Добавил:
Опубликованный материал нарушает ваши авторские права? Сообщите нам.
Вуз: Предмет: Файл:

Supersymmetry. Theory, Experiment, and Cosmology

.pdf
Скачиваний:
79
Добавлен:
01.05.2014
Размер:
12.6 Mб
Скачать

R-parity breaking 189

Table 7.1 Best limits on some of the couplings λijk . The bounds are obtained from experimental results on the following processes: neutrinoless double beta decay [ββ0ν], charge current universality in the quark sector [Vud], semileptonic decays of mesons [B→ Xq τ ν¯], atomic parity violation [QW (Cs)].

 

Coupling

Constraint

Process

 

 

λ111

3.3 104(mq˜/100 GeV)2(mg˜/100 GeV)1/2

 

ββ0ν

 

λ

 

, k = 2, 3

0.02(m ˜

/100 GeV)

 

Vud

 

 

11k

 

dkR

 

 

 

 

 

 

τ ν¯

 

 

λ

0.12(m˜

/100

GeV)

B

X

q

 

 

333

bkR

 

 

 

 

 

 

 

λ1j1

0.04(mu˜jL /100

GeV)

QW (Cs)

We refer the reader to more extensive reviews on the subject (see for example [27]) for complete lists of constraints. It should be stressed that, whereas the bounds on the λijk couplings scale like me˜kR , other bounds may have a more complicated dependence on the masses of the supersymmetric particles, as can be seen for example from Table 7.1

in the case of λijk.

The main phenomenological consequences of allowing for R-parity violations are as follows.

First, supersymmetric particles need not be produced in pairs, which lowers the threshold for supersymmetric particle production. This is certainly a welcome feature for experimental searches. On the other hand, the fact that, at the time of writing, no supersymmetric particle has been found thus finds a more natural explanation in the context of R-parity conservation.

Table 7.2 gives, for each class of high energy colliders, the resonant production mechanisms allowed by the di erent types of R-parity breaking couplings.

Second, the LSP is not stable and can decay in the detector. We note that, since the LSP is not stable, cosmological arguments about it being charge and color neutral drop. It could for example be a squark, a slepton, a chargino... We give in Table 7.3 the direct decays of supersymmetric fermions through R-parity violating couplings. Charginos and neutralinos decay into a fermion and a virtual sfermion which subsequently decays into standard fermions through R-parity violating couplings: this yields a three-fermion final state.

For example, in the case of a nonvanishing λijk coupling, the sneutrino R-parity violating partial width reads18

Γ ν˜i j+ k=

1

λijk2 mν˜i .

(7.108)

16π

The corresponding mean decay length L reads

L

= βγ

100 GeV

 

107

2

 

,

(7.109)

1 cm

mν˜i

 

λijk

where γ is the Lorentz boost factor.

18If the dominant coupling is λijk , replace in the following formulas λ2ijk with 3λijk2 , where 3 accounts for color summation.

190 Phenomenology of supersymmetric models: supersymmetry at the quantum level

Table 7.2 s-channel resonant production of sfermions at colliders in the case of R-parity violation.

Collider

Coupling

Sfermion

 

 

Process

 

e+e

λ1j1

ν˜µ, ν˜τ

1+ 1→ ν˜j

j = 2, 3

ep

λ

d˜R , s˜R , ˜bR

1uj

d˜kR

j = 1, 2

 

1jk

˜

+

 

 

 

 

u˜L , c˜L , tL

1 dk → u˜jL

 

pp¯

λijk

ν˜e, ν˜µ, ν˜τ

dkd¯j → ν˜i

 

 

 

e,˜ µ,˜ τ˜

 

 

¯

˜

j = 1, 2

 

 

uj dk

iL

 

λ

d,˜ s,˜ ˜b

u¯

i

d¯

d˜

j = k

 

ijk

˜

 

j

k

 

 

 

¯

¯

→ u˜i

j = k

 

 

u,˜ c,˜ t

dj dk

Table 7.3 Direct decays of supersymmetric particles via R-parity violating couplings.

