Hadron Blind Detector for low mass di-leptons in PHENIX Sasha Milov (PHENIX focus on HBD) Jun 20, 2006 Sasha Milov Phenix focus HBD June 20, 2006
Outline Ø Physics: ü Low mass pair measurements with di-electrons. Ø HBD principles of operation: ü Original ideas ü Their implementation ü Detector concept ü R&D Ø HBD progress ü Full scale prototype in PHENIX ü Detector under construction Ø Summary Sasha Milov Phenix focus HBD June 20, 2006 2
What are we looking at? Ø We want to distinguish between two different states of matter: ü Normal nuclear matter where quarks are confined in triplets ü s. QGP where quark interact “freely” with other quarks around Ø OK, we need a probe! ü Probe interacts in a differently way with different media Ø Oops! now! Better ü If probestill need tomedia it’s what is left from the probe! But we leaves the catch up no different as if it never been there. ü We need another probe Sasha Milov Phenix focus HBD June 20, 2006 3
What are we looking at? ØLet’s look at the probe closer. ü We are interested what happens to the daughter particles of our probe Ø Decays shall leave the media. If during their journey the interaction continues we loose the information! Ø So we need a probe, whose decay products go out without interacting with a media Sasha Milov Phenix focus HBD June 20, 2006 4
Why low mass di-electrons? Ø A unique probe to investigate newly discovered matter: ü Particles built with light quarks should have different properties in created media than in vacuum: mass, width, branching ratios. ü Short living particles, decaying before freeze-out. ü Decay products leave media undisturbed by strong interactions. Ø Decays we are interested in: ü ρ (m = 770 Me. V τ ~ 1 fm/c) e+eü ω (m = 782 Me. V τ ~20 fm/c) e+eü φ (m =1020 Me. V τ ~40 fm/c) e+e- Ø Here is the signal! Ø With HUGE combinatorial background Sasha Milov Phenix focus HBD June 20, 2006 5
Why low mass di-electrons? Signal quality we need Ø A unique probe to investigate newly discovered matter: ü Particles built with light quarks should have different properties in created media than in vacuum: mass, width, branching ratios. ü Short living particles, decaying before freeze-out. ü Decay products leave media undisturbed by strong interactions. Ø Decays we are interested in: ü ρ (m = 770 Me. V τ ~ 1 fm/c) e+eü ω (m = 782 Me. V τ ~20 fm/c) e+eü φ (m =1020 Me. V τ ~40 fm/c) e+e- NA 60 (CERN) Lower energy Ø Here is the signal! Ø With HUGE combinatorial background Sasha Milov Phenix focus HBD June 20, 2006 6
What to do with background? Ø Main difference between the signal and background is due to the mass of the primary particle. Ø Compare these two: φ 1020 Me. V π0 135 Me. V Same Momentum Ø Here is a solution: we need to eliminate close e+e- pairs which are due to background and use the rest for the analysis. Sasha Milov Phenix focus HBD June 20, 2006 7
Background sources? Ø In the decays contributing to the background: ü π 0 e + e- γ ü π 0 γ γ e + e- γ Ø Only one electron is detected in PHENIX and another is lost Ø To cut the background we need a new detector such that: ü It sees only electrons ü Located at the origin ü It does not produce its own background (is thin) ü… ü… ü… ~12 m Sasha Milov Phenix focus HBD June 20, 2006 8
rin g “Classic” Cherenkov Detector detector unit window 1 m+ r gas with n~1. 0006 mi rro o k eren Ch rticle a primary p t ligh v Ø Classic RICH (Ring Imaging Cherenkov Counter) has following parts ü gaseous radiator (n ~ 1. 0004 – 1. 0006) ü VUV mirror ü window Ca. F 2 (cheaper) Li. F (better) ü photo-detector (gaseous or PMT) Sasha Milov Phenix focus HBD HADES @ GSI June 20, 2006 9
New RICH for PHENIX? Ø This volume is in the magnetic field Ø If we put mirrors inside, where do we send light to? Ø Let’s get rid of mirrors and put detector right in the beam Ø Possible, but… ü it still must be thin ü it has to detect a single UV photon and be blind to all ionizing particles passing through it!!! Sasha Milov Phenix focus HBD June 20, 2006 10
