This action might not be possible to undo. Are you sure you want to continue?

By Spyros Konitopoulos

SUBMITTED FOR THE DEGREE OF DOCTOR OF PHILOSOPHY AT NATIONAL TECHNICAL UNIVERSITY OF ATHENS ATHENS, GREECE MAY 2009

NATIONAL TECHNICAL UNIVERSITY OF ATHENS DEPARTMENT OF SCHOOL OF APPLIED MATHEMATICAL AND PHYSICAL SCIENCES The undersigned hereby certify that they have read and recommend to the Faculty of Graduate Studies for acceptance a thesis entitled “ Non-Abelian Tensor Gauge Theory on lower rank ﬁelds” by Spyros Konitopoulos for the degree of Doctor of Philosophy.

Dated: May 2009

Research Supervisor:

Georgios Savvidis

Examing Committee:

George Zoupanos

George Koutsoumbas

ii

**NATIONAL TECHNICAL UNIVERSITY OF ATHENS IN COLLABORATION WITH THE NATIONAL CENTER OF SCIENTIFIC RESEARCH ”DEMOKRITOS”
**

Date: May 2009 Author: Title: Spyros Konitopoulos Non-Abelian Tensor Gauge Theory on lower rank ﬁelds Department: School of Applied Mathematical and Physical Sciences Degree: Ph.D. Convocation: May Year: 2009

Signature of Author

iii

I.) iv .To Giannis Vasilogiorgakis (R.P.

. . . 1. . . . . . . .2. . . . . . . . . . . . . .2 Vector Gauge Field . .3 Symmetric Tensor Gauge Field . . . . . . . . .2 ﬀ → TT . . . . . . . . .4 rank-2 tensor (part II) . . . . . . . . . . . . . . . . .4 Antisymmetric Tensor Gauge Field . .2. . . . . . . . . 2. . . . . . . . .Table of Contents Table of Contents Acknowledgements Introduction 1 Basic formulation 1. . . . . . . . . . . . . . . . . . . .1. . . . . . . . . .1. . . . . . . . . . . . . . . . . . . . .Gauge bosons . . . . . . . . . . . . . . . . . .4 Geometrical representation . . . . . . . . . . . 2 Unitarity of the theory at the non-interactive level 2. . . 2. . . . . . . . 1. . . . . . . . . . . . . . . . . . . . . .Fermions . . . . . . . . . . . . . . 2. . theory . . . .1. . . . . . v vii 1 10 10 19 22 28 31 31 32 33 34 36 37 42 44 48 52 53 58 58 62 63 . 3. . . . .1 Feynman rules . . .1 A method for counting the propagating modes of a gauge 2. .1. . . . .2 Propagator . . . . . . . . . . . . . . . .1 The method . . . . . 2. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2 Gauge Invariance . . . . . . .3 Geometrical Interpretation 1. 3 Unitarity of the theory at the interactive level-Processes 3. . . . . . . . . . . . . . . . . . . . . 2. . 3. . . 3. . . . . . . .2. . 2. . . . . . . .5 Rank-2 tensor gauge theory .2 rank-2 tensor (part I) . . . . . v . . . . . . . . . . .3 Squared Matrix Element . . . . . . . . . . . . . . . . . . . . . . .1 Yang-Mills . .1.1 The Feynman diagrams . . . . . . . . . . . . . . . . . . . . . .3 The Energy-Momentum Tensor . . . . . . . . . . 2. . . . . . 3. . .

. . . . . . . . . . . . . . . . . . . . . . . . . .2 Feynman diagrams . . . . . . . . . . .3. . . . . . . . . . . . . . . . . . . . . 3. . . . . vi . . . . . . . . . . . 3. . . .1. . . . . . . . . . . . . . . . . . . . . . . . . . . . .3. . . . . . .3. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3 e+ e− → W + W − . . . . . . . . . . . . . . . . . .5 Cross Section . . . . . .6 Helicity Structure . . . . . . . . . . . . . . . . . . . . . . . . .2. . . .2 Gauge Invariance . . . .4 eR eL → W W . . . . . . . . . . . . . . . . .3 e− . . . . . . . . . D. . . . . . .1. . . . . . . . . . . . . . . . . . . . . . .2. . . D. . . . . . . . . . . . . . 3. . . . . . . . . . . . .2 V V → V V . . . . . . . . . . . . . .1. . . . Bibliography . . . . . . . .3. . . . . . . . . . . . . . . . . . . .4 Evaluation of Traces 3. .1 Lagrangian . . . D. . .2. . . . . . . . . . . . . . . D. . . . . . . .2 Gauge invariance . . . . . . .3 Cross section . . . . . .2 Gauge invariance . . . . . . . . .2. . .3. . . . . D. . . .1 Feynman diagrams .3. . . . .3. . . . . . . . . . R eL → W W + − + − D. . . . . . . . . . . . . . .2. . . . . . . . . . . . . . . . . . . . 3. . . . . . . . . . . . . . D. . . . . . . . . . . . . + + − D. . . . . . . . . .3 Scattering amplitudes . 67 72 72 77 77 81 84 89 91 92 95 97 101 101 101 102 103 107 107 110 112 114 114 115 115 117 121 124 Conclusions Appendices A Free rank-2 Tensor Gauge Boson Equation Matrix B Elements of Group Theory C Evaluation of Traces D Processes D.1 f f → V V .1 Feynman diagrams . . . . . . . . . . . . .3 Helicity Amplitudes . . .2. . . .2.3 3. D. D. . . . . .4 Cross Section . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1 Feynman diagrams-the scattering amplitude D. . . . . . . . . .3. . . D. . . . . . VV-TT . . D. . . . . .

Judas Priest..Papagergiou for uncovering and illuminating my inner contradictions.W. I cannot resist the temptation to attribute my respect to some writers and artists whose inﬂuence was enormous to my cultural and mental evolution..Anderson. Demokritos.Dostoyevsky. S. S. I.Beethoven. J. I thank them all. T. This was the place where the seeds of doubt and method were sown upon me. His program was an essential prerequisite for my further programming. F.Freud. Psychotic Waltz. G.Kazantzakis. C.Gavra (opened my eyes to self-realization).S.Acknowledgements First of all I would like to thank my supervisor Prof. M. D.Marx.Beckett.Bonanos for letting me use his Mathematica programm for doing tensor algebra calculations.Nietzche. R.Bakunin. Ditsas the one who trusted and encouraged me to carry on. Greece May 24. this man realizes a rebirth of the notion of the teacher.Karidi (pushed me oﬀ the beaten track ). F. K.Savvidy for the pedagogical guidance and support. S.Zaglis. Athens. F. Th.Fripp.Castoriades. 2009 vii Spyros Konitopoulos . The least I can do is to express him my deepest appreciation. I would like to thank E. D.Dilari. Without his patience and tolerance I would not be able to get in touch with the Russian-school perspective.Tomaras he who initiated me in its secrets and P. L.Papaevagelou. The University of Crete has ”poisoned” my brain irreparably. N. S. P. Prof.Trahanas played an important role on my decision to focus on theoretical physics.Bach. In days of cultural crisis. E. Many thanks to Prof.Gabriel. Finally. where all aspects of human relations are overwhelmed by a commodity exchange value fetishism. People outside the physics’ community have their share on my present development.

Leptons and quarks are classiﬁed in 3 generations in each of which they are eigenstates of 5 irreducible representations of the symmetry group SU (3) × SU (2)L × U (1)Y . These ﬁelds mediate the interactions between the initial fermionic ﬁelds in accordance with the normative principles of locality. gives rise to 12 massless. a self consistent theory of the elementary particles and their interactions was constructed. The basic elements of the theory are the principle of local gauge invariance and the Higgs mechanism [8. In addition. 4. continuous symmetries of a theory one is straightforwardly led to the introduction of some new vector gauge ﬁelds with uniquely determined dynamics. 9.Introduction For more than 45 years the research in high energy physics has been conducted under the guidance of the gauge principle. 6. 11. bosonic ﬁelds and uniquely determine their dynamics. It can be seen. In the early ’70s. free. 2]. that the initial symmetry is spontaneously broken to SU (3) × U (1)em and that the 3 degrees 1 . massless fermions. 10]. Demanding the localization of the global. 12]. giving the vacuum a non-vanishing expectation value. They are the agents of the strong. weak and electromagnetic forces. causality and Lorentz invariance [1. 7. a complex doublet of scalar ﬁelds should accompany the initially. named the standard model (SM) [3. Demanding the theory be locally invariant under the symmetry group SU (3) × SU (2)L × U (1)Y . 5.

For perturbation theory to work. Up to energy scales of 100 GeV . The theory breaks down at the singularities at energies of the order of Mp ∼ 1019 GeV [15]. Z 0 parameters. and the fact that there is no experimental evidence against it. A renormalizable quantum theory of gravity has become a torturing accomplishment due to the divergences the theory exhibits. gravity escapes the uniﬁcation scheme. top quark etc. The Higgs mechanism completes its mission when the initial Lagrangian is supplemented by Yukawa interaction terms which will provide masses to the fermions. one will confront the so called naturalness problem [16]. There is however one piece that escapes the experimental conﬁrmation: The Higgs boson. Most of the predictions of the model have already been tested (neutral weak interactions. Hence. But even if one stays in lower energy scales. it is expected that the mass of the Higgs boson should be less than 1 T eV [16. there are theoretical arguments that the SM does not provide us with a complete theory of particle physics and their interactions. W ± . One ﬁnally ends up with a theory which needs 25 external input parameters to work. First of all.)[53].2 of freedom which would correspond to the massless Goldstone bosons are absorbed by linear combinations of the SU (2)L × U (1)Y gauge ﬁelds casting them massive. Fortunately this lies within the potentials of the LHC and hopefully it will be detected. This problem rises from the very existence of the fundamental . 17]. Despite the experimental conﬁrmations that the model has succeeded. the spontaneous symmetry breakdown is expected to occur at the regime of 250 GeV . the theory is more than successful.

the electric charge quantization etc. Something that would give an adequate explanation why there is such a big gap between the energy scale at which the spontaneously symmetry breakdown occurs and the scale at which new physics dominate. above which yet undiscovered physics prevail. The theory proposes that above the energy scale of 1015 GeV the symmetry group of the SM is uniﬁed to a semi simple group (usually SU (5)). This means that the Higgs’ mass is expected to be of the order of the cut oﬀ. One might legitimately guess that that there is a hidden symmetry which. But. small parameters. remain unanswered. Many questions concerning the choice of the speciﬁc pattern. in order for the perturbation theory to work -small coupling constants. a theory with such a huge parameter space (25 input parameters) cannot fulﬁll ones aesthetic demands. An interesting suggestion beyond the SM is provided by the grand uniﬁcation theory (GUT)[20]. broken in our energy region. gives rise to non-zero. It can be seen that the the Higgs mass receives quadratic divergent contributions from higher order diagrams. within the framework on the SM. is to make some high accuracy assignments (ﬁne tuning) that will drop the value of the Higgs’s mass down to the preferred energy scale. 19]. It seems that something is missing. It is proposed that the initial gauge symmetry breaks spontaneously at 1015 GeV to the standard model group SU (3) × SU (2)L × U (1)Y . Despite its magniﬁcent tested prediction of .3 scalars of the theory. The three interactions with their three coupling constants are combined in one. as it was mentioned above. while the particles are now classiﬁed in 2 representations of the symmetry group. The only way to overcome this problem. Finally. Along with the naturalness problem comes the hierarchy problem [18.the Higgs’ mass should be less than 1 GeV .

Self consistence of the and the chiral fermions(S= 1 2 theory demands the introduction of an additional complex scalar doublet. Also. a Supersymmetric partner should be introduced and that above an energy scale. rendering the model natural[23]. the world is symmetric under the interchange of ) bosons with fermions. which contradicts with experiment. Now. Not only does Supersymmetry provide stability and naturalness to the Standard Model. 22]. the value of the Higgs mass drops down to the scale where the spontaneous symmetry breakdown SU (3) × SU (2)L × U (1)Y → SU (2)L × U (1)Y occurs. Another problematic issue of the model is that the running coupling constants do not meet at an exact point. together with their fermionic superpartners (S= 1 ).4 the weak mixing angle. The gauge bosons (S=1) are accompanied with gauginos (S= 1 2 ) with complex scalars (S=0). though they come very close at 1014 GeV . thus overcoming the conﬂict with experiment. in such a way that they cancel the quadratic divergences. the greater of . the prediction for the life time of the proton extends to more that 1032 years. The theory is motivated by the naturalness and hierarchy problems of the SM. The role played by the partners is amazing! 2 They contribute in the higher order corrections of the Higgs mass. The contributions of the Supersymmetric partners to vacuum polarization lead to an accurate uniﬁcation of the running coupling constants at 1016 GeV . Many ambitions have been invested to Supersymmetic theories. The next revolutionary step comes with the advent of Supersymmetry[21. the theory predicts a proton lifetime of the order of 1031 years. It is suggested that for every SM particle. but also corrects the problems encountered in the framework of GUT.

Interpreting geometrically the tensor gauge ﬁelds as coeﬃcients of the expansion of an extended gauge ﬁeld over its tangent space. a short distance cut-oﬀ is provided in harmony with Lorentz invariance and causality [15]. Following the Yang-Mills paradigm. in the low energy limit of the open string theory with Chan-Paton charges [27]. the massless states can be identiﬁed with Yang-Mills quanta. Especially. making possible an extension of the symmetry breaking mechanism and thus of the electroweak theory to include particles of arbitrary high . A similar procedure has been followed in the description of fermionic and bosonic matter tensor ﬁelds. It is believed that this is the direction towards the uniﬁcation of gravity with all other forces. the extended gauge transformations for higher rank ﬁelds are suitably chosen. To solve the divergence problem of quantum gravity. It would be interesting if these states would appear in the spectrum of a Quantum Field Theory Lagrangian.5 which lies on its local version: Supergravity. 15]. it has been seen that there are two series of gauge invariant Lagrangians. in order to form a closed algebraic structure [31]. Under such a generalization. Assuming that the structural elements of nature are 1 dimensional objects (strings). This motivates a generalization of Yang-Mills symmetry to include non-Abelian tensor gauge ﬁelds[31]. each of which sums up all corresponding invariant Lagrangians of every rank [31]. it is suggested that the interactions are spread out by means of increasing the dimensions of the elementary particles. Quantum Field Theory is extended to a theory which includes in its description ﬁelds of arbitrary high spins. Many models in string theory predict an inﬁnite tower of particles of arbitrary high spin in their spectrum [14. A diﬀerent proposal for a theory of everything comes from string theory.

It was further seen that the linear combination of the two Lagrangians that describe the free rank-2 tensor gauge bosons. It remains an open subject if the theory so far described is self consistent.[35]. In the ﬁrst section. Gauge principle is generalized to tensor ﬁelds of rank-2 and 3 and two series of gauge invariant Lagrangians describing the gauge ﬁelds of rank-2 are constructed. Perhaps in a complete theory. It has been seen that these two Lagrangians are mapped to the original ones by a dual transformation. exhibits an enhanced gauge symmetry. two dual rank-2 Lagrangians have been constructed which are invariant under the full complementary gauge transformations [37]. Inspired from the complementary gauge symmetry the theory of free gauge ﬁelds exhibits. It is because of the enhancement of the gauge symmetry that the cancellation of the pure gauge modes is possible [31]. one has to include both the original series of gauge invariant Lagrangians and their dual counterparts.6 spins [31. . The same program has been followed in the series of Lagrangians that describe ﬁelds up to the fourth rank. In addition to the original gauge transformations of tensor ﬁelds. we present the fundamentals of Tensor Gauge Theory. we will focus our attention on tensor gauge ﬁelds of rank-2. On this thesis. the theory is shown to be symmetric under a complementary transformation for a special choice of the ratio of the weights with which the two gauge invariant Lagrangians participate in the linear combination. a condition which may possibly be proven enough to guarantee the unitarity of the theory [34]. The theory is extensively described and its self-consistence is examined. 33]. the total Lagrangian exhibits an enhancement of the gauge symmetry. It has been shown that for a suitable choice of the coeﬃcients.

As a next step. while the third describes an antisymmetric gauge ﬁeld of rank-2. Finally. we ﬁnally arrive at the same. conﬁrming from another point of view the fact that there are three physical propagating modes contributing to the energy of the free ﬁeld. Implementing this method in the case of rank-2 tensor gauge theory. previously mentioned result [36]. It is because of this symmetry enhancement that the cancellation of longitudinal modes occurs. is directed by the Bianchi identities of the free theory [36]. transversal degrees of freedom. After having speciﬁed the value of this ratio. we considered instructive to develop a general method for counting the propagating modes of a gauge theory which. one can determine the number of the physical propagating modes. the energy momentum tensor for the free 2nd rank tensor gauge ﬁeld is constructed. In the next section we pass from the free theory to the interactive. we check the unitarity of the theory on the free level. which behaves like a particle of spin 2. far from being arbitrary and imposed externally. leaves the gauge freedom of the theory unﬁxed. the generalization of the gauge principle to fermions of spin 3/2 is exposed and the full Lagrangian is constructed. the free theory (g = 0) exhibits an enhanced gauge symmetry [31]. Nevertheless. my the standard methods of gauge ﬁxing [31]. allowing the free propagation of three physical. since the number of the physical modes of a gauge theory is an experimentally tested result.7 Further. was the fact that the speciﬁc choice of the ratio g2 g2 = 1. Our aim is to . What was not known however. behaving like a scalar. and thus independent of the speciﬁc choice of the gauge. In previous articles it has been shown that for the special choice of the ratio g2 g2 = 1. based on the rank of the free equation matrix. Two of them are the polarizations of a symmetric tensor gauge ﬁeld of rank-2.

The brand new result for the process with 2 tensor gauge particles outgoing is: . The cross section for the annihilation process with two gauge bosons incoming. symmetric. On the other hand. tensor gauge bosons. 4 However. the cross section for the production of physical. ﬁrst through the annihilation of two fermions [44. the scattering angle dependence of the cross section is just sin2 θ. where the outgoing particles are longitudinal Ws. s 64d(r) Its sin2 θ behavior is dramatically diﬀerent from from the QED case. things bare much resemblance with the case of the standard electroweak theory process. In both the processes. where as outgoing particles one has two photons. one of the outgoing tensor particles is put in longitudinal polarization leaving all other particles transversal.8 check the unitarity on the interacting level. to check unitarity. 45] and next through the annihilation of two vector gauge bosons [46]. . In that case the cross section maximizes at small angles and has a minimum in perpendicular directions. should be compared to the annihilation process with the same incoming particles. This indicates the fact that contributions of higher rank particles should be added in the rank-2 propagator. we focus on the processes of pair production of rank-2. in the second process the scattering amplitude does not vanish. As it is known. For this. The cross section for the annihilation process where the colliding particles are SM fermions. exhibits a very simple scattering angle dependence: dσ = α2 C2 (r)C2 (G) sin2 θ dΩ. In both cases. In the ﬁrst case the scattering amplitude indeed vanishes for the choice of the L3/2 coeﬃcient f = g2 . but with two gauge bosons as ﬁnal products. transversal outgoing particles is calculated and the symmetry properties of every contributing helicity combination term are studied.

. preference is shown to perpendicular directions. where in the case of rank-2 tensor bosons. These results in the context of Tensor Gauge Theory are the main results to be defended on this thesis.9 2 dσ = g2 2 (G) 419 − 76 cos 2θ + 9 cos 4θ α 2 C2 dΩ s d(G) 512 As se know. in the Yang-Mills case the scattering favors forward and backward directions.

a geometrical picture of the theory is needed to be presented.1 Yang-Mills Yang and Mills proposed a generalization of the principle of local gauge invariance.2 the free gauge theory for tensors bosons of the second rank is constructed and a ﬁrst guess for the free Lagrangian is made. from the Abelian group U (1) to the group SU (2). with the aid of the Bianchi identities of the theory. The idea is to represent the fermionic wave function as an N-dimensional multiplet each element of which is a usual 4-component Dirac ﬁeld.4. 10 . In section 1. the coeﬃcient of the linear combination is evaluated. This is done in section 1. This extension can be made more general to include special unitary groups of arbitrary dimension (SU (N ))[2]. 1.3 where we ﬁnd that the most general Lagrangian which describes free rank-2 tensor gauge bosons is linear combination of two terms. To determine uniquely the free rank-2 tensor Lagrangian.Chapter 1 Basic formulation On this chapter we present the fundamentals of tensor gauge theory. In section 1. in an attempt to describe strong interactions [1].1 we give a brief review of Yang-Mills theory. In section 1.

ψN (1. .1) ¯ LF 1/2 = iΨ ∂ Ψ remains invariant under the global gauge transformation Ψ → Ψ = U (ξ )Ψ where U (ξ ) = eigξ is an element of the group SU (N ) . ta ] + [tc . This can be achieved via the introduction of a number of d(G) 1 gauge ﬁelds Aa . internal space. tc ]] + [tb .11 Ψ= The free fermionic Lagrangian ψ1 ψ2 . N-dimensional.1. [tb . [tc . . The transformation properties of the gauge ﬁelds are so chosen that the covariant derivative of Ψ Dµ Ψ = (∂µ − igAµ )Ψ. For SU (N ). tb ] = if abc tc and the Jacobi identity: [ta . tb ] = 0 (1.3) (1.2) Locality demands the constant parameters to be space-time dependent and the symmetry still to hold.1. 1 (1.4) The dimension of the adjoint representation of the group.1. [ta . d(G) = N 2 − 1 .1. These ﬁelds A = Aa ta can be used as a connection to parallel transport vectors inside this isospin.ξ a are arbitrary constant parameters and ta are the group Lie generators which satisfy the algebra: [ta .

a third term must be added to the Lagrangian which carries derivatives of the ﬁelds A.1. (Dµ Ψ) → (Dµ Ψ) = U (ξ )(Dµ Ψ) It can be easily seen that Aµ should transform as: i Aµ = U Aµ U −1 − (∂µ U )U −1 g Then the Lagrangian FB ¯ ¯ ¯ L1/2 = LF 1/2 + L1/2 = iΨ D Ψ = iΨ ∂ Ψ + g Ψ AΨ (1.5) (1. Until now the gauge ﬁelds appear non-dynamical. Dν ]Ψ) = U (ξ )[Dµ . Dν ]Ψ = −igGµν Ψ.1. It can be seen that [Dµ . where Gµν = ∂µ Aν − ∂ν Aµ − ig [Aµ . Dν ]Ψ. it is straightforward to show that the ﬁeld strength tensor transforms as the connection does but homogenously. Since ([Dµ .12 transforms under a gauge transformation the same way that Ψ does.7) is locally gauge invariant.1.8) . In order to become such. Gµν = U (ξ )Gµν U (ξ )−1 (1.9) (1.1. is the ﬁeld strength tensor. Aν ].1. The standard way to ﬁnd this is to compute the commutator of the covariant derivative on a vector. This must give the analogous Riemann tensor.6) (1.

11.11) Let us now be more speciﬁc in order to reveal some basic properties of the YangMills Lagrangian 1. 4 µν µν (1. Aµ = ta Aa µ .10) The full Yang-Mills Lagrangian is then given as follows: ¯ ∂ Ψ + gΨ ¯ AΨ − 1 Ga LY M = L1/2 + L1 = iΨ Ga . Having normalized the basis elements of the Lie space to unity: tr(ta tb ) = δ ab . the trace of the product of two strength tensors is gauge invariant.12) or in terms of the Lie generators ta . µν = Fµν + gf (1.13 Nevertheless.1.1.1. Gµν = ta Ga µν Gµν = Fµν − ig [Aµ . one gets a tr(Gµν Gµν ) = Ga µν Gµν (1.1. Aν ] . Indeed tr(Gµν Gµν ) = tr(Gµν Gµν ) = tr(Gµν Gµν ). First. we focus our attention on the free bosonic part: 1 Ga L1 = − Ga 4 µν µν where a abc b c Ga Aµ Aν .

which induces the homogenous transformation on the ﬁeld strength tensor: abc b δξ Ga Gµν ξ c µν = gf . To get the equations of motion one must vary the action over Aa µ. ∂ν We see that 1 c ∂Gc ∂L λρ = − Gλρ a ∂Aµ 2 ∂Aa µ ∂Gc λρ µ d µ d = gf acd (gρ Aλ − gλ Aρ ) a ∂Aµ Hence.14) (1. ξ ].13) . δξ Gµν = −ig [Gµν . where ζ = [η.15) ∂L ∂ (∂ν Aa µ) − ∂L =0 ∂Aa µ (1.6). ξ ].1. is: a abc b c δξ Aa Aµ ξ µ = ∂µ ξ + gf .14 The inﬁnitesimal form of (1. the gauge transformation under which the Lagrangian is invariant.1.1.1. ∂L c = gf acd Ad ρ Gµρ ∂Aa µ Analogously ∂L 1 c ∂Gc λρ = − Gλρ a ∂ (∂ν Aa ) 2 ∂ ( ∂ A ν µ) µ ∂Gc λρ ν µ ν µ ac = (gλ gρ − gρ gλ )δ a ∂ (∂ν Aµ ) (1. δξ Aµ = ∂µ ξ − ig [Aµ . [δη . δξ ]Aµ = −igδζ Aµ . ξ ]. It can be easily checked that the local gauge transformations form a closed algebraic structure.

