Faire du léger avec du lourd

La fragmentation nucléaire est ce qui se passe quand les noyaux se divisent, et cela se produit non seulement quand un hadron de haute énergie entre en collision avec un noyau, mais aussi dans les collisions périphériques entre ions lourds. La fragmentation des faisceaux d’ions lourds est utilisée dans plusieurs laboratoires pour produire des faisceaux d’ions légers. Le procédé a été introduit au CERN il y a 12 ans pour produire des ions légers destinés à l’expérience NA40 auprès du Supersynchrotron à protons. À présent, la collaboration NA61/SHINE a repris cette méthode, dans le cadre d’une étude systématique portant sur l’apparition du déconfinement dans les collisions d’ions lourds.

Nuclear fragmentation is the name given to the break-up of nuclei. It can happen when a high-energy hadron hits an intact nucleus. This is the process that is used to produce beams of exotic projectiles, such as radioactive nuclei, at CERN’s ISOLDE facility, which has served a worldwide community for many years (CERN Courier January/February 2012 p33). However, nuclear fragmentation also takes place in inelastic peripheral collisions between heavy ions, a process that is now being put to use to generate beams of light ions in the North Area at CERN.

In a heavy-ion collision, the nuclear matter is unstable outside the region where the interacting nuclei overlap – mainly because of the mismatch between shape and surface-energy – and it disintegrates into a mixture of different nuclei. The composition of the fragments produced, in terms of particle mass (A), charge (Z) and momentum, varies considerably from one collision event to another, even for fixed initial conditions in energy and impact parameter. This type of nuclear fragmentation has been studied extensively and found to occur over a range of incident energies, from as low as 20 MeV per nucleon up to highly relativistic energies. For a given collision system (that is, with specific values of A and Z for the projectile and target), the distributions of mass and charge of the nuclei in the final state are, to a good approximation, independent of the incident energy. The same independence is also true for the momenta of the produced ions in the rest frame of the corresponding parent nucleus. In the laboratory frame, however, the fragments experience an energy-dependent boost, which causes a forward-peaked angular distribution.

Fragmentation of beams of heavy nuclei is used at a variety of facilities, including GANIL, RIKEN, GSI and the National Superconducting Cyclotron Laboratory at Michigan State University (CERN Courier July/August 2008 p15). However, a different application of nuclear fragmentation was introduced 12 years ago at CERN, when beams of fragments with energies of 40A GeV/c and 158A GeV/c were produced in a primary carbon target and delivered to the North Area at the Super Proton Synchrotron (SPS).

Figure 1a shows the production cross-section of ion-fragment projectiles as a function of the fragment’s charge that was measured when lead nuclei at an energy of 158 GeV per nucleon collided with the carbon target (Cecchini et al. 2002, Thuillier et al. 2002). The results were in good agreement with model calculations and confirmed that there is a relatively high probability of producing ions with either low or high charge, giving rise to a U-shaped distribution of the kind previously observed at much lower energies (Trautmann et al. 1992, Schüttauf et al. 1996). At the same time a fragmented lead-ion beam was used by the NA49 experiment for physics, in which fragments with A/Z values close to two were transported to the experimental area. Charge measurements of the beam particles allowed "tagging" of the charge states Z = 6 or 14, corresponding to the 12C and 28Si ions whose interactions with the secondary target in NA49 were recorded and analysed. Fragmentation was also used to produce beams of mixed ions, with a large spread of combinations of A and Z, for the calibration of detectors such as the ring imaging Cherenkov counter for the Alpha Magnetic Spectrometer experiment in 2002 (Efthymiopoulos and Buenerd 2003).

The NA61/SHINE collaboration has recently revived this method with the aim of producing light-ion beams with increased purity (NA61/SHINE 2009). The work is part of an effort to study the onset of deconfinement in heavy-ion collisions and search for the critical point of hadronic/partonic matter by scanning systematically both in collision energy and in the size of the colliding nuclei (CERN Courier January/February 2012 p17). For the light-ion part of the programme, the collaboration decided to begin with a fragment beam, as primary light ions will become available in the North Area only in 2014. To create the light ions, a primary beam of lead ions from the SPS was directed towards a stationary target in the North Area, where a secondary beamline was tuned to transport projectile fragments with an optimized content of 7Be ions to NA61/SHINE, for studying the reaction 7Be + 9Be.

