The physics process that imposes the strictest performance requirements on the electromagnetic calorimeter is the intermediate mass Higgs decaying into two photons. Thus the benchmark against which the performance of the electromagnetic calorimeter (ECAL) is measured is the di-photon mass resolution. The mass resolution has terms that depend on the resolution in energy (E1,E2) and the two photon angular separation () and is given by

where denotes a quadratic sum, E is in GeV and is in radians. The energy resolution is usually parametrized as

where a is the stochastic term, b the constant and is the energy equivalent of noise. The noise term has two sources, namely electronics noise and the pileup energy. In order to achieve a good energy resolution all the contributing terms have to be kept small and should be of the same order at the relevant photon energies.

Both fully active and sampling ECALs have been considered. It is difficult to achieve a stochastic term much below in sampling ECALs without demanding strict mechanical tolerances. This sets the scale for the requirement of the constant term () and noise term ( MeV equivalent). However most fully active media have the potential to achieve stochastic terms of . Here the limitation is set by the control of systematics that build up the constant term. Achieving a constant term smaller than 0.5% for a large calorimeter is a challenging task.

A totally active ECAL using lead tungstate () crystals have been chosen. The reasons for the choice are given below.

6.1 Crystal Electromagnetic Calorimeter for CMS

The fundamental properties of some heavy scintillating crystals have been investigated during 1991-1993 by IHEP (Protvino) and INP (Minsk) in a framework of UNK projects, with a significant investment into the Russian industry. The PWO crystals were found best suited for the electromagnetic calorimetry and beam tests of the first samples were performed at the beginning of 1992 in IHEP [6.1].

The further study of these crystals including the basic properties investigations in the context of the application to electromagnetic calorimetry performed by IHEPINPAnnecy team in 1993-1994 at home institutes and at CERN, have demonstrated a very high potential of PWO in electromagnetic calorimetry. The recent results obtained from the beam tests of electromagnetic calorimeter prototypes by the CMS ECAL group at CERN are very close to that of "target" parameters for the CMS ECAL (see below). That has led to the decision of CMS collaboration to use PWO as the baseline option for the electromagnetic calorimeter. The groups involved in PWO study together with IHEP, INR, JINR and LPI groups which participated before in other options of the CMS ECAL intend to continue participation in the ECAL calorimeter project in a framework of RDMS collaboration.

6.2 Beam tests of EM calorimeter prototypes.

A series of the electromagnetic calorimeter prototypes beam tests was performed in 1992-1994 in the beams of U-70 (IHEP) and SPS (CERN). Energy and coordinate resolutions, light output, EM shower profile have been measured in a wide beam momentum range from 5 to 150 GeV. Several light readout options have been tested using photomultipliers, PIN diodes and avalanche photodiodes.

The best energy resolution was obtained with a 5 x 5 matrix with Philips XP1911 photomultipliers used for the light detection. The energy resolution can be parametrized as:

The coordinate resolution was measured in 26 GeV electron beam of U-70. It is mm for the matrix of 22 X 22 X 200 mm³ crystals with photomultiplier readout.

6.3 Detector overview

6.3.1 The Design of the Barrel Calorimeter

The barrel part of the electromagnetic calorimeter (EB) covers the pseudorapidity intervals The endcap electromagnetic calorimeters (EF) cover the intervals The gaps between the barrel and the endcaps are used to route the services of the tracker and preshower detectors. The EB granularity is 432-fold in and (108 X 2)-fold in . The front face of each crystal has a square section of 20.5 X 20.5 mm². All crystals are positioned on a cylinder of a radius of R = 1.44 m. The length of the crystals is 23 cm corresponding to 25.8 X0. The volume of the crystals in the barrel amounts to 9.75 m³ (81 tons).


Each half-barrel will be built from eighteen supermodules each subtending 20° in . This leads to a module size that is small enough for construction and assembly in several institutes. The aim is to have independent sub-units which can be assembled with the full set of electronics, monitoring systems and services. Various step by step tests can then be carried out as the assembly proceeds until calibration in test beams. This scenario allows considerable flexibility in the installation.

The individual supermodules will be suspended from rails which are mounted on the inner surface of the HCAL. Each supermodule would comprise four 'baskets' with almost equal numbers of crystals. ( Fig. 6.1 )

The Basket

The tension in the four side walls of the basket ( Fig. 6.1 and Fig. 6.2 ) is balanced by compression forces applied at the back of each crystal. This force keeps each crystal accurately in its position and renders partitions between the crystals unnecessary. The tension in the side walls maintains the basket's shape within the tolerance required for assembly and installation.

