ALICE selects gas electron multipliers for its new TPC

18 March 2016

The experiment plans a major upgrade for its TPC.


The ALICE experiment is devoted to the study of strongly interacting matter, where temperatures are sufficiently high to overcome hadronic confinement, and the effective degrees of freedom are governed by quasi-free quarks and gluons. This type of matter, known as quark–gluon plasma (QGP), has been produced in collisions of lead ions at the LHC since 2010. The detectors of the ALICE central barrel aim to provide a complete reconstruction of the final state of Pb–Pb collisions, including charged-particle tracking and particle identification (PID). The latter is done by measuring the specific ionisation energy loss, dE/dx.

The main tracking and PID device is the ALICE time projection chamber (TPC). With an active volume of almost 90 m3, the ALICE TPC is the largest detector of its type ever built. During the LHC’s Runs 1 and 2, the TPC reached or even exceeded its design specifications in terms of track reconstruction, momentum resolution and PID capabilities.

ALICE is planning a substantial detector upgrade during the LHC’s second long shutdown, including a new inner tracking system and an upgrade of the TPC. This upgrade will allow the experiment to overcome the TPC’s essential limitation, which is the intrinsic dead time imposed by an active ion-gating scheme. In essence, the event rate with the upgraded TPC in LHC Run 3 will exceed the present one by about a factor of 100.

The rate limitation of the current ALICE TPC arises from the use of a so-called gating grid (GG) – a plane of wires installed in the MWPC-based read-out chambers. The GG is switched by an external pulser system from opaque to transparent mode and back. In the presence of an event trigger, the GG opens for a time window of 100 μs, which allows all ionisation electrons from the drift volume to enter the amplification region. On the other hand, slow-moving ions produced in the avalanche process head back into the drift volume. Therefore, after each event, the GG has to stay closed for 300–500 μs to keep the drift volume free of large space-charge accumulations, which would create massive drift-field distortions. This leads to an intrinsic read-out rate limitation of a few kHz for the current TPC. However, it should be noted that the read-out rate in Pb–Pb collisions is currently limited by the bandwidth of the TPC read-out electronics to a few hundred Hz.

In Run 3, the LHC is expected to deliver Pb–Pb collision rates of about 50 kHz, implying average pile-up of about five collision events within the drift time window of the TPC. Moreover, many of the key physics observables are on low-transverse-momentum scales, implying small signal-over-background ratios, which make conventional triggering schemes inappropriate. Hence, the upgrade of the TPC aims at a triggerless, continuous read-out of all collision events. Operating the TPC in such a video-like mode makes it necessary to exchange the present MWPC-based read-out chambers for a different technology, which eliminates the necessity of active ion gating, also including complete replacement of the front-end electronics and read-out system.

The main challenge for the new read-out chambers is the requirement of large opacity for back-drifting ions, combined with high efficiency to collect ionisation electrons from the drift volume into the amplification region, to maintain the necessary energy resolution. To allow for continuous operation without gating, both requirements must be fulfilled at the same potential setting. In an extensive R&D effort, conducted in close co-operation with CERN’s RD51 collaboration, it was demonstrated that these specific requirements can be reached in an amplification scheme that employs four layers of gas electron multiplier (GEM) foils, a technology that was put forward by Fabio Sauli and collaborators in the 1990s.

A schematic view of a 4-GEM stack is shown on the previous page (figure 1). Optimal performance is reached in a setting where the amplification voltages ∆V across the GEMs increase from layer 1 to 4. This maximises the average number of GEMs that the produced ions have to pass on their way towards the drift volume, hence giving rise to minimal ion-escape probability. Moreover, the electron transparency and ion opacity can be optimised by a suitable combination of high and low transfer fields ET. Finally, the hole pitch of the GEM foils has proven to be an important parameter for the electron and ion transport properties, leading to a solution where two so-called standard-pitch GEMs (S, hole-pitch 140 μm) in layers 1 and 4 sandwich two GEMs with larger pitch (LP, hole-pitch 280 μm) in layers 2 and 3.

After being developed in small-size prototype tests in the laboratory, a full-size TPC inner read-out chamber (IROC) with 4-GEM read-out was built and tested in beams at the PS and SPS. To this end, large-size GEM foils were produced at the CERN PH-DT Micro-Pattern Technologies Workshop, in so-called single-mask technology (figure 2). As a main result of the test-beam campaigns, the dE/dx performance of the 4-GEM IROC was demonstrated to be the same as for the existing MWPC IROCs, and the stability against discharge is well suited for operation at the LHC in Run 3 and beyond.

After approval of the Technical Design Report by the LHC Experiments Committee, and an in-depth Engineering Design Review of the new read-out chambers in 2015, the TPC upgrade project is presently in its pre-production phase, aiming to start mass production this summer.

bright-rec iop pub iop-science physcis connect