Nuclear reactions, such as nucleon transfer, have been extensively
used in the past with stable ion beams as excellent tools to explain the
single-particle structure in nuclei and will continue to provide new detailed
spectroscopic information to understand the evolution of shell structure
far from stability with the availability of neutron-rich radioactive beams
of SPIRAL. At higher bombarding energies, and in particular for the study
of the loosely bound systems, a very promising and powerful complementary
spectroscopic tool is provided by nucleon-removal reactions for identifying
single-particle structure. Also referred to as ‘knockout’ reactions, these
direct reactions measure the probability of nucleon removal from the projectile
by observing the surviving beam-like fragments. A light target nucleus
acts as an absorptive disk, so that the nucleon is cleanly removed from
the projectile. The ground state of the projectile can be characterised
in terms of single-particle structure by the associated spectroscopic factor.
These reactions are studied in the intermediate energy regime ensuring
that the interaction with the target is peripheral. The nucleon is removed
by the interaction between the tail of its wavefunction, where it extends
beyond the core of the projectile nucleus, and the target nucleus. This
makes it obvious that this mechanism is an excellent probe of halo nuclei
and other weakly bound nuclei such as light nuclei near the neutron drip-line.
In these cases, the one(two)-neutron removal cross sections are about two(three)
orders of magnitude higher than for one(two)-neutron transfer reactions.
The applicability of knockout reactions remains to be clearly demonstrated
over a wide range of masses and binding energies, but for more bound systems
the cross sections will be closer to those for transfer.
The technique has been developed in the last five years
in conjunction with in-flight separated radioactive beams from fragmentation
reactions and used successfully, at energies above ~ 40 MeV/nucleon, to
measure spectroscopic factors in exotic p,sd-shell nuclei such as 8B [1,2],
14B, 11,12Be, 15,16,17,19C and 26,27,28P [3]. The projectile residues from
nucleon-removal reactions are generally observed using a high-resolution
spectrograph and the bound final states of the heavy residues are identified
by the associated g decays. The partial removal cross section, analysed
in eikonal reaction theory [4,5,6], determines the spectroscopic factor
while the shape of the measured longitudinal momentum distribution determines
the orbital angular momentum l. The method, originally developed at NSCL/MSU,
has an extremely high sensitivity, a great asset for experiments with radioactive
beams of very exotic nuclei (experiments have been carried out with intensities
as low as 1 ion/second [7]). A whole area of the nuclear chart, very far
from stability, is now open for study with this new technique. Very recently
the way has even been opened for the use of two-proton knockout reactions
[8,9]. The two-proton removal reaction is of special interest since it
leads straight ‘south’ in the nuclear chart, potentially to rare isotopes
that are difficult to access by other means, and because it promises a
handle on two-proton correlations in exotic nuclei.
A recent review article [3] has examined the agreement
between experimental and theoretical spectroscopic factors for the approximately
25 partial cross sections measured so far with the nucleon-removal technique
at NSCL. The result, although with a large error, was a scale factor not
very different from unity. Intuitively this is what one would expect for
neutron-halo states in which the neutron resides with a high probability
outside the core of the nucleus and may be thought as a quasi-free neutron.
This data suggests that the picture of a universal quenching factor close
to 0.5 for all nuclei independent of mass may not be the full story. More
precise and new experimental data, especially on exotic nuclei, is clearly
needed to investigate further this topical problem.
This technique, using nucleon-knockout reactions and
the detection of coincident g-rays, has become a major and most powerful
tool at the NSCL, GSI and RIKEN facilities to obtain detailed spectroscopic
information on nuclei far from stability. GANIL, with the SISSI device,
the SPEG spectrometer and EXOGAM array, constitutes an ideal facility to
perform spectroscopy of light neutron-rich nuclei using this technique
and would build on earlier surveying inclusive measurements [10,11].
Proposed and approved experiment
One puzzle in the region of light neutron-rich nuclei, which
is currently receiving a great deal of interest from both the theoretical
and experimental communities, is the particle instability of oxygen isotopes
starting from 25O, whereas the fluorine isotopes, with just one more proton,
are bound up to 31F. The reduction of the N = 20 shell gap [12] and
the appearance of a new shell gap at N = 16 [13] have been invoked to explain
this effect, as suggested by neutron separation-energy systematics, however
a quantitative understanding is still missing, for which detailed spectroscopic
information in this region is needed. We therefore propose to take advantage
of the high-intensity 36S primary beam at GANIL to produce intermediate-energy
light neutron-rich fragmentation radioactive beams with the SISSI device
in order to perform nucleon-removal reactions using the high-resolution
SPEG spectrometer and measuring coincident g-rays. This will enable us
to extract spectroscopic factors in this region and shed light on the N
= 16 shell closure. This proposal focuses on the one-neutron removal reactions
from 23O and 25F but data on the neighbouring nuclei will be obtained simultaneously
as demonstrated in ref. [10,11].
