CONSTRAINING HIGGS CP - PROPERTIES IN GLUON FUSION Matthew Dolan - PowerPoint PPT Presentation

CONSTRAINING HIGGS CP - PROPERTIES IN GLUON FUSION Matthew Dolan SLAC and University of Melbourne 1406.3322 with P . Harris, M. Jankowiak and M. Spannowsky

Introduction • Run I showed the Higgs boson is broadly SM-like • How can we constrain the CP-properties of the Higgs?

Introduction • Higgs an even eigenstate of CP in the SM • Many BSM theories include CP-odd scalars (pseudoscalars) • Or have CP-violation in the Higgs sector • Physical Higgs then not an eigenstate of CP Don’t ask ‘is the Higgs CP-even or odd’ but ‘how much’?

• Traditional analyses rely on angular correlations between decay products in X ! ZZ ! 4 ` Higgs-like state X θ 1 p j 1 Φ 1 p d φ e Q θ µ µ − V 1 µ − ( � e + X j β ˆ θ ` e z � ∆ φ Z Z V 2 θ � j α µ + θ h µ + e − θ e Φ ˆ e z Q ∆ φ j 2 ( � ) p θ � θ 2 p Or in correlations between tagging jets and decay products in weak boson fusion (WBF) From Englert et al, 1212.0840

Pseudoscalars do not have renormalisable couplings to massive vector bosons 0.6 0.6 0.6 1 d Γ 1 d σ 1 d σ + 0 Γ d ∆φ σ d ∆φ σ d ∆φ jj D5 0.4 0.4 0.4 - 0 D5 - 0 + D5 0 + SM - 0 0 SM D5 + 0.2 0.2 0.2 2 + + 0 0 + + + D5 2 SM 2 0 D5 0 0 0 0 2 4 6 0 2 4 6 0 2 4 6 Leading order scalar couplings are d=3 e hV µ V µ Leading order pseudoscalar couplings are d=5 e hV µ ν e V µ ν ecays and From Englert et al, 1212.0840

Results from ATLAS-CONF-2015-008 f 1 ( µ W − µ g L V 2 g H Z Z Z µ Z µ + g HWW W + c α  SM 0 = Sets constraints on f Z µ ν g − 1 1 c α  H Z Z Z µ ν Z µ ν + s α  AZ Z Z µ ν ˜ 4 Λ f W − µ ν g) − 1 1 µ ν W − µ ν + s α  AWW W + µ ν ˜ c α  HWW W + X 0 . 2 Λ mixing angles and higher dimension s α = sin α , c α = cos α operators suppressed by scale Λ Tree-level SM is κ SM = 1 , c α = 1 , Λ → ∞

How large should CP-violating effects be? 1 α Naive expectation: Λ ∼ 2 π v κ SM ∼ 1 , κ AV V ∼ 1 κ AV V = 1 Λ κ AV V ∼ α v 8 π ∼ 10 − 3 ˜ 4 κ AV V / κ SM ) tan α ∼ 10 − 3 tan α (˜ Coupling ratio Best fit value 95% CL Exclusion Regions Combined Expected Observed Expected Observed  HVV /  SM ˜ 0 . 0 − 0 . 48 ( −∞ , − 0 . 55] S [4 . 80 , ∞ ) ( −∞ , − 0 . 73] S [0 . 63 , ∞ ) ( ˜  AVV /  SM ) · tan ↵ 0 . 0 − 0 . 68 ( −∞ , − 2 . 33] S [2 . 30 , ∞ ) ( −∞ , − 2 . 18] S [0 . 83 , ∞ )

Information in Higgs production too BR ( h → ZZ ∗ ) and WBF negligible for a pure CP-odd state Gluon fusion increases by a factor ~9/4 Signal strength info rules out pure pseudoscalar at 4 σ 3.0 � 2.5 2.0 1.5 1 α < 0 . 76 (95% C.L.) 1.5 x u 0.5 1.0 1.5 1 0.5 0.5 Freitas, Schwaller 1211.1980 0.0 � 0.0 0.5 1.0 1.5 Djouadi, Moreau 1303.6591 Α

