Dark Matter Detection

1
Dark Matter Detection

• PHY 210 project
• Spring 2007
• Bogdan, Doug, Jay and Ragnhild

2
The idea

• WIMPs (Weakly Interacting Massive Particles) are
  hypothetical particles that offer one solution to
  the dark matter problem that the measured mass
  of galaxies is much lower than the mass of matter
  we can see
• WIMPs, by hypothesis, do not interact with
  electromagnetism, so cannot be seen directly. The
  idea is to detect them from nuclear recoils with
  some target nuclei in a detector.

3
In theory

• Build a detector to measure the recoil of the
  nuclei of some atoms from the dark matter.
  Calculate from this the relation between the
  event rate and the recoil energy, to infer
  various things about the structure of dark
  matter.
• Possible things to measure/inferMass of dark
  matter particleIncident kinetic energy of DM
  particleEvent rate/flux of DM through the earth
  (lab).

4
In practice

• Build a detector filled with Argon. Scintillation
  from nuclear recoil is picked up by
  photomultiplier tubes. After processing these
  signals through electronics/LabView, we can get
  the energy spectrum of the particles.

5
From those results

• After we have the spectrum (calibrated with known
  sources) we can find the energy of nuclear recoil
  and infer the event rate. From here, the
  equationdR/dER(R0/E0r)e-E(R)/E(0)rwill allow
  us to find DM mass, flux, etc., where R is the
  event rate per unit mass, ER is the recoil
  energy, E0 is the incident kinetic energy, R0 is
  the total event rate, and r is a kinematic factor
  involving the mass of the nucleus of the target
  particle and the mass of a DM particle.

6
Why Argon?

• Really, any old noble gas should do. The key is
  that we want something which is extremely stable,
  and scintillates under a fairly wide range of
  incident kinetic energies. Argon seems to fit
  the bill.
• In more serious experiments, Argon is most
  commonly used (in its liquid form) for exactly
  these properties. So, at least we have something
  to model our experiment after.

7
Setup
High-voltage supply for phototubes
Oscilloscope, so we can see whats going on
Fancy electronics
Container with Ar
Photomultiplier tubes
8
The detector itself
Plenty of electrical tape to ensure light doesnt
leak in
The container itself is a cylindrical steel tube
Photomultiplier tubes were glued into a G10
holder, which was screwed into the steel endpieces
One of the endpieces had an outlet used for
evacuating, and pumping in Ar
9
The electronics
The signal from the phototubes goes to the VME
Crate
From the VME crate, through a discriminator, and
then to a coincidence counter Signal from
phototube sent to computer, triggered on
coincidence.
(Thanks to prof. Cristiano Galbiati!)
10
High-voltage supply The voltage across the
phototubes was 2300 V
We viewed output from both tubes and from
coincidence counter on oscilloscope
11
Scintillation

• When we calibrated this apparatus, we used an
  organic scintillator with gamma rays from a few
  sources (137Cs, 133Ba, and 60Co). Scintillators
  emit photons with energies proportional to the
  incoming radiation, up to some scaling factors.

12
Some Challenges

• Container needs to be absolutely tight, and able
  to take vacuum, pressure (Solution? O-rings and a
  LOT of glue.)
• Must be absolutely dark - photons leaking in will
  both disturb signals and potentially burn out
  phototubes (Solution? A LOT of black tape.)

13
Technical problems

• This experiment turned out to be very
  challenging. At first we had trouble finding two
  phototubes that worked. Of our original two
  phototubes, we found out that one didnt work
  only after gluing it in. This caused a delay of
  about 5 days, as we had to remove it from the G10
  holder, and then glue a new one in.
  Unfortunately, we then found this second
  phototube sparked, burning out part of the VME
  crates electronics in the process. So we had to
  settle for a smaller second phototube, as we
  didnt have another tube the size of our first
  one. Overall, we ended up wasting about 3 weeks
  before having two working phototubes.
• Then, we had a bit of trouble running the Argon
  tests. When we tried putting the gas in at 12 psi
  one of the phototubes was pushed through its
  casing, which resulted in light coming in and the
  terminating of the run. We then tried a second
  run, with the gas at 7 psi, but about 3 hours
  into the experiment one of the phototubes
  registered a big weird-shaped pulse followed by
  no more activity. When we stopped the run, the
  other phototube was very hot and possibly burnt
  out. We believe it might have caused a spark
  which was the signal we picked up in Labview. We
  decided to try no more runs afterwards.

14
Normal pulse
Total number of pulses
Shape of the last pulse
Number in linear correspondence
to the energy of the last
pulse
Histogram of all the pulses
The histogram in semi-logarithmic scale
15
Kaboom pulse
Weird-shaped, very disturbing pulse
This number is one order of magnitude higher
than what we used to be getting
16
Close-up
17
Data (Pretty graphs)

• Barium

Cesium
All Sources
18
Main Graph
Cobalt was the best graph that we got. The
related peaks and decays should tell us the
energy spectrum we get from Cobalts gamma rays
(as well as a possible spectrum for muons).
Red Original Data Black Original data minus
no-source data Blue Muon data (no sources)

• Cobalt

19
Deep conclusions?

• Unfortunately, we never got a chance to actually
  take data with the argon. But were sure that if
  we had gotten data, it would turn the scientific
  world on its ear.
• One conclusion we do have is that experiments
  never go right the first time, but making
  mistakes isnt the end of the world. Also, that
  strange pulse shapes are very, very, very scary.
  Very.

20
Further development

• In terms of accuracy, this experiment has a lot
  of room to grow. A more shielded chamber,
  well-calibrated phototubes (preferably of the
  same size and type), and a better understanding
  of the underlying physics are all things which
  could be improved.
• In terms of the experimental method, were fairly
  spot-on. The experiments which are being
  performed now use the same basic setup PM tubes
  and scintillators, argon-filled chambers, and
  computerized data analysis. Such experiments
  also have a longer time scale, with many repeated
  measurements and runs being performed for hours
  at a time.

21
Credits

• Lyman Page
• Cristiano Galbiati
• Wei Chen
• Joe Horvath