PHY646 - Quantum Field Theory and the Standard Model

Even Term 2020

Dr. Anosh Joseph, IISER Mohali

LECTURE 46

Tuesday, April 14, 2020

(Note: This is an online lecture due to COVID-19 interruption.)

Topic: The Higgs Mechanism II: Non-Abelian Models.

The Higgs mechanism extends straightforwardly to systems with non-Abelian gauge symmetry.
It is not difficult to derive the general relation by which a set of scalar field vacuum expectation
values leads to the appearance of gauge boson masses. Let us work out the relation and then apply
it in a number of examples.
Consider a system of scalar fields φi that appear in a Lagrangian invariant under a symmetry
group G, represented by the transformation

φi → (1i αa ta )ij φj . (1)

It is convenient to write φi as real-valued fields, for example, writing n complex fields as 2n

real fields. Then the group representation matrices ta must be pure imaginary and, since they are
Hermitian, antisymmetric. Let us write
taij = iTija , (2)

so that the T a are real and antisymmetric.

If we promote the symmetry group G to a local symmetry, the covariant derivative on the φi is

Dµ φ = (∂µ − igAaµ ta )φ
= (∂µ + gAaµ T a )φ. (3)

Then the kinetic energy term for the φi is

1 1 1
(Dµ φi )2 = (∂µ φi )2 + gAaµ (∂µ φi Tija φj ) + g 2 Aaµ Abµ (T a φ)i (T b φ)i . (4)
2 2 2

Now let the φi acquire vacuum expectation values

hφi i = (φ0 )i , (5)

PHY646 - Quantum Field Theory and the Standard Model Even Term 2020

and expand the φi about these values. The last term in Eq. (4) contains a term with the structure
of a gauge boson mass,
1
∆L = m2ab Aaµ Abµ , (6)
2
with the mass matrix
m2ab = g 2 (T a φ0 )i (T b φ0 )i . (7)

This matrix is positive semi-definite, since any diagonal element, in any basis, has the form

m2aa = g 2 (T a φ0 )2 ≥ 0 (no sum). (8)

Thus, generically, all of the gauge bosons will receive positive masses. However, it may be that
some particular generator T a of G leaves the vacuum invariant

T a φ0 = 0. (9)

In that case, the generator T a will give no contribution to Eq. (7), and the corresponding gauge
boson will remain massless.
As in the Abelian case, the gauge boson propagator receives a contribution from toe Goldstone
bosons, which is necessary to make the vacuum polarization amplitude transverse. To compute this
contribution, we need the vertex that mixes gauge bosons and Goldstone bosons. This comes from
the second term in Eq. (4). When we insert the vacuum expectation value of the scalar field Eq.
(5), this term becomes
∆L = gAaµ ∂µ φi (T a φ0 )i . (10)

This interaction term does not involve all of the components of φ - only those that are parallel
to a vector T a φ0 for some choice of T a . These vectors represent the infinitesimal rotations of the
vacuum; thus the components φi that appear in Eq. (10) are precisely the Goldstone bosons. We
encountered the following Goldstone boson diagram for the Abelian case

kµ kν kµ kν
   
i
im2A g µν µ ν 2 µν
+ (mA k ) 2 (−mA k ) = imA g − 2 2 µν
= imA g − 2 . (11)
k k k

Now, using the fact that these bosons are massless, we can compute the counterpart, for the
non-Abelian case  i 
X 
gk µ (T a φ0 )j − gk ν
(T b
φ )
0 j (12)
k2
j

The sum runs over those components j with a nonzero projection onto the space spanned by
the T a φ0 , or equivalently well, over all j. See Fig. 1. This diagram is therefore proportional to the
mass matrix Eq. (7). Combining this expression with the contribution to the vacuum polarization

2/5
PHY646 - Quantum Field Theory and the Standard Model Even Term 2020

Figure 1: The Goldstone boson diagram for the non-Abelian case. It corresponds to
P  µ a i

ν b
j gk (T φ0 )j k2 − gk (T φ0 )j .

from Eq. (6), we find a properly transverse result

kµ kν
 
im2ab µν
g − 2 , (13)
k

where m2ab is given by Eq. (7). See Fig. 2.

kµ kν
Figure 2: The lading-order contributions to the vacuum polarization amplitude, im2ab g µν −


k2
.

Worked Example: SO(3) → SO(2) Breaking

Let us now apply this formalism to a specific example of non-Abelian gauge theories. There will be
one massive gauge boson for each broken-symmetry generator.
Consider an SO(3) gauge theory. We introduce 3 real scalars φi and the Lagrangian

1 a 2 1 1 λ
L = − (Fµν ) + (∂µ φi − igAaµ τija φj )2 + m2 φ2i − (φ2i )2 . (14)
4 2 2 4!

These scalars transform in the fundamental representation of SO(3). The potential is minimized
for
~ =v=
p
|hφi| 6m2 /λ. (15)

By an SO(3) transformation, we can pick the direction and phase so that hφ3 i = v and hφ1 i =

3/5
PHY646 - Quantum Field Theory and the Standard Model Even Term 2020

hφ2 i = 0. That is, without loss of generality, we take

   
* φ1 + 0
φ2  = 0 (16)
   

φ3 v

This vacuum is invariant under H = SO(2) ⊂ G = SO(3), which rotates φ1 and φ2 . Since
SO(2) has one generator and SO(3) has three, there will be two Goldstone bosons that are eaten
to form two massive gauge bosons. To see this explicitly, we can expand the Lagrangian in unitary
gauge. We find

1 2 a bµ
∆L = m A A
2 ab µ
1 2 a
= g (T φ0 )i (T b φ0 )i Aaµ Abµ
2
 
0
1 2 a b   a bµ
= g (0 0 v) T T 0 Aµ A . (17)
2
v

Upon using [Aaµ , Abµ ] = 0, we can symmetrize the matrix product T a T b . This gives
 
0
1 2 a b   a bµ
∆L = g (0 0 v) {T , T } 0 Aµ A . (18)
4
v

Plugging in the SO(3) generators,

     
0 0 0 0 0 1 0 −1 0
T 1 = 0 0 −1 , T 2 = i  0 0 0 , T 3 = 1 0 0 , (19)
     

0 1 0 −1 0 0 0 0 0

we see by explicit calculation that

 
0
a b  
(0 0 v) {T , T } 0
v

is only non-zero for a = b = 1 or a = b = 2. Thus,

1 a 2 1 2 1 1
L = − (Fµν ) + mA (Aµ Aµ + A2µ A2µ ), (20)
4 2

with m2A = g 2 v 2 , which describes two massive gauge bosons (corresponding to generators 1 and 2)
and one massless one (corresponding to the generator 3).

4/5
PHY646 - Quantum Field Theory and the Standard Model Even Term 2020

It is interesting to note that this model contains both massive and massless gauge bosons, with
the distinction between these bosons created by spontaneous symmetry breaking. If we interpret the
massive bosons as W bosons and the massless gauge boson a the photon, it is tempting to interpret
this theory as a unified model of weak and electromagnetic interactions. Georgi and Glashow
proposed this model as a serious candidate for the theory of weak interactions. However, Nature
chooses a different model, in which the spontaneously symmetry breaking pattern is SU (2)×U (1) →
U (1).

References
[1] M. E. Peskin and D. Schroeder, Introduction to Quantum Field Theory, Westview Press (1995).

5/5