Springer Series in

Nuclear
and Particle Physics
Springer Series in Nuclear and Particle Physics
Editors: Mary K. Gaillard . J. Maxwell Irvine . Erich Lohrmann . Vera Liith
Achim Richter

Hasse, R. W., Myers W. D.

Geometrical Relationships of Macroscopic Nuclear Physics

Belyaev, V. B.
Lectures on the Theory of Few-Body Systems

Heyde, K.L.G.
The Nuclear SheD Model

Gitman, D.M., Tyutin I.V.

Quantization of Fields with Constraints

Sitenko, A. G.
Scattering Theory

Fradkin, E. S., Gitman, D. M., Shvartsman, S. M.

Quantum Electrodynamics with Unstable Vacuum
Kris L. G. Heyde

The Nuclear Shell

Model

With 171 Figures

Springer-Verlag
Berlin Heidelberg New York London
Paris Tokyo Hong Kong Barcelona
Professor Dr. Kris L. G. Heyde
Laboratorium voorTheoretische Fysica en Laboratorium voor Kernfysica, Rijksuniversiteit Gent
Proeftuinstraat 86, B-9000 Gent, Belgium

Editor:
Professor Dr. J. Maxwell Irvine
Department of Theoretical Physics
The Schuster Laboratory, The University
Manchester, M139PL, United Kingdom

ISBN-13: 978-3-642-97205-8 e-ISBN-13: 978-3-642-97203-4

DOl: 10.1007/978-3-642-97203-4
Library of Congress-Cataloging-in-Publication Data. Heyde, Kris L. G., 1942-. The nuclear shell model 1
Kris Heyde. p. cm.-(Springer series in nuclear and particle physics) Includes bibliographical references.
1. Nuclear shell theory. 2. Nuclear models. 3. Nuclear structure. I. TItle. II Series. QC793.3.S8H48 1990
539.1'43-dc20 90-9596
This work is subject to copyright. All rights are reserved, whether the whole or part of the material is con-
cerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting,
reproduction on microfilms or in other ways, and storage in data banks. Duplication of this publication or
parts thereof is only permitted under the provisions of the German Copyright Law of September 9, 1965, in
its current version, and a copyright fee must always be paid. Violations fall under the prosecution act ofthe
German Copyright Law.
C Springer-Verlag Berlin Heidelberg 1990
Softcover reprint of the hardcover I st edition 1990
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.
2157/3150-543210 - Printed on acid-free paper
 To
Daisy, Jan, and Mieke
Preface

This book is aimed at enabling the reader to obtain a working knowledge of the
nuclear shell model and to understand nuclear structure within the framework of
the shell model. Attention is concentrated on a coherent, self-contained exposition
of the main ideas behind the model with ample illustrations to give an idea beyond
formal exposition of the concepts.
Since this text grew out of a course taught for advanced undergraduate and
first-year graduate students in theoretical nuclear physics, the accents are on a
detailed exposition of the material with step-by-step derivations rather than on
a superficial description of a large number of topics. In this sense, the book
differs from a number of books on theoretical nuclear physics by narrowing the
subject to only the nuclear shell model. Most of the expressions used in many
of the existing books treating the nuclear shell model are derived here in more
detail, in a practitioner's way. Due to frequent student requests I have expanded
the level of detail in order to take away the typical phrase " ... after some
simple and straightforward algebra one finds ... ". The material could probably
be treated in a one-year course (implying going through the problem sets and
setting up a number of numerical studies by using the provided computer codes).
The book is essentially self-contained but requires an introductory course on
quantum mechanics and nuclear physics on a more general level. Because of this
structure, it is not easy to pick out certain chapters for separate reading, although
an experienced practitioner of the shell model could do that
After introductory but necessary chapters on angular momentum, angular mo-
mentum coupling, rotations in quantum mechanics, tensor algebra and the calcu-
lation of matrix elements of spherical tensor operators within angular momentum
coupled states, we start the exposition of the shell model itself. Chapters 3 to 7
discuss the basic ingredients of the shell model exposing the one-particle, two-
particle and three-particle aspects of the nuclear interacting shell-model picture.
Mter studying electromagnetic properties (one-body and two-body moments and
transition rates), a short chapter is devoted to the second quantization, or occupa-
tion number representation, of the shell model. In later chapters, the elementary
modes of excitation observed in closed shell nuclei (particle-hole excitations)
and open shell nuclei (pairing properties) are discussed with many applications
to realistic nuclei and nuclear mass regions. In Chap. 8, a state-of-the-art illustra-
tion of present day possibilities within the nuclear shell model, constructing both
the residual interaction and the average field properties is given. This chapter
has a somewhat less pedagogical orientation than the first seven chapters. In the

VII
final chapter, some simple computer codes are included and discussed. The set
of appendices constitutes an integral part of the text, as well as a number of
exercises.
Several aspects of the nuclear interacting many-body system are not discussed
or only briefly mentioned. This is due to the choice of developing the nuclear
shell model as an in-depth example of how to approximate the interactions in a
complicated many-fermion system. Having studied this text, one should be able,
by using the outlined techniques, to study other fields of nuclear theory such as
nuclear collective models and Hartree-Fock theory.
This book project grew out of a course taught over the past 8 years at the
University of Gent on the nuclear shell model and has grown somewhat beyond
the original concept. Thereby, in the initial stages of teaching, a set of unpublished
lecture notes from F. Iachello on nuclear structure, taught at the "Kernfysisch
Versneller Instituut" (KVI) in Groningen, were a useful guidance and influenced
the first chapters in an important way. I am grateful for the many students who,
by encouraging more and clearer discussions, have modified the form and content
in almost every aspect. The problems given here came out of discussions with
them and out of exam sets: the reader is encouraged to go through them as an
essential step in mastering the content of the book.
I am most grateful to my colleagues at the Institute of Nuclear Physics in
Gent, in particular in the theory group, in alphabetic order, C. De Coster, J. Jolie,
J. Moreau, J. Ryckebusch, P. Van Isacker, D. Van Neck, J. Van Maldeghem, H.
Vincx, M. Waroquier, and G. Wenes who contributed, maybe unintentionally, to
the present text in an important way. More in particular, I am indebted to M.
Waroquier for the generous permission to make extensive use of results obtained
in his "Hoger Aggregaat" thesis about the feasability of performing shell-model
calculations in a self-consistent way using Skyrme forces. I am also grateful to
C. De Coster for scrutinizing many of the formulas, reading and critizing the
whole manuscript.
Also, discussions with many experimentalists, both in Gent and elsewhere,
too many to cite, have kept me from "straying" from the real world of nuclei. I
would like, in particular, to thank J.L. Wood, R. F. Casten and R.A. Meyer for
insisting on going ahead with the project and Prof. M. Irvine for encouragement
to put this manuscript in shape for the Springer Series.
Most of my shell-model roots have been laid down in the Utrecht school; I am
most grateful to P.J. Brussaard, L. Dieperink, P. Endt, and P.W.M. Glaudemans
for their experience and support during my extended stays in Utrecht.

Gent, March 1990 K.L.G. Heyde

VIII
Contents

Introduction 1

1. Angular Momentum in Quantum Mechanics ................ 4

1.1 Central Force Problem and Orbital Angular Momentum ..... 4
1.2 General Definitions of Angular Momentum ............... 12
1.2.1 Matrix Representations ......................... 13
!
1.2.2 Example for Spin Particles .................... 13
1.3 Total Angular Momentum for a Spin! Particle ........... 14
1.4 Coupling of Two Angular Momenta:
Clebsch-Gordan Coefficients ........................... 17
1.5 Properties of Clebsch-Gordan Coefficients ................ 20
1.6 Racah Recoupling Coefficients:
Coupling of Three Angular Momenta .................... 22
1.7 Symmetry Properties of 6j-Symbols . . . . . . . . . . . . . . . . . . . . 23
1.8 Wigner 9j-Symbols: Coupling and Recoupling
of Four Angular Momenta ............................. 25
1.9 Classical Limit of Wigner 3j-Symbols ................... 27
Short Overview of Angular Momentum Coupling Formulas 28

2. Rotations in Quantum Mechanics ......................... 31

2.1 Rotation of a Scalar Field-Rotation Group 0(3) ........... 31
2.2 General Groups of Transformations ..................... 35
2.3 Representations of the Rotation Operator ................. 37
2.3.1 The Wigner D-Matrices ........................ 37
2.3.2 The Group SU(2)-Relation with SO(3) ........... 38
2.3.3 Application: Geometric Interpretation
of Intrinsic Spin! . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
2.4 Product Representations and Irreducibility ................ 43
2.5 Cartesian Tensors, Spherical Tensors, Irreducible Tensors ... 45
2.6 Tensor Product ...................................... 47
2.7 Spherical Tensor Operators: The Wigner-Eckart Theorem ... 48
2.8 Calculation of Matrix Elements ............ . . . . . . . . . . . . 49
2.8.1 Reduction Rule I .............................. 50
2.8.2 Reduction Rule IT ............................. 51
Short Overview of Rotation Properties, Tensor Operators,
Matrix Elements . .. ..... ....... . ...... ... ...... . . .. ...... 52

IX
3. The Nuclear SheD Model .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
3.1 One-particle Excitations ............................... 54
3.1.1 Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
3.1.2 The Radial Equation and the Single-particle Spectrum:
the Harmonic Oscillator in the SheD Model ........ 61
3.1.3 illustrative Examples of Energy Spectra ........... 67
3.1.4 Hartree-Fock Methods: A Simple Approach ........ 70
3.2 Two-particle Systems: Identical Nucleons ................ 74
3.2.1 Two-particle Wavefunctions ..................... 74
3.2.2 Two-particle Residual Interaction ................ 77
3.2.3 Calculation of Two-Body Matrix Elements ......... 87
3.2.4 Configuration Mixing:
Model Space and Model Interaction .............. 101
3.3 Three-particle Systems and Beyond ..................... 108
3.3.1 Three-particle Wave Functions .................. 108
3.3.2 Extension to n-particle Wave Functions ........... 112
3.3.3 Some Applications: Three-particle Systems ........ 115
3.4 Non-identical Particle Systems: Isospin .................. 119
3.4.1 Isospin: Introduction and Concepts ............... 119
3.4.2 Isospin Formalism ............. . . . . . . . . . . . . . . . 121
3.4.3 Two-Body Matrix Elements with Isospin .......... 130

4. Electromagnetic Properties in the Shell Model ............... 136

4.1 General. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136
4.2 Electric and Magnetic Multipole Operators ............... 137
4.3 Single-particle Estimates and Examples .................. 139
4.4 Electromagnetic Transitions in Two-particle Systems ....... 145
4.5 Quadrupole Moments................................. 149
4.5.1 Single-particle Quadrupole Moment .............. 149
4.5.2 Two-particle Quadrupole Moment ................ 152
4.6 Magnetic Dipole Moment ............................. 153
4.6.1 Single-particle Moment: Schmidt Values .......... 153
4.6.2 Two-particle Dipole Moment .................... 156
4.7 Additivity Rules for Static Moments .................... 157

s. Second Quantization ..................................... 161

5.1 Creation and Annihilation Operators ... . . . . . . . . . . . . . . . . . 161
5.2 Operators in Second Quantization ....................... 165
5.3 Angular Momentum Coupling in Second Quantization ...... 169
5.4 Hole Operators in Second Quantization .................. 171
5.5 Normal Ordering, Contraction, Wick's Theorem ........... 174
5.6 Application to the Hartree-Fock Formalism ............... 177

x
6. Elementary Modes of Excitation:
Particle-Hole Excitations at Closed Shells ... . . . . . . . . . . . . . . . . 179
6.1 General. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179
6.2 The TDA Approximation ...............•............. 181
6.3 The RPA Approximation .............................. 186
6.4 Application of the Study of 1p - 1h Excitations: 160 ....... 192

7. Pairing Correlations: Particle-Particle Excitations

in Open-Shell Nuclei ........ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197
7.1 Introduction ........................................ 197
7.2 Pairing in a Degenerate Single j-Shell ................... 200
7.3 Pairing in Non-Degenerate Levels: Two-Particle Systems ... 204
7.4 n Particles in Non-Degenerate Shells: BCS-Theory ........ 206
7.5 Applications of BCS ................................. 217
7.5.1 Odd-Even Mass Differences, Elqp •••••.•••••••.•• 217
7.5.2 Energy Spectra ............................... 219
7.5.3 Electromagnetic Transitions ..................... 221
7.5.4 Spectroscopic Factors .......................... 226
7.6 Broken-Pair Model ................................... 229
7.6.1 Low-Seniority Approximation to the Shell Model ... 229
7.6.2 Broken-Pair or Generalized-Seniority Scheme
for Semi-Magic Nuclei ......................... 232
7.6.3 Generalization to Both Valence Protons and Neutrons 237
7.7 Interacting Boson-Model Approximation
to the Nuclear Shell Model ............................ 239

8. Self-Consistent Shell-Model Calculations . ......... ... .. . ... 254

8.1 Introduction ........................................ 254
8.2 Construction of a Nucleon-Nucleon Force: Skynne Forces 256
8.2.1 Hartree-Fock Bogoliubov (HFB) Formalism
for Nucleon-Nucleon Interactions
Including Three-Body Forces .................... 256
8.2.2 Application of HFB to Spherical Nuclei ........... 259
8.2.3 The Extended Skynne Force .................... 260
8.2.4 Parameterization of Extended Skynne Forces:
Nuclear Ground-State Properties ................. 265
8.3 Excited-State Properties of SkE Forces .... . . . . . . . . . . . . . . 271
8.3.1 Particle-Particle Excitations: Determination of X3 ••• 272
8.3.2 The Skynne Interaction as a Particle-Hole Interaction 279
8.3.3 Rearrangement Effects
for Density-Dependent Interactions
and Applications for SkE F<;>rces ................. 291

XI
9. Some Computer Programs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 298
9.1 Oebsch-Gordan Coefficients ........................... 298
9.2 Wigner 6j-Symbol ................................... 300
9.3 Wigner 9j-Symbol ................................... 303
9.4 Calculation of Table of Slater Integrals .................. 304
9.5 Calculation of c5-Matrix Element........................ 309
9.6 Matrix Oiagonalization ............................... 316
9.7 Radial Integrals Using Hannonic Oscillator Wave Functions. 320
9.8 BCS Equations with Constant Pairing Strength ............ 323

Appendix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 327
A. The Angular Momentum Operator in Spherical Coordinates . 327
B. Explicit Calculation of the Transformation Coefficients for
Three-Angular Momentum Systems ..................... 328
C. Tensor Reduction Formulae for Tensor Products ........... 329
O. The Surface-Delta Interaction (SOl) ..................... 331
E. Multipole Expansion of c5(rl - rz) Interaction ............ 335
F. Calculation of Reduced Matrix Element «1/2l)jIlY kll(1/21')j')
and Some Important Angular Momentum Relations ........ 338
G. The Magnetic Multipole Operator ....................... 342
H. A Two-Group (Degenerate) RPA Model ................. 343
I. The Condon-Shortley
and Biedenharn-Rose Phase Conventions:
Application to Electromagnetic Operators and BCS Theory 347
1.1 Electromagnetic Operators:
Long-Wavelength Form and Matrix Elements ........ 347
1.2 Properties of the Electromagnetic Multipole Operators
Under Parity Operation,TIme Reflection
and Hermitian Conjugation ....................... 348
1.3 Phase Conventions in the BCS Formalism ........... 353

Problems .................................................. 357

References ................................................. 367

Subject Index .............................................. 373

XII
Introduction

Approaching the atomic nucleus at low excitation energy (excitation energy less
than the nucleon separation energy) can be done on a non-relativistic level. H
we start from an A-nucleon problem interacting via a given two-body potential
Vi,j' the non-relativistic Hamiltonian can be written as

A A
H = L ti +~ L Vi,j ,
i=1 i,j=1

where ti is the kinetic energy of the nucleon motion. Much experimental evidence
for an average, single-particle independent motion of nucleons exists, a point of
view that is not immediately obvious from the above Hamiltonian. This idea acts
as a guide making a separation of the Hamiltonian into A one-body Hamiltonians
(described by an average one-body potential Ui) and residual interactions. This
can be formally done by writing

H = Ho + Hres, with
A
Ho = L {ti + Ud, and
i=1
A A
Hres =~ L Vi,j - L Ui .
i,j=1 i=1

It is a task to determine Ui as well as possible such that the residual interaction

Hres remains as a small perturbation on the independent A-nucleon system. This
task can be accomplished by modem Hartree-Fock methods where the residual
interaction with which one starts is somewhat more complicated such as the
Skyrme-type interactions (two-body plus three-body terms) which have been used
with considerable success. This process of going from the two-body interaction
Vi,j towards a one-body potential is drawn schematically.
Besides specific nucleon excitations in particular mass regions, coherent nu-
cleon motion appears that is not easily handled properly within a shell-model
basis. Nuclear collective vibrations, rotations and all possible transitional types
of excitations have been observed in rare-earth and actinide nuclei. In these situa-
tions, model concepts stemming from analogies with well-known classic systems
have been used and developed to a high level of art (Bohr-Mottelson). It is now
MICROSCOPIC MODELS

~ ~

,
I
I

\ r::Ii'-r.1
I-~---'_....L..---J'--....L.._"';" J ~ rj

~IBIHF

a)

MACROSCOPIC MODELS : SHAPE VARIABLES

~~
-Lj R 3

~
..;: );~; .~. ¥)~ .:
. _0.
··" '7
·, ·A .
2

(a) In a schematic way, we illustrate the fact that starting from the two-body interactions Vi,i in
the nuclear A-body problem, one can construct an average one-body field, expressed by Ui. In this
illustration, specific radial shapes for both Vi,i (short-range repulsion -long-range attractive one-pion
exchange tail OPEP) and Uj (Woods-Saxon type) are given. The connection is established through
the (Brueckner) Hartree-Fock method
(b) Dlustration of some specific collective modes of motion when a macroscopic (shape) model
is considered to describe the nuclear A-body problem. We illustrate both the case for vibrational
excitations (left-hand part) and for collective rotational motion (right-hand part). Here, the 3-axis
is the nuclear symmetry axis and R denotes the collective rotational angular momentum (j denotes
any intrinsic angular momentum)

a major task to try to bridge the space between pure shell-model methods and
macroscopic methods of nuclear collective motion. Over the last years, making
use of symmetry aspects of the nuclear many-body problem and its microscopic
foundations, the interacting boson model has helped to carry out this bridging
programme in an important way although many open problems remain to be
solved.
In this book, we shall mainly concentrate on the nuclear shell model proper
with, at the end, some more advanced topics. The main aim, however, is to
bring in the necessary elements of technique in order to understand and also
handle the nuclear shell model with some success. One consequence is that we

2
have to leave out collective nuclear models, but many techniques can easily be
adopted in that field, too, once the nuclear shell model has been worked through
in detail. In order to not interrupt the shell model discussion by technical ac-
counts of angular momentum, tensor operators, matrix elements and the like, we
start with introductory chapters on angular momentum in quantum mechanics
and on rotation in quantum mechanics. These chapters are self-contained but not
exhaustive, (proofs are usually left out) since they serve only as an avenue to
the shell model. In a long Chap. 3, we discuss the one-particle average field,
two-particle identical nucleon systems and their properties (wave function con-
struction, residual interactions, configuration mixing, ...), three-particle identical
nucleon systems, non-identical proton-neutron systems and isospin. Chapter 4
concentrates on electromagnetic properties in the nuclear shell model with major
emphasis on one- and two-particle moments (f,l, Q) and transition rates (E2 and
Ml transitions). In Chap. 5, we discuss second quantization methods to be able to
reformulate the nuclear shell model with its study of elementary modes of exci-
tation in a Hartree-Fock framework: particle-hole excitations near doubly-closed
shell nuclei (Chap. 6) and particle-particle excitations in open shells (pairing cor-
relations in Chap. 7). In this chapter, we briefly discuss some recent applications
of pairing aspects of the nuclear many-body system in a broken-pair and inter-
acting boson model description of low-lying collective excitations. In Chap. 8,
advanced topics on state-of-the-art shell model calculations, using a single inter-
action type in order to fix both the average field and the residual interactions,
are discussed. This chapter brings the reader into contact with present day shell
model methods. Problem sets as well as appendices that treat a number of more
technical problems are included at the end. Also, a set of elementary FORTRAN
programs are added that allow the calculation of the most-used angular momen-
tum coefficients (3j, 6j); calculation of radial integrals for harmonic oscillator
wave functions and a diagonalization program using the Jacobi method for di-
agonalizing small matrices. A code that calculates the two-body a-interaction
matrix elements, including spin exchange, is also given.
Note that in these chapters the subject is intentionally restricted to the nuclear
shell model to allow an in-depth, technical but also broad treatment. We hope to
thus be able to prepare students for their own research in this field.

3
1. Angular Momentum in Quantum Mechanics

Before starting a detailed discussion on the underlying mechanisms that establish

the nuclear shell model, not only for the single-particle degrees of freedom and
excitations but also in order to study nuclei where a number of valence nucleons
(protons and/or neutrons) are present outside closed shells, the quantum mechan-
ical methods are discussed in some detail. We discuss both angular momentum in
the framework of quantum mechanics and the aspect of rotations in quantum me-
chanics with some side-steps to elements of groups of transformations. Although
in these Chaps. 1 and 2, not everything is proved, it should supply all necessary
tools to tackle the nuclear shell model with success and with enough background
to feel at ease when manipulating the necessary "Racah"-algebra (Racah 1942a,
1942b, 1943, 1949, 1951) needed to gain better insight in how the nuclear shell
model actually works. These two chapters are relatively self-contained so that
one can work through them without constant referral to the extensive literature
on an~lar momentum algebra. More detailed discussions are in the appendices.
We also include a short summary of often used expressions for the later chapters.

1.1 Central Force Problem and Orbital Angular Momentum

A classical particle, moving in a central one-particle field U(r) can be described

by the single-particle Hamiltonian (Brussaard, Glaudemans 1977)
p2
H= 2m +U(r). (1.1)

In quantum mechanics, since the linear momentum p has to be replaced by the

operator -i Ii V, this Hamiltonian becomes

li2
H=-2m Ll + U(r) , (1.2)

where Ll is the Laplacian operator. The orbital angular momentum itself is defined
as

l=rxp, or (1.3)

1=-ilirxV, (1.4)

4
as the corresponding quantum mechanical angular momentum operator. The com-
ponents can be easily obtained in an explicit way by using the determinant no-
tation for I, i.e.,
i J k
x y z
-iii (1.5)
0 0 0
ox oy oz
where i,j, k denote unit vectors in the x, y and z direction, respectively.
The commutation rules between the different components of the angular mo-
mentum operator can be easily calculated using the relations

[x, Px] = XPx - pxx = iii , (1.6)

which leads to the results

(1.7)

with cyclic permutations.

We can furthermore define the operator that expresses the total length of the
angular momentum as
12 = [2x + [2y + [2z' (1.8)

which has the following commutation relations with the separate components

(i == x, y, z) . (1.9)

If we now try to determine the one-particle SchrOdinger equation that corre-

sponds to the central force problem of (1.1), we can use a shorthand method
for evaluating the operator 12. Starting from the commutation relations (1.6), one
can show that

12 = (r x p) . (r X p) = r2p2 - r(r . p) . p + 2i lir . p , (1.10)

and using

r 'p= -1'Ii r 0
- (1.11)
Or '
one obtains

12 = r2p2 + 1i2~ (r2~) . (1.12)

Or Or
The kinetic energy operator of (1.1) then becomes

p2 12 1i2
T= 2m = 2mr2 - 2mr2 or
0( 0) r2 or . (1.13)

5
We shall now briefly recapitulate the solutions to the central one-body SchrOdinger
equation, solutions that form a basis of eigenfunctions of the operators H, 12 and
lz simultaneously. So, we can write still in a rather general way that

H<p(r) =E<p(r) , (1.14)

(1.15)

The SchrOdinger equation (1.14) now becomes [using (1.15)]

(1.16)

Using a separable solution of the type

u(r)
<p(r) == R(r)Y(B, <p) = - . Y(B, <p) , (1.17)
r

the radial equation becomes

2m
2
- 1i- - - + [>..
Jlu(r)
dr2
2
-1i- + U(r) u(r) = Eu(r) ,
2mr2
1 (1.18)

and its solution, in particular, depends on the choice of the form of the central
potential U(r). This particular problem will be discussed in Chap. 3.
The eigenfunctions for the angular part of (1.16) can be obtained most easily
by rewriting the angular momentum [2 operator explicitly in a basis of spherical
coordinates (Fig. 1.1). One works as follows:

""
"" ,
" P(x.y.z)

I y
IP', I Fig. 1.1. The cartesian and spherical coordinates for the point
"
_ _ _ _ _ _ _ ::::.1
I P(r) (z,y,z) and (r,8,cp) for which the orbital angular
momentum is analyzed
x

6
 i) rewrite the cartesian components lx, l y , 1% as a function of the spherical
coordinates (r, 9, r.p)

Ix =i Ii (sin r.p! +cot9cOSr.p~) ,

ly =iii ( - cos r.p ! + cot 9 sin r.p ~) , (1.19)

1% = -iii :r.p •

For more details, see (Sect. 1.3).

ii) rewrite the length of the angular momentum 12 as a function of the spherical
coordinates
12 = [2x + [2y + [2%

= _1i2{ (sin r.p ! + cot 9 cos r.p ~) (sin r.p ! + cot 9 cos r.p ~)
+ (-cosr.p! +cot9sinr.p~) (1.20)

x (-cos r.p :9 + cot 9 sin r.p :r.p ) + :;2 }

= -Ii &}
2{ -sin1-9 -a9a(sm. 9-a9a) + -sin1-9 -ar.p2
-
. 2

We now determine the angular momentum eigenfunctions starting from

(1.15,20). Using a separable form
Y(9,r.p) = ~(r.p). 8(9) , (1.21)
one gets the two differential equations
cP~ 2
dr.p2 +m ~=O, (1.22)

(1.23)

In order to obtain these two equations, one uses a separation method for the
variables 9 and r.p as is discussed in introductory courses of quantum mechanics
(Fliigge 1974). The solutions to (1.22), using the condition of uniqueness of the
solutions, become

m=O, ±1, ±2, .... (1.24)

7
 e
Putting now ,.\ = 1(1 + 1) and = cos fJ, one recognizes in (1.23) the differential
equation for the associated Legendre polynomials p,m (Edmonds 1957), i.e., one
has (0 ~ Iml ~ 1)

p,m(e) = (1 - et /2 ~: p,(e) , (1.25)

with
1 d' I
p,(e) = 2'1! de' (e - 1) . (1.26)

(P, are the Legendre polynomials).

Finally, the solution to (1.15) becomes

with correct angular normalization, and using m ~ O. For negative values of m,

one has the relation

y,-m(fJ,<p) =(-l) mYr(fJ,<p)* .

These functions are well known as the spherical harmonics. Using the above
solutions, the angular momentum eigenvalue equations can be written
,2y,m(fJ, <p) = h21(l + l)y,m(fJ, <p) , (1.28)

lzy,m(fJ, <p) = hmy,m(fJ, <p) . (1.29)

We give here some of the most used spherical harmonics.

10°= y'4;
1 , (1.30)

[3 cos fJ ,
Yio = Vi;
(1.31)
y,±1
1
=..,.. [3e±i'P sin fJ ,
,vg;1

12° = J1~7r (3cos 2 fJ -1) ,

12±1 = =F/ffe±i'P cos fJ sin fJ , (1.32)

12±2 = J3~7r e±2i 'P sin2 fJ .

In Fig. 1.2, we illustrate some typical linear combinations, which are called the
s, p and d functions (Weissbluth 1974). These functions playa major role when
describing electronic bonds in molecule formation. These are the combinations

8
 z

x
S

z z

,.•.., , /
·4->,,/ ;
P y

/
r".J'y
X X

Px Py Pz
z z y

----':i:::----x ----.;:,.j~--y ---±"'---- x

Fig. 1.2. Polar diagrams for s (1 = 0), p (1 = 1) and d (1 = 2) angular wave functions. These
represent real combinations of the Yo0 , and Yt' Yi
spherical hannonics. The figure is taken from
CJ. Ballhausen and H.B. Gray "Molecular Orbital Theory", W.A. Benjamin, Inc.)

9
 Px TVfl(
= V(4; 2 - Yi1 + Yi-I) r = x ,

py TVfl2' Yi1+ Yi-I) r = y,

= V(4;
o
(

(1.33)

d zx = fJ.~ (-Yl + 12-1) 0 r2 = v3zx 0

The angular momentum eigenfunctions are a set of orthogonal functions on the

unit sphere expressed by

12'1f 1'1f Y,m * «() , cp)Y,r,n' «(), cp) sin () d() dcp =81l'8mm , 0 (1.34)

Also, an addition theorem exists

2:: 1L
+/
p/(COS()12) = y,m*(ill)y,m(il2) , (1.35)
m=-/

where ill. il2 are the angles «()I,CPI), «()2,CP2) defining the two directions and ()12
is the angle between the two direction vectors (llt(h
We now introduce angular momentum ladder operators l±, operators that are
linear combinations of the operators Ix, Iy but are very useful in setting up the
angular momentum algebra relations. Defining

(1.36)

we shall determine the action of these ladder operators acting on the spherical
harmonics. Therefore, we just need to evaluate the commutation relations of the
ladder operators among themselves and with lz. One can easily evaluate that

(1.37)

10
Using the spherical coordinates and the explicit forms of lx, ly, l z (1.17-19),
one can rewrite 1+, L and 1z as

1+ . (8 +
= h elIP 8B 8)
i cot B8<p

1
-
. (8
= -he-lIP - - icotB-
8B 8<p
8) (1.38)

lz = -ih :<p .

Knowing the explicit form of the spherical harmonics Yjm (B, <p), the action of the
operators l±, lz on these functions can, in principle, be calculated in a straightfor-
ward but tedious way. We shall discuss a more elegant method in evaluating the
action of l± on the spherical harmonics. We start from the eigenvalue equation
(1.29) on which we act with the operator 1+ giving
(1.39)
Now, using the commutation relations (1.37), this relation can be rewritten as
1+lzYr(B, <p) = (lz - h)/+Yjm(B, <p) = m hl+Yr(B, <p) , (1.40)
or,
(1.41)

This indicates that 1+ Yjm (B, <p) is an eigenfunction of the operator I z with eigen-
value (m + 1) h and thus the ladder operator 1+ effectively adds one unit h to
the original m-projection. Likewise, L substracts one unit h from the original
m-projection and gives

Iz(LYjm(B,<p») =(m-l)h(L1T(B,<p»). (1.42)

This allows for the relations
l+Yjm(B, <p) = o:(l, m)Yjm+l(B, <p) ,
(1.43)
LYjm(B, <p) = (3(l, m)Yjm-l(B, <p) .
The factors 0:(/, m), (3(l, m) can be determined by calculating the norm of ex-
pressions (1.43). Thus,

J yr*(B, <p)Ll+Yjm(B, <p)dD = 10:(/, m)1 2 , (1.44)

since the spherical harmonics form an orthonormal set of eigenfunctions. We can

now evaluate the operator expression Ll+ explicitly as follows. We start from
12 = I; + I; + 1; = ~ (/+L + L q + 1; , (1.45)
and using the commutation relations (1.37), this simplifies into
12 =Ll+ + lz(/z + h) , (1.46)
giving rise to the equality

11
 Ll+=12 -1 z (/z+ Ii.). (1.47)
Here, one also needs to impose the conditions
I+Y,' = L1';-' =0 .
This relation (1.47) used in (1.44) gives the result
1i.2 [1(1 + 1) - m(m + 1)] = la(/, m)1 2 , (1.48)
or
1+y,m(B, t,p) = Ii. { 1(1 + 1) - m(m + 1) }1/2y,m+l (B, t,p) . (1.49)
Similarly, for the other ladder operator L one has
Ly,m(B, t,p) = Ii. { 1(1 + 1) - m(m - 1) }1/2v,m-l(B, t,p) • (1.50)

1.2 General Definitions of Angular Momentum

In Sect. 1.1, we have derived the angular momentum operator l(lx, ly, lz) explic-
itly, starting from the one-body central force problem. This method only allows
for entire values of the angular momentum eigenvalue 1and m. It is now possible
to define angular momentum in a more general but abstract way starting from the
commutation rules (1.7,9). If we construct general operators J2, Ji(i =: x, y, z)
which fulfill the relations

(i=:x,y,z),
(1.51)

and cyclic permutations, the operator J2 defines a general angular momentum

operator. The eigenvectors are now defined as the abstract vectors in a Hilbert
space carrying two quantum numbers, i.e., the quantum number defining the
length j and the quantum number defining the projection of j(m), since the
quantum numbers corresponding to the full set of commuting operators define
the state vector uniquely. Thus, we have the eigenvector relations

J21j, m) = 1i.2 j(j + l)lj, m) , (1.52)

Jzlj,m) = li.mlj,m) . (1.53)

Using the ladder operators, we can also write

J±lj, m) = Ii.{j(j + 1) - m(m ± 1)}1/2Ij, m ± 1) , (1.54)

Jzlj, m) = li.mlj, m) , (1.55)

as the defining expressions for a general angular momentum operator.

12
1.2.1 Matrix Representations
In the discussion of Sect. 1.1, the angular momentum operators had an explicit
expression in terms of the coordinates and derivatives to these coordinates (dif-
ferential form). In the more general case, as discussed above, we can derive a
matrix representation of the operators J x , Jy, Jz or J+, J_, Jz and J2 within
the space spanned by the state vectors Ii, m). As an example, we use the five
state vectors Ii, m) for the case of i = 2(-2 :5 m :5 +2). We denote the state
vectors in column vector form as
1 0 0
0 1 0
0 0 , ... , 0 (1.56)
0 0 0
0 0 1

li,2) , Ii, 1) , ... , Ii, -2) .

The action of the ladder operators now leads to

J+li,m) =am,m+tli,m+ I} , (1.57)

where am,m+t defines the following matrix representation of J+, i.e.,
0 at,2 0 0 0
0 0 a2,3 0 0
J+~ 0 0 0 a3,4 0 (1.58)
0 0 0 0 a4,s
0 0 0 0 0
Similarly, one gets a matrix representation for J_ following from

Lli,m} = bm,m-tli,m -I}, (1.59)

and for J z since
(1.60)

1.2.2 Example for Spin ! Particles

The angular momentum representation for spin ! particles (electron, proton,
...), using the general method as outlined in Sect. 1.2, now gives in a simple
application the construction of the 2 x 2 spin matrices. We shortly recapitulate
!
the spin angular momentum commutation relations.

(1.61)
Defining s = 11,/20', these commutation relations become
(1.62)

13
The matrix representations are spanned in the two-dimensional space defined by
the state vectors

[~], [~], (1.63)

which correspond to the states Ii, +i} and Ii, -i}, respectively. One often
denotes the former by X~{~2 or a(s) and the latter by XI};/2 or (3(s) in litera-
ture on angular momentum (Edmonds 1957, de-Shalit, Talmi 1963, Rose, Brink
1967, Brossaard, Glaudemans 1977). The ladder operator relations (1.54) for the
specific case of spin 1/2 particles become

s+li, -i} = lili, +i} , (1.64)

s-li, +i} = lili, -i} ,
or
O"+I!, -!} = 21!, +!} ,
(1.65)
O"-li, +!} =21i, -i} .
One immediately gets the o"z, O"y and o"z "operators" as

o"z == ( 01)
1 0 ' O"y == (0i -i)
0 ' O"z == (10 -1
0) (1.66)

Finally, we give a number of interesting properties for the Pauli spin! matrices
without proof:
.)
1 O"z2 = O"y2 = O"z2 = n , (1.67)
where n denotes the 2 x 2 unit matrix.
ii) {o"z, O"y} = 0, (1.68)
and cyclic where {A, B} is the anti-commutator defined as AB + BA.
iii) (0- . A)(o- . B) =A . B + io- . (A x B) , (1.69)
if both A and B commute with 0-.

1.3 Total Angular Momentum for a Spin ~ Particle

The total wave function characterizing a particle with intrinsic spin! (electron,
proton, neutron, ... ) which, at the same time, carries orbital angular momentum
can be written as the product wave function

cp(r,o-) =.,p (Iml' !ms)

0.70)
=Rnl(r)Yiml(e,cp)X!,{~(o-) ,
14
where Rn,/(r) describes the solution of the radial SchrOdinger equation (1.18)
using a given potential U(r), n describes a radial quantum number counting the
number of nodes, and I the orbital angular momentum eigenValue. Furthermore,
~ml«(}, cp) (with -I ::; m/ ::; +l) describes the angular part of the wave function
and x!,{~(u) (m s = ±!) the intrinsic spin wave function. Since X!,{~ can be
written as a state vector in a two-dimensional space, it is more correct to speak
of (1.70) as a state vector than as a wave function.
The following eigenvalue equations are fulfilled for the state vectors (1.70):

12'I/J(lm/, !ms) = 1i21(l+ 1)'I/J(lm/, !ms) ,

lz'I/J(Zm/, !ms) = lim/'I/J(lm/, !ms) ,

(1.71)
82'I/J(lm/,!ms) = 1i2i'I/J(lm/, !ms) ,

sz'I/J(lm/, !ms) = Iims'I/J(lm/, !ms) .

We now define the operator

J = , + 8 = I + 1i/2u . (1.72)

Using the definitions for a general angular momentum operator (1.51), one can
show that the operator J(J 2, J x, J y , Jz) is indeed an angular momentum oper-
ator since the commutation relations

[lx, J y ] =iliJz , ... ,

(1.73)
[J 2 , Ji] =0,
hold (verify this explicitly).
By construction ,2, lz, 8 2 and Sz form a set of commuting operators (the
orbital angular momentum operators and the intrinsic angular momentum "spin"
operators act in totally different spaces, the former in the space of coordinates
(x, y, z), the latter in an abstract space, spanned by the unit vectors x!,{;).
One can now show that the operators J2, J z, ,2 and 8 2 also form a set of
commuting operators indicating that it is also possible to describe the full state of
the particle with orbital and intrinsic spin in a basis characterized with quantum
numbers relating to J2, J z , 12 and 8 2 , respectively. Since Jz = lz + sz, in the
above case, a fixed m-eigenvalue will occur but not necessarily a fixed m/, ms
value. This follows from the fact that J2 does not commute with lz or with Sz
but only with the sum lz + Sz. (The proof of this is left as an exercise.)
We now study the effect of acting with J2 on the state vectors that are
eigenvectors of the "uncoupled" (12, l z, 8 2 , sz) basis. Since one can write J2 as

(1.74)

acting on the vectors 'I/J(lm/, !m s) one obtains

15
 J 21/J(lm" !mS) =1i2{ 1(1 + 1)+ l +2m,ms}1/J(lm" !ms)
+a1/J (1 m, + 1, ! ms - 1) + IN (1 m, - 1, ! ms + 1) .
(1.75)
From (1.75) it becomes clear that the eigenvectors 1/J(lm" !m s) are not in general
eigenvectors of J2, although they are eigenvectors of ,2, 8 2 and Jz. The right
hand side of (1.75) represents a 2 x 2 matrix spanned by the configurations
1/J(lm" !m s) with a fixed value of total magnetic quantum number m (= m,+m s)'
By diagonalizing this matrix one obtains two eigenvalues of j, i.e., j = 1 and +!
j = 1- !.
Of course, in the two extreme cases (see the problem set) j = 1+ !,
m =1 +4
and j = 1 +!,
m = -1- 4,
only one component, 1/J(I,m, = I, !,
+!) m,
ms = and 1/J(I, = -1,4, ms = -!), respectively, results.
As an example, we take the case of 1 = 4, s = that can be combined to 4
form both the j = ~ and j = ~ total angular momenta. The states obtained are

0"(1 =4
0/ , s = 12' J' = 22' m =+2)
2

=1/J (I =4, m, =4, s = 4, ms =+4) ,

1/J (I =4, s = 4, j = ~, m = +~)
= a1/J (1 = 4, m, =4, s = 4, ms = -l)
+/J1/J(1=4,m,=3,s=4,ms =+l) ,
1/J(1=4, s=4,j=~, m=+~)
= IN (I =4, m, = 4, s = !, ma = -4)
-a1/J(Z=4,m,=+3,s=4,m a =+l) ,
(1.76)

1/J(1 = 4, s = 4, j = ~, m = -~)
=1/J(1=4, m, = -4, s = 4, ms = -4) .
In general, the eigenvectors of J2, J z, 12, 8 2, denoted by 1/J(ls = 4,
jm) can be
expanded in the eigenvectors of ,2, 1%. 8 2 , Sz 4
that are given by 1/J(lm" ms) as
follows:

1/J(ls = 4, jm) = L (1m" 4msl Zs = !,jm}1/J(lm" 4mB) . (1.77)

m"m.

The coefficients that establish the transformation of one complete basis to the
!,
other complete basis (1m" !mslls = jm) are denoted as Clebsch-Gordan

16
Tablet.t. Analytic expressions for the C1ebsch-Gordan coefficients appearing in (1.79) for coupling
the orbital angular momentum 1 with the intrinsic spin 8 = 1/2 to a total angular momentum
j=I±I/2

j m. =+4 m.= -4

1+ t C+I/2+m y/2 ( my/2

1+1/2 -
2/+ 1 + 21 1

1- 4 C+I/2- my/2 -C+I/2+m y/2

2/+ 1 2/+ 1

(Clebsch 1872, Gordan 1875) or vector-coupling coefficients. In the new, "cou-

pled" basis of (1.77) one has the following eigenvalue relations

12 tP (ls=!,jm) = 1i21(1+1)tP(ls=!,jm),
8 2tP(ls =!, jm)
= 1i2~tP(ls =!, jm) ,
(1.78)
J2tP(ls=!,jm) = 1i2j(j+ l)tP(ls = !, jm) ,

JztP(ls =!, jm) = IimtP (ls =!, jm) .

Written explicitly, the total state vector for a spin s =! fennion particle becomes
tP(nls=!,jm) =Rn1(r){(lm -!,! +!ll!,jm}Yim-1/2«(),cp)x:{~2(u)
1 . ) m+1/2 () 1/2 (u),
}
+ (1 m +'2'
1 1
'2 -'211 1'2' Jm Yi (,CP)x_1/ 2
(1.79)
with the (... I... ) Clebsch-Gordan coefficients given in Table 1.1.

1.4 Coupling of Two Angular Momenta:

Clebsch-Gordan Coefficients

In this section, we concentrate on the coupling of two distinct angular momenta,

but now for the more general case of angular momentum operators J 1, J 2 that
may have an m-projection Jtz, J2z representing both integer or half-integer
values. In that case, the total angular momentum operator is expressed by the
sum J = J 1 + J 2. The set of four commuting operators

(1.80)

are characterized by the common eigenvectors, expressed by the product state

vectors

17
 (1.81)

The other set of commuting operators

{J2, J", Jr, J~} , (1.82)

can also be characterized by a set of common eigenvectors, being linear combi-

nations of the eigenvectors (1.81). such that they form eigenvectors of J 2 • One
also denotes them as the "coupled" and "uncoupled" eigenvectors that are related
via the expression

Ijlh;jm} = (l.83)

The overlap coefficients (... I... ), going from one basis to the other are called
the Clebsch-Gordan coefficients. To simplify the notation one frequently uses an
abbreviation in the ket side of the bracket. i.e.,

(1.84)

In order to determine the relative phases for the Clebsch-Gordan coefficients,

we shall use the Condon-Shortley phase convention (Condon, Shortley 1935,
Edmonds 1957). which is basically defined as follows.
i) When acting with the ladder operator J± on the eigenvectors lilh;jm},
we define the phase as

J±lilh;jm} = eic5 n{v 1= m)v ± m + l)i/2 Ijlh;jm ± I} , (1.85)

with eic5 = +1.

ii) By acting with the operator J2 (J = Jl + J2) on the states with the
extreme projection quantum number M = jl + h or M = -jl - h, one gets the
result that
J 2 ljlh;jm = ±(jl + h)} = (jl + h)(jl + h + 1) n2 ljlh, jm = ±(jl + h)) ,
(l.86)
indicating that the state in (1.86) is an eigenstate of J2 and J" with eigenValues
VI + h)Vl + h + 1) n2 and M = ± nVt + h), respectively. Thus we get, by
applying (1.83) for this particular case
lith; j = jt + h, m = ±"(it + h)}
=eiDt lit ml = ±jl} Ih, m2 =±h} , (1.87)

which, with the choice eiDt = +1, gives the aligned Clebsch-Gordan coefficients
(jl ml = ±jl,hm2 = ±j2lj =it + h, m = ±(jt + h)) = +1 . (l.88)

18
 iii) By acting now with the ladder (lowering) operator J_ = Jt- + h-, on
(1.87), and relating that result to the explicit fonn of the state vector (1.83) with
j =it + h, m = jt + h - 1; we get
1-ljth; j =jt + h, m =it + h) = (Jt- + J2-)ljt,mt =jt)lh,m2 =h) ,
(1.89)
and
Ijth;j =jt + h, m =jt + h - 1)
= (itjt - l,hhli =jt + h, m =jt + h - 1)ljt,jt - 1)lh,h)
+ Utit,hh - Ilj = jt + h, m = jt + h - l)1jt,jt)lh,h - 1) .
(1.90)
This leads to the identification
Utit -l,hhli =jt + h, m = jt + j2 - 1) = (it/(jt + h)i/2
Utit,hh - Ilj = jt + h, m =jt + h - 1) = (h/(jt + h)i/2 • (1.91)

With the value of m = jt + h - 1, another state can be constructed, i.e.,

Ij = jt + h - 1, m = jt + h - 1) which should be constructed from the same
uncoupled states that appear in (1.90). By imposing the condition of orthog-
onality, one can deduce both the absolute value and the relative phase of the
Clebsch-Gordan coefficients

Utit - I, hhlj =it + h - 1, m =jt + h - 1), and

(1.92)
Utit, hh - Ilj = jt + h - 1, m = jt + h - 1) .
The absolute phases are now defined by the condition that for any given total J
and projection M one has

(JMIJIzIJ - 1, M) ~ O. (1.93)
This condition, written out for the Clebsch-Gordan coefficients making up the
states IJ M) and IJ - 1, M) becomes

(1.94)

The above condition (1.94) can be shown to be equivalent to the condition

(Brussaard 1967)
for each J. (1.95)
These are now the phase conditions of (i), (ii) and (iii) that uniquely define the
Clebsch-Gordan coefficients and thus also the coupled state vectors of (1.83).

19
1.5 Properties of Clebsch-Gordan Coefficients

Since the Clebsch-Gordan coefficients serve as expansion coefficients for a given

eigenvector in a specified ortho-normal basis, there exist orthogonality relations
that are given by

and
(1.96)
2::Ul m l ,j2m2Ijm)Ulm~ ,hm~\jm) = 8m1m~ 8m2m~ .
j,m

Interesting symmetry relations exist when interchanging the two angular momenta
that become coupled, e.g., (de-Shalit, Talmi 1963)
(1.97)
In interchanging either jl and j or j2 and j, more complex relations result. Mak-
ing use of the much more symmetric way of expressing the angular momentum
coupling coefficients, the Wigner 3j-symbol notation defined as (Wigner 1959,
de-ShaIit, Talmi 1963, Brussaard 1967)

( jt 0.98)
ml
symmetry properties under the interchange of any two angular momenta of the
set (jl, jz, h) become very simple:

( jl
ml ~3 ) = (:2 ~3 ~1) = (~3 ; :1
= (_I)i1 +i2+i3 (
mlit m3
J3
~22)
= (_I)it+i2+i3 (j3m3 m2
J2
~1) = ... , (1.99)

or, a phase factor +1 for an even permutation and a phase factor (_1)i1 +i2+i3 for
an odd permutation. Moreover, one gets the relation

Jl
( (1.100)
-ml
The former orthogonality relations (1.96) for the Clebsch-Gordan coefficients
now rewritten in terms of the Wigner 3j-symbols become
., ) 1
m313, = -2'
J3 +
18j 3 j'3 8m 3 m'3 , (1.101)

(1.102)

20
 .L (2j3 + 1) (mjll
}a,ma
J2
m2
j3) (jl
m3 mt
(1.103)
Extensive sets of tables of Wigner 3j-symbols exist (e.g. Rotenberg et al. 1959).
Explicit calculations of the Wigner 3j-symbol are easily performed using the
expression (de-ShaHt, Talmi 1963).

jl
( ml jz j3) = fJ (m 1 + m2, m3) ( ((jl + jz - j3)! (j2 + h - jl)!
m2 m3
X (j3 + jl - jz)!)/(it + jz + h + 1)!f2
X ((jl + ml)! (jl - ml)! (jz + m2)!(jz - m2)! (h + m3)! (h - m3)!i/2
X L(_I)il-h- m a+ t (t! (jl + jz - j3 - t)! (j3 - jz + ml + t)!
t

x (j3 - it - m2 +t)!(jl - ml - t)!(j2 +m2 - t)!fl . (1.104)

with (-m)! = 00 if m is positive, t is entire and O! = 1, so the following
conditions hold
t~O
jl + jz - h ~ t
- h + jz - ml ::; t
(1.105)
- h + jl + m2 ::; t
jl - ml ~ t
jz + m2 ~ t.
In Chap.9, a FORTRAN program is given that evaluates (1.104) numerically.
As an example, we evaluate the 3j -symbol

(j -m 0
1
m
j) .

The conditions (1.105) give the restrictions on t

t 0; 1 ~ t; - j + 1 + m ::; t; 0 ::; t;
~

j + m ~ t; 1 ~ t or t = 0, 1 .
Calculating in detail, one gets

j
( -m 1 j)=(2j-l)!/(2j+2)!i/2
0 m
X (0 - m)!(j + m)!(j + m)!(j - m)!i/2
x (_I)i- l - m [(0 - 1 - m)!(j + m)!fl - (0 - m)!(j + m - 1)!fl]
= (_I)i- m m /(j(j + 1)(2j + 1))1/2 .

21
1.6 Racah Recoupling Coefficients:
Coupling of Three Angular Momenta

In the case of a system described by three independent angular momentum op-

erators J .. J2, J3; one can again form the total angular momentum operator J
defined as
(1.106)

The six commuting operators

{Jf' J Iz , J~, hz, J~, hz} , (1.107)

have a set of common eigenvectors, the product vectors

(1.108)

For the three angular momentum operators, it is now possible to form three sets
of commuting operators:

{J 2,Jz, Jf, J~, J~, Jf2} , (1.109)

{J 2,Jz, Jf, J~, J~, J~} , (1.110)

{J 2, J z , Jf, J~, J~, Jf3} , (1.111)

with the eigenvectors

I(ith)J12h; JM} , (1.112)

lit (j2i3) 123; J M} , (1.113)

I(iti3) J13h; J M} , (1.114)

respectively.
It is possible to use a diagrammatic way of expressing the vector coupled
eigenstates (Brussaard, Glaudemans 1977), by using lines and arrows for a given
angular momentum and the order in which they are coupled. The intermediate
angular momentum is shown by the dashed line (Fig. 1.3).
Between the three equivalent sets of eigenvectors of (1.112-114), transforma-
tions that change from one basis to another can be constructed. We can formally
write for such a transformation (de-Shalit, Talmi 1963)

lit (j2i3) 123; JM) = 2::( (jth) Jt2h; Jlit (j2h)J23; J)

(1.115)
It can easily be shown that the transformation coefficients in (1.115) and in similar
relations do not depend on the projection quantum number M. Now by explicitly

22
 J J
Fig. 1.3. Graphical illustration of two possible ways to construct the angular momentum wave func-
tions for a system where three angular momenta are used, according to (1.112) and (1.1l3). The
angular momenta are represented by vectors, the intermediate momenta by dashed-line vectors

carrying out the recoupling from the states Ijl v2h)J23; J M} to the coupling
scheme IVlh)1t2h; J M} (Appendix B), one obtains the detailed fonn of the
recoupling coefficient of (1.115). In this particular situation, a full sum over all
magnetic quantum numbers of products of four Wigner 3j-symbols results. The
latter, defined as an angular momentum invariant quantity (no longer dependent
on the specific orientation of a quantization axis), the Wigner 6j-syrnbol, leads
to the following result (Wigner 1959, Brussaard 1967)
(it (hh)J23; JI(jlh)JI2h; J)
=(_I)it+h+ia+ J i 12 i 23 {;: ~ ~:}, (1.116)

(using the notation i == (2J + 1)1/2).

The precise definition of the 6j-symbol in tenns of the 3j-symbols reads
(de-Shalit, Talmi 1963)

{ jl
II
J2
h
h} =
&
L
all mi,m~
(_l)L'i;+El;+Em;+L'm: ( jl
~
J2
m2
h)
m3

X ( it 12 13 ) ( II J2 13 ) ( II h h )
-ml ~ -m~ -m~ -m2 m'3 m'I -m~ -m3
(1.117)

and very much resembles a "contraction of tensors" (one sums over projection
quantum numbers ml, m2, ... , m~, both of which always show up in different
3j-syrnbols with opposite sign). We show in Chap. 2 that, indeed, the 6j-syrnbol
is a full contraction not on cartesian but on spherical tensors (Wigner 1959).

1.7 Symmetry Properties of 6j-Symbols

Because of the very structure of the definition in (1.117), in each 6j-symbol four
angular momentum couplings have to be satisfied in order to be non-vanishing.

23
In shorthand notation, replacing the angular momenta with dots, one has the
couplings

c.-:-.·} {:_'./:}
{:~:-:} {:/:~:}. (1.118)

Here we quote some often used symmetry properties. A more detailed account
can be found in various texts (de-Shalit, Talmi 1963, Edmonds 1957, Rose, Brink
1967, Brussaard 1967, Brink, Satchler 1962)

i) { jt h j3} = 0 (1.119)
11 lz 13 '

unless the triangular (coupling) conditions vli2h), VI12h), (lllzj3), (l1i2h) are
fulfilled

m {~ ~ ~}={~ ~ ~}={~
)3
13
h}
lz

= {~: ~~ {:} = { {:
12
)2
~:} = ... , (1.120)

iii) orthogonality relation

2)2j + 1) {~1
" )3
~2
)4)
?,} {~1 )3)4
)2 )!,,} = bj' jl/(2j' + 1)-1 , (1.121)
J

iv) special case

{ ~1 )1
)2
~2 j3}
0
= (_I)i1 +i2+ia (31)~2)-1 b" "/ b" "/
J111 J2J2 '
(1.122)

v) Explicit form: Racah formula (de-Shalit, Talmi 1963)

{~:{~{: } = LlUti2j3)LlUI12 I3)Ll(Iti2 I3)Ll(III2h)

x L(-I)t(t+l)![(t-jl-h-h)!(t-jl-lz-h)!
t
X (t - 11 - j2 - h)! (t - It - lz - j3)! (jl + h + 11 + lz - t)!
X (j2 + j3 + lz + h - t)!(j3 + jl + h + 11 - t)!rl , (1.123)
with

Ll(abc) = [(a + b - c)!(b + c - a)!(c + a - b)! I(a + b + c + 1)!]1/2 , (1.124)

and the condition of having non-negative values of the integer in the factorial
expression in (1.123).

24
 1.8 Wigner 9j-Symhols: Coupling and Recoupling
of Four Angular Momenta

Similarly to the methods used in Sect. 1.6, we can construct the total angular
momentum operator corresponding to the sum of the four independent angular
momentum operators as

(1.125)

In constructing the total set of commuting operators one has in the uncoupled
representation,

{ J~, J}z, J~, hz, Ji, J3z, J~, J4z } , (1.126)

which have as eigenvectors the product vectors

(1.127)

In the coupled representation, one needs two intermediate angular momentum

operators for which a large choice exists. Coupling pairwise, one has three pos-
sibilities
J2, Jz, J~2' Ji4, J~, J~, Ji, J~ ,
J2, Jz, J~3' J~4' J~, J~, Ji, J~ , (1.128)
J2, Jz, J~4' Jt, J~, Ji, J~, J~ ,
(Fig. 1.4), with eigenvectors

I J,3
/
J

IJ,2
Fig. 1.4. Graphical illustration of possible ways to construct
.,/' the angular momentum wave functions for a system where
j, / /' j4
four angular momenta are used, according to (1.129). In the
lower part, we present the more general way of constructing
/.,/'/ J'23
the four-angular momentum system by specifying JI2, J123
as intermediate angular momenta, respectively
J

25
 I(jlh) J12 (j3j4) J34; JM) ,
I(jlh)J13(ilj4)h4; JM) , 0.129)
I(j!i4)Jt4(j2j3) J23; JM) ,
respectively.
There exist other possibilities, too, however, such as

J2,Jz, Jr2' Jr23' Jr, J~, J~, J~ , (1.130)

shown in Fig. 1.4. The latter method is probably the best adapted to extend
coupling to n angular momenta by successive coupling of an extra angular mo-
mentum to the former n - 1 system (Yutsis et al. 1962). Here, too, many possible
recoupling schemes and recoupling coefficients can be obtained (Edmonds 1957).
Here we only discuss recoupling between the states of (1.129) since they lead to
the Wigner 9j-symbol, e.g., (de-Shalit, Talmi 1963)

I(ith)J13(hj4)h4; JM) = L J13J24J12J34

x {;: ;: ~::} I(jlh) J12(hj4)J34; JM) . (1.131)

J13 124 J
Precise definitions of the Wigner 9j-symbol as a full contraction over products
of 6 3j-symbols can be found (Edmonds 1957, de-Shalit, Talmi 1963). In the
present context, where we shall concentrate on the nuclear shell model, we quote
a special case that often occurs:

{~1. ~2. Jk }= (_l)i2+ +ia+ Jk {hjl

J k
i2
j4 J}
J . (1.132)
~ n k k 0
Also, we point out that the general expression of (1.131) can be used when
recoupling from a (j j) coupling basis into an (LS) coupling basis if we consider
cases with two fermions. Thus one can relate the states 1(11 h)L(! !)S; J M) and
l(il !)jl (h !)i2; J M) by the transformation

1(llh)L G!) S; JM) = ~ LS]t32 {~

)1,)2 Jl
t ~}
J2 J
X I(It! )jl (h! )j2; J M) , (1.133)

a relation that gives the (jj) -+ (LS) basis transformation.

26
1.9 Classical Limit of Wigner 3j-Symbols

It is now possible to construct a classical (in the limit of large angular momenta)
model (Brussaard, Tolhoek 1957, Brussaard 1967) for angular momentum cou-
pling and thus also for the Wigner 3j- (and similarly for the 6j-, 9j-, 3nj-)
symbol. We make use of the fact that in quantum mechanics it is only possible
to specify both the length and the projection on a quantization axis of the angu-
lar momentum. Therefore, a precessing vector model results where for constant
precession velocity the azimuthal angle has a constant probability distribution.
Since the Oebsch-Gordan coefficients denote the expansion coefficients in an
orthonormal basis, the square can be interpreted as a probability. Thus, for the
uncoupled representation where Jr, JIz, Ji
and 12z are the commuting oper-
ators, the coefficients l(jlml,jzmzljm}IZ denote the probability that in a state
with fixed (jl ml) and (jzmz), a given value of (j, m) will result with j express-
ing the length of the angular momentum vector (correct only for large values
of j), Fig. 1.5. Similarly, IUlmtizmzljm}lz (Fig. 1.5) can be interpreted, for the
coupled basis where eigenstates of the operators J Z, J z , Jr, Ji
are considered,
as the probability that for given (j, m) the values ml and mz will result as pro-
jection quantum numbers relating to the angular momenta jl and jz, respectively.
One can even calculate this distribution in both cases from probability con-
siderations (Edmonds 1957 gives an explicit calculation). Extending the above
arguments, classical models can also be constructed for interpreting higher 3n - j
symbols (Brussaard 1967).

z-axis

z-axis

m1

Fig. 1.5. Graphical representation of two angular momenta i} and i2' shown as vectors that make
a precession around the z-axis with constant angular velocity (vector model). Using the addition to
a momentum i = i} + i2' the probability of obtaining a given value for the length j, given fixed
m} and m2 values, is given by the Clebsch-Gordan coefficient squared 1{i}m},j2m2Ijm > 12. If
the two vectors i} and i2 are coupled to form the total angular momentum i (which is a constant
of motion). the two vectors will make a precession around the direction of i. For fixed value of the
length of i and projection m, the projections m} and m2 can be obtained again as a probability
distribution given by the Clebsch-Gordan coefficient squared 1 < j}m},nm2Ijm > 12

27
Short Overview of Angular Momentum Coupling Formulas

One-particle central force motion-orbital angular momentum

lx, Iy, Iz: differential operators

=0
=i.~iik Itlk
= 2ltl z
= M+
=-ItL

l±/lm} = It(l(l + 1) - m(m ± 1))1/2/1, m ± I} .

General definition of angular momentum operator

via commutation relations

Differential operator representation Matrix representation

1=0,1,2, '"

Total angular momentum

j =1+8
J = JI + J2
{Jr, Jlz, J~,Jzz} and {J2, Jz, Ji, J~}
/jlj2;jm} = L (jlml,j2m2/jm}/ilml}/i2m2)

/jlmt)/j2m2 >= L(jlmt,jzm2/jm)/jtiz;jm) .

j,m

28
 Three angular momentum systems

J= JI +h+J3
{Jf, JIz, J~, J2%,Ji, J3%} ~ lit m I)lhm 2)lh m 3)
{J2, J%, Jf, J~, Ji, Jf2} ~ I(jIh)1t2h; JM)
{J 2, J>;, Jf, J~, Ji, Jf3} ~ I(jIi3)JI3h; JM)
{J 2, J%, Jf, J~, Ji, J~} ~ IjI (hh)J23; JM) °

Recoupling Wigner 6j-symbol I

{: :} =1)-1)- ( ") (

C 0) C 0)
ro-:-oo},{:~:_:},
{:_oo/:}, {:/:~:} °

Notation:

./

29
 Four angular momentum systems

J = JI + Jz +J3 +J4
{J~, JIz, Ji, J2z, Jt J3z, J~, J4%}

Basis states -+ litml}ljzm2}lhm3}lj4m4}

{ J 2, J z, J2I, J22, J23, two momenta } e.g. J212, J234

intennediate
angular

Recoupling-Wigner 9j-symbol

}E ( )( )( )
( )( )( ).
Notation:

30
2. Rotations in Quantum Mechanics

2.1 Rotation of a Scalar Field-Rotation Group 0(3)

In this section as well as in the rest of Chap. 2, we shall study the relation-
ships between the angular momentum operator and rotation of a physical system
described by a given wave function in more detail. We note that the angular mo-
mentum operator can act as a generator for rotations of a general scalar function.
We consider the change in a scalar function field f(r) because of a rotation
of this function field with respect to a fixed coordinate (x, y, z) system. This
is an active point of view characterizing the transformations. For the particular
case of a rotation about the z-axis through an angle orp (we denote this as a
vector, defined in the sense of the rotation, perpendicular to the plane in which
the rotation goes, Fig. 2.1).
We then know that a new function field F(r) is defined, but with the constraint
that

F(r + a) =f(r) , or (2.1)

F(r) =f(r - a)
=f(r) - a· Vf(r) + .. , . (2.2)

Fig. 2.1. Transformation of a point P(r) under a ro-

tation, in anti-clockwise direction, over an infinites-
imal angle ocp. The resulting rotation is character-
I y ized by a vector 011' (perpendicular on the rotation
plane in the sense of the rotation). The displace-
"I ment vector a then becomes (infinitesimal quanti-
'-J ties) a = 011' x r
x

31
The change of the function field in the same physical point r is expressed by

Sf(r) = F(r) - f(r)

= -Scpx r· Vf(r) + .. .
=-Scp . (r x V)f(r) + ... or

-i
Sf(r) = hScp . Lf(r) , (2.3)

up to lowest order in the infinitesimal angle IScpl. For a finite angle cp (which
can be divided into n infinitesimal rotations Scp = cpln) one gets

F(r) = lim (1 -
n-oo
~Ii cpn . L)n f(r)
= e-i/li.cp.L f(r) . (2.4)

We generally denote the operator that transforms the old scalar field f(r) into
the new function field F(r) in the same physical point by U R, with
(2.5)

[Note that although the derivation was given for a rotation Scp around the z-
axis, (2.4,5) hold generally for any rotation of an angle Icp I around a unit vector
In specifying the rotational plane (Edmonds 1957, Brink:, Satchler 1962, Rose,
Brink: 1967)].
The derivation, although slightly formal, can be illustrated in the example
of a temperature field T(x, y, z) (Fig. 2.2) where subsequent planes parallel to
the (z, y) plane are characterized by increasing temperature when approaching
the x = 0 point from the point at x = a (0° -+ 100°). If we rotate the physical
system, and thus the temperature field, a new field T'(x, y, z) at every point
is defined. We know, however, that for a given physical point P(r), the new
function field at the new coordinates has to be equal to the old function field at
the old coordinates or,

T'(x', y', z') = T(x, y, z) . (2.6)

As another example, we use the field

(2.7)

When rotating the function field over 45° anticlockwise one has
x' = x cos 45° - ysin45° ,
y' = x sin 45° + y cos 45° , (2.8)

and the new function F( r) becomes

F(r) =2axy. (2.9)

32
 T' (x.y.z)
z z
T (x.y.z)

y y

x =0

x
x
Fig.2.2. lllustration of the transfonnation properties of a scalar field [here we take a temperature
field, expressed by T(z, y, z)] under an active rotation of the system over an angle of 45° in anti-
clockwise direction around the z-axis. In the figure, we present the system as a "temperature" cube
with surfaces of given temperature: lOOo_50°--{)0. After rotation, a new temperature field T'(z, y, z)
results which describes the same physical system in the original system of axes (z, y, z)

Consider a point P(x = 3, y = 2) with a function value !(r) = 5a. The new
coordinates become x' = V2/2; y' = 5V2/2 and the new function value at
the new point becomes F(r') = 5a, too, illustrating the above discussion. It is
even possible to derive the new function form F(r) by using the more formal
definition of (2.5) applied to the present situation (see problem set).
The above discussion in deriving the rotation operator in an active image UR
has been using the fact that the rotation induced was related to orbital angular
momentum. A more general angular momentum operator J similarly induces a
rotation operator
UR=e -i/IiO/.·J , (2.10)
for a rotation about an angle lal around an axis defined by a unit vector In
(Fliigge 1974).
There thus exists a close relationship between the angular momentum opera-
tors (generators) and the rotations of a function field. Using some simple group
theoretical elements this can even be more transparent (Wigner 1959, Hamermesh
1962, Goldstein 1980, Gilmore 1974).
The group 0(3) describes the orthogonal transformations in three-dimensional
space. In defining a group structure we need the following rules for its elements
to hold
ab= c (product rule)
(ab)c = a(bc) (associativity)
ea = ae = a (a unit element e exists)
aa- I = a-Ia = 1 (the inverse element a-I exists) .

33
 Fig.2.3. Two-dimensional rotation group, characterized by one
y
parameter, the rotation angle 'fJ which transfonns a point P(x, y)
into the point P'(x', y'). This operation characterizes the SO(2)
rotation group

P(x.yl

We first consider the 0(2) group (0 ::; cp ::; 271"). With each point P(x, y), after
rotation new coordinates will correspond to the point P(x', y') (Fig. 2.3) or
x' = anx + al2Y ,
y' = a2lx + a22Y . (2.11)
Invariance of the length during rotation imposes
X,2 + y,2 = (arl + ~l )x2 + (ar2 + ~2) y2 + 2( all al2 + a2l a22)xy
=x2 +y2 . (2.12)
Thus it follows that

ar I + a~l = 1 , ar2 + a~2 = 1 , an al2 + a21 a22 = 0 , (2.13)

with, as a solution,
x' = x cos cp - y sin cp
y' = x sin cp + y cos cp . (2.14)
For the three-dimensional rotation group 0(3), a similar method can be used.
Then, 2. parameters aij show up with Q conditions, leaving J free parameters
that can be chosen as the Euler-angles specifying the rotation 0 == (01,02, (3)
(Goldstein 1980). In general, for the rotation group O(n) one has r = ~n(n - 1)
independent parameters.
For the rotation group 0(3), where the orbital angular momentum L(L x , L y ,
L z ) generates the rotations, the commutation relations
[Lx, Ly] = iliL z (and cyclic permutations) ,
[L2, Li] = 0, (2.15)

hold. These operators obey the structure of a Lie algebra for which one generally
has
(2.16)
c

with X a, Xb, and Xc the generators of the Lie algebra and C~b the structure
constants. If all C~b = 0 for all a, b, c one obtains an Abelian Lie algebra.

34
Invariant operators, (also called Casimir operators or Casimir invariants of the
Lie-group) satisfy the condition
[C,Xa] =0 for all a. (2.17)

The rotation FU
P in three-dimensional space generated by L, forms a Lie
group with L as the Casimir invariant operator [see (lachello 1980,1983) for
an elementary discussion].

2.2 General Groups of 1i-ansformations

According to the definitions in Sect 2.1, a group of general transformations can

be defined. The set of n x n square matrices A also form a group under certain
conditions:
- the set is closed,
- the rules of matrix multiplication guarantee associativity,
- the identity is the n x n unit matrix 1,
- the matrices are non-singular and thus A-I exists.
Matrix groups can be finite or infinite, discrete or continuous and can be
defined over the field of real (R), complex (C), ... numbers. Variables in the
real field will be denoted by z == (Xl, X2, ••• , Xi, ••• , xn) and in the complex
field by Z == (Zl, Z2, .. , , Zi, "', zn).
In the list below, we summarize the main matrix properties related to the
original matrix A:
A =A symmetric , A = -A * imaginary ,
A = -A skew symmetric, A = A+ hermitian ,
AA = 1 orthogonal , A = -A+ skew hermitian ,
A = A* real, A+ A = 1 unitary .
Continuous groups are among the more important groups with many applications
in physics (Iachello 1983).
i) The general linear group GL(n, C) defines the most general linear trans-
formation, characterized by 2n2 real parameters. If we restrict ourselves to real
transformations we get G L(n, R) with only n 2 elements and clearly
GL(n, C) :J GL(n,R). (2.18)
ii) The special linear group of transformations has the extra condition that
detlAI = +1, and so we obtain SL(n, C) with 2n2 - 2 parameters (real). For the
real group, we get SL(n, R) with n 2 - 1 parameters and
GL(n, C) :J SL(n, C) :J SL(n, R) . (2.19)
iii) The unitary group U(n, C) is a linear, complex transformation that keeps
the "length" E IZil2 invariant. The condition is also A+ A = 1 or

35
 L aikajk =Dii , (2.20)
k
and hence lai,il 2 ~ 1. The domain of n 2 parameters for U(n) is bounded and
closed (compact group). One could also define a transformation that leaves the
quantity
p p+q

-L IZil2 + L I il Z 2 =invariant, (2.21)

i=1 p+l

defining a group U(p, q). This group is non-compact We clearly have

GL(p+q,C)::> U(p,q) ,
(2.22)
GL(n, C) ::> U(n) .

iv) The special unitary groups now have, relative to iii), also detlAI = +1,
leaving n 2 - 1 parameters for SU(n). We can, similarly, construct the special
unitary SU(p, q) groups.
v) Orthogonal groups form an n(n - 1) parameter group that leave l:: zt
invariant (AA = 1) and are denoted by O(n, C) and have detlAI = ±1. The real,
x7
orthogonal groups O(n, R) leave the quantity l:: invariant, and the particular
subset with detA = +1 are the special orthogonal transformations SO(n, R) or,
in short, SO(n). As with the unitary groups we can also define the non-compact
orthogonal groups SO(p, q).
We will not discuss the symplectic group.
As an example of a non-compact group, we discuss the 1 + 1 dimensional
Lorentz group (Jachello 1983)
x' ="'(x - "'(P(ct)
(2.23)
ct' = -"'(px + "'(ct) ,
with
P = v/c and "'( = [1 - (v/c)2r 1/ 2 . (2.24)
Here, the invariant quantity is
x 2 _ c2t2 = xt2 _ c2 t '2 . (2.25)
The range of parameters is
1<",«00
- 00 < P"'( < +00 (2.26)
i_"'(2p2=1

and the transformation (2 x 2) matrix A is expressed as (Fig. 2.4)

A _ (COSh(} -sinh8) [with 8 =arctan (v/c)] . (2.27)

- -sinh8 cosh(}

36
 Fig. 2.4. Illustration of the variation of the elements of the trans-
fonnation matrix [A] in (2.27), describing the 0(1,1) group. The
cosh () and sinh () functions are illustrated

2.3 Representations of the Rotation Operator

2.3.1 The Wigner D-Matrices

Representations of the general rotation operator of (2.10) are fonned by a set of

square n x n matrices that follow the same group rules as the rotation operator
UR (Wigner 1959). A very well known representation is formed by the matrix
elements of UR in the basis spanned by the eigenvectors of J2 and J z, i.e., the
states Ij, m}.
The new, rotated state vector obtained by acting on Ij, m} with the operator
is called Ij, m}', or
Ij,m}' = URlj,m} . (2.28)
Inserting the full set of basis states Ij, m'}, we get the result that

Ij,m}' =L Ij,m'}(j,m'IURlj,m} . (2.29)

m'

The (2j + 1) x (2j + 1) matrices (j, m'le-i/lio·Jlj, m) are called the representation
matrices of the rotation operator UR, and are denoted by D~ m (R), the Wigner
D-matrices. '
D~ ,m(R) == (j, m'le-i/lia·Jlj, m) . (2.30)
For a rotation around the z-axis, the D-matrix reduces to a diagonal matrix,
D<i)
m',m
(R) =e-ima 6m',m· (2.31)
The representation matrices now fonn a group with the same group structure as
the rotation group 50(3).
As an example, we can obtain the transformation properties of the spherical
harmonics Y;m(8,<p). By using the same method as in (2.29) we obtain

(y,m(8,<p»)' = LD~"m(R)y,m/(8,<p). (2.32)

m'

Since the new functions (y,m(8,<p»' will be identical with y,m(R-1r), we can
write that

37
 (2.33)
m'
and the inverse transfonnation (since the D-matrices are unitary) reads
Y;m(9, If!) = L D~;,m(R)Yim' (9', If!') . (2.34)
m'
Here, (9, If!) and (9', If!') are the coordinates of the same physical point.
One can work this out for Y I and thereby detennine the transfonnation
matrices for the coordinates (x, y, z) of a given point P(x, y, z) using the fonn
of the D(1)(R) Wigner matrices.

2.3.2 The Group 5U(2).Relation with 50(3)

According to Sect. 2.2, the group of complex transformations that leave the ex-
pression lul2 + Ivl 2 invariant
u' =allU + a12V
v' = a21 U + a22V , (2.35)

fonn a U(2) group. The transfonnation has ~ real parameters. In matrix fonn
(2.35) can be rewritten

(2.36)

=(~ ~), (2.37)

or
ailan + aila21 =1 ,
aila12 + aila22 = 0, (2.38)
ai2an + ana21 =0 ,
ai2al2 + ana22 = 1 .
This forms the U(2) group. Now, ~ conditions reduce the set of ~ real parameters
to~. A further condition detA = 1 gives the special group SU(2) with

(2.39)

Now, only J. real parameters remain. The transformation acts on the state vectors
in an abstract space of spinors. If we use the notation

_, (u') _ (u)
U =
v
, ; u=
v
(2.40)

38
then one has

iL'=AiL . (2.41)

The SU(2) 3-parameter group and the parameters of the SO(3) group are related.
This relation can be illustrated in some more detail. Using the substitution

(2.42)

we study how (Xl, X2, X3) transform under the SU(2) transformation matrix A.
The conditions (2.38,39) lead to a simplified notation for A as

A= (al~. a~2) • (2.43)

-a12 an
The transformation for (Xl, X2, X3) becomes

X~ = u,2 = arl Xl + 2an al2X2 + ar2X3

X2' -
- u"v -
- *
-ana12XI +( * -
auan *) X2
a12a12 + ana12X3
** (2.44)
,
x3 = V t2 = a12
*2 x I -
2an
* al2x
* 2 + an
*2 X3 •

Using the linear combinations

(2.45)

the SU(2) transformation induces the following transformation of the coordinates

of the point P(x,y,z) -+ P(x',y',z')

X
, -_ 2
1 ( an
2 - al2
*2 - al2 *2) X + 2
2 + an i ( an
2 - al2
*2 + al2
2 - an
*2) Y

+ (an a I2 +ail a i2)z,

-1.(anal2 *
- aUal2 *)
z,

with

(2.46)

Therefore we can construct for any SU(2) matrix A, a corresponding SO(3)

matrix B.
We consider some special cases:
i) Rotation R(a,O,O) around the z-axis, corresponding to a diagonal SU(2)
matrix with al2 =ai2 =0:

39
 eia/2 0) (c?sa -sina
( -ia/2 => sm a cos a (2.47)
O e 0 o
ii) Rotation R(O, (3, 0) through (3 around the y-axis, corresponding to a choice
of a real SU(2) matrix:

( co~(3/2 sin (3/2) => ( co~

(3 ~ Si~ (3) . (2.48)
- sm {3/2 cos {3/2 -sm {3
. 0 cos (3
iii) The generalized matrix R(a, (3, -y) leads to the SU(2) matrix:

COS (3/2 ei/2(a+-y) sin (3/2 ei/2(a--y) )

( -sin{3/2ei/2(-y-a) cos (3/2e-i/2(a+-y) , (2.49)

and one can (tediously) construct the related SO(3) matrix B(a, (3, -y).
We use the relations
R(a,{3,-y) = R(a)R({3)R(-y)
(2.50)
SU(2; a{3-y) = SU(2; a)SU(2; (3)SU(2;-y)
Consider the rotation R(O, 0, 0) with

(~ ~)=>1, (2.51)

and the rotation R(27r, 0, 0) with

(~1 ~1) => 1. (2.52)

Thus, there is no one-to-one correspondence between the elements of the SU(2)

and SO(3) matrices. Two elements of SU(2) correspond to one element of
SO(3). This correspondence is called a homomorphic mapping (Hamennesh
1962, Wybourne 1974, Parikh 1978).

2.3.3 AppIica~ion: Geometric Interpretation of Intrinsic Spin l

Electrons and protons are spin !
fennions. They contain a property that was
called: " ... a peculiar type of double-valuedness, not describable in classical
tenns" (pauli 1925). This double-valuedness was illustrated experimentally in
the Stem-Gerlach experiments.
The spin eigenvectors have been represented in an abstract two-dimensional
space by using column matrices with two rows. In general

X= (~:) , (2.53)

where X+ and X _ are amplitudes, and can also be written as

40
 x = x+a + x-{3, where

(2.54)

We have defined the spin ! angular momentum operators (pauli-matrices) as

(2.55)

Intrinsic spin, described by using the abstract state vectors, has no, or only
limited, relation to the more intuitive notion of spin as a vector pointing in a
certain direction. This "classical" correspondence to an angular momentum vector
operator can be expressed in a vector model by sz, Sy, Sz and the expectation
values for the different spin! eigenvectors (sz), (Sy), (sz).
We can now put a link between the abstract rotations and geometric picture
in the following way: we calculate the expectation values of sz, Sy, Sz, for a
spin! eigenvector rotated over a general set of angles (a, (3, I) expressed via
the D-matrices. One then has

.
(2.56)
m'
with
(1/2) _ (cos (3/2e-i /2(a+"() - sin (3/2e-i /2(Ot-"(»)
D,
m. ,m. - sin {3/2e-1O/2()
"(-Ot cos {3/2e1O/2( Ot+"() •
(2.57)

The calculation leads to

(Sxh/2,1/2 =(1i/2) sin {3cos a
(Syh/2,1/2 =(1i/2) sin {3sin a (2.58)
(szh/2,1/2 = (1i/2)cos{3 ,
and
{Szh/2,-1/2 = -(1i/2) sin {3cos a
(Syh/2,-1/2 = -(1i/2) sin{3 sin a (2.59)
(Szh/2,-1/2 = -(1i/2) cos {3 .
Thus, we find the "classical" vector via the expectation values of the components
of the spin ! angular momentum operators in the spin ! and -!
eigenstates
(Fig. 2.5).
For the particular choice a = 1 = 0° ,

I!, !)'=cos {3/21!, !} + sin {3/21!, -!} (2.60)

I!, -!}' = - sin {3/21!, !} + cos {3/21!, -!} .
For {3 = 271", one has

41
 Fig. 2.5. Semi-classical representation of the spin
z
vector in a cartesian system. The classical vector
/ is given in the two possible states (double arrow)
corresponding to the two possible orientations of
the spin 1/2 angular momentum vector. The pro-
jections on the X-, y- and z-axis given in terms of
the rotation angles a, f3 are given by (2.58) and
(2.59), a denoting a rotation around the z-axis,
f3 a rotation around the y-axis

I y

\,~
/ \ ,
\1
x

(j)
.~
::>
i::'
z'"
15
~
~
U5
zw
~
~

0 20 40 60 80
a) b) CURRENT (rnA)

Fig. 2.6. Demonstration of spinor rotation for a spin 1/2 system: (a) Schematic diagram of a neutron
interferometer. The phase shift in path I is produced by inserting a magnetic field as shown. (b)
The magnetic field induces a rotation of the neutron spin (wave function) and changes its phase, a
rotation of 41[" produces a phase shift of 21[" [from (Werner 1980)]

112' 1}'
2
= _112' 1}
2
(2.61)
I!, -!Y = -I!, -!} ,
and only for f3 = 411" does the
state coincide with the original state. This phase
change has been observed experimentally in neutron interferometry [Fig. 2.6
(Werner 1980)].

42
2.4 Product Representations and Irreducibility

In Sect. 2.3, we have constructed the Wigner D~ m (R) representation matrices

for a general rotation of the system R(a, (j, 7). TIiese D-matrices transform the
eigenstates Ii, m} under a rotation of the physical system. Thus we can write for
the transformed eigenstates of a system lit, ml}

(2.62)

A second set of eigenstates for system ~ transform according to:

Ih,m2}' = LD~t,m2(R)lh,m~} . (2.63)

m~

The product states lit m I }Ihm2} then transform as

(Iit ml}lhm2})' = "L...." D~~)l' ml(R)D~~)2' m2(R)lil,m~}lh,~} (2.64)

m~,m;

(2.65)

where DUl Xh)(R) is the product matrix describing the product representations. It
is a direct product of D-matrices, also denoted by the notation A®B (Hamermesh
1962).
We will work this out in more detail for the particular situation of two spin
! particles: the eigenvectors are

and ( U2)
V2 '

(2.66)

The product transformation matrix D(I/2) ® D(I/2) becomes

an all allan
an an al2a21 a t2 a l2 )
allan (2.67)
a21 all a22an anan
a21 an a22a21 anan
Consider now the eigenstates

43
 1
cpo,o == I!, ! ;OO} = v'2(UIV2 - VIU2)

CPI,I == I!, !; ll} =UIU2 (2.68)

1
CPI,O == I!, !; to} = v'2 (UIV2 + VI U2)
CPI,-l == I!,!; I-I} =VIV2·
The transfonnation of the eigenvectors will also modify the product transfonna-
tion matrix. In shorthand notation: if cp describes the new basis and Z is the
matrix describing the transfonnation (2.68), we have

cp=Z·cP, (2.69)

and the general transfonnation (2.65)

cp' = A· cP , (2.70)

can be rewritten in the new basis as

Zcp' = ZAZ- I Z . cP , (2.71)

or
(2.72)

In the particular situation of two spin! particles, via (2.68) the matrix Z becomes

(0 -~ 1 0
~
0
0)
0
(2.73)
Z= 0 -L I 0 .
Vi V72
o 0 0 1
The product matrix ZAZ- I becomes (in the new basis) in shorthand notation
o o
(
CPOO)'
CPI:I = (b0 ct,l ct,o o
CJ,-l ) (CPOO
CPI:I ) . (2.74)
CPI,O 0 CO,1 CO,O CO,-l CPI,O
CPI,-l 0 C-I,l C-I,O C-I,-l CPI,-l
The product transfonnation D(I/2) ® D(I/2) is in this case reducible in a spin
1 and spin 0 system. In the more general case of a product of representation
matrices for particles with angular momentum it and h one obtains
D(it> ® D(h) = D(h+h) ED D(h+h- l ) ED ... ED D(lh-hl) , (2.75)

and the product transfonnation is reducible in blocks, corresponding to angular

momentum J = jl + h, jl + h - 1, ... Ijl - hi, and can be depicted diagram-
matically as in Fig. 2.7.

44
 Fig. 2.7. Diagrammatic illustration of the reducible
parts of the transfonnation Wigner D-matrix for the
product system of the angular momentum eigenvec-
tors lit ml)l12 m2). The blocks present the different
irreducible parts, corresponding to the momenta i
with lit - i21 ~ i ~ it + n

The detailed product elements D~~ m (R) are related to the separate elements
of the separate D-matrices via the Clebsch-Gordan series

(2.76)

A proof of (2.76) can be found in Hamermesh 1962, Edmonds 1957.

