+a
k
a
†
k
)
and using the commutator
[ a
k
, a
†
k
] = δ(
k −
k
)
gives
H =
d
3
k
N
k
+
1
2
ω
or
¸
k
N
k
+
1
2
ω
with
N
k
≡ a
†
k
a
k
(see Problems)
48 CHAPTER 3. FREE KLEINGORDON FIELD
We can also calculate the momentum
P as [Kaku 66, 67; Schwabl 279]
P =
d
3
k
N
k
+
1
2
k
or
¸
k
N
k
+
1
2
k
3.5.4 Normal order
References [Kaku68, Mosel 27, Schwabl 280]
The previously derived Hamiltonian is actually inﬁnite! This is beacuse the
term
d
3
k
1
2
ω =
1
2
d
3
k
k
2
+m
2
= ∞
This is one of the ﬁrst (of many) places where QFT gives inﬁnite answers.
Now because only energy diﬀerences are observable, we are free to simply
throw away the inﬁnite piece, and rewrite the Hamiltonian as
H =
d
3
k N
k
ω.
A formal way to always get rid of these inﬁnite terms (there is also one
in the previous expression for the momentum
P), is to introduce the idea of
normal order.
In a normal ordered product all annihilation operators are placed
to the right hand side of all creation operators.
Two colons :: are used to denote a normal ordered product. For example
[Schwabl 280]
: a
k
1
a
k
2
a
†
k
3
: = a
†
k
3
a
k
1
a
k
2
: a
†
k
a
k
+a
k
a
†
k
: = 2a
†
k
a
k
In calculating the Hamiltonian (see Problems) we arrived at
H =
d
3
k(a
†
k
a
k
+a
k
a
†
k
)
ω
2
and using the commutator
[ a
k
, a
†
k
] = δ(
k −
k
)
3.5. KLEINGORDON QFT 49
gave
H =
d
3
k(a
†
k
a
k
+a
†
k
a
k
+ 1)
ω
2
=
d
3
k(a
†
k
a
k
+ 1/2)ω
which is inﬁnite. If we deﬁne H to be normal ordered then
: H : =
d
3
k : (a
†
k
a
k
+a
k
a
†
k
) :
ω
2
=
d
3
k(a
†
k
a
k
+a
†
k
a
k
)
ω
2
=
d
3
ka
†
k
a
k
ω
which is ﬁnite. Note that normal ordering is equivalent to treating the boson
operators as if they had vanishing commutator. [Schwabl 280]. (nnn Are we
back to a classical theory?)
50 CHAPTER 3. FREE KLEINGORDON FIELD
3.5.5 Wave Function
[Kaku 6869]
Now that we have expanded the KG ﬁeld and Hamiltonian in terms of
creation and destruction operators, we need something for them to operate
on. These are just the manybody states introduced earlier [Bergstrom &
Goobar 289]
[ n
k
i
n
k
j
` =
¸
i
[n
k
i
` (3.7)
with
[n
k
i
` =
1
n
k
!
a
†
k
n
k
[0`
where the vacuum state is deﬁned via
a
k
[0` = 0
The physical interpretation [Bergstrom & Goobar 290] is provided by
H[ n
k
i
n
k
j
` =
¸
k
n
k
(
k)[ n
k
i
n
k
j
`
where (
k) = ¯hω(
k). Equation (3.7) is interpreted as a many particle state
where n
k
1
have momentum
k
1
, n
k
2
have momentum
k
2
etc.
Thus a 1particle state is written
[1
k
` ≡ [
k` = a
†
k
[0`
or
'1
k
[ ≡ '
k[ = '0[a
k
The states are normalized as
'
k[
k
` = δ(
k −
k
)
giving
'0[a
k
a
†
k
[0` = δ(
k −
k
).
3.6. PROPAGATOR THEORY 51
3.6 Propagator Theory
[Halzen and Martin 145150, Kaku, Bj RQM Chapter 6]
The KleinGordon equation for a free particle is
(P
2
+m
2
)φ = 0
or with the replacement P
2
→−p
2
in momentum space
(p
2
−m
2
)φ = 0
Let’s write the nonfree KGE as
(P
2
+m
2
)φ = J(x)
where J(x) is referred to as a source term. This is solved with the Green
function method by deﬁning a propagator ∆
F
(x −y) as
(P
2
+m
2
)∆
F
(x −y) ≡ −δ
4
(x −y)
so that the solution is
φ(x) = φ
0
(x) −
d
4
y ∆
F
(x −y)J(y)
where φ
0
(x) is the solution with J = 0. (see Problems) Deﬁne the Fourier
transform
∆
F
(x −y) ≡
d
4
k
(2π)
4
e
−ik·(x−y)
∆
F
(k)
Our usual method of solution for propagators or Green functions is to solve
for ∆
F
(k) and then do a contour integral to get ∆
F
(x − y) rather than
solving for ∆
F
(x −y) directly. The Problems show that
∆
F
(k) =
1
k
2
−m
2
52 CHAPTER 3. FREE KLEINGORDON FIELD
Example Derive the momentum space Green function for the
Schrodinger equation.
Solution The free particle Schrodinger equation is
−
¯h
2
2m
∇
2
−i¯h
∂
∂t
ψ = 0
and with the inclusion of a source this is
−
¯h
2
2m
∇
2
−i¯h
∂
∂t
ψ = J(x) ≡ −U(x)ψ(x, t)
or
+
¯h
2
2m
∇
2
+i¯h
∂
∂t
ψ = −J(x) = U(x)ψ(x)
Deﬁne the Green function
¯h
2
2m
∇
2
+i¯h
∂
∂t
G
0
(x −x
) = δ
4
(x −x
)
and deﬁne the Fourier transform
G
0
(x −x
) =
d
4
p
(2π)
4
e
−ip·(x−x
)
G
0
(p)
Now operate on this to give, and using p
µ
≡ (ω, p), with ¯ h = 1
¯h
2
2m
∇
2
+i¯h
∂
∂t
G
0
(x −x
) =
d
4
p
(2π)
4
¸
1
2m
(−i p)
2
+ i(−iω)] e
−ip·(x−x
)
G
0
(p)
= δ
4
(x −x
) =
d
4
p
(2π)
4
¸
−
p
2
2m
+ω
¸
e
−ip·(x−x
)
G
0
(p)
but recall that
δ
4
(x −y) =
d
4
k
(2π)
4
e
−ik·(x−y)
which implies
G
0
(p) =
1
ω − p
2
/2m
3.6. PROPAGATOR THEORY 53
Example Derive the position space Green function leaving the
pole on the Real axis.
Solution Substituting for G
0
(p) into G
0
(x−x
) we need to eval
uate (¯h = 1)
G
0
(x −x
) =
d
4
p
(2π)
4
1
ω − p
2
/2m
e
−ip·(x−x
)
=
d
3
p
(2π)
3
e
i p·(x−x
)
∞
−∞
dω
2π
e
−iω(t−t
)
ω − p
2
/2m
and we see that the integrand is singular at ω = p
2
/2m. This
simple pole is shown in the ﬁgure. We need to decide whether to
integrate in the upper half plane (UHP) or the lower half plane
(LHP). This is dictated by the boundary conditions as follows.
Write ω ≡ Re ω +i Im ω, so that
e
−iω(t−t
)
= e
i Re ω(t−t
)
e
+Imω(t−t
)
For t − t
> 0, the term e
Imω(t−t
)
will blow up for Im ω > 0
but will go to zero for Im ω < 0. Thus the boundary condition
t−t
> 0 dictates we integrate in the LHP. Similarly for t−t
< 0
we use the UHP. This is shown in the ﬁgure. [See also Halzen
and Martin 148]
X
Re ω
Im ω
t < t'
X
Re ω
Im ω
t > t'
ω=ω
0
=p
2
/2m
C
1
C
Now apply the Cauchy Residue Theorem, and remember coun
terclockwise integration is a positive sign. In the left ﬁgure the
54 CHAPTER 3. FREE KLEINGORDON FIELD
coutour does not enclose any poles so that
0 =
ω
0
−δ
−∞
+
C
1
+
∞
ω
0
+δ
+
C
and with
C
= 0 (Jordan’s lemma) and lim
δ→0
we have
∞
−∞
= −
C
1
Thus for t > t
G
0
(x −x
)
t>t
= −+πi
d
3
p
(2π)
3
e
i p·(x−x
)
1
2π
e
−iω
0
(t−t
)
= −
i
2
d
3
p
(2π)
3
e
i p·(x−x
)
e
−iω
0
(t−t
)
with ω
0
≡ p
2
/2m. Similarly for t < t
we obtain
G
0
(x −x
)
t<t
= −−πi
d
3
p
(2π)
3
e
i p·(x−x
)
1
2π
e
−iω
0
(t−t
)
= +
i
2
d
3
p
(2π)
3
e
i p·(x−x
)
e
−iω
0
(t−t
)
But this violates causality ! If the wave pulse is sent out at t
we
do expect a signal at a later time t > t
but certainly not at an
earlier time t < t
.
Because of the singularity, the integral is not well deﬁned until we specify
the limiting process. Let’s now try a diﬀerent method for evaluating the
integral that is consistent with the boundary condition, i.e. we want
G
0
(x −x
)
t<t
= 0.
3.6. PROPAGATOR THEORY 55
Example Evaluate the position space Green function by shifting
the pole oﬀ the real axis by a small amount , and take lim
→0
.
Solution We can shift the pole either above or below the real
axis. We choose to shift it below because then the upper contour
will enclose no poles and will give a zero integral consistent with
our boundary conditions. This is shown in the ﬁgure.
C
t>t’
t<t’
X
ω
0
 i ε
X X
ω
0
 i ε
For t > t
we have
∞
−∞
+
C
= 2πiΣ Residues and with
C
= 0 we
get
G
0
(x −x
)
t>t
= −2πi
d
3
p
(2π)
3
e
i p·(x−x
)
∞
−∞
dω
2π
e
−iω(t−t
)
ω − p
2
/2m+i
= −i
d
3
p
(2π)
3
e
i p·(x−x
)
e
−iω
0
(t−t
)
with lim
→0
e
(t−t
)
= 0, which is exactly double our previous answer.
For t < t
we get
G
0
(x −x
)
t<t
= 0
because no poles are enclosed and this now obeys the boundary
condition.
56 CHAPTER 3. FREE KLEINGORDON FIELD
An excellent discussion of all of these issues can be found in [Arfken 4th
ed. 417, 427; Landau 8488; Bj RQM 85; Cushing 315; Greiner QED 27;
Halzen and Martin 145150]
The key integral that we have been considering is of the form
I =
∞
−∞
dω
e
−iωt
ω −ω
0
and with the pole shifted below the real axis this is
I() =
∞
−∞
dω
e
−iωt
ω −ω
0
+i
and let’s summarize our results as
I
t>0
= −πi e
−iω
0
t
I
t<0
= +πi e
−iω
0
t
and
I()
t>0
= −2π i e
−iω
0
t
I()
t<0
= 0
When we leave the pole on the real axis the integral
∞
−∞
is actually the
Cauchy Principal Value deﬁned as [Arfken 4th ed., p. 417; Landau 88]
P
∞
−∞
f(x)dx ≡ lim
δ→0
¸
x
0
−δ
−∞
f(x)dx +
∞
x
0
+δ
f(x)dx
¸
with the pole at x = x
0
. As discussed by [Arfken 417, 418] the Cauchy
Principal value is actually a cancelling process. See also [Landau 8588].
This is because in the vicinity of the simple pole we have
f(x) ≈
a
x −x
0
which is odd relative to x
0
and so the large singular regions cancel out
[Arfken 4th ed., 417, 418].
The Principal Value Prescription (i.e. our ﬁrst method with the pole left
on the axis) and the i Prescription are related. Consider [Merzbacher, old
edition]
1
ω ±i
=
ω ∓i
ω
2
+
2
=
ω
ω
2
+
2
∓
i
ω
2
+
2
3.6. PROPAGATOR THEORY 57
and using [Merzbacher 3rd ed., p. 631]
δ(x) =
1
π
lim
→0
+
x
2
+
2
we have
1
ω ±i
=
ω
ω
2
+
2
∓iπ δ(ω)
The ﬁrst term on the right hand side becomes
1
ω
as → 0 except if ω = 0.
If f(ω) is a well behaved function we have [Merzbacher, old ed., p. 85]
lim
→0
∞
−∞
f(ω)
ω
ω
2
+
2
dω = lim
→0
¸
−
−∞
f(ω)
dω
ω
+
∞
f(ω)
dω
ω
+
−
f(ω)
ωdω
ω
2
+
2
= P
∞
−∞
f(ω)
dω
ω
+f(0) lim
→0
−
ωdω
ω
2
+
2
= P
∞
−∞
f(ω)
dω
ω
+ 0
where the last integral vanishes because the integrand is an odd function of
ω. Thus we can write [Merzbacher, old ed., p. 85]
lim
→0
1
ω ±i
= P
1
ω
∓iπ δ(ω)
and we can also write [Cushing 315] (with the pole at ω = ω
0
instead of
ω = 0)
P
f(ω)dω
ω −ω
0
=
f(ω)dω
ω −ω
0
±i
±iπ f(ω
0
)
58 CHAPTER 3. FREE KLEINGORDON FIELD
Example Verify the above formula for our previous integral
I =
dω
e
−iωt
ω−ω
0
.
Solution For t > 0 we had
P
e
−iωt
dω
ω −ω
0
= −πi e
−iω
0
t
and
e
−iωt
dω
ω −ω
0
+i
= −2πi e
−iω
0
t
Now f(ω
0
) = e
−iω
0
t
, so that
e
−iωt
dω
ω −ω
0
+i
+iπ f(ω
0
) = −2πi e
−iω
0
t
+πi e
−iω
0
t
= −πi e
−iω
0
t
= P
e
−iωt
dω
ω −ω
0
in agreement with the formula. For t < 0 we had
P
e
−iωt
dω
ω −ω
0
= +πi e
−iω
0
t
and
e
−iωt
dω
ω −ω
0
+i
= 0
so that
e
−iωt
dω
ω −ω
0
+i
+iπ f(ω
0
) = 0 +iπ e
−iω
0
t
= P
e
−iωt
dω
ω −ω
0
also in agreement with the formula.
3.6. PROPAGATOR THEORY 59
Example Show that the θ function deﬁned as
θ(t −t
) =
1 if t > t
0 if t < t
can be written as
θ(t −t
) = lim
→0
−1
2πi
∞
−∞
dω
e
−iω(t−t
)
ω +i
Solution We showed previously that with I() ≡
∞
−∞
dω
e
−iωt
ω−ω
0
+i
we get
I()
t>0
= −2πi e
−iω
0
t
I()
t<0
= 0
Replacing t with t −t
and ω
0
= 0 gives I =
dω
e
−iω(t−t
)
ω+i
and
I
t>t
= −2πi
I
t<t
= 0
which veriﬁes the above integral representation of θ.
Now let’s ﬁnish the job of evaluating the position space Green function
[Bj RQM 84, Greiner QED 27, Kaku 7374]
60 CHAPTER 3. FREE KLEINGORDON FIELD
Example Evaluate the position space Green function, consistent
with the causality boundary condition, in terms of θ(t −t
). Also
write the answer in terms of the plane wave states
φ
p
(x) ≡
e
−ip·x
(2π)
3/2
Solution To satisfy causality we use the i prescription to eval
uate the integral which gave
G
0
(x −x
)
t>t
= −i
d
3
p
(2π)
3
e
i p·(x−x
)
e
−iω
0
(t−t
)
and
G
0
(x −x
)
t<t
= 0
with ω
0
≡ p
2
/2m. This can be written
G
0
(x −x
) = −i
d
3
p
(2π)
3
e
i p·(x−x
)−iω
0
(t−t
)
θ(t −t
)
Using the plane wave states we have
φ
p
(x)φ
∗
p
(x
) =
1
(2π)
3
e
−ip·x
e
ip·x
=
1
(2π)
3
e
−ip·(x−x
)
=
1
(2π)
3
e
−i[ω(t−t
)− p·(x−x
)]
giving
G
0
(x −x
) = −iθ(t −t
)
d
3
p φ
p
(x)φ
∗
p
(x
)
[Bj RQM 86, Kaku 74]
Finally the d
3
p integrating gives (see Problems) [Bj RQM 86]
G
0
(x −x
) = −i
m
2πi(t −t
)
3/2
e
imx−x

2
2(t−t
)
θ(t −t
)
See footnote in [Bj RQM 86] discussing the Schrodinger equation
and the Diﬀusion equation.
3.6. PROPAGATOR THEORY 61
KleinGordon Propagator
We originally wrote the KleinGordon propagator ∆
F
(k) =
1
k
2
−m
2
but
then digressed on a lengthy discussion of the Schrodinger propagator in order
to illustrate the integration techniques. We now return to the KleinGordon
propagator. Consider the following i prescription
∆
F
(k) =
1
k
2
−m
2
+i
Write this as
1
k
2
−m
2
+i
=
1
k
2
0
−
k
2
−m
2
+i
=
1
k
2
0
−E
2
+i
with E ≡ +
k
2
+m
2
We can show (see Problems)
1
k
2
−m
2
+i
=
1
k
2
0
−E
2
+i
=
1
2k
0
¸
1
k
0
−E +i
+
1
k
0
+E −i
Recall that the momentum space propagator ∆
F
(k) was deﬁned in terms of
the Fourier transform
∆
F
(x −y) =
d
4
k
(2π)
4
e
−ik.(x−y)
∆
F
(k)
Performing the integration as before we get, with d
˜
˜
k ≡
d
3
k
(2π)
3
2ω
k
(see Prob
lems)
∆
F
(x −x
) = −iθ(t −t
)
d
˜
˜
k e
−ik·(x−x
)
−iθ(t
−t)
d
˜
˜
k e
ik·(x−x
)
= −iθ(t −t
)
d
˜
˜
k φ
k
(x)φ
∗
k
(x
) −iθ(t
−t)
d
˜
˜
k φ
∗
k
(x)φ
k
(x
)
62 CHAPTER 3. FREE KLEINGORDON FIELD
Example Show that i∆
F
(x −x
) = '0[Tφ(x)φ(x
)[0`
(Note this is also equal to
= '0[Tφ(x)φ
†
(x
)[0`
for Hermitian ﬁelds [BjRQF42].)
Solution The time ordered product is
TA(t
1
)B(t
2
) ≡
A(t
1
)B(t
2
) if t
1
> t
2
B(t
2
)A(t
1
) if t
2
> t
1
≡ θ(t
1
−t
2
)A(t
1
)B(t
2
) +θ(t
2
−t
1
)B(t
2
)A(t
1
)
Thus for the ﬁelds
Tφ(x)φ(x
) = θ(t −t
)φ(x)φ(x
) +θ(t
−t)φ(x
)φ(x)
or
'0[Tφ(x)φ(x
)[0` = θ(t −t
)'0[φ(x)φ(x
)[0` +θ(t
−t)'0[φ(x
)φ(x)[0`
Thus we want to evaluate '0[φ(x)φ(x
)[0` and '0[φ(x
)φ(x)[0`. Now
φ(x) =
d
˜
k (a
k
e
−ik·x
+a
†
k
e
ik·x
)
with d
˜
k ≡
d
3
k
√
(2π)
3
2ω
and ω ≡
k
2
+m
2
and ω
≡
k
2
+m
2
. Evaluate
'0[φ(x)φ(x
)[0`
=
'0[
(2π)
3
d
3
k
√
2ω
d
3
k
√
2ω
(a
k
e
−ik·x
+a
†
k
e
ik·x
)(a
k
e
−ik
·x
+a
†
k
e
ik
·x
)[0`
=
d
˜
k d
˜
k
'0[a
k
a
k
e
−i(k·x+k
·x
)
+a
k
a
†
k
e
−i(k·x−k
·x
)
+a
†
k
a
k
e
i(k·x−k
·x
)
+a
†
k
a
†
k
e
i(k·x+k
·x
)
[0`
However a
k
[0` = 0 and '0[a
†
k
= 0. Thus the 1st, 3rd and 4th terms are zero.
We are only left with
'0[a
k
a
†
k
[0` = δ(
k −
k
)
3.6. PROPAGATOR THEORY 63
Continuing
'0[φ(x)φ(x
)[0` =
1
(2π)
3
d
3
k
√
2ω
d
3
k
√
2ω
δ(
k −
k
)e
−i(ωt−
k·x−ω
t
+
k
·x
)
=
1
(2π)
3
d
3
k
2ω
e
i
k.(x−x
)−iω(t−t
)
=
1
(2π)
3
d
3
k
2ω
e
−ik·(x−x
)
Obviously we also have
'0[φ(x
)φ(x)[0` =
1
(2π)
3
d
3
k
2ω
e
−ik·(x
−x)
=
1
(2π)
3
d
3
k
2ω
e
ik·(x−x
)
Thus
'0[Tφ(x)φ(x
)[0` = θ(t −t
)'0[φ(x)φ(x
)[0` +θ(t
−t)'0[φ(x
)φ(x)[0`
= θ(t −t
)
d
3
k
(2π)
3
2ω
e
−ik·(x−x
)
+θ(t
−t)
d
3
k
(2π)
3
2ω
e
ik·(x−x
)
= i∆
F
(x −x
)
64 CHAPTER 3. FREE KLEINGORDON FIELD
Other Propagators
Recall that
φ(x) =
d
˜
k[a
k
e
−ik·x
+a
†
k
e
ik·x
]
≡ φ
+
(x) +φ
−
(x) [Kaku 152; Mandl & Shaw 44]
From this we can deﬁne other propagators (see Problems)
i∆
±
(x −y) ≡ [φ
±
(x), φ
∓
(y)] = ±
d
3
k
(2π)
3
2ω
k
e
∓ik·(x−y)
where ω
k
≡ +
k
2
+m
2
giving (see Problems)
i∆(x −y) ≡ [φ(x), φ(y)] = i∆
+
(x −y) +i∆
−
(x −y)
or just ∆(x) = ∆
+
(x) + ∆
−
(x). Thus one has [Mandl & Shaw 51]
∆(x) = ∆
+
(x) + ∆
−
(x) =
−1
(2π)
3
d
3
k
ω
k
sin k x
=
−i
(2π)
3
d
4
k δ(k
2
−m
2
)(k
0
)e
−ik·x
with (k
0
) =
k
0
k
0

=
+1 for k
0
> 0
−1 for k
0
< 0
Note also that ∆
−
(x) = −∆
+
(−x)
These propagators are related to the Feynman propagator by [Mandl
and Shaw 54]
∆
F
(x) = θ(t)∆
+
(x) −θ(−t)∆
−
(x)
or [Schwabl 283]
∆
F
(x −x
) = θ(t −t
)∆
+
(x −x
) −θ(t
−t)∆
−
(x −x
)
which can be written [Schwabl 283, Mandl and Shaw 54]
∆
F
(x) = ±∆
±
(x) for t
>
<
0
3.6. PROPAGATOR THEORY 65
Example Show that
∆
F
(x −x
) = θ(t −t
)∆
+
(x −x
) −θ(t
−t)∆
−
(x −x
)
where
i∆
F
(x −x
) = '0[Tφ(x)φ(x
)[0`
and
i∆
±
(x −y) ≡ [φ
±
(x), φ
∓
(y)].
Solution
i∆
F
(x −x
) = '0[Tφ(x)φ(x
)[0`
= θ(t −t
)'0[φ(x)φ(x
)[0` +θ(t
−t)'0[φ(x
)φ(x)[0`
Thus to do this problem we really need to show that
'0[φ(x)φ(x
)[0` = [φ
+
(x), φ
−
(x
)]
and
'0[φ(x
)φ(x)[0` = −[φ
−
(x), φ
+
(x
)]
Let’s only do the ﬁrst of these. The question is, how do these relations
come about ? The answer is easy. Commutators are just cnumbers, i.e.
classical functions or numbers. Thus '0[c[0` = c'0[0` = c for any cnumber.
Therefore
[φ
+
(x), φ
−
(x
)] = '0[[φ
+
(x), φ
−
(x
)][0`
= '0[φ
+
(x)φ
−
(x
)[0` −'0[φ
−
(x
)φ
+
(x)[0`
= '0[φ
+
(x)φ
−
(x
)[0`
because φ
+
[0` = 0.
Now φ = φ
+
+φ
−
. Thus
φ[0` = φ
+
[0` +φ
−
[0` = 0 +φ
−
[0` = φ
−
[0`
and similarly
'0[φ = '0[φ
+
giving
[φ
+
(x), φ
−
(x
)] = '0[φ(x)φ(x
)[0`
as required. (Proof of the other relation follows similarly).
66 CHAPTER 3. FREE KLEINGORDON FIELD
3.7 Complex KleinGordon Field
[Schwabl 285; Kaku 69; Huang 2425]
[W. Pauli and V. F. Weisskopf, Helv. Phys. Acta, 7, 709 (1934)]
Charged particles cannot be described with a real scalar ﬁeld because it
only has one component, whereas a minimum of two components is needed
to describe charge. Deﬁne
φ ≡
1
√
2
(φ
1
+iφ
2
) and φ
†
≡
1
√
2
(φ
1
−iφ
2
)
where φ
1
and φ
2
are real ﬁelds. The normalization factor
1
√
2
is chosen so
that φ
i
has the same renormalization as the real scalar ﬁeld discussed before.
A suitable classical Lagrangian is (with [φ[
2
= φφ
†
) [Huang 24]
L = ∂
µ
φ
†
∂
µ
φ −m
2
[φ[
2
=
1
2
2
¸
i=1
(∂
µ
φ
i
∂
µ
φ
i
−m
2
φ
2
i
)
which is just the sum of the Lagrangian for the real ﬁelds.
