Searching for neutral Higgs bosons in non-standard channels

Transcription

1 Searching for neutral Higgs bosons in non-standard channels Arjun Menon University of Oregon April 9, 2013

2 Motivation A SM-Higgs like resonance has been observed at the LHC The Higgs sector may also have extra scalars and pseudo-scalars. The ττ-channel is the standard mode of searching for such particles. Examples of models with suppressed A/H ττ rates: Enhanced b b couplings in 2HDM and MSSM. JHEP 1207 (2012) 091 w/ M. Carena, S. Gori, A. Juste, C. Wagner & L-T. Wang Enhance ZA couplings in NMSSM like models. JHEP 1302 (2013) 152 w/ S. Chang

3 Searching Non-Standard Higgses with enhanced b b rates

4 Higgs Sector in 2HDMs The Neutral components acquire vevs and their ratio is tan β = v u /v d. Neglecting CP violation in the Higgs sector, electroweak breaking leaves: 1 CP odd Higgs A 1 charged Higgs H ±, and 2 CP even Higgs bosons h, H One CP-even (SM-like) Higgs has SM strength couplings to gauge bosons. The other CP-even (Non-Standard) Higgs has suppressed couplings to gauge bosons.

5 Couplings to b-quarks and τ-leptons in 2HDMs General 2HDM Higgs fermions couplings are L Yuk = y u H u QU + yd H d QD + ỹu H d QU + ỹ d H u QD + y l H d LE + ỹl H u LE + h.c. d-type fermion couplings to Non-standard Higgses are: where for f = b, τ tan β f eff = g H/Af f m f v tan βf eff ɛ f = ỹf y f ( tan β 1 ɛ ) f 1 + ɛ f tan β tan β

6 Fermion couplings in the MSSM Including 1-loop effects, both quarks couple to both the Higgs bosons so that: L eff = d 0 RŶd[Φ 0 d Φ 0 2 * + Φ 0 u ( ) ˆɛ 0 + ˆɛ Y Ŷ uŷu ]dl 0 + h.c. Φ2 0 * d ~ L y d d ~ R u ~ R y u + u ~ L d L g ~ and have the structure: y u ~ g d R d L h ~ 1 y d h ~ 2 ɛ I 0 2α s 3π M 3µC 0 (m 2 d I 1, m 2 d I 2, M 2 3 ) ɛ Y 1 16π 2 A tµc 0 (m 2 t 1, m 2 t 2, µ 2 ) ɛ τ 3α 2 8π µm 2C 0 (M 2 τ 1, M 2 τ 2, M 2 1 ) d R Kolda, Babu, Buras, Roszkowski...

7 Non-standard Higgs boson production and decay b g A g b A g g b b Gunion et.al. 94, Balazs et.al, Diaz-Cruz et.al., & Huang et.al. 98, Campbell et.al. 03, Dawson et.al. 03 General b and τ couplings are g Abb m b tan βeff b ; g Aττ m τ tan βeff τ v v

8 contd... Enhanced production and decay modes: σ(b b A) 9(tan βeff b BR(A b b) )4 σ(b bh) SM (tan β eff τ )2 + 9(tan β eff b, )2 σ(gg, b b A) BR(A ττ) (tan β τ eff )2 (tan βeff b )2 σ(gg, b b h) SM (tan β eff τ )2 + 9(tan β eff b, )2 In the MSSM the b b channel has greater model dependence than τ τ. Carena et.al. 05

9 Non-Standard Higgs into 3b: Production and Decay tan βeff τ can be small compared to tan βb eff weaker reach in the ττ channel. The H/A b b can be enhanced enough to make it competitive with the clean τ τ channel. In addition to the 4b-final state we also have: g b b H/A 3b channel can be important at 14 TeV LHC for msugra Cao et.al. 09, Baer et. al. 11

10 Signal and Background Simulation Simulation used MG5 interfaced with Pythia 6.4. QCD background: Separately simulated the 3b+X and 2b+j+X where X= 1,2j Used k t matching, with matching scale of 30 GeV. Background separation into bbj and 3b samples does not model b jets with p T below 40 GeV very well. b-jets are clustered using anti-k T with R = 0.4. Jet energy smearing of 100%/ E/GeV. We assume a constant b-tagging efficiency of 60%, a c-jet mis-tag rate of 10% and a light-jet mis-tag rate of 1%. Low mis-tag rate of c- and light-jets leads to the bbj and 3b backgrounds being comparable