Supersymmetric

 

 

 

 

 

 

 

Couplings

 

 

 

 

particles

 

 

 

λijk

 

 

 

 

 

λijk

 

 

λijk

ν˜

 

 

 

+

 

 

 

d¯

 

d

kR

 

 

 

 

 

iL

 

 

 

jL kR

 

 

 

 

jL

 

 

 

 

 

˜

 

 

 

ν¯

 

 

 

u¯

jL

d

kR

 

 

 

 

iL

 

 

 

jL

kR

 

 

 

 

 

 

 

 

 

 

 

ν˜

 

 

 

+

 

 

 

 

 

 

 

 

 

 

 

 

 

 

jL

 

 

 

iL kR

 

 

 

 

 

 

 

 

 

 

 

 

 

 

˜

 

 

 

ν¯

 

 

 

 

 

 

 

 

 

 

 

 

 

 

jL

 

 

 

iL

kR

 

 

 

 

 

 

 

 

 

 

 

 

 

 

˜

ν

iL

, ν

jL

 

 

 

 

 

 

 

 

 

 

 

 

 

kR

 

 

jL

iL

 

 

 

 

 

 

 

 

 

 

 

 

 

u˜iR

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

¯

¯

 

 

 

 

 

 

 

 

+ d

 

 

 

djR dkR

u˜

jL

 

 

 

 

 

 

 

 

kR

 

 

 

 

 

 

 

 

 

 

 

 

 

 

iL

 

 

 

 

 

 

˜

 

 

 

 

 

 

 

 

ν¯iL dkR

 

 

 

 

djL

 

 

 

 

 

 

 

 

 

 

 

¯

˜

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

djR

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

u¯iR dkR

d˜

 

 

 

 

 

 

ν

iL

d

jL

, u

jL

u¯

iR

d¯

kR

 

 

 

 

 

 

 

 

 

 

 

iL

 

jR

χ0

i+ν¯j k, iνj k+, i+u¯j dk, iuj d¯k, u¯id¯j d¯k, uidj dk

 

 

ν¯i j+ k, νi jk+

ν¯id¯j dk, νidj d¯k

 

 

 

χ+

i+ j+ k, i+ν¯j νk, i+d¯j dk, i+u¯j uk, uidj uk, uiuj dk,

 

 

 

+

 

+

 

¯

 

 

 

 

 

 

¯

¯ ¯ ¯

 

 

ν¯i j

νk, νiνj k

ν¯idj uk

, νiuj dk

didj dk

Similarly, the partial width for a pure photino neutralino decaying with λijk is [99]

Γ = λ2

α

 

Mγ˜5

(7.110)

 

 

 

,

 

 

ijk 128π2

 

m4˜

 

 

 

 

 

 

where m˜ is the mass of the virtual slepton. The corresponding decay length is then

 

 

 

 

 

 

 

 

 

 

 

 

found to be

 

 

4

100 GeV 5

105

2

 

 

L

m˜

 

 

 

 

= 0.3γ

 

 

 

 

 

 

 

.

(7.111)

 

1 cm

100 GeV

Mγ˜

λijk

The issue of phases 191

Table 7.4 Charges of the fields and of the parameters under the symmetries U (1) and U (1)R of the minimal supergravity model.

 

 

 

 

 

Fields

 

 

 

 

 

Parameters

 

 

 

 

 

H

1

H

2

QU c

QDc

LEc

 

λ M

1/2

A

0

B µ

 

 

 

 

 

 

 

 

 

 

 

 

 

µ

 

 

U (1)

1

1

1

1

1

0

 

0

0

 

2

2

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

U (1)

1

1

1

1

1

1

 

2

2

 

2

0

R

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Thus, for couplings of order 105, one expects displaced vertices in the detector. For much smaller couplings, the neutralino decays outside the detector whereas for larger couplings, the primary and displaced vertices cannot be separated.

