Searching for additional Higgs bosons at the CERN Large Hadron Collider Trevor Vickey University of Sheffield, United Kingdom November 21, 2017 HEP Seminar University of Oxford
We have one 125 GeV Higgs boson Standard Model Higgs Standard Model predicted the existence of a single Higgs boson In the SM, the Brout-Englert-Higgs (BEH) mechanism gives mass to the W and Z Interactions of the fundamental particles with the Higgs field gives them their mass After decades of searching, Higgs boson discovered by ATLAS and CMS in July 2012 to-date, looks as predicted JHEP08 (2016) 045 JHEP08 (2016) 045 2
We have one 125 GeV Higgs boson Standard Model Higgs Standard Model predicted the existence of a single Higgs boson In the SM, the Brout-Englert-Higgs (BEH) mechanism gives mass to the W and Z Interactions of the fundamental particles with the Higgs field gives them their mass After decades of searching, Higgs boson discovered by ATLAS and CMS in July 2012 to-date, looks as predicted JHEP08 (2016) 045 ATLAS-CONF-2017-077 3
If the (light) Higgs mass is ~125 GeV, what next? Standard Model Higgs Beyond the SM Higgs The rest of this talk: Suppose that this is not the Standard Model Higgs Higgs with different couplings? 2HDM, Fermiophobic, Higgs impostor More complicated Higgs sector? MSSM, Doubly-charged Higgs, Composite Light scalar Higgs? NMSSM (Next-to-Minimal SUSY extension to SM) Hidden Higgs sector? Higgs to long-lived particles / decays into Dark Matter 4
If the (light) Higgs mass is ~125 GeV, what next? Standard Model Higgs Beyond the SM Higgs The rest of this talk: Suppose that this is not the Standard Model Higgs In SM just one SU(2) doublet is assumed 2HDMs from SUSY or axion models Higgs Triplet Models Can provide Majorana neutrinos with masses Add a Singlet to the SM Can provide a candidate for Dark Matter Hidden sector particles Candidates for Dark Matter Thankfully many of these scenarios are compatible with a 125 GeV Higgs Sadly not enough time today to discuss all of them! 5
Motivation for one theory: Supersymmetry Naturalness (Hierarchy Problem) Unification of the forces (gauge couplings) Provides a candidate for Dark Matter The Minimal Supersymmetric extension to the Standard Model (MSSM) is a Type II 2HDM at tree level 6
Two Higgs Double Models (2HDM) We might have an extended scalar sector (i.e., multiple Higgs bosons) originating from two Higgs doublets the 125 GeV particle could just be one of these Higgs bosons. Type 1 2HDM: All fermions couple to just one of the Higgs doublets Type II 2HDM: All up-type fermions couple to one Higgs doublet and all down-type fermions couple to the other Lepton-specific 2HDM: All quark couplings are like in the Type 1, lepton couplings are like in Type II Flipped 2HDM: All lepton couplings are like in the Type 1, quark couplings are like in Type II Tree-level coupling scale factors (BSM/SM) of a light Higgs boson to vector bosons, leptons, up-type and down-type quarks in 2HDMs The ratio of the vacuum expectation values for the two Higgs doublets = tan β The mixing angle between CP-even Higgs bosons = α 7
MSSM Higgs Sector Consider the case of an MSSM Higgs at the LHC 2 Higgs doublets give rise to 5 physical Higgs bosons: h, H (CP-even), A (CP-odd), H ± Enhanced coupling to 3 rd generation; strong coupling to down-type fermions (at large tanβ get strong enhancements to h/h/a production rates) Diagrams with bbϕ vertex enhanced proportional to tan 2 β where ϕ=h,h,a Can parameterize the masses of the Higgs bosons with two free parameters: tanβ and ma; maximum value of mh ~135 GeV H H, A mass h, A h Low m A High m A (h, A degenerate) (H, A degenerate) 8 8
125 GeV SM-like Higgs is Compatible with SUSY Standard Model Higgs Beyond the SM Higgs Higgs BR + Total Uncert 1-1 10-2 10 ττ cc bb γγ gg Zγ WW ZZ tt LHC HIGGS XS WG 2011 CERN-2012-002 -3 10 100 200 300 400 500 1000 M H [GeV] SM BRs (e.g., h at ~125 GeV) 9 For large ma SM BR values reached
The ATLAS and CMS experiments and the CERN Large Hadron Collider 10
The Large Hadron Collider (LHC) at CERN The world s highest-energy particle collider, just outside of Geneva, CH Proton-proton collider 27 km in circumference 14 TeV design CME Run-I: 7 TeV CME running in 2011 8 TeV CME running in 2012 Run-II 13 TeV CME running since 2015 (on-going) Home to four major experiments
