A report from the CMS experiment
New ways to detect long-lived particles (LLPs) are opening up avenues for searching for physics beyond the Standard Model (SM). LLPs could provide evidence for a hidden dark sector of particles that includes dark-matter candidates and could be studied via “portal interactions” with the visible universe. By employing the CMS experiment’s muon spectrometer in a novel way, the collaboration has recently deployed a powerful new technique for detecting LLPs that decay between 6 and 10 metres from the primary interaction point.
An LLP decaying in the endcap muon spectrometer volume should produce a particle shower when its decay products interact with the return yoke of the CMS solenoid. The secondary particles produced by the shower would traverse the gaseous regions of the cathode-strip chamber (CSC) detector and produce a large multiplicity of signals on the wire anodes and strip cathodes. Localised hits are reconstructed by combining these signals using a density-based clustering algorithm. This is the first time the CSC detectors have been used as a sampling calorimeter to try to detect and identify LLP decays.
Searching for CSC clusters with a sufficiently large number of hits suppresses background processes while maintaining a high efficiency for detecting potential LLP decays. The large amount of steel in the CMS return yoke nearly eliminates “punch-through” hadrons that are not fully stopped by the calorimeter, potentially mimicking the signature of an LLP. The largest remaining source of backgrounds is known LLPs produced by SM processes such as the neutral kaon, KL. These particles are copiously produced in LHC collisions and, on rare occasions, traverse the material without being stopped. Kaons are predominantly produced with much lower energies than the signal LLPs and therefore result in clusters with a smaller number of hits. Requiring clusters with more than 130 CSC hits suppresses these dominant background events to a negligible level (see figure 1).
This search improves on the previous best results by more than a factor of six
Using the full Run-2 dataset, the CMS collaboration detected no excess of particle-shower events above the expected backgrounds, setting constraints on a benchmark-simplified model of scalar LLP production mediated by the Higgs boson (a so-called Higgs portal model). This search improves on the previous best results by more than a factor of six (two) for an LLP mass of 7 GeV (≥ 15) GeV for a proper decay length (cτ) of the scalar larger than 100 m. It is the first to be sensitive to LLP decays with cτ up to 1000 m and masses between 40 and 55 GeV at branching ratios of the Higgs to a pair of LLPs below 20%.
This novel approach to identifying showers in muon detectors opens up an exciting new programme of searches for LLPs in a wide variety of theoretical models. Potential frameworks range from Higgs-portal models to other portals to a dark sector, including neutrinos, axions and dark photons. The on-going development of a dedicated Level-1 and High-Level Trigger focusing on particle showers detected in the CMS muon spectrometer promises an order of magnitude improvement in the discovery sensitivity for LLPs in the forthcoming run of the LHC.
CMS Collaboration 2021 arXiv:2107.04838.