Search for Anomalous Electroweak Gauge Boson Self-Interactions with ATLAS at the LHC

ATL-PHYS-PROC-2013-288

Search for anomalous electroweak gauge boson self-interactions with ATLAS at the LHC

Philipp Angera, on behalf of the ATLAS collaboration

aTechnische Universit¨atDresden

Abstract

The ATLAS collaboration has set√ limits on anomalous√ contributions to electroweak triple gauge boson couplings based on data produced at the LHC with both s = 7 TeV and s = 8 TeV.√ Sensitivity to measurements of anomalous electroweak quartic gauge boson couplings with an upgraded ATLAS detector at s = 14 TeV has been estimated for vector boson scattering and triple boson production in the context of eﬀective ﬁeld theories and searches for additional resonances. Keywords: LHC, ATLAS, Standard Model, electroweak, triple gauge coupling, TGC, aTGC, quartic gauge coupling, QGC, aQGC

1. Introduction than 93% the ATLAS detector performed very well during these periods. In the Standard Model of particle physics the electromagnetic and weak interactions are described by the electroweak theory, which is based on the spontaneously broken SU(2)L ⊗ U(1)Y gauge group. All aspects of this theory are in good agreement 2. Standard Model diboson production with the current experimental data. Interactions between fermions are mediated by bosonic fields describing massless (photon γ) and massive (W- and Z-bosons) As a first step towards measuring anomalous contributions to particles, the electroweak gauge bosons. The Standard Model electroweak gauge boson vertices, the measurement of the Stan- also predicts self-interactions between these bosons. Allowed dard Model production of two gauge boson in the final state as self-interactions are charged triple and quartic gauge boson ver- performed by ATLAS is presented. Triple gauge boson produc- tices: WWZ, WWγ, WWWW, WWZZ, WWZγ and WWγγ. tion has not been studied up to now, mainly due to the small The study of these vertices aim to test the Standard Model production cross-section. at the TeV scale. Such studies are possible with ATLAS at the LHC in channels involving the production of two or three gauge An overview of the measured production cross-sections cor- bosons in the final state. An important special case is vector bo- rected for leptonic branching fractions and comparison to the- son scattering (VBS) accessible in the (inclusive) two jet bin of oretical expectations can be seen in Figure 1. No significant electroweak diboson production. VBS has never been observed deviation from the Standard Model has been observed. before but plays an important role in understanding the electroweak symmetry breaking. In addition, the search for signals 105 ATLAS Preliminary of new physics at the high energy frontier is possible. A direct [pb] -1 35 pb search for physics beyond the Standard Model would involve total LHC pp s = 7 TeV σ 35 pb-1 Theory 4 the test of hypothetical models like Technicolor, Little Higgs or 10 Data (L = 0.035 - 4.6 fb-1) Extended Gauge Models. Here, the results of a more general LHC pp s = 8 TeV ansatz are collected and indirect searches for anomalous triple 3 10 Theory -1 and quartic gauge couplings are presented. 5.8 fb Data (L = 5.8 - 20 fb-1)

The reported results are based on data collected by the AT- 5.8 fb-1 2 LAS detector [1] during the data-taking periods 2011 and 2012. 10 1.0 fb-1 1.0 fb-1 13 fb-1 Two different center-of-mass energies of the√ colliding protons 4.6 fb-1 20 fb-1 4.6 fb-1 were used by the LHC during these years, s = 7 TeV and 10 -1 √ 2.1 fb s = 8 TeV. Analyses performed at both energies are pre- 4.6 fb-1 −1 sented based on different amounts of√ collected data: 35 pb to −1 −1 1 W Z tt t WW WZ Wt ZZ 4.6 fb integrated√ luminosity for s = 7 TeV and 5.8 fb to 20 fb−1 for s = 8 TeV. With a fraction of operational detector channels above 95% and data taking efficiencies higher Figure 1: Standard Model production cross section measurements, corrected for leptonic branching fractions, compared to the corresponding theoretical expectations [2].

