Documente Academic
Documente Profesional
Documente Cultură
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
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
3
3
3
5
7
10
14
17
20
23
24
30
33
35
38
39
40
41
43
45
48
48
49
51
52
53
vi
CONTENTS
1.12
1.13
1.14
1.15
1.16
1.17
1.18
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
57
59
62
64
66
69
78
81
84
90
90
92
97
102
102
106
108
109
114
117
120
2 Computational QFT
121
2.1 LanHEP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
2.2 CalcHEP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
2.3 LanHEP/CalcHEP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
3 Second quantization
3.1 Quantization of the nonrelativistic string .
3.1.1 The clasical string . . . . . . . . .
3.1.2 Quantization of the string . . . . .
3.1.3 Generalization to three dimensions
3.2 Quantization of the Klein-Gordon field . .
3.3 Fock space . . . . . . . . . . . . . . . . . .
3.4 Fock space for the harmonic oscillator . . .
3.5 Propagator . . . . . . . . . . . . . . . . .
3.6 Quantization of Fermions . . . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
131
132
132
138
146
147
153
153
156
157
vii
CONTENTS
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
173
173
175
178
180
183
189
190
191
193
193
194
198
7 Feynman Rules
7.1 Interaction picture . . . . . . . . .
7.2 Atomic decay . . . . . . . . . . . .
7.3 Yukawa interaction . . . . . . . . .
7.3.1 Yukawa interaction for Weyl
7.4 Wick Theorem . . . . . . . . . . .
7.4.1 Basics concepts . . . . . . .
7.4.2 Scalar propagator . . . . . .
7.4.3 Wick theorem . . . . . . . .
7.5 Scattering . . . . . . . . . . . . . .
199
199
205
209
214
214
214
217
224
227
. . . . .
. . . . .
. . . . .
fermions
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
8 Neutrinos
235
8.1 Weinberg operator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 235
9 Three body decays
9.1 Muon decay . . . . . . . . . . . . . .
9.1.1 Amplitude estimation . . . . .
9.1.2 Amplitude calculation . . . .
9.2 three body decays in radiative seesaw
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
237
237
237
240
249
viii
CONTENTS
9.A Sample point . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 256
9.B Preliminary discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 256
. . . . .
. . . . .
singlets
. . . . .
. . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
259
259
261
261
261
264
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 field theory, in Chapter 5 we introduce the Smatrix in the Scrodinger
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 7. In Chapter 6 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 7 we obtain the
Feynman rules used in two body decays directly from the first 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 fluently in reading technical texts in this language.
Some parts are still in Spanish.
This work have been partially supported by Dedicacion Exclusiva 2008-2009 project: RR 26663
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
1.1.1
Lagrangian Formulation
Ecuaciones de Euler-Lagrange
Definamos
,
x
En tres dimensiones, la accion de la se puede escribir como:
Z
S[, ] =
d4 xL(, )
=
(1.1)
(1.2)
donde d4 x = dt dx dy dz. Considere primero una variacion solo de los campos, tal que (x = x )
(x) = 0 (x) (x)
(1.3)
f
x +
x
(1.4)
L(0 , 0 ) = L( + , + ())
L
L
=L+
+
()
( )
( )
R
Z
Z
L
L
L
4
4
d x
dx
+
=
( )
( )
R
R
Z
Z
L
L
L
4
dx
S =
+
d = 0.
( )
R
( )
Donde hemos aplicado el Teorema de Gauss
Z
Z
3
Ad x =
A dS
V
(1.5)
(1.6)
(1.7)
( )
R
Como es cualquier posible variacion entre las fronteras de la hipersuperficie, el integrando debe
anularse y resultan las ecuaciones de Euler-Lagrange:
L
L
= 0.
(1.9)
( )
La densidad Lagrangiana
L0 = L + ((x))
(1.10)
donde (x) es cualquier funcion de los campos de la densidad Lagrangiana original, da lugar a la
Accion
Z
Z
Z
0
4
0
4
S =
d xL = d xL +
d4 x
R
ZR
ZR
4
= d x L + d
R
=S ,
(1.11)
para una hipersuperficie suficientemente grande. De modo que dos densidades lagrangianas que
difieran solo en derivadas totales dan lugar a la misma Accion.
Usando el principio de mnima accion en terminos del campo , tenemos que para la densidad
Lagrangiana (??)
"
2 #
2
1 1
,
(1.12)
L=
2
2 v
t
z
las ecuaciones de Euler-Lagrange (1.9)
L
L
L
0
=0
+ 3
(0 )
(3 )
L
+
=0
t (/t)
z (/z)
1
=0
v 2 t t
z z
1 2 2
2 =0 ,
v 2 t2
z
(1.13)
1.1.2
S[, , , ] =
d4 x L(, , , )
R
(1.16)
(
)
R
R
Z
L
L
d4 x
+
+
= 0.
(
)
( )
(1.17)
Usando de nuevo el Teorema de Gauss resultan las ecuaciones de Euler Lagrange para y
L
L
L
L
= 0,
= 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 hipersuperficie, los dos primeros terminos de la ec. (1.17) se anulan
y tendremos que para que S = 0:
Z
d4 x ( J ) = 0,
(1.19)
R
donde,
L
L
+
J =
( )
( )
Entonces J satisface la ecuacion de continuidad:
(1.20)
J = 0
(1.21)
J
+J=0
t
(1.22)
Z
J 0 3
J d3 x = 0,
d x+
t
ZV
ZV
J 0 3
d x + J dS = 0,
V t
S
(1.23)
Escogiendo una superficie suficientemente grande que abarque toda la fuente de densidad = J 0 , de
la corriente J, el segundo integrando es cero y
Z
d
d3 x = 0.
(1.24)
dt V
1.1.3
Para el caso de una simetra externas, por ejemplo la correspondiente a una traslacion espacio
temporal
x x0 =x + a
x =a
(1.26)
tenemos
0 (x0 ) = 0 (x + a)
0 (x)
0 (x) +
a
x
(1.27)
(1.28)
[(x) + (x)]a
x
(x)
(x) + (x) +
a ,
x
= [(x) + (x)] +
(1.29)
(1.30)
(x)
a .
x
(1.31)
(1.32)
(1.33)
(1.34)
En este caso, asumiendo que el campo satisface las ecuaciones de Euler-Lagrange y usando la ec. (1.32)
d4 x L((x), (x), x)
R
Z
ZR
4
d4 x L
d x L( + , + (), x + a )
=
R
Z
ZR
L
L
4
d x L+
+
() + ( L)a
d4 x L
(
)
R
ZR
L
L
+
() + ( L)a
=
d4 x
( )
ZR
L
L
4
dx
() + ( L)a
=
+
( )
( )
R
Z
L
4
d x
=
+ ( L)a
( )
R
Z
L
4
d x
+ La
=
( )
R
Z
L
4
a + La
=
d x
( )
R
Z
L
4
=
d x
a + L (a )
( )
R
Z
L
4
=
d x
( ) + L a
( )
R
S =
d x L( , , x )
(1.35)
Z
=
R
d4 x (T a ) = 0.
(1.36)
(1.37)
(1.38)
donde
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 ecuacion de continuidad:
T =
T0 = 0
(1.40)
Donde la densidad de Energa, o mas de forma mas general: la densidad Hamiltonina corresponde a
T00
L
L
(1.41)
(x)
= (x)
L.
(1.42)
t
Comparando con la expresion correspondiente en la formulacion Lagrangiana de la Mecanica Clasica,
tenemos que si (x) es la variable canonica, la variable canonica conjugada es (x)
H = T00 =
(x) =
L
.
((x)/t)
(1.43)
El teorema de Noether en este caso establece que la invarianza de la Accion bajo traslaciones temporales da lugar a la ecuacion de continuidad (1.38) para = 0
T0 = 0
(1.44)
(1.45)
De igual forma la invarianza bajo traslaciones espaciales de lugar a ecuaciones de continuidad para
cada componente = i (i = 1, 2, 3)
Ti = 0,
(1.46)
10
cuyas densidad de cargas conservadas, Ti0 , que en forma vectorial escribiremos como T0 , dan lugar
a la conservacion del momentum
Z
P=
d3 x T0 .
(1.47)
V
1.2
L
L
( ) + ( )
L
( )
( )
(1.48)
L(, , , ) =
+ V
2m
2
t
t
1
i
=
i i ( 0 0 ) + V .
2m
2
(1.49)
Aplicando las ecuaciones de Euler-Lagrange (1.18) para la funcion de onda obtenemos la ecuacion
de Scrodinger con ~ = 1:
L
L
L
L
L
0 =
= 0
+ i
.
(1.50)
( )
(0 )
(i )
Como
i
L
=
(0 )
2
L
1
=
i
(i )
2m
L
i
= 0 + V
i
L
=
(0 )
2
L
1
=
i
(i )
2m
L
i
=
0 + V .
2
= 0
+ i
i 0 + V
( )
2
2m
2
i
1
i
= 0 +
i i + 0 V .
2
2m
2
(1.51)
(1.52)
11
i =
t
1 2
+ V .
2m
(1.53)
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
= 0 = (ei 1) i
i .
Usando ademas la ec. (1.51) en la definicion de J 0 dada por la ec. (1.20), tenemos
L
L
0
+
J =
(0 )
(0 )
i
i
= (i) + (i )
2
2
= ,
(1.55)
(1.56)
(1.57)
y
L
L
+
J =
(i )
(i )
1
1
=
i (i) + (i )
i
2m
2m
i
=
(i i ) .
2m
i
(1.58)
(1.59)
(1.60)
12
En cuanto a las simetras externas, tenemos de la ec. (1.39) que da lugar a las ecuaciones de
continuidad (1.44)(1.46)
T0 = 0,
Ti = 0
(1.62)
Las cargas conservadas corresponden entonces a T00 y Ti0 . Usando las ecs. (1.51) en la ec. (1.48)
L
L
(i ) + (i )
(0 )
(0 )
i
i
Ti0 = (i ) + (i )
2
2
Ti0 =
Entonces, definiendo
T0 =
i
( )
2
(1.63)
(1.64)
Ademas
i
(( ) )
2
i
= i + ( ) .
2
T0 =
(1.65)
Integrando en el volumen
Z
V
T d x = i
Z
V
i
d x +
2
d3 x
(1.66)
(1.67)
De modo que hb
pi son las cargas conservadas asociadas al valor esperado el operador de momentum
b = i .
p
De otro lado
L
L
0 + 0
L
(0 )
(0 )
i
i
1
i
= 0 + 0
i i + ( 0 0 ) V
2
2
2m
2
1
=
i i V
2m
(1.68)
T00 =
(1.69)
13
H T00 =
Hd x =
( ) +
+ V d3 x
2m V
2m
V
V
Z
Z
1 2
1
( ) +
+ V d3 x
=
2m
2m
V
V
Z
i
1 2
=
hb
pi +
+ V d3 x
2m
2m
V
Z
1
2
+ V d3 x .
=
2m
V
(1.70)
(1.71)
Entonces
H
1 2
+V
2m
Hd x =
b = hHi.
b
d3 x H
Z
=
d3 x
b .
i = H
t
Podemos identificar entonces los operadores de energa y momentum.
b =i,
H
t
= i .
p
(1.72)
(1.73)
(1.74)
(1.75)
Retornando a la ec. (1.67), tenemos que para la solucion de partcula libre de la ecuacion de
Schrodinger
= A eikx ,
(1.76)
14
(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
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 finding 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 fixes 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 field 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.
1.3. LOCAL PHASE INVARIANCE IN THE SCRODINGERS
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 fields remain invariant. As we will see, this feature is deeply connected with the local
conservation of electric charge.
We start again with the Scrodinger Lagrangian as written in eq. (1.49):
i
1
+ V
L(, , , ) =
2m
2
t
t
1
i
=
i i ( 0 0 ) + V .
2m
2
(1.78)
This Lagrangian is not invariant under local phase changes of the wave function:
0 = ei(x)
= ei(x) + ei(x)
=ei(x) (i (x)) + ei(x)
=ei(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 ei(x) :
D D0 0 =( + X0 ) ei(x)
=ei(x) [i (x) + ] + X0 ei(x)
=ei(x) i (x) + + X0 .
(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 X0 = X i (x) .
(1.81)
(1.82)
16
Note that D transforms like the field , and because of this is called the covariant derivative of
. Similarly
(D ) (D )0 =( + X0 ) ei(x)
= i (x) + + X + i (x) ei(x)
= + X ei(x)
= (D ) ei(x) .
(1.83)
1
X ,
iq
(1.84)
(1.85)
The transformation properties of A can be obtained from the X transformation in eq. (1.81):
iqA iqA0 = iqA i (x)
1
A A0 = A (x) .
q
(1.86)
We define local gauge invariance as an arbitrary way of choosing the complex phase factor of a
charged field1 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 ) =
i
1
(Di ) Di [ D0 (D0 ) ] + V (x).
2m
2
(1.87)
where
1
A A0 = A (x) .
q
1
(1.88)
RELATIVISTA
1.4. NOTACION
17
This is just the gauge transformation which left the Electromagnetic fields invariant. In fact, the
new Lagrangian is now invariant under the local phase transformations
i
1
L L0 =
(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 field 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 definen como la transformaciones que dejan invariante al producto
escalar en el espacio de Minkowski definido como
2
a2 = g a a a a = a0 ai ai = a0 a a
(1.90)
(1.91)
1 0
0
0
0 1 0
0
{g } =
0 0 1 0
0 0
0 1
(1.92)
(1.93)
18
El inverso de la metrica es
{g } {g }1 = {g }
tal que
g g =
a = g a
and
(1.94)
(1.95)
a a0 = a
a a0 = a
(1.96)
a0 b0 =a b
g a0 b0 = g a b
g a b = g a b
g a b = g a b ,
(1.97)
da lugar a
g = g
{g } = { }T {g } .
or
En notacion matricial
g = T g .
(1.98)
(1.99)
(1.100)
= .
(1.101)
or
Since
1
(1.102)
= ,
(1.103)
= ,
(1.104)
the inverse of is
1
or
1
RELATIVISTA
1.4. NOTACION
19
a b a0 b0 = a bp
= a bp
= 1 a bp
= a bp
=a b .
Como un ejemplo de Transformacion de Lorentz considere un desplazamiento a lo largo del eje x
0 t+vx
t
t
cosh sinh 0 0
t
1v 2
x+vt
0
x
x
sinh
cosh
0
0
x
0 = 1v2 =
= { } {x } ,
{x } =
(1.105)
y
y y 0
0
1 0 y
z0
z
0
0
0 1
z
z
donde
cosh =
y, por ejemplo:
sinh = v,
and
1
.
1 v2
t + vx
t cosh + x sinh = (t + vx) =
.
1 v2
cosh sinh 0 0
1 0
0
0
cosh sinh 0 0
sinh cosh 0 0 0 1 0
sinh cosh 0 0
T g =
0
0
1 0 0 0 1 0 0
0
1 0
0
0
0 1
0 0
0 1
0
0
0 1
cosh sinh 0
0
cosh sinh 0 0
sinh cosh 0
0 sinh cosh 0 0
=
0
0
1 0 0
0
1 0
0
0
0 1
0
0
0 1
2
2
cosh sinh
cosh sinh cosh sinh 0
0
cosh sinh cosh sinh
sinh2 cosh2
0
0
0
0
1 0
0
0
0 1
=g
(1.106)
(1.107)
(1.108)
20
(1.109)
(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 =(x0 , x1 , x2 , x3 ) = (t, x, y, z) = (t, x)
p =(p0 , p1 , p2 , p3 ) = (E, px , py , pz ) = (E, p)
(1.111)
(1.112)
(1.113)
(1.114)
(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
RELATIVISTA
1.4. NOTACION
21
Para p = 0, es decir cuando la partcula esta en reposo se reduce a la famosa ecuacion E = mc2
(c = 1)
Del electromagnetismo tenemos
J = (J 0 , J) = (, J)
(1.116)
A = (A0 , A) = (, A)
(1.117)
,
,
,
=
,
,
,
x0 x1 x2 x3
x0 x1 x2 x3
, , ,
=
t x y z
=(0 , ) = ( 0 , )
=
=
, , ,
= (0 , )
x
t x y z
=
x
Por consiguiente:
=
(1.118)
(1.119)
(1.120)
Producto escalar:
a b = g a b = a0 b0 a1 b1 a2 b2 a3 b3 = a0 b0 ai bi = a0 b0 a b
(1.121)
Entonces
a0
+a
(1.122)
t
La ecuacion de continuidad J = 0 es un invariante bajo transformaciones de Lorentz: 0 J 0 =
J = 0 El operador cuadratico es, usando la ec. (1.90)
a =
= 0 0 2 =
2
2
2
2
t2 x2 y 2 z 2
(1.123)
(1.124)
b yp
dados en la ec. (1.75). Entonces
con H,
p = i = i( 0 , i ) = i(
, )
t
(1.125)
22
(1.126)
(1.127)
=(0 , i ) + iq(A0 , Ai )
=(0 , i ) + iq(A0 , Ai )
=(0 + iqA0 , i + iAi )
=(D0 , Di )
=(D0 , Di ) ,
(1.128)
Di = i + iqAi .
A A0 = A +
(1.129)
A0 A00 = A0
(1.130)
En notacion de cuadrivectores
0
, A +
A A = A
t
i
0
= A
, A + i
t
= A0 0 , Ai i
= A0 , Ai 0 , i
A A0 =A .
(1.131)
(1.132)
RELATIVISTA
1.4. NOTACION
1.4.2
23
The scalar field is defined by their properties under Lorentz transformation. In section 1.1.3 we
study the behavior of one scalar field under a spacetime translation. Under a general Lorentz
transformation
x x0 = x ,
(1.133)
Now we will study the effect of a Lorentz tranformation on the field (x), for example under a boost.
By definition the scalar field does not change by the Lorentz transformation, the functional form is
unaltered the scalar field still satisfy
(x) 0 (x0 ) = (x) .
(1.134)
0 (x0 ) = (1 x0 ) .
(1.135)
Therefore, for an arbitrary space-time point we have that the scalar field transforms under a Lorentz
transformation as
(x) 0 (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 x0 = 1 x
= x
=x ,
(1.137)
1
1 1
=
x
x0
(1.138)
1
1 1
=
,
x
x0
(1.139)
or
(1.140)
24
1
=
x0
x
0
1
,
=
(1.141)
The field A (x) transforms simultaneously as field and as vector under Lorentz transformation
A (x) A0 (x0 ) = A (1 x) .
1.5
(1.142)
We are now are in position to answer the following question: What is the most general Lagrangian
for a the fourcomponents field A compatible with Lorentz invariance and the gauge transformation
A A0 = A (x) ?
(1.143)
Definiendo
F = A A
G = A + A
El Lagrangiano que da lugar a una Accion invariante de Lorentz para el cuadrivector A es, hasta
derivadas totales y potencias en los campos de hasta dimension 4:
1
1
1
L = F F G G J A + m2 A A + 1 A (x)A (x)A (x) + 2 A A A A
4
4
2
(1.145)
(1.146)
25
A 1 x A 1 x A 1 x
= A 1 x A 1 x A 1 x
= A 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) L0 (x) = L(1 x)
(1.148)
Since the Action involves the integration over all the points, it is invariant under the Lorentz transformation. The J (x) does not involves the introduction a new vector field, because it will be identified
later as the 4current.
Terms like
K A (x)A (x)A (x) ,
(1.149)
=( A0 A0 )
= A A +
= A A +
=F
(1.151)
Si queremos que la Accion refleja las simetras de las ecuaciones de Maxwell debemos mantener
solo los terminos del Lagrangiano para A en (1.144) que sean invariantes hasta una derivada total.
Bajo una transformacion gauge, cada uno de los terminos
1
1
G G + m2 A A + 1 A A A + 2 A A A A + 3 F A A + 4 G A A + K (x)A A A
4
2
dan lugar a un L =
6 (algo) y la Accion no es invariante bajo la transformacion gauge. Para los
terminos restantes
1
L = F F J A ,
(1.152)
4
26
(1.153)
d4 x [ (J ) ( J )(x)]
Z
= d4 x( J )(x)
Z
Z
3
= d x
dt( J )(x) .
S =
(1.154)
In order to have S = 0 we need to assume for the while that J = 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
1
(1.155)
L = F F J A
4
es el mas general que da lugar a una Accion invariante de Lorentz e invariante gauge local.
The definition of F already includes the homogeneous Maxwell equations. To see this we note
first that the only non-zero F components are
(
F 0 = F i0 = 0
F =
(1.156)
F l = F ml = l
For = 0 we have
F i0 = i A0 0 Ai
A0 Ai
=(
)
xi
x0
A0 Ai
= ( i + 0 )
x
x
i
=E
(1.157)
27
A
.
t
(1.158)
= (lj mi li mj ) i Aj
= (lj mi li mj )i Aj
= (li mj lj mi )i Aj
Aj
= (li mj lj mi ) i
x
Aj
= lmk ijk i
x
= lmk ( A)k
= lmk B k ,
(1.159)
where
B = A.
(1.160)
Then we have
E1
{F } =
E 2
E3
0
E 1
=
E 2
E3
E 1
E 2
E 3
0
213 B 3 312 B 2
3
123 B
0
321 B 1
132 B 2 231 B 1
0
E 1 E 2 E 3
0
B 3 B 2
.
B3
0
B 1
B 2 B 1
0
B
,
t
A
t
(1.161)
28
and
B = ( A)
=0
which are just the homogeneous Maxwell equations. Therefore the expression
F = A A .
(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 F =( A A )( A A )
= A A A A A A + A A
=g g ( A A A A A A + A A ).
Entonces
F F =g g ( A + A A A
( A )
A A + A + A ).
=g g A + g g A g g A g g A
g g A g g A + g g A + g g A
= A + A A A A A + A + A
=4( A A )
F F = 4F
( A )
(1.163)
( A )
A
1
A
(F F ) + J
4
( A )
A
F + J
F
=0
=0
=0
= J.
(1.164)
29
Como era de esperarse una Accion 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
F = J .
(1.165)
+ F )
+ F )
+ F )
F )
(1.166)
(1.167)
and therefore
E = .
while for = k we have
(1.168)
30
F k = J k
i F ik + 0 F 0k = J k
i F ki 0 F k0 = J k
(ikj B j ) E k
= Jk
i
x
t
B j
E k
ijk i
= Jk
x
t
E k
( B)k
= J k. .
t
(1.169)
and therefore
B
E
= J.
t
(1.170)
where F = A A ,
(1.171)
B = 0,
E+
E = ,
1.5.1
(1.172)
(1.173)
(
F0i = F0 = g00 gij F 0j = F 0i
para = 0
para = i
(1.174)
T =
(1.175)
31
1
T00 = F 0 (0 A ) + F F + J A
4
=i
=0
1 z 0}| { 1 z i}| {
0
= F (0 A ) + F F0 + F Fi +J A
4
4
1 0
1
0
0
= F A0 F F0 + F F0 + F i Fi + J A .
4
4
(1.176)
=j
1 i
1 i0
1 z 0i}| { 1 z ji}| {
1 0
i0
F F0 + F F0 + F Fi = F Fi0 + F Fi0 + F F0i + F Fji
4
4
4
4
4
1 i0
1 i0
1
i0
= F Fi0 + F Fi0 + F Fi0 + F ji Fji
4
4
4
1 i0
1 ji
= F Fi0 + F Fji .
2
4
0
(1.177)
Ademas
F 0 A0 + J A = (A0 F 0 ) + A0 F 0 + J A
= (A0 F 0 ) A0 F 0 + J A
= (A0 F 0 ) A0 J 0 + J A
= i (A0 F 0i ) J A .
(1.178)
32
Entonces
1
1
T00 = i (A0 F 0i ) F i0 Fi0 + F ji Fji J A
2
4
1
1
= i (A0 F 0i ) + F i0 F i0 + F ji F ji J A,
2
4
1 i i 1
= E E + ijk B k ijl B l + i (A0 E i ) J A,
2
4
1 2 1
= E + kl B k B l + (A0 E) J A
2
2
1 2 1 2
= E + B + (A0 E) J A
2
2
(1.179)
(1.180)
(1.181)
En vista a la ec. (1.176), ya que la densidad Lagrangiana esta definida hasta una derivada total,
como (A0 E) = (A0 F 0 ), la densidad Hamiltoniana tambien estara definida hasta una derivada
total. De hecho, el Hamiltoniano es
Z
Z
1
3
2
2
d x (E + B ) +
d3 x (A0 E)
H=
2 V
V
Z
1
=
d3 x (E2 + B2 ),
2 V
(1.182)
y corresponde a la expresion 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. SCRODINGER
EQUATION IN PRESENCE OF THE ELECTROMAGNETIC FIELD
33
Pointing:
L
i A
(0 A )
= F 0 i A
Ti0 =
= F 0j (i Aj j Ai ) F 0j j Ai
= F 0j Fij F 0j j Ai
= F 0j F ij j (F 0j Ai ) + (j F 0j )Ai
=E j jik B k + j (E j Ai ) + (J 0 )Ai
= (E B)i (Ai E) Ai
(1.183)
P =
x Ti0
Z
=
ZV
P=
V
1.6
d x (E B) +
Z
V
d3 x (Ai E)
d3 x (E B) .
(1.184)
Scr
odinger Equation in presence of the electromagnetic
field
Once we have established the set of fields, as in this case , , and A , we should write the most
general Lagrangian. Therefore
L(, , , , A ) =
1 X
i
1
(Di ) Di [ D0 (D0 ) ] + V (x) F F J A .
2m i
2
4
(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 field, but with the normal derivative replaced by the
covariant one:
L(, , , , A ) =
1 X
i
1
(Di ) Di [ D0 (D0 ) ] F F .
2m i
2
4
(1.186)
34
1 X
i
1
(i + iqAi ) (i + iqAi ) [ (0 + iqA0 ) (0 + iqA0 ) ] F F
2m i
2
4
i
1
1 X
(i iqAi ) (i + iqAi ) [ (0 + iqA0 ) (0 iqA0 ) ] F F
=
2m i
2
4
1 X
=
i i iq Ai i + iqi Ai + q 2 Ai Ai
2m i
L=
i
1
[ 0 + iq A0 (0 ) + iqA0 ] F F
2
4
1 X
=
i i iq Ai i + iqi Ai + q 2 Ai Ai
2m i
i
1
[ 0 (0 ) + 2iq A0 ] F F .
2
4
(1.187)
1
i i iq Ai i + iqi Ai + q 2 Ai Ai
2m
i
1
[ 0 (0 ) + 2iq A0 ] F F .
2
4
L=
From this we can obtain the Euler-Lagrange equation for each field.
(1.188)
1.6. SCRODINGER
EQUATION IN PRESENCE OF THE ELECTROMAGNETIC FIELD
1.6.1
35
L
L
=0
( )
L
L
L
+ i
=0
0
(0 )
(i )
i
1 i
1
i
i
i
2
i
0
i + iqA
iqAi + q Ai A (0 + 2iqA0 ) =0
2
2m
2m
2
1
i0 qA0
i i + iqAi + iqAi i + iqAi =0
2m
1
(i + iqAi )( i + iqAi ) =0
i(0 + iqA0 )
2m
1 X
Di Di =0 ,
iD0 +
2m i
(1.189)
If we define
D iqA .
