Renormalization Group Invariants in the Minimal Supersymmetric Standard Model
Master’s Thesis September 2014
Tom van Daal [email protected]
Supervisor: Prof. dr. W.J.P. Beenakker
Department of Theoretical High Energy Physics Institute for Mathematics, Astrophysics and Particle Physics Radboud University Nijmegen
Contents
1 Introduction 5 1.1 Conventions...... 5
2 The Standard Model 7 2.1 IntroductiontotheStandardModel ...... 7 2.2 TheStandardModelfieldcontent...... 7 2.3 The Standard Model Lagrangian ...... 8 2.4 Flavormixing...... 10
3 Symmetries of the Standard Model 15 3.1 Noether’stheorem ...... 15 3.2 Poincar´esymmetries ...... 16 3.3 Gaugesymmetries ...... 16 3.4 Flavorsymmetries ...... 17 3.4.1 Flavor group without flavor mixing ...... 17 3.4.2 Flavor group with flavor mixing ...... 21
4 Supersymmetry 25 4.1 Introductiontosupersymmetry ...... 25 4.2 Motivation for supersymmetry ...... 26 4.3 Supersymmetric Lagrangians ...... 29 4.3.1 Chiralsupermultiplets ...... 29 4.3.2 Gaugesupermultiplets...... 30 4.3.3 Chiral and gauge supermultiplets combined ...... 31 4.4 Supersymmetrybreaking...... 32
5 The Minimal Supersymmetric Standard Model 33 5.1 TheMSSMfieldcontent...... 33 5.2 R-parity...... 34 5.3 The MSSM Lagrangian ...... 34 5.4 Softsupersymmetrybreaking ...... 37 5.5 Countingindependentparameters ...... 38 5.6 SymmetriesoftheMSSM ...... 42
6 Renormalization group techniques 45 6.1 What is renormalization? ...... 45 6.2 Renormalization group equations ...... 46 6.3 E↵ectivefieldtheories ...... 46 6.4 Renormalization group invariants ...... 47
7 One-loop -functions for MSSM scalar masses 51 7.1 The propagator for scalar fields ...... 51 7.1.1 Higgs doublets ...... 52 7.1.2 Scalarfields...... 52 7.2 One-loop corrections to scalar propagators ...... 53 7.2.1 Type 1 diagrams ...... 54 7.2.2 Type 2 diagrams ...... 55 7.2.3 Type 3 diagrams ...... 56 7.2.4 Type 4 diagrams ...... 58 7.3 Renormalization of scalar fields ...... 59 7.3.1 Field renormalization ...... 59 7.3.2 Mass renormalization ...... 60 7.4 Results...... 61
3 8 RG invariants in the MSSM 62 8.1 Constructing RG invariants in the canonical way ...... 62 8.2 RG invariants from a di↵erentperspective ...... 65
9 Conclusion 68
AIndices 69
B Unitary matrices 70
C Spinors 72 C.1 Diracspinors ...... 72 C.2 Weylspinors ...... 72
D Superpotential contributions to the MSSM Lagrangian 74
E Integrals 78
F One-loop mass counterterms for MSSM scalar fields 83
List of abbreviations 87
Acknowledgements 88
Bibliography 90
4 1 Introduction
The experiments at the Large Hadron Collider (LHC) at CERN, among other experiments, are searching for physics beyond the Standard Model of particle physics, since we know that the Standard Model cannot be the final story. Experimentalists are looking for signals that do not fit well in the Standard Model framework. One job for theorists, on the other hand, is to come up with new models that describe physics at even smaller length scales (or, equivalently, higher energy scales) that solve the current problems with the Standard Model. Over the last decades, many new theories have been described in literature, all coming with both pros and cons. In this thesis we will focus on one particular beyond the Standard Model theory: the so-called Minimal Supersymmetric Standard Model (MSSM). This is the most minimal extension to the Standard Model that respects an additional spacetime symmetry, called supersymmetry. In short, the MSSM predicts that each Standard Model particle has a supersymmetric counterpart with exactly the same quantum numbers and mass (except for the spin that di↵ers by half a unit). We already know that even the MSSM could not be the “theory of everything” though. Namely, since supersymmetric particles have not been found yet, this means that either supersymmety is simply not implemented in nature, or that supersymmetric particles are much heavier (at least beyond the reach of the LHC) than their Standard Model counterparts, i.e. supersymmetry is a broken symmetry of nature. In this thesis, of course, we will assume the latter. At energy scales many orders of magnitude higher than the typical LHC collider scale of (10 TeV), even beyond the range of validity of the MSSM, new physics would provide boundary conditionsO to the MSSM. Say that supersymmetric particles are discovered at the LHC and that some relevant new parameters are measured, would there then be a way to check which of those boundary conditions could be realized in nature? The answer is yes: quite recently, a new method has been devised that could probe these boundary conditions by making clever use of so-called renormalization group (RG) invariants using low energy data only (i.e. the new data coming from particle accelerators). Constructing these RG invariants as probes for high scale physics models can be a very tedious and complicated job though, and this has not been done yet for (the scalar sector of) the MSSM, but only for heavily constrained supersymmetric models (in these models typically about 80% of the MSSM parameters are “thrown away”). In this thesis we will attempt to construct RG invariant quantities for the full MSSM, i.e. without any simplifications. In this thesis a thorough analysis of symmetries of the MSSM will be performed. Most of this analysis will be carried out for the Standard Model though (as this theory is a bit simpler) and the results will then be generalized to the MSSM. Finally, an attempt will be made to unify symmetries of the MSSM with the concept of RG invariant quantities. If such a relation would turn out to exist, then it would be much easier to tell for a given (supersymmetric) theory if RG invariants could exist at all, and, if the answer is positive, how to construct them. After all, simplifying a certain theory (i.e. setting parameters equal to each other or to zero) generally means that symmetry is added to its Lagrangian. Could this in turn mean that also the number of RG invariants increases? In the first section we will discuss the Standard Model, mostly focussing on its Lagrangian and mixing of fermionic flavors. In section 2 we will present a thorough analysis of symmetries of the Standard Model. The obtained results for family space will be most important for the remainder of this thesis. Subsequently, the basic concepts of supersymmetry will be discussed and a general supersymmetric Lagrangian will be constructed. In the next section we will consider the MSSM and perform a careful count of its independent parameters. Also the symmetry analysis for the Standard Model will then be generalized to the MSSM. In the subsequent section, the concept of renormalization and the utility of RG invariants will be discussed. Section 7 will be the most technical part of this thesis. Here the one-loop RG equations for the MSSM scalar masses will be derived. Finally, an attempt will be made to construct RG invariants in the MSSM from these obtained RG equations and we will consider the role of symmetries in this context.