7.7The issue of phases

In our discussions above, we have not paid su cient attention to the fact that the parameters of a supersymmetric theory are complex. The presence of nontrivial complex phases leads to possible deviations from the standard phenomenology that has been presented until now.

These phases appear already at the level of the minimal supergravity model. We may take as parameters of this model19 M1/2, m0, A0 , Bµ and µ. One can redefine the gaugino and Higgs fields in order to make M1/2 and Bµ real. One is then left with two independent complex phases for A0 and µ, which one defines traditionally as

ϕA arg A0M1/2

, ϕB arg µBµM1/2

.

(7.112)

One may alternatively note [110] that, in the limit where both µ and the soft parameters Bµ, M1/2 and A0 vanish, the model acquires new abelian symmetries U (1) and U (1)R: the charges of the di erent superfields are given in Table 7.4. We note that the second symmetry is of a type known as R-symmetry, encountered in Chapter 4 or in Section 7.4.2: it does not commute with supersymmetry. Thus it is a symmetry of the scalar potential but not of the superpotential; similarly, gauginos transform whereas vector gauge fields are invariant. We will study in more details such symmetries in the next chapter.

If one allows the parameters of the model to transform as (spurion) fields, with charges given in Table 7.4, we may turn U (1) and U (1)R as symmetries of the full model. We note that we have the following invariant combinations:

M1/2µBµ, A0µBµ, A0M1/2.

Two of their phases are independent: they are ϕA and ϕB .

19This list is somewhat di erent from the usual one given for example in Section 6.8 of Chapter 6. We have replaced tan β by Bµ. Moreover, in Chapter 6, we assumed implicitly Bµ to be real and |µ| was fixed by electroweak radiative breaking: we were thus left only with an ambiguity with the sign of µ. Here it is the phase of µ which remains to be fixed.

192 Phenomenology of supersymmetric models: supersymmetry at the quantum level

These phases are constrained by experimental data, such as the electric dipole moment of the neutron. As an example, the down quark electric dipole moment receives a one-loop gluino contribution which is computed to be [56, 309]:

d

 

= m

 

 

3

|mg˜|

[ A

sin ϕ

 

+ µ

tan β sin ϕ

 

]

(7.113)

 

d 18π mq4˜

 

 

 

N

 

 

| |

 

A

| |

 

B

 

 

 

 

 

 

 

GeV

2

 

 

 

 

 

 

 

 

 

2

100

 

sin ϕA,B × 1023e.cm 6.3 × 1026e.cm,

 

 

 

m˜

 

where m˜ is a mass scale of the order of the gluino mass mg˜ or a squark mass mq˜, and A is the low energy value of the A-term (the experimental constraint is found in [144]). This shows that the ϕA,B have to be small. This need not be accidental. For example, if A0 and M1/2 arise from a single source of supersymmetry breaking, their phases might be identical, in which case ϕA = 0.

In the case of nonminimal models, other phases appear in the squark and slepton mass matrices. We will return to this question in Chapter 12 and see that a generic supersymmetric extension of the Standard Model has 44 CP violating complex phases!

Further reading

G.F. Giudice and R. Rattazzi, Theories with gauge-mediated supersymmetry breaking, Physics Reports 322 (1999) 419.

R. Barbier et al., R-parity violating supersymmetry, Physics Reports 420 (2005) 1.

Exercises

Exercise 1 We prove the cancellation of the one loop contributions to the magnetic operator in the supersymmetric QED model of Wess and Wess and Zumino [363], studied in Exercise 5 of Chapter 3.

Using Lorentz covariance, charge conservation and hermiticity, one may write the matrix element of the electromagnetic current in the Dirac theory in terms of

u¯2(p2)M µu1(p1) = u¯2

&γµF1(q2) − σµν

qν

(q2) + 5σµν

qν

(q2)' u1, (7.114)

 

F2

 

F3

m

m

where qµ = pµ2 − pµ1 and p21 = p22 = m2.

