Polarimeter Construction for measuring the depolarization of neutrons in Deuterium Matthew Bowers, Orlay Sint, Dr. Alexander Komives

1
Polarimeter Construction for measuring the
depolarization of neutrons in DeuteriumMatthew
Bowers, Orlay Sint, Dr. Alexander Komives
Polarimeter Construction and Testing Polarimete
rs are essentially large and powerful
electromagnets designed to maintain a high
magnetic field on the inside, but minimize
leakage to the outside. The metal for our
polarimeters was chosen for its properties to
hold a very strong field. To reduce field leakage
both top and bottom at flat plates and an iron
ring was placed around the outside. The effect is
an approximately 2 Tesla magnetic field which
cannot be detected by a compass from more than a
foot away and barely keeps a metal washer from
falling off.
Abstract Each of the fundamental forces has
constants associated with its strength. One of
the constants for the weak force is the r meson
coupling coefficient There is a proposed
experiment to measure this constant using
deuterium capturing polarized neutrons to make
tritium and polarized gamma rays. The calculation
of the r meson coupling coefficient requires that
the neutrons be polarized when they are captured.
But it is known that the absorption is not very
efficient so there is a possibility that before
it is captured the neutron will lose its spin.
The degree of neutron polarization can be
measured by passing the polarized gamma rays
through devices containing polarized electrons in
the form of magnetized metal. Two of these
devices, called polarimeters, have been built and
initial tests have been performed.
Cesium Iodide Scintillator and Photo Multiplier
Tubes
The detector is a Cesium Iodide Crystal (CsI)
with a photo multiplier tube (PMT) on each end.
CsI is a scintillator, which will release flashed
of light when photos enter it. The gamma rays in
this experiment have enough energy that the
electrons of the crystal are actually ionized,
recaptured, and de-excite producing a visible
photon. These photons then travel down the length
of the crystal and into the PMT. The PMT has a
clear lens with a transparently thin coating of
metal. The photos knock electrons off this layer
as they pass. The remained of the tube is a
series of metal plates with a high voltage. The
result is the electrons accelerate to a plate,
collide, producing more electrons, then
accelerate to the next plate. By the end enough
electrons have cascaded that there is a
discernable current which can be measured and
tabulated. The larger this current the larger the
initial gamma ray was.
This polarimeter has 493 turns of 16 gauge Hyslik
coated magnet wire. To produce the field we have
6 amps flowing at about 20 volts. This amount of
current produces a great deal of heat, so in
amongst the wire are 6 turns of 1/8 inch
refrigerator tubing to act as a cooling system.
The wire has a heat tolerance of 200 degrees but
we will be running for up to days at a time
during which quite a bit of heat can build up.
After an hour or so the outside of the
polarimeter is warm to the touch but to this
point we have not melted the insulation off of
our wires. Additionally, around the core there
are 4 turns of 18 gauge wire to act as a field
detector. When the main field is turned on or off
a current is induced across the 18 gauge wire.
This is a simple and effective way to check the
strength of the field and how long it takes to
change its polarity.
Background
The magnetic field produced by the polarimeter is
key to all of our measurements, but at the same
time it can potentially sabotage them. The
photomultiplier tubes use charges plates to
accelerate electrons, so by introducing another
field we cause the electrons to travel with a
different trajectory. This would not be too much
of a problem if the field were going to be
constant but it will be switched back and forth.
Moreover the nature of the experiment is to find
the difference in the number of gamma counts
independent of PMT effects. To this end we tested
extensively to find what kind of effect the
magnetic fields would have on our count rates.
Then to try to nullify these effects we used
magnetic shielding to protect the PMT from the
magnetic field of the polarimeter. The results of
these tests are highlighted below.
PMT time dependency and the 16 bit test
The plot to left shows eight ten second trials
all with a positive polarimeter polarity. All of
these lines should be all but identical but they
vary quite substantially. It seems that Photo
Multiplier Tubes have a regular time dependency.
To get around this measurements are taken in a
regularly osculating pattern. We found the most
success with a 16 bit pattern. To do this we took
16 consecutive measurements but the polarity of
the polarimeter went as follows - - - -
- - - - . To complete the analysis the s
and s are added according to their kind and
uses as a single measurement. The hope here is
that the variation in the PMT will average out
over all of these trial. We are confident that it
has.
Energy Resolution and Crystal Degradation
Testing The CsI crystals were chosen for this
project not only because they were available, but
also because they are a very fast crystal meaning
that more counts will be taken than with a slower
crystal such as NaI. The sacrifice of a faster
crystal however is a loss of energy resolution.
This is a measure of how well the crystal can
discern between gamma energies. In the charts
below a clear picture of the energy resolution is
shown. These histograms were produced with a
Nucleus brand multi-channel analyzer. A MCA
recieves the signals from the PMT and assigns a
channel to each different energy, this is known
as digitizing because the analog pulse comes in
from the PMT and all that is left is a digital
number. At top is a chart from a NaI crystal
with peaks for 137Cs, 22Na, and 60Co. At bottom
is the same measurements taken on the CsI
crystal. Notice how the NaI peaks are much
sharper. The most obvious difference is the fact
that the 2 60Co peaks merge into one on the CsI
chart.
