**
**## * Worlf Scientific *

## Lecture Notes in Physics Series - Vol. 40

# Phase Space Picture of Quantum Mechanics

### --- Group Theoretical Approach ---

### Y. S. Kim

Department of Physics

University of Maryland

College Park, Maryland 20742, USA

### M.E. Noz

Department of Radiology

New York University

New, York lOO16, USA

### PREFACE

Quantum mechanics can take different forms. The SchrOdinger picture of
quantum mechanics is very useful in atomic and nuclear physics. The
Heisenberg picture is
the basic language for the covariant formulation of quantum field theory. Is there
then any need for a new picture of quantum mechanics? This depends on whether
there are branches of physics where the Schrodinger or Heisenberg picture is less
than fully effective.
Quantum optics and relativistic bound-state problems are relatively new fields.
In quantum optics, we deal with creation and annihilation of photons and linear
superposition of multi photon states. It is possible to construct the mathematics
of harmonic oscillators in the Schrodinger picture to describe the photon's states.
However, the mathematics becomes complicated when we attempt to describe gen-
eralized coherent states often called the squeezed states. Is there a language simpler
than the Schrodinger picture?

Quantum field theory accommodates both the uncertainty principle and spe-
cial relativity. However, it is less than fully effective in describing bound-state
problems or localized probability distributions. It is possible to construct models
of relativistic hadrons consisting of quarks starting from the Schrodinger picture of
quantum mechanics. The question then is whether it is possible to formulate the
uncertainty relations in a covariant manner (Dirac 1927).

The phase-space picture of quantum mechanics provides the answer to these
questions. Starting from the Schrodinger wave function, it is possible to construct
a distribution function, often called the Wigner function, in phase space in terms
of the c-number position and momentum variables. In this picture, it is possible to
perform canonical transformations as in the case of classical mechanics. This will
bring us a deeper understanding of the uncertainty principle.

This phase-space picture of quantum mechanics is not new. The earliest ap-
plication of the Wigner phase-space distribution function was made in quantum
corrections to thermodynamics in 1932 (Wigner 1932a). Since then, the Wigner
function has been discussed in many branches of physics including statistical me-
chanics, nuclear physics, atomic and molecular physics, and foundations of physics.
However, it is difficult to see the advantage of using the Wigner function over the
existing method in those traditional branches of physics.

In this book, we discuss applications of the Wigner function in quantum optics
and the relativistic quark model which are relatively new subjects in physics and
which still need a basic scientific language. From the mathematical point of view,
the Wigner function for the ground-state harmonic oscillator is the basic language
for these new branches of physics. However, its symmetry properties constitute the
most interesting aspect of this new scientific language.

Indeed, the symmetry property of the Wigner function in phase space is that
of the Lorentz group. The Lorentz group is known to be a difficult subject to
mathematicians, because it is a non-compact group. To physicists, group theory is
a difficult subject when its representations have no physical applications. However,
the situation is quite the opposite when the representation can extract physical
implications.

In this book, we discuss the physical consequences of the symmetries of the
Wigner function in phase space. This book is written for those scientists and stu-
dents who wish to study the basic principles of the phase-space picture of quantum
mechanics and physical applications of the Wigner distribution functions. This
book will also serve a useful purpose for those who simply wish to study the physi-
cal applications of the Lorentz group.

We are indebted to Professor Eugene P. Wigner for encouraging us to formulate
a group theoretical approach to the phase-space picture of quantum mechanics.
Professor Wigner suggested the use of the light-cone coordinate system for the
covariant formulation of the Wigner function. Indeed, Chapter 10 of this book is
based on Professor Wigner's ideas. He suggested the possibility that the work of
Inonu and Wigner (1953) on group contractions be extended to study the space-
time geometry of relativistic particles (Kim and Wigner 1987a and 1990a). He also
suggested the use of the concept of entropy when the measurement process is less
than complete in a relativistic system (Kim and Wigner 1990c).

