Phenomenology of Trinification models
João Fonseca Seabra
Thesis to obtain the Master of Science Degree in
Engineering Physics
Supervisors: Prof. Doutor Filipe Rafael Joaquim
Prof. Doutor David Emanuel da Costa
Examination Committee Chairperson: Prof.a Doutora Maria Teresa Haderer de la Peña Stadler Supervisor: Prof. Doutor Filipe Rafael Joaquim Member of the Committee: Doutor Joaquim Inácio da Silva Marcos
November 2017
Acknowledgements
Besides all the effort that I had put into this work, it surely would not be the same withoutthe help of many people. First of all, I devote a special acknowledgement to Professor Filipe Joaquim, not only for his guidance and support during the development of this work, but also for being the person who really convinced me through his lectures that Particle Physics is an amazing subject. I am also very grateful to Professor David Costa for his guidance. Even after leaving IST, he was always available to answer my doubts and questions. I would also like to thank my colleagues in IST for all the great discussions and study sessions that I had the fun and the pleasure to share with them during these years. Finally, a big ”thank you” to my family for their everlasting patience and support.
i ii Resumo
A compreensão do fenómeno de violação da simetria Conjugação de carga e Paridade (CP) tem-se revelado como um grande desafio em Física de Partículas. De facto, uma forte evidência dequeo Modelo Padrão (MP) precisa de ser expandido reside na sua incapacidade de dar resposta a questões estreitamente relacionadas com violação de CP, tais como a assimetria bariónica do Universo. Na presente dissertação, estudamos extensões do MP com quarks vectorlike, dando particular ênfase aos modelos de Trinificação. Nestas Teorias de Grande Unificação, o sector fermiónico encontra-se associado a representações fundamentais do grupo E6, o que leva naturalmente ao aparecimento de um quark vectorlike em cada geração de fermiões. Como consequência da mistura que pode haver entre quarks vectorlike e quarks do MP, surgem novas fontes de violação de CP. Além disso, esta mistura afecta os acoplamentos dos quarks do MP a outras partículas, o que nos leva a analisar as restrições experimentais impostas aos quarks vectorlike. Discutimos também a possibilidade de encontrar violação espontânea de CP (VECP) em modelos de Trinificação. Ao contrário do MP onde violação de CP só pode ocorrer ao nível do Lagrangiano, al- gumas teorias que prevêem a existência de um maior número de campos escalares (tal como os modelos de Trinificação) permitem que a simetria CP possa ser quebrada espontaneamente pelo vácuo. Neste contexto, verificamos que a introdução de quarks vectorlike num modelo de Trinificação com VECP leva-nos a obter uma matriz Cabibbo-Kobayashi-Maskawa complexa, compatível com os resultados experimentais.
Palavras-Chave: Matriz Cabibbo-Kobayashi-Maskawa, Modelos de Trinificação, Quarks vector- like, Teorias de Grande Unificação, Violação de CP
iii iv Abstract
Understanding the phenomenon of violation of the symmetry Charge conjugation and Parity (CP) reveals itself as a great challenge in Particle Physics. In fact, a strong evidence that the Standard Model (SM) needs to be extended stems from its inability to provide an answer to questions deeply connected with CP violation, such as the baryonic asymmetry of the Universe. In this thesis, we study SM extensions with vectorlike quarks, giving special attention to Trini- fication models. In these Grand Unified Theories (GUTs), fermions are assigned to fundamental representations of the group E6, leading naturally to the appearance of one vectorlike quark in each generation of fermions. As a consequence of the mixing that may occur between vectorlike quarks and SM quarks, new sources of CP violation arise. Moreover, this mixing affects the couplings of SM quarks to other particles and, for this reason, we will look at the experimental constraints imposed on vectorlike quarks. We also discuss the possibility of having spontaneous CP violation (SCPV) in Trinification models. While in the SM CP violation can only occur at the Lagrangian level, some theories that predict the existence of a bigger number of scalar fields (such as Trinification models) allow CP to be broken spontaneously by the vacuum. In this context, we verify that the introduction of vectorlike quarks in a Trinification model with SCPV leads to a complex Cabibbo-Kobayashi-Maskawa matrix, in agreement with experimental data.
