Searches for supersymmetry at CMS in ﬁnal states with photons

Four searches for supersymmetry at CMS in ﬁnal states with photons are discussed in this article. The data analyzed corresponds to up to 19.7fb − 1 and was recorded at centre-of-mass energies of 7 and 8 TeV by the CMS experiment. Searches in ﬁnal states with photons are sensitive to various scenarios for physics beyond the standard model, especially gauge mediated supersymmetry breaking. To increase the sensitivity for various classes of models, four analyses with di ﬀ erent selections and background estimation methods are performed. Wino-like neutralino mixing scenarios are probed in ﬁnal states with at least one photon and jets, while ﬁnal states with at least two photons are most sensitive to bino-like neutralino mixing scenarios. A higgsino-like neutralino scenario is studied by selecting two photons with an invariant mass close to the Higgs mass. While all of these analyses explore events with high missing transverse energy, a fourth analysis covers also ﬁnal states with two photons and low missing transverse energy. No excess of data over the standard model expectation is observed, and upper limits on various signal cross sections are calculated.


Introduction
Searches for supersymmetry (SUSY) in final states with photons are especially sensitive to gauge mediated supersymmetry breaking (GMSB). As the lightest neutralino decays into a gravitino (G) and a standard model (SM) boson, the mixing of the lightest neutralino is important for the resulting final states. Three of the analyses presented here are optimized for the discovery of wino-, bino-, and higgsino-like neutralino scenarios, while the fourth analysis abandons the missing transverse energy ( E T ) requirement and is sensitive to various other scenarios for physics beyond the SM. The data is recorded by the CMS experiment [1].

Final states with one or two photons and E T
Events are selected with either a high energetic photon, hadronic activity, jets and E T , or with two photons Email address: kiesel@cern.ch (Maximilian Knut Kiesel) and E T [2]. The background of jets misidentified as photons and photon-jet production is estimated from a control sample with a photon-like jet instead of the photon. The ratio of selected events with photons to the control sample is propagated from events with low E T to the signal region with high E T . Kinematic dependencies of this ratio on photon p T or the recoil against the photons are considered.
To estimate the background from electrons misidentified as photons, events with electrons instead of photons are selected and scaled by the probability of an electron faking a photon. This probability is measured using the Z → ee resonance. For the single-photon analysis, initial and final state radiation is estimated using simulation. Figure 1 shows the event yields, background prediction and two signal scenarios for both signal selections.
No evidence for physics beyond the SM is found, and multiple statistically independent bins in E T are used to calculate limits at 95% confidence level using data Available online at www.sciencedirect.com

Final states with a Higgs boson and E T
Events containing two photons with an invariant mass close to the Higgs mass and two b-jets are selected [3]. To increase the sensitivity, these events are categorized with respect to having an additional b-jet, or the invariant mass of two b-jets being close to the Higgs mass. The background is estimated by a fit in the sidebands of the invariant mass spectrum of the di-photon system, as shown in Fig. 3.
The 95% confidence limits are calculated in bins of E T and combined for all event categories for higgsinolike neutralinos in the stop-higgsino mass plane. The cross section limit and the exclusion contour is shown in Fig. 4 for data corresponding to 19.7 fb −1 . Stop masses of 370 GeV and below are excluded.

Final states with two photons
A search without E T requirement can be sensitive to extra dimensions, heavy-flavor compositeness, little Higgs, R-parity violation, GMSB and stealth SUSY [4]. Events with two photons, at least four jets and the scalar sum S T of the transverse momenta of all jets, photons and E T larger than 700 GeV are selected. The signal region is seperated into events with exactly four jets and five or more jets. To estimate the background, two control regions are defined: A "jet multiplicity sideband" (JMSB) composed of events with two or three jets and S T > 600 GeV, and a "S T sideband" with four or more jets and 600 < S T < 700 GeV. Since the shape of the S T -spectrum was found to be independent of the jet multiplicity, the S T shape is taken from the JMSB and the normalization is taken from the S T sideband. The S T distribution, the background estimation, and a signal are shown in Fig. 5.
In Fig. 6, limits on stealth SUSY [5] cross sections as function of the squark mass are compared to the predicted stealth SUSY cross section for data corresponding to 4.96 fb −1 at a centre-of-mass energy of 7 TeV.  Figure 6: Stealth SUSY cross-section limit at the 95% CL as a function of Mq. The observed limit, median expected limit, and the prediction for a stealth SUSY model is shown.
Squark masses below 1400 GeV are excluded in this scenario.