Searches for rare radiative Higgs boson decays to light mesons with ICHEP 2018 (Seoul) 5th July 2018 Andy Chisholm (CERN and University of Birmingham) on behalf of the collaboration
Motivation - Why search for H M γ decays? 1 Why are the charm and light quark Yukawa couplings important? The light (u, d, s) and charm (c) quark 2mq couplings (y q = mq ) are very small, v 174 GeV new physics in this sector could easily have gone unnoticed! (small effect on Γ H ) However, H c c decays now constitute the largest part of the SM prediction for Γ H for which we have no experimental evidence To date, direct experimental evidence only available for 3rd generation Yukawa couplings! How can we study these couplings directly at the LHC? 38.9% 38.9% 0.01% 2.9% 58.2% 0.01% 2.9% 58.2% H b b H c c H s s H other Cartoon of SM 125 GeV H q q branching fractions, H uū/d d too small to show! Many indirect ways to study these coupling (e.g. global fits, pt H and y H distributions), though constraints are often quite model dependent Huge backgrounds make searches for inclusive H q q decays very challenging at LHC (though more promising for H c c, see search ) How can radiative Higgs boson decays to light mesons help?... Phys. Rev. Lett. 120 (2018) 211802 (arxiv:1802.04329)
H M γ Decays - Motivation 2 H M γ decays provide a clean probe of the charm and light quark Yukawa couplings at the LHC M is a vector (J PC = 1 ) light meson or quarkonium state such as J/ψ, ψ(2s), Υ(nS), φ(1020), ρ(770) Interference between direct (H q q) and indirect (H γγ ) contributions Direct amplitude (upper) provides sensitivity to the magnitude and sign of the Hq q couplings (e.g. M = J/ψ sensitive to Hc c coupling) Indirect amplitude (lower) makes dominant contribution to decay width, but not sensitive to Yukawa couplings H M γ M B (H J/ψ γ) = (2.99 ± 0.) 10 6 B (H ψ(2s) γ) = (1.03 ± 0.06) 10 6 B (H Υ(1S) γ) = (5.2 +2.0 1.7 ) 10 9 B (H φ γ) = (2.3 ± 0.1) 10 6 B (H ρ γ) = (1.7 ± 0.1) 10 5 H γ Phys. Rev. D 90, 113010 (2014) (arxiv:1407.6695) JHEP 1508 (2015) 012 (arxiv:1505.03870)
Z M γ Decays - Motivation 3 Z M γ decays could provide a stepping stone towards the observation of the Higgs decays at the LHC Analogous to Higgs boson decay, could provide a control channel for experimental analyses Proposed as a channel to test QCD factorisation Similar interference between direct (upper) and indirect (lower) contributions to the amplitude, relative contribution from indirect amplitude much lower than Higgs decay While these braching fractions are much lower than for the corresponding Higgs boson decays, Z bosons much more copiously produced than Higgs at the LHC... Z M γ M B (Z J/ψ γ) = (9.0 +1.5 1.4 ) 10 8 B (Z Υ(1S) γ) = (4.8 ± 0.3) 10 8 B (Z φ γ) = (1.0 ± 0.1) 10 8 B (Z ρ γ) = (4.2 ± 0.5) 10 8 Z γ Phys. Rev. D 97, 0009 (2018) (arxiv:1709.09320) JHEP 1504 (2015) 101 (arxiv:1501.06569)
