Ions for LHC (I-LHC) Project

Homepage LHC ion injector chain LEIR commissioning
Minutes of meetings Reports Links

The I-LHC Project has been completed successfully and first LHC runs have taken place. These pages are not updated any more, but are kept as reference as they contain general information on the project and minutes of meetings.


Introduction to the I-LHC and LEIR Projects



In addition to proton operation, the LHC machine will run a few weeks per year with ions to provide collisions for heavy ion experiments. Operation with Pb ions is part of the approved LHC program. Collisions between Pb ions and protons and collisions of lighter ions not yet approved, but are likely to be included in future LHC upgrades.

The aim of the I-LHC project is to investigate all aspects to be taken into account and all hardware additions necessary, in order to allow Pb ion collisions in addition to proton operation. Wherever possible, design choices should be taken with a view to later upgrades which allow LHC operation with lighter ions.

With the ion acceleration scheme avaible at present for SPS fixed target experiments, no Pb ion beam useful for LHC can be prepared. Amongst the proposals for an LHC ion injector chain were (i) the implementation of a Laser Ion Source (LIS), and (ii) an accumulator ring with strong electron cooling. The latter has been approved and is now part of the LHC scheme. An overview of the approved ion injector scheme for LHC, with accumulation in LEIR, is shown in Fig. 3.

LHC Limitations and Design Performance

The design performance and beam parameters have been defined taking into account the main limitations for LHC operation with Pb ions :

  • Electron Capture by Pair Production (ECPP) :
    One of the possible interactions during the encounter of two ions is the creation of an electron-positron pair. The electron may be captured by one of the ions, which in turn will be lost in the dispersion suppressor. The flux of of ions lost is proportional to the luminosity in the given experiment must remain below the threshold leading to quenches. Thus, ECCP sets directly an upper limit to the luminosity of about 0.5 1027 cm-2s-1 to 1.0 1027 cm-2s-1.
  • Sensitivity of Position Pick-Ups :
    The limited sensitivity of the Position Pick-ups requires a minimum of 4 107 Pb82+ ions per bunch for reliable measurements. Note that this intensity is relatively close to the design intensity per bunch.

The LHC ion operation scheme is adjusted to above limitations, and takes limitations along the injector chain into account as well. The parameters for nominal ion operation are given in Table 1. For the first ion runs, an special early operation scheme, with significant simplifications along the whole LHC ion chain, is proposed and the relevant parameter are given in Table 1 as well. The users of LHC ion collisions are interested in an early ion run with lower luminosity, but expect an increase in luminosity for later operation. Thus, early LHC ion operation using the early opertaion scheme with later upgrades to achieve design luminosity is a valuable approach.
Table 1: Design performance and parameter for LHC operation with Pb ions.
Parameter unit nominal operation early operation
Initial Luminosity cm-2s-1 1027 5 x 1025
Energy/nucleon TeV/u 2.76 2.76
Number of bunches   592 60
Bunch spacing ns 100  
b* m 0.5 1.0
Transverse normalized rms emittances mm 1.5 1.5
Transverse rms beam size mm 16 16
Luminosity half-life with 2/3 experiments hrs 4.7/3.1 9.4/6.2
The situation is summarized as well in Fig. 1, showing various options in a diagram combining intensities per bunch and luminosities. The nominal scheme and the early stage scheme for Pb ion operation of the LHC are compatible with LHC constraints. In addition, points corresponding to the performance achievable with the actual ion acceleration chain for SPS fixed target operation is plotted. It is evident that this existing hardware cannot satisfy the needs for LHC Pb ion operation. The intensity of a Pb ion beam delivered with the existing ion chain would be by far too low for the LHC beam instrumentation and yield luminosities of no interest for the users. An Oxygen ion beam compatible with LHC constraints could be delivered and would yield a high luminosity. However, such a scheme is of minor interest for the users even for an early LHC ion run and, thus, has been ruled out.

