Search for supersymmetry in final states with disappearing tracks in proton-proton collisions at $\sqrt{s}=13\mathrm{TeV}$

A search is presented for charged, long-lived supersymmetric particles in final states with one or more disappearing tracks. The search is based on data from proton-proton collisions at a center-of-mass energy of 13 TeV collected with the CMS detector at the CERN LHC between 2016 and 2018, corresponding to an integrated luminosity of $137{\mathrm{fb}}^{-1}$. The search is performed over final states characterized by varying numbers of jets, $b$-tagged jets, electrons, and muons. The length of signal-candidate tracks in the plane perpendicular to the beam axis is used to characterize the lifetimes of wino- and Higgsino-like charginos produced in the context of the minimal supersymmetric standard model. The $\mathrm{d}E/\mathrm{d}x$ energy loss of signal-candidate tracks is used to increase the sensitivity to charginos with a large mass and thus a small Lorentz boost. The observed results are found to be statistically consistent with the background-only hypothesis. Limits on the pair-production cross section of gluinos and squarks are presented in the framework of simplified models of supersymmetric particle production and decay, and for electroweakino production based on models of wino and Higgsino dark matter. The limits presented are the most stringent to date for scenarios with light third-generation squarks and a wino- or Higgsino-like dark matter candidate capable of explaining the observed dark matter relic density.

Figure 1

Representative diagrams for the simplified models considered in this analysis. From left to right: T6btLL, T6tbLL, and T5btbtLL (upper); and TChiWZ, TChiWW, and TChiW (lower). The shaded circles at the production vertices represent a sum over perturbative terms.

Figure 2

The distributions of simulated events used to train and validate the BDT classifiers. The left (right) column corresponds to the phase-0 (phase-1) detector and the upper (lower) row to the short-(long) track category. The uncertainty bars shown for the training samples indicate the Poisson uncertainties. No events appear outside of the regions shown.

Figure 3

Comparison of the $p_{\mathrm{T}}$ distributions of DTks in the ${\kappa }_{\mathrm{high}}^{\mathrm{low}}$ DY measurement control region for the data and background prediction for long (upper) and short (middle) showering tracks and in the ${\kappa }_{\mu \text{match}}^{\mu \mathrm{veto}}$ DY measurement control region for long muon tracks (lower). The left (right) column corresponds to the phase-0 (phase-1) detector. The uncertainty bars on the ratios in the lower panels indicate the fractional Poisson uncertainties in the observed counts. The gray bands show the fractional Poisson uncertainties in the sideband region counts, added in quadrature with the systematic uncertainties. The leftmost (rightmost) bin includes underflow (overflow).

Figure 4

Comparison of the $p_{\mathrm{T}}$ distributions of DTks in the high-$m_{\mathrm{T}}$ validation region for the data and background prediction for long (upper) and short (middle) showering tracks and for long muon tracks (lower). The left (right) column corresponds to the phase-0 (phase-1) detector. The uncertainty bars on the ratios in the lower panels indicate the fractional Poisson uncertainties in the observed counts. The gray bands show the fractional Poisson uncertainties in the sideband region counts, added in quadrature with the systematic uncertainties. The leftmost (rightmost) bin includes underflow (overflow).

Figure 5

Comparison of the $p_{\mathrm{T}}$ distributions of DTks in the ${\theta }_{\mathrm{low}}^{\mathrm{high}}$ QCD measurement control region for the data and background prediction for long (upper) and short (lower) tracks. The left (right) column corresponds to the phase-0 (phase-1) detector. The uncertainty bars on the ratios in the lower panels indicate the fractional Poisson uncertainties in the observed counts. The gray bands show the fractional Poisson uncertainties in the control region counts, added in quadrature with the systematic uncertainties. The leftmost (rightmost) bin includes underflow (overflow).

