The CRES RICH

1
The CRES RICH
CErenkov Ring Electron Spectrometer at the CERN
SPS (1990-2000) www.physi.uni-heidelberg.de/phy
si/ceres Presented by Peter Glässel, Heidelberg
(now ALICE)

• Univ. Heidelberg, H.J. Specht
• Weizmann Institute, Rehovot, I. Tserruya
• MPI-K Heidelberg, J.P. Wurm
• Politechnico di Milano, E. Gatti
• BNL, P. Rehak
• CERN, J. Schukraft

2
Outline

• CERES setup
• Radiators and windows
• Mirrors
• UV-photon detectors
• Understanding performance
• Summary

3
CERES RICH layout
axially symmetric
4
CERES in 1996
5
Radiator and UV windows

• CH4 at ambient pressure ? ? 32
• O2 and H2O about 1 ppm
• RICH1, L 90 cm
• window CaF2 in Al alloy
• UV bandwidth limited by radiator gas CH4
• RICH2, L 175 cm
• window quartz (Haereus Suprasil) in Invar frame,
  sliding O-ring seal vs. radiator tank
• bandwidth limited by quartz

6
UV1 CaF2 window

• 10 mm thick
• 10 sectors, each consisting of 4 pieces, glued
• Al alloy (cast) with tuned thermal expansion
  coeff
• border to CaF2 shaped like a bellow (1mm wall)

7
CERES RICH mirrors

• RICH1 1 mm CFK, replica-molded on glas master
• 60 cm ?
• Invar ring mount
• Reflectivity typ gt 85, 70 _at_ 140 nm
• Optical quality does not limit RICH resolution
• RICH2 ring-shaped, 10 sectors
• 1.8 m ?
• 6 mm glas, slumping method from flat glass (Andre
  Braem/ CERN)
• After software-correction for focal length
  variations does not limit RICH resolution

8
RICH1 Mirror

• 1 mm CFK
• replica on glas master
• 60 cm ?
• Invar ring mount
• Reflectivity
• avg. 85,
• 70 _at_ 140 nm
• Optical quality does not limit RICH resolution

9
RICH2 Mirror

• ring-shaped
• 10 sectors
• 1.8 m ?
• 6 mm glas, slumping method from flat glass (Andre
  Braem/ CERN)
• After offline-correction for focal length
  variations does not limit RICH resolution

10
The CERES UV detectors, 1st try
11
Assembly of UV2
12
UV2 detector
13
Why did 2-step (or 3-step) PPAC not work?

• Avalanches 5?107 108 e
  spark (Raether Criterion 1939)
• Independent of quencher ? Fonte, Pestov, Sauli
  NIM A305(1991)91
• Gas gain x dE/dx in typ. avalanche dimension
• 105 x 1000 e/mm x 1 mm 108
• Slow knock-on protons from shower neutrons do it!
• CERES in 32SAu 20 spark probability per
  interaction in UV2 (3 m2)
• Solution fundamentally different behavior of the
  multiplication schemes in the space-charge limit
• PPAC gain diverges
• MWPC gain saturates
• New scheme worked up to Pb-Au (? 20 dead time
  due to sparks)

Last stage must be a MWPC
14
The CERES UV detectors from 1992 on
15
UV1 multi-wire anode plane
16
Other UV detector specs

• Meshes stainless steel, 50 µm ?, 500x500 µm
  pitch, except first mesh (30 µm ? for higher
  transparency)
• Materials G10 frames, stainless steel, teflon,
  delrin, Cu, selected epoxy
• Resistive layer carbon-filled epoxy, 150 M?/
• MWPC anodes 30 µm ?, 2 (3) mm pitch
• 53.800 (48.400) pads
• He 6 CH4 TMAE(400 C) at ambient pressure,
  operated at 500 C
• Gas gain 2 4?105
• Detectors functioned for 7 beam periods

17
Event displays
Close-up of ee- pair
18
Event displayPb-Au
19
Understanding N0
from CERES Status Report to the SPSLC, CERN
SPSLC/94-02
20
Understanding Resolution
21
Cherenkov Ring Radial distribution

• Ring center resolution
• 1-dim, for tracking, momentum

?c 1.6 ?h / sqrt(N)
22
Single hit background
23
Summary

• TMAE is a manageable UV-converter
• Beware of spark limit for any gas multiplication
  scheme
• Theoretical performance limits can be closely
  reached in practice