1. TRDs for the third Millennium
TRD based on the usage of thin scintillators V.V.Berdnikov1, B.A.Dolgoshein1 , V.A.Kantserov1, A.P.Shmeleva2, V.V.Sosnovtsev3, V.O.Tikhomirov2, B.I.Zadneprovski3 1 National Research Nuclear University “MEPhI”, Moscow 2 P.N.Lebedev Physical Institute, Russian Academy of Science (LPI) 3 Central Research and Development Institute of Chemistry and Mechanics, Moscow (CRDICM)
TRDs for the third Millennium
4th Workshop on Advanced Transition Radiation Detectors for Accelerator and Space Applications.
Bari, Italy, September 14-16, 2011

2. Outline Motivation Possible approaches and constrains
Thin scintillator films
Experimental studies
Monte Carlo simulations
Summary and outlook
V.Tikhomirov. TRD based on thin scintillators. Workshop "TRDs for the third Millennium", September 2011, Bari, Italy

3. Motivation
Detection of TR in traditional TRDs is carried out by gaseous detectors filled by high-Z noble gas – usually Xe or Ar.
The usage of gas represents a serious obstacle e.g. in space apparatus – due to complexity of gas system, ageing problems etc.
The idea to use solid state detectors (e.g. scintillators) in TRDs seems very attractive.
V.Tikhomirov. TRD based on thin scintillators. Workshop "TRDs for the third Millennium", September 2011, Bari, Italy

4. Possible approaches and constrains
But: together with TR we also detect (parasitic) dE/dx ionization from primary particle. This circumstance leads to the presence of some optimum in detector thickness for better separation between TR-producing and TR-sterile identified particles. In practical cases, the optimal thickness of gaseous detectors is ~1 cm. Taking into account the density of gas and solid state, we can expect the optimal thickness of scintillator layer for TRD to be ~10 µm.
Two global problems are arose:
1. How to produce such a thin scintillator?
2. How to collect the light?
V.Tikhomirov. TRD based on thin scintillators. Workshop "TRDs for the third Millennium", September 2011, Bari, Italy

5. Light collection by vacuum PMT
Advantages:
Presumably rather high light collection efficiency
Not great number of channels due to large area of PMT window
Disadvantages:
Space consuming due to large size of PMTs along particle direction
Very large amount of material: multi-section TRD becomes an destructive detector
Presumably rather limited number of possible applications for such configuration
V.Tikhomirov. TRD based on thin scintillators. Workshop "TRDs for the third Millennium", September 2011, Bari, Italy

6. Light collection by WLS fibers
Advantages:
Compact
Acceptable material budget
Coordinate measurements are possible
No HV
Disadvantages:
Big number of channels – rather expensive
Main question: is effective light collection possible?
V.Tikhomirov. TRD based on thin scintillators. Workshop "TRDs for the third Millennium", September 2011, Bari, Italy

7. Light yield and collection factors
Detected number of photoelectrons on PMT (or number of SiPM cells): Nphe = E × R × ε1 × ε2 × ε3 where Nphe – number of photoelectrons; Е – energy loss (dE/dx or TR photon) in scintillator media, keV; R – scintillator yield, optical photons/keV; ε1 – efficiency of light collection from scintillator to PMT input window or to WLS; ε2 – efficiency of light reemission in WLS and light transfer to photodetector; ε3 – photodetector efficiency, photoelectrons/optical photons; Or: Nphe=Е×k where k will be named the (total) light collection, photoelectrons/keV
V.Tikhomirov. TRD based on thin scintillators. Workshop "TRDs for the third Millennium", September 2011, Bari, Italy

8. Light yield and collection factors (2)
It is obvious to use scintillators with large Z and high light yield.
Scintillators based on Lutetium are good candidates. For example, LSO (Lu2SiO5:Ce) has effective Z=66, light yield≈30 photons/keV, fast (~40 ns) response and emission spectrum (λmax=440 nm) suitable for detection by SiPMs.
Our preliminary estimation for WLS+SiPM configuration was: R=30 photons/keV for Lu-based scintillators, ε1=0.3, ε2=0.07, ε3=0.3, so k=0.2 photoelectrons/keV
Typical dE/dx loss in 10 µm of LSO ≈9 keV, mean energy of absorbed TR photons ≈15 keV, so we can expect ≈2 photoelectrons for dE/dx and ≈3 photoelectrons for TR in one TRD section.
Statistical fluctuations of such small number of detected photoelectrons are very large and will significantly smear the original distributions of dE/dx and (dE/dx+TR) energy losses for hadrons and electrons, making the rejection power worse.
V.Tikhomirov. TRD based on thin scintillators. Workshop "TRDs for the third Millennium", September 2011, Bari, Italy

