Download as pdf or txt
Download as pdf or txt
You are on page 1of 44

ANNUAL

REVIEWS Further
Quick links to online content
Annu.R�.Nuci'. Part. Sci. 1995.45:591-634
Copyright © 1995 by Annual R�iews Inc. All rights reserved

NUCLEAR HALOS

P. G. Hansenl and A. S. Jensen


Access provided by 2802:8010:470a:5b00:ddfc:f069:d425:9566 on 11/26/22. For personal use only.
Annu. Rev. Nucl. Part. Sci. 1995.45:591-634. Downloaded from www.annualreviews.org

Institute of Physics and Astronomy, Aarhus University, DK-8000 Aarhus C,


Denmark

B. Jonson
Department of Physics, Chalmers University of Technology, S-412 96
Goteborg, Sweden

CONTENTS
1. INTRODUCTION . . . . ..
. . . . . . .. . .
. . . . . . . . . . . . . ... . " ..................... 592
2. LOOKING BACK . . . . . . . . . .. . . . . . . . . . . . . . . . .. .. . . . . . . . . . . . . . . . . . . . . . . . . . . 595
3. GENERAL PROPERTIES . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . 598
3.1 Two-Body Systems .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 599
3.2 Three·Body Systems . . . . . . .. . . .. . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 600
3. 3 Mul tibody Neutron Halos . .. . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 603
3. 4 Bound Excited S tates . . .. . . .
. . . . . . . . . . .. .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 604
4. STRUCTURE AND REACTIONS . . . . . . . . .. . . . . . . . .
. . . . . . . . .. . . . . . . . . . . . . . . . 605
4.1 Reaction Mechanisms . . . . . . ...
. . . . . . . . ... . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . 605
4. 2 The Olle-Neutron Halo: IlBe . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 612
4. 3 The Two-Neutron Halo: 1 1Li . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 617
4. 4 The Continuum Problem: IOLi .. .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 622
5. f3 DECAY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 626
5.1 Transitions Between Discre te States . . . 627
5. 2 {J D e/ayed Par ticles . . . . . . . . . . . . . .. . . .. .. .. . . .. . . . .. .. .. . . . . . . . . .. .. ..
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

- 629
6. CONCLUDING REMARKS . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 631

KEY WORDS: exotic nuclei, radioactive beams, fragmentation. momentum distributions,


f3 decay

ABSTRACT

After giving a brief historical overview, we use simple few-body models to


illustrate the: basic properties of halo states. Giant halos may exist in certain

1Current address: National Superconducting Cyclotron Laboratory and Department of Physics


and Astronomy, Michigan State University, East Lansing, Michigan 48824-1321.
591
0163-8998/9511201-0591$5.00
592 HANSEN, JENSEN & JONSON

excited levels, although this is still very hypothetical. The main part of the paper
is dedicated to the momentum spectra of fragments from breakup and to the f3
decay of halo states.

1. INTRODUCTION

A novel structural feature called the: neutron halo has been found in a number
Access provided by 2802:8010:470a:5b00:ddfc:f069:d425:9566 on 11/26/22. For personal use only.

of light, extremely neutron-rich nuclei. A neutron halo is basically a threshold


Annu. Rev. Nucl. Part. Sci. 1995.45:591-634. Downloaded from www.annualreviews.org

effect resulting from the presence of a bound state close to the continuum.
The combination of the low neutron separation energy and the short range of
the nuclear force allows the neutron (or a cluster of neutrons) to tunnel into the
space surrounding the nuclear core so that neutrons are present with appreciable
probability at distances much larger than the normal nuclear radius. In this very
open structure, simple few-body or cluster models will largely account for the
most general properties of nuclear halos. In this review, we illustrate this aspect
using simple two- and three-particle systems but for the most part do not discuss
the theory of nuclear models of the halos, which has already been extensively
reviewed in the literature. An exciting possibility suggested by simple analytical
solutions of three-body models is that some two-neutron halos may possess a
bound excited state of even larger spatial dimensions, perhaps of the order
of 100 fm.
The intimate linkage of the halo to a particle threshold implies that the general
level spectrum of the nucleus has little direct bearing on the structure of the halo
state. To study this structure, it is Mcessary to tum either to the static properties
of the halo or, more often, to processes in which it is created or destroyed. In
the following, the momentum distributions of fragments produced in reactions
of halo states are discussed in detail. A second example is f3 decay to (or from)
halo states. We approach these topics from an essentially experimental and
phenomenological point of view.
The simplest example of a halo nucleus is 11 Be, which to a good approxi­
mation may be viewed as a two-body system consisting of a neutron coupled
to the quadrupole-deformed lOBe core. Its two (only) bound states are both
halo states (see Figure O. For the s state, the unnormalized external part of
the wavefunction is asymptotically X (T) = exp( -leT), with a reciprocal decay
length given by

Ie = J2/1'snlh 1.

expressed in terms of the reduced mass J-L and the neutron separation energy
Sn. The large neutron-core distance, which is of the order 1 1K, indicates that
many properties of 11Be are determined essentially by the asymptotic part of
the wavefunction and depend little: on the details of the core structure. The halo
NUCLEAR HALOS 593
0,6

. .
504keV �
0,4 320k'V 1l2

I EI
Access provided by 2802:8010:470a:5b00:ddfc:f069:d425:9566 on 11/26/22. For personal use only.

o
0,2 112+
Annu. Rev. Nucl. Part. Sci. 1995.45:591-634. Downloaded from www.annualreviews.org

ll
- Be
N " ,
.......
.....
........ . . .
I


......,
0


.......

-0,2

Is, Sn =
0.50 MeV
-0,4 Op. Sn = 0.18 MeV

-0.6 � �����-k���

o 5 10 15 20 25
r (fm)
Figure 1 The level scheme of II Be (inset) and the wavefunction x(r) = rR(r) for the two bound
states Is and Op (dashed line), which have rms radii of 6.0 and 5.7 fm, respectively. The core
radius is 2.5 fm.

structure has several very characteristic experimental features, such as large


cross sections for electric and nuclear dissociation processes, but probably the
most striking of these is that reactions in which the halo system is broken up
are associated with extremely narrow momentum distributions of the fragments
(see the example in Figure 2). The narrow distributions may be qualitatively
viewed as a direct consequence of the Heisenberg uncertainty principle. Proton
halos are equally possible but less pronounced, owing to the strong influence
of the Coulomb barrier.
The most inlteresting halo states are those that have an even number of neu­
trons. For example, the at particle will bind two neutrons but not one, and
four but not three. Viewed superficially, this selectivity is merely a result of
594 HANSEN, JENSEN & JONSON

the pairing energy. However, the additional binding cannot be calculated as a


perturbation; rather, it emerges when the properties of the combined system are
considered.
The continuum states situated just above the neutron threshold are of great
importance in elucidating the structure of the halo, and they give rise to crucial
final-state interactions for some breakup channels. The situation is identical
to that in more conventional studie:s of nuclear structure: The single-particle
Access provided by 2802:8010:470a:5b00:ddfc:f069:d425:9566 on 11/26/22. For personal use only.

states in odd systems are the key to understanding paired structures. As prime
Annu. Rev. Nucl. Part. Sci. 1995.45:591-634. Downloaded from www.annualreviews.org

examples of these structures we US!! 11 Li and 6He.

50

40

()
....... 30


:E

20

10

o
o

Figure 2 The width r ofthe longitudinal momentum distribution from the (1!Li,9Li) reaction
as a function ofthe atomic number Z ofthe target (43, 87,109). The insets show the measured
distributions transformed to the coordinate system traveling with the beam for (a) a Be target with
E(IlLi) = 66 MeV!u and (b) an AI target with E(lILi) = 280 MeV!u. The curves show, for
comparison, the estimated distribution for the (12C, WC) reaction based on measurements by Kidd
et aI (82).
NUCLEAR HALOS 595

The theory of nuclear halos, treated sporadically here, has been covered by
Zhukov et al (1), who primarily describe three-body aspects, and by Bertulani
et al (2), who focus on reactions. The recent surge of interest in the halo has
its origin in new and very powerful experimental techniques, which are the
subject of the companion paper by Geissel et al (3) in the present volume. On­
line mass separation and the separation of fast recoils have also been discussed
in several other reviews (4-8), and general properties of light, neutron-rich
Access provided by 2802:8010:470a:5b00:ddfc:f069:d425:9566 on 11/26/22. For personal use only.

nuclei are given in Reference 9. Experimental results on the halo have been
Annu. Rev. Nucl. Part. Sci. 1995.45:591-634. Downloaded from www.annualreviews.org

reviewed by Riisager ( 1 0) and Tanihata (11) and have been treated in a number
of conference contributions (12-16). However, we remind the reader that the
halo phenomenon itself has ancient roots in nuclear physics as well as some
parallels in atomic and molecular physics.

2. LOOKING BACK

The deuteron, with its rms neutron-proton distance of about 4 fm (comparable


to the values of Figure 1), is the forerunner of all halo states. This nucleus led
to the discovery of the main mechanisms for the reactions of halos, Coulomb
dissociation (17, 1 8), stripping ( 1 9), and diffraction dissociation (20). The
combination of all three mechanisms was later extended to the medium-energy
range (21) and recently applied to halo nuclei (2).
The hypeltriton, which consists of a proton, a neutron, and a A, has long been
known to have a A halo. The corresponding A separation energy is estimated
to be 0.08 ::J: 0.02 MeV, and the system undergoes Coulomb dissociation in
collisions with heavy nuclei (22). There are no new data on this interesting
nucleus, but it is still being remembered by theorists (23-25).
The tails of wavefunctions of loosely bound states play an important role in
stellar nucleosynthesis, since the capture rates are low and the reactions take
place at very large distances (of the order of several tens of fm). Nuclear halos
therefore enhance the matrix element. A particularly illuminating example,
that of the 495-keV excited 112+ state of 17F (to which we return in Section
5 . 1 and Figure 1 2), has been investigated by Rolfs (26). He found that the
radiative capture of a proton on 160, which is responsible for the breakout from
the carbon-niitrogen-oxygen (CNO) cycle, proceeds from a continuum p state
to this excited level and is strongly enhanced by the proton tail of the bound
state arising from its proton separation energy of only 1 05 keV. The enhanced
radiative width in 17F is but one example showing that the appearance of a cluster
substructure jin a nucleus may relax the forbiddenness of EI transitions if the
center of mass does not coincide with the center of charge. The well-developed
halo systems have El transition strengths of essentially one single-particle unit.
Alhassid et al (27) noted that this effect comes into play, although at a more
596 HANSEN, JENSEN & JONSON

modest level, already for much more strongly bound molecular clusters such
as the dineutron in 180 or the a in 218Ra.
The first clear-cut case of a neutron halo was reported by Millener et al (28).
Their experiment confirmed that the EI transition probability between the two
bound states of IIBe (Figure 1) was surprisingly large. The lifetime of 166 ± 15
fs translates into an E l strength of 0 .36 Weisskopf units [defined without the
(21 + 1 ) statistical factor], which means that this y ray is, by far, the fastest
Access provided by 2802:8010:470a:5b00:ddfc:f069:d425:9566 on 11/26/22. For personal use only.

known El transition between bound states in nuclear physics. These authors


Annu. Rev. Nucl. Part. Sci. 1995.45:591-634. Downloaded from www.annualreviews.org

wrote, "the experimental B(E1) can be understood only if the radial single­
particle matrix elements are evaluated taking into account the actual binding
energies of the single-particle orbits... (and) harmonic-oscillator wavefunctions
fail most dramatically in I I B e." In an extension of this work, Uchiyama & Mori­
naga (29) used asymptotic external wavefunctions to illustrate the disappear­
ance of E 1 hindrance and discussed the ensuing enhancement of direct neutron
capture. This latter mechanism may be important for multiple neutron capture
in a thermonuclear explosion (30) lmd in astrophysics. In the recent rush of
activities connected with the halo, the link to direct neutron capture has tended
to be forgotten, but the problem was taken up again not long ago by Otsuka
et al (3 1).
The current wave of interest in nuclear halos was triggered by a series of
experiments by Tanihata et al (32), who in 1985 used radioactive ion beams of
790 MeVlu from the fragment separator at the Lawrence Berkeley Laboratory­
Bevalac to measure interaction cross sections of light nuclei. These authors
found surprisingly large values for the isotopes 6,8He and II U and interpreted
this finding in terms of a large deformation or a long tail in the matter distri­
bution. The effect becomes more pmnounced at lower beam energies (33, 34),
where the nucleon-nucleon cross sections are larger.
A combination of optical and t3-decay measurements (35) at the on-line mass
separator ISOLDE at CERN demonstrated that for 9Li and II Li, the magnetic
dipole and electric quadrupole moments (36), and hence the proton distribu­
tions, are very similar. This finding proved that the increase in radius arises
from a neutron tail, which is in agreemel!t with the observation (37) that the
halo is formed as a consequence of the low binding energy of the last neutron
pair. It was also suggested (37, 38) that the halo would be associated with large
Coulomb dissociation cross sections for reactions with heavy targets, an effect
that was soon observed (39). Another demonstration of the structure of halo
states was given by Blank et al (40), who showed that in contrast to the in­
teraction cross sections (34), which increase monotonically along the isotopic
sequence 8,9,IIU, the cross sections for reactions that change the number of
protons in the projectile remain constant as a function of mass.
NUCLEAR HALOS 597

The momentum distributions of fragments from the breakup of halo states


are an important tool in studying the structure of the halo. The first experiment
of this type: was performed by Kobayashi et al (4 1), who found a surprisingly
narrow transverse momentum distribution of9Li fragments from a 790-MeV/u
11
Li beam. Even narrower distributions were subsequently found at lower
energies (42, 43) for momenta of neutrons and 9Li fragments.
The terms halo and skin have held several different meanings in the past (see
Access provided by 2802:8010:470a:5b00:ddfc:f069:d425:9566 on 11/26/22. For personal use only.

