1

Lecture Notes in Nuclear Structure Physics

B. Alex Brown
November 2005
National Superconducting Cyclotron Laboratory
and Department of Physics and Astronomy
Michigan State University, E. Lansing, MI 48824
CONTENTS 2
Contents
1 Nuclear masses 6
1.1 Masses and binding energies . . . . . . . . . . . . . . . . . . . . . . . 6
1.2 Q values and separation energies . . . . . . . . . . . . . . . . . . . . 10
1.3 The liquid-drop model . . . . . . . . . . . . . . . . . . . . . . . . . . 18
2 Rms charge radii 25
3 Charge densities and form factors 31
4 Overview of nuclear decays 40
4.1 Decay widths and lifetimes . . . . . . . . . . . . . . . . . . . . . . . . 41
4.2 Alpha and cluster decay . . . . . . . . . . . . . . . . . . . . . . . . . 42
4.3 Beta decay . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
4.3.1 Beta decay Q values . . . . . . . . . . . . . . . . . . . . . . . 52
4.3.2 Allowed beta decay . . . . . . . . . . . . . . . . . . . . . . . . 53
4.3.3 Phase-space for allowed beta decay . . . . . . . . . . . . . . . 57
4.3.4 Weak-interaction coupling constants . . . . . . . . . . . . . . 59
4.3.5 Double beta decay . . . . . . . . . . . . . . . . . . . . . . . . 59
4.4 Gamma decay . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
4.4.1 Reduced transition probabilities for gamma decay . . . . . . . 62
4.4.2 Weisskopf units for gamma decay . . . . . . . . . . . . . . . . 65
5 The Fermi gas model 68
6 Overview of the nuclear shell model 71
7 The one-body potential 77
7.1 General properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
7.2 The harmonic-oscillator potential . . . . . . . . . . . . . . . . . . . . 79
7.3 Separation of intrinsic and center-of-mass motion . . . . . . . . . . . 81
7.3.1 The kinetic energy . . . . . . . . . . . . . . . . . . . . . . . . 81
7.3.2 The harmonic-oscillator . . . . . . . . . . . . . . . . . . . . . 83
8 The Woods-Saxon potential 87
8.1 General form . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
8.2 Computer program for the Woods-Saxon potential . . . . . . . . . . . 91
8.2.1 Example for bound states . . . . . . . . . . . . . . . . . . . . 92
8.2.2 Changing the potential parameters . . . . . . . . . . . . . . . 93
8.2.3 Width of an unbound state resonance . . . . . . . . . . . . . . 94
8.2.4 Width of an unbound state resonance at a xed energy . . . . 95
9 The general many-body problem for fermions 97
CONTENTS 3
10 Conserved quantum numbers 101
10.1 Angular momentum . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
10.2 Parity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
10.3 Isospin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
11 Quantum numbers for the two nucleon system 107
11.1 Two neutrons or two protons . . . . . . . . . . . . . . . . . . . . . . 107
11.2 Proton-neutron . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
11.3 Two nucleons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
12 The Hartree-Fock approximation 110
12.1 Properties of single Slater determinants . . . . . . . . . . . . . . . . . 110
12.2 Derivation of the Hartree-Fock equations . . . . . . . . . . . . . . . . 111
12.3 Examples of single-particle energies . . . . . . . . . . . . . . . . . . . 113
12.4 Results with the Skyrme hamiltonian . . . . . . . . . . . . . . . . . . 115
12.4.1 Binding energies . . . . . . . . . . . . . . . . . . . . . . . . . 118
12.4.2 Single-particle energies . . . . . . . . . . . . . . . . . . . . . . 121
12.4.3 Rms charge radii and charge densities . . . . . . . . . . . . . . 122
12.4.4 Displacement energies . . . . . . . . . . . . . . . . . . . . . . 129
13 Angular Momentum and Tensor Algebra 136
13.1 Angular momentum coupling . . . . . . . . . . . . . . . . . . . . . . 136
13.1.1 Coupling of Two Angular Momenta . . . . . . . . . . . . . . . 136
13.1.2 Coupling of Three Angular Momenta . . . . . . . . . . . . . . 138
13.1.3 Coupling of Four Angular Momenta . . . . . . . . . . . . . . . 140
13.2 Tensors and reduced matrix elements . . . . . . . . . . . . . . . . . . 142
13.2.1 Special reduced matrix elements . . . . . . . . . . . . . . . . . 143
13.2.2 Products of tensor operators . . . . . . . . . . . . . . . . . . . 144
13.2.3 Other convensions for reduced matrix elements . . . . . . . . . 146
14 Single-particle electromagnetic moments 148
14.1 General results and notation . . . . . . . . . . . . . . . . . . . . . . . 148
14.2 General results for closed-shell and single-particle congurations . . . 149
14.3 Magnetic moments . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
14.4 Electric quadrupole moments . . . . . . . . . . . . . . . . . . . . . . 154
15 The Creation Operator Method 159
15.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159
15.2 Creation Operators and Wavefunctions . . . . . . . . . . . . . . . . . 159
15.3 Operators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162
15.4 Examples of a
+
and a Matrix Elements . . . . . . . . . . . . . . . . . 162
CONTENTS 4
16 Many-Body Wavefunctions 165
16.1 Wavefunctions in the m-scheme . . . . . . . . . . . . . . . . . . . . . 165
16.2 Wavefunctions in the J-scheme . . . . . . . . . . . . . . . . . . . . . 170
16.3 Angular momentum projection . . . . . . . . . . . . . . . . . . . . . . 173
16.3.1 Examples for (j = 5/2)
n
congurations . . . . . . . . . . . . . 176
16.4 General form for the matrix elements in the J scheme . . . . . . . . . 179
17 The Two-Body Hamiltonian 181
17.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181
17.2 Form of the two-body matrix elements . . . . . . . . . . . . . . . . . 182
17.3 General types of interactions . . . . . . . . . . . . . . . . . . . . . . . 183
17.4 Transformation from relative to center-of-mass coordinates . . . . . . 185
17.5 Simple potentials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 186
17.6 One boson exchange potentials . . . . . . . . . . . . . . . . . . . . . . 187
17.7 Numerical examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193
17.8 Density-dependent interactions . . . . . . . . . . . . . . . . . . . . . 193
18 Applications of the Two-Body Interactions 197
18.1 Interaction energies for closed shell, one-particle and one-hole congu-
rations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197
18.2 Interaction energies for diagonal two-particle congurations . . . . . . 201
18.3 Interaction energies for diagonal two-hole congurations . . . . . . . . 203
18.4 Interaction energies for particle-hole congurations . . . . . . . . . . 204
19 Conguration mixing 206
19.1 Two-particle congurations . . . . . . . . . . . . . . . . . . . . . . . 206
19.2 Application to
18
O . . . . . . . . . . . . . . . . . . . . . . . . . . . . 209
19.3 Many-particle congurations . . . . . . . . . . . . . . . . . . . . . . . 210
20 One-particle transfer 216
20.1 Fractional parentage coecients . . . . . . . . . . . . . . . . . . . . . 216
20.1.1 One orbit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 216
20.2 Many orbits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221
20.3 Spectroscopic factors . . . . . . . . . . . . . . . . . . . . . . . . . . . 222
20.3.1 Basic denitions and sum-rules . . . . . . . . . . . . . . . . . 222
20.3.2 Isospin dependence . . . . . . . . . . . . . . . . . . . . . . . . 223
20.3.3 Simple situations . . . . . . . . . . . . . . . . . . . . . . . . . 226
20.3.4 Center-of-mass corrections . . . . . . . . . . . . . . . . . . . . 227
20.3.5 Computation of shell-model spectroscopic factors . . . . . . . 228
20.4 Overlap functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229
20.4.1 Denition and properties . . . . . . . . . . . . . . . . . . . . . 229
20.4.2 Asymptotic properties . . . . . . . . . . . . . . . . . . . . . . 230
20.4.3 The well-depth prescription . . . . . . . . . . . . . . . . . . . 231
CONTENTS 5
20.4.4 Beyond the well-depth prescription . . . . . . . . . . . . . . . 232
21 Experiments related to spectroscopic factors 236
21.1 Experimental results for specic nuclei . . . . . . . . . . . . . . . . . 236
21.1.1
51
V
50
Ti . . . . . . . . . . . . . . . . . . . . . . . . . . . . 236
21.1.2 (
3
He,d) in the sd shell . . . . . . . . . . . . . . . . . . . . . . 237
21.1.3 (e,e

p) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 242
21.1.4 Proton decay . . . . . . . . . . . . . . . . . . . . . . . . . . . 245
21.1.5 Radioactive beams . . . . . . . . . . . . . . . . . . . . . . . . 247
21.2 Structure models for specic nuclei . . . . . . . . . . . . . . . . . . . 247
21.2.1
7
Li . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 248
21.2.2
12
C . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 249
21.2.3
16
O . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 249
21.2.4 The sd-shell . . . . . . . . . . . . . . . . . . . . . . . . . . . . 250
21.2.5
40
Ca and
48
Ca . . . . . . . . . . . . . . . . . . . . . . . . . . . 251
21.2.6
51
V . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251
21.2.7
208
Pb . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 252
21.3 Short-range correlations . . . . . . . . . . . . . . . . . . . . . . . . . 252
21.4 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 254
22 One-body transition operators and the OBTD 258
22.1 Isospin and proton-neutron formalism . . . . . . . . . . . . . . . . . . 259
22.2 OBTD for a single-orbital conguration . . . . . . . . . . . . . . . . . 260
22.3 OBTD for a two-orbital conguration . . . . . . . . . . . . . . . . . . 260
22.4 Scalar one-body matrix elements . . . . . . . . . . . . . . . . . . . . 262
23 Two-particle transfer operators 265
24 Two-body transition operators and the TBTD 269
24.1 Scalar two-body operators . . . . . . . . . . . . . . . . . . . . . . . . 270
24.2 Scalar TBTD for a single-orbital conguration . . . . . . . . . . . . . 271
24.3 Scalar TBTD for a two-orbital conguration . . . . . . . . . . . . . . 271
24.4 Sample calculations for interaction energies . . . . . . . . . . . . . . . 273
25 Electromagnetic transitions 275
25.1 Operators and transition rates . . . . . . . . . . . . . . . . . . . . . . 275
25.2 Moments in terms of electromagnetic operators . . . . . . . . . . . . 277
25.3 Nuclear matrix elements . . . . . . . . . . . . . . . . . . . . . . . . . 277
25.4 Applications to simple situations . . . . . . . . . . . . . . . . . . . . 279
25.4.1 Closed shell plus one particle . . . . . . . . . . . . . . . . . . 279
25.4.2 Single-orbit congurations . . . . . . . . . . . . . . . . . . . . 280
CONTENTS 6
26 Allowed beta decay 282
26.1 Formulation of allowed beta decay . . . . . . . . . . . . . . . . . . . . 282
26.2 Operators for allowed beta decay . . . . . . . . . . . . . . . . . . . . 283
26.2.1 Fermi decay . . . . . . . . . . . . . . . . . . . . . . . . . . . . 283
26.2.2 Gamow-Teller decay . . . . . . . . . . . . . . . . . . . . . . . 285
26.3 Sum rules . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 287
26.4 Eective operators for Gamow-Teller matrix elements . . . . . . . . . 288
1 NUCLEAR MASSES 6
1 Nuclear masses
1.1 Masses and binding energies
A basic quantity which can be measured for the ground states of nuclei is the atomic
mass M(N, Z) of the neutral atom with atomic mass number A and charge Z. Atomic
masses are usually tabulated in terms of the mass excess dened by
(N, Z) M(N, Z) uA, (1.1)
where u is the Atomic Mass Unit dened by u = M(
12
C)/12 = 931.49386 MeV/c
2
.
1
I will use the data from the 2003 compilation of Audi, Wapstra and Thibault [1].
Fig. (1.1) shows the position on the nuclear chart for these measured masses together
with the experimental error. There are 2127 nuclei measured with an accuracy of 0.2
MeV or better and 101 nuclei measured with an accuracy of greater than 0.2 MeV.
For heavy nuclei one observes several chains of nuclei with a constant N Z value
whose masses are obtained from the alpha-decay Q values.
Nuclear binding energy is dened as the energy required to break up a given
nucleus into its constituent parts of N neutrons and Z protons. In terms of the
atomic masses M(N, Z) the binding energy B(N, Z)
2
is dened by:
B(N, Z) = ZM
H
c
2
+NM
n
c
2
M(N, Z)c
2
, (1.2)
where M
H
is the mass of the hydrogen atom and M
n
is the mass of the neutron.
3
In
terms of the mass excess the binding energy is given by:
B(N, Z) = Z
H
c
2
+N
n
c
2
(N, Z)c
2
, (1.3)
where
H
c
2
= 7.2890 MeV and
n
c
2
= 8.0713 MeV.
How do we know that nuclei are made up of protons and neutrons? In the 1920s
when it was observed that nuclei decay by the emission of alpha particles, protons and
electrons one tried to make nuclear models out of constituent protons and electrons.
4
However, after the discovery of the neutron in 1932, it was observed that the
1
This and other constants can be found on the website:
http://physics.nist.gov/cuu/Constants/index.html
2
The binding energy will also be denoted by BE.
3
This binding energy also includes contribution from the Coulomb interaction between electrons
which is approximately given by 1.43x10
5
Z
2.39
MeV. On the scale of nuclear binding this can
usually be ignored. It is most important for heavy nuclei where, for example, for Z = 120 the
electronic contribution is 1.34 MeV.
4
Just because barks come out of dogs does not mean that dogs are made of barks. (Denys
Wilkinson).
1 NUCLEAR MASSES 7
Nuclei with measured masses
P
r
o
t
o
n

N
u
m
b
e
r
Neutron Number
0
20
40
60
80
100
120
0 20 40 60 80 100 120 140 160
0
0.2
0.4
0.6
0.8
>1.0 MeV error
Figure 1: Nuclei with measured masses. The solid line is for N = Z. The
dashed lines indicate the locations of the magic numbers 8, 20, 28, 40, 50, 82
and 126
atomic masses can be qualitatively understood by the contribution of the masses of
the individual protons and neutrons which make up the nucleus. To emphasize this
point the ratio
M(N, Z) [ZM
H
+NM
n
]
[ZM
H
+NM
n
]
is plotted in Fig. (1.2) as a function of mass. The largest deviation is found near
A = 60 where the total mass is only about one percent smaller than expected from
the sum of nucleon masses. The intrinsic properties of neutrons and protons inside
the nucleus are essentially the same as those of the free nucleons. Nuclear properties
are a result of these nucleons interacting with each other through the exchange of
mesons. At some level we will need to include small admixtures of other baryons.
We are interested in understanding the binding energy as a function of N and
Z. The total BE are shown in Fig. (1.3) as a function of A. One observes an overall
linear increase with A reaching a maximum value of about 2 GeV for the heaviest
nuclei. One can bring out more detail by plotting BE/A as in Figs. (1.4) and (1.5)
[Note that Fig. (1.4) is the inverse Fig. (1.2)]. The maximum as a function of A in
Fig. (1.4) are shown separately in the bottom of Fig. (1.6). These represent the nuclei
which are most stable and the most abundant in nature. An expanded portion for
the experimental for light nuclei is shown in Fig. (1.7). These are the nuclei at the
1 NUCLEAR MASSES 8
%

d
e
v
i
a
t
i
o
n

i
n

m
a
s
s
A
-1.0
-0.8
-0.6
-0.4
-0.2
0.0
0 50 100 150 200 250 300
Figure 2: Deviation of the atomic masses from that expected from the sum of
proton and neutron masses.
B
E

(
M
e
V
)
A
0
500
1000
1500
2000
2500
0 50 100 150 200 250 300
Figure 3: Binding energy as a function of A.
1 NUCLEAR MASSES 9
B
E
/
A

(
M
e
V
)
A
0
2
4
6
8
10
0 50 100 150 200 250 300
Figure 4: BE/A as a function of A.
Measured BE/A
P
r
o
t
o
n

N
u
m
b
e
r
Neutron Number
0
20
40
60
80
100
120
0 20 40 60 80 100 120 140 160
<6.8
7.2
7.6
8.0
8.4
8.8 MeV
Figure 5: BE/A as a function of N and Z.
1 NUCLEAR MASSES 10
top of the ridge-of-stability in binding energy or those at the bottom of the valley-
of-stability in mass. For a xed A value nuclei away from the valley of stability beta
decay until they reach the bottom of the valley. Nuclear structure models are used
to understand the overall features of these data as well as the deviations from the
averge.
The maximum in the binding energy per nucleon occurs for
58
Fe.
58
Fe represents
the most bound (lowest energy) state for nucleons. Thus fusion of two light nuclei
with a combined mass of A < 58 usually results in energy release. The fusion of
deuterium and tritium is the main reaction being investigated for controlled fusion
reactors. Other fusion processes are important for solar energy and for the creation
of elements up to A = 58 in stellar environments. The fallo in binding energy per
nucleon above A = 58 implies that most of these nuclei can spontaneously decay into
lighter products. The most common of these decay processes are alpha decay, where a
4
He is emitted, and ssion, where the nucleus breaks up into two roughly equal mass
fragments. The ssion products are usually accompanied by neutrons. Intermediate
decay modes, where light fragments such as
14
C are emitted, are also possible and
have also been observed, but their decay rate relative to alpha decay is extremely
small. Although most heavy nuclei have a positive Q value for spontaneous decay,
many of them have lifetimes on the order of the age of the universe and thus exist in
nature, due to the hindrance of tunneling through the Coulomb barrier.
1.2 Q values and separation energies
In this section we consider energy conservation for nuclear transformations that in-
clude, for example, the fussion of two nuclei a and b into the combined system c
[N, Z]
a
+ [N, Z]
b
[N, Z]
c
(1.4)
or the decay of nucleus a into two other nuclei b and c
[N, Z]
a
[N, Z]
b
+ [N, Z]
c
(1.5)
In general we consider the combinations

i
[N, Z]
i

f
[N, Z]
f
(1.6)
where N and Z are conserved.

i
N
i
=

f
N
f
and

i
Z
i
=

f
Z
f
(1.7)
1 NUCLEAR MASSES 11
B
E
/
A

(
M
e
V
)
A
0
2
4
6
8
0 50 100 150 200 250 300
exp
B
E
/
A

(
M
e
V
)

L
D
M

m
o
d
e
l
0
2
4
6
8
10
12
14
16
18
A
B
C
D
Figure 6: Bottom panel: Experimental values for the BE/A as a function of
A for the most stable nuclei. Top panel: Liquid-drop model for BE/A (curve
D) and shown in terms its components the volume term (A), the volume plus
surface terms (B), the volume plus surface plus Coulomb terms (C).
1 NUCLEAR MASSES 12
B
E
/
A

(
M
e
V
)
A
5
6
7
8
9
10
0 10 20 30 40 50 60 70
Figure 7: An expanded portion of the experimental values for BE/A (points
connected by a line). The liquid-drop model is shown by the dashed line.
This process is characterized by the Q value:
Q =

i
M(N
i
, Z
i
)c
2

f
M(N
f
, Z
f
)c
2
=

f
B(N
f
, Z
f
)

i
B(N
i
, Z
i
). (1.8)
Spontaneous decay involves a single initial nuclear state and is allowed if Q > 0. In
the decay, energy is released in the form of the kinetic energy of the nal products.
Reactions involving two initial nuclei and are endothermic (a net loss of energy) if
Q < 0; the reactions are exothermic (a net release of energy) if Q > 0.
We can consider the Q values associated with the removal of one or two nucleons
from a nucleus. These are conventionally dened in terms of the one-nucleon and
two-nucleon separation energies, S:
S
n
= Q
n
= B(N, Z) B(N 1, Z), (1.9)
S
p
= Q
p
= B(N, Z) B(N, Z 1), (1.10)
S
2n
= Q
2n
= B(N, Z) B(N 2, Z), (1.11)
and
S
2p
= Q
2p
= B(N, Z) B(N, Z 2). (1.12)
1 NUCLEAR MASSES 13
Nuclear drip lines
P
r
o
t
o
n

N
u
m
b
e
r
Neutron Number
0
20
40
60
80
100
120
0 20 40 60 80 100 120 140 160
0
0.2
0.4
0.6
0.8
>1.0 MeV error
HF1 inside the drip lines
AW outside the drip lines
Figure 8: Nuclei in the Audi-Wapstra-Thibault mass compilation together with
the HF1 prediction for nuclei out to the neutron and proton drip lines. Observed
nuclei outside the drip lines are indicated.
A negative value for one of these quantities means that the nucleus can spontaneously
decay by the emission of neutrons or protons. The boundary between positive and
negative values of S is called the drip line. Nuclei inside the drip lines are stable
to the spontaneous emission of nucleons, whereas those outside the drip line can
spontaneously decay by emission of one and/or two nucleons. The neutron drip line
represents a sharp boundary between those nuclei just inside drip line that beta decay
with a lifetime on the order of ms, and those outside the drip line that have lifetimes
on the order of 10
20
s associated with the strong interaction decay widths of MeV. For
heavy nuclei, the Coulomb barrier can greatly hinder the proton decay, and the the
lifetimes for the one or two proton decay of nuclei just outside the proton-drip line can
be comparable to or even longer than the lifetimes for beta decay. Fig. (1.8) shows
the nuclear chart for nuclei whose masses have been measured together with those
that are predicted to lie within the neutron and proton drip lines from a microscopic
nuclear model called HF1 [2]. An expanded version of this gure for light nuclei is
shown in Fig. (1.9). HF1 is a Hartree-Fock(*)
5
calculation where on the order of 10
parameters of the Hamiltonian are tted to the experimental binding energies. These
gures also show nuclei with measured masses which lie outside the drip lines. For
light nuclei, the masses given in the Audi-Wapstra-Thibault compilation are usually
5
The (*) indicates an idea or topic that will be covered latter in the book.
1 NUCLEAR MASSES 14
Nuclear drip lines
P
r
o
t
o
n

N
u
m
b
e
r
Neutron Number
0
20
40
60
0 20 40 60
0
0.2
0.4
0.6
0.8
>1.0 MeV error
HF1 inside the drip lines
AW outside the drip lines
Figure 9: An expanded version of Fig. (1.8) for light nuclei.
1 NUCLEAR MASSES 15
Measured one-neutron separation energies
P
r
o
t
o
n

N
u
m
b
e
r
Neutron Number
0
20
40
60
80
100
120
0 20 40 60 80 100 120 140 160
0
2
4
6
8
10
>12 MeV separation energy
Figure 10: Measured one-neutron separation energies.
obtained from the observation of resonances in nuclear reactions. As such, it would
be better to call these quantities resonance energies rather than masses or binding
energies. In addition, there are several light nuclei not shown in Fig. (1.9) whose
non-observation in experimental studies implies that they lie outside of the drip lines
[3]. For light nuclei the proton and neutron drip lines are experimentally established
only up to about A = 24. For heavier nuclei the proton-drip line is observed in
a few regions from the observation of proton decays with relatively long lifetimes.
The HF1 model gives a prediction for the neutron and proton drip lines. Between
A = 40 and A = 200 nearly half of the nuclei expected to be inside the drip lines have
not yet been observed in experiments. The properties for many of these unobserved
nuclei are critical for the understanding of nuclear models as well as the astrophysical
processes in element production. New generations of accelerators such are RIA (the
Rare Isotope Accelerator project in the US) are being planned to produce and study
these nuclei.
The systematics of the one-neutron separation energies, S
n
, are shown in Fig.
(1.10) for experiment and in Fig. (1.11) for the HF1 model. As one moves from the
proton to the neutron drip lines, the one-neutron separation energies decrease. This
decrease is not smooth but shows odd-even oscillations associated with the two-body
pairing nature of the strong interaction between neutrons(*).
1 NUCLEAR MASSES 16
HF1 one-neutron separation energies
P
r
o
t
o
n

N
u
m
b
e
r
Neutron Number
0
20
40
60
80
100
120
0 20 40 60 80 100 120 140 160
0
2
4
6
8
10
>12 MeV separation energy
Figure 11: HF1 results for the one-neutron separation energies.
The S
n
values vs neutron number for the even-even nuclei with N > Z are shown
in Fig. (1.12). Just after the magic numbers 28, 50, 82 and 126 there is sudden
decrease in the separation energy due to the fact that neutrons go into valence shells
which are loosely bound compared to those which have just been lled at the magic
numbers(*).
This jump at the magic numbers can be emphasized by taking the dierences in
one-neutron separation energy:
S
n
= B(N, Z) B(N 1, Z) [B(N + 1, Z) B(N, Z)]
= 2B(N, Z) B(N 1, Z) B(N + 1, Z). (1.13)
The values of S
n
for the even-even nuclei (N, Z) are shown in the bottom panel
of Fig. (1.12). One observes clear peaks at the magic numbers 50, 82 and 126. The
magic numbers 8, 20 and 28 also appear as peaks, but the peak for 20 goes away for
Z = 10 and 12. The value of peak height is related to the size of the shell gaps(*).
Similar results are found for the proton separation energies.
1 NUCLEAR MASSES 17
d
i
f
f
e
r
e
n
c
e
s

o
f

S
_
n

(
M
e
V
)
N
0
2
4
6
8
0 50 100 150
S
_
n

(
M
e
V
)
0
5
10
15
Figure 12: The top panel shows the one-neutron separation energies as a func-
tion of neutron number for even-even nuclei with N > Z. The lines connect
nuclei with the same Z value. The bottom panel shows the dierences between
one-neutron separation energies for the same set of even-even nuclei. The dashed
lines show the magic numbers 8, 20, 28, 50, 82 and 126.
1 NUCLEAR MASSES 18
1.3 The liquid-drop model
Nuclei are bound due to the overall attractive strong interactions between nucleons.
The strong interaction arises from the exchange of mesons. The interactions are short
ranged and occur mainly between neighboring nucleons. In addition, the nuclear
interaction saturates, resulting in a nearly constant interior nucleon density and a
surface radius approximately equal to 1.2 A
1/3
. The analogy of this situation with a
droplet of liquid, results in the liquid-drop model for the nuclear binding energies in
which the binding energy is expressed in the form:
B(N, Z) =
1
A
2
A
2/3

3
Z
2
A
1/3

4
(N Z)
2
A
. (1.14)
The four terms on the right-hand side are referred to as the volume, surface, Coulomb
and symmetry energy terms, respectively. The rst term represents the nearest neigh-
bor attractive interaction between nucleons, and the second term represents the cor-
rection due to the fact that the nucleons on the surface only interact with those in the
interior. The third term is due to the Coulomb repulsion between protons. The fourth
term called the symmetry energy arises because the proton-neutron strong interaction
is on the average more attractive than the proton-proton or neutron-neutron strong
interactions and because the total kinetic energy is minimized when N = Z. The
constant
3
=0.697 MeV is xed by the Coulomb interaction and the nuclear size.
Typical values of the liquid-drop constants that reproduce the average trends in the
experimental data are given by
1
= 15.49 MeV,
2
= 17.23 MeV and
4
= 22.6
MeV.
The comparison of the liquid-drop model for the most stable nuclei is shown in
Fig. (1.6). The total is shown with by the line (D). Combinations of individual terms
are also shown: the volume term (A), the volume plus surface terms (B), the volume
plus surface plus Coulomb terms (C). The term symmetric nuclear matter refers to
extrapolation of properties of nuclei with N = Z and with no Coulomb interaction
to innite sized nuclei where the surface can be ignored. Thus the symmetric nuclear
matter extrapolation of the liquid-drop model used for this example gives BE/A=
15.5 MeV.
The size of the symmetry energy term determines how the binding energy de-
creases as we move away from stability. To isolate this term we can examine the chain
of nuclei for a xed A value as a function of Z. Experimental values B(N, Z)/A vs Z
are shown in Fig. (1.13) for A = 100. The form of this curve for A = 100 is similar to
those for other A values. The binding energy has a maximum at Z
max
= 44. With-
out the Coulomb interaction between protons, the maximum would occur at N = Z
due to the symmetry energy (the
4
term). When the Coulomb interaction between
protons (the
3
term) is added, the peak is shifted to more neutron-rich nuclei. The
1 NUCLEAR MASSES 19
m
a
s
s

e
x
c
e
s
s

(
M
e
V
/
c
^
2
)

f
o
r

A
=
1
0
0
Z
-100
-90
-80
-70
-60
35 40 45 50
B
E

(
M
e
V
)

f
o
r

A
=
1
0
0
820
830
840
850
860
870
Figure 13: Binding energy per nucleon, BE/A, and the mass excess, , for
A = 100 as a function of Z. The points connected by a line are the experimental
values and the liquid-drop model for the binding energy is shown by the dashed
line.
1 NUCLEAR MASSES 20
oscillation in the binding energy curve in Fig. (1.13) is due to nuclear pairing inter-
action(*), which gives rise to the fact that nuclei with even numbers of protons or
neutrons are more bound than their neighboring (odd) nuclei. The droplet model
would predict that Z = 43 has the highest binding, but the pairing changes this to
Z = 44.
The stability of nuclei to beta decay is determined by the dierences in the mass
excess shown in this example for A = 100 in the bottom panel of Fig. (1.13). From
Eq. (1.2), the dierences in the mass excesses are related to dierences in the binding
energies by [(N, Z) (N, Z 1)]c
2
= [B(N, Z) B(N, Z 1)] [
n

H
]c
2
where [(n) (H)]c
2
= 0.782 MeV. Thus, for example, for A = 100 we nd two
stable isotopes in nature,
100
Mo and
100
Ru. If we consider the very weak double-beta
decay process,
100
Mo will eventually decay into
100
Ru. This occurs with a lifetime
which is greater than the age of the universe, but such rare decay modes have been
detected in experiments.
To go further, we are interested in the deviations between the experimental and
liquid-drop models for the binding energies, and in understanding these deviations
in terms of microscopic models. The dierence between the experimental and liquid-
drop model binding energies are Fig. (1.14). The dierence between the HF1 model
and the liquid-drop model is shown in Fig. (1.15). In light nuclei the deviations are
largest near the neutron drip line. This deviation for light nuclei can be improved
by adding another term to Eq. (1.14) related to the surface symmetry energy [4].
In Fig. (1.14) there are two strong peaks at A = 132 and A = 208 which are the
doubly closed shell nuclei
132
Sn and
208
Pb. There is also a hint of a peak forming
near
100
Sn.
The values for the dierence between experiment and the liquid-drop model plot-
ted vs N and Z are shown in Figs. (1.16) and (1.17), respectively. One observes
several types of deviations. When the dierences are plotted vs proton number Z as
in Fig. (1.16), one observes that the nuclei with Z=28, 50 and 82 are more bound than
average. These are the shell-model magic numbers for the closed-shell congurations
at these Z values. For neutrons, Fig. (1.17), the same magic numbers are observed
with the addition of N = 126. The small oscillations observed in these gures is
due to nuclear pairing(*). This can be removed by plotting the results for only the
even-even nuclei as shown in Figs. (1.18) and (1.19).
1 NUCLEAR MASSES 21
BE(exp) - BE(liquid-drop)
P
r
o
t
o
n

N
u
m
b
e
r
Neutron Number
0
20
40
60
80
100
120
0 20 40 60 80 100 120 140 160
-5
0
5
10
15
20 MeV
Figure 14: The dierence between the experimental and liquid-drop binding
energies as a function of N and Z.
BE(HF1) - BE(liquid-drop)
P
r
o
t
o
n

N
u
m
b
e
r
Neutron Number
0
20
40
60
80
100
120
0 20 40 60 80 100 120 140 160
-5
0
5
10
15
20 MeV
Figure 15: The dierence between the HF1 and liquid-drop binding energies as
a function of N and Z.
1 NUCLEAR MASSES 22
B
E
(
e
x
p
)

-

B
E
(
l
i
q
u
i
d

d
r
o
p
)

(
M
e
V
)
Z
-10
-5
0
5
10
15
20
25
0 50 100
Figure 16: The dierence between the experimental and liquid-drop binding
energies as a function of Z. The dashed lines show the magic numbers 28, 50
and 82.
B
E
(
e
x
p
)

-

B
E
(
l
i
q
u
i
d

d
r
o
p
)

(
M
e
V
)
N
-10
-5
0
5
10
15
20
25
0 50 100 150 200
Figure 17: The dierence between the experimental and liquid-drop binding
energies as a function of N. The dashed lines show the magic numbers 28, 50,
82 and 126.
1 NUCLEAR MASSES 23
B
E
(
e
x
p
)

-

B
E
(
l
i
q
u
i
d

d
r
o
p
)

(
M
e
V
)
Z
-10
-5
0
5
10
15
20
25
0 50 100
Figure 18: Same as in Fig. (1.16) but only for even-even nuclei.
B
E
(
e
x
p
)

-

B
E
(
l
i
q
u
i
d

d
r
o
p
)

(
M
e
V
)
N
-10
-5
0
5
10
15
20
25
0 50 100 150 200
Figure 19: Same as in Fig. (1.17) but only for even-even nuclei.
REFERENCES 24
References
[1] The Ame2003 atomic mass evaluation (II), G. Audi, A. H. Wapstra and C.
Thibault Nuclear Physics A 729, 337 (2003).
[2] F. Tondeur, S. Goriely and J. M. Pearson, Phys. Rev. C 62, 024308 (2000); S.
Goriely, J. M. Pearson and F. Tondeur, At. Data and Nucl. Data Tables 77, 311
(2001).
[3] M. Thoennessen, Rep. Prog. Phys., to be published.
[4] P. Danielewicz, Nucl. Phys. A 727, 233 (2003).
2 RMS CHARGE RADII 25
2 Rms charge radii
The root-mean-square (rms) charge radius has been measured for the ground states
of many nuclei. For a spherical charge density, (r), the mean-square radius is dened
by:

< r
2
> =
_
(r)r
2
d
_
(r) d
, (2.1)
and the rms radius is the square root of this quantity. Experimental values have
been compiled in [1]. The rms charge radii for the 687 nuclei shown in Fig. (2.1) are
plotted vs A in Fig. (2.2). This selection of nuclei is a result of the variety of ways in
which the charge radii have been measured. The radii for most stable nuclei (those on
the solid line in Fig. (2.1)) have been deduced from electron scattering form factors
and/or from the x-ray transition energies of muonic atoms. The relative radii for a
series of isotopes can be extracted from the isotope shifts atomic x-ray transitions.
The charge density of the nucleus,
ch
, with rms radius

R
ch
is given by the point
proton density in the nucleus,
p
, with rms radius

R
p
,convoluted with the charge
density of the proton,

R
o
, with rms radius

R
o
= 0.88 fm. Thus we have

R
ch
=
_

R
2
p
+

R
2
o
. (2.2)
and

R
p
=
_

R
2
ch

R
2
o
. (2.3)
This last equation is used to obtain the rms radii for point protons shown in Fig.
(2.3).
In the spherical liquid-drop model the density is given by:
(r) =
o
, for r < R = r
o
A
1/3
, and
(r) = 0, for r > R. (2.4)
The rms radius for this sharp-surface distribution is given by

R
d
=

3
5
R =

3
5
(r
o
A
1/3
). (2.5)
The liquid-drop model with r
o
= 1.185 fm is shown by the dashed line in Fig. (2.3).
The value of r
o
= 1.185 fm is chosen so that the line passes through the lower values of
the rms radii for heavy nuclei. (For the charge radii in Fig. (2.2) one needs r
o
= 1.20
fm.) It is observed that the data follow the liquid-drop line rather closely except for
light nuclei and some regions of heavy nuclei where the data is higher than the model.
2 RMS CHARGE RADII 26
The distribution of neutrons is not known so well but we will start out by assuming
that the rms radius for neutrons is the same as that for protons. In the liquid-drop
model the density for nucleons in the interior of the nucleus saturates to a value
o
.
We improve the sharp-surface model if we allow the surface to be diuse. The
typical way to introduce the diuseness is with the Fermi function shape:
(r) =

o
1 + exp[(r R)/a]
. (2.6)
The rms radius is given by:

R
f
=

3
5

R
2
+
7
3

2
a
2
(2.7)
The t to the data obtained with r
o
= 1.15 fm and a = 0.35 fm is shown by the solid
line in Fig. (2.3). This is a better overall t to light and heavy nuclei, but the data
still uctuate around the curve. In the Fermi distribution model for A nucleons the
interior density is given by

o
=
A
4
3
R
3
(1 +

2
a
2
R
2
)
. (2.8)
For large A we can neglect the

2
a
2
R
2
term, and if we assume that the rms radius for
neutrons is the same as that for the protons then R
3
= r
3
o
A and
o
=
3
4r
3
o
= 0.16
nucleons/fm
3
with r
o
= 1.15 fm. This is approximately the saturation density of sym-
metric nuclear matter. The assumption of equal rms radii for protons and neutrons
means that the interior density will be divided between protons and neutrons in the
ratio Z/N. So, for example, in
208
Pb, the interior nucleon density of 0.16 nucleons/fm
3
will be divided into (82/208)0.16 = 0.06 protons/fm
3
and (126/208)0.16 = 0.10
neutrons/fm
3
.
In order to bring out the dierences in the rms radii I show the experimental
values plotted vs neutron number in Fig. (2.4) and the dierence

R
exp

R
f
in Fig.
(2.5). One observes that the experimental rms radii are relatively small at the magic
numbers. This means that the nuclei associated with the magic numbers are more
compact than those away from the magic numbers. This increase in rms radii away
from the magic numbers is qualitatively related to nuclear deformation(*).
2 RMS CHARGE RADII 27
P r o t o n n u m b e r ( Z )
N
e
u
t
r
o
n

n
u
m
b
e
r

(
N
)
0
5
0
1
0
0
0
5
0
1
0
0
1
5
0
Figure 1: Nuclei with measured rms charge radii. There are a total of 687. The
dashed lines are for N = Z, and the magic numbers 28, 50, 82 and 126. the
solid line is drawn through those nuclei with the largest binding energy for a
given A value (the most stable nuclei).
2 RMS CHARGE RADII 28
r
m
s

c
h
a
r
g
e

r
a
d
i
u
s

(
f
m
)
A
2
4
6
0 50 100 150 200 250 300
Figure 2: The rms charge radii for 687 nuclei plotted as a function of the
atomic number A. The dashed line is the liquid-drop model with a sharp surface
R = r
o
A
1/3
with r
o
= 1.20 fm.
r
m
s

p
r
o
t
o
n

r
a
d
i
u
s

(
f
m
)
A
2
4
6
0 50 100 150 200 250 300
Figure 3: The rms proton radii for 687 nuclei plotted as a function of the
atomic number A. The dashed line is the liquid-drop model with a sharp surface
R = r
o
A
1/3
with r
o
= 1.185 fm. The solid line uses the form of Eq. (2.7) which
takes into account the diuseness.
2 RMS CHARGE RADII 29
r
m
s

p
r
o
t
o
n

r
a
d
i
u
s

(
f
m
)
N
2
4
6
0 50 100 150
Figure 4: The rms proton radii for 687 nuclei plotted as a function of the neutron
number N.
d
i
f
f
e
r
e
n
c
e

i
n

r
m
s

p
r
o
t
o
n

r
a
d
i
u
s

(
f
m
)
N
-0.2
-0.1
0.0
0.1
0.2
0 50 100 150
Figure 5: The rms proton radii for 687 nuclei with the diuse liquid-drop model
values of Eq. (2.7) subtracted. The magic numbers 28, 50, 82 and 126 are
indicated by the dashed lines.
REFERENCES 30
References
[1] H. de Vries, C. W. de Jager and C. de Vries, At. Data and Nucl. Data Tables 36,
495 (1987); G. Fricke, C. Bernhardt, K. Heilig, L. A. Schaller, L. Shellenberg, E.
B. Shera and C. W. de Jager, At. Data and Nucl. Data Sheets 60, 177 (1995).
3 CHARGE DENSITIES AND FORM FACTORS 31
3 Charge densities and form factors
Elastic electron scattering from nuclei has provided a great resource of experimen-
tal data with which to test models for nuclear ground states. It evolved from the
early determinations of rms charge radii, to much more precise measurements in the
70s and 80s which have provided nearly model-independent determinations of the
charge-density distributions of many nuclei. In these density distributions one can
observe oscillations in the interior density which represent the quantum waves in
the nucleus.
The charge-density normalization is given by:
_
(r)
ch
d = 4
_
(r)
ch
r
2
dr = Z, (3.1)
where Z is the number of protons in the nucleus. The charge probability density,
P(r)
ch
= 4r
2
(r)
ch
, (3.2)
represents the probability to nd Z protons at a given radius r from the center of the
nucleus.
As an example, the charge density measured for
208
Pb is shown in Fig. (3.1). The
charge density is shown it the top panel and the probabilitly density in the lower panel.
If one were to put the 208 nucleons of
208
Pb into a simple cubic lattice, a density of
0.16 nucleons/fm
2
corresponds to a lattice spacing of 1.85 fm. I have respresented this
situation in the upper panel of Fig. (3.1) by drawing Gaussian distributions for several
nucleons each of which has an rms radius of 0.88 fm and which are spaced a distance
of 1.85 fm. One observes in this situation that the overlap between nucleons is small.
(In nuclear models the nucleons are not conned to lattice sites but are described by
wave functions spread over the whole nuclear volume. Nucleons conned to lattice
sites would have zero-point kinetic energies that are an order of magnitude larger
than those allowed to spread over the nuclear volume.)
The ratio of the electron scattering cross section to the Rutherford cross section
(scattering from a point) as a function of momentum transfer, q, is related to the
plane-wave Fourier transform of the charge density:
F(q) =
1
Z
_
(r)
ch
e
i q
.
r
d =
4
Z
_
(r)
ch
j
o
(qr) r
2
dr, (3.3)
where
j
o
(qr) =
sin(qr)
qr
. (3.4)
The normalization in Eq. (3.3) is chosen to given F(q = 0) = 1. For light nuclei the
electron energy distortion is small, and the cross section is closely proportional to the
3 CHARGE DENSITIES AND FORM FACTORS 32
c
h
a
r
g
e

p
r
o
b
a
b
i
l
i
t
y

d
i
s
t
r
i
b
u
t
i
o
n
r (fm)
0
5
10
15
20
0 2 4 6 8 10
c
h
a
r
g
e

d
e
n
s
i
t
y
0.0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.1
Figure 1: Experimental charge-density prole (upper panel) and probability
prole (lower panel) for
208
Pb. The upper panel also shows the density proles
for some individual nucleons with an rms radius of 0.88 fm and spaced a distance
of 1.85 fm.
3 CHARGE DENSITIES AND FORM FACTORS 33
form factor [ F(q) [
2
which has minima corresponding to the zeros of F(q). For heavy
nuclei the electron energy distortion is larger, and the minima in the cross sections are
washed out. From a DWBA analysis of the electron scattering cross section one can
extract F(q) with good precision over the range of momentum transfers measured.
From the measured form factors one can obtain the charge-density with the in-
verse of Eq. (3.3):
(r)
ch
=
Z
2
2
_
F(q) j
o
(qr) q
2
dq. (3.5)
Since the form factor data are only determined up to some q
max
, the resulting charge-
density has some error band which depends upon the assumptions made about
F(q) for q > q
max
.
A common method for extracting the charge density from the electron scattering
form factors is to make a Fourier-Bessel expansion t [1] to the data. The Fourier-
Bessel expansion is given by:
(r)
ch
=
nmax

=1
a

j
o
(r/R), (3.6)
for r R and (r)
ch
= 0 for r > R, with the associated plane-wave transform:
F(q) =
4
Z
(qR)
2
q
3
nmax

=1

[j
o
( qR) j
o
( +qR)]. (3.7)
The parameters a

,
max
and R for a wide range of nuclei are given in Refs. [2] and
[3].
In order to obtain a qualitative understanding of how the form factors depend
upon the properties of the charge density, I show in Figs. (3.2) and (3.3) some results
for
208
Pb (Z = 82) based upon the Fermi distribution:
(r)
ch
=

o
1 + exp[(r R)/a]
(3.8)
The densities are shown in the upper panels and the associated form factors [ F(q) [
2
are shown in the lower panels.
In Fig. (3.2) I have xed the diuseness to be a = 0.4 fm and then varied R
to make the rms charge radius 5.30 fm (solid line), 5.50 fm (dashed line) and 5.70
fm (crosses). In Fig. (3.3) I have varied the diuseness from a=0.4 (solid line), 0.5
(dashed line) and 0.6 (crosses) but with R chosen to give a xed rms of 5.50 fm.
Note that an increase in the diuseness is associated with an increase in the
interior density and a large decrease in the form factor maxima at large q values. The
3 CHARGE DENSITIES AND FORM FACTORS 34
d
e
n
s
i
t
y
r (fm)
0.0
0.02
0.04
0.06
0.08
0 2 4 6 8 10
f
o
r
m

f
a
c
t
o
r
q (1/fm)
10
-9
10
-8
10
-7
10
-6
10
-5
10
-4
10
-3
10
-2
10
-1
10
0
0 1 2 3
Figure 2: Features of the fermi-function densities and the associated form factors
when the diuseness is constrained to a = 0.4 fm and the R is used to vary the
rms radius to be 5.30 fm (solid line), 5.50 (dashed line) fm and 5.70 fm (crosses).
3 CHARGE DENSITIES AND FORM FACTORS 35
d
e
n
s
i
t
y
r (fm)
0.0
0.02
0.04
0.06
0.08
0 2 4 6 8 10
f
o
r
m

f
a
c
t
o
r
q (1/fm)
10
-9
10
-8
10
-7
10
-6
10
-5
10
-4
10
-3
10
-2
10
-1
10
0
0 1 2 3
Figure 3: Features of the fermi-function densities and the associated form factors
when the rms radius is constrained to be 5.5 fm. The results are given for
diuseness values a = 0.4 fm (solid line), 0.5 fm (dashed line) and 0.6 fm
(crosses).
3 CHARGE DENSITIES AND FORM FACTORS 36
C
h
a
r
g
e

f
o
r
m

f
a
c
t
o
r
q (1/fm)
10
-9
10
-8
10
-7
10
-6
10
-5
10
-4
10
-3
10
-2
10
-1
10
0
0 1 2 3
C
h
a
r
g
e

f
o
r
m

f
a
c
t
o
r
10
-9
10
-8
10
-7
10
-6
10
-5
10
-4
10
-3
10
-2
10
-1
10
0
10
1
Figure 4: Experimental form factors for
16
O (solid line in the lower panel),
40
Ca
(dashed line in the lower panel),
92
Mo (solid line in the upper panel) and
208
Pb
(dashed line in the upper panel). The lines are plotted over the range of q values
measured.
3 CHARGE DENSITIES AND FORM FACTORS 37
c
h
a
r
g
e

d
e
n
s
i
t
y
r (fm)
0.0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.1
0 2 4 6 8 10
16O 40Ca 92Mo 208Pb
Figure 5: Experimental charge-density distributions for
16
O,
40
Ca,
92
Mo and
208
Pb. The results for the analysis of two dierent data sets for
208
Pb are shown
by the circles and crosses.
3 CHARGE DENSITIES AND FORM FACTORS 38
qualitative eects of oscillations in the densities on the form factors are discussed in
Ref. [4].
The form factors which have been measured for
16
O,
40
Ca,
92
Mo and
208
Pb are
shown in (3.4). The charge densities which have been extracted from these data are
shown in (3.5). One observes various oscillations in the charge density which is related
to the shell structure of nuclei. The surface gradually extends out for heavier nuclei
but the diuseness is rather constant as assumed in the Fermi distribution model.
The experimental errors are not given in Refs. [2] and [3], and I do not attempt
to show an experimental error band for the charge densities, the size of which is
strongly correlated with q
max
. For
208
Pb in Fig. (3.5) the density extracted from two
dierence sets of set are used to give an indication of the error. General aspects of
the statistical and model-dependent errors in these Fourier-Bessel ts are discussed
by Dreher et al., [1] and those related to specic experiments are discussed in the
original experimental papers. Generally the width of the error band increases toward
the center of the distribution (where the number of protons is small). For example
for
92
Mo [5] the statistical and model-dependent error is 1.5% at r = 2 fm and goes
down to 0.6% (about the size of the circles used to represent the experimental data
in the gures to be discussed below) for r = 4 fm.
REFERENCES 39
References
[1] B. Dreher, J. Friedrich, K. Merle, H. Rothhaas and G. Luhrs, Nucl. Phys. A235,
219 (1974).
[2] H. de Vries, C. W. de Jager and C. de Vries, At. Data and Nucl. Data Tables 36,
495 (1987).
[3] G. Fricke, C. Bernhardt, K. Heilig, L. A. Schaller, L. Shellenberg, E. B. Shera and
C. W. de Jager, At. Data and Nucl. Data Sheets 60, 177 (1995).
[4] P. G. Reinhard, J. Friedrich and N. Voegler, Z. Phys. A 316, 207 (1984).
[5] B. Dreher, Phys. Rev. Lett 35, 716 (1975)
4 OVERVIEW OF NUCLEAR DECAYS 40
4 Overview of nuclear decays
In 1896 Antoine Henri Becquerel found radiations which were spontaneously emitted
from uranium salts. In 1898 Maria and Pierre Curie found new elements latter called
Polonium and Radium that also spontaneously emitted radiations. These radiations
were observed to be bent by magnetic eld, and were named alpha () and beta ()
by Ernest Rutherford in 1899 after the observations that one was easily absorbed
(alpha) and one was more penetrating (beta). In 1900, Villard identied a third form
of penetrating radiation which could not be bent by a magnetic eld, and these were
called gamma () radiations.
In 1909, Ernest Rutherford and Thomas Royds established that alpha particles
were the nuclei of helium atoms and have atomic mass number A = 4 and nuclear
charge number Z = 2. Antoine Henri Becquerel (1899) and Walter Kaufmann (1902)
identied the beta radiation with the electron by its deection in electric and magnetic
elds. The basic theory for beta decay was constructed by Enrico Fermi in 1934 after
Wolfgang Paulis suggestion in the same year about the existence of the neutrino.
Gamma radiation was soon realized to be a high-energy form of the electromag-
netic radiation described by Maxwells equations. Today gamma radiation generally
refers to the high-energy region of the electromagnetic-radiation spectrum associated
with the decay of particles and nuclei. Gamma-ray energies range from a few keV
(where their energy range overlaps with those of X rays that are emitted in the decay
of atoms and molecules) up to 10
2
TeV (10
14
eV) for those found in cosmic rays. The
gamma ray, like the X ray, consists of photons that have no mass and no charge.
A group led by Lise Meitner and including Otto Hahn and Fritz Strassmann
began studying the products formed when uranium is bombarded by neutrons in the
1930s which led to the discovery of ssion in 1938-39. Fission is the induced or
spontaneous breakup of a heavy nucleus into two nuclei each of which has about one
half of the mass of the parent. The breakup is usually accompanied by the emission
of one or more neutrons.
Other forms of ground-state radioactivity we may consider are proton decay dis-
covered in 1981 [1], light cluster emission,
223
Ra
209
Bi +
14
C discovered in 1984
[2], two-neutrino double-beta decay discovered in 1987 [3], and two-proton decay dis-
covered in 2002 [4].
Since 1981 a many more cases of one-proton emission have been observed [5].
Also several other cases of light cluster emission have been observed including the
emission of
20
O,
24
Ne,
26
Ne,
28
Mg,
30
Mg,
32
Si and
34
Si [6].
4 OVERVIEW OF NUCLEAR DECAYS 41
Nuclear decays are characterized by their Q value and half-life. The lifetimes
we consider vary over an extremely wide range of magnitudes. In terms of x =
Log
10
[T
1/2
(sec)] it is useful to keep in mind the values of x 37 for the proton
lifetime limit, x 17 for the age of the universe (10 billion years), x 5 for one
day, x 7 (100 nsec) for the typical time it take to analyze a secondary beam from
the cyclotron, and x 21 for the lifetime corresponding to a resonance width of 1
MeV.
4.1 Decay widths and lifetimes
The decay of an emsemble of quantum states is descibed by the rate equation for the
number of nuclei present at time t.
dN
dt
= WN(t), (4.1)
where W is the transition rate or decay constant. The solution of this equation is the
exponential decay law
N(t) = N
o
e
Wt
. (4.2)
The time at which the number is redunced by half is the half-life
T
1/2
=
ln2
W
=
0.693
W
. (4.3)
The mean-lifetime is the average amount of time it takes to decay
=
_
te
Wt
dt
_
e
Wt
dt
=
1
W
=
T
1/2
ln2
=
T
1/2
0.693
. (4.4)
The time dependence of a decaying wavefunction involves both real and imaginary
parts
(t) = e
itWt/2
. (4.5)
where the average anergy is E
o
= h. The square of the Fourier transform of Eq.
(4.5) gives the probability to nd the state in energy E
P(E) =
1
4
2
1
(E E
o
)
2
+ ( hW/2)
2
. (4.6)
The value of E E
o
at which P(E) falls by a factor of two is E E
o
= hW/2. Thus
the full-width at half maximum denoted by is
= hW = h/. (4.7)
4 OVERVIEW OF NUCLEAR DECAYS 42
The Heisenberg uncertainty relation in terms of and is
= h = 6.58 10
22
MeV s = 6.58 10
4
eV ps. (4.8)
In nuclear physics we may describe the decay of a state in terms of its lifetime in units
of s, ms (10
3
), s (10
6
), ns (10
9
), ps (10
12
) or fs (10
15
), or alternatively with
Eq. (4.8) in terms of its width in MeV. Hitorically one usually the mean lifetime
for gamma decay of excited-state lifetimes and the half-life T
1/2
= ln(2) for ground-
state lifetimes for beta decay and alpha decay. (Note that the Table of Isotopes by
Firestone et al. uses T
1/2
for excited states.)
A given initial state may decay to several nal states. The total transition rate
is:
W =

f
W
i,f
, (4.9)
where W
i,f
is the partial decay rate to the particular nal state f. The branching
fraction to this state is:
6
b(i f) =
W
i,f
W
. (4.10)
When the total lifetime and the branching fraction for a given decay are known, we
can nd the partial lifetime
p
related to that specic decay channel by:

p
= /b. (4.11)
4.2 Alpha and cluster decay
Alpha decay occurs when a parent nucleus (A, Z) with atomic mass number A and
nuclear charge number Z spontaneously emits an alpha particle leaving a residual
(daughter) nucleus (A4, Z 2):
A
Z
(A4)
(Z 2) +
4
He. (4.12)
Alpha decay usually occurs from the nuclear ground state, but decay from excited
states may also occur. The alpha decay of a given parent nucleus often leads to
daughter nuclei that are themselves alpha or beta radioactive, thus giving rise to a
disintegration series. By 1935 the detailed decay schemes for three naturally occurring
series that started with
238
U (Z=92),
235
U, and
232
Th (Z=90) had been discovered.
The alpha particles observed for these naturally occurring decays have energies in the
range of 5 to 10 MeV, and were used as a source of projectiles for nuclear reaction
experiments until the use of particle accelerators took over in the 1940s.
6
The branching fraction is often denoted by BR and is given in percent.
4 OVERVIEW OF NUCLEAR DECAYS 43
Alpha decay Q values
P
r
o
t
o
n

N
u
m
b
e
r
Neutron Number
0
20
40
60
80
100
120
0 20 40 60 80 100 120 140 160
<0
2
4
6
8
10
12 MeV
Figure 1: Experimental values for the alpha decay Q values.
Q

(
M
e
V
)
N
-15
-10
-5
0
5
10
15
0 50 100 150
Figure 2: Experimental values for the alpha decay Q values.
4 OVERVIEW OF NUCLEAR DECAYS 44
Spontaneous alpha decay is allowed when the Q

value for the decay is positive.