11 Selection I vs Selection II Selection I: Exactly 3 b-tagged jets with p T > 60 GeV and η < 2.0. Selection II: Exactly 3 b-tagged jets with p b 1 T > 130 GeV, p b 2,3 T > 50 GeV and η < 2.0. Require M 12, M 13 or M 23 within 25 GeV window of Higgs mass. For tan β b eff = 30 fb 1 7 TeV LHC Selection I Selection II S/B S/ B S/B S/ B m A = 150 GeV m A = 200 GeV m A = 300 GeV m A = 400 GeV

12 Signal and Background Distributions for tan β = 30

13 Reach in the general 2HDM Model

14 The 3b vs ττ in the MSSM

15 CMS Analysis tanβ CMS preliminary fb obs exp (68%) exp (95%) exp (semi-leptonic) exp (all-hadronic) s = 7 TeV tanβ CMS preliminary fb obs exp (68%) exp (95%) exp (semi-leptonic) exp (all-hadronic) s = 7 TeV MSSM m h -max µ = +200 GeV 10 MSSM m h -max µ = -200 GeV [GeV/c ] m A [GeV/c ] m A HCP Nov. 2012

16 Conclusions The A ττ LHC search puts weak limits on regions of large tan βeff b and small tan βτ eff in 2HDMs. The A/H b b is a complementary channel that probes parametric scenarios of large tan βeff b. The reach of the A/H b b channel is limited by low S/B for low to moderate tan βeff b, but can be powerful at large tan βeff b.

17 Search for Non-Standard Higgs in the H ZA channel

18 Motivation: excess in the 2l+ 0,1 and 2 τ h s CMS 2011: 2.1 fb 7 TeV CMS-PAS-SUS CMS 2012: 4.8 fb 7 TeV arxiv:

19 Theoretical Implications of Signal The multi-lepton channel is sensitive to SM Higgs decay modes and with 5 fb 1 of data, the region 120 m h 150 GeV can be probed at 95% C.L. E. Contreras-Compana, et.al. 12 The CMS 2012 multi-lepton data puts limits on BR(t ch) < 2.7% N. Craig et.al. 12 It also leads to constraints on 2HDM s when multiple-channels from h, H, A and H ± decay modes. N. Craig et.al. 13

20 Example: The NMSSM The superpotential has the form with soft terms W = W Yuk + λĥuĥdŝ + κ 3 Ŝ3 V soft = m 2 H u H u 2 + m 2 H d H d 2 + m 2 S S ( m λ SH u H d m κ 3 S3) with m κ κa κ / 2 and m λ λa λ / 2 In the basis where scalar basis (hv, 0 Hv 0, hs) 0 and the pseudo-scalar basis (A 0 v, A 0 s) L Kin Higgs g ( ) 2 Z µ (c θa A 0 1 2c s θ A A 0 2 ) µ s 2β hv 0 + c 2β Hv 0 θw where the h v is direction that acquires a VEV. H Z τ + τ Has been studied in context of explaining LEP anomalies. Dermisek 08, Dermisek and Gunion 09

21 Higgs mass of Benchmark points Model λ κ t β A λ A κ A t µ eff M q (GeV) (GeV) (TeV) (GeV) (TeV) BM BM BM Model m H 0 1 m H 0 2 m A 0 1 m H ± g red. t th 0 1 g red. t th 0 2 (GeV) (GeV) (GeV) (GeV) BM BM BM

22 Higgs couplings of Benchmark points BR of H1 0 b b γγ WW ZZ A 0 1 A0 1 BM BM BM BR : γγ SM = ; WWSM = ; ZZSM = BR of H2 0 b b H1 0H0 1 ZA 0 1 A 0 1 A0 1 BM BM BM BR of A 0 1 ττ b b gg Signal Rate (µ) BM BM BM

23 Event Simulation Simulation used Pythia8.170 for pp collisions. Include the effects of ISR, FSR, multiple interactions and fragmentation. The Z-bosons were allowed to decay only into e, µ, τ. No detector simulator was used, but instead implemented an CMS-like τ h reconstruction algorithm. Trigger requirements: 1-lepton: muon (electron) has a p T > 35 (85) GeV 2-lepton: pt 1 20 GeV and p2 T 10 GeV. Lepton identification: p T 8 GeV and η 2.1. Lepton isolation: I Rel = E cone /E l 0.15, where E l = energy of lepton and E cone = energy in a R = 0.3 (0.4) for muons (electrons).

