Semiconductor detectors

1
Semiconductor detectors

• An introduction to semiconductor detector physics
  as applied to particle physics

2
Contents

• 4 lectures cant cover much of a huge field
• Introduction
• Fundamentals of operation
• The micro-strip detector
• Radiation hardness issues

3
Lecture 1 - Introduction

• What do we want to do
• Past, present and near future
• Why use semiconductor detectors

4
What we want to do - Just PPE

• Track particles without disturbing them
• Determined position of primary interaction vertex
  and secondary decays
• Superb position resolution
• Highly segmented ? high resolution
• Large signal
• Small amount of energy to crate signal quanta
• Thin
• Close to interaction point
• Low mass
• Minimise multiple scattering
• Detector
• Readout
• Cooling / support

5
Ages of silicon - the birth

• J. Kemmer
• Fixed target experiment with a planar diode
• Later strip devices -1980
• Larger devices with huge ancillary components
• J. Kemmer Fabrication of a low-noise silicon
  radiation detector by the planar process, NIM
  A169, pp499, 1980

6
Ages of Silicon - vertex detectors

• LEP and SLAC
• ASICs at end of ladders
• Minimise the mass inside tracking volume
• Minimise the mass between interaction point and
  detectors
• Minimise the distance between interaction point
  and the detectors
• Enabled heavy flavour physics i.e. short lived
  particles

7
ALEPH
8
ALPEH VDET (the upgrade)

• 2 silicon layers, 40cm long, inner radius 6.3cm,
  outer radius 11cm
• 300mm Silicon wafers giving thickness of only
  0.015X0
• S/N rF 281 z 171
• srf 12mm sz 14mm

9
Ages of silicon - tracking paradigm

• CDF/D0 LHC
• Emphasis shifted to tracking vertexing
• Only possible as increased energy of particles
• Cover large area with many silicon layers
• Detector modules including ASICs and services
  INSIDE the tracking volume
• Module size limited by electronic noise due to
  fast shaping time of electronics (bunch crossing
  rate determined)

10
ATLAS

• A monster !

11
ATLAS barrel

• 2112 Barrel modules mounted on 4 carbon fibre
  concentric Barrels, 12 in each row
• 1976 End-cap modules mounted on 9 disks at each
  end of the barrel region

12
What is measured

• Measure space points
• Deduce
• Vertex location
• Decay lengths
• Impact parameters

13
Signature of Heavy Flovours
Stable particles t gt 10-6 s Stable particles t gt 10-6 s ct
n 2.66km
m 658m
Very long lived particles t gt 10-10 s Very long lived particles t gt 10-10 s Very long lived particles t gt 10-10 s
p, K, KL0 2.6 x 10-8 7.8m
KS0, E, D0 2.6 x 10-10 7.9cm
Long lived particles t gt 10-13 s Long lived particles t gt 10-13 s Long lived particles t gt 10-13 s
t 0.3 x 10-12 91mm
Bd0, Bs0, Db 1.2 x 10-12 350mm
Short lived particles Short lived particles Short lived particles
p0, h0 8.4 x 10-17 0.025mm
r,w 4 x 10-23 10-9mm!!
14
Decay lengths
E.g. B ? J/Y Ks0
L
Secondary vertex
Primary vertex
L p/m c t

• By measuring the decay length, L, and the
  momentum, p, the lifetime of the particle can be
  determined
• Need accuracy on both production and decay point

15
Impact parameter (b)
b distance of closest approach of a
reconstructed track to the true interaction
point
b
beam
16
Impact parameter

• Error in impact parameter for 2 precision
  measurements at R1 and R2 measured in two
  detector planes
• af(R1 R2) and function of intrinsic resolution
  of vertex detector
• b due to multiple scattering in detector
• c due to detector alignment and stability

17
Impact parameter

• sb f( vertex layers, distance from main vertex,
  spatial resolution of each detector, material
  before precision measurement, alignment,
  stability )
• Requirements for best measurement
• Close as possible to interaction point
• Maximum lever arm R2 R1
• Maximum number of space points
• High spatial resolution
• Smallest amount of material between interaction
  point and 1st layer
• Good stability and alignment continuously
  measured and correct for
• 100 detection efficiency
• Fast readout to reduce pile up in high flux
  environments