Email address: philipp.anger@cern.ch (Philipp Anger)

Preprint submitted to Hepmad13 November 2, 2013 3. Anomalous triple gauge couplings Model independent aTGC can be introduced within an effective Lagrangian: Since triple gauge couplings are completely determined by V † µ ν †µ ν the Standard Model, any deviation in terms of anomalous triple LWWV ∝ g1 (WµνW V − WµνW V ) + gauge couplings (aTGC) would indicate the existence of new † µν λV † µ νρ s κV Wµ WνV + WρµWν V , (2) physics. At the LHC, the sensitivity to aTGC is driven by - m2 channel diboson production. The corresponding tree level dia- W gram can be seen in Figure 2. In the Standard Model, final states where V denotes either Z or γ. All parameters are fixed in with Zγ, γγ and ZZ final states cannot be produced through the V the Standard Model: g1 = κV = 1, λV = 0. When search- process represented in Figure 2 since just charged triple gauge ing for physics beyond the Standard Model, the parameters are couplings (WWZ, WWγ) are allowed. allowed to change. Therefore, five free parameters are introduced: ∆gZ = gZ − 1, ∆κ = κ − 1 and λ . q 1 1 V V V V Diboson WZ production was used to set limits on the three Z parameters λZ, ∆κZ and ∆g1 [3]. The influence of each of these parameters on the transverse momentum of the Z candidate is [a]TGC shown in Figure 3 for three different choices of parameters. Data are shown together with expected background and signal q events, assuming Standard Model. The colored lines indicate V the possible aTGC scenarios without unitarization. All param- Figure 2: A schematic Feynman diagram of TGC in diboson production. The eters change the high energy region while keeping the low en- circle indicates possible existence of anomalous contributions. V indicates elec- ergy regime. No evidence for any of the anomalous scenarios troweak gauge bosons W, Z or γ. The final states Zγ, γγ and ZZ are not allowed is found. Therefore limits on all parameters can be set. for this diagram in the Standard Model.

160 ATLAS 3.1. Unitarity 140 Data √ ∫ Ldt = 4.6 fb•1 s = 7 TeV SM WZ When adding aTGCs, unitarity can be violated at high s. 120 Bkg Events / bin σ This fact implies undefined probabilities leading to infinite stat + syst 100 λ Z = 0.05 cross-sections and therefore to unphysical results. In order to ∆κ Z = 0.57 obtain physical results, an unitarization procedure is needed. 80 ∆gZ = 0.10 1 Nevertheless, two scenarios have been used in ATLAS for 60 publishing results: 40 1. no unitarization 2. form-factor unitarization. 20 0 No unitarization introduces no model dependence and no as- 0 30 60 90 120 150 180 2000 sumptions, while the introduction of a form factor makes the pZ [GeV] T result less general. The form factor is chosen to be energy de- pendent in order to restore unitarity. A general form varies each Figure 3: Transverse momentum of the reconstructed Z candidate in data com- aTGC parameter α: pared to the sum of all backgrounds and Standard Model WZ production [3]. Possible un-unitarized aTGC contributions are stacked in addition for three dif- α ferent sets of parameters. α → α(ˆs) = . (1) sˆ n (1 + Λ2 ) FF One dimensional 95% confidence intervals on all three parameters accessible in WZ production are shown in Figure 4. Here, ΛFF is the energy scale of the form factor and n is a free parameter that is chosen differently for different final states: n = The results are shown without unitarization (ΛFF = ∞) and with 2 for WW and WZ analyses, n = 3 for ZZ analysis and n = 3 unitarization using form factors with n = 2 and ΛFF = 2 TeV. or 4 for Zγ final states.s ˆ is the center-of-mass energy of the As comparison unitarized results from CDF and D0 are shown outgoing bosons. which are similar to the ones obtained by ATLAS. Observed two dimensional 95% confidence regions on all 3.2. WWZ and WWγ vertices combinations of the three parameters can be seen in Figure 9. The results are based on un-unitarized parameters. The hori- There are just two triple gauge vertices in the Standard zontal and vertical lines inside each contour correspond to the Model: WWZ and WWγ. Both are fixed by the gauge theory. limits found in the one-dimensional fit procedure. They are accessible in WW, WZ1 and Wγ production. In addition to the WZ final state, also WW [4] and Wγ [5] analyses have set limits on aTGC. While WW diboson produc- 1Throughout this note, analyses with Z in the final state refer to Z or γ∗ final tion has access to the same set of parameters as WZ, it was states. possible to set limits on the remaining two parameters λγ and 2 ATLAS ATLAS, s = 7 TeV 4.6 fb•1, Λ = ∞ ATLAS, s = 7 TeV Z f5 4.6 fb•1, Λ = 2 TeV ATLAS Z ∆g CDF, s = 1.96 TeV 1 7.1 fb•1, Λ = 2 TeV γ D0, s = 1.96 TeV f5 CMS, s = 7TeV •1 4.1 fb , Λ = 2 TeV 5.0 fb•1, Λ = ∞ ATLAS, s = 7TeV 4.6 fb•1, Λ = ∞ λ + • Z W ±Z → l ±νl l ATLAS, s = 7TeV Z •1 95% C.L. f 4.6 fb , Λ = 3 TeV 4 LEP, s = 130•209 GeV 0.7 fb•1, Λ = ∞ D0, s = 1.96TeV 1.0 fb•1, Λ = 1.2 TeV ∆κ γ Z f4