Searches for Rare Exclusive Radiative H/Z Decays with 4 H/Z φ/ρ γ decays H/Z ψ(ns)/υ(ns) γ decays EUROPEAN ORGANISATION FOR NUCLEAR RESEARCH (CERN) EUROPEAN ORGANISATION FOR NUCLEAR RESEARCH (CERN) Submitted to: JHEP CERN-EP-2017-273 8th December 2017 Submitted to: PLB CERN-EP-2018-154 3rd July 2018 arxiv:1712.02758v1 [hep-ex] 7 Dec 2017 Search for exclusive Higgs and Z boson decays to and with the detector The Collaboration A search for the exclusive decays of the Higgs and Z bosons to a or meson and a photon is performed with a pp collision data sample corresponding to an integrated luminosity of up to 35.6 fb 1 collected at p s = 13 TeV with the detector at the CERN Large Hadron Collider. These decays have been suggested as a probe of the Higgs boson couplings to light quarks. No significant excess of events is observed above the background, as expected from the Standard Model. Upper limits at 95% confidence level were obtained on the branching fractions of the Higgs boson decays to and of 4.8 10 4 and 8.8 10 4, respectively. The corresponding 95% confidence level upper limits for the Z boson decays are 0.9 10 6 and 25 10 6 for and, respectively. arxiv:1807.00802v1 [hep-ex] 2 Jul 2018 Searches for exclusive Higgs and Z boson decays into J/ψ γ, ψ(2s) γ, and Υ(nS) γ at s = 13 TeV with the detector The Collaboration Searches for the exclusive decays of the Higgs and Z bosons into a J/ψ, ψ(2s), or Υ(nS) (n = 1, 2, 3) meson and a photon are performed with a pp collision data sample corresponding to an integrated luminosity of 36.1 fb 1 collected at s = 13 TeV with the detector at the CERN Large Hadron Collider. No significant excess of events is observed above the expected backgrounds, and 95% confidence-level upper limits on the branching fractions of the Higgs boson decays to J/ψ γ, ψ(2s) γ, and Υ(nS) γ of 3.5 10 4, 2.0 10 3, and (4.9, 5.9, 5.7) 10 4, respectively, are obtained assuming Standard Model production. The corresponding 95% confidence-level upper limits for the branching fractions of the Z boson decays are 2.3 10 6, 4.5 10 6 and (2.8, 1.7, 4.8) 10 6, respectively. -HIGG-20-13 2017 CERN for the benefit of the Collaboration. Reproduction of this article or parts of it is allowed as specified in the CC-BY-4.0 license. Accepted by JHEP (arxiv:1712.02758) Updates original search for H/Z φ γ decays with 10 more data and adds the ρ channel New for ICHEP! -HIGG-20-23 2018 CERN for the benefit of the Collaboration. Reproduction of this article or parts of it is allowed as specified in the CC-BY-4.0 license. Submitted to PLB (arxiv:1807.00802) Updates Run 1 search for H/Z ψ(ns)/υ(ns) γ with Run 2 dataset and adds ψ(2s) channel Phys. Rev. Lett. 117 (20), 111802 (arxiv:07.03400) Phys. Rev. Lett. 114 (2015), 121801 (arxiv:1501.03276) Analyses share similar event selection and signal/background modelling strategies, aim to describe them in parallel while highlighting the important differences...
H/Z φ/ρ γ - Sample and Event Selection 5 Focus on dominant decays φ K + K (B 49%) and ρ π + π (B 99%) and target high rate inclusive H and Z production Trigger and Sample Dedicated photon + di-track triggers implemented to identify distinctive event topology Collected up to 35.6 fb 1 s = 13 TeV pp dataset during the 2015 and 20 LHC runs M = {φ, ρ} Selection Require m K + K or m π + π consistent with φ or ρ meson mass (next slide) Minimum pt M requirement varying with m M γ from 40 47.2 GeV (to optimise both H and Z searches) Photon Selection Tight photon ID requirements Isolated in both tracker and calorimeter p γ T > 35 GeV φ(m, γ) > π/2 lead trk. pt Track Pair Selection > 20 GeV sub-lead trk. pt > 15 GeV Oppositely charged pair of tracks (no π/k/p ID applied) Isolated in tracker (accounting for neighboring track)
