Quantum Field Theory
Quantum Field Theory
Quantum Field Theory
Theory
Diego Restrepo
Instituto de Fsica
Universidad de Antioquia
2012
ii
License
All materials are licensed under the Creative Commons Attribution-Share Alike 3.0 United
States License.
iii
iv
Contents
Introduction 1
v
vi CONTENTS
5 Smatrix 171
5.1 The Smatrix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171
5.2 Relativistic and no relativistic normalizations . . . . . . . . . . . . . . . . . . . . . . 173
5.3 Process probability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
5.4 Cross Section . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177
5.4.1 2to2 cross section . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181
5.5 Decay Rates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187
5.5.1 Two body decays . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 188
5.6 Backup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189
8 Neutrinos 223
8.1 Weinberg operator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223
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.
This work have been partially supported by Dedicacion Exclusiva 2008-2009 project: RR 26663
1
2 CONTENTS
Chapter 1
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.
donde d4 x = dt dx dy dz. Considere primero una variacion solo de los campos, tal que (x = x )
f
f (x + x) = f (x) + x + (1.4)
x
3
4 CHAPTER 1. CLASSICAL FIELD THEORY
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
L= 2
, (1.12)
2 v t z
x x0 =x + a
x =a (1.26)
tenemos
0 (x0 ) = 0 (x + a) (1.27)
0 (x)
0 (x) + a (1.28)
x
= [(x) + (x)] + [(x) + (x)]a (1.29)
x
(x)
(x) + (x) + a , (1.30)
x
donde, por simplicidad, es de nuevo un campo real, y en el ultimo paso hemos despreciado un
termino de orden a . Entonces,
(x)
(x) 0 (x0 ) (x) = (x) + a . (1.31)
x
= ( )a , (1.32)
d 4 x0 = d 4 x (1.34)
En este caso, asumiendo que el campo satisface las ecuaciones de Euler-Lagrange y usando la ec. (1.32)
8 CHAPTER 1. CLASSICAL FIELD THEORY
y (1.9) tenemos
Z Z
4 0 0 0
S = d x L( , , x ) d4 x L((x), (x), x)
ZR R Z
4
= d x L( + , + (), x + a ) d4 x L
ZR R
Z
4 L L
d x L+ + () + ( L)a d4 x L
( )
ZR R
L L
= d4 x + () + ( L)a
( )
ZR
4 L L
= dx + () + ( L)a
R ( ) ( )
L
Z
4
= d x + ( L)a
R ( )
L
Z
4
= d x + La
R ( )
L
Z
4
= d x a + La
R ( )
L
Z
4
= d x a + L (a )
R ( )
L
Z
4
= d x ( ) + L a
R ( )
(1.35)
Z
= d4 x (T a ) = 0. (1.36)
R
1.1. LAGRANGIAN FORMULATION 9
Y por consiguiente
T a = 0, (1.37)
De modo que para cada , con a 6= 0, se satisface:
T = 0, (1.38)
donde
L
T = ( ) 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:
T0 = 0 (1.40)
Donde la densidad de Energa, o mas de forma mas general: la densidad Hamiltonina corresponde a
T00
L
H = T00 = 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)
L
(x) = . (1.43)
((x)/t)
El teorema de Noether en este caso establece que la invarianza de la Accion bajo traslaciones tem-
porales da lugar a la ecuacion de continuidad (1.38) para = 0
T0 = 0 (1.44)
cuya carga conservada corresponde a la energa
Z Z
3 0
H= d x T0 = d3 x H. (1.45)
V V
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 CHAPTER 1. CLASSICAL FIELD THEORY
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
= 0 = (ei 1) i (1.55)
i . (1.56)
Usando ademas la ec. (1.51) en la definicion de J 0 dada por la ec. (1.20), tenemos
0 L L
J = +
(0 ) (0 )
i i
= (i) + (i )
2 2
= , (1.57)
y
i L L
J = +
(i ) (i )
1 1
= i (i) + (i ) i
2m 2m
i
= (i i ) . (1.58)
2m
Entonces, normalizando apropiadamente la corriente escogiendo = 1, tenemos
J 0 = (1.59)
i
J= ( ) . (1.60)
2m
De acuerdo a la ec. (1.59), la cantidad conservada corresponde a la probabilidad de la funcion de
onda y normalizando apropiadamente la ec. (1.25)
Z
Q = d3 x = 1. (1.61)
V
12 CHAPTER 1. CLASSICAL FIELD THEORY
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
Ti0 = (i ) + (i )
(0 ) (0 )
i i
Ti0 = (i ) + (i ) (1.63)
2 2
Entonces, definiendo
i
T0 = ( ) (1.64)
2
Ademas
i
T0 = (( ) )
2
i
= i + ( ) . (1.65)
2
Integrando en el volumen
i
Z Z Z
0 3
T d x = i 3
d x + d3 x (1.66)
V V 2 V
De acuerdo a la ec. (1.61), la ultima integral es una constante y
Z Z
0 3
T d x = i d3 x
V ZV
hb
pi = pb d3 x (1.67)
V
De modo que hb
pi son las cargas conservadas asociadas al valor esperado el operador de momentum
b = i .
p (1.68)
De otro lado
L L
T00 = 0 + 0 L
(0 ) (0 )
i i 1 i
= 0 + 0 i i + ( 0 0 ) V
2 2 2m 2
1
= i i V (1.69)
2m
1.2. GLOBAL GAUGE INVARIANCE 13
Como las corrientes solo estan determinadas hasta un factor de proporcionalidad, definimos
1
H T00 = + V
2m
1 1 2
= ( ) + V . (1.70)
2m 2m
Integrando sobre el volumen y usando la ec. (1.67)
1 1 2
Z Z Z
3
Hd x = ( ) + + V d3 x
V 2m V V 2m
1 1 2
Z Z
= ( ) + + V d3 x
2m V 2m
V
i 1 2
Z
= hbpi + + V d3 x
2m V 2m
1
Z
= + V d3 x .
2
(1.71)
V 2m
Entonces
1 2
Z Z
3
H Hd x = +V d3 x
V V 2m
Z
= d3 x H
b = hHi.
b (1.72)
V
b =i,
H p = i . (1.75)
t
Retornando a la ec. (1.67), tenemos que para la solucion de partcula libre de la ecuacion de
Schrodinger
= A eikx , (1.76)
14 CHAPTER 1. CLASSICAL FIELD THEORY
Ejercicio: De la ec. (1.72) obtenega la densidad Hamiltoniana, y usando la ec. (1.41) encontrar la densidad
Lagrangiana (1.49).
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.
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):
1 i
L(, , , ) = + V (1.78)
2m 2 t t
1 i
= i i ( 0 0 ) + V .
2m 2
This Lagrangian is not invariant under local phase changes of the wave function:
0 = ei(x)
= ei(x) + ei(x)
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 .
(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)
D (D )0 = D0 0 =( + X0 ) ei(x)
Note that D transforms like the field , and because of this is called the covariant derivative of
. Similarly
(D ) (D )0 =( + X0 ) ei(x)
= + X ei(x)
= (D ) ei(x) . (1.83)
1
A X , (1.84)
iq
D = + iqA . (1.85)
The transformation properties of A can be obtained from the X transformation in eq. (1.81):
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:
1 i
L(, , , , A ) = (Di ) Di [ D0 (D0 ) ] + V (x). (1.87)
2m 2
where
1
A A0 = A (x) . (1.88)
q
1
like the electron field as described by the usual Scrodinger equation.
1.4. NOTACION RELATIVISTA 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
1 i
L L0 = (Di )0 (Di )0 0 (D0 )0 (D0 )0 0 + 0 V (x) 0
2m 2
1 i(x) i(x)
= (Di ) e e (Di )
2m
i
ei(x) ei(x) (D0 ) (D0 ) ei(x) ei(x) + ei(x) ei(x) V (x).
2
=L . (1.89)
To preserve invariance one notices that it is necessary to counteract the variation of with x, y,
z, and t by introducing the electromagnetic 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.
a g a (1.91)
a b = g a b = a0 b0 a b (1.93)
18 CHAPTER 1. CLASSICAL FIELD THEORY
El inverso de la metrica es
{g } {g }1 = {g } (1.94)
tal que
g g = and a = g a (1.95)
Bajo una transformacion de Lorentz.
a a0 = a (1.96)
a a0 = a
La invarianza del producto escalar en ec. (1.93)
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 (1.98)
En notacion matricial
g = T g . (1.99)
From eq. (1.98) we also have
g g =g g
= , (1.100)
or
= . (1.101)
Since
1
= (1.102)
the inverse of is
1
= , (1.103)
or
1
= , (1.104)
1.4. NOTACION RELATIVISTA 19
donde
1
cosh = sinh = v, and = . (1.106)
1 v2
y, por ejemplo:
t + vx
t cosh + x sinh = (t + vx) = . (1.107)
1 v2
El definido en la ec. (1.105) satisface la condicion en ec. (??),
cosh sinh 0 0 1 0 0 0 cosh sinh 0 0
sinh cosh 0 0 0 1 0 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 (1.108)
20 CHAPTER 1. CLASSICAL FIELD THEORY
Con esta notacion, el producto escalar de cuadrivectores puede expresarse como el producto escalar
de los dos vectores de cuatro componente a y a .
E =m
p =mv . (1.113)
E 2 p2 = 2 m2 (1 v 2 ) = m2 . (1.114)
p2 = p p = m2 = E 2 p2 (1.115)
De [?]
The intuitive understanding of this equation is that the energy of a particle is partially
due to its motion and partially due to the intrinsic energy of its mass. The application
to particle detectors is that if you know the mass of a particular particle, or if its going
so fast that its energy and momentum are both huge so that the mass can be roughly
ignored, then knowing the energy tells you the momentum and vice versa
1.4. NOTACION RELATIVISTA 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)
Del calculo vectorial
= , , , = , , ,
x x0 x1 x2 x3 x0 x1 x2 x3
= , , ,
t x y z
=(0 , ) = ( 0 , ) (1.118)
= = , , , = (0 , ) (1.119)
x t x y z
Por consiguiente:
= (1.120)
x
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 = +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)
2 2 2 2
= 0 0 2 = (1.123)
t2 x2 y 2 z 2
Los operadores de energa y momentum de la mecanica cuantica tambien forman un cuadrivector
p = (p0 , p) = (H,
b p) (1.124)
con H,
b y p dados en la ec. (1.75). Entonces
p = i = i( 0 , i ) = i( , ) (1.125)
t
22 CHAPTER 1. CLASSICAL FIELD THEORY
= A0 , Ai 0 , i
A A0 =A . (1.131)
Note that the eq. (1.131) can be written as
A A0 = A (x) (1.132)
which is just the transformation obtained in eq. (1.88).
1.4. NOTACION RELATIVISTA 23
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
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
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
= , (1.138)
x0 x
or
1 1 1
= , (1.139)
x0 x
The field A (x) transforms simultaneously as field and as vector under Lorentz transformation
A (x) A0 (x0 ) = A (1 x) . (1.142)
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
+ 3 F (x)A (x)A (x) + 4 G (x)A (x)A , . (1.144)
Ejercicio: Show that terms like A (x) A (x), and hence F F , transforms as
A 1 x A 1 x
(1.145)
Hint: use the Lorentz transformation properties of in eq. (1.141).
In the case of J A :
J (x)A (x) g J 0 (x)A0 (x) =g J 1 x A 1 x
= g J 1 x A 1 x
=g J 1 x A 1 x ,
(1.146)
1.5. VECTOR FIELD LAGRANGIAN 25
= 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 transfor-
mation. 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)
(for K constant) are not Lorentz invariant:
K A (x)A (x)A (x) K A0 (x)A0 (x)A0 (x) =K A 1 x A 1 x A 1 x . (1.150)
K (x)A (x)A (x)A (x) is Lorentz covariant but not gauge-invariant (see below).
Bajo la transformacion gauge (1.132)
F F 0 =( 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 CHAPTER 1. CLASSICAL FIELD THEORY
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
L = F F J A (1.155)
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)
1.5. VECTOR FIELD LAGRANGIAN 27
where
A
E = . (1.158)
t
while for = l we have
F ml = m Al l Am
= (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
0 E 1 E 2 E 3
E1 0 213 B 3 312 B 2
{F } =
E 2 3
123 B 0 321 B 1
E3 132 B 2 231 B 1 0
0 E 1 E 2 E 3
E 1 0 B 3 B 2
=E 2
. (1.161)
B3 0 B 1
E3 B 2 B 1 0
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 (1.163)
( A )
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 = 12 ( F + F )
= 12 ( F + F ) intercambiando ndices mudos
= 12 ( F + F ) conmutando derivadas
= 12 ( F F ) usando antisimetra de F
= 0,
J = 0 . (1.166)
F 0 = J 0
i F i0 = J 0
i0
F = J0
xi
E i
= J0 , (1.167)
xi
and therefore
E = . (1.168)
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
i
= Jk
x t
B j E k
ijk i = Jk
x t
E k
( B)k = J k. . (1.169)
t
and therefore
E
B = J. (1.170)
t
In this way the expression
F = J where F = A A , (1.171)
T00 = F 0 (0 A ) L
1
= F 0 (0 A ) + F F + J A
4
1
T00 = F 0 (0 A ) + F F + J A
4
=0 =i
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 . (1.176)
4 4
=0 =j
0 1 0 1 i i0 1 i0 1 z 0i}| { 1 z ji}| {
F F0 + F F0 + F Fi = F Fi0 + F Fi0 + F F0i + F Fji
4 4 4 4 4
1 i0 1 i0 1
= F Fi0 + F Fi0 + F Fi0 + F ji Fji
i0
4 4 4
1 i0 1 ji
= F Fi0 + F Fji . (1.177)
2 4
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 CHAPTER 1. CLASSICAL FIELD THEORY
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, suma tambien sobre i, j
2 4
1 i i 1
= E E + ijk B k ijl B l + i (A0 E i ) J A, suma tambien sobre i, j
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 (1.179)
2 2
1 1
H = E2 + B2 + (A0 E) . (1.180)
2 2
1 1 2
L = F F = E B2
(1.181)
4 2
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
1
Z Z
3 2 2
H= d x (E + B ) + d3 x (A0 E)
2 V V
1
Z
= d3 x (E2 + B2 ), (1.182)
2 V
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
Ti0 = i A
(0 A )
= F 0 i A
= 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)
Z Z Z
i 3
P = d x Ti0 = 3 i
d x (E B) + d3 x (Ai E)
V ZV V
P= d3 x (E B) . (1.184)
V
1 X i 1
L(, , , , A ) = (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:
1 X i 1
L(, , , , A ) = (Di ) Di [ D0 (D0 ) ] F F . (1.186)
2m i 2 4
34 CHAPTER 1. CLASSICAL FIELD THEORY
1 X i 1
L= (i + iqAi ) (i + iqAi ) [ (0 + iqA0 ) (0 + iqA0 ) ] F F
2m i 2 4
1 X i 1
= (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
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 . (1.187)
2 4
1
i i iq Ai i + iqi Ai + q 2 Ai Ai
L=
2m
i 1
[ 0 (0 ) + 2iq A0 ] F F . (1.188)
2 4
From this we can obtain the Euler-Lagrange equation for each field.
1.6. SCRODINGER EQUATION IN PRESENCE OF THE ELECTROMAGNETIC FIELD 35
If we define
D iqA . (1.190)
we have in components:
D i = i iqAi
D i = i + iqAi . (1.191)
1
iD0 = D D
2m
1 2
iD0 = D , (1.192)
2m
que corresponde a la ecuacion de Scrodinger con la derivada normal reemplazada por la derivada
covariante.
36 CHAPTER 1. CLASSICAL FIELD THEORY
Ejemplo:
Demuestre que la ec. (1.196) es covariante:
Como
0 = ei(x) (1.197)
Entonces
D 0 0 = [( iqA) i] ei(x)
= i()ei(x) + ei(x) iqAei(x) i()ei(x)
= ei(x) ( iqA)
= ei(x) (D) (1.198)
y
2
D 0 0 = D 0 (D 0 0 )
= [( iqA) i] ei(x) (D)
= i()ei(x) (D) + ei(x) (D) iqAei(x) (D) iei(x) (D)
= ei(x) ( iqA)(D)
= ei(x) (D 2 ) (1.199)
De la misma manera
0
D0 0 = ei(x) (D0 ) (1.200)
De modo que
D D0 0 = ei(x) (D ) (1.201)
y la derivada covariante del campo transforma como el campo. Tenemos entonces que
0 1 02 0
iD0 0 = D
2m
1 i(x) 2
iei(x) D0 = e D
2m
1 2
iD0 = D (1.202)
2m
En resumen, para
D = + iqA (1.203)
38 CHAPTER 1. CLASSICAL FIELD THEORY
y reemplazando q tenemos
A A0 = A (x)
0 = eiq(x)
D D0 0 = eiq(x) (D ) . (1.204)
En esta convencion q corresponde al generador de la transformacion y al parametro de la transfor-
macion.
Entonces
F = j
con
(
q =0
j = iq i i
(1.208)
2m
[(D ) D ] = i
Que incluye el termino corriente para una partcula cargada y es diferente de la corriente de proba-
bilidad en ec. (1.57). En otras palabras es la carga electrica la que se converva localmente.