•1 •0.8 •0.6 •0.4 •0.2 0 0.2 0.4 •0.6 •0.4 •0.2 0 0.2 0.4 0.6 0.8

Figure 5: 95% confidence regions on aTGC parameters for the ZZV (V = Z, γ) vertices as obtained by the ZZ final state analysis [6]. ATLAS results are com- Figure 4: 95% confidence intervals on unitarized aTGC parameters accessible pared to results from CMS, LEP and D0. No unitarization was applied for in WZ production obtained by ATLAS, CDF and D0 [3]. Un-unitarized inter- Λ = ∞ and a form factor unitarization was used for finite Λ. vals as obtained by ATLAS are also shown.

couplings, a unitarization similar to triple gauge boson studies ∆κγ using Wγ final states. The observed results of these analyses are shown in Figure 10. D0 confidence regions are shown (see 3.1) has to be applied. as comparison. For WW, also results of CDF, LEP and CMS are added to the plot. 4.1. Direct searches Additional resonances can be introduced in the context of 3.3. ZZZ, ZZγ and Zγγ vertices vector boson scattering. Following [8], they can be grouped Neutral vertices are not allowed in the Standard Model. Pos- in terms of√ spin and isospin. Prospects for the sensitivity of sible anomalous vertices are tested by ATLAS in ZZ and Zγ ATLAS at s = 14 TeV for additional resonances was investi- production. gated in [9]. In the ZZ final state analysis [6], the following effective La- q grangian was used: q V L ∝ V µβ α V σ µβ V ZZV f4 (∂µV )Zα(∂ Zβ) + f5 (∂ Vσµ)Ze Zβ. (3) q V V All four parameters f4 , f5 are zero in the Standard Model and V V limits on nonzero values are set. The results of the measure- q ments can be seen in Figure 5. ATLAS could improve the cur- q q V rent 95% C.L. limits of LEP and D0 by approximately one order of magnitude. Figure 6: Schematic Feynman diagrams of channels contributing to anomalous The Zγ final state was used to set limits on the ZZγ and Zγγ quartic gauge coupling measurements at the LHC: Weak boson scattering (left) vertices [5] that is also not allowed in the Standard Model. A and triple gauge boson production (right). V V different parametrization with four free parameters h3 and h4 was used [7]. 95% C.L. limits on these parameters are shown The sensitivity of vector boson scattering ZZ j j → ```` j j V production when adding a f0 resonance with spin S = 0, isospin in Figure 11 using a form factor unitarization with n = 3 (h3 ) V I = 2, mass m = 1 TeV, coupling to electroweak bosons and n = 4 (h4 ). No deviation from the Standard Model has been observed for g = 1.75 and width Γ = 50 GeV has been estimated using a the neutral triple gauge vertices. cut on m( j, j) > 1 TeV. The invariant mass of the four lepton system after this cut is shown in Figure 7. With an expected integrated luminosity of 300 fb−1 the discovery sensitivity has 4. Anomalous quartic gauge couplings been estimated to 1.7σ. With 3 ab−1 5.5σ can be reached. Possible deviations from the Standard Model quartic boson vertices are accessible at the LHC in electroweak gauge boson 4.2. Indirect searches scattering and triple vector boson production. A schematic view Instead of searching for additional direct contributions of of both processes is shown in Figure 6. new physics to the quartic gauge boson vertices in terms of ad- Changing Standard Model quartic gauge boson interactions ditional resonances, a search for the traces in the low energy can lead to violation of unitarity. When testing for anomalous region can be performed. The Standard Model Lagrangian is 3 analysis parameter int. lumi. int. lumi. ATLAS Preliminary SM VV [TeV−4] 300 fb−1 3 ab−1 50 (Simulation) Entries •1 4 L dt = 3000 fb Non•VV WZ j j → `ν`` j j fT,1/Λ 1.3 0.6 ∫ Diboson 40 W±W± j j → `±ν`±ν j j f /Λ4 10 4.5 SM VV + S,0 1.0 TeV Res 4 (g = 1.75) Zγγ fT,8/Λ 0.9 0.4 4 30 Zγγ fT,9/Λ 2.0 0.7

