Search for extended Higgs sectors in the $H \rightarrow hh$ and $A \rightarrow Zh$ channels in $\sqrt{s}=8$ TeV $pp$ collisions with multileptons and photons final states (HIG-13-025)

Abstract

Using a 19.5 fb$^{-1}$ data sample at $\sqrt{s} = 8$TeV collected by the CMS experiment at the LHC, a search for an extension of the Higgs sector to Two Higgs Doublet Models is presented. Decays $H \rightarrow hh$ and $A \rightarrow Zh$ of the heavy scalar and pseudo-scalar Higgs bosons, respectively, include the Standard Model-like Higgs h in the final state and lead to events with isolated leptons and photons.Observed multilepton events with or without diphoton candidates are organized into exclusive search channels based on event kinematics. The search channels are ordered by the amount of expected Standard Model background.Data-based estimation of the Standard Model backgrounds is emphasized, but data-validated simulations are also employed as appropriate. Observations are consistent with the Standard Model and lead to the exclusion of significant portions of Two Higgs Doublet Model parameter space presented in terms of the relevant parameters of the model.

Approved tables and plots (click on plot to get larger version)

Material in PAS

Figures

Figure	Caption
	Figure 1: Observed and expected limits with 1 and 2- $\sigma$ bands for $H\rightarrow hh$ in terms of $\sigma * BR$. These limits are based on multilepton and diphoton channels. Branching ratios for $h$ are assumed to have Standard Model values. No contribution from gg$\rightarrow$A$\rightarrow$Zh is considered in this limit.
	Figure 2: Observed and expected limits with 1 and 2- $\sigma$ bands for $A\rightarrow Zh$ in terms of $\sigma * BR$. These limits are based on multilepton and diphoton channels. Branching ratios for $h$ are assumed to have Standard Model values. No contribution from gg$\rightarrow$H$\rightarrow$hh is considered in this limit.
	Figure 3: Observed and expected limits with 1 and 2- $\sigma$ bands for $H\rightarrow hh$ in terms of $\sigma * BR$. These limits are based only on multilepton channels. Branching ratios for $h$ are assumed to have Standard Model values. No contribution from gg$\rightarrow$A$\rightarrow$Zh is considered in this limit.
	Figure 4: Observed and expected limits with 1 and 2- $\sigma$ bands for $A\rightarrow Zh$ in terms of $\sigma * BR$. These limits are based only on multilepton channels. Branching ratios for $h$ are assumed to have Standard Model values. No contribution from gg$\rightarrow$H$\rightarrow$hh is considered in this limit.
	Figure 5: Observed and expected limits on Heavy higgs of mass 300 GeV in Type I 2HDM. The parameters $\alpha$ an $\beta$ determine the cross section for $H$ production, the branching ratio ${\rm Br}(H \to hh)$ and the branching ratios ${\rm Br}(h \to WW, ZZ, \tau \tau, \gamma \gamma)$. Region below the observed limit and within the observed limit loop is excluded.
	Figure 6: Observed and expected limits on Heavy higgs of mass 300 GeV in Type II 2HDMs. The parameters $\alpha$ an $\beta$ determine the cross section for $H$ production, the branching ratio ${\rm Br}(H \to hh)$ and the branching ratios ${\rm Br}(h \to WW, ZZ, \tau \tau, \gamma \gamma)$. Region below the observed limit and within the observed limit loop is excluded.
	Figure 7: Observed and expected limits on $A$ of mass 300 GeV in Type I 2HDMs .The parameters $\alpha$ an $\beta$ determine the cross section for $H$ production, the branching ratio ${\rm Br}(A \to Zh)$ and the branching ratios ${\rm Br}(h \to WW, ZZ, \tau \tau, \gamma \gamma)$. Region below the observed limit is excluded.
	Figure 8: Observed and expected limits on$A$ of mass 300 GeV in Type II 2HDMs. The parameters $\alpha$ an $\beta$ determine the cross section for $H$ production, the branching ratio ${\rm Br}(A \to Zh)$ and the branching ratios ${\rm Br}(h \to WW, ZZ, \tau \tau, \gamma \gamma)$. Region below the observed limit is excluded.
	Figure 9: Observed and expected limits with 1 and 2- $\sigma$ bands on combined signal for Heavy Higgs and $A$ inType I 2HDMs ($m_{H}$ = $m_{A}$ = 300 GeV). The parameters $\alpha$ and $\beta$ determine the cross section for $H$ and $A$ production, the branching ratio Br(H $\rightarrow$ hh) and Br(A $\rightarrow$ Zh) and the branching ratios Br(h $\rightarrow$ WW, ZZ, $\tau \tau$, $\gamma \gamma$).Region below the observed limit is excluded.
	Figure 10: Observed and expected limits with 1 and 2- $\sigma$ bands on combined signal for Heavy Higgs and $A$ inType II 2HDMs ($m_{H}$ = $m_{A}$ = 300 GeV). The parameters $\alpha$ an $\beta$ determine the cross section for $H$ and $A$ production, the branching ratio Br(H $\rightarrow$ hh) and Br(A $\rightarrow$ Zh) and the branching ratios Br(h $\rightarrow$ WW, ZZ, $\tau \tau$, $\gamma \gamma$). Region below the observed limit is excluded.
	Figure 11: Invariant mass distribution of photons in channel with up to two (i.e. one or two) hadronic taus and two photons, 0-30 GeV MET.
	Figure 12: Invariant mass distribution of photons in channel with one lepton and two photons, 30-50 GeV MET bin. The blue fit is the one taken from the upto 2 tau+2photon channel and the red fit corresponds to an independent fit for this particular channel.
	Figure 13: Invariant mass distribution of photons in the channel with two leptons and two photons (0-30 GeV MET bin). The blue fit is the one taken from the (upto) 2tau + 2photon channel. The red line represents a fit excluding the zero bins and is for reference only, it is not used in the analysis.