e.g. 16), but in the context of exotic nuclei, the former now seems to be reserved
Annu. Rev. Nucl. Part. Sci. 1995.45:591-634. Downloaded from www.annualreviews.org

for the threshold phenomenon defined above. The possibility of a difference


between the! proton and neutron radii of stable nuclei has been studied for a
long time. According to Batty et al (44), such differences are small (of the
order 0. 1-0.2 fm). Larger effects are expected in radioactive nuclei. Tanihata
et al (45) have suggested using the term skin for a (dense) excess of neutrons
at the nuclear surface, typically extending about 1 fm beyond the protons. An
example of such a structure is 8He.
Other areas of quantum physics offer examples of structures that resemble
the nuclear halo (10, 14). Negative ions (see 46) share with the nuclear halo
marginal biJllding, the potential that decreases rapidly at large distances, and
the role played by correlations. The closest analogy to the one-neutron halo
is, however, provided by an electron bound in the field of an electric dipole,
which occurs if the dipole moment exceeds a certain limit (47). Only relatively
recently has atomic physics realized states of this nature, which may be of the
order of 100 A in radius. Lykke et al (48) observed these states in the form of
photodetachment resonances, and most recently a similar system was found for
which the halo is the ground state (49).
In the nuclear case, the halo nucleon has been squeezed out from the core
by the Pauli principle. Apparently analogous behavior has been observed in
solid-state physics (see 50) and references therein. In complexes of an alkali
atom with organic molecules called crown ethers, the valence electron, defined
by its spin density, predominantly is found far outside of all the atoms in the
complex.
The general interest in the three-body problem with weak short-range attrac­
tive forces has been discussed by Efimov (5 1), who points out that although
the three-body problem is usually considered too complex to be analyzed in
qualitative terms, there exists, in fact, for weakly bound systems a rather wide
class of tractable solutions. These solutions fulfill two conditions, namely, that
the size of the two-body system, represented by the magnitude of the s-wave
scattering length as, should be much larger than the range R of the force (this is
the condition for having a real or virtual state near zero energy) and that the ab­
solute energy of the three-body system should be much smaller than the Wigner
598 HANSEN, JENSEN & JONSON

limit, i.e. lEI «h2jmR2. These conditions are both met for halo nuclei. An
interesting possibility (52) to which we return in the next section is that some
three-body systems have one or seve:ral bound excited states of very large spa­
tial dimensions. The van der Waals force between He atoms would seemingly
allow molecular structures of this kind (53), and dimers and trimers of He have
been detected experimentally OP TOI!nnies, personal communication).
Access provided by 2802:8010:470a:5b00:ddfc:f069:d425:9566 on 11/26/22. For personal use only.

3. GENERAL PROPERTIES
Annu. Rev. Nucl. Part. Sci. 1995.45:591-634. Downloaded from www.annualreviews.org

Many-body nuclear theory, with its underlying framework rooted in the shell
model, has helped to further our understanding of the structure of the halo
states. The ordinary non-self-consistent, deformed shell model has been used
successfully to describe both ground states and high-spin states of "normal"
nuclei. However, the lack of self-consistency becomes critical when the struc­
ture changes rapidly as the dripline instability is approached. Improvements to
Hartree-Fock or to self-consistent de:nsity functional methods do not overcome
the basic problem. A better treatment of the continuum effects is achieved
with the Hartree-Fock-Bogoliubov (54) prescription, which shows a signifi­
cant change of shell structure when the dripline is approached and in which
correlations beyond pairing are ignored. All these mean field approximations
break down when the nuclear halo appears because the mean-field of the halo
nucleons differs from that of the core nucleons. In other words, the halo is too
dilute and contains too few nucleons to allow a mean-field description.
It is tempting to turn to the intera(:ting shell model, which in principle is cor­
rect, provided the interaction is conect. An overwhelming number of accurate
properties for the low-lying nuclear spectra are also obtained both for normal
nuclei and for nuclei near the dripline. However, treating the spatial extension
for halo nuclei remains problematk, since the shell model cannot accommo­
date in a natural way phenomena that require a precise knowledge of the remote
parts of the wavefunction. The neutron separation energy is therefore essential
for a correct description of the spatial extension. A generalization that includes
cluster aspects or correlations among groups of particles seems unavoidable,
since a complete and accurate description requires both adequate treatinent of
the core degrees of freedom and e:nough flexibility to include the necessary
correlations in the halo. The cluste:r orbital shell model (55), a translationally
invariant formulation with coordinates drawn from the core to the halo particle,
is at present the most promising method (56, 57), but the technical difficulties
increase substantially with the number of nucleons.
The complicated many-body models do not have firm predictive powers,
and experiment has been and at the time of this writing remains the prime
mover because the halo phenomenon is a delicately balanced threshold effect,
NUCLEAR HALOS 599

requiring a numerical precision that is best "put in by hand." A qualitative


assessment of the possibilities for new structures along the dripline is therefore
better obtained with simpler few-body models. The details of the many-body
system are then ignored, and focus is on few-body correlations, which in any
case are the key to our understanding of loosely bound and spatially extended
nuclear systems. Here we use schematic models to show various essential
connections. For large sizes and low binding energies, asymptotic properties
Access provided by 2802:8010:470a:5b00:ddfc:f069:d425:9566 on 11/26/22. For personal use only.

can be calculated. These properties are useful as reference units or landmarks


Annu. Rev. Nucl. Part. Sci. 1995.45:591-634. Downloaded from www.annualreviews.org

even if they are not realized in nature.

3.1 Two-Body Systems


Let us consider two particles interacting by a weak short-range potential. The
properties of possible bound states depend little on the details of the interaction,
provided the dominant part of the probability distribution lies outside the range
of the potenltial. The essence of the argument, which can be found in Reference
37, can be demonstrated by means of the asymptotic Yukawa wavefunction
defined by Equation 1 and leads to a mean square distance given by

(r 2 } ==
It r2exp( -2Kr)dr
00 ==
1 Tt2
== -- , 2.
J� exp( -2Kr)dr 2K2
-

4JLSn
which diverges as the separation energy (or K) decreases toward zero.
This result has been generalized to arbitrary radial moments, nonzero angular
momenta, and long-range potentials (58). When the binding energy approaches
zero, the normalized quantity (rV) converges for v < 21 1. For v > 21 1,
- -

it diverges as S;;v/2 for I == 0 and as S�21-'-V)/2 for I > O. Thus only states
with I == 0 and 1 have diverging mean square radii (v 2) in the limit of van­
=

ishing separation energy. The reason for this is that the probability of finding
the particle under the centrifugal barrier decreases exponentially with height
and width. Thus a bound state of given energy is pushed toward the attractive
pocket of the potential.
If both particles are charged, an additional repulsive Coulomb potential is
present. The resulting barrier is very thick owing to the long range (l/r) po­
tential, and 1he wavefunction is therefore confined even more than it would
be from the I:.:entrifugal barrier (l/r2). In fact, the thickness of the Coulomb
potential barrier is so large that all radial moments remain finite for all sep­
aration energies. The low-energy asymptotic large-distance behavior of the
radial wavefunction R(r) is easily derived from a Schrodinger equation, which
contains only the Coulomb potential and the kinetic energy operator (59, 60).
The solution is R(r) ex l Ir (r/rJ)'/4 exp(-2J17ij), where 1] Z,Z2e2 is the
=

product of the two charges. Extended proton halos therefore do not exist, al­
though a proton s state loosely bound to a light core (small charge) still may
600 HANSEN, JENSEN & JONSON

be significantly larger than normal nlilclear size. The maximum increase of the
rms radius is approximately a factor of two above normal values.
Families of two-body potentials with a given shape exhibit scaling properties,
i.e. all lengths in units of the potential range parameter R and all energies in
units of Tl2 //-LR2 lead to universal curves. In general, it is impossible to relate
the results of potentials of different shape in a global construction. However,
for short-range potentials and low en,ergies, a remarkably accurate scaling pre­
Access provided by 2802:8010:470a:5b00:ddfc:f069:d425:9566 on 11/26/22. For personal use only.

scription has been found (59, 61). The results for the lowest three angular
Annu. Rev. Nucl. Part. Sci. 1995.45:591-634. Downloaded from www.annualreviews.org

momenta are

3.

where RAn is the radius of a reference square well potential VAn defined as

f Vr3dr f VAnr3dr RAn2


=
-- 4.
f Vrdr = f VAnrdr 2 '

which is roughly related to the core mean square radius by (3/5)Rin = «(r2)core
+ ti), where ti r:::: 2-4 fm2 accounts for the range of the nucleon-nucleon interac­
tion. The first radial moment is chosen as normalization since the combination
decisive for the binding is VAnRin' as is well-known from the theory of the
deuteron.
A halo can be defined in terms of the probability, say >50%, of finding
the particle outside the range of the potential. An equivalent definition would
be provided by the requirement (r2) /Rin > 2, which for I = 0 and /-L = m
corresponds to snRin � 5 MeV fm2, as seen from the above scaling condition.

3.2 Th ree-Body Systems


Three particles interacting via shOlt-range two-body interactions can form a
variety of different structures (60). Leaving out all intrinsic structure, including
spins, a classification is sketched in Figure 3 with a notation indicating two
neutrons outside a core. We concentrate primarily on the Borromean region, a
term coined by Zhukov et al ( 1) to denote a bound three-body system for which
no binary subsystem is bound. This region exists for almost all potentials
independent of their shape (62). The most convenient coordinates for such
systems are the Jacobi coordinates, defined as the relative coordinates between
NUCLEAR HALOS 601
Access provided by 2802:8010:470a:5b00:ddfc:f069:d425:9566 on 11/26/22. For personal use only.
Annu. Rev. Nucl. Part. Sci. 1995.45:591-634. Downloaded from www.annualreviews.org

No bound < 2 -body

3 -body
jl
�___�__�____ � __ ��
2-

__�_
Od


Figure 3 Schematic classification of three-body states of the system (A + 2n) as a function of the
strengths of the two-body potentials. The thick curve separates the regions where the three-body
system is either Ibound or unbound, and the lines separate bound and unbound two-body systems.
In the Borromealll region ( I ), the three-body system is bound, but its two-body subsystems are
not.

two of the particles (x) and between their center of mass and the third particle
(y). The corresponding set of hyperspherical coordinates (p, a, Ox, Oy) is
defined by p = Jx2 + y2, the angles describing the directions Ox and Oy of x
and y, and by an angle ex defined by tgex = Ixl/lyl. The kinetic energy operator

(�+�� k2),
in these coordinates is then

T= _ � _
5.
2m dp2 p dp p2
602 HANSEN, JENSEN & JONSON

where the eigenvalues of the angular momentum operator f<.2 are given by
K(K + 4) in terms of the hypermoment K = 0,1, 2 . . .
The radial SchrOdinger equation at large distances p can for short-range
potentials be written as

(- d2 (K + 3/2)(K + 5/2) 2m )
+ E f (p) = 0, 6.
dp2 p2 - T1
Access provided by 2802:8010:470a:5b00:ddfc:f069:d425:9566 on 11/26/22. For personal use only.
Annu. Rev. Nucl. Part. Sci. 1995.45:591-634. Downloaded from www.annualreviews.org

which is of the same form as the two-body radial equation, for which the angular
momentum is now replaced by l* = K + 3/2. Thus, for vanishing separation
energy, (pV) is convergent for v < 2/* - 1 = 2K + 2, logarithmically divergent
for v = 2K + 2, and divergent as Si�+1-v/2) for v > 2(K + 1) . We see
that high K (or l*) confines the bound state to smaller distances because of the
repulsive effective angular momentum barrier. Even the lowest K (= 0) system,
which is comprised entirely of relative s states, has a finite effective centrifugal
barrier corresponding to l* = 3/2 (�:ee Equation 6). Thus the divergencies are
fewer and weaker for Borromean three-body systems than they are for two-body
systems. For two neutrons, the second radial moment converges for all K > °
and diverges logarithmically only �or K = 0, i.e. as In(S2nRinm/h2), where
S2n is the two-neutron separation energy.
The decisive divergence and the low-energy asymptotic properties are there­
fore determined by the lowest K in the expansion of the wavefunction. When
the dominating component corresponds to K larger than this minimum value,
the limiting behavior is reached only at very small energies.
A long-range potential such as thle Coulomb interaction between at least one
pair of particles changes the above conclusions. The longest-ranging potential
in Equation 6 will be inversely proportional to p and will therefore dominate
over the 1/p2 component for large p. The situation for the Borromean systems is
similar to that for a two-body system, and the asymptotic wavefunction is again
f ex (p/1J)1/4 exp(-2Jp/1J), whl!re YJ is now an effective Coulomb length
parameter. All radial moments are therefore again finite even for vanishing
separation energy.
For example, occurrence of a three-body halo could be defined by (rin)/R i n
> 2, where rAn is the distance between the neutron and the core. For K = 0,
this value is equivalent to s2nRin :::: 5-10 MeV fm2 , which is approximately the
same as for two-body systems. This coincidence is remarkable, since the two
systems are rather different, as evidenced by the divergence properties. Both
types of halos can therefore be expeGted to be equally abundant along the neutron
dripline. The necessary, but not sufficient, condition for occurrence of two- or
three-body halos can now be expre:ssed as the inequality SA2/3 :5 2 4 MeV-
involving the two- or three-body separation energy S and the mass number A.
NUCLEAR HALOS 603

3.3 Mu ltibody Neu tron H alos


Two pertinent questions arise. First, how is the transition from normal to
halo nuclear structure accomplished? Second, does halo structure exist with
more than two particles in the halo? The difference between the neutron and
proton densiities is directly related to the difference in the corresponding Fermi
energies. The resulting thickness of the so-called neutron skin increases with
the distance in the isotopic chain to the ,B-stable nuclei (63), and the neutron
Access provided by 2802:8010:470a:5b00:ddfc:f069:d425:9566 on 11/26/22. For personal use only.

