SUPERSYMMETRY An Introduction with Conceptual and Calculational Details H. J. W. Miiller-Kirsten A. Wiedemann Department of Physics, Uraversity of Kaiserslautern Kaiserslautern, West Germany World Scientific Singapore © New Jersey * Hona KonaPublished by World Scientific Publishing Co. Pte. Ltd. P.O. Box 128, Farrer Road, Singapore 9128 U.S.A. office: World Scientific Publishing Co., Inc. 687 Hartwell Street, Teaneck NJ 07666, USA Library of Congress Cataloging-in-Publication data is available. SUPERSYMMETRY — AN INTRODUCTION WITH CONCEPTUAL AND CALCULATIONAL DETAILS Copyright © 1987 by World Scientific Publishing Co Pte Ltd. All rights reserved. This book, or parts thereof, may not be reproduced in any form or by any means, electronic or mechanical, including photo- copying, recording or any information storage and retrieval system now Known or to be invented, without written permission from the Publisher. ISBN 9971-50-354-9 9971-50-355-7 pbk Printed in Singapore by Kim Hup Lee Printing Co. Pte, Ltd.PREFACE This text is a detailed version of material presented by both of us in seminars and lectures in the theory group of this department. Except for parts of Chapters 9 and 10 the material has also been covered in a series of seminars by one of us (M.-K.) in the Department of Physics of the University of Adelaide, Adelaide, Australia, in August and September 1985 and in the Department of Physics of Shanxi University, Taiyuan, China, in March and April 1987. The interest and criticism of the audiences at these depart- ments and, in particular, the support and enthusiasm of Professor A.W. Thomas (Adelaide) and Professor Zhang Jian- zu (Taiyuan) are gratefully acknowledged. The text was compiled with the belief that the majority of potential readers is more interested in actu- ally using or applying supersymmetry in some model theory than in painstakingly rediscovering the results of others for themselves. It seemed plausible, therefore, to revise various relevant concepts and in particular, to include the proof or verification of almost every formula. In this way the reader can select the problems he wants to tackle himself, compare his solutions with the calculations given here, and thus gain the confidence in his own calculations which he needs for his discussions of supersymmetry in other contexts. It has been our experience that (except for the last two chapters) the material presented here can be covered in a one-semester course for graduate or post gra- duate students with some knowledge of field theory.vi In compiling this text we have, of course, used pre- vious reviews. The choice of our sequence of topics was motivated by the lecture notes of Legovini’. standard texts which we have consulted are the monograph by Wess and Bagger? and the review by Fayet and Ferrara’, For the detailed treatment of the on-shell Wess-Zumino model we consulted the lecture notes of de Roo. In the text we do not discuss any experimental signatures of supersymmetry. For an introduction into this topic we refer to articles by Haber and kane®’®; further details can be found in the Proceedings of the Thirteenth SLAC Summer Institute on Particle Physics” and in the reviews by Nilles® and Dra- gon, Bllwanger and Schmidt®. As further general references we xefer to the nontechnical review by wess'°, to a very brief review of topics covered here by Campbell and Fogle- man’! and to the lectures of wess!*, rerrara!? ana witten'? A more advanced text is the book by Gates, Grisaru, Rocek and Siegel!>. the very readable review by Sohnius!® appeared after completion of the first draft of our text. Meanwhile several other texts have been published, each, however, with its emphasis in a different direction. We xefer here to the books by west'7, srivastava!® and Freund!?, For more specific topics we refer to the artic- le by Salam and Strathdee*° to the Proceedings of the 28th Scottish Universities’ Summer School in Physics”", and to the Proceedings of the NATO Advanced Study Insti- tute on Supersymmetry**, All considerations of this text refer to a four-dimensional Minkowski space. For the ba- sic technicalities in the context of supersymmetric quan- tum mechanics we refer to the work by Cooper and Freed- man*?, whereas those of two-dimensional field theories can be found in reference 24.Preface CONTENTS Introduction 1 Lorentz Group, Poincaré Group, SL(2,C), Dirac and Majorana Spinors The Lorentz Group The Poincaré Group SL(2,C), Dotted and Undotted Indices Spinor Algebra Calculations with Spinors Connection between SL(2,C) and L. The Fierz-Reordering Formula Further Calculations with Spinors Higher Order Weyl Spinors and their Representations t 4 Dirac and Majorana Spinors The Weyl Basis or Chiral Representation The Canonical Basis or Dirac Representation The Majorana Representation Charge Conjugation, Dirac and Weyl Representations Majorana Spinors Calculations with Dirac Spinors Calculations with Majorana Spinors XL 20 32 32 52 58 72 73 92 100 103 113 118 126 138 141 145viii 2 No-go Theorems and Graded Lie Algebras 2.1 The Coleman-Mandula Theorem and the Haag- Lopuszanski-Sohnius Theorem 149 2.2 Graded Lie Algebras 153 2.2.1 Lie Algebras 153 2.2.2 Graded Algebras 155 2.2.3 Graded Lie Algebras 156 2.3 The Graded Lie Algebra of the Group SU(2,C) 158 2.4 23 Graded Lie Algebras 167 2.5 Graded Matrices 179 3 The Supersymmetric Extension of the Poincare Algebra 3.1 The Supersymmetric Extension of the Poincaré Algebra in the Four-Component Dirac Formulation 189 3.2 The Supersymmetric Extension of the Poincaré Algebra in the Two-Component Weyl Formulation 208 4 Representations of the Super-Poincaré Algebra 4.1 Casimir Operators 210 4.2 Classification of Irreducible Representations 222 4.2.1 N = 1 Supersymmetry 222 4.2.2 N >1 Supersymmetry 233 5 The Wess-Zumino Model 5.1 The Lagrangian and the Equations of Motion 239 5.2 Symmetries 242 5.3 Plane Wave Expansions 251 5.4 Projection Operators 265 5.5 Anticommutation Relations 271 5.6 The Energy-Momentum Operator of the Wess-Zumino Model 288 5.6.1 The Hamilton Operator 2915.6.2 The Three-Momentum Py 5.7 Generators of Infinitesimal Supersymmetry Transformations Superspace Formalism and Superfields 6.1 Superspace 6.2 Differentiation with respect to Grassmann Numbers 6.3 Supersymmetry Transformations in the Weyl Formalism 1 Finite Supersymmetry Transformations 6.3.2 Infinitesima] Supersymmetry Transformations and Differential Operator Representations of the Generators Consistency with the Majorana Formalism Covariant Derivatives Projection Operators Constraints 6.8 Transformation Properties of Component Fields Constrained Superfields and Supermultiplets 7.1 Chiral Superfields 7.2 Vector Superfields