	Figure 14: Background breakdown vs MET for 4 lepton + OSSF1 off-Z + no hadronic Tau + no b-jet.Here, background and signal are stacked.
	Figure 15: Background breakdown vs MET for 4 lepton + OSSF1 off-Z + 1 hadronic Tau + no b-jet. Here, background and signal are stacked.
	Figure 16: Background breakdown vs MET for 4 lepton + OSSF2 off-Z + no hadronic Tau + no b-jet.Here, background and signal are stacked.
	Figure 17: Background breakdown vs MET for 4 lepton + OSSF2 off-Z + no hadronic Tau + at least 1 b-jet. Here, background and signal are stacked.
	Figure 18: Background breakdown vs MET for 4 lepton + OSSF1 on-Z + 1 hadronic Tau + at least 1 b-jet . Here, background and signal are stacked.
	Figure 19: Background breakdown vs MET for 4 lepton + OSSF2 on-Z + no hadronic Tau + at least 1 b-jet. Here, background and signal are stacked.
	Figure 20: Background breakdown vs MET for 4 lepton + OSSF2 on-Z + no hadronic Tau + no b-jet. Here, background and signal are stacked.
	Figure 21: Background breakdown vs MET for 4 lepton + OSSF1 on-Z + 1 hadronic Tau + no b-jet. Here, background and signal are stacked.