The energy of the emitted alpha particle is given by
E

= Q

M
d
/(M
d
+M

), (4.13)
where M
d
and M

are the masses of the daughter nucleus and alpha particle, respec-
tively, and the Q

is given in terms of binding energies B by:

Q

= B(N 2, Z 2) +B(2, 2) B(N, Z), (4.14)

were B(2, 2) = 28.296 MeV. The values obtained with the experimental binding
energies are shown in Figs. (4.1) and (4.2). The alpha-decay Q

value becomes
positive above Z 50 and, generally, for nuclei that are proton-rich compared to the
most stable. (An exceptional case is that of
8
Be for which the Q

value is positive
for the decay into two alpha particles.) Alpha decay is more important than other
light element emissions because of the relatively large binding energy of the alpha
particle together with its small Z value. The alpha decay of nuclear ground states
competes with beta decay and ssion decay. The alpha decay of nuclear excited states
competes, in addition, with gamma decay. Man-made heavy elements are often most
easily identied by their characteristic alpha decay series. For example, the element
with Z = 110 discovered in 1994 at GSI was identied from the detection of an alpha
particles with an energy of 11.13 MeV and a half-life of about 400 s in coincidence
with the alpha particle from the decay of the Z = 108 daughter, whose properties
had been studied in previous experiments.
The basic theory for alpha decay was developed by George Gamow and others in
1930. One postulates an alpha particle moving in the potential well of an attractive
strong interaction. The potential energy diagram for the
238
U decay is shown in Fig.
(4.3) The radius of the strong interaction potential, R
t
, is determined by the distance
between the centers when the surfaces of the alpha particle and the daughter nucleus
touch. This is given approximately by R
t
= R
d
+ R

, where R
d
= 1.2A
1/3
fm is the
radius the daughter nucleus and R

= 2.15 fm for the alpha particle. For illustrative

purposes the magnitude of the potential inside R
t
has been set to zero. The dashed
line shows the Q

value energy of 4.27 MeV. Beyond the distance R

t
the interaction
between the alpha particle and the daughter nucleus is determined by the repulsive
Coulomb potential V (r) = 2Z
d
e
2
/r, where Z
d
is the nuclear charge number of the
daughter nucleus.
The semi-classical picture developed by Gamow envisions an alpha particle mov-
ing back and forth classically inside the radius R
t
and hitting the potential barrier
at R
t
with a decay rate W
c
= v/(2R
t
) =
_
Q

/(2R
2
t
) where is the reduced mass
M
d
M

/(M
d
+ M

), and we estimate the velocity v by Q

=
1
2
v
2
. When the alpha
particle hits the potential barrier it has a probability, P, in the theory of quantum
4 OVERVIEW OF NUCLEAR DECAYS 45
V
(
r
)

(
M
e
V
)
r (fm)
-10
0
10
20
30
0 20 40 60 80 100 120
Figure 3: The potential energy diagram for the alpha decay of
238
U. The solid
line is the Coulomb barrier to the radius R
t
. the dashed line is the Q

value.
mechanics to tunnel though the barrier given by:
P = exp2
_
Rc
Rt
_
2[V (r) Q

]/ h
2
dr. (4.15)
The radius, R
c
= 2Z
d
e
2
/Q

, at which V (r) = Q

is referred to as the classical turning

radius, since classically an alpha particle approaching the daughter nucleus from a
large radius cannot go beyond this point. Integration gives
P = exp4Z
d
e
2
_
2/(Q

h
2
) [cos
1
(x) x

1 x
2
], (4.16)
where x =
_
R
t
/R
c
. The total alpha decay rate is W = W
c
P and the alpha-decay
half-life is given by
T
1/2
=
ln2
(W
c
P)
. (4.17)
The experimental half-lifes for the alpha decays of 119 (even-even) nuclei with
ground state angular momenta of J = 0 are compared to the Gamow estimate in
Fig. (4.4) One observes qualitative agreement with the Gamow estimate over 25
orders of magnitude in the half-life with a systematic trend for the experimental
half-lifes to be about 10-100 times longer than predicted. This hindrance can be
qualitatively interpreted as the probability that the actual many-body wave function
4 OVERVIEW OF NUCLEAR DECAYS 46
E
x
p
e
r
i
m
e
n
t
Theory
-10
0
10
20
-10 0 10 20
Figure 4: Comparison of experimental and theoretical values for log
10
[T
1/2
(sec)].
of the parent nucleus has an overlap with the wave function representing the alpha
particle plus daughter nucleus. The structure dependence in this hindrance factor can
be seen in Fig. (4.5), where the ratio of the experimental to theoretical decay rates
are shown. One observes a reduction in the experimental rate as one approaches the
magic number 126 and then a sudden jump up.
From the form of Eq. (4.16) one obtains with some approximations the Gieger-
Nuttall relation for alpha decay, Log
10
[T
1/2
] Z
d
/

. However a more accurate

empirical relation [7] which is able to reproduce the alpha decay half-lifes to within
about a factor of three is
log
10
[T
1/2
(sec)] = 51.37 + 9.54Z
0.6
d
/
_
Q

. (4.18)
where Z
d
is the charge of the daughter nucleus and Q

is in units of MeV. Generally

equation (4.18) has the same accuracy as Eq. (4.17), and neither equation contains
the structure dependence observed in Fig. (4.5).
Eq. (4.18) has been used to calculate the alpha-decay half-life for those nuclei in
Fig. (4.1). The results for all nuclei are shown in Fig. (4.6). Results based on the
HF1 model are shown (4.7). One observes that alpha decay is responsible for the
instability of nuclei above
208
Pb except for a small region of nuclei around
232
Th and
234238
U.
The Gamow model can be extended to the decay of non-zero spin nuclei by
4 OVERVIEW OF NUCLEAR DECAYS 47
A
l
p
h
a

d
e
c
a
y

r
a
t
e

(
e
x
p
/
t
h
)
N
0.0
0.1
0.2
80 90 100 110 120 130 140 150 160
P
r
o
t
o
n

n
u
m
b
e
r

Z
70
80
90
100
110
Figure 5: Ratio of the experimental to theoretical alpha decay rates (W) plotted
as a function of neutron number. The top part of the gure shows the nuclei
which are considered in the bottom part with the magic numbers Z = 82 and
N = 126 shown by the dashed lines.
4 OVERVIEW OF NUCLEAR DECAYS 48
Alpha decay Q values
P
r
o
t
o
n

N
u
m
b
e
r
Neutron Number
0
20
40
60
80
100
120
0 20 40 60 80 100 120 140 160
stable
x > 15
15 > x > 10
10 > x > 5
5 > x > 1
1 > x > -1
-1 > x > -3
-3 > x
Figure 6: Nuclei that are unstable to alpha decay. The symbols represent
dierent ranges of x = log
10
[T
1/2
(sec)] as calculated with equation (4.18) with
the experimental Q

values.
4 OVERVIEW OF NUCLEAR DECAYS 49
HF1 Alpha decay x values
P
r
o
t
o
n

N
u
m
b
e
r
Neutron Number
0
20
40
60
80
100
120
0 20 40 60 80 100 120 140 160
stable
x > 15
15 > x > 10
10 > x > 5
5 > x > 1
1 > x > -1
-1 > x > -3
-3 > x
Figure 7: Nuclei that are unstable to alpha decay. The symbols represent
dierent ranges of x = log
10
[T
1/2
(sec)] as calculated with equation (4.18) with
the HF1 Q

values.
4 OVERVIEW OF NUCLEAR DECAYS 50
14C decay Q values
P
r
o
t
o
n

N
u
m
b
e
r
Neutron Number
0
20
40
60
80
100
120
0 20 40 60 80 100 120 140 160
0
5
10
15
20
25
30
35 MeV
Figure 8: Experimental values for
14
C decay Q values.
Q

(
M
e
V
)
N
-30
-20
-10
0
10
20
30
40
0 50 100 150
Figure 9: Experimental values for
14
C decay Q values.
4 OVERVIEW OF NUCLEAR DECAYS 51
adding a centrifugal barrier to the potential V (r) corresponding to the relative angular
momentum L between the daughter nucleus and the emitted nucleus:
V
L
=
L(L + 1) h
2
2r
2
The model can be rened by considering a more realistic shape for the interior and sur-
face region of the potential. A more quantitative understanding of alpha decay relies
on using many-body wave functions for all three particles involved in the decay that
incorporate the individual proton and neutron degrees of freedom. The experimen-
tal data as well as the microscopic calculations indicate deviations from the Gamow
model that are related to change in shape and change in shell-structure between nu-
clei. The basic idea behind the Gamow model can also be used to calculate lifetimes
for proton and cluster decay [8]. The Q value systematics is the essential ingredient
which determined the existence and half-lives of these charged particle decay. For
example I show in Figs. (4.8) and (4.9) the Q values for
14
C decay. The large peak
in the Q values just above
208
Pb is what determines the the region of nuclei which
have been observed to have a
14
C decay branch, although even in the most likely case
cluster emission is extremely small ( 10
12
or less) compared to alpha decay.
4.3 Beta decay
The most elementary form of nuclear beta decay is that of the neutron into a proton,
an electron, and an electron antineutrino:
n p +e

+
e
. (4.19)
Nuclei are composed of protons and neutrons bound together by the strong in-
teraction. In the beta decay of nuclei, a given initial nuclear state
A
i
Z
i
is converted
into the ground state or an excited state of the nal nucleus
A
f
Z
f
, where Z
f
= Z
i
1.
The transition rate for nuclear beta decay is determined by the Q

value or energy
release and the structure of the initial and nal nuclear states.
Beta decays with the fastest rates occur when the leptons carry away =0 angular
momentum and are referred to as allowed transitions. Decays with >0 for the
leptions are referred to as forbidden transitions. The dependence upon the energy
release can usually be calculated to a precision of about 0.1 percent, and beta decay
thus provides a precise test of the strength of the weak interaction, as well as of the
internal structure of particles and nuclei. In the limit when Z is small and Q

is
large, the transition rate for allowed beta transitions is proportional to Q
5

.
In 1956 Tsung Dao Lee and Chen Ning Yang suggested that beta decay should
violate the principle of parity nonconservation, and they proposed an experiment to
4 OVERVIEW OF NUCLEAR DECAYS 52
test this idea. In 1957 parity nonconservation was conrmed by experiments carried
out by Wu, Ambler, Hayward, Hoppes and Hudson on the beta decay of
60
Co.
The modern theory of beta decay is based upon the Standard Model which uni-
es the weak and electromagnetic interactions. The Standard Model of beta decay
involves the W

bosons at an intermediate stage of the decay process. The most

elementary of these processes involved in nuclear beta decays are:
d u +W

u +e

+
e
, (4.20)
and
u d +W
+
d +e
+
+
e
, (4.21)
where u and d are the up and down quarks, respectively. These transformations
are examples of a larger class of transformations that involve all quarks and leptons.
Each step in the elementary decay process is proportional to the weak-interaction
coupling constant g. The Standard Model relates the value of g to the mass of the
W boson and value of the electric charge e. Also in the standard model, beta decay
is unied with a larger class of weak interaction processes that involve the Z boson
as an intermediate particle.
4.3.1 Beta decay Q values
Beta minus,

, decay involves the emission of an electron and electron antineutrino:

A
Z
A
(Z + 1) +e

+
e
. (4.22)
The Q value for

decay is given in terms of nuclear masses M and nuclear binding

energies BV by
Q(

) = [M(A, Z) M
1
(A, Z + 1) m
e
]c
2
= [M(A, Z) M(A, Z + 1)]c
2
= B(A, Z + 1) B(N, Z) +
nH
, (4.23)
where m
e
c
2
= 0.511 MeV is the mass of the electron,
M
1
(A, Z + 1) = M(A, Z + 1) m
e
(4.24)
is the mass of the nal nucleus with one electron missing, and

nH
=
n
c
2

H
c
2
= 0.782 MeV (4.25)
comes from the mass dierence between the neutron and the Hydrogen atom. In
these expressions we assume that the mass of the neutrino is zero and we ignore the
electronic binding energy.
4 OVERVIEW OF NUCLEAR DECAYS 53
Beta plus,
+
, decay involves emission of a positron and electron neutrino:
A
Z
A
(Z 1) +e
+
+
e
, (4.26)
The Q value for
+
decay is given by
Q(
+
) = [M(A, Z) M
+1
(A, Z 1) m
e
]c
2
= [M(A, Z) M(A, Z 1) 2m
e
]c
2
= B(A, Z 1) B(A, Z) 2m
e
c
2

nH
. (4.27)
where
M
+1
(A, Z 1) = M(A, Z 1) +m
e
(4.28)
is the mass of the nal nucleus with one extra electron.
Another form of beta decay that competes with
+
decay is electron capture
(EC) in which one of the atomic electrons is captured by the nucleus and an electron-
neutrino is emitted:
e

+
A
Z
A
(Z 1) +
e
. (4.29)
The Q value for electron capture decay is given by
Q(EC) = [M(A, Z) M(A, Z 1)]c
2
= B(A, Z 1) B(A, Z)
nH
. (4.30)
The experimental beta decay Q values are shown in Fig. (4.10) and those for the
HF1 predictions are shown in (4.11). The energy released in

or
+
decay is shared
between the recoiling nucleus, the electron and the neutrino. Since the nucleus is
heavy compared to the electron and neutrino, most of the energy is shared between
the electron and the neutrino with a probability distribution for each that can be be
accurately calculated. Usually only the electron or positron is detected, and it has a
range of kinetic energies ranging from zero up to Q

(the end-point energy), assuming

that the mass of the neutrino is zero. If the neutrino has a mass, the end-point energy
of the electron would be reduced. The end-point energy of the tritium beta decay
has been used to set a limit of about m

< 9 eV/c
2
for the mass of the electron
antineutrino. In the electron-capture process is all of the energy goes into that of the
neutrino. An experimental signature of electron capture is the X ray emitted when
the vacancy left by the inner electron that was absorbed by the nucleus is lled by
one of outer electrons.
4.3.2 Allowed beta decay
The allowed beta decay rate W between a specic set of initial and nal states is
given by:
W
i,f
= (f/K
o
)
_
g
2
V
B
i,f
(F

) +g
2
A
B
i,f
(GT

)
_
, (4.31)
4 OVERVIEW OF NUCLEAR DECAYS 54
Alpha decay Q values
P
r
o
t
o
n

N
u
m
b
e
r
Neutron Number
0
20
40
60
80
100
120
0 20 40 60 80 100 120 140 160
>0
2
4
6
8
10
>12 MeV
Figure 10: Experimental beta decay Q values.
4 OVERVIEW OF NUCLEAR DECAYS 55
HF1 Beta decay Q values
P
r
o
t
o
n

N
u
m
b
e
r
Neutron Number
0
20
40
60
80
100
120
0 20 40 60 80 100 120 140 160
>0
2
4
6
8
10
>12 MeV
Figure 11: HF1 beta decay Q values.
4 OVERVIEW OF NUCLEAR DECAYS 56
where f is dimensionless three-body phase-space factor which depends upon the beta-
decay Q value, and K
o
is a specic combination of fundamental constants:
K
o
=
2
3
h
7
m
5
e
c
4
= 1.8844 10
94
erg
2
cm
6
s. (4.32)
The refer to

decay of nucleus (A
i
, Z
i
) into nucleus (A
i
, Z
i
1). The weak-
interaction vector (V ) and axial-vector (A) coupling constants for the decay of neutron
into a proton are denoted by g
V
and g
A
, respectively.
The operators for Fermi and Gamow-Teller beta decay in terms of sums over
nucleons are:
O(F

) =

k
t
k
, (4.33)
and
O(GT

) =

k
t
k
. (4.34)
In terms of these operators the reduced matrix elements(*) in Eq. (4.31) are:
B
i,f
(F

) =
[< f[[O(F

)[[i >[
2
(2J
i
+ 1)
, (4.35)
and
B
i,f
(GT

) =
[< f[[O(GT

)[[i >[
2
(2J
i
+ 1)
=
[M
i,f
(GT

)]
2
(2J
i
+ 1)
, (4.36)
where
M
i,f
(GT

) =< f[[O(GT

)[[i > (4.37)

The matrix elements are reduced in orbital space and the (2J
i
+1) factor comes from
the average over initial M
i
states. The magnitude of reduced matrix element M(GT)
does not depend on the direction of the transition, i.e.,
[ M(GT, a b) [=[ M(GT, b a) [ (4.38)
whereas
B(GT, a b) =
(2J
b
+ 1)
(2J
a
+ 1)
B(GT, b a). (4.39)
The matrix elements obey the triangle conditions(*) J
f
= J
i
(J = 0) for Fermi and
(J
i
, j
i
, 1) (J = 1) for Gamow-Teller.
Historically one combines the partial half-life for a particular decay with the
calculated phase-space factor f to obtain from Eq. (4.31) an ft value given by:
ft
1/2
=
C
[B(F) + (g
A
/g
V
)
2
B(GT)]
(4.40)
4 OVERVIEW OF NUCLEAR DECAYS 57
where
C =
ln(2) K
o
(g
V
)
2
(4.41)
One often compiles the allowed beta decay rates in terms of a logft which stands
for log
10
of the ft
1/2
value.
4.3.3 Phase-space for allowed beta decay
The phase-space factors f are given by Wilkinson and collaborators [9], [10], [11].
The starting point is the analytic result for a nucleus of Z = 0 given by
f
Z=0
=
1
60
(2W
4
o
9W
2
o
8)p
o
+
1
4
W
o
ln(W
o
+p
o
) (4.42)
In this expression W
o
is the total electron (positron) endpoint energy in units of m
e
c
2
W
o
=
E
o
m
e
c
2
+ 1, (4.43)
and
p
o
=
_
(W
2
o
1) . (4.44)
The end point energy of the electron in units of MeV is
E
o
= (Q+E
i
) E
f
for

, (4.45)
and
E
o
= (Q+E
i
) E
f
2m
e
c
2
for
+
, (4.46)
where Q is the beta decay Q value, and E
i
and E
f
are the excitation energies of the
initial and nal states.
In the limit when E
o
>> m
e
c
2
:
f
Z=0

1
30
W
5
o
, (4.47)
which can be used as a qualitative estimate of the phase space factor and its depen-
dence on the decay energy.
To to higher level of accuracy one must take into account the distortion of the
electron energy due to the nuclear Coulomb eld and the eects of nuclear nite
size as well as other smaller corrections [10]. The nite size corrections are slightly
dierent for the Fermi and Gamow-Teller decays [10] and thus at this more precise
4 OVERVIEW OF NUCLEAR DECAYS 58
level one introduced two factors f = f
A
for the axial vector matrix element and f
V
for the vector matrix element. Eq. (4.40) written in terms of the ft value is:
ft
1/2
=
C
_
_
f
V
f
_
B(F) +
_
g
A
g
V
_
2
B(GT)
_
(4.48)
Several correction factors are applied to Eq. (4.42) to obtain f, the relation being
expressed in as:
f =
D

WM
f
Z=0
. (4.49)
The most important of these correction factors is
WM
which takes into account the
electron distortion factors. The values are tabulated by Wilkinson and Maceeld [10]
in the form

WM
= exp[

n=0,3
a
n
(lnE
o
)
n
], (4.50)
The calculation of
WM
uses electron wave functions for an atom generated from
a uniform spherical nuclear charge distribution whose radius is adjusted to t the
appropriate electron scattering and muonic x-ray data and which is corrected for the
screening of the atomic electrons. Also included in the values of
WM
are the eects
of the energy-dependent outer radiative correction to order and of the nite mass
of the nucleus. The parametrization of the values of
WM
by Eq. (4.50) is accurate
to better than 0.1% throughout [10].
The factor
R
incorporates the eects of the outer radiative correction to orders
Z
2
and Z
2

3
and is given [9], [10] by

R
= 1 + 3.67 10
4
[ Z [ +3.60 10
6
Z
2
. (4.51)
where Z is the proton number of the daughter nucleus. The factor
D
incorporates
the eects of the diuseness of the actual nuclear charge distribution [11]

D
= 1 + 1.8 10
5
[ Z [
1.36
1.2 10
6
[ Z [ W
o
. (4.52)
The Fermi phase-space factor f
V
is related [10] to the factor f by
f
V
=
V
f, (4.53)
where

V
= 1 (2/15)W
o
RZ (4/105)(W
o
R)
2
(4.54)
for beta decay, where = 1/137 and where we have used
R = 1.35A
1/3
fm (4.55)
The process of electron capture is combined with that of
+
decay in terms of a
total phase space factor:
f
+
= f(
+
) +f(ec). (4.56)
Electron-capture phase-space factors are given in the tables of Behrens and Janecke
[12].
4 OVERVIEW OF NUCLEAR DECAYS 59
4.3.4 Weak-interaction coupling constants
The values of the coupling constants for Fermi decay, g
V
, and Gamow-Teller decay,
g
A
, in the combinations in which they appear in Eq. (4.48) are obtained as follows.
For a 0
+
0
+
nuclear transition B(GT) = 0, and for a transition between T = 1
analogue states with B(F) = 2 Eq. (4.48) reduces to
C = 2t
1/2
f
V
. (4.57)
The partial half-lives and Q values for several 0
+
0
+
analogue transitions have
been measured to an accuracy of about one part in 10000. One obtains [11]:
C = 6170(4) (4.58)
This result, which together with the value of K
o
in Eq. (4.32), can be used with Eq.
(4.41) to obtain g
V
.
At the quark level g
V
= g
A
. But for nuclear structure we use the value obtained
from the neutron to proton beta decay [13]:
[ g
A
/g
V
[= 1.261(8). (4.59)
4.3.5 Double beta decay
Nuclear double-beta decay takes place in situations where a nucleus is energetically
stable to single-beta decay but unstable to the simultaneous emission of two electrons
(or two positrons). For example, for the nuclei with atomic mass number A=100,
100
Mo
100
Ru double-beta decay may occur. The Q values for double-beta decay
for those nuclei with single-beta decay Q values of 0.3 MeV or less is shown in Fig.
(4.12).
There are two types of double-beta decay: the standard (2e, 2) type in which
two neutrinos are emitted:
A
Z
A
(Z + 2) + 2e

+ 2
e
, (4.60)
or
A
Z
A
(Z 2) + 2e
+
+ 2
e
, (4.61)
and the (2e, 0) type in which no neutrinos are emitted:
A
Z
A
(Z + 2) + 2e

, (4.62)
4 OVERVIEW OF NUCLEAR DECAYS 60
double beta decay Q values
P
r
o
t
o
n

N
u
m
b
e
r
Neutron Number
0
20
40
60
80
100
120
0 20 40 60 80 100 120 140 160
0
1
2
3
4
5 MeV
Figure 12: Experimental double-beta decay Q values.
4 OVERVIEW OF NUCLEAR DECAYS 61
or
A
Z
A
(Z 2) + 2e
+
. (4.63)
The (2e, 2) double-beta decay mode has been observed in recent experiments. The
(2e, 0) double-beta decay mode has the unique signature that the total energy of the
two electrons (or positrons) must add up to the Q

value. The (2e, 0) decay mode is

being searched for experimentally, but has not yet been observed. The (2e, 0) decay
mode is possible only if the neutrinos have a nite mass, and the nonobservation of
this decay allows one to set an upper limit on the neutrino mass of about one eV/c
2
.
4.4 Gamma decay
For an electromagnetic transition from an initial nuclear state i (where the nucleus
is at rest) to nuclear state f, the momentum of the nucleus in state f (after the
transition) and the emitted gamma ray are equal and opposite. The nucleus recoils
with a kinetic energy T
f
= (E)
2
/(2m
f
c
2
) and the gamma ray has an energy E

=
E T
f
, where E = E
i
E
f
(the transition energy) is the rest-mass energy
dierence between initial and nal nuclear states. T
f
is much smaller than E and
thus to a good approximation E

= E. The energy and angular frequency of the

photon are related by E

= h. The wavelength is = hc/E

= 1237 MeV-fm/E

.
The electromagnetic transition between them can take place only if the emitted
gamma ray carries away an amount of angular momentum

such that

J
f
=

J
i
+

which means that [ J

i
J
f
[ J
i
+ J
f
where J =[

J [. Since the photon
has an intrinsic spin of one, transitions with =0 are forbidden, and hence gamma
transitions with J
i
= 0 J
f
= 0 are not allowed. A specic value determines the
multipolarity of the gamma radiation; =1 is called dipole, =2 is called quadrupole,
etc. In addition, when states can be labeled with a denite parity
i
=1 and
f
=1,
the transitions between them are restricted to the electric type of radiation when

f
(-1)

is even and the magnetic type of radiation when

i

f
(-1)

is odd.
The gamma transition rate is determined by the transition energy E, the multi-
polarity, and a factor that depends upon the details of the internal nuclear structure.
For example, the power for electric-dipole radiation (E1) from classical electromag-
netism is given by P =
4
e
2
d
2
/(3c
3
) where d is the average distance between the
positive and negative charge, = 2f, f is the frequency of the vibration in the dis-
tance between the positive and negative charge, and c is the speed of light. The E1
transition rate W(E1) (the number of gamma rays per second) is the power divided by
the energy per gamma ray (E

= h): W = P/E

=
3
e
2
d
2
/(3 hc
3
) = 2.9x10
15
E
3

d
2
MeV
3
fm
2
. The quantity d
2
depends upon the internal structure of the nuclear
states. An estimate for the lifetime associated with electric dipole radiation can be
4 OVERVIEW OF NUCLEAR DECAYS 62
obtained by taking a typical nuclear transition energy of E

= 1 MeV and a typical

nuclear size scale of d
2
= 1 fm
2
which gives a mean lifetime of = 1/W 0.3x10
15
s.
The lowest allowed multipolarity in the decay rate dominates over the next higher
one (when more than one is allowed) by several orders of magnitude. The most
common types of transitions are electric dipole (E1), magnetic dipole (M1), and
electric quadrupole (E2). Electromagnetic transition rates provide one of the most
unambiguous tests for models of nuclear structure. The strong interaction conserves
parity, and to the extent that the protons and neutrons are held together in the
nucleus by the strong interaction, their states can be labeled by a denite parity.
Since the weak interaction does not conserve parity, the weak interaction between
protons and neutrons leads to nuclear states that have a slightly mixed parity. The
electromagnetic decay between nuclear states that have a mixed parity gives rise
to mixed transitions (such as E1 plus M1), which produce circularly polarized
gamma rays (gamma rays in which the electric eld vector rotates around the axis of
propagation). Recent observations of circular polarized gamma rays have provided a
test of the weak interaction between protons and neutrons in the nucleus.
4.4.1 Reduced transition probabilities for gamma decay
The interaction of the electromagnetic eld with the nucleons can be expressed in
terms of a sum of electic and magnetic multipole operators with tensor rank(*)
O =

,
[O(E)

+O(M)

]. (4.64)
For a given E or M operator of rank the electromagnetic transition rate W
M
i
,M
f
,
for the transition between specic M states(*) is given by:
W
M
i
,M
f
,
=
_
8( + 1)
[(2 + 1)!!]
2
__
k
2+1
h
_
[< J
f
M
f
[ O()

[ J
i
M
i
>[
2
. (4.65)
In this expression k is the wave-number for the electromagnetic transition of energy
E

given by:
k =
E

hc
=
E

197 MeV fm
. (4.66)
W is explicitly labeled by the M values but depends implicitly on the wavefunctions
and operator.
The electric transition operator given by:
O(E) = r

( r) e
tz
e, (4.67)
4 OVERVIEW OF NUCLEAR DECAYS 63
were Y

are the spherical harmonics. Gamma transitions with =0 are forbidden

because the photon must carry o at least one unit of angular momentum. The e
tz
are the electric charges for the proton and neutron in units of e. For the free-nucleon
charge we would take e
p
= 1 and e
n
= 0, for the proton and neutron, respectively.
The magnetic transition operator is given by:
O(M) =
_

2g

tz
( + 1)
+s g
s
tz
_

[r

( r)]
N
=
_
(2 + 1)
_
[Y
1

2g

tz
( + 1)
+ [Y
1
s ]

g
s
tz
_
r
1

N
, (4.68)
where
N
is the nuclear magneton,

N
=
e h
2m
p
c
= 0.105 efm, (4.69)
and where m
p
is the mass of the proton. The g-factors g

tz
and g
s
tz
are the orbital and
spin g-factors for the proton and neutron, respectively. The free-nucleon values for
the g-factors are g

p
= 1, g

n
= 0, g
s
p
= 5.586 and g
s
n
= 3.826.
The total rate for a specic set of states and a given operator is obtained my
averaging over the M
i
states and summing over M
f
and :
W
i,f,
=
1
(2J
i
+ 1)

M
f
,M
f
,
W
M
i
,M
f
,
=
_
8( + 1)
[(2 + 1)!!]
2
__
k
2+1
h
_
[< J
f
[[O()[[J
i
>[
2
(2J
i
+ 1)
. (4.70)
The last factor in this equation is referred to as a reduced transition probability B
dened by:
B(i f) =
[< J
f
[[O()[[J
i
>[
2
(2J
i
+ 1)
. (4.71)
B depends upon the direction of the transition. For electromagnetic transitions J
i
is
the higher-energy initial state. But in Coulomb excitation the initial state is the lower
state, and one often uses the notation B() for this situation. If J
a
is the lower state,
J
b
is the higher state, and B() is given, then the value use for the electromagnetic
transitions J
b
J
a
is:
B(b a) =
(2J
a
+ 1)
(2J
b
+ 1)
B( a b). (4.72)
In general, a level may gamma decay by several multipoles to a given nal state:
W
i,f
=

[W
i,f
(E) +W
i,f
(M)]. (4.73)
4 OVERVIEW OF NUCLEAR DECAYS 64
One observes from the factors in Eq. (4.70) (with E

1 MeV) that the rate for

electric (magnetic) transitions with + 2 about seven orders of magnitude smaller
than those electric (magnetic) transitions for . Thus in most cases only the lowest
multipole for a given type of transition is important. Comparing rates for electric
and magnetic transitions we observe that the lowest allowed multipole may be equally
important. Thus (if allowed by the triangle condition) we need to consider mixed
transitions of the form (E1 and M2); (M1 and E2); . . .. For these mixed transition
we detine a mixing ratio by:

2
(M2/E1) =
W(M2)
W(E1)
, (4.74)
and

2
(E2/M1) =
W(E2)
W(M1)
. (4.75)
The convention is that the rate for the higher value is on top. By convention the
sign of the mixing ratio is taken as:
sign[(M2/E1)] = sign
< J
f
[[O(M2)[[J
i
>
< J
f
[[O(E1)[[J
i
>
, (4.76)
and
sign[(E2/M1)] = sign
< J
f
[[O(E2)[[J
i
>
< J
f
[[O(M1)[[J
i
>
. (4.77)
These mixing ratios can be measured from the angular distribution of the gamma rays
when the initial state is aligned (e.g. the M
i
state population is not uniform). For a
mixed transition the branching fraction, b, for a given type of transition is related to
. For example for a mixed M1 and E2 transition, the E2 branching fraction is:
b(E2) =
W(E2)
W(M1) +W(E2)
=

2
1 +
2
, (4.78)
and the M1 branching fraction is:
b(M1) =
W(M1)
W(M1) +W(E2)
=
1
1 +
2
. (4.79)
By convention the electric transition matrix elements are usually taken to be in
units of e fm

. The values of the magnetic matrix elements are usually taken to be

in units
N
fm
1
, where

N
=
e h
2m
p
c
= 0.105e fm (4.80)
4 OVERVIEW OF NUCLEAR DECAYS 65
is the nuclear magneton. It is convenient to make an explicit list of the factors which
relate the partial mean lifetime
p
and reduced transition probability B for the most
common types of transitions. For electric transitions:
B(E1) =
0.629
E
3

p
e
2
fm
2
MeV
3
fs) (4.81)
B(E2) =
816
E
5

p
e
2
fm
4
MeV
5
ps) (4.82)
B(E3) =
1760
E
7

p
e
2
fm
6
MeV
7
s) (4.83)
B(E4) =
5882
E
9

p
e
2
fm
8
MeV
9
s) (4.84)
B(E5) =
2.89 10
10
E
11

p
e
2
fm
10
MeV
11
s) (4.85)
B(E6) =
1.95 10
17
E
13

p
e
2
fm
12
MeV
13
s) (4.86)
and for magnetic transitions:
B(M1) =
56.8
E
3

2
N
MeV
3
fs (4.87)
B(M2) =
74.1
E
5

2
N
fm
2
MeV
5
ns) (4.88)
B(M3) =
0.1585
E
7

2
N
fm
4
MeV
7
s) (4.89)
B(M4) =
0.533 10
6
E
9

2
N
fm
6
MeV
3
s) (4.90)
4.4.2 Weisskopf units for gamma decay
In order to judge whether a transition is relatively weak or strong, one often gives the
B value in single-particle or Weisskopf units. This Weisskopf unit is an estimate of
the B value for a single-particle (proton or neutron) and how it depends upon mass.
By convention it is dened by:
B
W
(E) =
_
1
4
_
_
3
(3 +)
_
2
(1.2A
1/3
)
2
e
2
fm
2
, (4.91)
4 OVERVIEW OF NUCLEAR DECAYS 66
and
B
W
(M) =
_
10

_
_
3
(3 +)
_
2
(1.2A
1/3
)
22

2
N
fm
22
. (4.92)
The most commonly used are:
B
W
(E1) = 0.0645 A
2/3
e
2
fm
2
, (4.93)
B
W
(E2) = 0.0594 A
4/3
e
2
fm
4
, (4.94)
and
B
W
(M1) = 1.790
2
N
. (4.95)
Weisskopf units for the decay widths are:

W
(E2)(eV) = 4.9 10
8
A
4/3
[E

(MeV)]
5
(4.96)

W
(M1)(eV) = 2.1 10
2
[E

(MeV)]
3
(4.97)
REFERENCES 67
References
[1] S. Hofmann et al., Z. Phys. A 305, 111 (1982); O. Klepper et al., Z. Phys. A 305,
125 (1982).
[2] H. J. Rose and G. A. Jones, Nature 307, 345 (1984); D. V. Aleksandov et al.,
JEPT Lett. 40, 909 (1984).
[3] S. R. Elliott, A. A. Hahn and M. K. Moe, Phys. Rev. Lett. 59, 2020 (1987).
[4] M. Pfutzner et al., Eur. Phys. J. A14, 279 (2002); J. Giovinazzo et al., Phys. Rev.
Lett. 89, 102501 (2002).
[5] P. Woods and C. Davids, Annu. Rev. Nucl. Part. Sci. 47, 541 (1997).
[6] G. Ardisson and M. Hussonnois, Radiochim. Acta. 70/71, 123 (1995); A. A.
Oglobin et al., Phys. Rev. C 61, 034301 (2000).
[7] B. A. Brown, Phys. Rev. Lett. 69, 1034 (1992).
[8] S. B. Duarte et al., Atomic Data Nucl. Data Tables 80, 235 (2002).
[9] D. H. Wilkinson, Nucl Phys. A209, 470 (1973)
[10] D. H. Wilkinson and B. E. F. Maceeld, Nucl. Phys. A232, 58 (1974)
[11] D. H. Wilkinson, A. Gallman and D. E. Alburger, Phys. Rev. C18, 401 (1978)
[12] H. Behrens and J. Janecke, in Landolt-Bornstein, Numerical Data dn Func-
tional Relationships in Science and Technology, New Series, editor in cheif K. -H.
Hellwege, Group I: Nuclear Physics and Technology, Vol 4, ed. by H. Schopper
(Springer-Verlag) 1969.
[13] D. H. Wilkinson, Nucl. Phys. A377, 474 (1982)
5 THE FERMI GAS MODEL 68
5 The Fermi gas model
The fermi-gas model is used to obtain an approximate distribution for the kinetic
energy of nucleons conned to a volume V with density = A/V . It is evaluated
for the conditions present in nuclear matter and will give an approximate result for
the kinetic energy distribution in heavy nuclei. One assumes the lowest energy state
allowed by the Pauli exclusion principle. We will start with symmetric nuclear matter
(N = Z) with A nucleons conned to a box of length L on each side. The interaction
between nucleons enters indirectly in determining the saturation density and the size
of the box. The normalized wave function for a nucleon is given by:
(r) =
_
2
L
_
3/2
sin(k
x
x) sin(k
y
y) sin(k
z
z). (5.1)
The condition that the wave functions vanish on the boundary L means that:
k
x
=
n
x
L
(5.2)
with n
x
= 1, 2, 3 . . .. with similar conditions for k
y
and k
z
. Each state is represented
by a set of quantum numbers (n
x
, n
y
, n
z
) with energy:

i
=
h
2
k
2
2m
=
h
2
2m
_

L
_
2
(n
2
x
+n
2
y
+n
2
z
), (5.3)
and wave number
k
i
=

L
_
(n
2
x
+n
2
y
+n
2
z
) . (5.4)
From to the Pauli principle each state can be occupied by at most one identical
fermion. For nucleons we have protons and neutrons each of which can have spin up
or down, so the state degeneracy is g = 4. To get the lowest energy state we would
consider all states lled up to some radius:
n =
_
(n
2
x
+n
2
y
+n
2
z
) =
Lk

. (5.5)
The total number of states inside the radius n is equal to 1/8 of the volume of the
sphere:
N =
_
g
8
__
4
3
n
3
_
=
gL
3
k
3
6
2
, (5.6)
and the number of states inside the ring dk is:
dN =
gL
3
k
2
dk
2
2
. (5.7)
5 THE FERMI GAS MODEL 69
The maximum value for the wave number is called k
f
with the associated Fermi
energy:

f
=
h
2
k
2
f
2m
. (5.8)
The total number of states up to k
f
is the number of nucleons A:
A =
_
k
f
0
dN =
_
k
f
0
gL
3
k
2
dk
2
2
=
gL
3
k
3
f
6
2
. (5.9)
The total kinetic energy up to k
f
:
< T >=
_
k
f
0
h
2
k
2
2m
dN =
_
k
f
0
h
2
k
2
2m
gL
3
k
2
dk
2
2
=
_
1
5
_
_
h
2
2m
_
gL
3
k
5
f
2
2
. (5.10)
With Eq. (5.9) for A and Eq. (5.8) for
f
we have for the total kinetic energy:
< T >=
3
5
A
f
, (5.11)
and for the average kinetic energy:
< T >
A
=
3
5

f
. (5.12)
These results can be evaluated in terms of the density for symmetric nuclear
matter with Eq. (5.6):

o
=
A
L
3
=
g
6
2
k
3
f
, (5.13)
with
o
= 0.16 nucleons/fm
3
to obtain:
k
f
=
_
6
2

o
g
_
1/3
= 1.33 fm
1
, (5.14)

f
=
h
2
k
2
f
2m
= 36.7 MeV, (5.15)
and for the average kinetic energy:
< T >
A
=
3
5

f
= 22.0 MeV. (5.16)
For unequal numbers of protons and neutrons we take two overlapping Fermi
gases in the same size box with densities for protos and neutrons given by

op
=
Z
A

o
, (5.17)
5 THE FERMI GAS MODEL 70
and

on
=
N
A

o
, (5.18)
With the spin degeneracy for one kind of nucleon of g

= 2 we obtain for protons

k
fp
=
_
6
2
g

op
_
1/3
=
_
6
2
g

Z
A

o
_
1/3
=
_
g
g

Z
A
_
1/3
k
f
=
_
2Z
A
_
1/3
k
f
, (5.19)

fp
=
h
2
k
2
fp
2m
=
_
2Z
A
_
2/3

f
, (5.20)
and for the average kinetic energy:
< T >
p
Z
=
3
5

fp
. (5.21)
Likewise for the neutrons:
k
fn
=
_
2N
A
_
1/3
k
f
, (5.22)

fn
=
_
2N
A
_
2/3

f
, (5.23)
and for the average energy:
< T >
n
N
=
3
5

fn
. (5.24)
The total kinetic energy is thus given by
< T >=< T >
p
+ < T >
n
=
3
5
_
Z
_
2Z
A
_
2/3
+N
_
2N
A
_
2/3
_

f
. (5.25)
The numerical values for
208
Pb with Z = 82 and N = 126 are
fp
= 31.6 MeV,

fn
= 42.0 MeV, < T >= 4700 MeV and
<T>
A
= 22.8 MeV. The average kinetic
energy is larger than the value of 22.0 MeV obtained for N = Z (Eq. (5.16)).
One can expand Eq. (5.25) to rst order in [
(NZ)
A
]
2
to obtain:
< T >=
3
5
A
f
+
2
5

f
(N Z)
2
A
= 22 A+ 15
(N Z)
2
A
MeV. (5.26)
Comparing this with the total energy obtained empirically with the liquid drop model
(E = BE):
E =< T > + < V >= 16 A+ 23
(N Z)
2
A
MeV, (5.27)
we would deduce for the interaction energy:
< V >= 38 A+ 8
(N Z)
2
A
MeV. (5.28)
6 OVERVIEW OF THE NUCLEAR SHELL MODEL 71
6 Overview of the nuclear shell model
Nuclei are made up of protons and neutrons (nucleons) held together by the strong
interaction inside of a volume with a radius of a few Fermis (fm), (1 fm =10
15
m).
One might expect that the motions of these nucleons in this closely packed system
should be very complex because of the large number of frequent collisions. Niels Bohr
made the analogy of the nucleus to a table of billiard balls [1] and said that an analogy
with the motion of electrons in a one-body potential loses any validity for nucleons
in the nucleus [2]. But the data on nuclear properties increased rapidly, see Table
I, and by 1948 Maria G. Mayer was able to summarize the experimental data which
indicated that nuclei with 20, 50, 82 or 126 neutrons or protons were particularly
stable [3]. In 1949 the key role of the spin-orbit splitting in the one-body potential
was proposed by Maria Goeppert Mayer [4] and Otto Haxel, J. Hans D. Jensen and
Hans Suess [5]. This one-body potential model is the starting point for the nuclear
shell model. A short history of Mayers contributions is given in [6].
Table of Isotopes
Edition Year Authors Pages
1st 1940 Livingood and Seaborg 17
2nd 1944 Seaborg 32
3rd 1948 Seaborg and Perlman 82
4th 1953 Hollander, Perlman and Seaborg 182
5th 1958 Strominger, Hollander and Seaborg 320
6th 1967 Lederer, Hollander and Perlman 556
7th 1978 Lederer et al. 1523
8th 1996 Firestone et al. 2877
In the shell model, the quantum mechanical problem for the motion of one nucleon
in a nucleus is similar to that for the motion of an electron in the hydrogen atom,
except that overall scale is determined by the size of the nucleus (10
15
m) rather than
the size of the atom (10
10
m). Another important dierence between the atomic and
nuclear potentials is that the dependence of the potential on the relative orientation
of the intrinsic nucleon (electron) spin and its orbital angular momentum is much
stronger and opposite in sign for the nucleon compared to that for the electron.
The single-particle potential has eigenstates that are characterized by their single-
particle energies and their quantum numbers. The properties of a nucleus with a
given number of protons and neutrons are determined by the lling of the lowest
energy single-particle levels allowed by the Pauli exclusion principle which must be
obeyed in a system of identical Fermions (the nucleons in this case). The Pauli
exclusion principle allows only one proton or neutron to occupy a state with a given
6 OVERVIEW OF THE NUCLEAR SHELL MODEL 72
set of quantum numbers. The average nuclear potential arises from the short-ranged
attractive nucleon-nucleon interaction and is determined by the shape of the nuclear
density distribution.
Evidence for the validity of the nuclear shell model comes from the observation
of shell eects in experimental observables such as binding energy, size, spin, and
level density. In particular, the nuclear binding energy is not a smooth function
of proton and neutron number, but exhibits small uctuations. The deviation of
the experimental binding energies from the liquid-drop model was shown in Sec. 1.
The liquid-drop model binding energy is a smooth function of proton and neutron
number. When the liquid-drop values for the binding energies are subtracted from
the experimental values, the dierences show peaks at the magic numbers: N
m
=28,
50, 82 and 126. The peak indicates that the nuclei with these magic number are more
tightly bound than average. Those nuclei that are magic with respect to both neutron
and proton numbers are referred to as doubly-magic; an example is the nucleus
208
Pb
with N = 126 and Z = 82. Although not so obvious from the binding energy data
N
m
= 2, 6, 8, 14, 16, 20 and 32 are also magic numbers for some nuclei. In addition
N
m
= 34 has recently been proposed as a magic number for
54
Ca [7]. The occurrence
of magic numbers in nuclei is analogous to that observed for the properties of electrons
in atoms. However, for electrons the magic numbers are 2, 10, 18, 54, and 86.
The calculated single-particle energy levels appropriate for the neutrons in
208
Pb
are shown in Fig. (6.1). The potential arises from the average interaction of one
neutron with the 207 other nucleons. Since the nuclear force is short-ranged, the
shape of potential is similar to the nucleon density in
208
Pb which is experimentally
known to be close to the Fermi- or Woods-Saxon shape of
V (r) =
V
o
[1 + exp(r R)/a]
, (6.1)
where R 1.2A
1/3
fm and a 0.60 fm. The single-particle energy levels for a poten-
tial of approximately this shape and with a central depth of about V
o
= 50 MeV
are shown in the middle of Fig. (6.1). The number of neutrons that are allowed by
the Pauli principle to occupy one of these levels, the occupation number, is given by
the number in square brackets. In addition, each level is labeled by its cumulative
occupation number (the total number of neutrons needed to ll up to the given level)
and its n value. n is the radial quantum number (the number of times the radial
wave function changes sign) and is the angular momentum quantum number repre-
sented in the spectroscopic notations s, p, d, f, g, h, i and j for =0, 1, 2, 3, 4, 5, 6 and
7, respectively. Each value can have 2+1 m states and each m state can contain a
proton or neutron with spin up and spin down (s
z
=1/2). The occupation number
given by the Pauli principle is thus N
o
= 2(2 + 1).
The Woods-Saxon results are compared to the levels of an harmonic oscillator
6 OVERVIEW OF THE NUCLEAR SHELL MODEL 73
s
i
n
g
l
e
-
p
a
r
t
i
c
l
e

e
n
e
r
g
y

(
M
e
V
)

-50
-40
-30
-20
-10
0
10
[ 2] 2 N=0
[ 6] 8 N=1
[12] 20 N=2
[20] 40 N=3
[30] 70 N=4
[42] 112 N=5
[56] 168 N=6
[ 2] 2 0s
[ 6] 8 0p
[10] 18 0d
[ 2] 20 1s
[14] 34 0f
[ 6] 40 1p
[18] 58 0g
[10] 68 1d
[ 2] 70 2s
[22] 92 0h
[14] 106 1f
[ 6] 112 2p
[26] 138 0i
[18] 156 1g
[10] 166 2d
[ 2] 168 3s
[ 2] 2 0s1
[ 4] 6 0p3
[ 2] 8 0p1
[ 6] 14 0d5
[ 4] 18 0d3
[ 2] 20 1s1
[ 8] 28 0f7
[ 6] 34 0f5
[ 4] 38 1p3
[ 2] 40 1p1
[10] 50 0g9
[ 8] 58 0g7
[ 6] 64 1d5
[ 4] 68 1d3
[ 2] 70 2s1
[12] 82 0h11
[ 8] 90 1f7
[10] 100 0h9
[14] 114 0i13
[ 4] 118 2p3
[ 6] 124 1f5
[ 2] 126 2p1
[10] 136 1g9
[12] 148 0i11
[ 6] 154 2d5
[ 2] 156 3s1
[ 8] 164 1g7
[16] 180 0j15
[ 4] 184 2d3












































Harmonic
Oscillator
Potential
Woods-Saxon
Potential
Woods-Saxon
Plus Spin-Orbit
Potential
Neutron Single-Particle Energies
2
8
20
28
50
82
126
Figure 1: Neutron single-particle states in
208
Pb with three potential mod-
els, harmonic oscillator (left), Woods-Saxon without spin-orbit (middle) and
Woods-Saxon with spin orbit (right). The numbers in square brackets are the
maximum number of neutrons in that each level can contain, the following num-
ber is a running sum of the total. In addition the harmonic oscillator is labeled
by the major quantum number N = 2n +, the Woods Saxon is labeled by n,
and the Woods-Saxon with spin-orbit is labeled by n, , 2j.
6 OVERVIEW OF THE NUCLEAR SHELL MODEL 74
potential on the left-hand side of Fig. (6.1). Note that the degeneracy present in
the oscillator is broken in the Woods-Saxon potential with levels of larger coming
lower in energy.
The relative spacing of the neutron and proton levels for all nuclei are qualita-
tively similar to those shown in Fig. (6.1). (The overall spacing between levels goes
approximately as A
1/3
.) According to the Pauli principle, as neutrons are added to
nuclei they go into the lowest energy level not already occupied. When a nucleon
is added to a nucleus in which the neutron number is equal to one of the cumula-
tive occupation numbers, the neutron must be placed into a relatively higher-energy
(more loosely bound) state. Thus the nuclei with the highest relative binding energy
are those for which the proton or neutron number is equal to one of the cumulative
occupation numbers. A magic number occurs when there is a relatively large energy
gap above one of the cumulative numbers. The magic numbers are thus related to
the bunching of energy levels. The Woods-Saxon potential gives the correct magic
numbers for N
m
= 2, 8 and 20 but is incorrect for the higher values.
In 1949 Goeppert-Mayer and Haxel, Jensen and Suess postulated the existence
of an additional strong spin-orbit potential that could account for the observed magic
numbers. The spin-orbit potential has the form, V
so
(r)

.
s, where

is the orbital
angular momentum and s is the intrinsic spin angular momentum of the nucleon.
With the spin-orbit potential, m and s
z
are no longer good quantum numbers. The
orbital and spin angular momentum must be coupled to a denite total angular
momentum,

j =

+s. Eigenstates of the spin-orbit potential are determined by the
total angular momentum quantum number j = 1/2 (except j = 1/2 for =0)
and the quantum number m
j
associated with the z component of j. The expectation
value of

.
s can be obtained from the operator
j
2
= (

+s)
2
=
2
+s
2
+ 2

.
s. (6.2)
Hence,

.
s [
j
>=
1
2
(j
2

2
s
2
) [
j
>=
1
2
[j(j +1) ( +1) s(s+1)] [
j
> (6.3)
(the minus sign takes into account the observed sign of the < V
so
(r) >) which gives
<
j=+1/2
[

.
s [
j=+1/2
>=

2
(6.4)
and
<
j=1/2
[

.
s [
j=1/2
>= +
+ 1
2
(6.5)
Each j has 2j + 1 m
j
values and hence each j orbit can contain up N
o
= 2j + 1
protons or neutrons. The energy levels obtained when the spin-orbit potential is
6 OVERVIEW OF THE NUCLEAR SHELL MODEL 75
added to the Woods-Saxon potential are shown on the right-hand side of Fig. (6.1).
The dashed lines that connect to the middle of Fig. (6.1) indicate the eect of the
spin-orbit potential in splitting the states of a given value. The overall strength
of the spin-orbit potential has been determined empirically. Each level is labeled
by the occupation number (in square brackets), the cumulative occupation number,
and the values for n, , 2j (2j is twice the angular momentum quantum number j).
The values of the neutron number for which there are large gaps in the cumulative
occupation number now reproduce all of the observed magic numbers (as emphasized
by the numbers shown in the energy gaps on the right-hand side).
The average nuclear potential can be calculated microscopically from the nucleon-
nucleon interaction by using Hartree-Fock theory together with the Breuckner theory
for taking into account the repulsion at very short distances between the nucleons.
The strength of the spin-orbit potential for nucleons is much larger and opposite
in sign to spin-orbit potential for electrons atoms. The radial part of the spin-orbit
potential, V
so
(r), is largest at the nuclear surface and is often taken to be proportional
to the derivative of the Woods-Saxon form.
The essential physics behind the shell model is that the many-nucleon collisions
that might be expected are greatly suppressed in the nuclear ground states and low-
lying nuclear levels because the nucleons would be scattered into states which are
forbidden by the Pauli principle. At higher excitation energy the number of allowed
states becomes much greater and the nuclear properties indeed become complex and
chaotic.
The shell model in its simplest form is able to successfully predict the properties
of nuclei which are one nucleon removed or added to the one of the magic number.
The shell model can also be extended to include the more complex congurations that
arise for the nuclei with nucleon numbers that are in between the magic numbers. For
many applications these complex congurations can be taken into account exactly by
the diagonalization of a Hamiltonian matrix. In other cases approximations must be
used; these include the use of a deformed intrinsic single-particle potential, and the
use of group theory to classify the congurations. Current theoretical investigations
using the shell model focus on these complex congurations.
REFERENCES 76
References
[1] Niels Bohr, Science 86, 161 (1937).
[2] Niels Bohr, Nature, February, 1936.
[3] Maria G. Mayer, Phys. Rev. 74, 235 (1948).
[4] M. G. Mayer, Phys. Rev. 75, 1969 (1949).
[5] O. Haxel, J. H. D. Jensen and H. E. Suess, Phys. Rev. 75, 1766 (1949).
[6] K. E. Johnson, Maria Goeppert Mayer: Atoms, molecules and nuclear shells,
Physics Today 39, 44 (1986).
[7] M. Honma, T. Otsuka, B. A. Brown and T. Mizusaki, Phys. Rev. C 65, 061301
(2002).
7 THE ONE-BODY POTENTIAL 77
7 The one-body potential
7.1 General properties
The Schroedinger equation for a particle of mass m in a spin-independent central
potential, U
o
, is
H
o
[ ) = (T +U
o
) [ ) =

[ ), (7.1)
where T is the kinetic energy operator given in coordinate space by
T =
h
2
2m

2
=
h
2
2m
_
_
_
1
r
d
2
dr
2
r

2
r
2
_
_
_
, (7.2)
the ket [ ) is the single-particle wave function, and

is the eigenvalue. The coor-

dinate space solution of this equation has the form

(r ) =
R

(r)
r
Y

m

( r), (7.3)
where Y

are the spherical harmonics. The subscript stands for the set of quantum
numbers (n
r
, , m

), with n
r
being the radial quantum number. We will use the
quantum numbers n
r
= 0, 1, 2,. . . which indicate the number of times the radial wave
functions crosses the r axis. (Some other books use the convention which starts at
one, n
r
= 1, 2, 3,. . . ). The spherical harmonics are eigenfunctions of the
2
and
z
operators

2
Y

( r) = ( + 1)Y

( r), (7.4)
and

z
Y

m

( r) = m

( r), (7.5)
and are orthonormal functions
_
Y

( r) Y

( r) d =

. (7.6)
The radial wave function R

(r) is a solution of the one-dimensional radial equation

h
2
2m
R

(r) +
h
2
2m
( + 1)
r
2
R

(r) +U
o
(r)R

(r) =

(r), (7.7)
where R

(r) indicates the second derivative of R(r) with respect to r. The solution
to Eq. (7.7) is independent of m

.
For a particle with an intrinsic spin, s, the central potential may also contain a
spin-orbit term of the form
U
so
(r) = f
so
(r)

, (7.8)
7 THE ONE-BODY POTENTIAL 78
where = 2s. The total angular momentum,

j =

+s, is conserved and the wave
function takes the form

k,m
(r ) =
R
k
(r)
r
[Y

s
]
j
m
, (7.9)
where
s
is the intrinsic-spin wave function, k stands for the set of quantum numbers
(n
r
, , s, j), and the cross symbol denotes the Clebsch-Gordan product
[Y

s
]
j
m
=

,ms
, m

, s, m
s
[ j, m) Y

( r)
s
ms
. (7.10)
This equation denes the ordering and phase convention associated with spin-orbit
coupling used in this book. The intrinsic-spin wave functions are also assumed to be
orthonormal

s
ms
[
s

s
) =
ss

msm

s
. (7.11)
The operation of

on
k,m
(r ) gives
(

)
k,m
(r ) = [j(j + 1) ( + 1) s(s + 1)]
k,m
(r ) = a
so

k,m
(r ). (7.12)
Taking s =
1
2
and j =
1
2
,
a
so
= ( + 1) for j =
1
2
,
and
a
so
= for j = +
1
2
.
Thus R
k
(r) satises the equation

h
2
2m
R

k
(r) +
h
2
2m
( + 1)
r
2
R
k
(r) + [U
o
(r) +a
so
f
so
(r)]R
k
(r) =
k
R
k
(r). (7.13)
Eq. (7.13) has discrete solutions for bound states with <0 as well as continuum
solutions for >0. The normalization convention for bound states is
_
[
k,m
(r ) [
2
d =

,ms
m

,m

s
, m

, s, m
s
[ j, m), m

, s, m

s
[ j, m)

_
[ R
k
(r) [
2
dr
_
Y

( r)Y

( r) d
s
ms
[
s
m

s
). (7.14)
The spherical harmonics and intrinsic-spin wave functions are orthonormal and thus
_
[
k,m
(r ) [
2
d =
_
[ R
k
(r) [
2
dr = 1. (7.15)
7 THE ONE-BODY POTENTIAL 79
Eq. (7.15) denes the normalization of the radial wave functions. In addition, the
radial wave functions will be chosen to be real and positive at the origin. Observables
are, of course, independent of this phase convention, however, it is important to keep
the phase convention consistent throughout intermediate steps of a calculations which
involve R(r) or matrix elements of R(r). The quantity [ R
k
(r)/r[
2
is the radial prole
function the density observed for a thin slice at z = 0 in the xy plane. The quantity
P(r) =[ R
k
(r) [
2
is the radial probability distribution the probability of nding the
particle at a distance r from the origin. The = 0 states are the only ones with a
nonvanishing radial-prole at the origin.
There are a variety of numerical methods [1] for solving the radial dierential
equation for the eigenvalues and the eigenfunctions. As a function of increasing n
r
for a xed -value, the excitation energy relative to the ground state (gs)

gs
usu-
ally increases and the root-mean-square radius of the radial probability distribution
increases. Other properties will depend upon the details of the potential. For ex-
ample, the Coulomb potential with U
o
(r) = 1/r has energies proportional to 1/n
2
p
,
where n
p
= n
r
+ + 1 is the principal quantum number. In the following subsec-
tions, some results for the harmonic-oscillator and Woods-Saxon forms for the central
potential, which are commonly used in nuclear physics, are presented.
7.2 The harmonic-oscillator potential
A very useful form for U
o
(r) used in nuclear physics is the harmonic-oscillator (HO)
potential given by
U
HO
o
(r) =
1
2
m
2
r
2
, (7.16)
where m is the nucleon mass and is a parameter. This potential has nice analytical
properties, the most important of which is that the many-body HO Hamiltonian can
be exactly separated into a sum of intrinsic and center-of-mass terms. This will be
discussed in the next section. The single-particle energy levels are given by

= (N + 3/2) h, (7.17)
where N is the major-shell HO quantum number given by
N = 2n
r
+. (7.18)
The radial wave functions are given by
R

(r) =

_
2
nr+2
(2 + 2n
r
+ 1)!

(n
r
)! b
2+3
[(2 + 1)!!]
2
r
+1
e
r
2
/2b
2
7 THE ONE-BODY POTENTIAL 80

nr

k=0
(1)
k
2
k
(n
r
)! (2 + 1)!!
k! (n
r
k)! (2 + 2k + 1)!!
(r/b)
2k
, (7.19)
where
b =

h
m
, (7.20)
is the HO length parameter. Taking hc = 197 MeV-fm and mc
2
= 938 MeV
b
2
=
( hc)
2
(mc
2
)( h)
=
41.4 MeVfm
2
h
. (7.21)
The diagonal matrix element of r
2
can be obtained by using the virial theorem for
the HO potential,

[ U
HO
o
[

) =
1
2
m
2

[r
2
[

) =

[ T [

),
together with

[ U
HO
o
[

) +

[ T [

) = (N + 3/2) h
to give

[r
2
[

) = N[r
2
[ N) = (N + 3/2) b
2
. (7.22)
and

[ T [

) =
h
2
(N + 3/2) (7.23)
An analytic expression for the general matrix elements of r

can be found in [2] and

[3], and Fortran programs are given in [2] and [4].
The maximum number of spin
1
2
fermions which can occupy a given N value is
given by
D
N
= (N + 1)(N + 2). (7.24)
The total number of states which can be occupied up to N
max
is given by
N=Nmax

N=0
D
N
=
1
3
(N
max
+ 1)(N
max
+ 2)(N
max
+ 3). (7.25)
The harmonic oscillator parameter is conventionally chosen to reproduce the ob-
served mean-square charge radius. This is related to the mean-square radius for
protons given nucleus whose harmonic oscillator levels are lled up to N
max
r
2
)
p
=
N=Nmax

N=0
D
N
N[r
2
[ N)
N=Nmax

N=0
D
N
=
3
4
(N
max
+ 2)b
2
. (7.26)
7 THE ONE-BODY POTENTIAL 81
A common way of choosing the appropriate value of h is to compare Eq. (7.26)
to the observed proton mean-square radius as obtained from measured charge mean-
square radius by
r
2
)
p
= r
2
)
ch
r
2
)
o
. (7.27)
where r
2
)
o
= 0.77 fm
2
is the mean-square charge radius of the proton. For example,
for the nucleus
16
O, where the eight protons ll the N = 0 and N = 1 major shells,
N
max
= 1 and r
2
)
p
(
16
O) = (9/4)b
2
. The experimental root-mean-square (rms) charge
radius for
16
O is [5]
_
r
2
)
ch
= 2.71 fm, and hence r
2
)
p
= 6.57 fm
2
, b
2
= 2.92 fm
2
and h = 14.2 MeV for the point protons. Since there are equal numbers of neutrons
and protons in
16
O and since they should have about the same rms radii, it is a good
approximation to use the same oscillator parameter for the neutrons as was obtained
for protons.
For heavy nuclei, one can use the large N
max
approximation for the sum in Eq.
(7.25), Z (N
max
+ 2)
3
/3, to obtain from (7.26)
r
2
)
p

3
4
(3Z)
1/3
b
2
. (7.28)
For example, for
208
Pb (Z=82) the experimental rms charge radius is [5] 5.50 fm,
giving r
2
)
p
= 29.4 fm
2
, b
2
= 6.27 fm
2
and h
p
= 6.6 MeV for the point protons. It
is observed experimentally that the proton and neutron rms radii in
208
Pb are about
equal [6]: r
2
)
n
r
2
)
p
. The oscillator parameter for neutrons in
208
Pb, obtained by
using an rms radius of 5.5 fm and Eq. (7.28) with N in place of Z, is h
n
= 7.6 MeV.
Along the valley of stability the rms proton radii are approximately given by
1.08A
1/3
. Rewriting Eq. (7.28) in terms of A 2.5Z, one obtains for heavy nu-
clei h
p
39A
1/3
MeV. A better smooth approximation for the proton oscillator
parameter in both light and heavy nuclei is [7]
h
p
45A
1/3
25A
2/3
. (7.29)
7.3 Separation of intrinsic and center-of-mass motion
7.3.1 The kinetic energy
The many-body nuclear Hamiltonian for A nucleons with two-body interactions has
the form
H =
1
2m
A

i
p
2
i
+
A

i<j
V ([ r
i
r
j
[). (7.30)
7 THE ONE-BODY POTENTIAL 82
The coordinates refer to a xed point in space. For the internal structure of a given
nucleus we are not interested in the center-of-mass motion, and we would like to
rewrite the Hamiltonian as a sum of an intrinsic part H
int
and a center-of-mass part
H
cm
H = H
int
+H
cm
. (7.31)
The center-of-mass Hamiltonian depends only on the center-of-mass position

R and
momentum

Q:

R =
1
A
A

i
r
i
, (7.32)
and

Q =
A

i
p
i
. (7.33)
The intrinsic Hamiltonian depends upon the coordinates relative to the the center-
of-mass:

i
= r
i

R, (7.34)
and
q
i
= p
i

1
A

Q. (7.35)
The two-body interaction is explicitly a function of the center-of-mass coordinates

i

j
= r
i
r
j
. The kinetic energy part of H can be rewritten as
T = T
int
+T
cm
, (7.36)
where
T
int
=
1
2m
A

i
q
2
i
=
2
Am
A

i<j
( p
ij
)
2
, (7.37)
and
T
cm
=
1
2mA
Q
2
, (7.38)
where p
ij
= ( p
i
- p
j
)/2. Thus H separates into
H = H
int
+H
cm
,
with
H
int
=
1
2m
A

i
q
2
i
+
A

i<j
V ([
i

j
[) (7.39)
and
H
cm
= T
cm
=
1
2mA
Q
2
. (7.40)
The solution to
H = (H
int
+H
cm
) = E (7.41)
7 THE ONE-BODY POTENTIAL 83
has the form
=
int

cm
, (7.42)
with
H
int

int
= E
int

int
, (7.43)
H
cm

cm
= E
cm

cm
, (7.44)
and
E = E
int
+E
cm
. (7.45)
The coordinates
i
are not independent and therefore the equation for the intrinsic
Hamiltonian Eq. (7.43) is in general more dicult to solve than that for the full
Hamiltonian Eq. (7.41). Thus, one usually solves the full problem constrained is a
way which guarantees that the center-of-mass is in a known state, such as a plane
wave or the ground state of a harmonic oscillator. The intrinsic wave function is then
obtained by dividing by
cm
and the intrinsic energy is obtained by subtracting
E
cm
from E.
7.3.2 The harmonic-oscillator
In the nuclear many-body problem one divides the hamiltonian into a mean-eld
(single-particle) potential U plus a residual interaction W:
H

=
1
2m
A

i
p
2
i
+
A

i
U(r
i
) +
A

i<j
W([ r
i
r
j
[). (7.46)
In general U(r
i
) does not easily separate into an intrinsic and center-of-mass part.
However, in the special case of the harmonic-oscillator potential the separation can
be made analytically:
A

i
U
HO
(r
i
) =
1
2
m
2
A

i
r
2
i
=
1
2
m
2
A

2
i
+
1
2
A
2
mR
2
. (7.47)
Thus H

separates into
H

= H

int
+H

cm
, (7.48)
with
H

int
=
1
2m
A

i
q
2
i
+
1
2
M
2
A

2
i
+
A

i<j
W([
i

j
[) (7.49)
and
H

cm
=
1
2mA
Q
2
+
1
2
Am
2
R
2
, (7.50)
7 THE ONE-BODY POTENTIAL 84
Thus, if one solves Eq. (7.46) such a way that the center-of-mass is guaranteed to be
in a 0s oscillator state, the intrinsic wave function and energy can be obtained.
When the single-particle levels of the oscillator potential are lled in their lowest
energy state the center of mass must also be in its lowest energy energy state of 0s for
the nucleus with mass Am in the potential
1
2
Am
2
R
2
, with a center of mass energy
of
[ H

cm
[ ) =
3
2
h, (7.51)
Many-body states for which the center-of-mass is in the 0s ground state are referred
to as nonspurious states.
Excitations of nucleons across across major shells or from a lled major shell
to partially lled major shells are characterized by a single-particle energy change
E
sp
= N h, where
H
sp
=
1
2m
A

i
p
2
i
+
A

i
U
HO
(r
i
). (7.52)
This energy change may occur all in the intrinsic motion, all in the center-of-mass
motion, or partly in each.
Spurious states are formed when the center-of-mass is excited from its 0s ground
state into its excited states. These states can be explicitly constructed by operating
with

R on one of the nonspurious states in which the center-of-mass is in the 0s
ground state. The complete set of spurious states can be obtained from the angular
momentum coupling of all such spurious states with all of the nonspurious states. For
example, if one starts with the
16
O ground state, the 1 h excitations lead to negative
parity states. One of these states with J

=1

(and T=0) is spurious.