density distribution is characterized by bulk and surface properties similar to


Annu. Rev. Nucl. Part. Sci. 1995.45:591-634. Downloaded from www.annualreviews.org

those of normal nuclei. In contrast, the neutron halo is much more dilute.
It is instmctive to start with a system of d spatial degrees of freedom. For
a two- and three-body system, d = 3 and 6, respectively, when the three
center-of-mass coordinates are removed. The centrifugal barrier potential is
then 1i2(d 1 )(d 3) /(Smp2), where p is the hyperspherical lengtb coordinate
- -

(61). Thus halos seem to be hindered by large dimensions. However, angular


correlations among the halo particles may supply enough energy to overcome
this difficulty. In 11 Li, for example, the last neutron in a resonance (positive
energy state) relative to the9Li core paid an energy Ene but simultaneously gained
the interaction energy vnn as a result of the attraction of the other neutron. The
larger degeneracy in the 10Li system and n neutrons in the halo would give a
total energy of nEne + (1/2)n(n l)vnn. A positive binding energy is then
-

obtained if 2Ene < -en l)vnn. The total energy gain obtained here by
-

counting all possible pairwise interactions is presumably an overestimate, and


a linear dependence may be more realistic. A positive binding energy results if
Ene < -Vnn, which is also entirely possible.
Another approach to multineutron clusters (64) is illustrated by the example
of 11Li, in which two neutrons in a specific unbound confi guration (a dineutron)
are bound in the field of the 9Li core. Let us assume that the first excited
state in 11 Li is unbound and degenerate. More dineutrons coupled to 11 Li will
then lead to the same formal energy estimate as for individual neutrons, and a
positive binding energy may result for this correlated wavefunction. Of course,
units (clusters) other than the dineutron may be of interest as well. In all the
above examples, the radial wavefunction must be adjusted and the effects of
the center-of-mass correlation included. Both effects provide more binding
energy working against the effect of the Pauli principle. If specific numbers of
neutrons correlate better than others, a new type of shell structure may result
that could lead to particle-stable "reefs" relatively far from the dripline (64).
Established cases of systems with many neutrons outside a core are rather
scarce. The hydrogen isotopes give indications of resonances, but the He iso­
topes probably provide better insight into the possibilities of forming multi­
neutron configurations. The two-neutron separation energy increases from
604 HANSEN, JENSEN & JONSON

0.973 MeV in 6He to 2.14 MeV in sHe and, finally, two neutrons become
unbound by 1.2 MeV in IOHe. The two-neutron separation energy is about
1.4 MeV in 7He and 9He. Correlations are essential for the stabilization of the
heavy systems, and this pattern can be expected to recur in heavier elements.

3.4 Bound Exci ted States


Halos may appear in excited states as well as in ground states, as illustrated in
Access provided by 2802:8010:470a:5b00:ddfc:f069:d425:9566 on 11/26/22. For personal use only.

Figure 1. The frequency of occurrence is clearly in favor of excited states since


Annu. Rev. Nucl. Part. Sci. 1995.45:591-634. Downloaded from www.annualreviews.org

many normal nuclei have excited states of low angular momenta close to the
one- or two-particle thresholds. Over the years, various types of cluster states,
especially combinations of neutrons and Ot clusters, have been investigated in
light nuclei. In this subsection, we do not discuss these normal nuclei but
instead describe a new type of excited halo states.
As mentioned in Section 2, Efimov (52) suggested that an equal-mass three­
body system may have a series of loosely bound levels. If the two-body force
is characterized by the radius R
and the (very large) scattering length as, then
the size of the Nth state is of the order ReN1r, and the number of bound states
is of the order ll'-llnCiasl/R). States of a similar nature may be possible in
three-body neutron halo systems (65).
The nature of these states can be understood from the effective radial equation
(Equation 6), in which the centrifugal barrier term more generally can be written
as [A(p) - 1/4]/p2. The bound-state solution to the differential equation

(_ d2 + J...(p) 1/4 2m
f=0
- _

E) 7.
dp2 p2 1'12

is o f the form f ex p<X (Ot = 1/2±..(i.) in the region where the energy term can
be neglected, and the usual exponential fall-off is f ex exp(-Kp) for p-values
at which the centrifugal barrier term can be neglected. At smaller distances,
the wavefunction is modified by the influence of the finite potentials.
In Borromean systems held together by short-range potentials, the extreme
long-distance behavior of A is given by ..fi. == K + 2. The transition to this
region of exponential fall-off takes place when p is roughly the sum of the
three two-body scattering lengths, which for ordinary systems (although not
in the case considered here) is of the order of the ranges of the interactions.
However, in a large region of intermediate p-values, A can be negative and
constant provided at least two of the binary subsystems simultaneously (e.g.
by being identical) have virtual or bound s states close to zero energy and hence
have scattering lengths much larger than the range of the interaction. When
}" < 0, the radial wavefunction oscillates as

f ex p<X ex .jpsin(Hlnp). 8.
NUCLEAR HALOS 605

Different numbers of oscillation correspond to different excited states with the


same angular quantum numbers. The distance !:J.p between two nodes of an
oscillation is !:J.p = PI[exp(rr/H) - 1], where PI is the p-value of the
first node. For the first excitation, PI is a few times larger than the range of
the interactiions. A new bound state appears when A remains negative over a
sufficiently large interval to hold such an oscillation. If the negative A interval
is large enough to contain more of the mentioned p oscillations, more excited
Access provided by 2802:8010:470a:5b00:ddfc:f069:d425:9566 on 11/26/22. For personal use only.

states appeaI. Infinitely many bound states can be obtained in this way.
Annu. Rev. Nucl. Part. Sci. 1995.45:591-634. Downloaded from www.annualreviews.org

Already the first of these excited states is very large in size and has small
binding energy(65). Estimates in a simple model appropriate for light nuclei
such as "Li and 14Be are shown in Figure 4. The energies of the ground
state and of the first excited state are given as a function of the strength of the
neutron cone potential. All energies decrease with the increasing attraction.
The first excited state appears extremely close to the two-body threshold when
its excitation energy E 1 - Eo � 1 MeV corresponds to a neutron core scattering
length of about - 18000 fm. The energy E 1 decreases and disappears into the
two-body continuum for a scattering length of about 15 fm, corresponding to
a weakly bound neutron core system. Therefore, for this state, Sn = EAn -
EI
first increast:s and then decreases, as seen in the inset. The size corresponding
to separation energies of approximately 1 keV is several hundred fermis. These
states are most likely to appear when the neutron core subsystem is bound and
when the core spin is zero. The excited state has the same angular quantum
numbers as the ground state and has a node in the radial wavefunction placed
at p-values far beyond the range of the nuclear potential. Such extreme halo
states are fragile structures and will be difficult to create and detect.

4. STRUCTURE AND REACTIONS


A halo is characteristic of a bound state with suitable structure that lies very
near a threshold for particle emission. For this reason, the properties of other
states of the same nucleus are only of indirect relevance to the halo state. This
halo state can be studied via its diagonal matrix elements, accessible through
measurements such as nuclear moments (35, 36) or elastic scattering (66--68),
or via its non diagonal matrix elements, i.e. processes in which the halo system
is destroyed or created. The main body of information has, so far, come from
the latter method. The study of the momentum distributions of particles and
fragments from nuclear breakup processes has been especially rewarding.

4.1 Reacti on Mechanisms


In the energy range available from present-day fragment separators (typically
between 25 and 1000 MeVlu), nuclear reactions can initially be treated by
606 HANSEN, JENSEN & JONSON


10-3

-
2 �
t
Access provided by 2802:8010:470a:5b00:ddfc:f069:d425:9566 on 11/26/22. For personal use only.

10- r-
� ,
Annu. Rev. Nucl. Part. Sci. 1995.45:591-634. Downloaded from www.annualreviews.org

r
:E r
'-'" E
1
10- ;;...


Q)


100

t

-E
0
-


t �
2

101 r:- .....


� 1
I

l
c:::
<:

� I I I

6 7 I) 9 10 11 12
Strength of Potential ( MeV)
Figure 4 Energies of the two-body ground state EAn, the three-body ground state Eo, and the first
excited state E, (dashed lines) of the three-body system as functions of the core-neutron potential
strength. The inset shows the neutron separation energy for the excited state. The neutron-neutron
potential reproduce$ the low-energy s-wav,e scattering data, and the range of the neutron-core
potential is fixed. The core mass is nine timl�s the nucleon mass in this calculation.

simple means. The radioactive projectile may be assumed to move in a classical


trajectory, usually a straight line, and the Coulomb interactions are weak enough
to permit the use of perturbation theory. Because the projectile velocities are
much larger than the intrinsic veloci.ties of the halo neutrons and, for the highest
energies, larger than those of the core nUcleons, it is often convenient to use
the sudden approximation, first applied to the deuteron ( 19). Furthermore, the
relevant matrix elements tend to be dominated by contributions from large dis­
tances (see Figure 1 ), so that simpk, asymptotically correct wavefunctions will
account almost quantitatively for absolute cross sections and momentum distri­
butions. The presence of only one or, at most, a few bound states in the usual halo
NUCLEAR HALOS 607

systems also means that properties of the continuum, such as reaction cross sec­
tions and momentum distributions, can often be expressed in terms of sum rules
(37), as illustrated by the example of l IBe (69). A general discussion of sum
rules for mUltipole excitations in halo nuclei is presented by Sagawa et al (70).
The reactions of halo states can be divided into two categories according
to whether the impact parameter b is greater or smaller than the sum Rl +
R2 of the (:ore and target radii. For close encounters, the reaction in many
Access provided by 2802:8010:470a:5b00:ddfc:f069:d425:9566 on 11/26/22. For personal use only.

respects resembles a normal heavy-ion collision at high energies, which results


Annu. Rev. Nucl. Part. Sci. 1995.45:591-634. Downloaded from www.annualreviews.org

in complex fragmentation and deep-inelastic processes known as core breakup


reactions (69, 7 1). The distant encounters, i.e. those with b > Rl + R2, would
in a normal scenario allow only Coulomb excitation. In the presence of a halo,
which protrudes far beyond the nuclear surface, its neutron(s) may interact
strongly with the target and liberate the core via a stripping process. [Note that
here we use the word stripping in the sense in which it was first applied ( 1947) to
a beam of 190-MeV deuterons. Serber (72) recently commented on the origin
of this terminology.] However, in the Coulomb excitation of the halo, the recoil
momentum is imparted to the charged core, which leads to El excitations of
essentially single-particle strength. Such El transitions are not encountered in
the garden variety of nuclei because of the strong quenching of the E 1 strength
caused by the collective giant dipole resonance. Because the core of the halo
state emerges intact after the reaction in processes of the second category, we
refer to them as dissociation or halo removal reactions.
The selection according to impact parameter is implicit if the core fragment
is observed. In experiments in which neutrons are observed (42, 69, 7 1, 73-78)
this selection may be introduced as a coincidence requirement (single neutron­
exclusive reactions). The underlying assumption that the core will disintegrate
if a core target collision occurs was recently confirmed in a 11 Be experiment by
Nakamura et al (75). These authors used the Coulomb deflection of the core
fragment frOom a lead target to estimate the impact parameter and found that the
dissociation cross section vanished for b smaller than 10- 15 fm.
In the foUowing sections, we give many examples of how insight into the
structure and reactions of halos can be obtained from the momenta of core
fragments and from those of neutrons associated with dissociation and core
removal processes. Orr recently emphasized the complementary aspect of these
techniques (16).
4.1.1 FRAGlvlENTATION REACTIONS Over the past decade, great progress has
been made toward understanding the reactions of heavy ions at intermediate
and high en<:rgies (79). Of special relevance for the present subject are the
projectile-like fragments, which emerge in a narrow cone near 0° with a velocity
close to that of the projectile and with a momentum distribution that is close to
608 HANSEN, JENSEN & JONSON

isotropic in the projectile rest system (see 80-82). In this system, the momentum
distributions from the fragmentation of a stable projectile with mass number A
are well represented by single Gaussians with a characteristic spreading width
jAp(A - Ap)
a = ao y 9.
A-I '

where Ap is the mass number of �le fragment. The constant ao is of the


Access provided by 2802:8010:470a:5b00:ddfc:f069:d425:9566 on 11/26/22. For personal use only.

order 70-90 MeV/e over a wide range of beam energies (82) but drops rapidly
Annu. Rev. Nucl. Part. Sci. 1995.45:591-634. Downloaded from www.annualreviews.org

to much lower values (83) below 10 MeV/u. It can be related directly to


the Fermi energy (84) or to the motion of clusters (85) in a simple statistical
picture that interprets the kinetic energy distribution as having existed in the
projectile prior to the collision.
This picture immediately suggested that the narrow transverse momentum
distribution observed (41) for9Li fragments from the breakup of 1 1Li, as well as
the even narrower distributions found in later work, represent the characteristic
intrinsic momentum in the subsystem core plus halo neutrons. Qualitatively,
the space-momentum relationship is exactly what would be expected from
Heisenberg's uncertainty principle: The large size of the halo implies a narrow
momentum distribution. On closer examination, however, the quantitative rela­
tionships are more complicated. In �le following, we show that for 11Be, the fi­
nal moments are influenced significantly by the reaction mechanism and that the
results for 1 1 Li are modified by final-state interactions in the (9Li +n) subsystem.
4.1.2 DISSOCIATION REA CTIONS Dissociation reactions are the result of
Coulomb and nuclear interactions (2). A semiquantitative understanding that
includes both effects may be obtained from a simple black-disc picture of the
nuclear interaction. Modeled on Glauber's original discussion of the deuteron
(20), this analysis was applied to l I U (42,86) and later to l l Be (69, 74). The
basic idea is shown in the inset to Figure 5. Consider a coordinate system
CM situated in the center of mass of the system lOBe + n and traveling with
beam velocity v = fJe along the z axis. With the impact parameter b along the
positive x axis, the recoil wave vector imparted to the charged core due to the
Coulomb force is
Q
=_� 2Z1Z2e2 10.
b2 ne{J
Here, the Coulomb deflection and the longitudinal component have been ne­
glected. In the new center-of-mass system eM' formed after the collision, the
wave vector of the neutron is changed by q = -QAHI A, where the mass
number of the halo is AH-
At the same time, the target nucleus has swept through the halo and left a
"wound" 81/10, with a value of 1/10 in:>ide the tube swept by the target and a value
NUCLEAR HALOS 609

of zero outside the tube (see Figure 5), so the wavefunction in eM' after the
collision can be approximated by

11.

normalized to (I-Pa), where Pa = J lol/lol2 dr is the absorption probability.