While this book was being written, we received helpful comments and sug-
gestions from many of our colleagues, including K. Cho, D. Han, C. H. Kim, M.
Kruger, P. McGrath, H. S. Pilloff, L. Rana, Y. H. Shih, J. Soln, C. Van Hine, and
W. W. Zachary.

September 1990
YSK and MEN

- PHASE SPACE IN CLASSICAL MECHANICS (page 1)
- Hamiltonian Foml of Classical Mechanics
- Trajectories in Phase Space
- Canonical Transformations
- Coupled Harmonic Oscillators
- Group of Linear Canonical Transformations in Four-Dimensional Phase
Space
- Poisson Brackets
- Distributions in Phase Space.

- FORMS OF QUANTUM MECHANICS (page 19)
- Schrodinger and Heisenberg Pictures
- Interaction Representation
- Density-Matrix Formulation of Quantum Mechanics
- MixedStates 27
- Density Matrix and Ensemble Average
- Time Dependence of the Density Matrix

- WIGNER PHASE-SPACE DISTRIBUTION FUNCTIONS (page 37)
- Basic Properties of the Wigner Phase-Space Distribution Function
- Time Dependence of the Wigner Function
- WavePacketSpreads
- HamlonicOscillators
- Density Matrix
- Measurable Quantities
- Early and Recent Applications

- LINEAR CANONICAL TRANSFORMATIONS IN QUANTUM MECHANICS (page 57)
- Canonical Transformations in Two-Dimensional Phase Space
- Linear Canonical Transformations in Quantum Mechanics
- Wave Packet Spreads in Terms of Canonical Transformations
- HarmonicOscillators
- (2 + I)-Dimensional Lorentz Group
- Canonical Transformations in Four-Dimensional Phase Space
- The Schrodinger Picture of Two-Mode Canonical Transformations
- (3 + 2)-Dimensional de Sitter Group

- COHERENT AND SQUEEZED STATES (page 77)
- Phase-Number Uncertainty Relation
- Baker-Campbell-Hausdorff Relation
- Coherent States of Light
- Symmetry Groups of Coherent States
- Squeezed States
- Two-ModeSqueezedStates
- Density Matrix through Two-Mode Squeezed States

- PHASE-SPACE PICTURE OF COHERENT AND SQUEEZED STATES (page 99)
- InvariantSubgroups
- CoherentStates
- Single-Mode Squeezed States
- Squeezed Vacuum 107
- Expectation Values in terms of Vacuum Expectation Values
- Overlapping Distribution Functions
- Thomas Effect
- Two-Mode Squeezed States
- Contraction of Phase Space

- LORENTZ TRANSFORMATIONS (page 123)
- Group of Lorentz Transformations
- Little Groups of the Lorentz Group
- MasslessParticles
- Decomposition of Lorentz Transformations
- Analytic Continuation to the Little Groups for Massless
and Imaginary-Mass Particles
- Light-Cone Coordinate System
- LocalizedLightWaves
- Covariant Localization of Light Waves
- Covariant Phase-Space Picture of Localized Light Waves
- Uncertainty Relations for Light Waves and for Photons

- COVARIANT HARMONIC OSCILLATORS (page 145)
- Theory of the Poincare Group
- Covariant Harmonic Oscillators
- Irreducible Unitary Representations of the Poincare Group
- C-number Time-Energy Uncertainty Relation
- Dirac's Form of Relativistic Theory of "Atom"
- Lorentz Transformations of Harmonic Oscillator Wave functions
- Covariant Phase-Space Picture of Harmonic Oscillators

- LORENTZ-SQUEEZED HADRONS (page 167)
9.1 Quark Model
9.2 HadronicMassSpectra
9.3 Hadrons in the Relativistic Quark Model
9.4 Form Factors of Nucleons
9.5 Phase-Space Picture of Overlapping Wave Functions
9.6 Feynman's Parton Picture
9.7 Experimental Observation of the Parton Distribution

