Transport Model Simulations of Projectile Fragmentation Reactions at 140 MeV/nucleon

Transport Model Simulations of Projectile Fragmentation
Reactions at 140 MeV/nucleon

M. Mocko, 1 ' 2 M. B. Tsang, 2 D. Lacroix, 2 ' 3 A. Ono, 4
P. Danielewicz, 2 W. G. Lynch, 2 and R. J. Charity 5

1 Los Alamos National Laboratory, P. 0. Box 1663, Los Alamos, NM 87545, USA
2 National Superconducting Cyclotron Laboratory and Department of Physics
& Astronomy, Michigan State University, East Lansing, MI 48824, USA
3 Gaml, CEA, IN2P3-CNRS, B. P. 5027, F-14021 Caen Cedex, France
^Department of Physics, Tohoku University, Sendai 980-8578, Japan
5 Chemistry Department, Washington University, St. Louis, MO 63130, USA

Abstract

The collisions in four different reaction systems using 40 ' 48 Ca and 58,64 Ni isotope beams and a Be
target have been simulated using the Heavy Ion Phase Space Exploration and the Antisymmetrized
Molecular Dynamics models. The present study mainly focuses on the model predictions for the
excitation energies of the hot fragments and the cross sections of the final fragments produced in
these reactions. The effects of various factors influencing the final fragment cross sections, such
as the choice of the statistical decay code and its parameters have been explored. The predicted
fragment cross sections are compared to the projectile fragmentation cross sections measured with
the A1900 mass separator. At E/A = 140 MeV, reaction dynamics can significantly modify the
detection efficiencies for the fragments and make them different from the efficiencies applied to the
measured data reported in the previous work. The effects of efficiency corrections on the validation
of event generator codes are discussed in the context of the two models.

PACS numbers: 25.70.Mn

Keywords: projectile fragmentation, fragmentation reactions, fragmentation production cross section, level
density, event generator

1

I. INTRODUCTION

Projectile fragmentation is a well-established technique to produce beams of exotic nu-
clides used for various studies in fundamental nuclear physics. It is deployed in many facilities
around the world . Even though the first pioneering experiments were done in the

late 1970s at Berkeley Q, fl, and the fragmentation process is a fundamental decay mode
of highly excited nuclear systems [6( , the fragmentation reaction mechanism is not yet fully
understood. While there are many puzzling aspects to this phenomenon, it does display
some simplifying characteristics at high incident energies. For example, many experimental
observables in peripheral collisions at high energies (> 200 MeV/nucleon), such as mass,
charge, and multiplicity distributions, vary little with energy and target material. This,
so-called, limit ingfragment at ion behavior forms the basis of empirical parameterization of
the EPAX code [7]. It allows one to predict the mass and charge distributions of projec-
tile fragmentation reactions. Since it lacks the physics details of the reaction mechanism,
the predicted cross sections deviate rather significantly from the experimental data for very

10

111 ]. Understanding the physics of the

neutron-rich and very proton-rich nuclei (g, |9j,
projectile fragmentation is not only important for rare-isotope beam production purposes,
but also for understanding of fundamental nuclear physics processes involved in nuclear
collisions.

A campaign of four projectile fragmentation experiments was carried out at the National
Superconducting Cyclotron Laboratory at Michigan State University during 2002-2005 with
a goal to measure high-quality and comprehensive projectile fragmentation cross-section data
at intermediate energy. The data from eight different reaction systems yielded more than
1400 (1379 fragments +111 pick-up) measured cross sections. 140- MeV/nucleon lQf primary
beams of 40 Ca, 48 Ca, 58 Ni, and 64 Ni with 9 Be and 181 Ta targets were used in the present
experimental studies. The accuracy of these measurements provides benchmark quality sets
of data for testing reaction models [3, 13] as well as particle transport simulation codes

15

171]. In an

that are used in the design of accelerators and radiation shielding [14 .
effort to understand the underlying physical processes in projectile fragmentation reactions,
we have performed calculations using the macroscopic-microscopic Heavy Ion Phase Space
Exploration (HIPSE) model [3] and the sophisticated fully microscopic Antisymmetrized
Molecular Dynamics (AMD) model [3].

2

This paper is structured as follows. First we introduce the main features of the two
reaction models used in the present study in Section [III Then the properties of hot fragments
and comparisons of data and the calculated final fragments after decay are presented in
Section HTT1 The effect of sequential decays and the model dependence of the decay codes are
discussed in the same section. In addition, we discuss the effects of the detector efficiency
corrections when results from dynamical models are compared to data. Our results are
summarized in Section HVl

II. REACTION MODELS

Understanding of the reaction dynamics of fragment production in projectile fragmenta-
tion requires reaction models more sophisticated than phenomenological ones such as EPAX
or the widely used Abrasion- Ablation (AA) model 20, 21]. In the first step of the
AA model, collisions of spherical projectile and target nuclei are assumed. Nucleons in the
overlap region are "abraded" and their number depends on the impact parameter. In the
second step, the excited primary fragments decay. The model does not provide a mecha-
nism to calculate the excitation energy. For simplicity, the excitation energy is assumed to
be proportional to the number of abraded nucleons and sometimes adjusted to reproduce
the experimental data. The model's predictive power is also limited to cross sections and
cannot describe the velocity or momentum of the produced fragments. To understand the
dynamics of the reactions, we employ the Heavy Ion Phase Space Exploration and Antisym-
metrized Molecular Dynamics models, both of which describe the dynamical evolution of
the fragments during the collision. In the present work, there is no attempt to fit the data
by varying the model parameters, the nominal recommended values of the input parameters
are used in both calculations.