For the special situations where the excited conguration cannot be connected to
the original nonspurious conguration by the (

R)
n
operator, these states are also non-
spurious. The pure 0p
1/2
to 0d
5/2
excitation, for example, is completely nonspurious.
There are also special SU(3) congurations which are nonspurious [8].
In order to carry out the transformation of into
int
and
cm
one must be sure
that the basis states are complete with respect to all center-of-mass modes of a given
E
sp
. The center-of-mass wave functions can then be explicitly constructed within
the basis and projected out. Alternatively one can add a ctitious Hamiltonian
which acts only on the center-of-mass:
H

cm
=
_
Q
2
2Am
+
1
2
Am
2
R
2

3
2
h
_
. (7.53)
7 THE ONE-BODY POTENTIAL 85
The center-of-mass for the full Hamiltonian H + H

cm
will be in the 0s ground state
if the constant is made large enough so that the excitation energies of the center-
of-mass are larger than all of the intrinsic excitation energies of interest. Subtraction
of the constant
3
2
h means that E

cm
= 0 when the center-of-mass is in the 0s ground
state. Thus, the Hamiltonian
H

+H

cm
(7.54)
with H

given by Eq. (7.46) and H

cm
given by Eq. (7.53) can be used to solve the
nuclear many-body problem in terms of the xed space coordinates r
i
. This is the
method employed by the OXBASH shell-mode code [4] If one is starting with the
original Hamiltonian of Eq. (7.30), one can construct a Hamiltonian which generates
nonspurious states and is also corrected for the kinetic energy of the center-of-mass
by
H H
cm
+H

cm
, (7.55)
where H
cm
is given by Eq. (7.40).
REFERENCES 86
References
[1] J. R. Cash, A. D. Raptis and T. E. Simos, Jour. of Comput. Phys. 91, 413 (1990).
[2] K. L. G. Heyde, The Nuclear Shell Model, (Springer-Verlag, 1994)
[3] R. D. Lawson, Theory of the Nuclear Shell Model, (Clarendon Press, 1980).
[4] B. A. Brown, et al., the computer code OXBASH.
[5] B. A. Brown, C. R. Bronk and P. E. Hodgson, J. Phys. G10, 1683 (1984) and
references therin.
[6] B. A. Brown, S. E. Massen, J. I. Escudero, P. E. Hodgson, G. Madurga and J.
Vinas, J. Phys. G9, 423 (1983).
[7] J. Blomqvist and A. Molinari, Nucl. Phys. A106, 545 (1968).
[8] J. Millener et al., private communication.
8 THE WOODS-SAXON POTENTIAL 87
8 The Woods-Saxon potential
8.1 General form
The Woods-Saxon potential is a convenient phenomenological choice for the one-
body potential. It provides a model for the properties of bound-state and continuum
single-particle wavefunctions. The Woods-Saxon potential (or any other one-body
potential) cannot be used for the total binding energy since it is not based upon a
specic two-body interaction. The parameters of the Woods-Saxon are chosen for
a best t of nuclear single-particle energies and nuclear radii. The Woods-Saxon
potential is based upon the sum of a spin-independent central potential, a spin-orbit
potential, and the Coulomb potential:
V (r) = V
o
(r) +V
so
(r)

s +V
c
(r), (8.1)
where V
o
(r) is the spin-independent central potential:
V
o
(r) = V
o
f
o
(r), (8.2)
with a fermi shape
f
o
(r) =
1
1 + [exp(r R
o
)/a
o
]
, (8.3)
V
so
(r) is the spin-orbit potential:
V
so
(r) = V
so
1
r
df
so
(r)
dr
, (8.4)
with
f
so
(r) =
1
1 + [exp(r R
so
)/a
so
]
, (8.5)
and V
c
(r) is the Coulomb potential for protons based upon the Coulomb potential for
a sphere of radius R
c
:
V
c
(r) =
Ze
2
r
for r R
c
and
V
c
(r) =
Ze
2
R
c
_
3
2

r
2
2R
2
c
_
for r R
c
. (8.6)
The radii R
o
, R
so
and R
c
are usually expressed as:
R
i
= r
i
A
1/3
. (8.7)
8 THE WOODS-SAXON POTENTIAL 88
For nuclei with a neutron excess the protons will feel a stronger potential than
the neutrons, since the average proton-neutron potential is stronger than the average
neutron-neutron (or proton-proton) potential. Thus we take:
V
op
= V
0
+
(N Z)
A
V
1
for protons (8.8)
and
V
on
= V
0

(N Z)
A
V
1
for neutrons (8.9)
In principle, r
o
and a
o
could also be a little dierence for proton and neutrons in a
nucleus with N ,= Z. Thus the spin-indepedent potential could have six parameters
(and even more is any of them are allowed to take some additional mass dependence).
The values of these parameters are chosen to give an overall accounting of the observed
single-particle energies, the rms charge radii, and the electron scattering form factors.
The form of the spin-orbit interaction was originally taken from the form of the
spin-orbit interaction used for electrons in atoms which was derived from the Dirac
theory by Thomas and Frenkel [1]. For nuclear physics the spin-orbit interaction
ultimately comes from the nucleon-nucleon interaction, but there is still some debate
over whether or not relativistic eects are very important. The spin-orbit interaction
for nucleons in the nucleus has the opposite sign and is much larger than for electrons
in the atom. In the nuclear interior a nucleon sees an equal number of spin-up and
spin-down nucleons on both sides and the spin-orbit interaction must vanish. Thus,
the spin-orbit potential is peaked at the nuclear surface as in Eq. (8.4), and the
particular form of this equation gives a good overall description of the and mass
dependence observed experimentally. In analogy to the spin-independent potential,
one could introduce up to six parameters for the spin-orbit interaction. The strength
of the spin-orbit interaction is chosen to reproduce spin-orbit splittings (especially for
large ). For N ,= Z the strength V
so
could be dierent for protons and neutrons, but
it practice they are nearly the same. For the geometry one usually takes r
so
= r
o
and
a
so
= a
o
. Thus, the introduction of one parameter in the spin-orbit interaction V
so
gives a good overall accounting of the data.
A typical set of parameters for the Woods-Saxon potential is V
0
= 53 MeV,
V
1
= 30 MeV and V
so
= 22 MeV for the strengths, and r
o
= r
so
= 1.25 fm and
a
o
= a
so
= 0.65 fm for the geometry. For the Coulomb term the radius is a little
smaller with r
c
= 1.20 fm. One can nd in the literature many other sets of parameters
which are better for specic nuclei or mass regions.
8 THE WOODS-SAXON POTENTIAL 89
The gure on the next page from page 239 of Bohr and Mottelson Vol I [2] shows
the neutron single-particle energies for nuclei near the valley of stability obtained
with a set of Woods-Saxon parameters similar to those given above (note that the
radial quantum numbers in this gure start with one rather than with our convention
of zero). One observed that as A increases more single-particle states become bound
so that the energy of the most loosely bound lled orbit is aways around 8 MeV.
8 THE WOODS-SAXON POTENTIAL 90
Figure from BM
8 THE WOODS-SAXON POTENTIAL 91
8.2 Computer program for the Woods-Saxon potential
The program WSPOT.FOR can be used to obtain the single-particle energies and
single-particle radial wavefunctions for the bound states of the Woods-Saxon poten-
tial with quantum numbers n
r
, and j. (It could also be used for other potentials
if the program is modied.) If can also be used to calculate the nucleon scattering
cross sections for a given and j. values. The program is initialized with a typical
set of Woods-Saxon parameters:
V
0
= 53 MeV,
V
1
= 30 MeV,
V
so
= 22 MeV,
r
o
= r
so
= 1.25 fm,
a
o
= a
so
= 0.65 fm and
r
c
= 1.20 fm.
Some parameters can be modied in the input les. Other parameters can be
changed internally to the program and then program can be recompiled.
The inputs are given in the x.dai les. Run the pot.bat le by typing
pot x
This will leaves the output in the le x.dao. The pot.bat does the following:
del x.dai
copy x.dai pot.dai
wspot
rename pot.dao x.dao
del pot.dai
In the following I provide some sample inputs for calculating nucleon bound state
and scattering cross sections Some output is given in the screen and a more detailed
output is given in the le x.dao.
8 THE WOODS-SAXON POTENTIAL 92
8.2.1 Example for bound states
To obtain the bound states for neutrons in
40
Ca the input is:
40 20 1 0
0. 0. 0. 0.
0 0 1 0. 0. 0.
0 1 3 0. 0. 0.
0 1 1 0. 0. 0.
1 0 1 0. 0. 0.
0 2 5 0. 0. 0.
0 2 3 0. 0. 0.
0 0 0 0. 0. 0.
The rst line is (A, Z) of the nucleus and (a, z) of the nucleon. So a proton would be
(a, z) = (1,1). After the blank line the next lines are a list of n
r
, , 2j with a 0 in the
last line. The output will give
n
r
, , 2j, spe, k, kt, rms, rmst
for each orbit, where spe is the single-particle energy, k is the expectation value of the
kinetic energy < T >, and rms is

< r
2
>. The program will indicate if the orbit is
unbound. Kt means

i
(2j
i
+ 1) < T >
i
where the sum runs over all orbits up to the current one, with the assumption that
there are (2j
i
+ 1) nucleons in orbit j
i
. At the end this will tell you the total kinetic
energy for neutrons (if the complete set is included in the input). Rmst is obtained
from:
x
1
=

i
(2j
i
+ 1) < r
2
>
i
x
2
=

i
(2j
i
+ 1)
where the sum runs over all orbits up to the current one. Rmst =
_
(x
1
/x
2
) is the
total rms radius for all orbits up to the current one.
8 THE WOODS-SAXON POTENTIAL 93
8.2.2 Changing the potential parameters
In order to change the parameters of the central spin-independent potential the input
can be modied to be:
40 20 1 0
0. 0. 0. 0.
0 0 1 0.9 0.7 1.3
0 1 3 0. 0. 0.
0 1 1 0. 0. 0.
1 0 1 0. 0. 0.
0 2 5 0. 0. 0.
0 2 3 0. 0. 0.
0 0 0 0. 0. 0.
where V
N
= 0.9 modies the central potential strength to V

o
= V
N
V
o
, and a
o
and r
o
are changed to a

o
= 0.7 and r

o
= 1.3. These value are used for this and all following
orbits. Other parameters can be changed by modifying the program.
8 THE WOODS-SAXON POTENTIAL 94
8.2.3 Width of an unbound state resonance
For a given -j value the program calculates the phase shift (E) and the scattering
cross section (E) as a function of (positive) energy, where
(E) =
4(2 + 1)
k
2
[sin()]
2
where E = h
2
k
2
/2m (m is the mass of the nucleon).
As an example, the input for the 0d
5/2
proton resonance in
13
N.
12 6 1 1
0.1 10.0 0. 0.
0 2 5 0. 0. 0.
0 0 0 0. 0. 0.
The rst line is (A, Z) of the core nucleus and (a, z) of the nucleon. The next line
contains emin, emax for the energy interval over which to look for a resonance. The
next line contains n
r
, , 2j, and this is followed by 0 which stops program.
If a resonance peak is found in that interval, the program iterates over a series
of steps in which this energy interval is narrowed down until the full resonance curve
is found. It outputs EI and G where EI is the resonance energy peak and G= is the
full width at half maximum of the resonance. This particular input gives EI=0.99
MeV and =0.010 MeV. The output for the cross section vs E is given in the output
le POT.DAO.
For the scattering n
r
is not a quantum number and the value in the input has no
eect on the output. The value given on the detailed output in POT.DAO indicates
how often the wavefunction crosses zero. Usually the resonance is a quasi-bound
state corresponding to a particular the n
r
value of interest. One should check in the
output that the correct n
r
value is given. If not, the correct value may correspond to
a dierent energy range (or to a dierence value for V
N
discussed in the next section).
A resonance width is useful is the resonance is reasonably sharp. Broad reso-
nances will be asymmetry and the width itself is not a full measure of its nature.
Rather one must look at the full function of vs E. The program is probably not
accurate below about 0.1 MeV which is emin in the above input. Also narrow reso-
nances (if they exist) are at an energy below the Coulomb plus centrifugal barrier.
For these light nuclei this is not more than 10 MeV which is what determines emax
in the input.
8 THE WOODS-SAXON POTENTIAL 95
8.2.4 Width of an unbound state resonance at a xed energy
Experimentally one observes a 5/2
+
resonance at 3.547 MeV in
13
N with a one-proton
resonance energy of 1.60 MeV (the neutron separation energy is 1.60 MeV). This is
below the energy (0.99 MeV) obtained from the standard potential. We could ask
if the potential is slightly adjusted so that the resonance is exactly at 1.60 MeV then
what is the width? We need to slightly reduce the potential in order to increase the
resonance energy. We could so this by hand, for example:
12 6 1 1
0.1 6.0 0. 0.
0 2 5 0.96 0. 0.
0 0 0 0. 0. 0.
where the potential depth has been multiplied by 0.96 gives the resonance energy at
1.69 MeV (closer but not exactly the required value). We would then change V
N
by
trial and error until the resonance is at the correct position. However, the program
will automatically search for the correct V
N
by the following input:
12 6 1 1
1.60 1.60 0.8 1.2
0 2 5 0. 0. 0.
1. 2. 0. 0.
0 2 5 0. 0. 0.
0. 0. 0. 0.
The input 1.60,1.60,0.8,1.2 will search on V
N
over the interval 0.8 to 1.2 to put the
0d
5/2
resonance at exactly 1.6 MeV. Then the input 1.0,2.0 will search over the energy
interval 1.0 to 2.0 MeV with the new value of V
N
to nd the width of the resonance
at 1.6 MeV. The result is 64 keV which is not far from the observed width of 47(7)
keV.
REFERENCES 96
References
[1] Thomas, Nature 107, 514 (1926); Frenkel, Zeits. fur Phys. 37, 243 (1926).
[2] A. Bohr and B. R. Mottelson, Nuclear Structure, Vol. I, (W. A. Benjamin, 1969),
Nuclear Structure Vol. II, (W. A. Benjamin, 1975).
9 THE GENERAL MANY-BODY PROBLEM FOR FERMIONS 97
9 The general many-body problem for fermions
The Schroedinger equation for a system of n particles each of mass m in a central
potential, U
k
, which interact via a two-body potential, V
kl
, is
H [ >= E [ >, (9.1)
with
H =
_
n

k=1
(T
k
+U
k
)
_
+
_
_
_
n

k<l
V
kl
_
_
_
H
(0)
+W, (9.2)
where T
k
is the kinetic energy operator, ( h
2
/2m)
2
k
, and is the n-particle wave
function represented in coordinate space by (r
1
. . . r
n
). In applications to atomic
physics, U
k
is the Coulomb potential between the nucleus and the electrons. In
applications to nuclear physics one takes U
k
= 0. The coordinate space forms of U
k
and V
kl
will be denoted by U(r
k
) and V (r
k
r
l
), respectively
The standard technique for solving this equation involves rst solving the simple
problem
H
(0)
[
a
>= E
(0)
a
[
a
>, (9.3)
which has the solution
[
a
>=
n

k=1
[
k
>=

1
(r
1
)

2
(r
2
)

3
(r
3
) . . .
n
(r
n
), (9.4)
with
E
(0)
a
=
n

k=1

k
. (9.5)
The [ > are solutions of the single-particle equation
(T +U) [ >=

[ >, (9.6)
and label a represents one particular choice for the set of quantum numbers
k
. There
are innitly many dierent sets a which can be ordered by their total energy E
(0)
a
.
The states

are assumed to form a complete orthonormal single-particle basis:

<
k
[
l
>=
kl
The states are assumed to form a complete orthonormal basis for the solution of Eq.
(9.1). In a spherical basis would represent the set of quantum numbers (n
r
, , m

).
If the spin coupling is included the quantum numbers would be (n
r
, , j, m
j
).
9 THE GENERAL MANY-BODY PROBLEM FOR FERMIONS 98
The full exact many-particle problem can then be reformulated in terms of the
basis by expanding over the basis,
[ >=

a
c
a
[
a
>, (9.7)
and then solving the matrix equation

a
<
b
[ H [
a
> c
a
= Ec
b
. (9.8)
Alternatively, solutions can be obtained with perturbation theory. Nondegenerate
Rayleigh-Schroedinger perturbation theory gives
E
a
= E
(0)
a
+E
(1)
a
+E
(2)
a
. . . (9.9)
and
[
a
>=[
(0)
a
> + [
(1)
a
> . . . , (9.10)
where
E
(0)
a
=<
a
[ H
(0)
[
a
>, (9.11)
E
(1)
a
=<
a
[ W [
a
>, (9.12)
E
(2)
a
=

b=a
<
a
[ W [
b
><
b
[ W [
a
>
E
(0)
a E
(0)
b
, (9.13)
and
[
(0)
a
>=[
a
>, (9.14)
[
(1)
a
>=

b=a
<
b
[ W [
a
>
E
(0)
a E
(0)
b
[
b
> . (9.15)
By either method, the ingredients of the calculations are the matrix elements of one-
and two-body operators between the many-body basis states.
These methods of solution are limited by the fact that in practice one must deal
with nite matrices and/or that the perturbation series must converge. Improved
solutions by the above methods can often be achieved by adding and subtracting
single-particle potential in Eq. (9.2) to obtain
H =
_
n

k=1
(T
k
+U
k
+U

k
)
_
+
_
_
_
n

k<l
V
kl

k=1
U

k
_
_
_
H
(0)
+W. (9.16)
U

k
is chosen in order to minimize the size of o-diagonal matrix elements of W in
the basis, thus providing a means of truncating the matrix and/or improving the
convergence of the perturbation expansion. One systematic way of choosing U

k
is the
Hartree-Fock method.
9 THE GENERAL MANY-BODY PROBLEM FOR FERMIONS 99
When the particles are identical fermions, the wave functions must be totally
antisymmetric with respect to the interchange of the coordinates of any two particles.
This is arranged by constructing Slater determinants within the basis. For the two-
particle system the Slater determinant is given by
[
a
>=[ >=
1

(1)

(2)

(1)

(2)

=
1

(1)

(2)

(1)

(2) . (9.17)
or, in Dirac notation, as
[
a
>=[ >=
1

2
[ >
p
[ >
p
, (9.18)
where the subscript p denotes the product wave function given in coordinate space
by Eq. (9.4). For the n-particle system there are n! terms and the normalization is
1/

n! . The Slater determinant enforces the Pauli principle - the fact that no two
members of the set of quantum numbers
k
can be identical or the wave function will
vanish.
The matrix elements of H between these Slater determinants are straightforward
but tedious to evaluate. However, the diagonal matrix elements are particularly
simple. The diagonal matrix element for the one-body operator T
k
+U
k
is
<
a
[
n

k=1
(T
k
+U
k
) [
a
>=

< [ T +U [ >, (9.19)

where the sum over runs over the set of n occupied single-particle states in
a
, and
< [ T +U [ > stands for the single-particle matrix element

(r )
_
h
2
2m

2
+U(r )
_

(r ) d. (9.20)
The diagonal matrix element for the two-body operator V
kl
is
<
a
[

k<l
V
kl
[
a
>=

<
< [ V [ >
=
1
2

=
< [ V [ >=
1
2

< [ V [ > . (9.21)

where the sums over and run over the set of n occupied single-particle states in

a
. The summation over ( ,= ) can be replaced by the unrestricted summation
() because the two-particle wave function automatically vanishes when = .
9 THE GENERAL MANY-BODY PROBLEM FOR FERMIONS 100
The two-body matrix element is given by
< [ V [ >=< [ V [ >
p
< [ V [ >
p
=
_ _

(r
1
)

(r
2
)V (r
1
r
2
)

(r
1
)

(r
2
) d
1
d
2

_ _

(r
1
)

(r
2
)V (r
1
r
2
)

(r
1
)

(r
2
) d
1
d
2
(9.22)
where V (r
1
r
2
) = V (r
2
r
1
) has been used, and the subscript p indicates that the product
wave functions of Eq. (9.4) are used. The two terms in Eq. (9.22) are referred to as
the direct and exchange terms, respectively. Evaluation of the most general case will
be considered in the framework of the creation and destruction operators.
10 CONSERVED QUANTUM NUMBERS 101
10 Conserved quantum numbers
10.1 Angular momentum
The hamiltonian is a scalar in the spacial coordinates - the expectation value does
not depend on the orientation of the system. Thus the hamiltonian commutes with
total angular momentum operators:
[H,

J
2
] = 0, (10.1)
[H,

J
z
] = 0. (10.2)
The many-body wave functions are eigenstates of

J
2
and

J
z
with:

J
2
[ >= J(J + 1) [ >, (10.3)

J
z
[ >= M [ >, (10.4)
where there are (2J + 1) values of M going from M = J up to M = +J in integer
steps. Thus one of our goals is to construct wave functions which have good J and
J
z
. One can construct basis states which have good J and J
z
- this is called the
J-scheme. Sometimes one constructs a basis which has only good J
z
but not good
J - this is called the M-scheme. If this M-scheme basis is complete with regard to
rotations in the J-space, then the eigenstates will have good J and J
z
. For nuclei
the total angular momentum is sum of the orbital and spin angular momenta of all
nucleons.
10.2 Parity
The strong and electromagnetic hamiltonians conserve parity. Thus the eigenstates
can be broken down into two classes of states labeled by their parity =+1 or =-1,
and the hamiltonian does not mix these two classes. For nuclear structure the total
parity originates from the the intrinsic parity of the nucleon which is
int
=+1 and the
parities associated with the orbital angular momenta

= (1)

. The total parity is

the product over all nucleons
=
i

int
(i)

(i) =
i
(1)

i
(10.5)
One usually constructs basis states and eigenstates which which have a denite parity.
The weak hamiltonian does not conserve parity. When considered as a perturba-
tion the matrix elements of the weak hamiltonian are typically on the order of a few
10 CONSERVED QUANTUM NUMBERS 102
eV or less. Thus, parity mixing can always be calculated in rst-order perturbation
theory.
The eects of parity nonconservation can be observed in several types of experi-
ment in which the observable must be zero in the absence of parity nonconservation.
For example, when parity is conserved a L = 1 electromagnetic transition must ei-
ther be M1 if the two states do not change parity (
i

f
=+1), or E1 if the two states
change parity (
i

f
=-1). The angular distribution of both E1 and M1 gamma rays
is symmetric around the spin axis of the nucleus. When parity is not conserved then
there can be interference between E1 and M1 which leads to an left-right asymmetry
in the angular distribution.
10.3 Isospin
Since the spin of the proton and neutron are both 1/2 and the masses are nearly the
same, it is useful to think of them as members of an isospin doublet of the nucleon.
Thus they have isospin t = 1/2 and projection t
z
= 1/2 and t
z
= +1/2 for the
proton and neutron, respectively. The isospin operator

t for the isospin degrees of
freedom is analogous to the spin operator s for the spin degrees of freedom. For many
nucleons the total isospin and its z projection are given by

T =

t(i). (10.6)

T
z
=

t
z
(i). (10.7)
The strong interaction conserves isospin, that is, the expectation value of H
s
does
not depend on the orientation in isospin space. Thus H
s
commutes with the total
isospin operators:
[H
s
,

T
2
] = 0 (10.8)
[H
s
,

T
z
] = 0 (10.9)
The many-body wave functions are eigenstates of

T
2
and

T
z
with:

T
2
[ >= T(T + 1) [ >, (10.10)

T
z
[ >= T
z
[ >, (10.11)
where there are 2T + 1 values of T
z
going from T
z
= T up to T
z
= +T in integer
steps. For a nucleus with Z protons and N neutrons, T
z
= (N Z)/2. The sign
convention here is chosen so that the more common neutron-rich nuclei will have
positive T
z
(but other books may use a dierence convention). For a given T
z
value
10 CONSERVED QUANTUM NUMBERS 103
the minimum value of T is T
min
=[ T
z
[ and the maximum value is T
max
= A/2. The
lowest level in a nucleus usually has T = T
z
(there are a few exceptions for odd-odd
nuclei with N = Z). This is related to the fact that the nuclear interaction in the
T = 0 two-nucleon system is a stronger that in the T = 1 system (only the deuteron
has a bound state).
The experimental evidence for isospin conservation comes from a comparison of
the proton-proton, proton-neutron and neutron-neutron scattering lengths in their
T = 1 state, and the observation that the energy levels of nuclei can be organized
into T multiplets of (2T + 1) nuclei with T
z
= T up to T
z
= T.
One of the special consequences of isospin conservation is mirror symmetry. The
energy spectrum of a nucleus should be identical to that of another nucleus in which
the protons and neutrons numbers are interchanged. These are called mirror nuclei.
There are many examples of mirror nuclei whose spectra can be compared in this
way. The deviation from mirror symmetry is attributed to the Coulomb interaction.
The Coulomb hamiltonian does not conserve isospin since it only enters into the
interaction between protons. If the Coulomb interaction is treated as a perturbation,
then its o-diagonal matrix elements are typically on the order of 100 keV or less. For
many nuclear structure problems this is small (at least compared to the uncertainty
in the strong interaction matrix elements) and we can start with basis states and
eigenfunction which are constructed to have good isospin. Isospin non-conservation
is often treated in rst-order perturbation theory.
The diagonal matrix elements are relatively large. The diagonal energies can be
estimated from liquid-drop (ld) model for the Coulomb interaction for Z protons in
a sphere of radius R = r
o
A
1/3
:
E
c
(ld) =
3
5
e
2
Z
2
R
=
3e
2
5r
o
Z
2
A
1/3
. (10.12)
An isobaric mass multiplet for a state with isospin T are those (2T +1) states in
dierent nuclei which have T
z
= T up to T
z
= T. The isobaric mass multiplet equa-
tion (IMME) is an expression for the masses of these levels in terms of an expansion
in T
z
. We can also write the IMME in terms of binding energies:
BE(T, T
z
) = a +bT
z
cT
2
z
. (10.13)
where the signs are chosen so that a, b and c will be positive. The strong interaction
does not depend on T
z
, and only contributes to the rst term in Eqs. (10.13). When
the Coulomb interaction is treated in rst order-perturbation theory one can show
that its contribution to the binding energy does not have terms higher than T
2
z
(This
10 CONSERVED QUANTUM NUMBERS 104
comes from the fact that the Coulomb interaction can be written in terms of isospin
tensors of rank 0, 1 and 2. One then uses the Wigner-Eckart theorem to show T
2
z
is the highest power of T
z
which is allowed.) The b coecient of the IMME can be
determined from the binding energy dierences between mirror nuclei:
b =
BE(T, T
z
= T) BE(T, T
z
= T)
2T
. (10.14)
The liquid-drop estimate for the Coulomb contribution to these terms comes from
Eq. (10.12) with Z = (A/2) T
z
:
BE
c
(ld) =
3e
2
5R
_
A
2
4
AT
z
+T
2
z
_
. (10.15)
Thus
b(ld) =
3e
2
A
5R
and c(ld) =
3e
2
5R
. (10.16)
The experimental masses (binding energies) have been used to obtain the coe-
cients a, b and c in Eq. (10.13) with results given in Ref. [1]. States with T = 1/2 are
called isobaric doublets, and only the rst two two terms in Eq. (10.13) are needed
to t the two binding energies. States with T = 1 are called isobaric triplets, and
all three terms are needed. States with T = 3/2 are called isobaric quartets. For
quartets, the most general expansion in terms of T
z
would require a term of the form
dT
3
z
. However from the experimental binding energies the coecient d is zero within
experimental error, except for small values for A = 9 and A = 13 [1]. This fact is
consistent with the treatment of the Coulomb interaction as a rst-order correction
to the energy.
The b coecients obtained from the experimental binding energy are shown in
Fig. (10.1). They are compared with the prediction of the liquid drop model. The
liquid drop model is always higher than experiment. The reason is that the observed
displacement energy is related to a change in proton-neutron occupancy of the orbits
near the Fermi surface (the valence orbits that are the most loosely bound) which have
an rms radius which is larger than the average rms radius implicit in the liquid drop
model (R in the denominator of Eq. (10.16) is eectively larger). One can observe in
Fig. 1 some small downward glitches in the data at A = 16 and A = 40 corresponding
to the magic numbers 8 and 20, respectively. This is due to the increase in rms radius
of the orbit at the Fermi surface going from the p to sd (A = 16) and sd to pf
(A = 40) valence shells.
10 CONSERVED QUANTUM NUMBERS 105
b

(
M
e
V
)
A
0
2
4
6
8
10
12
0 10 20 30 40 50 60 70
Figure 1: b-coecients for the IMME. The symbols show the values obtained
from experimental binding energies for T = 1/2 (29 lled circles), T = 1 (24
squares), T = 3/2 (17 crosses) and T = 2 (9 pluses). The line is the prediction
of the liquid drop model.
REFERENCES 106
References
[1] W. Benenson and E. Kashy, Rev. Mod. Phys. 51, 527 (1979); M. S. Anthony, J.
Britz, J. R. Bueb and A. Pape, At. Data Nucl. Data Tables 33, 447 (1985).
11 QUANTUM NUMBERS FOR THE TWO NUCLEON SYSTEM 107
11 Quantum numbers for the two nucleon system
11.1 Two neutrons or two protons
There are no bound states for two neutrons. However, there is a strong low-energy
s-wave scattering resonance. First we will construct the antisymmetric wavefunction
for two neutrons in the =0 state. The wavefunctions are a product of the space, spin
and isospin parts:
f(r) Y

m
( r) (S) (T). (11.1)
Initially we will consider only the space and spin part. The parity associated with the
change of phase of the spherical harmonics Y when r is replaced by r is =(1)

.
For a two-particle system the symmetry under the interchange of the two particles
P
12
is related to the parity since
P
12
r = P
12
[r
1
r
2
] = [r
2
r
1
] = r. (11.2)
For l = 0, =+1, and the spatial wave function is symmetric under the inter-
change of the two neutrons. In order to make the entire wavefunction antisymmetric
the spin wave function must be antisymmetric. The spin wave functions
[ >=[ s = 1/2, s
z
= 1/2 >, (11.3)
[ >=[ s = 1/2, s
z
= 1/2 > . (11.4)
The neutron spins can be coupled to make wavefuncitons (S, S
z
) with total angular
momentum S by using the Clebsch-Gordan coecients < 1/2, s
z
, 1/2, s

z
[ S, S
z
>
(1, 1) =[ (1) > [ (2) >, (11.5)
(1, 0) =

1
2
[[ (1) > [ (2) > + [ (1) > [ (2) >] , (11.6)
(1, 1) =[ (1) > [ (2) >, (11.7)
(0, 0) =

1
2
[[ (1) > [ (2) > [ (1) > [ (2) >] , (11.8)
where (1) indicates that neutron 1 is in the state . The S = 1 wavefunctions are
symmetric under P
12
, and S = 0 is antisymmetric. Thus the spin symmetry under
P
12
is (1)
S+1
. To make the entire =0 wavefunction antisymmetric we must take the
S = 0 solution. Thus the quantum numbers for the = 0 state of the two neutrons
are:
1
S
0
: nn, = 0, S = 0, J = 0 and = +1. (11.9)
11 QUANTUM NUMBERS FOR THE TWO NUCLEON SYSTEM 108
A similar analysis for the scattering in the l = 1 state would give:
3
P
J
: nn, = 1, S = 1, J = 0, 1, 2 and = 1, (11.10)
where the rst symbol in Eqs. (11.9) and (11.10) indicates
2S+1

J
, with (2S +1) = 1
for the singlet state and (2S + 1) = 3 for the triplet state. The values of J are
restricted by the triangle condition for the coupling of and S.
In general the total symmetry under P
12
must be negative:
(1)
+S+1
= 1, (11.11)
which means that +S must be even for the two neutrons. If we add isospin to the
two-neutron system there is nothing new since the two neutrons must be in the state
(T = 1, T
z
= 1) =[ t = 1/2, t
z
= 1/2 (1) > [ t = 1/2, t
z
= 1/2 (2) >, (11.12)
which is symmetric under P
12
. The same results apply to the proton-proton system.
11.2 Proton-neutron
If we consider a proton and neutron in an =0 state there is no antisymmetry re-
quirement since these are not identical particle. Thus both S values are allowed. For
=0:
1
S
0
: pn, = 0, S = 0, J = 0 and = +1. (11.13)
3
S
1
: pn, = 0, S = 1, J = 1 and = +1. (11.14)
The J = 1 wave function corresponds to the bound state of state of the deuteron
which has J

= 1
+
. A similar set of states exist for higher values.
11.3 Two nucleons
When isospin is considered we also have to include the isospin coupling to T = 0 and
1 which in analogy with the spin coupling is:
(1, 1) =[ a(1) > [ a(2) >, (11.15)
(1, 0) =

1
2
[[ a(1) > [ b(2) > + [ b(1) > [ a(2) >] , (11.16)
(1, 1) =[ b(1) > [ b(2) >, (11.17)
11 QUANTUM NUMBERS FOR THE TWO NUCLEON SYSTEM 109
(0, 0) =

1
2
[[ a(1) > [ b(2) > [ b(1) > [ a(2) >] , (11.18)
where:
[ a >=[ t = 1/2, t
z
= 1/2 >, (11.19)
[ b >=[ t = 1/2, t
z
= 1/2 > . (11.20)
The symmetry under P
12
is (1)
T+1
.
The total symmetry under P
12
must be negative:
(1)
+S+1+T+1
= (1)
+S+T
= 1, (11.21)
which means that +S +T must be odd. Thus with isospin the proton-neutron state
with =0 and S = 0 must have T = 1 and it is the T
z
= 0 member of the T = 1
triplet, the other two being the neutron-neutron and proton-proton states
1
S
0
: T = 1, = 0, S = 0, J = 0 and = +1. (11.22)
The state with =0 and S = 1 must have T = 0
3
S
1
: T = 0, = 0, S = 1, J = 1 and = +1. (11.23)
For =1 and =2 we have
3
P
J
: T = 1, = 1, S = 1, J = 0, 1, 2 and = 1. (11.24)
1
P
1
: T = 0, = 1, S = 0, J = 1 and = 1. (11.25)
1
D
2
: T = 1, = 2, S = 0, J = 2 and = +1. (11.26)
3
D
J
: T = 0, = 2, S = 1, J = 1, 2, 3 and = +1. (11.27)
In general only the total angular momentum J is a good quantum number. For
S = 0, J = is unique. For S = 1, the wavefunction for a given J value can have
= J 1 or = J +1. Thus bound state of the deuteron with J

= 1
+
is in general
a mixture of the
3
S
1
and
3
D
1
states. =0 is the dominant part of the wavefunction
but there is some =2 (d-state) admixture.
12 THE HARTREE-FOCK APPROXIMATION 110
12 The Hartree-Fock approximation
12.1 Properties of single Slater determinants
Before deriving the Hartree-Fock equations is it useful to give some special cases of
some diagonal matrix elements. The eigenstates of H
(0)
can be ordered with respect
to the total energy E
(0)
, with the lowest energy state being the one in which all of the
particles occupy the lowest energy set of single-particle states allowed by the Pauli
principle. This lowest-energy state will be denoted by [ C >. (This notation derives
from the fact that it will be associated with the closed-shell conguration.) The n
single-particle states occupied in [ C > will be labeled by , , . . . . The total energy
E(C) of the state [ C > is
E(C) =< C [ H [ C >=

< [ T [ > +
1
2

< [ V [ > . (12.1)

I will use i and j to label specic states above or below the fermi surface which
will be considered explicitly in the summations. The total energy of a system with
the conguration [ C > plus one particle in the state i above the fermi surface (a
state unoccupied in [ C >) is
E(Ci) =< Ci [ H [ Ci >
= E(C)+ < i [ T [ i > +

< i [ V [ i > . (12.2)

The dierence between E(Ci) and E(C) will be denoted by, (i), the single-particle
energy:
(i) = E(Ci) E(C) =< i [ T [ i > +

< i [ V [ i > . (12.3)

The total energy of a system with the conguration [ C > plus particles in the states
i and j above the fermi surface is
E(Cij) =< Cij [ H [ Cij >
= E(C)+ < i [ T [ i > +

< i [ V [ i >
+ < j [ T [ j > +

< j [ V [ j > + < ij [ V [ ij >

= E(C) +(i) +(j)+ < ij [ V [ ij > (12.4)
The total energy of a system with the conguration [ C > with one particle
absent in the state i (normally occupied in [ C >) is
E(Ci
1
) =< Ci
1
[ H [ Ci
1
>
12 THE HARTREE-FOCK APPROXIMATION 111
= E(C) < i [ T [ i >

< i [ V [ i > . (12.5)

The dierence between E(Ci
1
) and E(C) will be denoted by, (i):
(i) = E(C) E(Ci
1
) =< i [ T [ i > +

< i [ V [ i >, (12.6)

which has the same form as the single-particle energy. Finally, the total energy of a
system with the conguration [ C > plus particles absent in the states i and j is
E(Ci
1
j
1
) =< Ci
1
j
1
[ H [ Ci
1
j
1
>
= E(C) (i) (j)+ < ij [ V [ ij > (12.7)
12.2 Derivation of the Hartree-Fock equations
In the Hartree-Fock approximation, E(C) is minimized with respect to variation of
the single-particle wave functions
i
(r ) [or equivalently with respect to
i
*(r )]. With
the coordinate space matrix elements of T and V , one obtains

i
(r )
_
E(C)

_
[

(r
1
) [
2
d
1
_
= 0
= T
i
(r )
i

i
(r ) +
1
2
_

(r
2
) V (rr
2
)
i
(r )

(r
2
) d
2
+

(r
1
) V (r
1
r )

(r
1
)
i
(r ) d
1

(r
2
) V (rr
2
)

(r)
i
(r
2
) d
2

(r
1
) V (r
1
r )
i
(r
1
)

(r ) d
1
_
, (12.8)
where are Lagrange multipliers which are introduced to enforce the normalization.
Using V (r
1
r
2
) = V (r
2
r
1
), this reduces to
T
i
(r ) +
_

(r
1
)V (rr
1
)

(r
1
) d
1
_

i
(r )

_
_

(r
1
)V (rr
1
)

(r )
_

i
(r
1
) d
1
=
i

i
(r ). (12.9)
This nonlocal dierential equation can be used to solve for
i
and
i
(r ). It can be
solved in an iterative fashion: (i) choose some initial guess for
i
(r ) and calculate the
integrals as a function of r, (ii) solve the dierential equation for
i
and
i
(r ), (iii)
recalculate the integrals, and (iv) iterate until the
i
and
i
(r ) converge.
12 THE HARTREE-FOCK APPROXIMATION 112
Multiplying Eq. (12.9) by

i
(r ) on both sides and integrating gives
< i [ T [ i > +

< i [ V [ i >=
i
= (i), (12.10)
where the Lagrange multiplier has been equated to the single-particle energy by com-
parison to Eqs. (12.3) and (12.6). The term

< i [ V [ i > can be identied as

the expectation value of the mean-eld potential U

.
< i [ U

[ i >=< i [ U
HF
[ i >=

< i [ V [ i > . (12.11)

Thus, the full Hartree-Fock hamiltonian is
H = H
(0)
+W, (12.12)
with
H
(0)
=

k
(T +U
HF
)
k
, (12.13)
and
W =

kl
V
kl

k
U
HF
k
. (12.14)
The state [ C > with the Hartree-Fock condition of Eq. (12.11) enforced will be
denoted by [
HF
>. The zeroth-order and rst-order matrix elements are
E
(0)
HF
=<
HF
[

k
(T +U
HF
)
k
[
HF
>=

(), (12.15)
and
E
(1)
HF
=<
HF
[ W [
HF
>=<
HF
[

kl
V
kl

k
(U
HF
)
k
[
HF
>
=
1
2

< [ V [ >

< [ U
HF
[ >
=
1
2

< [ V [ > . (12.16)

The total unperturbed energy can thus be expressed in several ways:
E
(0)
HF
+E
(1)
HF
=

()
1
2

< [ V [ >
=

< [ T [ > +
1
2

< [ V [ >
12 THE HARTREE-FOCK APPROXIMATION 113
=
1
2
_

() +

< [ T [ >
_
. (12.17)
It is to be noted that the Hartree-Fock condition does not make E
(1)
vanish. The
advantage of the Hartree-Fock procedure is that an important class of matrix elements
<
HF
[ W [ > which enter into the second-order corrections vanish namely, all
of those for which diers from
HF
by the addition of one particle above the fermi
surface and the removal of one particle below the fermi surface. These are called one-
particle one-hole, 1p-1h, congurations. The most important corrections are those in
which the diers from
HF
by addition of two-particle above the fermi surface and
the removal of two particles below the fermi surface (2p-2h congurations).
12.3 Examples of single-particle energies
The Hartree-Fock model works best for those nuclei where there is a large gap at the
fermi surface for both protons and neutrons. The total energies for the closed shell,
one-particle and one-hole congurations are the interaction energies E measured for
the respective, nuclei, where BE = E. For example, if we take
16
O as a doubly
closed shell nucleus, then energies obtained from the ground state binding energies
are:
E(
16
O) = E(C) = 127.619 MeV
E(
17
O) = E[C, 0d
5/2
neutron] = 131.763 MeV
E(
17
F) = E[C, 0d
5/2
proton] = 128.220 MeV
E(
15
O) = E[C, (0p
1/2
)
1
neutron] = 111.956 MeV
E(
15
N) = E[C, (0p
1/2
)
1
proton] = 115.492 MeV
where the n, , j values are inferred from the spin-parity of the odd-even ground states.
The experimental single-particle energies for these states are thus:
(0d
5/2
neutron) = E(
17
O) E(
16
O) = 4.144 MeV
(0d
5/2
proton) = E(
17
F) E(
16
O) = 0.601 MeV
(0p
1/2
neutron) = E(
16
O) E(
15
O) = 15.663 MeV
(0p
1/2
proton) = E(
16
O) E(
15
N) = 12.127 MeV
The single-particle energies for other states can be inferred form the energies E asso-
ciated with excited states in the A=15 and A=17 nuclei.
It will be shown in a homework that Eq. (12.17) is in fact not satised by ex-
perimental data in nuclear physics, when the are taken from experiment and when
12 THE HARTREE-FOCK APPROXIMATION 114
the kinetic energies are calculated. This means that higher-order corrections to the
Hartree-Fock are important. If these corrections are taken into account by using
an eective hamiltonian, this hamiltonian will need to include three-body and/or
density dependent terms. An example is the Skyrme interaction which includes a
density-dependent interaction.
12 THE HARTREE-FOCK APPROXIMATION 115
12.4 Results with the Skyrme hamiltonian
The Skyrme approximation [1], [2] is an s and pwave expansion of an eective
two-body interaction together with an swave density dependent interaction:
V
Skyrme
= t
0
(1 +x
0
P

) +
1
2
t
1
(1 +x
1
P

) (k

2
+ k
2
)
+t
2
(1 +x
2
P

) k

k +
1
6
t
3
(1 +x
3
P

(R)
+iW
0
(
i
+
j
) k k +V
Coul
, (12.18)
where = (r
i
r
j
), k = (1/2i)(
i

j
) is the relative momentum operator acting
on the wave function to the right and k

is the adjoint of k. P

is the spin-exchange
operator and R = (r
i
+ r
j
)/2. The form of the Skyrme interaction allows one to
calculate the potentials analytically in terms of the densities which makes the self-
consistent calculations fast.
The Skyrme interaction results in a non-local potential for protons (q = p) and
neutrons (q = n) given by U
q
(r) +U

q
(r) with
U
q
(r) = t
0
__
1 +
x
0
2
_

_
x
0
+
1
2
_

q
_
+
t
1
8
__
1 +
x
1
2
_
[2 3()]
_
x
1
+
1
2
_
[2
q
3(
q
)]
_
+
t
2
2
__
1 +
x
2
2
_
[2 + ()] +
_
x
2
+
1
2
_
[2
q
+ (
q
)]
_
+
t
3
6
__
1 +
x
3
2
_

_
x
3
+
1
2
_

q
_

W
0
2
(J +J
q
)
+U
Coul
q
(r) +U
so
q
(r)[ ] (12.19)
and
U

q
(r) =
_
t
1
4
__
1 +
x
1
2
_

_
x
1
+
1
2
_

q
_
+
t
2
4
__
1 +
x
2
2
_
+
_
x
2
+
1
2
_

q
__
, (12.20)
where the spin-orbit potential is:
U
so
q
(r) =
1
r
_
W
0
2
_
d
dr
( +
q
)
_
+
1
8
[(t
1
t
2
)J
q
]
1
8
[t
1
x
1
+t
2
x
2
] J
_
(12.21)
12 THE HARTREE-FOCK APPROXIMATION 116
The U

term of Eq. (12.20) can be combined with the kinetic energy operator to
write Eq. (12.9) in terms of the a Schroedinger-like equation with an eective mass:
_

h
2
2m

q
(r)
+U
q
(r)
_

i,q
(r) =
i,q

i,q
(r). (12.22)
where the eective mass is dened by:
h
2
2m

q
(r)
=
h
2
2m
+
t
1
4
__
1 +
x
1
2
_

_
x
1
+
1
2
_

q
_
+
t
2
4
__
1 +
x
2
2
_
+
_
x
2
+
1
2
_

q
_
(12.23)
The densities in these equations are

q
(r) =

[
q
(r) [
2
, (12.24)

q
(r) =

[
q
(r) [
2
(12.25)
J
q
(r) = i

q
(r)[
q
(r)], (12.26)
(r) =
p
(r) +
n
(r), (12.27)
(r) =
p
(r) +
n
(r), (12.28)
J(r) = J
p
(r) +J
n
(r), (12.29)
and
(f) =
1
r
_
d
2
dr
2
rf(r)
_
, (12.30)
where the derivative operates only inside the brackets.
Eq. (12.22) can be rewrittin in terms of the Schroedinger equation with an energy-
dependent potential [3]:
_

h
2
2m
q

2
+U

q
(r, )
_

i,q
(r) =
i,q

i,q
(r), (12.31)
where
U

q
(r, ) =
m

q
(r)
m
_
U
q
(r) +
1
2
_
d
2
dr
2
h
2
2m

q
(r)
_

q
(r)
2 h
2
_
d
dr
h
2
2m

q
(r)
_
2_
+
_
1
m

q
(r)
m
_

i,q
(12.32)
12 THE HARTREE-FOCK APPROXIMATION 117
Table 1: Values of the Skyrme parameters obtained with SKX .
Parameter SKX SKX error
0.5
t
0
1444.0 1.0
t
1
251.3 1.9
t
2
131.4 1.2
t
3
12043.9 18.1
W
o
149.4 2.7
x
0
0.364 0.021
x
1
0.521 0.171
x
2
0.131 0.013
x
3
0.088 0.051
The goal of the Skyrme HF formulation is to write the HF equations in terms
of a few parameters (the ten in Eq. (12.18)) that can be obtained from a least-
squares t to some selected set of experimental data. In principle one would like
to derive the Skyrme parameters from the experimental nucleon-nucleon interaction.
This involves understanding the eect of the trucation of the actual many-body wave
function to the closed-shell structure assumed in the derivation. It also requires an
understanding of the contributions of real three-body forces. There has not yet been
a quantitative derivation of the Skyrme parameters from the rst principles. The
Skyrme formulation is a specic type of density-functional model which have been
widely used in atomic, molecular and condensed matter physics. One can show that
such functionals exist even if they cannot be explicitly derived [4], [5].
The have been many attemps to obtain the Skyrme parameters from various
types of experimental data. In this book I will concentrate on two recent results.
One is the SKX hamiltonian which is obtained by applying the above equations
to the eleven closed-shell nuclei:
16
O,
34
Si,
40
Ca,
48
Ca,
48
Ni,
88
Sr,
100
Sn,
132
Sn and
208
Pb. The data include the binding energies of these nuclei, together with ve rms
charge radii and 65 single-particle energies. Ten parameters in Eq. (12.18) were
varied (although only six linear-combinations are well determined). The SKX spin-
orbit energy-density leaves out terms involving t
1
, t
2
, x
1
and x
2
, and a generalized
two-parameter spin-orbit force based upon the Hartree reduction was used [6]. SKX
uses the Friedman-Pandharipande neutron matter equation of state [7] as a constraint
(it has recently been shown that the neutron skin is sensitive to the properties of the
neutron equation of state [8], [9] ). SKX also introduces a new parameter which is
needed to reproduce the mirror displacement energies by the addition of a charge
symmetry breaking (CSB) interaction (SkXcsb [10] ). The values of the Skyrme
parameters for SKX are given in Table 1.
12 THE HARTREE-FOCK APPROXIMATION 118
Goriely et al. [11] obtained the parameter set called MSk7. In addition to
the formulation given above for closed-shell nuclei, the eect of nuclear pairing and
deformation were also taken into account. The MSk7 parameters are based on a t
to the binding energies of 1772 nuclei (the radii or single-particle energies were not
included). Of the ten parameters in Eq. (12.18), seven were varied (x
0
and x
1
were
xed at 0.5 and was xed at 1/3). The spin-orbit energy density retains the terms
involving t
1
, t
2
, x
1
and x
2
, a four-parameters -function pairing force was added, and
a two-parameter Wigner correction term was added. The MSk7 interaction was used
to calculate the binding energies and shapes of 9200 nuclei [11]. The results were
shown as the HF1 calculation in Chapter 1.
The relativistic mean-eld is an alternative to the Skyrme HF. The relativistic
description of nuclear systems uses a eld theoretical approach (quantum hydrody-
namics) where the interaction of nucleons is described by an exchange of mesons. [12]
I will use the non-linear parameter set NL3 [13] which gives a good description of
binding energies and radii.
12.4.1 Binding energies
The binding energies for the region of nuclei up to A = 60 vary by hundreds of MeV,
yet we consider theoretical calculations which reproduce experiment to the level of
several hundred keV to an MeV. The coast to coast situation for all nuclei between
the drip lines is illustrated in Fig. (12.1) where the BE obtained with the MSk7 HF
calculations [11] are shown for all nuclei between the proton and neutron drip lines
centered on Z = 20 (left-hand side) and N = 20 (right-hand side) and compared to
experiment where known.
The drip line is reached in each case when the derivative of the BE curve with
respect to proton or neutron number goes to zero. One observes in the bottom panels
of Fig. (12.1) an apparently featureless and smooth curve with the data in agreement
with theory. However, we are interested in a much higher level of detail which is
illustrated in the top panel by subtracting a smooth curve given by the liquid drop
model (LDM) from theory and experiment. This top curve brings out the detail
related the microscopic aspects of the nuclear shell model.
12 THE HARTREE-FOCK APPROXIMATION 119
B
E

(
M
e
V
)
N
100
200
300
400
6 10 14 18 22 26 30 34 38
Z=20
B
E

-

B
E
(
L
D
M
)

(
M
e
V
)
-5
0
5
10
15
20
Z
6 10 14 18 22 26 30
N=20
Figure 1: Theoretical (squares) and experimental (lled circles) binding energy
for Z = 20 as a function of neutron number (left-hand side) and for N = 20 as
a function of proton number (right-hand side). In the upper panels a smooth
curve given by the liquid drop model is subtracted from theory and experiment.
The magic numbers N = 20, N = 28 and Z = 20 are indicated by dashed lines.
12 THE HARTREE-FOCK APPROXIMATION 120
s
i
n
g
l
e
-
p
a
r
t
i
c
l
e

e
n
e
r
g
y

(
M
e
V
)

-24
-22
-20
-18
-16
-14
-12
-10
-8
-6
-4
-2
0
2
4
6
1s1
0d3
0f7
1p3
1p1
0f5
1s1
0d3
0f7
1p3
1p1
0f5
1s1
0d3
0f7
1p3
1p1
0f5
0p3
0p1
0d5
1s1
0d3
0p3
0p1
0d5
1s1
0d3
0p3
0p1
0d5
1s1
0d3
exp SKX NL3 exp SKX NL3
16
O
40
Ca
neutrons neutrons
Figure 2: Experimental and theoretical neutron single-particle energies for
16
O
and
40
Ca. The orbits are labeled by (n, , 2j), and the dashed line is the Fermi
energy.
s
i
n
g
l
e
-
p
a
r
t
i
c
l
e

e
n
e
r
g
y

(
M
e
V
)

-24
-22
-20
-18
-16
-14
-12
-10
-8
-6
-4
-2
0
2
4
6
0f7
1p3
1p1
0f5
0d5
1s1
0d3
0f7
1p3
1p1
0f5
0d5
1s1
0d3
0f7
1p3
1p1
0f5
0d3
1s1
0f7
0d5
0d3
1s1
0f7
1p3
1p1
0f5
0d5
0d3
1s1
0f7
1p3
1p1
0f5
exp SKX NL3 exp SKX NL3
protons neutrons
48
Ca
Figure 3: Experimental and theoretical proton and neutron single-particle en-
ergies for
48
Ca. The orbits are labeled by (n, , 2j), and the dashed line is the
Fermi energy.
12 THE HARTREE-FOCK APPROXIMATION 121
s
i
n
g
l
e
-
p
a
r
t
i
c
l
e

e
n
e
r
g
y

(
M
e
V
)

-24
-22
-20
-18
-16
-14
-12
-10
-8
-6
-4
-2
0
2
4
6
0g7
1d5
2s1
0h11
1d3
1f7
2p3
0h9
1f5
1p1
0g9
1d5
0g7
2s1
1d3
0h11
1f7
2p3
1f5
0h9
1p1
0g9
0g7
1d5
1d3
2s1
0h11
1f7
2p3
0h9
1f5
1p1
0g9
0g7
1d5
1d3
0h11
1p1
0g9
0g7
1d5
1d3
0h11
2s1
1p1
0g9
0g7
1d5
1d3
0h11
2s1
exp SKX NL3 exp SKX NL3
protons neutrons
132
Sn
Figure 4: Experimental and theoretical proton and neutron single-particle en-
ergies for
132
Sn. The orbits are labeled by (n, , 2j), and the dashed line is the
Fermi energy.
12.4.2 Single-particle energies
The experimental single-particle energies for
16
O,
40
Ca,
48
Ca,
132
Sn and
208
Pb are
compared with the SKX Hartree-Fock and NL3 Dirac Hartree calculations in Figs.
(12.2), (12.3), (12.4), and (12.5). Both mean-eld calculations are in qualitative
agreement with experiment. For light nuclei the NL3 results are in better, but for
heavy nuclei the SKX results are better. The dierence between SKX and NL3 is
mainly related to the eective mass (m

/m), which is about 1.0 for SKX and 0.6 for

NL3. The eective mass for the Skyrme interaction can be tuned by the parameters [2]
and those for SKX are determined primarily from the SPE of heavy nuclei where the
spacing around the Fermi surface requires an eective mass of about 1.0 [14]. For NL3
on the other hand, an eective mass of about 0.6 is intrinsic to the model. An eective
mass of 0.6 gives SPE in heavy nuclei which are spread out compared to experiment
as shown by NL3 in Figs. (12.4) and (12.5). Typical Brueckner G matrix interactions
also give an eective mass of about 0.6, and the reason for an empirical value of near
unity in heavy nuclei is attributed to conguration mixing due to coupling of the
single-particle states to surface vibrations [15], [16]. For a hamiltonian like SKX with
an eective mass of unity, these surface vibrations eectively included in terms of a
modied (renormalized) hamiltonian.
Experimental values of the SPE are usually used as input to shell-model calcula-
12 THE HARTREE-FOCK APPROXIMATION 122
s
i
n
g
l
e
-
p
a
r
t
i
c
l
e

e
n
e
r
g
y

(
M
e
V
)

-24
-22
-20
-18
-16
-14
-12
-10
-8
-6
-4
-2
0
2
4
6
1d5
0h11
1d3
2s1
0h9
1f7
0i13
2p3
1f5
2p1
1g9
0i11
0j15
1g7
1d5
0h11
1d3
2s1
0h9
1f7
0i13
1f5
2p3
2p1
1g9
0i11
0j15
1g7
1d5
0h11
1d3
2s1
0h9
1f7
0i13
1f5
2p3
2p1
0i11
1g9
1g7
0j15
0g7
1d5
0h11
1d3
2s1
0h9
1f7
0i13
1f5
0g7
1d5
0h11
1d3
2s1
0h9
1f7
0i13
0g7
1d5
0h11
1d3
2s1
0h9
1f7
0i13
exp SKX NL3 exp SKX NL3
protons neutrons
208
Pb
Figure 5: Experimental and theoretical proton and neutron single-particle en-
ergies for
132
Sn. The orbits are labeled by (n, , 2j), and the dashed line is the
Fermi energy.
tions. For nuclei near the drip lines where the experimental SPE are not known, one
needs a theoretical model for predicting or extrapolating the SPE. The HF parameter
model such as SKX and NL3 provide perhaps the most reliable way to extrapolate
the SPE from nuclei near stability to the most exotic nuclei near the drip lines. The
predictions for the
60
Ca and
78
Ni SPE are given in Figs. (12.6) and (12.7), respec-
tively. Other recent predictions for
60
Ca are given in [17]. It will remain for future
experiments to test these extrapolations.
As a guide to the ordering of the single-particle states as a function of N and Z,
I show in Fig. (12.8) the single-particle energies for nuclei with N = Z obtained from
with SKX . Fig. (12.9) shows the SKX single-particle energies for the calcium isotopes
as a function of neutron number. Beyond
60
Ca one observes that the 0g
9/2
orbital
becomes bound, and thus
70
Ca will be bound with SKX . The even-even calcium
nuclei between
60
Ca and
70
Ca will probably be bound due to pairing (which has been
neglected in this particular SKX HF calculation).
12.4.3 Rms charge radii and charge densities
In this section I will discuss the results for rms charge and charge densities obtained
with the Skyrme parameter sets SKX [14] and SKM* [18]
12 THE HARTREE-FOCK APPROXIMATION 123
s
i
n
g
l
e
-
p
a
r
t
i
c
l
e

e
n
e
r
g
y

(
M
e
V
)

-30
-28
-26
-24
-22
-20
-18
-16
-14
-12
-10
-8
-6
-4
-2
0
2
4
6
0d5
0s1
0d3
0f7
1p3
1p1
0f5
0g9
0d5
0s1
0d3
0f7
1p3
1p1
0f5
0g9
0d3
1s1
0f7
1p3
0f5
1p1
0g9
0d3
1s1
0f7
1p3
0f5
1p1
0g9
protons neutrons
60
Ca
SKX NL3 SKX NL3
Figure 6: Theoretical proton and neutron single-particle energies for
60
Ca. The
orbits are labeled by (n, , 2j), and the dashed line is the Fermi energy.
s
i
n
g
l
e
-
p
a
r
t
i
c
l
e

e
n
e
r
g
y

(
M
e
V
)

-30
-28
-26
-24
-22
-20
-18
-16
-14
-12
-10
-8
-6
-4
-2
0
2
4
6
1s1
0f7
1p3
1p1
0f5
0g9
1d5
1d3
2s1
1s1
0f7
0f5
1p3
1p1
0g9
1d5
2s1
1d3
1s1
0f7
1p3
0f5
1p1
0g9
1s1
0f7
0f5
1p3
1p1
0g9
protons neutrons
78
Ni
SKX NL3 SKX NL3
Figure 7: Theoretical proton and neutron single-particle energies for
78
Ni. The
orbits are labeled by (n, , 2j), and the dashed line is the Fermi energy.
12 THE HARTREE-FOCK APPROXIMATION 124
S
P
E

(
M
e
V
)

f
o
r

N
=
Z
N
-50
-45
-40
-35
-30
-25
-20
-15
-10
-5
0
5
0 10 20 30 40 50 60
0s1
0p3
0p1
0d5
0d3
1s1
0f7
0f5
1p3
1p1
0g9
0g7
Figure 8: Neutron single-particle energies for nuclei with N = Z as a function
of neutron number obtained with the SKX spherical Hartree-Fock calculation.
The vertical dashed lines show some of the magic number discussed in the text.
The large circles show the approximate Fermi energies for
16
O,
40
Ca,
56
Ni and
100
Sn. The orbits are labeled by (n, , 2j).
12 THE HARTREE-FOCK APPROXIMATION 125
S
P
E

(
M
e
V
)

f
o
r

Z
=
2
0
N
-50
-45
-40
-35
-30
-25
-20
-15
-10
-5
0
5
0 10 20 30 40 50 60
0s1
0p3
0p1
0d5
0d3
1s1
0f7
0f5
1p3
1p1
0g9
Figure 9: Neutron single-particle energies for the calcium isotopes obtained
with the SKX spherical Hartree-Fock calculation. The vertical dashed lines
show some of the magic number discussed in the text. The large circles show
the approximate Fermi energies for
28
Ca,
40
Ca,
48
Ca,
60
Ca and
70
Ca. The orbits
are labeled by (n, , 2j).
12 THE HARTREE-FOCK APPROXIMATION 126
E
x
p
e
r
i
m
e
n
t
Theory
2.4
2.8
3.2
3.6
4.0
4.4
4.8
5.2
2.4 2.8 3.2 3.6 4.0 4.4 4.8 5.2
Rms charge radii (fm)
SKM*
E
x
p
e
r
i
m
e
n
t
2.4
2.8
3.2
3.6
4.0
4.4
4.8
5.2
Rms charge radii (fm)
SKX
Figure 10: Comparison of experimental and theoretical rms charge radii for
the Skyrme interactions SKX (top panel) and SKM* (bottom panel).
The experimental and theoretical rms charge radii are compared in Fig. (12.10).
The experimental radii are from Refs. [19] and [20], with those with the smallest errors
selected in the case of more than one data set. The error in the experimental data is
typically smaller than the size of the data points. The Hartree-Fock results for the
two interactions SKM* (bottom panel) and SKX (top panel) are obtained with the
CM (fractional) occupations (those obtained with ESP occupations are essentially
the same). The excellent overall agreement between experiment and theory is not
surprising since these rms charge radii are used to constrain the values of the Skyrme
parameters. The deviations increase for light nuclei going up to 5% for
12
C. In general
one may expect the mean-eld approximation to be less valid for light nuclei.
In order to illustrate how the features of the charge density are built out of the
12 THE HARTREE-FOCK APPROXIMATION 127
d
e
n
s
i
t
y
r (fm)
0.0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0 2 4 6
1s
0s
0p
0d

