W. Greiner B.Miuller — J. Rafelski Quantum Electrodynamics of Strong Fields With an Introduction into Modern Relativistic Quantum Mechanics With 258 Figures Springer-Verlag Berlin Heidelberg New York TokyoProfessor Dr. Walter Greiner Professor Dr. Berndt Miller Institut fir Theoretische Physik der Johann-Wolfgang-Goethe Universitat D-6000 Frankfurt/Main, Fed. Rep. of Germany Professor Dr. Johann Rafelski Institute of Theoretical Physics and Astrophysics, University of Cape Town Rondebosch 7700, South Africa Editors Tullio Regge Istituto di Fisica Teorica Universita di Torino, C. so M. d’Azeglio, 46 Wolf Beiglbock 1-10125 Torino, Italy Institut fir Angewandte Mathematik Universitat Heidelberg Elliott H. Lieb Im Neuenheimer Feld 294 D-6900 Heidelberg 1 Fed. Rep. of Germany Department of Physics Joseph Henry Laboratories Princeton University Princeton, NJ 08540, USA Joseph L. Birman Walter Thirring Department of Physics, The City College Institut fur Theoretische Physik of the City University of New York der Universitat Wien, Boltzmanngasse 5 New York, NY 10031, USA A-1090 Wien, Austria ISBN 3-540-13404-2 Springer-Verlag Berlin Heidelberg New York Tokyo ISBN 0-387-13404-2 Springer-Verlag New York Heidelberg Berlin Tokyo Library of Congress Cataloging in Publication Data. Greiner, Walter, 1935~. Quantum electrodynamics of strong fields, (Texts and monographs in physics) Includes bibliographies and index. 1. Quantum electro- dynamics. 2. Heavy ion collisions, 3. Field theory (Physics) I. Milller, Berndt. 11. Rafelski, Johann, III. Title. IV. Title: Strong fields, V. Series, QC680.G73 1985 537.6 84-26824 This work is subject to copyright. All rights are reserved, whether the whole or part of the material is concerned, specifically those of transiation, reprinting, reuse of illustrations, broadcasting, reproduction by photocopy- ing machine or similar means, and storage in data banks. Under § 54 of the German Copyright Law, where copies are made for other than private use, a fee is payable to “Verwertungsgesellschaft Wort”, Munich. © Springer-Verlag Berlin Heidelberg 1985 Printed in Germany The use of registered names, trademarks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and there- fore free for general use. Typeseting: K+V Fotosatz, Beerfelden Offset printing and bookbinding: Konrad Triltsch, Graphischer Betrieb, Warzburg, 218373130-548210Preface The fundamental goal of physics is an understanding of the forces of nature in their simplest and most general terms. Yet there is much more involved than just a basic set of equations which eventually has to be solved when applied to specific problems. We have learned in recent years that the structure of the ground state of field theories (with which we are generally concerned) plays an equally funda- mental role as the equations of motion themselves. Heisenberg was probably the first to recognize that the ground state, the vacuum, could acquire certain prop- erties (quantum numbers) when he devised a theory of ferromagnetism. Since then, many more such examples are known in solid state physics, e. g. supercon- ductivity, superfluidity, in fact all problems concerned with phase transitions of many-body systems, which are often summarized under the name synergetics. Inspired by the experimental observation that also fundamental symmetries, such as parity or chiral symmetry, may be violated in nature, it has become wide- ly accepted that the same field theory may be based on different vacua. Practical- ly all these different field phases have the status of more or less hypothetical models, not (yet) directly accessible to experiments. There is one magnificent ex- ception and this is the change of the ground state (vacuum) of the electron-posi- tron field in superstrong electric fields. A whole new area of physics has devel- oped when it became clear in the late 1960s and early 1970s that the vacuum of QED in supercritical fields becomes charged, i. e. it undergoes a phase change from a neutral to a charged vacuum. Most important is the fact that experiments have been systematically proposed and performed to test these fundamental ideas in the laboratory. The proposed existence of superheavy quasimolecules and fi- nally of giant nuclear systems were milestones in this development, culminating in the discovery of sharp positron lines which probably signal the spontaneous decay of the vacuum and — at the same time — the existence of rather long-lived giant nuclei. It was a breathtaking development both in theory and experiments over the course of the last two decades, We shall trace the essential steps of this scientific endeavour in this book. Starting with an overview of the physical, historical and philosophical implica- tions (Chap. 1), a review of necessary elementary theoretical prerequisites (Chaps. 2, 3, 4), the Klein paradox in its original historical version (Chap. 5), we then develop systematically quantum electrodynamics of strong fields, always fo- cussing on the non-perturbative aspect of the issues (Chaps. 6, 8— 10). To attract the attention of graduate students also, Chap. 7 includes a synopsis of “or- dinary” quantum electrodynamics, i. e. QED of weak fields.vi Preface Formulation of the theory of quasimolecules and heavy-ion collision dynam- ics can be found in Chaps. 11 und 12. No theory stands by itself in physics, but must be proven right or wrong by experiments. The courageous and ingenious ex- periments performed over the course of many years are summarized and dis- cussed in Chap. 13. Theory is treated further in Chaps. 14-16. The difficult problem of vacuum polarization is discussed in detail in Chaps. 14 und 15, and many-body effects in strong fields, including the self-energy in high Z atoms, in Chap. 16. If bosons are bound in strong and supercritical potentials the phenomena change quantitatively and to some extent also conceptually. Chaps. 17—19 are devoted to these topics. Finally, in Chaps. 20 and 21, supercritical phenomena in other areas of physics, like superstrong gluon fields and supercritical gravitation- al fields, are presented to give an impression of the richness and variety of the applications of the idea of a supercritical vacuum. We also tried to exhibit a number of details of the theoretical developments in order to ease the learning process for students. Such sections are indicated by a vertical grey line in the margin. The last 15 years during which this research has been going on were most re- warding. We started with vague theoretical ideas which became more and more concrete and quantitative until they finally became real physics through experi- ments which were then compared in detail to theory. It is a pleasure to thank friends and colleagues, above all J. Reinhardt and G. Soff, and also P. Gartner, U. Heinz, J. Kirsch, U. Miller, W. Pieper, P. G. Reinhard, T. de Reus, P. Schliiter, R. K. Smith, D. Vasak, K. W. Wietschorke and several others who made important contributions. Without their energy, enthusiasm and reliable work, the complex theory of these involved processes would not be as developed as it is. We should equally like to thank a number of experimentalist friends and col- leagues, in particular J. Greenberg, P. Kienle, and W. Meyerhof, and also H. Backe, H. Bokemeyer, K. Bethge, F. Bosch, T. Cowen, A. Gruppe, E. Kanke- leit, C. Kozhuharov, D. Liesen, D. Schwalm, P. Vincent and their associates for their prodigous work and dedication. They followed a courageous path against many stumbling blocks, set backs and outside discouragements. The close contact with them was of mutual benefit and most gratifying. Finally we should like to thank our technical assistant B. Utschig and our secretaries R. Lasarzig, L. Schubert and J. Parsons for their patience and steady help in preparing the manuscript. One of us (B. M.) would also like to thank the secretarial staff of the Department of Physics at Vanderbilt University and of the Kellogg Radiation Laboratory at the California Institute of Technology, where part of the manuscript was written, for their help. Frankfurt am Main and Cape Town, Walter Greiner May 1985 Berndt Miller . Johann RafelskiContents 1, Introduction .. 2 1.1 The Charged Vacuum 1,2 From Theory to Experimental Verification . 1.2.1 Superheavy Quasimolecules ......... 1,2,2 Nuclear Sticking .. 1.2.3 K-Shell Ionization . 1.3 Theoretical Developments . 1.4 Historical Annotations on the Vacuum . . 1.4.1 The Concept of Vacuum ..... 1.4.2. The Vacuum in Strong Fields . 1.5 The Vacuum in Modern Quantum Physics 1.5.1 Pion Condensation ......... 1.5.2 Strong Gravitational Fields .. 1.5.3 Vacuum Structure of Strongly Interacting Fermions and Bosons . Bibliographical Notes . The Wave Equation for Spin-1/2 Particles 2.1 The Dirac Equation ..... 2.2 The Free Dirac Particle 2.3 Single-Particle Interpretation of Plane (Free) Dirac Waves 2.4 The Dirac Particle Coupled to Electromagnetic Fields — Non-Relativistic Limits and Spin of the Dirac Equation 2.5 Lorentz Covariance of the Dirac Equation . 2.5.1 Formulation of Covariance (Form Invariance) ............ 2.5.2 Determining the $(@) Operator for Infinitesimal Lorentz Transformations 2.5.3 The $(@) Operator for Finite Lorentz Transformations . 2.5.4 Finite, Proper Lorentz Transformations ......... 2.5.5 The $ Operator for Finite Lorentz Transformations 2.5.6 The Four-Current Density 2.6 Spinor Under Space Inversion (Parity Transformation) .. 2.7 Bilinear Covariants of Dirac Spinors 2.8 Gauge Invariant Coupling of Electromagnetic and Spinor Field .. Bibliographical Notesvul ‘Contents 3. Dirac Particles in External Potentials 3.1 A Dirac Particle in a One-Dimensional Square Well Potential . 3.2 A Dirac Particle in a Scalar, One-Dimensional Square Well Potential 3.3 A Dirac Particle in a Spherical Well . 3.4 Solutions of the Dirac Equation for a Coulomb and a Scalar 1/r Potential .. 3.4.1 Pure Scalar Potential 3.4.2 Pure Coulomb Potential . 3.4.3 Coulomb and Scalar Potential of Equal Strength (@’ = : 3.5 Stationary Continuum Waves for a Dirac Particle in a Coulomb Potential Bibliographical Notes . 4, The Hole Theory .......... 2.20. c ccc c cece eect e nent n eee 4.1 The “Dirac Sea” o 4.1.1 Historical Context . 4.2 Charge Conjugation Symmetry 4.3 Charge Conjugation of States in External Potential . 4.3.1 HistoricalNote .............- 4.4 Parity and Time-Reversal Symmetry 4.4.1 Parity Invariance 4.4.2 Time-Reversal Symmetry .. Bibliographical Notes 5. The Klein Paradox . 5.1 The Klein Paradox in the Single-Particle Interpretation of the Dirac Equation 5.2 Klein’s Paradox and Hole Theory Bibliographical Notes 6. Resonant States in Supercritical Fields 6.1 Resonances in the Negative Energy Continuum 6.2 One Bound State ee into One Continuum . 6.2.1 Filled K Shell . hee 6.2.2 Empty K Shell .. 6.3 Two and More Bound States Imbedded in One Continuum . 6.4 One Bound State Imbedded in Several Continua . 6.5 Overcritical Continuum States ... 6.5.1 Continuum Solutions for Extended Nuclei 6.5.2 Comments on the Point Nucleus Problem for Za > |x|. 6.5.3 The Physical Phase Shifts .............seeeeeeee 6.5.4 Resonances in the Lower Continuum for Z > Zer 6.5.5 The Vacuum Charge Distribution . 6.6 Some Useful Mathematical Relations . 6.6.1 A Different Choice of Phases .. Bibliographical Notes 59 59 67 1 719 83 84 86 86 1 92 92 95 102 104 104 105 106 1114 112 412 117 121 1221 10. Contents Quantum Electrodynamics of Weak Fields . 7.1 The Non-Relativistic Propagator . 7.2 The S Matrix .. 7.3 Propagator for Electrons and Positrons 7.4 Relativistic Scattering Theory .. Bibliographical Notes The Classical Dirac Field Interacting with a Classical Electromagnetic Field — Formal Properties . 8.1 Field Equations in Hamiltonian Form . 8.2 Conservation Laws ............05 8.3 Representation by Energy Eigenmodes 8.3.1 Time-Independent Potentials 8.3.2 Explicitly Time-Dependent Potentials 8.4 The Elementary Field Functions . Bibliographical Notes . Second Quantization of the Dirac Field and Definition of the Vacuum 9.1 Canonical Quantization of the Dirac Field 9.2 Fock Space and the Vacuum State(I) ... 9.3 Poincaré Invariance of the Quantum Theory 9.4 Gauge Invariance and Discrete Symmetries . 9.5 The Vacuum State (II) ... 9.6 The Feynman Propagator 9.7 Charge and Energy of the Vacuum (I) . 9.8 Charge and Energy of the Vacuum (II) . 9.9 Appendix: Feynman Propagator for Time-Dependent Fields ... Bibliographical Notes . Evolution of the Vacuum State in Supercritical Potentials ........... 10.1 The In/Out Formalism 10.2 Evolution of the Vacuum State : 10.3 Decay of a Supercritical K Vacancy — Projection Formalism ... 10.4 Decay of the Neutral Vacuum — Schrodinger Picture . 10.5 The Vacuum in a Constant Electromagnetic Field ... 10.6 Quantum Electrodynamics in Strong Macroscopic Field: 10.7 Klein’s Paradox Revisited + Bibliographical Notes ......... 0.0.0 cece cee e ener eee etree etn eee Superheavy Quasimolecules .. . 11.1 Heavy-Ion Collisions: General Remarks . 11.2 The Two-Centre Dirac Equation . 11.3 The Critical Distance Re, o Bibliographical Notes ....... 06... cece cee eee eee eee tee e ee en ene Ix 174 174x 12. 13. 14, 15. 16. Contents The Dynamics of Heavy-Ion Collisions .. 12.1 Rutherford Scattering .......... 12.2 Expansion in the Quasi-Molecular Basis 12.3 Heavy-Ion Collisions: A Quantal Description . 12.4 The Semiclassical Approximation .. 12.5 Collisions with Nuclear Interaction . 12.6 Status of Numerical Calculations . Bibliographical Notes Experimental Test of Supercritical Fields in Heavy-Ion Collisions .... 13.1 Establishing Superheavy Quasimolecules . 13.2 Positron Spectrometers .............. 13.2.1 The “Orange”-Type # Spectrometer . 13.2.2 Solenoidal Transport Systems . 13.3 Background Effects Creating Positrons . 13.4 Positron Experiments I: Gross Features 13.5 Positron Experiments II: Deep Inelastic Collisions 13.6 Positron Experiments III: Narrow Structures in the Positron Spectrum . 12.06... eee eeeee eee e eee 13.7 Giant Nuclear Systems and Spontaneous Positron Emission Bibliographical Notes .. i Vacuum Polarization 14.1 Vacuum-Current Density: Perturbative Expansion 14.2 Gauge Invariance and Vacuum Polarization .. 14.3 Charge Renormalization ............... 14.4 Explicit Form of the Polarization Function . 14.5 Vacuum Polarization Effects in Atoms Bibliographical Notes Vacuum Polarization: Arbitrarily Strong External Potentials ........ 15.1 Green’s Function for Arbitrarily Strong External Potentials 15.2 Vacuum Polarization Charge Density - 15.3 Vacuum Polarization in External Fields of Arbitrary Strengths . . 15.4 Real and Virtual Vacuum Polarization Bibliographical Notes Many-Body Effects in QED of Strong Fields ..................005 . 16.1 Self-Consistent Hartree-Fock Equations . : 16.2 Self-Energy Effectsin Atoms ...... 16.3 Self-Energy in Superheavy Atoms 16.4 Supercharged Vacuum 16.5 The Problem of a Supercritical Point Charge 16.5.1 Overcritical Single-Particle States ... 16.5.2 Screening Effects of the Vacuum Charge . . 16.5.3 Influence of Heavier Leptons .......... 00. ceeeeeeeReferences Subject Index Contents 16.6 Klein’s Paradox Revisited . Bibliographical Notes Bosons Bound in Strong Potentials . . 17.1 The Klein-Gordon Field ... 17.2 Alternate Form of the Klein-Gordon Equation .. Bibliographical Notes ........00.6ccccccsesseeececeeeeserseees . Subcritical External Potentials 2.2.2.0... 06.02 se eee eee eee 18.1 Quantization of the Klein-Gordon Field with External Fields ... 18.2 (Quasi) Particle Representation of the Operators ... 18.3 The Fock Space and Diagonalization of the Hamiltonian . 18.4 The Coulomb Problem for Spin-0 Particles Bibliographical Notes ........... 20.0.0... e cece eee eee eee eee ee . Overcritical Potential for Bose Fields .. 19.1 The Critical Potentials ... 19.2 The True Ground State and Bose Condensation . 19.3 Solutions of the Condensate Equations . Bibliographical Notes .. ). Strong Yang-Mills Fields 20.1 Quantum Chromodynamics . 20.2 Gluon Condensates in Strong Colour Fields . 20.3 The “Magnetic” Vacuum of QCD 20.4 Spontaneous Quark Pair Production and Fission of Quark Bags Bibliographical Notes .. Strong Fields in General Relativity 21.1 Dirac Particles in a Gravitational Field 21.2 Limiting Charge of Black Holes .. 21.3 Uniform Acceleration and Rindler Space . 21.4 Event Horizon and Thermal Particle Spectrum Bibliographical Notes .. XI 467 469 470 470 477 483 484 484 486 490 495 498 499 499 507 515 519 520 520 528 534 540 549 550 350 557 563 567 5711. Introduction The structure of the vacuum is one of the most important topics in modern theo- retical physics. In the best understood field theory, Quantum Electrodynamics (QED), a transition from the neutral to a charged vacuum in the presence of strong external electromagnetic fields is predicted. This transition is signalled by the occurrence of spontaneous e*e™~ pair creation. The theoretical implications of this process as well as recent successful attempts to verify it experimentally using heavy ion collisions are discussed. A short account of the history of the vacuum concept is given. The role of the vacuum in various areas of physics, like gravitation theory and strong interaction physics is reviewed. 1.1 The Charged Vacuum Our ability to calculate and predict the behaviour of charged particles in weak electromagnetic fields is primarily due to the relative smallness of the fine-struc- ture constant @ = 1/137. However, physical situations exist in which the coupling constant becomes large, e.g. an atomic nucleus with Z protons can exercise a much stronger electromagnetic force on the surrounding electrons than could be described in perturbation theory, and hence it is foreseeable that the new expan- sion parameter (Za) can quite easily be of the order of unity. In such cases non- perturbative methods have to be used to describe the resultant new phenomena, of which the most outstanding is the massive change of the ground-state struc- ture, i.e. of the vacuum of quantum electrodynamics. In fact, the fundamental understanding of the vacuum, its properties and phases is of more general interest. During the last decade in field theories the structure of the vacuum has been considerably elucidated. But the conjecture that the vacuum is not just an empty and trivial space domain and its philosoph- ical implications have a very long history dating back to the Greek philosophers of the Eleatic school and to Aristotle in particular. We shall come back to this in- teresting development and describe the historical roots below. Only with the advent of a quantum field theory did it become clear that a vacuum is an object capable of being subjected to physical experiments. The most fascinating aspect of the vacuum of quantum field theory, which is discussed extensively in this book, is the possibility that it allows for the creation of real particles in strong (static) external fields: the normal vacuum state is un- stable and decays into a new vacuum that contains real particles. This in itself is a2 1. Introduction deep philosophical and physical insight. But it is more than an academic prob- lem, for two reasons. Firstly, very strong electric fields are available for laboratory experiments and secondly, here is an example of a more general phe- nomenon encountered in strongly interacting quantum field theory, namely, analogous physical phenomena are encountered again when discussing the inter- action of other fundamental particles and fields. Quantum electrodynamics is commonly, but falsely, identified with its per- turbative expansion in powers of the fine-structure constant @ = e?