2.5 Cartesian Tensors, Spherical Tensors, Irreducible Tensors

In considering a vector X(Xt, X2, X3) with the above cartesian coordinates, under
a genuine rotation of the vector the new coordinates can be expressed as

X' = [A]X or xi = LAi,kXk. (2.77)

k

The components Ai,k constitute a real, orthogonal matrix with determinant +1.
If we now consider two vectors X(Xt, X2, X3) and Y(Yt, Y2, Y3), under a rotation
of the vectors the product quantities XkYI will transform like

xiyj = L Ai,kAj,IXkYI . (2.78)

k,l

It is also possible to have a non-separable quantity Fkl that transforms like the
product x k Yl, or

F[j =L Ai,kAj,IFkl . (2.79)

k,l

The quantities XkYI, Fkl, .•. transform like the components of a cartesian tensor
of rank 1. A more geneml tensor (cartesian) of rank r will then transform as

45
 FI j k ...
~
=,L Ai"Aj,mAk,n, ... F',m,n, ....
,"--0.,.-"
(2.80)
r ; r

In general, the object F is reducible with respect to the group of orthogonal

transformations SO(3), indicating that it is possible by using appropriate linear
combinations of the cartesian tensor components F"m,n, ... , to construct a number
of contributions that transform independently of each other under the group
SO(3) (Brink, Satchler 1962, Rose, Brink 1967, Weissbluth 1978).
For a cartesian tensor of rank 2, we can write (identity)
(2.81)
with
Ti,k == t8i,kTr(Fi,k) , (2.82)

A·I, k == !2 (F:I, k - F'ik ,.)I ' (2.83)

~~l =! (Fi,k + Fk,i) - Ti,k , (2.84)

with Ti k an invariant quantity, Ai k the components of a vector and ~o~ the five
compo~ents of a symmetric tenso~ of rank 2. '
H, as an example, we take the 9 components XiX j (XX, xy, ... , zz), we can
construct
= lr
r·I, k
3 8· k
2
I,

Ai,k = !(yz - zy), ... (2.85)

S i,k
(O) _
-
1( ) 1 21:
2 yz + zy - 3r Vi,k, ... ,

where Ti,k is proportional to the length of the scalar product l' . 1', Ai,k form
the three independent components of the vector product l' x l' and the ~o~ form
the five independent components of the symmetric tensor 1'1' - tr28i,k with zero
diagonal sum. It can be shown that in this example the quantities Ti,k, Ai,k and
~o~ transform under rotation according to the spherical harmonics of order 0, 1
and 2, i.e., Yo, Yl' and Y!, respectively (Weissbluth 1978).
The quantities Ti k, Ai k and Sn
transform among themselves (see e.g.
Weissbluth 1978) and are 'therefore 'called "irreducible" tensors of rank 0, 1
and 2. In general, an irreducible spherical tensor of rank: k is a set of 2k + 1
independent components ~k) (with -k ::; ", ::; +k) that transform under rota-
tion as the spherical harmonics of the same rank Ykl< do. Thus we obtain the
transformation law

(2.86)

The quantities S~k) == (-1)1< T~~+ also form an irreducible tensor of rank: k
(Brink, Satchler 1962, Edmonds 1957), a fact that can be proved by using the
relation for the D-matrices

46
 (2.87)

For example, for the radius vector r of a point P we have the cartesian coordi-
nates r(x, y, z) that form a cartesian tensor of rank 1. The spherical components
are r(r+l,r_l,ro), with

r+l = - ~(x +iy)

r-l = ~(x - iy)
(2.88)

ro = z,
form a spherical tensor of rank 1 and are proportional to Yj+l, Yj-l and Yjo,
respectively.

2.6 Tensor Product

By using the Clebsch-Gordan coefficients just as in angular momentum coupling,

spherical tensors of higher rank can be constructed in a systematic way. This
"building"-up process goes like (Brink, Satchler 1962, Edmonds 1957, de-Shalit,
Talmi 1963)

T~!a) =L (kl Kl, k2K2Ik3K3)Ti~1)Ti~2) , (2.89)

"1 t IC 2

or, in a short-hand notation

Ti!a) = [rkt> ® rk2) ]:a) . (2.90)

One can prove that the 2k3 + 1 components of 11!a)

form the components of a
spherical tensor of rank k3 if the quantities rkl) and rk2) are spherical tensors
of rank kl and k2, respectively.
The relation (2.90) can also be inverted to give

Ti~1)Ti~2) =L (kl Kl, k2K2I k3K3)Ti!a) . (2.91)

ka'' a
As an example, it is possible to show that the tensor product of the vectors r and
p, coupled to a tensor of rank 1, becomes proportional to the angular momentum
vector I. The exact relation is

(2.92)

A particular case is the tensor product of rank 0, thereby forming a scalar (invari-
ant) quantity. One therefore often uses the notation of a scalar product (de-Shalit,
Talmi 1963)

47
 -r k). U(k) == (_l)kk[-rk) ® U<k)tO) . (2.93)

By using the definition in (2.89) one then gets

-r k). U(k) = L(-l)le~k)U~~ . (2.94)

Ie

2.7 Spherical Tensor Operators: The Wigner-Eckart Theorem

Up to now, we have studied the transfonnation properties of wave functions, state

vectors or, more general, of tensor quantities. The operators used in quantum
mechanics that relate to observables can be classified according to their transfor-
mation properties: we speak of "scalar, vector, ... tensor of rank k" operators.
They are the key objects that appear when calculating measurable quantities that
can be expressed as matrix elements of a given spherical tensor operator acting
between eigenstates of the angular momentum J2 and J z:

J
{aljlm/l~k)lajm} = tfJ~/jlm/(z)~k)(Z)tfJajm(z)dz, (2.95)

where at the right hand side, z denotes all coordinates that are present in the
wave function and in the operator.
We now consider the general state vector ItfJ}. Under a rotation (characterized
by the rotation operator UR), the new state vector will be given by ItfJ}', via

The operator A(A;) is now called a vector operator if the matrix elements trans-
form as the components of a vector, i.e.,

j
(2.96)
{tfJIU1iA;URltfJ} = LR;,j{tfJIAjltfJ} .
j

This should hold for any ItfJ} vector, so we obtain, fonnally,

U1iA;UR = LR;,jAj , or
j
(2.97)
URA;U1i = LRj,;Aj .
j

We now call the 2k + 1 components '1!k) a spherical tensor operator if, likewise,
the components '1!k) transfonn according to the spherical harmonics of rank k,
thus
~k)1 = UR~k)U+ = '" D(k) (R)T k) (2.98)
" "R L.J Ie' ,Ie tr.'.

48
When calculating the matrix elements of spherical tensor operator components
T.ck ), the Wigner-Eckart theorem allows a separation in the part that only depends
on the projection quantum numbers (called the geometrical part) and another part
that depends on, e.g., the radial properties (and angular momentum properties) of
the operator and of the state vectors (Eckart 1930). The latter is called a reduced
matrix element defined by

(ajmlT.ck)la'j'm') =(_I)i- m (!m : ~,) (ajIlTk)lIa'j') . (2.99)

We give no proof here [see e.g. (Brussaard, Glaudemans 1977)]. Expression

(2.99) can also be rewritten as

(ajml~k)la'j'm') =G)-1(_1)2k{j'm', kltljm){ajIlTk)lIa'j') . (2.100)

[Here, and in further discussions, we use Jas a shorthand notation for (2j +1)1/2.]
We illustrate this by calculating the reduced matrix element {j IIi IIj}. We know
that

m = Umljzljm) = (_I)i- m (j
-m
1
It
j ) Ullillj)
m
=(_1)2<i- m )m(j(j + 1)(2j + 1»)-1/2Ullillj), or

Ullillj) = (j(j + 1)(2j + l)i/ 2 . (2.101)

Likewise, one can prove that

UIIRllj) =3 , (2.102)

where R is the unit-operator.

2.8 Calculation of Matrix Elements

In the general program of calculating matrix elements of spherical tensor oper-

ators, one deals with composite wave functions and operators, i.e., one handles
many-body nuclear wave functions and operators symmetrized in all the par-
ticipating particles. It should be possible to reduce such matrix elements to their
basic entities which are the reduced matrix elements of one-body spherical tensor
operators and the wave functions of a single system. The rules that allow such a
reduction use techniques of angular momentum coupling and recoupling (Racah
1942a,b,1943,1949,1951). We shall consider two distinct cases to which more
general situations can always be reduced.

49
2.8.1 Reduction Rule I
We consider the tensor operator as a tensor product of operators acting on two
independent subsystems

(2.103)

denoted by 1 and 2, respectively. Here 1 stands for all coordinates characteriz-

ing subsystem 1, i.e., rt, O"t, •••• In the same way, the state vectors will be
the angular momentum coupled eigenvectors of the two subsystems denoted by
ladt, a2h; J M}.
The reduced matrix element can then be reduced to the separate reduced
matrix elements and some recoupling coefficient. In particular one gets
{atjt, a2h; JIIT k)(I, 2)lIaUL a2j~; J'}
jt h J}
=jj'I: { j{ j~ J' {atitllrkl)lIaUD{a2hllrk2)lla2i~}.
kt k2 k
(2.104)
The way to obtain the above result is outlined in Appendix C. This method can
be used for the reduction of a matrix element into its basic ingredients. We now
consider some particular cases of the type (2.104).
i) H the operator r k)(1,2) reduces to a scalar product (i.e., k = 0), or

r k)(1) . U(k)(2) ,

the reduction fonnula (2.104) leads to

(adt, a2h; JIIT k)(1) . U(k) (2) II aU{ , a2j~; J'}
=(_l)i2+j~+J j {~! ~; Jk } (adlI1Tk)lIaUD{a2hIlU(k)lIa2j~}c5JJ' .
12 Jt
(2.105)
ii) H one of the operators is a unit operator, this means that the product tensor
operator is

T,.k) = [r k)(1) ® n]~), or

(2.106)
T,.k) = [n ® rk)(2)]~) .

Matrix elements can also be easily reduced with the result (we give some extra
intennediate steps in working this out)
{adt, a2h; JIIT(k)(1)lIaU{, a2j~; J'}
= (adt, a2h; JII [r k)(1) ® n] lIaU{, a2j~; J'}

50
 = JJ'(_1)it+i2+ J'+k {~, ~ Jt~ } {o:dtIl Tk)lIo:UD8J·21·,8
2 a 2a'2 .
(2.107)
We can similarly calculate the case when the tensor operator only acts on the
coordinates of the second system. We only give the result:

{o:dt, 0:2i2; JIITk )(2)lIo:U{, o:ii~; J'}

=JJ'(_1)it+i~+J+k{jtk J2
J'
J}{ 0:2J2·IITk)11 0:212
j~
'·'}8 .. 8
1tl~ at a~ •

(2.108)
2.8.2 Reduction Rule II
In contrast to rule I, it is also possible to construct a composite tensor operator
that acts on a single system, giving

(2.109)

Now the state vector is not a product state but only a single eigenstate Io:jm}.
In a way that is analogous to the method of rule I, we can derive the reduced
matrix element as (Brussaard, Glaudemans 1977)
{o:jIlT k)(1)lIo:'j'} = (_1)i+i'+kk
x L {~! k~ ~,} {o:jIlTkt)1I0:"j"}{0:"j"IIT k2)1I0:'j'}. (2.110)
a "."
,J J J J
By using the above rules, any matrix element can be reduced into a small number
of basic one-body reduced matrix elements. The Hamiltonian, being a scalar
operator (tensor operator of rank k = 0) can quite often be decomposed into a
sum over multiple operators

H(1,2) = LY ~(1)· Y~(2)f(rt,r2), (2.111)

~

and matrix elements can be calculated by using the reduction rule I. The operators
appearing in the electromagnetic one-body operators can mainly be reduced to
products of the basic matrix elements

(2.112)

from which all more complicated cases can be built.

51
Short Overview of Rotation Properties, Tensor Operators,
Matrix Elements

Rotation of scalar function field

x'=Rx
f'(x) = URf(x) -+ UR = e-i / Acp .L •

Representations of 0(3) and SU(2)

Ijm)' = URljm)
11m )' = LD!n',m(R)llm)
m'
1 = 0, 1,2, ....
SU(2) is homomorphic with SO(3)

m'
j = 0, 1, 1, ~, 2, ....

Product representations: Clebsch-Gordan series

it+h
D(it) ® D(h) = L D(I)
J=lil-hl
il+i2
D~\),ml (R)D~r,m/R) = L L (jl ml,h m2IJM)
J=lil-hl M,M'
X (jlmLhmiIJM')D~,M(R)

52
 Tensors, Spherical tensors, Irreducible tensors

FJ k = '"' Ai,IAj,mAk,n .•• FZ m n ...

., , , L..J ' - - " " '"
JO 000 0

--.....-.- l,m,n,ooo ______

r r r

Cartesian:
r(x,y, z)
p(Px,Py,Pz)
1(lx, Iy, lz)
productsxiXj, PiPj, XiPj : reducible.

Spherical:

r(r+l,r-t,ro) with r±l =1= ~(X±iy),ro=zo

Tensor product of spherical tensors

Wigner-Eckart theorem

Reduction rules

- rule I (see text)

- rule II (see text) °

53
 3. The Nuclear Shell Model

3.1 One-particle Excitations

3.1.1 Introduction
The basic assumption in the nuclear shell model is that, to first order, each
nucleon (proton or neutron) is moving in an independent way in an average
field. This is not so, a priori, since the nucleus constitutes an A -body problem
interacting via the nucleon-nucleon force in the nuclear medium. It is clear from
the very beginning that this nucleon-nucleon force will be different from the free
nucleon-nucleon interaction (Bohr, Mottelson 1969). As was already expressed
in the introduction to this book, the non-relativistic form of the nucleon-nucleon
interaction behaves as shown in Fig. 3.1. At large separations IT; - Tjl ';it l.5 -
2 fm, the force behaves according to a one-pion exchange potential (OPEP) which
has an analytic dependence on r = IT; - Tjl of (Bohr, Mottelson 1969)

V(r) = e- pr
f.tr
(1 + ~ + _3_) .
f.tr (f.tr)2
(3.1)

For small distances the attractive part turns over and becomes repulsive at dis-
tances r < 0.5 fm; this is the hard-core potential. At such short distances the
energy for interactions between nucleons becomes so high that non-relativistic
treatments are no longer justified. In this region, exchange of more pions or
heavy mesons is needed. It is probably even more correct to go to QeD where

v..
I,J

2 (fm)

"-
OPEP regIon Fig.3.1. Schematic illustration of the nucleon two-
body interaction Vi,j as a function of the nucleon
separation r = Iri - rj I. For large separation
(r ~ 1.5 - 21m), the OPEP tail results. For shott
distances, a shott range repulsive core shows up

54
nucleons are considered as being obtained from their quark constituents and so,
a nucleon-nucleon interaction process can be better depicted, as in Fig. 3.2.
One of the most unexpected features is still the very large nuclear mean free
path in the nuclear medium (Fig. 3.3).
In the study of the nuclear structure observed at low excitation energy (Ex <
8 MeV), two important aspects show up:
- how to handle (even in a non-relativistic way) the A-nucleon problem,
- how to describe the nuclear average field starting from the nucleon-nucleon
force V(ITi - Tjl) between free nucleons.
In the atomic case, a shell structure was shown to exist by N. Bohr. Starting
from an average Coulomb field V(r) = -Z e2 lr, the corresponding one-electron
SchrOdinger equation can be solved and the atomic orbits studied in detail. Ac-
cording to the Pauli principle, no more than two particles of intrinsic spin can !
be in a specific quantum state defined by the radial quantum number n, orbital

N N N

N N N N
a) b)

Fig. 3.2. The nucleon-nucleon interaction (a) on the level of QHD (quantum hadro-dynamics) where
the force is mediated by the exchange of pions (11"±, 11"0) between the interacting nucleons, or (b) on
the level of QCD (quantum chromo-dynamics) where a gluon is exchanged between one of the three
quarks constituting the nucleon. The left-hand side diagram is zero due to the colour selection rules,
the right-hand diagram contributes since colour selection rules are obeyed

•
•

Fig.3.3. The empirical mean free path ()..(fm»). The shaded

band denotes the range of values determined from reaction cross
sections for Ca, 'h and Pb in (Nadasen 1981), the solid line in-
dicates the representative values selected in (Bohr 1969) and the
data points are determined from optical potential fits for 208Pb
in (Nadasen 1981) [taken from (Negele 1983)]

55
Volts ,----,----,----r----,-----,---..,.-----,----.----,-"-t
~~--_r--_;---+_---_+----r_---~---_+---+_----_r_;
24
23
22
21
20
19
18
17
16

14
13
12
11
10

Fig. 3.4. Dependence of the ionization potential of the neutral atom on the atomic number Z [taken
from (Herzberg 1944)]

(l) and total (j) angular momentum. For each j, the 2j + 1 magnetic substates
with - j ~ m ~ j are degenerate and form a given shell structure. A number of
subshells now form a major shell. Atoms with a major closed-shell configuration
form configurations that are particularly stable against losing the last electron
(Figure 3.4 shows the ionization potentials of the elements).
In the nucleus a similar description seems to be possible. However, a number
of distinct differences to the atomic case arise:
i) The nuclear mean field is very different from the Coulomb potential. More-
over, strong spin-orbit coupling is shown to exist in nuclei (Sect. 3.1.3).
ii) In the nucleus both protons and neutrons are present.
iii) There is no preferential central point other than the center of mass in the
nucleus in contrast to the atomic field generated by the atomic nucleus.
Because of the above conditions the shell structure in the atomic nucleus will
be very different from the corresponding shell structure in the atom.
We now quote a number of nuclear properties that unambiguously point
towards nuclear shell structure and increased nuclear stability when either the
proton number (Z) and/or the neutron number (N) has a certain "magic" value.
i) Deviations of the nuclear mass (binding energy) from the mean, liquid drop
value exist (Fig. 3.5).

56
 £ -5
I
d.
~
L-l0
UJ
u 20 1.0 60 80 100
z
UJ PROTON NUMBER l
0:
UJ
u..
u..
Ci
5
VI
VI
<t -'.
L

-5

·10
L
20 40 60 80 100 120
NEUTRON NUMBER N

Fig. 3.5. Deviations of nuclear masses from their mean (liquid drop) values. and this as a function of neutron and proton number
[taken from (Myers 1966)]

01
......
 EXPERIMENTAL EVIDENCE FOR MAGIC NUMBERS

6-

___ Number of neutrons

Fig.3.6. The magic numbers, demonstrated by the excitation energy for the first excited state in
doubly-even nuclei (mainly a J1f = 2+ level) plotted as a function of the neutron number N [taken
from (Brussaard 1977)]

ii) Since nucleons couple into J1< = 0+ coupled pairs, the way to excite nuclei
(the excitation energy of the first excited state which is most often a J1< = 2+
state) as a function of neutron number (Fig. 3.6) again correlates very well
with the shell closures as obtained under i).
iii) Specific tests result when, in one-nucleon transfer reactions (pick-up or strip-
ping), a nucleon is taken out or added to a nucleus with given A (Z, N)
nucleon constitution. In the case of adding a proton to ~~Pbl26 via a eHe, d)
reaction, it is clearly observed that the extra proton is placed in very specific
nuclear shell model orbitals (Fig. 3.7).
The stable nucleon configurations so detennined are N (or Z) = 2, 8, 20, 28,
50,82, 126, (... ). These numbers can now be explained by starting from a one-
body SchrOdinger equation using a central average (and attractive) field U(r) to
which a strong spin-orbit interaction term (I· s has been added.
Suppose that <f'a(r) (a = n a, la, ja, m a, ...) are solutions to the SchrOdinger
one-body equation (here r is a notation for all coordinates, r == r, U, ...)
[T + U(r)]<f'a(r) = ca<f'a(r) . (3.2)

Here T describes the kinetic energy, U(r) the average field (Sect. 3.1.4) and
Ca the single-particle energy (recall that we do not write the spin and charge
coordinates explicitly).

58
 1hg12

Excitation energy In MeV

5 4 3 2 o

[""
350 3r, 2
1/ = 100

300

E
.,.E 250

...
0. '"
"'g"'""
200

...0..
'-

150
~
E
::>
Z
100

50

70 75 80 85 90
DIstance In em on the plates.

Fig. 3.7. Single-particle states in 209Bi, obtained from the pick-up reaction 208Pb eHe,dj209Bi. At the
angle of 8 = 100°, the 11kJ/2 ground-state level is less populated than the excited states but at other
angles, this situation can become reversed. Above the actual spectrum, the proton single-particle
states are drawn as a illustration [taken from (Mottelson 1967)]

Orthogonality demands that

J <p:(r)<Pb(r) dr =Dab. (3.3)

The model Hamiltonian for the A nucleons (taken as independent particles) can
then be written as
A A
Ho = :~::::CTi + U(ri») = L ho(i) . (3.4)
i=l i=l

59
The eigenfunctions of Ho are now of the product type
A
!liat,a2, ... , aA (rl, rz, ••• , r A) = II 4'a;(ri) , (3.5)
i=1
with the corresponding energy eigenvalue
A
Eo= Lea;. (3.6)
i=1
For a number of identical nucleons, the wave function (3.5) is not well con-
structed: the Pauli (exclusion) principle is not fulfilled. The correct wave function,
e.g., for two particles becomes
1
!lia1 ,a2(rl,rz) = Vi (4' al(rl)4' a2(rz) - 4'al(rz)4'a2(rl») , (3.7)

or, rewritten in (Slater) determinant form

tP ( ) - 1 (4'a 1(rl) 4'al (rz») (3.8)

at, a2 rl, rz - Vi 4' a2(rl) 4' a2(rz)

For an A-nucleon wave function, the generalization of (3.8) becomes an A-

particle Slater determinant.
We should note that the average field expressed by the potential U(r) is not
explicitly given. In fact, one has to start from the A-nucleon Hamiltonian
A 1 A
H = LTi + 2 L Vi,i , (3.9)
i=1 i,i=1
(restricting to two-body interactions only), to write the Hamiltonian that can be

t ,t t
expressed

H= [T, + U(r,)] + ( ~ Yo,j - U(r,») (3.10)

= Ho + Hres
A
L ho(i) + Hres ,
(3.11)
=
;=1

where Ho describes the motion of A nucleons, independent of each other in the

same average field. The smaller the effect of H res , the better the assumption
of an average, independent field becomes. The method of determining U(r),
starting from a known Vi,i and a Slater determinant A-nucleon wave function
that is a good approximation to the total ground state wave function for the full
Hamiltonian H, is carried out by using the Hartree-Fock method (Sect. 3.1.4).
Before that, however, we study the independent nucleon motion with a harmonic
oscillator potential U(r) = !mwzrz.

60
3.1.2 The Radial Equation and the Single-particle Spectrum:
the Harmonic Oscillator in the Shell Model
If we start from the central, one-body problem discussed in Chap. 1 (1.14-18)
the total wave function can be written as

c,o(r) = R(r)Y(9, c,o) ,

(3.12)
= u(r)Y(9,c,o) .
r
The equation governing the radial motion is then

li,2 tPu(r) 2 2
- 2m ~ + [1(1 + 1)1i, /2mr + U(r)]u(r) = Eu(r). (3.13)

In studying bound states (E < 0), conditions have to be imposed on the radial
solution u(r):

(3.14)
u(O) =0.

The normalization of the radial wave function leads to the integrals

1 00
R(r2)r2 dr 1
= 00
u2(r) dr =1 . (3.15)

Starting now from the harmonic oscillator potential U(r) = !mw 2r2, we obtain
the radial Laguerre equation, with the solution (Abramowitz, Stegun 1964)

= N k,I' r 1+1 e -vr L 'k+ / (2 v 2)

1 2
ukl () (3.16)
2
r r,

with Nk,1 a normalization factor, v = mw/21i, the oscillator frequency and the
Laguerre polynomial, given by (Abramowitz, Stegun 1964)
k
L 'k+1/ 2( x ) -_ "L.J ak' (l)k'
- x k' . (3.17)
k'=O

A number of radial solutions are illustrated for Z = 82, N = 126 for the neutron
motion. Although the wave functions in Fig. 3.8 are calculated for a more realis-
tic potential, a Woods-Saxon potential (Blomqvist, Wahlbom 1960), the overall
behavior is the same as for the harmonic oscillator potential.
The energy eigenValues corresponding with the eigenfunctions (3.16) are
given by

E =1tw (2k + 1+ n= 1tw (N + !) , (3.18)

with

61
 SJ?'rdr=1
OO 2 2

o
15 liz
05

a
20 r(fm)

-0.5

fD
rVl,nlj ()
r (fm)-V2
SJ?, r
00
2 2
dr =1
o
1i 13/2 1j 1512
1i 1112 2f 5/2
1h9/2
3p 1/2
0.5 29 7/2
3d 5/2

o
15 20 r(fm)

-0.5

=
Fig. 3.8. Neutron radial wave functions for A =208 and Z = 82 Unlj (r) (n 1,2, ...) [based on the
calculations with a Woods-Saxon potential by (Blomqvist 1960)] [taken from (Bohr, Mottelson 1969)]

62
 N =0,1,2, (major oscillator quantum number) ,
I = N, N - 2, ... , 1 or 0 (orbital quantum number) , (3.19)
k =(N -1)/2 (radial quantum number) .
Thus, the spectrum of eigenvalues presents a large number of degenerate (I, k)
quantum numbers corresponding to a fixed N major oscillator quantum number
(Fig. 3.9).
The radial quantum number more often used (de-Shalit, Talmi 1963) is related
to k via

n = k + 1 = (N - 1+ 2)/2 , (3.20)

and expresses the number of nodes of the radial wave function in the interval
(0, 00) including the node at the origin (excluding the one at infinity).
We now give a number of interesting properties of the Laguerre polyno-
mials that allow for an elegant calculation of the normalization factor Nk,l
(Abramowitz, Stegun 1964)

10 00
za e- z Lk(z)Lk,(z) dz =Okk' • r(k + a + 1)3/k! . (3.21)

We use (3.21) in calculating the norm Nk,l by putting z = 2vr2, r = (z/2v)1/2

and dr = dz/[2(2vz)1/2] and evaluate the integral

(0.4)
&(1.2) /Og (18)
-·1d (10) [70]
0.2,4 (2,0) "2s (2)

(0,3) _Of
(14)
3 1,3 [40]
(1.1) '1p (6)

(0.2) _Od (10)

2 0,2 (1.0) ... 1s (2) [20]

Fig. 3.9. Dlustration of the degenerate
hannonic oscillator energy spectrum up
-Op to N = 4. Besides the major shell quan-
10,1) (6) [8] tum nwnber (N), the (k, l) degeneracies
are drawn explicitly. Partial and cwnu-
lative occupation nwnbers are given in
- Os (2) [2] round and squared brackets, respectively
N=O 1=0 k=O,I=O

63
 =N;,d(2(2v)I+3/2) . r(l + ~ + k Y/k! = 1 or

Nk,l = (2(2v)I+3/2k!/ r(l + k + 3/2)3i/2 . (3.22)

Summarizing some properties:

r(z + 1) = zr(z), r(D = Vi ,

(3.23)
Lg(z) = r(a + 1), L~/\z) = Vi/2 .
The radial wave functions can be determined as follows, since
dk
Lk(z) = r(a + k + 1)/k! e Z z-a dz k (za+k e- Z ) , (3.24)

(k = entire). As an example, we illustrate the above equation for L~/\z):

L 11/ 2(Z) = ~.
2
Vi(~
2 2 _ z) .
(3.25)

A number of often used expressions relating to the calculation of radial integrals

are

(3.26)
(p > 0) .

Having determined the solutions to the radial equation for a single-particle

harmonic oscillator potential, we observe a large degeneracy in the orbitals.
Moreover, the nucleon numbers that form the stable configurations are N,
Z = 2, 8, 20, 40, 70 and do not agree with the experimental numbers. There
is a strong spin-orbit term (r) I . s that is modifying the single-particle spec-
trum in the right direction. We shall work this out in more detail. The original
Hamiltonian ho becomes

h = ho + (r)l . s . (3.27)

The consequences were originally worked out by Mayer (Mayer 1949,1950) and
Haxel, Jensen and Suess (Haxel et al. 1949).
The single-particle wave functions that were detennined in Chap. 1 are eigen-
functions of ho. Moreover, both the parallel and anti-parallel orientations corre-
spond to the same energy eigenvalue c:~lj since we have

(nlj, mlholnlj, m) = c:~lj , (3.28)

with
unl(r) [ 1/2 ](j)
(r,ulnlj,m}=-r- Y,(B,cp)0X (u) m' (3.29)

64
 the (I, !)j coupled single-particle wave function. By using the basis of (3.29)
the energy correction for the spin-orbital tenn is easily detennined since we can
express the spin-orbit tenn (r)l . s as

(3.30)

Thus we obtain
(0)
€nlj = €nlj
A
+ ~€nlj , (3.31)
with
L1€nlj = (nlj, ml(r)l . slnlj, m) , (3.32)
or

L1€nlj = ~ [j(j + 1) - l(l + 1) - ~] . (3.33)

We define

D= f u~,(r)(r) dr . (3.34)

This gives rise to a spin-orbit splitting of

D
L1€nlj=I+1/2 ="2 ·1
(3.35)
D
L1€nlj=I-1/2 = -"2(1 + 1) ,

as illustrated in Fig. 3. lOb.

A much used fonn of (r) is the derivative of the average U(r) potential and
is shown for a Woods-Saxon potential in Fig. 3. lOa. One can thus express (r)
as

dU(r)
err
/%. (1+1)

----1(----
U(r)
\ Dh· 1
=1+%

0) b)
Fig. 3.10. (a) We draw the possible radial fonn for the spin-orbit strength function «r) as detennined
by the derivative of a Woods-Saxon potential, described by (3.36). This function (r) peaks at the
nuclear surface. (b) The spin-orbit splitting between j =
1 ± ~ partners, according to (3.35), using
the factor D, with D == Ju!/(r)(r)dr [see (3.34)]
65
 I
5 d --{f'---- ~
6 9 --t\r----l~

1'~ ~
12. -.-:~:~~~--.-.-.: ----------~%-
SI ~ % 'Y2M
I ~~ ~

.2 ---::-j---~~--.---- -----~t-·-
50 ---~!df----- --~------- ----~---
~ 11

2. ---~~----.---~~--.-----l--.-.---~--
~ 2

20 -------------.-.-.---.-----.----------------------

~:---<
~

~
~ -{ ~
~
~

a ---~~-~--_------.-_~_-_---_-_=_- .------~-.-

'''- ____
Proton,
------ ~
Fig.3.11. A full single-particle spectrum, including tenns that split both the spin-orbit and angular
momentum degeneracies in the hannonic oscillator case of Fig. 3.9 (n = 1,2, ...). A level scheme
for both protons and neutrons is given [taken from (Klingenberg 1952)]

1 aU(r)
(r) =
2
v, ... ro· -
r
-a-
r
. (3.36)

A full single-particle spectrum, including a term proportional to 12 for splitting

the remaining degeneracies on (k,l) for given N in addition to the spin-orbit
interaction, is illustrated in Fig. 3.11. This figure gives a general idea of the

66
 p

Neutrons ~
p "- ~
p~

Fig.3.12. Oassical illustration of the existence of a spin-orbit interaction. A beam of unpolarized

neutrons is directed to 4He scattering nuclei. If the coefficient multiplying the spin-orbit interaction is
such as to favour parallel orientation, then neutrons passing on the lower side have their spin pointing
out of the plane (0), the neutrons on the upper side have their spin pointing into the plane (®). By
collimating the beam, a polarized beam is selected. For a 100% polarized beam and scattering on a
second 4He nucleus, all neutrons should pass on the lower side. This asymmetry is indeed observed
in realistic cases. This experiment is an example of the polarizer-analyzer set-up

nucleon shell structure with, this time, the correct "magic" numbers N, Z =
2, 8, 20, 28, 50, 82, 126, ....
Besides the correct reproduction of the stable nucleon configurations, there
exists experimental evidence for a spin-orbit term in the nucleon-nucleon inter-
action as observed from nucleon-nucleus scattering (Bohr, Mottelson 1969). In
a first collision of neutrons or protons with He nuclei of energy between 5 and
15 MeV, the beam is partially polarized. Classically, we can present the scattering
in the following way (Fig. 3.12). The potential that an incoming nucleon feels in
the nucleus is U(r) + (r) I· 8 with a negative value of v's in (3.36). Thus the
parallel orientation is favored by the scattering process. The nucleons scattered
to the right now have their spin preferentially in the forward direction; the ones
scattered to the left have their spin in the opposite direction. Using a diaphragm,
a partial polarization of the beam is obtained. By now using a second collision
process, right-left asymmetry is observed because of the partial polarization of
parallel spin-orbital orientation in the incoming beam. Although the above dis-
cussion is somewhat too simplified, the basic outcome is indeed observed and
so unambiguously indicates a term in the nuclear potential proportional to I· s.

3.1.3 Illustrative Examples of Energy Spectra

In the next two figures, we illustrate for both light nuclei (160 region) and heavy
nuclei e08Pb region) both the proton and neutron particle and hole excited states
at relatively low excitation energy (Figs. 3.13,14). Besides some extra states,
the concept of single-particle motion in an average field following the structure
obtained in Sect. 3.1.2 is indeed very well realized. Also, it is clear that in the
heavier 208Pb nucleus the average field is better determined compared to 16 0
where only 16 particles are present.

67
731 3/2+
7.16 5/2+ 6.85 7.12 ----1-
6.79 3/2+ 6.92 - - - - 2 +
P:ii~
===3-
6.33 3/2- 6.16 P3/~
3/2- 613
6.05 0+
5.30 \12+ 5/2(+)
5.28 5/ + 5.24
2 5.18 1/2 +

3.85 ---51z- 3.86 ---5;z-

3.06 ----112- 3.10 - - - I / Z -

087 _5_1.:..:/2,--- V2 +
0.50 _5_1....;/2,--- 1/2 +
P ii~ P I/~ dS/2 d 5/2
---'-':"-- 1/2- ----''-- \Iz- ---0+ --'-''-- 5/2 + --'-''--5jz+
-£..93=12.1 -£..93 = 15.7 £.93=4.14 £.93 = 0.60

I~Na 1: 07 1;0 9 I;Fa

Fig. 3.13. llIustration of both proton and neutron single-particle and single-hole states around 16 0.
On each nucleus, the binding energy difference .dB, relative to 160 is given. The levels where the
major single-particle (single-hole) character is concentrated are given with the quantum numbers (Ij)
although these quantities are not good quantum numbers in general (only J" is). The level scheme
is taken from (Bohr, Mottelson 1969)

h9/~
3.47--- 9/2-

2.71 --(9;z+)
f 7i~
2.34 - - - 7 /2"

113/2
. -I
d Si~
1.63--- 1312+ 1.67--- 5/2+ 1.60 -il311 13; + d 5/2 d 5/2
- - 2 /-_5/2+1.57--5/2+
h i1)2 156 _.-15/2-143-._ 15/2-
1.34--- 11/2- 142/ J 15/2 )15/2

17/2
0.89 - - - 7/z- i 11/2 i 1\12
0.80-- 1112+ 0.78-- 1111+
d j/~
0.35--- 3/2+
-1
P ii~ h912 99/2 99/2
---1/2----- ~1;z+ ---0+ _ - _ 9 /2 - _ - - 9 /2 + - - - 9 /2 +
-693 = 7.38 -693=8.03 93=1636 693=3.60 693=-14.7 693 = 3.94
T =45/2 T = 22 T = 43/2 T =45/2 T= 45/2

2~:Tl126 2~~Pb127
Fig. 3.14. See caption to Fig. 3.13, but now for the proton and neutron single-particle and single-hole
states around the nucleus 208Pb. Details on the origin of the experimental data can be found in the
Nuclear Data Sheets for the appropriate nuclei [Figure taken from (Bohr, Mottelson 1969)]

68
 We show, moreover, the variation of the single-particle states as a function
of nucleon number for the neutron bound states (Fig. 3.15), as well as an excerpt
for 208Pb (Fig. 3.16) around the Z = 82 and N = 126 closed shells. [The calcu-
lated levels result from a Woods-Saxon potential as studied by Blomqvist and
Wahlbom (Blomqvist, Wahlbom 1960).]
We now discuss how the average field U(r), used before in a rather
phenomenological way, can be determined from a microscopic starting point
(Hartree-Fock).

-10

rlh 11 /2
-3s '/2
'-2d l /2
-2d 5/2
>G/ -19 7/2
~
r: -19 9/ 2
ILl
-2P'/2
-2p l l2
-1f 5/2
-1fh

-----2s
f2
-30 _ _ _-_-_-_-_-_--ld l/2
-ld 512

~===::::============== '-lpl/2
-1 P 1f2

- _ _ _......:.:ls~1f2~_ _ _ _ _ _ _ _ _ _ _ lsIf2
-40

o 20 40 60 80 100 120 140 160 180 200 220

A
Fig.3.IS. Energies of neutron orbitals, calculated by CJ. Veje as quoted in (Bohr 1969). Use has
been made of a Woods-5axon potential U = V f(r) + Vi.(l· 8)r~(1/r)(d/dr)f(r), with f(r) having
= =
a Woods-5axon shape [1 +exp(r - Ro/a)]-l, and Ro roAl/3 (ro 1.27 fm) and a 0.67 fm. =
The potentials V and Vi. are given as V = (-51 + 33«N - Z)/A»MeV, Vi. = -O.44V, [taken
from (Bohr, Mottelson 1969)]

69
 0
W ;13/2
ii.
:::>
u
u
0
P/2
----
Z h 9/2
:::>
---------------- -----------

5 1;2

--- ~
d 3/2
0
w
, ii.
:::>
',,- U h11;2
U
0
d S/2 ~
10
1'/2---'/
/
/
""
h~------------

calc. obs. calc. obs.

NEUTRON STATES PROTON STATES
N=126 Z = 82
Fig. 3.16. The empirical values for the binding energy of a single nucleon with respect to 208Pb as
taken from the experimental one-nucleon separation energies (see the quantities .!1B in Fig. 3.14).
The calculated values have been taken from J. Blomqvist and S. Wahlbom (Blomqvist 1960) [Figure
taken from (Bohr, Mottelson 1969)]

3.1.4 Hartree-Fock Methods: A Simple Approach

Suppose nucleons fill up a number of nucleon orbitals CPa(r) such as to form a
density e(r) given in terms of the occupied single-particle states as

e(r) = :E cp;;(r)CPb(r) . (3.37)

bEF

Then the potential at a point r', generated because of the nucleon-nucleon two-
body interaction V(r,r') reads

70
 UH(r') =L
bEF
J CPb(r)V(r,r')CPb(r)dr. (3.38)

We denote by UH(r'), the Hartree tenn neglecting exchange effects, and this tenn
is used in the case of atoms. Within the atomic nucleus, U H(r') is the direct tenn
of the potential affecting the nucleon motion in the nucleus. The more correct,
single-particle SchrOdinger equation for the orbital cp;(r) now becomes

-;2 m
L1cp;(r) + L JCPb(r')V(r, r')CPb(r')dr' . cp;(r)
bEF

- L JCPb(r')V(r, r')CPb(r )cp;(r') dr' = CiCP;(r) . (3.39)

bEF
The second contribution on the left hand side takes into account the antisymmetry
for two identical nucleons, one moving in the orbital CPb(r'), the other in the
orbital CPi(r). The product wave function CPb(r')cp;(r) has to be replaced by
CPb(r')CPi(r) - CPb(r )cp;(r').
The above Hartree-Fock equations [since for every CPi(r) an analogous dif-
ferential equation is obtained, all coupling via the potential tenns] can be written
in shorthand fonn as

;,,2 L1cp;(r) + U H(r )CPi(r) -

- 2m J ' , ,=
U F(r, r )CPi(r ) dr c;CPi(r) , (3.40)

with

U H(r) =L
bEF
J CPb(r)V(r, r')CPb(r') dr' ,
(3.41)
U F(r, r') =L CPb(r')V(r, r')CPb(r) .
bEF
The iterative Hartree-Fock method now starts from an initial guess of the average
field, or of the wave functions, starting from the knowledge of V(r, r') to solve
the coupled equations (3.40) in order to detennine a better value for UH(r) and
UF(r,r'), the cp;(r) and Ci. One can thus proceed in this way until convergence
in the above quantities results. Schematically, one has
(0) (1) (2)
UH(F)(r) UH(F)(r) U H(F)(r)
t /' t /' t /'
cp~O)(r ) cp~l)(r) cp~2)(r ) (3.42)

/0) /1) /2)

• • •
At the end, a final field UH(r), wave function CPi(r), single-particle energy Ci is
obtained. It is now possible to prove that with the wave functions so obtained,
by calculating the energy expectation value

71
 (3.43)

with the !Pal ,a2, ••. , aA (rl, ... , r A) defined in (3.5), the minimal value is obtained.
In the present discussion, one assumes that the original one and two-body
Hamiltonian (3.9) does not contain strong short-range correlations nor density
dependent two-body interactions. In those cases, the variational aspect of Hartree-
Fock theory becomes lost (Ring, Schuck 1980, de-Shalit, Feshbach 1974). De-
tailed discussions on the determination of the Hartree-Fock energy and on vari-
ational approaches to the energy of an interacting many-body system can be
found in (Ring, Schuck 1980, Irvine 1972). Since it is not our aim to devote an
extensive treatment to those aspects of the nuclear A-body system, we refer the
reader to the above references. The situation with two-body density dependent
interactions will be discussed in some detail in Chap. 8.
There now exist many alternative methods in addition to the Hartree-Fock
SchrOdinger equations to derive this condition. We do not go into detail about
these aspects. We only mention that an often used method to determine the "best"
wave functions Cf'i(r) is to expand in a harmonic oscillator basis

Cf'i(r) =I: a~Cf'~·o·(r) , (3.44)

k

after which the coefficients a~ are detennined so that the total energy EHF (3.43)
is minimized.
The Hartree-Fock wave functions (and potential) indeed give a finn basis to
the independent particle shell-model approach to study nuclear excited states. Be-
sides the determination of the single-particle energies that can be compared with
the data and show good overall agreement throughout the nuclear mass table (see
Fig. 3.16 for a comparison in 208Pb), nuclear densities [charge densities eC(r)]
have also been determined and compared in detail. We do this for 160_208Pb
(Fig.3.17a) and compare these nuclei together to nuclear matter density for the
corresponding matter densities em(r) in the same region 160_208Pb (Fig.3.17b).
Calculations have been carried out using Skynne type of effective interactions
(see also Chap. 8) but other interactions give very similar results. Very recently,
using the difference of charge densities as obtained via (e, e') scattering exper-
iments at Saclay for 206Pb_20sTI, (Cavedon et al. 1982, Frois et al. 1983, Doe
1983)
e(206Pb) - eeosTI) = I: lCf'b(r)1 2(206Pb) - I: lCf'b(r)12(20STI)
bEF bEF

=1Cf'38~/2 (r)12 , (3.45)

the shape of the 381/2 orbit could unambiguously be determined (Fig. 3.18). This
gives a sound basis for the independent motion of nucleons as a very good picture
of the nuclear A-body problem. Thus this is a good point to study more detailed
features related to the residual nucleon-nucleon interactions remaining outside the

72
 SkE2-- SkE4----- Exp a)

0.1
Pc
0.1
4°Ca "'ea

5 6 7 , 6 7 , 6 7 ,

90Zr 01 0.1

005

7 , 0

b)

P nuclea,
matte,

9 10

Fig.3.17. (a) Charge densities for the magic nuclei 9OZr• 132Sn and 208Pb. The
160. 4Oea, 48Ca.
theoretical curves correspond to the effective interactions Sk:E2 and SkE4. respectively (see Chap. 8
for a detailed discussion on the extended Skynne forces) and are compared with the data. The units
for f}c are efm- 3 • with the radius r in fm. (b) Combined nuclear matter densities f}m (fm- 3 ) for
the above set of doubly-closed shell nuclei. Nuclear matter density f} F (nuclear matter) is given as
a "measure" for comparison

73
 "'e CHARGE DENSITY DIFFERENCE
~ 206pb _205 T1

THEORY z~ 0.01
ItJ

~ I~
is II~' ___EXPERIMENT
~O.OOE ~
'I
in
z I,
~~o.OO2 "I,
a:
< "~. •............' "
~ 0 ntIfHltt 1ttt "....................

o 2 4 6 8 10
RADIUS (frn)

Fig. 3.18. The nuclear density distribution for the least bound proton in 206Pb. The shell·model pre·
dicts the last (381/2) proton in 206Pb to have a sharp maximum at the centre, as shown at the left·hand
side. On the right·hand side the nuclear charge density difference (}c ~Pb)-{}c eosTI) <f'~ (r) = 8 1/2
is given [taken from (Frois 1983) and Doe 1983)]

average field (Hres). We shall study the two-particle and three-particle systems
(identical nucleons) and also address the proton-neutron systems, incorporating
isospin into the discussion. We also give some attention to the problem of the
effective nucleon-nucleon force acting in the nuclear medium as compared to the
free nucleon-nucleon force. In Chap. 8, we shall discuss a fully self-consistent
version of the shell-model approach and show the state-of-the-art possibilities
using present day high speed computers.

3.2 Two-particle Systems: Identical Nucleons

3.2.1 Two-particle Wavefunctions

The two-particle angular momentum wave function, following the methods of

Chap. 1, can be constructed as

t/J(jI(1)h(2); JM) . (3.46)

(c.p will always be the notation for a single-particle wave function, t/J and I[t for
composite wave functions). In what follows we denote with 1 == TI, 0"1, ••• all
coordinates of particle 1, and it is a notation for all quantum numbers necessary
to specify the single-particle state in a unique way jl == nl, 11, jl.
In describing the full Hamiltonian for a nucleus formed by a closed shell
system described by Ho and two extra identical valence nucleons described by

74
 Fig. 3.19. Splitting of the two-particle states
tf;(jIiz; J M)nas with IjI - jz I $ J $ jI + jz
through the residual nucleon-nucleon interaction
VI,Z. The unperturbed energy for the configura-
tions, E:h + E:j2 (all J) is given at the left-hand
side. The energy splitting is drawn schematically

the single-particle Hamiltonian ho(i) with V(rl, r2) as a two-body interaction,

the wave functions (3.46) are eigenfunctions of
2
H = Ho + L ho(i) , (3.47)
i=l

with energy eigenvalue Eo + c: il + c: j2. The energy shift induced by the residual
interaction now becomes

(3.48)

and will split the degeneracy in J for the (jlh)J multiplet of states (Fig. 3.19).
In the remaining discussion, we always shall discard the core-part Ho and only
consider the energy of the valence nucleons. We also recall that the wave function
(3.46) stands for

~(jl(1)h(2); JM) = L (jlml,h m2IJM)

(3.49)

In the case of identical particles (p - p; n - n) the wave function (3.46) needs

to be made explicitly antisymmetric under the interchange of all coordinates of
the two nucleons, however. We consider separately the cases (i) it f h and (ii)
jl = h. In the discussion below, when expressing and evaluating matrix elements
we shall always use the Dirac bra-ket notation and an index to show which matrix
elements are calculated. We also use the notation
as antisymmetrized wave function,
nas : normalized, anti symmetrized wave function,
no index: non-anti symmetrized wave function.
i) jl f 12- We construct
~as(ith; JM) =N L (itmt,hm2IJM)

(3.50)

75
Because of the explicit anti symmetrization, there is no need anymore to use the
coordinates in the wave function on which we put the index "as". The demand
for a normalized wave function results in N = 1/..fi. Equation (3.50) can be
rewritten as
tPnas(jtjz; JM) = ~ [tP(jti2; JM) - (-I)it+h- J tP (jzjt; JM)]. (3.51)

On the right hand side, we can also leave out coordinates by using the con-
vention of always coupling the quantum state for particle 1 with the quantum
state of particle 2 in that order reading from left to right! (standard convention
for carrying out Racah-Algebra within the nuclear shell-model) (de-Shalit, Talmi
1963).
ii) jt = j2. We now construct, as above,

x [CPjm 1 (l)cpjm2(2) - CPjm2(l)CPjml (2)]

=N' I: [(jmt,jm 2IJ M ) - (jm2,jmtIJM}]CPjml(I)CPjm2(2)

(3.52)
Thus, only even J values are obtained (J = 0, 2, ... , 2j - I) and N' = t.
As an example, we give the possible states for
- the Ids/ 2 Id3 / 2 two-particle states

J = I, 2, 3,4,

- the Ids/2 Ids/2 two-particle states

J =0,2,4.

The two-body matrix elements including the residual interaction can be writ-
ten as (it 1= jz)
LlE(jti2, J) == (jtj2; JMIVi2Ijti2; JM}nas
= (jti2; JMIVi2Ijtj2; JM)
- (_l)jl+h- J (jti2; JMIVi2Ijzjt; JM) , (3.53)

by assuming Vi2 = l'2t [where Vi2 == V(Tt, T2)] and if it = jz, one gets
LlE(i,J) == {i; JMIVi2Ii; JM} . (3.54)

76
Before discussing methods to evaluate the matrix elements (3.53,54) by using
the two-particle wave functions as constructed above, we first discuss in some
detail methods to get a better understanding of the effective interaction V(Tl, T2)
itself.

3.2.2 Two-particle Residual Interaction

The problem of an appropriate choice for the two-nucleon interaction, especially
when nucleons are surrounded by other nucleons (the nuclear medium) is a
difficult one (Brown 1964, Brown, Kuo 1967). It does not have a unique answer
since the "effective" force will be dependent on the particular model space that
one considers in handling low-lying excited states. Still, a number of avenues
have been followed in the past and we will discuss some of the more effective
ones.

a) Effective Two-Body Matrix Elements

Within this approach, no attempt is made to pin down the (radial) shape of
the interaction itself. Rather, the two-body matrix elements, together with the
single-particle energies, are taken as "free" parameters for use in the SchrOdinger
equation to describe the nuclear excited states (Brussaard, Glaudemans 1977).
Thus as parameters for a given model space one has
ci i , Ud2; JMIVi2Ihj4; JM} .
If one takes as an example the full sd space (in order to discuss nuclei between
16 0 and 40Ca), the parameters are the two relative energies c2S 1 / 2 -c1ds/2' c 2S1 / 2 -
c1da / 2 and the 63 two-body interaction matrix elements using as two-particle con-
figurations all two particle states Ijd2; JM} with jl,jz E (2S 1/ 2 , 1d3/ 2 , 1ds/ 2 ).
Since in the SchrOdinger equation for a given nucleus A(Z, N) with n valence
nucleons (protons and neutrons) 0 :::; n :::; 24, the eigenvalues Ero are functions
of the essential parameters given above, •

(3.55)

one can use iterative least-squares methods to get convergence to a final set
of two-body matrix elements and relative single-particle energies (Fig. 3.20).
This method has been in use for a long time, and with the advent of high-
speed computers has been applied to a large variety of nuclei. In particular the
Utrecht group (Brussaard, Glaudemans 1977) and Wildenthal-Brown (Wildenthal
1976,1985) have studied p shell nuclei and, more recently, the full sd shell using
a single set of two-body matrix elements. We illustrate this by some examples
for 27 AI, 28Si, 29Si in the middle of the sd shell for the excited states and for
the full sd shell including binding energies (two-neutron separation energies
(Figs. 3.21,22).
Eventually, of course, one should try to compare and understand the fitted
values with two-body matrix elements determined by other methods.

77
 THE LEAST-8QUARES FITTING PROCEDURE

selec t:
-.-
set up: calculate: JUdge:
Initial Inlerac \Ion f-+-- Hamiltonian I--+-- eigenvalues energy ~stop
parameters matrix elgenvec tors agreement
'--

poor

-.-
obtain: perform: construe t :
new Interaction f---.- least·squares linear

parameters fit to data. equations

Fig. 3.20. lllustration of the various steps necessary in order to deduce an effective interaction (the
two-body matrix elements (jti2; JIVi,2Ihi4; J) and single-particle energies E:j,) along the method
discussed in Sect. 3.2.13. Thereby a fit to experimental excitation energies and/or electromagnetic
properties can be imposed [taken from (Brussaard 1977)]

3-
t=====
====--
6 77
93~
33
79__---=::::::
>:... 1117
::;; 5§ 0__ 7----
;:5 5_ _
L?
0:
7 4----
I~=====
w II 7 ________
Z 5-----
w4
3
Z I
0
i=
<[

tix 3
w ~--
57 _ _ 3__

2_ _ 5 - -_ _
2
3__
3__

0 5 0----
27AI

Fig.3.21. Excited states for Z1 AI, 28 Si and 29Si. The ground-state energies are set equal in the
figure. The left-hand ending of each line contains the calculated number, the right-hand point the
corresponding experimental number. In odd-mass nuclei, the level is characterized by the value 2J.
Known negative parity states have been left out from the data. All other data are given (up to the
upper energy considered here) [taken from (Wildenthal 1985)]

78
 Fig. 3.22. Calculated and measured two-
neutron separation energies S2n along
sd-isotope chains. The lines connect
the theoretical points. The data are in-
dicated by dot, solid or open circles.
The diameters and their placement rel-
ative to the lines indicate the magni-
tude and directions of the differences
S2n(th)-52n(exp) [taken from (Wilden-
thai 1985)]

10 II 12 13 14 15 16 17 18 19 20
N-+

b) Realistic Interactions
In a completely different method, one tries to start from the free nucleon-nucleon
interaction and to incorporate the necessary modifications to obtain the appropri-
ate nuclear two-body interaction matrix elements. Pioneering work in this direc-
tion has been carried out by Brown and Kuo (Brown, Kuo 1967, Kuo, Brown
1966, Kuo et al. 1966).
Here we shall outline briefly how such "realistic" forces are determined,
forces that are obtained by fitting to the free nucleon-nucleon scattering observ-
abIes [phase shifts in all possible reaction channels c(ls):1 (EJab), polarization
data, ...J. We illustrate in Fig.3.23 those basic data in the pp, nn, pn case
(isospin T = 1 and T = 0 channels) for channels denoted by 28+1(1):1 quantum
numbers (1: relative orbital angular momentum, S: relative intrinsic spin, :1: total
relative angular momentum) (see also Fig. 3.24).
In general one describes processes up to EJab ~ 350MeV. Thus the most
important partial waves are 1 = 0, 1, 2 S, P and D-waves (MacGregor et al.
1968a, b, Wright et al. 1967). At higher energies the concept of a non-relativistic
scalar potential loses its precise definition. Some of the most popular potentials

79
 v (1,2) Fig. 3.23. Characterization of the possible
channels in a free nucleon-nucleon scatter-

..
ing process (15):1, T where 1, S,:1 and T
denote the relative angular momentum, the
total intrinsic spin, the total relative angu-
lar momentum and isospin (see Sect. 3.4),
N N respectively. The process can be described
* Spin state S =0,1 (for non-polarized quantities) by the phase
shifts 6(15):1, T as a function of the scat-
'* Charge sta te pp,pn,nn (isospinT) tering energy Elab
* Spatial stale l=even,odd

of this "realistic" type are the hard-core Hamada-Johnston potential (Hamada,

Johnston 1962), the Reid soft-core (Reid 1968) and the Tabakin (Tabakin 1964)
velocity dependent potentials.
One starts from a potential with analytic structure in the radial part that goes
into the one-pion exchange potential (OPEP) at distances Irl - r21 S:' 2fm but
with strengths that have to be fitted so as to describe the data as well as possible.
These general shapes, also dependent on intrinsic spin orientation, charge of the
interacting particles, etc., are dictated by general invariance principles (Ring,
Schuck 1980):
i) under exchange of the coordinates,
ii) translation
iii) Galilean
iv) space reflection
v) time reversal
vi) rotational, in coordinate space
vii) rotational, in charge space.
Thus for local forces not depending on the velocity, the central force is the
most important part:
Vdl,2) =Vo(r) + V u (r)O"I . 0"2 + V r (r)Tl . T2
(3.56)

(We shall discuss isospin where the T operators occur in detail in Sect. 3.4.)
The remaining local part is of tensor character and given by

VT(1,2) = [VTo(r)+VTr(r)Tl 'T2]512

with
3
512 = 2"
r
(0"1 . r) (0"2 . r) - 0"1 . 0"2 . (3.57)

80
 b

100

0.75 T=1 PHASE PARAMETERS

(J1
z
«
is 050
«
a::
J p,
t-
u. 025
::t: '0,
(J1

w MeV
(J1
« 0
400 E LAS
::t:
11.

-025

-0.50

b
tOO

T=O PHASE
PARAMETERS
0.75

(J1 HoV
Z
~ 0.50
0
«
a::
t-
u. 0.25
::t:
(J1

w MeV
(J1
« 0
::t: E LAs
11.

-0.25

-050

Fig. 3.24. The nuclear phase parameters for both the T = =

1 and T 0 channels and this for S, P
and D channels, up to a laboratory energy of Elab = 4OOMeV. The low-energy behavior (insert)
is detennined by the effective-range parameters, the data for E > 24 have been taken from (Arndt
1966). Phase shifts are given in radians [taken from (Bohr, Mouelson 1969)]

81
The important non-local term is the spin-orbit term

(3.58)

and a quadratic spin-orbit can also be added.

In all situations the radial shape is not determined by invariance principles.
The idea of Yukawa that the nucleon-nucleon interaction mainly derives from a
meson field theory (Yukawa 1935) leads to the Yukawa-shape as a fundamental
radial dependence given by the form

(3.59)

where 1/J.L == n/m7rc is the Compton wavelength of the pion.

We show in Fig. 3.25 the Hamada-Johnston potential (Hamada, Johnston
1962) as a good example of "realistic" interactions.
Even though the evaluation of two-body matrix elements of these interac-
tions becomes quite involved, they have been studied in detail (Kuo, Brown
1966, Brown, Kuo 1967). We compare in Fig. 3.26 the Tabakin matrix elements
((1g7/2)2; J MIVTabakinl(197/2)2; J M) with other matrix elements which we dis-
cuss below. One immediately observes that these "realistic" two-body matrix
elements can not be used as such in shell model calculations. Only after bringing
in the nuclear medium effects (a process called renormalization of the force) can
one perform a serious comparison. We refer the reader to Kuo and Brown for
such studies (Kuo, Brown 1966, Brown, Kuo 1967, Kuo 1974). In using realistic
interactions in order to describe the nuclear many-body system, a number of ex-
tra complications arise in comparison to the effective and schematic interactions.
Since most bare nucleon-nucleon forces contain an infinite, or at least a strong
repulsive core, perturbation theory and Hartree-Fock methods do not yield even
a good first approximation (Sect. 3.1.4). The Brueckner method (Brueckner et al.
1955,1958,1960,1962, Brown 1964) consists in replacing the nucleon-nucleon
potential by a G-matrix. Nuclei with few nucleons outside closed shells and
doubly-closed shell nuclei have been treated by using Brueckner-Hartree-Fock
methods (Ring, Schuck 1980, Brussaard, Glaudemans 1977, de-Shalit, Feshbach
1974, Brown 1964). Still, the strong tensor force causes considerable problems in
the development of a fully self-consistent microscopic theory of nuclear structure.
Along the line of the "realistic" forces, in recent years fully field-theoretical
forces have been constructed [the Paris potential (Lacombe 1975,1980), the Bonn
potential (Machleidt 1987)], where the exchange of the 7r-meson but also higher
mesons (w, f!, •..) is involved

c) Schematic Interactions
In contrast to the complicated process of bringing the free nucleon-nucleon force
into the nucleus, simple forces have been used that are immediately useful in
applications into a given mass region. One defines a simple radial shape leading
to a numerically rather simple calculation and determines the strength(s) so that

82
MeV Vc 1 MeV VLL
100 : 100
I
I
I
I

1\1~ \
- - HAMADA JOHNSTON - - HAMADA JOHNSTON
- - - OPEP

1 \
singlet odd \
I ,
I ,
1

J rlfm)
dfm)

-100 -100

MeV v"VLS : t
I I
1 I SPIN DEPENDENT POTENTIALS
I I
1 I
I I - - HAMADA JOfiNSTON
100 I I - - - OPEP
1 I
I I
I I
I 1
I , v, triplet odd
I (
I I
I \ /Vy triplet odd
I1 / / ',
\
\ "vLS triplet evell

....
'"
"'
',(

I Yy triplet even
Fig. 3.25. The ''realistic'' nucleon-nucleon po-
,"',
I
tentials as obtained from the analysis of T.
,
,,,
I v, triplet even Hamada and 1.0. Johnston (Hamada 1962) are
illustrated for both the central spin-orbit, tensor
and quadratic spin-orbit parts [see (3.56-58)].
I The dotted potentials (OPEP) correspond to the
I
-100 I one-pion exchange potential. We give on p.84
I

,,
I a box with the details of the Hamada-Johnston
potential [taken from (Bohr, Mottelson 1969)]

83
The Hamada-Johnston Potential

v = Vc(r) + VT(r)SI2 + VLs(r)1 . 5 + VLL(r)LI2 ,

with
3
S12 == 2" (0"1 . r) (0"2 . r) - 0"1 . 0"2 ,
r
L12 == (0"1 .0"2)12 - &[(0"1·1) (0"2.1) + (0"2 .1) (0"1 .1)] ,
== (bl,J + 0"1 .0"2)12 - (I . 5)2 .
The radial functions are, at large distances, restricted by the condition
of approaching the OPEP.

Vc(r)=VO(Tl·T2) (Ul·U2)Y(x)[I+acY (x)+bc y2 (x)] ,

VT(r)=VO(Tl·T2) (Ul·U2)Z(x)[I+aTY(x)+br y2 (x)] ,
VLS(r) = gLSvoy2(X) [1 + hsY(x)] ,
Z(x) [ 2 ]
VLL(r) = gLLvO-2- 1 + aLLY(x) + bLLY (x) ,
x

VO
1 P
= 3 ncm7rc = 3.65 MeV ,
2

x = (m 7r c)/n. r = r/1.43fm,
1
Y(x) = - exp( -x) ,
x

Z(x) = (1 +~ +~) .
x x2
Y(x) .

In addition, an infinite repulsion at the radius c = 0.49 fm (xc = 0.343),

is assumed.

The optimum, adjusted param- The values of the different potentials

eters are given in the table. at the hard core r = c have been deter-
Singlet Triplet Singlet Triplet mined as
even even odd odd Vc VT V LS V LL
ae 8.7 6.0 -8.0 -9.07 Singlet, even -1460 -42
be 10.6 -1.0 12.0 3.38 Triplet,even -207 -642 34 668
aT -05 -1.29 Singlet, odd 2371 -6683
bT 0.2 055 Triplet, odd -23 173 -1570 -1087
gLS 2.77 7.36
bLs -0.1 -7.1
gIJL -0.033 0.1 -0.1 -0.033
aLL 0.2 1.8 2.0 -7.3
bLL -0.2 -0.4 6.0 6.9

84
 00
k\' ' ,,
Jlt
t' ---- ,,
\, , , 6'

\
--- \
4'

\
I \
> I \ 2'
~ I
0-0.5 I
a:: \
w
z
w \
\
\
"-
"-
"-
"-
-10 "-
.. , O·
Tabakln SDl GaussIan
lal Ibl

Fig. 3.26. Comparison of two-body matrix elements for the (1 fI1 /z)z J configuration according to a
number of different interactions. We compare the surface-delta interaction (SDI) (see Appendix D),
the Gaussian interaction and the realistic, velocity-dependent Tabakin interaction (Tabakin 1964). (i)
For the Tabakin force, we give the total, bare matrix elements (a) and the Tabakin matrix elements
corresponding to the 1So channel only (b). (ii) The Gaussian interaction of Fig. 3.27, with the pro-
jection operators of (3.64) Le. V(r) = Vo(Ps + tPT), (with Vo = -35MeV and t = +0.2). (iii)
A SOl interaction, such that for the force Vsm(r) = -411" 48(rl - rz)8(rl - Ro), the product
AT = AT' C(Ro) = 0.25 MeV with C(Ro) == Jt.1g7/2 (Ro) ''1) IlZ

the nuclear structure can be well described. Eventually, the invariance principles
discussed above are also incorporated to suit certain specific purposes.
Often used potentials for VO, 2) are the Yukawa potential, e-}Lr / /-lr, square-
well potential, the Gaussian potential, e- JLr2 and 6 or surface-6 (SOl) interactions
(Brussaard, Glaudemans 1977): 6(r) or 6(r)6(n - Ro).
We give a combined illustration of various potentials in Fig.3.27. In the
brief Appendix. D, we point out that the SOl interaction is actually much better
founded than one would think and relates to properties of the nucleon-nucleon
free scattering process.
The schematic interactions all illustrate the short-range aspects of the nucleon-
nucleon interaction in the nucleus, with the 6-function form as an extreme case.
The latter form is particularly interesting to study the main features of the nuclear
two-body interaction matrix since one can work out most results in analytic form
when using the single-particle wave functions constructed in Sect 3.1.
In addition specific combinations of the intrinsic spin and charge properties
often occur, e.g.,
P(f == HI + 0'1 . 0'2) . (3.60)

85
 VARIOUS POTENTIAL SHAPES

>0 -02
.., ,
J

'0 - - - square well

-04
....
III
. eXpOnential
c: - - --Gaussian
:J -06
~ -Yukawa
r:.
g -0 ........ Saxon-Woods

t -12~--~--~----J---~~--~--~
o 2 3
~ r In units of Ro

Fig.3.27. Plots of the potentials, given below. In addition, the Woods-Saxon potential with a dif-
fuseness parameter a =0.7 is also shown [taken from (Brussaard, Glaudemans 1977)]
Square well V(r) = - Yo r $ &, ,
=0 r>&"
Exponential potential V(r) = - Yo exp ( - r / &,) ,
Gaussian potential V(r) = - Yo exp( _r2 /~) ,hlill
Yukawapotential V(r) = -Yoexp(-r/&,}/(r/&,).

This form is called the Bartlett term and is interesting for the following reason.
If we add the intrinsic spins of the interacting nucleons
8 = Sl + s2 , (3.61)
we have
tTl . tT2 = 2[8 sr - S~]
2 - , (3.62)
or for expectation value for an S = 0 or S = 1 state
(tTl· tT2}S = 2S(S + 1) - 3 . (3.63)
Thus, the operator PO" is an exchange operator since

PO"'ljJS=l = 'ljJS=l ,
PO"'ljJs=o = -'ljJs=o ,
and the two spin coordinates appear in a symmetric way in the S = 1 (parallel)
state and in an antisymmetric way in the S = 0 (antiparallel) state. From the
PO" 's, projection operators for the S = 1 (triplet) and S = 0 (singlet) two-nucleon
state can be defined:

Ps== IIs=o = HI - PO") ,

(3.64)
PT == IIS=l = HI + PO") .

86
 Similarly, an operator can be defined in charge space for the isospin operators,
giving the Heisenberg term p .. = 1(1
+ Tl . T2). Moreover, the Majorana term
which interchanges the spatial coordinates rl - r2 ---+ -(rl - r2), induced by
the operator p .. can be introduced.

3.2.3 Calculation of Two-Body Matrix Elements

a) Central Interactions

Starting from a general central interaction V(lrl - r2i) one can expand it within
a complete set of functions (e.g., with the Legendre polynomials). Thus, we can
express the interaction as
00

V(lrl - r21) = I>kh,r2)Pk(cosll12) , (3.65)

k=<l

with

(3.66)

From the normalized, antisymmetric two-particle wave functions constructed in

Sect. 3.2.1, we obtain for the diagonal matrix element
(ith; JMlv'i2li}j2; JM}nas = (jlh; JMlv'i2lilh; JM}dir
- (_I)it+h- J (j}j2; JMlv'i2lhil; JM}exch , (3.67)

with a direct and an exchange term. We carry out the calculation in some detail,
for both the direct and the exchange term:
The direct term can be written

(jli2; JMIVi2Iili2; JM}dir = Lik pk , (3.68)

k
with

(3.69)

and
pk = pk(nlit,n2h)
= J IUntlt(rl)Un212(r2)12vk(rl,r2) dq dr2. (3.70)

The expression for ik can be evaluated using the reduction rule I from Chap. 2,
and yields

ik=47r/(2k+l)'(-I)i2+it+J{~t
)2
i2
it k
J}
x (it IIY kllit}(hIIY kllh) . (3.71)

87
For the exchange matrix element we get, analogously
00

(ith;JMll'12lhit;JM}exch = LgkG k , (3.72)

k=O
with
gk = 471" /(2k + 1) . (_l)I+J { ~l ~2 Jk }
JI J2
x (jIIlYkllh}(hIlYklljl) , (3.73)
and

G k = G k (nlll' n2h) == J unlit {rt)U n2 12 (r2)U nl l l (r2)

X U n2 h (rl)Vk (rl' r2) drl dr2 . (3.74)
Thus the total matrix element becomes for
i) jl:f h

l1Eith ,J = L fkFk - (_l)il+h- J L 9k Gk , (3.75)

k k
and
ii) it =h
l1Ep,J = LfkFk . (3.76)
k

For a general central interaction when the vk(rl, r2) are determined, in most cases
one has to evaluate the Slater integrals G k and Fk numerically. If one chooses
a simpler schematic force like the 6(rl - r2) or the SOl 6(rl - r2)6(rl - Ro)
force, the integrals simplify significantly.
In Appendix F, we bring together the necessary Racah algebra expressions
needed to carry out the calculation of the (j II Y k IIj') reduced matrix elements as
well as the sums in (3.75,76). We derive in Appendix E the multipole expansion
for a zero-range 6(rl - r2) interaction.
The result is

6(rl-r2) = L6(rl-r2)/(rlr2) .(2k+l)/471".Pk(cosBI2), (3.77)

k

and so we easily get the vk(rl, r2) coefficient as

(3.78)

Thus, the above radial integrals Fk and Gk reduce to

88
 k 2k + 1 (X) 1 [ ]2 ......0
F = 4;- Jo r2 Unll l (r)Unzlz(r) dr = (2k + l)r ,

G k = 2~; 1L oo
r12 [Unlh(r)Unzlz(r)f dr =(2k+ 1)F' .
(3.79)

Furthennore, we use the explicit expression for the reduced matrix element
(jilY kllj), which becomes (Appendix F) after using the reduction rule I of
Chap. 2
((1j21)jIlY kll(1j21')j') =(-I)i- I / 23]'kj..a;

x (!! ~ {) ~(1 + (_I)'+I'+k) . (3.80)

Combining the above result with (3.71) we get the explicit fonn of the fk tenn
fk =(_I)2(it+iz)+J-I (2jl + 1)(2h + IH(I + (_1)k)
x {~l ~2
J2 JI
J}
k
(j\-2 k
0
~l)
2
(j1 -2
k
0
h)
! ' (3.81)

and the total direct contribution reads (3.79)

Lfk pk =F'(2jl + 1)(2h + 1) L(-I)J-I ~(1 + (_1)k)

k k

x (2k + 1) {~l
J2
~2
JI
J}
k
(j\
-2
k
0
~l) (hI
2 -2
k
0
~2)
2
.
(3.82)
Using the expressions of Appendix F where

k jc) ( jb k J2~d)
o 1
2 -! 0

=(_I)ib+id+ J ( ~! Ja
-2
I f) (t Jd
I
2 -=1) , (3.83)

L(2k + 1)(-I)k { ~a
k Jd
jb
Jc
J}
k
(jj
-2
k
0 t) (~! k
0 t)
= (_I)J- 2id (t Ja
-2
I ~) (t Jd
-2
I ~) , (3.84)

we get the direct tenn

(ith; J MIVi21jlh; J M}dir = F' (2jl + 1) (2h + I)!

x [(i, ~ i)' + (~ t !\)'] . (3.85)

89
In a completely analogous way one can calculate the exchange tenn on which
we give some intennediate results, i.e.,

9k=(_1)J+it+i2{~1 ~2 J} (2jl+1)(2h+l)
Jl J2 k

x (~\ ~ t) (~\ ~ ~) ~(I+(_1)It+12+k), (3.86)

with the exchange matrix element

LgkGk = Fl(2it + 1)(2h + 1)(_l)J+it+i2
k

x L!(I+(_1) ' t+ 12+ k) { ~1 ~2 Jk }

k
Jl J2

X(~\ ~ jt) (~2! k

o !
jl) (2k + 1) . (3.87)

Again using (3.83,84), the exchange matrix element finally becomes

(jlh; JMIVi2Ihjl; JM)exch = pO (2jl + 1)(2j2 + 1)

~Yl· (3.88)
The sum gives now when jl t= h,

i1Ejti2 ,J =Fl(2jl + 1)(2j2 + 1) (j!

x (1 + (_1) ' t+12+ J )/2 , (3.89)

o 2
----6+
---[.+

----2+

Fig. 3.28. lllustration of the two-

particle matrix elements £lEj2,J and this
for the j = ~ orbital. According to (3.90),
the matrix elements are scaled with the
corresponding Slater integrals 4pO. We
also construct (the right-hand side) the
_ _ _ 0+
£lEj2,J vs. J plot

90
Table 3.1. The two-body matrix elements of (3.90) and this for a j =
~ particle. The matrix elements
are expressed in units 4pO where pO denotes the corresponding Slater integral for the j ~ particle =
using a c5-residual iIi.teraction

J (} -2
j
I ~) ilEj 2,J/4Fl

0 -VI
2 J3.~.S ft =0.238
4 -J S111 ¥r =0.117
6 J 3.Sft.13 :9 = 0.058

and in the case jt = il,

FJ + 1)2
L1Ep ,J = T(2j (J_~' J
t
J)2 .
o (3.90)
-2
As an example, we show this in Fig. 3.28 for a j = 7/2 particle, combining to a
r = O+, 2+, 4+ and 6+ state (Table 3.1).
b) Examples

In Chap. 9 we provide a computer program that calculates the two-body matrix

elements in order to evaluate the a-function matrix elements. The Slater integrals
for the a-interaction FJ are calculated fully analytically. They are given together
with the code.

TilE TWO·PARTICLE MODEL FOR 180 p n

Fig, 3.29. A comparison between the-

ory and experiIi.Ient for both the biIi.d-
ing energy and the excited states in
ExlMeV) JTt
ISO (a) the spectrum for the neutron
355 . '
379

323
4' 360 4' (ld S/ 2)2 configuration only and (b)
2' the spectrum for two neutrons in the
full (lds/2 281/2 1d3/2) space. In both
241 2' cases, a MSDI interaction was used
198 2' to calculate the spectra. The Modified
SOl (MSDI) differs from the regu-
.1387Mev o • .1402MeVo • lar SOl force in the addition of terms
lal lb) VMSDI = Vsm + BTl' T2. For ob-
taining the spectra in (a) and (b), the
expenment theory
values AT=I = B = 25MeV/A were
used [taken from (Brussaard, Glaude-
mans 1977)]

91
 INFLUENCE OF THE CONFIGURATION SPACE
proton protons protons main
confIguratIon

3
S2
= 4 = 1 11312

77777777
2f712
1h9/2
=;+
77777777 J'
=H=
77777777

/-,,--.--------n--
-- --0'-

or' -~::::~~!,~~t)3~;~!~,:;~~ "3r'

-1r---

'--(2')--- 6'
__ - - - S ' - - - - - , - - - - - - S ' -
-e'---
2

ExlMeV) J'

161
1312·

~
-~:-~:~:~-~~~ ;: ;;~~-:':::-~:-
_,--2'-
S
>(
w _ 2 ' - - - - - - - .. - 2'---

!
,
090 712'

o ----9/2 -0'---------0'----------0'-

experiment theory (a) experiment theory (b)

210 p
84 °'26
Fig. 3.30. The effect of the possible configurations on the spectrum and, more in particular, on the
MSDI parameter AT=l is illusttated for 2~POl26 (see also Sect.3,2.4 for a detailed discussion on
configuration mixing), In case (a) only the configurations (lh 9/ 2)2, (lh9/2 2/7 /2) and (lh9/2 I i l3 / 2 )
are considered with a value of AT=l = 0,33 MeV, In case (b), both protons can be in the full space
(lh9/22/7/21i13/2) with now an optimum value of AT=l = O.l7MeV [taken from (Brussaard,
Glaudemans 1977)]

We compare the two-body matrix elements for several nuclei:

i) (Fig. 3.29): The (lds/ 2 )2 J matrix elements are compared with the exper-
18 0
imental spectrum. The agreement is much improved when we also take more
configurations into account (281/2. Id3 / 2 ) as well as the interactions among
the different configurations. This we shall discuss in detail in Sect. 3.2.4.
ii) 210pO (Fig. 3.30): Here we compare the effect of the possible configurations

92
 3.0 Experiment

16 5 ~ /,...---,4+
06 6.0 ,_'.--/

1
2.0

---
016
26 ,'-'-;.0='---'
/~,"""----

--" -----" ~---",--- ... - - - '

-----
r - - 2+
,-

.
;; 1.0
~

>-
Cl
a:
~ 0.0 - - - - - - - - - - - - - - 0 +
zWo 134Te 136Xe 138 80 140Ce 142Nd 144Sm 146Gd
52 54 56 58 60 62 64

S
~ 3.0 Theory
W

2.0

0.0 ------------0+

Fig. 3.31. We illustrate, for the N =

82 single closed-shell nuclei the energy spectra in even-even
nuclei S2Te, S4Xe, S6Ba, SgCe, 6ONd, 62Sm and 64Gd. We show the lowest 2+, 4+, 6+ levels that are
mainly formed from proton 2 quasi-particle (2qp) excitations (see Chap. 7 for a discussion on the the-
ory of pairing among n identical particles), The lifetime (TI/2 ) for the 6+ level is indicated in units J,LS

on the description of the energy spectrum and the effect of the model space
on the strength of the two-body MSDI interaction [Modified Surface Delta
Interaction (Brussaard, Glaudemans 1977)]. In one case. we only take the
(lh<;/2)2J, (lh9f22h/2)J and (lh 9/ 2 1i13/2)J configurations into account.
whereas in the other case. all two-body configurations within the full (1 h9 / 2 •
2h/2. 1i 13 / 2) space are taken into account.
iii) We also make a comparison for the N = 82 single-closed shell nuclei 134 Te_
146Gd (Fig. 3.31) where a two-quasi particle calculation was carried out. Full
details of the quasi-particle excitations and the pairing degree of freedom will
be discussed in Chap. 7. However. here one can clearly observe the typical
features in the 0+-2+-4+-6+ separation that show the short range correlations
in the nucleon-nucleon residual interaction.

c) Semi-Classical Interpretation
In comparing two-body matrix elements for many different mass regions. it is
interesting to compare the matrix elements relative to the average matrix element
(Fig. 3.32)

93
 IDENTICAL ORBIT SPECTRA 11 =1 2 = LI/2DELTA FORCE SPECTRUM
I> 0 T =1
'2:
-
:;
-I

-2

-3

T=O
·4

-s 120' 50' 0'

180' 150' 120' 90' 60' 30' O' 180'
e l2 0'2

Fig. 3.32. The relative matrix elements (i;JMIVi,2Ii;JM}/V [with V defined via (3.91) as a
function of the overlap angle 812 [see (3.94)]. Experimental values from (Anantaraman 1971) and
for a pure 6-force spectrum with j = ~ are given [taken from (Ring 1980)]

v = ~)2J + l)(Ud; JMIVt2IUd; JM}/ ~)2J + 1). (3.91)

J J
It is also possible to plot the matrix elements

(Ud; JMIV12IUd; JM}/V , (3.92)

versus the angle between the classical orbits for the two valence nucleons B12,
which is defined as
. . J2 ·2 ·2
31'12 -31-12
(3.93)
cos B12 = li1lii21 = 21i111i21 .
This results in

(cos B)
12 = J(J + 1) - j101 + 1) - 1202 + 1)
. (3.94)
2Jj101 + 1)h02 + 1)
In such a way one measures the overlap of the orbitals. For j1 s:' j2 (large
values) one obtains approximately 812 s:' 0 0 for J = j1 + j2 and 812 s:' 1800 for
J =0 (Fig. 3.33). The empirical results for a number of configurations (199/2)2,
(1 h /2)2, (2d3/2i, (299/2)2 do follow a single universal curve with the largest
attractive matrix element at Bl2 s:' 1800 (antiparallel spins) (Fig. 3.32). It indeed
follows that a a-interaction force, due to its short-range attractive characteristics
explains this experimental feature rather well.

94
 Fig. 3.33. The classical orbits for
two particles coupled in the par-
allel (012 = 0°) and antiparallel
(812 = 180°) way. The motion
is, in a schematic way, presented
by the arrow within the plane,
perpendicular to the angular mo-
mentum vectors

d) Moshinsky Transformation Method

In many cases of a central residual interaction depending only on the relative
nucleon separation r = Irl - r21, a method exists to separate the relative from
the center-of-mass coordinates when calculating the two-body matrix elements.
When using harmonic oscillator radial wave functions, this method is extremely
interesting.
IT we consider harmonic oscillator potentials for the interacting nucleons,
using the transformation
r=rl-r2, R=Hr l+r2) , (3.95)
one can rewrite

(3.96)
as

~(; ) w2r2 + ~(2m)w2 R2 . (3.97)

Similarly, by transforming the momenta to a relative (P) and center-of-mass total

(P) momentum, one can also rewrite the kinetic energy part such that

2~ (p~+p~) , (3.98)

becomes

(3.99)

Thus, the total Hamiltonian describing the unperturbed motion of two nucleons
in a harmonic oscillator potential becomes
H = Hho(l) + Hho(2) = Hre1 + Hcorn
1 p2
= p2jm+ 4mw2r2+ 4m +mw2R2, (3.100)

95
indicating that in the relative harmonic oscillator a mass of m /2 appears and for
the center-of-mass motion a mass of 2m.
The wave function describing the unperturbed motion of the two nucleons
can be written in Dirac ket notation

Inl/1ml)ln2/2m2) , (3.101)

with
(rlnlm) = R nl(r)Y;m(8, cp) , (3.102)

and

r = N n,l· r 1e -vr2L'+l/2(2
R nl () n-l vr2) . (3.103)
Here v = mw/21i and n = k + 1.
Since the total Hamiltonian is also equal to the sum of a relative (m/2) and
center-of-mass (2m) oscillator potential, the total two-particle wave function can
also be written as
(3.104)

where (n, I, m) and (N, A, MA) describe the quantum numbers of the relative and
center-of-mass wave function. It is possible to transfonn between the two equiv-
alent sets of basis functions (3.101, 104) by using the Moshinsky transfonnation
(Brody, Moshinsky 1960). One writes
Inlh, n2h; LM)
= L (nl,NA;Ll nl /l,n2 12;L)lnl,NA;LM). (3.105)
n,l,N,A
One can easily prove that the transfonnation coefficients are independent of the
projection quantum number M. During the transfonnation, a number of conser-
vation laws hold:

(i) = L =I + A ,
11 + 12 (3.106)
(ii) Enlll + En2l2 = Enl + ENA , (3.107)
or

2nl + II + 2n2 + h = 2n + I + 2N + A . (3.108)

We illustrate some of these transfonnation coefficients in Table 3.2. Applying
this Moshinsky transfonnation to the calculation of the two-body matrix elements
with a central force V(r), we can evaluate
M == (nl/1,n2h;LMIV(r)ln~/J,n2/2;LM)
= L
n,I,N,A
(nl /1,n2h;Llnl,NA;L)
nl,ll,N',A'

96
Table 3.2. D1ustration of transformation brackets from the product harmonic oscillator wave functions
to a product basis of relative times centte-of-mass (c.o.m.) oscillator wave functions. The table lists
= =
coefficients (nl/l' nzlz; Llnl, N A; L) [taken from (Brody 19(0)] with nl nz O. Note that in
the tables of Brody and Moshinsky. the total orbital angular momentum is denoted by A, the c.o.m.
=
angular momentum by L and that the radial quantum number n takes the values n 0, 1,2, ...

II Iz A n N L fl (I) T

0 0 0 0 0 0 0 1.00000000 1
0 0 0 0 1 0.70710678
0 1 0 0 -0.70710677 2
0 2 2 0 0 0 2 2 0.49999999
0 1 0 1 -0.70710679
0 2 0 0 0.49999999 3
0 3 3 0 0 0 3 3 0.35355340
0 1 0 2 -0.61237245
0 2 0 1 0.61237245
0 3 0 0 -0.35355340 4
0 4 4 0 0 0 4 4 0.24999999
0 1 0 3 -0.50000000
0 2 0 2 0.61237245
0 3 0 1 -0.50000000
0 4 0 0 0.24999999 5
0 5 5 0 0 0 5 5 0.17677670
0 1 0 4 -0.39528471
0 2 0 3 0.55901700
0 3 0 2 -0.55901699
0 4 0 1 0.39528471
0 5 0 0 -0.17677670 6
0 6 6 0 0 0 6 6 0.12500000
0 1 0 5 -0.30618623
0 2 0 4 0.48412292
0 3 0 3 -0.55901699
0 4 0 2 0.48412292
0 5 0 1 -0.30618622
0 6 0 0 0.12500000
0 0 0 1 0 2 0.70710679
0 1 0 1 0.00000000
1 0 0 0 -0.70710679 3
1 0 1 0 1 2 0.99999999
2 0 0 0 2 2 0.70710678
0 1 0 1 0.00000000
0 2 0 0 -0.70710678 3
2 0 0 1 1 3 0.40824829
0 0 2 0.23570226
0 1 1 0 -0.52704628
0 2 0 0.23570226
1 0 0 I -0.52704628
1 0 0 0.40824829 6
2 2 0 1 0 2 3 0.70710676
0 2 0 -0.70710676 2

97
/1 /2 ..\ n N L e (I) T

2 3 0 0 0 3 3 0.61237245
0 1 0 2 -0.35355340
0 2 0 1 -0.35355340
0 3 0 0 0.61237245 4
3 2 0 0 1 2 4 0.27386128
0 0 3 0.20000000
0 1 1 -0.45825757
0 2 0 2 0.00000000
0 2 1 0 0.41833000
0 3 0 -0.20000000
0 0 2 -0.41833000
1 0 1 0.45825757
1 2 0 0 -0.27386128 9
3 3 0 1 0 3 4 0.50000000
0 2 0 2 -0.70710677
0 3 0 0.50000000 3
3 4 0 0 0 4 4 0.50000001
0 1 0 3 -0.49999999
0 2 0 2 -0.0000000 1
0 3 0 1 0.49999999
0 4 0 0 -0.50000001 5
4 3 0 0 1 3 5 0.18898224
0 0 4 0.15152288
0 1 2 -0.37115375
0 2 0 3 -0.05976143
0 2 1 1 0.43915504
0 3 0 2 -0.05976143
0 3 1 0 -0.32732684
0 4 0 1 0.15152288
0 0 3 -0.32732684
1 0 2 0.43915504
2 0 1 -0.37115375
1 3 0 0 0.18898224 12
4 4 0 1 0 4 5 0.35355339
0 2 0 3 -0.61237243
0 3 0 2 0.61237243
0 4 0 -0.35355339 4
4 5 0 0 0 5 5 0.39528471
0 1 0 4 -0.53033006
0 2 0 3 0.25000000
0 3 0 2 0.24999999
0 4 0 1 -0.53033005
0 5 0 0 0.39528471 6
5 4 0 0 1 4 6 0.l3176157
0 0 5 0.11111111
0 1 3 -0.29l33579
0 2 0 4 -0.07273929
0 2 1 2 0.40458680
0 3 0 3 0.00000000

98
11 12 >. n I N L I} (I) T

0 3 1 -0.39086798
0 4 0 2 0.07273929
0 4 1 0 0.25230420
0 5 0 1 -0.11111111
1 0 0 4 -0.25230420
1 1 0 3 0.39086798
1 2 0 2 -0.40458680
1 3 0 1 0.29133579
1 4 0 0 -0.13176157 15
5 5 0 1 0 5 6 0.25000000
0 2 0 4 -0.50000000
0 3 0 3 0.61237245
0 4 0 2 -0.50000001
0 5 0 0.25000000 5

X {n~1i,n212;Lln'1',N'A';L}
{nl,NA;LMIV(r)ln'l',N'A';LM} . (3.109)

In the radial integral part at the right hand side of (3.109), the center-of-mass
part of the wave function drops out; we get orthogonality conditions ONN,OAA'
and an extra factor Oil', since in the relative wave function Inlm}, the angular
part described by the Y;m(9, <.p) spherical harmonics, also drops out (to show this
part of the calculation explicitly is left as an exercise). One then obtains

M = L {nl 1l,n212; Llnl,N A; L}

n,n',I,N,A

X {n~lLn2l2;Lln'l,NA;L}
X {n1IV(r)ln'1} . (3.110)
Here, the radial integral {nlIV(r)ln'l} can be put explicitly, in coordinate space,
as the integral

R == J R n, (r)V(r)R n ,,(r)r2 dr . (3.111)

Using the explicit form of the Laguerre polynomials

n-l
L~~{2(x)= L ak(-1)k x k, (3.112)
k=O
and a Gaussian type of residual interaction
V(r)=Voe- r 2/ ro,
2
(3.113)

the radial integral can be evaluated as the double sum (note that v ~ Vre! = v /2)

99
 L
(n-l)(n'-l)

R =Nn,INn',1 (_I)k+k' akak'

k,k'=O

X J 2 k+k' 2
r 21 e- vr 2 e- r 2/ ro2 (vr) r dr. (3.114)

We define the dimensionless coordinate x == v 1 / 2 r, and obtain

R = N n,1 N,n ,I ' " a p
~
p
J x2p e- x2 e- x2 /(vr~)x2 dx , (3.115)

where 1 ::; p::; 1+ n + n' - 2, ap = akak,(-I)k+k' (v)-(/+3/2).