16) (1.21) The last two terms give us the Feynman rules for the bosonic self-interactions. (1. In an Abelian theory (f abc = 0) this is not the case.1.19) where a c abc c jν = −gf abc Ab ∂µ (Ab µ Gµν − gf µ Aν ). (1.1.1. Equivalently. They manifest that the gauge bosons curry color and are sources of themselves. ∂L = Ga µν a ∂ (∂ν Aµ ) Thus.17) Or in terms of the Lie generators ta Dµ Gµν = ∂µ Gµν − ig [Aµ .1. (1.17) can be rewritten as a a ∂µ Fµν = jν (1. Expressed in terms of the As the free bosonic Lagrangian (1.(1.18) (1.(1. The 3-vertex (VVV) is (all momenta inwards): .1.15).12) appears as: 1 a a 1 2 abe cde a b c d b c L1 = − Fµν Fµν − gf abc (∂µ Aa f Aµ Aν Aµ Aν ν )Aµ Aν − g f 4 4 (1.1.1.1.16) ⇒ abc b c ∂ µ Ga Aµ Gµν = 0 µν + gf (1.1. Gµν ] = 0.1.15 Hence.1.20) The currents in the above equations stem from the fact that Yang-Mills equations are not linear.14).

24) . k2 (1. p.1. In the Feynman gauge it is: ¡ k µν .1.22) f abe f cde (ηµρ ηνσ − ηµσ ηνρ ) + d. µ b. σ q k p a. p. ρ ¡ ¡ c.23) The propagator of the gauge boson ﬁeld can be found by the standard FadeevPopov quantization method [13]. ν = gf abc ηµν (k − p)σ + ηνσ (p − q )µ + ησµ (q − k )ν (1. σ + f ace f bde (ηµν ηρσ − ηµσ ηρν ) + + f ade f bce (ηµν ηρσ − ηµρ ηνσ ) (1. µ 16 abc YM = Vµνσ (k.ab (k ) = − b. q ) = b. µ while the 4-vertex (VVVV) is: a.a. ν abcd = Vµνρσ = −ig 2 c. ν = DF i µν ab η δ .1. q ) = gf abc Fµνσ (k.

In addition to the pure bosonic vertices (1.1. one varies the full Yang-Mills Lagrangian (1.(1.26) or a a ∂µ Fµν = jν where a c abc µ c ¯ ν a jν (A.1.1.1. to get the equations of motion for Ψ and Aa µ . Now.17 Let us take into account the terms of the Lagrangian that include fermions. µ = igγµ ta (1.27) .23) the fermion-boson term ¯ AΨ gives us the vertex rule: gΨ ¡ a.1.1.11). Ψ) = −gf abc Ab ∂ (Ab µ Gµν − gf µ Aν ) − g Ψγ t Ψ. We need the terms: ∂L c ¯ a = gf acd Ad ρ Gµρ + g Ψγµ t Ψ ∂Aa µ ∂L = Ga µν a ∂ (∂ν Aµ ) ∂L ¯ = i ∂ Ψ + g AΨ ∂Ψ ∂L ¯ =0 ∂ (∂µ Ψ) Hence for the fermionic ﬁeld one has: i ∂ Ψ + g AΨ = 0 While for the gauge ﬁeld: abc b c ¯ γ ν ta Ψ = 0 ∂µ Ga Aµ Gµν + g Ψ µν + gf (1.22).25) (1.

28) . =D F k (1.18 The propagator of the pure fermionic ﬁeld is the well known: µ ¡ k ν ˜ µν (k ) = iηµν .1.

19

1.2

rank-2 tensor (part I)

Tensor Gauge theory is constructed under the restrictions posed by the fundamental principle of local gauge invariance. What we need for such a construction is to extend the gauge principle to higher spin gauge ﬁelds. On this thesis we will concentrate on the lowest non-trivial tensor gauge ﬁelds; those of the 2nd rank. We will see how this constructive procedure works [31, 32]. Actually, to describe the ﬁeld Aµν we will also need to introduce an auxiliary 3rd rank tensor gauge ﬁeld Aµνλ . Although the 2nd rank tensor ﬁeld have no symmetry properties between its indices, the 3rd rank tensor ﬁeld is symmetric under the interchange of its last two indices. The higher ﬁeld strength tensors are deﬁned as follows: 2nd-rank ﬁeld strength:

a a abc c b c Ga ( Ab µν,λ = ∂µ Aνλ − ∂ν Aµλ + gf µ Aνλ + Aµλ Aν )

(1.2.1)

or in terms of the Lie generators ta Gµν,λ = ∂µ Aνλ − ∂ν Aµλ − ig [Aµ , Aνλ ] + [Aµλ , Aν ] 3rd-rank ﬁeld strength:

a a abc c b c b c b c Ga ( Ab µν,λρ = ∂µ Aνλρ − ∂ν Aµλρ + gf µ Aνλρ + Aµλ Aνρ + Aµρ Aνλ + Aµλρ Aν )

= Dµ Aνλ − Dν Aµλ , (1.2.2)

(1.2.3) or Gµν,λρ = ∂µ Aνλρ − ∂ν Aµλρ − ig [Aµ , Aνλρ ] + [Aµλ , Aνρ ] + [Aµρ , Aνλ ] + [Aµλρ , Aν ] = Dµ Aνλρ − Dν Aµλρ − ig [Aµλ , Aνρ ] + [Aµρ , Aνλ ] . =

(1.2.4)

20

It should be noted that in the deﬁnition of the higher rank ﬁeld strength tensors apart from the tensor gauge ﬁelds of the same order, lower rank ﬁelds are also needed. We now come to the crucial point to deﬁne the gauge transformations for the higher rank tensor gauge ﬁelds. They should be deﬁned so as to form a closed algebraic structure. This means that the commutator of two extended gauge transformations should lead to another gauge transformation. To do this, we need to introduce higher

a a rank gauge parameters: The vector ξµ = ξµ t and the symmetric second rank tensor a a t . We deﬁne the extended gauge transformations as follows: ξµν = ξµν

δξ Aµ = ∂µ ξ − ig [Aµ , ξ ] δξ Aµν = ∂µ ξν − ig [Aµ , ξν ] + [Aµν , ξ ] δξ Aµνλ = ∂µ ξνλ = −ig [Aµ , ξνλ ] + [Aµν , ξλ ] + [Aµλ , ξν ] + [Aµνλ , ξ ] (1.2.5)

It is not hard to see that the above deﬁned extended gauge transformations form a closed algebraic structure. Indeed, the commutator of two gauge transformations acting on a 2nd rank tensor gauge ﬁeld is: [δη , δξ ]Aµν = −ig δη = −ig [Aµ , ξν ] + [Aµν , ξ ] − δξ [Aµ , ην ] − ig [Aµν , η ]

∂µ [η, ξν ] + [ην , ξ ] − ig [Aµ , ([η, ξν ] + [ην , ξ ])] − ig [Aµν , [η, ξ ]]

= −ig { ∂µ ζν − ig [Aµ , ζν ] − ig [Aµν , ζ ] } = −ig δζ Aµν This is again a gauge transformation with gauge parameters ζ, ζµ which are given by the following expressions: ζ = [η, ξ ], ζν = [η, ξν ] + [ην , ξ ].

The commutator of two gauge transformations acting on a rank-3 tensor gauge

21

ﬁeld is: [δη , δξ ]Aµνλ = −ig δη [Aµ , ξνλ ] + [Aµν , ξλ ] + [Aµλ , ξν ] + [Aµνλ , ξ ] − −δξ = −ig [Aµ , ηνλ ] + [Aµν , ηλ ] + [Aµλ , ην ] + [Aµνλ , η ] ∂µ [η, ξνλ ] + [ην , ξλ ] + [ηλ , ξν ] + [ηνλ , ξ ] − =

−ig [Aµ , ([η, ξνλ ] + [ην , ξλ ] + [ηλ , ξν ] + [ηνλ , ξ ])] − −ig [Aµν , ([η, ξλ ] + [ηλ , ξ ])] − ig [Aµλ , ([η, ξν ] + [ην , ξ ])] − ig [Aµνλ , [η, ξ ]] = −ig { ∂µ ζνλ − ig [Aµ , ζνλ ] − ig [Aµν , ζλ ] − ig [Aµλ , ζν ] − ig [Aµνλ , ζ ] } = = −igδζ Aµνλ , where ζ = [η, ξ ], ζν = [η, ξν ] + [ην , ξ ], ζνλ = [η, ξνλ ] + [ην , ξλ ] + [ηλ , ξν ] + [ηνλ , ξ ]. =

Note that the gauge parameters are not separately closed on each order. This happens only for the lowest, Yang-Mils, case. As we depart from this, we see that the parameters of a given order mix not only between themselves but also with those of lower orders. The gauge transformations (1.2.5) induce the homogenous gauge transformations on the ﬁeld strength tensors.

abc c b c δGa ( Gb µν,λ = gf µν,λ ξ + Gµν ξλ )

δξ Gµν,λ = −ig [Gµν , ξλ ] + [Gµν,λ , ξ ]

(1.2.6)

abc c b c b c b c ( Gb δGa µν,λρ = gf µν,λρ ξ + Gµν,λ ξρ + Gµν,ρ ξλ + Gµν ξλρ )

δξ Gµν,λρ = −ig [Gµν , ξλρ ] + [Gµν,λ , ξρ ] + [Gµν,ρ , ξλ ] + [Gµν,λρ , ξ ]

(1.2.7)

These can similarly be considered as appearing in the expansion of the big 1 .22 What remains now is to ﬁnd the Lagrangian which describes the dynamics of the 2nd rank tensor gauge ﬁeld.8) Indeed it can be easily checked that the above Lagrangian remains invariant under the gauge transformations (1.. 33. To do this. Aµν .2. e) = s=0 1 a A (x)La eλ1 .. . ∞ Gµν (x. we will prove that (1. 29.λ Gµν.6) and (1. One might guess that the most general one is the following: 1 a a 1 a L2 = − Ga µν. e) is the connection.3.2.3. e) the curvature tensor on an extended vector bundle X. eλs s! (1.7). .. e) = s=0 1 a Gµν.. .λρ . ∞ Aµ (x.λs . 36]. 30]. . a geometrical picture of the above construction is needed [31.λs . However.2) The extended gauge ﬁeld Aµ (x.. while the ﬁeld strength Gµν (x.Gauge bosons We can think of the tensor gauge ﬁelds Aµ .λ1 .. .λ and Gµν..λλ 4 4 (1.. e) over the tangent vector eµ [28. eλs s! µλ1 .. 1..2.λs (1.8) is not the most general Lagrangian that can be constructed out of the ﬁeld strengths Gµν.λ − Gµν Gµν.λs (x)La eλ1 . with a structure group G the elements of which U (ξ ) = eigξ(x.3 Geometrical Interpretation .2. as x-dependent coeﬃcients appearing in the expansion of the extended gauge ﬁeld Aµ (x. Aµνλ .1) The same holds for the ﬁeld strengths Ga µνλa .e) can be parametrized through the gauge parameters a ξλ .

.λs − ∂ν Aµλ1 ... .. .. e) = ∂µ Aν − ∂ν Aµ − ig [Aµ .. eλs eρ1 .. . .λs L eλ1 . ∞ ξ (x.λs )L eλ1 .λs = La eλ1 ...... .4) We will now prove that Gµν (x. .3) The extended algebra space is spanned by the inﬁnite many generators La λ1 . Aν ] ∂µ Aν − ∂ν Aµ − ig [Aµ .ρr λ1 . ..λs ) + s! s + gf ∞ abc k=0 n k c a Ab µλ1 .23 ξ (x....ρr [L ... .. eλs − s! ∞ −ig s ∞ 1 s! s n k b a b Aa µλ1 .λs − ∂ν Aµλ1 .. . . L ]eλ1 . ..3...λk Aνλk+1 .5) = s=0 1 a G La eλ1 . which obey the commutation relations: b abc c [L a Lλ1 . .λs The extended covariant derivative is deﬁned as: Dµ = ∂µ − ig Aµ . . e) over the tangent vector eµ ... eλs = Gµν (x..λs Ab νρ1 . eλs = (1.. eλs = k=0 = s=0 1 a (∂µ Aa νλ1 . e) s! µν.λs (1. eλk eλk+1 . eρr = s!r! −ig s..λs . . eλs s! λ1 .. Lρ1 . . L ]eλ1 .λs )L eλ1 . .λs − ∂ν Aµλ1 . e) = s=0 1 a ξ (x)La eλ1 .λk Aνλk+1 ..λs ρ1 ..r ∞ = s=0 1 a a (∂µ Aa νλ1 . eλs ..λ1 ..3.ρr ] = if (1. eλs − s! 1 a a b Aµλ1 .λs [L .... .. .3.. Aν ] = ∞ = s=0 1 a a (∂µ Aa νλ1 . ..

perm. Dν ]] = 0. Gνλ + Gνλ. ∂ν − ig Aν ] = −ig∂µ Aν + ig∂ν Aµ − g 2 [Aµ .9) (1.3. where Dµ = ∂µ − igAµ . The linear term in eρ gives: [Dµ . [Dν . if we expand the above equation over g. Aν ] = = −ig ∂µ Aν − ∂ν Aµ − ig [Aµ .ρ eρ ] + cyc.ρ + ∂λ Fµν. Gµν ] = 0 (1.3. Gνλ. We have. Gµν. Let us now expand equation (1. Aν ] = −ig Gµν (1. Gνλ ] + [Dν . Gλµ ] + [Dλ .6) is transformed into the generalized Bianchi identity [Dµ . Dλ ]] + [Dν . Dµ ]] + [Dλ .λ one can independently check the last identity and get convinced that it holds. Gµν and Gµν. Gνλ ] + [Dν . which with the aid of (1.6) The operators Dµ . Dλ obey the Jacobi identity: [Dµ .3. the zeroth order gives the Bianchi identity for the ﬁeld strength tensor Fνλ.7) over eρ up to linear terms.7) . [∂µ − igAµ − igAµρ eρ .ρ ] − ig [Aµρ . [Dλ . (1. Dν .8) Using the explicit form of the operators Dµ . Gλµ ] + [Dλ .ρ = 0.ρ + ∂ν Fλµ. Gλµ ] + [Dλ . Gµν ] = 0.3. Dν ] = [∂µ − ig Aµ . [Dµ .ρ ] − ig [Aνρ . Gνλ ] + [Dν . Now. + O(e2 ) = 0 In zero order the above equation gives the standard Bianchi identity in YM theory: [Dµ . Gµν ] = 0.ρ ] − ig [Aλρ .24 Now we see that [Dµ .3.ρ : ∂µ Fνλ. Gλµ.3.

3. U Aν U −1 − (∂ν U )U −1 = g g − = U (∂µ Aν − ∂ν Aµ )U −1 − ig (U Aµ Aν U −1 − U Aν Aµ U −1 ) + + i (∂µ U )∂ν U −1 − (∂ν U )∂µ U −1 + (∂µ U )U −1 ∂ν U −1 − (∂ν U )U −1 ∂µ U −1 = g (1. ξ ] (1. Aν ] = = (∂µ U )Aν U −1 + U (∂µ Aν )U −1 + U Aν (∂µ U −1 ) − −(∂ν U )Aµ U −1 − U (∂ν Aµ )U −1 − U Aµ (∂ν U −1 ) − i (∂µ ∂ν U )U −1 + (∂ν U )(∂µ U −1 ) − (∂ν ∂µ U )U −1 − (∂µ U )(∂ν U −1 ) − g i i −ig U Aµ U −1 − (∂µ U )U −1 . e) = U (ξ )Aµ (x.3. Gµν (x.13) .12) Similarly.10) = U (∂µ Aν − ∂ν Aµ − ig [Aµ . e) = ∂µ Aν − ∂ν Aµ − ig [Aµ . δ Aµ = ∂µ ξ − ig [Aµ .3. ξ ] (1. Lc ] = −gf bca ξ b Gµν La ⇒ La δ Gµν = igξ b Gµν c a δ Gµν = gf acb ξ b Gµν (1.3. e)U −1 (ξ ) − i ∂µ U (ξ ) U −1 (ξ ) g The gauge transformation over the Aµ (x.11) a c c [Lb . e). e) ﬁelds induces the corresponding homogenous gauge transformation on the ﬁeld strength tensors Gµν (x.25 We deﬁne the extended gauge transformation of the extended gauge ﬁeld Aµ (x. Indeed. Aν ])U −1 = U G µν U −1 For inﬁnitesimal transformations: Gµν = (1 + igξ )Gµν (1 − igξ ) = Gµν + igξ Gµν − ig Gµν ξ ⇒ δ Gµν = −ig [Gµν .

18) Indeed. La 2 (1. a abc c b c b c b c δAa (Ab µνλ = ∂µ ξνλ + gf µ ξνλ + Aµν ξλ + Aµλ ξν + Aµνλ ξ ).17) In a similar way we can obtain the homogenous gauge transformation induced on the ﬁeld strengths (1.2.16) For each order we have the equation for the gauge transformation of the tensor ﬁeld of the corresponding rank.6). We can now form invariant Lagrangians: a a Lµνλρ (x.13) one can compute term by term the gauge transformations of the higher rank tensor ﬁelds and reproduce the ones deﬁned in (1.2. a abc b c δAa Aµ ξ . (1.3. a c a c δ Lµνλρ = gf acb ξ b (Gµλ Gνρ + Gνρ Gµλ )=0 . . .26 Expanding (1. . Ineed. 1 a a a δ Aµ = δξ Aa µ + (δξ Aµν )eν + (δξ Aµνλ )eν eλ + .14) (1.2. µ = ∂µ ξ + gf a abc c b c δAa (Ab µν = ∂µ ξν + gf µ ξν + Aµν ξ ). L 2 1 a a ∂µ ξ = ∂µ ξ a + (∂µ ξν )eν + (∂µ ξνλ )eν eλ + . L 2 µνλ (1. . .3. ξ ] = gf abc + c b c b c Ab µ ξ + (Aµν ξ + Aµ ξν )eν + 1 b c c b c b c a (A ξ + Ab µν ξλ + Aµ ξνλ + Aµ ξνλ )eν eλ + .7). e) = Gµλ Gνρ (1.3. (1. .3.3.3.15) [Aµ .5).

λ2 + Gµλ. There are 15 ways of contracting them.22) (1.λ1 Gνρ.λ Gµλ.3.λ2 + Gµλ. e) to orders of e we can ﬁnd invariant Lagrangians for every rank. eλs ⇒ k=0 1 (2) (0) (1) Lµνλρ + Lµνλρ.λ1 + Gµλ.19) Comparing the terms.λ1 λ2 Gνρ (2) (1. .λ2 Gνρ.3. = 2 a a a a a = Ga µλ Gνρ + (Gµλ Gνρ. .27 Expanding L(x.λ1 eλ1 + Lµνλρ.λλ 4 4 and 1 a 1 a 1 a a a a LII 2 = Gµν. . there are only two independent scalar Lagrangians which we call: 1 a a 1 a a LI 2 = − Gµν. .λs (x)eλ1 . eλs = s! ∞ s=0 1 s! s s k a Ga µλ.λ1 λ2 eλ1 eλ2 + .. So we see: a Ga Lµνλρ (x.λs eλ1 ..3.23) .λ1 λ2 + Gµλ..3.ν Gµλ. we see that: a a a a a a a Lµνλρ. the Lagrangian which describes the dynamics of the free rank-2 tensor gauge ﬁeld must be a linear combination of the above Lagrangians: II L2 = g2 LI 2 + g2 L2 (1.3.λ − Gµν Gµν. .λ1 Gνρ )eλ1 + + 1 a a a a a a a Ga µλ Gνρ.νλ 4 4 2 (1.λ1 λ2 = Ga µλ Gνρ.λ1 λ2 Gνρ eλ1 eλ2 2 (1.λ1 λ2 + Gµλ.ν + Gµν..λk+1 . we have to contract the indices.λ2 Gνρ. e) = Gµλ νρ ⇒ ∞ s=0 1 (Lµνλρ )λ1 .λk Gνρ.λ1 + Gµλ.λ Gµν. Nevertheless..21) Hence.20) Since we need scalar Lagrangians. We are interested in Lagrangians describing 2nd rank tensors.λ + Gµν Gµλ..λ1 .λ1 Gνρ. .λ1 + Gµλ.

4 rank-2 tensor (part II) We are now ready to examine the properties of the rank-2 tensor Lagrangian (1. enhances the gauge symmetry at the free level and makes possible the cancellations of all the negative norm states.3.4. It will be seen that the primary constraints of the gauge system -Bianchi identities(1.σ a ) a ) a ∂ (∂λ Aa ) 2 ∂ ( ∂ A 2 g ∂ ( ∂ A ∂ ( ∂ A λ µν 2 λ µν λ µν ) µν = Ga µλ.4.τ ρσ.τ φ =0 ∂ (∂λ Aa µν ) ∂ L2 ∂ (∂λ Aa µν ) − ∂ L2 =0 ∂Aa µν (1.ρ ηµν + Gµρ.23). In the previous section it was seen that the Lagrangian which describes the dynamics of the rank-2 tensor gauge boson is: 1 a a g2 1 a 1 a 1 a a 1 a a Gµν.λ Ga L2 = − Ga µν.λλ + µλ.σ ρτ.τ ρσ.λ + Gρλ.ν Gµλ. ∂Gb ρσ.ν + Gµν.9) will determine the ratio of the coeﬃcients g2 g2 [36]. Further.1) Let us ﬁnd the Euler-Lagrange equations for the rank-2 tensor gauge ﬁeld Aa µν .3.λ Gµν.4.ρ ηνλ ) 2 g2 νλ.3.4) .τ ρτ.2) b ∂Gb ∂Fρσ. this particular choice of the ratio.ν − 1 g2 a a a (G + Ga µν.τ b b = − G + G + G ρσ.νλ 4 4 g2 4 4 2 (1.τ ρσ.28 1.4.µ = (1. as dictated by the zeroth order expansion over g of the Bianchi identity (1.8). ∂λ We see that ∂Gb ρσ =0 ∂ (∂λ Aa µν ) .τ = = (ηλρ ηµσ − ηλσ ηµρ )ηντ δ ab a ) ∂ (∂λ Aa ) ∂ ( ∂ A λ µν µν (1.λ − Gµν Gµν.3) ∂Gb ∂Gb ∂Gb ∂ L2 1 b 1 g2 ρσ. that stem from this Lagrangian.λ + Gµν Gµλ.

µ and (∂µ Fµλ. 2 g2 (1. taking g = 0.ν + ∂µ Fλν. of both terms in the equation are equal to zero separately. We see that a ∂λ ∂µ Fµν. due a to the antisymmetric properties of the ﬁeld strength tensor Fµν. over the ﬁrst free index ν .ρ ) = Jνλ (g ).µ + ηνλ ∂µ Fµρ.4. However.29 Just like in Yang-Mills case.λ a II ηνλ ∂µ Fµρ.4. Thus.ρ } = 0.5) Now. the equations of motion that describe the rank-2 tensor gauge boson are not linear.ρ } = 0. which arise from LI 2 and L2 respectively.λ − 1 g2 a a a a a (∂µ Fµλ.ν equation (1. Let us take the derivative over λ of the left-hand side of the equation (1.µ + ηνλ ∂µ Fµρ. We isolate all the g-terms which contribute to the sourceself current on the right hand side of the equation. it is not obvious to verify the cancellation alone from the antisymmetric properties of the ﬁeld strength tensor. It consists of two parts: ∂µ Fµν.µ + ηνλ ∂µ Fµρ.4.λ = 0.λ = 0. a ∂µ Fµν.µ + ηνλ ∂µ Fµρ.µ + ∂λ Fµν.5). There are two free indices in a a a a + + ∂λ Fµν.ν + ∂µ Fλν. let us ignore the self-interactions. let us take advantage of the Bianchi . Indeed.µ + ∂λ Fµν.5).ρ } = ∂λ {∂µ Fλν.ρ ).µ + ∂λ Fµν. as well as a a a a ∂ν {∂µ Fµλ. What about the second index λ? Here the cancellation is not that trivial. we have a ∂ν ∂µ Fµν. The derivatives.ν + ∂µ Fλν.µ + ∂µ Fλν.µ + ∂λ Fµν. as well as a a a a a a a ∂λ {∂µ Fµλ.λ under the exchange of µ and ν .

