Natural SUSY on Trial: Status of Higgsino Searches at ATLAS

Natural SUSY On Trial: Status of Higgsino Searches at ATLAS

Julia Gonski Harvard University

25 October 2018 US LHC User’s Association Annual Meeting SUSY (The Defendant)

• Provides uniﬁcation of forces, dark matter candidate, and solution to hierarchy problem

• Higgs mass quadratic corrections cancelled by new supersymmetric partners; higgsinos, stops and gluinos especially inﬂuential

➡ What do we search for ﬁrst?

25 October 2018 J. Gonski !2 SUSY Mass Spectrum (if we got to pick)

• Naturalness = reduce ﬁne ˜ Li, e˜i L˜i, e˜i tuning ⟹ 10% B˜ B˜ ˜ ˜ Q˜1,2, u˜1,2, d˜1,2 Q1,2, u˜1,2, d1,2 W˜ W˜ mass

˜bR ˜bR g˜ g˜ • Dark matter candidate! Dominantly (but not purely) t˜ t˜R ˜ ˜ L higgsino tL tR ˜bL ˜bL H˜ • Lightest three˜ higgsinos are compressedH (Δm ~ few GeV) natural SUSY decoupled SUSY natural SUSY decoupled SUSY 25 October 2018 J. Gonski !3

FIG. 1: Natural electroweak symmetry breaking constrains the superpartners on the left to be light. Meanwhile, the superpartners on the right can be heavy, M 1 TeV, without spoiling FIG. 1: Natural electroweak symmetry breaking constrains the superpartners on the left to be naturalness. In this paper, we focus on determining how the LHC data constrains the masses of light. Meanwhile, the superpartners on the right can be heavy, M 1 TeV, without spoiling the superpartners on the left. naturalness. In this paper, we focus on determining how the LHC data constrains the masses of the superpartners on the left. the main points, necessary for the discussions of the following sections. In doing so, we will try to keep the discussion as general as possible, without committing to the specific Higgs the main points, necessary for the discussionspotential of of the the MSSM. following We sections. do specialize In doing the discussion so, we will to 4D theories because some aspects try to keep the discussion as general asof fine possible, tuning without can be modified committing in higher to the dimensional specific Higgssetups. potential of the MSSM. We do specializeIn the a naturaldiscussion theory to of4D EWSB theories the because various contributions some aspects to the quadratic terms of the Higgs potential should be comparable in size and of the order of the electroweak scale v 246 GeV. of fine tuning can be modified in higher dimensional setups. ⇠ The relevant terms are actually those determining the curvature of the potential in the In a natural theory of EWSB the various contributions to the quadratic terms of the Higgs direction of the Higgs vacuum expectation value. Therefore the discussion of naturalness potential should be comparable in size and of the order of the electroweak scale v 246 GeV. ⇠ The relevant terms are actually those determining the curvature of the potential in the direction of the Higgs vacuum expectation value. Therefore the discussion of7 naturalness