Combining all intennediate results, the matrix element M can be written in
compact fonn
M= L (nl 1l,n2h;Llnl,NA;L)(nPLnili;Lln'l,NA;L)
n,n',I,
N,.1

X Nn,INn',1 LIp a{ r(p+~) , (3.116)

p

and where the Talmi integrals defined as Ip

I
p
=
r(p+ D
2 J x 2(p+l) e-(1+>.2)x 2 dx
' (3.117)

with -\ = (v1/ 2ro)-1, characterize the interaction. This Talmi integral can be
evaluated in closed fonn and gives
Ip = (1 + -\2) -p-3/2 . (3.118)
In Fig. 3.34 we plot the Talmi integrals Ip for the Gaussian interaction. This
method is very general, and Talmi integrals for, e.g., the Yukawa shape e- pr / {Lr,

! 5
Vo = 10 MeV

oS-t.
If)
-' Fig. 3.34. The nuclear Talmi integrals [see (3.117)
i!iC> 3 and (3.118)] for a residual interaction with a Gaus-
ill sian shape. The abcissa is indicated by 1/ A where
I-
Z 2 =
A (vlj2ro)-1 with v the harmonic oscillator pa-
rameter (v = mw/21i.) and ro the Gaussian shape
~
-'
<l: =
parameter of VCr) Vo exp( _r2 /r~). The values
1-1 h, ... , I13 are shown (in descending order). The
strength of the Gaussian interaction Vo = 10 MeV
o~~~~~~~~ is taken
0.0 05 1.0 15
--Y'A-

100
 the Coulomb shape 1/r, or any other force depending on r = ITI - T21 can be
obtained and evaluated (Talmi 1952).
In the above discussion we concentrated on the orbital part of the wave func-
tion. In actual situations, we calculate the two-body matrix elements
(ith; JMlv'i2lj3j4; JM) where jl, ... , j4 are notations for a full single-particle
wave function, containing besides the orbital part also the intrinsic spin part in
a coupled form (l~)j. In the problem set there is an exercise to go from the
above two-body matrix elements into the relative integrals via the Moshinsky
transformation brackets.

3.2.4 Configuration Mixing: Model Space and Model Interaction

In many cases when we consider nuclei with just two valence nucleons outside
closed shells, it is not possible to single out one orbital j. Usually a number of
valence shells are present in which the two nucleons can, in principle, move.
Let us consider the case of 18 0 with two neutrons outside the 16 0 core. In
the most simple approach the two neutrons move in the energetically most fa-
vored orbital, i.e., in the 2ds/ 2 orbital (Fig. 3.35). Thus we can only form the
(2ds /2)20+, 2+, 4+ configurations and then determine the strength of the residual
interaction v'il so that the theoretical 0+-2+-4+ spacing reproduces the experi-
mental spacing as well as possible (Fig. 3.29). The next step is to consider the
full sd model space with many more configurations for each J1r value. For the
J1r =0+ state, we have three configurations, i.e., the (2ds/2)20+, (3S 1/ 2)20+ and
the (2d3 / 2)20+ configurations. In the latter situation, the strength of the resid-
ual interaction Vi2 will be different from the first model space where only the

6r-----------------------,
E

Is 508

_4
><II
x
~3
0:
UJ
Z
UJ 2
t5
1=
t'!
(31 0.87_ _ _ _ _ __
x
UJ

0.00-------

Fig.3.35. The neutron single-panicle energies in

I~09 (relative to the Ids/ 2 orbital) for 281/2 and
Id3 / 2 orbitals. The energies are taken from the
experimental spectrum in 17 0

101
(2ds /2)20+ state was considered. Thus one generally concludes that the strength
of the residual interaction depends on the model space chosen, that is, Vi2 = Vi2
(model space) such that the larger the model space is, the smaller Vi2 will be-
come in order to get a similar overall agreement In the larger model spaces,
one will, in general, be able to describe the observed properties in the nucleus
better than with the smaller model spaces. This argument only relates to effective
forces using a given form, i.e., a Gaussian interaction, an (M)SDI interaction,
etc. for which only the strength parameter determines the overall magnitude of
the two-body matrix elements in a given finite dimensional model space. Thus
one should not extrapolate to the full (infinite dimensional initial) configuration
space in which the bare nucleon-nucleon force would be acting.
Thus, in the case of 180 where the 3S 1/ 2 and 2ds/ 2 orbitals separate from the
higher-lying 2d3 / 2 orbital, for the model spaces one has

r = 0+ -+ (1dS/2)~+' (3S1/2)~+
r = 2+ -+ (lds/2)~+' (ldS/22s1/2)2+
r = 3+ -+ (ld s/2 2s 1/2)3+
r = 4+ -+ (lds/2)!+ .
The energy eigenvalues for the J7r = 0+ states, for example, will be the corre-
sponding eigenvalues for the eigenstates to the Hamiltonian
H = Ho+Hres,
2
(3.119)
= Lho(i)+ Vi2,
i=l

where the core energy corresponding to the closed shell system Eo is taken as
the reference value.
The wave functions will, in general, be linear combinations of the possible
basis functions. This means that for J7r = 0+ we will get two eigenfunctions
n
I!liO+;l) = L ak,ll~~O); 0+) ,
k=l
n
I!liO+;2) = L ak,21~~O); 0+) , (3.120)
k=l

where for the particular case of 180 we define

I~~O);O+) == l(lds/ 2)2;0+) ,

I~~O);O+) == 1(2S1/2)2; 0+) .
Before turning back to the particular case of 180, we make the method more
general. If the basis set is denoted by I~~»)(k = 1, 2, ... , n), the total wave
function can be expanded as

102
 n
l!lip } = L akpltP~)} . (3.121)
k=l
The coefficients akp have to be determined by solving the SchrOdinger equation
for l!lip }, or

HI!lip} = Epl!lip} . (3.122)

In explicit form this becomes [using the Hamiltonian of (3.119)]

n n
(Ho + Hres) L akpltP~)} = Ep L akpltP~O)} , (3.123)
k=1 k=l
or
n
L(tP~O)IHo + HresltP~)}akp = Epa,p . (3.124)
k=l
Since the basis functions ItP~)} correspond to eigenfunctions of Ho with eigen-
values (unperturbed energies) E~), we can rewrite (3.124) in shorthand form
as
n
L H'kakp = Epa,p , (3.125)
k=l
with
0) I
_ ",(0)
H'k =.J!ii C,k + (tPl Hres ItPk(O)} . (3.126)

The eigenValue equation becomes a matrix equation

[H][A] = [E][A] . (3.127)

This forms a secular equation for the eigenvalues Ep which are determined from

... Hln
... H2n
=0. (3.128)

Hnl Hnn - Ep
This is a nth degree equation for the n-roots Ep (p = 1, 2, ... , n). Substitution of
each value of Ep separately in (3.125) gives a set of linear equations that can be
solved for the coefficients akp. The wave functions l!lip } can be orthonormalized
since
n
L akpakpl = Cppl . (3.129)
k=l

103
Out of (3.125) it now follows that
n
L alp' Hlkakp = Ep/jpp' , (3.130)
l,k=l

or, in matrix form

[A][H][A] = [E] , (3.131)

with

Equation (3.131) indicates a similarity transformation to a new basis that makes

[H] diagonal and thus produces the n energy eigenvalues. In practical situations
(n large), this process needs high-speed computers. A number of algorithms
exist for [H] (hermitian, real in most cases) matrix diagonalization which we
do not discuss here (Wilkinson 1965): the Jacobi method (small n or n ~ 50),
the Householder method (50 < n < 200), the Lanczos algorithm (n ~ 1000,
requiring the calculation of only a small number of eigenvalues, normally the
lowest lying ones). In cases where the non-diagonal matrix elements IHij I are on
the order of the unperturbed energy differences IE~O) - EjO)I, large configuration
mixing will result and the final energy eigenvalues Ep can be very different from
the unperturbed spectrum of eigenvalues E~O). If, on the other hand, the IHij I are
small compared to IE~O) - EJO) I, energy shifts will be small and even perturbation
theory might be applied.
Now to make these general considerations more specific, we discuss the case
of J7r = 0+ levels in 18 0 for the (ldsf22s l / 2 ) model space. As shown before, the
model space reduces to a two-dimensional space n = 2 (3.120) and the 2 x 2
energy matrix can be written out as

H= [2e 1d S/ 2 + ((ldS/Z)Z; O+IVizl (lds/z)z; 0+) ((ldS/Z)Z; o+lVizl (281/Z)Z; 0+) 1

(( 281/z) z; 0+ IVizl (lds/ 2 )z; 0+) 2o:Z 81 / 2 + (( 281/Z) z; O+IVizl ( 28 1/2) Z; 0+)
(3.132)
These diagonal elements yield the first correction to the unperturbed single-
particle energies 2£ld s / 2 and 2£28 1/2' respectively (the diagonal two-body interac-
tion matrix elements H11 and H 22 , Fig. 3.36). The energy matrix is Hermitian (for
a real matrix this means symmetric and H12 = H21 ) and, in shorthand notation
gives the secular equation

[ Hll - A H12] =0 (3.133)

HI2 H22 - A .

We can solve this easily since we get a quadratic equation in A

A2 - A(Hll + H22) - Hf2 + H11H22 =0 (3.134)

104
 18 0
8 10

...

1':
2e: 1d "-
3/2 "- ,
2E 1d3/2+(V3 > EO·
3

>
~ 6
>-
<!>
a:
w 4
z
W
<!>

-r
~ 2
0
z ..... .....
2 E 251/ 2 EO·
iii 2E251/2· <V2 > 2
a ...
2Eld "-
512 "-
"- ,
2E 1d 5/2+( ~>-..
EO·
1

Fig. 3.36. The solution of the secular equation for 0+ states in 180, shown in a diagrammatic way. On
the extreme left we show the unperturbed two-neutron single-particle energies 2€j. In the middle part
=
the diagonal interaction matrix elements (i; J" 0+IVi,2Ij2; J" 0+) for the three configurations =
are added and denoted by (VI), (Vz), (V3), respectively. On the right-hand side, the resulting energy
eigenvalues Eot (i = 1,2,3) from diagonalizing the energy matrix (3.132) are shown

with the roots

Hll + H22
A± = 2 ±"21 [(Htl - H22 ) +4H12
2 2 ]1/2
(3.135)

The difference LU == A+ - A_ then becomes

2 2 ]1/2
LU = [ (Hll - H22) + 4H12 , (3.136)

and is shown in Fig.3.37. Even for Hn = H22, the degenerate situation for
the two basis states, a difference of LlA = 2H12 results. It is as if the two
levels are repelled over a distance of H12. Thus 2H12 is the minimal energy
difference. In the limit of IHn -H22 I ~ IH12I, the energy difference LlA becomes
asymptotically equal to LlH == Hn - H22.
The equation for A± given in (3.135) is interesting with respect to perturbation
theory.

105
 Fig.3.37. The variation of the differ-
ence of eigenvalues Ll,X == ,X+ - ,X_
for the two-level model of (3.133) and
(3.136) as a function of the energy dif-

, ference between the unperturbed states

LlH == IHI1 - H221
t
,<
/"
~

<I
/
/
/
/
/
/
/
/
/

- - - IH n- H22I--

i) IT we consider the case IHn - H221 ~ IH121, then we see that one can
obtain by expanding the square root around Hn = H22

Hf2
Al = Hn + H nH + ... ,
- 22 (3.137)
Hf2
A2 = H22 + H H +. .. .
22 - n
[We use the expansion (1 + x)I/2 ~ 1 + + ....) !x
ii) One can show that if the perturbation expansion does not converge easily,
one has to sum the full perturbation series up to infinity. The final result of
this sum can be shown to be equal to the square root expression of (3.135)
(see Problem set).
It is also interesting to study (3.135) with a constant interaction matrix el-
ement HI2 when the unperturbed energies Hn, H22 vary linearly, i.e., Hn =
E~) + xa and H22 = E?) - Xb (Fig. 3.38). There will be a crossing point for the
unperturbed energies at a certain value of X = Xcrossing. However, the eigenvalues
El, Eh. will first approach the crossing but then change directions (no-crossing
rule). The wave functions are also interesting. First of all, we study the wave
functions analytically (the coefficients akp)'
We get

l!lil} = allltJ!~O)} + a211tJ!iO)}

(3.138)
1!li2} = a121tJ!~O)} + a221tJ!iO)} .
IT we use one of the eigenvalues, say AI, the coefficients follow from
(Hn - At)an + H12a21 =0, (3.139)

106
I
Fig. 3.38. The variation of eigen-
values ..\± [see (3.135)] obtained
from the two-level model as a
function of the parameter X which
+1 describes the variation of the un-
,<
perturbed energies Hu. H22. We
take a linear variation Hu =
Vl
w
:l
~
Ff;0) + Xa, H22 = 14°) - xb
z with an interaction matrix ele-
w
(!)
iii
ment Hlz. The variation (through
>- the crossing zone) of the wave
(!)
a:
w
function character is also indi-
z cated on the figure
w

o ----x--....-

Fig. 3.39. The change in energy separa-

tion .1 (.1 == H22 - Hll) to .1..\ (==
Ez - EI) from unperturbed to perturbed
speCtrunl in a two-level model. The spe-
cific form of the wave functions lqil} and
lqiz} is also given using the basis func-
tions 1'Ifo~O\ l'IfoiO)}

or
an -H12
-= (3.140)
a21 Hn - Al
The normalizing condition arl + a~l = 1 then gives
a21 = ( 1/(1 + (H12/(Hn - A))2))1/2 , (3.141)

and similar results for the other coefficients. In the situation that Hll = H22, the
absolute values of the coefficients all, a12, a21 and a22 all have absolute value
1/../i. These coefficients then also determine the wave functions of Fig. 3.39 at
the crossing point One can see in Fig. 3.38 that for the case of X = 0, one has

EI ~ Eft°) and l!lil) ~ 17jJ~O») ,

Ez ~ EiO) and 1!li2) ~ 17jJ~O») .

107
On the other hand, after the level crossing and in the region where again
IHll - Hnl ~ IH121, one has

El ~ H22 and 1!¥1) ~ l¢iO») ,

E7. ~ Hll and 1!¥2) ~ I¢~O») ,
so that one can conclude that the "character" of the states has been interchanged
in the crossing region, although the levels never actually cross!
In the realistic situations of ISO and 210pO (Figs. 3.29 and 3.30) (Brussaard,
Glaudemans 1977), the results are now the results of configuration mixing: in 18 0
(the Ids/ 2 and full Idsf2 , 281/2, Id3 / 2 spaces, respectively) and for 210pO [using
the (1h9/2)2, 1h9/2 217/2, Ih 9,2 1i13/2 and the full (lh 9/2, 217/2, li 13/ 2) spaces,
respectively]. In the case of 1 0, one observes a net improvement for the larger
model space when comparing with the data. Also, for 210po some improvement
can be observed for the larger model space. In both cases esO, 21OPO) one needs
to reduce the MSDI force strength when going to the larger space.

3.3 Three-particle Systems and Beyond

It is our purpose to construct fully antisymmetric and normalized many-particle

wave functions in order to be able, using the methods as outlined in Sect. 3.2,
to obtain energy spectra and wave functions for the many-particle nuclear
SchrOdinger equation. We first start from the three-particle system to show the
general methods before extending to the n-particle system.

3.3.1 Three-particle Wave Functions

First of all, we consider three-particle systems with the nucleons moving in the
orbitals ja, jb and jc [remember we use the notation j == (n, l,j)].
i) We consider the simplest situation of ja 'f jb =I jc. The anti symmetrized,
normalized wave function is constructed in a straightforward way as

AI" 2)-1)P ¢«ja(l)jb(2))JI2,jc(3); JM) , (3.142)

P

where AI" is the normalization and L: means a sum over all permutations of the
particle coordinates 1,2 and 3 over the orbitals ja,jb and jc with (-I)P = +1
for an even permutation of (1,2,3) and -1 for an odd permutation of (1,2,3).
So, (3.142) becomes

~{ ¢(ja(1)jb(2))J12,jc(3); JM) - ¢«ja(2)jb(1))JI2,jc(3); JM)

+ ¢«ja(2)jb(3))J12,jc(1); J M) - ¢«ja(3)jb(2))JI2,jc(1); J M)

108
 + 1/J«ja(3)jb(1»J12,jc(2); JM) -1/J([ja(1)jb(3)] Jt2,jc(2); JM)} .
(3.143)
ii) We consider the case (j a = j b = j) :f j c' In this case, the two-particle wave
function where ja and jb in (3.143) show up now contains two identical orbitals
j. Thus, the six tenns are reduced to only three tenns because of the symmetry
condition
1/J«j(2)j(1»J12,jc(3); JM)
= (-1)2 i -J12 1/J«j(1)j(2»Jt2,jc(3); JM) , (3.144)

with J12 even and (_1)2i = -1. Thus the total wave function of (3.143) reduces
to

~{ 1/J(j\12»J12,jc(3); JM) -1/J«j2(13»J12,jc(2); JM)

+ 1/J «j2(23) )J12, j c(1); J M) } . (3.145)

iii) The case where all three angular momenta become equal ja = jb = jc = j
is the most interesting one. It is possible to recouple the angular momenta in
(3.145) such that the particle coordinates in the three tenns always come in
the sequence 1,2,3 and in increasing order. In this case, the standard rules of
angular momentum algebra when calculating matrix elements are applicable. We
try to bring the second and third tenn in (3.145) in the order of the first tenn by
recoupling

1/J«j\23»J12,j(I); JM)

= 2:(-1) J12Jt2J12 {~
A AI J
~{:} 1/J«j2(12»JIz,j(3); JM) ,
J
J~2
and
1/J«j2(13»J12,j(2); JM)
= 2:( _1)J12+J~2+1 J12 J{2 {~ J
J
Jt2}
J{2
J~2

x 1/J«j2(12»JIz,j(3); JM) . (3.146)

By bringing these tenns together with the first tenn in (3.145) we always have
the particles in the order (1,2,3). Thus we can actually leave out the nucleon
coordinates. The state thus constructed is an antisymmetric state of three par-
ticles moving in a j-orbital and coupled to angular momentum J, denoted as
1/J(j3; J M)as. Written explicitly,

109
 x {~ ~ ~::} ]1/J( (i(12») Jh,j(3); JM) . (3.147)

The expansion on the right-hand side goes over all states Jh where one cou-
ples a third particle j(3) to an antisymmetric state of two particles in a single
j-shell P(12)Jh- The latter basis, a basis of three-particle wave functions an-
tisymmetrized in particle coordinates 1 and 2 but not in particle coordinate 3,
is much larger than the basis of three-particle states antisymmetric in all three
particle coordinates. In (3.147) one writes the 1/J(p; J M)nas through calculating
the overlap with all states 1/J«P(12»J12, j(3); J M), i.e., we project onto the sub-
space 1/J(p; J Mkas. The expansion (projection) coefficients [... ] do not form a
unitary transformation, therefore a special notation is used for the coefficients:

1/J(j3; JM1as =L [j2(J1)jJI}P J] 1/J(i(J1),j; JM) , (3.148)

Jl
even

and the coefficient [j2(J1)j JI}j3 J] is the cfp coefficient (coefficient of frac-
tional parentage) giving the projection of the state 1/J(p; JMkas on the basis
1/J(P(J1)j; JM). In the latter wave functions we leave out of the right-hand side
the nucleon coordinates since they are ordered in the sequence 1,2,3.

As an example we discuss the case of the (dS/ 2 l J configurations.

In making the two-particle configurations first, we have (dS / 2 o+, 2+, 4+ t
configurations. Next we couple a third dS/ 2 particle by just using angular mo-
mentum coupling, thereby constructing the basis I(dS/2)2 J12(d s/ 2); J M), i.e., we
have
l(ds/2)~+' dS/2; 5/2+)
l(ds/ 2h+, dS/ 2; 1/2+, ... ,9/2+)
l(ds/2)!+, ds/2; 3/2+ , ... , 13/2+)
This space has 12 states. Now using (3.147) to explicitly calculate the three-
particle cfp's for the (dS/ 2)3 J configurations, for all J12 intermediate values one
keeps just 3 fully antisymmetric states with Jff = 3/2+, 5/2+, 9/2+. If we call the
first space A (12 states) and the latter space B (3 states), then B is entirely within
space A and the states in B can be expanded using the projection coefficients of
the fully antisymmetric states on the basis spanning space A (Fig. 3.40). Making
this explicit for (dS/ 2 )3 J, we have

I(dS/2)3; J =~) = - ~1(dS/2)~+' dS/2; 5/2+)

+ v;,.1(ds/ 2
3v2
h+, ds/2; 5/2+)
+ ~1(dS/2)!+' ds/2; 5/2+) ,

110
 Fig. 3.40. Example for the construction of anti symmetric three-
Space A particle states in the (dS/Z)3 configuration. We indicate space
A: the larger space of states I(ds/zf JI, (ds/ z ); J M) antisym-
metric in two-particles only and the smaller space B: the space
l(ds/ Z)3; J M) of antisymmetrized three-particle configurations.
Space (B) is a subspace of space (A) and can be obtained via
a projection

Space B

I(ds/d; J = ~} = - ~1(ds/2~+' dS/ 2; 3/2+}

+ ~1(ds/2)!+' ds/2; 3/2+} ,
I(ds/d; J = ~} = ~1(ds/2~+' dS/ 2; 9/2+}
-
JIT 2
/1A1(ds/2)4+'
• +
ds/2, 9/2 } .
v14
In describing the fully antisymmetric three-particle states as in (3.148), in
most cases more than just one J state can be made from the j3 configuration:
an extra quantum number a will be needed to characterize the state uniquely.
H only one state can be formed, a is not necessary. In the latter case, the
wave function should be unique (except for an eventual overall phase factor)
and independent of the initial two-particle spin 1t2
[see (3.147)]. In making
the discussion somewhat more detailed, we first nonnalize the states 'Ij;(p; J M)
leading to normalized cfp coefficients

[l(J1)jJlllJ] = [c5 J1 ,J12 +2i1i12

{~ ~ ~12}] [3+6(2J12+1){~ ~ 1t2} ]-1/2 (3.149)

J J12
In the case J = j, we can start from the choice J12 = O. For j ~ ~ there is
only one antisymmetric state t/J(p; J = j) (by construction). For j > ~, this is
no longer the case.
If we now put J12 = 0, and fill in the explicit values of the Wigner 6j-symbols,
we get

[l(O)j J=jill J =j] = «2j - 1)/(3(2j + 1)i/2 ,

[l(J1)j J = jlli J = j] = -2«2J1 + 1)/(3(2j - 1)(2j + 1))1/2 , (3.150)

with J1 > 0, even.

111
 For the case of (dS/ 2)3, these expressions (3.150) reduce to the wave function

'IjJ( (ds/d (J12 = 0), J = ~) = V; 'IjJ( (dSf2)2 Jt =0, dS/2; J = ~)

v'5 (2
- '13 .J2'IjJ (ds/2) Jl = 2, dSf2; J = 2s)

- ~ 'IjJ ((ds/d Jl =4, ds/2; J = ~) .(3.151)

Here, we include J12 = 0 in the left-hand side to point out that we have
antisymmetrized starting from the J12 = 0 two-particle state. One can now verify
that starting from other states J12 = 2, 4 leads to the same wave function (3.151)
multiplied by a phase of -1. This will no longer be the case for j = ~ in
the (9 j2J configuration where different wave functions (cfp coefficients) result
when starting from J12 = 0 or from J12 = 2. In these cases, one has to construct
orthonormalized states explicitly.

3.3.2 Extension to n-particle Wave Functions

Using the discussion in Sect. 3.3.1, by using the building up principle one can
construct consecutively more complicated states, ending up with n-particle states

'ljJUna; JM)nas = L [jn-l (atJl)jJI}jn a J]

0I1,J1

(3.152)

where cfp coefficients are used to project from the fully anti symmetric n-particle
wave functions onto the space of wave functions antisymmetric in the first (n -
1) particles, coupled to the nth particle with the use of angular momentum
coupling. Thus, this is a 2 -+ 3 -+ ... n - 1 -+ n building up principle. One
can also carry out the process by using two-particle cfp coefficients where the
'IjJ(jn a; J M)nas wave functions are constructed from an anti symmetrized (n - 2)
and an antisymmetrized 2 particle wave function. This leads to an expression of
the form

'ljJU na ; JM)nas = L [jn-2(a 1Jl)j2(12) Jl}jnaJ]

0I1,J1,J2

(3.153)

Here the n -+ n - 2 cfp can be expressed in terms of n -+ n - 1 and

n - 1 -+ n - 2 one particle cfp coefficients (see problem set) with as a result

[jn-2( al Jt)j2( h) JI}jnaJ] = L [jn-2( al Jl)j J{I}jn-l aP:]

a~,J~

112
 (3.154)

Using the above wave functions, we are now in a position (using the reduction
rules I, n of Chap. 2 for calculating matrix elements of spherical tensor operators)
to evaluate all matrix elements necessary to set up a n-particle energy matrix and,
later on, to calculate one-body operator expectation or transition matrix elements
to test the wave functions against observable nuclear properties (electromagnetic
moments and transition rates, beta-decay transition probabilities, etc.).

i) One-Body Matrix Elements. Here, the general one-body operator for an n-

particle system reads
n
r,.k) = 2: f~k)(i) , (3.155)
i=t

where i denotes the particle coordinates (ri, «Ti, •..). We wish to evaluate the
reduced n-particle matrix element

(3.156)

which is related to the normal matrix element (jna; J MI~k)ljna'; J'M') via
the Wigner-Eckart theorem. Since ~k) is a sum over n particles and since
'IjJ(jna; J M) and 'IjJ(jna ,; J'M') are antisymmetrized wave functions, one can
write that
(jna; JMlf~k)(1)ljna'; J'M') + (jna; JMlf~k)(2)ljna'; J'M')
+ ... + (jna; JMlf~k)(n)ljna'; J'M')
=n{ra; J Mlf~k)(n)ljna'; J'M') . (3.157)

Using the n ~ n - 1 cfp coefficients, one can separate the (n - 1) particle part
out of the wave function and obtain after some simple algebra for the reduced
matrix element of (3.156)

(jnaJIIF(k)lljn a , J') =n 2: [jn-t (at1t)jJlljn a J]

01,J1

x [jn-t (at Jt)jJ'I}ra' J'] (_I)Jl+i+ J+k j j'

x {J, f ~} (jlllk)lIj) . (3.158)

Thus, the n-particle matrix elements are expressed in terms of the one-body
matrix elements (jlllk)lIj) only. This is an important result which we shall
discuss later in some detail.

113
ii) Two-Body Matrix Elements. Here, a general two-body operator is
n
V =L V(i,k).
i<k=l

The nuclear two-body interaction has this structure and we shall need these op-
erators in evaluating the energy for an n-particle system (ino:; JMlVlino:; JM)
or calculate the matrix elements needed to solve the n-particle secular equation.
Using similar arguments as under (i), since the two-body operator V =
V(1, 2) + V(1, 3) + ... V(2, 3) + ... V(n - 1, n), is symmetric in the coordinates of
the interacting nucleons and since the wave functions are antisymmetric in the
interchange of the coordinates of any two nucleons, one has

(ino:; JMlVljno:; JM) = n(n2- 1) (ino:; JMIV(n-l,n)lj no:; JM) .(3.159)

Now, using the n ~ n - 2 cfp coefficients, (3.159) can be reduced to

(ino:; JMlVljno:; JM) = n(n2- 1) L [jn-2(O:2h)P(J')JI}jno:Jr

Q2 J 2,J'

x {p; J'M'IVIP; J'M'} , (3.160)

where a reduction of the n-particle interaction matrix element is obtained as a

linear combination of two-body matrix elements only. If we now use the total
Hamiltonian for the n-particle system
n n

H = Ho + Hres = L ho(i) + L V(i,j) , (3.161)

i=1 i<i=1

the expectation value of the energy becomes

(ino:; JMIHljno:; JM) = nei + n(n2-1)

L [jn-2( O:2h)P(J')Jlljno:Jr {l; J'M'IVIP; J'M'} . (3.162)
Q2 J 2,J'
As we conclude this section, before discussing some applications, it becomes
clear that once we know the "basic" constituents: the single-particle matrix ele-
ments ei = (imlho<i)ljm), (illlk)lIj) for a general operator f~k) and the two-
body interaction matrix elements {P; J'M'IVIP; J'M'}, by using (3.158,162),
the related n-particle matrix elements can be obtained in order to study n-particle
nucleon systems when knowing the cfp coefficients. Thus one can test, e.g., the
consistency of the nuclear observables in the 117/2 shell model orbital (excita-
tion energies, transition rates): nuclei in between ~Ca20 and ~Niz8 starting from
a single set of two-body matrix elements {(lh/2)2; JMIVI(1h/2)2; J M) and
the one-body matrix elements {lh/2I1lk)1I1h/2}. This process is schematically
given in Fig. 3.41.

114
 Fig.3.41. Schematic illustration of how
the n-particle properties (energy matrix
elements and transition matrix elements)
are related to the one- and two-body
properties E:j, <P; J MIVi,zIP; J M} and
Ullf(k)lli} using the cfp coefficients for
an-+n-l-+n-2-+ ... -+2
decomposition

3.3.3 Some Applications: Three-particle Systems

Starting from the discussion at the end of Sect. 3.3.2, we are now able to calculate
spectra in nuclei having three particles outside a closed shell configuration. The
first application, of course, is the case where only a single configuration j is
important.
In this case,

(in; JMIHlin; JM) = 3cj +3 ~)p(J')jJl}i Jt AI' , (3.163)

I'

(with AJI == (j2; J'MlVlj2; J'M), the unperturbed energy 3cj can be left out
as a constant energy shift. For a (d5 / 2 )3 configuration, one has J = ~, ~ and 1,
Ao, A2 and ~. The corresponding cfp coefficients are given in Table 3.3.
Using these cfp, we obtain

J=~ -t i
~ Ao + A2 + A4 , 1
J=l2 15 A
-t 7" 2+".6~ , (3.164)
J=22 -t 9
14
A 2 + 14
33 A
4·

Now, by using a residual 6-interaction the two-particle matrix elements can be

obtained, using (3.90) as

Table 3.3. The one-particle coefficients of fractional parentage (cfp) for the (dS / Z)3 configuration.
So, we denote the coefficients as [(ds/z)z J' ds/ 2 JIHds/ 2 )3 J] and label the rows and columns by
J and JI, respectively

JI 0 2 4
J
5
2 -vII (31) 1
7z
3
2 -~ 1
~ -~14
9
2

115
 Fig. 3.42. The three-particle spectrum (ds /2)3
J1r (J'" = 5/2+, 3/2+. 9 /~ expressed in
tenDs of the two-body properties. The equa-
tions (3.165) are used with F' the Slater in-
tegral for the c5-interaction. Relative to the
unperturbed energy 3eds / 2 • the energies are
>-
o
a: - - - - - - - 9/2+ given in units FJj2 [with F~ =(2j + 1)Fl)
UJ
-0.371
z
UJ
o - - - - - - - 3/2+
z -0571
is
z
iii

- 1 0 0 0 - - - - - - - 5/2+

Ao = F&/2; A2 =8/35(F&/2); At =2/21 (Fa/2) ,

with
Fa = (2j + l)pO = 6pO .
This then leads to the interaction matrix elements
-t F0/2,
-t HF0/2) , (3.165)
-t ~~ (Fa/2) .
The result is shown in Fig. 3.42.
It is now possible to use the expressions (3.164) in a slightly different way.
If one knows in the nucleus with a two-particle (d5/ 2 )2 configuration, the experi-
mental 2+ and 4+ states and their excitation energy, then one can use (3.164) but
now interpreting the matrix elements AJ as differences from matrix elements,
i.e., AJI == AJ - Ao (Ao can not be determined from experiment but the AJI
are the relative or excitation energies for the corresponding J states). Thus one
obtains the relative energies in the three-particle case from the relative energies
in the two-particle case (see problem set).
We illustrate this for the relation ~n28 -t~ V28 (Fig. 3.43) where (Brussaard,
Glaudemans 1977)
i) We use theoretical values (MSDI-matrix elements Ao, ... A6) obtained from
a fit to the ~Ti spectrum in order to evaluate the @V spectrum. Agreement
is not very good.
ii) Now, using the empirical method where relative matrix elements A~, ... A:'
are taken from the ~n experimental spectrum, agreement becomes rather
good. This latter point proves that the spectra in SOn_51 V are rather consistent
with a single 1h /2 shell model configuration using an empirical set of two-
body matrix elements.

116
 THE (l f7/2}n MODEL FOR SOTi AND 5 I V
Ex(MeV) JIt

(s") 3 I I~/2-

Ex(MeV) JIt
2.S8 270
4" I~I~:
112"
3/2-

9/2-
6"
4" 9/2-
I ~
2" I.~3 u'>a
2" I~/2-

9/2-
1.19 1"2-
3/2-
114 3/2 -
0.9 3/2- ~/2-
0.98

0.36
~/2 - 0.32 ~/2 -

0" o· 7/2- 7/2- 7/2-

experiment MSDI empirical experiment MSDI
two-body me
50
Ti 51V
22 28 23 28

II
Fig. 3"43" The spectra of ~T~ and V 28 for the configurations (1 h /2)2 and (1 h /2)3. Both the
MSDI interaction (using a value of AT=1 = 25/AMeV) and the case of two-body empirical matrix
elements (taken from the experinlenta1 spectrum of son itself) are illustrated [taken from (Brussaard,
Glaudemans 1977)]

This discussion illustrates very well the methods discussed in Sect. 3.2.2.
We further illustrate three-particle spectra for some higher j orbitals: the
proton (199/2)3 spectrum and the neutron (197/2)3 spectrum (Fig. 3.44), (Table
3.4) (De Gelder et al. 1980).
From all the above illustrations, it is clear that the lowest-lying state is always
the (j)3 J = j state. This is easily explained since only for this J = j state, the
cfp for the intermediate two-particle 0+ state occurs. Now, the 0+ two-particle
matrix elements are by far the largest attractive matrix elements and still appear
in the three-particle spectra via the cfp coefficients.
In the above situation, if we consider besides the 117/2 orbital also the rela-
tively close-lying 2P3/2 and 115/2 orbitals, a more complicated situation results.
This is because as in the two-particle case with configuration mixing, we have
to construct all three-particle configurations for given J1r values. For the present
(117/2, 2P3/2' 115/2) space with three particles one obtains
(lh/d J1r =3/2- , 5/2- , 7/2- , 9/2- , 11/2- , 15/2-
(2P3/d pr = 3/2-
3
(1/5/2) J7r = 3/2- , 5/2- , 9/2-

117
 Fig. 3.44. We show the proton
Itt ,3 IV,g ,3 (1r199 / 2)3 and neutron (vl!17/2)3 en-
'9 912 7/2 ergy spectra with the proton and
21h- neutron two-body interaction deter-
mined as follows: the proton (1r 1!J9/2)2

t ISfz-
l'¥z·
ISIi
experimental spectrum of :Zr
matrix elements are taken from the
48 and
the neutron (vl!17 /2)2 matrix ele-
>
~2.0 ments from the experimental spec-
>- trum of 1~Th82 [taken from (De
I!)
a:: Gelder 1980)]
w v=3.9fi.
z
w
z 312- W
0
>= II/i IV;
:! 13fi
U
~,.o 51;
3h-

_ _ _ _ vol.9fi v.I.'h- _ _ __

Table 3.4. The cfp coefficients [(h/2)2J' h/2JIl<h/2)3J] for the three particle configurations.
We denote the rows and columns by J and J', respectively [taken from (Brussaard, Glaudemans
1977)]

J J'=O J'=2 J'=4 J' =6

3
2 0 0.463 -0.886 0.000
5
2 0 0.782 0.246 -0.573
7
2 0.500 -0.373 -0.500 -0.601
9 0.321 0.497
2 0 -0.806
11 0.527 -0.444 -0.725
"2 0
15 0.000 0.477 -0.879
"2 0

J1r =3/2-
r = 1/2- ,3/2- , 5/2- , 7/2-

In this manner one constructs the model space for each J1r, and using the tech-
niques discussed in Sects. 3.2.4, 3.3 one can calculate the energy matrix in order
to obtain the energy eigenvalues and the corresponding wave functions.

118
3.4 Non-identical Particle Systems: Isospin

Up to now we only considered those situations where identical nucleons outside

closed shells detennine the nuclear structure. In many cases, however, both pro-
tons and neutrons are present outside the closed shells. Therefore, we would like
to construct an extension of the methods outlined above so that the same general
structure remains. We shall introduce the concept of "charge" quantum number
or "isospin" quantum number, depending on which nuclei we shall treat. Before
constructing the nuclear wave functions including this new quantum number,
we shall point out the evidence that exists in nuclear properties to introduce the
concept of isospin as a valuable quantum number.

3.4.1Isospin: Introduction and Concepts

Protons and neutrons have almost identical mass (Llm/m 9:' 1.4 x 10-3 ) but
otherwise show an almost identical behavior in their nuclear interactions. In 1932,
Heisenberg proposed to consider protons and neutrons as two distinct fonns of
"nucleons", by using a double-valued variable called isospin that distinguishes
between the "proton" state (described by a projection quantum number tz =
-1) and a "neutron" state (described by a projection quantum number tz =
+1) (Heisenberg 1932). The isospin (or isotopic spin) fonnalism can now be
duplicated from the properties we studied in Chaps. 1,2 on intrinsic spin (s) of
the proton and the neutron. We first present some of the experimental evidence
for the equivalence of protons and neutrons in their nuclear interactions.
i) Low energy np scattering and pp scattering below E < 5MeV, after cor-
recting for Coulomb effects, is equal within a few percent in the 1S scatter-
ing channel (Arndt, MacGregor 1966, Wright et al. 1967, MacGregor et al.
1968a, b).
ii) Energy spectra in "mirror" nuclei are almost identical (Fig. 3.45). The small
differences are both a consequence of the difference in the Coulomb inter-
action energy and of specific nuclear wave functions. From this observation
one concluded that the exchange of protons and neutrons gives no modifica-
tion of the nuclear interaction energy or, that the substitution n - n {:} p- p;
n - p {:} P - n does not modify the interaction energy. This observation
implies the concept of charge symmetry in nuclear forces.
iii) Further infonnation on how the n - n, p - p forces relate to the n - p force
cannot be deduced from mirror nuclei. If we, therefore, study the triplet
of nuclei, e.g., ~~Si16, ~gPlS, ~2S14' it is immediately clear that within a
number of states (0+,2+) (after correcting for Coulomb energies) the nuclear
binding energies are equal in all three nuclei. From this observation a new
characteristic of the nuclear forces can be deduced. Taking as a core ~Si14'
the data show that the residual interaction energies due to n - n, p - p
and n - p interactions are equal in a number of states (Fig. 3.46). The above
leads to an even more stringent condition than obtained from (ii), i.e., charge

119
 THEA = 25 MIRROR NUCLEI Fig, 3,45. A comparison of the level schemes of
the A = 25 (25Mg_25 AI) mirror nuclei shows
the close similarity of the excitation energies for
the states with identical J7r values [taken from
(Brussaard, Glaudemans 1977)]

2.80 3/2'
2 . 7 4 - - - 7/2 +- ---~-_- _~;>',§l2~===:..;~!£~~r'
2.56 1/2' --=.
---- ___ ~2.~4~9~_ _~1~/2~'

1.96 5/2'
- -.__ 1.79
1.61 7/2' - - - - - - - 1.61

097 3/2'
- - --094
0.59 112'
- 045

_ _ _ _ _5""/~2' ______ _ _ _ _----'5"".:0..2'

25 A1
13 12

THE A = 30 ISOSPIN TRIPLET

E/MeV)

302 1
:0.2-=2....:4_ _-=-2_· . . . . ~ §~ - - - 1 ' 2'T, 1. . __ 2"'..:.2...:.1_ _--'2~·
2 72 2'
254 3'

197 3'

145 2" Fig. 3.46. The level spacings be-

tween the T = 1 isospin states
in the mass A = 30 chain eOSi-
071 1" 30P_30S) are very similar to the
0" 068 O"T'l 0"
corresponding spacings in 30Si,
30p and 30S. The states in lOp
1" where isospin is not given are
the isospin T = 0 states [taken
30
5, 30
P '5
30
5 '4
from (Brussaard, Glaudemans
14 1'6 15 16 1977)]

120
 independence in nuclear configurations that are possible in an n - n, n - p
and p - p interacting system. A number of states in 30p do not find a partner
in the 3OSi, 30S nuclei. This follows from the Pauli principle that excludes
the realization of a number of configurations in identical nucleon systems
(n - n, p - p) compared to the n - p (non-identical nucleons) system. Thus
the Pauli principle explains the large number of extra states in 3Op, as shown
in Fig. 3.46.

3.4.2 Isospin Formalism

a) General Properties of Isospin
We describe the neutron and proton by using a two-valued quantity just like the
intrinsic spin, but now in isospin space:

<Pn(r) = <p(r) (~) , (3.166)

<pp(r) = <p(r) (~) , (3.167)

where <p(r) describes the spatial wave function which is the same for the proton
and the neutron in the present case.
As for intrinsic spin, we can introduce Pauli isospin matrices r(Tx , T y , T z )

Tx = ( 01 01) ; Ty = (0i -i)

0 ; Tz = (10 0) '
-1 (3.168)

and define the isospin (t = ~) operator as

(3.169)

All other spin ~ algebra results and properties can be used again: we have

[tx, ty] = it z, and cyclic permutations,

(3.170)
[t2,ti] =0, i=x,y,z,

and t2, acting on the isospin spinors, gives as an eigenvalue t(t + 1). The proton
and neutron wave functions (3.166,167) are now eigenfunctions of t 2 and tz

tz<pp(r) = -~<pp(r) ,
(3.171)
tz<Pn(r) = ~<Pn(r) .
Now, the following relations hold

121
 HI + TZ)r,op(r) =0, (3.172)
HI - Tz)r,on(r) = 0,
! (1 + Tz)r,on(r) =r,on(r) ,
and we can introduce the charge operator

Q 1
-; == 2(I-Tz ) WIth
. Q
-; = (0 0)
0 1 (3.173)

The ladder operator properties hold here, too, and change proton into neutron
states (t+) or neutron into proton states (L), respectively:

=r,on(r) ,
t+r,op(r)
t+r,on(r)=0, (3.174)
Lr,on(r) = r,op(r) ,
Lr,op(r) =0 .

Extending to a many-nucleon system, the total isospin operators are constructed

according to the methods of adding angular momenta (Chap. 2) and we obtain

i=1
(3.175)
A
Tz = Ltz,i.
i=1

Acting on a many-nucleon eigenstate, the eigenvalues of T2 are T(T + 1) with

-T ::; Tz ::; +T. The total isospin will be integer or half-integer according to
whether A is even or odd. For each T value, there exists an isospin multiplet
with 2T+ 1 members, characterized by the z-component Tz since Tz varies from
-T to +T. The eigenvalue of Tz according to its definition in (3.175) becomes
~(N - Z). On the other hand, in a given nucleus A(Z, N) with a given T z
value one can find a number of states with different isospin T according to the
conditions
A
T=ITzl, ITzl+l, ... '"2.
This multiplet structure is illustrated in Fig. 3.47 for a number of nuclei. The
different members of an isospin multiplet denoted by the eigenvector IT, T z} are
connected via the ladder operators
A
T± = Tx ±iTy = Lt±(i) , (3.176)
i=1

and give as a result

122
 T=2
T~=-2 -1 o +1 +2

T =1
TZ=-l o +1

T=O

N= Z --N--

Fig. 3.47. lllustration of a series of levels in isotopic chains T = 0, 1,2, .... The variation with
the neutron excess N - Z = 2Tz is illustrated (isospin states in different nuclei of a given mass A
chain) and, for a given nucleus (given T z ) of the different isospin states T (ITz I ~ T ~ Aj2) as a
function of excitation energy

(3.177)

Consecutive action of T± changes Tz but does not affect the isospin T. This
means that within an isospin multiplet the spin-spatial wave function remains
constant; only the charge part changes as is expressed in the IT, T z } eigenvectors.
We now try to express the charge symmetry and charge independence prop-
erties of the nuclear interactions in a more formal way.
i) Conservation of charge implies

(3.178)

ii) Charge independence implies that all members of a given isospin multiplet
have the same energy, so that along a multiplet E = E(T) only and no
Tz-dependence shows up. Expressed in a mathematical way this demands
that

HIT, Tz} = E(T) IT, T z} ,

HT+IT, Tz} = E(T)T+IT, T z} , (3.179)
HT_IT,Tz} = E(T)T_IT,Tz}.

The conditions (3.179) imply that one has the commutator relation

(3.180)

indicating that H should be a scalar operator in isospin space. If we apply this

to a Hamiltonian describing N neutrons and Z protons with nuclear residual

123
interactions and the Coulomb interactions among protons, the Hamiltonian is
written

(3.181)

1
L L -ztz(i). LlmLl + '2 L
A li2 A li2 A
H ~- 2m Ll + V(i,j)
i<k=1 i=1 m i,i=1
A 2 A 2
+ L
i<k=1
4Iroe_rol-
• J
L ; (tz(i)+tz(j))/lri-ril
i<k=1
A
+ L e2 t z(i)t z(j)/l r i - ril· (3.182)
i<k=1

In the second expression m = (m n + m p )/2 and Llm is the difference Llm =

Im n - ml = Imp - mi. Here one observes that the Hamilton operator is not
a scalar in isospin space but contains tenns that are the T z = 0 component of
an isospin vector (rank 1) and of an isospin tensor (rank 2), indicating that the
actual many nucleon Hamiltonian is not fully charge independent. In shorthand,

(3.183)

in rank (2) (1) (0) with Tz = 0 in each case.

From this general Hamiltonian, an energy expression can easily be obtained
as

(T, TzIHIT, Tz) = (-~z ~ Jz) (TIIH(O)IIT)

+ (T 1 T) (TIIH(1)IIT)
-Tz 0 Tz

+ (T-Tz
2
0 Tz
T) (TIIH(2)IIT) . (3.184)

Here, the first tenn gives a constant value for given T, and expresses the charge
independent part of H. The second tenn (coming from the mass difference in
the kinetic energies of protons and neutrons and from the Coulomb tenn) in-
duces a linear dependence on T z , and the third tenn (Coulomb effect) introduces
a quadratic Tz dependence. The total dependence is shown schematically in
Fig. 3.48 for (T, T z ).

124
 Fig.3.48. Variation of the energy E(T, T z ) ==
(T, Tz IHIT, Tz) with Tz (= (N - Z)/2). The
different contributions from (3.184) are illus-
ttated: the constant energy coming from the
Ifo°) pan (isoscalar contribution), the linear
term coming from the Ifol)(isovector) term and
the quadratic tenn coming from the Ilf)
(ten-
sor of rank 2) term. The dependence on Tz (for
a given 1') values is given by

(-~ ~~) (-~ ~ ~)

- - TZ -
T
(-Tz 2
0 Tz
T) , respectively

b) Isospin Wave Functions

We first construct the two-nucleon isospin wave functions by using

Cf'n(r) = '~(:2 .
Cf'(r)
(3.185)
Cf'p(r) = ,~2/2 . Cf'(r) ,
with
1/2
,:!2 the isospin spinors fonnally corresponding to the spin eigenvectors
Xm ••
For the two-nucleon isospin eigenvectors, we construct

, (! !; TT%) = L (!t z, !t~ITTz},:!2(1)(:t:(2) , (3.186)

t%,t~

or
,(!!; T = 1, T z =+1) = '~(:2(1)(~{:2(2),
,(!!; T = 1, Tz = -1) = ,~2/2(1)(~2/2(2) ,
I' (1 1. - 1 T - 0) -_ v'2'+1/2(1)C
1 1/2 1/2 (2)
.. 2 2' T - , z - I/2

1 1/2 1/2 (3.187)

+ v'2CI/2(1)(+1/2(2) ,
I' (1 1.
.. 2 2' T -
_ 0, T _ 0) _ 1 1/2
% - -
1/2 (2)
v'2'+1/2(1)C 1/2

1 1/2 1/2
- v'2 C 112(1)(+1 12(2) ,
written explicitly. The first gives a two-neutron, the second a two-proton wave
function. The other two are linear combinations for a proton-neutron wave func-
tion, however, with a specific symmetry for the interchange of the quantum
numbers of the nucleons: T = 1 symmetric and T =0 antisymmetric.

125
 Consider now the more general case of a proton moving in the orbital (j a, ma)
and a neutron moving in the orbital (jb, mb). The wave function describing this
particular situation reads

tPpn(jaib; JM) = L (jama,ibmbIJM)CPjAmA(rp)CPjbmb(rn). (3.188)

In the nuclear potential (in particular for light and for medium heavy nuclei)
the situation with a proton moving in the orbital (jb, mb) and a neutron in the
orbital (ja, ma) is almost degenerate with the fonner case. This tells that the two
configurations,
tPpn(jaib; JM) ,
(3.189)
tPnp(jaib; JM) ,
are almost degenerate in energy. If we now diagonalize the residual proton-
neutron interaction Vpn , the degeneracy in (3.189) will be lifted and we get two
states with given symmetry tP~n(jaib; JM)

tP~n(jaib; JM) =N L (jama,ibmbI JM )

(3.190)

and we can also evaluate the energy shift between the states with different sym-
metry LlEj (Fig. 3.49). One calls the low-lying state with the spatial symmetric
wave function the T < state (lower isospin; T = 0 for a two-particle system) and
the high-lying state with the spatial antisymmetric wave function the T> state
(upper isospin; T = 1 for a two-particle system). Using now the two-particle
isospin wave functions of (3.187), one can rewrite (3.190) as

t AS

,
I

\
\
\
\
\
t
S
Fig. 3.49. lllustration of the mechanism of the proton-neutron interaction establishing the fonnation
of two states with a specific spatial symmetry character. Starting from the two unperturbed con-
figurations 1/Jpn(j"ib; J M) and t/Jnp(j"ib; J M) [see (3.189)], the proton-neutron force generates a
symmetric (S) and antisymmetric (AS) state [see (3.190)]. The lower, symmetric state corresponds
to the isospin T < (lower isospin, T = 0 for two-particle state) and the upper, anti symmetric state
corresponds to the isospin T> (upper isospin, T = 1 for two-particle state). The energy separation
is denoted by JlFff>

126
 (3.191)

If we now exchange the coordinates of particles 1 and 2 in (3.191), the physical

content of the wave function remains identical. Thus we get extra information that
can be removed by making linear combinations of (3.191) and (3.191) with 1 ¢:} 2
interchanged. We make linear combinations such that the final wave functions
obey a generalized exclusion principle: antisymmetry with respect to all (spatial,
spin, isospin) coordinates. Remember that this is not an extra assumption but
just a convenient formalism that allows us to handle n - n, p - p and n - p
systems consistently. The above combinations now are constructed as (using
a == ja,m a, ... )

tP;n(jaj,,; JM) = ~~) .. ·I··.}

X (<Pa(1)<p,,(2) + <Pa(2)<p,,(I»)((T =0, T", =0),
tP;n(jaj,,; JM) = ~~) .. ·I···}
X (<Pa(1)<p,,(2) - <Pa(2)<p,,(1»)((T = 1, T", = 0) , (3.192)

tPpp(jaj,,; JM) = ~~) ... I... }

X (<Pa(1)<p,,(2) - <Pa(2)<p,,(1»)((T = 1, T", = -1) ,
tPnn(jaj,,; JM) = ~ L(·· ·I···}
X (<Pa(1)<p,,(2) - <Pa(2)<p,,(1») ((T = 1, T", = 1) .
In the above, we have either a symmetric spatial-spin function (S) and an an-
tisymmetric isospin function (T = 0, T", = 0) or an antisymmetric spatial-spin
function (AS) and a symmetric isospin function (T = 1, T",) (see Fig. 3.50 for

nn n p p p Fig. 3.50. Schematic illustration of pos-

+t---+t---+t
sible combinations for the two-nucleon
configurations nn, pp, np. The nn
and pp only occur for S = 0, T =
Td.S,O 1 (taking a symmetric orbital wave
function, i.e. I = 0 wave). The n - p
(deuteron) occurs in both the T = 1,
S = 0 and T = 0, S = 1 (lowest
state) configurations

127
application to the deuteron). Note once again that we could have studied nuclei
without putting the generalized Pauli principle with isospin but just by using the
charge quantum numbers.
The outcome constructing all basis configurations according to the proton-
neutron valence numbers, is, however, in line with the isospin results.
To illustrate, we take the example of a light nucleus ~~SC21 where we consider
possible proton-neutron configurations (Fig. 3.51). Besides the 117 /2(P) 2P3/2(n)
basis state, we shall also have to take the Ih/2(n) 2P3/2(p) basis state into
account. In this small model space diagonalizing the nucleon-nucleon force, one
ends up with two classes of states: one set with a symmetric wave function in the
interchange of the charge coordinates, (T = 1) states that are high in energy, and
another set with antisymmetric wave functions (T = 0) that are low in energy.
Thus, if c1h/2(P) = c1h/2(n), C2P3/2(P) = c2p3/2(n) and all matrix elements are
exactly charge independent, the coefficients in the S and AS combination will
be equal to 1/../2, thus again forming the isospin structure. Slight differences in
the above conditions will induce some isospin mixing and thus, depending on
what one likes charge-quantum numbers or isospin constitutes a convenient basis
for obtaining the same physics.
If we now move to heavy nuclei with a neutron excess, the underlying
principles in constructing isospin cannot be used easily since configurations
'l/Jpn(jaib; JM) and 'l/Jnp(jaib; JM) are not at all degenerate any more: it costs a

0) IT v
2P3/2
---@)
•
• 1f 712

Fig. 3.51a, b. llIustration how, for a nucleus like

b) = =
~iSC2!. the isospin T 0 and T 1 states can be
realized for the 1h /2 21'3/2 configuration. (a) In
,r----- the upper part we show the two possible, almost
I T=1 degenerate proton-neutron configurations 1rlh /2
I V2P3/2 (e) and vlh/2 1r21'3/2 (0) configurations.
/ (b) In the lower part, we indicate how the two,
1f7l2(nI2P3/2(PJ!
nearly degenerate configurations form specific lin-
1f7/2 (pI 2P3/2(~\
ear combinations when diagonaiizing Vp,n. The
,, lowest state, the T = 0 state, gives a symmetric
=
spatial state, the upper one, the T 1 state, corre-
,---- sponds to an antisymmetric spatial state
T=O

128
 1t v Fig.3.S2. Construction, like in Fig.3.5t, for proton-neutron
configurations, but now for heavy nuclei with a large neutron
excess (a). In the upper pan, we have the 1rj vj' configu-
ration. Due to the neutron excess, the vj 1rj' configuration
does not exist, instead we have to make a core excited state
(vj)-l (vj') 1rj'. The latter state occurs ata much higher ex-
citation energy compared with the former one. In the lower
pan (b) we illustrate these two states corresponding with a
large energy difference ~Eco.e. The residual interaction Vp ...
will only induce minor admixtures. Isospin is not needed to
be introduced

01

,- ... ----
f j'(p) r'lnlj'l;'
I
I
: aE core
I
I
t_-,--_
bl j(p) J'lnl "'~'-_ __

lot of energy to create the "exchanged" state and isospin can be heavily broken
for low-lying states (Fig. 3.52). Here, one should better use a charge quantum
number formalism. It is, however, possible to define certain states that correspond
to a good isospin, i.e., configurations with two valence protons in orbitals that
correspond to filled orbitals in the neutron excess core, configurations with two
valence neutrons above the neutron excess and configurations with one valence
proton (corresponding to a neutron excess orbital) and one valence neutron above
the neutron excess (Fig. 3.53). The above states correspond to states with isospin
obtained by coupling the isospin of the neutron excess system Teare = (N - Z)j2,
to the isospin T = 1 of the two-particle system. In both cases, the isospin is max-
imal and corresponds to the projection T z or T =T z I
We shall illustrate the analogies and differences between a proton-neutron
formalism versus isospin formalism in Chap. 5 for particle-hole excitations in
16 0.
To finish the present discussion, we consider the case of two-nucleons (n -
n, p - p or p - n) in a single j-shell when constructing the wave functions
1/J(j2; J M, TTz ). Using the same method as in Sect. 3.2, we construct
1/J(i; JM,TTz) = N'(l- (_1)2 j -J+I-T)
x L (jm,jm'IJM) L {!t z, !t~, ITTz}
m,m' tz,t~

129
 ISO SPIN WI TH NEUTRON EXCESS

Tt v Tt v Tt v
••

Q) b) c)
Fig. 3.S3. States with fixed isospin T = Tz for nuclei with a neutron excess. In both case (a), (b) and
(c), it is impossible to change a proton into a neutron (via the T _ operator) since the correspond-
ing neutron orbitals are fully occupied in the core. So, states with maximal isospin result with T =Tz

x CPi m (1)CPi m/ (2)(!!2(1)(!{2(2) . (3.193)

These functions disappear except when J + T = odd, or
T = 1, J = even = 0, 2, ... , 2j - 1,
(3.194)
T = 0, J = odd = 1, 3, ... , 2j ,
with, as an example, the (117/2)2 (T = 1, J = 0,2,4,6) and (T = 0, J = 1,3,5,7)
configurations.

3.4.3 Two-Body Matrix Elements with Isospin

Using the same methods as discussed in Sect. 3.2.3, one can evaluate two-body
interaction matrix elements, now including isospin, with the result (using l+s = j
coupling)

r
(jIh;JM, TTz lVlhj4;JM, TTz ) = -~T)3t323334.
[( 1 + h"ith) (1 + h"i3i4) I/2
4)
(1 + (_1)lt+ 12+13+1 /2.

[(_1)it+ 12+I / 2 (ji ~2! ~) (_1)i3+ 14+I / 2 (j& ~) ~)

(1 - (_I)J+T+13+14) /2+ (1 +(_I)T) j2( _1)il+h
( ~I2 jl2 J)
-1
(_I)i3+i4 (jl ~4
2 2
J)].
-1
(3.195)

By using the above expressions we can also evaluate the matrix elements of the
interaction -47rvoh"(rl - r2)(1 + aUI • U2) by the concordance

130
 4°) =(1 +a)Ao, (3.196)
41) =(1 - 3a)Ao ,
with
Ao = Yo J Unll l (r)Un212(r)Unala(r)Un414(r) :2 dr •
This is true because we have S = 0, T = 1 and S = 1, T = 0 since the 6-
function only allows the spatially symmetric (L =even) wave functions to give
non-vanishing results.
One can verify that in the special case of a =0, jt =h, h = j4, jt:f h, the
result of Sect. 3.2.3 is reproduced.
Another case is for all ji = j with

(p; JM,TTzIVIP; JM,TTz) = -4T)(2j + 1)2~ [( 1!! ~Y

X(I-(_I)J+T)/2+(} i !lY(1+(-1)T)/2] (3.197)

In studying the above example for (1 h /2)~=O states, we obtain the spectrum
shown in Fig. 3.54. Here we see that the interaction is most attractive in the
parallel and anti-parallel angular momenta for T = 0 states. We can, more-
over, compare the above theoretical results with the 1~Sc nucleus where indeed,
the Ih/2(P) Ih/2(n) case occurs. In Sect.3.2.3, the results for T = 0 for a
6-interaction with large j are illustrated and do compare well with experimental
data on odd-odd nuclei.
In Fig. 3.55 we compare the low-lying spectra for ~Ca22 with 1~SC2t, and
observe that although many more states occur in 42SC, for the (1h/2)~=t; J=O,2,4,6
states, very similar results occur because of charge independence.

2 --1--
(1 f 7,,' 357
~ 0.0- ---.......!I~-
~
T=O
~-02 - 5+ -0.235
uJ 3+ -0.301
Z
uJ -04 _ 7+ -0.408
(!)
z
5 1+ -0524
~-0.6 -

~ -0.8
~--------------------~------------~
Fig. 3.54. Isospin T =0 two-body matrix elements for two particles in the 1h /2 orbital. The matrix
elements are expressed as LlEJ == (j2; J M, T = 0IVt,21j2; J M, T = 0) /4~O). On the right-hand
side, a corresponding LlEJ vs. J plot has been made

131
 42Ca 42SC 42Ti
20 22 21 21 22 20

3189
6' 3.043
6'
3.0
2752
4' 4+ 2676

->.
~
>-
Cl
ffi 2.0
z
w
z 1.555
Q 1525
I- 2'
<
I-
i3
x
w
10

7+ 0617
1+ 0.611

0+ _ _ _ __ 0+ _ _ _ __
0.0

Fig.3.55. Comparison of the data in 42ea. 42SC and 42Ti. We compare the T = I and T = 0
(I h /2)2 states. The conditions of charge independence are well followed when comparing 42ea and
42Ti. In 42SC• the extra states I+, r . (3+), (5+) are also given. In the case of 42ea and 42Ti. some
low-lying levels (0+, :Z+) have been left out since they do not correspond to the (Ih /2)2 configuration

The above particle-particle spectra in odd-odd nuclei are remarkably consis-

tent with a parabolic behavior. This works throughout the whole nuclear mass re-
gion (In nuclei, N = 81 and N = 83 nuclei) and has been known as the parabolic
rule (paar 1979, Van Maldeghem, Heyde 1985, Van Maldeghem 1988). Although
somewhat outside the scope of the present discussion, it has been shown that such
a parabolic dependence on spin J can be derived (Van Maldeghem 1988):

LlE(J) = -271" 9(p)9(n)( _l)ip+in+J { ~p ~n J}

5 In JP 2
x (jpllY 2I1jp}(jnIlY 2l1jn) , (3.198)
with 9(w) = u 2 (w) - v2 (w) and v2 (w), the occupation probability of the orbital
w (see Chap.7 for a discussion on how to determine the orbit occupation prob-
abilities). Writing out the Wigner 6j-symbol in explicit form as well as the Y 2
reduced matrix elements, one obtains for the "parabolic" rule
LlE(J) =- i{ [J(J + 1) - jp(jp + 1) - jn(jn + 1)]

+ [J(J + 1) - jp(jp + 1) - jn(jn + l)t}

132
 1.0 Fig.3.S6. Schematic illustration of the
["11"(199/2)-1 v(lh u /2)] proton-neutron
Ttlg~ V lh l1/2 y2
Ihl1/2 multiplet members, according to the para-
bolic rule of (3.199). It is the pairing fac-

!..
tor (see Chap.7) that causes a tum-over
of the parabola from convex to concave
0.5 in shape. Here, the notation "lI"(v) is used
for proton (neutron) orbitals
,.,......--- ....... ......
> ......
"-
~
> 0.75 \
"-
Cl
~ \
w
z
w 0
~
Cl
z \
0 \
z
in ""- /
...... ....... ___ -//025

-OS

-1.0 ,--::~~~~=--~~,--~_~_--'
026122030 42 56 72 90 110
-J(]+lI-

(3.199)
This expression clearly indicates a quadratic dependence of LlE(J) on the angular
momentum combination J(J + 1) of the proton-neutron multiplet members.
Examples for the Ig9;i(7r) x Ih9 / 2 (v) multiplet are shown first schematically
(Fig. 3.56) and then for the case of the odd-odd In nuclei (Fig. 3.57). For the
proton hole state one has Vf99/2 = 1, whereas for the neutron, a partial occupation
is included (Vfh ll / 2 f 0,1) and points towards using a neutron quasi-particle
excitation (see Chap. 7 for more details).
In this chapter we have learned how to study nuclei where the number of
nucleons outside closed shells is not too large to allow for an exact shell model
treatment. We have discussed methods to calculate the matrix elements needed to
set up the secular equation for the energy eigenvalues both for identical and non-
identical nucleons. We have also shown, once the wave functions are known, how
to calculate one-body expectation values that determine the nuclear observables
(half-lives of nuclear excited states, decay rates, moments, etc.). In particular, the
evaluation of these observables for electric and magnetic transitions and moments
will be outlined in Chap. 4.
Here we summarize the above procedure and philosophy to study a given
nucleus (N, Z).

133
 (bl
Experiment Theory

>a;
:::E

~
w
CD 0.4
~
15
z
co
0.2

Fig.3.57. (a) The experimental splitting of the [1r(I!19/Z)-lv(lh u /z)] multiplet members relative
to the 10- level as deduced from the level schemes in odd-odd In nuclei. (b) The same multiplet
but now for the calculated spectra using a 6·force [taken from (Van Maldeghem 1985)]

i) One detennines the nearby closed shells given by Zcl and Ncl so as to fix
the number of valence protons and neutrons.
ii) The number of active particles becomes np = (Z - Zcl) and nn = (N - Ncl)
where np and nn are the valence number of proton (neutron) particles (or
holes). Through (i) one can fix the single-particle orbitals i P ll i P2' ... and
inll in2' ... that detennine the properties at low energy when constructing
the model space.
iii) One constructs the model space and the configurations that span the space for
each J1r value. The basis configurations are denoted by I(jPliP2 '" iPnf p
Jp , (jnli n2 ... inn) nn I n; J M} which is a shorthand notation for construct-
ing the proton np particle state Jp multiplied to the neutron nn particle state
I n , both coupled to total spin J1r. This basis has n(J1r) basis configurations.
iv) Starting from the single-particle energies Cjpi ,Cjn. and the two-body matrix
elements for identical and non-identical nucleons: one builds up the energy
matrix [H] detennined by its elements Hij and diagonalizes the n x n energy
matrix. Thus one obtains the n energy eigenvalues and n corresponding
eigenfunctions.
v) With the wave functions tPi(J;i) and tP,(J;'), we calculate all possible
observables to compare with the data and so, in retrospect, can detennine
the wave functions and the basic input quantities that appear via point iv).
This method (i-v) sets a standard shell model calculation that is able to
give a good description of a large body of nuclear observables. When many

134
valence nucleons are present outside closed shells, we shall have to construct
good approximation schemes to the more general shell model calculations that
readily become unfeasible. Such approximation schemes shall be discussed in
Chap. 6 and Chap. 7 in detail.

135
4. Electromagnetic Properties in the Shell Model

4.1 General

Having studied the nuclear wave functions obtained from the nuclear secular
equation, one has a first test on a good description of the nuclear Hamiltonian
(average field, residual two-body interactions, ...). The study of nuclear decay
rates via gamma decay is a much better test, however, of the nuclear wave
functions obtained. Moreover, nuclear gamma-decay properties are some of the
best indicators of nuclear spin and parity (J1r) of excited states. In this chapter
we shall concentrate on the evaluation of transition rates and static moments.
We do not, however, discuss a derivation of the electric and magnetic multipole
operators themselves and refer the reader to (Brussaard, Glaudemans 1977) for
an efficient introduction.
Transition rates per unit time are given by
T(L) = 87rce2 Inc(L + 1)1 (L[(2L + 1)!!]2) k2L+1B(L) , (4.1)

where k = wi c ~ 11 R with R the nuclear radius, which means that we evaluate

transition rates in the long wavelength limit (the excitation energy of E"{ <
3 MeV). Here, L describes the multipolarity and B(L) the reduced transition
probability which actually carries the nuclear structure information.
If we start from an initial state lai; JiMi} going to a final state laf; JfMf}
with the operator O(LM), then the reduced transition probability means summing
for a given initial state (JiMi) over all final states (JfMf) and intermediate
values of the transition operator (L, M), i.e.,

B(Ji -? Jf;L) = L I(af; JfMfIO(LM)lai; JiMi} 12 , (4.2)

M,M I

and, when we apply the Wigner-Eckart theorem, this reduces to

(4.3)

when no particular polarization of the initial orientation (JiMi) is implied. If the

population of the initial substates Mi is dependent on Mi and given by certain
occupation numbers p(Mi), then (4.3) has to be modified.
The above arguments hold for pure states. If on the other hand, the initial
and final state are expanded in a basis such that

136
 lai;JiMi} = Lak(Ji)lki;JiMi},
k
(4.4)
laf;JfMf } = Lb,(Jf)llf;JfMf ),
,
then the reduced transition probability in (4.3) is generalized into the expression

B(Ji -+ Jf;L) = 2J.1+ 11 Lak(Ji)b,(Jf)(l; JfIl O (L)lIk; Ji)12. (4.5)

I k,'
In the latter, interference effects can result. In some cases (collective tran-
sitions), all partial contributions in the sum over (k, I) have the same phase,
resulting in enhanced transition rates.

4.2 Electric and Magnetic Multipole Operators

In the long wavelength approximation, the electric multipole operators are defined
for a continuous distribution of charge u(r) as (Brussaard, Glaudemans 1977)

O(el.,LM) =~ J u(r)rLyLM«(),<p)dr. (4.6)

Similarly, the magnetic mUltipole operator which is resulting from the convection
current j(r) and the magnetization current m(r), can be written as

O(mag.,LM) = 1/(L+ 1)I/ec JV [rLyt«(),<P)]

x {r X [j(r)+cV x m(r)1}dr. (4.7)
For a system of point charges with given values ei moving with momenta Pi
and intrinsic spin Si, the charge and current densities can be written as (Brussaard,
Glaudemans 1977)

u(r) =L eiDer - ri) ,

i

j(r) = Lei/mpiD(r - ri) , (4.8)

i

m(r) = Liiisic5(r - ri) .

i

This situation conforms to a nucleus consisting of A point nucleons (Z pro-

tons, N neutrons) characterized by their charge ei and magnetization moment iii.
Here, already, we anticipate effective charges and moments that can differ from
the free charges (e,O) for proton and neutron and (/-Lp, /-Ln) the free gyromagnetic
moments.
Using the general expression for the reduced transition probability in (4.3)
together with the electric and magnetic multipole operators where a system of

137
point particles is considered, we obtain for the electric and magnetic reduced
transition probabilities the following expressions.

i) Electric Transitions.

and
1
B(el.; Ji ~ Jf; L) = 2Ji + 1

x 1(Jfll L(ei/e)rfYL(Oi,CPi)IIJi)12 , (4.10)

i

where we sum over all A nucleons in the nucleus.

ii) Magnetic Transitions. In Appendix G, we give a derivation where the

magnetic multipole operator from (4.7) can be rewritten in a more convenient
form. For the system of point particles, this becomes

O(mag.,LM) = LVi
.
(rfyLM(Oi,CPi))' ~
e
•
1 eili _]
x [ (L + 1) me Ii + J.LiSi . (4.11)

Using the relation

V [rLy£'f (0, cp)] . V = [L(2L + 1)]1/2 r L-l [YL-l 0 V]~ , (4.12)

where V is a general vector operator. Further on, we shall measure the nuclear
moment iti in units of nuclear magnetons (n.m.) such that iti = (eli/2me)gs(i).
We also introduce the total angular momentum operator j = I + IT /2 = I + S.
Now the reduced magnetic transition probability becomes

B(mag., Ji ~ J f ; L) = 2Ji1+ II(J f lll;:: ~erf-l

x {-
e L+l
1[ '2 s
. 1(• .
ei - - . YL-l 0 J. ](L) + - g (z) - -ei . - -
e L+l
1)
x [YL_10Si](L)}(L(2L+1)i/2I1Ji)r. (4.13)

138
4.3 Single-particle Estimates and Examples

IT now the wave functions lai; JiMi} and lal; JIM/} are the single-particle
wave functions constructed in Sect. 3.1, the reduced transition probabilities
(4.10,13) become the single-particle estimates. This means also that the sum
will reduce to a single term: the remaining extra nucleon.

r(;Y ,
i) Electric Transitions. Starting from the expression

I
Bs.p. (el.,ji -. iJ; L) = 2ji1+ 1 (nl1liJllrLYLllni1di) (4.14)

and the value of the YL reduced matrix element as evaluated in Appendix F, the
above expression reduces to the closed form
Bs.p• (el.,ji -. iJ; L) = (2iJ + 1 )(211 + 1 )(21i + 1 )(2L + 1)/411"

X ( ~)2
e
.(110 L
0
li)2 {II
0 h
if
Ii
!} (L}2
L r ,
(4.15)

or, even more compact,

Bs.p.(el.,h -. iJ; L) = (2iJ + 1)(2L+ 1)/411"(e/e)2
x (~ ~~ ~y (rL}2. (4.16)

The radial integrals have been defined as

(rL) = J Un/l/r)rLUn;I;(r)dr . (4.17)

ii) Magnetic Transitions. Similarly, (4.13) reduces to the closed form

Bs.p.(mag.,ji -. iJ;L) = (h/mc)2(rL-l}2(I/IlYL_lllli}2
x (2it + 1)(2L + I)L [< _1);,+lJ +1/2 L: 1J;Vj;(j; + 1)

II
x { ji
iJ !} {iJ1 L ji} e f3
-; + y 2.

{I
Ii L - I ji L- 1
2I .JI
1
eLI } 2
x (gaL - - - )
eL+l
Ii
L-l
!
1 L
ji (4.18)

For free nucleons we have to use in these single-particle estimates,

ep = e, en = 0; ga(P) = 5.58, g.. (n) = -3.82 .

139
Table 4.1. The conversion factors used to relate the reduced EL and M L transition probabilities
(expressed in (fm)2L and (ti/2mc)2(fm)2L-2, respectively) to the total transition probabilities (ex-
pressed in units S-l). Here E-y is expressed in units of MeV

T(E1) = 1.59 X 1015 (E-y)3. B(E1)

T(E2) = 1.22 x 109 (E-y)5. B(E2)
T(E3) = 5.67 x 1()2 (E-y)7. B(E3)
T(E4) = 1.69 x 10-4 (E-y)9 . B(E4)
T(M1) = 1.76 x 1013 (E-y)3 . B(M1)
T(M2) = 1.35 x 107 (E-y)5. B(M2)
T(M3) =6.28 x 10° (E-y)7. B(M3)
T(M4) = 1.87 x 10- 6 (E-y)9 . B(M4)

Equations (4.16, 18) are single-particle estimates in the best way since the angular
momenta j i, j f still result in the final expressions. Later we shall try to define
estimates for the single-particle transition rates, values that will only depend on
the multipolarity L and no longer on ji, j f. Thus, these (Weisskopt) estimates
can be used as an easy measure of transition rates (Segre 1977).
First, we give in Table 4.1 the conversion factors between the reduced and
total transition rates. We express T(EL) and T(ML) in S-I, E"{ in MeV, B(EL)
in units (fm)2L and B(M L) in units (fi/2mcf . (fm)2L-2.
We now come to the Weisskopf estimate and point out the different restric-
tions that have to be imposed on (4.16,18) in order to remove the ji and it
dependence in them.
i) We use constant radial wave functions when calculating the radial integrals
(Fig. 4.1)
R(r) = C for O<r<R,
(4.19)
=0 for r ~ R,

and from the normalization condition, we find

C2 1R r2 dr =1 or C = (~3 )1/2 (4.20)

Now the radial integrals become

Fig.4.1. Representation of the approxima-

tions made on the radial wave function
--r_ Rnl(r) (constant value over the nuclear
interior), relative to a more realistic radial
dependence, when deriving the Weisskopf
estimate

140
 L [R L+2 ( 3 ) 3 L
(r ) = 10 r R3 dr = (L + 3) R . (4.21)

ii) Next we calculate the values for Bs.p.(el) and Bs.p.(mag) for the transitions
ji = L + !, it !.
= Under these restrictions, the single-particle transition rates
simplify significantly and we get

B (el. L)
s.p.,
= 2(2L471'+1) (L !+! !
-!
L)2 (_3_)2 R?-L
0 L +3 '
(4.22)
2 2
or,

W (L)
E
= Bs.p.·,
(el L) = (1.2)2L
471'
. (_3_)2 . A 2L / 3(fm)2L
L+3 .
(4.23)

For the magnetic transition rate, we note that the Wigner 6j-symbol

{II '
JI
i; Ii
I} =
2
L- 1
0
'

for the case that ji = L +! !.

and it = For that case, we have it ji = L I + +
and Ii; - itl = L. Thus only one tenn contributes to the magnetic transition
probability. Moreover, we use the restrictions (de-Shalit, Talmi 1963, Brussaard,
Glaudemans 1977)

e = e, [L (g _~e ._1
8 L+l
)]2 = ~2'
Then, the Weisskopf magnetic estimate reduces to

WM(L) =Bs.p.(mag.,L) = ~(1.2)2L-2 (L:3Y A(2L-2)/3

x (1i/2mc)2(fm)2L-2 . (4.24)

As an example we discuss the case of 170 for the 1/2+ --+ 5/2+ E2 transition.
In (4.23,24), all infonnation on ji and it is left out. When evaluating the
rate for the 1/2+ --+ 5/2+ E2 transition which is mainly a 281/2 --+ Ids/ 2 single-
particle transition (Fig. 4.2) in more detail, we use the more detailed expression
(4.16) and get

B (E2; 1/2+ --+ 5/2+) = 5.6 (~)2 (t il 2)2 (r2)2 ,

471' e 2 -2 0

= 2.
471'
(~)2
e
(r2)2 . (4.25)

Using a better estimate than the constant value of R(r), by using hannonic
oscillator wave functions, the radial integral can be evaluated to be

141
 Fig. 4.2. The E2 gamma transition in l"IO. deexciting the l/z+
0.87
level (mainly 281/2) into the ground state S/z+ level (mainly
Ids/2 ) in a single-neutron approximation

(r2) 1
= 00
r2U2s1/2(r)Uld5/2(r)dr ~ 12fm2 . (4.26)

Combining the results from (4.25,26), the E2 transition rate finally becomes

(4.27)

The experimental result is 6.3 fm4 such that an effective charge for the single-
neutron transition becomes e 9:' 0.43 e, a value which is quite different from the
free neutron charge en (free) = Of We would like to briefly discuss the concept
of an effective charge in the nucleus and the difference with the free nucleon
charges (electric and magnetic charges).
In considering the nucleus as an A-nucleon system, if we solved the corre-
sponding SchrOdinger equation, an A-nucleon wave function tP(l, 2, , ... , A; J M)
would result. The electromagnetic operators would be the sum of A one-body
operators since each of the A nucleons can induce the transition. The related
transition matrix elements would then read
A
(tPf(1,2, ... , A; JfMf)IL:0(LM;i)ltPi(l,2, ... , A; JiMi) . (4.28)
i=1

Working in a small model space, however, we like to reproduce the matrix

elements of (4.28) but now with model wave functions only using a restricted
number of nucleons (only one neutron in the case of 17 0 outside the closed 16 0
core) and the transition operator acting in the model space only. For a reduced
n-particle model space we have the matrix element
n
(tPr(1,2, ... , n; JfMf) I L:oefl'(LM;i)ltPi"(l,2, ... , n; JiMi)}. (4.29)
i=1

By equating the two matrix elements (4.28,29) where in (4.28) the free nucleon
charges ep = e, en = 0, ... occur, an implicit equation for the effective charges
in the model operator oefl'(LM; i) results once the model wave functions tP M
and the full A nucleon wave functions tP have been determined. This process
is illustrated in Fig.4.3. For the extreme case of 170, as discussed above, we

142
 Tt v Tt v

91,9 5 (free) 91,9 5 (effec live)

ep.e n Ifree) ep.enleffective)

FULL SPACE MODEL SPACE

Fig. 4.3. Illustration of the concept of effective charge and effective gyromagnetic factors
(el" en. g" g.) depending on the model space used. On the left-hand side, besides a core, many
valence protons and neutrons determine the nuclear low-lying properties with charges and gyromag-
netic factors almost the free nucleon values (if no core would be present). On the right-hand side,
the same physical situation is given but now, relative to a new reference state. In this case, only a
few particle-hole excitations determine the low-lying properties. As a draw-back, effective charges
and effective gyromagnetic factors have to be used, values that can differ much from the free values

go from a 17-nucleon problem (wave function and operators) to a I-neutron

problem. Therefore, the degrees of freedom of the 17-nucleon system have to be
"absorbed" in some way into the electromagnetic properties of the valence neu-
tron. This concept defines model operators and effective "charges" (electric and
magnetic) that are very much dependent on the model space. When the model
space extends towards the full space, we have to go over to the free "charges"
again. Thus, the charge for the E2 transition in 170 for the 1/2+ - t 5/2+ tran-
sition of en '::' 0.43 e reflects the neglect of the 16 0 core with 8 protons and 8
neutrons. We show this in Fig. 4.4. Experimental results for electric quadrupole
moments, E2 transitions and for magnetic dipole moments now have to be used
in deducing the neutron (proton) effective charges and magnetic moments.
One observes an important "state" (nlj) dependence of the effective "charges",
even for all cases where only a single-nucleon model is used to explain the ex-
perimental electromagnetic properties (Tables 4.2 and 4.3).

Fig.4.4. Single-particle picture of prop-

erties in 17 0 described by a single neu-
tron outside a 16 0 core and a 17 particle
(8 protons + 9 neutrons) picture of 17 0 in
order to describe the properties of 170.
The last neutron does not present a very
particular place in the latter picture

143
Table 4.2. Effective, electric charges as deduced from known electric quadrupole moments (a) and
E2-transition probabilities (b). In both cases, we concentrate on nuclear configurations that approach
single-particle (or single-hole) configurations as close as possible. Therefore. we select doubly.dosed
shell nuclei (:1::1 nucleon). Besides the data and the single-particle value. the effective charge (in units
e) is given. For detailed references on the data, see (Bohr, Mottelson 1969), from which the table
has been taken
(a) Quadrupole moments

Nucleus Ij Qobs(10-24 cm2) Qsp(10-24 cm2) eeff/e

1~O d S/2 -0.026 -0.066 0.40

~;K d- 1 0.09 -0.052 1.8

3/2
~Bi ~/2 -0.4 -0.26 1.6

(b) E2-ttansition probabilities

Nucleus (lj); (Ij)! B(E2)obs(e2fm4) B(E2).p(e2 fm4) eefI/e

1~N -1 -1 7.4 4.6 1.3
P3/2 P1/2
1~O 81 / 2 dS/2 6.3 35 0.42
1~F 81 /2 d S/2 64 43 1.2
gea P3/2 17/2 66 40 1.3
i~Sc P3/2 17/2 110 40 1.7
-1
~Pb fS/~ P1/2 70 81 0.9
-1 -1
~Pb P3/2 P1/2 80 110 0.85

~Pb 81 /2 dS/2 150 866 0.42

~Bi 17/2 ~/2 4O±2O 2.3 4 ± 1.5

In discussing electromagnetic transitions, the motion of a proton or neutron

is associated with a recoil of the rest of the nucleus since the center of mass
remains at rest. This effect is of particular importance in the study of electric
dipole (El) transitions. Here, a one-particle effective charge results, expressed
by the quantities

ep =(l - Z / A) . e , en = -(Z/ A) . e .
A detailed discussion of the derivation of these effective charges can be found
in Eisenberg, Greiner 1970, Ring, Schuck 1980.

144
Table 4.3. Effective. gyromagnetic factors. illustrated by a comparison of the observed and single-
panicle (using free g. and g, factors) magnetic dipole moments. For detailed references on the data.
see (Bohr. Mottelson 1969). from which the table has been taken

Nucleus Ij J.Iobs ,",sp

H -I
8 1/2 2.98 2.79
3He -I
8 1/2 -2.l3 -1.91
ISN -I -0.28 -0.26
Pl/2
ISO -I 0.72 0.64
PI/2
170 dSI2 -1.89 -1.91
17F dSI2 4.72 4.79
39K d- I 0.39 0.12
3/2
41Ca
17/2 -159 -1.91
sSCo
f7/~ 4.3 ±0.3 5.79
207Pb -I 0.59 0.64
PI/2
297Pb /.-1 0.65 ±0.05 1.37
SI2
209Bi ~/2 4.08 2.62

4.4 Electromagnetic Transitions in 1\vo-particle Systems

In discussing two-particle systems, we use the general tensor reduction expres-

sions discussed in Chap.2 and Sect. 3.3. Using the E2 transition operator in
shorthand notation

L j(2)(k) ,
2
F(2) == (4.30)
k=1

f
with 2)(k), the E2 operator of (4.9), we can derive the two-particle matrix
element

This more general expression, in the situation of a 2+ -+ 0+ E2 transition becomes

{l; J =0IlF(2) Ill; J' =2} =2(-1)2i v'5 {~ ~ ~} {jllf2)lIj}

=2(_1)2iv'5(_1)2i_1 1 {jllf2)lIj}
v'5 J1J+T '
= ~{jllf2)lIj} . (4.32)

145
We now have detennined the E2 reduced matrix element for single-particle states
to be

(j1lt<2)lIj) = (r 2 )(2j + 1) /5~ ({2

Y4; e
jl
-2
2) (_1)i- 1 / 2
0
, (4.33)

which becomes, when the explicit fonn of the Wigner 3j-symbol is inserted,
(j1lt<2)lIj)
_ 2 '+1 (5~ 3/4-j(j+1)
- {r )(2) )y 4; e (j(j + 1)(2j _ 1)(2j + 1)(2j + 3»1/2 . (4.34)

In this way, the reduced E2 transition probability for the 2+ --? 0+ transition
reads

or

( .. 2 + .2 +) _ (e)2 {r2)2 [3/4 - j(j + 1)]2

B E2,J (2 ) --? ) (0) - ~ - ; - j(2j _ 1)(2j + 3)(j + 1) . (4.35)

If we take the limiting value for j --? 00, this expression "loses" its specific j
dependence and gives

B(E2;O+ --? 2+).