H/Z φ/ρ γ - Event Characteristics (arxiv:1712.02758) 6 Events / 2 GeV 0.1 p K 2 T 0.08 0.06 0.04 0.02 Simulation p K 1 T γ p T H φ Before selection After selection 0 0 20 40 60 80 100 120 140 p [GeV] T Candidates / 0.002 GeV 3000 2500 2000 1500 1000 500 s γ 1 = 13 TeV, 35.6 fb Total signal efficiency from 17% (H φ γ) to 0.4% (Z ρ γ) 0 1.00 1.01 1.02 1.03 1.04 1.05 m + K K [GeV] Typically a very small angular separation between the two tracks from boosted φ/ρ decays Fit Result + φ K K Total Background Candidates / 0.022 GeV 10000 9000 8000 7000 6000 5000 4000 3000 2000 1000 s 1 = 13 TeV, 32.3 fb Events/0.0005 0.07 0.06 0.05 0.04 0.03 0.02 0.01 Simulation H φ γ Before selection After selection 0 0 0.01 0.02 0.03 0.04 0.05 R KK Fit Result ρ π + π Total Background 0 0.5 0.6 0.7 0.8 0.9 1.0 m π + π [GeV] Clear evidence in data of φ/ρ mesons in association with high p T photons in m K + K (left) and m π + π (right) distributions of M γ candidates
H/Z ψ(ns)/υ(ns) γ - Sample and Event Selection 7 Focus on the experimentally clean ψ(ns)/υ(ns) µ + µ decays and target high rate inclusive H and Z production Trigger and Sample Dedicated photon + single muon triggers implemented to identify distinctive event topology Collected 36.1 fb 1 s = 13 TeV pp dataset during the 2015 and 20 LHC runs M = {ψ(ns), Υ(nS)} Selection Require m µ + µ loosely consistent with ψ(ns) or Υ(nS) masses (next slide) Minimum pt M requirement varying with m M γ from 34 54.4 GeV, depending on channel (to optimise both H and Z searches) Photon Selection Tight photon ID requirements Isolated in both tracker and calorimeter p γ T > 35 GeV φ(m, γ) > π/2 p µ lead T p µ sub-lead T Di-muon Selection > 18 GeV > 3 GeV Oppositely charged pair of muons Isolated in tracker (accounting for neighboring muon track) L xy /σ Lxy < 3 to reject b ψ(ns)
H/Z ψ(ns)/υ(ns) γ - Event Characteristics (arxiv:1807.00802) 8 Clear evidence in data of ψ(ns) and Υ(nS) produced in association with high p T photons in m µ + µ distributions of M γ candidates Events / 0.04 GeV 1000 800 600 Fit ψ(ns) Background -1 s = 13 TeV, 36.1 fb Region: GR ψ(ns) γ channel σ = 54 ± 1 MeV Events / 0.1 GeV 100 80 60 Fit ϒ(nS) Background -1 s = 13 TeV, 36.1 fb Region: GR ϒ(nS) γ channel B category = 105 ± 18 MeV σ 1S 400 40 200 20 0 2.0 2.5 3.0 3.5 4.0 [GeV] m µ + - µ 0 8.0 8.5 9.0 9.5 10.0 10.5 11.0 11.5 12.0 [GeV] m µ + - µ Total signal efficiency varies from 23% (H Υ(nS) γ) to 11% (Z J/ψ γ)