Tables

Table	Caption
	Table 1: Observed yields for four lepton events from 19.5 $fb^{-1}$ data recorded in 2012. The channels are broken down by the number of and mass of any opposite-sign, same-flavor pairs (whether on or off Z), whether there are any b-jets present and the $E_{T}^{miss}$. Expected yields are the sum of simulation and data-driven estimates of backgrounds in each channel. The channels are exclusive.
	Table 2: Observed yields for three lepton events from 19.5 $fb^{-1}$ data recorded in 2012. The channels are broken down by the number of and mass of any opposite-sign, same-flavor pairs (whether on or off Z), whether there are any b-jets present and the $E_{T}^{miss}$. Expected yields are the sum of simulation and data-driven estimates of backgrounds in each channel. The channels are exclusive.
	Table 3: Observed yields for two lepton and two photon events from 19.5 $fb^{-1}$ data recorded in 2012. The channels are broken down the number of and mass of any opposite-sign, same-flavor pairs (whether on or off Z), and the $E_{T}^{miss}$. Only channels where invariant mass of photons lies in the higgs mass window (120-130 GeV) are considered. Expected yields are data-driven estimates of backgrounds in each channel. The channels are exclusive.
	Table 4: Observed yields for one lepton and diphoton events from 19.5 $fb^{-1}$ data recorded in 2012. The channels are broken down in bins of $E_{T}^{miss}$. There are no hadronic taus in these channels. Only channels where invariant mass of photons lies in the higgs mass window (120-130 GeV) are considered. Expected yields are data-driven estimates of backgrounds in each channel. The channels are exclusive.
	Table 5: Observed yields for up to two hadronic taus plus diphoton events from 19.5 $fb^{-1}$ data recorded in 2012. The channels are broken down in bins of $E_{T}^{miss}$. Only channels where invariant mass of photons lies in the higgs mass window (120-130 GeV)are considered. Expected yields are data-driven estimates of backgrounds in each channel. The channels are exclusive.
	Table 6: Observed yields for one lepton, one hadronic tau plus diphoton events from 19.5 $fb^{-1}$ recorded in 2012. The channels are broken down in bins of $E_{T}^{miss}$. Only channels where invariant mass of photons lies in the higgs mass window (120-130 GeV)are considered. Expected yields are data-driven estimates of backgrounds in each channel. The channels are exclusive.
	Table 7: The systematic uncertainties associated with this analysis.
	Table 7: This table shows the various decay modes of $h$. The combination of these decays considered for the analysis are marked with ``$\checkmark$" and those not considered for the analysis are marked with a ``X"
	Table 8: This table show the various decay modes of $h$ and Z boson. The combination of these decays considered for the analysis are marked with ``$\checkmark$" and those not considered for the analysis are marked with a ``X".
	Table 9: Various combinations of decay modes of two SM-like higgs and how they populate the search channels.
	Table 10: Various combinations of decay modes of Z boson and SM-like higgs and how they populate the search channels.
	Table 11: Couplings of the neutral Higgs bosons to SM fermions and massive gauge bosons as a function of $\alpha$ and $\beta$. This table has be taken from theory paper by Nathaniel Craig et al.(http://arxiv.org/abs/1207.4835). (Note: This is an erratum from the public PAS due to typos in this table in the PAS.)