)->00
= ~.
411'
(r2)2 (~)2
e
(4.36)

If we introduce the assumption of constant radial wave functions when evaluating

the radial integral in (4.36) as calculated before, we finally get

B(E2;0+ --? 2+)s.p. =0.30A4/ 3 fm 4 . (4.37)

This estimate is the extension of the Weisskopf estimate for single-particle transi-
tions and can be used in even-even nuclei to "measure" E2 0+ --? 2+ transitions.
The above estimate actually corresponds to (4.23) where we put L = 2 and take
the statistical factor of 5 == (21 f + 1) into account.
Applying the above calculation of (4.35) to the case of 180 where an E2
transition deexcites the 1.98 MeV 2+ level, using nlj = 1ds/2 , we obtain

B(E2;2+ --? 0+) = (r:)2 (;y 3~ , (4.38)

and using the radial integral

(r2) == 1 Uid
00
s / 2 (r)r
2dr ~ 12fm2 , (4.39)

a theoretical value of B(E2; 2+ --? 0+) = 10.5 (e/ e)2 fm4, results. If we consider
the effective charge for the neutron to not change very much in going from 170
to 180, for en = 0.43 e, a value of B(E2; 2+ --? 0+) = 1.9 fm4 results. Comparing

146
with the experimental value B(E2;2+ -+ O+)exp =7.05fm4 , a discrepancy of a
factor of 3 appears.
This can be interpreted in two ways:
i) By using a single (lds/2 )2 pure configuration, a neutron effective charge of
en =0.83 e is needed to reproduce the data. The variation when going from
a one-particle to a two-particle model space is completely within the spirit
of the concept of model or effective charges as discussed in Sect 4.3.
ii) We use the effective charge en = 0.43e but take into account that (ldS / 2 )2
is too strong a restriction in the model space for describing the 0+ and 2+
low-lying states. It was discussed in Sect. 3.2.4 that the Ids/2 and 281/2 con-
figurations will contribute to the 0+ and 2+ wave functions and we should
evaluate the B(E2; 2+ -+ 0+) transitions for configuration mixed wave func-
tions.
Thus for configuration mixing, in calculating the matrix element (J, II F(L) II Ji} ,
the initial and final states are described as linear combinations of basis states with
the same J, and Ji values.
Since we can write

IJiMi}= LCjj' (Ji)ljj'; JiMi} , (4.40)

j,j'
(4.41)
'11 .,,,
1 ,1
the transition matrix element becomes

L Cjjl (Ji)Cjll jill (J, ){j" jill; J,IIF(L)lIjj' ,Ji} . (4.42)

;,;'
j",;"'

Because of the many configurations that can contribute, depending on the c-

coefficients, coherence in the different contributions could result in transition
rates enhanced relative to the Weisskopf estimate (collective E2, E3 transitions
often occur in nuclei with many valence protons and neutrons outside closed
shells).
Making use of the reduction formulas in calculating matrix elements (Chap. 2)
and considering the antisymmetric character of the two-nucleon wave functions

Ijj'; JiMi} = [2(1 +6jjl)r1/ 2 [Ijj'; JiMi} - (_I)j+i'-J'Ij'j; JiMi)] ,

(4.43)
the transition one-body matrix element is evaluated as
(j"jlll; J,IIF(L)lIjj'; J i )
= [(1 +6j'ljlll)(1 +Ojj')r1/ 2 . J, Ji
X {(_I)jll+j"'+J,+L {~; ; j~'} (j"IIt'L)lIj}6j'jlll

147
 + (_l)i'+i"+J/+L {j;; ~{ { } (j"'lIl L)lIj'}oii"

.",
- (_l)Ji+ J/+ L { JJi f
J ."}
JL (j"'lIlL)lIj}°i'i"

+ (_l)i'+i"+L {~; ~{ jZ} (j"lIlL)lIj'}Oii lll • (4.44)

This expression simplifies considerably for the case that j = j' = j" = j"',
Jf = J and Ji = J'. Only transitions
i --. i, i --. jj', j,2 --. jj', jj' --. jj'
are possible. Transitions p --. j'2 cannot occur (for j:f j') since there is only
a one-body operator acting, indicating that only the quantum numbers of one
nucleon can change and the other nucleon acts as a spectator [the Kronecker
delta in (4.44) expresses this mathematically]. In Fig.4.5 a Goldstone diagram
illustrates the possible transitions in a two-particle configuration through this
one-body character (shown by the external line - - - ~).

)1.--- --oK

~--jj
j'

j' J'

)E---

J' Fig.4.5. Feynman-Goldstone diagrammatic

way to describe the electromagnetic transi-
tions, described by (4.43). Possible transi-
i i, i --> jj', jl2 --> jj'

~--j
J' tions are the

f.-_.
-->
and jj' --> jj' types. The fennion states are
depicted by a down-going line. The tran-
J
sition one-body operator which affects the
properties of a single nucleon at most, is
J' depicted by the dashed line and the vertex
(>E- - - )

148
4.5 Quadrupole Moments

As a general definition of the quadrupole moment, we use the expectation value of

(3z 2 - r 2 ), within the angular momentum state IJ M) with the maximal projection
J = M. Since (3z 2 - r2) is proportional to the E2 operator with M component
M = 0, the quadrupole moment is written as

_ . _ ru;; " ei O(Oi,IPi ).

Q = (J,M - Jlys ~ -;ri12 _
IJ,M - J), (4.45)
•
or, using the Wigner-Eckart theorem, as

J 2
Q= ( -J 0
J)
J
ru;;
"ei 2 ( )
Ys(JII~-;riY2 Oi,IPi IIJ)·
•
Filling out the Wigner 3j-symbol explicitly, one obtains the resulting expression
J(2J - 1) )1/2 ru;;
Q=
(
(2J + 1)(2J + 3)(J + 1) Ys
x (JII L. eei r ;Y2(Oi,IPi)IIJ) . (4.46)
•
4.5.1 Single-particle Quadrupole Moment
For the case of a pure single-particle configuration, we need the UII Y 211j) re-
duced matrix element, which becomes (Appendix F)

U11Y21lj) = (2j + 1) [5 3/4 - j(j + 1) . (4.47)

Y4; (j(j + 1)(2j - 1)(2j + 1)(2j + 3) il2

Now, putting the above reduced matrix element into (4.46) where J -+ j, the
final result simplifies very much to

(2j - 1) ~ ( 2) (4.48)
Qs.p.(j) = (2j + 2) e r .

So, for a particle moving outside a closed shell, a negative quadrupole moment
appears. This is because the particle density for a j, m = j orbital is localized
in the equatorial plane, giving rise to an oblate distribution. The absence of a
particle (a hole) (Chap.5) then corresponds to a prolate distribution giving a
positive quadrupole moment These types of motion are illustrated in Fig. 4.6.
As an example we take again 17 0 with a Idsl2 neutron particle moving
outside the 16 0 core in the equatorial plane. Inserting the appropriate spin values
we get
4e 2
Qs.p. ( Idsl2 ) = -7~(r ) . (4.49)

149
 z- axis 2- axis l' Ig. '1.0. ;)wgle-parU(;le lluaucUIJI.lle
moment for a nucleon moving in the
equatorial plane inducing an oblate
mass distribution, relative to the z-
axis (Q < 0). When a nucleon is
missing in the equatorial plane, an
"effective" hole distribution gives rise
to a prolate mass distribution relative
to the z-axis (Q > 0)

Q<O 0>0

:z - aXIs

~--- .........

Fig. 4.7. The dynamic effect of the equatorial nucleon, polariz-

ing the core around which the nucleon evolves, is illustrated.
Thereby, an effective larger quadrupole moment shows up since
the one-nucleon and core quadrupole moments add up and re-
inforce each other (coherence)

Using the value of the radial integral (4.39) and the effective neutron charge
en = 0.43e, a value of Qs.p.(lds/ 2 ) = -2.9fm2 results, a value that is in very
good agreement with the experimental value of Qexp(5/2+) = -2.6fm2. Thus
the E2 transition in 17 0 gives a charge consistent with the single-particle E2-
moment.
Looking back to Figs. 4.6, 7, it is tempting to assume that the odd particle
will "polarize" the core because of the nucleon-nucleon interaction and its short
range, demanding an extra induced oblate core re-shaping. This polarization can
be taken as the source of an extra or effective polarization charge that, for 17 0
becomes en = 0.43 e. In some regions (Fig. 4.8), very large extra charges (large
quadrupole moments) do appear (176Lu, 167Er) and indicate cases where the core
is extremely polarizable and coherence in the odd-nucleon motion is present.
Figure 4.8 shows the change of sign in Q for particle versus hole configurations.
The closed shell configurations 8, 20, 28, 50, 82, 126 are indicated by arrows.

150
 30

I"Lu
25
.OB

20

15

1
Q "'Vb
ZR l o "'HI
'H

"N

0
\

•. •
\

"0 U7AI:

•
-5
1620 28 82
t
126
U1Sb

-10
0 20 40
50
no. of odd nucleons
Fig. 4.8. Experimental reduced nuclear quadrupole moments as a function of the odd-nucleon num-
ber. The quantity Q/ Z ~ gives a measure of the nuclear defonnation independent of the size of the
nucleus [taken from (Segre 1977)]

151
4.5.2 Two-particle Quadrupole Moment
i) For r =0+ states, a trivial quadrupole moment Q =0 results.
ii) For J7r = 2+ states, by using the general expression (4.46) one obtains

Here again, we use the reduction formulas of Chap. 2 to get

(i; J = 2I1 F (2) IIi; J = 2) = 1O(_1)2j {~ ~ ~} Ullf(2)llj) . (4.51)

Combining this result with the single-particle quadrupole moment, one gets

Q(i.2+)=-4
, flO~(r2){j2 J~
VTe j}
2
x (2j + 1) 3/4 - j(j + 1) . (4.52)
(j(j + 1)(2j - 1)(2j + 1)(2j + 3)i/2

For the example of the 2+ level in 180, when considering this state as a (ld5/ 2 )2
pure configuration, we can evaluate the necessary quantities:

and
Q(2+) = 3.9 (e/e)fm2 •

Remember the positive sign of Q(2+) for a 2-particle configuration, in contrast

to the one-particle case with a negative sign. These results can be put into a
more general relationship between the one-particle and two-particle quadrupole
moment:

Q (2+) = _ 10 f2 { j 2 j}
V"35 2 j 2

x (
2j + 1)(2j + 3)(j +
j(2j _ 1)
1»)1/2 Q (j)
s.p.·
(4.53)

(The proof of this relation is an exercise in the problem set.)

Here, the tJ Wigner 6j-symbol has a positive sign and thus Q(2+) and
Qs.p. (j) have opposite signs. Looking at the mass distributions more qualitatively
as we did for the one-particle and one-hole motion, we obtain the result of
Fig. 4.9. Since the two larger momenta j add up to a relatively small spin J = 2,
and the quantization axis is the J-axis, the two-particle density distribution shows
a prolate distribution relative to the J =2 axis.

152
 Fig. 4.9. Semi-classic illustration of the quadrupole moment
resulting from the coupling of two nucleons, moving in a
relatively large single j-orbital and coupled to a small total
angular momentum (J" = ~). Since the nucleons, taken
separately, move in a plane perpendicular to the orientation
of the single-particle angular momentum vector j (with in-
trinsic oblate distribution), the total distribution is prolate
relative to the z-axis, defined by the total angular momen-
tum J-axis

4.6 Magnetic Dipole Moment

Here, according to convention, the magnetic dipole moment is defined as the

expectation value of the dipole operator in the state with maximum M projection
as
(4.54)

Here gl and gs are the orbital and spin gyromagnetic ratios. The operator in
(4.54) can be reshaped so that the total single-particle momentum jz,i occurs.
This gives
(4.55)

or, applying the Wigner-Eckart theorem

p,= J 1/2(JIiL{91(i)i;+(gs(i)-91(i))sdII J }. (4.56)

(J(J + 1)(2J + 1)) i

4.6.1 Single-particle Moment: Schmidt Values

Here we use the reduced matrix elements

(jliilij) = (j(j + 1)(2j + l)i/ 2 , (4.57)

(jllulij) = 0(2j + 1) { l
II
1
1 0
~' }
1
(!llull!}(21 + 1)1/2. (4.58)

153
The Wigner 9j-symbol with zero angular momentum reduces to a Wigner 6j-
symbol that can still be written out in detail so that one gets (Brussaard, Glaude-
mans 1977, de Shallt, Talmi 1963)

Ullo-lIj) = «2j + 1)/j(j + l)i/ z [j(j + 1) + i -1(1 + 1)] . (4.59)

Using the reduced matrix elements (4.57,58), the single-particle dipole moments
become finally
J1. (j = I+~) =jg,+ (gs - g,)/2,
(4.60)
J1. (j = 1- !) = jg, - (gs - g,)/2. j /(j + 1) .
These results are the Schmidt values and are illustrated in Fig. 4.10, using g,(n) =
0, g, (P) = 1, gs(n) = -3.82n.m. and gs(P) = 5.58n.m. Thus, for the odd-
proton, a steady, almost linear increase in J1. for both the j = 1+ ~ and j =1 - ~
orientations is present, in particular for large j values. This is not so for the
neutron moments. Most of the experimental dipole moments fall in between the
Schmidt lines, indicating again that the nucleons in the nucleus behave differently
from free nucleons outside of the nucleus. The effective moments or g-factors
are model quantities as discussed in Sect. 4.3.
For 170, the example of j = ~ (lds/ z ) gives a value of J1. = -1.91 n.m.
(experimental value: -1.89n.m.). We briefly discuss a more detailed case of the
magnetic dipole moments for the odd-mass 491n nuclei. Even though the neutron
number is changing over a large shell (50,82) the odd-proton hole moves in the
same orbit, i.e., the Ig9/ Z orbital. This is a j = 1+ ~ orientation. In Fig.4.l1,
we compare the experimental values which stay remarkably constant over a very
large interval of 105 ~ A ~ 127, with the theoretical single-particle calculations.
We use gs = 1 n.m and g~ff = 0.7 g~ to account for the effect of other nucleons
that are not explicitly considered in the model space in order to describe the
odd-mass In nuclei. Even though in the actual calculations core excitations of
the even-even Sn core were considered (the collective part which is very small
and is not discussed in detail, here) the odd-particle part reproduces the data very
well. We also see that here the orbital and spin part act in phase to produce the
total single-hole dipole moment.

Fig. 4.10. The Schmidt single-particle magnetic moments for (a) proton single-particle states and (b)
!
neutron single-particle states. Both the j = 1 ± extreme lines are drawn [see (4.48)]. The data
points. corresponding to effective g., gl values fill up the domain in between the two Sclunidt lines
for a given charge state [taken from (Blin-Stoyle 1956)]

154
 6
.. ",.
6

:"'rc
115
In.,U,•

.....
• Slv

"Co·
.2OlI..
4 141;'
4
.21"1

.'L. lSI .51"",

,./U:UI Sb
~ 3 Reo:'.,,-
ll a: 81 Rb .,21,
•'H
II, 66C ..,e 7l Ge
' ' "-.G.
79 '-i l3c•
8' ...

,-.,
2

2O"T, 75A1 :'59Tb

.l'p
.lSe,
:&K::O
0 ~"~
'n".
til
101
At
Ir
0
a) I~ 10~h
1/2 3/2 5/2 712 9/2
Spin

+15 +15

+1 +1

.'35. .SlZn

...
1890. •

'......
• '3'XI
""s.
191.... : 207,.,
+5 ",
+5
et7'Yb

0
.,.... 6'N,: 57 Fe
0

-5 77s.: 231", 141Sm : ISlEr -.5

11'01: 29$, ·20'Hg 173yb: ! 05 Pd .'45 Nd
131 . . : I1Jec.
• 123r. '''7Sm •
12'X-. 25Mg: 7"'r. .23S U
11Ssn: 126Te .73Ge
''M ......
e:"z,.
....
14~d. 83Kr •
-1 -1
"'T.:
111511: '17Sn
79s.
9'z,:8's,
.91z,
."let
-15 -15

'. e17 0

.....
-2 -2

1 =1 + t
b) -2.5
1/2 9/2
Spin
155
 6r
-----r------------------------------------- -
-
E ,- orbito I
c

::t. collect I ve
2 - = =. ~r-=' =-. = ==.. ~- --=-. =----=- = -=-.- ---=--== =- ----'--=

[-spin

~~O~4~1~06~~10~8~1~10~~1~12~~11~4--~11~6--~118~-1~270~12~2~~1724--~12-6-1~28·
---A -
Fig. 4.11. The experimental dipole moments (I' N) in the odd-A In nuclei with a single-hole moving
in the Ig9 / 2 orbital below the Z =50 shell closure [data taken from (Eberz 1987)]. The theoretical
values [the separate contributions from orbital (g/). spin (g.) and collective admixtures] are shown
= =
(using g/ IJJN, g~ff 0.7g~). The theoretical values have been derived in (Heyde 1978)

4.6.2 Two-particle Dipole Moment

i) For J = 0, we trivially have f.1 = 0,
ii) for J1r = 2+, using the general expression (4.56) we obtain

f.1(i;2+) = Hsu 2; J = 211 t

i=l
{g/(i)ii + (gs(i) - g/(i»)s;} IIi; J = 2) .
(4.61)
The two-particle to one-particle reduction fonnulas of Chap. 2, applied to a tensor
of rank 1, give

(i; J = 2I1F(1) IIi; J = 2) = 10 {~ ~ {} Ullf(1)lIi) . (4.62)

Combining the results of Sect. 4.6.1 with the outcome of (4.61,62), the two-
particle dipole moment reads

(4.63)

Here the moments are additive. For the J1r = 2+ state the two nucleons each
contribute their own gj factor and yield the value 2g j •

156
4.7 Additivity Rules for Static Moments

Magnetic dipole and electric quadrupole moments have been measured in many
odd-odd nuclei near closed shells (odd-odd In, odd-odd Sb, odd-odd n, ... nu-
clei). For such nuclei, starting again from a rather simple configuration for both
the odd-proton and the odd-neutron cases, the known moments and rather general
"additivity" rules can be derived by using simple angular momentum recoupling
techniques (de-Shalit, Talmi 1963), and have been used in determining the com-
posed moments.
If we call the eigenstate in the odd-proton nucleus IJp) with jl(Jp), Q(Jp), ...
the corresponding moments and IJn ) the eigenstate for the odd-neutron nucleus
with jl(Jn) , Q(Jn), ... the corresponding moments, under the assumption of
weak coupling in obtaining the eigenstate IJ) = IJp 0 I n; J) in the odd-odd
nucleus, one obtains the expressions

jl(J) = !... [jl(Jp) + jl(Jn) + (jl(Jp) _ jl(Jn»)

2 Jp In Jp In
Jp(Jp + 1) - In(Jn + 1)]
(4.64)
x J(J + 1) ,

Q(J) = (!J ~ ~) (_l)Jp +Jn+J . (2J + 1)

x [C;
(4.65)

These equations also apply to the coupling between identical nucleons in going
from the one-particle to the two-particle nucleus. Applying (4.64) to a system
of two identical nucleons, then jl(Jp) = jl(Jn) = jl(j), Jp = I n = j and (4.64)
reduces to (4.63) since
jl(J) = Jgj , (4.66)

an expression which applies for more general J, than just the value J = 2 that was
taken in deriving (4.63). Analogously, for the addition of quadrupole moments

in
with Q(Jp) = Q(Jn ) = Qs.p.(j) and Jp = I n = j, (4.65) reduces to

Q(J) = ~ ~ (-I)'" {~ ~} ; Q.pU) , (4.67)

157
 7
.<> ( • 3, exp, emp
eo ( - 4, exp, emp
6
IJ 0 .v
....<3
I • 5, exp,
( - 7, exp.,
emp
emp

E e _0 ( • 8, exp, emp

.s 5
., '\1-..
-.....~.~
- ':----:---¥---i <>---0
:::1.. 4

3 - - -- - - -{3-- - 8- - - 0- - - G -- -G--'El

104 106 108 110 112 114 116 118 120 122 124 126 128
mass number A
Fig.4.U. D1ustration of the magnetic dipole addition rules (4.64) in the case of the odd-odd In
nuclei. The data are taken from (Eben 1987). The full symbols represent the experimental values,
the open symbols the results according to the addition rules. The moments needed to apply the
addition rules are taken from the adjacent odd-proton (In) and odd-neutron (Sn) nuclei, when avail-
able. The symbols denote <>, J = 3; 0, J = 4; V, J = 5; <I, J = 7; D, J = 8 [taken from (Eben 1987)]

<3.
-'<3--<3
1.0 ~----~-.- ......

0.5
f
a
Ul

-0- - ·0'
o -0--
0.0

Fig.4.13. See caption to Fig.4.12, but now for the electric quadrupole moments and using (4.65)

an expression which, for J = 2, reduces to (4.52). Here again, (4.67) is more

genera1 since it applies to any J va1ues.
These methods have been tested with remarkable success in the odd-odd In
mass region by Eberz et a1. (Eberz et a1. 1987) for many magnetic dipole and
electric quadrupole moments. (See Eberz et a1. 1987 for a detailed discussion
on the possible configurations.) In such an approach, rather complicated proton
and neutron states are combined under the assumption that these states are not
modified very much when coupling to form the fina1 state in the odd-odd nucleus.
In odd-odd nuclei near closed shells, when rather few configurations with the
same Jp and I n va1ue are present in the odd-proton and odd-neutron nuclei,
respectively, the additivity rules are expected to work well. The application to
odd-odd In nuclei in the interva1 104 ~ A ~ 126 has been carried out for both
the J.t(J) and Q(J) va1ues and has considerable success in accounting for the
data, as is illustrated in Fig.4.12 and 4.13. These figures are taken from (Eberz
et a1. 1987).

158
 For the odd-neutron case, where the neutron number N varies over a rather
large interval 51 :5 N :5 79, a less unambiguous situation results. If we have
the odd-proton nucleus eigenstates IJ~i)} and the odd-neutron eigenstates IJ~P},
where we can have different (Jp , i) values and (In,j) values in a single nucleus,
weak-coupling may no longer hold, i.e., one obtains wave functions (Heyde 1989)

(4.68)

In this case, extra components from configuration mixing in the final nucleus
result. These terms give rise to extra "polarization" terms with respect to the
original zero-order term. A good qualitative estimate of configuration mixing is
obtained by studying the number of final states J in the odd-odd nucleus. If there
is only a single J state over an interval of ~ 1 MeV, weak-coupling will most
probably be a good approximation. If, on the other hand, many J levels result
at a small energy separation, chances for large configuration mixing are more
likely to occur.
In using the additivity method, it is of the utmost importance to use the odd-
mass moments as close as possible to the "unperturbed" odd-mass nuclei that are
used to carry out the coupling in obtaining the final odd-odd nucleus. In the odd-
odd In nuclei, more in particular for the (lg9ii(1l") 1hll / 2(V») 8- configurations,
some problems occur when comparing the measured moments and the "additiv-
ity" moments (Fig. 4.14). If one considers a pure 1g9ii proton-hole configuration
and a Ihll/2 neutron one-quasiparticle configuration (and linear filling of the
Ihll/2 orbital with n valence neutrons), then the dipole, respectively, quadrupole
moment would vary as (de-Shalit, Talmi 1963)

!
:a
0.5
4~In

0
Q
100

Fig. 4.14. Variation of 1'(81) and Q(81)

E in the odd-odd In nuclei (112 ~ A ~

-c::
~
126). The experimental data points (_)
are taken from (Eben 1987). The addi-
tivity moments (0) [see (4.64) and (4.65)]
D DaD are obtained, using the discussion in (Eben
1987)
114 116 118 120 122 124 126
---A _

159
 Jl(8-) =aJl (lg9i~) + bJl(lhll / 2 ) , (4.69)

Q(8-) = a'Q (lg9i~) + b'(12 - 2n)/1O(lhll /2) , (4.70)

which means a constant value for Jl(8-) and a linear increase in Q(8-) with n,
the number of neutrons filling the 1hll /2 orbital. In the more specific case of more
orbitals filling at the same time, some modifications to this simple dependence
on particle number can be expected (starting of filling the 381/2, 2d3 / 2 orbitals
before N = 76 and early filling of the Ihll/2 orbital before N = 64). This will
result due to the pair correlations and the resulting pair distribution of neutrons
over the five neutron single-particle states (Chap. 7).
We conclude this chapter, which is an intensive application of the rules and
methods of Racah-algebra to spherical tensor operators that describe the electro-
magnetic properties and observables in the nucleus. We studied some of the most
often used one- and two-particle transition rates and moments, concentrating on
the Ml and E2 operator. The methods, however, are general and can easily be
extended to other operators, too, starting from the general form of the electric
and magnetic multipole operators discussed in Sect.4.2. It is also possible to
describe other observables: Beta-decay processes can be discussed analogously
to electromagnetic decay processes. One can also address nucleon transfer reac-
tions. In all these cases, one always has to first construct the appropriate spherical
tensor operators when evaluating the nucleus matrix elements.

160
5. Second Quantization

5.1 Creation and Annihilation Operators

In Chaps. 3, 4, we have fonnulated the nuclear shell model using the coordinate
representation. This means that when considering identical particles in the nu-
cleus, the Pauli principle implies the explicit construction of antisymmetrized
wave functions in the exchange of all coordinates of any two of the A particles.
If we call, as before, the single-particle wave functions !pQ(ri, tTi, Ti), denoted
by !Pa(i) with 0: == n a, la,ja, rna, t zG , wave functions that constitute a complete
orthononnal set, then the A-particle antisymmetrized wave function reads

!Pal (A)
!pQ2(A)
1
. (5.1)

!PaA (A)
In such a wave function there is superfluous infonnation: the important point is
that the A single-particle orbitals 0:1, 0:2, ••• O:A are occupied with a nucleon. The
Pauli principle gives no extra infonnation: it just includes the antisymmetrization
meaning that any particle with coordinates 1 (r1, tTl, t1), 2, ... A could be in
any single-particle state.
Thus, we shall try to go to a fonnalism that gives the minimum of infonnation,
i.e., which orbitals are occupied [also called the occupation number representation
(Brussaard, Glaudemans 1977)] but has the Pauli principle built in, making use
of the properties of the detenninant wave function in (5.1).
We point out that the name "second quantization" fonnalism can be mis-
leading since we are not introducing quantum field theory nor the subsequent
quantization of standard quantum mechanics. The tenn "occupation number rep-
resentation" is more precise since one simply uses an alternative formulation of
the usual quantum mechanical description given in Chaps. 3,4. However, it turns
out to be a very useful way of handling an interacting many-body system such
as the atomic nucleus.
We denote a vacuum state with no particles by I}. By acting with an operator
a~i on this vacuum state, a particle in the quantum state O:i is created and is
denoted by

(5.2)

161
Thus we put a one-to-one correspondence between the single-particle wave func-
tions <Pai(1) and the states lai)

(5.3)

The A-particle state can then be obtained by the successive action of A such
operators, acting in a given order. The A particle state is made by acting of an
operator on the A - 1 particle state as

(5.4)

or, working this out

(5.5)

We now need a one-to-one correspondence between the states (5.1) and (5.5), or

(5.6)
The properties of the determinant wave function (5.1) indicate that it is multiplied
by a factor of -1 when interchanging any two rows. This leads to the following
properties of the operators a~i' i.e.,

(5.7)
or
(5.8)
From the definition of the many-particle states given by (5.5), this gives

(5.9)

for any A-particle state and any interchange of two rows. Thus we get

(5.10)

for any ai, aj, and {A, B} is the antic om mutator of the two operators A, B,
i.e., AB + BA.
Equation (5.10) expresses the Pauli principle within the occupation or
creation-operator formalism. The fact that two particles cannot move in the same
orbital is easily expressed by (5.10) putting ai == aj.
Using Hermitian conjugation, we obtain

(5.11)

with aai the Hermitian conjugate operator to a~i' We can easily rewrite the
above A-particle kets via Hermitian conjugation and obtain the results

162
 (at, a2, ... , aA I = (Iat, a2, ... , aA) r
= (I (a: A ••• a: a:J+
2

= ( laal aa2 ••• aaA • (5.12)

Thus, the operators aai acting to the left act again like creation operators. We
also get the Hermitian conjugate of the commutation relation (5.10) as
(5.13)

We now want to know what the meaning of the aai operators, acting to the right
is.
In order to obtain the result of the operator a A acting to the right on a general
state la,,8,,, ... ), we calculate the components of the vector

aAla,,8,,, ... } , (5.14)

in the basis formed by alII, 2, ... , A, ... particle states la', ,8', ,', ... } for all
, (.I' ,
a,f/,','" .
We know that
(a',,8',,', .. , la Ala,,8,,, ... ) = (a',,8',,', ... , >'la,,8,,, ... )
=0 if { a' , ,8' , ,', ... , >.} f {a, ,8, " ... }
(5.15)
= ± 1 if {a' , ,8' , ,', ... , >.} = {a,,8,,, ... } .
i) The case >. ¢. {a,,8,,, ... , w}. It follows that for the full set of states
{a',,8',,', ... ,w'} all matrix elements (a',,8',,', ... ,w'la Ala,,8,,, ... ,w)
vanish. Since all components of the vector aAla,,8,,, ... , w} vanish, the state
should have zero length and we note
aAla,,8,,, ... , w} = 0 for >.¢. {a,,8,,, ... , w} . (5.16)

ii) On the other hand, we have

(a,,8,,, ... , wla Ala,,8,,, ... , w, >.)
= (a,,8,,, ... , w,>'la,,8,,, ... , w,>.) = +1,
and we obtain the result
aAla,,8,,, ... , w,>.} = la,,8,,, ... , w} for >'¢'{a,,8,,, ... , w}. (5.17)
Thus the operator a A acting to the right on a given state acts as an annihilation
operator: if >. is unoccupied we get zero, if >. is occupied we can get the result
± 1 depending on the order of the operators. We give some examples:
aAI}=O,
aAI>'} = 1 for all >.,
a~I,8} = 0, (5.18)
a~la} = la,8} ,
apla,8} = la} .

163
 Having obtained the anticommutation relation for two creation or two
annihilation operators, we also want to derive the anticommutation relation
for a creation and an annihilation operator. We therefore use the following
method: We let the operators a~A aaA_l and aaA_l a~A act on a single state
lal, a2, ... , aA-2, aA-l}. We find
a~A aaA_llal, a2, '" , aA-l} = a~A lal, a2, ... , aA-2}
= lal, a2, ... , aA-2, aA} . (5.19)
Carrying out the same action but in opposite order, we obtain
aaA_l a: A la l,a2, ... , aA-2,aA-l}
= aaA_llal, a2, ... , aA-2, aA-l, aA}
= -aaA_llal, a2, ... , aA-2, aA, aA-l}
= -Ial, a2, ... , aA-2, aA} (5.20)
Adding (5.19) and (5.20) we obtain
(5.21)
for any state I ... }, giving that
{ a:., aaj } = 0 for ad aj . (5.22)
If ai = a j, the above arguments lead to
i) a:.aa. = 1 and aa.a:. =0, or
(5.23)
ii) a::'.aa· = 0 and aa.a::'. = 1 ,
...... , , • \A I

if the state ai is occupied or unoccupied, respectively. In conclusion, we can

state the basic content of the occupation number representation or the represen-
tation using creation and annihilation operators as given by the anticommutation
relations
{a:., a: j } =0,

{ aa. , aaj } = 0 , (5.24)

{a:., aaj } = lia.,aj ,

for any ai,aj.
We work out a few examples:
i) (Iaaapl) = (I }liap - (Iapaal) = liap .
ii) ( laaapa~a~ I) = liaoli p-y - lia-yli po .
iii) We call N a = a~aa the number operator since N a
acting on an occupied
orbital a gives 1, and the result 0 when acting on an unoccupied state a. Then
(5.25)
a a

is the total number operator.

164
5.2 Operators in Second Quantization

Having indicated that there exists a one-to-one correspondence between the Slater
determinant wave functions (5.1) and the states obtained via creation operators
acting on a vacuum state, we would also like to find out the way in which the one-
and two-body operators can be represented in second quantization (Brussaard,
Glaudemans 1977). We first study one-body operators.
A general, one-body operator for an A-nucleon system reads
A
0(1) =L O(rk, Uk, Tk) . (5.26)
k=1
(Below we use the notation rk == rk,uk,Tk, ... .)
The one-body matrix element for single-particle states CPa(r) and cpp(r) be-
comes

(a/O/,8) f
= cp~(r)O(r)cpp(r) dr . (5.27)

We shall demonstrate that the equivalent form to 0(1) in second quantization

reads

°=l)a/O/,8)a!ap .
a,P
(5.28)

We show the correctness of (5.28) by evaluating the one-body matrix elements

(al,a2, ... , aA/O/al,a2, ... , aA)
= L{a/0/,8){/aa1 aa2'" aaAa!apa!A'" a!2a!1/)' (5.29)
a,P
For each case a = ,8 = ai (i = 1,2, ... , A), the value of the matrix element
containing the creation and annihilation operators is +1. Thus the matrix element
(5.29) becomes
(5.30)
ai

Also, non-diagonal matrix elements different from zero can result if from the A
particle quantum numbers, A-I are identical in the bra and ket state, or,

(al,a2, ... , aA/O/"n,a2, ... , aA) = (al/0/'Yl) (at7~:'Yl)' (5.31)

One can illustrate the effect of one-body operators in a diagram (Fig. 5.1)
where the operators in the initial and final states are indicated by lines going
up and the one-body operator acts as an internal probe changing (eventually)
the quantum numbers of not more than one particle at the time. Thus we can
illustrate (5.30,31) schematically.

165
 a1

x- - x- + ..... + -x

a1

Vi
x-- = x--

Fig.S.l. Dlustration of the one-body operator matrix elements taken between the A-panicle states
lal, az, ... , aA) as obtained in (5.30) and (5.31). We use a Feynman-Goldstone diagrammatic way
to represent the above processes. (I) Diagonal elements (upper part) are given by the sum over A
one-body processes acting on each separate line al, 0!2, '" up to a A. The interaction zone of the
one-body operator with the A-panicle state is indicated by the "blob" (* - -). (iI) Non-diagonal
elements (lower part) where only a single one-body scattering process remains (al -+ 'Yl). Orthog-
onality takes the (A - 1)-particle state away

For two-body operators we have in coordinate space

A
0(2) = L O(r;,rj) (5.32)
;<j=l

and the two-body matrix elements

(a,BIOI'Yb}nas = f ~~(rl)~p(r2)0(rl' r2)(1 - H2)~-y(rl)~6(r2)drldr2 ,

(5.33)
with H2 an operator that exchanges the coordinates of particles 1 and 2. In
second quantization, the corresponding two-body operator becomes

o=1 L (a,BIOhb}nasa~apa6a-y. (5.34)

Ot,{J,-y,6
166
We verify, by a direct calculation, that (5.34) gives the correct two-body matrix
element when used within an A -particle state. We thus calculate
(al,a2, ... , aAIOlal,a2, ... , aA)
=t L (afjIOhl5)nas
0/./3.",(.6
x (I aO/ l a0/ 2 •••
+ + a"'(aO/
aO/ A aO/a/3a6 +
A •••
+ aO/
a0/ 2
+
l
I) • (5.35)

In the evaluation, in order to get non-vanishing results, we have to take

"I = ai , 15 =aj =ai , fj =aj

and a
or a =aj , fj =ai .

We can similarly change to "I = aj and 15 = ai, with again the two cases for
a and fj. When using the fact that the two-body matrix elements are evaluated
with antisymmetrized two-body wave functions one has

(5.36)

Thus in total, the matrix element becomes

(al,a2, ... , aAIOlal,a2, ... , aA)
=L (aiajIOlaiaj)nas . (5.37)

Similarly, we can obtain non-diagonal matrix elements

(al,a2, ... , aAIOI'YI,a2, ... , aA)

= L(al a jIOI'Yl a j)nas (al F'Yl) ,
(5.38)
(aI, a2, ... , aAIOI'YI, "12, ••• , aA)
=(al a210h1'Y2)nas (aI, a2 F "11, 'Y2) .
Here, too, by using the Goldstone diagrams one can show how a two-body
operator connects two incoming and two outgoing lines, leaving all other A - 2
nucleon lines unaffected (Fig. 5.2).

Fig. 5.2. lllustration of the two-body matrix element taken be-

tween the A-particle state la1. a2 • •..• a A} as obtained from
(5.38). Because we represent a non-diagonal element, only a
single process (with the two-body operator represented by the
diagram > - - <) scattering the state a1. a2 into the states
"'(1. ')'2 remains. The A - 2 nucleon state remains unaffected
by the two-body operator

167
 Having a two-body operator in coordinate space, it is always possible to
give the second quantized fonn corresponding to it by evaluating all two-body
matrix elements. It is also possible by using second quantization, to define an
interaction immediately in second quantization. Such an interaction is, e.g., the
pairing interaction which will be discussed in more detail in Chap. 7. For the
pairing interaction, there is no easy way to get a corresponding expression in
coordinate space. We shall comment on this point here.
A constant pairing interaction (Rowe 1970) in second quantization reads

v = -G L (_l)i+ m(_l)i+ m' aj,maj,_mai,-m,ai,m' , (5.39)

m,m'
(>0)

and indicates that a nucleon pair is annihilated in the states (j, m')(j, -m')
(a 0+ coupled pair) and scattered via the two-body interaction V into any other
substate (j, m)(j, -m) with constant strength G independent of the m-value. [In
Chap. 7, we shall discuss the particular relevance of the phase factors appearing
in (5.39).] This means fully isotropic scattering in the space of magnetic sub-
states m. One can easily show that the interaction (5.39) only affects the r = 0+
state and leaves all other J1r = 2+, 4+, ••• , (2j - 1)+ states degenerate at zero
energy. Thus the matrix giving the two-body matrix elements where column and
row label the magnetic substate quantum number has constant elements for the
pairing interaction
'm'
,
I j
I
j -1 ...

-]-1
m', I

j-.l1.
1 1 ...
1 ... . J
(-G) . (5.40)
.. .. ..
.. .
. ..
If we construct a similar matrix to (5.40) for a 6-interaction, scattering is almost
isotropic but with a slight preference for diagonal over non-diagonal scattering.
Speaking intuitively, the 6-interaction is the ultimate short-range interaction. We
see that the pairing interaction has an even "shorter" range by inspecting the scat-
tering matrix in the space of the different magnetic substate quantum numbers.
For very low multipole forces, e.g.,
V(1,2) = rrr~~(cos812) , (5.41)
one can show (see problem set) that the scattering matrix equivalent to (5.40) is
almost diagonal for large j-values.
In Fig. 5.3, we compare the 6- and pairing force spectrum for a (ji J con-
figuration. We also compare in Fig. 5.4 the scattering in m-space for the pairing
force compared to the low multipole (quadrupole) force.

168
 PAIRING 6-FORCE Fig. 5.3. Comparison of the two-panicle spec-
tra (j)2 J for a pairing force [defined in second
quantization in (5.39] and a 8-force

_ _ _ _ _ 0·

PAIRING Fig. 5.4. Comparison of the scatter-

ing in the space of magnetic sub-
Z-axis states (in m-space) [the matrix of
(5.40)] for a pure pairing force and
a low-multipole force [e.g. a ~
(cos (h2) force]. In case (I), scatter-
ing is fully isotropic i.e. the prob-
ability for scattering from a state
(m, -m) to any state (ml, -m / )
is equal, independent of m and mI.
In case (ii), scattering is preferen-
tially in the "forward direction",
this means that the final pair (ml,
-m /) obtained from (m, -m) is
such that m l ~ m

5.3 Angular Momentum Coupling in Second Quantization

In the above discussion, we have presented how to approximate the many-particle

states in describing the motion of A nucleons in the nuclear average field. As was
shown, however, one needs coupled angular momentum states. The same methods
of Chap.4 can be used here, too. Because of the anticommutation properties
of the creation and annihilation operators, the antisymmetry aspects that were
demanded by the Pauli principle in coordinate space are automatically included.
For two particles, one can construct the state

(5.42)

The condition for normalization then determines Nab. Performing this calculation
in detail, one has

169
 1 = (jajb; J Mljajb; J M}
-N 2
- ab L
mG,mb,m~,m~

X (lai4m~aibm~ajbmbalam41}. (5.43)

Using orthogonality properties of the Clebsch-Gordan coefficients and the

fact that the matrix element ( 1••• I} in (5.43) becomes

one gets
1 = N;b (1 - oab(-1)i4+ib- J )
=N;b(l +Oab) (J: even). (5.44)
Thus, the normalized state for two identical particles becomes

Ijajb; JM} = (1 + Oabr 1/ 2 L (jama,jbmbIJM}ajbmbalam41} , (5.45)

or, using the notation of tensor coupling for the two creation operators, one can
also write
(5.46)

If we include the possibility of non-identical nucleons by including the isospin

quantum numbers in the particle state, one has

X (~tZ4' ~tzbITMT}a;bmb,I/2t'baj4m4,1/2t.J}, (5.47)

or, in coupled notation

. .
I]a]b; JM, TMT} = - ()-1/2
1 + Oab
[a + ,1/2 0 +
a ib ,I/2
](J,T)
I}, (5.48)
i4 M M
, T

We can now use the above method to easily construct coupled (and antisym-
metrized) 3, 4, ... n particle states, by coupling the creation operators and cal-
culating the overlap and norms of the different possible states. For three-particle
states, one can construct the states

(5.49)

as a notation for
(jama, jbm blJt Ml} (Jl Ml, jcmc IJ M}

(5.50)

170
 In calculating the nonns of the states (5.49,50) and the overlaps, the non-
vanishing states will give the cfp coefficients for the three particle states that
were discussed in Sect. 3.3 (see problem set).
Not only the basis states but also the one- and two-body operators can be
represented in an angular momentum coupled form. We show this for the two-
body residual interaction that is written as

(5.51)

Now, by using the orthogonality of the Clebsch-Gordan coefficients, we can

carry out the angular momentum coupling in both the matrix element and the
operators, and using the definition o'i,m == (_1)i+m ai,-m (Sect. 5.4), one gets
for the residual interaction Hamiltonian

(5.52)

where, indeed, a scalar operator (coupling to tensor rank 0) results. Extension to

include isospin is straightforward.

5.4 Hole Operators in Second Quantization

If we consider a fully occupied single j-shell (with 2j + 1 particles, since all

magnetic substates m = -j, ... , m = +j are occupied) a closed shell has total
angular momentum J =0 and M =O. If we now consider the excitation of one
particle out of the occupied j-shell into a j'-orbital. a particle-hole excitation is
fonned with respect to the fully occupied j-shell (Fig. 5.5). There now exists a
method of using specific particle and hole creation operators. This allows using
a more economic method than considering a [(j)2i J = j,l] 2j + 1 particle
configuration (de-Shalit, Talmi 1963. Rowe 1970). We now define the closed
j-shell state as

- - - - - j' Fig. 5.5. Schematic representation of a full

shell (2j + 1 particles) and a hole state
(identical to a 2j particle state) in the
111111111111
orbital j (here, a particle needs to be-
'---v----' come excited into another configuration
2)+1 j' if conservation of particle number is
fulfilled)

171
 10) == Ilj+l; J =0, M =0) , (5.53)
and
10m)-1) oc Ij2j ; J = j,M = -m) , (5.54)

indicating that taking a particle out of the magnetic substate m, a 2j-particle

state remains with magnetic quantum number M = -m. We of course need to
have a better definition of the hole state.
We know that the particle creation operator acting on a vacuum state
I), aj,ml) transforms as a spherical tensor of rank j such that

Raj,ml) = 2::D~,m(R)aj,md) . (5.55)

m'

One can easily prove that a full shell like the state 10) with

10) == aj,jaj,j_l ... aj,_jl) , (5.56)

is invariant under rotation (J =0, M =0 state) and thus

RIO) = 10) . (5.57)

One can now ask for the transformation properties of the state aj,mIO). This
leads to a more tedious calculation (see problem set) but one finally obtains

Raj,mIO) = 2::(_l)m-m' D~!..,,_m(R)aj,mIO) , (5.58)

m'
or
R(-I)j+m aj,_mIO) = 2::D~,m(R)(-I)j+m'aj,_m'10). (5.59)
m'

Thus, the operator (_I)j+m aj,-m transforms as the creation operator aj,m
(Brown 1964).
We then define as the hole state
(5.60)

or, in second quantization notation for the hole creation operator

a-),m
· =
- (-I)j+ma·),-m· (5.61)

Thus, aj,m and aj,m both transform as spherical tensors of rank j under rota-
tion and can thus be coupled in the same way that we coupled angular momentum
eigenvectors or spherical tensor operators using Clebsch-Gordan coefficients. One
can easily show that creating a hole and creating again a particle in the same
state 0, m) leads to the closed shell. Formally, one has

172
 m

(5.62)

A general particle-hole state Ijajb"l; JM) then becomes, in second quantization,

liajb"l; JM) = L (jama,ibmbIJM)aib,mbaj.,mJO)

(5.63)

In the above way one can indeed economize the notation for low-lying excitations
in doubly-closed shell nuclei. We take the example of 160 where, in a Hartree-
Fock approach, the ground state is described by a 16-particle state

10) == 1160; O;s)

= (aip1 / 2,m=_1/2 ••• aiS 1/ 2,m=_1/2 aiS 1/ 2,m=+1/2)1I' 1)11'
X (aip1 / 2,m=_1/2 ... aiS 1/ 2,m=_1/2 aiS 1/ 2,m=+1/2) 1)11 .
II
(5.64)

A particle-hole excited state with a particle in the orbital Ids/2,m and a hole in
the IPI/2,+1/2 shell model state can then be written, for a proton p- h excitation,

(5.65)

and we do not need to write the 16 operators again (Fig. 5.6).

ldS/21-_----_-'r--=-
1Pl/2 I
1P3/2 ~H!--;OI-*---I/
15 1/2 I
1---_./
Fig.5.6. Representation of the 160
ground-state distribution of 8 nucle-
ons (protons and neutrons) over the
lowest 181/2, 1113/2, IPI/2 orbitals
(5.64). The state is indicated by 10).
1dS/2 In the lower pan (left-hand silk), an
excited 16 panicle state is shown rel-
1Pl/2 I--~:J----I
ative to the vacuum state of nucle-
1P3j2 ons. The same state can be repre-
15 1/2 1----i4'"'*---/ sented (right-hand silk) in a much
simpler way as a Ip - Ih Ip'0~
1ds/2 configuration, relative to the
160 excded particle - hole state new vacuum state 10)

173
 Including isospin in the formalism we have in exactly the same way the
operator
a.),m;I/2,t. = ·
- (_1)i+ + / +t· a),-m;I/2,-t.
m I 2
, (5.66)

and the operator in (5.66) acting on the closed shell produces a hole state

(5.67)

If we now consider particle-hole excitations that can be both a proton p - h and

a neutron p - h excitation, we can form specific linear combinations by vector
coupling in isospin.
Using the definition of (5.66,67), a proton p - h state becomes
1
Iph- I },. = v'2{ lph- I ; T = 1} -Iph- I ; T = O}} ,
(5.68)
1
Iph-I}v = - v'2{ Iph-I ; T = 1} + Iph- I ; T =O}} ,
and the symmetric and antisymmetric combinations of (5.68) give
1
Iph- I ; T =O} = v'2{ lph- I },. + Iph-I}v} ,
(5.69)
1
Iph- I ; T = 1} = v'2 {Iph- I },. -Iph-I}v} .

This is in agreement with the generalized Pauli principle: a symmetric spatial-

spin wave function in (5.69) is a T = 0 (antisymmetric in charge space) state
and vice-versa. We shall see in Chap. 6 when studying low-lying excited states
in doubly-closed shell nuclei that the lowest states are the T = 0 states and the
T = 1 levels are the higher-lying ones.

5.5 Normal Ordering, Contraction, Wick's Theorem

In the foregoing discussion we used the notation I} for the real vacuum state
and 10} for a closed-shell reference state. In the latter we can form particle-hole
excitations (Fig. 5.7).
Let us call hI, h2' h3 • ... the quantum numbers of the occupied states in 10}
and PI, Pl, Pl, ... the quantum numbers of the unoccupied states relative to the
reference state 10}. We then define new operators such that
t+ -
lOp = a+p ,

e == a
p p ,
(5.70)
fh == ah,
eh == at·
174
 Fig. 5.7. Separation of particle configurations (pI, Pl, .•. , Pi, ...)
-----P4 above theFenni level and hole configurations (hI, h2, ... , hj, ...)
P3 below the Fenni level. The reference state 10) corresponds to the
- - - - - P2 filling up to a sharp Fenni level with A nucleons
P,

