Comparison of Endcap CID-PID Klaus Föhl PID meeting 27/3/2007 panda-meeting Genova Focussing Lightguides Time-of-Propagation Proximity Focussing

Focussing Lightguides focal plane coord. [mm] lightguide number

Focussing Lightguides short focal plane 50mm ~1mm pixels needed optical errors exist thicker plate a problem 125mm disc-PMT focal plane 100mm pixel width 2-3mm benign optics thicker plate ok 200mm disc-PMT

Focussing Lightguides no LiF plate

Focussing Lightguides no LiF plate LiF plate performance decrease due to poor phi resolution for particle track near the edge, partial remedy to be investigated N.B. only rough matching of current geometry homework: for photon area of large angle tracks increase phi resolution by subdividing pixel length all calculations: =300nm-600nm Quantum Efficiency 20% n 0 =15.28/mm optimisation within the sum rule 6K pixels

Time-of-Propagation already investigated –excess radius –pixelisation effect  fall-off at larger angle to investigate –rectangular hole –different hole coatings  proper ray-tracing –several outer shapes –path ambiguties (with similar travel time) r=900mm and 1100mm r=900mm (23deg equivalent) 30ps 512 pixels 256 pixel 128 pixel not relevant not relevant

Time-of-Propagation hexagon with rectangular hole ToP disc Gießen Design Saclay version: hexagonal shape rectangular black hole reflective holeabsorbing hole 8 deg photon number corresponding to 54 tracks

Time-of-Propagation TOP  =70ps N 0 =344 n 0 =17.19/mm [ref: Markus Ehrenfried, Saclay talk] hexagon with rectangular hole circle black rectangle circle mirror rectangle hexagon mirror rectangle hexagon black rectangle circular with rectangular hole comparison: hexagon 960mm width or round disc 1100mm radius  t =30ps

Time-of-Propagation single photo timing crucial performance increase comes with more tracks in the time-angle-plane reflective holeabsorbing hole 16 deg these calculations: =400nm-800nm Quantum Efficiency 30% n 0 =17.19/mm per band:  n(group)= (inspired by [480nm-600nm]   n=

Proximity Focussing design variation with mirror and the expansion volume upstream radiator placed closer to EMC

Proximity Focussing C6F14+ CsI+GEM radiator 15mm expansion 135mm [no] mirror

Proximity Focussing detection plane needs to be larger than the radiator size to catch all photons on the possible cones or mirrors at the fringes to fold Cherenkov cone back onto the active area design variation with mirror and the expansion volume upstream radiator placed closer to EMC

Performance comparison focussing lightguide ToP C6F14+ CsI+GEM radiator 15mm expansion 135mm [no] mirror

Performance comparison radiator X0 [ n 0 [1/mm] N 0 N 0 *sin ^2 *geometry pixels photon rate..per area..per pixel Focussing 15mm SiO 2 12% /cos K 4E9 1/s 1.25E6 1/s/cm^2 0.66E6 1/s/pixel ToP 20mm SiO 2 16% /cos (75) 1K (?) 6E9 1/s 4.5E6 1/s/cm^2 (100%) 6E6 (?) 1/s/pixel Proximity 15mm C 6 F 10 7% ( +window ) E4 – 1.2E5 2.4E8 1/s ~1E4 1/s/cm^2 Cherenkov photons N=d*n 0 *sin C ^2=N 0 *sin C ^2 lightguide PMT 128* 25cm2 2E7 interactions/s ; multiplicity 6 ; Endcap region ~50% in CM  6E7 particles/s

Time-of-Propagation

Proximity Focussing

Photon yield – visible photons photons produced in radiator plus CsI quantum efficiency absorption plus quantum efficiency 20mm 10mm curves show photon yield for an energy interval starting at E_photon=5.4eV r.home.cern.ch/r/richrd26/www/hmpid/richsim.html material properties from RICHSIM web page at CERN

Simulation ingredients proper Cherenkov photons number and colour refractive index dispersion –Cherenkov angles –Snell's law absorption length quantum efficiency statistical analysis normal incidence particles only  maths simplification no angular straggling liquid without vessel no detector pixels (assumed to be small) Fresnel formula simplified (Brewster angle being close) perfect mirror Simplifications & Approximations C6F14 CsI mirror

Hit pattern response shape (1000 particles) photons 1 particle  =0.99 The particle distance is the average of the photon radial distances resulting from one charged particle. Particle distance mean and sigma are computed for samples of 1000 events  =1 and 1000 events  =0.99 and sigma separation & 4  -limit derived.

Performance comparison focussing lightguide ToP C6F14+ CsI+GEM radiator 15mm expansion 135mm [no] mirror

potential edge effects detection plane needs to be larger than the radiator size to catch all photons on the possible cones or mirrors at the fringes to fold Cherenkov cone back onto the active area design variation with mirror and the expansion volume upstream radiator placed closer to EMC

performance – radiator width

angle dependence preliminary and approximate calculation

source of material properties material data used is shown left and below from a RICHSIM web page at CERN r.home.cern.ch/r/richrd26/www/hmpid/richsim.html

material transmission radiator thickness

Comparison of Endcap PID

Time-of-Propagation TOP  =30ps N 0 =344 n 0 =7.64/mm