- SPACE-TIME GEOMETRY OF EXTENDED PARTICLES (page 193)
- Two-Dimensional Euclidean Group and Cylindrical Group
- Contractions of the Three-Dimensional Rotation Group
- Three-Dimensional Geometry of the Little Groups
- Little Groups in the Light-Cone Coordinate System
- Cylindrical Group and Gauge Transformations
- Little Groups for Relativistic Extended Particles
- Lorentz Transformations and Hadronic Temperature
- Decoherence and Entropy

- REPRINTED ARTICLES (page 217)
- E.P. Wigner, On the Quantum Correction for Thermodynamic
Equilibrium
- E.P. Wigner, On Unitary Representations of the Inhomogeneous
Lorentz Group
- P.A.M. Dirac, Unitary Representations of the Lorentz Group
- P.A.M. Dirac, A Remarkable Representation of the 3 + 2 de Sitter
Group

- REFERENCES (page 319)

### INTRODUCTION

The concept of phase space arises naturally from the Hamiltonian formulation of
classical mechanics, and plays an important role in the transition from classical
physics to quantum theory. However, in quantum mechanics, the position and mo-
mentum variables cannot be measured simultaneously. In the Schrodinger picture,
the wave function is written as a function of either the position or the momentum
variable, but not of both. For this reason, in quantum mechanics, the density ma-
trix (V on Neumann 1927 and 1955) replaces phase space as a device for describing
the density of states. It therefore appears that phase space is not a useful concept in
quantum mechanics. We disagree. The role of phase space in quantum mechanics
has not yet been fully explored.
Starting from the density matrix, is it possible to develop an algorithm of
quantum mechanics based on phase space? This question has been raised repeatedly
since the publication in 1932 of Wigner's paper on the quantum correction for
thermodynamic equilibrium (Wigner 1932a). Since it is not possible to measure
simultaneously position and momentum without error, it is meaningless to define
a point in phase space. However, this does not prevent us from defining an area
element in phase space whose size is not smaller than Planck's constant. Since the
measurement problem is stated in terms of the least possible value of the product
of the uncertainties in the position and momentum, it is of interest to see how the
uncertainty product can be stated in phase space.

The basic advantage of this phase-space picture of quantum mechanics is that
it is possible to perform canonical transformations, just as in classical mechanics.
The purpose of this book is to study the physical consequences derivable from
canonical transformations in quantum mechanics. Using these transformations, we
can compare quantum mechanics with classical physics in terms of many illustrative
examples. In addition, the phase-space picture of quantum mechanics is becoming
a new scientific language for modern optics which is a rapidly expanding field.
Furthermore, the Lorentz transformation in a given direction of boost is a canonical
transformation in the light-cone coordinate system. This allows us to state the
uncertainty relation in a Lorentz-invariant manner.

There are still many questions concerning the uncertainty relations for which
answers are not well known. For instance, in the Schrodinger picture, the free-
particle wave packet becomes widespread, and the uncertainty product increases
as time progresses or regresses. Is it possible to state the uncertainty
relation in terms of the quantity which remains constant? Can phase space provide an
answer to this question? The answer to this question is YES. In the phase space-
picture, the uncertainty is define in terms of the area which the Wigner distribution
function occupies. The spread of a wave packet is an area-preserving canonical
transformation in the phase-space picture of quantum mechanics.

Quantum optics is a rapidly expanding subject, and it is increasingly clear that
coherent and squeezed states of light will playa major role in a new understand-
ing of the uncertainty principle, and will provide innovations in high-technology
industrial applications. These optical states are minimum-uncertainty states, and
transformations among these state are therefore canonical transformations. Indeed,
the phase-space picture of quantum mechanics is the natural language for these
relatively new quantum states.