The EulerLagrange equations give the equations of motion [Schwabl
285]
(P
2
+m
2
)φ = 0 and (P
2
+m
2
)φ
†
= 0
with the conjugate momenta being [Kaku 70; Schwabl 285]
Π =
˙
φ
†
and Π
†
=
˙
φ
with the equal time commutation relations [Kaku 70; Schwabl 285]
[φ(x, t), Π(x
, t)] = [φ(x, t),
˙
φ
†
(x
, t)] = iδ(x −x
)
[φ
†
(x, t), Π
†
(x
, t)] = [φ
†
(x, t),
˙
φ(x
, t)] = iδ(x −x
)
and [Schwabl 285]
[φ(x, t), φ(x
, t)] = [Π(x, t), Π(x
, t)] = 0
[φ
†
(x, t), φ
†
(x
, t)] = [Π
†
(x, t), Π
†
(x
, t)] = 0
We Fourier expand each component φ
i
(x) exactly as before [Kaku 70; Huang
24]
φ
i
(x) =
d
˜
k (a
i
k
e
−ik·x
+a
†
i
k
e
ik·x
)
3.7. COMPLEX KLEINGORDON FIELD 67
where the commutation relations are [Kaku 70; Huang 25]
[a
i
k
, a
†
j
k
] = δ(
k −
k
)δ
ij
However rather than treating the i ﬁelds separately, we can combine them
as we did with the wave functions φ(x). Deﬁne [Kaku 70; Huang 25]
a
k
≡
1
√
2
(a
1
k
+ia
2
k
) and b
k
≡
1
√
2
(a
1
k
−ia
2
k
)
so that the Fourier expansion now becomes [Huang 25; Schwabl 285]
φ(x) =
d
˜
k (a
k
e
−ik·x
+b
†
k
e
ik·x
)
(Exercise: Prove this result) and for the other ﬁeld, obviously it is
φ
†
(x) =
d
˜
k(a
†
k
e
ik·x
+b
k
e
−ik·x
)
The new commutation relations are [Kaku 70; Schwabl 286; Huang 25]
[a
k
, a
†
k
] = [b
k
, b
†
k
] = δ
k
k
and
[a
k
, a
k
] = [b
k
, b
k
] = [a
k
, b
k
] = [a
k
, b
†
k
] = 0
There are now two occupationnumber operators, for particles a and for
particles b [Schwabl 286]
N
a
k
≡ a
†
k
a
k
and N
b
k
≡ b
†
k
b
k
where a
†
k
and a
k
create and annihilate a particles and b
†
k
and b
k
create and
annihilate b particles. The vacuum state is deﬁned by [Schwabl 286]
a
k
[0` = b
k
[0` = 0
The four momentum is [Schwabl 286; Mandl and Shaw 49; Greiner FQ 93]
P
µ
= (H,
P) =
¸
k
k
µ
(N
a
k
+N
b
k
)
giving
H =
¸
k
ω
k
(N
a
k
+N
b
k
)
and
P =
¸
k
k(N
a
k
+N
b
k
)
68 CHAPTER 3. FREE KLEINGORDON FIELD
3.7.1 Charge and Complex Scalar Field
Recall the real scalar ﬁeld Lagrangian L =
1
2
(∂
µ
φ∂
µ
φ − m
2
φ
2
) and the
complex scalar ﬁeld Lagrangian
L = ∂
µ
φ
†
∂
µ
φ −m
2
[φ[
2
=
1
2
2
¸
i=1
(∂
µ
φ
i
∂
µ
φ
i
−m
2
φ
2
i
)
with [φ[
2
≡ φφ
†
. The complex scalar ﬁeld Lagrangian (and action) is invari
ant under the transformation
φ(x) →e
iqθ
φ(x) φ
†
(x) →e
−iqθ
φ
†
(x)
which generates a U(1) symmetry [Kaku 70]. Recall that before we had (for
small α
i
=
i
)
η
r
(x) →η
r
(x) = e
iα
i
X
i
η
r
(x) ≈ (1 +i
i
X
i
)η
r
(x)
giving
δη
r
(x) = η
r
(x) −η
r
(x) = i
i
X
i
η
r
(x)
Our two ﬁelds are η
1
≡ φ and η
2
≡ φ
†
giving (with small θ = )
δφ(x) = iqφ(x) and δφ
†
(x) = −iqφ
†
(x)
Recall the Noether current
j
µ
≡
∂L
∂(∂
µ
η
r
)
∆η
r
−T
µν
δx
ν
with
T
µν
≡
∂L
∂(∂
µ
η
r
)
∂
ν
η
r
−g
µν
L
Also recall the relation between local and total variations
δη
r
(x) = ∆η
r
(x) −
∂η
r
∂x
µ
δx
µ
For δx
ν
= 0, we have δη(x) = ∆η(x) [Schwabl 272] and therefore
j
µ
=
∂L
∂(∂
µ
η
r
)
δη
r
=
∂L
∂(∂
µ
φ)
δφ +
∂L
∂(∂
µ
φ
†
)
δφ
†
=
∂L
∂(∂
µ
φ)
iqφ +
∂L
∂(∂
µ
φ
†
)
(−i)qφ
†
3.7. COMPLEX KLEINGORDON FIELD 69
and with L = ∂
µ
φ
†
∂
µ
φ −m
2
φφ
†
= g
µν
∂
ν
φ
†
∂
µ
φ −m
2
φφ
†
gives
∂L
∂(∂
µ
φ)
= g
µν
∂
ν
φ
†
= ∂
µ
φ
†
and
∂L
∂(∂
µ
φ
†
)
= ∂
µ
φ
Thus the Noether current is
j
µ
= (∂
µ
φ
†
)iqφ −(∂
µ
φ)iqφ
†
Once again we drop the constant factor and insert a minus sign to
deﬁne a new current [Kaku 71, Mosel 19, Schwabl 286, Huang 25]
j
µ
= iq[(∂
µ
φ)φ
†
−(∂
µ
φ
†
)φ]
which agrees exactly with the 4current derived previously!
The conserved charge (
dQ
dt
= 0) is therefore [Kaku 71, Mosel 19]
Q =
d
3
xj
o
= iq
d
3
x(
˙
φφ
†
−
˙
φ
†
φ)
= q
d
3
k(a
†
k
a
k
−b
†
k
b
k
)
= q
d
3
k(N
a
k
−N
b
k
)
(Exercise: Prove this result). Here j
o
matches exactly our expression for ρ
derived previously! [Kaku 71] gives an excellent discussion of the physical
interpretation of this charge. Also note that “charge” is used in a generic
sense [Huang 26] since electromagnetic coupling has not yet been turned on.
Compare this to our result from Chapter 1 which was (with probability
current density j
µ
= (ρ,
j ) =
¯
ψγ
µ
ψ )
Q =
d
3
xj
0
= q
d
3
x
¯
ψ γ
0
ψ
= q
d
3
xρ
where the probability density was ρ =
¯
ψγ
0
ψ = ψ
†
γ
0
γ
0
ψ = ψ
†
ψ. This was an
expression in terms of the classical Dirac ﬁeld ψ. Above we now have Q in
terms of the quantized ﬁeld operators a, a
†
, b, b
†
for the complex scalar ﬁeld.
70 CHAPTER 3. FREE KLEINGORDON FIELD
3.8 Summary
Two useful integrals are:
1
(2π)
3
d
3
xe
±i(
k−
k
).x
= δ(
k −
k
)
δ
4
(x −y) =
d
4
k
(2π)
4
e
−ik·(x−y)
3.8.1 KG classical ﬁeld
The massive KleinGordon Lagrangian is
L
KG
=
1
2
(∂
µ
φ∂
µ
φ −m
2
φ
2
)
giving the KGE (using p
2
→−P
2
)
(P
2
+m
2
)φ = 0
(p
2
−m
2
)φ = 0
The covariant momentum density is
Π
µ
= ∂
µ
φ
giving the canonical momentum
Π ≡ Π
o
=
˙
φ(x)
and Hamiltonian density
H =
1
2
[Π
2
+ (
∇φ)
2
+m
2
φ
2
]
and momentum
P = −
d
3
x
˙
φ(x)
∇φ(x)
The Fourier expansion of the KG ﬁeld is
φ(x, t) ≡ φ
+
(x) +φ
−
(x) =
d
˜
k
a
k
e
−ik·x
+a
†
k
e
ik·x
The conjugate momentum is
Π(x, t) =
˙
φ(x, t) ≡ Π
+
(x) + Π
−
(x)
= −i
d
˜
k ω
a
k
e
−ik·x
−a
†
k
e
ik·x
3.8. SUMMARY 71
In the above equations (with k
0
≡ ω = +
k
2
+m
2
)
d
˜
k ≡
d
3
k
2ω(2π)
3
=
2ω
(2π)
3
d
4
kδ(k
2
−m
2
) θ(k
o
)
The integrals are cast into discrete form with the replacement
d
3
k →Σ
k
.
3.8.2 KleinGordon Quantum ﬁeld
To develop the QFT we impose the equal time commutation relations
[φ(x, t), Π(x
, t)] = iδ(x −x
)
and
[φ(x, t), φ(x
, t)] = [Π(x, t), Π(x
, t)] = 0
which imply
[a(
k), a
†
(
k
)] = δ(
k −
k
)
and
[a(
k), a(
k
)] = [a
†
(
k), a
†
(
k
)] = 0
Deﬁning
N
k
≡ a
†
k
a
k
the Hamiltonian and momentum can be rewritten as
H =
d
3
k
N
k
+
1
2
ω
P =
d
3
k
N
k
+
1
2
k
The vacuum state is deﬁned via
a
k
[0` = 0
and the many body states are
[ n
k
i
n
k
j
` =
¸
i
[n
k
i
`
with
[n
k
i
` =
1
n
k
!
a
†
k
n
k
[0`
H[ n
k
i
n
k
j
` =
¸
k
n
k
(
k)[ n
k
i
n
k
j
`
where (
k) = ¯ hω(
k).
72 CHAPTER 3. FREE KLEINGORDON FIELD
3.8.3 Propagator Theory
The nonfree KGE is
(P
2
+m
2
)φ = J(x)
The Feynman propagator in position space is deﬁned as
(P
2
+m
2
)∆
F
(x −y) ≡ −δ
4
(x −y)
so that the solution to the nonfree KGE is
φ(x) = φ
0
(x) −
d
4
y ∆
F
(x −y)J(y)
where φ
0
(x) is the solution with J = 0. The momentum space Feynman
propagator is deﬁned through the Fourier transform
∆
F
(x −y) ≡
d
4
k
(2π)
4
e
−ik·(x−y)
∆
F
(k)
The usual method of solution for propagators or Green functions is to solve
for ∆
F
(k) and then do a contour integral to get ∆
F
(x − y) rather than
solving for ∆
F
(x −y) directly. We get
∆
F
(k) =
1
k
2
−m
2
which is the inverse of the momentum operator in the free KGE. This ﬁnally
leads to
∆
F
(x −x
) = −iθ(t −t
)
d
˜
k e
−ik·(x−x
)
−iθ(t
−t)
d
˜
k e
ik·(x−x
)
One can also show that
i∆
F
(x −x
) = '0[Tφ(x)φ(x
)[0`
= '0[Tφ(x)φ
†
(x
)[0`
where the second line is true for Hermitian ﬁelds [BjRQF42].
Other useful propagators are
i∆
±
(x −y) ≡ [φ
±
(x), φ
∓
(y)] = ±
d
˜
˜
ke
∓ik·(x−y)
giving
i∆(x −y) ≡ [φ(x), φ(y)] = −i∆
+
(x −y) +i∆
−
(x −y)
3.9. REFERENCES AND NOTES 73
with
∆(x) = ∆
+
(x) + ∆
−
(x) = −
d
3
k
(2π)
3
ω
k
sin k x
These other propagators are related to the Feynman propagator by
∆
F
(x) = θ(t)∆
+
(x) −θ(−t)∆
−
(x)
or
∆
F
(x) = ±∆
±
(x) for t
>
<
0
3.8.4 Complex KG ﬁeld
Not summarized; see text.
3.9 References and Notes
Good references for this chapter are [Teller, Bergstrom and Goobar, Roman,
Halzen and Martin, Muirhead, Leon, Goldstein].
74 CHAPTER 3. FREE KLEINGORDON FIELD
Chapter 4
Dirac Field
We have seen that the KGE gives rise to negative energies and nonpositive
deﬁnite probabilities and for these reasons was discarded as a fundamental
quantum equation. These problems arise because the KGE is nonlinear in
∂
∂t
, unlike the SE. Dirac thus sought a relativistic quantum equation linear
in
∂
∂t
, like the SE. We shall see that therefore Dirac was forced to invent a
matrix equation.
The ‘derivation’ of the DE presented here follows [Griﬃths, pg. 215].
Dirac was searching for an equation linear in E or
∂
∂t
. Instead of starting
from
p
2
−m
2
= 0
his strategy was to factor this relation, which is easy if we only have p
0
(i.e.
p = 0), namely (with p
0
= p
0
)
(p
0
−m)(p
0
+m) = 0
and obtain two ﬁrst order equations
p
0
−m = 0 or p
0
+m = 0
However it’s more diﬃcult if p is included. Then we are looking for some
thing of the form p
2
−m
2
= p
µ
p
µ
−m
2
= (β
µ
p
µ
+m)(γ
ν
p
ν
−m) where β
µ
and γ
ν
are 8 coeﬃcients to be determined. The RHS is
β
µ
γ
ν
p
µ
p
ν
+m(γ
ν
−β
ν
)p
ν
−m
2
For equality with LHS we don’t want terms linear in p
ν
, thus γ
ν
= β
ν
,
leaving
p
2
= p
µ
p
µ
= γ
µ
γ
ν
p
µ
p
ν
75
76 CHAPTER 4. DIRAC FIELD
i.e. (with (p
1
)
2
= (p
1
)
2
)
(p
0
)
2
−(p
1
)
2
−(p
2
)
2
−(p
3
)
2
= (γ
0
)
2
(p
0
)
2
+ (γ
1
)
2
(p
1
)
2
+ (γ
2
)
2
(p
2
)
2
+(γ
3
)
2
(p
3
)
2
+ (γ
0
γ
1
+γ
1
γ
0
)p
0
p
1
+(γ
0
γ
2
+γ
2
γ
0
)p
0
p
2
+ (γ
0
γ
3
+γ
3
γ
0
)p
0
p
3
+(γ
1
γ
2
+γ
2
γ
1
)p
1
p
2
+ (γ
1
γ
3
+γ
3
γ
1
)p
1
p
3
+(γ
2
γ
3
+γ
3
γ
2
)p
2
p
3
Here’s the problem; we could pick γ
0
= 1 and γ
1
= γ
2
= γ
3
= i but
we can’t get rid of the ‘cross terms’. At this point Dirac had a brilliant
inspiration; what if the γ’s are matrices instead of numbers? Since matrices
don’t commute we might be able to ﬁnd a set such that
(γ
0
)
2
= 1 (γ
1
)
2
= (γ
2
)
2
= (γ
3
)
2
= −1
γ
µ
γ
ν
+γ
ν
γ
µ
= 0 for µ = ν
Or, more succinctly, ¦γ
µ
, γ
ν
¦ = 2g
µν
which is called a Cliﬀord Algebra
where the curly brackets denote the anticommutator ¦A, B¦ ≡ AB + BA.
It turns out that this can be done, but the smallest set of matrices are 44.
These are
γ
0
≡ β =
I 0
0 −I
γ ≡ β α ≡
0 σ
−σ 0
γ
µ
≡ (β, β α)
Let’s introduce everything else for completeness
γ
5
≡ iγ
0
γ
1
γ
2
γ
3
=
0 I
I 0
σ
1
=
0 1
1 0
σ
2
=
0 −i
i 0
σ
3
=
1 0
0 −1
α ≡
0 σ
σ 0
β ≡
1 0
0 −1
γ
µ
†
= γ
0
γ
µ
γ
0
i.e. γ
0
†
= γ
0
, γ
i
†
= −γ
i
β
−1
= β and γ
5
= γ
5
These conventions are used by the following authors: [Griﬃths, Kaku,
Halzen & Martin].
Thus the Dirac equation is (p −m)ψ = 0 with A ≡ γ
µ
A
µ
. In coordi
nate space (using p
µ
→i∂
µ
) it is (i∂ −m)ψ = 0 . In noncovariant notation
it is
Hψ = i¯h
∂ψ
∂t
with H ≡ α p +βm
4.1. PROBABILITY & CURRENT 77
The DE can be written in terms of 4vectors (/ p −m)ψ = 0 and is therefore
manifestly covariant.
4.1 Probability & Current
For the SE and KGE we used SE
∗
and KGE
∗
to derive the continuity equa
tion. For matrices the generalization of complex conjugate (∗) is Hermitian
conjugate (†) which is the transpose of the complex conjugate. The DE is
(i/ ∂ −m)ψ = 0 = (iγ
µ
∂
µ
−m)ψ
and DE
†
is (using (AB)
†
= B
†
A
†
)
ψ
†
(i/ ∂ −m)
†
= 0
= ψ
†
(−iγ
µ†
∂
µ
−m) = 0 where ∂
†
µ
= ∂
µ
= ψ
†
(−iγ
0
γ
µ
γ
0
∂
µ
−m) = 0 using γ
µ†
= γ
0
γ
µ
γ
0
We want to introduce the Dirac adjoint (ψ is a column matrix)
¯
ψ ≡ ψ
†
γ
0
(
¯
ψ is a row matrix!)
Using γ
0
γ
0
= 1 we get
ψ
†
(−iγ
0
γ
µ
γ
0
∂
µ
−mγ
0
γ
0
) = 0
⇒
¯
ψ(iγ
µ
γ
0
∂
µ
+mγ
0
) = 0
and cancelling out γ
0
gives, the Dirac adjoint equation,
¯
ψ(i/ ∂ +m) = 0 ⇔
¯
ψ(i
←
/ ∂ +m) = 0
Some sloppy authors write this as (i/ ∂ +m)
¯
ψ = 0 but this cannot be because
¯
ψ is a row matrix! The notation
←
/ ∂ however means that / ∂ operates on
¯
ψ to
the left, i.e.
¯
ψ
←
/ ∂ ≡ (∂
µ
¯
ψ)γ
µ
.
It’s very important not to get confused with this. The DE and DE
†
are
explicitly
(i/ ∂ −m)ψ = 0 ⇔ iγ
µ
∂
µ
ψ −mψ = 0
¯
ψ(i/ ∂ +m) = 0 ⇔ i(∂
µ
¯
ψ)γ
µ
+m
¯
ψ = 0
78 CHAPTER 4. DIRAC FIELD
Now let’s derive the continuity equation. Multiply DE from the left by
¯
ψ
and DE
†
from the right by ψ. Now one could get very confused writing
¯
ψ(i/ ∂ −m)ψ = 0
¯
ψ(i/ ∂ +m)ψ = 0
whereas what is really meant is [Halzen & Martin, pg. 103]
¯
ψ(iγ
µ
∂
µ
ψ −mψ) = 0
(i(∂
µ
¯
ψ)γ
µ
+m
¯
ψ)ψ = 0
Adding these gives
¯
ψγ
µ
∂
µ
ψ + (∂
µ
¯
ψ)γ
µ
ψ = 0 = ∂
µ
(
¯
ψγ
µ
ψ)
giving
j
µ
= (
¯
ψγ
µ
ψ) ≡ (ρ,
j)
or
ρ =
¯
ψγ
0
ψ = ψ
†
ψ =
4
¸
i=1
[ψ
i
[
2
which is now positive deﬁnite! The 3current is
j =
¯
ψγψ
[see also Mosel 17, 34]
4.2 Bilinear Covariants
ψ can be written
¸
¸
¸
¸
ψ
1
ψ
2
ψ
3
ψ
4
¸
and one can try to construct a scalar, such as
ψ
†
ψ = (ψ
∗
1
ψ
∗
2
ψ
∗
3
ψ
∗
4
)
¸
¸
¸
¸
ψ
1
ψ
2
ψ
3
ψ
4
¸
= [ψ
1
[
2
+[ψ
2
[
2
+[ψ
3
[
2
+[ψ
4
[
2
but this is not
a Lorentz scalar (it’s got all + signs).
Rather, deﬁne the Dirac adjoint
¯
ψ ≡ ψ
†
γ
0
= (ψ
∗
1
ψ
∗
2
−ψ
∗
3
−ψ
∗
4
)
4.3. NEGATIVE ENERGY AND ANTIPARTICLES 79
and
¯
ψψ = [ψ
1
[
2
+[ψ
2
[
2
−[ψ
3
[
2
−[ψ
4
[
2
is a Lorentz scalar (see Bjorken and
Drell]. One can prove that the following quantities transform as indicated:
¯
ψψ scalar (1 component)
¯
ψγ
5
ψ pseudoscalar (1 component)
¯
ψγ
µ
ψ vector (4 component)
¯
ψγ
µ
γ
5
ψ pseudovector (4 component)
¯
ψσ
µν
ψ antisymmetric 2nd rank tensor (6 component)
where σ
µν
≡
i
2
(γ
µ
γ
ν
−γ
ν
γ
µ
).
4.3 Negative Energy and Antiparticles
4.3.1 Schrodinger Equation
The free particle SE is
−
¯h
2
2m
∇
2
ψ = i¯h
∂ψ
∂t
which has solution
ψ(x, t) = (C cos kx +Dsin kx) e
i
¯ h
Et
≡ ψ
E
=
Ae
ikx
+Be
−ikx
e
i
¯ h
Et
in 1dimension.
Substituting gives
¯hk
2
2m
ψ(x, t) = −Eψ(x, t) or
E +
¯h
2
k
2
2m
ψ = 0
yielding
E = −
¯h
2
k
2
2m
However it also has solution ψ(x, t) = (Ae
ikx
+Be
−ikx
)e
−
i
¯ h
Et
≡ ψ
−E
Substituting gives
¯h
2
k
2
2m
ψ(x, t) = +Eψ(x, t) or
E −
¯h
2
k
2
2m
ψ = 0
yielding
E = +
¯h
2
k
2
2m
80 CHAPTER 4. DIRAC FIELD
ψ
E
and ψ
−E
are diﬀerent solutions. The ﬁrst one ψ
E
corresponds to positive
energy and the second one ψ
−E
to negative energy. We are free to toss away
one solution as unphysical and only keep ψ
E
.
However for KGE and DE the same solution ψ gives both positive and
negative energy. Of course we are not free to toss away one energy because
it’s in the solution. Whereas in the SE we got two diﬀerent solutions for
positive and negative energy. If we want to get rid of negative energy in the
SE we toss away one solution. However in KGE and DE we always get both
positive and negative energy for all solutions. The only way to toss away
negative energy is to toss away all solutions; i.e. toss out the whole equation!
Also reason why Einstein did not reject SR; E = ±
p
2
+m
2
were
separate solutions; not part of same solution.
4.3.2 KleinGordon Equation
The free particle KGE is
(P
2
+m
2
)φ = 0
with
P
2
= −
∂
2
∂t
2
+∇
2
A solution is
φ = Ne
ip·x
= Ne
i(Et− p·x)
Substituting gives
(E
2
− p
2
+m
2
)φ = 0
which implies
⇒E
2
= p
2
+m
2
or
E = ±
p
2
+m
2
Thus the single solution φ = Ne
ip·x
has both positive and negative energy
solutions. Another solution is
φ = Ne
−ip·x
= Ne
−i(Et− p·x)
Substituting also gives the same as above, namely (E
2
− p
2
+m
2
)φ = 0 or
E = ±
p
2
+m
2
4.3. NEGATIVE ENERGY AND ANTIPARTICLES 81
and again the single solution φ = Ne
−ip·x
has both positive and negative
energy solutions.
The interpretation of these states is as follows. For p = 0 (particle at rest)
then E = ±m. For p = 0, there will be a continuum of states above and
below E = ±m [Landau, pg. 225], with bound states appearing in between.
In QM there would be transitions to the negative energy continuum to inﬁ
nite negative energy. This also happens with the Dirac equation. However,
the DE describes, automatically, particles with spin. Dirac’s way out of the
negative energy catastrophe was to postulate that the negative energy sea
was ﬁlled with fermions and so the ???
E < 0
E > 0
E
 mc
0
p = 0
2
mc
2
p = 0
However for the KGE the negative energies are a catastrophe.
Also we can now clearly see the problem with ρ as calculated with the
KGE. Recall ρ =
i¯h
2m
φ
∗ ∂φ
∂t
−φ
∂φ
∗
∂t
. For φ = e
±ip·x
we get ρ = ∓
¯h
m
φ
∗
φE
which gives negative ρ for positive or negative E. (Halzen & Martin, pg. 74]
82 CHAPTER 4. DIRAC FIELD
4.3.3 Dirac Equation
Let’s look for plane wave solutions of the form ψ(x) = w( p)e
−ip·x
. Substi
tuting into (i/ ∂ − m)ψ = 0 gives the momentum space DE (/ p −m)w = 0 .
Actually, let’s ﬁrst look at rest frame ( p = 0) solutions (RF). Writing the
DE as
Hψ = i¯h
∂ψ
∂t
with H = α p +βm
independent DE is Hψ = Eψ, which in the RF is
Hw = βmw =
mI 0
0 −mI
w = Ew,
The eigenvalues are E = m, m, −m, −m with eigenvectors [Halzen & Martin,
pg. 104]
¸
¸
¸
1
0
0
0
¸
¸
¸
¸
0
1
0
0
¸
¸
¸
¸
0
0
1
0
¸
¸
¸
¸
0
0
0
1
¸
Thus the DE has positive (E = +m) and negative (E = −m) energy solu
tions! (We shall look at p = 0 solutions in a moment)
Let’s summarize so far (Aitchison & Hey, pg. 71)
Probability (ρ) Energies for same solution ψ
SE + +
KGE − +, −
DE + +, −
Thus both the KGE and DE have negative energies.
The DE describes fermions (see next section). Dirac’s idea was that
the negative energy sea was ﬁlled with fermions and via PEP prevented the
negative energy cascade of positive energy particles. Dirac realized if given
particles in sea energy of 2mc
2
create holes. Dirac Sea not taken seriously
until positron discovered!! (1932 Anderson)
4.4. FREE PARTICLE SOLUTIONS OF DIRAC EQUATION 83
4.4 Free Particle Solutions of Dirac Equation
Before proceeding recall the following results:
σ
x
=
0 1
1 0
σ
y
=
0 −i
i 0
σ
z
=
1 0
0 −1
σ p =
P
z
p
−
p
+
−p
z
≡
p
z
p
x
−ip
y
p
x
+ip
y
−p
z
Let’s now look at p = 0 solutions (i.e. not in rest frame). As before, we look
for solutions of the form ψ(x) = w( p)e
−ip·x
Substituting into (i/ ∂ −m)ψ = 0
gives the momentum space DE (/ p − m)w = 0. Using / p =
E −σ p
σ p −E
gives, with w ≡
w
A
w
B
(/ p −m)w =
E −m −σ p
σ p −E −m
w
A
w
B
= 0 =
¸
(E −m) w
A
−o p w
B
σ p w
A
−(E +m) w
B
¸
giving
w
A
=
σ p
E −m
w
B
and w
B
=
σ p
E +m
w
A
Combining yields
w
A
=
(σ p )
2
E
2
−m
2
w
A
=
p
2
E
2
−m
2
w
A
because (σ p)
2
= p
2
. Thus
E
2
= p
2
+m
2
or E = ±
p
2
+m
2
Thus again we see the negative energy solutions ! (this time for the p = 0
DE)
Our ﬁnal solutions are
w =
w
A
w
B
=
w
A
σ· p
E+m
w
A
or
σ· p
E−m
w
B
w
B
with w
A
and w
B
left unspeciﬁed, which means we are free to choose them
[Aitchison & Hey 69] as
w
A
, w
B
=
1
0
or
0
1
84 CHAPTER 4. DIRAC FIELD
We have the following possibilities
Pick w
A
=
1
0
⇒w
B
=
o p
E +m
1
0
=
1
E +m
p
z
p
+
(1) p
±
≡ p
x
±ip
y
Pick w
A
=
0
1
⇒w
B
=
o p
E +m
0
1
=
1
E +m
p
−
−p
z
(2)
Pick w
B
=
1
0
⇒w
A
=
o p
E −m
1
0
=
1
E −m
p
z
p
+
(3)
Pick w
B
=
0
1
⇒w
A
=
o p
E −m
0
1
=
1
E −m
p
−
−p
z
(4)
But the question is, do we use E = +
p
2
+m
2
or E = −
p
2
+m
2
? Well,
for (1) and (2) we must use E = +
√
otherwise
1
E+m
blows up for p = 0.
For (3) and (4) we must use E = −
√
otherwise
1
E−m
blows up for p = 0.
The term E = +
√
is called the particle solution. The term E = −
√
is
called the antiparticle solution.
Rewrite as, with χ
(1)
≡
1
0
and χ
(2)
≡
0
1
w
(1)
( p) = N
¸
¸
¸
1
0
pz
E+m
p
+
E+m
¸
= N
1
σ· p
E+m
χ
(1)
w
(2)
( p) = N
¸
¸
¸
¸
0
1
p
−
E+m
−pz
E+m
¸
= N
1
o· p
E+m
χ
(2)
with E ≡ +
p
2
+m
2
and also
w
(3)
( p) = N
¸
¸
¸
pz
E−m
p
+
E−m
1
0
¸
= N
σ· p
E−m
1
χ
(1)
w
(4)
( p) = N
¸
¸
¸
¸
p
−
E−m
−pz
E−m
0
1
¸
= N
σ· p
E−m
1
χ
(2)
with E ≡ −
p
2
+m
2
4.4. FREE PARTICLE SOLUTIONS OF DIRAC EQUATION 85
or
w
(s)
( p) = N
1
o· p
E+m
χ
(s)
with E ≡ +
p
2
+m
2
w
(s+2)
( p) = N
σ· p
E−m
1
χ
(s)
with E ≡ −
p
2
+m
2
s = 1, 2
[Halzen & Martin, pg. 105]
But all free particles carry positive energy! Thus reinterpret w
(3)
and w
(4)
as positive energy antiparticle states
w
(s+2)
( p) = N
σ· p
−
√
p
2
+m
2
−m
1
χ
(s)
w
(s+2)
(− p) = N
−σ· p
−
√
p
2
+m
2
−m
1
χ
(s)
= N
σ· p
√
p
2
+m
2
+m
1
χ
(s)
.