20 Table 1: 5σ discovery significance regions for different dimension eight parameter contributing to aQGC in various final states for two scenarios of total integrated luminosity collected by ATLAS. 10

A summary in terms of 5σ discovery values can be seen in 0.2 0.3 0.4 0.5 0.6 1 Table 1.

m4l [TeV]

Figure 7: Invariant mass of the four lepton system for fully leptonic VBS ZZ j j 5. Summary production [9]. Non-ZZ diboson production is shown together with SM VBS ZZ production. The influence of a f0 resonance with m = 1 TeV, g = 1.75 and Thanks to the successful running of the LHC and smooth Γ = 50 GeV to the SM VBS ZZ production is shown in red. operation of the ATLAS detector, Standard Model diboson production has been measured and limits on anomalous triple gauge couplings have been set in many channels. No significant replaced by an effective Lagrangian extended by additional op- deviations from the Standard Model have been observed. In ad- erators [10]. Depending on the dimension of these operators, dition, studies of anomalous quartic gauge couplings in VBS they contribute to triple or quartic gauge boson vertices, or both. and triple gauge boson production are presented. MC stud- The lowest order affecting just the quartic vertices is eight. 18 ies have shown that ATLAS is sensitive to direct and indirect independent parameters can be added according to [10]. searches for physics beyond the Standard Model at 14 TeV. Sensitivity of ATLAS at 14 TeV for additional operators of dimension eight is estimated in [11]. Unitarity has been re- References stored using a cutoff unitarization discarding all events in the unitarity violating region of M(V, V). [1] The ATLAS Collaboration. The ATLAS experiment at the CERN Large As an example, Figure 8 shows the result for VBS in the Hadron Collider. JINST, 3:S08003, 2008. → [2] https://twiki.cern.ch/twiki/bin/view/AtlasPublic/ WZ j j `ν`` j j final state. The QCD induced SM process CombinedSummaryPlots [August 3, 2013]. is shown in yellow, EW induced processes including VBS are [3] The ATLAS Collaboration.√ Measurement of WZ production in proton- shown in blue. The contribution of an additional dimension proton collisions at s=7 TeV with the ATLAS detector. Eur. Phys. J., −4 C 72:2173, 2012. eight operator with parameter fT,1 = 1 TeV is shown in red. + − [4] The ATLAS√ Collaboration. Measurement of W W production in pp col- The largest deviation from the SM appears in the high mass lisions at s=7 TeV with the ATLAS detector and limits on anomalous region. WWZ and WWγ couplings. Phys. Rev. D, 87:112001, 2013. [5] The ATLAS Collaboration.√ Measurements of Wγ and Zγ production in pp collisions at s=7 TeV with the ATLAS detector at the LHC. Phys. Rev. D, 87:112003, 2013. ATLAS Simulation Preliminary VBS WZ (SM) 450 [6] The ATLAS√ Collaboration. Measurement of ZZ production in pp col- •1 lisions at s=7 TeV and limits on anomalous ZZZ and ZZγ couplings Entries 400 ∫ L = 3000 fb SM VBS WZ + •4 with the ATLAS detector. Journal of High Energy Physics, 03:128, 2013. fT1 = 1.0 TeV 350 [7] U. Baur and E. L. Berger. Probing the weak boson sector in Zγ production VBS WZ (SM) at hadron colliders. Phys. Rev. D, 47:4889–4904, 1993. 300 [8] A. Alboteanu, W. Kilian, and J. Reuter. Resonances and unitarity in weak 250 SM WZ QCD boson scattering at the LHC. JHEP, 0811:010, 2008. 200 [9] ATLAS Collaboration. Studies of vector boson scattering with an upgraded ATLAS detector at a high-luminosity LHC. Technical Report 150 ATL-PHYS-PUB-2012-005, CERN, Geneva, 2012. 100 [10] O.J.P. Eboli, M.C. Gonzalez-Garcia, and J.K. Mizukoshi. pp → j je±µ±νν and j je±µ±νν at O(α(em)6) and O(α(em)4α(s)2) for the study 50 of the quartic electroweak gauge boson vertex at LHC. Phys.Rev., D:073005, 2006. 0.6 0.7 0.8 0.9 1 [11] ATLAS Collaboration. Studies of vector boson scattering and triboson production with an upgraded ATLAS detector at a high-luminosity LHC. m3lν [TeV] Technical Report ATL-PHYS-PUB-2013-006, CERN, Geneva, 2013. Figure 8: Invariant mass of the WZ system for VBS WZ j j production [11]. Shown is QCD induced SM background and EW induced VBS production. −4 The contribution of an additional dimension eight parameter fT,1 = 1 TeV is shown in red.

