However, they are very fragile and could be easily damaged by sparks. A photo of MSGS damaged by sparks (L. Ropelewski, CERN detector seminar) ...
New micropattern sparkprotected gaseous detectors and their applications to High Energy Physics and Medicine
Currently a revolution is taking place in the development of gaseous detectors of photons and particles. Parallel platetype and wire-type detectors which dominated for years in high energy and space flight experiments are now being replaced by recent invented micropattern gaseous detectors.
Micropattern - means that the detectors and their electrodes are manufactured by a microelectronics technology and the gap between the anode and cathode electrode networks is very small (below mm)
The main designs of micropattern gaseous detectors
25-100μm
25-100μm Microstrip gaseous detectors
Microdot gaseous detectors
25μm
140μm
MICROMEGAS
GEM
Advantages of micropattern gaseous detectors: ● Lithographic technology is applied for their manufacturing (the same as for solid –state detectors). This offers simplicity and low cost ● High segmentation/granularity and thus potentials for high position resolution (granularity is comparable to solid –state detectors). This opens new avenues in applications such as medical, industrial and many others
However, they are very fragile and could be easily damaged by sparks
A photo of MSGS damaged by sparks (L. Ropelewski, CERN detector seminar)
Physics of discharges A simplified picture
Example of a traditional gaseous detector - a single wire counter hv -V
Q=An0 Amplifier
n0
Field lines
Anode wire
Townsend avalanche
Gas at 1atm
Typically A~105-106
Discharges in thin wire detectors
Primary avalanche
VUV photons
Secondary photoelectrons
Secondary avalanches
Geiger mode in quenched gases
Aγ=1 (Aγph=1 or Aγ+=1) cfront=106-107 cm/s
Geiger discharge is not damaging. One can observed signals~1V directly on 1MΩ input of the scope (no amplifier is needed)
Discharges in thick wire detectors
-V
Avalanche
Anode wire (grounded via amplifier)
Cylindrical cathode
Self-quenched streamer
Streamer
Streamers cannot propagate to the cathode because the electric field drops as 1/r
Transition to streamer occurs when An0≥Qmax=108electros
Strimers give huge amplitudes but the are not harmful as well
Signal’s amplitude in proportional and streamer modes
P. Fonte et al., INFN Insrum. Bull, SLAC-Journal ICFA-15-1, 1997
Discharges in parallel-plate geometry
An0 ≥Qmax=108 electrons - so called Raether limit. P. Fonte et al., INFN Insrum. Bull, SLAC-Journal ICFA-15-1, 1997
Streamer to spark transition in parallel-plate geometry: in the geometry with parallel field lines streamer propagated to the cathode and cause the spark -V a)
Streamer
-V b)
Spark
Kaput!
This discharge is harmful: all electric energy stored in the detector capacitance is released in the spark
There are limited number of ways to restrict the the discharge energy One of them is the use resistive electrodes instead of metallic
-V
Summary:
Aγ=1(Aγ+=1 or Aγph=1)-slow mechanism An0=108 electrons-fast mechanism
What type of breakdown dominates in micropattern gaseous detectors? Studies performed by P. Fonte et al., show that in most cases it is a fast breakdown (see: IEEE Trns Nucl Sci 45,1998,244 and IEEE 46,1999,321)
If it is a fast mechanism, will be Raether limit valid for micropattern gaseous detectors as well? Actually the answer was unknown. Note that Raether limit was established empirically by Raether for parallel-plate avalanche detectors with large amplification gaps (>a few mm) and for a few specific gas mixtures only.
What was well known that almost any type of micropattern gaseous detectors has maximum achievable gain of ~104(with 55Fe ) and all attempts to overcome this limit failed Only in ~1998 it was established that there is a charge limit in avalanche (similar to Raether limit), see: Y. Ivanchenkov et al., NIM A422,1999,300. If the total charge in avalanche overcome this limit An0>qmax breakdowns occur. It was an important new result
Why there are sparks in micropattern gaseous detectors?
Regions with parallel fields lines where any streamer, if appear, is unquenched and may reach the cathode
Because there are regions with parallel field lines, so streamers develop there by the same mechanism as in PPAC
“Raether” limit for micropattern gaseous detectors
Thus, one can claim that the “Raether” limit for micropattern gaseous detectors is An0=qmax~106-107electrons V. Peskov et al., IEEE Nucl. Sci. 48,2001,1070
Coming from the fact that in micropattern gaseous detectors qmax= An0=106-107 electrons one can explain why the max. achievable gain practically in all micropattern gaseous detectors is 104 ( for 55 Fe) and why it drops on a few orders in the case of alphas.