(1.190)
we have in components:
D i = i iqAi
D i = i + iqAi .
(1.191)
iD0 =
(1.192)
que corresponde a la ecuacion de Scrodinger con la derivada normal reemplazada por la derivada
covariante.
36
+ iqA0
i
t
i qA0
t
i q
t
b q
H
1 X
(i + iqAi )2
2m i
1 X
=
(i iqAi )2
2m i
1
( iqA)2
2m
i2
=
( iqA)2
2m
1
=
(i + qA)2
2m
1
(i qA)2
=
2m
1
=
(b
p qA)2 .
2m
=
(1.193)
In this way, the Scrodinger equation in presence of the electromagnetic field, can be obtained from
the original Scrodinger equation but with the minimum substitution:
b H
b q
H
b b
p
p qA .
(1.194)
2
(i qA) + qA0 .
i =
(1.195)
t
2m
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 transformacion de la funcion de onda, 0 , donde 0 satisface la ecuacion
1 02 0
0
iD0 0 =
D
2m
0
1
0 2
0
(i qA ) + qA 0 0 .
i =
t
2m
(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 significa que
mantiene la misma forma bajo una transformacion gauge.
1.6. SCRODINGER
EQUATION IN PRESENCE OF THE ELECTROMAGNETIC FIELD
37
Ejemplo:
Demuestre que la ec. (1.196) es covariante:
Como
0 = ei(x)
(1.197)
Entonces
D 0 0 = [( iqA) i] ei(x)
(1.198)
y
2
D 0 0 = D 0 (D 0 0 )
= [( iqA) i] ei(x) (D)
(1.199)
D0 0 = ei(x) (D0 )
De modo que
D D0 0 = ei(x) (D )
(1.200)
(1.201)
y la derivada covariante del campo transforma como el campo. Tenemos entonces que
1 02 0
D
2m
1 i(x) 2
iei(x) D0 =
e
D
2m
1 2
D
iD0 =
2m
0
iD0 0 =
(1.202)
En resumen, para
D = + iqA
(1.203)
38
y reemplazando q tenemos
A A0 = A (x)
0 = eiq(x)
D D0 0 = eiq(x) (D ) .
(1.204)
1.6.2
L
L
= 0.
( A )
A
(1.205)
L
=0
A
L
=0
Aj
1
[iq( i ) iq ( i ) + 2q 2 Ai ] =0
2m
iq
[( i ) ( i ) 2iq Ai ] =0
(F j )
2m
iq
(F j )
[( i ) iq Ai ( i ) iq Ai ] =0
2m
iq
(F j )
{[( i iqAi ) ] ( i + iqAi )} =0
2m
iq
(F j )
[(Di ) Di ] =0 ,
2m
(F j )
(1.206)
y otra para A0
L
=0
A0
(F 0 ) + q =0
(F 0 ) +
(1.207)
1.6. SCRODINGER
EQUATION IN PRESENCE OF THE ELECTROMAGNETIC FIELD
39
Entonces
F = j
con
(
q
=0
j = iq
[(Di ) Di ] = i
2m
(1.208)
Que incluye el termino corriente para una partcula cargada y es diferente de la corriente de probabilidad en ec. (1.57). En otras palabras es la carga electrica la que se converva localmente.
1.6.3
Conserved currents
0
0
= L
L
+ i
i
J =
=0
.
=i
i
i
J 0 = (iq) iq
2
2
=q ,
1
[(i iqAi ) iq iq (i + iqAi ) ]
2m
iq
Ji =
[(Di ) (Di )] .
2m
(1.209)
(1.210)
Ji =
(1.211)
When is fixed to 1 as in ec. (1.57) to define the probability, we get eq. (1.465).
It is worth to notice that for T00 , and Ti0 we should obtain
b =i q
H
t
b = i qA .
p
(1.212)
40
1.7
Ejemplo:
Muestre que los campos electromagneticos son invariantes bajo las siguientes transformaciones
A A0 = A +
0 =
(1.213)
Ya que
E E0 = +
= E
t
t
t
(1.214)
B B0 = A + = B
(1.215)
| {z }
=0
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 A0 =A
(1.216)
(1.217)
tal que
UI = 1
(x) = (x)
tal que
U 1 = ei(x)
Note que si
i
A A0 = U A U 1 + ( U )U 1
q
(1.218)
41
(1.219)
Entonces
i
A0 =[1 + i(x) + O(2 )]A [1 i(x) + O(2 )] + (i (x))[1 + i(x) + O(2 )][1 i(x) + O(2 )]
q
1
=A (x) + O(2 )
(1.220)
q
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 finita para
el foton pueden inferirse de un conjunto de postulados que hacen de las ecuaciones de Proca la u
nica
generalizacion posible de las ecuaciones de Maxwell [?].
Teniendo en cuenta solo el termino de masa en la ec. (1.155)
1
1
L = F F + m2 A A J A .
4
2
Usando las ecuaciones de Euler-Lagrange, tenemos
1
1 2
F F
m A A J A = 0
4
( A )
A 2
F + m2 A = J .
(1.221)
(1.222)
(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. (??),
42
emerge como una restriccion 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 A = 0 y las ec. (1.222), como:
F + m2 A = J
A A + m2 A = J
( + m2 )A =J
(1.224)
donde esta definido en la ec. (1.123). En ausencia de corrientes, cada una de las componentes del
campo vectorial satisface la ecuacion 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)
1
1
L = F F + m2 A A J A
4
2
1
1
= ( A A + A A A A A A ) + m2 A A J A
4
2
1
1 2
= A A m A A + J A ,
2
2
(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 solo del
espn de las partculas. El termino cuadratico en los campos corresponde al termino de masa, y el
u
ltimo corresponde a la interaccion del campo con una corriente. Cuando un Lagrangiano contiene
solo 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 seran llamadas
Lagrangiano de Interaccion. De este modo podemos reescribir el Lagrangiano (1.225) como
L = Lfree + Lint ,
donde,
1
1
Lfree = A A m2 A A
2
2
Lint = J A .
(1.226)
Debido a que la teora masiva ya no es invariante gauge, la condicion de Lorentz aparece automaticamente como la u
nica restriccion apropiada sobre el campo vectorial.
Una vez se toma en cuenta la condicion de Lorentz el campo masivo libre puede expandirse en
ondas planas con tres grados de libertad independientes de polarizacion. Dos de estos corresponden
43
a los dos estados transversos que aparecen en las ondas electromagneticas (A1 , A2 ), y el tercero (A3 )
corresponde a un estado longitudinal en la direccion del momento de la partcula [?].
Aunque hemos hecho el analisis de la ecuacion de Proca permitiendo un termino de masa para
el foton, 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 foton es m < 6 1017 eV
(1.1 1052 Kg) [?]. Debido al principio gauge local, desde el punto teorico se espera que la masa del
foton 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
1
1
1
=g
1
1
1
=
g
x 1 x
=g 1 x 1 x
= 1 x 1 x .
(1.229)
Since (x) 0 (x) = (1 x), under a Lorentz transformation the full Lagrangian transform as
L(x) L0 (x) = L(1 x)
(1.230)
Since the Action involves the integration over all the points, it is invariant under the Lorentz transformation.
44
(1.231)
2
2
+ m = .
t2
(1.232)
2
2
+ m = 0.
t2
(1.233)
Que corresponde a la ecuacion de Klein-Gordon (??). Una expresion escrita en terminos de productos
escalares de Lorentz se dice que esta en forma covariante.
De acuerdo a la ec. (1.226), tenemos
1
1
Lfree = m2 2
2
2
Lint =
T =
L
L ,
( )
(1.234)
(1.235)
45
1
H=
+ ()2 + m2 2 .
2
t
2
(1.236)
donde
L
(/t)
=
.
t
(x) =
(1.237)
La densidad de momentum es
L
i
(0 )
Ti0 = 0 i
T0 = .
t
Ti0 =
1.9.1
(1.238)
Complex scalars
En la seccion 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 mecanica cuantica
la funcion de onda compleja puede describir parcialmente a un electron cargado. Sin embargo la
funcion de onda del electron tambien debe ser generalizada para poder dar cuenta del espn. Esto
corresponde al funcion de onda de la ecuacion de Dirac en la seccion ??.
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
1
1
1
1
L = [ 1 1 m2 21 ] + [ 2 2 m2 22 ]
2
2
2
2
(1.239)
46
Si definimos
1 + i2
i2
1
=
,
2
2 =(1 + i2 )
2 =(1 i2 ).
=
then
(1.240)
and
(1.241)
Therefore
2( + ) =21
2( ) =2i2 .
Then
+
1 =
2
2 = .
2i
(1.242)
(1.243)
(1.244)
L
L
=0
( )
+ m2 = 0
( + m2 ) = 0,
(1.245)
47
(1.246)
De este modo tanto , como , satisfacen la ecuacion de Klein-Gordon. Cada campo ademas
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 conservada puede definida en la seccion ??, eq. (1.20)
L
L
+
( )
( )
J = + .
J =
(1.247)
0 = ei (1 + i)
= + i.
(1.248)
Entonces,
= i
= i .
(1.249)
(1.250)
J i( ),
(1.251)
= J i(
).
t
t
(1.252)
J = i( ),
(1.253)
Definimos J como
Como puede ser negativo no puede interpretarse como una probalidad, como se hizo con la funcion
de onda de la ecuacion de Scrodinger. Esto presento un obstaculo en la interpretacion inicial de la
ecuacion de Klein-Gordon. Sin embargo una vez se cuantiza el campo escalar la probabilidad de los
estados cuanticos queda bien definida [?].
48
1.10
(1.254)
does not left the Action invariant. To have a proper formulation of the quantum mechanics through
the general equation
i
,
= H
t
(1.255)
(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)aab 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 fields
(x) 0 (x0 ) =(x)
A (x) A0 = A (1 x)
(x) 0 (x) =S()(1 x)
1.11
Scalar field,
Vector field,
Spinor field.
(1.258)
Diracs Action
S ,
=H
t
(1.259)
49
(1.260)
In order to have a well defined 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
,
= H
t
(1.261)
like the Scrodinger Lagrangian. However, this automatically imply that the Lagrangian will be also
linear in the spacial derivatives. A pure scalar field cannot involve a Lorentz invariant term of only
first derivatives (see eq. (1.254)). Therefore the proposed field must have some internal structure
associated with some representation of the Lorentz Group. Therefore we build the Lagrangian for a
field of several components
1
2
(1.262)
= ..
.
n
1.11.1
Lorentz transformation
If the field 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) 0 (x) = S() 1 x .
(1.263)
One possible invariant could be the term (x)(x). However, under a Lorentz transformation we
should have S S. As we cannot assume that S() is unitary, the solution is to define the adjoint
spinor
= b .
(1.264)
0
(1.265)
0
(x)(x) (x) 0 (x) = 1 x S ()bS() 1 x
(1.266)
which transforms as
and,
50
The condition that must be fulfilled for Lorentz invariance of the Action is
S ()bS() =b ,
(1.267)
0
(x)(x) (x) 0 (x) = 1 x 1 x ,
(1.268)
and therefore,
and:
0
(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:
1
(1.270)
S 1 () S() = .
(1.271)
(1.272)
Where the coefficients have been already fixed by convenience. Since the Action is real, it is convenient
to rewrite this as
L = i m
1
= i + i m
2
i
i
= ( ) + i m
2
2
i
i
= ( ) m .
2
2
(1.273)
51
(1.274)
ya que
i
i
L =
b b m b
2
2
i 2
i
2
=
b b b b m b
2
2
i
i
b b b b m
=
2
2
i
i
m
=
2
2
1.11.2
De la ec. (??)
#
"
L
L
J0 =
+
(0 )
0
= i 0
(1.275)
(1.276)
= i.
(1.277)
J 0 = 0
(1.278)
de modo que
Por consiguiente
Para que J 0 pueda interpretarse como una densidad de probabilidad, se debe cumplir
b 0 = I
(1.279)
52
(1.280)
(1.281)
= 0 .
(1.282)
0 = 0
0 =1
S () 0 S() = 0 .
0 0 =
(1.283)
En general
#
"
L
L
+
J
( )
(i)
i
(i)
i
=
(1.284)
J = b .
(1.285)
1.11.3
Tensor momento-energa
L
L
L
0 + 0
(0 )
0
0 0 L
= i
T00 =
i i + m,
= i
p + m),
= (
= 0 ( p + m),
= H,
(1.286)
53
= 0 ( p + m)
H
(1.287)
= H
t
(1.288)
d3 x.
H
(1.289)
Ademas
L
L
i + i
(0 )
0
0 i
= i
Ti0 =
= (ii )
(1.290)
de modo que
h
pi =
1.11.4
d3 x
p
(1.291)
Ecuaciones de Euler-Lagrange
= H
t
(1.292)
con el Hamiltoniano dado en la ec. (1.289), que corresponde a un Lagrangiano de solo 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
L
=0
L
=0
i m = 0.
(1.293)
54
Expandiendo
i 0 0 + i i i m = 0
i 0 0 (i) m = 0,
+ m),
i 0 0 = ( p
de donde
i 0
= 0 ( p + m).
t
(1.294)
tenemos que
0
2
= 1.
(1.295)
De la ec. (1.287)
= 0 ( p + m),
H
(1.296)
= 0.
i H
t
(1.297)
El campo tambien debe satisfacer la ecuacion de Klein-Gordon. Podemos derivar dicha ecuacion
aplicando el operador
i H
t
2
+i
t2
i H =0
i H
t
t
i H
i
H = 0
t
t
!
H
iH
+ H
2 = 0
+ iH
t
t
t
2
+ H = 0.
t2
(1.298)
55
2 = (0 p + 0 m)(0 p + 0 m)
H
= (0 p)(0 p) + m0 p0 + m02 p + m2
(1.299)
Sea
= 0
i = i
i = i
(1.300)
2 = ( p)( p) + m p + m p + m2
H
= ( p)( p) + m( + ) p + m2
(1.301)
(A )2 =
X
i
Ai i +
X
i<j
Ai , Aj i j
(1.302)
56
[(A )] =
XX
Ai i Aj j
ij
XX
i j
Ai Aj
ij
i j Ai Aj
ij
!
=
X X
i Ai Ai +
i j Ai Aj +
i j Ai Aj
i>j
i<j
!
=
X X
i2
Ai Ai
"
X X
i j
Ai Aj
i<j
Aj Ai
#
i2
Ai Ai +
i j Ai Aj + Aj Ai
i<j
#
X
X
Ai Ai +
i j Ai , Aj
i<j
"
=
j i
j>i
"
=
#
X
i2
i2
A +
j
Ai , Aj
i<j
(1.303)
Entonces
2 =i2 p2i +
H
X
i<j
{i , j } pi pj + m(i + i )pi + m2
(1.304)
i 6= j
2 = 2 + m2
H
2
2
+m =0
t2
+ m2 = 0
(1.305)
(1.306)
(1.307)
57
i 2
= 1 0 i 0 i =
i 2
2
=1
i 2
= 1 = 1
i 0 + 0 i = i, 0 = 0
(1.308)
De modo que
i , j = 0 i 0 j + 0 j 0 i = 0
0 0 i j 0 0 j i = 0
i j
j i
+ =0
i j
, = 0
i 6= j
i 6= j
i 6= j
i 6= j
(1.309)
(1.310)
= ( 0 , i )
(1.311)
0 0 = .
(1.312)
donde
1.11.5
i m = 0,
(1.314)
= 0 .
(1.315)
De la ec. (1.312)
(
0 = 0
= 0 0
2
i = 0 i = i
=0
.
=i
(1.316)
58
Definiendo
5 = i0 1 2 3 ,
(1.317)
entonces,
52
52
52
52
52
52
= 0 1 2 3 0 1 2 3
= + 02 1 2 3 1 2 3
= + 1 2 3 1 2 3
= 2 3 2 3
=2 2 3 3
=1 .
52 = 1,
(1.318)
(1.319)
Teniendo en cuenta que 2 1 y conmuta con las demas matrices, tenemos por ejemplo
5 3
5 2
5 1
5 0
=i0 1 2 32 = 32 i0 1 2 = 3 i0 1 2 3 = 3 5
= i0 1 22 3 = 22 i0 1 3 = 2 i0 1 2 3 = 2 5
=i0 12 2 3 = 12 i0 2 3 = 1 i0 1 2 3 = 1 5
=i0 1 2 3 0 = 02 i1 2 3 = 0 5 .
(1.320)
De modo que
{ , 5 } = 0.
(1.321)
(1.322)
y por consiguiente
Tr = 0.
(1.323)
U U ,
(1.324)
De otro lado, si
59
para alguna matriz unitaria U , entonces corresponde a otra representacion de algebra de Dirac
en ec. (1.310), ya que
{
, } = U U , U U
= U { , } U
= 2g U U
= 2g 1.
(1.325)
(1.326)
i , j = 2 ij ,
(1.327)
y por lo tanto pueden identificarse con las tres matrices de Pauli. Como en general tenemos 4 matrices
independientes, su dimension mnima debe ser 4.
1.11.6
i = i ,
para
U = 0
(1.328)
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
i j
J , J = iijk J k .
The Pauli matrices are set of matrices satisfying this commutation relations:
i j
k
,
= i ijk
2 2
2
(1.329)
(1.330)
60
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.
Definiendo los generadores de SU (2) como
Ti =
i
,
2
(1.332)
i
i
1 + iT i i = 1 + i
i
i .
2
(1.333)
(1.334)
det (i ) = 1
{i , j } = 2ij I i2 = I
Tr i j = 2 ij
i j = iijk k + ij
(1.335)
Ademas
In [9]:
It is generally true that one can find matrix representations of a continuous group by
finding matrix representations of the generators of the group (which must satisfy the
proper commutation relations), then exponentiating these infinitesimal 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 commutation relations by writing the generators as differential operators; from the expression
J = x p = x (i) ,
the angular momentum commutation relations (1.329) follow straightforwardly.
(1.336)
61
(1.337)
=i(li mj lj mi )xi j
=i(xl m xm l ) .
(1.338)
Involving three generators. The generalization to four-dimensions give to arise three further generators J 0i :
J = i(x x ) .
(1.339)
(1.340)
From [9]:
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
J
= exp i
2
To find a representation of the usual boosts and rotations, consider a boost
0 t+vx
t
t
cosh sinh 0 0
t
1v 2
x+vt
0
x
x0 = 1v2 = sinh cosh 0 0 x = { } {x } ,
{x } =
y
y y 0
0
1 0 y
0
z
z
0
0
0 1
z
z
(1.341)
(1.342)
Since
cosh =
sinh =
X
2n
n=0
X
n=0
2n!
1 + O( 2 )
2n+1
+ O( 2 ) ,
(2n + 1)!
(1.343)
62
{ }xboost
0
0
1
0
0
0
0
1
0
0
0
.
0
1
Similarly a rotation by an infinitesimal angle = 3 along xyplane (or about the zaxis)
1 0 0 0
0 1 0
{ }xyrotation
0 1 0 .
0 0 0 1
(1.344)
(1.345)
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
(J )
2
(1.348)
1
= 1 + i bi + i ijk rjk ,
2
(1.349)
bi = iJ i0
1.11.7
rjk = iJ jk .
(1.350)
or S () 0 = 0 S 1 () .
(1.351)
63
In order to find a representation of the Lorentz Group in terms of the Dirac matrices we propose
1
S() = 1 + i B i + i ijk Rjk .
2
(1.352)
Instead of show the Lorentz invariance of the Dirac Action, we use the conditions derived from the
invariance, to find a representation in terms of the Dirac matrices for B i and Rjk . 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
1
Bi = 0 i
2
1
Rjk = j k ,
2
(1.354)
i
[ , ] .
4
(1.355)
S 0i =
i 0 i
= iB i
2
i i j
= iRij .
2
(1.356)
It is worth notices that in fact S satisfy the Lorentz algebra, and therefore are the generators of
the Lorentz group elements:
S
S() = exp i
2
i
1 S .
(1.357)
2
Another consistency check is
S () 0 S() = 0 ,
(1.358)
64
or equivalently
S () 0 = 0 S 1 ()
i
i
0
0
= 1 + S
1 + S
2
2
0
0
S = S .
(1.359)
2
2
0 = 0 ,
(1.360)
we have
i 0
[ , ]
4
i
= , 0
4
i 0
= ,
4
i
= [ , ] 0
4
= 0 S
S 0 =
1.11.8
(1.361)
Diracs Lagrangian
(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 coefficients. In short
65
5
= 2i [ , ]
Transformacion
Escalar (S)
Pseudoescalar (P)
Vector (V)
Vector axial (A)
Tensor antisimetrico (T)
N
umero
1
1
4
4
6
16
Escalar en Dirac
. Demostracion
1 x)S 1 () S()(1 x)
J (x) (x)
(x) (
1 x) (1 x)
= (
= J (1 x) .
(1.363)
(1.364)
i
24
(1.365)
and
det = 1 2 3 4 .
(1.366)
66
1.12
Electrodin
amica Cu
antica
Para hacer el Lagrangiano en ec. (1.313) invariante gauge local bajo U (1)Q , procedemos de la forma
usual. El campo transforma como
0 = ei(x)Q
i(x)Q ,
0 = e
(1.367)
(1.368)
donde e es la carga electrica del electron. De esta forma, si e es el campo que representa al electron
eQe = e(1)e = ee .
(1.369)
El Lagrangiano correspondiente a la interaccion de un fermion y el campo electromagnetico corresponde 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 Accion bajo un
cambio de fase local
(1.370)
L = (i D m) 41 F F ,
y es invariante bajo transformaciones locales U (1)Q . Desarrollando la expresion anterior, tenemos
L = [i ( ieQA ) m] 14 F F
A 1 F F .
= (i m) + eQ
4
(1.371)
Este Lagrangiano da lugar a la Accion de la teora conocida como Electrodinamica Cuantica (QED
de sus siglas en ingles).
tenemos
Aplicando las ecuaciones de Euler-Lagrange para ,
(i m) + eQ A = 0
(i i2 eQ A m) = 0
[i ( ieQA ) m] = 0
(i D m) = 0.
(1.372)
1.12. ELECTRODINAMICA
CUANTICA
67
Que corresponde a la ecuacion de Dirac en presencia del campo electromagnetico. Mientras que para
el campo A , tenemos
1
F F
A = 0
eQ
4
( A )
A
F = eQ
(1.373)
Definimos entonces la corriente electromagnetica generada por el fermion como
.
j = eQ
(1.374)
L
L
+
( )
= i (i(x)Q)
= (x)Q ,
(1.375)
(1.376)
(1.377)
Para la QED solo hay un termino de interaccion que es suficiente para explicar todos los fenomenos
electromagneticos y su interaccion con la materia. Este esta representado por el diagrama de Feynman
mostrado en la Figura 1.2
La repulsion electromagnetica esta representada por la figura 1.3. En la Figura (a) el primer
electron emite un foton y se dispersa, mientras que el segundo absorbe el foton y se dispersa en
la direccion opuesta. En la Figura (b) el primer electon absorve el foton emitido por el segundo
electron. Los dos diagrams se representa por uno u
nico con el foton en horizontal como se muestra
en la Figura (c).
68
Figure 1.3: Electromagnetic repulsion. The diagrams (a) and (b) are summarized in the diagram (c)
1.13. CROMODINAMICA
CUANTICA
1.13
69
Cromodin
amica Cu
antica
Los protones, neutrones, piones, kaones y demas hadrones, son partculas compuestas de constituyentes elementales llamados quarks. Por ejemplo los protones, neutrones y piones estan constituidos de quarks up y down. Los hadrones estan dividos en bariones, B, constituidos de tres quarks,
y los mesones, M , de dos. Para satisfacer el principio de exclusion de Pauli, y justificar el confinamiento de los hadrones, se requiere que cada quark contenga Nc 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, Nc = 3. De este modo cada quark
q = u, d, c, s, t, b viene en tres colores
q = q1 , q2 , q3 = qr , qb , qg ,
(1.378)
donde los u
ltimos subndices hacen referencia a los colores red, blue, green. De este modo los Bariones
y mesones estan descritos por combinaciones singletes de color del tipo qr qb qg y qr qr ,
1
B = |q q q i
6
1
M = |
q q i
3
(1.379)
Estos estados son singletes de color. Una de las determinaciones de Nc proviene del observable
(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 fmax
+
=1 (e e f f )
(e+ e + )
Pfmax PNc
R
f =u
f =u (e e f f )
(e+ e + )
Pfmax
R Nc
(1.381)
70
De este modo R esta dado por la suma de las cargas electricas al cuadrado
P 2
f Qf
R Nc
Q2
fmax
=Nc
Q2f
f =u
2 2
1 2
2
Nc [( 3 ) + 2( 3 ) ] = 3 Nc
= Nc [2( 32 )2 + 2( 1
)2 ] = 10
Nc
3
9
1 2
11
2 2
Nc [2( 3 ) + 3( 3 ) ] = 9 Nc
fmax = s
2 Nc = 3,
10
= 3 Nc = 3,
fmax = c
11
Nc = 3,
fmax = b
3
f = u, d, s, fmax = s
fmax = c
fmax = b
(1.382)
En la figura, tomada de [?], se muestra el grafico de R con respecto a s(la energa de centro de masa
de la colision). 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 Nc = 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 mas simple
es imponer la simetra SU (3)c , tal que tengamos un vector compuesto de 3 espinores de Dirac en el
espacio de color:
r
qr
(1.383)
=
= qb .
b
g
qg
El Lagrangiano de Dirac con invarianza gauge global SU (3), para un quark, se puede escribir como
donde
m,
Lglobal = i
(1.384)
a
.
= exp ia
2
(1.385)
a
,
2
(1.386)
1.13. CROMODINAMICA
CUANTICA
71
a b
c
,
= if abc ,
2 2
2
(1.387)
1
Tr (G G ) ,
2
(1.388)
donde
0 = U (x)
D (D )0 = U (x)D ,
(1.389)
con la matriz 3 3
a
U (x) = exp ia (x)
2
,
(1.390)
72
a a
G igs G
2
donde hemos definido la matriz 3 3 G , como
a
Ga
(G ) =
2
D = igs
(1.391)
(1.392)
D0 U = U (x)D .
Por consiguiente
D0 U = U D
(1.393)
(1.394)
D (D )0 = U D U 1
(1.395)
( igs G0 )U
( igs G0 )U
U + ( U ) igs G0 U
( U ) igs G0 U
igs G0 U
=D0 0
=U D U 1 U
=U ( igs G )
=U igs U G
= igs U G
= ( U ) igs U G ,
(1.396)
de modo que
1
( U ) + U G
igs
i
G0 = ( U )U 1 + U G U 1 .
gs
G0 U =
(1.397)
i
( U ) U .
gs
(1.398)
1.13. CROMODINAMICA
CUANTICA
73
Entonces
i
[i( e )e (1 if f )]
gs
i
=(c + ib b c )(1 id d )Gc [i( e )e (1 if f )]
gs
1
[c id c d + ib b c ]Gc + e e
gs
1
=[c ib (c b b c )]Gc + e e
gs
1
=a Ga i(if acb a )Gc b + a a
gs
1
a
acb
= Ga + a + f Gc b
gs
a G0 a (1 + ib b )c Gc (1 id d )
(1.399)
de donde
Ga G0 a Ga +
1
a + f abc Gb c
gs
(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 parametro de transformacion gauge a (x).
