A study of gas electron multiplier

VoL 15 No 5NUCLEAR SCIENCE AND TECHNIQUESOctober 2004a study of gas electron multiplierAN Shao-Hui, LI Cheng*, ZHOU Yi, XU Zi-ZonDepartment of Mondern Physics, University of Science and Technology of China, Hefei 230027)Abstract A new kind of gas detector based on gas electron multiplier (GEM) is studied for X-ray imaging of highluminosity. A single-GEM device is designed to test the property of gEM foil. The effective gain and counting capability of a double-Gem detector are measured by an X-ray tube with Cu target. An initial X-ray imaging experimentis carried out using a triple-GEM detector and the position resolution of less than 0. Imm is achieved. The 3D distribution of electrostatic field of GEM mesh is also presentedKeywords Gas electron multiplier, Effective gain, X-ray imagingCLC number TL8131 Introductionchemically in a photolithography process. Etchinboth sides simultaneously leads to the structure ofThe gas electron multiplier(GEM) is a two- double conical holes( 2). Fig. I shows the microscopicide copper-coated Kapton foil and perforated with a view of GEM with double conical holes. The typicalhigh density of holes. It acts as preamplifier for elec- structure of GEM holes is shown in Fig.2trons released by ionizing radiation in gas when suitThe distribution of electrostatic field of gemable potentials are applied. The GEM holes are etched with voltages applied is simulated by the 3Dmaxwellprogram based on the finite element method 3.In thesimulation program, we make the field intensity2k V/cm for the upper gap, 6k v/cm for the lower gapand gEm voltage 500V. The simulation result isyn in Figdrifting along the fieldlines in the drift gap, a gemwill collect nearly all the electrons into its holes. The-100vFig 1 GEM microscopic view200vFig2 Double conical structure of GIS-5um,7=50um,D=70m,d=60um,P=140um)Fig 3 3D electric field lines distribution of GemSupported by the National Natural Science Foundation of China(10075045)中国煤化工Received date: 2004-01-09CNMHGNo 5AN Shao-Hui et al. A study of gas electron multiplierfieldlines are strongly compressed up to larger than gle gEm is defined as M=(b+ls/(enR), where e and5k V/cm(shown in Fig 4) and the resulting high n are the electron charge and the number of ion pairselectric field causes electron avalanche inside gEm per conversion, R is the counting rate of X-ray. Theholeseffective gain is given by M=1s/(enR). The test resultsshow(see Fig. 6) that the real gain of single-GEM canup to 105. The effective gain is about 1/3 of the realwhen the transfer field is 3kV. 4102001501000D50100150200002334Zum40 F Vav-560VFig 4 Field intensity in holes of GEm6(kv.cm"2 Effective gain of single-gemFig 6 Dependence of lb and Is on transfer fieldThe resistance of a manufactured gem foil3 Double-gem detectorsshould be more than 500GQ2. All geMs are tested in anitrogen-filled box before used. A single gem deviceThe Multi-GEM detector is a parallel plate dewith a cathode, a piece of GEM foil and a Print Cir- vice with one or more GEM foils. We design a dou-cuit Board(PCB), is designed to test the quality of ble -Gem detector to acquire a large enough gain inGEM foils and to measure its effective gain. The order to detect the small number of photoelectronsschematic view is shown in Fig. 5. The test beam of Fig. 7 shows the schematic of a double-gem detectorX-rays is generated from an X-ray tube with Cu targetWith different negative voltages applied to the driftcathode, GEM top layer and GEM bottom layer, theCnft ectrode5 mm dnncurrents of Ib and Is are measured by ampere-meter as一GEMshown in Fig. 5, where Ib is the leak current on the2 mm transter 1bottom layer of GEM, Is is the signal current collected-GEM 2on the PCB readout, Edrift is the field intensity of the2 mm transfer 25mm drift gap, and Etransfer is the field intensity of the20 readou2mm transfer gap Because a part of fieldlines end toMets framethe bottom electrode of GEM, the number of electrons Fig.7 Schematic view of a double-GEM detectorcollected on the PCB is reduced. The real gain of sinThe PCB readout with orthogonal strips in x, y dimen-x-rarssions on both sides is used for particle localizationig. 8). The strips pitch on both sides is 0.8mm5mm cn gasThe width of the strips is 0. 