Search for Higgs Boson Decays to a Meson and a Photon

Search for Higgs boson decays to a meson and a photon

Konstantinos Nikolopoulos University of Birmingham on behalf of the ATLAS and CMS Collaborations

Precision Workshop - Rencontres du Vietnam ATLAS experiment at CERN 26 September 2016, Quy Nhon, Vietnam This project has received funding from the European Union’s 7th Framework Programme for research, technological development and demonstration under grant agreement no 334034 (EWSB) Higgs Mechanism: Scalar Couplings Structure Higgs-fermion interactions: YukawaBosonic sector:couplings + V EWSB gives mass to W , W , Z bosons H • Higgs interactions to vector bosons: defined by symmetryHiggs breaking couplings proportional to m2 gHV V • W /Z Higgs interactions to fermions: ad-hoc hierarchical Yukawa couplings∝mf 2m2 g = V V HVV v Higgs Mechanism: Scalar Couplings Structure Fermionic sector: Bosonic sector: f + V EWSB gives mass to W , W , Z bosons H After introducting Higgs field, can add • H • Higgs couplings proportional to m2 gHV V gHff¯ Yukawa terms to Lagrangian • W /Z 2 m Higgs couplings proportional to fermion mass 2 2mV f • 2mV ghV V = V ghff¯ = f¯ gHVV = mf v gHf f¯ = Yf = v Fermionic sector: v is Higgs field vacuum expectation value f • After introducting Higgs field, can add Loops (e.g. ,gluon)sensitivetoBSMphysics H • Yukawa couplings not imposed• by fundamental principle gHff¯ Yukawa terms to Lagrangian Higgs couplings Enhanced proportional Yukawa to fermion couplings mass in BSM scenarios • 4of39 ¯ [Phys. Rev. D80, 076002, Phys. Lett. B665 (2008) 79, Phys.Rev. D90 (2014) 115022,…] f mf Unitarityg ¯ = Y = bounds (through EFT) for fermion mass Hf f f v generation scale (1st/2nd generation) v is Higgs field vacuum expectation value • Loops (e.g. ,gluon)sensitivetoBSMphysics • 3 ⇤ <20 TeV 4of39 ⇠ smf

[Phys. Rev. Lett. 59, 2405 (1987); Phys.Rev. D71 (2005) 093009]

K. Nikolopoulos / Quy Nhon, 26 Sep 2016 / Search for Higgs boson decays to a meson and a photon 2 Higgs-fermion interactions: The story so far

4.5(3.4)σ

ATLAS-CONF-2015-044

JHEP 1504 (2015) 117 Progress in Higgs boson properties: mass known to 0.19% bosonic decays measured to ~20-30% In fermion sector, different picture: 3.2(3.5)σ τ-lepton: direct evidence by ATLAS and CMS for h→ττ e,µ: no evidence→non-universality t-quark: no firm evidence for ttH; indirect evidence b-quark: no evidence for h→bb in LHC; mild excesses c-quark: no direct evidence, loose bounds from h→bb

u/d/s-quarks: no direct searches available JHEP 1405 (2014) 104

K. Nikolopoulos / Quy Nhon, 26 Sep 2016 / Search for Higgs boson decays to a meson and a photon 3 We take m =125.9 0.4 GeV, and we obtain Γ(H γγ)=9.565 10−6 GeV from H ± → × the values of the Higgs-boson total width and branching fraction to γγ in Refs. [11, 12]. We estimate the uncertainties in the indirect amplitude along the lines that were suggested in footnote 2 of Ref. [8]. In Γ(H γγ), we take the uncertainty from uncalculated higher- → order corrections to be 1%, and the uncertainties that arise from the uncertainties in the

top-quark mass mt and the W -boson mass mW to be 0.022% and 0.024%, respectively. We take the uncertainties in theExclusive leptonic decay widths Decays to be 2.5% for h the→J/ψQandγ 1.3% for h→Qγ decays:the a Υclean. We probe estimate on the Yukawa uncertainties couplings in the of indirect 1st and amplitude 2nd generation from uncalculated quarks mass Q is a vector meson or quarkonium2 2 state corrections to be mV /mH . We have not included the effects of the uncertainty in mH , as it Two contributions: direct and indirect amplitude is expected that that uncertainty will be significantly reduced in Run II of the LHC. Direct amplitude: provides sensitivity to Yukawa couplings The uncertainties in the direct amplitude arise primarily from the uncertainties in φ , Indirect amplitude: larger contribution than direct amplitude 0 0 2 FIG. 1: The Feynman diagrams2 for the direct amplitude2 for H V + γ4 at order αs.Theshaded Destructive interferencev , and uncalculated corrections of order αs,orderαsv ,andorder→ v . We estimate the ⟨ ⟩ blob represents the quarkonium wave function. The momenta that are adjacent to the heavy-quark order-α2 correction to be 2%, the order-α v2 correction to be 5% for the J/ψ and 1.5% for s “Direct” contributionlines are defined in the text.s “Indirect” contribution the Υ,andtheorder-v4 correction to be 9% for the J/ψ and 1% for the Υ. The uncertainties Small angular separation of decay products in the direct amplitude that arise from the uncertainties in mc and mb are 0.6% in the case of the J/ψ and 0.1% in the case of the Υ, and so they are negligible in comparison withthe other uncertainties in the direct amplitude. Our results for the widths are7 meson decay 2 −10 Γ(H J/ψ + γ)= (11.9 0.2) (1.04 0.14)κc 10 GeV, (53a) products → FIG. 2: The Feynman± diagram− for the indirect± amplitude for×H V + γ.Thehatchedcircle Phys.Rev.2 →D90 (2014) 11, 113010 Higgs Γ[H Υ(1S)+γ]=!(3.33 0.03) (3.49 0.15)κ! 10−10 GeV, (53b) → represents top-quark! ± or W -boson− loops, and± the shaded!b blob× represents the quarkonium wave function. ExclusiveΓ[H Υ decays(2S)+γ ]=lead!(2 to.18 distinct0.03) experimental(2.48 0.11)κ !2 signatures10−10 GeV, (53c) → ! ± − ± b! × In the direct process, the Higgs boson decays into a heavy quark-antiquark (QQ¯) pair, photon High-pT isolated •quarkonium recoiling against 2 −10 Γ[H Υ(3S)+γ]=!(1.83 0.02) (2.15 0.10)κb! 10 GeV. (53d) high-pT→ isolated photonone of! which radiates± a photon− before± forming a quarkonium! × with the other element of the! pair. ! The SM values for the widths (κQ! =1)are ! K. Nikolopoulos / Quy Nhon, 26 Sep 2016 / Search for Higgs bosonIn the decaysindirect to process a meson, the Higgs and bosona photon decays through a top-quark loop or a vector-4 • ∗ ∗ boson loop to a γ and a γ (virtual+0.05 photon).− The8 γ then decays into a vector quarko- ΓSM(H J/ψ + γ)=1.17−0.05 10 GeV, (54a) nium.→ × Γ [H Υ(1S)+γ]=2.56+7.30 10−12 GeV, (54b) SMThe Feynman→ diagrams for the direct−2.56 and× indirect processes are shown in Figs. 1 and 2, Γ respectively.[H Υ(2 ItS is)+ the quantumγ]=8 interference.46+7.79 between10−12 theseGeV two, processes that provides phase(54c) SM → −5.35 × 3 Γ [H Υ(3S)+γ]=10.25+7.33 10−12 GeV. (54d) SM → −5.45 ×

7 We do not include results for the ψ(2S) because a value for v2 [ψ(2S)] does not exist in the literature ⟨ ⟩ and because it is likely that v2 for the ψ(2S) is so large that the theoretical uncertainties in the width would be very large.