0.0
0.01
0.02
0.03
0.04
1p
0f
Figure 11: The point-proton density (points) of
40
Ca in the SKX model with
a closed shell conguration. The total density is decomposed in terms of the
contributions from the individual lled orbitals. In the upper panel the densities
for protons in the orbitals just above the Fermi-surface of
40
Ca are shown, for
eight protons in the 0f
7/2
orbit and for four protons in the 1p
3/2
orbit.
specic shell model orbitals which are lled, we show in Fig. (12.11) the point-proton
density of
40
Ca (points) obtained with SKX with the assumption that the 0s, 0p, 1s
and 0d orbits are lled. The individual contributions of the lled orbits to the proton
density are shown. The 0s and 1s are the contributions from two protons in each of
these orbits respectively. The 0p indicates the sum of the four protons in the 0p
3/2
orbit and two protons in the 0p
1/2
orbit. The 0d indicates the sum of the six protons
in the 0d
5/2
orbit and four protons in the 0d
3/2
orbit. In the top panel the densities
associated with the (unlled) valence orbits above the Fermi-surface are shown; 0f
for eight protons in the 0f
7/2
orbit and 1p for four protons in the 1p
3/2
orbit.
For comparison between experimental and theoretial charge densities I consider
in Fig. (12.12) a set of data for nuclei that cover a wide mass range and which for
which there is the good electron scattering data:
28
Si from Ref. [21],
32
S from Ref.
[22] (circles) and Ref. [23] (squares),
40
Ca and
48
Ca from Ref. [24],
50
Ti and
52
Cr
from Ref. [25],
54
Fe from Ref. [26],
58
Ni from Refs. [27] (circles) and [26] (squares),
12 THE HARTREE-FOCK APPROXIMATION 128
c
h
a
r
g
e

d
e
n
s
i
t
y
r (fm)
0.0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.1
0.11
0.12
0.13
0.14
0.15
0.16
0.17
0.18
0.19
0 2 4 6 8 10 12 14
28
Si
32
S
40
Ca
48
Ca
50
Ti
52
Cr
54
Fe
58
Ni
88
Sr
90
Zr
92
Mo
204
Hg
208
Pb
Figure 12: Experimental charge-density distributions (lled circles) compared
with the SKX (dashed lines) and SKM* (solid lines) calculations. The charge
density units are e fm
3
. Beyond
28
Si the curves and data have been progres-
sively oset by 1 fm and 0.01 in the charge density. The data are discussed in
the text.
88
Sr from Ref. [28],
90
Zr from Ref. [29],
92
Mo from Ref. [30],
204
Hg from Ref. [31],
and
208
Pb from Refs. [32] (circles) and [33] (squares). The data are compared to the
SKX (dashed lines) and SKM* (solid lines) are calculations. Both SKX and SKM*
nicely reproduce the nucleus-dependent oscillations. The hamiltonian parameters are
obtained from ts to the rms charge radii, binding energies and excited state energies
of these nuclei. Thus the good agreement between experiment and theory for the
nucleus-dependent oscillations observed in Fig. (12.12) are not a result of a t to
these data but arise naturally from the underlying shell structure.
The main dierence between the SKX and SKM* results is that the interior den-
sity is about 5% higher with SKM* compared to SKX, with SKM* in best overall
agreement with experiment. Close inspection of the curves in Fig. (12.12) in the region
where the density falls o reveals a slightly larger surface diuseness for SKM* com-
12 THE HARTREE-FOCK APPROXIMATION 129
pared to SKX (the dashed line for SKX is slightly steeper than the line for SKM* in
the surface). The correlation between the increased interior height with the increased
diuseness (when the rms radius is the same) is consistent with what is expected
from the Fermi model shown in Fig. 3.2. The change between SKX and SKM* corre-
sponds to a a = 0.036 fm for
208
Pb. The change in the diuseness is connected to
the dierence in the density-dependent part of the hamiltonians, namely [(r)]
1/2
for
SKX and [(r)]
1/6
for SKM*. There is agreement between experiment and theory to
an accuracy of about 2% or better for r > 1 fm with SKM*. As discussed above, this
is about the level of the accuracy with which these densities can be experimentally
determined.
In the overview of Fig. (12.12) one can observe several interesting features asso-
ciated with how the quantum waves change with shell structure and mass. Between
28
Si and
32
S in the sd shell there is a large increase in the interior density due to
the lling of the 1s
1/2
orbital (see also [23]). Likewise between
204
Hg and
208
Pb there
is a large increase in the interior related to the lling of the 2s
1/2
orbital. Between
40
Ca and
48
Ca one observes a redistribution of the charge (proton) density due to
the interaction with the valence 0f
7/2
neutrons. The theoretical density distributions
for the sequence
48
Ca,
50
Ti,
52
Cr and
54
Fe show a smoothly varying trend due to
dominance of the proton 0f
7/2
subshell lling.
12.4.4 Displacement energies
The displacement energy is the binding energy dierence between mirror nuclei (those
with the same atomic number A but with the proton number Z and neutron number
N interchanged). For a given mass A and isospin T the displacement energy is:
D(A, T) = BE(A, T
<
z
) BE(A, T
>
z
), (12.33)
where T =[ T
<
z
[ =[ T
>
z
[, BE(A, T
<
z
) is the binding energy of the proton-rich nu-
cleus and BE(A, T
>
z
) is the binding energy of the neutron-rich nucleus. If the nuclear
force is charge symmetric, then this binding energy dierence can be related to the
well-understood Coulomb interaction between the protons. However, it was shown
by Nolen and Schier [34] that there is a systematic increase in the experimental
displacement energies compared to those calculated with a charge symmetric strong
interaction (the Nolen-Schier anomaly). In the usual HF calculation one has both di-
rect and exchange terms in the Coulomb-energy density functional. For the exchange
one uses the Fermi-gas approximation which is a good approximation to the exact
calculation [35]. The ground-state displacement energies obtained with the Coulomb
plus Coulomb-exchange HF approach (from the SKXce interaction of [14]) are shown
on the right-hand side of Fig. (12.13). One observes the systematic deviation between
experiment and theory associated with the Nolen-Schier anomaly. For the heaviest
12 THE HARTREE-FOCK APPROXIMATION 130
D
i
s
p
l
a
c
e
m
e
n
t

E
n
e
r
g
y

(
M
e
V
)

0
10
20
30
0 20 40 60
T=1/2
T=1
T=3/2
T=2 SKXce
Mass (A)
0 20 40 60
T=1/2
T=1
T=3/2
T=2 SKX

0 20 40 60
T=1/2
T=1
T=3/2
T=2 SKXcsb
Figure 13: Ground-state displacement energies as a function of A for SKXce
(with the normal Coulomb exchange term), SKX (without the exchange term)
and SKXcsb (with the exchange term and with a CSB interaction). The exper-
imental data are lled circles and the calculations are given as crosses.
nuclei the ratio shown in Fig. (12.14) goes to a constant value of about 1.06. It is
well known [34] that the displacement energies are sensitive to the rms charge radii;
the SKX interactions reproduce the experimental charge radii of
16
O,
40
Ca and
48
Ca
to better than one percent (see Fig. 10 of [14]). The displacement energies are also
sensitive to the rms radius of the valence orbits, and the SKX interactions give radii
for the d
5/2
orbit in
17
O and the 0f
7/2
orbit in
41
Ca which are within 2% of those
deduced from the magnetic electron scattering form factors [36].
In the t to closed-shell nuclei the displacement energy is represented by the
pair of nuclei
48
Ni-
48
Ca. The binding energy of
48
Ni is not measured but can be
extrapolated to within an uncertainly of a few hundred keV from the 0f
7/2
shell
displacement energy systematics [37], [38], [39], [40], [41]. The recent discovery of
48
Ni [42] is consistent with the mass obtained from the 0f
7/2
extrapolations.
In order to improve agreement with experiment it was found that the HF theory
could be improved in two ways. One of them consists of leaving out the Coulomb
exchange term, with the result shown in the middle panel of Figs. (12.13) and (12.14).
This may be interpreted as a correction from nuclear correlation (conguration mix-
ing) which happens to cancel the exchange term. This has been discussed in the
general HF framework by Bulgac and Shaginyan [43], [44], [45] in terms of a surface-
vibration contribution to the Coulomb correlation energy. I note that the relativistic
approach leaves out the Coulomb exchange by default, and that most Woods-Saxon
programs [46], [47] leave out the Coulomb exchange.
12 THE HARTREE-FOCK APPROXIMATION 131
R
a
t
i
o
0.8
0.9
1.0
1.1
1.2
SKXce SKX SKXcsb
D
i
f
f
e
r
e
n
c
e

(
M
e
V
)

-1
0
1
0 20 40 60
Mass (A)
0 20 40 60

0 20 40 60
Figure 14: Top: the ratio experiment/theory for the data shown in Fig. (12.13);
Bottom: the dierence experimenttheory. The symbols are lled circles for
T = 1/2, crosses for T = 1, squares for T = 3/2 and plus signs for T = 2.
Another way to improve agreement with experiment is to add a charge-symmetry
breaking (CSB) term to the Skyrme interaction which can be expressed as a change
to the proton-proton (pp) and neutron-neutron (nn) s-wave interactions:
V
pp
Skyrme
= t
0
(1 x
0
)(1 +x
a
)
and
V
nn
Skyrme
= t
0
(1 x
0
)(1 x
a
), (12.34)
where x
a
is a parameter to be determined.
The A = 48 closed shell nuclei require x
a
= 0.0140.002 for the interaction
called SKXcsb The results for the all displacement energies are shown on the right-
hand side of Figs. (12.13) and (12.14).
A correct description of the displacement energies within the mean eld approx-
imation is obviously important for understanding the position of the proton drip
lines. A = 99 is the heaviest T = 1/2 isobaric doublet for which the proton-rich
nucleus is expected to be bound. The calculated displacement energies for A = 99
are 13.54 MeV (SKXce), 14.03 MeV MeV (SKX) and 14.15 MeV (SKXcsb).
The introduction of the extra terms in the Skyrme hamiltonian which are needed for
the displacement energies, also has an inuence on the neutron drip line; for example
12 THE HARTREE-FOCK APPROXIMATION 132
the binding energy of
176
Sn is predicted to be 1158.0 MeV (SKXce), 1149.0 MeV
(SKX) and 1148.4 MeV (SKXcsb).
The SKXcsb interaction should ultimately be related to the CSB nucleon-nucleon
(NN) scattering data. Analysis of the NN scattering data together with a model for
the NN interaction gives [48], [49] a value of a
CSB
= a
pp
a
nn
= 1.5 0.5 fm for
the dierence in the pp and nn scattering lengths. Modern NN potentials such as
the AV18 [50] and CDbonn99 [51] are designed to reproduce this dierence. It is not
easy to interpret the CSB contribution to the displacement energies directly in terms
of a NN potential due the short-range nuclear correlations and their dependence on
the strong NN potential. Probably the most realistic way to do this is to consider the
CSB contribution to the displacement energies obtained with the variational Monte
Carlo approach for A = 7 [52] and the BHF approach for A = 15 and A = 17 [51].
For example, the CSB displacement energy for the A = 17, d
5/2
state is 92 keV with
AV18 [51] to be compared with 355 keV with SKXcsb. From these comparisons one
nds that the eect of the empirical CSB interaction obtained for SKXcsb is a factor
of 34 larger than expected from AV18 or CDbonn99. Thus one concludes that
either there is a signicant NNN or many-body CSB contribution whose origin is
unknown, or that a major part of the displacement energy anomaly is due to nuclear
correlations. Possible many-body CSB eects at the quark level have recently been
examined [53], [54].
The systematics associated with the Coulomb displacement energies can be used
to obtain theoretical binding-energies of proton-rich nuclei from the experimental
binding energies of neutron-rich nuclei. The displacement energies of Eq. (12.33) can
be modeled on smooth systematics [38], [39], [55]; shell-model conguration mixing
which contains the Coulomb and CSB interactions [37], [40], [41]; or on the mean-
eld models. One can combine the experimental binding energy for the neutron-
rich nucleus BE(A, T
>
z
)
exp
together with the calculated value for D(A, T) to give an
extrapolation for the proton-rich binding energy:
BE(A, T
<
z
) = D(A, T)
theory
+BE(A, T
>
z
)
exp
. (12.35)
For most of the nuclei out to the proton drip line the binding energy BE(A, T
>
z
)
exp
of the mirror neutron-rich nucleus is known to 100 keV or better. This method has
been used to predict the binding energies and the drip line for Z = 19 28 [37], [41],
[38] and Z = 28 38 [41], [56]. The latter calculations have been used [56] to study
the rapid-proton (rp) capture path in the astrophysical explosive hydrogen burning
process [57]. The rp-process in light nuclei depends upon theoretical calculations of
the displacement energies of ground and excited states and upon the spectroscopic
factors which enter into the (p,) reaction [58], [59].
REFERENCES 133
References
[1] T. R. H. Skyrme. Philos. Mag. 1, 1043 (1956); Nucl. Phys. 9, 615 (1959); 9, 635
(1959).
[2] D. Vautherin and D. M. Brink, Phys. Rev. C 5, 626 (1972).
[3] C. B. Dover and N. Van Giai, Nucl. Phys. A190, 373 (1972).
[4] W. Kohn and L. J. Sham, Phys. Rev. 140 A1133 (1965).
[5] R. M. Dreizler and E. K. U. Gross, Density Functional Theory: An Approach to
the Quantum Many-Body Problem, (Springer, Berlin, 1990).
[6] M. M. Sharma, G. Lalazissis, J. Konig and P. Ring, Phys. Rev. Lett. 74, 3744
(1995).
[7] B. Friedman and V. R. Pandharipande, Nucl. Phys. A361, 502 (1981).
[8] B. A. Brown, Phys. Rev. Lett. 85, 5296 (2000).
[9] S. Typel and B. A. Brown, Phys. Rev. C 64, 027302 (2001).
[10] B. A. Brown, W. A. Richter and R. Lindsay, Phys. Lett. B 483, 49 (2000).
[11] S. Goriely, F. Tondeur and J. M. Pearson, At. Data Nucl. Data Sheets 77, 311
(2001).
[12] J. D. Walecka, Ann. Phys. (N. Y.) 83, 491 (1974).
[13] G. A. Lalazissis, J. Konig and P. Ring, Phys. Rev. C 55, 540 (1997).
[14] B. A. Brown, Phys. Rev. C 58, 220 (1998).
[15] S. O. Backman, G. E. Brown and J. A. Niskanen, Phys. Rep. 124, 1 (1985).
[16] C. Mahaux, P. F. Bortignon, R. A. Broglia and C. G. Dasso, Phys. Rep. 120, 1
(1985).
[17] I. Hamamoto, H. Sagawa and X . Z. Zhang, Phys. Rev. C 64, 024313 (2001).
[18] J. Bartel, P. Quentin, M. Brack, C.Guet, and M. B. Hakansson, Nucl. Phys.
A386, 79 (1982).
[19] H. de Vries, C. W. de Jager and C. de Vries, At. Data and Nucl. Data Tables
36, 495 (1987).
REFERENCES 134
[20] G. Fricke, C. Bernhardt, K. Heilig, L. A. Schaller, L. Shellenberg, E. B. Shera
and C. W. de Jager, At. Data and Nucl. Data Sheets 60, 177 (1995).
[21] H. Miessen, Ph.D. Thesis, University of Mainz, 1982 (unpublished).
[22] D. Rychel, Ph.D. Thesis, University of Mainz, 1983 (unpublished)
[23] J. Wesseling, C. W. de Jager, L. Lapikas, H. de Vries, L. W. Fagg, M. N. Harakeh,
N. Kalantar-Nayestanaki, R. A. Lindgren, E. Moya De Guerra, and P. Sarriguren,
Phys. Rev. C55, 2773 (1997)
[24] H. J. Emrich, Ph.D. Thesis, University of Mainz, 1983 (unpublished)
[25] J. W. Lightbody, Jr., J. B. Bellicard, J. M. Cavedon, B. Frois, D. Goutte, m.
Huet, Ph. Leconte, A. Nakada, Phan Xuan Ho, S. K. Platchkov, S. Turck-Chieze,
C. W. de Jager, J. J. Lapikas, and P. K. A. de Witt Huberts. Phys. Rev. C27,
113 (1983)
[26] H. D. Wohlfahrt, Habilitationsschrift, University of Mainz, 1976 (unpublished)
[27] G. Beuscher, Ph.D. Thesis, University of Mainz, 1983 (unpublished)
[28] G. Stephan, Ph.D. Thesis, University of Mainz, 1976 (unpublished)
[29] H. Rothhaas, Ph.D. Thesis, University of Mainz, 1976 (unpublished)
[30] B. Dreher, Phys. Rev. Lett 35, 716 (1975)
[31] A. J. C. Burghardt, Ph.D. Thesis, University of Amsterdam, 1989 (unpublished)
[32] H. Euteneuer, J. Friedrich and N. Vogler, Nucl. Phys. A298, 452 (1978)
[33] J. L. Friar, J. Heisenberg and J. W. Negele, Proceedings of the June Workshop
in Intermediate Energy Electromagnetic Interactions, edited by A. M. Bernstein,
Massachusetts Institute of Technology (1977), p. 325
[34] J. A. Nolen and J.P. Schier, Ann. Rev. Nucl. Sci. 19, 471 (1969).
[35] C. Titin-Schnaider and P. Quentin, Phys. Lett. 49B, 397 (1974).
[36] T. Suzuki, H. Sagawa and A. Arima, Nucl. Phys. A536, 141 (1992)
[37] B. A. Brown, Phys. Rev. C 43, R1513 (1991); C 44, 924 (1991).
[38] B. J. Cole, Phys. Rev. C 54, 1240 (1996).
[39] Phys. Rev. C 58, 2831 (1998).
[40] W. E. Ormand, Phys. Rev. C 53, 214 (1996); Phys. Rev. C 55, 2407 (1997).
REFERENCES 135
[41] W. E. Ormand, Phys. Rev. C 55, 2407 (1997).
[42] B. Blank et al., Phys. Rev. Lett. 84, 1116 (2000).
[43] A. Bulgac and V. R. Shaginyan, Nucl. Phys. A601, 103 (1996).
[44] A. Bulgac and V. R. Shaginyan, Phys. Lett. B 469, 1 (1999).
[45] A. Bulgac and V. R. Shaginyan, Eur. Phys. J. A 5, 247 (1999).
[46] R. Sherr and G. Bertsch, Phys. Rev. C 32, 1809 (1985).
[47] B. A. Brown, A. Csoto and R. Sherr, Nucl. Phys. A597, 66 (1996).
[48] G. A. Miller, B. M. K. Nefkens and I. Slaus, Phys. Rep. 194, 1 (1990).
[49] G. Q. Li and R. Machleidt, Phys. Rev. C 58, 3153 (1998).
[50] R. B. Wiringa, V. G. J. Stoks and R. Schiavilla, Phys. Rev. C 49, 2950 (1995).
[51] C. Harzer, H. Muther and R. Machleidt, Phys. Lett. B 459, 1 (1999).
[52] B. S. Pudliner, V. R. Pandharipande, J. Carlson, S. C. Pieper, and R. B. Wiringa,
Phys. Rev. C 56, 1720 (1997).
[53] K. Tsushima, K. Saito and A. W. Thomas, Phys. Lett. B 465, 36 (1999).
[54] C. J. Horowitz and J. Piekarewicz, Phys. Rev. C 63, 01330(R), (2000).
[55] B. A. Brown and P. G. Hansen, Phys. Lett. B 391 (1996).
[56] B. A. Brown, R. Clement, H. Schatz, W. A. Richter, A. Volya and W. A. Richter,
Phys. Rev. C65, 045802 (2002).
[57] H. Schatz et al., Phys. Rep. 294, 167 (1998).
[58] H. T. Fortune, R. Sherr and B. A. Brown, Phys. Rev. C 61, 057303 (2000).
[59] H. Herndl, J. Goerres, M. Wiescher, B. A. Brown and L. Van Wormer, Phys.
Rev. C 52, 1078 (1995).
13 ANGULAR MOMENTUM AND TENSOR ALGEBRA 136
13 Angular Momentum and Tensor Algebra
13.1 Angular momentum coupling
The total angular momentum of a system is conserved. The total angular momentum
may be composed of the sum of orbital and intrinsic spins, or it may be composed of
the sum of the angular momenta

j for two or more nucleons. In this section

J will
denote a generalized angular momentum, it could be orbital, intrinsic-spin or some
combinations of both. Results for the coupling of up to four angular momenta and
the notation relating these to the 3j, 6j and 9j coecients will be summarized.
Explicit expressions for the 3j, 6j and 9j coecients are given in [1], [2] and [3].
Computer programs for these can be found in the books by Thompson [3] and [4] as
well as in the OXBASH shell-model computer package [5]. The 3j and 6j coecients
are tabulated in many references for small values of angular momentum and in [6]
for small and intermediate values of angular momentum. Some of the more useful
special relations and values for these coecients will be given below. Many others
are available in the literature [1] [2] [3] [7] [8] [9] [10].
13.1.1 Coupling of Two Angular Momenta
The coupling of two angular momenta will be considered rst. The total angular
momentum,

J =

J
1
+

J
2
, is conserved, and the quantum numbers J associated with
the vector

J take on the integer values from J
min
=[ J
1
J
2
[ to J
max
= J
1
+J
2
. This
condition on the J values will be referred to as the triangle condition, (J
1
J
2
J). The
sum of the J values in the triangle condition, J
t
= J
1
+J
2
+J in this case, is always
an integer, and the factor (1)
2Jt
= 1 can be multiplied in order to simplify a phase
factor.
The wave functions associated with the states [ J
1
J
2
JM > are linear combina-
tions of the states [ J
1
M
1
J
2
M
2
>
[ J
1
J
2
JM >=

M
1
M
2
[ J
1
M
1
J
2
M
2
>< J
1
M
1
J
2
M
2
[ J
1
J
2
JM >, (13.1)
where M
1
+M
2
= M. The Clebsch-Gordan coecients [1] are a specic normalization
and phase convention for the overlaps
< J
1
M
1
J
2
M
2
[ JM >< J
1
M
1
J
2
M
2
[ J
1
J
2
JM > .
In particular, they are all real and
< J
1
M
1
J
2
M
2
[ J
max
, M = J
max
>= 1.
13 ANGULAR MOMENTUM AND TENSOR ALGEBRA 137
Since they are real,
< J
1
M
1
J
2
M
2
[ JM >=< JM [ J
1
M
1
J
2
M
2
> . (13.2)
The Clebsch-Gordan coecients can be used, for example, to couple the two wave
functions
J
1
a
and
J
2
b
to the product wave function
c
(J
1
, J
2
)
J

c
(J
1
, J
2
)
J
M
=

M
1
M
2
< J
1
M
1
J
2
M
2
[ J
1
J
2
JM >
J
1
a,M
1

J
2
b,M
2
[
J
1
a

J
2
b
]
J
M
. (13.3)
The Clebsch-Gordan coecients obey the orthonormality conditions

J,M
< J
1
M
1
J
2
M
2
[ JM >< J
1
M

1
J
2
M

2
[ JM >=
M
1
M

M
2
M

2
, (13.4)
and

M
1
M
2
< J
1
M
1
J
2
M
2
[ JM >< J
1
M
1
J
2
M
2
[ J

>=
JJ

MM
. (13.5)
The symmetry properties of the Clebsch-Gordan coecients are
< J
1
M
1
J
2
M
2
[ JM >= (1)
J
1
+J
2
J
< J
2
M
2
J
1
M
1
[ JM >, (13.6)
< J
1
M
1
J
2
M
2
[ JM >= (1)
J
1
+J
2
J
< J
1
, M
1
J
2
, M
2
[ J, M >, (13.7)
and
< J
1
M
1
J
2
M
2
[ JM >= (1)
J
2
+M
2

2J + 1
2J
1
+ 1
< J
2
, M
2
JM [ J
1
M
1
> . (13.8)
The 3j coecient
_
J
1
J
2
J
M
1
M
2
M
_
is related to the Clebsch-Gordan coecient
by
< J
1
M
1
J
2
M
2
[ JM > (1)
J
1
J
2
+M

2J + 1
_
J
1
J
2
J
M
1
M
2
M
_
, (13.9)
where the sum of the M values in the bottom row must be zero: M
1
+M
2
M = 0.
The symmetry properties of the Clebsch-Gordan coecients are given compactly in
terms of those for the 3j coecients and can be summarized by
_
J
1
J
2
J
3
M
1
M
2
M
3
_
= (1)
p
_
J
a
J
b
J
c
M
a
M
b
M
c
_
, (13.10)
where p = J
a
+ J
b
+ J
c
when the 3j on the right is obtained by the interchange
of any two neighboring columns (a permutation of the columns) of the 3j on the
13 ANGULAR MOMENTUM AND TENSOR ALGEBRA 138
left. Since p always an integer, two such interchanges gives (1)
2p
= 1. In addition,
p = J
a
+ J
b
+ J
c
if the top row is unchanged but the bottom row is changed in sign.
(M
1
= M
a
, M
2
= M
b
and M
3
= M
c
). The orthogonality relations for the 3j
coecients which are equivalent to Eqs. (13.4) and (13.5) take the form

J,M
(2J + 1)
_
J
1
J
2
J
M
1
M
2
M
__
J
1
J
2
J
M

1
M

2
M
_
=
M
1
M

M
2
M

2
, (13.11)
and

M
1
,M
2
(2J + 1)
_
J
1
J
2
J
M
1
M
2
M
__
J
1
J
2
J

M
1
M
2
M

_
=
JJ

MM
. (13.12)
Useful expressions for some special cases are
_
J 0 J
M 0 M

_
=
(1)
JM

MM

2J + 1
, (13.13)
_
J 1 J
M 0 M

_
=
(1)
JM
M
MM

_
J(2J + 1)(J + 1)
, (13.14)
and
_
J 2 J
M 0 M

_
=
(1)
JM
[3M
2
J(J + 1)]
MM

_
(2J 1)J(2J + 1)(J + 1)(2J + 3)
. (13.15)
13.1.2 Coupling of Three Angular Momenta
In this section we use the notation [ (J
1
J
2
J
12
)J
3
JM > for the three particle wave
functions formed from the Clebsch-Gordan coupling of J
1
and J
2
to make J
12
and
then the Clebsch-Gordan coupling of J
12
to J
3
to make the total the total angular
momutum J and projection M. The wave functions may require other quantum
numbers for their complete specication, but only the explicit representation for J is
required in this section. The three angular momenta can be coupled in various ways.
For example,
[ (J
1
J
2
J
12
)J
3
JM >=

M
12
M
3
< J
12
M
12
J
3
M
3
[ JM >[ (J
1
J
2
J
12
M
12
)J
3
M
3
>
=

M
1
M
2
M
3
< J
12
M
12
J
3
M
3
[ JM >< J
1
M
1
J
2
M
2
[ J
12
M
12
>[ J
1
M
1
J
2
M
2
J
3
M
3
>,
(13.16)
13 ANGULAR MOMENTUM AND TENSOR ALGEBRA 139
and
[ J
1
(J
2
J
3
J
23
)J

>=

M
23
M

1
< J
1
M

1
J
23
M
23
[ J

>[ J
1
M

1
(J
2
J
3
J
23
M
23
) >
=

1
M

2
M

3
< J
1
M

1
J
23
M
23
[ J

>< J
2
M

2
J
3
M

3
[ J
23
M
23
>[ J
1
M

1
J
2
M

2
J
3
M

3
> .
(13.17)
The overlaps between these two ways of coupling are dened in terms of the 6j
coecients
< (J
1
J
2
J
12
)J
3
JM [ J
1
(J
2
J
3
J
23
)J

>=
=
JJ

MM

M
1
M
2
M
3
< J
12
M
12
J
3
M
3
[ JM >< J
1
M
1
J
2
M
2
[ J
12
M
12
>
< J
1
M
1
J
23
M
23
[ JM >< J
2
M
2
J
3
M
3
[ J
23
M
23
>
(1)
J
1
+J
2
+J
3
+J
_
(2J
12
+ 1)(2J
23
+ 1)
_
J
1
J
2
J
12
J
3
J J
23
_
, (13.18)
and thus
[ J
1
(J
2
J
3
J
23
)JM >=

J
12
(1)
J
1
+J
2
+J
3
+J
_
(2J
12
+ 1)(2J
23
+ 1)

_
J
1
J
2
J
12
J
3
J J
23
_
[ (J
1
J
2
J
12
)J
3
JM > . (13.19)
The allowed values of J are restricted by the triangle conditions (J
1
J
2
J
12
), (J
3
JJ
12
),
(J
3
J
2
J
23
) and (J
1
JJ
23
), as represented by the circles in
_
o o o
_
,
_
o
o o
_
,
_
o
o o
_
and
_
o
o o
_
.
The symmetry properties of the 6j coecient can be summarized by
_
J
1
J
2
J
3
J

1
J

2
J

3
_
=
_
J
a
J
b
J
c
J

a
J

b
J

c
_
, (13.20)
when the columns (a, b, c) are any permutation of the columns (1, 2, 3). In addition,
_
J
1
J
2
J
3
J

1
J

2
J

3
_
=
_
J

1
J

2
J
3
J
1
J
2
J

3
_
. (13.21)
An orthogonality condition for the 6j coecients is

J
3
(2J
3
+ 1)(2J
6
+ 1)
_
J
1
J
2
J
3
J
4
J
5
J
6
__
J
1
J
2
J
3
J
4
J
5
J

6
_
=
J
6
J

6
. (13.22)
13 ANGULAR MOMENTUM AND TENSOR ALGEBRA 140
Another useful relation for the 6j coecients is

k
(1)
k+k
1
+k
2
(2k + 1)
_
k
1
J

1
k
J

2
J
2
J
__
k
1
k
2
k
J

1
J
1
J

2
__
k
1
k
2
k
J

2
J
2
J

1
_
= (1)
J
1
+J
2
+J

1
+J

2
+J

1
+J

2
+J
_
J
1
J
2
J
J

1
J

2
k
1
__
J

1
J

2
J
J

1
J

2
k
2
_
. (13.23)
When one of the arguments is zero, the 6j coecient reduces to
_
J
1
J
2
J
3
J
4
J
5
0
_
=
(1)
J
1
+J
2
+J
3

J
5
J
1

J
4
J
2
_
(2J
1
+ 1)(2J
2
+ 1)
. (13.24)
13.1.3 Coupling of Four Angular Momenta
The 9j coecient is dened by the overlap between two ways of coupling four angular
momenta
< (J
1
J
3
J
13
)(J
2
J
4
J
24
)J [ (J
1
J
2
J
12
)(J
3
J
4
J
34
)J >

_
(2J
13
+ 1)(2J
24
+ 1)(2J
12
+ 1)(2J
34
+ 1)
_

_
J
1
J
2
J
12
J
3
J
4
J
34
J
13
J
24
J
_

_
J
1
J
2
J
12
J
3
J
4
J
34
J
13
J
24
J
_

_ . (13.25)
The symbol with nine arguments is the 9j coecient, and the [ ] symbol with nine
arguments will be referred to as the normalized 9j coecient. The allowed values of J
are restricted by the triangle condition associated with any row or column. The value
of the 9j coecient is unchanged for any even permutation of the rows or columns
and changes only by a phase factor (1)
J
1
+J
2
+J
12
+J
3
+J
4
+J
34
+J
13
+J
24
+J
for any odd
permutation of the rows or columns. An orthogonality condition for the normalized
9j coecients is

J
13
J
24
_

_
J
1
J
2
J
12
J
3
J
4
J
34
J
13
J
24
J
_

_
_

_
J
1
J
2
J

12
J
3
J
4
J

34
J
13
j
24
J
_

_ =
J
12
J

12

J
34
J

34
. (13.26)
Another sum rule which will be used later is (see page 367 in Ref. [2])

J
(2J + 1)
_

_
J
1
J
2
J
12
J
3
J
4
J
34
J
13
J
24
J
_

_
_

_
J

1
J

2
J

12
J

3
J

4
J

34
J
13
J
24
J
_

_
13 ANGULAR MOMENTUM AND TENSOR ALGEBRA 141
=

J
(2J + 1)
_
J
1
J
4
J
J
24
J
12
J
2
__
J
1
J
4
J
J
34
J
13
J
3
_

_
J
24
J
12
J
J

1
J

4
J

2
__
J
34
J
13
J
J

1
J

4
J

3
_
. (13.27)
(This last sum rule is related to the 12j coecient of the Second Kind [2]). When
one of the arguments is zero, the 9j coecient reduces to
_

_
J
1
J
2
J
12
J
3
J
4
J
34
J
13
J
24
0
_

_
=
(1)
J
2
+J
3
+J
12
+J
13

J
12
J
34

J
13
J
24
_
(2J
12
+ 1)(2J
13
+ 1)
_
J
1
J
3
J
13
J
4
J
2
J
12
_
. (13.28)
The 9j coecient can be expressed as a sum over 6j coecients

(1)
2J

(2J

+ 1)
_
J
1
J
3
J
13
J
24
J J

__
J
2
J
4
J
24
J
3
J

J
34
__
J
12
J
34
J
J

J
1
J
2
_
=
_

_
J
1
J
2
J
12
J
3
J
4
J
34
J
13
J
24
J
_

_
. (13.29)
13 ANGULAR MOMENTUM AND TENSOR ALGEBRA 142
13.2 Tensors and reduced matrix elements
A tensor T

of rank is a set of 2+1 operators, = , +1, . . . , , which obey

the following commutation relations with the angular momentum operators J

and
J
z
[J

, T

] = [J
x
iJ
y
, T

] =
_
( + 1) ( 1) T

1
, (13.30)
and
[J
z
, T

] = T

. (13.31)
The spherical harmonics, Y

, are the most obvious example of a tensor of rank .

Based on the components of Y
1

in cartesian coordinates, the components of a vector

r can be arranged into a tensor of rank = 1 as: r
1
0
= z, and r
1

= (x iy)/

2 .
The use of tensor operators reduces the M-state dependence of matrix elements to a
simple dependence on the Clebsch-Gordan or 3j coecient via the Wigner-Eckhart
theorem
< JM [ T

[ J

>= (1)
JM
_
J J

M M

_
< J[[T

[[J

>
= (1)
2
< J

[ JM >
< J[[T

[[J

>
_
(2J + 1)
, (13.32)
where < J[[T

[[J

> is referred to as the reduced matrix element. The reduced matrix

element convention followed in this book is common [1] [11] [9] [7] [8] [12] [13] but not
universal. By multiplying Eq. (13.32) on both sides by the 3j coecient, summing
over M, and M

and using the 3j sum rule of Eq. (13.12), one obtains the inverse
relation
< J[[T

[[J

>= (2 + 1)

MM

(1)
JM
_
J J

M M

_
< JM [ T

[ J

>
=

MM

(1)
JM
_
J J

M M

_
< JM [ T

[ J

> . (13.33)
If T

is a tensor of rank , then its Hermitian conjugate, (T

)
+
, is not a tensor
of rank . However, the quantity

(1)
p+
(T

)
+
(13.34)
is a tensor of rank . This can be proven by taking the Hermitian conjugate of both
sides of Eqs. (13.30) and (13.31) and then making the substitution
(T

)
+
= (1)
p

T

.
13 ANGULAR MOMENTUM AND TENSOR ALGEBRA 143
The arbitrary phase factor p is chosen for convenience. In order to ensure that the
phase factors are real, one can choose p = 0 for integer values of , and p = for
half-integer values of . When the phase factor is real, (1)
p
= (1)
p
.
The relationship between < J[[T

[[J

> and < J

[[

[[J > can be obtained by

considering
< JM [ T

[ J

>=< J

[ (T

)
+
[ JM >
+
=< J

[ (T

)
+
[ JM >

= (1)
p
< J

[

T

[ JM >

= (1)
p+J

_
J

J
M

M
_
< J

[[

[[J >

= (1)
p+J

M
_
J J

M M

_
< J

[[

[[J >

,
where Eq. (13.10) has been used in the last line together with M = M

+. Thus by
comparison with Eq. (13.32)
< J[[T

[[J

>= (1)
p+J

J
< J

[[

[[J >

, (13.35)
and
< J[[

[[J

>= (1)
JJ

+p
< J

[[T

[[J >

, (13.36)
where * indicates complex conjugation. The spherical harmonics have the property
that

Y

= Y

, and hence
< J[[Y

[[J

>= (1)
JJ

< J

[[Y

[[J >

. (13.37)
If wave functions are chosen to be real (as will be the case in this book), this expression
simplies further to
< J[[Y

[[J

>= (1)
JJ

< J

[[Y

[[J > . (13.38)

13.2.1 Special reduced matrix elements
Reduced matrix element can be calculated from Eq. (13.33) or more simply by eval-
uating the left-hand side of Eq. (13.32) for a particular choice of M, and M

for
which the 3j coecient does not vanish. Reduced matrix elements for many types of
operators can be found in the literature [7] [9] [1] [2]. Some commonly used reduced
matrix elements are
< J[[1[[J

>=
JJ

2J + 1 , (13.39)
13 ANGULAR MOMENTUM AND TENSOR ALGEBRA 144
< J[[J
=1
[[J

>=
JJ

_
J(J + 1)(2J + 1) , (13.40)
and
< [[Y

[[

>= (1)

(2 + 1)(2 + 1)(2

+ 1)
4
_

0 0 0
_
(13.41)
It is clear from Eqs. (13.13) and (13.32) (and from rotational invariance) that the
matrix element of a scalar is independent of M and can be written as
< JM [ T
=0
[ J

>=
JJ

MM
< J [ T
=0
[ J >=
JJ

MM

< J[[T
=0
[[J >

2J + 1
.
(13.42)
13.2.2 Products of tensor operators
The components of two tensors T

1
and U

2
can be combined to form a third tensor
S

by
S

= [T

1
U

2
]

2
<
1

2
[ > T

1
U

2
(13.43)
It is straightforward to show that the components of S

satisfy Eqs. (13.30) and

(13.31). Useful results for the reduced matrix elements involving products of tensor
operators will be discussed in this section. The dot product (e.g.

s) is related to
this tensor cross product by
T

1
U

1
= (1)

1
_
(2
1
+ 1) [T

1
U

1
]
0
. (13.44)
When T and U operate in the same space (e.g. [Y

1
( r) Y

2
( r)]

), the reduced
matrix element can be evaluated by summing over a complete set of intermediate
states J
c
[9]
< J
a
[[[T

1
U

2
]

[[J
b
>= (1)
Ja++J
b
_
(2 + 1)

Jc
_
J
b
J
a

1

2
J
c
_
< J
a
[[T

1
[[J
c
>< J
c
[[U

2
[[J
b
> . (13.45)
When T and U operate in dierent spaces (e.g.

s), one obtains [9]
< J
a
J
b
J[[[T

1
U

2
]

[[J
c
J
d
J

>=
_
(2J + 1)(2 + 1)(2J

+ 1)

_
J
a
J
b
J
J
c
J
d
J

1

2

_

_
< J
a
[[T

1
[[J
c
>< J
b
[[U

2
[[J
d
>, (13.46)
13 ANGULAR MOMENTUM AND TENSOR ALGEBRA 145
where T

1
operates in the space of [ J
a
> and [ J
c
>, and U

2
operates in the space
of [ J
b
> and [ J
d
>.
Special cases of Eq. (13.46) can be obtained from the reduction of the 9j coecient
given in Eq. (13.28). For = 0 Eq. (13.46) reduces to
< J
a
J
b
J[[[T

1
U

1
]
0
[[J
c
J
d
J

>=
JJ
(1)
J
b
+Jc+J+
1

2J + 1
2
1
+ 1

_
J
a
J
b
J
J
d
J
c

1
_
< J
a
[[T

1
[[J
c
>< J
b
[[U

1
[[J
d
> . (13.47)
For
1
= 0 and T

1
= 1, Eq. (13.46) reduces to
< J
a
J
b
J[[U

2
[[J
c
J
d
J

>= (1)
Jc+J
d
+J+
2
_
(2J + 1)(2J

+ 1)

JaJc
_
J
b
J
d

2
J

J J
a
_
< J
b
[[U

2
[[J
d
>, (13.48)
and for
2
= 0 and U

2
= 1, Eq. (13.46) reduces to
< J
a
J
b
J[[T

1
[[J
c
J
d
J

>= (1)
Ja+J
b
+J

+
1
_
(2J + 1)(2J

+ 1)

J
b
J
d
_
J
a
J
c

1
J

J J
b
_
< J
a
[[T

1
[[J
c
> . (13.49)
Eq. (13.49) can be used, for example, to obtain the single-particle j-coupled
matrix element of Y

< sj[[Y

[[

sj

>= (1)
+1/2+j

+
_
(2j + 1)(2j

+ 1)

j 1/2
_
< [[Y

[[

>, (13.50)
which can be combined with Eq. (13.41) and simplied [1] to obtain
< sj[[Y

[[

sj

>=
1
2
[1 + (1)
++

](1)
j+1/2

(2j + 1)(2 + 1)(2j

+ 1)
4
_
j j

1/2 0 1/2
_
. (13.51)
13 ANGULAR MOMENTUM AND TENSOR ALGEBRA 146
13.2.3 Other convensions for reduced matrix elements
Other conventions for reduced matrix elements analogous to Eq. (13.32) often used
in the literature are that of Lawson [14]
< JM [ T

[ J

>=< J

[ JM >< J[[T

[[J

>
L
, (13.52)
and that of Brink and Satchler [10], Towner [15] and Thompson [3]
< JM [ T

[ J

>= (1)
2
< J

[ JM >< J[[T

[[J

>
B
. (13.53)
An advantage of using Eq. (13.32) instead of one of the above alternatives is that Eq.
(13.38) only involves a phase factor, and a disadvantage is the extra

2J + 1 factor
which appears in Eqs. (13.39) and (13.42). When expressions involving reduced ma-
trix elements are taken from the literature, it is obviously important to be consistent
and to know which convention is being used.
REFERENCES 147
References
[1] A. R. Edmonds, Angular Momentum in Quantum Mechanics, (Princeton Univer-
sity Press, 1957).
[2] D. A. Varshalovich, A. N. Moskalev and V. K. Khersonskii, Quantum Theory of
Angular Momentum, (World Scientic, 1988).
[3] W. J. Thompson, Angular Momentum, (John Wiley and Sons, Inc., 1994)
[4] K. L. G. Heyde, The Nuclear Shell Model, (Springer-Verlag, 1994)
[5] B. A. Brown, et al., the computer code OXBASH.
[6] M. Rotenberg, R. Bivins, M. Metropolis and K. R. Wooten Jr., The 3-j and 6-j
Symbols, (Technology Press MIT, Cambridge, 1959).
[7] A. de Shalit and I. Talmi, Nuclear Shell Theory, (Academic Press, 1963).
[8] I. Talmi, Simple Models of Complex Nuclei, (Harwood Academic Publishers,
1993).
[9] P. J. Brussaard and P. W. M. Glaudemans, Shell Model Applications in Nuclear
Spectroscopy, (North Holland, 1977).
[10] D. M. Brink and G. R. Satchler, Angular Momentum, (Clarendon Press, Oxford,
1968).
[11] A. Messiah, Quantum Mechanics, Vol. 1, (North Holland, Amsterdam, 1975).
[12] A. Bohr and B. R. Mottelson, Nuclear Structure, Vol. I, (W. A. Benjamin, 1969),
Nuclear Structure Vol. II, (W. A. Benjamin, 1975).
[13] A. de Shalit and H. Feshbach, Theoretical Nuclear Physics Vol. I: Nuclear Struc-
ture, (John Wiley and Sons, 1974).
[14] R. D. Lawson, Theory of the Nuclear Shell Model, (Clarendon Press, 1980).
[15] I. S. Towner, A Shell Model Description of Light Nuclei, (Clarendon Press, 1977).
14 SINGLE-PARTICLE ELECTROMAGNETIC MOMENTS 148
14 Single-particle electromagnetic moments
14.1 General results and notation
Electromagnetic moments have the general form:
/
em
=< , J, M = J [

k
T

=0
(k) [ , J, M = J >, (14.1)
where T

is a one-body tensor operator of rank associated with the interaction of the

nucleus with the multipole components of the electromagnetic eld. This is a diagonal
matrix element, and by denition we take M to have its maximum value M = J.
The matrix elements with other M values are related to /
em
by the Wigner-Eckart
theorem:
< , J, M [

k
T

=0
(k) [ , J, M >=
< J, M, , = 0 [ J, M >
< J, J, , = 0 [ J, J >
/
em
. (14.2)
For a given J value, the allowed values of the are determined by the triangle
condition, (J, , J). In particular,
max
= 2J, and for J = 0 only = 0 is allowed.
There are two type of electromagnetic moments: the electric moments associated
with the static distribution of charge and the magnetic moments associated with the
magnetic currents. The parity conservation requires =even for the static moments
and =odd for the magnetic moments. The names relate to the greek for the 2

poles.
The lowest few are:
2
0
monopole
2
1
dipole
2
2
quadrupole
2
3
octupole
2
4
hexadecapole
2
5
triakontadupole
The electromagnetic moment for =0 is

k
e
k
which simply counts the number
of particles with charge e
k
. The operators for magnetic dipole (=1) and electric
quadrupole (=2) moments will be discussed in the following subsections. A compi-
lation of experimental magnetic and quadrupole moments is given on the web [1].
14 SINGLE-PARTICLE ELECTROMAGNETIC MOMENTS 149
14.2 General results for closed-shell and single-particle con-
gurations
We rst consider the angular momenta allowed for these simple one-dimensional con-
gurations, and then investigate the consequences for the electromagnetic moments.
For the closed-shell conguration [ C > there are several (n
r
, , j) orbits which are
completely lled. For each of these orbits there is an m degeneracy of (2j +1) states
(for protons or neutrons) and the total M
j
value is:
M
j
=

< j, m [

j
z
[ j, m >=
m=j

m=j
< j, m [

j
z
[ j, m >=
m=j

m=j
m = 0,
(The wavefunction [ j, m > also depends upon n
r
and , but as in this example,
sometimes they need not be written explicitly.) Thus each lled j orbit has J = 0.
The sum over all M
j
gives M = 0, and thus we have J

= 0
+
for the closed shell.
The parity is positive since the number of particles in each j orbit is even. Only the
= 0 moment contributes for the closed shell.
For the single-particle conguration [ Ci >, the extra nucleon will go into one of
the empty states (n
r
, , j, m) above the fermi surface. There are (2j + 1) allowed M
values from j to j. Thus the total angular momentum is J = j, and the parity is
(1)

. For the moments we need the Mstate with M = m = j.

For the single-hole conguration, [ Ci
1
>, the nucleon will be removed from one
of the lled states (n
r
, , j, m

) below the fermi surface. The M value of the state is:

M =

m=m

m = m

.
There are (2j + 1) values for M from j to j. Thus the total angular momentum is
J = j, and the parity parity is (1)

. If we want to have a many-body state with

M = J = j, then the nucleon must be removed from the single-particle state with
m

= j.
Next we consider the moments for >0. For one-particle outside of a closed-shell
conguration, [ >=[ Ci >, the moment is:
/(Ci)
em
=

a
< [ T

=0
[ > + < i [ T

=0
[ i >, (14.3)
where the sum over runs over the states lled in the closed shell. Since the closed
shell has J

= 0
+
the rst term in Eq. (14.3) is zero unless =0. For >0 we have
for the [ Ci > conguration:
/(Ci)
em
=< i [ T

=0
[ i >=< j, m = j [ T

=0
[ j, m = j > . (14.4)
14 SINGLE-PARTICLE ELECTROMAGNETIC MOMENTS 150
For the one-hole conguration:
/(Ci
1
)
em
=

a
< [ T

=0
[ > < i [ T

=0
[ i >, (14.5)
and for >0:
/(Ci
1
)
em
= < i [ T

=0
[ i >= < j, m = j [ T

=0
[ j, m = j > . (14.6)
Note that the particle is removed from the state m = j so that M = J = j. A
special result of the Wigner-Eckart theorem of Eq. (14.2) is:
< j, m = j [ T

=0
[ j, m = j >=
=
< j, j, , = 0 [ j, j >
< j, j, , = 0 [ j, j >
< j, m = j [ T

=0
[ j, m = j >
= (1)

< j, m = j [ T

=0
[ j, m = j > . (14.7)
Thus, for the hole states:
/(Ci
1
)
em
= (1)
+1
< j, m = j [ T

=0
[ j, m = j > . (14.8)
The one-body operator T

will have terms which operate separately on the space

and spin parts of the wave function. Thus, the general evaluation of the single-
particle moments starts with the decomposition of the state j into its space and spin
components:
[ , j, m >=

,ms
< , m

, s, m
s
[ j, m >[ , m

, m
s
,
where the notation [will be used for the state where m

and m
s
are good quantum
numbers (n
r
and s are implicit). The special cases we need for the moments are:
[ , j = +
1
2
, m = j >=[ , ,
1
2
, (14.9)
and
[ , j =
1
2
, m = j >=

1
2 + 1
[ , 1,
1
2

2
2 + 1
[ , ,
1
2
. (14.10)
The matrix elements of T

have the form:

< n
r
, , j = +
1
2
, m = j [ T

[ n
r
, , j = +
1
2
, m = j >
=, ,
1
2
[ T

[ , ,
1
2
, (14.11)
14 SINGLE-PARTICLE ELECTROMAGNETIC MOMENTS 151
and:
< n
r
, , j =
1
2
, m = j [ T

[ n
r
, , j =
1
2
, m = j >
=
_
1
2 + 1
_
, 1,
1
2
[ T

[ , 1,
1
2

+
_
2
2 + 1
_
, ,
1
2
[ T

[ , ,
1
2
. (14.12)
For these expressions I have used the fact that the electromagnetic operator T

is
diagonal in m
s
(it does not connect the states with m
s
=
1
2
and m
s
=
1
2
).
14.3 Magnetic moments
The magnetic moment operator is dened to be
T
=1
=0
=
z
[
z
g

tz
+s
z
g
s
tz
]
N
, (14.13)
where g

tz
and g
s
tz
are the orbital and spin g-factors for the proton (t
z
=
1
2
) and
neutron (t
z
=
1
2
). The free-nucleon values for the g-factors are g

p
= 1, g

n
= 0,
g
s
p
= 5.586 and g
s
n
= 3.826. The values of the magnetic moments are conventionally
taken to be in units of the nuclear magneton,

N
=
e h
2m
p
c
= 0.105 e fm (14.14)
where m
p
is the mass of the proton.
From Eq. (14.8), the single-particle and single-hole magnetic moments are the
same. From Eq. (14.11) for m = j and j = +
1
2
we have
(j = +
1
2
)

N
= g

tz
+
1
2
g
s
tz
, (14.15)
and from Eq. (14.12) for m = j and j =
1
2
we have
(j =
1
2
)

N
=
_
1
(2 + 1)
( 1) +
2
(2 + 1)

_
g

tz
+
_
1
2 + 1
_
1
2
_
+
2
2 + 1
_

1
2
_
_
g
s
tz
=
(2 1)( + 1)
(2 + 1)
g

tz

(2 1)
(4 + 2)
g
s
tz
. (14.16)
The traditional and most compact way to write these expressions is:

N
= j
_
g

tz

g
s
tz
g

tz
2 + 1
_
, (14.17)
14 SINGLE-PARTICLE ELECTROMAGNETIC MOMENTS 152
m
a
g
n
e
t
i
c

m
o
m
e
n
t
l value for j = l + 1/2
-4
-2
0
2
4
6
-1 0 1 2 3 4 5 6
odd proton
j = l + 1/2
1/2+ 3/2- 5/2+ 7/2- 9/2+ 11/2- 13/2+
l value for j = l - 1/2
-1 0 1 2 3 4 5 6 7
odd proton
j = l - 1/2
1/2- 3/2+ 5/2- 7/2+ 9/2- 11/2+
m
a
g
n
e
t
i
c

m
o
m
e
n
t
-4
-2
0
2
4
6
8
odd neutron
j = l + 1/2
1/2+ 3/2- 5/2+ 7/2- 9/2+ 11/2- 13/2+
odd neutron
j = l - 1/2
1/2- 3/2+ 5/2- 7/2+ 9/2- 11/2+
Figure 1: Experimental magnetic moments compared to the single-particle
(Schmidt) values. For a given j

value the data are plotted vs increasing neu-

tron or proton number withplus signs for positive parity and minus signs for
negative parity.
there the sign goes with j
1
2
. The g-factor is dened as /(
N
J) which for the
single-particle case gives:
g =
_
g

tz

g
s
tz
g

tz
2 + 1
_
. (14.18)
These single-particle magnetic moments are called the Schmidt values. The result for
neutrons in the state j = +
1
2
with g

n
= 0 reduces to
(j = +
1
2
, neutron) =
g
s
n

N
2
=
n
= 1.913
N
where
n
is the magnetic moment of the neutron.
The magnetic moments data from the compilation of Stone [1] are shown in Fig.
14 SINGLE-PARTICLE ELECTROMAGNETIC MOMENTS 153
(14.1) in comparison with the single-particle values. The data are shown for the odd-
neutron and odd-proton nuclei. For a given j

value one can associate the unique

and the data can be divided into two groups with j = +1/2 and j = 1/2. There
is a pronounced dierence between the odd-neutron and odd-proton plots that is
related to the inuence of the orbital contribution for protons. With few exceptions
all of the data for a given kind of nucleon lies in between the limits given by the
j = 1/2 single-particle values. The reason for this can be understood as a result
of conguration mixing byond the single-particle model.
There is a short-cut for deriving these results by using the projection theorem:
< j, m [ v
z
[ j, m >=
< j [ (

j
.
v) [ j >
j(j + 1)
< j, m [ j
z
[ j, m >, (14.19)
where v
z
is any vector operating in the j-space, and the matrix element of

j
.
v has been
written without the m to indicate that the scalar matrix element does not depend
upon m. This can be derived using the Wigner-Eckart theorem.
For the single-particle magnetic moment with m = j:
< j, m = j [
z
[ j, m = j >=
< j [ (

j
.
) [ j >
(j + 1)
. (14.20)
We can save some algebra by writing:

z

N
=
z
g

tz
+s
z
g
s
tz
= (
z
+s
z
) g

tz
+s
z
(g
s
tz
g

tz
) = j
z
g

tz
+s
z
(g
s
tz
g

tz
)
The matrix elements for the magnetic moment is then:
< j, m = j [
z
[ j, m = j >

N
=
< j [ j
2
g

tz
+ (

j
.
s)(g
s
tz
g

tz
) [ j >
(j + 1)
= jg

tz
+
< j [ (

j
.
s) [ j > (g
s
tz
g

tz
)
(j + 1)
The scalar matrix element of

j
.
s can be derived using the identity

j s =

to obtain

j
.
s =
j
2
+s
2

2
2
whose expectation value is:
< j [

j
.
s [ j >=
j(j + 1) +s(s + 1) ( + 1)
2
,
and after some simplication one obtains the same result as in Eq. (14.17).
14 SINGLE-PARTICLE ELECTROMAGNETIC MOMENTS 154
14.4 Electric quadrupole moments
The electric quadrupole operator is dened to be
T
=2
=0
=

Q = (3z
2
r
2
)e
tz
e =

16
5
r
2
Y
2
0
( r) e
tz
e, (14.21)
were Y

are the spherical harmonics. The e

tz
are the charges for the proton and
neutron in units of e. For the free-nucleon charge we would take e
p
= 1 and e
n
= 0,
for the proton and neutron, respectively. Although the initial operator for quadrupole
moments only acts upon the protons, we will keep the general expression in terms of
e
tz
because later we will introduce eective charges for the proton and neutron. By
convention the electric quadrupole moment is taken to be in units of e: Quadrupole
moments are usually quoted in units of e fm
2
or eb where b is the barn, b = 100
fm
2
(sometimes the e is implicit).
From Eq. (14.11) the single-particle quadrupole moment in the state j = +
1
2
is:
Q
e
=

16
5
< Y

[ Y
2
0
[ Y

> < r
2
> e
tz
. (14.22)
where
< r
2
>=
_
f
2
(r) r
4
dr =
_
R
2
nr,,j
r
2
dr.
The angular integral is given by Eqs. (13.32) and (13.41):
< Y

[ Y
2
0
[ Y

>= (1)

_
2
0
_
(2 + 1)

5
4
_
2
0 0 0
_
=

5
4
_

2 + 3
_
. (14.23)
Thus, the single-particle quadrupole moment in the state j = +
1
2
simplies to:
Q(i)
e
=
_
2
2 + 3
_
< r
2
> e
tz
. (14.24)
From (14.8), the single-hole quadrupole moment in the state j = +
1
2
is:
Q(i
1
)
e
=
_
2
2 + 3
_
< r
2
> e
tz
. (14.25)
The radial integrals can be evaluated with the chosen radial wavefunctions such as
harmonic-oscillator or Woods-Saxon.
14 SINGLE-PARTICLE ELECTROMAGNETIC MOMENTS 155
For the general case which includes both j = +
1
2
and j =
1
2
one obtains for
particle states:
Q(i)
e
=
_
2j 1
2j + 2
_
< r
2
> e
tz
. (14.26)
and for hole states:
Q(i
1
)
e
=
_
2j 1
2j + 2
_
< r
2
> e
tz
. (14.27)
A geometrical understanding of the sign of the quadrupole moment can be ob-
tained from the form of the quadrupole operator in cartesian co-ordinates (3z
2
r
2
) =
(2z
2
x
2
y
2
). Thus if the density is spherical Q = 0. If the density extends more
along the z axis it has a prolate shape with Q > 0, and if the density is concentrated
in the x y plan it has a oblate shape with Q < 0. The density distribution for
the state m

= is concentrated towards the x y plane and has Q < 0. For the

single-hole state the density associated with this m

= state is missing and the

remaining density has a prolate shape with Q > 0.
The experimental data for the quadrupole moments of the ground states of odd-
even nuclei are shown in Fig. (14.2). In order to qualitatively divide out the mass
dependence implied by the < r
2
> matrix elements in Eqs. (14.26) and (14.27),
the experimental moments can be divided by R
2
where R = 1.2A
1/3
. These scaled
quadrupole moments are shown in Fig. (14.3). The scaled data are shown for a region
of the nuclear chart in Fig. (14.4).
There are several places where the sign change between particle (oblate) and hole
(prolate) states can be observed, for example, for odd-protons in Fig. (14.5) just above
and below Z = 50. But overall most nuclei have positive Q moments characteristic
of a prolate shape.
In the single-particle model we would expect the Q moments for the neutrons to
be zero or at least small compared to protons. But one observes that the Q moments
for odd-proton and odd-nuetron nuclei are about the same. The single-particle values
for the scaled Q moments are on the order of unity. In contrast the data for many
regions of nuclei show Q which are up to an order of magnitude larger than this.
Both of these deviations from the single-particle model are signatures of conguration
mixing. Near the magic numbers this can be understood in terms of the interaction
between the valence nucleon and the core nucleons producing a core-polatization
that can be modeled in terms of an eective change for protons and neutrons. Away
from the closed shell the interaction between valence nucleons results in a collective
(coherent) motion between many nucleons that is qualitatively understood in the
deformed model for nuclei.
14 SINGLE-PARTICLE ELECTROMAGNETIC MOMENTS 156
Q

m
o
m
e
n
t

(
e

f
m
^
2
)
Z or N
-100
0
100
200
300
0 20 40 60 80 100 120 140 160
odd-protons
Q

m
o
m
e
n
t

(
e

f
m
^
2
)
-100
0
100
200
300
400
odd-neutrons
Figure 2: Experimental quadrupole moments for the ground states of odd-even
nuclei. The dashed lines show the magic numbers 8, 20, 28, 40, 50, 82 and 126.
14 SINGLE-PARTICLE ELECTROMAGNETIC MOMENTS 157
s
c
a
l
e
d

Q

m
o
m
e
n
t
Z or N
-3
-2
-1
0
1
2
3
4
5
6
7
8
0 20 40 60 80 100 120 140 160
odd-protons
s
c
a
l
e
d

Q

m
o
m
e
n
t
-3
-2
-1
0
1
2
3
4
5
6
7
8
9
odd-neutrons
Figure 3: Scaled experimental quadrupole moments for the ground states of
odd-even nuclei. The dashed lines show the magic numbers 8, 20, 28, 40, 50,
82 and 126.
REFERENCES 158

P
r
o
t
o
n

N
u
m
b
e
r
Neutron Number
40
60
80
50 70 90 110 130
-2
0
2
4
6
8
Figure 4: Scaled experimental quadrupole moments for the ground states of
odd-even nuclei for a region of nuclei. The dashed lines show the magic numbers
50 and 82 for protons and 82 and 126 for neutrons.
References
[1] http://www.nndc.bnl.gov/publications/preprints/nuclear-moments.pdf
15 THE CREATION OPERATOR METHOD 159
15 The Creation Operator Method
15.1 Introduction
A convenient method for taking into account the antisymmetry of the many-body
wavefunctions is to express them in second-quantized form in terms of creation and
destruction operators. The antisymmetry is enforced by the commutation relations
of the operators. The next few sections summarize the properties of the creation
operator and their use in constructing wavefunctions and operators. Some examples
are given in the last section.
15.2 Creation Operators and Wavefunctions
The operator a
+

creates the one-particle state [ >

a
+

[>=[ >, (15.1)

where [> represents the vacuum state. The label represents, for example, the set
of quantum numbers (n
r
jm) used in the spherical shell-model basis. The n-particle
state is formed from the product of these creation operators. For example, for n = 3
a
+

a
+

a
+

[>=[ > . (15.2)

We use the convention that the order of the labels on the right-hand side is the inverse
of that on the left hand side. The Hermitian conjugates of Eqs. (15.1) and (15.2) are
<[ (a
+

)
+
=< [ <[ a

, (15.3)
and
<[ (a
+

a
+

a
+

)
+
=<[ a

=< [ . (15.4)
The a
+
and a operators obey the commutation relations
a
+

, a
+

= a
+

a
+

+a
+

a
+

= 0, (15.5)
a

, a

= a

+a

= 0, (15.6)
and
a
+

, a

= a
+

+a

a
+

. (15.7)
Use of the creation operators a
+
and annhilliation operators a together with their
commutation relationships are completely equivalent to the results obtained from
15 THE CREATION OPERATOR METHOD 160
Slater determinants, as will be shown in the next few paragraphs. In particular, we
must show either that the commutation relations can be obtained directly from the
structure of the antisymmetric wavefunctions, or that the commutation relations give
the same result when applied to any wavefunction.
Antisymmetry of the Slater determinant means that
[
1
. . .
n
>= [
1
. . .
n
>, (15.8)
and
[
1
. . .
n
>= (1)
n
[
1
. . .
n
>, (15.9)
Non-vanishing of the wavefunctions in Eqs. (15.8) and (15.9) implies that all of the
labels are dierent. Eq. (15.8) implies that
a
+

a
+

[
1
. . .
n
> +a
+

a
+

[
1
. . .
n
>= 0. (15.10)
Since the wavefunction itself is assumed not to vanish, one must have
a
+

a
+

+a
+

a
+

= 0, (15.11)
which is the rst the rst commutation relation, Eq. (15.5). The Hermitian conjugate
of this gives the second commutation relation, Eq. (15.6). The case when = in Eqs.
(15.5) and (15.6) implies in particular that
a
+

a
+

= a

= 0. (15.12)
Before deriving Eq. (15.7), some intermediate results are useful, rst:
a
+

[
1
. . .
n
>= 0, (15.13)
a

[
1
. . .
n
>=[
1
. . .
n
>, (15.14)
a
+

[
1
. . .
n
>=[
1
. . .
n
>, (15.15)
and
a

[
1
. . .
n
>= 0, (15.16)
Eq. (15.13) follows from Eq. (15.12), and Eq. (15.15) follows from the basic denition
of Eq. (15.2). The results for operating with a to the right in Eqs. (15.14) and (15.16)
can be inferred from the normalization condition
< [ >=<[ a

[ >=

, (15.17)
which can be satised by the two relations
a

[ >= 0 if ,= (15.18)
15 THE CREATION OPERATOR METHOD 161
and
a

[ >=[> if = . (15.19)
In particular,
a

[>= 0, (15.20)
and
a

[ >=[> . (15.21)
Next, from Eqs. (15.13)-(15.16) it is straightforward to show that
a

a
+

[
1
. . .
n
>= 0, (15.22)
a
+

[
1
. . .
n
>=[
1
. . .
n
>, (15.23)
a

a
+

[
1
. . .
n
>=[
1
. . .
n
>, (15.24)
and
a
+

[
1
. . .
n
>= 0, (15.25)
Note that a

a
+

or a
+

cannot always be replaced by unity.