4 4 4 4 2 (1.λ − Gµν Gµν.ρ )= 2 g2 (1.λ − (1.4.ρ + ∂µ ∂λ Fνλ.9) derived in the previous section.4.30 identities (1.λ the above equation implies: g2 = g2 .4.µ + ∂µ ∂ν Fµρ. II Hence. Taking the derivative of the Bianchi identity over ∂µ and setting ν = ρ we get ∂ 2 Fµν.λ + Gµν Gµλ.ν Gµλ.4. with the aid of (1. the Bianchi identity forces LI 2 and L2 to appear with the same weight in the linear combination of L2 .µ + ∂ν ∂µ Fµρ.λ Gµλ.µ + ∂ 2 Fµν.ν + Gµν.8) .λ g2 a For arbitrary tensors Fµν.νλ .λ Gµν.3.7) = 1− g2 a ∂µ ∂λ Fµν. 1 1 a a 1 a 1 a 1 a a a a a L2 = − Ga µν.6) we get a 0 = ∂µ ∂λ Fµν.λλ + Gµν.µ ≡ 0 Hence.6) 1 g2 a a a (∂λ ∂µ Fλν.

the theory is endowed with an enhanced gauge symmetry [31].Chapter 2 Unitarity of the theory at the non-interactive level Now that we have the ﬁnal form of the Lagrangian which describes the dynamics of the rank-2 tensor gauge boson. 2. we can study its properties. It is because of this enhancement that the negative norm states do not appear in the experimentally testable quantities [36]. it will be instructive to present a general method for counting the propagating modes of a gauge theory.1 A method for counting the propagating modes of a gauge theory Before we begin the study of the properties of the free rank-2 tensor gauge theory. gauge ﬁxing independent method was considered an essential missing piece of the bibliography and we present 31 . A general. Our ﬁrst task will be to check the unitarity of the theory at the free level (g = 0). We will see that at the non-interactive level. Nevertheless. when the number of degrees of freedom becomes large such a procedure becomes rather cumbersome. In the bibliography this is usually done in a straightforward gaugeﬁxing approach.

Because the matrix operator H (k ) explicitly depends on the momentum kµ .32 it here.1) Analyzing the rankH of the matrix operator H one can observe that it depends on 2 2 = 0 . (2.1.1.on mass-shell momenta vector space consists of pure gauge ﬁelds.1.the . (2.oﬀ mass-shell momenta . Therefore the number of propagating modes can be calculated from the following relation of propagating modes = N (k )|k2 =0 − N (k )|k2 =0 = rankH |k2 =0 − rankH |k2 =0 . 2. then the vector space has the dimension N = d − r. When kµ the value of momentum square kµ 2 = 0 . .1 The method The vector space of independent solutions of a system of equations crucially depends on the rank of the equation matrix.2) Let us consider for illustration some important examples. If the matrix operator H . its rankH = r also depends on momenta and therefore the number of independent solutions N depends on momenta N (k ) = d − r(k ) . When kµ the vector space consists of pure gauge ﬁelds and propagating modes. the equation matrix of a free gauge theory in momentum space. has dimension d × d and its rank is rankH = r.

0. and the number of independent solutions increases: 4-1=3. then the rank of the matrix drops. the rank of the 4-dimensional matrix Hαγ is rankH |ω2 −k2 =0 = 3 and the number of independent solutions is 4-3=1. 0.4) where Aµ = eµ exp (ikx). If ω 2 − k 2 = 0. These three solutions of equations (2.33 2. 0.1. As one can see from the relation Hα γ (k )kγ = 0 this solution is proportional to the momentum eµ = kµ = (−ω. e(1) . k ) and the matrix operator H takes the form −k 2 0 0 −kω 0 ω2 − k2 0 0 γ Hα = .1. rankH |ω2 −k2 =0 = 1. This is a consequence of the gauge invariance of the theory eµ → eµ + akµ . 2 2 0 0 ω −k 0 2 kω 0 0 ω If ω 2 − k 2 = 0. γ γ 0 1 2 0 1 0 0 gauge) e( γ .2 Vector Gauge Field The kinetic term of Lagrangian which describes the propagation of free vector gauge ﬁeld is 1 K = − Fµν F µν 4 and the corresponding equation of motion in momentum space is γ Hα γ eγ = (−k 2 δα + kα k γ )eγ = 0.1.1. We can always choose the momentum vector in the third direction k µ = (ω. (2. 0.4) are −1 0 0 1 0 1 0 (2) =√ = .3) (2. e = . k ) and is a pure-gauge ﬁeld.

5) that respects the symmetry properties of Aµν .1.1.1. while the remaining two are the physical modes. c1 . perpendicular to the direction of the wave propagation. The corresponding matrix operator is: 1 2 Hααγ ´ γ ´ (k ) = {ηαα ´ ηγ γ ´ − (ηαγ ηα ´γ ´ + η αγ ´ ηαγ ´ )}k − ηαα ´ kγ kγ ´ − ηγ γ ´ kα kα ´ + 2 1 + (ηαγ (2.3 Symmetric Tensor Gauge Field The free gravitational ﬁeld is described in terms of a symmetric second-rank tensor ﬁeld hµν and is governed by the Einstein and Pauli-Fierz equation: ∂ 2 hνλ − ∂ν ∂ µ hµλ − ∂λ ∂ µ hµν + ∂ν ∂λ hµµ + ηνλ (∂ µ ∂ ρ hµρ − ∂ 2 hµµ ) = 0. (2. We see that the number of propagating modes is rankH |ω2 −k2 =0 − rankH |ω2 −k2 =0 = 3 − 1 = 2. as it should be.6) (2. . where a. which is invariant with respect to the gauge transformations δhµλ = ∂µ ξλ + ∂λ ξµ .7) ´ kγ kα ´ + ηα ´γ ´ kα kγ + ηαγ kα ´ kγ ´ + ηαγ ´ kα kγ ´) 2 and is a 10 × 10 matrix in four-dimensional space-time with the property Hααγ ´ γ ´ = Hααγ ´ γ ´ = Hαα ´γ ´ γ = Hγ γ ´ αα ´. The general solution at ω 2 − k 2 = 0 will be a linear combination of these three eigenvectors: (2) eγ = akγ + c1 e(1) γ + c2 e γ .34 from which the ﬁrst one is a pure gauge ﬁeld (∼ kγ ). 2. c2 are arbitrary constants.1.

0. These four symmetric solutions are pure-gauge tensor ﬁelds.1. This leaves us with 10 − 4 = 6 solutions. k 0 0 0 0 .1. ´ (2. 0 0 0 0 0 0 (2. 0 0 −ω 0 0 0 0 0 0 −ω 0 0 0 0 0 0 k −2ω 0 0 k 0 0 k 0 .1. then one can ﬁnd the following four linearly independent −ω 2 0 0 0 0 0 eγ γ ´ = 0 0 0 0 0 0 solutions: 0 0 −ω 0 −ω 0 . These are the four pure-gauge solutions (2. if again we choose the coordinate system such that k γ = (ω. 0 0 0 2 k 0 k 0 0 k 0 .1. 0.9) When ω 2 − k 2 = 0.8) pure-gauge ﬁeld solutions of the form (2. (1) (2) . 0 0 0 0 Thus the general solution of the equation on mass-shell is: eγ γ ´ = ξγ kγ ´ + ξγ ´ kγ + c1 eγ γ ´ + c2 e γ γ ´.6) and two additional symmetric solutions representing propagating modes: the helicity states of the graviton 0 0 0 0 0 0 0 1 0 0 0 0 (1) (2) eγ γ . then the rank of the matrix Hααγ ´ γ ´ (k ) drops and is equal to rankH |ω2 −k2 =0 = 4.1. eγ γ ´ = ´ = 0 0 −1 0 0 1 0 0 0 0 0 0 0 0 (2.10) 1 0 .6) eγ γ ´ = kγ ξγ ´ + kγ ´ ξγ as one can see from the relation γγ ´ Hαα (k )(kγ ξγ ´ + kγ ´ ξγ ) = 0. the rank of the 10-dimensional matrix Hαα ´ rankH |ω2 −k2 =0 = 6 and the number of independent solutions is 10 − 6 = 4. Indeed.35 γγ ´ (k ) is equal to If ω 2 − k 2 = 0. k ).

1.1. 40.1. ´ = −ω 0 0 k 0 0 0 0 0 −k 0 0 −k 0 solutions: 0 0 k 0 . The corresponding matrix operator is: Hααγ ´ γ ´ (k ) = − 1 2 (ηαγ ηα ´γ ´ − ηαγ ´ ηαγ ´ )k − 2 1 − (ηαγ ´ kγ kα ´ − ηα ´γ ´ kα kγ + ηαγ ´ kα kγ ´ − ηαγ kα ´ kγ ´) 2 (2. as it should be.1. 41.12) (2. c2 are arbitrary constants.1. These three antisymmetric solutions are pure-gauge ﬁelds. 0 0 0 −1 0 0 1 0 0 0 . 0. 2. Indeed. 42]: ∂ 2 Bνλ − ∂ν ∂ µ Bµλ + ∂λ ∂ µ Bµν = 0 which is invariant with respect to the gauge transformations δBµλ = ∂µ ηλ − ∂λ ηµ . one can ﬁnd the following three 0 0 ω 0 0 ω 0 0 0 0 0 −ω 0 0 eγ γ . the rank of the 6-dimensional matrix Hαα (k ) is equal to ´ rankH |ω2 −k2 =0 = 3 and the number of independent solutions is 6 − 3 = 3.14) 0 0 0 0 0 0 .36 where c1 . 0.4 Antisymmetric Tensor Gauge Field The antisymmetric second-rank tensor ﬁeld Bµν is governed by the equation [39. We see that the number of propagating modes is rankH |ω2 −k2 =0 − rankH |ω2 −k2 =0 = 6 − 4 = 2. γγ ´ If ω 2 − k 2 = 0. in the coordinate system k γ = (ω.11) that respects the symmetry properties of Bµν .13) and is 6 × 6 matrix in four-dimensional space-time with the property Hααγ ´ γ ´ = −Hααγ ´ γ ´ = −Hαα ´γ ´ γ = Hγ γ ´ αα ´. (2. k ). (2.

This leaves us with 6 − 2 = 4 solutions. We see that the number of propagating modes is rankH |ω2 −k2 =0 − rankH |ω2 −k2 =0 = 3 − 2 = 1. (A) 2.15) When ω 2 − k 2 = 0.λ Fµλ.12) and the antisymmetric solution representing the propagating mode: the helicity zero state eγ γ ´ (A) 0 0 0 0 (2.1. 4 4 4 (2.ν + Fµν.1. ´ (2. The coupling constant is set to zero (g = 0) and all all ﬁelds of rank other than 2 are ignored. These are the three puregauge solutions (2.16) 0 0 1 0 = .1.17) .5 Rank-2 tensor gauge theory We begin with the pure kinetic term of the 2-rank bosonic Lagrangian. 0 −1 0 0 0 0 0 0 Thus on mass-shell the general solution of the equation is: eγ γ ´ = kγ ηγ ´ − kγ ´ η γ + c3 e γ γ ´ .λ . where c3 is arbitrary constant. We have: 1 a 1 a 1 a a a a K = − Fµν.λ Fµν.1.ν Fµλ. then the rank of the matrix Hααγ ´ γ ´ (k ) drops and is equal to rankH |ω2 −k2 =0 = 2.1.12) eγ γ ´ = kγ ηγ ´ − kγ ´ ηγ as one can see from the relation γγ ´ Hαα (k )(kγ ηγ ´ − kγ ´ ηγ ) = 0.37 pure-gauge ﬁelds of the form (2.1. After this parenthetic discussion we shall turn to the tensor gauge theory.λ + Fµν.

the second and the forth terms and used the fact that the third term is identically equal to zero. Just from the symmetry properties of the ﬁeld strength tensor it is not obvious to see why the rest of the terms are equal to zero.λ .1. Nevertheless.ν 1 a 1 a a a a Fµν. The δ ˜ transformation of with respect to the complementary gauge transformation δ K is ˜K = − 1 F a ∂λ (∂µ η a − ∂ν η a ) + 1 F a ∂ν (∂µ η a − ∂λ η a ) + 1 F a ∂λ (∂µ η a − ∂λ η a ) = δ ν λ λ µ µ µ 2 µν. we can use the g-zeroth order Bianchi identities (1.λ + ∂ν Fλµ. (2.λ = 0.19) = 2 2 where we combined the ﬁrst. δ µν µ a Indeed.3.ν ∂λ (∂µ ηλ − ∂λ ηµ ) (2.1.λ transforms with respect to these transforma- tions as follows: a δξ Fµν.λ ∂λ ∂ν ηµ + Fµν.38 We notice here that apart from the the usual gauge symmetry a δξ Aa µν = ∂µ ξν the theory is also invariant under the complementary gauge transformation ˜η Aa = ∂ν η a .9) for the a ﬁeld strength tensor Fµν.λ 2 µν.18) Therefore the kinetic term K is obviously invariant with respect to the ﬁrst group of gauge transformations (δξ K = 0). (2.λ = 0.1. but it is less trivial to see that it is also invariant ˜.20) . a a ˜η F a δ µν. the ﬁeld strength tensor Fµν.λ = ∂λ (∂µ ην − ∂ν ηµ ).λ 2 µν.λ + ∂µ Fνλ. We use the identity in the form a a a ∂λ Fµν.

λ − (∂µ Fµλ.3.39 Ignoring total divergence terms. Is this enhanced gauge symmetry enough to guarantee the the self consistence of the theory? Is it enough to guarantee that the propagating modes it predicts are physical? To answer that we have to solve the equations of motion and study the solutions.ρ ) = 0.25) .1.1.19) is equivalent to ˜K = − 1 (∂λ F a )∂ν η a + (∂λ F a )(∂µ η a − ∂λ η a ) δ µ λ µν. the equations of motion for the free rank-2 bosonic Lagrangian (g = 0) are: 1 a a a a a ∂µ Fµν.23) + (2.1.18) when g2 = g2 .24) + Transforming the above equation to momentum space we get: γγ ´ Hαα (k ) Aa γγ ´ = 0.17) and (2.1.µ + ηνλ ∂µ Fµρ.1.17) is invariant under the general gauge transformation: a a a a Aa µν → Aµν = Aµν + ∂µ ξν + ∂ν ηµ . 2 Or in terms of the As: 1 a 1 a 1 a a ∂ 2 Aa − ∂ν ∂µ Aa νλ − Aλν µλ − Aλµ − ∂λ ∂µ Aνµ − Aµν 2 2 2 1 a 1 a 2 a ∂ν ∂λ Aa µµ − Aµµ + ηνλ (∂µ ∂ρ Aµρ − ∂ Aµµ ) = 0 2 2 (2.λ 2 Which with the aid of (2.ν µ µν.1.µ + ∂λ Fµν. (2.ν + ∂µ Fλν.21) ˜ transThis demonstrates the invariance of the action with respect to the δ and δ formations deﬁned by (1. (2.1. ´ (2.20) can be shown to nullify. equation (2. Hence the free Lagrangian (2.1.22) Now the question is the following.1. As we have seen in the previous section.

1.1.26) rankH |ω2 −k2 =0 = 9 and the number of linearly independent solutions is 16 − 9 = 7 [36]. These seven solutions are pure-gauge tensor potentials of the form eγ γ ´ = kγ ξγ ´ + kγ ´ ηγ . The matrix operator (2. It has the following where Hαα ´ form: 1 1 2 Hααγ ´ γ ´ (k ) = (−ηαγ ηα ´γ ´ + η αγ ´ ηαγ ´ + ηαα ´ ηγ γ ´ )k + ηαγ kα ´ kγ ´ + ηα ´γ ´ kα kγ 2 2 1 − (ηαγ ´ kα ´ kγ + ηαγ ´ kα kγ ´ + η αα ´ kγ kγ ´ + ηγ γ ´ kα kα ´ ). eγ γ ´ = −ω 2 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 ω 0 0 0 0 ω 0 0 (2.1. 0 0 0 0 0 0 0 0 0 0 k 0 0 k 0 0 0 0 0 0 0 0 0 0 (2. 0. 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 . . the rank of the 16-dimensional matrix Hαα (k ) is equal to ´ (2. In the reference frame. .29) which follows from the gauge invariance of the action and can also be checked explicitly. . k ). where k γ = (ω.27) 0 0 ω 0 ω 0 0 0 0 0 0 0 0 0 0 0 0 .26). 0 0 0 0 0 2 k 0 0 0 k k 0 0 0 0 ω 0 0 k 0 . γγ ´ If ω 2 − k 2 = 0. 2 with the property that Hααγ ´ γ ´ = Hγ γ ´ αα ´. .1.28) ω 0 0 k 0 0 0 0 as one can get convinced from the relation γγ ´ Hαα (k )(kγ ξγ ´ + kγ ´ ηγ ) = 0.40 γγ ´ (k ) is a matrix operator quadratic in momentum kµ . it has the form presented on Appendix A. ´ (2.1. in four-dimensional space-time is a 16×16 matrix. 0.

Indeed. These include the seven pure-gauge potentials (2.31) where c1 . e = ´ γγ ´ 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 1 0 . These are propagating modes of helicity-two (λ = ±2) and helicity-zero λ = 0 charged gauge bosons [31]. 0 − sin 2θ cos 2θ 0 0 0 0 0 0 − sin 2θ cos 2θ 0 = Λe(2) ΛT = .1. if we make a rotation around the z-axis 1 0 0 0 0 cos θ − sin θ 0 Λαβ = .27) plus three new solutions representing the physical propagating modes: 0 0 0 0 0 −1 0 0 (1) (2) eγ γ = . c3 are arbitrary constants. This leaves us with 16 − 6 = 10 solutions. c2 . We see that the number of propagating modes is three rankH |ω2 −k2 =0 − rankH |ω2 −k2 =0 = 9 − 6 = 3.28). (1) (2) (A) (2.1.1.30) Thus the general solution of the equation on mass-shell is: eγ γ ´ = ξγ ´ kγ + ηγ kγ ´ + c1 e γ γ ´ + c2 eγ γ ´ + c3 e γ γ ´ . 0 cos 2θ sin 2θ 0 0 0 0 0 . 0 sin θ cos θ 0 0 0 0 1 we shall get e(1) 0 0 0 0 e(2) 0 0 0 0 0 − cos 2θ − sin 2θ 0 = Λe(1) ΛT = . then the rank of the matrix Hααγ ´ γ ´ (k ) drops and is equal to rankH |ω2 −k2 =0 = 6.1. eA γγ ´ = 0 1 0 0 0 − 1 0 0 0 0 0 0 0 0 0 0 (2. (2.41 When ω 2 − k 2 = 0.

under the interchange (α ↔ α ).1) This can be decomposed into its symmetric. therefore it describes a helicity-zero state. Indeed. . 2.2 Propagator In the previous section we saw that the equation matrix for the pure kinetic part of the rank-2 tensor bosonic Lagrangian is: 1 1 1 2 Hαα. This result can also be derived from the consideration of the equations of motion for the symmetric and antisymmetric parts of the tensor gauge ﬁeld Aa µλ .42 Therefore the ﬁrst two solutions describe helicity λ = ±2 states. the third. antisymmetric solution remains invariant under a Lorentz transformation.β ´ = (−ηαβ ηα ´ + ηαβ ´ ηαβ ´ + η αα ´ )k + ´ ηβ β ´ β ´β g2 2 2 +ηαβ kα ´ + ηα ´ kα kβ − ´ kβ ´β 1 − (ηαα ´ + ηαβ ´ kα ´ + ηβ β ´ kα kα ´ kβ kβ ´ kβ + ηαβ ´ kα kβ ´) 2 (2. For the antisymmetric part of the tensor gauge ﬁelds Aa νλ the equation reduces to the equation which describes the propagation of massless boson with one physical polarization: the λ = 0 helicity state [40]. and antisymmetric parts. one can observe that for the symmetric part of the tensor gauge ﬁelds Aa νλ the equation reduces to the free equation of gravity [39]. On the other hand.2. which describes the propagation of massless tensor gauge bosons with two physical polarizations: the λ = ±2 helicity states. as it was done in [31].

7) .γ ´ (k ) = iηαγ ηα α.β ´ + ηαβ ´ ηαβ ´ )k + ´ − 2ηαα ´ ηβ β ´ (k ) = − (ηαβ ηα ´β ´ β g2 4 1 + (ηαβ kα ´ + ηαβ ´ kα ´ + ηα ´ kα kβ ) − ´ kβ ´ kβ + ηαβ ´ kα kβ ´β 4 1 − (ηαα ´ + ηβ β ´ kα kα ´ kβ kβ ´) 2 1 A H ´ β ´ (k ) = g2 αα.β β. There we will see this factor appears as the weight of the helicity zero quantum in the total energy of the rank-2 ﬁeld.3) Since. The equation that deﬁnes the propagator is: f ix Hα ´ γ ´γ ´ ´ (k )DF.2.43 1 1 S 2 Hαα.5) 1 a ∂ λ Aa µλ − ∂µ Aλλ = 0 2 (2.β 1 2 3 − (ηαβ ηα ´ − ηαβ ´ ηαβ ´ )k + ´β 4 1 + (ηαβ kα ´ + ηα ´ kα kβ ) ´ − η αβ ´ kα ´ kβ − ηαβ ´ kα kβ ´ kβ ´β 4 (2. 2 In this gauge one gets: 1 1 S.2. A convenient gauge is the following: 1 a ∂µ A a µλ − ∂λ Aµµ = 0.β ´ β (2.β ´ + η αβ ´ ηαβ ´ )k ´ − ηαα ´ ηβ β ´ (k ) = − (ηαβ ηα ´β ´ β g2 4 1 A.2.f ix 2 Hαα.6) It is remarkable that the antisymmetric component of H comes with a factor of 3 [46]. the matrix H is singular.2.f ix 1 2 Hαα. gauge ﬁxing is essential in the computation of the propagator.2) (2.2. In the next section the energy-momentum tensor of the rank-2 tensor theory will be given [36].β ´ − ηαβ ´ ηαβ ´ ) k ´ (k ) = 3 − (ηαβ ηα ´β ´ β g2 4 (2.4) (2.2.

1.γ γ ´ α.2.9) i S.2.31) into the energy-momentum of the tensor gauge ﬁeld theory.1) .11) 2.10) (2.f ix S Hα ( k ) D ( k ) = (ηαγ ηα ´γ ´ + η αγ ´ ηαγ ´ ) ´ ´ F.3.β β.3 The Energy-Momentum Tensor We would like to consider the contribution of the general solution (2. µν ¡ k We see that µν. This will test from another point of view the unitarity of the theory [36].β ´ β 2 (2.ab DF (k ) = − i 3g2 k2 (4η µλ η νρ + 2η µρ η νλ − 3η µν η λρ )δ ab (2.ab b.2.f ix A Hα ( k ) D ( k ) = (ηαγ ηα ´γ ´ − η αγ ´ ηαγ ´ ) ´ ´ F.λρ.8) This can be decomposed into a symmetric and an antisymmetric part: µν.44 from which we ﬁnd: a. One can expect that only transverse propagating modes eγ γ ´ ∼ c1 e γ γ ´ + c2 eγ γ ´ + c3 e γ γ ´ . (1) (2) (A) (2.β β.2. λρ = DF (k ) = = − 1 µλ νρ i µλ νρ µρ νλ µν λρ (η η − η µρ η νλ ) ( η η + η η − η η ) + 2 g2 k 3 (2.β ´ β 2 i A.γ γ ´ α.λρ.

