Data-driven Search for Long-Lived Particles via Displaced Vertices at CMS
We report on a search for pairs of displaced vertices that is sensitive to long-lived particles (LLPs) in a low-background environment. This search uses 137 fb$^{-1}$ of $\sqrt{s}$ = 13 TeV data collected by the CMS experiment and targets LLPs with mean proper decay lengths between 0.1 and 30 mm. Benchmark models are extended from the previous iteration to include exotic decays of the 125-GeV Higgs boson to a pair of LLPs, which decay hadronically to jets. The previous trigger targeted mass points of LLPs above 800 GeV and was unable to capture the sensitivity of this model. In this search, the same RPV SUSY models are probed as low as 200 GeV. Moreover, data-driven techniques from reconstructing displaced vertices to quantifying associated systematic uncertainties have been improved. To study vertexing efficiency, we employ a novel technique called "TrackMover," in which tracks are artificially displaced in data and simulated SM events to mimic the kinematics and topology of LLP decays. Low-mass LLPs decay to a pair of two soft quarks, posing an issue of no associated jets. Despite the challenge in certain models, this technique has been generalized for all our benchmark models. It could also be adapted to other CMS searches for displaced vertices. While the exclusion limit remains blinded, the displaced-vertex search is expected to continue providing complementary exclusion limits across models to other searches for LLPs with longer lifetimes.