/hc. Hence, the perturbative radiative corrections are normally referred to as the testing ground of. QED. However, this traditional view ignores the importance of under- standing the interaction of electrons with a strong source such as a heavy nucleus. Here the coupling constant becomes Za, where Ze is the electric charge of the nucleus. Since Za is not generally small, and also because understanding bound states requires non-perturbative treatment, we must also consider other phe- nomena than radiative phenomena. That there are effects of zeroth order, i.e. ‘yes or no’ effects, which are most fundamental aspects of QED, does not seem to be well known. The most profound of these effects is the change of the vacuum (more specifically, the electron-positron vacuum) in over-critical fields. Let us first illustrate here what is meant by these statements. One of the basic features of (relativistic) quantum field theory is the intrinsic many-body aspect of the theory. It becomes important as soon as a quantum field is considered in a strong external (classical, i.e. unquantized) field: particles may be produced spontaneously when there is a change of the vacuum caused by sufficiently strong and either strongly space- or time-dependent external fields. In the static case, even below the critical potential strengths, virtual particle pairs are generated as fluctuations of the vacuum, but because of the uncertainty relation the lifetime of the virtual pair is only of the order At ~ h/mg c?. However, if such a virtual pair is separated during this time by more than a Compton wavelength and if it has gained more kinetic energy than twice its rest mass, the pair may become real, as for overcritical potentials. Hence the strong background field has to extend over a sufficiently large region to lead to observable critical behaviour. Indeed, a potential difference of more than 1 MV in a Van de Graff machine, that is, over macroscopic distances, has a negligible probability to lead to particle production. However, if both the potential and its gradient are sufficiently strong, the particle creation process continues until either the potential differences in the ex- ternal field are reduced substantially or the Pauli principle prevents further par- ticle creation. Very often the new ground state will be charged, thus it has dy- namically broken charge symmetry. The new ground state is called ‘supercritical vacuum’ or ‘charged vacuum’ [Mii 72]. A precise definition of the vacuum in quantum electrodynamics is obtained considering the single-particle spectrum (Fig. 1.1) of electrons in an external field. The spectrum contains positive and negative energy states separated by an energy gap of 2moc?, containing the localized bound states. The negative energy states are y,, the positive energy states, yp). The quantum field operator at time t= 0 is defined as1.1 The Charged Vacuum 3 fe Fig. 1.1. Single-particle spectrum of the Dirac equation in an positive attractive external field continuum TGS [__ bound states =m} _ Fermi surface negative continuum Px, t=O = Y Spvplx)+ ¥ an Walk), (1.4) where the operators 5, are the annihilation operators for electrons and 47 the creation operators for positrons. This definition guarantees that the energy Ap of the electron-positron field is positive definite, i.e. (Chap. 9) Ay= YE nb} bp+ ¥|Enlat Gn. (1.2) 5 a The vacuum |0), characterized by the Fermi energy Zp, i.e. by the states |) with E, < Ey filled with electrons, is best illustrated using hole theory (Fig. 1.2). A hole in this “Dirac sea” is then interpreted as a positron. The position of the Fermi surface dividing the single particle states into ‘electronic’ and ‘positronic’ states can, for a given physical system, be determined experimentally by observing, e.g., the threshold for e*e~ pair production. Normally, such a vacuum |0) is neutral. Calculating the vacuum expectation value of the charge operator (Chap. 9) Sx) se (x, 0), P@, 01, (1.3) which is the zero component of the current four-vector b aS 18 yA, (1.4) _ ‘unbound (tree) states mec? bound states Ferm: surface Fig. 1.2. The neutral QED vacuum can be — nau viewed as negative energy states filled with elec- {ree states trons. The infinite charge is renormalized to zero ~mec?4 1, Introduction gives «0 (6/0) = apo) fee vi wale) E vb vio}. a5) This expression vanishes by symmetry for the field-free case because of the equal number and structure of n and p states. It leads to the well-known expression for the vacuum-polarization charge g foi(x) when these interactions are present. Vacuum polarization is now a very well understood phenomenon as, e.g., part of the Lamb shift, energy shifts in muonic atoms; even in weak fields the vacuum is a polarizable medium, characterized by occurrence of a displacement charge den- sity Qpat), which for weak fields does not contain a net charge: foxi@yd?x=0. (1.6) In Fig. 1.3 a qualitative illustration of this phenomenon is represented by the remarkable fact that upon renormalization the displacement charge density is positive nearer to the positive charge source. If the charge of the central nucleus is continuously increased, the spectrum of electrons near a point-like and an extended source looks as presented in Fig. 1.4. For point nuclei the so-called fine-structure formula results, which has no s states beyond Z = 137. This puzzling behaviour is postponed to the so-called critical charge Z~ 173 for extended nuclei, where the 1s state “dives”, i.e. joins the negative continuum. For even higher central charge at Z:, ~ 183, the 2p state dives, etc. The significance of this is the following. In the supercritical case the dived state is degenerate with the (occupied) nega- tive electron states. Hence spontaneous e*e™ pair creation becomes possible, where an electron from the Dirac sea occupies the additional state, leaving a hole in the sea which escapes as a positron while the electron’s charge remains near the b b, The vacuum polarization charge around a central nucleus (a) and around two colliding iclei (b). In (a) the static s electrons are shifted in energy somewhat due to modification of the Coulomb potential by Opel (this is part of the Lamb shift); in (b) the vacuum polarization charge is partially stripped off because of ionic motion. The vacuum polarization charge cannot readjust fast enough to the permanently changing position of the moving nuclei [So 77a}. Here “e” denotes the electron charge1.1 The Charged Vacuum 5 {) positive energy continuum Tiegative enéfay continua accupied with electrons Fig. 1.4. Lowest bound states of the Dirac equation for nuclei with charge Z. While the Sommerfeld fine-structure energies (~ — —) for x= —1 (5 states) end at Z= 137, the solutions for extended Coulomb potentials ( ) can be traced down to the negative-energy continuum reached at the critical charge Z,, for the 1s state. The bound states entering the continuum obtain a spreading width as indicated [Ma 72a] source [Pi 69]. This is a fundamentally new process, whereby the neutral vacuum of QED becomes unstable in supercritical electric fields. It decays within about 10-9 s into a charged vacuum [Mii 72a — c]. The charged vacuum is now stable due to the Pauli principle, that is the num- ber of emitted particles remains finite. The vacuum is first charged twice because two electrons with opposite spins can occupy the 1s shell. After the 21/2 shell has dived beyond Z;, = 185, the vacuum is charged four times, etc. This change of the vacuum structure is not a perturbative effect, as are the radiative QED effects (vacuum polarization, self-energy, etc.), but has all the features of a phase transi- tion. In the new charged vacuum the polarization charge contains a real compo- nent, not only a displacement charge density of net zero charge: \ Q’* = § (charged vacuum | 6|charged vacuum) d°x = 2e, 4e,... . (1.7) When displaying the vacuum charge as a function of the proton number Z of the source, the usual discontinuity associated with a phase transition is evident. The dashed line in Fig. 1.5 is smoothed, ignoring the atomic shell structures. We rec- Fig. 1.5. Vacuum charge Q* as function of nuclear charge. (. ) actual calculation; (———) smoothed behaviour6 1. Introduction Q) undercritical 5) overcritical pump Fig. 1.6a, b. The vacuum in the box is penetrated by the electric field of a central nucleus with charge — Ze; e being the charge of an electron. The nucleus is solely a spectator, furnishing the source of the electric field ognize a 2nd-order transition. The full lines are the actual results with strong shell effects. One can view Q™ as the order parameter of the transition. The physical significance of the vacuum charge is further vividly illustrated in Fig. 1.6. The space in the box is thought to be pumped empty by an elementary particle pump. (Theoreticians can “invent” such a tool.) Only the central nucleus can act as a source for the electric field left as a spectator. The empty space in the box represents the neutral vacuum. In the undercritical case (a) it is stable. In the supercritical case (b), however, two positrons are emitted into free states, travel- ling around in the box and simultaneously two electron charges are localized around the nucleus. The positrons, being in free states, are easily pumped away. The electron cloud left over is the charged vacuum, representing the stable ground state in the supercritical field case. With great effort (i.e., substantial supply of energy) the two vacuum charges can also be pumped out, but within about 10~'%s again two positrons are created and the vacuum charge cloud re- establishes itself. In the supercritical situation only the charged vacuum is stable. With these non-perturbative concepts well established, one still has to show that what has become of the radiative corrections is a negligible and well control- lable effect. In this context the following questions can be raised. 1) Can vacuum polarization hinder the vacuum decay, i.e. prevent diving of the bound levels? The answer is no. The energy shift due to vacuum polarization is of the order of a —10 keV (about one percent of the binding energy) at the diving point [Gyu 75, Ri 75]. 2) Can self-energy prevent diving? The answer is also no, Self-energy correc- tions shift the energy approximately + 10 keV at the critical charge as Soff et al. have established [So 82b]. 3) Can non-linear effects in the electromagnetic field or in the Dirac field eventually become so large in supercritical fields that vacuum decay is prevented or, at least, substantially shifted toward higher central charges? The answer is again no, if the excellent agreement between QED of weak fields and high preci- sion experiments (Lamb shift, 4 atoms, etc.) is not to be destroyed [So 73, Wa 74]. 4) What happens in the point limit, ie. when the radius of an overcritical charge distribution R -0. This very interesting problem was solved several years1,2 From Theory to Experimental Verification 7 72153 Fig. 1.7. As the radius R of the central charge shrinks to zero, the vacuum has an onion-shell type Z structure and becomes higher and higher charged, % remaining shielding the central charge. Thus point charges #5 Mecirons higher than 137 are prevented in QED Zz ZL shells of vacuum Srungtacher with the nucleus "Pac R wr ago [Ga 81], indicating that as R-0, more and more bound levels with x= +14 dive. The vacuum becomes higher and higher charged; the charge distribution of the vacuum in which the (zs, (+ 1) p?) shells are arranged like onion skins shrinks to a point, so that in the limit R- 0 the effective charge Z.; of the system, con- sisting of the central charge Z minus the vacuum charge Z,a, approaches 137: » Zyp= Z- Loe aa? 131. This is a most interesting result indicating that QED does not allow point charges with charge larger than 137 (Fig. 1.7). In other words, the coupling constant of QED (for point particles) can never become larger than one. If a charged funda- mental (point-like) spin-0 boson existed, the limiting charge of a point-like nucleus would be reduced to 137/2, because of the Klein-Gordon equation [Ba 81c]. 1.2 From Theory to Experimental Verification The crucial question is: Is the decay of the neutral vacuum an observable event? The question can be answered positively, but first the following concepts need to be established. i) In heavy-ion collisions near the Coulomb barrier, the electrons behave as if the source were the combined Coulomb field of both nuclei. Since the electrons’ motion is similar to the binding electron in a molecule this is a quasi-molecular state. However, the electronic binding is much too weak to keep the nuclei together. ii) Under certain conditions the nuclear interactions keep the colliding ions from separating again for 107s or more. Then the vacuum has time to decay if iii) there has been a substantial probability for establishing the neutral vacuum in the entrance channel, i.e. for a hole in the strongly bound K shell. We now discuss these three points in more detail.8 1. Introduction 1.2.1 Superheavy Quasimolecules Quasi-stable islands of superheavy nuclei were predicted around Z = 114 [Me 67, Ni 69a, Mo 68] and around Z = 164 [Mo 69], but there was little hope from the beginning that they could be produced in sufficient quantity. More importantly, the predicted proton number (and hence the electric field) was not sufficiently large to lead to supercriticality. The new idea was to use the slowness of nuclear motion as compared to the motion of inner-shell electrons when the colliding nuclei are close to each other during a collision. Thus a superheavy molecule or atom is simulated (Fig. 1.8) [Ra 71, Mii 72c]. Hence these intermediate systems were called superheavy quasimolecules. The adiabaticity conditions particularly are also fulfilled for the highly rela- tivistic inner-shell electrons in such systems [Ra 72]. Scheid and Greiner [Sche 70, Gr 70] began to study nuclear quasimolecules, in 1967 at the University of Virginia (where Pieper and Greiner worked on electrons in superheavy nuclei at the same time) to explain the nuclear molecular resonances observed by Bromley and collaborators [Br 60]. It was then a small step from the nuclear quasi- molecule to the electronic quasimolecule. Fig. 1.8. Basic idea for forming superheavy quasimolecules in heavy-ion collisions. Since the relative motion of the two nuclei is slow (Vion ~ ¢/20) compared to the motion of the inner electrons (V, = c), the latter are expected to orbit around both centres This proposal of the Frankfurt school had its precursors. An alternative to the theoretical method for calculating ion-atom collisions in Born approximation is the perturbed stationary state method [Mo 33], known for a longtime. Mott described in his excellent paper already fifty years ago the elements of the time- dependent molecular perturbation theory, using the eigenfunctions of the quasi- molecule, similar to the so-called Born-Oppenheimer approach. About the same time, Coates [Co 34] observed what are now called quasi-molecular x-rays: he interpreted the x-ray bumps observed as radiation stemming from an inter- mediate quasi-molecular collision system. As with many untimely discoveries, these results remained obscure, as the measurements did not have independent importance within the precision obtainable at that time. On investigating the vacuum stability of quantum electrodynamics in 1969/ 1970, the Frankfurt school first proposed that supercritical quasimolecules would be created in heavy-ion collisions and would be essential in understanding1.2 From Theory to Experimental Verification 9 QED of strong fields [Gr 69, Ra71, Mii 72a—c]. Similar ideas were also ex- pressed by Gershtein and Zel’dovich [Ge 70} but were not pursued further much beyond a general proposal. The need to understand relativistic particles in two centre potentials led the Frankfurt group to further work [Mii 73a, 76b] which has provided a basis for a quantitative description of the physical phenomena. Some milestones establishing the formation of superheavy quasimolecules are the following. a) Saris et al. [Sa 72] observed continuous x-ray spectra in various ion—atom collisions, which were interpreted as quasi-molecular L x-rays. Nearly simul- taneously, the GSI-Cologne group of Armbruster and colleagues [Ar 72] observed quasi-molecular M x-rays from the superheavy I + Au system. b) Meyerhof et al. measured broad continuous x-ray spectra in Br— Br collisions, which were interpreted as K-molecular x-rays [Me 73b]. These results were confirmed shortly thereafter by Davies and Greenberg [Da 74b]. ©) Greenberg et al. [Gr74] discovered the asymmetry of the quasi-molecular spectrum in Ni+ Ni collisions, theoretically predicted by Miller and Greiner [Mii 74a,b, Gr 76]. Asymmetry of M x-rays was simultaneously found by Kraft et al. [Kr 74]. These measurements constituted the basis of the final identification of quasi-molecular x-rays by Meyerhof [see e)], and for the systematic spectroscopy of quasiatoms by Stoller et al. [St 77]. d) Kaun and his group observed K-quasi-molecular x-rays from superheavy systems [Ka 74]. e) Meyerhof et al. rigorously confirmed the existence of quasi-molecules [Me 75] by measuring the Doppler shift of quasi-molecular x-rays (1975) using the asymmetry of the x-ray spectrum. f) O’Brien et al. isolated selected MO transitions to vacant 1 sq states using the cascade coherence between MO and K x-rays [O’Br 80]. 1.2.2 Nuclear Sticking While studying adiabaticity and the effects of the electron binding in nuclear collisions, the idea emerged that nuclear forces could also be crucially important in delaying the motion of colliding heavy ions. This phenomenon is well known in light nuclear systems, where nuclear molecules were first discovered by Bromley et al. in 1960 [Al160, Br 74] and theoretically founded by the nuclear two-centre shell model [Ho 69, Ma 72, Pr 70, Pa 81]. This development led us to suggest in 1977 that indeed the nuclear time delay would greatly influence the spontaneous positron spectrum from vacuum decay [Ra 78c]. If the two nuclei stick for one reason or another, the spontaneous positron decay line due to the decay of the neutral to the charged vacuum should be clearly visible. The longer the sticking time 7, the more pronounced the line structure should become, hence unambiguously proving the formation of the charged vaccuum. The evolution of the 1s and 2p states is shown in Fig. 1.9 qualitatively as a function of time, assuming that between times f, and f, the nuclei stick, leading to pronounced lines of positrons (insert) in the spectrum. Occasionally, when10 1. Introduction Fig. 1.9. Positron spectrum from overcritical heavy-ion collisions with sticking (formation of super- heavy nuclear systems). The colliding, sticking and separating nuclei are indicated by black circles (©). The 1s and 2p quasi-molecular levels are drawn as a function of time lo IS vacuum decay A ¢ Stokes satellites Fig. 1.10. Schematic positron spectrum from a long-living giant nuclear system. The expected Stokes and anti-Stokes lines are due to excited (mostly collective) states of the giant nuclear sys- tem. They are shown on top of the smooth dy- namic spectrum je AntrStokes /<— satellites Eee nuclear rotation is substantial, one may even notice the magnetic Zeeman split- ting as shown. Reinhardt extended these investigations and incorporated them in a general formulation of positron production processes in sub- and supercritical heavy-ion collisions [Re 81b]. In addition, it has been suggested that the giant nuclear system eventually formed in such sticking processes might itself be excited and decay by supercritical internal conversion [Re 81a, Mu 83c]. The internal conver- sion rate in such giant systems increases by many orders of magnitude compared. to that in ordinary heavy nuclei. Thus additional positron lines might appear at higher energies of the spectrum, yielding information on the nuclear structure of the giant nuclear system formed in the collision. A schematic form of the ex- pected spectrum is shown in Fig. 1.10. The dynamically induced part of the spec- trum is due to induced and direct positron production (shake-off of the vacuum polarization cloud).1.2 From Theory to Experimental Verification 11 1.2.3 K-Shell Ionization The experimental signature for the charged vacuum, viz. spontaneous positron creation, requires the existence of a vacancy in the supercritical bound state. In principle, one can produce an uranium nucleus with only one or even no electron(s) in the K shell by sending the beam through a thin stripping foil. However, the beam energy must exceed several hundred MeV per nucleon, far too much for nuclear collision experiments at energies of 5— 6 MeV per nucleon further down the beam line. (With the SIS 18 accelerator such experiments may become possible in a few years; the fully stripped heavy ions would be cooled in a storage ring and the naked heavy ion beams split and later crossed, as was pro- posed by W. Greiner and Ch. Schmelzer [Be 75].) The important step was to rec- ognize that K-shell vacancy could be produced with sufficient probability by ion- ization in the same collision prior to positron creation [Mii 72c]. This was disputed for several years, with estimates ranging from a probability of 1075 or less [Me 74, Fo 76] to 10~' [Bu 74], until a full-scale calculation predicted a K vacancy probability of several percent [Be 76b]. This prediction has been con- firmed by experiment [Gr 77]. Today, the dependence of the K-shell ionization probability on impact parameter has been measured for various