18 and κb ∆Υ(1S) = (0.948 0.040) + i(0.130 0.019) eff +0.0184 0.0015i, ± ± κγγ − ! " κb ∆Υ(2S) = (1.014 0.054) + i(0.141 0.022) eff +0.0207 0.0015i, (43) ± ± κγγ − # $ κb ∆Υ(3S) = (1.052 0.060) + i(0.148 0.025) eff +0.0221 0.0015i. ± ± κγγ − ! " Approximate expressions forκ ¯ρ0 ,¯κω andκ ¯φ have been given in (22) and (23). The constant terms in the above results show the tiny power-suppressed corrections. Only for the Υ(nS) states they reach the level of percent. Our complete expressionsfortheCP-oddcoefficients ∆˜ V are also given in Appendix E. It is a good approximation to only keep the direct contributions in these terms, which are likely to give rise to the dominant effects. Their coefficients are the same as in the expressions above, but withκ ¯q replaced by κ˜¯q and κb replaced byκ ˜b. It is interesting to compare our result for the quantities ∆V with corresponding expressions obtained by other authors. From [10] one can extract ∆ρ0 =(0.095 0.020) (2¯κu +¯κd)/3, ∆ =(0.092 0.021) (2¯κ +¯κ )and∆ =(0.130 0.027)¯κ , while from± [32] one can obtain ω ± u d φ ± s ∆ =(0.392 0.053)¯κ , ∆ =(1.048 0.046)κ , ∆ =(1.138 0.053)κ and ∆ = J/ψ ± c Υ(1S) ± b Υ(2S) ± b Υ(3S) (1.175 0.056)κb. These values are systematically higher than ours due to the fact that these authors± have not (or not fully) included QCD radiative corrections and RG evolution effects in the direct contributions. For the Υ(nS) states it is important to keep the small imaginary parts of the direct contributions, since in theExclusive SM the real parts alm Decaysost perfectly cancel h→ inQ theγ combinations 1 ∆V in (37). The result for ∆ω obtained in [10] misses the contribution from ω φ mixing and− contains a sign mistake in front ofκ ¯ . Note also that our predictions for the Substantial− recent% interest% from the theory communityd regarding branching ratio estimatesvector∆V parameters mesons, and we %feasibility: of find light% mesons are significantly more accurate than those obtained in [10]. 6 Br(h J/ψγ)=(2.95 0.07f 0.06direct 0.14h γγ) 10− , → ± J/ψ ± ± → · 4Phenomenologicalresults+1.75 9 Br(h Υ(1S) γ)=(4.61 0.06fΥ(1S) 1.21 direct 0.22h γγ) 10− , → ± − ± → · (45) We begin by quoting our benchmark results for the+0.75h V γ branching fractions9 in the SM. Br(h Υ(2S) γ)=(2.34 0.04fΥ(2S) 0.99 direct→ 0.11h γγ) 10− , For a Higgs mass→ of mh =(125.09 0±.024) GeV,− the SM value± of the→ h · γγ branching ratio 3 ± +0.75 → 9 is (2.28 0.Br(11)h 10−Υ(3[57].S) γ Using)=(2 this.13 result,0.04f weΥ(3S obtain) 1.12 direct for the0 decays.10h γγ into) 10 light− . vector mesons ± · → ± − ± → · JHEP 1508 (2015) 012 In these cases there is an extra source0 of theoretical uncertainty related to5 the calculation of the Br(h ρ γ)=(1.68 0.02fρ 0.08h γγ) 10− , direct contribution to the decay→ amplitude. Note± that± there is→ an· almost perfect cancellation 6 between the direct and indirectBr(h contributionsωγ)=(1.48 to0.03 thefω h 0.07Υh(nSγγ))γ 10decay− , amplitudes, and(44) as → ± ±→ → · a consequence the resulting branching ratios are roughly three orders6 of magnitude smaller Br(h φγ)=(2.31 0.03f 0.11h γγ) 10− , → ± φ ± → · than the h J/ψγ branching fraction. For comparison, we note that the branchingJHEP ratios 1504 (2015) 101 → +0.16 6 +1.74 9 +1.86 9 foundwhere in we [32] quote read separately (2.79 0.15) the10 uncertainties− for J/ψ,(0 due.61Decay to0 the.61 mode) vector-m10− foresonΥ(1 decayS), (2Branching constant.02 1.28)f ratioV10and− for the asymptotic LO +1.75 − 9 · − · − · Υ(2hS)and(2γγ branching.44 1.30 ratio,) 10− thefor latterΥ(3S being). We the find dominant good0 agreement0 source of with uncertain+0.09 the resultsty. Our reported predictions by 12 Z π γ (9.80 0.14 µ 0.03f 0.61a2 0.82a4 ) 10− 7.71 14.67 Z→Q→γ decays −also ·interesting − theseare authors systematically except for lower the and decay moreh accurateΥ(1S) γ than, where0 → those0 their obtained value+0 is. in about04 [1±0], a where factor± the7 smaller± values · 9 Z ρ γ (4.19 0.06 µ 0.16f 0.24a2 0.37a4 ) 10− 3.63 5.68 Experimentally0 unconstrained 5→ − 6 thanBr( ours.h Theρ γ)=(1 reason.9 is that0.15) we10 do− not,Br( neglecth ωγ the0)=(1→ imaginary.6 0 part.17)+0 o.f1003 the−± directand Br(± contributionh ±φγ)= · 8 → 6 ± · →2 Z ωγ ± (2.89 ·0.05 µ 0.15f 0.29→a2 0.25a4 ) 10− 2.54 3.84 − to LEP(3∆.Υ0(1 accuratelyS)0in.13) (42),10 which− measuredare prevents quoted. b-/c-quark While1 ∆Υ the(1S ) firstfrom two0 → becoming results are arbitrarily compatible+0.08 ± small. with± ours± within · 9 ± · − Z φγ (8.63 0.13 µ 0.41f 0.55a2 0.74a4 ) 10− 7.12 12.31 errors, there is a significant difference for the important mode h − φγ. For decays into heavy couplingsOur predictions (~1%) may also be compared with the0 → upper limits obtained+0±.14 from± a recent+0.39± first 8· ! ! Z J/ψγ →(8.02 0.15 µ 0.20f 0.36 σ) 10− 10.48 6.55 − − analysis light quark of these couplings rare decays less reported constrained! by the ATLAS ! 0 → collaboration.TheyareBr(+0.10 ±h J/+0.ψγ11 )·< 8 3 3 Z Υ(1S) γ (53.39 0.10 µ 0.08→f 0.08 σ) 10− 7.55 4.11 1.5 10 , Br(h Υ(1S) γ) < 1.3 10 ,Br(h Υ→(2S) γ) < 1.9 10 and− Br(h± Υ(3− S) γ)·< Sensitive− to BSM contributions − 0 − +0.02 +0.02 8 · 3 → · →Z Υ(4S) γ · (1.22 0.02 µ →0.13f 0.02 σ) 10− 1.71 0.93 1.3 10 , all at 95% CL [20]. It will require18 an improvement by a factor− 500 to− become − 0 → +0.18 ± +0.20 · 8 · Z Υ(nS) γ (9.96 0.19 µ 0.09f 0.15 σ) 10− 13.96 7.59 sensitive to the h J/ψγ mode in the SM, while→ the SM branching fractions− ± for the− decays· → K. Nikolopoulosh Υ(nS / )Quyγ are Nhon, out 26 ofSep reach 2016 / atSearch the for LHC. Higgs Nevertheless, bosonTable decays 4: Predicted to as a wemeson will branching and discuss a photon fractions below, forthese various decayZ Mγ 5decays, including error → 1 → modes allow for very interesting new-physics searches.estimates due With to scale 3 ab− dependenceof integrated (subscript luminosity, “µ”) and the uncertainties in the meson 8 about 1.7 10 Higgs bosons per experiment willdecay have constants been (“producedf”), the Gegenbauer by the end moments of the high- of light mesons (“an”), and the width × luminosity LHC run [11]. If the J/ψ is reconstructedparameters via of its heavy leptonic mesons decays (“σ”). into See textmuon for pairs, further explanations. + 7 the effective branching ratio in the SM is Br(h J/ψγ µ µ−γ)=1.8 10− , meaning that → → · + about 30 events can be expected per experiment. If also the decays into e e− can be used, p2 m2 then ATLAS and CMS can hope to collect aour combined case, on the sample other of hand, about= 120Z events.is equal A to detailed the mass of the decaying heavy gauge boson, in which case the above expression does not exhibit a 1/k2 pole, but is instead proportional discussion of the experimental prospects for reconstructing these events over the background to 1/m2 . Hence we conclude that A = 0 in (68). Note that in the limit k2 0 one obtains can be found in [9]. Concerning the h φγ decayZ mode, a reconstruction efficiency ϵ =0.75 → → from (69) φγ was assumed for the φγ final state in [10], which appears to us as an optimistic1 1 m assumption.2 + ln Z iπ +const. , (70) In the SM one expects about 400ϵφγ events per experiment in this mode,m2 ϵ meaningµ2 − that the Z ! " two experiments can hope to look at a combinedwhich is sample precisely of of several the form hu ofndred our (bare) events. hard-scattering Likewise, coefficients. 0 in the SM one expects about 2900ϵ 0 events per experiment in the decay mode h ρ γ. ρ γ → In Figure 6 we show our predictions for the3.4 ratio Phenomenological of branching fractions results (times 1000) defined eff ¯ eff in (37) in the plane of the parametersκ ¯V /κγγ and κ˜V /κγγ. We focus on the most interesting cases V = φ, J/ψ and Υ(1S). The correspondingWe are plotsnow ready for V to= presentρ0, ω would detailed look numerical very similar predictions for the various radiative decay to that for V = φ (apart from the overall scalemodes. of the We branching start with fractions), the decays of while the Z theboson, plots using for relation (35). Besides the input parameters already mentioned, we need the Z-boson mass m =(91.1876 0.0021) GeV and higher Υ(nS) resonances would look very similar to that for the Υ(1S) meson. For orientation, Z ± total width ΓZ =(2.4955 0.0009) GeV [45]. When squaring the decay amplitudes, we expand we mention that a value of 0.4 in these plots corresponds to a h ± V γ branching fraction of the resulting expressions consistently to first order in αs. The imaginary parts of the form 6 → about 10− , assuming that the h γγ branchingfactors in fraction (42) do is not SM enter like. at This this assumption order. Our results will be are presented in Table 4. Significant → implicit whenever we quote absolute branchinguncertainties ratios in below; our predictions otherwisethequotednumbers arise from the hadronic input parameters, in particular the meson decay constants (see Appendix B) and the various Gegenbauer moments. Their impact is explicitly19 shown in the table. Our error budget also includes a perturbative uncertainty, which we estimate by varying the factorization scale by a factor of 2 about the default value µ = mZ . All other uncertainties, such as those in the values of Standard Model parameters, are negligible. Note also that power corrections from higher-twist LCDAs are bound to be 2 2 negligibly small, since they scale like (ΛQCD/mZ ) for light mesons and at most like (mM /mZ ) 11 0 0 for heavy ones. The predicted branching fractions range from about 10− for Z π γ to 7 0 → about 10− for Z J/ψγ. In the last row, the symbol Υ(nS)meansthatwesumover the first three Υ states→ (n =1, 2, 3). Strong, mode-specific differences arise foremost from the