Most physicists these days learn classical mechanics from Goldstein's text book
(Goldstein 1980). However, Goldstein's book does not emphasize the importance
of linear canonical transformations, which are discussed in more advanced books
(Arnold 1978, Abraham and Marsden 1978, Guilemin and Sternberg 1984). In this
book, we shall discuss the group of linear canonical transformations in phase space
which is the inhomogeneous symplectic group (Han et ai. 1988). For a single pair of
canonically conjugate variables, the group is the inhomogeneous symplectic group
ISp(2), and it is ISp(4) for two pairs of conjugate variables.

If we do not take into account translations in phase space, the symmetry groups
become those of homogeneous symplectic transformations. The groups Sp(2) and
Sp( 4) are locally isomorphic to the (2 + 1 )-dimensional and (3 + 2 )-dimensional
Lorentz groups. Thus the study of the symmetries in phase space requires the study
of Lorentz transformations.

The Lorentz transformation is one of the most fundamental transformations in
physics, and this subject can be formulated in terms of the inhomogeneous Lorentz
group (Wigner 1939). Since this group governs the fundamental space-time sym-
metries of elementary particles, there are many papers and books on this subject
(Kim and Noz 1986). In this book, we treat Lorentz transformations as canonical
transformations.

One of the persisting question in modern physics is whether the uncertainty
relations can be Lorentz-transformed. Does Planck's constant remain invariant
under Lorentz transformations? Is localization of the probability distribution a
Lorentz-invariant concept? It is very difficult to answer these questions in the
Heisenberg or Schrodinger picture of quantum mechanics. The basic limitation of
these pictures is that they do not tell us how the uncertainty relations appear to
observers in different Lorentz frames. The question of whether quantum mechanics
can be made consistent with special relativity has been and still is the central issue
of modern physics.

We shall address this question within the framework of the phase-space pic-
ture of quantum mechanics. It is interesting to note that the Lorentz boost in a
given direction is a canonical transformation in phase space using the light-cone
variables. This allows us to state the uncertainty relations in a Lorentz-invariant
manner. Feynman's parton picture (Feynman 1969) and the nucleon form factors
are discussed as illustrative examples.

In the first two Chapters, we discuss the forms of classical mechanics and
quantum mechanics useful for the formulation of the Wigner phase-space picture
of quantum mechanics, which is discussed in detail in Chapters 3 and 4. Chapters
5 and 6 are for the applications of the Wigner function to coherent and squeezed
states of light. It is seen in these chapters that the study of the Wigner function
requires the knowledge of the Lorentz group.

In Chapters 7 and 8, we present a detailed discussion of the physical represen-
tations of the inhomogeneous Lorentz group or the Poincare group which governs
the fundamental space-time symmetries of relativistic particles. By constructing the
representation based on harmonic oscillators, we study the phase-space picture of
relativistic extended particles. Chapters 9 contains a detailed discussion of experi-
mental observation of Lorentz-squeezed hadrons. Finally, in Chapter 10, we discuss
some fundamental issues in space-time symmetries of relativistic system, including
the unification of space-time symmetries of massive and massless particles and the
entropy increase due to the incompleteness in measurements.

Since we are combining the Wigner function with group theory, we have
reprinted in the Appendix Wigner's 1932 paper on the Wigner function as well
as his 1939 paper on the representations of the inhomogeneous Lorentz group. The
study of phase space requires a knowledge of harmonic oscillators. P .A.M. Dirac
was interested in constructing representations of the Lorentz group based on four
dimensional harmonic oscillators. We have therefore included Dirac's 1945 paper
on the Lorentz group and his 1963 paper on the de Sitter group.

There are many other interesting subjects which can be studied within the
framework of the phase-space picture of quantum mechanics but are not discussed
in this book. However, there are now a number of review articles (Wigner 1971,
O'Connell 1983, Carruthers and Zachariasen 1983, Hillery et al. 1984, Balazs and
Jennings 1984, Littlejohn 1986) containing applications of the Wigner phase-space
distribution function to various branches of modern physics. The scope of this book
is limited to the simplest form of the Wigner function with maximum symmetry
applicable to the branches of physics in which the phase-space picture is definitely
superior to other forms of quantum mechanics.