Up to now, for the oxygen and fluorine isotopes near
N = 14, 15, 16 inclusive nucleon-knockout measurements have been performed
up to 24,25F and 23O, using the fragmentation of a 70 MeV/nucleon 40Ar
beam at GANIL by Sauvan et al [10,11]. The fragmentation of 36S will produce
higher yields than with 40Ar for the most neutron-rich nuclei so that a
detailed study of the structure of 23O and 25F can be performed as well
as new data extended to 24O and 26F. In the case of 23O and 25F with relatively
high neutron separation energies, the measured core momentum distributions
are clearly narrower than in the case of their neighbours due to the large
n2s1/2 admixtures expected in their ground state (see Figure 1) [10].
Recently a new measurement of inclusive one- and two-neutron
removal reactions of 23O was performed at RIKEN at 72 MeV/nucleon [14],
concluding to a Ip = 5/2+ ground state for 23O, in contradiction with the
shell-model prediction, which is interpreted by the authors by introducing
a structural change in the 22O core. These conclusions were recently contradicted
by Brown et al [15] who could explain the measured one-neutron removal
cross section of Kanungo et al [14] by performing a standard shell-model
and Glauber analysis which does not require any 22O core modification.
The authors of ref. [11] and ref. [16] also contradict these claims. Brown
et al [15] in particular underline the need for the measurement of
partial cross sections to individual final states from g-ray coincidences,
which the present proposal aims to perform.
A series of recent experiments at GANIL performed in-beam
g-ray spectroscopy in this region, using the fragmentation of 36S and second-step
fragmentation of the secondary beams subsequently produced, to study the
structure of excited states in 18,20C, 21-24O and 23,25F [17-22]. Level
schemes are suggested for 21,22O but no g-rays were observed from 23O and
24O, suggesting that their excited states lie above the neutron decay thresholds.
Following a recent experiment performed at GSI at 938 MeV/nucleon [16],
the measurements of Sauvan et al [10,11] for the longitudinal momentum
distribution were confirmed, in agreement with a Ip = 1/2+ ground state
for 23O with a dominant n2s1/2 Ä 22O(0+g.s.) single particle component.
The measurement of partial cross sections to individual final states in
22O was attempted by coincident g-ray detection using NaI detectors. Unfortunately
the energy resolution of the NaI detectors did not allow the 2.4 and 3.2
MeV g-rays in 22O to be resolved. Preliminary results on the spectroscopic
factor extracted for the sum of the excited states indicate a clear disagreement
with the shell-model predictions of ref. [15]. New experimental data with
better statistics and improved g-ray resolution is thus necessary to investigate
this issue.
The investigation of the fluorine isotopes promises to
be even more exciting. In the case of 25F, very little spectroscopic information
is known. So far the level scheme for 23F is known from a measurement using
the multinucleon transfer reaction 22Ne(18O,17F)23F [23], and a level scheme,
obtained from in-beam g-ray spectroscopy, has recently been proposed [20].
Excited states in 24F were identified for the first time in a b-g coincidence
experiment on LISE [24]. A first level scheme for 25F is reported in [20]
and [22]. The measurement of Sauvan et al [7,8] for the longitudinal momentum
distribution of 25F are confirmed by the GSI data, showing a narrow width.
Neither of these measurements included coincident g-ray detection,
which we propose to undertake in this new experiment in order to investigate
the possible increase of the contribution from n1d5/2. Detailed spectroscopic
information about 25F, located at N = 16, will be of crucial importance
to understand quantitatively the change in nuclear structure occurring
between the oxygen and fluorine isotopic chains and the possible appearance
of a new shell gap at N = 16.
Experimental details
For this type of measurements with knockout reactions the
energy of the secondary beams needs to be the highest possible, i.e corresponding
to the highest possible magnetic rigidity of the transfer line (Br = 2.88
Tm). The yields of the secondary beams produced in the fragmentation of
a 77.5 MeV/nucleon 36S primary beam with high-intensity (~ 1 kW) on a C
target using SISSI and transmitted to the SPEG spectrometer are estimated
to be approximately 60 pps for 23O (optimum C target thickness 590 mg/cm2)
and 200 pps for 25F (C target thickness 530 mg/cm2), to be compared with
about 1 pps for 25F in the experiment performed by Sauvan et al [10,11,25].
These yields have been estimated using the LISE code and corrected to take
account of rates measured in recent experiments.