ATLAS = A Toroidal Lhc ApparatuS CMS = Compact Muon Solenoid Major experiments at the LHC
LHC: Integrated Luminosity Luminosity a measurement of the number of collisions that can be produced in a detector per cm 2 and per second. Luminosity delivered over time (integrated luminosity) is expressed in inverse of cross section (i.e. 1/pb or pb -1 picobarn -1 ; 1/fb or 1fb -1 femtobarn -1 ). Higgs xs at 13 TeV ~52 pb. Performance in 2017 has been really impressive, especially considering heavy beam losses related to the 16L2 issue Recently exceeded the goal of 45 fb -1 delivered to each of the ATLAS and CMS experiments in 2017. 13 TeV pp running now finished for 2017. NB: 1 b = 10-24 cm 2 Just under 50 fb -1 delivered to each of ATLAS and CMS in 2017! 13
Neutral BSM Higgs Searches at the LHC MSSM ϕ=h/a/h 14
Neutral MSSM Higgs Searches Results frequently quoted as limits on cross-section x BR, but then also put into the context of different benchmark scenarios. For example: hmssm scenario: the measured value of 125 GeV can be used to predict masses and decay branching ratios of the other Higgs bosons mh mod scenario: top-squark mixing parameter is chosen such that the mass of the lightest CP-even Higgs boson is close to the mass of the one observed at the LHC (historical) Three recent ATLAS high-mass neutral Higgs searches using fermions to show you A/H to ττ using 36.1 fb -1 of 13 TeV pp collision data arxiv: 1709.07242 Inclusive A/H to ttbar using 20.3 fb -1 of 8 TeV pp collision data arxiv:1707.06025 A/H to ttbar + heavy flavor jets in 13.2 fb -1 of 13 TeV data ATLAS-CONF-2016-104 Theory projections for 2σ sensitivity are shown A/H to ττ and A/H to ttbar are each very important channels at mid- to high-ma 15 Djouadi, A., Maiani, L., Polosa, A. et al. JHEP 06 (2015) 168
ATLAS Run-II MSSM Higgs Search (A/H τ + τ - ) New ATLAS MSSM neutral Higgs search for 2017 uses 36.1 fb -1 of 13 TeV data Strongest limits to-date (no recent update of the search released from CMS as of yet) Improvement on the limits from the 2015 ATLAS result: Eur. Phys. J. C (2016) 76: 585 Can use different categories to target main production mechanisms and different decays: With or without b-quarks (b-jets; use of software for identification, i.e., b-tagging ) With or without hadronic tau lepton decays (tau-jets; use software for identification) Hadronic Jet b-quark decay (b-jet) Specialized ATLAS software to identify these 16 Hadronic Tau Lepton decay (tau-jet) In total, 4 categories are considered by the analysis
Post-fit Plots for the 4 Categories (A/H τ + τ - ) Background modeling simulation for backgrounds with true leptons faking taus Data-driven methods for some backgrounds where jets fake taus Fake-factors when Monte Carlo used for some backgrounds where jets can fake taus Discriminant used is the mt tot variable, as it offers good separation of signal from backgrounds due to fake τs 17
Run-II MSSM Neutral Higgs Search (A/H τ + τ - ) Systematics: τ energy scale, τ trigger, jet fake-related (lep-had), top modeling (had-had) Statistically combine the τlep-τhad and τhad-τhad channels for one exclusion limit We determine a σ x BR limit (A/H ττ) for gluon-fusion and b-associated production separately; exclusions range from ~0.9 pb to ~4.0 fb, depending on the Higgs mass and production mechanism. Also produce the limit in the hmssm benchmark scenario. 18
Run-I High-mass Higgs Search (A/H ttbar) We revisited a Run-1 ttbar resonance search that used 20.3 fb -1 of 8 TeV proton-proton collision data: JHEP 08 (2015) 148 This analysis uses the ttbar lepton+jets channel, and takes the interference between the signal and ttbar background production modes into account for the first time Monte Carlo samples used: A/H to ttbar signal: MadGraph5+Pythia6 Backgrounds: ttbar: Powheg-Box+Pythia6 ttbar + V: Madgraph5+Pythia6 single top: Powheg+Pythia6 W+jets and Z+jets: Alpgen+Pythia6 Diboson: Sherpa 19 MadGraph5 used for both Direct and Indirect A/H signal generation (Direct used; difference taken as a modeling systematic)