What Other Couplings Can Be Probed? • Scalar and pseudoscalar couplings to fermions and massless vector bosons arise at the same order Tree-level couplings to fermions f γ 5 f h ¯ h ¯ ff 1-loop couplings to gluons/photons d hG µ ν e g hG µ ν G µ ν G µ ν , r

What Other Couplings Can Be Probed? Will focus on CP-sensitive variables in Higgs production • • Production via gluon fusion arises at same order in both cases H H H (a) (c) (b) For decay see Felix and Marco’s talks

What Other Couplings Can Be Probed? Will focus on CP-sensitive variables in Higgs production • WBF amenable to angular analysis • Gauge-Higgs invariant mass in associated production • θ 1 LHC8 12000 + 0 j 1 - 0 10000 d + 2 Q V 1 Arbitrary Units ( � 8000 ∆ φ 6000 V 2 θ � 4000 Q j 2 ( � ) 2000 θ 2 0 500 1000 1500 2000 2500 M VX For decays see Felix and Marco’s talks Ellis, Sanz, You 1208.6002

What Other Couplings Can Be Probed? • Higgs plus two jet production is known to be sensitive to the Higgs CP properties through angular correlations in the jets • In particular differences between azimuthal angles ∆ φ jj -3 10 × 45 jj Φ pp jj H CP-even → ∆ /d m = 160 GeV CP-odd H σ 40 d CP-mixed σ 1/ 35 30 25 20 15 10 -150 -100 -50 0 50 100 150 ∆ Φ jj Klamke, Zeppenfeld ’07

We will consider a mixed CP-state with couplings ff = cos ↵ y f ¯ y f ¯ f f h + sin ↵ e f i � 5 f h . L h ¯ L hV V ⊃ cos α 2 m 2 hW µ W µ + cos α 2 m 2 W Z hZ µ Z µ v v This generates couplings to gluons L hgg = cos ↵ ↵ S µ ν G a,µ ν + sin ↵ ↵ S µ ν e 12 ⇡ v hG a 4 ⇡ v hG a G a,µ ν Mixing parametrised by angle α is pure CP-even α = 0 is pure CP-odd α = π / 2

Event Generation We generate signal using VBFNLO 2.6.3 at 8 and 14 TeV Gluon fusion generated at NLO WBF generated at LO Background using Sherpa 2.0.0 Generate Zjj (QCD + EW), W+jets and t ¯ t QCD multijets assumed to be flat across phase-space

Cross-Sections In the CP-odd limit the WBF cross-section vanishes at tree-level The CP-odd GF cross-section is larger than the CP-even case by 9/4 8 TeV GF cross-section (fb) 8 TeV WBF cross-section (fb) 14 TeV GF cross-section (fb) 14 TeV WBF cross-section (fb) α 0.00 250 467 1141 1481 0.30 278 426 1268 1351 0.60 352 318 1606 1009 0.90 447 181 2038 572 1.20 529 61 2411 194 We focus on h → ττ

Event Selection We consider four different final states: di-hadronic, semi-leptonic and leptonic (e+mu) Cuts designed to mimic ATLAS/CMS di-tau analyses µ τ h e τ h eµ τ h τ h p µ p e p lead T > 20 GeV T > 25 GeV > 20 GeV T lepton selection T > 45 GeV p τ p trail T > 30 GeV T > 30 GeV > 10 GeV p τ p τ T m µ kinematic selection p H m e T < 30 GeV b-tag veto with p b T > 100 GeV T < 30 GeV T > 20 GeV loose jet selection m jj > 500 GeV m jj > 500 GeV m jj > 500 GeV m jj > 500 GeV | ∆ η jj | > 3.5 | ∆ η jj | > 3.5 | ∆ η jj | > 3.5 | ∆ η jj | > 3.5 m jj > 700 GeV m jj > 700 GeV m jj > 700 GeV tight jet selection | ∆ η jj | > 4 . 5 | ∆ η jj | > 4.5 | ∆ η jj | > 4.5 p H p H p H T > 100 GeV T > 100 GeV T > 100 GeV CMS: 1401.5041 ATLAS-CONF-2013-108 updated to 1501.04943