A. Heavy Ion Phase Space Exploration model

The Heavy Ion Phase Space Exploration (HIPSE) model has been implemented to bridge
the gap between the statistical models, which reduce the description of the reaction to a
few important parameters, and fully microscopic models 18|, |22j . Based on a macroscopic-

microscopic "phenomenology," it accounts for both dynamical and statistical aspects of

3

nuclear collisions. The HIPSE model has been shown to describe central and semi-peripheral
collisions well. On the other hand, very peripheral reaction mechanisms such as knock-out,
break-up, or pick-up reactions, which require the inclusion of the intrinsic quantum nature
of nucleons, cannot be accounted for.

Nuclear reaction, as described by the HIPSE model [18j], can be separated into three
stages: approach of the projectile and the target nuclei, partition (formation of fragments),
and the cluster propagation phase (with an in-flight statistical decay). Classical two-body
dynamics of the center of masses of the target and the projectile nuclei is assumed in the
entrance channel. The macroscopic proximity potential, giving a realistic Coulomb barrier, is
used to describe the nucleus-nucleus potential at large distances. At small relative distances,
the nucleus-nucleus potential should become sharper when the beam energy increases. To
account for this effect, a phenomeno logical parameter, denoted by a a , has been introduced
which extrapolates from the adiabatic limit (a a < 0) to the sudden approximation (a a = 1).
At the minimal distance of approach, nucleons in each nucleus are sampled according to
a realistic zero temperature Thomas-Fermi distribution. The participant and spectator
regions are then obtained using simple geometrical considerations. Nucleons outside the
overlap region define the Quasi-Projectile and Quasi- Target spectators. Then two physical
effects, namely direct nucleon-nucleon collisions and nucleon exchange, are treated in a
simple way. When the beam energy increases, the effect of direct nucleon-nucleon collisions
becomes increasingly important. This effect is modeled by assuming that a fraction, x co u,
of the nucleons in the overlap region undergoes in-medium collisions. The main effect of
the in-medium collision is to slightly distort the Fermi motion hypothesis in the sampling.
Once the direct collisions are over, a fraction, x ex , of the nucleons in the overlap region is
exchanged between the two spectator nuclei, relaxing the pure participant-spectator picture.

After these preliminary steps, clusters are formed using a coalescence algorithm [18( and are
propagated according to a classical Hamiltonian using the same nucleus-nucleus potential
as in the approach phase. To incorporate the physics of low energy reaction (below the
Fermi energy), after a time denoted by tf roz , two fragments with relative separation less
than their fusion barrier distance fuse if their relative energy is below their Coulomb barrier.
This feature leads, in general, to a large Final State Interaction (FSI). Once all the FSIs
are processed, the nuclei cannot exchange particles anymore and a chemical freeze-out is
reached. At this stage, the total excitation energy can be determined event-by-event from

4

the energy conservation and assigned to clusters, which finally undergo in-flight decay.

The HIPSE model has only three adjustable parameters (a a , x ex , x co u). The values of
these parameters have been adjusted 22j for beam energies of 10, 25, 50, 80 MeV/nucleon.
Jsing simple functions, we extrapolated the values of a a = 0.55, x ex = 0.09, and x co u = 0.18
9j to our beam energy of 140 MeV/nucleon. In order to compare the HIPSE simulation
with results from the Antisymmetrized Molecular Dynamics (AMD) model (see below), the
time, tf roz , originally taken as 50 fm/c has been increased to 150 fm/c. We have checked
that this does not affect the final results.

The HIPSE model originally includes the in-flight decay based on an improved version
of the SIMON decay model 23(. In the present study, the phase space gene rated by HIPSE

24J which is known

before the in-flight decay is used as input to the GEMINI decay code
to give a better treatment of sequential decays of excited nuclei. As a consequence, some
spatial-temporal correlations may be lost. The influence of the decay code will be discussed
in more detail below.

B. Antisymmetrized Molecular Dynamics model

The Antisymmetrized Molecular Dynamics (AMD) model [19j, |25j has been chosen from
among many microscopic models to simulate the fragmentation reactions measured in our
experiments. As one of the most sophisticated transport models, it describes the nuclear
reaction at the microscopic level of interactions of individual nucleons. In the AMD model, a
potential is used to take into account all reaction processes involved in the complex heavy-ion
collisions.

The AMD wave function is given by a Slater determinant of Gaussian wave packets for
individual nucleons. Centroids of these wave packets are treated as dynamical variables.
An effective nuclear interaction determines the one-body motion of the wave packets by
the mean field. The correlations are introduced by expressing the many-body state as an
ensemble of many AMD wave functions, i.e., by adding stochastic terms to the equation of
motion. Nucleon-nucleon scattering is included as a stochastic process. The probabilities
of collisions are determined by the assumed in-medium cross sections of nucleon-nucleon
collisions. Furthermore, another stochastic term is considered in order to take into account
the change of the width and shape of the phase-space distributions of individual nucleons.

5

The single-particle wave functions in each channel are Gaussian wave packets with a fixed
width parameter, which is advantageous in describing the fragment formation. It should be
noted that the time evolution is solved independently for each channel, and the interference
of different channels is neglected.

In the present study, we employ a Gogny-type force (Gogny-AS (2g|) as the effective
nucleon-nucleon interaction and the free two-nucleon collision cross sections with a cut-off
at 150 mb as the in-medium nucleon-nucleon cross section.

