BEAM-LOSS INDUCED PRESSURE RISE OF LHC COLLIMATOR MATERIALS IRRADIATED WITH 158 GeV / u In 49 + IONS AT THE CERN SPS

During heavy ion operation, large pressure rises, up to a few orders of magnitude, were observed at CERN, GSI, and BNL. The dynamic pressure rises were triggered by lost beam ions that impacted onto the vacuum chamber walls and desorbed about 10 to 10 molecules per ion. The deterioration of the dynamic vacuum conditions can enhance charge-exchange beam losses and can lead to beam instabilities or even to beam abortion triggered by vacuum interlocks. Consequently, a dedicated measurement of heavy-ion induced molecular desorption in the GeV/u energy range is important for LHC ion operation. In 2003, a desorption experiment was installed at the SPS to measure the beam-loss induced pressure rise of potential LHC collimator materials. Samples of bare graphite, sputter coated (Cu, TiZrV) graphite, and 316 LN stainless steel, were irradiated under grazing angle with 158 GeV/u indium ions. After a description of the new experimental set-up, the results of the pressure rise measurements are presented, and the derived desorption yields are compared with data from other experiments.


INTRODUCTION
An intense experimental programme was started at the CERN Heavy Ion Accelerator (LINAC 3) in November 2000 to measure the beam-loss induced molecular desorption of Pb 53+ ions at 4.2 MeV/u in order to prepare the Low Energy Ion Ring (LEIR) vacuum system for operation with heavy ions under LHC-type conditions.Effective desorption yields of up to 2 × 10 4 molecules/ion were measured for standard stainless steel vacuum chambers [1].Initial pressure rises depended strongly on the surface preparation of the vacuum chamber and could be reduced by polishing, noble-metal coating, and continuous bombardment (beam scrubbing) with heavy ions [2].Similar high desorption rates and very strong pressure rises were also reported from GSI [3,4] and BNL [5,6].At RHIC, a yield of 1.5 × 10 7 molecules/ion was reported for Au 79+ ions with an energy of about 10 GeV/u.[6].This value is about three orders of magnitude higher than the yields measured previously at LINAC 3.This latter observation, indicating a strong effect of the ion energy on the desorption rate, motivated the SPS experiment described in this paper.The measurements were done with In 49+ ions at 158 GeV/u, an energy very close to the 177 GeV/u injection energy of Pb 82+ into the LHC.

EXPERIMENTAL SETUP
At the CERN SPS an experiment has been installed in October 2003 to measure the ion-induced desorption yields for In 49+ at 158 GeV/u.An overview of the experimental set-up is shown in Figure 1.The experiment consists of a rotatable 316 LN stainless steel vacuum chamber with an inner diameter of 400 mm and a height of 292 mm.This central chamber is pumped with a turbo molecular pumping (TMP) group, a 400 l/s sputter ion pump (SIP), and a 1200 l/s titanium sublimation pump (TSP).Pressure measurements are made with a Bayard-Alpert gauge (BAG) and a quadrupole residual gas analyzer (RGA), both calibrated.The RGA can be calibrated in situ via a gas injection valve.Four identical 316 LN stainless steel vacuum chambers, with an inner diameter of 156 mm and a length of 300 mm, are connected symmetrically (90º between each) to the central vacuum chamber.Each chamber is equipped with a different collimator sample, which were prepared in the following way.Three commercially available graphite blocks [7], each 200 mm long, 65 mm wide, and 20 mm thick, were first snow jet cleaned with CO 2 to remove the dust from the graphite surface and afterwards vacuum fired at 1000ºC for 2 h.One bare graphite block was kept under vacuum, while the second block was sputter coated with a nonevaporable getter (NEG) film of TiZrV (1.5 µm), and the third graphite sample was sputter coated with Cu (1.5 µm).The fourth sample was a machined 316 LN stainless steel block (same dimensions as the graphite) that was cleaned and vacuum fired at 950ºC for 2 h.Each sample was mounted on a motorized manipulator.The manipulators enable a vertical sample displacement of z = ± 25.0 mm with a resolution of 0.02 mm.The position z = 0 corresponds to the center of the vacuum system.All four collimators are fixed in the horizontal plane (x-y direction) and were carefully aligned in order to guarantee an ion impact angle of Θ = 35 mrad.A picture of an assembled sample inside a collimator chamber is shown in Figure 2.Each sample chamber also contains one electron detector which is mounted on a port perpendicular to the collimator surface.A description of the functionality of these electron detectors, which were successfully used during electron cloud studies at the SPS, can be found elsewhere [8].For the SPS ion-induced desorption experiment the standard electron detectors were modified and are now bakeable at 300ºC.The sample chambers are closed with specially made 100 µm thick 316 LN stainless steel windows, which allow the 158 GeV/u ion beam to pass the experimental set-up with negligible interaction.After assembly the vacuum system was baked at 300ºC for 48 h.A pressure of 7 × 10 -12 Torr was achieved 72 h after cool-down to room temperature.The experiment was then transported under vacuum to the SPS North Area, moved into the T4-H8 beamline, and fixed on a rail system, which allows the removal of the test-stand from the beam axis.Two filament scanners, which are part of the SPS North Area beam line instrumentation, are used to measure the beam position.