Keywords: Cabibbo-Kobayashi-Maskawa matrix, CP violation, Grand Unified Theories, Trinifi- cation models, Vectorlike quarks
v vi Contents
Acknowledgements i
Resumo iii
Abstract v
List of Figures ix
List of Tables xi
List of Abbreviations xiii
1 Introduction 1 1.1 The discrete symmetries C, P, CP and T ...... 2 1.2 CP violation in the SM ...... 4 1.3 From the SM to GUTs ...... 5
2 The Standard Model of Particle Physics 7 2.1 Field content and Lagrangian ...... 7 2.2 The Higgs mechanism ...... 10 2.3 Charged and Neutral interactions ...... 11 2.4 Fermion masses and mixing ...... 12 2.5 Open questions in Particle Physics ...... 16
3 SM extensions with Vectorlike Quarks 19 3.1 Vectorlike quark representations ...... 20
3.2 SM extensions with SU(2)L singlets of vectorlike quarks ...... 20 3.2.1 Mass matrix diagonalization ...... 23 3.2.2 Quark mixing and suppression of FCNCs ...... 25 3.2.3 Complex phases of the generalized CKM matrix ...... 25 3.3 Mixing with third generation of SM quarks ...... 26 3.4 Experimental status of vectorlike B quarks ...... 27 3.4.1 Production at the LHC and mass constraints ...... 27 3.4.2 Mixing constraints ...... 29
4 Trinification Models 31 4.1 Trinification group and field content ...... 32 4.2 Scalar Lagrangian and SSB ...... 34 4.3 Fermion masses and mixing ...... 42
vii 5 Concluding Remarks 47
Bibliography 49
A Feynman rules for Vectorlike Quarks 53
B Representations of the Trinification group 55
B.1 The G333 Generators ...... 55 B.2 24 and 27 Representations ...... 56
viii List of Figures
1.1 The status of weak interactions in 1966 ...... 1 1.2 Parity transformation ...... 2 1.3 Charge conjugation ...... 3 1.4 CP transformation ...... 3
2.1 Unitarity triangle ...... 16 2.2 Experimental constraints on the (ρ, η) plane to the CKM parameters ...... 17
3.1 Vectorlike B quark production ...... 27 3.1(a) Pair production ...... 27 3.1(b) Single production involving a Z boson ...... 27 3.1(c) Single production involving a W boson ...... 27 3.2 Branching ratios of the B quark ...... 28 3.3 Lower limits for the B quark mass in the branching ratio plane BR(B → W t) versus BR(B → Hb) ...... 28 3.3(a) Observed 95% Confidence Level ...... 28 3.3(b) Expected 95% Confidence Level ...... 28 3.4 Experimental constraints on the mixing between b and B quarks ...... 30
ix x List of Tables
2.1 The SM Gauge fields and their representation properties...... 7 2.2 Fermionic fields of the SM ...... 8 2.3 Fit results for the Wolfenstein parameters ...... 16
3.1 SM predictions and experimental results for the observables affected by changing the Zbb coupling ...... 29
4.1 Numerical values of the scalar potential and VEV parameters compatible with the
observation of a SM Higgs boson with mass mH ∼ 125 GeV ...... 40 4.2 Masses of quarks computed at the electroweak scale ...... 44 4.3 Best fits of the effective down-quark matrix of the Trinification model totheCKM parameters and down-quark masses ...... 46
xi xii List of Abbreviations
CC Charged Current
CKM Cabibbo-Kobayashi-Maskawa
CP Charge conjugation and Parity
EWSB Electroweak Symmetry Breaking
FCNC Flavor Changing Neutral Current
NC Neutral Current
QFT Quantum Field Theory
SCPV Spontaneous Charge conjugation and Parity Violation
SM Standard Model
SSB Spontaneous Symmetry Breaking
VEV Vacuum Expectation Value
xiii xiv 1 Introduction
While reflecting upon the status of weak interactions at the 1966 Berkeley Conference, Nicola Cabibbo used the cartoon shown in Fig. 1.1 to reinforce the idea that besides some good progresses were being made on the subject of weak interactions, there was a big problem to be solved.
Fig. 1.1: The status of weak interactions in 1966 according to N. Cabibbo (image taken from [1]).
The cartoon depicts pretty well how Charge-Parity (CP) violation was a big mystery of weak interac- tions two years after its discovery. But since then, more than half a century has passed and one may ask: Do we understand CP violation nowadays? We could start answering by saying that our knowledge on the subject has increased significantly. For many years we could only observe CP violation in the kaon system but today, we are able to observe it in the B-meson system too. Moreover, the explanation found in the Standard Model (SM) for CP violation in terms of the Cabibbo-Kobayashi-Maskawa (CKM) mechanism is in agreement with all measurements made up to date on those systems [2]. However, we should say at this point that the SM is not a complete theory and it leaves many open questions in Particle Physics, some concerned with CP violation. For example, we know that it occurs, but there is no definite explanation for its origin. On the other hand, it is well established that CP Violation is a crucial ingredient for baryogenesis [3], the process of dynamically generating the matter-antimatter asymmetry of the Universe. In the SM, the CKM mechanism fails to accommodate the observed asymmetry by several orders of magnitude [4]. We then conclude that the cartoon of Figure 1.1 is still relevant today, in spite of the discoveries we have made so far about CP violation. It also reveals the importance the subject of CP violation has. After all, if there was no matter-antimatter asymmetry in our Universe, we would not be here to
1 discuss it. The need of finding new sources of CP violation is the main motivation for us tostudySM extensions with vectorlike quarks. Even though there is no experimental evidence for their existence, there are several reasons for these particles to receive a lot of attention nowadays [5]. In a particular set of SM extensions known as Grand Unified Theories (GUTs), vectorlike quarks may arise naturally, leading to new sources of CP violation. This is what happens in Trinification models, which we will study in greater detail. The main purpose of this chapter is to introduce some of the most important concepts of this thesis. In the following section, we introduce the discrete symmetries C, P, CP and T, so that we can understand better what is CP violation and how it was discovered in the neutral kaon system. After that, we present briefly the development of the CKM mechanism over time and its importance for the construction of the SM. Since we will be mostly interested in Trinification models, the most fundamental properties of GUTs are introduced.
1.1 The discrete symmetries C, P, CP and T
In Quantum Field Theory (QFT), discrete symmetries under transformations of Parity (P), Charge conjugation (C) and Time reversal (T) play a fundamental role and our interest on them grew as we realized they could be broken. In order to better visualize these transformations, we will recreate their effect on M. C. Escher’s painting Day and Night whenever possible. For instance, we represent in Fig. 1.2 a Parity transformation, which corresponds to an inversion of all spatial coordinates, (x, y, z) → (−x, −y, −z).
Fig. 1.2: An example of a Parity transformation.
Notice that we obtain the image we would see of the painting reflected on a mirror. Although there is only one dimension involved, the inversion of all spatial coordinates in three dimensions can also be achieved through a mirror reflection, followed by a 180◦ rotation. Therefore, the question one might ask regarding parity symmetry is whether, for a given event, its reflected image can also occur in Nature or not. For a long time, physicists took for granted that all laws of physics conserve parity. However, this was questioned in 1956 by T. D. Lee and C. N. Yang [6] when they came up with the idea that weak interactions may violate that symmetry. Among several ideas to test their hypothesis, one based on the β-decay of Cobalt-60 was performed in 1957 by Wu et al. [7], who found a preferred direction for the emission of β radiation. This result clearly indicated that P may be violated, as Lee and Yang suggested. These two physicists were awarded with the Nobel prize in 1957 for their proposal of P violation in weak interactions. Symmetry under Charge conjugation means that particles and antiparticles behave in a similar way. In contrast with P and T, the C symmetry can only be defined in the context of a QFT, where antiparticles arise from the interpretation of negative energy solutions of the Dirac equation. The effect of C is illustrated in Fig. 1.3, through exchange of black and white (which represent opposite
2 charges) on the original painting Day and Night.