(The word absorption should not be taken literally; it implies that the halo
Access provided by 2802:8010:470a:5b00:ddfc:f069:d425:9566 on 11/26/22. For personal use only.

neutron will not appear in the forward direction with essentially beam velocity.)
The wavefunction symbolized in Figure 5 represents a mixture of the ground
Annu. Rev. Nucl. Part. Sci. 1995.45:591-634. Downloaded from www.annualreviews.org

state and excited states. The amplitude for an elastic process is the overlap

Yel = J l/Io(r')I/I(r)dr' = Yc - YC+N, 12.

which results in a Coulomb term and a Coulomb-plus-nuclear term. The prob­


ability of an elastic process Pel is the square of this quantity. We can now


B
t:l Diss.
=
o
.....
....

C,)
Q.)
if)
(/J
I
Inelast.
(/J

8 {
u Abs.
I
I ! , , ! I I ,
0. 1 ! ! 1 ! I !

o 20 40 60 80 100

Target Z

Figure 5 Measured and calculated (69) dissociation cross sections for the reaction (IlBe, lOBe)
at 41 MeV/u The solid lines (starting from below) are the calculated contributions from ab­
.

sorption, from inelastic reactions (Coulomb and nuclear reactions with final states consisting of
lOBe and a free neutron), and from the sum of the two (top). The inset illustrates the geometrical
model discussed in the text. The nuclear inelastic contribution is approximately equal to that from
absorption.
6 10 HANSEN, JENSEN & JONSON

calculate the probability of an inelastic process Pinel = 1 Pa - I Yel 1 2 that


-

(assuming that there are no bound excited states) must represent transitions of
a neutron to the continuum. In the absence of a Coulomb term, the absorption
and inelastic contributions are approximately identical, just as they are for free
neutrons, in which these contributions lead to a cross section twice the size
of the geometrical one. In this case, i.e. for a light target,the inelastic pro­
cess is referred to as diffraction dissociation and is associated with transverse
Access provided by 2802:8010:470a:5b00:ddfc:f069:d425:9566 on 11/26/22. For personal use only.

momentum components that reflect the radial dimension of the "wound." With
Annu. Rev. Nucl. Part. Sci. 1995.45:591-634. Downloaded from www.annualreviews.org

a simple Ansatz for the wavefunction and the geometry (see Section 4.2), the
reaction amplitudes may all be expressed analytically (69). The probabilities
P can be converted to cross sections by integration over the impact parameter
b with surface element 2rrbdb, giving the result shown in Figure 5.
4.1 .3 MOMENTUM DISTRIBUTIONS A more detailed picture of the structure
of the halo than the one based on integral cross sections can be obtained by
measuring the momentum distributions of the nucleons and nuclear fragments
emerging from the collision. An example of an experimental arrangement for
this purpose is given in Figure 6, whi.ch shows the equipment used in a series of
experiments (7 1 , 87, 88) at the GSI. An essential feature of these experiments
has been the use of coincidence requirements between neutrons and charged
fragments to select particular reaction channels.
In order to compare different measurements, it is instructive to characterize
them by parameters from simple statistical distributions, a problem that has been
discussed by Riisager (89). One advantage of a Gaussian is that the three- and
two-dimensional distributions factorize into one-dimensional distributions, all
characterized by the same spreading parameter a . For example, the distribution
along one dimension in space is unchanged by an integration over all or part of
the other two dimensions. If this relation holds, the experiments are insensitive
to the precise acceptance of the apparatus. Nature, however, is often less
simple.2
A Yukawa wavefunction, which corresponds to the external wavefunction of
a neutron in an s state, has the three-dimensional momentum distribution
r 1
W3 = 2rr 2 2 /4 + p2)2 ' 13.
(r

2 A number o f papers represent the momE:ntum distribution o f core fragments as a superposition


of a broad and a narrow Gaussian, i.e. the line is described by three free parameters (two widths and
one intensity ratio) in addition to the normalization. The simple relations for single Gaussians no
longer hold in this case. The broad component is sometimes interpreted as resulting from stripping
of core neutrons, which have high momenta, but this requires an additional mechanism in which a
halo neutron is recaptured by the excited core.
NUCLEAR HAWS 61 1

;:;
;; 1 ,5

!
� I
Access provided by 2802:8010:470a:5b00:ddfc:f069:d425:9566 on 11/26/22. For personal use only.
Annu. Rev. Nucl. Part. Sci. 1995.45:591-634. Downloaded from www.annualreviews.org

Figure 6 Experimental arrangement for the GSI experiment (7 1 , 87, 88). The direction of the
incoming beam, here 280-MeV/u HLi, is determined with multiwire drift chambers (MWDC) ( 1 ,
2). The charged fragments from the target (3) are analyzed using the ALADIN magnet (4), two
MWDCs (5), and a plastic wall (6). The position and flight time of the coincident neutrons are
measured with the large-area neutron detector (LAND) (7). The left inset shows the momentum
distribution oCthe 9Li fragment in the direction perpendicular to the pole plane of the magnet, and the
right inset depicts the radial momentum distribution of neutrons in coincidence with a 9Li fragment.

where the width parameter r == '2JiK is defined by Equation 1 and where the
three-dimensional volume element is dr 2
= df2p dp . In two dimensions with

surface element dA = d<PPrdpr , the radial momentum distribution is


r 1
W2 == 14.
41l' (r 2 /4 +
-
3/2 '
if)
After integration over two dimensions, the parallel momentum distribution
becomes a Breit-Wigner curve
r 1
WI == 15.
21l' r2/4 + pn
_ .

corresponding to a full width at half maximum (FWHM) of r. From the first


of the three �:xpressions, we see that a measurement of the same distribution
with a narrow aperture along a given axis (e.g. with Px == Py = 0) would give
a FWHM of 0.644r .
612 HANSEN, JENSEN & JONSON

To characterize a given experimental situation by a single distribution param­


eter is at best meaningful for the low-momentum components characteristic of
the halo wavefunction at large distances. The behavior of the Lorentzians at
high momenta is nothing but a reflection of the (unphysical) singularity of the
Yukawa wavefunction at the origin; consequently, the average energy becomes
divergent.
The systems with two neutrons and a core (such as 1 1 Li) present a more com­
Access provided by 2802:8010:470a:5b00:ddfc:f069:d425:9566 on 11/26/22. For personal use only.

plex problem, but we maintain for simplicity the Lorentzian width parameter
Annu. Rev. Nucl. Part. Sci. 1995.45:591-634. Downloaded from www.annualreviews.org

r as our reference value. This is not completely without physical foundation.


It has been shown by Zhukov & Jonson (90) that a strict three-body treatment
with asymptotic wave functions of Yukawa type leads to momentum distribu­
tions that are generalized Lorentzian functions. Over a wide range of momenta,
. these functions can be approximated by a single Lorentzian curve.
4.2 The One-N eu tron H alo: l lBe
The isotope 1 1 Be, the best-studied case of a single-nucleon halo, is useful for
clarifying the interplay between structure and reaction mechanism. In order to
understand the parity inversion (Figure 1), it is, as shown by Sagawa et al (90)
and Otsuka et al (91), essential to take into account the effects of quadrupole
core excitation and pairing blocking. The dominant part of the wavefunction is
then a single-particle lslj2 state with an approximately 20% admixture of Od5j2
coupled to the 2+ core excitation. This value agrees with the measured (92)
spectroscopic factor of 0.77. Several other recent papers (93-95) investigate
the structure of 1 1 Be and underline the role played by core excitations; however,
all also agree on the predominantly single-particle character of the two bound
states. In the simple estimates given in the following section, we approximate
tt
Be by a product wavefunction wilh an inert core and the asymptotic Yukawa
wavefunction for the halo neutron given in connection with Equation 1 . This
approximation underestimates the tail of the halo but this can be compensated
by applying a "finite-size correction" to the final cross sections (37, 69, 75).
Other simple wavefunctions with the correct asymptotic behavior have been
investigated by Bulboaca et al (96).

4.2. 1 DISSOCIATION REACTIONS The term dissociation reactions refers to re­


actions in which l OBe is observed as a final reaction product, either alone
or in coincidence with a neutron. In Figure 7, calculations are compared
with experimental results on neutron radial-momentum distributions and on
the excitation energy spectra (69, 74, 75, 88). For heavy targets, the low­
momentum part of the measured spectra is mainly due to Coulomb excitation
(2, 97) and can be accounted for aocurately in perturbation theory by assuming
that the reaction is a direct breakup process leading to the p-wave part of plane-
NUCLEAR HALOS 613

wave states (69, 75). For higher momenta, the nuclear contributions become
dominant. For the light targets Be and C (Figure 5), the nuclear interaction
dominates completely and shows up clearly (Figure 7) in the transverse mo­
mentum distributions (69, 88). However, this interaction is barely detectable
in the complete kinematics experiment (75), perhaps because of its limited
acceptance.
The momentum distributions may be obtained from the arguments given in
Access provided by 2802:8010:470a:5b00:ddfc:f069:d425:9566 on 11/26/22. For personal use only.

Section 4. 1.2. The wavefunction (Equation 1 1) of the reaction complex is a


Annu. Rev. Nucl. Part. Sci. 1995.45:591-634. Downloaded from www.annualreviews.org

superposition of the ground-state wavefunction and a mixture of excited states,


and the conresponding momentum distribution does not represent that of the
outgoing neutrons. The wavefunction of the decaying state may be obtained by
proj ecting out the ground-state wavefunction in the coordinate system eM' to

Figure 7 Radial momentum distributions of neutrons from the dissociation of 11Be. (a) Results
(69) at 41 MeV/u for a gold target compared with Coulomb excitation theory (short dashes, eE)
and with an estimate including Coulomb and diffraction dissociation in the sudden approximation
(solid lines, SUD). (b) The corresponding results for a Be target at 41 MeV/u (69) and for a C target
at 460 MeV/u (118) (open triangles, arbitrary normalization) are due to diffraction dissociation and
the curve represents a calculation (74). (c) Distributions in excitation energy have been measured
in a complete kinematics experiment (75) for Pb (open circles) and C (filled circles) targets.
614 HANSEN, JENSEN & JONSON

obtain 1/1 - Ye1 1/l0 , which clearly is orthogonal to 1/10 . This wavefunction is then
transformed back to CM to obtain the decaying state in the coordinate system
traveling with the beam
1/Id(r) = 1/I0(r) - 81/10(r) - Yel e-iq- r 1/l0(r) . 16.
This expression has the correct normalization Pinel given in Section 4. 1 .2. As
final-state interactions are neglected, the momentum distribution in CM is given
Access provided by 2802:8010:470a:5b00:ddfc:f069:d425:9566 on 11/26/22. For personal use only.

by the square of the Fourier transform of Equation 16, which under certain sim­
Annu. Rev. Nucl. Part. Sci. 1995.45:591-634. Downloaded from www.annualreviews.org

ple assumptions can be expressed analytically. These distributions, projected


to the space observed experimentally, are shown in Figure 7.
Thus the upshot of these experiffil�nts is that the momentum spectra of neu­
trons resulting from dissociation of I I Be never, strictly speaking, correspond
to the halo wavefunction in momentum coordinates, although the result for
Coulomb dissociation on a gold target comes close (74) and would correspond
to r = 63 MeV/c in good agreeme:nt with the value r = (8/1,sn)I/2 = 58.5
MeV/c expected from the neutron se:paration energy (see Section 4.1 .3). How­
ever, the dip near zero momentum in the left part of Figure 7 is the signature
of a final p state reached by EI excitation from the halo's ground state, and the
theoretical curve shows that Coulomb excitation supresses the high-momentum
components. The broad spectra observed for light targets are the signature of
diffraction dissociation, and their shapes convey no information about the halo.
Because diffraction dissociation is a transverse process involving mainly
the neutrons, the longitudinal momentum distributions are a cleaner tool for
investigation of the halo. Moreover, longitudinal distributions have less smear­
ing in the case of processes involving charged fragments since the Coulomb
contributions from the incoming and outgoing leg of the trajectory to a first
approximation cancel. Orr et al (43) demonstrated how to use the Michigan
State University fragment separator as a 0° energy-loss spectrometer to obtain
a resolution an order of magnitude better than the momentum bite of the sep­
arator. This technique was recently applied to I I Be by Kelley et al (98), who
found widths that within experimental errors are identical for light and heavy
targets; the average corresponds to a momentum parameter r of 43.6 ± 1 . 1
MeV/c . This value is much narrower than the theoretical intrinsic value of58.5
MeVIc and also narrower than the value of about 60 MeVIc obtained from
core breakup reactions discussed below. At first this discrepancy seemed to
suggest that our understanding was incomplete,especially since several theo­
retical papers (100-102) state that the longitudinal momentum distribution is
insensitive to the reaction mechanism and reflects the original wavefunction
of the halo system. A recent paper (103) takes a different view and proposes
that the narrow width for the II BI� longitudinal momentum distributions is
caused by two different reaction mel;hanisms: a Coulomb process for the heavy
NUCLEAR HALOS 615