Above are histograms of a Cs 137 source with the
Crystal and PMT covered with black felt. Though
they look the same you can see at left the purple
and blue points match up much better than at
right where the blue is to the right of the
purple. This comes from the positioning and
shielding on the polarimeter. To left the
polarimeter is about 34-40 cm away with magnetic
shielding surrounding it. On right the
polarimeter is about 10 cm away and there is no
shielding. This is the type of effect we can see
from the magnetic field. Quantitatively we
measured the centroids of the curves. The change
in centroid on the left is .07 channels, where at
right the change is 1.83 channels.
For our experiment to be a success we will need
to measure the asymmetry of the gamma rays. For
this reason we started these tests measuring the
asymmetry caused not by the filtering of gammas
but only by the changing magnetic field on the
PMT. Unfortunately after much analysis we decided
that our error was too large and we could not
take enough data fast enough with our current
equipment to see the changes that should be
there. For this reason we started to examine the
change in centroid from trial to trail. However
in a last attempt to see a non-zero asymmetry
because of the magnetic field we put aside the
polarimeter and used a completely unshielded
solenoid at about 5 cm from the PMT. Amazingly
the asymmetry was still consistent with zero when
error was taken into account. This trial however
was useful in reaffirming the was the magnetic
field changes the centroid.
Proposed Setup
This shows the general set up of our experiment.
Though the reaction is the same as that above we
are not concerned with the asymmetry due to the
weak interaction but only the degree of
polarization of the neutrons.
The data analysis for this test will be given to
show how we are drawing results. Just despite the
consistency of the centroid shift, we do not know
why it would happen. It is thought that if we
could run more faster tests with less dead time
we would be able to see the asymmetry after 50
some measurements. This is obviously not
practical until the process is automated.
CsI in iron box
Polarimeter
?
Sn
Numbers of
Interest The asymmetry is
the primary number of
interest though, as seen on
the chart is consistent
with zero. We anticipated a
non-zero value because of
the strength of the
magnetic field.
The
centroid is the center of
mass of the area under the
curve. Away from the presence of a magnetic field
(such as above left) the centroid does not move.
The stronger and nearer the field to the PMT the
more the centroid moves. Interestingly enough,
the centroid always moves in the same direction
when there is a field even when the polarity of
the field is changed. This is possibly caused by
a residual magnetic field on the magnet that is
not quite overcome by changing the polarity.
Perhaps testing with identical permanent magnets
would yield a result where the negative polarity
and positive polarities would be on opposite
sides of no field.
Polarized neutrons will be directed toward the
Deuterium target where they will be captured
producing tritium and 6.2 MeV gamma rays. The
gamma will then enter one of the polarimeters.
Deuterium Target
Vn
n
Polarized neutron beam
?
Polarimeter
CsI in iron box
In the polarimeter the gamma will collide with
the electrons of the metal and undergo Compton
scattering. Because of the strong magnetic field
(2 Tesla), some of the electrons will be spinning
in the same direction. If the electron is
spinning parallel to the spin of the gamma it
will be scattered away from the CsI detector. If
the spins are anti-parallel the gamma will pass
through to the CsI scintillator where they will
be detected and measured.
Conclusions It appears as though the crystals
have not degraded too much to be useful for our
measurements. They are in fact large enough to
pick up multiple instances simultaneously. The
Photomultiplier tubes work but in such a way that
a 16 bit sampling method is best to limit time
dependent detection efficiency. The crystals are
most effective when the radiation source is near
to their center. The cause of this is unknown but
thought to be geometric effect as the gammas are
more likely to hit the crystal if they occur near
the center. With current measuring techniques we
are unable to detect asymmetries, though error
has been as low as 4 x 10-3. Because of this we
were not able to see if the magnetic field of the
polarimeter has any effect on the asymmetry or
not. However, we were able to remove almost all
of the centroid change inherent to the presence
of a magnetic field using a combination of
spacing and magnetic shielding. It is important
to note that the asymmetries for these test
should be zero, as the gammas are not polarized.
It was expected, and observed on previous test
that the polarimeters would cause asymmetries in
the counts taken. It is also possible that these
polarimeters were well enough constructed that
those effects are not being seen with this set
up. It is much more likely, however, that we have
too much systematic error to be able to see what
is there. To further improve upon these results
an automated data collection system is being
constructed. With this we will be able to take
more shorter samples faster. These samples should
allow us to be more precise in the amount of time
we are collecting data, as well as eliminate some
of the time dependencies of the PMT.
Acknowledgments DePauw University, Science
Research Fellows, Dr. Komives Help Andy
Smith Equipment Dr. Kertzman, IU, NIST, Rea
Magnet Wire Papers
Part of the degradation of the crystals is that
they will absorb water out of the air. This water
will cloud the crystal and reduce its
effectiveness at detecting peaks. According to
previous testing over the last 10 years these
crystals have lost about 10 of their energy
resolution. This should not be a problem because
of the nature of our project but it is possible
to repair them by polishing off the clouded layer
with very fine sand. This would result in a
clearer crystal, but a loss in size which is also
important for gamma detection.