Fig. 1.3: An illustration of Charge conjugation, obtained by exchanging black and white on the original painting.
Soon after the discovery of Parity violation, it was shown by T. D. Lee, R. Oehme and C. N. Yang [8] that the breaking of C is implied in weak interactions if P is not conserved. On the other hand, Wu’s experiment did not find any evidence for violation of symmetry under C and P combined. This transformation we have mentioned is what we call CP transformation and its effect is illustrated in Fig. 1.4.
Fig. 1.4: CP transformation.
At this stage, we should emphasize that gravity, electromagnetic and strong interactions are invariant under all the discrete transformations we have mentioned. Until some years after the discovery of Parity violation, there was no evidence for CP breaking on weak interactions. Against the belief of most physicists, this scenario would change in 1964 when J. Christenson, J. Cronin, V. Fitch and R. Turlay [9] discovered CP violation in weak interactions. It is known experimentally that the decays of neutral kaons may occur with two different lifetimes,
0 −9 0 −6 τ(KS → 2π) = 9 × 10 s , τ(KL → 3π) = 5 × 10 s , where the subscripts S and L stand for ’short’ and ’long’, respectively. The final states of those decays transform differently under a CP transformation:
CP|ππi = |ππi , CP|πππi = −|πππi .
0 0 In other words, if there was CP conservation KL could only decay into three pions and KS into two. 0 In spite of occurring with a small branching ratio, it was shown that the process KL → 2π may also occur, confirming that CP is not a symmetry of weak interactions. For this discovery, Croninand Fitch received the Nobel Prize in 1980. Finally, a Time reversal transformation consists on the inversion of the time coordinate, t → −t. Invariance under this operation is observed if, for a given event, it is possible to find a similar one but evolving backwards in time. A very important connection between this discrete symmetry and the
3 remaining ones is established by the CPT Theorem. Demonstrated in the fifties by G. Lüders [10], W. Pauli [11] and J. Bell [12], this theorem states that the combination of C, P and T (CPT) must be conserved for any QFT. This demand of CPT conservation allows one to establish many relations between discrete symmetries. For instance, if one finds CP violation in weak interactions, the T symmetry must also be broken according to this theorem, as it was verified experimentally in 2012 on the B-meson system by the BaBar collaboration [13]. Then, we conclude that of all seven combinations involving the three discrete symmetries, CPT is the only one that might be conserved by all symmetries, because all others may be broken in weak interactions. In fact, no violation of CPT has been found so far [2].
1.2 CP violation in the SM
The discovery of CP violation in 1964 was not expected by theoretical physicists and it would not be an easy task to understand how one could incorporate it in the SM. In 1973, a successful explanation was put forward by M. Kobayashi and T. Maskawa [14] for the occurrence of CP violation in terms of the so-called CKM mechanism, which is closely connected with quark mixing in the SM. After the success of Fermi’s theory on the explanation of β-decay, physicists tried to extend it to other decays. In order to apply this universality to weak interactions, one should be able to describe all decays in terms of the same coupling constant as the one introduced by Fermi (we will call it g). Before proceeding, we must note that we are designating by β-decay the process
n → p + e + νe , (n = udd, p = uud) .
Nowadays, we understand this decay as one where a transition from a down-quark d to an up-quark u (u ↔ d) is involved. This transition forms a charged weak current (mediated by W ±) like all the others we will mention. For leptonic charged currents, it was straightforward to extend Fermi’s theory. Only transitions involving particles of the same generation (e ↔ νe and µ ↔ νµ) were observed, and they could be characterized nearly by the same coupling constant g as the one discovered by Fermi. However, the addition of the second generation of quarks (formed by c and s ones) would not be so straightforward. Since decays like
Λ → p + e + νe , Λ = (uds) , are observed, where a u ↔ s transition occurs, we conclude that the theory of weak interactions must accommodate transitions between quarks of different generations. In this context, we assume that u and c quarks do not couple to d and s respectively, but to superpositions of those two:
0 0 d = d cos θC + s sin θC , s = −d sin θC + s cos θC .
In other words, we say that the weak eigenstates d0 and s0 which participate on weak interactions can be obtained through a rotation by an angle θC (the Cabibbo angle) of the mass eigenstates d and s. This idea of quark mixing was introduced by N. Cabibbo [15] in 1963, and it was extremely important to restore universality of weak interactions. Indeed, decays with strangeness violation (as the one we presented above for Λ) are much more rare than others involving just one generation of particles. Since the probability of some given decay to occur is proportional to the square of its coupling constant, there was some difficulty to reconcile universality with strange particle decays. But
4 according to Cabibbo’s theory, all transitions we mentioned so far are related to g as follows:
e ↔ νe, µ ↔ νµ ⇒ g , u ↔ d ⇒ g cos θC , u ↔ s ⇒ g sin θC .
All the above predictions are supported by experimental results, so we can confirm that weak inter- actions can be characterized by a single constant g. Nonetheless, it is important to emphasize that ◦ the Cabibbo angle is not predicted by the SM, hence, its numerical value (θC ' 13 ) is fitted to be in agreement with experiment. The c quark and its coupling with s0 were postulated some years after Cabibbo’s work in the framework of the Glashow-Iliopoulos-Maiani (GIM) mechanism [16], which explains the suppression of Flavor Changing Neutral Currents (FCNCs) in the SM. But more important, two-quark mixing does not allow for CP violation. As it was found by Kobayashi and Maskawa in 1973, the idea of quark mixing had to be extended to three generations of quarks in order to get CP violation. In this case, instead of just one angle, the unitary transformation of quark fields requires at least three angles and one complex phase. Although we will come back to this in Chapter 2, we can already anticipate that this phase gives rise to CP violation in the SM. The validity of all the above obviously depended on the discovery of a third generation of quarks. The first evidence for its existence came out in 1977, whenthe b quark was discovered by the L. Lederman team at Fermilab [17]. The search for the top quark would take much longer, but it was finally found in 1995 by the CDF [18]O andD / [19] collaborations. This confirmed the prediction made by Kobayashi and Maskawa, who were awarded with the Nobel prize in 2008 for their successful explanation of CP violation.
