Activity report of ILD-TPC Asia group Yukihiro Kato Kinki University on behalf of ILD-TPC Asia group Contents 1. Introduction. Test beam 3. Analytic formula 4. Local field distortion 5. Effects of positive ion and gate devices 6. Cooling for electronics 7. Summary and future plan 1
ILD-TPC Asia group ILD Detector A group of ILD-TPC collaboration - Japan, China, Philippines - R&D of TPC with GEM radius 3.6m length 4.7m (max drift distance.m) B=3.5T
Performance goal of ILD-TPC Momentum Resolution σ(1/pt) = x10-5 (GeV -1 ) >00 sampling points along a track with a spatial resolution better than σrφ~100 µm over the full drift length of >m in B=3.5T (recoil mass, H µ + µ - ). High Efficiency -track separation better than ~mm to assure essentially 100% tracking efficiency for PFA in jetty events.high tracking efficiency also requires minimization of dead spaces near the boundaries of readout modules. Minimum material for PFA calorimeters behind, also to facilitate extrapolation to the inner Si tracker and the vertex detector Events/0. [GeV] 150 + Zhµ µ - X 100 50 s = 50 GeV -1 - + L int = 50 fb, P(e, e ) = (-0.8, +0.3) Signal+Background (MC) Fitted signal+background Fitted signal Fitted background 0 10 130 140 150 [GeV] M recoil Recoil mass measurement Particle Flow Algorithm 3
Why MPGD is good for ILD-TPC? MPGD (Micro Pattern Gas Detector) MPGD is a gaseous detector with gas amplification by the D micro pattern structure. Drift Volume Beam Advantage of MPGD against MWPC less ExB effect to rφ resolution in the high magnetic field (B=3.5T) fine two-dimension structure of the O(100µm) can achieve O(100µm) resolution and two track separation small dead signal region by the support frame Amplification Gap GEM Readout Pads Signal operation principle of GEM-TPC two MPGD candidates of ILD-TPC GEM micromegas ILD-TPC Asia group have proposed to use GEM and been doing R&D. 4
Large prototype and test beam Pad plane 19 pads/row 8 row/module 178 pads/row pad: (1.151.5) mm W X 5.6 mm H 0.1 mm gap, staggered layout Large prototype radius 70mm, length 610mm B = 1T (PCMAG) placed on DESY T-4 area (5 GeV electron beam ) End plate GEM sheet two GEM made of 100µm thick LCP Φ - 70µm, pitch - 140µm Purpose of Large prototype estimate to the TPC performance under the near condition of the real TPC compare the performance of the various readout modules 5
Beam test 010 and 01 DESY II test beam line T4 (electron pe = 5 GeV/c) Sep.1 - Sep.6 010 and Nov.19 - Dec.18 01 010 Data B=1T z-scan: (5cm - 55cm) Increase the partitions (-> 4) of GEM electrode to protect the readout electronics by discharge 01 version 010 version There were problems at 010 beam test 1. NO data of short drift length and various beam angle. large distortion around the module boundary 3. damaged readout electronics by discharge took below steps for 01 beam test 1. optimize HV of field shaper. increase the partitions of GEM electrode 3. add protection circuit to readout electronics 6 01 Data B=0T, B=1T z-scan: (.5cm - 50cm) x-scan: (-cm - cm) theta-scan: (-10,10 ) phi: (-10,10 ) different gains, shaping times more various data than 010 beam test
Result of beam test xy-resolution ( central module, row 17 ) 01 data 010 data MC (Neff = 8.5) 01 data is consistent with 010 data at short drift distance xy-resolution turns worse as drift length become long another group using large prototype at 01 observed same trend 7
Result of beam test (MC with real charge loss) If there is a real charge loss by some reasons, can the real charge loss explain the xy-resolution of 01 data? Compare 01 data and MC simulation with Neff = 8.5 (consistent with 010 data) [mm] x 0.3 0.5 01 data 010 data 0. 0.15-0.5%/cm charge loss Without charge loss 0.1 0.05 0 MC Neff = 8.5 Cd = 94.4µm/ (cm) 100 00 300 400 500 Drift Length [mm] -0.5%/cm charge loss can explain the change for worse of xy-resolution on 01 data charge loss may be caused by gas leak or bad O monitor 8
