Two applications of nuclear
physics research
Henry R. Weller
Duke University and Triangle Universities Nuclear Laboratory
HIgS PROGRAM
Jlab Users Meeting, 2012
•HIgS
•High Intensity g-ray Source
(HIgS)
–Located at the Duke Free Electron
Laser Laboratory
Part of the Triangle Universities
Nuclear Laboratory (TUNL)
–Intra-cavity Compton Back
Scattering of FEL photons by
electrons circulating in the Duke
Storage Ring
Jlab Users Meeting, 2012
HIgS g-ray beam generation
Provides circularly and linearly polarized, nearly monoenergetic g-rays from 2 to 100 MeV
Utilizes Compton backscattering to generate g-rays
RF Cavity
Optical Klystron
FEL
Booster Injector
Mirror
LINAC
Jlab Users Meeting, 2012
Jlab Users Meeting, 2012
Some typical beam intensities
Eg(MeV)
Beam on target (DE/E = 3%)
1-2
2 x 107 g/s
8 – 16
2 x 108 (total flux of 2 x 109)
20 – 45
8 x 106
50 – 100 4 x 106 (will increase by x3-5 in 2012)
160
1.2 x 107 (by 2015)
Jlab Users Meeting, 2012
Show HIgS animation.
Jlab Users Meeting, 2012
MEGa-ray project at Livermore
X-band linac (250 MeV) and Chirped pulse amplification laser (530 nm)
Intended to produce 1020 g/s x mm2 x mrad2 x 0.1% bandwidth
0.5 – 2.5 MeV
Jlab Users Meeting, 2012
Main application:
Isotope Identification via Nuclear Resonance Fluorescence
We proposed a search for new signals to identify fissionable materials.
Jlab Users Meeting, 2012
A New Method for Identifying Special Nuclear Materials Based
Upon Polarized (g,n) Asymmetries
A TUNL/HIgS Project funded by the Department of
Homeland Security through their Academic Research
Initiative program
H. R. Weller—PI
M. Ahmed and Y. Wu -- Co PIs
Jlab Users Meeting, 2012
Introduction
In fissionable nuclei, energetic neutrons are
produced even at energies below (g,n)
threshold.
Jlab Users Meeting, 2012
Photofission Overview
E = ~1-2 MeV
E = ~0.5 MeV/A
E = ~5.5 - 7 MeV
A = 232-240
Z = 90-94
A = ~140
Z = ~55
10-20 s after
fission
10-18 s after
fission
A = ~95
Z = ~37
t
E = ~1.0 MeV/A
Jlab Users Meeting, 2012
4/6/2012
11
Experiment Schematic
• Polarized γ ray induces
fission of target nuclei
• Prompt neutrons are
detected both parallel
and perpendicular to
the plane of
polarization of the
incident γ ray
4/6/2012
φ = 90°
φ = 0°
γ-ray beam
Jlab Users Meeting, 2012
Preliminary Exam for J. M. Mueller
12
Spectra of fission neutrons from 235U and 238U using ~ 6 MeV polarized
g rays. 100 grams gives ~600 cts. in 1 minute.
NO neutrons are produced when this beam strikes iron, lead, carbon,
oxygen, nitrogen, etc.
Jlab Users Meeting, 2012
Neutron production below (g,n) threshold
Running at a g-ray energy of ~6.0 MeV and looking
at neutrons above 2 MeV only produces counts for
fissionable nuclei, except for d, Li and 9Be. These
can be identified by their unique spectra.
This provides a very promising tool for
interrogation which will receive further study.
Jlab Users Meeting, 2012
We actually use an array of 22 detectors
Jlab Users Meeting, 2012
Our discovery!
The ratio for depleted Uranium 238 is 3-to-1, while it is 1-to-1 for
enriched Uranium 235.
Jlab Users Meeting, 2012
Polarized Photofission Simulation
1.
2.
3.
4.
Measured fragment angular distributions and pure E1 assumption
allowed calculation of the fragment polarization asymmetries.
Fragment mass distribution taken from typical double-humped
mass distributions. Fragment velocities obtained from total
energy released of ~165 MeV.
Neutrons were emitted isotropically in fragment rest frame. With
kinetic energies given by evaporation spectrum modified to take
account of mulitple neutron emission.