The original idea Ø The original idea by Y. Giomataris and G. Charpak 1991 (NIMA 310) ov p renk Che part ic zing Ioni HV le hoto n Mesh Ø Electrons from Cherenkov light are produced at the photocathode and amplified full way. Ø Electrons from primary ionization are produced at random and are amplified much less (exponential law). Photocathode Sasha Milov Phenix focus HBD June 20, 2006 11
Original idea modified Ø Problem: such setup requires a window. ü Radiator gas must be transparent. ü The avalanche in the gas contains as many photons as electrons ü What happens if photons shine back on photocathode? Ø We need to separate the radiator and detector volume by a window HV Ø And get even more problems… ü Windows are bulky ü Window is a perfect source of the Cherenkov photons from any particle! Ø Let’s get rid of the window! Sasha Milov Phenix focus HBD June 20, 2006 12
Original idea modified Ø Solution: Pull electron through! Ø Photomultipliers do exactly that. light conversion probability overall: product of two Photocathode electron extraction probability Ø But at a significant cost of efficiency ü Light conversion probability is the best at the upper surface ü Electron extraction probability is the best on the lower surface ü They work against each other resulting in small overall efficiency HV Ø Much better way to get electron on the same side and then pull it through Sasha Milov Phenix focus HBD June 20, 2006 13
Gas Electron Multiplier (GEM) 150μ Ø The original idea by F. Sauli (mid 90 s) US Patent 6, 011, 265 Ø HV creates very strong field such that the avalanche develops inside the holes Ø The same field is sufficient to pull electrons from the surface into holes Ø Photon feedback is not there: GEMs screen the photocathode Sasha Milov Phenix focus HBD June 20, 2006 14
The concept. Ø Take a GEM Ø Put a photocathode on top HV Ø Electron from Cherenkov light goes into the hole and multiplies Ø Use more GEMs for larger signal Ø Pick up the signal on pads Ø What about ionizing particle? Ø Mesh with a reverse bias drifts ionization away from multiplication area Ø Sensitive to UV and blind to traversing ionizing particles Sasha Milov Phenix focus HBD June 20, 2006 15
Some technical details. Ø Photocathode choice: ü There are no many options for solid photocathode. ü Cs. I evaporated onto surface is pretty much the only choice ü As transparent in UV as possible. ü As high refraction index as possible. ü But still usable as a detector working gas. Ø The best possible choice is CF 4 Ø Chemists in the room should throw a flag! Number of photons Ø Gas: (n-1)2/λ 2 λ ü What if moisture gets in the detector? ü H 20 + CF 4 + e- HF + X? Ø Physicists in the room may stay calm… ü Moisture in the detector kills UV transparency and Cs. I much before it kills the detector ü Monitoring gas transparency and humidity on PPM level is required Sasha Milov Phenix focus HBD June 20, 2006 16
Photocathode and gas. Ø Photocathode: ü Cs. I is an obvious choice. ü We are using INFN built evaporator, currently at Stony Brook to do this project. § High area, § High vacuum, § In-situ Q. E. control, § Zero exposure to open air. Ø Gas CF 4 (well known): ü Transparent up to 11. 5 e. V, makes perfect match to Cs. I ü Is a good detector gas. Sasha Milov Phenix focus HBD June 20, 2006 17
Photocathode and gas. Ø Photocathode: ü Cs. I is an obvious choice. ü We are using INFN built evaporator, currently at Stony Brook to do this project. § High area, Ø Gas CF 4 (was not really known): § High vacuum, ü Has high electron extraction § In-situ Q. E. control, probability § Zero exposure to open air. ü Has avalanche self quenching mechanism Ø Gas CF 4 (well known): ü Transparent up to 11. 5 e. V, makes perfect match to Cs. I ü Is a good detector gas. Sasha Milov Phenix focus HBD June 20, 2006 18
Gas purity. Ø Even CF 4 is transparent, oxygen and especially water can absorb UV light ü Water blocks the UV in the region of highest Cs. I efficiency. ü More water in presence of ionization can damage the detector physically. ü Oxygen is a good photo absorber too. ü Desired levels: § H 20 < 10 ppm § 02 < 5 ppm. Sasha Milov Transmission for high purity Ar and C Phenix focus HBD June 20, 2006 19