and Generalized Gauge Transformations 7.3 The Supersymmetric Field Strength Supersymmetric Lagrangians 8.1 Integration with respect to Grassmann Numbers 8.2 Lagrangians and Actions 8.2.1 Construction of Lagrangians from Scalar Superfields 8.2.2 Construction of Lagrangians from Vector Superfields 300 304 315 320 326 326 336 346 349 361 372 373 388 409 418 433 444 444 4559 Spontaneous Breaking of Supersymmetry 9.1 9.2 9.3 th 9.3.2 9.4 9.4.1 9.4.2 The Superpotential Projection Technique Spontaneous Symmetry Breaking The Goldstone Theorem Remarks on the Wess-Zumino Model The O'Raifeartaigh Model Spontaneous Breaking of Supersymmetry in the O'Raifeartaigh Model The Mass Spectrum of the O'Raifeartaigh Model 10 Supersymmetric Gauge Theories 10.1 10.2 10.3 10.4 Minimal Coupling Super Quantum Electrodynamics The Payet-Iliopoulos Model Supersymmetric Nonabelian Gauge Theory References Index 469 476 501 506 510 511 511 519 530 541 548 564 577xi INTRODUCTION Symmetries are of fundamental importance in the descrip- tion of physical phenomena. In the realm of particle phy- sics symmetries are believed to permit ultimately a classi- fication of all observed particles. A fundamental symme- try of particle physics, which has been firmly established both theoretically and experimentally is that of the Poin- caré group, i.e. of rotations and translations in four- dimensional Minkowski space. Besides this fundamental symmetry there are other socalled internal symmetries (such as the symmetry of the SU(3) flavour group) which have also been firmly established over the last few deca- des, although their manifestation in Nature is not exact. As is well known, the consistent search for more fundamen- tal symmetries led to the development of nonabelian gauge theories and the spectacular experimental confirmation of several predictions of the latter in recent years. In the course of time several attempts have been made to unify the space-time symmetry of the Poincaré group with the symmetry of some internal group. Such attempts have, however, been shown to be futile if the theory, which necessarily has to be a quantum field theory, is expected to satisfy certain basic requirements. In fact, the socalled "no-go" theorem of Coleman and Mandula 25 shows that if one makes the plausible assumptions of locality, causality, positivity of energy and finitenessxii of the number of particles (and one more technical assump- tion) the invariance group of the theory can at best be the direct product of the Poincaré group and a compact (internal) group, and this therefore does not offer a genuine unification of one group with the other. Now the generators of the Poincaré group satisfy well known commutation relations, and Noether's theorem relates these to conserved currents. In their turn the conserved currents are functions of relativistic fields. The commu- tation relations of the field operators which quantize these fields are therefore directly related to those of the generators. It was realized by Wess and Zumino”?*’that if one allows also anticommutation relations of generators of supersymmetry transformations which transform bosons into fermions and vice versa, then the unification of the space-time symmetries of the Poincare group with this internal symmetry can be achieved. The formal proof of this discovery, i.e. the proof that anticommuting gene- rators which respect the other assumptions of the theorem of Coleman and Mandula*°do exist, was established by Haag, Lopuszanski and Sohnius”® Supersymmetry thus arises as a symmetry which combines bosons and fermions in the same representation or multiplet of the enlarged group which encompasses both the transfor- mations of the Poincaré group and the appropriate super- symmetry transformations. Thus every bosonic particle must have a fermionic partner and vice versa. In view of the fact that such a spectrum of particles is not compa- tible with observation, supersymmetry must be badly broken at the level of presently available energies. Clearly only experimental observation can decide whether super- symmetry is indeed inherent in Nature. It can be argued that one of the most immediate ways to observe evidencexiii of supersymmetry is to see if there is a missing energy and momentum in the final ee” spectrum of the reaction ett ee F ett it —», eee + eee where @*, 27 and } are the supersymmetry partners of e", e and % respectively. If there is such a missing energy and momentum it could be that carried away by the neutral photinos # (charged supersymmetry particles at energies presently available would have been detected long ago). Since supersymmetry must be broken, the photinos Y would not be massless. However, supersymmetry does not only open the possi- bility of a much more complex spectrum of particles than heretofore envisaged; supersymmetry also has some intri- guing theoretical consequences which could make it a desirable symmetry. It is well known that a realistic quantum field theory in the traditional sense is plagued by the problem of ultraviolet divergences and the conse- quent necessity of renormalization. Supersymmetry, how- ever, provides a mechanism for the cancellation of such divergences in view of the same number of bosonic and fermionic degrees of freedom in each particle multiplet. Clearly such a built-in cancellation of divergent terms is a highly desirable feature of a quantum field theory. In Chapter 1 we begin with a recapitulation of basic aspects of the Lorentz group, including a discussion of Casimir operators and the classification of representa- tions in terms of their eigenvalues. We then consider the group SL(2,C) and its basic representations, i.e. the self-representation and the complex conjugate self-repre- sentation. The elements of the appropriate representa- tions spaces are the undotted and dotted Weyl spinors. In view of the importance of Weyl spinors throughout thexiv entire text, we consider these here in more detail than is generally done in the literature. We then introduce the concept of Grassmann number and perform some basic manipu- lations involving Weyl spinors, thereby deriving a number of useful formulae. In the subsequent section the connec- tion between SL(2,C) and the proper orthochonous Lorentz group is established. It is then natural to discuss four- component Dirac spinors and the Weyl representation. The connection with two-component Weyl spinors is obtained by introducing four-component Majorana spinors. Then again various formulae are derived which are useful in later calculations. Chapter 2 begins with a discussion of the "no-go" theorems of Coleman and Mandula” and Haag, Lopuszanski and Sohnius”®.the latter leads to a