L L
J = +
(
L
+ L =0
= L0 0
L
. (1.209)
i
+ i =i
i i
J 0 = (iq) iq
2 2
=q , (1.210)
1
Ji = [(i iqAi ) iq iq (i + iqAi ) ]
2m
iq
Ji = [(Di ) (Di )] . (1.211)
2m
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 b = i qA .
p (1.212)
t
40 CHAPTER 1. CLASSICAL FIELD THEORY
A A0 = A + 0 = (1.213)
t
Ya que
A
E E0 = + = E (1.214)
t t t
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)
Producto de Grupo
U1 U2 = ei[(x1 )+(x2 )] ei(x3 ) U (1)
Identidad:
(x) = 0 tal que UI = 1
Inverso
(x) = (x) tal que U 1 = ei(x)
Note que si
i
A A0 = U A U 1 + ( U )U 1 (1.218)
q
1.8. PROCA EQUATION 41
y si es suficientemente pequeno
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)
A A + m2 A = J
A A + m2 A = J
m2 A = J (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 CHAPTER 1. CLASSICAL FIELD THEORY
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 , (1.225)
2 2
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
ultimo 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 au-
tomaticamente como la unica 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
1.9. KLEIN-GORDON EQUATION 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 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 trans-
formation.
44 CHAPTER 1. CLASSICAL FIELD THEORY
El campo puede pensarse como proveniente de una fuente de la misma manera como el campo
electromagnetico surge de partculas cargadas. Como en el caso del electromagnetismo, en esta
seccion podemos considerar los campos sin preocuparnos de las fuentes. En tal caso tendremos una
teora en la cual el campo escalar juega el papel de partcula mediadora de la interaccion.
Si el campo escalar se generaliza para que pueda tener otros numeros cuanticos, como carga
electrica, entonces estos pueden ser las fuentes de las respectivas cargas y corrientes en la ecuaciones
para campos vectoriales. Esto se estudiara en la seccion ??. En tal caso podramos tener por ejemplo
atomos formados de partculas escalares que se excitan emitiendo fotones.
La ecuaciones de Euler-Lagrange para V () = dan lugar a:
( + m2 ) = .
2
2 2
+ m = . (1.232)
t2
Con el cuadrivector (1.125) podemos construir la siguiente ecuacion
p p = m2
i i = m2
= m2
2
2 2
+ m = 0. (1.233)
t2
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 = (1.234)
L
T = L , (1.235)
( )
1.9. KLEIN-GORDON EQUATION 45
T00 = 0 0 L
=(x)0 (x) L
1 1 1
= 0 0 0 0 i i + m2 2
2 2 ! 2
1 X 1
= 0 0 + i i + m2 2
2 i
2
" #
2
1 1
H= + ()2 + m2 2 . (1.236)
2 t 2
donde
L
(x) =
(/t)
= . (1.237)
t
La densidad de momentum es
L
Ti0 = i
(0 )
Ti0 = 0 i
T0 = . (1.238)
t
1 1 1 1
L = [ 1 1 m2 21 ] + [ 2 2 m2 22 ] (1.239)
2 2 2 2
46 CHAPTER 1. CLASSICAL FIELD THEORY
Si definimos
1 + i2
= then (1.240)
2
1 i2
= , and (1.241)
2
2 =(1 + i2 )
2 =(1 i2 ). Therefore
2( + ) =21
2( ) =2i2 . Then
+
1 = (1.242)
2
2 = . (1.243)
2i
1 1
L = [ ( + ) ( + ) m2 ( + )2 ]
4 2
1 1
+ i2 [ ( ) ( ) m2 ( )2 ]
4 2
1
= [ + + 2 m2 (2 + 2 ) + 2 ]
4
1
[ + 2 m2 (2 + 2 ) 2 ]
4
1
= [4 4m2 ]
4
L = m2 (1.244)
( + m2 ) = 0, (1.245)
1.9. KLEIN-GORDON EQUATION 47
( + m2 ) = 0. (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 con-
servada puede definida en la seccion ??, eq. (1.20)
L L
J = +
( ) ( )
J = + . (1.247)
U = ei 1 + i.
Entonces
U
0 = ei (1 + i)
= + i. (1.248)
Entonces,
= i (1.249)
= i . (1.250)
does not left the Action invariant. To have a proper formulation of the quantum mechanics through
the general equation
i = H , (1.255)
t
where S() is some spinorial representation of the Lorentz Group. We will check in next section if
a Action with a term like
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
i = HS , (1.259)
t
1.11. DIRACS ACTION 49
where
HS = (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 , (1.261)
t
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
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)
which transforms as
0
(x) (x) = 0 (x)b = 1 x S ()b ,
(1.265)
and,
0
(x)(x) (x) 0 (x) = 1 x S ()bS() 1 x
(1.266)
50 CHAPTER 1. CLASSICAL FIELD THEORY
The condition that must be fulfilled for Lorentz invariance of the Action is
S ()bS() =b , (1.267)
and therefore,
0
(x)(x) (x) 0 (x) = 1 x 1 x ,
(1.268)
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
(x) (x) 0 (x) 0 (x) = a 1 x Sab 1 Scd ()d 1 x
()bc
= 1 x 1 S 1 () S() 1 x
S 1 () S() = . (1.271)
L = i m , (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 . (1.273)
2 2
1.11. DIRACS ACTION 51
b = b
b2 = I
b b = (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
0 = ei i, (1.276)
de modo que
= i. (1.277)
Por consiguiente
J 0 = 0 (1.278)
Para que J 0 pueda interpretarse como una densidad de probabilidad, se debe cumplir
b 0 = I (1.279)
52 CHAPTER 1. CLASSICAL FIELD THEORY
La densidad de corriente es
J 0 . (1.280)
Que podemos interpretar como una densidad de probabilidad.
De la ec. (1.279), ya que la inversa de es unica:
b = 0 . (1.281)
L L
T00 = 0 + 0 L
(0 ) 0
= i 0 0 L
= i i i + m,
= ( p + m),
= 0 ( p + m),
= H, (1.286)
1.11. DIRACS ACTION 53
donde
H = 0 ( p + m) (1.287)
la ecuacion de Scrondinger de validez general es entonces:
i = H (1.288)
t
y, como en mecanica clasica usual Z
hHi = H d3 x. (1.289)
Ademas
L L
Ti0 = i + i
(0 ) 0
= i 0 i
= (ii ) (1.290)
de modo que Z
hpi = p d3 x (1.291)
Expandiendo
i 0 0 + i i i m = 0
i 0 0 (i) m = 0,
i 0 0 = ( p + m),
de donde
2
i 0 = 0 ( p + m). (1.294)
t
tenemos que
2
0 = 1. (1.295)
De la ec. (1.287)
H = 0 ( p + m), (1.296)
A este punto, solo nos queda por determinar los parametros .
La ec. (1.292) puede escribirse como
i H = 0. (1.297)
t
El campo tambien debe satisfacer la ecuacion de Klein-Gordon. Podemos derivar dicha ecuacion
aplicando el operador
i H
t
De modo que, teniendo en cuenta que H/t = 0,
i H i H = 0
t t
i H i H = 0
t t
!
2 H
+i + iH iH + H 2 = 0
t2 t t t
2
2
+ H = 0. (1.298)
t2
1.11. DIRACS ACTION 55
H 2 = (0 p + 0 m)(0 p + 0 m)
= (0 p)(0 p) + m0 p0 + m02 p + m2 (1.299)
Sea
= 0
i = i
i = i (1.300)
H 2 = ( p)( p) + m p + m p + m2
= ( p)( p) + m( + ) p + m2 (1.301)
2 2
X X
(A )2 = Ai i + Ai , Aj i j
(1.302)
i i<j
56 CHAPTER 1. CLASSICAL FIELD THEORY
Demostracion
XX
[(A )] = Ai i Aj j
ij
X X
= i j
Ai Aj
ij
XX
= i j Ai Aj
ij
!
2
X X X X
= i Ai Ai + i j Ai Aj + i j Ai Aj
i i<j i>j
!
i2
X X X X
= Ai Ai + i j
Ai Aj + j i
Aj Ai
i i<j j>i
" #
i2
X X X
Ai Ai + i j Ai Aj + Aj Ai
=
i i<j
" #
2
X X
i Ai Ai + i j Ai , Aj
=
i i<j
" #
i2 i2
X X
i j
Ai , Aj
= A + . (1.303)
i i<j
Entonces
X
H 2 =i2 p2i + {i , j } pi pj + m(i + i )pi + m2 (1.304)
i<j
De modo que
i , j = 0 i 0 j + 0 j 0 i = 0
i 6= j
0 0 i j 0 0 j i = 0 i 6= j
i j j i
+ =0 i 6= j
i j
, = 0 i 6= j (1.309)
{ , } + = 2g 1 (1.310)
donde
= ( 0 , i ) (1.311)
Definiendo
5 = i0 1 2 3 , (1.317)
entonces,
52 = 0 1 2 3 0 1 2 3
52 = + 02 1 2 3 1 2 3
52 = + 1 2 3 1 2 3
52 = 2 3 2 3
52 =2 2 3 3
52 =1 . (1.318)
52 = 1, (1.319)
Teniendo en cuenta que 2 1 y conmuta con las demas matrices, tenemos por ejemplo
5 3 =i0 1 2 32 = 32 i0 1 2 = 3 i0 1 2 3 = 3 5
5 2 = i0 1 22 3 = 22 i0 1 3 = 2 i0 1 2 3 = 2 5
5 1 =i0 12 2 3 = 12 i0 2 3 = 1 i0 1 2 3 = 1 5
5 0 =i0 1 2 3 0 = 02 i1 2 3 = 0 5 . (1.320)
De modo que
{ , 5 } = 0. (1.321)
Expandiendo el anticonmutador tenemos
5 = 5
5 5 =
Tr (5 5 ) = Tr
Tr (5 5 ) = Tr
Tr = Tr , (1.322)
y por consiguiente
Tr = 0. (1.323)
De otro lado, si
U U , (1.324)
1.11. DIRACS ACTION 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)
L = i 0 0 0 + i 0 i i = i 0 + i i i , (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.
Como i = 0 i 0 = i = i , podemos definir la representacion de paridad
0 = 0 , i = i , para U = 0 (1.328)
The Pauli matrices are set of matrices satisfying this commutation relations:
i j
k
, = i ijk (1.330)
2 2 2
60 CHAPTER 1. CLASSICAL FIELD THEORY
donde i
0 1 0 i 1 0
1 = 2 = 3 = (1.331)
1 0 i 0 0 1
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
i
Ti = , (1.332)
2
un elemento del Grupo puede escribirse como
i i
U = eiT i
1 + iT i i = 1 + i i . (1.333)
2
Como antes, i es el parametro de la transformacion.
Las matrices de Pauli y por consiguiente Ti satisfacen
i = i
Tr (i ) = 0 (1.334)
Ademas
det (i ) = 1
{i , j } = 2ij I i2 = I
Tr i j = 2 ij
i j = iijk k + ij (1.335)
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 commu-
tation relations by writing the generators as differential operators; from the expression
J = x p = x (i) , (1.336)
Involving three generators. The generalization to four-dimensions give to arise three further genera-
tors J 0i :
J = i(x x ) . (1.339)
[J , J ] =i(g J g J g J + g J ) . (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 (1.341)
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
0 x+vt
x
x0 = 1v2 = sinh cosh 0 0 x = { } {x } ,
{x } =
y y y 0 (1.342)
0 1 0 y
0
z z z 0 0 0 1 z
Since
X 2n
cosh = 1 + O( 2 )
n=0
2n!
X 2n+1
sinh = + O( 2 ) , (1.343)
n=0
(2n + 1)!
62 CHAPTER 1. CLASSICAL FIELD THEORY
1
= 1 + i bi + i ijk rjk , (1.349)
2
bi = iJ i0 rjk = iJ jk . (1.350)
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 . (1.352)
2
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 , (1.354)
2
which can be written in covariant form if we define
i
S = [ , ] . (1.355)
4
In fact, the six set of non-zero independently generators are
i 0 i i 0 i
S 0i = i 0 = = iB i
4 2
i i i j
ij
S = i j j i = = iRij .
(1.356)
4 2
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 CHAPTER 1. CLASSICAL FIELD THEORY
or equivalently
S () 0 = 0 S 1 ()
i 0 0 i
1 + S = 1 + S
2 2
0 0
S = S . (1.359)
we have
i 0
S 0 = [ , ]
4
i
= , 0
4
i 0
= ,
4
i
= [ , ] 0
4
= 0 S
(1.361)
(x)(x) , (1.362)
where is a 44 complex matrix. The most general such matrix can always be expanded
in terms of 16 independent 4 4 matrices multiplied by complex coefficients. In short
1.11. DIRACS ACTION 65
En la Tabla 1.1 se muestran las matrices de traza nula con sus propiedades de transformacion
bajo el Grupo de Lorentz. En la ultima se muestra el correspondiente escalar en el espacio de Dirac
. Demostracion
i
5 = (1.365)
24
and
det = 1 2 3 4 . (1.366)
66 CHAPTER 1. CLASSICAL FIELD THEORY
0 = ei(x)Q
0 = ei(x)Q , (1.367)
D = ieQA , (1.368)
donde e es la carga electrica del electron. De esta forma, si e es el campo que representa al electron
L = [i ( ieQA ) m] 14 F F
= (i m) + eQ A 14 F F . (1.371)
Este Lagrangiano da lugar a la Accion de la teora conocida como Electrodinamica Cuantica (QED
de sus siglas en ingles).
Aplicando las ecuaciones de Euler-Lagrange para , tenemos
(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
eQ =0
4 ( A ) A
F = eQ (1.373)
j = eQ . (1.374)
Lf ree = (i m) 14 F F
Lint =eQ A . (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 unico con el foton en horizontal como se muestra
en la Figura (c).
68 CHAPTER 1. CLASSICAL FIELD THEORY
Figure 1.3: Electromagnetic repulsion. The diagrams (a) and (b) are summarized in the diagram (c)
1.13. CROMODINAMICA CUANTICA 69
Los protones, neutrones, piones, kaones y demas hadrones, son partculas compuestas de consti-
tuyentes elementales llamados quarks. Por ejemplo los protones, neutrones y piones estan constitu-
idos de quarks up y down. Los hadrones 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 confi-
namiento 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 ultimos 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 1
B = |q q q i M = |q q i (1.379)
6 3
Estos estados son singletes de color. Una de las determinaciones de Nc proviene del observable
(e+ e hadrones)
R (1.380)
(e+ e + )
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
Pfmax PNc +
f =u =1 (e e f f )
R
(e+ e + )
Pfmax +
f =u (e e f f )
R Nc (1.381)
(e+ e + )
70 CHAPTER 1. CLASSICAL FIELD THEORY
De este modo R esta dado por la suma de las cargas electricas al cuadrado
P 2
f Qf
R Nc
Q2
fmax
X
=Nc Q2f
f =u
2 2 1 2 2
Nc [( 3 ) + 2( 3 ) ] = 3 Nc
f = u, d, s, fmax = s
= Nc [2( 32 )2 + 2( 1
3
)2 ] = 109
Nc fmax = c
1 2
2 2 11
Nc [2( 3 ) + 3( 3 ) ] = 9 Nc fmax = b
2 Nc = 3,
fmax = s
10
= 3 Nc = 3, fmax = c (1.382)
11
3
Nc = 3, fmax = b
En la figura, tomada de [?], se muestra el grafico de R con respecto a s(la energa de centro de masa
que va hasta una energa s 4 GeV que corresponden
de la colision). Se observan dos escalones, uno
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 senalan 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
= b
= qb .
(1.383)
g qg
El Lagrangiano de Dirac con invarianza gauge global SU (3), para un quark, se puede escribir como
Lglobal = i m, (1.384)
donde
a
0
= exp ia . (1.385)
2
a = 1, . . . , 8, a /2 son los ocho generadores de SU (3) y a son los parametros de la transformacion
global. Los generadores de SU (3)
a
a , (1.386)
2
1.13. CROMODINAMICA CUANTICA 71
satisfacen el algebra
a b c
, = if abc , (1.387)
2 2 2
donde f abc son las constantes de estructura fina de SU (3).