7 3 Example(EWKino(Spectrum 3 In(principle,(any(bino/wino/higgsino(mass(hierarchy(is(allowed Example(EWKino(Spectrum 3 3 Example(EWKino(Spectrum Example(EWKino(Spectrum In(principle,(any(bino/wino/higgsino(mass(hierarchy(is(allowed 3 ~ 0 ~ 0 ~ ± ~ In(principle,(any(bino/wino/higgsino(mass(hierarchy(is(allowed !3 !4 !2 3 3 H)(higgsino) In(principle,(any(bino/wino/higgsino(mass(hierarchy(is(allowed3 Example(EWKino(SpectrumExample(EWKino(SpectrumExample(EWKino(SpectrumSUSY Mass Spectrum Δm%~%few%–%tens%of%GeV ~Example(EWKino(Spectrum0 ~ 0 ~ ± ~ !3 !4 !2 ~ 0 ~ 0 ~ ± ~ (if we3H) got(higgsino to) pick)3 4 2 In(principle,(any(bino/wino/higgsino(mass(hierarchy(is(allowedIn(principle,(any(bino/wino/higgsino(mass(hierarchy(is(allowed30 October 2017 J. Gonski ! ! !12 ~ 0 ~ ± (higgsino) ~ 3 Δm%~%few%–%tens%of%GeV 3 2 H)1 Example(EWKino(SpectrumIn(principle,(any(bino/wino/higgsino(mass(hierarchy(is(allowed~ 0 ~!0 ~!± ~W)(wino) 3Δm%~%few%–%tens%of%GeV4 2 Example(EWKino(Spectrum ~ 0 ~ ± In(principle,(any(bino/wino/higgsino(mass(hierarchy(is(allowed~ ! ! ! (higgsino) • 2 1 (wino) ˜ Δm%~%few%hundreds%MeV%H) ~Example(EWKino(Spectrum0 ~ 0 ~ ± Naturalness~ ~!0 ~!0 ~ ±= reduceW)~ ﬁne Li, e˜i ˜ 3 In(principle,(any(bino/wino/higgsino(mass(hierarchy(is(allowed4 2 A Natural3 4 2 Electroweak Spectrum Li, e˜i ! ! ! 3H)(higgsino! !) ! Δm%~%few%hundreds%MeV%H)(higgsino) Δm%~%few%–%tens%of%GeV(assuming%heavy%sfermions%and%higgsinos) In(principle,(any(bino/wino/higgsino(mass(hierarchy(is(allowed30 October 2017 tuningJ. Gonski 10% 12 ~ 0 ~˜± ~ ˜ Δm%~%few%–%tens%of%GeV⟹ (assuming%heavy%sfermions%and%higgsinos)Δm%~%few%–%tens%of%GeV 2 B1 B Example(EWKino(Spectrum Gauginos ~ 0 ~ ! 0 ~ !Mass± Eigenstates˜~W) (˜wino) ˜ ˜ In(principle,(any(bino/wino/higgsino(mass(hierarchy(is(allowed⇢3 4 2 ~ Q~1,2, u˜1,2,~d1,2 ~~ Q1,2, u˜1,2, d1,2 ~ 0 ~ ± ~ ~~0 ~0 ±0~ 0 ~ ± ~ ~ ~ ! ! ~!0 ~˜ 0 ~Δ±m%~%few%hundreds%MeV%0 ± (~higgsino0 ) ˜ !2 !1 W)(wino) !!2 3!1!4 !2 B)W)(bino(H))wino(higgsino) ) !3 !W4 !2 !2 !H)1 !1 (higgsinoW)W)(bino(wino) ) In(principle,(any(bino/wino/higgsino(mass(hierarchy(is(allowedA Natural~ 0 ~ 0 ~ ± Electroweak~ Spectrum H) B) !3 !4 !2 Δm%~%few%hundreds%MeV%H)(higgsino) Δm%~%few%hundreds%MeV%Δm%~%few%–%tens%of%GeV ~ Δm%~%few%–%tens%of%GeV(assuming%heavy%sfermions%and%higgsinos)mass (gravitino) Δm%~%few%–%tens%of%GeVΔm%~%few%hundreds%MeV% (assuming%heavy%sfermions%and%higgsinos)Δm%~%few%–%tens%of%GeV (assuming%heavy%sfermions%and%higgsinos) G) ˜ ~ Gauginos Mass Eigenstates ˜ bR ~ 0 ~ ± ~ Appears%in%GMSB/GGM%Li, e˜i ˜b (assuming%heavy%sfermions%and%higgsinos)˜ G)(gravitino) !2 !⇢1 W)(wino) ~ R Li, e˜i ~ 0 ~ ~ ~ ~ ~ ~~0 ~0 ±0~ 0 ~ ± ~ ~ ~ 1 ~ 0 ~ 0 ~ ± 0 ± ~˜ 0 models,%mass%~keV !!2 3!1!4 !2 B)W)(bino(H))wino(higgsino) ! ) Δm%~%few%hundreds%MeV%B)!(3bino!) 4 !2 !2 !1 ~!B10 ~ ± (higgsinoW))(bino(wino~) ) B˜ Appears%in%GMSB/GGM% Mass 2 H)1 ˜ B)˜ (wino) ˜ ˜ (assuming%heavy%sfermions%and%higgsinos)! !~ Q1,2, u˜1~,2, d01,2 W) ~ Q1,2, u˜1,2, d1,2 Δm%~%few%hundreds%MeV%Δm%~%few%–%tens%of%GeV ~ ˜ Δm%~%few%hundreds%MeV%(gravitino) 1 g˜ ˜ models,%mass%~keV (assuming%heavy%sfermions%and%higgsinos) G)(gravitinog˜ ) Δm%~%few%–%tens%of%GeVW G) Δm%~%few%hundreds%MeV%! WB)(bino) ~ ~ 0 ~ ˜ Appears%in%GMSB/GGM% ~ 0 ~ ± ~ !1 (binoAppears%in%GMSB/GGM%) Li, e˜i (assuming%heavy%sfermions%and%higgsinos)˜ Mass G)(gravitino)