consideration of graded Lie algebras which we approach in successive steps by defining first the characteristics of a Lie algebra, then those of a graded algebra and finally those of a graded Lie algebra, i.e. the properties of grading, supersymme- trization and generalized Jacobi identities. As an example we calculate the graded Lie algebra of the algebra su(2,C). The final section of Chapter 2 deals with gra- ded matrices and their properties. Chapter 3 deals with the grading, i.e. supersymmetric extension of the Poincaré algebra. We demonstrate expli- citly that for the grading chosen all possible Jacobi iden- tities are satisfied. Having established the algebra of the Super-Poincaré group with the fermionic generators in the Dirac four-component form, we then decompose it into the appropriate relations of the Weyl two-component method. In Chapter 4 we use the method of Casimir operators to classify the irreducible representations of the Super- Poincaré algebra, and it is shown that supersymmetry im- plies an equal number of bosonic and fermionic degrees ofxv freedom. Chapter 5 deals with the most immediate field theore- tical realization of the Super-Poincaré algebra, the Wess- Zumino model, which is a field theory involving a scalar field, a pseudoscalar field and one spinor field, all with the same mass. We demonstrate by explicit calculation that the spinor charges of the theory, considered as linear operators in Fock space, satisfy the commutation and anti- commutation relations of the Super-Poincaré algebra. In Chapter 6 we introduce the concepts of superspace and superfields, and define differentiation with respect to Grassmann numbers. Then three different but related operators are constructéd which describe three different but equivalent actions of the supersymmetry group on func- tions in superspace. These operators define three diffe- rent types of superfields. By considering infinitesimal supersymmetry transformations we obtain the corresponding three differential operator representations of the fermi- onic generators of the Super-Poincaré group. Then cova- riant derivatives are introduced as a prerequisite for the construction of manifestly supersymmetric action inte- grals. These covariant derivatives also permit the defi- nition of projection operators. The search for irredu- cible representations of the Super-Poincaré algebra then becomes a search for solutions of constraint equations expressed in terms of these projection operators. ‘The final section of Chapter 6 is devoted to the derivation of the explicit supersymmetry transformations of the component fields of the supermultiplet. In this context it is seen that the highest order component field always transforms into a total Minkowski derivative and thus is a candidate for a supersymmetric Lagrangian density. In Chapter 7 we begin with an investigation of the con- straint equations which define left-handed and right-xvi handed chiral superfields (also known as scalar super- fields). Then vector superfields are defined by an appro- priate constraint equation, and the supersymmetric gene- ralization of the abelian gauge transformation is discus- sed. Finally left-handed and right-handed spinor super- fields are discussed which represent the components of the supersymmetric field strength for an arbitrary vector su- perfield. Chapter 8 deals with the construction of supersymmetric action integrals. We begin with the definition of inte- gration over Grassmann numbers. Then Lagrangians are con- structed from scalar superfields and from vector super- fields (i.e. the supersymmetric field strength). The case of the former is shown to contain the Wess-Zumino model as a special case, whereas the case of the latter yields the supersymmetric generalization of the pure Maxwell theory (i.e. with no interaction with matter fields) which contains in addition to the massless vector field also the massless spinor field of the photino. Chapter 9 deals with the spontaneous breaking of su- persymmetry. For the convenience of discussions the con~ cept of superpotential is introduced. In view of the ne- cessity of evaluating action integrals over superspace an equivalent and convenient Grassmann projection tech- nique is developed. Some general aspects of spontaneous symmetry breaking are then discussed and, in particular, the Goldstone theorem is established for the general case of the breaking of supersymmetry or some other symmetry. Finally the O'Raifeartaigh model, which is a specific theory involving three scalar superfields, is considered and the spectrum resulting from the spontaneous breaking of supersymmetry is investigated. In this case super- symmetry breaking results from the nonvanishing vacuum expectation value of some auxiliary field of a superfield.xvii Finally, in Chapter 10, we consider supersymmetric gauge theories. Introducing first global and local U(1) gauge transformations of scalar superfields and the cor- responding supersymmetric version of minimal coupling, we consider super quantum electrodynamics. We then investi- gate the Fayet-Iliopoulos mechanism of spontaneous brea- king of supersymmetry in which the latter results from the nonvanishing vacuum expectation value of the highest order component field of a vector superfield. The last section contains a brief introduction to nonabelian gauge trans- formations for superfields with the appropriate tensorial transformation properties.CHAPTER 1 LORENTZ GROUP, POINCARE GROUP, SL(2.C). DIRAC AND MAJORANA SPINORS 1.1 The Lorentz Group* A point in the space-time manifold is denoted by (xP) aa x? are the = (x°, x', x7, x?) where x° = t and x!, x space components of the four-vector x#. ‘The laws of physics are invariant under the Lorentz group. Transfor- mations of this group are linear transformations acting on four-vectors xy = AY x” a.) leaving the quadratic form 2 x" = xP a, = Tyke x” = (xe)*—(2)* (1.2) invariant, i.e. a Sections 1.1 and 1.2 serve mainly the purpose of com- pleteness, to define notation and to recollect some formulae which will be needed later in the text. The reader familiar with Sections 1.1 and 1.2 could start imme- diately with Section 1.3 .xt ai =Vuv KIB tv = Juv Mo eA AALE 4 Yor xP xt Tew AMM. 7 Toe (1.3) Here 7, yo diag (1, +1, -1, -1) is the metric tensor; it lowers ‘indices and its inverse he raises indices. Hence Proposition: The constraints at A=t!, {A°.[21 (1.4) define four disconnected pieces in the parameter space. Proof: The determinant of a product of matrices is the pro- duct of the determinants. Hence, taking the determinant of (1.3) yields [at A]*= 1 or &et A =H) Taking the oo-component of (1.3) we obtain Too = Tv APA”, eee and Too A’. A, a Tea A. A*. Ghee eee CA =] * (A) = 1+ CAE )* Hence gg (A’.) 