26 The LHC, ATLAS and CMS

LHCb ATLAS

CMS ALICE

LHC

High Quality data collected!

K. Nikolopoulos / Quy Nhon, 26 Sep 2016 / Search for Higgs boson decays to a meson and a photon 6 h/Z→J/ψγ and h/Z→Y(nS)γ (n=1,2,3)

Phys.Rev.Lett. 114 (2015) 121801 FIG. 1: The Feynman diagrams for the direct amplitude for H V + γ at order α0.Theshaded → s K. Nikolopoulos / Quy Nhon, 26 Sep 2016 / Search for Higgs boson decays to a meson and ablob photon represents the quarkonium wave function. The7 momenta that are adjacent to the heavy-quark lines are deﬁned in the text.

FIG. 2: The Feynman diagram for the indirect amplitude for H V + γ.Thehatchedcircle → represents top-quark or W -boson loops, and the shaded blob represents the quarkonium wave function.

In the direct process, the Higgs boson decays into a heavy quark-antiquark (QQ¯) pair, • one of which radiates a photon before forming a quarkonium with the other element of the pair.

In the indirect process, the Higgs boson decays through a top-quark loop or a vector- • boson loop to a γ and a γ∗ (virtual photon). The γ∗ then decays into a vector quarkonium.

The Feynman diagrams for the direct and indirect processes are shown in Figs. 1 and 2, respectively. It is the quantum interference between these two processes that provides phase

3 h/Z→J/ψγ and h/Z→Y(nS)γ: Mass Resolution

Y(1S)

Y(2S)

Y(3S) Simple event categorisation 4 detector-driven categories Muon pseudorapidity (×2) Photon conversion (×2) Mass resolution ~1.2-1.8%

unconverted/barrel unconverted/end-cap

Phys.Rev.Lett. 114 (2015) 121801 K. Nikolopoulos / Quy Nhon, 26 Sep 2016 / Search for Higgs boson decays to a meson and a photon 8 h/Z→J/ψγ and h/Z→Y(nS)γ: Background Inclusive quarkonium with jet “seen” as γ combinatoric background: small contribution contribution from Q+γ production Nonparametric data-driven background model Begin with loose sample of candidates Model kinematic and isolation distributions Generate “pseudo”-background events Apply selection to “pseudo”-candidates Y(nS)γ: also Z→µµγFSR from side-band fit

Phys.Rev.Lett. 114 (2015) 121801

K. Nikolopoulos / Quy Nhon, 26 Sep 2016 / Search for Higgs boson decays to a meson and a photon 9 Systematics Uncertainties - Signal and Background h/Z→J/ψγ and h/Z→Y(nS)γ: Systematics

Signal Yield Uncertainty: Several sources of systematic uncertainty on the H and Signal Yield Uncertainty: Several sources of systematic uncertainty on the h and Z Z signal yields are considered, all modeled with nuisance parameters in likelihood: signal yields are considered, all modelled by nuisance parameters in likelihood:

Source Signal Yield Uncertainty Estimated From Total H cross section 12% QCD scale variation and Total Z cross section 4% PDF uncertainties Calibration observable and Integrated Luminosity 2.8% vdM scan uncertainties Trigger Eciency 1.7% Photon ID Eciency Up to 0.7% Data driven techniques with + + Muon ID Eciency Up to 0.4% Z ` `, Z ` ` and ! + ! J/ µ µ events Photon Energy Scale 0.2% ! Muon Momentum Scale Negligible

Background Shape Uncertainty: Estimated from modifications to modeling procedure (e.g. shifting/warping input distributions), shape uncertainty included in likelihood as a shape morphing nuisance parameter Search for rare Higgs and Z boson decays to 32 / 32

K. Nikolopoulos / Quy Nhon, 26 Sep 2016 / Search for Higgs boson decays to a meson and a photon 10 h/Z→J/ψγ and h/Z→Y(nS)γ: Results

Multi-observable fit mµµγ, pTµµγ for J/ψγ mµµγ, pTµµγ,mµµ for Υ(nS)γ No significant excess above background observed

Phys.Rev.Lett. 114 (2015) 1121801

K. Nikolopoulos / Quy Nhon, 26 Sep 2016 / Search for Higgs boson decays to a meson and a photon 11 h/Z→J/ψγ and h/Z→Y(nS)γ: Results

Theory

Phys.Rev.Lett. 114 (2015) 12, 121801 95% CL upper limits on decay Branching Ratios: �(10-3) for Higgs boson (SM production) �(10-6) for Z boson Indicate non-universal Higgs boson coupling to quarks [Phys.Rev. D92 (2015) 033016, JHEP 1508 (2015) 012]

K. Nikolopoulos / Quy Nhon, 26 Sep 2016 / Search for Higgs boson decays to a meson and a photon 12 CMS h→J/ψγ

CMS search for h→J/ψγ extending the h→llγ studies used 19.7 fb-1 at 8 TeV Phys.Lett. B753 (2016) 341 Event Selection single muon and a photon, both pT>22 GeV |ηµ|<2.4, pTµ>23,4 GeV, pTµµ>40 GeV Dimuon invariant mass in vicinity of J/ψ |ηγ|<1.44, pTγ>40 GeV

µµ and γ isolation, ×103 0.9 fb-1 (13 TeV, 2016) 2.9 < mµµ < 3.3 GeV 160 CMS Preliminary ΔR(µ,γ)>1 for each muon 140 σ = 30MeV + - pµ µ > 16GeV 120 T muon impact parameter requirements DP-2016-027 Events/ 7 MeV 100 transverse <2mm 80 longitudinal <5mm 60 40 20 0 2.95 3 3.05 3.1 3.15 3.2 3.25 3.3 µ+µ- invariant mass [GeV]

K. Nikolopoulos / Quy Nhon, 26 Sep 2016 / Search for Higgs boson decays to a meson and a photon 5 13 CMS h→J/ψγ

background-only fit to the data

Fit over the 110-150 GeV mass range. Background: 2nd degree polynomial Signal: Crystal Ball + Gaussian Phys.Lett. B753 (2016) 341 No excess above background observed 95% CL upper limit BR(H→J/ψγ) < 1.5x10-3 → 540 times the SM prediction

K. Nikolopoulos / Quy Nhon, 26 Sep 2016 / Search for Higgs boson decays to a meson and a photon 14 Search for h/Z→φγ � New ATLAS analysis based on 2.7 fb-1 at 13 TeV collected in 2015 _ Higgs decay h→φγ sensitive to strange quark Yukawa coupling probing light quark Yukawa coupling was considered impossible at the LHC very challenging to access with inclusive H→ss decays! Looking for new physics through anomalous H→ss couplings possible in various BSM scenarios, would modify BR(h→φγ) Z→φγ not directly constrained by existing measurements Phys. Rev. Lett. 117, 111802 week ending PRL 114, 101802Supplementary (2015) Information:PHYSICAL http://atlas.web.cern.ch/Atlas/GROUPS/PHYSICS/PAPERS/HIGG-2016-05/ REVIEW LETTERS 13 MARCH 2015 implies real κ¯q, κV;γ, and κ¯qq κ¯Ã . Note that κ¯q and κ¯qq are 0 ¼ q0q 0 normalized to theThe SM ideab-quark is to Yukawa benefit coupling. from the The SM limit correspondsinterference to κγ κV 1of, and theκ¯ s“direct”ms=m andb 0.020, ¼−3 ¼ ¼ ≃ −4 κ¯d md=mb 1.0 ×“indirect”10 , κ ¯amplitudes!u mu=mb 4.7 × 10 . The¼ quark masses≃ are evaluated at¼ μ m ≃using NNLO Phys.Rev.Lett. 114 (2015) 101802h running in the MS¯ scheme with low-energy¼ inputs from

Ref. [14]. The κ¯qq0 vanish in the SM. Any deviations from these relations would signal the presence of new physics. Constraints from the current data.—In Ref. [7] the LHC inclusive production rate was used to place an indirect FIG. 1 (color online). Direct-amplitude diagram (left)6 and BR(h )=(2indirect-amplitude.31 0.03f diagram (right)0.11 contributingh ) to10h →ϕγ. bound on the charm Yukawa coupling. Here,! we adapt this ± ± ! · analysis to the otherJHEP 1508 Yukawa (2015) 012 couplings, κ¯ . The current i a multiplet that approximately conserves the flavor sym- ATLAS [15],CMS[16], and Tevatron [17] Higgs measure- metries, its contributions could be (partially) canceled by ments are includedK. Nikolopoulos (based on / Quy Tables Nhon, 13 26 and Sep 14 2016 of Ref./ Search[18] for), Higgs boson decays to a meson and a photon 15 other members of the multiplet. The latter could mostly as are the indirect constraints from the LEP electroweak decay to light quarks or have reduced production rates, thus precision measurements [19].Forsimplicity,correlations remaining unobserved. between the different measurements are neglected and Flavor-conserving photonic decays.—We begin with asymmetric uncertainties are symmetrized. The quark anti- h → ϕγ. The decay amplitude receives two dominant quark Higgs-fusion cross section is evaluated at next-to- contributions, which we denote as direct and indirect; leading order in α based on the bottom fusion cross section s see Fig. 1. The indirect contribution proceeds through obtained in Ref. [20] using MSTW parton distribution the hγγ coupling, followed by the fragmentation of γ → ϕ. functions [21]. Below, we check that our fit results are Ã In our analysis, we use the on-shell h → γγ amplitude (2). stable against uncalculated higher-order corrections by The error due to this is small, O m2 =m2 . Similarly, the varying our production cross sections by 40%, the estimated ð ϕ hÞ theoretical error at next-to-leading order [20]. The resulting indirect contribution from h → γZ is neglected, because it is suppressed by the off-shell Z. The direct amplitude shifts in the bounds on the κ¯i are extremely small. We begin with the flavor-conserving couplings. A naive involves a hard h → ss¯γ vertex, where an intermediate 2 s-quark line with an off-shellness Q2 ∼ O m2 is integrated χ fit to the data that fixes all Higgs couplings to their SM ð hÞ values, except for one of the up, down, or strange Yukawa out. Its evaluation is a straightforward application of QCD couplings at a time, leads to the 95% confidence level factorization [25]. The largest sensitivity to the Higgs– (C.L.) bounds strange quark coupling is due to the interference of the two amplitudes. (The direct amplitude by itself yields −11 κ¯u < 1.0; κ¯d < 0.9; κ¯s < 0.7: 3 BRh→ϕγ ∼ 10 in the SM.) However, the interference j j j j j j ð Þ only involves the real part of the coupling, Re κ¯ . Working ð sÞ If all of the Higgs couplings (including h → WW; in the limit of real κ¯s, the h → ϕγ decay amplitude is ZZ; γγ; gg; Zγ;bb¯ and ττ¯) are allowed to vary from their 2 SM values, we get the weaker 95% C.L. bounds ϕ Qse ϕ γ mb ϕ ϕ 4α fϕmh Mss ϵ × ϵ κ¯s f 1=uu¯ κγAγ ; ¼ 2 v ⊥h i⊥ þ πv mϕ κ¯u < 1.3; κ¯d < 1.4; κ¯s < 1.4: 4 j j j j j j ð Þ 6 ð Þ The 95% C.L. upper bounds obtained for the off- diagonal couplings, when modifying only a single where the first and second terms are the direct and indirect ϕ ϕ Yukawa coupling at a time (or allowing for modification contributions; f and 1=uu¯ are the decay constant and h i of the other Higgs couplings as above), are inverse moment⊥ of the light-cone⊥ distribution amplitude (LCDA) defined in Eq. (8), Qse −e=3 is the strange κ¯ < 0.6 1 ; 5 quark electric charge, and ε and¼ε are the γ and ϕ j qq0 j ð Þ ð Þ γ ϕ polarization vectors. We have used the definition μ μ for q; q0 ∈ u; d; s; c; b and q ≠ q0. The bounds are 10%– ϕ J 0 0 f m ϵ for the ϕ decay constant f , h j EMð Þj i ¼ ϕ ϕ ϕ ϕ 20% stronger for couplings only involving sea quarks, as where Jμ Q f¯γμf is the electromagnetic current. their slightly smaller direct production cross section does EM ¼ f f Note that for CP-violating couplings Mϕ is sensitive to the not compensate for the increased decay width. P ss Inclusive Higgs rate measurements cannot distinguish phase between Aγ and κ¯γ. The LCDA convolution integral is between the individual κ¯qq0 . The weakest indirect bound from low-energy observables is found to be κ¯ <8×10−2 bs 1 ϕ j j ϕ ϕ u [22] (see also Refs. [23,24]). However, such bounds are 1=uu¯ du ⊥ð Þ : 7 model dependent. For instance, if the Higgs boson is part of h i⊥ ¼ 0 u 1 − u ð Þ Z ð Þ 101802-2 h/Z→φγ: Analysis Strategy