superheavy quasimolecules, where it agrees excellently with theoretical calculations [Li 82]. In particular, observation of the tremendously enhanced K-vacancy creation in superheavy quasimolecules [Gr77, Ma78a, An78b] provided experimental proof that collisions of very heavy ions are an ideal testing ground for QED of strong fields. Use of the K-vacancy measurements to determine binding energies (suggested by Miiller et al. [Mii 78, So 78b]) was experimentally realized by Behncke et al. [Be 78, Li 80]. The first positron experiments with heavy ions establishing their non-nuclear origin and their energy dependence were carried out by Backe and collaborators in 1978 [Ba 78]. Coincidence experiments of positrons and heavy ions [Ba 78, Ko 79b] showed that positron creation is a non-perturbative phenomenon depending on nuclear charge like Z”° (“shake-off” of vacuum polarization). Observation of peaks in the coincidence positron spectrum in 1981 by Kien/e et al. [Ki83] and by Boke- meyer, Greenberg, Schwalm et al. [Bo 83a, Gr 83b] was a breakthrough in the search for spontaneous positron creation. The second group was the first to perform experiments with most kinematic parameters fully determined, which seems essential for reproducing the positron spectra. At about the same time, Backe and collaborators [Ba 83b] studied positron emission in heavy-ion colli- sions in which a deep inelastic nuclear reaction occurs. Triggering with fission products, they were able to observe time-delay effects in the positron spectrum due to nuclear “sticking” of the order of 1077's, thus also contributing to the observation of the spontanous decay of the vacuum. The last year (1983) has witnessed the repeated observation and full confirma- tion of the narrow line structure in the positron spectrum of various supercritical collision systems (U + U, U+Th, U+Cm) by Kienle, Clemente, Bosch, Kozhu- harov et al. [C184], and by Schwalm, Backe, Greenberg and associates12 1, Introduction AVIMeV] 228) 248, isomeric CO antiguration giont molecule ‘exotic incoming energy git uceus b ‘molecutor distance etermined by the Positron line as 1S | | 26 rtm] 2 30 40 50 Fig. 1.11. A quasi-stable giant nuclear system can be formed if the attractive nuclear force leads to a pocket in the heavy-ion potential. Here r is the separation of the two nuclei [Schw 83]. The absence of these structures in undercritical systems (Pb + U) is an important check that the positron lines indeed originate from the giant system and not from standard conversion processes in the uranium atoms, as Kienle’s group determined. Further such checks are necessary. The decisive test that the narrow line structure in the positron spectra does not originate from some ordinary processes (like pair conversion, monoenergetic pair conversion) lies in measuring the 6-electron and x-ray spectra. Both Green- berg’s and Kienle’s groups measured these spectra for U+Cm and U+U systems, respectively, during summer 1983, and found that these spectra are smooth. Cowan, Greenberg et al. even measured the Doppler shift of the posi- trons from the resonance structure and showed that those positrons are emitted from a system moving with the center-of-mass velocity of the compound system. This unseems to prove that the positron lines do not come from conversion of excited nuclear states. Once this possibility is excluded, analysis of the experimental data shows that the positron lines must come from a long-lived giant nuclear system with a life- time of T>5 x 10~™s. The line intensity requires only a rather small cross- section of a few millibarn to form the giant nuclear system in collision. The long lifetime of the combined nuclear system may be tentatively understood if the attractive nuclear force leads to a binding pocket in the internuclear potential (Fig. 1.11) [Se 84, Gr 83c]. The giant system can be formed only in the immediate neighbourhood of the barrier [He 836]. This is precisely the behaviour found in experiments where there is a narrow beam energy window of only 0.2—0.3 MeV per nucleon for observing the positron line structures. The positive proof that the origin of the observed positron lines is, indeed, the decay of the vacuum in the strong electric field of a giant nuclear system can be1,3 Theoretical Developments 13 further confirmed by a Doppler shift analysis of the line structures. If the emit- ting system moves with the velocity of the centre-of-mass of the two ions, the origin of radiation from the combined atomic system is proven. First experiments have been carried out, with the result that the line broadening is compatible with Doppler broadening from the centre-of-mass motion, but not from the motion of the separated nuclei [Bo 84]. Again, further such investigations deem necessary. 1.3 Theoretical Developments Although Sect. 1.4 treats the historical evolution of the idea that particle-anti- particle pairs can be created in strong fields, we note already here that this idea emerged repeatedly between 1930 and 1965, but was never considered worthy of further pursuit. This attitude changed from 1968-1971, when the theoretical schools in Frankfurt [Gr 69, Pi69, Ra 71, Mii72c] and Moscow [Ge 70, Ze 72] became aware that sufficiently strong electric fields could be realized under laboratory conditions in collisions of two very heavy atoms. Both groups were, in part, motivated in their studies by plans at Berkeley, Darmstadt and Dubna to build machines that could accelerate even the heaviest atoms, bringing them as close together as their nuclear radii. Vice versa, these ideas have particularly motivated the scientists in Germany to build the Unilac accelerator at GSI so that uranium can be lifted over the Coulomb barrier of the heaviest elements. An essential step in understanding the decay of the neutral vacuum into the charged vacuum was made by the description of the decay as an autoionization process of positrons [Mii72a—c, Ra 74]. Charged vacuum properties, such as the real vacuum polarization, energy spectrum of the emitted positrons, could be investigated. Fano’s autoionization theory [Fa 61] also enabled simple treatment of the time-dependence of the decay of the neutral vacuum during a heavy-ion collision [Mii 72c, Ra 74, Re 81b]. This is particulary important because super- critical electric fields exist only for a finite time in heavy-ion collisions. When solutions of the two-centre Dirac equation became available (Mii 73a, b], positron creation in heavy-ion collisions could be investigated in realistic situations. The probability of making a K vacancy in the course of the collision could be reliably calculated; it turned out to be of the order of several percent. Predictions about the background of dynamically induced positrons were later quantitatively confirmed by experiment. A statistical description of the charged, supercritical vacuum by Miiller and Rafelski [Mii 75] enabled the decay of the neutral vacuum to be viewed as the prototype of a phase transition in quantum field theory. The expectation value of the total charge in the vacuum state is the order parameter for the phase change. Further developments and considerable improvements of this path later allowed the limiting case of a point-like supercritical charge to be treated, which is shielded by the charge of the vacuum [Ga 81]. The fully dynamical calculations showed that heavy-ion collisions below the Coulomb barrier are so short that spontaneous positron emission is hidden in the14 1. Introduction background of dynamic processes of pair-production. Our suggestion that nucle- ar reactions could provide the time needed for the charged vacuum state to estab- lish itself [Ra 78c], which was initially ridiculed, has become the central idea for theoretical work. New methods to treat supercritical heavy-ion collisions devel- oped by Reinhardt et al. [Re 81b] enabled quantitative predictions for positron spectra with nuclear delay to be made [Re 81a, Mui 83c], which are crucial for comparisons with experiments. As the importance of nuclear reactions is increasingly recognized, present theoretical work tries to put this influence on a solid basis in the framework of a quantum mechanical reaction theory [He 83a, 84a,b, Re 83, To 83, 84]. 1.4 Historical Annotations on the Vacuum 1.4.1 The Concept of Vacuum Since the early Greek philosophers our view of the physical world has been dominated by certain paradigms, i.e. specific pictures, of selected physical entities. Such entities are space, time and matter as the basis of natural philos- ophy or, more specifically, of physics. Therefore, it is no surprise that our concept of “vacuum”, intimately connected with the picture of space, time and matter, is among the most fundamental issues in the scientific interpretation of the world. The picture of a vacuum has undergone perpetual modifications during the last twenty-five centuries as the available technologies have changed; often old, abandoned ideas have been resurrected when new information became acces- sible. Many aspects of today’s concept of the vacuum date back to ancient Greek philosophy, but have only recently been established by modern experiments. Two concepts alternatively formed the ancient Greek point of view concern- ing the vacuum: (i) “vacuum” is the void separating material objects and allowing their relative motion; and (ii) “empty space” is not possible, i.e. the vacuum was conceived as a medium. In the Pythagorean point of view, “air” was identified with the void as a seat of pure numbers. In the atomism of Democritos (400 BC), the vacuum manifests itself as “intervals that separate atom from atom and body from body, assuring their dis- creteness and possibility of motion”. Lucretius, (60 BC), writes in De Rerum Natura: “All nature then, as it exists, by itself is founded on two things: there are bodies and there is a void in which these bodies are placed and through which they move about”. For Plato (380 BC), a physical body was merely a part of space limited by geometrical surfaces containing nothing but empty space. For Plato physics was merely a geometry, a point of view, which finally led to Einstein’s geometric vision of the world and the modern theory of spaces of fibre bundles. Opposed to that attitude was the opinion of Parmenides (480 BC) and Melissos (350 BC), “according to whom the universe was a compact plenum, one continuous unchanging whole” [Me 71]. Aristotle (350 BC) also rejected the1.4 Historical Annotations on the Vacuum 15 Platonic, Democritian and Pythagorean concepts of space, and, avoiding the notion of empty space, he spoke of space as the total sum of all places occupied by bodies, and of place (topos) as that part of space whose limits coincide with the limits of the occupying body. Rejecting the vacuum, Aristotle insisted that the containing body has to be everywhere in contact with that which is “con- tained”. Empedokles performed an experiment with a “vacuum” around 450 BC, using the so-called clepsydra or water thief, a bronze sphere with an open neck and small holes in the bottom, used as a kitchen ladle at that time. Consequently, the idea of a natural “horror vacui” was developed, i.e. the belief that nature tries to avoid the formation of vacuum. However, with the invention of the air pump (around 1640) by Otto v. Guericke, and of the barometer by Toricelli (1643), it became clear that air can be removed from the interior of a vessel. The question remained as to what kind of vacuum is formed if all air molecules are pumped out of the vessel. Since then many different concepts of vacuum were developed by scientists, different vacua as carriers for various kinds of physical phenomena. Newton’s laws of mechanics require an absolute space for the principle of inertia to make sense. He writes in the Principia: “Absolute space, owing to its nature, remains the same and fixed regardless of any relationship to any substance”. Mach con- sidered Newton’s absolute space an empty philosophical notion which should be replaced by an operational device to construct inertial frames. That idea developed into the requirement that local inertial coordinate systems should be determined by the energy-momentum distribution in the universe (Mach’s prin- ciple). In Einstein’s theory of gravity it means that the bundle of local inertial frames of reference is determined by Einstein’s field equation when combined with suitable initial values of energy momentum and geometry on a space-like hypersurface. As an experimental consequence, e.g., it is predicted that near the surface of the earth a free gyroscope, with its axis oriented perpendicular to the rotational axis of the earth, should precess at a rate of 0.05 seconds of arc per year (Lense-Thirring effect). Attempts are underway to measure this precession by a gyroscope launched into orbit around the earth. When the wave nature of light had been firmly established, the hypothesis of the vacuum as an elastic medium, the “ether”, was developed parallel to the theory of elasticity worked out in the early nineteenth century by Navier, Cauchy, Poisson and others. From this point of view the ether had to be a very firm, elastic body to account for the high velocity of light, but it had to impose practically no friction on planetary motion. Michelson’s experiment (1881) showed that the velocity of light is not influenced by the motion of the earth with respect to the ether by quantities of order (v/c)”, and thus rejected the notion of ether as an absolute system of reference. This provided the basis for Einstein’s principle of special relativity, whereby all inertial systems are physically equiv- alent. Newton’s “absolute space” has lost most of its attributes, yet is not com- pletely abandoned in Einstein’s geometrodynamics. Unaccelerated frames are still “preferred” coordinate systems, although locally, within an infinitesimally small space-time volume, a freely falling system in a gravitational field may not be distinguished from a local inertial frame. When larger regions of space-time16 1. Introduction are considered, a true gravitational field is revealed by tidal forces and the radiation field of an “accelerated” charge can always be detected in the far zone. Relativistic quantum mechanics and quantum field theory with the possibility of pair creation laid the grounds for our present concept of the nature of a vacuum. In today’s language, a vacuum consists of a polarizable gas of virtual particles, fluctuating randomly. It is found that in the presence of strong external fields, the vacuum may even contain “real” particles. The paradigm of “virtual particles” not only expresses a philosophical notion, but directly implies observ- able effects: 1) The occurrence of spontaneous radiative emission from atoms and nuclei can be attributed to the action of the fluctuations of the virtual gas of photons. 2) The virtual particles cause effects of zero-point motion as in the Casimir effect. (Two conducting, uncharged plates attract each other in a vacuum environment with a force varying like the inverse fourth power of their separa- tion.) Hawking’s pair formation effect by a collapsing body may also be under- stood as an overcritical “decay of the vacuum” in strong gravitational fields or as a gravitational Casimir effect. 3) The electrostatic polarizability of virtual fluctuations can be measured in the Lamb shift and Delbriick scattering. The electron in the hydrogen atom is subject not only to the Coulomb potential ¢ of the nucleus but also to the fluc- tuation field of the virtual particles. If L is the typical dimension of the system, fluctuations in the electrostatic field strength are of the order (he)? Ae~ r: This field slightly displaces the electron by Ax from the usual classical orbit, leading to a shift of atomic energy levels of the order AE: AE= 4 (ay (V7) average + Precise calculations give a difference of 1057.9 MHz between the energies of the 2812 and 2 p;,2 levels in the hydrogen atom, in good agreement with experiment. Photon scattering by an external electric field (Delbriick scattering) caused by interaction of the photons with virtual electrons and positrons was first measured. by Wilson [Wi 53]. 4) The magnetic polarizability of the vacuum can be measured by observing the rotation of the polarization of laser light travelling through strong magnetic fields. Experiments are in progress to determine this effect [Ia 79]. 1.4.2 The Vacuum in Strong Fields The first indication that something strange would happen in strong electric fields came when O. Klein studied the reflection and penetration of electron waves on a potential barrier Vy, immediately after the first formulation of the Dirac1.4 Historical Annotations on the Vacuum A7 V(z) Viz) e fo e oe EV Ey, Ye, 4 a Zz z Fig. 1.12 Fig. 1.13 Fig. 1.12. The Klein paradox. If the barrier Vj becomes very high (Vj -+ 0»), 83% of the electron wave will penetrate into the potential Fig. 1.13, Sauter studied the Klein paradox for smoothed-out potential barriers Fig. 1.14. By introducing the quantized- field aspected using hole theory, Hund ex- plained the Klein paradox. The incoming electrons could knock out additional elec- trons at the barrier and positrons would move inside the potential equation in 1928 [K129]. He found that for very high potential barriers an unusually large number of electrons penetrate into the wall (about 83% for Vo 0). Moreover, Klein observed that these electrons had negative energy. This was later called Klein’s paradox, Fig. 1.12. Sauter [Sa 31] found essentially the same results in the more general case where the barrier is not sudden, but smoothed out, Fig. 1.13. If the potential step of order moc? appears over a distance of the Compton wavelength d= A,, penetration of electrons into the classically forbidden region ceases. In 1940 Hund [Hu 41, 54] introduced the quantized field aspect by discussing Klein’s paradox in the framework of hole theory, Fig. 1.14. The particles of negative energy in the high potential domain are then electrons from the vacuum. In this picture there can be more reflected electrons than incoming ones, because the incoming electrons can knock out an electron-positron pair at the surface of the wall, the electron joining the reflected electrons and the positron (hole) travelling inside the high potential. Particle creation by strong, infinitely extend- ed constant electrostatic fields (which is related to the Klein paradox) was quanti- tatively predicted by Heisenberg and Euler in 1936 [He 36]. Clearly, pair creation for Klein’s paradox is stimulated by the incoming electron beam. This introduces some time-dependence, which is responsible for knocking out pairs from the barrier. It is similar to induced pair creation,18 1, Introduction Fig. 1.15. Energy spectrum of a Dirac particle in a deep square well potential 3|m =¢ appearing in the time-dependent heavy-ion collision (shake-off of the vacuum polarization discussed above). It may be possible to create pairs by the time- dependence of an external gauge field since a quantized field can be viewed as consisting of infinitely many harmonic oscillators (“modes”). If the quantized field is then excited from its ground state (i.e. from the vacuum state), real particles are produced. In some cases these two possibilities of pair creation by external gauge fields are connected by a gauge transformation: a globally static external gauge field may represent a dynamic field if only the gauge is suitably chosen. One simple example is the constant electrostatic field described by the static potential A, = (E-x,O) and also by A, = (O, Et). Shortly before Hund’s contribution in 1939, Schiff, Snyder, and Weinberg, in Oppenheimer’s institute, discussed particles in a deep well [Schi40]. They found an energy spectrum similar to the behaviour as we understand it today, shown in Fig. 1.15 for the Dirac equation. When the depth of the square well potential exceeds 2m,c’, the energy of the lowest bound state becomes equal to ~ m.c”, and the bound level joins the states of the Dirac sea. They recognized that beyond this “critical” potential depth, the Dirac sea, i.e., the vacuum, contains more states than before. If these are all occupied by electrons, the ground state has non-zero charge. That these bound states “dive” into the negative energy continuum, keeping their identity to a large extent (up to a spreading) width, was not recognized. In 1945, Pomeranchuk and Smorodinsky [Po 45], in Landau’s institute, showed that a similar phenomenon occurs in the deep Coulomb potential of a superheavy nucleus with charge Z = 175 (at a nuclear radius of 8 fm). They accepted the interpretation given by Schiff et al. [Schi 40] that the ground state becomes charged, but they failed to recognize that this could occur only with the accompanying positron production. They write in their article: “If an electron happens to be in such a well (or in the field of such a charge), it will jump on this unoccupied level. The vacuum acquires a charge — e, and the electron, since this is the level of an unobservable vacuum, will pass into an unobservable state”. This is close to ideas of a charged vacuum, but that was not clearly recognized then. The ideas were still rather confused. The same is true for the investigation of Werner and Wheeler [We 58], who studied superheavy nuclei, more or less hypothetically from the present point of1.5 The Vacuum in Modern Quantum Physics 19 view. They also solved the Dirac equation for extended high-Z nuclei up to Z=170 and found that the K-binding energy can reach 2m,c?. The fact that beyond Z = 170 something fundamental happens was not noticed by them. The spontaneous positron emission in such a supercritical situation was first recog- nized in 1961 by Voronkov and Koleznikov [Vo 61]. This paper was rediscovered around 1977 by Armbruster who had been approached by Koleznikov at a con- ference in Dubna. Even though this short paper contained for the first time the correct interpretation of what happens at the critical point, it did not influence the later physical developments, because the idea of how to realize sufficiently strong fields in an atomic collision was missing and the theoretical concepts had not been further worked out. A relative step back was the paper dealing with the Klein paradox as regards atoms by Beck et al. [Be 63], in which they stated that if the binding energy for electrons in a potential becomes larger than 2m,c?, those states “will be occupied”. That they did not understand the physical implications of this phenomenon is hidden in a footnote in which they remark “that this occu- pation is — of course — not in the sense of a physical process, but according to some definition” they give. 1.5 The Vacuum in Modern Quantum Physics In modern quantum field theories the vacuum necessarily has to have certain Structures, most of which are not yet completely understood. Let us review here the best known cases [Le 81]. a) In some o* theories the vacuum is degenerate, because the V(@) potential looks as depicted in Fig. 1.16, and degenerate ground states are obviously possible with (@) = + do. vid) Fig. 1.16. The ground state (vacuum) in a ¢* theory is degenerate Fig. 1.17. The true vacuum of QCD is liquid- like, with virtual gluon and quark balls of a certain size forming and decaying. Real quarks can exist only within the simple vacuum; they are caught inside the bubbles20 1. Introduction b) In gauge theories of weak and electromagnetic interactions, it must be assumed that the vacuum contains Higgs fields to produce masses for the particles in a gauge-invariant way. Such Higgs fields have non-vanishing expec- tation values of fields in the ground state, such as (a). It is difficult to judge the physical reality of these constructions, but we do not know any better way to construct gauge-invariant, renormalizable theories allowing the gauge particles to obtain mass. However, it has been suggested that the Higgs fields express dynamic symmetry breaking. c) In quantum chromodynamics, the theory of strong interactions, two vacua have been discussed: i) The true vacuum expels chromodynamic field lines similarly to a supercon- ductor repelling magnetic field lines. Hence, an assembly of real quarks, which are the source of chromoelectric field lines, forms a bubble (bag) within the true vacuum. ii) For a perturbative vacuum the true vacuum cannot exist in this bubble because of the presence of chromoelectric fields. It is likely that the bag con- tains particles behaving as if they were in structureless simple vacuum, subject to the Jaws of perturbative quantum field theory, Fig. 1.17. d) Grand unified theories of all interactions require that a very large sym- metry group is broken sequentially, leading to the richness of the physical world around us. It is generally believed that the symmetry breaking may, at least in part, be dynamic and the vacuum structure is correspondingly complex and un- explored. The problem with the vacuum ideas described under (a) and (b) is that the structure of the vacuum is a priori predetermined by the choice of the Lagran- gian. In cases (c) and (d) the contrary applies. There is probably a more funda- mental Lagrangian, but we have no tools to explore the vacuum structures ina relatively easy and controllable way. But “experimenting” with the vacuum is necessary to substantiate the underlying ideas, or, put differently, to convert the models and vague ideas into physical pictures of consequence in understanding the world. This has been achieved by the description of the vacuum structure in the field theory best known today, i.e. the vacuum of quantum electrodynamics, which is the central issue of this book. The intense study of the problem of electrons bound to strong potentials has stimulated interest in a number of related problems. They can be divided into three groups: a) bosons bound to strong electromagnetic and nuclear potentials; b) fermions and bosons in strong gravitational fields; c) the vacuum structure of strongly interacting fermions and bosons (over- criticality in QCD). 1.5.1 Pion Condensation Let us first discuss the theory of bosons bound to strong potentials. Clearly, some conceptual features must be different since the Pauli exclusion principle1,5 The Vacuum in Modern Quantum Physics 21 ee SS Fig. 1.18. Qualitative features of eigenvalues from the Klein-Gordon equation for a square well potential I an \y/ \) i \ IS Ws / stabilizes the charged vacuum for fermions in overcritical fields; without it con- tinued positron production would ensue. This is the central problem when con- sidering bosons in strong external potentials. The first step here is to understand the single-particle spectrum of the Klein-Gordon equation. As already mentioned, the earliest investigation of the solutions of the Klein- Gordon equation with a strong external potential was carried out by Schiff et al. [Schi 40]. They solved the problem of the square well potential and found that the spectrum behaves qualitatively different from that of the Dirac equation as the potential strength is varied. They discovered that for a given state, there are two critical points; at a value V, an antiparticle state with the same quantum. numbers emerges from the negative energy continuum, while at V,, the particle and antiparticle states meet each other. No particle or antiparticle state with these particular quantum numbers is found above V.,, Fig. 1.18. Although this behaviour of the spectrum is suggestive for any potential, a different result was found later by Popov (1971) and discussed by Bawin and Lavine (1975) for long-range potentials. In particular, they found that the Coulomb potential of a finite size charge has an eigenvalue spectrum similar to that of the Dirac equation [Po 71, Ba 75]. Although the superbound pion system is theoretically a very interesting problem, there does not seem to be a prospect of experimental tests in atomic collisions. This can be seen by noting that an external potential comparable in strength to the mass of the pion Rm,a"' ~ 2000 is needed. Here R describes the nuclear size. This formula is valid only if Rm, >1 and therefore does not apply to electrons bound in the nuclear Coulomb field. Synder and Weinberg success- fully introduced a second quantization of the theory for potentials smaller than V., but made no attempt to treat the overcritical case [Sn 40]. The modern development of the subject which led to present understanding of the supercritical state for pions was stimulated by the work on the overcritical Dirac equation. Migdal [Mi 72] argued that to stabilize the vacuum in the over- critical Bose case, some further interaction effect must be included in the Hamil- tonian, such as a A g* term (see above). This is based on the consideration that as V approaches V,,, the energy necessary to make a meson pair vanishes, allowing, in principle, an infinite number of pairs to be produced. To stabilize the vacuum, a further positive definite part in the Hamiltonian is needed to stop the produc-22 1. Introduction tion of the condensate when a certain meson density is reached. In this context introduction of an arbitrary interaction is, however, quite unsatisfactory as the condensate and its inherent stability must follow from a picture of dynamic sym- metry breaking. This much more difficult problem was first approached by Klein and Rafelski [K1 75a, b, 77, Ra78b], who showed, in particular, that it is possible to consider the charge of the condensate itself as the stabilizing agent. They demonstrated that the Coulomb repulsion, which is always present in any charge distribution of the meson condensate, suffices to stabilize the condensate. Their treatment was formally complete for both the under- and overcritical cases. A full relativistic quasi-particle formalism was developed for V > V., and equations determining the Bose condensates were given. At that time it had not yet been realized that the 1s state of the Klein-Gordon equation joins the lower continuum for long-range potentials, much in the same way as for Dirac states [Ba 75]. Recognition of this feature of the Klein-Gordon spectrum led [K175b] to the conjecture that a charged condensate develops in the long-range supercritical Coulomb field, which was further substantiated and the properties of the charged condensate described in [K177, Ra78b]. In accordance with charge conservation, a large number of free antimesons would be produced in this process. Furthermore, in view of the possibility of weak decay of the charged vacuum, other mechanisms leading to pion condensates in external fields have also been described. Another form of “pion condensation” arises because nuclear matter itself may have excited states with the quantum numbers of a pion, i.e. OT -T=1 states. Now, as a function of nuclear density the effective nucleon-nucleon interaction generally changes. It turns out that such excited states (in normal nuclear matter) may become the ground state in dense nuclear matter. However, intensive searches for this effect were negative [Os 82], so that one must conclude that the critical density for this second type of pion condensation is beyond ex- perimental reach, or as has been argued very convincingly in recent years by Faessler, that such an effect does not exist at all [F4 83]. Therefore these develop- ments are not further discussed in this volume. 1.5.2 Strong Gravitational Fields The success of QED of strong fields raises the question whether further instabili- ties of the Dirac vacuum, similar to the electrostatic instability, can occur in nature. Although many instabilities of the Dirac vacuum can be visualized via energy level diagrams based on solutions of classical field equations, they should ultimately be described in the framework of quantum field theory. For this purpose one needs a manifestly covariant formulation of the quantized Dirac field in curved space-time enabling description of particle creation. Such a for- malism has, for example, been suggested by Rumpf [Ru 76], whereby pair crea- tion by external fields can be described by introducing four (instead of two) clas- ses of field modes, viz. ingoing and outgoing particle and antiparticle modes. This formalism, however, is not applicable to situations where a new stable ground state of the supercritical Dirac system develops. Then one has to rely on1.5 The Vacuum in Modern Quantum Physics 23 energy level diagrams. Stabilization of the supercritical vacuum can then be de- scribed by statistical methods first suggested for the highly charged supercritical vacuum in quantum electrodynamics [Mu 75]. For example, in the framework of generalized gravitation theory with spatial torsion, the Dirac vacuum develops a non-zero expectation value of spin in strong graviational torsion fields [So 79c, 82c]. Let us sketch the general procedure for studying Dirac particles in gauge fields. One starts with the free Dirac action of spin-+ particles that have (internal) degrees of freedom with values on a semisimple Lie group. Background fields are introduced as gauge fields describing the parallel transport of the spinor from one space point to the next. Typical gauge groups are U(1), forming the basis of QED, the Poincaré group P, underlying the Einstein-Cartan theory of gravita- tion, and the SU(m) groups, playing such an important role in modern theories of elementary interactions (Yang-Mills theories). It turns out that the nature of this group profoundly influences the stability or instability of the Dirac vacuum in a strong, globally static gauge field. In general, non-Abelian gauge theories such as the Einstein-Cartan theory or Yang-Mills theories introduce a complicated and rich vacuum structure into the Dirac system [So 82c]. The complications arise from the gauge freedom: gauge non-equivalent potentials may lead to identical field strengths but to different behaviour of the Dirac field. On the other hand, locally gauge-equivalent poten- tials may lead to different vacuum structures distinguishable by global, topo- logical properties. For example, a uniformly accelerated observer could measure particles in the Minkowski vacuum. Here, the transformation from an observer at rest to a uniformly accelerated observer is a gauge transformation in the sense of Poincaré symmetry of the theory, but leads to a new vacuum structure, the vacuum of the accelerated observer. The new vacuum structure of the accelerated observer means that the Fock space of those states actually measured by a physical detector depends on the coordinate system of the observer and on the topology of the submanifold of the entire space-time, which is causally connected with the observer. Thus there is the apparent paradox that the gravitating part of matter can be described by a co- variant energy-momentum tensor 7,,,, but that, on the other hand, the trajectory of the detector enters into the detection of field quanta, which is therefore a coor- dinate- (or gauge-)dependent phenomenon [Un 76]. This behaviour is particular- ly evident in the theory of fermions in Rindler space, as shown by Soffel, Miiller and Greiner [So 80b]. It turns out that the observer moving with a constant acceleration g in Minkowski space experiences a thermal flux of all kinds of fermions with effec- tive temperature T = g h/2ck (k is Boltzmann’s constant). This result, known for some time for particles with integer spin [Fu 73, Un 76], depends crucially upon the fact that the uniformly accelerated observer in Minkowski space has an event horizon: there are parts of Minkowski space-time with which the observer cannot communicate [Fu 73, Un 76]. Thus the geometric and topological structure of the submanifold naturally connected with the observer’s state of motion leads to the appearance of the thermal particle spectrum.24 1. Introduction 1.5.3 Vacuum Structure of Strongly Interacting Fermions and Bosons Consider that in a Gedankenexperiment the fine-structure constant is increased arbitrarily. At a certain point, if the bound positronium state reaches zero mass, an inherent instability of QED could occur [Ra 78a]. For a> a, ~ 1 muons, the heavy partners of electrons can no longer be produced as charged bare quanta, since the spontaneous production of an e*e~ pair would then become possible, leading to a neutralized muon state with the muon charge carried away by the electron field. At even greater values of a (probably about 1.5) one could argue that the electron field also develops an inherent instability and a complex vacuum structure develops. This has been explicitly shown in the context of the so-called Schwinger model [Schw 62]. In one-dimensional space this problem can be solved, but the instability appears already for any arbitrarily small coupling con- stant. However, as this example shows, the method developed for studying supercritical fields can be applied successfully to gauge theories, particularly to studying the structure of the QCD vacuum. Here is simultaneously a strongly coupled theory with the bose field having both attractive and repulsive self-inter- action. The inherent instability of the perturbative vacuum of this theory has been widely discussed [Ma 77a, b, 78b, Mi78, Ma 79a], but only recently has the new ground state based on the perturbative vacuum from gauge field condensates around a Fermi source been constructed [Mu 82, Ca 83]. This new ground state, which must be considered in a satisfactory theory, requires substantial screening of the colour charge of the source. From this study one learns that the essential aspect of the interaction is the supercritical attractive force between the vector gluons and the (average) colour magnetic field of the condensate when both are antiparallel. This property can be further exploited to construct a new, non-perturbative global vacuum state, based on an approximate evaluation of the zero point energy of the interacting gauge fields and in particular on the associated ‘effective Lagrangian’. In this context it was shown that the supercritical binding of vector gluons in a colour- magnetic field lowers the energy density of the interacting ground state below that of the perturbative state globally [Sa 77, Ni 78]. Using this qualitative model quantitatively, one obtains for the first time the true vacuum state which is indeed quite different from the perturbative (naive) vacuum in which (perturba- tive) quarks are found. An ultimate description of the true QCD vacuum state in terms of the perturbative fields, perhaps similar in idea to the BCS method, but certainly quite different in detail is still to be given. This problem is intrinsically related with the colour confinement problem of QCD. The general problems associated with the quark-gluon ground state are discussed in Chap. 21, but of course remain incomplete. Our intention is primarily to develop a modern view of the properties of the vacuum of particles in various strong fields. In particular, we want to convey that related phenomena are connected with a change in the ground state for many dif- ferent types of interaction. All these phenomena have in common that some critical strength of the interaction exists at which it becomes energetically more favourable to produce particles out of the old vacuum during the transition to a1.5 The Vacuum in Modern Quantum Physics 25 new ground state. Depending on the geometry of the fields, either one particle species is emitted to infinity while the quantum numbers of the other species are retained in the interaction region, or both ‘particles’ and ‘antiparticles’ (i.e., their charges) remain localized, providing a real vacuum polarization. In either case, the phenomenon screens the interaction strength. These considerations are particularly important for theories supercritical for any value of the coupling strength, such as QCD, for which it is therefore difficult to guess the correct ground-state wave function. The necessary condition under which the vacuum state (phase transition) can change is that the field distinguishes the particles according to some quantum number, such as electric charge, spin, colour. This condition may be expressed more formally by saying that the interaction must dynamically break some sym- metry of the non-interacting theory. Consequently, the changed ground state is characterized by the spontaneous global (over the typical domain of the inter- action) breaking of the symmetry, i.e. it is charged or carries spin, colour, etc. Therefore, the underlying subject of this volume is really dynamic symmetry breaking in strong (gauge) fields. But we emphasize that this phenomenon is typical for most interactions and not characteristic for a special one. Of course, decay into a charged vacuum state in strong electromagnetic fields plays a distinct role as the one example amenable to laboratory experiment. Bibliographical Notes ‘The fundamental ideas concerning the change of the vacuum state in strong electric fields are discus- sed in the following review articles: (Mu 76a, Re 77,84, Ra 78b, Br 78]. These ideas have been extended to other gauge fields, particularly to gravitational fields, in the review [So 82c]. A report on the status of experimental investigations up to 1977 is also contained in [Re 77]. ‘A more recent overview of the field has been given by Greiner in [Gr 83c]. Those interested in more details about these topics should consult Quantum Electrodynamics of Strong Fields (Proceedings of the Lahnstein conference 1981) {Gr 83a]. It contains a detailed account of the historical development and a survey of the status of experiments in the field up to 1981. For more general reading [Re 76a, Fu 79, Gr 80, 82] can be recommended.2. The Wave Equation for Spin-1/2 Particles Following Dirac’s original derivation the four-component relativistic wave equa- tion describing spin-1/2 particles is introduced. The energy spectrum, which exhibits two continua for both negative and positive energies, as well as the corre- sponding free spinor solutions are discussed. Finally we study in detail the trans- formation properties of Dirac-spinors under the Lorentz transformations and the construction of bilinear covariants. 