λ Fµλ.17): 1 a 1 a 1 a a a a + Fµν.3. the energy momentum tensor for the Aµν ﬁeld is: Tµν = ∂µ Aρσ In order to calculate the term the ﬁeld strength tensor: ∂Fµλ. eγ γ ´ ∼ ξγ ´ kγ + ηγ kγ ´.6) ∂L − ηµν L ∂ (∂ν Aρσ ) (2.σ ) − − ηµν Fρτ.3.ν Fµλ. The free Lagrangian has the form (2.5) (2.τ = (ηµν ηρλ − ηλν ηρµ )ηστ . will have no contribution.σ + 2 4 (2.σ + (Fνσ.1.1.τ Fρσ.τ Fρσ.ν ) + (Fνλ.ρ + + 2 4 1 1 (∂µ Aσσ Fνρ.4) (2.λ ηρσ − Fρλ.ν + ∂µ Fλν.45 will contribute to the energy-momentum of the gauge ﬁelds and that the longitudinal.λ + Fµν.3.ρ )=0 2 By deﬁnition.λ − (∂µ Fµλ.λ 4 4 4 and the equation of motion for the Aµν ﬁeld are (2. ∂ (∂ν Aρσ ) hence it is easy to see that 1 1 ∂L = −Fνρ.λ ηνσ ) ∂ (∂ν Aρσ ) 2 2 and ﬁnally get: 1 Tµν = − ∂µ Aρσ Fνρ.ρ − Fρσ.ρ − Fρσ.23): 1 a a a a a ∂µ Fµν.σ Fλσ.3.ρ − ∂µ Aρν Fρσ.2) ∂L ∂ (∂ν Aρσ ) we need the expression for the derivative of .3) (2.σ + ηµν Fρσ.ν Fµν.ν ) − ηµν Fλρ. pure gauge ﬁelds.µ + ηνλ ∂µ Fµρ.µ + ∂λ Fµν.3.λ L = − Fµν.τ + 4 1 1 ∂µ Aρσ (Fνσ.

σ ∂ρ Aµρ + Fσρ.3. Choosing 1 Ψµνρ = Aµσ Fνρ.σ ∂ρ Aµσ − (Fνσ.λ . 2 which fulﬁlls the property Ψµνρ = −Ψµρν .3. (2.λ + Aµν Fλρ.τ Fρσ.σ Fλσ.10) The sum of (2.ρ + Aµσ Fσρ.3. 2 (2.σ + ηµν Fρσ.8) We can use this freedom to express Tµν solely in terms of the ﬁeld strength tensor Fµν.ν ) − ηµν Fλρ.λ ).46 With the aid of (2. 2 4 (2.9) (2.ρ + Fρµ.4) we can get that 1 ∂ρ Ψµνρ = Fνρ.3.6) and (2.σ Fνρ.4) one can compute the derivative of the energy-momentum tensor Tµν over its second index ν and demonstrate that it is zero: ∂ν Tµν = 0.7) and the total fourmomentum of the system Pµ = Tµ0 dV.σ Fνρ.σ Fνσ.3.11) .3.ν ∂ρ Aµσ + Fνσ.σ − (Aµσ Fνσ.7) The energy-momentum tensor is not uniquely deﬁned because one can add any term of the form ∂ρ Ψµνρ Tµν → Tµν + ∂ρ Ψµνρ where Ψµνρ = −Ψµρν without changing its basic property (2. and using (2.ν Fρλ.3.3.τ + 4 1 1 (Fµρ.λ .σ Fσρ.3.ρ + Fµρ.10) gives the ﬁnal form of the energy momentum tensor expressed in terms of ﬁeld strength tensors 1 Tµν = − Fµρ.ρ + + 2 4 1 1 + (Fµσ.λ ) − ηµν Fρσ.ν + Aµρ Fνλ.σ ∂ρ Aµν ). (2.σ Fρλ.3.ρ ∂ρ Aµσ + Fσρ.

j = 1.1.13) (A) (i) (A) (i) (A) (A) (A) (A) Thus.12) as it should be in a massless and scale invariant theory.1.λ = i(−kµ eνλ + kν eµλ ) where eµν is a general solution (2. Using the following orthogonality relations: kµ k µ = 0.3. (i) (j ) (i) (j ) kµ eµλ = kλ eµλ = 0 eµλ eµλ = eµλ eλµ = δ ij . we see that only the transverse propagating modes contribute to the energymomentum of the ﬁeld. it is straightforward to see that 1 2 2 Tµν = kµ kν (c2 1 + c2 + 3 c3 ) 2 (2. First of all we can ﬁnd that Fµν.47 It is easy to see that the energy-momentum tensor is traceless T = Tµµ = 0 (2. 2 eµλ eµλ = 1.31) (2) (A) eµν = ξν kµ + ηµ kν + c1 e(1) µν + c2 eµν + c3 eµν . This only means that it can not be used for the calculation of angular momentum of the ﬁelds (see paragraph 32 and paragraph 96 of [43].3. As it is also obvious from the ﬁnal expression it is not symmetric Tµν = Tνµ . eµλ eλµ = −1 eµλ eµλ = eµλ eλµ = 0. 2. Now we can calculate the contribution of the general solution (2. As expected no pure gauge ﬁelds appear in the expression .31) into the energy and momentum of the free gauge ﬁeld. i = 1. f or i.

ξ ] (2.2. .4. (2.. (2. one can read out the transformations induced on every fermion ﬁeld of every rank. The transformations are: δξ ψ = igLa ξ a ψ.4.4) . e) over the unit tangent vector eλ [33].4. e).4.48 (2. term by term.3). e) = U (ξ )Ψ(x.. δξ Aµ = ∂µ ξ − ig [Aµ .2) Inﬁnitesimally. eλs s! (2. the transformation which fermions undergo. δξ Ψ = igξ Ψ . e) = s=0 1 ψ (x)λ1 .4 Geometrical representation .13). ∞ Ψi (x.3) Expanding the gauge equation (2.4. 2.Fermions The fermion spinor-tensor ﬁelds ψλ1 . e) → Ψ (x. a δξ ψλ = igLa (ξ a ψλ + ξλ ψ ).e) the ﬁelds are transformed as: Ψ(x. An interesting remark is the factor of three which accompanies the energy contribution of the antisymmetric ﬁeld. a a a δξ ψλρ = igLa (ξ a ψλρ + ξλ ψρ + ξρ ψλ + ξλρ ψ ). ... As we stated before. this coeﬃcient owes its existence in the form of the T-propagator derived in the previous section (2.1) Under the extended gauge group with elements U (ξ ) = eigξ(x.λs eλ1 . is coupled to the transformation of the gauge ﬁelds.3.λs belong to the same representation of the compact Lie group G and are considered to appear in the expansion of the extended fermion ﬁeld Ψi (x.9).

4. e)) = δ (Ψ)iγµ Dν Ψ + Ψiγµ δ (Dν Ψ) = 0 (2.7) (2. δ (Dµ Ψ) = δ (∂µ − ig Aµ )Ψ = = ∂µ (δ Ψ) − ig (δ Aµ )Ψ − ig Aµ δ Ψ = igξ (∂µ − ig Aµ )Ψ = = igξ Dµ Ψ Hence. ¯ LF µν (x.49 The Dirac Lagrangian. Indeed.5) On expanding the extended Lagrangian density over eλ on can get the gauge .4. e) = Ψiγµ Dν Ψ remains invariant under such transformations. ¯ ¯ δ (LF µν (x.6) (2.4.

4..4. . . . . e) = s=0 ∞ 1 ¯ ψλ . .λ eλ . . .. . ⇒ Lµν. ∞ LF µν (x.ρr La eρ1 . × ψ 2 1 2 1 2 1 ×iγµ ∂ν − ig Aν + Aνρ eρ + Aνρ1 ρ2 eρ1 eρ2 + .50 invariant Lagrangian density of every rank.(2) 1 ¯ ¯ρ γµ (i∂ν + gAν )ψλ + ψ ¯λρ γµ (i∂ν + gAν )ψ + ψλ γµ (i∂ν + gAν )ψρ + ψ 2 ¯ µ (i∂ν + gAν )ψλρ + g ψ ¯λ γµ Aνρ ψ + g ψ ¯ρ γµ Aνλ ψ + +ψγ ¯ µ Aνλ ψρ + g ψγ ¯ µ Aνρ ψλ + g ψγ ¯ µ Aνλρ ψ +g ψγ (2. eρs r! × = q =0 = ¯+ψ ¯λ eλ + 1 ψ ¯λ λ eλ eλ + ..9) 2 .8) To get scalar Lagrangians the indices must be contracted in pairs. . . .(2) ¯ LI 3/2 = Lµµλλ = ψλ γµ (i∂µ + gAµ )ψλ + ψγµ (i∂µ + gAµ )ψλλ + ψλλ γµ (i∂µ + gAµ )ψ + 2 2 1 ¯ µ Aµλλ ψ ¯λ γµ Aµλ ψ + g ψγ ¯ µ Aµλ ψλ + g ψγ +g ψ (2..σ eσ . There are two ways of doing this: 1¯ 1¯ F. = 2 × = O(0) + O(1) + ¯ µ 1 (∂ν − igAν )ψλρ − ig (Aνλ ψρ + Aνρ ψλ ) − ig Aνλρ ψ + + ψiγ 2 2 2 1 ¯ ¯ + ψ λ iγµ (∂ν − igAν )ψρ + ψρ iγµ (∂ν − igAν )ψλ − 2 ig ¯ ¯ρ iγµ Aνλ ψ + 1 ψ ¯λρ iγµ (∂ν − igAν )ψ eλ eρ − ψλ iγµ Aνρ ψ + ψ 2 2 +O(3) + ... eλs s! 1 s 1 1 ψσ . . . 2 1 × ψ + ψσ eσ + ψσ1 σ2 eσ1 eσ2 + . eσq q! 1 q 1 ∞ × iγµ ∂ν − ig r=0 1 a Aνρ1 . .λρ = F.

µλ = F. so that f1 = f . the total fermionic Lagrangian for spin.(2) 1 2 ¯µ γµ (i∂λ + gAλ )ψλ + ψ ¯λ γµ (i∂λ + gAλ )ψµ + ψ ¯µλ γµ (i∂λ + gAλ )ψ + +ψγ ¯ µ (i∂λ + gAλ )ψµλ + +ψ ¯µ γµ Aλλ ψ + g ψ ¯λ γµ Aλµ ψ + g ψγ ¯ µ Aλµ ψλ + g ψγ ¯ µ Aλλ ψµ + + gψ ¯ µ Aλµλ ψ + g ψγ (2. in what follows we shall assume f2 = 0.3/2 fermions is a linear combination of the above two Lagrangians: II L3/2 = f1 LI 3/2 + f2 L3/2 (2.51 LII 3/2 = Lµλ.10) Hence.4.4. . f2 for reasons of simplicity.11) Since in the case of the fermionic ﬁelds there are no Bianchi identities that would determine the ratio f1 .

To check the consistency of the theory. the theory which describes the dynamics of the rank-2 tensor gauge bosons respects unitarity at the free level (g = 0) and predicts 3 physical propagating modes. the scattering amplitudes of both processes will be calculated with one external particle in longitudinal polarization. After this purely theoretical test. A straightforward way to examine gauge invariance and hence unitarity. As shown in the previous chapter. It is expected that the gauge invariance of the theory will force both the amplitudes to vanish. is to consider the full theory up to rank-2 tensor bosons and spin 3/2 fermion ﬁelds and study scattering processes. As a next step we confront the gauge invariance of the theory at the interacting level (g = 0). 45]. On this thesis two processes will be thoroughly studied. That of the production of two tensor gauge bosons of helicity 2 via the annihilation of 2 spin 1/2 fermions [44. we have checked the self-consistency of the lower rank tensor gauge theory in the absence of interactions.Chapter 3 Unitarity of the theory at the interactive level-Processes Up to now. the scattering cross sections for both processes will be calculated and compared with those of the 52 . and that of the production of the same particles via the annihilation of two Yang-Mills vector gauge bosons[46].

ν Gµλ.λ + Gµν Gµλ.1) • spin 3/2 fermion propagator: a.1.νλ 4 4 4 4 2 1 1 ¯ µ (i∂µ + gAµ )ψλλ + ψ ¯λλ γµ (i∂µ + gAµ )ψ + ¯λ γµ (i∂µ + gAµ )ψλ + ψγ L3/2 = ψ 2 2 ¯λ γµ Aµλ ψ + g ψγ ¯ µ Aµλ ψλ + 1 g ψγ ¯ µ Aµλλ ψ (3. 3.λ Gµν.1.λ Gµλ.1 Feynman rules The total Lagrangian which describes Yang-Mills vector bosons. more familiar QCD and Standard model results [47].ab (k ) = 1 iηµν δ ab .λλ + Gµν. 3/2 fermions and their interactions is the following: L = LY M + g2 L2 + f L3/2 . µ ¡ ¡ µ k k ν ˜ µν (k ) = iηµν . rank-2 tensor gauge bosons. = D F k (3.2) +g ψ 2 From the above Lagrangian we can derive the following Feynman rules. • spin 1/2 fermion propagator: (3.1.1. spin 1/2. b.3) ˜ µν .4) .ν + Gµν.53 corresponding. ν = D F f k (3. where ¯ ∂ Ψ + gΨ ¯ AΨ − 1 Ga Ga LY M = L1/2 + L1 = iΨ 4 µν µν 1 a a 1 a 1 a 1 a a 1 a a a L2 = − Ga µν.λ − Gµν Gµν.

µ = igγ µ ta (3.6) • Vertex: FFV ¡ a.λρ. µ k k = − µν . λρ = DF (k ) = 1 µλ νρ i ab µλ νρ µρ νλ µν λρ δ ( η η + η η − η η ) + (η η − η µρ η νλ ) g2 k 2 3 (3.1. ν = DF (k ) = − i µν ab η δ k2 (3. µν ¡ ¡ a.5) µν.ab b.54 • vector propagator: • 2nd rank tensor propagator: a.1.7) .1.ab b.

p.9) = ηµν (k − p)σ + ηνσ (p − q )µ + ησµ (q − k )ν (3. q ) = if gγ µ ta ν (3.11) • . p. µ b. q ) = gf abc Fµνσ (k.1.1. q ) (3. p. ρ ¡ ¡ ¡ a.1.55 • Vertex: F F2 T • Vertex: VVV • Vertex: VVVV a. µ c. σ q k p a.8) abc YM = Vµνσ (k. ν YM Fµνσ (k. ν abcd = Vµνρσ = −ig 2 f abe f cde (ηµρ ηνσ − ηµσ ηνρ ) + d.1. σ +f ace f bde (ηµν ηρσ − ηµσ ηρν ) + +f ade f bce (ηµν ηρσ − ηµρ ηνσ ) (3. µν c.10) b.

1.νρ. ρρ abcd = Vµ. p. q ) − Fµλ. p. ν . p. µ 56 abc TTV = Vµλ. q ) = T T V (2) (ηνλ ηµρ + ηνρ ηµλ )(k − p)σ + ηµν ησλ (k − p)ρ + ηµν ηρσ (k − p)λ + +(ηνλ ηρσ + ηνρ ηλσ )(p − q )µ + ηνσ ηµλ (p − q )ρ + ηνσ ηµρ (p − q )λ + +(ηρµ ησλ + ηρσ ηµλ )(q − k )ν + ηµσ ηνλ (q − k )ρ + ηνρ ηµσ (q − k )λ (3. q ) 2 YM Fµλ.νρ.λλ .νρ.νρ. p. q ) T T V (1) (3.σ (k.νρ.σ (k.15) • Vertex: VVTT ¡ d. p.σ (k. p.νρ.16) b.13) (3. λλ a. q ) = Fµλ. νρ (3.λλ .ν. q ) b.12) 1 T T V (2) T T V (1) TTV Fµλ.σ (k.1.νρ.σ (k.1.σ (k. µλ ¡ k p c.ν.Vertex: TTVc.ρρ = ig 2 g2 V V T T.abcd Fµ. p. p.1. q ) = gg2 f abc Fµλ. q ) = ηλρ Fµνσ (k.14) Fµλ.1.σ (k.ρρ 2 (3. σ q a.

17) .ν.ρρ = (f ace f bde + f ade f bce )(2ηλρ ηλ ρ − ηλρ ηλ ρ − ηλλ ηρρ )ηµν + + (f abe f cde − f ade f bce )(2ηνρ ηλ ρ − ηνρ ηρλ − ηνλ ηρρ )ηµλ − − (f abe f cde + f ace f bde )(2ηλν ηλ ρ − ηλρ ηνλ − ηνρ ηλλ )ηµρ + + (f ade f bce − f abe f cde )(ηµλ ηλρ + ηµρ ηλλ )ηνρ − − (f ace f bde + f ade f bce )(ηµλ ηνρ + ηµρ ηνλ )ηλρ + + (f abe f cde + f ace f bde )(ηµλ ηρρ + ηµρ ηρλ )ηνλ (3.1.λλ .57 V V T T.abcd Fµ.

2.1). the explicit form of which have been presented in the previous subsection. We will ﬁnd that the transition amplitude is gauge invariant.2. 3. The diﬀerence between these processes is in the actual expressions of the corresponding interaction vertices. The process receives contribution from three Feynman diagrams which are similar to the QED and QCD diagrams for the annihilation processes with two photons or two gluons in the ﬁnal state (see D.2 ﬀ → TT On this section the leading-order diﬀerential cross section of spin-two tensor gauge ¯ → T T will be calculated boson production in the fermion pair annihilation process f f [44] and the angular dependence of the polarized cross sections for each set of helicity orientations of initial and ﬁnal particles will be analyzed [45].the transition amplitude vanishes [44]. (3.58 3.36) and compare them with the corresponding cross sections for photons and gluons in QED and QCD (D. That is. This means that the unphysical .1). we make the following assignments: .1 The Feynman diagrams Working in the center-of-mass frame.and the other one in an unphysical . Having made this consistency check.longitudinal polarization . if we the one of the outgoing tensor gauge bosons is taken to be in a physical transverse polarization .3).longitudinal polarization states are not produced in the scattering process. we shall calculate the polarized cross sections for each set of helicity orientations of the initial and ﬁnal particles (3.2. Below we shall present the Feynman diagrams for the given process and calculate the transition amplitude. as well as with the W-pair production in Electroweak theory (D.35).

59

x

q1

p−

E ©

θ '

p+

z

q2

Figure 3.1: Incoming and outgoing momenta in the center-of-mass frame

µ pµ − = E (1, 0, 0, 1), p+ = E (1, 0, 0, −1), µ µ q1 = E (1, sin θ, 0, cos θ), q2 = E (1, − sin θ, 0, − cos θ),

(3.2.1)

**¯ and q1,2 the momenta of the tensor where p± are the momenta of the fermions f, f
**

2 2 2 gauge bosons T T . All particles are massless p2 − = p+ = q1 = q2 = 0. In the

center-of-mass frame the momenta satisfy the relations p+ = −p− , q2 = −q1 and E− = E+ = ω1 = ω2 = E . The invariant variables of the process are: s = (p+ + p− )2 = (q1 + q2 )2 = 2(p+ · p− ) = 2(q1 · q2 ) = 4E 2 s t = (p− − q1 )2 = (p+ − q2 )2 = − (1 − cos θ) 2 s 2 2 u = (p− − q2 ) = (p+ − q1 ) = − (1 + cos θ), 2 where s = (2E )2 and θ is the scattering angle.

(3.2.2)

60

It is convenient to write the diﬀerential cross section in the center-of-mass frame with dσ = 1 |M |2 dΦ, 4(p+ · p− ) (3.2.3)

where the ﬁnal-state density for two massless tensor gauge bosons is dΦ = so that dσ = 1 1 |M |2 dΩ. 2s 32π 2 (3.2.4) d3 k1 d3 k2 1 (2π )4 δ (p+ + p− − k1 − k2 ) = dΩ, 3 3 (2π ) 2ω1 (2π ) 2ω2 32π 2

We shall calculate the polarized cross sections for this reaction, to lowest order in α = g 2 /4π . The lowest-order Feynman diagrams contributing to fermion-antifermion annihilation into a pair of tensor gauge bosons are given below. In order g 2 , there are three diagrams. Dirac fermions ψ are conventionally drawn as thin solid lines, and Rarita-Schwinger spin-vector fermions ψ µ by double solid lines. These diagrams are similar to the QCD diagrams for fermion-antifermion annihilation into a pair of vector gauge bosons (see D.1). The diﬀerence between these processes is in the actual expressions for the corresponding interaction vertices (3.1.12).

a, µλ

q1

;ab iMµλ,νρ A

=

p ¯+

¡

b, νρ if η λρ b ν ∗ t γ u(p− )e∗ µλ (q1 )eνρ (q2 ) p− − q2

q2 = (ig )2 v ¯(p+ )γ µ ta p−

(3.2.5)

a, µλ, q1

;ab iMµλ,νρ = B

p ¯+

a, µλ

iMµν,ab = G

= −(ig )2 v ¯(p+ )g2 f bac tc γσ

¡ ¡

p− q1 q2 q3 p ¯+ p−

61

q2 , b, νρ if η λρ a µ ∗ t γ u(p− )e∗ µλ (q1 )eνρ (q2 ) p− − q1

=

(ig )2 v ¯(p+ )γ ν tb

(3.2.6)

b, νρ

=

1 νρ,µλ,σ ∗ F (−q2 , −q1 , q3 )u(p− )e∗ µλ (q1 )eνρ (q2 ) 2 TTV q3 (3.2.7)

**The probability amplitude of the process can be written in the form
**

∗ Mµλ,νρ e∗ µλ (q1 )eνρ (q2 ) =

(ig )2 v ¯(p+ ) γ µ ta

f η λρ b ν f η λρ a µ t γ + γ ν tb t γ + p− − q 2 p− − q1 1 νρ,µλ,σ ∗ + ig2 f bac tc γσ 2 FT (−q2 , −q1 , q3 ) u(p− )e∗ µλ (q1 )eνρ (q2 ), q3 T V (3.2.8)

**where u(p− ) is the wave function of spin 1/2 fermion, v (p+ ) of antifermion, the ﬁnal
**

∗ tensor gauge bosons wave functions are e∗ µλ (q1 ) and eνρ (q2 ) and q3 = p− + p+ . The

Dirac and symmetry group indices are not shown.

62 3. We take the rank-2 tensor e∗ νρ (q2 ) in longitudinal polarization.σ (−q2 .9) Taking into account that the tensors representing rank-2 tensor gauge bosons must be traceless (η µν eµν (q ) = 0). The total scattering amplitude should cancel when one of the outgoing gauge bosons is taken longitudinal (eL ) and the other in transversal polarization (eT ).2. Taking again the polarization ∗ tensors e∗ νρ (q2 ) to be longitudinal. 4 ρσ (3. we can check unitarity at the interacting level imitating the standard procedure followed in the Yang-Mills case (Appendix D. Indeed.σ (−q2 .2. −q1 .8). −q1 . q2 eµλ (q2 ) = q1 eµλ (q1 ) = 0.8).2.2. In the center of-mass-frame the following relations should ∗ hold.2.2 Gauge Invariance At this point. using relations (3. 2 2 4 Then the last term becomes: ig2 f abc v ¯(p+ )tc γ σ u(p− ) 1 ∗ e ( q1 ) ξ ρ . q3 )(q 2ν ξ ρ + q 2ρ ξ ν )eµλ (q1 ) = −q3 ξ eρσ (q1 ) T T V (1) 3 2 ρ ∗ T T V (2) ξ eρσ (q1 ) Fνρ. and the tensor eµλ (q1 ) in transversal polarization: µ µ λ λ q1 eµλ (q1 ) = q1 eµλ (q1 ) = 0.10) Now let us consider the ﬁrst two terms in (3. q3 )(q 2ν ξ ρ + q 2ρ ξ ν )eµλ (q1 ) = − q3 2 2 q3 1 3 TTV 2ν ρ 2ρ ν µλ 2 ρ ∗ = − ξ ρ e∗ Fνρ.µλ.µλ. e∗ νρ (q2 ) = q2ν ξρ + q2ρ ξν . −q1 .2.states are not produced in the scattering process.9) for the wave function eµλ (q1 ) . let us consider the last term in (3. This is a consequence of the generalized Ward identity and expresses the fact that the unphysical . q3 )(q ξ + q ξ )e (q1 ) = −q3 ξ eρσ 1 − · ρσ (q1 ).σ (−q2 .µλ. (3.longitudinal polarization .1). we get: 2 ρ ∗ Fνρ.

3 Squared Matrix Element Using the explicit form of the vertex operator F νρ. (3.12) and the orthogonality properties of the tensor gauge boson wave functions: µ µ λ λ q1 eµλ (q1 ) = q1 eµλ (q1 ) = q2 eµλ (q1 ) = q2 eµλ (q1 ) = 0.63 and the Dirac equation for the wave-functions u(p− ) and v ¯(p+ ) we shall get: ρ fv ¯(p+ ){−ta tb γ µ + tb ta γ µ }u(p− )e∗ µρ (q1 )ξ = ρ = −if f abc v ¯(p+ )tc γ σ u(p− )e∗ ρσ (q1 )ξ .2.2. 2 q3 2 (3. µ µ λ λ q2 eµλ (q2 ) = q2 eµλ (q2 ) = q1 eµλ (q2 ) = q1 eµλ (q2 ) = 0.2. p− − q 2 tb γ ν + γ ν tb 1 λρ η 4 p− − q1 ta γ µ + +if abc c (q2 − q1 )σ µν λρ 1 µρ νλ ∗ (η η − η η ) u(p− )e∗ t γσ µλ (q1 )eνρ (q2 ).2.2.11) This term precisely cancels the contribution coming from the last term of the amplitude (3.2.13) where the last relations follow from the fact that q1 we shall get ∗ 2 Mµλνρ.1. (3.2. 3.ab e∗ ¯(p+ ) γ µ ta µλ (q1 )eνρ (q2 ) = (ig ) g2 v 1 λρ η 4 q2 in the process of Fig.D.12) Thus the cross term matrix element between transverse and longitudinal polarizations vanishes: MeT eL = 0. Our intention now is to calculate the physical matrix elements MeT eT for each set of helicity orientations of initial and ﬁnal particles.µλ.10) if we demand: f= g2 4 (3.14) .σ (3.2.