An axial force, between two to five times the crystal weight, is applied from the back of each crystal. The crystals press against the bottom plate of the baskets. ( Fig. 6.1 ) A capsule is attached to the rear face of the crystal to evenly distribute the applied force. It has a spherical end to ensure that no moment is applied to the crystal. The crystal front face is terminated by a spherical cap for the same reason. The inner face of the basket bottom plate has a moulded cover that matches the shape of the spherical cap.

The capsule houses the photodetector and its preamplifier board, the monitoring optical fibre, the temperature sensor, a thermal connection to ensure the cooling of the preamplifier board and the thermal regulation of the rear part of the crystals.

6.3.2 Endcap Calorimeter

The endcap PWO calorimeter (EF) occupies the space z=3200 - 3760 mm; and 1.653 to 2.61 in . In it is subdivided into 72 five degree sectors. In the radial direction each sector consists of 11 modules.

The pattern of crystals in the sector is shown in Fig.6.3, the total number of the crystals in the cell being 112. The total number of crystals/side is 8064. All the crystals have a length of 230 mm and a "quasipointing" geometry, i.e. are rotated away from the interaction point by 3° . To realize this geometry, 34 different crystal shapes are introduced, the maximum lateral size of a crystal not exceeding 29 mm. The total weight of the sector is ~ 112 kg, the weight of one detector is ~ 8 tons.

The endcap preshower detector (SF) is situated in front of the EF and has two layers of silicon strip detectors ( and directions) positioned after 3X0 of an absorber. The thickness of the silicon wafers is chosen of 300 µm, the pitch of strips is ~ 2 mm. Each or plane has 1248 detectors and in total it gives 4992 detectors for both endcaps. The total area or silicon detectors for endcap region is ~9 m² each side.

The present CMS Technical design also includes the Barrel preshower (SB) with coverage in range of with one layer of silicon wafers. The area of detectors for SB is ~ 30 m² (~ 8300 silicon strip detectors).

6.3.3 Mechanical structure

The mechanical design of the EF follows closely the one for the EB. The scheme of centering and pressing mechanisms for the crystals similar to the barrel solution (Fig. 6.4).

The calorimeter is attached to the front flange of the endcap HCAL (HF). On the front flange of the EF the SF preshower, with a weight of ~ 2 tons is located. The body of each EF sector is made of composite material. Inside it is subdivided into cells in which the crystals are installed and positioned with an accuracy of ~ 0.1 mm. The maximum gap between the crystals in one sector is ~ 0.3 mm and between the sectors is ~ 1.0 mm.

The SF detector is designed as an independent construction and could be handled separately from the EF. The main rigidity element of the preshower is the absorber made as stainless steel lead stainless steel sandwich. This absorber is used as the front flange of a preshower.

To provide full overlap in both and directions silicon strip detectors with dimensions 60 X 60 mm² are placed in four planes. ( Fig. 6.5 ) Two successive detector planes are attached on both sides of a cooling layer. Every cooling layer is divided into eighth azimuth sectors. Each sector contains 39 micromodules with detector and readout hybrid chip placed on both sides. Sectors are attached to the absorber.

For cooling of preshower detectors and read-out electronics sectors have inner cavity for liquid coolant. The temperature of silicon detectors should be ~ 5 C° at inner radii and ~ 10 C° at outer.

A serious problem is that of thermo-stabilization of the calorimeter with an accuracy of 0.2 Ci°. For that special aluminum panels which have cooling channels for liquid circulation are foreseen in the front and the back side of the calorimeter, together with an isolation material.

To provide the thermoinsulation between the SF and EF the layer of the polyethylene is used. This polyethylene should act as the neutron moderator too.

6.3.4 The cost estimate

The total estimated cost of the CMS ECAL is, at present, 79.2 MCHF, the EF calorimeter cost (including the SF preshower detector) is 16.3 MCHF (4.5 MCHF for the preshower).

6.4 RDMS participation in ECAL R&D phase

Taking into account a big experience of IHEP, INR, JINR groups in electromagnetic calorimetry, and INP in a basic crystal properties investigations and 4 years R&D efforts in crystal calorimetry and PWO in particular we are going to contribute significantly in all aspects of CMS ECAL R&D:

  • Crystal parameters optimization

(i.e. common R&D of the institutes and PWO crystals producers).