2 1 (wino) B) • Li, e˜i xsec ! ! Dark matter candidate! (assuming%heavy%sfermions%and%higgsinos)mass ~ ~ 0 ~ W) models,%mass%~keV ˜ 1 ˜ models,%mass%~keV ~ ˜b bR ! Δm%~%few%hundreds%MeV%B)(bino) Mass~B0 ~ ± B˜ ~ R Appears%in%GMSB/GGM% G)(gravitino) Mass 2 1 Dominantly˜ ˜ (but(wino notG)) purely(gravitino˜ )) ˜ t˜L t˜R (assuming%heavy%sfermions%and%higgsinos)! !~ Q1,2, u˜1~,2, d01,2 W) ~ Q1,2, u˜1,2, d1,2 ˜ (gravitino) 1 ˜~Appears%in%GMSB/GGM%0 ˜ ˜ ~ models,%mass%~keVAppears%in%GMSB/GGM% W G) higgsinoΔm%~%few%hundreds%MeV%! !tL1 WB)t(Rbino) (g˜bino) ˜bL ~ 0 ~ g˜ models,%mass%~keVB) !1 (bino) Appears%in%GMSB/GGM% ˜ Mass models,%mass%~keV

B) Mass xsec (assuming%heavy%sfermions%and%higgsinos)mass bL ~ models,%mass%~keV ˜ ˜b ~ ~ bR RMass ˜ (gravitino) Mass (gravitino) ˜ ˜ H G) G)(gravitinoweak scale) G) • Lightest three higgsinos aretL tR ~Appears%in%GMSB/GGM%0 ~ t˜L t˜R ˜ ˜ Appears%in%GMSB/GGM% 1 g˜ H bL Appears%in%GMSB/GGM% ! models,%mass%~keVB)(binocompressed˜) (Δm ~ few GeV) (< 400 GeV) g˜ bL models,%mass%~keVmodels,%mass%~keV Mass H˜ ~ natural SUSY decoupled SUSY Mass˜ Mass G)(gravitino) t˜L H t˜R t˜ t˜R natural SUSY decoupled SUSY L 25˜bL October 2018 Appears%in%GMSB/GGM%J. Gonski !4 ˜ natural SUSY decoupled SUSY bL models,%mass%~keV natural˜ SUSY decoupled SUSY Mass H H˜ FIG. 1: Natural electroweak symmetry breaking constrains the superpartners on the left to be light. Meanwhile, the superpartners on the right can be heavy, M 1 TeV, without spoiling FIG.natural 1: Natural SUSY electroweakFIG. symmetry 1: Naturaldecoupled electroweak breaking SUSY symmetry constrains breaking the constrains superpartners the superpartners on the on left the left to to be be light. Meanwhile, thenaturalness. superpartners In on this the paper, right can we be focus heavy, onM determining1 TeV, without how spoiling the LHC data constrains the masses of natural SUSY FIG. 1: Naturaldecoupled electroweak SUSY symmetry breaking constrains the superpartners on the left to be light. Meanwhile, the superpartnersnaturalness. In on this the paper, right we focus can on be determining heavy, howM the LHC1 TeV, data constrains without the spoiling masses of light. Meanwhile, the superpartners on the right can bethe heavy, superpartnersM 1 TeV, on without the left. spoiling the superpartners on the left. FIG. 1: Naturalnaturalness. electroweak symmetry In this paper, breaking we constrains focus the on superpartners determining on how the left the to LHCbe data constrains the masses of naturalness. In this paper, we focus on determining how the LHC data constrains the masses of light. Meanwhile,the superpartners the superpartners on on the right left. can be heavy, M 1 TeV, without spoiling FIG. 1: Natural electroweak symmetry breakingthe superpartners constrains the on the superpartners left. on the left to be the main points, necessary for the discussions of the following sections. In doing so, we will naturalness. In this paper, we focus on determiningthe main how points, the LHC necessary data constrains for the discussionsthe masses of of the following sections. In doing so, we will light. Meanwhile, the superpartners on the right can be heavy, M 1 TeV, without spoiling the superpartners on the left. try to keep the discussiontry to keep as general the discussionas possible, without as general committing as possible, to the specific without Higgs committing to the specific Higgs naturalness. In this paper, we focus on determining how the LHC data constrains the masses of the mainthe points, main necessary points, for necessary the discussionspotential for the of of the the discussions MSSM. followingpotential We