Signal Modeling (A/H ttbar) The signal process is simulated using the generator MadGraph5 v2.0.1 with the Higgs Effective Couplings Form Factor model (implements the production of scalar and pseudoscalar particles through loop-induced gluon fusion) Loop contributions from both bottom and top quarks are taken into account Signal shape is distorted from a simple Breit-Wigner peak, to a peak-dip structure Statistical interpretation of measured event rates in data are compared to the total sum of Signal + Interference + Background (S + I + B) The mass of the SM-like Higgs boson, h, is chosen to be 125 GeV and sin(α-β) is set to 1 (i.e., the alignment limit, where h has decays as expected in the SM) arxiv:1707.06025 20
Event Selection / Mass Reconstruction (A/H ttbar) Analysis targets the ttbar lepton+jets channel (one W to hadrons, one to leptons) Single electron or single muon triggers are used 2 categories (one for e; one for μ) One high pt electron or muon; high MET from the escaping neutrino; presence of at least 4 high pt jets in the event; at least one jet originating from b quarks must be tagged (70%); Sum of MET and mt > 60 GeV (multi-jets suppression) A chi-squared fit is used for assignment of the decay products, then mtt is reconstructed Events further classified depending on the b-tagged jet(s) assignment 3 categories 6 categories in total (2 lepton types) x (3 b-tagging classifications) 21
ATLAS High-mass Higgs Results (A/H ttbar) arxiv:1707.06025 No significant excess over Standard Model background expectation is observed in the Run-1 inclusive analysis at 8 TeV Type-II 2HDM shown ATLAS also looks into associated production of a heavy Higgs at 13 TeV using 13.2 fb -1. We set upper limits on the signal production cross section x BR for a heavy Higgs mass between 400 GeV and 1000 GeV using the ttbar final state. Limit for Type-I or Type-II 2HDM shown. 22 ATLAS-CONF-2016-104
ATLAS High-mass Searches (H WW or ZZ) ATLAS H WW lvqq (l=e,μ) uses 36.1 fb -1 of 13 TeV pp data Selection: e,μ + MET > 100 GeV, 1 jet Likelihood fit with dedicated control regions for ttbar and W+jets Signal Regions: VBF and ggf/qq (merged and resolved) Inclusion of a jet substructure variable for boosted boson vs. QCD separation H ZZ 4l and llvv (l=e,μ) with ATLAS use 36.1 fb -1 of 13 TeV pp data Uses the same selection as the SM search, but requires both Z bosons to be on-shell Signal Regions: VBF enriched category ggf enriched category Limits are set using various widths arxiv: 1710.07235/ATLAS-CONF-2017-058 arxiv:1708:04445/arxiv: 1708:09638 23
CMS High-mass Searches (H ZZ) CMS separates the searches into the H to ZZ to 4l and llvv final states The ZZ to 4l search is done as a function of Γ (with Γ < mx) on m4l The ZZ search in 4l uses a separate event categorization for ggf and VBF production Matrix element calculations are used to form several kinematic discriminants ZZ to 4 lepton ZZ to llvv CMS PAS HIG-16-033 24 CMS PAS B2G-16-023
Di-Higgs Searches at the LHC 25
Run-1 Di-Higgs Search Results (H hh) Di-Higgs production is very small in the Standard Model due to destructive interference 33.7 fb for proton-proton collisions with CME of 13 TeV (non-resonant) With physics Beyond the Standard Model, this can be enhanced by: Modified top Yukawa coupling or λhhh (non-resonant production) Resonant production: 2HDM H hh, Kaluza-Klein gravitons, etc. 26
ATLAS Run-1I Di-Higgs to bbbb search Public ATLAS Di-Higgs result for Run-1I uses 13.3 fb -1 of 13 TeV data Resolved search: Optimized for low-mass and non-resonant regimes 4 b-tagged jets (R=0.4), jet pairing using ΔRjj and mh constraints Boosted search: Optimized for high-mass resonances, where the two b-jets of the Higgs boson decay cannot be resolved (events splits into 2-, 3- and 4-tag categories) Main backgrounds to both searches are multi-jets and ttbar For the Resolved search, ttbar is modeled using Monte Carlo; multi-jets is data-driven For the Boosted search, multi-jets is data-driven, ttbar shape from Monte Carlo and normalization comes from data Largest systematic uncertainty come from the b-tagging ATLAS-CONF-2016-049 Results: Limit at 95% CL for pp hh bbbb) < 330 fb Excluded mass range: 360 GeV < m(gkk) < 860 GeV (expected 380 GeV < m(gkk) < 910 GeV) m2j = mass of two Higgs candidate system 27