(Ta, U) and . a nuclear process for the light (Be, Nb) targets. The argument may
be summarized as follows.
Coulomb dissociation treated in perturbation theory and with the approxima­
tion of a Yukawa wavefunction leads to an explicit expression for the double­
differential cross section (69), which can be transformed to a distribution on
the longitudinal momentum P II ' Calculations ( 103) for the beam energies 4 1
and 460 MeV/u give widths of 4 1 and 37 MeVIc, respectively, in good agree­
Access provided by 2802:8010:470a:5b00:ddfc:f069:d425:9566 on 11/26/22. For personal use only.

ment with the average value of 43 ± 2 MeVIc measure by Kelley et al (98) for
Annu. Rev. Nucl. Part. Sci. 1995.45:591-634. Downloaded from www.annualreviews.org

tantalum and uranium targets at a beam energy of 63 MeV/u. The origin of


the narrowing is simply the dominant transverse term in high-energy Coulomb
excitation, which makes the angular distribution peak at 90° to the beam.
Nuclear dissociation produces a l OBe fragment via two reaction channels,
neutron removal and diffraction dissociation. The analysis ( 103) shows that
both lead to approximately the same momentum distribution, given by the
square of thl� Fourier transform of the wound in the halo wave function 8 1/10
introduced in Equation 1 1 . In the approximation of a planar cutoff at the target
surface, this transform can be expressed analytically. After integration over the
unobserved x and y directions, the result turns out to be the intrinsic momentum
distribution jin one dimension give by Equation 15 multiplied by a correction

£7r/2 cos(lJ) exp[-2wl cos(tJ)]dtJ,


factor C 1 ( w)l defined by

C1 (w) = 17 .
.0
where the parameter is defined as w = (b - Rr )(r2 14 + p�) 1 /2In, expressed in
terms of the impact parameter b and the target radius. This shows that increasing
values of the: impact parameter lead to increasingly narrow distributions. As
an average over impact parameter, the cross section is distributed with a width
of 37 MeVIc, close to the measure value of 43.5 ± 1.5 MeV/c found by Kelley
et al (98) for the light targets. Furthermore, the model with a planar cutoff must
exaggerate the importance of the large impact parameters and underestimate the
width. The t:ssence of the argument is then that the nuclear reaction samples
the outer parts of the halo, which have a smaller momentum content. [This
argument would not hold for a Gaussian wavefunction, which may explain
some of, the disagreemnet with the theoretical papers ( 100-102).]
4.2.2 CORE BREAKUP REACTIONS A third technique for studying the neutron
momenta of the halo state is to observe them in coincidence with fragmentation
of the core (66, 68) (see Section 4.1. 1). The channels corresponding to impact
parameters smaller than the sum of the target and core radii were originally
thought to bt: too complex to have any direct bearing on the halo. However,
these channels are useful sources of information for three reasons. First, because
616 HANSEN, JENSEN & JONSON

a halo neutron has a probability distribution that extends far beyond the core, it
must also have an appreciable probability of steering clear of the debris from
the core-target conflagration. Second, the sudden approximation is applicable
to this process, so the neutron (viewed in the coordinate system following the
beam) may be expected to emerge with a momentum distribution characteristic
of the initial state. Third, the "wound" in the wavefunction does not change the
momentum greatly. In Equation 1 6, for example, for a large recoil momentum
Access provided by 2802:8010:470a:5b00:ddfc:f069:d425:9566 on 11/26/22. For personal use only.

q the elastic amplitude vanishes so that the final state is very similar to 1{Io.
Annu. Rev. Nucl. Part. Sci. 1995.45:591-634. Downloaded from www.annualreviews.org

Two experiments have applied this approach to I I Be. In the first, which used
a beam energy of 4 1 MeV/u (69), all events already contained in the dissoci­
ation cross sections were removed lby an anticoincidence requirement to give
what was referred to as restricted inclusive cross sections. These represent the
channel "single neutron plus anything different from l OBe" and yield identical
angular distributions for Be and Ti targets that seem to indicate a superposition
of a narrow and a broad component. The narrow distribution corresponds to
values of r of 59 and 6 1 MeV/C, respectively, while the broad distribution
presumably represents neutrons from the core breakup, a contribution that, in
principle, could have been determined in a separate measurement with lOBe
projectiles with the same velocity. Such an experiment is now in progress at
GANIL. Experiments in which 7.8Lj were selected as the exit channel have been
performed at the aSI (7 1) (see the 7Li data discussed in Section 4.4.2), with an
I I Be beam of 460 MeV/u leading to values for the r parameter of 67 ± 5 and
56 ± 4 MeV!c.
All in all, the momentum distributions from reactions of I I Be seem well in
hand. They suggest the presence of an s-state halo determined by the neutron
separation energy and characterized by an intrinsic momentum width parameter
f' close to 60 MeV/c. The momentum widths observed in experiments are
changed in characteristic ways by the reaction mechanism.

4.2.3 OTHER ONE-NUCLEON HA LOS The isotope 19C exhibits a longitudinal


momentum width for core fragments from reactions on a Be target of 44 ±
6 MeV/c that is similar to that of the I l Be isotope (104). The transverse
momentum width for neutrons from the core-removal reaction was measured by
Angelique et al (lOS), who found a r of 40 ± 3 MeV/c. The narrow distribution
suggests that the halo state again is an s state and is at least consistent with the
low neutron separation energy of (160 ± 1 10 keV), which is the weighted
average of the three most recent experiments (105).
The p state valence proton of8B is a candidate for a halo state, and the interest
in this problem is heightened by the role of this isotope in the solar neutrino
problem [see the discussion by Riisager ( 10, 106) and references therein] .
In recent experiments ( 107), the longitudinal momentum distribution for the
NUCLEAR HALOS 6 17

dissociation reaction to CBe + p) was measured and found to have a r of 8 1 ± 6


MeVIc, about one third of the value obtained for one-proton removal reactions
on the stable: targets 12C and 160. Experiments on the transverse momentum of
protons from 8B are in progress (B Blank, private communication). The proton
halo of the 495-keV s state of 17p is discussed in Sections 2 and 5 . 1 .

4.3 The Two-Neutron Halo: I I Li


Access provided by 2802:8010:470a:5b00:ddfc:f069:d425:9566 on 11/26/22. For personal use only.

The two-neutron halo has received more attention than any other problem related
Annu. Rev. Nucl. Part. Sci. 1995.45:591-634. Downloaded from www.annualreviews.org

to neutron hulos, and it is still not certain that the main qualitative elements of
this phenomlmon have been clearly identified. The dust will not settle for some
time, and all we can do at present is attempt to identify some trends. The special
position of I l Li is due to its very low two-neutron separation energy, recently
determined by Young et al (108) to be 295 ± 35 keY.

4.3 . 1DISSOCIATION REACTIONS The momentum widths for the core recoils
from dissociation reactions of this nucleus have been studied in many exper­
iments. Except for a few earlier values that may have been broadened by
experimental effects, the r values obtained in these experiments seem to be
consistenly close to 45 MeVIc (43, 87, 89, 109) (Section 4.1 .3), and at most, a
weak dependence on the target Z is observed (see Figure 2). In the following
survey, the emphasis is on the results for light targets, which are the easiest
to interpret. For a beryllium target, the longitudinal distributions give at 66
MeV/u the average result (43, 109) 46 ± 2 MeV/c. At 280 MeV/u, the longitu­
dinal result (87) is 49 ± 3 MeV/c, and the transverse (Py) is 45 ± 4 MeV/c. For
the neutrons, typical values are 25-35 MeV!c. Thus the ratio between the core
and neutron momentum widths at first sight seems close to the factor of ap­
proximately �fi that would be appropriate if the momenta were representative
of those preexisting in the halo and if the two neutrons were weakly correlated
(14, 89). The data have also been analyzed in such a way as to suggest very
strong correlations between the two neutrons ( I l l ) . It turns out that n�ither the
distributions of neutrons nor those of the core recoils can be interpreted in such
a simple manner.
The momentum widths for the neutrons, in fact, present two problems. First,
the momentum distributions at about 29 MeV!u (42, 73, 76) appear to be much
too narrow to agree with a halo rms radius of 5 fm (1 1 1 ). The measured value of
r is about 30 MeVIc according to Sackett et al (74), whose experiment has the
best granularity at small angles, and about 25 MeV/C for the other experiments
(73). Second, the data give almost the same shape for light and heavy targets
despite the fact that two different processes are involved: (nuclear) diffraction
dissociation for the light targets and Coulomb dissociation for the heavy targets
(see for example the very different results for the one-neutron halo in Figure 7).
618 HANSEN, JENSEN & JONSON

As a way around this problem, Barranco et al (86) suggested extending the


interpretation given for I I Be (74) with a contribution arising from the decay in
flight of particle-unstable l OU formed after stripping or diffraction of the other
halo neutron. [There are now indications that the I OU intermediate product is
a superposition of two states (see Section 4.4).] If this state is very close to
the threshold, the Be target gives ris'e to a narrow distribution superimposed on
Access provided by 2802:8010:470a:5b00:ddfc:f069:d425:9566 on 11/26/22. For personal use only.

the broad distribution from diffraction dissociation. The latter distribution is


unimportant at small angles, although it accounts for approximately one third
Annu. Rev. Nucl. Part. Sci. 1995.45:591-634. Downloaded from www.annualreviews.org

of the dissociation cross section on a light target. The distribution for the gold
target arises predominantly from Coulomb dissociation. Its resemblance to that
for Be must be considered a numerical coincidence.

4.3.2 CORE BREAKUP REACTIONS The results for neutrons in tenus of the r
parameter defined in Section 4.1 .3 from experiments at 28 and 280 MeV/u
are shown in Figure 8 and give a consistent average value of 43 ± 3 MeV/c,
well above the 25-30 MeV/c seen in dissociation reactions. Assuming that the
core breakup result represents the " true" value for neutrons, it is at first sight
surprising that this and the 9Li momentum width are identical. We suggest
below that the near equality in part reflects the (9Li + n) final-state interaction.
However, this result does not seem Ito be universal [see the otherwise analogous
case of 6He ( 1 12), discussed below].

4.3.3 COULOMB DISSOCIATION Although reactions with light targets seem to


be the simplest to interpret, the strong interest in Coulomb processes con­
tinues. This is not surprising sill(!e electric dipole transitions of low energy
and with a strength at the scale of one Weisskopf unit are unique to the halo
nuclei. Moreover, frequent speculations have indicated that a collective phe­
nomenon referred to as a soft El giant resonance could be involved, although
we prefer to interpret the E l transitions as predominantly a single-particle
phenomenon. With 11 U incident on a heavy target, the Coulomb effect is ap­
preciable even at 800 MeV/u (39), and at 29 MeV/u the measure of 5-barn
dissociation cross section is mainly electromagnetic. Several experiments have
been performed to measure the EI strength function, which for I l Li means
determining the momenta of the three dissociation products. These experi­
ments have led to the so-called kinematically complete measurements of the
breakup, which, as underlined by Orr (16), can be very sensitive to the exper­
imental acceptance and which may be far from complete in this sense. The
first experiment (76, 77), in fact, had a transverse momentum acceptance for
a single neutron of only 15 MeV/c. Widths in good agreement with those
given above, about 42 MeVIc for the charged fragment and 30 MeV/c for the
neutrons, were reported. Additionally, the excitation process was governed by
NUCLEAR HALOS 619

1 l Li Be/C :;iP Li
Access provided by 2802:8010:470a:5b00:ddfc:f069:d425:9566 on 11/26/22. For personal use only.

• + � +n+X
Annu. Rev. Nucl. Part. Sci. 1995.45:591-634. Downloaded from www.annualreviews.org


• T
- .

o 10 20 30 40 50 60
P
r
( MeV/c )
Figure 8 Radial momentum distributions of neutrons from I I Li in coincidence with charged
fragments (excluding 9 Li). The top data set is for reactions at 28 MeVlu in a Be target (42).
The distribution from 9Li projectiles has been subtracted to eliminate neutrons from core-target
interactions. Th: bottom data set is for reactions at 280 MeV/u in a C target (71). The curves are
two-dimensional Lorentzians with rs of 42 and 43 MeV/c. respectively.
620 HANSEN, JENSEN & JONSON

three-body phase space, and hence no evidence for correlations was present.
This lack of evidence is especially notable in the distribution of relative n-n
momenta, which fail to show a peak at low energy. A new 11 Li experiment
(78) with considerably improved acceptance has provided better data for the
El strength function. It demonstrat,es that the excitation energy primarily in­
volves motion of 9Li relative to the neutrons and yields no evidence for n-n
correlations.
Access provided by 2802:8010:470a:5b00:ddfc:f069:d425:9566 on 11/26/22. For personal use only.

An intriguing feature of the experiment by Ieki et al (76) is a velocity shift


Annu. Rev. Nucl. Part. Sci. 1995.45:591-634. Downloaded from www.annualreviews.org

along the beam axis, which gives th� neutrons lower velocity than the charged
fragment. The magnitude of the effect corresponds to that expected from a
simple picture involving deceleration of the projectile in the electric field of
the target, emission of the neutrons close to the target, and a reacceleration of
the (lighter) fragment to a velocity slightly above that of the beam. A very
similar effect has been observed for 1 1 Be by Nakamura et al (75), who used a
differential method to show that the: shift varies with the impact parameter, as
is expected of a Coulomb mechanism. The effect nonetheless remains difficult
to accommodate in theory [see (1 13--1 15)]. There are also other reports, in part
contradictory, of similar shifts. The problem is interesting and would probably
justify an effort to obtain more accurate data.