.. Deﬁne u(p, s) = u
(s)
(p) ≡ w
(s)
( p) u
(1,2)
(p) = w
(1,2)
( p)
v(p, s) = v
(s)
(p) ≡ w
(s+2)
(− p) v
(1,2)
(p) = w
(3,4)
(− p)
(Me & Kaku)
everyone else: v
(1,2)
(p) = w
(4,3)
(− p)
Note: Bj writes u(p, s) and w( p). Gross writes u( p, s)
u
(1)
= N
¸
¸
¸
1
0
pz
E+m
p
+
E+m
¸
u
(2)
= N
¸
¸
¸
¸
0
1
p
−
E+m
−pz
E+m
¸
v
(2)
= N
¸
¸
¸
pz
E+m
p
+
E+m
1
0
¸
v
(1)
= N
¸
¸
¸
¸
p
−
E+m
−pz
E+m
0
1
¸
all with E ≡ +
p
2
+m
2
= v
(1)
(Kaku) = v
(2)
(Kaku) Kaku diﬀerent
from everyone else
u = N
1
σ· p
E+m
χ v = N
σ· p
E+m
1
χ
86 CHAPTER 4. DIRAC FIELD
u
(1)
, v
(2)
→χ
(1)
=
1
0
u
(2)
, v
(1)
→χ
(2)
=
0
1
Kaku N =
E +m
2m
σ p =
p
z
p
−
p
+
−p
z
Halzen & Martin, pg. 107
ψ(x) = w( p)e
−ip·x
ψ(x) = w(− p)e
−i(−p·x)
⇒ψ(x) = u( p)e
−ip·x
ψ(x) = v(p)e
+ip·x
(/ p −m)w( p) = 0 (−/ p −m)w(− p) = 0
⇒ (/ p −m)u = 0 ⇒ (/ p +m)v = 0
The adjoints satisfy
¯ u(/ p −m) = 0 and ¯ v(/ p +m) = 0
Normalization [Muirhead, pg. 71ﬀ; Griﬃths, pg. 220 footnotes]
The most common conventions are, with E ≡ +
p
2
+m
2
everywhere be
low
Bjorken & Drell, Kaku, Greiner, Sterman
u
†
u =
E
m
⇒N =
E +m
2m
but spurious diﬃculties when m →0
Griﬃths, Halzen & Martin
u
†
u = 2E ⇒N =
√
E +m
Bogoliubov & Shirkov
u
†
u = 1
These oddlooking normalizations come by specifying ¯ uu. Using the result
¯ uγ
µ
u =
p
u
m
¯ uu (Problem 4.7) we have u
†
γ
0
γ
µ
u =
p
µ
m
¯ uu and with µ = 0 and
γ
0
γ
0
= 1 we get
u
†
u =
E
m
¯ uu
Thus ¯ uu = 1 ⇒u
†
u =
E
m
. Actualy the more general normalization, in this
case, is ¯ u
(r)
u
(s)
= δ
rs
(see Problem 4.8).
Alternatively we can specify ¯ vv = −1 , i.e. ¯ v
(r)
v
(s)
= −δ
rs
[Kaku, pg.
754]
4.5. CLASSICAL DIRAC FIELD 87
4.5 Classical Dirac Field
The Dirac Lagrangian is [Mosel 34]
L
D
=
¯
ψ(i/ ∂ −m)ψ
from which one can obtain the Dirac equation and its adjoint (see Problems).
NNN  now handwriting
4.5.1 Noether spacetime current
4.5.2 Noether internal symmetry and charge
4.5.3 Fourier expansion and momentum space
References: [Mosel 35, Greiner FQ 123, Peskin 52, Huang 123, Schwabl 290,
Kaku 86]
We now wish to expand the Dirac ﬁeld in terms of creation and anni
hilation operators. Most books just write down the answer (as we shall)
but [Greiner FQ 123] derives the result very clearly using the same method
that we used for the KleinGordon ﬁeld where one proves that the expansion
must contain two terms.
The best discussion as to why the Dirac creation and annihilation op
erators must obey anticommutation relations is given in [Peskin 5256] and
[Greiner FQ 129].
Because of the normalization of our Dirac spinors (see previous chapter)
we will have a diﬀerent normalization constant in our Fourier expansion of
the Dirac ﬁeld, as compared to the KG case. Because of this the measure
for fermions will be [IZ 114, 147, 703]
NNN  now handwriting
88 CHAPTER 4. DIRAC FIELD
4.6 Dirac QFT
4.6.1 Derivation of b, b
†
, d, d
†
Anticommutators
4.7 Pauli Exclusion Principle
4.8 Hamiltonian, Momentum and Charge in terms
of creation and annihilation operators
4.8.1 Hamiltonian
4.8.2 Momentum
4.8.3 Angular Momentum
4.8.4 Charge
4.9 Propagator theory
4.10 Summary
4.10.1 Dirac equation summary
4.10.2 Classical Dirac ﬁeld
4.10.3 Dirac QFT
4.10.4 Propagator theory
4.11 References and Notes
Mandl & Shaw, Teller, Sakurai QM, Leon, Merzbacher
S matrix and G function Bj RQM 83,97,100
S matrix without SE, Bj RQF 177
Chapter 5
Electromagnetic Field
5.1 Review of Classical Electrodynamics
5.1.1 Maxwell equations in tensor notation
5.1.2 Gauge theory
5.1.3 Coulomb Gauge
5.1.4 Lagrangian for EM ﬁeld
5.1.5 Polarization vectors
References: [Weidner and Sells, 1st ed.,p.975; GriﬃthsEM, 1st ed., p. 350;
Jackson, 2nd ed. p. 274; Schwabl 313; Mandl and Shaw 129; Guidry 86,87;
GreinerFQ 161, 177]. Note that Guidry and Schwabl are excellent.
In general transverse waves can be linearly or circularly polarized. A
nice elementary discussion can be found in [Weidner and Sells, 1st ed.,p.975;
GriﬃthsEM, 1st ed., p. 350]. A circularly polarized wave can be made out
of two linearly polarized waves if the two linear waves are out of phase. (If
they are in phase then they combine to form a linearly polarized wave at a
diﬀerent angle.) To produce a linearyly polarized wave on a rope just jiggle
the rope up and down for a linearly polarized wave in the vertical direc
tion. The circularly polarized wave can be produced by moving one’s hand,
which is holding the rope, in a circular fashion. One can rotate clockwise or
counterclockwise to produce circularly polarized states of opposie helicity.
An elliptically polarized wave is formed from two linearly polarized waves
out of phase but with each linearly polarized wave having a diﬀernt ampli
tude. The amplitudes are the same for a circular wave.
89
90 CHAPTER 5. ELECTROMAGNETIC FIELD
Just as the elliptically or circularly polarized wave can be created from
two linear polarized waves (linear combination of them), similarly a linearly
polarized wave can be built from a linear combination of two circularly
polarized wave. Thus any wave in general can be represented in terms of
linearly polarized basis states or circularly polarized basis states. This is
discussed from a mathematical point of view in [Jackson, 2nd ed. p. 274].
Even though it should be obvious, note that the treatment of polarization
vectors from the point of view of classical electrodynamics [Jackson] is the
same as the quantum ﬁeld treatments. That is, the polarization vectors used
are the same in the classical and quantum case.
In general the linear polarization basis states are described by real po
larization vectors
1
and v
2
. (As we shall see below there are only two,
not four, possible states for photons.) The circularly polarized states are
described by complex basis vectors, say written as
+
and
−
.
These states are written in terms of each other
±
=
1
√
2
(
1
±i
2
)
The circularly polarized states are complex, which is an easy way of
allowing for a phase diﬀerence between the two linearly polarized states out
of which the circular states are constructed. [Jackson, 2nd ed. p. 274]
Finally, in describing our quantum ﬁeld states for the photon it does not
matter whether we use the linearly polarized basis states or the circularly
polarized ones. Diﬀerent authors choose diﬀerent basis states. We will know
what states a particlar author is using by simply recognizing wether the states
are real (linear polarization) or complex (circular polarization).
5.2. QUANTIZED MAXWELL FIELD 91
5.1.6 Linear polarization vectors in Coulomb gauge
5.1.7 Circular polarization vectors
5.1.8 Fourier expansion
5.2 Quantized Maxwell ﬁeld
5.2.1 Creation & annihilation operators
5.3 Photon propagator
5.4 GuptaBleuler quantization
5.5 Proca ﬁeld
92 CHAPTER 5. ELECTROMAGNETIC FIELD
Chapter 6
Smatrix, cross section &
Wick’s theorem
Our Feynman diagram series is going to come from an expansion of the so
called Scattering Matrix (or Smatrix). Each term in this Smatrix expan
sion will represent a particular Feynman diagram. The Smatrix expansion
looks very similar to something which occurs in nonrelativistic quantum
mechanics, namely the expansion of the time evolution operator. Thus we
will study that ﬁrst.
6.1 Schrodinger Time Evolution Operator
Recall the timedependent Schrodinger equation (SE)
H[α(t)` = i¯h
∂
∂t
[α(t)`
Deﬁne the time evolution operator [Mandl & Shaw, pg. 101, Leon, pg. 63,
Sakurai]
[α(t)` ≡ (t, t
0
)[α(t
0
)`
and upon substitution into the SE gives
H(t, t
0
)[α(t
0
)` = i¯h
∂
∂t
(t, t
0
)[α(t
0
)`
but
∂
∂t
does not operate on [α(t
0
)` because t
0
is ﬁxed. Thus we can ‘cancel’
[α(t
0
)` to arrive at the operator equation
H(t, t
0
) = i¯h
∂
∂t
(t, t
0
)
93
94 CHAPTER 6. SMATRIX, CROSS SECTION & WICK’S THEOREM
A schematic solution is
(t, t
0
) = e
−
i
¯ h
H(t−t
0
)
However let’s solve the equation more rigorously. We solve for (t, t
0
) by
integrating both sides to give
(t, t
0
) −(t
0
, t
0
) =
1
i¯h
t
t
0
H(t
1
)(t
1
, t
0
)dt
1
and with the boundary condition
(t
0
, t
0
) = 1
we have
(t, t
0
) = 1 −
i
¯h
t
t
0
dt
1
H(t
1
)(t
1
, t
0
)
This is solved by iteration as follows
(t, t
0
) = 1 −
i
¯h
t
t
0
dt
1
H(t
1
)(t
1
, t
0
)
= 1 −
i
¯h
t
t
0
dt
1
H(t
1
)
¸
1 −
i
¯h
t
1
t
0
dt
2
H(t
2
)(t
2
, t
0
)
= 1 +
−i
¯h
t
t
0
dt
1
H(t
1
) +
−i
¯h
2
t
t
0
dt
1
t
1
t
0
dt
2
H(t
1
)H(t
2
)(t
2
, t
0
)
= 1 +
−i
¯h
t
t
0
dt
1
H(t
1
) +
−i
¯h
2
t
t
0
dt
1
t
1
t
0
dt
2
H(t
1
)H(t
2
)
¸
1 −
i
¯h
t
2
t
0
dt
3
H(t
3
)(t
3
, t
0
)
= 1 +
−i
¯h
t
t
0
dt
1
H(t
1
) +
−i
¯h
2
t
t
0
dt
1
t
1
t
0
dt
2
H(t
1
)H(t
2
)
+
−i
¯h
3
t
t
0
dt
1
t
1
t
0
dt
2
t
2
t
0
dt
3
H(t
1
)H(t
2
)H(t
3
) +
or
(t, t
0
) = 1 +
∞
¸
n=1
−i
¯h
n
t
t
0
dt
1
t
1
t
0
dt
2
t
n−1
t
0
dt
n
H(t
1
) H(t
n
)
6.1. SCHRODINGER TIME EVOLUTION OPERATOR 95
which is the ﬁrst version of our expansion of the time evolution operator.
Two other ways of writing this are
(t, t
0
) = 1 +
∞
¸
n=1
1
n!
−i
¯h
n
t
t
0
dt
1
t
t
0
dt
2
t
t
0
dt
n
T[H(t
1
) H(t
n
)]
≡ T e
−i/¯h
t
t
0
H(t
)dt
[Leon, pg. 64; Merzbacher, pg. 475] where the last expression is just a formal
way of writing the formula above [Leon, pg. 64] and recall
e
x
= 1 +x +
x
2
2!
+
x
3
3!
+ = 1 +
∞
¸
n=1
1
n!
x
n
. Hatﬁeld, pg. 5253 gives a good example as to why we introduce time
ordered product. See also Guidry, pg. 99.
6.1.1 Time Ordered Product
In the above formulae we introduced the time ordered product which we
now deﬁne for two terms
T[A(t
1
)B(t
2
)]
(2! combinations)
≡
A(t
1
)B(t
2
) if t
1
> t
2
B(t
2
)A(t
1
) if t
2
> t
1
≡ θ(t
1
−t
2
)A(t
1
)B(t
2
) +θ(t
2
−t
1
)B(t
2
)A(t
1
)
where
θ(x −y) ≡
1 if x > y
0 if x < y
The time ordered product of three terms is deﬁned as
T[A(t
1
)B(t
2
)C(t
3
)]
(3! combinations)
≡
A(t
1
)B(t
2
)C(t
3
) if t
1
> t
2
> t
3
A(t
1
)C(t
3
)B(t
2
) if t
1
> t
3
> t
2
B(t
2
)A(t
1
)C(t
3
) if t
2
> t
1
> t
3
B(t
2
)C(t
3
)A(t
1
) if t
2
> t
3
> t
1
C(t
3
)A(t
1
)B(t
2
) if t
3
> t
1
> t
2
C(t
3
)B(t
2
)A(t
1
) if t
3
> t
2
> t
1
≡ θ(t
1
−t
2
)θ(t
2
−t
3
)A(t
1
)B(t
2
)C(t
2
) +θ(t
1
−t
3
)θ(t
3
−t
2
)A(t
1
)C(t
3
)B(t
2
)
+θ(t
2
−t
1
)θ(t
1
−t
3
)B(t
2
)A(t
1
)C(t
3
) +θ(t
2
−t
3
)θ(t
3
−t
1
)B(t
2
)C(t
3
)A(t
1
)
+θ(t
3
−t
1
)θ(t
1
−t
2
)C(t
3
)A(t
1
)B(t
2
) +θ(t
3
−t
2
)θ(t
2
−t
1
)C(t
3
)B(t
2
)A(t
1
)
96 CHAPTER 6. SMATRIX, CROSS SECTION & WICK’S THEOREM
Notice that there are n! combinations for n operators. This explains the
1
n!
term appearing in the expansion for .
[do Problem 8.1]
6.2 Schrodinger, Heisenberg and Dirac (Interac
tion) Pictures
Reference: [Mandl and Shaw, pg. 22]
One can do classical mechanics with either the Newtonian, Lagrange or
Hamilton formulation of mechanics. In each formulation the equations are
diﬀerent. Although not exactly analogous, there are three popular ways to
work in quantum mechanics known as the Schrodinger, Heisenberg or Dirac
pictures. The Dirac picture is often also called the Interaction picture.
In the usual formulation of quantum mechanics via the Schrodinger equa
tion, i.e. in the Schrodinger picture, the operators are frozen in time and the
states evolve in time. The opposite is true in the Heisenberg picture.
In the Interaction picture, which we shall use extensively, the Hamilto
nian is split into a fre particle piece and an interaction piece
H ≡ H
0
+H
I
then H
I
evolves and H
0
is frozen.
Picture Operators State Vectors
Schrodinger Frozen Evolve
Heisenberg Evolve Frozen
Dirac (Interaction) Evolve Evolve
In time dependent perturbation theory one considers potentials like  =
E
0
cos ωt (e.g. oscillating electromagnetic ﬁeld) in the Schrodinger picture
(SP).  is part of H and so what does it mean to say that operators are
frozen in the SP? What we mean is that the state vectors [α` obey an
equation of motion H[α` = i¯h
∂
∂t
[α` and the operators do not. Viceversa for
the Heisenberg picture.
We shall label Schrodinger, Heisenberg, Interaction picture states as
[α`
S
, [α`
H
, [α`
I
and operators as O
S
, O
H
, O
I
.
Recall the Schrodinger picture of NRQM
H[α(t)`
S
= i¯h
∂
∂t
[α(t)`
S
6.2. SCHRODINGER, HEISENBERGANDDIRAC (INTERACTION) PICTURES97
and the Schrodinger state vectors evolve in time according to
[α(t)`
S
≡ (t, t
0
)[α(t
0
)`
S
where (t, t
0
) is our time evolution operator.
Deﬁne Heisenberg picture states as
[α`
H
≡ 
†
[α(t)`
S
= [α(t
0
)`
S
which is clearly frozen in time. (We have used 
†
= 1). Deﬁne Heisenberg
operators
O
H
(t) ≡ 
†
O
S

which clearly evolves in time because O
S
is frozen but 
†
and  carry time
dependence.
Now the important thing about the Heisenberg and Schrodinger pictures
is that expectation values remain the same in both pictures. We must have
this for the physics to be the same. Expectation values are the same, i.e.
S
'β(t)[O
S
[α(t)`
S
=
H
'β[O
H
(t)[α`
H
[do Problem 8.2]
6.2.1 Heisenberg Equation
In the Schrodinger representation the state vectors evolve in time and the
Schrodinger equation describes their time evolution. In the Heisenberg pic
ture the operators evolve in time, so what is the equation governing the
operator time evolution? It is called the Heisenberg equation of motion (for
operators). It is obtained by diﬀerentiating O
H
(t) ≡ 
†
O
S
 to give
[O
H
(t), H] = i¯h
d
dt
O
H
(t)
[do Problems 8.3 and 8.4]
6.2.2 Interaction Picture
We use the Dirac or Interaction representation if the Hamiltonian can be
split into two parts; a free particle piece H
0
and an interaction piece H
I
.
(Unfortunately we have the notation H
I
I
for the interaction Hamiltonian in
the Interaction picture.) Thus
H = H
0
+H
I
98 CHAPTER 6. SMATRIX, CROSS SECTION & WICK’S THEOREM
where, at this stage, H
I
means interaction Hamiltonian, not Interaction
picture.
Deﬁne a time evolution operator for H
0
alone as

0
≡ 
0
(t, t
0
) ≡ e
−
i
¯ h
H
0
(t−t
0
)
and
[α(t)`
I
≡ 
†
0
[α(t)`
S
and
O
I
(t) ≡ 
†
0
O
S

0
Now
H
I
0
= H
S
0
≡ H
0
(exercise: show this)
Manipulating the previous two equations we get
[O
I
(t), H
0
] = i¯h
d
dt
O
I
(t)
and
H
I
I
(t)[α(t)`
I
= i¯h
d
dt
[α(t)`
I
where
H
I
I
(t) = 
†
0
H
S
I

0
= e
i
¯ h
H
0
(t−t
0
)
H
S
I
e
−
i
¯ h
H
0
(t−t
0
)
[do Problems 8.5, 8.6, 8.7]
Now deﬁne a time evolution operator for H
I
alone as
[α(t)`
I
≡ 
I
[α(t
0
`
I
which leads to
H
I
I

I
= i¯h
d
dt

I
(6.1)
which has the rigorous solution

I
(t, t
0
) = 1 +
∞
¸
n=1
1
n!
−i
¯h
n
t
t
0
dt
1
t
t
0
dt
2
t
n−1
t
0
dt
n
T[H
I
(t
1
) H
I
(t
n
)]
Now deﬁne the Smatrix
S ≡ (∞, −∞)
6.3. CROSS SECTION AND SMATRIX 99
in other words
[α(∞)` ≡ S[α(−∞` = S[i`
where [i` is the initial state. Thus
S = 1 +
∞
¸
n=1
1
n!
−i
¯h
n
∞
−∞
dt
1
∞
−∞
dt
2
∞
−∞
dt
n
T[H
I
(t
1
) H
I
(t
n
)
or in covariant form
S = 1 +
∞
¸
n=1
1
n!
−i
¯h
n
d
4
x
1
d
4
x
2
d
4
x
n
T[H
I
(x
1
)H
I
(x
2
) H
I
(x
n
)]
which is the Dyson expansion of the Smatrix. This is an inﬁnite series, each
term of which gets represented as a Feynman diagram.
Note that this expansion of the Smatrix does not rely on the Schrodinger
equation but comes from the general interaction picture and the opera
tor equation of motion (6.1), which, although reminiscent of, is not the
Schrodinger equation but rather the operator equation of motion in the In
teraction picture [Bj RQF 177].
6.3 Cross section and Smatrix
Consider the reaction
1 + 2 →1
+ 2
+ 3
+. . . n
We shall be calculating an Smatrix element, but it always has common
factors. Pulling these out we are left with a quantity ´called the invariant
amplitude. A set of Feynman rules actually gives −i´ [Griﬃths]. Cross
sections are writtem directly in terms of ´. The relation between the S
matrix element and the invariant amplitude is given by [Greiner FQ 267,
Greiner QED 221]
< f[S[i >= i(2π)
4
δ
4
(p
1
+p
2
−
n
¸
i=1
p
i
) ´Π
2
i=1
N
i
2E
i
(2π)
3
Π
n
i=1
N
i
2E
i
(2π)
3
where the normalization factors are N
i
= 1 for scalar bosons and photons
and N
i
= 2m for fermions. [Greiner FQ 267]
100CHAPTER 6. SMATRIX, CROSS SECTION & WICK’S THEOREM
A set of Feynman rules actually gives −i´[Griﬃths, 1987]. The diﬀer
ential cross section is given directly in terms of the invariant amplitude as
[RPP, pg. 176; PPB, pg. 219; Griﬃths, pg. 198]
dσ =
(2π)
4
[´[
2
4
(p
1
.p
2
)
2
−m
2
1
m
2
2
dΦ
n
(p
1
+p
2
; p
3
. . . p
n+2
)
with [RPP, pg. 175; PPB, pg. 215; Griﬃths, pg. 198]
dΦ
n
(P; p
1
. . . p
n
) ≡ δ
4
P −
n
¸
i=1
p
i
n
¸
i=1
d
3
p
i
(2π)
3
2E
i
Thus for the reaction
1 + 2 →3 + 4 n
we have [Griﬃths, pg. 198]
dΦ
n
(p
1
+p
2
; p
3
p
n
) = δ
4
(p
1
+p
2
−p
3
−p
4
− p
n
)
d
3
p
3
(2π)
3
2E
3
d
3
p
4
(2π)
3
2E
4
d
3
p
n
(2π)
3
2E
n
giving [Griﬃths, pg. 198]
dσ =
(2π)
4
[´[
2
4
(p
1
p
2
)
2
−m
2
1
m
2
2
d
3
p
3
(2π)
3
2E
3
d
3
p
4
(2π)
3
2E
4
d
3
p
n
(2π)
3
2E
n
δ
4
(p
1
+p
2
−p
3
−p
4
− p
n
)
For the reaction
1 + 2 →3 + 4
dσ =
(2π)
4
[´[
2
4
(p
1
p
2
)
2
−m
2
1
m
2
2
d
3
p
3
(2π)
3
2E
3
d
3
p
4
(2π)
3
2E
4
δ
4
(p
1
+p
2
−p
3
−p
4
)
with [Griﬃths, pg. 200]
δ
4
(p
1
+p
2
−p
3
−p
4
) = δ(E
1
+E
2
−E
3
−E
4
)δ
3
( p
1
+ p
2
− p
3
− p
4
)
giving
dσ
d
3
p
4
=
(2π)
4
[´[
2
4
√
d
3
p
3
(2π)
3
2E
3
1
(2π)
3
2E
4
δ(E
1
+E
2
−E
3
−E
4
)δ
3
( p
1
+ p
2
− p
3
−p
4
)
6.4. WICK’S THEOREM 101
so that
dσ
d
3
p
4
d
3
p
4
= dσ =
2π[´[
2
8E
4
(p
1
p
2
)
2
−m
2
1
m
2
2
d
3
p
3
(2π)
3
2E
3
δ(E
1
+E
2
−E
3
−E
4
)
giving the Lorentz invariant diﬀerential cross section for production of par
ticle 3 as
dσ
d
3
p
3
/E
3
=
[´[
2
64π
2
E
4
(p
1
p
2
)
2
−m
2
1
m
2
2
δ(E
1
+E
2
−E
3
−E
4
)
NNN PUT IN GRIFFITHS STATISTICAL FACTOR S (EASY!)
ALSO WRITE DOWN Λ
6.4 Wick’s theorem
6.4.1 Contraction
6.4.2 Statement of Wick’s theorem
References and Notes
Mandl & Shaw, Teller, Sakurai QM, Leon, Merzbacher
S matrix and G function Bj RQM 83,97,100
S matrix without SE, Bj RQF 177
For a derivation of the Smatrix based on Green function techniques,
and in and outstates, see Bjorken and Dell, RQF 177; RQM 83, 97, 100.
102CHAPTER 6. SMATRIX, CROSS SECTION & WICK’S THEOREM
Chapter 7
QED
7.1 QED Lagrangian
7.2 QED Smatrix
7.2.1 First order Smatrix
103
104 CHAPTER 7. QED
7.2.2 Second order Smatrix
From the Dyson expansion, the second order term in the Smatrix is (with
¯h = 1) [GreinerFQ 238]
S
(2)
=
1
2!
(−i)
2
d
4
x
1
d
4
x
2
T[H(x
1
)H(x
2
)]
=
1
2!
(−iq)
2
d
4
x
1
d
4
x
2
T[ :
¯
ψ(x
1
)γ
µ
ψ(x
1
)A
µ
(x
1
) : :
¯
ψ(x
2
)γ
ν
ψ(x
2
)A
ν
(x
2
) : ]
=
(−iq)
2
2!
d
4
x
1
d
4
x
2
:
¯
ψ(x
1
)γ
µ
ψ(x
1
)
¯
ψ(x
2
)γ
ν
ψ(x
2
)A
µ
(x
1
)A
ν
(x
2
) : (a)
+
(−iq)
2
2!
d
4
x
1
d
4
x
2
:
¯
ψ(x
1
)γ
µ
ψ(x
1
)
¯
ψ(x
2
)γ
ν
ψ(x
2
)A
µ
(x
1
)A
ν
(x
2
) : (b)
+
(−iq)
2
2!
d
4
x
1
d
4
x
2
:
¯
ψ(x
1
)γ
µ
ψ(x
1
)
¯
ψ(x
2
)γ
ν
ψ(x
2
)A
µ
(x
1
)A
ν
(x
2
) : (c)
+
(−iq)
2
2!
d
4
x
1
d
4
x
2
:
¯
ψ(x
1
)γ
µ
ψ(x
1
)
¯
ψ(x
2
)γ
ν
ψ(x
2
)A
µ
(x
1
)A
ν
(x
2
) : (d)
+
(−iq)
2
2!
d
4
x
1
d
4
x
2
:
¯
ψ(x
1
)γ
µ
ψ(x
1
)
¯
ψ(x
2
)γ
ν
ψ(x
2
)A
µ
(x
1
)A
ν
(x
2
) : (e)
+
(−iq)
2
2!
d
4
x
1
d
4
x
2
:
¯
ψ(x
1
)γ
µ
ψ(x
1
)
¯
ψ(x
2
)γ
ν
ψ(x
2
)A
µ
(x
1
)A
ν
(x
2
) : (f)
+
(−iq)
2
2!
d
4
x
1
d
4
x
2
:
¯
ψ(x
1
)γ
µ
ψ(x
1
)
¯
ψ(x
2
)γ
ν
ψ(x
2
)A
µ
(x
1
)A
ν
(x
2
) : (g)
+
(−iq)
2
2!
d
4
x
1
d
4
x
2
:
¯
ψ(x
1
)γ
µ
ψ(x
1
)
¯
ψ(x
2
)γ
ν
ψ(x
2
)A
µ
(x
1
)A
ν
(x
2
) : (h)
In this expansion we have used the modiﬁcation to Wick’s theorem that
says that “no equal time contractions are allowe” (no e.t.c.  see before).
As previously mentioned Greiner states this somewhat diﬀerently [Greiner
238] but the result is the same. Greiner has a nice discussion [Greiner 238]
showing that this prescription eliminates the socalled “tadpole” diagrams.
Also in the above expansion we don’t include contributions of the form
ψ(x
1
)ψ(x
2
) =
¯
ψ(x
1
)
¯
ψ(x
2
) = 0
because they give zero contribution. [Schwabl 337, GreinerFQ 238]
We now need to introduce some additional Feynman diagrams corre
sponding to the above contractions. These are illustrated in the ﬁgure be
low. Note that the photon diagrams do not have an arrow associated with
them since each photon is its own antiparticle. [Greiner FQ 236]
7.2. QED SMATRIX 105
x
2
x
1
(a)
x
2
x
1
(b)
Fig. x.x
(a) Fermion contraction ψ(x
2
)
¯
ψ(x
1
) = iS
F
(x
2
−x
1
)
(b) Photon contraction A
µ
(x
2
)A
ν
(x
1
) = iS
F
(x
2
−x
1
)
Using the above diagrams and also the ones shown previously we are in
a position to draw the Feynman diagrams corresponding to all the terms in
the second order Smatrix. These are shown in the Figure below.
