Detailed Lecture Notes in Quantum Field PDF

Detailed lecture notes in Quantum Field
Theory
Diego Restrepo
Instituto de Fsica
Universidad de Antioquia
2012
ii
License
All materials are licensed under the Creative Commons Attribution-Share Alike 3.0 United
States License.
iii
iv
Contents
Introduction 1
1 Classical Field Theory 3
1.1 Lagrangian Formulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.1.1 Ecuaciones de Euler-Lagrange . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.1.2 Teorema de Noether para simetras internas . . . . . . . . . . . . . . . . . . . 5
1.1.3 Teorema de Noether para simetras externas . . . . . . . . . . . . . . . . . . . 7
1.2 Global gauge invariance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
1.3 Local phase invariance in the Scr odingers Lagrangian . . . . . . . . . . . . . . . . . . 14
1.4 Notaci on relativista . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
1.4.1 Ejemplos de cuadrivectores . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
1.4.2 Lorentz tranformation for elds . . . . . . . . . . . . . . . . . . . . . . . . . . 23
1.5 Vector eld Lagrangian . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
1.5.1 Energa del campo electromagnetico . . . . . . . . . . . . . . . . . . . . . . . . 30
1.6 Scr odinger Equation in presence of the electromagnetic eld . . . . . . . . . . . . . . 33
1.6.1 Euler-Lagrange equation for
. . . . . . . . . . . . . . . . . . . . . . . . . . 35
1.6.2 Euler-Lagrange equation for A
. . . . . . . . . . . . . . . . . . . . . . . . . . 38
1.6.3 Conserved currents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
1.7 Gauge Transformation Group . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
1.8 Proca Equation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
1.9 Klein-Gordon Equation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
1.9.1 Complex scalars . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
1.10 Lorentz transformation of the elds . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
1.11 Diracs Action . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
1.11.1 Lorentz transformation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
1.11.2 Corriente conservada y Lagrangiano de Dirac . . . . . . . . . . . . . . . . . . 51
1.11.3 Tensor momento-energa . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
1.11.4 Ecuaciones de Euler-Lagrange . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
v
vi CONTENTS
1.11.5 Propiedades de las matrices de Dirac . . . . . . . . . . . . . . . . . . . . . . . 57
1.11.6 Lorentz Group . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
1.11.7 Lorentz invariance of the Dirac Action . . . . . . . . . . . . . . . . . . . . . . 62
1.11.8 Diracs Lagrangian . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
1.12 Electrodinamica Cu antica . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
1.13 Cromodinamica Cu antica . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
1.13.1 Ecuaciones de EulerLagrange . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
1.13.2 Derivada covariante adjunta . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
1.14 Spontaneous symmetry breaking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
1.15 Fermiones quirales de cuatro componentes . . . . . . . . . . . . . . . . . . . . . . . . 90
1.16 Standard model Lagrangian . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
1.16.1 Spontaneous symmetry breaking in SU(3)
c
SU(2)
L
U(1)
Y
. . . . . . . . . 97
1.16.2 Yukawa Lagrangian . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
1.16.3 Fermion-gauge interactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
1.16.4 Self-interactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
1.16.5 Lagrangiano del modelo est andar para la primera generacion . . . . . . . . . . 108
1.16.6 Din amica de sabor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
1.17 Fenomenologa Electrodebil . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
1.17.1 Decaimientos debiles mediados por corrientes cargadas . . . . . . . . . . . . . 116
1.18 Calculo de procesos . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
2 Computational QFT 121
2.1 LanHEP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
2.2 CalcHEP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
2.3 LanHEP/CalcHEP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
3 Second quantization 131
3.1 Quantization of the nonrelativistic string . . . . . . . . . . . . . . . . . . . . . . . . . 131
3.1.1 The clasical string . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131
3.1.2 Quantization of the string . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
3.1.3 Generalization to three dimensions . . . . . . . . . . . . . . . . . . . . . . . . 146
3.2 Quantization of the Klein-Gordon eld . . . . . . . . . . . . . . . . . . . . . . . . . . 147
3.3 Quantization of Fermions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156
4 Quantization of the electromagnetic eld 165
4.1 Preliminaries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165
4.2 Quantization of the electromagnetic eld . . . . . . . . . . . . . . . . . . . . . . . . . 167
CONTENTS vii
5 Propagators 169
5.1 Scalars . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169
5.2 Fermions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169
5.3 Bosons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170
6 Smatrix 171
6.1 The Smatrix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171
6.2 Relativistic and no relativistic normalizations . . . . . . . . . . . . . . . . . . . . . . 173
6.3 Process probability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
6.4 Cross Section . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177
6.4.1 2to2 cross section . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181
6.5 Decay Rates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187
6.5.1 Two body decays . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 188
6.6 Backup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189
7 Two body decays 191
7.1 Particle decays . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191
7.2 Width decay . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 192
7.3 e
+
e
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196
8 Feynman Rules 197
8.1 Interaction picture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197
8.2 Atomic decay . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203
8.3 Yukawa interaction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207
8.4 Wick Theorem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 211
8.5 Scattering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 213
9 Three body decays 223
9.1 Muon decay . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223
9.1.1 Amplitude estimation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223
9.1.2 Amplitude calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 226
9.2 three body decays in radiative seesaw . . . . . . . . . . . . . . . . . . . . . . . . . . . 235
9.A Sample point . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 242
9.B Preliminary discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 242
10 Renomalization at 1-loop 245
10.1 Self-energy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 245
10.2 Other . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 249
Introduction
We have organized the topics in order of complexity, and, in the same spirit than in previous book
[1], we have tried to write the calculations as detailed as possible. In Chapter 3 we included the
building blocks of quantum eld theory, in Chapter 6 we introduce the Smatrix in the Scr odinger
Picture separating the kinematical and normalization factors from the matrix element. Then the
expressions for the decay rates and cross sections are obtained. The explicit calculation of the
matrix element from the expansion of the Smatrix to obtain the Feynman rules, is postponed to
Chapter 8. In Chapter 7 we use the Feynman rules necessary to calculates the matrix element, and
develop the techniques associated to the squaring of the matrix element. In Chapter 8 we obtain the
Feynman rules used in two body decays directly from the rst order expansion of the Smatrix in
the interaction picture. The subsequent chapters have applications of the techniques developed to
the calculation of tree-level, Chapter 9 and loop processes.
This notes are based in books [2], [3], [4]. In each Chapter or Section the main reference used
is cited. Also, we have included material developed by students Juan Alberto Yepez, Jose David
Ruiz

Alvarez. This notes are written in English, because at this level it is expected that any physics
student be uently in reading technical texts in this language.
This work have been partially supported by Dedicacion Exclusiva 2008-2009 project: RR 26663
1
2 CONTENTS
Chapter 1
Classical Field Theory
This chapter is a summary of the main topics developed in the course Hacia la teora cuantica
de campos [1]. We will introduce special relativity as the necessary ingredient to guarantee the
local conservation of electric charge in quantum mechanics. The symmetries of the electromagnetic
Lagrangian will be extended to include the electron, as one Dirac spinor. The resulting Quantum
Electrodynamics theory will be used as a paradigm to explain the other fundamental interactions.
1.1 Lagrangian Formulation
1.1.1 Ecuaciones de Euler-Lagrange
Denamos
=

x
, (1.1)
En tres dimensiones, la accion de la se puede escribir como:
S[,
] =
_
R
d
4
xL(,
) (1.2)
donde d
4
x = dt dxdy dz. Considere primero una variacion solo de los campos, tal que (x = x
)
(x) =
(x) (x) (1.3)

De otro lado, con x = x
x, la expansi on de Taylor para f(x + x) es

f(x + x) = f(x) +
f
x
x + (1.4)
3
4 CHAPTER 1. CLASSICAL FIELD THEORY
Para L, tenemos de la ec. (1.3)
L(
) = L( + ,
())
= L +
L
+
L
(
() (1.5)
Entonces, de imponer que S = 0, tenemos
S = S
S =
_
R
d
4
x L(
)
_
R
d
4
x L(,
)
=
_
R
d
4
x
_
L
+
L
(
()
_
=
_
R
d
4
x
_
L
_
L
(
)
___
+
_
R
d
4
x
_
L
(
_
S =
_
R
d
4
x
_
L
_
L
(
)
___
+
_
_
L
(
_
d
= 0. (1.6)
Donde hemos aplicado el Teorema de Gauss
_
V
Ad
3
x =
_
S
A dS (1.7)
generalizado a cuatro dimensiones. Como la variaci on de es cero sobre la hipersupercie resulta
_
R
d
4
x
_
L
_
L
(
)
___
= 0. (1.8)
Como es cualquier posible variacion entre las fronteras de la hipersupercie, el integrando debe
anularse y resultan las ecuaciones de Euler-Lagrange:
_
L
(
)
_
= 0. (1.9)
La densidad Lagrangiana
L
= L +
((x)) (1.10)
donde (x) es cualquier funcion de los campos de la densidad Lagrangiana original, da lugar a la
Acci on
S
=
_
R
d
4
x L
=
_
R
d
4
x L +
_
R
d
4
x
=
_
R
d
4
x L +
_
=S , (1.11)
1.1. LAGRANGIAN FORMULATION 5
para una hipersupercie sucientemente grande. De modo que dos densidades lagrangianas que
dieran solo en derivadas totales dan lugar a la misma Acci on.
Usando el principio de mnima acci on en terminos del campo , tenemos que para la densidad
Lagrangiana (??)
L =
1
2
_
1
v
2
_
t
_
2
z
_
2
_
, (1.12)
las ecuaciones de Euler-Lagrange (1.9)
0
_
L
(
0
)
_
+
3
_
L
(
3
)
_
=0
t
_
L
(/t)
_
+

z
_
L
(/z)
_
=0
1
v
2
t
_
t
_

z
_
z
_
=0
1
v
2
t
2

2
z
2
=0 , (1.13)
que corresponde a la ecuacion de onda.
Generalizando a tres dimensiones vemos que la ecuaci on para una onda propagandose a una
velocidad v,
1
v
2
t
2

2
= 0 , (1.14)
proviene de una densidad Lagrangiana (hasta derivadas totales)
L =
1
2
_
1
v
2
_
t
_
2

_
=
1
2
_
1
v
2
0

i
_
. (1.15)
1.1.2 Teorema de Noether para simetras internas
Para un campo complejo la ec. (1.2) se generaliza a
S[,
] =
_
R
d
4
x L(,
) (1.16)
Usando el mismo procedimiento, se obtiene
S =
_
R
d
4
x
_
L
_
L
(
)
___
+
_
R
d
4
x
_
L
_
L
(
)
___
+
_
R
d
4
x
_
L
(
)
+
L
(
)
_
= 0. (1.17)
Usando de nuevo el Teorema de Gauss resultan las ecuaciones de Euler Lagrange para y
_
L
(
)
_
= 0,
_
L
(
)
_
= 0. (1.18)
De otro lado, si asumimos que y
satisfacen las ecuaciones de EulerLagrange, en lugar de asumir

que y
se anulan sobre la hipersupercie, los dos primeros terminos de la ec. (1.17) se anulan
y tendremos que para que S = 0:
_
R
d
4
x (
) = 0, (1.19)
donde,
J
=
_
L
(
)
_
+
_
L
(
)
_
(1.20)
Entonces J
satisface la ecuaci on de continuidad:
= 0 (1.21)
J
0
t
+ J = 0 (1.22)
Integrando con respecto al volumen
_
V
J
0
t
d
3
x +
_
V
Jd
3
x = 0,
_
V
J
0
t
d
3
x +
_
S
J dS = 0, (1.23)
Escogiendo una supercie sucientemente grande que abarque toda la fuente de densidad = J
0
, de
la corriente J, el segundo integrando es cero y
d
dt
_
V
d
3
x = 0. (1.24)
Este resultado es conocido como Teorema de Noether.

Este establece que para toda transformacion
continua del tipo (1.3), debe existir una cantidad conservada, dQ/dt = 0, que en este caso corresponde
a
Q =
_
V
d
3
x. (1.25)
Figure 1.1: Traslaci on de funci on y coordenadas en una dimensi on: (x) =
(x
)
1.1.3 Teorema de Noether para simetras externas
Para el caso de una simetra externas, por ejemplo la correspondiente a una traslacion espacio
temporal
x
=x
+ a
=a
(1.26)
tenemos
(x
) =
(x + a) (1.27)

(x) +

(x)
x
(1.28)
= [(x) + (x)] +

x
[(x) + (x)]a
(1.29)
(x) + (x) +
(x)
x
, (1.30)
donde, por simplicidad, es de nuevo un campo real, y en el ultimo paso hemos despreciado un
termino de orden a
. Entonces,
(x)
(x
) (x) = (x) +
(x)
x
. (1.31)
Para una traslacion, (x) = 0, ver gura 1.1. De modo que
= (
)a
, (1.32)
y la transformaci on del campo como consecuencia de la traslaci on es
(x)
(x) = (x) + (x) = (x) (
(x))a
. (1.33)
Si a
es constante (un an alisis mas general es hecho en [?])

d
4
x
= d
4
x (1.34)
En este caso, asumiendo que el campo satisface las ecuaciones de Euler-Lagrange y usando la ec. (1.32)
y (1.9) tenemos
S =
_
R
d
4
x L(
, x
)
_
R
d
4
x L((x),
(x), x)
=
_
R
d
4
x L( + ,
(), x
+ a
)
_
R
d
4
x L
_
R
d
4
x
_
L +
L
+
L
(
() + (
L)a
_
R
d
4
x L
=
_
R
d
4
x
_
L
+
L
(
() + (
L)a
_
=
_
R
d
4
x
__
_
L
(
)
__
+
L
(
() + (
L)a
_
=
_
R
d
4
x
_
_
L
(
_
+ (
L)a
_
=
_
R
d
4
x
_
L
(
)
+La
_
=
_
R
d
4
x
L
(
+La
_
=
_
R
d
4
x
L
(
+L
(a
)
_
=
_
R
d
4
x
__
L
(
)
(
) +
L
_
a
_
(1.35)
=
_
R
d
4
x
(T
) = 0. (1.36)
Y por consiguiente
= 0, (1.37)
De modo que para cada , con a
,= 0, se satisface:
= 0, (1.38)
donde
T
=
L
(
)
(
L (1.39)
El tensor T
proviene de asumir la homogeneidad del espacio y el tiempo y es llamado el tensor de

momentumenerga.
Para una traslacion temporal: = 0, se genera entonces la ecuaci on de continuidad:
0
= 0 (1.40)
Donde la densidad de Energa, o m as de forma mas general: la densidad Hamiltonina corresponde a
T
0
0
1 = T
0
0
=
L
L (1.41)
= (x)
(x)
t
L. (1.42)
Comparando con la expresi on correspondiente en la formulaci on Lagrangiana de la Mecanica Clasica,
tenemos que si (x) es la variable can onica, la variable can onica conjugada es (x)
(x) =
L
((x)/t)
. (1.43)
El teorema de Noether en este caso establece que la invarianza de la Accion bajo traslaciones tem-
porales da lugar a la ecuaci on de continuidad (1.38) para = 0
0
= 0 (1.44)
cuya carga conservada corresponde a la energa
H =
_
V
d
3
x T
0
0
=
_
V
d
3
x 1. (1.45)
De igual forma la invarianza bajo traslaciones espaciales de lugar a ecuaciones de continuidad para
cada componente = i (i = 1, 2, 3)
i
= 0, (1.46)
cuyas densidad de cargas conservadas, T
0
i
, que en forma vectorial escribiremos como T
0
, dan lugar
a la conservacion del momentum
P =
_
V
d
3
x T
0
. (1.47)
Generalizando a un campo complejo
T
=
L
(
)
(
) + (
)
L
(
L (1.48)
1.2 Global gauge invariance
Haciendo = 1, el Lagrangiano que da lugar a la ecuaci on de Schr odinger es
L(,
) =
1
2m

i
2
_
t

_
+
V (1.49)
=
1
2m
i

i
2
(
0

0
) +
V .
Aplicando las ecuaciones de Euler-Lagrange (1.18) para la funci on de onda
obtenemos la ecuaci on
de Scrodinger con = 1:
0 =
_
L
(
)
_
=
0
_
L
(
0
)
_
+
i
_
L
(
i
)
_
. (1.50)
Como
L
(
0
)
=
i
2
L
(
0
)
=
i
2
L
(
i
)
=
1
2m
L
(
i
)
=
1
2m
i
(1.51)
L
=
i
2
V
L
=
i
2
0
+ V .
Entonces, reemplazando la ec. (1.51) en la ec. (1.50), tenemos
0 =
_
L
(
)
_
=
0
_
i
2
_
+
i
_
1
2m
i
2
0
+ V
_
=
i
2
0
+
1
2m
i
+
i
2
0
V . (1.52)
1.2. GLOBAL GAUGE INVARIANCE 11
Que puede escribirse como
i

t
=
_
1
2m
2
+ V
_
. (1.53)
El Lagrangiano en ec (1.49), y por consiguiente la Acci on, es invariante bajo una transformacion
de fase

= e
i
. (1.54)
Por consiguiente, de acuerdo al Teorema de Noether, debe existir una cantidad conservada. La
corriente conservada se obtine de la ec. (1.20). Para los campos y
, tenemos
=
= (e
i
1) i (1.55)
. (1.56)
Usando ademas la ec. (1.51) en la denicion de J
0
dada por la ec. (1.20), tenemos
J
0
=
_
L
(
0
)
_
+
_
L
(
0
)
_
=
i
2
(i) + (i
)
i
2
, (1.57)
y
J
i
=
_
L
(
i
)
_
+
_
L
(
i
)
_
=
1
2m
(i) + (i
)
1
2m
=
i
2m
(
i
i
) . (1.58)
Entonces, normalizando apropiadamente la corriente escogiendo = 1, tenemos
J
0
=
(1.59)
J =
i
2m
(
) . (1.60)
De acuerdo a la ec. (1.59), la cantidad conservada corresponde a la probabilidad de la funci on de
onda y normalizando apropiadamente la ec. (1.25)
Q
=
_
V
d
3
x = 1. (1.61)
En cuanto a las simetras externas, tenemos de la ec. (1.39) que da lugar a las ecuaciones de
continuidad (1.44)(1.46)
0
= 0,
i
= 0 (1.62)
Las cargas conservadas corresponden entonces a T
0
0
y T
0
i
. Usando las ecs. (1.51) en la ec. (1.48)
T
0
i
=
L
(
0
)
(
i
) + (
i
)
L
(
0
)
T
0
i
=
i
2
(
i
) +
i
2
(
i
) (1.63)
Entonces, deniendo
T
0
=
i
2
(
) (1.64)
Adem as
T
0
=
i
2
((
)
= i
+
i
2
(
) . (1.65)
Integrando en el volumen
_
V
T
0
d
3
x = i
_
V
d
3
x +
i
2
_
V
d
3
x (1.66)
De acuerdo a la ec. (1.61), la ultima integral es una constante y
_
V
T
0
d
3
x = i
_
V
d
3
x
p =
_
V
pd
3
x (1.67)
De modo que p son las cargas conservadas asociadas al valor esperado el operador de momentum
p = i. (1.68)
De otro lado
T
0
0
=
L
(
0
)
0
+
0
L
(
0
)
L
=
i
2
0
+
i
2

1
2m
i
+
i
2
(
0

0
V
=
1
2m
V (1.69)
1.2. GLOBAL GAUGE INVARIANCE 13
Como las corrientes solo est an determinadas hasta un factor de proporcionalidad, denimos
1 T
0
0
=
1
2m
V
=
1
2m
(
)
1
2m
2
+
V . (1.70)
Integrando sobre el volumen y usando la ec. (1.67)
_
V
1d
3
x =
1
2m
_
V
(
) +
_
V
1
2m
2
+ V
_
d
3
x
=
1
2m
_
V
(
) +
_
V
1
2m
2
+ V
_
d
3
x
=
i
2m
p +
_
V
1
2m
2
+ V
_
d
3
x
=
_
V
1
2m
2
+ V
_
d
3
x . (1.71)
Entonces
H
_
V
1d
3
x =
_
V
1
2m
2
+ V
_
d
3
x
=
_
V
d
3
x
H =
H. (1.72)
Que es un resultado bien conocido de la mec anica cuantica.
Como
H =
1
2m
p
2
+

V , (1.73)
podemos escribir la ec. (1.53) como
i

t
=

H . (1.74)
Podemos identicar entonces los operadores de energa y momentum.
H = i

t
, p = i . (1.75)
Retornando a la ec. (1.67), tenemos que para la soluci on de partcula libre de la ecuaci on de
Schr odinger
= Ae
ikx
, (1.76)
la condicion de normalizaci on en ec. (1.61) implica que [A[
2
= 1/L
3
, y
_
V
T
0
d
3
x = k. (1.77)
Ejercicio: De la ec. (1.72) obtenega la densidad Hamiltoniana, y usando la ec. (1.41) encontrar la densidad
Lagrangiana (1.49).
1.3 Local phase invariance in the Scrodingers Lagrangian
When we discuss the wave function (x), x represents the point in space at which we want to know
the value of the wave function. Since complex numbers are, well, complex, you cant represent
them by a position on a simple number line. Instead, the have to be represented by a point in a
twodimensional plot.
In addition the length of the arrow pointing to the complex number we also need an angle to
specify exactly how to draw the arrow pointing to the complex number. The observable is encoded
into the length of the arrow representing the value of the complex valued wave function at that point
of the spacetime. Its angle is unobservable.
The complex number (x) in the Scrodinger equation is just the number whose square is the
relative probability of nding the object at that point.
Now, suppose that you arbitrarily decide to make a change of phase of the wave function to
change, at every point in space, the angle of the complex number makes with the real axis. Here
is the critical point: Is this change phase is global, if the phase that you change the phase angle
is the same everywhere in space, the this change of phase will not destroy the delicate and essential
balance between the kinetic and potential energy in the Scrodinger equation.
However, in the view implemented by Einsteins relativity, the need to require that quantum
mechanical systems be unaltered only by global changes of phase seemed to be very unnatural. Once
you choose the phase of the wave function at one space-time point, the requirement of global phase
invariance xes it at all other space-time points:
As usually conceived however, this arbitrariness is subject to the following limitation: once one
choose [the phase of the wave function] at one spacetime point, one is then not free to make
any choices at other spacetime points.
It seems that it is not consistent with the localized eld concept that underlies the usual physical
theories. In the present paper we wish to explore the possibility of requiring all the interactions
to be invariant under independent [change of phases] at all space-time points.
Yang-Mills, Physical Review, 1954
1.3. LOCAL PHASE INVARIANCE IN THE SCR
ODINGERS LAGRANGIAN 15
This is similar to what happens in electromagnetic theory expressed in terms of scalar and vector
potentials. The can be changed by arbitrary functions in a such way that the measured electric
and magnetic elds remain invariant. As we will see, this feature is deeply connected with the local
conservation of electric charge.
We start again with the Scr odinger Lagrangian as written in eq. (1.49):
L(,
) =
1
2m

i
2
_
t

_
+
V (1.78)
=
1
2m
i

i
2
(
0

0
) +
V .
This Lagrangian is not invariant under local phase changes of the wave function:
_
e
i(x)
_
=
_
e
i(x)
_
+ e
i(x)
=e
i(x)
(i
(x)) + e
i(x)
=e
i(x)
[i
(x) +
] . (1.79)
In order to have a new Lagrangian invariant under local phase changes, or local gauge transformations,
we need to introduce a new term to compensate for the term arising from the derivate of e
i(x)
:
T
=(
+ X
)
_
e
i(x)
_
=e
i(x)
[i
(x) +
] + X
_
e
i(x)
_
=e
i(x)
_
i
(x) +
+ X
. (1.80)
The transformation condition of the new term X
, in order to compensate for the term arising from

the derivative of the local phase, i
(x), is just that

X
= X
(x) . (1.81)
Replacing back in Eq. (1.80) we have
T
(T
= T
=(
+ X
)
_
e
i(x)
_
=e
i(x)
[i
(x) +
+ X
(x)]
=e
i(x)
[
+ X
]
=e
i(x)
(T
) . (1.82)
Note that T
transforms like the eld , and because of this is called the covariant derivative of
. Similarly
(T
(T
=(
+ X
)
_
e
i(x)
_
=
_
i
(x) +
+ X
+ i
(x)
e
i(x)
=
_
+ X
e
i(x)
=(T
e
i(x)
. (1.83)
It is convenient to redene X
in terms of A
:
A

1
iq
X
, (1.84)
such that the covariant derivative can be conveniently written as
T
+ iqA
. (1.85)
The transformation properties of A
can be obtained from the X
transformation in eq. (1.81):

iqA
iqA
= iqA
(x)
A
= A
1
q
(x) . (1.86)
We dene local gauge invariance as an arbitrary way of choosing the complex phase factor of a
charged eld
1
at all space time points.
In this way, we can change the original Lagrangian for a new one which is invariant under local
phase transformations:
L(,
, A
) =
1
2m
(T
i
)
T
i

i
2
[
T
0
(T
0
)
] +
V (x). (1.87)
where
A
= A
1
q
(x) . (1.88)
1
like the electron eld as described by the usual Scrodinger equation.
1.4. NOTACI
ON RELATIVISTA 17
This is just the gauge transformation which left the Electromagnetic elds invariant. In fact, the
new Lagrangian is now invariant under the local phase transformations
L L
=
1
2m
(T
i
)
(T
i
)
i
2
_
(T
0
)
(T
0
)
V (x)
=
1
2m
(T
i
)
e
i(x)
e
i(x)
(T
i
)
i
2
_
e
i(x)
e
i(x)
(T
0
) (T
0
)
e
i(x)
e
i(x)
e
i(x)
e
i(x)
V (x).
=L. (1.89)
To preserve invariance one notices that it is necessary to counteract the variation of with x, y,
z, and t by introducing the electromagnetic eld A
. In this way, the electromagnetic interaction is

obtained as the result of impose local gauge invariance under U(1) (local phase transformations). To
fully implement the gauge principle, i.e, the paradigm to obtain the interactions as the result of the
gauge invariance, we need to introduce some concepts of special relativity to be developed below.
1.4 Notaci on relativista
Las transformaciones de Lorentz se denen como la transformaciones que dejan invariante al producto
escalar en el espacio de Minkowski denido como
a
2
= g
= a
0
2
a
i
a
i
= a
0
2
a a (1.90)
donde , = 0, 1, 2, 3, i = 1, 2, 3 y se asume suma sobre ndices repetidos. Adem as
a
(1.91)
Finalmente la metrica usada se dene como
g
=
_
_
_
_
1 0 0 0
0 1 0 0
0 0 1 0
0 0 0 1
_
_
_
_
(1.92)
donde g
denota la forma matricial del tensor g
.
El producto de dos cuadrivectores se dene en forma similar como
a
= g
= a
0
b
0
a b (1.93)
El inverso de la metrica es
g
1
= g
(1.94)
tal que
g
and a
= g
(1.95)
Bajo una transformaci on de Lorentz.
a
(1.96)
a
La invarianza del producto escalar en ec. (1.93)

a
=a
= g
= g
= g
, (1.97)
da lugar a
g
or g
T
g
_
. (1.98)
En notacion matricial
g =
T
g. (1.99)
From eq. (1.98) we also have
g
=g
, (1.100)
or
. (1.101)
Since
_
1
_
(1.102)
the inverse of is
_
1
_
, (1.103)
or
_
1
_
, (1.104)
1.4. NOTACI
ON RELATIVISTA 19
Example: Lorentz invariance
a
b
p
=
b
p
=
_
1
_
b
p
=
b
p
=a
.
Como un ejemplo de Transformacion de Lorentz considere un desplazamiento a lo largo del eje x
x
=
_
_
_
_
t
x
y
z
_
_
_
_
_
_
_
_
t
_
_
_
_
=
_
_
_
_
t+vx
1v
2
x+vt
1v
2
y
z
_
_
_
_
=
_
_
_
_
cosh sinh 0 0
sinh cosh 0 0
0 0 1 0
0 0 0 1
_
_
_
_
_
_
_
_
t
x
y
z
_
_
_
_
=
, (1.105)
donde
cosh = sinh = v, and =
1
1 v
2
. (1.106)
y, por ejemplo:
t cosh + x sinh = (t + vx) =
t + vx
1 v
2
. (1.107)
El
denido en la ec. (1.105) satisface la condici on en ec. (??),
T
g =
_
_
_
_
cosh sinh 0 0
sinh cosh 0 0
0 0 1 0
0 0 0 1
_
_
_
_
_
_
_
_
1 0 0 0
0 1 0 0
0 0 1 0
0 0 0 1
_
_
_
_
_
_
_
_
cosh sinh 0 0
sinh cosh 0 0
0 0 1 0
0 0 0 1
_
_
_
_
=
_
_
_
_
cosh sinh 0 0
sinh cosh 0 0
0 0 1 0
0 0 0 1
_
_
_
_
_
_
_
_
cosh sinh 0 0
sinh cosh 0 0
0 0 1 0
0 0 0 1
_
_
_
_
=
_
_
_
_
cosh
2
sinh
2
cosh sinh cosh sinh 0 0
cosh sinh cosh sinh sinh
2
cosh
2
0 0
0 0 1 0
0 0 0 1
_
_
_
_
=g (1.108)
Denotaremos los cuadrivectores con ndices arriba como
a
= (a
0
, a
1
, a
2
, a
3
) = (a
0
, a) (1.109)
Entonces el correspondiente cuadrivector con ndices abajo, usando la ec. (1.91), es
a
= (a
0
, a
1
, a
2
, a
3
) = (a
0
, a
1
, a
2
, a
3
) = (a
0
, a). (1.110)
Con esta notacion, el producto escalar de cuadrivectores puede expresarse como el producto escalar
de los dos vectores de cuatro componente a
y a
.
1.4.1 Ejemplos de cuadrivectores
x
=(x
0
, x
1
, x
2
, x
3
) = (t, x, y, z) = (t, x) (1.111)
p
=(p
0
, p
1
, p
2
, p
3
) = (E, p
x
, p
y
, p
z
) = (E, p) (1.112)
De la relatividad especial tenemos que
E =m
p =mv . (1.113)
Por lo tanto, ya que v
2
= v
2
= [v[
2
E
2
p
2
=
2
m
2
(1 v
2
) = m
2
. (1.114)
El invariante de Lorentz asociado a p
corresponde a la ecuacion de momento energa una vez se

identica la masa de una partcula con su cuadrimomentum
p
2
= p
= m
2
= E
2
p
2
(1.115)
De [?]
The intuitive understanding of this equation is that the energy of a particle is partially
due to its motion and partially due to the intrinsic energy of its mass. The application
to particle detectors is that if you know the mass of a particular particle, or if its going
so fast that its energy and momentum are both huge so that the mass can be roughly
ignored, then knowing the energy tells you the momentum and vice versa
1.4. NOTACI
ON RELATIVISTA 21
Para p = 0, es decir cuando la partcula est a en reposo se reduce a la famosa ecuaci on E = mc
2
(c = 1)
Del electromagnetismo tenemos
J
= (J
0
, J) = (, J) (1.116)
A
= (A
0
, A) = (, A) (1.117)
Del calculo vectorial
=
_

x
0
,

x
1
,

x
2
,

x
3
_
=
_

x
0
,

x
1
,

x
2
,

x
3
_
=
_

t
,

x
,

y
,

z
_
=(
0
, ) = (
0
, ) (1.118)
=

x
=
_

t
,

x
,

y
,

z
_
= (
0
, ) (1.119)
Por consiguiente:
=

x
(1.120)
Producto escalar:
a
= g
= a
0
b
0
a
1
b
1
a
2
b
2
a
3
b
3
= a
0
b
0
a
i
b
i
= a
0
b
0
a b (1.121)
Entonces
=
a
0
t
+ a (1.122)
La ecuaci on de continuidad
= 0 es un invariante bajo transformaciones de Lorentz:
= 0 El operador cuadr atico es, usando la ec. (1.90)

=
0
2
=

2
t
2

2
x
2

2
y
2

2
z
2
(1.123)
Los operadores de energa y momentum de la mecanica cu antica tambien forman un cuadrivector
p
= ( p
0
, p) = (
H, p) (1.124)
con

H, y p dados en la ec. (1.75). Entonces
p
= i
= i(
0
,
i
) = i(

t
, ) (1.125)
Las derivada covariante en la ec, (1.85) es en terminos de componentes:
T
0
=
0
+ iqA
0
T
i
=
i
iqA
i
. (1.126)
Deniendo D de la misma forma que el gradiente, tenemos
T
i
=
i
+ iqA
i
D =+ iqA. (1.127)
Podemos denir el cuadrivector
T
=(T
0
, D)
=(
0
, ) + iq(A
0
, A)
=(
0
,
i
) + iq(A
0
, A
i
)
=(
0
,
i
) + iq(A
0
, A
i
)
=(
0
+ iqA
0
,
i
+ iA
i
)
=(T
0
, T
i
)
=(T
0
, T
i
) , (1.128)
donde hemos usado
T
i
=
i
+ iqA
i
. (1.129)
Adem as A
tiene la transformacion gauge

A A
= A+ A
0
A
0
= A
0

t
(1.130)
En notacion de cuadrivectores
A
=
_
A
0

t
, A+
_
=
_
A
0

t
, A
i
+
i
_
=
_
A
0
0
, A
i
_
=
_
A
0
, A
i
_
0
,
i
_
A
=A
. (1.131)
Note that the eq. (1.131) can be written as
A
= A
(x) (1.132)
which is just the transformation obtained in eq. (1.88).
1.4. NOTACI
ON RELATIVISTA 23
1.4.2 Lorentz tranformation for elds
The scalar eld is dened by their properties under Lorentz transformation. In section 1.1.3 we
study the behavior of one scalar eld under a spacetime translation. Under a general Lorentz
transformation
x
, (1.133)
Now we will study the eect of a Lorentz tranformation on the eld (x), for example under a boost.
By denition the scalar eld does not change by the Lorentz transformation, the functional form is
unaltered the scalar eld still satisfy
(x)
(x
) = (x) . (1.134)
By using eq. (1.133) we have
(x
) = (
1
x
) . (1.135)
Therefore, for an arbitrary space-time point we have that the scalar eld transforms under a Lorentz
transformation as
(x)
(x) = (
1
x) . (1.136)
In order to check the Lorentz invariance of the scalar we need to obtain the Lorentz transformation
properties for
. It is convinient to invert eq. (1.133)

_
1
_
=
_
1
_
=x
, (1.137)
1
x
=
_
1
_
1
x
, (1.138)
or
1
x
=
_
1
_
1
x
, (1.139)
and the dention of the Lorentz transformation itself:
g
=
_
1
_
1
_
. (1.140)
From eq. (1.139) we can obtain the Lorentz transformation for
= /x
=
_
1
_
=
_
1
_
, (1.141)
The eld A
(x) transforms simultaneously as eld and as vector under Lorentz transformation

A
(x) A
(x
) =
(
1
x) . (1.142)
1.5 Vector eld Lagrangian
We are now are in position to answer the following question: What is the most general Lagrangian
for a the fourcomponents eld A
compatible with Lorentz invariance and the gauge transformation

A
= A
(x) ? (1.143)
Deniendo
F
El Lagrangiano que da lugar a una Acci on invariante de Lorentz para el cuadrivector A
es, hasta
derivadas totales y potencias en los campos de hasta dimensi on 4:
L =
1
4
F
1
4
G
+
1
2
m
2
A
+
1
(x)A
(x)A
(x) +
2
A
+
3
F
(x)A
(x)A
(x) +
4
G
(x)A
(x)A
, . (1.144)
Ejercicio: Show that terms like
(x)
(x), and hence F
, transforms as
1
x
_
1
x
_
(1.145)
Hint: use the Lorentz transformation properties of
in eq. (1.141).
In the case of J
:
J
(x)A
(x) g
(x)A
(x) =g
1
x
_
1
x
_
=
1
x
_
A
1
x
_
=g
1
x
_
A
1
x
_
, (1.146)
1.5. VECTOR FIELD LAGRANGIAN 25
in the case
(x)A
(x)A
(x):
(x)A
(x)A
(x)
(x
)A
(x
)A
(x
) =
_
1
_
1
x
_
A
1
x
_
A
1
x
_
=
1
x
_
A
1
x
_
A
1
x
_
=
1
x
_
A
1
x
_
A
1
x
_
,
(1.147)
and similarly for the other terms. Under a Lorentz transformation the full Lagrangian transform as
L(x) L
(x) = L(
1
x) (1.148)
Since the Action involves the integration over all the points, it is invariant under the Lorentz transfor-
mation. The J
(x) does not involves the introduction a new vector eld, because it will be identied
later as the 4current.
Terms like
K
(x)A
(x)A
(x) , (1.149)
(for K
constant) are not Lorentz invariant:

K
(x)A
(x)A
(x) K
(x)A
(x)A
(x) =K
1
x
_
A
1
x
_
A
1
x
_
. (1.150)
K
(x)A
(x)A
(x)A
(x) is Lorentz covariant but not gauge-invariant (see below).

Bajo la transformaci on gauge (1.132)
F
=(
)
=
=F
(1.151)
Si queremos que la Acci on reeja las simetras de las ecuaciones de Maxwell debemos mantener
s olo los terminos del Lagrangiano para A
en (1.144) que sean invariantes hasta una derivada total.

Bajo una transformaci on gauge, cada uno de los terminos
1
4
G
+
1
2
m
2
A
+
1
+
2
A
+
3
F
+
4
G
+K
(x)A
dan lugar a un L , =
(algo) y la Acci on no es invariante bajo la transformacion gauge. Para los

terminos restantes
L =
1
4
F
, (1.152)
usando la ec. (1.166), tenemos
L = L
L =
1
4
F
+
1
4
F
+ J
=J
+ J
(x) J
(J
) (
)(x) (1.153)
For the action
S =
_
d
4
x [
(J
) (
)(x)]
=
_
d
4
x(
)(x)
=
_
d
3
x
_

dt(
)(x) . (1.154)
In order to have S = 0 we need to assume for the while that
= 0. However we will see that

this is just a self-consistent condition.
In summary, if the electromagnetic current is conserved, then the Lagrangian is invariant under
the gauge transformation (1.143). Note that the Lagrangian density is not locally gauge invariant.
However, the action (and hence the theory) is gauge invariant.
Por lo tanto, el Lagrangiano
L =
1
4
F
(1.155)
es el m as general que da lugar a una Accion invariante de Lorentz e invariante gauge local.
The denition of F
already includes the homogeneous Maxwell equations. To see this we note

rst that the only non-zero F
components are
F
=
_
F
0
= F
i0
= 0
F
l
= F
ml
= l
(1.156)
For = 0 we have
F
i0
=
i
A
0
0
A
i
= (
A
0
x
i
A
i
x
0
)
= (
A
0
x
i
+
A
i
x
0
)
= E
i
(1.157)
where
E =
A
t
. (1.158)
while for = l we have
F
ml
=
m
A
l
l
A
m
= (
lj
mi
li
mj
)
i
A
j
= (
lj
mi
li
mj
)
i
A
j
= (
li
mj
lj
mi
)
i
A
j
= (
li
mj
lj
mi
)
A
j
x
i
=
lmk
ijk
A
j
x
i
=
lmk
(A)
k
=
lmk
B
k
, (1.159)
where
B = A. (1.160)
Then we have
F
=
_
_
_
_
0 E
1
E
2
E
3
E
1
0
213
B
3
312
B
2
E
2
123
B
3
0
321
B
1
E
3
132
B
2
231
B
1
0
_
_
_
_
=
_
_
_
_
0 E
1
E
2
E
3
E
1
0 B
3
B
2
E
2
B
3
0 B
1
E
3
B
2
B
1
0
_
_
_
_
. (1.161)
From eqs. (1.158), and (1.160)
E =

t
A
=
B
t
,
and
B = (A)
= 0
which are just the homogeneous Maxwell equations. Therefore the expression
F
. (1.162)
with the F
given in (1.161), is just an equivalent form for the homogeneous Maxwell equations.
The remaining Maxwell equations can be obtained from the Euler-Lagrange equations for A
:
Con miras a calcular las ecuaciones de Euler-Lagrange para el Lagrangiano en ec. (1.155), tenemos
F
=(
)(
)
=
=g
).
Entonces
)
F
=g
).
=g
+ g
+ g
+ g
=4(
)
F
= 4F
(1.163)
Usando la ec. (1.163), tenemos
_
L
(
)
_
L
A
= 0
1
4
_

(
)
(F
)
_
+ J
= 0
+ J
= 0
= J
. (1.164)
Como era de esperarse una Acci on invariante de Lorentz e invariante gauge local, expresada en
terminos del Lagrangiano (1.155), da lugar a la Teora Electromagnetica.
Tomando la derivada con respecto a en ambos lados tenemos
. (1.165)
De la parte izquierda de esta ecuacion tenemos
=
1
2
(
)
=
1
2
(
) intercambiando ndices mudos

=
1
2
(
) conmutando derivadas
=
1
2
(
) usando antisimetra de F
= 0 ,
Por consiguiente, la cuadricorriente J
es conservada:
= 0 . (1.166)
Again, for = 0, we have
F
0
= J
0
i
F
i0
= J
0
x
i
F
i0
= J
0
E
i
x
i
= J
0
, (1.167)
and therefore
E = . (1.168)
while for = k we have
F
k
= J
k
i
F
ik
+
0
F
0k
= J
k
i
F
ki
0
F
k0
= J
k
(
ikj
B
j
)
x
i

E
k
t
= J
k
ijk
B
j
x
i

E
k
t
= J
k
(B)
k
E
k
t
= J
k
. . (1.169)
and therefore
B
E
t
= J. (1.170)
In this way the expression
= J
where F
, (1.171)
is completely equivalent to the full set of Maxwell equations:
B = 0, E +
B
t
= 0 (1.172)
E = , B
E
t
= J. (1.173)
1.5.1 Energa del campo electromagnetico
Necesitamos la expresion para F
,
F
= g
_
F
0i
= F
0
= g
00
g
ij
F
0j
= F
0i
para = 0
F
ij
= F
i
= g
ik
g
jl
F
kl
= F
ij
para = i
(1.174)
De la ec. (1.39), se tiene
T
=
L
(
)
(
L
= F
L (1.175)
La energa del campo, corresponde a la componente T
0
0
:
T
0
0
= F
0
(
0
A
) L
= F
0
(
0
A
) +
1
4
F
+ J
Usando las ecuaciones (??), (??), (1.174)

T
0
0
= F
0
(
0
A
) +
1
4
F
+ J
= F
0
(
0
A
) +
1
4
=0
..
F
0
F
0
+
1
4
=i
..
F
i
F
i
+J
= F
0
A
0
F
0
F
0
+
1
4
F
0
F
0
+
1
4
F
i
F
i
+ J
. (1.176)
Tenemos dos partes
F
0
F
0
+
1
4
F
0
F
0
+
1
4
F
i
F
i
= F
i0
F
i0
+
1
4
F
i0
F
i0
+
1
4
=0
..
F
0i
F
0i
+
1
4
=j
..
F
ji
F
ji
= F
i0
F
i0
+
1
4
F
i0
F
i0
+
1
4
F
i0
F
i0
+
1
4
F
ji
F
ji
=
1
2
F
i0
F
i0
+
1
4
F
ji
F
ji
. (1.177)
Adem as
F
0
A
0
+ J
(A
0
F
0
) + A
0
F
0
+ J
(A
0
F
0
) A
0
F
0
+ J
(A
0
F
0
) A
0
J
0
+ J
=
i
(A
0
F
0i
) J A. (1.178)
Entonces
T
0
0
=
i
(A
0
F
0i
)
1
2
F
i0
F
i0
+
1
4
F
ji
F
ji
J A
=
i
(A
0
F
0i
) +
1
2
F
i0
F
i0
+
1
4
F
ji
F
ji
J A, suma tambien sobre i, j
=
1
2
E
i
E
i
+
1
4
ijk
B
k
ijl
B
l
+
i
(A
0
E
i
) J A, suma tambien sobre i, j
=
1
2
E
2
+
1
2
kl
B
k
B
l
+ (A
0
E) J A
=
1
2
E
2
+
1
2
B
2
+ (A
0
E) J A (1.179)
Entonces, en ausencia de corrientes
1 =
1
2
E
2
+
1
2
B
2
+ (A
0
E) . (1.180)
Similarmente la densidad Lagrangiano puede escribirse como
L =
1
4
F
=
1
2
_
E
2
B
2
_
(1.181)
En vista a la ec. (1.176), ya que la densidad Lagrangiana esta denida hasta una derivada total,
como (A
0
E) =
(A
0
F
0
), la densidad Hamiltoniana tambien estar a denida hasta una derivada
total. De hecho, el Hamiltoniano es
H =
1
2
_
V
d
3
x (E
2
+B
2
) +
_
V
d
3
x (A
0
E)
=
1
2
_
V
d
3
x (E
2
+B
2
), (1.182)
y corresponde a la expresi on conocida para la energa del campo electromagnetico. Hemos usado el
hecho que en ausencia de corrientes todo lo que entra a un volumen debe salir y por consiguiente las
integrales sobre el volumen de la divergencia de cualquier vector es cero.
Similarmente el momentum total del campo, en ausencia de corrientes, corresponde al vector de
1.6. SCR
ODINGER EQUATION IN PRESENCE OF THE ELECTROMAGNETIC FIELD 33