Analytic formula of Generalization toref.inclined Tra Ryo Yonamine s spatial resolution Ph.D thesis x (z; w, L tan, Cd, Nef f, N ef f, [f ]) Z +1/ = [A] + 1 Nef f [B] + [C] + X x d (aw) hhfa (x + y tan + w a diffusion-averaged & cluster 1 N ef f [D] N hg Obtained Knowledge [A] := x)i x iy x Analytical explanation is important Generalization of 1/ to understand the relation to drift perpendicular trac position average charge centroid systematics distance and spatial resolution in * + Spatial resolution consists of 4 Z +1/ X X x Exactly TPC. The four elements can be [B] := components. d (aw)fa (x + x) x)i xbehaviors General (aw) hfa (x + w σax [A] : systematics due to finite pad 1/ make influence of the spatial for per a x readout. diffusion-averaged charge centroid resolution. [A]z=0 + Generalization to Inclined Tracks displacement due to diffusion for a single electron d > 0.4 1 /w 1 if X disappears PR + (A) systematics due to finite pad [B] 1 same as G (z; w, L tan, C, N, N, [f ]) = [A] [B] + [C] + [D] Notation:Z Exactly + Y / f drift ef flength or x (long inclined tracks) [C] :=d ef (aw) dy sum Nef f hg(y)iy := P SI (k, y) g(y) N ef f readout N N a G perpendicular tr Y Y / [B] : diffusion effect (B) diffusion effect Z +1/ X L tan x [D] := Gas(x property whe (C) electronic noise [A]effect := d (aw)-hhf + y variance tan +projected x)i xto iyx-axis x for a primary Generalization cluster of [A] forvisible Effective only secondary a 1 w 1/ cluster 0 in the asymptotic formula ionization probability a diffusion-averaged - We found& that (D) primary cluster fluctuation perpendicular tracks [A] position average charge centroid Cd systematics z [D] + x = 0 + Z +1/ * X Generalized effe Nef f X [C] x Exactly same as with the [B] [B] := d (aw)f (x written + x) as σ0(aw) (x eff+. x)i x canabe =[A]hF /N z=0a number Drift tracks. lengthof electr w 1/ for perpendicular a a x We understood why Neff is much smaller than Normalized a seed electron diffusion-averaged charge centroid * N gain * for N + * P ki + + 3 Z (D) makes effort to hn i average electrons. ( ef [A]z=0 + d +1/ of f Xfor a single X N ) X displacement dueseed to electron x diffusion j=1 Gij 4 5 resolution of inclined N := d (aw) hf i (aw) hhf i i X [C] : electronic noise effect P P ef f a a same x x y as with the 1 N k Exactly [C] for i G w 1/ (aw) a a i=1 j=1 i=1 fluctuation track[c] := G y G N,k N N a [D] : primary cluster perpendicular tracks. * 3 if1 almost*constant a functionof drift +length Pas + σ0 ki N X L tan G ij φ is fixed. It vanished for φ=0. j=1 [D] := 5 Visible only when φ 0. variance projected a primary cluster 4 to x-axis forp P 1 Effective numbe N k i - We understood why N G ef f is much smaller than ij i=1 j=1 i=1 Slide by R.Yonamine effective number of seed electrons.g( N N,k primary clusters ef f Nef f ) 9 Normalized gain for a single primary cluster Normalized * gain * for a seed electron 9 + * Generalized effective 7 number of electrons. P + + 3 1
Analytic formula of Compare analytic formula and 010 test beam data (noise effect is assumed to negligible small) spatial resolution Ref. Ryo Yonamine s Ph.D thesis 010 data Analytic formula is consistent with test beam data The spatial resolution for the drift length can be understood. We can estimate the performance of the real TPC using the results of large prototype measurement. 10
Extrapolation to the ILD-TPC [mm] x 0.3 0.5 0. 0.15 0.1 (a) Ar-CF 4 -isoc 4 H 10 (95:3:), B=3.5T L=5.6mm W=1.15mm C d =30 µ m/ cm (E=30V/cm) (a) PRF =0.00 mm (LP1 configuration) (b) PRF =0.60 mm 0.05 (b) 0 0 00 400 600 800 1000 100 1400 1600 1800 000 00 400 Drift Length [mm] The expect performance by test beam is satisfied with the requirement of ILD-TPC 11
Local field distortion 010 test beam 01 test beam Distortion was still existed There was a big distortion Distortion size is smaller than 010 data because HV on field shaper was suitable. There were distortion at the GEM electrode gap (gap size: 00µm -> 1mm) Reason: The electric field was bend around module boundary because HV on field shaper were not suitable 1mm gap 00µm gap module boundary module boundary gap need to estimate the field uniformity The field uniformity will be checked using the track by UV laser system 1
Result of UV laser test laser laser Distortion is found near electrode boundary. Direction of distortion is opposite to the direction of laser. This problem is studied with the field simulation Field is bent to the electrode boundary on 01 version Residual of hit point for the track 1mm gap is too wide need new GEM with narrower gap drift region electrode boundary (1mm) GEM transfer region Result of simulation (01 GEM) 13
New GEM and UV laser system 500µm new GEM to suppress the distortion decrease boundary gap size (1mm -> 500µm) no boundary gap on front side front side back side Test plan with UV laser system measure the distortion of new GEM and compare to simulation measure the distortion of GEM with wire gate 60 60 GEM chamber 40 30 quartz window laser UV laser system test chamber (Nd-Yag λ=66nm) 14 event display (01 GEM)