18
Impact parameter
Blue 5mm Black 1mm (baseline) Green 0.5
mm Red 0.1 mm
Effect of extra mass and distance from the
interaction point
Lower Pt
GR Width Flux increase() to silicon Improvement of the IPres. wrt 1mm()
5mm -44 -38.1?0.9
0.5mm 14.1 5.8 ?0.7
0.1mm 27.7 10.0 ? 0.7
Guard Ring Width Impact on d0 Performances and
Structure Simulations. A Gouldwell, C Parkes, M
Rahman, R Bates, M Wemyss, G Murphy, P Turner and
S Biagi. LHCb Note, LHCb-2003-034
19
Why Silicon

• Semiconductor with moderate bandgap (1.12eV)
• Thermal energy 1/40eV
• Little cooling required
• Energy to create e/h pair (signal quanta) 3.6eV
  c.f Argon gas 15eV
• High carrier yield
• ? better stats and lower Poisson stats noise
• Better energy resolution and high signal
• ? no gain stage required

20
Why silicon

• High density and atomic number
• Higher specific energy loss
• Thinner detectors
• Reduced range of secondary particles
• Better spatial resolution
• High carrier mobility ? Fast!
• Less than 30ns to collect entire signal
• Industrial fabrication techniques
• Advanced simulation packages
• Processing developments
• Optimisation of geometry
• Limiting high voltage breakdown
• Understanding radiation damage

21
Disadvantages

• Cost ? Area covered
• Detector material could be cheap standard Si
• Most cost in readout channels
• Material budget
• Radiation length can be significant
• Effects calorimeters
• Tracking due to multiple scattering
• Radiation damage
• Replace often or design very well see lecture 4

22
Radiation length X0

• High-energy electrons predominantly lose energy
  in matter by bremsstrahlung
• High-energy photons by ee- pair production
• The characteristic amount of matter traversed for
  these related interactions is called the
  radiation length X0, usually measured in g cm-2.
• It is both
• the mean distance over which a high-energy
  electron loses all but 1e of its energy by
  bremsstrahlung
• the mean free path for pair production by a
  high-energy photon

23
Lecture 2 lots of details

• Simple diode theory
• Fabrication
• Energy deposition
• Signal formation

24
Detector p-i-n diode

• Near intrinsic bulk
• Highly doped contacts
• Apply bias (-ve on p contact)
• Deplete bulk
• High electric field
• Radiation creates carriers
• signal quanta
• Carriers swept out by field
• Induce current in external circuit
  ? signal

n contact ND1018cm-3
ND1012cm-3
p contact NA1018cm-3
25
Why a diode?

• Signal from MIP 23k e/h pairs for 300mm device
• Intrinsic carrier concentration
• ni 1.5 x 1010cm-3
• Si area 1cm2, thickness300mm ? 4.5x108
  electrons
• 4 orders gt signal
• Need to deplete device of free carriers
• Want large thickness (300mm) and low bias
• But no current!
• Use v.v. low doped material
• p rectifying (blocking) contact

26
p-n junction
Carrier density
p
n
(5)
(1)
(2)
Electric field
(6)
Dopant concentration
(3)
Electric potential
Space charge density
(4)
(7)
27
p-n junction

• take your samples these are neutral but doped
  samples p and n-
• bring together free carriers move
• two forces drift and diffusion
• In stable state
• Jdiffusion (concentration density) Jdrift
  (e-field)
• p area has higher doping concentration (in this
  case) than the n region

28
p-n junction

• Fixed charge region
• Depleted of free carriers
• Called space charge region or depletion region
• Total charge in p side charge in n side
• Due to different doping levels physical depth of
  space charge region larger in n side than p side
• Use n- (near intrinsic) ? very asymmetric
  junction
• Electric field due to fixed charge
• Potential difference across device
• Constant in neutral regions.

29
Resistivity and mobility

• Carrier DRIFT velocity and E-field
• mn 1350cm2V-1s-1 mp 480cm2V-1s-1
• Resistivity
• p-type material
• n-type material

30
Depletion width

• Depletion Width depends upon Doping Density
• For a given thickness, Full Depletion Voltage is
• W 300mm, ND ? 5x1012cm-3 Vfd 100V

31
Reverse Current

• Diffusion current
• From generation at edge of depletion region
• Negligible for a fully depleted detector
• Generation current
• From generation in the depletion region
• Reduced by using material pure and defect free
• high lifetime
• Must keep temperature low controlled