Fonte et al., NIMA419,1998,405 Y. Ivanchenkov et a.l, NIM A422, 1999,300
Some details
V. Peskov et al., IEEE Nucl. Sci. 48,2001,1070
For n0>50 electrons “Rather” limit works well, however for n0 106) were achieved. The resistive coating in turn protected the device and amplifiers from the occasional discharges.
Gain measurements 1,00E+07 1,00E+06
G a in
1,00E+05 1,00E+04 1,00E+03 1,00E+02 1,00E+01 1,00E+00 0
200
400
600
800
1000
Voltage (V)
Blue –PPAC combined with semitransparent CsI photocathode; Rose: “hybrid RPC” combined with Semitransparent CsI photocathode
Gas mixture: He+1%CH4+EF
1200
Rate characteristics:
Thus, RETGEM is suitable for rather low rate applications Preprints: physics/0602114 and LIP-2006-01)
Robust thick GEM In collaboration with CERN ICARUS group
Simplified lithographic prototypes of RETGEMS (5x5cm2) with CrO or CuO resistive coatings manufactured at CERN (R. De Oliveira Workshop) .
Design feature: coating around the holes was removed
Larger size (10x10cm2)prototypes
aQ T aQ T
Design feature: coating around the holes was removed
These oxides coatings, due to the small thickness, do not suppress sparks as strong as graphite coating but never the less the RETGEM is much more robust than bare TGEM
1. Why a thin layer does not fully protects against the sparks?
Dielectric
c~ε/d Metal In principle what matters is the amount of surface charge from the incoming avalanche that is needed to substantially modify the field in the gap, quenching the discharge. In a nutshell: the capacity per unit area. This will be high if the layer is thin or if the dielectric constant is high.
2. Why RETGEM is better protected against sparks than GEM?
GEM
RETGEM
E=V2C/2 For GEM V=500V, for RETGEM 1500V, the energy will be les 9/30~1/3. Suface charge~VC, in RETGEM it is~10 times less
Comparison of gains of GEM and RETGEM
O v e r a l g a in
Double-step operation of RETGEM –gains close to 106; thus single photoelectrons could be easily detected without reaching the breakdowns
1.00E+06 1.00E+05 1.00E+04 1.00E+03 1.00E+02 1.00E+01 1.00E+00 2800
Ar
Ar+20%CH4 Ar+10%CH4
3300
3800
4300
Total voltage (V)
4800
Gain of double RETGEM vs. a total voltage
Gain of triple GEM vs. a total voltage
O verla gain
Ar+20%CH4 1.00E+05
Ne
1.00E+03 1.00E+02
Ar
1.00E+04
1.00E+04 O v e r a l g a in
1.00E+06
1.00E+01
1.00E+03 1.00E+02 500
Ar+10%CH4
Ne
1.00E+00 0
1500
2500
Ar
3500
4500
Overal voltage (V)
Quantum efficiency measured with respect to a Hamamatsu calibrated photodiode: 23% at 187 nm
1000
2000
3000
Overal voltage
Quantum efficiency measured with respect to a Hamamatsu calibrated photodiode: 13.3% at 187 nm
We are considering a possibility to use windowless RETGEM for RICH as was earlier suggested for GEM detectors see (Z. Fraenkel et al., NIM., A546, 2005, 466 )
Another designs of micropattern detectors with resistive electrodes In collaboration with CERN ICARUS group
Hybrid detector: CuO coating combined with true resistive electrodes for ensuring the streamer mode of operation •
•
Hybrid RPC concept: – Resistive layer “quenches” the electron avalanche – Vetronite holes “limit” the photon propagation and after pulses Disadvantages – Choice of resistive material critically depending on rate and gain (resistive materials from Quadrant Technology, ranging from 105 to 1015 Ω-cm, under investigation)
Preliminary results: gains > 104 easily reached in a single step before the transition to a streamer mode
Resistive (oxided) electrode Vetronite
HV
Resistive plates Signal pickup
+++++++++++++++ ____________
Before
A charged particle entering the hole induces an avalanche, which develops into a spark. The discharge is quenched when all of the locally (~1 hole) available charge is consumed.
++++++ _____
++++++ _____
The discharged area recharges slowly through the high-resistivity plates.