(a)Show that parity conservation implies under F3 = 0. We will assume that parity is conserved from now on (hence we do not take into account weak interactions).

(b)Since (p/ − m)u(p) = 0, one may use the projectors (p/ ± m)/2m and write the decomposition above as

(p/2 + m)M µ(p/1 + m) = (p/2 + m) &γµF1(q2) − σµν qmν F2(q2)' (p/1 + m). (7.115)

Deduce that one can write the Pauli form factor F2(q2) as

 

F2

(q2) = q2(−q2 + 4m2) Tr γµ

−q2 + 4m2

m

(p/2 + m)M µ(p/1 + m)

,

 

 

m2

q2 + 2m2

Pµ

 

(7.116)

where P ≡ p1 + p2.

F2(q2) (a) = 2 F2(q2) (b) = 2 F2(q2) (c) = · · ·

Exercises 193

(c)We now consider the supersymmetric QED model studied in Exercise 5 of Chapter 3. The Lagrangian is given by its equation (3.58) in terms of the two Majorana spinors Ψ1 and Ψ2 and their supersymmetric partners (we add mass terms to give these fields a common mass m). Show that the magnetic operator is necessary of

¯

σµν Ψ2F

µν

+ h.c. in this model and write explicitly the indices (1 or

the form Ψ1

 

2) on the fermion and scalar lines of the Feynman graphs of Fig. 7.2.

(d)Compute the respective contributions of diagrams (a), (b), and (c) of Fig. 7.2 to the Pauli form factor (7.116) and show that they cancel.

Hints:

(a)Under parity uipi) = γ0ui(pi) up to a phase (see equation (A.107) of Appendix A).

(b)Use (7.116) to compute Tr [γµ(p/2 + m)M µ(p/1 + m)] and Tr [Pµ(p/2 + m)M µ(p/1 + m)].

¯

(c) Because of the Majorana nature (see (B.40) of Appendix B), Ψiσµν Ψi = 0.

(d)

Exercise 2 Show that, in the case where the hidden sector gauge symmetry is broken at an intermediate scale MI , the expression (7.66) for the gaugino condensates is

replaced by

λ¯hλh 2

M 6e24π2 S+S¯ /b0 ,

(7.117)

 

 

 

 

 

where one will express the scale M in terms of MU , MI , and the one-loop beta function coe cient b0 (resp. b0) of the corresponding gauge group above (resp. below) the scale MI .

Hints: Write

 

 

 

1

 

 

 

 

 

b

 

Λ

c

 

 

 

 

0 =

 

 

 

 

 

+

0

 

ln

 

 

,

 

g2(M

)

8π2

M

I

 

 

 

 

 

 

 

 

 

 

 

I

 

 

 

 

 

 

 

 

 

 

 

 

 

1

 

=

 

1

 

 

 

+

 

b0

ln

 

MI

 

,

g2(M

)

g2(M

 

)

8π2

M

 

 

 

 

 

 

 

U

I

 

 

 

U

 

 

 

 

 

 

 

 

 

 

 

to obtain M/MU = (MI /MU )1(b0/b0).

Exercise 3 Using Appendix C, write in components the Lagrangian

S =

1

d4y d2θ S W αWα + h.c.,

4

where S = S + 2θψS + θ2FS is a chiral superfield.

Hints: Use for example Exercise 4, question (a) of Appendix C.

L =

1

 

µν

 

 

 

i

µν ˜

 

 

i

µ

¯

1

 

2

8

SF

 

Fµν

8

SF

Fµν +

2

Sλσ

µλ +

4

SD

 

+

1

FS λλ +

1

 

 

ψS λD +

i

 

(ψS σµν λ) Fµν

+ h.c.

