1. GasPixel Transition Radiation Tracker
GasPixel Transition Radiation Tracker Victor Blanco Carballo, Max Chefdeville, Fred Hartjes, Martin Fransen, Serguei Morozov, Serguei Konovalov, Anatoli Romaniouk, Harry Van der Graaf
Motivation
Test beam prototype
MC model
Tracking
Particle ID
Timing
First test beam results
Plans

2. Motivation
Pixelization of the information from the particle track offers new possibilities to combine in one detector many features:
1. Precise coordinate measurements (well below diffusion limit). 2. Vector track reconstruction. 3. Very good multi-track resolution (not covered in this talk). 4. Powerful pattern recognition features (not covered in this talk). 5. dE/dX measurements. 6. Enhanced Transition Radiation separation. 7. Good timing properties
This technique definitely has a wide range of applications but
Inner Trackers of the collider experiments
may have a particular interest.

3. In addition to the particle identification
Transition Radiation Detector normally situated at the outer radii of the Inner Tracker. Because of large leaver arm this area is very much important for momentum measurements -> a precision tracking demand. h
Possible layout
Sensitive layers are interleaved with the TR radiator.
In addition to the particle identification
the Gas Pixel TRT will provide:
1. Precision space points X,Y
2. Vector direction f, h
Very important for the pattern recognition: f
f, h - cone to search for the partners

4. Gas Pixel TRT Test Beam Prototype, What we should expect to get
Everything starts from the first step :
Gas Pixel TRT Test Beam Prototype,
InGrid TimePix detector, 14 x 14 mm2
256 x 256 pixels
L = 18 mm
0.05 mm V0 V1
h or 90o - h
Minimum L for TR absorption is ~ 18 mm
TR-Radiator
Beam E
20o e-
Some Chamber Parameters
Vo~ 520 V
V1= 4500 V,
Edrift ~ 2000 V/cm
Gas mixture 70%Xe+30%CO2
Total drift time ~ 400 ns
Ion signal ~ 80 ns
Transverse diffusion sT ~200 mm/cm1/2
Longitudinal diffusion sL ~120 mm/cm1/2
What we should expect to get
from this prototype?

5. Toy MC simulation: model
Particle tracking including all interactions + geometry - GEANT3
Ionisation losses PAI model
TR generation ATLAS TRT code
Change of the parameters through the configuration files. It allows to set:
The test beam set-up configuration
Chamber parameters
Beam energy and particle sort
Beam parameters (position, spread)
Radiator parameters
Gas mixture composition
Drift velocity and diffusion
Gas gain fluctuation type (Exponential, Polya)
Ion collection time
Ion signal shape
Electronics response shape
Registration thresholds
Output files :
Pixelman compatible format
No FE electronics response
-Time
-Electrons
With FE electronics response:
-Time mode
-Time over Thr. mode
Colour map numbers correspond to
a hit arrival time- the coreect time would be (400-T)*10 ns

6. Signal parameters used in simulation
Electrons
Ions
80 ns
Electronics current d-function response
(variable parameter)
Pixel single electron current function
Response to one electron
180 ns T1 T2
T1 OR DT=T2-T1
are measured
Threshold (variable parameter)

7. The results of the analysis shown below are very preliminary!
The programme is the same for the MC and the experimental data.
Analysis parameters can be changed
through the configuration files.
Type of files to be analysed.
Events to process.
Parameters for tack reconstruction.
Threshold for TR cluster definition.
Some others.
2 events, pions 20 GeV
Red dots indicate reconstructed track.
Number of electrons per track slice (2 pixels)
Ionisation clusters
Threshold to detect clusters Y X
The results of the analysis shown below are very preliminary! MC
Pixels with the hits above the threshold.

8. Analysis: tracking (MC data source)
pions 20 GeV
Space point in f direction (X)
Space point in Z direction (Y)
s=17 mm
s=18 mm
Difference between real and reconstructed space point across the track.
Difference between real and reconstructed space point along the track.

9. Analysis: tracking (MC data source)
pions 20 GeV
Angle f
Angle h
s =0.65o
s =0.3o
Difference between real and reconstructed angle of the track.
Difference between real and reconstructed angle of the track.

10. Vector track reconstruction.

11. What we can measure with the beam? -> Pseudo track parameters
pions 20 GeV MC MC
s=33 mm
s =0.8o
Difference between reconstructed space points for two pseudo tracks.
Difference between angle of the track for two pseudo tracks.

