UNITARY REPRESENTATIONS OF THE POINCARE GROUP AND RELATIVISTIC WAVE EQUATIONS Y.OHNUKI (Nagoya University) Translated by S.KITAKADO T.SUGIYAMA (NAGOYA) World Scientific Singapore * New Jersey © Hong Kong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. UNITARY REPRESENTATIONS OF THE POINCARE GROUP AND RELATIVISTIC WAVE EQUATIONS Copyright © 1988 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 9971-50-250-X Printed in Singapore by JBW Printers & Binders Pte. Ltd.Preface to the English Edition This book was written in 1975 and published in Japanese in September of 1976. Since that time a lot of new developments have been made in the theory of elementary particles as seen for example in superstring models. In spite of this the role of the Poincaré group is still very fundamental in understanding relativistic particles. At least in this sense the publication of an English version could be worthwhile for students who are learning the theory of elementary particles. Though nothing new has been added, some trivial errors in the original book have been corrected through the translation. T would like to thank Prof. K. K. Phua who strongly recommended the publication of the English version. I am also extremely grateful to Drs. T. Sugiyama and S. Kitakado who have kindly carried out the laborious work of translation into English. January, 1988 Yoshio OhnukiPreface to the Original Edition ‘This book is based on my lectures for graduate students of physics, and was written with a considerable amount of addition. It is intended also for undergraduate students and readers in other fields. An attempt to unify quantum mechanics and special relativity was made by Dirac toward the end of the 1920’s, and the famous Dirac equation for an electron, together with relativistic wave equations for particles of finite mass and of arbitrary spin, was discovered. The studies along this line, also including massless particles, have been developed by many authors thereafter. However, the waves obeying this kind of equations are different from those of the first Dirac’s idea, and do not represent the probability amplitudes by themselves. For example, the probability wave function for a photon is neither the electric nor magnetic field which satisfies the Maxwell’s equations. In a sense, the state vector of a photon lies behind the Maxwell’s equations. Other particles are also in similar situations. In this sense, the so-called relativistic wave equations must be regarded only as indirect representations for the description of one-particle probability waves, and the forms of equa- tions themselves have a direct connection to relativistic matter waves, i.e., the equations of field theory. On the other hand, the idea in which quantum states of relativistic parti- cles are formulated directly without the use of wave equation, was proposed by Wigner (1938). From the point of view in which states of a free parti- cle are given by unitary irreducible representations of the Poincaré group, , the group formed by translations and Lorentz transformations in thevill Preface to the Original Edition Minkowski space, Wigner considered the problem of representations of the group, and revealed its structure. Although this laborious work, which was written as a mathematical paper on the representation of group, does not put much emphasis on concrete relations to physics, the presented argument has a significant meaning in understanding the relativistic picture of particles in quantum mechanics. As a matter of fact, if we get all unitary irreducible representations of the Poincaré group, we do have a complete knowledge of every free particle states and their behavior. It is a surprising fact that a sim- ple framework as the Poincaré group, when unified with quantum theory, fixes our possible picture of particles so severely and without any exception. As a natural consequence it must be recognized that Wigner’s idea had great in- fluences on basic understanding of subsequent relativistic quantum mechanics or elementary particle theories. Nevertheless, to the author’s knowledge, there is no textbook discussing the details of the relation between physics and unitary representations of the Poincaré group. Thus it may be meaningful to publish a book emphasizing this problem. At least to the author, this was one of the motivations of writing this text. The text is divided into three parts. The first five chapters cover the theory of unitary representations of the Poincaré group and their concrete contents, the connection to covariant formalism is considered in Chapters 6 and 7, and Chapter 8 is devoted to discussions of relations with quantum field theory. This last chapter has the smallest number of pages because it restricts itself only to the fundamental problems. Applications of field theory to the elementary particle theories are beyond the scope of