Pointing:
T
0
i
=
L
(
0
A
i
A
=F
0
i
A
=F
0j
(
i
A
j
j
A
i
) F
0j
j
A
i
=F
0j
F
ij
F
0j
j
A
i
=F
0j
F
ij
j
(F
0j
A
i
) + (
j
F
0j
)A
i
=E
j
jik
B
k
+
j
(E
j
A
i
) + (J
0
)A
i
=(E B)
i
(A
i
E) A
i
(1.183)
En ausencia de cargas y corrientes
P
i
=
_
V
d
3
x T
0
i
=
_
V
d
3
x (E B)
i
+
_
V
d
3
x (A
i
E)
P =
_
V
d
3
x (E B) . (1.184)
1.6 Scr odinger Equation in presence of the electromagnetic
eld
Once we have established the set of elds, as in this case ,
, and A
, we should write the most

general Lagrangian. Therefore
L(,
, A
) =
1
2m
i
(T
i
)
T
i

i
2
[
T
0
(T
0
)
] +
V (x)
1
4
F
.
(1.185)
If we further assume that all interactions are obtained from the covariant derivative, then we
need only consider the free Lagrangian of each eld, but with the normal derivative replaced by the
covariant one:
L(,
, A
) =
1
2m
i
(T
i
)
T
i

i
2
[
T
0
(T
0
)
]
1
4
F
. (1.186)
The expansion of the Lagrangian in terms of the eld ,
, and A
is
L =
1
2m
i
(
i
+ iqA
i
)
(
i
+ iqA
i
)
i
2
[
(
0
+ iqA
0
) (
0
+ iqA
0
)
]
1
4
F
=
1
2m
i
(
i
iqA
i
) (
i
+ iqA
i
)
i
2
[
(
0
+ iqA
0
) (
0
iqA
0
) ]
1
4
F
=
1
2m
i
_
i
iq
A
i
i
+ iq
i
A
i
+ q
2
A
i
A
i
i
2
[
0
+ iq
A
0
(
0
) + iqA
0
]
1
4
F
=
1
2m
i
_
i
iq
A
i
i
+ iq
i
A
i
+ q
2
A
i
A
i
i
2
[
0
(
0
) + 2iq
A
0
]
1
4
F
. (1.187)
By using the sum convention upon repeated indices we have
L =
1
2m
_
i
iq
A
i
i
+ iq
i
A
i
+ q
2
A
i
A
i
i
2
[
0
(
0
) + 2iq
A
0
]
1
4
F
. (1.188)
From this we can obtain the Euler-Lagrange equation for each eld.
1.6. SCR

1.6.1 Euler-Lagrange equation for
In particular for
we have
_
L
(
)
_
=0
0
_
L
(
0
)
_
+
i
_
L
(
i
)
_
=0
i
2
0

1
2m
i
_
i
+ iqA
i
1
2m
_
iqA
i
i
+ q
2
A
i
A
i
i
2
(
0
+ 2iqA
0
)
_
=0
i
0
qA
0

1
2m
_
i
_
i
+ iqA
i
_
+ iqA
i
_
i
+ iqA
i
_
=0
i(
0
+ iqA
0
)
1
2m
(
i
+ iqA
i
)(
i
+ iqA
i
) =0
iT
0
+
1
2m
i
T
i
T
i
=0 , (1.189)
If we dene
D iqA. (1.190)
we have in components:
D
i
=
i
iqA
i
D
i
=
i
+ iqA
i
. (1.191)
Then we have the new wave equation:
iT
0
=
1
2m
D D
iT
0
=
1
2m
D
2
, (1.192)
que corresponde a la ecuaci on de Scr odinger con la derivada normal reemplazada por la derivada
covariante.
Expandiendo esta ecuacion tenemos
i
_

t
+ iqA
0
_
=
1
2m
i
(
i
+ iqA
i
)
2
_
i

t
qA
0
_
=
1
2m
i
(
i
iqA
i
)
2
_
i

t
q
_
=
1
2m
(iqA)
2
H q
_
=
i
2
2m
(iqA)
2
=
1
2m
(i+ qA)
2
=
1
2m
(iqA)
2
=
1
2m
( p qA)
2
. (1.193)
In this way, the Scr odinger equation in presence of the electromagnetic eld, can be obtained from
the original Scrodinger equation but with the minimum substitution:
H q p p qA. (1.194)
De la ecuacion (1.193) podemos obtener la ecuacion de Sch odinger en presencia de un campo
electromagnetico
i

t
=
_
1
2m
(iqA)
2
+ qA
0
_
. (1.195)
Para que la mecanica cuantica sea consistente con las ecuaciones de Maxwell es necesario que las
transformaciones gauge (1.86) de los potenciales de Maxwell esten acompa nados por una transfor-
maci on de la funcion de onda,
, donde
satisface la ecuaci on
iT
=
1
2m
D
i

t
=
_
1
2m
(iqA
)
2
+ qA
0
_
. (1.196)
Como la forma de la ecuacion (1.196) es exactamente la misma que la forma de (1.195) entonces
ambas describen la misma fsica. Se dice que la ec. (1.195) es covariante gauge, lo que signica que
mantiene la misma forma bajo una transformaci on gauge.
1.6. SCR

Ejemplo:
Demuestre que la ec. (1.196) es covariante:
Como

= e
i(x)
(1.197)
Entonces
D
= [(iqA) i] e
i(x)
= i()e
i(x)
+ e
i(x)
iqAe
i(x)
i()e
i(x)
= e
i(x)
(iqA)
= e
i(x)
(D) (1.198)
y
D
= D
(D
)
= [(iqA) i] e
i(x)
(D)
= i()e
i(x)
(D) + e
i(x)
(D) iqAe
i(x)
(D) ie
i(x)
(D)
= e
i(x)
(iqA)(D)
= e
i(x)
(D
2
) (1.199)
De la misma manera
T
= e
i(x)
(T
0
) (1.200)
De modo que
T
= e
i(x)
(T
) (1.201)
y la derivada covariante del campo transforma como el campo. Tenemos entonces que
iT
=
1
2m
D
ie
i(x)
T
0
=
1
2m
e
i(x)
D
2
iT
0
=
1
2m
D
2
(1.202)
En resumen, para
T
+ iqA
(1.203)
y reemplazando q tenemos
A
= A
(x)

= e
iq(x)
= e
iq(x)
(T
) . (1.204)
En esta convenci on q corresponde al generador de la transformacion y al par ametro de la transfor-
maci on.
1.6.2 Euler-Lagrange equation for A
Para el campo A
tenemos
_
L
(
)
_
L
A
= 0 . (1.205)
Usando el Lagrangiano en (1.187) y el resultado de (1.164) tenemos
(F
)
L
A
= 0
que da lugar a dos conjuntos de ecuaciones, una para A
i
(F
j
) +
L
A
j
=0
(F
j
)
1
2m
[iq(
i
) iq
(
i
) + 2q
2
A
i
] =0
(F
j
)
iq
2m
[(
i
(
i
) 2iq
A
i
] =0
(F
j
)
iq
2m
[(
i
) iq
A
i
(
i
) iq
A
i
] =0
(F
j
)
iq
2m
[(
i
iqA
i
)
(
i
+ iqA
i
) =0
(F
j
)
iq
2m
[(T
i
)
T
i
] =0 , (1.206)
y otra para A
0
(F
0
) +
L
A
0
=0
(F
0
) + q
=0
(1.207)
1.6. SCR

Entonces
= j
con
j
=
_
q
= 0
iq
2m
[(T
i
)
T
i
] = i
(1.208)
Que incluye el termino corriente para una partcula cargada y es diferente de la corriente de proba-
bilidad en ec. (1.57). En otras palabras es la carga electrica la que se converva localmente.
1.6.3 Conserved currents
The 4-current can be obtained directly from the Noethers Theorem:
J
=
L
=
_
L
+
L
= 0
L
+
L
= i
. (1.209)
J
0
=
i
2
(iq) iq
i
2
=q
, (1.210)
J
i
=
1
2m
[(
i
iqA
i
)
iq iq
(
i
+ iqA
i
) ]
J
i
=
iq
2m
[(T
i
)
(T
i
)] . (1.211)
When is xed to 1 as in ec. (1.57) to dene the probability, we get eq. (1.465).
It is worth to notice that for T
0
0
, and T
0
i
we should obtain
H =i

t
q p =iqA. (1.212)
1.7 Gauge Transformation Group
Ejemplo:
Muestre que los campos electromagneticos son invariantes bajo las siguientes transformaciones
A A
= A+
=

t
(1.213)
Ya que
E E
= +

t

A
t

t
= E (1.214)
B B
= A+
. .
= 0
= B (1.215)
Esto implica que diferentes observadores en diferentes puntos del espacio, usando diferentes
calibraciones para sus medidas, obtienen los mismos campos. Las ecs. (1.213), corresponden a
transformaciones gauge locales
En notacion de cuadrivectores
A
=A
(1.216)
Sea U un elemento del Grupo de Transformaciones U(1):
U = e
i(x)
U(1) (1.217)
El Grupo est a denido por el conjunto innito de elementos U
i
= e
i(x
i
)
. Entonces
Producto de Grupo
U
1
U
2
= e
i[(x
1
)+(x
2
)]
e
i(x
3
)
U(1)
Identidad:
(x) = 0 tal que U
I
= 1
Inverso
(x) = (x) tal que U
1
= e
i(x)
Note que si
A
= U A
U
1
+
i
q
(
U)U
1
(1.218)
1.8. PROCA EQUATION 41
y si es sucientemente peque no
U = e
i(x)
1 + i(x) +O(
2
) U
1
= e
i(x)
1 i(x) +O(
2
) (1.219)
Entonces
A
=[1 + i(x) +O(

2
)]A
[1 i(x) +O(
2
)] +
i
q
(i
(x))[1 + i(x) +O(

2
)][1 i(x) +O(
2
)]
=A
1
q
(x) +O(
2
) (1.220)
which is just the eq. (1.86)
1.8 Proca Equation
Consideraremos ahora el efecto de adicionar un termino de masa a la teora de Maxwell. Los campos
vectoriales masivos juegan un papel importante en fsica. Campos como W
, Z
que median las

interacciones debiles son ejemplos de campos de este tipo. Las implicaciones de una masa nita para
el foton pueden inferirse de un conjunto de postulados que hacen de las ecuaciones de Proca la unica
generalizaci on posible de las ecuaciones de Maxwell [?].
Teniendo en cuenta s olo el termino de masa en la ec. (1.155)
L =
1
4
F
+
1
2
m
2
A
. (1.221)
Usando las ecuaciones de Euler-Lagrange, tenemos
1
4
_

(
)
F
_
1
2
m
2
A
_
= 0
+ m
2
A
= J
. (1.222)
Tomando la cuadridivergencia a ambos lados de la ecuacion y usando la ec. (??), tenemos
+ m
2
+ m
2
m
2
(1.223)
De este modo, en ausencia de corrientes, la ecuaciones de Proca dan lugar a la condicion de Lorentz.
De otro lado, si asumimos que la corriente se conserva, la condicion de Lorentz tambien aparece. Por
consiguiente, si la masa de campo vectorial es diferente de cero, la condicion de Lorentz, ec. (??),
emerge como una restricci on adicional que debe ser siempre tomada en cuenta. De este modo la
libertad gauge de las ecuaciones de Maxwell se pierde completamente en la ecuaciones de Proca, que
sin perdida de generalidad se pueden reescribir, usando
= 0 y las ec. (1.222), como:
+ m
2
A
= J
+ m
2
A
= J
(+ m
2
)A
=J
(1.224)
donde esta denido en la ec. (1.123). En ausencia de corrientes, cada una de las componentes del
campo vectorial satisface la ecuaci on de Klein-Gordon (??). Por consiguiente m corresponde a la
masa del campo vectorial A
.
Aplicando la condicion de Lorentz a la ec. (1.221), obtenemos el Lagrangiano de la Ecuacion de
Proca (1.224)
L =
1
4
F
+
1
2
m
2
A
=
1
4
(
) +
1
2
m
2
A
=
1
2
1
2
m
2
A
+ J
, (1.225)
donde hemos reabsorbido un signo global que no afecta las ecuaciones de movimiento. El primer
termino que incluye solo derivadas de los campos es llamado termino cinetico y dependen s olo del
espn de las partculas. El termino cuadr atico en los campos corresponde al termino de masa, y el
ultimo corresponde a la interacci on del campo con una corriente. Cuando un Lagrangiano contiene
s olo terminos cineticos y de masa diremos que el campo que da lugar al Lagrangiano es libre de
interacciones, o simplemente que es un campo libre. Las otras partes del Lagrangiano ser an llamadas
Lagrangiano de Interaccion. De este modo podemos reescribir el Lagrangiano (1.225) como
L = L
free
+L
int
,
donde,
L
free
=
1
2
1
2
m
2
A
L
int
= J
. (1.226)
Debido a que la teora masiva ya no es invariante gauge, la condicion de Lorentz aparece au-
tom aticamente como la unica restriccion apropiada sobre el campo vectorial.
Una vez se toma en cuenta la condici on de Lorentz el campo masivo libre puede expandirse en
ondas planas con tres grados de libertad independientes de polarizacion. Dos de estos corresponden
1.9. KLEIN-GORDON EQUATION 43
a los dos estados transversos que aparecen en las ondas electromagneticas (A
1
, A
2
), y el tercero (A
3
)
corresponde a un estado longitudinal en la direcci on del momento de la partcula [?].
Aunque hemos hecho el an alisis de la ecuaci on de Proca permitiendo un termino de masa para
el fot on, las implicaciones experimentales de una teora de este tipo dan lugar a restricciones muy
fuertes sobre la masa del foton[?]. El lmite actual sobre la masa del fot on es m < 6 10
17
eV
(1.110
52
Kg) [?]. Debido al principio gauge local, desde el punto teorico se espera que la masa del
fot on sea exactamente cero. En general, los campos vectoriales puede ser generados a partir de otras
cargas no electromagneticas y pueden ser masivos. El reto durante varias decadas fue entender como
las masa de los campos vectoriales de la interaccion debil podra hacerse compatible con el principio
gauge local.
1.9 Klein-Gordon Equation
De la componente escalar de la ecuaci on de Proca, (1.225), obtenemos la ecuacion de KleinGordon
para un campo escalar real = A
0
,
L =
1
2

1
2
m
2
2
+ (1.227)
Donde es la densidad de carga que actua como fuente del campo . El Lagrangiano m as general
posible que cuya acci on sea invariante de Lorentz, para el campo escalar real (x) es
L =
1
2

1
2
m
2
2
V () , (1.228)
donde V () es alguna funcion de con operadores de dimension menor o igual a 4. Para demostrar
la invarianza de Lorentz.
The kinetic part of Klein-Gordon Lagrangian transforms as
(x)
(x) g
(x)
(x)
=g
_
_
1
_
1
x
_
_
__
1
_
1
x
_
=
_
1
_
1
_
1
x
_
1
x
_
=g
1
x
_
1
x
_
=
1
x
_
1
x
_
. (1.229)
Since (x)
(x) = (
1
x), under a Lorentz transformation the full Lagrangian transform as
L(x) L
(x) = L(
1
x) (1.230)
Since the Action involves the integration over all the points, it is invariant under the Lorentz trans-
formation.
Ejercicio Demuestre que un termino
(x)a
(x) . (1.231)
con a
constante, no es invariante de Lorentz, de modo que en efecto, el Lagrangiano en de

Klein-Gordon en (1.228) es el m as general posible (hasta terminos de intercci on en (x).)
El campo puede pensarse como proveniente de una fuente de la misma manera como el campo
electromagnetico surge de partculas cargadas. Como en el caso del electromagnetismo, en esta
secci on podemos considerar los campos sin preocuparnos de las fuentes. En tal caso tendremos una
teora en la cual el campo escalar juega el papel de partcula mediadora de la interaccion.
Si el campo escalar se generaliza para que pueda tener otros n umeros cuanticos, como carga
electrica, entonces estos pueden ser las fuentes de las respectivas cargas y corrientes en la ecuaciones
para campos vectoriales. Esto se estudiara en la seccion ??. En tal caso podramos tener por ejemplo
atomos formados de partculas escalares que se excitan emitiendo fotones.
La ecuaciones de Euler-Lagrange para V () = dan lugar a:
(+ m
2
) = .
_

2
t
2

2
+ m
2
_
= . (1.232)
Con el cuadrivector (1.125) podemos construir la siguiente ecuaci on
p
= m
2
= m
2
= m
2
_

2
t
2

2
+ m
2
_
= 0. (1.233)
Que corresponde a la ecuaci on de Klein-Gordon (??). Una expresi on escrita en terminos de productos
escalares de Lorentz se dice que esta en forma covariante.
De acuerdo a la ec. (1.226), tenemos
L
free
=
1
2

1
2
m
2
2
L
int
= (1.234)
T
=
L
(
L, (1.235)
T
0
0
=
0
0
L
=(x)
0
(x) L
=
0
0

1
2
0

1
2
i
+
1
2
m
2
2
=
1
2
_
0
+
_
+
1
2
m
2
2
1 =
1
2
_
_
t
_
2
+ ()
2
_
+
1
2
m
2
2
. (1.236)
donde
(x) =
L
(/t)
=
t
. (1.237)
La densidad de momentum es
T
0
i
=
L
(
0
)
T
0
i
=
0
T
0
=
t
. (1.238)
1.9.1 Complex scalars
En la secci on anterior se trabajo con un campo escalar real que solo podra describir un pion neutro.
Para describir piones cargados debemos construir un campo escalar complejo. En mec anica cu antica
la funci on de onda compleja puede describir parcialmente a un electr on cargado. Sin embargo la
funci on de onda del electr on tambien debe ser generalizada para poder dar cuenta del espn. Esto
corresponde al funci on de onda de la ecuaci on de Dirac en la secci on ??.
De hecho, algunas consecuencias fsicas interesantes surgen si consideramos un sistema de dos
campos escalares reales,
1
y
2
, que tengan la misma masa m. Entonces
L =
1
2
[
1
2
m
2
2
1
] +
1
2
[
1
2
m
2
2
2
] (1.239)
Si denimos
=
1
+ i
2
2
then (1.240)
1
i
2
2
, and (1.241)
2 =(
1
+ i
2
)
=(
1
i
2
). Therefore
2( +
) =2
1
2(
) =2i
2
. Then
1
=
+
2
(1.242)
2
=

2i
. (1.243)
Reemplazando la ecuaciones (1.242) y (1.243) en la ec. (1.239), tenemos
L =
1
4
[
( +
( +
)
1
2
m
2
( +
)
2
]
+ i
2
1
4
[
)
1
2
m
2
(
)
2
]
=
1
4
[
+ 2
m
2
(
2
+
2
) + 2
1
4
[
m
2
(
2
+
2
) 2
]
=
1
4
[4
4m
2
]
L =
m
2
(1.244)
De la ec. (1.18) de la secci on ??,
De las ecuaciones de Euler-Lagrange para
, usando el Lagrangiano en ec. (1.244)
_
L
(
)
_
= 0
+ m
2
= 0
(+ m
2
) = 0, (1.245)
y de la ecuaciones de Euler-Lagrange para ,
(+ m
2
)
= 0. (1.246)
De este modo tanto , como
, satisfacen la ecuacion de Klein-Gordon. Cada campo adem as

corresponde a una partcula de masa m como en el caso de
1
y
2
Estamos ahora interesado en las simetras internas del Lagrangiano. Entonces la corriente con-
servada puede denida en la secci on ??, eq. (1.20)
J
=
L
(
)
+
L
(
)
J
. (1.247)
Adem as de la invarianza de Lorentz, el Lagrangiano en ec, (1.244) tambien es invariante bajo el
grupo de transformaciones U(1) denido en las seccion ??, pero con una fase constante
U = e
i
1 + i.
Entonces
= e
i
(1 + i)
= + i. (1.248)
Entonces,
= i (1.249)
= i
. (1.250)
Reemplazando en ec. (1.247)
J
i(
), (1.251)
y
= J
0
i(
t
). (1.252)
Denimos J
como
J
= i(
), (1.253)
Como puede ser negativo no puede interpretarse como una probalidad, como se hizo con la funci on
de onda de la ecuacion de Scrodinger. Esto present o un obstaculo en la interpretaci on inicial de la
ecuaci on de Klein-Gordon. Sin embargo una vez se cuantiza el campo escalar la probabilidad de los
estados cuanticos queda bien denida [?].
1.10 Lorentz transformation of the elds
Note again, that a term like
(x)a
(x) , (1.254)
does not left the Action invariant. To have a proper formulation of the quantum mechanics through
the general equation
i

t
=

H , (1.255)
with some, to be determined, relativistic Hamiltonian operator

H, we should be able to build a
Lagrangian with temporal derivatives of order one. Therefore, the Lorentz invariant requires all the
derivatives of order one.
Consider spinor elds, which transforms as
a
(x)
a
(x) = S
ab
()
b
(
1
x) , (1.256)
where S() is some spinorial representation of the Lorentz Group. We will check in next section if
a Action with a term like
a
(x)a
ab
b
(x) (1.257)
could be invariant under Lorentz transformations, for some internal representation of the Lorentz
Group.
In summary we have the following Lorentzs transformation properties for the elds
(x)
(x
) =(x) Scalar eld,

A
(x) A
(
1
x) Vector eld,
(x)
(x) =S()(
1
x) Spinor eld. (1.258)
1.11 Diracs Action
The Scrodinger equation can be written as
i

t
=

H
S
, (1.259)
1.11. DIRACS ACTION 49
where
H
S
= (1.260)
In order to have a well dened probabilty in relativistic quantum mechanics it is necessary that
Lagrangian be linear in the time derivative, in order to obtain the general Sccodinger equation:
i

t
=

H , (1.261)
like the Scr odinger Lagrangian. However, this automatically imply that the Lagrangian will be also
linear in the spacial derivatives. A pure scalar eld cannot involve a Lorentz invariant term of only
rst derivatives (see eq. (1.254)). Therefore the proposed eld must have some internal structure
associated with some representation of the Lorentz Group. Therefore we build the Lagrangian for a
eld of several components
=
_
_
_
_
_
2
.
.
.
n
_
_
_
_
_
(1.262)
1.11.1 Lorentz transformation
If the eld is to describe the electron. it must have spin and in this way it must transform under
some spin representation of the Lorentz Group
(x)
(x) = S()
_
1
x
_
. (1.263)
One possible invariant could be the term
(x)(x). However, under a Lorentz transformation we

should have
S. As we cannot assume that S() is unitary, the solution is to dene the adjoint
spinor
=
b . (1.264)
which transforms as
(x)
(x) =
(x)b =
1
x
_
S
()b , (1.265)
and,
(x)(x)
(x)
(x) =
1
x
_
S
()bS()
_
1
x
_
(1.266)
The condition that must be fullled for Lorentz invariance of the Action is
S
()bS() =b , (1.267)
and therefore,
(x)(x)
(x)
(x) =
_
1
x
_
1
x
_
, (1.268)
and:
(x)
(x) =
1
x
_
bS
1
()
=
_
1
x
_
S
1
() . (1.269)
A Action with a Lagrangian term linear in the derivatives, could be Lorentz invariant if, taking
into account:
(x)
(x)
(x)
(x) =
a
_
1
x
_
S
1
ab
()
bc
_
1
_
S
cd
()
d
_
1
x
_
=
_
1
x
_ _
1
_
_
S
1
()
S()
_
1
x
_
= (x)
(x) , (1.270)
if the following condition is satised:
S
1
()
S() =
. (1.271)
the most general Lagrangian for this eld is
L = i
m , (1.272)
Where the coecients have been already xed by convenience. Since the Action is real, it is convenient
to rewrite this as
L = i
m
=
1
2
_
i
_
+ i
m
=
i
2
(

i
2
+ i
m
=
i
2

i
2
(
m . (1.273)
Para que este nuevo Lagrangiano sea real se requiere que,
b
= b
b
2
= I
b
b =
(1.274)
ya que
L
=
_
i
2

i
2
b
_
m
b
=
_
i
2
b
2

i
2
b
2
b
_
m
b
=
_
i
2

i
2
b
_
m
=
_
i
2

i
2
_
m
1.11.2 Corriente conservada y Lagrangiano de Dirac

De la ec. (??)
J
0
=
_
L
(
0
)
_
+
_
L
_
_
= i
0
(1.275)
El Lagrangiano es invariante bajo transformaciones de fase globales, U(1)

= e
i
i, (1.276)
de modo que
= i. (1.277)
Por consiguiente
J
0
=
0
(1.278)
Para que J
0
pueda interpretarse como una densidad de probabilidad, se debe cumplir
b
0
= I (1.279)
La densidad de corriente es
J
0

. (1.280)
Que podemos interpretar como una densidad de probabilidad.
De la ec. (1.279), ya que la inversa de es unica:
b =
0
. (1.281)
se dene como la adjunta de :
=
0
. (1.282)
It is convenient at this point to summarize the properties for
0
:
=
0
_
0
_
2
=1
0
0
=
()
0
S() =
0
. (1.283)
En general
J
_
L
(
)
_
+

_
L
_
_
i
(i)
i
(i)
=

(1.284)
y
J
. (1.285)
1.11.3 Tensor momento-energa
T
0
0
=
L
(
0
)
0
+
0

_ L
= i
0
L
= i
i
+ m
,
=

( p + m),
=
0
( p + m),
=

H, (1.286)
donde
H =
0
( p + m) (1.287)
la ecuacion de Scrondinger de validez general es entonces:
i

t
=

H (1.288)
y, como en mecanica cl asica usual
H =
_

H d
3
x. (1.289)
Adem as
T
0
i
=
L
(
0
)
i
+
i

_
= i
(i
i
) (1.290)
de modo que
p =
_

p d
3
x (1.291)
1.11.4 Ecuaciones de Euler-Lagrange
Queremos que el Lagrangiano de lugar a la ecuacion de Scrondinger de validez general
i

t
=

H (1.292)
con el Hamiltoniano dado en la ec. (1.289), que corresponde a un Lagrangiano de s olo derivadas de
primer orden y covariante, en lugar del Hamiltoniano para el caso no relativista.
De hecho, aplicando las ecuaciones de Euler-Lagrange para el campo

al Lagrangiano en ec. (??)
,tenemos
_
L
_
_
= 0
L
= 0
i
m = 0. (1.293)
Expandiendo
i
0
0
+ i
i
i
m = 0
i
0
0
(i) m = 0,
i
0
0
= ( p + m),
de donde
i
0
2

t
=
0
( p + m). (1.294)
tenemos que
_
0
_
2
= 1. (1.295)
De la ec. (1.287)
H =
0
( p + m), (1.296)
A este punto, s olo nos queda por determinar los parametros
.
La ec. (1.292) puede escribirse como
_
i

t

H
_
= 0. (1.297)
El campo tambien debe satisfacer la ecuaci on de Klein-Gordon. Podemos derivar dicha ecuaci on
aplicando el operador
_
i

t

H
_
De modo que, teniendo en cuenta que

H/t = 0,
_
i

t

H
__
i

t

H
_
= 0
_
i

t

H
__
i
t

H
_
= 0
t
2
+ i
_

H
t
_
+ i

H
t
i

H
t
+

H
2
= 0
_

2
t
2
+

H
2
_
= 0. (1.298)
De la ec. (1.296), y usando la condici on en ec. (1.295), tenemos
H
2
= (
0
p +
0
m)(
0
p +
0
m)
= (
0
p)(
0
p) + m
0
p
0
+ m
2
0
p + m
2
(1.299)
Sea
=
0
i
=
i
i
=
i
(1.300)
H
2
= ( p)( p) + m p + m p + m
2
= ( p)( p) + m( + ) p + m
2
(1.301)
Sea A una matriz y en un escalar. Entonces tenemos la identidad
(A )
2
=
i
A
i
2
i
2
+
i<j
_
A
i
, A
j
_
j
(1.302)
Demostraci on
[(A )]
ij
A
i
i
A
j
j
=
ij
A
i
A
j
ij
j
A
i
A
j
i
2
A
i
A
i
i<j
j
A
i
A
j
i>j
j
A
i
A
j
_
=
i
2
A
i
A
i
i<j
j
A
i
A
j
j>i
i
A
j
A
i
_
=
i
2
A
i
A
i
i<j
j
_
A
i
A
j
+ A
j
A
i
_
_
=
_
i
2
_
A
i
A
i
_
i<j
j
_
A
i
, A
j
_
_
=
_
i
2
A
i
2
+
i<j
j
_
A
i
, A
j
_
_
. (1.303)
Entonces
H
2
=
2
i
p
2
i
+
i<j
i
,
j
p
i
p
j
+ m(
i
+
i
)p
i
+ m
2
(1.304)
(suma sobre ndices repetidos). Si
2
i
= 1
i
,
j
= 0 i ,= j
i
+
i
= 0 (1.305)
H
2
=
2
+ m
2
(1.306)
y reemplazando en la ec. (1.298) llegamos a la ecuaci on de Klein-Gordon para
_

2
t
2

2
+ m
2
_
= 0
_
+ m
2
_
= 0 (1.307)
En terminos de las matrices
las condiciones en ec. (1.305) son

_
0
_
2
= 1
_
i
_
2
= 1
0
i
=
_
i
_
2
= 1
_
i
_
2
= 1
0
+
0
i
=
_
i
,
0
_
= 0 (1.308)
De modo que
_
i
,
j
_
=
0
j
+
0
i
= 0 i ,= j
i
= 0 i ,= j
j
+
j
i
= 0 i ,= j
_
i
,
j
_
= 0 i ,= j (1.309)
Las ecuaciones (1.308)(1.309) pueden escribirse como
= 2g
1 (1.310)
donde
= (
0
,
i
) (1.311)
Adem as, de la ec. (1.283)
0
=
. (1.312)
Cualquier conjunto de matrices que satisfagan el algebra en ec. (1.310) y la condici on en ec. (1.312),
se conocen como matrices de Dirac. A se le llama espinor de Dirac.
En terminos de la matrices
, el Lagrangiano de Dirac y la ecuaci on de Dirac, son respectivamente

de las ecs. (??) y (??)
L =

(i
m) , (1.313)
i
m = 0, (1.314)
donde
0
. (1.315)
1.11.5 Propiedades de las matrices de Dirac
De la ec. (1.312)
=
0
=
0
= 0
=
0
2
i
=
i
= i
. (1.316)
Deniendo
5
= i
0
3
, (1.317)
entonces,
2
5
=
0
2
5
= +
2
0
2
5
= +
1
2
5
=
2
2
5
=
2
2
5
=1. (1.318)
2
5
= 1, (1.319)
Teniendo en cuenta que
2
1 y conmuta con las dem as matrices, tenemos por ejemplo
3
=i
0
2
3
=
2
3
i
0
2
=
3
i
0
3
=
3
2
=i
0
2
2
3
=
2
2
i
0
3
=
2
i
0
3
=
2
1
=i
0
2
1
3
=
2
1
i
0
3
=
1
i
0
3
=
1
0
=i
0
0
=
2
0
i
1
3
=
0
5
. (1.320)
De modo que
,
5
= 0. (1.321)
Expandiendo el anticonmutador tenemos
5
=
5
5
=
Tr (
5
5
) = Tr
Tr (
5
) = Tr
Tr
= Tr
, (1.322)
y por consiguiente
Tr
= 0. (1.323)
De otro lado, si

, (1.324)
para alguna matriz unitaria U, entonces
corresponde a otra representacion de algebra de Dirac

en ec. (1.310), ya que

=
_
U
, U
_
= U
= 2g
UU
= 2g
1. (1.325)
Claramente, la condicion en ec. (1.312) se mantiene para la nueva representaci on. Como
0
es
hermtica, siempre es posible escoger una representaci on tal que
0
U
0
U
sea diagonal. Como
2
0
= 1, sus entradas en la diagonal deben ser 1, y como Tr
0
= 0, debe existir igual n umero de
+1 que de 1. Por lo tanto la dimension de
0
(y de
) debe ser par: 2, 4, . . .. Para un fermion sin

masa
L = i
0
+ i
i
= i
0
+ i
i
, (1.326)
solo se requieren tres matrices 2 2 que satisfacen
_
i
,
j
_
= 2
ij
, (1.327)
y por lo tanto pueden identicarse con las tres matrices de Pauli. Como en general tenemos 4 matrices
independientes, su dimension mnima debe ser 4.
Como
i
=
0
0
=
i
=
i
, podemos denir la representacion de paridad

0
=
0
,
i
=
i
, para U =
0
(1.328)
1.11.6 Lorentz Group
We must build a representation of the Lorentz Group in the Dirac space of n dimensions. First, let
us consider a simpler group, corresponding to the rotation group in tree dimensions. The generators
are the angular momentum operators J
i
, which satisfy the commutation relations
_
J
i
, J
j
= i
ijk
J
k
. (1.329)
The Pauli matrices are set of matrices satisfying this commutation relations:
_
i
2
,

j
2
_
= i
ijk
k
2
(1.330)
donde
i
1
=
_
0 1
1 0
_

2
=
_
0 i
i 0
_

3
=
_
1 0
0 1
_
(1.331)
dividas por dos, corresponden a los generadores del Grupo. Las constantes de estructura del Grupo
corresponden a
ijk
. Como los generadores no conmutan, SU(2) es un Grupo de Lie no Abeliano.
Deniendo los generadores de SU(2) como
T
i
=

i
2
, (1.332)
un elemento del Grupo puede escribirse como
U = e
iT
i
i
1 + iT
i
i
= 1 + i
i
2

i
. (1.333)
Como antes,
i
es el par ametro de la transformaci on.
Las matrices de Pauli y por consiguiente T
i
satisfacen
i
=
i
Tr (
i
) = 0 (1.334)
Adem as
det (
i
) = 1
i
,
j
= 2
ij
I
2
i
= I
Tr
_
j
_
= 2
ij
j
= i
ijk
k
+
ij
(1.335)
In [8]:
It is generally true that one can nd matrix representations of a continuous group by
nding matrix representations of the generators of the group (which must satisfy the
proper commutation relations), then exponentiating these innitesimal representations.
For our present problem, we need to know the commutation relations of the generators of
the group of Lorentz transformations. For the rotation group, one can work the commu-
tation relations by writing the generators as dierential operators; from the expression
J = x p = x (i) , (1.336)
the angular momentum commutation relations (1.329) follow straightforwardly.
The last equation can be written as (summation of repeated indices)
J
k
= [x (i)]
k
=i
ijk
x
i
j
= i
ijk
x
i
j
(1.337)
J
lm

lmk
J
k
=i
lmk
ijk
x
i
j
=i(
li
mj
lj
mi
)x
i
j
=i(x
l
m
x
m
l
) . (1.338)
Involving three generators. The generalization to four-dimensions give to arise three further genera-
tors J
0i
:
J
= i(x
) . (1.339)
The six generators J
satisfy the algebra

[J
, J
] =i(g
+ g
) . (1.340)
From [8]:
Any matrices that are to represent this algebra must obey these same commutation rules.
The exponentiation of the generators give to arise to group elements
= exp
_
i
2
_
(1.341)
To nd a representation of the usual boosts and rotations, consider a boost
x
=
_
_
_
_
t
x
y
z
_
_
_
_
_
_
_
_
t
_
_
_
_
=
_
_
_
_
t+vx
1v
2
x+vt
1v
2
y
z
_
_
_
_
=
_
_
_
_
cosh sinh 0 0
sinh cosh 0 0
0 0 1 0
0 0 0 1
_
_
_
_
_
_
_
_
t
x
y
z
_
_
_
_
=
, (1.342)
Since
cosh =
n=0
2n
2n!
1 +O(
2
)
sinh =
n=0
2n+1
(2n + 1)!
+O(
2
) , (1.343)
one innitesimal boost along x is
xboost

_
_
_
_
1 0 0
1 0 0
0 0 1 0
0 0 0 1
_
_
_
_
. (1.344)
Similarly a rotation by an innitesimal angle =
3
along xyplane (or about the zaxis)
xyrotation

_
_
_
_
1 0 0 0
0 1 0
0 1 0
0 0 0 1
_
_
_
_
. (1.345)
In general we dene the six independent LorentzGroup parameters:
0i
=
i0

i
12
=
21

3

32
=
23

2

13
=
31

1
. (1.346)
The 4 4 matrices
(J
= i (
) , (1.347)
where and label which of the six matrices we want, while and label components of the
matrices. These matrices satisfy the commutations relations (1.340), and generate the three boosts
and three rotations of the ordinary Lorentz 4-vectors:
i
2
(J
(1.348)
= 1 +
i
b
i
+
1
2
ijk
r
jk
, (1.349)
b
i
=iJ
i0
r
jk
= iJ
jk
. (1.350)
1.11.7 Lorentz invariance of the Dirac Action
We need to satisfy the following conditions
S
1
()
S() =
()
0
S() =
0
or S
()
0
=
0
S
1
() . (1.351)
In order to nd a representation of the Lorentz Group in terms of the Dirac matrices we propose
S() = 1 +
i
B
i
+
1
2
ijk
R
jk
. (1.352)
Instead of show the Lorentz invariance of the Dirac Action, we use the conditions derived from the
invariance, to nd a representation in terms of the Dirac matrices for B
i
and R
jk
. As a consistency
check, the resulting representation would satisfy the Lorentz algebra. In this way, by using eq. (1.349)
and (1.352), we obtain from
S
1
()
S() =
, (1.353)
that
B
i
=
1
2
i
R
jk
=
1
2
k
, (1.354)
which can be written in covariant form if we dene
o
=
i
4
[
] . (1.355)
In fact, the six set of non-zero independently generators are
o
0i
=
i
4
_
0
_
=
i
2
i
= iB
i
o
ij
=
i
4
_
i
_
=
i
2
j
= iR
ij
. (1.356)
It is worth notices that in fact o
satisfy the Lorentz algebra, and therefore are the generators of

the Lorentz group elements:
S() =exp
_
i
2
_
1
i
2
. (1.357)
Another consistency check is
S
()
0
S() =
0
, (1.358)
or equivalently
S
()
0
=
0
S
1
()
_
1 +
i
2
0
=
0
_
1 +
i
2
_
o
0
=
0
o
. (1.359)
Taking into account that
0
=
_
0
_
2
0
_
2
0
=
0
, (1.360)
we have
o
0
=
i
4
[
0
=
i
4
_
0
=
i
4
_
0
=
i
4
[
]
0
=
0
o
(1.361)
1.11.8 Diracs Lagrangian
Para una matriz de n dimensiones existen n
2
matrices hermticas (o antihermticas) independientes.
Si se sustrae la identidad quedan n
2
1 matrices hermticas (o antihermticas) independientes de
traza nula. En el caso n = 2 corresponden a las 3 matrices de Pauli. En el caso de la ecuaci on de
Dirac se requieren 4 matrices independientes, por lo tanto deben ser matrices 4 4. En efecto para
n = 4 existen 15 matrices independientes de traza nula dentro de las cuales podemos acomodar sin
problemas las 4
.
De [15]:
All Dirac matrix elements will now be written in the form
(x)(x) , (1.362)
where is a 44 complex matrix. The most general such matrix can always be expanded
in terms of 16 independent 4 4 matrices multiplied by complex coecients. In short
Matriz Transformacion N umero Escalar en Dirac
1 Escalar (S) 1

5
Pseudoescalar (P) 1

Vector (V) 4

5
Vector axial (A) 4

=
i
2
[
] Tensor antisimetrico (T) 6

16
Table 1.1: Matrices
i
.
the matrices can be regarded as a 16dimensional complex vector space spanned by 16
matrices.
It is convenient to choose the 16 matrices,
i
, so that they have well dened transformation
properties under the Lorentz Transformations. Since the
s have such properties, we

are lead to choose the following 16 matrices for this basis:
En la Tabla 1.1 se muestran las matrices de traza nula con sus propiedades de transformacion
bajo el Grupo de Lorentz. En la ultima se muestra el correspondiente escalar en el espacio de Dirac
. Demostraci on
J
(x)

(x)
(x)
(
1
x)S
1
()
S()(
1
x)
=

(
1
x)
(
1
x)
=
(
1
x) . (1.363)
In [15]: Problem 5.4:
5
S
1
()
5
S() = (det )
5
(1.364)
The solution is in Appendix C. of Burgess book, by using
5
=
i
24
(1.365)
and
det =
4
. (1.366)
1.12 Electrodinamica Cuantica
Para hacer el Lagrangiano en ec. (1.313) invariante gauge local bajo U(1)
Q
, procedemos de la forma
usual. El campo transforma como

= e
i(x)Q
=

e
i(x)Q
, (1.367)
donde Q es el generador de carga electrica en unidades de la carga del electr on.
La derivada covariante se dene de manera que transforma de la misma forma que el campo,
introduciendo el campo gauge A
ieQA
, (1.368)
donde e es la carga electrica del electr on. De esta forma, si
e
es el campo que representa al electron
eQ
e
= e(1)
e
= e
e
. (1.369)
El Lagrangiano correspondiente a la interacci on de un fermi on y el campo electromagnetico cor-
responde al Lagrangiano de Dirac con la derivada normal reemplzada por la derivada covariante, y el
correspondiente termino cinetico invariante gauge y de Lorentz asociado al nuevo campo introducido
en la derivada covariante: A
. Este campo es necesario para compensar los cambios en la energa y

momentum que sufre el electron como consecuencia de imponer la invarianza de la Acci on bajo un
cambio de fase local
L = (i
m)
1
4
F
, (1.370)
y es invariante bajo transformaciones locales U(1)
Q
. Desarrollando la expresi on anterior, tenemos
L =

[i
ieQA
) m]
1
4
F
=

(i
m) + eQ
1
4
F
. (1.371)
Este Lagrangiano da lugar a la Acci on de la teora conocida como Electrodinamica Cu antica (QED
de sus siglas en ingles).
Aplicando las ecuaciones de Euler-Lagrange para

, tenemos
(i
m) + eQ
= 0
(i
i
2
eQ
m) = 0
[i
ieQA
) m] = 0
(i
m) = 0. (1.372)
1.12. ELECTRODIN
AMICA CU
ANTICA 67
Que corresponde a la ecuacion de Dirac en presencia del campo electromagnetico. Mientras que para
el campo A
, tenemos
1
4
_
F
)
_
eQ
= 0
= eQ
(1.373)
Denimos entonces la corriente electromagnetica generada por el fermion como
j
= eQ
. (1.374)
De nuevo, la aparici on de la interaccion electromagnetica es una consecuencia de la invarianza gauge
local.
El calculo directo de la corriente
J
=
L
(
)
+