The ground states of 40 ' 48 Ca and 58,64 Ni projectiles and 9 Be target were prepared by the

frictional cooling method [25J applied to the AMD wave function. The AMD simulations
were carried out for an impact parameter range of 0-10 fm and up to the time of 150 fm/c,
when the primary fragments are spatially well separated. These fragments, recognized by

a simple coalescence algorithm, with the associated excitation energy
the GEMINI code [24].

9| are decayed using

III. RESULTS

A. Primary fragments

The most direct comparison of different model calculations is to examine the properties of
the primary fragments before sequential decays occur. Fig. [1] presents the isotopic primary
fragment distributions obtained by the HIPSE (solid line) and AMD (dashed lines) models
for the 64 Ni+ 9 Be reaction system. In general, the total fragment cross sections from AMD
are higher resulting in higher isotope cross sections around the peak compared to HIPSE
results. For elements close to the projectile (Z > 25), both models predict very similar
isotope distributions with relatively narrow widths (bottom panels). The models start to
show increasing differences in the widths, the centroids, and the magnitudes of the cross
sections for the isotope distributions with increasing number of removed protons. With
more removed nucleons (upper panels), the centroids of the isotopic distributions from AMD
are shifted to less neutron-rich isotopes. As expected, the shifts are more pronounced for
neutron-rich projectiles of 48 Ca and 64 Ni In the AMD model, the centroids and widths
of the isotope distributions are expected to depend on the symmetry energy terms of the

effective interaction, as is the case for central collisions

27

28). In the HIPSE model,

6

experimental masses and empirical formula are involved in the computation of Q values and
excitation energy but no explicit density dependence of the symmetry energy is included.

B. Excitation energy

After nuclear collisions, the excited projectile-residue decays through emission of light
particles. The evolution of the decaying system depends on the excitation energy trans-
ferred. However, in projectile fragmentation experiments we detected fragments in forward
angles (±30 mrad) with velocities close to that of the projectile [9j]. The mean excitation
energy can be deduced only indirectly from theoretical models. In the simple Abrasion-

Ablation (AA) model 20], the excitation energy of primary fragments is assumed to be

proportional to the number of nucleons removed [9j, [21|, [29j . To best describe fragmentation
cross sections, both the excitation energy and its fluctuations are determined by fitting the
data [9]. Contrary to the AA models, the models considered here calculate the excitation
energy and its fluctuations. The HIPSE model defines the excitation energy of the residues
from energy conservation, and the AMD calculation defines the excitation energy by eval-
uating the expectation value of the Hamiltonian for the many-body wave function of the
primary fragment.

The average excitation energy per nucleon, E*/A, for primary fragments produced in the
fragmentation of 40,48 Ca and 58 > 64 Ni on 9 Be target is shown as open squares in Fig. [2]and[3]for
the AMD and HIPSE calculations, respectively, as a function of the primary fragment mass
number. The shaded regions show the root-mean-square (RMS) widths of the excitation
energy distributions. For reference, we calculated the residue excitation energy with a single
particle, microscopic Boltzman-Uehling-Uhlenbeck (BUU) equation [30]. The BUU results
are deterministic and give average values of the observable for a given impact parameter.
They are shown as solid lines in Fig. [2] and [3j The BUU results which do not extend to
small residue masses, exhibit trends more similar to the HIPSE calculations.

In the case of the AMD simulations, we notice a rather sharp rise of E*/A with the number
of removed nucleons close to the projectile. The excitation energy saturates around 4 MeV
after removal of about 10 nucleons. The saturation is inconsistent with the assumptions
used in AA models, which assume that excitation energy is proportional to the number of
abraded nucleons. In the HIPSE model, similar saturation values are obtained especially in

7

the case of the Ni isotopes. However, the excitation energy fluctuations are much larger in
the HIPSE model than the AMD model.

For residues close to the projectile (~ 0-10 abraded nucleons), the AMD calculation
produces systematically higher excitation energy. This could be due to different cluster
formation in the models. The simple nucleon sampling procedure used in HIPSE to con-
struct fragments may predict lower excitation for projectile-like fragments. As discussed
in Section IHI Dl the excitation energy profile may be related to the widths of the isotope
distributions.

Saturation of the excitation energy is not obvious in the fragmentation of the Ca isotopes
in the HIPSE calculations. The residue excitation energy increases with decreasing masses
but the rate of increase is much less for residues lighter than 30 for 40 Ca and 38 for 48 Ca
projectiles. Monotonic increase is observed in the reactions with 181 Ta targets in the HIPSE
calculations. The latter increase may be related to the increase of the maximum number of
nucleons involved in the collisions and exchange with the quasi-projectile when the targets
are changed from 9 Be to 181 Ta. However, different profiles of the mean excitation energy for
;he 9 Be and 181 Ta are not supported by the data, which show very little target dependence
9|. Unfortunately, we could not perform the AMD calculations involving 181 Ta targets to
check the observations seen in the HIPSE model calculations because of excessive CPU time
required.

C. Evaporation Codes

The primary fragments produced by different reaction models cannot be directly com-
pared to the experimental data, because the experimental observation of fragments is per-
formed after hundreds of nanoseconds, many orders of magnitude later, than the prompt
step simulated by the nuclear reaction models (HIPSE, AMD). Direct comparison with data
requires incorporating the sequential decay models as the second step after excited fragments
are formed.

n

Currently, there are no standardized sequential decay codes |3JJ. The GEMINI code,
widely used in performing sequential decay of hot fragments, calculates the decay of a
primary fragment by sequential binary decays. Monte Carlo technique is employed to follow
all decay chains until the resulting products are unable to undergo further decay. For the

8

purposes of the sequential decay calculations the excited primary fragments gen erated by

the HIPSE and AMD model calculations are taken as the compound nucleus [32

the GEMINI code. Hence, every primary fragment is decayed as a separate event 9J. For the

input to

evaporation of particles lighter than an alpha-particle, the Hauser-Feshbach |33j formalism
is applied. The liquid drop model with shell corrections 3J1 is used to calculate the masses

n

of all parent and daughter nuclei in the calculation. The Fermi gas [35( expression is used
to calculate the level density.