35mm on the upper layerHavnand 0. 65mm on the bottom layer to get the same exGEMarm transferstem supplying voltages to the douPCa readoutble-gem detector is similar to Fig. 5. The changes ofsame gas mixture with different GEM voltages areFig 5 Schematic view of single GEM devicemeasureedh中国煤化工9keVX- rays ofCNMHG292NUCLEAR SCIENCE AND TECHNIQUESVoL 15V.=450Vv=390vB020ArCo.30V.=350vkro290:10v450v10=1 W. amrtEu·E“4kvcm300320340360380400420440460480500520Vav IVFig 8 PCB readout(thickness of 0.3mm)Fig9 Comparison of effective gains for different gas compo-SFe. The detector can get a larger gain and workitionswith lower voltages in a gas mixture(Ar: CO2=70: 30)Argon: Co27030Eo" kv. amespecially. But the phenomenon of discharge is de4 W. ot一-370v300Vtected occasionally when both VGEMI and VGEM2 are=410V-430Vmore than 480V. The effective gain dependence onthe gap field is also measured. With Edrift and etransferlincreasing, the electron transparency of GEM meshis almost constant, while it changes linearily withEtransfer2(see Fig. 11). From the results, a better working condition is selected: Edrif=l kv. cm", Etransfer l. 24kv cm", and VGEM12-450V. Thus the effective3403060304004204404804305C0gain will be 1.4×10Fig 10 Dependence of effective gains on GEM voltages10fAco-70306[Arco, -7030EE“1kcmE6=410V4}vm=+410Ve4ls1k.am,E-A4kVcm-tVav.410V2123456789102345678910体vcm")Fig11 Effect of gap fields on effective gainAro2=7030vav“va-450VEor" kv.amx-ray court rate=2x10°sThe counting capability is measured using X-rayE4 kv. ctube under the same condition shown in Fig. 12. Thdefined as o=ls[= Mnert/S,where e and n are the electron charge and number oft is the time of irradithe beam intensity R is 5x10Hz and the irradiatedspot S is 0.5mm, and M is the effective gain. Whenthe integrated charge is up to 700uC' mm, thechange of pulse height(ADC) is less than 1%010020030040050060070oFig 12 Stability of the double-GEm detectorH中国煤化工CNMHGNo 5AN Shao-Hui et al. A study of gas electron multiplier2934 X-ray imaging222798a third gem foil is inserted between the Em2and the PCB readout for reducing the discharge probability and getting a stable gain to realize the X-rayImd gE Mh eco GEMer gap) aft bst mp isolWith edrift=lkv cm, Etransferl 23=4kV370V, and VGEM2. 3=450V, the effective gain will be4.5x10 when operated with X-ray of 8kev in thesame gas mixturea three-slit image is shown in Fig 13 using thePenon(mm)centre-of-gravity readout method. The slit perpen- Fig14(a) Position distribution on X-directiondicular to X-axis is 0.5mm. The two slits perpendicu180lar to y-axis are both 0.3mm with a 2mm distancenalanThe position projection of the slits on X-axis and160 F Maan013570.1394Y-axis are shown in Fig. 14(a),(b) respectively. Theaverage measured spacial resolution ox is 1800umand o, is 137 6um. Deducting the effect of slits widthwhich is 1536um for the slit of 0. 5mm and 922umfor the slit of 0.3mm based on the trapezium distribtion method, the real spacial resolution of the detec-tor is 94um in X direction and 102um in y direction9510105115t212513Postion (mm)Fig 14(b) Position distribution on Y-directionThe effective gain of triple-Gem detector canreach 5x10. USing the centre-of-gravity readoutmethod with 2D PCB, it can work well in X-ray im-aging, with a good spatial resolution better than0. 1mm after the electronic fluctuation be subtractedReferences1 Sauli F Nucl Instr Meth. 1997. A386: 531Fig 13 Slits imaged by the triple-GEM2 Benlloch J, Bressan A, Capeans M, et al. Nucl Instr Meth1998,A419:4105 Conclusion3 LI Cheng, ZHOU Yi, An Shaohui, et al. Electrostatic fieldsimulation and calculation of the gem(in Chinese). Pro-The performance of GEM devices is measured indetail. The effective gain of double-GEM can be up toceeding of the 11 China Symposium on ComputerApplication in Modern Science Technology, Changdao1.410 with each GEM voltage of 450V and suitablegap fieldsted in themixture of4 Bachmann S, Bressan A, Ropelewski L Nucl Instr MethAr: CO2 =70: 30. The multiplication and electrons col-lection of the GEM detector are two separate proc- 5 Altunbas C, Capeans M, Dehmelt K. CERN-EP, 2002-008esses which reduces the spacial charge effect. The6 LI Cheng, CHEN Hong-Fang, LE Yi, et al. HEP NP(indetector can work stably under X-rays with a high rateof10°HzmmTYH魏