CMS Run-1I Di-Higgs to bbττ search CMS uses 34.9 fb -1 of 13 TeV data in their di-higgs bbττ search Searching for resonant Higgs pair production Massive CP-even scalar decaying into an hh pair One Higgs to bb and the other into ττ (m ττ reconstructed using likelihood method) Analysis split into resolved and boosted topologies for the bb kinematics A multivariate discriminant is used to reject ttbar processes arxiv:1707.02909 28
Charged Higgs Searches at the LHC Charged Higgs H ± (or maybe H ±± ) 29
Charged Higgs bosons could be expected from the MSSM Higgs sector (or more generally 2HDM) H + Production: Light H + : pp tt bw bh + Heavy H + : gb th + and gg tbh + H + Decay: Light H + : Almost exclusively to τν (at low tanβ predominantly to cs) Heavy H + : tb; τν; χ + χ 0 Charged Higgs Searches Charged Higgs searches with taus: Use final states with τνjjb and τνjjbb tt [H ± b][wb] [τνb][qqb] gb [t][h ± ] [qqb][τν] gg [tb][h ± ] [qqbb][τν] 30
ATLAS Run-1I: Charged Higgs: H + τν For MH+ > mt, production in association with a top-quark is dominant Discriminating variable is mt Signal region: 1 hadronic tau with pt > 40 GeV; 3 jets with pt > 25 GeV ( 1 b-tagged jet); MET > 150 GeV Backgrounds: True hadronic taus using Monte Carlo which is normalized / checked in Control Regions with low mt; jet-to-tau fakes are estimated using a fake-factor method Observed hmssm exclusion: tanβ > 42-60 at mh+ = 200-540 GeV 31 ATLAS-CONF-2016-088
ATLAS Run-1I: Charged Higgs: H + tb For MH+ > mt, production in association with a top-quark is dominant Here the lepton + jets final state is used (lepton = e, μ) Discriminant formed from trained binary decision tree (BDT-score) using 12 input variables Signal Regions: 1 lepton with pt > 25 GeV; 5j3b, 5j 4b, 6j3b and 6j 4b Backgrounds: ttbar+jets production dominates, modeled using Powheg + Pythia6: reweight ttbar+light and ttbar+ 1c to NNLO and ttbar+ 1b to NLO Control Regions: 4j2b, 4j 3b, 5j2b and 6j2b HT had (sum of pt of selected jets) used for Control Region ATLAS-CONF-2016-089 32
CMS PAS HIG-16-031 CMS Run-1I: Charged Higgs: H + τν CMS result uses 12.9 fb -1 of 13 TeV data In many ways, CMS analysis very similar to that of ATLAS 33
Beyond Run-II 34
CERN LHC Plans for Run-3, 4-5, ATLAS and CMS pp running at 13 TeV now finished for 2017 Run again in 2018 From all of Run-II, each experiment expects to record ~150 fb -1 The long-term plan is to have about 3000-4000 fb -1 of integrated luminosity ATLAS and CMS will need upgrades Phase-1 Upgrade for L=2-3 x 10 34 cm -2 s -1 In the advanced stages starting production 35 Phase-II Upgrade for L=5-7.5 x 10 35 cm -2 s -1 In the design and prototyping stage
CERN LHC Plans for Run-3, 4-5, ATLAS and CMS pp running at 13 TeV now finished for 2017 Run again in 2018 From all of Run-II, each experiment expects to record ~150 fb -1 The long-term plan is to have about 3000-4000 fb -1 of integrated luminosity ATLAS and CMS will need upgrades 36 Djouadi, A., Maiani, L., Polosa, A. et al. JHEP 06 (2015) 168
Phase-II: All-Silicon Tracking Detector for ATLAS Increased radiation hardness Higher granularity to keep occupancies low Larger readout bandwidth capabilities Reduced material in front of calorimeters Extended coverage at high η ATLAS-TDR-025 37 JINST 3 (2008) S08003
Sheffield Semiconductor Development Facility Univ. of Sheffield has been participating in ATLAS ITk strips detector module pre-production 38
Movie Time: Our Hesse BJ-820 Wire-Bonder Here is Sheffield s BJ-820 automatic wire-bonder in action 39
Conclusions and Outlook ATLAS and CMS have a very active search program for Beyond the Standard Model Higgs bosons and we ve been exploring extended scalar sectors (among other BSM scenarios) Some Run-1 searches data not yet repeated at 13 TeV (e.g., MSSM ϕ μ + μ ) No hint of an extended Higgs sector either during Run-I or so far in Run-II Have now pushed constraints much further than previous searches (e.g., LEP or Tevatron) Even with a SM-like Higgs observed, BSM Higgs searches continue to be relevant (e.g., there are still regions of MSSM parameter space that are compatible with the observed Higgs at 125 GeV) Many exciting new results using Run-II data but this is only a fraction of what is available Impossible for me to touch on everything here many models and searches We have at least one new boson maybe more! These are very exciting times! h 0 A 0 H 0 H + H 40