this book, and are expected to be found in other appropriate textbooks. Needless to say, this is not a book of pure mathematics. So it does not take account of generality or exactness which are not considered to be necessary from a physical point of view, rather it takes care of developing the argument. as specifically as possible, Consequently, it involves somewhat tedious calcu- lations and slightly complicated handling of expressions. This is unavoidable because, in natural science, understanding the specific cases seems to be very important in any abstract argument. I have not chosen such an easy-going way of writing as rearranging contents collected from many literatures. Be- cause each literature has its own limitations, any effort of arrangement will not lead to a self-contained textbook. In addition, there were several prob- lems which were not covered by the existing literatures. Therefore, I argued to some extent in my own way, and derived each formula by myself. So I am afraid of unexpected errors or misunderstandings.Preface to the Original Edition ix A few sections in the primary plan had to be omitted because of lack of space. But the whole volume was written carefully so that a self-consistent style was kept till the end. I would like to accept reader’s comments as to whether the aim of the author has been accomplished, and hope to have an opportunity to improve the text by taking away a superfluity and making up an insufficiency. Further, notations used in the text follow the author’s private usage. Of course, this is not so essential, and readers should understand them in their own styles. Professor Mikio Namiki, who invited me to write this book, read the man- uscript and made valuable comments. Mr. Hiromi Katayama of Iwanami publishing Co. took on the management during the preparation of this book. I would like to express my sincere gratitude to these people. March 1976 Yoshio OhnukiContents Preface to the English Edition Preface to the Original Edition Chapter 1 Introduction §1.1 Transformation and Invariance §1.2 Poincaré Group and Free Particles Chapter 2. Lorentz Group §2.1 Double-valued Representations §2.2 Spinor Representations §2.3 Infinitesimal Transformations Chapter 8 Irreducible Representations of the Poincaré Group §3.1 Translational Transformations §8.2 Lorentz Transformations §3.3 Little Groups §3.4 Inreducible Representations Chapter 4 Unitary Representations of Little Groups $4.1 Rotation Group §4.2 Two-Dimensional Euclidean Group §4.3 Lorentz Group vii 26 26 31 33, 36 39 39 41 ATxii Contents $4.4 Three-Dimensional Lorentz Group §4.5 Classifications of Free Particles Chapter 5 Wigner Rotations $5.1 Particles with Finite Mass $5.2 Particles with Zero Mass 35.3 Particles with Imaginary Mass 35.4 Angular Momenta of Massless Particles Chapter 6 Covariant Formalism I — Massive Particles §6.1 Particles with Spin 0 §6.2 Dirac Particles $6.3 Particles with Higher Spin 36.4 Generalized Bargmann-Wigner Equations §6.5 Matrices §6.6 Discrete Transformations (a) Space Reflection (b) Time Reversal (c) Charge Conjugation §6.7 Other Covariant Formalisms (a) Spin 1 Particle (b) Spin 3 Particle (c) Generalization Chapter 7 Covariant Formalism Il — Massless Particles §7.1 Particles with Discrete Spin §7.2 Discrete Transformations §7.3 Covariant Inner Products (a) Spin 1 Particle (b) Spin $ Particle and Others $74 Particles with Continuous Spin (a) Particles in a Single-Valued Representation (b) Particles in a Double-Valued Representation Chapter 8 Quantized Fields §8.1 Quantum Theory of Matter Waves $8.2 Harmonic Oscillators $8.3. Scalar Fields 52 57 59 59 64 70 72 81 82 85 97 106 112 113 114 iit 118 119 121 124 135 135 143 148 148 151 156 156, 166 170 170 176 184Contents §8.4 Spin and Statistics §8.5 Poincaré Group and Free Fields References and Bibliography Index xiii 191 199 209 211Chapter 1 Introduction §1.4 ‘Transformation and Invariance Let us denote the state vectors (or simply states), which describe a sys- tem according to the rules of quantum mechanics, as |A), |B),.... For an arbitrary complex number ¢,c|A) is also a state vector, and the well known superposition principle for the state vectors indicates that 1¢) =|4)+|B) (1.1.1) is also a state of the system. An inner product ( A| B) is defined for arbitrary states |A) and |B). The product is a complex number fixed by these two states, and satisfies (A|B)* = (BIA), (1.1.2) (Ale|B) = ¢( A|B), (1.1.3) (A|A) 20, (1.1.4) and for |C) of (1.1.1), (D|C) =(D|A) +(D|B). (1.1.5) Here the symbol * denotes complex conjugate, and the equality in (1.1.4) stands only for |A) = 0. The norm of a state | A) is given by (A| A).