En un analisis similar al de la seccion 1.11.8 tenemos que la Accion invariante gauge local bajo
SU (3)c , se obtiene de reemplazar la derivada normal por la derivada covariante
1
Llocal = i D m Tr (G G ) , (1.388)
2
donde
0 = U (x)
D (D )0 = U (x)D , (1.389)
con la matriz 3 3
a
U (x) = exp ia (x) , (1.390)
2
72 CHAPTER 1. CLASSICAL FIELD THEORY
y
a a
D = igsG igs G (1.391)
2
donde hemos definido la matriz 3 3 G , como
a
(G ) = Ga (1.392)
2
D (D )0 =D0 0 = U (x)D
D0 U = U (x)D . (1.393)
Por consiguiente
D0 U = U D (1.394)
D (D )0 = U D U 1 (1.395)
Desarrollando a ambos lados
D (D )0 = D0 0 =D0 0
( igs G0 )U =U D U 1 U
( igs G0 )U =U ( igs G )
U + ( U ) igs G0 U =U igs U G
( U ) igs G0 U = igs U G
igs G0 U = ( U ) igs U G , (1.396)
de modo que
1
G0 U = ( U ) + U G
igs
i
G0 = ( U )U 1 + U G U 1 . (1.397)
gs
Como U es unitaria, la transformacion de los campos gauge puede escribirse como
i
G (G )0 = U G U 1 ( U ) U . (1.398)
gs
1.13. CROMODINAMICA CUANTICA 73
Entonces
i
a G0 a (1 + ib b )c Gc (1 id d ) [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
a 1 acb
= Ga + a + f Gc b (1.399)
gs
de donde
1
Ga G0 a Ga + a + f abc Gb c (1.400)
gs
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,
i a
G = [D , D ] G , (1.401)
gs 2 a
tenemos
i
G = [ 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)
74 CHAPTER 1. CLASSICAL FIELD THEORY
De modo que
G = G G igs [G , G ] , (1.403)
que se reduce al caso Abeliano cuando los bosones gauge conmutan. En terminos de componentes
b
a G a a c
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
a = Ga Ga + gs f Gb Gc Ga b c (1.405)
con
e = G G
G (1.406)
a a a
i 0 0
G G0 = 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 75
Tr a b =2 ab
1
Tr a b = ab , (1.410)
2
tenemos (suma sobre indices repetidos de SU (3))
Tr (G G ) Tr G 0 G 0 = Tr a G b b
a G
= Tr a b G b
a G
1
= ab G
a G
b
2
1
= G Ga . (1.411)
2 a
Expandiendo el Lagrangiano en ec. (1.388), tenemos
a a 1
L =i igs G m Tr (G G )
2 2
a 1
=i igs Ga m G Ga
2 4 a
a 1
=i m + gs Ga G Ga
2 4 a
a 1 e ea
=i m + gs Ga G G
2 4 a
1 e
gs Ga fade Gd Ge + gs f abc Gb Gc G
ea + gs2 f abc fade G Gc Gd Ge
b
4
=Lfree + Lgauge + LSI , (1.412)
donde
Lfree =i m
a 1 e ea
Lgauge =gs Ga G G
2 4 a
1
d e abc e a 2 abc d e
LSI = gs Ga fade G G + gs f Gb Gc G + gs f fade Gb Gc G G .
e (1.413)
4
Hemos divido el Lagrangiano en tres partes
76 CHAPTER 1. CLASSICAL FIELD THEORY
a gs g2
Lint = gs Ga f abc ( Ga Ga ) Gb Gc s f abc fade Gb Gc Gd Ge . (1.417)
2 2 4
From [18] (pag 136):
The quarks have an additional type of polarization that is not related to geometry. The
idiot physicists, unable to come up with any wonderful Greek words anymore, call this
type of polarization by the unfortunate name of color, which has nothing to do with
color in the nornal sense. At a particular time, a quark can be in one of three conditions,
or colorsR, G, or B (can you guess what they stand for?). A quarks color can be
changed when the quark emits or absorbs a gluon. The gluons come in eigth diffent types,
according to the colors they can couple with. For example, if a red quark changes to
green, it emits a red-antigreen gluona gluon that takes the red from quark and gives
it green (antigreen means the gluon is carrying green in the opposite direction). This
gluon could be absorved by a green quark, which changes to red (see Fig. 1.5). There are
eigth different possible gluons, such as red-antired, red-antiblue, red-antigreen, and so on
1.13. CROMODINAMICA CUANTICA 77
u d
green red
red
t
green
red green
u 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 unica 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
rr rb rg
br bb bg , with rr + bb + gg = 0
gr g b gg
78 CHAPTER 1. CLASSICAL FIELD THEORY
Figure 1.6: Triplegluon selfinteraction. The anticolors are the colors running back in time.
como
a
Ja
= gs q q . (1.418)
2
Note que tanto para la Electrodinamica Cuantica como para la Cromodinamica Cuantica la corriente
es vectorial. Para las interacciones debiles la estructura es mas complicada y requiere un
conocimiento mas profundo de la ecuacion de Dirac y sus soluciones.
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
G G gs
d d
= Ga gs fdbc Gb Gc
e
a
2 ( G ) 2
gs dbc e g s
+ f Gd ( ba Gc + Gb ca ) + f ibc fide (g g Gb Gc Gd Ge )
2
4
= G 1
ea gs fdbc G Gc ( da da ) gs a
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
= G ea gs f G G + gs fabc G G gs a
1 abc 1
b c b c
2 2 2
gs gs adb e b
f adc G e c
d G f Gd 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
+ f fida g g G G G . (1.420)
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
= f fdje Gj Gc G + f fdje Ge Gc G + f fdae Gb Gc G + f dbc fdea Gc Gb Ge
e j e
4 4 4 4
gs dac gs dca gs gs
= f fdje Gj Gc G + f fdje Ge Gc G + fdac f Gj Gce G + fdca f dje Ge Gj Gc
e j dje c
4 4 4 4
gs abc b gs abc b gs gs
= f G fcej Ge Gj f G fcej Ge Gj fabc G f Ge Gj fabc Gb f cej Ge Gj
b cej
4 4 4 4
b cej
= gs fabc G f Ge Gj (1.421)
80 CHAPTER 1. CLASSICAL FIELD THEORY
Entonces
L L
( Ga ) Ga
a
e abc e gs facd Gc f dej Ge Gj
= G a g s f Gb c gs
G gs f acd Gc Gd
2
acd c a
= G
a gs f G Gd gs = 0 . (1.422)
2
Entonces las Ecuaciones de Euler Lagrange para Ga , son
acd c a
G
a + gs f G Gd = gs . (1.423)
2
Definiendo
a
Ja = gs , (1.424)
2
La ec.(1.423) puede reescribirse como:
a
G
a = gs fabc Gb G
c + (1.425)
2
y usando el hecho que = :
G e + gs (fabc G G )
a = G b c
1
=0 + [gs (fabc Gb Gb ) + gs (fabc Gb Gc )]
2
1
= [gs (fabc Gb Gb ) + gs (fabc Gb Gc )]
2
1
= [gs (fabc Gb Gb ) + gs (facb Gc Gb )]
2
1
= [gs (fabc Gb Gb ) gs (fabc Gb Gc )]
2
=0 , (1.426)
como en el caso Abeliano, tenemos la corriente conservada
j = 0 , (1.427)
donde
a
ja = gs fabc Gb G
c + . (1.428)
2
El primer termino corresponde a las autointeracciones y el segundo a la corriente de color generada
por los quarks.
1.13. CROMODINAMICA CUANTICA 81
si G es un vector en el espacio SU (3) con las 8 componentes Ga , entonces podemos escribir (1.400)
como
1
G G0 = G + + G . (1.430)
gs
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)
G + gs G G = J (1.432)
con
a , b = ifabc c .
(1.435)
D = igs a Ga (1.436)
82 CHAPTER 1. CLASSICAL FIELD THEORY
En componentes
(D )ab G c
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
Ja = gs . (1.441)
2
Para escribir el Lagrangiano en forma vectorial en el espacio SU (3), debemos reescribir la trans-
formacion 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)
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
A (B C) = B (C A) = C (A B) , (1.445)
G a
a G = G
G G0 G0 G G + G (G ) + (G ) G
=G G + G ( G ) + (G ) G
=G G (G ) G + (G ) G
=G G . (1.446)
El Lagrangiano para los campos gauge, el cual puede generalizarse para cualquier teora SU (N ), es
1
Lgluon = Lgauge + LSI = G G J G , (1.448)
4
Da lugar la ecuaciones de Maxwell pero con la derivada normal reemplzada por la derivada covariante
D G =J , (1.449)
donde
D G = G + gs G G . (1.450)
Note que en el caso Abeliano, U (1), la derivada covariante del tensor de campo se reduce a la derivada
normal de dicho tensor. El termino extra en la derivada covariante da lugar a las autointeracciones
de los campos gauge.
Ejercicio:
Muestre que la derivada covariante de G , transforma como G .
84 CHAPTER 1. CLASSICAL FIELD THEORY
L = 21 V () (1.451)
con
V () = 12 2 2 . (1.452)
Este Lagrangiano es simetrico bajo la transformacion discreta .
Cuando 2 > 0, el campo tiene excitaciones alrededor del mnimo del potencial que cuestan
energa y dicho termino se interpreta como la masa de la partcula. Ver 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
que mantiene la simetra bajo la transformacion discreta . > 0 garantiza la aparicion de los
dos mnimos que se muestran el la figura 1.8. Si la energa es suficientemente alta como se muestra
en la figura 1.8, las excitaciones son simetricas con respecto al maximo del potencial y el termino en
2 no puede interpretarse como masa para la partcula escalar.
Sin embargo, si la energa es suficientemente baja como se muestra en la figura 1.9, las excitaciones
alrededor del mnimo dan lugar a la aparicion de un termino de masa para el campo escalar. Ademas,
dichas excitaciones no respetan la simetras . En tal caso decimos que la simetra ha sido
espontaneamente rota: aunque el Lagrangiano mantiene la simetra original, el vaco la rompe.
Para analizar cuantitativamente el espectro de partculas es necesario expandir el campo alrededor
del mnimo y determinar las excitaciones. Establezcamos en primer lugar los mnimos del potencial.
La V / = 0 da lugar a
2 + 3 = 0 (1.455)
(2 + 2 ) = 0, (1.456)
con extremos max = 0, y r
2
min hi v = . (1.457)
De hecho
2V
2
= 2 + 32 . (1.458)
p
= 0 corresponde a un maximo, mientras que la segunda derivada para = 2 / es 22 > 0
y corresponden a los mnimos. Expandiendo el campo alrededor del mnimo
(x) = H(x) + v (1.459)
86 CHAPTER 1. CLASSICAL FIELD THEORY
V () = 12 2 2 + 41 4
= 12 2 (H + v)2 + 41 (H + v)4
= 21 2 (H + v)2 + 41 (H + v)4
2
= 21 2 H 2 + 2vH + v 2 + 41 H 2 + 2vH + v 2
h 2 i
= 21 2 H 2 + 2vH + v 2 + 41 H 4 + 2H 2 2vH + v 2 + 2vH + v 2
= 21 2 H 2 + 2vH + v 2 + 41 H 4 + 4vH 3 + 2H 2 v 2 + 4v 2 H 2 + 4v 3 H + v 4
= 21 2 H 2 + 2vH + v 2 + 41 H 4 + 4vH 3 + 6H 2 v 2 + 4v 3 H + v 4
= 21 2 H 2 23 H 2 2 + 2 vH 2 vH + 21 2 v 2 41 2 v 2 + 14 H 4 + 4vH 3
V (H) = 12 22 H 2 + vH 3 + 41 H 4 + 14 2 v 2 ,
(1.460)
y
LH = 21 H H 1
22 H 2 vH 3 14 H 4 + constant.
2
(1.461)
V () = 21 v 2 2 + 41 4
= 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)
1.14. SPONTANEOUS SYMMETRY BREAKING 87
1 1 m2H 3 1 m2H 4
V (H) = m2H H 2 + H + H . (1.463)
2 2 v 8 v2
donde
m2H = 2 2 = 2v 2 (1.464)
Consideremos ahora un campo escalar complejo sin termino de masa, pero con potencial:
L = V () (1.465)
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 CHAPTER 1. CLASSICAL FIELD THEORY
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 :
1 1 1 2
L = i i 2 i i 2 i i , (1.467)
2 2 2
que es invariante bajo una simetra O(N )
i
i 0 = Rij j , (1.468)
[. . . ] there are not continuous symmetries for N = 1, while for N = 2 there is a single
direction of rotation. A rotation in N dimensions can be in any one of N (N 1) planes, so
the O(N )symmetric theory has N (N 1)/2 continuous symmetries. After spontaneous
symmetry breaking there are (N 1)(N 2)/2 remaining symmetries corresponding to
rotations of the (N 1) [non massive] fields. The number of broken symmetries is the
difference, N 1.
Goldstones theorem states that for every spontaneously broken continuous symmetry, the
theory must contain a massless particle.
In a global symmetry that is spontaneously broken the symmetry currents are still con-
served and interactions are similarly restricted [the Lagrangian keeps the symmetry], but
the vacuum state does not respect the symmetry and the particles do not form obvi-
ous symmetry multiplets. Instead, such a theory contains massless particles, Goldstone
bosons, one for each generator of the spontaneously broken symmetry. The third case
is that of a local, or gauge, symmetry. [. . . ] such a symmetry requires the existence of
a massless vector field for each symmetry generator, and the interactions among these
fields are highly restricted.
It is now only natural to consider a fourth possibility: What happens if we include both
local gauge invariance and spontaneous symmetry breaking in the same theory?
3
Sec. 11.1
4
Introduction to Chapter 20
1.14. SPONTANEOUS SYMMETRY BREAKING 89
En el caso de la Accion invariante gauge local bajo el Grupo U (1), tenemos el Lagrangiano (??):
2
L L0 = (D )0 0 (D )0 0 2 ( )0 0 ( )0 0 41 (F F )0
(1.471)
En adelante omitiremos las primas, aunque debe estar claro que se esta trabajando en el gauge
especfico de la ec. (1.470). Entonces
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 numero 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 CHAPTER 1. CLASSICAL FIELD THEORY
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
L PL
R PR . (1.477)
1.15. FERMIONES QUIRALES DE CUATRO COMPONENTES 91
Entonces
= L + R . (1.478)
Las matrices PL,R tienen las propiedades
2
PL + P R = 1 PL,R = PL,R PL,R = PL,R
PL P R = 0 PL,R = PL,R . (1.479)
Usando la ec. (1.321)
1 5 1 5
PL,R = = = PR,L (1.480)
2 2
Para escribir el Lagrangiano en termino de los nuevos L,R debemos tener en cuenta que
L,R = (PL,R ) 0 = PL,R 0 = 0 PR,L = PR,L (1.481)
L =i m
=i(PL + PR ) m(PL + PR )
=iPL + iPR mPL mPR
=iPL PL + iPR PR mPL PL mPR PR
=iPL PR + iPR PL mPL PL mPR PR
=iR R + iL L m(R L + L R ) . (1.482)
En terminos de espinores izquierdos y derechos de cuatro componentes la transformacion de paridad
tt x x L (t, x) R (t, x), R (t, x) L (t, x)
0 0 L (t, x) R (t, x), R (t, x) L (t, x) . (1.483)
Ademas L = r p (r) (p) = L, y como esta asociado al momento angular intrnsico,
entonces tambien
Entonces la transformacion de paridad da lugar a (sin tener en cuenta el cambio de argumento
en los campos que desaparece en la integral de la Accion)
R R = R 0 0 R + R R L 0 0 L L L
=L 0 0 L + L L
=L 0 0 0 0 L + L 0 0 L
=L 0 0 L + L L
=R R . (1.484)
92 CHAPTER 1. CLASSICAL FIELD THEORY
Entonces
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 electro-
magneticas ni fuertes, es decir, no se acopla con el foton o los gluones. Ademas dicha partcula
no puede tener masa de Dirac. En el caso del neutrino esto se entiende pues al no tener carga
electrica 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)
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
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,
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
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
YL =YL = YeL
YQ =YuL = YdL . (1.493)
El generador de carga electrica Q,b se va obtener a partir de una combinacion lineal del generador
diagonal de SU (2)L , T3 , y del generador de hipercarga, Y .
Bajo la simetra SU (2)L , los campos transforman como:
L L0 = exp(iT i i )L (1 + iT i i )L
Q Q0 = exp(iT i i )Q (1 + iT i i )Q
eR e0R = eR
uR u0R = uR
dR d0R = dR . (1.494)
donde
i
Ti = , (1.495)
2
y i son las matrices de Pauli dadas en la ec. (1.331).
Claramente el termino de masa me en la ec. (??) no es invariante bajo la simetra SU (2)L . El
Lagrangiano en la ec. (??), sin termino de masa, puede reescribirse de manera que exhba de forma
mas explicita la invarianza bajo SU (2)L como
+ (D ) D 2 ( )2
(he LeR + hd QdR + hu Qu e R + h.c)
=Lfermion + Lgauge + LW BH LYukawa . (1.500)
donde 2 < 0, y > 0,
e = i2 ,
(1.501)
LW BH =(D ) D 2 ( )2
LYukawa =he LeR + hd QdR + hu Que R + h.c (1.502)
96 CHAPTER 1. CLASSICAL FIELD THEORY
Bajo una transformacion gauge local las derivadas covariantes de los campos (y por consiguiente
los campos) transforman como:
D (D )0 = exp ii T i iY D
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
YL + Y + QeR =0
YQ + Y + QdR =0
YQ + Ye + YuR = YQ Y + QuR =0 ,
From this set of three equations we obtain the three doublet hypercharges:
1 1 1
YL = , Y = , YQ = . (1.504)
2 2 6
En el analisis anterior hemos fijado Ye = Y . Esto es debido a que si Q es un invariante SU (2)L ,
el termino Q tambien es un invariante de SU (2). Explcitamente
e Q =(i2 ) Q
0
= Q
0 +
uL
=
dL
=0 uL + dL
=12 Q1 2 + 21 Q2 1
=ab Qa b . (1.505)
1.16. STANDARD MODEL LAGRANGIAN 97
Sin perdida de generalidad los cuatro grados de libertad de , pueden escribirse en la forma
ij (x)T j 0
=e 1 [H(x) + v]
. (1.507)
2
V () = 2 + ( )2 (1.508)
se reduce a
1 1
V (H) = 2 (H + v)2 + (H + v)4 . (1.509)
2 4
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 CHAPTER 1. CLASSICAL FIELD THEORY
de manera que quedara un generador no roto correspondiente a una simetra remanente del vaco
U (1)Q
hi
SU (2)L U (1)Y U (1)Q . (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 , (1.512)
2
(H(x) + v)
que define el gauge unitario. En adelante sin embargo omitiremos las primas sobre los campos
transformados 0 y W
0
.