sections. do of of specialize the the InMSSM. following doing the discussion Weso, sections. we do will to specialize 4D theories In doing the because discussion so, some we aspects will to 4D theories because some aspects the superpartners on the left. try to keep the discussion as general asof finepossible, tuning without can be modified committing in higher to the dimensional specific Higgs setups. the main points,try necessary to keep for the the discussion discussions of as the general following as sections.of fine possible, tuning In doing without can so, bewe will modified committing in higher to the dimensional specific Higgssetups. potential of the MSSM. We do specializeIn the a naturaldiscussion theory to 4Dof EWSB theories the because various contributions some aspects to the quadratic terms of the Higgs try to keep the discussion as general as possible, without committingIn a to natural the specific theory Higgs of EWSB the various contributions to the quadratic terms of the Higgs potential of the MSSM. Wepotential do specialize should be comparable the discussion in size and to of the 4D order theories of the electroweak because somescale v aspects246 GeV. the main points, necessary for the discussionspotentialof fine of of tuning the the MSSM. following can be We modifiedsections. do specialize in In higher doing the discussion dimensionalso, we will to 4D setups. theories because some aspects ⇠ The relevant termspotential are actually should those be determining comparable the in curvature size and of of the the potential order of in the the electroweak scale v 246 GeV. try to keep the discussion as general asof finepossible,In tuning a naturalof without can fine be theory modifiedtuning committing of EWSB in can higher to be the the dimensionalmodified various specific contributions Higgsin setups. higher to dimensional the quadratic setups. terms of the Higgs ⇠ direction of the HiggsThe vacuum relevant expectation terms are value. actually Therefore those the discussion determining of naturalness the curvature of the potential in the potential of the MSSM. We do specializeIn thepotential a naturaldiscussion should theoryIn to a be natural4Dof EWSB comparable theories the theory because various in size of contributions EWSBsome and of aspects the the order to various the of quadratic the electroweak contributions terms of the scale Higgs tov the246 quadratic GeV. terms of the Higgs direction of the Higgs⇠ vacuum expectation value. Therefore the discussion of naturalness potentialThe relevant should be terms comparable are actually in size and those of the determining order of the electroweak the curvature scale ofv the246 potential GeV. in the of fine tuning can be modified in higher dimensionalpotential setups. should be comparable in size and of the order⇠ of the electroweak scale v 246 GeV. Thedirection relevant terms of the are Higgs actually vacuum those expectation determining value. the curvature Therefore of the potentialdiscussion in of the7 naturalness ⇠ In a natural theory of EWSB the various contributions to the quadratic terms of the Higgs direction ofThe the Higgs relevant vacuum terms expectation are value. actually Therefore those the determining discussion of naturalness the curvature of the potential in the potential should be comparable in size and of the order of the electroweak scale v 246 GeV. direction of the Higgs vacuum⇠ expectation value. Therefore the discussion of7 naturalness The relevant terms are actually those determining the curvature of the potential in the 7 direction of the Higgs vacuum expectation value. Therefore the discussion of7 naturalness