µλ. q3 ) u(p− ) × 2 TTV q3 1 λρ 1 λρ 2 ν b 4g a µ µ a 4g × (−ig ) u ¯(p− ) γ t t γ +γ t tb γ ν + p − − q2 p− − q1 1 νρ.µ λ . −q1 .σ − if abc tc γσ 2 FT (−q2 . q3 ) v (p+ ) q3 T V For unpolarized fermions-antifermion scattering the average over the fermion and antifermion spins is deﬁned as follows: |M|2 = Using the completeness relations us (p− )¯ us (p− ) = p− .64 The complex conjugate of the scattering amplitude (3. −q1 .µλ. spin 1/2 v s (p+ )¯ v s (p+ ) = p+ . q3 T V b 1 λρ η 4 a µ µ a 1 λρ g 4 . s s 11 22 |M |2 . −q1 . −q1 .8) is M∗µλνρ = (−ig )2 g2 u ¯ ( p− ) tb γ ν + p− − q2 p− − q1 1 νρ.σ +if abc tc γσ 2 FT (−q2 . −q1 . q3 ) .σ F (−q2 . q3 ) × q3 T V = × p− γ t + if a b c ν t γ +γ t tb γ ν + p− − q2 p− − q 1 1 ν ρ . q3 ) v (p+ ) q3 T V γ ν tb 1 λρ g 4 ta γ µ + γ µ ta 1 λρ g 4 and we can calculate now the squared matrix elements in the form Mµλνρ M∗µ λ ν ρ = (ig )2 v ¯(p+ ) γ µ ta − if abc tc γσ 1 λρ η 4 p− − q2 tb γ ν + γ ν tb 1 λρ η 4 p− − k1 ta γ µ − 1 νρ. and averaging over spins of the fermions we shall get: Mµλνρ M∗µ λ ν ρ 1 λρ 1 λρ η η g4 T r p+ γ µ ta 4 tb γ ν + γ ν tb 4 ta γ µ − 4 p− − q2 p− − q1 1 νρ.2.µλ.µλ.σ tc γσ 2 FT (−q2 .σ +if abc tc γσ 2 FT (−q2 .

q1 q ˜1 q1 q ˜1 µ µ where q1 = (ω1 . As a next step we shall calculate the sum over transversal tensor gauge bosons polarizations. The sum over transversal polarizations of the helicity-two tensor gauge boson is [48. The explicit form of the transversal polar- ization tensors.ab M∗µ λ ν = ρ .65 Contracting the last expression with the transversal on-shell polarization tensors of ∗ the ﬁnal tensor gauge bosons e∗ µα (k1 ) and eνβ (k2 ) we get the probability amplitude in the form: Mµλνρ.a b ∗ e∗ µλ (q1 )eνρ (q2 )eµ λ (q1 )eν ρ (q2 ) = 1 λρ 1 λρ η η g4 T r p+ γ µ ta 4 tb γ ν + γ ν tb 4 ta γ µ + 4 p− − q 2 p − − q1 1 1 + if abc tc γσ 2 (q2 − q1 )σ (η µν η λρ − η µρ g νλ ) × q3 2 p− γ t ν b 1 λρ η 4 p − − q2 t γ a µ +γ t µ a 1 λρ η 4 p− − q 1 tb γ ν − × − if a b c tc γσ 1 1 (q − q1 )σ (η µ ν η λ ρ − η µ ρ η ν λ ) 2 2 q3 2 ∗ × e∗ µλ (q1 )eνρ (q2 )eµ λ (q1 )eν ρ (q2 ). 31. is given by the . q1 ) and q ˜1 = (ω1 . 44] r r e∗ µλ (q1 )eµ λ (q1 ) = r q1µ q ˜1µ + q ˜1µ q1µ 1 q1λ q ˜1λ + q ˜1λ q1λ (−ηµµ + )(−ηλλ + )+ 2 q1 q ˜1 q1 q ˜1 +(−ηµλ + q1µ q ˜1µ + q ˜1λ q1µ q1µ q ˜1λ + q ˜1µ q1λ )(−ηλµ + )− q1 q ˜1 q1 q ˜1 ˜1µ q1λ q1µ q ˜1λ + q q1µ q ˜1λ + q ˜1µ q1λ −(−ηµλ + )(−ηµ λ + ) . when the momentum is aligned along the third axis. −q1 ).

0 1 2 √ . 0. 0. 0 From the kinematics of the process in Fig. 36]: e1 µλ 0. e = µλ 2 0. 0.66 matrices [48. 0 =√ 2 0. 31. 1. 0 . 0. 0. 0 0. 0. 0. 0.D. 1. d(r) d(r) 2 2 In the next section we shall evaluate these traces and sum over the polarizations. .a b ∗ e∗ µλ (q1 )eνρ (q2 ) eµ λ (q1 )eν ρ (q2 ) = (3. 2 where Eµµ = −ηµµ + q1µ q2µ + q2µ q1µ . 0. q ˜2µ = q1µ and the average over polarizations can be rewritten as 1 e∗ µλ (q1 )eµ λ (q1 ) = (Eµµ Eλλ + Eµλ Eλµ − Eµλ Eµ λ ). 1 0.15) Thus the average over tensor gauge boson polarizations gives Mµλνρ.ab M∗µ λ ν = ρ . 0. 0.2 it follows that ω2 = ω1 and k2 = −k1 therefore q ˜1µ = q2µ . 0 0.2.2. 1.2. 0 0. q1 · q2 (3.16) 1 λρ 1 λρ η η g4 T r p+ γ µ ta 4 tb γ ν + γ ν tb 4 ta γ µ + 4 p− − q 2 p − − q1 1 1 + if abc tc 2 ( q2 − q1 )(η µν η λρ − η µρ η νλ ) × q3 2 p− γ t t γ +γ t tb γ ν − p − − q2 p− − q 1 1 1 − if a b c tc 2 ( q2 − q1 )(η µ ν η λ ρ − η µ ρ η ν λ ) q3 2 ν b 1 λρ η 4 a µ µ a 1 λρ g 4 × 1 δ aa δ bb 1 (Eµµ Eλλ + Eµλ Eλµ − Eµλ Eµ λ ) (Eνν Eρρ + Eνρ Eρν − Eνρ Eν ρ ). 0. 0 0. −1.

2 2 = We can calculate the traces over the symmetry group indices using formulas from Appendix B: tr(f abc f abc tc tc ) = d(r)C2 (r)C2 (G) = d(G)C (r)C2 (G) = N (N 2 − 1) 2 .2.(3. The ﬁrst contribution can be evaluated in the following way: (GG∗ )µλνρ . We shall use convenient notations for the diﬀerent terms in the amplitude.4 Evaluation of Traces In order to evaluate the squared matrix element in the last expression (3.2. The whole amplitude will be expressed as a symbolic sum of three terms: M = A + B + G.5).2.2.16) we have to calculate the traces and then perform the summation over polarizations.7) so that the squared amplitude (3. (3.2.67 3. exactly corresponding to the three Feynman diagrams in (3. µλν ρ g4 1 1 T r{ p+ if abc tc 2 (η µν η λρ − η µρ η νλ )( q2 − q1 ) 2 4d (r) q3 2 1 1 p− (−i)f a b c tc 2 (η µ ν η λ ρ − η µ ρ η ν λ )( q2 − q1 )}δ aa δ bb = q3 2 4 T r{ p+ ( q2 − q1 ) p− ( q2 − q1 )} g tr(f abc f abc tc tc ) = 2 4d (r) (2q1 · q2 )(2q1 · q2 ) 1 1 (η µν η λρ − η µρ η νλ )(η µ ν η λ ρ − η µ ρ η ν λ ).2.16) shall have nine terms MM∗ = (A + B + G)(A + B + G)∗ .6).

ηλ ρ − 1 2 Using again expression (3.15) and scalar products (3. GG∗ = g4 d(r)C2 (r)C2 (G) sin2 θ.2.15). 4d (r) (−2p− q2 )(2q1 q2 ) 4 2 ∗ µλνρ .2) we can sum over the polarizations of tensor gauge bosons: BG∗ = 1 g4 d(r)C2 (r)C2 (G)(− sin2 θ).2.2. 2 Now it is easy to calculate summation over tensor gauge boson polarizations using expression (3.16) can be evaluated as follows: 1 λρ g4 µ a 4η = 2 T r { p+ [γ t (BG ) tb γ ν ] 4d (r) p− − q2 1 1 p− [−if a b c tc 2 (g µ ν η λ ρ − η µ ρ η ν λ )( q2 − q1 )]}δ aa δ bb = q3 2 4 T r{ p+ γ µ ( p− − q2 )γ ν p− ( q2 − q1 )} 1 λρ µ ν λ ρ g 1 η (η η = −i 2 tr(f abc ta tb tc ) − η µ ρ η ν λ ).2): p− ·(q2 −q1 )+p+ ·p− q1 ·q2 g4 d(r)C2 (r)C2 (G)8 p+ ·(q2 −q1 ) (2 4d2 (r) q1 q2 )(2q1 q2 ) 1 µρ νλ 1 µρ ν λ (η µν η λρ − 2 η η )(η µ ν η λ ρ − 2 η η )= = g4 d(r)C2 (r)C2 (G)2 sin2 4d2 (r) 1 µρ ν λ θ(η µν η λρ − 1 η µρ η νλ )(η µ ν η λ ρ − 2 η η ).18) .17) The next contribution in (3.C and the corresponding scalar products (3. 4d2 (r) (3.68 and then the traces of gamma matrices using relations from App.2.2. µ λ ν ρ and then using traces from Appendices B and C.2. 2 4d (r) 4 (3. we shall get: g 4 d(r)C2 (r)C2 (G) 4{η µν [p+ 4d2 (r) 2 · p− q1 · q2 + p+ · (q2 − q1 ) p− · q2 + p+ · q2 p− · (q2 − q1 )] + µ ν µ ν µ ν ν +q1 · q2 (pµ + p− − p− p+ ) + 2p− · (q2 − q1 )p+ p− + 2p+ · (q2 − q1 )p− p− } 1 λρ µ ν 1 η (η (−2p− q2 )(2q1 q2 ) 4 η µ ρ g ν λ ).2.

19) As one can get convinced.20) GA∗ = The sixth contribution is (BB ∗ )µλνρ = .69 The third contribution is 1 λρ g4 ν b 4η T r { p [ γ t ta γ µ ] + 4d2 (r) p− − q1 1 1 p− [−if a b c tc 2 (η µ ν η λ ρ − η µ ρ g ν λ )( q2 − q1 )]}δ aa δ bb = q3 2 4 g T r{ p+ γ ν ( p− − q1 )γ µ p− ( q2 − q1 )} 1 αβ µ ν λ ρ 1 = −i 2 tr(f abc tb ta tc ) η (η η − ηµ ρ ην λ ) 4d (r) (−2p− q1 )(2q1 q2 ) 4 2 (AG∗ )µλνρ . 2 4d (r) 4 (3.2. 2 4d (r) 4 g4 1 d(r)C2 (r)C2 (G)(− sin2 θ). the next two terms GB ∗ and GA∗ give similar contributions: GB ∗ = 1 g4 d(r)C2 (r)C2 (G)(− sin2 θ). µλν ρ (3.2.21) = 1 λρ 1 λρ g4 µ a 4η b ν ν b 4η T r p [ γ t t γ ] p [ γ t ta γ µ ] δ aa δ bb = + − 4d2 (r) p− − q2 p− − q 2 µ ν ν µ g4 a b b a T r { p+ γ ( p− − q2 )γ p− γ ( p− − q2 )γ } 1 λρ 1 λ ρ tr ( t t t t ) η η 4d2 (r) (2p− · q2 )(2p− · q2 ) 4 4 . µλν ρ = and can be evaluated in the similar way: g4 d(r)C2 (r)C2 (G) − 4× 2 4d (r) 2 × η νµ [−p+ · p− q1 · q2 + p+ · (q2 − q1 ) p− · q1 + p+ · q1 p− · (q2 − q1 )] + + × µ ν µ ν µ ν µ q1 · q2 (pν − p+ − p+ p− ) + 2p− · (q2 − q1 )p+ p− + 2p+ · (q2 − q1 )p− p− 1 λρ µ ν λ ρ 1 1 η (η η − η µ ρ η ν λ ).2. 4d (r) 4 ∗ (3. (−2p− q1 )(2q1 q2 ) 4 2 so that after summation over polarization we shall get: g4 1 (AG ) = 2 d(r)C2 (r)C2 (G)(− sin2 θ).

23) = 1 λρ η g4 T r{ p+ [γ µ tap4 tb γ ν ] 4d2 (r) q2 −− p− [γ µ ta 1 λ ρ η 4 tb p− −q1 γ ν ]}δ aa δ bb = = µ −q2 )γ νp− γ µ (p− −q1 )γ ν } 1 λρ 1 λ ρ g4 tr(ta tb ta tb ) T r{p+ γ (p− η 4η 4d2 (r) (2p− q2 )(2p− q1 ) 4 . 2 4d (r) 4 (3. µλν ρ g4 1 d(r)C2 (r)C2 (r) sin2 θ.22) The seventh contribution is identical with the sixth one and gives AA∗ = The eighth contribution is (BA∗ )µλνρ . (2p− q2 )(2p− q2 ) 4 4 g4 1 d(r)C2 (r)C2 (r) sin2 θ. 2 4d (r) 4 (3.2. It can be performed using the expression presented in C: g4 µ µ µ µ µµ ν ]+ d(r)C2 (r)C2 (r) 16pν − p− [p− p+ + p+ p− − p+ · p− η 2 4d (r) µ µ µ µν µν − ( p+ · q2 pµ + +8pν − − p− · q2 p+ )η − (p+ · q2 p− + p− · q2 p+ )η ν µµ (p+ · q2 pν − − p− · q2 p+ ) η µ µ µ µν µν +8pν − (p+ · q2 pµ + − − p− · q2 p+ ) η − (p+ · q2 p− + p− · q2 p+ )η ν µµ (p+ · q2 pν − − p− · q2 p+ ) η +8p+ · q2 p− · q2 [η µν η µ ν − η µν η µ ν + η µµ η νν ] After summation over the polarizations we shall get BB ∗ = 1 1 λρ 1 λ ρ η η .70 and involves trace of eight gamma matrices.2.

(3.2.2. Putting together all pieces of the squared matrix element (3.2.2. (3.19).24).16).2. we ﬁnally .24) and for the ninth contribution we shall get identically AB ∗ = 1 1 g4 d(r)C2 (r)(C2 (r) − C2 (G)) (− sin2 θ).2.2. (3.22).18).25) We are now in a position to calculate the total contribution to the squared matrix element (3. (3.2.20).25).71 and gives g4 d(r)C2 (r)(C2 (r) 4d2 (r) −1 C (G)) −4q1 · q2 p+ · p− η µν η νµ + 4p+ · q2 p− · q1 η µν η νµ + 2 2 +4p+ · q1 p− · q2 η µν η νµ − 4q1 · q2 p+ · p− η µν η µ ν + +4p+ · q2 p− · q1 η µν η µ ν + 4p+ · q1 p− · q2 η µν η µ ν + +4q1 · q2 p+ · p− η µµ η νν − 4p+ · q2 p− · q1 η µµ η νν − µ µ µν ν µ p+ p− + 4q1 · q2 η νµ pν −4p+ · q1 p− · q2 η µµ η νν − 4q1 · q2 η νν pµ + p− + 4 q1 · q2 η + p− − µ µ µν ν µ νν µ µ p + p− − p+ p− + 4q1 · q2 η µν pν −4q1 · q2 η νν pµ + p− + 8 p− · q2 η + p− − 8p− · q2 η µ ν νν µ µ µν ν µ p− p− − 4q1 · q2 η µ ν pµ −4q1 · q2 η µν pν + p− + + p− + 8p− · q2 η p+ p− − 8p+ · q2 η ν µµ ν ν ν µν µ ν p+ p− + +8p− · q1 η µ ν pµ p+ p− − 8p− · q1 η µν pµ + p− − 4q1 · q2 η + p − + 4 q1 · q2 η ν µν µ ν ν µν µ ν p− p− − +8p− · q1 η µµ pν p− p− + 8p+ · q1 η µν pµ − p− + 8 p + · q 2 η + p− + 8 p+ · q1 η ν νµ µ ν ν ν µ µ ν p+ p− + 4q1 · q2 η µν pµ −16p+ · p− η µν pµ + p− − − p− + 16p+ p− p− p− + 4q1 · q2 η µ ν ν µµ ν ν ν µν µ ν p− p− + 16pµ −4q1 · q2 η µµ pν + p− p− p− + p− + 8p+ · q2 η p− p− − 8p+ · q1 η 1 λρ 1 λ ρ 1 η 4η (2p− q2 )(2p− q1 ) 4 . After summation over polarizations we shall get BA∗ = g4 1 1 d(r)C2 (r)(C2 (r) − C2 (G)) (− sin2 θ) 2 4d (r) 2 4 (3. (3. (3. 2 4d (r) 2 4 (3.17).2.21). (3. (3.2.2.23).2.

2. 4π (3.2.3.2.2. s 64d(r) C2 (r)C2 (G) 64d(r) (3.26). as we discussed in the introduction.28) where for the SU (N ) group we have = (N 2 −1) . s 64d(r) g2 .26) 3.6 Helicity Structure As the next step we shall calculate the above matrix element in the helicity basis for initial fermions and ﬁnal tensor gauge bosons. This calculation of polarized cross . receives contribution from three Feynman diagrams shown in Fig. Plugging everything into our general cross-section formula (3. This process.2. 128N 3.1) yields the diﬀerential cross section in the center-of-mass frame: dσ = 1 1 1 g4 d(r)C2 (r) C2 (G) sin2 θ dΩ = 2 4d (r) 4 2s 32π 2 g 2 1 C2 (r)C2 (G) sin2 θ dΩ = = ( )2 4π s 64d(r) α2 C2 (r)C2 (G) = sin2 θ dΩ.2. 2 4d (r) 4 (3.?? and for the unpolarized fermion pairs the squared matrix element was presented above (3.27) where α= Thus the unpolarized cross section is dσ = α2 C2 (r)C2 (G) sin2 θ dΩ.72 obtain MM∗ = 1 g4 d(r)C2 (r) C2 (G) sin2 θ.5 Cross Section We can calculate now the leading-order diﬀerential cross section for the tensor gauge bosons production in the annihilation process.

29) 0 −1 .2.13).30) = 0. µν L ( q2 ) = µν R (q1 ). −E sin θ. E sin θ.2.30) are orthonormal ∗µλ R (q1 ) L (q1 )λν (3. and the tensor gauge boson wave functions for circular polarizations along the q1 direction are: 0 0 0 0 0 2 cos θ i cos θ − cos θ sin θ 1 µλ ( q ) = . E cos θ) and q2 = (E. µ µ where q1 = (E.2. 0. 0 √ 1 µ uL (p− ) = 2E . 0 0 0 √ 0 µ vL (p+ ) = 2E (3. The right. 0. −E cos θ). ∗µλ L (q1 ) L (q1 )µλ =1 and fulﬁl the equations (3. .73 sections is very similar to the one in QED [?]. 1 R 2 0 i cos θ − 1 − i sin θ 2 0 − cos θ sin θ −i sin θ sin θ 0 0 0 0 0 2 cos θ − i cos θ − cos θ sin θ 1 µλ ( q ) = . ∗µλ R (q1 ) R (q1 )µλ =1 . 1 L 2 0 − i cos θ − 1 i sin θ 2 0 − cos θ sin θ i sin θ sin θ It is easy to check that the wave functions (3. The helicity states for the second gauge boson are: µν R ( q2 ) = µν L (q1 ) .and left-handed spinors wave functions are: 0 √ 0 µ 2E uR (p− ) = 1 0 1 √ 0 µ vR (p+ ) = 2E 0 0 .2.

30) into the matrix element (3.29) and (3. (3.74 Now we can calculate all sixteen matrix elements between states of deﬁnite helic¯ ities. (3. 1 2 − µλ νρ 2 q3 2 2 so that all together they will give Mµλ. Similarly.2. using the helicity wave functions (3.2.32) 16d2 (r) 16 d(r) where the invariant operator C2 is deﬁned by the equation ta tb = C2 .33) 16d2 (r) 16 d(r) .14) we can ﬁnd the ﬁrst term (ig )2 g2 v ¯L (p+ ) {γ µ ta then the second one (ig )2 g2 v ¯L (p+ ) {γ ν tb 1 λρ η 4 R ∗R ta γ µ } uR (p− ) e∗ µλ (q1 )eνρ (q2 ) = 1 λρ η 4 R ∗R tb γ ν } uR (p− ) e∗ µλ (q1 )eνρ (q2 ) = − p− − q2 g2 g 2 a b t t sin θ. we must square the matrix element (3. Let us start with fR f L → TR TR .νρ RL ∗R ∗R µλ (q1 ) νρ (q2 ) = (ig )2 g2 4 [ta . By plugging the explicit expressions for the helicity wave functions (3.2.3.2.14) for these particular helicities Mµλ.νρ RL ∗R ∗R µλ (q1 ) νρ (q2 ) contains three terms. This gives |M|2 ¯ →RR = RL g4 g 4 C2 (r)C2 (G) 2 abc abd c d 2 tr ( f f t t ) sin θ = sin θ. tb ] − 2if abc tc sin θ = ig2 g 2 abc c f t 4 sin (3.8). 4 p− − q 1 g2 g 2 b a t t sin θ 4 and ﬁnally the third one (ig )2 g2 v ¯L (p+ ){if abc tc ig2 g 2 abc c ( q2 − q1 ) µν λρ 1 µρ νλ R ∗R ∗R ( η η − η η ) } u ( p ) e ( q ) e ( q ) = f t sin θ. The scattering amplitude (3.29).2.2. This gives |M|2 ¯ →LL = RL g4 g 4 C2 (r)C2 (G) abc abd c d 2 tr ( f f t t ) sin θ = sin2 θ.2. To compute the cross section.31) and then average over the symmetries of the initial fermions and sum over the symmetries of the ﬁnal tensor gauge bosons. (3.31) θ.2.2. we can calculate ¯ the amplitude fR f L → TL TL .

Indeed. fR fL ¯ → TL TL . The e+ e− → γγ annihilation cross section .2.34) ∗R ∗L µλ (q1 ) νρ (q2 ) .2. let us compare this result with the electron-positron annihilation into two transversal photons.2. s 64d(r) (3. Thus only four amplitudes out From this analysis it follows that the total spin angular momentum of the ﬁnal state is one unit less than that of the initial state. fL fR ¯ → TR TR .2.27) yields: dσfR fL = ¯ →TR TR g 4 C2 (r) C2 (G) 2 1 1 sin θ dΩ = 16 d(r) 2s 32π 2 α2 C2 (r)C2 (G) = sin2 θ dΩ. 128N Adding up all sixteen amplitudes and dividing by four.2. (3. therefore a unit of spin angular momentum is converted to the orbital angular momentum and the ﬁnal state is a P-wave. we recover the unpolarized cross section 3.32) into our general cross-section formula in the center-of-mass frame (3.η λρ of sixteen are nonzero: fR fL ¯ → TR TR . to average over the initial particle spins.36) dσfR fL ¯ →TR TR = dσfR fL ¯ →TL TL = dσfL fR ¯ →TR TR = dσfL fR ¯ →TL TL . We can calculate now the leading-order polarized cross sections for the tensor gauge boson production in the annihilation process.28.75 ¯ The amplitude fR f L → TR TL vanishes because the common factor to all three pieces of this amplitude .2. 4π For the rest of the helicities we shall get (3. fL fR ¯ → TL TL .35) where α = g2 . Plugging the matrix elements (3.is equal to zero. where for the SU (N ) group we have C2 (r)C2 (G) 64d(r) = (N 2 −1) . This cross section should be compared with the analogous annihilation cross sections in QED and QCD.