lo)-~~~~~
~~~~,,*h,
44<~~~~h2

The operator e~ (with a either p or h) now creates a particle (for unoccupied

states) and a hole (for occupied states), and, in all cases, we have

ea.l0} = 0, (for all a) (5.71)

{Ole~ =0. (5.72)

The anticommutation relations now remain for the new operators e~ or

{e~,ep} = 0,
{ea,(S} =0, (5.73)
{e~, e,8} = Sa,8 .
i) Normal Product (Fetter, Walecka 1971). A product of operators is in "normal"
order relative to a given reference state 10} when all creation operators are at
the left of the annihilation operators. In bringing a product of operators to the
normal order, we often have to carry out a number of permutations. One does not,
however, consider possible Kronecker 8-symbols when interchanging appropriate
creation and annihilation operators. The sign becomes ± 1 according to the nature
of the permutation needed to bring the original operator product in normal ordered
form. As a notation one uses N(ABC ...) or :ABC ... :.
As an example:

N (at a;l ap2 ah2) =N (~hl e;l eP2~2)

=-e+p1 e-+h2 eh
-
1ep2
+ +
= -apt ah2aht a p2 , (5.74)

One clearly sees that the result depends on the choice of the reference state.
An interesting outcome is that

{OIN(ABC ...)IO} =0 . (5.75)

175
ii) Contraction (Fetter, Walecka 1971). The contraction of two operators is the
expectation value of the operator product with res~ct to the reference state. The
result is a pure number and we use the notation AB. Thus, one has

f1;ap == (Ola:apIO) , (5.76)

~:ap~'Yali == -(Ola:a'YIO)apali , (5.77)

and the permutation factor is needed, ± 1 for even or odd permutation, in order
to bring the contracted operators together. One can also have contractions within
a normal product of operators, i.e.,

(5.78)

and

N(~:~pa~~lidea,) = -(Ola:aliIO) (OlapaeIO)N(a~a,) . (5.79)

We are now in a position to formulate Wick's theorem.

iii) Wick's Theorem (Fetter, Walecka 1971). A product of a number of operators
can be written as the sum of all contracted normal ordered products of the
operators (considering partially and fully contracted terms, also the uncontracted).
Thus we have
AIA2 A3 ... AN =N(Al' A2, A3 ... AN)
+ L N(AIA2 ... l'--u-.-..-'A p ... AN)
u<p
+ L I I i I
N ( AIA2 ... Au ... A'Y ... Ap ... Ali ... AN
)

a<fJ
-,<6
+ ...
+ fully contracted terms. (5.80)

As an example, we give

(5.81)

We do not give a proof but refer the reader to (March et al. 1967).
Relative to the reference state 10) we have also

(5.82)

176
5.6 Application to the Hartree-Fock Formalism

We start from the one- and two-body A-particle Hamiltonian describing the
kinetic energy and two-body interactions of A fermions in a nucleus as

(5.83)
OI,Y 0I,/3,y,6

Applying Wick's theorem to the Hamiltonian one has,

H=L(aITI'Y} {N( a~a"() +~"(}

01,"(

(5.84)

The contractions disappear, except in those cases ~"( where a = 'Y and
a, 'Y describe occupied states. If, moreover, we keep track of the antisymme-
try of the two-body matrix elements, i.e., (a.BIVI'Y0}nas = -(a.BlVlo'Y}nas =
-(.BalVhc5}nas = (.BalVlc5'Y}nas, and calling h, h', hit, ... the quantum number
of occupied states in the reference state IO), one can rewrite the Hamiltonian as

H = L(hITlh} +! L(hh'IVlhh'}nas
h h,h'

+ ~ { {aITh} + ~ {ahlVhh}_ } N ( a~a,)

+ l L (a.BIVI'Yo}nas N (a~apa6a"() . (5.85)
01,/3,"(,6

If we now take the expectation value of H relative to the reference state IO}
of occupied orbitals h, h', hit, ... , the normal products give no contribution and
one has the energy Eo

Eo == (OIHIO) = L(hITlh} +! L(hh'lVlhh'}nas . (5.86)

h h,h'

For an appropriate choice of the basis states la}, ... , the single-particle Hamilto-
nian can be diagonalized and one obtains (we consider the single-particle energy
ea to be independent of the magnetic quantum number rna)

(aITI'Y) + L(ahlVl'Yh}nas = eaoOl"( , (5.87)

h

177
anu
Ca = (aiT + Ula) , (5.88)

with the one-body average field defined as

(aIUla) == 'L)ahlVlah}nas . (5.89)

h

In this way, an average field is defined in tenns of the two-body interaction

according to (5.89). One can finally rewrite the Hamiltonian as

H = Eo + L caN ( a:a Q ) +1 L (a.BlVho}nasN (a:apa6a1')' (5.90)

01 01,/1,1',6

and the total Hamiltonian has thus been separated into a core tenn Eo, the
energy of the reference state 10), the single-particle energy contributions given
by the Ca, and the residual interaction given by the nonnal ordered product
N(a~apa6a1'). In spherically symmetric nuclei where the single-particle energy
Co. is independent of the orientation of the angular momentum ja, we denote the
single-particle energy as Ca (a = n a, la, ja and a == {a, rna}).
This application shows the power of Wick's theorem in separating the Hamil-
tonian in different tenns. The Hamiltonian (5.90) can now serve as a starting point
in describing the elementary excitation modes in (i) closed-shell nuclei and (ii)
nuclei with a number of valence nucleons in open shells. We shall now study
these two cases in some detail in Chaps. 6 and 7, respectively. There, the methods
of second quantization will be fully used. We shall also study both approximate
exactly solvable models as well as realistic cases using modem, effective interac-
tions. In Chap. 8 we shall discuss the state-of-the-art of shell model calculations
in a way similar to that in Chaps. 6, 7 but using only the Skynne interaction.

178
6. Elementary Modes of Excitation:
Particle-Hole Excitations at Closed Shells

6.1 General

Having derived in the final part of Chap.5 the second quantized expression
for the nuclear many-body Hamiltonian (containing one-body and two-body op-
erators), we shall now study how elementary excitations of the doubly-closed
shells, via particle-hole excitations, can give rise to the low-lying excited states
that have been observed in these nuclei. In this chapter we shall describe the
basic methods of treating the residual interactions, induced by the Hamiltonian
of (5.90), in doubly-closed shell nuclei. In addition to approximation methods
that highlight the salient features of these elementary excitation modes (exactly
solvable models), we shall also discuss a realistic application to the case of 16 0
as a doubly-closed shell nucleus.
We start from the Hamiltonian of (5.90),

H =Eo + I>a N ( a~aa ) + i a,{J,,,),,6

a
L (a.8IVI-yD}N (a~apa6a")') , (6.1)

where, for the remaining discussion, we shall leave out the reference energy
Eo corresponding to the static Hartree-Fock energy of the doubly-closed shell
nucleus. As a basis for treating the residual interaction [the third term of (6.1)],
we can construct the zero-order wave function of the diagonal part of H. For a
doubly-closed shell nucleus, this basis consists of all Ip - Ih, 2p - 2h, 3p -
3h, ... , np- nh configurations (Fig. 6.1) which we generally call the zero-order

+ ....

lp-lh 2p-2h 3p-3h

Fig. 6.1. Hierarchy of excitations, relative to a closed-shell nucleus, that form the basis for expanding
the actual wave functions .ya(J"). The states are divided according to the number of Ip - Ih com-
ponents: Ip -lh, 2p - 2h, 3p - 3h, .•. (both for protons and neutrons). In this way, a classification
according to unperturbed energy is made at the same time that can serve as a guide in 1rUI1cating the
model space

179
eigenstates "pp. Because of the nonnal ordering part N(ai,.acr), the eigenvalue
for a Ip - Ih state is ep - eh, for a 2p - 2h state ep + ep' - (eh + eh'), etc. We
then obtain the secular problem for detennining the total wave functions of H
by expanding in the basis "p p as

(6.2)

We should note that here {a, [3, ••• } are labels for identifying the zero-order np-
=
nh wave functions completely. Thus for a Ip-lh state the label [3 (p, h, J'tr) or
=
[3 (p, h, J1r, T). Using the wave function (6.2), this secular equation becomes
(Chap. 3)

L c1) b'IH 1[3} =Ecrc;cr) or

p
(6.3)
L {c1)b'IHol[3} + c1)b'IHresl[3}} = Ecrc;cr) .
p
Here, the part including Ho is diagonal and becomes E~P) . 5-yp, with E~P) the
unperturbed energy of the related np - nh basis states. The eigenvalue equation
can still be written as

L { (E~) - Ecr )5-yp + b'IHresl[3}} c~cr) =O. (6.4)

P
In principle, the dimension is infinite. Truncation effects can now effectively
reduce the dimensions to a reasonable number by, e.g., restricting to Ip - Ih,
2p - 2h configurations in light nuclei (160, 40ea, ••• ) and taking up to Ip - Ih
configurations only in the heavy nuclei e08Pb) where the oscillator shells are
much larger.
The matrix equation (detenninantal equation) related to (6.4) becomes, more
explicitly,

(E~1) - E) + {1IHresI 1} {nIHresI 1}

{IIHresI 2} {nIHresI 2}
=0 . (6.5)

{IIHresln}
Two major approximations for treating the doubly-closed shell nuclei are
(Rowe 1970, Ring, Schuck 1980): i) the Tamm-Dancoff approximation ('fDA);
ii) the Random-Phase approximation (RPA); and we shall discuss both methods
in some detail.

180
6.2 The TDA Approximation

If, as a ground state Ig}, we use the reference vacuum state IO} obtained by
filling all Hartree-Fock orbitals for a given doubly-closed shell nucleus with the
A nucleons present, then low-lying excited states will be obtained starting from
Ip - Ih configurations relative to this reference or vacuum state.
Thus, we start from the definitions

Ig} == IO} , (6.6)

and
(6.7)
as the Ip - Ih configurations. (We use the uncoupled basis. Angular momentum
coupling makes all expressions somewhat more complicated when evaluating the
necessary matrix elements, but all essential elements remain.) When carrying out
angular momentum coupling, the particle-hole states

(6.8)

are more conveniently defined by using the creation and annihilation operators
a-:-n and ai, respectively. In the uncoupled representation, the basis (6.7) is more
convenient since the extra phase factors (_I)ji+ m i do not appear to make ex-
pressions more cumbersome.
The secular equation in the 1p - 1h space becomes

HIIiw) =liwlliw} , (6.9)

with the eigenstate 11iw) at the energy Iiw, expressed as

(6.10)
m,i

More explicitly, (6.9) leads to the equation

(emi - Iiw )Xm,i + 2)mi- 1IHres lnr 1)Xn,j =0 , (6.11)

n,j
and
(6.12)

gives the diagonallp - Ih energy denoted in what follows as emi.

We can evaluate the interaction matrix element in (6.11) by using the explicit
form of the residual interaction Hres as

{mi-1IHreslnj-l} =~ 2:= {glaramN(a:a;a6a"()a~ajlg}{a.BIVI')'8}nas

01,13,"(,6
= {mjIVlin}nas . (6.13)

181
Thus the secular equation for the 1p - 1h subspace reads

(cmi -1iw)Xm,i+ 2)mjlVlin)nasXn,i =0. (6.14)

n,j
In coordinate space (r, 0" ) (or even (r, 0" , 7") when including the charge character
in the isospin fonnalism), the matrix element (mjlVlin)nas is the sum of a direct
and an exchange term and can be expressed as

JcP~(1)cp;(2)V(1,2)CPi(I)CPn(2)dld2
-JcP~(1)cp;(2)V(1,2)CPn(1)CPi(2)dld2 , (6.15)

where I, is a notation for r}, 0"1 (7"1). In diagrammatic form, the two terms
of (6.15) can be given as the direct and exchange term, using the methods of
Sect. 5.2. In the direct term, the given hole j-l and particle state n combine to
form the ground state which, through the residual interaction, is broken up again
into a particle-hole state mi- 1 • In the exchange part, the particle states m, n
scatter off the hole state (i-I, j-l) (Fig. 6.2).
Now in the actual calculations in TDA, the electromagnetic transition strength
to the low-lying collective excitations is in almost all cases underestimated. This
finds its origin in the asymmetry which is built in the approximation between
the ground state Ig) and the excited states 11iw) since the latter contain Ip-
Ih correlations whereas Ig) is a static reference state without any particle-hole
correlations. We shall later learn how to remedy this asymmetry within the RPA.
In Sect. 6.4 we will study the specific applications of a TDA calculation in
the case of 160 and thus, at present, will not discuss realistic cases. A general
outcome is that because of the short-range attractive character of the nucleon-
nucleon interaction and since the direct and exchange matrix elements are almost
equal in magnitude, the particle-hole interaction is attractive in light nuclei for
the T =0 states and repulsive in the T = 1 states (Rowe 1970).
This becomes clear when we consider the following. In light doubly-closed
shell nuclei, proton Ip-lh and neutron Ip-lh excitations are almost degenerate
in excitation energy, i.e.,

m V-'_7\ - n
-1
)
DIRECT EXCHANGE
Fig.6.2. The particle-hole matrix element (mi-1IH... lnj-l) (6.13), separated into the direct and
exchange contribution. In the direct term, a Ip - Ih is annihilated by H ... and created again in
another configuration (the Ip - Ih states can even change in charge character through the action of
H ... ). In the exchange term, the particle and hole states are scattered off each other via the interac-
tion: this is possible for equal charge character for mi- 1 and nr 1 only

182
 (6.16)

Having constructed the particle-hole T = 0 and T = 1 combinations in Sect. 5.4

(5.69), one has

(6.17)

Assuming an attractive particle-particle interaction one obtains for the respective

matrix elements

and
... (ph- 1 IHres lph- 1 )" = D, (6.18)

where D and E stand for direct and exchange tenn, respectively. The final
particle-hole interaction matrix elements for the states (6.17) read

(ph -IIHres Iph -1 )T=O =2D - E ,

(6.19)
(ph- 1 IHres lph- 1 )T=1 = -E.
Since for a short-range attractive force D ':!' E, the particle-hole interaction
becomes attractive in the T = 0 channel and repulsive in the T = 1 channel
(D ~ E < 0). These results will help us to construct and study an exactly
solvable model, when we put in the above ingredients on the particle-particle
and particle-hole interaction matrix elements.
Thus we discuss here the solvable model as proposed by Brown and Bolsterli
(Brown, Bolsterli 1959). From (6.11), we can write that

(Iiw - emi)Xm,i = L(mi-1IHreslnj-I)Xn,j . (6.20)

n,j
If we now assume a separable force that for the diagonal particle-hole interaction
matrix elements agrees with the results of (6.19), we can write

(mi- 1 IHreslnj-l) = -XDm,iD~,j . (6.21)

(With X > 0 we get the attractive T = 0 case and for X < 0 we obtain the
repulsive T = 1 case in a simple way.) Bringing this result into (6.20) we obtain

(Iiw - emi)Xm,i = -XDm,i L D~,jXn,j . (6.22)

n,j
In this sum, the variables (n,j) go over allip-lh configurations and a constant
value (called Nix) will result. Thus we can solve (6.22) for the Xm,i as
NDmi
Xm,i = (emi - h.w) . (6.23)

183
 The coefficient N can be determined through the normalization condition for
the wave function expanded in the 1p - 1h basis since

L IX m ,i12 = 1 , or
m,i

N-2 _ ~ IDm,i 12
-L..J 2'
(6.24)
m,i (Iiw - emi)
A dispersion relation for the eigenvalues Iiw is then obtained by summing

~ D*m,i X m,i = - ~ xlD m,il 2 ~ D* X

L..J. L..J. Iiw _ emi ~ n,j n,j, (6.25)
m,1 m,1 n,J

or,

(6.26)

The roots Iiw of this equation are now the crossing points between the left
hand side X-I and the right hand side which is an expression which goes to
infinity at the values Iiw =emi (the unperturbed Ip - Ih energies in the system,
Fig. 6.3).
If we take the degenerate case that all Ip - Ih energies emi == em - ei are
equal to e, all eigenvalues remain degenerate at Iiw =e except one:

Iiw =e - LxIDm,iI2. (6.27)

m,i

Thus, for X > 0 (attractive p - h interaction), the one state that is affected in
energy can become much lower in energy than the unperturbed Ip - Ih energy
(T = 0 states) whereas the case of X < 0 (repulsive p - h interaction) is the

Fig. 6.3. The TDA secular equation (6.26) illustrated in a schematic way. The left-hand side of (6.26)
(1/ x) is a straight-line. The right-hand side represents the curve with a number of roots (where E .. .
intersects the l/x line). The unpenurbed Ip-lh energies em; are illustrated here as el, e2, e3, .. ..
The line (X > 0) gives the low-lying collective isoscalar state (-), the line (X < 0) the high-lying
collective isovector state (0)

184
 ~I
ATTRACTIVE ~
II
-~-----
III 1'lw
EIII
- - ---lIL--
,II
II
',II REPULSIVE
Ii
III

Fig. 6.4. Limit of Fig. 6.3 where all em; become degenerate at a single point em; = e. The collective
root is given (its energy liw) by (6.27) and becomes attracted (e) or repelled (0) in excitation energy,
starting from the e value. All other roots remain degenerate at 1iw = e

one where the energy of the linear combination of Ip - Ih configurations gets

pushed up in excitation considerably above the unperturbed value of c (Fig. 6.4).
In the X-I> 0 (or T = 0) case, one can easily check from the wave functions
that the proton and neutron particle-hole excitations move in phase whereas
in the X-I < 0 case (T = 1), proton-neutron particle motion is out of phase
(Brown, Bolsterli 1959). It is the latter motion which goes up in energy in light
nuclei even above the neutron emission threshold and so becomes a resonance.
In particular, in light nuclei 160, 4OCa, ... , the J1r = 1- T = 1 resonance or giant
dipole resonance is obtained and can thus be described as a linear superposition
of Ip-lh configurations with all amplitudes X m ,; having the same phase so that
coherence in the electromagnetic decay or excitation probability results (Speth,
Van der Woude 1981). In this way, a connection between a purely collective
out- or in-phase motion of protons versus neutrons (Fig. 6.5) and a microscopic
Ip - Ih description could be obtained (Brink 1957). In Sect. 6.4, we discuss
the realistic case for 160. In Fig. 6.5, the simple, schematic representation of the
collective wave function (in a macroscopic and a microscopic way) is given as
well as the concentration of El excitation strength (expressed in % of the total
Ip-lh dipole sum rule) for 160 comparing both the unperturbed and interacting
Ip - Ih 1- spectra (Elliott, Flowers 1957).

185
 COLLECTI VE,
MACROSCOPIC

m +\--t9'+"~OLLECTIVE'
It V It V

MICROSCOPIC

L.o
UJ
:5
a::
20
::E
~ 80
UJ
5
IL
60
is l.0
'$.

I 20~~~~~-L~~~~~
Fig. 6.S. DIustration of the macroscopic and microscopic point of view in describing the nuclear
giant electric dipole collective state. In the collective model, the proton and neutron distributions
undergo a translational, dynamic oscillatory motion (electric dipole). In the microscopic model, a
linear coherent superposition of proton Ip - Ih and neutron Ip - Ih states makes up the collective
state (repulsive part of the IDA dispersion relation corresponding to X < 0). Some Ip - Ih config-
urations are given as an illustrative example. In the lowest pan, the unperturbed 1p - 1h spectrum
and interacting Ip - Ih spectrum in 16 0 (for 1- levels) is given. The height gives the % of the
dipole sum rule (proportional to the El transition probability 1(0IlD(El)1I1-)12 )

6.3 The RPA Approximation

We poinled out in the TDA approximation that a basic asymmetry between

the ground state Ig > and the excited Ip - Ih IIiw > states exists. In the RPA
approximation, we use a ground state that is no longer described by the reference
state 10 > but is treated on an equal footing with the excited states, i.e., p - h
excitations or correlations are also present. We leave out a precise description of

186
the new ground state but will define operators that connect the RPA ground state
to the excited states (Rowe 1970, Ring, Schuck 1980).
We define the creation operators Q~(m, i)
Q+(
a m,z =
.) - X(a) +
m ,iamai -
y(a) +
m ,iaiam, (6.28)
and the corresponding (Hennitian conjugate) annihilation operator Qa(m, i) as

(6.29)

The amplitudes Xm(a)." and y(a~

m,1
are in general complex and only when the ym(a), •.
amplitudes disappear is the IDA recovered from the RPA. The ground state is
now defined via the condition
(6.30)

for all a(m,i). The excited states are created as

(6.31)

From the definition of the creation operator Q~(m, i) in (6.28) and Q~(m, i)IO},
we extract the infonnation that the single-particle state m has to be occupied in
the state IO} in order for am to be able to annihilate that particle, but then a-:;'
cannot create another particle. So, arguing in tenns of a pure fennion language,
we come in conflict with the Pauli principle which is partly violated from the
definition of the operators Q~ and Qa. In a somewhat oversimplified way, we
could picture an RPA ground state as a configuration where besides 10}, 2p -
2h, ... , excitations outside 10} are present at the same time such that particles
(m, n, ... ) above the Fenni level can be annihilated and particle states below the
Fenni level (i,j ...) can be created (Fig. 6.6).

10> + L 2p-2h+ Ll.p-l.h+ L6p-6h + ...

Fig. 6.6. A possible description of the RPA ground-state IO} which is defined through (6.30). In order
to be able to create excited states via the operator Q~(m, i) of (6.28), 2p -2h, 4p -4h, 6p -6h, ...
excitations have to be present, relative to the system with a sharp distribution of nucleons up to the
Fermi level

187
 In RPA, one implies boson commutation relations for the Q~ and Qa opera-
tors and the approximation is therefore also called the Quasi-Boson Approxima-
tion (QBA) (Lane 1964), where one enforces the commutator conditionl

(6.32)

The derivation of the RPA secular equations is somewhat more complicated

than the derivation of the IDA secular equation. Imposing

(6.33)

one obtains the set of coupled equations for the X~)i and y~a~ amplitudes which
form the RPA secular equations. "

(emi - Ea)X~;i + L)mjlV/in)nasX~~J

n,j
+ 2:(mn/V/ij)nasY~~ =0,
n,j
(6.34)

n,j
+ 2:(ij/V/mn)nasX~~} = 0,
n,j
or, in matrix form

(6.35)

Here we have defined the matrices A and B via their elements labeled by the
particle-hole indices (m, i) as row and (n, j) as column indices

Ami,nj == emibmi,nj + (mj/V/in)nas ,

(6.36)
Bmi,nj == (mn/V/ij)nas ,
or
A mi,nj -= (mt.-I/H/ nJ.-1) ,
(6.37)
Bmi,nj == (( mi- I ) (nj-I) /H/D) .
Thus, the dimension of the RPA secular equation, given in (6.35), has become
twice that of the IDA. For each solution formally written as [X, Y] at an energy
Em a solution [Y*,X*] exists at the energy -Ea. The latter values are on
the "unphysical" branch since only positive energy eigenvalues can be taken to

1 This commutator relation has to be considered as an expectation value relative to the RPA ground
state.

188
 (I) (I)
Fig.6.7. Creation (I) [or annihilation (ll)] of a 2p - 2h configuration (mi-l)(nrl) out of the
ground state. This process. described by the matrix elements Bmi,nj is shown by the corresponding
Feynman-Goldstone diagrams

represent the actual excited states in atomic nuclei. IT we set the matrix B == 0,
the IDA secular equations result as a particular case. This means that those
processes expressed by matrix elements of B, i.e., the creation of a 2p - 2h
configuration from the static ground state IO}, can be obtained (Fig. 6.7).
In the same way as in IDA, a separable interaction can be used with the
properties of attractive matrix-elements (in T = 0 channel) and repulsive matrix
elements (in T = I channel) (Rowe 1970).
Again we take

(mi-11VInj-1) = -XDm,iD:,i ' (6.38)

and obtain from (6.34)

(6.39)

The expression between square brackets [...J has a constant value and is denoted
by N / x. Thereby, we obtain the result

NDmi
Xm,i = (emi - hw) ,
(6.40)
ND:'ni
Ym,i = (em,. + fW.l
~. -) .
Multiplying (6.39) by D:'n,i' respectively Dm,i, and summing over all (n,j)
particle-hole pairs, we obtain the dispersion equation for the energy eigenvalues
1iw as

(6.41)

189
 ~ PHYSICAL BRANCH Itlw>OI
W//fi'$$///.,
Fig. 6.S. The RPA eigenvalue equation (6.41) is shown in a schematic way. The left-hand side in
(6.41) is equal to the TDA equation (see Fig.6.3). The right-hand side is quadratic in li.w and re-
sults in both positive and negative solutions for the energies li.w. The physical branch is the sector
li.w > O. Because of the doubling of solutions, possibilities exist that the lower, collective solution
(e) becomes imaginary (no intersection with the l/x line)

This equation resembles the IDA secular equation. However, we have a

quadratic expression in Iiw in the denominator and also an extra factor of 2 in
the numerator. The factor 2 indicates that for the same strength X, the lowest
(or highest) RPA eigenvalue for Iiw is quite a bit lower (or higher) than the cor-
responding IDA eigenvall!e. We represent schematically the solutions to (6.41)
in Fig. 6.8. For the degenerate cases where all emi = e, one gets the eigenvalue
equation (6.41) into the form

(6.42)
m,i

If we take the particular case that (X > 0)

(6.43)

one obtains the lowest eigenvalue (RPA) at Iiw = O. On the contrary, for the
same strength X of (6.43), the lowest TDA eigenvalue occurs at Iiw = e/2. This
application again indicates that in the RPA, the lowest-lying collective in-phase
motion of protons and neutrons comes lower in energy. It can even become
so that for particular strong collective states (e.g., 3- in 40Ca) the lowest root
becomes imaginary. In Fig.6.9, we indicate besides the graphical solution to
the RPA dispersion relation of (6.41), the particular case of a strength in the
degenerate case given by (6.43).
We now define the quantity S (the "non-energy weighted" sum rule for the
relevant one body operator expressed by Dm,i),

S= LID m ,;j2. (6.44)

m,i

190
 \
TDA

Fig.6.9. Degenerate limit (all emi = e) of Fig.6.S for the RPA approximation. We illustrate the
particular situation that the value of X = XO is given by (6.43) and the lowest (collective) RPA root
equals 1i.w =O! ifullline). We draw on the same figure the IDA solution with the same value of x.
Here, the collective IDA root occurs at 1i.w(TDA) = e/2. So, for a given strength X, the RPA root
becomes lower than the corresponding IDA root

We obtain the approximate wave function for the collective excitation

(6.45)

Here, one observes that in most cases the amplitudes

y(l). ~ X(l) .
m,1 m,I'

because of the extra factor

(c -liw(l»)
2c ~ 1, (6.46)

in actual cases.
The transition probability of this collective state to the RPA ground state then
becomes generally

" [X~ ,;Dm ,; +

(1iw(l)IDIO) = '~ y.: ,iD~ ,i] = N*X . (6.47)
m,1

For the degenerate case and its collective, low-lying (or high-lying) state, the
transition matrix element (6.47) simplifies to

(6.48)

Using the expression for the normalization coefficient N and its relation to the
solution, this gives the transition probability
191
 (6.49)

Thus, the collectivity of this transition is expressed by the coherence of S in all

particle-hole contributions with the same sign (which one also obtains in the TDA
case), but now multiplied with an extra factor e/Iiw(l) which easily becomes 2.
This extra enhancement in the transition probability comes in particular from
the fact that the RPA ground state IO} contains p - h correlations and that an
excited state 11iw} can decay into the ground state in different ways by using a
one-body transition operator D.
In Appendix H we shall discuss a slightly more elaborate two-group RPA
model which, however, bears a resemblance to many actual situations in real
nuclei.
Before discussing a detailed study of 160, we illustrate the results of a study of
the 3 - collective isoscalar state in 208Pb as was studied by Gillet et al. (Gillet et al.
1966). As a function of the dimension of the lp-lh configuration space both the
lowest TDA and RPA eigenvalues are illustrated in Fig. 6.10. One first observes
that the RPA eigenvalue is always lower than the corresponding TDA eigenvalue
but that at the same time, for a given strength of the residual interaction (in this
case a Gaussian residual interaction has been used), the dimension of the 3-
configuration space strongly affects the particular excitation energy.

TDA

~ 3.0 Fig. 6.10. The position of the octupole Slale in 208Pb using
IDA and RPA as a function of the dimension of the config-

I RPA uration space. On the abscissa, we give both the number and
unperturbed energy of the 1p-l h configurations. A Gaussian
interaction V(r) = Yoexp(-,.2/1'2)[W + MP(r) + BPq +
H Pqp(r)] was used with Yo = -40MeV, I' = 1.68fm and
exchange admixture W, M, B and H as discussed in (Gillet
- EunperiurbedlMeVI-- 1966) [Fig.4. adapted from (Gillet 1966)]

6.4 Application of the Study of Ip - Ih Excitations: 16 0

For the doubly-closed shell nucleus 160, the low-lying excited states will be
within the space of Ip - Ih configurations where, due to the charge indepen-
dence of the nucleon-nucleon interaction and the almost identical character of
the nuclear average field, proton Ip - Ih and neutron Ip - Ih excitations are
nearly degenerate in unperturbed energy (Fig. 6.11). It is the small mass differ-

192
 Fig. 6.11. The model space used to carry
16 0 out the 1p - 1h calculations discussed in
8 8 Sect. 6.4. The full (sd) space for unoccupied
It V and (sp) space for occupied configurations
ld 3/2
are considered in constructing the relevant
energy matrices
25 1/2

ldS/2

1Pl/2
l P312

151/2

ence between proton and neutron and the Coulomb interaction that induce slight
perturbations on the isospin symmetry of the picture. Although we could work
in an isospin fonnalism we shall use the explicit difference between proton and
neutron 1p - 1h configurations and later check on the isospin purity of the eigen-
states. Since we know how to couple the proton Ip - Ih and neutron Ip - Ih
configurations to definite isospin,
1
Iph- 1; J1r)1r = J2 [lph- 1 ; J1r, T = 1) + Iph- 1 ; J1r, T = 0)] ,
(6.50)
Iph- 1 ;r)v= ~[lph-\r,T=I)-lph-l;J1r,T=O)],
we can, in the linear combination of the wave functions expressed in the (11", v)
basis, substitute expressions (6.50) and obtain an expansion in the isospin J1r, T
coupled basis. Fonnally, the wave function

li;r) = L c(ph- 1(e);iJ)lph- 1;J 1r )II' (6.51)

p,h,1I

where i denotes the rank number labeling the eigenstates for given J1r and e the
charge quantum number (e == 11", v) can be rewritten in the basis

li;r) =L c(ph-l;iJ,T)lph-l;r,T) , (6.52)

p,h,T

where now for each (p, h) combination we sum over T = 0 and T = 1 states.
From the diagonalization (see later) it follows that low-lying states are mainly
T =0 in character and the high-lying ones are T = 1 states.
The Hartree-Fock field for 16 0 was detennined by using the SkE4* in-
teraction (Chap. 8) with its proton and neutron single-particle orbitals, energies

193
 UNPERTURBED DIAGONAL TDA
160 pOl
8 8

/
, 1277 1278
T =1 (68%)
v 11.33/
..- T =0 (68%)
(2S1/2.1P~J2J 0- n 11.00 10.91
10.22'"
Fig.6.U. Using the SkE4* force (see Clap. 8) in detennining both the Hartree-Fock field (single-
particle energies €n/j) and the residual interaction, we present: (i) the unperturbed proton and neutron
0-(281/21PIA) configurations, (ii) the energies, when adding the diagonal ph- 1 matrix elements
(repulsive), (iii) the final TDA results after diagonalizing the energy matrix for J1< = 0-. lsospin
purity is also given in T (%)

and the same interaction SkE4* was used as a residual interaction (Waroquier
1982, Waroquier et al. 1983b). We illustrate schematically that ej(7r):f ej(V) in
Fig. 6.11. This is even more clear for the case of those 1P - 1h excitations that
form the 0- state. The (281/2, 1pil~)O- configurations are given in Fig. 6.12
with a net difference in unperturbed energy of L1e = 1.11 MeV. The diagonal
(ph - 1, 0-) matrix elements are slightly repulsive and finally, the interaction in
non-diagonal form is small. Thus, very pure proton ph- 1 and neutron ph- 1 states
result. When expressed in the basis of (6.52), this gives very impure isospin wave
functions (Fig. 6.12) with 68 % of the main component only.
Relating to the 1- states within the 11iw (L1N = 1) space, we can only form
the following 1- configurations where

1281/2(lpl/2r1; 1-) ,
1281/2 (1P3/2rl; 1-) ,
11d3/2(lPl/2r1; 1-) ,
11d3/2(1P3/2r1; 1-) ,
11ds/2(1P3/2r1; 1-) ,
where both proton and neutron 1p - Ih configurations can result, leading
to a 10-dimensional model space only. Using now the methods discussed in
Sects. 6.2, 3, one can set up the IDA and RPA eigenvalue equations. The results
for 0-, 1- ,2-,3 - and 4- are given in Fig. 6.13, where we give in all cases:
i) the unperturbed energy with the charge character of the particular ph- 1
excitation,
ii) the IDA diagonalization result (with isospin purity),
iii) the RPA diagonalization result,
iv) the experimental data.

194
 Ex(MeV)

SkE4'"

TDA unperturbed RPA EXP TDA unperturbed RPA EXP

~) T:l T:l (79',l, ~

" T:1
22 _ \ n ,r-- I

\ .. n ... :
•
- 22

20
- ~:~~~ T:O (73'/,0_~J-~ -
I- -----v
T=1 (60'10): ld1 lP~'
2 2
\ T=1
,,.--- I:L.!m:! 1d3
2' l P2'
3'\
,!..:.L- - 20
T= 1
-
I- T=O (5409\\
..
,IT =0 " ..1 1" ,
,'---- \,1""2 1P2 /
181- In : -"--....~ - 18
)C' p .>( T=O (82·!')/.~~... , /.~
I-
T=O (98'k)l
,. 53'1\
T,O (590 '1, 2i;;F'-'- -
16 I- ld_ lp_
2 2
'. T:O
==--- ld5 T:O
'2,- - 16
- <'..
I '

T:l
'n
(52·.4)~'.p T =1 -
14 - T:l
,....;..;~-
- 14
T:l (99'.4) T:l ld 1 1p .J:1
~-
12 - - . . ;-- T:O 2 2 I.:.Q _____ 12
- \I '
I
T:l (97'/.)
---'\\ A--- ,/~
T:l

- , n
)<:~:1
:
}<L-->: -
10 - / ld§ lpl"' \
! 2s1 lpr
J;9____ - 10
- :
, 2 2 '
\ ,: 2 2
\
\, -
8 _ M (99# \
, T:0(97'.10)/ ~
- 8
- ----
T:O

6_ CD :
~ - 6

SkE4* Sk E4 >t

TDA unperturbed RPA EXP TDA unperturbed RPA EXP

24 - -24
_ T:l
----.,
(68%)
ld 3 l pl ,.---
•
-
22 - \~ 2 ,- - 22
20 _
_ T:l (74'101 \'~'l T:l
):~
-
T:O,I (SO'lo)\" • p , T: 1 (94'10)
~ /r--
T:l ,-
_____ E - 20
T: 1 (647.~, 1
3" / T=1 \,
- ~:"IS2 lP"2,' r--- /
/
~-
18 T:O !73%l: '.' .---,!..-.. /T:O
I
_T:_0_(9_4°_10):"' _ _~ • T:O 4-
/
ToO 4- - 18
- T:O (81 'Io)1::,,~:,,:c,:g~ 4- • -)-.---
- 1d~ lp!' ' ld§ 3.
2 l Pi
1 -
16 - hI (76%) 2 2 T:l - 16
-------. .---
- \)c,~"" -
14 _ " p ,J(,
- 14
T: 1 (8S"oV ld11Pl' \ T:l
- ===>, 2 2 ,'= T:l
0:----\
(68'/.)
:~
T:l O· ~-
12 -
T:O 176%1\ n
T: 0 !86%1..,>< P .-":.. ~
"
T:O
I ,
- 12
- 1<1.§. lpll T:O (68'1o)~,--..-J..~ 1:2.........Q:" -
2 2

o
O· • -'
10 - -10
25 1 lp!:' r::i\ D\
- _T:_O__ ~ 2 - \S!..J+\!ij
8~ ________ ____________ ~~ ~ ____________~~~______~ 8
Fig. 6.13. Negative parity states in 160. Comparison between the IDA and RPA results using SkE4*
and the data is given. The unperturbed structure is indicated (n: neutron, p: proton). The isospin pu-
rity (%) is also given. Experimental levels, drawn with a dashed line, have predominant 3p - 3h
character [taken from (Waroquier 1983b)]

195
Table 6.1. The wave functions for the 3 - T = 0 and T = 1 states in 160, obtained using TDA
and the SkE4· interaction. The wave functions are expanded in both a charge basis, giving the
Iph -1; J""} II (/1 == p, n) components and an isospin basis, giving the Iph -1; JlI' ,T) components.
The total isospin purity is also expressed in the last colunm
p 117/2 I/s/2 1dsl2 It412 1ds/2 isospin
h- 1 18 -1 18 -1 1 -1 1 -1 1 -1 purity
112 1/2 P3/2 P3/2 Pl/2
3- p -0.01 +0.01 +0.15 -0.20 +0.72
Ell: =7.88 MeV n -0.02 +0.01 +0.15 -0.18 +0.60
T=O -0.03 +0.02 +0.22 -0.27 +0.93 99 lifo
T=1 +0.00 -0.00 +0.00 -0.01 +0.08 1%
3- p -0.03 +0.00 +0.25 -0.05 -0.64
Ell: =13.lOMeV n +0.03 +0.00 -0.13 -0.04 +0.71
T=O -0.00 +0.00 +0.09 -0.06 +0.05 1%
T=1 -0.04 -0.00 +0.26 -0.01 -0.96 99%

One observes, in particular for the lowest 1- , 3 - levels, the very pure T = 0
isospin character. Since the unperturbed proton and neutron 1p-1 h configurations
are almost degenerate, the diagonalization induces a definite symmetry (T = 0
or T = 1) according to the lower- or higher-lying states. In the case of 3-
(T = 0) and 3- (T = 1), we also give the wave functions in both the charge
quantum number basis and in the isospin quantum number basis (Table 6.1) to
illustrate the above general discussion on isospin (Sect. 3.4) and the discussion in
the beginning of this section. In Chap. 8, when discussing some state-of-the-art
shell-model calculations, we shall discuss some more examples as well as outline
the self-consistent methods used in determining both the average field and the
residual interaction.
One could, of course, discuss other topics related to the IDA and, in partic-
ular, to the RPA method. One can show that the RPA equations correspond to
the small amplitude limit of the time-dependent Hartree-Fock method (harmonic
approximation) and thus the term quasi-boson approximation is related to the
dynamics described by the RPA eigenvalue equations (Lane 1964, Ring, Schuck
1980, Rowe 1970). The RPA secular problem reduces to a non-Hermitian matrix
problem which we will not discuss in detail since it is outside the scope of this
book.

196
7. Pairing Correlations: Particle-Particle Excitations
in Open-Shell Nuclei

7.1 Introduction

Until now we always considered a sharp Fenni level with a very particular
distribution of level occupation probabilities. Up to the Fenni level, all levels
are fully occupied, above all are unoccupied. This confonns with a Hartree-Fock
static ground state with Ip - Ih, 2p - 2h, ... excitations as elementary modes
of excitation (see Sect. 6.2 for a discussion on the IDA approximation). A short
range force of sufficient strength is now able to scatter couples of particles across
the sharp Fenni level. For nuclei at or near to doubly-closed shell configurations,
this leads to a "diffuse" ground state with 2p-2h, 4p-4h, ... correlations, much
like the RPA ground state IO} as discussed in Sect. 6.3. When, however, many
nucleons outside closed shells are present, this pair scatter can, under certain
conditions, lead to a stable, smooth probability distribution for the occupation
of the single-particle orbitals. So, the pairing correlations set in and modify
the nuclear ground state distribution of nucleons in nuclei in a major way (see
Fig. 7.1).
In Chap. 7, we shall discuss these modifications in detail, starting from the
most simple, solvable model of a single j-shell with n particles, interacting with
the pairing interaction of Sect. 5.2 and going on to detailed pairing calculations.
Before attacking the new aspects brought into the nucleus by means of the
pairing correlations, we shortly recollect the major aspects that detennine the

~I
E(MeV) E
(MeV)
I I Vfh'Hh____--t
PAIR
SCATTERING
Fig. 7.1. Schematic distribution of nucleon pairs
in one case where all nucleons occupy the low-
lying orbitals pairwise up to the Fermi level
(closed shell) and in another case, where a
I .. smeared out pair distribution occurs
OCC PROS. OCG.PROS

197
nucleon motion in the nucleus (Rowe 1970). We have indicated how the Hartree-
Fock theory has succeeded in a separation of the A-body nuclear Hamiltonian
into
H = Ho+Hres, (7.1)
where Ho becomes the Hartree-Fock Hamiltonian HHF when going to the appro-
priate single-particle basis. The importance of this separation in (7.1) depends
on the strength of (Hres) compared to (HHF). In light nuclei like 16 0, 4OCa, ...
the energy gap between major oscillator shells, nwo, is such that
nwo ~ (Hres) . (7.2)
So, the Hartree-Fock ground state remains stable with respect to the residual
interactions and independent particle-hole excitations out of this Hartree-Fock
ground state are well defined (see Fig. 7.2).
For non-closed-shell nuclei (toSn with A 9t 110 - 120), the Hartree-Fock
solution to the ground-state nucleon distribution can sometimes lead to a small
gap so that

nwo <: (Hres) . (7.3)

In those cases, instability against pair excitations or deformation of the nucleus

could set in. We know that the two-body interaction V(i,j) cannot completely
be absorbed into the average Hartree-Fock field. We can ask what part exactly
goes into the average field.
In the Hartree-Fock approximation (Sect. 3.1), we have shown that

U(ri) = JV(r;,rj)e(rj)drj . (7.4)

If we now make a multipole expansion of the two-body interaction as

V(r;,rj) = Lf>'(r;,rj)
>',1'
2:: 1Y,{'(r;)Y{ (rj) , (7.5)

and substitute this expression in (7.4) one obtains

U(r;) = LUf(r;) , with (7.6)

>',1'

Fig. 7.2. The Hartree-Fock ground state

for doubly-even nuclei is rather stable
against 2p - 2h excitations (.\ <: 1)
in those cases that Awo (separation be-
tween major oscillator shells) is large
compared to the residual interaction en-
( 21'll1\, I ergy (Hnos)

198
 Uf(ri) =Yt (r;) 2:: 1JJA(r;,rj)Y{ (rj)e(rj)drj,
=Yt(r;)a~(r;) . (7.7)

So, also the average field presents different multipoles according to the folding
of the density with corresponding multipoles of the two-body residual interaction
V (r ;, r j ). For A = 0, we recover the spherical, central static field component.
The A = 1 part relates to an overall shift of the centre-of-mass of the nucleus
and does not relate to specific "internal" modes of the nuclear field. Higher
multipoles, A = 2, A = 3, ... represent the quadrupole, octupole, ... deformation
of the average field, respectively. The low multipoles of V(r;, r j) (A = 0, A = 2,
A = 3) are called the field-producing components (Bohr, Mottelson 1975, Ring,
Schuck 1980).
In competition with the "long-range" components of U(ri), short-range com-
ponents of the interaction scatter nucleons out of their individual orbitals. Con-
tributions to all A are obtained from short-range forces such as the 8( r i - r j)
zero-range force or the pairing interaction (defined in second quantization in
Sect. 5.2). These forces show the tendency to correlate nucleons in zero-coupled
pairs (J1r = 0+) and thus restore the spherical symmetry in the nucleus. This cor-
relation is well-known from the energy spectra in nuclei with just two nucleons
outside a doubly-closed shell configuration. We show, in Fig.7.3, the case of

- - - ....
(~ 11~/2!---1·
--().&) "--10' - - - - - - - l'
' , - - - - - - - 10'
-7.5

- --8'
---.. . - a• ..------- 8'
~
--6.0-
--6'
--L'
(29gll~~~:====&;:"-------
'~--L' -------- L
6:
>-
C)
---.?', Fig. 7.3. Typical example of the
a:
uJ
z \ nucleon-nucleon pairing prop-
uJ ' - - - - - - - 2'
C) --2' erties in a heavy nucleus 2~gPb
z with 2 valence neutrons outside
is
z- 8.5 f- the 208Pb doubly closed-shell
iii --,0'
, core. On the left hand side, the
\
experimental spectrum is indi-

H \H'N K·/o.f·firJ
';---v---J ~
cated. On the right-hand side
various theoretical spectra are
-9.01- (AI
,--,
\ (B) (el given, illustrated by the diagrams
--0' \
in columns (A), (B), and (C).
\
\
Here, (C) represents the more
210 Pb '--0' realistic situation, (A) the "bare"
EXP. 82 128 interaction (see Herling, Kuo
1972)

199
 E
(Mev
00 o even-even nuclei
0 x odd-A nuclei
0 00

1.0 00 o
0 o 0
0
000
o
0
o 0

0.5

x Xx x X

x x X
x_x_
X
X xx: X

0.1 x. x xX
x XX
)(
x x

150 170 190 210 230 250 A

Fig.7.4. Energies of the first excited intrinsic states in defonned nuclei as a function of the mass
number. The solid line represents the average distance between intrinsic levels in the odd-mass nuclei.
The figure contains data in the 150 < A < 190 and A > 228 regions. We have not included the
low-lying [( =0 states in even-even nuclei in the Ra, Th region since these states represent collective
odd-parity oscillations (taken from (Bohr, Mottelson, Pines 1958»

2~gPbl28 with two neutrons outside the 208Pb core, moving in the 299/2 orbital as
the lowest orbital. A comparison with the theoretical two-particle spectra as de-
termined by Herling and Kuo (Herling, Kuo 1972), using the Hamada-Johnston
force, is carried out.
This particular pairing effect was recognized before and introduced by Racah
in 1942 (Racah 1942a, 1942b), in the seniority scheme in atomic physics. Later,
the 0+ property of even-even nuclei throughout the whole nuclear mass region,
was recognized by M.G. Mayer in 1950 (Mayer 1950), and applied to the nucleus
by Flowers, Racah and Talmi in 1952 (Flowers 1952, Racah, Talmi 1952, Talmi
1952). Pairing was incorporated in a more elaborate way using the BCS theory
of superconductivity as applied to even-even atomic nuclei (Bohr et al. 1958)
from which we take their original illustration of the energy of the first excited
state in deformed nuclei (intrinsic state) for both odd-mass and even-even nuclei
(see Fig. 7.4).
We shall treat the salient features presented by not just one extra pair but by
many pairs outside closed-shell configurations.

7.2 Pairing in a Degenerate Single j-Shell

Starting from the above pairing aspects, it is tempting to use an interaction that
has this pairing property of acting only in J1r =0+ states. Knowing that a given

200
two-body interaction, given in coordinate space can be written in a second-
quantized fonn

(7.8)
OI,fJ,'Y,6
we can define an interaction Hamiltonian, immediately in second quantized fonn
as (see Sect. 5.2)

(7.9)

In the two-particle subspace (m-states), H has the matrix representation (with

j+! rows and columns)

11 11 11 .. . .J
H=-G ( . .. . (7.10)
. . .
. . .
It can be shown that the state

is the lowest energy eigenstate of (7.10). Here 10) denotes the closed-shell wave
!.
function and il is the shell degeneracy or il == j + The energy of the state
(7.11) is Eo = -Gil.
To solve now for the n-particle system the pairing interaction problem, we
define the pair creation operator

C'+_
~"-
1
1
-.,fli L(- l)j+m+a· 1m
a·+
1- m '
(7.12)
m>O

which creates the two-particle J1I" =0+ state. Using this operator S1, the f:lamil-
tonian (7.9) can be rewritten as

H = -GilSjSj . (7.13)

The commutation relations are

[Sj,Sj] = 1- ~ , (7.14)

where n is the number operator

n= Laimajm = L (aimajm + ai-maj-m) . (7.15)

m m>O

201
A class of eigenstates of H [see (7.9)] are now given by

[H, S;] = -GS;(t1 - n) = -G(t1 - n + 2)Sj . (7.16)

Starting from zero valence particles, one gets

HSjIO) = -Gt1SjIO) ,

H( S; flO) = -2G(t1 - 1)(S;flO) ,

n/2 G ( )n/2
H ( S; ) 10) = -"4 n(2t1 - n+2) S; 10) . (7.17)

Here all particles are coupled pairwise to J1r = 0+. They are described as seniority
v = 0 states, where the seniority quantum number v denotes the number of
unpaired particles. In shorthand we can write
n/2
In, v = 0) == ( S; ) 10) , (7.18)

and
G
Ev:{J(n) = -"4n(2t1 - n + 2) . (7.19)

One can generalize this procedure in adding to the Sj operator, the t1 - 1

operators Bj which create pairs coupled to angular momentum J(J t= 0) defined
as

Bj = L(-l)i+ m(jmj - mIJO)ajmaj_m . (7.20)

m>O

They obey the relation

[H,Bj] 10) =0. (7.21)

So we can construct a set of seniority v = 2 states

HBjIO) == Hln = 2, v = 2, J)= 0,
HSjBjIO) == H14, 2, J) = -G(t1 - 2)14,2, J) ,

<n/2-1) G
H ( Sj ) BjIO) == Hln, 2, J) = -"4(n - 2)(2t1 - n)ln, 2, J).
(7.22)
For v > 2 we cannot continue in this way since an overcomplete set of states
BjBjIO), ... arises. Still we can calculate the energy since a state with maximum
seniority v = n has energy zero or Hln, v = n) = O. We can now calculate
successively
202
 Hln, n - 2} = H (S;) In - 2, n - 2} = -G(n - n + 2)ln, n - 2} ,

Hln,n -4} = H(S;Yln -4, n -4} = -G(2n - 2n+6)ln,n - 4} ,

<n-V)/2 G
Hln,v} =H ( S; ) Iv,v) =-"4(n - v)(2n - n - v + 2)ln, v} ,
(7.23)
giving the general expression for the spectrum
G
Ev(n) - Eo(n) = "4v(2n - v + 2) . (7.24)

~-1 0
::E
>-
<!>
ffi
z
UJ

C>
Z
Ci
~-2 0
Fig. 7.5. The spectrum for a pure pairing
force within the 1h 11 /2 orbital i.e. for the
(lh11/2)n spectrum with seniority v = 0,
v = 2, v = 4 and v = 6 as a function of
particle number n. The pairing strength G
-3.0~_~_--'--~--7--:'::----:-: was chosen as G =0.25 MeV (see (7.24»
o

The spectrum relating to (7.24), is illustrated in Fig. 7.5 for the Ihll/2 shell
where we steadily fill the orbital with 2, 4, 6, 8, 10, 12 particles. The pairing
strength G was taken as G = 0.25MeV. Here, one again observes that the
J1r =0+ state is the lowest in all cases, next come the states with one "broken-
pair" (see Sect. 7.6) (seniority v =2), etc. It appears that the binding energy of the
pairs is maximal when the orbital is half-filled. It is also remarkable to observe
that the energy difference (7.24) is independent of n as shown in Fig. 7.5.
For a more realistic case: the N = 50 single-closed shell nuclei where the
proton number represents 4 holes, 6 holes, 8 holes, 10 holes respectively in the
Z = 50 proton core, the pairing properties are experimentally well established
(see the O+, 2+, 4+, 6+, 8+ spectra) (see Fig. 7.6) (Sau et al. 1983).

203
 90 Zr 921010 94 Ru 96 Pd

EXPERIMENT EXPERIMENT EXPERIMENT EXPERIMENT

110'1
5.0
112'\
1101 11r1

t
(91 (91
Ixt)
4.0
ITI
8· In
:>II r:
~
~
0::
30 4·

8· 8·\
W
Z 6· 15"\>
W 8·
5- 6~ 6·
z 5-
4·
Q 4·
~ 2.0 4·
O·
~
w l 2· 2·

/'0

0.0 0·_ __ 0· _ __ 0· _ __ 0· _ __

Fig. 7.6. The experimental spectra for the N = 50 single closed-shell nuclei 9OZr. 92Mo. 94Ru. 96Pd
=
up to an excitation energy of E", ':!!' 5 MeV. From 92Mo on, the seniority v 2 spectrum of nucleons
moving in the 199/2 orbital is clearly observed (taken from (Sau et al. 1983))

7.3 Pairing in Non-Degenerate Levels: Two-Particle Systems

If we now consider, first 2 (later n) particles in a number of non-degenerate

orbitals much of the simplicity of the above exact solvable model gets lost but
the main results of a possible classification in the energy spectra according to
seniority v = 0, v = 2, ... remain approximately realized in actual nuclei.
So, let us suppose that the different single-particle orbitals are such that
..dca < G on the average. Moreover, for the short-range interactions discussed
before, the J1r = 0+ matrix elements are attractive and much larger than the other
J1r matrix elements.
We shall, in order to make the discussion applicable to deformed nuclei too,
work in the basis where only the magnetic number is denoted. So, the orbitals we
use could be the Hartree-Fock orbitals in a deformed nucleus where the angular
momentum is no longer a good quantum number, but only its projection on the
symmetry axis for axially symmetric systems (Rowe 1970, Ring, Schuck 1980,
Bohr, Mottelson 1969) is. It is then always possible to expand in a spherical
basis (general case)

204
 i,m
(7.25)

where
a-l:],m
_ = (_I)i+ma-l:],-m'
-
(7.26)

and indicates for a~ the creation of a particle in the time-reversed state of p.

If one uses the particular choice of the orbital single-particle wave functions
i1y/(m)(r), then the time-reversal operator or the rotation operator, 'R. y(7r) (per-
forming a rotation of 7r around the y-axis) acting on a single-particle state Ii, m)
gives the same result. So, the above phase choice (Biedenharn-Rose (BR) phase-
convention) is quite often used in pairing theory (see also Appendix I) (Rowe
1970, Bohr, Mottelson 1969, Alder, Winther 1971, Waroquier et al. 1979, Brus-
saard 1970, Allaart 1971).
Using the above BR convention, one can obtain the same sign for all pairing
matrix elements, in the magnetic sub states

(pPlVlvii) = -G with G >0. (7.27)

So, pairing interactions are described by the Hamiltonian

H = LC/la~a/l - G L a;a~aiia/l' (7.28)

/I 1',/1>0

where C /I are the single-particle energies. If we now define the eigenstate

In) == S~IO) = Lx~a)a~atIO) , (7.29)

/1>0

then x~a) and Ea are determined via the secular equation

(7.30)
or
(2c/I - Ea )x~a) = G Lx~a) . (7.31)
1'>0

The formal solution so becomes

x(a) - N G (7.32)
/I - a2c/I-Ea '
with

(7.33)

205
 Fig.7.7. The IDA secular (or dispersion) equa-
tion for the pairing properties of two particles in a
number of non-degenerate orbitals at unperturbed
energies 2eI, 2e2, 2e3,... for the two-particle
configurations (see (7.34». The intersection be-
tween the horizontal line at 1/G with the right-
hand side of (7.34) gives the eigenvalues (roots)

The nonnalization of the state Io:} then leads to the condition

II

or,

.!.=:L
G
1
0 2cII - Eo
(7.34)
II>

Since G > 0, we obtain a low-lying 0+ state that comes down from the unper-
turbed energies 2ch 2c2, ... to a low-energy. This coherent combination of 0+
pairs is the new ground state that takes the short-range interaction optimally into
account (Fig. 7.7).
One could carry through the above discussion within a spherical single-
particle basis where the creation- and annihilation operators are characterized
by the quantum numbers (j, m).

7.4 n Particles in Non-Degenerate Shells: BeS-Theory

The situation where many pairs move in a number of non-degenerate orbitals is

most close to the actual nuclei one can observe e.g. the even-even Sn nuclei. Here,
the proton number remains at Z = 50 but the neutron number varies between
50 :::; N :::; 82, indicating that between 0 and 32 nucleons (0 to 16 pairs) are
distributed over the available five neutron orbitals 2ds/2, Ig7/ 2, Ihll/2' 3S1/2,
2d3 / 2 • Still, the first 2+ level remains remarkably stable around an excitation
energy of Ex ';;;t 1.2 MeV and needs to be explained as an important property of
the nuclear many-body problem (Fig. 7.8).
Within the most general case of n particles (or nj2 = N pairs), an exact
treatment as in Sect. 7.2 and 7.3 is no longer possible and different approx-
imation methods to obtain the nucleon pair distribution have been introduced:
(BCS) Bardeen-Cooper-Schrieffer theory (Bardeen et al. 1957), (OS) generalized
seniority (Talmi 1971), (BPA) broken-pair approximation (Allaart et al. 1988),

206
 Fig. 7.8. The systematics of the first 2i" level in the even-even Sn nuclei
(102 ~ A ~ 130)

c
en
<0
+ !::!

c
en
+ ..,
!::!

c
+ N
0_
...en
c
en
N
+ N
0-

c
en
+
0-
l'l

c
en
<0
+ :=

c
..,en
:=

c
en
+
0
~

c
en
N
+
0
:=

c
en
+ :g
0-

c
+ 0
...en
0-

c
en
+ N
o~

. .. . We shall later on discuss BPA in some detail since this approach easily gets
into contact with the spherical shell-model that was studied in detail before and
can make contact with the ffiM (Interacting Boson model) too (Sect. 7.7).
As a trial approach to the ground-state 0+ -wave function we use a product
state of n /2 pair states as discussed in Sect 7.3 and we obtain

207
 n/2
In) == (
L c.,(n)a:a; )
10} . (7.35)
.,>0
Then, one has to detennine the coefficients c.,(n) (for each n separately) such
as to study the variational problem

o{nIHln} =0 . (7.36)

The above wave function In} is no longer an independent-particle wave function

which makes the variational problem difficult to carry out in an analytic way.
A way out is to consider a more general wave function 10} which, projected
on the space of n/2 pairs, gives back the ground state In} of (7.35). This new
trial wave function,

10) == II (u., + v.,a:a; ) 10) , (7.37)

.,>0
is known as the BCS ground state, but has no constant particle number. At this
stage, the coefficients u., and v., are still to be detennined. Later, we shall show
that it turns out that v!
is the occupation probability of the levelv.
Carrying out the projection via the operator Pn , one gets

(7.38)

[Prove this relation (7.38).]

The state 10) has the peculiar property that it is the vacuum state for a new
kind of generalized fennion annihilation operator cl£' such that

cl£IO) =0 for all J.L • (7.39)

Here,
-
cl£ = ul£al£ - vl£aji+ . (7.40)

The BCS-transfonnation (canonical transfonnation) that transfonns from the par-

ticle creation and annihilation operators to the new "quasi" particle operators is
given by (Rowe 1970, Ring, Schuck 1980, Lane 1964, de-Shalit, Feshbach 1974)

C: = u.,a: - V.,aii ,
(7.41)
c., =u.,a., - v.,a; ,
and the inverse transfonnation
a: =u.,c: + V.,Cii , (7.42)
a., =U.,C., + v.,c; .
From (7.41) one observes, by inspection, that the new operators reduce to

208
 e: =a: , e" = a" (u" = 1, e" = 0) ,
e: = aji , e" = a~ (u" = 0, v" = -1) ,
(7.43)

so that u; and v;
can, although still in a loose way, be interpreted as occupation
probabilities. Far above the Fenni level of occupied states the quasi-particle
operators reduce to the regular particle operators. Far below the Fenni level
the quasi-particle creation (annihilation) operators become creation (annihilation)
operators of hole excitations. Near the Fenni level, some composite structure
results.
From the constraint of anticommutation of the new operators {e;, e,,} = 8,..",
one derives
(7.44)

We like to stress again that the new vacuum state IO} does not contain a fixed
particle number. In Fig. 7.9a and b, we respectively indicate the state IO}, the one
quasi-particle state e; IO}, as well as the population of the single-particle levels
c:" depending on the pairing force strength G.
The quantities v" (or u,,) are now detennined by perfonning a constrained
variational calculation. So, we minimize the "Hamiltonian"

1{ = H - An, (7.45)

1{ = l)C:" - A) (a:a" + a~aji) - G L a;a~aiia" . (7.46)

,,>0 ,..,,,>0
Here, A is a Lagrange multiplier, and is chosen such that the average particle
number agrees with the actual number of valence nucleons or

a) b)

G» p1 G « p1

Fig.7.9. (a) Representation, on a graph with the occupation probability distribution for the different
orbitals, of the quasi-particle vacuum state 10) (lfft-hand part) and of a one-quasi particle (lqp)
ct
excitation in the orbital 1/ denoted by 10). In such a case, the orbital 1/ is occupied with a single
particle since one has annihilated in 10) the state 1/ for that part which was occupied and created a
particle in 1/ for that part which was unoccupied. (b) Probability for the occupation of the orbitals
v(ev) for different ratios of the pairing strength G to the average distance (,,-1) between single-
particle levels. For G > ,,-I, an almost constant occupation for all levels 1/ is obtained. For
G 9!' ,,-I, a smeared out distribution and for G « ,,-I, a sharp distribution of full occupation up
to a given level (Fenni-level) is obtained

209
 =n,
(OlnlO) (7.47)
So, we get>. = >'(n) via the condition

! (OI'HIO) = 0 , (7.48)
or
(7.49)

indicating that>. is the chemical potential of the A-nucleon system.

In order to determine the quantities u v , vv, we start by rewriting the Hamil-
tonian 'H, using Wick's theorem, with respect to the new vacuum state 10) and
transforming the creation and annihilation operators in (7.46), making use of the
expressions (7.42).
In short, the expression turns out to have the structure
'H = Uo + Hll + Hw + H02 + Hres , (7.50)
where Uo is the energy, corresponding to the reference state 10}. Furthermore,
the other terms have as general structure
Hu ex c+c,
H20 ex c+c+ ,
Hres ex c+ c+ c+ c+ + c+ c+ c+ c + c+ c+ cc + hc . (7.51)
In more detail, the ground state energy reads

Uo = 2:)ev - >.) {~av +~ji}

v>O
- G "LJ {'+' '+' v + ap.ap.ajia
ap.ajiap.a 7 + r----J v } (7.52)
p.,v>o
If we now evaluate the different contractions, where also the terms ?a+ give
non-zero contributions, we obtain for the energy Uo

Uo = 2: [2(e v - >')v~ - Gv~] - G[Lu vvv]2 (7.53)

v>o v>O
Since, in all other terms of the Hamiltonian, normal order products of operators
appear, the variational problem of calculating (OI'HIO} reduces to evaluating the
above expression Uo and one obtains

o - - 0
-(01'H10) = -Uo =0. (7.54)
ovv ovv
This derivative leads to the condition

2(e~ - >,)uvvv = Ll(u~ - v;) , with (7.55)

210
 (7.56)

The new single-particle energy e~ (= ell - Gv!) contains the self-energy correc-
tion of a particle in a given orbital v interacting, via the constant pairing force,
with an extra pair of nucleons. It describes the changing single-particle energy
ell as a function of n when starting from the constant Hartree-Fock energy ell as
determined for a doubly-closed shell nucleus. In many BCS calculations where
a set of single particle energies is deduced from the data, the e~ are used from
the beginning.
From (7.55) one obtains the solutions

(7.57)

One now needs two equations to determine the chemical potential . x and the
quantity Ll. From (7.55) which is also called the "gap" equation, one can derive
an equivalent form

L [(e~ -..x) 2 + Ll ]-1/2 = G2 .

2 (7.58)
,,>0
The particle number condition then leads to

L [1 -
e'" -..x
2 1/2 1n.
= (7.59)
,,>0 [(e~ -..x) +Ll2]
In practice, one has now to solve (7.58) and (7.59) simultaneously for a given
set of energies ell, given n and pairing strength G for the unknowns ..x, Ll. This
highly non-linear set of two coupled equations can be solved using the iterative
Newton-Raphson method in a rapid way. Once Ll, . x known, all other quantities
v!, u!, e~ can be obtained. In Chap.9 we enclose a code for solving these
equations using a constant pairing force and a number of single-particle states.
The one-body part of the Hamiltonian Hu. together with the part Hw + H02
can be obtained by considering the single contractions in (7.46)

L(e" - ..x) {N(a!a" ) +N(a;aii)}

,,>0

-G L {¥~N(aiia")+N(a;a~)¥"
/l,,,>0
211
 (7.60)

(7.61)

the part corresponding to Hll reads

Hll = LE" (ctc" + C~Cji) • (7.62)

,,>0
Herein, we have used the definition

E" = [(e~ -,\) 2 + L12]1/2 . (7.63)

If we now could show that H02 + H']J) is not very important, it could be shown
that going to the new representation of quasi-particle operators, a set of new
elementary excitation modes are defined that absorb a large part of the residual
interaction (pairing part) into the new basis.
Evaluating H']J) + H02 from the remaining terms in (7.60), one obtains

H']J) + H02 = L [2(e~ -,\ )u"v" - L1(u; - v!)] (ctc~ + CjiC,,). (7.64)
,,>0
This term identically becomes zero since the variational problem implies (7.55)
and so the factor in square brackets in (7.64) disappears. In retrospect, one
can re-interpret the quasi-particle representation as that representation in which
the two-particle scattering processes across the Fermi level become absorbed in
defining a new basis and reference state. This condition [vanishing term in (7.64)]
implies the minimum in energy of Uo.
Finally, the total Hamiltonian becomes (relative to the energy Uo)

1-£ = L E" ( ctc" + C~Cji

,,>0
) + H40 + H31 + Hn + hc , (7.65)

with the latter terms (H40, H31, Hn, •.• ) to be considered as the residual inter-
action Hres relative to the new zero-order basis.
With (7.65) we are back to a well-known problem of the nuclear shell-model
with a number of valence nucleons in open shells. One can again consider the
ground state IO}, one-, two-, three-, ... quasi-particle excitations relative to this
ground state. In even-even nuclei one shall consider

o+ 2 qp excitations,

and in odd-mass nuclei

212
 NUCLEON DISTRIBUTION Un' 26

2d3 /2
351/2
1h 11/2
11M Mil ""
"""" ""
19 712 ---- MM 1111 MMIilI Mil 1111

2d5l2 -....;"~"--
"" "" "" II" Mil "" M" II" II" M M 101 1111

n=2 6 10 20 26

2d3/2 - - - -

1h'V2
351/2 - - - - - - l
I

19~2dl
2d5/2
1 I 1
I
_ _ _ ---1I _ _ _ -lI ___ -.J

-
OCCUPATION PROBABILITY
Fig. 7.10. lllustration of the distribution of a number of nucleons n(2 :5 n :5 26) over the five
orbitals 2dS / 2 • 197/2. Ih ll / 2 • 381/2 and 2d3 / 2 • In the upper part, we depict one (for each n) of the
many possible ways the n nucleons can be distributed over the five available orbitals. This number
of possible distributions grows very fast with increasing n. In the lower pan we give. for increasing
n. the optimal pair distribution according to the BeS prescription (7.54) of minimal ground-state
energy (011£10)

1 + 3qp excitations (Baranger 1960, 1961, Kuo et al1966a, 1966b) .

The above configurations, when pictured in a particle representation of all nu-

cleons outside the closed shell, correspond to highly complicated nucleon dis-
tributions. So, the enormous advantage here is that with numerical calculations
that are not more difficult than performing a two-particle shell-model calcula-
tion, one studies a large chain of even-even nuclei. In Fig.7.10, we illustrate
this for the even-even Sn nuclei where we consider a number of cases: in the
upper part we consider 2,6, 10,20 and 26 neutrons to be distributed over the
available five orbitals. Although we only draw the most trivial distribution we
should in principle draw, as n increases, a large increasing set of distributions
for describing the ground state wave function. In the lower part, the BCS pair
distribution, adapted to each case by solving the gap equations, gives already
within the 0 qp space only a very good description of the actual pair distribution.
In considering also the excited states, the 2 qp excitations have to be taken into
account, but even here this leads maximally (for pr = 2+) to a 9 dimensional
matrix to be diagonalized.

213
 Fig.7.11. The difference between single-
particle excitations (g"" - g"/) and one-

~
£1
I
1
E quasi particle excitations E"" - E"/, illus-
trated in a schematic way. The Fenni level
>. is also given. The single-particle energy
A difference is measured by the distance be-
tween two given single-panicle configura-
SINGLE -PARTICLE QUASI-PARTICLE tions. The one quasi-panicle energy differ-
{J, {J, ence follows from a geometrical construc-

, _:_~_l~_::~E:',
v" "~ =
tion since E" V(g" - >')2 + .12 (see the
insert in Fig. 7.11)

In odd nuclei, the one-quasi particle excitations are the corresponding low-
lying excited states that correspond to single-particle (or single-hole) excitations
at doubly-closed shell nuclei. Here, the effects of the pair distribution on the
odd-particle comes about via the change of the single-particle energy ev into the
one quasi-particle energy Ev. This becomes minimal L1 (for ev = ).).
In Fig. 7.11 we compare the distinctly different features between excitations in
odd-mass nuclei: starting from the Fermi level, single-particle energy differences
will always be larger than the corresponding quasi-particle energy differences
(because of the dominant L1 factor). We also indicate the geometric way to
interpret the one-quasi particle energy.
In even-even nuclei, the lowest excited states result at a minimal value of
Ex ~ 2.1 (which in most cases is of the order of 2 MeV) (see Fig. 7.12).

(MeV)

2.0
-----} 2qp excitations

------T-
I
I
I
1.0 2A I
I
I Fig. 7.12. Representation of the two quasi-particle (2qp) spec-
I trum in an even-even nucleus. The lowest possible 2qp exci-
I tations cannot show up below 2.1. For typical values of .1 9t
I O.S-1.0MeV, this indicates that most 2qp excitations in spher-
t ical, single closed-shell nuclei show up around Ex 9t 2 MeV

214
 Fig. 7.13. lliustration of the relation be-

t tween the pairing gap (..::1) or better, the

lowest one-quasi particle energy, corre-
sponding with the ground-state configu-
ration El qp (ground-state) and the bind-
ing energy of a number of nuclei. In
a linear aPlB'oximatioo, one obtains the
expression El qp (ground-state) ::: ..::1 =
1/2(BE(A + 2)+ BE(A) - 2BE(A + 1»
with BE(A) the binding energy in the nu-
cleus with A nucleons

even-even

I I I
A-2 A-l A A+l A.2
- - - MASS NUMBER A - -

The energy in an even-even nucleus and the adjacent odd-A nuclei is, in
lowest order described by Uo, Uo+ E", Uo+ E", and so by studying the difference
in binding energy in the ground states in adjacent nuclei, an empirical estimate
of Ll (upper limit) can be obtained (Bohr, Mottelson 1969). This relation is
expressed in Fig. 7.13.
Later on, in Sect. 7.5, we shall discuss some more realistic cases where con-
ditions like constant pairing matrix elements are relaxed and thus, a better agree-
ment with experiment can be realized.
Before, we discuss shortly the effect of Hres in the quasi-particle scheme.
The main effect is that the number of quasi-particles gets mixed because of
the structure of the different terms. H40 has 4 quasi-particle creation operators,
H31 three quasi-particle creation and one annihilation operator, etc., depicted
diagrammatically in Fig. 7.14. The diagrams H40 (H04), H31 (H13) relate to RPA-
type of correlations whereas H22 is the only quasi-particle number conserving
interaction. It has the same form as in the ordinary particle representation.
We give some attention to the existence of solutions to the BeS gap equations

Ll=GLu"v" ,
" (7.66)
n=2Lv; .
,,>0
These equations always contain the trivial solution Ll = 0, U" = 1,0 and
v" = 0, 1, corresponding with a sharp Fermi distribution. Sometimes, a non-
trivial solution Ll:f 0 exists. If G is too small, or Llc too large compared to
G, no BeS solution outside the trivial one exists. If we take a as the highest,
occupied and f3 as the lowest, unoccupied orbital, then, to fulfil the number
equation of the gap equations, A must occur in between COl and cp. If for such
value of A, G is so small that

215
Fig. 7.14. Diagrammatic represention of the different contributions to the residual interaction in the
quasi-particle representation (see (7.65)). The notation Hnm is such that n indicates (reading from
below the interaction vertex to above the vertex) the number of outgoing lines and m the number
of incoming lines or n indicates the number of qp creation operators and m the number of qp
annihilation operators. The tenns H40(H04) induce RPA-like correlations, H31(H13) the creation
(annihilation) of 2qp states and H22 the scattering of 2qp states

; L \e/l - .\1- 1 < 1, (7.67)

/1>0

then clearly no solution for

; L IC/I - .\1-1 [1 + ,12 2]-1/2 = 1, (7.68)

/1>0 (C/I -.\)
exists. This is precisely the case for closed-shell nuclei where pairing is so
weak that no scattering into the empty orbitals in the next major oscillator shell
is possible. For nuclei with a partly filled, degenerate j-shell, a BCS solution
(,1 f 0) always exists no matter how small G becomes.
As examples we quote (Rowe 1970):
i) 2s-1d shell nuclei: deformed nuclei exist but a non-trivial BCS solution does
not.
ii) single-closed shell Sn, N = 82 nuclei, ... : spherical solutions, BCS solutions
exist.
iii) rare-earth nuclei: deformed solutions with a non-trivial BCS solution on top
exist.
We finally point out that minimizing the energy Uo to find a BCS solution
does not immediately imply that we have obtained the lowest energy solution

216
 Fig. 7.1S. Schematic picture of the optimalization
H.F. of the total energy with respect to the Hartree-Fock
(HF) and BCS type of correlations in the ground-
state wave function. The two types are depicted
on two orthogonal axes (HF and BCS). Contour
lines of constant energy are shown. It is shown, al-
beit schematically. that the energy minima in the
Hartree-Fock ground state and BCS ground state
need not coincide and do not always fonn the low-
est energy minimum

to H in an absolute sense. It could be that, minimizing with respect to both

the structure of the Hartree-Fock orbitals and the pair distribution at the same
time (HFB), one could reach a deeper point on the energy surface. Fully self-
consistent HFB calculations have rarely been carried out. In the best of cases,
an iterative set of HF + BeS minimizations are carried out, a method that can in
most cases approach the absolute minimum in a good way (Ring, Schuck 1980)
(see Fig. 7.15).

7.S Applications of BeS

We shall now discuss a number of typical results that are obtained using the Bes
method, using this time the more realistic set of gap equations, in which (7.56)
has to be replaced by (we use spherical nuclei) (Kuo, Brown 1966, Kuo et al.
1966a, Heyde, Waroquier 1971, Waroquier, Heyde 1970,71)

(7.69)
c
We discuss (i) odd-even mass differences and one-quasi-particle energies, (ii)
energy spectra in even-even and odd-mass nuclei, (iii) electromagnetic transition
rates, (iv) spectroscopic factors, respectively.

7.5.1 Odd-Even Mass Differences, Elqp

Because of the pairing correlations, the last odd nucleon is less bound in odd-
mass nuclei compared to the even-even nuclei. A mass difference, in a linear
approximation, can be defined via the relation (Bohr, Mottelson 1969)
El qp = HBE(n + 1) + BE(n - 1) - 2BE(n») , (7.70)
where n is the number of valence nucleons in the odd-mass nucleus and BE is
the binding energy of the nucleus. If the ground state in the odd-mass nucleus
would always be a pure one-quasi particle (1 qp) state, then one has
217
 4,0 J' ,
b
b
W
(\J

1,0

N=28 Ni(Z=28) N=50 5n=(Z=50) N=82

~;
Pb(Z=82)

Fig. 7.16. Even-odd mass differences (see Fig.7.13). The dots are the experimental differences
BE(A + I)+BE(A - 1) - 2BE(A). The theoretical curves are 2Elqp(j), twice the energy of the
lowest lqp excitation for the odd-A IUlclei. Curves (a) and (b) are for G = 19/A and 23/A. Curve
(c), for the Pb nuclei only, corresponds with G = 30/ A (taken from (Kisslinger, Sorensen 1960»

(7.71)
if

We illustrate 2El qp, as obtained from (7.70) throughout the nuclear mass region
(Fig.7.16). The data are compared with constant pairing strength calculations
using G = 23/A (line b), G = 19/A (line a) and G = 30/A (line c) for the
Pb region only. This figure has been taken from (Kisslinger, Sorensen 1960)
and illustrates that pairing is able to account in a very systematic way for the
behaviour in odd-even mass differences.
We also illustrate, in some more detail, for the N = 82 nuclei, how the 1 qp
energies change with increasing Z· (53 ~ Z ~ 63) for given proton single-particle
energies (Fig.7.17) (see insert). The data and theoretical values are compared.
For more details on the force used in the calculation, we refer to (Heyde, Waro-
quier 1971, Waroquier, Heyde 1970,1971). The effect of filling particles and the
subsequent variation of A is very nicely illustrated in the interchange of 2ds /2
and 197/2 between Z = 57 and Z = 59 with, on top, a steady decrease of the
Ih ll / 2• 381/2 and 2d3/ 2 1 qp energies. A gap between the (2ds/ 2, Ig7/ 2) and
(lhll/2' 381/2, 2d3/ 2) orbitals is evident.

218
 EXPERIMENT
THEORY

>-
(!)
II:
W
Z
w 1.5
z
o
I-
~
U
x
w
1.0

0.5

0.0
135 1 137CS 139 L 11.1 Pr 143Pm 11.5 E
53 55 57 a 59 61 63 U

Fig. 7.17. Realistic calculation for the lowest 5/Z+. 7/Z+. 11/2-.3/2+ and 1/2+ excitations in the
N = 82 nuclei. that correspond to rather pure one-quasi particle configurations. Both the theoretical
and experimental energies. as a function of Z. are given. In the insert, we give the lowest quasi-
particle energy E,,{A). as determined from the odd-even mass differences. We compare the data with
the theoretical values

7.5.2 Energy Spectra

In even-even nuclei, as discussed before, one does not expect excited states to
lower below the value of 2..1 in the unperturbed energy spectra. Since ..1 ~
1 MeV, this means that above Ex > 2MeV, the level density should rapidly
increase, as is indeed the case for the N = 82 nuclei: 136Xe, 138Ba, 140ee (see
Fig. 7.18). There are, however, some states that descend into the gap, in particular
the J1r =2+ collective state, which occurs around Ex ~ 1.2-1.4MeV. Also, one
4+ and 6+ level decreases below the value 2..1 but it is indeed close to Ex ~ 2-
2.5 MeV that many levels are observed and calculated in the 0 + 2 qp calculation
using a Gaussian residual interaction within the IDA approximation.
219
 LEVEL- SCHEMES OF THE N=82 NUCLEI

54 Xe 82 56 8 °82 58 Ce 82

L
>Ql
:::E
3----
r---

4+ _ _ _
2"---
0"---

>- ,+
I.!> zf---
a:
w
z
w
s+J===
4' _ _ _
6+
Z 2.0
2

---
,.-
,.-
~ 0+ , ,.-
;5 6+
4+ 0" _ __ ---2"
U ---2"
--
X 2"---
W 2+

I
1.0

2QP EXP. 20P EXP. 2QP EXP.

00 0"_ _ _ _ _ _ _ _ _ _ 0" 0" _ _ _ _ _ _ _ _ _ _ 0" 0'_ _ _ _ _ _ _ _ _ 0"

Fig. 7.18. Level schemes in the single closed-shell N =

82 nuclei 136Xe, 138Ba and 140ee. We
compare the experimental spectra with the results of a 2qp calculation. In the calculations. a Gaussian
radial shape was used with spin exchange. expressed by the spin projection operators PSt Fr as
V(r) = - Voexp[-.Br 2 ](Ps +tPT)' In each case t = +0.2 was used and Vo obtained via the inverse
gap equation (IGE) method (see also (Waroquier. Heyde 1971»

In the odd-mass nuclei, on the other hand, no such gap shows up but here,
as the lowest levels, we encounter the 1 qp excitations E a , Eb, .... In Fig. 7.19,
we illustrate the case of 137 Cs, where we compare the data (Holm, Borg 1967,
Wildenthal 1969, Wildenthal et al. 1971), the results of a 1 + 3qp calculation
(Heyde, Waroquier 1971) and the exact shell-model calculation of WIldenthal,
using a SDI-force (Wildenthal 1969, Wildenthal et al. 1971). The comparison
between both calculations is striking but the 1 + 3 qp calculation needs much less
numerically involved methods.

220
 EXPERIMENT
r CS'2
.J!6. SHEll -MODEL
(9 HWILDENTHAL et alII) (GAUSS Wllh t =0.2) <SOil
Ex IMWAROQUER, KHEYDE I (BJ1WLDENTHAL)
IMeYI (~OLMIlI)
30
3Ii (Vi,¥.z·~Jl):Y~.I\Ii) (~~,IIJi) D
2852 __ ...... _5 2.948 1~· __
(~l
:y~
'12' ==\'-

Vi iii
~~:
_-tj
!IN HI 1l...2lL
M di-~:
, vze=:
sf/.
¥2'
-
$ - - - IIJi
7. -
2'»_
'12; -~-9t2
--~J!:l-:I ;~: ~.'
20 ~ 20
lrz- ,.
III
2, -,*-
'I'.!-=::So/i
~~1780
-=-.: jim %' 5 2 - - - - 7 ; -

I.
1m !

'ii '41O __ ~_ 1 :"-'//f.J.llll...___..~' :tz~-- -- ~

\1:1"--
--7/f..JJIL
---~
3h!.-- ---112
to (~) 0., --~f'i~ 0926
IW !fl' 10
1000 1j2'
Wl6J
(!til
--'--
!i'i ....
gZ!!!
~2~--

gt.5!j
Sfi 0162
%+---

00 y; 1fz' YI' 0.0

Fig. 7.19. Comparison of the TDA 1 + 3qp calculation. using a Gaussian interaction (see caption
to Fig.7.18) with spin exchange and the shell-model calculation of Wildenthal (Wildenthal 1969)
USing a SOl force with the data in 137es. The data are from (Wlldenthal et aI. 1971) and (Holm,
Borg 1967). Levels connected by dashed lines are well established states from one-nucleon transfer
reactions (taken from (Heyde, Waroquier 1971»

7.5,3 Electromagnetic Transitions

The electromagnetic transition operators, that were discussed in detail in Chap. 4,
and were one-body operators (Kuo, Brown 1966, Kuo et aI. 1966a, Brussaard
1970, Allaart 1971, Waroquier et aI. 1975)
(7.72)
a,p
can now be rewritten, using the quasi-particle transformation to the quasi-particle
operators c~, cp. Using the Condon-Shortley phase convention for the single-
particle orbital wave functions IT (also see Appendix I), one obtains for (7.72),
the expression

L().)-l (aIlO(,x) lib)! [(UaVb + (-I),xubva)

a,b
X (A+(ab, >.p.) + (-1)1' A(ab, >. - p.)]
+ L().)-l {a IIO(,x) lib} (UaUb - (-I),xvavb ) AO(ab, >.p.)
a,b
+ o,xo L{aIO~)L8}v~oap . (7.73)
ap
221
Here,

A+(ab,AIl)= L (jarna,jbrnbIAIl)Cla,macjh,mb' (7.74)

ma,mb
and
A(ab, All) == (A+(ab, All) r.
We also define

AO(ab, All) = L (-1)ih- mh(jarna,jb - rnbIAIl)C;',ma cib,mh .

The summation over greek variables means a summation over all quantum num-
bers i.e. 0: == n a, la,ja, rna, whereas the summation over roman variables ex-
cludes the magnetic quantum number rna i.e. a == n a, la,ja.
Within the CS (Condon-Shonley) phase convention, one can find solutions
to the BCS equations with all Va > 0, and the phase Ua = (_1)la(1 - v;)1/2.
Using the BR convention, one can find solutions with both U a, and Va taken as
positive quantities.
So, transitions 0 qp <=> 2 qp; 1 qp <=> 3 qp; 1 qp <=> 1 qp; 2 qp <=> 2 qp can oc-
cur. The corresponding reduction factors for the single-particle matrix elements,
that show up in (7.73), are given in Table 7.1. One observes, that in particular

Table7.1. The single-particle pairing reduction factors for various types of transitions and for the
electric and magnetic multipole operators, according to (7.73). The table is such that it corresponds
to the CS phase convention for defining the single-particle wave functions (i.e. one chooses UjVj =
(_1)1; IUjvjl). The relation with the BR phase convention is discussed in the text

Operator Type of transition Pairing factor

E,\ lqp - lqp (UjUj - VjVj)
M,\ lqp - lqp (UjUj + Vj Vj)
Q lqp - lqp (U~ - 0)
JJ lqp - lqp 1
E,\ Oqp - 2qp (UjVj +UjVj)
M,\ Oqp - 2qp (UjVj - UjVj)

EA(M A) transitions without (with) a change in qp number can be much retarded

(see the sign for destructive interference). Also the quadrupole moments Q can
become heavily reduced (at the middle of a single j-shell) changing sign at the
mid-shell configuration.
We now give a number of illustrations
i) The B(E2; 6j ~ 4j) values in the N = 82 nuclei (Fig. 7.20). Since we
have here a 2 qp ~ 2 qp transition, reduction factors u~ - v~ show up and
are responsible for the very large modulation in B(E2) values since we go
through the 2ds/2 and 197/2 orbitals, respectively. A more detailed discussion
is given in (Waroquier, Heyde 1974).

222
 ~ experiment
o Elliott (th. va2 )
• E 11'011 (exp. va2 )
100

0
10
~
.
N
.0

\
N·

...
\
+

+
f
.- ~
<D

\
UJ
ID

1.0 \
\
\

0.1 L - _ - L_ _-'-.'-1.J..J..L_ _-'-_...I..-'--_....J...._ _-'----'

l~;Te l~~Xe l~:Ba l~~Ce l~~Nd ~4;Sm l~~Gd

Fig.7.20. lllusttation of reduced E2 transition probabilities B(E2; 6+ - 4+) between mainly 2qp
configurations in the single-dosed shell N =82 nuclei. The data ( ~) are compared with calculations
using the Elliott interaction (Elliott 1968) and both theoretical (0) and experimental (.) values for
the occupation probabilities for the proton 58-82 single-particle orbitals

ii) We discuss, for odd-AN = 82

nuclei, the different transitions that can
deexcite from the 11/21 1 qp state (mainly) into the low-lying 7/2i and
5/2i levels that are also mainly 1 qp states (Table 7.2). This means that
we have to study M2 and E3 transitions. We discuss the results for 137 Cs-
145Eu and indicate the very stable results. Here, it is also clear that the
1 qp {:} 3 qp components give a rather important contribution which is, in all
cases correcting the 1 qp {:} 1 qp matrix elements towards the experimental
values (Waroquier, Van den Berghe 1972).

223
 Table7.2. The different transition rates from theJr
= 11/21 state to the 7/2T and 5/2T states. We also indicate the half-life estimate for the 11/21
level in the odd-proton N =82 nuclei (taken from (Waroquier. Van den Berghe 1972»
I\)

~ 137es 139La 141Pr 143Pm 14SEu

E-y(exp) in MeV 1.870 1.420 0.965 0.688 0.383

B(M2) Single part. estimate 62.56 62.86 63.16 63.45 63.76

in lQP-IQP 33.70 25.28 18.64 18.63 13.22
(eh/2M c)21m2 1 - 3QP - 1 - 3QP 14.25 11.64 10.27 11.41 9.583
Experiment 10.3 ± 0.7
11/2-
!
7/zt' Single part. estimate 3.47 x 102 5.95 x 102 1.28 x 103 2.49 X 103 7.79 x 1()3
P-y(M2;Ih.l1l2-197l2)
lQP-IQP 5.48 x 102 1.55 x 103 7.05 x 1()3 2.55 x 104 2.03 x lOS
P-y(E3;lh 11 / 2 _197/2)
1 - 3QP - 1 - 3QP 3.42 x 102 1.39 x 103 1.41 x 104 7.34 x 104 1.43 x 1()6
Experiment very large

11/2- B(E3) Single part. estimate 1.526 x 104 1.548 x 104 1.570 x 104 1.593 x 104 1.615 x 104
! in lQP-IQP 1.356 x 104 1.168 x 104 8.604 x 1()3 4.107 x 103 9.557 x 102
5/zt' e2/m 6 1 - 3QP - 1 - 3QP 1.266 x 104 1.065 x 104 6.450 x 1()3 3.542 x 103 8.292 x 102
Experiment 1.098 x 104

E-y(exp) in MeV 1.414 1.254 1.110 0.959 0.716

Branching-ratio lQP-IQP 119 61 21 22 30

1 - 3QP - 1 - 3QP 54 31 15 15 25
P-y(1h ll l 2- 197l2)
Experiment 9 4.7 ±0.5
P-y(lh ll / 2 - 2d r./2)

lQP-lQP 0.0658 0.345 3.13 17.0 431

T1/2 of the Ihu/2 1 - 3QP - 1 - 3QP 0.154 0.737 5.58 27.3 591
level in ns Experiment 0.5 ~ T;j~2 26.0± 1.2
 • =TDA
A = EXPERIMENT
• = SHElJ.:-MJOEL ESTIMATE (\1=2)

Fig. 7.21. The gyromagnetic g-factor for the J{ = 4t

state in N = 82 nuclei, calculated using a
proton 2qp IDA calculation. The dashed-dot line results from a pure, seniority v = 2 shell-model
wave-function for the J't = 4+ level. In the case of a completely filled (or completely empty) single
j-shell, only J7r ::: 0+ states are formed in a pure shell-model picture. In this case g(O+) is given
(taken from (Waroquier, Heyde 1971»

iii) Dipole moments g(4+) are presented for the N = 82 nuclei and compared
to the pure shell model data, in the same way as we carried this out in (iii)
(Fig. 7.21). The point in Ce is remarkable as is the agreement with the data.
This clearly points out the importance of pairing correlations in obtaining
rather good wave functions for these single-closed shell nuclei (Waroquier,
Heyde 1971).
iv) Quadrupole moments in even-even N = 82 nuclei are presented in Fig. 7.22.
We compare, for the Q(2j) value with a simple-shell model estimate
assuming a consecutive filling of the 2ds/2 and 197 / 2 orbitals. Already from
Ba on, important deviations start to show up (Waroquier, Heyde 1971).

225
 Jiig.7.22. 1be electric quadrupole
• = TOA
moment for the J[ = 2! level
" = EXPERIMENT in the N = 82 nuclei (solid
• = SHEU:-KIOEL ESTIMATEIV=2)
line). The dashed line conneclS
020 Q2+ for a seniority v = 2 shell-
A model wave-function for the 2'"
/\ state. For a fully occupied (or
/ \ empty) shell, the same remark as
in Fig. 7.21 holds. The separate
0.1 contributions for (1m /2)~+ and
(2d5/2)~+ componenlS to Q(2!) are
also drawn (light, solid line) (taken
from (Waroquier, Heyde 1971))

ao

-020 I I I
Sn Te Xe
50 82 52 82 54 82

7.S.4 Spectroscopic Factors

Spectroscopic factors in one-nucleon transfer reactions are very important to

provide infonnation on the position and possible fragmentation of nuclear single-
particle configurations (Brussaard, Glaudemans 1977, Kuo, Brown 1966, Kuo et
al. 1966a).
IT we describe the transfer of one-nucleon between two nuclei, the lighter
nucleus being described by the wave function 'ljJ(Jo, Mo) and the heavier nucleus
by 'ljJ(J, M), one defines the spectroscopic factor as

S(J;jJo) = 1('ljJ(J)!laj!l'ljJ(Jo)}12 2/+ 1' (7.75)

or

S(J;jJo) = I('ljJ(JM)I!P(jJo; JM)}12 , (7.76)

if !P(j Jo; J M) is the state obtained by coupling the odd nucleon j with the core
Jo to angular momentum J. If now the final state 'ljJ(J) would be precisely the
odd particle j coupled to the core Jo, without fragmentation, the overlap integral
would be unity and we should obtain a spectroscopic factor of 1.

226
Table7.3. Single-nucleon transfer pairing reduction factors

Target Fmal nucleus Pairing factor

Z even, A even Z,A+l u~

Z even, A even Z,A-l v~
Z even, A odd Z,A+l J
Z even, A odd Z,A-l u~1

Suppose now that the lighter nucleus is an even-even nucleus, so t{J(Jo, Mo) ==
10) and that in the heavier nucleus we have a 1 qp configuration t{J(J, M) _
cj,m 10), then the spectroscopic factor becomes
(7.77)

We give the other combinations in Table 7.3, where all possible transitions are
encountered
We discuss now
i) occupation probabilities deduced from stripping reactions for nuclei through-
out the nuclear mass region (Cr-Nd). We compare the data with the theo-
retical value (Table 7.4) of

Table 7.4. Comparison between the experimental and theoretical occupation factors from one-nucleon
stripping to the 0+ ground-state of spherical nuclei throughout the nuclear mass region. The theoretical
value is Sj =(2; + l)vj . lei 012 where the latter coefficient gives the single-particle component in
the odd-A nuclear wave function. describing the ground-state with angular momentum; (taken from
(Sorensen, Lin 1966»

Target J I Sexp SIbeory

S3er 3/2 1 0.91 0.97

S7pe 1/2 1 0.072 ± 0,01 0.()67
61Ni 3/2 1 2.0±0.3 1.7
67Zn 5/2 3 2.2 ± 0.3 2.9
77Se 1/2 1 0.68 ± 0,01 0.83
91Zr 5/2 2 1.44±0.21 1.6
9sMo 5/2 2 2.48 ±0.37 2.7
99,101Ru 5/2 2 2.74±0.4 3.3
10sPd 5/2 2 1.74 ± 0.25 2.8
l1SSn 1/2 0 1.08 ±0.16 1.03
117Sn 1/2 0 1.4±0.2 1.2
119Sn 1/2 0 1.3 ±0.2 1.4
12STe 1/2 0 1.2 ± 0.3 0.99
13S,137Ba 3/2 2 2.4 ± 0.3 2.6
143, 14SNd 7/2 3 2.4± 1.2 1.4

227
 ,
I~ SjO'4)
1-
+ n +
'--", 1=5/2
11
1=7/2 -
........
o.5 " ' .....
"~ ''-...
I I I --~"-~ I I I I I
,_ .............. -.."
-
',- _-- /'
.~- ~
...,? ::.

~
........
---~'
o.S
+
d=11/2
11 - 11
1=3/2
I I I I I I I J I L

'Cr- ,1'---. .....

',- ./-:l
I
_____ EXPERIMENT

' ,_ _ --Il
_ _ TAMM-DANCOf"f"
o.5
+
d=1/2
1{

0 I I I I I I I I I I
137 139 141 143 145 137 139 141 143 145
----A ----A
Fig.7.23. Spectroscopic factors for eHe,d) stripping to the strongest, excited J1r = 1/2+, 3/2*",
5/2+,7/2+ and 11/2- states in nuclei with a single closed-shell at N = 82 (137 $ A $ 145).
The data (dashed line) and IDA 2qp calculation (full lines) are compared (see (7.77)) (taken from
(Waroquier, Heyde 1970))

Sj =(2j + I)Vj2(C1,o.)2 , (7.78)

where Cl,o indicates the single-particle component in the ground state j,

as obtained from quasi-particle core coupling calculations (Sorensen, Lin
1966). This can give, in some cases, a large extra reduction over the 1 qp
pure result of (2j + l)vl.
ii) We compare detailed values for U] for the lowest 5/2+, 7/2+, 11/2-,3/2+
and 1/2+ levels, as obtained from eHe,d) reactions (Fig. 7.23). Comparison
of the IDA 1 + 3 qp calculation with the data is presented and proves to be
very good (Waroquier, Heyde 1970).
iii) The occupation probabilities for the 2ds/2, 197/2 orbitals when filling these
orbitals with 4 to 13 particles (136Xe to 14SEu). We compare the extreme
shell-model estimates (dashed line) with different calculations and the data
(Fig. 7.24) (Heyde, Waroquier 1971).

228
 SHELl:MOIJEL ESTIMATE
V2 A EXPERIMENTAL DATA
1 • THEORETICAL (GAUSSIAN FORCE)
1.0
I 19f'2
I
i ./

.-------
I
I ./
I
I -.---- ij/
I I:
I
+ /,

t
0.75 I

t
;, 2dSiz
I • ;;
/
I
I
I
I 'I
I • /
I
0.5 / /
/ /
/

./;/
•

! /
,
I
U25 /
/
/
: " ,/
/

!
-t----·
/ /
/

"
Fig. 7.24. The experimentally determined occupation probabilities for the 1!/7/2 and 2dS / 2 proton
orbitals in the single-closed shell N = 82 nuclei (Wildenthal et al. 1971). A comparison with the
theoretical values, obtained by solving the BeS equations, is carried out. Extreme Qinear) filling of
the 1!/7/2 and 2ds/2 orbitals, in sequence, in a naive shell model is also presented

7.6 Broken-Pair Model

7.6.1 Low-Seniority Approximation to the Shell Model

Before considering the broken-pair model as such, we first consider for a system
with n valence particles that are distributed over a number i of single-particle
orbitals a slightly different approximation to the full shell model calculation. We
call this the low-seniority shell-model (LSSM) (Racah 1942a, 1943, Racah, Talmi
1952, Talmi 1971) and recall some of the important ingredients of the model as
compared to the general BCS method that was discussed in Sect. 7.4. We shall
discuss the case with only one type of nucleons: extensions will be discussed
later in this section.

229
 We thereby start again from the Hamiltonian describing the interaction in an
identical nucleon system

H =I: eaa:aor +! I: {a,8lVh8}nasa:a~a6a'Y. (7.79)

or or,fJ,'Y,6
Considering n valence (identical) nucleons outside a doubly-closed core, the
shell-model problem of constructing the n-particle basis states with fixed Jft
quantum number, where nucleons are distributed over i orbitals, very soon (with
n and i increasing) leads to prohibitively large energy matrices to be diagonalized.
The BCS method, where a particular trial wave function was assumed (Sect. 7.4)
is one approach but has the drawback of not having a correct particle number.
The LSSM is a method that overcomes this point but still truncates the "model
space" to tractable dimensions.
In the description of semi-magic nuclei, where one kind of active nucle-
ons in the valence shells is considered, the pair scattering matrix elements
{(ja)2; O+IVI(jci; O+) are by far the largest due to the short-range nature of the
interaction (see Fig. 7.25). Therefore, the low-lying states in the energy spectrum
of even-even nuclei are expected to be composed mainly of Jft = 0+-coupled
pairs. This concept is also supported by experimental evidence on ground-state
spin of even-even semi-magic nuclei. They occur considerably lower than the
Jft :f 0+ states. Therefore, it is meaningful to characterize the shell-model config-
urations of n particles distributed over the i orbitals by the number of particles
combined in 0+ pairs. For this purpose, the seniority quantum number was in-
troduced (called v), denoting the number of unpaired particles. Thus (n - v)j2

:>
<II
~
.....
z
w
:::;:
w 0
...J
W
x
ii:
~
0..
ci..
...J
«
~ -1 2
w
:::;: c = '2991:'
22
• = '1 h91.i
ii:
w
0..
x Fig. 7.2.5. Some typical two-body matrix elements
w for the 199/2(0), 1h9 / 2(e) and 299 / 2(0) orbitals.
o = '199//
The large preference of the formation of J1r = 0+-

I -2 and, to a smaller extent, of J1r = ~ pairs, is

illustrated. All other (higher J1r values) remain

-
almost degenerate at zero binding energy
4 6 8
l

230
angular momentum 0+ - coupled pairs are present in a seniority v configuration
with n particles. An orthononnal set of v = 0 basis states can be constructed
using algebraic properties of quasi-spin operators (Kennan 1961, Kennan et al.
1961, Helmers 1961, Watanabe 1964, Arima, Kawasada 1964). These operators
can be constructed from the S; (7.12), by using a slightly different nonnalization,
i.e. one calls

s+(j) = .;n;S; . (7.80)

The operator s+(j) is related to the pair creation operator

AjM(ab) == [a! ® at]~ , (7.81)

via the condition

v
1
L"+(j) _ _ 'IfA+ooJ
-.../iVJ~i
( .2)
. (7.82)

One so obtains for a system of n identical nucleons the shell-model v = 0 state

InI2 ni.
, ... , 2 ' V
=O} = IIi [(il(nk/-n
k=I
k k /2)!]I/2 L"+(k)n,,/210)
2 )!{}k! v .
(7.83)

Here, the index k runs over all shells where some ordering is assumed.
Furthennore ilk are the specific degeneracies ilk = jk + 1/2 and nk/2 the
number of pairs in the kth valence orbital. This satisfies
nk
T ::; ilk,
i (7.84)
L
k=I
nk
2
=~.
2
The total seniority v is now defined as the sum of the seniorities v k for the
particles in the subshells or
i
v= LVk. (7.85)
k=I

The general expression for a nonnalized state with nk = qk + Vk particles in the

shell k and with seniority Vk is

(7.86)

where IVk, Vk) is the nonnalized state with Vk particles in the shell k and seniority
Vk. The nonnalization factor is

if (7.87)

231
 qk/2 ::; fh - Vk •

For qk > fh - Vk, the state (7.86) vanishes. As an example one can write a
nonnalized v =2 basis state with unpaired nucleons in the shells a and b as

(7.88)
where now the relations

(7.89)

and
n'
2k ::; fh - bk,a - bk,b , (7.90)

hold.
The relation (7.86) can now be used to calculate matrix elements of shell-
model operators between low-seniority states.
The dimension of the v = 0 space is equal to the number of different ways in
which n/2 pairs can be distributed over the valence subshells. This is typically
of the order of a hundred for medium-weight and heavy semi-magic nuclei with
five or six subshells. When more shells are included the number increases very
rapidly. The size of the model space also increases considerably with increasing
seniority. It is important to note that not all distributions of the pairs over the
shells are equally important. One may adopt a special type of linear combinations,
which can be written in the fonn of a state with n/2 identically distributed pairs.
With this type of states one finds almost perfect overlap with the v = 0 shell-
model ground sstate (Macfarlane 1966). This is the rationale of the broken-pair
model.

7.6.2 Broken-Pair or Generalized-Seniority Scheme for Semi-Magic Nuclei

The ground state for a system of two identical valence nucleons is known to
have J1I: = 0+. In the shell model it can be represented as

S+IO) = L cpas+(a)IO) , (7.91)

a

where the coefficients CPa characterize the distribution of the pair over the valence
orbitals a. This distribution is produced by the short-range part of the shell-model
residual interaction. The broken-pair scheme is now based on the observation that
there is only one state (7.91) which has an energy much lower than all the other

232
states, even though there may be several single-particle orbits that have almost
degenerate single-particle energies. So one concludes that the interaction singles
out one specific coherent pair structure (7.91), henceforth called S-pair, while
other, orthogonal, superpositions of J7r = 0+ configurations for two particles are
as high in energy as the lowest J7r::f 0+ states. Because the two-body shell-model
interaction favours the S-pair structure so strongly, one may now suppose that
the ground state of nuclei with several valence nucleons is also predominantly
composed of S-pairs only. This idea finds support in the observation that the
lowest part of the spectrum for a semi-magic nucleus with several valence pairs
is similar to the spectrum for one valence pair (see also Fig. 7.25). A possible
interpretation is that in the lowest states of nuclei with several pairs all but one
pair are S-pairs while the last pair has similar configurations as for the nucleus
with a single valence pair. The number of nucleons that do not occur in S-pairs
will be called generalized seniority v g, in analogy with the seniority concept of
the previous Sect. 7.6.1. One also speaks of broken pairs referring to particle
pairs which are not S-pairs.
A shell-model basis which is labelled with this generalized-seniority quantum
number Vg (or equivalently with the number of broken pairs) must be generated
by explicit construction. For a system of n identical nucleons, this is done by
adopting a step by step procedure, starting from the Vg = 0 or zero-broken-pair
(Obp) state, which is built from S-pairs only

(7.92)

Here No is a non-trivial normalization factor. It has often been demonstrated

that this multi-S-pair state is an excellent approximation to the low-seniority
shell model LSSM or even to the exact shell-model ground state (Gambbir et al.
1973). This is true only if the coefficients 'Pa in the S-pair operator (7.91) are
determined by minimization of the multi-pair state (7.92). It turns out that these
coefficients 'Pa do not depend very much in practice on the pair number nj2.
Unless stated otherwise, we shall mean by generalized seniority, the number of
particles which do not occur in S-pairs and where the 'Pa are calculated for each
specific nucleus. Thus Vg is related to the number of broken S-pairs nbp for a
system with even particle number as

(7.93)

by definition.
Starting with this Vg = 0 (Obp) state (7.92), one then constructs the Vg = 2
(lbp) states by replacing one S-pair creation operator in (7.92) by a two-particle
creation operator (7.81), obtaining the 1bp configuration

(7.94)

It is to be understood that for J7r = 0+ one constructs linear combinations of

such states which are orthogonal to the Vg = 0 state (7.92). This is similar to the

233
seniority truncation scheme, with the difference that here the structure of n /2 - 1
pairs is frozen. This in fact is responsible for a large reduction of the dimensions
of the broken-pair configuration space. As the structure of (7.94) is governed by
the pair operator AiM the number of Vg :5 2 (zero or one bp) states is equal to
the number of two-particle shell-model configurations, irrespective of the total
number of valence nucleons n under consideration. Similarly the Vg = 4 (2bp)
states are constructed as linear combinations of

(7.95)

orthogonalized to the previously obtained set of v 9 :5 2 states. The dimension of

the Vg :5 4 space is equal to that for the four-particle shell-model problem. By
such a procedure one may continue to obtain a complete hierarchy of broken-
pair basis states characterized by the quantum number Vg or equivalently the
number of broken pairs according to the definition (7.93). One hopes that this
classification of states provides a good truncation scheme, such that one may
truncate the shell model space to include Vg :5 2 or Vg :5 4 only. Obviously
the Vg :5 n space has the same dimension as the full shell-model space. The
hope that one may apply a Vg truncation in calculations of low-lying states of
semi-magic nuclei is mainly based on the empirical relation
(7.96)

where .1 is half the energy gap between the ground states and the lowest (non-
collective) excited states. One finds L1 ~ 1.5 MeV for Ni isotopes and N = 50
isotones, .1 ~ 1.2 MeV for Sn nuclei and for N = 82 isotones while L1 ~
0.9 MeV for the Pb isotopes. If the quantity 2.1 is interpreted as the energy that
is required to break up each next pair, (7.96) may also be valid for Vg > 2.
In that case Vg truncation corresponds to a rather well-defined energy truncation
of the shell-model space. It should be emphasized, however, that (7.96) is only
a rough estimate which may be violated for collective states which are pushed
downwards in energy by the coherent action of multipole forces. It is also unclear
whether (7.96) may indeed be applied for the case of large v g • Nevertheless, a Vg
truncation scheme seems the only reasonable way to treat the shell-model problem
for semi-magic nuclei with many valence nucleons. It has been demonstrated
(Allaart, Boeker 1971) that the Vg :5 2 model provides a very good tractable
approximation of the v :5 2 shell model, which involves much larger dimensions
of the model spaces. In Table 7.5 these dimensions are listed for a typical case.
Indeed, Vg :5 4 calculations have actually been performed and are discussed
by (Allaart et al. 1988), while it has only recently been feasible to carry out
v :5 4 shell-model calculations for heavy semi-magic nuclei and then still with
additional truncations (Scholten 1983).
A slightly different notation of broken-pair states that has been used quite
often by Gambhir et al. (Gambhir et al. 1969, 1971, 1973) is

with (7.97)

234
Table7.5. Dimensions of model spaces with definite generalized seniority. The example of 112Sn is
shown with 12 valence particles in the orbitals 197/2. 2ds/ 2 • 2~/2' 381/2 and Ihu/2 (taken from
(Allaart et aI. 1988»
J7r "9 =0 "9 =2 "9 =4 "9 =6 "9 =8 total. all "9
0+ 4 45 467 3318 56,907
2+ 9 157 1967 15149 267,720
4+ 7 190 2854 23372 426,558
~ 3 158 3006 26588
g+ 2 133 2630 22668

16+ 3 255 4577

22- 2 184
22+ 118
31+
31- 4IJ "9 -- 12

= II( til (n/2)!

1 (s+t/
+ - 2 (7.98)
Tn/2 k Uk .

We illustrate the above broken-pair model with the calculation of the N =50
semi-magic nuclei 88S r and 90Zr (Fig.7.26). The calculations were carried out
by Allaart et al. (Allaart et al. 1988), using a model space of 10 shells listed in
Table 7.6 and using a Gaussian force.

V(lr, -,.,0 =V(lliE +tPro) exp { - GI>' -,.,I)'} . (7.99)

The interaction parameters used were chosen as V = -23MeV, t = 0.6 sup-

plemented by a triplet-even force between protons and neutrons with strength
VlE = -20MeV and a range parameter such that J.l1I =0.90 where II is the har-
monic oscillator parameter. The largest 1bp components are listed with a coding
of 1 == 2Pl/2, 3 == 2P3/2' 5 == 115/2, 9 == 199/ 2. The symbol "s'" means that no
specific component is dominating the excited 0+ state.

235
 6 88 Sr 90 Zr
8+
... ,
6+
...... ,... ...' - - - 1.00 (9,9)
"'--- 0.99 (9,9)

5
2-
--- 096 (5,9) 2-
--- 0.75 ( 5,9)

6- 0.&6 (3,9)
4+ 3- ------ ~1.00 (5,9)
7- --, ,0.99 (5.9) 0.78 (5.9)
3- ........- - - - - , 0 . 8 8 (5,9) 5- 'Z.:..-::::;::< 0.94 (5.9)
5- _ _ _ -"=-<==:::"0.95 (5.9) '0.86 (5.9)
><::i 4 _ -- 0.77 (5.3) 6- 4- '0.96 (3.9)

~
4- _6_ _ - - - 0.94
- - - - - - 0.98
(3.9)
(3.9)
5-
--- 0.94 (3,9)

-------
1.00 (5,7)

t ~-----==== 0.93
5- (3,9) 8' _ - - - - - 1.00 (9.9)
6' __ - 0.99 (9.9)
1.00 (3.1)
0.86 (5.1) / - - - 0.98 (9.9)
"0' ~--
------ "s'"
/

""
4'
3
3- 4- 3-
- - - - - - 0.89 (3.9) ""...- - _ _ _ 0.63 (3.9)
...... - - - 0.99 (1.9)
5- 0.99(1,9)
2' _ - - - - - - 0.76 (9,9)
,, _ _ _ 0.86 (3, /)
2 2'
,
"" 0' "s'"
EXP. 1bp EXP. 1 bp

Fig. 7.26. One-broken-pair (IBPA) spectra for ~Srso and :gZrso. The calculations were carried
out in a model space of 10 subshells, listed in Table 7.6 and using a Gaussian force V(r) =
Yo(Ps + tPT)exp{-rz/pZ} (sometimes supplemented with a triplet-even force (ViE) between
protons and neutrons). The parameters were Yo = -23MeV, t = 0.6, ViE = -20MeV and a
range parameter pv = 0.9 with v the hannonic oscillator range. The largest IBP components are
listed with the coding 1 == 2Pl/Z; 3 == 2P3/z; 5 == 1/5/Z; 9 == 199/Z' The symbol "S'" indicates an
unspecified mixture of 0+ pairs, orthogonal to the 0+ ground state, taken from (Allaart et a1. 1988)

236
Table7.6. Single-particle energies and occupation probabilities for N = 50 isotones in a (broken-)
pair model calculation. These were deduced from data on the adjacent odd-mass isotones with
interaction parameters, given in the caption to Fig. 7 .26. The neutron energies were the same for both
nuclei (taken from (Allaart et aI. 1988))

nlj 1/~ IIi 2p~ 2p!

88S r e,,(MeV) -14.0 -11.2 -10.2 -8.8
protons occ. % 98 95 85 23
90Zr e,,(MeV) -16.5 -14.0 -13.5 -11.5
protons occ.% 98 97 94 66
neutrons e,,(MeV) -18.5 -15.5 -12.5 -11.5
occ.% 100 100 100 100

Ig~ 2d~
2 3s! 24 Ig~ Ih¥
88S r -7.4 -3.5 -2.8 -2.0 -1.5 +0.3
prolOns 4.5 0.6 0.3 0.4 0.5 0.2
90Zr -10.5 -7.8 -7.0 -5.5 -4.5 -3.0
protons 10.5 1.3 0.5 0.6 0.5 0.3
neutrons -11.1 -6.4 -5.3 -4.2 -3.7 -2.4
100 0 0 0 0 0

7.6.3 Generalization to Both Valence Protons and Neutrons

One may generalize the broken-pair scheme to systems with both valence protons
and neutrons in a straightforward manner by forming simply the product of proton
and neutron broken-pair basis states. The zero-broken-pair state

(7.100)

may now be obtained by minimizing its energy, including the coupling between
the proton and neutron distributions due to the proton-neutron interaction Hamil-
tonian
H 1rv = 2:= (7l'1VIIV17l'2v2)a~t a!t a V2 a 1r2 • (7.101)
1rl , 7r 2 ,"1 ,"2

We have used the symbols 71' for protons and v for neutrons explicitly. The
energy minimization condition for (7.100) results in a set of coupled equations
for the S-pair structure coefficients <pp and <pn (Egmond, Allaart 1983). It should
now be remarked, however, that the zero-broken-pair state (7.100) may not in
general be a good approximation for the ground state of nuclei with both open
proton and neutron shells. Especially where the strong proton-neutron quadrupole
force induces large admixtures of components with several broken pairs in the
ground state. Empirically it is known that nuclei with several valence protons
and neutrons are soft against quadrupole vibrations or even become deformed.
Such a deformed state corresponds to a mixture of states with many broken

237
pairs in the spherical representation. For such a case the model becomes less
interesting. One should then work in a deformed representation. However, the
generalization presented here may still be of interest for two reasons. First,
one may hope that it is still applicable when either the proton or the neutron
number of valence particles (or holes) is so small, typically one or two, that
deformation effects are still negligible. We shall discuss a few applications for
such cases. Secondly, this may be useful in the microscopic interpretation of
the phenomenological Interacting Boson Model (Sect. 7.7), for one may assume
that the S-pairs, together with a coherent pair structure with angular momentum
J =2, called D-pair, still playa prominent role in deformed nuclei. The broken-
pair model with states built from these S- and D-pairs may then be useful in
the microscopic analysis of the boson model parameters from a shell model
microscopic viewpoint We shall discuss this relation in Sect. 7.7.
Some applications when both proton and neutron broken pairs can be present
are the following:
i) Studies for odd-mass nuclei with Vg = 1 and Vg = 3 can also be carried
out. This is similar to a Vg = 0 + 2 calculation. For 89y, we compare such
a Vg $ 3 calculation (Fig.7.27). We indicate the angular momentum and
parity as 2J1r. In the first calculation, only the proton Ifs/2' 2[>3/2, 2Pt/2
and Ig9 / 2 orbitals are included in the Vg $ 3 model space. The spectrum
called "core-excitation" includes all shells between magic numbers 20 and

Fig. 7.27. Influence of core-exci-

tations on the low-lying levels in
~Yso in a broken-pair (Vg ::; 3)
model calculation. The numbers
:=====:::::-- near the levels indicate twice the
angular momenta. In the first cal-
culation, the proton 115/2.21'3/2.

----- 2Pl/2 and 199/2 shells are in-

cluded in the Vg ::; 3 model space.
The middle spectrum includes all
shells between magic numbers 20
and 82 for both protons and neu-
o 1-
b'Ons. A Gaussian, effective inter-
Vg ,3 Vg~3WITH action (see Fig. 7.26) was used.
VALENCE ONLY CORE-EXCITATION EXP The data are from (Kocher 1975)
(taken from (Allaart et aI. 1988»

238
 6r------------------------------------------------------2-0~·,

~15-,5""
5 15- 16 1r

-;s= 16'

~ ---w ,
2--
11-
4

>()I ~
:E

t
10-
010 rere- 1 0--
3
4,0·8·
? J.r'O+6+ •
,~8

2 4' 0 _ _8_ "

6' g- 0 - - - -6-.
'~ 5- g-
7- '~ ---Z o. 3 7-
2 (1.2)·(3~ --L 6 __4' 45 _ _T
~ ~ 47 _ _
5-
0. 4 + __0'
7 __2'

__2'

EXP.

Fig. 7.28. Broken-pair calculation for 196Pb. For angular momenta J < 13, as many as 20 subshel1s
were included for both proton and neutron orbitals with, however, the restriction that only one
neutron pair was broken or only one proton Ip - Ih pair was present. The percentages of these
proton Ip - Ih configurations are listed. For J > 13, only valence neutrons in the 82 < N < 126
shel1 were considered in a two broken-pair model (v g = 4). The data are from (Van Ruyven et aJ.
1986) and (Roulet et al. 1977) (taken from (Al1aart et aJ. 1988»

82 for both protons and neutrons. A Gaussian effective interaction has been
used (Allaart et al. 1988).
ii) For the even-even Z = 82 closed shell 196Pb nucleus, for angular momenta
J < 13, as many as 20 subshells have been included for both proton and neu-
tron orbitals, but with the restriction that only one neutron pair was broken
or only one proton particle-hole pair present (Fig. 7.28). The percentages of
the proton p - h configurations are listed. For J > 13, only valence neutrons
in the 82 < N < 126 shell were considered in a 2bp (v 9 = 4) calculation
(Allaart et al. 1988).

7.7 Interacting Boson-Model Approximation

to the Nuclear Shell Model

Starting from the broken-pair model where both proton and neutron broken pairs
can be present, it was alluded to in Sect. 7.6, that the proton-neutron interaction
will almost inevitably mix the proton and neutron seniority strongly. In nuclei

239
where such proton-neutron broken-pair excitations will be dominant (for nuclei
with both open proton and neutron shells), one actually enters those regions of
nuclei where the quadrupole degree of freedom starts to playa dominant role in
determining nuclear structure. Collective quadrupole vibrational excitations and
rotational motion and the transitional forms in between determine the low-lying
excited states in those nuclei. Besides the geometric or shape variable models
conceived by A. Bohr, B. Mottelson and J. Rainwater (Bohr 1951, 1952, 1976,
Bohr, Mottelson 1953, 1955, 1969, 1975, Mottelson 1976, Rainwater 1950, 1976)
(a topic which is outside the scope of the present book), an alternative description
based on symmetries in an interacting boson model, a model where s (L = 0)
and d (L = 2) bosons are considered, has been introduced by Arima and Iachello
(Arima, Iachello 1975a, 1975b, 1976, 1978a, 1978b, Iachello, Arima 1988). Both
models, although coming from different assumptions, give a good description of
those nuclei where the quadrupole degree of freedom is dominant. This is exactly
the region of open shell proton-neutron systems, where the broken-pair model,
just discussed, is probably able to give some justification of the Interacting-Boson
model (IBM). Both the Bohr-Mottelson model and the mM have been discussed
in many detail, in particular in recent years in a number of review articles and
some books, and it is not the purpose to go again into these topics in much detail.
Even though the original ideas of the Interacting Boson model (IBM) are highly
rooted within concepts of dynamical symmetries in nuclei, we shall not elaborate
on that interesting aspect of the mM.
We like, however, to mention, just shortly, how the concept of symmetries
in physics has always been a guideline to unify seemingly different phenomena.
So, let us quote some of the major steps using symmetry concepts in describing
different aspects of the nuclear many-body system.

1932: The concept of isospin symmetry, describing the charge independence

of the nuclear forces by means of the isospin concept with the SU(2)
group as the underlying mathematical group (Heisenberg 1932). This is the
simplest of all dynamical symmetries and expresses the invariance of the
Hamiltonian against the exchange of all proton and neutron coordinates.
1936: Spin and isospin are combined by Wigner into the SU(4) supermultiplet
scheme with SU(4) as the group structure (Wigner 1937). This concept has
been extensively used in the description of light a-like nuclei (A =4 x n).
1948: The spherical symmetry of the nuclear mean field and the realization of
its major importance for describing the nucleon motion in the nucleus
has been put forward by Mayer (Mayer 1949), Haxel, Jensen and Suess
(Haxel et al. 1949).
1958: Elliott remarked that in some cases, the 'average nuclear potential could
be depicted by a deformed, harmonic oscillator containing the SU(3) dy-
namical symmetry (Elliott 1958, Elliott, Harvey 1963). This work opened
a first possible connection between the macroscopic collective motion and
its microscopic description.

240
1942: The nucleon residual interaction amongst identical nucleons is particularly
strong in r = 0+ and 2+ coupled pair states. This "pairing" property is
a corner stone in accounting for the nuclear structure of many spherical
nuclei near closed shells in particular. Pairing is at the origin of seniority,
itself related with the quasi-spin classification and group as used first
by Racah in describing the properties of many-electron configurations in
atomic physics (Racah 1943).
1952: The nuclear deformed field is a typical example of the concept of spon-
taneous symmetry breaking. The restoration of the rotational symmetry,
present in the Hamiltonian, leads to the formation of nuclear rotating spec-
tra. These properties were discussed before in a more phenomenological
way by Bohr and Mottelson (Bohr 1951, 1952, Bohr, Mottelson 1953).
1974: The introduction of dynamical symmetries in order to describe nuclear
collective motion starting from a many-boson system with only 8 (L =0)
and d (L =2) bosons is introduced by Arima and Iachello (Arima, Iachello
1975a, 1975b, 1976, 1978a, 1979). The relation to the nuclear shell model
and its underlying shell-structure has been studied extensively (Otsuka et
al. 1978b). These boson models have been giving rise to a new momentum
in nuclear physics research. These symmetries are depicted schematically
in Fig. 7.29.

1932 : Isotopic spin Symmetry----o-.--

1936 : Spin Isospin symmetry-H-++

1942 : Seniority - pairing

1948 : Spherical central field

1952 : Collective model

1958 : Quadrupole SU(3)

symmetry J=O J=2

1974 : Interacting Boson

model symmetries
It ~
tIt ~
j J=O J=2
Fig. 7.29. Pictorial representation of the
Bose-Fermi symmetries
most important nuclear symmetries de-
veloped over the years

241
 We like, however, to point out in which way, starting from the shell-model
techniques, described in this chapter (Sects. 7.1 to 7.6), contact can be made with
the proton-neutron mM model, called mM-2. Thereby, we shall in particular
point out in an almost schematic way, the different approximations and trunca-
tions needed on the exact shell-model calculation with all valence nucleons, in
order to come within the mM-2 model space.
First, however, we shall discuss a simpler model, that allows for the onset of
quadrupole motion in nuclei near vibrational regions of the mass table. Thereby,
we shall point towards the major importance of the quadrupole proton-neutron
interaction in establishing smooth, quadrupole collective motion.
The nuclear mean field gives rise to a number of highly stable magic nu-
cleon configurations so that nucleons outside these configurations can be treated
as valence nucleons that mainly determine the low-lying nuclear collective de-
grees of freedom. This process is a first act of truncation of the large shell-model
space into a truncated valence shell-model space. The residual nucleon-nucleon
interaction in the Oliw model space now selects mainly J1r =0+ and 2+ coupled
pairs, and therefore, will separate a highly coherent set of pairs for the low-lying
configurations as a starting point when describing nuclear quadrupole collective
motion (Fig. 7.25). As such, in more realistic cases with many valence nucleons
outside closed shells, a BeS pair distribution (or 0+ pair distribution with correct
particle number) is a very good starting point in describing the distribution of
nucleons over the available fermion orbitals (Sects. 7.6 and 7.7).
Starting from a second quantized form of the paired states, using (7.12)

Cf+ _ 1 " ( l)i+m + + (7.102)

.:Ji - In ~ - aimai-m ,
vil m>O
[with il == j +!
the degeneracy and aim a fermion creation operator in the orbit
denoted by (j, m) == (n, I,j, m)] and

Dj = L(-I)i- m(jmj - mIJO}ajmaj_m , (7.103)

m>O

for a single j-orbit, the lowest 0+ and 2+ excitations for protons and neutrons
separately can be described in the shell model as

(7.104)

(7.105)

242
 Fig. 7.30. The unperturbed seniority
~--2+ tI = 2 proton and neutron shell-
model configurations for situations
f Vnv 2 with n". (and nIl) proton (neutron)
----- v,,)
particles in a single i .. shell, re-
spectively (see (7.106) as well as
1Vnv the two J'" = ~ levels obtained
l1 after diagonalizing the quadrupole
proton-neutron force -K-Q .. . Q"
in the two-level model of (7.106)
and (7.109) (taken from (Heyde, Sau
---o~ 1986»

In the proton-neutron coupled basis, there are two independent 2+ basis states
i.e.

12~) = [(S; )<n,,/2)-1 ()] ()<n~/2)

Dj 2+ ® S; 10) ,
(7.106)
12~) = [( S; )
<n~ /2)-1 ( )] ()<n" /2) 10) .
Dj 2+ ® S;
If the energy of the lowest 2+ states in the proton and neutron space are nearly
equal [what is often the case in experimental situations e.g. Ex(2j) in Sn and
Ex(2i) in N = 82 nuclei] and taken as
(7.107)
then the eigenstates for the coupled proton-neutron system follow from diagonal-
izing in the 2 x 2 model space, the residual proton-neutron interaction (Fig. 7.30).
Using a quadrupole proton-neutron interaction -KQ7r· Qv
and using the Racah
algebra reduction formulae (de-Shalit, Talmi 1963), one obtains for the matrix
element
(v7rv ) = (2~1- KQ7r . QvI2~)
1/2
K N7r Nv 1/2 4
=-S(N7r(1- n7r)Nv (1- nv))
( )
(n7r- 1)(nv- 1)
(7.108)
The energy matrix becomes

(V7rv )] (7.109)
..0 ,
~2+

with as lowest eigenvalue, the expression

243
 0.----------------------------, Fig. 7.31. The energy of the lowest Jr =
zr level (see (7.110», in units of (,./5).
F (where F is defined in (7.108», plot-
/1 2 ted as a function of the product N w N v
for nuclei in the region 1 :5 N w :5 8.
x 3 Here, we denote with N w (Nv)thenum-
[] 4 her of proton (neutron) pairs or Nfl =
ne/2(e == 1I',v) (taken from (Heyde,
v 5
Sau 1986»
o 6

5~~ __ ~~ __~~L_~_ _~~_ _~

o 4 8 U ffi ~ ~ U n ~
- - - N l t NV -

(7.110)

The corresponding wave function reads

(7.111)

So, the lowest 2+ state is a symmetric linear combination of the unperturbed

states and thus seniority becomes strongly mixed in the final 2+ state. Also, the
energy eigenvalue presents a specific N 1tNv dependence shown in Figs. 7.31 and
7.32 which for small N 1t /{}1t, Nv/{}v is almost like ../N1t N v . Recently Casten
et al. (Casten et al. 1981, Casten 1985a, 1985b, 1985c) have pointed out the very
striking observation that experimental quantities such as E z (2t), E z (4j)/E z (2j)
lie on a smooth curve when plotted as a function of the product N 1tNv. At least
in the vibrational regime, the above simple shell-model calculation predicts such
a behaviour for E z (2j).

244
 Fig. 7.32. A detailed fit, using (7.110),
1.0 2t
to the lowest J[ = level in the even-
--<>- Experiment even Te, Xe, Ba and Ce nuclei. Here,
__ Theory we use the degeneracies {}" = {}v = 16

f
_0.8
and a constant value of (/t/5). F =
0.365MeV for all nuclei (taken from
~ (Heyde, Sau 1986»
::;:
>
C>
a::
UJ
z06
UJ
Te
Z
o
;::
;:! Xe
U
x04
UJ

I 0.2

---N--

Since we like to go beyond the vibrational region and also if we like to

handle more realistic cases where particles move in many, non-degenerate shells,
we have to use more general pair creation operators e.g. we use

~ = I>~j"S+(j7r) ,
j"
D+7r,JI. = '~ a. . D-+:11(,171"'1"
" iJ}'1f:,171" . (7.112)
jr,j7l'1

G+1r,P. = '"
~
'V' .
I}w,}",'
G-+:J'fr,}"Ir"p.'
.
31(,j.,,:,

with

S+(. ) = JQj" (a-+: a+' )(0)

J7r J2 J." 3" 0 '
(7.113)
-1/2 )(2)
D-+:·
]",],,/1'
= (
1 + b'3d"/)
. ( a+']" a+'3,,1 I' '

and where the sums go over all single-particle states. Remind that the operator S;
used in (7.112) and within the generalized seniority scheme (see also Sect. 7.6)
and in the further discussion, has a different normalization i.e. for a single j -shell
S; == ..;JiS;' For degenerate orbitals, all Otj are equal (Otj = J!(j + !» but for
more realistic situations, deviations occur and the optimum pair distribution will
have to be determined from a variational procedure for each nucleus (and in prin-
ciple for each excited state). Using all possible angular momenta S, D, G, ... ,
a huge shell-model basis still results.
245
Here, Our Next Step of Truncation Comes in. We truncate the huge space of pair
states (S!! 1014 down to 1()3 for e.g. lS6S m) such that only the S and D pairs
are considered (Otsuka et al. 1978a, 1978b). The argument of truncation here is
again founded by the fact that in studying low-lying collective quadrupole states
we first of all have to take into account the most strongly bound pair states for
the nuclear many-body system. The basis functions are now written as

(7.114)

with
n",
N s" +Nd" =T =N""
(7.115)
n"
Nsp +Nd p = T =N",

where n",(n,,) are the number of valence protons (neutrons) respectively.

As a technical point, we shortly mention that there are problems with the
construction of orthonormal states when many D-pairs are present, e.g.

(S;)N. (D;)~~MIO), and,

(S;)N.+1 (r;tMIO) , (7.116)

where (r;hM is a notation for operators which create 2(Nd - 1) nucleons (e =

7C', v) (e.g., the state ID2, J =O) is not orthogonal to the state ISZ, J =O}) (Rowe
1970). Thus, calculations using fermion pairs are still cumbersome. Although the
fermion pair commutator relations resemble the boson ones, there are corrections
that vanish only in the small n limit.
Here, the Third Approximation Comes in. We map the space built from S, D
fermion pairs and the operators in the S, D space to a corresponding s, d-boson
space (Fig. 7.33). Thus, the corresponding boson state of (7.114) reads

Fig. 7.33. A schematic representation of the mapping proce-

dure, explained in (7.117, 118), in order to construct the boson
Hamiltonian and boson transition operators. Here, F denotes
B the fennion complete space, B the 5, D-pair fermion subspace
and B the corresponding ad-boson space (taken from (Arima,
lachello 1984»

246
Table7.7. Correspondence between lowest seniority sn fermion states and the corresponding sd-
boson states (taken from (Arima, Iachello 1984»
Fennion space B Boson space 8
n= O,V =0 10) N =O,nd =0 10)
n= 2,v =0 stlO) N= 1,nd =0 sfjO)
v=2 DfjO) nd = 1 dtlO)
n=4,v=0 S t2 10) N=2,nd =0 st210)
v=2 stntlO) nd = 1 st dfjO)
v=4 n t2 10) nd =2 dt210)

(7.117)

where now IO} is the boson vacuum state (see Table 7.7 for the fermion-boson
correspondence for the lowest seniority states). The method used starts by equat-
ing matrix elements between corresponding fermion and boson states and thereby
we obtain implicit equations for the boson operator
(7.118)
Thereby, we ask that the lowest eigenvalues and transition matrix elements re-
main equal in the boson space as compared with the fermion space. Thus, we
carry out a boson mapping, not a boson expansion. As a result one gets a num-
ber of implicit equations determining the boson-truncated Hamiltonian H B as
well as the boson truncated electromagnetic operator 0 B as a function of the
fermion quantities (single-particle energies, two-body fermion interaction matrix
elements, fermion charges). This mapping, for a pairing-plus-quadrupole Hamil-
tonian, is now described.
Using (7.118), fermion matrix elements for low-seniority states can be cal-
culated using Racah algebra. Since the starting Hamiltonian for discussing
quadrupole collective motion consists of a pairing-plus-quadrupole part we dis-
cuss both terms separately (Otsuka et al. 1978b).
i) Pairing Term. This part of the Hamiltonian, for a pure pairing force between
identical nucleons (1 «(1 == 71", v) can be written as
(7.119)
Since H does not mix seniority, Ns and Nd will be conserved separately,
and we only need the two-body matrix-elements for all states from the (j)2 and
(j)4 configurations in order to determine, via (7.118), the boson Hamiltonian
uniquely. After some calculations, one gets

HB =c s 8+8 + Cda:- • d+ ~ ~ CL (a:- a:-fL) • (d J)(L)

(7.120)

247
with the coefficients given by

es = (0)2 J =01 Hpair 10)2 J =O) = -G . n ,

Cd = (0)2 J = 21 Hpair 10)2 J =2) =0 ,
CL = (0)4 JI Hpair 10)4 J) - 2ed = 0 , (7.121)
Uo = (0)4 J =01 Hpair 10)4 J =O) - 2es =2G ,
U2 = (0)4 J =21 Hpair 10)4 J =2} - es - cd =2G .

Thus, for a proton-neutron system and considering the pairing term only, a
boson Hamiltonian

(7.122)

(n s " nd, are the number operators) is obtained which can be rewritten (since
es < Cd) as

(7.123)

ii) Quadrupole Term. We map the fermion quadrupole operator, using the ma-
trix element mapping method. Since the quadrupole operator has both senior-
ity changing and non-changing parts, two different terms result in the boson
quadrupole operator. For the seniority changing boson term one can write

QB = ao( d!" s)(2) + L a~L) ((d!" d!")(L)

L
dt) s + ... . (7.124)

By mapping the corresponding fermion and boson matrix elements we get the
equality

(7.125)
or

(7.126)

Similarly. the non-changing seniority part becomes

Q'JJ =Po (d!" J)(2) + L ,8~L,L') [( d!" d!")(L) (d J)(L')] (2) + .... (7.127)
LL'
Mapping matrix elements gives
(sN-1dIlQ'JJlIsN-1d) = {(~t-l D+; J =2I1QFII(~t-lD+; J =2) ,
(7.128)
or

(7.129)

248
The higher order coefficients a~L), ,8~L,L') can, in principle, be detennined using
the same method as outlined here.
Combining all parts, the full image (in lowest order) of the fermion quadrupole
operator becomes

Q<J} = Ke ((d+ s + s+ J)(2) + Xe(d+ d+f2») , (7.130)

with Ke = ao and Xe = f30/ao.

A rather general ffiM-2 Hamiltonian can now be written as
H B = Eo + Cd" nd" + Cd. nd. + K Q (2)
11'.
Q(2)
V + TT V.
y 11'11' + VV +
M lI'V, (7.131)

for which Eo, Cd", Cd., K, XlI', Xv are related to the underlying shell-model struc-
ture that we now study. Here Cd" (cd.) is the proton (neutron) d-boson energy,
nd .. (nd.) the proton (neutron) d-boson number operator, Q~) the quadrupole
boson operator consisting of two terms [see (7.133») with Xe describing the rel-
ative strength and VlI'lI" V vv , MlI'v describe remaining boson residual interaction
terms. In particular, the last term MlI'v describes the Majorana term which we
do not discuss here. Using the above mapping, microscopic estimates for these
quantities ce == cd~ - cs~, K, Xe are obtained (Scholten 1980) and are illustrated
in Fig. 7.34.
We have determined C = Cd - Cs = Gil and this quantity has to be a constant
value independent of the particle number (Fig. 7.34). For the quadrupole operator
parameters, the following behaviour results:
i) the total strength K = K,KlI'Kv should be given by the function (Fig. 7.34)

(7.132)

(here K, is the strength of the original fermion quadrupole interaction).

ii) the relative value Xe then, is given by the expression

(7.133)

We show the dependence of c, K and X e for a series of isotones in the

50-82 shell where all orbitals are considered as one large degenerate
j = 31/2 shell (il = 16) (Fig.7.34) (Scholten 1980). More detailed cal-
culations studying the parameter behaviour have been given by Duval,
Barrett (for two j-shells) (Duval, Barrett 1981a, 1981b) and by Pittel et
al. (for many non-degenerate shells) (pittel et al. 1982) where the ma-
jor trends in c, K, Xe are retained. A full calculation along the above
lines, using a delta interaction and a quadrupole proton-neutron force was
carried out by T. Otsuka (Otsuka et al. 1978a, 1978b). Results for the
case nll' = 6 (0 ::; nv ::; 14) are presented in Fig.7.35a for a single
j (j = 31/2) shell approximation together with a fit (using c, K, Xe'
VlI'lI" Vvv as parameters) for the Ba nuclei (Fig.7.35b). Thus, it appears

249
 Fig.7.34. The dependence of the important
mM parameters e, 11:, Xv on the neutron num-
ber nv as deduced using the shell-model the-
ory with degenerate single-panicle orbitals
(taken from (Scholten 1980»

-------
0·-- -- -- -- -- -- __ __

~
~
Ie
-0.2

-0.4

·1

X"

-I

!58 66 74 82
Neutron ruriler

that many nuclear properties in medium-mass and heavy-mass regions

are rather insensitive to details and only depend on gross features such as
- the existence of closed shells at 50, 82, 126
- the number of valence protons n 1r and neutrons nil.
So, we have shown how, starting from general shell-model techniques and
using the pairing and quadrupole force components, a rather interesting approxi-
mation to the shell-model could be determined. The basic approximations are the
S, D pair truncation and subsequently, the boson (8, d) space mapping. A boson
Hamiltonian is obtained which for the lowest-lying levels is determined via the
shell-model S, D basis and shell-model interactions. This Hamiltonian is able to
describe a large class of collective states in medium heavy and heavy nuclei.

250
 E E
(a) (MeV)
(MeV) n".=6

4 4 20

3 3
0; 4+I
4+,
2:
2 2 10

2; 2;

o 8 16 24 32
n"
Fig. 7.35. (a) The energy spectrum of even-even nuclei, for a fixed proton number n .. = 2N.. = 6
and varying neutron number 0 ~ nv ~ 32 in the single-j shell approximation. We give the J" = O+,
2+, 4+ and 6+ levels (taken from (Otsuka et al. 1978a». (b) Calculated energy spectra in the even-
even Ba (Z = 56) nuclei. The circles (2+), squares (6+), triangles (4+) and diamonds (0+) denote the
experimental data (taken from (Arima, Iachello 1984»

08202850 82 126 184

E(MeV)
I

o~~+---~--~------~------~
K(MeV)
-02

+1
Xv 0
-I
Fig.7.36. Behavior of the mM-2 parameters
+1
e, K, and Xv as a function of N .. and of X.. as
X". 0 a function of the nucleon number (taken from
-I
(Wood 1983) and (Wood 1987»
082028 50 82 126 184
Nucleon Number

One can now also, from here on, consider the parameters e, K, XU, ••• , as free
parameters that are determined in order to describe nuclear excited states opti-
mally (Wood 1983, 1987) (Fig.7.36). In order to carry out such a program, the
general Hamiltonian H B needs to be diagonalized in the boson basis of (7.117).
A general code has been written by T. Otsuka for performing this task (NPBOS)

251
2

50 58 66 74 82 50 66
58 74 8250 58 66 74 82
Neutron number
Fig. 7.37. Comparison between a typical mM-2 calculated spectrum for even-even Xe nuclei and the
data (taken from (Puddu et at 1980»

and I show some typical results (Fig. 7.37). The general strategy for carrying out
an mM-2 calculation is schematically depicted in Figs. 7.38 and 7.39.
So, by this small Sect 7.7, with its relation to the general problem of treating
a large number of valence protons and neutrons outside of a doubly-close shell
configuration, it should have become clear that, within a number of reasonable
approximations, a nice link between the nuclear shell-model as such and a boson
model description of low-lying collective quadrupole excitations can be made.

FULL FERMION SHELL-MODEL VALENCE

SPACE

A.Z.N;closed shells

I
PAIR APROXI MA liON

!
S.D PAIR TRUNCATION

1
s.d BOSON MAPPING

+ HAMILTONIAN HB

I Fig. 7.38. Schematic representation of

PARAMETERS the approximations underlying the inter-
Ed' "It. Xn.X v.... acting boson model when starting from
the complete shell-model valence space

252
 1T V

_ _ 2dlllz
-3sl/z d.". _ _ _ L=2 d v _ L=2
- l h ll12
-- -IO"l
-- """**"- 2d Iz
'Iz
~2 N,,=7

~ Proton Bosons Neutron Bosons

Fig.7.39. Schematic outline of how to carry out an actual mM-2 calculation for a nucleus like
l~Xeot: (i) First determine the nearest closed shell (50,50), (ii) Determine the number of bosons
(valence particle (hole) number divided by two), (iii) Make an estimate of the major parameters
e, It, X.. and Xv, according to (7.132,133) (taken from (Arirna, IacheUo 1984»

253
8. Self-Consistent Shell-Model Calculations

8.1 Introduction

In the preceding chapters we have discussed the necessary methods to study the
nuclear structure of nuclei throughout the nuclear mass table, excluding strongly
defonned nuclei (Bohr, Mottelson 1975) which are outside the scope of the
present presentation. We have discussed the short-range (pairing) properties of
atomic nuclei as illustrated most nicely near closed-shell nuclei. We have ap-
plied the concept of pairing to nuclei with many valence nucleons outside a
single-closed shell nucleus, or even to nuclei with a number of valence protons
and neutrons outside doubly-closed shell nuclei (Chap.7). Also, at closed shells,
particle-hole excitations show up as elementary modes of excitation and have
been studied in a IDA and RPA approximation (Chap. 6). In the study of the
latter chapter, extensive use was made of the methods of second quantization,
developed in Chap. 5. In most applications, we started from an average field that
was detennined in a phenomenological way [harmonic oscillator one-body po-
tential (Chap. 3)] and a residual nucleon-nucleon interaction was used (effective
matrix elements, schematic interaction or realistic interaction) which was not
self-consistently detennined with the one-body potential.
The method of detennining the average one-body potential from the nucleon-
nucleon interaction (Hartree-Fock method) was shortly outlined in Chap. 3. This
method, in principle, allows for a self-consistent study of nuclear structure starting
from a given nucleon-nucleon force Vii. This method is a rather ambitious one
since it faces the problem of detennining at the same time the average nucleonic
properties (binding energy, nuclear radii, density distributions, ...) and the excited
states in each nucleus (with the constraint of studying spherical nuclei near closed
shells only). This ambitious task is schematically drawn in Fig. 8.1 pointing out
that three orders of magnitude should be bridged by a single nucleon-nucleon
interaction. Because of the need to describe nuclear average properties, the force
should establish correct saturation properties, a condition that was not needed for
describing the excitation spectra of nuclei with a few valence nucleons outside
closed shells. So, the nucleon-nucleon force will be parameterized with a relative
small number of parameters, to be detennined the same throughout the nuclear
mass region. In the next few sections, we shall outline the construction of a
nucleon-nucleon interaction of Skynne type (Skyrme 1956, Vautherin, Brink
1972, Negele, Vautherin 1972) with a subsequent application to ground-state

254
 6
1.5 I- 2+
)j+
1.0 I- 4+
'2
0.51- 2

0,- 0+

1000

>CIJ
:::E

>-
C> 500
a:
w
z
w
C> -90
Z
is
z
iii

2D 50 250 A

Fig. 8.1. Schematic illustration of a self-consistent calculation where, with a single nucleon-nucleon
interaction, the global properties (nuclear binding energies, radii, densities, ...) and the local prop-
erties (excited states in each nucleus, see insert) are determined. The difference in energy scale of a
factor 1()3 is presented on the energy scale with the ''magnifier''

properties. Next, we shall illustrate the value of the Skynne forces as an effective
interaction for describing detailed nuclear properties.
The spirit of this last chapter will be slightly different from the early chapters
where a fully deductive method with specific and detailed derivations was carried
out. Here, we shall discuss and present more the state-of-the-art that can be
reached in self-consistent shell-model calculations. This Chap. 8 will rely heavily
on developments that were done in the theory group of the Nuclear Physics
Institute in Gent.

255
8.2 Construction of a Nucleon-Nucleon Force: Skyrme Forces

8.2.1 Hartree-Fock Bogoliubov (HFB) Formalism

for Nucleon-Nucleon Interactions Including Three-Body Forces
The most general form of a Hamiltonian where besides two-body forces, also
three-body forces occur is

H = T + !2 'L..-t
" If;I,). + !6 'L..-t
" W·I,),. .. L
'
(8.1)
i,j i,j,k

with T the kinetic energy, Vij the two-body and Wi,j,k the three-body force.
Using second-quantization, this expression (8.1) can be rewritten as (Chap. 5)

H =L{o:ITI,}a:a..,+ 1 L {0:,8IVI,o}~a:apa6a..,
cr,.., cr,p,..,,6
+k L {0:,8eIWI,oJ.t}nas~:apa;a/Ja6a.., . (8.2)
Q,fJ,'Y
6,C,1J

Here, 0:,,8, ... denote the single-particle quantum numbers characterizing the
orbitals and nas means normalized antisymmetrized matrix elements.
The vacuum of the a: operators is the state with no particles present I}. It is
sometimes more interesting to define a new reference state (a doubly-closed shell
nucleus), a state which acts as a new vacuum state for more general operators
than the air (c~, being linear, unitary combinations of the air) with the condition

(8.3)

(8.4)

Using now Wick's theorem (Chap. 5), the Hamiltonian (8.2) can be rewritten as
follows (condensed form)

(8.5)

with

H(O) = L{o:IT + !U + !U'I,}ucr.., +! L(Ll + Ll')crpKpcr , (8.6)

cr,.., cr,p

H(2) =L{o:IT+ U+ !U' +PI,}N(a:a..,)

cr,..,

+! L [(Ll + Ll')crpN (a:ap) + hC] , (8.7)

cr,p

256
 H(4)=! L (a.8IV+V'hD}N(a:apa6a..,)
a,fJ,..,,6
+ A L [(a.8eIW I1'DJL}rWIl:6..,N (a:apa!a,,) + he] , (8.8)
a,fJ,..,,6
and
H(6) = 316 L (a.8eIWhDJL}nasN (a:apa!a"a6a..,) . (8.9)
Q,/J."t
6,c,,,

Since H(O) only contains contractions, the expectation value of the Hamiltonian
in the new vacuum becomes
gO) = H(O) = (OIHIO) . (8.10)

In the above expression for the Hamiltonian, we have used the following notation:

(8.11)

(8.12)

(aIUh) = L(a.8IVI1'D}nasl.JfJ6 , (8.13)

fJ,6
(8.14)
fJ,6
which represents a part of the average field coming from the three-body forces,
since V' is defined through the relation

(8.15)
e,,,
.1a{3 = !L (a.8IVI1'D}nasll:6-y , (8.16)
..,,6
denotes the pairing potential and similarly

!
.1~fJ = L(a.8IV'I1'D}nas Il:6.., , (8.17)
..,,6
is a contribution to the pairing potential due to the three-body forces.

(8.18)
fJ,e,6,,,
then is a pairing contribution, typical for the three-body forces, and making up
part of the single-particle energy.
The HFB method now tries to determine that linear transformation (8.4)
which minimizes the HFB ground-state energy E(O). As was shown, in a more
simple case in Chap. 7, the term H(2) will become diagonal in the new basis c~.

257
 i) One can now, starting from the given basis a~ and the corresponding known
wave functions Cf'a(r) try to detennine the transfonnation quantities X and
Y which accomplish the above task (Waroquier 1982). This procedure results
in the HFB secular equations

L(aIT + U + ~U' + PI!')Xi-y + L(Ll + Ll')a,8Yi,8 = XiaEi ,

~ ,8
(8.19)
LhlT + U + ~U' + Pla)Yi~ + L(Ll + Ll')~,8Xi,8 = -YiaEi .
~ ,8

Solving these equations (including the three-body force through the U', P
and Ll' fields), H(2) becomes diagonal, E(O) minimal and H(4) and H(6) are
the remaining residual interactions (prove that E(O) is minimal, supposing
that the HFB equations (8.19) are fulfilled).
ii) Using the theorem of Bloch and Messiah (Bloch, Messiah 1962), it is possi-
ble to show that the general, linear transfonnation of (8.4) results in pairing
effects so that each state is connected via pairing to just one other state. In
this situation, (8.4) can be reduced to the fonn
(8.20)

Now, the single-particle basis a and the amplitudes Ua , Va are unknown and
have to be detennined through the condition of getting H(2) in diagonal fonn
within the new representation. Using (8.20), the H(2) in (8.7) becomes
H(2) = Hf~) + H!li/ , (8.21)
with

Hg) = L ba~( uau~ - VaV~) - (Ll + Ll')a"Y(uaV~ + VaU~)]C~c~, (8.22)

and

H~~ = L [1]a~UaV~ + ~(L\ + L\')a"Y· (uau~ - VaV~)] (c::yca + c~c~) .(8.23)

a,~

Here, we have introduced the matrix element

1]a~ == (aiT + U + ~U' + PI!') . (8.24)

The necessary ~d sufficient condition for minimizing H with res~ect to the
reference state 10) is that H!li/ disappears. At the same time, HU becomes
diagonal, leading to the BCS equations (Chap. 7)
1]a~( uau~ - Vav~) - (Ll + Ll')a"Y( uav~ + vau~) = Ea8a~ ,
(8.25)
1]a~UaU~ + ~(Ll + Ll')a"Y(uaU~ - VaV~) =0 .
The solution to these equations (perfonned self-consistently) diagonalizes
Hg), and takes the matrices f!a~ and Kpa in their canonical fonn i.e.

258
 (8.26)

The ground-state energy E(O) then becomes

(8.27)
a a
with

Ca = (alUla) , (8.28)

c~ = !(aIU'la) , (8.29)
self-energy corrections.

8.2.2 Application of HFB to Spherical Nuclei

For spherical nuclei near doubly-closed shells, the pair scattering from occupied
to the much higher-lying unoccupied orbitals is almost non-existing and the HFB
formalism reduces to the spherical HF formalism, for which one has to determine
the HF single-particle states 'Pa(r).
Whenever pairing effects are non-negligible, one most often uses a two-step
procedure: (i) solve the HF equations with given BCS quantities in order to
determine the HF states 'Pa(r) and, (ii) a second minimization is carried out
[with the above 'Pa(r)] to solve for the new BCS quantities. Using (i) and (ii)
now in an iterative way leads to the HF + BCS method.

a) Hartree-Fock (HF) Approximation

Here now, the tensor KPa disappears and ea'Y = oa'Y (if a is occupied) and
ea'Y = 0 (if a is unoccupied.)
This now leads to

H(O) = L (aiT +!U + 1U'la) , (8.30)

(~pied)
and

H(2) = L(aIT+ U+ !U'I'Y)N( a~a'Y) . (8.31)

a,'Y

We have studied the above equations in Chap. 5 where the new basis reduces to

a: = CO
a: =
(a: occupied),
(8.32)
c~ (a: unoccupied).
The state a means the "paired" state to a or the time-reversed state to a (with
some extra phase factors, depending on the phase convention used: Condon-
Shortley or Biedenham-Rose) [see (Rowe 1970) and Appendix 1].

259
The tenn H(2) then becomes

H(2) = ~::>~ N (a~aQ) , (8.33)

Q

with e~ the Hartree-Fock energies. The ground-state energy now becomes (using
the presence of three-body forces)

FfO) = "L.J eHFQ

_ 12 La..fJ
(a,BlVla,B}nas
(~) (occupied)

(8.34)
CIt,{J,-y
(occupied)

The remaining tenns H(4) and H(6) then describe the residual interactions
that couple the different particle-hole excitations relative to the vacuum state (a
doubly-closed shell nucleus) IO}.

b) Solving the HF+BCS Equations in a Spherical Single-Particle Basis

In the most general case, the solution of the HF + BCS equations, in an iterative
way, as discussed before is still very complicated (Waroquier et al. 1979, Waro-
quier 1982). Using now a spherical single-particle basis (and the Condon-Shortley
(CS) convention) one has
ra} =sQI- a} , (8.35)
with
(8.36)
and
(8.37)
In the BCS equations, using the CS convention, one can choose Va > 0 and
U a = lu a l(-1)lc with a resulting qp transfonnation

(8.38)
U sing the above restrictions, the BCS equations can be simplified very much
(Waroquier 1982) and solved for the quantities ua(e), va(e), ... with e = 7r, V
(proton, neutron) in the HF + BCS iterative way.

8.2.3 The Extended Skyrme Force

Trying to construct a nucleon-nucleon force that is apt to describe both the nuclear
global and local properties, using a parameterization that remains constant over
the whole nuclear mass region is a most challenging task. Thereby we start from
a Skynne type force (Skynne 1956) as discussed in a basic article of Vautherin
and Brink (Vautherin, Brink 1972). Here, however, only nuclear ground-state

260
properties are concentrated on. Going beyond these global aspects of the nucleus,
one needs to cope with the specific pairing correlations in order to describe
excited states in nuclei. Moreover, one has to treat this short-range aspects of
the force in a consistent way with the other, saturation aspects of the nucleon-
nucleon force. A number of calculations were performed with this aspects in
mind (Vautherin, Brink 1972, Beiner et al. 1975, Liu, Brown 1976, Krewald et
al. 1977).
One can now try to add extra terms, with one or two extra parameters to be
fixed in order to reproduce pairing aspects properly and lead to realistic two-body
matrix elements.
The suggested extension has two-body and three-body interactions: the two-
body part contains an extra zero-range density-dependent term and in the tbree-
body part, velocity-dependent terms are added. This becomes, in a schematic
way
2 - body part : V(rl' r2) = V(O) + V(1) + V(2)
/' + v(ls) + VCoul. + (1 - x3H'o ,
SkE (8.39)

3 - body part: W(rl' r2, r3) = X3 Wo(rl, r2, r3)

+ WI (rl' r2, r3, kl, k2, k3) ,
where v(O), v(1), V(2) and V(ls) have the same structure as the original Skyrme
force (Vautherin, Brink 1972):

i) V(O) = to(l +xoPIT)h(rl - r2) ,

ii) V(1) = ~tl [h(r l - r2)k 2 + kl2h(rl - r 2)] ,
(8.40)
iii) V(2) = t2k' . h(rl - r2)k ,
iv) V(ls) = iWO(O"I + 0"2) • (k' x h(rl - )k) .
r 2

Here, k denotes the momentum operator, acting to the right

(8.41)

and

(8.42)

acting to the left. The spin-exchange operator PIT has been defined in Chap. 3
and the Coulomb force has its standard form. The density-dependent zero-range
force Vi> reads

(8.43)

In the three-body part, one has the original term Wo

261
 (8.44)

to which a velocity dependent zero-range tenn Wt is added, given as

Wt = it4 [( k~~ + k~ + k~ )S(rl - r2)S(rl - r3)

+ S(rl - r2)S(rl - r3) (kr2 + kt + k~l)] . (8.45)

We point out that only a fraction X3 of the original three-body tenn Wo is retained,
but a fraction (1 - x3)l'o of a density-dependent two-body force is added.
One can show that both interactions Yo and Wo contribute in the same way
to the binding energy in "time-reversal invariant" systems (even-even nuclei).
Thereby, the parameter X3 remains as a "free" parameter relative to the nucleon
ground-state properties and has to be detennined such that properties of excited
states can be well described.
One can express the ground-state energy corresponding to the Skyrme force
given above in (8.39) in tenns of a number of elementary density functions as

EftF = J
H(r) dr , (8.46)

where H(r) describes the Hamilton density [expressed in tenns of (!q(r), Tq(r),
M q(r), Sq(r), V q(r) and Jq(r): the nucleon density, the kinetic energy density,
the current density, the spin density, the spin-kinetic density and the spin-current
density function, respectively] (Waroquier 1982).
Minimization of the HF ground-state energy against independent variations
of the single-particle wave functions Cf',,(r) leads to

SH(r) = 0 for SCf',,(r) (all a), (8.47)

and results into a differential equation for the Cf',,(r). For spherical nuclei, the
radial part can be separated from the standard angular part and results in a radial
differential equation for the part Cf',,(r).
For spherical even-even nuclei (time-reversal invariant systems), the densities
Sq(r), Mq(r) and Vq(r) disappear and the contribution of Yo and Wo to the
Hartree-Fock ground-state energy becomes equal i.e.

(8.48)

with (!p(r), (!tot(r) the proton and total nucleon density, respectively. This does
not at all imply that both interactions Yo and Wo are identical with respect to
evaluating the corresponding two-body matrix elements.
We give, in detail, the Hamiltonian density as a function of the remaining
densitiy (!q(r), Tq(r) and Jq(r):

262
 .
kinetlc 1;,2 ()
energy --+ 2m T r ,
V(O) --+ !to [(1 + xo/2) l'~t - (! + xo) (l'! + l'!)]
V(1) --+ Itl [2l'totTtot -l'pTp -l'nTn] - f2tI
x [2l'tot V 2l'tot - l'p V2l'p - l'n V2l'n] + I~ tl [J! + J!] ,
V(2) It2 [2l'totTtot + l'pTp + l'nTn] + 3~t2
--+

X [2l'tot V 2l'tot + l'p V2l'p + l'n V2l'n] - I~ t2 [J! + J!] ,

V(ls) --+ -!W& [l'totV, Jtot + l'P V· J p + l'n V· I n] ,

VCoul --+ !l'p VC(r) ,
(1 - X3)Vo --+ HI - X3)t3l'pl'nl'tot ,
X3 Wo --+ !X3t3l'pl'nl'tot ,

WI --+
t4[ 2 122 122
24 -l'pl'n V l'tot - '2l'p V l'n - '2l'n V l'p

+ 2Ttotl'pl'n + Tnl'! + Tpl'! + !l'tot V l'p . V l'n

+ ll'p(V l'nl + ll'n (V l'pl + !J!l'n + !J!l'p] ,

and

VC(r) =J l'p(r') Ir ~ r'l dr' . (8.49)

The stationary condition for the ground-state energy now leads to

6(E~) = 6«(HFIHIHF}) =6 J H(r)dr=O. (8.50)

Since single-particle wave functions need to be normalized, (8.50) leads to a

constrained extremum problem

6 J[H(r) - Le~l'~~l'~~(r)] dr =0. (8.51)

Ctq,q

The specific form of the Skyrme force (zero-range terms) makes this variational
problem more tractable since we know H(r) as a function of the elementary
density function. This results into

6H(r) =f( 6l'q, 6Tq, 6J q)

=L [2:,2(r)6Tq(r)+Uq(r)6l'q(r)+Wq(r)'6Jq(r)], (8.52)
q q

263
where the coefficient m;(r), Uq(r) and W q(r) define the nucleon effective mass,
the potential Uq(r) and spin-orbit potential W q(r), respectively. Here, we only
give the effective mass in detail
1;,2 1;,2 1 1
2m;(r) =2m q + 4" (tt + t2) l!tot(r) + 8 (t2 - tt) eq(r)

(2
1 t4 etot(r) - eq(r) .
+ 24 2) (8.53)

The other terms Uq(r) and W q(r) are given in (Waroquier 1982).
Finally, one obtains the differential equation

-V· [2:;(r) VCf'a 9(r)] + [Uq(r) - iWq(r)· (V xu)] Cf'a 9(r)

= e~Cf'a9(r) . (8.54)

This equation resembles very much a one-body SchrOdinger equation with effec-
tive mass m;(r). Using the spherical shell-model basis for expressing Cf'a 9(r), a
simplified radial equation results where now, the functions 1;,2/2m;, Uq and W q
become functions of the radial variable r, only.
Since we have now

(8.55)

7,(') = 4~ p2i. + 1)v:, W~:' y + I. (l. + 1) /r' . '1':, (r)1

= TQ(r) , (8.56)

Jq(r)=- 4~L)2ja+l)v;9 [la(la+l)+~-ja(ja+l)]

a9
1 2
X -Cf'a (r)l r
r 9
= Jq(r) , (8.57)

the purely radial equation for Cf'a 9 becomes finally

cPCf'a (r) [2m;(r) d ( 1;,2 )

+
----'-:.-:9!:__
dr2 1;,2
-
dr 2m;(r)
+-
r
2] -drd Cf' (r )
a9

+ [2";.~(r) (e~ _U,(r) + ~W,(r) [I.{l. + 1)

+~-ja(ja+l)]) - :2 la (la + 1)] Cf'a9(r) =0 , (8.58)

with

264
Before coming to a discussion of how to detennine the parameters in the SkE-
forces, we like to make the following points.
We denote, in a shorthand notation, the extended Skyrme force as

(I)
v + (1- X3)Va (8.59)
X3WO+Wl
with

Va = 1i3(1 + Pu )u( (rl + rz)/2 )c(rl - rz) ,

(8.60)
Wo = t3c(rl - rz)c(rl - r3) .
The original Skyrme force looks like

Ol) 1;0 I· (8.61)

The big difference between (1) and (II) is:

- the addition of a velocity-dependent three-body force,
- the introduction of the fraction parameter X3.
In time-reversal invariant systems, Va and Wo contribute the same energy to
Ef!P and so, (I) and (m) give rise to the same HF equations, with (III) defined
as the force

(8.62)
OIQ 1 WO:WI I·
This observation does not at all imply that the force parameterizations (I) and
(III) would be identical. On the contrary, the X3 parameter plays a major role in
detennining, in case (1), correct pairing two-body matrix elements.

8.2.4 Parameterization of Extended Skyrme Forces:

Nuclear Ground-State Properties
We shall determine the parameters that appear in the SkE parameterization of
(8.39) by insisting on a good reproduction of both nuclear matter and ground-state
properties in even-even nuclei.
Nuclear matter is described as a medium with infinite dimensions, a unifonn
density and an equal number of non-interacting protons and neutrons, without
distinction (the Coulomb force turned off).
The nuclear matter binding energy per nucleon (E / A) amounts to -16 MeV.
Variations in the density e around the equilibrium density, as a function of the
quantity a == (N - Z)/A, can be written as

265
 E/A(e, ci) = (E/A)eq. + &K' ( e ~FeF y
+ aT' o? . (8.63)

The coefficient K' is the nuclear incompressibility, defined by

K' = e~ (fPE;A)
8e IFflF

= !k 2 (fPE/A) =!K (8.64)

9 F 8k 2 k=kF
9'
where kF denotes the Fermi momentum.
The numerical value of K is uncertain and lies between 100 and 300MeV.
The coeffcient aT is called the isospin symmetry energy term. An experimental
value around 25 to 30 MeV is estimated. In the Fermi theory of nuclear matter
aT can be expressed as (cF = (1i,2j2m)k~)

(8.65)
with a resulting value of aT = 12.3 (±0.51) MeV.
In extending the non-interacting Fermi nuclear matter theory into a theory
of an interacting system, the Landau description of nuclear matter is obtained
(Landau 1956). Using the SkE forces of (8.39), a value of the binding energy
per nucleon is obtained as

(E)
A run =
/i,2 3 2 3 3 ( ) 2
2m SkF + gtoeF + 80 3tl + 5t 2 eFkF
1 2 3 2 2
+ 16t3eF + 160t4eFkF . (8.66)

The saturation condition for the nucleon-nucleon force leads to an extra condition
on the SkE parameters

/i,2 6 3 2 1 ( ) 4
2m SkF + 471"2 tokF + 871"2 3tl + 5t2 kF
1 5 1 7
+ 671"4 t3 k F + 1571"4 t4kF = O. (8.67)

(8.68)

We have now three equations with the unknown quantities to, (3tl + 5h), t3 and
t4 for a given set of nuclear matter properties kF' (E/A)run and I<.
The nucleon effective mass becomes
/i,2 /i,2 1 1
--
2m*
= -2m + -(3tl
16
+5h)eF + -t4e~.
32
(8.69)

266
If t4 'f 0, the parameters K and (m* 1m) can vary independent of each other
(this property can be used to reproduce the correct level density near the Fermi
surface in finite nuclei). So, given (EIA)nm, K and (m* 1m) in nuclear matter,
the quantities to, t3, t4 and (3tl + 5t2) can be determined uniquely. Separate
values tl, t2 can only be determined making use of properties in finite nuclei.
In a Thomas-Fermi model for the nuclear density, one can derive the expression
for the nuclear mean-square radius {r2} as

2 3 (911" A)2/3 1 [
{r }T-F = 5 -8-
7 2 ( 8 )2/3 2
k~ 1 - 8111" 911"A kF(9t l - 5t2) Ito
1.
(8.70)
Finally, the spin-exchange parameter Xo can be determined from the isospin
symmetry energy coefficient a r since

h2 1 2 tok~ ( ) t2k~ t3 6 t4 8
ar = 2m '3 kF - 1211"2 1 + 2xo + 911"2 - 3611"4 kF - 13511"4 kF . (8.71)

Combining the above results, it is clear that the SkE forces have eight param-
eters to, tl, t2, t3, t4, xo, X3 and the strength of the spin-orbit force Woo These
parameters should determine a large number of experimental quantities (finite
nuclei, nuclear matter).
In aftrst step, the quantities to, t3, t4 and (3tl +5t2) are extracted from nuclear
matter properties «EIA)nm, kF' K and m*).
In a second step, the quantity 9tl - 5t2 is fitted using properties of finite
nuclei, such as the nuclear radii in a Thomas-Fermi model.
In a third step, the parameter Xo is determined from the isospin symmetry
energy a r •
In a fourth step, self-consistent HF calculations for 160, 40Ca, 48Ca, 9°Zr,
132Sn and 208Pb are carried out. From these calculations, a value for Wo is
extracted as Wo = 120 MeV fms. In this step, a very good reproduction of (E I A)
is imposed and, to a minor extent, the reproduction of the HF single-particle
energies, more in particular in the vicinity of the Fermi level.
One performs steps 1 to 4 a number of times in order to obtain an idea on
the influence of the nuclear matter properties on the Hartree-Fock self-consistent
aspects of finite nuclei. In this way, parameter regions can be delimitated. Finally,
four parameterizations SkE1, SkE2, SkE3 and SkE4 (Table 8.1) are retained,
giving good Hartree-Fock results (Table 8.2). In Fig. 8.2, we show a nomogram,
illustrating the parameter interval for constant kF = 1.33fm-1 and (EIA) =
-16MeV. For given value of m* and K, independent of each other, to, t3, t4
and (3tl + 5t2) can be uniquely determined. The dashed line illustrates the SkE2
parameter set. We also illustrate, in Figs. 8.3 and 8.4, the HF single-particle
energies in 160 and 208Pb. In the above tables and figures, we also give the
results of the SkIll force, the Skyrme force used originally (with X3 = 1 and
t4 = 0) (Beiner et al. 1975). From the above results, a number of rather general
remarks can be made:

267
Table8.I. illustration of a number of parameterizations (SkE2 and SkE4) yielding suitable values of
nuclear matter and ground-state quantities for doubly-closed shell nuclei. We also compare with the
original Skynne interaction Skill (Beiner et al. 1975, Waroquier et al. 1983a)

to tl tz t3 :1:0 W/0
(MeVfm3) (MeV fIns) (MeV fIns) (MeV fm6) (MeV fIns)
SkE2 -1299.30 802.41 -67.89 19558.96 0.270 120
SkE4 -1263.11 692.55 -83.76 19058.78 0.358 120
Skill -1128.75 395.0 -95.0 14000.0 0.45 120
t.t J( (E/A>mn kF m*/m aT
(MeV fIn8) (MeV) (MeV) (fIn-I) (MeV)
SkE2 -15808.79 200 -16.0 1.33 0.72 29.7
SkE4 -12258.97 250 -16.0 1.31 0.75 30.0
Skill 0.0 356 -15.87 1.29 0.76 28.2

TableS.2. Binding energies per nucleon E/A (MeV), proton and neutron points nus radii rp and
r,. (fIn) and charge nus radii rc (fIn), corresponding to various Skynne parameterizations after
self-<:onsistent Hartree-Fock calculations (Waroquier et al. 1983a)

E/A rp r,. rc E/A rp r,. rc

160 90Zr
SkE2 -7.92 2.63 2.60 2.68 SkE2 -8.67 4.17 4.24 4.21
SkE4 -7.96 2.65 2.62 2.70 SkE4 -8.71 4.22 4.29 4.26
Skill -8.03 2.64 2.61 2.70 Skill -8.69 4.26 4.31 4.30
exp -7.98 2.71- exp -S.71 4.27c
40ea 13ZSn

SkE2 -8.56 3.37 3.31 3.42 SkE2 -8.36 4.62 4.84 4.66
SkE4 -8.59 3.40 3.35 3.46 SkE4 -8.36 4.68 4.89 4.71
Skill -8.57 3.41 3.36 3.46 Skill -8.36 4.73 4.90 4.78
exp -8.55 3.36e 3.48b exp -8.36
48Ca 208Pb
SkE2 -8.63 3.39 3.56 3.44 SkE2 -7.S7 5.41 5.57 5.45
SkE4 -8.65 3.43 3.59 3.47 SkE4 -7.87 5.47 5.62 5.50
SkID -8.69 3.46 3.60 3.50 Skill -7.87 5.52 5.64 5.56
exp -8.67 3.54e 3.48b exp -7.87 5.5

- (Sick, Mc Carthy 70) b (Wohlfart et al. 78)

C (Alkhazov et al. 76) d (Sick 73)

e (Shlomo, Schaeffer 79)

268
 E k -I !,
A"-IS MeV; F" 1.33 fm
[MeV.fm8 J
t
5000.

-10000.

-15000.

-20000.
05

25000.
L~~OO----~~~00---s--~20~0~0----~IOO~0----~0~--~1~0~000~.--~2~0~OOO~.--~~~OOO~·
13!I.Stzl [MeV. fm ] !3[MeV. fm S]
• •
Fig. 8.2. Nomogram illustrating the range of parameters in m* 1m. (3tl +Stz). t3 and 4 for a constant
value kp = 1.33fm- 1 and (EIAkn =
-16MeV. The dashed line indicates the parameterization
SkE2 (K is expressed in units of MeV) (Waroquier et aJ. 1979)

EHF
IMeVI
1Me I
SkEI 5IcE2 SkE3 SkE' Exp Sk rn SkEI SkE2 SkE3 SkE' Exp SklII
o ~12' _ _ ......l!!!i.
o
'---------,,-.
2!!a... __ -~
~-------,/ ,,~
--------, ,---..
-10
® ® ~--....
®
--------- "

®
'\ 1ds/2
~
-10

,---, ~

~,..-.. ... -----"-""

~ ~' '-------------,
---------'
-20 ~ -20
~,.-. ....

-30

,r - - - - -
,- -30

-40
~'
~I
,,--- - -40

Fig.8.3. Single-particle states in 160 from a self-consistent Hartree-Fock calculation using various
Skynne force parameterizations (Waroquier 1982)

269
 Pb
82 126
PROTON NEUTRON

SkEI SkE2 SkE3 SkE4 Exp

=ES
-30 '
....!!!a/ ~ _ _
,,..----""
:.= ~
--...Jlli2 -30
---..lUG
~/ ~
~ I===':~
~<= ~
'40
I ~ ~
-40

~f -2!!LL------/
f PJ/2 / '

"
Fig. 8.4. Same caption of Fig. 8.3, but for 208Pb (Waroquier 1982)

i) Deep-lying orbitals are more bound, using SkE1, compared to the other
forces.
i) The level density around the Fermi energy is almost equal in the SkE2, SkE3
and SkE4 parameterization, pointing out that k F and J{ can compensate each
other's influence on the level density.
iii) All four parameterizations give a good reproduction of self-consistent bind-
ing energies. The force SkE4 gives the better results, however.
iv) Saturation properties can best be studied in comparing the nucleon density
distributions (see Fig. 3.17). The density, in particular for heavy nuclei, ap-
proaches the equilibrium density eF in nuclear matter (e F = 0.159 fm -3 for
SkE2).
v) The Skyrme force is a non-local force. The most important consequence
comes about in a large variation of the effective mass ratio m* 1m (Fig. 8.5):
a low-value in the nucleus, even up to the nuclear surface. This non-locality
is needed to reproduce the deep-lying single-particle orbitals. The level-
density near the Fermi energy is too low in general. To reproduce correctly
the single-particle spacings, one needs a value m* 1m S:' 1 near the Fermi
level and m* 1m S:' 0.6 for deep-lying orbitals. Figure 8.5 illustrates this be-
haviour. Since experimental single-particle energies have been extrapolated
from one-nucleon transfer reactions, not taking into account coupling of the
single-particle motion to the nuclear surface collective modes of motion, an
exact reproduction should not be imposed.

270
m"
m 2a8 Pb
Potential
(MeV)
10.

Wn
1.0. a
Wp
0.9 -10.

0.8 -20.
m" neutron
m
0..7 proton -30.

0..6 -40.
Up
-50.

Un -60.

a 9 10 11 12
r (1m)
Fig.8.S. Effective mass ratio m* 1m, the potential Uq (MeV) (q =proton or neutron) and the spin-
orbit potential Wq in 208 Pb, using the SkE2 parameterization (Waroquier 1982)

As a general conclusion, we find that SkE2, SkE4 and the original Sklll
force reproduce nuclear matter and global aspects of finite nuclei very well. A
later discrimination shall have to be made when comparing local aspects in finite
nuclei (Sect. 8.3).

8.3 Excited-State Properties of SkE Forces

Having constructed an extended Skyrme force in order to describe nuclear matter

and ground-state properties in doubly-closed shell nuclei (binding energy, radii,
single-particle spectra, saturation properties), we have determined only seven of
the eight free parameters and retained two rather good parameterizations (SkE2
and SkE4).
In the present Sect. 8.3, we shall use the extra parameter X3 for determining
the forces such that correct pairing properties result throughout the whole nuclear
mass region starting from the known two-particle spectra in nuclei like 180, 42Ca,
SOCa, 134Te, ... . Since these nuclei are mainly studied in a small model space
consisting of an open shell with just two valence nucleons, effective forces and
effective operators have to be determined (Chaps.3,4). Using specific forces
that are fitted to such a small model space (using schematic interactions or
using empirical two-body matrix elements as discussed in Sect. 3.2.2) a good
reproduction can quite often be obtained. These forces, however, can vary very
much when used in different parts of the nuclear mass region. Realistic forces on
the other hand, determined from the free nucleon-nucleon scattering properties,
generate "bare" two-body matrix elements that are almost always not very well

271
adapted to be used in a small model space. The realistic force has to be modified
into a corresponding effective force by considering (through perturbation theory
mainly) effects of configurations outside of the model space (3p - Ih, 4p - 2h
excitations, ...). Also the SkE2 and SkE4 forces have to be corrected for this
"core-polarization" in order to be apt for describing nuclei with just two valence
nucleons in a small model space.
So, we shall study subsequently how to determine the fraction parameter
X3 from particle-particle correlations (Sect. 8.3.1), the particle-hole excitations in
doubly-closed shell nuclei (Sect. 8.3.2) with applications to giant multipole reso-
nances and finally (Sect. 8.3.3) we shall study some effects due to rearrangement
in the density-dependence of the residual interaction.

8.3.1 Particle-Particle Excitations: Determination of :1:3

The case of two particles outside of a closed shell has been studied in detail
in Sect. 3.2 (in coordinate space) and in Sect. 7.3 (second quantization). Here,
we shortly discuss how the effect of configurations outside of the chosen model
space of two-particle configurations can be taken into account in an effective
way into the two-particle model space thereby "renormalizing" the original force
and single-particle properties.
Relative to the situation with a Fermi level, in the single-particle energy spec-
trum (Fig. 8.6) and denoting the single-closed shell as the A-nucleon reference
state IHF), the realistic eigenvector for the A + 2 nucleus can be written as

IA+2;a) =L cmn(a)a~a:IHF)
m<n

+ L
m<n<r
Cmnr,i(a)a~a:a;aiIHF)
i

+ L
m<n<r<_
Cmnrs,ij(a)a~a:a;a:aiajIHF) + '" (8.72)
i<i

-*""~-- -*- -*---

•• + ....
Fig. 8.6. The different components in the wave function of (8.72) for a nucleus with two valence
nucleons outside of a closed shell configuration: besides the two-particle component, the 3p - 1h,
4p - 2h, ... components are also indicated

272
 If we now force the model space to contain only the specific two-particle
configurations, relative to IHF}, the model eigenvector has to be represented by

IA + 2; a}Model = L' c~n(a)a~a~IHF} , (8.73)

m<n

where ' indicates that we only sum over two-particle states with expansion
coefficients c![n(a). We can define a projection operator that projects onto the
two-particle model space such that

IA + 2; a}Model = PIA + 2; a} . (8.74)

Using this operator, the eigenvalue equation in the small model space can be
expressed formally as
[EO' - H(O) - (em + en)] c~n(a)
= L '(mnIVeff(EO')lm'n'}nasc~'n,(a), (8.75)
m'<n!

with
1
Veff(EO') = V + VQ EO' _ H(O) _ QVQ QV , (8.76)

and (Q + P = 1).
Now, the "effective" interaction becomes dependent on the energy EO' itself.
Using an expansion of (8.76), the effective interaction can be rewritten as

(8.77)
or
(8.78)

which is an equivalent expression.

Using the SkE forces, we have to introduce the construction of such an
effective model interaction in order to reproduce the two-particle spectra and
determine the parameter X3.
At present we do not go into details of how to determine from (8.75)
and (8.77) the appropriate two-body matrix elements (Kuo, Brown 1966,1968,
Brown, Kuo 1967, Kuo 1974, 1981).
It occurs that intermediate 3p-lh configurations in (8.77) determine the ma-
jor corrections to the bare SkE matrix elements in order to construct an effective
two-particle model space (Kuo, Brown 1966, Brown, Kuo 1967, Kuo 1974) with
as major characteristics:
i) a large increase of the pairing matrix elements and, to a lesser extent, of the
J7r = 2+ matrix elements,
ii) a slight reduction of the high J7r matrix elements.

273
 ol
000 bl

Fig.8.7. (a) Schematic illustration of the "bare" two-nucleon interaction and the ''polarized'' two-
nucleon interaction where long-range correlations in the core are induced, modifying the original.
"bare" interaction matrix elements (shown in the figure as a macroscopic deviation of the core from
its original. spherical shape). (b) The "bare" single particle motion in an average field and a "dressed"
single particle. dragging in its motion the polarization effect in the core (shown in the figure as a
macroscopic deviation of the core, from its spherical shape following the extra nucleon in its motion
through the nucleus)

One can depict this 3p - 1h intermediate state as a polarization of the nuclear

core. This effect induces a long-range attractive part [quite similar to the effect of
the P2 (cos (h 2) force] acting over distances of the order of the nuclear dimensions
(Fig.8.7a). Detailed calculations of the 3p - 1h correction indeed show that the
intermediate J1r = 2+ states contribute most of the correction.
The effective force can then be specified by the two-body matrix elements
(ab; JMIVelfled; JM) nqq... =(ab; JMIV + V' led; JM) nqq...
+ (ab; JMIl'Jp-lh!cd; JM}nqq...
+ (ab; J MIl-4p-2h led; J M) naB
qq
• (8.79)

Evaluation of the matrix elements occurring in (8.79) is discussed in (Kuo,

Brown 1968, Waroquier 1982).
We have to remember that, due to the use of an effective model space, the
zero-order Hamiltonian also becomes modified into
H(O}
elf = H(O} + ""e:elf
a
L...Jmmm,
+ a (8.80)
m>F

where e:C;; f e:!!;. The energies are mainly determined from fits to nuclei with one
nucleon outside of the closed-shell nucleus under consideration. We call these
nucleons, "dessed" nucleons that interact in the model space via the effective
interaction Velf (see Fig.8.7b).
Using the above method and the self-consistently determined single-particle
orbitals and matrix elements one adjusts X3 such that, after taking into account
the above "renormalization" effects, energy spectra in closed-shell nuclei ±2
particles are well reproduced.
In a first step, we study some general properties of two-body matrix ele-
ments, using the SkE4 force for the (ld S/ 2 )2 configuration (which applies to
18 0). The pure three-body version (X3 = 1) yields repulsive matrix elements

274
 x)= 0.265 Hamada-
after Johnston
[MeV) renormalisation renormalisation

4-
0 2- 4-
4- t;+ 4-
0- t
-1 2-
2- 2-
-2
0- 0- 0-

-3

Fig.8.8. Antisymmetric and normalized neutron two-body matrix element in 18 0. using the SkFA
parameterization «lds/ z )z; J MIV + V'I(1ds/ z)z;J M)oas (Waroquier et aI. 1983b)

(Fig. 8.8). The situation X3 = 0 gives a too wide spectrum compared to the 18 0
data. For a suitable X3 value, a spectrum which is very much comparable to a
realistic Hamada-Johnston (Hamada, Johnston 1962) potential results. A similar
conclusion is obtained after diagonalizing in a full two-particle space and taking
into account renormalization effects (3p - Ih, 4p - 2h). It is now of the utmost
importance to use a constant X3 parameter value throughout the whole mass re-
gion. This can indeed be accomplished using effective (dressed) single-particle
energies instead of the self-consistent HF energies.
Finally, a set of values

X3 = 0.43 SkE2
X3 = 0.265 SkE4

results.
Before discussing some typical examples, using these forces and comparing to
other forces, we point out that all calculations are performed in a self-consistent
way (except that in a number of cases adjusted single-particle energies were
used). We point out, once again the successive steps in the calculations:
i) For given A(N, Z), the spherical HF calculations are carried out determining
€~, c,oa(r),
ii) We calculate the SkE particle-particle "two-body" matrix elements using the
above self-consistent wave functions,
iii) We renormalize the two-body matrix elements,
iv) We construct the energy matrix, diagonalize it in the model space and de-
termine energy spectra, transition probabilities, etc.....

275
 Using the above outline, shell-model calculations have been carried out for
most nuclei with two proton (-neutron) particles (-holes) outside of a doubly-
closed shell configuration i.e. 180, 42Ca, SOea, S8Ni, 134Te, l30Sn (Waroquier
1982, Waroquier et al. 1983b). Besides the typical influence of the X3 parameter
in SkE force, one quite often uses the SkE force with an empirical set of single-
particle energies, denoted by an asterisk (SkE*). An extensive discussion of the
180, 42Ca, and 130Sn spectra is carried out in (Waroquier 1982, Waroquier et al.
1983b). The combined energy spectra are shown in Fig. 8.9, using the SkE*2
force (X3 = 0.43). Only positive parity states are retained. The levels in 42Ca,
drawn with dashed lines have mainly a 4p - 2h character and are outside of the
two-particle model spaces, discussed here. This figure gives a good indication of
the overall agreement for a large number of nuclei throughout the nuclear mass
region.
As a typical illustration, we give the detailed comparison for 134Te where
the two protons move in the 50-82 proton valence model space. The bare SkE
two-body matrix elements are corrected for the 3p - 1h core polarization effects.
Here, we compare the SkE results with results obtained using an effective
Gaussian two-body force. Such an effective force has been used in the study of
the N = 82 nuclei with good results (Waroquier 1970, Waroquier 1971, Heyde
1971). In Fig. 8.10, we compare the pairing two-body matrix elements, using
the bare and renormalized SkE2 force and a Gaussian force (Vo = -39 MeV,
t = +0.2 and t = -0.7) where t denotes the triplet to singlet spin ratio. We
remark that:
- the behaviour is very similar for all orbitals,
- the Gaussian matrix elements, using a value t = -0.7, are remarkably similar
to the SkE2 matrix elements and thus the SkE2 force approximates very
well effective interactions constructed to study these N = 82 nuclei more
particularly.
In Fig. 8.11, we compare the 134Te energy spectra and, apart from a somewhat
too large 0+ -2+ energy separation using the SkE forces, a very detailed similarity
between all interactions (SkE2, SkE4, Gaussian with t = +0.2) results. Moreover,
when comparing, in Table 8.3, the 11g7/22ds/2; JM) matrix elements (1+ ::;
J1r ::; 6+), one observes that the relative energy differences are all very similar
(although different in absolute value). Only the Tabakin interaction gives a too
much compressed spectrum and the Gaussian force with t = -0.7 approximates
best the SkE matrix elements.
As a general conclusion on the study of the SkE force as a particle-particle
interaction it has been pointed out that the fraction parameter X3 could be deter-
mined. This value was fixed at 0.43 and 0.265 for SkE2 and SkE4, respectively.
It was found that X3 tends to a constant value throughout the whole nuclear mass
region. At this stage, all parameters of the SkE force are determined. One might
hope that the forces SkE2 and SkE4 also give correct results in odd-mass nuclei.
We have, moreover, indicated that the SkE forces very much behave as "real-
istic" interactions: specific core-polarization effects renormalize the "bare" matrix

276
 ~
Ex 8 8
(MeV) 7 7
I
6 6

*5 5
SkE2
(x3=O.431 4 4
3++5+ L
3 t4_-
+0+ tNb 3
5+
2 4+ 5+ 2
2+

0+
0+
o 0+ o

t
Ex
8
42
Ca
b) 50
Ca
c) 134
Te
e)
8

(MeV)7 7

I6 6

5 5
Exp.
4 4
0+
5+~:' __ _

3 4+ 3
.2:" __ _
2 0+ 2
2+----- tJ_S+_
r
0+ 0+ 0+
o 0+
o

Fig. 8.9. Two-particle spectra of some (doubly-closed shell + two nucleons) nuclei, using the
SkE2*(X3 = 0.43) force. Only positive parity states are retained. The dashed lines in the 42Ca
spectrum, in the lower pan of the figure, denote mainly 4p - 2h excitations not observed in the theo-
retical 2p shell-model calculations. SkE2* points towards the use of adjusted single-particle energies
(Waroquier et al. 1983b): • (Fortune 1978), b (Void et aI. 1978a, 1978b), C (Bjerregaard et al. 1967,
BrogJia 1968, Tape 1975), d (Bertin et al. 1969, Start 1970), e (Kerek et al. 1972)

277
 " SkE 2 (bare)

t
la+lc /
-<aaolvl ccO>QS (-1)
/ " SkE2 (eft.)
MeV
15 I / GAUSS (t=+O,2)
, / GAUSS (t=-O,7)
/

10 1= 3s1
2
2= 2dl
2
3=2d 5
05 2"
4= 19 I
2
5 =1h 11
"2
1 1 1 1 2 2
2 3 4 5 2 3
Fig_ 8.10. The proton pairing matrix elements in the orbitals for the 50 - 82 N = 4 harmonic
oscillator shell and this for 134 Te. We compare the "bare" and renormalized SkE4 matrix elements
with two-body matrix elements for a Gaussian, effective interaction (using Va = -39MeV and spin
triplet to singlet ratio t = +<l.2 and -0.7) (Waroquier 1982)

Ex
(MeV)
SkE2 SkE4 Guuss u) exp b)
1,):043) 1'1,0265 ) II :01)

l' I'
--=:::--
l'·~' ~),~--

3 f.,'-- t:
~
t'5'
r
6' 6'
"~
§'L- 6'

2
6' 6'
,'---
6' 6'
" " ~,

l' t. t
l'

o 0' 0'

0'
0'

Fig, 8.11. Positive parity states in 134Te. We compare the data (b), taken from (Kerek et al. 1972)
with calculations using an effective Gaussian interaction (a) (Sau et al. 1980) and with self-consistent
calculations using the SkE2 and SkE4 parameterizations (Waroquier 1982)

278
TableS.3. Diagonal two-body matrix elements for the multiplet (l97/z2ds/z) in 134Te

(197/z2ds/z; J MIVIJ97/z2ds/zJ M)aas 1+ 2+ 3+ 4+ 5+ 6+

=0.43)
SkE2 (:Il3 +0.52 +0.12 +0.37 +0.15 +0.38 -0.20
=0.265)
SkFA (:Il3 +0.46 +0.11 +0.37 +0.16 +0.40 -0.20
Tabakin -0.07 +0.05 -0.12 -0.03 -0.11 -0.28
Gauss (t =0.2) -0.20 -0.19 -0.11 -0.21 -0.10 -0.61
Gauss (t =-0.7) +0.70 +0.37 +0.24 +0.04 +0.11 -0.35

elements for use in a highly truncated shell-model space. The 3p-l h corrections,
especially in light nuclei, are very important in order to provide the necessary
long-range quadrupole component for the renormalized SkE forces. It is also
remarkable that such a good agreement with phenomenological interactions, ad-
justed to a particular mass region, results.
It is obvious that for light nuclei, not all low-lying levels in the two-particle
shell-model calculations can be reproduced. There are a number of states that are
mainly outside of the small two-particle model space. Such levels contain large
deformed components and are mainly from a 4p - 2h origin.

8.3.2 The Skyrme Interaction as a Particle-Hole Interaction

a) General Introduction
In the present section, we shall study the above (Sect. 8.3.1) Skyrme force param-
eterizations SkE2 and SkE4 for use in particle-hole calculations in doubly-closed
shell nuclei.
Besides the microscopic picture of a particle-hole interaction, one can also
take a more macroscopic point of view. A linear combination of particle-hole
states in a doubly-closed shell nucleus can be seen as an excitation of the core:
density fluctuations appear in the average field which give rise to a macroscopic
particle-hole interaction.
We expect that the SkE forces, determined in Sect. 8.3.1, will also give rise to
a correct behaviour in the particle-hole channel. To test this, we have studied the
doubly-closed shell nuclei 160, 4OCa, 48Ca, 9OZr, 132Sn, 146Gd and 208Pb in a self-
consistent way. Application to 16 0 as an illustration of particle-hole calculations
was already discussed in detail in Sect. 6.4. Here, we shall concentrate on another
nucleus 40Ca, but first we recall the standard IDA and RPA methods that are
used to calculate the nuclear structure in doubly-closed shell nuclei.
In the IDA approximation (Sect. 6.2), we use a one-particle one-hole basis
relative to the Hartree-Fock ground state to span the basis of the related config-
uration space. Thereby, in an angular momentum coupled scheme, we use the
1p - 1h operators
Iph- I ; JM} == A;h(JMIHF}
= I: {jpmp,jh - mhlJM}

279
 (8.81)

In the RPA approach, a more general ground state (containing particle-hole exci-
tations) is considered for which one defines collective excited states by (Sect. 6.3)

Q~,aIO} = 1!li(Ja)} , (8.82)

and the RPA vacuum state through the condition that

(8.83)

for all (J, a). In this way, each excited state can be well approximated by the
explicit operator
Q~Ma = 2: {Xp;h;(J,a)A;;h;(JM)
i
-Yp;h;(J,a)(-I)J-M Ap;h;(J - M)} (8.84)

The corresponding RPA secular equation becomes

HQ~,aIO} = IiwJ,aQ~,aIO} , (8.85)

or
(8.86)

This problem cannot be solved in general, since one does not know the exact
structure of the RPA ground-state 10}. For a given interaction, however, one could
improve on this by calculating the ground-state admixtures. Thereby, improved
or extended RPA calculations could be carried out (Waroquier 1983c). We only
know that 10} is a vacuum for the Qj a operators.
One can rewrite (8.86) using the commutator expression

(8.87)

with the normalization condition of

(01 [Q J,{h Q~,a ] 10) = Sa{J . (8.88)

With the following approximations:

i) Use the form of (8.84) for the Q~,a operator,
ii) Replace in (8.87) the RPA ground state 10} by the Hartree-Fock ground
state IHF}, one can express the secular equation in (8.87) in a more explicit way

( Q+E R ) (X(J,a»)
R* (Q + E)*Y(J, a)

= IiwJ,a (0I -I0) (X(J,a»)

Y(J,a) , (8.89)

280
where Q and R are the angular momentum coupled representation of the Q and
R matrix elements of Sect. 6.3 and are given by

(8.90)

where H(4) is given in (8.8).

Using the RPA approximation conditions

(Ola;iapj 10) - t 0 , (8.91)

(0 1at, ahj 10) - t 6ii ,
the angular momentum Q and R matrix elements become

Qii = I)2Jt + 1) { ~Pi ~hj

J1 JPj Jhi
1} (_l)ihi+iPj -J 1

x (Pihi; JdV + V'lhiPi; Jt)nas , (8.92)

and
Rii =L (2J1 + 1) { ~Pi Jpj JJ1} (_l)ihi+ipi +J-J1
J1 Jh) Jhi
x (PiPi; J11V + V'lhih i ; J1)nas . (8.93)

The RPA secular equations can be rewritten as

( Q+E
-R*
R
-(Q+E)*
) (X(J,o:)) (X(J,
y(J,o:) =nwJ,o: Y(J,
0:)) ,
0:) (8.94)

This is the standard RPA secular equation for a non-hermitian matrix. If nw J,o:
is an eigenvalue, then -nwj,o: is also an eigenvalue with eigenvectors

( X(J, 0:)) and ( Y*(J, 0:))

Y(J,o:) X*(J,o:) ,

respectively.
Before concentrating on some applications, we discuss shortly (i) the elimi-
nation of the spurious centre-of-mass motion, and (ii) the electromagnetic decay
properties in single-closed shell nuclei.
i) The nuclear A-body Hamiltonian can be separated in a part that describes
the motion of the centre of mass of the A-nucleon system and the A-I internal
coordinates. The Hartree-Fock ground state of a doubly-closed shell nucleus has
the centre-of-mass motion in its ground state. An excited state of the centre-
of-mass motion consists of a linear combination of 1p - 1h excitations and is
admixed with the internal 1p-1h excitations (with the centre-of-mass motion in
its ground state). One has now to project out this spurious centre-of-mass motion

281
which, for the 1P - 1 h space, results in a single 1- state that can be constructed
explicitly.
ii) For the electromagnetic decay properties, we use the expressions as derived
in Chap. 4. Within the RPA, this results into the transition probability (reduced)
from an excited state (J, a) to the 0+ ground state

B(el.; J,a -+ 0;8) = 2/+ 1CJL1 ~[Xpihi(J,a)

I

+ (-l) JYpi hi(J, a)] (PiIlO(el.; L)lIhi }1 2 , (8.95)

°
with the reduced single-particle matrix element for the (el.; L) operator defined
in Sect. 4.3.
Similarly, the more complicated expression between an excited state (J, a)
and (J, (3) can be derived.
We also shortly mention the non-energy weighted (new) and energy-weighted
(ew) sum rule since they give infonnation on the collectivity of certain excited
states.
a) The new sum rule is defined as

Snew(OL) = L B ( OL; O:S -+ J, a)

'"
=CJL L(!PoIOjMOJMI!Po} , (8.96)
M
with

(8.97)

where l!Po} denotes the ground-state. According to the appropriate approximation

this stands for IHF} or IO}.
Within IDA, this leads to

(8.98)

with the IDA fulfilling the "non-energy-weighted" sum rule.

Within the RPA, this gives, on the other hand

S~:(el.; L) =?: I[X;ihi(J,a)+ (-l)JYP~hi(J,a)]

I,'"
x (PiIlO(el.; L)lIhi}1 2 cJL . (8.99)

Here, the RPA does not fulfil the new sum rule. Within an identical model space,
one has

(8.100)

282
 b) The energy-weighted (ew) sum rule is defined as

Sew(el.; L) = L IiwJ,O'B(el. L;O~ - J, a) (8.101)

0'

and can be rewritten as

Sew(el.; L) = ~OJL L(!liol [[OiM' H] ,OJM] l!lio) , (8.102)

M

with l!lio) defined as in (8.96).

Within IDA, only a part of the Hamiltonian H [the part Q + E of (8.89)]
is taken into account. The IDA does not fulfil the ew sum rule. In the RPA,
however, one can prove (Waroquier 1982) that the ew sum rule is fulfilled.

b) Description of Low-lying Collective Vibrations

Having recalled the necessary elements and expressions for testing the quality
of the particle-hole character of the Skyrme forces, we shall now discuss some
applications to the doubly-closed shell nuclei 160, 40Ca. For 160, we refer to
Chap. 6.
We now concentrate on 4°Ca: a nucleus with an analogous structure to 160.
In Figs. 8.12, 13 we present detailed IDA and RPA calculations, using the SkE4*
interaction for the 3-, 5- and 0-,1-,2-,4- levels, respectively.
One observes that the low-lying collective 3- and 5- levels are mainly T = 0
states. For the other states, the percentage of isospin character T is presented in
all cases (Figs. 8.12, 13).
For 11iw 1p - 1h excitations, the hole moves in the N = 2 shells (2s1d
orbitals) and the particle in the N = 3 shell (lf2p orbitals). This restricted
spac~ is denoted by the indices (7-10) in Fig. 8.12. Extending to 21iw 1p - 1h
excitations, we also consider hole states in the N = 1 shell (lp orbitals) and
particle states in the N = 4 shell (lg2d3s orbitals). This space is denoted by
the indices (7-16) in Fig. 8.12. One clearly observes the influence of this model
space increase on excitation energies for 3- and 5- levels. The influence is
negligible, except for the collective 3- and 5- levels. Some states of more
complex character (3p - 3h) do, however, occur at low excitation energy, levels
that cannot be described within the model space chosen here.
Comparing IDA with RPA, the most clear-cut difference is the imaginary
eigenvalue of the lowest 3 -, T = 0 state in RPA. This results for SkE2 and
SkE4 both in the small and larger model space. For all other states, even the
collective 5- level, the TDA-RPA differences are minor. Also, use of the realistic
Tabakin interaction (Dieperink et al. 1968) results in an imaginary root for the 3 - ,
T = 0 level. This deficiency in solving the RPA secular equations is a constant
<:
problem in the description of low-lying (Ex 4 MeV) strongly collective states.
A possible explanation can be found in the neglection of ground-state correlations

283
~ SkE'· ~
.
MeV MeV
12 - unperturbed RPA EXP

I
TO~
'" - 12 1
(7-161 (7-101 (7-101 (7-161

- T-l (79%1 __ T'1

•, T,1 _. T-l (71%1
-
I
\ 4
10 - \• 'nD "
'I
(92%1 IT=1l
-10
T-l (91·'.1 -' T-l
'1• /:',R_ _
'- ',T-J
~ 't!
-- T-l
- ",'e,2f.l2s¥ .,~,
/, 2 - J '\
.~;
-
T-O (79%'._ T-g t I '2pl1d!' ",\T=O T-O (78"10'
8- 2 2 • -8
.'l,.:
""
164%1._ T_l T=1
T-l 1T=1 T.l
(74%1 __ T,O • 169%1

-
r-g n
-
/',IaO _. T-O (82%1 (T-OI
:\r. ~ 2 ,I' (1=01
" ~ T:rl
:' If.! ld~
6r- :1 2 2 -6
"
"
r-
"
:' -
"
~
T-O
'f- T-O (99,4%1,/
• T-O
xxxxxxXXXXX''''l
imaginary
T'O (QQ,8"101
XXXXxxxxxxxxx
imaginary
T=O - 4

r- -
2r- CD -2

SkE4·
TOA RPA EXP
'"
A

(7-16 I (7-10' unperturbed (7-10 I 17-161

- (98"1oL T-l
-
T·l ,
I
,T-l
,
T-l (99"10' T=1
I
,
8- I
,, -8
,,,
I
•\
,
- (P ~
n
\ '
.~
-
1f1 1d!' \
6- , , 2 2 I
-6
,,
I
\
T-O (98"101 T·O , I
•\
- \
\T.O -
..... T-O (99% 1 T=O
-4
4-
C9
Fig. 8.12. The J7r = 3- and 5- states in 4Oea. We compare the IDA and RPA results, both in a
small (lAw excitations with the particle in the lf2p shell) and a large (311"" excitations with the
particle in the 192d3s shell) configuration space. The RPA eigenvalue for the strongly collective
3iT =0 state becomes imaginary (Waroquier 1982)

284
 I-
TOA
SkE4"
unperturbed
,.~.,
: 2p!2S~ :
RPA EXP TOA
SkE4"
RPA EXP
- •I
MeV

12 I- T=I (92"';~)L---.. ~ T-I (92"1., -12

:'"
)f'tP---:~\Y
\1,---
.
II- ,It" 3 r' XiI
"II T=O - I
TaO ISS'll:,
---~-} ~'
2P-2 2 "'"2
'11---
(54"1.,
." I :r' I.
10 I- ~iI 2P2 Id 2 ~ - 10
II I~\ ho (90"1.,
T.nl IS~ll , .. jM" 150"10,
91-~1 :r
t'f.. ..P-___"1 - 9
:: 2Pi td,·' ~
8 1- II \ -8
" ,
71-
~
1 \
\M(98"1o'
y,o - 7

61- ~ -6

51- o - 5

SkE4" SkE4"
TDA RPA EXP TDA RPA EXP
-
unperturbed unperturbed

- 12
~"' ,T=I (70"10,
III- \ ' - I

\,,---
I '
10!- ,..1 (82"'\' i T-I (80%'
I, ;
'
\ n , - 10
T=O (70"lol : - - - - -..., '.0 (70"10,
~~~;l.
" n " T:O (80'7.) .!:.ll....- -~'
91- 1f1251' - - - ,
'p I
I
- 9
2 2 , 2p.i Idi
2 2
81- To! (60"7' T.I (60.,., !:L- - 8
\ n
71- ~:.c:~')..~' - 7
\f2 Idl' '
61- 2 2 -6