32
Capacitance

• Capacitance is due to movement of charge in the
  junction
• Fully depleted detector capacitance defined by
  geometric capacitance
• Strip detector more complex
• Inter-strip capacitance dominates

33
Noise

• Depends upon detector capacitance and reverse
  current
• Depends upon electronics design
• Function of signal shaping time
• Lower capacitance ? lower noise
• Faster electronics ? noise contribution from
  reverse current less significant

34
Fabrication

• Use very pure material
• High resistivity
• Low bias to deplete device
• Easy of operation, away from breakdown, charge
  spreading for better position resolution
• Low defect concentration
• No extra current sources
• No trapping of charge carriers
• Planar fabrication techniques
• Make p-i-n diode
• pattern of implants define type of detector
  (pixel/strip)
• extra guard rings used to control surface leakage
  currents
• metallisation structure effects E-field mag ?
  limits max bias

35
Fabrication stages

• Stages
• dopants to create p- n-type regions
• passivation to end surface dangling bonds and
  protect semiconductor surface
• metallisation to make electrical contact

• Starting material
• Usually n-
• Phosphorous diffusion
• P doped poly n Si

n- Si
36
Fabrication stages

• Deposit SiO2
• Grow thermal oxide on top layer
• Photolithography etching of SiO2
• Define eventual electrode pattern

37
Fabrication stages

• Form p implants
• Boron doping
• thermal anneal/Activation
• Removal of back SiO2
• Al metallisation patterning to form contacts

38
Fabrication

• Tricks for low leakage currents
• low temperature processing
• simple, cheap
• marginal activation of implants, cant use IC
  tech
• gettering
• very effective at removal of contaminants
• complex

39
Energy Deposition

• Charge particles
• Bethe-Bloch
• Bragg Peak
• Not covered
• Neutrons
• Gamma Rays
• Rayleigh scattering, Photo-electric effect,
  Compton scattering, Pair production

40
Charge particles- concentrating on electrons

• At ?? ? 3 dE/dx minimum independent of absorber
  (mip)
• Electrons
• mip _at_ ?1 MeV
• Egt50 MeV radiative energy loss dominates

Momentum transferred to a free electron at rest
when a charged particle passes at its closest
distance, d. integrate over all possible values
of d
41
Well defined range

• at end of range specific energy loss increases
• particle slows down
• deposit even more energy per unit distance

Bragg Peak
E 5 MeV in Si (increasing charge)
R (?m) p 220 ? 25 16O 4.3
Useful when estimating properties of a device
42
Energy Fluctuation

• Electron range of individual particle has large
  fluctuation
• Energy loss can vary greatly - Landau
  distribution
• Close collisions (with bound electrons)
• rare
• energy transfer large
• ejected electron initiates secondary ionisation
• Delta rays - large spatial extent beyond particle
  track
• Enhanced cross-section for K-, L- shells
• Distance collisions
• common
• M shell electrons - free electron gas

43
e/h pair creation

• Create electron density oscillation - plasmon
• requires ?17 eV in Si
• De-excite almost 100 to electron hole pair
  creation
• Hot carriers
• thermal scattering
• optical phonon scattering
• ionisation scattering (if E gt 3/4 eV)
• Mean energy to create an e/h pair (W) is 3.6
  eV in Si (Eg 1.12 eV ? 3 x Eg)
• W depends on Eg therefore temperature dependent

44
Delta rays

• Proability of ejecting an electron
• with E ? T as a function of T
• b) Range of electron as a
• function of energy in silicon

45
Displacement from d-electrons

• Estimate the error
• Assume 20k e/h from track
• 50keV d-electron produced perpendicular to track
• Range 16mm, produces 14k e/h
• Assume ALL charge created locally 8mm from
  particles track

46
Consequences of d-electrons

• Centroid displacement

• Resolution as function of pulse height

47
Consequence of d-electrons
45º
45º
15 ?m
E.g. CCD
300 ?m
Most probable E loss 3.6keV 10 proby of 5keV
? pulls track up by 4 ?m
E.g. Microstrip
Most probable E loss 72keV 10 proby of 100keV
? pulls track up by 87 ?m
48
Signal formation

• Signal due to the motion of charge carriers
  inside the detector volume the carriers
  crossing the electrode
• Displacement current due to change in
  electrostatics (c.f. Maxwells equations)

• Material polarised due to charge introduction
• Induced current due to motion of the charge
  carriers
• See a signal as soon as carriers move