Gas monitoring system. Input and two outputs measured by switching gas Input and two outputs measured by moving mirror. More expensive but clearly better Sasha Milov Phenix focus HBD June 20, 2006 20
The design. Made of 2 units with R~60 cm, the volume is filled with CF 4 magnetic field is turned off Electrons emit Cherenkov light is registered by 12 photo-detectors in each unit Signal is read out by 94 pads in each unit, pad size ~ size of a circle Accumulating ~36 photoelectrons from each primary electron, while most other operational RICHes have ~15 or less. High statistics allows to separate 2 close electrons even if their signals overlay! 36 72 Number of photoelectrons Sasha Milov Phenix focus HBD June 20, 2006 21
Event display (simulation). Sasha Milov Phenix focus HBD June 20, 2006 22
Mechanical parts and PCB final design. Quick MC shows no difference with standard cells Sasha Milov Phenix focus HBD Entrance window frames are ready, the window itself to be tight between them June 20, 2006 23
Final detector construction Ø Final HBD under construction. ü Mechanical parts for the cage and GEM on frames are being produced at the Weizmann Institute in Israel. ü First GEMs were shipped from WIS to SUNYSB on Sunday for Cs. I coating. ü Final electronics (Nevis Columbia) is waiting test results. Inner side ~1. 2 m Detector panels glued on a jig Outer (copper) side Sasha Milov Phenix focus HBD June 20, 2006 24
Detector construction Pictures taken this week !!! Sasha Milov Phenix focus HBD June 20, 2006 25
Detector construction Glove box Assembly frame GEM testing box Sasha Milov Phenix focus HBD June 20, 2006 26
GEM combinatorics Ø GEM gain varies across the surface. ü We see ~10% variations gain across the surface of a single GEM. ü Variation of a triple GEM stack follow the product of three single GEMs: G 123 ~ G 1 x. G 2 x. G 3 ü All GEMs were surveyed at the WIS and put on a database. ü We have to find the best combination. Ø Little problem: ü 24 Cs. I coated GEMs and 48 uncoated make 24 x 48! / 2! (48 -2)! ~27, 000 combination ü Choosing 24 best pairs out of 27, 000! / 24! (27, 000 -47)!. . . ü Forget it! ü But one can still do better Sasha Milov Phenix focus HBD June 20, 2006 27
The full scale prototype. ADC time slices w. r. t. to the beam-beam trigger Sasha Milov Phenix focus HBD June 20, 2006 28
The full scale prototype. Gas gain changes with time the detector is under HV and may be something else Sasha Milov Phenix focus HBD June 20, 2006 29
Summary: Ø PHENIX detector at RHIC can investigate a new class of probes to study newly invented s. QGP. Ø It requires to built a revolutionary detector based on novel ideas and techniques. Figure of merit for such detector is 6 times more than any detector built so far. Ø The concept of a new detector has been developed over last several years backed up by extensive R&D and simulations. Ø The final detector is now in construction. The prototype installed in PHENIX right now and is taking data. Ø Full HBD will be installed during next physics run. Sasha Milov Phenix focus HBD June 20, 2006 30
Highlights: Ø Cs. I quantum efficiency and CF 4 transparency make a perfect match. Ø Amplification field is sufficient to pull all electrons from GEM surface into holes. Ø GEM acts as a semitransparent photocathode providing high Q. E. Ø GEM geometry eliminates photon feed back from amplification region onto the photocathode. Ø Self quenching mechanism works for GEMs operating in pure CF 4 Ø High electron extraction efficiency is measured into pure CF 4 Ø Given optical purity is observed CF 4 chemical activity is not an issue Sasha Milov Phenix focus HBD June 20, 2006 31
BACKUPS Sasha Milov Phenix focus HBD June 20, 2006 32
Скачать презентацию Hadron Blind Detector for low mass di-leptons in
1097c0cb6b9b66fa0eedc0a6925d9168.ppt