51- - 5

Fig.S.13. The J7r =0-,1- ,2- and 4- states in 4OCa. We compare the TDA and RPA calculations,
including 3Aw excitations (Waroquier 1982)

285
Table8.4. The single-particle energies used for the Ip - Ih calculations in 4Oea. We compare with
other sets used in calculations by (Dieperink et aI. 1968) (I). (Hsieh et aI. 1975) (ll) and (Gloeckner.
Lawson 1975) (llI)

SkFA SkFA'
self-consistent adjusted I II III

P n P n P n

11'3/2 -24.72 -25.38 -U,.72 -27.38

IPI/2 -20.98 -2154 -22.98 -23.54
Ids/ 2 -12.48 -12.84 -13.48 -13.84 -13.30 -15.67 -13.31 -13.63
Id3/ 2 -5.95 -6.12 -6.60 -7.10 -7.30 -6.37 -7.27 -7.24
28 1/2 -7.03 -7.34 -9.10 -9.40 -9.80 -9.17 -9.82 -9.91

117/2 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0

I/s/2 +7.74 +8.33 +5.74 +6.33 +6.2 +6.5 +6.5 +6.5
21'3/2 +4.09 +4.67 +2.09 +2.67 +2.0 +2.1 +1.89 +2.18
2PI/2 +5.81 +6.65 +3.81 +4.65 +4.1 +3.9 +3.59 +4.06
199/ 2 +11.76 +12.52 +9.76 +10.52

that are present in the RPA ground-state IO} when deriving the RPA secular
equations. This is corroborated independently by the observation of a very low-
lying 0+ state in 40Ca (Ex = 3.35MeV) which is of a 2p - 2h origin (pairing
vibration).
There are now a number of possibilities to remedy for the above deficiency
in the RPA method:
i) One could enlarge the unperturbed (1 h /2 1d'3N 1p-1 h energy by 1.4 MeV
(Table 8.4). The 3- lowest eigenvalue now becomes real (at Ex = 1.64MeV)
pointing towards a highly critical situation in determining the correct excitation
energy for these low-lying collective states. This is clear in Sect. 6.3 (Figs. 6.8,9)
where the RPA eigenvalues are presented as a function of 1/X. A slight change in
X can render the lowest root real or imaginary. Also, a slight change in the lowest
unperturbed Ip - 1h energy can make this change in character (real, imaginary)
of the lowest root possible.
ii) As a direct consequence of (i), the X3 fraction parameter which determines
the magnitude of the two-body matrix elements can make the above change
in character of the lowest collective 3 - root possible. The precise influence
[in the small model space (7-10)] of X3 is presented in Fig. 8.14. An increase
in X3 (an increase in the importance of the three-body force Wo) lowers the
collective character of the interaction. At a value X3 = 0.45 (SkE4) the 3-,
T =0 eigenvalue becomes real again. One can, however, not generally conclude
that with increasing X3 value, the agreement between theory and experiment
improves substantially.

286
 I,ocal
SkE4"
MeV UNPERTURBED EXP

9 9
r T.l !.!:!...-
8 4-T=1 )"T:l 8

r T.O 7
7 z
r,
ToO
" T:O [Idl
..c:P- 7----,,]_I,...-J··, 4- T=O ]-Y.O
112 ld 1 1\ 2-M
6 1\
II 6
I I 4- T:2
I I
:I \I ... ·-_0
..:~;..:,,:.:-_ _,,"
5 5
\ " s· T.O "" .. " 5-T:0

, I
I
I
I
I
I
]- T:O
,
I
I
I
I
3 : I
3
I
:I
2 I 2
\ _~)"~~_o_____
~ )- T:O 3· T:O /"
xxxxxxxxxxxxxx._ .... .xxxxxJCxxxxxxxx,JuO,...
imaginary imaginary

Fig.8.14. Influence of the fraction parameter :1:3 on the low-lying lp - lh excitations in 40Ca as
obtained from the RPA calculations using the small (lliw) configuration space (Waroquier 1982)

iii) The above two suggestions can be used when calculating the 40Ca energy
spectrum with a phenomenological or schematic interaction for which the strength
is fitted so as to obtain good agreement for a number of nuclei in a specific model
space. Using realistic forces, however, (as SkE2 and SkE4) this freedom does
not exist so that the method of approximation (IDA, RPA, ... ) itself has to be
corrected (Sect 8.3.3).

c) Giant Multipole Resonances

Having studied the low-lying collective excitations (vibrations) and non-collective
excitations in doubly-closed shell nuclei, an extension towards a description of
giant multipole resonances, using the SkE effective interactions, is attempted.
Since we always consider bound-state wave functions, widths cannot be calcu-
lated. However, the problem of location of the dipole resonance energies and
strengths can well be studied. In the light of the results in Sect 8.3.2b, such a
calculation is most interesting.

287
 >.
~

o >
•
~ ~ ~

1\1
0
U
...
0

--- ."
Q. -4 ~
~

....
0 1\1
-
N-
c: , U
;>-, •, ~
~
~
-/
,, "
c
0'"
NI ...
.. --... ---~ ...'\\ ...'" 0...
0 >.
~
5!
0
N

N
II)
c::
~
>
~
.
c:
;;:. ~,.
/ ~

.J--- .
0
N
"0 >
.
----- - ~ ...
C!)

;!
~

..........
-.......... , ~ .t:I
-:::o!:--''---''--'--......... o , a.
\
~
\ 52
M·"Sf(J.,!..~!131 8 \
I
N

o
N

'"
~S/IJ-'S'1b!1318

"'~'----'..........- - - ' - - ' O

0"'.us/U..!!o~318

.........,"'!:;-"--""--""---'-~o 0

M~US/! _I..;!!O ! 1318

:g
",.us/! _12 ~O'1318
Fig. 8.1S. Giant dipole resonances (GDR) in doubly-closed shell-nuclei as obtained from a self-
consistent RPA calculation using the SkEA force parameterization (Waroquier et al. 1983b)

As a first application we show the dipole strength distribution, using SkE4,

in Fig. 8.15 for 160, 40Ca, 48Ca, 9OZr, 132Sn, 146Gd and 208Pb. The excitation
energy is very well reproduced in all cases using the RPA method with the SkE4
interaction, in a self-consistent way. For a detailed discussion of both the giant
dipole resonance and other, higher-multipole resonances in 160_208Pb, we refer
to (Waroquier 1982, Waroquier et al. 1983b).