2.1 The Dirac Equation In 1928, Dirac searched for a covariant relativistic wave equation of Schrédinger form [Di 28] notin 2.1 i 7 vy. (2.1) Because of the difficulties encountered then in interpreting the Klein-Gordon equation £? b= (6c? +m) or 7 2 a= (- WV? + mbct)b, (2.2) which did not allow for a positive definite TrobeBiiy density, Dirac required that (2.1) take a positive definite probability density @ ~ ¥*¥. The charge- density interpretation for the Klein-Gordon case was not yet known and, besides, would have made little physical sense, because the * and 2~ mesons as charged spin-0 particles had not yet been discovered [Gr 81b, Bj 65]. Since an equation of the form (2.1) is linear in the time derivative, linearity in the space derivatives is also required (homogeneity in space-time) and hence (2.1) must more specifically be of the form in OP a) MEG, OF 4 5, OF a OF) tBmetyl AY. 2.3) ax! ax? ax? The as yet unknown coefficients & and # cannot be c numbers, because otherwise (2.2) would not be form invariant, not even against simple space rotations. Probably &, and f are matrices (operators), indicated by the circumflex (“) above2.1 The Dirac Equation 27 the symbols. Therefore ¥ cannot be a scalar function, but must be a column vector wi@. 0) wee, y= | 29 | (2.4) yn, t) Let us already call it a spinor, even though the precise definition of a spinor is given below (2.90— 96). The positive definite probability density is then of the form " = yt =(y*. wt Ps Wa} x ay. ox, 1) = PPX, d) = (yi, yi... wr) : = Mi wi, t). (2.5) YN. To prove that (x, ¢) of (2.5) can be interpreted as a probability density, g(x, t) must be the time component of a current four vector, which obeys the continuity equation. The dimension N of the spinor is determined below. From (2.3, 4) we conclude that &;and # have to be N x N square matrices, so that on both sides of (2.3) a column vector with N components appears. This simply means that the Schrédinger-type equation (2.3) is, in fact, a system of N coupled linear differen- tial equations for the spinor components y,(x,¢), = 1,2...N. Using this, (2.3) can be written as +, 9vo_ he N/, 8 3 2X4 iA—2 = —_ — + 62—, + &: + MgC or i t\ ext ax 8. cue : 2 Oows N = EL ADave o=1,2...N. (2.6) oa Equation (2.3) is simply a short-hand notation of (2.6). To proceed, one naturally requires 4a) the correct energy-momentum relation for a relativistic particle, i.e. E?=¢*p?+ mict, (2.7) b) the Hamiltonian Ay of (2.3) should be hermitian, ) the continuity equation for the density (2.5), and a) Lorentz covariance (i.e. Lorentz form invariance) for (2.3), or respectively, (2.6). To fulfil requirement (a), every component yw, of the spinor ¥ has to obey the Klein-Gordon equation (2.2), i.e. ne Ovo _ (=We2v2+ mc4) (2.8) oe oC VWo- 72B 2, The Wave Equation for Spin-1/2 Particles Now, iteration of (2.3) yields pee Ae t itt ey ar? na 2 ax! ax! +m free 5 uh Ray o* e+ Binge! y (2.9) and comparing (2.8, 9) leads directly to the following conditions for the matrices @ and G0) + G6) = 25,1, ap+ pa; =0 (2.10) @ =P These anticommutation relations define an algebra for the four matrices. In order to fulfil condition (b) these matrices have to Hermitian: a= 4, B=8. Hence the eigenvalues of the matrices 4; and f are real, and since a? =f and BP = {| we conclude that the eigenvalues can only be +1. This is most easily seen in the eigenrepresentation of the matrices, where the matrices are diagonal. For example, & has then the form (2.11) AiO... 0 ,.|0 Ab 0 a 0 |? 0 0... 0Ay where A‘,(v=1,2...N) are the eigenvalues of &;. Because of 00...0 (Ai? 0 - 0 @ai- {919 : |_jo ayy? : i :01 0 | ~|: o |? Ow... 01 OE O(Ay)? then (iP =1 @ Al= +1. (2.12) Furthermore, the anticommutation relations of (2.10) a= - Baik indicate that the trace of the matrices @ and 2 must vanish. Since tr AB = tr BA tré)= +trp'a,= +trpa,h = -tra,2.1 The Dirac Equation 29 and hence tr @ = 0. Similarly, tr # = 0. Now, since the trace of a matrix equals the sum of its eigenvalues and because of (2.12), each of the matrices @;and f has to have an equal number of positive (+1) and negative (— 1) eigenvalues and the dimension of these matrices has to be even. The smallest even dimension N = 2 must be excluded, because in two dimensions only 3 anticommuting matrices exist, namely the 3 Pauli matrices 6;. We therefore try N = 4 and easily find an explicit representation for the &; and B matrices, obeying all conditions (2.10): , (0 6 en) an(9 2), a-(4 A) (2.13) where the three well-known Pauli matrices 6;(i= 1,2,3) and the 2 x 2 unit matrix Aare #-(9! ae ( 07 “\n oJ? °\i of? (2.14) * 1 0 10 G3= » = 0-1 o1 and obey the anticommutation relations [Bj 64, Gr 81b]: 6i8;+ 66; = 25,1. (2.15) Next we construct the current density and verify its continuity equation. Therefore (2.3) is multiplied from the left by the adjoint spinor ¥t= (vt V3, WE, wi), which gives 3 invt® yo FE § wa, 2 wi mge? viBy. 2.16) or ik 8x The Hermitian conjugate of (2.3) is 3 mal ae alte vp. (2.17) Multiplied from the right by ¥, it gives t ino? p= -Fe 5 ay’ ar 5 eat moe PRY, (2.18) i K=1 Ox* where the Hermiticity of the Dirac matrices (@f= &, B*= B) has been used. Subtracting (2.18) from (2.16) then yields cpt = Fe eB Se ( Pla) (2.19) or 82 gL a divj = 0 (2.20)30 2. The Wave Equation for Spin-1/2 Particles if we identify o=Yy and jkacHtay (2.21) as the probability density and current density respectively. The former is positive definite, as required. From the continuity equation (2.20) follows immediately ax yt y= [divi x= —[j-dF=0, (2.22) ory v F where Fis the surface of the volume V. The last step in this equation is valid for infinite volumes with the surface at infinity where the wave functions ¥(r—+ 0) vanish. Hence (2.22) expresses the conservation of total probability with time. We have tacitly assumed that j = {/',j?,/°} is a three-vector, i.e. transforms under three dimensional rotations like a vector in 3-space. This, however, must still be proven. One also expects that J= {cai} = coi P 7} (2.23) forms a four-vector, whose components transform via a Lorentz transformation from one inertial frame to another. This, as well as the form invariance of the Dirac equation (which is called covariance) under Lorentz transformations must still be demonstrated before we can accept the Dirac equation as a proper relativistic wave equation, Sect. 2.5. Note that (2.13) gives a special representation of the Dirac matrices and hence of the Dirac equation. The choice of the matrices G,, of (2.13) is not unique, however. This can be seen by realizing that any unitary transformation § leads to new matrices a =8$@8"', pr =Sps"', (2.24) which also fulfil (2.10). As an example we calculate the first of the anticommuta- tors (2.10) and find Gi} & + & &| = 86,8~'Sa,S~'+ $4)8-'Sa,5~' = S(Gj6;+ G6) S~' = §25,1) 8"! = 26,1. 2.2 The Free Dirac Particle To get a preliminary understanding of the physical content of the Dirac equation, let us investigate the solutions of the free equation (2.3), of the equation without potentials. We rewrite it in the form2.2 The Free Dirac Particle 34 1, OW rs 2B ih—-= (cB + moe BDv=A;Y, (2.25) it where p = {6', p”, 6°} = —ih{(8/8x"'), (8/8x?), (8/8x")} is the momentum oper- ator. Stationary solutions of (2.25) are found with the ansatz P(x, 1) = V(x) exp (- <= «) (2.26) with which (2.25) becomes E(x) = Aes. (2.27) The quantity ¢ characterizes the time development of the stationary wave function (x, ¢). For the Dirac equation it is identical with the energy [Gr 81b]. For many applications it is useful to split up the four-spinor (2.4) into two two-spinors g and y according to "ws =|" )/-({9 p= wl ( °) : (2.28) Wa Using now the explicit form (2.13) for the Dirac matrices, (2.27) becomes a %)=c(°%).6(% )+me(* °\(? (2.29) x 0 x o-1)\x ir a 2 eg=cé-prxt+me lg, Bx+mo : 9 2.30) ex=cé-po-—moc lx. Since [f, H;]_ = 0, states with well-defined linear momentum may arise (°)=(%)e(4o-s). (2.31) x Xo h which, after insertion into (2.30), yield the same equations as (2.30) for (go/xo) after replacing the operator # by the momentum p (which is a c number). Order- ing the resulting equations yields (e— moc?) 1 go— C6 - p xo = 0 lo 0 PX (2.32) C6 +p gt (et myc’) 1x =0, and leads to non-trivial solutions only if the determinant32, 2. The Wave Equation for Spin-1/2 Particles (e—moc?)1_ -cé-p (2.33) -cé:p (e+mpyc’)1 ¢ » vanishes. Using now the well-known relation for Pauli matrices [Gr 81b] (6-A)(@-B)=A-B1+ié-(A xB) (2.34) gives e=mbct+cp?, or e=4E,=AE, with (2.35) E,=+Verp +m, A= 41. The two signs A = +1 correspond to positive and negative energy solutions for the Dirac equation. For given ¢ = AE, one obtains from (2.32) c(é-p) w= wy. 2.36) OT pe + e 239) Thus yo is determined, if gp is known. We may choose most generally oo= (it) mu (2.37) uz with the normalization to =utu=ut Su =1 PoM% = UU =Ufuy+utu,=1. Here uy,u2 are c numbers. The complete solutions of the free Dirac equation, which are also eigenstates of the momentum operator f, now read A exp po x- £0 Yr th=N, | cé-p RARE LOR RESEDARAAND 8 (2.38) SF moc+ AE, Van Here A= +1 characterizes the positive and negative energies, respectively (€ = AE,), and N, is a normalization factor, which is readily determined from Jae.) Pyrat, thd?x = 54, d@—p') (2.39) to be 2 z moc +e moc’ t+ AE, N,= |/ Moc te = | f TOC Ae (2.40) 4 l 2e | 2AE, Note that the radicand is always positive. The energy spectrum corresponding to the solutions ¥, ,(x, ¢) is given by2.2 The Free Dirac Particle 33 Fig. 2.1. Energy spectrum of the free Dirac equation Epa = AEp = AY mpct+ c*p? (2.41) and is depicted in Fig. 2.1. It shows a gap of 2c? separating the positive energy continuum (€p1 = + / mech cp’) from the negative energy continuum (¢),_ = — J mbcr + cp 2), Interpretation of the states with A= —1 is considered in Sect. 2.3. Here we note that all states (2.38) are obviously also eigenfunctions of the linear momentum operator p: BYpa=P Pi. (2.42) For each momentum p two kinds of states exist, namely those with positive energy (A = +1) and those with negative energy (A = —1), both having the same absolute energy |é),=1| = |€p,--1| = Ep. We shall now show that still another quantum number, the helicity, exists, which helps classify the free solutions completely. First note that the operator Fe-(§ 3) (2.43) 0 6 commutes with the free Dirac Hamiltonian Ay [cf. (2.25)] and with the momentum operator [A,,2-6]-=0, [6,£-p]_ =0. ew) Now sa42-4(¢ ) (2.45) can be considered as the four-dimensional generalization of the spin operator § = hG/2 of the Pauli theory. The helicity operator A, is now defined as the projection of § along the momentum # (Fig. 2.2), i.e. j.- fg. 6.9.8 A,=—2) == B. (2.46) 2 IAl |p|34 2. The Wave Equation for Spin-1/2 Particles z e Fig. 2.2a,b. Electrons with (a) positive and (b) negative helicity. The double arrow symbolizes spin, and ~ rotation of the spinning electron s 7 b) : With (2.44) it is easily verified that A, also commutes with A; and p: IAs, A-=0, [4,6] =0. Without loss of generality we can assume that the spinor wave moves in the z direction. Then (2.47) p={0,0,p} and 2 h As=8e= > Es aif &0 (2.48) 2\0 6 1000 _h|0-100 “Z/0010 0 0 0-1 with the eigenvalues + A/2. The eigenvectors of A, are immediately recognized to be wm) (ey (2). (9 J, (2.49) 0 0 ny uy where uy = (( ). u_y= 7 are the standard eigenvectors of 6,. The clas- sification of the free states (2.38) of the Dirac equation can be completed by noting explicitly that Ug Pr ro = Na cé.p ug moc’ + AE, v0] L0.-1840] , (2.50) Asati, o=ti,2.3 Single-Particle Interpretation of Plane (Free) Dirac Waves 35 and that states with o= +1 have positive and negative helicity, respectively. Obviously «Pp.a0| Ppta'a') = O12! Sa0' (Dz — Pz) « (2.51) 2.3 Single-Particle Interpretation of Plane (Free) Dirac Waves Here we shall briefly discuss the single-particle interpretation of the Dirac equa- tion and its solutions. It is already evident from the occurrence of negative energy solutions that a single-particle interpretation in its full significance is not possible. Let us first note that the time-development factor e = AE» of (2.35) must be interpreted as energy. This can be proven by establishing the Lagrangian density gninee wrine WV: &— mol V BY, (2.52) from which the Dirac equation (2.25) can be derived in the standard way by applying the Euler-Lagrange field equations. From (2.52) also the energy- momentum tensor 7,,, and in particular its Too component and hence the energy of the system can be calculated. One finds [Gr 81b], see also Sect. 8.1, E=JTy(%p,1,0)°X = AEp= €. (2.53) This clearly indicates that the negative time-development factor e= —E, cannot be reinterpreted as positive energy (as for the Klein-Gordon equation) but must be negative energy. A single-particle theory cannot explain this observation at all. Dirac’s hole theory solves this dilemma. Assume that real electrons are described only by the positive energy states of (2.38), i.e. by the states Ya-+1,0- The states with negative energy shall all be occupied by electrons, Fig. 2.3. Because of the Pauli principle this also prevents an electron in a positive energy state from losing energy by some mechanism (e.g. radiation) and so dropping deeper and deeper into negative energy states. Naturally the question arises about the physical meaning of this Dirac sea. It is shown below (Chap. 4) that a hole in this sea represents a positron (antipar- ticle of an electron) and the sea itself characterizes the vacuum. It is, however, 4e electron ing state of positive energy me Pmee® Fig. 2.3. The states with negative energy (E < — mic?) are occupied by electrons and form the so-called Dirac sea, representing the vacu- um, because observable electrons occur generally only with positive energy36 2. The Wave Equation for Spin-1/2 Particles already clear that the interpretation of the negative energy solutions of the Dirac equation leads us away from the single-particle to the many-body interpretation. Figure 2.3 demonstrates this aspect vividly. Prediction of antiparticles has been one of the outstanding triumphs of relativistic quantum theory. 2.4 The Dirac Particle Coupled to Electromagnetic Fields — Non-Relativistic Limits and Spin of the Dirac Equation The electromagnetic fields are described by the four-potential AMX) = {AoX), AQ}, (2.54) which is coupled minimally to the particle through the substitution. (2.55) in order to ensure gauge invariance. Hence the Dirac equation with electromag- netic potentials becomes «(ar 7 a) PQ) = [<# (6 -£A «) + mc! Fe) (2.56) act c¢ ¢ or ins? = [<« (6 -£A 0) + eA(x) + al PO) = (Art Ain) PX) - The particle and field interaction is contained in the interaction Hamiltonian Ai = ~ c+ A(x) +eAo(x) = - Lv A(x) + eAolx), 57) c c where ax 1 a x, A]_ = — Ix, Ail - at A] ih x, Ail 4 4 Bune c2 c a =— |x, c&- p+ Bmoc*]_ = — |x, &- f]- (2.58) ih ih =cé is the relativistic velocity operator. Expression (2.57) is identical with the classical interaction of a point particle with electromagnetic fields. Let us now study the non-relativistic limit of (2.56) by again decomposing the four-spinor ¥ into the two spinors @ and x according to2.4 The Dirac-Particle Coupled to Electromagnetic Fields 37 p= ( s) : (2.59) Xx and rewriting (2.56) as ) + ma( 4p 2(%)\0g-4(% int (2) =c0 #( 1) +eAvo( ‘With the ansatz (abe) the rest energy is split off and (2.60) becomes in (2) c0-2(2) redo ?)—2mer(°). (2.61) at \ x ° x x Now consider the lower component of this equation first and assume that the kinetic energy is much smaller than the rest energy, i.e. |ih87/8t|< |moc?x | and also that the Coulomb energy is much smaller than the rest energy: leAo(x)x| <|moc?x|. Then RS Bet -Se ) . (2.60) Gh 2moc x= Q> (2.62) indicating that in the non-relativistic limit the x component of the four-spinor is much smaller than the g component: Inserting (2.62) into the equation for the upper component of (2.61) gives jn 2O@) FRG FR Os canny oa). (2.63) ot 2m With (2.34) the first term on the right-hand side is readily simplified: Gt G-R= +I (AXA) 2 = p-£a00) +ié- [(-av- £400) x (19-<4e)| Cc Cc Cc 2 = (6-409) - 2 ne.(V XA) c c -( e i eh =(p-—A(x)) -—4-B. ¢ c38 2. The Wave Equation for Spin-1/2 Particles Thus (2.63) becomes e in 29) _ ¢ € ) __eh ot 2mo 2moc &-B+eAo(x)} glx). (2.64) This is the well-known Pauli equation [Gr 81b]. The two components of g describe the two spin degrees of freedom. From non-relativistic quantum mechanics it is known that the Pauli equation describes particles with spin 1/2. Hence we can conclude that also the Dirac equation — as the relativistic gener- alization of the Pauli equation — has to describe spin-1/2 particles. This is especially so, since the spin, as an intrinsic property, should be present indepen- dently whether the particle moves with relativistic or non-relativistic velocities. 2.5 Lorentz Covariance of the Dirac Equation A proper relativistic theory must be Lorentz covariant, i.e. invariant in form when changing from one inertial frame to another. Lorentz transformations are briefly recapitulated, and for a deeper group theoretical discussion [Bj 64, Gr 81b] are recommended. Two observers A and B sitting in different inertial frames describe the same physical event in their respective space-time coordinates. The coordinates observer A uses to describe the event are x“, those of observer B for the same event are x“, Both coordinates are connected by a Lorentz transformation 3 (Y= YL apxt wax, x4 =O x4x7,x5} = (e649 z), (2.65) =0 7 X y= (Xo, X1,X25X3} = (Ct, — x, —Y, ~Z}, which is a linear, homogeneous transformation and the coefficients a, depend only on the relative velocity and on the spatial orientation of the two coordinate systems. The Lorentz transformations (2.65) preserve the distance between two space-time points, which can be expressed differentially by invariance of the length element ds? = dx" dx, = gydx"dx", — Quy = diag fl, —1,-1,-1}=9, (2.66) as dx"dx, = dx" dx,. (2.67) This yields ! dx" dxj, = afajdx" dx, = dx"dx,= djdx"dx,, and hence aay = 5 . (2.68)2.5 Lorentz Covariance of the Dirac Equation 39 wie) w(x) x°sct x’sct’ A System 5 B System x x (ga me) yx) =0 and (iv et 7 me) y'@)=0 ax! ax’ Fig. 2.4. Both A and B observe the same form of the Dirac equation for their respective spinors y(x) and y'(x') These are the orthogonality relations for the Lorentz transformations (L.T.). Proper and improper Lorentz transformations are distinguished; because of (eras)? = 1, which follows from (2.68), one can conclude that either det(af)= +1 (properL.T.) oo *) 7 p ) oF (2.69) det(a#) = —1 (improper L.T.) . The former are the proper L.T.’s and the latter improper. The proper Lorentz transformations consist of the group of those coordinate transformations from one inertial frame into another one which can be constructed from the identity by continuous infinitesimal operations (rotations, translations, etc.). The improper L.T.’s contain either a (discrete) space inversion or a (discrete) time inversion. Such discrete transformations cannot be constructed from the identity by succes- sive infinitesimal transformations. The determinant of their transformation coefficients is therefore negative. Our task is now to connect the measurerfients of observer A with those of observer B. Each observer performs the measurements in her own inertial frame. More precisely, a relation (transformation law) between y'(x’) and y(x) is required (Fig. 2.4). For a given w(x) for A, the as yet unknown transformation law must enable yw’ (x') of observer B to be calculated. The Lorentz covariance requirement now means that y(x) in the A system and y’(x') in the B system have to obey the respective Dirac equations in these systems. The Dirac equations in both systems must have the same form to ensure that both systems are equal in the sense of the principle of relativity. In the following it is more convenient to denote the Dirac equation in four- dimensional form. In this way the symmetry between the time coordinate x” = ct and the space coordinates is better exhibited. Starting with (2.25), we multiply it from the left by f/c, which gives for y(x) 8 ka. (ane + EPauin 7 me) v@e)=0,40 2. The Wave Equation for Spin-1/2 Particles and with the definitions =f, y'=faj, (@=1,2,3) (2.70) finally : 3 1 8 2 8 3 8 in( p92 4 yt yrs yA) w—mcw=0. 