The selection and transport of a specific ion species from a fragmented heavy-ion (208Pb) beam is not straightforward. The secondary beamlines in the North Area are designed to transport particles emerging from the primary targets to the experiments. They basically consist of two large spectrometers, which can select particles with a range in rigidity (momentum-to-charge ratio, Bρ ≈ 3.31γA/Z) of ±1.5%. The desired ions produced in the fragmentation of the primary beam will be immersed in a variety of other nuclei that have a similar mass-to-charge ratio and, therefore, a rigidity value within the beam acceptance. Moreover, overlaps in rigidity occur not only for ions with the same mass-to-charge ratio but also for neighbouring elements. This is because the momentum of the ions varies as a result of the nuclear Fermi motion of the fragments. Without Fermi motion, the fragments would leave the interaction region almost undisturbed, with the same velocity (or momentum per nucleon) as the incident lead ions. Instead, the Fermi motion, which depends on the masses of the fragment and the projectile, can spread the longitudinal momenta of light nuclear fragments by up to 3–5% – i.e. much, more than the beam acceptance.

The 7Be ion was chosen for the beam for the NA61/SHINE experiment because it has no long-lived near-neighbours, thus allowing the production of a light-ion beam with a large proportion of the desired ions. The near neighbours to 7Be are its isotopes 6Be and 8Be and nuclei with a charge-difference of one and a similar mass-to-charge ratio (e.g. 5Li,9B). Furthermore 7Be has more protons (Z = 4) than neutrons (N = 3). Such nuclear configurations are disfavoured with increasing nuclear mass because a surplus of protons causes a Coulomb repulsion that cannot be balanced by the attractive potential of the smaller number of neutrons. Figure 1b shows ion rates in the fragment beam delivered to NA61/SHINE. It indicates that 7Be fragments are accompanied mainly by deuterons and helium ions, whose rigidity overlaps with that of the wanted ions because of the Fermi motion. A counter-example for the choice of ion-species would have been a nucleus with a mass-to-charge ratio of two, which would be accompanied by a range of stable or long-lived nuclei from 2D up to 56Ni.

At low energies, the insertion of a "degrader" into the beamline improves the separation of the desired ions (Münzenberg et al. 1992, Geissel et al. 1995), profiting from the double spectrometer configuration of the secondary beamline. The first spectrometer selects ions within a rigidity range that maximizes the proportion of wanted ions produced by the primary fragmentation target; on passing through the degrader, a piece of material introduced at the spectrometer’s focal point, these ions lose energy in a charge-dependent way. The second spectrometer then separates the ions spatially according to their charge so that they can then be selected by using a thin collimator slit.

The drawback of this method is a loss of beam intensity, through both the nuclear interactions and the beam blow-up caused by multiple scattering in the material of the degrader, which rises with increasing thickness. So the high separation power is accompanied by a high loss of intensity. Furthermore, for a given degrader thickness both the nuclear cross-section and the energy loss are energy independent to a large extent. This means that the separation power (ΔE/E) increases with decreasing energy.

NA61/SHINE is located on the H2 beamline in the North Area, where lead ions from the SPS are focused onto a primary beryllium fragmentation target, 180 mm long. In passing through the target the lead beam undergoes collisions, mostly peripheral, with the light target-nuclei. Part of the resulting mixture of nuclear fragments is captured by the beamline, which is tuned to a rigidity that maximizes the ratio of the created 7Be to all ions. Figure 2 shows the layout of the H2 beamline with its two-step spectrometer. The optional degrader (a copper plate either 1 cm or 4 cm in thickness) is located between the two spectrometer sections. The composition of the ion beam can be monitored by scintillation counters that measure the charge (Z2) and time-of-flight of the ions. The latter allows the determination of the mass (A) of the ions for momenta lower than 20 GeV/c per nucleon.

Investigations of fragment separation in the H2 beamline took place during test-beam time in 2010, using a 13A GeV/c lead beam incident on the primary target and with the 4 cm degrader in place. Figure 3 shows, for a given rigidity setting, the charge distributions detected with the collimator set to optimize the selection of either 7Be or 11C ions. During running in 2011 the NA61/SHINE collaboration used the configuration without degrader to record a total of 6 × 106 7Be + 9Be collisions at beam momenta of 158A GeV/c, 80A GeV/c and 40A GeV/c. A typical charge spectrum for a fragment beam selected by the spectrometer is indicated in figure 1b. With an incident beam from the SPS of several 108 lead ions per spill, typical beam intensities at NA61/SHINE were 5000 to 10,000 7Be particles per spill, with 10 to 20 times as many unwanted ions (Efthymiopoulos et al. 2011).

A second period with a 7Be beam is scheduled for autumn this year. It will be devoted to data-taking at beam momenta of 30A GeV/c, 20A GeV/c and 13A GeV/c. The latter is close to the lower limit of what is possible given the characteristics of the SPS accelerator and the external beamlines.