Fig. Feynman diagrams for 2nd order Smatrix
NNN identical to GreinerFQ, ﬁg 8.5, p. 239
106 CHAPTER 7. QED
7.2.3 First order Smatrix elements
We previously considered the ﬁrst order Smatrix, and we wish now to evlu
ate Smatrix elements. For deﬁniteness let’s consider the diagram of Fig.
7.4 (b) which corresponds to an electron radiating a photon. In that case
the initial state would be creation of an electron from the vacuum, i.e.
[i >≡ b
†
k
1
,s
1
[0 >
and the ﬁnal state consists of the scattered electron together with the pro
duced photon, i.e.
[f >≡ b
†
k
1
,s
1
a
†
k,λ
[0 >
or
< f[ ≡< 0[a
k,λ
b
k
1
,s
1
We wish to evaluate the matrix element < f[S[i > and the ﬁrst order S
matrix S
(1)
contains 8 terms. However all but one (the second) of these terms
will be zero, if the [i > and < f[ states above are used. To illustrate this
let’s evaluate the ﬁrst of the eight < f[S
(1)
[i > terms, denoted < f[S
(1)
1
[i >.
7.3. CASIMIR’S TRICK & TRACE THEOREMS 107
7.2.4 Second order Smatrix elements
Electronelectron (Moeller) scattering
7.2.5 Invariant amplitude and lepton tensor
Electronmuon scattering
Invariant amplitude
7.3 Casimir’s trick & Trace theorems
7.3.1 Average over initial states / Sum over ﬁnal states
Polarized ﬁnal states / Unpolarized initial staes
Unpolarized initial and ﬁnal states
7.3.2 Casimir’s trick
2
Contents
1 Lagrangian Field Theory 1.1 Units . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.1.1 Natural Units . . . . . . . . . . . . . . . . . . . . . 1.1.2 Geometrical Units . . . . . . . . . . . . . . . . . . 1.2 Covariant and Contravariant vectors . . . . . . . . . . . . 1.3 Classical point particle mechanics . . . . . . . . . . . . . . 1.3.1 EulerLagrange equation . . . . . . . . . . . . . . . 1.3.2 Hamilton’s equations . . . . . . . . . . . . . . . . . 1.4 Classical Field Theory . . . . . . . . . . . . . . . . . . . . 1.5 Noether’s Theorem . . . . . . . . . . . . . . . . . . . . . . 1.6 Spacetime Symmetries . . . . . . . . . . . . . . . . . . . . 1.6.1 Invariance under Translation . . . . . . . . . . . . 1.6.2 Angular Momentum and Lorentz Transformations 1.7 Internal Symmetries . . . . . . . . . . . . . . . . . . . . . 1.8 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.8.1 Covariant and contravariant vectors . . . . . . . . 1.8.2 Classical point particle mechanics . . . . . . . . . . 1.8.3 Classical ﬁeld theory . . . . . . . . . . . . . . . . . 1.8.4 Noether’s theorem . . . . . . . . . . . . . . . . . . 1.9 References and Notes . . . . . . . . . . . . . . . . . . . . . 2 Symmetries & Group theory 2.1 Elements of Group Theory . . . . . . . . . . 2.2 SO(2) . . . . . . . . . . . . . . . . . . . . . 2.2.1 Transformation Properties of Fields 2.3 Representations of SO(2) and U(1) . . . . . 2.4 Representations of SO(3) and SU(1) . . . . 2.5 Representations of SO(N) . . . . . . . . . . 3 7 7 7 10 11 12 12 14 15 18 24 24 25 26 29 29 29 29 30 32 33 33 33 34 35 35 36
. . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . 87 4. .6. . . . . . . . . .3 KleinGordon QFT Hamiltonian . . . . . .4 3 Free KleinGordon Field 3. . . . .1 Derivation of b. . d.3. . . . . . . . . .8. . . .5. . . . .3 Dirac Equation . . . a† Commutators . . . . .7. . . . . . . . . . . 83 4. . . 3. . . . . . . .3 Propagator Theory . . . . . . . . . . . . . . . . . . . . . . . . 3. . . . .9 References and Notes . . . . . 87 4. . 3. . 3. . . . . . . . . . . . .2. . . . . . . . . . . . . . . .2 KleinGordon Quantum ﬁeld . . . . . . . . . . . . . . . . . . . . 78 4. . . . . . .4 Normal order . . .1 Schrodinger Equation . . .2 Probability and Current . . . . . . . 79 4. . . 87 4. . 3. . . . . . . . . . . . . 88 4. . . . . . . . .5. . . .8. . . . . 3. .3. . . . . . 3. . 79 4. . . . . . . . . .1 Indirect Derivation of a. . . . . . . . 82 4. . . . . . . . . .5 KleinGordon QFT . . . 77 4. CONTENTS 37 37 39 39 40 41 42 45 45 47 47 48 50 51 66 68 70 70 71 72 73 73 . . . . . 3. . . . . . . . . . . .8 Hamiltonian. . .4 Free Particle Solutions of Dirac Equation . . . . . . . . . . . . . . . . . . . . . . . . . 3. . . . . . . . . 3. . 80 4. . . . . . 3.1 Noether spacetime current . . . . . . . . . . . .3 Negative Energy and Antiparticles . . . . . . . . . . . . . . . . . . 88 4. . . . . d† Anticommutators . . Momentum and Charge in terms of creation and annihilation operators . .5. . . . . . . . . . . . .1 KleinGordon Equation . . . . . .3 Classical Field Theory .8. . . . . . . . . . . 87 4. .7 Pauli Exclusion Principle . . . . . . 3. . . . . . .2 KleinGordon Equation . . . .5. . . . . . . . . .5. . .8. . . . . . 88 4. . 3. . . 3. . . . b† . . . . . . . . . 4 Dirac Field 75 4. . . .2 Bilinear Covariants . . . . . . . . .5. . . . . . 3. .6 Dirac QFT . . .8 Summary . . . . . . . . . . . . . . . . .6 Propagator Theory . . . . . .2. 3.2 Noether internal symmetry and charge .3 Fourier expansion and momentum space . . . . . . . . . . . 3. . . . . . . . . . .4 Fourier Expansion & Momentum Space . . . . . . . . . . . . . .1 Schrodinger equation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4 Complex KG ﬁeld . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2 KleinGordon Equation . . . . . . . .5. . . . . . . . . . . . . . . .5. . . . .1 Probability & Current . . . . 3. . . . .5 Wave Function . . . . . . . . . . 88 . 3. . . . . .3. . . . . . .2 Direct Derivation of a. . . . .7 Complex KleinGordon Field .1 KG classical ﬁeld . . . . . .1 Charge and Complex Scalar Field . . . . . . . . . . . . . a† Commutators 3.5 Classical Dirac Field .
. . . . . . . . . . . . . . . . . . . . . . . . .1 Maxwell equations in tensor notation . . . . . .1 Schrodinger Time Evolution Operator . . .1 Contraction . . . . . . . . . . . . . . .1. . . . . . 6. . .2 Statement of Wick’s theorem . 5. . . . . .3 Angular Momentum . . . .8. . .1. .6 Linear polarization vectors in Coulomb gauge 5. Heisenberg and Dirac (Interaction) Pictures 6. . .4 Wick’s theorem . . . . . . . . .1. . . .3 Cross section and Smatrix . . . . . . . 4. . . . . 6. . . . . . . . . . . . . . . . . . . . . . . . . . .1 Heisenberg Equation . . . . . . . . . . . . . . . . . . . .3 Coulomb Gauge . . . . . . .1. . . 6. . . . . 4. . . . . . . .2 Classical Dirac ﬁeld . . . . . . . . . . cross section & Wick’s theorem 6. . . .2 Interaction Picture . . . . . . . . . . . . . . . . . . . .1. . . 5. . . . . . . . . . . . . . . . . . . . . . . .1. . . 93 .11 References and Notes . . . . . .10 Summary . . . . . . . . . . 101 . . . . . . . . . . . . . . . . . . .8 Fourier expansion . . . . 93 . . 96 . . . . . . . .7 Circular polarization vectors .1 Creation & annihilation operators . 4. . .1 Dirac equation summary 4. . . . . . . . . . . . . . . . . . . . . . . . . . .2 Momentum . . . . . . . . . . 6. . 5. .8.4 GuptaBleuler quantization .1.1. . . . . . . . . . . . . . . . . . .2. . 5. . . .2. . . . . . . . . . . . . . . . 99 . . . . . . . 4. 4. . . .4. .1 Hamiltonian . .5 Polarization vectors . . . 5. . . . . . . 101 . . . 5. . . . . . . .1 Review of Classical Electrodynamics . . . . . .10. . . . . . . . . . . . . . .10. . . . . 95 . . . . . .8. . . . . . . . .10. . . . . . . 5. 5. .2. . . . . . .4 Lagrangian for EM ﬁeld . . . . . . .4 Propagator theory . . . 97 .5 Proca ﬁeld . .4. 6 Smatrix. . . . . . . . . . . 6. . . . . . . . . . . . . .2 Quantized Maxwell ﬁeld . . . . . .CONTENTS 4. . . . . . . .4 Charge . .3 Photon propagator . . . . . . .1 Time Ordered Product . . . . . . . . .1. . . . . . . . . . . . . . . 5. . 4.3 Dirac QFT . . . . . . . . . . . . . . . . . . . 97 . . . . . .8. . . . . . . . . . 5. . . . 101 . . . . . . .9 Propagator theory . . . . . . . . . . . . . . . . . . . . 5. . . . . . . . . . . . . . 5 88 88 88 88 88 88 88 88 88 88 88 89 89 89 89 89 89 89 91 91 91 91 91 91 91 91 5 Electromagnetic Field 5. . . . .2 Schrodinger. . .10. . . . .2 Gauge theory . . . 5. . . . . 6. . 4. . . . . . 4. . 6. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4. . . . .
. .2. . . . . . . 107 . . . . . . . . . . . . . 7. . . .3. . .2 Casimir’s trick .2.2. . . . . . .2. . . . . . . .1 First order Smatrix . . 103 . . . . . . . . . . . . . . 103 . . . .1 Average over initial states / Sum over ﬁnal states 7. . . 107 . . . . . .6 CONTENTS 103 . . . 107 7 QED 7. . . . . . . . . .3 First order Smatrix elements . 106 . . . . . . . . 7. . . 7. 107 . 7. . . . . . . . . 7. .4 Second order Smatrix elements . . . . . . . . . . . 7. 7. . . . .5 Invariant amplitude and lepton tensor . . . . . . . . . . . . . . . . . . 104 . .2 Second order Smatrix . . . 103 . . . .2.3 Casimir’s trick & Trace theorems .1 QED Lagrangian . . .2 QED Smatrix . . . . . . . . . . . . . . 7. . . . 107 . .3.
namely units and concentrate on the units most widely used in particle physics and quantum ﬁeld theory (natural units). 1 eV = 1.055 × 10−34 J sec = 6.Chapter 1 Lagrangian Field Theory 1. The formulas ¯ considerably simplify if we choose a set of units. i. called natural units where c and h are set equal to 1.6 × 10−19 J h ¯ c = 197M eV f m 1f m = 10−15 m 1barn = 10−28 m2 1mb = . the particles are both relativistic and quantum mechanical and therefore our formulas have lots of factors of c (speed of light) and h (Planck’s constant).1 Natural Units In particle physics and quantum ﬁeld theory we are usually dealing with particles that are moving fast and are very small. In natural units the units of length. seconds.e. because these days it is likely that students will study this subject as well. seconds (sec).582 × 10−22 M eV sec m c = 3 × 108 sec . kilogram (kg). the basic quantities of length.1. We also mention the units used in General Relativity. mass and time are centimeters (cm). 7 . gram (g). or in MKS units these are meters (m). Some useful quantities are [PPDB]: h h ¯ ≡ 2π = 1.1f m2 1.1 Units We start with the most basic thing of all. ¯ In CGS units (often also called Gaussian [Jackson appendix] units). mass and time are all expressed in GeV.
If c ≡ 1 ⇒ sec = 3 × 1010 cm We can now derive the other conversion factors for natural units. one can deduce the units of mass from E = mc2 .99792458 × 108 m/sec and h = 1. LAGRANGIAN FIELD THEORY Example With c ≡ 1. These are sec = 1. in which h ¯ is also set equal to unity. show that sec = 3 × 1010 cm. Solution c = 3 × 1010 cm sec−1 .5821220 × ¯ −25 GeV sec. ) 10 .61 × 1026 GeV (The exact values of c and h are listed in the [Particle Physics Booklet] as ¯ c = 2.52 × 1024 GeV −1 m = 5.05457266 × 10−34 Jsec= 6.07 × 1015 GeV −1 kg = 5. Once the units of length and time are established.8 CHAPTER 1.
1. One simply substitutes the conversion factors from before. Natural units are also often used in cosmology and quantum gravity [Guidry 514] with G given above as G = M12 . the best known value of G is [PPDB] G = 6.7 × 10−39 GeV −2 1 = 2 MP l where the Planck mass is deﬁned as MP l ≡ 1. Let’s now get to the problem. Solution It is interesting to note that the value of G is one of the least accurately known of the fundamental constants.52 × 1024 GeV −1 )2 = 6. which contains far fewer decimal places than the other fundamental constants. UNITS 9 Example Deduce the value of Newton’s gravitational constant G in natural units. Pl . Whereas.67 × 10−11 m3 kg −1 sec−2 6.07 × 1015 GeV −1 )3 = (5.67 × 10−11 (5.22 × 1019 GeV .51099906M eV /c2 or the ﬁne structure constant as α = 1/137. say the mass of the electron is known as [Particle Physics Booklet] me = 0.67259 × 10−11 m3 kg −1 sec−2 .0359895 and c and h are known to many decimal ¯ places as mentioned above. namely G = 6.1.61 × 1026 GeV )(1.
67 × 10−8 cm3 g −1 sec−2 and when c ≡ 1 we have sec = 3 × 1010 cm giving G = 6. In gravitation (General Relativity). Example Evaluate G when c ≡ 1. quantum eﬀects become important at length scales approaching LP l .6 × 10−66 cm2 . In natural units c = h = 1 and we deduce that G = M12 as ¯ Pl in a previous Example.6 × 10 energies (or masses) approach the Planck mass MP l . Recall that in natural units everything was expressed in terms of GeV .1. . gravity becomes important when LP l ≡ 1.67 × 10−8 cm3 g −1 (3 × 1010 cm)−2 = 7.4 × 10−29 cm g −1 Now imposing G ≡ 1 gives the geometrical units sec = 3 × 1010 cm g = 7.10 CHAPTER 1. Solution G = 6. In geometrical units everything is expressed in terms of cm.4 × 10−29 cm It is important to realize that geometrical and natural units are not compatible. In particle physics. LAGRANGIAN FIELD THEORY 1.2 Geometrical Units In classical General Relativity the constants c and G occur most often and geometrical units are used with c and G set equal to unity.67 × 10−11 m3 kg −1 sec−2 = 6. (see Problems) Note that in these units h = L2 l where ¯ ¯ P −33 cm. In geometrical units c = G = 1 we deduce that h = 2.
∂i ) = ( o . denoted by the covariant symbol ∂µ and deﬁned via ∂ ∂ ∂ ∂ ≡ ∂µ = (∂o . Ai ) = (Ao . −A) noting that Ao = Ao Thus Aµ = (Ao . ) ) ∂ ∂ µ = ( ∂t . ∂xµ ∂x ∂x ∂t The contravariant operator ∂ µ is given by ∂ ∂ ∂ ∂ ∂ ∂ ≡ ∂ µ = (∂ o . A) Covariant vectors are formed by “lowering” the indices with the metric tensor as in Aµ = gµν Aν = (Ao .1. i ) = ( . Ai ) = (Ao . −A) Now we discuss derivative operators. COVARIANT AND CONTRAVARIANT VECTORS 11 1.2. The metric tensor that is used in this book is η µν = ηµν = 1 0 0 0 0 −1 0 0 0 0 −1 0 0 0 0 −1 Contravariant vectors are written in 4dimensional form as Aµ = (Ao .− ) ∂xµ ∂xo ∂xi ∂x ∂x ∂t Thus ∂ ∂µ = ( ∂t .2 Covariant and Contravariant vectors The subject of covariant and contravariant vectors is discussed in [Jackson]. In this section we summarize the basic results. ∂ i ) = g µν ∂ν = ( . which students should consult for a thorough introduction.− i) = ( . ) = ( o. −Ai ) = (Ao . A) Aµ = (Ao . − ) .
) If m = 0 then Fi = m¨i = mai . For conservative forces F = − U ˙ ˙ q where U is the scalar potential.1 Classical point particle mechanics EulerLagrange equation dp dt Newton’s second law of motion is F = or in component form (for each component Fi ) Fi = dpi dt where pi = mqi (with qi being the generalized position coordinate) so that ˙ dpi = mqi + m¨i .3.1) Finally. LAGRANGIAN FIELD THEORY The length squared of our 4vectors is A2 ≡ Aµ Aµ = Aµ Aµ = A2 − A2 o and ∂ 2 ≡ ∂µ ∂ µ ≡ P2 = ∂2 − ∂t2 2 (1. note that with our 4vector notation. Rewriting Newton’s law we have − dU d = (mqi ) ˙ dqi dt . (Here and throughout this book we use the notation ˙ ˙ q dt dx x ≡ dt . the usual quantum mechanical replacements pi → i¯ ∂ i ≡ −i¯ h h and po → i¯ ∂ o = i¯ h h can be succintly written as pµ → i¯ ∂ µ h giving (with h = 1) ¯ p2 → −P2 ∂ ∂t 1.3 1.12 CHAPTER 1.
˙ 1 In freshman physics T = T (qi ) = 2 mqi and U = U (qi ) such as the harmonic ˙ ˙2 1 2 . A more rigorous derivation is based on the calculus of variations [HoKim47. consider a variation of the path [Schwabl262. Bergstrom284] as follows. To see the consequences of this. Huang54. generalized velocities qi and time t. qi . That is in freshman physics T is a function only oscillator U (qi ) = 2 kqi of velocity qi and U is a function only of position qi . In classical point particle mechanics the action is S≡ t2 L(qi . Thus ˙ ˙ ˙ ˙ Newton’s law is dpi dt ∂L d ∂L = ( ) ∂qi dt ∂ qi ˙ Fi = with the canonical momentum deﬁned as pi ≡ ∂L ∂ qi ˙ The next to previous equation is known as the EulerLagrange equation of motion and serves as an alternative formulation of mechanics [Goldstein].e. ˙ According to Hamilton’s principle. qi ) ≡ T − U where T is the kinetic energy. the action has a stationary value for the correct path of the motion [Goldstein36]. BjRQF6] qi (t) → qi (t) ≡ qi (t) + δqi (t) . t)dt ˙ t1 where the Lagrangian L is a function of generalized coordinates qi . i.Goldstein37.3. δS = 0 for the correct path.1. qi ) = ˙ ˙ ∂L dU ∂L dT T (qi ) − U (qi ). CLASSICAL POINT PARTICLE MECHANICS 13 Let us deﬁne the Lagrangian L(qi . Thus L(qi . It is usually written d ∂L ∂L )− =0 ( dt ∂ qi ˙ ∂qi or just pi = ˙ ∂L ∂qi We have obtained the EulerLagrange equations using simple arguments. It follows that ∂qi = − dqi and ∂ qi = dqi = mqi = pi .
3.2) where the boundary term has vanished because δqi (t1 ) = δqi (t2 ) = 0. Hamilton’s equations of motion immediately follow as ∂H = qi ˙ ∂pi ∂L Now L = L(pi ) and ∂H = − ∂qi so that our original deﬁnition of the canon∂qi ical momentum above gives − ∂H = pi ˙ ∂qi . qi ) ˙ ˙ p2 p2 p2 (1. qi ) = 2m + U (qi ) = T + U which is the ˙ total energy. 1. qi ) ≡ pi qi − L(qi .2 Hamilton’s equations We now introduce the Hamiltonian H deﬁned as a function of p and q as H(pi .14 CHAPTER 1. The subsequent variation in the action is (assuming that L is not an explicit function of t) t2 δS = t1 ( ∂L ∂L δqi + δ qi )dt = 0 ˙ ∂qi ∂ qi ˙ with δ qi = ˙ d dt δqi and integrating the second term by parts yields ∂L δ qi dt = ˙ ∂ qi ˙ ∂L d(δqi ) ∂ qi ˙ ∂L ∂L = δqi t2 − δqi d( ) t1 ∂ qi ˙ ∂ qi ˙ d ∂L = 0 − δqi ( )dt dt ∂ qi ˙ (1. We are left with t2 δS = t1 ∂L d ∂L − ( ) δqi dt = 0 ∂qi dt ∂ qi ˙ which is true for an arbitrary variation δqi indicating that the integral must be zero. which yields the EulerLagrange equations. LAGRANGIAN FIELD THEORY subject to the constraint δqi (t1 ) = δqi (t2 ) = 0.3) ∂L ˙2 ˙ ˙ For the simple case T = 1 mqi and U = U (qi ) we have pi ∂ qi = mqi so that 2 ˙ i i i T = 2m and pi qi = m so that H(pi .
In the next section we shall show how to derive the EulerLagrange equations from the action deﬁned as S≡ Ldt which again is clearly not covariant. ∂µ φ) and with L ≡ Ld3 x. A covariant form of the action would involve a Lagrangian density L via S≡ Ld4 x = Ld3 xdt ∂L where L = L(φ. The generalized coordiante q has been replaced by the ﬁeld variable φ and the discrete index i has been replaced by a continuously varying index x. consider an arbitrary variation of the ﬁeld [Schwabl 263. ˙ ˙ Let us replace the qi by a ﬁeld φ ≡ φ(x) where x ≡ (t. Goldstein548]. Ryder 83. BjRQF13] φ(x) → φ (x) ≡ φ(x) + δφ(x) . If there is more than one scalar ﬁeld φi then the EulerLagrange equations are ∂L ∂µ ∂(∂µ φi ) − ∂L ∂φi =0 To derive the EulerLagrange equations for a scalar ﬁeld [HoKim48.35. The term − ∂qi in the Euler∂L Lagrange equation gets replaced by the covariant term − ∂φ(x) .1. CLASSICAL FIELD THEORY 15 1. Mandl & Shaw 30. In both special and general relativity we always seek covariant equations in which space and time are given equal status. an example of which may be the Higgs boson. Thus the ﬁeld φ considered below can be thought of as an inﬂaton. Such a ﬁeld is called an inﬂaton. x). Thus one can guess that the covariant generalization of the point particle EulerLagrange equation is ∂µ ∂L ∂L − =0 ∂(∂µ φ) ∂φ which is the covariant EulerLagrange equation for a ﬁeld φ.39. Any time d ∂ derivative dt should be replaced with ∂µ ≡ ∂xµ which contains space as well as time derivatives.4. The EulerLagrange equations above are clearly not covariant because special emphasis is placed on d ∂L time via the qi and dt ( ∂ qi ) terms.4 Classical Field Theory Scalar ﬁelds are important in cosmology as they are thought to drive inﬂation. a Higgs boson or any other scalar boson.
16 CHAPTER 1. implying that the integrand must be zero. yielding the EulerLagrange equations. In analogy with the canonical momentum in point particle mechanics. we deﬁne the covariant momentum density Πµ ≡ ∂L ∂(∂µ φ) so that the EulerLagrange equations become ∂µ Πµ = ∂L ∂φ The canonical momentum is deﬁned as Π ≡ Π0 = ∂L ˙ ∂φ The energy momentum tensor is (analagous to the deﬁnition of the point particle Hamiltonian) Tµν ≡ Πµ ∂ν φ − gµν L . We need the result δ(∂µ φ) = ∂µ δφ = ∂ δφ ∂xµ which comes about because δφ(x) = φ (x) − φ(x) giving ∂µ δφ(x) = ∂µ φ (x) − ∂µ φ(x) = δ∂µ φ(x) showing that δ commutes with diﬀerentiation ∂µ . but the ﬁnal result is (see Problems) X2 δS = X1 ∂L ∂L δφ d4 x = 0 − ∂µ ∂φ ∂(∂µ φ) which holds for arbitrary δφ. Integration by parts on the second term is a bit more complicated than before for the point particle case. LAGRANGIAN FIELD THEORY again with δφ = 0 at the end points. The variation of the action is (assuming that L is not an explicit function of x) X2 δS = X1 ∂L ∂L δφ + δ(∂µ φ) d4 x = 0 ∂φ ∂(∂µ φ) where X1 and X2 are the 4surfaces over which the integration is performed.
4. ˙ Π=φ B) Evaluating ∂L = −m2 φ. Solution A) The covariant momentum density is more easily evaluated by rewriting LKG = 1 (g µν ∂µ φ∂ν φ − m2 φ2 ). Thus for the KleinGordon ﬁeld we have Πα = ∂ α φ ˙ giving the canonical momentum Π = Π0 = ∂ 0 φ = ∂0 φ = φ. CLASSICAL FIELD THEORY with the Hamiltonian density H≡ Hd3 x 17 ˙ H ≡ T00 = Πφ − L In order to illustrate the foregoing theory we shall use the example of the classical. C) Derive expressions for the energymomentum tensor and the Hamiltonian density. In momentum space p2 = −P2 . massive scalar ﬁeld. Thus 2 ∂L α α α α Πµ = ∂(∂µ φ) = 1 g µν (δµ ∂ν φ + ∂µ φδν ) = 1 (δµ ∂ µ φ + ∂ ν φδν ) = 2 2 1 α 2 (∂ φ + ∂ α φ) = ∂ α φ. the EulerLagrange equations give ∂φ the ﬁeld equation as ∂µ ∂ µ φ + m2 φ or (P2 + m2 )φ = 0 ¨ φ− 2 φ + m2 φ = 0 which is the KleinGordon equation for a free. massive KleinGordon ﬁeld. Example The massive KleinGordon Lagrangian density is 1 LKG = (∂µ φ∂ µ φ − m2 φ2 ) 2 1 ˙2 = [φ − ( φ)2 − m2 φ2 ] 2 A) Derive expressions for the covariant momentum density and the canonical momentum. thus (p2 − m2 )φ = 0 . B) Derive the equation of motion in position space and momentum space.1.
section 2. pg. 588] and [Greiner FQ. Many books [Kaku] discuss Noether’s theorem in a piecemeal fashion. The theory below closely follows [Greiner FQ 40]. Section 127. pg.5 Noether’s Theorem Noether’s theorem provides a general and powerful method for discussing symmetries of the action and Lagrangian and directly relating these symmetries to conservation laws. We prefer to use the notation of [Goldstein] for ﬁelds. The discussions by [Goldstein] and [Greiner] treat the symmetries of both the Lagrangian and the action as well. 39]. for example by treating internal and spacetime symmetries separately.4. Another excellent discussion of this general approach is presented in [Schwabl. namely ηr (x) or ηr (xµ ) rather than using .) 2 C) The energy momentum tensor is Tµν ≡ Πµ ∂ν φ − gµν L = ∂µ φ∂ν φ − gµν L 1 = ∂µ φ∂ν φ − gµν (∂α φ∂ α φ − m2 φ2 ) 2 ˙ Therefore the Hamiltonian density is H ≡ T00 = φ2 − 1 (∂α φ∂ α φ− 2 2 φ2 ) which becomes [Leon] m 1 1 1˙ H = φ2 + ( φ)2 + m2 φ2 2 2 2 1 2 = [Π + ( φ)2 + m2 φ2 ] 2 1. 268].4.2. It is better to develop the formalism for all types of symmetries and then to extract out the spacetime and internal symmetries as special cases.1). although the general spacetime and internal symmetries are properly treated together. section 12. The best discussion of this approach is in [Goldstein. pg. LAGRANGIAN FIELD THEORY ∂ (Note that some authors [Muirhead] deﬁne P2 ≡ 2 − ∂t2 different from (1. In what follows we rely on the methods presented by [Goldstein].18 CHAPTER 1. so that they write the KleinGordon equation as (P2 − m2 )φ = 0 or (p2 + m2 )φ = 0. However note that the discussion presented by [Schwabl] concerns itself only with the symmetries of the Lagrangian.