Effects of Positive Ions 1.There are a lot of ions in TPC at the ILC experiment..these ions in the drift region make the distortion of electric field. 3.The distortion of electric field disturb the drift electron path. Estimate the effects Simulation of primary ion by A.Vogel for 100 bunch X-ings averaged over Z > x30 for 3k bunches 1 bunch train Primary Ions ( associated with track) secondary Ions ( associated with GEM) Solved the Poisson equation for the simulation ion density distribution with proper boundary conditions and then estimated the distortion of drift electron trajectory by the Langevin equation (D.Arai and K.Fujii) distortion of track by positive ions Not OK For the secondary ions from the amplification, we need an ion gate device for the ion feed back ratio of > 10-3 at the gas gain of 1000 15
Ions gates for ILD-TPC Requirement of ILD-TPC - ions feedback must be smaller than 10-3 (ie. no ions from MPGD) - Gate can be open for 1 msec and be closed following 199 msec. - ion can drift < 1cm 3 candidates Gate: open close wire GEM micromesh e MPGD e ion MPGD ~1cm known technology, local change of E directional, structure wire tension, ExB E E MPGD ion MPGD potential symmetry, local change of E electron transmission MPGD ion MPGD symmetry, simplicity electron transmission global E change 16
Wire gate module To decrease the dead region by support structures, need to put the wire radially Wires can create field distortions wire gate Radial wire structure - no radial support structures: minimizes dead regions - ExB in the wire direction -> minimizes distortions Prototype were built for test - 30µm wires, mm pitch - spot welded on stainless steel frame - only one potential : no alternate potential closed gate scheme Test with Fe55 Take Fe55 spectra for different HV configurations fixed drift field (30V/cm) change transfer field lower Et => lower transparency Edrift (30V/cm) Etrans next to laser test charge distribution peak distribution 17
Study of GEM gate Problem of current GEM gate electron transparency to low current gate - 35% at B=3.5T current GEM gate Φ90µm, pitch 140µm, t 14µm => 37% geometrical aperture Design of large aperture GEM gate Honeycomb structure 10µm wide, 100µm pitch - 81% aperture - difficult to build B=3.5T Large aperture gate simulation B=0T large aperture design by ANSYS Geometrical aperture is the key parameter for electron transparency in high B field 18
High transparency GEM gate How to build a high transparency gate? Key item is to keep large apertures A mechanically sound GEM gate with large aperture might be feasible - with thicker, harder metal - with wider holes -> 400µm holes, 40µm structure? influence of hole size to ion suppression influence of thicker metal influence of thickness 19
Cooling for electronics -phase CO system at KEK Cooling channel R&D Test bench CO pipes.1%/(345(6&6.0 FPGA (RCU) " " "#$%&'()*"+, " X 71%/(6&6.(#""6(&'(%8.(,&99#.(:;<(&'-/.0&'=(%8.( 9&0%'-.(;.%>..'(5(?3?(;#"-@0A SALTRO 71%/(-""#&'=("'(%8.(69(0&9.("'#< B8.(,&99#.(#""6(-"$#9(6./860(;.(&'%.=/%.9(&'( Y Schematic view of piping and heat sink -phase CO cooling system (Delivered at NIKHEF) Circulation system We will also set up the system at DESY beam test area for cooling test with S-ALTRO16 Thermal simulation 0
Summary ILD-TPC Asia group is working on R&D of GEM TPC Analytic formula for GEM-TPC explains the result of test beam well. The expect performance by test beam is satisfied with ILD-TPC performance. Solution of the distortion problem is important. The laser test system will check the distortion of new GEM module. Development of gate device is a key issue for ILD-TPC because of secondary ion effect. Especially, geometrical aperture is the key parameter in high B field. The wire gate is a default option, but MPGD gate fits better to GEM module. Study of cooling system is in progress. Tracking software is developed for non-uniform B field (Bo Li s talk) 1
Future Plan for TDR Design the GEM module with no (or small) distortion and test it. Continue the GEM gate study to get larger (>70%) electron transmission (= large aperture geometry). Wire gate is a backup option. Taking the beam test with 3.5T magnet. R&D of cooling and electronics are continued in cooperation with ILD-TPC collaboration. Development of common tracking software for ILD-TPC We have to do lots of things for the realization of ILD-TPC