9. Scintillator films production
Several samples of thin scintillator films were produced at two Russian Institutes: Central Research and Development Institute of Chemistry and Mechanics (CRDICM) and Institute of Solid State Physics of Russian Academy of Science (ISSP).
Films were produced as a suspension of sub-micron powder of scintillator crystals in optical transparent media.
LuBO3:Ce and Lu2SiO5:Ce with different concentration of Ce were used as scintillating crystals.
Many experimental studies with film samples were carried out both with vacuum PMTs and (WLS fibers + SiPM) configurations.
Films produced by CRDICM show better characteristics (uniformity, transparency, light yield). Later in this presentation only films made by CRDICM are considered.
V.Tikhomirov. TRD based on thin scintillators. Workshop "TRDs for the third Millennium", September 2011, Bari, Italy

10. Scintillator films production (2)
Stage 1 : production of LuBO3 (Lu2SiO5 ) : Ce3+ powder
Initial components: Lu2O3, H3BO3,, HNO3, Ce(NO3)3∙6H2O, NH3OH
Initial components: Lu2O3 Na2SiO3,, HNO3,
Ce(NO3)3∙6H2O, NH3OH
Sol-gel method
Air drying at T= oC
Crystallization at T=950 oC – 4 h
LuBO3 powder, granules size: µm
Lu2SiO5 powder, granules size:0.3=0.8 µm
V.Tikhomirov. TRD based on thin scintillators. Workshop "TRDs for the third Millennium", September 2011, Bari, Italy 10

11. Scintillator films production (3)
Stage 2 : production of scintillator films
Thin Al base + optical epoxy compound + Lu2SiO5:Ce3+ powder, 24 h hardening
Film thickness: µm can be obtaibed; a few cm2 area
Lu2SiO5 concentration: 4-50 mg/cm2 can be obtained
Samples of scintillator films – LuBO3-based (left) and Lu2SiO5-based (right) – on the top of sheet with printed numbers and on the ruler. The photo illustrate the transparency of the produced scintillator films.
Side-view cut off of scintillator film with visible Lu2SiO5 granules inside
V.Tikhomirov. TRD based on thin scintillators. Workshop "TRDs for the third Millennium", September 2011, Bari, Italy

12. Spectra, yield and timing of the films
a) Scintillation spectra for
different concentration of Ce
b) Total scintillation intensity
as a function of Ce concentration
V.Tikhomirov. TRD based on thin scintillators. Workshop "TRDs for the third Millennium", September 2011, Bari, Italy

13. Spectra, yield and timing of the films (2)
c) Signal timing for 1 mol% Ce
d) Signal timing for mol% Ce
V.Tikhomirov. TRD based on thin scintillators. Workshop "TRDs for the third Millennium", September 2011, Bari, Italy

14. Measurements with vacuum PMT
Experimental setup
PMT calibration:
4.5 channels/photoelectron
Achieved yield:
k = 1.1 photoelectron/KeV
55Fe
241Am
V.Tikhomirov. TRD based on thin scintillators. Workshop "TRDs for the third Millennium", September 2011, Bari, Italy

15. Silicon Photomultipliers (SiPMs)
SiPMs are photodetector with matrix of p-n junction cells with area ~50x50 µm2 which are working practically independently in self-quenching Geiger regime.
Each cell has a gain ~106 and can detect a single photon. The output signal from SiPM is proportional to number of detected photons (if number of photons is much less than number of cells).
Rather high efficiency – like vacuum PMTs or better: ~30%.
Fast (~100 ps), low operational voltage (~ V), very compact, non-sensitive to magnetic fields.
R 50 h
pixels
Ubias 50V Al
Depletion
Region
2 m
substrate
Resistor ~1 MOhm
20m
42m
V.Tikhomirov. TRD based on thin scintillators. Workshop "TRDs for the third Millennium", September 2011, Bari, Italy

16. SiPM Hamamatsu S10362-33-05С Main SiPM parameters:
Effective active area: 3x3 mm2
Pixel size: 50x50 µm2
Number of pixels: 3600
Gain: 7.5∙105
Peak sensitivity wavelength: 440 nm
Operating voltage range: 70±10 V
Dark count (25 oC): 6 Mcps
Time resolution (FWHM): ps
V.Tikhomirov. TRD based on thin scintillators. Workshop "TRDs for the third Millennium", September 2011, Bari, Italy