For neutron-rich projectiles, the final isotope distributions are found to be sensitive to the
selection of the level density parameter, a. To demonstrate this effect, results from the hot
fragments produced in AMD simulations of the collisions of the 48 Ca beam on a 9 Be target
using two different level density parameterizations are shown in Fig. HI The solid lines are
calculated using a = A/12 MeV" 1 and the dashed lines depict the calculation using a = A/8
MeV -1 . The calculation using a = A/ 12 MeV -1 results in wider isotope distributions and
shifting of the peaks towards the more neutron-rich isotopes. Such shifts are also observed in

h

the case of the neutron-rich 64 Ni beam [9J . The change of the level density parameter while
keeping all other parameters of the sequential decay constant corresponds to an effective
change of temperature of the decaying system. In this picture, decaying compound nucleus
with higher temperature (a = A/12 MeV -1 ) leads to wider isotope distributions with more
neutron-rich fragments. On the other hand, the decay of the less neutron-rich fragments
such as those produced in the 40 Ca and 58 Ni induced reactions will not be affected very
much. Unless otherwise noted, all sequential decay calculations in this paper use a = A/ 10
MeV- 1 .

In addition to the level density parameter, the final fragment distributions also depend

on the evaporation code used

311 ] . For example, if the SIMON decay code is used instead of

GEMINI, the fragment distributions are different. The dashed lines in Fig. [5] are predictions
from HIPSE coupled with the decay code SIMON while the solid lines are predictions from
HIPSE using GEMINI as the decay code for the 48 Ca+ 9 Be reaction. In general, results
from GEMINI reproduce the cross sections of the neutron deficient fragments consistently
better than calculations using SIMON to decay the hot fragments. Detailed comparisons of
different decay codes have been published in Ref. [3jJ. In particular, SIMON code failed to
exhibit the isoscaling behavior which is reproduced by other statistical codes. Nonetheless,
the uncertainties introduced by different decay codes or using different parameters in the

9

decay codes can be as large as the differences of the results between the AMD and HIPSE
models.

D. Fragment cross sections

In this section we compare the experimentally determined reaction cross sections with
the final fragment cross sections predicted by the HIPSE and AMD models. To minimize
the influence of different sequential decay codes, GEMINI is used as the evaporation code
to decay the excited primary fragments created in both models.

Experimentally, only fragments with velocities that match the acceptance of the fragment
separator are measured. Ideally, such experimental constraints have to be compensated so
that the experimental cross-section data can be compared directly with predicted cross
sections from theoretical models. This is especially true for models such as the EPAX pa-
rameterization of the fragment cross sections and the abrasion-ablation models, which do
not contain dynamic information about the collisions. However, the detection efficiency co-
efficients used in converting the measured (raw) cross sections to total cross sections depend
on the assumptions of the transmission efficiency and angular distributions of the fragments.
In general, fragment transmission through the magnetic spectrometer used is obtained from
ion-optics simulations and the transmission efficiency is better than 95%. On the other hand,
it is very difficult to estimate the fragment angular distributions with certainty. In the ex-
periment, only the momentum distributions accepted within the spectrograph are measured.

3e measured easily and their distributions are

a.

Transverse momentum distributions cannot
normally estimated using parameterizations

The open squares in Fig. [6] show the mass dependence of the transmission correction
factor, e, used in Ref. {9], [juj] to obtain the published fragment cross sections for the four
reactions studied here. From dynamical models, one can calculate the correction factors by
constructing the ratios of fragment cross sections filtered by the experimental acceptance,
^filtered, to the calculated fragment cross sections, a m odei'-

£th = ° filtered/ " model' (1)

The corresponding correction factors, Eth, obtained from the models, a solid line for HIPSE
and a dashed line for AMD, are quite different from e (open squares), used to correct the

10

experimental data, suggesting the angular distributions assumed in e do not agree with the
angular distributions described by the models. The differences are model dependent and
largest for lighter fragments. Fig. E] illustrates that it may not be appropriate to com par e
calculated results directly to the published data as is customarily done Q, fl HQ-
As the correction factors are model dependent, it is more accurate to compare filtered
calculations to uncorrected (raw) experimental cross sections. To illustrate the differences
in comparing corrected data with unfiltered theoretical results (Fig. [7J and raw data with
filtered calculations (Fig. [8]), we plot the mass distributions for the four systems studied
here. The data are shown as open symbols and lines are predictions from various models.
Since the detection efficiency decreases with the mass of the detected fragments, lighter
masses that are less than half of the projectile masses have the largest corrections. This
is expected as the lighter masses have a larger velocity or momentum spread in nearly all
models. For reference, the dotted lines in Fig. [7] are EPAX predictions. Agreement with the
EPAX results is better for the 48 Ca+ 9 Be and 58 Ni+ 9 Be reactions as the mass distributions
of both of these reactions were used to extract the EPAX parameters jzj].