However there are other transformations (such as internal symmetries or gauge symmetries) that change the shape of the wave function at a single point. namely L (ηr (x ). We wish to consider how the Lagrangian and action change under a coordinate transformation xµ → xµ ≡ xµ + δxµ Let the corresponding change in the ﬁeld (total variation) be [Ryder83. ∂ν ηr (x ). ∂ν ηr (x ). ∂ µ η(x). Thus the local variation is deﬁned as (same as before) ηr (x) ≡ ηr (x) + δηr (x) This follows from the assumption of form invariance [Goldstein 589]. x) → L (ηr (x ). x ) ∂ with ∂ν ≡ ∂x ν The assumption of form invariance [Goldstein 589] says that the Lagrangian has the same functional form in terms of the transformed quantities as it does in the original quantities.1. namely the change in coordinates from x to x and also the change in the shape of the function from η to η . x) where ∂ µ η(x) ≡ ∂η(x) ∂xµ and 1 L (x ) ≡ L(η (x ). spinor or vector ﬁeld components.5. The notation ηr (x) is completely general and can refer to scalar. ∂ν ηr (x ). ∂ µ η (x ). x ) 1 . Schwabl263] ηr (x ) ≡ ηr (x) + ∆ηr (x) and the corresponding change in the Lagrangian L (x ) ≡ L(x) + ∆L(x) with L(x) ≡ L(η(x). ∂ν ηr (x). x ) (Note: no prime on L on right hand side) where ∂ µ η (x ) ≡ ∂η (x ) ∂xµ Notice that the variations deﬁned above involve two transformations. x ) = L(ηr (x ). In general the Lagrangian gets changed to L(ηr (x). NOETHER’S THEOREM 19 φr (x) or ψr (x) because the latter notations might make us think of scalar or spinor ﬁelds.
20 CHAPTER 1. We do this because the second term is second order involving both ∂η and δxµ .. LAGRANGIAN FIELD THEORY The local and total variations are related via δηr (x) = ηr (x) − ηr (x) = ηr (x) − ηr (x ) + ηr (x ) − ηr (x) = −[ηr (x ) − ηr (x)] + ∆ηr (x) Recall the Taylor series expansion f (x) = f (a) + (x − a)f (a) + . ∂f (x) = f (a) + (x − a) + ..) From the deﬁnition δη(x) ≡ η (x) − η(x) it is obvious that (see before) ∂ ∂η(x) δη(x) = δ ∂xµ ∂xµ . (It turns out δ does commute but ∆ does not.. Thus ﬁnally we have the relation between the total and local variations as (to ﬁrst order) δηr (x) = ∆ηr (x) − ∂ηr δxµ ∂xµ Now we ask whether the variations ∆ and δ commute with diﬀerentiation..  ∂x x=a or f (x) − f (a) ≈ (x − a) which gives η(x ) − η(x) ≈ (x − x) ∂η(x ) x =x ∂x ∂η ∂η ≡ (x − x) = δx ∂x ∂x ∂ηr δxµ ∂xµ ∂f ∂a Thus δηr (x) = ∆ηr (x) − To lowest order ηr ≈ ηr .
∂ν ηr (x). ∂ν ηr (x ). ∂ν ηr (x). NOETHER’S THEOREM 21 ∂ showing that δ “commutes” with ∂ µ ≡ ∂xµ . but has an additional term.e. x) − Ω d4 x L(ηr (x). ∂ν ηr (x). as in (see Problems) [Greiner FQ41] ∂ ∂η(x) ∂η(x) ∂δxν ∆η(x) = ∆ + ∂xµ ∂xµ ∂xν ∂xµ Let us now study invariance of the action [Goldstein 589. The assumption of scale invariance [Goldtein 589] says that the action is invariant under the transformation 2 (i. xµ ) ≡ S Demanding that the action is invariant. ∂ν ηr (xµ ). x ) − Ω d4 x L(ηr (x). Greiner FQ 41]. However ∆ does not commute.5. x) = 0 In the ﬁrst integral x is just a dummy variable so that d4 x L(ηr (x).1. transformation of an ignorable or cyclic coordinate) S ≡ = Ω Ω d4 x L (ηr (xµ ). we have (in shorthand notation) δS ≡ d4 x L (x ) − d4 x L(x) ≡ 0 Ω Ω Note that this δS is deﬁned diﬀerently to the δS that we used in the derivation of the EulerLagrange equations. . x) = 0 Ω which [Goldstein] uses to derive current conservation. Using L (x ) ≡ L(x) + ∆L(x) gives δS ≡ d4 x ∆L(x) + Ω Ω d4 x L(x) − Ω d4 x L(x) = 0 We transform the volume element with the Jacobian d4 x = using x µ ∂x µ 4 d x ∂xν = xµ + δxµ which gives [Greiner FQ 41] 2 Combining both form invariance and scale invariance gives [Goldstein 589] δS ≡ S − S = Ω d4 x L(ηr (x ). ∂ν ηr (xµ ). xµ ) d4 x L(ηr (xµ ).
but because the region of . ∂δxµ 4 )d x ∂xµ = (1 + to ﬁrst order only.. 1+ ∂δx3 ∂x3 d4 x . LAGRANGIAN FIELD THEORY ∂x d4 x = d4 x = ∂xν µ 1+ ∂δx0 ∂x0 ∂δx1 ∂x0 1 . Using ∂x the relation between local and total variations gives δS = = ∂L ∂δxµ δxµ ) + d4 x L(x) µ ∂x ∂xµ Ω Ω ∂ d4 x {δL(x) + µ [L(x)δxµ ]} ∂x Ω d4 x (δL(x) + Recall that L(x) ≡ L(ηr (x). ∂ µ ηr (x)). This expression for δL is substituted back into δS = 0. . . . Thus the variation in the action becomes δS = Ω (1 + 4 ∂δxµ 4 )d x ∆L(x) + ∂xµ 4 (1 + Ω ∂δxµ 4 )d x L(x) − ∂xµ Ω d4 x L(x) = ∂δxµ d x ∆L(x) + d x L(x) ∂xµ Ω Ω µ to ﬁrst order. . . ∂δx0 ∂x1 1 + ∂δx1 ∂x .. The second order term ∂δxµ ∆L(x) has been discarded... because δ “commutes” with ∂ µ ≡ δL = = Now add zero. Now express the local variation δL in terms of total variations of the ﬁeld as δL = = ∂L ∂L δηr + δ(∂ µ ηr ) ∂ηr ∂(∂ µ ηr ) ∂L ” + ∂ µ δηr ∂(∂ µ ηr ) ∂ ∂xµ . δηr + ∂L ∂ µ δηr ∂(∂ µ ηr ) ∂L ∂L ∂L δηr − ∂ µ δηr + ∂ µ µη ) ∂ηr ∂(∂ r ∂(∂ µ ηr ) ∂L ∂L ∂L − ∂µ δηr + ∂ µ δηr µη ) ∂ηr ∂(∂ r ∂(∂ µ ηr ) Note: the summation convention is being used for both µ and r.22 CHAPTER 1.
so that the second term becomes ∂L ∂ηr ∆ηr − δxν + Lδxµ = 0 ∂µ µη ) ∂(∂ r ∂xν which is the continuity equation ∂ µ jµ = 0 with [Schwabl 270] jµ ≡ ∂L ∂L ∆ηr − ∂ν ηr − gµν L δxν µη ) ∂(∂ r ∂(∂ µ ηr ) ≡ ∂L ∆ηr − Tµν δxν ∂(∂ µ ηr ) with the energymomentum tensor deﬁned as [Scwabl 270] Tµν ≡ ∂L ∂ν ηr − gµν L ∂(∂ µ ηr ) The corresponding conserved charge is (See Problems) Q≡ such that d3 x j0 (x) dQ =0 dt Thus we have j0 (x) is just the charge density j0 (x) ≡ ρ(x) This leads us to the statement. Thus the integrand is ∂L ∂L ∂L − ∂µ δηr + ∂ µ δηr + Lδxµ = 0 µη ) ∂ηr ∂(∂ r ∂(∂ µ ηr ) The ﬁrst term is just the EulerLagrage equation which vanishes. The spatial integral over this current’s zero component yields a conserved charge. [Mosel 16] . the integrand itself has to vanish.1. For ηr use the relation between local and total variations. NOETHER’S THEOREM 23 integration is abritrary.5. Noether’s Theorem: Each continuous symmetry transformation that leaves the Lagrangian invariant is associated with a conserved current.
LAGRANGIAN FIELD THEORY 1. ∂L ∂ν ηr − gµν L ∂(∂ µ ηr ) The above expression for Tµν we wrote Tµν = πµ ∂ν φ − gµν L with πµ = ∂L ∂(∂ µ φ) dP ν =0 dt is the same result we obtained before where (1. In 4dimensional notation [GreinerFQ 43] P ν = (E.24 CHAPTER 1. P ) = with d3 x T 0ν = constant. T02 . Thus Tµν will have 4 conserved charges corresponding to T00 . Dropping oﬀ the constant factor ν lets us write down a modiﬁed current (called the energymomentum tensor) Tµν ≡ with ∂ µ Tµν = 0 In general jµ has a conserved charge Q ≡ d3 x jo (x). The simplest spacetime symmetry is 4dimensional translation invariance. 1. Super symmetries relate both of these. The shape of the ﬁeld does not change.5) .6. so that ∆ηr = 0 (which is properly justiﬁed in Schwabl 270) giving the current as jµ = − ∂L ∂ηr − gµν L µ η ) ∂xν ∂(∂ r ν with ∂ µ jµ = 0.4) (1. T03 which are just the energy E and momentum P of the ﬁeld.6 Spacetime Symmetries The symmetries we will consider are spacetime symmetries and internal symmetries.1 Invariance under Translation µ. involving space translation and time translation. T01 . [GreinerFQ 43] Consider translation by a constant factor xµ = xµ + µ and comparing with xµ = xµ + δxµ gives δxµ = µ .
6.2 Angular Momentum and Lorentz Transformations NNN: below is old Kaku notes.1. and we get the conserved current Mρµν = T ρν xµ − T ρµ xν with ∂ρ Mρµν = 0 and the conserved charge M µν = with d3 x M0µν d µν M =0 dt d ij dt M For rotations in 3d space. Need to revise.6. Before for spacetime translations we had δxµ = aµ and ∂φ therefore δφ = ∂xµ δxµ = δxµ ∂µ φ = aµ ∂µ φ. the angular momentum. = 0 corresponds to conservation of . the Lorentz transformation is δxµ = δφ = δ∂ρ φ = µ ν νx µ ν ν x ∂µ φ µ ν ν x ∂µ ∂ρ φ Now repeat same step as before. SPACETIME SYMMETRIES 25 1. Copying this. Schwabl and Greiner are best (they do J=L+S) Instead of a simple translation δxi = ai now consider a rotation δxi = ij xj . Lorentz transformations are a generalisation of this rotation. namely a δxµ = µ ν xν .
.. where the latter approximation is for inﬁnitessimal group elements αi = i .. The Xi are linearly independent Hermitian operators (there are N of them) which satisfy the Lie algebra [Xi .αN ) = eiαi Xi ≈ 1 + i i Xi + . we can write any group element as U (α1 .7 Internal Symmetries [Guidry 9192] One of the important theorems in Lie groups is the following : Theorem: Compact Lie groups can always be represented by ﬁnitedimensional unitary operators. [Tung p.26 CHAPTER 1. LAGRANGIAN FIELD THEORY 1. as in [Schwabl 272] η(x) → η (x) = eiαi Xi η(x) ≈ (1 + i i Xi )η(x) giving δη(x) = η (x) − η(x) = i i Xi η(x) ¯ Consider the Dirac equation (i ∂ − m)ψ = 0 with 4current. Xj ] = i fijk Xk where fijk are the structure constants of the group. Noether’s theorem now tells us tht we can ﬁnd a corresponding conserved charge and conserved current. . α2 . α2 . j µ = ψγ µ ψ.190] Thus using the notation U (α1 ... These group elements act on wave functions. The Noether current. the Dirac Lagrangian L is invariant under the transformation ψ → ψ = eiαi Xi ψ. / ¯ ∂ − m)ψ where ψ ≡ ψ † γ 0 . for αi = constant. ¯ This is derived from the Lagrangian L = ψ(i / Now. .173. with δxν = 0 and therefore δη(x) = ∆η(x) [Schwabl 272]. is jµ = = ∂L δηr ∂(∂ µ ηr ) ∂L i i Xi ηr ∂(∂ µ ηr ) .αN ) for an element of an Nparameter Lie group (αi are the group parameters).. This is the signiﬁcance of group theory in quantum mechanics.
INTERNAL SYMMETRIES 27 and again dropping oﬀ the constant factor i deﬁne a new curent (and throw in a minus sign so that we get a positive current in the example below) µ ji = − ∂L i Xi ηr ∂(∂µ ηr ) which obeys a continuity equation µ ∂µ ji = 0 µ Example Calculate ji for the isospin transformation eiαi Xi for the Dirac ¯ / ¯ Lagrangian L = ψ(i ∂ − m)ψ ≡ ψ(iγ µ ∂µ − m)ψ Solution µ ji = − ∂L iXi ψ ∂(∂µ ψ) and ∂L ¯ = ψ i γµ ∂(∂µ ψ) µ ¯ ji = −iψγ µ iXi ψ giving or µ ¯ ji = ψγ µ Xi ψ where Xi is the group generator.7. (Compare this to the ordinary Dirac ¯ probabilty current j µ = (ρ.1. j ) = ψγ µ ψ). .
d3 x ψ † ψ = u† u = 2E . Thus the conserved current is ¯ j µ = q ψγ µ ψ ¯ where J µ = (ρ. Write the U (1) group elements as U (θ) = eiθq where the charge q is the generator.28 CHAPTER 1. The conserved charge is [Mosel 17. BjRQM 9].34] Q = = q d3 x j 0 = q d3 x ρ ¯ d3 x ψ γ 0 ψ ¯ (Note that ρ = ψγ 0 ψ = ψ † γ 0 γ 0 ψ = ψ † ψ = Σ4 ψi 2 where ψi is each i=1 component of ψ. Thus ρ is positive deﬁnite. This is explained very clearly in [Gross 122124]. j ) = ψγ µ ψ is just the probability current for the Dirac equation. Note that if ψ is normalized so that [Strange 123] d3 x ρ = d3 x ψ † ψ = 1 then we have Q = q as required.34. LAGRANGIAN FIELD THEORY From the previous example we can readily display conservation of charge.) See also [Mosel 17. However often diﬀerent normalizations are used for the Dirac wave functions [Muirhead 72. Halzen & Martin 110]. For example [Halzen & Martin 110] have d3 x ρ = See also [Griﬃths 223].
8 1.1.8.8. the EL equations are =0 The covariant momentum density is Πµ ≡ and the canonical momentum is Π ≡ Π0 = ∂L ˙ ∂φ ∂L ∂(∂µ φ) .3 Classical ﬁeld theory ∂L ∂µ ∂(∂µ φi ) − ∂L ∂φi For classical ﬁelds φi . − ) 1. qi ) ˙ ˙ giving Hamilton’s equations ∂H = qi ˙ ∂pi − ∂H = pi ˙ ∂qi (1. A) Aµ = (Ao .2 Classical point particle mechanics The point particle canonical momentum is pi = and the EL equations are d ∂L ∂L )− =0 ( dt ∂ qi ˙ ∂qi The point particle Hamiltonian is H(pi .8.1 Summary Covariant and contravariant vectors Aµ = (Ao .8. ) ∂ ∂ µ = ( ∂t . qi ) ≡ pi qi − L(qi . SUMMARY 29 1. −A) Contravariant and covariant vectors and operators are and ∂ ∂µ = ( ∂t .6) ∂L ∂ qi ˙ 1.
j0 (x) ≡ ρ(x) If we consider the spacetime symmetry involving invariance under translation then we can derive Tµν from jµ . For the KleinGordon Lagrangian this becomes Tµν = ∂µ φ∂ν φ − gµν L The momentum is (with E ≡ H) P ν = (H.8) where dP = 0..αN ) = eiαi Xi ≈ 1 + i i Xi + ..30 CHAPTER 1. .8.. P ) = ν d3 x j0 (x) (1.4 Noether’s theorem The conserved (∂ µ jµ = 0) Noether current is jµ ≡ with Tµν ≡ ∂L ∂ν ηr − gµν L ∂(∂ µ ηr ) ∂L ∆ηr − Tµν δxν ∂(∂ µ ηr ) The conserved ( dQ = 0) charge is dt Q≡ which is just the charge density. LAGRANGIAN FIELD THEORY The energy momentum tensor is (analagous to point particle Hamiltonian) Tµν ≡ Πµ ∂ν φ − gµν L with the Hamiltonian density ˙ H ≡ T00 = Πφ − L 1.7) d3 x T 0ν = constant (1. The result for Tµν agrees with that given above. α2 . dt For internal symmetries the group elements can be written U (α1 . ..
for an internal symmetry (δxν = 0 and therefore δη = ∆η) becomes ∂L µ ji = − i Xi ηr ∂(∂µ ηr ) For the Dirac Lagrangian. SUMMARY where [Xi .1.8. with αi = constant. invariant under eiαi Xi . this becomes µ ¯ ji = ψγ µ Xi ψ . Xj ] = i fijk Xk The group elements act on wave functions η → η = eiαi Xi η ≈ (1 + i i Xi )η giving δη = η − η = i i Xi η 31 The Noether current.
1) are [Guidry. Griﬃths. Guidry.511514. 345. 3536]. For a complete introduction to covariant and contravariant vectors see [Jackson] The most commonly used metric is ηµν = 1 0 0 0 0 −1 0 0 0 0 −1 0 0 0 0 −1 This metric is used throughout the present book and is also used by the following authors: [Greiner. Aitchison and Hey. pg. pg. MTW. pg. Chang. 526531. Misner.1. pg. Jackson. pg. pg. 1213. pg. 345. pg. Sterman. 514. Mandl and Shaw. Peskin & Schroeder. Thorne and Wheeler. Mandl & Shaw. Ryder. Halzen and Martin. Ryder] . 9697. Itzykson & Zuber. Mandl and Shaw.1) are [Aitchison and Hey. 511514. Griﬃths.9 References and Notes General references for units (Section 1. pg. Another less commonly used metric is ηµν = 1 0 0 0 0 −1 0 0 0 0 −1 0 0 0 0 −1 This metric is not used in the present book. Griﬃths.1. pg. pg.32 CHAPTER 1. Kaku. Guidry. pg.2) are [Guidry. The best references for the classical ﬁeld ELE and Noether’s theorem are [Schwabl. 526531. Bjorken and Drell. Halzen and Martin. References for Natural units (Section 1. LAGRANGIAN FIELD THEORY 1. 9697. pg. 811821. pg. 3536]. 1213] and references for Geometrical units (Section 1. Halzen & Martin and Gross]. Aitchison & Hey. but it is used by [Weinberg and Muirhead].
2. 3638 33 .2 SO(2) In SO(2) the invariant is x2 + y 2 . We write x y or xi = Oij xj For small angles this is reduced to δx = θy and δy = −θx or δxi = θ where ij ij j = cos θ sin θ sin θ cosθ x y x is antisymmetric and 12 =− 21 =1 Read Kaku p.Chapter 2 Symmetries & Group theory 2.1 Elements of Group Theory SUSY nontrivially combines both spacetime and internal symmetries.
Lorentz boost transformations and also the discrete transformations of space reﬂection. SYMMETRIES & GROUP THEORY 2. time reversal and spacetime reﬂection. The general Lorentz transformation is written [Mosel 21. i. HoKim 28. U U † = 1 implying U † = U −1 . Elbaz 192. Greiner FQ 95. we impose < α β >=< α β > which means that U is unitary.e.1 Transformation Properties of Fields References: [Kaku 38.2. with expectation value < O >≡< αOα >=< α O α > This gives < O >=< αOα >=< αU † O U α > O = U †O U or O = U OU † giving the transformation rule for the operator as To summarize. 96. Now consider transformation of an operator O.34 CHAPTER 2.1) . Kaku 50] xµ = Λµν xν (2. Mosel 21] Consider a transformation U which transforms a quantum state [Elbaz l92] α >≡ U α > < α =< α U † To conserve the norm. which include spatial rotation. if a quantum state transforms as [Elbaz 192] ψ = Uψ or α >= U α > then an operator transforms as O = U OU † Fields can be grouped into diﬀerent categories [Mosel 21] according to their behavior under general Lorentz transformations.
4 Representations of SO(3) and SU(1) SO(3) leaves x2 + x2 + z 2 invariant. it just transforms in the same way as an ordinary 4vector.2) Λµν or aµν for rotation. . Read Kaku 4245. Kaku 39] φµ (x ) = aν φν (x) ≡ U φµ (x )U † µ i.3.2) as φ (x ) = U φ(x )U † = φ(x) A vector ﬁeld transforms as [HoKim 30. Kaku 38] φ (x ) = U φ(x )U † Thus a scalar ﬁeld transforms under (2. A scalar ﬁeld transforms under (2. space inversion is very nicely discussed in [HoKim 1922]. Peskin 496] 2.1) or (2.e. Thus the group U(1) can be deﬁned as the set of transformations that leave ψ ∗ φ invariant [Kaku 742. 2. boost. 741748 SO(2) can be deﬁned as the set of transformations that leave x2 + y 2 invariant. HoKim 26. REPRESENTATIONS OF SO(2) AND U(1) 35 but let’s write it more generally (in case we consider other transformations) xµ = aµν xν (2.3 Representations of SO(2) and U(1) Read Kaku 3942. Greiner 95.2. 96]. A U(1) transformation can be written ψ = eiθ ψ = U ψ.2) as φ (x ) = φ(x) [Mosel 21. This will leave the inner product ψ ∗ φ invariant. There is a homomorphism between SO(2) and U(1). Now if φ(x) is an operator then its transformation is also written as [Greiner 96.
5 Representations of SO(N) Students should read rest of chapter in Kaku.m.36 CHAPTER 2. NNN NOW do a. . SYMMETRIES & GROUP THEORY 2.
i. The two prime relativistic wave equations are the KleinGordon equation (KGE) and the Dirac equation.) Our quantum wave equation (both relativistic and nonrelativistic) is written ˆ ˆ Hψ = Eψ with ˆ ˆ ˆ H ≡T +U and ∂ ˆ E = i¯ h ∂t p2 ˆ 2m 2 ˆ Nonrelativistically we have T = − h ¯2 2m with p = −i¯ ˆ h ∂ ψ ∂t giving + U ψ = i¯ h ˆ The timeindependent Schrodinger equation simply has Eb instead of E where Eb is the binding energy. − h ¯2 2m 2 + U ψ = Eb ψ 37 .Chapter 3 Free KleinGordon Field NNN write general introduction 3.e. (However these are only valid for 1particle problems whereas the Schrodinger equation can be written for many particles.1 KleinGordon Equation Relativistic Quantum Mechanics (RQM) is the subject of studying relativistic wave equations to replace the nonrelativistic Schrodinger equation.
h 526528) ¨ φ − 2 φ + m2 φ = 0 Recall the wave equation (with y ≡ ∂ 2 y/∂x2 . ﬁrstly the operator −¯ 2 2 + m2 is nonlocal [Landau 221] h making it very diﬃcult to work with in coordinate space (but actually it’s easy in momentum space) and secondly the equation is not manifestly covariant [Gross. FREE KLEINGORDON FIELD However in special relativity we have (with c = 1) E =T +m and E 2 = p 2 + m2 giving T = p2 + m2 − m With the replacement p → −i¯ . see Halzen & Martin. 92]. the relativistic version of the free particle h ˆ (U = 0) Schrodinger equation would be (T ψ = Eb ψ) −¯ 2 h 2 + m2 − m ψ = E b ψ In the nonrelativistic case E = Eb . giving −¯ 2 h 2 + m2 ψ = Eψ which is called the Spinless Salpeter equation. 12. y ≡ ∂ 2 y/∂t2 ) ¨ y − 1 y=0 ¨ c2 2 where c is the wave velocity. 13. Squaring the Spinless Salpeter operator gives the KleinGordon equation (KGE) (−¯ 2 h 2 + m2 )φ = E 2 φ = −¯ 2 h ∂2 φ ∂t2 or (with ¯ = c = 1. There are two problems with this equation.38 CHAPTER 3. Thus the KGE (c = 1) is (with φ − 1 ¨ φ = m2 φ c2 = ∂ 2 /∂x2 ) . but relativistically E = Eb + m. pg. pg. Aitchison & Hey. pg.
2 Probability and Current The KleinGordon equation was historically rejected because it predicted a negative probability density.2. In order to see this let’s ﬁrst review probability and current for the Schrodinger equation.2. The KGE is written in manifestly covariant form as (P2 + m2 )φ = 0 which in momentum space is (using p2 → −P2 ) (p2 − m2 )φ = 0 A quick route to the KGE is with the relativistic formula p2 ≡ pµ pµ = m2 (= p2 ) giving p2 − m2 = 0 and (p2 − m2 )φ = 0 and p2 → −P2 giving (P2 +m2 )φ = 0. PROBABILITY AND CURRENT 39 which is like a massive (inhomogeneous) wave equation.1 Schrodinger equation ¯2 h 2m The free particle Schrodinger equation (SE) is − 2 ψ = i¯ h ∂ψ ∂t The complex conjugate equation is (SE∗ ) h2 2 ∗ ¯ ∂ψ ∗ ψ = −i¯ h 2m ∂t ∗ and SE∗ by ψ Multiply SE by ψ − − h ¯2 ∗ ψ 2m 2 ψ = i¯ ψ ∗ h ∂ψ ∂t . Finally.) 3. 3. Then the KG example will be easier to understand. The KGE can be written in terms of 4vectors.3. (p2 −m2 )φ = 0 and is therefore manifestly covariant. note that the KGE is a 1particle equation! −1 1 (Note that some authors [Muirhead] use g µν = and 1 1 ∂2 so have P2 ≡ 2 − ∂t2 and (P2 − m2 )φ = 0 or (p2 + m2 )φ = 0 for the KleinGordon equation.