1.3 From the SM to GUTs
The construction of the SM relies on the principle that its Lagrangian must be invariant under a specific set of local symmetries established by a gauge group. In order to guarantee localgauge invariance, one must introduce gauge fields in the theory, which will act as mediators of interactions. A theory with these properties is designated by gauge theory and they have been playing a fundamental role on the description of interactions between elementary particles. The simplest example of a gauge theory is Quantum Electrodynamics (QED), which describes electromagnetic interactions between photons and electrically charged matter. This theory is based on the Abelian group U(1), meaning that the Lagrangian of QED must be invariant under field transformations of the form ψ → eiqα(x)ψ, where α(x) is the phase in a point x of spacetime and q is the electric charge. The photon field is introduced in order to preserve gauge invariance. This procedure can be generalized for non-Abelian groups, namely SU(n). Gauge theories based on these groups are often designated by Yang-Mills theories [20] and they have applied in the description of weak and strong interactions. The first step for the formulation of the SM was taken in 1961 by S. Glashow [21], whenhepro- posed a unified theory of weak and electromagnetic interactions based on the gauge group SU(2)⊗U(1). When this was done, there was a big issue around it. The gauge bosons introduced in the theory were all massless and one could not add to the Lagrangian mass terms for the mediators of weak inter- actions. The Lagrangian would be no longer gauge invariant and, consequently, one could not make any predictions based on the theory because it would not be renormalizable. This was the motivation to introduce Spontaneous Symmetry Breaking (SSB) in the SM. Discovered by Y. Nambu [22] in the framework of Particle Physics, SSB consists on the breaking of the gauge symmetry by the vacuum state of the theory, generating in this way mass terms for gauge bosons without breaking Lagrangian
5 invariance. For some time, it was thought one could not use this in the SM since SSB gives rise to massless bosons, not observed experimentally. In fact, this is a consequence of the Goldstone theo- rem [23], which establishes the appearance of massless scalar particles (known as Goldstone bosons) in the same number as the number of generators of the broken gauge group. However, we can solve this problem through the Higgs mechanism [24–27]. Proposed in 1964, it consists on the appearance of mass terms for gauge bosons after SSB, thanks to the (would-be) Goldstone bosons which one expects to arise from the breaking. In other words, the latter are not observed because they are ’absorbed’ as longitudinal degrees of freedom by the particles that become massive. The Higgs mechanism was later incorporated in Glashow’s theory by S. Weinberg [28] and A. Salam [29] in order to solve the problem of the massless gauge bosons associated to weak interactions. Finally, G. t’Hooft and M. Veltman [30] proved in 1971 that gauge theories with SSB (as the one built by Glashow, Weinberg and Salam) are renormalizable, establishing this way the SM of Particle Physics. Extending the gauge group of electroweak interactions with an extra SU(3) for Quantum Chro- modynamics (QCD), which describes strong interactions, everything we said about the SM in this section can be summarized by:
SU(3) ⊗ SU(2) ⊗ U(1) −−→SSB SU(3) ⊗ U(1) | {z } | {z } |{z} QCD Electroweak QED i.e., for relatively high energies, one has a gauge theory invariant under the group SU(3)⊗SU(2)⊗U(1), in which electromagnetic and weak forces are unified. After SSB, the mediators of weak interactions become massive and the gauge group becomes SU(3)⊗U(1). The subgroup SU(3) remains unbroken, so the mediators of strong interactions (the gluons) are massless, whereas the electroweak group is broken into U(1), which is the group of QED that we described above. Despite the success of the SM, it was soon realized that perhaps the SM was not the ultimate theory of Nature. One of the reasons for this belief motivated the appearance of GUTs and it is related to the lack of unification between strong and electroweak interactions in the SM. Then, asthe name suggests, GUTs are SM extensions where the gauge group SU(3)⊗SU(2)⊗U(1) is embedded into a larger one, such as SU(5) [31], SO(10) [32], E6 [33], or the one which characterizes trinified models, SU(3)⊗SU(3)⊗SU(3) [34]. Since the GUT gauge group is always larger than the SM group, more gauge bosons have to be introduced to preserve local gauge invariance. Furthermore, the representations of these groups have more components than the SM fields can fill up, so new scalars and fermions mustbe added as well. These postulated particles acquire masses of the order of the unification scale through SSB of the GUT group, leaving SU(3)⊗SU(2)⊗U(1) as a residual symmetry, so that all phenomena observed at low-energy scales is still (successfully) explained by the SM. In general, GUTs are able to solve many problems in the SM, making at the same time predictions that were not observed experimentally up to date. Although we will focus on the issues of CP Violation, we will be able to observe some of these aspects of GUTs by studying Trinification models. But before that, we review the SM in Chapter 2, where all concepts introduced above will be explained more rigorously. After that, Chapter 3 is dedicated to vectorlike quarks, and to show why are these particles so attractive. We will study their properties, their present experimental status, and some examples of SM extensions where these particles are postulated. In Chapter 4 we study Trinification models, where we will see how vectorlike quarks naturally arise. Finally, in Chapter 5, we draw the main conclusions of our work.
6 The Standard Model 2 of Particle Physics
Since its formulation in the sixties by Glashow, Weinberg and Salam, the SM has survived to many experimental tests and it is still the best theory we have to describe strong, weak and electromagnetic interactions. Our goal in this chapter is to review the SM. We will explore in more detail the properties of the CKM matrix, due to its connection with CP violation. In the end, we discuss briefly some of the open questions in Particle Physics, which call for extensions of the SM.