+ - Exploit φ→Κ Κ decays, BR=49% Small angular separation of Distinctive topology decay products Pair of collimated high-pT isolated tracks recoils against high-pT isolated photon meson decay Enabled by dedicated trigger (Sep 2015) products Higgs Photon (pTγ>35 GeV) and isolated di-track

(at least one pT>15 GeV) consistent with mφ photon Efficiency ~80% w.r.t offline selection Phys. Rev. Lett. 117, 111802

Z→µµ candidate with 25 reconstructed vertices from the 2012 run. Only good quality tracks with pT>0.4GeV are shown K. Nikolopoulos / Quy Nhon, 26 Sep 2016 / Search for Higgs boson decays to a meson and a photon 16 + Event Selection I - K K Selection ! h/Z→φγ: Event Selection

No ⇡/K/p particle ID available for tracks in relevant pT range - assume K hypothesis Tracks for all tracks No particle-ID available at momentum range, all tracks considered Kaons Leading track pT > 20 GeV and sub-leading track pT > 15 GeV

Require looking consistency for with two mass: oppositemKK chargedm < 20 MeV tracks; leading pT>20GeV, sub-leading pT>15 GeV | | Require di-track fractional trackconsistent isolation (additionalwith φ-meson tracks within massR within< 0.2) of 20 di-track MeV pair < 0track-based.10 isolation applied The di-track di-track system system transverse must momentum satisfy: must satisfy:

40 GeV, for mKK 91 GeV KK  p > 40 + 5/34 (mKK 91) GeV, for 91 GeV < mKK < 125 GeV T 8 ⇥ <>45 GeV, for mKK 125 GeV Photons:> “Tight” identification criteria pTγ>35 GeV Phys. Rev. Lett. 117, 111802 Search for H, Z with ATLAS 9 / 17 |ηγ|<2.47 and not in 1.37<|ηγ|<1.52! Isolated (calorimeter- and track-based) Δφ(K+K-,γ)>0.5 Total signal acceptance/efficiency h→φγ→KKγ ~ 18% Ζ→φγ→KKγ ~ 8%

Full event selection w/o mKK requirement

K. Nikolopoulos / Quy Nhon, 26 Sep 2016 / Search for Higgs boson decays to a meson and a photon 17 h/Z→φγ: Efficiency and Resolution

Phys. Rev. Lett. 117, 111802

Inclusive analysis Total signal efficiency: 18% for Higgs boson 8% for Z boson Muon Mass resolution ~1.8%

K. Nikolopoulos / Quy Nhon, 26 Sep 2016 / Search for Higgs boson decays to a meson and a photon 18 h/Z→φγ: Background

Dominated by QCD production γ+jet and multi-jet events Exclusive “peaking” backgrounds (e.g. h/Z→µµγFSR) estimated to be negligible Nonparametric data-driven model; same procedure as in h/Z→J/ψγ

Background Region loose selection (loose +PTKK>45 GeV selection)

Phys. Rev. Lett. 117, 111802

loose selection loose selection +γ-isolation +ΚΚ-isolation

K. Nikolopoulos / Quy Nhon, 26 Sep 2016 / Search for Higgs boson decays to a meson and a photon 19 Systematic Uncertainties

Signal Yield Uncertainty: Several sources of systematic uncertainty on the H and Z signal yields are considered, all modeled with nuisance parameters in likelihood: h/Z→φγ: Results

Source H/Z Yield Uncertainty Estimated From QCD scale variation and Total H cross section 12% PDF uncertainties Total Z cross section 5.5% ATLAS Measurement Calibration observable and Integrated Luminosity 5% vdM scan uncertainties Photon ID Eciency 2.5% Data driven techniques with Photon Energy Scale 0.3% + + Z ` ` and Z ` ` Trigger Eciency 2% ! ! Tracking Eciency 6% Tracking studies within dense jets

Background Shape Uncertainty: Estimated from modiﬁcations to modeling T procedure (e.g. shifting pKK and neglecting the weakest correlation included in the model),Final shape uncertainty discriminant included in likelihood is m asΚΚγ a shape morphing NP Method described in: EPJC 73 (2013) 2518 (arXiv:1302.4393) † 95% confidence levelSearch upper for rare Higgs limit and Z boson using decays to CLs31 / 32 with profile likelihood test statistic Largest observed excess at ~100GeV; 2σ effect No significant H or Z signal observed, Branching ratio limits at the level of 10-3 (10-6) for Higgs (Z) boson decays