H/Z M γ - Signal Modelling 9 γ p M Dedicated Signal MC Simulations Higgs and Z boson production modelled with POWHEG-BOX v2 + PYTHIA8 MC generators (PYTHIA8 alone for VH production) Predictions scaled to LHC Higgs XS WG YR4 total cross-sections and measurement of s = 13 TeV inclusive Z cross-section H, Z {K + K, π + π, µ + µ } γ implemented accounting for helicity effects in M decays Signal shape in m M γ modelled with Gaussian or Voigtian PDFs, with parameters fit to MC Resolution in m M γ around 1.6 1.8% for all final states (for both H and Z decays) arxiv:10.07922 Phys. Lett. B 759 (20) 601 (arxiv:03.09222) p Candidates / 0.5 GeV γ K + K Candidates / 0.5 GeV + K γ K 3 10 2.2 2.0 1.8 + K H φ(k ) γ 1.6 Inclusive Signal MC 1.4 Signal Fit 1.2 1.0 0.8 0.6 0.4 0.2 0.0 110 115 120 125 130 135 140 350 300 250 200 150 100 50 m + K [GeV] K γ + Z φ(k K ) γ Signal MC Signal Fit 0 75 80 85 90 95 100 105 110 115 Inclusive m + K [GeV] K γ Figures from arxiv:1712.02758
H/Z φ/ρ γ - Background Modelling (arxiv:1712.02758) 10 Background Composition Background dominated by QCD production of photon+jet and multi-jet events -Driven Background Model Begin with large background dominated sample of loose M γ events with relaxed p T and isolation requirements (on both M and γ) Model kinematic and isolation distributions and generate M γ pseudo-candidates Apply nominal selection to pseudo- candidates to model background in signal region Distributions smoothed with Gaussian Kernal Density Estimation technique to provide good description of the background Model validated with both dedicated control regions and events where M candidate di-track mass is in sidebands of φ/ρ peak /Model Candidates / 2 GeV /Model Candidates / 2 GeV 450 400 350 300 250 200 150 100 50 φ Sideband Region 1 = 13 TeV, 35.6 fb 0 1.5 1.0 0.5 50 100 150 200 250 300 [GeV] 600 500 400 300 200 100 ρ Sideband Region s Background Model Model Shape Uncertainty m K + K γ 1 = 13 TeV, 32.3 fb 0 1.5 1.0 0.5 50 100 150 200 250 300 m π [GeV] π γ s Background Model Model Shape Uncertainty
H/Z ψ(ns)/υ(ns) γ - Background Modelling (arxiv:1807.00802) 11 Two main distinct sources of background: non-resonant production of M γ candidates and resonant Z µ + µ γ production Events / 2.50 GeV 250 200 150 1 s = 13 TeV, 36.1 fb Region : GR ψ(ns) γ Background model Z FSR Model uncertainty Events / 2.50 GeV 120 100 80 60 1 s = 13 TeV, 36.1 fb Region : GR ϒ(nS)γ B category Background model Z FSR Model uncertainty 100 40 50 20 /Model 1.5 1.0 0.5 50 100 150 200 250 300 m µ γ [GeV] + µ /Model 1.5 50 100 150 200 250 300 1.0 0.5 50 100 150 200 250 300 m µ γ [GeV] µ m µ + µ γ distribution for background control region for J/ψ γ candidates (left) and Υ(nS) γ candidates (right) Non-resonant background: modelled with same data-driven method used in H/Z φ/ρ γ analysis shown earlier and validated in a similar manner Z µ + µ γ background: shape determined from simulation (fitted Voigtian PDF), normalisation determined from fit to data
H/Z M γ Decays - Statistical Analysis and Uncertainties 12 Statistical Analysis Quantify presence of a signal with an un-binned likelihood fit with the m M γ distribution used as the primary S/B discriminant In H/Z ψ(ns)/υ(ns) γ case, the m µ + µ distribution also used (2D fit) to gain sensitivity to the individual ψ/υ(ns) states and control Z µ + µ γ backgound Limits set with CL s formalism and profile likelihood ratio test statistic Source Signal Yield of Uncertainty [%] Uncertainty φ/ρ ψ/υ H Signal Modelling 6.3 7.2 Z Signal Modelling 2.9 5.7 Integrated Luminosity 3.4 2.1 Photon ID Efficiency 2.5 1.4 Trigger Efficiency 2.0 2.0 Tracking Efficiency 6.0 Muon ID Efficiency 2.8 Background Modelling Uncertainties Estimated from modifications to modeling procedure (e.g. shifting p γ T or tilting φ(m, γ) distributions) Shape uncertainties included in likelihood as m M γ shape morphing nuisance parameters Signal Modelling Uncertainties Typically dominanted by H/Z cross-section and theory uncertainties