2.1 Field content and Lagrangian
As seen in Section 1.3, the SM is a gauge theory based on the group
GSM = SU(3)c ⊗ SU(2)L ⊗ U(1)Y , (2.1) where the subscripts c, L and Y label each subgroup corresponding to color, left-handedness and hypercharge, respectively. We characterize all SM fields by their quantum numbers (r3, r2, y), so that each field is assigned to the representation r3 of SU(3)c and r2 of SU(2)L. Its hypercharge y is an eigenvalue of the operator
Y = Q − T3 , (2.2) where Q is the electric charge and T3 the third component of weak isospin.
In order to guarantee gauge invariance under GSM, each of its subgroups has associated to it the same number of gauge fields as the number of generators, corresponding to eight, three and onefor SU(3), SU(2) and U(1) respectively. Then, we conclude that there are twelve gauge bosons in the SM, assigned to the adjoint representations, as presented in Table 2.1. Regarding the fundamental constituents of matter, quarks and leptons, their fields are grouped in three generations with identical properties (apart from their masses), described by their left and
Group Gauge fields (r3, r2, y)
a SU(3)c Gµ (a = 1,..., 8) (8,1,0) i SU(2)L Wµ (i = 1, 2, 3) (1,3,0)
U(1)Y Bµ (1,1,0)
Table 2.1: The SM Gauge fields and their representation properties.
7 Generations Notation i = 1 i = 2 i = 3 (r3, r2, y)
α α α α u u c t qα = i (3, 2, 1/6) iL d d s b Quark i L L L L uα uα cα tα (3, 1, 2/3) fields iR R R R dα dα sα bα (3, 1, −1/3) iR R R R νli νe νµ ντ Lep- `i = (1, 2, −1/2) L l e µ τ ton i L L L L
fields liR eR µR τR (1, 1, −1)
Table 2.2: Fermionic fields of the SM. The α index in quark fields accounts for their color.
right-handed components,
1 − γ5 1 + γ5 ψL = ψ , ψR = ψ , γ5 = iγ0γ1γ2γ3 , (2.3) 2 2 where γi are the Dirac matrices. These chiral fields ψL,R transform differently under the subgroup
SU(2)L of GSM. This is illustrated in Table 2.2, where we also establish the notation used from now on to describe the whole fermionic content of the SM. The field hypercharges given in Table 2.2 were computed using Eq. (2.2) considering that:
2 1 Q(ν ) = 0 ,Q(l ) = −1 ,Q(uα) = ,Q(dα) = − , li i i 3 i 3 (2.4) 1 1 T (ν , u ) = ,T (l , d ) = − ,T (l , u , d ) = 0 . 3 liL iL 2 3 iL iL 2 3 iR iR iR
In order to generate mass for gauge bosons and fermions through Electroweak Symmetry Breaking (EWSB), we also need to add a complex scalar doublet to the SM: ! ! φ+ 1 ξ + iξ Φ = = √ 1 2 , Φ ∼ (1, 2, 1/2) , 0 (2.5) φ 2 ξ3 + iξ4 with φ+ and φ0 being, respectively, the charged and neutral components of the field Φ. Along with the addition of gauge fields, interactions between them and the other SM fields arise from demanding local gauge invariance. These interactions are read from the covariant derivatives,
8 3 X a a X i i 0 Dµ = ∂µ − igc GµT − ig WµI − ig BµY, (2.6) a=1 i=1 which must replace the usual partial derivatives in the scalar and fermion kinetic terms. The param- 0 eters gc, g and g are the coupling constants of the SU(3)c, SU(2)L and U(1)Y groups, respectively. Depending on the representation assignments r2,3, one must perform the following replacements:
a i a λ i τ r3 = 3 → T = , r2 = 2 → I = , 2 2 (2.7) a i r3 = 1 → T = 0 , r2 = 1 → I = 0 .
In other words, T a and Ii are the generators of the SU(3) and SU(2) groups on a given representation.
8 If the fields are assigned to the fundamental representation of these groups, λa and τ i are defined by the Gell-Mann and Pauli matrices, respectively. In contrast, T a and Ii are zero if the fields transform as singlets under SU(3) and SU(2). All information about the field dynamics and interactions is encoded in the full SM Lagrangian1:
LSM = LGauge + LΦ + LMatter + LYuk . (2.8)
Here, LGauge comprises the gauge kinetic terms. The pure gauge-field terms which preserve gauge invariance are provided by the field strength tensors,
a a a abc b c Gµν = ∂µGν − ∂ν Gµ + gsf GµGν ,
i i i ijk j k Wµν = ∂µWν − ∂ν Wµ + g WµWν , (2.9)
Bµν = ∂µBν − ∂ν Bµ ,
abc ijk where f are the structure constants of SU(3) and is the rank-3 Levi-Civita tensor. LGauge is then written as: 1 a aµν 1 i iµν 1 µν − LGauge = G G + W W + Bµν B . (2.10) 4 µν 4 µν 4 Further inspection of this Lagrangian reveals that it contains not only the kinetic terms, but also interactions involving exclusively gauge fields.