In Fig. [HI the filtered results, obtained by applying the experimental acceptance cut
of 30 mrad to the simulated events from both the HIPSE and AMD models agree with
the raw data much better, especially for light fragments. For the HIPSE model, the drop
of fragment yields around the projectiles is due to inadequate fluctuations in the most
peripheral collisions. Except for 40 Ca induced reactions, the AMD model predicts higher
fragment yields for the other three reactions and the HIPSE model reproduces the overall
magnitude of the cross sections.

Fig. l9TfT2l present comparisons of the measured isotopic cross sections without experimen-
tal efficiency corrections from the fragmentation of the 40 Ca, 48 Ca, 58 Ni, and 64 Ni primary
beams on 9 Be targets. They are presented in terms of isotopic distributions as a function of
neutron excess, N — Z. Individual panels in Fig. l9TfT2lshow isotopic distributions for differ-
ent elements labeled with their respective chemical symbols. (Note that these experimental
cross sections are different from those published in Ref. 0, [l0]. The previously published
cross sections were corrected for angular and momentum transmission inefficiency based on
parameterization without the knowledge of the collision dynamics.)

The HIPSE and AMD models are stochastic calculations. The lower limit of calculated
cross sections depends on the number of simulated events, which were approximately 100,000

11

and 20,000 for each reaction for the HIPSE and AMD models, respectively. Overall the
peaks of the isotope distributions are described well by both calculations for fragmentation
of 40,48 Ca and 58,64 Ni beams. Except for the 40 Ca+ 9 Be reaction, the AMD-predicted cross
sections for the other three reactions are consistently higher than the data and the HIPSE
model predictions. There are many input parameters and model details in the AMD model,
which affect cross sections. For example, the cross sections may be sensitive to the transport-
model input parameters such as the in-medium nucleon-nucleon cross sections. The nuclear
structure and density profile of the projectile and target nuclei may also affect the fragment
cross sections. More calculations with the AMD model will be needed to understand this
discrepancy further.

In the case of neutron-deficient projectiles such as 40 Ca the HIPSE model predicts larger
cross sections of the fragments with mass close to the projectile which in turn leads to
over-estimations of the neutron-deficient isotopes.

The isotope cross-section distributions calculated by the HIPSE model are generally wider
than the experimental data for all investigated projectiles. Fig. [13] shows the RMS widths
of the isotope distributions obtained from the data (open squares), HIPSE (solid line),
and AMD (dashed line). The experimental widths are much better described by the AMD
model. The wider isotopic widths predicted by the HIPSE model may be related to the larger
fluctuations of the excitation energy of the primary fragments. Comparison of Fig. [2|3l and[T3l
suggests that the isotopic widths are correlated to the mean excitation energy of the primary
projectile-like residues produced in the models. The discrepancies in the isotopic widths
are the largest for the 40 Ca+ 9 Be reactions and least for the 48 Ca+ 9 Be collisions. Similar
discrepancies are observed in the mean excitation energy of the projectile-like particles.

IV. SUMMARY

We carried out an extensive study of the projectile fragmentation reactions using the
macroscopic-microscopic Heavy Ion Phase Space Model and the fully microscopic Antisym-
metrized Molecular Dynamics Model. Even though these models were not developed to
describe the projectile fragmentation process, the agreement between predictions and data
is reasonable, especially when one considers that there is no effort to vary model parameters
to fit the data and that these models were not developed to describe fragmentation reaction

12

mechanisms.

These models go beyond the phenomenological models such as the EPAX or the Abrasion-
Ablation model in describing the dynamics of the reactions as well as in predicting the
excitation energy of the primary fragments. Both models give similar dependence of the
excitation energy profiles as a function of removed nucleons from the projectile in reactions
with Be targets. The saturation of the excitation energy is contradictory to the assumptions
used in the AA models. The HIPSE model is able to reproduce the overall magnitude of the
mass distributions, but the AMD model predictions are more consistent with the shape of
the measured isotope distributions. The calculated final cross-section distributions are the
results not only of the primary (fast) step of the nuclear reaction (modeled by HIPSE or
AMD), but also of the secondary (slow) step, modeled by the statistical evaporation code
(GEMINI). At 140 MeV/nucleon, the final fragment distributions are influenced significantly
by the sequential decays. Hence, it is imperative to better understand the de-excitation part
of the nuclear collision, the evaporation process, if we want to put the dynamical nuclear-
collision calculations to a more stringent test.

Due to the Monte-Carlo nature of the transport models, it is impractical to use these
models to estimate the yield of the rare isotopes. However, all current models, including
EPAX parameterization, cannot predict sufficiently accurately the yields of rare isotopes
with extremely low cross sections. Dominance of the sequential decay processes suggests that

parameterization based on statistica

of the yields of these rare nuclei

13

model considerations gives more accurate predictions

36J.

Finally, models such as AMD include information about transport mechanisms and their
parameters. Of particular interest are asymmetry term of the nuclear equation of state, the
in-medium nucleon-nucleon collisions, and cluster formation. Our analysis suggests that in
addition to the cross section, other measured experimental quantities such as the momentum
distributions may provide important constraints to these transport quantities.

Acknowledgments

A. Ono and D. Lacroix thank the National Superconducting Cyclotron Laboratory at
Michigan State University for the support and hospitality during their sabbatical stays in
2005-2006 (A.O.) and 2006-2007 (D.L.). We would like to acknowledge the High Perfor-

13

mance Computing Center [37] at Michigan State University for facilitating CPU-intensive
calculations with the AMD code. This work is supported by the National Science Foundation
under Grant Nos. PHY-01-10253, PHY-0606007 and DE-FG02-04ER41313.