288
d) General Outlook: SkE As a Particle-Hole Interaction
In the preceding subsections, a detailed study of the adequacy of the SkE force
as a particle-hole interaction has been carried out. It was shown that after having
fully detennined the SkE interaction by fitting the fraction parameter X3 to spectra
of two-particle (two-hole) valence nucleon configurations, good agreement has
been obtained in describing doubly-closed shell nuclei. Many features have been
satisfactorily described. We recall the main results:
i) Description of isospin purity in the collective low-lying vibration J1r = 3-
and 5-, T = 0 states and in the giant multipole resonances in spite of the use of
self-consistent orbitals.
ii) The truncation to a lp - lh model space is mainly sufficient to describe
low-lying negative-parity collective vibrations as well as positive-parity states
(the latter in heavy nuclei only). However, one needs an enlargement of the
model space with 3p - 3h configurations to satisfy the energy-weighted sum
rules (except for the dipole strength, which almost exhausts the EWSR), to de-
scribe higher-lying resonances (~ 40MeV in 160, ~ 20MeV in 208Pb), to re-
e
produce all experimentally observed negative-parity states in light nuclei 6 0,
40Ca ... ). Some low-lying 3p - 3h states exhibit strongly rotational features, and
are predominantly composed by the coupling of the collective octupole with the
quadrupole vibration. The energy position of the multipole resonances is mainly
unaffected by this model-space truncation.
iii) The excitation energy of the giant dipole resonance is fairly well repro-
duced. In light nuclei one observes a separation of the dipole strength in two
states due to the effect of the spin-orbit splitting, while a much more pronounced
fragmentation of the strength is encountered in heavy nuclei, still concentrated
into a restricted energy region (L1Ex ~ 5 MeV). The latter can probably be
attributed to the momentum dependence of the considered Skynne interaction.
iv) In the description of the full level schemes, we have noticed that adjusted
single-particle energies (SkE*) give the better agreement in the model space
truncated to Ip - Ih configurations. However, we remark a better reproduction
of the giant multipole resonances when complete self-consistency is taken into
account (SkE). An explanation of this peculiar behaviour is obtained as follows.
Brown, Dehesa and Speth (Brown et al. 1979) have pointed out that, due to
a dynamical dependence of the effective nucleon mass in nuclear particle-hole
excitations, unperturbed single-particle levels corresponding to an effective mass
m* 1m ~ 0.60 - 0.64 should be employed in order to push the dipole state to a
sufficiently high energy, whereas for low-lying excited states a value m* 1m ~ 1
seems more appropriate. Since the SkE interaction, inside the nuclear interior,
corresponds to low values m* 1m ~ 0.7 (Fig. 8.16), it is quite consistent that
the self-consistent HF single-particle energies give a much better description of
the GDR excitation energies, whereas adjustments (SkE*) need to be introduced
for reproducing low-lying Ip - Ih excitations in doubly-closed shell nuclei,
corresponding more to m* 1m ~ 1. Concluding, a different set of unperturbed
energies is desirable in RPA calculations depending on each J1r state being
calculated.
289
m*r--------------------------------------------------------,
m

1.0 ~-----__l1.0

0.9 0.9

0.8 0.8

0.7 0.7
Sk E2

0.6'--__--L..._ _---L_ _-'--_ _--'-_ _- - - ' ' - -_ _- ' -_ _--'-_ _ _ _-'--_ _- L -_ _- - - '_ _ _ _- ' - -_ _---'0.6
2 3 4 5 6 7 8 9 10 11 r(fm)

Fig. 8.16. The effective m· 1m ratio in 208Pb using the SkE force, for protons (p) and neutrons (n)
separately (Waroquier et al. 1983b)

v) The unnatural-parity states (J1r = 2-,4-) are less satisfactorily described

than the natural-parity levels using SkE. This feature is even more manifest when
using realistic interactions.
vi) A serious defect of the SkE interaction remains the existence of imaginary
solutions in the standard RPA. It should be emphasized that use of the realistic
Tabakin potential also gives rise to an imaginary solution for the low-lying col-
lective octupole state. When using phenomenological residual interactions, the
force strength acts as a "renormalization" parameter, which is adjusted in order
to reproduce excitation energies according to the approximation used (TDA or
RPA). Using the IDA adjusted values for the force parameters in the RPA calcu-
lation, one should also obtain an imaginary solution for the collective Jj = 3} ,
T =0 state. When calculating the particle-hole interaction starting from a realis-
tic potential or from an effective force such as SkE, one has no such additional
parameters to adjust. Thus, it is likely to attribute the defect of imaginary solu-
tions to the neglect of explicitly taking into account ground-state correlations in
the derivation of the standard RPA secular equation. This argument is supported
by studies of Rowe (Rowe 1968, 1970, Rowe et al. 1971). In order to account for
possible 2p - 2h correlations in the physical ground state Iwo}, Rowe deduced
renormalized single-particle energies and densities, which reduce the matrix el-
ements of the standard RPA secular equation. This renormalized RPA method
clearly indicates that the imaginary solution becomes real and that the other so-
lutions remain almost unaltered. Moreover, the thus obtained Jt = 3} excitation
energy approaches the IDA value. Rowe's results support the possibility that

290
the SkE interaction could give rise to a real and satisfactory eigenvalue for the
lowest octupole vibrational excitation, if the explicit ground state J!lio) is consid-
ered, without altering significantly the other higher-lying states and modes. Part
of such a calculation has already been performed (Waroquier et al. 1983c) with
the SkE interaction by evaluating the 2p - 2h correlations in the ground state of
16 0 and 40Ca. The conclusions of this study strongly invalidate the Hartree-Fock
ground state for the physical correlated ground state in solving the equations of
motion, when describing low-lying collective excitations with realistic potentials
or effective interactions, adjusted in self-consistent Hartree-Fock calculations.
The appearance of imaginary solutions can therefore be attributed to a defect of
the standard RPA, rather than to a defect of the SkE interaction.

8.3.3 Rearrangement Effects for Density.Dependent Interactions

and Applications for SkE Forces
In order to obtain correct saturation of the nuclear binding energy and density in
a self-consistent Hartree-Fock calculation, it has been shown (Moszkowski 1970,
Negele 1970, Vautherin, Brink 1972, Campi, Spring 1972, Beiner et al. 1975,
Kohler 1976, Dechargee, Gogny 1980, Waroquier et al. 1983a) that the effective
interaction should contain a density-dependent two-body force or a three-body
term, which, for spin-saturated even-even nuclei behaves like a density-dependent
two-body force. The effect of including a density dependence is most significant
at the nuclear surface, but becomes significantly suppressed in the nuclear interior.
All density-dependent effective nucleon-nucleon interactions used in the literature
exhibit density-dependent parts which bear a striking resemblance to each other
and which are of zero or very short-range nature. When exciting the nuclear
A-body system a subsequent variation in the nucleon density results. Through
the density dependence of the nucleon-nucleon interaction, this implies also a
modification of the average field in which the interacting nucleons move. This
extra term, modifying the average Hartree-Fock field is called the rearrangement
potential energy correction. This extra term will also modify the HFB equations
(8.19,25).
We shall discuss the importance of the rearrangement terms, using the SkE
forces in a number of numerical applications: they drastically effect the collective
states in e.g. doubly-closed shell nuclei.
Since the two-body interaction V shows an explicit density-dependence, the
interaction matrix elements (a,BlV[elh'6)nas will show an extra dependence on
the modifications of the average field through the rearrangement terms implied by
exciting the nuclear system. A detailed discussion how to evaluate any interaction
matrix element occurring in shell-model calculations of an even-even nucleus,
including these rearrangement terms, is discussed by Waroquier et al. (Waroquier
et al. 1987).
When applied to the SkE force, we use an expression with the three-body
component

291
 Wo = X3t3c5(rl - r2)c5(rl - r3) , (8.103)
and a density-dependent two-body force

Vo = (1 - 1
x3)6t3(1 + Pa )c5(rl - r2)e (rl
-2 +r2)
- (8.104)

The advantage of the partitioning in a fraction X3 and (1 - X3) for the three-
body and density-dependent two-body force has been discussed in Sect. 8.2. Here
we present how important the rearrangement effects are on the particle-hole
matrix elements of doubly-closed shell nuclei, and how they affect the collective
properties in an RPA calculation. Due to the particular structure of the above
density-dependent SkE force (1 - X3)Vo, rather simple expressions for the T =
o and T = 1 Q and R matrix elements in the RPA secular equations result
[(8.92-8.94); (Waroquier et al. 1987)]. Using the assumption for the densities
ep = en = !!'tot,
which is not entirely correct, one simplifies the discussion
of rearrangement effects in the RPA considerably. The T = 1 matrix elements
remain unaffected while in the T = 0 channel only natural parity states acquire
an additional repulsive contribution (t3 > 0). This correction is not negligible, as
shown in Fig. 8.17. Here, we depict the Q- and R-matrix elements of the major
particle-hole configurations in 4OCa, as a function of the fraction parameter X3.
The strongly attractive QT~ matrix elements and the repulsive RT~ matrix
elements are reduced considerably, reducing the collective features of the T = 0
states. In Sect. 8.3, we observed a number of problems with the lowest T = 0, 3-
eigenvalue when solving the RPA equations. In Fig. 8.18, we illustrate the effect
the inclusion of rearrangement has on the low-lying T = 0, 3 - and 5- levels. The
collective 3- level now obtains a real eigenvalue and good agreement for 40Ca
is obtained in a self-consistent way (no adjustment of single-particle energies).
One can conclude that the defect attributed to the SkE force parameterization
(Waroquier et al. 1983b) is resolved by taking into account rearrangement effects
properly. Further we stress that excitation energies and strength distributions for
giant multipole resonances are hardly affected so that all conclusions concerning
SkE as given in Sect. 8.3 remain.
It is also so that rearrangement effects, proven to be important in the particle-
hole channel for doubly-closed shell nuclei, have to be taken into account when
discussing the particle-particle aspects of the SkE forces. We therefore apply the
above methods to a discussion of 18 0 and 42Ca (Fig. 8.19). No attempt has been
made to adjust single-particle spectra in an optimal way. Only a slightly smaller
X3 value was taken in the SkE2 parameterization. The higher X3 values lead to
more compressed energy spectra (Sect. 8.3.1). The rearrangement terms reduce
core-polarization corrections induce a serious compression in the two-particle
energy spectra, shown in Fig. 8.19. This effect is counterbalanced by choosing a
somewhat smaller X3 value in the force. This explains why good results could be
obtained in Sect. 8.3.1 in the absence of rearrangement effects. In the particle-
hole RPA no such compensating effects occur and here, rearrangements are really
needed to improve the comparison between experiment and calculations.

292
 Q (1I1h.ld3h.1f7h.ld:Yz.JT) R (1f7/z.ld3/Z.If'¥2.ld3fz.JTI
MeV
2.0 40Ca 40 Ca

......
"- ......
......
1. 1.0 "- ......
...... 73~M
......
X:!:0.43
...... ......
t 0
......
0 x3

-10 ,,
3.0 3.0 "",
""
""
20 2.0 "" ,
"
",
"
10 10 /'"
-5-,1=0 "
,,
x :0.43 ,
"
31
0 X3 00 XJ
0.5

-10 -1

-2.0 --- --- ---

Fig. 8.17. Influence of rearrangement on a particular Q- and R matrix element in 40Ca in the J7r =3-
and 5 - states using the SkE2 parameterization. The dashed line represents the situation without
rearrangement corrections (Waroquier et al. 1987)

We finally would like to report on a very extensive and state-of-the art self-
consistent calculation, using all of the above ingredients, in the single-closed
shell even-even nucleus 116Sn. The model space here is taken to contain rather
complex configurations of 4qp character. We take this nucleus 116Sn since a
detailed set of experiments was performed including 116Sn(p,p'), 116Sn(e, e'),
1l5IneHe,d) and l15In(a, t) (Van der Werf 1986).
In the present calculation [see also (Waroquier et al. 1987) for more details
on the formalism] we consider neutron 2qp, proton Ip - Ih across Z = 50 and
coupled (Ip - h)1r 181 (2qp)v configurations. These calculations are performed

293
 40 Ca
MeV
Fig. 8.18. Low-lying negative
EXP RPA RPA parity (J1r, T) states in 4OCa.
.rearr. without The theoretical levels corre-
rearr.
spond to a fully self-consistent
9 2";1 (97'/.1 2";1197"101 RPA calculation using the SkE2
S,l
T,l
5~1 (99.,.1 - - - 5-1 (100.,.1 (:1:3 =0.43) force. The num-
8 3;1 bers between brackets denote
- - - 3 ' 1 (ao"!.1 ___ 3~1I85"!.1
4~1 the isospin purity (Waroquier et
7 4-,1 (88'/.1 n(88.,.1
---2";0 - - - 2;0 (97.,.1 2·0(97"101 aI. 1987)
4;0(88"1.1 4~0(88"101
6
4:0
5-,0 (99"101
5
---5~0
4 3~ 0 (100"!.1
---3;0
3

2
5;0(100"101
lOClOClOCXlOC 3; 0 1100"!.1
imaginary

18 0 42ea
exp WIth WIthout exp WIth WIthout
rearrangement rearrangement rearrangement rearrangement
Ex Ex
4+ (MeV)
(MeV)
4+ T 8
8

4+ 3+ 0+
7 4+ /s+ 7
2+ 3+
2+ 4+
3+ 4+
6 6
5+ 2+
3+ 2+ 0+
3+ 5+
5 3+ 4+ 3+ 5
4+ 4+
0+ 2+ 6+
2+ 2+
4 2+ 0+ 4+ 4
0+
4+ 4+
6+
3 2+ 6+ 3
4+
4+ 2+
2+
2 2+ 2
2+ 2+

_ _ _ _ 0+ _ _ _ _ 0+ _ _ _ _ 0+ _ _ _ _ 0+ _ _ _ _ 0+
o o
Fig. 8.19. Two-particle spectra in 180 and 42Ca using the SkE2 (:1:3 = 0.37) force. Only positive
parity states are retained. The dominant low-lying 4p - 2h excitations, observed in the experintental
spectra, are not inserted in the figure as they could not be reproduced in a two-particle model space
(Waroquier et aI. 1987)

294
 POSITIVE PARITY STATES NEGATIVE PARITY STATES
MeV
4r-

r-
7+--- 7; _ _ _
+---
8----
4+--_ 10+_____ r
9-=
10:,,- 8+---
- ~+- 4+---
6+---
9----
-
(6+), 6+ 3----
2+'L---
4+---
~t:' 5+---
8-~
r--- 1-,
(2+I W ) 4----
2+--- (7-)--- 6-'---
4+--- 5-
3- (3+)---
1,2+;:;---- 2+---
(2+) _ _ _ 4+ _ _ _ T---
4+--- 3+-- 6----
2+-_ _
2+--- T,>---
(1+)--- 6----
- (0+)---
0+--- 5----
-
4+---
5----
2+---
~:'-=-
1+
3----

0+---
2- -
0+---

EXP TH EXP TH.

Fig. 8.20. Experimental and calculated low-lying positive and negative parity states for 116 Sn. Levels
belonging to the "intruder" band, based on a proton 2p - 2h excitation across the Z = 50 closed
shell, are not included in the figure (Waroquier et al. 1987)

fully self-consistently using SkE forces including also rearrangement effects and
provides a serious test of the above interaction. Only a slight change in the
single-particle energies has been considered: these little changes can be justified
through the missing core-polarization corrections on the level of the one-nucleon
motion in the nucleus (dressed single-particle states).
A detailed evaluation of all necessary matrix elements in the above basis
is presented in (Waroquier et al. 1987). Here, we present some of the salient
features to highlight the above shell-model calculations, using Figs. 8.20,21.
We give in Table 8.5 the proton single-particle and neutron quasi-particle
energies (and occupation probabilities v 2 ) that were used in the determination of
the 116Sn spectrum of Fig. 8.20. Here, low-lying positive and negative parity states
are displayed separately to make comparison easier. The data on the deformed
proton 2p - 2h, which are fully outside of the present model space, have been

295
 a) b) d) e)
MeV 2qp 2qp 1p- lh lp _lh C ) lp-l h lp-lh Full MeV
+
2qp 2qp calculation
2qp 2qp
5 + + + 5
1.00 1.00 098 ITtz'®v) (Ttz:,'®v) (Tt, 0 vi
0.00 000 0.D2 I
0.00 0.00 0.00
9. 0.94

----
'\
3- '\ 0.01 0.92
1
1.00 1.00 \ '\~ 0.01
000 0.00 \ .... ~
0.07
4 0.00 0.00 \ 0.87 4
'\ 0.00 0.86

--
\
'\
'.
0.77
0.79 0.79 "",
'\
---
0.13 0.00
~ 9-1
0.23
0.00 0.20 0.20
0.77
r1
0.01 0.01 0.77
0.18 0.18
3 005 0.05 3

t1 - - - - - -1.00
1.00
-.. ."',
.
2 0.00 0.00 '-0.91', 2
0.00 0.00 0.09 '\
0.00 ,'----
0.89
0.06 0.88 0.87 0.87
0.05 0.06 0.06
0.06 0.06
0.07 0.07

Fig. 8.21. Excitation energies of the J[' and =2t, 3. 9.

levels as a function of the dimension of the
model space: (a) only neutron 2qp excitations in the 50 - 82 shell are included, (b) space of (a) but
including also the 199/ 2 orbital, (c) the coupled configurations are included with the proton Ip - 1h
configurations coupled to J" = 2+, (d) space of (c) but now with J7r = all natural parity spins. The
numbers near each level refer to the wave function square amplitudes of (from top to bottom) 2qp,
1p - Ih and (lp - Ih)" 0 (2qp)v configurations (Waroquier et al. 1987)

Table8.5. Proton single-particle, neutron one quasi-particle energies (in MeV) and occupation prob-
abilities

e(p)

117/2 -5.00
1/5/2 -3.00
2P3/2 -2.00
2P1/2 -1.00
199/2 0.00 5.00 0.99
197/2 4.30 2.15 0.84
2ds/ 2 4.40 2.20 0.92
1hll/2 5.90 2.10 0.12
2d3/ 2 5.20 1.40 0.25
38 1/2 5.40 1.00 0.55

296
left out (Heyde et al. 1983, 1985). There is a general good agreement though a
detailed identification has not been attempted.
It is interesting to discuss in some more detail the influence of the dimension
of the model space chosen on the excitation energy of some of the collective
states (2j, 31) and on the "stretched" 91 level (Fig. 8.21). Here, it is shown
that in general the excitation energy is lowered, proportional to the dimension
of the model space improving the agreement with the data. For the 91 , a shift
of almost 1 MeV occurs in going from a neutron 2qp space to the full space, an
effect which is very important to get a correct description of states in 116Sn.
In calculating spectroscopic strengths for one-proton transfer reactions one
should have a description of the 115In nucleus within a model space, similar to
the one in 116Sn. Such calculations are in progress (Waroquier 1989).

297
9. Some Computer Programs

In this chapter we shortly discuss a set of simple computer programs that allow,
at the level of assimilating the present material, calculations that go beyond
some analytical evaluations or academic questions. Even though this collection
is not very high-brow, the necessary Wigner 3nj symbols are included. We give
a code to evaluate the Slater integrals and the matrix elements of a a-residual
interaction. These ingredients allow for the construction of some simple secular
equations. To this need we include a diagonalization code, based on the Jacobi
method for diagonalizing matrices of small dimensions. We furthermore include,
for the calculation of transition rates the computation of radial integrals of any
power of r(r L ) weighted with the harmonic oscillator wave functions. Finally, a
BCS code, using a constant pairing force is included for finding the occupation
probabilities in a given number of single-particle orbitals.

9.1 Clebsch-Gordan Coefficients

Here, with the function program VN02BA (jt,jz,j, mt, m2, m), we evaluate the
Clebsch-Gordan coefficient

(9.1)

with parameters it, il, j, mt, m2, m in the function to be used having double
the physical value of the angular momentum quantum number. This particular
convention is also used in Sects. 9.2 and 9.3, when evaluating 6j and 9j-symbols.
A small main program is added which includes the input set. In this main
program, a common block NN06FCIFCf(40) is generated which contains the
quantities FCf(I) = (I-l)!/loI, values to be used in the calculation of other
Wigner nj-invariants too. It is the dimension of this array FCf(DIMENSION)
which constrains the angular momenta to be used in evaluating the Clebsch-
Gordan coefficient

298
PROGRAM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . .

FUNCTION VN02BA(J1,J2,J,M1,M2,M)
INTEGER Z,ZMIN,ZMAX,FASE
COMMONjVN06FC/FCT(50)
CC-O.O
IF(M1+M2-M)20,1,20
1 IF(IABS(M1)-IABS(J1»2,2,20
2 IF(IABS(M2)-IABS(J2»3,3,20
3 IF(IABS(M)-IABS(J»4,4,20
4 IF(J-J1-J2)5,5,20
5 IF(J-IABS(J1-J2»20,6,6
6 ZMIN-O
IF(J-J2+M1)7,8,8
7 ZMIN--J+J2-M1
8 IF(J-J1-M2+ZMIN)9,10,10
9 ZMIN--J+J1+M2
10 ZMAX-J1+J2-J
IF(J2+M2-ZMAX)11,12 ,12
11 ZMAX-J2+M2
12 IF(J1-M1-ZMAX)13 ,14,14
13 ZMAX-J1-M1
14 DO 15 Z-ZMIN,ZMAX,2
JA-Z/2+1
JB-(J1+J2-J-Z)/2+1
JC-(J1-M1-Z)/2+1
JD-(J2+M2-Z)/2+1
JE-(J-J2+M1+Z)/2+1
JF-(J-J1-M2+Z)/2+1
FASE-«-1)**(Z/2»
F2-FASE
15 CC-CC+F2/{FCT(JA)*FCT(JB)*FCT(JC)*FCT(JD)*FCT(JE)*FCT(JF»
JA-(J1+J2-J)/2+1
JB-(J1-J2+J)/2+1
JC-(-J1+J2+J)/2+1
JD-(Jl+M1)/2+1
JE-(J1-M1)/2+1
JF-(J2+M2)/2+1
JG-(J2-M2)/2+1
JH-(J+M)/2+1
n-(J-M)/2+1

299
 JJ-(J1+J2+J+2)/2+1
Fl-J+1
CC-SQRT(F1*FCT(JA)*FCT(JB)*FCT(JC)*FCT(JD)
1*FCT(JE)*FCT(JF)*FCT(JG)*FCT(JH)*FCT(J1)/FCT(JJ»*CC
20 VN02BA-CC/SQRT(10.0)
RETURN
END

PROGRAM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . .

COMMONjVNO 6FC/FCT (40)

FCT(1)-1.
DO 400 1-2,40
400 FCT(1)-FCT(1-1)*(I-1.)/10.
4 READ(1,2)J1,J2,J3,M1,M2,M3,1CON
2 FORMAT(712)
X-VN02BA(J1,J2,J3,M1,M2,M3)
WRITE(6,3)J1,J2,J3,M1,M2,M3,X
3 FORMAT(1H ,614,E22.8)
1F(1CON.EQ.O)GOTOS
GOTO 4
5 STOP
END

9.2 Wigner 6j-Symbol

Here, with the function program VN02B9 (jl ,j2, j3, II, /Z, h), we evaluate the
Wigner 6j-symbol

{ jl
11
jz
/z
i3}
lJ '
(9.2)

where again, double the physical value of the angular momentum quantum num-
bers are used.
This function VN02B9 uses the function VN02BB (jl,jz,j3, FCT) in order
to calculate the Wigner 6j-symbol.
As with the calculation of the Clebsch-Gordan coefficient, the array FCT(40)
within the common block NN06FC/FCT(40) has to be used.

300
PROGRAM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. .

REAL*8 FUNCTION VN02B9(J1,J2,J3,L1,L2,L3)

IMPLICIT REAL*8 (A-H,O-Z)
COMHONjVN06FC/FCT(40)
Cc-o.O
IF(J1+J2-J3) 20,1,1
1 IF(IABS( J1-J2)-J3) 2,2,20
2 IF (J1+J2+J3-2*«J1+J2+J3)/2» 20,3,20
3 IF(J1+L2-L3) 20,4,4
4 IF (IABS(J1-L2)-L3) 5,5,20
5 IF (J1+L2+L3-2*«J1+L2+L3)/2» 20,6,20
6 IF(L1+J2-L3) 20,7,7
7 IF(IABS(L1-J2)-L3) 8,8,20
8 IF(L1+J2+L3-2*«L1+J2+L3)/2» 20,9,20
9 IF(L1+L2-J3) 20,10,10
10 IF(IABS(L1-L2)-J3) 11,11,20
11 IF(L1+L2+J3-2*«L1+L2+J3)/2» 20,12,20
12 OMEGA-o. 0
IF(J3) 37,38,37
37 IF(L3) 40,39,40
38 VN02B9-(-1.)**«J1+L2+L3)/2)/SQRT«FLOAT(J1)+1.)*(FLOAT(L2)+1.»
GO TO 41
39 VN02B9-(-1.)**«J1+J2+J3)/2)/SQRT«FLOAT(J1)+1.)*(FLOAT(J2)+1.»
GO TO 41
40 IWMIN-J1+J2+J3
IF(IWMIN-J1-L2-L3) 13,14,14
13 IWMIN-J1+L2+L3
14 IF(IWMIN-L1-J2-L3) 15,16,16
15 I~IN-L1+J2+L3
16 IF(IWMIN-L1-L2-J3) 17,18,18
17 IWMIN-L1+L2tJ3
18 IWMAX-J1+J2+I.l+L2
IF(IWMAX-J2-J3-L2-L3) 22,22,23
23 lWMAX-J2+J3+L2+L3
22 IF (IWMAX-J1-J3-L1-L3) 24,24,25
25 IWMAX-J1+J3+L1+L3
24 IF(IWMIN-IWMAX) 26,26,20
26 DO 701 IW - IWMIN,IWMAX,2
IW1-IW/2+2
IW2-(IW-J1-J2-J3)/2+1

301
 IW3-(IW-J1-L2-L3)/2+1
IW4-(IW-L1-J2-L3)/2+1
IW5-(IW-L1-L2-J3)/2+1
IW6-(J1+J2+L1+L2-IW)/2+1
IW7-(J1+J3+L1+L3-IW)/2+1
IW8-(J2+J3+L2+L3-IW)/2+1
IF(IW-4*(IW/4» 30,31,30
31 PH-l.O
GO TO 35
30 PH--l.O
35 OMEGA-0MEGA+PH*FCT(IW1)/FCT(IW2)/FCT(IW3)/FCT(IW4)/FCT(IW5)
1/FCT(IW6)/FCT(IW7)/FCT(IW8)
701 CONTINUE
CC-0MEGA*VN02BB(J1,J2,J3,FCT)*VN02BB(J1,L2,L3,FCT)*VN02BB(L1,J2,L3
1,FCT)*VN02BB(L1,L2,J3,FCT)
20 VN02B9-CC*10.0
41 RETURN
END
REAL*8 FUNCTION VN02BB(J1,J2,J3,FCT)
IMPLICIT REAL*8 (A-H,O-Z)
DIMENSION FCT(40)
IW1-(J1+J2-J3)/2+1
IW2-(J1-J2+J3)/2+1
IW3-(-J1+J2+J3)/2+1
IW4-(J1+J2+J3+2)/2+1
FDELTA-SQRT(FCT(IW1)*FCT(IW2)*FCT(IW3)/FCT(IW4»
VN02BB-FDELTA/3.16227765
RETURN
END

PROGRAM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..

IMPLICIT REAL*8 (A-H,O-Z)

COMMONjVN06FC/FCT(40)
FCT(1)-1.
DO 400 1-2,40
400 FCT(I)-FCT(I-1)*(I-1.)/10.
4 REAO(1,2)J1,J2,J3,L1,L2,L3,ICON
2 FORMAT(7I2)
X-VN02B9(J1,J2,J3,L1,L2,L3)
WRITE(6,3)J1,J2,J3,L1,L2,L3,X

302
3 FORMAT(1H ,6I4,D22.12)
IF(ICON. EQ. 0) GOT05
GOTO 4
5 STOP
END

9.3 Wigner 9j-Symbol

Below, we give the function program that evaluates the Wigner 9j-symbol,
WINEJ Un,j12,j13,izl,j22,j23,i31,i32,i33) and is used to evaluate

~n ~12 ~13}
{ 321 322 323 . (9.3)
hI jn h3
The input parameters are double the physical value of the corresponding angular
momentum quantum numbers. The evaluation uses a summation over products
of three Wigner 6j coefficients. Again, the array FCT(40) has to be defined in a
main program via the common block, NN06FC/FCT(40).

PROGRAM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . .

FUNCTION WINEJ(J11,J12,J13,J21,J22,J23,J31,J32,J33)
COMMONjVN06FC/FCT(40)
WINEJ-O.
IF(J11+J21-J31)2,1,1
1 IF(IABS(J11-J21)-J31)3,3,2
3 IF(J11+J21+J31-2*«J11+J21+J31)/2»2,4,2
4 IF(J21+J22-J23)2,5,5
5 IF(IABS(J21-J22)-J23)6,6,2
6 IF(J21+J22+J23-2*«J21+J22+J23)/2»2,7,2
7 IF(J31+J32-J33)2,8,8
8 IF(IABS(J31-J32)-J33)9,9,2
9 IF(J31+J32+J33-2*«J31+J32+J33)/2»2,10,2
10 IF(J11+J12-J13)2,11,11
11 IF(IABS(J11-J12)-J13)12,12,2
12 IF(J11+J12+J13-2*«J11+J12+J13)/2»2,13,2
13 IF(J12+J22-J32)2,14,14
14 IF(IABS(J12-J22)-J32)15,15,2
15 IF(J12+J22+J32-2*«J12+J22+J32)/2»2,16,2
16 IF(J13+J23-J33)2,17,17
17 IF(IABS(J13-J23)-J33)18,18,2

303
18 IF(JI3+J23+J33-2*«JI3+J23+J33)/2»2,19,2
19 KMIN-MAXO(IABS(JII-J33),IABS(J32-J21),IABS(J23-JI2»
KMAX-MINO(Jll+J33,J32+J21,J23+JI2)
DO 28 K-KMIN,KMAX,2
28 WINEJ-WINEJ+(-I)**K*(FLOAT(K)+I.)*VN02B9(JII,J21,J31,J32,J33,K)*
IVN02B9(JI2,J22,J32,J21,K,J23)*VN02B9(JI3,J23,J33,K,Jll,J12)
2 RETURN
END

9.4 Calculation of Table of Slater Integrals

In the present program, a list of Slater integrals pO (see (3.79»

FJ = 417r J Unll l (r)Un212(r)Unala(r)Un414 (r)r- 2 dr , (9.4)

is evaluated, Slater integrals that occur for a l5(r2 - rl) residual interaction.
The quantum numbers are introduced as n(i),l(i) with n the radial quantum
number (n = 1,2, ...) and I the orbital angular momentum. Even though in the
actual integral (9.4) the radial quantum number n varies as n = 0, 1,2, ... ; in the
input to the program the more standard notation n' == n + 1 = 1, 2, 3, ... is used.
According to a given maximal number of radial wave functions, given by the
variable IMAX, all Slater integrals are evaluated automatically. The strength of
the interaction is determined by the variable VEFF and the harmonic oscillator
range by the variable NU(== v) with as definition
v= mw/2h, (9.5)

or in more easy-to-handle units

41A- 1/ 3 me2
v= 2(he)2 (9.6)

where me2 is the nucleon rest mass and he the combination 197 MeV fm such
that v has indeed the dimension of fm -2.
The radial integral is evaluated via an analytic~ expression, in closed form,
in the function program SLATER. This function uses a number of coefficients,
calculated in the function VN07 AL, needed to evaluate the Slater integrals.
The expressions used are
FJ =~N
87r nlll N n212 N nala N n414 (2v)-(lt+ 2+ a+ 4)/2-3/2
1 l 1

X 1 00
x(ll+12+la+14)/2+1/2e-2xVnlll(X)Vn212(X)

X Vnala (x )V n4 14 (x )dx , (9.7)

with

304
 _ [2 1- n+2(2v)I+3/2 . (21 + 2n + I)!!] 1/2
(9.8)
Nnl - J7r[(21 + l)!!Fn! '

~ k k n! (21 + I)!! k
v nl(x)=L)-I) ·2 (n-k)!k!(2/+2k+l)!!x , (9.9)
k=O

and

r = N nl' r 1+1 e _IIr Vnl (2 vr2) .

Unl ()
2
(9.10)

Using the integrals

1 00
x P e- 2X dx = 2-(P+l)p!

1 00
xP+l/2 e - 2x dx = 2-(2p+S/2)J7r(2p+ I)!!
(9.11)

nJ
with p a positive number, one finally gets

F" =~ -'v'f2 IJ 2-(·;+1;)1' { ~} [IJ(21, + 2n, + I)!! I'

nl n2 n3 n4 (_l)K . 2- K (2K +2L)!!

x 'L...J
" 'L...J
" 'L...J
" 'L...J
" TI{(n' - k·)"k·"(2/·+2k· + I)"} . (9.12)
k1=O k =O k =O k =O i
2 3 4 ' , .. , . . , , ..

Here we have defined

K=kl+k2+kJ+k4,
L = (II + h + h + 14) + ! (9.13)
2 2'
The numbers between accolades

{~}
mean that one uses Vi or J7r when L is integer or half-integer, respectively.
PROGRAM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. .

PROGRAM SlATMAT
REAL NU
REAL INTI, INT2
DIMENSION N(IO),L(IO)
COMMONjVN07CB/NU,AKNL(4,4,8) ,ANL(4,8) ,INTI(20) ,INT2(20)
C
C SlATMAT MAKES A LIST OF SlATER INTEGRALS

305
C
CALL VN07AL
WRITE(3,360)NU
360 FORMAT(lH ,'NU-' ,FlO.S/)
READ(l,l)IMAX,VEFF
1 FORMAT(I2,F7.2)
WRITE(3,7)IMAX,VEFF
7 FORMAT(lH ,'MAX' ,14, 'VEFF' ,FlO.5/)
DO 2 I-l,IMAX
READ(l,3)N(I),L(I)
WRITE(3,4)N(I),L(I)
3 FORMAT(2I2)
4 FORMAT(lH ,'N' ,14, 'L' ,14)
2 CONTINUE
C
C
C
DO 5 I-l,IMAX
DO 5 J-I,IMAX
DO 5 K-l,IMAX
DO 5 M-K, IMAX
IF(lO*I+J.GT.lO*K+M)GOTO 5
IF«-l)**(L(L)+L(J».NE.(-l)**(L(K)+L(M»)GOTO 5
NA-N(I)-l
NB-N(J)-l
NC-N(K)-l
ND-N(M)-l
LA-L(I)
LB-L(J)
LC-L(K)
LD-L(M)
X=SLATER(NA, l..t.&, NB, LB ,NC, LC ,ND, LD)·,\-VEfF
WRITE(3,6)N(I),LA,N(J),LB,N(K),LC,N(M),LD,X
6 FORMAT(lH, '<' ,413,' lVI' ,413, '>' ,FlO.5)
5 CONTINUE
STOP
END
FUNCTION SLATER(NI,LI,NII,LII,NK,LK,NKK,LKK)
C
C FUNCTION SUBPROGRAM CALCULATES SLATER-INTEGRALS FOR A DELTA FORCE
C

306
C USING FOUR RADIAL HARMONIC OSCILLATOR WAVEFUNCTIONS WITH QUANTUM-
C NUMBERS (NI,LI),(NII,LII),(NK,LK),(NKK,LKK) WHERE N-O,l,2,3.
C
C (NI,LI),(NII,LII), (NK,LK) , (NKK,LKK): THE FOUR SETS OF QUANTUM
C NUMBERS OF THE 4 RADIAL WAVE FUNCTIONS WITH N-O,l,2,3 AND
C L-O,l,2,3,4,5,6,7
C INT: THE SLATER-INTEGRAL (INCLUDING DIVISION BY 4*PI )
C
COMMONjVN07CB/NU,AKNL(4,4,8) ,ANL(4,8),INT1(20) ,INT2(20)
REAL NU
REAL INT1,INT2
ROM-O.
L-LI+LII+LK+LKK
NIl-NI+l
NIIl-NII+l
NKl-NK+l
NKKl-NKK+l
LIl-LI+l
LIIl-LII+l
LKl-LK+l
LKKl-LKK+l
Ll-(L+l)/2
IF(L/2.EQ.L/2.) GO TO 1
DO 2 I-l,NIl
DO 2 II-l,NIIl
DO 2 K-l,NKl
DO 2 KK-l,NKKl
L2-Ll+I+II+K+KK-4
2 ROM-ROM+AKNL(I,NI1,LI1)*AKNL(II,NII1,LII1)*AKNL(K,NK1,LK1)*AKNL(KK
l,NKK1,LKK1)*INT1(L2)
GO TO 3
1 DO 4 I-l,NIl
DO 4 II-l,NIIl
DO 4 K-l,NKl
DO 4 KK-l,NKKl
L2-L/2+I+II+K+KK-3
4 ROM-ROM+AKNL(I,NI1,LI1)*AKNL(II,NII1,LII1)*AKNL(K,NK1,LK1)*AKNL(KK
1 ,NKKl,LKK1)*INT2(L2)
3 SOM-ROM
SLATER-ANL(NI1,LI1)*ANL(NII1,LII1)*ANL(NK1,LK1)*ANL(NKK1,LKK1)*SOM
1*(2*NU)**1.5/78.95683523

307
 RETURN
END
SUBROUTINE VN07AL
C
C SUBROUTINE THAT CALCULATES A NUMBER OF COEFFICIENTS NECESSARY WHEN
C CALCULATING THE SLATER INTEGRALS
C
C FAKl(N),N-l,4: FAKl(N)-(N-l)!
C FAK2(N),N-l,ll: FAK2(N)-(2N-l)!!
C ANL(N,L),N-l,4, L-l,8:
C ANL(N,L)-SQRT(2**(L-N+2)*(2*NU)**(L+l/2)*(2L+2N-3)!!)/SQRT(SQRT(PI)
C *«2L-l)!!)**2*(N-l)!)
C AKNL(K,N,L),K-l,4, N-l,4, L-l,8:
C AKNL(K,N,L)-(-2)**(K-l)*(N-l)!*(2L-l)!!/(K-l)!*(N-K)!*(2L+2K-3)!!)
C INTl(I) ,1-1,20: INTl(I)-I!/2**(I+l)
C INT2(I) ,1-1,20: INT2(I)-SQRT(PI)*(2I-l)!!/2**(2I+l/2)
C NU-M*OMEGA/2*H-BAAR
C
REAL NU
REAL INTl, INT2
COMMONjVN07CB/NU,AKNL(4,4,8) ,ANL(4,8) ,INTl(20) ,INT2(20)
DIMENSION FAKl(21),FAK2(20)
READ(l,l) NU
1 FORMAT(FlO.8)
FAKl(l)-l.
DO 2 IR-l,20
2 FAKl(IR+l)-FAKl(IR)*IR
FAK2(1)-1.
DO 3 IR-l,19
3 F~_~2(IR+l)-FAK2(IR)*(2*IR+l)

DO 4 N-l,4
DO 4 L-l,8
LNS-L+N
FI-FAK2(LNS-l)/(FAK2(L)**2*FAKl(N»
ANL(N,L)-SQRT(2.**(L-N+2.)*Fl)
DO 4 K-l,N
KNS-N-K
LKS-L+K
4 AKNL(K,N,L)-(-2)**(K-l)*FAKl(N)*FAK2(L)/(FAKl(K)*FAKl(KNS+l)*FAK2(
ILKS-I»
DO 5 1-1,20

308
 INT1(I)-FAK1(I+1)/2**(I+1)
5 INT2(I)-1.25331413731550*FAK2(I)/4.**I
RETURN
END

$ KRIS
$ ASSIGN SYS$INPUT FOR001
$ ASSIGN SYS$OUTPUT FOR003
$ R SIATMAT. EXE
0.1
10 100.00
1 0
1 1
2 0
1 2
2 1
1 3
3 0
2 2
1 4
1 5

9.S Calculation of 6-Matrix Element

Making use of the above program SLATER, the two-body matrix elements using
a residual interaction of the type

(9.14)

are calculated. We evaluate proton-neutron matrix elements as

(nplpjp)(nn1njn); J MIVI(n~l~j~)(n~l~j~); J M}
=JnJ~JpJ~(-I)ip+i~+lp+l~(2J + 1)-1 (jp!, jn - !IJO}(j~!,j~ - !IJO)
x [! { 1 - (_I)in+i~+lp+l~ S(jnjp, J)S(j~j~, J)(4J(J + 1))-1 }
-a [1 + (_I)'n+ 'p +J ]] V(eft)pO , (9.15)
with
S(jpjn, J)=(2jp + 1) + (-I)ip+i n+J (2jn + 1) . (9.16)
Here, we used the coupling order I + ! =i. Note that the order of coupling the
angular momenta is important.

309
 The choice of other quantities is discussed in Sect. 9.4 where the evaluation
of Slater integrals has been discussed.
A typical example of an input file is given where we have the Ih9/2 proton
and li13/2 neutron orbital as configurations that are used in order to detenmne
the matrix elements

(9.17)

Remark: The code can even be used to calculate two-body matrix elements for
a b"-interaction in the specific case of (j2; J MIVIi,2; J M) matrix elements. In
that case, the variable IDENT has to be chosen equal to 1.

PROGRAM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..
PROGRAM DELT
REAL NU. NORMI. NORM2
REAL INTI. INT2
INTEGER PAR
COMMON/SHELL/E(20).N(20).L(20).J(20).IT(20).NUM(20)
COMMON/PARAM/ALFA.VEFF.IDENT
COMMONjVN06FC/FCT(40)
COMMONjVN07CB/NU.AKNL(4.4.8).ANL(4.8).INTl(20).INT2(20)
C
C MATRIX ELEMENTS IN ORDER <J(P)J(N).JJIVIJ(P·)J(N').JJ>
C (L+S)J IN THAT ORDER
C R(N.L) POSITIVE AT THE ORIGIN
C N-l.2.3 •.... AS INPUT (N-0.l,2 in code)
C JA-2*J DOUBLE PHYSICAL VALUE
C LA SINGLE PHYSICAL VALUE
C V--VEFF*(l-ALPHA+ALPHA*SIGMA(P)*SIGMA(N»*DELTA FUNCTION
C NU-(M*OMEGA)/(2.*HBAR)
C
C GOOD ESTIMATES ARE VEFF-400;ALPHA-0.2
C IDENT-l IDENTICAL PARTICLES
C
CALL VN07AL
WRITE(3,360)NU
360 FORMAT(lH .'NU-'.FIO.8/)
READ(l,400)ALFA.VEFF.IDENT
400 FORMAT(2F8.4,I2)
WRITE(3,365)ALFA,VEFF
365 FORMAT(lH .'ALFA-'.F8.4.'VEFF-',F8.4)
FCT(l)-l ...

310
 DO 50 1-2,40
FCT(I)-FCT(I-l)*(I-l.)/lO.
50 CONTINUE
READ(l,3l)IMAX
31 FORMAT(I2)
DO 38 I-l,IMAX
READ(l,32)E(I),N(I),L(I),J(I),IT(I),NUM(I)
32 FORMAT(F5.3,5I4)
WRITE(3,33)E(I),N(I),L(I),J(I),IT(I),WJM(I)
33 FORMAT(lH ,'E-',F5.3,' NLJ-',314,'/2 IT-',I2,'NUM-',I4)
38 CONTINUE
60 READ(1,34)Il,I2,I3,I4,JP,ICON
34 FORMAT(6I2)
X-DELTA(Il,I2,I3,I4,JP)
WRITE(3,35)X
35 FORMAT(lH ,'MATRIX EL.-' ,E20.8)
IF(ICON.EQ.O)GOTO 60
STOP
END
FUNCTION DELTA(Il,I2,I3,I4,JP)
COMMON/PARAM/ALFA,VEFF,IDENT
COMMON/SHELL/E(20),N(20) ,L(20),J(20),IT(20) ,NUM(20)
REAL NORM1, NORM2
NA-N(Il)-l
NB-N(I2)-1
NC-N(I3)-l
ND-N(I4)-1
A1-0.S
IF(IDENT.EQ.l)Al-l.
IF(JP.EQ.O.AND.IDENT.NE.l)GOTO 106
IF(JP.GE.l.AND.IDENT.NE.l)GOTO 101
106 ZL-O.
GOTO 102
101 ZL-S(J(I2),J(Il),JP)*S(J(I4),J(I3),JP)/(4.*JP*(JP+1.»
102 XL-SQRT«J(Il)+1.)*(J(I2)+1.)*(J(I3)+1.)*(J(I4)+1.»*(-l)**«J(Il)
1+J(I3»/2+L(Il)+L(I3»/(2.*JP+1.)*
2VN02BA(J(I2),J(Il),2*JP,l,-l,O)*
3VN02BA(J(I4),J(I3),2*JP,l,-l,O)*
4(Al*(1.-(-1)**«J(I2)+J(I4»/2+L(Il)+L(I3»*ZL)-
5ALFA*(1.+{-1)**(L(Il)+L(I2)+JP»)*VEFF*
6SLATER(NA,L(Il),NB,L(I2),NC,L(I3),ND,L(I4»

311
 NORH1-1.
NORH2-1.
IF(IDENT.NE.1)GOTO 104
IF(I1.EQ.I2)NORM1-1./SQRT(2.)
IF(I3.EQ.I4)NORH2-1./SQRT(2.)
104 XL-XL*NORH1*NORM2
DELTA-XL
RETURN
END
FUNCTION S(IA.IB.IC)
C HULPFUNKTIE
S-IA+1+(IB+1)*«-1)**«IA+IB)/2+IC»
RETURN
END
FUNCTION VN02BA(J1.J2.J.M1.M2.M)
COMMON/VN06FC/FCT(40)
INTEGER Z.ZMIN.ZMAX.FASE
C VN02BA DENOTES THE CLEBSCH-GORDAN COEFFICIENT
C <J1.M1.J2.M2 I J1.J2.J.M>
C
CC-O.
IF(M1+M2-M)20.1.20
1 IF(IABS(M1)-IABS(J1»2.2.20
2 IF(IABS(M2)-IABS(J2»3.3.20
3 IF(IABS(M)-IABS(J»4.4.20
4 IF(J-J1-J2)5.5.20
5 IF(J-IABS(J1-J2»20.6.6
6 ZMIN-O
IF(J-J2+M1)7.8.8
7 ZMIN--J+J2-M1
8 IF(J-J1-M2+ZMIN)9.10.10
9 ZMIN--J+J1+M2
10 ZMAX-J1+J2-J
IF(J2+M2-ZMAX)11.12.12
11 ZMAX-J2+M2
t2 IF(J1-M1-ZMAX)13.14.14
13 ZMAX-J1-M1
14 DO 15 Z-ZMIN.ZMAX.2
JA-Z/2+1
JB-(J1+J2-J-Z)/2+1
JC-(J1-M1-Z)/2+1

312
 JD-(J2+M2-Z)/2+1
JE-(J-J2+Ml+Z)/2+1
JF-(J-JI-M2+Z)/2+1
FASE-«-I)**(Z/2»
F2-FASE
15 CC-CC+F2/(FCT(JA)*FCT(JB)*FCT(JC)*FCT(JD)*FCT(JE)*
lFCT(JF»
JA-(Jl+J2-J)/2+1
JB-(JI-J2+J)/2+1
JC-(-Jl+J2+J)/2+1
JD-(Jl+Ml)/2+1
JE-(JI-Ml)/2+1
JF-(J2+M2)/2+1
JG-(J2-M2)/2+1
JH-(J+M)/2+1
JI-(J-M)/2+1
JJ-(Jl+J2+J+2)/2+1
Fl-.1+1
CC-SQRT(Fl*FCT(JA)*FCT(JB)*FCT(JC)*FCT(JD)*FCT(JE)*
IFCT(JF)*FCT(JG)*FCT(JH)*FCT(JI)/FCT(JJ»*CC
20 VN02BA-CC /SQRT(10.)
RETURN
END
FUNCTION SLATER(NI,LI,NII,LII,NK,LK,NKK,LKK)
C
C FUNCTION SUBPROGRAM CALCULATES SLATER INTEGRALS FOR A DELTA FORCE
C
C USING FOUR RADIAL HARMONIC OSCILLATRO WAVEFUNCTIONS WITH QUANTUM
C NUMBERS (NI,LI) ,(NII,LII), (NK,LK),(NKK,LKK) WHERE N-O,I,2,3.
C
C (NI,LI),(NII,LII),(NK,LK), (NKK,LKK): THE FOUR SETS OF QUANTUM
C NUMBERS OF THE 4 RADIAL WAVEFUNCTIONS WITH N-O,I,2,3 AND
C L-O,I,2,3,4,5,6,7
C INT: THE SLATER-INTEGRAL (DIVISION BY 4*PI INCLUDED)
C
COMMONjVN07CB/NU,AKNL(4,4,8) ,ANL(4,8),INTl(20) ,INT2(20)
REAL NU
REAL INTl,INT2
ROM-O.
L-LI+LII+LK+LKK
NIl-NI+1

313
 NIIl-NII+l
NKI-NK+l
NKKI-NKK+l
LIl-LI+l
LIIl-LII+l
LKI-LK+l
LKKI-LKK+l
Ll-(L+l)/2
IF(L/2.EQ.L/2.) GO TO 1
DO 2 I-l,NIl
DO 2 II-l,NIIl
DO 2 K-l,NKl
DO 2 KK-l,NKKl
L2-Ll+I+II+K+KK-4
2 ROM-ROM+AKNL(I,NIl,LIl)*AKNL(II,NIIl,LIIl)*AKNL(K,NKl,LKl)*AKNL(KK
1,NKKl,LKKl)*INTl(L2)
GO TO 3
1 DO 4 I-l,NIl
DO 4 II-l,NIIl
DO 4 K-l,NKl
DO 4 KK-l, NKKl
L2-L/2+I+II+K+KK-3
4 ROM-ROM+AKNL(I,NIl,LIl)*AKNL(II,NIIl,LIIl)*AKNL(K,NKl,LKl)*AKNL(KK
1,NKKl,LKKl)*INT2(L2)
3 SOM-ROM
SLATER-ANL(NIl,LIl)*ANL(NIIl,LIIl)*ANL(NKl,LKl)*ANL(NKKl,LKKl)*SOM
1*(2*NU)**1.5/78.95683523
RETURN
END
SUBROUTINE VN07AL
C
C SUBROUTINE CALCULATES A NUMBER OF COEFFICIENTS NECESSARY FOR THE CAL
C CULATIONS OF THE SLATER INTEGRALS
C

C FAKl(N),N-l,4: FAKl(N)-(N-l)!
C FAK2(N),N-l,11: FAK2(N)-(2N-l)!!
C ANL(N,L),N-l,4,.L-l,8:
C ANL(N,L)-SQRT(2**(L-N+2)*(2*NU)**(L+l/2)*(2L+2N-3)!!)/SQRT(SQRT(PI)
C *«2L-l)! !)**2*(N-l)!)
C AKNL(K,N,L),K-l,4, N-l,4, L-l,8:
C AKNL(K,N,L)-(-2)**(K-l)*(N-l)!*(2L-l)!!/(K-l)!*(N-K)!*(2L+2K-3)!!)

314
C INT1(I) ,1-1,20: INT1(I)-I!/2**(I+1) .
C INT2(I) ,1-1,20: INT2(I)-SQRT(PI)*(2I-1)!!/2**(2I+1/2)
G NU-H*OMEGA/2*H-BAAR
C
REAL NU
INT1,INT2
REAL
COMMON/VN07CBINU,AKNL(4,4,8) ,ANL(4,8) ,INT1(20) ,INT2(20)
DIMENSION FAK1(21),FAK2(20)
READ(l,l) NU
1 FORMAT(FlO.8)
FAK1(1)-1.
DO 2 IR-1,20
2 FAK1(IR+1)-FAK1(IR)*IR
FAK2(1)-1.
DO 3 IR-1,19
3 FAK2(IR+1)-FAK2(IR)*(2*IR+1)
DO 4 N-1,4
DO 4 L-1,8
LNS-L+N
F1-FAK2(LNS-1)/(FAK2(L)**2*FAK1(N»
ANL(N,L)-SQRT(2.**(L-N+2.)*F1)
DO 4 K-1,N
KNS-N-K
LKS-L+K
4 AKNL(K,N,L)-(-2)**(K-1)*FAK1(N)*FAK2(L)/(FAK1(K)*FAK1(KNS+1)*FAK2(
1LKS-1»
DO 5 1-1,20
INT1(I)-FAK1(I+1)/2**(I+1)
5 INT2(I)-1.25331413731550*FAK2(I)/4.**I
RETURN
END

$ ASSIGN SYS$INPUT FOR001

$ ASSIGN SYS$OUTPUT FOR003
R DELTA.EXE
0.083
0.5000300.0000
2
0.000 1 5 9 1 1
0.000 1 6 13 o 2
1 2 1 2 2 0

315
 1 2 1 2 3 0
1 2 1 240
1 2 1 2 5 0
1 2 1 2 6 0
1 2 1 2 7 0
1 2 1 280
1 2 1 2 9 0
1 2 1 210 0
1 2 1 211 1

9.6 Matrix Diagonalization

Here, we include a short program to diagonalize real symmetric matrices using the
Jacobi method. In the present option, the dimensions have been put at 20 x 20.
These numbers can be easily enlarged. However, since the Jacobi method is
mainly used for small matrices, it is advised not to enlarge dimensions beyond
100 x 100.
Besides the subroutine VN0108 (N), we include a main program that intro-
duces the matrix A to be diagonalized in lower diagonal form (see input example).
In the output all eigenvalues and corresponding eigenvectors are provided.

PROGRAM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . .
PROGRAM DIAGOT
C TEST OF DIAGONALISATION OF EIGENVALUES AND
C EIGENVECTORS OF SMALL SYMMETRIC REAL MATRICES
C UP TO DIMENSION 20
COMMONjVNOIB2/A(20,20),IDIM,S(20,20)
READ(I,I)IDIM
1 FORMAT(I2)
WRITE(3,2)IDIM
2 FORMAT(IH " DIMENSION OF MATRIX-',I4/)
DO 5 I-l,IDIH
READ(I,3)(A(I,J),J-l,I)
3 FORMAT(10F8.5/10F8.5)
WRITE(3,4)(A(I,J),J-l,I)
4 FORMAT(IH ,'MATRIX',10F8.5/10F8.5)
5 CONTINUE
DO 10 I-l,IDIM
DO 10 J-l,I
A(J, I)-A(I ,J)
316
10 CONTINUE
CALL VNOI08(IOIM)
DO 20 I-l,IOIM
WRITE(3,6)A(I,I)
WRITE(3,7)(S(J,I),J-l,IOIM)
7 FORMAT(lH ,'EIGENVECTOR'/10F9.5/10F9.5)
6 FORMAT(lH ,'EIGENVALUE' ,F9.5)
20 CONTINUE
STOP
END

$ASSIGN SYS$INPUT FOROOI

$ASSIGN SYS$INPUT FOR003
R DIAGOT.EXE
10
3.00000
0.05000 2.80119
0.00000 0.00000 2.80830
0.00000 0.00000 0.00000 2.98050
0.00000 0.00000 0.00000 0.00000 2.98561
0.00000 0.00000 0.00000 0.00000 0.00000 2.99318
0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 2.07925
0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 2.09859
0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 2.12036
0.00000 0.00000 0.00000 0.00000 U.UOUOO 0.00000 0.00000 0.00000 0.00000 2.59965

PROGRAM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. .
SUBROUTINE VNOI08(N)
C VNOI08 OIAGONALISES SQUARE MATRICES UP TO THE ORDER OF 20*20
C
COMMONjVNOIB2/A(20,20),IOIM,S(20,20)
IF(N-l)21,21,22
22 DO 1 I-l,N
S(I,I)-I.
Il-I+l
IF(1l.GT.N)GOTO 60
DO 1 J-Il,N
S(1,J)-O.
1 S(J,1)-S(1,J)
60 NUL-O
NBMAX-I000
EPS-I.OE-3

317
 IN-3
EF-10.
DO 2 I-l,IN
EPS-EPS/EF
DO 2 NB-NUL,NBKAX
IF(NB)4,5,4
4 IF(IDR+IDI) 2 ,2,5
5 IDR-o
IDI-o
DO 33 II-2,N
JI-II-l
DO 33 JJ-l,JI
C-A(II,JJ)+A(JJ,II)
D-A(II,II)-A(JJ,JJ)
IF(ABS(C)-EPS)6,7,7
6 CC-l.
SS-O.
GOTO 8
7 CC-D/C
IF(CC)10,9,9
9 SIG-l.
GOTO 12
10 SIG--l.
12 COT-CC+SIG*SQRT(l.+CC*CC)
SS-SIG/SQRT(l.+COT*COT)
CC-SS*COT
IDR-IDR+1
8 E-A(II,JJ)-A(JJ,II)
CH-l.
SH-o.
IF(ABS(E)-EPS)14,13,13
13 CO-CC*CC-SS*SS
SI-2.*SS*CC
H-O.
G-o.
HJ-o.
DO 15 K-l,N
IF(K-II)16,15,16
16 IF(K-JJ)18,15,18
18 H-H+A(II,K)*A(JJ,K)-A(K,II)*A(K,JJ)
Sl-A(II,K)*A(II,K)+A(K,JJ)*A(K,JJ)
S2-A(JJ,K)*A(JJ,K)+A(K,II)*A(K,II)

318
 G-G+S1+S2
HJ-HJ+S1-S2
15 CONTINUE
D-D*CO+C*SI
H-2.*H*CO-HJ*SI
F-(2.*E*D-H)/(4.*(E*E+D*D)+2.*G)
IF(ABS(F)-EPS)14,17,17
17 CH-1./SQRT(1.-F*F)
SH-F*CH
IDI-IDI+1
14 C1-CH*CC-SH*SS
C2-CH*CC+SH*SS
Sl-CH*SS+SR*CC
S2-SH*CC-CH*SS
IF(ABS(Sl)+ABS(S2»20,33,20
20 DO 30 L-1,N
A1-A(L,II)
A2-A(L,JJ)
A(L,II)-C2*A1-S2*A2
A(L,JJ)-C1*A2-S1*A1
A1-S(L,II)
A2-S(L,JJ)
S(L,II)-C2*A1-S2*A2
30 S(L,JJ)-C1*A2-S1*A1
DO 31 L-1,N
A1-A(II,L)
A2-A(JJ,L)
A(II,L)-C1*A1+S1*A2
31 A(JJ,L)-C2*A2+S2*A1
33 CONTINUE
2 CONTINUE
TAUSQ-O.
N1-N-1
DO 150 I-1,N1
II-I+1
DO 150 J-II,N
150 TAUSQ-TAUSQ+(A(I,J)+A(J,I»**2
RETURN
21 S(l,l)-l.
RETURN
END

319
9.7 Radial Integrals Using Harmonic Oscillator
Wave Functions

Here, we evaluate the radial integrals, using the hannonic oscillator wave func-
tions, as discussed in (9.10) of this chapter. The integral, however, is evaluated
using dimensionless variables.
The function that evaluates these integrals is
VNOI05(Nl, Ll, LAMB, N2, L2)
where Nl, Ll (N2, L2) denote the radial (starting at Nl = 0, 1, ...) and the orbital
quantum number, respectively.
The integral becomes

J ,,~
Un I (x)x unl(x)dx -
-[ r(n + l)r(n' + 1)
r( n+ 1 +t-T )r(' n + 1 +t-T ')
] 1/2 I 'I
T.T.

X L r(t+O'+ 1)
O'!(n - O')!(n' - O')!(O' + T - n)!(O' + T' - n')!
, (9.18)
tT

with
T = !(l' - I + oX)
T' = !(l - I' + oX)

t = !(I + I' + oX + 1) . (9.19)

The sum on 0' is restricted via the condition

n}> >{n-T (9.20)

n' - 0' - n' - T'

Some integrals are

{n, 1- llrln, I} = (n + 1+ !)1/2,

{n-l,l+llrln,l} =n1/ 2,
(9.21)
{n + 1, l-llrln,l} = (n+ 1)1/2,
{n,l+llrln,l} =(n+l+~)1/2,

{n, llr2ln, I} =2n + I + ~ = N +~,

{n - 1, llr2ln, I} =(n(n + I + !»1/2,
{n, 1- 21r21n, I} =«n + 1+ !)(n + I _ !»1/2, (9.22)
{n + 1,1- 21r21n, I} =(n(n _ 1»1/2,
{n + 2, 1 - 21r21n, I} =«n + 1)(n + 2»1/2.

320
PROGRAM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..
FUNCTION VN0105(Nl,Ll,LAMB,N2,L2)
C VN0105(Nl,Ll,LAMB,N2,L2) -(Nl,LlfRHO**LAMB/N2,L2)
C WHERE IN, L) IS THE RADIAL PART OF THE WAVE FUNCTION
C ASSOCIATED WITH THE SYMMETRICAL HARHONICAL
OSCILLATOR POTENTIAL AND
WHERE RHO IS THE REDUCED R-COORDINATE
1 NUL-O
IF(N2-Nl)2,3,3
2 NA-Nl
Nl-N2
N2-NA
NA-Ll
Ll-L2
L2-NA
3 S-O.
NNl-2*Nl
NN2-2*N2
DO 9 K-l,Nl,l
KK-2*K
Tl-VN0107 (NN1)/VN0107 (KK)
Ml-2*(Nl-K+l)
M2-2*(N2-Nl+K)
T2-VN0107(NN2)/(VN0107(Ml)*VN0107(M2»
M3-2*(Nl-K+l)+Ll+L2+LAMB+l
T3-VN0107(M3)
T4-1.
T5-1.
IF(K-l)31,5,31
31 AL-Ll-L2-LAMB
KA-K-2
DO 4 I-NUL,KA,l
T4-T4*(AL/2.+FLOAT(I»
4 CONTINUE
5 IF«K.EQ.l).AND.(Nl.EQ.N2»GOTO 7
BL-L2-Ll-LAMB
KB-K-2+N2-Nl
DO 6 J-NUL,KB,l
T5-T5*(BL/2.+FLOAT(J»
6 CONTINUa
7 S-S+(Tl*T2*T3*T4*T5)*10.**(FLOAT(LAMB)/2.

321
 1+FLOAT(-2*K+NI-N2+2»
9 CONTINUE
11 N5-2*(Nl+Ll)+1
NG-2*(N2+L2)+1
10 SQ-VNOI07 (NNl)*VN0107(NN2)*VN0107 (N5)*VNOI07 (NG)
SQI-SQRT(SQ)
12 VNOI05-(S*FLOAT«-1)**(Nl+N2»)/SQl
RETURN
END
FUNCTION VNOI07(N)
C VNOI07(N)-GAMMA(N/2)/(10**(N/2-1»
C WHERE GAMMA IS THE WELL KNOWN GAMMA-FUNCTION AND
C WHERE N IS AN INTEGER
C FCT(M)-(M-l)!/(10**(M-l»
C WHERE M IS A POSITIVE INTEGER
COMMON/VNOGFC/FCT(40)
10 1-0
1 K-N/2
IF(K*2-N)3,2,12
2 IF(N)15,15,21
21 GA-FCT(K)
IF(I)90,90,13
3 K-N/2
IF(K)31,G,31
31 Kl-2*K
GA-(FCT(Kl)*SQRT(31.4159»/(FCT(K)*2.**(2*K-l»
IF(I)90,90,13
6 GA-SQRT(31.4159)
IF(I) 90,90,13
12 1-1
N-N*(-1)+2
GOTO 3
13 G-GA
N-N*(-1)+2
FA-(-1)**«l-N)/2)
GA-(FA*31.4159)/G
GOTO 90
90 VNOI07-GA
15 RETUHN
END

322
9.8 BeS Equations with Constant Pairing Strength

In the present, simple BCS code, we solve the two coupled algebraic equations,

L [(ej - >i + Ll2 ] -1/2 = ~ ,

j
(9.23)
~ (j + ~) [1 - [(ej _ e~)-;+>'Ll2]1/2 ] = n ,
J

where G and n are input values, G the pairing strength and n the number of
identical valence nucleons outside of the closed-shell configurations. For G, we
can use estimates as discussed in Sect. 7.5.1.
Here, in (9.23), the sum on j goes over all single-particle orbitals.
From solving the BCS equations which is done in the program BCS one gets
>. and Ll as output values and subsequently, one easily derives v;,
and uJ
and
Ej.
In solving for the BCS equations, a typical input set is given with
i) STREN, NUMl, !MAX
STREN: the pairing strength G
NUMl: the number of valence nucleons n
!MAX: the number of single-particle orbitals considered.
ii) J(I), E(I)
We give as input 2 x j, e j for the different single-particle orbitals with j the
angular momentum and e j the single-particle energy corresponding to given
J.
iii) FER, DEL
We have to give as input an estimate of the Fermi level energy >. and of the
gap Ll. If, incidentally, the increments ti>. and tiLl are such that the solution
diverges (i.e. the new value of Ll becomes negative), a new set of starting
values for >., Ll has to be used as an input.
The present example starts from the set

2h/2' Ih9/2' li13/2' 3[>3/2,215/2 and 3Pl/2 ,

respectively, with energies -0.075,0.599, 1.392,3.156,3.530 and 4.700 (MeV),
respectively.

323
PROGRAM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . .
DIMENSION E(20),J(20) ,VKW(20) ,EE(20)
300 READ(l,l)STREN,NUM1,IMAX,IWRITE,ICON1,ICON2
1 FORMAT(F6.3,5I2)
WRITE(6,51)STREN,NUM1,IMAX
51 FORMAT(lH ,'STRENGTH-',F6.3,'NUMBER',I4,'LEVELS',I3/)
IF(ICON1.EQ.1)GOTO 200
DO 3 I-1,IMAX
READ(l,4)J(I),E(I)
WRITE(6,5)J(I),E(I)
4 FORMAT(I2,F6.3)
5 FORMAT(lH ,12,'/2 E -',F6.3/)
3 CONTINUE
200 READ(l,6)FER,DEL
6 FORMAT(2F6.3)
KT-O
FX-O.
FY-o.
50 Fl--2.0/STREN
F2-(-1)*FLOAT(NUM1)
FlI.-O.
F2I.-0.
F1D-0.
F2D-0.
DO 7 I-1,IMAX
Zl-«E(I)-FER)**2+DEL**2)**0.5
Z2-Z1*«E(I)-FER)**2+DEL**2)
F1D-F1D-DEL*(J(I)+1.)/2./Z2
F1I.-F1L+(E(I)-FER)*(J(I)+1.)/2./Z2
F2D-F2D+(E(I)-FER)*DEL*(J(I)+1.)/2./Z2
F2I.-F2L+DEL**2*(J(I)+1.)/2./Z2
FI-F1+(J(I)+1.)/2/Z1
F2-F2+(1.-(E(I)-FER)/Zl)*(J(I)+1.)/2
IF(IWRITE.EQ.O)GOTO 7
WRITE(6,52)Zl,Z2,FID,FIL,F2D,F2L,F1,F2
52 FORMAT(lH ,8E16.6)
7 CONTINUE
AI-FIL*F2D-FID*F2L
A2-F2*FID-Fl*F2D
A3-Fl*F2L-F2*FlL
DLAM-A2/Al

324
 DDEL-A3/A1
FER-FER+DLAM
DEL-DEL+DDEL
IF(IWRITE.EQ.O)GOTO 11
WRITE(6,53)A1,A2,A3,DLAM,DDEL,FER,DEL,KT
53 FORMAT(lH ,7E16.6,I4/)
11 PREC-0.001
IF(ABS(F1).LT.PREC.AND.ABS(F2).LT.PREC)GOTO 40
KT-KT+1
FVER1-ABS(F'X-F1)
FVER2-ABS(FY-F2)
F'X-F1
FY-F2
IF(IWRITE.EQ.O)GOTO 13
WRITE(6,54)FVER1,F'X,F1,FVER2,FY,F2
54 FORMAT(lH ,6E16.6/)
13 IF(DEL.LT.O)GOTO 200
GOTO 50
C CALCULATE VKW AND QP ENERGY FROM FER AND DEL
C

40 WRITE(6,56)FER,DEL
56 FORMAT(lH ,'FERMI LEVEL' ,F8.5,'GAP ',F8.5111)
DO 15 I-1,IMAX
ZA-«E(I)-FER)**2+DEL**2)**0.5
ZB-E(I)-FER
VKW(I)-0.5*(1.-ZB/ZA)
EE(I)-ZA
WRITE(6,55)VKW(I),EE(I)
55 FORMAT(lH ,'VKW-',F8.5,'lQP ENERGY-' ,F8.5,I2,'/2 'II)
15 CONTINUE
IF(ICON2.EQ.1.AND.ICON1.EQ.1)GOTO 80
IF(ICON2.EQ.0)GOTO 300
80 STOP
END

$ ASSIGN SYS$INPUT FOR001

$ ASSIGN SYS$OUTPUT FOR006
$ R BCS
0.195 1 6 1
7-0.075
9 0.599

325
 13 1. 392
3 3.156
5 3.530
1 4.700
1.000 1.000
1.000 0.200

326
Appendix

A. The Angular Momentum Operator in Spherical Coordinates

The transfonnation from cartesian to spherical coordinates for the angular mo-
mentum operator components (Lx, Ly and L z) can be evaluated in detail. Starting
from Chap. 1, Fig. 1.1, one has the relations

r =(x 2 + l + z2)1/2 ,
I{J = tg- 1 ! ,
x (A.l)
_ (x2 + y2)1/2
9=tg 1 --2-
Z

The derivatives of r, I{J and 9 towards the cartesian coordinates x, y, z become,

respectively
8r . {)
8x = sm u . cos I{J ,

8r . {) .
-8y = smu· Slnl{J , (A.2)

8r
- =cos9
8z '
81{J 1 sin I{J
8x = - ; sin 9 '
81{J 1 cos I{J
(A.3)
8y =; sin 9 '
81{J = 0
8z '
and
89 1
-=-cos9·cosl{J
8x r '
89 1 {).
-=-cosu·Slnl{J (A.4)
8y r '
89=_!sin9.
8z r

327
Now, we calculate

L
Z
=0: (x~ -y~)
ayax' (A.5)

or, written out in detail

. ()
L z=o: { rsm [. (). 1 (). a
1 coscp
coscp sm smcpar +;cos smCPa()+;' sin()
a]
acpa

-
. (). [. () a 1 () a 1 sincp a ]} , (A.6)
rsm smcp sm coscp ar +;cos coscp a() -; sin() acp
or

Lz =0: (r sin2 () cos cp sin cp :r + sin () cos () sin cp cos cp !

+ cos cp
2 a - r sm. 2() cos cp sm• cp ara
acp
.() ().
- sm cos sm cp cos cp a()a. a)
+ sm cp acp ,
2
(A.7)
or
a
Lz = 0: acp . (A.S)

Completely similar calculations can be carried out to determine the other com-
ponents Ly and L z • Subsequently, one easily determines the value of L2 ==
L; +L; +L;.
B. Explicit Calculation of the Transformation Coefficients
for Three-Angular Momentum Systems

An explicit calculation of the transformation (1.115) for a system of three inde-

pendent angular momenta goes as follows.
First, we uncouple the coupled basis states Iii (hh)J23; J M) explicitly via
the expression
Ijl(hh)J23; JM} = L (hm2,hm3IJ23 M23)

x (jlml, J23 M23IJM) Ijlml} Ihm2} U3 m3} . (B.l)

Coupling now the angular momenta (iIh)J12 and then (JI2h)J in order to
recover the transformation (1.115) one gets

Ijl(hj3)J23;JM) = L
ml,m2,mS I
(hm2,hm3IJ23M23)
M12. M 23. J 12

x (jlml, J23M23IJM)(iIml,hm2IJI2MI2}
x (JI2MI2,hm3IJM}I(jlh)JI2h; JM) . (B.2)

328
In the latter expression (B.2), we can also sum over the magnetic quantum number
M, and since the transformation coefficients are independent of M, one divides
by the factor (2J + 1).
Rewriting (B.2) in terms of the Wigner 3j-symbols we get the following
expression for the transformation (1.115) (we make use of the notation 3-
(2j + 1)1/2 throughout the appendices, as well as in the text)

Ijl(hh)J23; JM} = L{ L
J12 all magn.
(2J + l)-lJ12 J23 JJ
qn

X (_I)it-h- M12(_I)J12-ia- M
X (_I)i2-ia- M2a(_I)it- J2a-M

i2 h
X ( jl
ml m2
1t2
-M12
) ( J12
M12 m3 -~)
x(h m2
)3
m3
J23 ) (
-M23
jl
ml
J23
M23 -~) }
X l(jli2)JI2h; J M} . (B.3)

Now, making use of the quantity between curly brackets, where the sum over
all magnetic quantum numbers over a product of four Wigner 3j-symbols is
observed, and (1.117), one finally gets the result (1.115) back as

Ijl(j2h)J23; J M} =L(_I)it+h+ia+J JI2 J23

jl
X { h
J12} l(j1)2.)J12)3;
hJ J23 . JM}
. (B.4)

Thereby, we have proved the transformation relation and obtain the result in
terms of a Wigner 6j-symbol.

C. Tensor Reduction Formulae for Tensor Products

Here, we give an explicit derivation of the reduction formula (2.104). The method
used is rather general and can also be used for deriving other reduction formulae.
We start from the standard matrix element to which we apply the Wigner-
Eckart theorem

M =(adl' a2i2; J MI~k) laUi, a~j~; J'M'}.

=(_I)J-M (J k J,)
-M K M'
X (adt, a2h; JIIT k)lIaUL a~j£; J'} . (C.l)

329
We can now work out the left-hand part of (C.l) by uncoupling the wave functions
for the combined system and also uncoupling the tensor operator T£k) into its
uncoupled form. One gets
M = 2)jlml,hm2IJM)(j~mLj~m2IJI M')
X {kl 11:1 , k211:2IkK){alitmll~~dlaU~mD
x (a2hm2IT~~2)la~j~m~) . (C.2)

If we apply in (C.2) the Wigner-Eckart theorem again, one obtains for M, the
expression, in short-hand
kl
M= "{... I... X... I ... ){ ... I ... )(_l)i1 - ml ( jl j{ )
L..J -ml 11:1 m~

x (_l)i2- m 2 (h -m2
k2
11:2
mj~2') {atilWr<kd ll a UD
x (a2hIlTk2)lIa~j~) . (C.3)

In (C.2) and (C.3), the sum goes over the magnetic quantum numbers {ml' m2,
m~, m2, 11:1, 1I:2}. We rewrite the above expressions using the Wigner-3j-symbols
and make use of the orthogonality relations for 3j-symbols to bring the 3j-
symbol in the right-hand side of (C.l) so as to calculate the reduced matrix
element itself. We so obtain
{a til ,a2jz; JIITk)lIaUL a2j~; J' }

= L
all
(;1 /:2 -~) (~~ ~, 2
J' ) (kl
-MI 11:1
k2
11:2
k)
-II:

(
mi

x (jl kl jt) ( iz k2
11:2
j~)
m2
J
-M
k J' )
M'
-ml 11:1 m~ -m2 II:

x jj 1 k(_l)it- m l(_l)i2- m 2
X (-l)it -i2- M(-l)i; -i;-M' (_l)k 1-k2-K( _l)J-M
x {atilI1T(kdllaU~){a2izIlT(k2)lIa2i~} , (C.4)

_ "
- L..J
(jl
ml
h
m2 -M
J) (
m~
jt j~
m~ -M'
JI) (kl
11:1
k2
11:2
k ) jjlk
-II:
all mi

XC!~1 :: ~~) (}~2 :~ ~2) (-~

JI
M'
x (-lrase{atill1rkdllaUD{a2hIlT(k2)lIa2j~) , (C.5)
where
Phase =jl - h + j~ - j~ + kl - k2 + J + jl + jz + j~ + j~ + J'
+ kl + k2 + k + J + J' + k - M - M' - II: - M - ml - m2. (C.6)
Since we know how the Wigner 9j-symbol is defined in terms of a sum over all
magnetic quantum numbers mi, we use the definition

330
 r J21
J31
i12
J22 in
.
h2 J33
j,,} = L (
all m

X (
Jll
mn

i31
m31
J12 i13) ( hI
m12 m13
in h3 ) ( in
mn m33
m21

mll
J22
m22

hI
m21
in )
mn

i31 )
m31
X ( i12 i22 in ) ( it3 In i33) . (C.7)
m12 m22 m32 m13 mn m33

So, from (C.6), one finally obtains

. JIIT k ) IIalJl,
,., a2J2;
,., J')

r
( . a2J2;
alJl,
J2.,
=ii'A: i{ 32 { } (",jlllT"') lI"bl){mh liT"') II";';;)· (C.S)
kl k2

D. The Surface-Delta Interaction (SOn

The SOl interaction has some very interesting properties which makes it a quite
"realistic" interaction (paessler 1967) although, at first glance, it looks like any
effective interaction:

i) Interaction Between Nucleons at the Nuclear Surface. The local kinetic

energy T for a nucleon within the nuclear potential U(r) is much larger in the
central region compared to the nuclear surface region (see Fig. 0.1), since for a
nucleon moving at the nuclear fenni surface, the total energy E is E =T+U(r),
a constant value. The nucleon scattering process is mainly a 1 = 0 wave (S-
wave) process of nucleons that interact in a "frontal" way. The relative kinetic
energy for nucleon-nucleon scattering in the nuclear interior is now much larger
compared to the nuclear surface region. It appears that, for S-wave scattering,
e
the phase shift for 1 So scattering S +l l.7) increases with decreasing collision
energy. From the Born approximation theory to nucleon-nucleon scattering, one
has (Fetter, Walecka 1971)

tgc5 = - ~; 1 V(r)i~(kr)r2dr
00
, (0.1)

with
h,2 k2
E=-
m
the energy available in the relative system at which the phase parameters have to
be evaluated. The c5(1 So) phase is illustrated in Fig. 0.2 and illustrates indeed the
importance of scattering at the lower energy (nuclear surface region) compared
to the higher energy (nuclear interior) scattering process.

331
 --r -
10

O~------------------------------~~--~

"'"'-SURFACE --

6 -20
a::
OJ
z
OJ -30 ::: 45 MeV

-40

05Ro Ro
RADIAL DISTANCE

Fig. D.I. A Woods-Saxon potential is used to illustrate the relative kinetic energy. The well depth is
~ 50 MeV. The diffuseness parameter is ao = 0.64 fm and the radius parameter is 1.25 fm. The mass
number is A = 185. The local kinetic energy of a particle near the Fermi surface is about 45 MeV.
At the nuclear surface, the kinetic energy is appreciably lower (taken from (Faessler, Plastino 1967»

VI
OJ
OJ
a::
0
OJ
40
a

...
10

LL
J:
VI 20
OJ
VI
«
::z::
0..

O~--------~----------~--~~-----+--~

100 200 300

ENERGY [MeV]

Fig. D.2. The singlet 1= 0 wave (ISO) phase shift for proton-proton scattering (in a qualitative way)
as a function of the laboratory bombarding energy

Of course the Born-integrals are not yet the relative integrals of the nuclear
relative two-body interaction V(r) (with the nucleon-nucleon relative coordinate)
since in the integral the j/(kr) spherical Bessel functions show up and not the
relative harmonic oscillator wave functions.

332
ii) Relation Between Spherical Bessel Functions and Harmonic Oscillator
Functions (Abramowitz 1964). One can write the harmonic oscillator functions
as
1 n2
R () = [2 - + (21 + 2 n + 1)" ..v 3/2] 1/2 -vr2/2( 2)1/2
nl r ?r l / 2 • n![(21 + 1)!!]2 e vr

X1Ft (-n;1+~;vr2) , (D.2)

with 1Ft (ex, f3, x) the confluent, hypergeometric series and v = mw /2fi. There
now exists a relation (for small r-values and large n(n -+ 00» that

1Fl (-n; I + ~; vr2) = r (1 + ~) e vr2 / 2

X (:21( 2n + 1 + :2
3) vr2
)-1/2-1/4

X
2)1/2 ( 2 ( 2n + 1+:2
( -; 3) vr2
)1/4

X jl«vv,)1/2 r ) (D.3)
_ (21 + I)!! vr2/2 (2 1 ~)-1/2
- 21/2 e n+ +2
x (vr 2)-1/2jl«VV,)1/2 r ) . (D.4)

This now gives the relation

(21 + I)!! .
Rnl(r) = Nnl (2(2n + 1+ 3/2»1/2 J1 (kr) , (D.S)

with k2 == vv'.
This relation follows through identification of the relative energy, which is,
fi 2 k2 /m and also 1i.w(2n + I + 3/2), such that, since v = mw/2fi, one has that

-;;
2
k =2 ( 2n + I + 3)
:2 = v' . (D.6)

The relation (D.S) has been tested in the mass region A ';;;f 140. We illustrate
it for Rls(r), R2s and jo(kr) (Fig. D.3), and, indeed, for small r-values (clearly
up to the nuclear radius Ro = 1.2 A l / 3 fm) this relation (D.5) is well fulfilled.
Agreement increases with increasing n, and for equal n with increasing I.
The reduced integrals which appear in the calculation of Sect 3.2.3d, when
we couple to the relative spin wave function 1(1/21/2); SMs) lead to

(nl, S); .JTIV(r)l(nl, S); .JT) = 1 ~1(r)V(r)r2dr,

00
or (D.7)

333
 ~~---------------------------------------------------,

-.,

·.21----------------------------------------------------~

Fig. D.3. Comparison of the radial hannonic oscillator wave functions R1a(r). R2.(r) with the related
spherical Bessel function jo(kr). The precise relation between both functions is expressed by the
relation (D.S). The ordinate is in units fm- 3/ 2

2'-n+I(21 + 2n + 1)!!1iw· km
(nl, 5); .JTIV(r)l(nl, 5); .JT} = 1'(1/2. n!(v /)I+I/2Ji2

X 1 V(r)j~(kr)r2dr
00
. (0.8)

Using the Born-approximation, relating the integral in (0.8) to the phases de-
scribing scattering in the (I, S);.JT channel, one can finally write the radial
integrals describing nuclear structure as
(nl, 5); .JTIV(r)l(nl, 5); .JT}
2'-n+I(21 + 2n + 1)!!1iw
=- 1'(1/2 . n!(v/)I+I/2 tg6(E) . (0.9)

From (0.6) (giving v'), the energy for scattering E = Ji2 k2 1m (the energy that is
present in the relative nucleon-nucleon motion) and Iiw =41 A -I /3 MeV, one can
now determine the relative integrals over the nuclear, relative 2-body interaction
V(r). One can use the tabulated values of Arndt, Wright and McGregor (see
Table 0.1) for 6(E) (McGregor 1968a).

iii) A Delta-Force Expression. The free nucleon-nucleon interaction indeed has

a short range. Even though the detailed form of the nucleon-nucleon interaction in
the nuclear medium probably deviates from the free nucleon-nucleon interaction,
the short-range property remains a major characteristic. This can be expressed
most easily using a 8-function force 6(rl - r2).

334
 SINGLE·PARTICLE RADIAL DENSITIES

1
harmonic OSCillator
0.4

0.2

...
-..
J
E
C
.;. 0
.;;
I
:i
11
..1~/
I '.\
\
\
Saxon·WOOdS

C : I \

t
0
GI
:. i

02

---
2 4 6 8 10
~rln fm
Fig. D.4. The probability density r2 R!,(r) for a neutron in the Id, 28, If or 2p single-panicle
state calculated for the nucleus 29Si. The plots illustrate that for both the hannonic oscillator and
Woods-Saxon potential, the probabilities near the nuclear surface r = Ro do not depend strongly
(state-independence) on n nor I (taken from (Brussaard, Glaudemans 1977»

iv) State Independence of R",(r =

~). If one calculates the probability
of finding a nucleon between the distance r, r + dr, one gets the expression
41rr2 dr R;,(r). For r < Ro, this function is strongly dependent on n and 1. For
r 9t Ro, however, this probability is almost constant (see Fig. D.4), and thus the
radial integral, using the surface delta force 6(rl - r2)6(rl - Ro), simplifies to

(D. 10)

and can even be approximated by a constant (non state-dependent) quantity A

(Brussaard, Glaudemans 1977).