Additional Material for HIG-13-025

Figure	Caption
	Figure 1: $\sigma$*BR (H$\rightarrow$hh) contours for TYPE I 2HDM.Parameters $\alpha$ and $\tan\beta$ give Heavy higgs's couplings to SM fermions and massive gauge bosons. This figure is similar to one from theory paper by Nathaniel Craig et al. (http://arxiv.org/abs/1305.2424), the only difference is plotting of $\tan\beta$, instead of $\beta$, on the vertical axis.
	Figure 2: $\sigma$*BR (H$\rightarrow$hh) contours for TYPE II 2HDM. Parameters $\alpha$ and $\tan\beta$ give Heavy higgs's couplings to SM fermions and massive gauge bosons. This figure is similar to one from theory paper by Nathaniel Craig et al. (http://arxiv.org/abs/1305.2424), the only difference is plotting of $\tan\beta$, instead of $\beta$, on the vertical axis.
	Figure 3: $\sigma$*BR (A$\rightarrow$Zh) contours for TYPE I 2HDM. Parameters $\alpha$ and $\tan\beta$ give A particle's couplings to SM fermions and massive gauge bosons. This figure is similar to one from theory paper by Nathaniel Craig et al. (http://arxiv.org/abs/1305.2424), the only difference is plotting of $\tan\beta$, instead of $\beta$, on the vertical axis.
	Figure 4: $\sigma$*BR (A$\rightarrow$Zh) contours for TYPE II 2HDM. Parameters $\alpha$ and $\tan\beta$ give A particle's couplings to SM fermions and massive gauge bosons. This figure is similar to one from theory paper by Nathaniel Craig et al. (http://arxiv.org/abs/1305.2424), the only difference is plotting of $\tan\beta$, instead of $\beta$, on the vertical axis.
	Figure 5: Ratio of SM-like higgs branching ratio in TYPE I 2HDM to SM higgs branching ratio for h$\rightarrow$WW.This figure is similar to one from theory paper by Nathaniel Craig et al. (http://arxiv.org/abs/1305.2424), the only difference is plotting of $\tan\beta$, instead of $\beta$, on the vertical axis.
	Figure 6: Ratio of SM-like higgs branching ratio in TYPE I 2HDM to SM higgs branching ratio for h$\rightarrow$ZZ. This figure is similar to one from theory paper by Nathaniel Craig et al. (http://arxiv.org/abs/1305.2424), the only difference is plotting of $\tan\beta$, instead of $\beta$, on the vertical axis.
	Figure 7: Ratio of SM-like higgs branching ratio in TYPE I 2HDM to SM higgs branching ratio for h$\rightarrow \tau\tau$. This figure is similar to one from theory paper by Nathaniel Craig et al. (http://arxiv.org/abs/1305.2424), the only difference is plotting of $\tan\beta$, instead of $\beta$, on the vertical axis.
	Figure 8: Ratio of SM-like higgs branching ratio in TYPE I 2HDM to SM higgs branching ratio for h$\rightarrow \gamma\gamma$. This figure is similar to one from theory paper by Nathaniel Craig et al. (http://arxiv.org/abs/1305.2424), the only difference is plotting of $\tan\beta$, instead of $\beta$, on the vertical axis.
	Figure 9: Ratio of SM-like higgs branching ratio in TYPE I 2HDM to SM higgs branching ratio for h$\rightarrow$bb. This figure is similar to one from theory paper by Nathaniel Craig et al. (http://arxiv.org/abs/1305.2424), the only difference is plotting of $\tan\beta$, instead of $\beta$, on the vertical axis.
	Figure 10: Ratio of SM-like higgs branching ratio in TYPE II 2HDM to SM higgs branching ratio for h$\rightarrow$WW. This figure is similar to one from theory paper by Nathaniel Craig et al. (http://arxiv.org/abs/1305.2424), the only difference is plotting of $\tan\beta$, instead of $\beta$, on the vertical axis.
	Figure 11: Ratio of SM-like higgs branching ratio in TYPE II 2HDM to SM higgs branching ratio for h$\rightarrow$ZZ. This figure is similar to one from theory paper by Nathaniel Craig et al. (http://arxiv.org/abs/1305.2424), the only difference is plotting of $\tan\beta$, instead of $\beta$, on the vertical axis.
	Figure 12: Ratio of SM-like higgs branching ratio in TYPE II 2HDM to SM higgs branching ratio for h$\rightarrow \tau\tau$. This figure is similar to one from theory paper by Nathaniel Craig et al. (http://arxiv.org/abs/1305.2424), the only difference is plotting of $\tan\beta$, instead of $\beta$, on the vertical axis.
	Figure 13: Ratio of SM-like higgs branching ratio in TYPE II 2HDM to SM higgs branching ratio for h$\rightarrow \gamma\gamma$. This figure is similar to one from theory paper by Nathaniel Craig et al. (http://arxiv.org/abs/1305.2424), the only difference is plotting of $\tan\beta$, instead of $\beta$, on the vertical axis.
	Figure 14: Ratio of SM-like higgs branching ratio in TYPE II 2HDM to SM higgs branching ratio for h$\rightarrow$bb. This figure is similar to one from theory paper by Nathaniel Craig et al. (http://arxiv.org/abs/1305.2424), the only difference is plotting of $\tan\beta$, instead of $\beta$, on the vertical axis.
	Figure 15: Background breakdown vs MET for 3 lepton + OSSF1 below-Z + 1 hadronic Tau + no b-jet. Here, background and signal are stacked.
	Figure 16: Background breakdown vs MET for 3 lepton + OSSF1 below-Z + 1 hadronic Tau + at least 1 b-jet. Here, background and signal are stacked.
	Figure 17: Background breakdown vs MET for 3 lepton + no OSSF + 1 hadronic Tau + no b-jet. Here, background and signal are stacked.
	Figure 18: Background breakdown vs MET for 3 lepton + no OSSF + no hadronic Tau + at least 1 b-jet. Here, background and signal are stacked.
	Figure 19: Background breakdown vs MET for 3 lepton + OSSF1 below-Z + no hadronic Tau + no b-jet. Here, background and signal are stacked.
	Figure 20: Background breakdown vs MET for 3 lepton + OSSF1 below-Z + no hadronic Tau + at least 1 b-jet. Here, background and signal are stacked.
	Figure 21: Background breakdown vs MET for 4 lepton + OSSF2 Two Z + no hadronic Tau + at least 1 b-jet. Here, background and signal are stacked.
	Figure 22: Background breakdown vs MET for 4 lepton + OSSF2 TwoZ + no hadronic Tau + no b-jet. Here, background and signal are stacked.
	Figure 23: Background breakdown vs MET for 3 lepton + OSSF1 on-Z + 1 hadronic Tau + no b-jet. Here, background and signal are stacked.
	Figure 24: Background breakdown vs MET for 3 lepton + OSSF1 on-Z + 1 hadronic Tau + at least 1 b-jet. Here, background and signal are stacked.
	Figure 25: 95$\%$ C.L. upper limits on $\sigma * BR$ for $H\rightarrow hh$ and $A\rightarrow Zh$.