Pressure rise measurements
At the beginning of the experiment the In 49+ ion beam was carefully aligned and passed the vacuum system without hitting the collimators, which were kept in their parking position, i.e., 18.5 mm below the beam axis.The pressure remained at about 6.2 × 10 -12 Torr.The beam size was measured with the filament scanners to 2 × 2 mm 2 full width half maximum and 4 × 4 mm 2 full width.
For pressure rise measurements only one collimator was moved from its parking position into the beam axis and bombarded under 35 mrad grazing incidence with ∼1.5 × 10 6 In 49+ ions/spill.The spill length was 6.2 s and the SPS cycle length 19.2 s.Pressure readings of the Bayard-Alpert gauge were taken every ∼150 ms.The results obtained for the Cu/graphite sample are shown in Figure 3.A surprisingly small pressure increase was clearly identified when the 158 GeV/u indium beam was hitting the sample (see Fig. 3a).The measured pressure rise pattern fits very well with the time the collimator is bombarded (6.2 s) and the period (13 s) without ion beam (see Fig. 3c) Very similar pressure rises were measured for the other three collimator materials, i.e., for bare graphite, TiZrV/graphite, and 316 LN stainless steel.

Desorption yields
The measured pressure rises, shown exemplary in Figure 3, are used to derive the desorption yields of the four LHC-type collimator materials.The effective desorption yield η eff (molecules/ion) is given by where ∆P is the pressure rise under ion bombardment, S is the pumping speed in l/s, is the number of impacting indium ions per second, k B is the Boltzmann constant, T is the temperature (300 K) and G is a constant (≈ 3.2 × 10 19 at 300 K), that converts gas quantities (Torr × l) into number of molecules.
The pressure rises ∆P (N 2 equivalent), measured with the calibrated Bayard-Alpert gauge, had to be corrected (factor 2.2) because of the time constant (10 s) of the used electrometer.This value is larger than the 6.2 s spill length which corresponds to the ion bombardment period.Both time constants explain the fact that the pressure rise curves (see Fig. 3) did not stabilize but show a nontypical "zigzag" behaviour.
A pumping speed of S ≈ 1270 l/s (for CO) was measured in situ after the experiment.We obtain the following η eff values for the different samples: η eff (Graphite) ≈ 55800 molecules/In 49+ ion η eff (Cu/Graphite) ≈ 111500 molecules/In 49+ ion η eff (TiZrV/Graphite) ≈ 111500 molecules/In 49+ ion η eff (316 LN stainless steel) ≈ 37200 molecules/In 49+ ion It has been verified by Monte Carlo simulations that a thermal contribution of the SPS indium beam can be neglected for the desorption yield estimation.

Comparison with other desorption data
An overview of available data for heavy-ion induced desorption yields, measured as a function of the ion energy, is shown in Figure 4. Effective molecular desorption yields, derived from machine experiments (AGS: Au 31+ , SIS 18: U 28+ , RHIC: Au 79+ ), are compared with dedicated "test-stand" experiments (LINAC 3: Pb 53+ , HLI: Pb 27+ , Zn 10+ ).The new SPS data fit well with the low energy LINAC 3 data obtained for bulk 316 LN stainless steel and OFE copper vacuum chambers (η eff ≈ 2 × 10 4 molecules/Pb 53+ ion).The wide spread of LINAC 3 desorption yields (see Fig. 4) is explained by the strong influence of the surface preparation (different polishing and coatings) of the stainless steel vacuum chambers [2].We conclude that the effective molecular desorption yield of heavy ions does not depend strongly on the ion energy as suggested by the RHIC result.More ion desorption measurements in the range of ∼10 MeV/u up to ∼100 GeV/u are proposed to verify this conclusion.Fig. 4: Overview of heavy-ion induced desorption data including the SPS measurements with In 49+ ions at 158 GeV/u.The measurements were made at CERN [1,2], GSI [3,4], and BNL [5,6].The ion impact angles (perpendicular, mrad, µrad) of the experiments are different.

SUMMARY AND CONCLUSION
The present study represents the first ever desorption experiment with heavy ions at an energy well above 10 GeV/u.In 49+ ions with 158 GeV/u bombarding LHC-type collimator material under 35 mrad only led to a very small pressure rise in the 10 -12 to 10 -11 Torr range.Effective beam-loss induced desorption yields of (4 − 11) × 10 4 molecules per In 49+ ion were measured for different materials.No dramatic pressure rise (several orders of magnitude) was observed and the reported RHIC desorption yield of 1.5 × 10 7 molecules/Au 79+ ion is not confirmed.The influence of the ion type, ion charge state, and especially the impact angle, on the desorption yield remains to be investigated.

Fig. 1 :
Fig. 1: Schema of the ion-induced desorption experiment at the CERN SPS.

Fig. 2 :
Fig. 2: Pictures of the inner part of one collimator chamber of the SPS desorption experiment.The aligned (Θ = 35 mrad) Cu/graphite collimator is shown in its "parking position", i.e., 18.5 mm below the beam axis.The ion beam has to pass the centered conductance (diameter 30 mm), which is made of a vacuum fired 100 µm thick 316 LN stainless steel foil.

Fig. 3 :
Fig. 3: Pressure rises of the Cu/graphite collimator bombarded under 35 mrad with 158 GeV/u In 49+ ions at the SPS.The intensity was ∼1.5 × 10 6 ions/spill with a spill length of 6.2 s. a) pressure rise when the sample was moved at 15:40 h into the ion beam, b) pressure rise during continuous bombardment, c) magnification of b).