Similarmente, definiendo la matriz 3 3,
G =
a
i
[D , D ] G
,
gs
2 a
(1.401)
tenemos
i
[ igs G , igs G ]
gs
i
= [( igs G ) ( igs G ) ( igs G ) ( igs G ) ]
gs
i
gs2 G G igs [ (G ) + G ] + gs2 G G + igs [ (G ) + G ]
=
gs
i
= {( ) gs2 (G G G G ) igs [( G ) ( G )]
gs
igs [G + G G + G ]}
=[ G G igs (G G G G )]
={ G G igs [G , G ]}
(1.402)
G =
74
(1.403)
que se reduce al caso Abeliano cuando los bosones gauge conmutan. En terminos de componentes
c
a
a
b
a G
a = Ga Ga igs [ Gb , Gc ]
=a Ga a Ga igs [b , c ]Gb Gc
=a Ga a Ga igs (ia fabc )Gb Gc
=a Ga a Ga + a gs fabc Gb Gc .
(1.404)
Por consiguiente
abc
e + gs f abc G G ,
G
Gb Gc G
a = Ga Ga + gs f
a
b c
(1.405)
e = G G
G
a
a
a
(1.406)
con
A diferencia del caso Abeliano G ya no es invariante bajo transformaciones gauge
G G0
i 0 0
D ,D
gs
i
=
U D U 1 , U D U 1
gs
= U G U 1 .
(1.407)
Note que con la definicion (1.401), la derivada covariante de la matrix G , transforma como la
matrix G
D G (D G )0 = U D G U 1 .
(1.408)
Para poder obtener un invariante bajo transformaciones gauge a partir del producto G G ,
debemos utilizar la traza
Tr (G G ) Tr G 0 G 0 = Tr U G U 1 U G U 1
= Tr U G G U 1
= Tr U 1 U G G
= Tr (G G ) .
(1.409)
1.13. CROMODINAMICA
CUANTICA
Teniendo en cuenta la normalizacion de las matrices de Gell-Man
Tr a b =2 ab
1
Tr a b = ab ,
2
75
(1.410)
b b
Tr (G G ) Tr G 0 G 0 = Tr a G
a G
b
= Tr a b G
a G
1
b
= ab G
a G
2
1
Ga .
= G
2 a
Expandiendo el Lagrangiano en ec. (1.388), tenemos
a a
1 Tr (G G )
L =i igs G m
2
2
igs a Ga m
1 G Ga
=i
2
4 a
m
+ gs
a Ga 1 G
=i
Ga
2
4 a
e G
ea
m
+ gs
a Ga 1 G
=i
2
4 a
1 e
ea + gs2 f abc fade G Gc Gd Ge
gs Ga fade Gd Ge + gs f abc Gb Gc G
b
4
=Lfree + Lgauge + LSI ,
(1.411)
(1.412)
donde
m
Lfree =i
e G
ea
a Ga 1 G
Lgauge =gs
2
4
1
d e
2 abc
d e
abc e a
e
LSI =
gs Ga fade G G + gs f Gb Gc G + gs f fade Gb Gc G G .
4
Hemos divido el Lagrangiano en tres partes
(1.413)
76
1
( Ga Ga ) Ga Ga Ja Ga ,
4
(1.414)
dende
Ja = gs
a
,
2
(1.415)
es la nueva corriente conservada de interaccion fuerte que surge como consecuencia de la invarianza gauge local SU (3); y
Una parte de auto-interacciones gauge:
gs abc ea gs2 abc
f G Gb Gc + f fade Gb Gc Gd Ge
2
4
gs abc
g2
= f ( Ga Ga ) Gb Gc s f abc fade Gb Gc Gd Ge .
2
4
LSI =
(1.416)
g2
gs
a
Ga f abc ( Ga Ga ) Gb Gc s f abc fade Gb Gc Gd Ge .
2
2
4
(1.417)
1.13. CROMODINAMICA
CUANTICA
77
green
t
red
red
red
green
green
d
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 estan determinadas en terminos de una u
nica constante de acoplamiento
gs . Las autointeracciones gauge pueden explicar aspectos de la interaccion fuerte como la libertada
asintotica, que consiste en que las interacciones fuertes se vuelven mas debiles a distancias cortas.
En terminos de ndices de color la corriente, y las otras partes del Lagrangiano, pueden escribirse
2
r
r rb r
g
b
r bb b
g ,
g
r g b g
g
with
r
r + bb + g
g=0
78
Figure 1.6: Triplegluon selfinteraction. The anticolors are the colors running back in time.
como
Ja
= gs q q
a
2
.
(1.418)
Note que tanto para la Electrodinamica Cuantica como para la Cromodinamica Cuantica la corriente
1.13.1
Ecuaciones de EulerLagrange
Sigiendo los mismos procedimientos anteriores debemos llegar a los siguientes resultados. Para el
campo
(i D m) = 0 ,
(1.419)
1.13. CROMODINAMICA
CUANTICA
79
L
L
a
( G )
Ga
1
a
d
d
e
= Ga gs fdbc Gb Gc
G G gs
a
2
( G )
2
g
gs dbc e
s
+ f Gd ( ba Gc + Gb ca ) + f ibc fide (g g Gb Gc Gd Ge )
2
4
1
ea gs fdbc G Gc ( da da ) gs a
= G
b
2
2
gs dac e c gs dba e b
+ f Gd G + f Gd G
2
2
gs ibc
+ f fide g g ( ba Gc Gd Ge + Gb ca Gd Ge + Gb Gc da Ge + Gb Gc Gd ea )
4
1
1
abc
ea gs f G G + gs fabc G G gs a
= G
b c
b c
2
2
2
gs
gs adb e b
c
e
f adc G
f Gd G
d G
2
2
gs
gs
gs
+ f iac fide g g Gc Gd Ge + f iba fide g g Gb Gd Ge + f ibc fiae g g Gb Gc Ge
4
4
4
gs ibc
b c d
(1.420)
+ f fida g g G G G .
4
gs iac
gs
gs
f fide g g Gc Gd Ge + f iba fide g g Gb Gd Ge + f ibc fiae g g Gb Gc Ge
4
4
4
gs ibc
b c d
+ f fida g g G G G
4
gs iac
gs
gs
gs
= f fide Gd Gc Ge + f iba fide Ge Gb Gd + f ibc fiae Gb Gc Ge + f ibc fida Gc Gb Gd
4
4
4
4
gs dac
gs dca
gs dbc
gs
e
j
e
= f fdje Gj Gc G + f fdje Ge Gc G + f fdae Gb Gc G + f dbc fdea Gc Gb Ge
4
4
4
4
gs dac
gs dca
gs
gs
e
j
dje
c
= f fdje Gj Gc G + f fdje Ge Gc G + fdac f Gj Gce G + fdca f dje Ge Gj Gc
4
4
4
4
gs abc b
gs
gs
gs abc b
b cej
= f G fcej Ge Gj f G fcej Ge Gj fabc G f Ge Gj fabc Gb f cej Ge Gj
4
4
4
4
b cej
(1.421)
= gs fabc G f Ge Gj
80
Entonces
L
L
( Ga )
Ga
abc
a
e gs facd Gc f dej Ge Gj
e
g
f
G
G
gs f acd Gc G
= G
s
a
b c gs
d
2
a
acd c
= G
G Gd gs = 0 .
a gs f
2
Entonces las Ecuaciones de Euler Lagrange para Ga , son
acd c
G
G Gd = gs
a + gs f
a
.
2
(1.422)
(1.423)
Definiendo
Ja = gs
a
,
2
(1.424)
= gs
fabc Gb G
c
a
(1.425)
(1.426)
(1.427)
donde
ja
= gs
fabc Gb G
c
a
.
+
2
(1.428)
1.13. CROMODINAMICA
CUANTICA
1.13.2
81
Toda el algebra de SU (3) se puede escribir en notacion vectorial en terminos de vectores de 8 componentes 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 representacion adjunta de SU (3).
Definiendo el producto vectorial de SU (3) como
(A B)a = fabc Ab B c ,
(1.429)
si G es un vector en el espacio SU (3) con las 8 componentes Ga , entonces podemos escribir (1.400)
como
G G0 = G +
1
+ G .
gs
(1.430)
Podemos escribir tambien la ec. (1.405) en terminos de vectores en el espacio SU (3) como:
G = G G + gs G G ,
(1.431)
(1.432)
j = gs G
Gb
+ .
2
(1.433)
(1.434)
a , b = ifabc c .
(1.435)
con
(1.436)
82
En componentes
(D )ab =ab igs (c )ab Gc
=ab gs fcab Gc
=ab + gs facb Gc .
(1.437)
c
(D )ab G
b =ab Gb + gs facb G Gb
(D )ab G
b = Ga + gs (G G )a
(D G )a = G
a + gs (G G )a ,
(1.438)
D G = G + gs G G .
(1.439)
Entonces las las Ecuaciones de Euler Lagrange para Ga , en (1.423) se pueden escribir como
D G =J ,
(1.440)
a
.
2
(1.441)
Para escribir el Lagrangiano en forma vectorial en el espacio SU (3), debemos reescribir la transformacion gauge de G en terminos de vectores de SU (3). Como
G G0 =U G U
=(1 + ib b )c Gc (1 ib b )
(1.442)
Ga G0 Ga + f abc Gb c .
(1.443)
Note que en el caso Abeliano fabc = 0, el tensor correspondiente es invariante gauge, como ocurre el
caso electromagnetico. En notacion de vectores de SU (3):
G G0 G + G .
(1.444)
1.13. CROMODINAMICA
CUANTICA
83
(1.445)
a
G G0
G
a G = G
G0 G G + G (G ) + (G ) G
=G G + G ( G ) + (G ) G
=G G (G ) G + (G ) G
=G G .
(1.446)
1 G G .
L = i igs G m
2
4
(1.447)
El Lagrangiano para los campos gauge, el cual puede generalizarse para cualquier teora SU (N ), es
1
Lgluon = Lgauge + LSI = G G J G ,
4
(1.448)
Da lugar la ecuaciones de Maxwell pero con la derivada normal reemplzada por la derivada covariante
D G =J ,
(1.449)
D G = G + gs G G .
(1.450)
donde
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 .
84
1.14
(1.451)
V () = 12 2 2 .
(1.452)
con
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 figura 1.7. En Teora
Cuantica 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. Ademas
el alejamiento del campo del punto de simetra del potencial no cuesta energa. Por consiguiente en
ese caso, el termino de interaccion
V () = 12 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 () = 21 2 2 + 14 4
2 < 0, > 0
(1.454)
85
(1.455)
(1.456)
2
.
2V
= 2 + 32 .
2
(1.457)
(1.458)
p
2 / es 22 > 0
(1.459)
86
(1.460)
y
LH = 21 H H
1
2
22 H 2 vH 3 14 H 4 + constant.
(1.461)
= 12 v 2 H 2 + 2vH + v 2 + 41 H 4 + 4vH 3 + 6H 2 v 2 + 4v 3 H + v 4
=v 2 H 2 + vH 3 + 41 H 4 + constant.
(1.462)
87
(1.463)
m2H = 2 2 = 2v 2
(1.464)
donde
Consideremos ahora un campo escalar complejo sin termino de masa, pero con potencial:
L = V ()
V () = 2 + ( )2
2 < 0, > 0
(1.465)
(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 suficientemente 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.
88
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) [9]3 :
2
1
1
1
L = i i 2 i i 2 i i ,
2
2
2
(1.467)
i 0 = Rij j ,
(1.468)
Sec. 11.1
Introduction to Chapter 20
89
En el caso de la Accion invariante gauge local bajo el Grupo U (1), tenemos el Lagrangiano (??):
L = (D ) D 2 ( )2 41 F F
(1.469)
=
(1.470)
=e
e
2
2
2
L L0 = (D )0 0 (D )0 0 2 ( )0 0 ( )0 0 41 (F F )0
(1.472)
(1.473)
LA = 41 F F + 12 g 2 v 2 A 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
mA = 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 boson 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 final en la ec. (1.473) no aparece el boson 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 boson
de Higgs.
90
1.14.1
Superconductivity
A review ot the use of the Proca Equations for a massive photon in superconductivity is given in [21].
A popularization review along this lines is in the book of Frank Wilczek The Lightness of Being
(see Additional references).
The photon mass inside a superconductor is 1011 GeV (or 1/1000 of the electron mass accordin
to [21]). Also from the article in Beamline 10 m y M = ~/c
There are two important length scales in a superconductor. The first measures how efficientrly
the condensate expels a magnetic field. In fact, the expulsion is not
Additional references:
The Lightness of Being: Mass, Ether, and the Unification of Forces, Frank Wilczek, http:
//www.amazon.com/The-Lightness-Being-Unification-Forces/dp/0465018955
http://www.scholarpedia.org/article/Englert-Brout-Higgs-Guralnik-Hagen-Kibble_
mechanism_(history)
Elementary Particle Physics: Volume 1: Quantum Field Theory and ..., Volume 1 By Yorikiyo
Nagashima, ElementaryParticlePhysics:Volume1:QuantumFieldTheoryand...,Volume1ByYorikiyoNa
From Superconductors to Supercolliders by LANCE DIXON http://www.slac.stanford.
edu/pubs/beamline/26/1/26-1-dixon.pdf
Electrodynamics of Superconductors http://www.physics.buffalo.edu/phy514/w11/index.
html
1.15
Sea
1 5
2
1 + 5
PR
.
2
PL
(1.476)
Ademas
L PL
R PR .
(1.477)
91
Entonces
= L + R .
(1.478)
PL + P R = 1
PL,R
= PL,R .
PL P R = 0
(1.479)
(1.480)
(1.481)
L =i m
=i(PL + PR ) m(PL + PR )
(1.482)
x x
R (t, x) L (t, x)
R (t, x) L (t, x) .
(1.483)
=L 0 0 L + L L
=L 0 0 0 0 L + L 0 0 L
L
=L 0 0 L + L
=R R .
(1.484)
92
Entonces
L L0 =iR R + iL L m(R L + L R ) ,
(1.485)
donde = U 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 transformacion de paridad.
La existencia de ambos espinores L,R garantizan que el Lagrangiano de Dirac es invariante bajo
la transformacion de paridad.
La corriente de la electrodinamica cuantica en ec. (1.374) (o la de la cromodinamica cuantica,
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 interaccion vectorial viola paridad y no puede tener ni interacciones electromagneticas 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 solo requiere dos grados de libertad independientes.
De otro lado, si una determinada interaccion, como es el caso de la interaccion debil, solo participa
la componente izquierda de la ec. (1.486), esta corresponde a una interaccion del tipo
L L = PR PL = PL
1 5
2
= 12 ( 5 ) ,
(1.487)
1.16
The known matter is build from the elementary set of particles defined in table
we further define the color triplets of quarks as
u1
d1
u = u 2
d = d2
u3
d3
(1.488)
93
quarks up quark
u1 , u2 , u3
down quark d1 , d2 , d3
Charge
1
0
2/3
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
The free Lagrangian containing this particles can be written as
Lfree =ie e me ee + i + iu u mu uu + id d md dd
=ieL eL + ieR eR me (eR eL + eL eR ) + iL L
+ iR R + iuL uL + iuR uR
me (uR uL + uL uR )idL dL + idR dR me (dR dL + dL dR ) .
(1.489)
donde,
L,R =PL,R ,
uL,R =PL,R u,
eL,R =PL,R e
dL,R =PL,R d .
(1.490)
Corrientes VA
En las interacciones debiles solo participan las partes izquierdas de los campos. Esto nos permite
prescindir del R , pues no tiene carga electrica, fuerte, o debil
Lfree =ieL eL + ieR eR me (eR eL + eL eR ) + iL L
+ iuL uL + iuR uR
mue(uR uL + uL uR ) + idL dL + idR dR md (dR dL + dL dR ) .
(1.491)
94
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 definiendo los dobletes:
L
u
L
Q= L ,
(1.492)
eL
dL
De otro lado, La invarianza bajo U (1)Y requiere que
YL =YL = YeL
YQ =YuL = YdL .
(1.493)
Q Q0 = exp(iT i i )Q (1 + iT i i )Q
eR e0R = eR
uR u0R = uR
dR d0R = dR .
(1.494)
donde
Ti =
i
,
2
(1.495)
(1.496)
Para obtener la interacciones del modelo estandar, reemplazamos las derivadas normales por derivadas
covariantes.
95
a
i
1
1
14 G
a G 4 Wi W 4 B B ,
(1.497)
Donde
D igs
i
a
Ga ig Wi ig 0 Y B .
2
2
(1.498)
donde
a
a
, a = 1, 2, . . . , 8
2
i
T i , i = 1, 2, 3
2
Y
8 generadores de SU (3)c
3 generadores de SU (2)L
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 Wi . To solve this problem,
we postulate a new complex scalar doublet with four degree of freedom:
+
1 + i2
=
=
.
(1.499)
0
3 + i4
The + and 0 superindexes for just for later convenience. The full Lagrangian involving those fields
are
L =iQ D Q + iL D L + ieR D eR + idR D dR + iuR D uR
a
i
1
1
14 G
a G 4 Wi W 4 B B
+ (D ) D 2 ( )2
e R + h.c)
(he LeR + hd QdR + hu Qu
(1.500)
e = i2 ,
(1.501)
a
i
1
1
Lgauge = 41 G
a G 4 Wi W 4 B B
LW BH =(D ) D 2 ( )2
e R + h.c
LYukawa =he LeR + hd QdR + hu Qu
(1.502)
96
Bajo una transformacion gauge local las derivadas covariantes de los campos (y por consiguiente
los campos) transforman como:
D L (D L)0 = exp ii T i iYL D L
D Q (D Q)0 = exp ia a ii T i iYQ D Q
D (D )0 = exp ii T i iY D
D eR (D eR )0 = exp (iYeR ) D eR = exp (iQeR ) D eR
(1.503)
donde QeR = 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 interaccion entre fermiones
y el campo escalar, correspondiente a una interaccion de Yukawa: LeR y QdR son invariantes bajo
transformaciones SU (3)c SU (2)L U (1)Y si
YQ + Ye + YuR
YL + Y + QeR =0
YQ + Y + QdR =0
= YQ Y + QuR =0 ,
From this set of three equations we obtain the three doublet hypercharges:
1
YL = ,
2
Y =
1
,
2
YQ =
1
.
6
(1.504)
0
=
Q
uL
0
+
=
dL
=0 uL + dL
=12 Q1 2 + 21 Q2 1
=ab Qa b .
(1.505)
97
(1.506)
Sin perdida de generalidad los cuatro grados de libertad de , pueden escribirse en la forma
0
ij (x)T j
=e
.
(1.507)
1 [H(x) + v]
2
El potencial escalar, definido por
V () = 2 + ( )2
(1.508)
1
1
V (H) = 2 (H + v)2 + (H + v)4 .
2
4
(1.509)
se reduce a
1.16.1
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
espontanea de la simetra SU (2)L U (1)Y
0
ij T j
=e
(1.510)
1 (H(x) + v)
2
+
0
1 1 + i3
2i
1 (H(x) + v)
2i
1 i3
2
2
+
+
i H + vi
1
=
1
2 2 (H + v i3 H i3 v)
G+
= 1
.
(H(x) + v iG0 )
2
Para SU (2)L U (1)Y tenemos cuatro generadores y cuatro bosones gauge. De acuerdo a la parametrizacion
en ec. (1.510) esperamos que aparezcan tres bosones de Goldstone y un campo de Higgs con masa,
98
de manera que quedara un generador no roto correspondiente a una simetra remanente del vaco
U (1)Q
hi
(1.511)
Se espera entonces que el espectro consista de un boson de Higgs, tres bosones gauge masivos, y
un boson gauge sin masa.
Podemos hacer una transformacion gauge similar a la de la ec. (??) sobre el campo , tal que
0
0
,
(1.512)
= 1
(H(x) + v)
2
que define el gauge unitario. En adelante sin embargo omitiremos las primas sobre los campos
0
transformados 0 y W
.
Comenzaremos analizando la parte escalar del Lagrangiano del Modelo dada en la ec. (1.502)
1
0
0
LW BH =
D
D
V (H) ,
(1.513)
H(x) + v
H(x) + v
2
donde V (H) dado en la ec. (1.463), incluye el termino de masa para el boson de Higgs (1.464):
m2H = 2 2 = 2v 2
(1.514)
Como
W =
1 0 i
1 1 0
0 1
1
2
W +
W +
W3
1 0
2 i 0
2 0 1
W1 i W2
3
W
2
1
2
=
2
1
2 W + i W
3
2
W
2
3
1
W
2W+
2W W3
2
!
1
1 W +
W
3
2
= 12
.
1
3
W
2
2
Ti Wi
1
=
2
(1.515)
D corresponde a la matrix 2 2
D =
1
gW3 + g 0 Y
2
i2 gW
!
i2 gW+
.
i 12 gW3 + g 0 Y B
(1.516)
99
Entonces
i2 gW+ (H + v)
.
H i 21 gW3 + g 0 Y B (H + v)
D =
(1.517)
De modo que
i2 gW + (H + v)
1
LW BH =
H i 21 gW3 + g 0 Y B (H + v)
2
i2 gW+ (H + v)
V (H)
H i 21 gW3 + g 0 Y B (H + v)
1 i
(H
+
v)
gW
(H
+
v)
H
+
i
gW
+
g
Y
B
=
3
2
2 2
i2 gW+ (H + v)
V (H)
H i 21 gW3 + g 0 Y B (H + v)
1
= g 2 W W+ (H + v)2 V (H)
4
1
+ H + i 21 gW3 + g 0 Y B (H + v)
2
H i 21 gW3 + g 0 Y B (H + v)
1
= V (H) + g 2 W W+ (H + v)2 +
4
2
1
1
21 gW3 + g 0 Y B (H + v)2
+ H H +
2
2
donde la u
ltima lnea corresponde a la magnitud del n
umero complejo:
H i 21 gW3 + g 0 Y B (H + v)
(1.518)
(1.519)
Entonces
1
LW BH = H H V (H)
2
gv 2
1
1
+
W W+ + g 2 W W+ H 2 + vg 2 W W+ H + LZAH ,
4
4
2
(1.520)
donde
LZAH =
1 1 2 3 1 0
0
02 2
1
g
W
W
gg
Y
W
B
gg
Y
W
B
+
g
Y
B
B
3
3
3
2
2
2 4
H 2 + 2vH + v 2
(1.521)
100
LZAH
3
g 2 gg 0
1
W
2
2
W3 B
H
+
2vH
+
v
=
2
B
gg 0 g 0
8
(1.522)
Sea
V =
cos W sin W
sin W cos W
=p
g2 + g02
g g0
,
g 0 g
(1.523)
con tan W = g 0 /g, tal que g sin W = g 0 cos W , como en la ec. (??). Note que V es una matrix
ortogonal que satisface V V T = V T V = 1. Si (ver ec. (??)),
W3
B
=V
Z
A
o
Z
A
3
W
=V
B
T
(1.524)
entonces
LZAH
2
3
1
g
gg 0
W
2
2
T
T
3
B VV
H
+
2vH
+
v
V
V
= W
2
B
gg 0 g 0
8
2
0
1
g
gg
Z
= Z A V T
H 2 + 2vH + v 2
V
0
02
A
gg
g
8
(1.525)
3
1
g 2 gg 0
g + gg 0 2
g 2 g 0 g 0 3
g g0
V = 2
2
2
g 0 g
gg 0 g 0 2
g + g 0 2 +g 2 g 0 g 2 g 0 gg 0 + gg 0
3
1
g g0
g + gg 0 2 g 2 g 0 g 0 3
= 2
g 0 g
0
0
g + g02
4
1
g + g 2 g 0 2 + g 2 g 0 2 + g 0 4 g 3 g 0 + gg 0 3 g 3 g 0 gg 0 3
= 2
0
0
g + g02
2
02
g +g 0
=
(1.526)
0
0
101
LZAH
1 g2 + g02
=
Z Z H 2 + 2vH + v 2
2
4
2
1 g
=
1 + tan2 W Z Z H 2 + 2vH + v 2
2 2
2
1
g
=
Z Z H 2 + 2vH + v 2
2 2 cos W
2
2
1
gv
1
g
=
Z Z +
Z Z H 2
2 2 cos W
2 2 cos W
2
g
+
vZ Z H
2 cos W
(1.527)
2
Z Z H +
v Z Z H
+
2 2 cos W
2 cos W
1 2
1
1
+ mW W W+ + m2W W W+ + m2Z Z Z ,
2
2
2
(1.528)
donde:
Masas gauge:
mW =
gv
2
y
mZ =
mZ =
gv
,
2 cos W
mW
.
cos W
V (H) = 21 m2H H 2 + vH 3 + 41 H 4
1
m2
1 m2H 4
= m2H H 2 + H H 3 +
H
2
2
4 2v
2v
H
H2
1
= m2H H 2 1 +
+ 2 .
2
v
4v
(1.529)
(1.530)
(1.531)
102
m2H = 22 = 2v 2 .
(1.532)
3
W
cos W sin W
Z
=
,
sin W cos W
A
B
(1.533)
g sin W = g 0 cos W .
(1.534)
tal que
1.16.2
Yukawa Lagrangian
(1.535)
definiendo
hf v
mf =
2
(1.536)
tenemos
LYukawa =me ee + md dd + mu uu +
1.16.3
me
md
mu
eeH +
ddH +
uuH .
v
v
v
(1.537)
Fermion-gauge interactions
(1.538)
103
Los terminos de interaccion generados por la simetra gauge para el campo L son:
iL D L iL L =iL (igTi Wi ig 0 YL B )L
=L
+ gT3 W + g YL B L
0
2 W
0 W+
3
0
g
=iL
L
+
gT
W
+
g
Y
B
L
L
3
L
0
2 W
eL W+
g
=L
+ LAZL
2 L W
g
= L eL W+ + eL L W + LAZL
2
=LW L + LAZL ,
(1.539)
donde
g
LW L = L eL W+ + eL L W
2
LAZL =L gT3 W3 + g 0 YL B L
Generalizando para todos los campos:
g
LW L L eL W+ + uL dL W+ + h.c .
2
(1.540)
(1.541)
=L [gT3 cW Z + gT3 sW A g 0 YL sW Z + g 0 YL cW A ] L
=L [(gcW T3 g 0 sW YL ) Z + (gsW T3 + g 0 cW YL ) A ] L ,
LAZL =L
gcW T3 g
YL Z + (gsW T3 + gsW YL ) A L
cW
=gsW L [(cot W T3 tan W YL ) Z + (T3 + YL ) A ] L .