The most general scalar Lagrangian LΦ invariant under GSM is:
µ † LΦ = (D Φ) (DµΦ) − V (Φ) , (2.11) where V (Φ) is the Higgs potential, which contains all non-kinetic dim-4 combinations of the scalar field. Namely, in the SM, 2 † † 2 V (Φ) = µ Φ Φ + λ Φ Φ . (2.12)
The last two pieces of LSM are concerned with fermions. LMatter contains their kinetic terms, interactions with gauge bosons, and it can be written in the form LMatter = Lq + L` (the subscripts q and ` stand for ’quarks’ and ’leptons’, respectively) with2
3 X µ µ µ Lq = i qjLγ DµqjL + ujRγ DµujR + djRγ DµdjR , (2.13) j=1 3 X µ µ L` = i `jLγ Dµ`jL + ljRγ DµljR . (2.14) j=1
Notice that all fermions of the SM are massless with the Lagrangians we have defined at this point. In fact, mass terms for fermions cannot be added to the Lagrangian directly. Since left and right-handed components of fermions transform differently under SU(2)L, a Dirac mass term,
mDirac = −m(f LfR + f RfL) , (2.15) for a fermion f would break explicitly the gauge symmetry. However, these particles may acquire mass through SSB due to Yukawa interactions, in which fermions couple to the Higgs field. These
1 Gauge fixing and Fadeev-Popov terms are being omitted in LSM. 2Since we are focusing only on electroweak interactions, we omit color indices in quark fields. We also use the definition ψ ≡ ψ†γ0.
9 couplings can be read from the Yukawa Lagrangian,
u d ` − LYuk = Yij qiLΦ ujR + Yij qiLΦ djR + Yij `iLΦ ljR + H.c., (2.16)
∗ u d l where Φ = iτ2Φ and Yij , Yij and Yij are general complex 3 × 3 matrices, which contain most of the free parameters of the SM.
2.2 The Higgs mechanism
The SM vacuum is identified by minimizing the potential V (Φ) of Eq. (2.12). The vacuum energy must be non-zero in order to achieve SSB. To ensure charge neutrality, only neutral scalar fields can acquire a non-vanishing Vacuum Expectation Value (VEV), which is possible for the lower isospin component of the Higgs doublet, ! 0 h0|Φ|0i ≡ hΦi = . (2.17) v
Hence, the minimum of the potential is given by the condition
r ∂V (Φ) µ2 = 0 ⇒ |v| = − . (2.18) ∂Φ hΦi 2λ
In any case, one must consider λ > 0, otherwise the potential would not be bounded from below. Regarding the parameter µ, one must consider µ2 < 0, otherwise the VEV vanishes and SSB does not occur. Of the four gauge fields associated to SU(2)L ⊗ U(1)Y , three become massive and, by the Goldstone theorem, a residual unbroken symmetry must exist. This can be found by analyzing the symmetries of the vacuum. Given a gauge transformation eiαGhΦi ' (1 + iαG)hΦi, the vacuum is invariant under G if GhΦi = 0. This is not verified for any generator of SU(2) ⊗ U(1) as long as the VEV is non-zero. However, considering the action of the electric charge operator defined through Eq. (2.2) on hΦi, we notice that " ! !# ! ! ! ! h τ3 i 1 1 0 1 1 0 0 1 0 0 0 QhΦi = Y (Φ) + hΦi = + = = . (2.19) 2 2 0 1 2 0 −1 v 0 0 v 0
This reveals that the generator of the unbroken symmetry is the electric charge and, since it is diagonal, it must be associated to a U(1) group. This proves what we pointed out in Chapter 1 regarding EWSB, when we said that the electroweak group SU(2) ⊗ U(1) is broken into the U(1) of electromagnetism. Before proceeding, it is important to check that if one applies the obtained electric charge operator on the scalar doublet defined in Eq. (2.5), we get: ! ! ! 1 0 Φ+ Φ+ Q Φ = = . (2.20) 0 0 Φ0 0
This confirms that Φ+ is indeed a charged scalar field whereas Φ0 is neutral. Considering now small oscillations around the vacuum, we reparametrize Φ as: ! ξ(x)a τ a 0 Φ(x) = exp i H(x) , (2.21) v 2 v + √ 2 where H(x) and ξ(x)a are real scalar fields. Before EWSB, one is free to perform gauge transformations as long as they leave the Lagrangian invariant. Then, three of the four degrees of freedom owned by
10 the field Φ can be removed by performing an SU(2) transformation which cancels the exponential of Eq. (2.21): ! ξ(x)a τ a 0 Φ(x) → exp −i Φ(x) = H(x) . (2.22) v 2 v + √ 2 This transformation take us to the unitary gauge, with only one physical field H(x) (the Higgs boson field) remaining after EWSB. Disregarding constant, kinetic and interaction terms, we obtaingauge boson and Higgs mass terms by replacing the transformed Φ of Eq. (2.22) in LΦ. With this, ! ! 1 1 g2 −gg0 W 3 LW,Z,H = g2v2(W 1µW 1 + W 2µW 2) + v2 3µ µ µ − µ2H2 . Mass µ µ W B 0 02 (2.23) 2 4 −gg g Bµ
The mass spectrum of the gauge bosons can be read defining the mass eigenstates
1 2 3 3 ± Wµ ∓ Wµ Aµ = sin θW W + cos θW Bµ ,Zµ = cos θW W − sin θW Bµ ,W = √ , (2.24) µ µ µ 2 where θW is the weak mixing angle, related to the coupling constants by
g0 g sin θW = , cos θW = . (2.25) pg2 + g02 pg2 + g02
± 3 Rewriting Eq. (2.23) in terms of Aµ, Zµ and Wµ one finds :
2 2 2 2 02 W,Z,H g v µ+ − 1 v (g + g ) µ 1 2 2 L = W W + Z Zµ − 2µ H . (2.26) Mass 2 µ 2 2 2
From here, it is clear that the gauge bosons W ±, Z and the Higgs Boson H acquire masses
r 2 02 gv g + g gv p 2 mW ± = √ , mZ = v = √ , mA = 0 , mH = −2µ . (2.27) 2 2 2 cos θW
The beauty and success of the Higgs mechanism is evident from Eq. (2.27). It not only generates ± mass for the three gauge fields (Zµ and Wµ ), but it also keeps the photon field Aµ massless. One also verifies by counting the number of degrees of freedom (twelve before and after SSB) that inlinewith what was discussed in Section 1.3, the three Goldstone bosons ’gauged-away’ in Eq. (2.22) re-emerged as longitudinal polarizations of the massive gauge bosons.