 

2

 

2

 

4

2

2

194 Phenomenology of supersymmetric models: supersymmetry at the quantum level

 

Exercise 4 : Consider the potential Ve

of (7.24) where V (0) is given by (5.13).

 

(a) Express V /∂Si|min, i = 1, 2 in terms of the parameters of V (0), v1, v2 (vi

and

Si are defined in (7.34); the minimum considered is the one of Ve ).

 

 

(b) Express

 

 

2

e

and

 

 

 

2

 

e

in

(7.35) in terms of these parameters and the

 

MS

 

 

 

2MP

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2

V

 

 

 

 

 

 

 

 

 

 

 

 

 

derivatives

V

 

,

 

V

 

 

 

and

 

 

 

 

.

 

 

 

 

 

 

 

 

 

 

 

∂S ∂S

 

 

 

 

∂P ∂P

 

 

 

 

 

 

 

 

 

 

 

 

 

 

∂S

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

min

 

i j

 

 

 

 

 

 

 

i

j

min

 

 

 

 

 

 

 

 

 

(c) Show that

 

i

 

 

min

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MP2

e

 

 

 

 

sin

2

β

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

sin β cos β

m¯ A2 ,

 

 

 

 

 

 

 

 

 

 

 

= sin β cos β

 

 

cos2 β

 

 

 

 

where m¯ A is the loop-corrected pseudoscalar mass.

 

 

 

 

 

 

 

 

(d) Deduce (7.36).

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Hints:

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

(b)

 

 

2Ve

 

 

 

 

 

 

 

 

 

 

 

 

1 δij V

min

 

 

2V

min .

 

 

 

 

 

 

 

 

 

 

 

v1v2

 

+

 

 

 

 

 

∂Pi

∂Pj

min = −Bµ vivj

2 vi

 

 

∂Si

 

∂Pi∂Pj

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

(c)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MS2

 

e

 

2V

min + (1)i+j

 

 

MP2

 

e

2V

 

 

g2

+ g 2

.

 

ij

= ∂Si∂Sj

 

 

ij

∂Pi∂Pj

min +

 

2

vivj

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

8

Dynamical breaking. Duality

A motivation for supersymmetry is the stability of the electroweak breaking scale under quantum corrections arising from fundamental physics at a much higher scale (naturalness problem). One may however turn the argument around: in a theory with a very large fundamental scale, how does one generate a scale as low as the supersymmetry breaking scale (hierarchy problem)? Again, the supersymmetry breaking mechanism should be devised in order to provide a rationale for such a behavior.

We will argue in what follows that nonperturbative dynamics may be the answer: there are several examples where nonperturbative phenomena account for the generation of very large ratios of scales. This is referred to under the generic name of dynamical breaking. As we will see, supersymmetry provides new tools to control quantum fluctuations in the nonperturbative regime. One will uncover some duality relations between two regimes of a given theory, or two regimes of two di erent theories. This duality is often of a strong/weak coupling nature and is thus very promising to study strongly interacting theories.

The theories that we will consider have a dynamical scale below which some of the symmetries are broken. In the low energy regime, below this scale, the fields are usually composite bound states of the fundamental fields. This leaves the possibility that the fields of the supersymmetric theory at low energy (such as the MSSM studied in Chapter 5) are not fundamental. Supersymmetry may thus be realized at a preonic level.

8.1Dynamical supersymmetry breaking: an overview

8.1.1Introduction

The expression “dynamical breaking of a symmetry” refers to the spontaneous breaking of this symmetry when it is generated by a nonperturbative dynamics.