Transformed neutrons into lab frame and counted detector hits to
generate polarized neutron asymmetries.
Jlab Users Meeting, 2012
Polarized Angular Distributions
Jlab Users Meeting, 2012
18
Polarized Angular Distributions
• At Eγ = 6.0 MeV
– 235U: Rf ~ 1
– 239Pu: Rf ~ 1
– 238U: Rf ~ 10
– 232Th: Rf ~ 150
• Significant expected Rf in photofission of
even-even actinides may lead to
significant Rn
Jlab Users Meeting, 2012
4/6/2012
19
Simulated Results
Jlab Users Meeting, 2012
20
Simulated Results
Jlab Users Meeting, 2012
21
Summary
We have discovered that a linearly polarized g-ray beam
can be used to distinguish between enriched and
depleted Uranium.
The ratio of neutrons parallel vs perpendicular to the
plane of polarization is ~3 for 238U and ~1 for 235U.
Studies continue on targets:
232Th, 233U, 237Np, 239,240Pu.
Measurement of prompt neutron polarization asymmetries in
photofission of 235,238U, 239Pu, and 232Th,
J. M. Mueller et al., Phys. Rev. C 85, 014605 (2012).
Jlab Users Meeting, 2012
Jlab Users Meeting, 2012
Development of an aneutronic Fusion Reactor
Based upon the 11B (p,a)aa reaction
Tri-Alpha Energy, Inc. –Foothills Ranch,Ca.
Controlled fusion in a field reversed
configuration and direct energy conversion.
(US patent: 7459654)
The 11B (p,a)aa reaction at low energies is dominated by the 2resonance at 675 keV which has a width of 300 keV. The Q-value
of the reaction is 8.58 MeV (Ein (cm)x 14).
Jlab Users Meeting, 2012
The 11B (p,a)aa reaction
When 675 keV protons strike 11B, a resonance in 12C is
formed at 16.6 MeV having a width of 300 keV.
This resonance decays by emitting a primary a-particle,
leading to the first excited state of 8Be which is 3 MeV
above the ground state. This state decays into two
secondary a-particles.
The state-of-the-art understanding of this reaction came
from a 1987 study which “proved” that the reaction
yielded one high energy primary a-particle, and two low
energy secondary ones.
Jlab Users Meeting, 2012
Tri-Alpha Energy Corporation
We have proven that this is incorrect!
But first—a few words on the concepts which
underlie the aneutronic fusion reactor presently
being developed at Tri-Alpha, Inc.
Jlab Users Meeting, 2012
Magnetic Field map of the Field Reversed Configuration
Jlab Users Meeting, 2012
Annular layer of plasma rotates around the null surface (86).
Collisions between injected p’s and 11B create “high energy”
alpha particles. Relative energy of protons to 11B is about 600
keV.
Jlab Users Meeting, 2012
The direct energy conversion system utilizes an Inverse Cyclotron
Converter (ICC). D’s are replaced by four semi-cylindrical
electrodes. Energy is removed from the alphas as they spiral past
the electrodes connected to a resonant circuit.
Jlab Users Meeting, 2012
The design of a 100 MW reactor is underway.
Test “shots” to demonstrate plasma
confinement are in progress.
Jlab Users Meeting, 2012
Brief History of the 11B (p,a)aa reaction
The history of this reaction is almost as long as the history of nuclear
physics itself.
Lord Rutherford studied the reaction over 75 years ago at ~200 keV at
the Cavendish Laboratory, measuring the ranges of particles coming
from the target. In his 1933 Proceedings of the Royal Society paper
he states:
“we might anticipate that the most probable mode of disintegration
would be for the three a-particles to escape symmetrically with
equal energies”.
(Proc. Roy. Soc. Lond. A 141, 259 (1933).)
Jlab Users Meeting, 2012
Three years later (1936), Dee and Gilbert, also of the Cavendish,
published the results of their expansion chamber studies of the
11B (p,a)aa reaction. Their 300 keV studies led them to conclude:
“the common mode of disintegration is into two [alpha] particles
which proceed at angles of 150o to 180o relatively to one another,
the third particle receiving very little energy”
The track images shown in their paper strongly support this
interpretation.