17. WLS fibers Double cladding WLS fibers square in section 1x1 mm2
Produced by Saint-Gobian Crystals
7% of own light collection
9 fibers in row under scintillator film were connected to single 3x3 cm2 SiPM
V.Tikhomirov. TRD based on thin scintillators. Workshop "TRDs for the third Millennium", September 2011, Bari, Italy 17

18. Measurements with WLS and SiPM
Experimental setup
V.Tikhomirov. TRD based on thin scintillators. Workshop "TRDs for the third Millennium", September 2011, Bari, Italy

19. Measurements with WLS and SiPM (2)
Assuming Poisson distribution, the mean number of photoelectrons is estimated to be 1.84 for 16 KeV Am241 source, i.e.
achieved yield: k = 0.12 photoelectron/KeV
V.Tikhomirov. TRD based on thin scintillators. Workshop "TRDs for the third Millennium", September 2011, Bari, Italy

20. Monte Carlo simulations
Initial motivation for MC was a possible usage of Sci-TRD in gamma spectrometer for the future GAMMA-400 space experiment. The TRD should provide additional rejection factor between protons and electrons ~50.
The MC program is based on GEANT3 and additional TR generation module, provided by P.Nevski.
Detailed geometry description, all physical processes (dE/dx, TR absorption, Auger and photo electrons propagation etc) were considered.
Scintillation yield, light collection and photodetector efficiency were taken into account via single factor k mentioned above.
Statistical fluctuations of number of photoelectrons on photodetector were also taken into account.
V.Tikhomirov. TRD based on thin scintillators. Workshop "TRDs for the third Millennium", September 2011, Bari, Italy

21. Monte Carlo simulations (2)
The following effects were not taken into account in MC:
Scintillator granules difference in shape and size (each granule is presented as an ideal sphere 1 µm in diameter)
Possible local non-uniformity of granules distribution inside optical compound (random, but uniform distribution is assumed in MC)
Digitization & noise (rejection power is calculated via the number of detected photoelectrons)
Parameters for optimization: number of TRD sections at fixed total length 50 cm; radiator parameters; scintillator granules size; total thickness of scintillator granules etc.
Two methods of signal processing are presented here:
a) Sum of amplitudes from all TRD sections
b) “Cluster counting” method: based on number of TRD sections where signal exceeds some predefined threshold: cheaper solution – requires a discriminator only in each of electronic channel.
V.Tikhomirov. TRD based on thin scintillators. Workshop "TRDs for the third Millennium", September 2011, Bari, Italy

22. Detector parameters optimization
Optimal total thickness of scintillator granules: 5-15 µm
Optimal size of granules: 1-2 µm
Optimal radiator parameters: µm polyethylene foils with µm spacing
Rejection factor ~100 can be achieved
(24 TRD sections and light collection factor k=0.2 is assumed here)
V.Tikhomirov. TRD based on thin scintillators. Workshop "TRDs for the third Millennium", September 2011, Bari, Italy

23. Detector parameters optimization (2)
Light collection
efficiency, obtained
so far k=0.12 phe/keV
Optimal number of TRD sections (at fixed 50 cm length) seems to be around 12.
Rejection power crucially depends on light collection efficiency: low light collection leads
to pure rejection due to mentioned statistical fluctuations in number of photoelectrons
V.Tikhomirov. TRD based on thin scintillators. Workshop "TRDs for the third Millennium", September 2011, Bari, Italy

24. Summary and outlook
TRD detector based on usage of thin scintillator films (Sci-TRD) is proposed.
Several samples of scintillator films were produced and tested.
Clear signals from gamma sources were detected by using vacuum PMT and (WLS fibers + SiPM) link.
Corresponding Monte Carlo simulations of proposed Sci-TRD have been done. By optimistic scenario, electron/hadron rejection factor ~100 can be achieved.
The efficiency of light collection remains the main factor which restricts the expected rejection power of the Sci-TRD.
Work to produce new scintillator samples and find new approaches in light collection scheme to be continued.
V.Tikhomirov. TRD based on thin scintillators. Workshop "TRDs for the third Millennium", September 2011, Bari, Italy

25. Backup slides
LSO
V.Tikhomirov. TRD based on thin scintillators. Workshop "TRDs for the third Millennium", September 2011, Bari, Italy 25

26. Backup slides
V.Tikhomirov. TRD based on thin scintillators. Workshop "TRDs for the third Millennium", September 2011, Bari, Italy 26

27. Backup slides
V.Tikhomirov. TRD based on thin scintillators. Workshop "TRDs for the third Millennium", September 2011, Bari, Italy 27

28. Backup slides
V.Tikhomirov. TRD based on thin scintillators. Workshop "TRDs for the third Millennium", September 2011, Bari, Italy 28