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)
2 H(x) + v H(x) + v
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
10 1 1 0 i 1 1 0
W = Ti Wi = 1
W + 2
W + W3
21 0 2 i 0 2 0 1
W1 i W2
3
W 2
1 2
=
2 W + i W
1 2
3
2 W
2
3
2W+
1 W
2 2W W3
!
1 1 W +
W3
= 12 2
1 3 . (1.515)
2
W 2
W
D corresponde a la matrix 2 2
!
1
gW3 + g 0 Y i2 gW+
i 2
B
D = . (1.516)
i2 gW i 12 gW3 + g 0 Y B
1.16. STANDARD MODEL LAGRANGIAN 99
Entonces
i2 gW+ (H + v)
D = . (1.517)
H i 21 gW3 + g 0 Y B (H + v)
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 1 0
= gW (H + v) H + i 2
gW 3 + g Y B (H + v)
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
1 1 2
+ H H + 21 gW3 + g 0 Y B (H + v)2 (1.518)
2 2
donde la ultima lnea corresponde a la magnitud del numero complejo:
H i 21 gW3 + g 0 Y B (H + v)
(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 , (1.520)
4 4 2
donde
1 1 2 3 1 0 1 0 02 2
LZAH = g W 3 W gg Y W
3 B gg Y W
3 B + g Y B B
2 4 2 2
H 2 + 2vH + v 2
(1.521)
100 CHAPTER 1. CLASSICAL FIELD THEORY
g 2 gg 0
3
1 W 2 2
LZAH = W3 B 2 H + 2vH + v (1.522)
8 gg 0 g 0 B
Sea
g g0
cos W sin W 1
V = =p , (1.523)
sin W cos W g2 + g02 g 0 g
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. (??)),
3
W3 Z Z T W
=V o =V (1.524)
B A A B
entonces
gg 0
2 3
1 3
T g T W 2 2
LZAH = W B VV 2 V V H + 2vH + v
8 gg 0 g 0 B
0
2
1 g gg Z
= Z A V T H 2 + 2vH + v 2
0 02 V (1.525)
8 gg g A
g 2 gg 0 g + gg 0 2 g 2 g 0 g 0 3 g g0
3
T 1
V V = 2
gg 0 g 0 2 2
g + g 0 2 +g 2 g 0 g 2 g 0 gg 0 + gg 0
2
g 0 g
g + gg 0 2 g 2 g 0 g 0 3 g g0
3
1
= 2
g + g02 0 0 g 0 g
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
4
1
= 2
g + g02 0 0
02
2
g +g 0
= (1.526)
0 0
1.16. STANDARD MODEL LAGRANGIAN 101
1 g2 + g02
Z Z H 2 + 2vH + v 2
LZAH =
2 4
1 g
2
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 (1.527)
2 cos W
Retornando a la ec. (1.520), tenemos tenemos
2
LW BH = (D ) D 2
1
= H H V (H)
2
1 1
+ g 2 W W+ H 2 + vg 2 W W+ H
4 2
2 2
1 g 2 g
+ 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 , (1.528)
2 2 2
donde:
Masas gauge:
gv gv
mW = mZ = , (1.529)
2 2 cos W
y
mW
mZ = . (1.530)
cos W
V (H) = 21 m2H H 2 + vH 3 + 41 H 4
1 m2 1 m2H 4
= m2H H 2 + H H 3 + 2
H
2 2v 4 2v
H2
1 H
= m2H H 2 1 + + 2 . (1.531)
2 v 4v
102 CHAPTER 1. CLASSICAL FIELD THEORY
con
m2H = 22 = 2v 2 . (1.532)
3
W cos W sin W Z
= , (1.533)
B sin W cos W A
tal que
g sin W = g 0 cos W . (1.534)
definiendo
hf v
mf = (1.536)
2
tenemos
me md mu
LYukawa =me ee + md dd + mu uu + eeH + ddH + uuH . (1.537)
v v v
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 (gT1 W1 + gT2 W2 + gT3 W3 + g 0 YL B )L
0 W+
g 3 0
=L + gT3 W + g YL B L
2 W 0
0 W+
g 3 0
=iL L + L gT3 W + g YL B L
2 W 0
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
(1.540)
Generalizando para todos los campos:
g
LW L L eL W+ + uL dL W+ + h.c .
(1.541)
2
Usando la ec. (1.533)
LAZL =L [gT3 (cW Z + sW A ) + g 0 YL (sW Z + cW A )] L
=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 , (1.542)
donde cW = cos W , sW = sin W . Usando la relacion entre g y g 0 (1.534):
s2W
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 . (1.543)
Como el generador asociado a A debe ser el generador de carga electrica, tenemos que
e = g sin W (1.544)
104 CHAPTER 1. CLASSICAL FIELD THEORY
Usando los acoplamientos gauge de los quarks con los gluones (1.413), de los fermiones con el
W (1.541) y con Z y A (1.548) para expandir Lfermion en (1.538), tenemos
Lfermion =iQ D Q + iL D L + ieR D eR + idR D dR + iuR D uR
=iuL uL + iuR uR + idL dL + idR dR
+ ieL eL + ieR eR + iL L
a a a a
+ gs uL uL + uR uR + dL dL + dR dR Ga
2 2 2 2
g
+ L eL W+ + uL dL W+ + h.c
2
X e h i
+ F 3 2s2W Q b L F Z
F =Q,L,eR ,dR ,uR
2cW sW
+ e eL Q be eL + eR Q
b e eR
+uL Qbu uL + uR Q b u u R + dL Q
b d dL + dR Q
bd dR A . (1.549)
Para escribir este Lagrangiano en terminos de espinores de 4 componentes, tomemos algunos
casos especficos:
1.16. STANDARD MODEL LAGRANGIAN 105
u d e e
2vf 1 sin2 W
8
3
1 + sin2 W
4
3
1 1 + 4 sin2 W
2af 1 1 1 1
Table 1.3: Acoplamientos de corrientes neutras
L eL W+ = PR PL eW+
= PL2 eW+
= PL eW+
1
= (1 5 )eW+ , (1.550)
2
h i
2 b 2 b 2 b
Q 3 2sW QQ Q 2sW uR Qu uR 2sW dR Qd dR Z
" !
1 2s2W Q bu 0 uL
= u L dL
0 2 b
1 2sW Qd dL
i
2s2W uR Q
bu uR 2s2 dR Q
W
bd dR Z
donde
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 CHAPTER 1. CLASSICAL FIELD THEORY
Lfermion =iu u + id d + ie e + iL L
a a
+ gs u u + d d+ Ga
2 2
g X e
+ (1 5 )eW+ + u (1 5 )dW+ + h.c + f (vf af 5 ) f
2 2 f =u,d,,e
2c s
W W
+ e e Q be e + u Qu u + d Qd d A
X X a
= if f + gs q qGa
f =u,d,,e q=u,d
2
g
+ (1 5 )eW+ + u (1 5 )dW+ + h.c
2 2
X e
+ f (vf af 5 ) f
f =u,d,,e
2c s
W W
X
+e f Qbf f A , (1.553)
f =u,d,,e
1.16.4 Self-interactions
El Lagrangiano gauge
Lgauge = 14 G a 1 i 1
a G 4 Wi W 4 B B
(1.554)
1.16. STANDARD MODEL LAGRANGIAN 107
Lgauge = 14 F F 14 Z Z 21 (FW e G
) (FW ) 14 G a
ea
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
e2 h
+ 2
+ +
i
2 2 + +
W W W W W W e cot W 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 abc e a 2 abc d e
gs Ga fade G G + gs f Gb Gc G + gs f fade Gb Gc G G , (1.555)
4
donde
(FW ) = W+ W+
e = G G .
G (1.556)
a
108 CHAPTER 1. CLASSICAL FIELD THEORY
if ( mf ) f
X
L1 gen =
f
41 F F 14 Z Z 21 (FW ) (FW ) 14 G e ea
a G
H2
1 1 2 2 H
+ 2 H H mH H 1 + + 2
2 v 4v
H H2
2 + 1 2
+ mW W W + mZ Z Z 1+2 + 2
2 v v
a
f Qf f A
X X
+ gs q q Ga + e
q
2 f
e
f (vf af 5 )f Z
X
+
2 cos W sin W f
" #
g X mf
f (1 5 )f 0 W+ + h.c + ff H
X
+
2 2 f f
v
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
e2 h
+ 2
+ +
i
W 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 (1.557)
4
1.16. STANDARD MODEL LAGRANGIAN 109
Con
1 1 2 1
YLi = YQ i = YeiR = 1 YuiR = YdiR = . (1.559)
2 6 3 3
De los procesos entre familias, es decir de cambio de sabor, sabemos que
Los bosones gauge cargados W decaen siempre a leptones de la misma generacion y con la
misma intensidad.
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
hD ea b
ij dRi ab Qj + h.c
a b
hD
ij ab dR i Qj + h.c
e
a b
hD
ij ab (d ) 0 Qj + h.c ,
R i
e (1.561)
110 CHAPTER 1. CLASSICAL FIELD THEORY
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
e 1 21
L hD
11 (dR )1 0 Q11 + . . .
r
hD 0 r
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
D H(x)+v L
dR h 0 + h.c
2 dL
D H(x) + v
dR h dL + h.c
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 . (1.563)
2
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
dR,L 0j = (VR,L
D
)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 )ki MijD (VRD )jl = mD
k kl (1.565)
D D D D
VLD D
M M VLD = VLD MD VR
D
V R M VL
=MD D
diag Mdiag
2
= MDdiag , (1.566)
1.16. STANDARD MODEL LAGRANGIAN 111
donde MD D D D
diag = diag(m1 , m2 , m3 ). Similarmente
D 2
VR MD MD VRD
= MD
diag . (1.567)
=mD
k dLk dRk (1.568)
Para las diferentes combinaciones de terminos de corrientes
uL 0i dL 0i =uLk (VLU )ki (VLD )il dLl
=Vkl uLk dLl
L 0i eL 0i =L 0i (VLE )ij eLj
=L 0i (VLE )ij eLj
=Lj eLj (1.569)
Donde hemos definido la matriz de CabibboKobayashiMaskawa (CKM) como
V =VLU VLD
X X X X
V V =VLD VLU VLU VLD = 1 Vij Vjk = ik Vji Vjk = ik |Vji |2 = |Vij |2 = 1
j j j j
(1.570)
y los autoestados debiles de los neutrinos como
L 0i = (VLE )ij Lj (1.571)
Con esta definicion, las corrientes debiles cargadas para los leptones siguen siendo universales. Sim-
ilarmente
uL 0i uL 0i =uLk (VLU )ki (VLU )il uLl
=kl uLk uLl
=uLk uLk (1.572)
112 CHAPTER 1. CLASSICAL FIELD THEORY
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).
1.16. STANDARD MODEL LAGRANGIAN 113
Teniendo en cuenta estos resultados podemos escribir finalmente el Lagrangiano completo del
Modelo Estandar en la Gauge Unitario, para
if ( mf ) f
X
LSM =
f
41 F F 14 Z Z 12 (FW ) (FW ) 41 G e ea
a G
H2
1 H
+ 12 H H m2H H 2 1 + + 2
2 v 4v
H H2
2 + 1 2
+ mW W W + mZ Z Z 1+2 + 2
2 v v
a
f Qf f A
X X
+ gs q q Ga + e
q
2 f
e
f (vf af 5 )f Z
X
+
2 cos W sin W f
" #
g X
+
X
+
+ l (1 5 )lW + Vij ui (1 5 )dj W + h.c
2 2 l=e ij
X mf
+ ff H
f
v
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
e2 h
+ 2
+ +
i
2 2 + +
W W W W W W e cot W 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 abc e a 2 abc d e
gs Ga fade G G + gs f Gb Gc G + gs f fade Gb Gc G G . (1.574)
4
donde ml = 0.
114 CHAPTER 1. CLASSICAL FIELD THEORY
donde i = gi2 /(4). Esto tiene la ventaja de usar las tres cantidades experimentales mejor medidas.
Las relaciones
m2W
sin2 W =1 , m2W sin2 W = (1.576)
m2Z 2GF
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 su-
ficientes 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 [?]
Br(Z e+ e ) 3.363(4)
= 0.222 (1.582)
Br(Z bb) 15.12(5)
detalles del calculo de las amplitudes. Teniendo en cuenta que el canal Z tt esta cerrado
(Z e+ e )
Br(Z e+ e ) =
total
(|ve |2 + |ae |2 )
=P P2 P3
l [(|vl |2 + |a |2 ) + (|v |2 + |a |2 )] + N [
l l l c i=1 (|v ui
|2 + |a |2 ) +
ui i=1 (|vdi | + |adi | )]
2 2
(|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 3
s4 W
3.43% (1.583)
(W e e )
Br(W e e ) = (1.584)
total
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
total = (W l l ) + Nc [(W u1 di ) + (W u2 di )]
l i
X
=(W e e ){3 + Nc [|V1i |2 + |V1i |2 ]}
i
=(W e e )(3 + 2Nc )
(1.585)
entonces
1
Br(W e e ) = = 11.1% (1.586)
3 + 2Nc
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
inv (Z l l )
= 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.
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 numero 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
1 q q
D
e = g 2 . (1.589)
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 pequenas que mW ,
esto se reduce a
De = g . (1.590)
m2W
118 CHAPTER 1. CLASSICAL FIELD THEORY
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
g2 g
L= [ (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 ] , (1.591)
2
donde
GF g2
= 2
2 8mW
g24
= 2 2
8g v
1
= 2, (1.592)
2v
y
1/2
v= 2GF . (1.593)
G
L = [p (1 1.265 )n] [e (1 5 )e ] . (1.594)
2
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 ] . (1.595)
2
120 CHAPTER 1. CLASSICAL FIELD THEORY
De modo que G = GF V11 = GF cos C , donde C es el angulo de Cabbibo. Una vez se tienen en
cuenta correcciones electrodebiles se obtiene el valor |V11 | = 0.97418(27)[?]. Las magnitudes de los
elementos de la matriz CKM son[?]
0.97419 0.2257 0.0359
V 0.2256 0.97334 0.0415 1 (1.596)
0.00874 0.0407 0.999133
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
After download the source code from the http://theory.sinp.msu.ru/~pukhov/calchep.html to
some DIR,
Note that the tar.gz file name depends on the current version. At the moment of this writing
this was lhep311.tar.gz. To directly download this file use:
$ wget http://theory.sinp.msu.ru/~semenov/lhep311.tar.gz
121
122 CHAPTER 2. COMPUTATIONAL QFT
where \$ is to indicate that the command is to be written in the shell of your Linux session1 .
To uncompress the file:
$ make
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
An introduction to scientific computing is at http://gfif.udea.edu.co/cf
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 CHAPTER 2. COMPUTATIONAL QFT
5 use file_tex.
6
7 prtcprop pdg.
8
19 particle_type
20 particle/Antiparticle: (name, property name=VALUE, ...).
21
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 125
vector
A/A: (photon, gauge).
spinor e1:(electron),
e2:(muon, mass Mm = 0.1057),
e3:(tau-lepton, mass Mt = 1.777).
SetTexName([EE=e]).
prtcls1.tex:
vars1.tex
Parameter Value Comment
EE 0.31333 Electromagnetic coupling constant (1/128)
Mm 0.1057 mass of muon
Mt 1.777 mass of tau-lepton
2.3. LANHEP/CALCHEP 127
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 CHAPTER 2. COMPUTATIONAL QFT
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)
with D0 = V D, we can write the eq. (2.2) in the mass eigenstates as
LY DMDdia PR DH + h.c (2.3)
where
MDdia = V MD V (2.4)
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 (2.5)
Expanding we get . . . which is just the expresion in the Standard Model file
130 CHAPTER 2. COMPUTATIONAL QFT
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].
H|Statei
b = E|Statei (3.1)
One step further is to consider the wave function as the eigenstate of the operatorfield acting on
certain Fock states
|Fock
b Statei = ||Fock Statei , (3.2)
131
132 CHAPTER 3. SECOND QUANTIZATION
Like that usual quantum mechanical observable, the wave function will have an uncertainty. The
Fock states are the states under which the classical wave function can be obtained with a small
uncertainty
= hFock State||Fock
b 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 (3.4)
v 2 t2 z
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
also that 1/v 2 = /T . It is worth to stress that at the Lagrangian level, which is the sum of each
micro macro
l L = Nl
m = m/l
k T = kl
i (t) = (zi , t) (z, t)
Table 3.1: From micro to macro
individual oscillator Lagrangian, it is the sum of the kinetic and potential oscillator energy. However,
the Lagrangian density only have the kinetic term for the scalar field
1 1 0 3
L= 0 + 3
2 v2
21 . (3.6)
vc=1
3.1. QUANTIZATION OF THE NONRELATIVISTIC STRING 133
Note that only in the case v = c this Lagrangian can be written in a covariant form. Moreover, the
scalar 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
an ei(n tkn z) + an ei(n tkn z)
(t, z) = (3.7)
n
2 n L
n2 = v 2 kn2 (3.8)
kn = kn . (3.11)
Therefore
n = n . (3.12)
L = 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)
134 CHAPTER 3. SECOND QUANTIZATION
We now will check the origin of the normalization factor. For simplicity we work with one spatial
dimension. By using eq. (3.7)
X v
(z, t) = [an n (z, t) + an n (z, t)] , (3.16)
n=
2n
1
[E] = [a]
[E]1/2 [E]1
=E 1/2 [a] (3.17)
In this way Z L
dz n (z, t)m (z, t) = nm . (3.23)
0
Moreover Z L
dz n (z, t)m (z, t) = n,m e2in t . (3.24)
0
En tal caso de Z L
H= H dz . (3.25)
0
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 transfor-
mation
L
H = T00 = L (3.27)
(x)
= (x) L. (3.28)
t
where
L
(x) = (3.29)
((x)/t)
is the canonical conjugate variable (conjugate momentum) of the canonical variable (x).
