Cornell ILC Global ILC

WWS Home

"Scintillator-based Hadron Calorimetry"

1. Overview

The CALICE collaboration pursues an integrated approach to the development of electromagnetic and hadronic calorimetry within the particle flow (PFLOW) concept. In this concept, the jet energy resolution required to identify heavy bosons by their hadronic final state di-jet mass is achieved by reconstructing each particle in the jet individually. This imposes high demands on the imaging quality of the calorimeters and thus requires high longitudinal and transverse granularity.

The goal of this project is to develop the hadron calorimeter (HCAL) on the basis of scintillator as active material. With the advent of novel high gain silicon photo-sensors – so-called SiPMs - the high segmentation required for PFLOW reconstruction can be realized with scintillators at reasonable cost. The energy response of scintillators allows to trade amplitude resolution versus granularity and thus to optimize the cost of the readout electronics, too. In addition to the classical analogue readout, semi-digital concepts with few threshold bit information or a purely digital approach are also followed. A scintillator HCAL is a promising candidate for all PFLOW based detector concepts (SiD, LDC and GLD).

In a collaborative effort, the group is presently building a cubic-meter size scintillator steel calorimeter read out by 8000 SiPMs, for a combined testbeam program with the CALICE silicon tungsten electromagnetic calorimeter (ECAL), to be carried out at CERN and FNAL in 2006 – 2008. The setup includes a tail catcher module with scintillator strips, using the same readout components as the HCAL. The goal of the testbeam effort is the proof-of-principle of the PFLOW approach to calorimetry with a scintillator-based HCAL, and to collect hadron shower data with unprecedented granularity, needed to validate the simulations and develop the reconstruction algorithms.

The implementation of the technology for a full-size ILC detector is not yet addressed with the testbeam prototype. For example, the overall-cost relevant thickness of the active readout layer has not yet been optimized. These issues will be the focus of the R&D program for the forthcoming years, with the goal to propose a scintillator HCAL, and to demonstrate its feasibility with a realistic, scalable prototype, by the time of the GDE technical design report. To meet this goal, a significant increase of funding will be required.

2. Previous results

Multi-pixel Geiger mode avalanche photo diodes (SiPMs) have only recently become available in larger quantities from Russian industry. These millimeter-size devices can be mounted directly on scintillator tiles and can operate with moderate bias voltage in high magnetic field. With typically 1000 independently quenched pixels on a common load they provide a signal proportional to the number of pixels fired by registered photons and a gain comparable to that of vacuum photo-tubes. Nowadays, the technology is followed by several suppliers around the world and is also driven by non-HEP applications, e.g. astrophysics or medical imaging.

A subgroup of this project (Czech, German and Russian institutes) has built a first small scintillator steel prototype (the “minical”) read out with 100 SiPMs, and successfully tested it in the DESY electron beam. An important conclusion from the published results was that the inherent non-linearity of the individual SiPMs, due to the finite number of pixels, can be controlled such that linearity and resolution of the calorimetric response is as good as with conventional APD or PMT readout, which were tested with the same setup. The results and the positive operational experience have established the SiPM as baseline for the future developments in this project.

3. Testbeam prototype and test program

The development and construction of a cubic-meter size calorimeter with 8000 readout channels with the given resources was only possible with a high degree of task-sharing and common use of resources not only within the scintillator HCAL project, but also with the other CALICE activities. The granularity has been optimized such that the analogue and the semi-digital approach (proposed by NIU) can be tested simultaneously. The readout chain will use the DAQ developed by the UK groups for the SiW ECAL from the ADC onwards, and the absorber stack and mechanics is designed to be also used with gaseous active detector layers.

The SiPMs are developed, produced and mounted at MEPHI in collaboration with PUSLAR, Moscow. Characterization and quality control of the SiPMs is performed at ITEP, where they are assembled and tested with the scintillator tiles provided by ITEP. Construction of the full active modules takes place at DESY where also the on-detector electronics boards were developed and produced. The central front end electronics component is an 18-channel ASIC developed at LAL on the basis of the ECAL chip. Electronics for LED monitoring and slow control is provided by the Prague group. The mechanical structure (stack and movable table) is a DESY project. The first HCAL modules have been assembled and successfully tested in the DESY electron beam in summer 2005.

The tail catcher and muon tracker with scintillator strip and SiPM readout has been developed and constructed at NIU. It was recently integrated into the HCAL front end and DAQ system at DESY and will be further commissioned and beam-tested in 2006 at FNAL.

The HCAL detector shall be completed and further commissioned at DESY, with support from MEPHI, NIU and Hamburg University, to be ready for first testbeam data taking in 2006 at CERN. Running with electrons and hadrons is foreseen, in the widest available energy range, with the HCAL alone and in conjunction with the ECAL. The testbeam program shall be continued in 2007 at FNAL, to allow direct comparison with other HCAL options.

Altogether, it will be necessary to collect a data volume of order of 108 events. This data set will provide novel in sight into the details of hadronic shower development and serve to validate the simulation programs. Presently these simulations hampered by model uncertainties which are far too large for a reliable detector optimization.

For calibration, analysis, simulation and reconstruction the collaboration is heavily relying on a structured software environment developed by CALICE in close interaction with the ILC software and simulations group. This should facilitate the feedback from the testbeam effort into the overall detector optimization with respect to ILC physics benchmark analyses.

The operational experience being collected during this construction and commissioning phase and to be expected from the first testbeam data taking is most valuable for the further development of the scintillator HCAL. All contributing institutes intend to further pursue R&D in their field of expertise.

4. Future R&D

The next phase of R&D must address the implementation of the novel SiPM technology into a large scale detector concept. This cannot be deduced from previous experience and must be developed before a realistic scintillator based HCAL is proposed for the ILC detector.

A key issue is to further consolidate the SiPM technology. This requires studies of long-term stability and radiation hardness, for example, and a further industrialization of production and quality control, to cope with the large total number of several million sensors ultimately needed.

There is a considerable potential for optimizing the performance of the sensors, e.g. their efficiency or spectral sensitivity, which would result in important simplifications of the coupling between sensor and scintillator and – or alternatively - allow to reduce the scintillator thickness.

Front-end electronics which optimally exploits the fast SiPM response still needs to be developed. Highly integrated state-of-the art ASICs should be placed inside the detector volume and deliver already digitized data on a small number of readout lines. Power dissipation issues raised in such an advanced concept need to be addressed.

The calibration concept will receive important input from the testbeam experience. A robust and redundant, but cost-effective system based on light injection, radioactive sources and / or ionizing particle events must be included in the detector design.

Altogether, an electro-mechanical concept, with a thin integrated readout layer (scintillator, photodetectors and electronics) needs to be developed and validated with a scalable, realistic prototype, corresponding to a section of a possible ILC HCAL. This project is followed by the European and Russian groups and receives support from the EU, which, however, will only cover part of the needs for the R&D outlined here.

In order to evaluate the benefits, in terms of electronics cost, of the semi-digital approach, a separate readout design effort should be undertaken, as part of the integration concept study. This is being proposed by US groups, but not yet sufficiently funded.

R&D for the purely digital variant has just started in Canada with the construction of a small single-layer prototype, with the aim of proceeding towards multi-layer prototyping later-on. The project has many issues in common with the already mentioned R&D, but the smallest cell size requires special solutions, e.g. for the photo-sensor coupling, and of course for the electronics, therefore extra resources will be required to study this option, too.