1.1 Conventions Throughout this thesis we will make use of the following conventions:
We adopt natural units where ~ = c = 1. With this choice, mass is the only dimension that •
5 is left, i.e.: 1 1 [mass] = [energy] = [length] =[time] , (1) where the square brackets denote the mass dimension. We make use of the Einstein summation convention, which means that all repeated indices • are summed over (unless explicitly stated otherwise). The various types of indices that are used in this thesis are defined in appendix A. For the Minkowski metric ⌘ we choose the convention ⌘ = diag(1, 1, 1, 1). • µ⌫ µ⌫ We define the contravariant spacetime and energy-momentum four-vectors to be • xµ =(t, ~x) ,pµ =(E,p~ ) , (2)
and the (ordinary) derivative is defined as
@ @ . (3) µ ⌘ @xµ
The n n identity matrix will be denoted by In, and the most important matrices in family • space will⇥ be indicated in boldface (to make the text more readable).
It is important to be aware of the fact that the sections on the Standard Model make use • of Dirac spinors to describe the fermionic fields, while the sections on supersymmetry and renormalization group techniques use Weyl spinors instead. The di↵erences between Dirac and Weyl spinors are explained in appendix C. The abbreviation “h.c.” is short for “Hermitian conjugate”. •
6 2 The Standard Model
The Standard Model of particle physics incorporates all known fundamental particles and their (non-gravitational) interactions and forms the heart of particle physics. In this section we briefly review the main aspects of the Standard Model. We will discuss its particle content and provide the most important mathematical structures. Much more on the Standard Model can be found in any textbook on quantum field theory or particle physics, e.g. in [1].
2.1 Introduction to the Standard Model The Standard Model is a quantum field theory that describes the dynamics of all known funda- mental particles up to an incredible accuracy. It provides the framework for three out of the four (known) fundamental forces of nature: the electromagnetic, weak and strong interactions. The Standard Model was completed in the early 1970s, but not fully experimentally verified until the discovery of the Higgs boson at the Large Hadron Collider (LHC) in 2012 [2, 3]. Despite the great success of the Standard Model, there are also a couple of (significant) short- comings. One of them being the complete absence of gravity. It is expected that at the Planck scale gravity becomes comparable in strength to the other forces, which implies that quantum gravity e↵ects can no longer be ignored1. In other words, the Standard Model breaks down and a reconciliation of general relativity and quantum mechanics is needed. This is why the Standard Model cannot be the theory of everything; it provides an excellent e↵ective framework for “low” energy physics, but for higher energies (i.e. shorter distance scales) we simply need a new theory. An important beyond the Standard Model candidate theory is supersymmetry. In section 4 the ideas behind supersymmetry are elaborately discussed, as well as the shortcomings of the Standard Model and how supersymmetry claims to solve these.