We have then,
Z L
1 1 L
Z
H= 2 dz + dz
2v 0 t t 2 0 z z
X
= n an an (3.30)
n=
X v
= [in an n (z, t) + in an n (z, t)] ,
t n= 2n
X ivn
= [an n (z, t) an n (z, t)] , (3.31)
n=
2 n
136 CHAPTER 3. SECOND QUANTIZATION
1 X n m
2
= [an n (z, t) an n (z, t)] [am m (z, t) am m (z, t)] (3.32)
v t t n,m= 2 n m
X n m
= [an am n m an am n m an am n m + an am n m ] (3.33)
n,m=
2
n m
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)] , (3.34)
n=
2 n
X v 2 kn km
= [an n (z, t) an n (z, t)] [am m (z, t) am m (z, t)] (3.35)
z z n,m= 2 n m
X v 2 kn km
= [an am n m an am n m an am n m + an am n m ] (3.36)
n,m=
2
n m
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)
with
k
2 = (3.42)
m
This equation of motion can be obtained from the Lagrangian
1
L = T V = [mq 2 kq 2 ] (3.43)
2
138 CHAPTER 3. SECOND QUANTIZATION
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)
mn qn + i pn
an =
2mn
mn n i pn
q
an =
2mn
(3.49)
mn 1
an + an =2 qn = 2mn qn qn = (an + an )
2mn 2mn
2i i 2mn imn
an an = pn = pn pn = (an an ) (3.50)
2mn mn 2mn
3.1. QUANTIZATION OF THE NONRELATIVISTIC STRING 139
In quantum mechanics the classical objects qn and pn are promoted to operators which satisfy the
commutation relation
[b
qn , pbm ] =imn qn , qb m ] = [pbn , pb m ] =0 .
[b (3.51)
imm
qn , pbm ] =
[b {[b am ] [b
an , b an , b an , b
am ] + [b am ] + [b am ]}
an , b
2mn 2mm
imm
qn , pbm ] =
[b {[b am ] 2[b
an , b am ] + [b
an , b an , b
am ]} (3.52)
2 mn mm
am = n,m
an , b
b [b
an , b
am ] = b
an , b
am = 0 , (3.53)
Since
X 1
n , (3.56)
n=
2
140 CHAPTER 3. SECOND QUANTIZATION
This procedure is consistent since the related physics quantities arise from energy differences, no
from absolute energy determinations.
h i X 1
H, b
b am = n ab nb an + ,b
am
n=
2
X h i
= n ab nban , b
am (3.58)
n=
we have
h i
X h i
H, am =
b b n ab n , b
am ban + ab n [b
an , b
am ]
n=
X
= n nmb
an
n=
= m b
am (3.60)
h i
X h i h i
b ab m =
H, n ab n , ab m ban + ab n b
an , ab m
n=
X
= n ab n nm
n=
=m ab n (3.61)
3.1. QUANTIZATION OF THE NONRELATIVISTIC STRING 141
If |mn i is an eigenstate of H
b with eigenvalue En
b n i = En |mn i
H|m (3.62)
then
H
bban |mn i = b b nb
an H an |mn i
= (En n ) b
an |mn i
(3.63)
= (En + n ) ab n |mn i
(3.64)
an |0i =0
b
a n =0 ,
h0|b (3.65)
such that
h0|0i = 1 (3.66)
b : |0i = 0 .
:H (3.67)
142 CHAPTER 3. SECOND QUANTIZATION
We define the state whose energy is larger tha the energy of |0i by one quantum ~n by
a n |0i
|1n i b
h1n | =h0|b an (3.68)
|1n i is an Hamiltonian eigenstate of energy n :
b : |1n i =n a an |1n i
:H n
= n |1n i
= n 1|1n i , (3.69)
where we have made explicit that we have a quantum of energy ~. The normalized state is
an ab n |0i
h1n |1n i =h0|b
h i
=h0| ban , ab n |0i
=h0|0i
=1 . (3.70)
Similarly, the state with energy 2~ is
1 2
ab n |0i =|2n i
2
1
an )2 =h2n |
h0| (b (3.71)
2
with normalization
1
h2n |2n i = h0|b an ab n ab n |0i
an b
2
1
an ab n |1n i
= h1n |b
2
1 h i
= 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 (3.73)
m!
3.1. QUANTIZATION OF THE NONRELATIVISTIC STRING 143
1 m1
ab n ab n |0i =|mn i
m!
p
(m 1)! 1 m1
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
ab n |(m 1)n i = m|mn i
ab n |mn i = m + 1|(m + 1)n i (3.74)
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:
bn = ab nb
N an (3.77)
If N
bn |mn i = c|mn i, where c must be real because N
bn is Hermitian
1+c=m+1 (3.78)
and
N
bn |mn i = mn |mn i (3.79)
144 CHAPTER 3. SECOND QUANTIZATION
N
bn |mn 1i =(mn 1)|mn 1i (3.82)
an |mi = C |mn 1i
b (3.83)
an |mn i =
b mn |mn 1i (3.85)
such that
Nn |ml i = 0 n 6= l , (3.88)
3.1. QUANTIZATION OF THE NONRELATIVISTIC STRING 145
we have that
hmn | : H
b : |mn i = mn n = mn ~n . (3.89)
Therefore, once we have proper normalized states and renormalized Hamiltonian, the energy of an
state with m quantum ( of frequency n ) is just m times the energy of the one quanta of energy ~n .
Note that
h0| : H
b : |0i = 0 . (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
right of all creation operators. For an operator X,b its normal-ordered product will be denoted as
: X : . Using this algorithm on the expression of eq. (3.55), we find that
b
1 X
: H: =
b a nb
n : (b an ab n ) :
an + b
2 n=
1 X
= a nb
n (b an + ab nb
an )
2 n=
X
= n ab nb
an (3.91)
n=
= 0 , (3.92)
L = 12 , (3.93)
is
X 1 ipn x ipn x
(t, x) = a n e + a n e ,
n
2En L3
X 1
= p
3
a(nx ,ny ,nz ) exp{i[E(nx ,ny ,nz ) t px x py y pz z]}
nx ,ny ,nz
2E (nx ,n y ,n z ) L
o
+a(nx ,ny ,nz ) exp{i[E(nx ,ny ,nz ) t px x py y pz z]} , (3.94)
where in natural units the wave number can be identified with the momentum, p = k. In eq. (3.94)
2
En =p0n pi = ni (3.95)
L
2 q 2
En = nx + n2y + n2z (3.97)
L
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)
1
[E] = p [an ]
[E][E]3
=[E][an ] , (3.99)
am = n,m
ban , b [b
an , b
am ] = b
an , b
am = 0 , (3.100)
( + m2 ) = 0 , (3.101)
L = 21 12 m2 2 , (3.102)
The solution is the same that for the case m = 0 in eq. (3.94), but the new dispersion relation is
3 Z
X L
d3 p (3.105)
n
2
148 CHAPTER 3. SECOND QUANTIZATION
From
Z
d3 p (3) (p q) =1 (3.106)
where
2 2
pi = ni qi = mi , (3.108)
L L
we have
Z X
d3 p (3) (p q) = nm
n
X 2 3 X
(3) (p q) = n,m
n
L n
3
2
(3) (p q) =n,m . (3.109)
L
In this way
3
(3) L
(p q) = p,q , (3.110)
2
V
3 (0) = . (3.113)
(2)3
3.2. QUANTIZATION OF THE KLEIN-GORDON FIELD 149
Using eq. (3.111), we can write the commutation relations (3.53) in the continuum as
3
2
aq =
(3) (p q)
ap , b [b
ap , b
aq ] = b
ap , b
aq = 0 . (3.116)
L
b
Note that again ap is dimensionless. It is customary to write the general solution (3.115) with
a0p = L3 ap . (3.117)
Then
1
Z
0 ipx 0 ipx
d3 p
(t, x) = p a p e + a p e . (3.118)
(2)3 2Ep
aq = (2)3 (3) (p q)
0 0 0 0 0 0
ap , b
b ap , b
b aq = b aq = 0 .
ap , b (3.119)
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
1
Z
ipx ipx
(t,
b x) = d3 p p ba p e + a
b p e (3.121)
(2)3 2Ep
with
h i h i
bap , ab q = (2)3 (3) (p q) [b aq ] = ab p , ab q = 0 .
ap , b (3.122)
The conjugate momentum can be obtained from the Klein-Gordon Lagrangian in eq. (3.102), by
using eq. (3.29)
h
1 b2
i
(x)
b = ( )
2 0
(0 )b
=0 b
1
Z
ipx ipx
= d3 p
p iE pba pe + iE pba p e
(2)3 2Ep
r
i Ep
Z
3 ipx ipx
= dp b
a p e + a p e
(2)3 2
b
(3.123)
3.2. QUANTIZATION OF THE KLEIN-GORDON FIELD 151
s
i Z i Ep0 h ipx
h Z i
0 0
(t, x), (t, y) = d p d3 p0
b b 3
ap e ap eipx , b
+b ap0 eip y
ap0 eip y + b
2(2)6 Ep
b
s
i Ep0
Z Z
= d3 p d3 p0
2(2)6 Ep
nh 0
i h 0
io
bap eipx + b
ap eipx , bap0 eip y + b ap eipx + b ap0 eip y
ap eipx , b
s
i Ep0 nh ipx
Z Z i
ip0 y
= d3 p d3 p0 a p e , ba p0 e
2(2)6 Ep
b
h i h i h io
ipx ip0 y ipx ip0 y ipx ip0 y
+ b ap e , b ap e0 + b ap e ,bap 0 e + b ap e , b ap0 e
s
i Ep0 n i(px+p0 y)
Z Z
= d p d3 p0
3
e [bap , bap0 ]
2(2)6 Ep
0 0
h i 0
h io
ei(pxp y) b ap0 + ei(px+p y) b ap0
ap0 + ei(pxp y) b ap , b
ap , b ap , b . (3.124)
s
i Z i Ep0 n i(pxp0 y) h
h Z i 0 o
(t, x), (t, y) = d p d3 p0 ap0 ei(pxp y) b
3
b b e ap , b ap , b
ap0
2(2)6 Ep
b
s
i Ep0 h i(pxp0 y) (3)
Z Z i
0
= d3 p d3 p0 e (p p0 ) + ei(pxp y) (3) (p0 p)
2(2)3 Ep
s
i Ep0 (3)
Z Z h i
0 0
= d3 p d3 p0 (p p0 ) ei(pxp y) + ei(pxp y) . (3.125)
2(2)3 Ep
152 CHAPTER 3. SECOND QUANTIZATION
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.
1
Z
H=
b d3 p Ep ab pb ap ab p
ap + b (3.129)
2
h i
H,
b bap = 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
creates the quanta of momentum p of the field ,
b while b
ap is the annihilation operator for a field
quantum with momentum p. From [4]:
3.2. QUANTIZATION OF THE KLEIN-GORDON FIELD 153
What was the positive energy component of the classical field now annihilates the quan-
tum, and the negative energy component now creates the quantum. This quantum is
what we call particle of positive energy.
We can now construct the Fock space following the standard procedure for the harmonic oscillator:
we interpret bap as destruction operators and bap as creation operators, and we define a vacuum state
|0i as the state annihilated by all destruction operators, so for all p
ap |0i = 0 .
b (3.131)
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
: H : = d3 p Ep ab pb
b ap (3.132)
h0| : H
b : |0i = 0 . (3.133)
A possible normalization factor for the Fock one-particle state is (|pi |1p i)
1
|pi = b ap |0i
V
1
ap
hp| =h0|b (3.134)
V
This state contains one quantum of the field b with momenta p = (Ep , p). Such states have positive
norm, since
1
hp|p0 i = h0|b ap0 |0i
ap b
V
1
= h0|b ap0 b
ap b ap0 b
ap |0i
V
1
= h0|[b ap0 ]|0i
ap , b
V
(2)3 (3)
= (p p0 )
V
3
2
= (3) (p p0 ) (3.135)
L
154 CHAPTER 3. SECOND QUANTIZATION
Table 3.2: From discret to continuos, where pi = 2ni /L, and qi = 2mi /L ,
Similarly we can define many particle states. If a state has N particles with all different momenta
p1 , p2 , . . . , pN , it is defined by
1
|p1 , . . . , pN i = N/2
ab p1 ab pN |0p1 , . . . , 0pN i
V
1
N/2
ab p1 ab pN |0i
V
(3.137)
On the other hand, if we want to construct a state with m particles of momentum p, we must have
a Fock state similar to (3.73)
1 1 m
|mp i = ab p |0i (3.138)
V m/2 m!
From[4]
The vacuum, together with single particles states (3.134) and all multiparticle states
(3.137), (3.138), constitute a vector space which is calles the Fock space. The creation
and annihilation operators act on this space.
3.2. QUANTIZATION OF THE KLEIN-GORDON FIELD 155
It is convinient to define:
(x)
b = b+ (x) + b (x) (3.139)
where
1
Z
b+ (x) = d3 p p ap eipx
b
(2)3
2Ep
1
Z
b (x) = d3 p 3
p ap eipx .
b (3.140)
(2) 2Ep
The effect of the operator field, b (x), on the one particle state |pi
(x)|pi (3.141)
will be important for the evaluation of Smatrix elements in Chapter 7.
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 . (3.142)
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 . (3.143)
From the commutations relations and eq. (3.122) we find that
hp|qi = (2Ep )1/2 (2Eq )1/2 h0|b aq |0i
ap b
= (2Ep )1/2 (2Eq )1/2 h0| b aq |0i
ap , b
= (2Ep )1/2 (2Eq )1/2 (2)3 (3) (p q)
=2Ep (2)3 (3) (p q) . (3.144)
The factors (2Ep )1/2 in eq. (3.143) 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
hp|qi =2En L3 n,m
=2En V n,m . (3.145)
156 CHAPTER 3. SECOND QUANTIZATION
3.3 Propagator
With conventions
(x)
b = b+ (x) + b (x) (3.146)
where
1
Z
b+ (x) = d3 p p ap eipx
b
3
(2) 2Ep
1
Z
b (x) = d3 p p ap eipx .
b (3.147)
3
(2) 2Ep
+ m2 (x) = J(x)
(i m)(x) = 0 (3.148)
L = i m (3.149)
where
= 0 (3.150)
{ , } = 2g 1 (3.151)
See [1]. If we assume a plane wave solution like the wave function of the Scrodinger equation
eiEt , after sustition in eq. (3.148), we have
i 0 (iE) m =0
0 E m =0 (3.152)
3.4. QUANTIZATION OF FERMIONS 157
Moreover, we know that if satisfy the Dirac algebra, the matrices obtained after the unitary
transformation
e =U U
s.t U =U 1 (3.154)
e } = U U, U U
,
{e
=U { , } U
=2g U U
=2g (3.155)
In this way we can always choose U such that 0 be diagonal. Because the restrictions in eq. (3.153)
this implies that in this representation we have
0 1 0
= (3.156)
0 1
where the 1 and 0 are the 2 2 identity and null matrix respectively. Replacing back in eq. (3.152)
we have
Em 0
=0
0 E m
E = m. (3.157)
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) v1 (p)eipx = + (x) + (x) ,
(3.158)
v2 (p)eipx
where
with
u1 (p) 0
u2 (p) 0
us (p) =
0
vs (p) =
v1 (p)
(3.160)
0 v2 (p)
Checking this solutions to eq. (3.148) we have
(i 0 0 + i i i m)+ (x) =0
(i 0 0 + i m)+ (x) =0
( 0 E p m)+ (x) =0
( p m)+ (x) =0
(p m)+ (x) =0
(p m)us (p) =0 (3.161)
and
(p + m)vs (p) =0 (3.162)
This equations can also be written as
[(p m)us (p)] =0
us (p)( p m) =0
us (p) 0 p mus (p) 0 =0
us (p) 0 p mus (p) 0 =0
us (p)(p m) =0 (3.163)
vs (p)(p + m) =0 (3.164)
At zero momentum, E = m and
us (0)(p m) =0
us (0) 0 ( 0 p0 + i pi m) =0
us (0) 0 ( 0 p0 m) =0
us (0)(E 0 m) =0
us (0)m =us (0) 0 m
us (0) =us (0) 0 ,
3.4. QUANTIZATION OF FERMIONS 159
therefore
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. (3.151) 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
Eq. (3.165) 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
since the belong to differente eigenvalues. Note that the two us (0) and the two vs (0) are degenerate.