_
= i
(i(x)Q) = (x)Q
, (1.375)
y para la ecuaci on de Dirac, a diferencia de la ecuacion de Schr odinger, la corriente de probabilidad
tiene la misma forma que la corriente electromagnetica.
De esta manera podemos reescribir el Lagrangiano en terminos de un Lagrangiano libre y otro
de interacci on
L = L
free
+L
int
, (1.376)
L
free
=
(i
m)
1
4
F
L
int
=eQ
. (1.377)
Para la QED s olo hay un termino de interacci on que es suciente para explicar todos los fenomenos
electromagneticos y su interacci on con la materia. Este esta representado por el diagrama de Feynman
mostrado en la Figura 1.2
La repulsi on electromagnetica esta representada por la gura 1.3. En la Figura (a) el primer
electr on emite un fot on y se dispersa, mientras que el segundo absorbe el foton y se dispersa en
la direcci on opuesta. En la Figura (b) el primer elect on absorve el fot on emitido por el segundo
electr on. Los dos diagrams se representa por uno unico con el fot on en horizontal como se muestra
en la Figura (c).
Figure 1.2: Feynman rule for QED
Figure 1.3: Electromagnetic repulsion. The diagrams (a) and (b) are summarized in the diagram (c)
1.13. CROMODIN
AMICA CU
ANTICA 69
1.13 Cromodinamica Cuantica
Los protones, neutrones, piones, kaones y dem as hadrones, son partculas compuestas de consti-
tuyentes elementales llamados quarks. Por ejemplo los protones, neutrones y piones estan constitu-
idos de quarks up y down. Los hadrones est an dividos en bariones, B, constituidos de tres quarks,
y los mesones, M, de dos. Para satisfacer el principio de exclusion de Pauli, y justicar el con-
namiento de los hadrones, se requiere que cada quark contenga N
c
cargas diferentes, llamadas cargas
de color, de manera que la carga de color de un hadron sea cero. Muchos resultados experimentales
respaldan la existencia de tres cargas de color para cada quark, N
c
= 3. De este modo cada quark
q = u, d, c, s, t, b viene en tres colores
q
= q
1
, q
2
, q
3
= q
r
, q
b
, q
g
, (1.378)
donde los ultimos subndices hacen referencia a los colores red, blue, green. De este modo los Bariones
y mesones est an descritos por combinaciones singletes de color del tipo q
r
q
b
q
g
y q
r
q
r
,
B =
1
[q
M =
1
[ q
(1.379)
Estos estados son singletes de color. Una de las determinaciones de N
c
proviene del observable
R
(e
+
e
hadrones)
(e
+
e
)
(1.380)
Para f = u, d, s, c, b, t, (en orden de masa) tenemos que para una energa donde se pueden producir
hadrones compuestos de hasta quarks f
max
R
fmax
f=u
Nc
=1
(e
+
e
)
(e
+
e
)
R N
c
fmax
f=u
(e
+
e
f

f)
(e
+
e
)
(1.381)
De este modo R esta dado por la suma de las cargas electricas al cuadrado
R N
c
f
Q
2
f
Q
2
=N
c
fmax
f=u
Q
2
f
=
_
_
N
c
[(
2
3
)
2
+ 2(
1
3
)
2
] =
2
3
N
c
f = u, d, s, f
max
= s
N
c
[2(
2
3
)
2
+ 2(
1
3
)
2
] =
10
9
N
c
f
max
= c
N
c
[2(
2
3
)
2
+ 3(
1
3
)
2
] =
11
9
N
c
f
max
= b
=
_
_
2 N
c
= 3, f
max
= s
10
3
N
c
= 3, f
max
= c
11
3
N
c
= 3, f
max
= b
(1.382)
En la gura, tomada de [?], se muestra el graco de R con respecto a
s (la energa de centro de masa

de la colisi on). Se observan dos escalones, uno que va hasta una energa
s 4 GeV que corresponden

a f = u, d, s, con un R 2, y otro hasta

s 40 GeV que corresponde a f = u, d, s, c, b, con un
R 3.7 11/3. Los dos valores de R son compatibles con los esperados de la ec. (1.382). Como
referencia tambien se se nalan los valores para N
c
= 4 (en rojo).
Si queremos que el color sea una carga conservada como la carga electrica, esta debe ser la
consecuencia de una simetra gauge local. Para tener tres cargas diferentes la posibilidad m as simple
es imponer la simetra SU(3)
c
, tal que tengamos un vector compuesto de 3 espinores de Dirac en el
espacio de color:
=
_
_
g
_
_
=
_
_
q
r
q
b
q
g
_
_
. (1.383)
El Lagrangiano de Dirac con invarianza gauge global SU(3), para un quark, se puede escribir como
L
global
= i
, (1.384)
donde

= exp
_
i
a
a
2
_
. (1.385)
a = 1, . . . , 8,
a
/2 son los ocho generadores de SU(3) y
a
son los parametros de la transformaci on
global. Los generadores de SU(3)

a
2
, (1.386)
1.13. CROMODIN
AMICA CU
ANTICA 71
Figure 1.4: Datos para R
satisfacen el algebra
_
a
2
,

b
2
_
= if
abc
c
2
, (1.387)
donde f
abc
son las constantes de estructura na de SU(3).
En un analisis similar al de la secci on 1.11.8 tenemos que la Acci on invariante gauge local bajo
SU(3)
c
, se obtiene de reemplazar la derivada normal por la derivada covariante
L
local
= i
1
2
Tr (G
) , (1.388)
donde

= U(x)
T
(T
= U(x)T
, (1.389)
con la matriz 3 3
U(x) = exp
_
i
a
(x)
a
2
_
, (1.390)
y
T
ig
s
a
2
G
a
ig
s
G
(1.391)
donde hemos denido la matriz 3 3 G
, como
(G
=
_
a
2
_
G
a
(1.392)
Este Lagrangiano da lugar a la interacci on fuerte y es llamado el Lagrangiano de la Cromodin amica
Cu antica, o el Lagrangiano de la QCD de sus siglas en Ingles.
De (1.389), tenemos
T
(T
=T
= U(x)T
U = U(x)T
. (1.393)
Por consiguiente
T
U = UT
(1.394)
T
(T
= UT
U
1
(1.395)
T
(T
= (
ig
s
G
)U =U(
ig
s
G
T)
U
+ (
U) ig
s
G
U =U
ig
s
G
U) ig
s
G
U =ig
s
UG
ig
s
G
U =(
U) ig
s
UG
, (1.396)
de modo que
G
U =
1
ig
s
(
U) + UG
=
i
g
s
(
U)U
1
+ UG
U
1
. (1.397)
Como U es unitaria, la transformaci on de los campos gauge puede escribirse como
G
(G
= UG
U
1
i
g
s
(
U) U
. (1.398)
1.13. CROMODIN
AMICA CU
ANTICA 73
Entonces
a
G
a
(1 + i
b
b
)
c
G
c
(1 i
d
d
)
i
g
s
[i(
e
)
e
(1 i
f
f
)]
=(
c
+ i
b
c
)(1 i
d
d
)G
c

i
g
s
[i(
e
)
e
(1 i
f
f
)]
[
c
i
d
d
+ i
b
c
]G
c
+
1
g
s
e
=[
c
i
b
(
c
c
)]G
c
+
1
g
s
e
=
a
G
a
i(if
acb
a
)G
b
+
1
g
s
a
=
a
_
G
a
+
1
g
s
a
+ f
acb
G
b
_
(1.399)
de donde
G
a
G
a
G
a
+
1
g
s
a
+ f
abc
G
c
(1.400)
que se reduce al caso Abeliano cuando las constates de estructura son cero. Como era de esperarse
cada campo gauge tiene asociado un par ametro de transformaci on gauge
a
(x).
Similarmente, deniendo la matriz 3 3,
G
=
i
g
s
[T
, T
]

a
2
G
a
, (1.401)
tenemos
G
=
i
g
s
[
ig
s
G
ig
s
G
]
=
i
g
s
[(
ig
s
G
) (
ig
s
G
) (
ig
s
G
) (
ig
s
G
) ]
=
i
g
s
_
g
2
s
G
ig
s
[
(G
) + G
+ g
2
s
G
+ ig
s
[
(G
) + G
]
_
=
i
g
s
(
) g
2
s
(G
) ig
s
[(
) (
)]
ig
s
[G
+ G
+ G
]
=[
ig
s
(G
)]
=
ig
s
[G
, G
] (1.402)
De modo que
G
ig
s
[G
, G
] , (1.403)
que se reduce al caso Abeliano cuando los bosones gauge conmutan. En terminos de componentes
a
G
a
=
a
a
ig
s
[
b
G
b
,
c
G
c
]
=
a
a
ig
s
[
b
,
c
]G
b
G
c
=
a
a
ig
s
(i
a
f
abc
)G
b
G
c
=
a
a
+
a
g
s
f
abc
G
b
G
c
. (1.404)
Por consiguiente
G
a
=
a
+ g
s
f
abc
G
b
G
c

G
a
+ g
s
f
abc
G
b
G
c
, (1.405)
con
a
=
a
(1.406)
A diferencia del caso Abeliano G
ya no es invariante bajo transformaciones gauge

G
=
i
g
s
_
T
, T
=
i
g
s
_
UT
U
1
, UT
U
1
= UG
U
1
. (1.407)
Note que con la denicion (1.401), la derivada covariante de la matrix G
, transforma como la
matrix G
(T
= UT
U
1
. (1.408)
Para poder obtener un invariante bajo transformaciones gauge a partir del producto G
,
debemos utilizar la traza
Tr (G
) Tr
_
G
_
=Tr
_
UG
U
1
UG
U
1
_
=Tr
_
UG
U
1
_
=Tr
_
U
1
UG
_
=Tr (G
) . (1.409)
1.13. CROMODIN
AMICA CU
ANTICA 75
Teniendo en cuenta la normalizaci on de las matrices de Gell-Man
Tr
_
b
_
=2
ab
Tr
_
b
_
=
1
2
ab
, (1.410)
tenemos (suma sobre indices repetidos de SU(3))
Tr (G
) Tr
_
G
_
=Tr
_
a
G
a

b
G
b
_
=Tr
_
b
_
G
a
G
b
=
1
2
ab
G
a
G
b
=
1
2
G
a
G
a
. (1.411)
Expandiendo el Lagrangiano en ec. (1.388), tenemos
L =i
ig
s
a
2
G
a
_
m
1
2
Tr (G
)
=i
ig
s
a
2
G
a
_
m
1
4
G
a
G
a
=i
+ g
s
a
2
G
a
1
4
G
a
G
a
=i
+ g
s
a
2
G
a
1
4
G
a
1
4
_
g
s
a
f
ade
G
d
G
e
+ g
s
f
abc
G
b
G
G
a
+ g
2
s
f
abc
f
ade
G
b
G
c
G
d
G
e
_
=L
free
+L
gauge
+L
SI
, (1.412)
donde
L
free
=i
L
gauge
=g
s
a
2
G
a
1
4
G
a
L
SI
=
1
4
_
g
s
a
f
ade
G
d
G
e
+ g
s
f
abc
G
b
G
G
a
+ g
2
s
f
abc
f
ade
G
b
G
c
G
d
G
e
_
. (1.413)
Hemos divido el Lagrangiano en tres partes
El Lagrangiano libre de Dirac
Una parte gauge que puede escribirse como un Lagrangiano electromagnetico:
L
gauge
=
1
4
(
a
)
_
G
a
G
a
_
J
a
G
a
, (1.414)
dende
J
a
= g
s
a
2
, (1.415)
es la nueva corriente conservada de interaccion fuerte que surge como consecuencia de la invar-
ianza gauge local SU(3); y
Una parte de auto-interacciones gauge:
L
SI
=
g
s
2
f
abc
G
a
b
G
c
+
g
2
s
4
f
abc
f
ade
G
b
G
c
G
d
G
e
=
g
s
2
f
abc
(
a
) G
b
G
c
g
2
s
4
f
abc
f
ade
G
b
G
c
G
d
G
e
. (1.416)
que se desaparecen en el caso Abeliano.
El Lagrangiano de interacci on es:
L
int
= g
s
a
2
G
a
g
s
2
f
abc
(
a
) G
b
G
c
g
2
s
4
f
abc
f
ade
G
b
G
c
G
d
G
e
. (1.417)
From [17] (pag 136):
The quarks have an additional type of polarization that is not related to geometry. The
idiot physicists, unable to come up with any wonderful Greek words anymore, call this
type of polarization by the unfortunate name of color, which has nothing to do with
color in the nornal sense. At a particular time, a quark can be in one of three conditions,
or colorsR, G, or B (can you guess what they stand for?). A quarks color can be
changed when the quark emits or absorbs a gluon. The gluons come in eigth dient types,
according to the colors they can couple with. For example, if a red quark changes to
green, it emits a red-antigreen gluona gluon that takes the red from quark and gives
it green (antigreen means the gluon is carrying green in the opposite direction). This
gluon could be absorved by a green quark, which changes to red (see Fig. 1.5). There are
eigth dierent possible gluons, such as red-antired, red-antiblue, red-antigreen, and so on
1.13. CROMODIN
AMICA CU
ANTICA 77
Figure 1.5: Quarkgluon interaction
(youd think thered be nine, but for technical reasons, onw is missing)
2
. The theory is
not very complicated. The complete rule of gluons is: gluons couple with things having
colorit just requires a little bookkeeping to keep track of where the colors go. There
is, however, an interesesting possibility created by this rule: gluons can couple with other
gluons (see Fig. 1.6).
El primer termino da lugar a interacciones de cambio de color de quarks como la que se ilustra
en la Figura 1.5
Mientras que el segundo y tercer termino dan lugar a autointeracciones de los gluones como se
muestra en la Figura 1.6
Todas las interacciones est an determinadas en terminos de una unica constante de acoplamiento
g
s
. Las autointeracciones gauge pueden explicar aspectos de la interaccion fuerte como la libertada
asint otica, que consiste en que las interacciones fuertes se vuelven m as debiles a distancias cortas.
En terminos de ndices de color la corriente, y las otras partes del Lagrangiano, pueden escribirse
2
_
_
r r r
b r g
b r b
b b g
g r g
b g g
_
_
, with r r + b
b + g g = 0
Figure 1.6: Triplegluon selfinteraction. The anticolors are the colors running back in time.
como
J
a
= g
s
q
a
2
_
. (1.418)
Note que tanto para la Electrodin amica Cu antica como para la Cromodin amica Cu antica la corriente
es vectorial. Para las interacciones debiles la estructura es m as complicada y requiere un

conocimiento m as profundo de la ecuaci on de Dirac y sus soluciones.
1.13.1 Ecuaciones de EulerLagrange
Sigiendo los mismos procedimientos anteriores debemos llegar a los siguientes resultados. Para el
campo
(i
m) = 0 , (1.419)
1.13. CROMODIN
AMICA CU
ANTICA 79
_
L
(
G
a
)
_
L
G
a
a

1
2
g
s
f
dbc
G
b
G
G
a
)
_
G
d
G
d
_
_
g
s
a
2

+
g
s
2
f
dbc
d
(
ba
G
c
+ G
b
ca
) +
g
s
4
f
ibc
f
ide
(g
G
b
G
c
G
d
G
e
)
=
a

1
2
g
s
f
dbc
G
b
G
c
(
da
da
)
_
g
s
a
2

+
g
s
2
f
dac
d
G
c
+
g
s
2
f
dba
d
G
b
+
g
s
4
f
ibc
f
ide
g
ba
G
c
G
d
G
e
+ G
b
ca
G
d
G
e
+ G
b
G
c
da
G
e
+ G
b
G
c
G
d
ea
)
=
a

1
2
g
s
f
abc
G
b
G
c
+
1
2
g
s
f
abc
G
b
G
c
_
g
s
a
2

g
s
2
f
adc
d
G
c
g
s
2
f
adb
d
G
b
+
g
s
4
f
iac
f
ide
g
G
c
G
d
G
e
+
g
s
4
f
iba
f
ide
g
G
b
G
d
G
e
+
g
s
4
f
ibc
f
iae
g
G
b
G
c
G
e
+
g
s
4
f
ibc
f
ida
g
G
b
G
c
G
d
. (1.420)
Desarrollando los cuatro ultimos terminos, tenemos
g
s
4
f
iac
f
ide
g
G
c
G
d
G
e
+
g
s
4
f
iba
f
ide
g
G
b
G
d
G
e
+
g
s
4
f
ibc
f
iae
g
G
b
G
c
G
e
+
g
s
4
f
ibc
f
ida
g
G
b
G
c
G
d
=
g
s
4
f
iac
f
ide
G
d
G
c
G
e
+
g
s
4
f
iba
f
ide
G
e
G
b
G
d
+
g
s
4
f
ibc
f
iae
G
b
G
c
G
e
+
g
s
4
f
ibc
f
ida
G
c
G
b
G
d
=
g
s
4
f
dac
f
dje
G
j
G
c
G
e
+
g
s
4
f
dca
f
dje
G
e
G
c
G
j
+
g
s
4
f
dbc
f
dae
G
b
G
c
G
e
+
g
s
4
f
dbc
f
dea
G
c
G
b
G
e
=
g
s
4
f
dac
f
dje
G
j
G
c
G
e
+
g
s
4
f
dca
f
dje
G
e
G
c
G
j
+
g
s
4
f
dac
f
dje
G
j
Gc
e
G
c
+
g
s
4
f
dca
f
dje
G
e
G
j
G
c
=
g
s
4
f
abc
G
b
f
cej
G
e
G
j

g
s
4
f
abc
G
b
f
cej
G
e
G
j

g
s
4
f
abc
G
b
f
cej
G
e
G
j

g
s
4
f
abc
G
b
f
cej
G
e
G
j
=g
s
f
abc
G
b
f
cej
G
e
G
j
(1.421)
Entonces
_
L
(
G
a
)
_
L
G
a
a
g
s
f
abc
G
b
G
c
_
g
s
a
2
g
s
f
acd
G
c
d
g
s
f
acd
G
c
f
dej
G
e
G
j
=
a
g
s
f
acd
G
c
d
g
s
a
2
= 0 . (1.422)
Entonces las Ecuaciones de Euler Lagrange para G
a
, son
a
+ g
s
f
acd
G
c
d
= g
s
a
2
. (1.423)
Deniendo
J
a
= g
s
a
2
, (1.424)
La ec.(1.423) puede reescribirse como:
a
= g
s
_
f
abc
G
b
c
+

a
2

_
(1.425)
y usando el hecho que
a
=
+ g
s
(f
abc
G
b
G
c
)
=0 +
1
2
[g
s
(f
abc
G
b
G
b
) + g
s
(f
abc
G
b
G
c
)]
=
1
2
[g
s
(f
abc
G
b
G
b
) + g
s
(f
abc
G
b
G
c
)]
=
1
2
[g
s
(f
abc
G
b
G
b
) + g
s
(f
acb
G
c
G
b
)]
=
1
2
[g
s
(f
abc
G
b
G
b
) g
s
(f
abc
G
b
G
c
)]
=0 , (1.426)
como en el caso Abeliano, tenemos la corriente conservada
= 0 , (1.427)
donde
j
a
=g
s
_
f
abc
G
b
c
+

a
2

_
. (1.428)
El primer termino corresponde a las autointeracciones y el segundo a la corriente de color generada
por los quarks.
1.13. CROMODIN
AMICA CU
ANTICA 81
1.13.2 Derivada covariante adjunta
Toda el algebra de SU(3) se puede escribir en notaci on vectorial en terminos de vectores de 8 com-
ponentes asociados al espacio de los generadores de SU(3). Este nos permitira entender como las
autointeracciones gauge emergen tambien de la derivada covariante cuando se escribe en la repre-
sentaci on adjunta de SU(3).
Deniendo el producto vectorial de SU(3) como
(AB)
a
= f
abc
A
b
B
c
, (1.429)
si G
es un vector en el espacio SU(3) con las 8 componentes G
a
, entonces podemos escribir (1.400)
como
G
= G
+
1
g
s
+G
. (1.430)
Podemos escribir tambien la ec. (1.405) en terminos de vectores en el espacio SU(3) como:
G
+ g
s
G
, (1.431)
donde G
es el vector en el espacio SU(3) con las 8 componentes G
a
.
De igual forma, podemos escribir (1.423) en forma vectorial como
+ g
s
G
= J
(1.432)
y la corriente conservada como
j
=g
s
_
G
G
b
_
. (1.433)
Como G
es una matrix 8 8, su derivada covariante debe estar en la representaci on adjunta de

SU(3)
(
a
)
bc
= i (f
a
)
bc
, (1.434)
con
_
a
,
b
= if
abc
c
. (1.435)
En esta representacion la derivada covariante queda
T
ig
s
a
G
a
(1.436)
En componentes
(T
)
ab
=
ab
ig
s
(
c
)
ab
G
c
=
ab
g
s
f
cab
G
c
=
ab
+ g
s
f
acb
G
c
. (1.437)
Aplicada sobre la componente G
b
, queda
(T
)
ab
G
b
=
ab
b
+ g
s
f
acb
G
c
b
(T
)
ab
G
b
=
a
+ g
s
(G
)
a
(T
)
a
=
a
+ g
s
(G
)
a
, (1.438)
podemos escribir la derivada covariante de G
= (G
1
, , G
8
) como
T
+ g
s
G
. (1.439)
Entonces las las Ecuaciones de Euler Lagrange para G
a
, en (1.423) se pueden escribir como

T
=J
, (1.440)
donde el vector en espacio SU(3) J
, tiene por componentes

J
a
= g
s
a
2
. (1.441)
Para escribir el Lagrangiano en forma vectorial en el espacio SU(3), debemos reescribir la trans-
formaci on gauge de G
en terminos de vectores de SU(3). Como

G
=UG
=(1 + i
b
b
)
c
G
c
(1 i
b
b
) (1.442)
podemos realizar los mismos pasos que en (1.399). El resultado es
G
a
G
a
+ f
abc
G
b
c
. (1.443)
Note que en el caso Abeliano f
abc
= 0, el tensor correspondiente es invariante gauge, como ocurre el
caso electromagnetico. En notaci on de vectores de SU(3):
G
+G
. (1.444)
1.13. CROMODIN
AMICA CU
ANTICA 83
Utilizando la propiedad cclica del triple producto escalar
A (BC) = B (CA) = C (AB) , (1.445)
podemos construir el invariante
G
a
G
a
= G
+G
(G
) + (G
) G
=G
+G
( G
) + (G
) G
=G
(G
) G
+ (G
) G
=G
. (1.446)
El Lagrangiano de la QCD escrito en forma de vectores de SU(3) es
L = i
ig
s
2
G
_
m
1
4
G
. (1.447)
El Lagrangiano para los campos gauge, el cual puede generalizarse para cualquier teora SU(N), es
L
gluon
= L
gauge
+L
SI
=
1
4
G
, (1.448)
Da lugar la ecuaciones de Maxwell pero con la derivada normal reemplzada por la derivada covariante
T
=J
, (1.449)
donde
T
+ g
s
G
. (1.450)
Note que en el caso Abeliano, U(1), la derivada covariante del tensor de campo se reduce a la derivada
normal de dicho tensor. El termino extra en la derivada covariante da lugar a las autointeracciones
de los campos gauge.
Ejercicio:
Muestre que la derivada covariante de G
, transforma como G
.
Figure 1.7: V () =
1
2
2
con
2
> 0
1.14 Spontaneous symmetry breaking
Escribamos el Lagrangiano para una partcula escalar real de masa m como
L =
1
2
V () (1.451)
con
V () =
1
2
2
. (1.452)
Este Lagrangiano es simetrico bajo la transformacion discreta .
Cuando
2
> 0, el campo tiene excitaciones alrededor del mnimo del potencial que cuestan
energa y dicho termino se interpreta como la masa de la partcula. Ver gura 1.7. En Teora
Cu antica de Campos al estado de mnima energa se le llama el vaco y las excitaciones alrededor del
vacio corresponden a las partculas.
Si
2
< 0, no existe un mnimo del potencial alrededor del cual el campo pueda oscilar. Adem as
el alejamiento del campo del punto de simetra del potencial no cuesta energa. Por consiguiente en
ese caso, el termino de interacci on
V () =
1
2
2
< 0, (1.453)
no puede interpretarse como un termino de masa en el Lagrangiano dado por la ec. (1.451).
Consideremos ahora el potencial
V () =
1
2
2
+
1
4
2
< 0, > 0 (1.454)
1.14. SPONTANEOUS SYMMETRY BREAKING 85
Figure 1.8: V () =
1
2
2
+
1
4
4
con
2
< 0, y > 0. Simetra ex acta
que mantiene la simetra bajo la transformacion discreta . > 0 garantiza la aparici on de los
dos mnimos que se muestran el la gura 1.8. Si la energa es sucientemente alta como se muestra
en la gura 1.8, las excitaciones son simetricas con respecto al m aximo del potencial y el termino en
2
no puede interpretarse como masa para la partcula escalar.
Sin embargo, si la energa es sucientemente baja como se muestra en la gura 1.9, las excitaciones
alrededor del mnimo dan lugar a la aparicion de un termino de masa para el campo escalar. Ademas,
dichas excitaciones no respetan la simetras . En tal caso decimos que la simetra ha sido
espont aneamente rota: aunque el Lagrangiano mantiene la simetra original, el vaco la rompe.
Para analizar cuantitativamente el espectro de partculas es necesario expandir el campo alrededor
del mnimo y determinar las excitaciones. Establezcamos en primer lugar los mnimos del potencial.
La V/ = 0 da lugar a
2
+
3
= 0 (1.455)
(
2
+
2
) = 0, (1.456)
con extremos
max
= 0, y
min
v =
_
. (1.457)
De hecho
2
V
2
=
2
+ 3
2
. (1.458)
= 0 corresponde a un m aximo, mientras que la segunda derivada para =
_
2
/ es 2
2
> 0
y corresponden a los mnimos. Expandiendo el campo alrededor del mnimo
(x) = H(x) + v (1.459)
Figure 1.9: V () =
1
2
2
+
1
4
4
con
2
< 0, y > 0. Simetra espont aneamente rota.
V () =
1
2
2
+
1
4
4
=
1
2
2
(H + v)
2
+
1
4
(H + v)
4
=
1
2
2
(H + v)
2
+
1
4
(H + v)
4
=
1
2
2
_
H
2
+ 2vH + v
2
_
+
1
4
_
H
2
+ 2vH + v
2
_
2
=
1
2
2
_
H
2
+ 2vH + v
2
_
+
1
4
_
H
4
+ 2H
2
_
2vH + v
2
_
+
_
2vH + v
2
_
2
_
=
1
2
2
_
H
2
+ 2vH + v
2
_
+
1
4
_
H
4
+ 4vH
3
+ 2H
2
v
2
+ 4v
2
H
2
+ 4v
3
H + v
4
=
1
2
2
_
H
2
+ 2vH + v
2
_
+
1
4
_
H
4
+ 4vH
3
+ 6H
2
v
2
+ 4v
3
H + v
4
=
1
2
2
H
2
3
2
H
2
2
+
2
vH
2
vH +
1
2
2
v
2
1
4
2
v
2
+
1
4
_
H
4
+ 4vH
3
V (H) =
1
2
_
2
2
_
H
2
+ vH
3
+
1
4
H
4
+
1
4
2
v
2
, (1.460)
y
L
H
=
1
2
H
1
2
_
2
2
_
H
2
vH
3
1
4
H
4
+ constant. (1.461)
Entonces H adquiere una masa 2
2
y no es invariante bajo H H.
Otro metodo es usar las ecuaciones de mnimo
2
= v
2
, para eliminar un par ametro del
potencial:
V () =
1
2
v
2
2
+
1
4
4
=
1
2
v
2
_
H
2
+ 2vH + v
2
_
+
1
4
_
H
4
+ 4vH
3
+ 6H
2
v
2
+ 4v
3
H + v
4
=v
2
H
2
+ vH
3
+
1
4
H
4
+ constant. (1.462)
Figure 1.10: Potential for complex scalar eld
Podemos escribir el potencial en terminos del nuevo campo como
V (H) =
1
2
m
2
H
H
2
+
1
2
m
2
H
v
H
3
+
1
8
m
2
H
v
2
H
4
. (1.463)
donde
m
2
H
= 2
= 2v
2
(1.464)
Consideremos ahora un campo escalar complejo sin termino de masa, pero con potencial:
L =
V () (1.465)
V () =
2
+ (
)
2
2
< 0, > 0 (1.466)
La simetra del Lagrangiano corresponde a U(1) global. Este potencial corresponde al sombrero
mexicano, como se ilustra en la Figura 1.10. Para una energa sucientemente baja de manera que
el campo deba oscilar alrededor del mnimo aparecen dos tipos de excitaciones. Una sobre las paredes
que cuestan energa y corresponden a un campo escalar masivo como en el caso anterior, y otra a lo
largo de la circunferencia de mnimo, que corresponde a una partcula escalar sin masa, y es llamada
boson del Golstone.
El Lagrangiano escalar complejo es equivalente al Lagragiano de dos campos escalares reales con
los mismos parametros. Para un conjunto de N campos reales tenedremos (suma sobre i) [8]
3
:
L =
1
2
1
2
1
2
2
_
i
_
2
, (1.467)
que es invariante bajo una simetra O(N)
i
= R
ij
j
, (1.468)
para cualquier matriz N N ortogonal R. El analisis para N = 2 da lugar a un boson de Goldstone.
El analis para N > 2 es el mismo y por cada campo real que se introduzca aparece un nuevo bos on
de Goldstone [8]:
[. . . ] there are not continuous symmetries for N = 1, while for N = 2 there is a single
direction of rotation. A rotation in N dimensions can be in any one of N(N1) planes, so
the O(N)symmetric theory has N(N 1)/2 continuous symmetries. After spontaneous
symmetry breaking there are (N 1)(N 2)/2 remaining symmetries corresponding to
rotations of the (N 1) [non massive] elds. The number of broken symmetries is the
dierence, N 1.
Entonces tenemos el siguiente teorema [8]
Goldstones theorem states that for every spontaneously broken continuous symmetry, the
theory must contain a massless particle.
Also from [8]
4
In a global symmetry that is spontaneously broken the symmetry currents are still con-
served and interactions are similarly restricted [the Lagrangian keeps the symmetry], but
the vacuum state does not respect the symmetry and the particles do not form obvi-
ous symmetry multiplets. Instead, such a theory contains massless particles, Goldstone
bosons, one for each generator of the spontaneously broken symmetry. The third case
is that of a local, or gauge, symmetry. [. . . ] such a symmetry requires the existence of
a massless vector eld for each symmetry generator, and the interactions among these
elds are highly restricted.
It is now only natural to consider a fourth possibility: What happens if we include both
local gauge invariance and spontaneous symmetry breaking in the same theory?
3
11.1
4
Introduction to Chapter 20
En el caso de la Acci on invariante gauge local bajo el Grupo U(1), tenemos el Lagrangiano (??):
L = (T
)
2
1
4
F

2
< 0 and > 0 (1.469)
Para obtener directamente el espectro despues de la ruptura espont anea de simetra es conveniente
usar la transformacion gauge de la ec. (??). Haciendo (x) = (x):

= e
i(x)
e
i(x)
_
H(x) + v
2
_
=
H(x) + v
2
(1.470)
L L
=
_
(T
(T
2
(
_
(
1
4
(F
=
1
2
_
H + igA
(H + v)
H igA
(H + v)]
1
2
2
(H + v)
2
1
4
(H + v)
4
1
4
F
(1.471)
En adelante omitiremos las primas, aunque debe estar claro que se esta trabajando en el gauge
especco de la ec. (1.470). Entonces
L =
1
2
H
1
2
2
(H + v)
2
1
4
(H + v)
4
+
1
2
g
2
A
(H + v)
2
1
4
F
. (1.472)
Usando la ec. (1.460)
L = L
H
+L
A
+
1
2
g
2
A
H
2
+ g
2
vA
H, (1.473)
donde L
H
esta dado por la ec. (1.461) y
L
A
=
1
4
F
+
1
2
g
2
v
2
A
. (1.474)
Teniendo en cuenta la ec. (1.221) para el Lagrangiano de Proca, vemos que como consecuencia de la
ruptura espontanea de simetra el campo gauge ha adquirido una masa
m
A
= gv. (1.475)
El mecanismo completo mediante el cual, a partir de un Lagrangiano invariante gauge local, los
bosones gauge adquieren masa se llama mecanismo de Higgs [?]. La partcula escalar que adquiere
masa se llama Higgs, mientras que el bos on de Goldstone es absorbido por campo gauge como modo
longitudinal.
El n umero de grados de libertad independientes en el Lagrangiano original en la ec. (1.469) es
cuatro. Correspondientes a los dos grados de libertad del boson gauge no masivo y los dos del campo
escalar complejo. En el Lagrangiano nal en la ec. (1.473) no aparece el bos on de Goldstone. Sin
embargo esto no es un problema porque dicho Lagrangiano tambien tiene cuatro grados de libertad
correspondientes a los tres grados de libertad del boson gauge masivo y al grado de libertad del bos on
de Higgs.
1.15 Fermiones quirales de cuatro componentes
Sea
P
L

1
5
2
P
R

1 +
5
2
. (1.476)
Adem as
L
P
L
R
P
R
. (1.477)
Entonces
=
L
+
R
. (1.478)
Las matrices P
L,R
tienen las propiedades
P
L
+ P
R
= 1 P
2
L,R
= P
L,R
P
L,R
= P
L,R
P
L
P
R
= 0 P
L,R
= P
L,R
. (1.479)
P
L,R
=
1
5
2

1
5
2
=
P
R,L
(1.480)
Para escribir el Lagrangiano en termino de los nuevos
L,R
debemos tener en cuenta que
L,R
= (P
L,R
)
0
=
P
L,R
0
=
0
P
R,L
= P
R,L
(1.481)
L =i
m
=i(P
L
+ P
R
)
m(P
L
+ P
R
)
=iP
L
+ iP
R
mP
L
mP
R
=iP
L
P
L
+ iP
R
P
R
mP
L
P
L
mP
R
P
R
=iP
L
P
R
+ iP
R
P
L
mP
L
P
L
mP
R
P
R
=i
R
R
+ i
L
L
m(
R
L
+
L
R
) . (1.482)
1.15. FERMIONES QUIRALES DE CUATRO COMPONENTES 91
En terminos de espinores izquierdos y derechos de cuatro componentes la transformaci on de paridad
t t x x
L
(t, x)
R
(t, x),
R
(t, x)
L
(t, x)
0

0

L
(t, x)
R
(t, x),
R
(t, x)
L
(t, x) . (1.483)
Adem as L = r p (r) (p) = L, y como
esta asociado al momento angular intrnsico,

entonces tambien
Entonces la transformaci on de paridad da lugar a (sin tener en cuenta el cambio de argumento

en los campos que desaparece en la integral de la Acci on)
R
=
R
R
+
R

R

L
L

L
=
L
L
+
L

L
=
L
L
+
L
0

L
=
L

0
L
+
L

L
=
R

R
. (1.484)
Entonces
L L
=i
R

R
+ i
L

L
m(
R
L
+
L
R
) , (1.485)
donde
= U
, con U =
0
. Como las dos representaciones dan lugar a la misma fsica,
podemos decir que la Accion en terminos de espinores L, R de cuatro componentes es invariante bajo
la transformaci on de paridad.
La existencia de ambos espinores
L,R
garantizan que el Lagrangiano de Dirac es invariante bajo
la transformaci on de paridad.
La corriente de la electrodinamica cuantica en ec. (1.374) (o la de la cromodinamica cu antica,
ec. (1.418)) conservan paridad ya que, siguiendo los mismos pasos que en la ec. (1.482)
=
L
L
+
R
R

L

L
+
R

R
. (1.486)
Si para alguna partcula, como es el caso del neutrino, no existe la componente derecha, entonces
la correspondiente interacci on vectorial viola paridad y no puede tener ni interacciones electro-
magneticas ni fuertes, es decir, no se acopla con el foton o los gluones. Ademas dicha partcula
no puede tener masa de Dirac. En el caso del neutrino esto se entiende pues al no tener carga
electrica s olo requiere dos grados de libertad independientes.
De otro lado, si una determinada interacci on, como es el caso de la interaccion debil, solo participa
Type Name Symbol Charge
leptons electron e 1
neutrino 0
quarks up quark u
1
, u
2
, u
3
2/3
down quark d
1
, d
2
, d
3
1/3
Table 1.2: Elementary fermions. The symbol represent both the particle, e.g e
, as the antiparticle,
e.g, e
+
. The lectric chage is given in units of the electron chage e
la componente izquierda de la ec. (1.486), est a corresponde a una interaccion del tipo
L
= P
R
P
L
=
P
L
_
1
5
2
_
=
1
2
(
5
) , (1.487)
que de acuerdo a la asignacion en la Tabla corresponde a una corriente VA.
1.16 Standard model Lagrangian
The known matter is build from the elementary set of particles dened in table
we further dene the color triplets of quarks as
u =
_
_
u
1
u
2
u
3
_
_
d =
_
_
d
1
d
2
d
3
_
_
(1.488)
The free Lagrangian containing this particles can be written as
L
free
=ie
e m
e
ee + i
+ iu
u m
u
uu + id
d m
d
dd
=ie
L
e
L
+ ie
R
e
R
m
e
(e
R
e
L
+ e
L
e
R
) + i
L
L
+ i
R
R
+ iu
L
u
L
+ iu
R
u
R
m
e
(u
R
u
L
+ u
L
u
R
)id
L
d
L
+ id
R
d
R
m
e
(d
R
d
L
+ d
L
d
R
) . (1.489)
donde,
L,R
=P
L,R
, e
L,R
=P
L,R
e
u
L,R
=P
L,R
u, d
L,R
=P
L,R
d . (1.490)
1.16. STANDARD MODEL LAGRANGIAN 93
Corrientes VA
En las interacciones debiles s olo participan las partes izquierdas de los campos. Esto nos permite
prescindir del
R
, pues no tiene carga electrica, fuerte, o debil
L
free
=ie
L
e
L
+ ie
R
e
R
m
e
(e
R
e
L
+ e
L
e
R
) + i
L
L
+ iu
L
u
L
+ iu
R
u
R
mue(u
R
u
L
+ u
L
u
R
) + id
L
d
L
+ id
R
d
R
m
d
(d
R
d
L
+ d
L
d
R
) . (1.491)
Simetra global SU(3)
c
SU(2)
L
U(1)
Y
En el contexto de las interacciones debiles un e
L
es completamente equivalente a un campo
L
.
Es decir, el Lagrangiano debe ser invariante bajo una transformacion SU(2)
L
de esos campos. La
diferencia entre ellos son sus respectivas cargas electricas y sus masas. Asumiendo que ambos campos
tienen una misma hipercarga, asociada a una nueva simetra Abeliana U(1)
Y
, podramos esperar que
la corriente electromagnetica apropiada pueda obtenerse a partir del Grupo semisimple SU(2)
L

U(1)
Y
. Ademas las respectivas masas se podran obtener a partir del mecanismo de Higgs.
La simetra SU(2)
L
entre las partes izquierdas del neutrino y el electron, y entre las partes
izquierdas de los quarks up y down, se establece deniendo los dobletes:
L
_
L
e
L
_
Q =
_
u
L
d
L
_
, (1.492)
De otro lado, La invarianza bajo U(1)
Y
requiere que
Y
L
=Y
L
= Y
e
L
Y
Q
=Y
u
L
= Y
d
L
. (1.493)
El generador de carga electrica

Q, se va obtener a partir de una combinacion lineal del generador
diagonal de SU(2)
L
, T
3
, y del generador de hipercarga, Y .
Bajo la simetra SU(2)
L
, los campos transforman como:
L L
=exp(iT
i
i
)L (1 + iT
i
i
)L
Q Q
=exp(iT
i
i
)Q (1 + iT
i
i
)Q
e
R
e
R
= e
R
u
R
u
R
= u
R
d
R
d
R
= d
R
. (1.494)
donde
T
i
=

i
2
, (1.495)
y
i
son las matrices de Pauli dadas en la ec. (1.331).
Claramente el termino de masa m
e
en la ec. (??) no es invariante bajo la simetra SU(2)
L
. El
Lagrangiano en la ec. (??), sin termino de masa, puede reescribirse de manera que exhba de forma
m as explicita la invarianza bajo SU(2)
L
como
L
free
=ie
L
e
L
+ ie
R
e
R
+ i
L
L
+ iu
L
u
L
+ iu
R
u
R
+ id
L
d
L
+ id
R
d
R
=iL
L + iQ
Q + ie
R
e
R
+ iu
R
u
R
+ id
R
d
R
. (1.496)
Simetra gauge local SU(3)
c
SU(2)
L
U(1)
Y
Para obtener la interacciones del modelo est andar, reemplazamos las derivadas normales por derivadas
covariantes.
Proponemos entonces el Lagrangiano
L =iQ
Q + iL
L + ie
R
e
R
+ id
R
d
R
+ iu
R
u
R
1
4
G
a
G
a
1
4
W
i
W
i
1
4
B
, (1.497)
Donde
T
ig
s
a
2
G
a
ig
i
2
W
i
ig
Y B
. (1.498)
donde

a
2
, a = 1, 2, . . . , 8 8 generadores de SU(3)
c
T
i

i
2
, i = 1, 2, 3 3 generadores de SU(2)
L
Y generador de U(1)
Y
All the particles in this Lagrangian are massless. It is only good for the gluons and the Abelian
gauge boson, but is no realist for the fermions of the weak gauge bosons W
i
. To solve this problem,

we postulate a new complex scalar doublet with four degree of freedom:
=
_
0
_
=
_
1
+ i
2
3
+ i
4
_
. (1.499)
The + and 0 superindexes for just for later convenience. The full Lagrangian involving those elds
are
L =iQ
Q + iL
L + ie
R
e
R
+ id
R
d
R
+ iu
R
u
R
1
4
G
a
G
a
1
4
W
i
W
i
1
4
B
+ (T
)
2
(h
e
Le
R
+ h
d
Qd
R
+ h
u
Q
u
R
+ h.c)
=L
fermion
+L
gauge
+L
WBH
L
Yukawa
. (1.500)
donde
2
< 0, y > 0,
= i
2
, (1.501)
L
fermion
=iQ
Q + iL
L + ie
R
e
R
+ id
R
d
R
+ iu
R
u
R
L
gauge
=
1
4
G
a
G
a
1
4
W
i
W
i
1
4
B
L
WBH
=(T
)
2
L
Yukawa
=h
e
Le
R
+ h
d
Qd
R
+ h
u
Q
u
R
+ h.c (1.502)
Bajo una transformaci on gauge local las derivadas covariantes de los campos (y por consiguiente
los campos) transforman como:
T
L (T
L)
= exp
_
i
i
T
i
iY
L
_
T
L
T
Q (T
Q)
= exp
_
i
a
a
i
i
T
i
iY
Q
_
T
Q
T
(T
= exp
_
i
i
T
i
iY
_
T
e
R
(T
e
R
)
= exp (iY
e
R
) T
e
R
= exp (iQ
e
R
) T
e
R
T
d
R
(T
d
R
)
= exp (i
a
a
iY
d
R
) T
d
R
= exp (i
a
a
iQ
d
R
) T
d
R
T
u
R
(T
u
R
)
= exp (i
a
a
iY
u
R
) T
u
R
= exp (i
a
a
iQ
u
R
) T
u
R
. (1.503)
donde Q
e
R
= 1, etc, son las cargas electricas asociadas a los campos.
Para los campos del Lagrangiano, debemos asegurarnos de que todos los terminos invariantes
gauge locales y renormalizables sean considerados. De hecho un termino de interacci on entre fermiones
y el campo escalar, correspondiente a una interacci on de Yukawa: Le
R
y Qd
R
son invariantes bajo
transformaciones SU(3)
c
SU(2)
L
U(1)
Y
si
Y
L
+ Y
+ Q
e
R
=0
Y
Q
+ Y
+ Q
d
R
=0
Y
Q
+ Y
+ Y
u
R
= Y
Q
Y
+ Q
u
R
=0 ,
From this set of three equations we obtain the three doublet hypercharges:
Y
L
=
1
2
, Y
=
1
2
, Y
Q
=
1
6
. (1.504)
En el an alisis anterior hemos jado Y
= Y
. Esto es debido a que si Q es un invariante SU(2)

L
,
el termino

Q tambien es un invariante de SU(2). Explcitamente
Q =(i
2
Q
=
_

0
Q
=
_
+
_
_
u
L
d
L
_
=
0
u
L
+
d
L
=
12
Q
1
2
+
21
Q
2
1
=
ab
Q
a
b
. (1.505)
Bajo una transformaci on SU(2)
L
=
ab
Q
b
=
ab
U
ac
U
bd
Q
c
d
=
cd
det UQ
c
d
=
cd
Q
c
d
=
Q. (1.506)
Sin perdida de generalidad los cuatro grados de libertad de , pueden escribirse en la forma
=e
i
j
(x)T
j
_
0
1
2
[H(x) + v]
_
. (1.507)
El potencial escalar, denido por
V () =
2
+ (
)
2
(1.508)
se reduce a
V (H) =
1
2
2
(H + v)
2
+
1
4
(H + v)
4
. (1.509)
1.16.1 Spontaneous symmetry breaking in SU(3)
c
SU(2)
L
U(1)
Y
Retornando al doblete de Higgs del modelo estandar en la ec. (1.507), los cuatro grados de libertad
de , pueden escribirse en forma polar con la parte real neutra desplazada para generar la ruptura
espont anea de la simetra SU(2)
L
U(1)
Y
=e
i
j
T
j
_
0
1
2
(H(x) + v)
_
(1.510)
1
2
_
1 + i
3
2i
+
2i
1 i
3
__
0
1
2
(H(x) + v)
_
=
1
2
_
i
+
H + vi
+
1
2
(H + v i
3
H i
3
v)
_
=
_
G
+
1
2
(H(x) + v iG
0
)
_
.
Para SU(2)
L
U(1)
Y
tenemos cuatro generadores y cuatro bosones gauge. De acuerdo a la parametrizaci on
en ec. (1.510) esperamos que aparezcan tres bosones de Goldstone y un campo de Higgs con masa,
de manera que quedar a un generador no roto correspondiente a una simetra remanente del vaco
U(1)
Q
SU(2)
L
U(1)
Y
U(1)
Q
. (1.511)
Se espera entonces que el espectro consista de un bos on de Higgs, tres bosones gauge masivos, y
un boson gauge sin masa.
Podemos hacer una transformaci on gauge similar a la de la ec. (??) sobre el campo , tal que