(1.542)
(1.543)
Como el generador asociado a A debe ser el generador de carga electrica, tenemos que
e = g sin W
(1.544)
104
(1.545)
de modo que
b = (T3 + Y )L =
QL
Q 0
0 0
L=
L.
0 Qe
0 1
(1.546)
La ec. (1.545), se conoce como la relacion Gell-MannNishijima, y establece la condicion que se debe
satisfacer para obtener apropiadamente la QED a partir de la interaccion electrodebil asociada al
grupo semisimple SU (2)L U (1)Y . De esta forma
bL LA
LAZL =eL (cot W T3 tan W YL ) LZ + eL Q
h
i
bL T3 LZ + eL Q
bL LA
=eL cot W T3 tan W Q
h
i
e
bL LZ + eL Q
bL LA .
=
L 3 2s2W Q
2cW sW
Generalizando para los otros campos, tenemos
i
h
X
e
bL F Z + eF Q
b L F A .
LAZL
F 3 2s2W Q
2c
W sW
F =Q,L,e ,d ,u
R
(1.547)
(1.548)
Usando los acoplamientos gauge de los quarks con los gluones (1.413), de los fermiones con el
(1.541) y con Z y A (1.548) para expandir Lfermion en (1.538), tenemos
(1.549)
105
2vf
2af
u
1 sin2 W
1
8
3
d
1 + sin2 W
1
4
3
e
1
1
e
1 + 4 sin2 W
1
L eL W+ = PR PL eW+
= PL2 eW+
= PL eW+
1
= (1 5 )eW+ ,
2
(1.550)
h
i
b
2
b
2 b
2
Q 3 2sW QQ Q 2sW uR Qu uR 2sW dR Qd dR Z
"
!
bu
1 2s2W Q
uL
0
= u L dL
2 b
dL
0
1 2sW Qd
i
bu uR 2s2 dR Q
bd dR Z
2s2W uR Q
W
= u (1 5 )u d (1 5 )d 2sW u Qu u + d Qd d Z
2
2
1
1
1
1
2
= u
2sW Qu 5 u + d
2sW Qd + 5 d Z
2
2
2
2
(1.551)
= u (vu au 5 ) u + d (vd ad 5 ) d Z ,
donde
vf =T3f 2 sin2 W Qf
af =T3f 5
(1.552)
Los valores explcitos para vf y af en el modelo estandar, estan dados en la Tabla 1.3.
Usando las expresiones para pasar de fermiones L, R a los fermiones de Dirac de cuatro compo-
106
Lfermion =iu u + id d + ie e + iL L
a
a
+ gs u u + d d+ Ga
2
2
X
g
e
+ (1 5 )eW+ + u (1 5 )dW+ + h.c +
f (vf af 5 ) f
2c
s
2 2
W W
f =u,d,,e
be e + u Qu u + d Qd d A
+ e e Q
X
X
a
gs q qGa
=
if f +
2
q=u,d
f =u,d,,e
g
+ (1 5 )eW+ + u (1 5 )dW+ + h.c
2 2
X
e
+
f (vf af 5 ) f
2c
s
W W
f =u,d,,e
X
bf f A ,
f Q
(1.553)
+e
f =u,d,,e
1.16.4
Self-interactions
El Lagrangiano gauge
a
i
1
1
Lgauge = 14 G
a G 4 Wi W 4 B B
(1.554)
107
e
ea
Lgauge = 14 F F 14 Z Z 21 (FW
) (FW ) 14 G
a G
h
i
+
+
ie cot W (FW ) W Z (FW ) W Z + W W Z
h
i
ie (FW
) W+ A (FW ) W A + W W+ F
h
i
e2
+
2
+
+
W
W
W
W
e2 cot2 W W+ W Z Z W+ Z W Z
2
2 sin W
e2 cot2 W 2W+ W A Z W+ A W Z W+ Z W A
e2 W+ W A A W+ A W A
1 e
ea + gs2 f abc fade G Gc Gd Ge ,
gs Ga fade Gd Ge + gs f abc Gb Gc G
(1.555)
b
4
donde
(FW ) = W+ W+
e = G G .
G
a
(1.556)
108
1.16.5
Recopilando los resultados para LW BH (1.528), LYukawa (1.537), Lfermion (1.553), y Lgauge (1.555),
tenemos para f = e , e, u, d; q = u, d [con f 0 = e (d) para f = e (u) ]
L1 gen =
X
f
if ( mf ) f
e
ea
) (FW ) 14 G
41 F F 14 Z Z 21 (FW
a G
1 2 2
H
H2
1
+ 2
+ 2 H H mH H 1 +
2
v
4v
1 2
H H2
2
+
+ mW W W + mZ Z Z
1+2 + 2
2
v
v
X
X
a
q
f Qf f A
+ gs
q Ga + e
2
q
f
X
e
+
f (vf af 5 )f Z
2 cos W sin W f
"
#
X mf
X
g
+
f (1 5 )f 0 W+ + h.c +
ff H
v
2 2 f
f
h
i
+
+
ie cot W (FW ) W Z (FW ) W Z + W W Z
i
h
ie (FW
) W+ A (FW ) W A + W W+ F
h
i
e2
+
2
+
+
W
W
W
W
W
2 sin2 W
e2 cot2 W W+ W Z Z W+ Z W Z
e2 cot2 W 2W+ W A Z W+ A W Z W+ Z W A
e2 W+ W A A W+ A W A
1 e
ea + g 2 f abc fade G G Gd Ge .
gs Ga fade Gd Ge + gs f abc Gb Gc G
s
b c
4
(1.557)
109
1.16.6
Din
amica 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
i
e
L
L
L
Li = i :
L1 =
L2 =
L3 = L
eiR : eR , R , R
eL
eL
L
L
i
uL
uL
cL
t
Qi = i :
Q1 =
Q2 =
Q3 = L
uiR : uR , cR , tR
dL
dL
sL
bL
diR : dR , sR , bR .
(1.558)
Con
YLi =
1
2
YQ i =
1
6
YeiR = 1
YuiR =
2
3
YdiR =
1
.
3
(1.559)
0
0
0i
0i
0i
0i
L =i
Qi D Qi + Li D Li + eR D eR + dR D dR + uR D uR
i
0
0
0
U 0e
0
D 0
(hE
ij Li eR j + hij Qi dR j + hij Qi uR j + h.c)
i
41 Wi W
14 B B
+ (D ) D 2 + ( )2 .
(1.560)
Para aclarar la notacion, obviando de momento la definicion definitiva de hij y las primas sobre los
campos, consideremos el Lagrangiano de Yukawa para el sector down
L hD
ij dRi Qj + h.c
ea b
hD
ij dRi ab Qj + h.c
a b
e
hD
ij ab dR i Qj + h.c
a b
e
hD
ij ab (d ) 0 Qj + h.c ,
R i
(1.561)
110
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
1
e 1 2
L hD
11 (dR )1 0 Q11 + . . .
r
0 r
hD
11 dR dL + . . .
(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 (en el gauge unitario)
L dR hD Q + QL hD dR
u
H(x)+v
L
D
dR h 0
+ h.c
2
dL
H(x) + v
D
dL + h.c
dR h
2
hD
hD v
dR H(x)dL + dR dL + h.c
2
2
D
h
dR H(x)dL + dR MD dL + h.c .
2
(1.563)
La matrix 3 3 M3 es en general una matriz compleja no diagonal, la cual se debe diagonalizar con
una transformacion biunitaria (de similaridad). Retornado a la ec. (1.560), tenemos que para definir
apropiadamente la masa de los quarks, rotamos de los autoestados de interaccion a los autoestados
de masa con la matrices unitarias
0
D
dR,L 0j = (VR,L
)jk dR,L k
D
dR,L j = dR,L k (VR,L
)kj
(1.564)
Tal que
D
D
(VR,L
)ij (VR,L
)jk = ik
VLD
M M
VLD
D
VLD MD VR
D
=MD
diag Mdiag
2
,
= MD
diag
(1.565)
D
D D
V R M VL
(1.566)
111
D
D
= MD
.
VR
MD MD VR
diag
(1.567)
0
=dL i MijD dR 0j
=dLk mD
k kl dRl
=mD
k dLk dRk
(1.568)
(1.569)
V =VLU VLD
X
j
Vij Vjk = ik
X
j
Vji Vjk = ik
X
j
|Vji |2 =
X
j
|Vij |2 = 1
(1.570)
L 0i = (VLE )ij Lj
(1.571)
Con esta definicion, las corrientes debiles cargadas para los leptones siguen siendo universales. Similarmente
(1.572)
112
De modo que todas las corrientes neutras permanecen universales despues de la redefinicion 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 Estandar 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).
113
Teniendo en cuenta estos resultados podemos escribir finalmente el Lagrangiano completo del
Modelo Estandar en la Gauge Unitario, para
f =e , , , e, , , u, c, t, d, s, b;
LSM =
X
f
q =u, c, t, d, s, b;
l =e, ,
(1.573)
if ( mf ) f
e
ea
) (FW ) 41 G
41 F F 14 Z Z 12 (FW
a G
1
H
H2
+ 12 H H m2H H 2 1 +
+ 2
2
v
4v
1 2
H H2
2
+
+ mW W W + mZ Z Z
1+2 + 2
2
v
v
X
X
a
f Qf f A
q
q Ga + e
+ gs
2
q
f
X
e
+
f (vf af 5 )f Z
2 cos W sin W f
"
#
X
X
g
+
+
l (1 5 )lW +
Vij ui (1 5 )dj W + h.c
2 2 l=e
ij
X mf
ff H
+
v
f
h
i
) W+ Z (FW ) W Z + W W+ Z
ie cot W (FW
i
h
+
+
ie (FW ) W A (FW ) W A + W W F
h
i
e2
+
2
+
+
2
2
+
+
W
W
W
W
W
W
e
cot
W
W
Z
Z
W
Z
W
Z
2 sin2 W
e2 cot2 W 2W+ W A Z W+ A W Z W+ Z W A
e2 W+ W A A W+ A W A
1 e
d e
2 abc
d e
abc e a
gs Ga fade G G + gs f Gb Gc G + gs f fade Gb Gc G G .
(1.574)
4
donde ml = 0.
114
1.17
Fenomenologa Electrod
ebil
El Lagrangiano del Modelo contiene los parametros gs , g, sin W , v, mH . Alternativamente uno puede
escoger como parametros, en lugar de g, sin W , v [?]
GF = 1.166 371(6) 105 GeV2
1 = 137.035 999 679(94)
mZ = 91.1876(20) GeV
s (mZ ) = 0.1176(20) .
(1.575)
donde i = gi2 /(4). Esto tiene la ventaja de usar las tres cantidades experimentales mejor medidas.
Las relaciones
sin2 W =1
m2W
,
m2Z
m2W sin2 W =
2GF
(1.576)
determinan entonces
sin2 W =0.212
mW =80.94 GeV
(1.577)
sin2 W =0.233
mW =79.84 GeV
(1.578)
Los valores medidos son sin2 W = 0.23149(13), mW = 80.398(25) GeV, y pueden ser reproducidos por
el modelo estandar una vez se tienen en cuenta correcciones perturbativas inducidas por partculas
virtuales.
El acelerador e+ e LEP, que funciono hasta desde 1998 hasta el 2000 [?], opero a energas suficientes para producir millones de Z. Combinado con otros resultados experimentales, se pudo
verificar todo el Lagrangiano del Modelo Estandar hasta un nivel del 1 por mil. Con excepcion de
las interacciones asociadas con el Higgs.
La universalidad de los decaimientos del Z esta soportada por los resultados experimentales
siguientes donde solo se muestran los decaimientos leptonicos 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)%
(1.579)
1.17. FENOMENOLOGIA ELECTRODEBIL
115
Br(W ) = 10.59(15),
Br(W ) = 11.44(22)
(1.580)
Br(Z qq)
Nc (|vq |2 + |aq |2 )
h
2 1 i
2
1
2 + 2 sin W + 4
i
= h
2
Nc 21 + 23 sin2 W + 14
0.338 Nc = 2
0.776
= 0.225 Nc = 3
Nc
0.169 Nc = 4
(1.581)
15.12(5)
Br(Z bb)
(1.582)
116
detalles del calculo de las amplitudes. Teniendo en cuenta que el canal Z tt esta cerrado
Br(Z e+ e ) =
(Z e+ e )
total
(|ve |2 + |ae |2 )
P2
P3
2 + |a |2 ) + (|v |2 + |a |2 )] + N [
2 + |a |2 ) +
2
2
[(|v
|
(|v
|
l
l
c
u
u
i
i
l
l
l
i=1
i=1 (|vdi | + |adi | )]
(|ve |2 + |ae |2 )
3[(|ve |2 + |ae |2 ) + (|ve |2 + |ae |2 )] + 3[2(|vu |2 + |au |2 ) + 3(|vd |2 + |ad |2 )]
(|ve |2 + |ae |2 )
21|ae |2 + 3[|ve |2 + |ve |2 ] + 3[2|vu |2 + 3|vd |2 ]
(1 + 4s2 W )2 + 1
21 + 3[(1 + 4s2 W )2 + 1] + 3[2(1 83 s2 W )2 + 3(1 + 43 s2 W )2 ]
2 8s2 W + 16s4 W
42 80s2 W + 320
s4 W
3
=P
=
=
=
=
3.43%
(1.583)
(W e e )
total
(1.584)
donde, teniendo en cuenta que los canales a top estan cerrados, y usando la condicion de unitariedad
de la matriz CKM en ec. (1.570), tenemos
X
X
(W l l ) + Nc
[(W u1 di ) + (W u2 di )]
total =
i
=(W
=(W
X
e e ){3 + Nc
[|V1i |2 + |V1i |2 ]}
e e )(3 + 2Nc )
(1.585)
entonces
Br(W e e ) =
1
= 11.1%
3 + 2Nc
(1.586)
Una mejor prediccion de dichos resultados en el contexto del Modelo Estandar requiere tener en
cuenta las correcciones radiativas.
1.17. FENOMENOLOGIA ELECTRODEBIL
117
El ME tambien tiene una prediccion concreta para la amplitud del Z a neutrinos, inv :
P
(Z l l )
inv
= l
l
(Z e+ e )
N (Z e e )
=
(Z e+ e )
N (|ve |2 + |ae |2 )
|ve |2 + |ae |2
2N
=
(1 + 4 sin2 W )2 + 1
(
5.865 N = 3
,
(1.587)
7.819 N = 4
mientras que el valor medido experimentalmente para esta cantidad 5.942(16) [?], es una evidencia
muy fuerte de que solo exiten tres neutrinos livianos.
1.17.1
Decaimientos d
ebiles mediados por corrientes cargadas
(1.588)
Esto da lugar a los posibles diagramas para decaimientos de leptones a bosones virtuales, y bosones
a leptontes mostrados en la figura 1.11. Las flechas representan el flujo de n
umero leptonico. La
flecha 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 e , mostrado
en la figura 1.12 El propagador para el boson W de momentum q resulta ser
q
q
1
e =
g 2
.
(1.589)
D
q 2 m2W
mW
Para los propositos actuales la obtencion de este resultado no es necesaria, el punto importante es
que cuando los momentum de las partculas iniciales y finales son mucho mas peque
nas que mW ,
esto se reduce a
e = g .
D
(1.590)
m2W
118
1.17. FENOMENOLOGIA ELECTRODEBIL
119
Este resultado se entiende facilmente 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 W0 , tiene el signo opuesto.
El Lagrangiano efectivo para el decaimiento del muon, e e es entonces
g
g2
[
(1 5 )] 2 [
e (1 5 )e ]
8
mW
2
g
= 2 [
(1 5 )] [
e (1 5 )e ]
8mW
GF
= [
(1 5 )] [
e (1 5 )e ] ,
2
L=
(1.591)
donde
g2
GF
= 2
2 8mW
g24
= 2 2
8g v
1
= 2,
2v
(1.592)
y
v=
1/2
2GF
.
(1.593)
(1.594)
con GF dado en la ec. (1.575) y G = 1.10 105 GeV2 . La corriente hadronica tiene la forma
V1.26A. El factor 1.26 puede entenderse como debido a las correcciones a nivel hadronico de una
corriente que es de la forma VA a nivel del quarks, como en la ec. (1.574). A nivel de quarks el
decaimiento del neutron (udd) al proton (uud) corresponde al decaimiento de uno de los quarks down
del neutron d ue e
GF
L = V11 [
u (1 5 )d] [
e (1 5 )e ] .
2
(1.595)
120
0.97419 0.2257
0.0359
(1.596)
V 0.2256 0.97334 0.0415 1
0.00874 0.0407 0.999133
1.18
C
alculo de procesos
Chapter 2
Computational QFT
There are several tools which allows for the generation of models of particle physics models like
LanHEP [19]
http://theory.sinp.msu.ru/~semenov/lanhep.html,
or FeynRules [20]
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
121
122
The input of LanHEP are files were the Lagrangian of some model is written in a symbolic way.
Then, the LanHEP executable process the input files and generates four outputfiles which are the
input for the CalcHEP program. For example, in the LanHEP dir
$ ./lhep stand.mdl
Here ./lhep command, search for the file in the defaul directory mdl/stand.mdl. If there are no erros
printed, for files 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
To use CalcHEP you must first create a directory with the required files. This is achieved with
the CalcHEP command
1
2.3. LANHEP/CALCHEP
123
$ ./mkUsrDir YourModel
A directory YourModel is created with several files and directories inside. By default, a models
directory is created with two set of .mdl files, corresponding to two versions of the Standard Model:
$ ls YourModel/models/
func1.mdl lgrng1.mdl prtcls1.mdl vars1.mdl
func2.mdl lgrng2.mdl prtcls2.mdl vars2.mdl
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 files in the mdl directory of LanHEP must be modified in order to generate the
proper CalcHEP input files. 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
then copy the *1.mdl files to the sm/models, and from the sm CalcHEP directory run ./calchep.
In order to understand the structure of the LanHEP files consider the following skeleton:
124
1
2
3
4
5
use file_tex.
6
7
prtcprop pdg.
8
9
10
11
12
13
14
15
16
17
18
19
20
particle_type
particle/Antiparticle: (name, property name=VALUE, ...).
21
22
23
24
25
26
27
In line 1, N is an integer that will identify the four output files. The file in line 5 will contain the
LATEX definitions of the used particles. In lines 7-15, the format of the table prtclN.mdl, as required
by CalcHEP, is defined: 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 final commands in 26 and 27 is to check the consistency of the defined model. As a
simple illustration consider the simple case of QED:
model QED: e, mu tau/1.
use qed3g_tex.
prtcprop pdg.
2.3. LANHEP/CALCHEP
%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.
125
126
ea eb A
e ab
P
A
ea
a
a
aP
A
ea
a
a
e ab
e ab
prtcls1.tex:
Name
Spin EM charge
photon
1
0
electron
1/2
1
muon
1/2
1
tau-lepton 1/2
1
Color Comment
1
gauge
1
1
1
vars1.tex
Parameter
EE
Mm
Mt
Value
0.31333
0.1057
1.777
Comment
Electromagnetic coupling constant (1/128)
mass of muon
mass of tau-lepton
127
2.3. LANHEP/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
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
128
129
2.3. LANHEP/CALCHEP
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 + ) = 5 104 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:
LY D0 MD PR D0 H + h.c ,
(2.2)
(2.3)
MDdia = V MD V
(2.4)
where
Investing the equation 2.4 and replacing in (2.2) we can write in the interaction eigenstates:
LY (D0 V )(MDdia V )PR D0 H + h.c
Expanding we get . . . which is just the expresion in the Standard Model file
(2.5)
130
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 field. 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
Schrodinger 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 defined probabilities
for the states of the system. In section xx we start by building the Fock states associated to a massless
not relativistic scalar field. Then we generalize the results to a massive scalar field satisfying the
Klein-Gordon equation.
Some parts of the discussion were based in some topics of chapters 4-6 of [2].
In general, the formalism of second quantization is usefull to describe the states of an undetermined number of particles and interactions which do not conserve particle number. In addition
to high-energy physics where any number of particles may be created or annihilated during a collision process, in statistitical physics it becomes useful to describe a macroscopic body using the
grand-canonical statistical ensemble, in which the number of particles is allowed to fluctuate. In
condensed-matter states the interactions may modify also the number of various excitation quanta,
such as phonons. A more general formalism to discuss this systems is developed in Appendix ??.
131
132
3.1
3.1.1
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.
b
H|Statei
= E|Statei
(3.1)
One step further is to consider the wave function as the eigenstate of the operatorfield acting on
certain Fock states
b
|Fock
Statei = ||Fock Statei ,
(3.2)
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
b
= hFock State||Fock
Statei
(3.3)
This happens when the number of quanta of the Fock state is big enough. In fact, a state with a
definite number of quanta has a infinity uncertainty [16].
Eq. (3.2) is the basis for the calculation of cross section and decay widths in quantum field theory.
Now we will study how to define a such Fock state for a scalar field.
We have already see in Chapter 1 of [1] that a string have a collective wave motion that is
described by a continuous field, which satisfies the familiar one-dimensional wave equation
1 2 2
2 =0
v 2 t2
z
(3.4)
This equation can be derived following two different paths. The first 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
L
= 0.
(3.5)
( )
In the first 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
133
macro
L = Nl
= m/l
T = kl
(z, t)
Note that only in the case v = c this Lagrangian can be written in a covariant form. Moreover, the
scalar field (z, t) have nothing to do with the individual oscillators. An specific solution for (z, t)
would represent one specific oscillation mode of the string. It turn out that this specific 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
X
v
(3.8)
p
n = +|v| |kn |
(3.9)
2n
L
(3.10)
Note that
kn = kn .
(3.11)
134
Therefore
n = n .
(3.12)
(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 = 12 12 m2 2
(3.14)
m2 = 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) =
1
[E]1/2 [E]1
(3.16)
[a]
=E 1/2 [a]
(3.17)
[a] = [E]1/2
(3.18)
we define
1
n (z, t) = ei(n tkn z)
L
y las funciones n satisfacen las siguientes condiciones de normalizacion
Z L
Z
1 L
(3.19)
(3.20)
135
When n = m
L
dz
1
=
L
=1
dz
0
(3.21)
ei(n m )t
1
ei2(nm) 1
L
i(kn km )
=0
=
(3.22)
In this way
Z
(3.23)
(3.24)
Moreover
En tal caso de
Z
H=
0
H dz .
(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 define the Legendre
transformation
H = pq L ,
(3.26)
the Hamiltonian density can be obtained from the Lagragian density trough the Legendre transformation
L
L
(x)
L.
= (x)
t
H = T00 =
where
(x) =
L
((x)/t)
(3.27)
(3.28)
(3.29)
136
is the canonical conjugate variable (conjugate momentum) of the canonical variable (x).
We have then,
Z L
Z
1 L
1
dz
+
dz
H= 2
2v 0
t t
2 0
z z
X
=
n an an
(3.30)
n=
X
ivn
X
1
n m
=
[an n (z, t) an n (z, t)] [am m (z, t) am m (z, t)]
2
v t t n,m= 2 n m
n m
[an am n m an am n m an am n m + an am n m ]
2 n m
n,m=
(3.31)
(3.32)
(3.33)
X
v
=
[ikn an n (z, t) ikn an n (z, t)] ,
z n= 2 n
X
ivkn
=
[an n (z, t) an n (z, t)] ,
2
n
n=
X
v 2 kn km
=
[an n (z, t) an n (z, t)] [am m (z, t) am m (z, t)]
z z n,m= 2 n m
v 2 kn km
[an am n m an am n m an am n m + an am n m ]
n
m
n,m=
(3.34)
(3.35)
(3.36)
137
1 X
n m
H=
dz
[an am n m an am n m an am n m + an am n m ]
2 n,m= 0
2 n m
Z L
1 X
v 2 kn km
dz
[an am n m an am n m an am n m + an am n m ]
+
2 n,m= 0
2 n m
=
1 X n m
an am n,m e2in t an am n,m an am n,m + an am n,m e2in t
2 n,m= 2 n m
1 X v 2 kn km
an am n,m e2in t an am n,m an am n,m + an am n,m e2in t
2 n,m= 2 n m
n n
n n 2in t
1 X n n
2in t
an an e
+
(a an + an an )
a a e
=
2 n= 2 n n
2 n n n
2 n n n n
1 X v 2 kn kn
kn kn
kn kn 2in t
2in t
+
an an e
+
(a an + an an )
a a e
2 n= 2 n n
2 n n n
2 n n n n
(3.37)
Since n = n and kn = kn
1 X 1
H=
(n2 + v 2 kn2 )an an e2in t + (n2 + v 2 kn2 )(an an + an an )
2 n= 2n
+ (n2 + v 2 kn2 )an an e2in t
(3.38)
1 X
H=
n (an an + an an )
2 n=
(3.39)
X
n=
n an an
X
n=
n |an |2
(3.40)
In this way, the factor 2n in eq. (3.16), is a convenient choice of normalization for the coefficients
an 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.
138
3.1.2
q +
(3.41)
with
2 =
k
m
(3.42)
(3.43)
(3.45)
Comparing eq. (3.45) with Eq. (3.40) we see that the complex number an can be written as (~ = 1)
an =c1 qn + ic2 pn
(3.46)
139
mn
mn
c1 =
=
2mn
2
(3.47)
c2 =
1
2mn
(3.48)
mn qn + i pn
2mn
mn qn i pn
an =
2mn
an =
(3.49)
mn
qn = 2mn qn
an + an =2
2mn
2i
i 2mn
an an =
pn =
pn
mn
2mn
1
qn =
(an + an )
2mn
imn
(an an )
pn =
2mn
(3.50)
In quantum mechanics the classical objects qn and pn are promoted to operators which satisfy the
commutation relation
[b
qn , pbm ] =imn
[b
qn , qb m ] = [pbn , pb m ] =0 .
(3.51)
(3.52)
(3.53)
h
i
v
b
an ei(n tkn z) + ab n ei(n tkn z) ,
2n L
n=
(3.54)
140
In terms of operators b
an and ab n the Hamiltonian from eq. (3.39) can be written as
X
b =1
an + b
an ab n )
H
n (b
a nb
2 n=
i
h
1 X
=
an + b
an , ab n
n 2b
a nb
2 n=
X
1
an +
=
n ab nb
2
n=
(3.55)
Since
X
n=
1
,
2
(3.56)
X
1
1
:H: =
n ab nb
an +
2
2
n=
n=
n ab nb
an
(3.57)
n=
This procedure is consistent since the related physics quantities arise from energy differences, no
from absolute energy determinations.
b b
H,
am
1
,b
am
=
n ab nb
an +
2
n=
h
i
X
=
n ab nb
an , b
am
(3.58)
n=
(3.59)
141
we have
h
i
i
h
X
b
an , b
am ]
am b
an + ab n [b
H, b
am =
n ab n , b
n=
n nmb
an
n=
= m b
am
(3.60)
h
i
h
i
i
X
b
n ab n , ab m b
an + ab n b
an , ab m
H, ab m =
n=
n ab n nm
n=
=m ab n
(3.61)
b with eigenvalue En
If |mn i is an eigenstate of H
b n i = En |mn i
H|m
(3.62)
then
bb
b nb
H
an |mn i = b
an H
an |mn i
= (En n ) b
an |mn i
(3.63)
b
an |mn i is also an eigenstate with eigenvalue En n . Moreover,
b
H ab n |mn i = ab n H + n ab n |mn i
= (En + n ) ab n |mn i
(3.64)
142
h0|b
a n =0 ,
(3.65)
such that
h0|0i = 1
(3.66)
(3.67)
We define the state whose energy is larger tha the energy of |0i by one quantum ~n by
|1n i b
a n |0i
h1n | =h0|b
an
(3.68)
(3.69)
where we have made explicit that we have a quantum of energy ~. The normalized state is
h1n |1n i =h0|b
an ab n |0i
h
i
=h0| b
an , ab n |0i
=h0|0i
=1 .