*) *)We use this definition in this text. In mathematics, however, the norm is usually defined by f(A] A).2 Unitary Representations of the Poincaré Group and... Another important property of the state vectors is that the probability of getting a particular result in an observation is expressed in terms of them. That is, when a physical quantity is observed in a state | A) of the system, the transition probability from | A) into an eigenstate |B) is given by Ppa =|(BIA)P , (1.1.6) where | A) and | B) are both normalized ((A|A) = (B|B) =1). Assuming these properties for state vectors, let us consider transforma- tions of the states. To do this, let state vectors | A’),| B’),... correspond to |A),|B),... respectively, and consider replacements |A)— 14‘), |B) 18), (1.1.7) Needless to say, the transformed states must satisfy (1.1.2)~-(1.1.5). Further requirement of invariance of the theory under the transformation leads, at least, to invariances of the superposition principle and of the definition of probability.) In other words, if |C’) is given by (1.1.1) with arbitrary | A), and |B), then [C') =|4)+ |B), (1.1.8) KBLAYP = (BLADE, (1.1.9) must be satisfied. At this stage, either the relation (BIA) =(B"|4') (1.1.10) (B|A)=(A'|B') (1.1.11) can be derived in the following way. Let |A) = |B) in (1.1.9), then (1.1.4) gives (A|A) = (A’|A’). Since this relation holds for any | A), setting (C|C) = (C"|C") and substituting (1.1.1) and (1.1.8) for |C) and |C"), we get Re(B|A) =Re(B'| A’), (1.1.12) “)This may be a definition of the invariance in a broader sense.Introduction 3 where Re(B| A) is the real part of (B|A), and Im(B| A) in the following equation is the imaginary part. Substituting (1.1.12) into (1.1.9) we immedi- ately get Im(B| A) = Im(B’| 4’) (1.1.13) or Im(B|A) =—Im(B’|A’) . (1.1.14) Equations (1.1.12), (1.1.13) and (1.1.14) are nothing but Eqs. (1.1.10) and (1.1.11). Therefore, if an operator D, which relates |A),|B),... to] 4’),|B’),... is introduced by |4')=D|A), |B')=D|B), (1.1.15) then D is restricted to a unitary operator in the case of (1.1.10) by the in- variance of the inner product, and to an anti-unitary operator in the case of (1.1.11). A transformation of state vectors may be considered as a result of some external action to the system that causes the transition of the states, and it may also be considered to be caused by a transformation of the framework describing the system, e.g., a change of the direction of the coordinate axis.*) In either case, however, if a set of transformations {a,b,c,...) leaves the theory invariant, the state vectors are transformed by the corresponding unitary or anti-unitary operators D(a), D(b), D(c),... . In particular, when {a,b,¢,...} forms a group, and when ab = c, let us assume the following relation: D(a)D(b) = D(c) . (1.1.16) Then we call D(a), D(d),... representations of the group G, and refer that the state vectors transform according to the group G. If no change of state vectors is considered as one of the transformation, i.e., the identity trans- formation, its representation can be expressed by 1.°*) This is, of course, a unitary representation. If G is a continuous group and if any element g of *)in order to avoid confusion, we consider here that state vectors are transformed, but dynamical variables are not. **) This does not mean that there exists no other representation of the unit element of G than 1. Because there is a case where D(e) = 1 does not hold for € given by a succe: application of transformations on the identity. The so-called multi-valued represent belongs to this case (cf. §2.1).4 Unitary Representations of the Poincaré Group and G is obtained by successive multiplication of infinitesimal transformations on the identity, D(g) is written as a product of corresponding representations of the infinitesimal transformations. Hence, according to the continuity of rep- resentation, we can conclude that D(g) is always unitary. This means that an anti-unitary representation is possible only in a case with some discontinuous transformation. In quantum mechanics, the transformation of time reversal is a typical and familiar example of such a transformation (cf. §6.6). §1.2 Poincaré Group and Free Particles In the world of elementary particles, the theory of special relativity is necessary to describe the motion of particles, because the motion frequently approaches the light velocity. Quantum mechanics and special relativity are two fundamental prerequisites for the following arguments. Before entering on the main subject, let us give some remarks on the notation which will be used in the text. Let ty = (21,22,29,24) and yy = (vi,y2,¥9,ya) be 4-vectors in the 4 Lorentz space, and write an inner product )> zyy, a8 at Tye = xy + Lava = XY — Doyo» (1.2.1) where the boldface letter x denotes a 3-vector whose components are 2; (i =1,2,3), and xy denotes an inner product of x and y defined by (1.2.2) further z4 and yg are written as t=t20, yw=tyo- (1.2.3) In the case of general 4-dimensional tensor, the index 4 also can be changed to 0 by taking the imaginary factor i outside. The