2.3 Charged and Neutral interactions
Expanding the covariant derivatives in Eq. (2.14) in terms of the mass eigenstates defined in Eq. (2.24), and omitting the couplings with gluons, we can express the fermion Lagrangian as
f f f f L = Lkin + LCC + LNC , f = q, ` . (2.28)
f f The first term corresponds to kinetic terms of fermionic fields, whereas LCC and LNC are related to charged currents (CC) and neutral currents (NC), respectively. The designations ’charged’ and ’neutral’ stand in this context for the electric charge of gauge bosons involved in the interactions with
3 1 In contrast with both the Higgs Boson and the Z, we cannot remove a factor of 2 from the mass term of the W because this is a complex field.
11 ± f fermions. Then, interactions involving the Wµ fields stem from the Lagrangian LCC as
` g µ − L = √ ei γ W νiL + H.c., (2.29) CC 2 L µ
q g µ + L = √ ui γ W di + H.c.. (2.30) CC 2 L µ L
On the other hand, the couplings of fermions to the photon field Aµ and the Zµ field are encoded in f LNC. These are
` µ g X µ i i LNC = Q(ei) g sin θW eiγ Aµei + ψiγ (gV − gAγ5)Zµψi , (2.31) cos θW ψi=ei,νi
q X µ g X µ i i LNC = e Q(ψi) ψiγ Aµψi + ψiγ (gV − gAγ5)Zµψi , (2.32) cos θW ψi=ui,di ψi=ui,di where the first term of Eq. (2.31) was written in such a way that one can find the relation between the electric charge e and the weak mixing angle:
0 e = g sin θW = g cos θW . (2.33)
In Eq. (2.32) we have introduced the vector (V ) and axial-vector (A) couplings of the Z boson to fermions, which are respectively defined as
i 2 i gV ≡ T3 − 2 Qi sin θW , gA ≡ T3 , (2.34) being the charges Q(ui) and Q(di) given in Eq. (2.4). Defining the electromagnetic currents for quarks and leptons as q µ X µ ` µ µ Jem = Q(ψi) ψiγ ψi ,Jem = Q(ei) eiγ ei , (2.35)
ψi=ui,di it will be useful for us to write the NC Lagrangians in the following way:
` ` µ g h µ µ 2 ` µi LNC = eJem Aµ + νiLγ νiL − eiLγ eiL − 2 sin θW Jem Zµ , (2.36) 2 cos θW
q q µ g µ µ 2 q µ LNC = eJem Aµ + uiLγ uiL − diLγ diL − 2 sin θW Jem Zµ . (2.37) 2 cos θW
2.4 Fermion masses and mixing
After EWSB, the Yukawa Lagrangian of Eq. (2.16) becomes H u d l −LYuk = v + √ (ui Y uj + di Y dj + li Y lj ) + H.c. 2 L ij R L ij R L ij R (2.38) H ≡ LMass + Lint .
H While the Lagrangian Lint contains the couplings of fermions to the Higgs boson, the mass Lagrangian LMass is given by u d l − LMass = uiLMijujR + diLMijdjR + liLMijljR + H.c., (2.39) where we have introduced the mass matrices
u u d d l l Mij = vYij ,Mij = vYij ,Mij = vYij . (2.40)
12 Notice that there is no mass matrix for neutrinos. As one can see from Table 2.2, right-handed neutrino fields are not included in the SM. Thus, taking Eq. (2.15) into account, one isnotableto construct Dirac neutrino mass terms in this context4. This leads us to conclude that neutrinos are massless in the SM. In general, mass matrices like those defined in Eq. (2.39) are general complex matrices. Asa consequence, we need to transform the fields in order to bring the mass terms to the physical basis, i.e. u 0 u 0 uL → VL uL , uR → UR uR , d 0 d 0 dL → VL dL , dR → UR dR , (2.41) l 0 l 0 lL → VL lL , lR → UR lR , where V u,d,l and U u,d,l are unitary matrices which relate flavor and mass eigenstates (primed fields). These have been chosen such that they diagonalize M u, M d and M l through the following bi-unitary transformations: u† u u u VL M UR = diag(mu, mc, mt) ≡ Mf , d† d d d VL M UR = diag(md, ms, mb) ≡ Mf , (2.42) l † l l l VL M UR = diag(me, mµ, mτ ) ≡ Mf . Once we apply these transformations to Eq. (2.39), we arrive to:
3 X u 0 0 d 0 0 l 0 0 − LMass = Mfii uiLuiR + Mfii diLdiR + Mfii liLliR + H.c.. (2.43) i=1
The NC Lagrangians given in Eqs. (2.31) and (2.32) remain invariant under the same transformations of the fields. Thus, we conclude that there are no flavor changing neutral currents (FCNCs) attree level in the SM. Regarding leptonic CCs, the rotation of the fields in Eq. (2.29) leads to
` g 0 l ∗ µ − L = √ e V γ νiLW + H.c.. (2.44) CC 2 iL L µ
l 0 Since neutrinos are massless, one has freedom to perform the rotation νiL → V νiL, cancelling the ` effect of transforming the charged lepton fields. As aresult, LCC remains invariant and we conclude that there is no lepton mixing in the SM. The quark sector behaves differently, as we saw in Section 1.2. q Bringing the fields to the mass basis on LCC, given in Eq. (2.30), we find g Lq = √ u0 V u∗γµV dd0 W + + H.c. CC 2 iL L L iL µ Vud Vus Vub (2.45) g 0 µ 0 + u† d = √ u γ VCKMd W + H.c.,VCKM ≡ V V = V V V , 2 iL iL µ L L cd cs cb Vtd Vts Vtb where VCKM is the CKM matrix, which describes quark mixing in the SM. The appearance of CP violation in the CKM matrix can be understood by counting the number of independent parameters of a general complex square matrix U with dimension n2. Without any constraint, this matrix has 2n2 independent parameters. In order to make the matrix U unitary (UU † = 1), one has to impose n2 conditions, so the number of independent parameters becomes 2n2 − n2 = n2. Considering now an n × n orthogonal matrix (analog to an unitary matrix but with only real parameters), we can observe that it has n(n − 1)/2 Euler angles. This will be the number of
4For electrically neutral fermions like neutrinos, one could also build Majorana mass terms. However, the existence of these terms in the Lagrangian imply lepton number violation.