Phys. Rev. Lett. 117, 111802

K. Nikolopoulos / Quy Nhon, 26 Sep 2016 / Search for Higgs boson decays to a meson and a photon 20 h/Z→Qγ: in the future

HL-LHC is a Higgs boson factory

�(200M) Higgs bosons ATLAS-PHYS-PUB-2015-043 ATLAS HL-LHC projections for h/Z→J/ψγ Nice and, relatively, clean final state Small branching ratio, few events expected At SM sensitivity large contribution from h→µµγFSR~3×h→J/ψγ and (Z→µµγFSR for Z) Sensitive to “anomalous” h→γγ loop; use ratio to h→γγ The results presented in Tables 2 and 4 demonstrate that the introduction of a simple multivariate analysis provides a 20% improvement in the expected100 limits. ATLAS ATLAS Simulation 900 Simulation Preliminary Preliminary Expected branching ratio limits=14 at TeV, 95% 300 CL fb-1 800 s=14 TeV, 3000 fb-1 6 7 (H 80J/ ) [ 10 ] (Z DataJ/ (Bkg.) [ Only) 10 ] Data (Bkg. Only) B ! B !S+B Fit 700 S+B Fit Cut Based Multivariate Analysis CutBackground Based Background 1 +81 Events / ( 2 GeV ) +69 +2.7 Events / ( 2 GeV ) 300 fb 185 153 H7 Signal.0 × 100 600 H Signal × 100 52 60 43 2.0 1 +24 +19 Z Signal+1. 9× 10 Z Signal × 10 3000 fb 55 15 44 12 4.4 1.1 500 Standard Model expectation 400 40 6 7 (H J/ ) [ 10 ] (Z J/ ) [ 10 ] B ! B ! 300 2.9 0.2 0.80 0.05 ± ± 20 200 Table 2: The expected branching ratio limit at 95% CL for 300 fb 1 and 3000 fb 1 scenarios. The Standard Model 100 expectations are also reported for comparison. Not reviewed, for internal circulation only 60 80 100 120 140 160 180 60 80 100 120 140 160 180

Expected branching ratio limit atm 95%µµ γ (GeV) CL m µµ γ (GeV) Bkgd. Syst. Unc. Scenario 2% Not reviewed, for internal circulation only 180 K. Nikolopoulos / Quy Nhon,6 26 Sep 2016 / Search for Higgs boson decays to a meson and a photon 21 (H J/ ) [ 10 ] Median 1 2 B ! ATLAS Simulation 1600 ATLAS Simulation 160 +21 +51 Cut Based Analysis 52 Preliminary14 24 Preliminary +18 s=14 TeV,+43 300 fb-1 1400 s=14 TeV, 3000 fb-1 Multivariate Analysis140 43 12 20 Data (Bkg. Only) Data (Bkg. Only) 7 (Z J/ ) [ 10 ] Median 1 S+B2 Fit 1200 S+B Fit B ! 120 +1.7 Background+3.7 Background Cut Based AnalysisEvents / ( 2 GeV ) 4.3 1.2 2.0 Events / ( 2 GeV ) H Signal × 100 1000 H Signal × 100 100 Z Signal × 10 Z Signal × 10 1 Table 3: Comparison of the expected branching ratio limit at 95% CL for 3000 fb , assuming the alternative800 background systematic uncertainty scenario. 80 60 600 Expected limit at 95% CL 40 ⇥ B 400 (pp H) (H J/ ) [fb] ! ⇥ B ! Cut20 Based Multivariate Analysis 200 1 +2.9 +2.4 300 fb 10.4 4.5 8.6 3.7 1 0+0.9 20 40 60+0. 807 100 120 140 0 20 40 60 80 100 120 140 3000 fb 3.1 1.3 2.5 1.0 p µµ γ (GeV) p µµ γ (GeV) T T 1 Table 4: The expected limits at 95% CL on the Higgs cross section times branching fraction for 300 fb and 1 3000 fb scenarios. (a) (b)

+ µ µ m + p Figure 1: µ µ (upper plots) and T (lower plots) projections of the simultaneous ﬁt. The pseudo-data 1 1 correspond to the expected event yields for 300 fb (a) and 3000 fb (b). In the ﬁgure, for reference only, the Higgs and Z signal are shown assuming SM branching ratio enhanced by factors of 100 and 10, respectively.

5 Summary

100 ATLAS ATLAS Simulation 900 Simulation Preliminary Preliminary s=14 TeV, 300 fb-1 800 s=14 TeV, 3000 fb-1 80 Data (Bkg. Only) Data (Bkg. Only) S+B Fit 700 S+B Fit Background Background Events / ( 2 GeV ) H Signal × 100 Events / ( 2 GeV ) 600 H Signal × 100 60 Z Signal × 10 Z Signal × 10 500 400 40 300

20 200 100 Not reviewed, for internal circulation only 60 80 100 120 140 160 180 60 80 100 120 140 160 180

m µµ γ (GeV) m µµ γ (GeV) 180 ATLAS Simulation 1600 ATLAS Simulation 160 Preliminary Preliminary s=14 TeV, 300 fb-1 1400 s=14 TeV, 3000 fb-1 Yukawa140 sector is the least explored (and Data (Bkg. Only) Data (Bkg. Only) S+B Fit 1200 S+B Fit 120 Background Background motivated)Events / ( 2 GeV ) partH Signal ×of 100 theEvents / ( 2 GeV ) 1000 Standard ModelH Signal × 100 100 Z Signal × 10 st nd Z Signal × 10 800 →80 Particularly for 1 /2 generation.

60 600 LHC upgrade timescale 40 400 New20 Physics could be200 hiding here! HL-LHC upgrade proposed • 0 20 40 60 80 100 120 140 0 20 40 60 80 100 120 140 1 p µµ γ (GeV) p µµ γ (GeV) Goal to collect 3000 fb by 2035 T T A number of suggestions appearing in literature(a) currently: exclusive (b)decays, + µ µ m + p Figure 1: µ µ (upper plots) and T (lower plots) projections of the simultaneous ﬁt. The pseudo-data charm tagging, Higgs1 boson1 kinematic correspond to the expected event yields for 300 fb (a) and 3000 fb (b). In the ﬁgure, for reference only, the Higgs and Z signal are shown assumingproperties, SM branching ratio enhanced etc. by factors of 100 and 10, respectively.

New field of study in Higgs sector; experimental and theoretical ingenuity required to elucidate this corner of the SM!

5 CorrespondingRun I today proposals for upgradesRun II of the LHCRun III experiments HL-LHC •K. Nikolopoulos / Quy Nhon, 26 Sep 2016 / Search for Higgs boson decays to a meson and a photon 22 Central feature of ATLAS upgrade programme a new, all silicon tracking system

36 of 39 Additional Slides

K. Nikolopoulos / Quy Nhon, 26 Sep 2016 / Search for Higgs boson decays to a meson and a photon 23 3.1.2 Higgs production at hadron machines