21 The second of constraints (1.4) distinguishes socalled or- thochronous Lorentz transformations with /\°, >] and u nonorthochronous Lorentz transformations with A", < I Proposition: the matrices (/\M, )form a noncompact Liegroup, the Lorentz group L:= 0643) = {A €GLOERIIN GA =7F with Lie algebra ot, 3)!= {OLE Magy CRI| UT =-7N7$ where GL(4,R) denotes the set of all invertible 4x4-matri - ces with real components, and Maya (R) is the set of all 4x4- matrices with real elements. Proof: From Lie algebra theory we know that each A€ 0(1,3) can be written in the form A(t) = exp Ct0t) where t is a real parameter and OLE o(1,3) is an element of the Lie algebra. Matrices of 0(1,3) are subject to the condition NEIDAI=7 Inserting the above expression we obtain Lene (£00)I7 4 [exp (tay) =7- and considering Be {Lemp tou 7 Lexb EOI Fp .5=° we obtain the condition for Lie algebra elements, since the Lie algebra of any Lie group is isomorphic to the tangent space at the identity of the group. It follows that Lheptoy] yp cecny] lene ayy ge explicr)| which leads to mht =o iit ay + YA=o ot =-70l4 vote off,3) In summary we have the following classification: Let A be any invertible 4x4-matrix with real elements, i.e. A € GL(4,R), then (i) The full Lorentz group is= T, L = 0(53) = SNEGLERIAYA =7$ (ii) Proper Lorentz transformations are La t= S00,3)= {A€0013) | ot A=+ IF Ly being a subgroup of L. (iii) Improper Lorentz transformations are Lir= [A€00,3)|detN =-1F L_ is not a subgroup of L, since the identity element is not an element of L_ . Note, however, that discrete trans— formations such as time- or space-reflection are elements of LL. (iv) Orthochronous Lorentz transformations are = ° itr= {A€00,3)|N% 2 +18 * is a subgroup of L. (v) Nonorthochronous Lorentz transformations are Ls= fre oU,3)[A°e < 13 (vi) The restricted Lorentz group is t= LIN Le e = [A€ 01,3) [det N=tl, A, 2+ f This subgroup of L is also called the proper orthochronous Lorentz group; it does not contain time- or space-reflec~ tions. Remark: Lorentz transformations which are orthochronous map the forward light-cone onto itself and AVE L* maps the backward light-cone onto itself. Generators of the Lorentz group - In the neighbourhood of the identity 15,(,,3) 1 @ Lorentz transformation AV€L4 can be written N= Tyne FO (1.6) Inserting this expression into (1.3) we obtainTaw M*, Avs Tuo {St 2h US gtore F f v K m vp Tew f"9 Sete & eat ee Sg Nf @ Loe + oat @r.0 1 o = - wo a, o — (1.7) i.e. the infinitesimal parameters > are antisymmetric in @ and @. Now ath Ae x? (with (1.1) ) Hence (1+0)" 52” (with (1.6) ) = xk+ oF x” 1) and the variation of ial due to infinitesimal transforma-— tions is Tx re ein v = oF ix with antisymmetric infinitesimal parameters as demonstrated in (1.7). On the other hand we may consider the vector representation of the restricted Lorentz group bho and we write ve L, in the form = [exp(-$ OF MH IT, oe where ca 4x4-matrices(M @ —-) constfeste a basis of the Lie algebra 0(1,3), to be verified later. (M¢g)are anti- symmetric in @ and G~ and the factor i is chosen in such a way that the(M @@) are hermitian. For infinitesi- mal transformations, i.e. €9@© infinitesimal, we con-sider Sxt a xia aM, ae (now use (1.8) ) and we conclude oy fe oF ~ fae Mic) ao (1.10) Therefore the generators Moe of the Lorentz group have the matrix form & (Merde es Tew Se 5 av oe ) (1.11) Checking: i v, -£0°"(M, a vrZ wt Cfevd gf Foo) = £ toh, 2, FF) = ee (with (1.7) ) We now derive (1.11) explicitly. According to the definition of the Lie algebra 0(1,3), any OL € 0(1,3) satisfies the relation =- 7 CL 7 Explicitly this is a, oO 2, ‘ Boz ae 3 T Zig 2% Ue U3 @o Fy 22 423 Zo 23, 232 #33I 925%; 2 %3 I =-| ~! O |/2, e Gi %3 1° o 7! 275 Az, %1r Ge o -! “V] Ge 43, 252 % i which leads to the set of equations Boo Fly =z, = %33 =O fio = tor , C20 = Zoz , %3e = %e3 42, = -Zo , 43,7 -%3 , C32 =~ X23 Therefore a typical matrix O€ € 0(1,3) has the form 7 20, Seg %oa =| °° “22 U3 fo ig. ° ~@a3 fo3 Sz ag We choose a basis of 0(1,3) of the following form ooo oe ooo*°o oeoeco M,:= mf? Oo of 1 oo O- , MoM, Co Cee ootlo Oo-l co Oo °o of oo M:=] ° 2 He Ni=[ 1 2 22 3 of OOo}7 1 Bite tte te oo oO oo oOoo 1 0 Qo Oo | [e909 © e 0 © o © N= le ©°0 Nes oo ee 00 e°o fe oo The three matrices My generate the SO(3) subgroup of $0(1,3) the three matrices Nj, i = 1,2,3 generate Lorentz boosts. As can be shown by explicit calculation, these generators obey the following commutation relations: TMs, Mj] = je Mg CMe, My] = + €eyk Ny [N,, NyI N — £efk Me where Eujk is totally antisymmetric in i,j,k. The matrices M,, i =1,2,3, are antisymmetric, i.e. oT os . MT =- M, whereas the Lorentz boost generators N, are symmetric: Te . NeT = Na We now construct hermitian matrices ée i= Mg 745 1,23 (4M, Jt =- iM =< Mg = de and antihermitian matrices K,i= o ’ Ch) = Toe I= Tool) rN) (K),= Ne (K, = 5 o=-A(N,Jo* -« and all other elements of this matrix are zero. Therefore the explicit form of [on pv] is © i 0 0 [KJ = E 0 0 Oo oO ° ° oO ° K, and K, are given by13 oo1i°0 ooo 8 oo0 0 o 090 0 [Kw] a -i 000 il), ooo 0 oo0°90 -i ooo Hence in this form the six matrices 0% Sete 2 {3 i=1,2,3 form a set of hermitian generators of the Lorentz group. The Lie algebra 0(1,3) describes the Lorentz group locally and is determined by the commutator of the basis elements. Proposition: The generators Mur of the Lorentz group obey the following commutation relation [Mus Me] = (Nee Msge Tee My, - Tua Mus + Isp) (1.13) Proof: Using (1.11) and (1.11') we have [Myv, Moola a ir = Mevleg Mae ~ se Dey Mew) 5 fa “Cpa Tog 7 Tet Wad Yox- Lp 5,7) _ - lee Nee Yar Pee)” (LE Yo Tpys SF) pa To De To lo © log +¢ ~ Tp od x 1973 + Ope Pre faz ) Toa 164 Cie YouTev a) Tat Vax Ina t Tov Vora Trepp)al [Yee Va Tag” lor VP ) Toe Nua Tara Teor Nya) + Ya « (Force Pug ~Tev Gua) lee Nae Pre Tov Tagg) ] cs ‘ Te Toe Cong Pylon? . lenge (Pee Dea ~ Yo« Pag) : ve a Vc 7073 ~ New 7g? ~6 Dy gle) iy et 3 Hk few Tea” Tow Tpys) fh? My 6-Jac +2 Te CM wre )ueg « Wa Mig lag< ag Moca ec a Dropping the matrix indices %,/3, we obtain (1.13). Next we want to rederive (as a consistency check) the commutation relations of the generators Kye wy Starting from the commutator (1.13). We had Mian = Emni dc (1.14) Mo.