(3.70)
143
(3.71)
with normalization
1
an b
an ab n ab n |0i
h2n |2n i = h0|b
2
1
= h1n |b
an ab n |1n i
2
h
i
1
= h1n | b
an , ab n + ab nb
an |1n i
2
1
= (h1n |1n i + h0|0i)
2
=1 .
(3.72)
By induction we get
1 m
ab n |0i =|mn i
m!
(3.73)
p
ab n ab n
|0i =|mn i
m!
(m 1)!
s
(m 1)!
ab |(m 1)n i =|mn i
m(m 1)! n
r
1
ab |(m 1)n i =|mn i
m n
(3.74)
144
or,
hmn |b
a n = m + 1h(m + 1)n |
(3.75)
From this expressions we can check that number operator can be defined from:
hmn |b
a n ab n |mn i =(m + 1)h(m + 1)n |(m + 1)n i
(3.76)
(3.77)
(3.78)
bn |mn i = mn |mn i
N
(3.79)
and
(3.80)
bnb
N
an |mn i =(mn 1)b
an |mn i
(3.81)
(3.82)
(3.83)
145
such that
b
an |mn i =
mn |mn 1i
hmn |mn i = 1
b: =
:H
X
n=
bn
n N
(3.84)
(3.85)
(3.86)
(3.87)
Nn |ml i = 0
n 6= l ,
b : |mn i = mn n = mn ~n .
hmn | : H
(3.88)
(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
b : |0i = 0 .
h0| : H
(3.90)
The general procedure to redefine the zero of energy such that the vacuum energy vanishes is called
normal ordering. We define a normal-ordered product by moving all annihilation operators to the
b its normal-ordered product will be denoted as
right of all creation operators. For an operator X,
b : . Using this algorithm on the expression of eq. (3.55), we find that
:X
1 X
b
n : (b
a nb
an + b
an ab n ) :
: H: =
2 n=
=
=
1 X
n (b
a nb
an + ab nb
an )
2 n=
X
n=
n ab nb
an
(3.91)
146
3.1.3
Taking into account that En = ~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)
(3.93)
is
1
ipn x
ipn x
a
e
+
a
e
,
n
n
2En L3
n
X
1
p
=
a(nx ,ny ,nz ) exp{i[E(nx ,ny ,nz ) t px x py y pz z]}
2E(nx ,ny ,nz ) L3
nx ,ny ,nz
o
+a(nx ,ny ,nz ) exp{i[E(nx ,ny ,nz ) t px x py y pz z]} ,
(t, x) =
(3.94)
where in natural units the wave number can be identified with the momentum, p = k. In eq. (3.94)
En =p0n
pi =
2
ni
L
(3.95)
147
(3.96)
(3.97)
(3.98)
this solution must have units of energy in natural units. To obtain the dimensions of an , we just
check the dimensions in both sides of eq. (3.94)
[E] = p
[E][E]3
=[E][an ] ,
[an ]
3.2
(3.99)
(3.100)
It is convenient to put the system into a box of size L, so that the total volume is finite. 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 infinite volume.
The Klein-Gordon equation for a real scalar field (Chapter 3. [1])
( + m2 ) = 0 ,
(3.101)
L = 21 12 m2 2 ,
(3.102)
148
The solution is the same that for the case m = 0 in eq. (3.94), but the new dispersion relation is
En2 = p2n + m2 .
(3.103)
L
2
3 Z
d3 p
(3.105)
d3 p (3) (p q) =1
(3.106)
From
Z
and taking into account that
X
n,m = n,m =1 ,
(3.107)
where
pi =
2
ni
L
qi =
2
mi ,
L
(3.108)
we have
Z
d3 p (3) (p q) =
X 2 3
L
2
L
3
(3) (p q) =
nm
n,m
(3) (p q) =n,m .
(3.109)
In this way
(3)
(p q) =
L
2
3
n,m ,
(3.110)
149
(3.111)
V
.
(2)3
Z
L3
ipx
ipx
p
= d3 p
a
+
a
e
pe
p
(2)3 2Ep
Using eq. (3.111), we can write the commutation relations (3.53) in the continuum as
3
2
(3) (p q)
[b
ap , b
aq ] = b
ap , b
aq = 0 .
b
ap , b
aq =
L
(3.112)
(3.113)
(3.114)
(3.115)
(3.116)
Note that again ap is dimensionless. It is customary to write the general solution (3.115) with
(3.117)
a0p = L3 ap .
Then
Z
(t, x) =
d3 p
1
0 ipx
0 ipx
p
a
e
+
a
e
.
p
p
(2)3 2Ep
(3.118)
150
0 0
b
ap , b
aq = (2)3 (3) (p q)
0 0 0 0
b
ap , b
aq = b
ap , b
aq = 0 .
(3.119)
(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 field to a hermitian operator means to promote ap to an operator; thus
Z
b x) =
(t,
d3 p
1
ipx
ipx
p
b
a
e
+
a
b
e
p
p
(2)3 2Ep
(3.121)
with
h
i
b
ap , ab q = (2)3 (3) (p q)
h
i
[b
ap , b
aq ] = ab p , ab q = 0 .
(3.122)
The conjugate momentum can be obtained from the Klein-Gordon Lagrangian in eq. (3.102), by
using eq. (3.29)
b
(x)
=
1
b2
( )
2 0
b
(0 )
=0 b
Z
1
ipx
ipx
p
e
= d3 p
iE
a
+
iE
a
pb
pe
pb
p
(2)3 2Ep
r
Z
i
Ep
3
ipx
ipx
b
a
e
+
b
a
e
= dp
p
p
(2)3
2
(3.123)
151
i
2(2)6
i
Ep0 h ipx
0
0
b
ap e
+b
ap eipx , b
ap0 eip y + b
ap0 eip y
Ep
i
d3 p0
2(2)6
Ep0
Ep
Z
h
i Z
3
b
b
(t, x), (t, y) = d p d3 p0
Z
=
d3 p
nh
i h
io
0
0
b
ap eipx + b
ap eipx , b
ap0 eip y + b
ap eipx + b
ap eipx , b
ap0 eip y
s
Z
Z
i
Ep0 nh ipx
i
ip0 y
0
b
a
e
,
b
a
e
= d3 p d3 p0
p
p
2(2)6 Ep
h
i h
i h
io
ipx
ip0 y
ipx ip0 y
ipx ip0 y
0
+ b
ap e , b
ap e
+ b
ap e
,b
ap 0 e
+ b
ap e , b
ap0 e
s
Z
Z
Ep0 n i(px+p0 y)
i
3
e
[b
ap , b
ap0 ]
= d p d3 p0
2(2)6 Ep
io
i
h
h
0
0
0
ei(pxp y) b
ap , b
.
ap0 + ei(pxp y) b
ap , b
ap0
ap , b
ap0 + ei(px+p y) b
(3.124)
i
2(2)6
i
o
Ep0 n i(pxp0 y) h
0
e
b
ap , b
ap0 ei(pxp y) b
ap , b
ap0
Ep
i
d3 p0
2(2)3
i
Ep0 h i(pxp0 y) (3)
0
e
(p p0 ) + ei(pxp y) (3) (p0 p)
Ep
i
d3 p0
2(2)3
h
i
Ep0 (3)
0
0
(p p0 ) ei(pxp y) + ei(pxp y) .
Ep
Z
i Z
3
b
b
(t, x), (t, y) = d p d3 p0
Z
=
Z
=
d3 p
d3 p
(3.125)
152
(3.126)
(3.127)
(3.128)
The same expression is obtained for the original field 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 field 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.
Z
1
b
d3 p Ep ab pb
ap + b
ap ab p
(3.129)
H=
2
h
i
b
H, b
ap = Epb
ap
h
i
b ab p = +Ep ab p
H,
(3.130)
The analogy between the simple harmonic oscillator and the field is now complete. Therefore ab p
b while b
creates the quanta of momentum p of the field ,
ap is the annihilation operator for a field
quantum with momentum p. From [4]:
What was the positive energy component of the classical field now annihilates the quantum, and the negative energy component now creates the quantum. This quantum is
what we call particle of positive energy.
153
3.3
Fock space
Given the Hilbert space of single-particles H, to construct the space of states of variable particle
number, consider the collection of all possible spaces of n identical particles for either bosons ( = 1
or fermions = 1, Hn . In particular the one-dimensional H0 space is defined by
H0 = {|0i; C}
(3.131)
where |0i is called the vacuum state. A state in which the number of particles is not fixed, e.g n ,
is given by the sequences (|0i = |(0)i)
|i = {|(n)i}n ,
|i = {|(n)i}n ,
(3.132)
with properties
|i + |i = {|(n)i + |(n)i}n
|i = {|(n)i}n
X
h(n)|(n)i .
h|i =
(3.133)
n=0
The collection of all vector of the form (3.132) which are of finite norm
X
h(n)|(n)i < ,
h|i =
(3.134)
n=0
b
A(n)|(n)i
.
(3.135)
n=0
3.4
We can now construct the Fock space following the standard procedure for the harmonic oscillator:
we interpret b
ap as destruction operators and b
ap as creation operators, and we define a vacuum state
|0i as the state annihilated by all destruction operators, so for all p
b
ap |0i = 0 .
(3.136)
154
We normalize the vacuum with h0|0i = 1. The vacuum is the state which contains no particles and
no antiparticles either,
The normal ordered Hamiltonian is
Z
b : = d3 p Ep ab b
(3.137)
:H
p ap
such that, as in discrete case
b : |0i = 0 .
h0| : H
(3.138)
A possible normalization factor for the Fock one-particle state is (|pi |1p i)
1
|pi = b
ap |0i
V
1
hp| =h0|b
ap
(3.139)
V
This state contains one quantum of the field b with momenta p = (Ep , p). Such states have positive
norm, since
1
ap b
ap0 |0i
hp|p0 i = h0|b
V
1
= h0|b
ap b
ap0 b
ap0 b
ap |0i
V
1
= h0|[b
ap , b
ap0 ]|0i
V
(2)3 (3)
=
(p p0 )
V
3
2
=
(3) (p p0 )
(3.140)
L
With this normalization, the limit to discrete case is straightforward:
h1n |1m i = n,m
(3.141)
155
hn,m
i
b
b
(x),
(y)
= i (3) (x y)
[b
an , ab m ] = n,m
|1n i = ab n |0i
Continuum
R 3
L 3
dp
2
3
2
(3) (p q)
hL
i
b
b
(x),
(y)
= i (3) (x y)
3 (3)
[b
ap , ab q ] = 2
(p q)
L
|pi = b
ap |0i
hp|qi =
2 3
L
(3) (p q)
Continuum ab 0p
R 3
L 3
dp
2
3
2
(3) (p q)
hL
i
b
b
(x),
(y)
= i (3) (x y)
[b
ap , ab q ] = (2)3 (3) (p q)
b
ap |0i
3 (3)
(p q)
hp|qi = 2
L
|pi =
1
V
Table 3.2: From discret to continuos, where pi = 2ni /L, and qi = 2mi /L ,
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)
1
1 m
|mp i = m/2
ab p |0i
(3.143)
V
m!
From[4]
The vacuum, together with single particles states (3.139) and all multiparticle states
(3.142), (3.143), constitute a vector space which is calles the Fock space. The creation
and annihilation operators act on this space.
It is convinient to define:
b
(x)
= b+ (x) + b (x)
(3.144)
where
1
p
b
ap eipx
2Ep
Z
1
p
b (x) = d3 p
b
ap eipx .
3
(2) 2Ep
Z
b+ (x) =
d3 p
(2)3
(3.145)
The effect of the operator field, b (x), on the one particle state |pi
(x)|pi
will be important for the evaluation of Smatrix elements in Chapter 7.
(3.146)
156
Another choice of normalization is the Lorentz invariant one, to be used later. In this case, the
Fock state of N particles with all different momenta p1 , p2 , . . . , pN , is obtained acting on the vacuum
with the creation operators,
|p1 , . . . , pn i (2Ep1 )1/2 . . . (2Epn )1/2 b
ap1 . . . b
apn |0i .
(3.147)
The factors (2Ep1 )1/2 are a convenient choice of normalization. In particular, the one-particle states
are
|pi = (2Ep )1/2 b
ap |0i .
(3.148)
(3.149)
The factors (2Ep )1/2 in eq. (3.148) have been chosen so that in the above product the combination
Ep (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 affected we must consider only Ep (pz kz ).
Use the form of the Lorentz transformation of Ep , pz , together with the property of the Dirac delta
(f (x)) = (x x0 )/f 0 (x0 ) [2].
Using (3.111) we have in a finite box
3.5
(3.150)
b
(x)
= b+ (x) + b (x)
(3.151)
Propagator
With conventions
where
1
b
ap eipx
d3 p p
3
(2) 2Ep
Z
1
b (x) = d3 p p
b
ap eipx .
3
(2) 2Ep
Z
b+ (x) =
(3.152)
157
3.6
Quantization of Fermions
(3.153)
(3.154)
= 0
(3.155)
{ , } = 2g 1
(3.156)
where
See [1]. If we assume a plane wave solution like the wave function of the Scrodinger equation
eiEt , after sustition in eq. (3.153), we have
i 0 (iE) m =0
0 E m =0
From the Dirac matrices properties we have
0 = 0
2
(3.157)
Tr 0 =0 .
=1
(3.158)
Moreover, we know that if satisfy the Dirac algebra, the matrices obtained after the unitary
transformation
e =U U
s.t U =U 1
(3.159)
(3.160)
158
In this way we can always choose U such that 0 be diagonal. Because the restrictions in eq. (3.158)
this implies that in this representation we have
1 0
0
=
(3.161)
0 1
where the 1 and 0 are the 2 2 identity and null matrix respectively. Replacing back in eq. (3.157)
we have
Em
0
=0
0
E m
E = m.
(3.162)
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
u1 (p)eipx
u2 (p)eipx
(x)
(3.163)
v1 (p)eipx = + (x) + (x) ,
v2 (p)eipx
where
+ (x) us (p)ei(Etpx)
(x) vs (p)ei(Etpx)
(3.164)
with
u1 (p)
u2 (p)
us (p) =
0
0
0
0
vs (p) =
v1 (p)
v2 (p)
(3.165)
(3.166)
159
(3.167)
us (p)( p m) =0
(3.168)
vs (p)(p + m) =0
(3.169)
us (0) 0 ( 0 p0 + i pi m) =0
us (0) 0 ( 0 p0 m) =0
us (0)(E 0 m) =0
therefore
0 us (0) = + us (0)
0 vs (0) = vs (0)
(3.170)
From [5]
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. (3.156) 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 two eigenvectors with eigenvalue +1 and two with
1
160
Eq. (3.170) 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
us (0)vs0 (0) = 0
(3.171)
since the belong to differente eigenvalues. Note that the two us (0) and the two vs (0) are degenerate.
We define
us (0) s
vs (0) s
(3.172)
where the munis sign in s is just a convention. We define the normalized eigenvectors and such
that
s s0 =ss0
s s0 =ss0
s s0 = 0
1/2 =
0
0
1/2
0
1
=
0
0
1/2
(3.173)
0
0
=
1
0
1/2
0
0
=
0
1
(3.174)
To obtain the spinors for any value of p we know that they must satisfy eqs. (3.166), (3.167),
and, reduce to eq. (3.172) when p 0. The result is
us (p) =Np (p + m) s
vs (p) =Np (p + m) s
(3.175)
Choosing
1
Np =
E+m
(3.176)
(3.177)
(3.178)
we obtain
161
(3.179)
(3.180)
The spinors also satisfy some completeness relations (For details see [5])
X
us (p)
us (p) = p + m
(3.181)
vs (p)
vs (p) = p m
(3.182)
X
s
p + m
,
2m
with properties
[ (p)]2 = (p) .
+ = + =0
+ + =1 .
(3.183)
162
We define
= 23 31 12
p
p =
|p|
1 p
,
(p) =
2
where
[ (p)]2 = (p) .
Moreover
[ , ] =0 ,
and
p us (p) =sus (p)
p vs (p) = svs (p) .
We define also the chiral operator
L
(L )2
L + L
L L
1 5
=
2
=L
=1
=0 .
p s
m
n =
s+
n0 p
.
E+m
163
As with the scalar field, we write the Dirac field as an integral over momentum space of the plane
wave solutions, with creation and annihilation operators as coefficients,
Z
X
1
ipx
ipx
p
(3.184)
(p)v
(x) = d3 p
a
+
b
s (p)e
s (p)us (p)e
s
(2)3 2Ep s=1,2
Z
d3 xH
Z
L
3
= dx
L
Z
= d3 x i 0 0 i 0 0 i i i + m
Z
= d3 x i i i + m
Z
= d3 x (i + m) .
H=
(3.185)
Since that
X
1
p
as (p) (i + m) us (p)eipx
3
(2) 2Ep s=1,2
+bs (p) (i + m) vs (p)eipx
Z
X
1
p
as (p) ( p + m) us (p)eipx
= c d3 p
3
(2) 2Ep s=1,2
+bs (p) ( p + m) vs (p)eipx .
(i + m) =
d3 p
(3.186)
(3.187)
164
0 Ep
i i
p m us (p) =0
!
i i
p + m vs (p) =0 ,
(3.188)
( m) us (p) = 0 Ep us (p)
( + m) vs (p) = 0 Ep vs (p) ,
(3.189)
( + m) us (p) =0 Ep us (p)
( m) vs (p) =0 Ep vs (p) .
(3.190)
(i + m) =
d3 p
X
1
ipx
ipx
p
a
b
(p)
E
v
(p)e
. (3.191)
s (p)0 Ep us (p)e
0
p
s
s
(2)3 2Ep s=1,2
165
Therefore
Z
d3 x (i + m)
Z
Z
i
X h
1
0
0 ip0 x
0
0 ip0 x
3
3 0
p
as0 (p )us0 (p )e
+ bs0 (p )vs0 (p )e
0 (i + m)
= dx dp
3
(2) 2Ep0 s0 =1,2
Z
Z
i
X h
1
0
0 ip0 x
0
0 ip0 x
0 (p )v 0 (p )e
p
= d3 x d3 p0
a
(p
)u
(p
)e
+
b
0
0
0
s
s
s
(2)3 2Ep0 s0 =1,2 s
Z
X
1
ipx
ipx
p
(p)
a
b
d3 p
0 Ep vs (p)e
s (p)0 Ep us (p)e
s
(2)3 2Ep s=1,2
s
Z
Z
Z
i
d3 x
Ep X h 0 0 ip0 x
3 0
3
0
0 ip0 x
0 (p )v 0 (p )e
=
d
p
d
p
a
(p
)u
(p
)e
+
b
0
0
s
s
s
2(2)6
Ep0 s,s0 =1,2 s
as (p)us (p)eipx bs (p)vs (p)eipx
s
Z
Z
Z
Ep X
d3 x
3 0
3
=
d
p
d
p
2(2)6
Ep0 s,s0 =1,2
h
i
0
0
as0 (p0 )as (p)us0 (p0 )us (p)ei(p p)x bs0 (p0 )bs (p)vs0 (p0 )vs (p)ei(pp )x
s
Z
Z
3
Ep X
d
p
= d3 p0
2(2)3 Ep0 s,s0 =1,2
Z
Z
d3 x i(pp0 )x
d3 x i(p0 p)x
0
0
0
0
e
bs0 (p )bs (p)vs0 (p )vs (p)
e
as0 (p )as (p)us0 (p )us (p)
(2)3
(2)3
s
Z
Z
3
d
p
Ep X
= d3 p0
2(2)3 Ep0 s,s0 =1,2
h
i
as0 (p0 )as (p)us0 (p0 )us (p)ei(Ep0 Ep )t (3) (p p0 ) bs0 (p0 )bs (p)vs0 (p0 )vs (p)ei(Ep Ep0 )t (3) (p p0 )
Z
i
d3 p X h
0
=
(p)a
(p)u
(p)u
(p)
b
(p)b
(p)v
(p)v
(p)
a
0
0
0
s
s
s
s
s
s
s
2(2)3 s,s0 =1,2 s
Z 3
i
Xh
dp
=
as0 (p)as (p) bs0 (p)bs (p) .
(3.192)
(2)3
E p
s=1,2
H=
In order to obtain the quantization relations could see that if commutation relations are used we
166
could get
d3 p X
E
a
(p)a
(p)
b
(p)b
(p)
p
s
s
s
s
(2)3 s=1,2
Z
:H: =
(3.193)
The minus sign arise from the anticommutation relations, so that a real spinor field, where bs (p) =
as (p) is automatically zero. Even after normal ordering, this Hamiltonian could give to arise negative energy eigenvalues, which is a serious problem. If instead we assume that the creation and
annihilation operators satisfy anticommutation relations
ar (p), as (q) = br (p), bs (q) = (2)3 rs (3) (p q)
(3.194)
With this relations and taking into account
(x) =
L
= i 0 = i
(0 )
(3.195)
we obtain
{(x, t), (y, t)} = i (3) (x y)
With the anticommutators the normalordered Hamiltonian is
Z
d3 p X
:H: =
E
a
(p)a
(p)
+
b
(p)b
(p)
p
s
s
s
s
(2)3 s=1,2
(3.196)
(3.197)
Moreover
Z
: Q : =q
Z
=q
d3 x : :
d3 p
X
as (p)as (p) bs (p)bs (p)
(3.198)
s=1,2
With this definition as (p) creates particles of charge q, while bs (p) creates antiparticles of charge
q. In a similarly way to eq. (3.144), the most general free particle solution to Dirac equation is
b
(x)
= b+ (x) + b (x)
X
1
p
as (p)us (p)eipx
3
(2) 2Ep s=1,2
Z
X
1
p
b (x) = d3 p
bs (p)vs (p)eipx
(2)3 2Ep s=1,2
Z
b+ (x) =
(3.199)
d3 p
(3.200)
167
3
(2) 2Ep s=1,2
(3.201)
In this way as and as are the creation and annihilation operators for particles, while bs and bs are
the creation and annihilation operators for antiparticles.
It is clear then that the one particle state is
r
1
|e (p, s)i
a (p)|0i
(3.202)
V s
while the one antiparticle state is
r
+
|e (p, s)i
1
b (p)|0i .
V s
(3.203)
168
Chapter 4
Quantization of the electromagnetic field
4.1
Preliminaries
If we impose charge conservation: J = 0, the Proca Equations can be written without lost of
generality as ( 2.4 of [1])
( + m2 )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 p2 = m2 ,
(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 field:
( + m2 ) =
(4.4)
(4.5)
170
(4.6)
where is the charge density of the field which is the source for , and Lint is the interaction
Lagrangian.
In the same section it is shown that this Lagrangian give to arise to the Yukawa interaction
1 emr
.
4 r
(4.7)
F = qE + qv B,
(4.8)
V (r) =
In a similar way, when the Lorentz force
is interpreted in terms of quantum mechanical operators ( 3.3 of [1]) we have the canonical momentum
D = + iqA .
(4.9)
Now, if we force the Scrodinger 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:
D (D )0 = ei(x) D .
(4.10)
(4.11)
where A is a new field that compensates the changes form the derivate. From this it can be shown
that the same identity is valid for all of the powers
n
[(D )0 ] = ei(x) (D )n .
(4.12)
(4.13)
Therefore, the new field tranform like the electromagnetic field, and the modified Lagrangian for the
fields and A
L=
1
1
i
(D) D D0 D0 F F ,
2m
2
4
(4.14)
171
give to arise to the Scrodinger equation in presence of the electromagnetic field plus the Maxwell
with the explicit current
(
q
=0
j = iq
.
(4.15)
[() 2iq A] = i
2m
We now turn to tue quantization of the electromagnetic field
4.2
L
0 = 0.
A
(4.16)
(4.17)
A = 0.
(4.18)
With the Lorentz gauge a new term is added to the Lagrangian which contains the time derivative
of A0 , while in the Coulomb gauge the quantity A0 may be eliminated from the Lagrangian.
In the Coulomb gauge, we have for the A0 component
2 A0 = .
(4.19)
i j
A (r , t), (r, t) = i ij 2 3 (r r0 ) 3T (r r0 ) .
(4.20)
172
These commutation relations between the creation and annihilation involve only the independent
degrees of freedom.
The Hamiltonian obtainded from the Lagrangian is
b=
H
n an, an, ,
(4.21)
kn an, an, ,
(4.22)
n,
X
n,
We now shown that the particles which emerges from the quantization of the electromagnetic field
(the photons) have spin one. To obtain these results, it is necessary to discuss the behavior of these
fields under rotations.
To follow the non-relativistic part of this course we recommend now go directly from section 5.1
to 5.3 where the Smatrix is defined and the probability calculated. In Section 5.5 there is the
general formula for decay. In section 7.1 the perturbative expansion of the Smatrix is presented.
Finally, in section 7.2 an application for the interaction of a non-relativistic atom with radiation, is
given in the context of radiative decay.
The expansion in the Coulomb gauge is
XZ
e (p)
ipx
ipx
3
p
a (p) e
+a (p) e
(4.23)
dp
A=
(2)3 2Ep
(4.24)
Chapter 5
Smatrix
We will use the Smatrix formulation to obtain the decay rates and cross section formulas.
5.1
The Smatrix
(5.1)
|a, ti = HS |a, ti ,
t
(5.2)
is
i
where HS is said to be in the Schrodinger picture where the time dependence is carried out by the
states |a, ti. In this way HS is independent of time. Therefore, the solution to this equation is
|a, ti =eiHS (tti ) |a, ti i ,
(5.3)
since
i
(5.4)
(5.5)
174
CHAPTER 5. SMATRIX
we have that in the Scrodinger picture defined by eq.(5.3), the state of a system evolves with time
|a, ti =U (t, ti )|a, ti i
(5.6)
eiHS (tti ) |ai ,
where |a, ti i, at an initial time ti , is an eigenstate of a set of conmuting operators, and is denoted
simply by |ai. Similarly |bi = |b, tf i at an final time tf .