7 7 SUSY Mass Spectrum (if we got to pick)

• Naturalness = reduce ﬁne ˜ Li, e˜i L˜i, e˜i tuning ⟹ 10% B˜ B˜ ˜ ˜ Q˜1,2, u˜1,2, d˜1,2 Q1,2, u˜1,2, d1,2 W˜ W˜

B˜ W˜ H˜ mass

˜bR ˜bR χ˜± χ˜± χ˜0 χ˜0 χ˜0 χ˜0 2 1 g˜ 4 3 2 1 g˜ • Dark matter candidate! Dominantly (but not purely) t˜ t˜R ˜ ˜ L higgsino tL tR ˜bL ˜bL H˜ weak scale • Lightest three˜ higgsinos are compressedH (Δm ~ few GeV) (< 400 GeV) natural SUSY decoupled SUSY natural SUSY decoupled SUSY 25 October 2018 J. Gonski !5

7 SUSY Mass Spectrum (if we got to pick)

• Naturalness = reduce ﬁne ˜ Li, e˜i L˜i, e˜i tuning ⟹ 10% B˜ B˜ ˜ ˜ Q˜1,2, u˜1,2, d˜1,2 Q1,2, u˜1,2, d1,2 W˜ W˜

B˜ W˜ H˜ mass

7 SUSY Mass Spectrum (if we got to pick)

• Naturalness = reduce ﬁne ˜ Li, e˜i L˜i, e˜i tuning ⟹ 10% B˜ B˜ ˜ ˜ Q˜1,2, u˜1,2, d˜1,2 Q1,2, u˜1,2, d1,2 W˜ W˜

B˜ W˜ H˜ mass

˜bR ˜bR χ˜± χ˜± χ˜0 χ˜0 χ˜0 χ˜0 2 1 g˜ 4 3 2 1 g˜ • Dark matter candidate! Dominantly (but not purely) t˜ t˜R ˜ ˜ L higgsino tL tR ˜bL ˜bL H˜ weak scale • Lightest three˜ ewkinos are compressedH (Δm ~ few GeV) (< 400 GeV) natural SUSY decoupled SUSY natural SUSY decoupled SUSY 25 October 2018 J. Gonski !7

7 Higgsino Searches at ATLAS DRAFT • Final state: 2 opposite sign same j q j ` ﬂavor soft leptons p q p W * ˜ 0 • Challenging! ˜1± 0 ` ˜1 ˜1 • Low production ˜0 cross section 0 1 ˜0 ˜2 `˜ 1 Z * • Small Δm → low p ` p MET, very soft ` ` leptons (a) (b)

0 Figure 1: Diagrams representing the two lepton final state of (a) associated electroweakino 2 1± and (b) slepton pair `` production in association with an initial state radiation jet. In addition to (a), this analysis is also sensitive H H to 0 0 and production. 25 October 2018 J. Gonski2HH1 1± 1⌥ !8 H H H H 54 muon spectrometer (MS). The ID provides precision tracking of charged particles in pseudorapidity region Not reviewed, for internal circulation only 55 ⌘ < 2.5, consisting of pixel and microstrip silicon subsystems within a transition radiation tracker. An | | 56 insertable B-layer [37] was added for ps = 13 TeV data-taking to improve tracking performance. These 57 are immersed in a 2 T axial magnetic field provided by a superconducting solenoid. High-granularity 58 lead/liquid-argon electromagnetic sampling calorimeters are used for ⌘ < 3.2. Hadronic energy deposits | | 59 are measured in a steel/scintillator tile barrel calorimeter in ⌘ < 1.7. Forward calorimeters extend the | | 60 coverage to 1.5 < ⌘ < 4.9 regions for both electromagnetic and hadronic measurements. The MS | | 61 comprises trigger and high-precision tracking chambers spanning ⌘ < 2.4 and ⌘ < 2.7, respectively, | | | | 62 surrounded by three large superconducting toroidal magnets. Events of interest are selected using a two- 63 level trigger system [38], consisting of a first-level trigger implemented in hardware, which is followed by 64 a software-based high-level trigger.