2.40 CHAPTER 3. 3.2 KleinGordon Equation ∂2 ∂t2 The free particle KGE is (P2 + m2 )φ = 0 or (using P2 = ∂2φ − ∂t2 ∂ 2 φ∗ − ∂t2 ∂2φ − φ∗ ∂t2 2 − 2) φ + m2 φ = 0 and the complex conjugate equaiton is (KGE∗ ) 2 ∗ φ + m2 φ∗ = 0 Multiplying KGE by φ∗ and KGE∗ by φ gives φ∗ φ 2 φ + m2 φ∗ φ = 0 ∂ 2 φ∗ − φ 2 φ∗ + m2 φφ∗ = 0 ∂t2 and subtract these equations to give φ∗ ∂2φ ∂ 2 φ∗ − φ 2 − φ∗ ∂t2 ∂t 2 φ+φ 2 ∗ φ =0 . FREE KLEINGORDON FIELD − h ¯2 ∂ψ ∗ h ψ 2 ψ ∗ = −i¯ ψ 2m ∂t and subtract these equations to give − h ¯2 ∗ (ψ 2m h ¯2 2m 2 ψ−ψ 2 ψ ∗ ) = i¯ ψ ∗ h = i¯ h ∂ ∗ (ψ ψ) ∂t ∂ψ ∂ψ ∗ +ψ ∂t ∂t = − or · [ψ ∗ ψ − ( ψ ∗ )ψ] · [ψ ∗ ψ − ( ψ ∗ )ψ] = 0 ∂ρ ∂t ∂ ∗ h ¯ (ψ ψ) + ∂t 2mi which is just the continuity equation ρ ≡ ψ∗ψ j≡ + · j = 0 if h ¯ [ψ ∗ ψ − ( ψ ∗ )ψ] 2mi which are the probability density and current for the Schrodinger equation.
e. 227] Notice how this problematic ρ comes about because the KGE is 2nd order ∂ in time. Kaku. i. We have ∂ρ and ρ itself constains ∂t . See [Landau.3. Thus for ρ≡ + ∂φ i¯ h ∂φ∗ φ∗ −φ 2m ∂t ∂t The problem with ρ (in both expressions above) is that it is not positive deﬁnite and therefore cannot be interpreted as a probability density.3. pg. Let’s remind outselves of these results. This is one reason why the KGE was discarded. note that we can form a 4vector KG current jµ = i¯ h ∗ µ (φ ∂ φ − φ∂ µ φ∗ ) 2m 3. Chapter 3] Some of the key results for the free and real KleinGordon ﬁeld were worked out in an Example in Chapter 1. . j≡ which is −¯ h 2mi h ¯ φ∗ φ − φ φ∗ 2mi = ∂ρ ∂t −i¯ ∗ h φ φ − φ φ∗ 2m · j = 0 to hold we must have times j above. Chapter 13. By the wave. this does not happen with ∂t the SE or DE.3 Classical Field Theory Reference: [Schwabl. CLASSICAL FIELD THEORY = ∂ ∂φ ∂φ∗ φ∗ −φ ∂t ∂t ∂t − · φ∗ φ − φ φ∗ · j = 0 if 41 which is the continuity equation ∂ρ/∂t + ρ ≡ φ∗ ∂φ ∂φ∗ −φ ∂t ∂t or j µ = φ∗ ∂ µ φ − φ∂ µ φ∗ j ≡ φ φ∗ − φ∗ φ but to get this to match the SE wave function we should deﬁne j in the same way. (Note: because ρ can be either positive or negative it can be interpreted as a charge density.
to be determined later.42 CHAPTER 3. Jose and Saletin 589] As with the nonrelativistic case we expand plane wave states as φ(x. Substitute φ(r. t) = but now with ˜ dk = Nk d3 k where Nk is a normalization constant. the momentum operator of the KleinGordon ﬁeld is (see Problems) P =− ˙ d3 x φ(x) φ(x) 3. FREE KLEINGORDON FIELD The massive KleinGordon Lagrangian was 1 LKG = (∂µ φ∂ µ φ − m2 φ2 ) 2 which gave the equation of motion in position and momentum space as (P2 + m2 )φ = 0 (p2 − m2 )φ = 0 The covariant momentum density was Πµ = ∂ µ φ giving the canonical momentum ˙ Π ≡ Πo = φ(x) The Hamiltonian density was (with H ≡ d3 x H) 1 H = [Π2 + ( φ)2 + m2 φ2 ] 2 Finally. t)eik·x + m2 φ = 0 . t) into the KGE P 2 + m2 φ = 0 2 ∂t − 2 ˜ dk a(k.4 Fourier Expansion & Momentum Space [See Greiner FQ 76 .
t) is real. with Auxilliary equation r2 + ω2 = 0 or r = ± − iω The two solutions ± are crucial ! They can be interpreted as positive and negative energy. Thus [Teller. In the nonrelativistic (NR) case we only got one solution. t) giving ˜ dk a(k)ei(k·x−ωt) + c(k)ei(k·x+ωt) = Rewrite this as ˜ dk a(k)ei(k·x−ωt) + c(−k)e−i(k·x−ωt) = ˜ dk a† (k)e−i(k·x−ωt) + c† (−k)ei(k·x−ωt) ˜ dk a† (k)e−i(k·x−ωt) + c† (k)e−i(k·x+ωt) . t) = ˜ dk a(k)ei(k·x−ωt) + c(k)ei(k·x+ωt) √ (Remember that ω ≡ ω(k) = k 2 + m2 ) Now the Schrodinger wave function ψ(x. FOURIER EXPANSION & MOMENTUM SPACE giving ˜ ¨ dk a + k 2 a + m2 a eik·x = 0 Deﬁning ω ≡ ω(k) = √ 43 k 2 + m2 and requiring the integrand to be zero gives a + ω2a = 0 ¨ which is a 2nd order diﬀerential equation. 67] a(k.4.) Thus [Greiner. 77] φ(x. or as particle and antiparticle. pg. pg. t) = φ∗ (x. but the KGE does not. t) = c(k)eiωt + a(k)e−iωt giving our original expansion as φ(x. (The Schrodinger equation has an i in it. t) = φ† (x. t) is complex but the KG wave function φ(x.3.
Mandl and Shaw 44. t) = ˜ dk a(k)ei(k·x−ωt) + c(−k)e−i(k·x−ωt) gives [same conventions as Kaku 152. Schwabl 278] φ(x. continuum normalization] where k · x ≡ k µ xµ = k o xo − k · x = ωt − k · x. t) ≡ φ+ (x) + φ− (x) = = = k (3.) Obviously then c(−k) = a† (k) c† (−k) = a(k) Rewriting our expansion as φ(x.1) ˜ dk a(k)ei(k·x−ωt) + a† (k)e−i(k·x−ωt) ˜ dk a(k)e−ik·x + a† (k)eik·x √ 1 a e−ik·x + a† eik·x k 2ωV k [see Greiner FQ 79 for box vs. We are simply reversing the direction (magnitude) of k. This does not aﬀect the volume element d3 k. Jose & Saletan 588].2) ≡ Π+ (x) + Π− (x) Note that we are still studying classical ﬁeld theory when we Fourier expand these classical ﬁelds [Goldstein 568. The conjugate momentum is ˙ Π(x. FREE KLEINGORDON FIELD where we have made the substitution k → −k in two terms. t) = φ(x.44 CHAPTER 3. k is the magnitude of k. . it cannot. (This does not mean k → −k. t) = −i = −i ˜ dk ω a(k)ei(k·x−ωt) − a† (k)e−i(k·x−ωt) ˜ dk ω a(k)e−ik·x − a† (k)eik·x (3.
t)] = [Π(x. Π− (x . Π(x . t)] = iδ(x − x ) and [φ(x. t). t).5. t). a† (k )] = 0 (3. t)] = [φ+ (x. a(k )] = [a† (k). Π+ (x . φ(x .3) 3. a† Commutators [Greiner. a† ] = 1 To develop the QFT we impose similar relations. t). t). Π(x . a† (k )]e−i(k·x−k ·x ) + [a† (k). t). t)] = [φ+ (x.5) (3. We expect the usual results [a(k).5 KleinGordon QFT Recall the commutation relations from nonrelativistic quantum mechanics. t).6) (3. KLEINGORDON QFT 45 3. namely [x.3. . but this time for ﬁelds.1 Indirect Derivation of a. Π− (x . t)] = 0 where we had ˙ Π=φ We now need to ﬁnd the commutation relations for a(k) and a† (k). t)] + [φ− (x. t)] = −i ˜ dk ˜ dk ω [a(k). a† (k )] = δ(k − k ) and [a(k). pg. t)] + [φ+ (x. t) + Π− (x .4) (3. Π+ (x . The equal time commutation relations [φ(x. a(k )]e−i(k·x+k ·x ) − [a(k). t).5. a† (k )]ei(k·x+k ·x ) ] where k · x = ωt − k · x and t ≡ t and ω ≡ ω(k ) = k 2 + m2 . t). p] = i¯ h and [a. t) + φ− (x. a(k )]ei(k·x−k ·x ) − [a† (k). Π+ (x . 77] To check whether our expection above is correct evaluate [φ(x. Π(x . t)] +[φ− (x.
6) gives [φ(x. 77.46 CHAPTER 3.5) and (3. Π(x . Kaku 6465] d3 k = d4 k δ(k 2 − m2 ) θ(k 0 ) 2ω gives ˜ dk = 2ω d4 kδ(k 2 − m2 ) θ(k o ) (2π)3 (which is diﬀerent from IZ 114. Π(x . t). pg. a† (k )] = (2π)3 2ω k 1 (2π)3 2ω in which δ 3 (k − k ) [Greiner. Here we have used the result. Using the result [IZ114. but same as Kaku). footnote]. . t)] = −i ˜ dk ˜ dk ω −δ(k − k )e−i(k·x−k ·x ) −δ(k − k)ei(k·x−k ·x ) =i 2 d3 k Nk ω eik·(x−x ) + e−ik·(x−x ) Now if [Greiner. 1 (2π)3 d3 k e±ik·(x−x ) = δ(x − x ) With the above normalization we have ˜ dk ≡ d3 k 2ω(2π)3 (which is diﬀerent to IZ114. t). t)] = i 1 ω eik·(x−x ) + e−ik·(x−x ) 2ω(2π)3 = iδ(x − x ) d3 k 1 2ω(2π)3 as required. but same as Kaku) Note: Some authors [IZ114] use the normalization Nk = case [a(k). FREE KLEINGORDON FIELD Inserting (3. pg. 77] Nk = then we obtain [φ(x.
5. KLEINGORDON QFT 47 3.6) [do Problem 5.1) and (3.2 Direct Derivation of a.2) gives [see Problems] ak = ⇒ a† = k 1 2ω(2π)3 1 2ω(2π)3 d3 x eik·x [ωφ(x) + iΠ(x)] d3 x e−ik·x [ωφ† (x) − iΠ† (x)] where ak ≡ a(k) and φ(x) ≡ φ(x.5) and (3. t) and k · x ≡ ωt − k · x and by direct evaluation of the commutators we arrive at (3.5. The commutators are then derived directly. a† Commutators The best way to obtain the commutators directly is to invert the Fourier expansions to obtain a and a† in terms of φ and Π.7] But remember for KGE φ is real and therefore φ = φ† and Π = Π† giving a† = k 1 2ω(2π)3 d3 x e−ik·x [ωφ(x) − iΠ(x)] 3.3 KleinGordon QFT Hamiltonian The 2nd quantized ﬁeld Hamiltonian is (see Problems) H= and using the commutator [ ak . a† ] = δ(k − k ) k ω d3 k (a† ak + ak a† ) k 2 k gives H= or k d3 k Nk + Nk + 1 ω 2 1 ω 2 with Nk ≡ a† ak k (see Problems) .3. Inverting (3.5.
FREE KLEINGORDON FIELD We can also calculate the momentum P as [Kaku 66. Mosel 27.5. Schwabl 280] The previously derived Hamiltonian is actually inﬁnite! This is beacuse the term 1 1 d3 k ω = d 3 k k 2 + m2 = ∞ 2 2 This is one of the ﬁrst (of many) places where QFT gives inﬁnite answers. In a normal ordered product all annihilation operators are placed to the right hand side of all creation operators. Schwabl 279] P = or k d3 k Nk + Nk + 1 k 2 1 k 2 3. a† ] = δ(k − k ) k d3 k(a† ak + ak a† ) k k ω 2 . Now because only energy diﬀerences are observable. Two colons :: are used to denote a normal ordered product. is to introduce the idea of normal order. 67.48 CHAPTER 3. and rewrite the Hamiltonian as H= d3 k Nk ω. For example [Schwabl 280] : ak1 ak2 a† : = a† ak1 ak2 k3 k3 : a† ak + ak a† : = 2a† ak k k k In calculating the Hamiltonian (see Problems) we arrived at H= and using the commutator [ ak .4 Normal order References [Kaku68. A formal way to always get rid of these inﬁnite terms (there is also one in the previous expression for the momentum P ). we are free to simply throw away the inﬁnite piece.
3. KLEINGORDON QFT gave H = = d3 k(a† ak + a† ak + 1) k k 49 ω 2 d3 k(a† ak k + 1/2)ω which is inﬁnite. [Schwabl 280]. (nnn Are we back to a classical theory?) .5. If we deﬁne H to be normal ordered then :H: = = = d3 k : (a† ak + ak a† ) : k k ω 2 d3 k(a† ak + a† ak ) k k ω 2 d3 ka† ak ω k which is ﬁnite. Note that normal ordering is equivalent to treating the boson operators as if they had vanishing commutator.
nk2 have momentum k2 etc. Equation (3.7) is interpreted as a many particle state h where nk1 have momentum k1 .5. These are just the manybody states introduced earlier [Bergstrom & Goobar 289]  · · · n ki · · · n kj · · · = nki (3. FREE KLEINGORDON FIELD 3.7) i with nki = 1 nk ! a† nk k 0 where the vacuum state is deﬁned via ak 0 = 0 The physical interpretation [Bergstrom & Goobar 290] is provided by H · · · nki · · · nkj · · · = k nk (k) · · · nki · · · nkj · · · where (k) = ¯ ω(k). we need something for them to operate on.5 Wave Function [Kaku 6869] Now that we have expanded the KG ﬁeld and Hamiltonian in terms of creation and destruction operators. Thus a 1particle state is written 1k ≡ k = a† 0 k or 1k  ≡ k = 0ak The states are normalized as kk = δ(k − k ) giving 0ak a† 0 = δ(k − k ).50 CHAPTER 3. k .
6. This is solved with the Green function method by deﬁning a propagator ∆F (x − y) as (P2 + m2 )∆F (x − y) ≡ −δ 4 (x − y) so that the solution is φ(x) = φ0 (x) − d4 y ∆F (x − y)J(y) where φ0 (x) is the solution with J = 0. PROPAGATOR THEORY 51 3. The Problems show that ∆F (k) = k2 1 − m2 . Kaku.6 Propagator Theory [Halzen and Martin 145150.3. (see Problems) Deﬁne the Fourier transform d4 k −ik·(x−y) ∆F (x − y) ≡ e ∆F (k) (2π)4 Our usual method of solution for propagators or Green functions is to solve for ∆F (k) and then do a contour integral to get ∆F (x − y) rather than solving for ∆F (x − y) directly. Bj RQM Chapter 6] The KleinGordon equation for a free particle is (P2 + m2 )φ = 0 or with the replacement P2 → −p2 in momentum space (p2 − m2 )φ = 0 Let’s write the nonfree KGE as (P2 + m2 )φ = J(x) where J(x) is referred to as a source term.
52
CHAPTER 3. FREE KLEINGORDON FIELD
Example Derive the momentum space Green function for the Schrodinger equation. Solution The free particle Schrodinger equation is − h ¯2 2m
2
− i¯ h
∂ ψ=0 ∂t
and with the inclusion of a source this is − or + h ¯2 2m h ¯2 2m
2
− i¯ h
∂ ψ = J(x) ≡ −U (x)ψ(x, t) ∂t ∂ ψ = −J(x) = U (x)ψ(x) ∂t
2
+ i¯ h
Deﬁne the Green function h ¯2 2m
2
+ i¯ h
∂ G0 (x − x ) = δ 4 (x − x ) ∂t
and deﬁne the Fourier transform G0 (x − x ) = d4 p −ip·(x−x ) e G0 (p) (2π)4
¯ Now operate on this to give, and using pµ ≡ (ω, p), with h = 1 h ¯2 2m
2
+ i¯ h
∂ G0 (x − x ) = ∂t = δ 4 (x − x ) =
d4 p 1 (−ip)2 + i(−iω)] e−ip·(x−x ) G0 (p) 4 2m (2π) d4 p p2 − + ω e−ip·(x−x ) G0 (p) (2π)4 2m
but recall that δ 4 (x − y) = which implies G0 (p) = 1 ω − p 2 /2m d4 k −ik·(x−y) e (2π)4
3.6. PROPAGATOR THEORY
53
Example Derive the position space Green function leaving the pole on the Real axis. Solution Substituting for G0 (p) into G0 (x − x ) we need to evaluate (¯ = 1) h G0 (x − x ) = = d4 p 1 e−ip·(x−x ) 4 ω − p 2 /2m (2π) d3 p ip·(x−x ) e (2π)3
∞ −∞
dω e−iω(t−t ) 2π ω − p 2 /2m
and we see that the integrand is singular at ω = p 2 /2m. This simple pole is shown in the ﬁgure. We need to decide whether to integrate in the upper half plane (UHP) or the lower half plane (LHP). This is dictated by the boundary conditions as follows. Write ω ≡ Re ω + i Im ω, so that e−iω(t−t ) = ei Re ω(t−t ) e+ Im ω(t−t ) For t − t > 0, the term eIm ω(t−t ) will blow up for Im ω > 0 but will go to zero for Im ω < 0. Thus the boundary condition t−t > 0 dictates we integrate in the LHP. Similarly for t−t < 0 we use the UHP. This is shown in the ﬁgure. [See also Halzen and Martin 148]
Im ω
Im ω t < t'
ω=ω0=p2/2m X
C1 C
t > t'
Re ω
X
Re ω
Now apply the Cauchy Residue Theorem, and remember counterclockwise integration is a positive sign. In the left ﬁgure the
54
CHAPTER 3. FREE KLEINGORDON FIELD coutour does not enclose any poles so that
ω0 −δ ∞
0= and with
C
−∞
+
C1
+
ω0 +δ
+
C
= 0 (Jordan’s lemma) and lim we have
δ→0 ∞ −∞
=−
C1
Thus for t > t G0 (x − x )t>t = − + πi = − i 2 d3 p ip·(x−x ) 1 −iω0 (t−t ) e e (2π)3 2π d3 p ip·(x−x ) −iω0 (t−t ) e e (2π)3
with ω0 ≡ p 2 /2m. Similarly for t < t we obtain G0 (x − x )t<t = − − πi = + i 2 d3 p ip·(x−x ) 1 −iω0 (t−t ) e e (2π)3 2π d3 p ip·(x−x ) −iω0 (t−t ) e e (2π)3
But this violates causality ! If the wave pulse is sent out at t we do expect a signal at a later time t > t but certainly not at an earlier time t < t .
Because of the singularity, the integral is not well deﬁned until we specify the limiting process. Let’s now try a diﬀerent method for evaluating the integral that is consistent with the boundary condition, i.e. we want G0 (x − x )t<t = 0.
3.i ε C ∞ −∞ t>t’ X ω0 . . G0 (x − x )t<t = 0 For t < t we get because no poles are enclosed and this now obeys the boundary condition. This is shown in the ﬁgure.i ε For t > t we have get G0 (x − x )t>t + C = 2πiΣ Residues and with C = 0 we = −2πi = −i e−iω(t−t ) d3 p ip·(x−x ) ∞ dω e 2 (2π)3 −∞ 2π ω − p /2m + i 3p d eip·(x−x ) e−iω0 (t−t ) (2π)3 with lim e →0 (t−t ) = 0. →0 Solution We can shift the pole either above or below the real axis. which is exactly double our previous answer. t<t’ X X ω0 . PROPAGATOR THEORY 55 Example Evaluate the position space Green function by shifting the pole oﬀ the real axis by a small amount . and take lim .6. We choose to shift it below because then the upper contour will enclose no poles and will give a zero integral consistent with our boundary conditions.
417. Halzen and Martin 145150] The key integral that we have been considering is of the form ∞ I= −∞ dω e−iωt ω − ω0 and with the pole shifted below the real axis this is ∞ I( ) = −∞ dω e−iωt ω − ω0 + i and let’s summarize our results as It>0 = −πi e−iω0 t It<0 = +πi e−iω0 t and I( )t>0 = −2π i e−iω0 t I( )t<0 = 0 When we leave the pole on the real axis the integral ∞ −∞ is actually the Cauchy Principal Value deﬁned as [Arfken 4th ed. 418] the Cauchy Principal value is actually a cancelling process.. Greiner QED 27. our ﬁrst method with the pole left on the axis) and the i Prescription are related. This is because in the vicinity of the simple pole we have f (x) ≈ a x − x0 which is odd relative to x0 and so the large singular regions cancel out [Arfken 4th ed..56 CHAPTER 3. Consider [Merzbacher. 418]. Cushing 315. Landau 88] ∞ P −∞ f (x)dx ≡ lim x0 −δ −∞ ∞ δ→0 f (x)dx + x0 +δ f (x)dx with the pole at x = x0 . Landau 8488. 427. p. The Principal Value Prescription (i. old edition] 1 ω i ω i = 2 = 2 ω±i ω + 2 ω + 2 ω2 + 2 . See also [Landau 8588]. 417. FREE KLEINGORDON FIELD An excellent discussion of all of these issues can be found in [Arfken 4th ed. Bj RQM 85. 417. As discussed by [Arfken 417.e.
Thus we can write [Merzbacher.. p. PROPAGATOR THEORY and using [Merzbacher 3rd ed.. p. old ed. 85] ∞ lim →0 −∞ f (ω) ω ω2 + dω = lim 2 ∞ dω dω + f (ω) →0 ω ω −∞ ωdω + f (ω) 2 ω + 2 − ∞ ωdω dω = P f (ω) + f (0) lim 2+ 2 →0 − ω ω −∞ ∞ dω = P f (ω) +0 ω −∞ − f (ω) where the last integral vanishes because the integrand is an odd function of ω. 85] lim 1 1 =P →0 ω ± i ω iπ δ(ω) and we can also write [Cushing 315] (with the pole at ω = ω0 instead of ω = 0) f (ω)dω f (ω)dω P = ± iπ f (ω0 ) ω − ω0 ω − ω0 ± i .3. 631] δ(x) = we have 1 lim π →0+ x2 + 2 57 1 ω = 2 ω±i ω + 2 iπ δ(ω) 1 The ﬁrst term on the right hand side becomes ω as → 0 except if ω = 0.6. p.. If f (ω) is a well behaved function we have [Merzbacher. old ed.
For t < 0 we had P and e−iωt dω = +πi e−iω0 t ω − ω0 e−iωt dω =0 ω − ω0 + i e−iωt dω + iπ f (ω0 ) = 0 + iπ e−iω0 t ω − ω0 + i e−iωt dω = P ω − ω0 also in agreement with the formula. Solution For t > 0 we had P and e−iωt dω = −πi e−iω0 t ω − ω0 e−iωt dω = −2πi e−iω0 t ω − ω0 + i e−iωt dω + iπ f (ω0 ) = −2πi e−iω0 t + πi e−iω0 t ω − ω0 + i = −πi e−iω0 t e−iωt dω = P ω − ω0 Now f (ω0 ) = e−iω0 t .58 CHAPTER 3. so that in agreement with the formula. so that . FREE KLEINGORDON FIELD Example Verify the above formula for our previous integral e−iωt I = dω ω−ω0 .
PROPAGATOR THEORY 59 Example Show that the θ function deﬁned as θ(t − t ) = can be written as θ(t − t ) = lim −1 →0 2πi ∞ −∞ 1 if t > t 0 if t < t dω e−iω(t−t ) ω+i ∞ −∞ Solution We showed previously that with I( ) ≡ we get I( )t>0 = −2πi e−iω0 t I( )t<0 = 0 Replacing t with t − t and ω0 = 0 gives I = It>t = −2πi It<t = 0 e dω ω−ω0 +i −iωt dω e −iω(t−t ) ω+i and which veriﬁes the above integral representation of θ.6. Now let’s ﬁnish the job of evaluating the position space Green function [Bj RQM 84.3. Greiner QED 27. Kaku 7374] .
in terms of θ(t − t ). Also write the answer in terms of the plane wave states φp (x) ≡ e−ip·x (2π)3/2 Solution To satisfy causality we use the i prescription to evaluate the integral which gave G0 (x − x )t>t = −i and G0 (x − x )t<t = 0 with ω0 ≡ p 2 /2m. consistent with the causality boundary condition.60 CHAPTER 3. . Kaku 74] Finally the d3 p integrating gives (see Problems) [Bj RQM 86] G0 (x − x ) = −i m 2πi(t − t ) 3/2 1 −ip·x ip·x 1 −ip·(x−x ) e e = e 3 (2π) (2π)3 1 −i[ω(t−t )−p·(x−x )] e (2π)3 d3 p φp (x)φ∗ (x ) p e imx−x 2 2(t−t ) θ(t − t ) See footnote in [Bj RQM 86] discussing the Schrodinger equation and the Diﬀusion equation. FREE KLEINGORDON FIELD Example Evaluate the position space Green function. This can be written d3 p ip·(x−x )−iω0 (t−t ) e θ(t − t ) (2π)3 d3 p ip·(x−x ) −iω0 (t−t ) e e (2π)3 G0 (x − x ) = −i Using the plane wave states we have φp (x)φ∗ (x ) = p = giving G0 (x − x ) = −iθ(t − t ) [Bj RQM 86.
PROPAGATOR THEORY KleinGordon Propagator 61 1 We originally wrote the KleinGordon propagator ∆F (k) = k2 −m2 but then digressed on a lengthy discussion of the Schrodinger propagator in order to illustrate the integration techniques. We now return to the KleinGordon propagator.(x−y) e ∆F (k) (2π)4 d3 k (2π)3 2ωk ˜ ˜ Performing the integration as before we get. Consider the following i prescription ∆F (k) = Write this as k2 k2 1 − m2 + i 1 1 1 = 2 = 2+i 2 − k 2 − m2 + i −m k0 − E 2 + i k0 with E ≡ + k 2 + m2 We can show (see Problems) 1 1 1 1 1 = = 2 + k 2 − m2 + i 2k0 k0 − E + i k0 + E − i k0 − E 2 + i Recall that the momentum space propagator ∆F (k) was deﬁned in terms of the Fourier transform ∆F (x − y) = d4 k −ik. with dk ≡ lems) (see Prob ∆F (x − x ) = −iθ(t − t ) = −iθ(t − t ) ˜ ˜ dk e−ik·(x−x ) − iθ(t − t) ˜ ˜ dk φk (x)φ∗ (x ) − iθ(t − t) k ˜ ˜ dk eik·(x−x ) ˜ ˜ dk φ∗ (x)φk (x ) k .3.6.
3rd and 4th terms are zero. Evaluate 0φ(x)φ(x )0 d3 k d3 k 0 √ √ (a e−ik·x + a† eik·x )(ak e−ik ·x + a† eik ·x )0 = k k (2π)3 2ω 2ω k = ˜ ˜ dk dk 0ak ak e−i(k·x+k ·x ) + ak a† e−i(k·x−k ·x ) k +a† ak k e i(k·x−k ·x ) + a† a† k k e i(k·x+k ·x ) 0 However ak 0 = 0 and 0a† = 0.) Solution The time ordered product is T A(t1 )B(t2 ) ≡ A(t1 )B(t2 ) if t1 > t2 B(t2 )A(t1 ) if t2 > t1 ≡ θ(t1 − t2 )A(t1 )B(t2 ) + θ(t2 − t1 )B(t2 )A(t1 ) Thus for the ﬁelds T φ(x)φ(x ) = θ(t − t )φ(x)φ(x ) + θ(t − t)φ(x )φ(x) or 0T φ(x)φ(x )0 = θ(t − t ) 0φ(x)φ(x )0 + θ(t − t) 0φ(x )φ(x)0 Thus we want to evaluate 0φ(x)φ(x )0 and 0φ(x )φ(x)0 . Thus the 1st. FREE KLEINGORDON FIELD Example Show that i∆F (x − x ) = 0T φ(x)φ(x )0 (Note this is also equal to = 0T φ(x)φ† (x )0 for Hermitian ﬁelds [BjRQF42]. Now φ(x) = ˜ with dk ≡ √ d3 k (2π)3 2ω ˜ dk (ak e−ik·x + a† eik·x ) k and ω ≡ k 2 + m2 and ω ≡ k 2 + m2 .62 CHAPTER 3. k We are only left with 0ak a† 0 = δ(k − k ) k .
6.(x−x )−iω(t−t ) e 2ω d3 k −ik·(x−x ) e 2ω 63 . PROPAGATOR THEORY Continuing 0φ(x)φ(x )0 = = = Obviously we also have 0φ(x )φ(x)0 = = Thus 0T φ(x)φ(x )0 = θ(t − t ) 0φ(x)φ(x )0 + θ(t − t) 0φ(x )φ(x)0 d3 k d3 k = θ(t − t ) e−ik·(x−x ) + θ(t − t) eik·(x−x ) (2π)3 2ω (2π)3 2ω = i∆F (x − x ) 1 (2π)3 1 (2π)3 d3 k −ik·(x −x) e 2ω d3 k ik·(x−x ) e 2ω 1 (2π)3 1 (2π)3 1 (2π)3 d3 k d3 k √ √ δ(k − k )e−i(ωt−k·x−ω t +k ·x ) 2ω 2ω d3 k ik.3.
Thus one has [Mandl & Shaw 51] ∆(x) = ∆+ (x) + ∆− (x) = = −i (2π)3 −1 (2π)3 d3 k sin k · x ωk d4 k δ(k 2 − m2 ) (k0 )e−ik·x with (k0 ) = +1 for k0 > 0 −1 for k0 < 0 − (x) = −∆+ (−x) Note also that ∆ These propagators are related to the Feynman propagator by [Mandl and Shaw 54] ∆F (x) = θ(t)∆+ (x) − θ(−t)∆− (x) k0 k0  = or [Schwabl 283] ∆F (x − x ) = θ(t − t )∆+ (x − x ) − θ(t − t)∆− (x − x ) which can be written [Schwabl 283.64 Other Propagators Recall that φ(x) = CHAPTER 3. φ (y)] = ± d3 k e (2π)3 2ωk ik·(x−y) where ωk ≡ + k 2 + m2 giving (see Problems) i∆(x − y) ≡ [φ(x). FREE KLEINGORDON FIELD ˜ dk[ak e−ik·x + a† eik·x ] k + − ≡ φ (x) + φ (x) [Kaku 152. Mandl & Shaw 44] From this we can deﬁne other propagators (see Problems) i∆± (x − y) ≡ [φ± (x). φ(y)] = i∆+ (x − y) + i∆− (x − y) or just ∆(x) = ∆+ (x) + ∆− (x). Mandl and Shaw 54] ∆F (x) = ±∆± (x) for t > 0 < .
