Abstract

1
Enter The DRAGON Investigating the 13C(p,?)14N
reaction Aaron M. Bebington a, 1 (and the DRAGON
Collaboration b) aUniversity Of Surrey,
Guildford, Surrey, England bTRI-University
Meson Facility (TRIUMF), Vancouver, BC, Canada
Abstract The 13N(p,?)14O reaction is very
important for our understanding of explosive
astrophysical sites, such as novae and
supernovae. This reaction determines the
conditions under which the CNO cycle changes to
the Hot CNO cycle. If temperatures are hot
enough, 13N will capture a proton before it has
chance to beta decay, forming 14O which initiates
the HCNO cycle. The beta decay of 14O (t1/2
70.6secs) is much quicker than the beta decay of
13N (t1/2 9.97mins), which means that the HCNO
cycle produces energy much faster. The DRAGON
collaboration at TRIUMF plans to measure the
cross-section of the 13N(p,?)14O reaction at
energies around the Gamow window, relevant to
novae temperatures. This region of energy is
lower than the resonance peak energy, which has
been measured previously. As 13N is radioactive,
and is very close in mass to 13C (a difference of
0.002383 amu), a pure 13N beam is difficult to
produce, because 13C will contaminate the beam.
Initially we studied the 13C(p,?)14N reaction so
that its contribution could be compensated for
when studying the 13N(p,?)14O reaction. The
13C(p,?)14N reaction was used to probe the DRAGON
not only because it has similar properties, but
because 13C(p,?)14N measurements have been made
before by King et al2.
Data Analysis from the 13C(p,?)14N
reaction Figure 1 shows a typical coincidence
gamma energy spectrum from a DRAGON run of the
13C(p,?)14N reaction. A coincidence gamma is one
that is associated with a recoil heavy ion of 14N
as detected in the end detector of DRAGON. The
c?0 means that the data put into this spectrum is
from the most energetic coincidence gamma ray
detected by a single BGO per event by the BGO
gamma array. The main peak will correspond to the
energy of the gamma rays cascading from the 8MeV
excited state to the ground state. The various
other peaks are from either the cascade gammas
to other excited states, or from the main 8MeV
gammas that did not deposit all of their energy
into a single BGO. The data analyzer used by
DRAGON is a MIDAS program which looks at runs
online and offline. Analyzing a run offline means
that we can pass the run through the analyzer
many times, and by having made changes to the
online data base (ODB), we can eliminate more and
more unwanted background events. These changes in
the ODB mean that we can also look at spectrums
not set up in the online ODB. For example,
instead of looking at the most energetic
coincidence gamma per event, we can look at the
sum of all gammas that trigger a BGO per event
(see figure 2). By summing the gammas we have
eliminated the lower energy peaks which were
triggered by 8MeV gammas depositing their energy
over more than one BGO. On analysis of the 14N
recoils, we see what appears to be clipping of
lower energy recoils, in the peak (figure 3). The
13C(p,?)14N reaction has a large cone angle of
approximately 19mrad, which is beyond the design
limits of DRAGON (approximately 16mrad).
Therefore, some recoils will not make it through
the beam tubes out of the gas target box, but
will be clipped, staying in the gas target box.
To find out what percentage of recoils werent
making it to the end detector, we needed to
create and run GEANT simulations of DRAGON and
this reaction.
DRAGON The DRAGON (Detector of Recoils And
Gammas Of Nuclear reactions) is situated at
TRIUMF, Canadas National Laboratory for Particle
and Nuclear Physics, which houses the worlds
largest cyclotron. DRAGON was designed to measure
radiative capture reactions in inverse kinematics
using a hydrogen or helium gas target. The DRAGON
system is basically a 21m recoil mass
spectrometer which can create elements via proton
or alpha capture reactions, and then separates
them based on mass, in two stages. Beam enters a
windowless gas target box, which is surrounded by
a closely-packed array of 30 gamma detectors made
of BGO (Bismuth Germanium Oxide) scintillation
crystals. A series of pumps are found either side
of the entrance and exit to the target, and are
used to keep the entrance and exit in vacuum
(10-7 Torr), allowing the beam to pass cleanly
through the target.
On leaving the target, the products (recoils) of
the nuclear reaction (together with original
beam, known as leaky beam) enter the first stage
of the mass spectrometer. The mass spectrometer
is made up of magnetic dipoles (M), magnetic
quadrupoles (Q), magnetic sextupoles (S), and
electrostatic dipoles (E), arranged in a two
stage mass separation (QQMSQQQSE)(QQSMQSEQQ).
The magnetic dipoles are used in such a way as to
separate out the charge state of interest, and
the quadrupoles and sextupoles focus the beam
through the spectrometer. The electrostatic
dipoles are used in such a way as to separate the
recoils from the leaky beam using their different
momentums. At the end of DRAGON is a choice of
two end detectors, a double-sided-silicon-strip
detector (DSSSD) and an ionization chamber (IC).