=
_
0
1
2
(H(x) + v)
_
, (1.512)
que dene el gauge unitario. En adelante sin embargo omitiremos las primas sobre los campos
transformados
y W
.
Comenzaremos analizando la parte escalar del Lagrangiano del Modelo dada en la ec. (1.502)
L
WBH
=
1
2
_
T
_
0
H(x) + v
__
_
0
H(x) + v
_
V (H) , (1.513)
donde V (H) dado en la ec. (1.463), incluye el termino de masa para el boson de Higgs (1.464):
m
2
H
= 2
= 2v
2
(1.514)
Como
W
= T
i
W
i
=
1
2
_
0 1
1 0
_
W
1
+
1
2
_
0 i
i 0
_
W
2
+
1
2
_
1 0
0 1
_
W
3
=
1
2
_
_
_
_
W
3
2
W
1
i W
2
2
W
1
+ i W
2
2
W
3
_
_
_
_
1
2
_
W
3
2W
+
2W
W
3
_
=
_
1
2
W
3
1
2
W
+
2
W

1
2
W
3
_
. (1.515)
T
corresponde a la matrix 2 2
T
=
_
i
_
1
2
gW
3
+ g
Y B
_

i
2
gW
+
2
gW
i
_
1
2
gW
3
+ g
Y B
_
_
. (1.516)
Entonces
T
=
_

i
2
gW
+
(H + v)
H i
_
1
2
gW
3
+ g
_
(H + v)
_
. (1.517)
De modo que
L
WBH
=
1
2
__

i
2
gW
+
(H + v)
H i
_
1
2
gW
3
+ g
_
(H + v)
__
_

i
2
gW
+
(H + v)
H i
_
1
2
gW
3
+ g
_
(H + v)
_
V (H)
=
1
2
_
i
2
gW
(H + v)
H + i
_
1
2
gW
3
+ g
_
(H + v)
_
_

i
2
gW
+
(H + v)
H i
_
1
2
gW
3
+ g
_
(H + v)
_
V (H)
=
1
4
g
2
W
W
+
(H + v)
2
V (H)
+
1
2
_
H + i
_
1
2
gW
3
+ g
_
(H + v)
H i
_
1
2
gW
3
+ g
_
(H + v)
=V (H) +
1
4
g
2
W
W
+
(H + v)
2
+
+
1
2
H +
1
2
_
1
2
gW
3
+ g
_
2
(H + v)
2
(1.518)
donde la ultima lnea corresponde a la magnitud del n umero complejo:
_
H i
_
1
2
gW
3
+ g
_
(H + v)
(1.519)
Entonces
L
WBH
=
1
2
H V (H)
+
_
gv
4
_
2
W
W
+
+
1
4
g
2
W
W
+
H
2
+
1
2
vg
2
W
W
+
H +L
ZAH
, (1.520)
donde
L
ZAH
=
1
2
_
1
4
g
2
W
3
W
3

1
2
gg
3
B
1
2
gg
3
B
+ g
2
Y
2
_
2
_
H
2
+ 2vH + v
2
_
(1.521)
Haciendo Y
= 1/2 como en la ec. (1.504),

L
ZAH
=
1
8
_
W
3
B
_
_
g
2
gg
gg
2
__
W
3
_
_
H
2
+ 2vH + v
2
_
(1.522)
Sea
V =
_
cos
W
sin
W
sin
W
cos
W
_
=
1
_
g
2
+ g
2
_
g g
g
_
, (1.523)
con tan
W
= g
/g, tal que g sin

W
= g
cos
W
, como en la ec. (??). Si (ver ec. (??)),
_
W
3
_
= V
_
Z
_
(1.524)
entonces
L
ZAH
=
1
8
_
Z
_
_
V
T
_
g
2
gg
gg
2
_
V
_ _
Z
_
_
H
2
+ 2vH + v
2
_
(1.525)
V
T
_
g
2
gg
gg
2
_
V =
1
g
2
+ g
2
_
g
3
+ gg
2
g
2
g
3
+g
2
g
g
2
g
gg
2
+ gg
2
__
g g
g
_
=
1
g
2
+ g
2
_
g
3
+ gg
2
g
2
g
3
0 0
__
g g
g
_
=
1
g
2
+ g
2
_
g
4
+ g
2
g
2
+ g
2
g
2
+ g
4
g
3
g
+ gg
3
g
3
g
gg
3
0 0
_
=
_
g
2
+ g
2
0
0 0
_
(1.526)
L
ZAH
=
1
2
_
g
2
+ g
2
4
_
Z
_
H
2
+ 2vH + v
2
_
=
1
2
_
g
2
_
2 _
1 + tan
2
W
_
Z
_
H
2
+ 2vH + v
2
_
=
1
2
_
g
2 cos
W
_
2
Z
_
H
2
+ 2vH + v
2
_
=
1
2
_
gv
2 cos
W
_
2
Z
+
1
2
_
g
2 cos
W
_
2
Z
H
2
+
_
g
2 cos
W
_
2
vZ
H (1.527)
Retornando a la ec. (1.520), tenemos tenemos
L
WBH
=(T
_
2
=
1
2
H V (H)
+
1
4
g
2
W
W
+
H
2
+
1
2
vg
2
W
W
+
H
+
1
2
_
g
2 cos
W
_
2
Z
H
2
+
_
g
2 cos
W
_
2
v Z
H
+
1
2
m
2
W
W
W
+
+
1
2
m
2
W
W
W
+
+
1
2
m
2
Z
Z
, (1.528)
donde:
Masas gauge:
m
W
=
gv
2
m
Z
=
gv
2 cos
W
, (1.529)
y
m
Z
=
m
W
cos
W
. (1.530)
V (H) =
1
2
m
2
H
H
2
+ vH
3
+
1
4
H
4
=
1
2
m
2
H
H
2
+
m
2
H
2v
H
3
+
1
4
m
2
H
2v
2
H
4
=
1
2
m
2
H
H
2
_
1 +
H
v
+
H
2
4v
2
_
. (1.531)
con
m
2
H
= 2
2
= 2v
2
. (1.532)
_
W
3
_
=
_
cos
W
sin
W
sin
W
cos
W
__
Z
_
, (1.533)
tal que
g sin
W
= g
cos
W
. (1.534)
1.16.2 Yukawa Lagrangian
In the Unitary gauge
L
Yukawa
=h
e
Le
R
+ h
d
Qd
R
+ h
u
Q
u
R
+ h.c
=
1
2
_
h
e
(e
L
e
R
+ e
R
e
L
) + h
d
(d
L
d
R
+ d
R
d
L
) + h
u
(u
L
u
R
+ u
R
u
L
)
[H(x) + v]
=
v
2
_
h
e
ee + h
d
dd + h
u
uu
_
_
H(x)
v
+ 1
_
, (1.535)
deniendo
m
f
=
h
f
v
2
(1.536)
tenemos
L
Yukawa
=m
e
ee + m
d
dd + m
u
uu +
m
e
v
eeH +
m
d
v
ddH +
m
u
v
uuH . (1.537)
1.16.3 Fermion-gauge interactions
De la ec. (1.502) tenemos
L
fermion
=iQ
Q + iL
L + ie
R
e
R
+ id
R
d
R
+ iu
R
u
R
. (1.538)
Los terminos de interacci on generados por la simetra gauge para el campo L son:
iL
L iL
L =iL
(igT
i
W
i
ig
Y
L
B
)L
=L
(gT
1
W
1
+ gT
2
W
2
+ gT
3
W
3
+ g
Y
L
B
)L
=L
_
g
2
_
0 W
+
0
_
+ gT
3
W
3
+ g
Y
L
B
_
L
=iL
2
_
0 W
+
0
_
L + L
_
gT
3
W
3
+ g
Y
L
B
L
=L
2
_
e
L
W
+
L
W
_
+L
AZL
=
g
2
_
e
L
W
+
+ e
L
L
W
+L
AZL
=L
WL
+L
AZL
, (1.539)
donde
L
WL
=
g
2
_
e
L
W
+
+ e
L
L
W
L
AZL
=L
_
gT
3
W
3
+ g
Y
L
B
L (1.540)
Generalizando para todos los campos:
L
WL

g
2
_
e
L
W
+
+ u
L
d
L
W
+
+ h.c
. (1.541)
L
AZL
=L
[gT
3
(c
W
Z
+ s
W
A
) + g
Y
L
(s
W
Z
+ c
W
A
)] L
=L
[gT
3
c
W
Z
+ gT
3
s
W
A
Y
L
s
W
Z
+ g
Y
L
c
W
A
] L
=L
[(gc
W
T
3
g
s
W
Y
L
) Z
+ (gs
W
T
3
+ g
c
W
Y
L
) A
] L, (1.542)
donde c
W
= cos
W
, s
W
= sin
W
. Usando la relaci on entre g y g
(1.534):
L
AZL
=L
__
gc
W
T
3
g
s
2
W
c
W
Y
L
_
Z
+ (gs
W
T
3
+ gs
W
Y
L
) A
_
L
=gs
W
L
[(cot
W
T
3
tan
W
Y
L
) Z
+ (T
3
+ Y
L
) A
] L. (1.543)
Como el generador asociado a A
debe ser el generador de carga electrica, tenemos que

e = g sin
W
(1.544)
donde e es la carga electrica del electr on, y el generador de carga
Q = T
3
+ Y , (1.545)
de modo que
QL = (T
3
+ Y )L =
_
Q
0
0 Q
e
_
L =
_
0 0
0 1
_
L. (1.546)
La ec. (1.545), se conoce como la relacion Gell-MannNishijima, y establece la condici on que se deje
satisface para obtener apropiadamente la QED a partir de la interaccio on electrodebil asociada al
grupo semisimple SU(2)
L
U(1)
Y
. De esta forma
L
AZL
=eL
(cot
W
T
3
tan
W
Y
L
) LZ
+ eL
Q
L
LA
=eL
_
cot
W
T
3
tan
W
_
Q
L
T
3
__
LZ
+ eL
Q
L
LA
=
e
2c
W
s
W
L
3
2s
2
W
Q
L
_
LZ
+ eL
Q
L
LA
. (1.547)
Generalizando para los otros campos, tenemos
L
AZL

F=Q,L,e
R
,d
R
,u
R
e
2c
W
s
W
F
3
2s
2
W
Q
L
_
FZ
+ eF
Q
L
FA
. (1.548)
Usando los acoplamientos gauge de los quarks con los gluones (1.413), de los fermiones con el
W
(1.541) y con Z
y A
(1.548) para expandir L

fermion
en (1.538), tenemos
L
fermion
=iQ
Q + iL
L + ie
R
e
R
+ id
R
d
R
+ iu
R
u
R
=iu
L
u
L
+ iu
R
u
R
+ id
L
d
L
+ id
R
d
R
+ ie
L
e
L
+ ie
R
e
R
+ i
L
L
+ g
s
_
u
L
a
2
u
L
+ u
R
a
2
u
R
+ d
L
a
2
d
L
+ d
R
a
2
d
R
_
G
a
+
g
2
_
e
L
W
+
+ u
L
d
L
W
+
+ h.c
F=Q,L,e
R
,d
R
,u
R
e
2c
W
s
W
F
3
2s
2
W
Q
L
_
FZ
+ e
_
e
L
Q
e
e
L
+ e
R
Q
e
e
R
+u
L
Q
u
u
L
+ u
R
Q
u
u
R
+ d
L
Q
d
d
L
+ d
R
Q
d
d
R
_
A
. (1.549)
Para escribir este Lagrangiano en terminos de espinores de 4 componentes, tomemos algunos
casos especcos:
e
L
W
+
= P
R
P
L
eW
+
P
2
L
eW
+
P
L
eW
+
=
1
2
(1
5
)eW
+
, (1.550)
u d
e
e
2v
f
1
8
3
sin
2
W
1 +
4
3
sin
2
W
1 1 + 4 sin
2
W
2a
f
1 1 1 1
Table 1.3: Acoplamientos de corrientes neutras
_
Q
3
2s
2
W
Q
Q
_
Q2s
2
W
u
R
Q
u
u
R
2s
2
W
d
R
Q
d
d
R
_
Z
=
_
_
u
L
d
L
_
_
1 2s
2
W
Q
u
0
0 1 2s
2
W
Q
d
_
_
u
L
d
L
_
2s
2
W
u
R
Q
u
u
R
2s
2
W
d
R
Q
d
d
R
_
Z
=
_
u
L
u
L
d
L
d
L
2s
2
W
_
(u
L
+ u
R
)
Q
u
(u
L
+ u
R
) +
_
d
L
+ d
R
_
Q
d
(d
L
+ d
R
)
_
Z
=
_
1
2
u
(1
5
)u
1
2
d
(1
5
)d 2s
2
W
_
u
Q
u
u + d
Q
d
d
_
_
Z
=
_
u
__
1
2
2s
2
W
Q
u
_
1
2
5
_
u + d
__
1
2
2s
2
W
Q
d
_
+
1
2
5
_
d
_
Z
=
_
u
(v
u
a
u
5
) u + d
(v
d
a
d
5
) d
, (1.551)
donde
v
f
=T
f
3
2 sin
2
W
Q
f
a
f
=T
f
3

5
(1.552)
Los valores explcitos para v
f
y a
f
en el modelo estandar, est an dados en la Tabla 1.3.
Usando las expresiones para pasar de fermiones L, R a los fermiones de Dirac de cuatro compo-
mentes, y las ecuaciones (1.550), (1.551) tenemos
L
fermion
=iu
u + id
d + ie
e + i
L
L
+ g
s
_
u
a
2
u + d
a
2
d+
_
G
a
+
g
2
2
_
(1
5
)eW
+
+ u
(1
5
)dW
+
+ h.c
f=u,d,,e
e
2c
W
s
W
f
(v
f
a
f
5
) f
+ e
_
e
Q
e
e + u
Q
u
u + d
Q
d
d
_
A
f=u,d,,e
if
f +
q=u,d
g
s
q
a
2
qG
a
+
g
2
2
_
(1
5
)eW
+
+ u
(1
5
)dW
+
+ h.c
f=u,d,,e
e
2c
W
s
W
f
(v
f
a
f
5
) f
+ e
f=u,d,,e
f
Q
f
fA
, (1.553)
donde Q
f
est an dadas en la Tabla 1.2 y v
f
, a
f
en la Tabla 1.3.
1.16.4 Self-interactions
El Lagrangiano gauge
L
gauge
=
1
4
G
a
G
a
1
4
W
i
W
i
1
4
B
(1.554)
se puede escribir como
L
gauge
=
1
4
F
1
4
Z
1
2
(F
W
)
(F
W
)
1
4
G
a
ie cot
W
_
(F
W
)
W
+
(F
W
)
+ W
W
+
_
ie
_
(F
W
)
W
+
(F
W
)
+ W
W
+
e
2
2 sin
2
W
_
_
W
+
_
2
W
+
W
+
W
_
e
2
cot
2
W
_
W
+
W
+
_
e
2
cot
2
W
_
2W
+
W
+
W
+
_
e
2
_
W
+
W
+
1
4
_
g
s
a
f
ade
G
d
G
e
+ g
s
f
abc
G
b
G
G
a
+ g
2
s
f
abc
f
ade
G
b
G
c
G
d
G
e
_
, (1.555)
donde
(F
W
)
W
+
W
+
. (1.556)
1.16.5 Lagrangiano del modelo estandar para la primera generacion
Recopilando los resultados para L
WBH
(1.528), L
Yukawa
(1.537), L
fermion
(1.553), y L
gauge
(1.555),
tenemos para f =
e
, e, u, d; q = u, d [con f
= e (d) para f =
e
(u) ]
L
1 gen
=
f
i
f (
m
f
) f
1
4
F
1
4
Z
1
2
(F
W
)
(F
W
)
1
4
G
a
+
1
2
H
1
2
m
2
H
H
2
_
1 +
H
v
+
H
2
4v
2
_
+
_
m
2
W
W
W
+
+
1
2
m
2
Z
Z
__
1 + 2
H
v
+
H
2
v
2
_
+ g
s
q
q
a
2
_
q G
a
+ e
Q
f
fA
+
e
2 cos
W
sin
W
(v
f
a
f
5
)fZ
+
g
2
2
_
(1
5
)f
W
+
+ h.c
_
+
f
m
f
v
ffH
ie cot
W
_
(F
W
)
W
+
(F
W
)
+ W
W
+
_
ie
_
(F
W
)
W
+
(F
W
)
+ W
W
+
e
2
2 sin
2
W
_
_
W
+
_
2
W
+
W
+
W
_
e
2
cot
2
W
_
W
+
W
+
_
e
2
cot
2
W
_
2W
+
W
+
W
+
_
e
2
_
W
+
W
+
1
4
_
g
s
a
f
ade
G
d
G
e
+ g
s
f
abc
G
b
G
G
a
+ g
2
s
f
abc
f
ade
G
b
G
c
G
d
G
e
_
. (1.557)
1.16.6 Dinamica de sabor
El Modelo Estandar esta compuesto de las siguientes tres familias de fermiones i = 1, 2, 3. A cada
familia se le asigna una carga de sabor diferente
L
i
=
_
i
L
e
i
L
_
: L
1
=
_
e
L
e
L
_
L
2
=
_
L
_
L
3
=
_
L
_
e
i
R
: e
R
,
R
,
R
Q
i
=
_
u
i
L
d
i
L
_
: Q
1
=
_
u
L
d
L
_
Q
2
=
_
c
L
s
L
_
Q
3
=
_
t
L
b
L
_
u
i
R
: u
R
, c
R
, t
R
d
i
R
: d
R
, s
R
, b
R
. (1.558)
Con
Y
L
i
=
1
2
Y
Q
i
=
1
6
Y
e
i
R
=1 Y
u
i
R
=
2
3
Y
d
i
R
=
1
3
. (1.559)
De los procesos entre familias, es decir de cambio de sabor, sabemos que
No se han observado procesos de corrientes neutras que cambian sabor.
Los bosones gauge cargados W
decaen siempre a leptones de la misma generaci on y con la

misma intensidad.
Proponemos entonces el Lagrangiano
L =i
i
_
Q
i
+ L
i
+ e
R
i
e
R
i
+ d
R
i
d
R
i
+ u
R
i
u
R
i
_
(h
E
ij
L
i
e
R
j
+ h
D
ij
Q
i
d
R
j
+ h
U
ij
Q
u
R
j
+ h.c)
1
4
W
i
W
i
1
4
B
+ (T
+ (
)
2
. (1.560)
Para aclarar la notaci on, obviando de momento la denici on denitiva de h
ij
y las primas sobre los
campos, consideremos el Lagrangiano de Yukawa para el sector down
L h
D
ij
d
Ri
Q
j
+ h.c
h
D
ij
d
Ri
ab
a
Q
b
j
+ h.c
h
D
ij
ab
d
R
a
Q
b
j
+ h.c
h
D
ij
ab
(d
R
)
a
Q
b
j
+ h.c , (1.561)
donde i, a, , son ndices en los espacios de familia, SU(2)
L
, SU(3)
c
y de Dirac, respectivamente.
Por ejemplo el primer termino de la sumatoria
L h
D
11
(d
R
)
1
1
1
Q
2
11
+ . . .
h
D
11
d
R
r
0
d
r
L
+ . . . (1.562)
corresponde a la interaccion de Yukawa del quark down rojo (r) con un campo escalar complejo
neutro en carga electrica pero de isospn debil 1/2. En forma compacta la primera expresion en la
ec. (1.561) puede escribirse como
L d
R
h
D
Q+Q
L
h
D
d
R
(1.563)
Retornado a la ec. (1.560), tenemos que para denir apropiadamente la masa de los quarks, rotamos
de los autoestados de interacci on a los autoestados de masa con la matrices unitarias
d
R,L
j
= (V
D
R,L
)
jk
d
R,L
k
d
R,L
j
= d
R,L
k
(V
D
R,L
)
kj
(1.564)
Tal que
(V
D
R,L
)
ij
(V
D
R,L
)
jk
=
ik
(V
D
L
)
ki
M
D
ij
(V
D
R
)
jl
= m
D
k

kl
(1.565)
Con deniciones similares para los campos u
i
y e
i
.
L
Yukawa
d
L
i
h
D
ij
v
2
d
R
j
=d
L
i
M
D
ij
d
R
j
=d
Lk
(V
D
L
)
ki
M
D
ij
(V
D
R
)
jl
d
Rl
=d
Lk
m
D
k

kl
d
Rl
=m
D
k
d
Lk
d
Rk
(1.566)
Para las diferentes combinaciones de terminos de corrientes
u
L
d
L
i
=u
Lk
(V
U
L
)
ki
(V
D
L
)
il
d
Ll
=V
kl
u
Lk
d
Ll
e
L
i
=
L
(V
E
L
)
ij
e
Lj
=
L
i
(V
E
L
)
ij
e
Lj
=
Lj
e
Lj
(1.567)
Donde hemos denido la matriz de CabibboKobayashiMaskawa (CKM) como
V =V
U
L
V
D
L
V
V =V
D
L
V
U
L
V
U
L
V
D
L
= 1
j
V
ij
V
jk
=
ik

j
V
ji
V
jk
=
ik

j
[V
ji
[
2
=
j
[V
ij
[
2
= 1
(1.568)
y los autoestados debiles de los neutrinos como
i
= (V
E
L
)
ij
Lj
(1.569)
Con esta denicion, las corrientes debiles cargadas para los leptones siguen siendo universales. Sim-
ilarmente
u
L
u
L
i
=u
Lk
(V
U
L
)
ki
(V
U
L
)
il
u
Ll
=
kl
u
Lk
u
Ll
=u
Lk
u
Lk
(1.570)
De modo que todas las corrientes neutras permanecen universales despues de la redenicion de los
campos fermionicos. A este resultado, basado en la unitariedad de las transformaciones biunitarias se
le llama Mecanismo GIM. En muchas extensiones del Modelo Est andar las matrices que transforman
los fermiones a sus autoestados de masa no son unitarias y dan lugar a corrientes debiles neutras que
cambian sabor (FCNC de sus siglas en ingles).
Teniendo en cuenta estos resultados podemos escribir nalmente el Lagrangiano completo del
Modelo Estandar en la Gauge Unitario, para
f =
e
,
, e, , , u, c, t, d, s, b; q =u, c, t, d, s, b; l =e, , (1.571)

L
SM
=
f
i
f (
m
f
) f
1
4
F
1
4
Z
1
2
(F
W
)
(F
W
)
1
4
G
a
+
1
2
H
1
2
m
2
H
H
2
_
1 +
H
v
+
H
2
4v
2
_
+
_
m
2
W
W
W
+
+
1
2
m
2
Z
Z
__
1 + 2
H
v
+
H
2
v
2
_
+ g
s
q
q
a
2
_
q G
a
+ e
Q
f
fA
+
e
2 cos
W
sin
W
(v
f
a
f
5
)fZ
+
g
2
2
_

l=e

l
(1
5
)lW
+
ij
V
ij
u
i
(1
5
)d
j
W
+
+ h.c
_
+
f
m
f
v
ffH
ie cot
W
_
(F
W
)
W
+
(F
W
)
+ W
W
+
_
ie
_
(F
W
)
W
+
(F
W
)
+ W
W
+
e
2
2 sin
2
W
_
_
W
+
_
2
W
+
W
+
W
_
e
2
cot
2
W
_
W
+
W
+
_
e
2
cot
2
W
_
2W
+
W
+
W
+
_
e
2
_
W
+
W
+
1
4
_
g
s
a
f
ade
G
d
G
e
+ g
s
f
abc
G
b
G
G
a
+ g
2
s
f
abc
f
ade
G
b
G
c
G
d
G
e
_
. (1.572)
donde m
l
= 0.
1.17. FENOMENOLOG
IA ELECTROD
EBIL 113
1.17 Fenomenologa Electrodebil
El Lagrangiano del Modelo contiene los par ametros g
s
, g, sin
W
, v, m
H
. Alternativamente uno puede
escoger como parametros, en lugar de g, sin
W
, v [?]
G
F
= 1.166 371(6) 10
5
GeV
2
1
= 137.035 999 679(94)
m
Z
= 91.1876(20) GeV
s
(m
Z
) = 0.1176(20) . (1.573)
donde
i
= g
2
i
/(4). Esto tiene la ventaja de usar las tres cantidades experimentales mejor medidas.
Las relaciones
sin
2
W
=1
m
2
W
m
2
Z
, m
2
W
sin
2
W
=

2G
F
(1.574)
determinan entonces
sin
2
W
=0.212
m
W
=80.94 GeV (1.575)
Si se usa (M
Z
) 1/128 entonces
sin
2
W
=0.233
m
W
=79.84 GeV (1.576)
Los valores medidos son sin
2
W
= 0.23149(13), m
W
= 80.398(25) GeV, y pueden ser reproducidos por
el modelo est andar una vez se tienen en cuenta correcciones perturbativas inducidas por partculas
virtuales.
El acelerador e
+
e
LEP, que funcion o hasta desde 1998 hasta el 2000 [?], oper o a energas su-
cientes para producir millones de Z. Combinado con otros resultados experimentales, se pudo
vericar todo el Lagrangiano del Modelo Estandar hasta un nivel del 1 por mil. Con excepci on de
las interacciones asociadas con el Higgs.
La universalidad de los decaimientos del Z est a soportada por los resultados experimentales
siguientes donde solo se muestran los decaimientos lept onicos del Z diferentes de cero [?]
(Z e
+
e
) = 83.92(12) MeV (Z
+
) = 83.99(18) MeV (Z
+
) = 84.08(22) MeV
Br(Z e
+
e
) = 3.363(4)%, Br(Z
+
) = 3.366(7)%, Br(Z
+
) = 3.370(8)%
(1.577)
Mientras que para el W
, en %,
Br(W

e
e
) = 10.65(17), Br(W
) = 10.59(15), Br(W
) = 11.44(22)
(1.578)
La diferencia de
respecto a los otros representa un efecto a 2.8. La universalidad de los

acoplamientos lept onicos de W puede comprobarse tambien indirectamente a traves de los decaimien-
tos debiles mediados por corrientes cargadas. Los datos actuales verican la universalidad de los
acoplamientos de corrientes cargadas lept onicas al nivel del 0.2% [?]. Sin necesidad de entrar en
detalles de los c alculos de las amplitudes de decaimiento, podemos usar el hecho de que ellas son
proporcionales a los acoplamientos al cuadrado correspondiente, de modo que un cociente entre am-
plitudes de decaimiento es igual, en primera aproximaci on, a los cocientes de los acoplamientos al
cuadrado. Tendremos en cuenta ademas que el Branching es la amplitud de decaimiento a un canal
especico divido por la suma de las amplitudes de decaimiento a todos los canales posibles.
Para los decaimientos del Z el Modelo Estandar predice, ademas de la ausencia de eventos del
tipo Z e
+
, que para un cierto l = e, , , o q = d, s, b

Br(Z l
+
l
)
Br(Z qq)

([v
l
[
2
+[a
l
[
2
)
N
c
([v
q
[
2
+[a
q
[
2
)
=
_
_
1
2
+ 2 sin
2
W
_
2
+
1
4
_
N
c
_
_
1
2
+
2
3
sin
2
W
_
2
+
1
4
_
0.776
N
c
=
_
_
0.338 N
c
= 2
0.225 N
c
= 3
0.169 N
c
= 4
(1.579)
Para ser comparado con el resultado experimental de por ejemplo
Br(Z e
+
e
)
Br(Z
bb)
=
3.363(4)
15.12(5)
0.222 (1.580)
que de nuevo da lugar al N
c
= 3, que seguiremos tomando en adelante.
Los Branchings de decaimiento en la ec. (1.578) y ec. (1.577) pueden ser calculados sin entrar en
1.17. FENOMENOLOG
IA ELECTROD
EBIL 115
detalles del c alculo de las amplitudes. Teniendo en cuenta que el canal Z
tt esta cerrado
Br(Z e
+
e
) =
(Z e
+
e
total
=
([v
e
[
2
+[a
e
[
2
)
l
[([v
l
[
2
+[a
l
[
2
) + ([v
l
[
2
+[a
l
[
2
)] + N
c
[
2
i=1
([v
u
i
[
2
+[a
u
i
[
2
) +
3
i=1
([v
d
i
[
2
+[a
d
i
[
2
)]
=
([v
e
[
2
+[a
e
[
2
)
3[([v
e
[
2
+[a
e
[
2
) + ([v
e
[
2
+[a
e
[
2
)] + 3[2([v
u
[
2
+[a
u
[
2
) + 3([v
d
[
2
+[a
d
[
2
)]
=
([v
e
[
2
+[a
e
[
2
)
21[a
e
[
2
+ 3[[v
e
[
2
+[v
e
[
2
] + 3[2[v
u
[
2
+ 3[v
d
[
2
]
=
(1 + 4s
2
W
)
2
+ 1
21 + 3[(1 + 4s
2
W
)
2
+ 1] + 3[2(1
8
3
s
2
W
)
2
+ 3(1 +
4
3
s
2
W
)
2
]
=
2 8s
2
W
+ 16s
4
W
42 80s
2
W
+
320
3
s
4
W
3.43% (1.581)
Para W
tenemos por ejemplo

Br(W

e
e
) =
(W

e
e
total
(1.582)
donde, teniendo en cuenta que los canales a top est an cerrados, y usando la condici on de unitariedad
de la matriz CKM en ec. (1.568), tenemos
total
=
l
(W

l
l
) + N
c
i
[(W
u
1
d
i
) + (W
u
2
d
i
)]
=(W

e
e
)3 + N
c
i
[[V
1i
[
2
+[V
1i
[
2
]
=(W

e
e
)(3 + 2N
c
)
(1.583)
entonces
Br(W

e
e
) =
1
3 + 2N
c
= 11.1% (1.584)
Una mejor predicci on de dichos resultados en el contexto del Modelo Est andar requiere tener en
cuenta las correcciones radiativas.
inv
l
=
l
(Z
l
l
)
(Z e
+
e
)
=
N
(Z
e
e
)
(Z e
+
e
([v
e
[
2
+[a
e
[
2
)
[v
e
[
2
+[a
e
[
2
=
2N
(1 + 4 sin
2
W
)
2
+ 1
_
5.865 N
= 3
7.819 N
= 4
, (1.585)
mientras que el valor medido experimentalmente para esta cantidad 5.942(16) [?], es una evidencia
muy fuerte de que s olo exiten tres neutrinos livianos.
1.17.1 Decaimientos debiles mediados por corrientes cargadas
De la corrientes cargadas para leptones tenemos
L
cc

g
2
2
_
l

l
(1
5
)lW
+
+

l
(1
5
)
l
W
_
(1.586)
Esto da lugar a los posibles diagramas para decaimientos de leptones a bosones virtuales, y bosones
a leptontes mostrados en la gura 1.11. Las echas representan el ujo de n umero lept onico. La
echa de tiempo es de izquierda a derecha. Al lado izquierdo del vertice entran partculas y salen
antipartculas. Mientras que al lado derecho entran antip artculas y salen partculas Del primer y
cuarto diagrama obtenemos el diagrama de Feynman para el decaimiento

e
, mostrado
en la gura 1.12 El propagador para el bos on W de momentum q resulta ser
=
1
q
2
m
2
W
_
g
m
2
W
_
. (1.587)
Para los prop ositos actuales la obtencion de este resultado no es necesaria, el punto importante es
que cuando las masas de las partculas iniciales y nales son mucho mas peque nas que m
W
, esto se
reduce a
=
g
m
2
W
. (1.588)
1.17. FENOMENOLOG
IA ELECTROD
EBIL 117
Figure 1.11: Diagramas de Feynman para las corrientes cargadas
Figure 1.12: diagrama de Feynman para el decaimiento

e
Este resultado se entiende f acilmente cuando se compara con el propagador de una partculas escalar
masiva 1/(q
2
M
2
) 1/M
2
. Las componentes espaciales de W
con = 1, 2, 3, a bajas energas

tienen el mismo propagador que el de una partcula escalar, mientras W
0
, tiene el signo opuesto.
El Lagrangiano efectivo para el decaimiento del muon,

e
es entonces
L =
g
2
8
[
(1
5
)]
g
m
2
W
[ e
(1
5
)
e
]
=
g
2
8m
2
W
[
(1
5
)] [ e
(1
5
)
e
]
=
G
F
2
[
(1
5
)] [ e
(1
5
)
e
] , (1.589)
donde
G
F
2
=
g
2
8m
2
W
=
g
2
4
8g
2
v
2
=
1
2v
2
, (1.590)
y
v =
_
2G
F
_
1/2
. (1.591)
De otro lado, para el decaimiento , n pe

e
, de acuerdo a la gura 1.13, tenemos
L =
G
2
[ p
(1 1.26
5
)n] [ e
(1
5
)
e
] . (1.592)
con G
F
dado en la ec. (1.573) y G
= 1.10 10
5
GeV
2
. La corriente hadr onica tiene la forma
V1.26A. El factor 1.26 puede entenderse como debido a las correcciones a nivel hadr onico de una
corriente que es de la forma VA a nivel del quarks, como en la ec. (1.572). A nivel de quarks el
decaimiento del neutron (udd) al prot on (uud) corresponde al decaimiento de uno de los quarks down
del neutron d ue

e
L =
G
F
2
V
11
[ u
(1
5
)d] [ e
(1
5
)
e
] . (1.593)
1.18. C
ALCULO DE PROCESOS 119

Figure 1.13: Decaimiento del neutr on.
De modo que G
= G
F
V
11
= G
F
cos
C
, donde
C
es el angulo de Cabbibo. Una vez se tienen en
cuenta correcciones electrodebiles se obtiene el valor [V
11
[ = 0.97418(27)[?]. Las magnitudes de los
elementos de la matriz CKM son[?]
V
_
_
0.97419 0.2257 0.0359
0.2256 0.97334 0.0415
0.00874 0.0407 0.999133
_
_
1 (1.594)
1.18 Calculo de procesos
Chapter 2
Computational QFT
There are several tools which allows for the generation of models of particle physics models like
LanHEP [18]
http://theory.sinp.msu.ru/
~
semenov/lanhep.html,
or FeynRules [19]
http://feynrules.phys.ucl.ac.be/ .
This kind of programs are able to generate the output required for other programs which make
the calculation of Feynman diagrams and integration over multi-particle phase space. CalcHEP:
http://theory.sinp.msu.ru/
~
pukhov/calchep.html
for example, is able to calculate cross section and decays widths at tree level.
In this chapter we will illustrate the use LanHEP+CalcHEP
2.1 LanHEP
After download the source code from the http://theory.sinp.msu.ru/
~
pukhov/calchep.html to
some DIR,
Note that the tar.gz le name depends on the current version. At the moment of this writing
this was lhep311.tar.gz. To directly download this le use:
$ wget http://theory.sinp.msu.ru/~semenov/lhep311.tar.gz
121
122 CHAPTER 2. COMPUTATIONAL QFT
where \$ is to indicate that the command is to be written in the shell of your Linux session
1
.
To uncompress the le:
$ tar -zxvf DIR/lhep311.tar.gz
Go to the created directory
$ cd lanhep311
To compile and create the executable of the program (lhep):
$ make
The input of LanHEP are les were the Lagrangian of some model is written in a symbolic way.
Then, the LanHEP executable process the input les and generates four outputles which are the
input for the CalcHEP program. For example, in the LanHEP dir
$ ./lhep stand.mdl
Here ./lhep command, search for the le in the defaul directory mdl/stand.mdl. If there are no erros
printed, for les are created:
ls *4.mdl
func4.mdl lgrng4.mdl prtcls4.mdl vars4.mdl
2.2 CalcHEP
The installation of CalcHEP is similar. In Ubuntu you must be sure to have libx11-dev package, in
addion to the C and Fortran compilers:
$ sudo apt-get install libx11-dev build-essential gfortran
In the CalcHEP directory:
$ make
To use CalcHEP you must rst create a directory with the required les. This is achieved with
the CalcHEP command
1
An introduction to scientic computing is at http://gfif.udea.edu.co/cf
2.3. LANHEP/CALCHEP 123
$ ./mkUsrDir YourModel
A directory YourModel is created with several les and directories inside. By default, a models
directory is created with two set of .mdl les, corresponding to two versions of the Standard Model:
$ ls YourModel/models/
From the YourModel directory in CalcHEP, run the command
./calchep
A new window must appear with the info of CalcHEP and the loaded models in YourModel/models.
To navigate through this window, use the arrows keys and the <ESC> key to navigate back into the
menus.
2.3 LanHEP/CalcHEP
The sample .mdl les in the mdl directory of LanHEP must be modied in order to generate the
proper CalcHEP input les. From the LanHEP directory
$ mkdir sm
$ cd sm
$ wget --no-check-certificate \
https://github.com/rescolo/LanHEP/raw/release/sm/sm.mdl
$ wget --no-check-certificate \
https://github.com/rescolo/LanHEP/raw/release/sm/sm_tex.mdl
$ ../lhep sm.mdl
The four CalcHEP input les:
are then created.
From CalcHEP directoty:
$ ./mkUsrDir sm
$ cd sm/models
$ rm *
then copy the *1.mdl les to the sm/models, and from the sm CalcHEP directory run ./calchep.
In order to understand the structure of the LanHEP les consider the following skeleton:
1 model MODEL NAME/N.
2 % The coments are either this way
3 /* or this other way */
4
5 use file_tex.
6
7 prtcprop pdg.
8
9 prtcformat fullname: Full Name ,
10 name: P ,
11 aname: aP ,
12 pdg: number ,
13 spin2, mass, width, color, aux,
14 texname:> LaTeX P name <,
15 atexname:> LaTeX aP name < .
16
17 parameter VAR = VALUE : Description.
18
19 particle_type
20 particle/Antiparticle: (name, property name=VALUE, ...).
21
22 lterm Write here the Lagrangian in a LaTeX--like format
23
24 prtcprop pdg:(Particle=PDF code,...).
25
26 SetEM(A,EE). %check charge conservation
27 CheckHerm.
In line 1, N is an integer that will identify the four output les. The le in line 5 will contain the
L
A
T
E
X denitions of the used particles. In lines 7-15, the format of the table prtclN.mdl, as required
by CalcHEP, is dened: A new column with the PDG number for the particle. In line 17, the general
form to declarate a variable is established, while the lines 19-20 are the generic declaration for a
particle. The nal commands in 26 and 27 is to check the consistency of the dened model. As a
simple illustration consider the simple case of QED:
model QED: e, mu tau/1.
use qed3g_tex.
prtcprop pdg.
%prtc1.mdl is one of the output files of LanHEP. To make
% it compatible with CalcHEP we need to change their format
% to include the PDG particle number in the third column
prtcformat fullname: Full Name ,
name: P ,
aname: aP ,
pdg: number ,
spin2, mass, width, color, aux,
texname:> LaTeX P name <,
atexname:> LaTeX aP name < .
parameter EE = 0.31333 : Electromagnetic coupling constant (<->1/128).
vector
A/A: (photon, gauge).
spinor e1:(electron),
e2:(muon, mass Mm = 0.1057),
e3:(tau-lepton, mass Mt = 1.777).
% fermion interaction with gauge fields
lterm anti(psi)*gamma*(i*deriv - EE*A)*psi
where
psi=e1;
psi=e2;
psi=e3.
% gauge bosons Lagrangian
lterm -F**2/4 where
F=deriv^mu*A^nu-deriv^nu*A^mu.
%set PDG particle numbers:
prtcprop pdg:(A=22,e1=11, e2=13, e3=15).
SetEM(A,EE). %check charge conservation
CheckHerm.
where the required le qed3g_tex.mdl is
SetTexName([e1=e,E1=\\bar{e}]).
SetTexName([e=e,E=\\bar{e}]).
SetTexName([e1.c=e^c,E1.c=\\bar{e}^c]).
SetTexName([e.c=e^c,E.c=\\bar{e}^c]).
SetTexName([e2=\\mu,E2=\\bar{\\mu}]).
SetTexName([e3=\\tau,E3=\\bar{\\tau}]).
SetTexName([m=\\mu,M=\\bar{\\mu}]).
SetTexName([l=\\tau,L=\\bar{\\tau}]).
SetTexName([EE=e]).
SetTexName([Me=M_e, Mm=M_\\mu, Mt=M_\\tau]).
Running with the option -tex:
../lhep qed3g.mdl
the following output is generated
lgrng1.tex
Fields in the vertex Variational derivative of Lagrangian by elds
e
a
e
b
A
ab

a

b
A
ab

a

b
A
ab
prtcls1.tex:
P aP Name Spin EM charge Color Comment
A
photon 1 0 1 gauge
e
a
e
a
electron 1/2 1 1
a

a
muon 1/2 1 1
a

a
tau-lepton 1/2 1 1
vars1.tex
Parameter Value Comment
EE 0.31333 Electromagnetic coupling constant (1/128)
Mm 0.1057 mass of muon
Mt 1.777 mass of tau-lepton
Figure 2.1: CalcHEP welcome window
With the command
$ ../lhep qed3g.mdl
the same les are generated by in the format of CalcHEP.
parameter is for constants to be exported to tables, while let is only for internal LanHEP variables.
In CalcHEP
\$ ./mkUsrDir qed3g
\$ cd qed3g/models
\$ rm *
#copy the files: func1.mdl, lgrng1.mdl prtcls1.mdl, vars1.mdl here
\$ cd ..
\$ ./calchep
The window in Fig. 2.1 After hit <Enter>, the window with the model should appears as shown in
Fig. 2.2 To test that the model was loaded without errors:
QED: e, mu tau -> Edit Model -> Check Model
A message with The model is OK, should popup.
After returning to the model window in Fig. 2.2, we could calculate some process:
QED: e, mu tau -> Enter Process
and enter the process e1,E1 -> e2,E2 (e
+
e
) as shown in the Fig: 2.3 After <Enter>, the

window to calculate the process should appears, as in Fig. 2.4 In addition to View diagrams, we can
calculate the process with the sequence
Figure 2.2: CalcHEP model window
Figure 2.3: CalcHEP process window
Figure 2.4: CalcHEP calculation window
Squaring technique -> Symbolic calculations -> C-compiler
Then a new window with the process details should appears, as displayed in Fig. 2.5 After adjust
the input parameters at your convenience, we could just calculate the process with, in this case:
Easy 2-2, to obtain the result displayed in 2.6 e.g, for center of mass energy of 14 TeV (7 TeV per
beam) we could have:
(e
+
e
) = 5 10
4
pb (2.1)
Exercise: Repeat the previous calculation, but for one center of mass energy of 200 GeV.
For the Standard Model the Yukawa Lagrangian that couple the down fermions with the boson
scalar is written in the interaction basis:
L
Y
D
M
D
P
R
D
H + h.c , (2.2)
with D
= V
D, we can write the eq. (2.2) in the mass eigenstates as

L
Y
DM
dia
D
P
R
DH + h.c (2.3)
where
M
dia
D
= V M
D
V
(2.4)
Investing the equation 2.4 and replacing in (2.2) we can write in the interaction eigenstates:
L
Y
(D
)(M
dia
D
V )P
R
D
H + h.c (2.5)
Expanding we get . . . which is just the expresion in the Standard Model le
Figure 2.5: CalcHEP calculation window
Figure 2.6: CalcHEP results window
Chapter 3
Second quantization
Two key ingredients to formulate the Quantum Field Theory (QFT) are the quantization of systems
in which the particles can be created and destroyed (quamtum theory of radiation) and the behavior
of relativistic systems. When both ingredients are present the particles can be understood as the
excited modes of certain eld. When the particles in a system are not relativistic, the formalism of
creation and annihilation operators is just an alternative method to describe the Hamiltonian of the
Schr odinger equation. In relativistic systems however, the existence of negative energy states force
the construction of new quantum states, the Fock states, in order to have proper dened probabilities
for the states of the system. In section xx we start by building the Fock states associated to a massless
not relativistic scalar eld. Then we generalize the results to a massive scalar eld satisfying the
Klein-Gordon equation.
Some parts of the discussion were based in some topics of chapters 4-6 of [2].
3.1 Quantization of the nonrelativistic string
3.1.1 The clasical string
In conventional quantization the energy of one state is interpreted as the possible eigenstates of an
Hamiltonian operator acting on the states of the system.
H[State = E[State (3.1)