2.2 The Standard Model field content In this subsection we outline the Standard Model field content. The Standard Model contains both fermions (the matter particles) and bosons. The fermionic sector consists of three families of quarks and leptons. The quarks come in two di↵erent types: charge2 +2/3, the up-type quarks (up, charm and top); and charge 1/3, the down-type quarks (down, strange and bottom). Likewise, also the leptons come in two di↵erent types: charge 1, the charged leptons (electron, muon and tau); and the neutral neutrinos (electon neutrino, muon neutrino and tau neutrino). The three fundamental interactions described by the Standard Model are mediated by spin- 1 gauge bosons. Gauge bosons arise from the theory by requiring local invariance under gauge transformations corresponding to a certain gauge group. The full gauge group of the Standard Model is SU(3)C SU(2)L U(1)Y. The Standard Model gauge groups are Lie groups and the number of gauge bosons⇥ equals⇥ the number of generators. The gauge bosons “live” in the adjoint representation of the gauge group, while the fermions that couple to the gauge bosons “live” in the fundamental representation. The various representations are referred to with their dimensions indicated in boldface. The particles that are in the fundamental representations 2 and 3 are called doublets and triplets respectively. A fermion that is uncharged under a certain gauge group is in 1 (the trivial representation) and is called a singlet. The first gauge group, SU(3)C, forms the underlying symmetry group of the strong interac- tion. Its corresponding eight gauge bosons are the massless gluons. The “C” stands for color, the quantum number that indicates sensitivity to the strong interaction. The colored particles in the Standard Model, i.e. the particles that participate in the strong interaction, include quarks (color triplets) and gluons (color octets). The second group, SU(2)L U(1)Y, describes the electroweak sector of the Standard Model. The electroweak interaction is⇥ mediated by the W1,2,3 and B bosons. The W1,2,3 bosons, that come from the SU(2)L part, only couple to particles with a non-zero weak isospin (the weak isospin doublets), or di↵erently stated, only to particles that have a left-handed chirality (hence
1One of these e↵ects is that around the Planck scale the wavelength of a particle typically becomes shorter than its Schwarzschild radius due to Heisenberg’s uncertainty principle. This implies that studying these short distance scales by performing high energy particle collisions, inevitably results in black hole production. 2When we write just “charge”, we mean electrical charge.
7 3 the “L” in the name of the group) . The B boson corresponds to the U(1)Y part and couples to all particles with a non-vanishing weak hypercharge (Y ). From experiment we know that certain particles are massive while others are massless. This means that we need to add mass terms to the Standard Model Lagrangian. If we do this in the simplest possible way we run into trouble. Namely, these mass terms spoil the gauge invariance of the theory. As the principle of local gauge invariance provides the basis of the theory, we better respect that. The easiest way to add mass terms to the Lagrangian in a gauge invariant way, is by introducing a complex SU(2) doublet scalar field. If this spin-0 field has a non-zero vacuum (i.e. ground state) expectation value, then after nature has chosen a ground state, the electroweak symmetry is spontaneously broken and the Standard Model particles have acquired a mass. This is called the Higgs mechanism, first described by Higgs, Brout and Englert (cf. [4,5]). In terms of groups, electroweak symmetry breaking is described as follows:
SU(3) SU(2) U(1) SU(3) U(1) (4) C ⇥ L ⇥ Y ! C ⇥ Q After electroweak symmetry breaking, the complex scalar doublet gives rise to one (massive) physical state called the Higgs field. This breaking also mixes the electroweak gauge boson eigen- states to yield the familiar mass eigenstates that we observe in experiments: the massless photon, the mediator of the electromagnetic interaction; and the massive W± and Z bosons, the media- tors of the weak interaction. The electromagnetic interaction is characterized by the group U(1)Q, where “Q” stands for the electrical charge. Its mediator, the photon, couples to all particles that carry an electrical charge. This is the Standard Model in a nutshell. All Standard Model fields (before electroweak symmetry breaking) and their gauge group representations are given in table 1. Note that we have included the right-handed neutrino. In this thesis we assume that the right- handed neutrino completely decouples from the other particles.
Name Field SU(3) SU(2) U(1) C ⇥ L ⇥ Y L =(⌫ e )(1, 2, 1 ) L L L 2 leptons ⌫R (1, 1, 0) e (1, 1, 1) R 1 QL =(uL dL)(3, 2, 6 ) quarks 2 uR (3, 1, 3 ) d (3, 1, 1 ) R 3 + 0 1 Higgs boson =( )(1, 2, 2 ) Ga (8, 1, 0) gauge bosons W a0 (1, 3, 0) B (1, 1, 0)
Table 1: The fields of the Standard Model and their gauge group representations. The first two numbers in the last column (those indicated in boldface) refer to the dimension of the gauge group representations, while the last number denotes the eigenvalue of the weak hypercharge generator Y .
2.3 The Standard Model Lagrangian Any quantum field theory is usually described by a Lagrangian. A Lagrangian generally involves several fields and contains kinetic and interaction terms. The Euler-Lagrange equations – origi- nating from the principle of stationary action – give the equations of motion for the fields. Now let
3In appendix C we define the nomenclature for the left- and right-handed fermionic fields.
8 us consider the Lagrangian that describes the Standard Model. We will do this for the unbroken case and piece by piece. The names of the fields are given in table 1.
Kinetic terms for the gauge fields and interactions among themselves: • 1 1 1 L = Ga Gaµ⌫ W a0 W a0µ⌫ B Bµ⌫ , (5) gauge 4 µ⌫ 4 µ⌫ 4 µ⌫ where the field strength tensors are defined as follows:
Ga = @ Ga @ Ga g f abcGb Gc , (6a) µ⌫ µ ⌫ ⌫ µ s µ ⌫ W a0 = @ W a0 @ W a0 g✏a0b0c0 W b0 W c0 , (6b) µ⌫ µ ⌫ ⌫ µ µ ⌫ B = @ B @ B . (6c) µ⌫ µ ⌫ ⌫ µ The coupling constants pertaining to the strong, weak and electromagnetic interactions are denoted a0b0c0 abc respectively by gs,g,g0. The symbols ✏ and f are the completely antisymmetric structure constants of SU(2) and SU(3) respectively.