In the Standard Model, the main production mechanisms for Higgs particles at hadron colliders make use of the fact that the Higgs boson couples preferentially to the heavy particles, that is the massive W and Z vector bosons, the top quark and, to a lesser extent, the bottom quark. The four main production processes, the Feynman diagrams of which are displayed in Fig. 3.1, are thus: the associated production with W/Z bosons [241,242], the weak vector boson fusion processes [112,243–246], the gluon–gluon fusion mechanism [185] and the associated Higgs production with heavy top [247,248]orbottom[249,250]quarks:

associated production with W/Z : qq¯ V + H (3.1) −→ 3.1.2 Higgs production at hadron machines vector boson fusion : qq V ∗V ∗ qq + H (3.2) −→ −→ gluon gluon fusion : gg H (3.3) In the Standard Model, the main production mechanisms for Higgs particles− at hadron −→ associated production with heavy quarks : gg,qq¯ QQ¯ + H (3.4) colliders make use of the fact that the Higgs boson couples preferentially to the heavy −→ particles, that is the massive W and Z vector bosons, the top quark and, to a lesser extent, q the bottom quark. The four main production processes,q¯ the Feynman diagrams ofV which areq V displayed in Fig. 3.1, are thus: the associated production with W/ZV ∗ bosons [241,242], the ∗ H weak vector boson fusion processes [112,243–246], the gluon–gluon fusion• mechanism [185] • and the associated Higgs production with heavy top [247,248]orbottom[249,250]quarks: V ∗ q H q SM Higgs boson3.1.2 Higgs production production at hadron3.1.2 machines Higgs production at atthe hadron machinesLHC q associated production with W/Z : qq¯ V + H (3.1) In the Standard Model, the mainIn production the Standard mechanisms Model,−→ the for main Higgs production particles at mechanisms hadron for Higgs particles at hadron vector boson fusion : g qq V ∗V ∗ qq + H (3.2)g Q colliders make use of the fact thatcolliders the Higgs make boson use of couples−→ the fact preferentially that−→ the Higgs to the boson heavy couples preferentially to the heavy particles, that is the massive W and Z vector bosons, the top quark and,H to a lesser extent,86% gluon gluonparticles, fusion that : is thegg massiveHWQ and Z vector bosons,(3.3) the top quark and, toH a lesser extent, − −→ mH = 125 GeV pp the bottom quark. The four mainthe production bottom processes, quark. The the four Feynman main¯ production diagrams• of processes, which are the Feynman• diagrams of which are H (NNLO+NNLL QCD associated+ NLO EW) production with heavy quarks : g gg,qq¯ QQ + H (3.4)g ¯ displayed in Fig.s 3.1,= 8 are TeV thus: thedisplayed associated in Fig. production 3.1, are w thus:−→ith W/Z thebosons associated [241,242], production the with W/Z bosons [241,242],Q the weak vector boson fusion processesweak [112,243–246], vector boson the fusion gluon–gluon processes fusion [112,243–246], mechanism the [185] gluon–gluon fusion mechanism [185] Figure 3.1: The dominant SM Higgs boson production mechanisms in hadronic collisions. 10 and the associated Higgs productionand with the heavyassociated top [247,248 Higgs production]orbottom[249,250]quarks: withq heavy top [247,248]orbottom[249,250]quarks: q¯ V q

LHC HIGGS XS WG 2014 V V There are also several mechanisms for the pair production of the Higgs particles H+X) [pb] associated∗ production with W/Z :associatedqq¯ productionV + H∗ with W/Z :(3.1)7%q q¯ V + H (3.1) −→ H −→ pp qqH (NNLO QCD + NLO EW) vector boson fusion : qq HiggsV ∗vectorV pair∗ productionbosonqq + H fusionmH : = :pp(3.2) 125qq GeVHHV+∗VX∗ qq + H (3.2)(3.5) • −→ •−→ −→−→ −→ pp gluon gluon fusion : gg HgluonV ∗ gluon fusion :(3.3)gg H (3.3) 1 pp WH (NNLO QCD + NLOq EW) − q −→ (pp bbH (NNLO QCD in 5FS, NLO QCD in 4FS) Hand the relevant sub–processes are− the gg HH mechanism,−→ which proceeds through heavy pp associated production with heavy quarks : gg,qq¯ QQ¯ + qH (3.4) ¯ ZH (NNLO QCD +NLO EW) associated production with heavy quarks→ : gg,qq¯ QQ + H (3.4) top and bottom quark loops−→ [251,252], the associated double−→production with massive gauge

bosons [253, 254], qq¯ HHV ,andthevectorbosonfusionmechanismsqq V ∗V ∗ pp → q q → → g q¯ HHqq [255,g 256];q see also Ref. [254].Q However, because of the suppression by the additional -1 ttH (NLO QCD) V q¯ V q electroweak couplings, they haveV much smaller production crossV sections than the single 10 V ∗H V ∗ ∗ H 5% ∗ Q H H Higgs production mechanisms listed above. mH = 125 GeV • • • • • • g g V ∗ ¯ V q H q q QH 117 q ∗ q q -2 10 Figure 3.1: The dominant SM Higgs boson production mechanisms in hadronic collisions. g g g Q g Q There are also several mechanisms forH the pair production of the HiggsH particles 1.5% Q Q H H • • • mH = 125 GeV• 80 90 100 200g Higgs pair production300 : pp gHH + X ¯ (3.5) g−→ Q g Q¯ MH [GeV] and the relevant sub–processes are the gg HH mechanism, which proceeds through heavy Figure 3.1: The dominant SM HiggsFigure→ boson 3.1: production The dominant mechanis SMms Higgs in hadronic boson production collisions. mechanisms in hadronic collisions. top and bottom quark loops [251,252], the associated double production with massive gauge There are also several mechanismsThere for the are pair also production several mechanisms of the Higgs for particles the pair production of the Higgs particles bosons [253, 254], qq¯ HHV ,andthevectorbosonfusionmechanismsqq V ∗V ∗ → → → HHqq [255, 256]; see also Ref.Higgs [254]. pair However, production because : pp of theHHHiggs supp+ X pairression production by the : additionalpp (3.5)HH + X (3.5) −→ electroweak couplings, they have much smaller production cross sections than the−→ single and the relevant sub–processes areand the thegg relevantHH mechanism, sub–processes which are proceeds the gg throughHH mechanism, heavy which proceeds through heavy Higgs production mechanisms listed above. → → top and bottom quark loops [251,252], the associated double production with massive gauge K. Nikolopoulos / Quy Nhon, 26 Sep 2016 / Search for Higgs boson decays to a mesontop and bottomand a photon quark loops [251,252], the associated double24production with massive gauge bosons [253, 254], qq¯ HHV ,andthevectorbosonfusionmechanismsbosons [253, 254], qq¯ HHV ,andthevectorbosonfusionmechanismsqq V ∗V ∗ qq V ∗V ∗ → 117 → → → → → HHqq [255, 256]; see also Ref. [254].HHqq However,[255, 256]; because see alsoof the Ref. supp [254].ression However, by the because additional of the suppression by the additional electroweak couplings, they haveelectroweak much smaller couplings, production they cross have sections much smaller than the production single cross sections than the single Higgs production mechanisms listedHiggs above. production mechanisms listed above.