[1] H. Geissel, P. Armbruster, K. H. Behr, A. Briinle, K. Burkard, M. Chen, H. Folger,
B. Franczak, H. Keller, O. Klepper, B. Langenbeck, F. Nickel, E. Pfeng, M. Pfiitzner,
E. Roeckl, K. Rykaczewski, I. Schall, D. Schardt, C. Scheidenberger, K.-H. Schmidt,
A. Schroter, T. Schwab, K. Siimmerer, M. Weber, G. Miinzenberg, T. Brohm, H.-G. Clerc,
M. Fauerbach, J. J. Gaimard, A. Grewe, E. Hanelt, B. Knodler, M. Steiner, B. Voss, J. Weck-
enmann, C. Ziegler, A. Magel, H. Wollnik, J. P. Dufour, Y. Fujita, D. J. Vieira, and B. Sherrill.
Nucl. Instrum. Methods Phys. Res., Sect. B, 70:286, 1992.

[2] T. Kubo, M. Ishihara, N. Inabe, H. Kumagai, I. Tanihata, and K. Yoshida. Nucl. Instrum.
Methods Phys. Res., Sect. B, 70:309-319, 1992.

[3] B. M. Sherrill. Prog. Theor. Phys., 146:60-69, 2002.

[4] H. H. Heckman, D. E. Greiner, P. J. Lindstrom, and F. S. Bieser. Science, 174:1130-1131,
1971.

[5] H. A. Grander, W. D. Hartsough, and E. J. Lofgren. Science, 174:1128-1129, 1971.
[6] W. G. Lynch. Ann. Rev. Nucl. Part. Sci., 37:493-535, 1987.
[7] K. Siimmerer and B. Blank. Phys. Rev. C, 61:034607, 2000.

[8] M. Notani. Projectile Fragmentation Reactions and Production of Nuclei near the Neutron

Drip-line. PhD thesis, University of Tokyo, 2000.
[9] M. Mocko. Rare Isotope Production. PhD thesis, Michigan State University, 2006.

[10] M. Mocko, M. B. Tsang, L. Andronenko, M. Andronenko, F. Delaunay, M. A. Famiano, T. Gin-
ter, V. Henzl, D. Henzlova, H. Hua, S. Lukyanov, W. G. Lynch, A. M. Rogers, M. Steiner,
A. Stolz, O. Tarasov, M. J. van Goethem, G. Verde, M. S. Wallace, and A. Zalessov. Phys.
Rev. C, 74:054612, 2006.

[11] M. Notani, H. Sakurai, N. Aoi, H. Iwasaki, N. Fukuda, Z. Liu, K. Yoneda, H. Ogawa, T. Teran-
ishi, T. Nakamura, H. Okuno, A. Yoshida, Y. Watanabe, S. Momota, N. Inabe, T. Kubo, S. Ito,
A. Ozawa, T. Suzuki, I. Tanihata, and M. Ishihara. Phys. Rev. C, 76:044605, 2007.

[12] G. Chaudhuri, S. Das Gupta, W. G. Lynch, M. Mocko, and M. B. Tsang. Phys. Rev. C,

14

76:067601, 2007.

[13] M. B. Tsang, W. G. Lynch, W. A. Friedman, M. Mocko, Z. Y. Sun, N. Aoi, J. M. Cook,
F. Delaunay, M. A. Famiano, H. Hui, N. Imai, H. Iwasaki, T. Motobayashi, M. Niikura,
T. Onishi, A. M. Rogers, H. Sakurai, H. Suzuki, E. Takeshita, S. Takeuchi, and M. S. Wallace.
Phys. Rev. C, 76:041302(R), 2007.

[14
[15
[16

[17;

[18
[19
[20

[21
[22
[23
[24

[25
[26

[27;

[28

[29
[30
[31

H. Iwase, K. Niita, and T. Nakamura. J. Nucl. Sci. Tech., 39:1142-1151, 2002.
|nttp : //www-ap . f nal . gov7M ARS.
http : //mcnpx . lanl . gov,

R. M. Ronningen. In Hadronic Shower Simulation Workshop, volume 896, pages 71-80. AIP
Conference Proceedings, 2007.

D. Lacroix, A. Van Lauwe, and D. Durand. Phys. Rev. C, 69:054604, 2004.
A. Ono and H. Horiuchi. Prog. Part. Nucl. Phys., 53:501-581, 2004.

J. D. Bowman, W. J. Swiatecki, and C. F. Tsang. Abrasion and Ablation of heavy ions. LBL
Report, LBL-2908, 1973.

J.-J. Gaimard and K.-H. Schmidt. Nucl. Phys. A, 531:709-745, 1991.

|http: //caeinf o . in2p3 . f r/theorie/hipse .htm

D. Durand. Nucl. Phys. A, 541:266-294, 1992.

R. J. Charity, M. A. McMahan, G. J. Wozniak, R. J. McDonald, L. G. Moretto, D. G. Srantites,
L. G. Sobotka, G. Guarino, A. Pantaleo, L. Fiore, A. Gobbi, and K. D. Hildenbrand. Nucl.
Phys. A, 483:371-405, 1988.

A. Ono, H. Horiuchi, T. Maruyama, and A. Ohnishi. Phys. Rev. Lett, 68:2898-2900, 1992.
A. Ono, P. Danielewicz, W. A. Friedman, W. G. Lynch, and M. B. Tsang. Phys. Rev. C,
68:051601, 2003.

A. Ono, P. Danielewicz, W. A. Friedman, W. G. Lynch, and M. B. Tsang. Phys. Rev. C,
70:041604, 2004.

A. Ono. In Sixth Latin American Symposium on Nuclear Physics and Applications, vol-
ume 884, page 292. AIP Conference Proceedings, 2007.