Kinetic terms for the fermions and gauge interactions: •
Lfermions = iLLDL/ L + i⌫R@⌫/ R + ieRDe/ R + iQLDQ/ L + iuRDu/ R + idRDd/ R , (7) where all fermionic fields are multiplets in family space (more on this in the next subsection). The gauge covariant derivatives are given by:
1 a0 a0 1 D L = @ + igW ig0B L , (8a) µ L µ 2 µ 2 µ L ✓ ◆
D e = @ ig0B e , (8b) µ R µ µ R ✓ ◆
1 a a 1 a0 a0 1 D Q = @ + ig G + igW + ig0B Q , (8c) µ L µ 2 s µ 2 µ 6 µ L ✓ ◆ 1 a a 2 D u = @ + ig G + ig0B u , (8d) µ R µ 2 s µ 3 µ R ✓ ◆ 1 a a 1 D d = @ + ig G ig0B d . (8e) µ R µ 2 s µ 3 µ R ✓ ◆ Here a0 (the Pauli matrices) and a (the Gell-Mann matrices) are twice the generators of the SU(2) and SU(3) Lie algebras respectively. The conventions for the left- and right-handed projec- tions of the Dirac spinors are found in appendix C.
Yukawa couplings: • 2 2 L = iL y ⇤⌫ L y e iQ y ⇤u Q y d +h.c., (9) Yukawa L ⌫ R L e R L u R L d R where y⌫ , ye, yu, yd are general complex 3 3 matrices in family space, called the Yukawa coupling matrices. These couplings are discussed in⇥ the next subsection in much more detail. Note that we are taking the neutrinos massive here by including a right-handed neutrino multiplet ⌫R.This means that we simply treat the neutrinos the same as the other fermions4. In the sections on supersymmetry, however, the neutrinos are taken massless.
Kinetic term for the Higgs field and Higgs self-interactions: • µ L =(D )† D V, (10) Higgs µ where the gauge covariant derivative for the Higgs doublet is given by:
1 a0 a0 1 D = @ + igW + ig0B . (11) µ µ 2 µ 2 µ ✓ ◆ 4Within the Standard Model the right-handed neutrino would be a Dirac particle, while in some beyond the Standard Model theories it also carries a Majorana component.
9 The Higgs potential V = V ( ) is a function of the Higgs doublet and given by: | | V ( )= µ2 2 + 4 , (12) | | h| | | | 2 where , µh > 0. The minimum of this potential is located at µ v = h = , (13) | | p2 p2 where v µh/p represents the vacuum expectation value of the Higgs doublet (i.e. 0 0 = v/p2). The⌘ groundstate of the Higgs doublet is actually degenerate and each state leads toh || the|| samei physics. Once nature has chosen one of these vacua, the electroweak symmetry is spontaneously broken. The Higgs doublet then takes the following form in the unitarity gauge:
1 0 = , (14) p2 v + h ✓ ◆ with h denoting the real physical Higgs scalar field for which 0 h 0 = 0. The sum of (5), (7), (9) and (10) gives the full Lagrangianh | describing| i the Standard Model, except for one part: the so-called gauge fixing terms. The gauge fields cannot be identified with physical objects as they contain redundant degrees of freedom; that is to say, applying gauge transformations to these fields does not change the physics. Hence we need to choose a certain gauge5 to be able to describe physics in a unique way, and this is called gauge fixing. This means that additional terms, that are arbitrary to some extent, must be added to the Standard Model Lagrangian. These gauge fixing terms, however, do not change the symmetry analyses in this thesis and can therefore be safely ignored. So far we have mainly considered the Standard Model before breaking of the electroweak symmetry and the parts of the Standard Model Lagrangian have been written in terms of gauge eigenstates only. After electroweak symmetry breaking, however, it is convenient to rewrite the Lagrangian in terms of mass eigenstates. The U(1)Y and SU(2)L gauge fields mix and yield the following electroweak vector boson mass eigenstates:
+ 1 1 2 1 1 2 W = W iW ,W = W + iW , µ p2 µ µ µ p2 µ µ A =sin✓ W 3 + cos ✓ B ,Z= cos ✓ W 3 sin ✓ B , (15) µ w µ w µ µ w µ w µ where the weak mixing angle ✓w is defined as g tan ✓ 0 . (16) w ⌘ g Besides the electroweak vector bosons, it also turns out to be convenient to redefine the fermionic fields. This is the topic of the next subsection. Table 2 provides an overview of the Standard Model parameters before electroweak symmetry breaking. In the next subsection we will see that not all of these parameters are physical; some degrees of freedom can be gauged away by redefining the fields.
2.4 Flavor mixing As has been discussed briefly in de previous section, the Standard Model fermions come in three di↵erent families (or generations). The first family of quarks and leptons is constituted by the up and down quarks and, respectively, by the electron and its corresponding neutrino. The next family consists of fermions with exactly the same quantum numbers, only their masses are larger. The third family is formed by the heaviest fermions6. Some Standard Model interactions involve fermions of di↵erent families, this is called flavor mixing. In this subsection we will see where flavor mixing comes from and how it can be parametrized.
5For certain applications, of course, some gauges prove handier than others. 6Although the mass hierarchy of the neutrinos is as yet unknown.