φ− (x )] = 0φ(x)φ(x )0 0[φ+ (x). Solution i∆F (x − x ) = 0T φ(x)φ(x )0 = θ(t − t ) 0φ(x)φ(x )0 + θ(t − t) 0φ(x )φ(x)0 Thus to do this problem we really need to show that 0φ(x)φ(x )0 = [φ+ (x). Commutators are just cnumbers. classical functions or numbers. Therefore [φ+ (x). φ− (x )] and 0φ(x )φ(x)0 = −[φ− (x).6. Now φ = φ+ + φ− .3. PROPAGATOR THEORY 65 Example Show that ∆F (x − x ) = θ(t − t )∆+ (x − x ) − θ(t − t)∆− (x − x ) where i∆F (x − x ) = 0T φ(x)φ(x )0 and i∆± (x − y) ≡ [φ± (x). The question is. (Proof of the other relation follows similarly). how do these relations come about ? The answer is easy. φ (y)]. φ+ (x )] Let’s only do the ﬁrst of these. φ− (x )]0 0φ+ (x)φ− (x )0 − 0φ− (x )φ+ (x)0 0φ+ (x)φ− (x )0 as required. φ− (x )] = = = because φ+ 0 = 0. . Thus φ0 = φ+ 0 + φ− 0 = 0 + φ− 0 = φ− 0 and similarly 0φ = 0φ+ giving [φ+ (x). Thus 0c0 = c 00 = c for any cnumber.e. i.
66
CHAPTER 3. FREE KLEINGORDON FIELD
3.7
Complex KleinGordon Field
[Schwabl 285; Kaku 69; Huang 2425] [W. Pauli and V. F. Weisskopf, Helv. Phys. Acta, 7, 709 (1934)] Charged particles cannot be described with a real scalar ﬁeld because it only has one component, whereas a minimum of two components is needed to describe charge. Deﬁne 1 1 φ ≡ √ (φ1 + iφ2 ) and φ† ≡ √ (φ1 − iφ2 ) 2 2
1 where φ1 and φ2 are real ﬁelds. The normalization factor √2 is chosen so that φi has the same renormalization as the real scalar ﬁeld discussed before. A suitable classical Lagrangian is (with φ2 = φφ† ) [Huang 24]
L = ∂µ φ† ∂ µ φ − m2 φ2 1 = 2
2
(∂µ φi ∂ µ φi − m2 φ2 ) i
i=1
which is just the sum of the Lagrangian for the real ﬁelds. The EulerLagrange equations give the equations of motion [Schwabl 285] (P2 + m2 )φ = 0 and (P2 + m2 )φ† = 0 with the conjugate momenta being [Kaku 70; Schwabl 285] ˙ ˙ Π = φ† and Π† = φ with the equal time commutation relations [Kaku 70; Schwabl 285] ˙ [φ(x, t), Π(x , t)] = [φ(x, t), φ† (x , t)] = iδ(x − x ) ˙ [φ† (x, t), Π† (x , t)] = [φ† (x, t), φ(x , t)] = iδ(x − x ) and [Schwabl 285] [φ(x, t), φ(x , t)] = [Π(x, t), Π(x , t)] = 0 [φ† (x, t), φ† (x , t)] = [Π† (x, t), Π† (x , t)] = 0 We Fourier expand each component φi (x) exactly as before [Kaku 70; Huang 24] ˜ φi (x) = dk (a e−ik·x + a† eik·x )
ik ik
3.7. COMPLEX KLEINGORDON FIELD where the commutation relations are [Kaku 70; Huang 25] [aik , a† ] = δ(k − k )δij
jk
67
However rather than treating the i ﬁelds separately, we can combine them as we did with the wave functions φ(x). Deﬁne [Kaku 70; Huang 25] 1 1 ak ≡ √ (a1k + ia2k ) and bk ≡ √ (a1k − ia2k ) 2 2 so that the Fourier expansion now becomes [Huang 25; Schwabl 285] φ(x) = ˜ dk (ak e−ik·x + b† eik·x )
k
(Exercise: Prove this result) and for the other ﬁeld, obviously it is φ† (x) = ˜ dk(a† eik·x + bk e−ik·x )
k
The new commutation relations are [Kaku 70; Schwabl 286; Huang 25] [ak , a† ] = [bk , b† ] = δkk
k k
and
[ak , ak ] = [bk , bk ] = [ak , bk ] = [ak , b† ] = 0
k
There are now two occupationnumber operators, for particles a and for particles b [Schwabl 286] Nak ≡ a† ak and Nbk ≡ b† bk
k k
where and ak create and annihilate a particles and b† and bk create and k annihilate b particles. The vacuum state is deﬁned by [Schwabl 286] ak 0 = bk 0 = 0 The four momentum is [Schwabl 286; Mandl and Shaw 49; Greiner FQ 93] P µ = (H, P ) =
k
a† k
k µ (Nak + Nbk )
giving H=
k
ωk (Nak + Nbk )
and P =
k
k(Nak + Nbk )
68
CHAPTER 3. FREE KLEINGORDON FIELD
3.7.1
Charge and Complex Scalar Field
1 µ 2 (∂µ φ∂ φ
Recall the real scalar ﬁeld Lagrangian L = complex scalar ﬁeld Lagrangian L = ∂µ φ† ∂ µ φ − m2 φ2 = 1 2
2
− m2 φ2 ) and the
(∂µ φi ∂ µ φi − m2 φ2 ) i
i=1
with φ2 ≡ φφ† . The complex scalar ﬁeld Lagrangian (and action) is invariant under the transformation φ(x) → eiqθ φ(x) φ† (x) → e−iqθ φ† (x)
which generates a U (1) symmetry [Kaku 70]. Recall that before we had (for small αi = i ) ηr (x) → ηr (x) = eiαi Xi ηr (x) ≈ (1 + i i Xi )ηr (x) giving δηr (x) = ηr (x) − ηr (x) = i i Xi ηr (x) Our two ﬁelds are η1 ≡ φ and η2 ≡ φ† giving (with small θ = ) δφ(x) = iq φ(x) Recall the Noether current jµ ≡ with Tµν ≡ ∂L ∂ν ηr − gµν L ∂(∂ µ ηr ) ∂ηr δxµ ∂xµ ∂L ∆ηr − Tµν δxν ∂(∂ µ ηr ) and δφ† (x) = −iq φ† (x)
Also recall the relation between local and total variations δηr (x) = ∆ηr (x) −
For δxν = 0, we have δη(x) = ∆η(x) [Schwabl 272] and therefore jµ = = = ∂L δηr ∂(∂ µ ηr ) ∂L ∂L δφ + δφ† µ φ) ∂(∂ ∂(∂ µ φ† ) ∂L ∂L iq φ + (−i)q φ† µ φ) ∂(∂ ∂(∂ µ φ† )
Also note that “charge” is used in a generic sense [Huang 26] since electromagnetic coupling has not yet been turned on. Mosel 19. Mosel 19] dt Q = = q = q d3 x jo = iq k ˙ ˙ d3 x(φφ† − φ† φ) k d3 k(a† ak − b† bk ) d3 k(Nak − Nbk ) (Exercise: Prove this result). j ) = ψγ µ ψ ) Q = = q d3 x j 0 = q d3 x ρ ¯ d3 x ψ γ 0 ψ ¯ where the probability density was ρ = ψγ 0 ψ = ψ † γ 0 γ 0 ψ = ψ † ψ. Huang 25] jµ = iq[(∂µ φ)φ† − (∂µ φ† )φ] which agrees exactly with the 4current derived previously! The conserved charge ( dQ = 0) is therefore [Kaku 71. b† for the complex scalar ﬁeld. .3. Above we now have Q in terms of the quantized ﬁeld operators a. Here jo matches exactly our expression for ρ derived previously! [Kaku 71] gives an excellent discussion of the physical interpretation of this charge.7. Compare this to our result from Chapter 1 which was (with probability ¯ current density j µ = (ρ. COMPLEX KLEINGORDON FIELD and with L = ∂µ φ† ∂ µ φ − m2 φφ† = gµν ∂ ν φ† ∂ µ φ − m2 φφ† gives ∂L = gµν ∂ ν φ† = ∂µ φ† ∂(∂ µ φ) and ∂L = ∂µ φ ∂(∂ µ φ† ) 69 Thus the Noether current is jµ = (∂µ φ† )iq φ − (∂µ φ)iq φ† Once again we drop the constant factor and insert a minus sign to deﬁne a new current [Kaku 71. Schwabl 286. a† . b. This was an expression in terms of the classical Dirac ﬁeld ψ.
t) ≡ φ+ (x) + φ− (x) = The conjugate momentum is ˙ Π(x.8. FREE KLEINGORDON FIELD 3. t) ≡ Π+ (x) + Π− (x) = −i ˜ dk ω ak e−ik·x − a† eik·x k ˜ dk ak e−ik·x + a† eik·x k .8 Summary 1 (2π)3 Two useful integrals are: d3 xe±i(k−k ). t) = φ(x.x = δ(k − k ) d4 k −ik·(x−y) e (2π)4 δ 4 (x − y) = 3.1 KG classical ﬁeld 1 LKG = (∂µ φ∂ µ φ − m2 φ2 ) 2 The massive KleinGordon Lagrangian is giving the KGE (using p2 → −P2 ) (P2 + m2 )φ = 0 (p2 − m2 )φ = 0 The covariant momentum density is Πµ = ∂ µ φ giving the canonical momentum ˙ Π ≡ Πo = φ(x) and Hamiltonian density 1 H = [Π2 + ( φ)2 + m2 φ2 ] 2 and momentum P =− ˙ d3 x φ(x) φ(x) The Fourier expansion of the KG ﬁeld is φ(x.70 CHAPTER 3.
Π(x . t)] = iδ(x − x ) To develop the QFT we impose the equal time commutation relations and [φ(x. t).2 KleinGordon Quantum ﬁeld [φ(x. φ(x . SUMMARY In the above equations (with k 0 ≡ ω = + k 2 + m2 ) ˜ dk ≡ d3 k = 2ω(2π)3 2ω d4 kδ(k 2 − m2 ) θ(k o ) (2π)3 71 The integrals are cast into discrete form with the replacement d 3 k → Σk .8. ¯ .3. a† (k )] = δ(k − k ) [a(k). t). t)] = 0 which imply and Deﬁning [a(k). Π(x . a† (k )] = 0 Nk ≡ a† ak k the Hamiltonian and momentum can be rewritten as 1 ω H = d3 k Nk + 2 1 P = d3 k Nk + k 2 The vacuum state is deﬁned via ak 0 = 0 and the many body states are  · · · n ki · · · n k j · · · = i nki nk with nki = H · · · nki · · · nkj · · · = 1 nk ! k a† k 0 nk (k) · · · nki · · · nkj · · · where (k) = hω(k). a(k )] = [a† (k).8. t)] = [Π(x. t). 3.
FREE KLEINGORDON FIELD 3. Other useful propagators are i∆± (x − y) ≡ [φ± (x).72 CHAPTER 3. This ﬁnally leads to ∆F (x − x ) = −iθ(t − t ) One can also show that i∆F (x − x ) = = 0T φ(x)φ(x )0 0T φ(x)φ† (x )0 ˜ dk e−ik·(x−x ) − iθ(t − t) ˜ dk eik·(x−x ) where the second line is true for Hermitian ﬁelds [BjRQF42].8.3 Propagator Theory (P2 + m2 )φ = J(x) The nonfree KGE is The Feynman propagator in position space is deﬁned as (P2 + m2 )∆F (x − y) ≡ −δ 4 (x − y) so that the solution to the nonfree KGE is φ(x) = φ0 (x) − d4 y ∆F (x − y)J(y) where φ0 (x) is the solution with J = 0. φ(y)] = −i∆+ (x − y) + i∆− (x − y) ˜ ˜ dke ik·(x−y) . φ (y)] = ± giving i∆(x − y) ≡ [φ(x). The momentum space Feynman propagator is deﬁned through the Fourier transform ∆F (x − y) ≡ d4 k −ik·(x−y) e ∆F (k) (2π)4 The usual method of solution for propagators or Green functions is to solve for ∆F (k) and then do a contour integral to get ∆F (x − y) rather than solving for ∆F (x − y) directly. We get ∆F (k) = 1 k 2 − m2 which is the inverse of the momentum operator in the free KGE.
Muirhead. Halzen and Martin.3.9 References and Notes Good references for this chapter are [Teller. Roman. REFERENCES AND NOTES with ∆(x) = ∆+ (x) + ∆− (x) = − d3 k sin k · x (2π)3 ωk 73 These other propagators are related to the Feynman propagator by ∆F (x) = θ(t)∆+ (x) − θ(−t)∆− (x) or ∆F (x) = ±∆± (x) for t >0 < 3. .9. see text. Leon.8. Goldstein].4 Complex KG ﬁeld Not summarized. Bergstrom and Goobar. 3.
FREE KLEINGORDON FIELD .74 CHAPTER 3.
Chapter 4 Dirac Field We have seen that the KGE gives rise to negative energies and nonpositive deﬁnite probabilities and for these reasons was discarded as a fundamental quantum equation. These problems arise because the KGE is nonlinear in ∂ ∂t . ∂ Dirac was searching for an equation linear in E or ∂t . Then we are looking for something of the form p2 − m2 = pµ pµ − m2 = (β µ pµ + m)(γ ν pν − m) where β µ and γ ν are 8 coeﬃcients to be determined. which is easy if we only have p0 (i. p = 0).e. namely (with p0 = p0 ) (p0 − m)(p0 + m) = 0 and obtain two ﬁrst order equations p0 − m = 0 or p0 + m = 0 However it’s more diﬃcult if p is included. The ‘derivation’ of the DE presented here follows [Griﬃths. The RHS is β µ γ ν pµ pν + m(γ ν − β ν )pν − m2 For equality with LHS we don’t want terms linear in pν . Dirac thus sought a relativistic quantum equation linear ∂ in ∂t . like the SE. leaving p2 = pµ pµ = γ µ γ ν pµ pν 75 . unlike the SE. Instead of starting from p2 − m2 = 0 his strategy was to factor this relation. We shall see that therefore Dirac was forced to invent a matrix equation. thus γ ν = β ν . 215]. pg.
γ ν } = 2g µν which is called a Cliﬀord Algebra where the curly brackets denote the anticommutator {A. At this point Dirac had a brilliant inspiration. It turns out that this can be done.76 i. In noncovariant notation it is ∂ψ Hψ = i¯ h ∂t with H ≡ α · p + βm . what if the γ’s are matrices instead of numbers? Since matrices don’t commute we might be able to ﬁnd a set such that (γ 0 )2 = 1 (γ 1 )2 = (γ 2 )2 = (γ 3 )2 = −1 γ µ γ ν + γ ν γ µ = 0 for µ = ν Or. B} ≡ AB + BA. DIRAC FIELD (p0 )2 − (p1 )2 − (p2 )2 − (p3 )2 = (γ 0 )2 (p0 )2 + (γ 1 )2 (p1 )2 + (γ 2 )2 (p2 )2 +(γ 3 )2 (p3 )2 + (γ 0 γ 1 + γ 1 γ 0 )p0 p1 +(γ 0 γ 2 + γ 2 γ 0 )p0 p2 + (γ 0 γ 3 + γ 3 γ 0 )p0 p3 +(γ 1 γ 2 + γ 2 γ 1 )p1 p2 + (γ 1 γ 3 + γ 3 γ 1 )p1 p3 +(γ 2 γ 3 + γ 3 γ 2 )p2 p3 Here’s the problem. more succinctly. γ 0 = γ 0 .e. These are γ0 ≡ β = I 0 0 −I γ ≡ βα ≡ 0 −σ 0 I σ 0 γ µ ≡ (β. but the smallest set of matrices are 4 × 4. βα) Let’s introduce everything else for completeness γ 5 ≡ iγ 0 γ 1 γ 2 γ 3 = σ1 = α≡ 0 σ 0 1 σ 0 1 0 β≡ σ2 = 0 −i i 0 I 0 σ3 = † 1 0 0 −1 1 0 0 −1 and γµ = γ0γµγ0 † † i. {γ µ . In coordinate space (using pµ → i∂ µ ) it is (i∂ − m)ψ = 0 . (with (p1 )2 = (p1 )2 ) CHAPTER 4. Kaku. γ i = −γ i γ 5 = γ5 β −1 = β These conventions are used by the following authors: [Griﬃths. Thus the Dirac equation is (p − m)ψ = 0 with A ≡ γ µ Aµ .e. Halzen & Martin]. we could pick γ 0 = 1 and γ 1 = γ 2 = γ 3 = i but we can’t get rid of the ‘cross terms’.
PROBABILITY & CURRENT 77 The DE can be written in terms of 4vectors (/ − m)ψ = 0 and is therefore p manifestly covariant. the Dirac adjoint equation. ψ ∂ ≡ (∂µ ψ)γ µ . i.4. It’s very important not to get confused with this.e. ¯ ∂ ψ(i/ + m) = 0 ¯ / ⇔ ψ(i ∂ + m) = 0 ← ¯ (ψ is a row matrix!) ¯ Some sloppy authors write this as (i/ + m)ψ = 0 but this cannot be because ∂ ← ¯ ¯ ψ is a row matrix! The notation ∂ however means that ∂ operates on ψ to / / ¯/ ¯ the left. For matrices the generalization of complex conjugate (∗) is Hermitian conjugate (†) which is the transpose of the complex conjugate.1 Probability & Current For the SE and KGE we used SE∗ and KGE∗ to derive the continuity equation. The DE is (i/ − m)ψ = 0 = (iγ µ ∂µ − m)ψ ∂ and DE† is (using (AB)† = B † A† ) ∂ ψ † (i/ − m)† = 0 † (−iγ µ† ∂ − m) = 0 where ∂ † = ∂ =ψ µ µ µ = ψ † (−iγ 0 γ µ γ 0 ∂µ − m) = 0 using γ µ† = γ 0 γ µ γ 0 We want to introduce the Dirac adjoint (ψ is a column matrix) ¯ ψ ≡ ψ†γ 0 Using γ 0 γ 0 = 1 we get ψ † (−iγ 0 γ µ γ 0 ∂µ − mγ 0 γ 0 ) = 0 ¯ ⇒ ψ(iγ µ γ 0 ∂µ + mγ 0 ) = 0 and cancelling out γ 0 gives. 4. The DE and DE† are explicitly (i/ − m)ψ = 0 ∂ ¯ ∂ + m) = 0 ψ(i/ ⇔ ⇔ iγ µ ∂µ ψ − mψ = 0 ¯ ¯ i(∂µ ψ)γ µ + mψ = 0 ← .1.
pg.78 CHAPTER 4. deﬁne the Dirac adjoint ∗ ¯ ψ ≡ ψ † γ 0 = (ψ1 ∗ ψ2 ∗ − ψ3 ∗ − ψ4 ) . 103] ¯ ψ(iγ µ ∂µ ψ − mψ) = 0 ¯ ¯ (i(∂µ ψ)γ µ + mψ)ψ = 0 Adding these gives ¯ ¯ ¯ ψγ µ ∂µ ψ + (∂µ ψ)γ µ ψ = 0 = ∂µ (ψγ µ ψ) giving ¯ j µ = (ψγ µ ψ) ≡ (ρ. Rather. 34] 4. Now one could get very confused writing ¯ ∂ ψ(i/ − m)ψ = 0 ¯ ∂ ψ(i/ + m)ψ = 0 whereas what is really meant is [Halzen & Martin. Multiply DE from the left by ψ and DE† from the right by ψ. DIRAC FIELD ¯ Now let’s derive the continuity equation. such as ψ1 ψ 2 ∗ ∗ ∗ ∗ ψ † ψ = (ψ1 ψ2 ψ3 ψ4 ) = ψ1 2 + ψ2 2 + ψ3 2 + ψ4 2 but this is not ψ3 ψ4 a Lorentz scalar (it’s got all + signs).2 Bilinear Covariants ψ can be written ψ1 ψ2 ψ3 ψ4 and one can try to construct a scalar. j) or ¯ ρ = ψγ 0 ψ = ψ † ψ = 4 ψi 2 i=1 which is now positive deﬁnite! The 3current is ¯ j = ψγψ [see also Mosel 17.
3 4. ¯ = Ae i Substituting gives h ¯ k2 ψ(x. t) or 2m yielding E=+ h ¯ 2 k2 2m E− h ¯ 2 k2 2m ψ=0 .3. t) = (Aeikx + Be−ikx )e− h Et ≡ ψ−E Substituting gives h ¯ 2 k2 ψ(x. t) or 2m yielding E=− h ¯ 2 k2 2m i E+ h ¯ 2 k2 2m ψ=0 ¯ However it also has solution ψ(x. 4. NEGATIVE ENERGY AND ANTIPARTICLES 79 ¯ and ψψ = ψ1 2 + ψ2 2 − ψ3 2 − ψ4 2 is a Lorentz scalar (see Bjorken and Drell]. t) = +Eψ(x. t) = −Eψ(x.4.3. One can prove that the following quantities transform as indicated: ¯ ψψ ¯ ψγ 5 ψ ¯ ψγ µ ψ ¯ ψγ µ γ 5 ψ ¯ ψσ µν ψ scalar pseudoscalar vector pseudovector antisymmetric 2nd rank tensor (1 (1 (4 (4 (6 component) component) component) component) component) i where σ µν ≡ 2 (γ µ γ ν − γ ν γ µ ). t) = (C cos kx + D sin kx) e h Et ≡ ψE i ikx + Be−ikx e h Et in 1dimension.1 Negative Energy and Antiparticles Schrodinger Equation − ¯2 h 2m 2 The free particle SE is ψ = i¯ h ∂ψ ∂t which has solution ¯ ψ(x.
4. toss out the whole equation! Also reason why Einstein did not reject SR.2 KleinGordon Equation (P2 + m2 )φ = 0 The free particle KGE is with P2 = − A solution is φ = N eip·x = N ei(Et−p·x) Substituting gives (E 2 − p 2 + m2 )φ = 0 which implies ⇒ E 2 = p 2 + m2 or E = ± p 2 + m2 Thus the single solution φ = N eip·x has both positive and negative energy solutions. We are free to toss away one solution as unphysical and only keep ψE . Whereas in the SE we got two diﬀerent solutions for positive and negative energy. E = ± p2 + m2 were separate solutions. However for KGE and DE the same solution ψ gives both positive and negative energy. However in KGE and DE we always get both positive and negative energy for all solutions. If we want to get rid of negative energy in the SE we toss away one solution. DIRAC FIELD ψE and ψ−E are diﬀerent solutions. i. The only way to toss away negative energy is to toss away all solutions. namely (E 2 − p 2 + m2 )φ = 0 or E = ± p 2 + m2 ∂2 + ∂t2 2 . not part of same solution. Another solution is φ = N e−ip·x = N e−i(Et−p·x) Substituting also gives the same as above.80 CHAPTER 4. The ﬁrst one ψE corresponds to positive energy and the second one ψ−E to negative energy.e. Of course we are not free to toss away one energy because it’s in the solution.3.
For p = 0. pg. This also happens with the Dirac equation. with bound states appearing in between. NEGATIVE ENERGY AND ANTIPARTICLES 81 and again the single solution φ = N e−ip·x has both positive and negative energy solutions. the DE describes.4. pg. there will be a continuum of states above and below E = ±m [Landau. Recall ρ = 2m φ∗ ∂φ − φ ∂φ . For φ = e±ip·x we get ρ = m φ∗ φE ∂t ∂t which gives negative ρ for positive or negative E. Also we can now clearly see the problem with ρ as calculated with the ∗ i¯ h h ¯ KGE. automatically. 74] E>0 p=0 p=0 . (Halzen & Martin. 225].mc2 E<0 However for the KGE the negative energies are a catastrophe.3. The interpretation of these states is as follows. Dirac’s way out of the negative energy catastrophe was to postulate that the negative energy sea was ﬁlled with fermions and so the ??? E mc 2 0 . In QM there would be transitions to the negative energy continuum to inﬁnite negative energy. However. For p = 0 (particle at rest) then E = ±m. particles with spin.
∂ p Actually. −m with eigenvectors [Halzen & Martin. pg. Dirac Sea not taken seriously until positron discovered!! (1932 Anderson) . let’s ﬁrst look at rest frame (p = 0) solutions (RF). 104] 1 0 0 0 0100 0010 0 0 0 1 Thus the DE has positive (E = +m) and negative (E = −m) energy solutions! (We shall look at p = 0 solutions in a moment) Let’s summarize so far (Aitchison & Hey. − Thus both the KGE and DE have negative energies. The DE describes fermions (see next section). − +. 71) SE KGE DE Probability (ρ) + − + Energies for same solution ψ + +. DIRAC FIELD 4.3 Dirac Equation Let’s look for plane wave solutions of the form ψ(x) = w(p)e−ip·x .3. which in the RF is Hw = βmw = mI 0 0 −mI w = Ew. pg. m. Dirac realized if given particles in sea energy of 2mc2 create holes. Dirac’s idea was that the negative energy sea was ﬁlled with fermions and via PEP prevented the negative energy cascade of positive energy particles.82 CHAPTER 4. The eigenvalues are E = m. −m. Writing the DE as ∂ψ Hψ = i¯ h with H = α · p + βm ∂t independent DE is Hψ = Eψ. Substituting into (i/ − m)ψ = 0 gives the momentum space DE (/ − m)w = 0 .
which means we are free to choose them [Aitchison & Hey 69] as wA .4 Free Particle Solutions of Dirac Equation 0 −i i 0 ≡ Before proceeding recall the following results: σx = 0 1 1 0 Pz p+ σy = p− −pz σz = 1 0 0 −1 σ·p= pz px + ipy px − ipy −pz Let’s now look at p = 0 solutions (i. we look for solutions of the form ψ(x) = w(p)e−ip·x Substituting into (i/ − m)ψ = 0 ∂ E −σ · p gives the momentum space DE (/ − m)w = 0. with w ≡ wB (/ − m)w = p giving wA = Combining yields wA = (σ · p )2 p2 wA = 2 wA E 2 − m2 E − m2 E−m −σ · p σ·p −E − m wA wB =0= (E − m) wA − o · p wB σ · p wA − (E + m) wB σ·p σ·p wB and wB = wA E−m E+m because (σ · p)2 = p 2 .4. not in rest frame). FREE PARTICLE SOLUTIONS OF DIRAC EQUATION 83 4.e. Using p = p / σ·p −E wA gives.4. As before. wB = 1 0 or 0 1 . Thus E 2 = p 2 + m2 or E = ± p 2 + m2 Thus again we see the negative energy solutions ! (this time for the p = 0 DE) Our ﬁnal solutions are w= wA wB = wA σ·p E+m wA or σ·p E−m wB wB with wA and wB left unspeciﬁed.