After
Array (5x5cm2) of needle-CuO electrodes-operating in gas volume self-quenched streamer mode Coupling of a TGEM with a • needle array and oxided-Cu (or resistive) layer
Z, mm
•
Advantages: – much longer discharge path along hole walls – Streamers selfquenched by geometry – Resistive coating additionally ensures spark protection
Top electrode (CuO)
Streamer
Preliminary results: gains >> 104 easily reached before transition to the selfquenched streamers Poor resolution (~50%FWHM)
needle
Vetronite X, mm
•
Disadvantages: – Critical adjustment of needle height and shape: affecting gain uniformity
Applications
Devices in clinics at present: scintillators array combined with a film
Collimator grid
Film
Advantages: Low cost High enough position resolution Disadvantage: This is an “integrating”detector
X-ray source
Object
Scintillators
In last decade there has been a fast development of X-ray imaging techniques as a result of which traditional films began to be replaced by electronic devices. The main efforts were concentrated on the development of 2D pixelized imaging systems. However, for some applications, for example airport security devices or radiology, 1D scanning systems were developed as well. X-ray source
X-ray source
Slit collimator Object
Object
2-D pixelized detectors
I-D detector
Direction of scan
Scanning systems •
•
•
Scanning systems allow a very efficient rejection of the scattered radiation, improving contrast and the signal to noise ratio. The other important feature of the scanning systems is that they have much simpler electronics and as a result a possibility of simultaneous photon counting and measuring of their energy. The cost of the scanning system is usually much lower than for the 2-D imagers.
•
•
For these reasons they are very attractive for mammographic applications. Nowadays advanced mammographic scanning systems which are under tests in Laboratory and some of them even in clinics,are either solid state (Si,GaAs) or high pressure gaseous detectors. Each of these detectors has advantages and disadvantages, for example high pressure devices are bulky and finally much more expensive than gaseous detectors operating at 1atm.
A modified version of parallelplate micropattern detector was developed for X-ray imaging T. Francke et al., NIM A471,2001,85
Mammographic phantom
Low contrast object # 16 was possible to resolve
The main advantage: the dose was 5-10 times less
Comparison to solid –state mammographic scanners: 1.Lower price (thus large surfaces are possible) 2.Better signal to noise ratio due to the avalanche multiplication 3.Beter rejection of scattered/Compton photons
Tracking and high rate timing RPC
A possible application of low resistivity RPC for tracking
T.Francke et al., NIM, A508, 2003, 83
High rate timing RPC
Antistatic plastics:108-1012Ωcm
L. Lopes et al., NIM A533,2004,69
Measurements
L. Lopes et al., NIM A533,2004,69
Key results
L. Lopes et al., NIM A533,2004,69
UV visualization
Hyperspectroscopy
Environmental applications: results in IR and visible spectra obtained with MCPs
•The Reagent Research Center in Moscow developed and used this method for the analysis of earth surface. • For example, it is used now for Gasprom Inc. tasks: observation of oil pipes and spills from planes and helicopters
Necessity for UV
Object
Light source
Lens
Hyperspectrometer
Slit
MgF2
UV light
“Mesh” -type coating of the outer surface
v Spacers CsI coating Groovs
Readout strips or pads
Pestov glass ASIC readout
J.-M. Bidault et al., Presented at the Imaging-2006 Conference
Example of the object used
2500
2500
2000
2000
1500
1500
Counts
1000
c)
500
1000 500
81
76
71
66
61
56
51
46
41
36
31
26
21
16
1
79
73
67
61
55
49
43
37
31
25
19
7
13
1
6
0
0
11
a)
Counts
1D hyperspectroscopic images
Channel number
Channel num ber
2500
Counts
2000
b)
•
1500
c) A scan across the dark and light sticker (the RPC was turned on 90°).
1000 500
73
67
61
55
49
43
37
31
25
19
13
7
1
0 Channel num ber
a) "dark " yellow sticker, b) " light " yellow sticker
Note: with a solar blind PMs it was impossible to detect this radiation due to the long wavelength background
Flame visualization in day light conditions
Fire detection with image intensifier combined with a narrow-band UV filer
UV Only Image
Example of images of fire in visible (top on the left figure and bottom on the right figure a) and in the UV region (vise versa). One can see that in the latest case the signal to noise ratio is orders of magnitude higher See: http://www.daycor.com and G. F. Karabadjak et al., Preprint AIAA-2000-0105, 2000
Industrial applications: visualisation of corona discharges and sparks at day-light conditions •
Example: a UV image of the HV corona discharge obtained with image intensifies combined with a narrow band filters. During the day time this corona is practically undetectable in the visible region of spectra
In spite of impressive UV images which one can obtain with image intensifiers combined with narrow-band UV filters these detectors have several drawbacks, for example: small size of the MCPs used, high noise(bialkali photocathodes), bad transmission of narrow band filters for UV ( typically only 10%) See: http://www.daycor.com
Images of flames fully illuminated area 900
Countin rate (Hz)
800 700 600 500 400 300 200 100 0 1
9
17
25
33
41
49
57 65
73
81
89
97 105 113
Channel number
Digital images of candle placed 15 away from the detector It is 100-1000 times more sensitive than commercial UV flame detectors ( 100 times in direct sunlight and 1000 times in fully illuminated rooms) ●Very efficient detector for spark detection, ● Low cost
Comparison of our detector with commercially available
Comparison with UV image itensifiers •
•
Measurements performed in Reagen Research Center demonstrated that sensitivity of the UV image intensifier and a gaseous detector with CsI photocathode are almost the same: both detectors were able to detect a cigarette lighter on a distance of 30m. Thus in some applications photosensitive RPCs could be cheap and simple alternative to the UV image intensifiers.
