Search for new physics in the $\mu \mu +e/\mu +{\mathrm{E̸}}_T$ channel with a low-$p_T$ lepton threshold at the Collider Detector at Fermilab

A search for new physics using three-lepton (trilepton) data collected with the CDF II detector and corresponding to an integrated luminosity of 976 ${\mathrm{pb}}^{-1}$ is presented. The standard model predicts a low rate of trilepton events, which makes some supersymmetric processes, such as chargino-neutralino production, measurable in this channel. The $\mu \mu +\ell$ signature is investigated, where $\ell$ is an electron or a muon, with the additional requirement of large missing transverse energy. In this analysis, the lepton transverse momenta with respect to the beam direction ($p_T$) are as low as $5\mathrm{GeV}/c$, a selection that improves the sensitivity to particles that are light as well as to ones that result in leptonically decaying tau leptons. At the same time, this low-$p_T$ selection presents additional challenges due to the non-negligible heavy-quark background at low lepton momenta. This background is measured with an innovative technique using experimental data. Several dimuon and trilepton control regions are investigated, and good agreement between experimental results and standard-model predictions is observed. In the signal region, we observe one three-muon event and expect $0.4\pm 0.1\mu \mu +\ell$ events from standard-model processes.

Figure 1

Chargino-neutralino production through an $s$-channel $W$ boson (a) and a $t$-channel squark propagator (b). The $t$ channel is suppressed in scenarios with very massive squarks.

Figure 2

Chargino and neutralino decays through gauge bosons (a,b) and through sleptons (c,d). The branching fractions depend on the masses of the sleptons, which always decay to charged leptons, unlike the gauge bosons. The leptonic signature consists of three leptons in both cases.

Figure 3

Fit of $\mathrm{HF}+\mathrm{DY}+\mathrm{\text{fakes}}$ dimuon mass distribution to the observed data for opposite-charge dimuons. The HF normalization is the only free parameter of the fit. The blue (light gray) filled histogram is the HF and the red (dark gray) is the fakes. The thick line represents the total background, which is almost exclusively DY, HF, and fakes at the opposite-charge dimuon level. The hatched areas indicate the total uncertainty on the prediction.

Figure 4

Fit of $\mathrm{HF}+\mathrm{DY}+\mathrm{\text{fakes}}$ dimuon mass distribution to the observed data for same-charge dimuons. The HF normalization is the only free parameter of the fit. The blue (light gray) filled histogram is the HF and the red (dark gray) is the fakes. The thick line represents the total background, which is constituted almost exclusively by HF and fakes for same-charge dimuon pairs. The hatched areas indicate the total uncertainty on the prediction.

Figure 5

The control and signal regions used in our analysis are defined in the dimuon mass vs ${\mathrm{E̸}}_T$ plane, with the extra requirement of low ($\le 1$) or high ($>1$) jet multiplicity. In this paper we show results for the control regions that result from the inversion of one of the three main kinematic selections (dimuon mass, missing transverse energy, and jet multiplicity), with the addition of a $Z$-mass control region (Control z) and a low-${\mathrm{E̸}}_T$ control region (Control a). The control regions above are defined for low jet multiplicity, unless otherwise stated.

Figure 6

Dimuon mass and ${\mathrm{E̸}}_T$ distributions for the SM background in the dimuon control regions z (a,b), a (c,d), and b (e,f). The background histograms are stacked. The CDF data are indicated by points with error bars.

Figure 7

Dimuon mass and ${\mathrm{E̸}}_T$ for the SM background in the dimuon control regions c (a,b), d (c,d), and dimuon signal region (e,f). The background histograms are stacked. The CDF data are indicated by points with error bars.

Figure 8

Kinematic variables for the SM background and the SIG2 SUSY signal, in the trilepton signal region. The background histograms are stacked; the signal histogram is not. The CDF data are indicated by points with error bars.

Figure 9

The trimuon event in the transverse view of the central CDF detector. Tracks with transverse momenta above $1\mathrm{GeV}/c$ are shown.

Figure 10

The trimuon event in the $\eta -\varphi$ view. Calorimeter transverse energies above 1 GeV are shown. The longer bars correspond to the track momenta of the three muons (high $\varphi$). The calorimeter energy depositions $E_1$ and $E_2$ are mainly electromagnetic and could be associated with two photons.