2.71 (> ax? f ax! 7 ax? ’ os)y — a Even though the y” are matrices, they are denoted without the operator sign (), because confusion is not possible. Also, the four-vector x” is simply denoted by x, as before, X= k xh xix ex, (2.72) so that the notation is simplified. The y” matrices allow a more elegant formula- tion of the anticommutation relations (2.10), namely yy Pyt= 294, (2.73) where 1 is the 4 x4 unit matrix. The y' (i= 1,2,3) are unitary ((y')'= (y')~') and anti-Hermitian (y'! = — y'). These statements follow from QiP= -1= -yyi* and hence Of=O) t= ~y!. (2.74) On the other hand, the y° matrix is Hermitian (t= y), as can be seen from OP = 1 yy (2.75) The y“ can also be denoted explicitly using (2.13, 70): : o_{1 0 ): r-(3 4): (2.76) A further shorthand notation used from time to time is the Feynman dagger, defined through 040 8 ial 040 Aw yAy=gwy'At= yA ¥ yA'=yAl-y-A, i=t and the Nabla dagger 08 1 35:8 8 vin y aa —+ Ty oO y 8 p= ty Ve (2.77) 8ct it 8x' c¢ Bt : 7 With that the free Dirac equation (2.71) can be simply written as2.5 Lorentz Covariance of the Dirac Equation 4 GAY - moc) yx) =0, (2.78) or with the four-momentum operator 6, = ih 0/8x": (B- moc) w(x) = 0. (2.79) The Dirac equation (2.56) with minimal coupling to the electromagnetic four- potential A“ reads in this notation (6 = <4 a me) y(x) =0. (2.80) Both 6“ and A“ are four-vectors, and consequently this is also true for the differ- ence p“—eA“/c. Therefore, in the following discussion of covariance, we restrict ourselves to the free Dirac equation (2.79) without loss of generality. 2.5.1 Formulation of Covariance (Form Invariance) Covariance of the Dirac equation (2.79) has two requirements. 1) A prescription allowing observer B to construct her own spinor y'(x’) from spinor w(x) reported by observer A must exist. The y’(x’) of B has to describe the same physical state as the w(x) of A. 2) According to the principle of relativity the physical principles, i.e. the basic physical equations, are the same in every inertial systemi> Hence y'(x’) must be the solution of a Dirac equation, which in x’ coordinates has the same form as (2.79), namely (inv aoe - me) vx!) =0. (2.81) According to the relativity principle the y’” must obey the same anticom- mutation relations (2.73), otherwise the inertial frames of A and B would not be equivalent. Thus phy ey y= 29th, (2.82) and also Otay", Gite ny, ie which means that for the same reasons the same Hermiticity and anti-Hermiticity relations for the Dirac matrices must be valid for both observers as in (2.74). ‘This, by the way, also assures Hermiticity of the free Dirac Hamiltonian (2.25) in the x’ system [Gr 81b, Go 55]42 2. The Wave Equation for Spin-1/2 Particles At = erty" (in 7 ) + y9mgc*, (2.84) ax’® Le. pt =A}. (2.85) Ina longer algebraic proof [Go 55, Gr 81b] one can show that all 4x4 matrices y'#, which obey (2.82, 83), are equivalent up to a unitary transformation U, i.e. y*=Oty40, Ot=07'.” (2.86) Now, since unitary transformations do not change the physics, without loss of generality we can use in the reference frame of observer B the same y matrices as in the reference frame of A. Hence we can set py“ = yp“, and (2.81) becomes (p' — moc) y'(x') =0, (2.87) where now p=ihyy =. (2.88) ax” Let us now turn to determining the transformation between y’(x’) and y(x). It must be linear, because Dirac equations (2.79, 87) are linear in the wave functions. Hence v'&') = ' 4x) = S@ yo) = S@ya"'x'), (2.89) with @ denoting the Lorentz transformation matrix @ = (a)). Here §@ isa4x4 matrix depending on ay, and acts on the four components of the spinor y(x). Via a’ it depends on the relative velocities and relative orientations of the coordinate systems of A and B. The inverse operator §~1(@) has to exist, because A must also be able to construct her spinor y(x) from the spinor y'(x’) of B (relativity principle: equal opportunity for both physicists). Therefore y@) = S*@y'&')= S$" @y'@x). (2.90) Because of (2.89) it follows that ve) =S@)y'@')=S@)y'@x), (2.91) and therefore $a) = 8(@"'). (2.92) This is a condition for §, which can now be constructed. To that purpose we start with the Dirac equation (2.78) of A: [ines - me] w(x) =02.5 Lorentz Covariance of the Dirac Equation 43 and insert w(x) from (2.90), which gives i - 3 - ryt [iss 1(a) agit mocS ‘| y'(x')=0. Multiplication with §(@) from the left gives in$(@) y"8-"(@) © — moe wi(x')=0, (2.93) Ox" using $(4)$~ 1@) = 1. Now we express 8/8x“ in terms of x’ coordinates (2.65): Bx? ox Be? BET ais and insert this into (2.93): [ims y"S~*(@)ayy aor me| v'(x')=0. (2.95) Because of the form invariance of the Dirac equation, (2.95) has to be identical with (2.87). Comparing both leads us to conclude $(@) y"S- "(aay = Y (2.96) or, equivalently, ay y" = $(4) y"S-"(@). 2.97 This is the most important equation for determining the $(@) operator. In fact, determining §(@) gives the solution of (2.97). Condition (2.97) holds for all Lorentz transformations, i.e. for proper and improper L.T.’s because in this deduction det (af) = +1 has not been used. As soon as we have shown that a solution of (2.97) exists, and as soon as we have constructed §(4), we have proven the Lorentz covariance of the Dirac equation. It will be shown that the group property §(@4') = §(@) §(@’) also holds. At this stage a spinor can be precisely defined, up to now introduced only as a column vector with four components. Jn general a wave function wir) va)=] oi wale) is called a Lorentz spinor with four components, if y(x) transforms according to (2.90) with the condition (2.97). Such a 4 spinor is also called a bispinor, because it is built up of two 2-spinors:44 2. The Wave Equation for Spin-1/2 Particles ( wv i) yale) wax) ) war) For §@ one has to expect novel structures, i.e. structures uncommon to, e.g., tensor analysis, because bilinear forms of the spinor field y(x) such as the four- current (2.21) yx) = fcoJ*}={ev'wewtaty} (k= 1,2,3) have to be four-vectors. The following shows how this is achieved. In general it is easier to construct the various group operators of a symmetry group for infinitesimal transformations. The operators for finite transforma- tions (rotations, translations, etc.) are then built up from consecutive infinitesi- mal operations. This procedure is adapted to construct $(4) and, therefore, to investigate first infinitesimal Lorentz transformations, given by the matrix a= 6,+A@,, where (2.98) 4o=—-Ao"”, (2.99) as can be easily concluded from the orthogonality relations (2.68) by neglecting terms of order (A w*”)?: ahal = 67 = (54+ Awt)(d2+ Aws) = 51+ HAwst fAwy =d7+Awlt+dAays. Because of (2.99) 6 independent elements 4w*” exist. Each of these group parameters (rotation angles) generates an infinitesimal Lorentz transformation. We investigate two examples here, after which it is straight-forward to proceed similarly for other cases: a) Aw = ~4o"=-—AB+0; allother Aw”=0. We then have Ao = 91,4a” = gn 4@? = -Aw= +ABp = Aw" =Aws=-Awl and Aw} = gig4w"=0 forall i=1,2,3, etc. Because of (2.65)2.5 Lorentz Covariance of the Dirac Equation 45 ct ot’ Fig. 2.5. Illustration of (2.100). The x! system moves with velocity Av=cAp Av = cAB relative to the x system x x! (&')! = (Bi+ A) 5152+ Aw}535,)x" = (Gj- ABST Sj A BOS.) x", which is explicitly (Fig. 2.5) (8°)! =x Apxt=x9- 42 x1, ¢ (x)! =x! A Bx? = (x7)! =x’, (2.100) (3) =x, The inertial frame of observer B (x’ system) therefore moves against the inertial frame of A (x system) with velocity 4 v = c4 along the positive x! axis, which follows from 1 1 ey =02 2 =X = AP aap. x° ct o¢ b) Secondly, 4aj=-A40"=Aw"=4g; allother 4w””=0. Here the Lorentz transformation (2.65) reads (%')! = (i+ 515; gt 535; (— 4 g)) x" or explicitly (=x, (x!)'=x!+Agx?, (@)' =x?-Agx', ey =x?. (2.101) This transformation is immediately recognizable as a rotation through the angle Ag around the z axis, which would read for finite angles (x')' = cos gx! +singx?, 2.102 (x?)' = -singx'+cosgx?. i )46 2. The Wave Equation for Spin-1/2 Particles 2.5.2 Determining the §@) Operator for Infinitesimal Lorentz Transformations Let us now return to determining the operator S@ = S(A0”). We expand § in a power series of A w”” and stop after linear terms, so that S(4w”) =41 SA a") = 14 ond wo”, Guy = Gms wot = Sy. (2.103) The factor i/4 is chosen to allow for a simple structure of the yet unknown coef- ficients 6,,,. One should clearly note that each of the six quantities Gy, is itself a 4x4 matrix! The same is true, of course, for the operator §. Inserting now (2.103, 98) into condition (2.97) gives a system of equations for 6,,: (8i+ Ae) y= (: - ead o*) Y (: + 7 ap o) or Aay"= 74 wo Gap y"— y"Gap)- (2.104) In the last step terms quadratic in 4 w”” have been neglected. Using the antisym- metry equation (2.99) 4w””= —Aw"™ it is straightforward to derive from (2.104) the relation [Gaps »"]- = —2iGare- 9p ¥a) » (2.105) which determines Gy. Since 6, has to be antisymmetric in a and , see (2.103), it is plausible to try the ansatz a i Sap= > er yal» (2.106) which indeed solves (2.105), as a small calculation demonstrates [Gr 81b]. With that the §(@) operator for infinitesimal Lorentz transformations is now deter- mined and according to (2.103) reads S§(4@") = 1- 6y4o"= 142 lp nl-40". (2.107)2.5 Lorentz Covariance of the Dirac Equation 47 2.5.3 The §(@) Operator for Fi We now construct §(@) for finite, proper Lorentz transformations. To obtain the finite Lorentz transformation (2.65) from the infinitesimal one (2.98), first Aw), is denoted in the form ite Lorentz Transformations Aaj= Aoki» (2.108) where Aw is the infinitesimal parameter of the Lorentz group (“infinitesimal rotation angle”) around an axis in m direction. The 4x4 matrix (space and time components) for a unit Lorentz rotation around the 7 axis is Qa) ie For example, the Lorentz rotation (2.100) is a Lorentz transformation with a velocity along the x axis Av = c 4 which can easily be brought to the form (2.108): ah = dy+ Aw. = d1+ Awl.) = 51-4 B(d152+ 6353) , and thus with Aw= —- AB 0-1 0 0 oz= —(8ta2+ 35 = | ~4 0 0 eo (2.109) 0000 Obviously only the matrix elements C= Edo= — "= + (Ey = 4 are non-vanishing. The following relations also hold: 1000 r= 10 5b 8] and t=. (2.110) 0000 2.5.4 Finite, Proper Lorentz Transformations The algebraic properties (2.110) are useful for determining according to (2.65, 98) finite, proper L.T.’s from infinitesimal ones (2.108). From (2.100) ’ a (ey = tim (1428) (4428) 2.008 Neve Nay N Jas Nv = lim [(: +21) | xl = (ey x = [cosh (wf,) + sinh (wf, Vx" Now a = [(: + fol? + Loly* a) + (Se + oly" + | xe 2! 4! i! 3! ny48 2. The Wave Equation for Spin-1/2 Particles ot wo pyr, @ pa a, 7 -|(0 a5 + tet ptapte (i) | x" = [1-(h)?+ (cosh w)(f,)? + (sinh w) (F,))x4 . This reads explicitly x coshw —sinhw 0 0 x? x!) _ | -sinhw coshw0 0 x! ree 0 10] | x? ]> x 0 o 01) Lx or x°" = (cosh w)x°~ (sinh w)x! = cosh w(x°— (teh w)x!) , Wh (gt 0. 1 1 0: x°' = —(sinh @)x°+(cosh w)x* = cosh w(x'—(tghw)x”), gait The finite Lorentz rotation angle w can now easily be expressed in terms of the relative velocity v, = cA B, of the two inertial systems. Namely, the origin of the x' systems is described in the x system by x''=0 = coshw(x'~(tghw)x°), from which 1 1 <= = +: =tsho=2 (2.112) x ct c follows. Hence cosh w = OP = (2.113) / cosh? w— sinh? w V1 ~tgh? wo 1-2 and therefore (2.111) becomes 0, _ x°- Bx! (2.114) which is the Lorentz transformation for finite velocities v = cf. This procedure for constructing finite Lorentz transformations from infinitesimal ones can be2.5 Lorentz Covariance of the Dirac Equation 49 generalized to motions and rotations along an arbitrary axis. Six matrices (gen- erators) (f,)4 exist corresponding to the six independent Lorentz transformations. They are the four-dimensional generalizations of the well-known generators of three-dimensional space rotations (angular momentum operators) known from non-relativistic quantum mechanics [Bj 64, Gr 81b]. 2.5.5 The § Operator for Finite Lorentz Transformations Now, finally, we establish the spinor transformation operator for finite “rotation angles” starting from (2.103, 108). We shall perform infinitely many such infini- tesimal transformations, one after the other, so that : N v1) = $(@) v2) = im (: “74 ath”) ve) waa b ona yw). (2.115) For the special “rotation” (2.109), which corresponds to a Lorentz transforma- tion along the x axis v'@')= exp {- q olny" + &10 aa} ve) = exp {- F coldin +1) + G,0(— »3| v@) = exp {- + wea) w(x) =SLv@), (2.116) because, according to (2.103) 649 = — Go, and according to (2.109) ("= 41 and (f,)'°= — Similarly, for the rotation around the z axis using of (2.101) Aa,= Aoki» 00 00 Ci = | 4 00 oo4ro eccoo v' x’) = exp +e 06 y(F.)" ‘ven [- of - = PLA 2(f,)'? + Gr (F,)"1 i vx)50 2. The Wave Equation for Spin-1/2 Particles i. = exp {- i 1G i2(~1) + G21 (+ oi} yx) = exp lz ea, v(x) = exp G 0a") wx) =Sgy(x). (2.117) With the explicit 6% (2.106) we easily find igayy iff 0 %\/ 0 ®\_/ 0 &\/ 0 & gly orl- CS, 0 Vie ) an YC, 0 )| _i[f-am o \_/-ne o 21, 0 ea 0 ~a6, 6162-6264 0 0 6 G2— 6264 _i/2i0 \_/ 60 (2.118) 2\0 2143 0 8)’ : where the commutation relations (2.15) were used for the 2 x2 Pauli matrices. The similarity between the spinor transformation (2.117) and the transforma- tion of two-component Pauli spinors under space rotations [Gr 81b] should be noted: ee) =219( 50-4) g(x). (2.119) 2 y \ —i a Accordingly, the w”” can be interpreted as covariant angle-variables, because they enter the Lorentz transformations (2.98) in a similar manner as the com- ponents of the rotation vector @ enter the three-dimensional space rotations. The occurrence of half the angles in the transformation law (2.117) causes the doubled value of the spinor transformations. A spinor transforms into itself after arotation through 47 (not 2z!). Because of this, physical observables in a spinor theory (like the Dirac theory) have to be bilinear in y (or of even power in y). Only in this way can an observable turn into itself under a rotation through 27, as experience shows. According to (2.115) the transformation operator Sg for space rotations of spinors is Sp(@y) -o0(- ia0°), 57=1,2,3. (2.120) It is a unitary operator, because the 6, (i, j = 1,2, 3) are Hermitian and Sh= exp (49 a) =exp (ve) =S,'. (2.121)2.5 Lorentz Covariance of the Dirac Equation st For true Lorentz transformations (“Lorentz boosts”, i.e. transformations to a moving coordinate system) this does not hold. This can be seen, for example, for the Lorentz transformation operator $, as it appears in (2.116): Si-e(-L04) -e( +24) =S[+S,'. (2.122) Here we used Gor = Com 1590) = BB — Bia B) = (eat &) = ids. It is not difficult to show that in this case $1 = yoSly0. (2.123) and that this relation holds both for true Lorentz transformations and space rotations, i.e. that in general [Bj 64, Gr 81b] S-'= Sy. (2.124) This formula will be used repeatedly. 2.5.6 The Four-Current Density In (2.20—23) the four-current density was introduced U= Us levlycytay}=(eyty yy}, or simply J) = cyt) yy"wed =cVX) "WO, w= ty, (2.125) but it was not shown that it is really a four-vector. We can do this now by demon- strating that (2.125) transforms under Lorentz transformations as JME) = cy x) py yty'&) = cy"(x) Sty yt Sy(x) =cy"(x)y°S!p"Sw(x) [because of (2.124)] (2.126) = catyt(x) py’ye) [because of (2.97)] = ap j"(x). This is indeed the transformation law of a four-vector as in (2.65). Also the con- tinuity equation (2.20) can now be denoted in Lorentz covariant form as BM 9, (2.127) 8x"52 2. The Wave Equation for Spin-1/2 Particles This proves in particular that the probability density JP) = co(x) = cw yx) is indeed the time component of a four vector. In the following it turns out to be useful to abbreviate the often occurring combination vayty®. (2.128) Here y is called the adjoint spinor. Its transformation law under Lorentz trans- formations follows from (2.89, 123): WD = WTO)” = (Syenty= yeasty? =Y@yS'= pas". (2.129) 2.6 Spinor Under Space Inversion (Parity Transformation) Next we consider the improper Lorentz transformation of space inversion defined by x'=—-x, (2.130) t=t. The corresponding transformation matrix is 1000 0-1 00 y ai=| 9 0-1 0 | 79" (2.131) 0 0 0-1 Also in this case the Dirac equation must be covariant, because (2.130) is nothing else than a special case of the general Lorentz transformation (2.65). Therefore, all the general considerations from the last section can be taken over, except those based on infinitesimal transformations. The space inversion (2.130) cannot be constructed by a sequence of infinitesimal operations out of the identity. The corresponding operator for spinor transformation is called B (from “parity”), i.e. in this special case § = P. The general condition (2.97) also holds now and reads ayy = Py Pm or 4 as ayy" = Bal yP ! @ ds! = PY g%y"P-! @ Bl y°B = gy". v=0 Setting a P'y'P = gry’. (2.132) v, this finally becomes2.6 Spinor Under Space Inversion (Parity Transformation) 53 One should note that no summing over v occurs on the rhs. This equation has the simple solution Baeity®, Pt=eoity?, (2.133) where is still an arbitrary phase. In analogy to the proper Lorentz transforma- tions (2.118), where only a rotation through 4 z yields the identity transformation for the spinor, we request that four inversions reproduce the spinor, i.e. Pty = y= ee tys ety, 2.134) then (e'i%t= 1. Hence c= +1, +i. (2.135) The operator P of (2.133) is unitary: Bh =¢-ivy° = Bt, (2.136) and also obeys (2.124) Bos yPty?, (2.137) The spinor transformation under space inversion reads explicitly ve) = y'(-x, 1) = Py) = ely, n). (2.138) In the non-relativistic limit y reduces to v( 4) and therefore becomes an eigenstate of P. In general, with 0 we) = | 2b) wax) |? vate) (2.139) wi) wi) Bf] 20) | _ pie | v2) ws) — wax) wa) — ax) Thus, the eigenstates with positive energy in the non-relativistic limit have opposite P eigenvalues to those of negative energy. One speaks, therefore, of opposite “inner” parity of these states. Other improper Lorentz transformations still exist, the most important of which is time reversal, Sect. 4.3.54 2. The Wave Equation for Spin-1/2 Particles 2.7 Bilinear Covariants of Dirac Spinors Sixteen linearly independent 4x4 matrices must exist, here denoted by Cag, = 1,2...16. (2.140) It turns out that it is possible to construct 16 such linearly independent matrices from the Dirac matrices and their products. They are B=4, l= ys Fiv= Gw= — Fy (1) [4] {6 Paipy yysyey, LP=ys% (2.141) {1 (4] The numbers in brackets [---] indicate the number of a particular type of matrices. The superscripts S, V, T, P, A stand for scalar, vector, tensor, pseudo- scalar and axial vector, respectively. This nomenclature will become transparent below. These matrices (2.141) have some important properties. a) For each /*” y= +14. (2.142) This is obvious for £%, also for (¥)* = y= g (because of (2.73)), and can be checked straightforwardly for all other matrices (2.141). b) For each / (except /*S) at least one /” exists such that repr = -P"p". (2.143) av v This is illustrated just for FY, for which (YY = % for v + u because of the fjy and similarly for all anticommutation relations (2.73). Also fj,f° = -I other matrices. Relation (2.143) is important, because from it and (2.142) follows aft= —P"prpn = + Phy? and therefore, since tr4B = tr BA, eur = 1 hF"y = hh")? = (2.144) In other words, all the matrices (2.141) except f° have vanishing trace. ©) For given / and /* with a + b one can always find a /" + f such that Far? = fF", (no summation over n!) (2.145) where f2 is a c number. This is proved by a few examples: + : . FYAY = yyy = Mee ee wee = -idy= -ify.2.7 Bilinear Covariants of Dirac Spinors 55 The f factor is in this case f= —i. Similarly, one finds ~iPe Iu = —idmte for w=t a i : oo TiPo= > Yom Pr¥e) = 4 +47 Iau = idle — for w= +i, x+u,0,t for w+o#t, etc. Properties (a—c) allow the linear independence of the covariants (2.141) to be proved quite elegantly. Suppose a relation of the form La,t" =0 (2.146) n exists. Then, after multiplication with 7” + FS and taking the trace, 0= Yagtr"F" = ant" + Fa, fer n n=m =a,(+1) +0, because of (2.142, 144). Hence a, = 0 for all m +s. For f= /* similarly O=0 Laff" =atr(P+ ¥ agtrl” n nss =4a, +0, and thus also @, = 0. Hence all coefficients a, vanish necessarily if (2.146) holds, which means that *” are linearly independent. In other words, each 4x4 matrix (2.141) cannot be expressed through the others. Next we investigate the behaviour of bilinear forms of the type Yeah" we) (2.147) under Lorentz transformations. We shall need the obvious relation yyst psy! = 0 (2.148) from which follows [955 Gl - = — > (Psu Yu Yo Pu) — Nu Yo— PvPu) Ys) = 0 + (2.149) We know from (2.115) that spinors transform under proper Lorentz transforma- tions with the operator §(@) = exp ( - 4 en0%" = exp [ - = wan "| 7 (2.150) Because of (2.149) it follows that (S(4), 75]. = 0. (2.151)56 2. The Wave Equation for Spin-1/2 Particles The spinors are transformed under space inversion with the operator [see (2.133)] Bae'ty®, and it is immediately obvious that (B, y]_ =0 (2.152) holds. It can easily be computed that Wy’) = WTR w'&) = VI@)STYS w(x) = ylony’S'S ya) = Va) Ve), (2.153) showing that in all Lorentz systems V@) vo) = oP ye) (2.154) has the same value. It is an invariant or scalar. Similarly, it can be checked {Gr 81b] that UO) ysw’(™) = WO)S'ysS yx) = det(at) W&) ys yw) is a pseudoscalar, Vee’) yw! (%') = ane) phy) is a vector; UO) yyy!) = det (ah) ay x) y*y" W(x) (2.155) is a pseudovector; Y (Gy! (') = anapy(x) SG yx) is a tensor of second rank. “Pseudo” indicates that the corresponding quantity behaves as under proper Lorentz transformation, but changes sign under improper Lorentz transformations. 