65 3 Collision data and simulated event samples

66 The search uses pp collision data at ps = 13 TeV from the LHC [39], collected by the ATLAS detector in miss 67 2015 and 2016. Events are selected using triggers requiring signiﬁcant ET , with thresholds that depend miss 68 on the run period. The triggers are close to fully ecient for events with an oine-reconstructed ET 1 69 greater than 200 GeV. The data sample corresponds to an integrated luminosity of 36.1 fb with a relative 70 uncertainty of 2.1%, derived using methods similar to those described in Ref. [40].

71 Event samples from Monte Carlo (MC) simulation are used to model both the signal and speciﬁc processes 72 of the SM background. For the SUSY signal, two sets of simpliﬁed models [41–43] are used to guide the 73 design of the analysis: one based on Higgsino LSPs involving compressed electroweakino production, 74 and the other a model of compressed slepton pair production. In addition, a third model assuming the 75 production of wino-like gauginos decaying to a bino-like LSP (referred to as “wino-bino” model in the 76 following) is considered for the interpretation of the results.

77 Signal cross-sections are calculated to next-to-leading order (NLO) in the strong coupling, and next-to- 78 leading logarithm accuracy for soft-gluon resummation using R v1.0.7 [44–46].

20th September 2017 – 00:06 5 Higgsino Searches at ATLAS DRAFT • Final state: 2 opposite sign same j q j ` ﬂavor soft leptons p q p W * ˜ 0 • Challenging! ˜1± 0 ` ˜1 ˜1 - Low production ˜0 cross section 0 1 ˜0 ˜2 `˜ 1 Z * - Small Δm → low p ` p ETmiss, very soft ` ` leptons (a) (b)

0 Figure 1: Diagrams representing the two lepton final state of (a) associated electroweakino 2 1± and (b) slepton pair `` production in association with an initial state radiation jet. In addition to (a), this analysis is also sensitive H H to 0 0 and production. 25 October 2018 J. Gonski2HH1 1± 1⌥ !9 H H H H 54 muon spectrometer (MS). The ID provides precision tracking of charged particles in pseudorapidity region Not reviewed, for internal circulation only 55 ⌘ < 2.5, consisting of pixel and microstrip silicon subsystems within a transition radiation tracker. An | | 56 insertable B-layer [37] was added for ps = 13 TeV data-taking to improve tracking performance. These 57 are immersed in a 2 T axial magnetic field provided by a superconducting solenoid. High-granularity 58 lead/liquid-argon electromagnetic sampling calorimeters are used for ⌘ < 3.2. Hadronic energy deposits | | 59 are measured in a steel/scintillator tile barrel calorimeter in ⌘ < 1.7. Forward calorimeters extend the | | 60 coverage to 1.5 < ⌘ < 4.9 regions for both electromagnetic and hadronic measurements. The MS | | 61 comprises trigger and high-precision tracking chambers spanning ⌘ < 2.4 and ⌘ < 2.7, respectively, | | | | 62 surrounded by three large superconducting toroidal magnets. Events of interest are selected using a two- 63 level trigger system [38], consisting of a first-level trigger implemented in hardware, which is followed by 64 a software-based high-level trigger.

65 3 Collision data and simulated event samples

77 Signal cross-sections are calculated to next-to-leading order (NLO) in the strong coupling, and next-to- 78 leading logarithm accuracy for soft-gluon resummation using R v1.0.7 [44–46].