The secondary beams will be transported to the SPEG spectrometer
where the reactions will take place on a light secondary target (~ 150
mg/cm2 of C). The incident secondary ions will be tracked in small drift
chambers positioned between the SPEG analyser and the secondary reaction
target in order to correct the measured outgoing-fragment scattering angle
by the incident beam angle (this should, as demonstrated in ref. [10,11],
allow us to reconstruct the transverse momentum distributions). The high-resolution
energy-loss spectrometer SPEG operated in a dispersion-matched energy-loss
mode will allow the measurement of the momentum distributions of the fragments
and of the nucleon-removal cross sections (a resolution in the momentum
measurements of dp/p = 3.5´10-3 was obtained in ref. [10]). The fragments
will be identified in the focal plane of SPEG using the standard detection
systems (energy-loss measurement in an ionisation chamber and residual
energy measurement as well as time-of-flight measurement in a plastic detector).
The longitudinal momentum distributions will be determined from the position
measurements using the two standard large-area stripped-cathod drift chambers.
The nucleon-removal cross sections will be determined from the number of
detected ions which have reacted and the number of incident ions (for which
monitoring of the incident beam intensity will be provided by the beam-tracking
drift chambers). One can assume a nearly 100 % acceptance of the core fragments
in SPEG [10]. To detect the prompt coincident g-rays we propose to use
the EXOGAM germanium-detector array located, in a close geometry, around
the secondary reaction target. A mechanical support similar to the one
developed to hold EXOGAM detectors at the end of LISE is envisaged. The
photopeak efficiency of the so-called Gamma-Cube configuration (full suppression
shield, 4 CLOVERS at 68.3 mm) is 10 % at 1.3 MeV [26]. The experiment could
also be performed with the Chateau de Cristal barium fluoride array, in
particular for the detection of the higher-energy g-rays. In the recent
in-beam spectroscopy experiments of [17-22] performed with SPEG,
the g-rays were detected using 74 BaF2 detectors of the Chateau de Cristal
and 4 Ge detectors, with an efficiency of 25 % at 1.3 MeV. Simulations
will be run to determine the optimum g-ray detector arrangement to combine
good efficiency and energy resolution.
References
[1] D. Cortina Gil et al, Phys. Lett. B 529 (2002) 36.
[2] D. Cortina Gil et al, Nucl. Phys. A 720 (2003) 3.
[3] P.G. Hansen et al, Nucl. Phys. A 693 (2001) 133,
section 4 and references therein.
[4] J.A. Tostevin, J. Phys. G 25 (1999) 735.
[5] J.A. Tostevin, Nucl. Phys. A 682 (2001) 320.
[6] J.A. Tostevin et al, Phys. Rev. C 66 (2002) 024607.
[7] V. Maddalena et al, Phys. Rev. C 63 (2001) 024613.
[8] D. Bazin et al, Phys. Rev. Lett. 91 (2003) 012501.
[9] D. Warner, Nature 425 (9 Oct. 2003) 570.
[10] E. Sauvan et al, Phys. Lett. B 491 (2000) 1.
[11] E. Sauvan et al, in press in Phys. Rev. C, arXiv:nucl-ex/0307019.
[12] Y. Utsono et al, Phys. Rev. C 60 (1999) 054315.
[13] T. Otsuka et al, Phys. Rev. Lett. 87 (2001) 082502.
[14] R. Kanungo et al, Phys. Rev. Lett. 88 (2002) 142502.
[15] B.A. Brown et al, Phys. Rev. Lett. 90 (2003) 159201.
[16] D. Cortina Gil et al, Frontiers of Nuclear Structure
(Berkeley, 2002), edited by P. Fallon and R. Clark (2003) American Institute
of Physics, p315.
[17] O. Sorlin, Nucl. Phys. A 685 (2001) 186c.
[18] M. Stanoiu, Thèse de Doctorat de l’Université
de Caen (2003).
[19] M. Stanoiu et al, submitted to Phys. Rev. C, private
communication.
[20] O. Sorlin, proceedings of the International Conference
on the Labyrinth in Nuclear Structure (Crete, 2003), to be published as
a special issue of the AIP, private communication.
[21] D. Guillemaud-Mueller, Eur. Phys. J. A 13 (2002)
63.
[22] M. Bellegiuc et al, Nucl. Phys. A 682 (2001) 136c.
[23] N.A. Orr et al, Nucl. Phys. A 491 (1989) 457.
[24] A. Reed et al, Phys. Rev. C 60 (1999) 024311.
[25] E. Sauvan, Thèse de Doctorat de l’Université
de Caen, LPCC T 00-01 (2001).
[26] http://www.ganil.fr/exogam/ (section ‘Définition
du Projet’, subsection ‘Performances’).
[27] B.A. Brown, http://www.nscl.msu.edu/~brown/decay/gamma.htm.
[28] B.A. Brown, Prog. Part. Nucl. Phys. 47 (2001) 47
and references therein.
![[----]](greenline.gif){kind=small}