(Proc. Royal Soc. Lond. A 154, 279 (1936).)
Jlab Users Meeting, 2012
Brief History
Fast forward to 1987. Of course many experimental and
theoretical works were performed in the interim. But the
work of Becker, Rolfs and Trautvetter (Z. Phys. A327, 341
(1987)) established the modern view of this reaction:
“The reaction mechanisms have been studied via
kinematically complete coincidence measurements
showing that the reaction proceeds predominantly by
a sequential decay via 8Be”
Jlab Users Meeting, 2012
Primary and secondary distributions from Becker et al. The
two are about equal, leading to the factor of “2” in
determining the cross section.
Jlab Users Meeting, 2012
The 11B(p,a)aa reaction
“We want to know the energy and
location of every outgoing alpha
particle”
Tri-Alpha Energy Inc.
Jlab Users Meeting, 2012
Results of measuring the a-yields as a function of
incident proton and outgoing a-energies.
Previous results of Becker et al. indicated that only two alphas were
observed in finite detectors. We find 3. NACRE assumed 2 in its
Astrophysical S-factor compilation for this reaction. (This needs
correcting)
As will be shown later, the number of alphas is a function of both the
proton energy and the alpha energy.
To avoid this issue, we plot Counts instead of cross section, normalized
to the integrated luminosity and solid angle. Cross sections can be
obtained by dividing by the expected number of alpha particles, but
the absolute number of a’s is what is important to Tri-Alpha.
Jlab Users Meeting, 2012
Cross Section for the 11B(p,a)8Beaa Reaction
Although 3 a-particles are emitted, the number in a given energy interval
depends upon the reaction dynamics.
Therefore, instead of cross section, we report X = Counts/Luminosity
ie. X = Counts cm2/sr
Nt Np dW
Nt = # of target nuclei/cm2
Np = # of incident protons
dW = detector solid angle
X has the same units as differential cross section, but the # of a-particles
has not been divided out. It allows the user to find the number of
outgoing alpha particles an any energy and angle where X is known for
any specified Luminosity.
Jlab Users Meeting, 2012
Results showing the dominant 0.675 MeV resonance
Jlab Users Meeting, 2012
Two-step Reaction Simulation
Use experimental a-a phase shifts to generate the resonance
line shape of 8Be (2+). A pure single level assumption allows
correction for potential scattering using hard-sphere phase shifts.
d2 = - f2 + tan-1 ( G/2______)
ER + DR – E
The second term is what we are after.
f2 is the “hard-sphere” phase shift…easily calculated.
This avoids having to compute the energy dependence of G and
D. It all comes from the experimentally determined values of the
elastic scattering nuclear phase shifts d2.
Jlab Users Meeting, 2012
Two-step Reaction Simulation
Weight this with the probability that the a escapes from the 2resonance in the 12C nucleus (the penetrability for l=1 (or 3) alpha
particles) to get the probability distribution for populating the 8Be*
first excited state which subsequently decays into two secondary
a particles having equal energies in the 8Be* frame.
Randomly distribute the primary and secondary a's in the centerof-mass frame while preserving proper kinematics and boosts
back to the lab frame.
Jlab Users Meeting, 2012
Angular distributions
We need to know the form of the angular
distribution of both the primary alphas, and
of the secondary ones wrt. to the axis
defined by the primary alphas.
These can be calculated using the Zbar
formalism of Blatt and Biedenharn.
Jlab Users Meeting, 2012
For the 2- state at 0.675 MeV
For the 2- (un-natural parity) state at 0.675 MeV, we
have l=0 and l’=1 (according to Becker et al.).
This immediately leads to an isotropic angular
distribution for outgoing a’s, as observed
experimentally.
The distribution of the secondary a’s wrt. the
internal primary axis is found to be s(q) = 1 - P2
(cosq), which is proportional to sin2q.
This is when the fun begins!
Jlab Users Meeting, 2012
Two-step Reaction Simulation
Using these cm angular distribution functions, the
resulting outgoing a’s were transformed into the
lab system and checked to see if they would
enter any of our detectors.
If so, their energy was logged for the event,
leading to an energy distribution for each
detector.
Finally, it was possible to identify each as being
either a primary or a secondary a.