Figure 6

Comparison of the $p_{\mathrm{T}}$ distributions of DTks in events with one electron and one DTk, in a validation region with $m_{\mathrm{T}}<110\mathrm{GeV}$ and $m_{\mathrm{DTk},\ell }\notin [65,110]\mathrm{GeV}$, for the data and background prediction for long (upper) and short (lower) tracks. The left (right) column corresponds to the phase-0 (phase-1) detector. The uncertainty bars on the ratios in the lower panels indicate the fractional Poisson uncertainties in the observed counts. The gray bands show the fractional Poisson uncertainties in the control region counts, added in quadrature with the systematic uncertainties. The leftmost (rightmost) bin includes underflow (overflow).

Figure 7

Comparison in the baseline region for the long-track DTk category between the data and prefit predicted SM background for the $N_{\text{jet}}$ (upper left), $N_{b\text{-jet}}$ (upper right), $p_{\mathrm{T},\mathrm{hard}}^{\mathrm{miss}}$ (middle left), $N_e$ (middle right), $N_{\mu }$ (lower left), and $m_{\text{DTk};\mathrm{d}E/\mathrm{d}x}$ (lower right) distributions. The uncertainty bars on the ratios in the lower panels indicate the fractional Poisson uncertainties in the observed counts. The gray bands show the fractional Poisson uncertainties in the control region counts, added in quadrature with the systematic uncertainties. The leftmost (rightmost) bin includes underflow (overflow). For purposes of illustration, results from the T6tbLL and T5btbtLL models are shown, where the first and second numbers in parentheses indicate the squark (or gluino) mass and the LSP mass, respectively, in GeV.

Figure 8

Comparison in the baseline region for the short-track DTk category between the data and prefit predicted SM background for the $N_{\text{jet}}$ (upper left), $N_{b\text{-jet}}$ (upper right), $p_{\mathrm{T},\mathrm{hard}}^{\mathrm{miss}}$ (middle left), $N_e$ (middle right), $N_{\mu }$ (lower left), and $m_{\text{DTk};\mathrm{d}E/\mathrm{d}x}$ (lower right) distributions. The uncertainty bars on the ratios in the lower panels indicate the fractional Poisson uncertainties in the observed counts. The gray bands show the fractional Poisson uncertainties in the control region counts, added in quadrature with the systematic uncertainties. The leftmost (rightmost) bin includes underflow (overflow). For purposes of illustration, results from the T6tbLL and T5btbtLL models are shown, where the first and second numbers in parentheses indicate the squark (or gluino) mass and the LSP mass, respectively, in GeV.

Figure 9

Comparison between the data and SM background predictions for the 49 search regions. The left (right) plot shows the prefit (postfit) background predictions. The uncertainty bars on the ratios in the lower panels, shown for bins with nonzero entries, indicate the fractional Poisson uncertainties in the observed counts. The gray bands show the fractional Poisson uncertainties in the control region counts, added in quadrature with the systematic uncertainties. For purposes of illustration, results from the T6tbLL and T5btbtLL models are shown, where the first and second numbers in parentheses indicate the squark (or gluino) mass and the LSP mass, respectively, in GeV.

Figure 10

Observed 95% C.L. upper limits on the signal cross sections (colored area) versus the bottom or top squark and neutralino mass for the T6tbLL (upper) and T6btLL (lower) model for a chargino proper decay length $c\tau$ of 10 (left column) or 200 (right column) cm. Also shown are black (red) contours corresponding to the observed (expected) lower limits, including their uncertainties, on the squark and neutralino masses.

Figure 11

Observed 95% C.L. upper limits on the signal cross sections (colored area) versus the gluino and neutralino mass for the T5btbtLL model for a chargino proper decay length $c\tau$ of 10 (left column) and 200 (right column) cm. Also shown are black (red) contours corresponding to the observed (expected) lower limits, including their uncertainties, on the gluino and neutralino masses.

Figure 12

Observed 95% C.L. upper limits on the signal cross sections (colored area) versus the chargino-LSP mass difference and the mass of the chargino for the wino (left) and Higgsino (right) DM models. The black contours indicate the boundary where the observed upper limit equals the cross section of fully degenerate electroweakino production. The corresponding expected limits are shown by the red contours. The green lines represent the set of model points corresponding to the pure wino and pure Higgsino models where only radiative corrections to the mass splitting are assumed. Chargino lifetimes are based on two-loop calculations.