We define
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 (3.168)
To obtain the spinors for any value of p we know that they must satisfy eqs. (3.161), (3.162),
and, reduce to eq. (3.167) when p 0. The result is
Choosing
1
Np = (3.171)
E+m
we obtain
The spinors also satisfy some completeness relations (For details see [5])
X
us (p)us (p) = p + m (3.176)
s
X
vs (p)vs (p) = p m (3.177)
s
[ (p)]2 = (p) .
+ = + =0
+ + =1 .
+ us (p) =us (p)
+ vs (p) =0
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
2
1 5
L =
2
2
(L ) =L
L + L =1
L L =0 .
162 CHAPTER 3. SECOND QUANTIZATION
Z
H= d3 xH
L
Z
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) . (3.180)
Since that
1
Z X
(i + m) = d3 p 3
p as (p) (i + m) us (p)eipx
(2) 2Ep s=1,2
+bs (p) (i + m) vs (p)eipx
1
Z X
= c d3 p 3
p as (p) ( p + m) us (p)eipx
(2) 2Ep s=1,2
+bs (p) ( p + m) vs (p)eipx .
(3.181)
3.4. QUANTIZATION OF FERMIONS 163
0 p0 + i pi m us (p) =0
0 p0 + i pi + m vs (p) =0 ,
(3.182)
!
X
0 Ep i pi m us (p) =0
i
!
X
0 Ep i pi + m vs (p) =0 , (3.183)
i
( m) us (p) = 0 Ep us (p)
( + m) vs (p) = 0 Ep vs (p) , (3.184)
( + m) us (p) =0 Ep us (p)
( m) vs (p) =0 Ep vs (p) . (3.185)
1
Z X
ipx
d3 p ipx
(i + m) = p a s (p)0 Ep us (p)e b s (p)0 Ep vs (p)e . (3.186)
(2)3 2Ep s=1,2
164 CHAPTER 3. SECOND QUANTIZATION
Therefore
Z
H= d3 x (i + m)
1
Z Z X h i
0 0 ip0 x 0 0 ip0 x
= d x d3 p0
3
a 0 (p )u 0 (p )e + b s0 (p )v 0 (p )e 0 (i + m)
(2)3 2Ep0 s0 =1,2 s s s
p
1
Z Z X h i
0 0 ip0 x 0 0 ip0 x
= d x d3 p0
3
a 0 (p )u 0 (p )e + b s0 (p )v 0 (p )e 0
(2)3 2Ep0 s0 =1,2 s s s
p
1
Z X
ipx
d3 p ipx
p a s (p)0 Ep us (p)e b s (p) 0 Ep vs (p)e
(2)3 2Ep s=1,2
s
d3 x Ep X h 0 0 ip0 x
Z Z Z i
3 0 3 0 0 ip0 x
= d p d p a 0 (p )u 0 (p )e + b s 0 (p )v 0 (p )e
2(2)6 Ep0 s,s0 =1,2 s s s
In order to obtain the quantization relations could see that if commutation relations are used we
could get
d3 p X
Z
:H: = E p a s (p)a s (p) b s (p)b s (p) (3.188)
(2)3 s=1,2
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 neg-
ative energy eigenvalues, which is a serious problem. If instead we assume that the creation and
annihilation operators satisfy anticommutation relations
L
(x) = = i 0 = i (3.190)
(0 )
we obtain
d3 p X
Z
:H: = Ep as (p)a s (p) + b s (p)b s (p) (3.192)
(2)3 s=1,2
Moreover
Z
: Q : =q d3 x : :
Z X
d3 p as (p)as (p) bs (p)bs (p)
=q (3.193)
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. (3.139), the most general free particle solution to Dirac equation is
(x)
b = b+ (x) + b (x) (3.194)
166 CHAPTER 3. SECOND QUANTIZATION
1
Z X
b+ (x) = d3 p
3
p as (p)us (p)eipx
(2) 2Ep s=1,2
1
Z X
b (x) = d3 p 3
p bs (p)vs (p)eipx (3.195)
(2) 2Ep s=1,2
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 (3.197)
V s
while the one antiparticle state is
r
+ 1
|e (p, s)i b (p)|0i . (3.198)
V s
Chapter 4
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)
167
168 CHAPTER 4. QUANTIZATION OF THE ELECTROMAGNETIC FIELD
1 1
Lfree = m2 2
2 2
Lint = , (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
V (r) = . (4.7)
4 r
In a similar way, when the Lorentz force
F = qE + qv B, (4.8)
is interpreted in terms of quantum mechanical operators ( 3.3 of [1]) we have the canonical momen-
tum
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)
This suggest to make the minimum replacement
D = + iqA , (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)
From eqs. (4.10) and (4.11) the tranformation of A can be obtained:
1
A A0 = A . (4.13)
q
Therefore, the new field tranform like the electromagnetic field, and the modified Lagrangian for the
fields and A
1 i 1
L= (D) D D0 D0 F F , (4.14)
2m 2 4
4.2. QUANTIZATION OF THE ELECTROMAGNETIC FIELD 169
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)
2m
[() 2iq A] = i
A = 0 , (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 ) .
j 0
(4.20)
170 CHAPTER 4. QUANTIZATION OF THE ELECTROMAGNETIC FIELD
These commutation relations between the creation and annihilation involve only the independent
degrees of freedom.
The Hamiltonian obtainde from the Lagrangia is
X
H
b= n an, an, , (4.21)
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.
Chapter 5
Smatrix
We will use the Smatrix formulation to obtain the decay rates and cross section formulas.
is
i |a, ti = H|a, ti . (5.2)
t
The solution to this equation is
since
i |a, ti =i(i)HeH(tti ) |a, ti i
t
=H|a, ti . (5.4)
171
172 CHAPTER 5. SMATRIX
we have that in 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
|a, ti =U (t, ti )|ai
|a, ti =eiH(tti ) |ai , (5.6)
where |a, ti i, at an initial time ti , is an eigenstate of a set of conmuting operators, and is denoted
simply |ai. Similarly |bi = |b, tf i.
We have then
hb, tf |a, tf i =hb|a, tf i
=hb|eiH(tf ti ) |a, ti i
=hb|eiH(tf ti ) |ai , (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 eiH(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)
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)
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
In explicit calculations there will be an overall Dirac delta factor imposing energymomentum con-
servation. In order not to write explicitly the Dirac delta each time we compute a matrix element of
iT , it is convenient to 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 j
Mf i = M (p1 , . . . , pn ; k1 , . . . , kn ) . (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 .
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
and the normalization constant is fixed by the condition that there is one particle in the volume V ,
Z Z
2
1= 3
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 . (5.20)
V
Wave functions with different momenta are orthogonal, and therefore
Z
d3 x p 1 (x)p2 (x) = p1 ,p2 (5.21)
V
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
Z V
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.
5.3. PROCESS PROBABILITY 175
In relativistic QFT this normalization is not the most convenient, because the spatial volume
V is not relativistically invariant, and therefore the condition one-particle per volume V is not
invariant. A more convenient Lorentz invariant form was introduced in eq. (3.145)
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)
and of course for a multiparticle state
" n
#
Y
|p1 , . . . , pn iR = (2Ep V )1/2 |p1 , . . . , pn iNR (5.25)
i=1
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 i
n n
Y 1/2
Y 1/2
= (2Epi V ) 2Ekj V hp1 . . . pn |iT |k1 . . . kn iNR
i=1 j=1
n n
!
Y 1/2
Y 1/2 X X
= (2Epi V ) 2Ekj V (2)4 (4) pi ki iMf i
i=1 j=1 i i
(5.26)
Therefore
n n
Y 1/2
Y 1/2
Mf i = (2Epi V ) 2Ekj V Mf i (5.27)
i=1 j=1
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
2
X X
hp1 . . . pn |iT |k1 . . . km iNR = (2)4 (4) pi kj iMf i
(5.29)
i j
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)
Similarly we regularize also the time interval, saying that the time runs from T /2 to T /2 so that
Then
! 2
2
X
hp1 . . . pn |iT |k1 . . . km iNR = (2)4 (4) p
ki iMf i
j
!
X
=(2)4 (4) p kj V T Mf i (5.32)
i
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
x
2nj nj = , . . . , 1, 0, 1, . . .
kj = , nyj = , . . . , 1, 0, 1, . . . (5.33)
L
z
nj = , . . . , 1, 0, 1, . . .
X XXX
= (5.34)
kj nx
j nyj nzj
In the large-volume limit for each particle we can write, using eq. (3.105)
V
X Z
3
d3 kj , (5.35)
k
(2)
j
5.4. CROSS SECTION 177
The probability for the process of an initial particle decaying into n final particles is then
! n
2 1 d3 kj
Z Z X Y
4 (4)
1 = . . . (2) p kj T |Mf i | (5.38)
i
2Ep j=1 (2)3 2Ekj
On the other hand the probability for a process with two initial particles colliding into n final particles
is
! n
1 1 Y d3 kj
Z Z X
4 (4) 2
2 = . . . (2) p1 + p2 kj V T |Mf i | (5.39)
i
2Ep1 V 2Ep2 V j=1 (2)3 2Ekj
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
PS = . (5.40)
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 . (5.41)
Choosing t such that the volume contains just one particle, we can write
n = 1/V , (5.42)
or
1 = nvtA . (5.43)
Ps
= . (5.44)
nvt
5.4. CROSS SECTION 179
The density of particles of the incident state is normalized to one particle in the entire volume, so
that n = 1/V . Therefore
! n
1 1 1 Y d3 kj
Z Z X
4 (4) 2
= . . . (2) p1 + p2 kj |Mf i | . (5.46)
v i
2Ep1 2Ep2 j=1 (2)3 2Ekj
In general, as both particles may be moving we could use the relative velocity between them, vrel ,
! n
1 1 1 Y d3 kj
Z Z X
4 (4) 2
= . . . (2) p1 + p2 kj |Mf i | . (5.47)
vrel i
2Ep1 2Ep2 j=1 (2)3 2Ekj
In a frame where p1 and p2 are along the same line, this reduces to
p1 p 2
vrel = . (5.48)
E1 E2
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 | . (5.49)
I
vrel = (5.50)
E1 E2
where
q
I = (p1 p2 )2 m21 m22 (5.51)
180 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 (5.52)
Since
m21 m22 =(E12 p21 )(E22 p22 )
=(E12 E22 p21 E22 E12 p22 + p21 p22 ) (5.53)
q
I= p21 E22 2E1 E2 p1 p2 + E12 p22 + (p1 p2 )2 p21 p22 (5.54)
If
(p1 p2 )2 p21 p22 = 0 (5.55)
that implies that p1 and p2 are colineals,
q
I = p21 E22 2E1 E2 p1 p2 + E12 p22
p
= (p1 E2 p2 E1 )2
=|p1 E2 p2 E1 | (5.56)
I p1 p 2
vrel = = (5.57)
E1 E2 E1 E2
To simplify the notation we set Ei = Epi =, and Ef = Epf . Moreover, the differential cross
section is
2
!
X X 1 Y d3 kf
d =(2)4 (4) pi pf |Mf i |2
i=1 f
4vrel E1 E2 f
(2)3 2Ef
1
= (2)4 |Mf i |2 dn (p1 , p2 ; k1 , . . . , kn ) (5.58)
4vrel E1 E2
where
! n
X Y d3 kj
d(n) (p1 , p2 ; k1 , k2 , . . . , kn ) = (4) p kj 3 2E
. (5.59)
j j=1
(2) k j
We keep the diferential notation both for d, and d until the last integration have been made.
5.4. CROSS SECTION 181
p01 2 d|p01 |d
4(2)6 d(2) = (E1 + E2 E10 E20 ) (5.66)
E10 E20
As
q
|p01 | = E10 2 m1 2 (5.67)
d|p01 | 2E10
=
dE10
q
2 E10 2 m1 2
E10
= (5.68)
|p01 |
In this way, we can write, in general
and
|p01 |E10 dE10
4(2)6 d(2) = (E1 + E2 E10 E20 ) d
E10 E20
|p0 |dE 0
= (E1 + E2 E10 E20 ) 1 0 1 d (5.70)
E2
From the function in Eq. (5.63) we have that in the CM frame
(
CM p1 = p2
p1 + p2 p01 p02 = 0 (5.71)
p01 = p02
we have
2 2
p01 =p02
2 2 2 2
E10 m01 =E20 m02 , (5.73)
5.4. CROSS SECTION 183
q
E20 = E10 2 m01 2 + m02 2 (5.74)
In this way we can express E20 in terms of E10 in Eq. (5.70). Moreover, we can define the center of
mass energy as
s = E1 + E2 (5.75)
q
0 02 02 02
s E1 E1 m1 + m2 (5.76)
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
X (z zn )
(f (z)) = (5.77)
n
|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
(x x0 )
(f (x)) = (5.78)
|f 0 (x0 )|
where
q
f (x) = s x x2 m01 2 + m02 2 (5.79)
Therefore
(x x0 ) |p01 (x)|
Z
6 (2)
4(2) d =d dx
|f 0 (x0 )| E20 (x)
1 |p01 (x0 )|
=d 0
|f (x0 )| E20 (x0 )
(5.80)
we have
m01 2 m02 2 + s
f 0 (x0 ) = r 1
(m01 2 +m02 2 +s)
2
s s
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
= , (5.85)
s + m02 2 m01 2
and
s + m02 2 m01 2
(f (E10 )) = (E10 x0 ) (5.86)
2s
Replacing the expression for x0 in (5.83) into Eq. (5.81) 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 ) = (5.87)
2 s
5.4. CROSS SECTION 185
we have
To further evaluate Eq. (5.60), we need to express vrel and E1 E2 in terms of s and the masses.
Concerning vrel , from Eq. (5.57), 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 (5.93)
186 CHAPTER 5. SMATRIX
Replacing back Eqs. (5.88), and (5.93) into Eq. (5.60), we have
1 2 6 (2)
d = |Mf i | 4(2) d (5.94)
64 2 E1 E2 vrel
p
d 1 [s (m01 + m02 )2 ][s (m01 m02 )2 ]
= 2
|Mf i |2 (5.95)
d 64 E1 E2 vrel 2s
In the CM frame
s =E1 + E2
q q
= p1 + m1 + p22 + m22
2 2
q q
= p1 + m1 + p21 + m22
2 2
(5.97)
q
s =2p21
+ + m21 m22 + 2 p41 + (m21 + m22 )p21 + m21 m22
q
s (2p21 + m21 + m22 ) =2 p41 + (m21 + m22 )p21 + m21 m22 (5.98)
s2 2s(2p21 + m21 + m22 ) + [2p21 + (m21 + m22 )]2 =4(p41 + (m21 + m22 )p21 + m21 m22 )
s2 2s(2p21 + m21 + m22 ) + 4p41 + 4p21 (m21 + m22 ) + (m21 + m22 )2 =4(p41 + (m21 + m22 )p21 + m21 m22 )
4sp21 + s2 2s(m21 + m22 ) + (m21 + m22 )2 =4m21 m22
4sp21 + s2 2sm21 2sm22 + m41 + m42 + 2m21 m22 =4m21 m22
4sp21 + s2 2sm21 2sm22 + m41 + m42 2m21 m22 =0 (5.99)
and, finally
1/2
[s (m01 + m02 )2 ][s (m01 m02 )2 ]
d 1
= |M|2 (5.104)
d 64 2 s [s (m1 + m2 )2 ][s (m1 m2 )2 ]
or, in terms of the kinematic function defined in eq. (5.89)
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. (5.38) we have
! Z n
2 1 d3 kj
Z Z X Y
4 (4)
= . . . (2) p kj dt |Mf i | (5.107)
i
2Ep j=1 (2)3 2Ekj
188 CHAPTER 5. SMATRIX
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
d 4 (4)
X 1 2
Y d3 kj
d = (2) p kj |Mf i |
dt j
2Ep j=1
(2)3 2Ekj
(2)4
= |Mf i |2 d(n) (p; k1 , k2 , . . . , kn ) (5.109)
2Ep
where
! n
X Y d3 kj
d(n) (p; k1 , k2 , . . . , kn ) = (4) p kj 3 2E
. (5.110)
j j=1
(2) k j
(2)4
d = |Mf i |2 d(n) (p; k1 , k2 , . . . , kn ) (5.111)
2Ep
p =0 M = Ep (5.112)
Therefore
(2)4
d = 6
|Mf i |2 4(2)6 d(2) (p; p1 , p2 )
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
(5.113)
32 M
5.6. BACKUP 189
where
d3 p1 d3 p2
4(2)6 d(2) (p; p1 , p2 ) = (4) (p p1 p2 ) (5.114)
E1 E2
The Dirac delta in eq. (5.110) can be written in the CM frame as
(4) (p p1 p2 ) = (3) (p p1 p2 )(E E1 E2 )
= (3) (p1 + p2 )(M E1 E2 ) (5.115)
and,
6 (2) d3 p1 d3 p2
(3)
4(2) d (p; p1 , p2 ) =(M E1 E2 ) (p1 + p2 ) (5.116)
E1 E2
Comparing with eq. (5.65) we see that the two quantities are the same after the replacing s M,
p01 p1 and p02 p2 . Therefore we have from eq. (5.91)
1/2 (M 2 , m22 , m21 )
4(2)6 d(2) = d (5.117)
2M 2
Replacing back in eq. (5.113)
1/2
d 1 2 (M 2 , m22 , m21 )
= |M f i |
d 32 2 M 2M 2
1
= |Mf i |2 1/2 (M 2 , m22 , m21 ) (5.118)
64 2 M 3
By using eq. (5.92) we can write this expression also as
d 1
= |Mf i |2 2M |p1 |
d 64 2 M 3
|p1 |
= |Mf i |2 (5.119)
32 2 M 2
as usually written in several texts.
