Following the recommendation of the C&SR 2019, crystal collimation was successfully integrated in the WP5 upgrade baseline as an in-kind item to mitigate schedule issues with the 11 T dipoles which have shown potential reliability issues for long-term operation in the accelerator. A detailed investigation program is currently underway to understand the root cause of these issues and to define an updated strategy for the 11 T production program by the end of 2022. Following the decision in 2020 to defer the installation of the 11 T dipoles until after LS2, it is now planned to use crystal primary collimators (TCPC) for Pb ion cleaning in Run3. Two TCPCs are currently under construction at CERN for installation before the start of Run3 to replace two units in IR7 with performance issues. Two more units, planned to be provided through an in-kind collaboration with PNPI and IHEP, are foreseen for installation at the end of 2022 to complete the upgrade of the present system. The work at CERN advanced well during the first half of 2021 and some components are already available (see Figure 1). This will guarantee a new crystal-based collimation system for the Run3 operation with ions.

This innovative scheme for beam collimation uses high-purity silicon crystals, mechanically bent to a desired angle, instead of standard primary collimators. In specific conditions, positively-charged hadron particles that impinge on the crystal remain trapped in the potential well of adjacent lattice planes – this phenomenon is referred to as planar channeling – and experience a kick equal to the crystal bending. Typically, 50 urad are needed for collimation at the HL-LHC, achieved with only 4 mm-long crystals bent to a curvature radius of 80 m: this is equivalent to magnetic fields of about 300 T! A total of at least four crystals is needed for horizontal and vertical collimation of each beam.

The effect on the beam halo is shown in Figure 2: with crystal primary collimators, halo particles (green line) are directly steered onto existing secondary collimators (blue) that serve as halo absorbers. The process is “cleaner” than the interaction with standard primary collimators and results in lower dispersive losses in the dispersion suppressor (DS) downstream of IR7. In this sense, crystal collimation competes for Pb ion operation with the DS cleaning upgrade based on the installation of TCLD collimators: in the latter, losses are cured locally making space for a collimator with 11 T dipoles, while crystals reduce the overall dispersive leakage outside of IR7 for Pb ion beam operation.

Crystal collimation has been part of the WP5 R&D studies since 2014, following important preliminary work done by the UA9 collaboration. A complete crystal collimation test stand was installed in IR7 to study the halo-cleaning performance at high energy, in particular for heavy-ion beams. Beam tests in Run2 showed that the cleaning performance in the DS improves by more than a factor 3. This performance is not achieved with all crystals, and, in addition, the present setup is not designed for long-term reliable operation. Therefore, the described upgrade during LS2 is needed to ensure the efficiency of this mitigation measure and allow Pb ion operation at intensities with the nominal HL-LHC Pb parameters.