Uni Karlsruhe (TH) Institut für Experimentelle Kernphysik Magnetic Field Studies of a Time Projection Chamber with GEM-Technology a) 4 se 3 N n o i s s Se alk 2 T BERKELEY LAB a) a) B. Ledermann , M. Ball , J. Kaminski , S. Kappler , a) c) b) Th. Müller , M. Ronan , P. Wienemann so l a e DESY b) a) IEKP, Karlsruhe, Germany b) DESY, Hamburg, Germany c) LBNL, Berkeley, USA A GEM-TPC for the future Linear Collider: The GEM-Technology: The Gas Electron Multiplier (GEM) consists of a very thin, two-side metal-clad polymer foil (56 µm total thickness), which is perforated with a high density of photolithographically etched holes. As illustrated in Figure 1, these holes are placed in a triangular pattern with a typical pitch of 140 µm and a typical diameter of 70 µm. electrons ions 1 Figure 1: Electron microscope photograph of a ´Standard Geometry´ GEM One proposal for the future Linear Collider project is the TESLA project at DESY (Hamburg, Germany). The central tracker of choice is the Time Projection Chamber (TPC). Parameters of the TESLA-TPC: Radius 161.8 cm; Length: 2x 250cm; Favoured gas mixture: Ar:CH4:CO2-93:5:2 (so-called TDR-Gas) 6 5 copper kapton copper 70 µm 2 4 3 Figure 2: Schematic view of the electric field lines in a GEM, showing also the principle of ion feedback suppression copper kapton copper 3 Principle of a GEM-TPC: As seen in the right picture, an ionizing particle creates a track of electron-ion-pairs. The electrons drift with constant velocity along the field lines towards the endplate, where they are amplified by the GEMs and finally collected by the pads. With The 2D information from the readout plus the drift time, a precise 3D image of the track can be reconstructed readout electronics 4 5 140 µm Applying a potential difference (usually 300-500 V) between upper and lower metal layer creates a strong electric field inside the holes and makes proportional gas amplification possible. Due to the electric field configuration, electrons are mostly released into the volume below, the backdrifting ions are mostly absorbed by the upper metal layer suppressed ion feedback STAR readout electronics: 320 channels The low probability for electrical sampling rate: 19.66 MHz discharges of GEMs in the gas mixture and their insensitivity to aging grant stable operation even in the harsh radiation environment of hadronic beam experiments. GEMs and pad structure y Track Gas mixture: typically Ar:CH4:CO2-93:5:2 (so-called TDR-Gas) 25 cm Figure 4: Different cluster width for different magnetic fields; definiton of a, b Variation of the sampling rate effects not only the long. spatial resolution, but also, through Transverse and longitudinal spatial cross-correlation, the transverse spatial resolution. resolution worsen due to diffusion and due to pad-size limitation. 186.3cm Measurements in high magnetic fields: To evaluate the performance of a GEM-TPC in high magnetic fields our TPC prototype was put inside the former compensation magnet of ZEUS (at DESY), a superconducting solenoid with a 28cm bore and a magnetic field of up to 5.5 T. Measurement of diffusion coefficients: A magnetic field parallel to the 0.4 0.3 0.2 s 2 = s 02 + D 2 x s 0 » 450 m m 0.1 0 20 15 10 5 spatial resolution in µm transverse diffusion coefficient in µm/ cm squared cluster width in mm² electric field forces the drifting electrons on a helix and thus the electron diffusion and the resulting cluster size is reduced (Figure 4). Measured diffusion coefficients are in good agreement with MagBoltz simulations. transverse spatial resolution for different sampling rates Transverse diffusion coefficients Squared cluster width at B = 2T 0.5 450 MagBoltz simulation 400 Experimental data 100 0 100 80 60 300 250 200 1 2 3 4 5 TDR 4T 19cm 0 0 Bernhard Ledermann 1 2 3 4 5 6 7 alpha-cut in ° 600 i*ts_length 2*pad_length 550 500 450 600 TDR 4T 19cm 2 4 6 8 10 12 14 16 20 25 0 staggered geometry 3and1 geometry 500 combs geometry 400 300 200 5 10 15 20 25 drift distance in cm 18 alpha in ° ledermann@iekp.fzk.de 0 5 10 15 20 beta in ° staggered geometry 800 3and1 geometry 750 combs geometry 2 mm 700 650 600 450 12 0 25 850 500 expected for the pad-size limited region: 2mm = 577µm 100 300 longitudinal spatial resolution for different pad geometries 550 350 50 0 15 drift distance in cm To find the best suiting pad geometry three special geometries (see Figure 5) were tested in the magnet. The results show that the 3and1 geometry has advantages in transverse resolution and the comb geometry in longitudinal resolution. 400 0 10 transverse spatial resolution for different pad geometries longitudinal spatial resolution for different angles 100 20 5 Testing different pad geometries at B = 4T in the pad-size limited region: 50 150 40 600 100 magnetic field in T 350 700 400 150 spatial resolution in µm 120 800 500 0 i*pad_width 2*pad_length 160 200 drift distance in cm 400 180 120 transverse spatial resolution for different angles spatial resolution in µm spatial resolution in µm 140 200 250 25 450 220 Sampling rate: 12 MHz Sampling rate: 16 MHz Sampling rate: 19.66 MHz Sampling rate: 25 MHz Sampling rate: 33 MHz Sampling rate: 36 MHz 140 300 The following figures show the dependence of the transverse spatial resolution on the angle a (Figure 4). By application of hard cuts on a you can reach transverse spatial resolutions of 50-60 µm. Also shown is the depenb=arcsin( ) a=arcsin( ) dence of the longitudinal spatial resolution on the angle b. Steps in resolution can be explained by inclusion of one more pad or time slice to the track. 160 Sampling rate: 12 MHz Sampling rate: 16 MHz Sampling rate: 19.66 MHz Sampling rate: 25 MHz Sampling rate: 33 MHz Sampling rate: 36 MHz 240 longitudinal spatial resolution for different sampling rates 350 Dependencies of spatial resolution on different angles at B = 4T: transverse spatial resolution for different angle-cuts a readout boards: 32x8 pads of 1.27x12.5 mm (normal pads) 30x12 pads of 2x6 mm (different pad geometries) Amplification: 2 GEMs typical gain: 3000 scintillator 2 x 6 mm scintillator 1 t b spatial resolution in mm µ spatial resolution in µm solenoidal magnetic field drift volume Advantages of a GEM-TPC: - Intrinsically suppression of ions released into the drifting region (~10-3) - Almost evanescent distortions due to ExB-effects - Higher granularity - Small signal in transverse direction - Truly 3D-detector (no ambiguities) - Minimum amount of material - Good dE/dx resolution (~5%) TPC prototype: length: 25cm inner diameter: 20cm spatial resolution in mm Figure 3: Schematic view of magnet setup at DESY diffusion 0 5 10 15 400 20 25 drift distance in cm 0 5 10 15 20 25 drift distance in cm Figure 5: Three different pad geometries: staggered, 3and1, combs http://www-ekp.physik.uni-karlsruhe.de/~lederman IEEE, Rome 2004