36) shows dramatically diﬀerent behaviour . .76 [50] in the high-energy limit is dσγγ = α2 1 + cos2 θ dΩ s sin2 θ (3. The quark pair annihilation cross section into two transversal gluons q q ¯ → gg in the leading order of the strong coupling αs is dσgg = 2 αs C2 (r)C2 (r) 1 + cos2 θ C2 (G) [ (1 + cos2 θ)]dΩ − 2 s d(r) 4 C ( r ) sin θ 2 (3. The production cross section of spin-two gauge bosons (3. (3. 36]. The high energy production of longitudinal gauge bosons is [53] + − = dσe+ e− →W0 W0 1 + sin4 θw α2 [ ] sin2 θ dΩ.sin2 θ .with its maximum at θ = π/2 and decrease for small angles. One can only speculate that at high enough energies. we may observe the standard spin-one gauge bosons together with new spin-two gauge bosons [31.2.2.38) and also has a minimum at θ = π/2 and increases for small scattering angles [52].37) except very small angles of order me /E .35).2.2. s 256 sin4 θw cos4 θw (3.39) where cos θw = mw mz and it is similar to the spin-two transversal gauge boson production (3. To predict the threshold energy one should ﬁrst construct a massive theory and even in that case the corresponding Yukawa couplings most probably will be unknown. The cross section has a minimum at θ = π/2 and then increases for small angles [51]. It is also instructive to compare this result with the angular dependence of the W-pair production in Electroweak theory.35).2.2. may be at LHS energies and above the threshold.

1) .3. Working in the center-of-mass frame.2.13).2.2 are momenta of the vector bosons V + V and q1. and µ q1 = E (1. 0. − cos θ).27). where k1. cos θ) . 3. sin θ. (3. Next. µ k2 = E (1. Then we shall calculate the polarized cross sections for each set of helicity orientations of the initial and ﬁnal particles (see formulas (3.12). (D.31)).3. we make the following assignments: µ k1 = E (1. (3.3. 0. −sinθ.3.3. (D.26).3. we will check the on mass-shell gauge invariance of the scattering amplitude taking one of the outgoing tensor bosons in longitudinal polarization.1.3. Below we shall present the Feynman diagrams for the given process and the expressions for the corresponding transition amplitudes. −1).77 3. 0.30)) and compare them with the corresponding cross section of the vector gauge bosons V + V → V + V in Yang-Mills theory (see formulas (D. The process is illustrated in Fig. 0. µ q2 = E (1.14) and (3.1 Feynman diagrams The scattering process is illustrated in Fig. 0.3.3.3 VV-TT Our intention in this subsection is to calculate the leading-order diﬀerential cross section of spin-two tensor gauge boson production by a pair of vector gauge bosons in the process V + V → T + T and to analyze the angular dependence of the polarized cross sections for each set of helicity orientations of initial and ﬁnal particles [46].2.1 and receives contribution from four Feynman diagrams.3. 0.2 momenta of the tensor .28) and (3.

3.1) We shall calculate the polarized cross sections for the reaction V + V → T + T . 2 where s = (2E )2 and θ is the scattering angle.2: Incoming and outgoing momenta in the center-of-mass frame 2 2 2 2 = 0.78 x q1 k1 E © θ ' k2 z q2 Figure 3. All particles are massless k1 center-of-mass frame the momenta satisfy the relations k1 = −k2 . q2 = −q1 . The invariant variables of the process are: s = 2(k1 · k2 ). The lowest-order Feynman diagrams contributing to the process of the annihilation of a pair of vector bosons to a pair of tensor gauge . In the = q2 = q1 = k2 gauge bosons T + T . 32π 2 (3. 2s 32π 2 1 dΩ. It is convenient to write the diﬀerential cross section in the center-of-mass frame with the tensor bosons produced into the solid angle dΩ as dσ = where the ﬁnal-state density is dΦ = 1 1 |M|2 dΩ. s t = − (1 − cos θ). to the lowest order in α = g 2 /4π . 2 s u = − (1 + cos θ).

ρρ q1 e. µ eda = Vσσ ρρ ¡ ¡ λλ µ (−p. µ p q2 l.el (p) Vρρ τ τ ν (−q2 . −q2 . k1 ) ∆σσ (p ) Vλλ el (3.cd = I eca = Vσσ For the second diagram the amplitude is: c. For the ﬁrst diagram we shall get: c. there are four diagrams.3) . k2 ) eµ e∗ q2 . ν τ τ ν (−q1 . p ν ∗λλ ρρ . σσ k1 a. In order g2 g 2 . k1 ek2 eq1 µ (−p ττ clb . τ τ = k2 b.79 bosons are shown below. τ τ = k2 b. ν iMab. Vector gauge bosons V are conventionally drawn as thin wave lines and tensor gauge bosons T as double wave lines. λλ .3.3. d.2) p l. σσ k1 a. k1 ) σσ τ τ dlb DF. ρρ = iMab. q1 q2 . k2 ) ν ∗λλ ρρ eµ e∗ q2 .cd II e. λλ d. The probability amplitude of the process can be written as a sum of four terms corresponding to each diagram. p. −q1 . k1 ek2 eq1 (3.

−q1 .6) Our intention is to calculate the physical matrix elements in the helicity basis for initial vector and ﬁnal tensor gauge bosons.3.For the third diagram we get: c. k1 . −q2 .3.3.cd IV = a.4) = µ abdc ν ∗λλ ρρ Vµνρρ e∗ λλ (k1 . (3. This calculation of polarized cross . ν (3. d. p ) ek1 ek2 eq1 2 Finally for the forth diagram we have: c. λλ q2 l. ρρ = b. µ ν where eµ k1 is the wave function of the ﬁrst vector boson and ek2 of the second. ρρ q1 q2 iMab. µ µ eab στ dcl ν ∗λλ ∗ρρ = Vσµν (−p . b.cd = III p a. The λλ ρρ ﬁnal tensor gauge bosons wave functions are e∗ and e∗ q1 q2 . el (p ) Vρρ λλ τ (−q2 . ¡ ¡ k1 e. ν (3. τ q1 iMab. −q1 ) ek1 ek2 eq1 q2 . λλ d. k2 ) DF eq . σ k2 k1 k2 80 .5) The total amplitude is a sum of four terms: iM = iMI + iMII + iMIII + iMIV . k2 .

eR (q )eR (q )µν = 1. (3. i. i. 1.2 Gauge Invariance To check on mass-shell gauge invariance of the amplitude (3. − sin θ ) 1 eµν L (q1 ) = 2 (0. − sin θ).9) Before we proceed to the calculation of the physical matrix elements of the amplitude. 0) 2 . − cos θ.and left-handed vector wave functions are: eR (k1 )µ = 1 √ (0.3. − cos θ. we must ﬁrst guarantee the on mass shell gauge invariance of it. The right. i.3. i. i.8) are orthonormal: ∗µν ∗µν µν e∗ R (q )eL (q )µν = 0. (3. (3. 1.6) let one of the tensor gauge boson wave function be longitudinal: ρ ρ ρ ρ eρρ q2 = q2 ξ + q2 ξ . 0) 2 eR (k2 )µ = eL (k1 )µ eL (k2 )µ = eR (k1 )µ . cos θ. q2 eµν (q2 ) = q2 eµν (q2 ) = 0.3. eL (k1 )µ = 1 √ (0.2). cos θ. eL (q )eL (q )µν = 1 and fulﬁl the equations µ µ ν ν q1 eµν (q1 ) = q1 eµν (q1 ) = 0.3. eL (q2 ) = eR (q1 ).81 sections is very similar to the gluon-gluon scattering in QCD (D.8) The helicity states for the second tensor gauge boson are: µν µν µν eµν R (q2 ) = eL (q1 ) .7) and the tensor gauge boson wave functions for circular polarizations along the q1 direction are: 1 eµν R (q1 ) = 2 (0. 3. −i. sin θ ) (0. (0. .3.3. sin θ). . It is easy to check that the wave functions (3.

3. = .3.82 On mass-shell gauge transformations should fulﬁll the following conditions: 2 q2 = 0. tr eq2 = 0.10) MIII (LL → RI ) = 1 ig2 g 2 (f ace f bde − f ade f bce )E sin θ ξ1 cos θ − iξ2 − ξ3 sin θ 4 = = (f ace f bde − f ade f bce )M3 (3. In that case we shall get the following amplitudes (see ??): iMI (LL → RI ) = θ θ θ 1 ig2 g 2 f ace f bde E cos3 (ξ0 − ξ3 ) cos − (ξ1 + iξ2 ) sin 2 2 2 2 = = f ace f bde M1 iMII (LL → RI ) = 1 θ θ θ ig2 g 2 f ade f bce E sin3 (ξ0 + ξ3 ) sin − (ξ1 − iξ2 ) cos 2 2 2 2 = = f ade f bce M2 (3.11) ig 2 g2 ace bde (f f + f ade f bce )E sin θ cos θ iξ2 − ξ1 cos θ + ξ3 sin θ iMIV (LL → RI ) = 4 = (f ace f bde + f ade f bce )M4 . q2 · eq2 = 0. These equations are satisﬁed if q2 · ξ = 0 and therefore ξ0 = ξ1 sin θ + ξ3 cos θ. To see explicitly how the cancellation between diagrams takes place let us take the incoming vector bosons left polarized and one of the tensor bosons right polarized.

(3. Nevertheless. sin θ cos θ iξ2 − ξ1 cos θ + ξ3 sin θ . 2 2 2 cos3 sin θ ξ1 cos θ − iξ2 − ξ3 sin θ . 1 . 2M2 + M3 + M4 = 0. M2 − M3 + M4 = 0. (3. 2 2 2 θ θ θ sin3 (ξ0 + ξ3 ) sin − (ξ1 − iξ2 ) cos . one can straightforwardly see that the above equations are not satisﬁed. This indicates that the truncation we have implemented to the total Lagrangian.3. so as to include only tensors of the 2nd rank is not legitimate. One may also assume that the neglected terms would result in Tprop → −2Tprop and (V V T T ) → −(V V T T ). in order to arrive at a meaningful result. one can see that the following equations hold.13) Using the explicit expressions (3.12). but of the same order in perturbation theory should be included. (3. 2M1 − M3 + M4 = 0.3.12) For the total amplitude we shall get iM(LL → RI ) = f ace f bde M1 + M3 + M4 + f ade f bce M2 − M3 + M4 . In what follows.14) This means that the terms that have been neglected would result in doubling the rank-2 tensor propagator and reversing the sign of the TTV vetex 1 .3.3. and it nulliﬁes if M1 + M3 + M4 = 0.83 where M1 = M2 M3 M4 ig 2 g2 E 2 ig 2 g2 = E 2 ig 2 g2 = E 4 ig 2 g2 = E 4 θ θ θ (ξ0 − ξ3 ) cos − (ξ1 + iξ2 ) sin . Terms of higher rank.

(3.15).2).22).3. The scattering amplitude (3.ab DF (k ) = − 2 µλ νρ i µλ νρ µρ νλ µν λρ 2( η η + η η − η η ) − (η η − η µρ η νλ ) 2 g2 k 9 (3.5) we can ﬁnd their explicit form.(3. (3.7).3 Helicity Amplitudes Now we can calculate all sixteen matrix elements between states of deﬁnite helicities.3.3. vertices (1. For the t-channel amplitude (3.3.15) 3.8) into the matrix elements (3.9).1. iMI (RL → RL) = 0.3. iMI (LR → RL) = 0 i ace bde iMI (RL → LL) = iMI (RL → RR) = −g 2 g2 4 f f (1 + cos θ)(3 − 2 cos θ) iMI (RL → LR) = 0.3. we shall get the following sequence of sixteen polarization amplitudes: i ace bde iMI (LL → LL) = iMI (RR → RR) = −g 2 g2 3 f f (1 + cos θ)(2 − cos θ) 4 iMI (LL → LR) = iMI (RR → LR) = 0 iMI (LL → RL) = iMI (RR → RL) = 0 i ace bde iMI (LL → RR) = iMI (RR → LL) = −g 2 g2 8 f f (7 − 2 cos θ + cos 2θ) i ace bde iMI (LR → LL) = iMI (LR → RR) = −g 2 g2 4 f f (1 + cos θ)(3 − 2 cos θ) iMI (LR → LR) = 0.1.1.3.(3.3. the eﬀective propagator: µν.4) and (3.3.84 we shall assume the presence of those terms and use instead of (2.λρ.3.16) and helicity wave functions (3.3.2).6) for any particular choice of helicities contains four terms.2.3. By plugging the explicit expressions for propagators (1.16) .24).3).12). (3. (3.1. (3.

4) we get: iMII (RL → RL) = 0. (3.3.3. For the s-channel diagram (3.3.85 For the u-channel diagram (3.18) . iMIII (RL → RL) = 0. iMII (LR → RL) = 0 i ade bce iMII (RL → LL) = iMII (RL → RR) = −g 2 g2 4 f f (1 − cos θ)(3 + 2 cos θ) iMII (RL → LR) = 0.3.3) the amplitudes are: i ace bde iMII (LL → LL) = iMII (RR → RR) = −g 2 g2 3 f f (1 − cos θ)(2 + cos θ) 4 iMII (LL → LR) = iMII (RR → LR) = 0 iMII (LL → RL) = iMII (RR → RL) = 0 i ade bce f f (7 + 2 cos θ − cos 2θ) iMII (LL → RR) = iMII (RR → LL) = −g 2 g2 8 i ade bce iMII (LR → LL) = iMII (LR → RR) = −g 2 g2 4 f f (1 − cos θ)(3 + 2 cos θ) iMII (LR → LR) = 0.17) i (f ace f bde − f ade f bce ) cos θ iMIII (LL → LL) = iMIII (RR → RR) = −g 2 g2 2 iMIII (LL → LR) = iMIII (RR → LR) = 0 iMIII (LL → RL) = iMIII (RR → RL) = 0 i iMIII (LL → RR) = iMIII (RR → LL) = −g 2 g2 2 (f ace f bde − f ade f bce ) cos θ iMIII (LR → LL) = iMIII (LR → RR) = 0 iMIII (LR → LR) = 0. (3. iMIII (LR → RL) = 0 iMIII (RL → LL) = iMIII (RL → RR) = 0 iMIII (RL → LR) = 0.

3. iMIV (RL → RL) = 0. VL VR → TL TL .20) iMLL→LL = − iMLL→RR = − (3. VR VL → TR TR . VL VL → TR TR . ig2 g 2 ace bde f f (3 + cos θ) + f ade f bce (3 − cos θ) .6) summing the corresponding terms from each diagram: ig2 g 2 ace bde f f (4 + 5 cos θ + cos 2θ) + 4 + f ade f bce (4 − 5 cos θ + cos 2θ) . VR VR → TL TL . We can get the total helicity amplitudes (3.3. VL VR → TR TR .22) (3.19) VR VR → Thus only eight amplitudes out of sixteen are nonzero: VL VL → TL TL . TR TR .21) iMLR→LL = − .3. 4 ig2 g 2 (1 + cos θ)(2 − cos θ)f ace f bde + 4 + (1 − cos θ)(2 + cos θ)f ade f bce .3.3.3.86 And ﬁnally for the contact diagram (3. iMIV (LR → RL) = 0 i iMIV (RL → LL) = iMIV (RL → RR) = g 2 g2 4 (f ace f bde + f ade f bce ) sin2 θ iMIV (RL → LR) = 0.5) we get the following polarization amplitudes: i iMIV (LL → LL) = iMIV (RR → RR) = g 2 g2 4 (f ace f bde + f ade f bce ) sin2 θ iMIV (LL → LR) = iMIV (RR → LR) = 0 iMIV (LL → RL) = iMIV (RR → RL) = 0 i (f ace f bde + f ade f bce ) sin2 θ iMIV (LL → RR) = iMIV (RR → LL) = g 2 g2 4 i iMIV (LR → LL) = iMIV (LR → RR) = g 2 g2 4 (f ace f bde + f ade f bce ) sin2 θ iMIV (LR → LR) = 0. (3. VR VL → TL TL . (3.

s 128d(G) (3. + cos 2θ).3.3.27) dσLR→LL = where α = g2 .3. s 512d(G) (3. Adding up all sixteen helicity amplitudes and dividing the result by four in order to average over the initial boson spins we can get unpolarized cross section.22) and then average over the symmetries of the initial bosons and sum over the symmetries of the ﬁnal tensor gauge bosons.3.21). (3.1) yields: 2 dσLL→LL = g2 2 (G) α 2 C2 (124 − 23 cos 2θ + 3 cos 4θ) dΩ.20).3.3. (3. Helicity cross-sections. Plugging matrix elements (3. Thus summing over helicities 1 4d(G)2 col.3.28) Unpolarized cross section.24). into our general cross-section formula in the center-of-mass frame (3.25) 1 d(G)2 col 2 |MLL→RR |2 = g2 2 (G) g 4 C2 (55 32 d(G) 1 d(G)2 2 2 col |MLR→LL | = g2 2 (G) g 4 C2 (61 128 d(G) − 32 cos 2θ + 3 cos 4θ).3.3.3. (3.3.23) (3. s 128d(G) 2 α 2 C2 (G) (55 + cos 2θ) dΩ.24) (3.87 To compute the cross section.hel |M|2 = 1 4d(G)2 4 col 2|MLL→LL |2 + 2|MLL→RR |2 + 4|MLR→RR |2 = 419 − 76 cos 2θ + 9 cos 4θ (3.3.29) 2 g = g2 128 2 (G) C2 d(G) . 4π 2 g2 2 (G) α 2 C2 (61 − 32 cos 2θ + 3 cos 4θ) dΩ.23). (3.25).3.3. where the invariant operator C2 is deﬁned by the equation ta ta = C2 1. (3. We can calculate now the leading-order polarized cross sections for the tensor gauge boson production V + V → T + T .26) 2 dσLL→RR = g2 (3. This gives: 1 d(G)2 col 2 |MLL→LL |2 = g2 2 (G) g 4 C2 (124 32 d(G) − 23 cos 2θ + 3 cos 4θ). we must square matrix elements (3.

One can only speculate that at high enough energies.88 for unpolarized cross section we shall get 2 dσ = g2 2 (G) 419 − 76 cos 2θ + 9 cos 4θ α 2 C2 dΩ.30) has characteristic behavior with its maximum at θ = π/2 and decrease for small angles.15).31) This cross section increases at small angles θ ∼ 0.3.3.2. Indeed. This cross section should be compared with the analogous cross section in QCD. The production cross section of spin-two gauge bosons (3. we may observe the standard spin-one gauge boson together with its new partner.30) shows dramatically diﬀerent behavior with its maximum in the transverse direction at θ = π/2 and decrease in forward and backward directions. spin-two gauge boson. may be at LHC energies.30) where for the SU (N ) group we have = N2 . The V + V → V + V cross section is dσ = 2 (G) (4 − sin2 θ)3 α 2 C2 dΩ. let us compare this result with the gluon-gluon scattering cross section formula (D. s d(G) 32 sin4 θ (3. . s d(G) 512 2 (G) C2 d(G) (3. π and therefore the scattering is mostly going into forward and backward directions and has its minimum in transverse direction at θ = π/2.3.3. (N 2 −1) The production cross section of tensor gauge bosons (3.

e(2) µν = 0 1 0 1 0 0 0 0 0 0 g2 g2 is uniquely 0 0 .Conclusions We conclude this thesis by summarizing the basic results in which the author contributed. are correspond to the helicities of a symmetric 0 0 0 −1 e(1) µν = 0 0 0 0 tensor boson ﬁeld of rank 2 0 0 0 0 0 0 0 0 0 1 . 0 0 while the third to the degree of freedom of an antisymmetric ﬁeld which behaves like a scalar. We saw. Solving the equations of motion of the free rank-2 tensor gauge theory. it was found that there are three physical propagating modes. eA µν 0 0 0 0 0 0 1 0 = 0 −1 0 0 0 0 0 0 89 . The ratio of the coeﬃcients of those terms determined by the Bianchi identities of the free theory (g = 0). that the Lagrangian which describes rank-2 tensor gauge boson ﬁelds consists of two parts. The ﬁrst two. We speciﬁed on the ﬁelds of rank 2 in the theory of non Abelian tensor gauge bosons and examined carefully the problem of unitarity.

to the interacting case two processes were thoroughly examined. That of the pair production of 2 symmetric rank-2 tensor gauge boson particles through the annihilation of a pair of S. In the f f → T T case. Turning.90 The antisymmetric propagating mode contributes with a factor of three in the energy momentum tensor of the free rank-2 tensor gauge ﬁeld.were calculated by means of the Mathematica formalism. . The known S. while as we saw ( d ) shows Ω V V →V V Ω V V →T T preference in transverse directions. cross sections. As dσ dσ we know. In the V V → T T case. processes were carefully studied and the well known results -Ward identities.M. processes. The calculations were performed with Mathematica and the results were compared with more familiar S.M. 4 For this value the cross section was found to be: α2 C2 (r)C2 (G) sin2 θ dΩ. dσ = s 64d(r) showing a simple sin2 θ dependence. we got: 2 dσ = g2 2 α 2 C2 (G) 419 − 76 cos 2θ + 9 cos 4θ dΩ s d(G) 512 which appears a totally diﬀerent behavior with the QCD process V V → V V .2. if we chose f = g2 . . as follows: 2 2 Tµν ∼ kµ kν (c2 1 + c2 + 3c3 ). ( d ) maximizes at small angles. fermions and that of the same pair production through the annihilation of two vector gauge bosons. it was found that the scattering amplitude is gauge invariant.M. a result which is also visible in the propagator equation (2. it was found that the scattering amplitude becomes gauge invariant only if we add the terms including higher rank ﬁelds we have initially truncated. With this on mind.9).

Appendices 91 .

92 .

93 Appendix A Free rank-2 Tensor Gauge Boson Equation Matrix The matrix operator for non-Abelian tensor gauge theory (2.0.26) is of dimension 16 × 16 and has the following explicit form: 0 0 0 0 0 k2 2 2 −k 2 0 0 0 0 2 −k 2 0 0 0 0 0 0 −k 2 0 0 0 0 kω 2 0 k2 2 0 0 0 0 −kω 0 0 0 0 −k 2 0 0 0 0 0 0 0 kω 2 0 0 −kω 0 0 0 k2 2 0 0 0 kω 2 0 0 0 0 kω 2 0 0 0 0 0 0 2 −k 2 0 0 −k2 0 0 −kω 0 0 0 0 0 kω 2 0 0 2 −ω 2 0 0 − kω 2 0 0 0 0 0 0 1 (k2 − ω 2 ) 2 0 − kω 2 0 0 0 0 0 0 0 0 −k 2 + ω 2 0 0 1 (k2 − ω 2 ) 2 0 0 0 0 2 −ω 2 0 0 0 − kω 2 0 k2 2 0 kω 0 0 ω2 0 0 0 0 0 0 0 0 0 0 0 0 0 0 −k2 0 0 −kω 0 0 kω 2 0 0 2 −k 2 0 0 0 0 0 1 (k 2 − ω 2 ) 2 0 0 −k 2 + ω 2 0 0 0 0 0 0 2 −ω 2 0 0 − kω 2 0 1 (k2 − ω 2 ) 2 0 0 0 0 0 0 − kω 2 0 0 2 −ω 2 0 0 − kω 2 0 0 0 0 0 kω 0 0 ω2 0 0 0 0 0 0 0 0 kω 2 0 0 2 −ω 2 0 0 kω 2 0 0 0 0 0 0 kω 0 0 − kω 2 0 0 0 0 0 0 2 −ω 2 0 ω2 0 0 0 0 kω 0 0 0 2 −ω 2 0 0 − kω 2 0 0 2 −ω 2 0 0 ω2 0 0 0 0 0 0 0 0 0 0 0 0 0 (A.1) 0 0 .1.

94 and allows to calculate its rank as a function of momenta. .

d(G) tr(ta tb ) = C (r)δ ab where d(r) is the dimension of the representation r. 95 .. By convention the i and a are indices of the symmetry group G.. C2 (N ) = N2 − 1 . tb ] = if abc tc ... A number of fermions ψ i is equal to the dimension d(r) of the representation r: i = 1. . d(r).. The number of gauge bosons Aa is equal to the number d(G) of generators of the group G: a = 1. and satisfy the relation C (r) = d(r) C2 (r). . The invariant operators C (r) and C2 (r) are deﬁned by the equations ta tb = C2 (r)1. where the structure constants f abc are totally antisymmetric. For the fundamental N and adjoint G representations of the SU (N ) groups we have C (N ) = 1/2..Appendix B Elements of Group Theory The gauge group matrices ta form a representation r of the Lie group G. 2N C (G) = 1/2 = C2 (G) = N. d(G). The matrices ta obey the algebra [ta .