Fig. 1 : LHC limitations and luminosities

Need for Accumulation

The Pb ion intensities achievable with the existing ion accelerator chain are far below the needs for LHC. Even with various improvements along the chain (e.g. upgrade of the existing ECR source, new PSB injection with stacking in the vertical phase space as well, recombination of bunches in the PS), there is no hope to satisfy the needs for LHC. Two schemes to provide the Pb ion beam for LHC have been tested in extensive experimental investigations, namely a Laser Ion Source (LIS), and accumulation with electron cooling in a low energy LEAR-like synchrotron. Despite significant effort and investments, it has not been possible to proof that a reliable injector chain for LHC ion operation based on LIS is feasible. On the other hand, various measurements and a proof of principle experiment performed at the LEAR machine allow to extrapolate that, with accumulation, the needs for LHC Pb ion operation can be satisfied. Thus, it has been decided to convert the previous LEAR machine into LEIR, a dedicated ion accumulator ring for LHC. In addition, various upgrades and modifications are necessary along the whole accelerator chain for LHC ion operation.

Experimental results in LEAR and Extrapolation to a Pb Accumulator for LHC

Extensive experiments in view of ion accumulation have been performed in the LEAR machine from 1994 to 1996. Those tests gave valuable information and have influenced the final design of the LEIR machine. Notably :
  • Life-times in presence of electron cooling and choice of the charge state :
    During the first experiments, an unexpected short life-time of Pb53+ (charge state initially envisaged for ion accumulation) has been observed in presence of the cooling electron beam. This is caused by a large cross section for capture of an electron from the cooling electron beam. Thus, life-times for neigbouring charge states, which can be delivered from the Linac in similar quantities, have been measured as well and turned out to be longer. Finally, it has been decided to accumulate Pb54+ ions rather than Pb53+ ions as envisaged initially.
  • Role of the dispersion and the betatron function :
    In cooling down time versus electron cooler current measurements with different lattices, it was found that (i) a finite dispersion enhances cooling rates and that (ii) intermediate betatron functions (around 5 m) yield fastest cooling. Both these observations are contrary to initial expectations. The lattice parameters at the electron cooler of the LEIR machine have been defined taking these observations into account. The betatron functions will be bH = bV = 5 m. Since, due to the injection process (with stacking in momentum as well), the momentum spread of the injected beam will be large, the nominal dispersion at the cooler is D = 0 m. However, with the LEIR geometry and hardware, a small (negative) dispersion is possible.
  • Beam loss induced vacuum degradation :
    A dramatic reduction of the ion life-time has been observed during accumulation tests. This could be traced back to a degradation of the vacuum caused by desorption of molecules from the vacuum chamber surface due to lost ions. Following these observations in LEAR, systematic measurements of ion impact induced outgassing have been done with a dedicated set-up installed at the end of the ion linac 3. In addition, during these tests a "scrubbing effect", i.e. a reduction in ion impact desorption rate after continuous bombardment has been observed. For LEIR, one will rely on this "scrubbing effect" in addition to a careful design of the vacuum system, aiming at dynamic pressure of just a few 10-12 Torr.
For a proof of the principle, accumulation tests have been performed with a modified LEAR machine in 1997. The electron cooler has been moved to another straight section, allowing to better adjust the lattice parameters at the injection and at the electron cooler. The time evolution of the intensity, accumulated with an electron cooler current limited to only 105 mA due to technical problems (instabilities) of the electron gun, is shown in Fig. 2. After about 1.6 s (four Linac pulses), the time available for accumulation in the future LEIR machine, an intensity of a about 3.4 108 ions, i.e. about one third of what is required for LHC, is accumulated. Thus, a factor 3 must be gained in accumulation rate. Furthermore, one notices that the accumulated intensity saturates quickly, because the beam life-time is reduced (from observing the intensity decay between injections). This is caused by beam loss induced outgassing mentioned above.