H/Z φ/ρ γ Decays - Results (arxiv:1712.02758) 13 Events / 1 GeV 250 200 1 s=13 TeV, 35.6 fb Events / 1 GeV 1000 800 1 s=13 TeV, 32.3 fb Invariant mass distributions of K + K γ (left) and π + π γ (right) candidates Limit on B(H φ γ) improved by almost 2 w.r.t. earlier search (arxiv:07.03400) / Fit 150 100 50 Background Fit ±1σ Background 4 B(H φγ)=4.8 10 6 B(Z φγ)=0.9 10 1.2 1 0.8 80 90 100 110 120 130 m + K K γ [GeV] Observable / Fit 600 400 200 1.2 1 0.8 Background Fit ±1σ Background 4 B(H ργ)=8.8 10 6 B(Z ργ)=25 10 80 90 100 110 120 130 m π + π [GeV] γ 95% CL Upper Limit Expected Observed B (H φ γ) (4.2 +1.8 1.2 ) 10 4 4.8 10 4 B (H ρ γ) (8.4 +4.1 2.4 ) 10 4 8.8 10 4 B (Z φ γ) (1.3 +0.6 0.4 ) 10 6 0.9 10 6 B (Z ρ γ) (33 +13 9 ) 10 6 25 10 6 World s first constraint on light quark Yukawa couplings from searches for H ρ γ decays! (limit only 52 B SM)
H/Z ψ(ns) γ Decays - Results (arxiv:1807.00802) 14 Events/2.5 GeV 60 50 40 30-1 s=13 TeV, 36.1 fb Background fit ±1σ Combinatoric ψ(ns) background Z FSR -3 B(H ψ(ns)γ)=10-6 B(Z ψ(ns)γ)=10 Events/0.05 GeV 100 80 60 Background fit ±1σ -1 Combinatoric s=13 TeV, 36.1 fb Z FSR ψ(ns) background -3 B(H ψ(ns)γ)=10-6 B(Z ψ(ns)γ)=10 20 40 10 20 /Bkgd 1.5 0.5 1 50 100 150 200 250 300 m µ + - µ γ [GeV] /Bkgd 1.5 0.5 1 2 2.2 2.4 2.6 2.8 3 3.2 3.4 3.6 3.8 4 4.2 m µ - [GeV] + µ Projections of fit to µ + µ γ (left) and µ + µ (right) invariant mass distributions Observable 95% CL Upper Limit Expected Observed B (H J/ψ γ) (3.0 +1.4 0.8 ) 10 4 3.5 10 4 B (H ψ(2s) γ) (15.6 +7.7 4.4 ) 10 4 19.8 10 4 B (Z J/ψ γ) (1.1 +0.5 0.3 ) 10 6 2.3 10 6 B (Z ψ(2s) γ) (6.0 +2.7 1.7 ) 10 6 4.5 10 6 World s first limit on H/Z ψ(2s) γ decays! Limit on B(H J/ψ γ) improved by factor 4 w.r.t. Run 1 result!
H/Z Υ(nS) γ Decays - Results (arxiv:1807.00802) 15 Projections of fit to µ + µ γ (left) and µ + µ (right) invariant mass distributions Limits on B(H Υ(nS) γ) improved by around a factor 2 w.r.t. Run 1 Almost perfect cancellation between (in-)direct amplitudes for SM y b, rate highly sensitive to y b /yb SM! Events/5 GeV /Bkgd 0 140 120 100 80 60 40 20 1.5 0.5 1-1 s=13 TeV, 36.1 fb B category Background fit ±1σ Combinatoric ϒ(nS) background Z FSR -3 B(H ϒγ)=10-6 B(Z ϒγ)=10 50 100 150 200 250 300 m µ - γ [GeV] + µ Events/0.1 GeV /Bkgd 45 40 35 30 25 20 15 10 5 1.5 0.5 1 Background fit ±1σ Z FSR Combinatoric ϒ(nS) background -3 B(H ϒγ)=10-6 B(Z ϒγ)=10-1 s=13 TeV, 36.1 fb B category 8 8.5 9 9.5 10 10.5 11 11.5 12 - [GeV] 95% CL Upper Limit Observable Expected Observed B (H Υ(1S) γ) (5.0 +2.4 1.4 ) 10 4 4.9 10 4 B (H Υ(2S) γ) (6.2 +3.0 1.7 ) 10 4 5.9 10 4 B (H Υ(3S) γ) (5.0 +2.5 1.4 ) 10 4 5.7 10 4 B (Z Υ(1S) γ) (2.8 +1.2 0.8 ) 10 6 2.8 10 6 B (Z Υ(2S) γ) (3.8 +1.6 1.1 ) 10 6 1.7 10 6 B (Z Υ(3S) γ) (3.0 +1.3 0.8 ) 10 6 4.8 10 6 m µ + µ Limits on H Υ(nS) γ correspond to constraint y b /y SM b < O(10)!