We have then
hb, tf |a, tf i =hb|a, tf i
(5.7)
is the amplitude for the process in which the initial state |ai evolves into the final state |bi. In the
limit tf ti , the operator eiHS (tf ti ) is called the Smatrix. Therefore S is an operator that
maps an initial state to a final state
|ai S|ai ,
(5.8)
an the scattering amplitudes are given by its matrix elements, hb|S|ai. Observe that
ha| ha|S ,
(5.9)
(5.10)
so that SS = S S = 1.
More rigorously, if ha|ai = 1, and |ni is a complete set of states, the probability that |ai evolves
into |ni, summed over all |ni, must be 1,
X
|hn|S|ai|2 = 1.
(5.11)
n
|hn|S|ai|2 =
X
n
ha|S |nihn|S|ai
!
=ha|S
X
n
|nihn| S|ai
=ha|S S|ai
=1 ,
(5.12)
175
and we conclude that SS = S S = 1. The unitarity of the Smatrix express the conservation of
probability. It is also convenient to define the T matrix, separating the identity operator,
S = 1 + iT
(5.13)
hp1 . . . pn |S|k1 . . . kn i
(5.14)
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 defined the operator
T from S = 1 + iT . We assume that none of the initial momenta pj coincides a final momentum
ki . 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 final momenta coincide, the matrix element of the identity operator between
these states vanishes, and we need actually to compute the matrix element of iT
hp1 . . . pn |iT |k1 . . . kn i
(5.15)
In explicit calculations there will be an overall Dirac delta factor imposing energymomentum conservation. In order not to write explicitly the Dirac delta each time we compute a matrix element of
iT , it is convenient to define a matrix element Mf i from the matrix element,
!
X
X
hp1 . . . pn |iT |k1 . . . kn i = (2)4 (4)
pj
kj iMf i .
(5.16)
j
(5.17)
More generally, the initial and final states are labeled also by the spin states of the initial and final
particles.
So, instead of S or T , the quantity to be calculated is Mf i , but this need first to be relativistically
normalized, in which case it will be denoted as Mf i .
5.2
We first consider a system in a cubic box with spatial volume V = L3 . At the end of the computation
V will be sent to infinity. It is sometimes convenient to put the system into a box of size L, so that the
176
CHAPTER 5. SMATRIX
total volume V = L3 is finite. This procedure regularizes divergences coming from the infinite-volume
limit or, equivalently, from the small momentum region, and is an example of an infrared cutoff. In
a finite box of size L, imposing periodic boundary conditions on the fields, the momenta take the
discrete values p = 2n/L with n = (nx , ny , nz ) 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) = Ceipx ,
(5.18)
and the normalization constant is fixed by the condition that there is one particle in the volume V ,
Z
Z
2
3
1=
d x |p (x)| =
d3 x p (x)p (x)
V
V Z
=|C|2
d3 x
V
=|C|2 V ,
(5.19)
and
1
p (x) = eipx .
V
Wave functions with different momenta are orthogonal, and therefore
Z
d3 x p 1 (x)p2 (x) = p1 ,p2
(5.20)
(5.21)
R
Writing p (x) = hx|pi and using the completeness relation V d3 x|xihx| = 1, we can write this as
Z
NR
hp1 |p2 i =hp1 |
d3 x|xihx|p2 i
V
Z
3
=
d xhp1 |xihx|p2 i
ZV
=
d3 x p 1 (x)p2 (x)
V
=p1 ,p2 .
(5.22)
The superscript NR reminds us that the states have been normalized according to the conventions
of non-relativistic quantum mechanics.
177
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. (3.150)
hp1 |p2 iR = 2Ep1 V p1 ,p2
(5.23)
Therefore the difference between the relativistic and non-relativistic normalization of the one-particle
states is, comparing eqs. (5.22) and (5.23)
|piR = (2Ep V )1/2 |piNR
(5.24)
|p1 , . . . , pn i =
n
Y
#
1/2
(2Ep V )
i=1
|p1 , . . . , pn iNR
(5.25)
We denote by Mf i , defined in eq. (5.16), the scattering amplitude between the initial state with
momenta q1 , . . . , qn and the final state with momenta p1 , . . . , pn , with non-relativistic normalization
of the states, and by Mf i the same matrix element with relativistic normalization of the states. Then
from eq. (5.16)
!
X
X
(2)4 (4)
pi
ki iMf i =hp1 . . . pn |iT |k1 . . . kn iR
i
n
Y
1/2
(2Epi V )
i=1
n
Y
i=1
n
Y
2Ekj V
j=1
1/2
n
Y
X
X
1/2
1/2
(2Epi V )
2Ekj V
(2)4 (4)
pi
ki
j=1
!
iMf i
(5.26)
Therefore
Mf i =
n
Y
(2Epi V )
i=1
1/2
n
Y
2Ekj V
j=1
1/2
Mf i
(5.27)
n
Y
i=1
(2Epi V )
n
Y
1/2
j=1
!
2Ekj V
1/2
(2)4 (4)
X
j
pj
X
j
kj
iMf i . (5.28)
178
5.3
CHAPTER 5. SMATRIX
Process probability
NR
4 (4)
= (2)
X
i
pi
kj
iMf i
(5.29)
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
2
!
X
X
2
hp1 . . . pn |iT |k1 . . . km iNR = (2)4 (4)
pi
kj iMf i
i
(5.30)
and we are confronted with the square of the delta function. To compute it, we recall that we are
working in a finite spatial volume and, from eq. (3.112)
(2)3 (3) (0) = V
(5.31)
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
(5.32)
Then
2
!
X
2
hp1 . . . pn |iT |k1 . . . km iNR = (2)4 (4) p
ki iMf i
j
!
X
=(2)4 (4) p
kj V T Mf i
(5.33)
Moreover we must sum over all final states. In the discrete limit, since we are working in a finite
volume V , the sum over all final states corresponds to the sum over the possible discrete values of
the momenta of the final particles
nj = , . . . , 1, 0, 1, . . .
2nj
kj =
,
(5.34)
nyj = , . . . , 1, 0, 1, . . .
L
z
nj = , . . . , 1, 0, 1, . . .
179
XXX
nyj
nx
j
(5.35)
nzj
In the large-volume limit for each particle we can write, using eq. (3.105)
Z
X
V
d3 kj ,
3
(2)
k
(5.36)
...
k1
X
hp1 . . . pn |iT |k1 . . . km iNR 2
km
!
=
X
k1
X
X
X
(2)4 (4)
...
pi
kj
i
km
V T |Mf i |2
!
V d k1
V d km
...
(2)4 (4)
pi
kj V T |Mf i |2
3
3
(2)
(2)
i
! i
Z
Z
m
X
X
Y
V d3 kj
2
4 (4)
.
= . . . (2)
pi
kj V T |Mf i |
(2)3
i
i
j=1
Z
...
(5.37)
Z
...
!
(2)4 (4)
X
i
Z
=
Z
...
pi
kj
m
Y
V d3 kj
V T |Mf i |2
(2)3
j=1
!
(2)4 (4)
X
i
pi
X
i
kj
V T |Mf i |2
n
Y
i=1
1
2Epi V
n
Y
(2Epi V )
m
Y
1/2
i=1
!2
2Ekj V
1/2
j=1
m
Y
d3 kj
(2)3 2Ekj
j=1
The probability for the process of an initial particle decaying into n final particles is then
!
Z
Z
n
X
1 Y
d3 kj
1 = . . . (2)4 (4) p
kj T |Mf i |2
2Ep j=1 (2)3 2Ekj
i
(5.38)
(5.39)
On the other hand the probability for a process with two initial particles colliding into n final particles
is
!
Z
Z
n
X
1
1 Y
d3 kj
2 = . . . (2)4 (4) p1 + p2
kj V T |Mf i |2
(5.40)
3 2E
2E
V
2E
V
(2)
p
p
k
1
2
j
j=1
i
180
CHAPTER 5. SMATRIX
5.4
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 fill the area A randomly Assuming that an interaction
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
(5.41)
PS = .
A
Now suppose we have a parallel beam with density of particles n and velocity v towards the target.
In time t, this beam fills a volume
V = Avt .
(5.42)
Choosing t such that the volume contains just one particle, we can write
n = 1/V ,
(5.43)
1 = nvtA .
(5.44)
Ps
.
nvt
(5.45)
or
181
E1 E2
(5.49)
In fact, for not relativistic particles, where Ei = mi , this coincides with the usual relative velocity
m1 v1 m2 v2
vrel =
E1
E2
= |v1 v2 | .
(5.50)
The most general formula for the relative velocity is
vrel =
I
E1 E2
(5.51)
where
q
I = (p1 p2 )2 m21 m22
(5.52)
182
CHAPTER 5. SMATRIX
In general
q
I = (E1 E2 p1 p2 )2 m21 m22
q
= E12 E22 + (p1 p2 )2 2E1 E2 p1 .p2 m21 m22
(5.53)
Since
m21 m22 =(E12 p21 )(E22 p22 )
=(E12 E22 p21 E22 E12 p22 + p21 p22 )
I=
q
p21 E22 2E1 E2 p1 p2 + E12 p22 + (p1 p2 )2 p21 p22
(5.54)
(5.55)
If
(p1 p2 )2 p21 p22 = 0
(5.56)
p1
I
p
2
=
=
E1 E2
E1 E2
(5.57)
(5.58)
To simplify the notation we set Ei = Epi =, and Ef = Epf . Moreover, the differential cross
section is
!
2
X
X
Y d3 kf
1
d =(2)4 (4)
pi
pf
|Mf i |2
4vrel E1 E2
(2)3 2Ef
i=1
f
f
= (2)4
1
|Mf i |2 dn (p1 , p2 ; k1 , . . . , kn )
4vrel E1 E2
(5.59)
where
!
d(n) (p1 , p2 ; k1 , k2 , . . . , kn ) = (4) p
X
j
kj
n
Y
d3 kj
.
3 2E
(2)
k
j
j=1
(5.60)
We keep the diferential notation both for d, and d until the last integration have been made.
183
5.4.1
d =
(5.61)
(p
+
p
p
)
1
2
1
2
4(2)6
E10 E20
d3 p0 d3 p0
= (4) (p1 + p2 p01 p02 ) 4(2)6 0 1 0 2
E1 E2
(5.62)
We now will find an expression for cross section in the center of mass frame (CM)
The center of mass (CM) frame is defined by the condition
p1 + p2 = 0
(5.63)
(4) (p + p2 p01 p02 ) = (3) (p1 + p2 p01 p02 )(E1 + E2 E10 E20 )
(5.64)
In the CM frame
(4) (p + p2 p01 p02 ) = (3) (p01 + p02 )(E1 + E2 E10 E20 )
(5.65)
Mf i in integration does not depend on |p01 | or |p02 | as the final momentum is fixed by the initial
momentum whenever the final states have only two particles. In this way the integration on p02 can
be evaluated directly for d(2) . Replacing back in Eq. (5.61)
d3 p0 d3 p0
4(2)6 d(2) = (3) (p01 + p02 )(E1 + E2 E10 E20 ) 0 1 0 2
E1 E2
3 0 Z
dp
d3 p 0
=(E1 + E2 E10 E20 ) 0 1 (3) (p01 + p02 ) 0 2
E1
E2
3 0
dp
=(E1 + E2 E10 E20 ) 0 10
E1 E2
(5.66)
184
CHAPTER 5. SMATRIX
p01 2 d|p01 |d
E10 E20
(5.67)
As
|p01 |
E10 2 m1 2
(5.68)
d|p01 |
2E10
q
=
dE10
2 E10 2 m1 2
=
E10
|p01 |
(5.69)
(5.70)
and
|p01 |E10 dE10
d
E10 E20
|p0 |dE 0
= (E1 + E2 E10 E20 ) 1 0 1 d
E2
(5.71)
(5.72)
(5.73)
we have
2
p01 =p02
2
(5.74)
185
q
E10 2 m01 2 + m02 2
(5.75)
In this way we can express E20 in terms of E10 in Eq. (5.71). Moreover, we can define the center of
mass energy as
s = E1 + E2
(5.76)
Using The energy part of function in Eq. (5.64) can be written as
q
0
02
02
02
s E1 E1 m1 + m2
(5.77)
As established before, Mf i in this case in independent of |p01 |, and the integration on E10 can be done
directly only for d(2) . The integral is easily performed using the identity
(f (z)) =
X (z zn )
n
(5.78)
|f 0 (zn )|
where zn are the zeroes of f (z). In this case, this function is a function of the integration variable
E10 , with only one zero
(f (x)) =
(x x0 )
|f 0 (x0 )|
(5.79)
where
f (x) = s x
q
x2 m01 2 + m02 2
(5.80)
Therefore
6
4(2) d
(2)
(x x0 ) |p01 (x)|
dx
|f 0 (x0 )| E20 (x)
1
|p01 (x0 )|
=d 0
|f (x0 )| E20 (x0 )
Z
=d
E20 (x0 )
q
= x20 m01 2 + m02 2
(5.81)
(5.82)
186
CHAPTER 5. SMATRIX
q
s x0 x20 m01 2 + m02 2 = 0
2
2
s 2 s x0 + x20 = x20 m01 + m02
2
2
s 2 s x0 = m01 + m02
(5.83)
with solution
s + m01 2 m02 2
x0 =
2 s
(5.84)
(5.85)
we have
f 0 (x0 ) =
m01 2 m02 2 + s
r
1
2
m01 2 + m02 2 s
=
1
m01 2 + m02 2 + s
m01 2 + m02 2 s + m01 2 m02 2 s
=
m01 2 + m02 2 + s
2s
,
=
s + m02 2 m01 2
(5.86)
and
(f (E10 ))
(E10
x0 )
s + m02 2 m01 2
2s
(5.87)
Replacing the expression for x0 in (5.84) into Eq. (5.82) we have (See deltaxn.nb for additional
details)
p
[s (m01 m02 )2 ][s (m01 + m02 )2 ]
p01 (x0 ) =
2 s
s m01 2 + m02 2
E20 (x0 ) =
(5.88)
2 s
187
1
q
|f 0 (x0 )|
x20 m01 2 + m02 2
p
[s (m01 m02 )2 ][s (m01 + m02 )2 ]
s m01 2 + m02 2
= d
2s
s m01 2 + m02 2
p
[s (m01 m02 )2 ][s (m01 + m02 )2 ]
= d
2s
4(2)6 d(2) = d
(5.89)
(5.90)
s m02 + m01
2
(5.91)
we have
6
4(2) d
(2)
(5.92)
Moreover
p01
=
2 s
(5.93)
To further evaluate Eq. (5.61), we need to express vrel and E1 E2 in terms of s and the masses.
Concerning vrel , from Eq. (5.58), evaluated in CM frame
p1
p
2
E1 E2 vrel =E1 E2
E1 E2
p1
p1
+
=E1 E2
E1 E2
= |p1 | (E1 + E2 )
= |p1 | s
(5.94)
188
CHAPTER 5. SMATRIX
Replacing back Eqs. (5.89), and (5.94) into Eq. (5.61), we have
d =
1
2
6
(2)
|M
|
4(2)
d
f
i
64 2 E1 E2 vrel
d
1
=
|Mf i |2
2
d
64 E1 E2 vrel
(5.95)
(5.96)
(5.97)
In the CM frame
s =E1 + E2
q
q
2
2
= p1 + m1 + p22 + m22
q
q
2
2
= p1 + m1 + p21 + m22
=2p21
m21
m22
(5.98)
q
+ 2 p41 + (m21 + m22 )p21 + m21 m22
+
+
q
s (2p21 + m21 + m22 ) =2 p41 + (m21 + m22 )p21 + m21 m22
s2 2s(2p21 + m21 + m22 ) + [2p21 + (m21 + m22 )]2
s2 2s(2p21 + m21 + m22 ) + 4p41 + 4p21 (m21 + m22 ) + (m21 + m22 )2
4sp21 + s2 2s(m21 + m22 ) + (m21 + m22 )2
4sp21 + s2 2sm21 2sm22 + m41 + m42 + 2m21 m22
4sp21 + s2 2sm21 2sm22 + m41 + m42 2m21 m22
p21 =
(5.99)
(5.101)
189
|p1 | =
2 s
1/2 (s, m22 , m21 )
=
2 s
(5.102)
1p
[s (m1 + m2 )2 ][s (m1 m2 )2 ]
2
(5.103)
(5.104)
and, finally
1
d
=
d
64 2 s
1/2
|M|2
(5.105)
5.5
(5.106)
Decay Rates
ki
iMf i
(5.107)
where the initial state is a single particle of momentum p and mass M , while the final state is
given by n particles of momenta ki and masses mi , i = 1, . . . , n. We are therefore considering a decay
process.
By using eq. (5.39) we have
! Z
Z
Z
n
X
Y
d3 kj
2 1
4 (4)
p
kj
dt |Mf i |
(5.108)
= . . . (2)
2Ep j=1 (2)3 2Ekj
i
190
CHAPTER 5. SMATRIX
Z
...
4 (4)
(2)
X
i
kj
n
Y
1
d3 kj
2
|Mf i | dt
2Ep
(2)3 2Ekj
j=1
(5.109)
Finally we define the decay rate d as the decay probability in which in the final state the jth
particle has momentum between kj and kj + dkj per unit time
!
n
Y
X
d3 kj
1
d
2
4 (4)
= (2)
p
kj
|Mf i |
d
dt
2Ep
(2)3 2Ekj
j=1
j
(2)4
|Mf i |2 d(n) (p; k1 , k2 , . . . , kn )
=
2Ep
(5.110)
where
!
d(n) (p; k1 , k2 , . . . , kn ) = (4) p
kj
n
Y
d3 kj
.
3 2E
(2)
k
j
j=1
(5.111)
5.5.1
(2)4
|Mf i |2 d(n) (p; k1 , k2 , . . . , kn )
2Ep
(5.112)
We now consider the decay of particle of mass M decaying into two particles of 4momenta p1 , p2
and masses m1 , m2 . In the CM frame the initial momentum satisfy
p =0 M = Ep
(5.113)
Therefore
(2)4
|Mf i |2 4(2)6 d(2) (p; p1 , p2 )
6
2M [4(2) ]
1
= 3
|Mf i |2 4(2)6 d(2) (p; p1 , p2 )
2 M (2)2
1
=
|Mf i |2 4(2)6 d(2) (p; p1 , p2 )
2
32 M
d =
(5.114)
191
5.6. BACKUP
where
4(2)6 d(2) (p; p1 , p2 ) = (4) (p p1 p2 )
d3 p1 d3 p2
E1 E2
(5.115)
(5.116)
and,
d3 p1 d3 p2
4(2) d (p; p1 , p2 ) =(M E1 E2 ) (p1 + p2 )
E1 E2
6
(2)
(3)
Comparing with eq. (5.66) we see that the two quantities are the same after the replacing
p01 p1 and p02 p2 . Therefore we have from eq. (5.92)
4(2)6 d(2) = d
(5.117)
s M,
(5.118)
(5.119)
(5.120)
5.6
Backup
Perturbation theory is developed more easily using the Hamiltonian formalism. We therefore consider
a general field theory with a Hamiltonian
H = H0 + Hint
(5.121)
192
CHAPTER 5. SMATRIX
where H0 is the free Hamiltonian and Hint is the interaction term. The interaction term will be
considered small. For instance in QED
Z
Z
3
Hint = d x Hint = d3 x Lint
(5.122)
with
Lint = eA
(5.123)
The smallnes of the interaction follows from the fact that the parameter which turns out to be
relevatn for the perturbation expansion is = e2 /4 1/137.
SS =(1 + iT )(1 iT )
= 1 + i(T T ) + T T = 1 ,
T T = i(T T ) .
(5.124)
(5.125)
hb|T
X
hb|T |niha|T |ni = i (hb|T |ai ha|T |bi )
n
X
n
Tbn Tan
= i (Tba Tab
).
(5.126)
if a = b
|Tan |2 = i Im Taa .
(5.127)
Chapter 6
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 7 we will obtain the corresponding
Feynman rules from the Smatrix expansion.
6.1
Particle decays
Particle decay [6] is the spontaneous process of one elementary particle transforming into other
elementary particles. During this process, an elementary particle becomes a different particle with
less mass and an intermediate particle such as W boson in muon decay.
For a particle of a mass M , the differential decay width according Eq. (5.114), is
dn =
(2)4
|M|2 d(n) (P ; p1 , p2 , . . . , pn )
2M
(6.1)
n
X
i=1
pi )
n
Y
i=1
d3 pi
(2)3 2Ei
!
.
(6.2)
We will keep the d notation until all the integrals get evaluated.
The two-body decays in eq. (5.119) is
d
1
=
|Mf i |2 1/2 (M 2 , m22 , m21 )
2
3
d 64 M
193
(6.3)
194
6.2
Width decay
(6.4)
For the H f 1 f2 decay. The interaction between the Higgs boson with fermions1 is given by the
Yukawa interaction term [1]
(v + H)
(f R fL + f L fR )
2
Gf H
Gf v
= ff ff
2
2
1/2
ff
= mf f f mf GF 2
LHiggs = Gf
(6.5)
Such as the electro has acquired a mass me = Gf / 2. On the other hand the coupling to be
assigned to the process vertex is Gf 2 or mf /v =.
The decay process H f f , is displayed in Fig. 6.1
The Feynman rules, to be explained in Chapter 7 are indicated in Fig. 6.2.
In this way the scattering amplitude is
1/2
iM = imf GF 2
u(s1 , p1 )v(s2 , p2 ).
(6.6)
where p1 , s, p2 y s2 are the momentum and spines of fermion and antifermion respectively.
For the general case
iM = ihu(s1 , p1 )v(s2 , p2 ).
1
In this case we consider only electrons, by the formula is easy generalizable to other fermions
(6.7)
195
= v (s2 , p2 )( 0 u(s1 , p1 ))
= v (s2 , p2 )( 0 u(s1 , p1 ))
= v(s2 , p2 )u(s1 , p1 ).
Squaring M, and summing over possible polarization states of final particles, we have
X
X
|M|2 = h2
(u(s1 , p1 )v(s2 , p2 ))(v(s2 , p2 )u(s1 , p1 )).
s1 ,s2
s1 ,s2
(6.8)
196
The several sums in Ec. (6.8) can be calculated by expressing the products uv y vu en in terms of
their components, as follow
X
(u(s1 , p1 )v(s2 , p2 ))(v(s2 , p2 )u(s1 , p1 ))
s1 ,s2
s1 ,s2
s1 ,s2
u (s1 , p1 )u (s1 , p1 )
v (s2 , p2 )v (s2 , p2 )
s2
= (p 1 + mf ) (p 2 mf )
= Tr[(p 1 + mf )(p 2 mf )].
(6.9)
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 + m1 )(p 2 m2 )]
= Tr[( p1 + m1 )( p2 m1 )]
= Tr[ p1 p2 m2 p1 + m1 p2 m1 m2 ]
= p1 p2 Tr[ ] 4m1 m2
= 4g p1 p2 4m1 m2
= 4(p1 p2 m1 m2 ).
where m1 , m2 are the final masses, and
X
|M|2 = 4h2 (p1 p2 m1 m2 ).
s1 ,s2
(6.10)
197
M 2 E12 E22
2
(6.11)
p1 p2 m2f = E1 E2 p1 p2 m1 m2
= E1 E2 + p21 m1 m2
M 2 E12 E22
=
+ p21 m1 m2
2
1
=
M 2 m21 p21 m22 p21 + p21 m1 m2
2
1
M 2 m21 m22 2m1 m2
=
2
1 2
=
M (m1 m2 )2
2
(6.12)
s1 ,s2
(M
,
m
,
m
)
M
(m
+
m
)
1
2
2
1
d 32 2 M 3
After the integration
(6.13)
dCM = 4 2 we have
=
h2 1/2 2 2 2 2
2
(M
,
m
,
m
)
M
(m
+
m
)
1
2
2
1
8M 3
(6.14)
For m1 = m2 = mf
1/2
4m2f
(M
=M 1
M2
2
4m2f
2
2
M (m1 + m2 ) =M 1
M2
1/2
R 2
0
R
0
sin d = 4
, m22 , m21 )
(6.15)
198
and therefore
h2
(H f f ) = M
8
4m2f
1
M2
3/2
(6.16)
In the case of the standard model Higgs with mass MH decaying to fermion pair, according to the
Lagrangian in eq. (6.5)
In the limit mf MH
6.3
MH m2f GF
(H f f ) =
4 2
this expression reduces to
3/2
m2f
14 2
,
MH
(6.17)
MH m2f GF
(H f f ) =
.
4 2
(6.18)
e2
v(k2 ) u(k1 ) v(k2 ) u(k1 )
s
(6.19)
e + e +
L=
Chapter 7
Feynman Rules
When the case of interacting fields are considered, the particles can be created, destroyed and scattered. In essence this requires solving the coupled non-linear field equations for given conditions.
This is an extremely difficult 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 7.1 we write the Smatrix expansion derived in Chapter 5, in the interaction picture. In
section 7.3 we show how to use the Wick expansion to calculate Smatrix elements involving scalars
and spinors.
The considerations which allow to write a formal series of the evolution operator are completely
general and apply just as well the perturbative calculation of the scattering operator as to the Green
functions of the many-body problem in Appendix ??.
7.1
Interaction picture
(7.1)
(7.2)
200
(7.3)
Given the state |a, tiS in the SP, in the Heisenberg picture (HP) we defined the state
|aiH = U |a, tiS = |aiS
(7.4)
(7.5)
Hence, the transformation from HP to SP is unitary. At t = ti , states and operators in the two
pictures are the same. We see from Eq. (7.4) that in the HP state vectors are constant in time, while
from Eq. (7.5) the Heisenberg operators evolve with time. Is convenient to keep the temporal label
in the Heisenberg states
|aiH = |a, ti iH
(7.6)
Eq. (7.5) ensures the invariance of matrix elements and commutation relations:
S hb, t| O
OS , P S = c OH (t), P H (t) = c
(7.7)
(7.8)
where c is a constant.
Differentiation of Eq. (7.5)
d
d
d H
O (t) =
U OS U + U OS U
dt
dt
dt
S
S
=iH U O U + U O U (iH)
= i(OH H HOH ) ,
(7.9)
d H
O (t) = OH (t), H
dt
(7.10)
The interaction picture (IP) arises if the Hamiltonian is split into two parts
H = H0 + HI .
(7.11)
201
In quantum field theory HI will describe the interaction between two fields, themselves described by
H0
IP is related to the SP by the unitary transformation
Ui Ui (t, ti ) = eiHi (tti ) ,
(7.12)
(7.13)
OI (t) = U0 OS U0 .
(7.14)
in this way,
and
Thus the relation between IP and SP is similar to that between HP and SP, but with the unitary
transformation U0 involving only the noninteracting Hamiltonian H0 . Note that both the vector
states as the operators in the IP are time-dependent.