2.8 Gauge Invariant Coupling of Electromagnetic and Spinor Field Electrodynamics is gauge invariant, meaning that the electrodynamic laws, i.e. the Maxwell equations, are not changed by a gauge transformation of the vector potential A, (x) Ab(x) = Ay(x) + oe : (2.156)2.8 Gauge Invariant Coupling of Electromagnetic and Spinor Field 57 where x(x) is an arbitrary function. To ensure invariance of quantum mechanics under gauge transformations two more conditions are necessary. 1) The particle field (spinor field) y must transform under gauge transforma- tions according to ' ie y' (x) = exp [ = 7 09| wi), (2.157) i.e. w(x) obtains a space-time dependent phase x(x). 2) Then it is necessary to couple the y and A, fields minimally, meaning that everywhere in the equation of motion terms of the form f, y(x) are replaced by Bue) > (+. - £4.00) vi). (2.158) This then introduces (minimal) coupling of the two fields. For the Dirac equa- tion, where A, appears in the form »“f,, the minimal coupling term is given by Alin = 24 ulX). (2.159) One immediately recognizes that these requirements of gauge transformations and couplings indeed leave quantum mechanics gauge invariant, because ~ 2 at \ wo alin - & ox _ ie (>. £41) w' (x) (i» OxF £(4,+-24)) w(x) v0 Je x00] +, 9 e ie = h—-—A -— a ( ox? .) veo] cl he x00] (2.160) = [(0.- <4.) v)| es] - a xo] : Since the common phase factor exp[— (ie/Ac) x(x)] appears now in all terms of a wave equation, it drops out. In particular, for the Dirac equation (2.79) e y (0-41) y'(x)- mocy'(x) =0, (2.161) which is, because of (2.160), equivalent to yt (0.- £a,) w(x)— mocy(x) = 0. (2.162)58 2. The Wave Equation for Spin-1/2 Particles This then proves that y’ (x) together with A/,(x) obeys the same wave equation as w(x) together with A,,(x). The theory is thus gauge invariant. This fact shall be used often below. Bibliographical Notes In Chap. 2 we have summarized essential facts of elementary relativistic quantum mechanics. More depth is given by the standard text books [Bj 64, Sc 79, Ba 80b, Be 71, Gr 81b]3. Dirac Particles in External Potentials Chapter 3 treats the spectra and wave functions for a Dirac particle in various external potentials. It deepens our understanding and widens the range of applicability of the Dirac equation. 3.1 A Dirac Particle in a One-Dimensional Square Well Potential A potential of depth Vo and extension —a/2 2m. In Fig. 3.2 four energy domains can be distinguished. 1) E>mp. Free electrons travel from left to right and are scattered by the (attractive) potential. If the length of the potential well a is an integer multiple of the wavelength in domain II, a potential resonance appears, i.e. the probability for finding the electron within the well is especially large. 2) mo+ Vo< E < mo. The bound states are found within this energy range, char- acterized by an exponentially decaying probability in domains I and III. 3) -—mo+ Vo< E < —imo. Chapter 4 shows that negative energy states are to be interpreted as wave functions of antiparticles, ie. of positrons. Incoming positrons (continuum states with negative energy) “feel” a repulsive potential62 3. Dirac Particles in External Potentials 4 ® i (®t ® PAYLYMIADDDDMN ; 2 ANE b ta c Tree *Vy mee NS went 7 Tey 1 [etfs +N WA) SSS SS Nene es i , me? Fig. 3.2 eo Fig. 3.3 Fig. 3.2. The potential depth ¥ is smaller than 2m c. The qualitative behaviour of the large com- ponents of the wave function is shown for a few typical energies: (a) shows an electron wave function of positive energy, (b, c) represent bound states. The wave functions (d, e) are typical for the negative energy continuum Fig. 3.3. The potential depth ¥ is larger than 2m. The qualitative behaviour of wave functions with various energies is indicated in analogy to Fig. 3.2 (because of their opposite charge), at which they will be scattered. Since the prob- ability of finding such a particle in domain II decreases exponentially, the reflec- tion of the positrons is large (the transmission decreases with a and increases with |E)- 4) E< —mo+ Ko. The positrons are scattered at the repulsive potential. Again, potential resonances might appear if the extension of well a is an integer multiple of the wavelength in domain II. Figure 3.3 illustrates the case for |V| > 2s for which an additional energy interval with mo+Vo pair creation” (Chap. 6). Thus, while bound states in the energy domain — mp < E < mo can be empty without causing instability of the total system, it is not possible to keep bound states which “dived” into the negative energy continuum empty for a long time: they will be filled spontaneously. In other words, the hole in such states has a finite lifetime and the state therefore a finite decay width. It will be shown that in the energy interval mg+ Vo < E < — mono energetically sharp bound states (as for states in the interval —mg=E< mo) exist. The wave function of such states obtains a resonating structure, peaking around the expected binding energy of the bound state. After these qualitative considerations we shall now specify the statements made above via the solution of (3.8 — 11). We distinguish two cases: a) |E|>mo, i.e. kyisreal b) |El| 1. Hence, in this energy domain, the wave function vanishes identically. Equation (3.17) can be solved approximately by a graphical method, which suffices to obtain a qualitative survey about the behaviour of the bound states. We proceed with that first; afterwards we determine the exact energy spectrum of the bound states by numerical solution of (3.17). We know that ett, (ky= in). 3.18) ky, cotk2a = | &a) cot kxa——_> —, a Rend’ G Ski moo _ mo) . ky V V3 -2mVo, For Vo< —2mo the limit f (kK.+0) becomes imaginary. Hence for (3.17) a graph as depicted in Fig. 3.4 can be drawn. We see immediately that always at least one bound state exists, independent of the depth V of the potential, agreeing with the corresponding non-relativistic problem [Me 70], but contrasting to the corre- sponding three-dimensional problem; because of the angular momentum barrier, not every three-dimensional well contains a bound state, but only potential wells deeper than certain depth MY do. The numerical solution of (3.17) can be obtained for various sets of param- eters Vo, a. A resulting energy spectrum is shown in Fig. 3.5 for a= 10a. = 10A/moc. It is seen that, for increasing |¥| more and more bound states appear. At Vy = —2.04 moc? the potential becomes overcritical, i.e. the lowest bound state “dives” into the lower continuum, in which it becomes a resonance, as can be seen by inspecting the transmission coefficient (see below). Let us now consider the scattering states. Again several energy domains exist: a) k,and y are real for E> mand for E < —mo+ Vo. In the overcritical case the additional domain m+ Vo< E < — mp exists where the conditions are ful- filled, \ b) Kk and y are imaginary for Vo— my < E< V+imo.3.1 A Dirac Particle in a One-Dimensional Square Well Potential 65 Fig. 3.4 Fig. 3.5 Fig. 3.4. Graphical solution of (3.17). The intersections of the (~-—, —) yield the bound state energies Eigenvalue spectrum of electrons bound in a one-dimensional square well potential of 10 a,. (~ — —) show those energies at which the “dived states” resonate, as can be deduced from the maxima of the transmission coefficient (Fig. 3.6) At first we choose one of the constants A, A',C,C’ arbitrarily assuming that there is no wave impacting on the potential from the right-hand side, i.e. C’ = 0. Thus, C can simply be interpreted as that fraction of a wave with amplitude A moving from the left through the potential well. The term proportional to A’ is the wave reflected at the potential. It is now possible to define a transmission coefficient J and a phase shift d by Fes. (3.20) This means that the outgoing wave is reduced in its amplitude by the factor VT and shifted in its phase relative to the incoming wave by 6. From (3.10) for real kz 2 bg o2\2 ~1 T= |C =| cos*kpa + ity sin?k,a A 2y iy e @.21) |! +( » ) son si. 2y The phase shift 6 follows from [see (3.10)] aaaee 2 1 gibe-ikie = coskya—i( 1+? ) sink, (3.22) Vr 2y which can be written as66 3. Dirac Particles in External Potentials Fig. 3.6. The transmission coefficient for scattering states of a one-dimen- sional square well potential of depth V=-3mgc? and extension a= 10 A,. The energies of the bound states are indicated by (— - —) (1+?) sinkpa. (3.23) 2y (isin (6 — kya) +008 (5— kya)) = coskya—i VF Separating real and imaginary parts and eliminating VT yields 2 ~tan(kya—6) = 14” tankga, (3.24) or, respectively, 2 be kya-arctan (452 tank) . (3.25) y If, on the other hand, kis imaginary, (3.10) gives instead of (3.21) the following transmission coefficient 2 -1 r=|1+ (422) sintea] <1, (8.26) 2r where y = if’and k= ix. Instead of (3.25), the phase shift now reads _p = tanh ns) 7 (3.27) é= ka-arcan(# The transmission coefficient for a potential depth Vy) = —3:mpo and an exten- sion a= 10A,= 10 A/moc is plotted in Fig. 3.6. We have immediately chosen an overcritical potential, because the undercritical case will be the same as before except for the non-appearance of states in the domain mo+ % mo and in the negative energy continuum for energies above the potential barrier |E| > |Vo- mo| (E < —mo+ Vo) appear. Positrons (electrons of negative energy) with small kinetic energy (— 79+ Vo < E a/2). The Dirac equation then reads in the various domains68 3. Dirac Particles in External Potentials II: (@-f+Pmo)w=Ey, Ti: (@- p+ Bomt+h))w=Ev, (Y<0). (3.28) In contrast to vector coupling, which yielded in domain II the replacement of E by (E— Vo), we now have to replace mo by (17+ Vo) in this domain. While the vector coupling of the potential affected the electron (E > 0) and positron (E < 0) wave functions differently (if electrons were attracted by the potential, the positrons were repelled by the same potential and vice versa) and hence yielded an asymmetric energy spectrum (bound states existed for only one kind of particle, either for electrons or for positrons), the scalar coupling acts on particles and antiparticles alike. For vector coupling of a potential the particle couples to the charge (which is different for particles and antiparticles) while for scalar coupling the particle couples to the mass (which is the same for particles and antiparticles). In the last case we thus expect a symmetric energy spectrum, i.e. for electrons as well positrons bound states exist in the same potential. Therefore we expect an overcritical behaviour already for Vo < ~ mo; even in this relatively shallow potential electronic and positronic states may cross in principle. What happens afterwards is discussed in detail further below. Let us now proceed as for vector coupling. Again the momentum in domains T and III is given by Ki = E?—(m)’, (3.29) while in domain II Ki= E*-(mgt Ky)’. (3.30) We can also again denote the conditions for continuity of the wave function at x= —a/2 and x = a/2, respectively. In analogy to vector coupling we define ky (E+mo+ Vo) _ = 3.31 Etm) ke (E+ m0) sate Note that y1/y if E+ —E. With the same notation as for vector coupling (Sect. 3.1), (3.10) can be used, realizing that y is now given by (3.31). Thus, (4.) 1 ( (14 pele Cy pyeltti tae, A’) Gy \ == yy(e724— ei4y, = — y)(ei24— -ik20y Cc (14 pelea (4 — yeni thade ce}: (3.32) These equations are invariant under the transformation E + —E (i.e. y+ 1/y), which reflects the particle — antiparticle symmetry.3.2. A Dirac Particle in a Scalar, One-Dimensional Square Well Potential 69 The following cases are discussed separately: 1) |E|mo (scattering states of the electron and positron continuum) . We first consider the bound states, for which the following eigenvalue equation holds AEP gnikga — IT? gina (3.33) 1-y 1t+y This equation has solutions only as long as kp is real, i.e. |E| > |mo+ Vo or, equivalently E > mo+ Vo, for electrons and E < —(mo+ Vo) for positrons. Under these circumstances (3.33) again yields tankya = vs (3.34) where y=if’(TeR) and kycot ka = (a + ») =f(ka), (3.35) x where k, = ix;, respectively. Also this equation is, of course, symmetric in + E. One has for [Vol Vo+ VE <0 for |Wl +, — (corresponds to ky //—2 myc? V— Va), (3.37) 1E|> mg 2 2 ky) = Moot Vor hy <= 2mpeV-VA). (3.38) /—2mo¥o~Va-K >0 for |Kl>m . (3.36) so for KW >-2m, on [mV vey Famv-v \ | i ' iV 1 V A 1 I a @ ' I 1 t q q 1 Fig. 3.8. The graphical solution of (3.36)70 3. Dirac Particles in External Potentials heim, Fig. 3.9, Energy spectrum of the Dirac equation with a scalarly coupled one-dimen- sional square well potential of extension a= 104, as a function of the potential depth V a See a 9 1 =2¥ehing As long as |V|< mo, the graphical solution of (3.35) is similar to that for the vector-coupling case (Fig. 3.4), except that it must be continued symmetrically to negative energies, Fig. 3.8. Now, if |V| increases, the intersection of f(k2) with the ordinate axis in Fig. 3.8 shifts upwards. Correspondingly the lowest eigenvalues shifts to smaller values of ky. For (mmo Vo+ V3)/\/—2moVo~ Va > 1/a the lowest eigenvalue dis- appears at k, = 0, i.e. at |E| = |7m+ Vo| > 0. Hence the lowest bound state never reaches the energy E = 0. This can also be expressed in the following way: the deepest bound electron state will never cross a corresponding positron state, i.e. one which can be traced back to lower continuum. If |Vo|is further increased, the ordinate f(x) is also increased and, therefore, the higher lying states approach the eigenvalues k, = n7/a with the eigenenergies E* = (nn/a)"+(mo+Vo)*. For |Vo|—>27o f(k = 0) diverges, and the energies approach 2 an #- (22) +mb>m. a This means that for |Vo|->27 the bound states disappear. With increasing potential depth all bound states disappear, one after another. This property is graphically depicted in Fig. 3.9, which was obtained by numerically solving (3.35). This energy diagram is typical for a square well potential. One can in- vestigate this problem also in three dimensions and include a scalar Coulomb- type potential, Sect. 3.4 and [So 73a]. In this case all bound states approach the energy |E|=0 asymptotically (a' + &), so that again no crossings between electron and positron states appear (see Fig. 3.13). We thus find — in contrast to vector-coupled potentials — that for scalar coupled potentials spontaneous pair creation never appears, independent of the depth of the potential. This property can be traced back to the absolute positivity of the Dirac-Hamiltonian for scalar-coupled potentials [So 82c]. The qualitatively different behaviour of bound states for differently coupled a/r potentials can easily be understood by observing that due to scalar coupling the particles obtain an effective mass3.3 A Dirac Particle in a Spherical Well a Fig. 3.10. The effective, r-dependent mass for two scalarly coupled @/r potentials v=8 T a === ve Baa) merc? = |myc?+ V(r) |. (3.39) For the a@/r potential this effective mass is schematized in Fig. 3.10, together with some bound states. A region with mere < mo, in which bound states are possible, always exists. For increasing coupling strength @ the bound state wave functions shift towards larger r values; at the same time the effective potential minimum becomes broader, so that the energy values |E| are lowered and for a> approach |EZ|=0. Because merp=0 always, then |E|20 always, so that electron and positron states can never cross. It is true that for a— o the energy gap between electron and positron states decreases, but it never becomes zero. Hence energy- less spontaneous pair production is never possible. Let us shortly qualitatively discuss the scattering solutions with |E|> mo, again considering C’ = 0 (Sect. 3.1, after (3.19)], investigating the transmission coefficient T= |C/A|?. In the domain |E|> mp naturally |E|>mo+Vo, but generally not |E |? > (mo+ Vo)”, if Vo < — 2m. This means that for (Vy << — 27m) the electrons “feel” a potential barrier if they have energies in the energy interval My < E<|mot Vol. A similar result holds also for positrons. Thus, while for | Vo| <2 resonances appear in both continua, similar to those in Fig. 3.6 for energies E> mo, for |Vo| < 2m an additional energy domain arises, where electrons and positrons “feel” a potential barrier, thus lowering the transmission coefficient significant- ly. This corresponds to the situation of Fig.3.5 in the energy interval —MtVRo72 3. Dirac Particles in External Potentials In this section we shall at first keep c and fh explicitly in the equations, later h=c=1 will be introduced again. First the kinetic energy operator @-f is transformed into spherical polar coordinates using the identity V =e,(e,- V)—e,X (e;XV) =e, 9 i 2 xf or r with the orbital angular momentum operator (in units of h) f= This yields i@x V) = -i(re,xV). a-p= ~ia 8 -1aexty. or or Using the identity (@-A)(@-B) =A -B+it-(A xB) for A =e,, B= £ leads to @,(a-£) = e,- £ +if-(e,xL£) =i¢-(,xL) and, by multiplying from the right by ys = ( i ‘) to &,(6-£) =i@-(e,xL). Therefore (3.43) becomes a-p= ia, +i %e-D or r = -1a,(S+4-4,), or rior where. a= BG-L£+1). For the eigenfunctions of (3.40) we make the following ansatz: =( 90) x39) ai Can eae): (3.42) (3.43) 3.44) (3.45) (3.46) (3.47) (3.48) (3.49) thus separating the r and ¢-g motions. The 2-spinors z¥ fulfil [Ro 61, Gr 81b]:Bayt = —xxh, Bax = HK bas Sth = Ut, x 3.3 A Dirac Particle in a Spherical Well =(-1,1,-2,2,-3,3...). Hence the two coupled differential equations for g(r) and f(r) follow [E-V(r)— moc} g(r) = - (4+4) 4] so), [E-V(r) + moc] f) Instead of using f(r) and u(r) =rg(r), u(r) =rf(r), for which the differential Ir =(2414% -($+4+4) 00. g(r), it is often convenient to deal with equations read ae E+mc?- V(r) f(a) = r ea ar \ u(r) -E-me-viny) % u(r) r For constant potentials [V(r) = — Vo] (3.53) has the following solutions: a) If h?k?c* =(E+ Vo) )— mbc4 > 0: uy(r) = rlay ji, (kr) + am, (kr), x kr u(r) = — ———_. |x| E+ Vo+ moc’ where rae x for 7 —x-1 for —* for ba { x1 for lai (kr) + ann (kr), x>0 x<0’ -x>0 -x<0° b) If A?K?c? = mict— (E+ Vo)" > 0: u(r) =r |/ 2£" n Kr u2(r) =, 7 E+Vot moc’ [61K 1,412(Kr) + bal s1n{Kr)) , / [= 81K) ainKr)+ bly sinlKn). 73 (3.50) (3.51) (3.52) (3.53) (3.54) (3.55) (3.56) (3.57)74 3. Dirac Particles in External Potentials The), n,are the regular and irregular spherical Bessel functions, respectively; Kisiny T1412 are the modified spherical Bessel functions. Their asymptotic behaviour is given by . 1 1 2) —> ————-z', 1 Ginn 1 get’ 1 1 —I, Zz) —> ——____2", l 2g 2 <> Grant [Ekin > (-"On-Diiz", 2z 770 2 2 hvsin@) —> ~£ | 22 ne n | Eko ae 2Z pares Zz Before continuing, let us denote the explicit form of the x#(v, 9) of (3.49, 50): (3.58) nz) rye —(2/-1)1! 3.59) (3.60) XB, 9) = Lo etis un mm, w) Vy ym 9) Xi G.61) = 1/2, +1/2 where ia=(5) and wi?=(1) (3.62) are the standard spinors. Our aim now is to determine the bound states [Pi 69]. For those E > Vo+ mo, —imo