117 117 3.1.2 Higgs production at hadron machines

associated production with W/Z : qq¯ V + H (3.1) −→ vector boson fusion : qq V ∗V ∗ qq + H (3.2) 3.1.2 Higgs production at hadron machines −→ −→ 3.1.2 Higgs production at hadron machines gluon gluon fusion : gg H (3.3) In the Standard Model, the main production mechanisms for Higgs particles at hadron − −→ associated production with heavy quarks : gg,qq¯ QQ¯ + H (3.4) collidersIn make the Standard use of the Model, fact the that main the production Higgs boson mechanisms couples for pr Hieferentiallyggs particles to at the hadron heavy −→ colliders make use of the fact that the Higgs boson couples preferentially to the heavy particles, that is the massive W and Z vector bosons, the top quark and, to a lesser extent, particles, that is the massive W and Z vector bosons, the top quark and, to a lesser extent, q the bottom quark. The four main production processes, the Feynman diagrams ofq¯ which are V q the bottom quark. The four main production processes, the Feynman diagrams of which are V displayeddisplayed in Fig. in 3.1, Fig. are 3.1, thus: are thus: the associated the associated production production w withithW/ZW/Z bosons [241,242], [241,242], the the V ∗ ∗ H weak vector boson fusion processes [112,243–246], the gluon–gluon fusion3.1.2 mechanism Higgs production [185] at hadron machines weak vector boson fusion processes [112,243–246], the gluon–gluon fusion mechanism [185] • • and the associatedand the associated Higgs Higgsproduction production with with heavy heavy top top [247,248 [247,248]orbottom[249,250]quarks:]orbottom[249,250]quarks:In the Standard Model, the main production mechanisms for Higgs particlesV at hadron q H q ∗ colliders make use of the fact that the Higgs boson couples preferentially to theq heavy associated production with W/Z : qq¯ V + H particles, that is the(3.1) massive W and Z vector bosons, the top quark and, to a lesser extent, associated production with W/Z : qq¯ −→V + H (3.1) Higgs−→ theboson bottom quark. The four at main productionthe processes, LHC the Feynman diagrams of which are vector boson fusion : qq V ∗V ∗ qq + H (3.2) vector boson fusion : qq −→V ∗V ∗ −→displayedqq + inH Fig. 3.1,(3.2) are thus: the associated production with W/Z bosons [241,242], the gluon gluon fusion : gg−→ H −→ g (3.3) g Q gluon gluon− fusion : gg −→H weak vector boson fusion(3.3) processes [112,243–246], the gluon–gluon fusion mechanism [185] pp H (NNLO+NNLL QCD associated+ NLO EW) productions= 8 TeV− with heavy quarks : gg,q−→q¯ QQ¯ and+ H the associated(3.4) Higgs production withH heavy top [247,24886%]orbottom[249,250]quarks: associated production with heavy quarks : gg,qq¯ −→ QQ¯ + H (3.4) Q H −→ mh = 125 GeV 10 associated production• with W/Z : qq¯ V + H • (3.1) q g g−→ Q¯ q¯ LHC HIGGS XS WG 2014 q V q vector boson fusion : qq V ∗V ∗ qq + H (3.2) H+X) [pb] V −→ −→ pp qqH (NNLOq¯ QCD + NLO EW) V ∗ V q ∗ H 7% gluon gluon fusion : gg H (3.3) FigureV 3.1: The dominant SM Higgs− boson production−→ mechanisms in hadronic collisions. pp V ∗ ∗ associatedmh = 125 production GeV with heavy quarks : gg,qq¯ QQ¯ + H (3.4) 1 pp WH (NNLO QCD + NLO EW) H (pp bbH (NNLO QCD in 5FS, NLO QCD in 4FS) • • −→ pp

ZH (NNLO QCD +NLO EW) V q H q There∗ are also several mechanisms for the pair production of the Higgs particles • • q q V ∗ q¯ V q pp q H q -1 ttH (NLO QCD) Higgs pair production : pp HHV + X (3.5) 10 q V ∗ 5%−→ ∗ H g g Q • mh = 125 GeV • H and the relevant sub–processes are the gg HH mechanism,V which proceeds through heavy Q H q H → q ∗ g g top and bottomQ quark loops [251,252], the associated double productionq with massive gauge -2 • • 10 g g ¯ H bosons [253,Q 254], qq¯ HHV ,andthevectorbosonfusionmechanismsqq V ∗V ∗ Q H → → → HHqq [255, 256];g see also Ref. [254]. However, becauseg of the suppQression by the additional Figure 3.1: The dominant• SM Higgs boson production mechanis• ms in hadronic collisions. g g Q¯ H 80 90 100 200 300 electroweak couplings, theyQ have much smaller production crossH sections than the single There are also several mechanisms for the pair production of the Higgs particles • • MH [GeV] 1.5% Higgs productiong mechanisms listed above. g Q¯ Figure 3.1: The dominant SM Higgs boson productionm mechanish = 125ms GeV in hadronic collisions. Higgs pair production : pp HH + X (3.5) Phys.Rev.Lett 114 (2015) 191803 −→ Figure 3.1: The dominant SM Higgs boson production117 mechanisms in hadronic collisions. There are also several mechanisms for the pair production of the Higgs particles and the relevant sub–processes are the gg HH mechanism, which proceeds through heavy → There are also several mechanisms for the pair production of the Higgs particles δtopm/m~0.19% and bottom quarkHiggs loops pair [251,252], production the associated : pp doubleHH +productionX with massive gauge(3.5) bosons [253, 254], qq¯ HHV ,andthevectorbosonfusionmechanisms−→ qq V ∗V ∗ Higgs pair production : pp HH + X (3.5) → → → −→ HHqq [255, 256]; see also Ref. [254]. However, because of the suppression by the additional and the relevant sub–processes are the gg HH mechanism, which proceedsand the throughrelevant sub–processes heavy are the gg HH mechanism, which proceeds through heavy → → top and bottomelectroweak quark couplings, loops [251,252], they have themuch associated smaller production double production cross sectionstop with and than massivebottom the singlequark gauge loops [251,252], the associated double production with massive gauge Higgs production mechanisms listed above. bosons [253, 254], qq¯ HHV ,andthevectorbosonfusionmechanismsqq V ∗V ∗ bosons [253, 254], qq¯ HHV ,andthevectorbosonfusionmechanismsqq V ∗V ∗ → → → → HHqq [255,→ 256]; see also→ Ref. [254]. However, because of the suppression by the additional HHqq [255, 256]; see also Ref. [254]. However, because117 of the suppression by the additional electroweak couplings, they have much smaller production cross sections than the single electroweak couplings, they have much smaller production cross sectionsHiggs productionthan the mechanisms single listed above. Higgs production mechanisms listed above. 117 117 JHEP 08 (2016) 045

K. Nikolopoulos / Quy Nhon, 26 Sep 2016 / Search for Higgs boson decays to a meson and a photon 25 SM Higgs boson decays 1 WW bb

gg LHC HIGGS XS WG 2013 10-1 ZZ

10-2 Higgs BR + Total Uncert Z

10-3 µµ

-4 10 90 200 300 400 1000 MH [GeV] mH~125 GeV gives access to several decay channels Gauge bosons: γγ, ΖΖ*, WW*, Zγ Fermions: bb, ττ, µµ K. Nikolopoulos / Quy Nhon, 26 Sep 2016 / Search for Higgs boson decays to a meson and a photon 26 h/Z→J/ψγ and h/Z→Y(nS)γ: Mass Resolution barrel endcap

converted photon

Phys.Rev.Lett. 114 (2015) 121801

unconverted photon

mass resolution ~1.2-1.8%

K. Nikolopoulos / Quy Nhon, 26 Sep 2016 / Search for Higgs boson decays to a meson and a photon 27 h/Z→J/ψγ and h/Z→Y(nS)γ: Mass Resolution

Phys.Rev.Lett. 114 (2015) 121801

K. Nikolopoulos / Quy Nhon, 26 Sep 2016 / Search for Higgs boson decays to a meson and a photon 28 h/Z→J/ψγ and h/Z→Y(ns)γ

K. Nikolopoulos / Quy Nhon, 26 Sep 2016 / Search for Higgs boson decays to a meson and a photon 29 h/Z→J/ψγ and h/Z→Y(ns)γ

Phys.Rev.Lett. 114 (2015) 12, 121801

K. Nikolopoulos / Quy Nhon, 26 Sep 2016 / Search for Higgs boson decays to a meson and a photon 30