Upload
others
View
3
Download
0
Embed Size (px)
Citation preview
Silicon’s Limitations
Chris Kenney
April 28, 2011
Page 2
Key Parameters
• Material properties: mobilities, band gap, dielectric constant is 12!
• Geometry of device
• Operation: fields, temperature
GLAST/FERMI prior to launch
Page 3
Optical Light Absorption
• Very sensitive to wavelength
0.01
0.1
1
10
100
0 300 600 900 1200 1500
Abs
optin
Dep
th (m
icro
ns)
Wavelength (nm)
Optical Absorption
Page 4
X-ray Absorption
0.01
0.1
1
10
100
0 2000 4000 6000 8000 10000 12000 14000 16000 18000 20000 Abs
orpt
ion
Dep
th (m
icro
ns)
Energy (eV)
X-Ray Absorption
Page 5
Electron Absorption 3 keV electrons from Casino Almost all energy deposited in a sphere with a 100 nm diameter
Page 6
Electron Absorption
Page 7
Direct Sensing
• Primary, science particle is absorbed in silicon
• Match
Page 8
Current pulse
3D sensor Field null point
Page 9
Avalanche Mulitplication
• Intrinsically limited to thin regions • Electron multiplication factor much
higher than for holes • Requires careful device design
and fab • Can have slower recovery time • Can add extra noise
Page 10
Mobility - Electrons
Colder is faster Electrons
Page 11
Electrons go the distance
• Time for complete signal charge higher for holes
Page 12
Electrons go the distance
Collect electrons on the far face, if there is one. Minimize hole drift distance and maximize electron drift distance for a given thickness
p+
n+
i
3D Sensor
Use 200 microns thick sensor
100 micron pitch
50 micron n-to-p electrode spacing
Trench electrodes = uniform electric field
p+
n+
p+
n+
i i i
Delta Rays
• Changes total charge
• 3D scales waveform
• Planar introduces waveform distortion
• Large energy deltas will always degrade the timing
p+
n+
p+
n+
i i i
p+
n+
i
Page 15
Entrance Face
• Dielectrics – bad for electrons • Metals – bad for photons and
electrons • Heavily doped silicon – bad for
both – inefficient and slower • Anti-reflection coatings critical for
optical photons
Page 16
Diamond
• Higher mobilities • Lower dielectric constant • Higher bandgap • Optical transparency • Easy of fabrication?
30 nanometers
Platinum (3000A) on Mo (30 A) 18 Microns Thick Diamond
Collaboration with D. Pickard Nat. Univ. Singapore
ESRF Test
Placed in synchrotron beam
Attached to fast, discrete amplifier
Recorded bunch spacing period
Scope Trigger Count
I n t e r v a l (ns)
0 5
176.01
176.02
176.03
176.04
176.05
176.06
176.07
10 20 30 0
measured: mean=176.036ns σ = 16ps
σ = 16 ps
Collaboration with J. Morse of ESRF
Page 18
Signal to Noise
• Noise often limits the achievable resolution
• Encourages indirect sensing after multiplicative amplification of the primary
• Capacitance can be critical • Entrance face loss of electron energy
Page 19
Ideal
• Secondary electron accelerated to 3 keV – decent signal
• 300 nm sensor thickness • All singal charge collected within 5
picoseconds • Current pulse will have a rise time
several times faster
Page 20
Summary
• Match absorption thickness to particle
• Maximize electric fields
• Run cold
• Choose between direct and indirect sensing
• Entrance-face dead layer must be minimized
• Have electrons transit the long way
• Beware of capacitance
Page 21
Personnel
Sherwood Parker, Gary Varner, John Morse, Ed Westbrook, Al Thompson, Jasmine Hasi, Cinzia Da Via, Angela Kok, Giovanni Anelli, Dan Pickard, Niels van Bakel