W. A. Friedman, M. B. Tsang, D. Bazin, and W. G. Lynch. Phys. Rev. C, 62:064609, 2000.
P. Danielewicz. Nucl. Phys. A, 673:375, 2000.

M. B. Tsang, R. Bougault, R. J. Charity, D. Durand, W. A. Friedman, F. Gulminelli, A. Le
Fevre, Al. H. Raduta, Ad. R. Raduta, S. Souza, W. Trautmann, and R. Wada. Eur. Phys. J.

15

A, 30:129-139, 2006.
[32] N. Bohr. Nature, 137:344, 1936.

[33] W. Hauser and H. Feshbach. Physical Review, 87:366-373, 1952.
[34] H. J. Krappe, J. R. Nix, and A. J. Sierk. Phys. Rev. C, 20:992, 1979.
[35] H. A. Bethe. Physical Review, 50:332-341, 1936.

[36] M. Mocko, M. B. Tsang, Z. Y. Sun, L. Andronenko, M. Andronenko, F. Delaunay, M. A.

Famiano, W. A. Friedman, V. Henzl, D. Henzlova, H. Hui, X. D. Liu, S. Lukyanov, W. G.

Lynch, A. M. Rogers, and M. S. Wallace. Europhys. Lett, 79:12001, 2007.
[37] http://hpc.msu.edu

16

10

1

■^io- 2

$ 10 3
co 1Q 2
o
O

10" 1

& N

i i i | i i i i | i i i i | i i i

// \ -

; sc ;

Ti

v

1 1 1 1 ■ 1 ■ 1 1 1 ■ 1 ■ i ■ 1 ■

; 64 Ni+ 9 Be :

1 1 1 1 1 1 11 1 1 1 1 1 1 11 1
<^*~~\

s \\

/ V

1 1 1 I 1 1 1 1 I 1 1 1 1 I 1 1 1

1 1 1 1 1 1 1 1 1 1 '1 1 1 1 1 1

... i .... i .... i .. .

Fe

, , . i . , , , i , . , . i , , ,

. , , i , , , , i , , . , i , , ,

Ni \ "

, . , i , , , , i , . , , i , . ,

5 10

5 10

5 10

5 10

Neutron excess N-Z

FIG. 1: (Color online) Primary fragment isotopic distributions for the 64 Ni+ 9 Be reaction system
plotted as a function of neutron excess, N — Z. Solid and dashed lines show calculations by HIPSE
and AMD models, respectively.

17

Residue mass A Residue mass A

FIG. 2: (Color online) The mean excitation energy per nucleon, E*/A, plotted as a function of the
mass number of the primary fragments for reactions of 40 Ca (top left), 48 Ca (bottom left), 58 Ni
(top right), and 64 Ni (bottom right) with 9 Be target. The mean excitation energy per nucleon
calculated by the AMD model is shown as open squares connected by the dashed line to guide
the eye. The shaded region depicts fluctuations in the excitation energy per nucleon expressed in
terms of one standard deviation around the mean. For reference, the BUU calculation results are
plotted as a solid line.

18

Residue mass A Residue mass A

FIG. 3: (Color online) The mean excitation energy per nucleon, E*/A, plotted as a function of the
mass number of the primary fragments. Same convention as Fig. [2] is used.

19

10 2

lo 1

E

g 10" 2

S 10 2
w

CO

o 1

O

10" ;

ua+ be

7 n\

y \

! Al ]

1 1 1 i 1 1 1 1 i 1 1 1 1 i 1 ■ 1 '

Si

p

S

' ' ' i ' ' ' ' i ' ' ' ' i ' ' '

' ' ' i ' ' ' ' i ' ' ' ' i ' ' '
/ \

i
i

' ' ' i ' ' ' ' i ij' ' i ' ' '
i

J

CI

Ar

K

Ca

5 10

5 10

5 10

5 10

Neutron excess N-Z

FIG. 4: Isotopic cross-section distributions for 48 Ca+ 9 Be for 13 < Z < 20 elements calculated
using the AMD model coupled to GEMINI decay calculations with level density parameter a = A/8
MeV -1 (dashed line) are compared to calculations with a = A/12 MeV -1 (solid line).

5 10 5 10 5 10 5 10

Neutron excess N-Z

FIG. 5: Isotopic cross-section distributions of 48 Ca+ 9 Be reactions calculated by HIPSE with GEM-
INI decay (solid line) are compared to HIPSE coupled to SIMON decay (dashed line).

20

FIG. 6: (Color online) Transmission correction, used to correct the experimental data based on
parameterization [loj], is plotted as a function of the fragment mass number (open squares) for
all investigated reaction systems. The solid and dashed lines show the transmission correction
deduced from the HIPSE and AMD calculations, respectively.

21

10 2

^ 40 Ca+ 9 Be A rV D

D ,°.. ° Corrected Data _

" .... E pAX
| -HIPSE
- -AMD -

iii
58 Ni+ 9 Be

-V °n
V □

I ' ' ' ' I ' ' ' ' I ' ' ' ' I ' ' ' '
r 48 Ca + 9 Be

- g 9 " n D -

: .*•..-"••■' □□□ :

H 1 1 1 1 1 1 1 1 1 1

r 64 Ni+ 9 Be

riv □ □
□

I ... I ... I

l_l I I I I I I I I I I I I I I I I I I J I I I I I I l_

10 20 30 40 20 40

Mass number A

FIG. 7: (Color online) Mass cross-section distributions (open squares) for four reaction systems are
compared to predictions from two reaction models, HIPSE (solid lines) and AMD (dashed lines).