Fig. 2 : Intensity versus time measured during Pb ion accumulation tests in LEAR.
The Linac 3 repetition time was 400 ms.
Measures to gain the factor 3 for the intensity accumulated after 1.6 s needed for the LEIR project are :
  • Increase of the current delivered by the Linac :
    An upgrade of the ECR ion source is planned and an increase in current by a factor 1.5 to 2 is expected. To be fully effective, the beam emittances must not increase in order not to reduce the injection efficiency into LEIR.
  • Multiturn injection with stacking in vertical phase space as well :
    During the accumulation tests, a multiturn injection with simultaneous stacking in momentum in horizontal phase space has been applied. For the LEIR project, stacking in vertical phase space as well is planned. An increase of the intensity injected per multiturn injection by an additional factor 2 to 3 is expected. However, the resulting larger vertical emittance (compared to stacking in momentum and horizontal phase only) may result in somewhat longer cooling down times.
  • State-of-the-art electron cooler :
    A new electron cooler is constructed for the LEIR project. Amongst other improvements ("cooler" electron beam by expansion, a better quality of the magnetic field in the solenoid...), the gun will deliver higher stable electron currents than those used during the tests. This decreases the cooling down times and, thus, more injections are possible within the time available for accumulation.

Overview of the accelerator chain for LHC Pb ion operation

The whole accelerator chain for Pb ion operation of the LHC is depicted in Fig. 3 and Fig. 4.

Fig. 3 : Sketch of the accelerator chain for LHC Pb io operation.

Fig. 4. : Longitudinal structure of the Pb beam along the accelerator chain for LHC ion operation.
The main stages, necessary to provide the nominal LHC Pb ion beam, along the chain are :
  • Linac 3 :
    Production of the beam in an upgraded ECR ion source. After a first spectrometer, the Pb27+ beam with an intensity of 200 mA Pb27+ is bunched and accelerated in a RFQ. The beam is further accelerated in 3 RF tanks to 4.2 MeV/u. At the end of the Linac, the beam is stripped and the desired charge state Pb54+ is selected in a filter line. For the special LEIR multiturn injection, energy ramping, i.e. change of the energy during the Linac pulse is necessary. This will be done with a new cavity to be installed and the already existing debuncher (Note that, for the accumulation tests in LEIR, energy ramping has been provided with the debuncher only leading to a (slightly) larger momentum spread).
  • LEIR :
    In LEIR, 4 to 5 Linac pulses will be accumulated in order to build up the necessary intensity. Ingredients for short accumulation times are multiturn injection with stacking in all three (momentum, horizontal and vertical) phase spaces and strong electron cooling. After the accumulation at 4.2 MeV/u the beam will be bunched on harmonic number 2, and accelerated to 72 MeV/nucleon. Finally, the two bunches, each one containing 4.5 108 ions, will be transferred to the PS. The beam injected into and extracted from LEIR passes through a common transfer line, leading to some constraints.
  • PS :
    Amongst acceleration, complicated RF gymnastics (numerous changes of the harmonic number and bunch splittings) are necessary to provide the bunch pattern needed for transfer to the SPS. The bunch spacing necessary for LHC must be generated already in the PS. In addition every LHC bunch is split into a bunchlet pair distant by 5 ns. At every LEIR/PS cycle 4 bunchlet pairs, each one corresponding to one LHC bunch, are transferred towards the SPS. In the transfer line, the beam passes through an upgraded (minimizing emittance blow-up due to multiple Coulomb scattering) stripping insertion in order to fully ionize the Pb ions.
  • SPS
    On a long injection plateau, up to 13 LEIR/PS batches are accumulated. The space charge tune shift at SPS injection, which is expected to be a limitation, is reduced by transferring every LHC bunch split into a bunchlet pair. This, in turn, implies that bunchlet pairs must be recombined before extraction to build LHC bunches.
  • LHC :
    Finally 12 SPS batches, each one containing up to 48 bunches, are injected in one LHC ring. After about 20 minutes to fill both LHC rings, the beams will be accelerated and brought into collision.

Last update on Nov 1st, 2007 by S. Gilardoni or C. Carli