4N 1 (N 2 − 1) .96 The traces over symmetry group indices now can be evaluated: 1 N (N 2 − 1) 1 d(r)C2 (r)C2 (G) = d(G)C (r)C2 (G) = . tr(ta tb ta tb ) = d(G)C (r)C2 (r)(C2 (r) − C2 (G)) = − 2 4N −itr(f abc ta tb tc ) = . 2 2 4 N (N 2 − 1) . tr(f abc f abc tc tc ) = d(r)C2 (r)C2 (G) = d(G)C (r)C2 (G) = 2 (N 2 − 1)2 tr(ta tb tb ta ) = d(r)C2 (r)C2 (r) = d(G)C (r)C2 (r) = .

They are and k2 GG ∼ T r{ p+ ( k2 − k1 ) p− ( k2 − k1 )} = 8[p+ · (k2 − k1 ) p− · (k2 − k1 ) + p+ · p− k1 · k2 ].Appendix C Evaluation of Traces In this appendix we shall perform calculation of traces which appear in the squared matrix element (3.16). (RG∗ )µν ∼ T r{ p+ γ ν ( p− − k1 )γ µ p− ( k2 − k1 )} = = 4{g νµ [−p+ · p− k1 · k2 + p+ · (k2 − k1 ) p− · k1 + p+ · k1 p− · (k2 − k1 )] + µ ν µ ν µ ν µ + k1 · k2 (pν − p+ − p+ p− ) + 2p− · (k2 − k1 )p+ p− + 2p+ · (k2 − k1 )p− p− }. These terms can be ignored. Therefore we shall calculate the traces up to the longitudinal terms which are proportional to the µ ν vectors k1 .2. because after contraction with the corresponding transverse wave functions of the tensor gauge bosons eµα (k1 ) and eνβ (k2 ) they give zero contribution. (LG∗ )µν ∼ T r{ p+ γ µ ( p− − k2 )γ ν p− ( k2 − k1 )} = = 4{g µν [p+ · p− k1 · k2 + p+ · (k2 − k1 ) p− · k2 + p+ · k2 p− · (k2 − k1 )] + µ ν µ ν µ ν ν + k1 · k2 (pµ + p− − p− p+ ) + 2p− · (k2 − k1 )p+ p− + 2p+ · (k2 − k1 )p− p− }. The traces under consideration have terms proportional to µ ν the momentum of the tensor gauge bosons k1 and k2 . 97 .

. (LL∗ )µνµ ν ∼ T r{ p+ γ µ ( p− − k2 )γ ν p− γ ν ( p− − k2 )γ µ } = µ µ µ µ µµ ν = {16pν ]+ − p − [p− p+ + p+ p − − p + · p − g µ µ µ µν µν − (p+ · k2 pµ + + 8 pν − − p− · k2 p+ )g − [(p+ · k2 p− + p− · k2 p+ )g ν µµ + ( p + · k2 pν ]+ − − p− · k2 p+ )g µ µ µ µν µν + + 8 pν − (p+ · k2 pµ − − p− · k2 p+ )g − [(p+ · k2 p− + p− · k2 p+ )g ν µµ + ( p + · k2 pν ]+ − − p− · k2 p+ )g + 8p+ · k2 p− · k2 [g µν g µ ν − g µν g µ ν + g µµ g νν ]}. (RR∗ )µνµ ν ∼ T r{ p+ γ ν ( p− − k1 )γ µ p− γ µ ( p− − k1 )γ ν } = µ νν ν ν ν ν ]+ = {16pµ − p− [p+ p − + p− p+ − p+ · p− g ν ν νµ ν ν µ + − (p+ · k1 pν + 8 pµ − [(p+ · k1 p− + p− · k1 p+ )g − − p− · k1 p+ )g µ νν + (p+ · k1 pµ ]+ − − p− · k1 p+ )g ν ν ν + 8 pµ − [(p+ · k1 p− + p− · k1 p+ )g µ ν νµ − (p+ · k1 pν + − − p− · k1 p+ )g µ νν + (p+ · k1 pµ ]+ − − p− · k1 p+ )g + 8p+ · k1 p− · k1 [g νµ g µ ν − g νµ g µν + g νν g µµ ]}. (GR∗ )µν ∼ T r{ p+ ( k2 − k1 ) p− γ µ ( p− − k1 )γ ν } = = 4{g µν [−p+ · p− k1 · k2 + p+ · (k2 − k1 ) p− · k1 + p+ · k1 p− · (k2 − k1 )] + µ ν µ ν ν ν µ + k1 · k2 (pµ + p− − p− p+ ) + 2p− · (k2 − k1 )p+ p− + 2p+ · (k2 − k1 )p− p− }.98 (GL∗ )µν ∼ T r{ p+ ( k2 − k1 ) p− γ ν ( p− − k2 )γ µ } = = 4{g µν [p+ · p− k1 · k2 + p+ · (k2 − k1 ) p− · k2 + p+ · k2 p− · (k2 − k1 )] + µ ν µ ν µ ν ν + k1 · k2 (pµ + p− − p− p+ ) + 2p− · (k2 − k1 )p+ p− + 2p+ · (k2 − k1 )p− p− }.

99 (LR∗ )µνµ ν ∼ T r{ p+ γ µ ( p− − k2 )γ ν p− γ µ ( p− − k1 )γ ν } = = −4k1 · k2 p+ · p− g µν g νµ + 4p+ · k2 p− · k1 g µν g νµ + 4p+ · k1 p− · k2 g µν g νµ − 4k1 · k2 p+ · p− g µν g µ ν + 4p+ · k2 p− · k1 g µν g µ ν + 4p+ · k1 p− · k2 g µν g µ ν + 4k1 · k2 p+ · p− g µµ g νν − 4p+ · k2 p− · k1 g µµ g νν − µ 4p+ · k1 p− · k2 g µµ g νν − 4k1 · k2 g νν pµ + p− + µ νµ ν µ 4k1 · k2 g µ ν pν p + p− − + p− + 4k1 · k2 g µ νν µ µ p+ p− + 4k1 · k2 g νν pµ + p− + 8p− · k2 g µ µν ν µ 4k1 · k2 g µν pν p+ p− − + p− − 8p− · k2 g µ µν ν µ 4k1 · k2 g µν pν + p− + 8p− · k2 g p+ p− − µ µν µ ν p+ p− + 8p+ · k2 g νν pµ − p− − 4k1 · k2 g ν µν µ ν 8p− · k1 g µ ν pµ p+ p− − + p− + 4k1 · k2 g ν µµ ν ν 8p− · k1 g µν pµ p+ p− + + p− − 4k1 · k2 g ν µν µ ν 8p− · k1 g µµ pν p− p− + + p− + 8p+ · k1 g ν µν µ ν 8p+ · k1 g µν pµ p− p− − − p− + 8p+ · k2 g ν ν µ µ ν 16p+ · p− g µν pµ − p− + 16p+ p− p− p− + µν µ ν ν 4k1 · k2 g νµ pµ + p− + 4k1 · k2 g p+ p− − ν µν µ ν 4k1 · k2 g µµ pν + p− + 8p+ · k2 g p− p− − µ µ ν ν ν 8p+ · k1 g µµ pν − p− + 16p+ p− p− p− . .

(C.1) All these traces have been calculated with the use of the Mathematica program [49]. .100 (RL∗ )µνµ ν ∼ T r{ p+ γ ν ( p− − k1 )γ µ p− γ ν ( p− − k2 )γ µ } = = −4k1 · k2 p+ · p− g µν g νµ + 4p+ · k2 p− · k1 g µν g νµ + 4p+ · k1 p− · k2 g µν g νµ − 4k1 · k2 p+ · p− g µν g µ ν + 4p+ · k2 p− · k1 g µν g µ ν + 4p+ · k1 p− · k2 g µν g µ ν + 4k1 · k2 p+ · p− g µµ g νν − 4p+ · k2 p− · k1 g µµ g νν − µ 4p+ · k1 p− · k2 g µµ g νν − 4k1 · k2 g νν pµ + p− + µ µ µν pν 8p− · k2 g νν pµ + p− − 4k1 · k2 g + p− + µ µ νµ 8p− · k2 g µ ν pν pν + p− + 4k1 · k2 g + p− − µ µ νν 8p− · k2 g νµ pν pµ + p− + + p− − 4k1 · k2 g µ µν ν 4k1 · k2 g µν pν p+ pµ − − + p− + 4k1 · k2 g µ µν ν pµ 8p+ · k2 g νν pµ + p− + − p− + 4k1 · k2 g ν µµ ν 4k1 · k2 g µν pµ pν + p− − 4k1 · k2 g + p− + ν νµ ν 8p+ · k2 g µ ν pµ pµ − p− + 4k1 · k2 g + p− − ν µν µ 8p− · k1 g νµ pµ p+ pν + p− − 4k1 · k2 g − + ν µµ ν 8p− · k1 g µν pµ pν + p− − 4k1 · k2 g + p− + ν νµ ν 8p− · k1 g µµ pν pµ − p− + + p− + 8 p + · k1 g ν νµ ν pµ 8p+ · k2 g νµ pµ − p− + − p− − 16p+ · p− g µ µ ν ν ν 8p+ · k1 g µν pµ − p− + 16p+ p− p− p− − µ µ ν ν ν 8p+ · k1 g µµ pν − p− + 16p+ p− p− p− .0.

1 ff → V V For reasons of completeness we present analytically the full calculation of the cross section for the process f f − V V .1. using the Feynman rules (1. µ = iMµν.1.23).1 Feynman diagrams-the scattering amplitude The procedure is governed by the Yang-Mills Lagrangian (1. (1. ν k2 p− = (ig )2 v ¯(p+ )γ µ ta i γ ν tb u(p− ) (D.1. ¡ b.22).Appendix D Processes D.1) p− − k2 101 . D.11). (1.ab A k1 p ¯+ for the t-channel diagram.1.27). One gets: a.1.24).1. The expressions for the three diagrams that contribute to the process can be easily written down. (1. where a fermion and an anti-fermion are annihilated to two vector gauge bosons [47].1.

µ.1. To get the scattering amplitude one sums the three channels and get: iMµν.4) D. µ iMµν.ab + iMµν. while all others are kept in physical-transverse .2) k2 − p+ = (ig )2 v ¯(p+ )γ ν tb p− b.ab = iMµν.ab = B for the u-channel diagram and a.a.1. ν i γ µ ta u(p− ) (D. one of the outgoing vector bosons is put in longitudinal polarization.ab G = = ig v ¯(p+ )γρ tc u(p− )( ¡ ¡ p ¯+ k1 k2 k3 p ¯+ p− 102 k2 .1.b. ν = −i )gf abc η µν (k2 − k1 )ρ + η ρµ (k1 + k3 )ν − η νρ (k2 + k3 )µ 2 k3 (D.k1 iMµν.ab = A B G = (ig )2 v ¯(p+ ) γ µ ta −γρ tc ( i i γ ν tb + γ ν tb γ µ ta − p− − k2 k2 − p+ 1 abc µν )f [η (k2 − k1 )ρ + η ρµ (k1 + k3 )ν − η νρ (k2 + k3 )µ ] u(p− ) 2 k3 (D.2 Gauge invariance To check the self-consistency of Yang-Mills theory.1.3) for the s-channel diagram.ab + iMµν.

1.1. D.1. It is expected that the amplitude vanishes.3 Cross section Now take both the outgoing vector bosons transversal.5) The simpliﬁcation of the third term stems from the fact that µ µ ∗ e∗ µ (k1 )k1 = eµ (k1 )k2 = 0 Using the Dirac equation for both p− and p+ . iMµν.1.ab = (ig )2 v ¯(p+ ) γ µ ta i i γ ν tb + γ ν tb γ µ ta + p− − k2 k2 − p+ 1 ∗ + ( k1 − k2 )tc ( 2 )f abc η µν u(p− )e∗ µ (k1 )eν (k2 )(D. tb ] − γ µ tc f abc u(p− )e∗ µ (k1 )k2ν = (ig ) v µ (k1 ) (D. e∗ ν (k2 ) = k2ν i i k2 tb + k2 tb γ µ ta − γ µ tc f abc u(p− )e∗ µ (k1 ) p− − k2 k2 − p+ (D.103 modes.ab e∗ ¯(p+ ) −iγ µ [ta .7) 2 k3 Hence.1. the ﬁrst two terms simplify and the above relation becomes: 2 iMµν. Let us take: e∗ µ (k1 ) : transversal then 2 iMµν.ab e∗ ¯(p+ )( k1 − k2 )tc u(p− )( µ (k1 )eν (k2 ) = (ig ) v G 1 abc µν ∗ )f η eµ (k1 )e∗ ν (k2 ) (D. using the generators’ Lie algebra.8) k3 . The third diagram simpliﬁes to ∗ 2 iMµν.6) which obviously cancels.ab e∗ ¯(p+ ) γ µ ta µ (k1 )k2ν = (ig ) v .

colors ∗ ∗ ∗ ∗ × pols ∗ Eµνλρ (k1 k2 ) = = A A + A B + B A + B B + A G + B G + G∗ A + G∗ B + G∗ G.1.1.ab µν. ˜ k·k (D. one gets: 1 1 2 |M| = 4d(r)2 4d(r)2 ∗ M∗µν.ab = Mµν.104 or.11) ∗ where Eµνλρ (k1 k2 ) = e∗ µ (k1 )eν (k2 )eλ (k1 )eρ (k2 ). (D.ab = M∗ + M∗ + M∗ = A B G ν µ ν − γ ν k2 )γ µ a a γ (2p+ − k2 γ ) b t +t t + = (−ig ) u ¯(p− ) t −2p− · k2 2p+ · k2 1 + i( k1 − k2 )tc ( 2 )f abc η µν v (p+ )eµ (k1 )eν (k(D.ab M∗µν. ˜2µ = k1µ . using the Dirac equation Mµν. k .ab = A B G = (ig )2 v ¯(p+ ) ta ν µ ν (2pν γ µ (2pν + − γ k2 )γ a − − k2 γ ) b t + tb t − −2p− · k2 2p+ · k2 1 ∗ − i( k1 − k2 )tc ( 2 )f abc η µν u(p− )e∗ µ (k1 )eν (k2 )(D.ab spins.10) 2) k3 2 ν b (2p− Summing over the outgoing colors and transverse polarizations and averaging over the incoming spins and colors.ab Mλρ.9) k3 To arrive to the ﬁnal expression for the cross section one needs to square the scattering amplitude. From the completeness relations for the polarization vectors: eµ (k )e∗ ν (k ) = −ηµν + pols ˜ν + k ˜µ kν kµ k .ab + Mµν. −k ) and from the kinematics of the process: where k ˜1µ = k2µ k . The complex conjugate is: µν.1.ab + Mµν.12) ˜µ = (k 0 .1.ab µν.

1.16) .1. Let us see how the calculation goes for A∗ A. k2 ).13) (k1 · k2 )2 −ηµλ + from where we see that the tensor Eµνλρ (k1 . k2 ) is symmetric under the interchange (µν ) ←→ (λρ).105 one has: Eµνλρ (k1 . (D.14) where we have used the completeness relations: us (p)¯ us (p) = s s v s (p)¯ v s (p) = p. 2 4(p− · k2 ) (D. (D.15) Using the Dirac algebra of the gamma matrices one ﬁnds: ρ ν µ λ ρ tr p− (2pν − − γ k2 )γ p+ γ (2p− − k2 γ ) = ν λ µ λ µ λµ + = 16pρ − p− p− p+ + p+ p− − (p+ · p− )η µ µ λρ λ µρ − [(p+ · k2 )pλ + + 8 pν − − (p− · k2 )p+ ]η − [(p+ · k2 )p− + (p− · k2 )p+ ]η ρ µλ + [(p+ · k2 )pρ − − (p− · k2 )p+ ]η µ νλ λ λ µν + 8 pρ − [(p+ · k2 )pµ + − [(p+ · k2 )p− + (p− · k2 )p+ ]η − − (p− · k2 )p+ ]η ν µλ + [(p+ · k2 )pν − − (p− · k2 )p+ ]η + 8(p+ · k2 )(p− · k2 )[η λρ η µν − η µν η µρ + η µλ η νρ ] (D. k2 ) = k2ν k1ρ + k1ν k2ρ k1µ k2λ + k2µ k1λ −ηνρ + = k1 · k2 k1 · k2 ηµλ ηνρ = ηµλ ηνρ − (k2ν k1ρ + k1ν k2ρ ) − (k1µ k2λ + k2µ k1λ ) + k1 · k2 k1 · k2 1 + (k1µ k2λ + k2µ k1λ )(k2ν k1ρ + k1ν k2ρ ). A∗ A = 1 4d2 (r) g 4 tr(tb ta ta tb ) 1 ρ ν µ λ ρ tr p− (2pν − − γ k2 )γ p+ γ (2p− − k2 γ ) Eµνλρ (k1 .1.1.

(D.17) The remaining contributions can be evaluated in a similar manner.1.20) G∗ G = g4 C2 (G)C2 (r)4 sin2 θ 2 4d (r) (D.1. Putting all these together.18) B∗B = 1 + cos θ g4 d(r)C2 (r)C2 (r)8 (1 − cos θ + cos 2θ) 2 4d (r) 1 − cos θ g4 1 d(r)C2 (r)C2 (G)(−4) sin2 θ 4d2 (r) 2 (D.106 Also. from the group properties of the algebra generators (see Appendix A): 2 tr(tb ta ta tb ) = C2 (r)d(r).21) The diﬀerential scattering cross section for the process. = 2 2 dΩ 256π E 4N = where ag = g2 . A∗ B = B ∗ A = g4 d(r)C2 (r)[C2 (r) − C2 (G)]8 cos2 θ 4d2 (r) (D. in the center of mass frame and for an SU (N ) gauge symmetry group is the following: ( dσ 1 1 ( 2 )c.1. averaged over the initial states of the fermions (spins and colors) and summed over the ﬁnal states of the gauge bosons (polarizations and colors).22) 2 N −1 N3 2 spins.19) A∗ G = R∗ G = G∗ A = G∗ B = (D. one ﬁnds: A∗ A = 1 − cos θ g4 d ( r ) C ( r ) C ( r )8 (1 + cos θ + cos 2θ) 2 2 4d2 (r) 1 + cos θ (D.1.m. 4π |M|2 ) = θ N2 (1 + cos2 θ) − 1 .1.1.colors 2 αg 1 + cos2 2 4s sin θ .

The contribution of the tchannel diagram can be expressed in the form: c. −q2 .1 Feynman diagrams Four Feynman diagram contribute into this scattering. −q1 ) − 2 η στ δ ef Vνρτ λ (q1 )eρ (q2 ) p ig 2 ace bde f f [ηµσ (k1 + p)λ + ησλ (q1 − p)µ − ηλµ (q1 + k1 )σ ] × = p2 × [ηνρ (k2 + q2 )σ − ηρσ (q2 + p)ν + ησν (p − k2 )ρ ] × ∗ × eµ (k1 )eν (k2 )e∗ λ (q1 )eρ (q2 ) = ¡ q1 q2 e. µ b.2 VV →VV Here we shall review the well known result for the three-level gluon scattering V + V → V + V [47] in order to compare it with the tensor scattering considered above: V + V → T + T [46].2. p)eµ (k1 )eν (k2 )e∗ = Vµσλ (k1 . ν i bdf ∗ aec (k2 . λ d. −p.2. D.1) . σ p f. ρ iMI = k1 a.107 D. τ k2 = = ig 2 ace bde f f × p2 [(q1 − p) · e(k1 )]e∗µ (q1 ) + [(k1 + p) · e∗ (q1 )]eµ (k1 ) − (q1 + k1 )µ [e(k1 ) · e∗ (q1 )] × ∗ ∗ −[(p + q2 ) · e(k2 )]e∗ µ (q2 ) + [(p − k2 ) · e (q2 )]eµ (k2 ) + (q2 + k2 )µ [e(k2 ) · e (q2 )] (D.

2. −q2 ) − 2 η στ δ ef Vνλτ (k2 . τ = k2 = ig 2 ade bce f f × p2 [(q2 − p ) · e(k1 )]e∗µ (q2 ) + [(k1 + p ) · e∗ (q2 )]eµ (k1 ) − (q2 + k1 )µ [e(k1 ) · e∗ (q2 )] × ∗ ∗ −[(p + q1 ) · e(k2 )]e∗ µ (q1 ) + [(p − k2 ) · e (q1 )]eµ (k2 ) + (q1 + k2 )µ [e(k2 ) · e (q1 )] (D. ν i bcf aed ∗ = Vµσρ (k1 .that of the u-channel: c. −q1 . −p . d. q1 iMII = k1 a. λ. σ p 108 q2 .2) . µ b. ρ f. p )eµ (k1 )eν (k2 )e∗ λ (q1 )eρ (q2 ) p ig 2 ade bce f f [ηµσ (k1 + p )τ + ησρ (q2 − p )µ − ηρµ (q2 + k1 )σ ] × = p2 × [ηνλ (k2 + q1 )σ − ηλσ (q1 + p )ν + ησν (p − k2 )λ ] × ∗ × eµ (k1 )eν (k2 )e∗ λ (q1 )eρ (q2 ) = ¡ e.

λ iMIII = b.3) . −q1 )eµ (k1 )eν (k2 )eλ (q1 )eρ (q2 ) p ig 2 abe cde f f [ηνσ (k2 + p )µ − ησµ (p + k1 )ν + ηµν (k1 − k2 )σ ] × = p2 × [ησρ (p + q2 )λ + ηρλ (q1 − q2 )σ − ηλσ (q1 + p )ρ ] × ∗ × eµ (k1 )eν (k2 )e∗ λ (q1 )eρ (q2 ) = a. ρ = ig 2 ade bce f f × p2 [(p + k2 ) · e(k1 )]eµ (k2 ) − [(p + k1 ) · e(k2 )]eµ (k1 ) + (k1 − k2 )µ [e(k1 ) · e(k2 )] × ∗ ∗ ∗ ∗ [(p + q2 ) · e∗ (q1 )]e∗ µ (q2 ) − [(p + q1 ) · e (q2 )]eµ (q1 ) + (q1 − q2 )µ [e (q1 ) · e (q2 )] (D. −p . k1 ) − 2 η στ δ ef Vτ ρλ (p .The s-channel contribution is: c. τ q2 p e. ν i f dc bea ∗ ∗ = Vνσµ (k2 . µ ¡ q1 f.2. σ k1 k2 109 d. −q2 .

0.2. eL (k2 )µ = √ (0. ν (f abe f cde − f bce f ade )[e(k1 ) · e∗ (q1 )][e(k2 ) · e∗ (q2 )] − − (f abe f cde − f ace f bde )[e(k1 ) · e∗ (q2 )][e(k2 ) · e∗ (q1 )] + + (f ace f bde − f ade f bce )[e(k1 ) · e(k2 )][e∗ (q1 ) · e∗ (q2 )] (D.2. 1. − sin θ) .2. µ = −ig 2 × ¡ q1 q2 e k1 k2 . i. −sinθ. In the center of mass frame the kinematics of the process are as follows: µ k1 = k (1. 0.2. cos θ. 0) .2. 0. − cos θ) 1 1 eL (k1 )µ = √ (0. 1) µ = k (1. one right-handed and one longitudinal outgoing gluons.2). i.3). . −1) µ q2 = k (1. (D. (D. λ iMIV = a. ρ = b.1). (D. 0.Finally. 0. eI (q2 )µ = q2 . 0. L and I stand for right-handed. 2 where R. Substituting the above expressions in equations (D. 110 d. the contact diagram is: c. 1.4) we .4) D. sin θ.2 Gauge invariance We check gauge invariance with two left-handed incoming gluons. cos θ) q1 µ k2 = k (1. −i. left-handed and longitudinal polarizations. 0) 2 2 1 µ eR (q1 )µ = √ (0.2.

111 x q1 k1 E © θ ' k2 z q2 Figure D.1: Incoming and outgoing momenta in the center-of-mass frame get: iMI (LL → RI ) = −igk 2 f ace f bde iMII (LL → RI ) = igk 2 f ade f bce (1 − cos θ) sin θ √ 2 2 (1 + cos θ) sin θ √ 2 2 √ iMIII (LL → RI ) = igk 2 f abe f cde 2 sin θ sin θ (1 + cos θ) sin θ ade bce (1 − cos θ) sin θ √ √ iMIV (LL → RI ) = −ig 2 f abe f cde √ +f ace f bde −f f 2 2 2 2 2 Each of the terms in iMIV is combined with the ﬁrst three diagrams to give ig 2 iM(LL → RI ) = √ k sin θ f abe f cde + f ace f dbe + f ade f bce 2 which vanishes by Jacobi identity. The forth diagram is combined with each of the ﬁrst three diagrams to give the cancellation. . In a similar way one can show that the scattering amplitude vanishes for every possible combination of helicities provided that one particle is in longitudinal polarization. The cancellation is rather delicate.

