by Panos Charitos. Published: 23 January 2013

Django Manglunki studied physics engineering in the Ecole Polytechnique in Brussels and arrived at CERN in October 1983, almost 30 years ago. By the end of his studies, CERN was looking to hire some young physicists and engineers to start working on the operations team of LEAR. Django decided to apply although it had been only one week since he had graduated and finally he got the position.

He recalls “I started working with heavy ions when I worked for LEAR, the Low Energy Antiproton Ring. At that time we worked a lot with heavy ions and we devised one of the first schemes for accumulating heavy ions. It was a longitudinal scheme obviously different from the one that is now used in LEIR to fill the LHC. Moreover, at that time it was oxygen and not heavy ions that we are currently accelerating at the LHC.”

Django is currently working as one of the supervisors of the SPS machine but also of the LEIR machine. For each of the accelerators there are one (or more) operators who work on shifts 24/24, and an engineer who supervises the smooth running of the machine for one week at a time. Django shares his time daily between LEIR and the SPS.

But this is only part of his job. He is also the project leader for heavy ions being responsible for all the machines until the ions fill the rings of the LHC. He smiles when he mentions that LHC is one of his primary “customers”! Apart from LHC, his team also delivers heavy-ions to other experiments, NA61 in the North Area which is also a very demanding customer as they need lots of different types of ions.

Apart from these tasks, Django spends much of his time in thinking about the future and what his team can deliver based on the things that are coming. “On the fixed-target site will be all the new heavy-ion species for the NA61 (Argon, Xenon) which we hope we will deliver before 2017 before they switch to lead. In addition I am working on ways in which the lumonisity at the LHC can be upgraded following the recent ALICE upgrade plans. We are looking for an upgrade of the luminosity by a factor of two or three compared to the current performance or five to six compared to the design report of the LHC”.

Fig.1 Django Manglunki at the CERN Control Centre.

Following heavy-ions from the source until their injection to the LHC

The injection complex used for heavy ions was re-vamped in the early ‘80s for the needs of fixed targets experiments in the North Area of CERN (NA48, NA49). It has been modified in the beginning of 2000 in order to conform to the needs of the LHC experiments.

Django, explains that CERN currently has an ion source which is an ECR (Electron Cyclotron Resonance) source that was built in Grenoble.

This source provides partially stripped lead ions centred around Pb29+ although we are also getting a broader spectrum from Pb22+ to Pb32+. Immediately after the source there is a spectrometer used to select one of the charged states which in our case is the Pb29+.This is followed by the linear structures which accelerate the ions to 4.2 MeV/n : a Radio Frequency Quadrupole (RFQ) and then an Inter-digital Linear Accelerator play this role.

“This machine operates at 5 Hz which means that you can send one burst of particles every 200 msec and this is the current mode of operation. Once accelerated to 4.2 MeV/n the beam is sent through a first stripping stage in which the particles pass through a very thin stripper that consists of a 0.3 μm carbon foil which is less than the wavelength of the visible light!”

After passing through a spectrometer the Pb54+ ions are selected and then the beam is injected into LEIR, the Low Energy Ion Ring that has replaced the old Low Energy Antiproton (LEAR). Django notes that “We had to modify LEAR in order to accumulate the highest possible density of particles.”


The role of LEIR is to transform a series of long (200 μs) low intensity ion pulses from Linac3 into shorter bunches with higher brightness. This is done by using a multi-turn injection technique that works in the three directions of the phase-space but also by applying electron cooling and accumulating bunches in order to increase the intensity of the beam. Each Pb54+ linac pulse is injected with 70% efficiency by stacking 70 turns into horizontal, vertical (by an inclined electrostatic septum) and longitudinal (by energy ramping) phase space. In the transfer-line there is a ramping cavity in order to cover the whole phase space in every direction as it distributes different energies depending on whether one is at the front or at the tail of the incoming beam. In that way we make sure that the injected beam fills the machine in all three directions of the phase-space.

Then you have to apply electron cooling in order to shrink the beam and hence increase the density. The electron cooler strongly reduces the phase space volume of the beam in less than 200 ms and decelerates it into a stack sitting slightly inside the central orbit. Electron cooling is a key element as it produces the required beam brightness which is a factor of 30 times higher than for fixed-target ion operation.

After accumulating 7 similar pulses the beam is adiabatically captured in 2 bunches and accelerated to 72 MeV/n. In order to understand how the continuum of particles in LEIR is split in 2 bunches one has to remember that the number of bunches is harmonic and given by the ratio of the frequency over the revolutionary frequency of the beams (i.e. if the RF is 1 HZ and we apply a frequency of 3 Hz then we get 3 bunches etc). This step is important in order to accelerate the beams and be able to inject them to SPS and finally to the LHC.

Fig.2 The LEIR facility at CERN. (Image: CERN)

PS and the line to SPS

The PS is as straightforward accelerator in which bunches from the LEIR are injected and accelerated before being injected to the Super Proton Synchrotron (SPS). In this step one has to be cautious in keeping proper distance between the bunches because otherwise particles in the PS can start behaving in an unusual way. This distance needs to be a multiple product of 25; the bunches are distanced further apart to 200 ns before been injected to the SPS.

At the exit of the PS the particles have energy of 5.9 GeV/n and go through another stripper that finally provides the required Pb82+. The stripper consists of a thin (0.8 mm) aluminium foil which is needed to get a fully stripped beam. All the particles extracted at high energy from the PS are going through this stripper and this means that depending on whether we are stripping heavy ions or protons delivered to other experiments the stripper has to be moved in and out of the particles path. Specifically, It moves in at the beginning of the first ion cycle and moves out at the beginning of the first proton cycle!

Finally, after being extracted from the PS, the Pb 82+ go through the SPS. The PS sends two bunches which are expected next to each other at a very long flat bottom of 40 sec and this is repeated for 12 injections before we accelerate lead ions at a total energy of 177 GeV/n and injecting them to the LHC.

Fig.3 CERN’s accelerator complex is a succession of particle accelerators. (Image:CERN)

Preparing LHC for proton – lead collisions

For the proton-lead run, one of the rings will be filled with protons and the other with lead ions. The switch between the two fillings is made in less than a minute and most of the complex can deliver both beams at the same time.

Django notices: “As long as we are having p-p or ion-ion collisions we don’t have many constraints on the filling scheme because if you make a filling scheme for one ring then you use the same for the other ring. If you have two different species you have to make sure that you will have the same patterns inside the SPS because ultimately it is the SPS that sends bunches to the LHC.”

There were indications that there would be problems to keep the beam quality on the injection scheme and deal with other difficulties in the SPS and LEIR. However, a very sophisticated scheme was designed to mitigate these problems and this finally resulted in the successful proton-lead pilot run in September 2012 and the current full proton-lead run just before the LHC long shutdown 1.