The standard example is quantum chromodynamics: keeping only the three light quark flavors u, d and s, QCD is described by the action

 

 

 

3

 

 

 

 

1

 

 

S =

d4x −

Tr F µν Fµν + f =1 q¯f

µDµ − mqf

qf

(8.1)

4

where Fµν is the gluon field strength and Dµqf the SU (3) covariant derivative of the quark field qf . In the limit of vanishing quark masses mqf 0, this action is invariant under a global symmetry SU (3)L × SU (3)R, known as the chiral symmetry (independent SU (3) rotations respectively on the left and the right chirality quark fields). Such

196 Dynamical breaking. Duality

a symmetry is not observed in the spectrum: it would predict either a massless proton or a chiral partner for the proton. It is therefore spontaneously broken: the QCD vacuum breaks SU (3)L × SU (3)R. It is indeed believed that the nonperturbative QCD dynamics yields nonvanishing quark condensates Uf g ≡ q¯fL qgR which break spontaneously SU (3)L × SU (3)R into a diagonal SU (3)V (identical SU (3) rotations for both quark chiralities), the symmetry which allowed Gell-Mann to classify hadrons. The

¯

octet of pseudoscalar mesons, π±, π0, K0, K0, K± and η, yields the corresponding Goldstone bosons1. They may be interpreted as a variation in space and time of the background values and are thus described by the SU (3) spacetime-dependent matrix

Uf g (xµ).

The proton mass mp is a consequence of the nonperturbative QCD dynamics:

mp Λ = MP e8π2/(bgs2)

(8.2)

where b is the one-loop beta function and gs the value of the strong coupling at the Planck scale MP . As noted in Chapter 1, this allows us to explain the hierarchy of scale between Λ (or mp) and MP . We note that the QCD Lagrangian does not contain any dimensionful parameter in the limit of vanishing quark masses. It is only at the quantum level that a low energy mass scale is generated.

In the case of supersymmetric models, we must face a similar problem, i.e. generate, through spontaneous supersymmetry breaking, masses for supersymmetric particles which are much smaller than the Planck scale (otherwise, the prime reason for introducing supersymmetry in the Hamiltonian – the problem of naturalness – is lost). It is thus natural to study thoroughly the possible rˆole of dynamical breaking.

As we will now see, supersymmetry turns out to be an advantage, in the sense that it helps to control nonperturbative e ects. This should not be seen as a complete surprise since supersymmetry was precisely devised to get a better handle on quantum fluctuations.

8.1.2The power of holomorphy

We have seen in the preceding chapters that holomorphy is a fundamental property of the superpotential which summarizes the interactions of chiral supermultiplets (the scalar components of which are denoted here generically by φ): W depends on φ but not on φ . This is related to the chiral nature of the fermion fields associated with the scalars φ. One may generalize this property to account also for the dependence in the couplings.

To be explicit, let us consider the traditional Wess–Zumino model discussed in some details in Chapters 1 and 3. The superpotential reads (cf. (3.17) of Chapter 3)

W (φ) = 1 2

+ 1

λφ3.

(8.3)

2

2

 

 

We will extend the notion of holomorphy by considering the couplings themselves as background fields. In other words, we write m = M and λ = L and require W to be a holomorphic function of the scalar fields M and L as well as of φ.

1Restoring quark mass terms, which explicitly break the symmetry, generates a nonvanishing mass for these bosons.

Dynamical supersymmetry breaking: an overview 197

The justification may be found in string theory where, as we will see in Chapter 10, there is a single fundamental scale, the string scale MS and all dimensionless parameters may be expressed in terms of vacuum expectation values of scalar fields. For example, in the weakly coupled heterotic string theory, the gauge coupling g and the Planck scale MP are expressed in terms of the vacuum expectation value of the string dilaton S:

1

=

S ,

(8.4)

 

g2

MP2 =

S MS2.

(8.5)

More generally, our requirement of generalized holomorphy for the superpotential is simply a translation of the fact that, in a supersymmetric Feynman diagram which involves the interactions described by the superpotential, there is no possibility to find a coupling λ or a mass m .

Once couplings are interpreted on the same footing as fields, one may attribute them quantum numbers. For example, the superpotential (8.3) is invariant under the abelian symmetries U (1) × U (1)R, with quantum numbers given in Table 8.1.