24 τ h reconstruction τ h reconstruction: 1-pronged track with p T 8.0 GeV. τ h isolation: E ann /E cone 0.15 where, E ann = energy in 0.1 < R 0.3 E cone = energy in R 0.1. Τ 1.0 ΤΤ ArcTan p T T

25 Z ττ efficiency in Efficiency Ε: 2 Τh Channel ma GeV ma GeV ma GeV H H mh GeV ɛ = Number of events to pass cuts Number of events generated

26 Toy-Model for τ h reconstruction m H = 200 GeV and m A = 10 GeV pt GeV Π p T Cut CM Π Isolation Annulus CM θ CM = Angle of π + in rest frame of A when τ + π + ν τ p T is measured in the H rest frame R = the angle between the two charged tracks.

27 Limits of signal due to CMS data Due to low statistics we assume a Poisson distribution for the number of events. We assume the background errors are gaussian The maximum allowed number of signal events at 95% C.L. (S95 Max ) is found by solving db Γ(N [ ] obs + 1, S95 Max + B) 1 exp (B µ B) 2 0 N obs! N B 2σB 2 = 0.05 The bounds on σ sig, normalized to σ SM is µ i 95 S95 i Max σ HSM BR(Z l + l ) ɛ i L

28 contd CL Limit Μ fb 1 Projected 95 CL Limit Μ ma GeV ma GeV m H GeV 7 TeV 8 TeV m H GeV 1-τ h constraint is the strongest due to large ɛ 1τh and N bkg N CMS obs

29 H and A Mass reconstruction in the 2τ h channel Transverse Mass: ma T = pv 2 + 2(E V E+ T pt V pt + ) m T H = (p V + p Z ) 2 + 2((E V + E Z )E T + (pt V + pt Z ) pt + ) where m T i m i Barr et. al., 2009 Collinear Mass: Solve kinematics under assumption that neutrinos are collinear with the visible momenta λ 1 p T V 1 + λ 2 p T V 2 = p T +. where by assumption λ i s are positive. Ellis et. al., 1987

30 H and A Trial Mass Reconstruction in the 2τ h channel The 8 kinematic constraint equations are: p 2 ν 1 = 0 = p 2 ν 2 ( pν1 + p V1 ) 2 = m 2 τ = ( p ν2 + p V2 ) 2 m 2 A = ( p ν1 + p V1 + p ν2 + p V2 ) 2 m 2 H = ( p Z + p ν1 + p V1 + p ν2 + p V2 ) 2 p x ν 1 + p x ν 2 = p x + p y ν 1 + p y ν 2 = p y + However 10 unknowns p νi, m H and m A. Solve for the mean values of m H and m A where solutions exist.

31 Comparison of Mass reconstructions Events GeV Events GeV H H Events GeV Events GeV A A

32 Latest CMS analysis CMS-PAS-SUS But visible p τ T 20 GeV reduced efficiencies.

33 Conclusion The possibility of enhanced H ZA Z τ + τ decay exists. The NMSSM example scenario needs low tan β and large pseudo-scalar mixing. The efficiencies for detecting such a scenario are the largest in the 1τ h and 2τ h channel. The shape of the efficiency curves is due to an interplay between the isolation and min(p T ) cuts. For low m A a boosted τ strategy similar to Englert et. al., 11 may be needed.

34 contd... 1-τ h is the most constraining of the channels. The projected reach with 30 fb 1 CMS data could probe a large region interesting parameter space. For such decays the trial mass reconstruction method is more efficent than the transverse and collinear approaches. The phenomenology of non-standard Higgs bosons can be quite rich and appear in many channels other than ττ.