4 Z Z Z κ λ λ •1 ∆ •1 •1 ATLAS ∫L dt = 4.6 fb ATLAS ∫L dt = 4.6 fb ATLAS ∫L dt = 4.6 fb s = 7 TeV 0.05 s = 7 TeV 0.05 s = 7 TeV 0.5

0 0 0

•0.05 •0.05 •0.5 • • • W ±Z → l±νl+l , 95% C.L. W ±Z → l±νl+l , 95% C.L. W ±Z → l±νl+l , 95% C.L.

•0.1 •0.05 0 0.05 0.1 •0.1 •0.05 0 0.05 0.1 •0.5 0 0.5 ∆gZ ∆gZ ∆κ 1 1 Z

Figure 9: Two dimensional 95% conﬁdence regions for all combinations of parameters accessible in WZ production [3]. No unitarization has been applied.

ATLAS ATLAS, s = 7 TeV D0 (Wγ), s = 1.96 TeV pp → lν γ 4.6 fb•1, Λ = ∞ 4.2 fb•1, Λ = 2 TeV 95% CL ATLAS, s = 7 TeV D0 (WW,WZ,Wγ), s = 1.96 TeV 95% C.L. limits from WW production 4.6 fb•1, Λ = 6 TeV 8.6 fb•1, Λ = 2 TeV

ATLAS (4.6 fb-1, Λ=6 TeV) LEP ATLAS (4.6 fb-1, Λ=∞) ∆κ CMS (36 pb-1, Λ=∞) Z LEP (2.6 fb-1) CDF (3.6 fb-1, Λ=2 TeV) -1 D0 (1 fb , Λ=2 TeV) λγ

λZ

∆κγ ∆gZ 1

-0.6 -0.4 -0.2 0 0.2 0.4 0.6 •0.6 •0.4 •0.2 0 0.2 0.4 0.6 Coupling Strength

Figure 10: 95% confidence regions in all possible aTGC parameters as obtained by the WW final state analysis (left, [4]) and the Wγ final state analysis (right, [5]).

ATLAS ATLAS, s = 7 TeV CDF, s = 1.96 TeV ATLAS ATLAS, s = 7 TeV CDF, s = 1.96 TeV + • + • pp → l l γ, pp → νν γ 4.6 fb•1, Λ = ∞ 5.1 fb•1, Λ = 1.5 TeV pp → l l γ, pp → νν γ 4.6 fb•1, Λ = ∞ 5.1 fb•1, Λ = 1.5 TeV 95% CL ATLAS, s = 7 TeV D0, s = 1.96 TeV 95% CL ATLAS, s = 7 TeV D0, s = 1.96 TeV 4.6 fb•1, Λ = 3 TeV 7.2 fb•1, Λ = 1.5 TeV 4.6 fb•1, Λ = 3 TeV 7.2 fb•1, Λ = 1.5 TeV LEP

Z Z h3 h4

γ γ h3 h4

•0.2 •0.15 •0.1 •0.05 0 0.05 •0.001 0 0.001 Coupling Strength Coupling Strength

Figure 11: 95% conﬁdence regions in aTGC parameters for the ZVγ (V = Z, γ) vertices as obtained by the Zγ ﬁnal state analysis [5]. Unitarized and un-unitarized ATLAS results are compared to unitarized results from LEP.