10 Name Physicaldescription Mathematicaldescription #Parameters
gs,g,g0 gauge couplings real numbers 3 y , y , y , y Yukawacouplingmatrices complex3 3matricesinf.s. 72 ⌫ e u d ⇥ µh Higgs symmetry breaking parameter real number 1 Higgs self-coupling strength real number 1 Total: 77
Table 2: Naive counting of parameters in the unbroken Standard Model. The abbreviation “f.s.” stands for “family space”.
The so-called Yukawa terms in the Standard Model Lagrangian couple the Higgs doublet to the fermionic fields. In the unbroken Standard Model, the part of the Lagrangian that describes these couplings reads (cf. eq. (9)):
2 2 L = iL y ⇤⌫00 L y e00 iQ y ⇤u00 Q y d00 +h.c., (17) Yukawa L00 ⌫ R L00 e R L00 u R L00 d R where the primes refer to particular choices of basis (we will come back to this in a bit). Note that also in this subsection the fermionic fields (LL00 ,⌫R00 ,eR00 etc.) actually denote family multiplets. In fact, all fermionic fields in this section are family multiplets. So with eR00 , for example, we denote (with a slight abuse of notation) the family multiplet constituted by the right-handed electron, muon and tau: eR00 e00 µ00 . (18) R ⌘ 0 R 1 ⌧R00 The electroweak symmetry of the Standard@ ModelA is broken once we expand the Higgs doublet around a chosen vacuum state (i.e. ground state). Eq. (14) gives the Higgs doublet after electroweak symmetry breaking in the unitarity gauge. Plugging this particular form of the Higgs doublet in eq. (17) gives us the Yukawa terms in the broken Standard Model: 1 LYukawa = (v + h) ⌫00y⌫ ⌫00 + e00 yee00 + u00 yuu00 + d00 ydd00 +h.c. . (19) p2 L R L R L R L R ⇣ ⌘ A priori, there is no reason to assume that no mixing takes place between the di↵erent families and thus the Yukawa couplings y (with = ⌫, e, u, d) are general complex 3 3 matrices. The components of the diagonalized Yukawa coupling matrices are proportional to⇥ the masses of the fermions. To diagonalize these matrices, the fermionic fields undergo unitary basis transformations, i.e. we switch from a description in terms of fermionic gauge (“double primed”) eigenstates to mass (“single primed”) eigenstates. The other terms in the Lagrangian do not change their form under these changes of basis, except for the terms that couple the fermions to the W boson. Here the unitary transformations on the fermionic fields result in the unitary quark and lepton mixing matrices. Before we derive the form of these mixing matrices, let us first find the basis transformations that diagonalize the Yukawa coupling matrices. To diagonalize the Yukawa coupling matrices y (we will suppress the label in this derivation), we start by considering the Hermitian matrices yy† and y†y. Any Hermitian matrix can be diagonalized by a unitary similarity transformation and the resulting diagonal matrix has real entries only. Let U, V U(3), then yy† and y†y are diagonalized as follows: 2
yy† U †yy†U D, y†y V †y†yV D, (20) ! ⌘ ! ⌘ where D is a diagonal matrix with non-negative real entries. The latter can be seen as follows:
D = U †yy†U
= U †y†U † U †y†U. (21)
11 For convenience we define A U †y†U,then ⌘
[D]jj = [A]ij⇤ [A]ij i X = [A] 2 0 . (22) | ij| i X From the transformations given by (20), it follows that
yy† = UDU† , y†y = VDV† . (23)
From the first equation of (23) we infer that
y = UfW † , (24) where f is a diagonal matrix with [f] [D] and W U(3). Using eq. (24) yields jj ⌘ jj 2 p y†y = WDW† . (25)
From the second equation of (23), it follows that we can pick W = V . Now we arrive at the conclusion: U †yV = f . (26)
This means that the Yukawa coupling matrices y are diagonalized by a so-called bi-unitary transformation and the diagonalized matrices f have non-negative real entries only. The diag- onalization of the Yukawa coupling matrices amounts to applying basis transformations to the fermionic fields. From eqs. (19) and (26) it follows that it is useful to transform a fermionic field in the following way:
00 U † 00 0 , 00 V † 00 0 . (27) L ! L ⌘ L R ! R ⌘ R Writing eq. (19) in terms of the new fermionic bases (indicated by a single prime), yields: 1 LYukawa = (v + h) ⌫0f⌫ ⌫0 + e0fee0 + u0fuu0 + d0fdd0 . (28) p2 The way that we have combined here the left- and right-handed fields will be illustrated later (cf. eq. (59)). The Yukawa terms are now written in terms of the fermionic mass (“single primed”) eigenstates rather than the gauge (“double primed”) eigenstates. We can write the rest of the Standard Model Lagrangian in terms of these new bases too. For most terms this simply amounts to “removing one prime”. Let us for example consider the part of the Lagrangian that describes the interaction between right-handed up-type quarks and the Z boson (ignoring constant prefactors):
µ µ ZµuR00 uR00 = ZµuR0 Vu† VuuR0 µ = ZµuR0 uR0 . (29) So this term in the Lagrangian retains its form when we write it in terms of the new fermionic bases. The only part of the Lagrangian (besides the Yukawa terms) that does change form, is the coupling of fermions to the W boson. Here the unitary transformations that act on the fermionic fields result in the unitary flavor mixing matrices:
g + µ + µ LW = W ⌫00 e00 + W u00 d00 +h.c. 0 p2 µ L L µ L L ⇣ ⌘ g + µ + µ = W ⌫0 U † Uee0 + W u0 U † Udd0 +h.c. p2 µ L ⌫ L µ L u L ⇣ ⌘ g + µ + µ = W ⌫0 V 0† e0 + W u0 V 0d0 +h.c. , (30) p2 µ L ` L µ L q L ⇣ ⌘ where we have defined the matrices V`0 and Vq0 as
