LLP production processes considered in this paper, where the $\tilde{\chi}^{0}_{1}$ represents the LLP: (a) di-LLP production via a scalar particle $h^0$; (b) non-resonant, direct LLP production from quark interactions, where $X$ is a stable particle, with mass identical to the LLP. The LLP decays into a muon and two quarks: $\tilde{\chi}^{0}_{1} \rightarrow \mu^{+} q_{i} q_{j}( \mu^{-} \bar{q_{i}} \bar{q_{j}})$.

Distributions from data compared to simulated $ b \overline b $ events (blue) and the simulated signal h125-chi40-10p (red), after preselection. From (a) to (j): muon transverse momentum; muon impact parameter; muon isolation; the calorimetric energy, $\rm E_{calorimeters}$, associated with the muon normalised by the muon energy, $\rm E_{muon}$; the number of tracks used to reconstruct the LLP vertex including the muon; the radial distance to the beam line of the reconstructed vertex; longitudinal and radial vertex fit errors, $\sigma_Z$ and $\sigma_R$; reconstructed transverse momentum and mass of the LLP candidate. The distributions from simulated events are normalised to the data.

Reconstructed invariant mass of the LLP candidates. Subfigures (a), (c), and (e) correspond to the signal selections which assume the models h80-chi30-10ps, h200-chi20-10ps, and the non-resonant model chi30-10ps, respectively. Subfigures (b), (d), and (f) are the corresponding distributions for candidates selected in the background region. The results of the fits are superimposed.

Expected (open dots and 1$\sigma$ and 2$\sigma$ bands) and observed (full dots) cross-section times branching fraction upper limits (95% CL) as a function of m$_{\tilde{\chi}^{0}_{1} }$ for the resonant production processes with $ m_{h^0} =125\text{ Ge V /}c^2 $, and, from (a) to (g), $\tau_{\tilde{\chi}^{0}_{1} }$ of 5, 10, 20, 30, 50, 100, and 200 $\text{ ps}$ .

Expected (open dots and 1$\sigma$ and 2$\sigma$ bands) and observed (full dots) cross-section times branching fraction upper limits (95% CL) as a function of $\tau_{\tilde{\chi}^{0}_{1} }$ for the resonant production with $ m_{h^0} =125\text{ Ge V /}c^2 $, and, from (a) to (e), $m_{\tilde{\chi}^{0}_{1} }$ of 20, 30, 40, 50, and 60 $\text{ Ge V /}c^2$ .

Expected (open dots and 1$\sigma$ and 2$\sigma$ bands) and observed (full dots) cross-section times branching fraction upper limits (95% CL) as a function of $ m_{h^0}$ and, from from (a) to (f), $m_{\tilde{\chi}^{0}_{1} }$ of 10, 20, 30, 40, 50, and 60 $\text{ Ge V /}c^2$ .

Expected (open dots and 1$\sigma$ and 2$\sigma$ bands) and observed (full dots) cross-section times branching fraction upper limits (95% CL), (a): as a function of $\tau_{\tilde{\chi}^{0}_{1} }$ with $m_{\tilde{\chi}^{0}_{1} }=30\text{ Ge V /}c^2 $, (b): as a function of $m_{\tilde{\chi}^{0}_{1} }$ with $\tau_{\tilde{\chi}^{0}_{1} }=10\text{ ps} $. The processes are from direct, non-resonant, LLP production.

Animated gif made out of all figures.

Signal detection efficiency, in percent, after preselection, $\varepsilon _{presel}$, including the geometrical acceptance, and after MVA selection, $\varepsilon $, and numbers of fitted signal and background events in the signal region, $N_{\rm s}$ and $N_{\rm b}$, for the different signal hypotheses for the LLP non-resonant production models (masses units are $\text{ Ge V /}c^2$ , lifetimes in $\text{ ps}$ ). The last column gives the value of $\chi^2$ per degree of freedom, ndf, from the fit.

Contributions to the relative systematic uncertainties. The indicated ranges cover the theoretical models considered. The contributions from the signal and background models used in the LLP mass fit are treated separately.

Upper limits at 95% CL on the production cross-section times branching ratio for signal models with a non-resonant production. Masses are given in $\text{ Ge V /}c^2$ , lifetimes in $\text{ ps}$ , cross-sections in $\text{ pb}$ .

Supplementary material full pdf