Jlab Users Meeting, 2012
Failure of the two-step model at the 2- resonance at
0.675 MeV (G = 300 keV).
Jlab Users Meeting, 2012
What went wrong?
When the 2- resonance in 12C decays to the first excited 2+
state of 8Be, the primary a-particle can have orbital
angular momentum of 1 or 3 units.
(must be odd; add vectors so 2 + 3 can equal 1 thru 5)
The 1987 work assumed the value of 1. We tried 3!
Jlab Users Meeting, 2012
To change to l=3 outgoing primary alpha particles
requires two changes in our simulation
1.
Replace the l=1 penetrabilites by l=3 ones when
generating the line shape of the 8Be 2+ state.
2.
Use the angular distribution for the secondary alpha
particles when the primary alphas have l=3.
This is obtained by considering
(S=2- + l=3  2+  S’=0 + l’=2
From Zbar(3232;2L) Zbar(2222;0L) PL(cosq)
We obtain: s(q) = C [1.0 + 2/7 P2(cosq) – 9/7 P4(cosq)]
Jlab Users Meeting, 2012
Which looks like:
Jlab Users Meeting, 2012
The effect of the l=3 penetrability on the resonance line
shape: a much narrower peak!
(down from 1.5 to ~0.7 MeV!!!)
Jlab Users Meeting, 2012
Success of the two step model with l=3
primary alpha particles
Jlab Users Meeting, 2012
And the angular distribution remains isotropic as observed.
Jlab Users Meeting, 2012
What can additional experiments
tell us?
We performed a coincidence experiment using position
sensitive solid state detectors to increase the solid angle
and provide data over an extended angular range.
Two detectors which each subtended an angular range of
+/- 8o were employed. Measured angular resolution was
0.2 degrees.
The detectors were placed symmetrically on the left and
right sides of the beam. An opening angle range of 100o
– 180o was scanned.
Jlab Users Meeting, 2012
A comparison of the coincidence spectra on the two resonances
at the same lab a-a opening angle of 150 degrees is quite
revealing!
Jlab Users Meeting, 2012
Coincidence spectra for the 0.675 MeV resonance (top) and the 2.64 MeV
(bottom) indicating two high energy alpha particles at 0.675 MeV. The lab a-a
opening angle is 150o
Jlab Users Meeting, 2012
Testing the model
The simulation was asked to predict the
coincidence counting rate as a function of
opening angle. Coincidence data were taken
for opening angles ranging from 100o-to-180o.
Jlab Users Meeting, 2012
Both a-particles were required to have energies greater than 3 MeV.
Excellent agreement is observed.
Jlab Users Meeting, 2012
Conclusions
We have found that the two-step sequential model with l
= 1 primary a-particles describes the data at the 2.64
MeV 3- resonance very well, but that l=3 primary aparticles are required to describe the data at the 0.675
MeV 2- state.
The l=3 assumption predicts the existence of two highenergy a-particles at an opening angle centered at
155o –as originally proposed by Dee and Gilbert in
1936! Our coincidence measurements confirm this
result.
Jlab Users Meeting, 2012
Huge impact on the reactor design!
Our discovery of TWO high energy a-particles is
having a huge impact on the reactor design.
They are much easier to extract and convert
more efficiently into electricity.
Tri-Alpha Energy is thrilled!
(our results are published in Physics Letters B
696 (2011) 26-29 and The Journal of Fusion
Energy (Springerlink.com), M.C. Spraker et al.,
2011)
Jlab Users Meeting, 2012
Questions remain….
Why does the 2- resonance at 16.6 MeV decay via l=3?
This resonance is assigned T=1, based upon its em decay
properties. Also, the other members of the isospin triplet
are observed in 12B and 12N.
How does this T=1 state decay into 3 alphas?
We are presently measuring secondary reaction cross
sections for TriAlpha Energy including 11B(a,a)11B,
11,10B(a,n) and 10B(p,alpha)7Be.
Jlab Users Meeting, 2012
Jlab Users Meeting, 2012
Jlab Users Meeting, 2012
Jlab Users Meeting, 2012
Jlab Users Meeting, 2012
Jlab Users Meeting, 2012
Jlab Users Meeting, 2012