  • Crystal parameters control

Design and development of the test system for the complete set of essential crystal parameters measurement. The system will be used during the technology optimization R&D and later as an output control during the calorimeter cells mass production.

  • Crystal treatment.

Cutting and polishing technology optimization for the best possible light output and light collection uniformity.

  • Light collection optimization.

Systematic study of all calorimeter cell design options which are essential for the light collection i.e.:

  1. wrapping, painting;
  2. crystal-photodetector optical coupling;
  3. light concentrator.
  • Photodetector

The current baseline option of ECAL photodetector is Avalanche Photo Diode (APD). Hamamatsu and EG&G APD's were used to detect the scintillation light from the crystals in the matrices tested in CERN beams in 1993 and 1994. A large area APD from Minsk could be considered if promising results will be obtained from R&D. The first samples have been tested with PWO matrix during the 1994 beam tests.

  • Preshower

Significant part of the preshower price is the cost of Si strip detectors. JINR together with Russian Research and Production Association ELMA (Zelenograd) under CERN RD35 [6.2] activity program designed new technology of Si detector production. At the present moment the detector production price is 5.5 $/cm² and further improvement of the detector production technology is needed.

Second aspect of the preshower activity is the performing of beam-tests of new preshower read-out system adequate to LHC requirements.

  • Development of the monitoring system.

Adequate monitoring system is essential to achieve the designed parameters of the calorimeter. This is especially true for the endcap region where the effects of the radiation damage must be monitored.

  • Beam tests of the prototypes at U-70 and SPS beams.
  • Radiation hardness tests of the detector components at INR MMF charged particle beams and JINR neutron beams.

6.5 RDMS responsibilities in ECAL during the construction phase.

6.5.1 Responsibilities of the Institutes

Contribution of RDMS in the ECAL project during several years makes natural the future participation in all the aspects of the CMS ECAL design and construction.

In the frame of CMS ECAL Organization structure we are going to participate in the following aspects of the ECAL project:

  • Regional centre activity (see later).
  • Crystal production and treatment.
  • Mechanical design of the ECAL, barrel and endcaps.
  • Monitoring and calibration.
  • Electronics (mainly low-noise preamplifiers).
  • Software.
  • Simulation.
  • Production of the silicon strip detectors for SB and SF preshowers.
  • Assembling of the main parts of the Preshower.

6.5.2 Involvement of the Russian industry

The distinguishing feature of the CMS ECAL project is a deep involvement of the Russian industry. This is mainly because at present Russia has a substantial production capacity of the PWO crystals. There are several plants which in total may produce all the necessary 13 m³ of crystals, the main are:

  • Bogoroditsk plant, Tula region. It has 100 furnaces.
  • Sergatch plant, Tula region, 50 furnaces.
  • Appatity plant, Murmansk region, 180 furnaces.

The total cost of the crystals is 23 MCHF, and it is obvious that this will be an attractive project for the Russian industry.

The silicon strip detectors (substantial fraction) should be produced at Russian Research and Production Association ELMA (Zelenograd). Further assembly of micromodules and cooling sectors for SF should be performed at JINR (Dubna).

6.5.3 Regional Center organization in Russia.

Due to essential contribution of the RDMS institutes to the ECAL project and deep involvement of the Russian industry to the production of about 50 000 of PWO crystals it is proposed to organize a Regional Center in Russia (IHEP, Protvino).

The organization and the activity of the Regional Center will need close contact with CMS ECAL management and will be based mainly on a collaboration between IHEP and INR with their obligations to prepare 1/3 - 1/4 of the crystals produced in Russia for final assembling at CERN as described below:

IHEP will be responsible for:

  • test of the crystals;
  • wrapping, painting or coating of crystals;
  • mounting of photodetectors and very front end electronics on the crystals;
  • measurement of cell characteristics;
  • crystal installation in ECAL modules including the electronics, cooling and monitoring systems and testing of this modules before sending them to CERN for final assembly in the experiment;
  • beam calibration and beam-tests in Russia of some part of crystals.

    INR will be responsible for:

    • test and control of photodetectors;
    • test and control of preamplifiers;
    • transportation of crystals from the Regional Center to CERN.

    Other institutes involved in ECAL project will also participate in Regional Center activities.