One step further is to consider the wave function as the eigenstate of the operatoreld acting on
certain Fock states
[Fock State = [[Fock State , (3.2)

131
132 CHAPTER 3. SECOND QUANTIZATION
Like that usual quantum mechanical observable, the wave function will have an uncertainty. The
Fock states are the states under which the classical wave function can be obtained with a small
uncertainty
= Fock State[
[Fock State (3.3)

This happens when the number of quanta of the Fock state is big enough. In fact, a state with a
denite number of quanta has a innity uncertainty [15].
Eq. (3.2) is the basis for the calculation of cross section and decay widths in quantum eld theory.
Now we will study how to dene a such Fock state for a scalar eld.
We have already see in Chapter 1 of [1] that a string have a collective wave motion that is
described by a continuous eld, which satises the familiar one-dimensional wave equation
1
v
2
t
2

2
z
2
= 0 (3.4)
This equation can be derived following two dierent paths. The rst is to decomposing the string
into individual oscillators for which the usual Lagrangian formalism can be used. The second is just
by formulating certain Lagrangian density from which the equation of motion can be obtained by
using the Euler-Lagrange equation
_
L
(
)
_
= 0 . (3.5)
In the rst approach the string is considered to be composed of N oscillators coupled together by
springs with a spring constant k. At certain time t, the displacement of the oscillator i at time t
is represented by
i
(t). In Table 3.1 it is displayed the corresponding macroscopic quantities. Note
also that 1/v
2
= /T. It is worth to stress that at the Lagrangian level, which is the sum of each
micro macro
l L = Nl
m = m/l
k T = kl
i
(t) = (z
i
, t) (z, t)
Table 3.1: From micro to macro
individual oscillator Lagrangian, it is the sum of the kinetic and potential oscillator energy. However,
the Lagrangian density only have the kinetic term for the scalar eld
L =
1
2
_
1
v
2
0
+
3
vc=1
1
2
. (3.6)
3.1. QUANTIZATION OF THE NONRELATIVISTIC STRING 133
Note that only in the case v = c this Lagrangian can be written in a covariant form. Moreover, the
scalar eld (z, t) have nothing to do with the individual oscillators. An specic solution for (z, t)
would represent one specic oscillation mode of the string. It turn out that this specic frequency
mode corresponds to an particle state, that does have not connection with the physical particles in
the string.
The most general discrete solution to the wave equation (3.4) is the Fourier decomposition
(t, z) =
n
v
2
n
L
_
a
n
e
i(ntknz)
+ a
n
e
i(ntknz)
_
(3.7)
where the dispersion relation is
2
n
= v
2
k
2
n
(3.8)
where
n
is denite positive:
n
= +[v[
_
[k
n
[ (3.9)
To satisfy the boundary conditions we must have
k
n
=
2n
L
(3.10)
Note that
k
n
= k
n
. (3.11)
Therefore
n
=
n
. (3.12)
In three dimensions and with v = c = 1, the Lagrangian can be written as
L =
1
2
(3.13)
This Lagrangian is still covariant after the addition of a function of . An interesting case is just the
addition of the mass term the most general solution to the KleinGordon equation is
L =
1
2

1
2
m
2
2
(3.14)
which give to arise to Klein-Gordon
_
m
2
_
= 0 . (3.15)
We now will check the origin of the normalization factor. For simplicity we work with one spatial
dimension. By using eq. (3.7)
(z, t) =
n=
v
2
n
[a
n
n
(z, t) + a
n
(z, t)] , (3.16)
[E] =
1
[E]
1/2
[E]
1
[a]
=E
1/2
[a] (3.17)
[a] = [E]
1/2
(3.18)
we dene
n
(z, t) =
1
L
e
i(ntknz)
(3.19)
y las funciones
n
satisfacen las siguientes condiciones de normalizaci on
_
L
0
dz
n
(z, t)
m
(z, t) =
1
L
_
L
0
dz e
i(ntknz)
e
i(mtkmz)
=
1
L
_
L
0
dz expi[(
n
m
)t (k
n
k
m
)z]
=
e
i(nm)t
L
_
L
0
dz e
i(knkm)z
(3.20)
When n = m
_
L
0
dz
n
(z, t)
m
(z, t) =
1
L
_
L
0
dz
=1 (3.21)
For n ,= m, 2(n m) is an even integer and then
_
L
0
dz
n
(z, t)
m
(z, t) =
e
i(nm)t
L
e
i(knkm)z
i(k
n
k
m
)
L
0
=
e
i(nm)t
L
1
i(k
n
k
m
)
_
e
i2(nm)
1
_
=0 (3.22)
In this way
_
L
0
dz
n
(z, t)
m
(z, t) =
nm
. (3.23)
Moreover
_
L
0
dz
n
(z, t)
m
(z, t) =
n,m
e
2int
. (3.24)
En tal caso de
H =
_
L
0
1dz . (3.25)
From the analysis of the Theorem of Noether in chapter 1 of [1] we have, that in a similar way to the
usual Lagrangian formulation, where the canonical conjugate variable is used to dene the Legendre
transformation
H = p q L, (3.26)
the Hamiltonian density can be obtained from the Lagragian density trough the Legendre transfor-
mation
1 = T
0
0
=
L
L (3.27)
= (x)
(x)
t
L. (3.28)
where
(x) =
L
((x)/t)
(3.29)
is the canonical conjugate variable (conjugate momentum) of the canonical variable (x).
We have then,
H =
1
2v
2
_
L
0
dz

t
t
+
1
2
_
L
0
dz

z
z
=
n=
n
a
n
a
n
(3.30)
t
=
n=
v
2
n
[i
n
a
n
n
(z, t) + i
n
a
n
(z, t)] ,
=
n=
iv
n
2
n
[a
n
n
(z, t) a
n
(z, t)] , (3.31)
1
v
2
t
=
n,m=
m
2
m
[a
n
n
(z, t) a
n
(z, t)] [a
m
m
(z, t) a
m
(z, t)] (3.32)
=
n,m=
m
2
m
[a
n
a
m
m
a
n
a
m
m
a
n
a
m
+ a
n
a
m
] (3.33)
z
=
n=
v
2
n
[ik
n
a
n
n
(z, t) ik
n
a
n
(z, t)] ,
=
n=
ivk
n
2
n
[a
n
n
(z, t) a
n
(z, t)] , (3.34)
z
=
n,m=
v
2
k
n
k
m
2
m
[a
n
n
(z, t) a
n
(z, t)] [a
m
m
(z, t) a
m
(z, t)] (3.35)
=
n,m=
v
2
k
n
k
m
2
m
[a
n
a
m
m
a
n
a
m
m
a
n
a
m
+ a
n
a
m
] (3.36)
Using eqs. (3.23), and (3.24)
H =
1
2
n,m=
_
L
0
dz

n
m
2
m
[a
n
a
m
m
a
n
a
m
m
a
n
a
m
+ a
n
a
m
]
+
1
2
n,m=
_
L
0
dz
v
2
k
n
k
m
2
m
[a
n
a
m
m
a
n
a
m
m
a
n
a
m
+ a
n
a
m
]
=
1
2
n,m=
m
2
m
_
a
n
a
m
n,m
e
2int
a
n
a
m
n,m
a
n
a
n,m
+ a
n
a
n,m
e
2int
+
1
2
n,m=
v
2
k
n
k
m
2
m
_
a
n
a
m
n,m
e
2int
a
n
a
m
n,m
a
n
a
n,m
+ a
n
a
n,m
e
2int
=
1
2
n=
_

n
n
2
n
a
n
a
n
e
2int
+

n
n
2
n
(a
n
a
n
+ a
n
a
n
)

n
n
2
n
a
n
a
n
e
2int
_
+
1
2
n=
_
v
2
k
n
k
n
2
n
a
n
a
n
e
2int
+
k
n
k
n
2
n
(a
n
a
n
+ a
n
a
n
)
k
n
k
n
2
n
a
n
a
n
e
2int
_
(3.37)
Since
n
=
n
and k
n
= k
n
H =
1
2
n=
1
2
n
_
(
2
n
+ v
2
k
2
n
)a
n
a
n
e
2int
+ (
2
n
+ v
2
k
2
n
)(a
n
a
n
+ a
n
a
n
)
+ (
2
n
+ v
2
k
2
n
)a
n
a
n
e
2int
(3.38)
Finally, using eq. (3.8)
H =
1
2
n=
n
(a
n
a
n
+ a
n
a
n
) (3.39)
Since a
n
and a
n
are classical quantities that commutates, the Hamiltonian is
H =
n=
n
a
n
a
n
=
n=
n
[a
n
[
2
(3.40)
In this way, the factor
2
n
in eq. (3.16), is a convenient choice of normalization for the coecients
a
n
which guarantees the Hamiltonian.
To quantize the string, we need to promote H to an operator. In canonical quantization we need
to identify the proper conjugates variables. For this purpose it is convenient to write eq. (3.40) as
the Hamiltonian of an harmonic oscillator.
3.1.2 Quantization of the string
This Hamiltonian can be rewritten as a sum of independent oscillators Hamiltonians. Consider an
harmonic oscillator of frequency . The equation of motion for F = kq is
q +
k
m
q =0
q +
2
q =0 (3.41)
with
2
=
k
m
(3.42)
This equation of motion can be obtained from the Lagrangian
L = T V =
1
2
[m q
2
kq
2
] (3.43)
And the Hamiltonian can be obtained from eq. (3.26)
H =p q L
=
p
2
m

1
2
p
2
m
+ kq
2
=
1
2
_
p
2
m
+ m
2
q
2
_
=
1
2m
_
p
2
+ m
2
2
q
2
_
(3.44)
For a set of independent oscillators we have
H =
n
1
2m
_
p
2
n
+ m
2
2
n
q
2
n
_
H =
n
_
1
2m
n
p
2
n
+
m
n
2
q
2
n
_
(3.45)
Comparing eq. (3.45) with Eq. (3.40) we see that the complex number a
n
can be written as ( = 1)
a
n
=c
1
q
n
+ ic
2
p
n
(3.46)
a
n
a
n
= c
2
1
q
2
n
+ c
2
2
p
2
n
(3.47)
c
1
=
m
n
2
=
m
n
2m
n
c
2
=
1
2m
n
(3.48)
a
n
=
m
n
q
n
+ i p
n
2m
n
a
n
=
m
n
q
n
i p
n
2m
n
(3.49)
a
n
+ a
n
=2
m
n
2m
n
q
n
=
2m
n
q
n
q
n
=
1
2m
n
(a
n
+ a
n
)
a
n
a
n
=
2i
2m
n
p
n
=
i
2m
n
m
n
p
n
p
n
=
im
n
2m
n
(a
n
a
n
) (3.50)
In quantum mechanics the classical objects q
n
and p
n
are promoted to operators which satisfy the
commutation relation
[ q
n
, p
m
] =i
mn
[ q
n
, q
m
] = [ p
n
, p
m
] =0 . (3.51)
This implies that the objects a
n
and a
n
, are also operators
[ q
n
, p
m
] =
im
m
2m
n
2m
m
[a
n
, a
m
] [a
n
, a
m
] + [a
n
, a
m
] + [a
n
, a
m
]
[ q
n
, p
m
] =
im
m
2
m
n
m
m
[a
n
, a
m
] 2[a
n
, a
m
] + [a
n
, a
m
] (3.52)
If the operators a
n
and a
n
satisfy the commutation relations
_
a
n
, a
=
n,m
[a
n
, a
m
] =
_
a
n
, a
= 0 , (3.53)
then we recover equations (3.51). The scalar eld is now an operator
n=
v
2
n
L
_
a
n
e
i(ntknz)
+ a
n
e
i(ntknz)
_
, (3.54)
In terms of operators a
n
and a
n
the Hamiltonian from eq. (3.39) can be written as
H =
1
2
n=
n
( a
n
a
n
+a
n
a
n
)
=
1
2
n=
n
_
2 a
n
a
n
+
_
a
n
, a
n
__
=
n=
n
_
a
n
a
n
+
1
2
_
(3.55)
Since
n=
n
1
2
, (3.56)
it is convenient to renormalize the Hamiltonian as
:

H: =
n=
n
_
a
n
a
n
+
1
2
_
n=
1
2
=
n=
n
a
n
a
n
(3.57)
This procedure is consistent since the related physics quantities arise from energy dierences, no
from absolute energy determinations.
_
H, a
m
_
=
n=
n
__
a
n
a
n
+
1
2
_
, a
m
_
=
n=
n
_
a
n
a
n
, a
m
_
(3.58)
By using the identity
[AB, C] = [A, C] B + A[B, C] (3.59)
we have
_
H, a
m
_
=
n=
n
__
a
n
, a
m
_
a
n
+ a
n
[a
n
, a
m
]
_
=
n=
nm
a
n
=
m
a
m
(3.60)
_
H, a
m
_
=
n=
n
__
a
n
, a
m
_
a
n
+ a
n
_
a
n
, a
m
__
=
n=
n
a
nm
=
m
a
n
(3.61)
If [m
n
is an eigenstate of

H with eigenvalue E
n
H[m
n
= E
n
[m
n
(3.62)
then
Ha
n
[m
n
=
_
a
n
H
n
a
n
_
[m
n
=(E
n
n
) a
n
[m
n
(3.63)
a
n
[m
n
is also an eigenstate with eigenvalue E
n
n
. Moreover,
H a
n
[m
n
=
_
a
H +
n
a
n
_
[m
n
=(E
n
+
n
) a
n
[m
n
(3.64)
a
n
[m
n
is also an eigenstate with eigenvalue E
n
+
n
.
As stablished in [4]
In other words, the operator a
n
seems to annihilate a quantum of energy, of amount
n
,
from the state. On the other hand, a
n
creates a quantum of energy
n
. In this sense,
they are the are the annihilation and the creation operators, respectively. [...]
The ground state can be denoted by [0 = [0
n
. Since this is state of lowest energy, the
annihilation operator a
n
, acting on it, cannot produce a state of lower energy. Thus, this
state must be totally annihilated by the operation of a
n
:
a
n
[0 =0
0[ a
n
=0 , (3.65)
such that
0[0 = 1 (3.66)
The energy if the ground state can be xed to zero:
:

H : [0 = 0 . (3.67)
We dene the state whose energy is larger tha the energy of [0 by one quantum
n
by
[1
n
a
n
[0
1
n
[ =0[a
n
(3.68)
[1
n
is an Hamiltonian eigenstate of energy
n
:
:

H : [1
n
=
n
a
n
a
n
[1
n
=
n
[1
n
=
n
1[1
n
, (3.69)
where we have made explicit that we have a quantum of energy . The normalized state is
1
n
[1
n
=0[a
n
a
n
[0
=0[
_
a
n
, a
n
_
[0
=0[0
=1 . (3.70)
Similarly, the state with energy 2 is
1
2
_
a
n
_
2
[0 =[2
n
0[
1
2
(a
n
)
2
=2
n
[ (3.71)
with normalization
2
n
[2
n
=
1
2
0[a
n
a
n
a
n
a
n
[0
=
1
2
1
n
[a
n
a
n
[1
n
=
1
2
_
1
n
[
_
a
n
, a
n
_
+ a
n
a
n
[1
n
_
=
1
2
(1
n
[1
n
+0[0)
=1 . (3.72)
By induction we get
1
m!
_
a
n
_
m
[0 =[m
n
(3.73)
From here we have
1
m!
a
n
_
a
n
_
m1
[0 =[m
n
_
(m1)!
m!
1
_
(m1)!
a
n
_
a
n
_
m1
[0 =[m
n
(m1)!
m(m1)!
a
n
[(m1)
n
=[m
n
_
1
m
a
n
[(m1)
n
=[m
n
n
[(m1)
n
=
m[m
n
n
[m
n
=
m + 1[(m + 1)
n
(3.74)
or,
m
n
[ a
n
=
m + 1(m + 1)
n
[ (3.75)
From this expressions we can check that number operator can be dened from:
m
n
[ a
n
a
n
[m
n
=(m + 1)(m + 1)
n
[(m + 1)
n
m
n
[1 + a
n
a
n
[m
n
=(m + 1)
m
n
[1 +

A
n
[m
n
=(m + 1) (3.76)
In this way, the number operator as
A
n
= a
n
a
n
(3.77)
If

A
n
[m
n
= c[m
n
, where c must be real because

A
n
is Hermitian
1 + c = m + 1 (3.78)
and
A
n
[m
n
= m
n
[m
n
(3.79)
From here, we can calculate the eigenvalues of a
n
. Since
A
n
a
n
=
_

A
n
, a
n
_
+a
n

A
n
=
_
a
n
, a
n
_
a
n
+a
n

A
n
+ a
n
[a
n
, a
n
]
=a
n
+a
n

A
n
=a
n
_

A
n
1
_
(3.80)
A
n
a
n
[m
n
=(m
n
1)a
n
[m
n
(3.81)
Since the state
A
n
[m
n
1 =(m
n
1)[m
n
1 (3.82)
has the same eigenvalue, therefore
a
n
[m = C
[m
n
1 (3.83)
where C
is a number to be determined from the normalization condition

m
n
[ a
n
a
n
[m
n
=[C
[
2
m
n
1[m
n
1
m
n
[
A
n
[m
n
=[C
[
2
[C
[
2
= m
n
(3.84)
a
n
[m
n
=
m
n
[m
n
1 (3.85)
such that
m
n
[m
n
= 1 (3.86)
Eq. (3.57) can be rewritten as
:

H: =
n=
n

A
n
(3.87)
Noting also that
A
n
[m
l
= 0 n ,= l , (3.88)
we have that
m
n
[ :

H: [m
n
= m
n
n
= m
n
n
. (3.89)
Therefore, once we have proper normalized states and renormalized Hamiltonian, the energy of an
state with m quantum ( of frequency
n
) is just m times the energy of the one quanta of energy
n
.
Note that
0[ :

H: [0 = 0 . (3.90)
The general procedure to redene the zero of energy such that the vacuum energy vanishes is called
normal ordering. We dene a normal-ordered product by moving all annihilation operators to the
right of all creation operators. For an operator

X, its normal-ordered product will be denoted as
:

X: . Using this algorithm on the expression of eq. (3.55), we nd that
:

H: =
1
2
n=
n
: ( a
n
a
n
+a
n
a
n
):
=
1
2
n=
n
( a
n
a
n
+ a
n
a
n
)
=
n=
n
a
n
a
n
(3.91)
From [16] (pag. 121):
These idea carry over to quantum eld theory, but with a dierent interpretation. In
quantum mechanics we are talking about a single particle state [m
n
and energy levels
E
n
= (n +1/2). The creation and annihilation operators move the state of the particle
up and down in energy from the ground.
In quantum eld theory, we take the notion of number operator literally. The state [n
is not a state of a single particle, rather is an state of the eld with N particles present.
The background state which is also the lowest energy state is a state of the eld with 0
particles (but the eld is still there). The creation operator a
n
adds a single quantum
(a particle) to the eld, while the annihilation operator a
n
destroys a single quantum
(removes a single particle) from the eld. As we will see, in general there will be creation
operators and annihilation operators for particles as well as for antiparticles.
These operators will be functions of momentum. The elds will become operators which
will be written as sums over annihilation and creation operators.
3.1.3 Generalization to three dimensions
Taking into account that E
n
=
n
=
n
, when = 1, the most general solution to the generalization
to three dimensions of the wave equation with velocity of propagation c = 1
= 0 , (3.92)
obtained from the three dimension Lagrangian
L =
1
2
, (3.93)
is
(t, x) =
n
1
2E
n
L
3
_
a
n
e
ipnx
+ a
n
e
ipnx
_
,
=
nx,ny,nz
1
_
2E
(nx,ny,nz)
L
3
_
a
(nx,ny,nz)
expi[E
(nx,ny,nz)
t p
x
x p
y
y p
z
z]
+a
(nx,ny,nz)
expi[E
(nx,ny,nz)
t p
x
x p
y
y p
z
z]
_
, (3.94)
where in natural units the wave number can be identied with the momentum, p = k. In eq. (3.94)
E
n
=p
0
n
p
i
=
2
L
n
i
(3.95)
where p
0
= E
n
, and the solution satises the dispersion relation
p
2
n
= p
2
n
= c
2
E
n
= E
2
n
. (3.96)
The Energy will always be chosen to be positive
E
n
=
2
L
_
n
2
x
+ n
2
y
+ n
2
z
(3.97)
Since the Action is dimensionless,
S =
_
d
4
x m
2
2
[1] = [E]
4
[E]
2
[]
2
[] =([S]/[E]
2
)
1/2
= [E] , (3.98)
3.2. QUANTIZATION OF THE KLEIN-GORDON FIELD 147
this solution must have units of energy in natural units. To obtain the dimensions of a
n
, we just
check the dimensions in both sides of eq. (3.94)
[E] =
1
_
[E][E]
3
[a
n
]
=[E][a
n
] , (3.99)
and therefore a
n
is dimensionless.
The canonical quantization in eqs. (3.53) can be generalized to
_
a
n
, a
=
n,m
[a
n
, a
m
] =
_
a
n
, a
= 0 , (3.100)
3.2 Quantization of the Klein-Gordon eld
It is convenient to put the system into a box of size L, so that the total volume is nite. According
eq. (3.10), in this case the frequency is discret. However particles like the photon or electron have
frequencies in a continuum range. Therefore we need to establish relations that allows extrapolate
the discrete results into the continuum, and also we will need to take the limit of innite volume.
The Klein-Gordon equation for a real scalar eld (Chapter 3. [1])
(
+ m
2
) = 0 , (3.101)
can be obtained from the Lagrangian
L =
1
2

1
2
m
2
2
, (3.102)
The solution is the same that for the case m = 0 in eq. (3.94), but the new dispersion relation is
E
2
n
= p
2
n
+ m
2
. (3.103)
and therefore m can be interpreted as the mass of eld .
We assume that can have frequencies in the continuum. In this way the most general solution
is obtained after replacing the summatory by an integral
_
dp
n
p =
n
p
n+1
p
n
=
2
L
n
n + 1 n =
2
L
n
(3.104)
_
L
2
_
3
_
d
3
p (3.105)
From
_
d
3
p
(3)
(p q) =1 (3.106)
and taking into account that
n,m
=
n,m
=1 , (3.107)
where
p
i
=
2
L
n
i
q
i
=
2
L
m
i
, (3.108)
we have
_
d
3
p
(3)
(p q) =
nm
n
_
2
L
_
3
(3)
(p q) =
n,m
_
2
L
_
3
(3)
(p q) =
n,m
. (3.109)
In this way
(3)
(p q) =
_
L
2
_
3
p,q
, (3.110)
and we get that in the continuum limit
_
L
2
_
3
n,m

(3)
(p q) (3.111)
In particular, this implies that
(2)
3
(3)
(p = 0) L
3
= V (3.112)
3
(0) =
V
(2)
3
. (3.113)
This expression can be also obtained from the denition
3
(p) = lim
V
_
1
(2)
3
_
V
d
3
x e
ipx
_
, (3.114)
before taking the limit to innity.
Therefore, in the continuum the solution in eq. (3.94) can be written as
(t, x) =
_
L
2
_
3
_
d
3
p
1
_
2E
p
L
3
_
a
p
e
ipx
+ a
p
e
ipx
_
=
_
d
3
p
L
3
(2)
3
_
2E
p
_
a
p
e
ipx
+ a
p
e
ipx
_
(3.115)
Using eq. (3.111), we can write the commutation relations (3.53) in the continuum as
_
a
p
, a
=
_
2
L
_
3
(3)
(p q) [a
p
, a
q
] =
_
a
p
, a
= 0 . (3.116)
Note that again a
p
is dimensionless. It is customary to write the general solution (3.115) with
a
p
=
L
3
a
p
. (3.117)
Then
(t, x) =
_
d
3
p
1
(2)
3
_
2E
p
_
a
p
e
ipx
+ a
e
ipx
_
. (3.118)
and the commutation relations in eq. (3.116) can be written as
_
a
p
, a
= (2)
3
(3)
(p q)
_
a
p
, a
=
_
a
p
, a
= 0 . (3.119)
In what follows we will drop out the prime in a
p
.
The basic principle of canonical quantization is to promote the eld and its conjugate momentum
to operators, and to impose the equal time commutation relation
_
(t, x),
(t, y)
_
= i
(3)
(x y)
_
(t, x),
(t, y)
_
=
_
(t, x),
(t, y)
_
= 0 . (3.120)
We will now check that the commutation relations in eq. (3.119) will just generate the equal time
commutation relations in eq. (3.120).
Promoting the real eld to a hermitian operator means to promote a
p
to an operator; thus
(t, x) =
_
d
3
p
1
(2)
3
_
2E
p
_
a
p
e
ipx
+ a
p
e
ipx
_
(3.121)
with
_
a
p
, a
q
_
= (2)
3
(3)
(p q) [a
p
, a
q
] =
_
a
p
, a
q
_
= 0 . (3.122)
The conjugate momentum can be obtained from the Klein-Gordon Lagrangian in eq. (3.102), by
using eq. (3.29)
(x) =

(
0
)
_
1
2
(
0
)
2
_
=
0
=
_
d
3
p
1
(2)
3
_
2E
p
_
iE
p
a
p
e
ipx
+ iE
p
a
p
e
ipx
_
=
_
d
3
p
i
(2)
3
_
E
p
2
_
a
p
e
ipx
+a
p
e
ipx
_
(3.123)
Using the expressions for

, and

, in terms of a
p
, a
p
, the commutation relation (3.120) reads
_
(t, x),
(t, y)
_
=
_
d
3
p
_
d
3
p
i
2(2)
6
E
p
E
p
_
a
p
e
ipx
+a
p
e
ipx
, a
p
e
ip
y
+a
e
ip
y
_
=
_
d
3
p
_
d
3
p
i
2(2)
6
E
p
E
p
__
a
p
e
ipx
+a
p
e
ipx
, a
p
e
ip
y
_
+
_
a
p
e
ipx
+a
p
e
ipx
, a
e
ip
y
__
=
_
d
3
p
_
d
3
p
i
2(2)
6
E
p
E
p
__
a
p
e
ipx
, a
p
e
ip
y
_
+
_
a
p
e
ipx
, a
p
e
ip
y
_
+
_
a
p
e
ipx
, a
e
ip
y
_
+
_
a
p
e
ipx
, a
e
ip
y
__
=
_
d
3
p
_
d
3
p
i
2(2)
6
E
p
E
p
_
e
i(px+p
y)
[a
p
, a
p
]
e
i(pxp
y)
_
a
p
, a
p
+ e
i(pxp
y)
_
a
p
, a
_
+ e
i(px+p
y)
_
a
p
, a
__
. (3.124)
Taking into account the eqs. (3.122), then
_
(t, x),
(t, y)
_
=
_
d
3
p
_
d
3
p
i
2(2)
6
E
p
E
p
_
e
i(pxp
y)
_
a
p
, a
_
e
i(pxp
y)
_
a
p
, a
p
_
=
_
d
3
p
_
d
3
p
i
2(2)
3
E
p
E
p
_
e
i(pxp
y)
(3)
(p p
) + e
i(pxp
y)
(3)
(p
p)
_
=
_
d
3
p
_
d
3
p
i
2(2)
3
E
p
E
p
(3)
(p p
)
_
e
i(pxp
y)
+ e
i(pxp
y)
_
. (3.125)
(3)
(p p
) forces p = p
, which also means E

p
= E
p
, and since x
0
= y
0
= t, we have
_
(t, x),
(t, y)
_
=
_
d
3
p
_
d
3
p
i
2(2)
3
E
p
E
p
(3)
(p p
)
_
e
i[t(EpE
p
)px+p
y]
+ e
i[t(EpE
p
)px+p
y]
_
=
_
d
3
p
i
2(2)
3
_
e
i(px+py)
+ e
i(px+py)
=
_
d
3
p
i
2(2)
3
_
e
ip(xy)
+ e
ip(xy)
. (3.126)
Since
(3)
(x y) =
_
d
3
p
(2)
3
e
ip(xy)
=
(3)
(x +y) =
_
d
3
p
(2)
3
e
ip(x+y)
=
_
d
3
p
(2)
3
e
ip(xy)
, (3.127)
then
_
(t, x),
(t, y)
_
= i
(3)
(x y) . (3.128)
The same expression is obtained for the original eld operator in eq. (3.115) if the commutation
relations (3.116) are used. Moreover eq. (3.128) is covariant [4].
Note that the commutation relations for the real scalar eld in (3.122) are equivalent to that of
a collection of independent harmonic oscillators, with one oscillator for each value of the momentum
p.
Previous equations for the Hamiltonian still holds.
H =
1
2
_
d
3
p E
p
_
a
p
a
p
+a
p
a
p
_
(3.129)
_
H, a
p
_
= E
p
a
p
_
H, a
p
_
= +E
p
a
p
(3.130)
The analogy between the simple harmonic oscillator and the eld is now complete. Therefore a
p
creates the quanta of momentum p of the eld

, while a
p
is the annihilation operator for a eld
quantum with momentum p. From [4]:
What was the positive energy component of the classical eld now annihilates the quan-
tum, and the negative energy component now creates the quantum. This quantum is
what we call particle of positive energy.
We can now construct the Fock space following the standard procedure for the harmonic oscillator:
we interpret a
p
as destruction operators and a
p
as creation operators, and we dene a vacuum state
[0 as the state annihilated by all destruction operators, so for all p
a
p
[0 = 0 . (3.131)
We normalize the vacuum with 0[0 = 1. The vacuum is the state which contains no particles and
no antiparticles either,
The normal ordered Hamiltonian is
:

H: =
_
d
3
p E
p
a
p
a
p
(3.132)
such that, as in discrete case
0[ :

H: [0 = 0 . (3.133)
A possible normalization factor for the Fock one-particle state is ([p [1
p
)
[p =
1
V
a
p
[0
p[ =0[a
p
1
V
(3.134)
This state contains one quantum of the eld

with momenta p
= (E
p
, p). Such states have positive
norm, since
p[p
=
1
V
0[a
p
a
[0
=
1
V
0[a
p
a
a
p
[0
=
1
V
0[[a
p
, a
][0
=
(2)
3
V

(3)
(p p
)
=
_
2
L
_
3
(3)
(p p
) (3.135)
With this normalization, the limit to discrete case is straightforward:
1
n
[1
m
=
n,m
(3.136)
The results are summarized in Table 3.2.
Discret Continuum Continuum a
n
_
L
2
_
3
_
d
3
p
_
L
2
_
3
_
d
3
p
n,m
_
2
L
_
3
(3)
(p q)
_
2
L
_
3
(3)
(p q)
_
(x),
(y)
_
= i
(3)
(x y)
_
(x),
(y)
_
= i
(3)
(x y)
_
(x),
(y)
_
= i
(3)
(x y)
[a
n
, a
m
] =
n,m
[a
p
, a
q
] =
_
2
L
_
3
(3)
(p q) [a
p
, a
q
] = (2)
3
(3)
(p q)
[1
n
= a
n
[0 [p = a
p
[0 [p =
1
V
a
p
[0
1
n
[1
m
=
n,m
p[q =
_
2
L
_
3
(3)
(p q) p[q =
_
2
L
_
3
(3)
(p q)
Table 3.2: From discret to continuos, where p
i
= 2n
i
/L, and q
i
= 2m
i
/L,
Similarly we can dene many particle states. If a state has N particles with all dierent momenta
p
1
, p
2
, . . . , p
N
, it is dened by
[p
1
, . . . , p
N
=
1
V
N/2
a
p
1
a
p
N
[0
p
1
, . . . , 0
p
N
1
V
N/2
a
p
1
a
p
N
[0
(3.137)
On the other hand, if we want to construct a state with m particles of momentum p, we must have
a Fock state similar to (3.73)
[m
p
=
1
V
m/2
1
m!
_
a
p
_
m
[0 (3.138)
From[4]
The vacuum, together with single particles states (3.134) and all multiparticle states
(3.137), (3.138), constitute a vector space which is calles the Fock space. The creation
and annihilation operators act on this space.
It is convinient to dene:
(x) =

+
(x) +

(x) (3.139)
where
+
(x) =
_
d
3
p
1
(2)
3
_
2E
p
a
p
e
ipx
(x) =
_
d
3
p
1
(2)
3
_
2E
p
a
p
e
ipx
. (3.140)
The eect of the operator eld,

(x), on the one particle state [p
(x)[p (3.141)
will be important for the evaluation of Smatrix elements in Chapter 8.
Another choice of normalization is the Lorentz invariant one, to be used later. In this case, the
Fock state of N particles with all dierent momenta p
1
, p
2
, . . . , p
N
, is obtained acting on the vacuum
with the creation operators,
[p
1
, . . . , p
n
(2E
p
1
)
1/2
. . . (2E
pn
)
1/2
a
p
1
. . . a
pn
[0 . (3.142)
The factors (2E
p
1
)
1/2
are a convenient choice of normalization. In particular, the one-particle states
are
[p = (2E
p
)
1/2
a
p
[0 . (3.143)
From the commutations relations and eq. (3.122) we nd that
p[q =(2E
p
)
1/2
(2E
q
)
1/2
0[a
p
a
q
[0
=(2E
p
)
1/2
(2E
q
)
1/2
0[
_
a
p
, a
[0
=(2E
p
)
1/2
(2E
q
)
1/2
(2)
3
(3)
(p q)
=2E
p
(2)
3
(3)
(p q) . (3.144)
The factors (2E
p
)
1/2
in eq. (3.143) have been chosen so that in the above product the combination
E
p
(3)
(p q) appears, which is Lorentz invariant. To see this perform a boost along zaxis. Since
the transverse components of the momentum are no aected we must consider only E
p
(p
z
k
z
).
Use the form of the Lorentz transformation of E
p
, p
z
, together with the property of the Dirac delta
(f(x)) = (x x
0
)/f
(x
0
) [2].
Using (3.111) we have in a nite box
p[q =2E
n
L
3
n,m
=2E
n
V
n,m
. (3.145)
3.3 Quantization of Fermions
We consider now the Dirac equation
(i
m)(x) = 0 (3.146)
that can be obtained from the Lagrangian
L = i
m (3.147)
where
=
0
(3.148)
and the matrices satisfy the Dirac algebra
= 2g
1 (3.149)
See [1]. If we assume a plane wave solution like the wave function of the Scr odinger equation
e
iEt
, after sustition in eq. (3.146), we have
i
0
(iE) m =0
0
E m =0 (3.150)
From the Dirac matrices properties we have
_
0
_
=
0
_
0
_
2
=1 Tr
0
=0 . (3.151)
Moreover, we know that if
satisfy the Dirac algebra, the matrices obtained after the unitary
transformation

=U
U s.t U
=U
1
(3.152)
also satisfy the Dirac algebra. To check this note that

=
_
U
U, U
U
_
=U
U
=2g
U
=2g
(3.153)
3.3. QUANTIZATION OF FERMIONS 157
In this way we can always choose U such that
0
be diagonal. Because the restrictions in eq. (3.151)
this implies that in this representation we have
0
=
_
1 0
0 1
_
(3.154)
where the 1 and 0 are the 2 2 identity and null matrix respectively. Replacing back in eq. (3.150)
we have
_
E m 0
0 E m
_
=0
E =m. (3.155)
so that from the four wave functions that compose the full Dirac spinor , two of them are of positive
energy and the other two of negative energy. The Dirac spinor has four components, in this way we
expect four independent solutions. Let us represent solutions in the form
(x)
_
_
_
_
u
1
(p)e
ipx
u
2
(p)e
ipx
v
1
(p)e
ipx
v
2
(p)e
ipx
_
_
_
_
=
+
(x) +
(x) , (3.156)
where
+
(x) u
s
(p)e
i(Etpx)
(x) v
s
(p)e
i(Etpx)
(3.157)
with
u
s
(p) =
_
_
_
_
u
1
(p)
u
2
(p)
0
0
_
_
_
_
v
s
(p) =
_
_
_
_
0
0
v
1
(p)
v
2
(p)
_
_
_
_
(3.158)
Checking this solutions to eq. (3.146) we have
(i
0
0
+ i
i

i
m)
+
(x) =0
(i
0
0
+ i m)
+
(x) =0
(
0
E p m)
+
(x) =0
(
m)
+
(x) =0
(
p m)
+
(x) =0
(
p m)u
s
(p) =0 (3.159)
and
(
p + m)v
s
(p) =0 (3.160)
This equations can also be written as
[(
p m)u
s
(p)]
=0
u
s
(p)(
m) =0
u
s
(p)
0
p
mu
s
(p)
0
=0
u
s
(p)
0
mu
s
(p)
0
=0
u
s
(p)(
p m) =0 (3.161)
v
s
(p)(
p + m) =0 (3.162)
At zero momentum, E = m y
0
u
s
(0) = + u
s
(0)
0
v
s
(0) =v
s
(0) (3.163)
From [?]
Consider the matrix
0
. It is a 4 4 matrix, so it has four eigenvalues and eigenvectors.
It is hermitian, so the eigenvalues are real. In fact, from Eq. (3.149) we know that its
square is the unit matrix, so that its eigenvalues can only be 1. Since
0
is traceless, as
we have proved in 3, there must be tow eigenvectors witht eigenvalue +1 and tow with
1
Eq. (3.163) shows that at zero momentum, the uspinors and the vspinors are simply eigenstates
of
0
with eigenvalues +1 and 1. Of course this guaranteses that
u
s
(0)v
s
(0) = 0 (3.164)
since the belong to dierente eigenvalues. Note that the two u
s
(0) and the two v
s
(0) are degenerate.
We dene
u
s
(p)
s
v
s
(0)
s
(3.165)
where the munis sign in
s
is just a convention. We dene the normalized eigenvectors and such
that
s
=
ss

s
=
ss
s
= 0 (3.166)
In this way we have
1/2
=
_
_
_
_
1
0
0
0
_
_
_
_

1/2
=
_
_
_
_
0
1
0
0
_
_
_
_

1/2
=
_
_
_
_
0
0
1
0
_
_
_
_

1/2
=
_
_
_
_
0
0
0
1
_
_
_
_
(3.167)
To obtain the spinors for any value of p we know that they must satisfy eqs. (3.159), (3.160),
and, reduce to eq. (3.165) when p 0. The result is
u
s
(p) =N
p
(
p + m)
s
v
s
(p) =N
p
(
p + m)
s
(3.168)
Choosing
N
p
=
1
E + m
(3.169)
we obtain
u
s
(p)u
s
(p) = v
s
(p)v
s
(p) = 2E
ss
(3.170)
u
s
(p)v
s
(p) = v
s
(p)u
s
(p) = 0 (3.171)
In terms of the conjugate spinors
u
s
(p)u
s
(p) =2m
ss
v
s
(p)v
s
(p) =2m
ss
(3.172)
u
s
(p)v
s
(p) = v
s
(p)u
s
(p) = 0 (3.173)
The spinors also satisfy some completeness relations (For details see [?])
s
u
s
(p) u
s
=
p + m (3.174)
s
v
s
(p) v
s
=
p m (3.175)
The solutions to the free Dirac equations are
particle
(x) =
1
_
2E
p
V
u
s
(p)e
ipx
antiparticle
(x) =
1
_
2E
p
V
v
s
(p)e
ipx
(3.176)
As with the scalar eld, we write the Dirac eld as an integral over momentum space of the plane
wave solutions, with creation and annihilation operators as coecients,
(x) =
_
d
3
p
1
(2)
3
_
2E
p
s=1,2
_
a
s
(p)u
s
(p)e
ipx
+ b
s
(p)v
s
(p)e
ipx
(3.177)
H =
_
d
3
x1
=
_
d
3
x
_
L
L
_
=
_
d
3
x
_
i
0
0
i
0
0
i
i
i
+ m
_
=
_
d
3
x
_
i
i
i
+ m
_
=
_
d
3
x (i + m) . (3.178)
Since that
(i + m) =
_
d
3
p
1
(2)
3
_
2E
p
s=1,2
_
a
s
(p) (i + m) u
s
(p)e
ipx
+b
s
(p) (i + m) v
s
(p)e
ipx
= c
_
d
3
p
1
(2)
3
_
2E
p
s=1,2
_
a
s
(p) ( p + m) u
s
(p)e
ipx
+b
s
(p) ( p + m) v
s
(p)e
ipx
. (3.179)
From eqs. (3.159), and (3.160) we have
_
0
p
0
+
i
p
i
m
_
u
s
(p) =0
_
0
p
0
+
i
p
i
+ m
_
v
s
(p) =0 , (3.180)
_
0
E
p
i
p
i
m
_
u
s
(p) =0
_
0
E
p
i
p
i
+ m
_
v
s
(p) =0 , (3.181)
( m) u
s
(p) =
0
E
p
u
s
(p)
( + m) v
s
(p) =
0
E
p
v
s
(p) , (3.182)
( + m) u
s
(p) =
0
E
p
u
s
(p)
( m) v
s
(p) =
0
E
p
v
s
(p) . (3.183)
Replacing back in eq. (3.179), we have
(i + m) =
_
d
3
p
1
(2)
3
_
2E
p
s=1,2
_
a
s
(p)
0
E
p
u
s
(p)e
ipx
b
s
(p)
0
E
p
v
s
(p)e
ipx
. (3.184)
Therefore
H =
_
d
3
x (i + m)
=
_
d
3
x
_
d
3
p
1
(2)
3
_
2E
p
=1,2
_
a
(p
)u
(p
)e
ip
x
+ b
s
(p
)v
(p
)e
ip
x
_
0
(i + m)
=
_
d
3
x
_
d
3
p
1
(2)
3
_
2E
p
=1,2
_
a
(p
)u
(p
)e
ip
x
+ b
s
(p
)v
(p
)e
ip
x
_
_
d
3
p
1
(2)
3
_
2E
p
s=1,2
_
a
s
(p)
0
E
p
u
s
(p)e
ipx
b
s
(p)
0
E
p
v
s
(p)e
ipx
=
_
d
3
x
2(2)
6
_
d
3
p
_
d
3
p
E
p
E
p
s,s
=1,2
_
a
(p
)u
(p
)e
ip
x
+ b
s
(p
)v
(p
)e
ip
x
_
_
a
s
(p)u
s
(p)e
ipx
b
s
(p)v
s
(p)e
ipx
=
_
d
3
x
2(2)
6
_
d
3
p
_
d
3
p
E
p
E
p
s,s
=1,2
_
a
(p
)a
s
(p)u
(p
)u
s
(p)e
i(p
p)x
b
s
(p
)b
s
(p)v
(p
)v
s
(p)e
i(pp
)x
_
=
_
d
3
p
_
d
3
p
2(2)
3
E
p
E
p
s,s
=1,2
_
a
(p
)a
s
(p)u
(p
)u
s
(p)
_
d
3
x
(2)
3
e
i(p
p)x
b
s
(p
)b
s
(p)v
(p
)v
s
(p)
_
d
3
x
(2)
3
e
i(pp
)x
_
=
_
d
3
p
_
d
3
p
2(2)
3
E
p
E
p
s,s
=1,2
_
a
(p
)a
s
(p)u
(p
)u
s
(p)e
i(E
p
Ep)t
(3)
(p p
) b
s
(p
)b
s
(p)v
(p
)v
s
(p)e
i(EpE
p
)t
(3)
(p p
)
_
=
_
d
3
p
2(2)
3
s,s
=1,2
_
a
(p)a
s
(p)u
(p)u
s
(p) b
s
(p)b
s
(p)v
(p)v
s
(p)
_
=
_
d
3
p
E p
(2)
3
s=1,2
_
a
(p)a
s
(p) b
s
(p)b
s
(p)
_
. (3.185)
In order to obtain the quantization relations could see that if commutation relations are used we
could get
: H: =
_
d
3
p
(2)
3
s=1,2
E
p
_
a
s
(p)a
s
(p) b
s
(p)b
s
(p)
(3.186)
The minus sign arise from the anticommutation relations, so that a real spinor eld, where b
s
(p) =
a
s
(p) is automatically zero. Even after normal ordering, this Hamiltonian could give to arise neg-
ative energy eigenvalues, which is a serious problem. If instead we assume that the creation and
annihilation operators satisfy anticommutation relations
_
a
r
(p), a
s
(q)
_
=
_
b
r
(p), b
s
(q)
_
= (2)
3
rs
(3)
(p q) (3.187)
With this relations and taking into account
(x) =
L
(
0
)
= i
0
= i
(3.188)
we obtain
(x, t),
(y, t) = i
(3)
(x y) (3.189)
With the anticommutators the normalordered Hamiltonian is
: H: =
_
d
3
p
(2)
3
s=1,2
E
p
_
a
s
(p)a
s
(p) + b
s
(p)b
s
(p)
(3.190)
Moreover
: Q: =q
_
d
3
x:
:
=q
_
d
3
p
s=1,2
_
a
s
(p)a
s
(p) b
s
(p)b
s
(p)
(3.191)
With this denition a
s
(p) creates particles of charge q, while b
s
(p) creates antiparticles of charge
q. In a similarly way to eq. (3.139), the most general free particle solution to Dirac equation is
(x) =