= 7 Ke (1.15) Equation (1.13) reads for Mae O, ven, waz y i,j = 1,2,3 CMoz, Mord = CR, Ky] (with (1.15) ) ro “loo Me; ae Yo; Mio Leo Mog +e; M,.) =-4£ 155 Mis (using (1.13) ) = ~£Edik dheHence Dre, Rice fayle Lt (1.16) For f= 0, P= i, Q=h, © = 1; 1,5,k = 1,2,3 we have [Moz, Mae J=-LKz Extn Im] (with (1.14), (1.15) ) ~4 Ton Mie ~ Toe ke Tighe tie Mom) “4A DéR Ret OMe R Qe (sth 1.13) ) “(Ea Ken Seghe) Ng =- Sm) oy Hence = gem [ke, Jm]=~* (Eight Sek, ) (1.16") Then, using Extn kem = 2 Sam a7) we obtain (with (1.17) and (1.16') ) Epentatml Ki, Fm] = 245m lke, dm] = 20K InT = -E eu E-Fe%) = 4 Een Kg th Egin ky = 2kEing Ke Hence . : C - Cc Ke, ged = X 3 (A, 2) (1.25) in a vector space V such that 9 A2,%) 9 CA,,%)= 9 As, A,2%*%) ¥'CN,e) = 9 (AS -A-e Infinitesimally we can write (id, = identity in v) af L£ eo 9A, a) Reddy - SO GMO ea? (1.27 where Ogg =—~©g-o are six infinitesimal parameters leading to an infinitesimal Lorentz transformation € Lt (1.26) and Qe denotes four infinitesimal parameters, leading to an infinitesimal translation.22 Now MPS — Me? and PM are generators of Lorentz transformations and translations respectively in the corresponding representation. Consider (using (1.26) ) g'CA2) 9 (A299 CA, 0) = 97,2) 9(A'A, a") g(A-0) 9 (AN, 2) 4 (a-” 'n, n-'e') (1.27') For infinitesimal (Aia'je pt the left hand side gives TA, 2g (A5a)g (A,e) (now use (1.27) ) = GA, Vfddy~ & Opeol ON, La PHYA 2) = ely £ £0faog'A, o) Ms CA, 0) + a g-(A, oP g CA, 0) The right hand oe of (1.27') may be expanded as GUNA, A“a!) = — EMA) MOT ; ae a "pa = ay - EY Ke vr”? o MAT, #4 (NI fi, Pe ardy- Lin NG a. Me Lal AM PE Hence we obtain q7'(A,2) MEYg (A,o)= NN MOm g'(A,0)P* 4 (A,0)= AM, - The first ane states that mee is an antisymmetric tensor operator under Ly 7 the second relation shows that (1.28) PP is a vector operator under if » Now consider infi- nitesimal A € uf 3 for the first of equations (1.28) we obtain 97'(A, 0) M9 (A, 0) = (A 0 MB g (A, 0)= (ddy + Lar @ MOS) MO” (idly - S09, °F) mH tye LEMAR, mP>3 On the other hand As Ne MOT (a4 oh (S94 OF) ME = MBM 4 Of, SE MER SE ov MET = MEY 4 cols MEME Ca¥ MEO = vv ~ es TET Wp, MOM 47 6, gM = a o v ey) M x fye fae +7Fe we M + 4re 02, 9 MES 7 Fee, Ne} ant vy = MOM Selly SHE MO Eng o> ~7P EMEC GM? EFT Hence vr PPP MEY eo pyer [MeSmeyj= —4 (Jeenmeey VAMET, fe 7 TRO MPO Kho 4% [MM MOS ]a-2 Tye meen repay) Thus we have rederived the commutation relation of so(1,3), ive. equation (1.13). The second of equations (1.28) leads to 9-'(A,0) POG (A, 0) = (dy + F@uv MEY) P Cy ~ 4 On Me )24 = PO+ & Ons [mer Pez = Aah (using (1.28) ) CSCy + OF) P™ = pls of PY = pe + 7 ef Ory PY = PEI Oy yP Me 10°C POY = ple Le, {cep 70rPe } N and we obtain ‘i v C MEY POT -ACQ POP 7 PH) (1-29) In summary we see that the Poincaré algebra is given by the following set of commutators CMrv, Meel= -“ Che My Tuo v2 ~ ua Me + Pv M ya) [Mp BI =< C1 Re - oe) fu, mIJ=¢ (1.30) Proposition: P* = PR, Pf#& is a Casimir operator of the Poincaré algebra (1.30), i.e. CMpe, P*J Tae { (1.31) Ch. P*] =o Proof: The second of relations (1.31) is trivial. Hence consider the first. Using (1.29) we have25 [My P77 = [Myo POP,7 =[Muv, PIP, + POL My, BI =[Mpr, 7¢°R IP + POL My, Pod 1 ee Tae Ra Ing Ra Re ~ Pie Chea Po = Tra Fee) oe ae BoB tA 1 Toe fake ne Py te eee =-£ &% PLP, +i SRP, -x ALP, +P, Pe =-<0% BI-LCRR] =o The second Casimir operator of the Poincaré algebra (1.30) can be constructed from the Pauli-Ljubanski polarization vector defined by yer Heel P Wy t= 2 Enye © M (1.32) where Env e@ is the Levi-Civita tensor with o123 = + 1+ Proposition:The Pauli-Ljubanski vector is invariant under translations, i.e. (i) CP, WJ] =0 (1.33a) It can be written (44) w& = CL, PF] (1.33)where r te Euver Meryem and is a vector under it . (iii) [Myy, W.] = -4 Pega o> Paya Me ) (1.34) Proof: (i) Proof of (1.33a): Using (1.32) and (1.29) we have CheyWod = OF; Evers POMTT #3 fysrc lL Pu, POMETY] = a eveet Mer {LPS PCIMe e + POLPT METZ = Lycee Nag POPY OP PE Ter pep & Leupp POP™ — evorp P HAE is Tile £fEyone POPS Evrue? “pe — Eve op POPE ~ EvoppP "pe 3 feta of (1.33b) + CL, PAI ALF Eger s MOMIS, PHD =e fear l Mea Pe PR] $ faare {Mee [ears rede LMP eyo =< Egare IMP CL) (yt PE 7 pr) ~E(7t pbk pt) mre} F Lfgrs 7 Th my Ope27 Bears 7 SE MEP Ee ear Ph 0d Te eye PA MES fF =7eX apsM@BPS+ eee PME Now MY PFap Ey — Cg TPG. yBSp* ) ae with € ae = Eas MBPS at “cas P AM commutes with the generators Pa Mag lies 24 27- [Muv, W*J= 2, Che, W j=e (1.36) Proof: Using (1.34) we have C Mev, wJ = CMpv, we w. 1a) =[Mzy, WeJWe + Wel Mev, wed = ~* ure WwW, » Tue wd Wee WOO wr Yat.) = 4 Wy Whe Whe Wt Whe Wy Wy Wee) = Oo With (1.33) the second of relations (1.36) is trivial,i.e.30 2 Ch. ,W = [hw We WEE, WJ © (with (1.33) ) Proposition: The Pauli-Ljubanski polarization vector has the additional property W, Piso (1.37) Proof: w,P* = Z fuvpe PMC TPH (with (1.32) ) ‘ ¥ = Zepyee PPE MOT SE page PY TPO -££pverP fone (with (1.30) ) = pnp PY re naes Ot Eos PIP MOT + & Envy PYRO & eK PPE = F fuer LP? peqmer = © (with (1.30) ) The representation theory of the Poincaré group has been discussed extensively in the literature using the formalism of induced representations and the concept of little groups. We do not go into any detail here, therefore, and simply quote the following resuits??/34; The unitary (infinite-dimensional) representations of the Poincaré group can be split into three main classes. These are: a) p= pets me o, Ww =~ m’ sist) (1.38) The eigenvalue of the second Casimir operator W* is - m? s(s+1) where s denotes the spin which assumes dis- crete values s = 0, 1/2, 1, ..- ~ From (1.37) one deduces that in the rest frame (PH = (m, ©) ) the zero component of the Pauli-Ljubanski vector must vanish, and the space31 components in the rest frame are given by | W.=t fs0j k pe shh “a such that Te _ W*s-W =-m S 7 (1.39) where ; 1 Stebst*® Sy (1.40) is the spin operator. This representation is specified in terms of the mass m and spin s. Physically a state ina representation (m,s) corresponds to a particle of rest mass m and spin s; moreover, since the spin projection s, can take on any value from -s to +s, massive particles fall into (2s+1)-dimensional multiplets. ii) P= 0, Wao (1.41) In this case W and P are linearly dependent: Wye = AP (1.41a) The constant of proportionality is called the helicity and is equal to +8 where s = 0, 1/2, 1, .-. is the spin of the representation. The time component of W F is W? = FeCTR PR. Mr — oe Pg ed (1.12) ) (1,42) so that (1.41a) aap ea ae ° which is the definition of the helicity of a massless (1.43) particle. Examples of particles which fall into this category are the photon with spin 1 and helicity states : 1,and the neutrino with spin 1/2 and helicity states S42. ue iii) P* = 0, W L. (1.44) This type of representation describes a particle of rest Mass zero with an infinite number of polarization states labeled by the continuous variable (9. These represen-32 tations do not seem to be realized in nature. Remark: For this case the calculations of case i) do not work since we cannot make a transformation to the rest frame. However, We can always transform to a system where Pe = C%,0,0, Po) If G%, is the eigenvalue of W,+ (1.37) implies On PK = We? pe LF PP eee A 2 Thus in this case the eigenvalues of the Casimir operator w can assume any value. 