Published: 20 March 2013

This article is based on excerpts from an article written by J. Schukraft and C. Fabjan. It documents the main design choices and the close to 20 years preparation, detector R&D, construction and installation of ALICE.

ALICE, which stands for A Large Ion Collider Experiment, is very different in both design and purpose from the other experiments at the LHC. Its main aim is the study of head-on collisions between heavy nuclei, at first mainly lead on lead at the top energy of the LHC. In these reactions, the LHC’s enormous energy – collisions of Pb nuclei are 100 times more energetic than those of protons – heats up the matter in the collision zone to a temperature which is 100,000 times higher than the temperature in the core of our sun. Nuclei and nucleons melt into their elementary constituents, quarks and gluons, to form for a brief instant the primordial matter which filled the universe until a few microseconds after the Big Bang.

The hot reaction zone expands at almost the speed of light, and in the process cools, breaks up and condenses back into a plethora of ordinary, composite matter particles.

The ALICE detector has to measure as many as possible of the escaping particles, totalling up to several tens of thousands in each of these ‘little bangs’, and record their number, type, mass, energy and direction, in order to infer from the debris the existence and properties of matter under the extreme conditions created during the instance of the collision. >

ALICE will also take data with protons in order to compare them with the heavy ion results, both to look for the chances between the two types of beams and to characterise the global event structure of proton reactions with its very different and complementary set of detectors.


History and Challenges


The physics program of high-energy heavy-ion collisions and the search for the ‘Quark-Gluon Plasma’(QGP), the primordial matter of the Universe, started in 1986 at the CERN SPS accelerator and, simultaneously, at the Brookhaven AGS in the US. The first set of detectors, many of them put together from equipment used in previous generations of experiments, could actually only use rather light ion beams (from Oxygen to Silicon) and what today would be considered a rather modest energy at these fixed target machines (5 - 20 GeV centre-of-mass energy). Already the following year, in 1987, during a workshop to choose CERN’s next accelerator project amongst different contenders, the possibility of using both heavy ions as well as protons was mentioned for the machine which was to become the LHC. In 1990, when in the US the Relativistic Heavy Ion Collider (RHIC) - dedicated to heavy ion physics - was approved and a call for experiments was issued, the European community faced the decision to either participate at RHIC or focus its resources on the LHC. The schedules for RHIC and LHC were, at the time, quite comparable and therefore a sequential exploitation of both machines seemed impossible. After a series of workshops and discussions, which looked both at the physics potential of these machines and at different detector concepts, the decision was made, correct as it would seem with hindsight, to participate from Europe at a modest scale at RHIC and to start in parallel a dedicated design and R&D effort for a large general purpose heavy ion detector at LHC. This left the community with a busy schedule and thinly stretched resources: ongoing data analysis of the light ion program , constructing a new generation of experiments for the heavy ion program starting at CERN with Pb beams in 1994, designing and building detectors for RHIC (in operation since 2000), and R&D for an ambitious experiment at LHC. All this went on in parallel and involved many of the same actors and groups, but it did pay off handsomely in a rich program and fast progress.

Designing a dedicated heavy ion experiment in the early 90’s for use at the LHC some 15 years later posed some daunting challenges: In a field still in its infancy, it required extrapolating the conditions to be expected by a factor of 300 in energy and a factor of 7 in beam mass. The detector therefore had to be both ‘general purpose’ – able to measure most signals of potential interest, even if their relevance may only become apparent later – and flexible, allowing additions and modifications along the way as new avenues of investigation would open up. In both respects ALICE did quite well, as it included a number of observables in its initial menu whose importance only became clear after results appeared from RHIC, and various major detection systems where added over time, from the muon spectrometer in 1995, the transition radiation detector in 1999, to a large jet calorimeter added as recently as 2007.



A simulated high multiplicity event detected in the ALICE tracking detectors.


Other challenges relate to the experimental conditions expected for nucleus-nucleus collisions at the LHC. The most difficult one to meet is the extreme number of particles produced in every single event, which could be up to three orders of magnitude larger than in typical proton-proton interactions at the same energy and a factor two to five still above the highest multiplicities measured at RHIC. The tracking of these particles was therefore made particularly safe and robust by using mostly three-dimensional hit information with many points along each track (up to 150) in a moderate magnetic field (too strong a field would both mix up the particles and exclude the lowest energy ones from being observed).

In addition, a large dynamic range is required for momentum measurement, spanning more than three orders of magnitude from tens of MeV to well over 100 GeV. This is achieved with a combination of very low material thickness (to reduce scattering of low momentum particles) and a large tracking lever arm of up to 3.5 m (resolution improves at high momentum with the square of the measurement length), thus achieving good resolution at both high and low momentum with modest field.

And finally, Particle Identification (PID) over much of this momentum range is essential, as many phenomena depend critically on either particle mass or particle type. ALICE therefore employs essentially all known PID techniques in a single experiment, as discussed in some detail in the following sections.