49
Signal

• Simple diode
• Signal generated equally from movement through
  entire thickness
• Strip/pixel detector
• Almost all signal due to carrier movement near
  the sense electrode (strips/pixels)
• Make sure device is depleted
• under strips/pixels
• If not
• Signal small
• Spread over many strips

50
Lecture 3 Microstrip detector

• Description of device
• Carrier diffusion
• Why is it (sometimes) good
• Charge sharing
• Cap coupling
• Floating strips
• Off line analysis
• Performance in magnetic field
• Details
• AC coupling
• Bias resistors
• Double sides devices

51
What is a microstrip detector?

• p-i-n diode
• Patterned implants as strips
• One or both sides
• Connect readout electronics to strips
• Radiation induced signal on a strip due to
  passage under/close to strip
• Determine position from strip hit info

52
What does it look like?

• P contact on front of n- bulk
• Implants covered with thin thermal oxide (100nm)
• Forms capacitor 10pF/cm
• Al strip on oxide overlapping implant
• Wirebond to amplifier

• Strips surrounded by a continuous p ring
• The guard ring
• Connected to ground
• Shields against surface currents
• Implants DC connected to bias rail
• Use polysilicon resistors MW
• Bias rail DC to ground

53
Resolution

• Delta electrons
• See lecture 2
• Diffusion
• Strip pitch
• Capacitive coupling
• Read all strips
• Floating strips

54
Carrier collection

• Carriers created around track F ? 1mm
• Drift under E-field
• p strips on n- bulk
• p -ve bias
• Holes to p strips, electrons to n back-plane
• Typical bias conditions
• 100V, W300mm E3.3kVcm-1
• Drift velocity e 4.45x106cms-1 h1.6x106cm-1
• Collection time e7ns, h19ns

55
Carrier diffusion

• Diffuse due to conc. gradient dN/dx
• Gaussian
• Diffusion coefficient
• RMS of the distribution
• Since D ? m tcoll ? 1/m
• Width of distribution is the same for e h
• As charge created along a strip
• Superposition of Gaussian distribution

56
Diffusion

• Example for electrons
• tcoll 7ns T20oC
• s 7mm
• Lower bias ? wider distribution
• For given readout pitch
• wider distribution ? more events over gt1 strip
• Find centre of gravity of hits ? better position
  resolution
• Want to fully deplete detector at low bias
• High Resistivity silicon required

57
Resolution as a f(V)
Spatial resolution as a function of bias
Vfd 50V

• Vlt50V
• charge created in undeleted region lost, higher
  noise
• Vgt50V
• reduced drift time and diffusion width less
  charge sharing more single strips

58
Resolution due to detector design

• Strip pitch
• Very dense
• Share charge over many strips
• Reconstruct shape of charge and find centre
• Signal over too many strips ? lost signal (low
  S/N)
• BUT
• FWHM 10mm
• Limited to strip pitch ?20mm
• Signal on 1 or 2 strips

59
Two strip events

• Track between strips
• Find position from signal on 2 strips
• Use centre of gravity or
• Algorithm takes into account shape of charge
  cloud (eta, ?)
• Track mid way Q on both strips
• best accuracy
• Close to one strip
• Small signal on far strip
• Lost in noise

60
Capacitive coupling

• Strip detector is a C/R network
• Cstrip to blackplace 10x Cinterstrip
• Csb Cis ? ignore Csb
• Fraction of charge on B due to track at A

B
A
C
61
Floating strips

• 20mm strip pitch ? s2.2mm

• Large Pitch (60mm)
• Intermediate strip

1/3 tracks on both strips Assume s 2.2mm 2/3 on
single strips s 40/?12 11.5mm Overall s
1/3 x 2.2 2/3 x 11.5 8.4mm
60mm
20mm
20mm
20mm
20mm
Capacitive charge coupling 2/3 tracks on both
strips NO noise losses due to cap coupling 1/3
tracks on single strips s 2/3 x 2.2 1/3 x
20/?12 3.4mm
62
Off line analysis

• Binary readout
• No information on the signal size
• Large pitch and high noise
• Get a signal on one strip only

ltxgt 0
P(x)
-½ pitch ½ pitch
63
Centre of Gravity

• Have signal on each strip
• Assume linear charge sharing between strips

PHL
PHR

• Q on 2 strips x 0 at left strip

P
x

• e.g. PHL 1/3PHR

64
Eta function

• Non linear charge sharing due to Gaussian charge
  cloud shape

PHL
PHR
More signal on RH strip than predicted with
uniform charge cloud shape Non-linear function
to determine track position from relative
pulse heights on strips
P
x
65
Eta function