13 mixing angles we need to parametrize our unitary matrix, while the remaining n(n+1)/2 independent parameters are complex phases which arise in U. However, some of these phases are unphysical and can be removed. To see this, we must observe that the Lagrangian shown in Eq. (2.45) remains invariant under the quark rephasing
iθu iθu 0 † 0 Θu = diag e 1 ,..., e n , uL → uL Θu , uR → Θu uR , (2.46) 0 † 0 d d d → Θ d , iθ1 iθn dL → dL Θd , R d R Θd = diag e ,..., e , as long as we change U simultaneously, according to
0 † U = Θu U Θd . (2.47)
At this point, one might expect to identify 2n phases as unphysical. However, a global rephasing redefining all quarks by the same phase leaves U unchanged. As a result, we cannot use this transfor- mation to remove a phase from U, so our matrix has to be parametrized by n(n + 1)/2 − (2n − 1) = (n − 1)(n − 2)/2 complex phases. Our discussion regarding flavor mixing and CP violation in Sec- tion 1.2 becomes now clear. Indeed, a quark mixing matrix with only two generations (n = 2) needs to be parametrized by one angle only and no complex phases. In contrast, the existence of three generations of quarks (n = 3) forces us to parametrize the CKM matrix with three angles and one complex phase, which is responsible for all CP-violating phenomena in the SM. From all possible parametrizations for VCKM, the one proposed by Chau and Keung [35] became standard. Namely,
−iδ 1 0 0 c13 0 s13e c12 s12 0 VCKM = 0 c23 s23 0 1 0 −s12 c12 0 iδ 0 −s23 c23 −s13e 0 c13 0 0 1 (2.48) −iδ c12c13 s12c13 s13e −iδ −iδ = −s12c23 − c12s23s13e c12c23 − s12s23s13e s23c13 , −iδ −iδ s12s23 − c12c23s13e −c12s23 − s12c23s13e c23c13 where sij ≡ sin θij and cij ≡ cos θij control the mixing between generations i and j, and δ is the
CP-violating phase. Considering the parameter λ ≡ sin θC ' 0.22 (θC is the Cabibbo angle), and taking experimental data into account, the CKM matrix can be approximately written as
1 λ λ3 2 VCKM ' λ 1 λ . (2.49) λ3 λ2 1
Following the observed hierarchy among the elements of VCKM, Wolfenstein [36] developed a parametriza- tion (perhaps the most useful) defined in terms of powers of λ. Introducing the real parameters A, ρ and η, the mixing angles θij and the phase δ, these are expressed by [37, 38]:
|Vus| 2 Vcb iδ ∗ 3 s12 = λ = , s23 = Aλ = λ , s13e = V = Aλ (ρ + iη) , (2.50) p 2 2 ub |Vud| + |Vus| Vus to all orders in λ. The Taylor expansion of all the CKM matrix elements up to order O(λ4), obtained
14 after replacing Eqs. (2.50) into Eq. (2.48), gives the Wolfenstein parametrization of the CKM matrix:
1 − λ2/2 λ Aλ3(ρ − iη) 2 2 4 VCKM = −λ 1 − λ /2 Aλ + O λ . (2.51) Aλ3(1 − ρ − iη) −Aλ2 1
Observing now carefully the constraints we need to impose to guarantee the unitarity of a matrix,
† † X ∗ X ∗ VV = V V = 1 ⇒ VikVjk = VkiVkj = δij , (2.52) k k one finds six vanishing combinations (i 6= j) that express the orthogonality between two different columns or rows. These may be represented as unitarity triangles in the complex plane and, in spite of their different shapes, they have all the same area, equal to half of the Jarlskog invariant J [39]. This phase-convention independent measurement of CP violation is defined through
∗ ∗ X Im VijVklVil Vkj = J ikmjln , (2.53) m,n where ikm is the rank-3 Levi-Civita tensor. The Jarlskog invariant is written in terms of the standard and the Wolfenstein parametrizations as
2 2 6 J = c12c23c13s12s23s13 sin δ ' A λ , (2.54) thus we conclude that a non-vanishing CKM matrix necessarily requires J 6= 0. The most commonly used unitarity triangle is shown in Figure 2.1 and it arises from the condition,
∗ ∗ ∗ VudVub + VcdVcb + VtdVtb = 0 , (2.55)
∗ where we normalize each term to VcdVcb, so that one side is aligned with the real axis and its apex is given to all orders in λ by √ ∗ 2 VudV 1 − λ (ρ + iη) ρ + iη = − ub = √ √ . (2.56) ∗ 2 4 2 4 2 VcdVcb 1 − A λ + A λ 1 − λ (ρ + iη)
One can also notice that the series expansions of the parameters ρ and η we have introduced can be used to improve the unitarity of the Wolfenstein parametrization shown in Eq. (2.51), simply by doing the replacements ρ → ρ and η → η. Finally, we present some of the most important experimental data concerned with the CKM matrix. The fit results for the Wolfenstein parameters defined in Eq. (2.50) are shown in Table 2.3.Forthe magnitude of all nine CKM elements, the fit results are
+0.00011 0.97434−0.00012 0.22506 ± 0.00050 0.00357 ± 0.00015 |VCKM| = 0.22492 ± 0.00050 0.97351 ± 0.00013 0.0411 ± 0.0013 , (2.57) +0.00032 0.00875−0.00033 0.0403 ± 0.0013 0.99915 ± 0.00005