+
(x) +

(x) (3.192)
+
(x) =
_
d
3
p
1
(2)
3
_
2E
p
s=1,2
a
s
(p)u
s
(p)e
ipx
(x) =
_
d
3
p
1
(2)
3
_
2E
p
s=1,2
b
s
(p)v
s
(p)e
ipx
(3.193)
The Fourier expansion for antiparticles is
+
(x) =
_
d
3
p
1
(2)
3
_
2E
p
s=1,2
b
s
(p) v
s
(p)e
ipx
(x) =
_
d
3
p
1
(2)
3
_
2E
p
s=1,2
a
s
(p) u
s
(p)e
ipx
(3.194)
In this way a
s
and a
s
are the creation and annihilation operators for particles, while b
s
and b
s
are
the creation and annihilation operators for antiparticles.
It is clear then that the one particle state is
[e
(p, s)
_
1
V
a
s
(p)[0 (3.195)
while the one antiparticle state is
[e
+
(p, s)
_
1
V
b
s
(p)[0 . (3.196)
Chapter 4
Quantization of the electromagnetic eld
4.1 Preliminaries
If we impose charge conservation:
= 0, the Proca Equations can be written without lost of

generality as ( 2.4 of [1])
(+ m
2
)A
= J
. (4.1)
where A
= (, A).
The right side of the equation can be obtained after replacing the quantities in the equation for
energy-momentum conservation
E
2
p
2
= m
2
, (4.2)
p
= i
u
. (4.3)
This suggest that quantum mechanics is a key ingredient to understand the local conservation of
electric charge, as we will see later.
For the scalar part we have the Klein-Gordon equation of an real scalar eld:
(+ m
2
) = (4.4)
which can be obtained from the Lagrangian ( 3.1 of [1])
L = L
free
+L
int
, (4.5)
165
166 CHAPTER 4. QUANTIZATION OF THE ELECTROMAGNETIC FIELD
L
free
=
1
2

1
2
m
2
2
L
int
=, (4.6)
where is the charge density of the eld which is the source for , and L
int
is the interaction
Lagrangian.
In the same section it is shown that this Lagrangian give to arise to the Yukawa interaction
V (r) =
1
4
e
mr
r
. (4.7)
In a similar way, when the Lorentz force
F = q E + q v B, (4.8)
is interpreted in terms of quantum mechanical operators ( 3.3 of [1]) we have the canonical momen-
tum
+ iqA
. (4.9)
Now, if we force the Scr odinger equation to be invariant under local phase changes ( 3.4 of [1]),
we need to replace the normal derivate by the covariant derivate which must transform as the wave
equation:
T
(T
= e
i(x)
T
. (4.10)
This suggest to make the minimum replacement
T
+ iqA
, (4.11)
where A
is a new eld that compensates the changes form the derivate. From this it can be shown
that the same identity is valid for all of the powers
[(T
]
n
= e
i(x)
(T
)
n
. (4.12)
From eqs. (4.10) and (4.11) the tranformation of A
can be obtained:
A
= A
1
q
. (4.13)
Therefore, the new eld tranform like the electromagnetic eld, and the modied Lagrangian for the
elds and A
L =
1
2m
(T)
T
i
2
_
T
0

_
T
0
1
4
F
, (4.14)
4.2. QUANTIZATION OF THE ELECTROMAGNETIC FIELD 167
give to arise to the Scrodinger equation in presence of the electromagnetic eld plus the Maxwell
with the explicit current
j
=
_
q
= 0
iq
2m
[()
2iq
A] = i
. (4.15)
We now turn to tue quantization of the electromagnetic eld
4.2 Quantization of the electromagnetic eld
Here we follow closelly [15] chapter 2.
in the electromagnetic Lagrangian the generalized momentum conjugate to the time component
of the four-vector potential is zero
0
=
L
_
A
0
t
_ = 0 . (4.16)
Therefore, we cannot quantizate the A
0
eld.
The arbitrariness associated with the gauge freedom (4.13) must be removed so that the eld can
be uniquely specied everywhere. Two popular choices for the gauge xing, are the Lorentz gauge
= 0 , (4.17)
and Coulomb gauge
A = 0 . (4.18)
With the Lorentz gauge a new term is added to the Lagrangian which contains the time derivative
of A
0
, while in the Coulomb gauge the quantity A
0
may be eliminated from the Lagrangian.
In the Coulomb gauge, we have for the A
0
component
2
A
0
= . (4.19)
For the A
i
component
_
A
j
(r
, t), (r, t)
= i
_
ij
2
_
3
(r r
)
3T
(r r
) . (4.20)
168 CHAPTER 4. QUANTIZATION OF THE ELECTROMAGNETIC FIELD
These commutation relations between the creation and annihilation involve only the independent
degrees of freedom.
The Hamiltonian obtainde from the Lagrangia is
H =
n,
n
a
n,
a
n,
, (4.21)
while the momentum operator is
p =
n,
k
n
a
n,
a
n,
, (4.22)
We now shown that the particles which emerges from the quantization of the electromagnetic eld
(the photons) have spin one. To obtain these results, it is necessary to discuss the behavior of these
elds under rotations.
To follow the non-relativistic part of this course we recommend now go directly from section 6.1
to 6.3 where the Smatrix is dened and the probability calculated. In Section 6.5 there is the
general formula for decay. In section 8.1 the perturbative expansion of the Smatrix is presented.
Finally, in section 8.2 an application for the interaction of a non-relativistic atom with radiation, is
given in the context of radiative decay.
Chapter 5
Propagators
5.1 Scalars
i(x x
) = 0[T(x)(x
)[0 (5.1)
(x x
) =
_
d
4
p
(2)
4
e
ip(xx
)
(p) (5.2)
(p) =
1
p
2
m
2
+ i
(5.3)
5.2 Fermions
iS
F
(x x
) = 0[T
(x)
(x
)[0 (5.4)
S
F
(x x
) =
_
d
4
p
(2)
4
e
ip(xx
)
S
F
(p) (5.5)
S
F
(p) =

p + m
p
2
m
2
+ i
=
1
p m + i
. (5.6)
169
170 CHAPTER 5. PROPAGATORS
5.3 Bosons
D
(p) =
1
k
2
+ i
_
g
(1 )
p
p
2
_
(5.7)
Chapter 6
Smatrix
We will use the Smatrix formulation to obtain the decay rates and cross section formulas.
6.1 The Smatrix
The Scrodinger equation for the wave function of some state a
[a, t
a
(t) , (6.1)
is
i

t
[a, t = H[a, t . (6.2)
The solution to this equation is
[a, t =e
iH(tt
i
)
[a, t
i
, (6.3)
since
i

t
[a, t =i(i)He
H(tt
i
)
[a, t
i
=H[a, t . (6.4)
Dening the time-evolution operator as
U(t, t
i
) = e
iH(tt
i
)
, (6.5)
171
172 CHAPTER 6. SMATRIX
we have that in In the Scr odinger picture dened by eq.(6.3), the state of a system evolves with time
[a, t =U(t, t
i
)[a, t
i
[a, t =U(t, t
i
)[a
[a, t =e
iH(tt
i
)
[a , (6.6)
where [a, t
i
, at an initial time t
i
, is an eigenstate of a set of conmuting operators, and is denoted
simply [a. Similarly [b = [b, t
f
.
We have then
b, t
f
[a, t
f
=b[a, t
f
=b[e
iH(t
f
t
i
)
[a, t
i
=b[e
iH(t
f
t
i
)
[a , (6.7)
is the amplitude for the process in which the initial state [a evolves into the nal state [b. In the
limit t
f
t
i
, the operator e
iH(t
f
t
i
)
is called the Smatrix. Therefore S is an operator that
maps an initial state to a nal state
[a S[a , (6.8)
an the scattering amplitudes are given by its matrix elements, b[S[a. Observe that
a[ a[S
, (6.9)
a[a = 1 a[S
S[a = 1 , (6.10)
so that SS
= S
S = 1.
More rigorously, if a[a = 1, and [n is a complete set of states, the probability that [a evolves
into [n, summed over all [n, must be 1,
n
[n[S[a[
2
= 1. (6.11)
On the other hand we can write
n
[n[S[a[
2
=
n
a[S
[nn[S[a
=a[S
n
[nn[
_
S[a
=a[S
S[a
=1 , (6.12)
6.2. RELATIVISTIC AND NO RELATIVISTIC NORMALIZATIONS 173
and we conclude that SS
= S
S = 1. The unitarity of the Smatrix express the conservation of

probability. It is also convenient to dene the T matrix, separating the identity operator,
S = 1 + iT (6.13)
Consider a generic Smatrix element
p
1
. . . p
n
[S[k
1
. . . k
n
(6.14)
For notational simplicity the states are just labeled by their momenta, but all our considerations can
be generalized to the case in which the spin is taken into account. We have also dened the operator
T from S = 1 + iT. We assume that none of the initial momenta p
j
coincides a nal momentum
k
i
. This eliminates processes in which one of the particles behaves as a spectator and does not
interact with the other particles. In the language of Feynman diagrams to be explained later, this
means that we will consider only connected diagrams. Therefore, if we restrict to the situation in
which no initial and nal momenta coincide, the matrix element of the identity operator between
these states vanishes, and we need actually to compute the matrix element of iT
p
1
. . . p
n
[iT[k
1
. . . k
n
(6.15)
In explicit calculations there will be an overall Dirac delta factor imposing energymomentum con-
servation. In order not to write explicitly the Dirac delta each time we compute a matrix element of
iT, it is convenient to dene a matrix element M
fi
from the matrix element,
p
1
. . . p
n
[iT[k
1
. . . k
n
= (2)
4
(4)
_
j
p
j
j
k
j
_
iM
fi
. (6.16)
The labels i, f refer to the initial and nal states. Explicitly
M
fi
= M(p
1
, . . . , p
n
; k
1
, . . . , k
n
) . (6.17)
More generally, the initial and nal states are labeled also by the spin states of the initial and nal
particles.
So, instead of S or T, the quantity to be calculated is M
fi
, but this need rst to be relativistically
normalized, in which case it will be denoted as /
fi
.
6.2 Relativistic and no relativistic normalizations
We rst consider a system in a cubic box with spatial volume V = L
3
. At the end of the computation
V will be sent to innity. It is sometimes convenient to put the system into a box of size L, so that the
total volume V = L
3
is nite. This procedure regularizes divergences coming from the innite-volume
limit or, equivalently, from the small momentum region, and is an example of an infrared cuto. In
a nite box of size L, imposing periodic boundary conditions on the elds, the momenta take the
discrete values p = 2n/L with n = (n
x
, n
y
, n
z
) a vector with integer components. In non-relativistic
quantum mechanics a one-particle state with momentum p in the coordinate representation is given
by a plane wave
p
(x) = Ce
ipx
, (6.18)
and the normalization constant is xed by the condition that there is one particle in the volume V ,
1 =
_
V
d
3
x [
p
(x)[
2
=
_
V
d
3
x
p
(x)
p
(x)
=[C[
2
_
V
d
3
x
=[C[
2
V , (6.19)
and
p
(x) =
1
V
e
ipx
. (6.20)
Wave functions with dierent momenta are orthogonal, and therefore
_
V
d
3
x
p
1
(x)
p
2
(x) =
p
1
,p
2
(6.21)
Writing
p
(x) = x[p and using the completeness relation
_
V
d
3
x[xx[ = 1, we can write this as
p
1
[p
2
NR
=p
1
[
_
V
d
3
x[xx[p
2
=
_
V
d
3
xp
1
[xx[p
2
=
_
V
d
3
x
p
1
(x)
p
2
(x)
=
p
1
,p
2
. (6.22)
The superscript NR reminds us that the states have been normalized according to the conventions
of non-relativistic quantum mechanics.
6.3. PROCESS PROBABILITY 175
In relativistic QFT this normalization is not the most convenient, because the spatial volume
V is not relativistically invariant, and therefore the condition one-particle per volume V is not
invariant. A more convenient Lorentz invariant form was introduced in eq. (3.145)
p
1
[p
2
R
= 2E
p
1
V
p
1
,p
2
(6.23)
Therefore the dierence between the relativistic and non-relativistic normalization of the one-particle
states is, comparing eqs. (6.22) and (6.23)
[p
R
= (2E
p
V )
1/2
[p
NR
(6.24)
and of course for a multiparticle state
[p
1
, . . . , p
n
R
=
_
n
i=1
(2E
p
V )
1/2
_
[p
1
, . . . , p
n
NR
(6.25)
We denote by M
fi
, dened in eq. (6.16), the scattering amplitude between the initial state with
momenta q
1
, . . . , q
n
and the nal state with momenta p
1
, . . . , p
n
, with non-relativistic normalization
of the states, and by /
fi
the same matrix element with relativistic normalization of the states. Then
from eq. (6.16)
(2)
4
(4)
_
i
p
i
i
k
i
_
i/
fi
=p
1
. . . p
n
[iT[k
1
. . . k
n
R
=
n
i=1
(2E
p
i
V )
1/2
n
j=1
_
2E
k
j
V
_
1/2
p
1
. . . p
n
[iT[k
1
. . . k
n
NR
=
n
i=1
(2E
p
i
V )
1/2
n
j=1
_
2E
k
j
V
_
1/2
(2)
4
(4)
_
i
p
i
i
k
i
_
iM
fi
(6.26)
Therefore
M
fi
=
n
i=1
(2E
p
i
V )
1/2
n
j=1
_
2E
k
j
V
_
1/2
/
fi
(6.27)
6.3 Process probability
Consider the matrix element of iT in (6.16)
p
1
. . . p
n
[iT[k
1
. . . k
m
NR
= (2)
4
(4)
_
i
p
i
j
k
j
_
iM
fi
(6.28)
Assume for the moment that all particles are indistinguishable. The rules of quantum mechanics tell
us that the probability of this process is obtained by taking the square module of the amplitude
p
1
. . . p
n
[iT[k
1
. . . k
m
NR
2
=
(2)
4
(4)
_
i
p
i
j
k
j
_
iM
fi
2
(6.29)
and we are confronted with the square of the delta function. To compute it, we recall that we are
working in a nite spatial volume and, from eq. (3.112)
(2)
3
(3)
(0) = V (6.30)
Similarly we regularize also the time interval, saying that the time runs from T/2 to T/2 so that
(2)
4
(4)
(0) = V T (6.31)
Then
p
1
. . . p
n
[iT[k
1
. . . k
m
NR
2
=
(2)
4
(4)
_
p
j
k
i
_
iM
fi
2
=(2)
4
(4)
_
p
i
k
j
_
V TM
fi
(6.32)
Moreover we must sum over all nal states. In the discrete limit, since we are working in a nite
volume V , the sum over all nal states corresponds to the sum over the possible discrete values of
the momenta of the nal particles
k
j
=
2n
j
L
,
_
_
n
x
j
= , . . . , 1, 0, 1, . . .
n
y
j
= , . . . , 1, 0, 1, . . .
n
z
j
= , . . . , 1, 0, 1, . . .
(6.33)
k
j
=
n
x
j
n
y
j
n
z
j
(6.34)
In the large-volume limit for each particle we can write, using eq. (3.105)
k
j
V
(2)
3
_
d
3
k
j
, (6.35)
6.4. CROSS SECTION 177
The decay probability in (6.32) can be written as
=
k
1
. . .
km
p
1
. . . p
n
[iT[k
1
. . . k
m
NR
2
=
k
1
. . .
km
(2)
4
(4)
_
i
p
i
i
k
j
_
V T [M
fi
[
2
=
_
. . .
_
V d
3
k
1
(2)
3
. . .
V d
3
k
m
(2)
3
(2)
4
(4)
_
i
p
i
i
k
j
_
V T [M
fi
[
2
=
_
. . .
_
(2)
4
(4)
_
i
p
i
i
k
j
_
V T [M
fi
[
2
m
j=1
V d
3
k
j
(2)
3
. (6.36)
=
_
. . .
_
(2)
4
(4)
_
i
p
i
i
k
j
_
V T [/
fi
[
2
m
j=1
V d
3
k
j
(2)
3
_
n
i=1
(2E
p
i
V )
1/2
m
j=1
_
2E
k
j
V
_
1/2
_
2
=
_
. . .
_
(2)
4
(4)
_
i
p
i
i
k
j
_
V T [/
fi
[
2
n
i=1
1
2E
p
i
V
m
j=1
d
3
k
j
(2)
3
2E
k
j
(6.37)
The probability for the process of an initial particle decaying into n nal particles is then
1
=
_
. . .
_
(2)
4
(4)
_
p
i
k
j
_
T [/
fi
[
2
1
2E
p
n
j=1
d
3
k
j
(2)
3
2E
k
j
(6.38)
On the other hand the probability for a process with two initial particles colliding into n nal particles
is
2
=
_
. . .
_
(2)
4
(4)
_
p
1
+ p
2
i
k
j
_
V T [/
fi
[
2
1
2E
p
1
V
1
2E
p
2
V
n
j=1
d
3
k
j
(2)
3
2E
k
j
(6.39)
6.4 Cross Section
Consider a large number point-like projectiles directed to an area A that includes a solid target of
area , as displayed in Fig. , such that alll the ll the area A randomly Assuming that an interaction
A
Figure 6.1: Cross section probability

will occur (with 100% probability) if the projectile hits the solid, and not at all (0% probability) if
it misses, the total interaction probability for the single projectile will be
P
S
=

A
. (6.40)
Now suppose we have a parallel beam with density of particles n and velocity v towards the target.
In time t, this beam lls a volume
V = Avt . (6.41)
Choosing t such that the volume contains just one particle, we can write
n = 1/V , (6.42)
or
1 = nvtA. (6.43)
replacing back in (6.40) we have
=
P
s
nvt
. (6.44)
P
s
is just the decay probability in eq. (6.39). Therefore
=

2
nvT
=
1
nvT
_
. . .
_
(2)
4
(4)
_
p
1
+ p
2
i
k
j
_
V T [/
fi
[
2
1
2E
p
1
V
1
2E
p
2
V
n
j=1
d
3
k
j
(2)
3
2E
k
j
=
1
nvV
_
. . .
_
(2)
4
(4)
_
p
1
+ p
2
i
k
j
_
[/
fi
[
2
1
2E
p
1
1
2E
p
2
n
j=1
d
3
k
j
(2)
3
2E
k
j
. (6.45)
The density of particles of the incident state is normalized to one particle in the entire volume, so
that n = 1/V . Therefore
=
1
v
_
. . .
_
(2)
4
(4)
_
p
1
+ p
2
i
k
j
_
[/
fi
[
2
1
2E
p
1
1
2E
p
2
n
j=1
d
3
k
j
(2)
3
2E
k
j
. (6.46)
In general, as both particles may be moving we could use the relative velocity between them, v
rel
,
=
1
v
rel
_
. . .
_
(2)
4
(4)
_
p
1
+ p
2
i
k
j
_
[/
fi
[
2
1
2E
p
1
1
2E
p
2
n
j=1
d
3
k
j
(2)
3
2E
k
j
. (6.47)
In a frame where p
1
and p
2
are along the same line, this reduces to
v
rel
=
p
1
E
1
p
2
E
2
. (6.48)
In fact, for not relativistic particles, where E
i
= m
i
, this coincides with the usual relative velocity
v
rel
=
m
1
v
1
E
1
m
2
v
2
E
2
=[v
1
v
2
[ . (6.49)
The most general formula for the relative velocity is
v
rel
=
I
E
1
E
2
(6.50)
where
I =
_
(p
1
p
2
)
2
m
2
1
m
2
2
(6.51)
In general
I =
_
(E
1
E
2
p
1
p
2
)
2
m
2
1
m
2
2
=
_
E
2
1
E
2
2
+ (p
1
p
2
)
2
2E
1
E
2
p
1
.p
2
m
2
1
m
2
2
(6.52)
Since
m
2
1
m
2
2
=(E
2
1
p
2
1
)(E
2
2
p
2
2
)
=(E
2
1
E
2
2
p
2
1
E
2
2
E
2
1
p
2
2
+p
2
1
p
2
2
) (6.53)
I =
_
p
2
1
E
2
2
2E
1
E
2
p
1
p
2
+ E
2
1
p
2
2
+ (p
1
p
2
)
2
p
2
1
p
2
2
(6.54)
If
(p
1
p
2
)
2
p
2
1
p
2
2
= 0 (6.55)
that implies that p
1
and p
2
are colineals,
I =
_
p
2
1
E
2
2
2E
1
E
2
p
1
p
2
+ E
2
1
p
2
2
=
_
(p
1
E
2
p
2
E
1
)
2
=[p
1
E
2
p
2
E
1
[ (6.56)
v
rel
=
I
E
1
E
2
=
p
1
E
1
p
2
E
2
(6.57)
To simplify the notation we set E
i
= E
p
i
=, and E
f
= E
p
f
. Moreover, the dierential cross
section is
d =(2)
4
(4)
_
2
i=1
p
i
f
p
f
_
1
4v
rel
E
1
E
2
[/
fi
[
2
f
d
3
k
f
(2)
3
2E
f
= (2)
4
1
4v
rel
E
1
E
2
[/
fi
[
2
d
n
(p
1
, p
2
; k
1
, . . . , k
n
) (6.58)
where
d
(n)
(p
1
, p
2
; k
1
, k
2
, . . . , k
n
) =
(4)
_
p
j
k
j
_
n
j=1
d
3
k
j
(2)
3
2E
k
j
. (6.59)
We keep the diferential notation both for d, and d until the last integration have been made.
6.4.1 2to2 cross section
The the 2to2 cross section is
d =
(2)
4
4v
rel
E
1
E
2
[/
fi
[
2
d
2
(p
1
, p
2
; p
1
, p
2
)
=
2
4
4
2
8
6
4v
rel
E
1
E
2
2
8
6
[/
fi
[
2
d
n
(p
1
, p
2
; k
1
, . . . , k
n
)
=
1
2
6
2
v
rel
E
1
E
2
[/
fi
[
2
_
4(2)
6
d
2
(p
1
, p
2
; p
1
, p
2
)
=
1
64
2
v
rel
E
1
E
2
[/
fi
[
2
_
4(2)
6
d
2
(p
1
, p
2
; p
1
, p
2
)
(6.60)
where, as in eq. (6.110)
4(4)
6
d
(2)
(p
1
, p
2
; p
1
, p
2
) =
4(4)
6
4(2)
6
(4)
(p
1
+ p
2
p
1
p
2
)
d
3
p
1
E
1
d
3
p
2
E
2
=
(4)
(p
1
+ p
2
p
1
p
2
) 4(2)
6
d
3
p
1
E
1
d
3
p
2
E
2
(6.61)
We now will nd an expression for cross section in the center of mass frame (CM)
The center of mass (CM) frame is dened by the condition
p
1
+p
2
= 0 (6.62)
The function in Eq. (??)
(4)
(p + p
2
p
1
p
2
) =
(3)
(p
1
+p
2
p
1
p
2
)(E
1
+ E
2
E
1
E
2
) (6.63)
In the CM frame
(4)
(p + p
2
p
1
p
2
) =
(3)
(p
1
+p
2
)(E
1
+ E
2
E
1
E
2
) (6.64)
/
fi
in integration does not depend on [p
1
[ or [p
2
[ as the nal momentum is xed by the initial
momentum whenever the nal states have only two particles. In this way the integration on p
2
can
be evaluated directly for d
(2)
. Replacing back in Eq. (6.60)
4(2)
6
d
(2)
=
(3)
(p
1
+p
2
)(E
1
+ E
2
E
1
E
2
)
d
3
p
1
E
1
d
3
p
2
E
2
=(E
1
+ E
2
E
1
E
2
)
d
3
p
1
E
1
_

(3)
(p
1
+p
2
)
d
3
p
2
E
2
=(E
1
+ E
2
E
1
E
2
)
d
3
p
1
E
1
E
2
(6.65)
4(2)
6
d
(2)
= (E
1
+ E
2
E
1
E
2
)
p
1
2
d[p
1
[d
E
1
E
2
(6.66)
As
[p
1
[ =
_
E
1
2
m
1
2
(6.67)
d[p
1
[
dE
1
=
2E
1
2
_
E
1
2
m
1
2
=
E
1
[p
1
[
(6.68)
In this way, we can write, in general
[p[ d[p[ = E dE (6.69)
and
4(2)
6
d
(2)
= (E
1
+ E
2
E
1
E
2
)
[p
1
[E
1
dE
1
E
1
E
2
d
= (E
1
+ E
2
E
1
E
2
)
[p
1
[dE
1
E
2
d (6.70)
From the function in Eq. (6.63) we have that in the CM frame
p
1
+p
2
p
1
p
2
= 0
CM
_
p
1
= p
2
p
1
= p
2
(6.71)
Squaring the rst expression, and taking into account that
p
1
=
_
E
1
2
m
1
2
(6.72)
we have
p
1
2
=p
2
2
E
1
2
m
1
2
=E
2
2
m
2
2
, (6.73)
E
2
=
_
E
1
2
m
1
2
+ m
2
2
(6.74)
In this way we can express E
2
in terms of E
1
in Eq. (6.70). Moreover, we can dene the center of
mass energy as
s = E
1
+ E
2
(6.75)
Using The energy part of function in Eq. (6.63) can be written as
s E
_
E
1
2
m
1
2
+ m
2
2
_
(6.76)
As established before, /
fi
in this case in independent of [p
1
[, and the integration on E
1
can be done
directly only for d
(2)
. The integral is easily performed using the identity
(f(z)) =
n
(z z
n
)
[f
(z
n
)[
(6.77)
where z
n
are the zeroes of f(z). In this case, this function is a function of the integration variable
E
1
, with only one zero
(f(x)) =
(x x
0
)
[f
(x
0
)[
(6.78)
where
f(x) =
s x
_
x
2
m
1
2
+ m
2
2
(6.79)
Therefore
4(2)
6
d
(2)
=d
_
(x x
0
)
[f
(x
0
)[
[p
1
(x)[
E
2
(x)
dx
=d
1
[f
(x
0
)[
[p
1
(x
0
)[
E
2
(x
0
)
(6.80)
where from Eqs. (6.72), (6.74),
p
1
(x
0
) =
_
x
2
0
m
1
2
E
2
(x
0
) =
_
x
2
0
m
1
2
+ m
2
2
(6.81)
The zero is obtained from
s x
0
_
x
2
0
m
1
2
+ m
2
2
= 0
s 2
s x
0
+ x
2
0
= x
2
0
m
1
2
+ m
2
2
s 2
s x
0
= m
1
2
+ m
2
2
(6.82)
with solution
x
0
=
s + m
1
2
m
2
2
2
s
(6.83)
As (See deltaxn.nb for additional details)
f
(x) =
x
_
x
2
m
1
2
+ m
2
2
1 (6.84)
we have
f
(x
0
) =
m
1
2
m
2
2
+ s
s
_
(m
1
2
+m
2
2
+s)
2
s
1
=
m
1
2
+ m
2
2
s
m
1
2
+ m
2
2
+ s
1
=
m
1
2
+ m
2
2
s + m
1
2
m
2
2
s
m
1
2
+ m
2
2
+ s
=
2s
s + m
2
2
m
1
2
, (6.85)
and
(f(E
1
)) = (E
1
x
0
)
_
s + m
2
2
m
1
2
2s
_
(6.86)
Replacing the expression for x
0
in (6.83) into Eq. (6.81) we have (See deltaxn.nb for additional
details)
p
1
(x
0
) =
_
[s (m
1
m
2
)
2
][s (m
1
+ m
2
)
2
]
2
s
E
2
(x
0
) =
s m
1
2
+ m
2
2
2
s
(6.87)
Replacing Eqs. (6.85), and (6.87) in Eq. (6.80) we have
4(2)
6
d
(2)
= d
1
[f
(x
0
)[
_
x
2
0
m
1
2
_
x
2
0
m
1
2
+ m
2
2
= d
_
s m
1
2
+ m
2
2
2s
_
_
[s (m
1
m
2
)
2
][s (m
1
+ m
2
)
2
]
s m
1
2
+ m
2
2
= d
_
[s (m
1
m
2
)
2
][s (m
1
+ m
2
)
2
]
2s
(6.88)
Dening the kinematic two particle function
(a, b, c) (a b + c)
2
4ac (6.89)
and taking into account that
_
s m
2
2
+ m
1
2
_
2
4sm
1
2
= [s (m
1
m
2
)
2
][s (m
1
+ m
2
)
2
] (6.90)
we have
4(2)
6
d
(2)
= d
1/2
(s, m
2
2
, m
1
2
)
2s
(6.91)
Moreover
p
1
=
1/2
(s, m
2
2
, m
1
2
)
2
s
(6.92)
To further evaluate Eq. (6.60), we need to express v
rel
and E
1
E
2
in terms of s and the masses.
Concerning v
rel
, from Eq. (6.57), evaluated in CM frame
E
1
E
2
v
rel
=E
1
E
2
p
1
E
1
p
2
E
2
=E
1
E
2
p
1
E
1
+
p
1
E
2
=[p
1
[ (E
1
+ E
2
)
=[p
1
[
s (6.93)
Replacing back Eqs. (6.88), and (6.93) into Eq. (6.60), we have
d =
1
64
2
E
1
E
2
v
rel
[/
fi
[
2
_
4(2)
6
d
(2)
(6.94)
d
d
=
1
64
2
E
1
E
2
v
rel
[/
fi
[
2
_
[s (m
1
+ m
2
)
2
][s (m
1
m
2
)
2
]
2s
(6.95)
By using Eq. (6.93)
d
d
=
1
64
2
[p
1
[
s
[/
fi
[
2
_
[s (m
1
+ m
2
)][s (m
1
2
m
2
2
)]
2s
(6.96)
In the CM frame
s =E
1
+ E
2
=
_
p
2
1
+ m
2
1
+
_
p
2
2
+ m
2
2
=
_
p
2
1
+ m
2
1
+
_
p
2
1
+ m
2
2
(6.97)
s =2p
2
1
+ m
2
1
+ m
2
2
+ 2
_
p
4
1
+ (m
2
1
+ m
2
2
)p
2
1
+ m
2
1
m
2
2
s (2p
2
1
+ m
2
1
+ m
2
2
) =2
_
p
4
1
+ (m
2
1
+ m
2
2
)p
2
1
+ m
2
1
m
2
2
(6.98)
s
2
2s(2p
2
1
+ m
2
1
+ m
2
2
) + [2p
2
1
+ (m
2
1
+ m
2
2
)]
2
=4(p
4
1
+ (m
2
1
+ m
2
2
)p
2
1
+ m
2
1
m
2
2
)
s
2
2s(2p
2
1
+ m
2
1
+ m
2
2
) + 4p
4
1
+ 4p
2
1
(m
2
1
+ m
2
2
) + (m
2
1
+ m
2
2
)
2
=4(p
4
1
+ (m
2
1
+ m
2
2
)p
2
1
+ m
2
1
m
2
2
)
4sp
2
1
+ s
2
2s(m
2
1
+ m
2
2
) + (m
2
1
+ m
2
2
)
2
=4m
2
1
m
2
2
4sp
2
1
+ s
2
2sm
2
1
2sm
2
2
+ m
4
1
+ m
4
2
+ 2m
2
1
m
2
2
=4m
2
1
m
2
2
4sp
2
1
+ s
2
2sm
2
1
2sm
2
2
+ m
4
1
+ m
4
2
2m
2
1
m
2
2
=0 (6.99)
p
2
1
=
(s m
2
1
2m
2
m
1
m
2
2
) (s m
2
1
+ 2m
2
m
1
m
2
2
)
4s
(6.100)
6.5. DECAY RATES 187
[p
1
[ =
_
[s (m
1
+ m
2
)
2
][s (m
1
m
2
)
2
]
2
s
=
1/2
(s, m
2
2
, m
2
1
)
2
s
(6.101)
Replacing Eq. (6.101) back in Eq. (6.93) we have
E
1
E
2
v
rel
=
1
2
_
[s (m
1
+ m
2
)
2
][s (m
1
m
2
)
2
] (6.102)
Replacing Eqs. (6.102), and (6.88) in Eq. (6.60)
d =
1
64
2
[/
fi
[
2
d
2s
(2)
[s (m
1
+ m
2
)
2
][s (m
1
m
2
)
2
]
[s (m
1
+ m
2
)
2
][s (m
1
m
2
)
2
]
(6.103)
and, nally
d
d
=
1
64
2
s
_
[s (m
1
+ m
2
)
2
][s (m
1
m
2
)
2
]
[s (m
1
+ m
2
)
2
][s (m
1
m
2
)
2
]
_
1/2
[/[
2
(6.104)
or, in terms of the kinematic function dened in eq. (6.89)
d
d
=
1
64
2
s
1/2
(s, m
2
2
, m
1
2
)
1/2
(s, m
2
2
, m
2
1
)
[/[
2
(6.105)
6.5 Decay Rates
Consider the matrix element of iT in (6.16)
p[iT[k
1
. . . k
n
NR
= (2)
4
(4)
_
p
i
k
i
_
iM
fi
(6.106)
where the initial state is a single particle of momentum p and mass M, while the nal state is
given by n particles of momenta k
i
and masses m
i
, i = 1, . . . , n. We are therefore considering a decay
process.
=
_
. . .
_
(2)
4
(4)
_
p
i
k
j
_
__
dt
_
[/
fi
[
2
1
2E
p
n
j=1
d
3
k
j
(2)
3
2E
k
j
(6.107)
Therefore the dierential probability is
d =
_
. . .
_
(2)
4
(4)
_
p
i
k
j
_
1
2E
p
[/
fi
[
2
dt
n
j=1
d
3
k
j
(2)
3
2E
k
j
(6.108)
Finally we dene the decay rate d as the decay probability in which in the nal state the jth
particle has momentum between k
j
and k
j
+ dk
j
per unit time
d
d
dt
= (2)
4
(4)
_
p
j
k
j
_
1
2E
p
[/
fi
[
2
n
j=1
d
3
k
j
(2)
3
2E
k
j
=
(2)
4
2E
p
[/
fi
[
2
d
(n)
(p; k
1
, k
2
, . . . , k
n
) (6.109)
where
d
(n)
(p; k
1
, k
2
, . . . , k
n
) =
(4)
_
p
j
k
j
_
n
j=1
d
3
k
j
(2)
3
2E
k
j
. (6.110)
and the dierential decay width in the center of mass frame
d =
(2)
4
2E
p
[/
fi
[
2
d
(n)
(p; k
1
, k
2
, . . . , k
n
) (6.111)
6.5.1 Two body decays
We now consider the decay of particle of mass M decaying into two particles of 4momenta p
1
, p
2
and masses m
1
, m
2
. In the CM frame the initial momentum satisfy
p =0 M = E
p
(6.112)
Therefore
d =
(2)
4
2M[4(2)
6
]
[/
fi
[
2
4(2)
6
d
(2)
(p; p
1
, p
2
)
=
1
2
3
M(2)
2
[/
fi
[
2
4(2)
6
d
(2)
(p; p
1
, p
2
)
=
1
32
2
M
[/
fi
[
2
_
4(2)
6
d
(2)
(p; p
1
, p
2
)
(6.113)
6.6. BACKUP 189
where
4(2)
6
d
(2)
(p; p
1
, p
2
) =
(4)
(p p
1
p
2
)
d
3
p
1
E
1
d
3
p
2
E
2
(6.114)
The Dirac delta in eq. (6.110) can be written in the CM frame as
(4)
(p p
1
p
2
) =
(3)
(p p
1
p
2
)(E E
1
E
2
)
=
(3)
(p
1
+p
2
)(M E
1
E
2
) (6.115)
and,
4(2)
6
d
(2)
(p; p
1
, p
2
) =(M E
1
E
2
)
(3)
(p
1
+p
2
)
d
3
p
1
E
1
d
3
p
2
E
2
(6.116)
Comparing with eq. (6.65) we see that the two quantities are the same after the replacing
s M,
p
1
p
1
and p
2
p
2
. Therefore we have from eq. (6.91)
4(2)
6
d
(2)
= d
1/2
(M
2
, m
2
2
, m
2
1
)
2M
2
(6.117)
Replacing back in eq. (6.113)
d
d
=
1
32
2
M
[/
fi
[
2

1/2
(M
2
, m
2
2
, m
2
1
)
2M
2
=
1
64
2
M
3
[/
fi
[
2
1/2
(M
2
, m
2
2
, m
2
1
) (6.118)
By using eq. (6.92) we can write this expression also as
d
d
=
1
64
2
M
3
[/
fi
[
2
2M[p
1
[
=
[p
1
[
32
2
M
2
[/
fi
[
2
(6.119)
as usually written in several texts.
6.6 Backup
Perturbation theory is developed more easily using the Hamiltonian formalism. We therefore consider
a general eld theory with a Hamiltonian
H = H
0
+ H
int
(6.120)
where H
0
is the free Hamiltonian and H
int
is the interaction term. The interaction term will be
considered small. For instance in QED
H
int
=
_
d
3
x 1
int
=
_
d
3
x L
int
(6.121)
with
L
int
= eA
(6.122)
The smallnes of the interaction follows from the fact that the parameter which turns out to be
relevatn for the perturbation expansion is = e
2
/4 1/137.
SS
=(1 + iT)(1 iT
)
= 1 + i(T T
) + TT
= 1 ,
(6.123)
TT
= i(T T
) . (6.124)
Inserting a complete set of states we have
b[TT
[a =i(b[T[a b[T
[a)
b[T
_
n
[nn[
_
T
[a =i
_
b[T[a (a[T[b)
n
b[T[na[T[n
=i (b[T[a a[T[b
n
T
bn
T
an
=i (T
ba
T
ab
) .
(6.125)
if a = b
[T
an
[
2
= i ImT
aa
. (6.126)
Chapter 7
Two body decays
In this chapter we use directly the Feynman rules for Fermions to carry out the calculation of the decay
of the standard model Higgs into a pair of fermions. In chapter 8 we will obtain the corresponding
Feynman rules from the Smatrix expansion.
7.1 Particle decays
Particle decay [5] is the spontaneous process of one elementary particle transforming into other
elementary particles. During this process, an elementary particle becomes a dierent particle with
less mass and an intermediate particle such as W boson in muon decay.
For a particle of a mass M, the dierential decay width according Eq. (6.113), is
d
n
=
(2)
4
2M
[/[
2
d
(n)
(P; p
1
, p
2
, . . . , p
n
) (7.1)
The phase space can be determined from Eq. (6.110)
d
(n)
(P; p
1
, p
2
, . . . , p
n
) =
4
(P
n
i=1
p
i
)
_
n
i=1
d
3
p
i
(2)
3
2E
i
_
. (7.2)
We will keep the d notation until all the integrals get evaluated.
The two-body decays in eq. (6.118) is
d
d
=
1
64
2
M
3
[/
fi
[
2
1/2
(M
2
, m
2
2
, m
2
1
) (7.3)
191
192 CHAPTER 7. TWO BODY DECAYS
7.2 Width decay
Reglas de Feynman:
We consider now a general Yukawa interaction term
L
int
= hHf
1
f
2
(7.4)
For the H f
1
f
2
decay. The interaction between the Higgs boson with fermions
1
is given by the
Yukawa interaction term [1]
L
Higgs
= G
f
(v + H)
2
(f
R
f
L
+ f
L
f
R
)
=
G
f
v
2
ff
G
f
H
2
ff
= m
f
ff m
f
_
G
F
2
_
1/2
ff (7.5)
Such as the electro has acquired a mass m
e
= G
f
/
2. On the other hand the coupling to be

assigned to the process vertex is G
f
2 or m
f
/v =.
The decay process H ff, is displayed in Fig. 7.1
Figure 7.1: Diagrama de proceso H ff
The Feynman rules, to be explained in Chapter 8 are indicated in Fig. 7.2.
In this way the scattering amplitude is
i/= im
f
_
G
F
2
_
1/2
u(s
1
, p
1
)v(s
2
, p
2
). (7.6)
where p
1
, s, p
2
y s
2
are the momentum and spines of fermion and antifermion respectively.
1
In this case we consider only electrons, by the formula is easy generalizable to other fermions
7.2. WIDTH DECAY 193
Figure 7.2: Reglas de Feynman del proceso H ff
For the general case
i/= ihu(s
!
, p
1
)v(s
2
, p
2
). (7.7)
with nal masses m
1
, m
2
Now, having into account that
0
=
0
(u(s
1
, p
1
)v(s
2
, p
2
))
= v
(s
2
, p
2
)(u(s
1
, p
1
))
= v
(s
2
, p
2
)(u
(s
1
, p
1
)
0
)
= v
(s
2
, p
2
)(
0
u(s
1
, p
1
))
= v
(s
2
, p
2
)(
0
u(s
1
, p
1
))
= (v(s
2
, p
2
)u(s
1
, p
1
)).
Squaring /, and summing over possible polarization states of nal particles, we have
s
1
,s
2
[/[
2
= h
2
s
1
,s
2
(u(s
1
, p
1
)v(s
2
, p
2
))(v(s
2
, p
2
)u(s
1
, p
1
)). (7.8)
The several sums in Ec. (7.8) can be calculated by expressing the products uv y vu en in terms of
their components, as follow
s
1
,s
2
(u(s
1
, p
1
)v(s
2
, p
2
))(v(s
2
, p
2
)u(s
1
, p
1
))
=
s
1
,s
2
(u
(s
1
, p
1
)v
(s
2
, p
2
))(v
(s
2
, p
2
)u
(s
1
, p
1
))
=
s
1
,s
2
(u
(s
1
, p
1
)u
(s
1
, p
1
))(v
(s
2
, p
2
)v
(s
2
, p
2
))
=
s
u
(s
1
, p
1
)u
(s
1
, p
1
)
s
2
v
(s
2
, p
2
)v
(s
2
, p
2
)
= (
p
1
+ m
f
)
p
2
m
f
)
= Tr[(
p
1
+ m
f
)(
p
2
m
f
)]. (7.9)
Taking into account that Tr[
] = 0, and from the commutation relations for
matrices
Tr[
] = tr[
+ 2g
]
= Tr[
] + 2g
Tr[1]
= Tr[
] + 2g
4 (Tr[AB] = Tr[BA])
Tr[
] = 4g
.
In this way
Tr[(
p
1
+ m
1
)(
p
2
m
2
)]
= Tr[(
1
+ m
1
)(
2
m
1
)]
= Tr[
1
p
2
m
2
1
+ m
1
2
m
1
m
2
]
= p
1
p
2
tr[
] 4m
1
m
2
= 4g
1
p
2
4m
1
m
2
= 4(p
1
p
2
m
1
m
2
).
and
s
1
,s
2
[/[
2
= 4h
2
(p
1
p
2
m
1
m
2
).
From eq. (6.115)
M =E
1
+ E
2
[p
1
[ = [p
2
[ (7.10)
7.2. WIDTH DECAY 195
Therefore
E
1
E
2
=
M
2
E
2
1
E
2
2
2
(7.11)
p
1
p
2
m
2
f
= E
1
E
2
p
1
p
2
m
1
m
2
= E
1
E
2
+p
2
1
m
1
m
2
=
M
2
E
2
1
E
2
2
2
+p
2
1
m
1
m
2
=
1
2
_
M
2
m
2
1
p
2
1
m
2
2
p
2
1
_
+p
2
1
m
1
m
2
=
1
2
_
M
2
m
2
1
m
2
2
2m
1
m
2
_
=
1
2
_
M
2
(m
1
m
2
)
2
Therefore, the scattering amplitude is
s
1
,s
2
[/[
2
= 2h
2
_
M
2
(m
1
+ m
2
)
2
(7.12)
Replacing back in eq. (7.5)
d
d
=
h
2
32
2
M
3
1/2
(M
2
, m
2
2
, m
2
1
)
_
M
2
(m
1
+ m
2
)
2
(7.13)
After the integration in d
CM
= 4 we have
=
h
2
8M
3
1/2
(M
2
, m
2
2
, m
2
1
)
_
M
2
(m
1
+ m
2
)
2
(7.14)
For m
1
= m
2
= m
f
1/2
(M
2
, m
2
2
, m
2
1
) =M
2
_
1
4m
2
f
M
2
_
1/2
_
M
2
(m
1
+ m
2
)
2
=M
2
_
1
4m
2
f
M
2
_
(7.15)
and therefore
=
h
2
8
M
_
1
4m
2
f
M
2
_
3/2
(7.16)
In the case of the standard model Higgs with mass M
H
decaying to fermion pai, according to the
Lagrangian in eq. (7.5)
(H ff) =
M
H
m
2
f
G
F
4
2
_
1 4
m
2
f
M
2
H
_
3/2
, (7.17)
In the limit m
f
M
H
this expression reduces to
(H ff) =
M
H
m
2
f
G
F
4
2
. (7.18)
7.3 e
+
e
L =
e
2
s
_
v(k
2
)
u(k
1
)
_
v(k
2
)
u(k
1
)
(7.19)
Chapter 8
Feynman Rules
When the case of interacting elds are considered, the particles can be created, destroyed and scat-
tered. In essence this requires solving the coupled non-linear eld equations for given conditions.
This is an extremely dicult problem which has only been solved in perturbation theory.
In the Heisenberg picture, which we have so far been using, this program is still very complex, and
it was decisive for the successful development of the theory to work instead in the interaction picture.
In section 8.1 we write the Smatrix expansion derived in Chapter 6, in the interaction picture. In
section 8.3 we show how to use the Wick expansion to calculate Smatrix elements involving scalars
and spinors.
8.1 Interaction picture
This part is based in [3]. In the Schr odinger Picture (SP) the time dependence is carried by the
states according to the Scrodinger equation
labeleq : 84fi