V 0 U †U , V 0 U †U . (31) ` ⌘ e ⌫ q ⌘ u d
12 Now let us try to find explicit parametrizations for these matrices.
How many parameters are needed to parametrize V`0 and Vq0? As both matrices are unitary, we first consider a general unitary matrix (cf. appendix B for more information about unitary matrices). The dimension of the unitary group of degree n is n2, so an arbitrary U U(3) is 2 parametrized by nine real parameters, including three rotation angles: ✓12, ✓23, and ✓13; and six phases: ↵i (i =1, 2), j (j =1, 2, 3), and . Without loss of generality, an arbitrary U U(3) can be written as [6]: 2 U = AV B , (32) where 10 0 ei 1 00 A = 0 ei↵1 0 ,B= 0 ei 2 0 (33) 0 1 0 1 00ei↵2 00ei 3 and @ A @ A i c12c13 s12c13 s13e V = s c c s s ei c c s s s ei s c , (34) 0 12 23 12 23 13 12 23 12 23 13 23 13 1 s s c c s ei c s s c s ei c c 12 23 12 23 13 12 23 12 23 13 23 13 with c cos ✓ and@ s sin ✓ . A ij ⌘ ij ij ⌘ ij Any 3 3 unitary matrix can be written in the form given by eq. (32), thus also V 0 and V 0: ⇥ ` q
V 0 A V B , V 0 A V B . (35) ` ⌘ ` ` ` q ⌘ q q q
However, some degrees of freedom in V`0 and Vq0 are not physical and can be removed. The phase factors that appear in the matrices A`,B`,Aq,Bq can be completely “absorbed” by the quark and lepton fields without changing the physics. The absorption of phases by the lepton multiplets goes as follows: ⌫0 B ⌫0 ⌫ ,e0 A†e0 e , (36) L ! ` L ⌘ L L ! ` L ⌘ L and for the quark fields we have:
u0 A†u0 u ,d0 B d0 d . (37) L ! q L ⌘ L L ! q L ⌘ L
As a consequence, the matrices A`,B`,Aq,Bq that are a priori contained in V`0 and Vq0 are com- pletely absorbed by the fermionic fields. Instead of nine real parameters, the resulting flavor mixing matrices are parametrized by only four parameters each. We define the lepton mixing matrix, also called the Pontecorvo–Maki–Nakagawa–Sakata (PMNS) matrix, as7:
i ` c12c13 s12c13 s13e V = s c c s s ei ` c c s s s ei ` s c , (38) ` 0 12 23 12 23 13 12 23 12 23 13 23 13 1 s s c c s ei ` c s s c s ei ` c c 12 23 12 23 13 12 23 12 23 13 23 13 @ A ` ` with cij cos ✓ij and sij sin ✓ij. The standard parametrization of the quark mixing matrix, also called the⌘Cabibbo–Kobayashi–Maskawa⌘ (CKM) matrix, reads:
i q c12c13 s12c13 s13e V = s c c s s ei q c c s s s ei q s c , (39) q 0 12 23 12 23 13 12 23 12 23 13 23 13 1 s s c c s ei q c s s c s ei q c c 12 23 12 23 13 12 23 12 23 13 23 13 q @ q A with cij cos ✓ij and sij sin ✓ij. The CKM⌘ and PMNS⌘ matrices are unitary 3 3 matrices that describe the mixing between the di↵erent quark and lepton families respectively.⇥ The coupling of leptons and quarks to the W boson is the only part of the Standard Model that exhibits mixing of fermionic families. In terms of these new fermionic bases (without primes), eq. (30) becomes:
g + µ + µ LW = W ⌫LV † eL + W uL VqdL +h.c. (40) 0 p2 µ ` µ ⇣ ⌘ Naturally, the terms in the Lagrangian that do not involve mixing of flavors should be invariant under the transformations given by (36) and (37), as otherwise new parameters would enter the
7Assuming that neutrinos are Dirac particles rather than Majorana particles.
13 Standard Model. To this end, the right-handed projections of the fermionic fields have to transform the same as their left-handed counterparts, as can be inferred from the Yukawa terms given by eq. (28)8. In terms of the “unprimed” bases, the Yukawa terms take the form: 1 LYukawa = (v + h) ⌫f⌫ ⌫ + efee + ufuu + dfdd . (41) p2 A summary of the unitary transformations that act on the fermionic fields is given by table 3.
Fermionic fields
⌫00 B U †⌫00 ⌫ L ! ` ⌫ L ⌘ L ⌫00 B V †⌫00 ⌫ R ! ` ⌫ R ⌘ R e00 A†U †e00 e L ! ` e L ⌘ L e00 A†V †e00 e R ! ` e R ⌘ R u00 A†U †u00 u L ! q u L ⌘ L u00 A†V †u00 u R ! q u R ⌘ R d00 B U †d00 d L ! q d L ⌘ L d00 B V †d00 d R ! q d R ⌘ R Table 3: Transformations of the fermionic fields.