The commutation relation of Eq. (15.7) can now be proven by showing that it
holds for its application on all possible wavefunctions. The results for the case = :
(a
+

+a

a
+

) [
1
. . .
n
>=[
1
. . .
n
>, (15.26)
and
(a
+

+a

a
+

) [
1
. . .
n
>=[
1
. . .
n
>, (15.27)
are obtained immediately from Eqs. (15.22) (15.25). For ,= the results
(a
+

+a

a
+

) [
1
. . .
n
>= 0, (15.28)
and
(a
+

+a

a
+

) [
1
. . .
n
>= 0, (15.29)
follow immediately from Eqs. (15.13) and (15.16). The nal case is
(a
+

+a

a
+

) [
1
. . .
n
>= (a
+

a
+

a
+

a
+

) [
1
. . .
n
>
= (a
+

a
+

) [
1
. . .
n
>= 0, (15.30)
where (15.24) has been used to replace the a

a
+

by unity. Eqs. (15.26) (15.30) are

equivalent to Eq. (15.7).
15 THE CREATION OPERATOR METHOD 162
15.3 Operators
One- (O) and two-body (T) operators can be expressed in terms of these creation
and destruction operators:

O =

< [ O [ > a
+

, (15.31)
and

T =
1
4

< [ T [ > a
+

a
+

. (15.32)
Note the order of and at the end of this equation for T are dierent from their
order in [ >. Angular momentum coupled versions of these operators will be given
in a following chapter.
A basic one-body operators is the scalar number operator
N =
n

i=1
1 = n,
which in second-quantized form becomes

N =

< [ 1 [ > a
+

a
+

. (15.33)
When operating on a state to the right,

N will give unity if the state is occupied
[Eq. (15.23)] and zero if it is unoccupied [Eq. (15.25)], and the sum will thus give the
total number of particles, n.
The creation operator method is well suited for computer codes, since the basis
states can be represented by binary bit patterns, with (1) denoting an occupied state
and (0) an unoccupied state. The operations amount to binary additions and keeping
track of phase factors. Many computer codes have been developed which make use of
these techniques. These computations are based on some nite set of single-particle
quantum numbers (
1
, . . .,
n
). It is convenient to choose some arbitrary but xed
ordering for these. For example, if one considers only the state with j = 3/2, there
are four m-states which could put in the order (3/2, 1/2, 1/2, 3/2). The complete
set of many-body states for two particles in this basis is then, (0011), (0101), (1001),
(0110), (1010) and (1100).
15.4 Examples of a
+
and a Matrix Elements
This section will provide a few explicit examples for evaluating matrix elements of
operators in the a
+
-a representation. The rst examples are considered from the point
of view of algebraically interchanging the a
+
and a operators until one has a [>= 0.
15 THE CREATION OPERATOR METHOD 163
a) A normalization:
< i [ k >=<[ a
i
a
+
k
[>=<[
ik
a
+
k
a
i
[>=
ik
. (15.34)
b) The elementary one-body matrix element:
< i [

O [ k >=

< [ O [ >< i [ a
+

[ k >
=

< [ O [ ><[ a
i
a
+

a
+
k
[>
=

< [ O [ ><[ a
i
a
+

[
k
a
+
k
a

] [>
=

< [ O [ ><[ a
i
a
+

k
[>
=

< [ O [ ><[
i

k
[>=< i [ O [ k > .
This is a tedious way to get an identity. However, the power of second quantization is
apparent when the technique is applied to a more general case with help of a computer
code. Next we consider some examples which are closer to how second quantization
is used in computer programs.
c) The elementary one-body matrix element a repeat of (b) above:
< i [

O [ k >=

< [ O [ k >< i [ a
+

a
k
[ k >
=

< [ O [ k >< i [ a
+

[ >=< i [ O [ k >< i [ i >=< i [ O [ k > . (15.35)

d) A more general diagonal one-body matrix element:
< ikl[

O [ ikl >=

< [ O [ >< ikl[a

+

[ ikl >
=

_
< [ O [ i >< ikl[a
+

[ kl >
< [ O [ k >< ikl[a
+

[ il > + < [ O[l >< ikl[a

+

[ ik >
_
15 THE CREATION OPERATOR METHOD 164
=< i [ O [ i >< ikl[ikl > + < k [ O [ k >< ikl[ikl >
+ < l [ O [ l >< ikl[ikl >
=< i [ O [ i > + < k [ O [ k > + < l [ O [ l >, (15.36)
which is result expected for a diagonal matrix element. The notation in the
above wavefunctions is to indicate that the single-particle state which normally would
occupy this position is empty.
f) An o-diagonal one-body matrix element
< ikl[

O [ ikm >=

< [ O [ >< ikl[a

+

[ ikm >
=

a
_
< [ O [ i >< ikl[a
+

[ km >
< [ O [ k >< ikl[a
+

[ im > + < [ O [ m >< ikl[a

+

[ ik >
_
=< i [ O [ i >< ikl[ikm > + < k [ O [ k >< ikl[ikm >
+ < l[O [ m >< ikl[ikl >=< l[O [ m > . (15.37)
16 MANY-BODY WAVEFUNCTIONS 165
16 Many-Body Wavefunctions
16.1 Wavefunctions in the m-scheme
For a spherical potential the single-particle wavefunctions are labeled by their radial,
orbital angular momentum, and total angular momentum quantum numbers, n
r
,
and j, respectively. This set of quantum numbers will be denoted by k (n
r
, , j).
Each j value has (2m+1) m-states, and the associated single-particle wavefunctions
will be labeled by (km). It is useful to think of k as unique numerical sequence
of numbers associated with the complete set of single-particle states. A particular
choice for this labeling which is often used [1] is k = 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, . . ., for
the sequence 0s
1/2
, 0p
3/2
, 0p
1/2
, 0d
5/2
, 0d
3/2
, 1s
1/2
, 0f
7/2
, 0f
5/2
, 1p
3/2
, 1p
1/2
, . . . . The k
value for a given nj can be computed from:
k =
1
2
[(2n +)(2n + + 3) 2j + 3]. (16.1)
The formalism developed thus far is basically all that is needed for an M-scheme
calculation [2]. In the M-scheme one starts with a set of basis states for a given
M value
M =

, (16.2)
where the sum is over the m values for the occupied states. In general there are
an innite number of basis states , but a for a given situation one truncates the
number of states based upon those that are lowest in unperturbed energy. Since the
Hamiltonian is diagonal in M one need only consider the subset of basis states with
a single value of M in the construction of the many-particle wavefunctions . The
many-body matrix elements of the relevant one- and two-body operators can then be
calculated with the techniques of second quantization. Many computer codes have
been written to do this.
A basis state with a given value of M does not in general have a denite (good)
value of the total angular momentum J. However, since the Hamiltonian is spherically
symmetric, the Hamiltonian is also diagonal in J. Thus, the eigenvalues of H, which
are linear combinations of the basis, will automatically have good a J value with
J M, as long as the basis contains the complete set of states that are connected
by the

J
2
operator. The J value can be determined by calculating the expectation
value < [

J
2
[ >.
We can calculate the total number of states for a given J, the J-dimension D(J),
from the M-scheme dimensions d(M). This is based upon the fact that for each J
16 MANY-BODY WAVEFUNCTIONS 166
Table 1: M values for the [(j = 5/2)
2
] conguration for identical nucleons.
The xs under columns headed by 2m indicate that the state is occupied, and
the total M value is given on the right-hand side.
2m 5 3 1 -1 -3 -5
M
x x 4
x x 3
x x 2
x x 1
x x 0
x x 2
x x 1
x x 0
x x -1
x x 0
x x -1
x x -2
x x -2
x x -3
x x -4
state there must be (2J + 1) M-states. Thus we nd:
D(J) = d(M = J) d(M = J + 1). (16.3)
The meaning of Eq. (16.3) is that the number of extra M = J states compared to
to the number of M = J + 1 states must be the number of states with angular
momentum J. Since d(M) = d(M) one only has to consider M 0. Eq. (16.3) will
be illustrated with with some examples. The results for these examples only depend
upon the (j, m) quantum numbers, and thus the (n
r
, ) values are not given explicitly.
Table (16.1) gives all possible M-states for the [(j = 5/2)
2
] conguration for
identical nucleons. Due to the Pauli principle we cannot have M = 5, and thus J = 5
is not allowed. The M-scheme dimensions for this case are given in Table (16.2).
From the number of states for each M, one can make up Table (16.2). M
max
= 4
means that the highest J value is J = 4 and this will account for nine of the M states,
M= 4, 3, 2, 1, 0, -1, -2, -3 and -4. J = 3 is not allowed since there is only one M = 3
state which must go with J = 4. The extra M = 2 state means that there is a state
with J = 2 which accounts for ve more M states M= 2, 1, 0, -1, and -2. J = 1 is
not allowed since all of the M = 1 states are now used, and the extra M = 0 states
means that there is one state with J = 0. Thus, the allowed J values are 0, 2 and 4.
A second example is given for the [(j = 5/2)]
3
conguration given in Table (16.3).
16 MANY-BODY WAVEFUNCTIONS 167
Table 2: Table of M-scheme dimensions for [(j = 5/2)
2
].
M d(M)
4 1
3 1
2 2
1 2
0 3
-1 2
-2 2
-3 1
-4 1
Table 3: M0 values (M 0) for the [(j = 5/2)
3
] conguration for identi-
cal nucleons. The xs under columns headed by 2m indicate that the state is
occupied, and the total M value is given on the right-hand side.
2m 5 3 1 -1 -3 -5
2M
x x x 9
x x x 7
x x x 5
x x x 3
x x x 5
x x x 3
x x x 1
x x x 1
x x x 3
x x x 1
From the multiplicity of the M values one can deduce that only J = 3/2, 5/2 and 9/2
are allowed.
In general, the maximum J-value allowed for a [k
n
] conguration is given by the
sum of the n largest possible m values,
J
max
=
n

i=1
m
max
i
. (16.4)
In the [(j = 5/2)
3
] example, J
max
= 5/2 + 3/2 + 1/2 = 9/2. Lawson [3] discusses
other rules which can be deduced from these counting procedures. In particular, the
k
n
state with J = J
max
1 is not allowed by the Pauli principle.
A third example is given for the [(j = 5/2)
2
(j = 1/2)] conguration in Table
16 MANY-BODY WAVEFUNCTIONS 168
Table 4: M values (M 0) for the [(j = 5/2)
2
(j = 1/2)
1
] conguration for
identical nucleons. The xs under columns headed by 2j and 2m indicate that
the state is occupied, and the total M value is given on the right-hand side.
2j 5 5 5 5 5 5 1 1
2m 5 3 1 -1 -3 -5 -1 1
2M
x x x 9
x x x 7
x x x 7
x x x 5
x x x 5
x x x 3
x x x 3
x x x 1
x x x 1
x x x 5
x x x 3
x x x 3
x x x 1
x x x 1
x x x 1
(16.4). One nds that J = 9/2, 7/2, 5/2, 3/2 and 1/2 are allowed. The restrictions
on the J values due to the Pauli principle are only related to those associated with
putting several identical nucleons into one j-orbit. Thus these allowed J values can
also be deduced by the all possible values allowed by the triangle condition of coupling
j = 1/2 to the J = 0, 2 and 4 states allowed for two particles in the j =5/2 state.
A forth example is given for the [(j = 5/2)
1
(j = 1/2)
2
] conguration in Table
(16.5). One nds only J = 5/2, as expected from the coupling of j = 5/2 to the
closed-shell [(j = 1/2)
2
] conguration.
If there is only one M value allowed, the wavefunctions of the J state is given by
that of the single M state. If there is more that one M value allowed, then one must
diagonalize the hamiltonian in a space which has the dimension d(M). The states of
good J will be linear combinations of the M states. For a given J value of interest,
one usually chooses M = J, since the number of M states is minimized in this case.
However, one could also choose any M [ J [ value.
Alternatively, the linear combination of M-states with good J can be calculated
using angular-momentum projection methods. This is the method used by the code
OXBASH to construct a matrix with dimension D(J) corresponding to the states of
16 MANY-BODY WAVEFUNCTIONS 169
Table 5: M values (M 0) for the [(j = 5/2)
1
(j = 1/2)
2
] conguration for
identical nucleons. The xs under columns headed by 2j and 2m indicate that
the state is occupied, and the total M value is given on the right-hand side.
2j 5 5 5 5 5 5 1 1
2m 5 3 1 -1 -3 -5 -1 1
2M
x x x 5
x x x 3
x x x 1
good J in terms of the M-scheme basis.
A partition is dened as a specic distribution of the particles into the allowed
(active) set of k states. The examples discussed above include the three partitions
allowed for putting three particles into the j = 5/2 and j = 1/2 states, namely
[(j = 5/2)
3
], [(j = 5/2)
2
(j = 1)] and [(j = 5/2)(j = 1/2)
2
]. These are all of the
partitions allowed for three particles in these two orbits. The rst step in a shell
model calculation is to specify the number of particles, the number of active orbits,
and then to make a list of the complete set of partitions. Then for each partition
we calculate D
p
(J) and the total J-dimension is obtained by summing D
p
over all
partitions. The calculation may proceed with the full set of partitions (a full-space
calculation) or it may be restricted to some subset of the partitions (a truncated
calculation).
Compilations of D(J) for j
n
congurations for some values of j are available in
the literature [4], and computer codes are available for the general case [1], [5], [6].
16 MANY-BODY WAVEFUNCTIONS 170
16.2 Wavefunctions in the J-scheme
In order to reduce the size of the calculation it is useful to form a basis which has a
denite J value. It is also useful to express the many-particle matrix elements in a way
in which the trivial M-state dependence is taken into account. This section discusses
the structure of the angular momentum-coupled wave functions. The structure of the
operators and matrix elements, is considered in the next section.
The two-particle wavefunction in which particles in the states k
1
and k
2
with
angular momentum

j
1
and

j
2
, respectively, are coupled to a total angular momentum

J is given in terms of the product wavefunctions by

[ k
1
k
2
JM >
p
=

m
1
m
2
< j
1
m
1
j
2
m
2
[ JM >[ k
1
m
1
k
2
m
2
>
p
, (16.5)
where M = m
1
+ m
2
, and the subscript p indicates the product wavefunctions. As
usual, the allowed values of J are from J
min
=[ j
1
j
2
[ to J
max
= j
1
+ j
2
in integer
steps. The result for the Slater determinant is not as simple, but takes the form
[ k
1
k
2
JM >= N
12

m
1
m
2
< j
1
m
1
j
2
m
2
[ JM >[ k
1
m
1
k
2
m
2
>, (16.6)
where N
12
can be evaluated by taking the norm of both sides
< k
1
k
2
JM [ k
1
k
2
JM >= 1
= N
2
12

m
1
m
2
m
3
m
4
< j
1
m
1
j
2
m
2
[ JM >< j
1
m
3
j
2
m
4
[ JM >

1
2
_
< k
1
m
1
k
2
m
2
[ k
1
m
3
k
2
m
4
>
p
< k
1
m
1
k
2
m
2
[ k
2
m
4
k
1
m
3
>
p
< k
2
m
2
k
1
m
1
[ k
1
m
3
k
2
m
4
>
p
+ < k
2
m
2
k
1
m
1
[ k
2
m
4
k
1
m
3
>
p
_
= N
2
12

m
1
m
2
m
3
m
4
< j
1
m
1
j
2
m
2
[ JM >< j
1
m
3
j
2
m
4
[ JM >

m
1
m
3

m
2
m
4

m
1
m
4

m
2
m
3

k
1
k
2

= N
2
12
_

m
1
m
2
[< j
1
m
1
j
2
m
2
[ JM >[
2

k
1
k
2

m
1
m
2
< j
1
m
1
j
1
m
2
[ JM >< j
1
m
2
j
1
m
1
[ JM >
_
= N
2
12
1
k
1
k
2
(1)
J2j
1
= N
2
12
1 +
k
1
k
2
(1)
J
, (16.7)
where the symmetry property of the Clebsch-Gordan coecient has been used. Thus,
[ k
1
k
2
JM > vanishes if J is odd and k
1
= k
2
. When the two-particle wavefunction
does not vanish
N
12
=
1
_
(1 +
k
1
k
2
)
. (16.8)
16 MANY-BODY WAVEFUNCTIONS 171
The inverse of Eq. (16.6) is
[ k
1
m
1
k
2
m
2
>=
1
N
12

JM
< JM [ j
1
m
1
j
2
m
2
>[ k
1
k
2
JM > . (16.9)
Also note that
[ k
2
k
1
JM >= (1)
j
1
+j
2
J+1
[ k
1
k
2
JM >, (16.10)
where a phase factor of (1) arises from the antisymmetry of the two-particle wave-
function.
In terms of the a
+
operators introduced in Sec. 16.7, the angular momentum
coupled two-particle wavefunction is
[ k
1
k
2
JM >= N
12

m
1
m
2
< j
1
m
1
j
2
m
2
[ JM > a
+
k
2
m
2
a
+
k
1
m
1
[>
= N
12

m
1
m
2
< j
1
m
1
j
2
m
2
[ JM > a
+
k
1
m
1
a
+
k
2
m
2
[>
= N
12
[a
+
k
1
a
+
k
2
]
J
M
[> . (16.11)
In general, when more than two particles are coupled to good angular momentum,
one nds that N = 1 as long as each particle is in a dierent k state. When more than
two particles go into a single k state, the normalization factor becomes increasingly
complicated. However, as discussed in the previous section, the number of allowed
states for a given J value are easily deduced from the number of M values.
The wave functions associated with these allowed J values can be obtained with
the projection operators discussed in Sec. 16.12. The n-particle wavefunction will be
denoted by
[ k
n
JM > Z
+
(k
n
JM) [>, (16.12)
where is the additional quantum number which must be introduced when there is
more than one state for a given value of J, and where Z
+
are the linear combinations
of a
+
operators which create the antisymmetric n-particle state. (In the literature,
[7] one sometimes nds the notation Z in used place of Z
+
.) In some cases may be
a label associated with the seniority quantum number [8] or some group theoretical
classication. It will be assumed that these states are constructed to be orthonormal.
Computer programs are available for calculating the number of allowed J values in a
k
n
conguration. [9], [1]
When the n particles are distributed over two k-states, the wavefunctions take
the form
[ nJM >=[ (k
n
1
1

1
J
1
)(k
n
2
2

2
J
2
)JM >
16 MANY-BODY WAVEFUNCTIONS 172
=

M
1
M
2
< J
1
M
1
J
2
M
2
[ JM > Z
+
(k
n
2
2

2
J
2
M
2
)Z
+
(k
n
1
1

1
J
1
M
1
) [>
= (1)
n
1
n
2
[Z
+
(k
n
1
1

1
J
1
) Z
+
(k
n
2
2

2
J
2
)]
J
M
[>, (16.13)
where n = n
1
+ n
2
, and where the operators Z
+
are dened in Eq. (16.12) above.
The phase factor of (1)
n
1
n
2
arises from commuting n
1
particles in state k
1
with n
2
particles in state k
2
. Hermitian conjugation of this equation gives
< nJM [=< (k
n
1
1

1
J
1
)(k
n
2
2

2
J
2
)JM [
=

M
1
M
2
< J
1
M
1
J
2
M
2
[ JM ><[ Z(k
n
1
1

1
J
1
M
1
)Z(k
n
2
2

2
J
2
M
2
)
=<[ [Z(k
n
1
1

1
J
1
) Z(k
n
2
2

2
J
2
)]
J
M
. (16.14)
The values and multiplicities of J in Eq. (16.13) are those allowed by the vector
coupling of all possible

J
1
and

J
2
.
When the n particles are distributed over many k states, the wavefunctions take
one of the possible forms allowed by the successive vector coupling of the angular
momenta. As an example, for three states
[ nJM >=[ [(k
n
1
1

1
J
1
)(k
n
2
2

2
J
2
)J
12
][k
n
3
3

3
J
3
]JM >, (16.15)
where n = n
1
+ n
2
+ n
3
. For a given total number of particles n, the distribution
n
1
, n
2
. . . among the k states will be referred to as the partition.
16 MANY-BODY WAVEFUNCTIONS 173
16.3 Angular momentum projection
The angular-momentum projection operator can be used to construct explicit linear
combinations of the M-scheme Slater determinants which have good total angular
momentum. These linear combinations can thus be used to construct a Hamiltonian
matrix which is block diagonal in J. Before discussing the angular-momentum pro-
jection operator, the results for the simpler types of angular-momentum operators
will be reviewed.
The single-particle states [ k, m > are eigenfunctions of the j
2
and j
z
operators
j
2
[ km >= j(j + 1) [ km > (16.16)
and
j
z
[ km >= m [ km > . (16.17)
The j
2
operator can be written in terms of the raising and lowering operators j

j
2
= j

j
+
+j
2
z
+j
z
= j
2
= j
+
j

+j
2
z
j
z
, (16.18)
where
j

[ km >=
_
j(j + 1) m(m1) [ k, m1 > . (16.19)
The total angular momentum is a one-body operator given by a sum over the angular
momenta of the particles

J =
n

k=1

j
k
. (16.20)
J
2
is a two-body operator given in terms of the vector components of Eq. (16.20) by
J
2
= J

J
+
+J
2
z
+J
z
. (16.21)
The second-quantized form for the one-body vector operators are

J
z
=

< [ J
z
[ > a
+

a
+

km
ma
+
km
a
km
, (16.22)
and

km
_
j(j + 1) m(m1) a
+
k,m1
a
km
. (16.23)
One can operate with

J
2
on any M-state Slater determinant to determine whether or
not it has good angular momentum. For example, for the closed-shell conguration
of a given single-particle state dened by
[ C
k
>= a
+
kj
a
+
k,j1
. . .a
+
k,j
[> (16.24)
16 MANY-BODY WAVEFUNCTIONS 174
one obtains

J
z
[ C
k
>= M [ C
k
>=

km
m a
+
km
a
km
[ C
k
>
=
j

m=j
m [ C
k
>= 0 [ C
k
>, (16.25)
and

[ C
k
>=

km
_
j(j + 1) m(m1) a
+
k,m1
a
km
[ C
k
>= 0, (16.26)
since the state with quantum numbers (k, m1) is already occupied in [C
k
>. Hence
M = 0 and by Eq. (16.21) J = 0. That is, the closed-shell conguration is spherically
symmetric. The general closed-shell conguration
[ C >=

a
[ C
ka
> (16.27)
also has J = M = 0, and thus only partially lled shells contribute to non-zero
angular momenta.
The closed-shell plus one-particle conguration (the single-particle conguration)
can be represented by
[ Ck
a
> a
+
kama
[ C > . (16.28)
This state is also an eigenstate of

J
2
and

J
z
with eigenvalues j
a
(j
a
+ 1) and m
a
,
respectively.
The closed-shell minus one-particle conguration (the single-hole conguration)
with one particle absent in the state (k

) is given by
[ C(k

)
1
>= (1)
jm
a
km
[ C > (16.29)
Operation of

J
z
and

J
2
on this state gives

J
z
[ C(k

)
1
>=
m=j

m=j,m=m
m [ C(k

, m

)
1
>
= m

[ C(k

, m

)
1
> . (16.30)
and

J
2
[ C(k

, m

)
1
>= j

(j

+ 1) [ C(k

, m

)
1
> . (16.31)
Hence, the single-hole conguration as dened by Eq. (16.29) has J = j

and M =
m

. The single-hole conguration with J = j

and M = m

is given by the
closed-shell conguration with one particle absent in the state (k

, m

)
[ Ck
1

> (1)
j+m
a
k,m
[ C > a
km
[ C >, (16.32)
16 MANY-BODY WAVEFUNCTIONS 175
where a is the tensor operator of rank j

.
In general, multi-particle M-states [

(M)> are not eigenstates of

J
2
(they do
not have a good J value). However, states with good J are linear combinations of
the M-states:
[ (JM) >=
d(M)

=1
c

(M) > . (16.33)

Conversely, the M-states are linear combinations of the J-states:
[
a
(M) >=
Jmax

JM
D(J)

=1
d
J
[ (JM) > . (16.34)
The d(M) and D(J) are the M-scheme and J-scheme dimensions, respectively. Any
conguration which has d(M) = 1 must have good angular momentum J = M
max
,
since there is only one term in Eqs. (16.33). In particular, the state [ (M
max
)> with
M = M
max
has J = M
max
. This can also be seen from the fact that

J
+
[ (M
max
) >= 0, (16.35)
and hence

J
2
[ (M
max
) >= (J
2
z
+J
z
) [ (M
max
) >
= (M
max
+ 1)M
max
[ (M
max
) > . (16.36)
An operator that projects out the component with J = J
o
leaving the component
with J = J
i
unchanged is

P
J
i
Jo

J
2
J
o
(J
o
+ 1)
J
i
(J
i
+ 1) J
o
(J
o
+ 1)
. (16.37)
This can be seen by operating with

P on the wave function of Eq. (16.34)

P
J
i
Jo
[
a
(M) >=
Jmax

JM,J=Jo
D(J)

=1
d

J
[ (JM) >, (16.38)
where
d

J
=
J(J + 1) J
o
(J
o
+ 1)
J
i
(J
i
+ 1) J
o
(J
o
+ 1)
d
J
. (16.39)
That is, the component with J = J
o
is missing and the component with J = J
i
has
d

= d.
The operator, which projects out all of the unwanted components with J = J
o
leaving only one component with J = J
i
, is thus
P
J
i

Jmax

Jo=M,Jo=J
i

P
J
i
Jo
=
Jmax

Jo=M,Jo=J
i

J
2
J
o
(J
o
+ 1)
J
i
(J
i
+ 1) J
o
(J
o
+ 1)
. (16.40)
16 MANY-BODY WAVEFUNCTIONS 176
P
J
i
acting on the wave function of Eq. (16.34) gives
P
J
i
[
a
(M) >=
D(J)

=1
d
J
i
[ (J
i
M) > . (16.41)
Since each of the states (JM) is a linear combination of M-states, the sum in the
above equation is also be a linear combination of M-states
D(J)

=1
d
J
i
[ (J
i
M) >=
d(M)

b=1
e
b
[
b
(M) > . (16.42)
Evaluation of the real coecients e
b
will be illustrated below by some examples.
When d(M) > D(J), the set of J-states obtained from operating P
J
i
on all of the
M-states is overcomplete. Also it may happen the projection from dierent M-states
produces states which are not linearly independent (as in the above examples). Thus,
one must carefully choose D(J) out of the total set of d(M) projected states which
will form a complete basis. In addition, when D(J) > 1, the projected states will not
necessarily be orthogonal, and they must be orthogonalized by the Gram-Schmidt
orthogonalization procedure.
In general, the angular-momentum projection operator can be applied to cong-
urations with particles distributed over several single-orbital states. The projection
operator cannot change the partition (the number of particles in each single-orbital
state). Thus the overlaps between the good J states in dierent partitions is auto-
matically zero. Only one partition need be considered at a time for constructing the
good J states.
The projection operator is unitary,
P
J
i
P
J
i
= P
J
i
, (16.43)
and it commutes with the spherically symmetric Hamiltonian operator,
[H, P
J
i
] = 0. (16.44)
Thus, matrix elements of the Hamiltonian operator with the states of good J can be
obtained by projecting on only one side
<
a
[ P
J
i
HP
J
i
[
b
>=<
a
[ HP
J
i
P
J
i
[
b
>=<
a
[ HP
J
i
[
b
> . (16.45)
16.3.1 Examples for (j = 5/2)
n
congurations
Example (i): For the (j = 5/2)
2
conguration there are two M-states with M = 2
[ a >= a
+
1/2
a
+
5/2
[> (16.46)
16 MANY-BODY WAVEFUNCTIONS 177
and
[ b >= a
+
1/2
a
+
3/2
[> . (16.47)
(a
+
1/2
a
+
j=5/2,m=1/2
, etc.) These must by associated with the J-states that have
J =2 or 4. From Eq. (16.4), J = 4 is the maximum J-value allowed, and due to the
Pauli principle [Eq. (16.7)] J = 3 is not allowed. Operation with

J
2
gives

J
2
[ a >= 15 [ a > +

45 [ b > (16.48)
and

J
2
[ b >=

45 [ a > +11 [ b > . (16.49)

The linear combination of [ a > and [ b > that has J = 2 can be obtained by
projecting out the states with J
o
=3 and 4 by operating on [ a > or [ b > with P
J=2
and projecting out the states with J
o
=3 and 4; in particular [ a > gives
P
J=2
[ a >=

P
J=2,Jo=3

P
J=2,Jo=4
[ a >=
_

J
2
12
6 12
__

J
2
20
6 20
_
[ a >
=
_

J
2
12
6 12
_

5
14
_
_
_

5
14
[ a >

9
14
[ b >
_
_
_
=
5
14
[ a >

45
14
[ b >=

5
14
_
_
_

5
14
[ a >

9
14
[ b >
_
_
_
. (16.50)
(Note that the operation of projecting out J
o
= 3 had no eect because J
o
= 3 is
forbidden by the Pauli principle.) The normalization factor taken out in front in the
last line shows that the probability of nding the component J = 2 component in
state [ a > is 5/14 (and thus the probability to nd the J = 4 component must be
9/14). In this two-particle example the factors inside the brackets in the last line
must be within a phase factor of the result obtained with Eq. (16.6) with N
12
=
1

2
1

2
(2) < 5/2, 1/2, 5/2, 5/2 [ 2, 2 >=

5
14
(16.51)
and
1

2
(2) < 5/2, 1/2, 5/2, 3/2 [ 2, 2 >=

9
14
, (16.52)
where the factor of (2) comes from the fact that there are two terms in Eq. (16.6)
which contribute to a given antisymmetric M-scheme basis state. The normalized
wave function is
P
J=2
[ a >
N
a
=

5
14
[ a >

9
14
[ b > (16.53)
16 MANY-BODY WAVEFUNCTIONS 178
The same result (up to a phase factor) would be obtained fromP
J=2
acting on the state
[ b >. In general the projection operator applied to any two-particle conguration
gives results which are equivalent to those in Sec. 16.2. However, the projection
operator can also be applied in the general case of the n-particle conguration to
obtain wave functions with good J, as in the next example.
Example (ii): From Table 16.1 it is seen that there are two M-states for the (j = 5/2)
3
conguration with M = 5/2
[c >= a
+
3/2
a
+
3/2
a
+
5/2
[> (16.54)
and
[ d >= a
+
1/2
a
+
1/2
a
+
5/2
[> (16.55)
The operation of

J
2
on these states gives

J
2
[c >=
67
4
[c > +
32
4
[ d >
and

J
2
[ d >=
67
4
[c > +
32
4
[ d > (16.56)
Since there is no J
o
= 7/2 component, the state with J = 5/2 can be obtained by
projecting out the J
o
= 9/2 component from state [c >
P
J=5/2
[c >=
(

J
2

99
4
)
(
35
4

99
4
)
[c >=
1

2
_

2
[c > +
1

2
[ d >
_
. (16.57)
The normalized projected wave functions is given by
P
J=5/2
[c >
N
c
=
_

2
[c > +
1

2
[ d >
_
. (16.58)
16 MANY-BODY WAVEFUNCTIONS 179
16.4 General form for the matrix elements in the J scheme
In this section, a few general results for the matrix elements for creation operator in
the J-scheme are given. These will be used in subsequent sections when the matrix
elments of the J-coupled forms are derived.
In addition to the angular momentum recoupling, attention must be paid to the
phase factors which occur due to the reordering of particles between groups. Exam-
ples of these phase factors for the wave functions themselves appear above in Eqs.
(16.10) and (16.13). As an example for the particle-number-dependent phase factors
in matrix elements of the creation and destruction operators, consider a matrix ele-
ment involving the wave functions of Eqs. (16.13) and (16.14) with operators O(k
i
n

i
)
which involve n

i
creation and/or destruction operators for the state k
i
< (k
n
1
1

1
J
1
)(k
n
2
2

2
J
2
)JM [ O(k
1
n

1
)O(k
2
n

2
) [ (k
n

1
1

1
J

1
)(k
n

2
2

2
J

2
)J

>
=

M
i
M

i
< J
1
M
1
J
2
M
2
[ JM >< J

1
M

1
J

2
M

2
[ J

>
<[ Z(k
n
1
1

1
J
1
M
1
)Z(k
n
2
2

2
J
2
M
2
)O(k
1
n

1
)O(k
2
n

2
)
Z
+
(k
n

2
2

2
J

2
M

2
)Z
+
(k
n

1
1

1
J

1
M

1
) [>
= (1)
n
2
n

1
+n

1
(n
2
+n

2
+n

2
)

M
i
M

i
< J
1
M
1
J
2
M
2
[ JM >< J

1
M

1
J

2
M

2
[ J

>
<[ Z(k
n
1
1

1
J
1
M
1
)O(k
1
n

1
)Z
+
(k
n

1
1

1
J

1
M

1
)
Z(k
n
2
2

2
J
2
M
2
)O(k
2
n

2
)Z
+
(k
n

2
2

2
J

2
M

2
) [> . (16.59)
The phase factor arises from rst commuting the state operator Z for n
2
with the op-
erator O for n

1
, and next commuting the state operator Z
+
for n

1
with the operators
for n

2
, n

2
and n
2
. The phase factor can be simplied by noting that n
i
+ n

i
+ n

i
is
always even. This can be seen by considering n

i
as a sum of the number of creation,
c

i
, and destruction, d

i
, operators. It is clear that the total number of destruction
operators, n
i
+d

i
, must equal the total number of creation operators, n

i
+c

i
. Since
n
i
+d

i
= n

i
+c

i
, and both are integers, the sum n
i
+d

i
+n

i
+c

i
= n
i
+n

i
+n

i
must
be even. This also means that (1)
n

i
= (1)
n
i
+n

i
. Thus the phase factor becomes
(1)
n
2
n

1
+n

1
(n
2
+n

2
+n

2
)
= (1)
n
2
n

1
= (1)
n
2
(n
1
+n

1
)
. (16.60)
A general angular momentum recoupling result that is often needed is the matrix
element for an operator which acts only on particles in a set of active states (n
a
J
a
)
and not on those in a set of inactive states (n
i
J
i
)
< (n
i

i
J
i
)(n
a

a
J
a
)J[[O

(n

a
)[[(n

i
J

i
)(n

a
J

a
)J

>
REFERENCES 180
= (1)
J

i
+J

a
+J+
_
(2J + 1)(2J

+ 1)
_
J
a
J

a

J

J J
i
_
< n
a

a
J
a
[[O

(n

a
)[[n

a
J

a
>
n
i
n

J
i
J

i
. (16.61)
where Eq. 13.48 has been used and the number-dependent phase factor from Eq.
(16.60) is (1)
n(2n
i
)
= 1. The matrix element with the active and inactive states
interchanged is the same except for the phase factor; from Eq. 13.49
< (n
a

a
J
a
)(n
i

i
J
i
)J[[O

(n

a
)[[(n

a
J

a
)(n

i
J

i
)J

>
= (1)
Ja+J
i
+J

+
_
(2J + 1)(2J

+ 1)
_
J
a
J

a

J

J J
i
_
< n
a

a
J
a
[[O

(n

a
)[[n

a
J

a
>
n
i
n

J
i
J

i
. (16.62)
References
[1] W. D. M. Rae, A. Etchegoyen and B. A. Brown, OXBASH, the Oxford-Buenos
Aries-MSU Shell-Model Code, Michigan State University Laboratory Report no.
524.
[2] R. R. Whitehead, A. Watt, B. J. Cole and I. Morrison, Advances in Nuclear
Nuclear Physics 9, 123, (1977); L. M. Mackenzie, A. M. Macleod, D. J. Berry and
R. R. Whitehead, Comput. Phys. Comm. 48, 229 (1988).
[3] R. D. Lawson, Theory of the Nuclear Shell Model, (Clarendon Press, 1980).
[4] T. Sebe and M. Harvey, Enumeration of Many-Body States of the Nuclear Shell
Model with Denite Angular Momentum and Isobaric Spin with Mixed Single
Particle Orbits, ARCL, Chalk River preprint 3007 (1968).
[5] D. K. Sunko, Phys. Rev. C33, 1811 (1986).
[6] J. P. Draayer and H. T. Valdes, Comput. Phys. Comm. 36, 313 (1985).
[7] P. J. Brussaard and P. W. M. Glaudemans, Shell Model Applications in Nuclear
Spectroscopy, (North Holland, 1977).
[8] A. de Shalit and I. Talmi, Nuclear Shell Theory, (Academic Press, 1963).
[9] L. B. Hubbard, Comput. Phys. Commun. 1, 453 (1970).
17 THE TWO-BODY HAMILTONIAN 181
17 The Two-Body Hamiltonian
17.1 Introduction
In this section we will discuss the methods used to calculate two-body hamiltonian
matrix elements for use in microscopic nuclear structure studies. The starting point
for the two-body interaction is the data on nucleon-nucleon (NN) scattering and its
interpretation in terms of NN potentials. For very light nuclei A 8 one can carry
out calculations with Greens Function Monte-Carlo techniques in which the NN
interaction is a direct input [1]. The result is that the theoretical binding energies
are about 25% smaller than experiment. This corresponds to the interaction energy
being about 5% to small (there is a large cancellation between kinetic and potential
energy). To do better one has to introduce phenomenological three-body interactions
for which direct experimental data does not exist.
Calculations for heavier nuclei are based upon a sub-space (a model space) within
the shell-model basis. For these calculations the NN interaction must be renormalized
to take into account the nite basis. This renormalization is usually divided into two
steps. The rst step uses Brueckner theory to deal with strongly repulsive short-
range part of the interaction. This theory uses the NN interaction to obtain G
matrix elements. The resulting G matrix elements can be expressed in terms of an
eective one-boson exchange potential such as M3Y [2] or HKT [3]. Alternatively one
may take only the low-momentum (low-k) part of the NN interaction to calculate the
two-body matrix elements. In this section we will discuss methods used to calculate
two-body interactions from eective potential models.
The second step is to renormalize the G matrix elements to take into account
congurations which are omitted from the model space that are not included in the
Brueckner G matrix. The are called core-polarization corrections and the two-body
matrix elements are called G
eff
. Calculations with G
eff
are rather successful for the
spectra of nuclei with a few valence nucleons removed from the closed shells [4], [5].
For systems with three or more valence nucleons, the agreement with experiment
can be improved by using eective potentials or eective two-body matrix elements
V
eff
that are determined from the binding energy data. The dierences between
G
eff
and V
eff
are not well understood, but part of it may come from the omitted
three-body interactions.
17 THE TWO-BODY HAMILTONIAN 182
17.2 Form of the two-body matrix elements
Antisymmetric J-coupled two-body matrix elements are the primary ingredients for
the setting up the Hamiltonian matrix. With the results of Sec. 16.2 they can be
evaluated by expanding them into direct and exchange terms involving the product
wavefunctions:
< k
1
k
2
J [ V [ k
3
k
4
J >
= N
12
N
34

m
i
< j
1
m
1
j
2
m
2
[ JM >< j
3
m
3
j
4
m
4
[ JM >< k
1
m
1
k
2
m
2
[ V [ k
3
m
3
k
4
m
4
>
= N
12
N
34

m
i
< j
1
m
1
j
2
m
2
[ JM >< j
3
m
3
j
4
m
4
[ JM >
< k
1
m
1
k
2
m
2
[ V [ k
3
m
3
k
4
m
4
>
p
< k
1
m
1
k
2
m
2
[ V [ k
4
m
4
k
3
m
3
>
p
, (17.1)
where the subscripte p indicates that this is the product wavefunction. With the
symmetry properties of the Clebsch-Gordan coecients the sums of m
i
can be carried
to obtain:
< k
1
k
2
J [ V [ k
3
k
4
J >=
1
_
[(1 +
k
1
k
2
)(1 +
k
3
k
4
)]

_
< k
1
k
2
J [ V [ k
3
k
4
J >
p
(1)
j
3
+j
4
J
< k
1
k
2
J [ V [ k
4
k
3
J >
p
_
. (17.2)
Since the interaction V is Hermitian and the matrix elements are real, one has
< k
3
k
4
J [ V [ k
1
k
2
J >=< k
1
k
2
J [ V [ k
3
k
4
J > . (17.3)
In addition, one has from the wavefunction symmetry:
< k
1
k
2
J [ V [ k
4
k
3
J >= (1)
j
3
+j
4
J
< k
1
k
2
J [ V [ k
3
k
4
J >, (17.4)
and
< k
2
k
1
J [ V [ k
3
k
4
J >= (1)
j
1
+j
2
J
< k
1
k
2
J [ V [ k
3
k
4
J > . (17.5)
Thus it is sucient to calculate the two-body matrix element for only one-ordering
of k
1
, k
2
, k
3
and k
4
.
When isospin is introduced the factors associated with j and J are doubled with
those involving t =
1
2
and T and the last four equations become:
< k
1
k
2
JT [ V [ k
3
k
4
JT >=
1
_
[(1 +
k
1
k
2
)(1 +
k
3
k
4
)]

_
< k
1
k
2
JT [ V [ k
3
k
4
JT >
p
(1)
j
3
+j
4
J+1T
< k
1
k
2
JT [ V [ k
4
k
3
JT >
p
_
,
(17.6)
17 THE TWO-BODY HAMILTONIAN 183
< k
3
k
4
JT [ V [ k
1
k
2
JT >=< k
1
k
2
JT [ V [ k
3
k
4
JT >, (17.7)
< k
1
k
2
JT [ V [ k
4
k
3
JT >= (1)
j
3
+j
4
J+1T
< k
1
k
2
JT [ V [ k
3
k
4
JT >, (17.8)
and
< k
2
k
1
JT [ V [ k
3
k
4
JT >= (1)
j
1
+j
2
J+1T
< k
1
k
2
JT [ V [ k
3
k
4
JT > . (17.9)
The M-scheme two-body matrix elements can be expressed in terms of these
J-coupled two-body matrix elements
< k
1
m
1
k
2
m
2
[ V [ k
3
m
3
k
4
m
4
>=
_
(1 +
k
1
k
2
)(1 +
k
3
k
4
)

JM
< JM [ j
1
m
1
j
2
m
2
>< JM [ j
3
m
3
j
4
m
4
>< k
1
k
2
J [ V [ k
3
k
4
J > . (17.10)
When isospin is introduced this becomes
< (k
1
m
1
t
z
)(k
2
m
2
t

z
) [ V [ (k
3
m
3
t
z
)(k
4
m
4
t

z
) >=
_
(1 +
k
1
k
2
)(1 +
k
3
k
4
)

JMTTz
< JM [ j
1
m
1
j
2
m
2
>< JM [ j
3
m
3
j
4
m
4
>
< TT
z
[
1
2
t
z
1
2
t

z
>< TT
z
[
1
2
t
z
1
2
t

z
>< k
1
k
2
JT [ V [ k
3
k
4
JT > . (17.11)
Conservation of charge means that there are two basic types of interactions. Those
with (t
z
, t

z
) = (p, p) or (t
z
, t

z
) = (n, n) which must have T = 1, and those with
(t
z
, t

z
) = (p, n) which are a linear combination of T = 0 and T = 1.
17.3 General types of interactions
The general form of the two-body potential consists of central (c), spin-orbit (s) and
spin-tensor (t) components:
V
c
= f
c
(r), (17.12)
V
s
= f
s
(r)

L

S, (17.13)
and
V
t
= f
t
(r) S
12
. (17.14)
In these equations,

S = (
1
+
2
)/2, (17.15)

L is relative angular momentum operator, and S

12
is the spin-tensor operator
S
12
= 3(
1
r)(
2
r) (
1

2
), (17.16)
17 THE TWO-BODY HAMILTONIAN 184
where r is the relative coordinate
r = r
1
r
2
,
with magnitude r =[ r [. (17.17)
In order to carry out the tensor algbra it is useful to rewite the potential in terms
of tensor operators U
(q)
which act on the spacial coordinates and X
(q)
which act on
the spin coordinates:
V =

q
V
q
=

q
U
(q)
X
(q)
, (17.18)
The interaction components are specied by q = 0 for central, q = 1 for spin-orbit
and q = 2 for spin-tensor.
In order to carry out the tensor algebra one can rewrite the interaction of Eqs.
(17.12)-(17.14) in the form of Eq. (17.18) with the spacial parts U given by:
U
(0)
(r) = f
c
(r), (17.19)
U
(1)
(r) = f
s
(r)

L, (17.20)
and
U
(2)
(r) = f
t
(r) Y
(2)
( r), (17.21)
and the spin parts X given by
X
(0)
= 1, (17.22)
X
(1)
=

S, (17.23)
and
X
(2)
=

24
5
[
1

2
]
(2)
. (17.24)
To evaluate the space and spin components of the matrix elements of Eq. (17.18)
we need to transform from jj to LS coupling:
< k
a
k
b
JT [ V [ k
c
k
d
JT >
p
=

LSL

a
1/2 j
a

b
1/2 j
b
L S J
_

_
_

c
1/2 j
c

d
1/2 j
d
L

J
_

_
< k
A
k
B
LSJT [ V [ k
C
k
D
L

JT >
p
. (17.25)
k
A
is shorthand for the set of quantum numbers (n
a

a
) and k
a
is shorthand for the
set (n
a

a
j
a
), etc. The large square bracket is the LS to jj transformation coecient
given by Eq. 14.25. With Eq. (17.18), these LS coupled matrix elements are then
expressed in terms of reduced matrix elements:
< k
A
k
B
LSJT [ V [ k
C
k
D
L

JT >
p
17 THE TWO-BODY HAMILTONIAN 185
=

q
< k
A
k
B
LSJT [ V
q
[ k
C
k
D
L

JT >
p
=

q
(1)
S+L

+J
_
L S J
S

q
_
< k
A
k
B
L[[U
(q)
[[k
C
k
D
L

>< ST[[X
(q)
[[S

T > . (17.26)
17.4 Transformation from relative to center-of-mass coordi-
nates
The shell-model wavefunctions are expressed in terms of the coordinates r
1
and r
2
with
respect to a xed center, but the NN interaction depends on the relative coordinate
r. The transformation between center-of-mass and relative coordinates can be carried
out in an oscillator basis. The reason is that the oscillator hamiltonian is seperable
in both coordinates:
H =
p
2
1
2m
+
p
2
2
2m
+
1
2
m
2
r
2
1
+
1
2
m
2
r
2
2
=
Q
2
2M
+
1
2
M
2
R
2
+
p
2
2
+
1
2

2
r
2
, (17.27)
where M = 2m and = m/2 (m is the nucleon mass). The quantum numbers of the
usual shell-model basis are:
[ k
1
k
2
L >=[ (n
1

1
)(n
2

2
)L >, (17.28)
and the quantum numbers associated with the relative coordinates are:
[ (n)(N)L >, (17.29)
where (n, ) are the radial and angular-momentum quantum numbers associated with
the relative coordinate r, and (N, ) go with the center-of-mass coordinate

R. The
transformation between these two sets of coordinates is given by:
[ k
1
k
2
L >=[ (n
1

1
)(n
2

2
)L >=

nN
< nNL [ n
1

1
n
2

2
L > [ (n)(N)L > .
(17.30)
These transformation coecients are called Moshinsky or Talmi-Moshinsky brackets
after the authors who rst calculated their properties. Note that changing the mass
from m to = m/2 for the relative motion means that the relative radial wavefunc-
tions use an oscillator parameter which is eectively smaller than by a factor of
two.
For the matrix element of Eq. (17.26) the spatial wavefunctions [ k
A
k
B
> and
[ k
C
k
D
> are expressed in terms of the relative (n) and center-of-mass (N) coor-
dinates using the Talmi-Moshinsky transformation:
< k
A
k
B
L[[U
(q)
[[k
C
k
D
L

>=

nn

NN

< nN [ k
A
k
B
L >
17 THE TWO-BODY HAMILTONIAN 186
< n

[ k
C
k
D
L

>< nNL[[U
(q)
(r)[[n

> . (17.31)
With Eq. 13.49 this last integral reduces to:
< nNL[[U
(q)
(r)[[n

>= (1)
++L

+q
_
(2L + 1)(2L

+ 1)

q
L

L
_
< n[[U
(q)
(r)[[n

>
N N

. (17.32)
With the U(r) given by Eqs. (17.19)-(17.21) the relative integrals reduce to:
< n[[U
(0)
(r)[[n

>=< n [ f
c
(r) [ n

>

2 + 1

, (17.33)
< n[[U
(1)
(r)[[n

>=< n [ f
s
(r) [ n

>
_
( + 1)(2 + 1)

, (17.34)
and
< n[[U
(2)
(r)[[n

>
= (1)

< n [ f
t
(r) [ n

>
_
(2 + 1)(2

+ 1)
_
2

0 0 0
_
. (17.35)
The central matrix element (q = 0) of Eq. (17.26) simplies to
< k
A
k
B
LSJT [ V
0
[ k
C
k
D
L

JT >
p
=
=

nn

NN

< nN [ k
A
k
B
L >< n

[ k
C
k
D
L

>
< n [ f
c
(r) [ n

>
N N

LL

S,S
. (17.36)
17.5 Simple potentials
The two-body Coulomb is given by the central interaction:
V
coul
= e
2
/r. (17.37)
It is independent of spin and acts only between protons (T = 1).
The simplest short-ranged interaction might be approximated by a delta function:
V
delta
= W(r), (17.38)
with a strength W which may depend upon the total spin S. Only =

=0 con-
tributes to the integral. The spacial symmetry of two nucleons in an =0 state is
symmetric, and thus S +T must be odd to make the total two-nucleon wavefunction
17 THE TWO-BODY HAMILTONIAN 187
antisymmetric. Thus the delta function only acts in the (S, T) = (0, 1) and (1, 0)
states. The scalar operator for spin that selects out the S components is
(
1

2
) [ S = 0 >= 3,
and
(
1

2
) [ S = 1 >= 1. (17.39)
Thus the most general delta-function interaction can be witten:
V
delta
= W[1 + (
1

2
)](r), (17.40)
with typical strength values of W = 500 MeV fm
3
and = 0.12. Thus for (S, T) =
(0, 1) W(1 3) = 180 MeV fm
3
, and for (S, T) = (1, 0) W(1 + ) = 560 MeV
fm
3
.
17.6 One boson exchange potentials
The one-pion exchange potential (OPEP) is the longest ranged part of the NN in-
teraction due to the small mass of 140 MeV for the pion. The form of the OPEP
interaction is:
V
OPEP
(r) = g
2
(
1

2
) [(
1

)(
2

)]f(r), (17.41)
where g is the pion-nucleon coupling constant, and f(r) is the Yukawa form:
f(r) =
e
(r/)
r/

. (17.42)
The range

is given by

= h/(m

c) = 1.414 fm. (17.43)

Eq. (17.41) can be expanded and written in terms of the central and spin-tensor
operators
V
OPEP
(r) =
g
2
3
(
1

2
)
_
S
12
_
3
r
2
+
3

r
+
1

_
f(r) +
(
1

2
)

f(r) 4(
1

2
)(r)
_
.
(17.44)
In the G matrix one takes only the low-k components and the (r) term is dropped.
The isospin operator is analogous to those for the spin in Eq. (17.39)
(
1

2
) [ T = 0 >= 3,
and
(
1

2
) [ T = 1 >= 1. (17.45)
17 THE TWO-BODY HAMILTONIAN 188
Thus, the central OPEP potential which contains the the term (
1

2
)(
1

2
) gives
a relative weighting of 9, 3, 3, and 1 to the (S, T) channels (0,0), (0,1), (1,0) and
(1,1), respectively. The overall strength of the OPEP interaction as determined from
pion-nucleon scattering data is (g
2

) = 42 MeV fm
3
.
Another way to write the OPEP strength is in terms of the projection operators
P
S,T
= P
S
P
T
. These operators project out the part with specic S and T values. For
the spin part:
P
S=0
[ S = 0 >= 1, P
S=0
[ S = 1 >= 0,
P
S=1
[ S = 0 >= 0, P
S=1
[ S = 1 >= 1, (17.46)
or in terms of the spin operators:
P
S=0
=
1 (
1

2
)
4
,
P
S=1
=
3 + (
1

2
)
4
, (17.47)
with a similar set of equations for the isospin projection. Thus we could write in
terms of P
S,T
:
(
1

2
)(
1

2
) = P
11
3P
01
3P
10
+ 9P
00
.
Hosaka, Kubo and Toki (HKT) [3] parameterized their G-matrix in terms of one-
boson-exchange potential (OBEP) forms [6]. In general the functions f(r) can depend
on the value of the total spin S and isospin T. Thus we introduce these two indicies
as well strength parmeters W which depend on the channel; central (c), spin-orbit
(s) or spin-tensor (t); the range index, i, and S and T:
f
c,S,T
(r) = P
S,T

i
W
c,i,S,T
e
x
i
x
i
, (17.48)
f
s,S,T
(r) = P
S,T

i
W
s,i,S,T
_
1 +
1
x
i
_
e
x
i
x
2
i
, (17.49)
and
f
t,S,T
(r) = P
S,T

i
W
t,i,S,T
_
1 +
3
x
i
+
3
x
2
i
_
e
x
i
x
i
. (17.50)
The x
i
is r divided by the range parameter
i
:
x
i
= r/
i
. (17.51)
The HKT parameterization requires four ranges for the central component:
i
= 0.20,
0.33, 0.50 and 1.414 fm; two ranges for the tensor component:
i
= 0.25 and 1.414
fm; and two ranges for the spin-orbit component:
i
= 0.25 and 0.40 fm. The range
17 THE TWO-BODY HAMILTONIAN 189
W
(
r
)
r (fm)
0.0
0.1
0 2 4 6
n=0,l=0
n=0,l=2
f
(
r
)

x

r
^
2
-150
-100
-50
0
50
potential for S=0, T=1
one-pion
total
W
(
r
)

x

f
(
r
)

x

r
^
2
-12
-8
-4
0
4
8
integral from 0 to r
n=0,l=0
n=0,l=2
Figure 1: The central S = 0, T = 1 channel for the the HKT interaction.
1.414 fm is that for the one-pion exchange. The shorter ranges are associated with
two-pion, rho and omega meson exchange. The strength parameters W were obtained
from a t to G-matrix elements calculated from the Paris N-N potential [7] and are
given in Table 2 of [3].
Contributions to the integral
< n [ f
c
(r) [ n >=
_

n,
(r)f
c
(r)
n,
r
2
dr
for the central part of the HKT interaction are shown in Figs. (17.1) (17.4). The
bottom panels show the contribution W =
2
n,
in the integrand from the relative
radial wavefunctions y, the middle panels show the radial part of the potential mul-
tiplied by r
2
and the top panels show the integral evaluated out to radius r. Note
that the orbital angular momentum of the relative wavefunction is restricted by the
symmetry condition that +S +T is odd for a two-particle wavefunction. Thus, for
17 THE TWO-BODY HAMILTONIAN 190
W
(
r
)
r (fm)
0.0
0.1
0 2 4 6
n=0,l=0
n=0,l=2
f
(
r
)

x

r
^
2
-150
-100
-50
0
50
potential for S=1, T=0
one-pion
total
W
(
r
)

x

f
(
r
)

x

r
^
2
-12
-8
-4
0
4
8
integral from 0 to r
n=0,l=0
n=0,l=2
Figure 2: The central S = 1, T = 0 channel for the the HKT interaction.
17 THE TWO-BODY HAMILTONIAN 191
W
(
r
)
r (fm)
0.0
0.1
0 2 4 6
n=0,l=1
f
(
r
)

x

r
^
2
-50
0
50
100
150
potential for S=0, T=0
one-pion
total
W
(
r
)

x

f
(
r
)

x

r
^
2
-12
-8
-4
0
4
8
integral from 0 to r
n=0,l=1
Figure 3: The central S = 0, T = 0 channel for the the HKT interaction.
17 THE TWO-BODY HAMILTONIAN 192
W
(
r
)
r (fm)
0.0
0.1
0 2 4 6
n=0,l=1
f
(
r
)

x

r
^
2
-50
0
50
100
150
potential for S=1, T=1
one-pion
total
W
(
r
)

x

f
(
r
)

x

r
^
2
-12
-8
-4
0
4
8
integral from 0 to r
n=0,l=1
Figure 4: The central S = 1, T = 1 channel for the the HKT interaction.
17 THE TWO-BODY HAMILTONIAN 193
a short range potential, the most important contributions are for =0 in the (S, T)
= (0,1) and (1,0) channels.
The longest ranged part of the HKT interaction corresponds to the one-pion-
exchange (OPEP) potential. The OPEP contribution to the bare G-matrix provides
about 1/3 of the attractive interaction needed for the (S, T) = (0,1) and (1,0) chan-
nels of the central component. A common feature of this as well as other OBEP
parameterizations is a strong attractive medium-range part which is partly cancelled
out by a repulsive short-range part.
17.7 Numerical examples
In Table 1 and 2 I compared some typical values of the two-body matrix elements
obtained with the potentials discussed in the proceeding section.
V
con
= 1,
V
coul
is the Coulomb potential of Eq. (17.37),
V
del
is the delta-function potential of Eq. (17.40),
V
OPEP
is the one-pion exchange from Eq. (17.44),
V
G
is the HKT G-matrix potential from Eq. (17.48).
17.8 Density-dependent interactions
Phenomenological density-dependent interactions have long played an important role
in Hartree-Fock calculations [8], [9], [10] as well as in shell-model calculations [11],
[12], [13]. From a more fundamental point of view, density dependence enters into
the calculations of the G-matrix in nite nuclei [14]. There also has been a renewed
interest in density dependence because of recent developments with relativistic models
[15]. For a density-dependent interaction there is an additional dependence on the
center-of-mass coordinate:

R = (r
1
+r
2
)/2. (17.52)
The spacial functions U
(q)
are extended to include a dependence on R:

U
(q)
(R, r) = U
(q)
(r) F
q
(R), (17.53)
and Eq. (17.32) will become:
< nNL[[

U
(q)
[[n

>= (1)
++L+q
_
(2L + 1)(2L

+ 1)
17 THE TWO-BODY HAMILTONIAN 194
Table 1: Values of some central potential matrix elements < k
1
k
2
JT [ V [
k
3
k
4
JT > (in MeV) with h=13.9 MeV. They are labeled by k = 4 for 0d
5/2
and k = 6 for 1s
1/2
.
k
1
k
2
k
3
k
4
J T V
con
V
coul
V
del
V
OPEP
V
G
4 4 4 4 0 1 1 0.493 3.109 0.873 2.916
4 4 4 4 2 1 1 0.416 0.711 0.236 1.052
4 4 4 4 4 1 1 0.387 0.296 0.063 0.567
4 4 4 6 2 1 0 0.029 0.496 0.183 0.687
4 4 6 6 0 1 0 0.046 1.068 0.291 0.833
4 6 4 6 2 1 1 0.420 0.740 0.420 1.523
4 6 4 6 3 1 1 0.388 0.000 0.156 0.088
6 6 6 6 0 1 1 0.452 2.529 0.951 2.514
4 4 4 4 1 0 1 2.954 0.527 1.027
4 4 4 4 3 0 1 1.658 0.038 1.317
4 4 4 4 5 0 1 2.591 0.834 3.740
4 4 4 6 3 0 0 1.269 0.268 1.487
4 4 6 6 1 0 0 1.277 0.199 0.916
4 6 4 6 2 0 1 0.864 0.520 0.239
4 6 4 6 3 0 1 2.159 0.804 3.370
6 6 6 6 1 0 1 4.426 0.951 3.784
Table 2: Values of some spin-orbit (s) and spin-tensor (t) potential matrix
elements < k
1
k
2
JT [ V [ k
3
k
4
JT > (in MeV) with h=13.9 MeV. They are
labeled by k = 4 for 0d
5/2
and k = 6 for 1s
1/2
.
k
1
k
2
k
3
k
4
J T V
OPEP
(t) V
G
(t) V
G
(s)
4 4 4 4 0 1 1.078 0.732 0.343
4 4 4 4 2 1 0.089 0.043 0.186
4 4 4 4 4 1 0.190 0.139 0.049
4 4 4 6 2 1 0.052 0.071 0.025
4 4 6 6 0 1 0.000 0.000 0.000
4 6 4 6 2 1 0.334 0.245 0.025
4 6 4 6 3 1 0.239 0.175 0.124
6 6 6 6 0 1 0.000 0.000 0.000
4 4 4 4 1 0 0.977 0.450 0.011
4 4 4 4 3 0 0.657 0.414 0.028
4 4 4 4 5 0 0.198 0.183 0.034
4 4 4 6 3 0 0.076 0.094 0.006
4 4 6 6 1 0 0.743 0.460 0.000
4 6 4 6 2 0 0.124 0.153 0.011
4 6 4 6 3 0 0.086 0.109 0.057
6 6 6 6 1 0 0.000 0.000 0.000
17 THE TWO-BODY HAMILTONIAN 195

q
L

L
_
< n[[U
(q)
(r)[[n

>< N [ F
q
(R) [ N

>

. (17.54)
A typical form for the density dependence is:
D(R) = 1 +A
d
F(R)
B
d
, (17.55)
where and
F(R) = 1 + exp[(R R
o
)/a]
1
, (17.56)
where A
d
and B
d
are constants to be chosen. Initially we take R
o
= 1.1A
1/3
fm
(with A=16 for the sd shell) and a = 0.6 fm. In the original Skyrme Hartree-Fock
calculations [16] and work by Migdal [13], linear density dependence (B
d
= 1) was
assumed. Note that F(R) is the Fermi function which approaches unity as R goes to
zero.
It is interesting to compare this density dependence with the commonly used
surface-delta interaction (SDI) for the central component [11], [12]. For a zero-range
interaction the radial dependence of the two-body matrix element reduces to the
integral [11]
I =
_
D(R)
A
(R)
B
(R)
C
(R)
D
(R)R
2
dR. (17.57)
For a surface-delta interaction, one could evaluate this integral with D(R) = (RR
o
)
for some value of R
o
near the surface. However, in practice one usually replaces the
integral by unity times a phase factor which takes into account the fact that the radial
wavefunctions (R) conventionally are dened to be positive near the origin):
I = (1)
na+n
b
+nc+n
d
. (17.58)
A more realistic form of the density dependence is given by D(R) in Eq. (17.55)
with A
d
= 1 (i.e., an interaction which goes smoothly to zero in the nuclear inte-
rior). In Table (17.1) the radial matrix elements for the sd shell calculated with a
density-independent zero-range interaction and with various versions of the density-
dependent, zero-range interactions are compared. Very large dierences among the
various models is noted. Also,it is noted that the linear density dependence gives re-
sults which are closest to the commonly used constant approximation of Eq. (17.58).
Many of the successful phenomenological global interactions, such the D1 interaction
of Decharge and Gogny [9], have a density-dependence which makes the interaction
much smaller in the nuclear interior than on the surface.
REFERENCES 196
References
[1] B. R. Wiringa, S. C. Pieper, J. Carlson and V. R. Pandharipande, Phys. Rev. C
62, 014001 (2000).
[2] N. Anantaraman, H. Toki, G. F. Bertsch, Nucl. Phys. A398, 269 (1983).
[3] A. Hosaka, K. I. Kubo, H. Toki, Nucl. Phys. A444, 76 (1985). Note the following
corrections: in Eq. (2) the in Eq. (2) should be replaced by s, in Eq. (3) the
Y
LS
(x) term should have the factor e
x
/x
2
rather than e
x
/x, and in Table 3 the
range should be 0.40 rather than 0.50.
[4] T. T. S. Kuo, Ann. Reviews of Nucl. Sc. 24, 101 (1974).
[5] J. Shurpin, T. T. S. Kuo and D. Strottman, Nucl. Phys. A408, 310 (1983).
[6] R. Bryan and B. L. Scott, Phys. Rev. C177, 1435 (1969).
[7] M. Lacombe, B. Loiseau, J. M. Richard, R. Vinh Mau, J. Cote, P. Pires and R.
de Tourreil, Phys. Rev. C 21, 861 (1980).
[8] Nguyen van Giai and H. Sagawa, Phys. Lett. 106B, 379 (1981).
[9] J. Decharge and D. Gogny, Phys. Rev. C21, 1568 (1980).
[10] J. Friedrich and P. G. Reinhard, Phys. Rev. C33, 335 (1986).
[11] P. J. Brussaard and P. W. M. Glaudemans, Shell Model Applications in Nuclear
Spectroscopy (North-Holland, 1977).
[12] P. W. M. Glaudemans, P. J. Brussaard and B. H. Wildenthal, Nucl. Phys. A102,
593 (1967).
[13] A. B. Migdal, Nuclear Theory; The Quasiparticle Method (Benjamin, New York,
1968); Theory of Finite Fermi Systems and Applications to Atomic Nuclei (Wiley,
New York, 1967).
[14] J. W. Negele, Phys. Rev. C1, 1260 (1970).
[15] M. R. Anastasio, L. S. Celenza, W. S. Pong and C. M. Shakin, Physics Reports
100, 328 (1983); L. S. Celemza, W. S. Pong and C. M. Shakin, Phys. Rev. Lett.
47, 3 (1981).
[16] D. Vautherin and D. M. Brink, Phys. Rev. C5, 626 (1972).
18 APPLICATIONS OF THE TWO-BODY INTERACTIONS 197
18 Applications of the Two-Body Interactions
18.1 Interaction energies for closed shell, one-particle and
one-hole congurations
The total energy for a diagonal conguration in the M-scheme is the sum of kinetic
energy and interaction energy contributions.
E = E
K
+E
I
, (18.1)
where the kinetic energy is
E
K
=

< [ T [ >, (18.2)

and the interaction energy is
E
I
=
1
2

,
< [ V [ > . (18.3)
In this section we will derive the interaction energies for the closed-shell and closed-
shell plus-or-minus one-particle congurations in terms of the J-coupled two-body
matrix elements. For a closed shell the sum runs over all of the lled states labeled
by (k, m), where k stands for (n
r
, , j). In terms of explicit sums over the k orbits the
kinetic energy is
E
K
(C) =

ka,ma
< k
a
m
a
[ T [ k
a
m
a
>=

ka
N
a
< k
a
[ T [ k
a
>, (18.4)
where the factor of
N
a
= (2j
a
+ 1) (18.5)
comes from number of degenerate m
a
states which have the same kinetic energy. The
interaction energy is
E
I
(C) =
1
2

ka,k
b
,ma,m
b
< k
a
m
a
k
b
m
b
[ V [ k
a
m
a
k
b
m
b
> . (18.6)
We can replace the M-scheme two-body matrix element with the J-scheme result
given in Eq. 17.10:
E
I
(C) =
1
2

ka,k
b
,ma,m
b
[1 +
ka,k
b
]

JMJ

< JM [ j
a
m
a
j
b
m
b
> < J

[ j
a
m
a
j
b
m
b
>
< k
a
k
b
J [ V [ k
a
k
b
J > . (18.7)
18 APPLICATIONS OF THE TWO-BODY INTERACTIONS 198
With

ma,m
b
< JM [ j
a
m
a
j
b
m
b
> < J

[ j
a
m
a
j
b
m
b
>=
J,J

M,M
this becomes:
E
I
(C) =
1
2

ka,k
b
[1 +
ka,k
b
]

JM
< k
a
k
b
J [ V [ k
a
k
b
J > . (18.8)
Note that
1
2

ka,k
b
[1 +
ka,k
b
] is equivalent to
kak
b
. Also < V > does not depend on
M and there are (2J + 1) values of M. Thus we have
E
I
(C) =

kak
b

J
(2J + 1) < k
a
k
b
J [ V [ k
a
k
b
J > . (18.9)
The average (monopole) interaction energy between two k orbits is dened as:

V
kak
b
=

J
(2J + 1) < k
a
k
b
J [ V [ k
a
k
b
J >

J
(2J + 1)
. (18.10)
When k
a
,= k
b
we have
J
(2J + 1) = N
a
N
b
, where N = (2j + 1) is the lled shell
orbit occupancy. This represents the total number of two-body interactions when one
particle is in orbit k
a
and the other particle is in orbit k
b
. When both particles are
in the same orbit k
a
, then
J
(2J +1) = N
a
(N
a
1)/2, which represents the number
of two-body interactions for the particles within orbit k
a
. In terms of these average
interactions the total closed-shell interaction energy is:
E
I
(C) =

ka
N
a
(N
a
1)
2

V
kaka
+

ka>k
b
N
a
N
b

V
kak
b
, (18.11)
where N = (2j + 1).
We can use these average interactions to calculate the energy associated with
one-hole and one-particle congurations. Thus for one-particle in orbit k
i
outside the
closed-shell we sum over the average interactions with the lled orbits k
a
multiplied
by the total number of particles N
a
= (2j
a
+ 1) in that orbit:

k
i
= E(C k
i
) E(C) =< k
i
[ T [ k
i
> +

ka
N
a

V
ka,k
i
. (18.12)
For one hole in orbit k
i
the result is:

k
i
= E(C) E(C k
1
i
) =< k
i
[ T [ k
i
> +(N
i
1)

V
k
i
,k
i
+

ka=k
i
N
a

V
ka,k
i
, (18.13)
where the factor in front of the

V
k
i
,k
i
term is the dierence between the total number
of interactions for the lled shell, N
a
(N
a
1)/2, and the total number of interactions
for a lled shell minus one particle, (N
a
1)(N
a
2)/2.
18 APPLICATIONS OF THE TWO-BODY INTERACTIONS 199
When isospin is included the number of nucleons in the lled shell becomes:
N

a
= (2t + 1)(2j
a
+ 1) = 2(2j
a
+ 1), (18.14)
where t = 1/2. The closed-shell kinetic energy is:
E
K
(C) =

ka,ma
< k
a
m
a
[ T [ k
a
m
a
>=

ka
N

a
< k
a
[ T [ k
a
>, (18.15)
the closed-shell interaction energy is:
E
I
(C) =

kak
b

J,T
(2T + 1)(2J + 1) < k
a
k
b
JT [ V [ k
a
k
b
JT >, (18.16)
the average interaction is:

V
kak
b
=

JT
(2J + 1)(2T + 1) < k
a
k
b
JT [ V [ k
a
k
b
JT >

JT
(2J + 1)(2T + 1)
, (18.17)
the single-particle energy is:

k
i
=< k
i
[ T [ k
i
> +

ka
N

V
ka,k
i
, (18.18)
and the single-hole energy is:

k
i
=< k
i
[ T [ k
i
> + (N

i
1)

V
k
i
,k
i
+

ka=k
i
N

V
ka,k
i
. (18.19)
As an example, we can calculate the interaction energy for
16
O with the matrix
elements of the low-k NN interaction derived by Bogner et al. [1] (the sum of the
central, spin-orbit and spin-tensor terms). We take oscillator potential with h=14
MeV provides an approximation to the Hartree-Fock potential. The required two-
body matrix elements are given in Table 1. The total sum from Eq. (18.16) is E
I
= 325.3 MeV. The total kinetic energy from Sec. 7.2 is E
K
= 18 h = 252 MeV.
We should also subtract the center-of-mass kinetic energy E
K
(cm) =
3
4
h = 10.5
MeV. Finally we can add the total Coulomb energy by using Eq. (18.9) to sum the
proton-proton two-body matrix elements of the Coulomb interaction to get E(coul)
= 13.8 MeV. (The proton-proton matrix elements are given by the T = 1 Coulomb
matrix elements in Table 1.) Thus the total energy is:
E = E
I
+E
K
E
K
(cm) +E(coul) =
= 325.3 + 252 10.5 + 13.8 MeV = 70.0 MeV. (18.20)
18 APPLICATIONS OF THE TWO-BODY INTERACTIONS 200
Table 1: Values of the some V (lowk) and Coulomb two-body matrix elements
used in this section. The orbits are labeled by k = 1 for 0s
1/2
k = 2 for 0p
3/2
,
k = 3 for 0p
1/2
, and k = 4 for 0d
5/2
.
k
1
k
2
k
3
k
4
J T V
lowk
V
coul
1 1 1 1 1 0 7.86
1 1 1 1 0 1 6.72 0.67
average 7.29 0.67
1 2 1 2 1 0 1.07
1 2 1 2 2 0 7.86
1 2 1 2 1 1 3.80 0.59
1 2 1 2 2 1 1.00 0.44
average 2.87 0.50
1 3 1 3 0 0 7.86
1 3 1 3 1 0 4.47
1 3 1 3 0 1 1.88 0.44
1 3 1 3 1 1 0.88 0.52
average 2.18 0.50
2 2 2 2 1 0 1.39
2 2 2 2 3 0 3.95
2 2 2 2 0 1 3.30 0.56
2 2 2 2 2 1 1.37 0.47
average 2.22 0.48
2 3 2 3 1 0 5.54
2 3 2 3 2 0 5.05
2 3 2 3 1 1 0.53 0.44
2 3 2 3 2 1 2.50 0.49
average 2.63 0.47
3 3 3 3 1 0 2.26
3 3 3 3 0 1 0.63 0.50
average 1.44 0.50
4 4 4 4 1 0 0.28
4 4 4 4 3 0 0.72
4 4 4 4 5 0 2.94
4 4 4 4 0 1 1.79
4 4 4 4 2 1 1.15
4 4 4 4 4 1 0.51
18 APPLICATIONS OF THE TWO-BODY INTERACTIONS 201
This should be compared with the experimental value of E = BE = 127.6 MeV.
That is, an dierence of 58 MeV (corresponding to 18% in E
I
(lowk)). To do better we
need to include congurations beyond the closed-shell such as two-particle two-hole
admixtures. Also perhaps three-body interactions must be added.
One can also use the two-body matrix elements from Table 1 to obtain the single-
hole energies for 0s
1/2
, 0p
3/2
and 0p
1/2
. For example, for the 0p
1/2
we would use Eq.
(18.19) with

V
33
= 1.44 MeV,

V
23
= 2.63 MeV, and

V
13
= 2.18 MeV to get
3

V
33
+ 8

V
23
+ 4

V
13
= 34.1 MeV for the interaction energy. The kinetic energy is
5
4
h = 17.5 MeV and the single-hole energy from the sum is 16.58 MeV. This is not
too far from the experimental value of
n
= [BE(
16
O) BE(
15
O)] = 15.66 MeV
for the neutron.
To get the 0p
1/2
proton single-hole energy we would also include the contribution
from the Coulomb interaction between protons. From Eq. (18.13) and Table 1 this
is (0.50) + 4(0.47) + 2(0.50) = 3.38 MeV (for the sum over 0p
1/2
, 0p
3/2
and 0s
1/2
,
respectively). The proton 0p
1/2
hole energy is thus 16.58 3.38 = 13.20 MeV
compared to the experimental value of
p
= [BE(
16
O) BE(
15
N)] = 12.13 MeV.
Finally we could also compare the calculated value for
p

n
= 3.38 MeV (just the
Coulomb energy term) with the experimental value of 3.54 MeV.
18.2 Interaction energies for diagonal two-particle congu-
rations
The diagonal matrix element of the hamiltonian for two-particles outside of a closed
shell as given by Eq. 12.4 is:
E(C + 2) =< C [ H [ C >= E(C) +

+ < [ V [ >, (18.21)

where C represents all of the orbitals in the closed shell, and where are interpreted
in Hartree-Fock theory as the single-particle energies for a single-particle outside the
closed shell. In the M-scheme the and could be in two dierent m values in the
same k = (n
r
, , j) orbital or in two dierent orbitals k
a
and k
b
. It is useful to dene
an energy relative to that of the closed shell:
E = E(C + 2) E(C) =

+ < [ V [ >, (18.22)

These energy dierences provide a dierent kind of test for the shell-model methods.
In the J-scheme with both particles in the same k orbit we would have:
E = E(C + 2) E(C) = 2
k
+ < k
2
J [ V [ k
2
J > . (18.23)
18 APPLICATIONS OF THE TWO-BODY INTERACTIONS 202
For example, as a model for
18
O we could use Eq. (18.23) with the observed single-
particle energy of the lowest 5/2
+
state in
17
O.

0d
5/2
= [BE(
17
O) BE(
16
O)] = 4.14 MeV (18.24)
For two neutrons in j = 5/2 we can have J = 0, 2 and 4. If we take the lowest states
with these spins in
18
O (the 2
+
and 1.98 MeV and the 4
+
at 3.55 MeV) has having
these congurations we have from the experimental binding energies:
E
exp
(0
+
) = [BE(
18
O), 0
+
) BE(
16
O)] = 12.19 MeV,
E
exp
(2
+
) = [BE(
18
O), 2
+
) BE(
16
O)] = 10.21 MeV,
and
E
exp
(4
+
) = [BE(
18
O), 4
+
) BE(
16
O)] = 8.64 MeV. (18.25)
These can be compared to those calculated with the experimental single-particle
energy of 4.14 MeV and the theoretical two-body matrix elements from Table 1:
E
th
(0
+
) = 2(4.14) 1.79 = 10.07 MeV,
E
th
(2
+
) = 2(4.14) 1.15 = 9.43 MeV,
and
E
th
(4
+
) = 2(4.14) 0.51 = 8.79 MeV. (18.26)
The deviation between experiment and theory is largest for J = 0. This dierence
is mainly due to mixing with other congurations as will be discussed in the next
section.
This simple model for two-particles in one orbit outside of a closed shell works
best when that one orbit is isolated in energy from the others. The best examples
of these T = 1 two-particle spectra are found in
50
Ti for two 0f
7/2
protons outside
of
48
Ca,
92
Mo for two 0g
9/2
protons outside of
90
Zr and
210
Po for two 0h
9/2
protons
outside of
208
Pb.
In this one-orbit approximation, we can use the equations to deduce experimen-
tal values for the (T = 1) two-body matrix elements:
< k
2
J [ V [ k
2
J >
exp
= [E(C + 2) E(C)] 2
= [E(C + 2) E(C)] 2[E(C + 1) E(C)]
= [BE(C + 2) BE(C)] + 2[BE(C + 1) BE(C)]. (18.27)
In the example for
18
O
< (0d
5/2
)
2
J = 0 [ V [ (0d
5/2
)
2
J = 0 >
exp
= 3.91 MeV,
18 APPLICATIONS OF THE TWO-BODY INTERACTIONS 203
< (0d
5/2
)
2
J = 0 [ V [ (0d
5/2
)
2
J = 0 >
exp
= 1.93 MeV,
and
< (0d
5/2
)
2
J = 0 [ V [ (0d
5/2
)
2
J = 0 >
exp
= 0.36 MeV. (18.28)
These can be compared to the theoretical values of 1.79, 1.15 and 0.51 MeV,
respectively.
18.3 Interaction energies for diagonal two-hole congurations
The diagonal matrix element of the hamiltonian for two-particles outside of a closed
shell as given by Eq. 12.7 is:
E(C 2) =< C
1

1
[ H [ C
1

1
>= E(C)

+ < [ V [ >,
(18.29)
where C represents all of the orbitals in the closed shell, and where are interpreted
in Hartree-Fock theory as the single-particle energies for the one-hole conguration.
The energy relative to that of the closed shell is:
E = E(C 2) E(C) =

+ < [ V [ >, (18.30)

The formalism for two-holes is identical to that of two-particles except for the sign in
front of . This result can be generalized to many holes; the formalism for n holes is
the same as that for n particles except for the sign in front of .
In the J-scheme for two holes in the same orbit:
E = E(C 2) E(C) = 2
k
+ < k
2
J [ V [ k
2
J > . (18.31)
A comparison of Eqs. (18.23) and (18.31) shows that the energy spectra for the hole
conguration is the same as that of the particle conguration, as long as the two-body
matrix elements are the same. Experimental values for the two-body matrix elements
can be derived for the properties of the hole nuclei by:
< k
2
J [ V [ k
2
J >
exp
= [E(C 2) E(C)] + 2
= [E(C 2) E(C)] + 2[E(C) E(C 1)]
= [BE(C) BE(C 2)] 2[BE(C) BE(C 1)]. (18.32)
One of best examples can be found in the nucleus
54
Fe with two holes in the 0f
7/2
orbit relative to a closed shell for
56
Ni.
18 APPLICATIONS OF THE TWO-BODY INTERACTIONS 204
18.4 Interaction energies for particle-hole congurations
The diagonal matrix element of the hamiltonian for a particle-hole state is:
E(ph) =< C
1
[ H [ C
1
>= E(C) +

<

[ V [

>, (18.33)
where C represents all of the orbitals in the closed shell, and where are interpreted
in Hartree-Fock theory as the single-particle energies.

b is the m-scheme wavefunction
for the state
[

>= (1)
j

+m

[ k

, m

> . (18.34)
The energy of the particle-hole state coupled to total angular momentum J relative
to that of the closed shell is:
E(ph, J) = E(ph) E(C) =

+V (ph, J), (18.35)

where V (ph, J) is as linear combinations of two-body matrix elements known as the
Pandya transformation [2]
V (ph, J) =

(2J

+ 1)
_
j

J
j

_
< k

[ V [ k

> . (18.36)
Extension to isospin formalism gives:
E(ph, JT) = E(ph) E(C) =

+V (ph, JT), (18.37)

where
V (ph, JT) =

(2J

+ 1)(2T

+ 1)
_
j

J
j

__
1/2 1/2 T
1/2 1/2 T

_
< k

[ V [ k

> . (18.38)
Puttng in the values of the isospin factors one has:
V (ph, JT = 0) =

(2J

+ 1)
_
j

J
j

_
[
3
2
< k

= 1 [ V [ k

= 1 >
1
2
< k

= 0 [ V [ k

= 0 >],
(18.39)
and
V (ph, JT = 1) =

(2J

+ 1)
_
j

J
j

_
[
1
2
< k

= 1 [ V [ k

= 1 > +
1
2
< k

= 0 [ V [ k

= 0 >].
(18.40)
REFERENCES 205
References
[1] S. Bogner, T. T. S. Kuo, L. Coraggio, A. Covello and N. Itaco, Phys. Rev. C 65,
051130(R) (2002).
[2] S. P. Pandya, Phys. Rev. 103, 956 (1956).
19 CONFIGURATION MIXING 206
19 Conguration mixing
19.1 Two-particle congurations
In
17
O one observes that the 1s
1/2
orbit lies only 0.87 MeV above the 0d
5/2
ground
states. Thus both of these orbits should be considered for the two-particle congura-
tion in
18
O. For J

= 0
+
we take the two J-coupled basis states:
[
1
>=[ C, aa J >=[ C, (0d
5/2
)
2
J >,
and
[
2
>=[ C, bb J >=[ C, (1s
1/2
)
2
J >, (19.1)
where C indicates all orbits in the
16
O closed shell. The ground state will be a
linear combination of these two basis states which are determined by diagonalizing
the hamiltonian matrix:
_

H
11

H
12

H
21

H
22
_
, (19.2)
where

H
11
=< C, aa J [ H [ C, aa J >= E(C) + 2
a
+ < aa J [ V [ aa J >,

H
22
=< C, bb J [ H [ C, bb J >= E(C) + 2
b
+ < bb J [ V [ bb J >,
and

H
12
=

H
21
=< C, aa J [ H [ C, bb J >=< aa J [ V [ bb J > . (19.3)
For the o-diagonal matrix element,

H
21
=

H
12
, since the matrix is real and hermitian.
Since the two basis states dier by the change of two particles between orbits a and b,
the only contribution to H
12
comes from the operator

a
+

a
+

acting to move
the two particles between orbits a and b (if H operates on a state other than those
in a and b the remaining overlaps of the form < a [ b > will give zero).
The core energy E(C) appears as a constant in the diagonal of the matrix. Thus
we can consider the energies E relative to the closed-shell energy given by the eigen-
values of
H =

H E(C)
as given by the matrix
_
H
11
H
12
H
21
H
22
_
, (19.4)
H
11
= 2
a
+ < aa J [ V [ aa J >,
H
22
= 2
b
+ < bb J [ V [ bb J >,
19 CONFIGURATION MIXING 207
and
H
12
=< aa J [ V [ bb J > . (19.5)
We will assume that the matrix is arranged such that:
H
11
< H
22
. . . ,
e.g. H
11
is the lowest diagonal energy. The eigenvalues E for this two-by-two matrix
are given by the two solutions of the quadratic equation obtained from determinant
[ H E [:
E
1/2
=
1
2
_
H
22
+H
11

_
(H
22
H
11
)
2
+ 4H
2
12
_
. (19.6)
The subscripts 1(2) indicate the eigenstates which are associated with
1(2)
when
H
12
= 0. We can write the energies as:
E
1
= H
11
,
and
E
2
= H
22
+, (19.7)
where
=
1
2
_
_
(H
22
H
11
)
2
+ 4H
2
12
(H
22
H
11
)
_
, (19.8)
where > 0. In terms of the eigenfunctions are given by:
[
1
>=
1
N
[[
1
> (/H
12
) [
2
>] ,
and
[
2
>=
1
N
[[
2
> +(/H
12
) [
1
>] , (19.9)
with normalization:
N
2
= 1 + (/H
12
)
2
Note that [
i
> are also eigenstates. Thus there are random phase factors
in the wave functions which must be kept consistent throughout the calculations of
physical observables (which always involve a given wavefunction twice).
It is useful to look at various limits of these results. When the states start out
degenerate H
11
= H
22
, then = H
12
, and the states are repelled to have a total
separation energy of E
2
E
1
= 2, and the eigenstates are equal mixtures of the
basis states [
1
> and [
2
>.
When the diagonal separation energy is large compared to o-diagonal matrix
element; H
22
H
11
[ H
12
[, then
E
1
= H
11
H
11

H
2
12
H
22
H
11
,
19 CONFIGURATION MIXING 208
Table 1: Values of the some V (lowk) two-body matrix elements used in this
section. The orbits are labeled j = 5 for 0d
5/2
, j = 3 for 0d
3/2
and j = 1 for
1s
1/2
.
j
1
j
2
j
3
j
4
J T V
lowk
5 5 5 5 0 1 1.79
5 5 3 3 0 1 3.47
5 5 1 1 0 1 0.83
3 3 3 3 0 1 0.37
3 3 1 1 0 1 0.68
1 1 1 1 0 1 2.53
5 5 5 5 2 1 1.15
5 5 5 3 2 1 0.44
5 5 5 1 2 1 0.63
5 5 3 3 2 1 0.69
5 5 3 1 2 1 0.60
5 3 5 3 2 1 0.42
5 3 5 1 2 1 0.23
5 3 3 3 2 1 0.87
5 3 3 1 2 1 0.69
5 1 5 1 2 1 1.27
5 1 3 3 2 1 0.80
5 1 3 1 2 1 1.51
3 3 3 3 2 1 0.28
3 3 3 1 2 1 0.07
3 1 3 1 2 1 0.66
5 5 5 5 4 1 0.51
5 5 5 3 4 1 1.09
5 3 5 3 4 1 2.15
E
2
= H
22
+ H
22
+
H
2
12
H
22
H
11
, (19.10)
[
1
>[
1
>
H
12
H
22
H
11
[
2
>,
[
2
>[
2
> +
H
12
H
22
H
11
[
1
>, (19.11)
where the normalization in the wavefunction is ignored since the last term is small.
These are equivalent to the perturbation theory results given in Eq. 9.13 and 9.15,
respectively. When H
12
= 0 the eigenstates start out at energies of H
11
and H
22
and
as H
12
is turned on the states repel each other by an additional amount 2.
19 CONFIGURATION MIXING 209
19.2 Application to
18
O
We can apply these results to the
18
O J

= 0
+
state by using the experimental
single-particle energies:

5
=
0d
5/2
= [BE(
17
O, 5/2
+
, gs) BE(
16
O] = 4.14 MeV,
and

1
=
1s
1/2
= [BE(
17
O, 1/2
+
, 0.87 MeV) BE(
16
O] = 3.27 MeV,
where we have introduced a short-hand notation (2j +1). The basis states for J = 0
are:
[
1
>=[ 55J >=[ (0d
5/2
)
2
J = 0 >,
and
[
2
>=[ 11J >=[ (1s
1/2
)
2
J = 0 >, (19.1)
The single-particle energies together with the low-k two-body matrix elements from
Table 1 give the hamiltonian matrix elements:
H
11
= 2
5
+ < 55J [ V [ 55J >= 2(4.14) 1.79 = 10.07 MeV
H
22
= 2
1
+ < 11J [ V [ 11J >= 2(3.27) 2.53 = 9.07 MeV
and
H
12
=< 55J [ V [ 11J >= 0.83 MeV, (19.12)
The lowest eigenenergy is 10.54 MeV which is an improvement over the single-orbit
(0d
5/2
) result of 10.07 MeV, but still not as low as the experimental value of 12.19
MeV.
For the 0
+
state the next conguration to consider would be
[
3
>=[ 33J > . (19.13)
for the 0d
3/2
orbit. If we take the 3/2
+
state at 5.08 MeV as the single-particle state
in
17
O, its single-particle energy is:

3
=
0d
3/2
= [BE(
17
O, 3/2
+
, 5.08 MeV) BE(
16
O] = 0.94 MeV,
The hamiltonian matrix is then a 3x3 matrix of the form:
_

_
H
11
H
12
H
13
H
21
H
22
H
23
H
31
H
32
H
33
_

_ , (19.14)
19 CONFIGURATION MIXING 210
where in addition to the terms given in Eq. (19.12) we need:
H
33
= 2
3
+ < 33J [ V [ 33J >= 2(0.94) 0.37 = 1.51 MeV,
H
13
=< 55J [ V [ 33J >= 3.47 MeV,
and
H
23
=< 11J [ V [ 33J >= 0.68 MeV. (19.15)
The lowest eigenvalue of the 3x3 matrix is 11.44 MeV which is getting close to
the experimental value of 12.19 MeV. To get better results we will need to include
congurations which will break the closed shell of
16
O.
From the matrix elements given in Table 1 we can also consider conguration
mixing for J = 2. There are ve ways to make J = 2:
[
1
>=[ 55J >,
[
2
>=[ 53J >,
[
3
>=[ 51J >,
[
4
>=[ 33J >,
and
[
5
>=[ 31J > . (19.16)
There will be o-diagonal matrix elements in which one-particle in moved from one
orbit to another. For example:
H
12
=< 55J [ V [ 53J > . (19.17)
Single-particle matrix elements of the kinetic energy (or the Hartree-Fock potential
U
HF
) are non-zero only for the diagonal M-scheme matrix elements of the type < [
T [ >, and thus can only enter into the diagonal part of the hamiltonian matrix.
Matrix elements of the type, < [ T [ > with ,= are zero since the one-body
scalar operators cannot change m or j.
19.3 Many-particle congurations
The basic ideas explored in the previous section for the conguration mixing of two
particles can be extended to many-particle systems. We start with a J-coupled basis
of the form:
[ nJ >=[ [(k
n
1
1

1
J
1
)(k
n
2
2

2
J
2
)J
12
][k
n
3
3

3
J
3
]J >,
[ n

J >=[ [(k
n

1
1

1
J

1
)(k
n

2
2

2
J

2
)J

12
][k
n

3
3

3
J

3
]J >, (19.18)
19 CONFIGURATION MIXING 211
where n = n
1
+ n
2
+ n
3
= n

1
+ n

2
+ n

3
. For a given total number of particles n,
the distribution n
1
, n
2
. . . among the k states is called the partition. The maximum
number of particles in one orbit is n = (2j +1). The indices distinguish the various
basis states with the same J value.
The diagonal hamiltonian matrix elements have the form:
< nJ [ H [ nJ >= n
1

k
1
+n
2

k
2
+n
3

k
3
+

i
D
i
(, , J) < V >
i
, (19.19)
and the o-diagonal matrix elements have the form:
< nJ [ H [ n

J >=

i
D
i
(,

, J) < V >
i
. (19.20)
The < V >
i
represent the list of possible two-body matrix elements in the model
space, and the D
i
are numerical coecients obtained from the matrix elements of

a
+

a
+

. For example if we take the closed-shell conguration:

[ nJ >=[ [k
2j
1
+1
1
J = 0 >,
then D
i
= (2J

+1) and V
i
=< k
2
1
J

[ V [ k
2
1
J

> for the allowed J

values. If we take
the two-particle conguration:
[ nJ >=[ [k
2
1
J >,
then there is only one non-zero value of D with D
i
= 1 for V
1
=< k
2
1
J [ V [ k
2
1
J >. In
general one must write an Mscheme or a Jscheme shell-model computer code to
calculate D
i
. If isospin is introduced, then all of the J in the equations are replaced
by (J, T) and the maximum number of particles in one orbit is n = 2(2j + 1).
Note that the o-diagonal matrix element does not contain the single-particle en-
ergies for the same reasons as discussed in the two-particle example. The o-diagonal
matrix element if zero if the set of numbers (n
1
, n
2
, n
3
) diers from (n

1
, n

2
, n

3
) by the
change of more than two particles. Thus if the basis states are ordered by their diag-
onal energy, the matrix will be banded with regions o-diagonal regions in which
are zero. An example of this banded matrix is shown in Fig. 1 for the 0
+
, T=0 levels
for 12 particles in the sd-shell (0d
5/2
,0d
3/2
,1s
1/2
) model space for the nuclus
28
Si [1].
The J, T matrix dimension is 839.
For the energy level spectrum I will use the example of the 0
+
, T = 0 states
obtained for the nucleus
24
Mg with 8 nucleons in the (0d
5/2
,0d
3/2
,1s
1/2
) model space.
For this calculation there are 35 partitions and
max
= 325 basis states. In the sd-shell
there are 63 independent (J, T) two-body matrix elements, so the sums over i in Eqs.
(19.19) and (19.20) is over 63 terms. The hamiltonian matrix is shown in Fig. 1 where
19 CONFIGURATION MIXING 212
Figure 1: The hamiltonian matrix for the 0
+
, T = 0 sd-shell basis in
28
Si. The
points represent the non-zero matrix elements. The matrix is symmetric and
only the bottom half is shown.
REFERENCES 213
one observes the bands of zeros corresponding to the regions where the partitions dier
by the change of more than two particles. Fig. 2 shows the eigenvalues has a function
of the strength of the o-diagonal part of the matrix. 100 percent corresponds to the
standard strength. Fig. 2 shows an expanded region of the results in Fig. 2. One
observes an overall level repulsion with the lowest state decreasing monotonically in
energy. The levels never cross but always show a repulsion typical of the two-level
mixing model.
References
[1] V. Zelevinsky, B. A. Brown and M. Horoi, Phys. Rep. 276, 85 (1996).
REFERENCES 214
Figure 2: The spectrum of 0
+
states in
24
Mg as a function of the strength (in
percent) of the o-diagonal part of the hamiltonian matrix.
REFERENCES 215
Figure 3: An expanded part of the previous gure.
20 ONE-PARTICLE TRANSFER 216
20 One-particle transfer
20.1 Fractional parentage coecients
20.1.1 One orbit
Observables for the removal or addition of a nucleon from a specic initial state to a
specic nal state are related to the matrix elements of the creation and destruction
operators. In this section we will study the basic properties of these matrix elements
and their sum-rules. The reduced matrix elements of the creation and destruction op-
erators are used to dene the spectroscopic factors associated with nuclear reactions.
They are also the building blocks for the more complicated operators associated with
one-body (such as electromagnetic and beta decay) and two-body transition ampli-
tudes.
The creation operator a
+
km
is a tensor of rank j since it creates the single-particle
state [ km > (k stands for the set of single-particle quantum numbers nj). The
destruction operator a
km
is not a tensor of rank j, however, by Eq. [13.34],
a
km
(1)
j+m
[a
+
k,m
]
+
= (1)
j+m
a
k,m
(20.1)
is a tensor of rank j. The inverse of Eq. (20.1) is
a
km
= (1)
jm
a
k,m
. (20.2)
Eq. [13.35] (with the matrix elements being real and p = j) can be used to relate the
reduced matrix elements involving a to those involving a
+
< k
n1

[[ a
k
[[k
n
J >= (1)
j+J

J
< k
n
J[[a
+
k
[[k
n1

>, (20.3)
All many-body matrix elements of a
+
can be reduced to these involving a single
k-state.
Wave-function expansion relations and sum-rules for the states in the n1 particle
system can be obtained by operating with the number operator

N
k
=

m
a
+
km
a
km
(20.4)
on the k
n
conguration and then inserting a complete set of states with n1 particles

N
k
[ k
n
JM >=

m
a
+
km
a
km
[ k
n
JM >= n [ k
n
JM >
=

a
+
km
[ k
n1

>< k
n1

[ a
km
[ k
n
JM > .
20 ONE-PARTICLE TRANSFER 217
The matrix element of a
k,m
can be reduced with the Wigner-Eckhart theorem
< k
n1

[ a
km
[ k
n
JM >= (1)
jm
< k
n1

[ a
k,m
[ k
n
JM >
= (1)
jm+J

_
J

j J
M

m M
_
< k
n1

[[ a
k
[[k
n
J >,
and by Eq. (20.3) the reduced matrix element of a can be converted in a reduced
matrix element of a
+
to obtain the nal result:

N
k
[ k
n
JM >= n [ k
n
JM >=

(1)
mM

+J
_
J

j J
M

m M
_
a
+
km
[ k
n1

>< k
n
J[[a
+
k
[[k
n1

> . (20.5)
We can thus expand the k
n
wavefunction in terms of those in the k
n1
basis:
[ k
n
JM >= (1)
n

< k
n
J[[a
+
k
[[k
n1

>
n
_
(2J + 1)
[Z
+
(k
n1

) a
+
k
]
J
M
[> . (20.6)
A phase factor of (1)
n1
arises from commuting a
+
with the n 1 particles in the
state k
n1
.
A sum-rule for the matrix elements of a
+
can be obtained by multiplying both
sides of Eq. (20.5) by < k
n

[ to obtain
n
JJ

MM

(1)
mM

+J
_
J

j J
M

m M
_
< k
n

[ a
+
km
[ k
n1

>< k
n
J[[a
+
k
[[k
n1

>
=

_
J

j J
M

m M
__
J

j J

m M

_
< k
n

[[a
+
k
[[k
n1

>< k
n
J[[a
+
k
[[k
n1

>
=

JJ

MM

(2J + 1)

< k
n

[[a
+
k
[[k
n1

>< k
n
J[[a
+
k
[[k
n1

>, (20.7)
where Eq. [13.12] has been used for the summation over 3j coecients. Thus one
nds the sum-rule:

< k
n

[[a
+
k
[[k
n1

>< k
n
J[[a
+
k
[[k
n1

>= n(2J + 1)

JJ
. (20.8)
Thus:

[< k
n
J[[a
+
k
[[k
n1

>[
2
=

[< k
n1

[[ a
k
[[k
n
J >[
2
= n(2J + 1). (20.9)
20 ONE-PARTICLE TRANSFER 218
The matrix elements in which the sum over nal states is normalized to unity are
historically called coecients of fractional parentage [1] (CFP) dened by:
< j
n
J [j
n1

>
< k
n
J[[a
+
k
[[k
n1

>
_
n(2J + 1)
. (20.10)
The j is used in the CFP rather than k, in order to emphasize the fact that these
coecients depend only on j and not on n and . In terms of CFP the sum-rule of
Eq. (20.9) becomes:

[< j
n
J [j
n1

>[
2
= 1. (20.11)
Thus, the square of the one-particle CFP is the probability of taking one particle in
the state k from the conguration [ k
n
J > leaving and leaving it in the conguration
[ k
n1

>.
A sum-rule related to the connection between states in the n and n + 1 particle
systems can be obtained with the number operator written in the form:

N
j
=

m
(1 a
jm
a
+
jm
). (20.12)
Multiplying this on the state k
n
and inserting a complete set of states of k
n+1
gives:

[< k
n+1

[[a
+
k
[[k
n
J >[
2
= (2J + 1)(2j + 1 n). (20.13)
Combining Eqs. (20.9) and (20.13) we obtain a sum-rule for the sum over all states
in both k
n1
and k
n+1
:

[< k
n1

[[ a
k
[[k
n
J >[
2
+

[< k
n+1

[[a
+
k
[[k
n
J >[
2
= (2J + 1)(2j + 1). (20.14)
When there is only one initial and nal state, the CFP are unity. The basic
examples are:
[< j
1
, J = j [j
0
, J

= 0 >[= 1, (20.15)
[< j
2
J [j
1
, J

= j >[= 1, (20.16)
where J is even, and
[< j
2j+1
, J = 0 [j
2j
, J

= j >[= 1. (20.17)
The wavefunctions for a closed-shell minus n particles can be expressed in terms
of those for holes:
[ k
n
J >[ k
2j+1n
J > . (20.18)
20 ONE-PARTICLE TRANSFER 219
The matrix elements of a for the hole-congurations are algebraically the same as the
matrix elements of a
+
for particle congurations:
< k
n

[[ a
k
[[k
(n1)
J >=< k
n

[[a
+
k
[[k
n1
J > . (20.19)
Thus from Eq. (20.3) we can relate the matrix elements of a
+
for hole states to those
for particle states:
< k
(n1)
J[[a
+
k
[[k
n

>= (1)
j+J

J
< k
n

[[ a
k
[[k
(n1)
J >
= (1)
j+J

J
< k
n

[[a
+
k
[[k
n1
J >, (20.20)
or
< k
2j+1(n1)
J[[a
+
k
[[k
2j+1n

>= (1)
j+J

J
< k
n

[[a
+
k
[[k
n1
J > . (20.21)
In terms of the one-particle CFP this becomes
< j
(n1)
J [j
n

>< j
2j+1(n1)
J [j
2j+1n

>
= (1)
j+J

_
n(2J

+ 1)
(2j + 2 n)(2J + 1)
< j
n

[j
n1
J > . (20.22)
We can convert the special results of Eqs. (20.15) and (20.16) for particles to
those for holes. When n = 1, J

= j, and J = 0 one obtain the same result as in Eq.

(20.17). For the case when n = 2 and J = j one obtains
[< j
2j
, J = j [j
2j1
J

>[=

_
(2J

+ 1)
j(2j + 1)
[< j
2
J

[j
1
, J = j >[=

_
(2J

+ 1)
j(2j + 1)
,
(20.23)
where J

is even. For example when we take one particle from the [(j = 5/2)
5
, J = 5/2]
conguration the probability to leave the system in the states [(j = 5/2)
4
, J

] is
2
30
10
30
and
18
30
, for J

= 0, 2 and 4, respectively.
There are a variety of methods and computer programs for calculating the one-
particle CFP [2], [3], [4], and they are tabulated in the literature [1], [5], [6] for small
values of n and j. Also the projection operator method discussed in the next section
can be used to calculate these CFP. All matrix elements in the J-scheme can be
reduced to equations involving these one-particle CFP, and thus they are a primary
input to any calculation or computer code that utilizes a J-coupled basis. As an
example, the one-particle CFP for the (j = 5/2)
3
and (j = 7/2)
3
conguration [1] are
given in Tables (20.1) and (20.2).
The results for specic CFP given in the text and tables implicitly have asso-
ciated a choice for the phase factors of the initial and nal wavefunctions. It is
important to keep these phase factors consistent throughout the intermediate stages
of a calculation, especially when the CFP are combined with the outputs of other
programs.
20 ONE-PARTICLE TRANSFER 220
Table (20.1). One-particle CFP for the (j = 5/2)
3
conguration.
JJ

0 2 4
3/2 0 -
_
5
7
_
2
7
5/2 -
_
2
9
_
5
18
_
1
2
9/2 0
_
3
14
-
_
11
14
Table (20.2). One-particle CFP for the (j = 7/2)
3
conguration.
JJ

0 2 4 6
3/2 0
_
3
14
-
_
11
14
0
5/2 0
_
11
18
_
2
33
-
_
65
198
7/2 -
_
1
4
_
5
36
_
1
4
_
13
36
9/2 0
_
13
126
-
_
50
77
_
49
198
11/2 0 -
_
5
18
_
13
66
-
_
52
99
15/2 0 0
_
5
22
-
_
17
22
20 ONE-PARTICLE TRANSFER 221
20.2 Many orbits
The reduced matrix elements of the creation and destruction operators for the full
many-body wave functions have the form:
<
A1
f

f
J
f
[[ a
k
[[
A
i

i
J
i
>= (1)
j+J
f
J
i
<
A
i

i
J
i
[[a
+
k
[[
A1
f

f
J
f
> . (20.24)
One can use the number operator to derive the sum-rules:

f
J
f
[<
A1
f

f
J
f
[[ a
k
[[
A
i

i
J
i
>[
2
= (2J
i
+ 1) < n
k
>
i
, (20.25)
and

f
J
f
[<
A+1
f

f
J
f
[[a
+
k
[[
A
i

i
J
i
>[
2
= (2J
i
+ 1)[(2j + 1) < n
k
>
i
]. (20.26)
In these expressions < n
k
>
i
is the number of particle in orbit k in state
i
. If one
has a pure conguration this number is an integer, but for a mixed conguration this
is the average number of particles in orbit k. For example, if one has a basis states
from a single partition, e.g.
[ nJ >
i
=[ [(k
n
1
1

1
J
1
)(k
n
2
2

2
J
2
)J
12
][k
n
3
3

3
J
3
]J >
i
,
then
< n
ka
>
i
= n
a
,
for a = 1, 2 and 3. If one has a wavefunction that contains two or more dierent
partitions, e.g.
[ >
i
= c [ >
i
+c

>
i
,
where
[ >
i
=[ [(k
n
1
1

1
J
1
)(k
n
2
2

2
J
2
)J
12
][k
n
3
3

3
J
3
]J >,
and
[

>
i
=[ [(k
n

1
1

1
J

1
)(k
n

2
2

2
J

2
)J

12
][k
n

3
3

3
J

3
]J >, (20.27)
then
< n
ka
>
i
= c
2
n
a
+c
2
n

a
,
for a = 1, 2 and 3.
20 ONE-PARTICLE TRANSFER 222
20.3 Spectroscopic factors
20.3.1 Basic denitions and sum-rules
The spectroscopic is dened in terms for the reduced matrix elements of a
+
by:
S =
[<
A
J[[a
+
k
[[
A1

>[
2
(2J + 1)
=
[<
A1

[[ a
k
[[
A
J >[
2
(2J + 1)
, (20.28)
where the (2J + 1) factor is by convention associated with the heavier mass A. S
implicitly depends upon all quantities in the expressions (explicit labels are added
when they are needed).
When the wavefunction is restricted to one orbital k with angular momentum j
the spectroscopic factor is related to the coecient of fractional parentage given in
Eq. [20.10] by:
S = n [< j
n
J [j
n1

>[
2
. (20.29)
The cross section for reactions involving the removal of a particle (proton or
neutron) from nucleus A is proportional to the full matrix element of a summed over
m and the nal M-states M
f
, and averaged over the initial M-states M
i
:

1
2J
i
+ 1

M
i
,M
f
,m
[<
A1
f

f
J
f
M
f
[ a
k,m
[
A
i

i
J
i
M
i
>[
2
=

M
i
,M
f
,m
_
J
f
j J
i
M
f
m M
i
__
J
f
j J
i
M
f
m M
i
_
[<
A1
f

f
J
f
[[ a
k
[[
A
i

i
J
i
>[
2
(2J
i
+ 1)
=
[<
A1
f

f
J
f
[[ a
k
[[
A
i

i
J
i
>[
2
(2J
i
+ 1)
= S. (20.30)
From the reaction theory one usually calculates a single-particle cross section
sp
(i.e. assuming that S = 1) for the removal of one particle (proton or neutron) in orbit
k and thus:

= S
sp
. (20.31)
Similarly the cross section for the addition of a particle to nucleus A is obtained from
the full matrix element of a
+
summed over m and M
f
and averaged over M
i
:

1
2J
i
+ 1

M
i
,M
f
,m
[<
A+1
f

f
J
f
M
f
[ a
+
k,m
[
A
i

i
J
i
M
i
>[
2
=

M
i
,M
f
,m
_
J
f
j J
i
M
f
m M
i
__
J
f
j J
i
M
f
m M
i
_
[<
A+1
f

f
J
f
[[a
+
k
[
A
i

i
J
i
>[
2
(2J
i
+ 1)
20 ONE-PARTICLE TRANSFER 223
=
[<
A+1
f

f
J
f
[[a
+
k
[[
A
i

i
J
i
>[
2
(2J
i
+ 1)
=
(2J
f
+ 1)
(2J
i
+ 1)
S, (20.32)
where S is the spectroscopic factor associated with the removal of a particle from
A+ 1 to A. Thus the cross section for the addition of a particle is given by:

+
=
(2J
f
+ 1)
(2J
i
+ 1)
S
sp
, (20.33)
where the spectroscopic factor S is for the removal of a nucleon from (A + 1) to A,
and the extra J factors take into account the dierent M state averaging for the two
type of reactions.
The particle removal sum-rule for spectroscopic factors is obtained from Eq.
[20.26] by summing over all states in the (A1) nucleus (denoted by f):

f
S
i,f,k
=< n
k
>
i
[A (A1)], (20.34)
and the particle addition sum-rule for spectroscopic factors is obtained from Eq.
[20.27] by summing over all states in the (A+ 1) nucleus (denoted by f+):

f+
(2J
f
+ 1)
(2J
i
+ 1)
S
i,f,k
= (2j + 1) < n
k
>
i
[A (A+ 1)], (20.35)
where < n
k
>
i
is the average occupation of protons or neutrons in orbit k in the
initial state i [with a maximum possible value of (2j + 1)]. The total sum-rule is:

f
S
i,f,k
+

f+
(2J
f
+ 1)
(2J
i
+ 1)
S
i,f,k
= (2j + 1). (20.36)
If we also sum Eq. (20.34) over all orbits we have a sum-rule for the total number of
particles (protons or neutrons) in the nucleus:

f,k
S
i,f,k
= Z or N [A (A1)]. (20.37)
20.3.2 Isospin dependence
The spectroscopic factors S as dened in Sec. 21.1 apply to protons or neutrons
without regard to isospin. One can generalize the results to nucleons in the state
(J, T) by adding the isospin labels. Eq. (20.28) becomes:
S(T) =
[<
A
JT[[[a
+
k
[[[
A1

>[
2
(2J + 1)(2T + 1)
=
[<
A1

[[[ a
k
[[[
A
J >[
2
(2J + 1)(2T + 1)
,
(20.38)
20 ONE-PARTICLE TRANSFER 224
where the triple bar matrix element indicates that it is reduced in both space and
isospin. The sum-rules can be applied to the nucleon occupancies, e.g.

f
S
i,f,k
(T) =< n
k
>
i
[A (A 1)], (20.39)
where < n
k
>
i
is the average occupation of nucleons in orbit k in the initial state i
(with a maximum possible value of 2(2j + 1)), and

f,k
S
i,f,k
(T) = A [A (A1)], (20.40)
When isospin is introduced we can also make explicit relationships between proton
and neutron spectroscopic factors and nucleon spectroscopic factors. Starting with
Eq. (20.28) but adding the explicit T
z
dependence:
S(t
z
) =
[<
A
JTT
z
[[a
+
k,tz
[[
A1

z
>[
2
(2J + 1)
, (20.41)
where our convention is (t, t
z
) = (
1
2
,
1
2
) for a neutron and (t, t
z
) = (
1
2
,
1
2
) for a proton,
we use the Wigner-Eckhart theorem in isospin space to obtain:
S(t
z
) =
_
T
1
2
T

T
z
t
z
T

z
_
2
[<
A
JT[[[a
+
k
[[[
A1

>[
2
(2J + 1)
=[< TT
z
[ T

z
tt
z
>[
2
[<
A
JT[[[a
+
k
[[[
A1

>[
2
(2J + 1)(2T + 1)
, (20.42)
for the removal of a nucleon and:
S(t
z
) =[< TT
z
[ T

z
tt
z
>[
2
[<
A+1
JT[[[a
+
k
[[[
A

>[
2
(2J + 1)(2T + 1)
, (20.43)
for the addition of a nucleon. Thus in shorthand notation:
S(t
z
) C
2
S(T), (20.44)
where C
2
is the square the isospin Clebsch and S(T) is the nucleon spectroscopic
factor given by Eq. (20.38). In these expressions (T, T
z
) are always associated with
the heavier nucleus.
Although S(t
z
) are always the basic quantities measured, for nuclei not too far
from N = Z it is often convenient to calculate S(T) in isospin formalism and then
use Eq. (20.44) to convert to S(t
z
). In addition, some tables of experimental values
and compilations [7] are given in terms of S(T) by dividing the experimental S(t
z
)
20 ONE-PARTICLE TRANSFER 225
value by C
2
. With good isospin there is an isospin selection rule given by the triangle
condition (T
f
,
1
2
, T
i
).
For an initial nucleus with N = Z, S(t
z
) = S(t
z
) by mirror symmetry. If we
add a nucleon to (T = 0, T
z
= 0) then we must go to T
f
= 1/2 with C
2
= 1. If we
remove a nucleon from (T = 0, T
z
= 0) then we go to T
f
= 1/2 with C
2
=
1
2
. In a
neutron-rich nucleus, proton removal involves only one isospin value in the nal state:
(T, T
z
) proton (T +
1
2
, T
z
+
1
2
) = (T
>
, T
z
+
1
2
), (20.45)
(transitions to the nal state with higher isospin are forbidden by the triangle condi-
tion). Neutron removal can go to two isospin values:
(T, T
z
) neutron (T +
1
2
, T
z

1
2
) = (T
>
, T
z

1
2
), (20.46)
that are the isobaric analogues of the (T +
1
2
, T
z
+
1
2
) states, and
(T, T
z
) neutron (T
1
2
, T
z

1
2
) = (T
<
, T
z

1
2
). (20.47)
The T
>
and T
<
are referred to as the T-upper and T-lower states, respectively.
The C
2
values for these three types of transitions are:
[< T, T
z
[ T +
1
2
, T
z
+
1
2
, t,
1
2
>[
2
=
_
2T + 1
2T + 2
_
, (20.48)
[< T, T
z
[ T +
1
2
, T
z

1
2
, t, +
1
2
>[
2
=
_
1
2T + 2
_
, (20.49)
and
[< T, T
z
[ T
1
2
, T
z

1
2
, t, +
1
2
>[
2
= 1. (20.50)
The explicit relations between S(t
z
) and S(T) are:
S(proton) =
_
2T + 1
2T + 2
_
S(T
>
), (20.51)
S(>, neutron) =
_
1
2T + 2
_
S(T
>
) =
_
1
2T + 1
_
S(proton), (20.52)
and
S(<, neutron) = S(T
<
). (20.53)
Thus, in neutron-rich nucleus one can measure proton spectroscopic factors, by
neutron removal to the T
>
states. Although these T
>
states lie at a high excitation
energy and are a often unbound to proton decay, their widths are small because
20 ONE-PARTICLE TRANSFER 226
their nucleonic decay is isospin forbidden. These type of transitions have only been
observed in nuclei not too far from N = Z since where the (2T +1) factor is not too
large.
French-Macfarlane sum-rules [8] are obtained from these results. For proton
removal to T
>
= T
i
+
1
2
(T
z
= T
i
+
1
2
) states:

f
S(proton) =< n
k,proton
>
i
, (20.54)
for neutron removal to the isobaric analogues of T
>
(T
z
= T
i

1
2
) states:

f
S(>, neutron) =
< n
k,proton
>
i
2T + 1
. (20.55)
For the total neutron removal:

f
S(neutron) = S(<, neutron) +S(>, neutron) =< n
k,neutron
>
i
, (20.56)
and thus for neutron removal to the T
<
(T
z
= T
i

1
2
) states:

f
S(<, neutron) =< n
k,neutron
>
i

< n
k,proton
>
i
2T + 1
. (20.57)
20.3.3 Simple situations
If we make a simple model for the initial and nal states the values for the spectro-
scopic factors are sometimes easy to calculate. The results given in Sec. 20 for the
special values of CFP in simple situations can be used to obtain the spectroscopic
factors for the corresponding cases.
For example,
16
O might be assumed to be the closed-shell conguration (0s
1/2
)
4
(0p
3/2
)
8
(0p
1/2
)
4
. One nucleon removal to A = 15 would then go to only three states
each of which has a denite sum: 1/2

with S(0p
1/2
) = 4, 3/2

with S(0p
3/2
) = 8
and 1/2
+
with S(0s
1/2
) = 4. One proton or neutron removal for T
i
= 0 going to
T
f
= 1/2 has C
2
= 1/2 and the sum rules are 2, 4 and 2, respectively, the same
result one would obtain from the proton and neutron occupancies. (Center-of-mass
corrections for these quantities will be discussed below.)
The spectroscopic factor for adding a 0d
5/2
neutron to the closed-shell congu-
ration for
16
O to make
17
O is S
0d
5/2
= 1 and the stripping sum rule of Eq. (20.35)
(J
f
= 5/2 and J
i
= 0) gives 6 which is the number of 0d
5/2
neutron holes outside of
16
O.
20 ONE-PARTICLE TRANSFER 227
20.3.4 Center-of-mass corrections
The center-of-mass (CM) correction to spectroscopic factors is closely associated with
the problem of spurious states. Lets consider again the
16
O to A = 15 transition
with a closed-shell conguration (0s
1/2
)
4
(0p
3/2
)
8
(0p
1/2
)
4
for
16
O. We take a single-
particle basis that corresponds to the motion of the nucleons around a xed center.
In actuality, we should take them with respect to the CM of of the A 1 nucleons.
However, with harmonic-oscillator wave functions, the CM of the nucleus is in a 0s
state as long as the nucleons ll the lowest possible oscillator states. There are eects
due to CM motion, but they are the simple ones associated with the CM being on its
0s state. We will refer to these as non-spurious states. All oscillator states of the
form (0s
1/2
)
4
(0p
3/2
)
n
(0p
1/2
)
m
are non-spurious and in particular the A = 15 the 1/2

and 3/2

states formed in this way are non-spurious.