1.3 SL(2,C), Dotted and Undotted Indices 1.3.1 Spinor Algebra We consider the special linear Lie group in 2 dimen- sions with complex parameters su(2,¢) := [Me GL(2,c) | det M= 41} (1.45) A linear representation of this group is a map from SL(2,C) into the automorphism group of a certain vector space F ? This means Mé SL(2,¢) —~ DCM) (1.46) b The reader who wants to refresh his memory of defini- tions of mathematical terms without delving into mathema- tical texts is advised to consult the article by A.S. Sciarrino and P. Sorba 35 which is a "Junior Dictionary" of group theory concepts commonly met in particle physics.33 the automorphism group being defined as the set of linear bijective maps from F to F, the group multiplication being the composition of maps. As usual we demand the represen- tation properties DlAsiezeo) = 1 (1-47) Dem) Dem) = DCM, -M,7, VM, M, € SL(2,¢) where 151,(2,¢) is the unit element of SL(2,C) and Ip the identity map in the vector space F. 5 Let «p be any element of F and [2&3 the canonical ni = Z Ye Kar non (1.48) and representation matrices act on “P in the following way: dimF basis in F. Then dimFé ta Dim) + = Z, a En (1.49) where , imF = 2 Cm) y- tn ret Pn % (1.50) The (D," ()) axe dim F x dim F - dimensional matrices called representation matrices; dim F is called the dimen- sion of the representation. Two representations Dp? are called equivalent if an invertible dim F x dim F - matrix U can be found, such that pcm) supmuy, uéGLCF) a SL(2,C) admits two inequivalent spinor representations, as we shall see. a) The self-representation The self-representation is defined by Demj:= M , YM € SLC2,c) (1.52)34 The dimension of the representation space is therefore two. According to (1.50) elements of the representation space eF transform under the self-representation as d= M8 +g iy A,B =1,2 (1.53) In the literature the elements P6F transforming accor- ding to (1.53) under SL(2,C) are called left-handed Weyl spinors, and this representation is denoted by (1/2, 0)36/37 b) The complex conjugate self-representation The socalled complex conjugate self-representation is defined by DIM:=M*, VME SL(2,C) (1.54) and as in case a) the dimension of the representation space is two. We call the corresponding representation space F . Elements @ of this representation space transform under $L(2,C) according to —/ - eee eee = *)-B 5 = de = CM*); +g, AB=',2 (1.55) where Gg EF . the elements ¥q € F transforming according to (1.55) under SL(2,C) are called right-handed Weyl spinors, and the representation is denoted by (0, 1/2). Left- and right-handed Weyl spinors are related by complex conjugation; i.e. taking “Pa @F then! (4q)* oy pA 6éF (see also (1.177) ) (1.56) It should be observed that this equation does not exhibit the same index structure on both sides. As long as we deal only with Weyl spinors this does not give rise to difficulties. A relation consistent with (1.56) which does exhibit the same index structure on both sides will be given in Section 1.4.4 . The self-representation and its complex conjugate representation are inequivalent, i.e. it is not possible to find a 2x2-matrix C such that M=cmM*c-!35 Proposition: The representation Dim) =:m7!T (1.57) is equivalent to the self-representation (1.52). Proof: We have to show the existence of a matrix € such that éMe's= M7'T, € € c(2,c) (1.58) Let é & -(e Se fae e) (Eps) Ex 22 and “le (Egg )7! CS Pi ae = Zee ~€z, 4 and M My Mrz Ma, Mae Then MoT = —_ ! Moo. cle Ms, ee) dat “Mar My —Min Ma since M € SL(2,C) and so det M = +1. Then we have to show that ! (& o] Mu oe) E22 o até Ex, £22 May Maz = E22 Ex i Meo ~May “Min My36 Evaluating the product of the three matrices we have My €u Sen Mia En Ex, + 2) Se Exe — Maz Er2 Ex, My €e €227 Mia £24 My Ege Myo a2) Mi Ee Ea, + Miz Ey Eg - Mg, Ea tMeeE ye faz + Me Se2 ete)" — Mu Fra & it MizEar 22 Mz, £12 & 2 I 2 f Manz -Mar Mia My which leads to the following set of equations M262 £21 Mu 2q F202. ~ Mya En En, + Mey Se S227 =M,, ote 2 z My fat Enz Mia, Eni + May E22 — M22 £22 € 2 2 + My2 &2 Ee = - Mi, &teE ~My Eu Era + Miz Su Eze ~ M2, FS — -M,, det & 2 ~My Et2 221 # M201 E227 M21 Ee Egat Mg Sen = M, até From the first equation we deduce by equating coefficients of May on both sides7 €y = £22 =e and — §,, &2, = et € From the second 2 €2, = det € and from the third 2 Ey = be E Then a — 828415 EE = Eg i.e. = a -2= f2r We choose E,g= —f so that €2,= 7! and we have (Eqs) = C os (248)! (1.59) where the right hand side defines a matrix with upper indi- ces. Then the matrix € with upper indices is o1r\, T (48) = (4 .) = Ens) (1.60) such that (£48) (2 48)-'s g.27' = tone implies Ol)\fo -I\_ (: ° -1of/{,; of \e tl or in mixed form AB — oA BCc_ EnPeg. = S%c 5 EgpE P= HS T BC c cape =- Sp (1.61) With this index convention, i.e. writing € with upper and £& 7! with lower indices, so that € plays the38 role of a metric, we can write (1.58) AB c = iT )A ESB Mg Seen = (MT) (1.62) M-atT : roe Thus is equivalent to M. Multiplying (1.62) by ca from the left and by €?* from the right we obtain AB DF tT )4_ DF EmaE Myo Ecpe | = €eq (MIT) "DE i.e : A DE Se? Myo &F = Egg (MT) € MeF = fga(M!7) 4g £7F where we used (1.61). Now, taking a hy €F we know from (1.53) that if t= Ma 4g F Exo (M~'T)% EP? y, (using (1.63) ) Multiplying this equation by €£4 srom the left and again using (1.61) we arrive at (1.63) fA! rj. -28 € t, =(M yD f % (1.64) We now define a Weyl spinor with contravariant spinor index by Awa fe AB ¥ iS Eé tn (1.65) Then (1.64) reads 4A pia . (m-!T) B ys (1.66) Proposition: The representation DIM) =: M*T (1.67) is equivalent to the complex conjugate representation (1.54)39 Proof: As in (1.58) we have to search for a 2x2-matrix = such that EM*=e-! i mente (1.68) In the same manner as for (1.58) one can show that > =: AB za( ime and (1.69) =—-! o -l = 2 Eee é (; ° AB In index Saiheahceniial (1.68) is written e48 * Ce. 3 we-IT)A Mego cag = ONS toy Multiplying from the fe by €24 and from the right by eDF ana using AB. = oA. E° £86 o% 2 -- a (1.71) E.-g Be . 