12. Particle ID: use TR and dE/dX information
TB prototype:
dE/dx total
Number of primary
Collisions vs gamma
Pions
Plato Fremi : Mean 4
Min. ionising: Mean 2.7
Number of primary collisions per cm of Xe
Factor 1.36
ln g MC
Relativistic rise vs
d-electron energy
dE/dx, arb. units
Number of clusters above ~1 keV
Plato Fremi : Mean 2.7
Min. ionising: Mean 1.2
For 2 kev d-electrons
Factor of 2
Detailed knowledge of the ionisation structure of the track gives more information for the particle ID.
B. Dolgoshein, NIM A326 (1993) 434

13. TR generation. Set up as shown above. Number of photos per eventMC MC
Average:
2.5 TR photons.
No TR photons:
125 out of 1000
20 keV
TR spectrum
Number of photos per event

14. TR registration: The main issue is to separate dE/dX from TR
Use the low:
X-ray absorption probabilities.
Something we don’t know yet (see Exp. Data).
1+2 == 3
13% 6% 3% 9% MC
Pions
Electrons
+ TR
Total energy deposition in the chamber.
B. Dolgoshein, NIM A326 (1993) 434

15. TR registration Event 2003 TR clusters Time information
Amplitude information

16. TR registration TR clusters
Event
TR clusters MC
Threshold to detect candidates to the TR clusters
(30 in this units).
Total number of electrons in the area with R=sT is calculated
All the rest is treated as dE/dX

17. TR registration TR absorption in the chamber. MC MC
This distribution is flat
for the d-electron
Number of photons
Absorption position in the chamber, cm.
X-ray energy, keV.
Number of the TR photons absorbed in
the chamber vs depth.

18. Electron/Pion separation
Electrons
Pions MC MC
0-bin =96 out of 1000
No TR 125
0-bin =>753 out of 1000
Some electrons without TR
Produced clusters
Total number of electrons in the clusters for electrons and pions (20 GeV)
Now we have detailed information about the track structure
How to use it?

19. Electron/Pion separation
The best is likelihood method but other algorithms almost as powerful can be used.
Demonstrative exercise (far to be a complete analysis)
Two classes of events:
Events with 0 or 1 cluster.
Events with more then 1 clusters.
0 and 1 cluster events
1 cluster events
408 e
963 pions
224 pi
316 e MC MC
Size of the cluster
Normalised dE/dX
Size of the cluster
Pixel position
Electrons are coloured

20. Electron/Pion separation 2 and more clusters
37 pi (3.7%) e (59%)
Pions are in pink MC MC
Sum of the clusters
Normalised dE/dX
Size of the maximum cluster
Pixel position for the cluster with lowest energy
Very simplified approach gives pion rejection factor >10 at
electron efficiency slightly less than 90% for already one radiator-chamber set
Up to now NO Time information used yet (apart of ToT).

21. Timing. Very important to separate bunch crossings in the collider experiments Strongly affected by electronics shaping time MC
140 ns
25 ns
Electronics
Shaping function
Arrival time for
the first electron
20 ns
100 ns
Timing jitter at the pixel
threshold of 1 electron
Timing jitter at the pixel
threshold of 0.1 electron
140 ns
peaking time
100 ns
50 ns

22. At low threshold total jitter can be reduced below 25 ns.
Timing.
25 ns peaking time MC MC
Timing jitter at the pixel
threshold of 1 electron
Timing jitter at the pixel
threshold of 0.1 electron
25 ns
55 ns
At low threshold total jitter can be reduced below 25 ns.
But there is a question about dynamic range of the ToT measurement!

23. Amplitude measurement saturation. Threshold 0.1 el.
Events 2051 and 2052 measured with electron counting and ToT methods (MC)
Number of electrons MC
Time over Threshold
Peak suppressed
Detailed simulations are required to reach a reasonable balance!

24. First tests with the beam (6 GeV).
Should be considered as a preparatory step for a real tests beam.
Very little time.
Time measurements only (no ToT information).
Some problem with the chamber (E-field distortion).
Tracking (experiment).
Pseudo track reconstruction
Space point
Angle

25. Particle Identification
Pions (left) Electrons (right)
TR clusters
Most of the TRT clusters are out of track!?
TR angle = 1/ NO
Multiple scattering angle = 5/ NO
Photo-electron path UNLIKELY
To be understood
and used
if it is not artefact!
Results of the analysis based on
the cluster presence information only :
Electrons ID efficiency: ~90%
Pions Rejection factor: ~10

26. Brief summary:
First estimate of the Transition Radiation Tracker properties based on the Pixel technology show large potential of this detector which combines precision tracking, particle Identifications and powerful pattern recognition features.
Simplified MC which contains all basic physics processes allows to make detector optimization and realistically estimate its performance (yet for the test beam set up).
Analysis program is the same for MC and tests beam data what allows to get fast feed back and use it as an express online tool.
For the further steps a good quality of the test beam data are required to tune MC and Data Analysis Model and optimize the detector parameters and FE read out.
Test beams: May, October-November 2008
These are the firsts steps on the long way to a real detector!

27. Back-up slides
TR clusters

28. Electron Drift Velocity and Transverse Diffusion.
Gas properties 20%, 30% and 40% CO2
Electron Drift Velocity and Transverse Diffusion.

29. Electron Longitudinal Diffusion.
Gas properties
Electron Longitudinal Diffusion.
Lorentz angle is rather large.
In a real detector its effect must be taken
into account as a part of the calibration procedure.

30. TR registration Event 2004 TR clusters Time information
Amplitude information

31. Particle Identification
ToT is absolutely necessary for particle ID
Calibration is a very important step
X-ray run with ToT shown reasonable energy resolution.
Sample of the events with 6 keV events (Fe55) and spectrum
for all calibration data.
sE/E~13%