Even better sensitivity one can expect in the case of gaseous detectors combined with CsTe photocathodes:
One can see that even CsI with the highest possible QE (rose) poorly overlaps the fire with emission spectra (red), whereas CsTe (blue)overlaps quiet well. In this figure a green curve is the sun spectra at the ground level
Visible pc: SbCs coated with a CsI protective layer (I. Rodionov et al., presented at the Im.-2006 Conf.)
First tests with CrO covered by SbCs photocathode (preliminary)
10
70
SbCs
1 QE (%)
0
100
200
300
400
500
600
0.1
SbCs/CsI
C u rre n t ( m ic ro A )
60 50 40 30 20 10 0 0
0.01 Wavelength (nm)
QE measured one hour later after the photocathode evaporation (brown curve) Green curve QE for SbCs/CsI photocathode
1000
2000
3000
4000
5000
Time (min)
Photocurrent changes with time (SbCs photocathode)
Noble liquid TPC
The principle of the LAr/Xe TPC •
Gamma
Neutron
-V
Photodetector
Photpodetector
LXe Track VUV
Avalanches
VUV Track
Position-sensitive charge detector
The detection of primary scintillation light in combination with the charge or secondary scintillation signals is a powerful technique in determining the events “t=0” as well as particle/photon separation in large mass TPC detectors filled with noble gases and/or condensed noble gases.
Long longitudinal muon track crossing the cathode plane
1.5 m
Right Chamber
18 m
1.5 m
Left Chamber
Cathode
Track Length = 18.2 m
Top View
dE/dx = 2.1 MeV/cm
3D View
3-D 3-Dreconstruction reconstructionofofthe thelong longtrack track dE/dx dE/dxdistribution distributionalong alongthe thetrack track
A. Rubbia detector
ZEPLIN II Æ ZEPLIN IV 30 kg Æ 1000 kg
The Thelatest latestdesign designas asat at DM2002 DM2002
The idea is to replace PMs by RETGEMs
Chamber design for study of operation of RETGEMs at cryogenic temperatures (CERN, ICARUS group) Gas scintillation chamber
Ar gas
241Am
Drift mesh
55Fe
Ar gas
RETGEMs
LAr 78-300 K
Temperature controlled cryostat
CsI
RETGEM coated with an CsI photocathode was installed inside the LAr test chamber (CERN, ICARUS group)
Temperature controlled cryogenic vessel
RETGEM coated with an CsI photocathode installed inside the LAr test chamber
Gains of single and double steps RETGEM operating at cryogenic temperatures RETGEM (1mm thick) with CsI pc placed above LAr
RETGEM (1mm thick) with CsI pc in Ar
100K
1.00E+04
300K
1.00E+02
Lar
1.00E+00 1000
1500
2000
2500
Overall voltage (V)
O v e ra ll g a in
Overall gain
1.00E+06
3000
1.00E+06 LAr
300K
1.00E+04 1.00E+02 1.00E+00 1000
1500
2000
2500
Overall voltage (V)
3000
“50 liters” ICARUS test chamber PM
Light amplification A
Ar gas
Alphas
LAr LAr
Dewar
With this device we could study the operation of CsI photocathode (monitor its QE and stability) in gas in the temperature interval 100300K as well as in LAr
Could resistive electrode approach be applied to other micropattern detectors ?
Recently appeared on the market Bruker x-ray detector (Pos. resol. 120 μm, rate 5105Hz/mm2)
1mm
D. Khazins et al., IEEE Nucl. Sci 51,2004,943
Micromegas “now it is spark protected!”-Y. Giomataris, private communication
Conclusions: ●We have developed and extensively tested several types of micropattern gaseous detectors with resistive electrodes These detectors are spark –protected and at the same time have good rate characteristics ● Some of them were already successfully used in several applications ● We believe that next generation of micropattern detectors will be based on resistive electrode approach
Collaborators T. Francke and XCounter team, Stockholm C. Icobeaeus, M. Danielsson and KTH –Stockholm team P.Fonte, LIP, Coimbra, Portugal J.-M.Bidault, O. Zanet, P.Gally, Pole University, Paris P. Pietrpoalo, L.Periale, P. Pichhi, CERN- CARUS Di Mauro, P. Martinengo, E. Nappi, ALICE-HMPD group Ecole des Mins team, St. Etienne I. Rodionov, Reagent, Moscow R. De Oliveira, CERN Workshop