You might also like

• Application-Driven Quantum and Statistical Physics - A Short - Jean-Michel Gillet - 2019 - WSPC (Europe) - 9781786345578 - Anna's Archive

  

  100% (3)

  Application-Driven Quantum and Statistical Physics - A Short - Jean-Michel Gillet - 2019 - WSPC (Europe) - 9781786345578 - Anna's Archive

  335 pages
• General Physics 1: Quarter 1 - Week 1

  

  No ratings yet

  General Physics 1: Quarter 1 - Week 1

  37 pages
• An Introduction To Gauge Theories 1st Ed by Cabibbo, Maiani and Benhar PDF

  

  100% (1)

  An Introduction To Gauge Theories 1st Ed by Cabibbo, Maiani and Benhar PDF

  321 pages
• Clifford Algebras An Introduction Garling CUP 2011

  

  No ratings yet

  Clifford Algebras An Introduction Garling CUP 2011

  210 pages
• Sakurai, Advanced Quantum Mechanics

  

  100% (15)

  Sakurai, Advanced Quantum Mechanics

  344 pages
• An Introduction Into The Feynman Path Integral PDF

  

  No ratings yet

  An Introduction Into The Feynman Path Integral PDF

  94 pages
• Huang K. - Quantum Field Theory

  

  100% (16)

  Huang K. - Quantum Field Theory

  446 pages
• Introduction To SR and GR by Hans Stephani PDF

  

  80% (5)

  Introduction To SR and GR by Hans Stephani PDF

  418 pages
• Ta-Pei Cheng, Ling-Fong Li-Gauge Theory of Elementary Particle Physics Problems and Solutions-Oxford University Press (2000)

  

  100% (2)

  Ta-Pei Cheng, Ling-Fong Li-Gauge Theory of Elementary Particle Physics Problems and Solutions-Oxford University Press (2000)

  315 pages
• Field Quantization - Walter Greiner

  

  100% (6)

  Field Quantization - Walter Greiner

  458 pages
• GR 4 Science Reviewer - Water & States of Matter

  

  100% (13)

  GR 4 Science Reviewer - Water & States of Matter

  2 pages
• Advanced Quantum Mechanics

  

  100% (1)

  Advanced Quantum Mechanics

  371 pages
• Field Quantization - Walter Greiner & Joachim Reinhardt

  

  No ratings yet

  Field Quantization - Walter Greiner & Joachim Reinhardt

  458 pages
• Advanced-Quantum-Mechanics - Sakurai PDF

  

  No ratings yet

  Advanced-Quantum-Mechanics - Sakurai PDF

  344 pages
• Landau L.D., Lifshitz E.M. - The Classical Theory of Fields. Volume 2-Butterworth - Heinemann (1980)

  

  100% (1)

  Landau L.D., Lifshitz E.M. - The Classical Theory of Fields. Volume 2-Butterworth - Heinemann (1980)

  438 pages
• Joel Franklin-Classical Field Theory

  

  100% (2)

  Joel Franklin-Classical Field Theory

  218 pages
• FieldTh StatisticalPhysics

  

  No ratings yet

  FieldTh StatisticalPhysics

  332 pages
• Quantum Mechanics, Special Chapters - Walter Greiner

  

  83% (6)

  Quantum Mechanics, Special Chapters - Walter Greiner

  402 pages
• Quantum Mechanics An Introduction - Walter Greiner

  

  No ratings yet

  Quantum Mechanics An Introduction - Walter Greiner

  512 pages
• Advanced Quantum Mechanics, Notes Based On Online Course Given by Leonard Susskind - Lecture 1

  

  No ratings yet

  Advanced Quantum Mechanics, Notes Based On Online Course Given by Leonard Susskind - Lecture 1

  7 pages
• 40 - The Quantized Electromagnetic Field PDF

  

  No ratings yet

  40 - The Quantized Electromagnetic Field PDF

  21 pages
• Miransky, Dynamical Symmetry Breaking in Quantum Field Theories

  

  100% (2)

  Miransky, Dynamical Symmetry Breaking in Quantum Field Theories

  550 pages
• Witten Reflections On The Fate of Spacetime

  

  No ratings yet

  Witten Reflections On The Fate of Spacetime

  7 pages
• A Guide To Quantum Field Theory

  

  100% (3)

  A Guide To Quantum Field Theory

  112 pages
• Advanced Quantum Field Theory

  

  100% (1)

  Advanced Quantum Field Theory

  111 pages
• Quantum Weinberg PDF

  

  No ratings yet

  Quantum Weinberg PDF

  7 pages
• Classical Mechanics, Point Particles and Relativity - Walter Greiner

  

  100% (6)

  Classical Mechanics, Point Particles and Relativity - Walter Greiner

  510 pages
• CM Merged

  

  No ratings yet

  CM Merged

  230 pages
• Relativistic Quantum Field Theory

  

  100% (1)

  Relativistic Quantum Field Theory

  248 pages
• Dirac and The Principles of Quantum Mechanics - K.gottfried

  

  No ratings yet

  Dirac and The Principles of Quantum Mechanics - K.gottfried

  11 pages
• The Supersymmetric Dirac Equation

  

  100% (4)

  The Supersymmetric Dirac Equation

  216 pages
• Porat A Gentle Introduction To Tensors 2014 PDF

  

  No ratings yet

  Porat A Gentle Introduction To Tensors 2014 PDF

  87 pages
• Shahab Sahraee, Peter Wriggers - Tensor Calculus and Differential Geometry For Engineers With Solved Exercises-Springer (2023)

  

  No ratings yet

  Shahab Sahraee, Peter Wriggers - Tensor Calculus and Differential Geometry For Engineers With Solved Exercises-Springer (2023)

  686 pages
• A Modern Course in Quantum Field Theory: December 2018

  

  0% (1)

  A Modern Course in Quantum Field Theory: December 2018

  27 pages
• Introduction To Classical Field Theory

  

  100% (1)

  Introduction To Classical Field Theory

  235 pages
• Andrei Smilga - Lectures On Quantum Chromodynamics

  

  No ratings yet

  Andrei Smilga - Lectures On Quantum Chromodynamics

  352 pages
• W. Galleas - The Bethe Ansatz Equations For Reflecting Magnons

  

  No ratings yet

  W. Galleas - The Bethe Ansatz Equations For Reflecting Magnons

  18 pages
• Quantum Mechanics - An Accessible Introduction - Robert J Scherrer - 2, 2024 - WSPC - 9789810984854 - Anna's Archive

  

  100% (2)

  Quantum Mechanics - An Accessible Introduction - Robert J Scherrer - 2, 2024 - WSPC - 9789810984854 - Anna's Archive

  385 pages
• QFT Lecture Notes

  

  100% (1)

  QFT Lecture Notes

  175 pages
• Notes On Relativistic Quantum Field Theory: A Course Given by Dr. Tobias Osborne

  

  No ratings yet

  Notes On Relativistic Quantum Field Theory: A Course Given by Dr. Tobias Osborne

  97 pages
• How To Become A GOOD Theoretical Physicist

  

  100% (1)

  How To Become A GOOD Theoretical Physicist

  4 pages
• String Theory

  

  100% (2)

  String Theory

  39 pages
• Advanced Quantum Physics

  

  100% (1)

  Advanced Quantum Physics

  31 pages
• Lectures On General Relativity

  

  100% (1)

  Lectures On General Relativity

  63 pages
• Tensor and General Relativity

  

  100% (1)

  Tensor and General Relativity

  88 pages
• Lattice Gauge Theories An Introduction PDF

  

  100% (1)

  Lattice Gauge Theories An Introduction PDF

  603 pages
• Cambridge Quantum Mechanics

  

  100% (1)

  Cambridge Quantum Mechanics

  45 pages
• Josiah Dykstra-Essential Cybersecurity Science - Build, Test, and Evaluate Secure Systems-O'Reilly Media (2015)

  

  100% (1)

  Josiah Dykstra-Essential Cybersecurity Science - Build, Test, and Evaluate Secure Systems-O'Reilly Media (2015)

  266 pages
• Classical Mechanics - John Taylor - Free Download, Borrow, and Streaming - Internet Archive

  

  No ratings yet

  Classical Mechanics - John Taylor - Free Download, Borrow, and Streaming - Internet Archive

  3 pages
• General Relativity

  

  No ratings yet

  General Relativity

  94 pages
• FirstCourseGR Notes On Schutz2009 PDF

  

  100% (2)

  FirstCourseGR Notes On Schutz2009 PDF

  299 pages
• (Cambridge Monographs On Mathematical Physics) Olivier Babelon, Denis Bernard, Michel Talon - Introduction To Classical Integrable Systems - Cambridge University Press (2007)

  

  100% (1)

  (Cambridge Monographs On Mathematical Physics) Olivier Babelon, Denis Bernard, Michel Talon - Introduction To Classical Integrable Systems - Cambridge University Press (2007)

  616 pages
• Advanced Quantum Field Theory

  

  100% (3)

  Advanced Quantum Field Theory

  267 pages
• Intermediate Quantum Mechanics

  

  100% (2)

  Intermediate Quantum Mechanics

  89 pages
• Application-Driven Quantum and Statistical Physics - A Short - Jean-Michel Gillet - 2018 - World Scientific Publishing Europe LTD - 9781786345547 - Anna's Archive

  

  No ratings yet

  Application-Driven Quantum and Statistical Physics - A Short - Jean-Michel Gillet - 2018 - World Scientific Publishing Europe LTD - 9781786345547 - Anna's Archive

  300 pages
• Useful Relations in Quantum Field Theory

  

  100% (1)

  Useful Relations in Quantum Field Theory

  30 pages
• Hestnes 1986 - Clifford Algebra and The Interpretation of Quantum Mechanics

  

  No ratings yet

  Hestnes 1986 - Clifford Algebra and The Interpretation of Quantum Mechanics

  18 pages
• Quantum Electrodynamics of Strong Fields - Walter Greiner, Berndt Mueller, Johann Rafelski

  

  No ratings yet

  Quantum Electrodynamics of Strong Fields - Walter Greiner, Berndt Mueller, Johann Rafelski

  606 pages
• Texts and Monographs in Physics: Series Editors

  

  100% (2)

  Texts and Monographs in Physics: Series Editors

  604 pages
• Advanced Quantum Mechanics Sakurai

  

  100% (2)

  Advanced Quantum Mechanics Sakurai

  344 pages
• Advanced Quantum Mechanics - Sakurai

  

  100% (1)

  Advanced Quantum Mechanics - Sakurai

  344 pages
• Advanced Quantum Mechanics by J. J. Sakurai (Z-Lib - Org) - Text

  

  No ratings yet

  Advanced Quantum Mechanics by J. J. Sakurai (Z-Lib - Org) - Text

  344 pages
• Advanced Quantum Mechanics

  

  No ratings yet

  Advanced Quantum Mechanics

  344 pages
• Thermodynamics and Statistical Mechanics - Walter Greiner

  

  92% (12)

  Thermodynamics and Statistical Mechanics - Walter Greiner

  480 pages
• Quantum Mechanics, An Introduction - Walter Greiner

  

  100% (10)

  Quantum Mechanics, An Introduction - Walter Greiner

  512 pages
• Quantum Mechanics, Symmetries - Walter Greiner

  

  100% (2)

  Quantum Mechanics, Symmetries - Walter Greiner

  258 pages
• Classical Mechanics, Systems of Particles and Hamiltonian Dynamics - Walter Greiner

  

  100% (4)

  Classical Mechanics, Systems of Particles and Hamiltonian Dynamics - Walter Greiner

  563 pages
• Mechanical Vibration Individual Assignment - I Final

  

  No ratings yet

  Mechanical Vibration Individual Assignment - I Final

  3 pages
• Moving Coil Ammeter & Voltmeter

  

  No ratings yet

  Moving Coil Ammeter & Voltmeter

  3 pages
• Genscie Lesson Plan W4

  

  No ratings yet

  Genscie Lesson Plan W4

  16 pages
• Aits 1920 CRT Iv Jeem PDF

  

  No ratings yet

  Aits 1920 CRT Iv Jeem PDF

  20 pages
• II Puc Centum Schedule - 2024-25

  

  No ratings yet

  II Puc Centum Schedule - 2024-25

  70 pages
• 250 MM 350 MM: Check For Crack Width As Per Is 456-2000

  

  No ratings yet

  250 MM 350 MM: Check For Crack Width As Per Is 456-2000

  3 pages
• Experiment 3: Sublimation and Melting Point Determination

  

  No ratings yet

  Experiment 3: Sublimation and Melting Point Determination

  4 pages
• Vibration Analysis Rotating Equipment

  

  100% (17)

  Vibration Analysis Rotating Equipment

  161 pages
• Tutorial-1 (EEL 207) : August 7, 2013

  

  No ratings yet

  Tutorial-1 (EEL 207) : August 7, 2013

  1 page
• Science Worksheets

  

  No ratings yet

  Science Worksheets

  2 pages
• HSDT Test Specification

  

  No ratings yet

  HSDT Test Specification

  32 pages
• Water Potential Word Problems

  

  No ratings yet

  Water Potential Word Problems

  2 pages
• Spectroscopia Raman

  

  No ratings yet

  Spectroscopia Raman

  6 pages
• E5 Sol

  

  No ratings yet

  E5 Sol

  32 pages
• Notes 1

  

  No ratings yet

  Notes 1

  10 pages
• Dr. Vijaya Khader Dr. MC Varadaraj: Development Team

  

  No ratings yet

  Dr. Vijaya Khader Dr. MC Varadaraj: Development Team

  28 pages
• Lecture 28 Light Emitting Diode-III Output Characteristics

  

  No ratings yet

  Lecture 28 Light Emitting Diode-III Output Characteristics

  28 pages
• Stress Transformation: Tractors, Linear Algebra, and Circles Within Circles

  

  No ratings yet

  Stress Transformation: Tractors, Linear Algebra, and Circles Within Circles

  43 pages
• Crack Tip Opening Displacement (Ctod) in Single Edge Notched Bend (Sen (B) )

  

  No ratings yet

  Crack Tip Opening Displacement (Ctod) in Single Edge Notched Bend (Sen (B) )

  328 pages
• ODE Mathematical Physics

  

  No ratings yet

  ODE Mathematical Physics

  9 pages
• Fundamentals of Quantum Information Theory: Michael Keyl

  

  No ratings yet

  Fundamentals of Quantum Information Theory: Michael Keyl

  120 pages
• Galilean Transform

  

  No ratings yet

  Galilean Transform

  4 pages
• Lifeguard, Snell's Law, Fermat's Principle and Brachistochrone - ENG+ITA

  

  No ratings yet

  Lifeguard, Snell's Law, Fermat's Principle and Brachistochrone - ENG+ITA

  12 pages
• Flashcards - CP 16 Determine The Value of An Unknown Mass Using The Resonant Frequencies - Edexcel Physics A-Level

  

  No ratings yet

  Flashcards - CP 16 Determine The Value of An Unknown Mass Using The Resonant Frequencies - Edexcel Physics A-Level

  25 pages
• Grounding Grid System Calculation PDF

  

  No ratings yet

  Grounding Grid System Calculation PDF

  39 pages
• Hetero Assignment

  

  No ratings yet

  Hetero Assignment

  9 pages
• Geothermal Production Engineering Skill

  

  No ratings yet

  Geothermal Production Engineering Skill

  9 pages