4.3.4 DISSOCIATION OF 6He In a parallel theoretical study of the fragmenta­


tion of 6He at a beam energy of 400 MeV/u on a carbon target, Korshenin­
nikov & Kobayashi ( 1 12) reached conclusions similar to those of Barranco
et al (86) by comparing experimental data on momentum distributions of a
particles and neutrons (Figure 9). The isotope 6 He, with its two-neutron sep­
aration energy of 973 keY, is less of a halo state than l l Li, but it has two
essential advantages: well-known basic a-n and non interactions and good
wavefunctions built on microscopk three-body models (see 1). Furthermore,
the a width ra, which is of the order of 120 MeV/c, agrees well with the
calculated intrinsic momentum function. The distribution for the neutrons has
a broad component characteristic of the interaction of the quasifree neutrons
with the carbon target, i.e. of diffraction dissociation, as discussed in Section
4.2. In this context, the observed distribution is broadened over that for free
neutrons by the effect of the intrinsic neutron momenta in 6He (1 1 2). The
width of the narrow momentum component for the neutrons is of the order of
80 MeVIc, and the calculated intrinsic neutron momentum width is broader
by almost a factor of two, a discrepancy that can be understood from the in­
fluence of the final-state interaction caused by the P3/2 resonance situated in
5He 890 keY abo�e the threshold (a + n). Finally, the mechanism involving
elastic scattering of as from the carbon target is suppressed by roughly an or­
der of magnitude relative to neutron scattering. This supports the idea that the
NUCLEAR HALOS 62 1

I I
I '
.--._--r--.--�--�'�

-
...-,
I

C,)
2 f-
8
l, ,
/" , "
....... 6 ,
+ ! ! •

Access provided by 2802:8010:470a:5b00:ddfc:f069:d425:9566 on 11/26/22. For personal use only.

• • -
,
� 4 + • •
Annu. Rev. Nucl. Part. Sci. 1995.45:591-634. Downloaded from www.annualreviews.org

'-" t • • •
t • •
+oJ •
2 •
I=: .. •
� • •

.... -. •
.-4
....-...., -.�.. . •
0
� -200 0 200
.


..
..
r- •
..
..-

-
....-

-
0 I I I I I

- :2 0 0 - 1 60 - 1 20 -80 -40 0 4 0

P (MeV/c)
x

Figure 9 Experimental and theoretical ( 1 12) transverse momentum distributions from fragmen­
tation of 6 He at a beam energy of 400 MeVlu . The inset is the result for detected rt particles. The
main figure shows the distribution of neutrons in the x direction and integrated over the z and y
directions. The: broad component is due to diffraction dissociation, and the curve results from a
calculation. The narrow component reflects the intrinsic neutron momentum as modified by the
(rt + n) final-state interaction.

exclusive core events may be used to select the impact parameter (see Section
4. 1), especially when it is remembered that the ex is much more durable that 9Li
and lOBe.

4.3.5 OTHER CASES Fewer experiments have been performed on other halo
nuclei. Zahar et al ( 1 1 6) determined the longitudinal and transverse momentum
distributions of 12Be fragments from breakup of 14Be on a carbon target. Three
consistent re'sults have the average value of [' 92 ± 3 MeVIe, which is in =

good agreement with the high two-neutron separation energy of 1 .3 1 MeV.


The neutron angular distribution in coincidence with 12Be fragments gives
rn = 47 ± 5 MeVie (73). Too little is known about the structure of these
622 HANSEN, JENSEN & JONSON

nuclei and of 13Be to permit an interpretation of this factor of two at present.


Theoretical predictions indicate that the 13Be nucleus has a 112+ state very low
in the spectrum (1 17).
The isotope 8He, to which we return in Section 5, may be viewed as an a
particle coupled to a four-neutron Skill. Calculations ( 1 1 8) have been compared
with experimental data and show that a five-body cluster-orbital shell model
approximation (COSMA) can account for the momentum distributions of 6He
Access provided by 2802:8010:470a:5b00:ddfc:f069:d425:9566 on 11/26/22. For personal use only.

fragments. Those of the neutrons, on the other hand, are modified by the
Annu. Rev. Nucl. Part. Sci. 1995.45:591-634. Downloaded from www.annualreviews.org

(6He + n) final-state interaction, as they are for the examples discussed above.
Experiments on the parallel momentum of core fragments from 6,sHe are under
way (JJ Kolata, private communication). Finally, new results ( 1 19) have been
obtained for 6,sHe with elastic proton scattering in inverse kinematics, which
measures the matter radii and has confirmed the presence of a thick neutron
skin in SHe.

4.4 The Continuum Problem: 10Li


The Borromean halo states (1) are situated at the threshold that separates the
discrete and the continuous spectrum. They are held together by forces that
in the neighboring nuclei give rise t.o scattering states, and their decay may,
as we have just seen, leave the strong imprint of those states in the form of
final-state interactions. It is therefore imperative to understand the continuum
well. We discuss here recent experiments that shed new light on the neutron
single-particle states of I OLi and, via these states, on the structure of the I I Li
halo.

4.4.1 COMPLEX REACTIONS Shell-model theory ( 1 20, 1 2 1 ) as well as early


three-body calculations of \ 1 Li (122, 123) had correctly suggested early on that
the ISI /2 state would come very low in the I OLi spectrum corresponding to the
inversion in I I Be, the neighboring nucleus with seven neutrons (see Figure 1).
However, many theoretical calculations (not discussed here) had assumed that
the two valence neutrons would be in the 0P I /2 state corresponding to the normal
orbital for the last two neutrons of the P shell. This assumption had given rise
to too broad intrinsic momentum distributions.
W hat almost certainly is the p state was observed experimentally as a single
asymmetric peak in the multinucleon transfer reactions 9Be(13C,1 2N) I OLi and
13C(14C,17F) I OLj by Bohlen et al ( 1 24), who chose to interpret this state as a dou­
blet at 0.42 (1 +) MeV and 0.80 MeV (2+), and in the reaction I I BeLi,12B)IOLi
by Young et al ( 1 25), who found the shape to be consistent with a single
peak at 0.56 ± 0.06 MeV. We suspect that the poorly resolved peak observed
earlier ( 126) at 0. 1 5 ± 0. 15 MeV in the reaction "B(Jr-,p) also is the p
state.
NUCLEAR HALOS 623

Kryger et al (127) applied a technique called sequential neutron decay spec­


troscopy ( 1 28) to the IOLi problem. They found in the relative velocity spec­
trum of the fragments (9Li + n) near 0° (Figure 10) a narrow peak close to
zero velocity. They interpreted this peak as evidence for the ground state, in
which case it would most likely represent the missing s state. Alternatively,
the peak could be an s or p resonance situated near 2.5 MeV and decaying to
the first excited state of 9Li at 2.7 MeV. The first interpretation supported by
Access provided by 2802:8010:470a:5b00:ddfc:f069:d425:9566 on 11/26/22. For personal use only.

weak evidence in the spectrum measured by Young et al (125) was used by


Annu. Rev. Nucl. Part. Sci. 1995.45:591-634. Downloaded from www.annualreviews.org

Thompson &: Zhukov (129) who performed calculations for l l Li that adjusted
the energy and gave reasonable radii. They found that an approximately equal

l l LI + C -+n + 9L1 + X
I +
f
300 - Sum ( a• • . 20 rm )

f E (1.1). 0.42 MeV ,


p
..... ..... ..
C/l 200 DtrrracUon
I::
OJ
......

-
I::
....;
10 20 30 40 50
& P, (MeV/c)
1 00

o
- 4 -2 2 4

Figure 10 The relative velocity spectrum measured by Kryger et aI (127) for n + 9Li coincidence
events from the breakup of 1 8 0 at 80 MeV/u on a C target. The inset (88) is the measured
distribution at 280 MeV/u of radial momentum for neutrons from the breakup of 1 1 Li, also shown
in Figure 6. The' fully drawn curves assume that IOLi has an s ground state with scattering length
-20 fm and a p-wave resonance at 0.42 MeV and that s and p components contribute equally to
the I I Li halo. Note that the calculated velocity spectrum has not been corrected for resolution and
detection efficiency.
624 HANSEN, JENSEN & JONSON
Access provided by 2802:8010:470a:5b00:ddfc:f069:d425:9566 on 11/26/22. For personal use only.
Annu. Rev. Nucl. Part. Sci. 1995.45:591-634. Downloaded from www.annualreviews.org

1 1
Be + C -7 n + AZ + X

0.0 1 �--�--�----� �� �__�__J-__ ���____��


o 10 20 30 40 50 0 10 20 30 40 50 60
(MeV/c) Pr
(MeV/c)
Pr

Figure 11 The distributions (arbitrary scale) of radial momentum of neutrons in coincidence with
selected fragments from reactions of460-MeVlu 11 Be (88). The lOBe data. also shown in Figure 7,
represent diffraction dissociation. The two {l�lly drawn curves in the left diagram are calculations
described in the text and assuming for the s..wave scattering length as the values 0 (upper curve)
and -20 fm (lower curve). In the right diagram. the top curve (dots) shows that a low-lying p state
(E = 0.05 MeV) and a high-lying s state (as == -2 fm) cannot explain the narrow distribution.
To illustrate the effect of the 9 Li recoil term two pairs of theoretical curves are shown for as == -5
and -SO fm. with (fully drawn curves) and without the inclusion of this correction calculated for
O"Q = 100 MeV/c.

admixture of (lSI /2 ) 2 and (OPI /2 ) 2 components gave a much better agreement


with the narrow momentum distributions measured for the 9Li recoils. The
measured, very narrow neutron disbributions (42, 77), which had not been con­
sidered in previous calculations, now also fell into place as an effect of the
strong interaction in the s state and tied in with the interpretation by Barranco
et al (86).

4.4.2 SINGLE NUCLEON STRIPPING With the equipment shown in Figure 6,


I
IOLi was produced in reactions of the halo nuclei I Be and II Li with a carbon
target at energies of 460 and 280 MeV/u , respectively. In both cases, a narrow
radial neutron distribution (Figures 10 and 1 1) characterized by a r of about
36 MeV/c was found by selecting 9Li+n as the exit channel. This channel
NUCLEAR HALOS 625

must represent the state described by Kryger et al. It must also be the ground
state because stripping of a loosely bound 1 1 Li halo neutron does not excite
the core to 2.7 MeV. The properties of the I OLi ground state can now be in­
ferred from the results with the \1 Be beam (88) and used to draw conclusions
about the \1Li halo. This analysis reveals that the identity of the two widths
hides a more complex reality; both distributions are broadened considerably but
for different reasons. The gist of the argument hinges on the sudden approx­
Access provided by 2802:8010:470a:5b00:ddfc:f069:d425:9566 on 11/26/22. For personal use only.

imation. Since the halo preexists in a defined quantum state of the collision
Annu. Rev. Nucl. Part. Sci. 1995.45:591-634. Downloaded from www.annualreviews.org

complex, the experiment maps its expansion in final eigenstates of the system
eli + n).
If we consider first the results for the I I Be beam (Figure 11), it is striking
that the widths in coincidence with the three fragments lOBe, 7Li, and 9Li are
so different (r 98, 67, and 36 MeVIc, respectively). The first two values
=

were discussed in Section 4.2 as examples of diffraction dissociation and of the


core breakup mechanism, respectively. The lower set of data in the left part of
the figure represents the modification of the original I I Be s state momentum
distribution by the (n + 9Li) final-state interaction. These data may be analyzed
as follows: If the recoil imparted by the stripping of the proton can be neglected,
the halo neutron will (according to the sudden approximation) still be in an
s state (not an eigenstate of IOLi), and its decay demonstrates that the state

described by Kryger et al (127) is an s state as well. This state remains the


dominating channel, even when the recoil is taken into account.
In analogy with the Coulomb recoil Q (defined by Equation 10), we introduce
a nuclear recoil imparted by the stripping reaction. The distribution in the new
center-of-mass system eM' is now found by expanding the wavefunction of the

r rptm (r')ljr* (r')dr ' ,


reaction complex (given as in Equation 11, the "wound" is neglected):

atm
J
I I

= 18.

where the \p:c;. are continuum eigenfunctions of the eli + n) potential, adjusted
to reprodUl:e the desired scattering length as of the I = 0 virtual state and the
resonance energies in other I channels. (This parametrization is convenient
because there is no real s resonance.) The, expansion coefficients in Equa­
tion 17 determine the distribution in eM'. The distribution actually observed
is obtained by transforming back to eM and averaging over Q, taken to be
isotropic.
The lowest solid curve in the left part of Figure 11 assumes an I 0 state with
=

scattering length as -20 fm, an I 1 resonance at 0.42 MeV, and isotropic


= =

distribution of the core recoils with a Gaussian spread (JQ = 100 MeV/c. (A
scattering length of -20 fm corresponds to a virtual state at approximately
50 keY.) The curve for 7Li assumes the same parameters except for as = 0 and
626 HANSEN, JENSEN & JONSON

corresponds to a good approximation to the undisturbed momentum distribution


of the halo. In the top curve in the right part of Figure 1 1, the energies of the
I = 0, 1 states are reversed, and it is seen that the assumption of a low-lying p
state does not explain the experiment2l1 result. Other curves are used to examine
the effect of the recoil term.
The neutron distribution from I I Li measured in coincidence with a 9Li frag­
ment is shown in the inset to Figure 10. The stripping of a halo neutron is a
Access provided by 2802:8010:470a:5b00:ddfc:f069:d425:9566 on 11/26/22. For personal use only.

gentle process that imparts essentially no recoil momentum, and yet a calcula­
Annu. Rev. Nucl. Part. Sci. 1995.45:591-634. Downloaded from www.annualreviews.org

tion (88) shows that a scattering length of the order of -20 fm would lead to a
much narrower distribution than that observed.3 This problem is resolved if we
assume that the initial state contains equal amounts4 of (1s l /2 ) 2 and (OP I /2 ) 2
components so that also the excited P' state is populated. This estimate, shown
in Figure 10, is very rough: It assum(�s two final states instead of the (at least)
four actually present; the scattering length is not well known.

4.4.3 OTHER CASES The experiment. discussed here is but one example of the
experiments that can be performed with existing radioactive beams and modern
spectrometers. Another example is provided by a study of the unbound IOHe
in single-proton stripping of I l Li ( 1 31; see also 132). The state is a resonance
1 .2 MeV above threshold and has a width less than 1 .2 MeV for the decay to
(8He + 2n). The extension of continuum studies to heavier dripline nuclei is
now essentially a question of beam intensity and detection efficiency. The next
generation of experiments is expected to show much progress in this direction.

