D0 Run2b Upgrade Plans

1
Schematic Design of a 3D Silicon Tracker
Richard Partridge Brown University
Victoria ALCWG Meeting July 28 - 31, 2004
2
Why 3D?

• In principle, a pixel vertex detector and a 2D
  outer tracker can do much of the LC physics
  program
• Vertex detector does 3D track finding, outer
  tracker pins down momentum
• Steve Wagners pattern recognition studies show
  efficient tracking is possible for short-lived
  particles
• It may be possible to adequately track long-lived
  particles by combining outer tracker and
  calorimeter measurements
• Nevertheless, there are strong reasons for a 3D
  outer tracker with stand-alone tracking
  capabilities
• Provides redundancy/robustness in the event
  backgrounds are worse than anticipated or vertex
  detector doesnt meet performance goals
• Almost certainly needed in forward direction
• Robust measurement of long-lived particles (KS,
  L, new physics?)
• Z-coordinate can help resolve ambiguities in
  track-finding
• Outer tracker can be operated/commissioned
  independent of vertex detector

3
Proposed Design Requirements

• Provide robust 3D tracking capability that
  matches calorimeter coverage for current Silicon
  Detector
• qmin 110 mr
• Stay within current Silicon Detector boundaries
• r 126 cm
• z 165 cm
• Provide sufficient tracking layers for
  independent pattern recognition
• Can form a helix from every combination of 3 hits
• Add two layers to suppress fake tracks (
  occupancy2 suppression)
• Add an additional layer to allow a missing hit on
  the track (non-working detectors, inefficiency,
  cracks, poorly measured clusters, etc.)
• Total layers required 6

4
Tracker Layout

• Barrel Disk geometry
• 6 Barrel layers have axial and stereo
  measurements
• 6 Disk layers on each end have two stereo
  measurements (x-u, x-v, or u-v)
• Low-angle tracking is one of the challenges for
  this detector
• Large backgrounds close to beams due to pair and
  two-photon backgrounds
• Backgrounds fall off rapidly with increasing
  radius
• Solution is to push disks out to moderately large
  z to reach low angles without entering
  high-background region
• Schematic Design
• Many choices to be made in developing the tracker
  layout/design
• Goal is to show a specific layout and some
  performance measures
• Have chosen to keep layout simple/uniform rather
  than trying to make detailed optimizations
• Much simulation and mechanical design work needed
  to develop a real tracker design

5
Tracker Layout II

• Occupies volume 12 lt r lt 126 cm, z lt 165 cm

6
Tracker Geometry

• Barrel
• Disk

7
Forward Disk Layout

• All strips oriented in a single direction
• Disk measures either x, u, or v coordinate of
  track
• Disk layer has two measuring planes to make 3D
  hits (x-u, x-v, or u-v)
• Hexagon honeycomb possesses a 6-fold rotational
  and 6-fold reflection symmetries
• Strip direction breaks symmetry

8
Why Hexagons?

• Hexagon symmetry leads naturally to x, u, and v
  coordinate measurements
• Hexagons are one of the few shapes that can be
  used to tile a detector
• Only 1-2 different sensor types are needed for
  entire disk
• Various options for small-angle region may or
  may not want half-hexagons nearest beampipe
• Hexagon utilizes 83 of silicon area
• Square detectors would require 30 more wafers to
  cover the same area
• Wedge, rectangular detectors are even less
  efficient in silicon usage
• Can be subdivided into 60 wedges to allow
  construction of smaller sub-assemblies
• 3 types of wedges needed, differing in the strip
  orientation

9
Hexagon Sensor

• 120 mm width face-to-face
• Fits in 139.5 mm maximum fiducial diameter for
  Hamamatsu 6 wafers
• Strips parallel to face
• Strip length varies from 68 - 136 mm
• 2048 readout strips with 58.5 mm pitch
• Allows for 1 mm guard ring around outside edge
• Each hexagon read out independently
• Use similar readout as is proposed for Silicon
  Detector calorimeter
• 2nd metalization layer to route sensor signals to
  readout chip
• Bump bond readout chip to sensor
• 2048 channels in a single readout chip (or 2
  chips with 1024 channels)

10
Sensor Layout

• Strips run parallel to one pair of edges
• Every 20th strip shown below

11
Bend Coordinate Resolution

• Assume 7 mm hit resolution

12
Radial Coordinate Resolution
13
Barrel Layout
14
Barrel Sensors

• Rectangular sensors 80mm long by 114 mm wide
• Fits in 139.5 mm maximum fiducial diameter for
  Hamamatsu 6 wafers
• 2048 strips per sensor, 54.7 mm strip pitch
• Allows for 1 mm guard ring around outside edge
• Incorporate intermediate strips to increase
  charge sharing?
• 20 axial sensors and 20 stereo sensors per phi
  segment
• Get small angle stereo by rotating sensor
• Gang pairs of sensors together for readout to
  reduce material, power dissipation
• Effective strip length 16 cm

15
Tracker pT Resolution

• 90 Tracks, 7 mm hit resolution in strips, 5 mm
  in pixels
• 2 RL per layer in strips, 0.12 in pixels

16
Impact Parameter Resolution
17
Error Ellipse Example

• Pixel and outer trackers give complementary
  information

18
Wafer Count

• 5040 axial wafers in barrel ladders
• 5040 stereo wafers in barrel ladders
• 8784 wafers in disks
• 18864 wafers total
• Automated fabrication will be critical!

19
Summary

• Many advantages to having a 3D outer tracker
• Reduced risk, track long-lived particles, more
  pattern recognition info
• Schematic design of a 3D tracker presented
• Hexagonal sensors seem like the natural geometry
  for the forward region
• Central region uses conventional rectangular
  sensors
• Excellent momentum and impact parameter
  resolution
• Requires lots of silicon (but silicon is in our
  name)