For (3) and (4) we must use E = − √ √ The term E = + is called the particle solution. do we use E = + p 2 + m2 or E = − p 2 + m2 ? Well. with χ(1) ≡ and χ(2) ≡ 0 1 w(1) (p) = N 1 0 pz E+m p+ E+m =N =N 1 σ·p E+m χ(1) w(2) (p) = N 0 1 p− E+m −pz E+m 1 o·p E+m χ(2) with and also E ≡ + p 2 + m2 pz E−m p+ E−m =N =N σ·p E−m σ·p E−m w(3) (p) = N 1 0 p− E−m −pz E−m 1 χ(1) w(4) (p) = N 0 1 1 χ(2) with E ≡ − p 2 + m2 . √ 1 for (1) and (2) we must use E = + otherwise E+m blows up for p = 0.84 We have the following possibilities Pick wA = Pick wA = Pick wB = Pick wB = 1 0 0 1 1 0 0 1 o·p E+m o·p ⇒ wB = E+m o·p ⇒ wA = E−m o·p ⇒ wA = E−m ⇒ wB = 1 0 0 1 1 0 0 1 CHAPTER 4. 1 0 Rewrite as. √ 1 otherwise E−m blows up for p = 0. The term E = − is called the antiparticle solution. DIRAC FIELD 1 E+m 1 = E+m 1 = E−m 1 = E−m = pz p+ p− −pz pz p+ p− −pz (1) p± ≡ px ± ipy (2) (3) (4) But the question is.
s) u(1) = N 1 0 pz E+m p+ E+m (2) u =N 0 1 p− E+m −pz E+m v (2) = N pz E+m p+ E+m 1 0 v (1) = N p− E+m −pz E+m 0 1 all with E ≡ + p 2 + m2 = v (1) (Kaku) = v (2) (Kaku) Kaku diﬀerent from everyone else 1 σ·p E+m u=N χ v=N σ·p E+m 1 χ .2) (p) = w(3. Gross writes u(p. Deﬁne u(p.3) (−p) Note: Bj writes u(p. s) = u(s) (p) ≡ w(s) (p) u(1.2) (p) = w(1. FREE PARTICLE SOLUTIONS OF DIRAC EQUATION or w(s) (p) = N w(s+2) (p) = N 1 o·p E+m σ·p E−m 85 χ(s) with E ≡ + p 2 + m2 χ(s) 1 with E ≡ − p 2 + m2 s = 1. s) and w(p). 2 [Halzen & Martin.4) (−p) (Me & Kaku) everyone else: v (1.4. pg.2) (p) v(p. s) = v (s) (p) ≡ w(s+2) (−p) v (1.4. .2) (p) = w(4. 105] But all free particles carry positive energy! Thus reinterpret w(3) and w(4) as positive energy antiparticle states w (s+2) (p) = N √ − √ σ·p p 2 +m2 −m 1 χ(s) = N χ(s) √ σ·p p 2 +m2 +m w(s+2) (−p) = N − −σ·p p 2 +m2 −m 1 1 χ(s) ..
Kaku. is u(r) u(s) ¯ E m. pg. pg. v (2) → χ(1) = u(2) .e.7) we have u† γ 0 γ µ u = u† u = E uu ¯ m and with µ = 0 and = 1 we get Thus uu = 1 ⇒ u† u = ¯ case.86 u(1) . pg. DIRAC FIELD 1 0 0 1 pz p+ p− −pz σ·p= Halzen & Martin. Griﬃths. Halzen & Martin √ E+m These oddlooking normalizations come by specifying uu. Actualy the more general normalization. Sterman u† u = E ⇒N = m E+m but spurious diﬃculties when m → 0 2m u† u = 2E ⇒ N = Bogoliubov & Shirkov u† u = 1 pµ ¯ m uu Griﬃths. Alternatively we can specify v v = −1 . 71ﬀ. in this = δrs (see Problem 4. 220 footnotes] The most common conventions are. i. v (r) v (s) = −δrs [Kaku. Using the result ¯ uγ µ u = ¯ γ0γ0 pu ¯ m uu (Problem 4. v (1) → χ(2) = Kaku N = E+m 2m CHAPTER 4. pg. Greiner.8). ¯ ¯ 754] . with E ≡ + p2 + m2 everywhere below Bjorken & Drell. 107 ψ(x) = w(−p)e−i(−p·x) ψ(x) = w(p)e−ip·x −ip·x ⇒ ψ(x) = u(p)e ψ(x) = v(p)e+ip·x (/ − m)w(p) = 0 p (−/ − m)w(−p) = 0 p p ⇒ (/ + m)v = 0 p ⇒ (/ − m)u = 0 The adjoints satisfy u(/ − m) = 0 ¯p and v (/ + m) = 0 ¯p Normalization [Muirhead.
Huang 123.5. Peskin 52.3 Noether spacetime current Noether internal symmetry and charge Fourier expansion and momentum space References: [Mosel 35.5. Greiner FQ 123.5.now handwriting 4.5 Classical Dirac Field The Dirac Lagrangian is [Mosel 34] ¯ ∂ LD = ψ(i/ − m)ψ from which one can obtain the Dirac equation and its adjoint (see Problems). The best discussion as to why the Dirac creation and annihilation operators must obey anticommutation relations is given in [Peskin 5256] and [Greiner FQ 129].now handwriting . 703] NNN . NNN . Because of this the measure for fermions will be [IZ 114. CLASSICAL DIRAC FIELD 87 4. Because of the normalization of our Dirac spinors (see previous chapter) we will have a diﬀerent normalization constant in our Fourier expansion of the Dirac ﬁeld. Most books just write down the answer (as we shall) but [Greiner FQ 123] derives the result very clearly using the same method that we used for the KleinGordon ﬁeld where one proves that the expansion must contain two terms. 147.1 4. Kaku 86] We now wish to expand the Dirac ﬁeld in terms of creation and annihilation operators.2 4. as compared to the KG case. Schwabl 290.4.5.
10. b† .100 S matrix without SE.11 References and Notes Mandl & Shaw. DIRAC FIELD 4.2 4. d.8. Leon. Sakurai QM.3 4.4 Pauli Exclusion Principle Hamiltonian.10.10 4.7 4.88 CHAPTER 4. Bj RQF 177 .2 4.3 4.10. Momentum and Charge in terms of creation and annihilation operators Hamiltonian Momentum Angular Momentum Charge 4. Teller.4 Propagator theory Summary Dirac equation summary Classical Dirac ﬁeld Dirac QFT Propagator theory 4.8 4.1 4.8.1 4. d† Anticommutators 4.8.8.6 4.6. Merzbacher S matrix and G function Bj RQM 83.9 4.10.97.1 Dirac QFT Derivation of b.
1. Mandl and Shaw 129.1.4 5.2 5. 274. 350]..1. The amplitudes are the same for a circular wave.1 5.) To produce a linearyly polarized wave on a rope just jiggle the rope up and down for a linearly polarized wave in the vertical direction. 2nd ed.. 1st ed. Guidry 86. which is holding the rope.975. in a circular fashion. An elliptically polarized wave is formed from two linearly polarized waves out of phase but with each linearly polarized wave having a diﬀernt amplitude.3 5.87. 177]. The circularly polarized wave can be produced by moving one’s hand. A nice elementary discussion can be found in [Weidner and Sells.p. p. In general transverse waves can be linearly or circularly polarized. 89 . (If they are in phase then they combine to form a linearly polarized wave at a diﬀerent angle. 350. GriﬃthsEM. 1st ed. Schwabl 313. A circularly polarized wave can be made out of two linearly polarized waves if the two linear waves are out of phase.1. 1st ed.1 5. 1st ed.1. One can rotate clockwise or counterclockwise to produce circularly polarized states of opposie helicity.p. Note that Guidry and Schwabl are excellent.. p.5 Review of Classical Electrodynamics Maxwell equations in tensor notation Gauge theory Coulomb Gauge Lagrangian for EM ﬁeld Polarization vectors References: [Weidner and Sells. p. GreinerFQ 161. GriﬃthsEM.Chapter 5 Electromagnetic Field 5. Jackson..975.
the polarization vectors used are the same in the classical and quantum case. 274] Finally. 2nd ed. p. not four. Thus any wave in general can be represented in terms of linearly polarized basis states or circularly polarized basis states. Even though it should be obvious. (As we shall see below there are only two. We will know what states a particlar author is using by simply recognizing wether the states are real (linear polarization) or complex (circular polarization).) The circularly polarized states are described by complex basis vectors. In general the linear polarization basis states are described by real polarization vectors 1 and v 2 .90 CHAPTER 5. p. [Jackson. in describing our quantum ﬁeld states for the photon it does not matter whether we use the linearly polarized basis states or the circularly polarized ones. Diﬀerent authors choose diﬀerent basis states. similarly a linearly polarized wave can be built from a linear combination of two circularly polarized wave. note that the treatment of polarization vectors from the point of view of classical electrodynamics [Jackson] is the same as the quantum ﬁeld treatments. possible states for photons. say written as + and − . 274]. This is discussed from a mathematical point of view in [Jackson. ELECTROMAGNETIC FIELD Just as the elliptically or circularly polarized wave can be created from two linear polarized waves (linear combination of them). . which is an easy way of allowing for a phase diﬀerence between the two linearly polarized states out of which the circular states are constructed. These states are written in terms of each other ± 1 =√ ( 2 1 ± i 2) The circularly polarized states are complex. That is. 2nd ed.
2.6 5.5.2 5.1.5 Photon propagator GuptaBleuler quantization Proca ﬁeld .4 5.2.3 5.1 Quantized Maxwell ﬁeld Creation & annihilation operators 5.1.7 5. QUANTIZED MAXWELL FIELD 91 5.1.8 Linear polarization vectors in Coulomb gauge Circular polarization vectors Fourier expansion 5.
ELECTROMAGNETIC FIELD .92 CHAPTER 5.
101. t0 ) ∂t . Thus we can ‘cancel’ α(t0 ) to arrive at the operator equation h HU(t. Each term in this Smatrix expansion will represent a particular Feynman diagram. Sakurai] α(t) ≡ U(t. namely the expansion of the time evolution operator.Chapter 6 Smatrix. t0 )α(t0 ) ∂t ∂ but ∂t does not operate on α(t0 ) because t0 is ﬁxed. t0 )α(t0 ) and upon substitution into the SE gives HU(t. The Smatrix expansion looks very similar to something which occurs in nonrelativistic quantum mechanics. cross section & Wick’s theorem Our Feynman diagram series is going to come from an expansion of the socalled Scattering Matrix (or Smatrix). Leon. pg. 63. t0 )α(t0 ) = i¯ h ∂ U(t. t0 ) = i¯ 93 ∂ U(t.1 Schrodinger Time Evolution Operator ∂ α(t) ∂t Recall the timedependent Schrodinger equation (SE) Hα(t) = i¯ h Deﬁne the time evolution operator [Mandl & Shaw. Thus we will study that ﬁrst. pg. 6.
94 CHAPTER 6. t0 ) dt2 H(t1 )H(t2 ) × dt1 t0 t0 t2 × 1− =1+ −i h ¯ t i h ¯ t t0 dt3 H(t3 )U(t3 . CROSS SECTION & WICK’S THEOREM A schematic solution is ¯ U(t. t0 ) by integrating both sides to give U(t. SMATRIX. t0 )dt1 t0 dt1 H(t1 )U(t1 . t0 ) = 1 we have U(t. t0 ) t0 t1 dt1 H(t1 ) + t0 −i h ¯ t 2 dt1 t0 t2 t1 dt2 H(t1 )H(t2 ) dt3 H(t1 )H(t2 )H(t3 ) + · · · + or U(t. We solve for U(t. t0 ) − U(t0 . t0 ) t0 dt1 H(t1 ) 1 − t i h ¯ t1 dt2 H(t2 )U(t2 . t0 ) = and with the boundary condition U(t0 . t0 ) = 1 + ∞ n=1 −i h ¯ 3 dt1 t0 t0 dt2 t0 −i h ¯ n t t1 dt1 t0 t0 dt2 · · · tn−1 t0 dtn H(t1 ) · · · H(tn ) . t0 ) = 1 − i h ¯ t 1 i¯ h t H(t1 )U(t1 . t0 ) = e− h H(t−t0 ) i However let’s solve the equation more rigorously. t0 ) = 1 − =1− =1+ =1+ i h ¯ t t0 i h ¯ t dt1 H(t1 )U(t1 . t0 ) t0 −i h ¯ −i h ¯ dt1 H(t1 ) + t0 t dt1 H(t1 ) + t0 −i h ¯ −i h ¯ 2 t t1 dt1 t0 2 t t0 t1 dt2 H(t1 )H(t2 )U(t2 . t0 ) t0 This is solved by iteration as follows U(t.
6. 64. Two other ways of writing this are U(t. SCHRODINGER TIME EVOLUTION OPERATOR 95 which is the ﬁrst version of our expansion of the time evolution operator. pg. pg. 64] and recall ∞ 1 n x2 x3 e =1+x+ + + ··· = 1 + x 2! 3! n! n=1 x . t0 ) = 1 + 1 n! n=1 t t0 ∞ −i h ¯ H(t )dt n t t dt1 t0 t0 dt2 · · · t t0 dtn T [H(t1 ) · · · H(tn )] ≡T e −i/¯ h [Leon. 99. Hatﬁeld. pg. Merzbacher. 5253 gives a good example as to why we introduce time ordered product.1. pg. 475] where the last expression is just a formal way of writing the formula above [Leon. pg.1.6. See also Guidry.1 Time Ordered Product In the above formulae we introduced the time ordered product which we now deﬁne for two terms T [A(t1 )B(t2 )] (2! combinations) ≡ A(t1 )B(t2 ) if t1 > t2 B(t2 )A(t1 ) if t2 > t1 ≡ θ(t1 − t2 )A(t1 )B(t2 ) + θ(t2 − t1 )B(t2 )A(t1 ) where θ(x − y) ≡ 1 if x > y 0 if x < y The time ordered product of three terms is deﬁned as A(t1 )B(t2 )C(t3 ) A(t1 )C(t3 )B(t2 ) B(t )A(t )C(t ) 2 1 3 T [A(t1 )B(t2 )C(t3 )] ≡ B(t2 )C(t3 )A(t1 ) (3! combinations) C(t3 )A(t1 )B(t2 ) if if if if if C(t3 )B(t2 )A(t1 ) if t1 t1 t2 t2 t3 t3 > t2 > t3 > t1 > t3 > t1 > t2 > t3 > t2 > t3 > t1 > t2 > t1 ≡ θ(t1 − t2 )θ(t2 − t3 )A(t1 )B(t2 )C(t2 ) + θ(t1 − t3 )θ(t3 − t2 )A(t1 )C(t3 )B(t2 ) +θ(t2 − t1 )θ(t1 − t3 )B(t2 )A(t1 )C(t3 ) + θ(t2 − t3 )θ(t3 − t1 )B(t2 )C(t3 )A(t1 ) +θ(t3 − t1 )θ(t1 − t2 )C(t3 )A(t1 )B(t2 ) + θ(t3 − t2 )θ(t2 − t1 )C(t3 )B(t2 )A(t1 ) .
pg. the Hamiltonian is split into a fre particle piece and an interaction piece H ≡ H 0 + HI then HI evolves and H0 is frozen. This explains the term appearing in the expansion for U. U is part of H and so what does it mean to say that operators are frozen in the SP? What we mean is that the state vectors α obey an equation of motion Hα = i¯ ∂t α and the operators do not. OI .1] 6. CROSS SECTION & WICK’S THEOREM Notice that there are n! combinations for n operators. α H . We shall label Schrodinger. OH . Although not exactly analogous.e. Picture Schrodinger Heisenberg Dirac (Interaction) Operators Frozen Evolve Evolve State Vectors Evolve Frozen Evolve In time dependent perturbation theory one considers potentials like U = E0 cos ωt (e. Lagrange or Hamilton formulation of mechanics. there are three popular ways to work in quantum mechanics known as the Schrodinger. The opposite is true in the Heisenberg picture.g. which we shall use extensively. Heisenberg or Dirac pictures. oscillating electromagnetic ﬁeld) in the Schrodinger picture (SP). α I and operators as OS .96 CHAPTER 6.2 Schrodinger. Heisenberg and Dirac (Interaction) Pictures Reference: [Mandl and Shaw. In each formulation the equations are diﬀerent. SMATRIX. Heisenberg. Interaction picture states as α S . Recall the Schrodinger picture of NRQM Hα(t) S = i¯ h ∂ α(t) ∂t S . In the usual formulation of quantum mechanics via the Schrodinger equation. The Dirac picture is often also called the Interaction picture. Viceversa for h∂ the Heisenberg picture. i. 22] One can do classical mechanics with either the Newtonian. 1 n! [do Problem 8. In the Interaction picture. in the Schrodinger picture. the operators are frozen in time and the states evolve in time.
i. Deﬁne Heisenberg picture states as α H ≡ U † α(t) S = α(t0 ) S which is clearly frozen in time.1 Heisenberg Equation In the Schrodinger representation the state vectors evolve in time and the Schrodinger equation describes their time evolution. S β(t)OS α(t) S = H βOH (t)α H [do Problem 8. H] = i¯ h [do Problems 8. I (Unfortunately we have the notation HI for the interaction Hamiltonian in the Interaction picture.2.2] 6.e. SCHRODINGER.2. t0 )α(t0 ) S where U(t.6. Deﬁne Heisenberg operators OH (t) ≡ U † OS U which clearly evolves in time because OS is frozen but U † and U carry time dependence.3 and 8.2 Interaction Picture We use the Dirac or Interaction representation if the Hamiltonian can be split into two parts. It is obtained by diﬀerentiating OH (t) ≡ U † OS U to give [OH (t). t0 ) is our time evolution operator. We must have this for the physics to be the same.2. In the Heisenberg picture the operators evolve in time. so what is the equation governing the operator time evolution? It is called the Heisenberg equation of motion (for operators). Expectation values are the same. a free particle piece H0 and an interaction piece HI . Now the important thing about the Heisenberg and Schrodinger pictures is that expectation values remain the same in both pictures. HEISENBERG AND DIRAC (INTERACTION) PICTURES97 and the Schrodinger state vectors evolve in time according to α(t) S ≡ U(t. (We have used UU † = 1).4] d H O (t) dt 6.) Thus H = H 0 + HI .
not Interaction picture. CROSS SECTION & WICK’S THEOREM where.1) which has the rigorous solution UI (t. 8. −∞) .6. SMATRIX. HI means interaction Hamiltonian. t0 ) = 1 + 1 n! n=1 ∞ −i h ¯ n t t dt1 t0 t0 dt2 · · · tn−1 t0 dtn T [HI (t1 ) · · · HI (tn )] Now deﬁne the Smatrix S ≡ U(∞. t0 ) ≡ e− h H0 (t−t0 ) i and α(t) and Now I † ≡ U0 α(t) S † OI (t) ≡ U0 OS U0 I S H0 = H 0 ≡ H0 (exercise: show this) Manipulating the previous two equations we get [OI (t). 8.7] Now deﬁne a time evolution operator for HI alone as α(t) which leads to I HI UI = i¯ h I ≡ UI α(t0 d UI dt I (6.98 CHAPTER 6.5. at this stage. H0 ] = i¯ h and I HI (t)α(t) I d I O (t) dt d α(t) dt = i¯ h i I where † S I S ¯ ¯ HI (t) = U0 HI U0 = e h H0 (t−t0 ) HI e− h H0 (t−t0 ) i [do Problems 8. Deﬁne a time evolution operator for H0 alone as ¯ U0 ≡ U0 (t.
CROSS SECTION AND SMATRIX in other words α(∞) ≡ Sα(−∞ = Si where i is the initial state..3 Cross section and Smatrix Consider the reaction 1 + 2 → 1 + 2 + 3 + . A set of Feynman rules actually gives −iM [Griﬃths]. is not the Schrodinger equation but rather the operator equation of motion in the Interaction picture [Bj RQF 177]. although reminiscent of. Thus S =1+ 1 n! n=1 ∞ 99 −i h ¯ n ∞ −∞ ∞ dt1 −∞ dt2 · · · ∞ −∞ dtn T [HI (t1 ) · · · HI (tn ) or in covariant form S =1+ 1 n! n=1 ∞ −i h ¯ n d4 x1 d4 x2 · · · d4 xn T [HI (x1 )HI (x2 ) · · · HI (xn )] which is the Dyson expansion of the Smatrix. Note that this expansion of the Smatrix does not rely on the Schrodinger equation but comes from the general interaction picture and the operator equation of motion (6. which. but it always has common factors.6. This is an inﬁnite series. Greiner QED 221] n < f Si >= i(2π)4 δ 4 (p1 + p2 − i=1 pi ) M Π2 i=1 Ni Πn 2Ei (2π)3 i=1 Ni 2Ei (2π)3 where the normalization factors are Ni = 1 for scalar bosons and photons and Ni = 2m for fermions. The relation between the Smatrix element and the invariant amplitude is given by [Greiner FQ 267.n We shall be calculating an Smatrix element. [Greiner FQ 267] . Pulling these out we are left with a quantity M called the invariant amplitude. 6.. each term of which gets represented as a Feynman diagram.1).3. Cross sections are writtem directly in terms of M.
219. 215. pg. pg. CROSS SECTION & WICK’S THEOREM A set of Feynman rules actually gives −iM [Griﬃths. . p3 · · · pn ) = δ 4 (p1 + p2 − p3 − p4 − · · · pn )× d 3 p3 d 3 p4 d 3 pn × ··· (2π)3 2E3 (2π)3 2E4 (2π)3 2En giving [Griﬃths. p1 . The diﬀerential cross section is given directly in terms of the invariant amplitude as [RPP. . 198] dσ = (2π)4 M2 4 (p1 . pn ) ≡ δ 4 P − i=1 pi i=1 d 3 pi (2π)3 2Ei Thus for the reaction 1 + 2 → 3 + 4···n we have [Griﬃths. 176. Griﬃths. pg. . 1987]. .100CHAPTER 6. PPB. pn+2 ) with [RPP. 200] δ 4 (p1 + p2 − p3 − p4 ) = δ(E1 + E2 − E3 − E4 )δ 3 (p1 + p2 − p3 − p4 ) giving dσ 1 (2π)4 M2 d3 p3 = √ δ(E1 +E2 −E3 −E4 )δ 3 (p1 +p2 −p3 −p4 ) d 3 p4 4 (2π)3 2E3 (2π)3 2E4 . 175. 198] n n dΦn (P . Griﬃths. 198] dΦn (p1 + p2 . pg. pg. 198] dσ = d 3 p3 d 3 p4 3 3 4 (p1 · p2 )2 − m2 m2 (2π) 2E3 (2π) 2E4 1 2 d 3 pn ··· δ 4 (p1 + p2 − p3 − p4 − · · · pn ) (2π)3 2En 1+2→3+4 dσ = d 3 p3 d 3 p4 δ 4 (p1 + p2 − p3 − p4 ) 2 − m2 m2 (2π)3 2E3 (2π)3 2E4 4 (p1 · p2 ) 1 2 (2π)4 M2 (2π)4 M2 For the reaction with [Griﬃths. PPB. SMATRIX. pg. p3 . pg.p2 )2 − m2 m2 1 2 dΦn (p1 + p2 . pg. pg.
6. RQM 83. Bj RQF 177 For a derivation of the Smatrix based on Green function techniques. RQF 177. Teller.and outstates. and in. Leon. Sakurai QM. Merzbacher S matrix and G function Bj RQM 83.4 6. see Bjorken and Dell.4.4.97. 97.4.100 S matrix without SE. WICK’S THEOREM so that 101 dσ 3 d 3 p3 2πM2 d p4 = dσ = δ(E1 + E2 − E3 − E4 ) 3 d 3 p4 8E4 (p1 · p2 )2 − m2 m2 (2π) 2E3 1 2 giving the Lorentz invariant diﬀerential cross section for production of particle 3 as dσ M2 = δ(E1 + E2 − E3 − E4 ) d3 p3 /E3 64π 2 E4 (p1 · p2 )2 − m2 m2 1 2 NNN PUT IN GRIFFITHS STATISTICAL FACTOR S (EASY!) ALSO WRITE DOWN Λ 6.2 Wick’s theorem Contraction Statement of Wick’s theorem References and Notes Mandl & Shaw. 100.1 6. .
102CHAPTER 6. SMATRIX. CROSS SECTION & WICK’S THEOREM .
2.1 QED Lagrangian QED Smatrix First order Smatrix 103 .2 7.1 7.Chapter 7 QED 7.
Note that the photon diagrams do not have an arrow associated with them since each photon is its own antiparticle.see before). [Greiner FQ 236] . the second order term in the Smatrix is (with h ¯ = 1) [GreinerFQ 238] S (2) = = = + + + + + + + 1 (−i)2 2! 1 (−iq)2 2! (−iq)2 2! (−iq)2 2! (−iq)2 2! (−iq)2 2! (−iq)2 2! (−iq)2 2! (−iq)2 2! (−iq)2 2! d4 x1 d4 x2 T [H(x1 )H(x2 )] ¯ ¯ d4 x1 d4 x2 T [ : ψ(x1 )γµ ψ(x1 )Aµ (x1 ) : : ψ(x2 )γν ψ(x2 )Aν (x2 ) : ] ¯ ¯ d4 x1 d4 x2 : ψ(x1 )γµ ψ(x1 )ψ(x2 )γν ψ(x2 )Aµ (x1 )Aν (x2 ) : ¯ ¯ d4 x1 d4 x2 : ψ(x1 )γµ ψ(x1 )ψ(x2 )γν ψ(x2 )Aµ (x1 )Aν (x2 ) : ¯ ¯ d4 x1 d4 x2 : ψ(x1 )γµ ψ(x1 )ψ(x2 )γν ψ(x2 )Aµ (x1 )Aν (x2 ) : ¯ ¯ d4 x1 d4 x2 : ψ(x1 )γµ ψ(x1 )ψ(x2 )γν ψ(x2 )Aµ (x1 )Aν (x2 ) : ¯ ¯ d4 x1 d4 x2 : ψ(x1 )γµ ψ(x1 )ψ(x2 )γν ψ(x2 )Aµ (x1 )Aν (x2 ) : ¯ ¯ d4 x1 d4 x2 : ψ(x1 )γµ ψ(x1 )ψ(x2 )γν ψ(x2 )Aµ (x1 )Aν (x2 ) : ¯ ¯ d4 x1 d4 x2 : ψ(x1 )γµ ψ(x1 )ψ(x2 )γν ψ(x2 )Aµ (x1 )Aν (x2 ) : ¯ ¯ d4 x1 d4 x2 : ψ(x1 )γµ ψ(x1 )ψ(x2 )γν ψ(x2 )Aµ (x1 )Aν (x2 ) : (a) (b) (c) (d) (e) (f ) (g) (h) In this expansion we have used the modiﬁcation to Wick’s theorem that says that “no equal time contractions are allowe” (no e.c. QED 7. GreinerFQ 238] We now need to introduce some additional Feynman diagrams corresponding to the above contractions. Greiner has a nice discussion [Greiner 238] showing that this prescription eliminates the socalled “tadpole” diagrams. .104 CHAPTER 7.2. These are illustrated in the ﬁgure below.t.2 Second order Smatrix From the Dyson expansion. [Schwabl 337. Also in the above expansion we don’t include contributions of the form ¯ ¯ ψ(x1 )ψ(x2 ) = ψ(x1 )ψ(x2 ) = 0 because they give zero contribution. As previously mentioned Greiner states this somewhat diﬀerently [Greiner 238] but the result is the same.
These are shown in the Figure below.x ¯ (a) Fermion contraction ψ(x2 )ψ(x1 ) = iSF (x2 − x1 ) (b) Photon contraction Aµ (x2 )Aν (x1 ) = iSF (x2 − x1 ) Using the above diagrams and also the ones shown previously we are in a position to draw the Feynman diagrams corresponding to all the terms in the second order Smatrix. 239 .2. Feynman diagrams for 2nd order Smatrix NNN identical to GreinerFQ. QED SMATRIX 105 x2 x2 x1 (a) x1 (b) Fig.5. x. Fig. p. ﬁg 8.7.
106 CHAPTER 7. 7. f >≡ b† or < f  ≡< 0ak. To illustrate this (1) let’s evaluate the ﬁrst of the eight < f S (1) i > terms.2. . and we wish now to evluate Smatrix elements.s1 0 > and the ﬁnal state consists of the scattered electron together with the produced photon.4 (b) which corresponds to an electron radiating a photon.s1 a† 0 k1 . QED 7.3 First order Smatrix elements We previously considered the ﬁrst order Smatrix.s1 k. i >≡ b† k1 . if the i > and < f  states above are used.e.e. For deﬁniteness let’s consider the diagram of Fig. i. denoted < f S1 i >. In that case the initial state would be creation of an electron from the vacuum.λ bk 1 . However all but one (the second) of these terms will be zero. i.λ > We wish to evaluate the matrix element < f Si > and the ﬁrst order Smatrix S (1) contains 8 terms.
5 Invariant amplitude and lepton tensor Electronmuon scattering Invariant amplitude 7.4 Second order Smatrix elements Electronelectron (Moeller) scattering 7.3. CASIMIR’S TRICK & TRACE THEOREMS 107 7.2 Casimir’s trick .2.3 7.2.1 Casimir’s trick & Trace theorems Average over initial states / Sum over ﬁnal states Polarized ﬁnal states / Unpolarized initial staes Unpolarized initial and ﬁnal states 7.3.7.3.
This action might not be possible to undo. Are you sure you want to continue?
We've moved you to where you read on your other device.
Get the full title to continue reading from where you left off, or restart the preview.