Let us briefly summarize what has happened to the degrees of freedom in the Yukawa sector of the Standard Model. A priori, the Yukawa matrices y (with = ⌫, e, u, d) are general complex 3 3 matrices, thus giving rise to 4 18 = 72 independent parameters. However, many parameters can⇥ be absorbed by redefinitions of⇥ the fermionic fields. After applying appropriate basis transfor- mations to these fields (cf. table 3), we end up with diagonalized (real) Yukawa matrices f (with = ⌫, e, u, d) and two flavor mixing matrices V`, Vq that are both described by four parameters. This means that the number of independent physical parameters in the Yukawa sector of the Standard Model is 4 3+2 4 = 20. Table 4 summarizes the parameter counting in the Standard Model after having carried⇥ out⇥ all the redefinitions of the fields. From this table we infer that the Standard Model has 25 actual free parameters9. Electroweak symmetry breaking, of course, does not change the number of independent parameters.
Name Physicaldescription Mathematicaldescription #Parameters
gs,g,g0 gauge couplings real numbers 3 f , f , f , f diag. Yukawa coupling matrices real diag. 3 3matricesinf.s. 12 ⌫ e u d ⇥ V , V flavor mixing matrices unitary 3 3matrixinf.s. 8 ` q ⇥ µh Higgs symmetry breaking parameter real number 1 Higgs self-coupling strength real number 1 Total: 25
Table 4: All parameters of the unbroken Standard Model. The abbreviations “diag.” and “f.s.” stand for “diagonal” and “family space” respectively.
8 The (diagonal) matrices Ai,Bi (with i = `, q)commutewiththediagonalizedYukawacouplingmatricesf (with = ⌫, e, u, d). 9Note that we are excluding the strong CP violating angle. For simplicity and good phenomenological reasons, we have set this parameter to zero.
14 3 Symmetries of the Standard Model
Symmetries of nature are transformations on a physical system that leave the physics unchanged. Symmetry principles play a very important role with respect to the laws of physics. More precisely, all continuous symmetries of a physical system give rise to conserved quantities. In this section we study the continuous symmetries of the Standard Model gauge theory. These continuous symme- tries can be divided into three types: Poincar´e, gauge and flavor symmetries. The first two types form the cornerstones of the Standard Model, while the last one constitues a class of accidental symmetries. After a short discussion on the importance of symmetries in field theory, we will list the symmetries of the Standard Model and their corresponding conservation laws.
3.1 Noether’s theorem The relation between symmetries and conserved quantities in field theory is given by Noether’s theorem. This very important theorem was proved by Noether in 1915 and today this theorem and its derivation are found in any standard textbook on field theory, e.g. in [1]. What does this theorem say? Let us consider a set of fields i that satisfy the Euler–Lagrange equations for the Lagrangian L = L ( i , @µ i ): { } { } @L @L = @ , (42) @ µ @(@ ) i ✓ µ i ◆ and let these fields undergo the following continuous transformations: + ↵ , (43) i ! i i where ↵ is infinitesimal. The equations of motion of the fields i do not change as long as the action corresponding to the Lagrangian is left invariant under these transformations. In that case, the field transformations form a symmetry of the system. For the Lagrangian this means that it can change at most by a four-divergence: L L + ↵@ Gµ , (44) ! µ which means that µ ↵@µG = ↵ L @L @L = ↵ + (@ ) @ i @(@ ) µ i ✓ i µ i ◆ @L @L @L = ↵@ + ↵ @ . (45) µ @(@ ) i @ µ @(@ ) i ✓ µ i ◆ i ✓ µ i ◆ The second term vanishes due to the Euler-Lagrange equations for the fields i (cf. eq. (42)), so we are left with µ @µj =0, (46) where the conserved current jµ is given by
µ @L µ j = i G . (47) @(@µ i) This is the mathematical form of Noether’s theorem and the conserved currents are called Noether currents. Moreover, a conserved current implies a locally conserved charge defined by
C (t) d~x j 0(x) , (48) V ⌘ ZV as d d C (t)= d~x j 0(x) dt V dt ZV = d~x ~ ~j(x) r· ZV = d~s ~j(x) , (49) · Z@V
15 for any volume V with boundary @V . In the following subsections we list all continuous symmetries of the Standard Model and relate these to their conserved charges that follow from Noether’s theorem.
3.2 Poincar´esymmetries Minkowski spacetime is invariant under certain transformations. These symmetries come in three di↵erent types: Translations in time and space. • Rotations in space. • Boosts connecting two uniformly moving bodies. • These symmetries, called Poincar´esymmetries, are global and continuous. What are, according to Noether’s theorem, the corresponding conservation laws? Invariance under translations in time and space give conservation of energy and (linear) momentum respectively, and rotational invariance of space implies conservation of angular momentum. Invariance under boosts gives that the field’s center of energy travels at constant velocity (which is not very interesting from a physics’ point of view). The full symmetry group of Minkowski spacetime is called the Poincar´egroup P and it contains the Lorentz group SO(1, 3) (rotations and boosts) and the group of translations R1,3. The Lorentz group is a subgroup of dimension six and the group of translations is a normal subgroup of dimension four and every element of the Poincar´egroup can be written uniquely as a product of a Lorentz transformation and a translation. This means that P is the so-called semidirect product of R1,3 and SO(1, 3), or in mathematical notation:
1,3 P = R o SO(1, 3) . (50) The dimension of the Poincar´egroup is 4 + 6 = 10. An equation that is invariant under Poincar´etransformations is called relativistically invariant. All sensible descriptions of nature should be relativistically invariant, which is why this principle has been one of the cornerstones of the Standard Model.