Summary γ p M p is pioneering the study of rare exclusive radiative Higgs boson decays at the LHC, leading to new constraints on the light and charm quark Yukawa couplings! Analogous searches for rare Z boson decays also pursued as potential new angle to study QCD and stepping stone on path towards SM Higgs sensitivity Limits imply constraints on charm and light quark Yukawa couplings complementary to other studies (e.g. H c c and pt H / y H distributions) Latest analyses now able to exclude H M γ branching fractions at 10 4 level Highlights include improvement in B (H J/ψ γ) by factor 4 and world s first information on the decays H ψ(2s) γ and H ρ γ These experimental studies and the associated theoretical work represent an important emerging subfield of Higgs physics!
Additional Slides
Detector 18 Inner Detector (ID): Silicon Pixels and Strips (SCT) with Transition Radiation Tracker (TRT) η < 2.5 New for Run 2! - Insertable B-Layer (IBL) - additional inner-most pixel layer (r = 33 mm) and lower x/x 0 beam pipe LAr EM Calorimeter: Highly granular + longitudinally segmented (3-4 layers) Two Level Trigger: Level 1 (Hardware) + High Level (Software) trigger
H/Z ψ(ns)/υ(ns) γ Decays - Fit Result (arxiv:1807.00802) 19 J/ψ region ψ(2s) region m µ + µ γ projections Events/2.5 GeV Events/0.05 GeV 30 25 20 15 10 5-1 s=13 TeV, 36.1 fb 2.9 < m µ - < 3.3 GeV + µ Background fit Combinatoric ψ(ns) background Z FSR -4 B(H J/ψγ)=3.5 10-6 B(Z J/ψγ)=2.2 10 50 100 150 200 250 300 m µ + - µ γ [GeV] 50 40 30 20 10 Background fit -1 s=13 TeV, 36.1 fb Z FSR 86 < m µ - Combinatoric γ < 96 GeV + µ ψ(ns) background -6 B(Z J/ψγ)=2.3 10-6 B(Z ψ(2s)γ)=4.5 10 2 2.2 2.4 2.6 2.8 3 3.2 3.4 3.6 3.8 4 4.2 m µ + - µ [GeV] Z region Events/2.5 GeV Events/0.05 GeV 14 12 10 8 6 4 2-1 s=13 TeV, 36.1 fb 3.5 < m µ - < 3.9 GeV + µ Background fit Combinatoric ψ(ns) background Z FSR -4 B(H ψ(2s)γ)=19.8 10-6 B(Z ψ(2s)γ)=4.5 10 50 100 150 200 250 300 m µ + - µ γ [GeV] 20 18 14 12 10 8 6 4 2 Background fit -1 s=13 TeV, 36.1 fb Z FSR 120 < m µ - Combinatoric γ < 130 GeV + µ ψ(ns) background -4 B(H J/ψγ)=3.5 10-4 B(H ψ(2s)γ)=19.8 10 2 2.2 2.4 2.6 2.8 3 3.2 3.4 3.6 3.8 4 4.2 m µ + - µ [GeV] H region m µ + µ projections
Prospects for H J/ψ γ in a HL-LHC scenario (ATL-PHYS-PUB-2015-043) 20 Run 1 H J/ψ γ analysis projected to s = 14 TeV scenario with 300(0) fb 1 Events / ( 2 GeV ) 900 Simulation Preliminary 800 s=14 TeV, 3000 fb 700 600 500 400 300 200 100 1 (Bkg. Only) S+B Fit Background H Signal 100 Z Signal 10 Expected branching ratio limit at 95% CL B (H J/ψγ) [ 10 6 ] B (Z J/ψγ) [ 10 7 ] Cut Based Multivariate Analysis Cut Based 300 fb 1 185 +81 52 153 +69 43 7.0 +2.7 2.0 3000 fb 1 55 +24 15 44 +19 12 4.4 +1.9 1.1 Standard Model expectation B (H J/ψγ) [ 10 6 ] B (Z J/ψγ) [ 10 7 ] 2.9 ± 0.2 0.80 ± 0.05 60 80 100 120 140 0 180 m µµγ (GeV) Events / ( 2 GeV ) 00 Simulation Preliminary 1400 s=14 TeV, 3000 fb 1200 1000 800 Optimistic scenario with MVA analysis still only sensitive to B (H J/ψ γ) at 15 SM value with 3000 1 fb 1 (Bkg. Only) S+B Fit Background H Signal 100 Z Signal 10 New ideas likely required to reach SM sensitivity in a HL-LHC scenario with this channel! 600