2c ABe Si we obtain : *)-B = E52 we-IT)C, -DB (M*)4 Ae (M BE (1.72) Using (1.55) we have the transformation property of dotted Weyl spinors +; fr (m*),8B; = She (METS Sg £78 Z, s. é. Multiplying this equation from the left by € | we obtain EA gt eS upe-I TE, o DB 7 EG, =e eps (me-IT)S #;40 ~iT)E- PBS. = (METERS EOE Pe cette 1.70 ) Defining dotted spinors with contravariant indices by DA AB G psa e 5 B (1.73) we conclude that dotted spinors with contravariant indices transform under M*™ ~'7 according to pla =(M¥IT)AR PB (1.74) Summarizing we have: For any M @ $L(2,C) the matrix M, its complex conjugate M™ , the inverse of its transpose Mm’ ~' and the inverse of its hermitian conjugate (m*)~' all re- present SL(2,C) . Two-component spinors with covariant and contravariant dotted or undotted spinor indices transform under SL(2,C) as follows: a = Ma? ve (1.75a) wtA. (mT) 4g $B (1.75b) a! as P= MQ? PE (1.75) Pan (m*-IT)AS Es (1.754) The raising and lowering of spinor indices has to be understood in the following way: i) Given any spinor which transforms under SL(2,C) in the self-representation, 4 , we can construct a contra~ variant spinor YA which transforms under uw! 7; Prag hB +n =- 4 54 (1.76a) ii)ror a Weyl spinor which transforms under wo! 7, pA ty = S43 ¥® (1.76b)4 iii) Given any Weyl spinor which transforms under $L(2,C) according tom™ , & cele oe Asom TL. BA A= gAB - - -€ pane 3 a tg (1.76¢) iv) For a Weyl spinor in the representation (0, 1/2) which transforms under (MY ~') F222, G8 A aa The raising and lowering of spinor indices is performed with the help of the matrices €, & —¢ which also connect the two equivalent representations M and M'™ . Hence the matrix € plays the role of a metric in the spinor (1.764) space F. Explicitly the raising and lowering of the spinor indices is given by i) pA = £48 Lg lee! Bbga ety = since €!/* 5:41 (see (1.60) ), yrs £78 bg = ctlpa-4F ii) = : ta = Ean PB t : , = €1g eae £2? =i? since E,, = -I, ¢, = £2, b' = ! iii) DAs cA g Fla efSGg= €'* $i = F wnere €!* mrp l (see (1.69) ) pig et G=-%sine ERS -1y Eg P= This is a consistent set of equations. We now define dual spaces F® , F¥* and establish the connection between Y* ana @ . As explained in Sec- tion 1.3.1 we have four types of two-component Weyl spinors transforming according to four different representations of SL(2,C), where M, M' ", anam™, m™ 71T equivalent representations. describe i) Any € F transforms under the self-representation (1.52) as t Pa =M,Bbe (1.53) Such spinors are characterized by a lower index. ii) From (1.57) we know that spinors transforming accor- ding to tA = -IT)A B yas (mT e¢ (1.66) form an equivalent representation of SL(2,C). Such spinors carry upper undotted indices and belong to re + iii) A spinor transforming according to the complex conju~ gate self-representation is given by ~«EF and is characterized by a lower dotted index, =! = = *)- B ee (1.55) ; (M*), 8 Pg iv) Finally an equivalent representation of SL(2,C) is given by spinors which transform according to Piha (met Ag FS (1.74)43 such spinors carry upper dotted indices and belong to F*. A mathematical framework for these different types of spi- nors is given by the following considerations. Considering F as the vector space of two-component spi- nors Yq ,we may construct the dual space F * in the follo- wing way. According to linear algebra, the elements of F* are linear maps ¢ from F to C: b:F—-C such that for all Yé F Pop) i= bd, Ec (1.76e) Hence according to our index convention we may interpret two-component spinors with upper (undotted) indices as elements of the dual space F* , i.e. der, Prerx The asterisk on F% should not be confused with complex conjugation; it is simply the mathematical symbol for the dual space. In addition we know from (1.65) how we can correlate an element of F to the corresponding element of r*. che £-matrix may be considered as a map (4B): F —» F* ty > Pan cABe, The inverse map is, of course, given by the inverse matrix, i.e. the & -matrix with lower indices (Eqg)i F*—> F pA % =Eng $8 Interpreting the composition on the right hand side of (1.76e) as matrix multiplication, spinors with upper undotted indices represent rows and those with lower indi- ces represent columns. In the same way one can consider the space Pas the44 vector space of two-component dotted spinors with upper in- dices PAe ee and we have Oh) Fe > FCF) = a Sec (1.7 6£) Hence a io Per (sts), GA er* Thus for dotted spinors we have to assign to $4 a row and to spinors with upper dotted indices columns. We then have Furthermore the correct description of the transition from P to F*® is given by complex conjugation and multiplica- tion by the matrix @ © (see (1.176) or Section 1.3.3 where the @ -matrices are introduced) (SOAR : F —> F*™ So we set 7 7 ache > a AA Ch) * = ¢A (1.769) The inverse map is easily found to be og (PB) = (1.76h) In agreement with (1.88) (to be given later) we also have —- * > yh = Hs Fe BA and45 Bx PA = PPK oS, These relations show the connection between Yq and pa 5 Thus the four complex numbers Va GA ,a=1,24K = 1, 2, do,in fact,define only four real independent nun- bers which we can take to be real Y Pa A=1,23A = 1,2. We now consider dual maps. From linear algebra we re- call the following results. Let V and W be vector spaces over a field K. For every linear map Piv—~>w we may construct the socalled "dual map" i, * * ps W*_» V by the following prescription: Let «PE W™ so that f is a map ei WwW —> KCeR,C) Then we define the dual map by * = e*(H) = feo This is to be understood in the following way. Applying the left hand side onto a vector v7 @ V we have vEeV P™*Ch) Cc where *C YW) is by construction an element of V* and this element maps € V/ onto the real or complex num- bers. The right hand side of the equation then gives piv pew* ceV EO geew Wea mec Hence the dual map qb * is defined in such a way that the left and right hand sides of the equation give the same element in K when applied to an element of V. Further- more, a linear map q> then implies a Linear dual map ¢*46 It is important to note that if @ is a map from Vv to W, then the dual map c™* maps w* onto v* , i.e. dual maps have the “opposite” direction. Describing qd) and pb* as matrices, we see that if «fp corresponds to a matrix A then the dual map q>* is given by the transposed matrix a™ . vor a proof of this statement we refer to books on linear algebra. With the help of the mathematical concept of a dual map we may explain the transformation properties of the various types of two-component Weyl spinors; within this formalism we find a natural explanation of (1.53), (1.66), (1.55) and (1.74). We start by considering (1.53). Formally Mi? Fo .F Ya— & = Mabon i.e. M is a map from the vector space F to the vector space F. According to the above formalism the dual map is given by M'sE* —> F* Moos mT (the dual map is the transposed matrix). Here we choose as stated earlier the inverse of the dual map instead of the dual map itself as shown in Fig.1. The figure shows that w'™ can be re- expressed in terms of a sequence of maps. We start with P € r*. applying the matrix €4 we obtain the corres- ponding element Pa= EygYBot r. then with =F Pp we obtain the transformed element Yas M4B ¥g= M4, B Eg PC of F. Again applying the metric g AB we finally obtain the transformed element of F*. The result of this compo- sition of maps has to be the same as applying wm’? to wA E@F* vence we conclude = AB (MIT)4q = E49 MC EcD47 Yaz €a0 P? Yh =My® Dy=M,? Enc WF Sse aeeee eee AB Ene E Mott F*—<+—________________ * 7 T 1A ' y M prs ef? Up = eABYC D = EA? Mae Eco HIMT/AR Ye Fig. 1: Directions of dual maps The map M : Fe» F induces a corresponding map in the ° space F which can be constructed from M using the general prescription of complex conjugation, multiplication by the matrix @° and contraction with the metric € . Starting with c 2 B th eMzetg , Yee F we obtain the corresponding transformation in F¥;48 FA = (HAA CH AY* (E°)AA (Cm, By, )* = CE)AA (my, 8)* ph = CED 4A (m4, 2) "ey; Fe using (1.769) and (1-76h). Contracting this equation with the metric £54 ,we obtain the transformation in F , i.e. eos “ pis : CFA (14, 8G Mye 85 FF soit 2" ee wstkes gz; 0? = <8 eA EE) oe sce Dy v Defining 7 cae (M¥)gB = 246 CRY A (M,B)™ Co gg E98 then J . = : e = ¥*)- BaD: ¥, Pe a(M*)5 803 5 ME: FF in agreement with (1.55). The relation between the various Weyl spinors and their respective representation spaces is shown in Fig. 2.49 rsh," do4* VA Paz pr oni Ena AB € pA: ¢g° AA a” Dy= Osa pe* Fig. 2 : Weyl spinors and their respective representation spaces50 Remark: For the matrix m! m2 (M,8) = ; ‘ é SL(2,c) A m, 1 m,* the above definition of ( (m*); 8) implies fas * 2 metize m* 1 m;! m 7 pers 2 efemaemet mef se m1 nobis ml, 7 2° ' Note that the above unconventional definition of M™®™ is due to qur index convention; the connection between a and apa plays a particular role. Again the transformation in F leads to a dual map in the dual space F*. This corresponding dual map can be derived by contraction with the metric. Given Fi a(MYzb Fs and contracting both sides with AB we obtain ss eOes- Fe EAB CM) 5° Se? by = EAB (ng) 8¢ Ege; = hb (M86 ees GF = s48(m 5° £ce P ar(MeIT)Ag pe (1.55) where (met AR p= EACCM*)- Degs MNT: Ey FH51 This is in agreement with (1.74). Again for ME& SL(2,C) with m! m2 B) _ 1 1 (M,8) = the elements of the matrix M¥~17 are given by t a m, m, ? m2 - x2 Reis ? msi im, ‘. rt 2 2. wm i mos In deriving the matrix M™ we see that we cannot describe complex conjugation as a linear map from F to F . This can, of course, also be shown explicitly by considering a similarity transformation as in the previous equivalence proofs, and showing that the appropriate matrix does not exist. (We shall see later that such a linear transfor- mation exists in the four-dimensional Dirac formalism). Therefore F and F define two inequivalent representation spaces for SL(2,C). On the other hand elements of the corresponding dual spaces F*and ?can be obtained with the metrics E4p and=&fZe respectively. These are matrix representations of linear maps and therefore F and pe B*, are equivalent representation spaces as well as F and52 1.3.2 Calculations with Spinors We now consider some explicit calculations with spi- nors and derive several useful formulae. 1) Proposition: The quadratic form (WX) = PAX (1.774) is invariant under transformations of SL(2,C). Proof: We have to show Ce (Y'x') = Consider (using | (1.75a) aa x) 75b) ) (YX = wptAn’, 17 B (M-'T)Ag PBM,CX, Be m- PE CM~')g4 Mas Xe = Bb bgo%Xe = $8xp Similarly = CX) paisa a x4 ii) Proposition: The quadratic form CBX):= VR X (1.77) is invariant under SL(2,C). Proof: We have (using (1. es and (1.75a) ) (BX) = By HA (10), BBs (eit), FE ®, os (maTIA, x¢ i ¥, Ze Tz i i N n ti a53 =($xX) (using (1.77b) ) Remark: For calculations with Weyl spinors it is a useful convention to sum undotted indices from the upper left to the lower right ("north-west to south-east") and dotted in- dices from the lower left to the upper right ("south-west to north-east") as indicated in (1.77a) and (1.77b). Thus we write i (ox) BA Kq na (BE) = BAR* G70) However, expressions such as (1.76b), 1.e. Pas <4 yB show that in general summations in the other directions are also possible. iii) Postulate: We postulate that the components of spinors are Grassmann variables (also called Grassmann numbers or Grassmann parameters) which are not to be confused with anticommuting operators such as the spinor charges Q, to be defined later; i.e. we demand £44, P8}= 5 ashy at v4 P8he0 ind al (1.79) {%4, KEQS{KA, KS HLA, LB feo If we require the PA "s to anticommute, an expression iike(p W) does not vanish. Thus (pp) gad, (with (1.77) ) E45 Y, $, (with (1.76a) ) = ef t,%, tet ty, 4% aa b,, sixe Le -EKl*o4r/ I£ the components Ya were required to commute, i.e. if C4, W7 = 0, ten (peo. stead we assume that i] rT54 the fq 's anticommute with each other and with other Grassmann variables such as fermion fields and spinor charges. This requirement is, of course, nothing but an application of the spin statistics theorem: half-integer spin quantities obey Fermi-Dirac statistics whereas inte- ger spin quantities obey Bose-Einstein statistics. iv) Proposition: For anticommuting Y's the quadratic form (1.77) is symmetric on exchange of spinors, i.e. CPX) = Cx) (1.80) Proof: We have CPx) = pA Xa (using (1.77) ) “Xa (using (1,79a) ) =~ Egg xPerc ye = — yBroT AC xB) EE = xB Soy (using (1.61) ) = xB the (using (1.77) ) = (x) (using (1.76a,b) ) v) Proposition: The quadratic form for dotted spinors, i.e. (1.78), is symmetric on exchange, i.e. te (PX) = (x ®) (1.81) ig (P and X are Grassmann variables. Proof: We have ($x) = ye ® 7 (using (1.78) ) =-% iG; (using (1.79b) ) =-~cf8y, Sie ge (using (1.76¢,4) )N N vi) Proposition: If @ is @*:= (ee) 3*:= (68) and hence (@) Oo ete. Proof: We have 6° = (00) = 08, = 042,08 = 06'S. 07 +07&,0' = -6'67+e670' =-26'6* -20'e* 267 ©; 55 (using (1.71) ) (using (1.78) ) a Grassmann variable (1.82) (asing (1.77) ) (using (1.76b) ) (using (1.79) ) (Observe:" @ squared" on the left, component "@ two" on the right in our notation!) Similarly 5.64. 57485. = Qe = O:E og using first (1.76c) and then (1.69) . 26;0,56 vii) Proposition: For Grassmann variables @ and @ we have e498 =-~£2%8 (ee) led Hos = zB cea) (1.83b) GAGb = 4 E48 (66) (1.836) tle ps G3 =-ze jp COO) igen Proof: We shall see later (cf. (1.129c,d)) that De _ Dd Dec (1.129¢) Ege" = Ao SB°- 5S ‘i and tee 3 - 2 ~2 9 Pn 5-6 S29 5-0S5-¢ eAa€ A 58°" OSB (1.1294) These formulae are needed in proving (1.83 a,b,c,d). i) Consider the expression on the right hand side of (1,.83a) = - £€48 (90) =-£ £48 6° a, =-$e4F £07 eC e? ¢ e-£(E BAL 4bp%e ‘@2 =-§ (e89%- e408) = e468 ii) Consider the expression on the left hand side of (1.83b) + S|} O88 *facEap OO” =-— $eqgcEav€E My is given by the following trace relation