The DSSSD measures energy, position, and
time-of-flight. The IC measures energy and change
in energy (?E). The IC was used for the
13C(p,?)14N experiment, and will be used for the
future 13N(p,?)14O experiment, because it can be
used to separate out all the contaminate
elements, using ?E measurements.
GEANT GEANT is a Detector Description and
Simulation Tool. It is a program that simulates
the way in which elementary particles pass
through matter. It was originally designed for
High Energy Physics but is also today used in
medical and biological sciences, and
astronautics. The main applications of GEANT for
High Energy Physics are the tracking of particles
through an experimental setup for the simulation
of detector response, and the graphical
illustration of the setup and of the particle
trajectories. We used GEANT to create a replica
of the DRAGON separator, for simulations of
different astrophysical reactions, such as the
13C(p,?)14N reaction.
Figure 1
Figure 2
Figure 3
BGO simulations Due to the many excited states of
14N, and hence the large amount of gamma
cascades, I would start the analysis of
13C(p,?)14N by concentrating solely on the 8MeV
ground state gamma, which could be compare with
King et al2. But how could I separate out the
ground state gammas from the cascades? The GEANT
simulation of DRAGONs BGO array3 was used to
calculate the percentage of 8MeV gammas that
deposited all of their energy in a single BGO.
However, the BGO gamma array only covers 92 of
the solid angle of the gas target, and hence not
all gammas are registered. Of the gammas that did
register, 85.3 deposited their energy in a
single BGO, and 13.9 deposited their energy in a
BGO and its neighbour. From the diagram of the
simulated BGO array to the right, you can see
that it is very complex, so defining a
neighbouring BGO is difficult. To simplify, the
GEANT simulation was updated to use a cuboid
technique, where by if a BGO fires and another
fires a certain distance away which is within the
cube, then it is said to be a neighbouring BGO.
13C(p,?)14N simulations By creating an input file
for the 13C(p,?)14N reaction, I was able to
simulate this reaction through DRAGON to compare
with the actual data. Figure 3 shows a recoil
spectrum with a peak energy value of around 5MeV
for an actual DRAGON run. Simulating the same
conditions with the DRAGON GEANT simulation gave
a peak energy of 6.55MeV (figure 4). This 1.5MeV
energy difference was believed to happen as the
recoils pass through the entrance window (a mylar
foil) of the ionization chamber.
To test this theory, I started working on
creating an ionization chamber within GEANT for
our DRAGON simulation. Other motivations for
simulating the ionization chamber were to
a) get a proper estimate of energy straggling, b)
find out what anode the recoil ion stops in, c)
get a proper energy spectrum, d) compare with the
real data and estimate the acceptance loss, e)
simulate the correct geometry features of the
energy loss, f) test recoils in different
pressures within the ionization chamber.
Creating an ionization chamber in the DRAGON
simulation Using schematic diagrams of DRAGONs
actual ionization chamber, I was able to simulate
a simple ionization chamber with cuboids and
cylinders. A lot of FORTRAN code was needed for
the simulation and their were a lot of problems
with the designing. After months of work, the
simulated ionization chamber was operational, and
the DRAGON simulation was able to track recoil
particles through it.
Figure 4
Summary Initially, the 13C(p,?)14N reaction
experiment was used as an acceptance test for
DRAGON. We were pushing the limits of angular
acceptance of the DRAGON, due to the large
19.4mrad cone angle for this reaction. Also, the
very small cross-section meant a simulation was
necessary to investigate the acceptance loss
specific to this reaction. My simulations of the
13C(p,?)14N reaction in DRAGON, with the new
ionization chamber, are still continuing, with
histogram updates, and running with different
mistunes of distance, angle and percentage, of
the beam in the gas target. Once complete, my
analysis of the 13C(p,?)14N data compared with
analysis from my 13C(p,?)14N simulations, will
provide DRAGON with sufficient enough results to
be able to compensate for this reaction occurring
with the 13N(p,?)14O reaction. My creation of the
ionization chamber in the DRAGON simulation will
not only aid the DRAGONeers in distinguishing the
different elements in their future 13N(p,?)14O
data, but also help in the analysis of future
reaction studies when the ionization chamber is
used.
Acknowledgements I would like to thank Dr Chris
Ruiz, Dr Alison Laird, Dr Sabine Engel, Dario
Gigliotti, and Mike Lamey, for their close help
and support, throughout this project, and their
friendship during my year at TRIUMF. Also, I like
to thank Professor John DAuria for giving me
this excellent opportunity to come to this
facility, and experience nuclear astrophysics
outside of the classroom.
1 Authors Email Address ph91ab_at_surrey.ac.uk,
2 J. King et al., Nuclear Physics A 567 (1994)
354-376, 3 D.Gigliotti, Masters thesis,
University of Northern British Columbia (in
preparation) 2003