Starting from a non-spurious state we can make 1 h excited states either by an
intrinsic 1 h excitation leaving the CM in the 0s state, or by the 1 h excitation of the
CM state from 0s to 1p. These latter are spurious states in the nuclear shell model.
A method for generating spurious states it to construct a two-body hamiltonian cor-
responding to the oscillator CM motion and then diagonalize this in the shell-model
basis. If the coecient in front of the CM hamiltonian is large enough, the spurious
states will be pushed to a high excitation energy and are prevented from mixing into
the low-lying (non-spurious) states.
The lowest 1 h excitation for A = 15, 1/2
+
corresponds to a mixture of (0s)
3
(0p)
8
and (0s)
4
(0p)
7
(1s0d)
1
congurations (e.g. the notation 0p stands for all possible
arrangements involving the 0p
1/2
and 0p
3/2
). There are a total of 18 basis states and
diagonalization of the CM hamiltonian results in two spurious states that are linear
combinations of these basis states.
We can calculate the summed spectroscopic strength from the closed-shell con-
guration of
16
O to the non-spurious A = 15 states. The result is S = 12 for the 0p
shell and S = 3.2 for the 0s shell - the remaining S = 0.8 for the 0s shell is in the two
spurious states. However, the sum rule of S = 16 still applies to the wave-functions
the intrinsic frame, and in order to recover this result we must make the following
correction to the 0p spectroscopic factors [9]:
S
0p

A
(A1)
S
0p
. (20.58)
where A = 16 in this example. This correction is the largest at the lower end of the
0p shell, e.g. (7/6) for the 0p-shell A=7 to A=6 spectroscopic factors. The general
20 ONE-PARTICLE TRANSFER 228
derivation and result is given in [9]:
S
k

_
A
(A1)
_
2nr+
S
k
. (20.59)
For example, for the removal of the 0d
5/2
in
17
O to
16
O, the CM correction factor is
(17/16)
2
= 1.129, and for the removal of the 0f
7/2
in
48
Ca to
47
Ca the CM correction
factor is (48/47)
3
= 1.065. [The rst Variational Monte Carlo (VMC) calculations [10]
for the
16
O(e,e

p) reaction indicated that the CM correction had the eect of reducing

the 0p spectroscopic factor by about 12 percent, however subsequently an error was
found [11] and the new VMC result is consistent with the 7 percent enhancement
expected from from the CM correction.]
20.3.5 Computation of shell-model spectroscopic factors
In simple situations, the spectroscopic factors can be obtained from the sum rules.
In particular, pickup from a closed shell conguration has a value of 2(2j + 1) for
nucleons or (2j +1) for protons or neutrons. Pickup from a state which is represented
by single-particle outside of a closed shell is unity. Often the j
n
conguration allows
only a single state with a given J value and the spectroscopic factor in this situation
can be calculated with the use of tabulated coecients of fractional parentage. For
complicated situation one must use a shell-model computer code. For example, in
Oxbash [12] one rst calculates the wave functions and then takes the one-particle
overlaps. The outputs are in les labeled *.LSA for the spectroscopic amplitudes and
*.LSF for the spectroscopic factors.
20 ONE-PARTICLE TRANSFER 229
20.4 Overlap functions
20.4.1 Denition and properties
Spectroscopic factors are related to the expansion of the wave function for a specic
state
A
i
in the initial nucleus with A nucleons in terms of a summation over the
complete set of states
A1
f
in the nal nucleus with A1 nucleons:

A
i
=

f,,j

i,f,,j
(r )
A1
f
, (20.60)
In the reaction for the removal of particles from state
i
to a specic state
f
one
requires the overlap function :
<
A1
f
[
A
i
>=

,j

i,f,,j
(r ), (20.61)
where an explicit summation over all possible and j values of the single-particle
overlap function is made. The spectroscopic amplitude A is dened by the normal-
ization of the overlap function:
A
i,f,,j
=
_

i,f,,j
(r ) d, (20.62)
and the related spectroscopic factor is:
S
i,f,,j
=[ A
i,f,,j
[
2
. (20.63)
The explicit dependence on and j is kept because the angular distributions in a
particular reaction depend upon and j, and thus is it possible to separate each of
these components in a given transition. Often the dependence on j is small in which
case one measures the sum over the two possible j values for a given :
S
i,f,
=

j
S
i,f,,j
. (20.64)
For example, for a J
i
J
f
=
3
2
+
2
+
transition both 0d
5/2
and 0d
3/2
could
contribute and one will measure:
S
=2
= S
0d
5/2
+S
0d
3/2
The radial size of the spectroscopic amplitude can be characterized in terms of
its mean-square radius:
< r
2
>
A
=
_
(r ) r
2
d
_
(r ) d
. (20.65)
20 ONE-PARTICLE TRANSFER 230
One can expand the overlap function in terms of a complete set of single-particle wave
functions:

i,f,,j
(r ) =

nr
B
i,f,k

k
(r ) (20.66)
where the are normalized to unity and k stands for (n
r
, , j). The single-particle
states are given by:

k
(r ) =
R
k
(r)
r
[Y
()
( r)
s
]
j
, (20.67)
where
s
is the intrinsic-spin wave function. For the single-particle states
k
, one
could take a basis of states generated from a mean-eld potential. In practical terms
one often approximates the mean-eld with harmonic-oscillator or Woods-Saxon po-
tentials. With the general expansion:
S
i,f,l,j
=

nr
[ B
i,f,k
[
2
. (20.68)
One often assumes that sum in Eq. (20.66) is restricted to only one radial quantum
number n
r
, in which case B = A, and S =[ A [
2
=[ B [
2
.
The possible and j values are restricted by the total angular momenta J and
parities of the initial and nal states. In general one must use a complete set of single-
particle states in the overlap function. But the quantum numbers associated with j
are restricted by the spin and parities of the initial and nal states. In particular
parity is conserved, and if J

i
= 0
+
then j = J
f
and
f
= (1)

.
The overlap function is exact to the extent that the many-body calculation is
exact and the sum in Eq. (20.66) extends over all (n
r
) values. However, simplied
nuclear models usually lead to the use of only a few (n
r
) values related to orbits near
the fermi surface. For example, a Hartree-Fock calculation for
17
O may be based
upon a closed-shell conguration for
16
O plus a neutron in the 0d
5/2
orbital. Then
the overlap function for the
17
O to
16
O reaction has A
0d
5/2
= 1 and is given by
(r ) =
0d
5/2
(r ) (the Hartree-Fock single-particle wave function).
In order to meaningfully use spectroscopic factors we require that the reaction
theory reproduce the observed dependence on projectile, energy and angle. The
spectroscopic factor is a single number whose value should not depend on the reaction.
When this is not the case, the reaction model or the model assumed for the overlap
function must be questioned. A theoretical review is given by Bang et al. [13].
20.4.2 Asymptotic properties
We can consider the overlap in Eq. (20.62) as a function of r. At large r where the
nuclear interactions are negligible this overlap must be governed by the kinematical
20 ONE-PARTICLE TRANSFER 231
asymptotics appropriate for the energy dierence E = E
f
E
i
(E = BE where
BE is the binding energy). For example for
17
O(5/2
+
) to
16
O(0
+
), E = 4.14 MeV
and there is an exponential decay appropriate for a d wave bound by 4.14 MeV.
In some reactions such as (
3
He,d) and (d,
3
He), the optical potentials are such
that only the part of the overlap function near the nuclear surface is important. In
the extreme case (e.g. at incident energies far below the Coulomb barrier) it is only
the asymptotic part of the overlap function which is important. The assumptions
about the potential only inuence the overall normalization, N
,j
, of the asymptotic
wave function:

i,f,,j
(r ) N
,j
f

(r), (20.69)
where f
l
(r) is an asymptotic form independent of the strong potential that depends
upon due to the the centrifugal barrier. It is usually taken as the Whittaker function
[14]. In the single-particle model the aymptotic behaviour is that of the single-particle
radial wavefunction
R
k
(r)
r
N
k
f

(r), (20.70)
The radial part of the overlap integral for large r is:
<
A1
f
[
A
i
>

,j
A
i,f,,j
N
,j
f

(r). (20.71)
In this case the nuclear structure part of the reaction cross section for a given (, j)
depends only on the square of the asymptotic normalization A
i,f,,j
N
,j
.
20.4.3 The well-depth prescription
In some cases the nuclear model may explicitly contain the correct asymptotic behav-
ior, such as a Hartree-Fock model for
17
O in which the last neutron is bound by 4.14
MeV. But usually many-body nuclear models are developed which do not explicitly
contain the correct asymptotic form. For example, a typical shell-model calculation
for
18
O consists of calculating the sd-shell conguration mixing and total energy by
evaluating the G matrix elements of a two-body residual interaction in an harmonic-
oscillator basis. The implicit radial wave functions used are the same for both
17
O
and
18
O and have the incorrect (oscillator) asymptotics for both the
17
O to
16
O and
the
18
O to
17
O overlaps.
The general method for generating an overlap function with the correct asymp-
totic form is to start with a realistic mean-eld potential for the single-particle wave
functions and then to modify this potential such that the single-particle () energy
is equal to the actual energy dierence (E). One way to do this is to multiply the
mean-eld potential by a constant such that the = E. This is the well-depth
20 ONE-PARTICLE TRANSFER 232
prescription. Historically, the well-depth prescription was easy to apply numerically
and has become the default method in many codes and analyses. For example, for
the
18
O(gs) to
17
O(5/2
+
gs) reaction one would increase the potential depth in order
to bind the 0d
5/2
orbital by the experimental energy dierence E = 12.19 MeV.
Likewise for the
18
O(gs) to
17
O(1/2
+
0.87 MeV excited state) reaction one would
increase the potential for the 1s
1/2
orbital in order to give the experimental energy
dierence of E = 13.06 MeV.
In cases where the single-particle wave function generated by the well-depth pre-
scription is close to that of a realistic mean-eld (e.g. the multiplying factor is not
too dierence from unity) this method is a reasonable procedure. This usually ap-
plies to cases where the spectroscopic factor is near the sum-rule limit. But when
the spectroscopic factor is small, this prescription may be questioned. For example,
in the (0d
5/2
)
3
conguration there are states with J =3/2, 5/2 and 9/2. The overlap
function between (0d
5/2
)
2
, J = 0 and (0d
5/2
)
3
, J = 3/2 must vanish (S = 0) since the
J = 3/2 does not match the j = 5/2 of the transferred particle. A nonvanishing spec-
troscopic factor would result from a small (0d
5/2
)
2
(0d
3/2
) admixture into the (0d
5/2
)
3
wave functions. The single-particle potential needed to give the correct asymptotics
for this small component is quite dierent from the mean-eld potential. Thus rather
than multiply the mean-eld potential it may be more appropriate to add a surface
peaked term to the mean-eld potential as discussed in the next section.
A common way to implement the well-depth prescription is to generate an over-
lap function from the single-particle wave functions obtained from a Woods-Saxon
potential. For the well-depth prescription, one would adjust the strength of the cen-
tral potential to obtain a single-particle wave function
j
which has a single-particle
energy equal to the energy dierence between the initial and nal states under con-
sideration. This wave function goes into the reaction theory to calculate a cross
section, and then from comparison to experiment one deduces the spectroscopic fac-
tor S. However, one should keep in mind that the value for S depends upon the
assumptions made about the reaction models and about the overlap function.
20.4.4 Beyond the well-depth prescription
One would like to carry out an experiment which is sensitive to the shape as well as
the magnitude of the overlap function. But the nuclear reactions are generally only
sensitive to the surface or asymptotic part, and thus the assumption about the xed
Woods-Saxon shape cannot be easily tested. As mentioned above the prescription
based on the adjustment of the depth of a central potential is appropriate only when
the transition under consideration is actually close to single-particle in nature. When
the single-particle strength is fragmented due to residual interactions (beyond the
20 ONE-PARTICLE TRANSFER 233
mean-eld) one should consider additions to the central potential which are related
to the residual interaction.
Pinkston and Satchler [15] discussed the situation for the (p,d) reactions in the
0f
7/2
shell region, for example
48
Ti(p,d) leading to the lowest T
<
= 3/2 and T
>
= 5/2
states in
47
Ti which dier in separation energy by 7.3 MeV. This isospin splitting of
the 0f
7/2
orbit can be related to an isovector residual interaction which peaks near
the nuclear surface and might be modeled on the derivative of the central Woods-
Saxon potential. Thus the shape as well as the strength of the central potential must
be considered. In terms of the original Woods-Saxon shape, the surface potential
has the eect of eectively increasing the radius r
o
for the T
>
state (in contrast
to the well-depth prescription where the well-depth V
o
is increased). This results
in a relatively larger radius for the overlap function, a larger DWBA cross section,
and hence a smaller spectroscopic factor. Similar considerations regarding the shape
of the potential are related to the collective model. [15] More detailed models of
these residual-interaction eects have discussed some of which involve solving a set
of coupled equations for the bound state [16], [17], [18], [19], [20].
The examples above apply to removal from a partially lled orbit. In this situa-
tion the energy dierence E for the transition in question is usually larger than the
single-particle energy. The well-depth prescription thus gives an rms radius for the
overlap function which is smaller than the rms radius of the single-particle state. The
surface eect of the residual interaction correction leads to an overlap function which
has a relatively larger rms radius (back in the direction the single-particle rms radius).
In the 0f
7/2
shell it was found that a practical prescription is to x the Woods-Saxon
well depth and radius to give the correct energy dierence and a state-independent
rms radius for the overlap function [21]. The spectroscopic factors obtained with the
residual interaction corrections are typically up to a factor of two smaller than than
those obtained with the well-depth prescription.
The other extreme is stripping to a nearly lled orbit. For example
40
Ca(d,p)
41
Ca
leading to the 3/2
+
state (at 2.04 MeV). The spectroscopic factor would be zero in
the limit where the 0d
3/2
orbit is lled in
40
Ca. The energy dierence is E = 10.4
MeV as compared to the single-particle energy of about = 15.6 MeV (e.g. the
40
Ca -
39
Ca binding energy dierence). The well-depth prescription thus leads to an
rms radius for the overlap function (E = 10.4 MeV) which is larger than that
of the single-particle state (E = 15.6 MeV). The surface property of the residual
interaction leads to an overlap function which has a relatively smaller rms radius [22]
(again back in the direction of the single-particle rms radius).
REFERENCES 234
References
[1] A. de Shalit and I. Talmi, Nuclear Shell Theory, (Academic Press, 1963).
[2] L. B. Hubbard, Comput. Phys. Commun. 1, 225 (1970); D. Zwart, Comput. Phys.
Commun. 38, 656 (1985); L. D. Skouras and S. Kossionides, Comput. Phys. Com-
mun. 39, 197 (1986).
[3] A. Novoselsky, J. Katriel and R. Gilmore, Jour. Math. Phys. 29, 1368 (1988).
[4] X. Ji and M. Vallieres, Phys. Rev. C 35, 1583 (1987).
[5] R. D. Lawson, Theory of the Nuclear Shell Model, (Clarendon Press, 1980).
[6] B. F. Bayman and A. Lande, Nucl. Phys. 77, 1 (1966); I. S. Towner and J. C.
Hardy, Nucl. Data Tables A6, 153 (1969); L. B. Hubbard, Nucl. Data Tables A9,
85 (1971); S. Shlomo, Nucl. Phys. A184, 545 (1972).
[7] Spectroscopic Factors for Single Nucleon Transfer for A = 21-44, P. M. Endt,
Atomic and Nuclear Data Tables 19, 23 (1977).
[8] J. B. French and M. H. Macfarlane, Nucl. Phys. 26, 168 (1961).
[9] A. E. L. Dieperink and T. de Forest, Phys. Rev. C 10, 543 (1974); M. Ichimura,
A. Arima, E. C. Halbert and T. Terasawa, Nucl. Phys. A204, 225 (1973).
[10] M. Radici, S. Bo, S. C. Pieper and V. R. Pandharipande, Phys. Rev. C 50,
3010 (1994).
[11] D. Van Neck, M. Waroquier, A. E. L. Dieperink, S. C. Pieper and V. R. Pand-
haripande, Phys. Rev. C 57, 2308 (1998).
[12] W. D. M. Rae, A. Etchegoyen and B. A. Brown, OXBASH, the Oxford-Buenos
Aries-MSU Shell-Model Code, Michigan State University Laboratory Report no.
524.
[13] J. M. Bang, F. G. Gareev, W. T. Pinkston and J. S. Vaagen, Phys. Rep. 125,
253 (1985).
[14] P. G. Hansen and J. A. Tostevin, Ann. Rev. Nucl. Part. Sci. 53, 219 (2003).
[15] W. T. Pinkston and G. R. Satchler, Nucl. Phys. 72, 641 (1965).
[16] E. Rost, Phys. Rev. 154, 994 (1967).
[17] T. Tamura, Phys. Rev. Lett. 19, 321 (1967); Phys. Rev. 165, 1123 (1968).
REFERENCES 235
[18] A. Prakash, Phys. Rev. Lett. 20, 864 (1968); A. Prakash and N. Austern, Ann.
of Phys. 51, 418 (1969).
[19] J. L. Hutton, Nucl. Phys. A210, 567 (1968).
[20] R. J. Philpott, W. T. Pinkston and G. R. Satchler, Nucl. Phys. A119, 241
(1968).
[21] A. Moalem, J. F. A. van Heinen and E. Kashy, Nucl. Phys. A307, 277 (1978).
[22] W. T. Pinkston, R. I. Philpott and G. R. Satchler, Nucl. Phys. A125, 176 (1969).
21 EXPERIMENTS RELATED TO SPECTROSCOPIC FACTORS 236
21 Experiments related to spectroscopic factors
21.1 Experimental results for specic nuclei
There have been many experimental papers related to the extraction of spectroscopic
factors. The details obtained have been and continue to be tremendously important
in guiding the our understanding of nuclear structure. In this section we will give a
few examples of the wide range of data available.
We will mainly look at examples where nucleons are removed from the nucleus.
Reactions such as (d,
3
He) or (p,d) are called pickup reactions since a nucleon is
picked up by the light projectile and removed from the target nucleus. Reactions such
as (e,e

p) are called knockout reactions since the electron hits the proton and knocks
it out of the nucleus. The collision of radioactive beams with light and heavy targets
can also knockout protons or neutrons from the radioactive beam. In addition, the
spontaneous decay of a proton or neutron from an unbound state removes a nucleon
from the nucleus.
We will also look at examples for spectroscopic factors when nucleons are added
to the nucleus. Reactions such as (
3
He,d) or (d,p) are often called stripping re-
actions since since a nucleon is stripped from the light projectile and added to the
target nucleus. Sometimes the spectroscopic factors for the a given pair of states
can be studied in both stripping and pickup reactions. For consistency the same
spectroscopic factor should be obtained from both reactions.
21.1.1
51
V
50
Ti
One of the best examples for proton removal spectroscopic factors in a one-orbit model
is given by the reaction
51
V
50
Ti. In this mass region the 0f
7/2
orbit is rather well
isolated. In this model the structure of
51
V is a closed shell, (0f
7/2
)
8
, for neutrons and
(0f
7/2
)
3
for protons. For this conguration there is only one J = 7/2 state, and the
spectroscopic factors can be obtained immediately from the Table of CFP together
with the relationship:
S = n [< j
n

i
J
i
[j
n1

f
J
f
>[
2
. (21.1)
This gives the values in the rst column of Table [22.1] whose sum is three - the
total number of protons in the 0f
7/2
. This particular structure leads to nontrivial
values for the spectroscopic factors which are in excellent agreement with the original
51
V(d,
3
He) (52 MeV) experiment (Table 22.1.a) [1]. An analysis of
51
V(d,
3
He) data
at several energies [2] shows that absolute spectroscopic factors are sensitive to the
21 EXPERIMENTS RELATED TO SPECTROSCOPIC FACTORS 237
Table 1: Proton spectroscopic factors for
51
V
50
Ti
J
f
(f
7/2
)
3
model (d,
3
He) (d,
3
He) (e,e

p) (d,
3
He)
(a) (b) (c) (c)
0 0.75 0.73 0.40(1) 0.37(3) 0.30
2 0.42 0.39 0.21(1) 0.16(2) 0.15
4 0.75 0.64 0.42(1) 0.33(3) 0.26
6 1.08 1.05 0.66(1) 0.49(4) 0.39
sum 3.00 2.81 1.69 1.35(7) 1.10
a) 1967 analysis of Hinterberger et al. [1] for 52 MeV data.
b) 1976 analysis of Craig et al. [2] for 80 MeV data.
c) 2001 analysis of Kramer et al. [3].
reaction model, but the relative spectroscopic factors for the various J
f
values are
rather insensitive. A more recent nite-range DWBA analysis of 80 MeV data results
in spectroscopic factors which are about 0.6 of those expected from the 0f
7/2
shell
model (Table 22.1.b). This reduction from unity is consistent with the more recent
(e,e

p) results (Table 22.1c) which will be discussed in more detail below.

21.1.2 (
3
He,d) in the sd shell
The sd-shell provides one of the best examples of how shell-model conguration mixing
can be applied consistently over a wide mass range. One can nd an eective two-
body interaction [4] which is able to reproduce the energies of about 1000 levels for
nuclei over the mass range A=17-39 to within an rms deviation of a few hundred keV
[5]. The electromagnetic and beta decay properties of these levels are also rather well
understood in this sd-shell model [6].
Endt has made a compilation of spectroscopic factors for the sd-shell [7]. Ver-
notte et al. [8] has made an analysis of spectroscopic factors in the sd-shell with a
unied analysis of (
3
He,d) (nucleon removal) data obtained with a
3
He energy of 25
MeV. The data themselves are rather precise and the statistical error is relatively
small. Most of the uncertainty associated with the extracted spectroscopic factors is
related to the reaction-model interpretation. The data provides information for 39
transitions between specic initial and nal states over the mass region A=17-40. The
analysis depends upon a zero-range DWBA analysis for which one must provide op-
tical potentials for the
3
He and d. The overlap function is based upon the well-depth
method.
The initial analysis was carried out with Woods-Saxon shape parameters r
o
=
21 EXPERIMENTS RELATED TO SPECTROSCOPIC FACTORS 238
S

(
e
x
p
)
0
1
2
3
4
5
normal spin-orbit
S

(
e
x
p
)

/

S

(
t
h
)
S (th)
0.5
1.0
1.5
0 1 2 3 4
reduced-radius spin-orbit
S (th)
0 1 2 3 4 5
Figure 1: Theoretical vs experimental spectroscopic factors for 39 transitions
in the sd shell. The parameters of the Woods-Saxon potential are r
o
= 1.25
fm and a = 0.65 fm. The results on the left-hand side were obtained with the
normal spin-orbit potential r
so
= r
o
and a
so
= a
o
. The results on the right-hand
side were obtained with the reduced-radius spin-orbit potential r
so
= 1.00 fm
and a
so
= 0.52 fm. The points are for 0d
5/2
(lled circles), 0d
3/2
(crosses) and
1s
1/2
(squares).
1.25 fm and a = 0.65 fm. Initially the same shape parameters were used for the spin-
orbit potential. The experimental and theoretical spectroscopic factors are shown
on the left-hand side of Fig. (21.1) (In [8] the results for the nucleonic spectroscopic
factors are given, but here I have converted these to proton spectroscopic factors by
dividing by the C
2
factor.) In the top one observes good agreement in terms of the
variation from large to small values. The largest value for 0d
5/2
(lled circle) corre-
sponds to the
27
Al (5/2
+
gs) to
28
Si(0
+
gs) transition. In the simplest (0d
5/2
)
n
model
the theoretical value would be six, but this gets reduced to 3.61 due to conguration
mixing in the full sd shell. The largest value for 0d
3/2
(cross) is for the
39
K(3/2
+
)
40
Ca(0
+
) transition. In the simple (0d
3/2
)
n
model as well as in the full sd-shell model
it has a theoretical value of four.
In the bottom of Fig. (21.1) the ratios are plotted vs the theoretical value. The
ratios for 0d
3/2
(crosses) are systematically higher than those for 0d
5/2
(lled circles).
A way to correct this problem was noted by Vernotte et al., and I will discuss this
in some detail to illustrate the problems which can arise in the model dependence
21 EXPERIMENTS RELATED TO SPECTROSCOPIC FACTORS 239
V
(
r
)

(
M
e
V
)
r (fm)
-60
-40
-20
0 1 2 3 4 5 6
V
(
r
)

(
M
e
V
)
-60
-40
-20
WS
P
(
r
)
10
-2
10
-1
10
0
r (fm)
0 1 2 3 4 5 6 7
SKX
Figure 2: Bound state potentials and radial wave functions for the Woods-
Saxon (left-hand side) and SKX Skyrme Hartree Fock (right-hand side). The
bottom panel shows the cental (line) and spin-orbit (dots) potentials for A = 28.
The middle panel shows the eective total potentials for 0d
5/2
(line) and 0d
3/2
(dashes) required to bind the orbitals by 3 MeV. The top panel shows the
radial density probability for 0d
5/2
(line) and 0d
3/2
(dashes) resulting from the
eective potentials.
in the analysis. Namely, the ratios can be improved if the value of r
so
used in the
Woods-Saxon potential is reduced from 1.25 to 1.00 fm. To better understand how
this works, the central and spin-orbit Woods-Saxon potentials are plotted on the
bottom left-hand of Fig. (21.2). The central potential is the one required to bind a 0d
orbit by 3 MeV (no spin-orbit). The spin-orbit potential plotted represents the total
change in potential between the 0d
5/2
and 0d
3/2
orbits which is required to reproduce
the spin-orbit energy dierence of about 6 MeV. The well-depth prescription is now
used to generate a wave function for the 0d
5/2
and 0d
3/2
orbits with a xed binding
energy of 3 MeV (an arbitrary but typical value) by adjusting the strength V
o
of the
central potential. The eective total potential (the sum of the adjusted central and
spin-orbit terms and excluding Coulomb) for the 0d
5/2
and 0d
3/2
are shown in the
middle part of Fig. (21.2). The square of the radial wave functions are shown at the
top. The eective potential for 0d
5/2
has an eectively larger radius than that for
0d
3/2
and the resulting 0d
5/2
radial density is larger than that 0d
3/2
.
21 EXPERIMENTS RELATED TO SPECTROSCOPIC FACTORS 240
The (
3
He,d) cross section is mainly sensitive to the overlap function near the
surface, and essentially only the asymptotic normalization is important. On the left-
hand side of Fig. (21.2) this asymptotic normalization is 25% larger for 0d
5/2
compared
to 0d
3/2
which arises from the spin-orbit potential. Thus for a given separation energy,
the spin-orbit potential causes the extracted spectroscopic factor for 0d
5/2
to be 25%
less than that for 0d
3/2
(in the direction of the deviation noted in Fig. (21.1)).
Vernotte et al. were able to improve this situation by reducing the range of the
spin-orbit potential to from 1.25 fm to 1.00 fm (the ratio of the diuseness to the
range of the spin-orbit potential was xed at 0.52). This results in eective potentials
for 0d
5/2
and 0d
3/2
that are nearly the same at the nuclear surface. The resulting
spectroscopic factors shown in Fig. (21.1) have a more consistent average ratio for
0d
5/2
and 0d
3/2
.
This analysis brings into focus the model dependence of extracting spectroscopic
factors. One could use other models for the bound states. I show, as an example,
the potentials which one obtains from the SKX Skyrme Hartree-Fock model [9] at
the bottom of Fig. (21.2), together with the eective potentials obtained by adjusting
the central part of the HF potential to obtain binding energies of 3 MeV for 0d
5/2
and 0d
3/2
. It turns out that the spin-orbit potential from the Skyrme interaction is
peaked at a smaller radius compared to the conventional Woods-Saxon model and
that this leads to very similar eective potentials near the surface. In this sense the
Skyrme HF justies the empirical result obtained by Vernotte et al. The reason why
the Skyrme spin-orbit potential peaks at a smaller radius is that it is obtained from
the derivative of the matter density which has a smaller eective radius than the
potential.
At small radii there are also large dierences between the Woods-Saxon and
Skyrme potentials, (21.2), but this does not have much inuence on the analysis of
spectroscopic factors. This small radius behavior has an inuence on the interior
density which can be determined from the charge density measured in elastic elec-
tron scattering. Some charge densities extracted from experiments are compared with
Hartree-Fock calculations [11] based upon the SKX [9] and SKM* [10] Skyrme inter-
actions are shown in Fig. [12.12]. For the sd-shell the occupancies from the sd-shell
model calculations were used. The good agreement between experiment and theory
together with the relatively strong nucleus to nucleus variation tends to conrm the
Skyrme HF approach.
The possible uncertainty in the spin-orbit potential is part of a larger question as
to how accurately the central potential and the resulting asymptotic normalizations
can be calculated. In terms of the (
3
He,d) reaction Vernotte et al. pointed out the
strong dependence of the extracted spectroscopic factors on the range of the central
potential. They noted that an increase of 0.01 fm in r
o
decreased the extracted
21 EXPERIMENTS RELATED TO SPECTROSCOPIC FACTORS 241
c
h
a
r
g
e

d
e
n
s
i
t
y
r (fm)
0.0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.1
0.11
0.12
0.13
0.14
0.15
0.16
0.17
0.18
0.19
0 2 4 6 8 10 12 14
28
Si
32
S
40
Ca
48
Ca
50
Ti
52
Cr
54
Fe
58
Ni
88
Sr
90
Zr
92
Mo
204
Hg
208
Pb
Figure 3: Charge densities for a range of nuclei. The lled circles are the
experimental densities obtained from a Fourier-Bessel analysis of the electron
scattering form factor data. The Hartree-Fock results are shown for the SKX
[9] and SKM* [10] Skyrme interactions.
spectroscopic factors by 3-4% (see Fig. 8 in [8]). Since the reaction takes place at the
nuclear surface and depends mainly on the asymptotic normalization, there is very
little dependence of the shape of the angular distributions on r
o
.
There are also issues related to the reaction model. Vernotte et al. used the
convenient zero-range form of DWBA, but they noted that non-locality corrections
and nite-range correction would decrease the overall spectroscopic factors by about
25%. Thus to attempt to summarize the sd-shell data in terms of absolute spectro-
scopic factors I show in Fig. (21.4) the values obtained with the new reduced-radius
spin-orbit interaction and multiplied by 0.75 are plotted as the ratio of experiment
over theory vs the mass value.
21 EXPERIMENTS RELATED TO SPECTROSCOPIC FACTORS 242
R
a
t
i
o
A
0.0
0.5
1.0
1.5
16 18 20 22 24 26 28 30 32 34 36 38 40
Figure 4: Ratio of the experimental to theoretical spectroscopic factors for
(A 1)(
3
He,d)A in the sd shell. The experimental values obtained from the
reduced-radius spin-orbit potential model are multiplied by 0.75. (Center-of-
mass corrections are not included in the theory.) Filled circles are for 0d
5/2
crosses are for 0d
3/2
and squares are for 1s
1/2
.
21.1.3 (e,e

p)
The (e,e

p) reaction is perhaps the most direct way to measure the proton spectro-
scopic factor. It is largely determined by the well understood electromagnetic interac-
tion and the only hadronic reaction theory involved is for the nal-state interactions
of the protons as they leave the nucleus. [12], [3] Unlike the reactions such as the
(
3
He,d) and (d,
3
He) discussed above which depend mainly on the surface part of the
overlap function, the (e,e

p) reaction is sensitive to the overlap function at all radii.

The shape of the momentum proton distributions is sensitive to the radial shape of
the overlap function. Thus, one can adjust the well depth and the radius of the po-
tential to give the separation energy and the shape of the momentum distribution.
The analysis of many data have recently been carried out Kramer et al. [3] and I give
the some of the resulting proton spectroscopic factors in Table [22.2].
Table [22.2] also gives the spectroscopic factors obtained from the conventional
analysis of (d,
3
He) data based upon zero-range DWBA and a Woods-Saxon potential
with a radius of r
o
= 1.25 fm. The spectroscopic factors deduced from (e,e

p) are
systematically smaller than those from the old (d,
3
He) analysis by a factor of 0.6-0.7.
They are also about this much smaller than that expected from a shell-model or sum-
rule estimate. An explanation of this is that the old (d,
3
He) analysis is based upon a
simple (zero-range) reaction theory and a shape for the form factor which are rather
arbitrarily chosen to give the expected (sum-rule) result. Kramer et al., also carry out
a new analysis of the (d,
3
He) data based on a nite-range DWBA and with Woods-
Saxon wave functions for the overlap function with r
o
adjusted to give the correct
shape of the (e,e

p) momentum distributions. A non-locality correction [3] was also

21 EXPERIMENTS RELATED TO SPECTROSCOPIC FACTORS 243
Table 2: Proton spectroscopic factors for a range of nuclei.
Nuclei E
x
J
f
(e,e

p)(a) (e,e

p) (d,
3
He)(a) (d,
3
He)(a)
(MeV) NR(b) R(c) (old) (new)
12
C
11
B 0.00 3/2

1.72(11) 2.98 1.72

2.12 1/2

0.26(2) 0.69 0.27

5.02 3/2

0.20(2) 0.31 0.11

16
O
15
N 0.00 1/2

1.27(13) 1.46(e) 2.30 1.02

6.32 3/2

2.25(22) 2.8(e) 3.64 1.94

9.93 3/2

0.133(15)(d)
10.70 3/2

0.222(4)(d)
30
Si
29
Al 0.00 5/2
+
2.21(20)(g) 3.96(g)
31
P
30
Si 0.00 0
+
0.40(3) 0.62 0.36
40
Ca
39
K 0.00 3/2
+
2.58(19) 3.32(f) 3.70 2.30
2.52 1/2
+
1.03(7) 1.65 1.03
48
Ca
47
K 0.00 1/2
+
1.07(7) 1.55 0.96
0.36 3/2
+
2.26(16) 4.16 2.39
51
V
50
Ti 0.00 (0,2,4,6)
+
1.35(7) 2.81 1.10
208
Pb
207
Tl 0.00 1/2
+
0.98(9) 1.40(8)(h) 1.8 1.5
0.35 3/2
+
2.31(22) 2.92(16)(h) 3.8 2.2
1.35 11/2

6.85(68) 7.7 5.4

1.67 5/2
+
2.93(28) 3.5 3.1
3.47 7/2
+
2.06(20) 3.5 2.9
a) From Kramer et al. [3] except where noted.
b) Nonrelativistic analysis from Kramer et al. [3] except where noted.
c) Relativistic analyses.
d) From Leuschner et al. [13].
e) From Gao et al. [14].
d) From Wessling et al. [15].
g) From Jin et al. [16].
h) From Udias et al. [17].
21 EXPERIMENTS RELATED TO SPECTROSCOPIC FACTORS 244
applied to the Woods-Saxon wave functions. This adjustment of bound state wave
functions has the eect of increasing the rms radius of the orbitals, increasing the
asymptotic normalization, and hence reducing the spectroscopic factors. The total
eect of changing the reaction model and overlap function is to bring the spectroscopic
factors deduced from the new (d,
3
He) analysis into much better agreement with those
from (e,e

p) (see Table [22.2]).

The values of r
o
which are needed to t the (e,e

p) momentum distributions [3]

are typically in the range 1.20 to 1.30 fm - close to the average accepted value of 1.25
fm. But there are exceptional cases, such as the value of r
o
= 1.65 required for the
relatively weak
12
C to
11
B(1/2

, 2.1 MeV) transition.

For consistency we can compare the spectroscopic factors obtained from the nu-
cleon removal reactions in Table [22.2] to those from the (
3
He,d) nucleon addition
reaction of Vernotte [8]. The results based on the discussion in the previous sec-
tion with the reduced-radius spin-orbit potential and then multiplied by 0.75 are
C
2
S = 0.49 for
30
Si
31
P and C
2
S = 2.69 for
39
K
40
Ca. These are close to those
obtained in (e,e

p) but somewhat larger than those from the new (d,

3
He) analysis
given in Table 3.
Most of the (e,e

p) data in Table [22.2] is for the pickup of states below the fermi
surface. One of the few (e,e

p) data available for states above the fermi surface is

from the
209
Bi data of Branford et al. [18]. The data for
209
Bi(e,e

p)
208
Pb(gs) are
not very good, but from the most accurate data points in Fig. 2 of [18] one would
interpret this experiment as giving C
2
S = 1 for 0h
11/2
knockout. as expected in the
extreme single-particle shell model. Branford et al. also study the excited states in
208
Pb corresponding to knocking out the hole states given in Table [22.2] and leaving
208
Pb in multiplet of states corresponding to the one-particle one-hole congurations.
From a comparison to the
208
Pb(e,e

p)
207
Tl data, they conclude that the model of
209
Bi based on a 0h
9/2
proton coupled to an inert
208
Pb core has a high degree of
validity.
It is important that the spectroscopic factor extracted from the (e,e

p) data does
not depend on the electron energy and/or the reaction model which is used to un-
derstand the data. For the (e,e

p) data typical of those for Table [22.2], the electron

energy is about 500 MeV and the outgoing proton kinetic energy is about 100 MeV.
The relevant variable is the four-momentum squared, Q
2
, transferred by the virtual
photon to the knocked-out nucleon which is about 0.2 (GeV/c)
2
in this case. For
these low Q
2
a proton optical potential was used for the nal-state interactions. The
12
C(e,e

p) data has been taken and analyzed over a wide range of Q

2
. At high Q
2
values (0.8 and higher) one usually uses a Glauber model for the nal-state interac-
tions of the protons. At an intermediate Q
2
of about 0.6 both the optical model and
the Glauber model can be used. Lapikas et al. [19] found that the two methods of
21 EXPERIMENTS RELATED TO SPECTROSCOPIC FACTORS 245
analysis gave dierent spectroscopic factors. For sum of the three 0p shell states in
Table [22.2], the proton spectroscopic factors deduced from the optical-model analysis
is about 2.2 (consistent with Table [22.2]) but the Glauber model analysis gave 3.56
(much closer to the simple shell model result of 4). However, more recently Radici
et al. [20] have analyzed a similarly wide range of data for
16
O(e,e

p) and nd a
consistently small spectroscopic factor over the whole range of Q
2
values.
Data for
40
Ca(e,e

p) with 460 MeV electrons have been analyzed in the DWIA

(distorted-wave impulse approximation) bound state (overlap functions) from the
relativistic potential and also a relativistic model for the scattering state and the
current operators [16]. The spectroscopic factor of 3.32 obtained for the 3/2
+
state
is 25% larger than the nonrelativistic results of Lapikas given in Table [22.2]. A
relativistic DWIA analysis has also been carried out for
16
O(e,e

p) with 2.4 GeV

electrons [14]. The proton spectroscopic factors obtained are S(1/2

, gs) = 1.46 and

S(3/2

, 6.32MeV ) = 2.8. These are about 20% larger than those of Table [22.2].
There is also a relativistic DWIA analysis of the
208
Pb(e,e

p) data [17] which gives

a proton spectroscopic factors of 1.40(8) and 2.92(16) for the 1/2
+
and 3/2
+
states,
respectively, again considerably larger than the nonrelativistic-analysis of 0.98(9) and
2.31(22), respectively. In [17] the dierence between the nonrelativistic and relativistic
results in mainly attributed to the proton optical potential which produces relatively
more absorption in the relativistic model and hence gives rise to larger spectroscopic
factors.
21.1.4 Proton decay
The widths for proton or neutron decay are related the spectroscopic factors. The
usual method for calculating these decay widths is to multiply the single-particle
width for the decay of a nucleon in orbital j with the spectroscopic factor:
= (C
2
S)
sp
(21.2)
where
sp
is the single-particle width calculated from the scattering of a nucleon from
a central potential. Usually a Woods-Saxon potential is used and the well depth V
o
is adjusted to reproduce the experimental centroid of the state.
Millener [21] has analyzed the proton decay widths for several cases in the mass
region A = 11 16 and his results are summarized in Table [22.3]. He used a Woods-
Saxon potential with r
o
= 1.25 fm and a
o
= 0.60 fm. The calculated widths are
based upon spectroscopic factors obtained with the Millener-Kurath interaction [22]
(which are similar to those obtained with WBP [23]). There is good overall agreement
between experiment and theory.
21 EXPERIMENTS RELATED TO SPECTROSCOPIC FACTORS 246
Table 3: Observed and calculated [21] proton decay widths for 1 h states
which have dominantly 1s
1/2
or 0d
5/2
character. (Center-of-mass corrections
are not included in the theory.)
Nucleus J

th

exp
Dominant Component
11
N
1
2
+
1467 1440 81% s
1/2
(gs)
12
N 2

87 118(14) 70% s
1/2
(gs)
1

894 750(250) 76% s

1/2
(gs)
1

311 260(30) 77% s

1/2
(
1
2

)
13
N
1
2
+
32 31.7(8) 89% s
1/2
(gs)
3
2
+
84 115(5) 87% s
1/2
(2
+
)
5
2
+
2
7 11 74% s
1/2
(2
+
)
14
O 1

28.8 30(1) 76% s

1/2
(gs)
15
F
1
2
+
934 1000(200) 98% s
1/2
(gs)
16
F 0

21 40(20) 100% s
1/2
(gs)
1

78 < 40 97% s
1/2
(gs)
11
N
5
2
+
535 600(50) 67% d
5/2
(gs)
12
N 3

220 220(25) 88% d

5/2
(gs)
4

610 744(25) 82% d

5/2
(gs)
3

280 180(23) 70% d

5/2
(
1
2

)
13
N
9
2
+
258 280(30) 91% d
5/2
(2
+
)
7
2
+
3 9.0(5) 92% d
5/2
(2
+
)
5
2
+
45 47(7) 80% d
5/2
(gs)
14
O 3

13.4 16(2) 84% d

5/2
(gs)
2

38.2 41(2) 66% d

5/2
(gs)
15
F
5
2
+
222 240(30) 93% d
5/2
(gs)
16
N 2

3.5 40(30) 96% d

5/2
(gs)
3

12 < 15 96% d
5/2
(gs)
21 EXPERIMENTS RELATED TO SPECTROSCOPIC FACTORS 247
An increase of 0.01 fm in r
o
increases the calculated single-particle with by about
1%, and an increase of 0.01 fm in a
o
increases the calculated single-particle with by
about 2% [21]. Thus for example a change of a
o
from 0.60 to 0.65 would increase the
calculated width by about 10%.
21.1.5 Radioactive beams
It has recently become possible to carry out one-nucleon transfer reactions with ra-
dioactive beams. One way is to do the traditional reactions such as (p,d) in inverse
kinematics. A new way which has been developed is to knockout a loosely bound
nucleon in a reaction on a light target such as
9
Be. The spectroscopic factor for a
specic nal state is obtained by analyzing the momentum distribution of the (A1)
fragment in coincidence with gamma rays. A summary [24] of the spectroscopic fac-
tors obtained from the analysis of several experiments at the NSCL is shown in Fig.
(21.5). The spectroscopic factors are obtained with an overlap function calculated
with the traditional Woods-Saxon well-depth method.
21.2 Structure models for specic nuclei
The mean-eld model for nuclear structure corresponds to the extreme single-particle
value for the spectroscopic factor. The residual interaction results in correlations in
the nuclear wave function beyond the mean eld which will change the spectroscopic
factors. We can attempt to distinguish between long- and short-range correlation.
By long range correlation I mean the mixing of the congurations near the fermi
surface. This results in a spread of the orbital occupations and fractionation of the
spectroscopic strength into many states up to about 10 MeV in the spectrum. Short-
range interactions can scattering the nucleons into very high single-particle states and
spread the spectroscopic strength up to as high as 100 MeV in excitation. In this
section I will discuss the results in terms of these long- and short-range correlations.
Shell-model conguration mixing takes into account the residual interaction for
single-particle states near the fermi surface (the valence states). This type of con-
guration mixing is mainly related to the long-range correlations. The short-range
correlation enters implicitly in terms of a renormalization of the two-body interaction
(the G matrix). I will rst concentrate on the spectroscopic factors obtained from
shell-model calculations. The ratio of the experimental spectroscopic factors to the
extreme single-particle model will be denoted by R and the ratio to the congura-
tion mixed shell-model spectroscopic factors will be denoted by R
s
. The subscript
s means that its deviation from unity will be attributed to the eect of short-range
21 EXPERIMENTS RELATED TO SPECTROSCOPIC FACTORS 248
0 0.5 1 1.5
C
2
S(theo)
0
0.5
1
1.5
C
2
S
(
e
x
p
)
l=0
l=1
l=2
Figure 5: Comparison of experimental and calculated spectroscopic factors for
reactions at approximately 60 MeV/nucleon leading to specic nal levels in
the nuclei
25,26,27
Si [25].
10
Be [26],
11
Be [27],
13
B [28] and
14,15,16,18
C [29].
Circles, triangles and squares correspond to =0,1,2, respectively. The dashed
line corresponds to R = 1.
correlations. There is actually no clear boundary between what we can attribute to
long and short range correlations, but this will serve as a working denition. Also
as we will discuss, there may be other types of eects such as clustering left out of
the shell-model calculations which may contribute to R
s
. The notation R

s
will be
used when using (e,e

p) experimental data interpreted with a relativistic model the

reaction. A summary of R values from the results discussed below is given in Table
[22.4].
21.2.1
7
Li
The
7
Li(e,e

p) reaction has recently been studied by Lapikas et al. [30] The spectro-
scopic factors for nonrelativistic analysis of the
7
Li(e,e

p) data summed over the 0

+
and 2
+
states in
6
He is 0.58(5). The 0p sum-rule value is 1.00 and typical 0p shell
conguration mixing leads to about 0.95 in these lowest two states. Thus we would
21 EXPERIMENTS RELATED TO SPECTROSCOPIC FACTORS 249
Table 4: Ratio between experimental and shell-model spectroscopic factors
Nucleus J
f
R(a) R
s
(b) R

s
(c)
7
Li 3/2

0.52 0.52
12
C 1/2

,3/2

0.50 0.51 0.84

16
O 1/2

,3/2

0.60 0.66 0.81

30
Si 5/2
+
0.37 0.58
40
Ca 3/2
+
0.64 0.70 0.90
48
Ca 3/2
+
0.56 0.75
51
V (0,2,4,6)
+
0.45 0.50
208
Pb 3/2
+
0.58 0.65 0.82
a) Non-relativistic analysis of Kramer et al. [3] compared to the extreme single-
particle model.
b) Non-relativistic analysis of Kramer et al. [3] compared to valence shell-model
calculations.
c) Relativistic or Glauber model analyses compared to valence shell-model calcula-
tions.
infer R
s
= (6/7) (0.58)/(0.95) = 0.52, where (6/7) is the center-of-mass correction.
21.2.2
12
C
12
C can easily be treated in the full 0p-shell. The WBP interaction [23] gives a ground
state wave function with 48% (0p
3/2
)
8
, 41% (0p
3/2
)
6
-(0p
1/2
)
4
, 4.7% (0p
3/2
)
5
-(0p
1/2
)
3
and 5.5% (0p
3/2
)
4
-(0p
1/2
)
4
. The proton occupancies are 3.41 for 0p
3/2
and 0.59 for
0p
1/2
. The spectroscopic strength gets split into C
2
S = 3.16 to the 3/2
11
B ground
state, 0.58 to a 1/2

state at 1.85 MeV and 0.19 to a 3/2

state at 4.31 MeV. The

small remainder of 0.07 is fragmented over many states above 10 MeV in excitation.
A common feature of all calculations within the 0p shell with a variety of eective
interactions is that most (3.9 out of 4.0) of the p-shell sum-rule strength is contained
in the lowest three states in
11
B. The sum of the spectroscopic strength obtained from
the nonrelativistic analysis of the
12
C(e,e

p) data for the lowest three states in

13
B is
2.18 (Table [22.2]) giving R
s
= (11/12) (2.18)/(3.9) = 0.51. The Glauber model
analysis gives R

s
= (11/12) (3.56)/(3.9) = 0.84.
21.2.3
16
O
For
16
O it is possible to go up to 4 h beyond the closed-shell conguration explicit
conguration mixing [31]. Conguration mixing for
16
O with the WBP interaction
21 EXPERIMENTS RELATED TO SPECTROSCOPIC FACTORS 250
[23] gives a ground state wave function with 49% closed-shell conguration and 39%
2 h (mainly 2p-2h) and 12% 4 h (mainly 4p-4h). The proton occupation numbers
are 2.00 (0s
1/2
), 3.70 (0p
3/2
), 1.68 (0p
1/2
), 0.36 (0d
5/2
), 0.17 (0d
3/2
) and 0.07 (1s
1/2
)
(with about 0.01 in the 1p0f shell). For the
16
O
15
N transition, the transition to
the 1/2

ground state has C

2
S=1.65 (most of the 0p
1/2
strength) and the transition
to the lowest 3/2

state has C
2
S=3.29 (89% of the 0p
3/2
strength. The ratio of the
spectroscopic factors obtained from the nonrelativistic analysis of the (e,e

p) data
(Table [22.2]) to these calculated values is R
s
(1/2

, gs) = (15/16) (1.27)/(1.65) =

0.72 and R
s
(3/2

, 6.32MeV ) = (15/16) (2.25)/(3.29) = 0.64, or R

s
= 0.66 for the
sum. The relativistic analysis gives a higher value of R

s
= 0.81.
21.2.4 The sd-shell
One of the most complete models available is for conguration mixing in the sd shell.
In Fig. (21.6) I show the proton occupation numbers which result from the wave
functions obtained with the USD interaction in the full sd-shell basis. They are
compared to the extreme single-particle (ESP) model. For
28
Si in the middle of the
sd-shell the ESP model corresponds to a closed-shell (0d
5/2
)
12
conguration with a
proton occupation of number of 6 and a proton spectroscopic model of 6. In the full
sd-shell model with 839 basis states, the ground-state wave function has only 22%
of the (0d
5/2
)
12
state. However, the other basis states contain some nucleons in the
0d
5/2
orbital and the proton occupancy is 4.62. Thus the sum-rule strength for 0d
5/2
pickup in the full sd-shell model is 4.62. The transition to the
27
Al ground state
gets 3.61 of this sum and the rest gets fragmented over several excited states in
27
Al.
The spectroscopic factor from the (
3
He,d) of Vernotte et al. (with the 0.75 DWBA
reduction) of 2.75 is 25% smaller that the sd-shell value.
The
28
Si(e,e

p) reaction was studied at low resolution in one of the rst exper-

iments of this type [32], and an unexpectedly large spectroscopic factor of 5.5 for
protons in 0d orbit inferred (it would be nice to have a new experiment on
28
Si).
In place of
28
Si, we can consider the
30
Si(e,e

p)
29
Al result from [15] which gives an
experimental spectroscopic factor of 2.21(20) for the transition to the ground state
which when compared to the sd-shell calculation of 3.79 gives R
s
= 0.58(6).
The ratio of experiment to theory for 39 transitions in the sd-shell is plotted vs
mass in Fig. (21.4). For the 0d states one observes a mass-dependent ratio from about
R
s
= 0.9 at the lower end to R
s
= 0.6 at the upper end. An interpretation of this
will be discussed below.
21 EXPERIMENTS RELATED TO SPECTROSCOPIC FACTORS 251
p
r
o
t
o
n

o
c
c
u
p
a
t
i
o
n

n
u
m
b
e
r
Z
0
1
2
3
4
5
6
7
6 8 10 12 14 16 18 20 22
Figure 6: Proton occupation numbers for the even-even N = Z sd shell nuclei
as a function of mass. The lines are those expected in the extreme single-particle
shell model and the symbols are those from the USD interaction: 0d
5/2
(lled
circles), 0d
3/2
(crosses) and 1s
1/2
(squares).
21.2.5
40
Ca and
48
Ca
One can treat conguration mixing in
40
Ca in terms of (0d
3/2
,0f
7/2
)
n
conguration
outside of an assumed closed shell for
32
S. The ground state obtained with the HJMW
interaction [33] has 76% (0d
3/2
)
8
, 21% (0d
3/2
)
6
-(0f
7/2
)
2
, 2.4% (0d
3/2
)
4
-(0f
7/2
)
4
, and
0.13% (0d
3/2
)
2
-(0f
7/2
)
6
. The 0d
3/2
proton occupancy is 3.73 and and most of this
(3.70) goes to the spectroscopic factor for the
39
K 3/2
+
ground state. Taking the
(e,e

p) experimental value from Table [22.2] gives R

s
= 0.70 for the nonrelativistic
model and R

s
= 0.90 for the relativistic model. The 0f
7/2
occupancy is 0.27 and 0.18
of this goes to the lowest
39
K 7/2

state at 2.34 MeV.

For
48
Ca the HJMW ground state has has 92% (0d
3/2
)
8
-(0f
7/2
)
8
, 7% (0d
3/2
)
6
-
(0f
7/2
)
10
and 0.17% (0d
3/2
)
4
-(0f
7/2
)
12
. The 0d
3/2
proton occupancy is 3.85 and and
3.00 of this goes to the spectroscopic factor of the
47
K 3/2
+
state. Taking the (e,e

p)
experimental value of 2.26 from Table [22.2] gives R
s
= 0.75.
21.2.6
51
V
Calculations can be carried out in the full pf shell [34]. The dimension for the
51
V
ground state is large, 938,626 J=7/2, T=t5/2 states, but the wave function is still
21 EXPERIMENTS RELATED TO SPECTROSCOPIC FACTORS 252
dominated by the 0f
7/2
congurations and the average occupation proton number
obtained with the FPD6 interaction [35] is about 2.68. Most of this into the spectro-
scopic factors in the lowest states of
50
Ti given in Table [22.2]. The ratio of experiment
(Table [22.2]) to the full pf shell value is R
s
= 0.50.
21.2.7
208
Pb
It has recently become possible to consider the nucleus
208
Pb in a 24 orbit model
space with 2p-2h mixtures into the closed-shell ground state [36]. The resulting wave
function is 32% closed shell plus 68% 2p-2h. The resulting proton occupation numbers
for the orbits observed in Table [22.2] are 7.91 (0g
7/2
), 5.88 (1d
5/2
), 11.83 (0h
11/2
),
3.85 (0d
3/2
) and 1.91 (2s
1/2
). Although the ground state is only 32% closed shell the
occupations are close to the (2j + 1) ESP limit. Of this total occupation about 90%
(with the exception of 0g
7/2
) goes into the lowest state of each spin. About 60% of the
0g
7/2
strength goes into the lowest state with the rest fragmented over more highly
excited states. The average ratio of experiment to theory is to shell-model theory
is about R
s
= 0.65 for the nonrelativistic analysis and R

s
= 0.82 for the relativistic
analysis.
21.3 Short-range correlations
Summarizing the results from previous sections, we nd that there usually a reduc-
tion in the spectroscopic strength compared to valence shell-model calculations. The
reduction appears to be largest for cases in which a nucleon is picked up from a closed-
shell conguration leading to a single-particle state which is below the fermi surface.
The size of the reduction has some reaction model-dependence ranging from an av-
erage value of R
s
= 0.65 for the nonrelativistic analysis of (e,e

p) data to R

s
= 0.85
for a relativistic analysis of the same data. I will express this result in terms of a
correction factor
s
dened by:
R
s
= (1
s
) (21.3)
There is a trend for the 0d orbit within the sd-shell, Fig. (21.6) for a smaller
correction factor (
s
0.1) at the beginning of the shell (A=17) where the single-
particle state is above the fermi surface to a larger correction factor (
s
0.4) at the
upper end of the shell (A=39) where the state is below the fermi surface. [However,
the mass dependence observed in Fig. (21.4) may also be related to the incorrect
use of a constant value of r
o
= 1.25 fm for all sd-shell nuclei.] Other examples for
spectroscopic factors of states above the fermi surface where the correction factor is
21 EXPERIMENTS RELATED TO SPECTROSCOPIC FACTORS 253
small are those obtained from the radioactive beam studies [24] and those obtained
from the proton decay of unbound states discussed above.
One might expect that the correction factor due to short-range correlations to
depend on the radial size of the overlap function relative to the size of the core density.
When the overlap function is composed of orbitals below the fermi surface, the nucleon
being removed will be spatially close to the core nucleons and the interaction may
result in a relatively large short-range correlation. On the other hand if the nucleon
being removed is above the fermi surface, there is a smaller spacial overlap with
the core and the short-range correlation will be smaller. A semi-quantitative model
for a hard-core potential of radius a
h
has been derived by Birse and Clement [37].
(Birse and Clement [37] also discuss the eect of short-range correlations on the spin-
dependent sum rules [38].) The correction is given in terms of an integral containing
the valence density,
v
(r), and the total matter density,
m
(r) =
p
(r) +
n
(r):

s
=
4
3
a
3
h
_

v
(r)
m
(r)r
2
dr (21.4)
(The Pauli principle leads to a dependence on the core density that is proportional
to [37]
1
2

tz
(r) +
tz
(r) where t
z
is the isospin of the transferred nucleon and
tz
(r)
are the densities of the core protons/neutrons. In Eq. (21.4) I ignore the factor of
1
2
.)
In Fig. (21.7) I plot the components of this integral for a selection of neutron valance
orbits in the closed-shell congurations of
16
O and
40
Ca obtained from SKX Skyrme
Hartree-Fock calculations [9]. For
16
O the valence states below the fermi surface are
0s
1/2
bound by -30 MeV and 0p
3/2
bound by -18 MeV. For
16
O the valence states above
the fermi surface are 0d
5/2
bound by 5 MeV (a typical value at the beginning of the
shell) and 1s
1/2
bound by 0.5 MeV (the value of its separation energy in
11
Be). For
40
Ca the valence state below the fermi surface is the 0d
3/2
bound by 14.8 MeV. The
relative value of the integrals are are 0.200, 0.144, 0.097, 0.048 and 0.160 for 0s
1/2
,
0p
3/2
, 0d
5/2
, 1s
1/2
and 0d
3/2
, respectively. Thus we nd that a signicant binding
energy dependence for the short-range corrections with the most loosely bound states
having the smallest values.
Variational Monte-Carlo (VMC) calculations should contain both long and short
ranged correlations. VMC calculations for the
7
Li(e,e

p) reaction give good agreement

with the data. The small R
s
value in this case may be inuenced by the dierence in
structure between
7
Li and
6
He caused by the small binding (and large neutron radius)
of
6
He. The alpha-triton clustering aspects of
7
Li may also enter. Both of these are
relatively long-range eects which are special to light nuclei and which should be
taken into account within the VMC. Thus the relative importance of short-range
correlations in the VMC calculations for
7
Li is not clear.
The VMC calculation [39], [40] for
16
O(e,e

p) gives R
th
= 0.90 for the ratio to
the extreme single-particle model (without center-of-mass corrections) for the 0p
3/2
21 EXPERIMENTS RELATED TO SPECTROSCOPIC FACTORS 254

r (fm)
0.0
0.05
0.1
0.15
0 1 2 3 4 5 6
16O 0s1/2

0.0
0.05
0.1
0.15
16O 0p3/2

0.0
0.05
0.1
0.15
16O 0d5/2
P
(
r
)
0.0
0.05
0.1
0.15
16O 1s1/2

0.0
0.05
0.1
0.15
0.2
40Ca 0d3/2
Figure 7: Components of the overlap function of Eq. (21.4). The solid line is the
matter density,
m
(r), (normalized by 4
_

m
(r) r
2
dr = A). The dashed line
is r
2

v
(r) (in arbitrary units) and the dotted line is r
2

m
(r)
v
(r) (in arbitrary
units).
orbit. The role of the various types of short-range correlations has been studied in the
correlated-basis-function theory [41] where it was found that tensor correlations are
most important. Since the VMC calculations do not include the eect of fragmenta-
tion of the hole strength, but should contain the long-range correlations, we should
compare this to the ratio obtained from the sum of the 3/2

states in Table [22.2],

e.g. R = (15/16) (2.60)/(4.0) = 0.61. Use of the relativistic (e,e

p) reaction model
[14] would bring this up to about R

= 0.76. If we assume that the VMC calculation

does not include the long-range correlation included in the shell-model calculation
then we should compare R
th
= 0.90 to the value of R

s
= 0.81 found in the previous
section.
21.4 Conclusions
It appears possible to quantitatively understand spectroscopic factors in terms of
a sum of short and long-long-range correlations for the overlap function and with
the relativistic analysis of the (e,e

p) data. When the spectroscopic factors obtained

REFERENCES 255
from the relativistic reaction theory are compared to valence shell-model calculations
(which account for the long-range correlations) one obtains an average value of about
R

s
= 0.85 for states below the fermi surface. This 15% reduction could be attributed
to short-range correlations. Calculations based upon the NN interaction [39], [40]
can account for most of this, R
th
s
= 0.90.
The long-range correlations have the eect of fragmenting the spectroscopic
strength over many states in the A+1 and A1 nuclei, and are strongly dependent
on the spacing and occupancy of the single-particle states around the fermi surface.
Conguration mixing with empirical two-body matrix elements or renormalized G
matrix interaction can account for the observed fragmentation, but the details are
sensitive to the use of the correct single-particle energies. The details of the fragmen-
tation and comparison to experiment is one of the most direct ways of testing nuclear
structure models.
References
[1] F. Hinterberger et al., Z. Phys. 202, 236 (1967).
[2] J. N. Craig, N. S. Wall and R. H. Bassel, Phys. Rev. Lett. 36, 656 (1976).
[3] G. J. Kramer, H. P. Blok and L. Lapikas, Nucl. Phys. A679, 267 (2001).
[4] B. H. Wildenthal, Prog. in Part. and Nucl. Physcis 11, 5 (1984).
[5] http://www.nscl.edu/

brown/sde.htm
[6] B. A. Brown and B. H. Wildenthal, Ann. Rev. Nucl. Part. Sci. 38, 29 (1988).
[7] Spectroscopic Factors for Single Nucleon Transfer for A = 21-44, P. M. Endt,
Atomic and Nuclear Data Tables 19, 23 (1977).
[8] J. Vernotte et al., Nucl. Phys. A571, 1 (1994).
[9] B. A. Brown, Phys. Rev. C 58, 220 (1998).
[10] J. Bartel, P. Quentin, M. Brack, C. Guet and M. B. Hakansson, Nucl. Phys.
A386, 79 (1982).
[11] W. A. Richter and B. A. Brown, submitted for publication.
[12] J. W. A. den Herder et al., Phys. Rev. Lett. 57, 1843 (1986).
[13] M. Leuschner et al., Phys. Rev. C 49, 955 (1994).
REFERENCES 256
[14] J. Gao et al., Phys. Rev. Lett. 84, 3265 (2000).
[15] J. Wesseling et al., Nucl. Phys. A547, 519 (1992).
[16] Y. Jin, J. K. Zhang, D. S. Onley and L. E. Wright, Phys. Rev. C 47, 2024 (1993).
[17] J. M. Udias, P. Sarriguren, E. Moya de Guerra, E. Garrido and J. A. Caballero,
Phys. Rev. C 48, 2731 (1993).
[18] D. Branford et al., Phys. Rev. C 63, 014310 (2000).
[19] L. Lapikas, G. van der Steenhoven, L. Frankfurt, M. Strikman and M. Zhalov,
Phys. Rev. C 61, 064325 (2000); L. Frankfurt, M. Strikman and M. Zhalov, Phys.
Lett. B 503, 73 (2001).
[20] M. Radici, W. H. Dickho and E. Roth Stoddard, Phys. Rev. C 66, 014613
(2002).
[21] D. J. Millener, to by published.
[22] D. J. Millener and D. Kurath, Nucl. Phys. A255, 315 (1975).
[23] W. K. Warburton and B. A. Brown, Phys. Rev. C 46, 923 (1992).
[24] P. G. Hansen and B. M. Sherrill, Reactions and Single-Particle Structure of
Nuclei Near the Drip Lines, Nucl. Phys. A (to be published).
[25] A. Navin et al., Phys. Rev. Lett. 81, 5089 (1998).
[26] T. Aumann, et al., Phys. Rev. Lett. 84, 35 (2000).
[27] A. Navin et al., Phys. Rev. Lett. 85, 266 (2000).
[28] V. Guimaraes et al., Phys. Rev. C 61, 06409 (2000).
[29] V. Maddalena et al., Phys. Rev. C 63, 024613 (2001).
[30] L. Lapikas, J. Wesseling and R. B. Wiringa, Phys. Rev. Lett. 82, 4404 (1999).
[31] E. K. Warburton, B. A. Brown and D. J. Millener, Phys. Lett. B293, 7 (1992).
[32] J. Mougy et al., Nucl. Phys. A262, 461 (1976).
[33] S. T. Hsieh, X. Ji, R. Mooy and B. H. Wildenthal, AIP Conference Proceedings
142, 357 (1986).
[34] M. Honma, private communication.
REFERENCES 257
[35] W. A. Richter, M. G. van der Merwe, R. E. Julies and B. A. Brown, Nucl. Phys.
A523, 325 (1991).
[36] B. A. Brown, Phys. Rev. Lett. 85, 5300 (2000).
[37] M. C. Birse and C. F. Clement, Nucl. Phys. A351, 112 (1981).
[38] C. F. Clement, Nucl. Phys. A213, 469 (1973); C. F. Clement, Nucl. Phys. A213,
493 (1973); C. F. Clement and S. M. Perez, Phys. Lett. 81B, 269 (1979); C. F.
Clement and S. M. Perez, Nucl. Phys. A362, 86 (1981).
[39] M. Radici, S. Bo, S. C. Pieper and V. R. Pandharipande, Phys. Rev. C 50,
3010 (1994).
[40] D. Van Neck, M. Waroquier, A. E. L. Dieperink, S. C. Pieper and V. R. Pand-
haripande, Phys. Rev. C 57, 2308 (1998).
[41] D. Van Neck, L. Van Daele, Y. Dewulf and M. Waroquier, Phys. Rev. C 56,
1398 (1997).
22 ONE-BODY TRANSITION OPERATORS AND THE OBTD 258
22 One-body transition operators and the OBTD
The M-scheme one-body transition operator is:

< [ O

[ > a
+

The tensor coupled form for this operator is:

kk

< k

[[O

[[k

>

mm

(1)
jm
_
j

_
a
+

kk

< k

[[O

[[k

>
[a
+
k
a
k

_
(2 + 1)
. (22.1)
Where stands for (n

), and k

stands for (n

). It is convenient to
express the reduced matrix element for the n-particle wave function in the form of
a product over one-body transition densities (OBTD) times reduced single-particle
matrix elements
< f[[

[[i >=< nJ[[

[[n

>=

kk

OBTD(fik

) < k

[[O

[[k

>, (22.2)
where the OBTD is given by
OBTD(fik

) =
< nJ[[[a
+
k
a
k

[[n

>
_
(2 + 1)
. (22.3)
The labels i and f are a short-hand notation for the initial and nal state quantum
numbers (n

) and (nJ), respectively. The OBTD represents in a compact form,

the most general information needed to calculate the matrix elements of one-body
operators between a given initial and nal state. The OBTD can be calculated in
an M-scheme basis by dividing a nonvanishing expectation value of the M-scheme
matrix element by the 3j factor from the Wigner-Eckhart theorem
< J[[[a
+
k
a
k

[[J

>=
< JM [ [a
+
k
a
k

[ J

>
(1)
JM
_
J J

M M

_
(22.4)
The OBTD can be evaluated in a J-coupled basis by inserting a complete set of
states with (n 1) particles between the a
+
and a operators in Eq. (22.3) with Eq.
[13.45] to obtain:
OBTD(fik

) = (1)
J++J

_
J

J
j

_
22 ONE-BODY TRANSITION OPERATORS AND THE OBTD 259
< nJ[[a
+
k
[[(n 1)

>< (n 1)

[[ a
k

[[n

>
=

(1)
J++J

+j

_
J

J
j

_
< nJ[[a
+
k
[[(n 1)

>< n

[[a
+
k

[[(n 1)

> . (22.5)
These last two matrix elements can also be expressed as one-particle CFP. The explicit
forms of the OBTD for the cases of a single-k conguration and for a two-orbital
conguration are derived in the next two sections.
22.1 Isospin and proton-neutron formalism
The formulae in the previous section are immediately applicable to the proton-neutron
formalism. In this case the labels for the orbits k implicitly include the proton or
neutron label for the orbital. For example for

decay or (p,n) reactions we would

destroy a neutron (q = 1/2) and create a proton (q

= 1/2):
OBTD(fik
,q
k
,q
) =
< nJ[[[a
+
k,q
a
k

,q
]

[[n

>
_
(2 + 1)
. (22.6)
To obtain the equivalent expression in terms of a reduced matrix element in
isospin formalism we would label the wavefunctions explicitly with their (T, T
z
) values
and the operator in terms of the isospin rank (T, T
Z
) and then use the Wigner-
Eckhart theorm in isospin space to obatin:
OBTD(fik
,q
k
,q
) =
< nJTT
z
[[[a
+
k,q
a
k

,q
]
,T,Tz
[[n

z
>
_
(2 + 1)
= (1)
TTz
_
T T T

T
z
T
z
T

z
_
< nJT[[[a
+
k,q
a
k

,q
]
,T
[[n

>
_
(2 + 1)
=< T

z
TT
z
[ TT
z
>

_
(2T + 1)
(2T + 1)
< nJT[[[a
+
k,q
a
k

,q
]
,T
[[n

>
_
(2 + 1)(2T + 1)
=< T

z
TT
z
[ TT
z
>

_
(2T + 1)
(2T + 1)
OBTD(fik

T) (22.7)
where the last line contains the one-body transition density in isospin formalism:
OBTD(fik

T) =
< nJT[[[a
+
k
a
k

]
,T
[[n

>
_
(2 + 1)(2T + 1)
. (22.8)
22 ONE-BODY TRANSITION OPERATORS AND THE OBTD 260
22.2 OBTD for a single-orbital conguration
The OBTD for a single-orbital conguration is given by
OBTD(f, i) =

(1)
J++J

+j

_
J

J
j

_
< k
n
J[[a
+
k
[[k
n1

>< k
n

[[a
+
k

[[k
n1

> (22.9).
The only possibility for the set (k

) in the summation in Eq. (22.5) is (k, k) and

one has in terms of the one-particle CFP:
OBTD(fi) = n
kk

k
_
(2J + 1)(2J

+ 1)

(1)
J++J

+j

_
J

J
j j J

_
< j
n
J [j
n1

>< j
n

[j
n1

> . (22.10)
22.3 OBTD for a two-orbital conguration
First, the two-orbital conguration in which the initial and nal states have the same
partition is considered
< i [ < n

[=< (k
n
1
1

1
J

1
)(k
n
2
2

2
J

2
)J

[ (22.11),
and
< f [ < nJ [=< (k
n
1
1

1
J
1
)(k
n
2
2

2
J
2
)J [ (22.12).
Consideration of Eq. (22.5) shows that there are two possible intermediate states
[ a >[ (n 1)

(a) >=[ (k
n
1
1
1

1
J

1
)(k
n
2
2

2
J

2
)J

>, (22.13)
and
[ b >[ (n 1)

(b) >=[ (k
n
1
1

1
J

1
)(k
n
2
1
2

2
J

2
)J

> . (22.14)
The OBTD will be a sum of the two terms
OBTD(fi) = OBTD
a
(fi) + OBTD
b
(fi). (22.15)
For state (a)
OBTD
a
(fi) =

1
J

2
J

2
J

(1)
J++J

+j

_
J

J
j
1
j
1
J

_
< (k
n
1
1
J
1
)(k
n
2
2
J
2
)J[[a
+
k
[[(k
n
1
1
1
J

1
)(k
n
2
2
J

2
)J

>
22 ONE-BODY TRANSITION OPERATORS AND THE OBTD 261
< (k
n
1
1
J

1
)(k
n
2
2
J

2
)J

[[a
+
k

[[(k
n
1
1
1
J

1
)(k
n
2
2
J

2
)J

> (22.16)
(the labels are implicit in the wave functions) and the set () must be (1, 1). Ap-
plication of Eq. [13.49] together with the sum-rule of Eq. [13.23] for the 6j coecients
gives
OBTD
a
(fi) =
()(1,1)

1
J

2
J

2
J

(1)
J++J

+j
1
+J
1
+J
2
J

1
J

2
(2J

+ 1)
_
(2J + 1)(2J

+ 1)

J
2
J

2
J

_
J

J
j

__
J
1
J

1
j
1
J

J J
2
__
J

1
J

1
j
1
J

2
_
< k
n
1
1

1
J
1
[[a
+
k
1
[[k
n
1
1
1

1
J

1
>< k
n
1
1

1
J

1
[[a
+
k
1
[[k
n
1
1
1

1
J

1
> (22.17)
= n
1

()(1,1)

J
2
J

2
_
(2J + 1)(2J

+ 1)(2J
1
+ 1)(2J

1
+ 1)

_
J

J
J
1
J

1
J
2
_

1
J

1
(1)
J

j
1
+J

2
J

1
_
J
1
J

1

j
1
j
1
J

1
_
< j
n
1
1

1
J
1
[j
n
1
1
1

1
J

1
>< j
n
1
1

1
J

1
[j
n
1
1
1

1
J

1
> . (22.18)
The contribution from intermediate state [ b > is related to this result by interchang-
ing the subscripts 1 and 2, and multiplying by the phase factor (1)
J
1
+J
2
J+J

1
+J

2
J

to obtain
OBTD
b
(fi) = n
2

()(2,2)

J
1
J

1
_
(2J + 1)(2J

+ 1)(2J
2
+ 1)(2J

2
+ 1)

_
J

J
J
2
J

2
J
1
_

2
J

2
(1)
j
2
+J
1
+J

2
J
2
+JJ

2
_
J
2
J

2

j
2
j
2
J

_
< j
n
2
2

2
J
2
[j
n
2
1
2

2
J

2
>< j
n
2
2

2
J

2
[j
n
2
1
2

2
J

2
> . (22.19)
For the two-orbit case, there is also a OBTD connecting an initial state and with a
nal state which diers by a change of one particle from one orbit to another
< i [ < nJ [=< (k
n
1
1

1
J
1
)(k
n
2
2

2
J
2
)J [ (22.20),
and
< f

[ < n

[=< (k
n
1
1
1

1
J

1
)(k
n
2
+1
2

2
J

2
)J

[ (22.21).
Consideration of Eq. (22.5) shows that there is only one possible type of intermediate
state
[ a >[ (n 1)

>=[ (k
n
1
1
1

1
J

1
)(k
n
2
2

2
J

2
)J

> . (22.22)
The OBTD is obtained by application of Eqs. [13.48] and [13.49] together with the
sum rule of Eq. [13.29]
OBTD(f

i) =
()(1,2)
(1)
j
2
+J
2
J

2
_
n
1
(n
2
+ 1)
22 ONE-BODY TRANSITION OPERATORS AND THE OBTD 262

_
(2J + 1)(2J

+ 1)(2J
1
+ 1)(2J

2
+ 1)
_

_
J

2
J

1
J

J
2
J
1
J
j
2
j
1

_

_
< j
n
1
1

1
J
1
[j
n
1
1
1

1
J

1
>< j
n
2
+1
2

2
J

2
[j
n
2
2

2
J
2
> . (22.23)
The matrix elements for the other possible cases where the nal-state partition diers
from the initial-state partition by the moving of one particle can be derived in an
analogous manner and will have a form similar to Eq. (22.23). The OBTD for the
case when two or more particles are moved between partitions vanishes.
22.4 Scalar one-body matrix elements
For the Hamiltonian one is interested in matrix elements of scalar operators. For the
special case of a scalar (=0) one-body operator the results simplify and it is useful
to dene a new quantity OBTDS related to the unreduced matrix elements by
< nJ [

O
=0
[ n

>=

kk

OBTDS(fik

) < k

[ O
=0
[ k

>, (22.24)
where OBTDS is given by
OBTDS(fik

) =
jj

JJ

_
(2j

+ 1)
(2J + 1)
OBTD(fik

= 0)
=

_
(2j

+ 1)
(2J + 1)
< nJ[[[a
+
k
a
k

]
=0
[[n

>=
_
2j

+ 1 < nJ [ [a
+
k
a
k

]
=0
[ n

>
=
1
(2J + 1)

jj

JJ

< nJ[[a
+
k
[[(n 1)

>< n

[[a
+
k

[[(n 1)

> .
(22.25)
For a single-orbital conguration considered in Eq. (22.10), OBTDS reduces to
OBTDS(fi) = n
kk

JJ

< j
n
J [j
n1

>
< j
n

[j
n1

>= n

JJ
. (22.26)
where the last line follows from the sum rule of Eq. [20.8]. Similarly for a two-orbital
conguration one obtains
OBTDS
a
(fi) = n
1

()(1,1)

J
1
J

J
2
J

2
, (22.27)
OBTDS
b
(fi) = n
2

()(2,2)

J
1
J

J
2
J

2
, (22.28)
22 ONE-BODY TRANSITION OPERATORS AND THE OBTD 263
and
OBTDS(f

i) =
()(1,2)

j
1
j
2

JJ
(1)
J+J

1
+J

2
_
n
1
(n
2
+ 1)
(2J
1
+ 1)(2J

2
+ 1)
_
J

2
J
2
j
1
J
1
J

1
J
_
< j
n
1
1

1
J
1
[j
n
1
1
1

1
J

1
>< j
n
2
+1
2

2
J

2
[j
n
2
2

2
J
2
> . (22.29)
A special case of Eq. (22.25) is obtained for k

= k

= k
o
OBTDS(fi, k

= k

= k
o
)
_
2j
o
+ 1 < nJ [ [a
+
ko
a
ko
]
=0
[ n

J >
=
_
2j
o
+ 1

m
< j
o
mj
o
, m [ 0, 0 >< nJ [ a
+
kom
a
ko,m
[ n

J >
=

m
< nJ [ a
+
kom
a
kom
[ n

J >=< nJ [

N
ko
[ n

J >, (22.30)
where in the last line the single-state number operator

N
ko
=

m
a
+
kom
a
kom
. (22.31)
has been introduced. In general [ nJ > are not eigenstates of

N
ko
. However, they
can be expanded in terms of basis state
a
each of which has a xed partition (an
integer number of particles n
o
(a) in state k
o
) and hence are eigenstates of

N
ko
;
[ nJ >=

a
c
a
[
a
>, (22.32)
and

N
ko
[
a
>= n
o
(a) [
a
> . (22.33)
Thus one obtains
OBTDS(fi, k

= k

= k
o
) =< nJ [

N
ko
[ n

J >
=

ab
(c
b
)

a
<
b
[

N
ko
[
a
>=

a
(c
a
)

a
n
o
(a). (22.34)
In particular, for the diagonal matrix element, f = i (

= and c

a
= c
a
),
OBTDS(f = i, k

= k

= k
o
) =< nJ [

N
ko
[ nJ >
=

a
[c
a
[
2
n
o
(a) < p
o
>
nJ
. (22.35)
where < p
o
>
nJ
is the average number of k
o
particles in the state [ nJ >.
Eqs. (22.999), (22.999), and (22.999) are consistent with the general result of Eqs.
(22.35). One should not confuse the condition k

= k

with the condition j

= j

. For
22 ONE-BODY TRANSITION OPERATORS AND THE OBTD 264
example, the pair of states k = 3 [(nj) = (0, 1, 1/2)] and k = 6 [(nj) = (1, 0, 1/2)]
obviously fullls the latter condition but not the former condition, and thus one must
use Eq. (22.25) to evaluate OBTDS. However, often the calculation is restricted to a
set of active orbits for which j

= j

also implies that k

= k

, and hence one can

use the simpler results of Eqs. (22.34) and (22.35) for all scalar matrix elements in
the model space. This is the case, for example, when the model space is restricted to
a single major harmonic oscillator shell.
A useful sum rule for the one-particle transfer matrix elements can be obtained
from a comparison of Eqs. (22.25) and (22.35)

[< nJ[[a
+
ko
[[(n 1)

>[
2
= (2J + 1) < p
o
>
nJ
. (22.36)
A related sum rule can be obtained by rewriting the number operator as

N
ko
=

m
(1 a
kom
a
+
kom
). (22.37)
By methods similar to the above derivation one obtains

[< (n + 1)

[[a
+
ko
[[nJ >[
2
= (2J + 1)(2j
o
+ 1) < p
o
>
nJ
(2J + 1) < h
o
>
nJ
. (22.38)
where < h
o
>
nJ
is the average number of k
o
holes in the state [ nJ >.
23 TWO-PARTICLE TRANSFER OPERATORS 265
23 Two-particle transfer operators
The operator that creates an antisymmetric state of two particles coupled to angular
momentum J is dened by
A
+
(k

J
o
M
o
) [>=[ k

J
o
M
o
>, (23.1)
and hence from Sec. 16.2:
A
+
(k

J
o
M
o
) = N
12

mm

< j

[ J
o
M
o
> a
+
k

a
+
km
= N
12

mm

< j

[ J
o
M
o
> a
+
km
a
+
k

= N
12
[a
+
k
a
+
k

]
Jo
Mo
, (23.2)
where N
12
=
1

(1+
kk

)
. The two-particle destruction operator is
A(k

J
o
M
o
) = A
+
(k

J
o
M
o
)
+
=
= N
12

mm

< j

[ J
o
M
o
> a
km
a
k

. (23.3)
The tensor form of the two-particle destruction operator is

A(k

J
o
M
o
) = (1)
JoMo
A
+
(k

J
o
, M
o
)
+
= N
12
(1)
JoMo

mm

< j

[ J
o
, M
o
> a
km
a
k

= N
12
[ a
k
a
k

]
Jo
Mo
.
(23.4)
Matrix elements involving

A can be converted into those involving A
+
:
< (n 2)J[[

A(k

J
o
)[[n

>
= (1)
Jo+JJ

< n

[[A
+
(k

J
o
)[[(n 2)J > . (23.5)
All matrix elements of A
+
can be reduced to two types of matrix elements in-
volving active states. The rst type with = involves a matrix element related to
the two-particle CFP, denoted by < j
n
J [j
2
(J
o
); j
n2

> and dened by:

< j
n
J[[A
+
(kkJ
o
)[[j
n2

>

n(n 1)(2J + 1)
2
< j
n
J [j
2
(J
o
); j
n2

> . (23.6)
There are a variety of notations in the literature for the two-particle CFP and the
particular choice made in Eq. (23.6) is made to reect the structure of the A
+
matrix
element. Note that j is used in the two-particle CFP rather than k in order to
23 TWO-PARTICLE TRANSFER OPERATORS 266
emphasize the fact that it does not depend on the (n) quantum number. Wave
function expansion relations and sum rules for the two-particle CFP can be derived
by applying the two-particle number operator

JoMo
A
+
(kkJ
o
M
o
)A(kkJ
o
M
o
)
=
1
2

JoMom
i
< jm
1
jm
2
[ J
o
M
o
>< jm
3
jm
4
[ J
o
M
o
> a
+
km
2
a
+
km
1
a
km
3
a
km
4
=
1
2

m
1
m
2
a
+
km
2
a
+
km
1
a
km
1
a
km
2
=
1
2
(

N
2
k


N
k
) (23.7)
to the k
n
conguration and inserting a complete set of states with (n 2) particles
to obtain
1
2
(

N
2
k


N
k
) [ k
n
JM >=
1
2
n(n 1) [ k
n
JM >
=

Jo
< k
n
J[[A
+
(kkJ
o
)[[k
n2

>
_
(2J + 1)
[Z
+
(k
n2

) A(kkJ
o
)]
J
M
[>, (23.8)
and

Jo
< j
n

[j
2
(J
o
); j
n2

>< j
n
J [j
2
(J
0
); j
n2

>=

JJ
. (23.9)
The square of the two-particle CFP is the probability that the removal of two particles
in state k coupled to angular momentum J
o
from the state [ k
n
(J) > will leave the
system in the state [ k
n2
(

) >. Since the CFP are normalized to unity, the

square of the CFP for a given initial state leading only one possible nal state is
unity, and thus
[< j
2
J [j
2
(J
o
); j
0
, J

= 0 >[=
JJo
. (23.10)
One nds for the two-particle CFP
< j
(n2)
J [j
2
(J
o
); j
n

>< j
2j+1(n2)
J [j
2
(J
o
); j
2j+1n

>
= (1)
Jo+J

_
n(n 1)(2J

+ 1)
[(2j + 3 n)(2j + 2 n)(2J + 1)]
< j
n

[j
2
(J
o
); j
n2
J > .
(23.11)
The case when n = 2 and J = 0 gives
[< j
0
, J = 0 [j
2
(J
o
); j
2
J

>[
=

2J

+ 1
j(2j + 1)
[< j
2
J

[j
2
(J
o
); j
0
, J = 0 >[=

2J

+ 1
j(2j + 1)

J

Jo
. (23.12)
23 TWO-PARTICLE TRANSFER OPERATORS 267
For the k
3
conguration the matrix element of A
+
can be related to the matrix element
of a
+
:
< k
3
J[[a
+
k
[[k
2
J

>= (1)
J+J

+j
< k
3
J[[A
+
(kkJ

)[[kj >, (23.13)

which in terms of CFP is
< j
3
J [j
2
J

>= (1)
J+J

+j
< j
3
J [j
2
(J

); j
1
j > . (23.14)
The two-particle CFP can also be reduced to a sum over one-particle CFP by
inserting a complete set of intermediate states with (n 1) particles between the a
+
of Eq. (23.2) (with
jj
= 1):
< k
n
J[[A
+
(kkJ
o
)[[k
n2

>
=
1

2
(1)
J+Jo+J

_
2J
o
+ 1

_
J

J J
o
j j J

_
< k
n
J[[a
+
k
[[k
n1

>< k
n1

[[a
+
k
[[k
n2

>
=
1

2
(1)
J+Jo+J

+1
_
(2J
o
+ 1)(2J + 1)n(n 1)

2J

+ 1
_
J

J J
o
j j J

_
< j
n
J [j
n1

>< j
n1

[j
n2

> .
(23.15)
In terms of CFP this becomes
< j
n
J [j
2
(J
o
); j
n2

>= (1)
J+Jo+J

+1
_
2J
o
+ 1

2J

+ 1
_
J

J J
o
j j J

_
< j
n
J [j
n1

>< j
n1

[j
n2

> .
(23.16)
In the second type of matrix element for A
+
, two dierent k-states are active and
Eqs. (23.2) (with
kk

= 0) can be used to express the matrix element in terms of

one-particle CFP
< (k
n

)(k
n

)J

[[A
+
(k

J
o
)[[(k
n1

)(k
n

)J

>
= (1)
n

+1
_
(2J

+ 1)(2J
o
+ 1)(2J

+ 1)

_
J

J
o
_

_
< k
n

[[a
+
k
[[k
n1

>< k
n

[[a
+
k

[[k
n

>
(23.17)
= (1)
n

+1
_
n

(2J

+ 1)(2J
o
+ 1)(2J

+ 1)(2J

+ 1)(2J

+ 1)
23 TWO-PARTICLE TRANSFER OPERATORS 268

_
J

J
o
_

_
< j
n

[j
n1

>< j
n

[j
n

> . (23.18)
The phase factor arises from the (1)
n

factor of Eq. [16.60] times the (1) from Eq.

(23.2).
24 TWO-BODY TRANSITION OPERATORS AND THE TBTD 269
24 Two-body transition operators and the TBTD
The M-scheme form for the two-body transition operator is:

=
1
4

< [ T

[ > a
+

a
+

(24.1)
The tensor coupled form for this two-body operator is:

=
1
4

kk

kk

mm

mm

< k

[ T

[ k

> a
+
km
a
+
k

a
k

a
km
(24.2)
=
1
4

kk

kk

JoMoJ

o
M

o
_
(1 +
kk

)
_
(1 +
kk

) < k

J
o
M
o
[ T

[ k

o
M

o
>

mm

mm

< J
o
M
o
[ j

>< J

o
M

o
[ j

> a
+
km
a
+
k

a
k

a
km
(24.3)
=
1
4

kk

kk

JoMoJ

o
M

o
< k

J
o
M
o
[ T

[ k

o
M

o
>
(1 +
kk

)(1 +
kk

)A
+
(k

J
o
M
o
)A(k

o
M

o
) (24.4)
=
1
4

kk

kk

JoJ

o
< k

J
o
[[T

[[k

o
>
(1 +
kk

)(1 +
kk

)
[A
+
(k

J
o
)

A(k

o
)]

2 + 1
(24.5)
=

kk

kk

JoJ

o
< k

J
o
[[T

[[k

o
>
[A
+
(k

J
o
)

A(k

o
)]

2 + 1
. (24.6)
Eq. (24.3) is obtained by the M-scheme expansion of the J-coupled two-body matrix
elements, Eq. [16.9], Eq. (24.4) comes from the denition of the two-particle operators
in Eqs. [23.2] and [23.3], and Eq. (24.5) comes from the tensor coupling of the A
+
and
A opertors. Also note in Eq. (24.5) that the (1) obtained from commuting a
k

with a
km
is cancelled by the (1) from Eq. [23.2]. It is convenient to express the
reduced matrix element for the n-particle wave function in the form of a product over
two-body transition densities (TBTD) times reduced two-particle matrix elements
< nJ[[

[[n

>=

kk

,kk

JoJ

o
TBTD(fikJ
o
J

o
) < k

J
o
[[T

[[k

o
>, (24.7)
24 TWO-BODY TRANSITION OPERATORS AND THE TBTD 270
where the TBTD is given by
TBTD(fikJ
o
J

o
) =
< nJ[[[A
+
(k

J
o
)

A(k

o
)]

[[n

>

2 + 1
. (24.8)
We can evaluate these in terms of matrix elements of the two-particle transfer oper-
ators by interting a complete set of states for the (n 2) particle system
TBTD(fikJ
o
J

o
) = (1)
J++J

_
J

J
J
o
J

o
J

_
< nJ[[A
+
(k

J
o
)[[(n 2)

>< (n 2)

[[

A(k

o
)[[n J

>
=

(1)
J++J

+J

o
_
J

J
J
o
J

o
J

_
< nJ[[A
+
(k

J
o
)[[(n 2)

>< n

[[A
+
(k

o
)[[(n 2)

> . (24.9)
The dependence of the TBTD on k

, k

, k

and k

is abreviated by k, and the depen-

dence of the TBTD on the initial and nal states is abreviated by i and f, respectively.
The TBTD represents in a compact form, the most general information needed to
calculate the matrix elements of two-body operators between a given initial and nal
state. The results given in Sec. 23 can be used to reduce the matrix elements of A
+
to one-particle and two-particle CFP. The TBTD can be evaluated for specic con-
gurations following the procedures given in Sec. 22 for the OBTD. The analogous
results for a scalar two-body operator will be given in the next section.
24.1 Scalar two-body operators
For the special case of a scalar (=0) two-body operator, it is useful to dene a new
quantity TBTDS related to the unreduced matrix elements by
< nJ [

T
=0
[ n

>
=

kk

kk

Jo
TBTDS(fikJ
o
) < k

J
o
[ T
=0
[ k

J
o
>, (24.10)
where TBTDS is given by
TBTDS(fikJ
o
) =
JoJ

JJ

_
(2J
o
+ 1)
(2J + 1)
TBTD(fikJ
o
J
o
= 0)
= (2J + 1)
1

< nJ[[A
+
(k

J
o
)[[(n 2)

>
24 TWO-BODY TRANSITION OPERATORS AND THE TBTD 271
< n

J[[A
+
(k

J
o
)[[(n 2)

> . (24.11)
This result is most often used with respect to the scalar two-body interaction V
< nJ [ V [ n

J >=

kk

kk

,Jo
TBTDS(fikJ
o
) < k

J
o
[ V [ k

J
o
>, (24.12)
In the next few subsections, the TBTDS for some simple congurations will be eval-
uated.
24.2 Scalar TBTD for a single-orbital conguration
First consider the evaluation of the TBTDS for a single-orbital conguration
TBTDS(fikJ
o
) = (2J + 1)
1

,J

< k
n
J[[A
+
(k

J
o
)[[k
n2

>
< k
n

J[[A
+
(k

J
o
)[[k
n2

> . (24.13)
The only possibility for the set (k

) in the summation in Eq. (24.12) is (kkkk)

and, in terms of the two-particle CFP:
TBTDS(fikJ
o
) =
(kk

kk

),(kkkk)
n(n 1)
2

< j
n
J [j
2
(J
o
); j
n2

>< j
n

J [j
2
(J
o
); j
n2

> . (24.14)
24.3 Scalar TBTD for a two-orbital conguration
For the second example, consider the interaction energy coecient for the two-orbital
conguration in which the initial and nal states have the form
< i [ < n

J [=< (k
n
1
1
J

1
)(k
n
2
2
J

2
)J [,
and
< f [ < nJ [=< (k
n
1
1
J
1
)(k
n
2
2
J
2
)J [ .
Consideration of Eq. (24.11) shows that there are three possible intermediate states
[ a >[ (n 2)

>=[ (k
n
1
2
1
J

1
)(k
n
2
2
J

2
)J

>,
[ b >[ (n 2)

>=[ (k
n
1
1
1
J

1
)(k
n
2
1
2
J

2
)J

>,
24 TWO-BODY TRANSITION OPERATORS AND THE TBTD 272
and
[c >[ (n 2)

>=[ (k
n
1
1
J

1
)(k
n
2
2
2
J

2
)J

> .
The TBTDS is given by Eq. (24.12) with a sum of three terms corresponding to these
three types intermediate states
TBTDS = TBTDS
a
+ TBTDS
b
+ TBTDS
c
.
In case (a), the set () must be (1, 1, 1, 1), in case (b), the set () must
be (1, 2, 1, 2), and in case (c), the set () must be (2, 2, 2, 2). For case (a), the
TBTDS can be obtained after application of Eq. [13.49] and the sum rule of Eq.
[13.27] for the 6j coecients
TBTDS
a
=
(),(1,1,1,1)
(2J + 1)
1

1
J

2
J

2
J

(1)
2J

2
+J

1
+J
1
+2J+2Jo

2
,

J
2
J

2
J

2
< k
n
1
1
J
1
[[A
+
(k
1
k
1
J
o
)[[k
n
1
2
1
J

1
>< k
n
1
1
J

1
[[A
+
(k
1
k
1
J
o
)[[k
n
1
2
1
J

1
>
(2J + 1)(2J

+ 1)
_
J
1
J

1
J
o
J

J J
2
__
J

1
J

1
J
o
J

J J
2
_
=
(),(1,1,1,1)
n
1
(n
1
1)
2

J
2
J

J
1
J

1
J

1
< j
n
1
1

1
J
1
[j
2
1
(J
o
); j
n
1
2
1

1
J

1
>< j
n
1
1

1
J
1
[j
2
1
(J
o
); j
n
1
2
1

1
J

1
> . (24.15)
Similarly the contribution from intermediate state (c) is
TBTDS
c
=
(),(2,2,2,2)
n
2
(n
2
1)
2

J
1
J

J
2
J

2
J

2
< j
n
2
2

2
J
2
[j
2
2
(J
o
); j
n
2
2
2

2
J

2
>< j
n
2
2

2
J
2
[j
2
2
(J
o
); j
n
2
2
2

2
J

2
> . (24.16)
The contribution from the intermediate state [ b > is obtained by application of Eq.
[13.46]
TBTDS
b
=
(),(1,2,1,2)
(2J
o
+ 1)n
1
n
2
_
(2J
1
+ 1)(2J
2
+ 1)(2J

1
+ 1)(2J

2
+ 1)

1
J

2
J

2
< j
n
1
1

1
J
1
[j
n
1
1
1

1
J

1
>< j
n
2
2

2
J
2
[j
n
2
1
2

2
J

2
>
< j
n
1
1

1
J

1
[j
n
1
1
1

1
J

1
>< j
n
2
2

2
J

2
[j
n
2
1
2

2
J

2
>

(2J

+ 1)
_

_
J
1
J
2
J
J

1
J

2
J

j
1
j
2
J
o
_

_
_

_
J

1
J

2
J
J

1
J

2
J

j
1
j
2
J
o
_

_
. (24.17)
24 TWO-BODY TRANSITION OPERATORS AND THE TBTD 273
There are two types of o-diagonal two-orbital terms. One in which the nal
states have partitions which dier from the initial state by the change of one particle
< f

[=< (k
n
1
1
1

1
J
1
)(k
n
2
+1
2

2
J
2
)J [,
and another in which the nal states have partitions which dier from the initial state
by the change of two particles
< f

[=< (k
n
1
2
1

1
J
1
)(k
n
2
+2
2

2
J
2
)J [ .
In analogy with the derivation given above for the diagonal case, explicit formulaes
for the TBTDS can be derived. The result for the case [ f

> is given in [1]. Although

the general analytical results for the many-orbital case become quite complicated, [2],
[3] they are easily obtained numerically by using an M-scheme shell-model code or a
J-projection shell-model code such as OXBASH [4].
24.4 Sample calculations for interaction energies
As a sample application, the interaction energy for the (j = 5/2)
3
conguration will
be calculated. This is obtained from Eqs. (24.12) and (24.14) by making use of the
relationship between one- and two-particle CFP as given by Eq. [23.14] together with
the one-particle CFP in Table [20.1]. For J = 9/2 one obtains
TBTDS(J = 9/2, J
o
= 2) =
9
14
,
and
TBTDS(J = 9/2, J
o
= 4) =
33
14
,
and thus
< j
3
, J = 9/2 [ V [ j
3
, J = 9/2 >=
9
14
< j
2
, J
o
= 2 [ V [ j
2
, J
o
= 2 >
+
33
14
< j
2
, J
o
= 4 [ V [ j
2
, J
o
= 4 > .
As a check, note that for a constant two-body interaction, V = C
< j
2
, J
o
= 2 [ C [ j
2
, J
o
= 2 >=< j
2
, J
o
= 4 [ C [ j
4
, J
o
= 4 >= C,
giving
< j
3
, J = 9/2 [ C [ j
3
, J = 9/2 >= 3C =
n(n 1)C
2
,
where n = 3.
REFERENCES 274
As an example of the application for the two-orbital conguration, the interaction
energy for the conguration
[ [(5/2)
2
, J
1
= 4], [(1/2)
1
, j
2
= 1/2], J = 9/2 >
is obtained with TBTDS
a
= 1 (J

1
= 0), TBTDS
c
= 0, and
TBTDS
b
= 18(2J
o
+ 1)
_
5/2 4 5/2
1/2 J
o
9/2
_
,
giving
< J = 9/2 [ V [ J = 9/2 >=< (5/2)
2
, 4 [ V [ (5/2)
2
, 4 >
+0.17 < 5/2, 1/2, 2 [ V [ 5/2, 1/2, 2 >
+1.83 < 5/2, 1/2, 3 [ V [ 5/2, 1/2, 3 > .
References
[1] I. Talmi, Simple Models of Complex Nuclei, (Harwood Academic Publishers,
1993).
[2] J. B. French, E. C. Halbert, J. B. McGrory and S. S. M. Wong, Advances in
Nuclear Physics 3, 193 (1969).
[3] J. Q. Chen, A. Novoselsky, M. Valliers and R. Gilmore, Phys. Rev. C 39, 1088
(1989).
[4] B. A. Brown, et al., the computer code OXBASH.
25 ELECTROMAGNETIC TRANSITIONS 275
25 Electromagnetic transitions
25.1 Operators and transition rates
The interaction of the electromagnetic eld with the nucleons can be expressed in
terms of a sum of electic and magnetic multipole operators with tensor rank
O =

,
[O(E)

+O(M)

]. (25.1)
The total rate for a specic set of states and a given operator is given by:
T
i,f,
=
_
8( + 1)
[(2 + 1)!!]
2
__
k
2+1
h
_
B(i f), (25.2)
where k is the wave-number for the electromagnetic transition of energy E

given by:
k =
E

hc
=
E

197 MeV fm
. (25.3)
The last factor in Eq. (25.2) is referred to as a reduced transition probability B
dened by:
B(i f) =
[< J
f
[[O()[[J
i
>[
2
(2J
i
+ 1)
. (25.4)
With our denition of the reduced matrix element,
[< J
f
[[O()[[J
i
>[
2
=[< J
i
[[O()[[J
f
>[
2
. (25.5)
B depends upon the direction of the transition by the factor of (2J
i
+ 1). For elec-
tromagnetic transitions J
i
is that for the higher-energy initial state. But in Coulomb
excitation the initial usually taken as the ground state, and one can use the notation
B() for this situation.
The electric transition operator given by:
O(E) = r

( r) e
tz
e, (25.6)
were Y

are the spherical harmonics. Gamma transitions with =0 are forbidden

because the photon must carry o at least one unit of angular momentum. The e
tz
are the electric charges for the proton and neutron in units of e. For the free-nucleon
charge we would take e
p
= 1 and e
n
= 0, for the proton and neutron, respectively.
Although the bare operator acts upon the protons, we will keep the general expression
in terms of e
tz
in order to incorporate the eective charges for the proton and
25 ELECTROMAGNETIC TRANSITIONS 276
neutron, which represent the center-of-mass corrections and the average eects of the
renormalization from wavefunction admixtures outside the model space.
The magnetic transition operator is given by:
O(M) =
_

2g

tz
( + 1)
+s g
s
tz
_

[r

( r)]
N
=
_
(2 + 1)
_
[Y
1

2g

tz
( + 1)
+ [Y
1
s ]

g
s
tz
_
r
1

N
, (25.7)
where
N
is the nuclear magneton,

N
=
e h
2m
p
c
= 0.105 efm, (25.8)
and where m
p
is the mass of the proton. The g-factors g

tz
and g
s
tz
are the orbital and
spin g-factors for the proton and neutron, respectively. The free-nucleon values for
the g-factors are g

p
= 1, g

n
= 0, g
s
p
= 5.586 and g
s
n
= 3.826. We may use eective
values for these g-factors to take into account the truncation of the model space.
The most probable types of transitions are E1, E2 and M1. The E1 transition
operator is given by Eq. (25.6) with =1:
O(E1) = r Y
(1)

( r) e
tz
e =

3
4
r e
tz
e, (25.9)
The E2 transition operator is given by Eq. (25.6) with =2:
O(E2) = r
2
Y
(2)

( r) e
tz
e, (25.10)
The M1 transition operator is given by Eq. (25.7) with =1 and Y
0
= 1/

4 :
O(M1) =

3
4
[

tz
+s g
s
tz
]
N
. (25.11)
The selection rules are given by the triangle condition for the angular momenta
in Eq. (25.5), (J
i
, J
f
, ). The electromagnetic interaction conserves parity, and
the elements of the operators for E and M can be classied according to their
transformation under parity change:
POP
1
=
O
O. (25.12)

O
= (1)

for Y

,
O
= 1 for the vectors r,

and p, and
O
= +1 for pseudo
vectors

l = r p and . For a given matrix element we have:
<
f
[ O [
i
>=<
f
[ P
1
POP
1
P [
i
>=
i

O
<
f
[ O [
i
> . (25.13)
25 ELECTROMAGNETIC TRANSITIONS 277
The matrix element will vanish unless
i

O
= +1. Thus the transitions are divided
into two classes, the ones which do not change parity change
i

f
= +1 which go by
the operators with
O
= +1:

f
= +1 for M1, E2, M3, E4 . . . , (25.14)
and the ones which do change parity change
i

f
= 1 which go by the operators
with
O
= 1:

f
= 1 for E1, M2, E3, M4 . . . . (25.15)
25.2 Moments in terms of electromagnetic operators
The operator for electromagnetic moment can be expressed in terms of the electromag-
netic transition operators. By the parity selection rule of Eq. (25.14), the moments
are nonzero only for M1, E2, M3., E4, . . .. The most common are:
=

4
3
< J, M = J [ O(M1) [ J, M = J >
=

4
3
_
J 1 J
J 0 J
_
< J[[O(M1)[[J >, (25.16)
and
Q =

16
5
< J, M = J [ O(E2) [ J, M = J >
=

16
5
_
J 2 J
J 0 J
_
< J[[O(E2)[[J > . (25.17)
25.3 Nuclear matrix elements
Electromagnetic transitions and moments depend upon the reduced nuclear matrix
elements < f[[O()[[i >. With the formalism of Sec. 23, these can be expressed as a
sum over one-body transition densities times single-particle matrix elements:
< f[[O()[[i >=

kk

OBTD(fik

) < k

[[O()[[k

>, (25.18)
where the OBTD is given by
OBTD(fik

) =
< f[[[a
+
k
a
k

[[i >
_
(2 + 1)
. (25.19)
25 ELECTROMAGNETIC TRANSITIONS 278
The labels i and f are a short-hand notation for the initial and nal state quantum
numbers (n
i
J
i
) and (n
f
J
f
), respectively. Thus the problem is divided into two
parts, one involving the nuclear structure dependent one-body transition densities
OBTD, and the other involving the reduced single-particle matrix elements (SPME).
The SPME for E operator of (25.6) is given by:
< k
a
[[O(E)[[k
b
>= (1)
ja+1/2
[1 + (1)
a++
b
]
2

(2j
a
+ 1)(2 + 1)(2j
b
+ 1)
4
_
j
a
j
b
1/2 0 1/2
_
< k
a
[r

[ k
b
> e
tz
e. (25.20)
The SPME for the spin part of the magnetic operator of Eq. (25.7) is:
< k
a
[[O(M, s)[[k
b
>=
=
_
(2 + 1) < j
a
[[[Y
1
s ]

[[j
b
>< k
a
[r
1
[ k
b
> g
s
tz

N
=
_
(2 + 1)
_
(2j
a
+ 1)(2j
b
+ 1)(2 + 1)
_

a
1/2 j
a

b
1/2 j
b
1 1
_

_
<
a
[[Y
1
[[
b
>< s[[s[[s >< k
a
[r
1
[ k
b
> g
s
tz

N
, (25.21)
where
< s[[s[[s >=
_
3/2 . (25.22)
The SPME for the orbital part of the magnetic operator of (25.7) is:
< k
a
[[O(M, )[[k
b
>=
=
_
(2 + 1)
+ 1
< j
a
[[[Y
1

[[j
b
>< k
a
[r
1
[ k
b
> g

tz

N
=
_
(2 + 1)
+ 1
(1)
a+1/2+j
b
+
_
(2j
a
+ 1)(2j
b
+ 1)

_

a

b

j
b
j
a
1/2
_
<
a
[[[Y
1

[[
b
>< k
a
[r
1
[ k
b
> g

tz

N
, (25.23)
where
<
a
[[[Y
1

[[
b
>= (1)
+a+
b
_
(2 + 1)
b
(
b
+ 1)(2
b
+ 1)

_
1 1

b

a

b
_
<
a
[[Y
1
[[
b
>, (25.24)
25 ELECTROMAGNETIC TRANSITIONS 279
with
<
a
[[Y
1
[[
b
>= (1)
a

(2
a
+ 1)(2
b
+ 1)(2 1)
4
_

a
1
b
0 0 0
_
. (25.25)
For the M1 operator of (25.11) the radial matrix element is:
< k
a
[r
0
[ k
b
>=
na,n
b
, (25.26)
and the SPME simplify to:
< k
a
[[O(M1, s)[[k
b
>=

3
4
< j
a
[[s [[j
b
>
na,n
b
g
s
tz

N
=

3
4
(1)
a+ja+3/2
_
(2j
a
+ 1)(2j
b
+ 1)
_
1/2 1/2 1
j
b
j
a

a
_
< s[[s [[s >
a,
b

na,n
b
g
s
tz

N
, (25.27)
where
< s[[s [[s >=
_
3/2 ,
and
< k
a
[[O(M1, )[[k
b
>=

3
4
< j
a
[[

[[j
b
>
na,n
b
g

tz

N
=

3
4
(1)
a+j
b
+3/2
_
(2j
a
+ 1)(2j
b
+ 1)
_

a

b
1
j
b
j
a
1/2
_
<
a
[[

[[
b
>
na,n
b
g

tz

N
, (25.28)
where
<
a
[[

[[
b
>=
a,
b
_

a
(
a
+ 1)(2
a
+ 1) .
Thus the M1 operator can connect only a very limited set of orbits, namely those
which have the same n and values.
25.4 Applications to simple situations
25.4.1 Closed shell plus one particle
For a closed shell plus one particle one nds that OBTD=1 and the only term con-
tributing to the sum (for >0) comes from the transition between two specic particle
states with J = j (since J
c
= 0)
< J
f
= j
f
[[O()[[J
i
= j
i
>=< k
f
[[O()[[k
i
>, (25.29)
25 ELECTROMAGNETIC TRANSITIONS 280
Table (25.1). Coecients C for k
2
E2 transitions.
transition (3/2)
2
(5/2)
2
(7/2)
2
(9/2)
2
(11/2)
2
(13/2)
2
2 0 0.800 0.914 0.952 0.970 0.979 0.985
4 2 0.630 0.950 1.114 1.207 1.265
6 4 0.433 0.771 0.990 1.132
8 6 0.308 0.612 0.841
10 8 0.229 0.491
12 10 0.177
and the reduced transition probability for this cases is:
B() =
[< k
f
[[O()[[k
i
>[
2
(2j
i
+ 1)
. (25.30)
25.4.2 Single-orbit congurations
For a closed shell plus n particles in a single state k these expressions (for >0)
reduce to:
< k
n
,
f
, J
f
[[O()[[k
n
,
i
, J
i
>= OBTD(fik) < k[[O()[[k >, (25.31)
OBTD(fik) = n
_
(2J
f
+ 1)(2J
i
+ 1)

J
(1)
J
f
++J+j

_
J
i f

j j J
_
< j
n

f
J
f
[j
n1
J >< j
n

i
J
i
[j
n1
J > . (25.32)
For n = 1 this simplies to the equivalent to Eq. (25.29)
< J
f
= j[[O()[[J
i
= j >=< k[[O()[[k >, (25.33)
which is the reduced matrix element which can be used in Eqs. (25.16) and (25.17)
for the single-particle moments of the state k.
For n = 2, the CFP are unity and the sum only contains J = j in which case Eq.
(25.32) simplies to
OBTD(fik) = (1)
J
f
++1
n
_
(2J
f
+ 1)(2J
i
+ 1)
_
J
i
J
f

j j j
_
. (25.34)
25 ELECTROMAGNETIC TRANSITIONS 281
The reduced transition rate becomes:
B() = n
2
(2J
f
+ 1)
_
J
i
J
f

j j j
_
2
[< k[[O()[[k >[
2
. (25.35)
= n
2
(2J
f
+ 1)(2j + 1)
_
J
i
J
f

j j j
_
2
[< k[[O()[[k >[
2
(2j + 1)
= C(J
i
, J
f
, j)
[< k[[O()[[k >[
2
(2j + 1)
, (25.36)
where the last line is written in terms of a coecient which depends upon J
i
and J
f
times a reduced single-particle transition rate. These coecients for some j value
are given in Table 1.
26 ALLOWED BETA DECAY 282
26 Allowed beta decay
The study of nuclear beta decay provides information both about the nature of the
weak interaction and about the structure of nuclear wave functions. The types of
beta decay can be classied by the angular momenta carried away by the electron and
neutrino. The most important are those for =0 which are referred to as allowed
beta decay. There are two type of allowed beta decay Fermi (F) and Gamow-
Teller (GT). The operator associated with Fermi decay is proportional to the isospin
raising and lowering operator. As such it can only connect isobaric analogue states
and it provides an exacting test of isospin conservation in the nucleus. The operator
associated with Gamow-Teller decay also contains the nucleon spin operator. Since
the total spin S is not a good quantum number, Gamow-Teller beta decay goes in
general to many nal states and provides a sensitive test of shell-model conguration
mixing in the nucleus.
26.1 Formulation of allowed beta decay
The allowed beta decay rate T between a specic set of initial and nal states is given
by
T
i,f
= (f/K
o
)
_
g
2
V
B
i,f
(F

) +g
2
A
B
i,f
(GT

)
_
, (26.1)
where f is dimensionless three-body phase-space factor which depends upon the beta-
decay Q value, and K
o
is a specic combination of fundamental constants
K
o
=
2
3
h
7
m
5
e
c
4
= 1.8844 10
94
erg
2
cm
6
s. (26.2)
The refer to

decay of nucleus (A
i
, Z
i
) into nucleus (A
i
, Z
i
1). The weak-
interaction vector (V ) and axial-vector (A) coupling constants for the decay of neutron
into a proton are denoted by g
V
and g
A
, respectively.
The total decay rate for a given initial state is obtained by summing the partial
rates over all nal states
T =

f
T
if
, (26.3)
with the branching fraction to a specic nal state
b
if
=
T
if
T
. (26.4)
Beta decay lifetime are usually given in terms of the half-life with a total half-life of
T
1/2
= ln(2) =
ln(2)
T
. (26.5)
26 ALLOWED BETA DECAY 283
The partial half-life for a particular nal state will be denoted by t
1/2
t
1/2
=
T
1/2
b
if
. (26.6)
Historically one combines the partial half-life for a particular decay with the
calculated phase-space factor f to obtain from Eq. (26.1) an ft value given by
ft
1/2
=
C
[B(F

) + (g
A
/g
V
)
2
B(GT

)]
(26.7)
where
C =
ln(2) K
o
(g
V
)
2
(26.8)
One often compiles the allowed beta decay rates in terms of a logft which stands
for log
10
of the ft
1/2
value.
The values of the coupling constants for Fermi decay, g
V
, and Gamow-Teller
decay, g
A
, in the combinations in which they appear in Eq. (26.7) are obtained as
follows. For a 0
+
0
+
nuclear transition B(GT) = 0, and for a transition between
T = 1 analogue states with B(F) = 2 Eq. (26.7) reduces to
C = 2t
1/2
f. (26.9)
The partial half-lives and Q values for several 0
+
0
+
analogue transitions have
been measured to an accuracy of about one part in 10000. With the phase space
factors discussed in Sec. 4.3.3 (including the correction
V
), one obtains [1]
C = 6170(4) (26.10)
This result, which together with the value of K
o
in Eq. (26.2), can be used with Eq.
(26.8) to obtain g
V
.
At the quark level g
V
= g
A
. But for nuclear structure we use the value obtained
from the neutron to proton beta decay [2]
[ g
A
/g
V
[= 1.261(8). (26.11)
26.2 Operators for allowed beta decay
26.2.1 Fermi decay
The operator for Fermi beta decay in terms of sums over the nucleons is
O(F

) =

k
t
k
. (26.12)
26 ALLOWED BETA DECAY 284
The matrix element is
B(F) =[< f [ T

[ i >[
2
, (26.13)
where
T

k
t

(26.14)
is the total isospin raising and lowering operator for total isospin constructed out of
the basic nucleon isospin raising and lowering operators
t

[ n >=[ p >, t

[ p >= 0,
and
t
+
[ p >=[ n >, t
+
[ n >= 0. (26.15)
The matrix elements obey the triangle conditions J
f
= J
i
(J = 0). The Fermi
operator has
O
= +1, and thus the initial and nal nuclear states must have
i

f
=
+1 for the matrix element to be nonzero under the parity transform.
When isospin is conserved the Fermi matrix element must obey the isospin tri-
angle condition T
f
T
i
(T = 0), and the Fermi operator can only connect isobaric
analogue states. For

decay
T

[
i
, J
i
, M
i
, T
i
, T
zi
>
=
_
(T
i
(T
i
+ 1) T
zi
(T
zi
1) [
i
, J
i
, M
i
, T
i
, T
zi
1 >, (26.16)
and
B(F

) =[<
f
, J
f
, M
f
, T
f
, T
zi
1 [ T

[
i
, J
i
, M
i
, T
i
, T
zi
>[
2
= [T
i
(T
i
+ 1) T
zi
(T
zi
1)]

f
,

J
i
,J
f

M
i
,M
f

T
i
,T
f
. (26.17)
For
+
we have
B(F
+
) =[<
f
, J
f
, M
f
, T
f
, T
zi
+ 1 [ T
+
[
i
, J
i
, M
i
, T
i
, T
zi
>[
2
= [T
i
(T
i
+ 1) T
zi
(T
zi
+ 1)]

f
,

J
i
,J
f

M
i
,M
f

T
i
,T
f
. (26.18)
For neutron-rich nuclei (N
i
> Z
i
) we have T
i
= T
zi
and thus
B(F

)(N
i
> Z
i
) = 2T
zi
= (N
i
Z
i
)

f
,

J
i
,J
f

M
i
,M
f

T
i
,T
f
, (26.19)
and
B(F
+
)(N
i
> Z
i
) = 0. (26.20)
For proton-rich nuclei (Z
i
> N
i
) we have T
zi
= T
i
and thus
B(F
+
)(Z
i
> N
i
) = 2T
zi
= (Z
i
N
i
)

f
,

J
i
,J
f

M
i
,M
f

T
i
,T
f
, (26.21)
and
B(F

)(Z
i
> N
i
) = 0. (26.22)
26 ALLOWED BETA DECAY 285
26.2.2 Gamow-Teller decay
The operator for Gamow-Teller beta decay in terms of sums over the nucleons is
O(GT

) =

k
t
k
. (26.23)
The reduced matrix elements is
B
i,f
(GT

) =
[< f[[O(GT

)[[i >[
2
(2J
i
+ 1)
=
[M
i,f
(GT

)]
2
(2J
i
+ 1)
, (26.24)
where
M
i,f
(GT

) =< f[[O(GT

)[[i > (26.25)

The matrix elements are reduced in orbital space and the (2J
i
+1) factor comes from
the average over initial M
i
states. The magnitude of reduced matrix element M(GT)
does not depend on the direction of the transition, i.e.,
[ M(GT, a b) [=[ M(GT, b a) [ (26.26)
whereas
B(GT, a b) =
(2J
b
+ 1)
(2J
a
+ 1)
B(GT, b a). (26.27)
The matrix elements obey the triangle (J
i
, j
i
, J = 1). The Gamow-Teller operator
has
O
= +1, and thus the initial and nal nuclear states must have
i

f
= +1
for the matrix element to be nonzero under the parity transform. When isospin
is conserved the Gamow-Teller matrix elements obey the isospin triangle condition
(T
f
, T
i
, T = 1).
In second-quantized form the GT

operator has the form

O(GT

) =

< [ t

[ > a
+
,p
a
,n
, (26.28)
where a
,n
destroys a neutron in state and a
+
,p
creates a proton in state . The
J-coupled form is
O(GT

) =

kak
b
< k
a
, p[[t

[[k
b
, n >
[a
+
ka,p
a
k
b
,n
]

_
(2 + 1),
(26.29)
where =1 for the GT operator. The reduced transition probability for the transition
from an initial state i to a nal state f is given by
B(GT

) =

kak
b
< k
a
, p[[t

[[k
b
, n > OBTD(k
a
, k
b
, f, i), (26.30)
26 ALLOWED BETA DECAY 286
where
OBTD(k
a
, k
b
, f, i) =
< f[[[a
+
ka,p
a
k
b
,n
]

[[i >
_
(2 + 1)
(26.31)
The analogous equations for GT
+
are
O(GT
+
) =

< [ t
+
[ > a
+
,n
a
,p
, (26.32)
where a
,p
destroys a proton in state and a
+
,n
creates a neutron in state . The
J-coupled form is
O(GT
+
) =

kak
b
< k
a
, n[[t
+
[[k
b
, p >
[a
+
ka,n
a
k
b
,p
]

_
(2 + 1),
(26.33)
and the reduced transition probability is
B(GT
+
) =

kak
b
< k
a
, n[[t
+
[[k
b
, p > OBTD(k
a
, k
b
, f, i), (26.34)
where
OBTD(k
a
, k
b
, f, i) =
< f[[[a
+
ka,n
a
k
b
,p
]

[[i >
_
(2 + 1)
(26.35)
The reduced single-particle matrix elements are given by
< k
a
, p[[t

[[k
b
, n >=< k
a
, n[[t
+
[[k
b
, p >= 2 < k
a
[[s [[k
b
>, (26.36)
where the matrix elements of s are given by
< k
a
[[s [[k
b
>=< j
a
[[s [[j
b
>
na,n
b
= (1)
a+ja+3/2
_
(2j
a
+ 1)(2j
b
+ 1)
_
1/2 1/2 1
j
b
j
a

a
_
< s[[s [[s >
a,
b

na,n
b
,
(26.37)
with
< s[[s [[s >=
_
3/2 ,
The matrix elements of s has the selection rules
a,
b
and
na,n
b
. Thus the orbits
which are connect by the GT operator are very selective; they are those in the same
major oscillator shell with the same value. The matrix elements such as 1s
1/2
0d
3/2
which have the allowed j coupling but are zero due to the coupling are called
-forbidden matrix elements.
26 ALLOWED BETA DECAY 287
26.3 Sum rules
Sum rules for Fermi and Gamow-Teller matrix elements can be obtained. The sum
rule for Fermi is obtained from the sum

f
[B
i,f
(F

) B
i,f
(F
+
)] =

f
_
[< f [ T

[ i >[
2
[< f [ T
+
[ i >[
2
_
The nal states f in the T

matrix element go with the Z

f
= Z
i
+ 1 nucleus and
those in the T
+
matrix element to with the Z
f
= Z
i
1 nucleus. One can explicitly
sum over the nal states to obtain

f
[< i [ T
+
[ f > < f [ T

[ i > < i [ T

[ f > < f [ T
+
[ i >]
=< i [ T
+
T

T
+
[ i >=< i [ 2T
z
[ i >= (N
i
Z
i
). (26.38)
The sum rule for Gamow-Teller is obtained as follows.

f,
[< f [

k,
t
k
[ i >[
2

f,
[< f [

k,
t
k+
[ i >[
2
=

f,
< i [

k,
t
k+
[ f > < f [

,
t
k

[ i >

f,
< i [

k,
t
k
[ f > < f [

,
t
k

+
[ i >
=

_
< i [
_

k,
t
k+
__

,
t
k

k,
t
k
__

,
t
k

+
_
[ i >
_
=

< i [

2
k,
[t
k+
t
k
t
k
t
k+
] [ i >= 3 < i [

k
[t
k+
t
k
t
k
t
k+
] [ i >
= 3 < i [ T
+
T

T
+
[ i >= 3 < i [ 2T
z
[ i >= 3(N
i
Z
i
). (26.39)
We have used the fact that
2
x
=
2
y
=
2
z
= 1. When k ,= k

the operators commute

and cancel. These sum rules hold for each M
i
value and thus also hold for B(F) and
B(GT) when we take an average over the M
i
values. Thus

f
[B
i,f
(F

) B
i,f
(F
+
)] = (N
i
Z
i
), (26.40)
and

f
[B
i,f
(GT

) B
i,f
(GT
+
)] = 3(N
i
Z
i
). (26.41)
The sum-rule for the Fermi matrix elements applies even when isospin is not
conserved. When isospin is conserved we recover the results given by Eqs. (26.17)
26 ALLOWED BETA DECAY 288
and (26.18). For and N > Z we usually have T
i
= T
zi
which means that B(F
+
) = 0
and we can use Eq. (26.40) to obtain B(F

) = (N
i
Z
i
) (the same as Eq. (26.19))
for the transition to the isobaric analogue state. For N = Z(T
zi
= 0) and T
i
= 0
we have from Eqs. (26.17) and (26.18) B(F
+
) = B(F

) = 0, and for T
i
= 1 we have
B(F
+
) = B(F

) = 2. Fermi transitions which would be zero if isospin is conserved

are called isospin-forbidden Fermi transitions.
When N > Z there are some situations where one has B(GT
+
) = 0, and then
from (26.41) we have B(GT

) = 3(N
i
Z
i
). In particular for the

decay of the
neutron we have B(F

) = 1 and B(GT

) = 3.
26.4 Eective operators for Gamow-Teller matrix elements
There are several reasons why the free-nucleon calculations may dier from experi-
ment. In reality the nuclear wave functions are more complicated than the theoretical
model we use in that they incorporate nucleon degrees of freedom beyond the sd-shell
space. In addition, non-nucleonic degrees of freedom involving the delta isobars and
mesons in the nucleus may be important in the observed rates. The calculation of
the corrections corresponding to these processes have been the subject of many the-
oretical investigations. Most recently these were carried out by Towner and Khanna
[3]. For references and comparisons to the earlier works, see [3] and [4]
Since the factor g
2
A
appears in front of B(GT) one might parameterize the eec-
tive matrix in terms of an eective g
A
value one speaks of a renomalization of the
axial vector current in the nuclear medium. Equivalently one can express the renor-
malization in terms of corrections to the GT operator. It is convenient to express the
eective GT operator as:
O(GT

)
eff
= (s +O
GT
)t

, (26.42)
where
O
GT
=
s
s +

+
p
p(s d), (26.43)
and where
p = (8)
1/2
[Y
(2)
(r) s]
(1)
. (26.44)
The coecients characterize the renormalizations which are needed when working
within the sd-shell model space. The reduced single-particle matrix elements for the
individual operator components s,

and p are given in [4].

and
p
turn out to be relatively small compared to
s
. Thus to a good approx-
imation we may take the eective operator as just:
O(GT

)
eff
= (1 +
s
)st

, (26.45)
REFERENCES 289
A small
p
value means that the -forbidden 0d
3/2
1s
1/2
single-particle matrix
element is not zero but is relatively small.The weak branch for the beta decay of the
39
Ca 3/2
+
state to the 1/2
+
state in
39
K has been measured [5]. In the sd shell model
these levels are just the 0d
3/2
and 1s
1/2
single-particle states and the ft value for this
transition thus directly provides a value for the -forbidden matrix element or for the
parameter
p
(s-d). From the A = 39 datum a value of [
p
(s d) [= 0.017 0.003 is
obtained. The
p
value extracted from a t to many sd-shell GT decays [6] is
p
=
+0.021(8). The value of
p
is positive as expected theoretically [3].
References
[1] D. H. Wilkinson, A. Gallman and D. E. Alburger, Phys. Rev. C18, 401 (1978)
[2] D. H. Wilkinson, Nucl. Phys. A377, 474 (1982)
[3] I. S. Towner and F. C. Khanna, Nucl. Phys. A399, 334 (1983).
[4] B. A. Brown and B. H. Wildenthal, Phys. Rev. C28, 2397 (1983).
[5] E. C. Adelberger, J. L. Osborne, H. E. Swanson and B. A. Brown, Nucl. Phys. A
417, 269 (1984).
[6] B. A. Brown and B. H. Wildenthal, Atomic Data Nucl. Data Tables 33, 347
(1985). The B(GT) as dened in the present work is related to the M(GT) of this
references by B(GT) = [M(GT)/1.251]
2
/(2J
i
+ 1).