Kinematic Distributions 1 GeV) rad) Bkgs Bkgs 0.9 Higgs(WBF) -1 Higgs(WBF) (0.1 -1 0.8 (100 ggH+2j ( = 0.0) α ggH+2j ( = 0.0) α 1 jj φ 0.7 ggH+2j ( = 0.6) ggH+2j ( = 0.6) α α ∆ jj /d /dm ggH+2j ( = 1.2) ggH+2j ( = 1.2) α α 0.6 σ d σ -1 d 0.5 σ -1 σ 0.4 -1 10 0.3 0.2 0.1 -2 0 10 500 1000 1500 2000 2500 3000 -3 -2 -1 0 1 2 3 (rad) m (GeV) ∆ φ jj jj 5 | |/2) jj η Bkgs /d| Bkgs 4.5 jj φ σ ∆ d /dsin(| 4 Higgs(WBF) -1 Higgs(WBF) σ 3.5 ggH+2j ( = 0.0) ggH+2j ( = 0.0) α α σ d 3 -1 1 ggH+2j ( = 0.6) σ α ggH+2j ( = 0.6) α 2.5 ggH+2j ( = 1.2) α ggH+2j ( = 1.2) α 2 1.5 1 0.5 -1 10 4 4.5 5 5.5 6 6.5 7 7.5 8 8.5 9 9.510 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 η sin(| |/2) ∆ φ jj jj Most sensitive variable is ∆ φ jj = φ y> 0 − φ y< 0

is pretty optimal ∆ φ jj = φ y> 0 − φ y< 0 Trained a BDT to discriminate between two gluon fusion samples with and α = 0 α = 1 . 2 Signal Efficiency Signal Efficiency 1 1 0.8 0.8 0.6 0.6 (bdt) α (bdt) α 1.5 1.5 (sin(| |)) α ∆ φ (sin(| |)) α ∆ φ 0.4 0.4 1.5 jj 1.5 jj (bdt) α (bdt) α 0.6 0.2 0.6 0.2 (sin(| |)) (sin(| |)) α ∆ φ α ∆ φ 0.6 0.6 jj jj 0 0 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Background Efficiency Background Efficiency 8 TeV 14 TeV

Also trained a BDT to discriminate between GF+WBF signal and sum of backgrounds A category-based analysis using only m ττ , ∆ φ jj , m jj , ∆ η jj does about as well as the BDT trained on full set of variables 2 GeV) Bkgs 1.8 -1 1.6 (20 Higgs(VBF) 1.4 T ggH+2j ( = 0.0) /dp α 1.2 σ d ggH+2j ( = 0.6) α -1 1 σ ggH+2j ( = 1.2) α 0.8 0.6 0.4 0.2 0 0 50 100 150 200 250 300 350 400 p leading Jet (GeV) T

Constraints 5 Significance Significance m and sin(| /2|) loose( ) ∆ φ α m and sin(| /2|) loose( ) ∆ φ α τ τ τ τ jj jj m and sin(| /2|) tight( ) -1 ∆ φ α m and sin(| /2|) tight( ) ∆ φ α 4.5 14 50 fb 14 TeV -1 τ τ 20 fb 8 TeV τ τ jj jj mva and sin(| /2|) tight( ) ∆ φ α mva and sin(| /2|) tight( ) ∆ φ α jj jj m and sin(| /2|) loose ∆ φ m and sin(| /2|) loose ∆ φ 4 τ τ τ τ jj 12 jj m and sin(| /2|) tight ∆ φ m and sin(| /2|) tight ∆ φ τ τ τ τ jj jj mva vs sin(| /2|) tight ∆ φ mva vs sin(| /2|) tight 3.5 ∆ φ jj jj 10 m loose m loose τ τ τ τ 3 m tight m tight τ τ τ τ 8 2.5 2 6 1.5 4 1 2 0.5 0 0 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 α α Dashed: Significance of total signal over SM background Solid: Exclusion significance relative to case α = 0 with 50/fb at 14 TeV α ≤ 0 . 7

Constraints ) -1 Exclusion (fb m and Loose ∆ φ τ τ 3 10 m and Tight ∆ φ τ τ MVA and Tight ∆ φ 2 10 σ 2 10 1 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 α Expected exclusion limit as a function of integrated luminosity at 14 TeV