• Measured tracks as a function of incident
  particle flux

• Measured and predicted particle position

66
Lorentz force

• Force on carriers due to magnetic force
• Perturbation in drift direction
• Charge cloud centre drifts from track position
• Asymmetric charge cloud
• No charge loss is observed
• Can correct for if thickness B-field known

vh
E
H
qL
ve
67
Details

• Modern detectors have integrated capacitors
• Thin 100nm oxide on top of implant
• Metallise over this
• Readout via second layer
• Integrated resistors
• Realise via polysilicon
• Complex
• Punch through biasing
• Not radiation hard
• Back to back diodes depleted region has high R

68
Details

• Double sided detectors
• Both p- and n-side pattern
• Surface charge build up on n-side
• Trapped ve charge in SiO
• Attracts electrons in silicon near surface
• Shorts strips together
• p spray to increase inter-strip resistance

69
Lecture 4 Radiation Damage

• Effects of radiation
• Microscopic
• Macroscopic
• Annealing
• What can we do?
• Detector Design
• Material Engineering
• Cold Operation
• Thin detectors/Electrode Structure 3-D device

70
Effects of Radiation

• Long Term Ionisation Effects
• Trapped charge (holes) in SiO2
• interface states at SiO2 - Si interface
• Cant use CCDs in high radiation environment
• Displacement Damage in the Si bulk
• 4 stage process
• Displacement of Silicon atoms from lattice
• Formation of long lived point defects clusters

71
Displacement Damage

• Incoming particle undergoes collision with
  lattice
• knocks out atom Primary knock on atom
• PKA moves through the lattice
• produces vacancy interstitial pairs (Frenkel
  Pair)
• PKA slows, reduces mean distance between
  collisions
• clusters formed
• Thermal motion ?98 lattice defects anneal
• defect/impurity reactions
• Stable defects influence device properties

72
PKA

• Clusters formed when energy of PKAlt 5keV
• Strong mutual interactions in clusters
• Defects outside of cluster diffuse form
  impurity related defects (VO, VV, VP)
• e ? dont produce clusters

73
Effects of Defects
EC
e
e
e
e
h
h
h
EV
Generation
Recombination
Trapping
Compensation
Effective Doping Density
Leakage Current
Charge Collection
74
Reverse Current

• I ??Volume
• Material independent
• linked to defect clusters
• Annealing material independent
• Scales with NIEL
• Temp dependence

? 3.99 ? 0.03 x 10-17Acm-1 after 80minutes
annealing at 60?C
75
Effective Doping Density

• Donor removal and acceptor generation
• type inversion n ? p
• depletion width grows from n contact
• Increase in full depletion voltage
• V ? Neff

? 0.025cm-1 measured after beneficial anneal
76
Effective Doping Density

• Short-term beneficial annealing
• Long-term reverse annealing
• temperature dependent
• stops below -10?C

77
Signal speed from a detector

• Duration of signal carrier collection time
• Speed ? mobility field
• Speed ? 1/device thickness
• PROBLEMS
• Post irradiation mobility lifetime reduced
• ?? lower ? longer signals and lower Qs
• Thick devices have longer signals

78
Signal with low lifetime material

• Lifetime, ?, packet of charge Q0 decays
• In E field charge drifts
• Time required to drift distance x
• Remaining charge
• Drift length, L ? mt
• mt is a figure of merit.

79
Induced charge

• Parallel plate detector
• In high quality silicon detectors
• ? ? 10ms, ?e 1350cm2V-1s-1, E 104Vcm-1
• ? L ? 104cm (d 10-2cm)
• Amorphous silicon, L ? 10?m (short lifetime, low
  mobility)
• Diamond, L ? 100-200?m (despite high mobility)
• CdZnTe, at 1kVcm-1, L ? 3cm for electrons, 0.1cm
  for holes

80
What can we do?

• Detector Design
• Material Engineering
• Cold Operation
• Electrode Structure 3-D device

81
Detector Design

• n-type readout strips on n-type substrate
• post type inversion ? substrate p type ?
  depletion now from strip side
• high spatial resolution even if not fully
  depleted
• Single Sided
• Polysilicon resistors
• Wlt300?m thick ? limit max depletion V
• Max strip length 12cm ? lower cap. noise