The ALICE design evolved from the Expression of Interest (1992) via a Letter of Intent (1993) to the Technical Proposal (1996) and was officially approved in 1997. The first ten years were spent on design and an extensive R&D effort. Like for all other LHC experiments, it became clear from the outset that also the challenges of heavy ion physics at LHC could not be really met (nor paid for) with existing technology. Significant advances, and in some cases a technological break-through, would be required to built on the ground what physicists had dreamed up on paper for their experiments. The initially very broad and later more focused, well organised and well supported R&D effort, which was sustained over most of the 1990’s, has lead to many evolutionary and some revolutionary advances in detectors, electronics and computing.



In the ALICE experimental cavern



ALICE Detector Overview


ALICE is usually referred to as one of the smaller detectors, but the meaning of ‘small’ is very relative in the context of LHC: The detector stands 16 meters tall, is 16 m wide and 26 m long, and weights in at approximately 10,000 tons. It has been designed and built over almost two decades by a collaboration which currently includes over 1000 scientists and engineers from more than 100 Institutes in some 30 different countries. The experiment consists of 17 different detection systems, each with its own specific technology choice and design constraints.

A schematic view of ALICE is shown in Fig.1. It consists of a central part, which measures hadrons, electrons, and photons, and a forward single arm spectrometer that focuses on muon detection. The central ‘barrel’ part covers the direction perpendicular to the beam from 45° to 135° and is located inside a huge solenoid magnet, which was built in the 1980’s for an experiment at CERN’s LEP accelerator. As a warm resistive magnet, the maximum field at the nominal power of 4 MW reaches 0.5 T. The central barrel contains a set of tracking detectors, which record the momentum of the charged particles by measuring their curved path inside the magnetic field. These particles are then identified according to mass and particle type by a set of particle identification detectors, followed by two types of electromagnetic calorimeters for photon and jet measurements. The forward muon arm (2°-9°) consists of a complex arrangement of absorbers, a large dipole magnet, and fourteen planes of tracking and triggering chambers.



Fig 1: Computer model of the ALICE detector. The different sub-detectors are labelled with their acronym, which are not always explained but listed both in the figure and the text to help locating the various components of ALICE.


Two features of ALICE, not found in the other LHC detectors, are at the origin of several unusual installation challenges:

The asymmetric layout with the large magnetic Muon Spectrometer constraints the installation of and access to the central detectors: the 200 tons of ‘central’ detectors can only be placed from the opposite side; these central detectors placed inside the L3 magnet can only be supported from the mechanically rigid iron crowns at the ends of the magnet, separated by a distance of 15 meters.

These seemingly innocent issues kept several brilliant and creative engineers occupied for the better part of five years. The asymmetry imposed by the Muon spectrometer on the overall design was not accepted lightly. The alternative implied however displacing the 900-ton muon dipole magnet, the 300 ton muon filter, the muon detectors together with a large section of the delicate Be-vacuum chamber. The final verdict was rather clear: the complex, delicate one-sided installation represented the lesser evil.

First, the Muon Spectrometer was installed in its final, fixed position together with the hadron absorber. A ‘ballet’ of ITS and TPC motions had to be minutely orchestrated to allow the installation of the vacuum chamber, the independent Pixel and ITS detectors and, finally, the TPC.

Connecting the detectors to cables, gas and cooling lines required to place these detectors at various intermediate positions. This was not only complex and delicate, it was potentially dangerous: the movement of the detectors caused large, up to 5 mm vertical deformations of the supports, while the vacuum chamber, attached to the detectors, was limited to excursions of less than 2.5. mm. This installation scenario could literally ‘make or break’ the ALICE experiment; it was reviewed by many committees, dress-rehearsed on the surface with many of the final components, monitored with strain gauges, feeler gauges, cross-checked by survey teams, engineers and physicists… It took the better part of nine months during 2007 to install these systems. Using a variety of tools from the inevitable duct tape to a dentist’s drill for final dimensional adjustments, the operation was completed successful and in time as can be seen in Fig 15, which shows the experiment shortly before completion in early 2008, essentially ready to accept beams from the LHC.



Front view of the L3 magnet, with its doors partially open. A bridge guides services, power cables, fluids and gases into the central detector. The silicon tracker is no longer visible and even the large TPC is mostly obscured. The stainless-frame structure which supports most of the central detectors is partially filled with TOF and TRD modules.



Looking forward

New technology, skilful engineering, and critical design decisions have led to a state-of-the-art detector that will be up to the task of observing the primordial matter created by heavy ion collisions in the LHC. ALICE is the first truly universal ‘general purpose’ heavy ion experiment, which combines in a single detector most of the capabilities assigned in the past to several more specialised experiments. Incorporating the fruits of many years of R&D effort dedicated specifically to meeting the numerous challenges posed by the physics of nuclear collisions at the LHC, it is ready and well prepared, after more than 15 years of design and construction, to explore the ‘little bang’ and enter ALICE’s wonderland of physics at the LHC.



Excitement in the ALICE control room during the first collisions (November 2009).



The article was originally prepared for the experiment chapter of the book ‘The Large Hadron Collider: A marvel technology’, EPFL-Press Lausanne, Switzerland, 2009 (Editor: L. Evans).