t
[a, t
S
= i
d
dt
[a, t
S
= H[a, t
S
(8.1)
With the solution given in Eq. (??)
[a, t
S
= U(t, t
i
)[a
S
. (8.2)
where U is the unitary operator [see Eq. (??)]
U U(t, t
i
) = e
iH(tt
i
)
. (8.3)
Given the state [a, t
S
in the SP, in the Heisenberg picture (HP) we dened the state
[a
H
= U
[a, t
S
= [a
S
(8.4)
197
198 CHAPTER 8. FEYNMAN RULES
Si O
S
in an operator in the SP, the corresponding Heisenberg operator is dened as
O
H
(t) = U
O
S
U (8.5)
Hence, the transformation from HP to SP is unitary. At t = t
i
, states and operators in the two
pictures are the same. We see from Eq. (8.4) that in the HP state vectors are constant in time, while
from Eq. (8.5) the Heisenberg operators evolve with time. Is convenient to keep the temporal label
in the Heisenberg states
[a
H
= [a, t
i
H
(8.6)
Eq. (8.5) ensures the invariance of matrix elements and commutation relations:
S
b, t[ O
S
[a, t
S
=
S
b, t[ UO
H
(t)U
[a, t
S
=
H
b, t
i
[O
H
(t)[a, t
i
H
(8.7)
_
O
S
, P
S
= c
_
O
H
(t), P
H
(t)
= c (8.8)
where c is a constant.
Dierentiation of Eq. (8.5)
d
dt
O
H
(t) =
_
d
dt
U
_
O
S
U + U
O
S
d
dt
U
=iH U
O
S
U + U
O
S
U(iH)
=i(O
H
H HO
H
) , (8.9)
gives the Heisenberg equation of motion
i
d
dt
O
H
(t) =
_
O
H
(t), H
(8.10)
The interaction picture (IP) arises if the Hamiltonian is split into two parts
H = H
0
+ H
I
. (8.11)
In quantum eld theory H
I
will describe the interaction between two elds, themselves described by
H
0
IP is related to the SP by the unitary transformation
U
i
U
i
(t, t
i
) = e
iH
i
(tt
i
)
, (8.12)
8.1. INTERACTION PICTURE 199
in this way,
[a, t
I
= U
0
[a, t
S
, (8.13)
and
O
I
(t) = U
0
O
S
U
0
. (8.14)
Thus the relation between IP and SP is similar to that between HP and SP, but with the unitary
transformation U
0
involving only the noninteracting Hamiltonian H
0
. Note that both the vector
states as the operators in the IP are time-dependent.
Dierentiating Eq. (8.14) gives the dierential equation of motion operators in the IP:
i
d
dt
O
I
(t) =
_
O
I
(t), H
0
(8.15)
Substituting Eq. (8.13) into the Scr odinger Eq. (??), one obtains the equation of motion of state
vectors in the IP, If the system is described by a time-dependent state vector [(t)
i
d
dt
[a, t
S
=H
S
[a, t
S
i
d
dt
(U
0
[(t)) =H
S
U
0
[(t)
i
_
d
dt
U
0
_
[(t) + iU
0
d
dt
[(t) =H
S
U
0
[(t)
U
0
H
0
[(t) + iU
0
d
dt
[(t) =H
S
U
0
[(t)
U
0
H
0
[(t) + iU
0
d
dt
[(t) =(H
0
+ H
S
I
)U
0
[(t)
iU
0
d
dt
[(t) =H
S
I
U
0
[(t)
i
d
dt
[(t) =U
0
H
S
I
U
0
[(t) (8.16)
i
d
dt
[(t)
I
= H
I
I
[(t)
I
, (8.17)
where, as in Eq. (8.14)
H
I
I
= e
iH
S
0
(tt
i
)
H
S
I
e
iH
S
0
(tt
i
)
(8.18)
is the interaction Hamiltonian in the IP, with H
S
I
and H
S
0
being the interaction and free-eld Hamil-
tonian in the SP. From now on we shall omit the labels I, used in the equations to distinguish the
IP, as we shall be working exclusively in the IP in what follows.
Eq. (8.17) is a Scrodinger-like equation with the time dependent Hamiltonian H
I
(t). With the
interaction switched o (i.e. we put H
I
= 0), the state vector is constant in time. The interaction
leads to the state [(t) changing with time. Given that the system is in a state [i at an initial time
t = t
i
, i.e.
[(t
i
) = [i , (8.19)
the solution of Eq. (8.17) with this initial condition gives the state [(t) of the system at any other
time t. It follows from the Hermicity of the operator H
I
(t) that the time development of the state
[(t) according to Eq. (8.17) is a unitary transformation. Accordingly it preserves the normalization
of states
(t)[(t) = const. (8.20)
and, more generally, the scalar product.
Clearly the formalism which we are here developing is not appropriate for the description of
bound states but it is particularly suitable for scattering processes. In a collision processes the state
vector [i will dene an initial state, long before the scattering occurs (t
i
= ), by specifying a
denite number of particles, with denite properties and far apart from each other so that they do
not interact. (For example [i would specify a denite number of electrons, and positrons with given
momenta and spins). In the scattering process, the particles will come close together, collide (i.e
interact) and y apart gain. Eq. (8.17) determines the state [(t) into which the initial state
[() = [i , (8.21)
evolves at t = , long after the scattering is over and all particles are for apart again. The Smatrix
relates [() to () and is dened by
[() = S[() = S[i , (8.22)
A collision can lead to many dierent nal states [f, and all these possibilities are constrained
within [().
The transition probability is given by
[f[()[
2
= [f[S[i[
2
S
2
fi
, (8.23)
where S
fi
is the corresponding probability amplitude.
8.1. INTERACTION PICTURE 201
In order to calculate the Smatrix we must solve Eq. (8.17) for the initial condition (8.19). These
equations can be combined into the integral equation
d[(t) =idt H
I
(t)[(t)
_
|(t)
|()
d[(t) =i
_
t
dt
1
H
I
(t
1
)[(t
1
)
[(t) [() =i
_
t
dt
1
H
I
(t
1
)[(t
1
)
(8.24)
[(t) = [i i
_
t
dt
1
H
I
(t
1
)[(t
1
) . (8.25)
In the limit t
[() = S
(0)
[i i
_

dt
1
H
I
(t
1
)[(t
1
) . (8.26)
where
S
(0)
= 1 . (8.27)
From Eq. (8.25) we can obtain [(t
1
) at next order:
[(t
1
) =[i i
_
t
1
dt
2
H
I
(t
2
)[(t
2
) . (8.28)
This equation then can be solved iteratively. If H
I
is small we can solve this equation by iteration
[(t) = [i + (i)
_
t
dt
1
H
I
(t
1
)[i + (i)
2
_
t
dt
1
_
t
1
dt
2
H
I
(t
1
)H
I
(t
2
)[(t
2
) . (8.29)
In the limit t
[(t) =
_
S
(0)
+ (i)
_

dt
1
H
I
(t
1
)
_
[i + (i)
2
_

dt
1
_
t
1
dt
2
H
I
(t
1
)H
I
(t
2
)[(t
2
)
=
_
S
(0)
+ S
(1)
_
[i + (i)
2
_

dt
1
_
t
1
dt
2
H
I
(t
1
)H
I
(t
2
)[(t
2
) , (8.30)
where
S
(1)
= (i)
_

dt
1
H
I
(t
1
) . (8.31)
The next order of Eq. (8.29) is
[(t) =[i + (i)
_
t
dt
1
H
I
(t
1
)[i + (i)
2
_
t
dt
1
_
t
1
dt
2
H
I
(t
1
)H
I
(t
2
)
_
[i + (i)
_
t
2
dt
3
H
I
(t
3
)[i + (i)
2
_
t
2
dt
3
_
t
3
dt
4
H
I
(t
3
)H
I
(t
4
)[(t
4
)
_
(8.32)
[(t) =[i + (i)
_
t
dt
1
H
1
(t
1
)[i + (i)
2
_
t
dt
1
_
t
1
dt
2
H
I
(t
1
)H
I
(t
2
)[i
+ (i)
3
_
t
dt
1
_
t
1
dt
2
_
t
2
dt
3
H
I
(t
1
)H
I
(t
2
)H
1
(t
3
)[i
+ (i)
4
_
t
dt
1
_
t
1
dt
2
_
t
2
dt
3
_
t
3
dt
4
H
I
(t
1
)H
I
(t
2
) H
I
(t
3
)H
I
(t
4
)[(t
4
) (8.33)
In the limit t
[(t) =
_
S
(0)
+ S
(1)
+ S
(2)
+ S
(3)
_
[i
+ (i)
4
_

dt
1
_
t
1
dt
2
_
t
2
dt
3
_
t
3
dt
4
H
I
(t
1
)H
I
(t
2
) H
I
(t
3
)H
I
(t
4
)[(t
4
) (8.34)
where
S
(2)
=(i)
2
_

dt
1
_
t
1
dt
2
H
I
(t
1
)H
I
(t
2
)
S
(3)
=(i)
3
_

dt
1
_
t
1
dt
2
_
t
2
dt
3
H
I
(t
1
)H
I
(t
2
)H
1
(t
3
) (8.35)
and so on we obtain the Smatrix
S =
n=0
S
(n)
=1 +
n=1
(i)
n
n!
_

dt
1
_
t
1
dt
2
. . .
_
t
n1
dt
n
H
I
(t
1
)H
I
(t
2
) . . . H
I
(t
n
) . (8.36)
8.2. ATOMIC DECAY 203
8.2 Atomic decay
Here we follow closelly [15] chapter 3.
For the atomic decay at rst order in perturbation theory, we have
S
(1)
= i
_

dt [H
I
(t)[ . (8.37)
where the Hamiltonian in the interaction picture is
H =H
A
+ H
EM
+ H
I
(t) = H
0
+ H
I
(t) , (8.38)
where, following the denition of the interaction picture
a
(r, t) =e
iH
A
t
a
(r)
=U
A
(t)
a
(r) , (8.39)
H
I
(t) = U
1
A
H
I
(t)U
A
. (8.40)
The several terms of the Hamiltonia are
H
A
=
p
2
e
2m

Z
r
e
, (8.41)
H
EM
=
1
2
_
d
3
r
_
:
2
(r, t) : + : B
2
(r, t) :
_
, (8.42)
H
I
(t) =
_
e
2m
_
p
e
A(r
e
, t) +A(r
e
, t) p
e
+
e
2
2m
A
2
(r
e
, t)
__
, (8.43)
and = e
2
/(4). The states as dened as
[a, n =
a
(r
e
)[n . (8.44)
The scalar product of the atomic states requires an integration over the coordinate r
e
a
, n
[a, n =
_
d
3
r
a
(r
e
)
a
(r
e
)n
[n. (8.45)
For one photon decay of an initial atomic state a into a nal atomic state b and a photon of
energy
n
and polarization , the states are
[ =[, 0 =
a
(r
e
)[0
[ =[b, 1
n
=
b
(r
e
)[1
n
, (8.46)
where
[1
n
= a
n
[0 (8.47)
is the one-photon state with frequency
n
and polarization . Since
1
n
[A
2
[0 1
n
[(a
n
a
n
+ a
n
a
n
+ a
n
a
n
)[0 = 0[1
n
= 0 (8.48)
and
(p
e
A(r
e
, t) +A(r
e
, t) p
e
)
a
(r
e
) =i (
e
A(r
e
, t) +A(r
e
, t)
e
)
a
(r
e
)
=i (
e
A)
a
+A
e
a
+A
e
=i 2A
e
a
, (8.49)
where A = 0 was used.
With all of this we have a simplied formula
S
ba
=i
_

dtb, 1
n
[U
1
A
_
ie
m
A(r
e
, t)
e
_
U
A
[a, 0
S
ba
=i
_

dtb, 1
n
[U
1
A
_
ie
m
A(r
e
, t)
e
_
U
A
[a, 0
S
ba
=i
_

dtb, 1
n
[e
iE
b
t
_
ie
m
A(r
e
, t)
e
_
e
iEat
[a, 0
=
_

dt e
i(E
b
Ea)t
_
d
3
r
e
1
n
[
b
(r
e
)
e
m
(A(r
e
, t)
e
)
a
(r)[0
=
_

dt e
i(E
b
Ea)t
e
m
_
d
3
r
e
b
(r
e
)
e
a
(r) 1
n
[A(r
e
, t)[0. (8.50)
8.2. ATOMIC DECAY 205
Since
1
n
[A(r
e
, t)[0 =
2
n
L
3
n
1
n
[a
[0e
ik
n
xe
=
2
n
L
3
n
1
n
[a
[0e
iknxe
=
1
2
n
L
3
n
1
n
[1
n
e
ik
n
xe
=
1
2
n
L
3
n
e
intknre
. (8.51)
Inserting this expression into (8.50) gives
S
ba
= i2(E
b
+
n
E
a
)
1
L
3
M
ba
, (8.52)
where the decaya amplitude M
ba
is
M
ba
= i
e
m
1
2
n
_
d
3
r
e
e
iknre
b
(r
e
)
n

e
a
(r) . (8.53)
In most atomic decays, the energy of the emmited photon, which is equal to
n
= E
b
E
a
, is much
less than 1/R, where R is the size of the atomic system, and hence the maximum range of the integral
over r
e
. In this case, the dipole approximation
e
iknre
1 (8.54)
is extremely good. Then we have
M
ba
=
e
m
1
2
n
n
p
ba
, (8.55)
where
p
ba
= i
_
d
3
r
e
b
(r
e
)
e
a
(r
e
). (8.56)
The diferential decay rate, and using
[2(E
f
E
i
)]
2
= 2(E
f
E
i
)T , (8.57)
we have
W
ba
= lim
T
[S
ba
(T/2, T/2)[
2
T
= lim
T
1
T
2(E
b
+
n
E
a
)T
1
L
3
[M
ba
[
2
=2(E
b
+
n
E
a
)
1
L
3
[M
ba
[
2
(8.58)
Then
W
ba
=
2
V
(E
b
+
n
E
a
)
e
2
2
n
m
2
[
n
p
ba
[
2
. (8.59)
Summing over all nal photon states to get the total a b decay rate gives
W
ba
=
n,
2
V
(E
b
+
n
E
a
)
e
2
2
n
m
2
[
n
p
ba
[
2
=
n
2
V
(E
b
+
n
E
a
)
e
2
2
n
m
2
[
n
p
ba
[
2
. (8.60)
Using
V
(2)
3
_
d
3
k (8.61)
W
ba
=
_
d
3
k
(2)
2
(E
b
+
n
E
a
)
e
2
2m
2
[
n
p
ba
[
2
=
e
2
2m
2
_
d
3
k
(2)
2
(E
b
+
n
E
a
)[
n
p
ba
[
2
=
e
2
2m
2
_
[k[
2
d[k[
_
d
d
3
k
(2)
2
(E
b
+ E
a
)[
n
p
ba
[
2
. (8.62)
By using [k[ = , we have
W
ba
=
e
2
2m
2
_
d

(2)
2
(E
b
+ E
a
)
_
d[
n
p
ba
[
2
=
e
2
4
_

2m
2
_
_
d[
n
p
ba
[
2
, (8.63)
8.3. YUKAWA INTERACTION 207
where = E
b
E
a
. Integrating over all directions

k of the outgoing photon
_
d[
n
p
ba
[
2
=
8
3
[p
ba
[
2
. (8.64)
Then the total rate for the decay of the state a into b is
W
ba
=
e
2
4
_
4
3m
2
_
[p
ba
[
2
. (8.65)
8.3 Yukawa interaction
As a concrete example, we take a theory with a fermion eld and scalar eld, which interact via the
Yukawa interaction [4]:
L
int
= h. (8.66)
Let the quantum of the eld be denoted by B, since the particle is a boson. The quanta of the
fermionic eld will be called electrons. The mass of B is M, and the mass of the electron by
m. Suppose M > 2m, so that kinematically it is possible to have the B particle decay into an
electron-positron pair. The process is denoted by
B(k) e
(p) + e
+
(p
) , (8.67)
where k, p, p
are the 4momenta of the particles.

For the interaction Hamiltonian we have
1
I
= h : : (8.68)
where the required ordered product will be explained in next section. The term linear in the inter-
action Hamiltonian in the Smatrix. It is
S
(1)
=
=ih
_
d
4
x : : . (8.69)
S
(1)
= ih
_
d
4
x : (
+
+
)(
+
+
)(
+
+
) : . (8.70)
L
int
=h :
_
+
+

_
(
+
+
) (
+
+
) :
= :

+
+

+
+

+
+

+
+

:
=
+
+ (8.71)
To check that only the ordered terms are dierent from zero we can analyse the full terms for
initial and nal states dened as [i = [0
, 0
, 1
y f[ = 1
, 1
, 0
[.
+
[n
[n 1
[
+
n + 1
[ (8.72)
y
[n
[n + 1
n 1
[ (8.73)
Y lo mismo tendremos bien sea para un campo fot onico o fermi onico.
El Langrangiano de interaccion de nuestro interes est a dado por
L
int
= h
(8.74)
Que en terminos de las componentes + y de los campos se puede expresar como
De desarrollo del langrangiano en las componentes de los campos, vemaos que terminos con-
tribuyen al elemento de matriz
1
, 1
, 0
+
[0
, 0
, 1
, 2
, 0
[0
, 0
, 0
= 0
1
, 1
, 0
[0
, 0
, 1
, 2
, 0
[0
, 0
, 2
= 0
1
, 1
, 0
+
[0
, 0
, 1
, 0
, 0
[0
, 0
, 0
= 0
1
, 1
, 0
[0
, 0
, 1
, 0
, 0
[0
, 0
, 2
= 0
1
, 1
, 0
+
[0
, 0
, 1
, 2
, 0
[0
, 0
, 0
= 0
1
, 1
, 0
[0
, 0
, 1
, 2
, 0
[0
, 0
, 2
= 0
1
, 1
, 0
+
[0
, 0
, 1
, 0
, 0
[0
, 0
, 0
, = 0
1
, 1
, 0
[0
, 0
, 1
, 0
, 0
[0
, 0
, 2
= 0
8.3. YUKAWA INTERACTION 209
Figure 8.1: Feynman diagrams for B e
+
e
The only term that contributes to the matrix element of the process is
ih
_
d
4
x
+
. (8.75)
Let us dene the oneparticle states as in eq. (??)
[B(p)
_
1
V
a
p
[0 (8.76)
From eq. (??)
[e
(p, s)
_
1
V
a
s
(p)[0
[e
+
(p, s)
_
1
V
b
s
(p)[0 , (8.77)
Using the commutation relations, our states are then normalized as
B(p)[B(p
) =
(2)
3
V

3
(p p
)
e
(p, s)[e
(p
, s
) =
(2)
3
V

ss

3
(p p
)
e
+
(p, s)[e
+
(p
, s
) =
(2)
3
V

ss

3
(p p
) (8.78)
As established in Sec. 3.1, it is convenient to work in the discrete limit where (??)
3
(0) =
V
(2)
3
. (8.79)
Now we can write down the action of various eld operators on dierent one particles states. Using
the Fourier decomposition of the scalar eld in eq. (??), and taking into account that a
p
[0 = 0, we
have
+
(x)[B(k) =
_
d
3
p
1
(2)
3
_
2
p
a
p
e
ipx
[B(k)
=
_
d
3
p
1
(2)
3
_
2
p
a
p
e
ipx
1
V
a
k
[0
=
_
d
3
p
1
(2)
3
_
2
p
V
e
ipx
[a
p
, a
k
][0 . (8.80)
By usinbg the commutation relations in eq. (??) we have
+
(x)[B(k) =
_
d
3
p
(3)
(p k)
_
2
p
V
e
ipx
[0 (8.81)
+
(x)[B(k) =
1
2
k
V
e
ikx
[0 (8.82)
Similarly, we have
+
(x)[B(k) =
1
2
k
V
e
ikx
[0
+
(x)[e
(p, s) =
1
_
2E
p
V
u
s
(p)e
ipx
[0
+
(x)[e
+
(p
, s
) =
1
_
2E
p
V
v
s
(p
)e
ip
x
[0 , (8.83)
where
k
and E
p
represent the energies of the scalar and the electron for the 3-momenta in the
subscripts.
Similarly, for the adjoint operators
B(k)[
(x) =0[
1
2
k
V
e
ikx
e
(p, s)[
(x) =0[
1
_
2E
p
V
u
s
(p)e
ipx
e
+
(p
, s
)[
(x) =0[
1
_
2E
p
V
v
s
(p
)e
ip
x
, (8.84)
8.4. WICK THEOREM 211
In the lowest order the only term which contributes to the matrix element is the term shown in
Eq. (8.75) The matrix element at rst order in Eq. (8.95), between the initial and the nal state is
then
S
(1)
fi
= ih
_
d
4
x
(p)e
+
(p
B(k)
_
. (8.85)
Using Eqs. (8.83)(8.84), we obtain
S
(1)
fi
=(ih) u
s
(p)v
s
(p
)
_
d
4
x e
i(p+p
k)x
1
2
k
V
1
_
2E
p
V
1
_
2E
p
V
. (8.86)
Since
_
d
4
x e
i(p+p
k)x
= (2)
4
4
(k p p
) , (8.87)
we obtain
S
(1)
fi
=
_
1
2
k
V
1
_
2E
p
V
1
_
2E
p
V
_
(2)
4
4
(k p p
) [(ih) u
s
(p)v
s
(p
)] (8.88)
Comparing with Eq. (??) we have therefore that the relativistic matrix element is
i/
fi
= (ih) u
s
(p)v
s
(p
) , (8.89)
and everything else is the history presented in Chapter 7.
8.4 Wick Theorem
From [4]. The normal ordering procedure involved putting all the annihilation operators to the right
of all creation operators so that it annihilates the vacuum. But the time ordering raises complications
because in it all operators at earlier times must be further to the right. So creation operators at later
times would be to the right of annihilation operators at later times, contrary to what we need for
normal ordering. The advantage of normal ordered products is that their expectation values vanish
in the vacuum.
If H
I
contains an even number of fermion factors, we can use the timeordered product T. . .
of n factors to write this expression in the equivalent form. For S
(2)
we have for example
_

dt
1
_

dt
2
TH
I
(t
2
)H
I
(t
2
) =
_

dt
1
_

dt
2
(t
2
t
1
)H
I
(t
2
)H
I
(t
2
) +
_

dt
1
_

dt
2
(t
1
t
2
)H
I
(t
1
)H
I
(t
2
)
(8.90)
_

dt
1
_

dt
2
TH
I
(t
2
)H
I
(t
2
) =2
_

dt
1
_
t
1
dt
2
H
I
(t
2
)H
I
(t
2
) . (8.91)
S =1 +
n=1
(i)
n
n!
_

dt
1
_

dt
2
. . .
_

dt
n
TH
I
(t
1
)H
I
(t
2
) . . . H
I
(t
n
) , (8.92)
In terms of the Hamiltonian density, we have
S = 1 +
n=1
(i)
n
n!
_

_
d
4
x
1
d
4
x
2
. . . d
4
x
n
T1
I
(x
1
)1
I
(x
2
) . . . 1
I
(x
n
) , (8.93)
In the above perturbation formalism the states [i and [f are, as usual, eigenstates of the unperturbed
free-eld Hamiltonian H
0
. As such can be introduced inside the integrals
S
fi
=f[S[i
=1 +
n=1
(i)
n
n!
_

_
d
4
x
1
d
4
x
2
. . . d
4
x
n
f[ T1
I
(x
1
)1
I
(x
2
) . . . 1
I
(x
n
)[i . (8.94)
For example, at rst order
S
(1)
fi
=f[S
(1)
[i
=f[ i
_
d
4
x
1
T1
I
(x
1
)[i
=i
_
d
4
x
1
f[ : 1
I
(x
1
) : [i . (8.95)
In order to evaluate this integrals we need to write the time ordered product in terms of the elds.
This can done by induction. We start by considering the simple no trivial case with two scalar elds
T(x
1
)(x
2
) =: (x
1
)(x
2
) : +(x
1
)(x
2
) (8.96)
The same expression can be obtained for fermions. Generalizing the results for n scalar or fermion
elds, but with an even number of fermions elds, we have the Wick theorem
T(x
1
)(x
2
)(x
3
) (x
n
) = : (x
1
)(x
2
)(x
3
) (x
n
) : +
+ (x
1
)(x
2
) : (x
3
) (x
n
) :
+ : (x
1
) (x
2
)(x
3
) (x
n
) : + (8.97)
For details of the full result see for example [4].
8.5. SCATTERING 213
8.5 Scattering
From the previous calculation we have
S
(n)
=
(i)
n
n!
_

_
d
4
x
1
d
4
x
2
. . . d
4
x
n
T1
I
(x
1
)1
I
(x
2
) . . . 1
I
(x
n
) . (8.98)
The relevant term for the scattering
e
(p
1
) + e
(p
2
) e
(p
1
) + e
(p
2
) (8.99)
is
S
(2)
=
(i)
2
2!
_ _
d
4
x
1
d
4
x
2
T1
I
(x
1
)1
I
(x
2
)
=
(ih)
2
2!
_ _
d
4
x
1
d
4
x
2
T: ()
x
1
()
x
2
=
(ih)
2
2!
_ _
d
4
x
1
d
4
x
2
: ()
x
1
()
x
2
: +
(ih)
2
2!
_ _
d
4
x
1
d
4
x
2
: ()
x
1
()
x
2
: +
(8.100)
The rst term corresponds to two disconnected Feynman diagrams that does not contribute to the S
matrix. For the process at hand, we want terms where four fermionic operators are not contracted,
corresponding to the particles in the initial and nal states. The second term in the previous
expansion of the Wick theorem is the only satisfying this requirement. In this way
S
(2)
(e
+
e
e
+
e
) =
(ih)
2
2!
_ _
d
4
x
1
d
4
x
2
(x
1
)(x
2
) : ()
x
1
()
x
2
: (8.101)
The Wick contraction can be written as:
(x
1
)(x
2
) =0[T(x
1
)(x
2
)[0
=i
F
(x
1
x
2
) (8.102)
since
(x) =
+
(x) +
(x) , (8.103)
T [(x
1
), (x
2
)] = (t
1
t
2
)(x
1
)(x
2
) + (t
2
t
1
)(x
2
)(x
1
) (8.104)
0[T [(x
1
), (x
2
)] [0 =(t
1
t
2
)0[(x
1
)(x
2
) + (t
2
t
1
)0[(x
2
)(x
1
)[0
=(t
1
t
2
)0[
+
(x
1
)
(x
2
) + (t
2
t
1
)0[
+
(x
2
)
(x
1
)[0
=0[(t
1
t
2
)(x
1
)
+
(x
2
) + (t
2
t
1
)0[
+
(x
2
)
(x
1
)[0 (8.105)
with
+
(x) =
_
d
3
p
1
(2)
3
_
2
p
a
p
e
ipx
(x) =
_
d
3
p
1
(2)
3
_
2
p
a
p
e
ipx
, (8.106)
we have
0[T((x
1
)(x
2
))[0 =(t
1
t
2
)0[
_
d
3
p
1
(2)
3
_
2E
p
1
a
p
1
e
ip
1
x
1
_
d
3
p
2
(2)
3
_
2E
p
2
a
p
2
e
ip
2
x
2
[0
+ (t
2
t
1
)0[
_
d
3
p
2
(2)
3
_
2E
p
2
a
p
2
e
ip
2
x
2
_
d
3
p
1
(2)
3
_
2E
p
1
a
p
1
e
ip
1
x
1
[0
=(t
1
t
2
)
_ _
d
3
p
1
d
3
p
2
(2)
6
_
2E
p
1
_
2E
p
2
e
ip
1
x
1
e
ip
2
x
2
0[ a
p
1
a
p
2
[0
+ (t
2
t
1
)
_ _
d
3
p
2
d
3
p
1
(2)
6
_
2E
p
2
_
2E
p
1
e
ip
2
x
2
e
ip
1
x
1
0[ a
p
2
a
p
1
[0
(8.107)
=(t
1
t
2
)
_ _
d
3
p
1
d
3
p
2
(2)
6
_
2E
p
1
_
2E
p
2
e
ip
1
x
1
e
ip
2
x
2
0[ a
p
1
a
p
2
[0
+ (t
2
t
1
)
_ _
d
3
p
2
d
3
p
1
(2)
6
_
2E
p
2
_
2E
p
1
e
ip
2
x
2
e
ip
1
x
1
0[ a
p
2
a
p
1
[0
(8.108)
On the other hand
: ()
x
1
()
x
2
:= :
(x
1
)
(x
1
)
(x
2
)
(x
2
) : . (8.109)
From the Fourier expansions in eqs. (??), (??) we have that a
s
and a
s
are the creation and
annihilation operators for particles. As we have only particles (and not antiparticles) in the initial
and nal states, the only non-zero contribution of the ordered product in eq. (8.109) must have the
order a
a a. As
+
and
are associated to a and a
respectively, the only non-zero contribution

from the ordered fermion product is
8.5. SCATTERING 215
: ()
x
1
()
x
2
:=
(x
1
)
(x
2
)
+
(x
1
)
+
(x
2
) . (8.110)
The relevant Smatrix element then reads
S
(2)
fi
=
(ih)
2
2
_ _
d
4
x
1
d
4
x
2
e
(p
1
)e
(p
2
)[i
F
(x
1
x
2
)
(x
1
)
(x
2
)
+
(x
1
)
+
(x
2
)[e
(p
1
)e
(p
2
)
=
(ih)
2
2
_ _
d
4
x
1
d
4
x
2
i
F
(x
1
x
2
)
e
(p
1
)e
(p
2
)[
(x
1
)
(x
2
)
+
(x
1
)
+
(x
2
)[e
(p
1
)e
(p
2
)
=
(ih)
2
2
_ _
d
4
x
1
d
4
x
2
_
d
4
q
(2)
4
i
F
(q)e
iq(x
1
x
2
)
e
(p
1
)e
(p
2
)[
(x
1
)
(x
2
)
+
(x
1
)
+
(x
2
)[e
(p
1
)e
(p
2
) (8.111)
The two particle Fock state is, after proper normalization
[e
(p
1
)e
(p
2
) =
1
V
a
s
(p
2
)a
s
(p
1
)[0 (8.112)
Therefore
+
(x
1
)
+
(x
2
)[e
(p
1
)e
(p
2
) =
_
d
3
k
2E
k
V
_
d
3
k
2E
k
V
u
(k)u
(k
)e
ikx
1
e
ik
x
2
a
s
(k)a
s
(k
)a
s
(p
2
)a
s
(p
1
)[0 (8.113)
+
(x
1
)
+
(x
2
)[e
(p
1
)e
(p
2
) =
1
_
2E
p
1
V
1
_
2E
p
2
V
_
u
(p
1
)u
(p
2
)e
ip
1
x
1
e
ip
2
x
2
u
(p
2
)u
(p
1
)e
ip
2
x
1
e
ip
1
x
2
[0
(8.114)
Following similar steps, we nd
e
(p
1
)e
(p
2
)[
(x
1
)
(x
2
) =
1
_
2E
p
1
V
1
_
2E
p
2
V
0[
_
u
(p
1
) u
(p
2
)e
ip
1
x
1
e
ip
2
x
2
u
(p
2
)u
(p
1
)e
ip
2
x
1
e
ip
1
x
2
_
(8.115)
As expected, the nal result can be written in term of three dierent factors: the momentum con-
servation, normalization, and the relativistic amplitude
S
(2)
fi
= i(2)
4
4
_
i=1,2
p
i
f=1,2
p
f
_

i=1,2
1
2E
i
V
f=1,2
1
_
2E
f
V
/
fi
(8.116)
where
/
fi
= (ih)
2
_
u
(p
2
) u
(p
1
)
F
(p
1
p
2
)u
(p
1
)u
(p
2
) u
(p
1
) u
(p
2
)
F
(p
1
p
1
)u
(p
1
)u
(p
2
)
(8.117)
The two contributions are displayed in Fig. 8.2 Since
F
(q) =
1
q
2
m
2
(8.118)
In the limit q
2
m
2
F
=
1
m
2
(8.119)
/
fi
=
h
2
m
2
_
u
(p
2
)u
(p
1
) u
(p
1
)u
(p
2
) u
(p
1
)u
(p
1
) u
(p
2
)u
(p
2
)
=
h
2
m
2
[ u(p
2
)u(p
1
) u(p
1
)u(p
2
) u(p
1
)u(p
1
) u(p
2
)u(p
2
)] (8.120)
For one interaction of type we should have
/
fi
=
h
2
m
2
[ u(p
2
)u(p
1
) u(p
1
)u(p
2
) u(p
1
)u(p
1
) u(p
2
)u(p
2
)] (8.121)
For the interaction of a fermion pair with W
, we know from the standard model Lagrangian [1],

that
g
2
2
(1
5
) (8.122)
Therefore in this case
=
(1
5
) (8.123)
8.5. SCATTERING 217
e
(p
1
)
e
(p
2
) e
(p
1
)
e
(p
2
)
x
1
x
2
B(p
1
p
2
)
e
(p
1
)
e
(p
2
) e
(p
2
)
e
(p
1
)
x
1
x
2
B(p
1
p
1
)
Figure 8.2: fermion scattering
For p M
2
W
the analysis is similar to the previous one with
W
(x
1
)W
(x
2
) =0[TW
(x
1
)W
(x
2
)[0
_
d
4
q
(2)
4
g
M
2
W
e
iq(x
1
x
2
)
(8.124)
For the process
e
(p) +
(k)
(p
) +
e
(k
) (8.125)
the global coupling for p M
2
W
is
g
2
8M
2
W
=
G
F
2
(8.126)
After the replacement G
F
/
2 h
2
/m
2
, we have
S
(2)
fi
= i(2)
4
4
(p
1
+ p
2
p
1
p
2
)
1
2E
1
V
1
2E
2
V
1
_
2E
1
V
1
_
2E
2
V
/
fi
(8.127)
where
/
fi
=
G
F
2
u
e
(p
2
)u
e
(p
1
) u
(p
1
)u
(p
2
) (8.128)
The corresponding Feynman diagram is shown in Fig. 8.3 Therefore we have
/
fi
=
G
F
2
u
e
(p
2
)
(1
5
)u
e
(p
1
) u
(p
1
)
(1
5
)u
(p
2
) (8.129)
We now must sqaure the scattering amplitude, /, and summing up over nal spin states, and
averaging over the intial spin states, as we did in Eq. (??). The result that will be obtained in detail
in Chapter 9 for the muondecay is
[/[
2
= 64 G
2
F
(p
1
p
2
)(p
1
p
2
) (8.130)
From Eq. (??)
d
d
=
1
64
2
s
_
s m
2
s m
2
e
_
[/[
2
(8.131)
8.5. SCATTERING 219
e
(p
1
)
(p
2
)

(p
1
)
e
(p
2
)
x
1
x
2
W
(p
1
p
2
)
Figure 8.3: scattering with four fermions
The center of mass (CM) frame is dened by the condition in Eq. (??):
p
1
+p
2
= 0 (8.132)
The function in Eq. (8.127)
(4)
(p
1
+ p
2
p
1
p
2
) =
(3)
(p
1
+p
2
p
1
p
2
)(E
1
+ E
2
E
1
E
2
) (8.133)
implies
p
1
+p
2
p
1
p
2
= 0
CM
_
p
1
= p
2
p
1
= p
2
(8.134)
Moreover
s = E
1
+ E
2
(8.135)
In the CM frame
s =(E
1
+ E
2
)
2
=
_
_
p
2
1
+ m
2
1
+
_
p
2
2
+ m
2
2
_
2
=
_
_
p
2
1
+ m
2
e
+
_
p
2
1
+ m
2
e
_
2
(8.136)
Therefore
E
2
= [p
1
[ (8.137)
We already have the expression for [p
1
[ as given in eq. (??). In this case m
2
= 0, and m
1
= m
e
, so
that
[p
1
[ =
s m
2
e
2
s
(8.138)
From (8.137)
s =E
1
+ E
2
=E
1
+[p
1
[ (8.139)
E
1
=
s [p
1
[
=
s +
s + m
2
e
2
s
=
2s s + m
2
e
2
s
=
s + m
2
e
2
s
(8.140)
Then, by using Eqs. (8.134), (8.137) and (8.138), and (8.140), we have
p
1
p
2
=E
1
E
2
p
1
p
2
=E
1
[p
1
[ +p
2
1
=
(s m
2
e
)(s + m
2
e
)
4s
+
(s m
2
e
)
2
4s
=
(s m
2
e
)
4s
(s + m
2
e
+ s m
2
e
)
=
1
2
(s m
2
e
) (8.141)
As p
2
2
= p
2
2
= 0, we have from function
(p
1
+ p
2
)
2
=(p
1
+ p
2
)
2
(p
1
+ p
2
)
2
=(p
1
+ p
2
)
2
p
2
1
+ 2p
1
p
2
+ p
2
2
=p
1
2
+ 2p
1
p
2
+ p
2
2
p
2
1
+ 2p
1
p
2
=p
1
2
+ 2p
1
p
2
m
2
e
+ 2p
1
p
2
=m
2
+ 2p
1
p
2
(8.142)
8.5. SCATTERING 221
p
1
p
2
= p
1
p
2
1
2
(m
2
m
2
e
) (8.143)
p
1
p
2
=
1
2
(s m
2
) (8.144)
Replacing back in Eq. (8.130) and then in Eq. (8.131) we have
d
d
=
1
64
2
s
_
s m
2
s m
2
e
_
64G
2
F
1
2
(s m
2
e
)
1
2
(s m
2
) (8.145)
d
d
=
G
2
F
4
2
(s m
2
)
2
s
(8.146)
=
G
2
F
(s m
2
)
2
s
(8.147)
Note that s.
Chapter 9
Three body decays
9.1 Muon decay
For a three body decay we have from eq. (7.1)
d =
1
(2)
5
1
2M
[/[
2
4
(P p
1
p
2
p
3
)
d
3
p
1
2E
1
d
3
p
2
2E
2
d
3
p
3
2E
3
=
1
(2)
5
1
2M
d
3
p
1
2E
1
_
[/[
2
4
(P p
1
p
2
p
3
)
d
3
p
2
2E
2
d
3
p
3
2E
3
(9.1)
9.1.1 Amplitude estimation
Since / is dimensionless, the amplitude averaged over spins for decay must be
[/[
2
= CG
2
F
m
4
. (9.2)
We use
C =
1
2
(2 2 1 1) = 2 (9.3)
The rst factor is for the initial average and the factor are for the number of spin states of , e and
the two neutrinos.
223
224 CHAPTER 9. THREE BODY DECAYS
Consider rst the integral
_

4
(P p
1
p
2
)
d
3
p
1
2E
1
d
3
p
2
2E
2
=
_
(E E
1
E
2
)
3
(Pp
1
p
2
)
d
3
p
1
2E
1
d
3
p
2
2E
2
=
_
(E E
1
E
2
)
d
3
p
2
4E
1
E
2
_

3
(Pp
1
p
2
)d
3
p
1
=
_
(E E
1
E
2
)
d
3
p
2
4E
1
E
2
_

3
(p
1
+p
2
P)d
3
p
1
=
_
(E E
1
E
2
)
d
3
p
2
4E
1
E
2
_

3
[p
1
(Pp
2
)]d
3
p
1
(9.4)
using
_
(x x
0
)dx = 1 (9.5)
we have
_

4
(P p
1
p
2
)
d
3
p
1
2E
1
d
3
p
2
2E
2
=
_
(E E
1
E
2
)
d
3
p
2
4E
1
E
2
_

4
(P p
1
p
2
)
d
3
p
1
2E
1
d
3
p
2
2E
2
=
_
(E E
1
E
2
)
p
2
2
d[p
2
[d
4E
1
E
2
(9.6)
Since [p
1
[ = [p
2
[ we have
E = E
1
+ E
2
=(m
2
1
+p
2
1
)
1/2
+ (m
2
2
+p
2
2
)
1/2
=(m
2
1
+p
2
2
)
1/2
+ (m
2
2
+p
2
2
)
1/2
(9.7)
dierentiating this equation with respect to p
2
dE
d[p
2
[
=
1
2
_
2[p
2
[
(m
2
1
+p
2
2
)
1/2
+
2[p
2
[
(m
2
2
+p
2
2
)
1/2
_
=[p
2
[
_
1
E
1
+
1
E
2
_
=[p
2
[
_
E
1
+ E
2
E
1
E
2
_
(9.8)
Therefore
d[p
2
[ =
dE
[p
2
[
_
E
1
E
2
E
1
+ E
2
_
(9.9)
9.1. MUON DECAY 225
replacing back in eq. (9.6)
_

4
(P p
1
p
2
)
d
3
p
1
2E
1
d
3
p
2
2E
2
=
_
dE
[p
2
[
(E E
1
E
2
)
p
2
2
d
4E
1
E
2
_
E
1
E
2
E
1
+ E
2
_
=
_
dE (E E
1
E
2
)
[p
2
[ d
4(E
1
+ E
2
)
=
_
[p
2
[
4E
d (9.10)
For a relativistic particle [p
2
[ E
2
= E/2 and
_

4
(P p
1
p
2
)
d
3
p
1
E
1
d
3
p
2
E
2
= 2 (9.11)
Applying this result to eq. (9.1) we have
d =
1
(2)
5
1
2M
d
3
p
1
2E
1
_
[/[
2
4
(P p
1
p
2
p
3
)
d
3
p
2
2E
2
d
3
p
3
2E
3
=
1
(2)
5
1
2M
d
3
p
1
8E
1
[/[
2
_

4
(P p
1
p
2
p
3
)
d
3
p
2
E
2
d
3
p
3
E
3
=
1
(2)
5
1
2M
d
3
p
1
8E
1
[/[
2
(2)
=
G
2
F
m
4
8(2)
4
m
E
1
p
2
1
d[p
1
[
_
d
G
2
F
m
3
8(2)
4
E
1
E
2
1
dE
1
(4)
G
2
F
m
3
4(2)
3
E
1
dE
1
(9.12)
As the maximum value of E
1
is m
/2