82
Multiguard rings

• Enhance high voltage operation
• Smoothly decrease electric field at detectors
  edge

back plane bias
Poly
strip bias
Guard rings
V
83
Substrate Choice

• Minimise interface states
• Substrate orientation lt100gt not lt111gt
• Lower capacitive load
• Independent of ionising radiation
• lt100gt has less dangling surface bonds

84
Metal Overhang

• Used to avoid breakdown performance deterioration
  after irradiation

2
SiO2
p
(1)
(2)
n
1
n
Breakdown Voltage (V)
4?m
0.6?m
p
Strip Width/Pitch
lt111gt after 4 x 1014 p/cm2
85
Material Engineering

• Do impurities influence characteristics?
• Leakage current independent of impurities
• Neff depends upon O2 and C

86
O2 works for charged hadrons

• Neff unaffected by O2 content for neutrons
• Believed that charge particle irradiation
  produces more isolated V and I

V O ? VO V VO ? V2O V2O ? reverse
annealing High O suppresses V2O formation
87
Charge collection efficiency

• Oxygenated Si enhanced due to lower depletion
  voltage
• CCI 5 at 300V
• after 3x1014 p/cm2

CCE of MICRON ATLAS prototype strip detectors
irradiated with 3 1014 p/cm2
88
ATLAS operation
Damage for ATLAS barrel layer 1
Use lower resistivity Si to increase lifetime in
neutron field Use oxygenated Si to increase
lifetime in charge hadron field
89
Cold Operation

• Know as the Lazarus effect
• Recovery of heavily irradiated silicon detectors
  operated at cryogenic temps
• observed for both diodes and microstrip detectors

90
The Lazarus Effect

• For an undepleted heavily irradiated detector
• Traps are filled ? traps are neutralized
  Neff compensation (confirmed by experiment)
  B. Dezillie et al., IEEE
  Transactions on Nuclear Science, 46 (1999) 221

where
91
Reverse Bias
Measured at 130K - maximum CCE CCE falls with
time to a stable value
92
Cryogenic Results

• CCE recovery at cryogenic temperatures
• CCE is max at T 130 K for all samples
• CCE decreases with time till it reaches a stable
  value
• Reverse Bias operation
• MPV 5000 electrons for 300 mm thick standard
  silicon detectors irradiated with 2?1014 n/cm2
  at 250 V reverse bias and T77 K
• very low noise
• Forward bias is possible at cryogenic
  temperatures
• No time degradation of CCE in operation with
  forward bias or in presence of short wavelength
  light
• same conditions MPV 13000 electrons

93
Electrode Structure

• Increasing fluence
• Reducing carrier lifetime
• Increasing Neff
• Higher bias voltage
• Operation with detector under-depleted
• Reduce electrode separation
• Thinner detector ? Reduced signal/noise ratio
• Close packed electrodes through wafer

94
The 3-D device

• Co-axial detector
• Arrayed together
• Micron scale
• USE Latest MEM techniques
• Pixel device
• Readout each p column
• Strip device
• Connect columns together

95
Operation
-ve
-ve
-ve
ve
ve
-ve
SiO
2

p

h

h
Bulk
n
E
W2D
-
e
-
e

n
W3D
Equal detectors thickness W2DgtgtW3D
ve
E
Carriers drift total thickness of material
Carriers swept horizontally Travers short
distance between electrodes
Proposed by S.Parker, Nucl. Instr. And Meth. A
395 pp. 328-343(1997).
96
Advantages

• If electrodes are close
• Low full depletion bias
• Low collection distances
• Thickness NOT related to collection distance
• No charge spreading
• Fast charge sweep out

97
A 3-D device

• Form an array of holes
• Fill them with poly-silicon
• Add contacts
• Can make pixel or strip devices
• Bias up and collect charge

98
Real spectra

• At 15V
• Plateau in Q collection
• Fully active

Very good energy resolution
99
3-D Vfd in ATLAS

• 3D detector!

100
Summary

• Tackle reverse current
• Cold operation, -20?C
• Substrate orientation
• Multiguard rings
• Overcome limited carrier lifetime and increasing
  effective doping density
• Change material
• Increase carrier lifetime
• Reduce electrode spacing

101
Final Slide

• Why?
• Where?
• How?
• A major type
• A major worry