5. P DECAY

Although the momentum distributions from the breakup of halo states are at
present the main tool for studying the halo wavefunction, f3 decay can provide
information that would be difficult to obtain by other means. We give here two
illustrations of this: (a) The transition rates between discrete states provide spe­
cific and quantitative information and may be the only way to obtain such data if
the halo is an excited level. (b) The high Qp values and low particle separation
energies of the halo states allow, analogously to the reaction experiments, a
study of the coupling to the continuum.

3The low momentum of the second s neutron may, at least in part, explain why the neutrons
and the core recoils have similar momentum Wildths (see Section 4.3.2).
41t is perhaps interesting to compare the scale of the configuration mixing found here with that
of the atomic three-body problem. Calculationii for the He atom cited in Reference 130 show that
the increase in binding of about 1 .3% required for agreement with experiment implies 99.51 % s2,
0.48% p2, and 0.01 % d2. In atomic physics, the exact agreement between calculated and measured
energies provides confirmation of the model. III nuclear physics, the argumentation is necessarily
more involved.
NUCLEAR HALOS 627

In both applications, a crucial advantage is that the fJ-decay interaction is


well understood and very specific. The allowed decays manifest themselves
by full ("superallowed") strengths only for transitions that connect members of
a Wigner supermultiplet. Although this symmetry is broken by the Coulomb
force, which normally shifts the strength outside the energy window, as well as
by spin-dependent forces, these shifts are small for light nuclei. Qualitatively,
these small shifts may explain why very fast transitions are encountered sys­
Access provided by 2802:8010:470a:5b00:ddfc:f069:d425:9566 on 11/26/22. For personal use only.

tematically in the isotopes 6,8He and 9, I l Li ( 133-1 37). Sagawa et al ( 1 38) have
Annu. Rev. Nucl. Part. Sci. 1995.45:591-634. Downloaded from www.annualreviews.org

discussed a specific model accounting for these shifts. In another picture of fJ


decay of a neutron halo, the spatial separation of the halo neutrons from the
core enables them to decay essentially as free particles, so that, for example, a
dineutron will form a deuteron. This "submultiplet" model is the weak inter­
action equivllient of the emergence (see Section 2) of electric dipole transitions
of full single-particle strength.
5.1 Transitions Between Discrete States
The well-known superallowed fJ transition between the ground states of 6He
and 6Li is an example of a transition between a bound dineutron and a bound
deuteron. This transition exhausts about 80% of the Gamow-Teller (GT) sum
rule value of 6, a result that can be accounted for by a three-body model (1).
Poves et al (139) have suggested that the as yet undiscovered and very likely
unstable nudeus 280 would have a large dineutron component in the ground
state, decaying to a high-lying deuteron cluster in 28F.
The first excited state in 17p at 495 keY was discussed in Section 2. This
case is not available to radioactive beam experiments and has been studied
experimentally (140) in the first-forbidden fJ decay of 17Ne (Figure 1 2). The
intensity of the y transition gives a branching ratio of ( 1 .65 ± 0. 16)%, which
is almost a factor two above what would be calculated from the well-known
mirror decay of 17N to the first excited state of 170. This is the largest mirror
asymmetry recorded to date for a bound state and can be explained (140) in
terms of the structure of the proton states in 17Ne and 17F.
In Sectioll1 4.4 we discussed various estimates that indicate that the halo of
II
Li is a superposition of roughly equal amounts of 0P l /2 and lS /2. Suzuki &
1
Otsuka (141l) have proposed that such information may be obtained indepen­
dently from the GT transition 1 1 Li(312- , g.s.) --+ 1 1 Be( 1/2- , 320 keY). They
performed a calculation that took into account the influence of the halo, meson
exchange currents, and mixing of the sd-shell configurations on the f3 decay
rate and found that these three effects contribute coherently to the observed
logft value:. The largest contribution comes from the halo, and these authors'
calculation suggests, in agreement with other results cited above, that 50-
60% of (OP /2)2 configuration is found in the I l Li ground-state wavefunction
I
628 HANSEN, JENSEN & JONSON
Access provided by 2802:8010:470a:5b00:ddfc:f069:d425:9566 on 11/26/22. For personal use only.
Annu. Rev. Nucl. Part. Sci. 1995.45:591-634. Downloaded from www.annualreviews.org

0 - 1 sec

1 - 4 sec 495 _.....,..._.J

1 00
( keY)

495 keV
5 1 1 keY

sao 520

Ener,gy ( keV)

Figure 12 Y spectrum (140) from the f3 deca,y of l7 Ne (T1 /2 = 109 ms) studied at the ISOLDE
PSB Facility at CERN. A single I -GeV proton ]pulse, containing about 3 . 1 0 13 particles, bombarded
a MgO target every 4 . 8 s. The y spectrum measured during the first second after the pulse is shown
as a histogram. The background recorded between the pulses is shown as a dotted line. The inset

shows the relevant part of the l7 Ne decay scheme.


NUCLEAR HALOS 629

based on an experimentally determined feeding to the 320-keV state of 9%


( 142). The forbidden f3 transition to the 1 1 Be ground state may actually be
an independe:nt solution to the same problem, but the experiment is very diffi­
cult.

5.2 fJ-Delayed Particles


Access provided by 2802:8010:470a:5b00:ddfc:f069:d425:9566 on 11/26/22. For personal use only.

The halo nudei have large energy windows for f3 decay, and the separation
energies for particles or clusters of particles in the daughter nuclei are low.
Annu. Rev. Nucl. Part. Sci. 1995.45:591-634. Downloaded from www.annualreviews.org

Therefore, the decays proceed via emission channels in the continuum, which
are often quite complex and which are referred to as f3-delayed particles ( 143).
The first case of f3-delayed deuterons was that of 6He, which had a measured
branching ratio of only (7.6 ± 0.6) - 1 0-6 (133, 1 36), whereas an R-matrix anal­
ysis yields a much larger value ( 1 33). Several theoretical studies of this decay
have been performed (144-147), and the spectral shape at high deuteron ener­
gies is well reproduced in the calculations. This result is not surprising, since
the shape to a large extent is given by phase-space and penetrability factors.
The low intensity of the deuteron branch appears to arise from a cancellation
effect ( 1 36, 148). Assuming a decay to the virtually excited ground state of
6Li, which subsequently breaks up into an ex particle and a deuteron, Borge
et al (1 36) suggested a second decay route with opposite ordering, i.e. an initial
decay of 6He: into a virtual ex + 2n state followed by a f3 decay of the dineu­
tron to a deuteron. The interference introduced within this model explains the
intensity of the deuterons. Barker (148) used a single-level R-matrix treatment
to show that this picture becomes reasonable when the relative contributions
from the internal and external regions to the GT matrix element are taken into
account. Experimental data extending to lower energies could help clarify the
situation.
Another interesting case is 8He, which has a surprisingly strong (8.0 ±
0.5) · 10-3 branch ( 136) of f3 -delayed tritons (Figure 13). This nucleus, with its
two-neutron separation energy of2. 14 MeV and four-neutron separation energy
of 3. 1 1 MeV, is a candidate for a neutron skin. R-matrix calculations (1 36, 1 37)
show that triton emission proceeds essentially via a single relatively narrow 1 +
state at 9.3-MeV excitation in sLi. The f3 feeding determined from the triton
branch alone gives a reduced GT transition probability BGT of 5. 18 correspond­
ing to almost one half of the GT sum rule. This result indicates then that the
8He ground state has a large overlap with an ex particle and a neutron cluster.
Zhukov et al (149) recently proposed a five-body cluster orbital shell model
approximation to describe the ground-state wavefunction of SHe as ex + 4n. An
analytical neutron correlation function between the four neutrons was derived.
This correlation function shows a number of maxima that represent the most
630 HANSEN, JENSEN & JONSON

240
Access provided by 2802:8010:470a:5b00:ddfc:f069:d425:9566 on 11/26/22. For personal use only.
Annu. Rev. Nucl. Part. Sci. 1995.45:591-634. Downloaded from www.annualreviews.org

200

1 60
>
v
.!I::
\()
C'l
....... J
C/l 1 20
.....
t: �
::l

j
0
U

80
l
1
40
1

o 500 1 000 1 500 2000 2500 3000

Triton Energy (keV)

Figure 13The experimental and calculated ,B-delayed triton spectrum from 8 He (1 36). The
intensity of the ,B-delayed triton branch is (8.0 ± 0.5) . 10- 3 . The continuous line shows a single­
level R-matrix fit to the spectrum assuming that the decay proceeds via one intermediate state at
8
9.3 MeV excitation in U. The inset shows two configurations with maximal probability for the
angular part of the spatial correlation function according to Reference 1 36. These are two examples
of configurations with high probability for ,B decay to the triton channel.
NUCLEAR HALOS 63 1

probable configurations (inset of Figure 13), which resemble clusters of three


or four neutrons. These configurations might explain the large delayed triton
branch.

6. CONCLUDING REMARKS
At the time of this writing, research involving the halo neutron is progress­
Access provided by 2802:8010:470a:5b00:ddfc:f069:d425:9566 on 11/26/22. For personal use only.

ing very rapidly. A number of working hypotheses that just two years ago
Annu. Rev. Nucl. Part. Sci. 1995.45:591-634. Downloaded from www.annualreviews.org

seemed more or less generally accepted have turned out to be wrong or in­
complete, and it would, by extrapolation, seem that more surprises must be in
store. A few firm values for quantities that relate directly to the halo structure,
such as certain momentum widths and f3 decay rates, have been established,
and we have begun to understand how other values, seemingly of a similar
nature, are filtered and distorted by the reaction mechanisms in nuclear colli­
sions and ,B decay. The final-state interactions in fragmentation (Section 4.4)
and the quenching of the deuteron branch from 6 He (Section 5 .2) are cases
in point. Armed with accumulating experience and an already good level
of theoretical understanding, we should in the coming years be ready to ap­
proach thl! problem of correlations in the wavefunctions. However, if this
subject is to appeal to physicists outside of nuclear physics, it will require more
than the confirmation of an overcomplicated experiment by an overcomplicated
theory. We hope that such a demonstration will be technically possible in the
future.
An exciting possibility in the longer term is the existence of giant halos
(Section 3 .4), marginally bound excited states of the three-body system with
sizes running into hundreds of fermis. Theory has yet to tell us where it would
pay to look, and experiment has provided only vague ideas of how to go about
such a search, although cross sections of many thousands of barns could ease
the task of detection. This problem poses a great challenge.

ACKNOWLEDGMENTS

We are indebted to Nigel A Orr, Karsten Riisager, Jan S Vaagen, and Mikhail
V Zhukov for many valuable discussions. We have benefited from discussions
with T Andersen, RH Dalitz, and TA Kaplan about halo analogues. PGH
appreciates support from Aarhus Universitets Forskningsfond as well as the
hospitality extended to him by CERN in 1994-95.

ny Annual Review chapter, as well as any article cited in an Annual Review chapter,
may be purchasedfrom the Annual Reviews Preprints and Reprints service.
1-800-347-8007; 415-259-5017; email: arpr@class.org
L
632 HANSEN, JENSEN & JONSON

Literature Cited

1 . Zhukov MV, et aI. Phys. Rep. 23 1 : 1 5 1 32, Tanihata I, et aI. Phys. Rev. Lett. 55:2676
( 1 993) ( 1 985); Phys. Lett. B 1 60:380 ( 1 985)
2 . Bertulani CA, Canto LF, Hussein MS. 33. Fukuda M, et aI. Phys. Lett. B 268:339
Phys. Rep. 226:281 ( 1 993) ( 1 99 1 )
3 . Geissel H, Miinzenberg G. Riisager K. 34. Blank B, e t al. Nucl. Phys. A 555:408
Annu. Rev. Nucl. Part. Sci. 45: 1 63-203 ( 1 993); Blank B, et aI. Z. Phys. A 340:41
( 1 995) ( 1 99 1 )
4. Hansen PG. Annu. Rev. Nucl. Part. Sci. 35. Arnold E , e t aI. Phys. Lett. B 197:3 1 1
Access provided by 2802:8010:470a:5b00:ddfc:f069:d425:9566 on 11/26/22. For personal use only.

