Kara Hoffman

1
A Gedankenexperiment in Beam Profiling
Kara Hoffman Mark Oreglia
2
(No Transcript)
3
A promising bolometric material platinum

• high Z large dE/dx
• temperature-resistivity function very steep at
  20K
• should be more sensitive than the nickel and
  graphite prototypes previously tested in electron
  beams at Argonne

Platinum TCR curve
4
Specific application the linac test facility
Corresponding resistivity change in bolometer
strip
GEANT3 simulation
5
Bolometry summary

• Advantages
• doesnt disturb the beam
• relatively inexpensive
• robust
• Drawbacks
• must be applied to absorber window for heat
  sinking could be an issue mechanically/safetywis
  e and cannot be removed or replaced
• small signal, particularly for more diffuse beams
• metal strips provide challenge in large
  electromagnetic noise environment
• large thermal time constants

6
Diamond is prized for more than just its sparkle
(high refractive index)
low leakage I
very fast readout
no p-n junction needed
low capacitance
no cooling
hard
rad hard, strong
insensitive to gs lgt220nm
Makes a great particle detector!
The RD42 collaboration (CERN) has been developing
diamond (primarily) as a microvertex detector.
7
Anatomy of a diamond substrate microstrip
detector
Essentially a very compact solid-state ionization
chamber.
8
Polycrystalline CVD Diamond
induced charge
dx distance e-holes drift apart
m carrier mobility, t carrier lifetime
growth side

• Charge collection efficiency effected by
• grain boundaries
• in grain defects

substrate side
9
Can we read out both sides of the detector?
Naïve answer is yes, however, holes are generally
less mobile than electrons. Charge collection
efficiency could be much lower.

• Efficiency may not be an issue in a high
  intensity beam.
• Diamond has been successfully implemented as a
  pixel detector, but that complicates DAQ.
• Could flip polarity and alternate coordinates, if
  necessary.

E. Milani University of Rome
10
Radiation Hardness

• RD42 has irradiated diamond to proton fluences of
  and they still function

RD42
11
Annealing

• Pumping passivates traps, actually improving
  charge collection efficiency up to a fluence of

Initially, smaller signals become larger and
response becomes more uniform.
RD42
Collected charge (e-)
Charge threshold above which 90 of events fall
RD42
12
Electronics

• Power supply/amplifier must maintain bias
  voltage while reading out a potentially large
  signal.
• High bandwidth Fast electronics are desirable to
  exploit the excellent timing characteristics.
• Wide dynamic range Large variation in
  intensities to be measured
• German nuclear scientists have developed such an
  amplifier DBA-II (Diamond Broadband Amplifier)
  It can be done.

We have a 1 GHz bandwidth Tektronix oscilloscope
with 400 ps rise time, 10 GS/s sample rate. Good
enough?
Electronics is not a showstopper. Level of
complexity depends on timing resolution desired.
13
A (Destructive) Test
Open Questions
Place a small (1sq. cm) diamond detector with a
single electrode on each side in a proton beam.

• Can we simultaneously read 2 coordinates from the
  same detector?
• How will they perform with such a large
  instantaneous particle flux? Will the response be
  linear or is there some saturation point?
• What kind of time resolution can we achieve?
• Quantify rad-hard. No one has irradiated them
  to a fluence where they had no signal.

Monitor both electrodes and compare signal
strength for electrons and holes.
Continue until signal completely disappears.
Remetalize electrodes and repeat process to
determine whether the diamond is toast or the
electrodes simply vaporized.
14
Potential advantages/payoffs

• sensitive (2 coordinate?) measurement
• relatively huge signal
• fast (subnanosecond 40ps) response might allow
  temporal beam profiling, in addition to current
  and position measurements
• free standing-accessible
• low Z- no beam loss (lt0.1)
• could be implemented quickly
• RD42 has already developed a vendor (DeBeers)
• could have broader applications
• for other beams single particle efficiency,
  high bandwidth make it suitable for transfer
  lines, etc.
• measuring Lab G death rays and other RF
  cavities

15
Now back to the linac test facility

• Ed Black and I have sketched a robotic arm to
  sweep the diamond sensor through the beam
• Allows us to pull sensor out of the beam, thus
  increasing sensor life by minimizing radiation
  exposure.
• A single sensor can be used to sample the beam at
  several different radii, thus minimizing cost
  while still allowing us to make a full 3s
  measurement.

16
Diamond quality cut, clarity COST!
Industrial diamond is manufactured primary for
heat sinking or optics. For HEP, the figure of
merit is charge collection distance.

• Size of single crystals
• Density of traps
• Purity
• Polishing
• Uniformity of thickness and response

Price range Electronics grade (P1 diamond, and
others) 100/cm2 Tracking grade (DeBeers only)
gtgt1000/cm2
We dont need single particle efficiency to see a
beam. Crappy diamond will probably work.
17
Needs?
Component Supplier Cost
2 CVD diamond wafers 5x5 cm each DeBeers Industrial (via U. of Toronto) Waiting for a quote from W. Trischuk
8 mm x 8 mm diamond square P1 Diamond 150 - cost goes down with volume
Metal lead sputtering Rutgers U. lt300
data acquisition electronics U. Chicago -
We probably dont need this quality.
18
Final Thoughts

• I am negotiating with RD42 to borrow/purchase a 1
  sq. cm piece of tracking grade diamond, and Im
  purchasing some electronics grade diamond for
  destructive test.
• I believe we can beam test (or nuke rather)
  some samples on a short time scale.
• If electronics grade stuff works, it could be
  disposable.
• This talk probably no longer belongs in an
  absorber review.

19
Extra slide comparison of physical properties of
polycrystalline and single crystal diamond