Differentiating Eq. (7.14) gives the differential equation of motion operators in the IP:
i
d I
O (t) = OI (t), H0
dt
(7.15)
Substituting Eq. (7.13) into the Scrodinger Eq. (7.1), 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
i
d
|a, tiS =H S |a, tiS
dt
d
(U0 |(t)i) =H S U0 |(t)i
dt
d
d
i
U0 |(t)i + iU0 |(t)i =H S U0 |(t)i
dt
dt
d
U0 H0 |(t)i + iU0 |(t)i =H S U0 |(t)i
dt
d
U0 H0 |(t)i + iU0 |(t)i =(H0 + HIS )U0 |(t)i
dt
d
iU0 |(t)i =HIS U0 |(t)i
dt
d
i |(t)i =U0 HIS U0 |(t)i
dt
i
(7.16)
202
d
|(t)iI = HII |(t)iI ,
dt
(7.17)
(7.18)
is the interaction Hamiltonian in the IP, with HIS and H0S being the interaction and free-field Hamiltonian 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. (7.17) is a Scrodinger-like equation with the time dependent Hamiltonian HI (t). With the
interaction switched off (i.e. we put HI = 0), the state vector is constant in time. The interaction
leads to the state |(t)i changing with time. Given that the system is in a state |ii at an initial time
t = ti , i.e.
|(ti )i = |ii ,
(7.19)
the solution of Eq. (7.17) with this initial condition gives the state |(t)i of the system at any other
time t. It follows from the Hermicity of the operator HI (t) that the time development of the state
|(t)i according to Eq. (7.17) is a unitary transformation. Accordingly it preserves the normalization
of states
h(t)|(t)i = const.
(7.20)
(7.21)
evolves at t = , long after the scattering is over and all particles are for apart again. The Smatrix
relates |()i to () and is defined by
|()i = S|()i = S|ii ,
(7.22)
A collision can lead to many different final states |f i, and all these possibilities are constrained
within |()i.
203
(7.23)
|(t)i
|()i
|(t)i |()i = i
(7.24)
|(t)i = |ii i
(7.25)
In the limit t
|()i = S
(0)
|ii i
(7.26)
where
S (0) = 1 .
(7.27)
t1
(7.28)
This equation then can be solved iteratively. If HI is small we can solve this equation by iteration
|(t)i = |ii + (i)
t
2
t1
dt1
(7.29)
204
In the limit t
Z
(0)
|(t)i = S + (i)
t1
dt1
dt2 HI (t1 )HI (t2 )|(t2 )i
dt1 HI (t1 ) |ii + (i)
Z t1
Z
2
(0)
(1)
dt2 HI (t1 )HI (t2 )|(t2 )i ,
dt1
|ii + (i)
= S +S
(7.30)
where
S
(1)
dt1 HI (t1 ) .
= (i)
(7.31)
Z t3
Z t2
Z t2
2
dt3
dt4 HI (t3 )HI (t4 )|(t4 )i
|ii + (i)
dt3 HI (t3 )|ii + (i)
t
2
t1
Z t
Z t1
Z t2
3
dt1
+ (i)
dt2
dt3 HI (t1 )HI (t2 )H1 (t3 )|ii
Z t
Z t1
Z t2
Z t3
4
+ (i)
dt1
dt2
dt3
dt4 HI (t1 )HI (t2 ) HI (t3 )HI (t4 )|(t4 )i
(7.32)
(7.33)
In the limit t
|(t)i = S (0) + S (1) + S (2) + S (3) |ii
Z
Z t1
Z t2
Z
4
+ (i)
dt1
dt2
dt3
t3
(7.34)
where
S
(2)
=(i)
S (3) =(i)3
t1
dt1
Z t1
dt1
(7.35)
205
S (n)
n=0
=1 +
X
(i)n
n=1
7.2
n!
t1
dt2 . . .
dt1
(7.36)
tn1
Atomic decay
(7.37)
(7.38)
(7.39)
(7.40)
HA =
HEM
HI0 (t)
=
e
2m
1
=
2
Z
p2e
,
2m
re
d3 r : 2 (r, t) : + : B2 (r, t) : ,
e2 2
pe A(re , t) + A(re , t) pe +
A (re , t)
,
2m
(7.41)
(7.42)
(7.43)
206
(7.44)
(7.45)
For one photon decay of an initial atomic state a into a final atomic state b and a photon of
energy n and polarization , the states are
|i =|, 0i = a (re )|0i
|i =|b, 1n i = b (re )|1n i ,
(7.46)
|1n i = an |0i
(7.47)
where
(7.48)
and
(pe A(re , t) + A(re , t) pe ) a (re ) = i (e A(re , t) + A(re , t) e ) a (re )
= i {(e A) a + A e a + A e a }
= i {2A e a } ,
where A = 0 was used.
With all of this we have a simplified formula
Z
ie
1
Sba = i
dthb, 1n |UA A(re , t) e UA |a, 0i
m
Z
ie
1
Sba = i
dthb, 1n |UA A(re , t) e UA |a, 0i
m
Z
ie
iEb t
Sba = i
dthb, 1n |e
A(re , t) e eiEa t |a, 0i
m
Z
Z
e
=
dt ei(Eb Ea )t d3 re h1n |b (re ) (A(re , t) e ) a (r)|0i
m
Z
Z
e
=
dt ei(Eb Ea )t
d3 re {b (re )e a (r)} h1n |A(re , t)|0i.
m
(7.49)
(7.50)
207
n0 0
n0 0
1
0
1
n h1n |1n ieikn0 xe
3
2n L
1
=
n ein tkn re .
2n L3
(7.51)
(7.52)
(7.53)
In most atomic decays, the energy of the emmited photon, which is equal to n = Eb Ea , is much
less than 1/R, where R is the size of the atomic system, and hence the maximum range of the integral
over re . In this case, the dipole approximation
eikn re 1
(7.54)
e 1
n pba ,
m 2n
(7.55)
d3 re b (re )e a (re ).
(7.56)
where
pba = i
(7.57)
208
we have
|Sba (T /2, T /2)|2
T
T
1
1
= lim 2(Eb + n Ea )T 3 |Mba |2
T T
L
1
=2(Eb + n Ea ) 3 |Mba |2
L
Wba = lim
(7.58)
Then
Wba =
e2
2
(Eb + n Ea )
|n pba |2 .
2
V
2n m
(7.59)
Summing over all final photon states to get the total a b decay rate gives
Wba =
X 2
n,
(Eb + n Ea )
X X 2
e2
| pba |2
2n m2 n
(Eb + n Ea )
e2
|n pba |2 .
2
2n m
(7.60)
Using
X
n
(2)3
d3 k
d3 k
e2
Wba =
(Eb + n Ea )
|n pba |2
2
2
(2)
2m
Z
d3 k
e2 X
= 2
(Eb + n Ea )|n pba |2
2m
(2)2
Z
Z
e2 X
d3 k
2
= 2
|k| d|k| d
(Eb + Ea )|n pba |2 .
2m
(2)2
(7.61)
XZ
(7.62)
Wba = 2
d
(Eb + Ea ) d|n pba |2
2m
(2)2
Z
e2 X
d|n pba |2 ,
=
4 2m2
(7.63)
209
Then the total rate for the decay of the state a into b is
e2
4
Wba =
|pba |2 .
4 3m2
7.3
(7.64)
(7.65)
Yukawa interaction
As a concrete example, we take a theory with a fermion field and scalar field, which interact via the
Yukawa interaction [4]:
Lint = h .
(7.66)
Let the quantum of the field be denoted by B, since the particle is a boson. The quanta of the
fermionic field 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+ (p0 ) ,
(7.67)
(7.68)
where the required ordered product will be explained in next section. The term linear in the interaction Hamiltonian in the Smatrix. It is
S (1) =
d4 x : : .
(7.69)
d4 x : ( + + )(+ + )(+ + ) : .
(7.70)
= ih
(1)
= ih
210
hn |+ hn + 1 |
(7.71)
|n i |n + 1 i
hn | hn 1 |
(7.72)
and
and,
hf | =h1, 1 , 0 |
(7.74)
Evaluating the expectation values for each term in eq. (7.73) for these initial and final states, we
have
h1, 1 , 0 |+ + + |0, 0 , 1 i h2, 2 , 0 |0, 0 , 0 i = 0
h1, 1 , 0 |+ + |0, 0 , 1 i h2, 2 , 0 |0, 0 , 2 i = 0
h1, 1 , 0 |+ + |0, 0 , 1 i h2, 0 , 0 |0, 0 , 0 i = 0
h1, 1 , 0 |+ |0, 0 , 1 i h2, 0 , 0 |0, 0 , 2 i = 0
h1, 1 , 0 | + + |0, 0 , 1 i h0, 2 , 0 |0, 0 , 0 i = 0
h1, 1 , 0 | + |0, 0 , 1 i h0, 2 , 0 |0, 0 , 2 i = 0
h1, 1 , 0 | + |0, 0 , 1 i h0, 0 , 0 |0, 0 , 0 i =
6 0
211
: := + + .
(7.75)
In general, we have the following result: The ordered product of a set of operators is such that all
the creation operators are on the left and destruction operators are in the right while keeping the
original order consistent with the initial and final states.
h1, 1, , 0, 0| : 1 2 n1 n : |0, 0, , 1, 1i h1, 1, , 0, 0|a1 a2 an1 an |0, 0, , 1, 1i (7.76)
The only term that contributes to the matrix element of these process is therefore
ih
d4 x + .
(7.77)
In the language of second quantization it is said that at the interaction point in Fig. 7.1, the
scalar is destroyed by + and one electron ( ) and a positron ( ) are created.
Let us define the oneparticle states as in eq. (3.139)
r
|B(p)i
1
a |0i
V p
(7.78)
1
|e (p, s)i
a (p)|0i
V s
r
1
+
|e (p, s)i
b (p)|0i ,
V s
(7.79)
212
3 (0) =
V
.
(2)3
(7.80)
(7.81)
Now we can write down the action of various field operators on different one particles states. Using
the Fourier decomposition of the scalar field in eq. (3.145), and taking into account that ap |0i = 0,
we have
Z
1
p
b
ap eipx |B(k)i
+ (x)|B(k)i = d3 p
3
(2) 2p
Z
1
1
p
= d3 p
b
ap eipx b
ak |0i
3
(2) 2p
V
Z
1
p
eipx [b
ap , b
ak ]|0i .
= d3 p
(7.82)
3
(2) 2p V
By usinbg the commutation relations in eq. (3.122) we have
Z
+ (x)|B(k)i =
(3) (p k) ipx
d3 p p
e
|0i
2p V
(7.83)
1
eikx |0i
2k V
(7.84)
+ (x)|B(k)i =
Similarly, we have
1
eikx |0i
2k V
1
+ (x)|e (p, s)i = p
us (p)eipx |0i
2Ep V
1
0
+ (x)|e+ (p0 , s0 )i = p
vs0 (p0 )eip x |0i ,
2Ep0 V
+ (x)|B(k)i =
(7.85)
213
where k and Ep represent the energies of the scalar and the electron for the 3-momenta in the
subscripts.
Similarly, for the adjoint operators
1
hB(k)| (x) =h0|
eikx
2k V
1
us (p)eipx
he (p, s)| (x) =h0| p
2Ep V
1
0
he+ (p0 , s0 )| (x) =h0| p
vs0 (p0 )eip x ,
2Ep0 V
(7.86)
In the lowest order the only term which contributes to the matrix element is the term shown in
Eq. (7.77). The matrix element at first order in eq. (7.31), between the initial and the final state is
then
Z
(1)
Sf i = ih d4 x e (p)e+ (p0 ) + B(k) .
(7.87)
Using Eqs. (7.85) and (7.86), we obtain
(1)
Sf i
d4 x ei(p+p k)x
=(ih)
us (p)vs0 (p )
1
1
1
p
p
.
2k V 2Ep V 2Ep0 V
(7.88)
Since
Z
(7.89)
we obtain
"
(1)
Sf i =
1
1
1
p
p
2k V 2Ep V 2Ep0 V
#
(2)4 4 (k p p0 ) [(ih)
us (p)vs0 (p0 )]
(7.90)
Comparing with Eq. (5.28) we have therefore that the relativistic matrix element is
iMf i = (ih)
us (p)vs0 (p0 ) ,
and everything else is the history presented in Chapter 6.
(7.91)
214
7.3.1
Left-handed fermion:
d3 p
p
(2)3 2Ep
Z
d3 p
p
(eL ) =
(2)3 2Ep
eL =
i
y (p, s)a(p, s) eipx +x (p, s)a (p, s) eipx .
(7.92)
Right-handed fermion
d3 p
p
eR =
(2)3 2Ep
Z
d3 p
p
(eR ) =
(2)3 2Ep
y 0
(p, s)b(p, s) eipx +x0
(p, s)b (p, s) eipx .
+ |e
L i =xaa |0i
0
+ | e
R i =x bb |0i
+ | e
L i =y aa |0i
0
+ |e
R i =y bb |0i
(7.93)
(7.94)
7.4
Wick Theorem
7.4.1
Basics concepts
M2 =yy 0
(7.95)
Return back to the interpretation of interaction in terms of one exchanged particle, as illustrated
again in Fig. 7.2
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
215
(a)
(b)
x2
x1
x1
(c)
x2
t(x2)<t(x1)
t(x1)<t(x2)
Figure 7.2: Electromagnetic repulsion. The diagrams (a) and (b) are summarized in the diagram (c)
we need for normal ordering. The advantage of normal ordered products is that their expectation
values vanish in the vacuum.
If HI 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
dt1
dt1
dt1
(7.96)
(
0
(t2 t1 ) =
1
t1 > t2
,
t1 < t2
(
0
(t1 t2 ) =
1
t2 > t1
.
t2 < t1
(7.97)
216
In this way.
Z
t2
dt2
dt1
Z
+
(
((
((((
(
(
(
((
dt(1 (t
dt2
2 t1 )HI (t2 )HI (t1 )
(
((((
(
(
Z
t1
Z 2t1
dt1
Z t2
(
(((
((((
(
(
((
+
dt1 ((dt(2 (t
1 t2 )HI (t1 )HI (t2 ) .
((t1
(
(
(
(7.98)
The integral in dt01 is for t1 > t2 , with t2 fixed, so that (t2 t01 ) = 0 for t01 > t2 . Similarly, The
integral in dt02 is for t2 > t1 , with t1 fixed, so that (t1 t02 ) = 0 for t02 > t1 . Therefore,
Z
Z
Z
Z t2
Z
Z t1
dt1
dt2 T{HI (t1 )HI (t2 )} =
dt2
dt1 HI (t2 )HI (t1 ) +
dt1
dt2 HI (t1 )HI (t2 )
Z
Z t1
Z
Z t1
dt1
dt1
=
dt2 HI (t1 )HI (t2 ) +
dt2 HI (t1 )HI (t2 )
Z
Z t1
dt1
dt2 HI (t1 )HI (t2 ) .
(7.99)
=2
X
(i)n
dtn T{HI (t1 )HI (t2 ) . . . HI (tn )} ,
dt1
dt2 . . .
S =1 +
n!
n=1
In terms of the Hamiltonian density, we have
Z
Z
X
(i)n
S =1+
d4 x1 d4 x2 . . . d4 xn T{HI (x1 )HI (x2 ) . . . HI (xn )} ,
n!
n=1
(7.100)
(7.101)
In the above perturbation formalism the states |ii and |f i are, as usual, eigenstates of the unperturbed
free-field Hamiltonian H0 . As such can be introduced inside the integrals
Sf i =hf |S|ii
Z
Z
X
(i)n
=1 +
d4 x1 d4 x2 . . . d4 xn hf | T{HI (x1 )HI (x2 ) . . . HI (xn )}|ii .
n!
n=1
(7.102)
217
(7.103)
In order to evaluate this integrals we need to write the time ordered product in terms of the fields.
This can done by induction. We start by considering the simple no trivial case with two scalar fields
7.4.2
Scalar propagator
With
(x1 )(x2 ) h0|T {(x1 )(x2 )}|0i
(7.104)
(7.105)
since
(x) = + (x) + (x) ,
(7.106)
(7.107)
h0|T [(x1 ), (x2 )] |0i =(t1 t2 )h0|(x1 )(x2 ) + (t2 t1 )h0|(x2 )(x1 )|0i
=(t1 t2 )h0|+ (x1 ) (x2 ) + (t2 t1 )h0|+ (x2 ) (x1 )|0i
=h0|(t1 t2 )(x1 )+ (x2 ) + (t2 t1 )h0|+ (x2 ) (x1 )|0i
(7.108)
with
Z
+ (x) =
1
p
dp
b
ap eipx
3
(2) 2p
3
Z
(x) =
d3 p
1
p
b
ap eipx ,
2p
(2)3
(7.109)
218
we have
Z
d3 p1
d3 p2
ip1 x1
p
p
h0|T ((x1 )(x2 ))|0i =(t1 t2 )h0|
a
a
p eip2 x2 |0i
p1 e
(2)3 2Ep1
(2)3 2Ep2 2
Z
Z
d3 p2
d3 p1
ip2 x2
p
p
+ (t2 t1 )h0|
a
a
p eip1 x1 |0i
p2 e
(2)3 2Ep2
(2)3 2Ep1 1
Z Z
d3 p1 d3 p 2
p
p
ap 1 a
p2 |0i
=(t1 t2 )
eip1 x1 eip2 x2 h0|
6
(2) 2Ep1 2Ep2
Z Z
d3 p d3 p1
p 2 p
+ (t2 t1 )
ap 2 a
p1 |0i
eip2 x2 eip1 x1 h0|
6
(2) 2Ep2 2Ep1
Z
(7.110)
d3 p d3 p2
p 1 p
eip1 x1 eip2 x2 h0|
ap 1 a
p2 |0i
6
(2) 2Ep1 2Ep2
Z Z
d3 p d3 p1
p 2 p
eip2 x2 eip1 x1 h0|
ap 2 a
p1 |0i
+ (t2 t1 )
6
(2) 2Ep2 2Ep1
=(t1 t2 )
Z Z
(7.111)
224
7.4.3
Wick theorem
T {(x1 )(x2 )} =: (x1 )(x2 ) : + (x1 )(x2 )
(7.112)
227
7.5. SCATTERING
Generalizing the results for n scalar or fermion fields, but with an even number of fermions fields,
we have the Wick theorem
T {(x1 )(x2 )(x3 ) (xn )} = : (x1 )(x2 )(x3 ) (xn ) : +
+ (x1 )(x2 ) : (x3 ) (xn ) :
+ : (x1 ) (x2 )(x3 ) (xn ) : +
= : (x1 )(x2 )(x3 ) (xn ) : +
h
i
+ : (x1 )(x2 )(x3 ) (xn ) : +perm
h
i
+ : (x1 )(x2 ) (x3 )(x4 )(x5 ) (xn ) : +perm
+
(7.113)
7.5
Scattering
(7.114)
(7.115)
is
S
(2)
Z Z
(i)2
=
d4 x1 d4 x2 T{HI (x1 )HI (x2 )}
2!
Z Z
(ih)2
=
d4 x1 d4 x2 T{: ()x1 ()x2 }
2!
Z Z
Z Z
(ih)2
(ih)2
4
4
=
d x1 d x2 : ()x1 ()x2 : +
d4 x1 d4 x2 : ()x1 ()x2 : +
2!
2!
(7.116)
The first 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,
228
corresponding to the particles in the initial and final states. The second term in the previous
expansion of the Wick theorem is the only satisfying this requirement. In this way
Z Z
(ih)2
(2) +
+
S (e e e e ) =
(7.117)
d4 x1 d4 x2 (x1 )(x2 ) : ()x1 ()x2 :
2!
On the other hand
(7.118)
From the Fourier expansions in eqs. (??), (??) we have that as and as are the creation and
annihilation operators for particles. As we have only particles (and not antiparticles) in the initial
and final states, the only non-zero contribution of the ordered product in eq. (7.118) must have the
order a a a a. As + and are associated to a and a respectively, the only non-zero contribution
from the ordered fermion product is
(x1 )+
: ()x1 ()x2 := (x1 ) (x2 )+
(x2 ) .
(7.119)
(x1 )+
he (p01 )e (p02 )|iF (x1 x2 ) (x1 ) (x2 )+
(x2 )|e (p1 )e (p2 )i
Z
Z
(ih)2
=
d4 x1 d4 x2 iF (x1 x2 )
2
(7.120)
(7.121)
Therefore
Z
d3 k
d3 k
0
=
u (k)u (k0 )eikx1 eik x2
2Ek V
2Ek0 V
0
as (k)as (k )as (p2 )as (p1 )|0i
Z
(7.122)
229
7.5. SCATTERING
1
1
p
+
(x1 )+
(x2 )|e (p1 )e (p2 )i = p
2Ep1 V 2Ep2 V
u (p1 )u (p2 )eip1 x1 eip2 x2 u (p2 )u (p1 )eip2 x1 eip1 x2 |0i
(7.123)
Following similar steps, we find
1
1
(7.124)
As expected, the final result can be written in term of three different factors: the momentum conservation, normalization, and the relativistic amplitude
!
Y
X
X
Y
1
1
(2)
q
Sf i = i(2)4 4
p0f
pi
Mf i
(7.125)
2Ei V f =1,2 2E 0 V
i=1,2
i=1,2
f =1,2
f
where
Mf i = (ih)2 u (p02 )
u (p01 )F (p1 p02 )u (p1 )u (p2 ) u (p01 )
u (p02 )F (p1 p01 )u (p1 )u (p2 )
(7.126)
The two contributions are displayed in Fig. 7.3 Since
F (q) =
1
m2
(7.127)
1
m2
(7.128)
q2
In the limit q 2 m2
F =
h2 0
Mf i = 2 u (p2 )u (p1 )
u (p01 )u (p2 ) u (p01 )u (p1 )
u (p02 )u (p2 )
m
h2
= 2 [
u(p02 )u(p1 )
u(p01 )u(p2 ) u(p01 )u(p1 )
u(p02 )u(p2 )]
m
(7.129)
230
e (p1)
e (p2)
x1
B(p1 p2 )
e (p2)
e (p1)
x2
e (p1)
e (p1)
x1
B(p1 p1 )
e (p2)
x2
Figure 7.3: fermion scattering
e (p2)
231
7.5. SCATTERING
For one interaction of type we should have
Mf i =
h2
u(p02 )u(p2 )]
u(p01 )u(p2 ) u(p01 )u(p1 )
[
u(p02 )u(p1 )
2
m
(7.130)
For the interaction of a fermion pair with W , we know from the standard model Lagrangian [1],
that
g2
(1 5 )
2 2
(7.131)
= (1 5 )
(7.132)
2
For p MW
the analysis is similar to the previous one with
e
2
(2)4 MW
(7.133)
(7.134)
2
is
the global coupling for p MW
g2
GF
=
2
8MW
2
(7.135)
1
1
1
1
p 0 p 0 Mf i
2E1 V 2E2 V 2E1 V 2E2 V
(7.136)
where
GF
Mf i = ue (p02 )ue (p1 )
u (p01 )u (p2 )
2
The corresponding Feynman diagram is shown in Fig. 7.4 Therefore we have
(7.137)
232
e (p1)
e(p2)
x1
W (p1 p2 )
x2
(p2)
(p1)
GF
Mf i = ue (p02 ) (1 5 )ue (p1 )
u (p01 ) (1 5 )u (p2 )
2
(7.138)
We now must sqaure the scattering amplitude, M, and summing up over final 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
|M|2 = 64 G2F (p1 p2 )(p01 p02 )
(7.139)
s m2
s m2e
|M|2
(7.140)
The center of mass (CM) frame is defined by the condition in Eq. (??):
p1 + p2 = 0
(7.141)
(4) (p1 + p2 p01 p02 ) = (3) (p1 + p2 p01 p02 )(E1 + E2 E10 E20 )
(7.142)
233
7.5. SCATTERING
implies
(
p1 = p2
p1 + p2 p01 p02 = 0
p01 = p02
CM
(7.143)
Moreover
s = E1 + E2
(7.144)
s = (E1 + E2 )2
2
q
q
2
2
2
2
p1 + m1 + p2 + m2
=
q
2
q
2
2
2
2
=
p1 + me + p1 + me
(7.145)
In the CM frame
Therefore
E2 = |p1 |
(7.146)
We already have the expression for |p1 | as given in eq. (??). In this case m2 = 0, and m1 = me , so
that
s m2e
|p1 | =
2 s
(7.147)
From (7.146)
s =E1 + E2
=E1 + |p1 |
E1 = s |p1 |
s + m2e
= s+
2 s
2s s + m2e
=
2 s
s + m2
= e
2 s
(7.148)
(7.149)
234
Then, by using Eqs. (7.143), (7.146) and (7.147), and (7.149), we have
p1 p2 =E1 E2 p1 p2
=E1 |p1 | + p21
(s m2e )(s + m2e ) (s m2e )2
=
+
4s
4s
(s m2e )
(s + m2e + s m2e )
=
4s
1
= (s m2e )
2
(7.150)
(7.151)
1
p01 p02 = p1 p2 (m2 m2e )
2
(7.152)
1
p01 p02 = (s m2 )
2
(7.153)
G2F (s m2 )2
(7.154)
(7.155)
(7.156)
Chapter 8
Neutrinos
8.1
Weinberg operator
c =C T
Lc =CL .
The Majorana mass term should be of the form Lc L . Since L has I3 = 1/2, the Majorana mass
T
term has I3 = 11 . With L = L eL
Lc i2 L (3, 2) .
One would need an isotriplet scalar field (3, 2), which either elemental or composite. The term
H T i2 H (3, 2) ,
can play the role of the composite triplet, where H = H + H 0
Can the operator
f
Lc i2 L H T i2 H
M
1
235
T
236
CHAPTER 8. NEUTRINOS
e
= Lc ecL
,i L , L
eL
eL
L
= ecL eL Lc L , i Lc L + ecL eL , Lc eL + ecL L
Replacing back Lc H + , ecL H 0 , L H + , and eL H 0 , we have
H T i2 H = H 0 H 0 H + H + , i H 0 H 0 + H + H + , H + H 0 + H + H + .
Lc i2 L H T i2 H
= 2 ecL eL H + H + Lc L H 0 H 0 + Lc eL H + H 0 + ecL L H + H 0
= 2 Lc H 0 ecL H + H 0 L H + eL
0
L
H
H 0 H +
= 2 Lc ecL
eL
H +
We have the Weinberg operator
f
Lc i2 L H T i2 H + h.c
2M
f c e e
=
LH
H L + h.c
M
f
Lc H 0 ecL H + H 0 L H + eL + h.c
=
M
f
12 Lc1 H2 + 21 Lc2 H1 (12 L1 H2 + 21 L2 H1 ) + +h.c
=
M
f
= Lca Hc Lb Hd ac bd + h.c .
M
L=
(8.1)
Chapter 9
Three body decays
9.1
Muon decay
(P
p
)
1
2
3
(2)5 2M
2E1 2E2 2E3
Z
3
1
1 d p1
d3 p2 d3 p3
2 4
=
|M| (P p1 p2 p3 )
(2)5 2M 2E1
2E2 2E3
d =
(9.1)
9.1.1
Amplitude estimation
Since M is dimensionless, the amplitude averaged over spins for decay must be
|M|2 = CG2F m4 .