E. Multipole Expansion of 6(rl - r2) Interaction

In obtaining a multipole expansion of the delta effective interaction 6(rl - r2),

we start from the integral representation

335
 TableD.t. PlJase.shift analysis (with error bars) for a laboratory energy interval 5MeV ~ Ellb ~
4OOMeV. The phase shifts are given in degrees (taken from (Mac Gregor et al. 1968a»

Lab
energy
(MeV) ISO ID2 IG4 'Po '1\ '1'2 "2
S S4.6S ± 0.03 0.06 ± 0.00 0.00 ± 0.00 1.77±0.02 -1.09±0.01 0.29±0.01 -0.06±0.00
10 S4.97±0.07 0.20 ± 0.00 O.OO±O.OO 3.83 ±0.04 -2.32 ± 0.01 0.80±0.02 -0.23±0.00
IS S3.01±0.09 0.38±0.00 O.ot ±O.OO S.61 ±0.07 -3.41 ±0.02 1.41 ± 0.03 -0.44 ± 0.00
20 SO.7S ± 0.11 0.S7±0.01 0.03 ±O.OO 7.09±0.10 -4.36 ± 0.03 2.07 ± 0.04 -0.66±0.01
25 48.S1 ± 0.11 0.77±0.01 O.OS±O.OO 8.28±0.12 -S.2O ± 0.04 2.7S ± 0.04 -0.87±0.01
30 46.36±0.1l 0.9B±0.01 0.07 ± 0.00 9.23±0.14 -S.9S±0.OS 3.43±0.OS -1.0B ± O.ot
40 42.37±0.12 1.3B ± 0.02 0.12±0.00 10.S4±0.IB -7.28±0.OS 4.76±0.OS -1.4S ±0.02
SO 3B.7B ± 0.13 1.77 ±0.02 0.17±0.00 11.25 ± 0.20 -B.4S ±0.06 6.02±0.OS -1.76±0.02
60 3S.SS±0.14 2.1S±0.03 0.23±0.00 I1.S1 ± 0.22 -9.51 ±0.06 7.19 ± O.OS -2.03±0.03
70 32.62±0.16 2.S3±0.04 0.29±0.00 11.4S ± 0.23 -10.51 ±0.07 8.27 ± O.OS -2.25 ± 0.03
BO 29.94±0.IB 2.B9±0.04 0.3S±0.00 11.13 ± 0.23 -11.47 ± 0.07 9.26±0.OS -2.43 ± 0.04
90 27.48±0.20 3.25±0.OS 0.41 ±0.01 10.62±0.23 -12.39 ± 0.07 10.17 ± O.OS -2.S7±0.04
100 25.21 ±0.21 3.60±0.OS 0.47 ±0.01 9.97±0.23 -13.29 ± 0.07 10.99 ± O.OS -2.68 ± 0.04
120 21.0B±0.23 4.27±0.06 0.S9±0.01 B.36±0.23 -IS.02 ± 0.07 12.41 ±0.06 -2.S3±0.04
140 17.3S±0.25 4.91 ±0.07 0.71 ±0.02 6.49 ± 0.24 -16.70 ± 0.09 13.SB±0.06 -2.91±0.04
160 13.96±0.27 S.S2±0.OB 0.B2±0.03 4.50 ± 0.26 -IS.33 ± 0.11 14.S3 ±0.07 -2.91 ±O.OS
IBO 10.71 ±0.29 6.10 ± 0.10 0.93±0.03 2.44 ± 0.30 -19.91 ±0.13 IS.30± O.OB -2.B7 ± 0.06
200 7.SS ± 0.31 6.66±0.11 1.04±0.04 0.3B ± 0.34 -21.46 ± 0.16 15.91 ±0.09 -2.79±0.OB
220 4.51 ±0.34 7.19 ± 0.12 US ±O.OS -1.6S ± 0.40 -22.96 ± 0.20 16.39 ± 0.10 -2.68 ± 0.09
240 1.46±0.3B 7.69 ±0.14 1.25 ±0.06 -3.64±0.46 -24.43 ± 0.24 16.77 ± 0.12 -2.5S ± 0.12
260 -1.57 ± 0.43 8.17±0.16 1.3S ±0.07 -S.S7±0.S3 -2S.86 ± 0.27 17.04±O.IS -2.39 ± 0.14
280 -4.62±0.SO 8.63 ±0.17 1.4S ±0.08 -7.43 ± 0.60 -27.26 ± 0.30 17.24± 0.18 -2.23 ± 0.17
300 -7.67 ± 0.59 9.07±0.19 I.SS ±0.09 -9.22 ± 0.69 -28.62 ± 0.34 17.37 ± 0.21 -2.0S±0.20
320 -10.7S ± 0.71 9.49±0.21 1.6S ±O.IO -10.93±0.76 -29.94 ± 0.37 17.44 ± 0.24 -1.86±0.23
340 -13.8S ± 0.84 9.89±0.23 1.74 ± 0.11 -12.S7 ± 0.83 -31.23±0.4O 17.46±0.27 -1.67 ± 0.26
360 -16.97 ± 0.99 10.28 ±0.25 1.83 ± 0.12 -14.12 ± 0.90 -32.49 ± 0.43 17.44±0.31 -1.47 ±0.29
380 -20.11 ± US 10.64±0.27 1.92 ± 0.13 -IS.61 ± 0.97 -33.72 ± 0.46 17.38 ±0.3S -1.27± 0.32
400 -23.27 ± 1.33 11.00 ± 0.29 2.01 ± 0.14 -17.02± 1.04 -34.91 ± 0.49 17.28 ± 0.39 -1.07 ± O.3S

(E.I)

(E.2)

We use the plane-wave expansions

(E.3)
',m
and
(E.4)
",m'
Combining these results, one obtains

o(rt - r2) = (2~)3 ,L, j(167r2)i'+" (-I)"j,(krt)j,,( kr2)

I,l,m,m

X Y, (rt)Y" m'· (r2)Y, m· (k)Y" m''''

m A A A 2
(k)k dk dk .
A
(E.5)

336
 3Fz 3F3 3F4 £4 3H4 3H5 'H6
O.OO±O.OO -0.01±0.00 0.00 ± 0.00 0.00 ± 0.00 O.OO±O.OO O.OO±O.OO 0.00 ± 0.00
0.01 ±O.OO -0.04±0.00 O.OO±O.OO O.OO±O.OO O.OO±O.OO O.OO±O.OO 0.00 ± 0.00
0.04±0.00 -0.10 ± 0.00 0.01±0.00 -0.02±0.00 O.OO±O.OO O.OO±O.OO 0.00 ± 0.00
0.07±0.00 -0.17±0.00 0.01±0.00 -0.03±0.00 0.00 ± 0.00 -0.01 ± 0.00 0.00 ± 0.00
0.10 ± 0.00 -0.26±0.00 0.02±0.00 -0.06±0.00 0.00 ± 0.00 -0.02±0.00 O.OO±O.OO
0.14 ± 0.00 -0.35 ± 0.00 0.04 ± 0.00 -O.OS±O.OO 0.01 ±O.OO -0.03 ± 0.00 O.OO±O.OO
0.22±0.00 -O.SS±O.OO O.OS ± 0.00 -0.14 ± 0.00 0.02 ± 0.00 -0.06±0.00 0.00 ± 0.00
0.30 ± 0.01 -0.7S±0.01 0.13±0.00 -0.21 ±O.OO 0.03±0.00 -0.10±0.00 0.01 ±O.OO
0.37 ±O.o1 -0.94±0.01 O.lS±O.01 -0.28 ± 0.00 0.04 ± 0.00 -0.14±0.00 0.01 ±O.OO
0.43 ±0.02 -1.12± 0.02 0.25 ±0.01 -0.35 ± 0.00 0.06±0.00 -0.19±0.00 0.02 ± 0.00
0.4S±0.02 -1.28 ± 0.02 0.32±0.01 -0.42±0.01 O.OS±O.OO -0.24±0.01 0.03 ± 0.00
0.S3±0.03 -1.44±0.03 0.40±0.01 -0.49±0.01 0.10 ±O.o1 -0.29±0.00 0.04 ± 0.00
0.S6±0.04 -l.S8± 0.04 0.48 ± 0.02 -O.SS±O.Ol 0.12 ± 0.01 -0.3S±0.01 O.OS±O.OO
0.61 ±0.06 -l.S3± 0.06 0.65 ± 0.03 -0.66 ± 0.01 0.16 ± 0.01 -0.46±0.02 O.OS±o.ol
O.64±O.OS -2.04±0.OS 0.83 ± 0.03 -0.77±0.02 0.20 ± 0.02 -0.57 ± 0.04 0.11 ±0.01
0.64±0.10 -2.22±0.11 1.01 ±0.04 -0.86 ± 0.03 0.25±0.03 -0.6S ± 0.05 0.15 ±0.01
0.62±0.13 -2.37 ± 0.13 1.20±0.OS -0.93±0.04 0.29±0.04 -0.7S ± 0.07 0.19 ± 0.02
0.S9±0.lS -2.S0±0.lS 1.39 ±0.06 -1.00 ± 0.05 0.34±0.OS -O.SS±O.09 0.23±0.03
0.55 ± 0.17 -2.61 ±O.lS loSS ± 0.07 -1.06 ± 0.06 0.3S±0.07 -0.9S ± 0.11 0.27±0.03
0.49±0.20 -2.71±0.20 1.77 ± O.OS -1.11 ±0.07 0.42 ± 0.08 -LOS ± 0.14 0.32±0.04
0.43 ±0.22 -2.79±0.23 1.96±O.09 -1.15 ±0.08 0.46±0.10 -1.17±0.16 0.36 ± 0.05
0.37±0.24 -2.8S±0.25 2.14±0.10 -1.19 ± 0.09 0.SO±0.12 -1.26±0.19 0.41 ±0.06
0.30±0.27 -2.91 ± 0.27 2.32±0.11 -1.22±0.10 0.S4±0.14 -1.34±0.22 0.46±0.06
0.23±0.29 -2.96±0.30 2.S0± 0.12 -1.24±0.11 0.58 ±0.16 -1.43 ± 0.25 0.51 ±0.07
0.15 ± 0.31 -3.01 ±0.32 2.67±0.13 -1.26±0.12 0.62±0.lS -1.51 ± 0.28 0.S7±0.08
O.OS±O.34 -3.04 ± 0.34 2.8S±0.14 -1.28 ±0.14 0.6S±0.20 -loSS ± 0.31 0.62 ± 0.09
0.00 ± 0.36 -3.07±0.37 3.01 ±O.lS -1.29 ± 0.15 0.69±0.22 -1.66± 0.34 0.67±0.10
-0.08±0.38 -3.10±0.39 3.18 ± 0.16 -1.30±0.16 0.72±0.24 -1.73 ± 0.3S 0.72±0.1l

This can be integrated using the angular element dk to give

6(r1 - r2) =
1611"2
811"3 '"'
L...J J. . 2 A
]1(kn)]I(kr2)k dky;m(r1)}';m (r2) •
• A
(E.6)
I,m

The integral J... dk gives the result (11" /2)6(r1 - r2)/(n r2),. and one obtains
finally

(E.?)

or

(E.8)

which leads to
21 + 1 6(r1 - r2)
V, (r1,r2) -_ -4-11" r1r2
. (E.9)

337
F. Calculation of Reduced Matrix Element
(1/21); Ilykll(I/21');')
and Some Important Angular Momentum Relations

Here, besides a calculation of the reduced matrix elements for the spherical
harmonics taken between single-particle states (n, I,j), we also discuss some of
the necessary angular momentum coefficient relations, needed in the calculation
of the two-body matrix elements for a a-interaction (see Sect. 3.2.3).
We start from the recoupling expressions, transforming among the equivalent
three-angular momentum basis states,

l(jlh)JI2j); J M) =L(_1)i1+h+i3+ Jil2i13

J 23

X { j.l
J3 J13 I'Jl (j2J3')J13; JM) .
hJ J12} (F.l)

This now can be rewritten as

L { L Ulml,hm2IJI2MI2){JI2MI2,j)m3IJM)}
M12 all mi
X Ijlml}lhm2)Ij)m3}

= L { L (hm2,j)m3IJ13M13)(jlml, J13 M13IJM}iI2i13

J 23 ,M23 all mi

Herefrom, by re-ordering the summation symbols, one obtains

LUI ml ,hm2IJI2MI2)(J12MI2,j)m3IJ M)
M12

= L (-l)il+h+ia+Ji12i13{~~ ~ ~~}
J 23 ,M23

x (hm2,j3m3IJ13M13)(jlmt,J13M13IJM). (F.3)

By putting everything in terms of Wigner 3j-symbols, one finally gets the par-
ticular relation

J J13)
-M M13

338
 =L(_1)J12+m l+m a (~1
M12
X (j3 m3
J J12 )
-M M12
(F.4)

Using the orthogonality relations for Wigner 3j-symbols, the summation on

(J23, M23) can be carried out by multiplying both sides of (F.4) with the 3j-
symbol

( jl J J23 )
ml -M M23
and summing over (ml,-M). Then one gets the relation

i3
( m3 h
m2
J23)
-M23
{hit i2 J23}
J J12
= L (_1)it+J+J12-M-ml-M12 (~1 /:2 J12 )
-M12
M,ml,M12

X(~3 -~ j:2) (}~1 ~ -~23) (F.5)

Putting the above results together, relations of the following type have been
obtained, relations which we denote in a symbolical way by

{: :}=fi(-1)Phase(: : :)(: . :)(: . :)(: :) ,

(F.6)

{...
. .
}(...
..•
) ="(_1)Phase'
L..J
(...
. •.
) (...
. .
3(mi)
) (.... .'),
(F.7)

J~' {: • ;,} ( : • : ; , ) (: -~, ) (2J' + 1)

,

= ~ (: -~ ) (: . ~) (F.8)

Calculation of {(i')jIlY kll<i")j')

i) Using the reduction formula type (I) of (2.104) and since Y k only acts on the
angular part of the single-particle wave function, one gets

(Gl) jllY kll Gl') j'} = (_1i/ 2+1'+j+k] .],

x{), f, !}{IIIYklll'}, (F.9)

339
 ii) We have now to evaluate (lilY kill'). Therefore, we start from the normal
matrix element and apply the Wigner-Eckart theorem. So, we obtain

(ImlYtll'm') =J yr· (r)Yt(r)Y,?,' (r)dr

=(_l)'-m (_~ : ~,) (IllY kill') . (F.lO)

We apply some properties of the spherical harmonics. In particular, we use the

property (Brink 1962)

y,ma(r)y;mb(r)
a b
=" la~
~v4rr
',m

X
( la
0
I
0 (F.ll)

So, the matrix element (F.10) becomes

J y,m· (r)Yt(r)Y,?,' (r)dr

=(_l)m Jy,-m(r)Y,?" (r)Yk"(r)dr

= (_l)m J" v'4,r

~
n'k' (I k' I') ( I
k',,,'
0 0 0 -m
k'
Ii'
I' )
m'

(F. 12)
The latter angular integral gives a Kronecker delta 6kk' ·6"", and (F.12) reduces
to

-1 m n' k'
( ) v'4,r 0 0 0
(I k I') (I k 1') -m Ii m' .
(F.13)

Identifying (F.13) with (F.lO) one immediately has

n'k
(lilY k III') = (_1)' v'4,rOOO' (I k I') (F. 14)

ii) Combining the results of (F.9) and (F.14), the total reduced matrix element
becomes

((!I) jilY kll (!I') j') = (_1)1/2+"+'+j+k3~k

x (~ ~ ~) {}, f, t}· (F.1S)

We now use the relation (F.S) which leads to

340
 (~ ~ ~) {;, ! ~} = (_I)i+i'+{1/2)-(3/2)

( 1 l j) (j, k j)
x 0 -l l l 0 -l
X ( ~!
.,
I I') +
1
0 (_I)i+i'+(1/2)+(3/2)

X ( 1 21 J. ) ( .J, k J. )
o l -l -l 0 l
X () !! ~), (F. 16)

or, combining the right-hand side in more compact form

(-lj k
O!
j') (1 0
l j) (j' l
-l l -l!
I' )
0
X (_I)k+l [1 + (_I)'+I'+k] . (F. 17)

Since we know the explicit expression of the Wigner 3j-symbols

(-lj' I I')
0
! (_1)1'+1
= v'2(21' + 1) ,
(F.18)

we can rewrite (F.17) as

. k .') ( 1)'+I'+k+l 1
( J1 J1 - ~ (1 + (_l)'+I'+k) . (F. 19)
-2 0 2 11' 2
This finally results into

((~[) IIY kll (!l') j') = (_I)i-l/2~

X (!! ~ )) ~(1 + (_I)'+I'+k) . (F.20)

In the calculation of this appendix, we have used the conventions:

i) (1 j21)j coupling in that order,
ii) changing the order to (llj2)j gives an extra phase factor (_I)'+I'+I-i-i'. It
is that particular convention (that of coupling 1+ s to form j, in that order)
that is used throughout the main text except if stated otherwise. The factor
[1 + (_I)l+l'+k] induces a selection rule since one should have

Parity {I + ['} == Parity operator {k} , (F.21)

for non-vanishing matrix elements.

341
G. The Magnetic Multipole Operator

In the present Appendix G, we derive a better "handable" expression for the spin
part in the magnetic multipole operator of Sect 4.2, as expressed by equation
(4.7). We start from (Brussaard, Glaudemans 1977)

f V(rLYf!). {(r x V) x uildr

= f V .A - f V . «r x V) x ui)rLYf! dr . (G.1)

The first part on the right-hand side of (G.1), which is equal to the surface integral

fA. dS , (G.2)

disappears since the spin contribution vanishes through an arbitrary large surface
S surrounding the nucleus.
So, (G.1) becomes

-f V x (r x V)· UirLYf!dr . (G.3)

We now evaluate V x (r x V). We consider the component j, or

{V x (r x V}j = L: CjklClmnOkrmOn , (G.4)

klmn
with
Cjkl =+1
for an even permutation of (1,2,3)

Cjkl=-1

for an odd permutation of (1, 2, 3)

Cjkl =0
in all other cases.
We also have

L: CabcCade = 6 bd • 6 ce - 6 be • 6 cd • (G.5)
a

So, one writes

{V x (r x V)}j = L: (6jm 6 kn - 6jn6km)okrmOn

k,m,n

= L:0krjOk - L:0krkOj
k k

342
 =L(ri)kok +8jkOk)
k

- L(OjOkrk - Ok 8jk) . (G.6)

k

This now becomes in vector notation

(G.7)
or
{V x (r x V)} = rLl- V(V· r - 2) = rLl- V(1 + r· V) . (G.8)
Putting this into the original magnetic multipole operator one gets the result

j(r. (TiLlrLYI! - V(1 + r . V)rLyLM . (Ti)dr . (G.9)

Now, one has

ilrLYLM =0,
(G. 10)
r . VrLYI! =LrLYLM ,
so that finally one obtains the result

j(L + l)Vr L y LM . (Tidr . (G.ll)

H. A Two-Group (Degenerate) RPA Model

With a two-gap, unperturbed Ip - Ih spectrum, we mean that the Ip - Ih

configurations form mainly two groups (both groups containing a number of
Ip - Ih configurations that are close in unperturbed Ip - Ih excitation energy)
with a large interval between the two groups, compared to the average energy
separation within the groups themselves.
The unperturbed spectrum can be drawn schematically in Fig. H.l. Here, we
call the particle-hole configurations m, i(l) and m, i(2) to differentiate between
the groups for a given particle-hole excitation. The analysis goes along the same
lines as in Sect 6.3. The equations (6.39), (6.40) and (6.41) remain valid. We
shall, however, separate the sum over p - h states into two parts explicitly, in
order to be able to make an easy connection to the situation of Fig. H.l. The
dispersion relation corresponding to (6.41) now contains two gaps, characterized
by el and e2, respectively, and thus points towards the existence of a second,
collective excitation nw2 (see Fig. H.2).
We study the solution for nwl, Iiwz in the limit of two groups at the unper-
turbed energy el and e2, respectively. So, (6.41) now becomes

343
 E1
'hW
~

E2

m. i (11 m,i (2)

Fig. H.l. The schematic. unperturbed spectrum for Ip - 1h states within a two-group system of levels,
centered around the energy el and e2. respectively. We have separated the particle-hole notation m, i
into two groups, characterized by an extra group index k i.e. m, i(1) and m, i(2)

m,i(1) m,i(2)

'IIW

El
e:2
(
Fig. H.2. The two-group RPA secular equation. corresponding to the unperturbed spectrwn of Fig, H.l.
The two collective roots are denoted by liwl (lowest one) and AWl (the intennediate root)

1 2e }S} 2e2~
- - + --;;,.-:;;.....;;;-=- (H.l)
X - (er - Ii.w~) (e~ - Ii.w~) •

Solution for n.wt.

To a first approximation, we replace 1i.w0t in the second tenn
by ct. We then obtain
1
li.wf = er - 2Xe}S}
1-
2 S
e2 2X
(H.2)

(e~ - er)
We remark that the second group of unperturbed states near the energy C2 acts
coherently with the first group in lowering the excitation energy.

344
 The nonnalization Nl becomes

N [81 4clc282 ] -1/Z (H.3)

1 ~ (cl -1i.w1 )2 + (c~ - cf}Z '

and the wave functions become

(H.4)

where the notation means

(H.5)

the particle-hole X-amplitudes for particle-hole components from group (1) and
group (2), respectively. A similar notation for the Y amplitudes is used. For the
second root, we then obtain, in a similar way amplitudes

(H.6)

Now, also the second group of levels (at s:' cz) contributes to the collectivity of
the root 1i.w1 and does so in the ratio (cz - cl)/(cl - liwl), relative to the first
group of levels. For a not too large difference cz - Cl, the influence of the second
group is almost as large as for the first group. In second order perturbation theory,
we can calculate the transition probability

l( liwlIDIO) IZ ~ 281 [1 + cz8z [c~ -liw~] [2c~Cz~ -Iiw~liw~ ci]

II
liwl c181 Cz - liw 1 1

+ c~8!
Zr<2
Cl.:11
[dZ _liw~]Z
Cz -
liwz [1 +
1
[ciZ -liwi]Z
Cz -
liwz + ...
1
(H.7)

Since
2c~ - liwi - ci > 2(c~ - ci) > 0 ,

all contributions in (H.7) add up coherently.

Solution for li.wz. We can now, to a first approximation, replace liwOt in the first
tenn of (H.l) by cz. So, one obtains

Iiw! = c~ - 2xcz8z [
1+
~ j.
1
8
IX
(H.8)

(c~ - cf)

Here, one remarks that the collective excitation at liwz is mainly created by a
coherence effect from the states near to cz but that the group of levels near Cl

345
counteracts this coherence. This incoherence between the two groups at e2 and
et can be made even more explicit.
The nonnalization factor N2 is obtained as

N [S2 4ete2.n ] -t/2 (H.9)

2 ~ (e2 _ hw-z)2 + (e~ - e~)2 '
with, subsequently

X2 .(2) f'V Dm,iN2 .

m,l - e2 - hw-z '
(H.IO)
y,? .(2) f'V
D*m,l·N2
m,l - 2e2 '
The minus sign shows the orthogonality between the solutions at n.wt and hw-z.
Similarly as for n.wt the transition probability for the state at n.w2 becomes

1(hw-zIDIO) 12 ~~S2[1- etSt [e~2-~~] [2+ (e~2-~~)]

II
hw-z e2S2 e2 - et (e2 - et)

+ e~Sf
2r12 2 [1 + [e~2_n.w~]2
[e~2_n.w~]2 2 +... , (H. 11)
e2~2 e2 - et e2 - et
where now not all contributions act coherently in the sum. The second solution at
hw-z therefore is less collective than the first solution. The bigger the gap e2 - et
the stronger the collectivity in the second group becomes.
As an application we study the octupole eigenvalues (3- levels) in t46Gd
(Waroquier 1982). In Fig. H.3 we show both the unperturbed 1p - 1h spectrum

o
I
2I 4I

II

Fig. H.3. The realistic case of the unper-

turbed Ip - Ih spectrum with angular mo-
mentum J" = 3- (neutron Ip - Ih con-
figurations: short lines and proton 1p - 1h
configurations: long lines) up to an energy
of E ~ 2OMeV. In the lower part, the RPA
solutions and the E3 transition probability
in 146Gd B(E3; 3i --+ or) are indicated

346
(indicating separately the proton and neutron lp-lh states) and the RPA roots.
The lowest solution at ~ 1.8 MeV is most collective and contains the higher-lying
Ip-lh states in a coherent way. Near 5.6 MeV a next, less collective solution is
obtained, falling exactly in the gap between the unperturbed states near 3 MeV
and near 7.5 MeV. This is a clear, realistic illustration of the discussion given
above.

I. The Condon-Shortley and Biedenharn-Rose Phase

Conventions: Application to Electromagnetic Operators
and BCS Theory

In the present appendix, we shall discuss the two often used phase conventions in
defining the single-particle wave functions (Condon, Shortley 1935, Biedenharn,
Rose 1953): the Condon-Shortley and Biedenharn-rose phase conventions. Quite
some attention has been given in (Rowe 1970), to the possible subtle effects
related with the use of these phase conventions. Also, Brussaard and Allaart
(Brussaard 1970, Allaart 1971a) have been concentrating on the effect of the
two phase choices when incorporating pairing correlations (BCS-theory) in the
description of nuclear excited states, more in particular concerning the calculation
of electromagnetic decay properties. We shall give here a consistent discussion on
the phase choices and their use when calculating electromagnetic decay properties
using BCS theory.

Ll Electromagnetic Operators: Long-Wavelength Form

and Matrix Elements
Using the discussion of Chap.4 and the results of (Waroquier et al. 1975), the
terms contributing to electric multipole transitions can be denoted by

1 ~
O(el., LM) = Iiw (e
~ -; ) [L
i
,
ri YLM (ri), H , A ]
(I.1)

Ii) (2L+3
O'(el.,LM) = -ik (2mc
1 ) (2L + 1 )1/2
L+l

"2 (e
X ~
i
-; ) L(£+I,1) (A
riL+lyM ri ) . Pi , (1.2)
,
O"(el.,LM)= (L~I) (2!C) '2;g.(i)l(rf
,
Y LM (r i»·si, (1.3)

and for the magnetic multipole transitions, we have

O(mag.,LM) = - (L!I) (2!c) '2;, (;) V(rfYf!(ri»·li, (1.4)

347
 (1.5)

(with w = kc). Here, we define the vector spherical hannonics (within the
Condon-Shortley phase convention) as

Yj(I,I)(r) == L {lmllJ.tljm}Yjml(r)e p
m",.
where e p denote the spherical unit vectors (I' = 0, ±1).
An important application is now the evaluation of transition probabilities
using the above operators between an initial state Ii} and a final state If}. Here,
difficulties arise when evaluating the so-called single-particle matrix elements of
the mnltipole transition operators, even in the long-wavelength approximation.
The electric transition operator O(el., LM) still contains many-body effects due
to the presence of the nuclear A-body Hamiltonian H. This problem can be solved
(Allaart 1971a) by replacing the Hamilton operator, acting on the eigenstates
Ii}, If} by its eigenvalues Ei and Ec such that

{fIO(el.,LM)li} = '&~Ef{floeff(el.,LM)li}, (1.6)

where Oeff(el., LM) reduces to the usual, effective electric or Coulomb multipole
operator of (4.9). This effective operator now becomes a one-body operator. We
draw the attention to the difference
O(el., LM) f Oeff(el., LM) , (I.7)

even in the case of photon emission where the factor (.& - Ec) / 1iw reduces to
+1.

1.2 Properties of the Electromagnetic Multipole Operators

Under Parity Operation, Time Reflection and Hermitian Conjugation
Parity. The parity operator can be associated with a unitary transformation op-
erator p, which inverts the spatial coordinates, i.e. one has
PriP-1 = -ri
PPiP-1 = -Pi' Ps iP- 1 = Si , (1.8)
PY[I(ri)p- 1 = (-I)LyLM (ri).
Then, we easily derive the parity transformation properties for the multipole
operators as for (I.2}-(l.3) having the parity (_I)L while (1.4), (1.5) have the
parity (_I)L+l.

Time Reversal. The classical equations of motion for an interacting A-nucleon

system have to be invariant with respect to time-reversal. The time-reversal oper-
ator T only affects kinematical quantities such as the momentum P or spin s of

348
the particles. We can associate many properties with the time-reversal operator,
so we are free to make a specific choice of phases.
Let .,p(r, t) be the eigenfunction of the SchrOdinger equation

.1,( )
H ( r, t)'I" r, t = 1.Ii, o.,p(r,
at
t)
. (1.9)

We define the time-dependent wave functions .,pR(r, t) == 7.,p(r, -t) = .,p*(r, -t)
(with no spin), as the time-reversed wave function of .,p(r, t). It satisfies the
original SchrOdinger equation, provided

7H(r, -t)7- 1 =H*(r, -t) =H(r, t) . (1.10)

The operator 7 means in this case simply complex conjugation. So, a time-
reversal operator can be defined associated with the properties: (i) the Hamilto-
nian remains invariant under time reflection, (ii) the function obtained by oper-
ating with 7 on .,p(r, t) remains a solution to the original SchrOdinger equation,
(iii) the operator 7 can be considered as a combination of two operations: com-
plex conjugation (K) followed by a unitary transformation (7 = UT K), which
will be necessary in the case of wave functions with spin.
Let now Ir) be a state vector, without spin, invariant under time reflec-
tion Tlr) = Ir). Since 7 = UTK, we can also determine the effect of time
reflection in the momentum representation Ip). Here, the momentum p trans-
forms into -p under time reversal. The orbital angular momentum " the spin
8 and total angular momentum i transform into -l, - 8 and -i, respectively.
Since 7i7- 1 = -i, the operator 7 anticommutes with the total angular mo-
mentum i. Therefore, it is convenient to combine 7 with a rotation R through
an angle 11" about an axis perpendicular to the z-axis, say R y • State vectors,
characterized by the angular momentum j and projection m and being discrete
eigenstates of a time-reversal invariant Hamiltonian are therefore also eigen-
functions of the time-reversal operator 7. The eigenvalues of this operator are
related to the phases of the basic state vectors (Rowe 1970). Two phase conven-
tions are generally used: the Condon-Shortley (CS) (Condon, Shortley 1935) and
Biedenharn-Rose (BR) (Biedenharn, Rose 1953) phase conventions. The former
assumes the state vectors Inljm) to be invariant under the 'PR7 operation ('P:
parity operator); the latter assumes Inljm) invariant under the R7 operation, a
combination of Ry(1I") with the time-reversal operator. Following the BR conven-
tion, we derive Tlnljm) = R-1Inljm). The rotation around the y-axis can be
a positive or negative rotation. We shall always use positive rotation (a rotation
that advances a right-handed screw along the axis of rotation). With this choice
Ry(1I") =exp(-i1l"Jy ), which reduces to exp(-i1l"uy /2) = -iuy for intrinsic spin.
This operation iu y can now be associated with the unitary transformation UT
and thus
7 =iUyK . (1.11)
So, in a condensed form, we can write

349
 Condon-Shortley Biedenharn-Rose
'PRTlnljm) = Inljm) , RTI~) = I~) .
The state vector Inlj, m) can be decomposed, using 1- 8 coupling

CS : Inljm) = L (/ml' !m8Ijm)Rnl(r}y,ml(r}x!,{~ ,

(1.12)

where yr == i'y,m. Applying the transformation of spherical harmonics under

rotation Ry(7r) (see Chap.2), we can easily derive the expressions

CS : Tlnljm)= (_l}l+i+mlnlj - m) ,
(1.13)
BR : TI~) = (_I}i+ml~) .

Let now T.ck ) be a spherical tensor, which transforms according to the irreducible
representation D<f;~MI(Ot, (3, ,) of the rotation group (Chap. 2). The transforma-
tion under time reflection is related to

(1.14)

The phase factor C'T is independent of K. but still depends on the choice of
phases for T.ck ). It is evident that the choice of CT can result into real matrix
elements of T~k) between wave functions described in either the CS or the BR
phase convention. As the time-reversal operator is anti-unitary and since a scalar
product remains invariant under a unitary transformation, it follows that

(1.15)

Applying this property to the matrix elements between single-particle states

Inlltil ml) and In2lzjzm2) one has

(1.16)

and we obtain

(1.17)

The C1>h are phases related to the phase convention in which the spherical har-
monics are described and are defined as

Tlnljm) =Cph(-I)i+mlnlj - m) , (1.18)

where Cph = (_1)' in the CS and C1>h = 1 in the BR phase convention (see (1.13».
+'
Relation (1.17) means that matrix elements of T.c k ) are real if CT = (_1)' 1 2 in
the CS convention and if C'T =+1 in the BR convention.

350
 Applying now the above properties, one can derive the following time-
reversal properties

TrnT- 1 = r n ,
TY!:(frn)T- 1 = (-I)MyL- M (r n)(cT = (_I)L) ,
(1.19)
TPnT - 1 = -Pn' TVnT- 1 = V n , TsnT- 1 = -Sn,
TinT-I = -In,
and thus, the transformation properties for the electromagnetic multipole opera-
tors (1.1) to (1.5). They are given in Table 1.1.

Table 1.1. Intrinsic properties of the electromagnetic multipole operators. The c' factors are, of course,
meaningless for the electric multipole operators O(el., LM) since this operator contains, in general,
many-body effects. With O(el., LM), we mean the sum of operators (1.1) to (1.3), with Oeff(el., LM)
the sum of the operators (12), (1.3) and the standard Coulomb operator (Chap.4, equation (4.9»

r,.k) Parity Tr,.k)Tl CT (T~k»+ CH C'

CS BR

O(el.,LM) (_I)L
(_I)L+l
(-I)MO(el.,L - M) (_I)L
(_I)L+l
(_I)M+l0(el.,L - M) -I - -
(_I)L+l
O(mag.,LM) (-I)M+lO(mag., L - M) (_I)M O(mag., L - M) +1 +1
Oeff(el.,LM) (_I)L (_I)Moelf(el.,L - M) (_I)L (_I)Moelf(el,L - M) +1 +1 (_I)L

Hermitian Conjugation. An irreducible tensor of rank k transforms under rota-

tion according to
nT(k)n -I
IC
=~ r k)D(k)
L-t K/ K.' ,It
(0' (3 .... )
, " ,
(1.20)
K'

where (0', (3,,) correspond to the Euler angles, n(O'{3,) to a unitary transforma-
tion. Taking the Hermitian conjugation of (l.20) and using the unitary property
of n we get

n(T,.k)tn- 1 = L(T~~)t D~~~:(O', (3,,) . (1.21)

K'

Applying properties of the rotation operators and changing the sign of '" and ",',
it follows that (-I)K(T!~)+ transforms under rotation in the same way as the
T~k), or, we can write

n(-I)K(T!~tn-l = L(-I)K'(T!~,tD~~K(O',{3,,). (1.22)

K'

So, the Hermitian adjoint ~k)+ of the tensor operator ~k), can be defined as

(1.23)

351
where CH is a phase factor independent of /'i,. It can be fixed by the additional
condition for self-adjoint operators

(I. 24)

'The momentum p, the orbital angular momentum I, the spin II are self-adjoint
operators. 'The Hennitian conjugation of the vector spherical hannonics simplifies
by taking the conjugation of the ordinary spherical hannonics and the spherical
unit vectors e", appearing in the definition

y,m'(rn)e/l = ~)lml' IJLlim)Yj(i,l)(r n) , (I.25)

j,m

and results into

m (Arn
(y;(1,1) »+ = (- 1)'+l-;+my-m (A )
;(1,1) rn . (I. 26)

Since LyLM(r n) and V x LyLMe"n) can be expressed in terms of the vector

spherical harmonics, we can also write that

(LyLM(rn)t = (_I)L+M+l LYiM(r n),

(I.27)
(V x LyLM(r n»+ = (_I)L+M+IV x LyL-M(-;'n).

The Hermitian conjugate of the operators (I.l}-(l.5) can easily be derived (see
Table 1.1).
The interchange of initial and final states in the reduced single-particle matrix
elements is related to the behavior of the operator T(k) under Hermitian conjuga-
tion. Using the equations (1.23), (1.24) and applying the Wigner-Eckart theorem,
we derive

We can finally obtain the relation

(nl1till1rk)lln2lzh) = c'( -1)it -i2 (n2lzhll rk )llnl l tid , (I.29)

where we have defined

(I.30)

Hence, the coefficient c' only depends on the considered angular momentum
phase convention and on the convention satisfying the tensor operator k ) under r
time reflection and Hermitian conjugation. We remind that Cph = (_1)1 in the CS
convention and Cph = 1 in the BR convention. So, in Table 1.2, we give the
properties of the reduced matrix elements corresponding to the electromagnetic
multipole operators (I.l}-(1.5) in both the CS and BR phase conventions. This
indicates that in the more standard CS convention matrix elements are real fer
the operators (1.1}-(1.5) and Oeff(el., LM) requiring of course the selection rule
11 + 12 + L even (or odd) for electric (or magnetic) operators to be fulfilled. In

352
Table 1.2. Propenies of the reduced matrix elements of the electromagnetic multipole operators. The
state vectors Inljrn) are described in the considered phase convention (CS: Condon-Shonley 1935,
BR: Biedenham-Rose 1953)

~k) BR
I< CS

{real if Leven
O(el..,LM) real
imago if L odd

{ real if Leven
O"rr(el., LM) real
imago if L odd

O(mag.,LM) real {real if L odd

imago if L even

iLO(el., LM) {real if Leven

real
imago if L odd

iLO"fI(el., LM) {real if Leven

real
imago if L odd

iL-10(mag., LM) {real if L odd real

imago if Leven

the BR convention one has real matrix elements for L even (or odd) for electric
(magnetic) multipole operators, respectively. For the operators iLO(el., LM) and
i(L-l)O(mag., LM), the matrix elements, using the BR convention become all
real! (see Table 1.2).

1.3 Phase Conventions in the BCS Formalism

In second quantization, state vectors, single-particle tensor operators etc. are
represented in terms of creation- and annihilation operators (see Chap. S)

CS BR
a~IO) =lnI11I1jl1ml1) ' a~IO) =I~) .
The state vectors are defined in (1.12) in such a way as to transform under time
reversal according to their corresponding phase convention. The relation between
creation and annihilation operators in the different phase conventions is given by

(1.31)

The transformation under time reflection then becomes

353
 CS .• Ta+T-
cr
1 = _(_I)'a s a a+- a '
TsO/a_O/T- 1 = (-I) 'a aO/ ,
(1.32)
BR .• Ta+T-
0/
1 = -s 0/ 0.+-0/'
T sO/a_O/ T- 1 = 0.0/ .
Here, SO/ denotes the phase factor (_I)i4- m 4.
In a quasi-particle representation, using the BCS transfonnation (see Chap. 7),
one has
CS : c~ = uaa~ -
sO/vaa_O/ ,
(1.33)
BR : c~ = uaa~ - sO/vaa_O/ .

The same transfonnation (1.32) under time-reversal for the quasi-particle oper-
ator exists as well as the relation (1.31) between these operators. The different
amplitudes U a and Va are related by

(1.34)

Signs can be associated with the amplitudes after studying the expression for the
energy gap in the BCS fonnalism (Baranger 1960):
'I

CS : Lla =L ~cucvcG(aacc;O),
c Ja
'I
(1.35)
BR : Lia = L ~c ucvcG(aacc; 0) ,
c Ja

where G(abcd; 1) denotes the antisymmetric two-body matrix element (the pre-
cise relation is G(abcd; J) = -1/2«1 +6ab )(1 +6cd »1/2{ab; JMlVlcd; JM}nas).
The relation between the pairing matrix elements in both phase conventions reads

G(aacc; 0) = (_1)14 -Ie G(aacc; 0) , (1.36)

and results into

(1.37)
The residual pairing matrix elements G( aacc; 0) are usually real and positive. The
gap equations (1.35) can be transfonned into a real non-symmetric eigenvalue
problem by introducing the relation 2uava = Ll a/ Ea. In the BR convention, the
positive definite matrix [M], appearing in the secular equation

Lia = LMacLic, with

c

- = 2"1'
Mac
l1c E1 . G(aacc;O)
-
, (1.38)
Ja c

354
Table 1.3. Phases of the BCS parameter in the Condon-Shortley (CS) and Biedenham-Rose (BR)
phase conventions, respectively

CS BR

Ua ( _1)IG +1
Va +1 +1
da ( _1)IG +1
Ea +1 +1

always involves an eigenvector with positive components and it only has one
such eigenvector, applying the Frobenius theorem (W"illdnson 1965). In the BR
convention the gaps Lia will hence all be positive and so too the amplitudes U a
and Va. On the other hand, in the CS convention, the energy gap ..1 a and the
amplitude U a can have a negative sign if the parity of the orbitalla is odd. All
results are given in Table 1.3. We remark that the sign of ..1 a has no significance
since in the BCS expression one always has combinations ..1:
(or u a ..1 a ).
Any single-particle tensor T,.k) of rank k can then be expressed, using second
quantization and the quasi-particle operators, as the sum of three distinct terms:

(1.39)
a

~k)(2) =~ I:(nalajallrk)llnbhjb}
k ab
X (UaUb - c'(-I)k vavb ) X AO(ab, k",) , (1.40)

T~k)(3) = ~ I:(nalajallrk)lInblbjb}
k ab
X (UavbA+(ab, k",) + UbVa( _l)k+1C A(ab, k - ",» , (1.41)
with

AO(ab, h) = I: (jama,jb - mblk",}spc:cp . (1.42)

The expressions are the same in the BR convention, on condition that all symbols
are described within the BR convention. The operators AO, A+ and A denote
two-quasi-particle creation operators (see (Waroquier, Heyde 1970) and (Heyde,
Waroquier 1971». The phase relations are

A+(ab, k",) = i'G+'b .J+(ab, k",) ,

(1.43)
A(ab, k - ",) = (_i)'G+'b .J(ab, k - ",) .

355
We apply this onto the enhanced E2 transitions in spherical even-even nuclei as
an example. For E2 transitions between a 2qp state and the BCS Oqp vacuum
state IO}, one has contributions from the term (1.41) only and obtains
- rnlk)
{OI.L~ A +(ab, JiMi)IO)
-

.
=OJ;koM!_,,(-I)k+"[vaUb + c'(-I)kuavb]l/k{jaIITk)lIjb} . (1.44)

The pairing factor in (1.44) now becomes (vaub+(-I)kuavb) and (IVaUbl+luavbl>

in the CS and BR conventions respectively and thus the same final numerical
factors result, irrespective of the phase convention used!

356
Problems

Chapter 1
1.1 Prove the relation that the square of the angular momentum operator 12
can be written as (see (1.10,11, 12»

12 =r2p2 + Ii? :r (r2 !) .

1.2 Show that the operator J = 1 + 8, where 1 describes the orbital angular
momentum operator and 8 the intrinsic spin angular momentum operator
for a spin 1/2 particle, constitutes an angular momentum operator.
1.3 Derive an explicit form of the 2P3/2l m =+1/2 wave function in terms of
the spin and orbital angular momentum wave functions.
1.4 Show that the total angular momentum operator J2 for a nucleon (obtained
by coupling the orbital and intrinsic spin angular momentum operator) can
be diagonalized in the basis Ilm,}11/2m s }. Determine the eigenvalues and
show that the corresponding eigenfunctions are also eigenfunctions of the
Hamiltonian H = Ho + al . B with

p2 1 22
Ho = 2m + :2 mw r .

1.5 Prove the orthogonality relations for Clebsch-Gordan coefficients (see

(1.96».
1.6 Show that the recoupling of (1.131) indeed leads to a Wigner 9 - j symbol
(with no extra phase factor).
1.7 Show that the recoupling coefficients {(jlh)J12h; JMljl(j2h)J23; JM)
that describe the transformation between states with different coupling
order in systems with three angular momenta, are independent of M.
1.8 Determine the relative weights for the S = 0 and S = 1 intrinsic spin
components in the l(lds/2)2; J =2} two-particle wave function.
1.9 Discuss the classical limit of the Wigner 6 - j symbols, according to the
methods of Sect. 1.9.
1.10 Calculate the probability density P(j) (Le. the probability that the length
of j lies between j and j + dj is P(j)dj) if we suppose that, according to
the upper part of Fig. 1.5, j 1 rotates at constant rate about the z-axis with
respect to j2. Then P(j) is inversely proportional to dj / dt.

357
Chapter 2

2.1 Derive the transfonned function of F(x, y) = a(x 2_y2) as F'(x, y) =2axy
(the new function), using the fonnal definition

F'(x, y) = exp [-*<PzLz] F(x, y) ,

for a rotation of <pz =45 0 around the z-axis.

2.2 Derive (2.10) i.e. show that the rotation operator for a general angular
momentum operator J and a rotation over an angle 0: around an axis,
specified by the unit vector becomes :nO',
UR(O:) =exp [-*o::n a . J] .
2.3 Derive the transfonnation matrix acting on the coordinates of a point
P(x, y) for a rotation around the z-axis over an angle <p, using the trans-
fonnation properties of the spherical harmonics YiM (r) as given in (2.34).
2.4 Show, by explicit calculation, that the transfonnation matrix in (2.74), for
a system of two spin-l/2 particles decouples into an S = 0 and S = 1
part, when coupling the two spin-l/2 systems.
2.5 Detennine the representation 2 x 2 matrix exp(-if3CT y /li) in the space
spanned by the spin 1/2 eigenvectors.
2.6 Given are the D-functions for spin-l/2. Study the effect of rotation of
1800 , 3600 , 7200 around the x-axis concerning the transfonnation prop-
erties of the spin-l/2 eigenvectors x!i:
(see figure). Discuss your results.

TableP.l. The Wigner Dl/~ m,m

(Ot, (3, -y) rotation matrix representation for the basis of the
intrinsic spin 8 = 1/2 states

m +1/2 -1/2

m'

+1/2 e- iOl / 2 cos«(3/2)e-i-y/2 _e- iOl / 2 sin«(3/2)ei-y/2

-1/2 eiOl /2 sin«(3 /2)e-i-y /2 eiOl /2 cos«(3 /2)ei-Y /2

358
 2.7 Prove, as an example of a tensor product, for the rank 1 tensors r(l) and
p(l), that ,(1) = -iV2[r(l) ® p(1)](1).
2.8 Derive the reduction rule n (2.110) using the methods of Sect. 2.8.1.

Chapter 3
3.1 The three-dimensional harmonic oscillator can also be solved by separat-
ing the eigenvalue equation for the energy into the cartesian coordinates
(x, y, z). Show that in this case cp(x, y, z) becomes a product of three
one-dimensional oscillator wave functions (product of Hermite functions).
Show that a relation exists with the solutions of the eigenvalue equation
in spherical (r,8,cp) coordinates (see (3.16».
3.2 Derive the Hartree-Fock equations from the variational condition, ex-
pressed by (3.43).
3.3 Derive the antisymmetrized and normalized two-body matrix elements
(jajb; JMlVljcid; JM}nas where V is a central two-body force V(r) us-
ing the Moshinsky transformation brackets.
3.4 Study the three-level model, where two levels are degenerate at the energy
e and only interact via the intermediate of a third level at energy e' with a
strength V (see figure). Study the energy eigenvalues and wave functions
as a function of Lle/V (with Lle = Ie' - el).

(3)
E'

<V> /
1\ ,,
I
, <V>
I \
(1 )
t. ~
(2)
E

3.5 Show that perturbation theory summed to all orders in calculating the
energy of the level (1) with unperturbed energy Hu, by including the
interaction with a second level at Hn(H22 > Hu) with strength H12
gives the result

\ _ Hll + H22
1\- - 2
1 [(H
- 2: 11 - H)2
22 + 4H212]1/2 .
3.6 Construct the three-particle wave functions for a (j)3 J =j configuration
with j ~ 7/2. We give the normalized cfp coefficients

[p(Jl)jJnl J] =
[ SJI JO + 2ioil {J~ J ~ol}]
j
[3 +6(2jo + 1) { ~ JJ. ~oo}] 1/2
Jj
.
and

359
 {j, j, ~} =(-1)i+i'+J<33,)-1 .

3.7 Show, by explicit calculation, that the (lg9/2)3 J =9/2 configuration has
two independent states with J = j, depending on the original angular
momentum of the two-particle configuration (Jo) one starts from.
3.8 a) Show that the interaction energy in an n-particle state IjnaJ M} with
general n(n ~ 2) can be expressed as a function of the two-body anti-
symmetrized matrix elements.
b) Show that in the calculation of a three-particle spectrum, starting from
a two-body spectrum, one only needs the relative matrix elements

AJ1 = (i; 1t IV Ii; Jl) - =olVli; Jl =O) .

(i; Jl
3.9 Given are n degenerate levels jl,j2 ... ,jn(eit = eh = ... = ein).
De-
termine the energy spectrum for two particles in the configuration space
with Jff: = 0+ if the interaction matrix elements

(h)2; J =01V1(jk)2; J =O) =-; JIJk (G > 0) .

Determine also the wave function corresponding with the lowest Jff: = 0+
eigenstate.
3.10 In a nucleus, the two valence nucleons move in the 199 / 2 and 2Pl/2
orbitals. The energy spectrum of 0+ states looks like shown in the figure
(a). If only the 2Pt /2 orbital would be considered, only one }'If: =0+ state
shown in the figure (b) results. Determine the relative energy difference

0+ 1. 7888

0+ 0.8944

0+ 0.0000

o. b.

between the 199 / 2 and 2Pl/2 single-particle energies Lle. Determine also
the wave functions for the 0+ ground state in case b.
Given are the two-body matrix elements

(ja)2; J =01V.sI(jc)2; J =O) =FJ

2
(2ja + 1)(2jc + 1)(_1)iG+i.+ 1G +1•

x
( ja
1/2 -1/2 0
ja 0) (1/2 jc jc
-1/2 0
0) .
360
3.11 We give the two-particle spectrum in l~DY82 with Z = 64, N = 82 as
a closed-shell configuration. The relative spectrum is Eo+ = 0, Ez+ =
1.677MeV, E4 + = 2.427MeV, E6 + = 2.731 MeV, Es+ = 2.8323 MeV
and E lO + = 2.9177 MeV.
Calculate the three-particle spectrum in l:~Ho82. Compare the calcu-
lated spectrum with the experimental situation E1l/2- = 0, E l 5/2- =
1.560 MeV, E l9 / 2- = 2.287 MeV, En/2- = 2.594 MeV, Ez7/2- =
2.738 MeV. Discuss the agreement (or disagreement) between the data
and the calculated spectrum. Enclosed: table of (11/2)3 cfp coefficients.
3.12 The spectrum of ~Ti28 is given in the figure, with r =0+,2+,4+ and 6+
spectrum.
a) The cfp (7/2)3 coefficients are given in the table. Calculate the spectrum
in IIV28 starting from the experimental spectrum in the figure.
b) In the 5lV spectrum, a number of positive parity states ( r =
1/2+,3/2+) occur that cannot be described within a (117/2)3 spectrum.
Give some suggestions for the possible origin of these states.
3.13 Derive' the explicit (T, T z ) dependence of the isoscalar, isovector and
isotensor (rank 2) term energy E(T, T z) defined as (T, TzIHIT, T z).
3.14 Show, by acting with the isospin lowering operator T_ = L:L(i) on the
wave function corresponding with the nucleon distributions of Fig. 3.53,
that these wave functions have a good isospin T = Tz (maximal isospin).

I. 0

50 T 51
22 '28 23 V28

320 6+

30

1.+ 2.70 15/2-

.
:;
~
2.68
2.68
255
3/2+
1/2+
21.1 3/r
>-
<:>
0:
~20
w
z 1.S1 9/r
~
I-
<!
I-
155 2+ 1.61 11/r
0
x
w

1.0
0.93 3/2-

0 3 2 ' - - - - 5/r
____ 0 _ _ _ _ 7/2-
0.0

361
 Table P.2. The cfp coefficients [(11/2)2J\, 11/2;JI}(11/2)3Jj and this for the J = 3/2, 5/2, 7/2, 9/2, 11/2, 13/2, 15/2, 17/2, 19/2,21/2, 23/2 and 27/2 states.
Co)
CI>
For J = 9/2, 11/2 and 15/2, two-independent states can be constructed (two rows)
J • 1712
N

VI JI v =3
( ~ )3 .........
J. 3/2 4 0.463CH
t -C.57e6~5
e -0.37222~
VI JI V • 3 J • 11/2 Ie 0.558128
.......••
-0.H6131 VI JI • I V • 3 19/2
O. 819C~9
...•..•..
o e -0.527e46 e. VI JI • 3
J. 5/2 2 2 C.2357e2 e.26Ct;8 .........
2 ~ 0.316228 -C.67C2'4
........ 2 6 O. Hoe5e 0.IOC946 2 4 -0.5604~~
2 e 0.43461~ o .61C H~
VI JI 2 t C.21ce42
• 3 2 Ie 0.483C46 -C.Hf8B C.~060H
•..•....• 2 e
2 Ie C.621C53
2 ~ -0.770911
-0.28~26e
J • 13/2
2 t
2 8 0.5t9gec J • 21/2
VI JI • 3
J. 712 VI JI • 3
2 2 0.4638C8
2 4 -0.477255
VI JI • 3 2 f -0.4672ge 6 C.383482
......... e -C.H41l~
2 8 -0.0857~1
Ie O.470"~1
-0.550~82
2 Ie 0.575664
2
2 ~ 0.250813
2 6 0.507519 23/2
0.613561 J • 15/2
2 e
VI JI V • 3 VI JI v
J. 9/2
........ 2 6 -0.545331
2 2 -0.356966 -C.4]~1~1
VI JI V • 3 2 e C.4CIC22
• 3 2 ~ 0.292l4e -c.e~e212
.•....•.. 6 -0.311HC e.70e245 2 IC 0.736C68
2
-e.6~BI2 2 8 0.8185H -0.eC6146
2 -0.221937 C.566C47
0.3551~7 -0.~9Ht9 2 Ie C.I418C5
2 ~ J • 2712
2 6 -0.7e07H -C.C3e8~5
2 e 0.460985 -0.eH8CO
2 Ie 0.050H6 C.~BllO VI JI V • 3
.........
2 8 -C.486664
1 IC C.873589
 TableP.3. The cfp coefficients [(7/2)2(VIJI), 7/2; JI}(J /2)3, vJ) and this for J = 7/2,
31l. 51l. 9/2, IIIl and 15/2. The extra quantum number v, VI denotes the seniority quantum
number

~
0 2

0 2 4 6

7
2
I
-2 :i2
6
I
2
:LIT
6

3 3
2 v'! -v'lf
5
2 ~
3 2 VI -3~
2
~ -~
7
2 3 14 3J22
11
T -~ A 2~
3 11
15
T -VI Vi
3.15 Construct the isospin wave functions for a two-nucleon system (n - n,
n -p,p- p):
a) Calculate the expectation value for the above wave functions of the
interaction

V(l,2) = - Vo!(r + 7"1 • 7"2) •

b) Determine the energy contribution for the above operator in the
(117 /2)2 configurations in 42Ca, 42SC and 42Ti.

Chapter 4

4.1 Derive a general relation between the one-particle and the two-particle
(J1r =2+) electric quadrupole moment.
4.2 Study the sign of the two-particle electric quadrupole moment for a (j)2 J
configuration as a function of J.
4.3 Derive the magnetic dipole moment for a two-particle configuration
J.L(j2, 1) as a function of the one-particle magnetic moment J.L(j). Study
the dependence on J (magnitude and sign).
4.4 Derive the additivity rules (the equivalent of (4.64) and (4.65» for general
ML and EL moments.

Chapter 5
5.1 Calculate the energy spectrum for the two-particle configurations IP; J M)
(for J = 0,2, ... , 2j - 1) starting from the pairing interaction (5.39).

363
 Discuss the results and compare with the energy spectrum resulting from
as-interaction.
5.2 Derive the matrix (5.40) for a S-force interaction and also for a P2(COS 81 2 )
force. Show that for the latter force in the limit oflarge j(j ..... 00), forward
scattering results preferentially with m' =m.
5.3 Derive the angular momentum coupled form of the two-body interaction
of (5.52) in second quantization.
5.4 Determine the particle-hole two-body matrix elements

in terms of the particle-particle two-body matrix elements.

5.5 Show that the quadrupole moment for a hole state and for a particle
state have opposite sign but the same absolute value (neglecting any core
contribution).

Chapter 6

6.1 Derive the transition probability for the collective and non-collective
states, obtained from the schematic TDA model, to the ground state
1(0IDlliw1}12 and 1(0IDlliwdegenerate}12. Show the coherence of the dif-
ferent p - h components in the transition probability from the collective
state.
6.2 Derive the secular equation for the angular momentum coupled p - h rep-
resentation Iph -1; J M} in the case of the TDA and RPA approximations.
6.3 Study the general properties of the RPA non-hermitian eigenvalue problem
(normalization, orthogonality, metric, ...).
6.4 Starting from the schematic TDA (or RPA) model, show that the excitation
energy of the collective root is inversely proportional, for a given strength,
with the model space dimension.

Chapter 7

7.1 Derive the commutation relations for the S-pair fermion creation operator
(7.14)
[Sj,Sjl- =1-njfl.
7.2 Derive the number projected BCS wave function (7.38) starting from the
more standard BCS wave function

7.3 Show that the condition

364
 indeed leads to the BCS equations for u", v" and this for given G and ell
values.
7.4 Derive the angular momentum coupled (for constant pairing force strength
G) form of the BCS equations (7.56).
7.5 Show that, for a given E" spectrum of one-quasi particle energies and
given particle number n, the BCS equations can be inverted to give the
ell and force strength G (use Frobenius method).
7.6 Derive a general form of the electromagnetic operator, expressed in quasi-
particle operators, for the Biedenharn-Rose (BR) phase convention. How
do the pairing reduction factors of Table 7.1 change formally.
7.7 Derive the quadrupole-quadrupole matrix element of (7.108) and the en-
ergy eigenvalues when diagonalizing the 2-level energy matrix of (7.109).

365
References

Abramowitz, M., Stegun, I.A.: Handbook of Mathematical Functions (1964) Dover, New York
Alder, K., Wmther, A. (1971): Phys. Lett 348, 357
Alkhazov, GD. et al. (1976): Nucl. Phys. A274, 443
AI1aart, K. (1971): Nucl. Phys. A174, 545
Al1aart, K., Boeker, E. (1971): Nucl. Phys. A168, 630
Allaan, K., Boeker, E., Bonsignori, G., Savoia, M., Gambhir, Y.K. (1988): Phys. Repts. 169, 209
Anantaraman, N., Schiffer, I.P. (1971): Phys. Lett. 378, 229
Arima, A., Kawasada, H. (1964): I. Phys. Soc. lap. 19, 1768
Arima, A., Iachello, F. (1975): Phys. Lett. 578, 39
Arima, A., Iachello, F. (1975): Phys. Rev. Lett. 35, 1069
Arima, A., Iachello, F. (1976): Ann. Phys. (N.Y.) 99, 253
Arima, A., Iachello, F. (1978): Ann. Phys. (N.Y.) 111, 201
Arima, A., Iachello, F. (1978): Phys. Rev. Lett. 40, 201
Arima, A., Iachello, F. (1979): Ann. Phys. (N.Y.) 123, 436
Arima, A., Iachello, F. (1984): in Advances in Nuclear Physics, Vol. 13 (139)
Arndt, R.A., Mac Gregor, M.H. (1966): Phys. Rev. 141, 873
Ballhausen, CJ., Gray, H.B. (1964): Molecular Orbital Theory (W.A. Benjamin, New York)
Baranger, M. (1960): Phys. Rev. 120, 957
Baranger, M. (1961): Phys. Rev. 122, 992
Bardeen, I., Cooper, L.N., Schrieffer, I.R. (1957): Phys. Rev. 108, 1175
Beiner, M., Flocard, H., Van Giai, N., Quentin, P. (1975): Nucl. Phys. A238, 29
Bertin, M.C., Benczer-Koller, N., Seaman, G.E., MacDonald, I.R. (1969): Phys. Rev. 183, 964
Biedenharn, L.C., Rose, M.E. (1953): Revs. Mod. Phys. 25, 729
Bjerregaard, I.H., Hansen, 0., Nathan, 0., Chapman, R., Hinds, S., Middleton, R. (1967): Nucl. Phys.
AI03,33
Blin-Stoyle, RJ. (1956): Revs. Mod. Phys. 28, 75
Bloch, C., Messiah, A. (1962): Nucl. Phys. 39, 95
Blomqvist, I., Wahlbom, S. (1960): Arkiv. Fys. 16, 545
Bohr, A. (1951): Phys. Rev. 81, 134
Bohr, A. (1952): Mat. Fys. Medd. Dan. Vid. Selsk. 26, nO 14
Bohr, A., Mottelson, B. (1953): Mat. Fys. Medd. Dan. Vid. Selsk. 27, n016
Bohr, A., Mottelson, B. (1955): Dan. Mat. Fys. Medd. Dan. Vid. Selsk. 30, nOl
Bohr, A., Mottelson, B., Pines, D. (1958): Phys. Rev. 110, 936
Bohr, A., Mottelson, B. (1969): Nuclear Structure, Vol. 1 (W.A. Benjamin, New York)
Bohr, A., Mottelson, B. (1975): Nuclear Structure, Vol. 2 (W.A. Benjamin, New York)
Bohr, A. (1976): Revs. Mod. Phys. 48,365
Brink. D.M. (1957): Nucl. Phys.4, 215
Brink. D.M., Satchler, G.R. (1962): Angular Momentum (Oxford Univ. Press, London)
Brody, T.A., Moshinsky, M. (1960): Tables o/Transformation Brackets (Monografias delo Instituto
de Fisica, Mexico)
Broglia, R.A., Federman, P., Hansen, 0., Hehl, K., Riedel, C. (1968) Nucl. Phys. AI06, 421
Brown, G.E., Boisterli, M. (1959): Phys. Rev. Lett. 3, 472

367
Brown, G.E. (1964): Unified Theory of Nuclear Models (North-Holland, Amsterdam)
Brown, G.E., Kuo, T.T.S. (1967): Nue!. Phys. A92, 481
Brown, G.E., Jackson, AD. (1976): The Nucleon-Nucleon Interaction (North-Holland, Amsterdam)
Brown, G.E., Dehesa, J.S., Speth, J. (1979): Nue!. Phys. A330, 290
Brueckner, K.A., Eden, RJ., Francis, N.C. (1955): Phys. Rev. 100, 891
Brueckner, K.A., Gammel, JL. (1958): Phys. Rev. 109, 1023
Brueckner, K.A., Thieberger, R. (1960): Phys. Rev. Lett. 9, 466
Brueckner, K.A., Masterson, K.S. (1962): Phys. Rev. 128, 2267
Brussaard, P J., Tolhoek, H.A. (1957): Physica 23, 955
Brussaard, PJ. (1967): Ned. T. Natuurkunde 33, 202
Brussaard, PJ. (1970): Nue!. Phys. A158, 440
Brussaard, PJ., Glaudemans, P.WM. (1977): Shell-Model Applications in Nuclear Spectroscopy
(North-Holland, Amsterdam)

Campi, X., Sprung, D. (1972): Nue!. Phys. A 194,401

Casten, R.F., Warner, D.D., Brenner, D.S., Gill, R.L. (1981): Phys. Rev. Lett. 47, 1433
Casten, R.F. (1985): Phys. Lett. 152B, 145
Casten, R.F. (1985): Phys. Rev. Lett. 54, 1991
Casten, R.F. (1985): Nue!. Phys. A443, 1
Cavedon, J.M. et a1. (1982): Phys. Rev. Lett. 49, 978
Clebsch, A. (1872): Theorie der Biniiren algebraischen Formen (Teubner, Leipzig)
Condon, E.U., ShortIey, G.H. (1935): The Theory of Atomic Spectra (Cambridge University Press)
Ch.ill
Decharg~, J., Gogny, D. (1980): Phys. Rev. C21, 1568
De Gelder, P., et a1. (1980): Nue!. Phys. A337, 285
de-Shalit, A., Talrni, I. (1963): Nuclear Shell Theory (Academic, New York)
de-Shalit, A., Feshbaeh, H. (1974): Theoretical Nuclear Physics (Wiley, New York)
Dieperink, A.EL., Leenhouts, H.P., Brussaard, P J. (1968): Nue!. Phys. A116, 556
DOE Nne!. Science Adv. Committee, Dec. 1983: A Long Range Planfor Nuclear Science, A report
by the NSF
Duval, P.D., Barrett, B.R. (1981): Phys. Rev. Lett. 46, 1504
Duval, P.D., Barrett, B.R. (1981): Phys. Rev. C24, 1272

Eberz, J. et a!. (1987): Nne!. Phys. A464, 9

Eckart, C. (1930): Revs. Mod. Phys. 2, 305
Edmonds, A.R. (1957): Angular Momentum in Quantum Mechanics (Princeton University Press)
Egmond, A. Van, Allaart, K. (1983): Prog. Part. Nnel. Phys. 9, 405
Eisenberg, J.M., Greiner, W. (1970): Nuclear Theory, Vo!. 2 (North-Holland, Amsterdam) p. 93
Elliott, J.P, Flowers, B.H. (1957): Proc. Roy. Soc. A242, 57
Elliott, J.P. (1958): Proc. Roy. Soc. A245, 128,562
Elliott, J.P., Harvey, M. (1963): Proe. Roy. Soc. Al72, 557
Elliott, J.P., Jackson, A.D., Mavrornatis, H.A., Sanderson, E.A., Singh, B. (1968): Nne!. Phys. A121,
241
Endt, P., Van der Leun, C. (1973): Nne!. Phys. Al14, 1

Faessler, A., Plastino, A. (1967): Nnovo Cim. 478,297

Fetter, A.L., Walecka, J.D. (1971): Quantum Theory of Many-Particle Systems (Me Graw-HiII, New
York)
Flowers, B.H. (1952): Proc. Roy. Soc. Al12, 248
Fliigge, S. (1974): Practical Quantum Mechanics (Springer, Berlin-Heidelberg)
Fortune, H.T. (1978): Phys. Lett. 768, 267
Frois, B., et a!. (1983): Nue!. Phys. A396, 40ge

Gambhir, Y K., Rimini, A., Weber, T. (1969): Phys. Rev. 188, 1573

368
Gambhir, Y K., Rimini, A., Weber, T. (1971): Phys. Rev. C3, 1965
Gambhir, YX., Rimini, A., Weber, T. (1973): Phys. Rev. C7, 1454
Gillet, V., Green, AM., Sanderson, E.A. (1966): Nue!. Phys. 88, 321
Gilmore, B. (1974): Lie Groups, Lie-Algebras and Some of Their Applications (Wiley-Interscienee,
New York)
Gloeckner, D.H., Lawson, RD. (1975): Phys. Leu. 56B, 301
Goldstein, H. (1980): Classical Mechanics (Addison-Wesley, Reading, MA)
Gordan, P. (1875): Uber das Formensystem biniirer Formen (feubner, Leipzig)
Hamada, T., Johnston, I. (1962): Nue!. Phys. 34,382
Hamermesh, M. (1962): Group Theory and its Application to Physical Problems (Addison-Wesley,
Reading, MA)
Haxel, 0., Jensen, J.HD., Suess, HE. (1949): Phys. Rev. 75, 1766
Heisenberg, W. (1932): Z. Phys. 77, 1
Helmers, K. (1961): Nuel. Phys. 23, 594
Herling, G.H., Kuo, T.T.S. (1972): Nue!. Phys. A181, 113
Herzberg, G. (1944): Atomic Spectra and Atomic Structure (Dover, New York)
Heyde, K., Waroquier, M. (1971): Nue!. Phys. A167, 545
Heyde, K., Waroquier, M., Meyer, R.A. (1978): Phys. Rev. C17, 1219
Heyde, K., Van Isaeker, P., Waroquier, M., Wood, J.L., Meyer, R.A. (1983): Phys. Repts. 102, 291
Heyde, K., Van Isaeker, P., Casten, R.F., Wood, JL. (1985): Phys. Leu. 1558, 303
Heyde, K., Sau, J. (1986): Phys. Rev. C33, 1050
Heyde, K. (1989): Mod. Phys. A9, 2063
Holm, G., Borg, S. (1967): Ark. Phys. 36, 91
Hsieh, S.T., Knfipfle, K.T X., Mairle, G., Wagner, GJ. (1975): Nue!. Phys. A243, 380
Iaehello, F. (1980): in Nuclear Spectroscopy, Lect. Notes Phys., Vol. 119, edited by G. Bertsch, D.
Kurath (Springer, Berlin-Heidelberg)
Iaehello, F. (1983): Lie Groups, Lie Algebras and Some Applications (Trento Lecture Notes), U .T.F 97
Iaehello, F., Arirna, A. (1988): The Interacting Boson Model (Cambridge Univ. Press)
Irvine, JM. (1972): Nuclear Structure Theory (Pergamon, Oxford)
Kerek, A., Holm, G.B., Borg, S., De Geer, L.E. (1972): Nue!. Phys. A195, 177
Kerman, AX. (1961): Ann. Phys. (N.Y.) 12, 300
Kerman, AX., Lawson, RD., MacFarlane, M.H. (1961): Phys. Rev. 124, 162
Kisslinger, L.S., Sorensen, R.A. (1960): Mat. Fys. Medd. Dan. Vid. Selsk. 32, n09
Klingenberg, P.F.A. (1952): Revs. Mod. Phys. 24, 63
Kocher, D.C. (1975): Nuc!. Data Sheets 16, 445
Kohler, H.S. (1976): Nue!. Phys. A258, 301
KrewaId, S., Klemt, V., Speth, J., Faessler, A. (1977): Nue!. Phys. A281, 166
Kuo, T.T.S., Brown, GE. (1966): Nue!. Phys. 85, 40
Kuo, T.T.S., Baranger, E., Baranger, M. (1966): Nue!. Phys. 79, 513
Kuo, T.T.S., Baranger, E., Baranger, M. (1966): Nue!. Phys. 81, 241
Kuo, T.T.S. (1967): Nue!. Phys. A90, 199
Kuo, T.T.S., Brown, GE. (1968): Nuel. Phys. A114, 241
Kuo, T.T.S. (1974): Ann. Rev. Nucl. Sci. 24, 101
Kuo, T.T.S. (1981): in Topics in Nuclear Physics I, (Proceedings, Beijing, China) Lect. Notes. Phys.,
Vo!. 144 (Springer, Berlin-Heidelberg)
Lacombe, M., et aI. (1975): Phys. Rev. DU, 1495
Lacombe, M., et aI. (1980): Phys. Rev. C21, 861
Lane, AM. (1964): Nuclear Theory, (Benjamin, New York)
Landau, LD. (1956): JETP (Sov. Phys.) 3, 920
Lederer, CM., Hollander, JM., Perlman, I. (1968): Table of Isotopes (Wiley, New York)
Liu, K.F., Brown, GE. (1976): Nue!. Phys. A265, 385

369
MacFarlane, M.H. (1966): in Lectures in Theoretical Physics, Vol. VIIIc, edited by PD. Kunz, D.A.
Lind, W.E. Brittin (Univ. of Colorado Press, Boulder) 583
MacGregor, M.H., Arndt, R.A., Wright, R.M. (1968): Phys. Rev. 169, 1128
MacGregor, M.H., Arndt, R.A., Wright, R.M. (1968): Phys. Rev. 169, 1149
Macbleidt, R., Holinde, K., Elster, Ch. (1987): Phys. Repts. 149, 1
March, N.H., Young, W.H., Sampanthar, S. (1967): The Many-Body Problem in Quantum Mechanics
(Cambridge Univ. Press, London)
Mayer, M.G. (1949): Phys. Rev. 7S, 1969
Mayer, M.G. (1950): Phys. Rev. 78, 22
Moszkowski, S.A. (1970): Phys. Rev. C2, 402
Mouelson, B.R. (1967): in Topics in Nuclear Structure Theory (Nordita, Copenhagen), No. 288
MoUclson, B.R. (1976): Revs. Mod. Phys. 48, 375
Myers, WD., Swiatecki, WJ. (1966): Nucl. Phys. 81,1
Nadasen, A., et al. (1981): Phys. Rev. cn, 1023
Negele, J.W. (1970): Phys. Rev. C1, 1260
Negele, J.W., Vautherin, D. (1972): Phys. Rev. CS, 1472
Negele, J.W. (1983): Comm. Nucl. Part. Phys. 12, 8
Otsuka, T., Arima, A., Iachello, F., Th1rni, I. (1978): Phys. Leu. 76B, 139
Otsuka, T., Arima, A., Iachello, F. (1978): Nucl. Phys. A309, 1
Paar, V. (1979): Nucl. Phys. A331, 16
Parikh, J.C. (1978): Group Symmetries in Nuclear Structure (Plenum, New York)
Pauli, W. (1925): Z. Phys. 31, 373
Pittel, S., Duval, PD., Barrett, B.R. (1982): Ann. Phys. (N.Y.) 144,168
Puddu, G., Scholten, 0., Otsuka, T. (1980): Nucl. Phys. A348, 109
Racah, G. (1942): Phys. Rev. 61, 186
Racah, G. (1942): Phys. Rev. 62,438
Racah, G. (1943): Phys. Rev. 63, 367
Racah, G. (1949): Phys. Rev. 76, 1352
Racah, G. (1951): Lectures on Group Theory (lnst. for Adv. Study, Princeton)
Racah, G., Talmi, I. (1952): Physica 18, 1097
Rainwater, J. (1950): Phys. Rev. 79,432
Rainwater, J. (1976): Revs. Mod. Phys. 48, 385
Reid, R.V. (1968): Ann. Phys. (N.Y.) SO, 411
Ring, P, Schuck, P. (1980): The Nuclear Many-Body Problem (Springer, Berlin-Heidelberg)
Rose, HJ., Brink, D.M. (1967): Revs. Mod. Phys. 39, 306
Rotenberg, M., Bivins, R., Metropolis, N., Wooten Jr., J.K. (1959): The 3-j and 6-j Symbols (M.I.T.
Technology Press, Cambridge, MA)
Roulet, C., et al. (1977): Nuc1. Phys. A28S, 156
Rowe, DJ. (1968): Revs. Mod. Phys. 40, 153
Rowe, DJ. (1970): Nuclear Collective Motion (Methuen, New York)
Rowe, DJ., U11ah, N., Wong, S.S.M., Parikh, J.C., Castel, B. (1971): Phys. Rev. C3, 73
Sau, J., Heyde. K .• CMry. R. (1980): Phys. Rev. C21. 405
Sau. J•• Heyde. K .• Van Maldeghem, J. (1983): Nucl. Phys. A410. 14
Scholten, o. (1980): The Interacting Boson Model and Some Applications, Ph. D. Thesis, Univ. of
Groningen, The Netherlands
Scholten, O. (1983): Phys. Rev. C28, 1783
Segr~, E. (1977): Nuclei and Particles (Benjamin, Reading, MA)
Shlomo, S., Schaeffer, R. (1979): Phys. Lett. 838, 5
Sick, I., McCarthy, I.S. (1970): Nucl. Phys. AlSO, 631
Sick, I. (1973): Nucl. Phys. Al08, 557

370
Skynne, T B.R. (1956): Philos. Mag. I, 1043
Sorensen, R.A., Lin, ED. (1966): Phys. Rev. C14l,729
Speth, 1., Van der Woude, A. (1981): in Repts. Prog. Phys. 44, 719
Start, D.P., Anderson, R., Carlson, L.E., Robertson, A.G., Grace, M.A. (1970): Nucl. Phys. AI62,49

Tabakin, F. (1964): Ann. Phys. (N.Y.) 30, 51

TaImi, I. (1952): Helv. Phys. ACl2S, 185
Talmi, I. (1971): Nucl. Phys. AI72, 1
Tape, J.W., Ulrickson, M., Benczer, N., MacDonald, J.R. (1975): Phys. Rev. CI1, 2125
Van der Werf, S.Y., et al. (1986): Phys. Lett. 116B, 372
Van Maldeghem, J., Heyde, K. (1985): Phys. Rev. C3l, 1067
Van Maldeghem, J. (1988): Ph. D. Thesis, University of Gent, Belgium
Van Ruyven, JJ., et al. (1986): Nucl. Phys. A449, 579
Vautherin, D., Brink, D.M. (1972): Phys. Rev. CS, 626
VoId, P.B .. et al. (1978): Phys. Lett. 72B, 311
VoId, P.B., et al. (1978): Nucl. Phys. A302, 12
Waroquier, M., Heyde, K. (1970): Nucl. Phys. AI44, 481
Waroquier, M., Heyde, K. (1971): Nucl. Phys. AI64, 113
Waroquier, M., Vanden Berghe, G. (1972): Phys. Lett. 41B, 267
Waroquier, M., Heyde, K. (1974): Z. Phys. 268, 11
Waroquier, M., Heyde, K., Vmcx, H. (1975): Physica BOA, 465
Waroquier, M., Heyde, K., Vmcx, H. (1976): Phys. Rev. CI3, 1664
Waroquier, M., Sau, J., Heyde, K., Van Isacker, P., Vmcx, H. (1979): Phys. Rev. CI9, 1983
Waroquier, M., Heyde, K. (1981): fJ8 Session d'Etudes Bie1l1Ulle de Physique Nucllaire (Aussois), S17
Waroquier, M. (1982): Hoger Aggregaatsthesis, University of Gent, unpublished
Waroquier, M., Heyde, K., Wenes, G. (1983): Nucl. Phys. A404, 269
Waroquier, M., Wenes, G., Heyde, K. (1983): Nucl. Phys. A404, 298
Waroquier, M., Bloch, J., Wenes, G., Heyde, K. (1983): Phys. Rev. C28, 1791
Waroquier, M., Ryckebusch, J., Moreau, J., Heyde, K., Blasi, N., Van der Werf, S.Y., Wenes, G.
(1987): Phys. Repts. 148, 249
Waroquier, M.: Priv. comm. and to be published
Watanabe, H. (1964): Progr. Theor. Phys. 32, 106
Weissbluth, N. (1974): Atoms and Molecules (Academic, New York)
Werner, S.A.E. (1980): Phys. Today 11, 24
Wigner, B.P. (1937): Phys. Rev. 51,106
Wigner, B.P. (1959): Group Theory and its Application to the Quantum Mechanics of Atomic Spectra
(Academic, New York)
Wildenthal, BB. (1969): Phys. Rev. Lett. 22, 1118
Wildenthal, BB., Mc Grory, J.B., Halbert, B.C., Graber, HD. (1971): Phys. Rev. C4, 1708
Wildenthal, BB. (1976): Varenna Lectures 69, 383
Wildenthal, BB. (1985): in Int. Symp. on Nuclear Shell Models, edited by M. Valli~res, B.H. Wilden-
thal (World-Scientific, Singapore) 360
Wilkinson, JB. (1965): The Algebraic Eigenvalue Problem (Clarendon, Oxford)
Wohlfart, HD., Shera, B.B., Hoehn, M.V., Yamazaki, Y.. Fricke, G., Steffen, R.M. (1978): Phys.
Lett. 73B, 131
Wuod, JL. (1983): Nucl. Phys. A396, 245c
Wood, J L.: Private communication
Wright, R.M., Mac Gregor, MB., Arndt, R.A. (1967): Phys. Rev. 159,1422
Wybourne, B.G. (1974): Classical Groupsfor Physicists (J. Wiley, New York)
Yukawa, H. (1935): Proc. Phys. Math. Soc. Jpn. 17,48
Yutsis, A.P., Levinson, I.B., Vanagas, V.V. (1962): Mathematical Apparatus of the Theory ofAngular
Momentum (Israel Program for Scientific Translations, Jerusalem)

371
Subject Index

Additivity relations 157 Clebsch-Gordan coefficient

Angular momentum - (see also Wigner 3j-symbol) 16ff
- addition 15ff, 17ff - program 298
- coupling formulas 28ff Coefficient of fractional parentage (cfp)
- definition 4ff - n -> n - 1 particle cfp 112
- generalization 12 - n -> n - 2 particle cfp 112
- illustration for spin If2 13 - three particle cfp 110
- matrix representation 13 Computer programs 298ff
- spherical representation 327 Condon-ShortJey phase convention
Annihilation operator 161ff - angular momentum coupling 18
Antisymmetry - em. operators 221,347ff
- many-particle system 112ff - second quantization 260
- matrix element 76 Configuration
- particle wavefunction 75 - application to 18 0 100ff
- definitions 101
- mixing 101ff,147fI
BCS (Bardeen-Cooper-Schriefer)
Contraction 176
- applications 217ff
Core polarization 272,276,292
- gap equations 211,215
Coulomb interaction 124
- ground-state 208
Creation operator 161ff
- method 206 Current density 137,262
- pair distributions 213
- phase conventions 353ff
Delta-interaction 88fI,334
- program 323 - multipole expansion 335
- transformation 208 - program 309
Biedenharn-Rose Phase convention 205, Density distributions
222,347 - charge 72
Binding energy 57,72,255 - in (e, e') scattering 74
Bloch-Messiah theorem 258 - mass 73
Born approximation 331 Density-dependent interactions 256,261,
Broken-pair model 203,206, 229ff 265,291
Brown-Bolsterli model 183 Dipole (El)
- giant resonances 186,288
Central interaction - IDA mode 185
- potential 4 Dipole moment 153ff,225
- two-body matrix elements 87ff Diagonalization
Charge - energy matrix 103
- density 73 - program 316
- effective 142ff D-matrix (see Wigner-D matrix) 37
- independence 119
- operator 122 Effective
- symmetry 119 - charge 142ff, 150
Chemical potential 211 - mass 264,289fI

373
- two-body interactions 77,273 - symmetry properties 240
Eigenvalue equations 104 - truncation properties 246
Electric multipole Interactions
- operator 137 - effective 77,116,117
- phases 347ff - realistic 79,84
- quadrupole moment 149,226 - schematic 82, 87ff, 132ft"
- single-particle estimate 139 Irreducible tensor operator 43,45
- transitions 138, 221ff, 282 Isospin
- two-particle properties 145 - algebra 121ff
Electromagnetic transitions 136ff,347ff - Clebsch-Gordan coefficient 125
Elementary excitations 179ft' - introduction 119
(EWSR) Energy-weighted sum rule 283 - multiplet 123
Exchange matrix element 76, 182 - neutron excess 130
Extended Skynne force 260ff,265 - reduced matrix element 124
- second quantization 170
- spinor 121
Fermi level 179,197,211,214
- symmetry energy 266
Fermi momentum 266
Fractional parentage (see cfp) 110, 112
Frobenius theorem 355 Kinetic energy density 262ft"
Kuo-Brown interaction 273
Gap (BCS) equations 211
Gaussian interaction 85,99,235,276 Ladder operators 10ft'
Generalized seniority 232 Laguerre polynomials 63,99
g-factor 153ff Lanczos algorithm 104
Giant multipole resonances 287 Landau description 266
Goldstone diagrams 166, 167 Lifetime 136, 142, 146
Ground-state Long-range interaction 168,199
- binding energy (see binding energy) 57, LS coupling wave functions 26
72,255
- correlations 186,283 Magic numbers 56
Majorana force 249
Hamada-Johnston potential 83,84 Magnetic multipole
Hamilton density 262 - dipole moment 153
Harmonic oscillator - operator 137,342
- potential in shell-model 61 - phases 347
- radial integral program 320 - single-particle estimate 139
- relative motion 96 - state subspace 169
Hartree-Fock - transitions 138
- multipole field decomposition 198 - two-particle dipole moment 156
- second quantization 177 Mapping of matrix elements 247ff
- simple approach 70ft' Matrix elements 52ff
- variational approach 72 Mean-free path (nucleon) 55
Hartree-Fock-Bogoliubov 256ff Mirror nuclei 120
Hermitian conjugation 351ff Model space 101ff, 116ff, 142ff
HF + BeS equations 260 Moshinsky transformation 95ff
Hole operators 171 MSDI 91,93, 102, 116
Householder method 104 Multipole expansion 87,198

Interacting Boson Model NEW (non-energy weighted sum rule) 282

- matrix element mapping 247ff Newton-Raphson method 211
- parameters 249ft' N",N" scheme 244
- shell-model truncation 245ff Non-local force 270

374
Nonnal ordering 175 Realistic interactions 79ff
Nuclear Rearrangement effects 291ff
- incompressibility 266 Reduced matrix element 49ff
- mass 57 - reduction rule I 50
- stability 56 - reduction rule IT 51
Nucleon density 262 - single-particle matrix element 338ff
Number operator 201 Renonnalization 274
Rotation
Oblate distribution 150 - matrices (see Wigner D-matrices) 37
Odd-even mass difference 215, 217ff - operator 32, 37,172
Odd-odd nuclei 132ff, 158 - properties 52ff
One-body matrix element 113,139,165 - scalar field (0 (3)) 31ff
OPEP-potential 54 RPA (Random Phase approximation) 186ff,
280ff
Pair creation operator 201,231 - two-group degenerate model 343ff
Pairing
- correlations 197ff Schematic interactions 82, 85ff
- degenerate single j-shell 200ff Schematic model
- interaction 168 - particle-hole IDA 183ff
- non-degenerate levels 204ff,206ff - particle-hole RPA 189ff
- phase properties 353ff Schmidt values 153ff
- potential 257 SOl-interaction 91,220,221,331ff
- schematic models 200ff,205ff Second quantization
- Skyrme interactions 271ff,275 - angular momentum coupling 169
- spectroscopic factors 226ff - comparison with first quantization 165ff
- strength 218 - diagrammatic expressions 166,167
- two quasi -particle spectra 219 - hole states 171
Parabolic rule 132ff - operators 161,165,221
Parity operator 348 - rotation properties 172
Particle-hole - spectroscopic factors 226ff
- application to 16 0 192ff Seniority 200,202,229,231
- excitations 179ff Semi-classical interaction 93
- Skyrme interactions 279ff Self-energy corrections 211
- states 173 Self-consistent shell-model 254ff
Particle-particle excitations 197, 272ff Shon-range interaction 168,182,199,230
Pauli-spin matrices 13ff Single-particle
Perturbation expansion 106 - estimates 139
Phase parameters 331,336 - quadrupole moment 149
Polarization charge 150 - spectra (illustrations) 68ff
Prolate distribution 150 Skyrme
Proton-neutron multiplet 133ff - effective interactions 194, 254ff
- forces 265ff,275
Quadrupole moment - parametrization 265ff
- one-particle 149 Slater integrals 88
- two-partiCle 152 - program 304
Quadrupole-quadrupole interaction 243 SO(3)group 38ff
Quasi-Boson approximation (QBA) 188 Spectroscopic factors 226ff
Quasi-particle excitations 208, 212ff, 256ff Spherical Bessel function 333
Quasi-spin 231 Spherical harmonics 8ff
Spin
Radial integral 87ff, 146 - current density 262
- program 320 - density 262
Radial single-particle equation 61ff,264 - eigenvector (spinor) 14

375
- exchange operator 86,261 - matrix elements (many-particle) 114
- geometry of intrinsic spin 1{2 40ff - matrix elements (two-particle) 74ff
Spin-orbit Two-particle
- experimental indication 67 - energy spectra 91ff,131ff,199,276ff
- splitting 65 - isospin 125ff
SU(2) group 38ff - residual interactions 77ff,261
Symmetry properties - wave functions 74ff
- isospin 125ff
- Wigner 6j-symbol 23ff Unit tensor operator 49

Talmi integral 100

Vacuum
IDA (Tamm-Dancoff approximation) 181ff
- expectation value 174ff, 210, 263
Tensor - state 161,174,181,186,208
- cartesian 45 Variational problem 256ff,263
- irreducibility 43,45 Vector coupling coefficients (see Wigner
- operators 52ff nj-symbols)
- product 43,47 Velocity dependent forces 262
- reduction 329ff Vibrational excitations
- spherical 48 - closed-shell nuclei 283ff
Thomas-Fermi model 267 - open-shell nuclei 239ff
Three-body forces 256,261
Three-particle
- wave functions lOSff Weisskopf estimate 140
- interaction matrix elements 115ff Wick's theorem 176,210,256
Time-reversal operator 348ff Wigner
Transformation - classical limit of nj-symbols 27
- general group transformations 35ff - D-matrices 37
- Moshinsky 95ff - Eckart theorem 48, 136
Truncation - 3j-symbol and properties 20ff
- effective charge 142 - - program 299
- ffiM model space 246 - 6j-symbol and recoupling 22ff,109,328
- shell-model space 101,242 - - program 300
Two-body - 9j-symbol and recoupling 25ff
- matrix elements (isospin) 130 -- program 303

376
 Springer Series In
Nuclear
and Particle Physics
Enquiries and manuscripts should be addressed to the editor working on your
subject (or to the managing editor)

Editors
Mary K. Gaillard Department of Physics
The University of California
Berkeley, CA 94720, USA
J. Maxwell Irvine Department of Theoretical Physics
The Schuster Laboratory
The University
Manchester, M13 9PL, United Kingdom

Erich Lohrmann Il.lnstitut fOr Experimentalphysik

Universitat Hamburg
Luruper Chaussee 149
2000 Hamburg, Fed. Rep. of Germany

Vera Liith SLAC

Stanford University
PO Box 4349
Stanford, CA 94309, USA

Achim Richter Institute of Nuclear Physics

Technical University of Darmstadt
Schlossgartenstrasse 9
6100 Darmstadt, Fed. Rep. of Germany

Managing Editor
Ernst F.Hefter Physics Editorial
Springer-Verlag
Postfach 105280
6900 Heidelberg 1, Fed. Rep. of Germany