G
2
F
m
3
4(2)
3
_
m/2
0
E
1
dE
1
=
G
2
F
m
3
4(2)
3
m
2
8
, (9.13)
or
=
3
4
G
2
F
m
5
192
3
. (9.14)
9.1.2 Amplitude calculation
The Standard Model Lagrangian includes
L =
2g
2
(
eL
e
L
W
+
+ e
L
eL
W
L
W
+
+
L
L
W
)
=
g
2
(
e
P
L
eW
+
+ e
P
L
e
W
P
L
W
+
P
L
)
=
g
2
2
(
e
(1
5
)eW
+
+ e
(1
5
)
e
W
(1
5
)W
+
(1
5
)
)
(9.15)
where
(
e
P
L
eW
+
= e
P
L
(
e
)
= e
P
L
e
W
= e
P
R
e
W
= e
0
0
P
L
e
W
= e
P
L
e
W
(9.16)
We can build the eective Lagrangian
Applying the Feynman rules to the diagram in g. 9.1 we have the amplitude
/=
ig
2
8
u
3
(1
5
)u
1
_
g
+ q
/M
2
W
q
2
M
2
W
_
u
4
(1
5
)v
2
(9.17)
where u (v) is for an incoming particle and u ( v) is for an ongoing particle (antiparticle).
The Dirac equations for spinors u and v are
(
p m)u = 0 (
p + m)v = 0
u(
p m) = 0 v(
p + m) = 0 . (9.18)
In this way
1
M
2
W
(1
5
)u
1
u
4
(1
5
) =
1
M
2
W
(1 +
5
)
qu
1
u
4
q(1
5
)
=
m
m
e
M
2
W
(1 +
5
)u
1
u
4
(1
5
) (9.19)
the term in q
can be safely neglected. The term q

2
is m
2
is small compared with m

2
W
. Therefore
/=
ig
2
8M
2
W
u
3
(1
5
)u
1
u
4
(1
5
)v
2
=
iG
F
2
u
3
(1
5
)u
1
u
4
(1
5
)v
2
(9.20)
9.1. MUON DECAY 227
Figure 9.1: Tree level diagram for muon decay
/ is a dimensionless scalar. The relevant coupling is
g
2
8M
2
W
=
G
F
2
. (9.21)
The conjugate is, following the same steps that in eq. (9.16)
/
=
ig
2
8M
2
W
[ u
3
(1
5
)u
1
]
[ u
4
(1
5
)v
2
]
=
ig
2
8M
2
W
[ u
1
(1
5
)u
3
] [ v
2
(1
5
)u
4
] . (9.22)
Multiplying / and /
we have
[/[
2
=
g
4
64M
4
W
[ u
3
(1
5
)u
1
u
1
(1
5
)u
3
]
[ u
4
(1
5
)v
2
v
2
(1
5
)u
4
]
=
g
4
64M
4
W
L
(9.23)
where
L
=
_
u
3
(1
5
)
1
u
1
(1
5
)
=
_
u
(1
5
)
v
2
v
(1
5
)
u
4
(9.24)
L
=
_
u
3
u
3
(1
5
)
1
u
1
(1
5
)
=
_
(u
3
u
3
)
(1
5
)
(u
1
u
1
)
(1
5
)
=Tr [(u
3
u
3
)
(1
5
)(u
1
u
1
)
(1
5
)] (9.25)
Using
s
u(p, s) u(p, s) = (
p + m)
s
v(p, s) v(p, s) = (
p m) (9.26)
s
L
=Tr [(
p
3
)
(1
5
)(
p
1
+ m
(1
5
)]
=p
3
Tr
_
(1
5
)(p
+ m
(1
5
)
_
=p
3
Tr
_
(
5
)(p
(1
5
) + m
(1
5
))
_
=p
3
Tr
_
p
(1
5
) p
(1
5
)
+m
(1
5
) m
(1
5
)]
=p
3
Tr
_
p
5
p
+ p
5
+m
5
m
+ m
5
] (9.27)
as the trace of an odd number of matrices is zero, we have
s
L
=p
3
p
1
Tr
_
5
+
2
5
=2p
3
p
1
Tr [
(1
5
)] (9.28)
Similarly
s
M
= 2p
4
p
2
Tr
_
(1
5
)
(9.29)
9.1. MUON DECAY 229
substituting back in eq. (9.23) we have,
[/[
2
=
g
4
64M
4
W
4p
3
p
1
p
4
p
2
Tr [
(1
5
)] Tr
_
(1
5
)
=
g
4
64M
4
W
4p
3
p
1
p
4
p
2
(64
)
=
g
4
64M
4
W
4 64(p
3
p
4
)(p
1
p
2
)
=
4g
4
M
4
W
(p
3
p
4
)(p
1
p
2
)
=4
_
8
g
2
8M
2
W
_
2
(p
3
p
4
)(p
1
p
2
)
=4
_
8
G
F
2
_
2
(p
3
p
4
)(p
1
p
2
)
=128 G
2
F
(p
3
p
4
)(p
1
p
2
)
=256
G
2
F
2
(p
3
p
4
)(p
1
p
2
) . (9.30)
The demonstration of the used TrTr identity can be found in Appendix B. of [7].
The spinaveraged dierential decay width for

e
is
d =
1
(2)
5
1
2E
1
d
3
p
3
2E
3
_
1
2
[/[
2
_
4
(p
1
p
2
p
3
p
4
)
d
3
p
2
2E
2
d
3
p
4
2E
4
=
1
2E
1
1
2
[/[
2
1
(2)
5
d
3
p
2
8E
2
4
(p
1
p
2
p
3
p
4
)
d
3
p
3
E
3
d
3
p
4
E
4
=
1
2
4g
4
M
4
W
1
(2)
5
2E
1
(p
1
p
2
)(p
3
p
4
)
d
3
p
2
2E
2
4
(p
1
p
2
p
3
p
4
)
d
3
p
3
2E
3
d
3
p
4
2E
4
=
2g
4
16(2)
5
M
4
W
E
1
E
4
p
1
p
4
d
3
p
4
I
(9.31)
where the covariant integral I
on the neutrino momentum is

I
=
_
p
3
p
2
4
(p p
2
p
3
)
d
3
p
2
E
2
d
3
p
3
E
3
. (9.32)
The variable p in ec. (9.32) is dened as p = p
1
p
4
= p
2
+ p
3
. Moreover
p
2
=p
2
2
+ p
2
3
+ 2p
2
p
3
=m
2
e
+ m
2
+ 2p
2
p
3
2p
2
p
3
g
=2g
3
p
2
p
=2p
3
p
2
. (9.33)
I
must have the form

I
= g
A(p
2
) + p
B(p
2
) . (9.34)
Dening the itegral I as follows
I =
_

4
(p p
2
p
3
)
d
3
p
2
E
2
d
3
p
3
E
3
, (9.35)
Since
m
2
0 =E
2
p
2
E
2
=p
2
(9.36)
and in addition the integral I is covariant, we choose to evaluate it in the rest frame of the two
neutrinos [p
2
[ = [p
3
[, which implies E
2
= E
3
.
I =
_
(E E
2
E
3
)
3
(p p
2
p
3
)
d
3
p
2
E
2
d
3
p
3
E
3
=
_
(E E
2
E
3
)
d
3
p
2
E
2
E
3
_

3
(p p
2
p
3
)d
3
p
3
. .
=1
=
_
(E 2E
2
)
E
2
2
p
2
2
d[p
2
[d
=
_
(E 2E
2
)
E
2
2
E
2
2
dE
2
(4)
=4
_

_
2
_
E
2
E
2
__
dE
2
=4
1
2
_

_
E
2
E
2
_
dE
2
=2 (9.37)
9.1. MUON DECAY 231
then multiplying (6.13) by g
and p
successively gives, using eq. (9.33)

g
= 4A + p
2
B =
_
p
3
p
2
4
(p p
2
p
3
)
d
3
p
2
E
2
d
3
p
3
E
3
=
p
2
2
I = p
2
In order to compute p
, we make use of the fact that it is a Lorentz invariant quantity, so

that we may evaluate it in any reference frame. In particular, we can evaluate it in the rest frame of
the neutrinos involved in this process. This means that p = p
2
+ p
3
= (p
0
, 0) and E
2
= E
3
p
= p
2
A + p
4
B (9.38)
= p
_
d
3
p
2
E
2
d
3
p
3
E
3
p
3
p
2
4
(p p
2
p
3
)
=
_
d
3
p
2
E
2
d
3
p
3
E
3
E
3
p
0
E
2
p
0
4
(p p
2
p
3
)
= (p
0
)
2
_
d
3
p
2
d
3
p
3
4
(p p
2
p
3
) (9.39)
= (p
0
)
2
_
d
3
p
2
(p
0
2E
2
)
= (p)
2
_
dE
2
E
2
2
d
1
2
(
p
0
2
E
2
) = 4
p
2
2
_
p
2
2
_
2
=
p
4
2
(9.40)
where the usual tricks have been used to simplify the integrals, using the delta function inside.
Therefore
A =
p
2
4
( B) (9.41)
p
4
4
( B) + p
4
B =
p
4
2
4

B
4
+ B =
2
3B
4
=
4
B =
3
(9.42)
A =
p
2
4
(

3
)
=
p
2
4
(
2
3
)
=
p
2
6
(9.43)
I
=

6
_
g
p
2
+ 2p
_
. (9.44)
Substituting back in eq. (9.31) we have
d =
2g
4
16 6(2)
5
M
4
W
E
1
E
4
p
1
p
4
(g
p
2
+ 2p
)d
3
p
4
d =
2g
4
16 12(2)
4
M
4
W
E
1
E
4
[(p
1
p
4
)p
2
+ 2(p p
1
)(p p
4
)]d
3
p
4
d =
2g
4
192(2)
4
M
4
W
E
1
E
4
[(p
1
p
4
)p
2
+ 2(p p
1
)(p p
4
)]d
3
p
4
(9.45)
For further evaluation we will use the rest frame of the decaying muon. In this frame the four
momentum are
p
1
=(m
, 0)
p
4
=(E
4
, p
4
)
p =p
1
p
4
= (m
E
4
, p
4
)
p
2
=E
2
p
2
= m
2
2m
E
4
+ (E
2
4
p
2
4
) = m
2
+ m
2
e
2m
E
4
(9.46)
Moreover
p
1
p
4
=m
E
4
p p
1
=m
2
E
4
p p
4
=m
E
4
E
2
4
+p
2
4
= m
E
4
m
2
e
p
2
4
= m
2
e
= E
2
4
p
2
4
p
2
4
= E
2
4
m
2
e
[p
4
[ =(E
2
4
m
2
e
)
1/2
d[p
4
[
dE
4
=
1
2
2E
4
(E
2
4
m
2
e
)
1/2
=
E
4
[p
4
[
d[p
4
[ =
E
4
[p
4
[
dE
4
d
3
p
4
=p
2
4
d[p
4
[ d = [p
4
[E
4
dE
4
d (9.47)
9.1. MUON DECAY 233
Substituting back in eq. (9.45) we have
d =
2g
4
192(2)
4
M
4
W
m
[p
4
[ dE
4
d[(m
2
+ m
2
e
2m
E
4
)m
E
4
+ 2(m
2
E
4
)(m
E
4
m
2
e
)] (9.48)
Neglecting electron mass we have [p
4
[ = E
4
, and
d =
2g
4
(4)
192(2)
4
M
4
W
m
E
4
dE
4
[(m
2
2m
E
4
)m
E
4
+ 2(m
2
E
4
)m
E
4
]
=
2 2g
4
192(2)
3
M
4
W
m
E
2
4
[m
2
2m
E
4
+ 2m
2
2m
E
4
] dE
4
=
4g
4
192(2)
3
M
4
W
E
2
4
_
3m
2
4m
E
4
dE
4
=
4g
4
192(2)
3
M
4
W
m
2
E
2
4
_
3 4
E
4
m
_
dE
4
=
4g
4
192(2)
3
M
4
W
m
4
4
_
2E
4
m
_
2
_
3 2
_
2E
4
m
__
m
2
d
_
2E
4
m
_
=
4g
4
192(2)
3
M
4
W
m
5
8
_
2E
4
m
_
2
_
3 2
_
2E
4
m
__
d
_
2E
4
m
_
(9.49)
E
4
varies from 0 to E
max
4
can be obtained from (m
e
= 0)
p
1
p
4
= p
2
+ p
3
. (9.50)
The square of he factor on the left is
(p
1
p
4
)
2
=p
2
1
+ p
2
4
2p
1
p
4
=m
2
+ m
2
e
2p
1
p
4
. (9.51)
We have then using eqs. (9.47)(9.50)
2p
1
p
4
=m
2
+ m
2
e
(p
1
+ p
4
)
2
2m
E
4
=m
2
+ m
2
e
(p
2
+ p
3
)
2
m
2
(p
2
+ p
3
)
2
. (9.52)
(p
2
+ p
3
)
2
is the invariant mass squared of the
+
e
system, which ranges from 0 to m
2
. For
(p
2
+ p
3
)
2
= m
we have E
min
4
= 0, while for (p
2
+ p
3
)
2
= 0 we have E
max
4
= m
/2. The missing

integration on d is in the variable x such that
x =
2E
m
, x
min
= 0 , x
max
=
2E
max
m
= 1 . (9.53)
Therefore
=
4g
4
192(2)
3
M
4
W
m
5
8
_
1
0
x
2
[3 2x] dx
=
4g
4
192(2)
3
M
4
W
m
5
8
1
2
=
g
4
192
3
8M
4
W
m
5
4
=
g
4
64M
4
W
2
192
3
m
5
=
G
2
F
2
2
192
3
m
5
=
G
2
F
192
3
m
5
(9.54)
From Eq. (9.49), we also have
d =
4g
4
192(2)
3
M
4
W
E
2
4
_
3m
2
4m
E
4
dE
4
=
g
4
32M
4
W
4
6(2)
3
E
2
4
_
3m
2
4m
E
4
dE
4
=
G
2
F
2
3 8
3
E
2
4
_
3m
2
4m
E
4
dE
4
=
G
2
F
12
3
E
2
4
3m
2
_
1
4
3
E
4
m
_
dE
4
(9.55)
d
dE
4
=
G
2
F
12
3
m
2
E
2
4
_
3
4E
4
m
_
(9.56)
9.2. THREE BODY DECAYS IN RADIATIVE SEESAW 235
Without neglect the electron mass we have
d =
2g
4
192(2)
4
M
4
W
m
[p
4
[ dE
4
d[(m
2
+ m
2
e
2m
E
4
)m
E
4
+ 2(m
2
E
4
)(m
E
4
m
2
e
)]
=
4g
4
192(2)
3
M
4
W
m
dE
4
(E
2
4
m
2
e
)
1/2
[m
3
E
4
+ m
2
e
m
E
4
2(m
E
4
)
2
+ 2m
3
E
4
2(m
E
4
)
2
2m
2
m
2
e
+ 2m
m
2
e
E
4
]
d =
4g
4
192(2)
3
M
4
W
m
dE
4
(E
2
4
m
2
e
)
1/2
[3m
3
E
4
+ 3m
2
e
m
E
4
4(m
E
4
)
2
2m
2
m
2
e
] (9.57)
The decay width, in terms of x = m
e
/m
is
=
4g
4
192(2)
3
M
4
W
m
_
m(1+x
2
)/2
me
(E
2
4
m
2
e
)
1/2
[(3m
2
+ 3m
2
e
4m
E
4
)m
E
4
2m
2
m
2
e
] dE
4
=
4g
4
192(2)
3
M
4
W
m
m
6
16
I (x) , I(x) = 1 8x
2
24x
4
ln(x) + 8x
6
x
8
=
G
2
F
m
5
192
3
I (x)
=
_
G
F
2
_
2
m
5
96
3
I (x) , (9.58)
9.2 three body decays in radiative seesaw
We have the Lagrangian [14]
L =
ab
h
j
N
j
P
L
L
a
b
+ h.c.
=h
j
N
j
P
L
L
1
2
h
j
N
j
P
L
L
2
1
+ h.c.
=h
j
N
j
P
L
0
h
j
N
j
P
L
l
+
+ h.c (9.59)
where
(N
j
P
L
l
+
)
= l
P
L
0
N
j
=

l
P
R
0
N
j
(9.60)
Therefore
L =h
j
N
j
P
L
0
h
j
N
j
P
L
l
+
+ h
P
R
N
j
0
h
j
l
P
R
N
j
=
1
2
_
h
j
N
j
(1
5
)
0
h
j
N
j
(1
5
)l
+
+ h
(1 +
5
)N
j
0
h
j
l
(1 +
5
)N
j
(9.61)
Figure 9.2: Tree level diagram for N
j
decay
Applying Feynman rules to the diagram in g.2 N
j
(p
1
) l
(p
3
)h
+
, h
+
l
+
(p
2
) + N
i
(p
4
).
we have the amplitude
/=ih
j
u
3
(1
5
)u
1
_
1
q
2
M
2
_
h
i
u
4
(1
5
)v
2
ih
j
u
3
(1
5
)u
1
_
1
q
2
M
2
_
h
i
u
4
(1
5
)v
2
iH
ij
M
2
u
3
(1
5
)u
1
u
4
(1
5
)v
2
(9.62)
where
H
ij
= h
j
h
i
+ h
i
h
j
(9.63)
/
=
iH
ij
M
2
[ u
3
(1
5
)u
1
]
[ u
4
(1
5
)v
2
]
=
iH
ij
M
2
[ u
1
(1 +
5
)u
3
][ v
2
(1 +
5
)u
4
] . (9.64)
Multiplying / and /
we have
[/[
2
=
H
2
ij
M
4
[ u
3
(1
5
)
1
u
1
(1 +
5
)
3
][ u
4
(1
5
)
2
v
2
(1 +
5
)
4
]
=
H
2
ij
M
4
[u
3
u
3
(1
5
)
1
u
1
(1 +
5
)
][u
4
u
4
(1
5
)
2
v
2
(1 +
5
)
]
=
H
2
ij
M
4
[(u
3
u
3
)
(1
5
)
(u
1
u
1
)
(1 +
5
)
][(u
4
u
4
)
(1
5
)
(v
2
v
2
)
(1 +
5
)
]
=
H
2
ij
M
4
Tr[(u
3
u
3
)(1
5
)(u
1
u
1
)(1 +
5
)] Tr[(u
4
u
4
)(1
5
)(v
2
v
2
)(1 +
5
)] (9.65)
Using eq. (9.26), and neglecting charged fermion masses
[/[
2
=
H
2
ij
M
4
Tr[
p
3
(1
5
)(
p
1
+ M
j
)(1 +
5
)] Tr[(
p
4
+ M
i
)(1
5
)
p
2
(1 +
5
)]
=
H
2
ij
M
4
LM (9.66)
L =Tr[(
p
3
p
3
5
)(
p
1
+
p
1
5
+ M
j
+ M
j
5
] (9.67)
L =Tr[
p
3
p
1
+
p
3
p
1
5
+ M
j
p
3
+ M
j
p
3
p
3
p
1
p
3
p
1
p
3
5
M
j
+ M
j
5
]
=2 Tr[
p
3
p
1
]
=2p
3
p
1
Tr[
]
=8p
3
p
1
g
=8(p
3
p
1
) (9.68)
Similarly
M = 8(p
4
p
2
) (9.69)
Therefore
[/[
2
=
H
2
ij
M
4
64(p
3
p
4
)(p
1
p
2
)
[/[
2
=
H
2
ij
4M
4
4 64(p
3
p
4
)(p
1
p
2
)
(9.70)
In this way, comparing with eq. (9.30), the results for the moun decay can be directly used after the
replacements
g
4
64M
4
W
H
2
ij
4M
4
g
4
M
4
W
16H
2
ij
M
4
M
j
x =
m
e
m
M
i
M
j
. (9.71)
The decay width is according eq. (9.58)
(N
j
l
N
i
) =
16H
2
ij
M
4
4
192(2)
3
M
j
M
6
j
16
I (x)
=
(h
j
h
i
+ h
i
h
j
)
2
2M
4
M
5
j
192
3
I (x) (9.72)
where
I(x) = 1 8x
2
24x
4
ln(x) + 8x
6
x
8
, x =
M
i
M
j
. (9.73)
Similarly the decay through
0
is
(N
j

N
i
) =
(h
j
h
i
+ h
i
h
j
)
2
2M
4
0
M
5
j
192
3
I (x) (9.74)
In this way, for example for N
2
(N
2
l
l
+
N
1
) =
h
2
2
h
2
1
+ h
2
1
h
2
2
+ 2h
2
h
1
h
2
h
1
2M
4
M
5
2
192
3
I (x)
=
h
2
2
h
2
1
+h
2
1
h
2
2
+ 2h
2
h
1
h
2
h
1
2M
4
M
5
2
192
3
I (x) (9.75)
(N
2
l
l
+
N
1
) =
h
2
2
h
2
1
+h
2
1
h
2
2
+ 2(h
2
h
1
)
2
2M
4
M
5
2
192
3
I (x)
=
h
2
1
h
2
2
+ (h
1
h
2
)
2
M
4
M
5
2
192
3
I (x) (9.76)
In general
(N
j
l
l
+
N
i
) =
h
2
i
h
2
j
+ (h
i
h
j
)
2
M
4
M
5
j
192
3
I
_
M
i
M
j
_
(N
j

N
i
) =
h
2
i
h
2
j
+ (h
i
h
j
)
2
M
4
0
M
5
j
192
3
I
_
M
i
M
j
_
(9.77)
For x i and j
Br(N
j
l
l
+
N
i
)
Br(N
j

N
i
)
=
M
4
0
M
4
(9.78)
while for
Br(N
3

N
2
)
Br(N
3
l
l
+
N
1
)

h
2
2
h
2
3
+ (h
2
h
3
)
2
h
2
1
h
2
3
+ (h
1
h
3
)
2
M
4
M
4
0
I(M
2
/M
3
)
Br(N
3
l
l
+
N
2
)
Br(N
3
l
l
+
N
1
)

h
2
2
h
2
3
+ (h
2
h
3
)
2
h
2
1
h
2
3
+ (h
1
h
3
)
2
I(M
2
/M
3
) (9.79)
For N
2
the total decay width is
tot
(N
2
) =
_
h
2
1
h
2
2
+ (h
1
h
2
)
2
M
5
2
192
3
I
_
M
1
M
2
_
_
1
M
4
+
1
M
4
0
_
(9.80)
And the individual branchings through
given by eq. (9.72).

For N
3
we have several possibilities for signals with charged leptons. The cleanest one is when
N
3
decay only through
through an intermediate N
2
.
The branching of N
3
to two charged leptons plus missing energy is either
Br(N
3
l
N
1
) (9.81)
where the N
3
is reconstructed, or
Br(N
3

..
0
l
N
1
) = Br(N
3

N
2
) Br(N
2
l
N
1
) (9.82)
that seem to be very dicult to reconstruct. This also seem to be an irreducible background for
Br(N
2
l
N
1
) (9.83)
To get rid of processes like the one in eq. (9.82) must be Br(N
3

N
2
) is suppressed. This
happens if
I(M
2
/M3) 1. In this case the mutilepton signal for N
3
is also suppressed. Clearly this
happens for M
2
M
3
as I(x) is a sharpest function which controls the kinematical suppression.
We show below for an specic point that even for M
3
M
2
20 GeV, we can have the Branching
in eq. (9.81) suciently large.
M
0
In appendix 9.A, it is shown a full set of yukawas consistent with neutrino physics. For this
solution
Br(
+
N
3
)
Br(
+
N
1
)
0.61
Br(
+
N
2
)
Br(
+
N
1
)
0.37
Br(
+
N
1
) 0.51 Br(
+
N
2
) 0.19 Br(
+
N
3
) 0.30 (9.84)
Below we estimate the branchings to N
3
l
l
+
N
1
or N
3

N
2

l
+
N
1
. For this we
need the Branchings for N
2
l
l
+
N
1
compared with Branching to N
2

N
1
. In general this is
From this, the visible decays are using eq. (9.78)
Br(N
2
l
l
+
N
1
)
Br(N
2

N
1
)
0.758
Br(N
2
l
l
+
N
1
) = 0.431 (9.85)
On the other hand the chanels for N
3
are N
3
l
l
+
N
1
, N
3

N
1
, N
3
l
l
+
N
2
, and N
3

N
2
. From eqs. (9.78) (9.79)
Br(N
3

N
2
)
Br(N
3
l
l
+
N
1
)
0.0812
Br(N
3
l
l
+
N
2
)
Br(N
3
l
l
+
N
1
)
0.0615
Br(N
3

N
1
)
Br(N
3
l
l
+
N
1
)
1.320 (9.86)
Br(N
3
l
l
+
N
1
)
1
1 + 0.0812 + 0.0615 + 1.320
= 0.406
Br(N
3

N
1
) 0.536
Br(N
3

N
2
) 0.030
Br(N
3
l
l
+
N
2
) 0.025 (9.87)
The expected background for N
2,3
l
l
+
N
1
is
Br(N
3

N
2
)
Br(N
2
l
l
+
N
1
) 0.030 0.431 = 0.013 (9.88)
We have that
tot
(N
2
) =
_
h
2
1
h
2
2
+ (h
1
h
2
)
2
M
5
2
192
3
I
_
M
1
M
2
_
_
1
M
4
+
1
M
4
0
_
vis
(N
2
N
1
)
(N
2
l
l
+
N
1
)
=
h
2
1
h
2
2
+ (h
1
h
2
)
2
M
4
M
5
2
192
3
I
_
M
1
M
2
_
vis
(N
3
N
1
)
(N
3
l
l
+
N
1
)
=
h
2
1
h
2
3
+ (h
1
h
3
)
2
M
4
M
5
3
192
3
I
_
M
1
M
3
_
invis
(N
3
N
2
)
(N
3

N
2
)
=
h
2
2
h
2
3
+ (h
2
h
3
)
2
M
4
0
M
5
3
192
3
I
_
M
2
M
3
_
.
From above equations we can obtain the following observable:
Br
invis
(N
3
N
2
) Br
vis
(N
2
N
1
)
Br
vis
(N
3
N
1
)
=
h
2
2
h
2
3
+(h
2
h
3
)
2
M
4
0
M
5
3
192
3
I
_
M
2
M
3
_
h
2
1
h
2
2
+(h
1
h
2
)
2
M
4
M
5
2
192
3
I
_
M
1
M
2
_
h
2
1
h
2
3
+(h
1
h
3
)
2
M
4
M
5
3
192
3
I
_
M
1
M
3
_
tot
(N
2
)
=
h
2
2
h
2
3
+ (h
2
h
3
)
2
h
2
1
h
2
3
+ (h
1
h
3
)
2
I
_
M
2
M
3
_
1
M
4
0
_
1
M
4
0
+
1
M
4
_
=
h
2
2
h
2
3
+ (h
2
h
3
)
2
h
2
1
h
2
3
+ (h
1
h
3
)
2
I
_
M
2
M
3
_
1
_
1 +
M
4
0
M
4
_
9.A Sample point
write(32,*) (h(i,1),i=1,3),(h(i,2),i=1,3),(h(i,3),i=1,3)
-0.00188878597 0.000780236776 0.000248251388
-0.000352494763 -0.000180683976 -0.00122443053
0.000392272581 0.00120920029 -0.0012245638
So that
h
2
1
4.238 10
6
h
2
2
1.656 10
6
h
2
3
3.116 10
6
h
1
h
2
2.208 10
7
h
1
h
3
1.015 10
7
h
2
h
3
1.143 10
6
(9.89)
h
2
1
h
2
2
+ (h
1
h
2
)
2
7.067 10
12
h
2
1
h
2
3
+ (h
1
h
3
)
2
1.321 10
11
h
2
2
h
2
3
+ (h
2
h
3
)
2
6.465 10
12
(9.90)
The spectrum consistent with neutrino data is
M
1
6.16918656 KeV M
2
22.8695451 GeV M
3
43.126911 GeV
M
0 139.1382 GeV M
149.1382 GeV (9.91)

I(M
1
/M
3
) 1 I(M
2
/M
3
) 0.126 (9.92)
9.B Preliminary discussion
One interesting possibility in view of the large invisible direct decay, like N
3

N
1
, is to get the
observables from the missing plus one energetic lepton (coming from
+
) signal. May be decays like
+
l
+
N
3
l
+

E
T
+
l
+
N
2
l
+

E
T
(9.93)
Once
0
R,I
, or
are produced the full list of signals is: For
production. The decay to N

j
is
(
+
l
+
N
j
) =
3h
2
j
16M
1/2
_
M
2
, M
2
j
, m
2
_
_
1
M
2
j
+ m
2
M
2
_
(9.94)
9.B. PRELIMINARY DISCUSSION 243
(
+
l
+
N
j
) =
3h
2
j
16M
1/2
_
M
2
, M
2
j
, m
2
_
_
1
M
2
j
+ m
2
M
2
_
(9.95)
with
1/2
_
M
2
, M
2
j
, m
2
_
=
_
_
M
2
+ M
2
j
m
2
_
2
4M
2
M
2
j
_
1/2
(9.96)
Neglecting m
with respect to N
2,3
, we have for j = 2, 3
1/2
_
M
2
, M
2
j
, m
2
_
M
2
_
_
1 +
M
2
j
M
2
_
2
4M
2
j
M
2
_
1/2
M
2
_
1 + 2
M
2
j
M
2
4M
2
j
M
2
_
1/2
M
2
_
1 2
M
2
j
M
2
_
1/2
M
2
_
1
M
2
j
M
2
_
(9.97)
Therefore
(
+
l
+
N
j
)
3h
2
j
M
16

_
_
_
1
M
2
j
M
2
_
2
j = 2, 3
1 j = 1
3h
2
j
M
16

_
_
_
1 2
M
2
j
M
2
_
j = 2, 3
1 j = 1
(9.98)
In this way
tor
(
+
) =
j
(
+
l
+
N
j
)
3M
16
_
h
2
1
+h
2
2
_
1 2
M
2
2
M
2
_
+h
2
3
_
1 2
M
2
3
M
2
__
(9.99)
Br(
+
N
j
)
Br(
+
N
i
)
=
(
+
l
+
N
j
)
(
+
l
+
N
i
)
h
2
j
h
2
i
1 2M
2
j
/M
2
1 2M
2
i
/M
2
h
2
j
h
2
i
(1 2M
2
j
/M
2
)(1 2M
2
i
/M
2
)
1
h
2
j
h
2
i
(1 2M
2
j
/M
2
)(1 + 2M
2
i
/M
2
h
2
j
h
2
i
_
1 2
_
M
2
j
M
2
i
M
2
__
(9.100)
For three branchings we should have
a + b + c = 1
1 +
b
a
+
c
a
=
1
a
a =
1
1 + b/a + c/a
(9.101)
In this way
Br(
+
N
1
) =
1
1 +
Br(
+
N
3
)
Br(
+
N
1
)
+
Br(
+
N
3
)
Br(
+
N
1
)
(9.102)
From eq.
Br (N
3
N
1
)
Br(N
3

..
N
2
)
= (9.103)
Chapter 10
Renomalization at 1-loop
In this chapter we study the renomalization of the mass terms and the gauge coupling in the context
of a Yang-Mills theory
10.1 Self-energy
El Lagrangiano de interaccion entre gluones y fermiones y entre fotones y fermiones es
L
int
=g
s
q
q
a
2
_
q G
a
+ e
Q
f
fA
=g
s
q
q
(T
a
) q G
a
+ e
Q
f
fA
=g
s
q
q
(T
a
)
G
a
+ e
f
i
(Q
f
ij
)f
j
A
, (10.1)
Generalizando para un acoplamiento entre un boson gauge y un fermi on, tenemos
L
int
= g
a
(T
c
)
ab
b
G
c
. (10.2)
245
246 CHAPTER 10. RENOMALIZATION AT 1-LOOP
S
(2)
= S
(2)
=
(i)
2
2!
_ _
d
4
x
1
d
4
x
2
T1
I
(x
1
)1
I
(x
2
)
=
(ig)
2
2!
_ _
d
4
x
1
d
4
x
2
T:
_
b
G
c
(T
c
)
ab
x
1
_
e
G
f
(T
f
)
de
x
2
:
=
(ig)
2
2!
_ _
d
4
x
1
d
4
x
2
:
_
b
G
c
(T
c
)
ab
x
1
_
e
G
f
(T
f
)
de
x
2
: +
+
(ig)
2
2!
_ _
d
4
x
1
d
4
x
2
:
_
b
G
c
(T
c
)
ab
x
1
_
e
G
f
(T
f
)
de
x
2
: + (10.3)
S
(2)
= S
(2)
=
(i)
2
2!
_ _
d
4
x
1
d
4
x
2
T1
I
(x
1
)1
I
(x
2
)
=
(ig)
2
2!
_ _
d
4
x
1
d
4
x
2
T:
_
G
c
b
(T
c
)
ab
x
1
_
G
f
e
(T
f
)
de
x
2
:
=
(ig)
2
2!
_ _
d
4
x
1
d
4
x
2
:
_
G
c
b
(T
c
)
ab
x
1
_
G
f
e
(T
f
)
de
x
2
: +
+
(ig)
2
2!
_ _
d
4
x
1
d
4
x
2
:
_
G
c
b
(T
c
)
ab
x
1
_
G
f
e
(T
f
)
de
x
2
: + (10.4)
The self-energy diagram give to arise
i
ab
(p) = g
2
4d
_
d
d
k
(2)
d
p k m
k
2
(T
c
)
ad
(T
c
)
db
(10.5)
i
ab
(p) = (T
c
)
ad
(T
c
)
db
(p) (10.6)
(p) = ig
2
4d
_
d
d
k
(2)
d
p k + m)
[(p k)
2
m
2
]k
2
(10.7)
Changing variables
(p) = ig
2
4d
_
1
0
dz
_
d
d
k
(2)
d
p(1 z) k
+ m)
[k
2
m
2
z + p
2
z(1 z)]
2
(10.8)
Applyng the formulas
_
d
d
p
1
p
2
+ 2pq M
2
=i
d/2
( d/2)
()
1
[q
2
M
2
]
d/2
_
d
d
p
p
p
2
+ 2pq M
2
=i
d/2
( d/2)
()
q
[q
2
M
2
]
d/2
(10.9)
10.1. SELF-ENERGY 247
We get in the rst place that the integration in k is zero. Moreover, dening
M
2
= m
2
z p
2
z(1 z) (10.10)
and taking into account that q = 0, we have
_
d
d
k
1
[k
2
m
2
z + p
2
z(1 z)]
2
= i
d/2
(2 d/2)
(2)[m
2
z + p
2
z(1 z)]
2d/2
(10.11)
Since
(n) = (n 1)! (10.12)
so that (2) = 2, then
_
d
d
k
1
[k
2
m
2
z + p
2
z(1 z)]
2
= i
d/2
(2 d/2)
[m
2
z + p
2
z(1 z)]
2d/2
(10.13)
Let
= 4 d

2
= 2 d/2 (10.14)
In this way we are interested in the limit 0:
(p) =ig
2
4d
(2 d/2)
_
1
0
dz
i
d/2
(2)
d
p(1 z) + m)
[m
2
z + p
2
z(1 z)]
2d/2
=g
2
4d
(2 d/2)(4)
d/2
_
1
0
dz

p(1 z) + m)
[m
2
z + p
2
z(1 z)]
2d/2
=g
2
(
2
)
2d/2
(2 d/2)
(4)
2d/2
(4)
2
_
1
0
dz

p(1 z) + m)
[m
2
z + p
2
z(1 z)]
2d/2
=g
2
(
2
)
/2
(/2)
(4)
/2
(4)
2
_
1
0
dz

p(1 z) + m)
[m
2
z + p
2
z(1 z)]
/2
=
g
2
(/2)
16
2
_
1
0
dz
p(1 z) + m)
_
m
2
z + p
2
z(1 z)
4
2
_
/2
(10.15)
From the Dirac algebra in d-dimensions
=d
=(2 d)
= (2 + )
(10.16)
Therefore
(p) =
g
2
(/2)
16
2
_
1
0
dz (
p(2 + )(1 z) + dm)

_
m
2
z + p
2
z(1 z)
4
2
_
/2
=
g
2
(/2)
16
2
_
1
0
dz [2
p(1 z) + (4 )m +
p(1 z)]
_
m
2
z + p
2
z(1 z)
4
2
_
/2
=
g
2
(/2)
16
2
_
1
0
dz 2
p(1 z) + 4m + [
p(1 z) m]
_
m
2
z + p
2
z(1 z)
4
2
_
/2
=
g
2
(/2)
16
2
_
1
0
dz 2
p(1 z) 4m[
p(1 z) m]
_
m
2
z + p
2
z(1 z)
4
2
_
/2
(10.17)
Since
(/2) =
2
+O()
a
=1 + ln a + (10.18)
(p)
g
2
16
2
_
2

__
1
0
dz 2
p(1 z) 4m[
p(1 z) m]
_
1

2
ln
_
m
2
z + p
2
z(1 z)
4
2
__
=
g
2
16
2
_
1
0
dz
__
2

_
(2
p(1 z) 4m)
_
2

_
[
p(1 z) m]
_
_
1

2
ln
_
m
2
z + p
2
z(1 z)
4
2
__
g
2
16
2
_
1
0
dz
_
2
[2
p(1 z) 4m] [2
p(1 z) 4m] 2[
p(1 z) m]
_
_
1

2
ln
_
m
2
z + p
2
z(1 z)
4
2
__
g
2
16
2
_
1
0
dz
_
2
[2
p(1 z) 4m] [2
p(1 z) 4m] 2[
p(1 z) m]
[2
p(1 z) 4m] ln
_
m
2
z + p
2
z(1 z)
4
2
__
(10.19)
Since
_
1
0
dz =1
_
1
0
(1 z)dz = z
z
2
2
1
0
=
1
2
(10.20)
10.2. OTHER 249
we have
(p)
g
2
16
2
_
2
p 4m] [
p 4m] [
p 2m]
2
_
t
0
dz[
p(1 z) 2m] ln
_
m
2
z + p
2
z(1 z)
4
2
__
=
g
2
8
2
[
p 4m]
g
2
16
2
_
p + 4m
p + 2m2
_
t
0
dz[
p(1 z) 2m] ln
_
m
2
z + p
2
z(1 z)
4
2
__
=
g
2
8
2
[
p 4m] +
nite
(p) (10.21)
where
nite
(p) =
g
2
16
2
_
p(1 + ) 2m(1 + 2) + 2
_
t
0
dz[
p(1 z) 2m] ln
_
m
2
z + p
2
z(1 z)
4
2
__
(10.22)
Finally
ab
(p) =
g
2
8
2
[
p 4m] (T
c
T
c
)
ab
+
ab
nite
(p) (10.23)
For SU(N) we have
(T
c
T
c
)
ab
=
N
2
1
2N

ab
C
2
(F)
ab
(10.24)
where
C
2
(F) =
N
2
1
2N
(10.25)
is the Casimir for SU(N). In this way
ab
(p) =
g
2
8
2
C
2
(F) [
p 4m] +
ab
nite
(p) (10.26)
10.2 Other
The other loop calculation and the renormalization procedure are explained in scan of the following
notes: http://gfif.udea.edu.co/cf/Renormalization.pdf (PDF 13M)
Bibliography
[1] Diego Restrepo, Haca la teora cu antica de campos, curso web, http://gfif.udea.edu.co
[2] M. Maggiore, A Modern introduction to quantum eld theory, Oxford University Press, 2005.
(Oxford Series in Physics, 12. ISBN 0 19 852073 5)
[3] F. Mandl and G. Shaw, Chichester, Uk: Wiley ( 1984) 354 P. ( A Wiley-interscience Publica-
tion)
[4] A. Lahiri and P. B. Pal, A rst book of quantum eld theory, Harrow, UK: Alpha Sci. Int.
(2005) 380 p
[5] Wikipedia article: http://en.wikipedia.org/wiki/Particle_decay
[6] M. Klasen, Int. J. Mod. Phys. C 14 (2003) 1273 [arXiv:hep-ph/0210426].
[7] Electroweak Interactions, Peter Renton
[8] Michael E. Peskin and Daniel V. Schroeder. An introduction to Quantum Field Theory, Addison-
Wesley Publishing Company(1995), p. 101
[9] ibdg 1, p. 105
[10] Francis Halzen and Alan D. Martin. Quarks & Leptons: An introductory Course in Modern
Particle Physics, John Wiley & Sons(1984), p. 89
[11] Chris Quigg. Gauge theory of the Strong, Weak and Electromagnetic Interactions, Westview
Press(1997), p. 110
[12] ibdg 4, p. 130.
[13] ibdg 5, p. 325.
251
252 BIBLIOGRAPHY
[14] D. Aristizabal Sierra, J. Kubo, D. Restrepo, D. Suematsu and O. Zapata, Radiative seesaw:
Warm dark matter, collider and lepton avour violating signals, Phys. Rev. D 79 (2009) 013011
[arXiv:0808.3340 [hep-ph]].
[15] Relativistic Quantum Mechanics and Field Theory, Franz Gross, John Wiley & Sons, INC. 1993
[16] D. McMahon, Quantum eld theory demystied: A self-teaching guide, SPIRES entry New
York, USA: McGraw-Hill (2009) 299 p
[17] R. P. Feynman, QED. The Strange Theory Of Light And Matter, Princeton, Usa: Univ. Pr.
( 1985) 158 P. ( Alix G. Mautner Memorial Lectures)
[18] A. Semenov, LanHEP - a package for the automatic generation of Feynman rules in eld
theory. Version 3.0, Comput. Phys. Commun. 180 (2009) 431 [arXiv:0805.0555 [hep-ph]], http:
//feynrules.phys.ucl.ac.be/.
[19] N. D. Christensen and C. Duhr, FeynRules - Feynman rules made easy, Comput. Phys. Com-
mun. 180, 1614 (2009) [arXiv:0806.4194 [hep-ph]], http://feynrules.phys.ucl.ac.be/.

Detailed Lecture Notes in Quantum Field PDF

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Detailed Lecture Notes in Quantum Field PDF

Uploaded by

Copyright:

Available Formats

Detailed lecture notes in Quantum Field

(x) (x) (1.3)

x, la expansi on de Taylor para f(x + x) es

satisfacen las ecuaciones de EulerLagrange, en lugar de asumir

satisface la ecuaci on de continuidad:

(x) = (x) + (x) = (x) (

es constante (un an alisis mas general es hecho en [?])

proviene de asumir la homogeneidad del espacio y el tiempo y es llamado el tensor de

, in order to compensate for the term arising from

(x), is just that

can be obtained from the X

transformation in eq. (1.81):

. In this way, the electromagnetic interaction is

denota la forma matricial del tensor g

La invarianza del producto escalar en ec. (1.93)

denido en la ec. (1.105) satisface la condici on en ec. (??),

corresponde a la ecuacion de momento energa una vez se

= 0 es un invariante bajo transformaciones de Lorentz:

= 0 El operador cuadr atico es, usando la ec. (1.90)

tiene la transformacion gauge

. It is convinient to invert eq. (1.133)

(x) transforms simultaneously as eld and as vector under Lorentz transformation

compatible with Lorentz invariance and the gauge transformation

El Lagrangiano que da lugar a una Acci on invariante de Lorentz para el cuadrivector A

(x), and hence F

constant) are not Lorentz invariant:

(x) is Lorentz covariant but not gauge-invariant (see below).

en (1.144) que sean invariantes hasta una derivada total.

(algo) y la Acci on no es invariante bajo la transformacion gauge. Para los

= 0. However we will see that

already includes the homogeneous Maxwell equations. To see this we note

) intercambiando ndices mudos

Usando las ecuaciones (??), (??), (1.174)

ODINGER EQUATION IN PRESENCE OF THE ELECTROMAGNETIC FIELD 33

, we should write the most

ODINGER EQUATION IN PRESENCE OF THE ELECTROMAGNETIC FIELD 35

ODINGER EQUATION IN PRESENCE OF THE ELECTROMAGNETIC FIELD 37

ODINGER EQUATION IN PRESENCE OF THE ELECTROMAGNETIC FIELD 39

=[1 + i(x) +O(

(x))[1 + i(x) +O(

que median las

= 0 y las ec. (1.222), como:

constante, no es invariante de Lorentz, de modo que en efecto, el Lagrangiano en de

, usando el Lagrangiano en ec. (1.244)

, satisfacen la ecuacion de Klein-Gordon. Cada campo adem as

) =(x) Scalar eld,

(x)(x). However, under a Lorentz transformation we

1.11.2 Corriente conservada y Lagrangiano de Dirac

las condiciones en ec. (1.305) son

, el Lagrangiano de Dirac y la ecuaci on de Dirac, son respectivamente

1 y conmuta con las dem as matrices, tenemos por ejemplo

corresponde a otra representacion de algebra de Dirac

sea diagonal. Como

) debe ser par: 2, 4, . . .. Para un fermion sin

satisfy the algebra

satisfy the Lorentz algebra, and therefore are the generators of

] Tensor antisimetrico (T) 6

s have such properties, we

. Este campo es necesario para compensar los cambios en la energa y

s (la energa de centro de masa

s 4 GeV que corresponden

ya no es invariante bajo transformaciones gauge