22

FIG. 8: (Color online) Uncorrected cross-section mass distributions (open squares) for four reaction
systems are compared to filtered events from two reaction models, HIPSE (solid lines) and AMD
(dashed lines).

23

10 2

AC\ Q

40 Ca+ 9 Be

r \ -

:/S :

/ V

a d*
' 1

: B :

— i — i — i — i — i — i — i — i — i —

C □ :

N n

1 B

B ! .

10 2

jd
E

I'l'i'i — i— >—
f

''1' — '''I'

: A* :

c10" 1
o

o

□

F n "

'a □

Ne

□

□

Na n

-a

Ma □ ■

ivi y 1-1

wio 2

CO
CO

o

L.

O

icr 1

— , — , — 1 — , — , — , — , — | — , —

:A :

□

- □

' ' i ' ' ' ' i '

r i

_ □

' ' i ' ' ' ' i '

' ' 1 ' ' ' ' 1 '

J- B B

-,• -j -

a

□

Al

Si D :

p

s A -

10 2
10

— , — , — i — , — , — , — , — i — ^~

' ' i ' ' ' ' i '

' A

r

A

. □

10" 2

CI

A

Ar

, , i , . , , i ,

K

, , i , , , , i ,

Ca

, , i , , , , i ,

05050505

Neutron excess N-Z

FIG. 9: (Color online) Fragmentation (open squares) and nucleon pick-up (open triangles) cross
sections of 40 Ca+ 9 Be reactions are compared to calculations by HIPSE (solid line) and AMD
(dashed line) models. The error bars in the data are smaller than the symbols.

24

10

1

10" 2
10

-O

.91 0" 2
o

CD
CO

8 10

o

6 1

oa+ be

: A

: B D

□

C □

A :

□

N

?

A ^

o □

1 1 i 1 1 1 1 i 1 1 1 1 i ■

□

F

i i 1 i i i i 1 i T 1 i i 1 '

1 1 1 1 1 1 1 1 1 1 1 1 1 1

A

□

Ne

i i i i i i i i i i i i i i

A

□

Na

1 1 1 1 1 1 1 1 1 1 1 1 1 1

A :

- □ □

□

Mg D

TV

- Al D :

?! .

- □ □ -
□

Si □ -

^ 1 i 1 1 1 1 i 1 1 1 1 1 1

A:

Q □
" P

' ' i ' ' ' ' i ' ' ' ' i '

f\ ■

- a -

o □
S

□

- D CI

□ Ar

, , i , , , , i , , , , i ,

' ' 1 ' ' ' ' 1 ' ' ' ' 1 '

a A

□ K *■

, V i , , , , i , , , , i ,

■ ■ i ■ ' ' ■ i ■ r ■ i ^

A

□

Ca

, □ i , , , , i , , , , i ,

10"'

10" 1

5 10

5 10 5

Neutron excess N-Z

10 5 10

FIG. 10: (Color online) Fragmentation (open squares) and nucleon pick-up (open triangles) cross
sections of 48 Ca+ 9 Be reactions are compared to calculations by HIPSE (solid line) and AMD
(dashed line) models. The error bars in the data are smaller than the symbols.

25

10

1

10" 2
10

1

10" 2
10

E
o

O 10
CD

co 1
w

CO

o

O10" 2

10
1

10" 2
10 2

10" 1

58 Ni+ 9 Be

B

Ne

□

i ' ' ' ' i

Na g

□ □

A

□

Al

□

□ V

n Fe

□

Co

5 5

Neutron excess N-Z

FIG. 11: (Color online) Fragmentation (open squares) and nucleon pick-up (open triangles) cross
sections of 58 Ni+ 9 Be reactions are compared to calculations by HIPSE (solid line) and AMD
(dashed line) models. The error bars in the data are smaller than the symbols.

26

10

1

10" 2
10

1

10" 2
10

1

E

~io 2
o

o 10

CD

CO !

CO
CO

o

O10" 2

10
1

10" 2
10 2

10" 1

^Ni^Be

□

B

P n

i ' ' ' ' i

i ' ' ' ' i

A

ci □

□

Sc

V

i 1 1 1 1 1 1 1 11 1

□

□ □
Mn

Fe

'Co

5 10 5 10 5 10

Neutron excess N-Z

5 10

FIG. 12: (Color online) Fragmentation (open squares) and nucleon pick-up (open triangles) cross
sections of 64 Ni+ 9 Be reactions are compared to calculations by HIPSE (solid line) and AMD
(dashed line) models. The error bars in the data are smaller than the symbols.

27

<

en

DC

1.4
1.2
1

=u 0-8

c 0.6
o

"5 2.4

€ 2.2

o 1-8
g-1.6
O 1.4
^ 1.2
1

0.8

40 Ca+ 9 Be

Q-

H 1 1 1 1 1 1 1 1 1 1 h

58 Ni+ 9 Be

□□□□□□□ nx _

H — I — I — I — — I — I — I — I — — I — I — I — I — — I — I — I — I — — I — I — ^

48 Ca+ 9 Be

64 Ni+ 9 Be

□ n an D

□

□

10

15 10 15

Nuclear Charge Z

20 25

FIG. 13: (Color online) Widths of the isotope distributions shown in Fig. [9 HT21 expressed in terms
of standard deviation (RMS) are plotted as a function of nuclear charge, Z, for all investigated
reaction systems. Experimental data are shown with open squares, solid and dashed lines depict
the HIPSE and AMD simulations, respectively.

28