Additional Material from SUS-13-002 and SUS-13-003 (Links to Public Results)

Figure	Caption
	Figure 1: The $S_\text{T}$ distribution in an opposite sign e$\textrm{$\mu$}$ $\textrm{t}\bar{\textrm{t}}$ control region. The uncertainties in the ratio plot below include both statistical and systematic uncertainties.
	Figure 2: Distribution for H$_{\text{T}}$ in the opposite sign e$\textrm{$\mu$}$ dilepton $\textrm{t}\bar{\textrm{t}}$ control region.
	Figure 3: Distributions for E$^{miss}_{\text{T}}$ in the opposite sign e$\textrm{$\mu$}$ dilepton $\textrm{t}\bar{\textrm{t}}$ control region.
	Figure 5: Distributions for E$^{miss}_{\text{T}}$ in the WZ control region.
	Figure 6: 4-lepton mass distribution for low-E$^{miss}_{\text{T}}$, low-H$_{\text{T}}$ ZZ control region.
	Figure 7: 3-muon invariant mass showing asymmetric internal photon conversion.
	Figure 8: Efficiency ratio vs Rd$_{\textrm{xy}}$ ("b-ness of events") for muons and electrons.
	Figure 9: f$_\textrm{t}$-f$_\textrm{sb}$ for taus with p$_{\textrm{T}}$ between 40 and 60 GeV (f$_\textrm{t}$ is the fake-rate for taus and f$_\textrm{sb}$ is inversely proportional to jet activity).
	Figure 10: f$_\textrm{t}$-f$_\textrm{sb}$ for taus with p$_{\textrm{T}}$ between 20 and 40 GeV (f$_\textrm{t}$ is the fake-rate for taus and f$_\textrm{sb}$ is inversely proportional to jet activity).

Signal MC Generation

For signal, all LHE files are made using MADGRAPH 4.4.5 and then decayed and showered through PYTHIA 6 in CMSSW532patch4. We simulate gluon fusion production of a Heavy Higgs boson decaying to a pair of SM-like Higgs bosons, $gg \to H \to hh$ for $H$ mass ranging from 260 GeV to 360 GeV. Each $h$ can decay to $WW$, $ZZ$, $\tau\tau$, $bb$ and $\gamma\gamma$ ($h\rightarrow WW$, $h\rightarrow ZZ$, $h\rightarrow \tau\tau$, $h\rightarrow bb$, $h\rightarrow \gamma\gamma$) , for example, as shown in figure below. We generate a separate LHE file for each combination of $h$ decay modes, for each mass point of $H$ .

Figure 11: Heavy higgs decays to 2 SM higgs. One of the SM higgs decays to WW, other to ZZ

We also simulate gluon fusion production of a heavy pseudo-scalar Higgs boson decaying to a SM-like Higgs boson and $Z$-boson, $gg \to A \to Zh$, for masses of $A$ between 260 GeV to 360 GeV. Z-boson can decay leptonically or otherwise and $h$ can decay to $WW$, $ZZ$, $\tau\tau$ or $\gamma\gamma$.

Each production and decay topology is generated exclusively, and combined according to the $H$, $A$, and $h$ branching ratios.These signal samples are generated model independently. To interpret the results in Two Higgs Doublet Model (2HDM) scenario we use cross sections obtained from the SusHi (http://arxiv.org/abs/1212.3249) program and branching ratios for SM-like Higgs calculated using 2HDMC (http://arxiv.org/abs/0902.0851).