29:69 ( 1 979) ( 1 987)


36. Arnold E, et aI. Phys. Lett. B 28 1 : 1 6
Annu. Rev. Nucl. Part. Sci. 1995.45:591-634. Downloaded from www.annualreviews.org

5. Detraz C, Viera DJ. Annu. Rev. Nucl. Part.


Sci. 39:407 ( 1 989) ( 1 992)
6. Tanihata I. See Ref. 7, p. 443 ( 1 989) 37. Hansen PG, Jonson B. Europhys. Lett.
7. Bromley DA. Treatise on Heavy-Ion Sci­ 4:409 ( 1 987)
ence, Vol. 8. New York: Plenum ( 1989) 38. Bertulani CA, BauT G. Nucl. Phys. A
8. Mueller AC, Sherrill BM. Annu. Rev. 480: 6 1 5 ( 1 988)
. Nue/. Part. Sci. 43:529 ( 1 993) 39. Kobayashi T, et al. Phys. Lett. B 232:5 1
9. Ogloblin AA, Penionzhkevich, YEo See ( 1 989)
Ref. 1 37, p. 26 1 ( 1 989) 40. Blank B, et aI. Z. Phys. A 343:375 ( 1 992)
1 0. Riisager K. Rev. Mod. Phys. 66: I I O�i 4 1 . Kobayashi T, et al. Phys. Rev. Lett.
( 1 994) 60:2599 (1988)
1 1 . Tanihata 1. Progr. Nucl. Part. Phys. In 42. Anne R, et aI. Phys. Lett. B 250: 19 ( 1 990)
press ( 1 995) 43. Orr NA, et al. Phys. Rev. Lett. 69:2050
1 2. Tanihata l . Nucl. Phys. A 520:4 1 1 c ( 1 990). ( 1 992)
A522:275c ( 1 99 1 ) 44. Batty 0, et aI. Adv. Nucl. Phys. 1 9 : 1 ( 1 989)
1 3 . Kobayashi T. Nue/. Phys. A 538:343c 45. Tanihata I, et al. Phys. Lett. B 289:261
( 1 992), A553:465c ( 1 993) ( 1 992)
1 4. Hansen PG. Nue/. Phys. A 553:89c ( 1 993) 46. Andersen T. Phys. Scr. T34:23 ( 1 99 1 )
1 5 . Jonson B. Nue/. Phys. A 574 : 1 5 I c; Nuel. 47. Levy-Leblond JM. Phys. Rev. 1 5 3 : 1
Phys. A 583:733 ( 1 995) ( 1 967)
1 6. Orr NA. Proc. CLUSTER-94, Strasbourg, 48. Lykke KR, Mead RD, Lineberger We.
France, 1 994. To be published ( 1 995) Phys. Rev. Lett. 52:2221 ( 1 984)
1 7. Oppenheimer JR. Phys. Rev. 47:845 49. Desfrancois C, et aI. Europhys. Lett. 26:25
( 1 935) ( 1 994)
1 8. Dancoff SM. Phys. Rev. 72: 1 63, 1 0 1 7 50. Rencsoc R, Kaplan TA, Harrison JF. 1.
( 1 947) Chern. Phys. 98:9758 ( 1 993)
1 9. Serber R. Phys. Rev. 72: 1 008 ( 1 947) 5 1 . Efimov VM. Comments Nucl. Part. Phys.
20. Glauber RJ. Phys. Rev. 99: 1 5 1 5 ( 1 955) 19:271 ( 1 990)
2 1 . Fwdt G. Phys. Rev. D 2:846 ( 1 970) 52. Efimov VM. Sov. 1. Nucl. Phys. 12:589
22. Bohm G, Wysotzki F. Nucl. Phys. B ( 1 970)
1 5 : 628 ( 1 970); Bohm G, et aI. Nucl. Phys. 53. Cornelius T, GWckle W. 1. Chern. Phys.
B 4:5 1 1 ( 1 968) 85:3906 ( 1 986)
23. Lyuboshits VL. Sov. 1. Nucl. Phys. 5 1 :648 54. Dobaczewski J, et aI. Phys. Rev. Lett.
( 1 990) 72:981 ( 1 994)
24. Congleton JG. 1. Phys. G 1 8 :339 ( 1 992) 55. Tosaka Y, Suzuki Y. Nucl. Phys. A 5 12:46
25. Miyagawa K, G10ckle W. Phys. Rev. C ( 1 990)
48:2576 ( 1 993) 56. Varga K, Suzuki Y, Lovas RG. Nucl. Phys.
26. Rolfs e. Nucl. Phys. A 2 1 7:29 ( 1 973) A 571 :447 ( 1 994); Varga K, Suzuki Y,
27. Alhassid Y, Gai M, Bertsch G. Phys. Rev. Ohbayasi Y. Phys. Rev. C 50: 1 89 ( 1 994)
Lett. 49: 1 482 ( 1 982); Gai M, et aI. Phys. 57. Baye D, Suzuki Y, Descouvemont P. Prog.
Lett. B 2 15:242 ( 1 989) Theor. Phys. 9 1 :27 1 ( 1 994)
28. Millener OJ, et aI. Phys. Rev. C 28:497 58. Riisager K, Jensen AS, M!'Iller P. Nucl.
( 1 983) Phys. A 548:393 ( 1 992)
29. Uchiyama T, Morinaga H. Z. Phys. A 59. Fedorov DV, lensen AS, Riisager K. Phys.
320:273 ( 1 985) Rev. C 49:20 1 ( 1 994)
30. Hofmann DC. Ark. Fys. 36:533 ( 1 967) 60. Fedorov DV, lensen AS, Riisager K. Phys.
3 1 . Otsuka T, et aI. Phys. Rev. C 49: R2289 Rev. C 50:2372 ( 1 994)
( 1 994) 6 1 . Fedorov DV, Jensen AS, Riisager K. Phys.
NUCLEAR HALOS 633

Lett. B 3 12: 1 ( 1 993) 90. Zhukov MV. Jonson B. Nucl. Phys. A


62. Richard J-M. Fleck S. Phys. Rev. Lett. 589: 1 ( 1 995)
7 3 : 1 464 ( 1 994) 9 1 . Sagawa H, Brown BA. Esbensen H. Phys.
'
63. Fukunishi N. Otsuka T. Tanihata I. Phys. Lett. B 309 : 1 (1993)
Rev. C 48: 1648 ( 1 993) 92. Otsuka T. Pukunishi N. Sagawa H. Phys.
64. Jensen AS. Riisager K. Phys, Lett. B Rev. Lett. 79: 1 385 ( 1 993)
264:238 (199 1 ) ; Nuc!. Phys. A 537:45 93. Zwieglinski B. et al. Nucl. Phys. A
( 1 992) 3 1 5 : 124 ( 1 979)
65. Fedorov lOV. Jensen AS. Phys. Rev. Lett. 94. Nunes PM, Thompson IJ. Johnson RC . J.
7 1 :4 1 03 ( 1 993); Fedorov DV. Jensen Phys. G. In press ( 1 995)
Access provided by 2802:8010:470a:5b00:ddfc:f069:d425:9566 on 11/26/22. For personal use only.

AS. Riisager K. Phys. Rev. Leu. 73:281 7 95. Vinh Mau N. IPNO-TH 94190; and to be
( 1994) published ( 1 995)
Annu. Rev. Nucl. Part. Sci. 1995.45:591-634. Downloaded from www.annualreviews.org

66. Kolata JJ. et aI. Phys. Rev. Lett. 69:263 1 96. Muta A. Otsuke T. RIKEN-AF-NP-188;
( 1 992) and to be published ( 1 995)
67. Lewitowicz M. et aI. Nucl. Phys. A 97. Bu1boaca I. Lassaut M, Lombard RP. Or­
562:301 ( 1 993) say preprint; and to be published ( 1995)
68. Moon CoB. et al. Phys. Lett. B 297:39 98. Winter A, Alder K. Nucl. Phys. A 3 1 9 : 5 1 8
( 1 993) ( 1979)
69. Anne R. et al. Nucl. Phys. A 575 : 125 99. Kelley JH. et al. Phys. Rev. Lett. 74:30
( 1 994) ( 1 995)
70. Sagawa H, Takigawa N, Giai NY. Nucl. 100. Bertulani CA, McVoy KV. Phys. Rev. C
Phys. A 543:575 ( 1 992) 46:2638 ( 1 992)
7 1 . Nilsson T. et al. Europhys. Lett. 30: 1 9 1 0 1 . Sagawa H, Takigawa N. Phys. Rev. C
( 1 995) 50:985 ( 1 994)
72. Serber R. Annu. Rev. Nucl. Part. Sci. 44: 1 102. Barranco p. Vigezzi E. Broglia RA.
( 1 994) Momentum Distributions in Halo Nu­
73. Riisager K, et al. Nucl. Phys. A 540:365 clei, NTGMI-95-2; and to be published
( 1 992) ( 1 995)
74. Anne R. etal. Phys. Lett. B 304:55 ( 1 993) 103. Hansen PG. See Ref. 150 ( 1 995)
75. Nakamura T, et al. Phys. Lett. B 3 3 1 :296 104. Bazin D, et al. MSUCL-968; and to be
( 1 994) published ( 1 995)
76. Ieki E, et al. Phys. Rev. Lett. 70:730 ( 1993) 105. Angelique FC. et aI. See Ref. 150 ( 1 995)
77. Sackett D. et al. Phys. Rev. 48: 1 1 8 ( 1993) 106. Audi G. Wapstra AH. Nucl. Phys. A 565: I
78. Shimoura S. et al. Phys. Lett. B 348:29 ( 1 993)
( 1 995) 107. Riisager K. Jensen AS. Phys. Lett. B 30 1 :6
79. Peshbal:h H. Theoretical Nuclear Physics, ( 1 992)
Nuclear Reactions. New York: Wiley & 108. Schwab W, et al. Z. Phys, A 350:283
Sons ( 1 99 1 ) ( 1 995)
80. Souliotis GA, e t al . Phys. Rev. C 46: 1 383 109. Young 8M, et al. Phys. Rev. Lett. 7 1 :4 1 24
( 1 992) ( 1 993)
8 1 . Morrisey DJ. Phys. Rev. C 39:460 ( 1 989) 1 1 0. Orr NA,etal. Phys. Rev. C 5 1 :3 1 1 6 ( 1 995)
82. Kidd JM. et al. Phys. Rev. C 37:26 1 3 I I I . Tanihata I. Phys. Lett. B 287:307 ( 1 992)
( 1 988) 1 1 2. Korsheninnikov AA. Kobayashi T. Nucl.
83. Gelbke CK. et al. Phys. Lett. B 70:4 1 5 Phys. A 567:97 ( 1 994)
( 1 977) 1 1 3 . Typel S. Baur G. Nuc!. Phys. A 573:486
84. Goldhaber AS. Phys. Lett. B 53:306 ( 1 994)
( 1 974) 1 /4. Baur G, Bertulani CA. Kalassa DM. Nucl.
85. Friedman WA. Phys. Rev. C27:569 ( 1 983) Phys. A 550:527 (1 992)
86. Barrarlco P, Vigezzi E. Broglia RA. Phys. 1 / 5. Bertsch GP, Bertulani CA. Nucl. Phys. A
Lett. B 3 1 9:387 ( 1 993) 556: 1 36 ( 1 993)
87. Humbert F. et al. Phys. Lett. B 347: 1 98 1 1 6. Zahar M. et al. Phys. Rev. 48:R 1 484
( 1 995) ( 1 993)
88. Zinser M. et al. Phys. Rev. Lett. 75: 1 7 1 9 1 1 7. Descouvemont P. Phys. Lett. B 3 3 1 :271
( 1 995) ( 1 994)
89. Riisager K. Residence in Forbidden Re­ 1 1 8. Korsheninnikov AA. et a), Europhys. Lett.
gions. Summary paper of thesis defended 29:359 ( 1 995)
for th� Doctorate of Science degree, Fac­ 1 19 . Neumaier S, et aI' Proc. Int. ConI
ulty of Science, Aarhus Univ., p. 76 Nucleus-Nucleus Collisions, 5th.
(1994) Taormina, Italy, 1994 ( 1994)
634 HANSEN, JENSEN & JONSON

120. Barker FC, Hickey GT. 1. Phys. G 3:L23 1 36. Borge MIG, et aI. Nucl. Phys. A 560:664
( 1 977) (1993)
1 2 1 . Hees AGM, Glaudemans PWM. Z. Phys. 1 37. Barker FC, Warburton EK. Nucl. Phys. A
A 3 1 5:223 ( 1 984) 487:269 (1988)
122. Johannsen L, Jensen AS, Hansen PG. 1 38. Sagawa H, Hamamoto I, Ishihara M. Phys.
Phys. Lett. B 244:357 ( 1990) Lett. B 303:2 1 5 (1993)
123. Zhukov MV, et aI. Phys. Lett. B 265: 1 9 139. Poves A, et aL Z. Phys. A 347:227 ( 1994)
( 1 99 1 ) 1 40. Borge MJG, et aI. Phys. Lett. B 3 1 7:25
124. Bohlen HG, e t aI. Z. Phys. A344:381 ( 1 993)
( 1 993) 1 4 1 . Suzuki T, Otsuka T. Phys. Rev. C 50:R555
Access provided by 2802:8010:470a:5b00:ddfc:f069:d425:9566 on 11/26/22. For personal use only.

125. Young BM, et aI. Phys. Rev. C 49:279 ( 1994)


( 1994) 142. Bjornstad T, et aI. Nucl. Phys. A 359:1
Annu. Rev. Nucl. Part. Sci. 1995.45:591-634. Downloaded from www.annualreviews.org

126. Amelin AI, et aI. Sov. J. Nucl. Phys. (1981)


52:782 ( 1 990) 143. Hansen po, Jonson B. In Particle Emis­
127. Kryger RA, et aI. Phys. Rev. C 47:R2439 sion from Nuclei, ed. DN Poenaru, M
( 1 993) Ivascu, 3: 157 ( 1 988)
128. Deale F, et aI . Nucl. Instrum. Methods A 144. Descouvemont P, Leclercq-Willain C. 1.
258:67 ( 1 987) Phys. G 1 8:L99 (1992)
129. Thompson 11, Zhukov MV. Phys. Rev. C 145. Zhukov MV, et aI. Phys. Re v. C 47:2937
49: 1904 (1 994) ( 1 993)
1 30. Slater JC. Quantum Theory of Matter; :p. 146. Cs6t6 A, Baye D. Phys. Rev. C 49: 8 1 8
434. New York: McGraw-Hill (1968) ( 1 994)
1 3 1 . Korsheninnikov AA, et aI. Phys. Lett. B 147. Varga K, Suzuki Y. Ohbayasi Y. Phys. Rev.
326:31 ( 1 994) C 50: 1 89 (1994)
1 32. Ostrowski AN, et at. Phys. Lett. B 338: 1 3 148. Barker FC. Phys. Lett. B 322 : 1 7 ( 1 994)
( 1 994) 149. Zhukov MV, Korsheninnikov AA, Smed­
133. Riisager K, et aI. Phys. Lett. B 235:30 berg M. Phys. Rev. C 50:RI ( 1994)
( 1 990) ISO. Guillemaud-Mueller D, ed. Proc Int.
1 34. Nyman G, et aI. Nucl. Phys. A 5 1 0 : 1 89 Con/. Exotic Nuclei Atomic Masses, Ar­
(1 990) Ies. France. June 19-23, 1995. In press
1 35. Borge MJG , et al. Z. Phys. 340:255 (1991) ( 1 995)

You might also like