(9.2)
1
C = (2 2 1 1) = 2
2
(9.3)
We use
The first factor is for the initial average and the factor are for the number of spin states of , e and
the two neutrinos.
237
238
d3 p1 d3 p2
(E E1 E2 ) 3 (P p1 p2 )
2E1 2E2
Z
3
d p2
(E E1 E2 )
3 (P p1 p2 )d3 p1
4E1 E2
Z
d3 p2
3 (p1 + p2 P)d3 p1
(E E1 E2 )
4E1 E2
Z
d3 p2
(E E1 E2 )
3 [p1 (P p2 )]d3 p1
4E1 E2
(9.4)
using
Z
(x x0 )dx = 1
(9.5)
we have
Z
d3 p1 d3 p2
d3 p2
(P p1 p2 )
= (E E1 E2 )
2E1 2E2
4E1 E2
Z
Z
3
3
d p1 d p2
p22 d|p2 |d
4
(P p1 p2 )
= (E E1 E2 )
2E1 2E2
4E1 E2
Z
(9.6)
(9.7)
(9.8)
Therefore
dE
d|p2 | =
|p2 |
E1 E2
E1 + E2
(9.9)
239
dE
p22 d
E1 E2
(E E1 E2 )
|p2 |
4E1 E2 E1 + E2
|p2 | d
dE (E E1 E2 )
4(E1 + E2 )
|p2 |
d
4E
(9.10)
(9.11)
(P
p
)
d =
1
2
3
(2)5 2M 2E1
2E2 2E3
Z
3
1
1 d p1
d3 p2 d3 p3
2
4
=
|M|
(P p1 p2 p3 )
(2)5 2M 8E1
E2 E3
3
1
1 d p1
=
|M|2 (2)
5
(2) 2M 8E1
Z
G2F m4
2
=
p d|p1 | d
8(2)4 m E1 1
G2F m3
E1 dE1
4(2)3
(9.12)
E1 dE1
4(2)3 0
G2F m3 m2
=
,
4(2)3 8
or
=
3 G2F m5
.
4 192 3
(9.13)
(9.14)
240
9.1.2
Amplitude calculation
2g
(eL eL W+ + eL eL W + L L W+ +
L L W )
L=
2
g
= (e PL eW+ + e PL e W + PL W+ +
PL W )
2
g
= (e (1 5 )eW+ + e (1 5 )e W + (1 5 )W+ +
(1 5 ) W )
2 2
(9.15)
where
(e PL eW+ ) = e PL (e ) W = e PL 0 e W = e 0 0 PR 0 e W
= e 0 0 PL e W = e PL e W
We can build the effective Lagrangian
Applying the Feynman rules to the diagram in fig. 9.1 we have the amplitude
2
g + q q /MW
ig 2
u3 (1 5 )u1
u4 (1 5 )v2
M=
2
8
q 2 MW
(9.16)
(9.17)
(p + m)v = 0
v(p + m) = 0 .
(9.18)
In this way
1
1
q (1 5 )u1 u4 q (1 5 ) = 2 (1 + 5 )qu
4 q(1
1u
5 )
2
MW
MW
m me
=
(1 + 5 )u1 u4 (1 5 )
2
MW
(9.19)
the term in q q can be safely neglected. The term q 2 is m2 is small compared with m2W . Therefore
ig 2
u3 (1 5 )u1 u4 (1 5 )v2
2
8MW
iGF
= u3 (1 5 )u1 u4 (1 5 )v2
2
M=
(9.20)
241
GF
g2
= .
2
8MW
2
(9.21)
The conjugate is, following the same steps that in eq. (9.16)
ig 2
[
u3 (1 5 )u1 ] [
u4 (1 5 )v2 ]
2
8MW
ig 2
M =
[
u1 (1 5 )u3 ] [
v2 (1 5 )u4 ] .
2
8MW
M =
g4
[
u3 (1 5 )u1 u1 (1 5 )u3 ]
4
64MW
[
u4 (1 5 )v2 v2 (1 5 )u4 ]
g4
=
L M
4
64MW
(9.22)
|M|2 =
(9.23)
where
L = u3 (1 5 ) u1 u1 (1 5 ) u3
M = u4
(1 5 ) v2 v2
(1 5 ) u4
(9.24)
242
(9.25)
Using
X
u(p, s)
u(p, s) = (p + m)
X
s
v(p, s)
v (p, s) = (p m)
(9.26)
L = Tr [(p 3 ) (1 5 )(p 1 + m ) (1 5 )]
h
i
Tr (1 5 )(p1 + m ) (1 5 )
h
i
=p3 Tr ( 5 )(p1 (1 5 ) + m (1 5 ))
h
=p3 Tr p1 (1 5 ) p1 5 (1 5 )
=p3
+m (1 5 ) m 5 (1 5 )]
h
=p3 Tr p1 p1 5 p1 5 + p1 5 5
+m m 5 m 5 + m 5 5 ]
(9.27)
L =p3 p1 Tr 5 5 + 52
=2p3 p1 Tr [ (1 5 )]
(9.28)
Similarly
X
s
M = 2p4 p2 Tr (1 5 )
(9.29)
243
|M|2 =
(9.30)
The demonstration of the used TrTr identity can be found in Appendix B. of [8].
The spinaveraged differential decay width for e e is
1
1 d3 p3 1 X
d3 p 2 d3 p4
2
4
d =
|M|
(p
p
)
1
2
3
4
(2)5 2E1 2E3 2
2E2 2E4
3
3
X
1 1
d p3 d3 p4
1 d p2 4
=
|M|2
(p
p
)
1
2
3
4
2E1 2
(2)5 8E2
E3 E4
4
3
1 4g
1
d p2 4
d3 p3 d3 p4
(p
p
)(p
p
)
(p
p
)
=
1
2
3
4
1
2
3
4
4
2 MW
(2)5 2E1
2E2
2E3 2E4
4
2g
p p d3 p4 I
=
4
5
16(2) MW
E1 E4 1 4
(9.31)
p3 p2 4 (p p2 p3 )
d3 p2 d3 p3
.
E2 E3
(9.32)
244
g p p =2g p3 p2
p p =2p3 p2 .
(9.33)
(9.34)
4 (p p2 p3 )
d3 p2 d3 p3
,
E2 E3
(9.35)
Since
m2 0 =E2 p2
E2 =p2
(9.36)
and in addition the integral I is covariant, we choose to evaluate it in the rest frame of the two
neutrinos |p2 | = |p3 |, which implies E2 = E3 .
Z
d3 p 2 d3 p3
3
I = (E E2 E3 ) (p p2 p3 )
E2 E3
Z
Z
d3 p2
= (E E2 E3 )
3 (p p2 p3 )d3 p3
E2 E3
{z
}
|
=1
Z
(E 2E2 ) 2
=
p2 d|p2 |d
E22
Z
(E 2E2 ) 2
=
E2 dE2 (4)
E22
Z
E
=4 2 E2
dE2
2
Z
1
E
=4
E2
dE2
2
2
=2
(9.37)
245
g I = 4A + p B =
p3 p2 4 (p p2 p3 )
d3 p2 d3 p3
p2
= I = p2
E2 E3
2
In order to compute p p I , 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 = p2 + p3 = (p0 , 0) and E2 = E3
p p I = p2 A + p4 B
Z 3
d p2 d3 p3
=p p
p3 p2 4 (p p2 p3 )
E2 E3
Z 3
d p2 d3 p3
E3 p0 E2 p0 4 (p p2 p3 )
=
E2 E3
Z
0 2
= (p )
d3 p2 d3 p3 4 (p p2 p3 )
Z
0 2
= (p )
d3 p2 (p0 2E2 )
2
Z
1 p0
p2 p2
2
2
= (p)
dE2 E2 d ( E2 ) = 4
2 2
2
2
4
p
=
2
(9.38)
(9.39)
(9.40)
where the usual tricks have been used to simplify the integrals, using the delta function inside.
Therefore
p2
A = ( B)
(9.41)
4
p4
( B) + p4 B
4
B
+B
4
4
3B
4
B
p4
2
=
2
=
4
=
3
=
(9.42)
246
( )
4
3
p2 2
= ( )
4 3
p2
=
6
A=
I =
g p2 + 2p p .
6
(9.43)
(9.44)
d =
(9.45)
For further evaluation we will use the rest frame of the decaying muon. In this frame the four
momentum are
p1 =(m , 0)
p4 =(E4 , p4 )
p =p1 p4 = (m E4 , p4 )
p2 =E 2 p2 = m2 2m E4 + (E42 p24 ) = m2 + m2e 2m E4
(9.46)
Moreover
p1 p4 =m E4
p p1 =m2 m E4
=
=
2
2
1/2
dE4
2 (E4 me )
|p4 |
E4
dE4
d|p4 | =
|p4 |
d3 p4 =p24 d|p4 | d = |p4 |E4 dE4 d
(9.47)
247
2g 4
|p4 | dE4 d[(m2 + m2e 2m E4 )m E4 + 2(m2 m E4 )(m E4 m2e )] (9.48)
4
4
192(2) MW m
2g 4 (4)
E4 dE4 [(m2 2m E4 )m E4 + 2(m2 m E4 )m E4 ]
4
192(2)4 MW
m
2 2g 4
=
m E42 [m2 2m E4 + 2m2 2m E4 ] dE4
4
3
192(2) MW m
2
4g 4
2
=
E
3m
4m
E
dE4
4
4
4
192(2)3 MW
4g 4
E4
2 2
m E4 3 4
=
dE4
4
192(2)3 MW
m
m4 2E4 2
2E4
m
2E4
4g 4
32
d
=
4
192(2)3 MW
4
m
m
2
m
2
5
4
m 2E4
4g
2E4
2E4
=
32
d
4
192(2)3 MW
8
m
m
m
d =
(9.49)
(9.50)
(9.51)
(9.52)
(p2 + p3 )2 is the invariant mass squared of the + e system, which ranges from 0 to m2 . For
(p2 + p3 )2 = m we have E4min = 0, while for (p2 + p3 )2 = 0 we have E4max = m /2. The missing
integration on d is in the variable x such that
x=
2E
,
m
xmin = 0 ,
xmax =
2Emax
= 1.
m
(9.53)
248
Therefore
Z
m5 1 2
4g 4
=
x [3 2x] dx
4
192(2)3 MW
8 0
m5 1
4g 4
=
4
192(2)3 MW
8 2
5
4
m
g
=
4
3
192 8MW 4
2
g4
m5
=
4
64MW 192 3
G2 2
= F
m5
3
2 192
G2
= F 3 m5
192
(9.54)
4g 4
E42 3m2 4m E4 dE4
4
3
192(2) MW
2
4
g4
2
E
3m
=
4m
E
dE4
4
4
4
32MW 6(2)3
2
G2F 2
2
E
3m
4m
E
dE4
=
4
4
3 8 3
4 E4
G2
= F 3 E42 3m2 1
dE4
12
3 m
d =
d
G2F 2 2
4E4
=
m E 3
dE4 12 3 4
m
(9.55)
(9.56)
249
d =
d =
4g 4
dE4 (E42 m2e )1/2 [3m3 E4 + 3m2e m E4 4(m E4 )2 2m2 m2e ]
4
192(2)3 MW
m
(9.57)
9.2
=hj N j PL L1 2 hj N j PL L2 1 + h.c.
=hj N j PL 0 hj N j PL l + + h.c
(9.59)
(N j PL l + ) = l PL 0 Nj = l PR 0 Nj
(9.60)
where
Therefore
L =hj N j PL 0 hj N j PL l + + hj PR Nj 0 hj l PR Nj
1
= hj N j (1 5 ) 0 hj N j (1 5 )l + + hj (1 + 5 )Nj 0 hj l (1 + 5 )Nj (9.61)
2
250
Applying Feynman rules to the diagram in fig.2 Nj (p1 ) l (p3 )h+ , h+ l+ (p2 ) + Ni (p4 ).
we have the amplitude
1
M = ihj u3 (1 5 )u1
hi u4 (1 5 )v2
q 2 M2
1
ihj u3 (1 5 )u1
hi u4 (1 5 )v2
q 2 M2
iHij
u3 (1 5 )u1 u4 (1 5 )v2
M2
(9.62)
where
Hij = hj hi + hi hj
iHij
[
u3 (1 5 )u1 ] [
u4 (1 5 )v2 ]
2
M
iHij
=
[
u1 (1 + 5 )u3 ][
v2 (1 + 5 )u4 ] .
M2
(9.63)
M =
(9.64)
251
2
Hij
|M| =
[
u3 (1 5 ) u1 u1 (1 + 5 ) u3 ][
u4 (1 5 ) v2 v2 (1 + 5 ) u4 ]
4
M
2
2
Hij
=
[u3 u3 (1 5 ) u1 u1 (1 + 5 ) ][u4 u4 (1 5 ) v2 v2 (1 + 5 ) ]
4
M
2
Hij
=
[(u3 u3 ) (1 5 ) (u1 u1 ) (1 + 5 ) ][(u4 u4 ) (1 5 ) (v2 v2 ) (1 + 5 ) ]
M4
2
Hij
Tr[(u3 u3 )(1 5 )(u1 u1 )(1 + 5 )] Tr[(u4 u4 )(1 5 )(v2 v2 )(1 + 5 )]
M4
(9.65)
2
Hij
LM
M4
(9.66)
L = Tr[(p 3 p 3 5 )(p 1 + p 1 5 + Mj + Mj 5 ]
(9.67)
L = Tr[p 3 p 1 + p 3 p 1 5 + Mj p 3 + Mj p 3 5 p 3 5 p 1 p 3 5 p 1 5 p 3 5 Mj + Mj 5 ]
=2 Tr[p 3 p 1 ]
=2p3 p1 Tr[ ]
=8p3 p1 g
=8(p3 p1 )
(9.68)
Similarly
M = 8(p4 p2 )
(9.69)
Therefore
|M|2 =
2
Hij
64(p3 p4 )(p1 p2 )
M4
|M|2 =
2
Hij
4 64(p3 p4 )(p1 p2 )
4M4
(9.70)
252
In this way, comparing with eq. (9.30), the results for the moun decay can be directly used after the
replacements
2
Hij
g4
4
64MW
4M4
2
16Hij
g4
4
MW
M4
m Mj
Mi
me
.
x=
m
Mj
(9.71)
2
16Hij
Mj6
4
I (x)
M4 192(2)3 Mj 16
(hj hi + hi hj )2 Mj5
=
I (x)
2M4
192 3
(9.72)
where
I(x) = 1 8x2 24x4 ln(x) + 8x6 x8 ,
x=
Mi
.
Mj
(9.73)
(hj hi + hi hj )2 Mj5
I (x)
2M40
192 3
(9.74)
(N2 l l+ N1 ) =
(N2 l l+ N1 ) =
=
(9.75)
(9.76)
253
Mi
Mj
Mi
Mj
(Nj
l l+ Ni )
(9.77)
(9.78)
while for
P
Br(N3 N2 )
I(M2 /M3 )
+
h21 h23 + (h1 h3 )2 M40
Br(N3 l l N1 )
P
Br(N3 l l+ N2 ) h22 h23 + (h2 h3 )2
P
I(M2 /M3 )
2 2
+
h1 h3 + (h1 h3 )2
Br(N3 l l N1 )
(9.79)
M1
M2
"
1
1
+ 4
4
M M0
#
(9.80)
(9.81)
Br(N3 |{z}
l l N1 ) = Br(N3 N2 ) Br(N2 l l N1 )
(9.82)
that seem to be very difficult to reconstruct. This also seem to be an irreducible background for
Br(N2 l l N1 )
(9.83)
To get rid of processes like the one in eq. (9.82) must be Br(N3 N2 ) is suppressed. This
happens if
254
I(M2 /M 3) 1. In this case the mutilepton signal for N3 is also suppressed. Clearly this
happens for M2 M3 as I(x) is a sharpest function which controls the kinematical suppression.
We show below for an specific point that even for M3 M2 20 GeV, we can have the Branching
in eq. (9.81) sufficiently large.
M M0
In appendix 9.A, it is shown a full set of yukawas consistent with neutrino physics. For this
solution
Br( + N3 )
0.61
Br( + N1 )
Br( + N1 ) 0.51
Br( + N2 )
0.37
Br( + N1 )
Br( + N2 ) 0.19
Br( + N3 ) 0.30
(9.84)
N
)
2
0.0812
0.0615
+
+
Br(N3 l l N1 )
Br(N3 l l N1 )
P
Br(N3 N1 )
P
1.320
(9.86)
+
Br(N3 l l N1 )
X
Br(N3 l l+ N1 )
1
= 0.406
1 + 0.0812 + 0.0615 + 1.320
Br(N3 N1 ) 0.536
Br(N3 N2 ) 0.030
Br(N3 l l+ N2 ) 0.025
(9.87)
255
Br(N3 N2 )
We have that
M25
tot (N2 ) = h21 h22 + (h1 h2 )2
I
192 3
X
vis (N2 N1 )
(N2 l l+ N1 )
=
vis (N3 N1 )
=
invis (N3 N2 )
=
h21 h22
M1
M2
M1
M2
M1
M3
M2
M3
"
1
1
+
M4 M40
(N3 l l+ N1 )
h21 h23
+ (h1 h2 )2 M25
I
M4
192 3
+ (h1 h3 )2 M35
I
M4
192 3
(N3 N2 )
h22 h23
+ (h2 h3 )2 M35
I
M40
192 3
4
4
3
3
192
M3
192
M2
M 0
M
=
2
2
5
h1 h3 +(h1 h3 )2 M3
1
I M
tot (N2 )
192 3
M3
M 4
h2 h2 + (h2 h3 )2
M2
1
= 22 23
I
h1 h3 + (h1 h3 )2
M3
1
1
4
M0 M 4 + M 4
0
h2 h2 + (h2 h3 )2
M2
1
= 22 32
I
2
M 40
h1 h3 + (h1 h3 )
M3
1 + M4
(9.88)
256
9.A
Sample point
write(32,*) (h(i,1),i=1,3),(h(i,2),i=1,3),(h(i,3),i=1,3)
(9.89)
(9.90)
M2 22.8695451 GeV
M 149.1382 GeV
I(M1 /M3 ) 1
9.B
M3 43.126911 GeV
(9.91)
(9.92)
Preliminary discussion
One interesting possibility in view of the large invisible direct decay, like N3 N1 , is to get the
observables from the missing plus one energetic lepton (coming from + ) signal. May be decays like
+ l+ N3 l+ E T
+ l+ N2 l+ E T
(9.93)
0
Once R,I
, or are produced the full list of signals is: For production. The decay to Nj is
3h2j 1/2
Mj2 + m2
+
+
2
2
2
( l Nj ) =
M , Mj , m 1
(9.94)
16M
M2
257
l+ Nj )
3h2j 1/2
Mj2 + m2
2
2
2
=
M , Mj , m 1
16M
M2
(9.95)
with
i1/2
h
2
1/2 M2 , Mj2 , m2 = M2 + Mj2 m2 4M2 Mj2
(9.96)
#1/2
4Mj2
M2
1/2
Mj2 4Mj2
2
M 1 + 2 2
M
M2
1/2
Mj2
2
M 1 2 2
M
Mj2
2
M 1 2
M
M2 , Mj2 , m2 M2
"
Mj2
1+ 2
M
2
(9.97)
Therefore
2 2
M
j
X
1 2
( + l+ Nj )
16
1
Mj2
2
3hj M
12 2
M
16
1
3h2j M
j = 2, 3
j=1
j = 2, 3
j=1
(9.98)
In this way
tor ( + ) =
X
j
( + l+ Nj )
M22
M32
3M 2
2
2
h1 + h2 1 2 2 + h3 1 2 2
16
M
M
(9.99)
258
P
+
+
Br( + Nj )
( l Nj )
P
=
+
+
Br( + Ni )
( l Ni )
h2j 1 2Mj2 /M2
2
hi 1 2Mi2 /M2
h2j
(1 2Mj2 /M2 )(1 2Mi2 /M2 )1
h2i
h2j
2 (1 2Mj2 /M2 )(1 + 2Mi2 /M2 )
hi
2
Mj Mi2
h2j
2 12
hi
M2
(9.100)
1
1 + b/a + c/a
1+
Br( + N3 )
Br( + N1 )
(9.101)
In this way
Br( + N1 ) =
1
+
Br( + N3 )
Br( + N1 )
(9.102)
From eq.
Br (N3 N1 )
=
Br(N3 |{z}
N2 )
(9.103)
Chapter 10
One-loop neutrino masses
Here we calculate the one-loop neutrino mases in the mass-basis
10.1
Extra particles
We assume to have -scalar and n-fermions with the mass-eigenstate couplings (Weyl spinors)
0
L = yin i n S + ymi
n i S + h.c + mn n n + m2 S2 +
(10.1)
From left to right in Figure 10.1
iij (p)
In the limit p 0
d4k
(yin ) iSF (k) (yjn ) iF (p + k)
(2)4
Z 4
yin yjn
dk
k + mn
=
2
2
2
2
16
i k mn (p + k)2 m2S
Z
Mij =
yin yjn
2
2
m
B
0;
m
,
m
0
n
n
S .
16 2
(10.3)
p+k
p
L
(10.2)
p
L
260
where B0 0; m2n , m2S is the B0 Passarino-Veltman function [?]
B0 0; m2n , m2S
d4 k
1
2
2
2
i k mn k 2 m2S
1
= 2
A0 m2n A0 m2S ,
2
mn mS
Z
and
A0 m2 = m2 + 1 ln m2 /2
with
=
2
ln + E ,
B0 0; m2n , m2S =
m2n
yin yjn
mn [cte() + f (mn , mS )] ,
16 2
(10.5)
where
m2S ln m2S m2n ln m2n
f (mn , mS ) =
m2n m2S
(10.6)
261
10.2. APPLICATIONS
10.2
Applications
10.3
There [?]
yin =hi N3n
(10.7)
and
Mij
3
X hi hj X
10.3.1
16 2
(N3n )2 mn f (mS , mn ) .
(10.8)
n=1
Zee
In the Zee model we can work in the Higgs-basis with hH1 i = v/ 2, hH2 i = 0 [?]. In that basis the
scalar potential is
1
V =21 H1 H1 + 22 H2 H2 [23 H1 H2 + H.c.] + 1 (H1 H1 )2
2
1
+ 2 (H2 H2 )2 + 3 (H1 H1 )(H2 H2 ) + 4 (H1 H2 )(H2 H1 )
2
1
2
+
5 (H1 H2 ) + [6 (H1 H1 ) + 7 (H2 H2 )]H1 H2 + H.c.
2
+ 2h |h+ |2 + h |h+ |4 + 8 |h+ |2 H1 H1 + 9 |h+ |2 H2 H2
+ 10 |h+ |2 (H1 H2 + H.c.) + H1 H2 h .
(10.9)
We define
G+
H1 =
Re (H10 ) + iG0 / 2
H+
H2 =
Re (H20 ) + iA0 / 2
(10.10)
0 2
2
0 4
=
Re H1 + 1 Re H1 +
Re (H10 ) Re (H10 ) 1
2
3
1
=221 Re H10 + 1 4 Re H10 +
2
(10.11)
262
3
V
= 23 2 Re H10 + 6 2 Re H10 +
0
Re (H2 )
(10.12)
=0 ,
(10.13)
we have
1
21 + 1 =0
2
1
3 + 6 =0
2
(10.14)
1
3 = 6 .
2
(10.15)
or
1
21 = 1
2
Replacing back in the potential, we have for the charged scalars that In the basis = (G , H , h )
the squared-mass matrix for the charged Higgs states is given by
Lcharged = M2C + ,
(10.16)
where
0
0
0
MH2 v/ 2 , and
M2C = 0
0 v/ 2
M233
(10.17)
1
MH2 =22 + v 2 3
2
M233 =2h + v 2 8 .
By defining the mass basis state as S = G , h.1 , h
2 , then after the rotation
S = R ,
(10.18)
(10.19)
where
1
0
0
R = 0 cos sin
0 sin cos
(10.20)
263
2 +
Lcharged = M12 h+
1 h1 + M2 h2 h2 +
(10.21)
The relevant terms for neutrino masses in the mass eigenstate basis reads
+
+
L =02 cos (eR ) L h
1 2 sin (eR ) L h2 + fji (eL )j (L )i sin h1 + fji (eL )j (L )i cos h2
cl (eR ) eL + h.c + M 2 h+ h + M 2 h+ h
+M
1 1 1
2 2 2
(10.22)
Therefore
ckk fjk sin
Mij (02 )ik cos M
(10.23)
Li T C (2fik ) eLk R2 h+
+
mk eLk eRk + h.c ,
(10.24)
Therefore
yik =Oik R1
yjk =2fik R1
(10.25)
where
cos sin
R=
sin cos
(10.26)
264
Mij =
=
=
X X 2Oik R1 fjk R2
k
X Oik fjk
k
8 2
X Oik fjk
k
8 2
X Oik fjk
k
16 2
16 2
mk [cte() + f (mn , mS )]
mk [R11 R21 cte() + R11 R21 f (mk , M1 ) + R12 R22 cte() + R12 R22 f (mk , M2 )]
mk cos sin [f (mk , M1 ) f (mk , M2 )]
mk sin 2 [f (0, M1 ) f (0, M2 )]
X Oik fjk
(10.27)
sin 2 X
(fik mk Ojk + fjk mk Oik ) [f (mk , M1 ) f (mk , M2 )]
(4)2 k
(10.28)
10.3.2
Minimal Zee
L=
(10.30)
265
Therefore
tan
2
mk ik R1
v
=2fik R1
yik =
yjk
(10.31)
Mij
X fjk mk ik sin 2
k
(4)2
fji mi sin 2
ln
=
(4)2
which is non-zero only for i 6= j.
ln
M22
M12
M22
M12
.
(10.32)
266
Bibliography
[1] Diego Restrepo, Haca la teora cuantica de campos, curso web, http://gfif.udea.edu.co
[2] M. Maggiore, A Modern introduction to quantum field 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 Publication)
[4] A. Lahiri and P. B. Pal, A first book of quantum field theory, Harrow, UK: Alpha Sci. Int.
(2005) 380 p
[5] Pal, P. Representation-independent manipulations with Dirac matrices and spinors,
arXiv:physics/0703214v2 [physics.ed-ph]
[6] Wikipedia article: http://en.wikipedia.org/wiki/Particle_decay
[7] M. Klasen, Int. J. Mod. Phys. C 14 (2003) 1273 [arXiv:hep-ph/0210426].
[8] Electroweak Interactions, Peter Renton
[9] Michael E. Peskin and Daniel V. Schroeder. An introduction to Quantum Field Theory, AddisonWesley Publishing Company(1995), p. 101
[10] ibdg 1, p. 105
[11] Francis Halzen and Alan D. Martin. Quarks & Leptons: An introductory Course in Modern
Particle Physics, John Wiley & Sons(1984), p. 89
[12] Chris Quigg. Gauge theory of the Strong, Weak and Electromagnetic Interactions, Westview
Press(1997), p. 110
[13] ibdg 4, p. 130.
267
268
BIBLIOGRAPHY