3.3 Gauge symmetries Another type of symmetries is given by the gauge symmetries, which provide the basis for the Standard Model. As explained in the previous section, gauge symmetries are internal symmetries that, once we make them local and require invariance, give rise to gauge bosons and interactions mediated by these gauge bosons in a natural way. The full gauge group of the Standard Model is the twelve-dimensional Lie group SU(3)C SU(2)L U(1)Y. According to Noether’s theorem, all twelve generators must give rise to a conserved⇥ quantity.⇥ Each fundamental interaction of nature (that is described by the Standard Model) corresponds to a particular gauge group. The strong interaction is connected to the Lie group SU(3). Invariance under local SU(3) transformations give – through Noether’s theorem – conserved quantities called color (C1,...,C8). The electroweak interaction, on the other hand, is related to the Lie group SU(2) U(1). The corresponding conserved quantities again follow from Noether’s theorem and ⇥ are called weak isospin (T1,T2,T3) and weak hypercharge (Y ). They correspond to the groups SU(2) and U(1) respectively. After the electroweak symmetry has been broken (cf. eq. (4)), weak isospin and weak hypercharge have combined into a new quantum number called electrical charge (Q), given by Q = T3 + Y. (51) Electrical charge corresponds to a U(1) Lie group and is a conserved quantity of the broken Standard Model. Both Poincar´eand gauge symmetries have served as guiding principles for the construction of the Standard Model. Some other symmetries of the Standard Model are rather accidental; they just happen “to be there”. These symmetries are found in the family space of the Standard Model and are elaborately discussed in the next subsection.
16 3.4 Flavor symmetries In this subsection we consider the purely accidental (global) continuous symmetries of family space. Our aim is to find the symmetry group of family space, called the flavor group.Wewill do this in the context of the Standard Model, and, for simplicity, consider the broken rather than the unbroken version. At the end of this subsection we will comment on the implications of our results for the unbroken Standard Model and notice that nothing really changes. To determine the flavor group, we will distinguish between various cases: with and without flavor mixing, as well as di↵erent fermionic mass spectra10.
3.4.1 Flavor group without flavor mixing We will first consider the Standard Model without flavor mixing. This simply means that we omit the CKM and PMNS matrices from our description. Another way of looking at this is saying that we aprioriassume the Yukawa coupling matrices to be diagonal. To find the flavor group, we first have a look at the kinetic terms of the fermions. If denotes a fermionic field11,thenitskinetic term in the Lagrangian reads: free L = i @/ . (52)
Let M GL(3, C) be a (general) linear transformation in family space that acts on the fermionic fields as2 follows: M . (53) ! This means that eq. (52) transforms as
i @/ i M† @/M = i M † M @/ . (54) ! ⇣ ⌘ Requiring invariance of (54) under the transformation M , leads to the restriction:
M † M = I3 , (55) which means that M U(3). However, this is not the final story. Let us have a closer look at the field . The Standard Model2 is a chiral theory, which means that left- and right-handed fermions (transforming independently under Lorentz transformations) live in di↵erent representations of the Standard Model gauge group (cf. appendix C for more information on this and conventions for notation). In eq. (52) the field can be split into its left- and right-handed projections, as
µ µ = (PL + PR) µ 2 2 = PL + PR µ µ = PR PL + PL PR µ µ = L L + R R , (56) so that eq. (52) can be written as
free L = i L@/ L + i R@/ R . (57)
As a consequence, the left- and right-handed projections of the fermionic fields in eq. (57) can transform independently: let M L,MR U(3), then 2 M L , M R . (58) L ! L R ! R These transformations leave the kinetic term for invariant. This is not only true for the kinetic terms though, also the interactions between fermions and vector bosons are of this form. However, as can be inferred from eq. (40), the neutrinos, charged leptons, up- and down-type quarks cannot transform independently: the left-handed charged leptons and neutrinos form doublets, as well as the left-handed up- and down-type quarks. These doublets are independent of each other though.
10The masses of the Standard Model fermions are varied over this analysis rather than taken fixed. 11Note that as before these fields denote family multiplets.
17 Not all interactions decouple left- and right-handedness completely. The (only) terms in the Standard Model Lagrangian that mix left- and right-handed fields are the Yukawa interactions given by eq. (41). As
= (PL + PR) 2 2 = PL + PR
= R L + L R , (59) the Yukawa interaction for a field can be written as
Yukawa 1 L = (v + h) f R + f L . (60) p2 L R This means that the flavor group must be such that L and R do not transform independently. Let L R M ,M U(3) be the transformation matrices that act on the left- and right-handed projections 2 L R 12 of , like in (58). Eq. (60) then transforms under M and M as :
f M L†f M R . (61) L R ! L R L R If we require invariance of this term under M and M , it follows that
M L†f M R = f M Lf f M R =0. (62) ()
Before we study the solutions of this equation, let us first have a closer look at f . The diagonalized Yukawa coupling matrix f is of the form