length of a 3-vector x is written as Ix] = vo2 - (1.2.4) The expression x x y is a vector product of x and y, and its components are 3 (xy) = Do sina» (1.2.5) aIntroduction 5 where €,,, is completely antisymmetric with respect to the indices #, 7 and k, and €193 = 1. We adopt the natural unit in which the Planck constant h = h/2n and the light velocity ¢ are both 1: hac=l. (1.2.6) When z, describe the position of a point in 4-dimensional space-time, 2 indicates the time ¢. Although in this text both zp and ¢ are used to describe the parameter for the time, these are of course the same. For other symbols we shall mention where they appear. If Lorentz transformed z, are written as 2), = Ayp2y, the relation 21,2), = Ty2y leads to AwAyp = bp + (1.2.7) This formula implies that the transposed matrix AT of the 4x4 matrix A is equal to the inverse matrix A~1, and thus it can be rewritten as AvpApu = bps (1.2.8) where As; (i, 7 = 1,2,8) and Aga are real, and Ajg and Aas are purely imaginary numbers. . Hereafter, A is restricted to such A that can be obtained by successive applications of infinitesimal Lorentz transformations on the identity transfor- mation. From (1.2.7) we find that the determinant of A satisfies (det(A))? = land that A2, = 1- x (Ava)? > 1. By continuity of the transformation, A is further restricted to oe det(A) =1, (1.2.9) Awl. (1.2.10) The Lorents transformations defined by (1.2.7)-(1.2.10) obviously form a group, which is called the continuous Lorents group or the orthochronous proper Lorentz group. We call it simply the Lorentz group in the text. Con- sidering the condition (1.2.7), we can easily find that such A is character- ined by six real parameters, and A varies with continuous variation of these parameters. For the discussions of the world of elementary particles this is, however, not sufficient, and another transformation which translates the coordinates as i (1.2.11)6 Unitary Representations of the Poincaré Group and... has to be introduced. This transformation combined with the previous Lorentz transformation obviously forms a continuous group, which is called the proper Poincaré group or the continuous Poincaré group. We call it the Poincaré group for simplicity hereafter. Taking account of (1.2.11) we find that the Poincaré group is a continuous group characterized by ten real pa- rameters. If the theory is invariant under this transformation, and if the state vectors are transformed according to the Poincaré group, their representations must be unitary according to the argument of the previous section. In other words, the state vectors are transformed according to a unitary representation of the Poincaré group. The following discussion will be focused on the case where the states particularly represent free particles. At an instant of time sufficiently before or sufficiently after the scattering, we can suppose that particles move separated from each other. Then the interaction between the particles becomes so small that we cau neglect it completely, and the particles behave as free particles without getting any effect from each other. As a result, the naming of each particle, the electron or the deuteron for example, becomes possible in such a system. However, it must be noticed that such a naming of a one-particle state is usually based on the invariance of it under the coordinate transformations of the Poincaré group. For example, when there is a free electron, we call the object an electron either in the Lorentz transformed reference frame or in another reference frame which has a different origin. That is, unitary representations of the Poincaré group play a role of connecting various state vectors of the same free particle. In this sense, if we consider a superposition of two state vectors which cannot be connected by any transformation of the Poincaré group, we cannot regard it as a state vector of a free particle. In other words, any two state vectors of a given free particle can be transformed mutually by an appropriate transformation of the Poincaré group. Assuming that a free particle is such an object, let us proceed to the following discussions. However there may remain still a problem whether we consider free particles in a larger transformation group containing discrete transformations such as the space reflection which connects the right-handed frame with the left-handed frame. But in the theory of special relativity, the requirement of the invariance under discrete transformations does not exist. Therefore we consider free particles in the above sense with weak restriction, and will discuss later their behavior under the discrete transformations. If U is a unitary operator and D is a unitary representation of the Poincaré group UDU' is also a unitary representation of the group, where the symbol } denotes Hermitian conjugate. Two representations D and UDUt are called