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 (5.120)
190 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
Hint = d x Hint = d3 x Lint
3
(5.121)
with
Lint = eA (5.122)
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 ,
(5.123)
T T = i(T T ) . (5.124)
if a = b
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.
(2)4
dn = |M|2 d(n) (P ; p1 , p2 , . . . , pn ) (6.1)
2M
n n
!
X Y d3 pi
d(n) (P ; p1 , p2 , . . . , pn ) = 4 (P pi ) . (6.2)
i=1 i=1
(2)3 2Ei
We will keep the d notation until all the integrals get evaluated.
The two-body decays in eq. (5.118) is
d 1
= |Mf i |2 1/2 (M 2 , m22 , m21 ) (6.3)
d 64 2 M 3
191
192 CHAPTER 6. TWO BODY DECAYS
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)
LHiggs = Gf (f R fL + f L fR )
2
Gf v Gf H
= ff ff
2 2
1/2
= mf f f mf GF 2 ff (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
1/2
h = mf GF 2 in the standard models Now, having into account that 0 = 0
(u(s1 , p1 )v(s2 , p2 ))
= v (s2 , p2 )(u(s1 , p1 ))
= v (s2 , p2 )(u (s1 , p1 ) 0 )
= 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 )). (6.8)
s1 ,s2 s1 ,s2
194 CHAPTER 6. TWO BODY DECAYS
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
X
= (u (s1 , p1 )v (s2 , p2 ))(v (s2 , p2 )u (s1 , p1 ))
s1 ,s2
X
= (u (s1 , p1 )u (s1 , p1 ))(v (s2 , p2 )v (s2 , p2 ))
s1 ,s2
X X
= u (s1 , p1 )u (s1 , p1 ) v (s2 , p2 )v (s2 , p2 )
s 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
Therefore
M 2 E12 E22
E1 E2 = (6.11)
2
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
Therefore, the scattering amplitude is
X
|M|2 = 2h2 M 2 (m1 + m2 )2
(6.12)
s1 ,s2
d h2 1/2 2 2 2
2 2
= (M , m2 , m1 ) M (m1 + m 2 ) (6.13)
d 32 2 M 3
R
After the integration dCM = 4 2 we have
h2 1/2 2 2 2 2 2
= (M , m2 , m1 ) M (m1 + m2 ) (6.14)
8M 3
For m1 = m2 = mf
1/2
4m2f
1/2 2
(M , m22 , m21 )
=M 1 2
M2
4m2f
2 2
2
M (m1 + m2 ) =M 1 (6.15)
M2
2
R 2 R
0
d 0
sin d = 4
196 CHAPTER 6. TWO BODY DECAYS
and therefore
3/2
h2 4m2f
(H f f ) = M 1 (6.16)
8 M2
In the case of the standard model Higgs with mass MH decaying to fermion pair, according to the
Lagrangian in eq. (6.5)
3/2
MH m2f GF m2f
(H f f ) = 14 2 , (6.17)
4 2 MH
In the limit mf MH this expression reduces to
MH m2f GF
(H f f ) = . (6.18)
4 2
6.3 e + e +
e2
v(k2 ) u(k1 ) v(k2 ) u(k1 )
L= (6.19)
s
Chapter 7
Feynman Rules
When the case of interacting fields are considered, the particles can be created, destroyed and scat-
tered. 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.
197
198 CHAPTER 7. FEYNMAN RULES
OH (t) = U OS U (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
Eq. (7.5) ensures the invariance of matrix elements and commutation relations:
S hb, t| O
S
|a, tiS = S hb, t| U OH (t)U |a, tiS = H hb, ti |OH (t)|a, ti iH (7.7)
S S
O , P = c OH (t), P H (t) = c
(7.8)
where c is a constant.
Differentiation of Eq. (7.5)
d H d d
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
i (7.10)
dt
The interaction picture (IP) arises if the Hamiltonian is split into two parts
H = H0 + HI . (7.11)
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
in this way,
and
OI (t) = U0 OS U0 . (7.14)
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:
d I
O (t) = OI (t), H0
i (7.15)
dt
Substituting Eq. (7.13) into the Scrodinger Eq. (??), one obtains the equation of motion of state
vectors in the IP, If the system is described by a time-dependent state vector |(t)i
d
i |a, tiS =H S |a, tiS
dt
d
i (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 (7.16)
dt
d
i |(t)iI = HII |(t)iI , (7.17)
dt
where, as in Eq. (7.14)
S S
HII = eiH0 (tti ) HIS eiH0 (tti ) (7.18)
200 CHAPTER 7. FEYNMAN RULES
is the interaction Hamiltonian in the IP, with HIS and H0S being the interaction and free-field Hamil-
tonian in the SP. From now on we shall omit the labels I, used in the equations to distinguish the
IP, as we shall be working exclusively in the IP in what follows.
Eq. (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.
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
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
A collision can lead to many different final states |f i, and all these possibilities are constrained
within |()i.
The transition probability is given by
In order to calculate the Smatrix we must solve Eq. (7.17) for the initial condition (7.19). These
equations can be combined into the integral equation
Z t
|(t)i = |ii i dt1 HI (t1 )|(t1 )i . (7.25)
In the limit t
Z
(0)
|()i = S |ii i dt1 HI (t1 )|(t1 )i . (7.26)
where
S (0) = 1 . (7.27)
This equation then can be solved iteratively. If HI is small we can solve this equation by iteration
Z t Z t Z t1
2
|(t)i = |ii + (i) dt1 HI (t1 )|ii + (i) dt1 dt2 HI (t1 )HI (t2 )|(t2 )i . (7.29)
In the limit t
Z Z Z t1
(0) 2
|(t)i = S + (i) dt1 HI (t1 ) |ii + (i) dt1 dt2 HI (t1 )HI (t2 )|(t2 )i
Z Z t1
(0) (1) 2
= S +S |ii + (i) dt1 dt2 HI (t1 )HI (t2 )|(t2 )i , (7.30)
202 CHAPTER 7. FEYNMAN RULES
where
Z
(1)
S = (i) dt1 HI (t1 ) . (7.31)
Z t Z t Z t1
2
|(t)i =|ii + (i) dt1 H1 (t1 )|ii + (i) dt1 dt2 HI (t1 )HI (t2 )|ii
Z t Z t1 Z t2
3
+ (i) dt1 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 (7.33)
In the limit t
where
Z Z t1
(2) 2
S =(i) dt1 dt2 HI (t1 )HI (t2 )
Z Z t1 Z t2
S (3) =(i)3 dt1 dt2 dt3 HI (t1 )HI (t2 )H1 (t3 ) (7.35)
p2e Z
HA = , (7.41)
2m re
1
Z
d3 r : 2 (r, t) : + : B2 (r, t) : ,
HEM = (7.42)
2
e2 2
e
HI0 (t) = pe A(re , t) + A(re , t) pe + A (re , t) , (7.43)
2m 2m
The scalar product of the atomic states requires an integration over the coordinate re
Z
ha , n |a, ni = d3 r a0 (re )a (re )hn0 |ni.
0 0
(7.45)
204 CHAPTER 7. FEYNMAN RULES
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
where
and
ie
Z
Sba = i dthb, 1n |UA1 A(re , t) e UA |a, 0i
m
Z
1 ie
Sba = i dthb, 1n |UA A(re , t) e UA |a, 0i
m
Z
ie
Sba = i dthb, 1n |e iEb t
A(re , t) e eiEa t |a, 0i
m
Z
e
Z
= dt ei(Eb Ea )t d3 re h1n |b (re ) (A(re , t) e ) a (r)|0i
m
Z
e
Z
= dt ei(Eb Ea )t d3 re {b (re )e a (r)} h1n |A(re , t)|0i. (7.50)
m
7.2. ATOMIC DECAY 205
Since
X 1 0
h1n |A(re , t)|0i =
3
n0 h1n |an0 0 |0ieikn0 xe
n0 0
2n0 L
X 1 0
=
3
n0 h1n |an0 0 |0ieikn xe
n0 0
2n0 L
1
= 3
n h1n |1n ieikn0 xe
2n L
1
= n ein tkn re . (7.51)
2n L3
1
Sba = i2(Eb + n Ea ) Mba , (7.52)
L3
where the decaya amplitude Mba is
e 1
Z
Mba = i d3 re eikn re b (re )n e a (r) . (7.53)
m 2n
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)
where
Z
pba = i d3 re b (re )e a (re ). (7.56)
we have
|Sba (T /2, T /2)|2
Wba = lim
T T
1 1
= lim 2(Eb + n Ea )T 3 |Mba |2
T T L
1
=2(Eb + n Ea ) 3 |Mba |2 (7.58)
L
Then
2 e2
Wba = (Eb + n Ea ) 2
|n pba |2 . (7.59)
V 2n m
Summing over all final photon states to get the total a b decay rate gives
X 2 e2
Wba = (Eb + n Ea ) | pba |2
n,
V 2n m2 n
X X 2 e2
= (Eb + n Ea ) 2
|n pba |2 . (7.60)
n
V 2n m
Using
V
X Z
d3 k (7.61)
n
(2)3
XZ d3 k e2
Wba = 2
(Eb + n Ea ) 2
|n pba |2
(2) 2m
e2 X d3 k
Z
= 2 (Eb + n Ea )|n pba |2
2m (2)2
e2 X d3 k
Z Z
2
= 2 |k| d|k| d (Eb + Ea )|n pba |2 . (7.62)
2m (2)2
Then the total rate for the decay of the state a into b is
e2
4
Wba = |pba |2 . (7.65)
4 3m2
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
HI = h : : (7.68)
where the required ordered product will be explained in next section. The term linear in the inter-
action Hamiltonian in the Smatrix. It is
S (1) =
Z
= ih d4 x : : . (7.69)
Z
(1)
S = ih d4 x : ( + + )(+ + )(+ + ) : . (7.70)
208 CHAPTER 7. FEYNMAN RULES
Lint = h : + + (+ + ) (+ + ) :
= : + + + + + + + + + + + + + +
+ + + + + :
= + + (7.71)
To check that only the ordered terms are different from zero we can analyse the full terms for
initial and final states defined as |ii = |0 , 0 , 1 i y hf | = h1 , 1 , 0 |.
+ |n i |n 1 i hn |+ hn + 1 | (7.72)
|n i |n + 1 i hn | hn 1 | (7.73)
Lint = h (7.74)
Que en terminos de las componentes + y de los campos se puede expresar como
De desarrollo del langrangiano en las componentes de los campos, vemaos que terminos con-
tribuyen al elemento de matriz
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
h1 , 1 , 0 | |0 , 0 , 1 i h0 , 0 , 0 |0 , 0 , 2 i = 0
7.3. YUKAWA INTERACTION 209
The only term that contributes to the matrix element of the process is
Z
ih d4 x + . (7.75)
r
1
|B(p)i a |0i (7.76)
V p
(2)3 3
hB(p)|B(p0 )i = (p p0 )
V
0 0 (2)3
he (p, s)|e (p , s )i = ss0 3 (p p0 )
V
3
(2)
he+ (p, s)|e+ (p0 , s0 )i = ss0 3 (p p0 ) (7.78)
V
As established in Sec. 3.1, it is convenient to work in the discrete limit where (??)
V
3 (0) = . (7.79)
(2)3
210 CHAPTER 7. FEYNMAN RULES
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. (??), and taking into account that ap |0i = 0, we
have
1
Z
+ (x)|B(k)i = d3 p 3
p ap eipx |B(k)i
b
(2) 2p
1 ipx 1
Z
= d3 p 3
p a
b p e b ak |0i
(2) 2p V
1
Z
= d3 p 3
p eipx [b ak ]|0i .
ap , b (7.80)
(2) 2p V
By usinbg the commutation relations in eq. (??) we have
(3) (p k) ipx
Z
+ (x)|B(k)i = d3 p p e |0i (7.81)
2p V
1
+ (x)|B(k)i = eikx |0i (7.82)
2k V
Similarly, we have
1
+ (x)|B(k)i = 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 , (7.83)
2Ep0 V
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
he (p, s)| (x) =h0| p us (p)eipx
2Ep V
1 0
he+ (p0 , s0 )| (x) =h0| p vs0 (p0 )eip x , (7.84)
2Ep0 V
7.4. WICK THEOREM 211
In the lowest order the only term which contributes to the matrix element is the term shown in
Eq. (7.75) The matrix element at first order in Eq. (7.95), between the initial and the final state is
then
Z
(1)
Sf i = ih d4 x e (p)e+ (p0 ) + B(k) .
(7.85)
we obtain
" #
(1) 1 1 1
Sf i = p p (2)4 4 (k p p0 ) [(ih)us (p)vs0 (p0 )] (7.88)
2k V 2Ep V 2Ep0 V
Comparing with Eq. (??) we have therefore that the relativistic matrix element is
iMf i = (ih)us (p)vs0 (p0 ) , (7.89)
and everything else is the history presented in Chapter 6.
(i)n
X Z Z Z
S =1 + dt1 dt2 . . . dtn T{HI (t1 )HI (t2 ) . . . HI (tn )} , (7.92)
n=1
n!
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
(i)n
X Z Z
=1 + d4 x1 d4 x2 . . . d4 xn hf | T{HI (x1 )HI (x2 ) . . . HI (xn )}|ii . (7.94)
n=1
n!
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
T {(x1 )(x2 )} =: (x1 )(x2 ) : + (x1 )(x2 ) (7.96)
The same expression can be obtained for fermions. 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 ) : + (7.97)
7.5 Scattering
From the previous calculation we have
(i)n
Z Z
(n)
S = d4 x1 d4 x2 . . . d4 xn T{HI (x1 )HI (x2 ) . . . HI (xn )} . (7.98)
n!
The relevant term for the scattering
is
(i)2
Z Z
(2)
S = d4 x1 d4 x2 T{HI (x1 )HI (x2 )}
2!
(ih)2
Z Z
= d4 x1 d4 x2 T{: ()x1 ()x2 }
2!
(ih)2 (ih)2
Z Z Z Z
4 4
= d x1 d x2 : ()x1 ()x2 : + d4 x1 d4 x2 : ()x1 ()x2 : +
2! 2!
(7.100)
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,
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
(ih)2
Z Z
(2) + +
S (e e e e ) = d4 x1 d4 x2 (x1 )(x2 ) : ()x1 ()x2 : (7.101)
2!
The Wick contraction can be written as:
since
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 (7.105)
with
1 1
Z Z
+ (x) = dp3
3
p ap eipx
b (x) = d3 p 3
p ap eipx ,
b (7.106)
(2) 2p (2) 2p
we have
d3 p1 d3 p2
Z Z
ip1 x1
h0|T ((x1 )(x2 ))|0i =(t1 t2 )h0| p a p1 e p ap eip2 x2 |0i
(2)3 2Ep1 (2)3 2Ep2 2
d3 p2 d3 p1
Z Z
ip2 x2
+ (t2 t1 )h0| p a p2 e p ap eip1 x1 |0i
(2)3 2Ep2 (2)3 2Ep1 1
d3 p d3 p 2
Z Z
=(t1 t2 ) p 1 p eip1 x1 eip2 x2 h0|ap1 ap2 |0i
6
(2) 2Ep1 2Ep2
d3 p d3 p1
Z Z
+ (t2 t1 ) p 2 p eip2 x2 eip1 x1 h0|ap2 ap1 |0i
6
(2) 2Ep2 2Ep1
(7.107)
d3 p d3 p2
Z Z
=(t1 t2 ) p 1 p eip1 x1 eip2 x2 h0|ap1 ap2 |0i
(2)6 2Ep1 2Ep2
d3 p d3 p1
Z Z
+ (t2 t1 ) p 2 p eip2 x2 eip1 x1 h0|ap2 ap1 |0i
(2)6 2Ep2 2Ep1
(7.108)
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. (7.109) 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
7.5. SCATTERING 215
: ()x1 ()x2 := (x1 ) (x2 )+ (x1 )+ (x2 ) . (7.110)
The relevant Smatrix element then reads
(ih)2
Z Z
(2)
Sf i = d4 x1 d4 x2
2
he (p01 )e (p02 )|iF (x1 x2 ) (x1 ) (x2 )+
(x1 )+ (x2 )|e (p1 )e (p2 )i
(ih)2
Z Z
= d4 x1 d4 x2 iF (x1 x2 )
2
he (p01 )e (p02 )| (x1 ) (x2 )+
(x1 )+ (x2 )|e (p1 )e (p2 )i
(ih)2 d4 q
Z Z Z
= d4 x1 d4 x2 iF (q)eiq(x1 x2 )
2 (2)4
he (p01 )e (p02 )| (x1 ) (x2 )+
(x1 )+ (x2 )|e (p1 )e (p2 )i (7.111)
The two particle Fock state is, after proper normalization
1
|e (p1 )e (p2 )i = as (p2 )as (p1 )|0i (7.112)
V
Therefore
d3 k d3 k
Z Z
0
+ (x1 )+ (x2 )|e (p1 )e (p2 )i = u (k)u (k0 )eikx1 eik x2
2Ek V 2Ek0 V
0
as (k)as (k )as (p2 )as (p1 )|0i (7.113)
1 1
+ (x1 )+ (x2 )|e (p1 )e (p2 )i = p p
2Ep1 V 2Ep2 V
u (p1 )u (p2 )eip1 x1 eip2 x2 u (p2 )u (p1 )eip2 x1 eip1 x2 |0i
(7.114)
Following similar steps, we find
1 1
he (p01 )e (p02 )| (x1 ) (x2 ) = p 0 p 0
2Ep1 V 2Ep2 V
h 0 0 0 0
i
h0| u (p01 )u (p02 )eip1 x1 eip2 x2 u (p02 )u (p01 )eip2 x1 eip1 x2
(7.115)
216 CHAPTER 7. FEYNMAN RULES
As expected, the final result can be written in term of three different factors: the momentum con-
servation, normalization, and the relativistic amplitude
!