Let us pause for a moment to discuss the di erence of status of the two abelian symmetries. The U (1) symmetry is of a standard nature and commutes with supersymmetry: if (φi, ΨiL ) are the bosonic and fermionic components of a chiral multiplet of charge qi, they both transform under U (1) as:

φi → eiqiαφi, ΨiL → eiqiα ΨiL .

(8.6)

The superpotential W must be invariant under such a symmetry. On the other hand, U (1)R is known as a R-symmetry, a concept which we have encountered several times: it does not commute with supersymmetry and, as such, plays a central rˆole in all discussions of supersymmetry and supersymmetry breaking. One can show [see Section 4.1 of Chapter 4] that, in the case of N = 1 supersymmetry, the only symmetries which may not commute with supersymmetry are abelian symmetries. Under such a symmetry, fermion and boson components of a given multiplet (φi, ΨiL ) of charge ri transform di erently:

φi → eiriα φi

, φi → e−iriα φi

 

ΨiL → ei(ri1)α ΨiL ,

ΨicR → e−i(ri1)αΨicR

(8.7)

and the superpotential W has nonzero charge r = 2.

Table 8.1

 

U (1)

U (1)R

φ

1

1

m

2

0

λ

3

1

W

0

2

198 Dynamical breaking. Duality

This can be checked on the full Lagrangian term

 

1 2W

 

 

 

1 e2

ΨciR ΨjL → −

2

 

∂φi∂φj

2

 

ei(ri+rj )α

(3.30) of Chapter 3. For example2, the

2W

 

 

 

 

ei(ri+rj 2)α ΨciR

ΨjL

 

∂φi∂φj

 

is invariant because the superpotential W has a charge r = 2.

The superpotential (8.3) that we started with is the tree level form. One may expect quantum corrections, in particular at the nonperturbative level, that would completely change this form. However, quantum fluctuations respect the U (1) × U (1)R symmetry

and thus W (φ) has the general form

 

 

 

 

 

 

 

 

 

nφn+2

 

cn

λφ

n

 

Wnp(φ) =

 

cn

λ

≡ mφ2

 

 

.

(8.8)

n

mn−1

n

m

We may take the limit λ → 0, m → 0 with λ/m fixed. Since λ → 0, the perturbative

result (8.3) holds:

1

1

 

 

 

 

 

 

λφ

 

Wnp(φ) = 2

 

 

+

 

 

 

 

 

(8.9)

2

3

m

to all orders in λ/m. Hence c0 = 1/2, c1 = 1/3 and cn = 0 for n ≥ 2,

 

Wnp(φ) = 21 2 + 31

λφ2.

(8.10)

Thus, the superpotential is nonrenormalized, a well-known result which is shown to hold even at the nonperturbative level. Let us stress that holomorphy, extended to masses and couplings, has been the key assumption to prove this result, and R- symmetries are an important tool that can be used because the superpotential transforms nontrivially under them.

It is through such arguments, specific to supersymmetry, that one gets a better handle on nonperturbative dynamics. This is why dynamical breaking may be studied more thoroughly in the context of supersymmetry, as we will see in the following sections. Before doing so, let us stress that, in the context of N = 1 supersymmetry, arguments of holomorphy apply only to interactions (i.e. superpotential) and not to kinetic terms: there is still much freedom arising from wave function renormalization of the fields. One has to go to N = 2 supersymmetry in order to get a handle on this wave function renormalization as well, and to prove much more stringent results. This is what we will do at the end of this chapter.

8.1.3Flat directions and moduli space

We have stressed in Chapter 7 (Section 7.3) the importance of flat directions of the potential, i.e. valleys where the scalar potential vanishes. Since along these directions global supersymmetry is not broken, the corresponding degeneracy is not lifted by perturbative quantum e ects. Only nonperturbative e ects may lift these flat directions.

2[For a general treatment using superfields, see Section C.2.3 of Appendix C.]