AKM Page 1

1
CdZnTe Substrate Photoluminescencein HgCdTe
Arrays

• A. Mainzer

2
CdZnTe Behavior with ProtonIrradiation

• Analyzed 31 files from Ames/Bernie Rauscher
  containing full-frame darks from H1RG w/ 5 mm
  detector material
• Data appear as 1024x1024x2 FITS cubes a
  signal frame and a noise frame.
• Signal frame average of 5 frames
• Noise frame standard deviation of 5 frames
• Each frame has 8 samples (Fowler 4?)
• Easiest to spot CR strikes in noise frames, since
  hot/bad pixels dont show up as much.

3
Example CR Hit Glow
4
Two Examples
Hits seem to disappear after each frame, although
these are 5 frames averaged together, so its not
possible to check this. Persistence effect
unknown on a single-frame basis.
5
Amplitudes in Noise Frame

• Median level in clean part of frame 0.0003
• Size of 5x noise region 2400 pixels
• Size of 17x noise region
• 300 pixels
• Regions assymetric due to CR propagation thru
  substrate _at_ oblique angle
• Contours
• 0.001
• 0.0015
• 0.002
• 0.005
• 0.007
• 0.010

6
Another Hit

• Median level in clean part of frame 0.0003
• Size of 5x noise region 510 pixels
• Size of 17x noise region
• 80 pixels
• Regions asymmetric due to CR propagation thru
  substrate _at_ oblique angle
• Contours
• 0.001
• 0.0015
• 0.002
• 0.005
• 0.007
• 0.01
• 0.05
• 0.1

7
A Simulated WISE Frame

• Based on expected fluence of 1 hit/cm2/sec, we
  expect 40 hits/frame.
• This image simulates the appearance of a typical
  WISE frame w/o substrate removal ? YUCK

8
Scientific Impacts - Sensitivity

• Based on an estimated 0.5 billion sources in the
  sky, we expect 7000 point sources per WISE frame
  on average. Since most of sky is sparser, 2400
  is a reasonable.
• I added 2400 fake stars w/ FWHM 3 pixels to the
  simulated image ran a simple aperture
  photometry routine w/ inner radius 4 sky
  apertures 7-15 pixels.
• Measured in a clear (hit-free) region, the median
  SNR 11.
• Median of entire image (including hits) was SNR
  5, due to irregular background and increased
  noise caused by CRs.
• 1/2400 sources was not detected due to inability
  of photometry routine to fit a Gaussian profile
  since source landed on CR
• Although this is only a single WISE frame, not a
  stack of 8, this is a significant hit to
  sensitivity. Asteroid science will be
  particularly impacted.
• Area affected is large enough that many sources
  will suffer from overlapping CR hits.

9
Scientific Impacts - Reliability

• Production of significant numbers of false faint
  sources would negatively impact brown dwarf
  science, as many candidates will be single-band
  detections in band 2 ? cant use cross-band
  identification schemes
• An estimate of impact on reliability
• If 2500 pixels are affected/CR, and we get
  30-50 CR/frame (based on 18 mm pixels and 1
  CR/cm2/sec fluence), we can expect 10 of all
  pixels are affected per frame
• If we assume that a source must appear in 5/8
  frames to be counted as a real detection, the
  probability of overlapping regions is 1e-5
• Multiply this by 1024x1024 pixels to get per
  pixel probability 10
• Now assume that we require a source to cover 10
  pixels to appear as a point source ? 1 false
  source per set of 8
• We survey 40,000 sq deg / (47 arcmin2 FOV)
  65,000 looks at the same region
• ? There will be 60,000 false sources detected
• VERY rough calculation lacking details of
  precise source extraction algorithms