(2)
X X Y 1 Y 1
Sf i = i(2)4 4 pi p0f q Mf i (7.116)
i=1,2 f =1,2 i=1,2
2E i V f =1,2 2E 0
V 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 )
(7.117)
In the limit q 2 m2
1
F = (7.119)
m2
h2 0
u (p2 )u (p1 )u (p01 )u (p2 ) u (p01 )u (p1 )u (p02 )u (p2 )
Mf i = 2
m
h2
= 2 [u(p02 )u(p1 )u(p01 )u(p2 ) u(p01 )u(p1 )u(p02 )u(p2 )] (7.120)
m
For one interaction of type we should have
h2
Mf i = 2
[u(p02 )u(p1 )u(p01 )u(p2 ) u(p01 )u(p1 )u(p02 )u(p2 )] (7.121)
m
For the interaction of a fermion pair with W , we know from the standard model Lagrangian [1],
that
g2
(1 5 ) (7.122)
2 2
Therefore in this case
= (1 5 ) (7.123)
7.5. SCATTERING 217
e (p1) x1 e (p2)
B(p1 p2 )
x2
e (p2) e (p1)
e (p1) x1 e (p1)
B(p1 p1 )
x2
e (p2) e (p2)
2
For p MW the analysis is similar to the previous one with
g2 GF
2
= (7.126)
8MW 2
After the replacement GF / 2 h2 /m2 , we have
(2) 1 1 1 1
Sf i = i(2)4 4 (p1 + p2 p01 p02 ) p 0 p 0 Mf i (7.127)
2E1 V 2E2 V 2E1 V 2E2 V
where
GF
Mf i = ue (p02 )ue (p1 )u (p01 )u (p2 ) (7.128)
2
The corresponding Feynman diagram is shown in Fig. 7.3 Therefore we have
GF
Mf i = ue (p02 ) (1 5 )ue (p1 )u (p01 ) (1 5 )u (p2 ) (7.129)
2
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
s m2
d 1
= |M|2 (7.131)
d 64 2 s s m2e
7.5. SCATTERING 219
e (p1) x1 e(p2)
W (p1 p2 )
(p2) x2 (p1)
The center of mass (CM) frame is defined by the condition in Eq. (??):
p1 + p2 = 0 (7.132)
(4) (p1 + p2 p01 p02 ) = (3) (p1 + p2 p01 p02 )(E1 + E2 E10 E20 ) (7.133)
implies
(
p1 = p2
CM
p1 + p2 p01 p02 = 0 (7.134)
p01 = p02
Moreover
s = E1 + E2 (7.135)
In the CM frame
s = (E1 + E2 )2
q q 2
2 2 2 2
= p1 + m1 + p2 + m2
q q 2
2 2 2 2
= p1 + me + p1 + me (7.136)
220 CHAPTER 7. FEYNMAN RULES
Therefore
E2 = |p1 | (7.137)
We already have the expression for |p1 | as given in eq. (??). In this case m2 = 0, and m1 = me , so
that
s m2
|p1 | = e (7.138)
2 s
From (7.137)
s =E1 + E2
=E1 + |p1 | (7.139)
E1 = s |p1 |
s + m2e
= s+
2 s
2s s + m2e
=
2 s
s + m2
= e (7.140)
2 s
Then, by using Eqs. (7.134), (7.137) and (7.138), and (7.140), 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 ) (7.141)
2
As p22 = p02 2 = 0, we have from function
(p1 + p2 )2 =(p01 + p02 )2
(p1 + p2 )2 =(p01 + p02 )2
2 2
p21 + 2p1 p2 + p22 =p01 + 2p01 p02 + p02
2
p21 + 2p1 p2 =p01 + 2p01 p02
m2e + 2p1 p2 =m2 + 2p01 p02 (7.142)
7.5. SCATTERING 221
1
p01 p02 = p1 p2 (m2 m2e ) (7.143)
2
1
p01 p02 = (s m2 ) (7.144)
2
Replacing back in Eq. (7.130) and then in Eq. (7.131) we have
s m2
d 1 1 1
= 64G2F (s m2e ) (s m2 ) (7.145)
d 64 s s me
2 2 2 2
d G2 (s m2 )2
= F2 (7.146)
d 4 s
G2F (s m2 )2
= (7.147)
s
Note that s.
222 CHAPTER 7. FEYNMAN RULES
Chapter 8
Neutrinos
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) ,
T
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
Any SU (2) spinor , T i2 satisfy that
both 1 2 and T1 i 2 are invariant.
both 1 2 and T1 i 2 transform as vectors.
223
224 CHAPTER 8. NEUTRINOS
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
(8.1)
0
H L
= 2 Lc ecL H 0 H +
H + eL
1 1 2 4 d3 p1 d3 p2 d3 p3
d = |M| (P p 1 p 2 p 3 )
(2)5 2M 2E1 2E2 2E3
3
1 1 d p1 d3 p2 d3 p3
Z
2 4
= |M| (P p1 p2 p3 )
(2)5 2M 2E1 2E2 2E3
(9.1)
We use
1
C = (2 2 1 1) = 2 (9.3)
2
The first factor is for the initial average and the factor are for the number of spin states of , e and
the two neutrinos.
225
226 CHAPTER 9. THREE BODY DECAYS
we have
d3 p1 d3 p2 d3 p2
Z Z
4
(P p1 p2 ) = (E E1 E2 )
2E1 2E2 4E1 E2
3 3
d p1 d p2 p22 d|p2 |d
Z Z
4
(P p1 p2 ) = (E E1 E2 ) (9.6)
2E1 2E2 4E1 E2
Since |p1 | = |p2 | we have
ig 2
M= 2
u3 (1 5 )u1 u4 (1 5 )v2
8MW
iGF
= u3 (1 5 )u1 u4 (1 5 )v2 (9.20)
2
9.1. MUON DECAY 229
L = u3 u3 (1 5 ) u1 u1 (1 5 )
= (u3 u3 ) (1 5 ) (u1 u1 ) (1 5 )
Using
X X
u(p, s)u(p, s) = (p + m) v(p, s)v(p, s) = (p m) (9.26)
s s
X
L = Tr [(p 3 ) (1 5 )(p 1 + m ) (1 5 )]
s
h i
=p3 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 )
+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)
X
L =p3 p1 Tr 5 5 + 52
s
=2p3 p1 Tr [ (1 5 )] (9.28)
Similarly
X
M = 2p4 p2 Tr (1 5 )
(9.29)
s
9.1. MUON DECAY 231
g4
|M|2 = 4p3 p1 p4 p2 Tr [ (1 5 )] Tr (1 5 )
4
64MW
g4
= 4
4p3 p1 p4 p2 (64 )
64MW
g4
= 4
4 64(p3 p4 )(p1 p2 )
64MW
4g 4
= 4 (p3 p4 )(p1 p2 )
MW
2
g2
=4 8 2
(p3 p4 )(p1 p2 )
8MW
2
GF
=4 8 (p3 p4 )(p1 p2 )
2
=128 G2F (p3 p4 )(p1 p2 )
G2
=256 F (p3 p4 )(p1 p2 ) . (9.30)
2
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 d3 p3 1 X d3 p 2 d3 p4
1 2 4
d = |M| (p 1 p 2 p 3 p 4 )
(2)5 2E1 2E3 2 2E2 2E4
3
1 1 X 1 d p2 4 3
d p3 d3 p4
= |M|2 (p 1 p 2 p 3 p 4 )
2E1 2 (2)5 8E2 E3 E4
4 3
1 4g 1 d p2 4 d3 p3 d3 p4
= 4
(p 1 p 2 )(p 3 p 4 ) (p 1 p 2 p 3 p 4 )
2 MW (2)5 2E1 2E2 2E3 2E4
4
2g
= p p d3 p4 I (9.31)
5
16(2) MW 4
E1 E4 1 4
d3 p2 d3 p3
Z
I = p3 p2 4 (p p2 p3 ) . (9.32)
E2 E3
232 CHAPTER 9. THREE BODY DECAYS
d3 p2 d3 p3 p2
Z
2
g I = 4A + p B = p3 p2 4 (p p2 p3 ) = 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 (9.38)
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 ) (9.39)
Z
0 2
= (p ) d3 p2 (p0 2E2 )
2
1 p0 p2 p2
Z
2 2
= (p) dE2 E2 d ( E2 ) = 4
2 2 2 2
4
p
= (9.40)
2
where the usual tricks have been used to simplify the integrals, using the delta function inside.
Therefore
p2
A = ( B) (9.41)
4
p4 p4
( B) + p4 B =
4 2
B
+B =
4 4 2
3B
=
4 4
B = (9.42)
3
234 CHAPTER 9. THREE BODY DECAYS
p2
A= ( )
4 3
p2 2
= ( )
4 3
p2
= (9.43)
6
g p2 + 2p p .
I = (9.44)
6
Substituting back in eq. (9.31) we have
2g 4
d = 4
p1 p4 (g p2 + 2p p )d3 p4
16 6(2) MW E1 E4
5
2g 4
d = 4
[(p1 p4 )p2 + 2(p p1 )(p p4 )]d3 p4
16 12(2)4 MW E1 E4
2g 4
d = 4
[(p1 p4 )p2 + 2(p p1 )(p p4 )]d3 p4 (9.45)
192(2)4 MW E1 E4
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
p p4 =m E4 E42 + p24 = m E4 m2e
p24 = m2e = E42 p24 p24 = E42 m2e
|p4 | =(E42 m2e )1/2
d|p4 | 1 2E4 E4
= 2
=
dE4 2 (E4 me )2 1/2 |p4 |
E4
d|p4 | = dE4
|p4 |
d3 p4 =p24 d|p4 | d = |p4 |E4 dE4 d (9.47)
9.1. MUON DECAY 235
Therefore
4g 4 m5 1 2
Z
= 4
x [3 2x] dx
192(2)3 MW 8 0
4g 4 m5 1
= 4
192(2)3 MW 8 2
5
g 4 m
= 3 4
192 8MW 4
g4 2
= m5
64MW 192 3
4
G2 2
= F 3
m5
2 192
G2
= F 3 m5 (9.54)
192
4g 4
E42 3m2 4m E4 dE4
d = 3 4
192(2) MW
g4 4 2
2
= 4
E 4 3m 4m E4 dE4
32MW 6(2)3
G2F 2 2
2
= E 4 3m 4m E4 dE4
3 8 3
G2
4 E4
= F 3 E42 3m2 1 dE4 (9.55)
12 3 m
G2F 2 2
d 4E4
= m E 3 (9.56)
dE4 12 3 4 m
9.2. THREE BODY DECAYS IN RADIATIVE SEESAW 237
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 (9.62)
M2
where
Hij = hj hi + hi hj (9.63)
iHij
M = 2
[u3 (1 5 )u1 ] [u4 (1 5 )v2 ]
M
iHij
= [u1 (1 + 5 )u3 ][v2 (1 + 5 )u4 ] . (9.64)
M2
9.2. THREE BODY DECAYS IN RADIATIVE SEESAW 239
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
2
Hij
|M|2 = 64(p3 p4 )(p1 p2 )
M4
2
Hij
|M|2 = 4 64(p3 p4 )(p1 p2 )
4M4
(9.70)
240 CHAPTER 9. THREE BODY DECAYS
In this way, comparing with eq. (9.30), the results for the moun decay can be directly used after the
replacements
2
g4 Hij
4
64MW 4M4
2
g4 16Hij
4
MW M4
m Mj
me Mi
x= . (9.71)
m Mj
The decay width is according eq. (9.58)
2
16Hij 4 Mj6
(Nj l l Ni ) = I (x)
M4 192(2)3 Mj 16
(hj hi + hi hj )2 Mj5
= I (x) (9.72)
2M4 192 3
where
Mi
I(x) = 1 8x2 24x4 ln(x) + 8x6 x8 , x= . (9.73)
Mj
Similarly the decay through 0 is
(hj hi + hi hj )2 Mj5
(Nj Ni ) = I (x) (9.74)
2M40 192 3
In this way, for example for N2
X X h22 h21 + h21 h22 + 2h2 h1 h2 h1 M 5
(N2 l l+ N1 ) = 4
2
3
I (x)
2M 192
h22 h21 + h21 h22 + 2h2 h1 h2 h1 M25
= I (x) (9.75)
2M4 192 3
In general
h2i h2j + (hi hj )2 Mj5
X Mi
(Nj l l+ Ni ) = I
M4 192 3 Mj
h2i h2j + (hi hj )2 Mj5
X Mi
(Nj Ni ) = I (9.77)
M40 192 3 Mj
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
242 CHAPTER 9. THREE BODY DECAYS
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 ) Br( + N2 )
0.61 0.37
Br( + N1 ) Br( + N1 )
Br( + N1 ) 0.51 Br( + N2 ) 0.19 Br( + N3 ) 0.30 (9.84)
X 1
Br(N3 l l+ N1 ) = 0.406
1 + 0.0812 + 0.0615 + 1.320
X
Br(N3 N1 ) 0.536
X
Br(N3 N2 ) 0.030
X
Br(N3 l l+ N2 ) 0.025 (9.87)
9.2. THREE BODY DECAYS IN RADIATIVE SEESAW 243
We have that
" #
M25
2 2 2 M1 1 1
tot (N2 ) = h1 h2 + (h1 h2 ) I +
192 3 M2 M4 M40
X
vis (N2 N1 ) (N2 l l+ N1 )
So that
h21 h22 + (h1 h2 )2 7.067 1012 h21 h23 + (h1 h3 )2 1.321 1011
h22 h23 + (h2 h3 )2 6.465 1012 (9.90)
+ 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
with
h 2 i1/2
1/2 M2 , Mj2 , m2 = M2 + Mj2 m2 4M2 Mj2 (9.96)
Therefore
2 2
Mj
3h2j M
1 2 j = 2, 3
X
( + l+ Nj ) M
16
1 j=1
Mj2
2
3hj M 12 2
j = 2, 3
M
16 1 j=1
(9.98)
In this way
X
tor ( + ) = ( + l+ Nj )
j
M22 M32
3M 2 2 2
h1 + h2 1 2 2 + h3 1 2 2 (9.99)
16 M M
246 CHAPTER 9. THREE BODY DECAYS
+ +
Br( + Nj )
P
( l Nj )
= P
Br( + Ni ) ( l Ni )
+ +
a+b+c=1
b c 1
1+ + =
a a a
1
a= (9.101)
1 + b/a + c/a
In this way
1
Br( + N1 ) = Br( + N3 ) Br( + N3 )
(9.102)
1+ Br( + N1 )
+ Br( + N1 )
From eq.
Br (N3 N1 )
= (9.103)
N2 )
Br(N3 |{z}
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 Publica-
tion)
[4] A. Lahiri and P. B. Pal, A first book of quantum field theory, Harrow, UK: Alpha Sci. Int.
(2005) 380 p
[9] Michael E. Peskin and Daniel V. Schroeder. An introduction to Quantum Field Theory, Addison-
Wesley Publishing Company(1995), p. 101
[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
247
248 BIBLIOGRAPHY
[15] D. Aristizabal Sierra, J. Kubo, D. Restrepo, D. Suematsu and O. Zapata, Radiative seesaw:
Warm dark matter, collider and lepton flavour violating signals, Phys. Rev. D 79 (2009) 013011
[arXiv:0808.3340 [hep-ph]].
[16] Relativistic Quantum Mechanics and Field Theory, Franz Gross, John Wiley & Sons, INC. 1993
[17] D. McMahon, Quantum field theory demystified: A self-teaching guide, SPIRES entry New
York, USA: McGraw-Hill (2009) 299 p
[18] R. P. Feynman, QED. The Strange Theory Of Light And Matter, Princeton, Usa: Univ. Pr.
( 1985) 158 P. ( Alix G. Mautner Memorial Lectures)
[19] A. Semenov, LanHEP - a package for the automatic generation of Feynman rules in field
theory. Version 3.0, Comput. Phys. Commun. 180 (2009) 431 [arXiv:0805.0555 [hep-ph]], http:
//feynrules.phys.ucl.ac.be/.
[20] N. D. Christensen and C. Duhr, FeynRules - Feynman rules made easy, Comput. Phys. Com-
mun. 180, 1614 (2009) [arXiv:0806.4194 [hep-ph]], http://feynrules.phys.ucl.ac.be/.
[21] Christianto V., Smarandache F., and Lichtenberg F., A Note of Extended Proca Equations
and Superconductivity, Progress in Physics, 1 (2009) 40 [viXra:1003.0054]