2.3 Scattering amplitudes Now let us calculate the scattering amplitudes for each possible combination of physical-transverse polarizations.6) . sin4 2 (D.2. − 24 cos θ + cos 2θ) cot2 sin2 θ sin4 cos4 θ 2 θ 2 θ 2 iMI (LL → LR) = iMI (LL → RR) = iMI (LR → LR) = iMI (LR → RL) = for the u-channel diagram: iMII (LL → LL) = ig 2 ace bde f f 4 ig 2 ace bde f f 2 ig 2 ace bde f f 2 ig 2 ace bde f f 2 3+cos θ sin2 θ 2 3+cos θ sin2 θ 2 3+cos θ sin2 θ 2 θ .5) ig 2 ade bce f f (39 8 + 24 cos θ + cos 2θ) tan2 sin2 θ cos4 cos4 θ 2 θ 2 θ 2 iMII (LL → LR) = iMII (LL → RR) = iMII (LR → LR) = iMII (LR → RL) = ig 2 ade bce f f 4 ig 2 ade bce f f 2 ig 2 ade bce f f 2 ig 2 ade bce f f 2 3−cos θ cos2 θ 2 3−cos θ cos2 θ 2 3−cos θ cos2 θ 2 θ sin4 2 . One has the following equalities: iM(LL → LL) = iM(RR → RR). (D.112 D. iM(LL → RR) = iM(RR → LL).2. For the t-channel diagram: iMI (LL → LL) = ig 2 ace bde f f (39 8 iM(LR → RL) = iM(RL → LR). iM(LR → LR) = iM(RL → RL). iM(LL → LR) = iM(RR → LR) = iM(LR → LL) = iM(LR → RR) = = iM(LL → RL) = iM(RR → RL) = iM(RL → LL) = iM(RL → RR). Because of parity and crossing symmetry from the 16 possible amplitudes only 5 are diﬀerent.

2.8) iMLR→LR = 2ig 2 cos2 θ θ θ θ cot f ace f bde + tan f ade f bce cot 2 2 2 2 1 f ace f bde + 1 − cos θ (D. iMLL→RR = 0.9) (D.7) (f ace f bde + f ade f bce ) sin2 θ iMIV (LL → LR) = − ig 4 θ θ iMIV (LL → RR) = −ig 2 f abe f cde cos θ + f ace f bde (1 − cos4 2 ) + f ade f bce (1 − sin4 2 ) 2 iMIV (LR → LR) = ig 2 (f ace f bde + f ade f bce ) cos2 θ iMIV (LR → RL) = ig 2 (f ace f bde + f ade f bce ) sin2 θ.11) 1 + cos θ together with iMLL→LR = 0.2. So for the total amplitudes one has: iMLL→LL = 4ig 2 1 f ace f bde + 1 − cos θ 1 f ade f bce 1 + cos θ (D.113 for the s-channel diagram: iMIII (LL → LL) = −ig 2 f abe f cde cos θ iMIII (LL → LR) = 0 iMIII (LL → RR) = −ig 2 f abe f cde cos θ iMIII (LR → LR) = 0 iMIII (LR → RL) = 0 and for the contact diagram: θ θ ) + f ade f bce (1 − cos4 2 ) iMIV (LL → LL) = −ig 2 f abe f cde cos θ + f ace f bde (1 − sin4 2 (D.2. Thus only 3 × 2 = 6 of 16 helicity amplitudes are nonzero.10) iMLR→RL = ig 2 (1 − cos θ)2 1 f ade f bce (D.2. .2.

C2 (G) d(G) sin4 θ (D.2.14) Using these formulas and (3.2.25) are decreasing at θ ∼ 0. π and increasing in the transverse direction θ ∼ π/2. 4 d(G) sin θ 2 (D.1) one can easily get the cross section: 2 α 2 C2 (G) (4 − sin2 θ)3 dσ = dΩ.14) is that they increase at small angles θ ∼ 0.3.16)-(3.12) g4 2 3 + cos2 θ |MLR→LR | = C2 (G) (1 + cos θ)4 . d(G) sin4 θ (D.2.3.2. .2.(D. D.4 Cross Section Squaring the matrix elements one can get |MLL→LL |2 = 3 + cos2 θ 16g 4 2 C2 (G) .3.3 e+e− → W +W − Since the diﬀerential cross section for the production of two transversal gauge bosons ¯ → T T is similar to the standard electroweak result in the of spin 2 in the process f f process e+ e− → W + W − .12) .13) |MLR→RL |2 = g4 2 3 + cos2 θ (1 − cos θ)4 .3. it is instructive to expose here the complete calculation for the scattering amplitude and the cross section for this process.114 D.5)-(D. π and therefore the scattering is mostly going into forward and backward directions.8) with the corresponding helicity amplitudes for the tensor gauge bosons (3.15) It is also instructive to compare the above helicity amplitudes (D.23)-(3.19). s d(G) 32 sin4 θ (D.2. In contrast to that behavior tensor gauge boson amplitudes (3. The characteristic feature of the squared amplitudes (D.2.2.3.2.

− Fµν 4 4 where ΨL e = L ψν L ψe (D. Since the weak vector bosons should be massive.2 Feynman diagrams In the process contribute 3 diagrams: .1 Lagrangian The total Lagrangian that uniﬁes electromagnetism with the weak interactions is symmetric under local gauge transformations generated by the group U (1)Y × SU (2)L .2) D.1) 1 ¯L µ L µ+ JW = √ (ψ ν γ ψe ) 2 1 ¯L µ L µ− JW = √ (ψ e γ ψν ) 2 1 µ ¯L γ µ ( 1 )ψ L + ψ ¯L γ µ (− 1 + sin2 θw )ψ L + ψ ¯R γ µ (sin2 θw )ψ R ψ JZ = ν ν e e e e cos θw 2 2 ¯e γ µ ψe J µ = −ψ em Y Fµν = ∂µ Bν − ∂ν Bµ a a Fµν = ∂µ A a ν − ∂ν Aµ + g abc c Ab µ Aν (D.3.115 D. The Lagrangian is: ¯R i ∂ψ R + +g (W + J µ+ + W − J µ− + Z 0 J µ ) + eAµ J µ − ¯ L i ∂ ΨL + ψ L = Ψ e e e e µ W µ W µ Z em 1 a a 1 Y Y Fµν − Fµν Fµν + LHiggs .3. the SU (2) gauge symmetry must be spontaneously broken.3.3.

5) e+ the two cases separately.3. µ k+ (D. W − iMII = k− e− Since v ¯L (odd # of γ s)uR = v ¯R (odd # of γ s)uL = 0. µ (D.3. W − iMI = e− ν. W − iMII = e− ν. the polarizations of the + − + incoming electrons can be combined in just two ways: e− R eL and eL eR .3. enables us to neglect w mz the mass of the fermion and terms of the order O( m . The high energy limit considered here. E . We will treat ¡ ¡ ¡ k− k+ γ p− p ¯+ k− k+ Z0 p− p ¯+ ν p ¯− p+ 116 W +.ν.4) e+ W +. µ (D.3) e+ W +. E ).

E ) .8) since only left-handed electrons participate in weak interactions.117 D. 0.3.3. −E ) µ µ k− = (E.3. k cos θ) . −k sin θ. k+ = (E. −k cos θ). (D.3.3 + + − e− R eL → W W iMI = v ¯R γλ uR (−ie) −i ie q2 ∗ η µν (k− − k+ )λ − η νλ (q + k− )µ + η λµ (q + k+ )ν e∗ µ (k+ )eν (k− ) (D. 0. Adding these together we get: iM = iMI + iMII + iMIII = = ie2 v ¯R γλ uR m2 z q 2 (q 2 − m2 z) η µν (k− − k+ )λ − η νλ (q + k− )µ + η λµ (q + k+ )ν × (D. 0. p+ = (E. 0.9) ∗ × e∗ µ (k+ )eν (k− ). √ uR (p− ) = 0 0 0 0 √ 2E . vR (p+ ) = 2E 1 0 0 −1 . where q = k− + k+ The kinematics of the process in the center-of-mass frame are brieﬂy sketched below: µ pµ − = (E. k sin θ. 0.3. 0.7) iMIII = 0. where E = k 2 + m2 w.6) iMII = v ¯R γλ uR (ie) −i ie − m2 Z ∗ η µν (k− − k+ )λ − η νλ (q + k− )µ + η λµ (q + k+ )ν e∗ µ (k+ )eν (k− ) q2 (D.

118 x k− p− E © θ ' p+ z k+ Figure D.2: Incoming and outgoing momenta in the center-of-mass frame 1 0 cos θ 0 0 0 0 0 1 µ √ e∗ ( k ) = − R 2 0 0 sin θ 1 1 cos θ ∗µ = √ = eL (k+ ) 2 0 1 0 −i −i 0 − sin θ 0 cos θ 0 − sin θ 0 1 cos θ ∗µ ∗µ eL (k− ) = √ = eR (k+ ) 2 i − sin θ 1 0 0 0 k k 0 cos θ 0 sin θ 0 E sin θ 1 1 µ e∗ = I (k− ) = mw 0 m w 0 1 0 0 0 0 − sin θ 0 cos θ E E cos θ .

k− · eL.3.10) E (4E 2 − m2 z) + + − − + + − iM(e− R eL → WL WR ) = iM(eR eL → WR WL ) = 0 (D.L (k± ) · eI (k± ) = 0 that ¯+ ) qu(p− ) = 0.R (k− ) = 0 that ∗ e∗ R.13) .I (k+ ) = 0.3.I (k− ) = k+ · eL. v (p and with the aid of the program presented on the next subsection we get: + + − 2 iM(e− R eL → WL WL ) = 2ie m2 z ∗ (¯ vR k− uR )(e∗ L+ · eL− ) = q 2 (q 2 − m2 ) z ie2 m2 z = − k sin θ ≈ 0 (D.119 1 µ e∗ I (k+ ) = mw Using the fact that k −E sin θ 0 −E cos θ ∗ ∗ ∗ k− · e∗ L.11) − + + iM(e− R eL → WR WR ) = − ie2 m2 z k sin θ ≈ 0 E (4E 2 − m2 z) (D.3.R.R.R (k+ ) = k+ · eL.3.12) + iM(e− R eL → WI+ WI− ) = 2ie 2 m2 z 2 2 q (q − m2 z) × ∗ ∗ ∗ × (¯ vR k− uR )(e∗ vR ( e∗ I + · eI − ) + [¯ I + − eI − )uR ](k+ · eI − ) ≈ ≈ −ie2 m2 z sin θ 2m2 w (D.

120 m2 z + + − ∗ 2 iM(e− e → W W ) = − 2 ie (¯ vR e∗ R− uR )(k− · eI + ) ≈ R L I R 2 2 2 q (q − mz ) √ 2 mz mz ≈ −ie2 (1 + cos θ) (D.3.16) ≈ −ie2 4 E mw 2 m2 z + + − ∗ 2 iM(e− e → W (¯ vR e∗ W ) = 2 ie L+ uR )(k+ · eI − ) ≈ R L L I q 2 (q 2 − m2 ) z √ 2 mz mz (1 + cos θ) (D.15) ≈ ie2 4 E mw + iM(e− R eL → + WR WI− ) m2 z ∗ = 2ie (¯ vR e∗ R+ uR )(k+ · eI − ) ≈ 2 2 2 q (q − mz ) √ 2 mz mz (1 − cos θ) (D.3.14) 4 E mw m2 z + + − ∗ 2 (¯ vR e∗ iM(e− e → W W ) = − 2 ie L− uR )(k− · eI + ) ≈ R L I L q 2 (q 2 − m2 ) z √ 2 mz mz (1 − cos θ) (D.3.17) ≈ ie2 4 E mw .3.

121

D.3.4

− + − e+ R eL → W W

Now we have contributions from the three diagrams.

iMI = v ¯L γλ uL (−ie)

−i ie q2

∗ η µν (k− − k+ )λ − η νλ (q + k− )µ + η λµ (q + k+ )ν e∗ +µ e−ν

(D.3.18)

iMII = v ¯L γλ uL

ie2 1 1 (− + sin2 θw ) 2 × 2 q − m2 sin θw 2 z

∗ × η µν (k− − k+ )λ − η νλ (q + k− )µ + η λµ (q + k+ )ν e∗ +µ e−ν

(D.3.19)

iMIII = −

1 ig 2 ∗ [¯ vL e∗ + ( p− k− ) e− uL ] 2 (p − k− )2

(D.3.20)

Adding the ﬁrst two contributions we get: i(MI + MII ) = ie2 (¯ vL γλ uL ) 1 2 q − m2 z m2 1 z − 2 q 2 sin2 θw ×

∗ × η µν (k− − k+ )λ − η νλ (q + k− )µ + η λµ (q + k+ )ν e∗ +µ e−ν

(D.3.21)

w,z 2 Neglecting terms of O( mE ) we get (see program in the next subsection):

122

+ + − iM(e− L eR → WL WL ) = 0

+ + − iM(e− L eR → WL WR ) = −

ig 2 cos θ − 1 2 cos θ − 1 ig 2 cos θ + 1 2 cos θ − 1

sin θ

+ + − iM(e− L eR → WR WL ) = −

sin θ

+ + − iM(e− L eR → WR WR ) = 0

+ iM(e− L eR

→

WI+ WI− )

ig 2 mz =− 4 mw

2

sin θ

+ iM(e− L eR

→

− WI+ WL )

ig 2 mw =− √ 2 2 E

1 mz (1 + cos θ) + 2 mw

2

(1 + cos θ)

ig 2 mw + + − √ iM(e− e → W W ) = − L R I R 2 2 E

(1 + cos θ) −

1 mz 2 mw

2

(1 − cos θ)

ig 2 mw + + − √ W ) = iM(e− e → W I L R L 2 2 E

(1 + cos θ) −

1 mz 2 mw

2

(1 − cos θ)

ig 2 mw + + − √ e → W W ) = iM(e− R I L R 2 2 E

(1 + cos θ) +

1 mz 2 mw

2

(1 + cos θ)

123

In the high energy limit the total, squared scattering amplitude averaged over the initial electron spins will be: mz 1 g4 1−2 |M|2 = 4 spins 16 mw dσ dΩ 1 1 (4π )2 16E 2

2

+

5 mz 4 mw

4

+

2(1 + cos2 θ) sin2 θ 2 (1 − cos θ)

=

CM

1−

mw E

2

×

1 4

|M|2

dσ dΩ

CM

a2 1 mz g = 1−2 2 16 (4E ) mw

2

5 mz + 4 mw

4

+

2(1 + cos2 θ) sin2 θ (1 − cos θ)2

**The cross section for production of two longitudinal Ws is: dσ dΩ where g =
**

+ − (e− e+ → W0 W0 )= CM

1 + sin4 θw α2 sin2 θ dΩ, 256 s sin4 θw cos4 θw

e sin θw

and cos θw =

mw . mz

What is remarkable is that the angular depen-

¯ → T T process. dance (sin2 θ) is exactly as the f f

Phys. Nucl. [3] T. Goldstone. ”Gauge Theory of Elementary Particle Physics” Oxford University Press 1984 [4] S. [7] S. Phys.L. Maiani.. Lett. Utiyama. Theor. Rev. Cheng and L. Phys. ”Aspects of Symmetry” Selected Erice Lectures (1988) [11] M. Li. Phys. [2] R. Rev.W. Kobayashi and T. ”Invariant Theoretical Interpretation of Interaction”. (1961) 154 [9] P. Glashow. ed. 1968). Lett. Rev. ”CP-Violation in the Renormalizable Theory of Weak Interaction” Prog. Rev. [6] S. I. Phys. 101 (1956) 1597. D2 (1970) 1285. Salam. Coleman. 96 (1954) 191. Weinberg. 49 (1973) 652 124 .Svartholin (Almquist and Forlag. Stockholm. Higgs. ”Elementary Particle Theory”. Yang and R. N. 19 (1967) 1264. 12 (1964) 132. Mills.N. [8] J.Bibliography [1] C. Phys. ”A Model For Leptons” Phys. Maskawa. Glashow. Rev. Nuovo Cim 19. [5] A. 22 (1961) 579. ”Conservation of Isotopic Spin and Isotopic Gauge Invariance”. Iliopoulos and L. [10] S.

[19] E. Schwarz and E. Phys. Rev. Nucl.P. 72 (1981) 185-385 [21] M. J. An Introduction to the Bosonic String: Introduction to the Bosonic String v. D 13 (1976) 974. 110 (1984) 1-162 [17] S. B33. [22] E. [25] H. De Vega and A. Nucl. ”Supersymmetry. Georgi. Cambridge Monographs on Mathematical Physics 1987 [15] J. N. Nilles. B 253 (1991) 319.1. 1”. Green. Lett. The Zero Tension Limit Of The Superstring. Supergravity And Particle Physics” Phys. Phys. Veltman. Rep. Phys. Phys. (1967) 29 [14] M. B. Rev. Nucl. Langacker. 25B. (1971) 167. Rep. Popov. B35. Nicolaidis. ”Grand Uniﬁed Theories And Proton Decay” Phys. Lindstrom. Weinberg. Witten Superstring Theory: Introduction v. Phys. . Lett. Cambridge Monographs on Mathematical Physics 2004 [16] H.t Hooft. Dimopoulos. Phys. Sundborg and G. Phys. B 193 (1981) 150 [24] U. J. ”A Supersymmetry Primer” hep-ph/9709356 [18] S. B 12 (1981) 437. Faddeev and V. Acta Phys.125 [12] G. D 14 (1976) 1667 [20] P. Martin. Witten. B 188 (1981) 513. B 295 (1992) 214. Lett. Phys. Strings in strong gravitational ﬁelds. Gildener. [13] D. Theodoridis. ”String Theory:: Volume 1. H. (1971) 173 . Pol. Polchinski. [23] S.

B 552 (2003) 72. Generalization of Yang-Mills theory” arXiv:hep-th/0510258v2 [35] G. K. arXiv:hep-th/0403165. Phys. Savvidy and T. ”Interaction of non-Abelian tensor gauge ﬁelds” arXiv:0804. Phys. A 21 (2006) 4931-4977. ”Conformal Invariant Tensionless Strings” Phys. [hep-th/0505033]. Savvidy. [31] G. B 10 (1969) 516 [28] G.II” Int. Lett. Tsukioka ”Gauge Invariant Lagrangian for Non-Abelian Tensor Gauge Fields of Fourth Rank” [36] S. Lett. Savvidy. K.E. Savvidy. correspondence with SO(D. ”Propagating modes of a non-Abelian tensor gauge ﬁeld of second rank ” J.D) sigma model”. [33] G. Savvidy.M. J. Bakas and C. ”A dual lagrangian for non-Abelian tensor gauge ﬁelds” Phys. Konitopoulos and G. Paton and H.[hep-th/0510258] [32] G. ”Tensionless strings. Phys.126 [26] I. Lett. ”Generalized Veneziano Model With Isospin” Nucl. Mod. ”Tensionless strings: Physical Fock space and higher spin ﬁelds” Int. [27] J. Mod. B 615 (2005) 285.2003 [34] G. A 19 (2004) 3171. A 41 (2008) 355402 [37] J. On the tensionless limit of gauged WZW models. Phys. [29] G. Lett. Sourdis. Savvidy. Chan. [arXiv:hep-th/0310085]. Savvidy. Savvidy. ”Non-Abelian tensor gauge ﬁelds: Generalization of Yang-Mills theory” Phys. J. Phys. Savvidy. Savvidy. Barrett and G. ”Non-Abelian tensor gauge ﬁelds and extended current algebra. [30] G. ”Non Abelian Tensor Gauge Fields I. B 625 (2005) 341. B 652 (2007) 141-145 .

B 120 (1977) 301 [42] Y. Fazio. Kalb and P. 85 (2009) 51001 [45] S. Lett. Nambu. [41] P. Einstein. Duality transformation of non-Abelian tensor gauge ﬁelds. Nucl. A 23 (2008) 999 [39] A. Preuss. The Classical Theory of Fields. ”Magnetic And Electric Conﬁnement Of Quarks”. Konitopoulos. Westview Press ABP. Konitopoulos and G. Savvidy.van Nieuwenhuizen.G.127 [38] S. and G.0847 [46] S.M. Phys. ”Brs Gauge And Ghost Field Supersymmetry In Gravity And Supergravity”. R. Townsend and P. 1975) [44] S. Mod. Akad. Schroeder. Konitopoulos and G.J. 48 (1915) 844. 2 (Butterworth Heinemann. Ramond. Savvidy.4345 [47] M. Massive And Massless Yang-Mills And Gravitational Fields. Phys. ”Production of charged spin-two gauge bosons in gluon-gluon scattering ”. Rev.D. Nucl. arXiv:0804. Peskin and D. Rept. Sitzungsber. Savvidy. Wiss. ”Tensor gauge boson production in high-energy collisions”.K. Guttenberg and G. An Introduction to Quantum Field Theory. Classical direct interstring action. Oxford. [40] M. ”Die Feldgleichungen der Gravitation”. Landau and E. . EPL. 23 (1976) 250 [43] L. van Dam and M. Phys. arXiv:0812. Veltman. Savvidy. Phys. B 22 (1970) 397. Course of Theoretical Physics v. (1995) [48] H. Lifshitz. Phys. ”Production of Spin-Two Gauge Bosons”. D 9 (1974) 2273.

Phys.128 [49] U. D 34 (1986) 3286 [52] F.phys. [53] W.rug. Mod. Phil. Phys. Boyer and A. C. http://wwwth. Derrick et al. ”W Boson Production In E+ E. Dirac.Positron Physics At Petra”. B 119 (1977) 125. Nucl.matrix algebra and Fierz transformations in arbitrary dimensions”.Collisions In The Weinberg-Salam Model”. ”A Mathematica package for performing Gamma. Duinker ”Review Of Electron . Phys. M. M. Proc. J. Rev. Gran.8 T eV ”.nl gran/ [50] P. Buras. . [51] P. A . D 48 (1993) 998. Soc. 54 (1982) 325. Rev. 26 (1930) 361. Phys. Cambr. ”Experimental Study Of The Reactions e+ e− → e+ e− and e+ e− → γγ at 29-Gev”. Alles. Abe et al ”Measurement of the dijet mass distribution in p p ¯ collisions at √ s = 1. Rev.

- Bioethics
- Περί Ομοφωνίας
- Analytical Anarchism
- Heidegger Der Spiegel
- Anarchism and Organization
- Women in the Spanish Civil War
- Τέοντορ Αντόρνο - Μαξ Χόρκχαϊμερ Adorno & Horkheimer-Διαλεκτική του διαφωτισμού-Νήσος (1996)
- Foucault
- Θεωρία του Xάους και Διαλεκτική
- 53927212-αθήνα-οχυρωμένη-πόλη
- ΤΑ ΣΤΑΥΡΟΔΡΟΜΙΑ ΤΟΥ ΛΑΒΥΡΙΝΘΟΥ
- Kouvelakis -The Greek Cauldron

Sign up to vote on this title

UsefulNot usefulPhd Thesis

Phd Thesis

- Effective field theory of W (infinity) Strings and Kac-Moody Algrebras
- Nan Zhao et al- Observable Topological Effects of Mobius Molecular Devices
- principles-of-solid-mechanicstqw_darksiderg.pdf
- PhysRevB.84.195122.pdf
- 50559_appa
- Introduction to Tensor Calculus for General Relativity - MIT
- Identity Recognition Using Edge Suppression Method
- Syllabus.pdf
- Modeling M-Theory Vacua via Gauged S-Duality
- r5.03.22
- Fiction
- Supersymmetry and String Theory. Beyond the Standard Model - Michael Dine - 1ed
- Godel Emergence Life
- Exceptional Strings, Branes and Octonionic Gravity
- Lecture Light Cone
- quantum field theory by Peskin_Chap03 solution
- Subatomic Particles
- Lecture 27
- Pitkanen - Notion of Hyperdeterminant in TGD (2010)
- 10.1.1.167
- Higgs Taus ATLAS - Nunes 12
- Higgs Boson Time Travel
- A Student’s Guide to Geophysical Equations (W. Lowrie 2011).pdf
- AnindaSinha_31.12.08
- J. Niederle and A. G. Nikitin- Relativistic wave equations for interacting, massive particles with arbitrary half-integer spins
- DeRoeck
- The Resource / Volume 3 Issue 4
- Loop
- International Journal of Engineering Research and Development
- silvestre-brac_PhysRevC.26.2640.pdf
- Tensor Gauge Theory of rank 2-tensor fields

Are you sure?

This action might not be possible to undo. Are you sure you want to continue?

We've moved you to where you read on your other device.

Get the full title to continue

Get the full title to continue reading from where you left off, or restart the preview.