20th September 2017 – 00:06 5 Higgsino Searches at ATLAS DRAFT ISR kick = • Final state: 2 ETmiss trigger opposite sign same j q j ` flavor soft leptons p q p W * ˜ 0 • Challenging! ˜1± 0 ` ˜1 ˜1 - Low production ˜0 cross section 0 1 ˜0 ˜2 `˜ 1 Z * - Small Δm → low p ` p ETmiss, very soft ` ` leptons (a) (b) Low mll ~ Δm 0 Figure 1: Diagrams representing the two lepton final state of (a) associated electroweakino 2 1± and (b) slepton pair `` production in association with an initial state radiation jet. In addition to (a), this analysis is also sensitive H H to 0 0 and production. 25 October 2018 J. Gonski2HH1 1± 1⌥ !10 H H H H 54 muon spectrometer (MS). The ID provides precision tracking of charged particles in pseudorapidity region Not reviewed, for internal circulation only 55 ⌘ < 2.5, consisting of pixel and microstrip silicon subsystems within a transition radiation tracker. An | | 56 insertable B-layer [37] was added for ps = 13 TeV data-taking to improve tracking performance. These 57 are immersed in a 2 T axial magnetic field provided by a superconducting solenoid. High-granularity 58 lead/liquid-argon electromagnetic sampling calorimeters are used for ⌘ < 3.2. Hadronic energy deposits | | 59 are measured in a steel/scintillator tile barrel calorimeter in ⌘ < 1.7. Forward calorimeters extend the | | 60 coverage to 1.5 < ⌘ < 4.9 regions for both electromagnetic and hadronic measurements. The MS | | 61 comprises trigger and high-precision tracking chambers spanning ⌘ < 2.4 and ⌘ < 2.7, respectively, | | | | 62 surrounded by three large superconducting toroidal magnets. Events of interest are selected using a two- 63 level trigger system [38], consisting of a first-level trigger implemented in hardware, which is followed by 64 a software-based high-level trigger.

65 3 Collision data and simulated event samples

77 Signal cross-sections are calculated to next-to-leading order (NLO) in the strong coupling, and next-to- 78 leading logarithm accuracy for soft-gluon resummation using R v1.0.7 [44–46].

20th September 2017 – 00:06 5 Current Exclusion

[1]

[1] ATLASSummaryPlots 25 October 2018 J. Gonski !11 Current Exclusion

LEP2 [1]

[1] ATLASSummaryPlots 25 October 2018 J. Gonski !12 Current Exclusion

[1]

2L compressed

[1] ATLASSummaryPlots 25 October 2018 J. Gonski !13 Current ExclusionATL-PHYS-PUB-2017-019

What if the LSP is dominantly higgsino? [1]

• For a pure higgsino state: Δm~100 MeV + has long lifetime ~10-11sec χ1 ⇒

• Enough to sometimes go through the detector but then decay to χ 0 and soft pions 1

• Disappearing track signature!

• Main background: catastrophic interactionDisappearing track with the detector Pure higgsino/wino state: Δm~100 MeV ⇒ long lifetime ~10-11s [2] • Provides excellent [2]limits ATL-PHYS-PUB-2017-19 at small Δm! [1] ATLASSummaryPlots 25 October 2018 J. Gonski !14

39 Current Exclusion

[1]

Relic density disfavors pure higgsino LSP! (Well-Tempered Neutralino) [3]

[3] arXiv:0601041 [1] ATLASSummaryPlots 25 October 2018 J. Gonski !15 Areas of Improvement

[1]

New working point gives us muons down to 3 GeV (2019)

[1] ATLASSummaryPlots 25 October 2018 J. Gonski !16 Areas of Improvement

[1]

More luminosity

New working point gives us muons down to 3 GeV (2019)

[1] ATLASSummaryPlots 25 October 2018 J. Gonski !17 What’s Next?

• Natural SUSY is still well-motivated! (& experimental constraints are weakest in the electroweak sector) - Fine tuning < 10% possible for m( g˜ ) < 2.5 TeV and m( t˜ ) < 1.5 TeV • Next analysis: 1L + track, softer leptons, HL-LHC! (2026+)

200 400

25 October 2018 J. Gonski !18 What’s Next?

• Natural SUSY is still well-motivated! (& experimental constraints are weakest in the electroweak sector) - Fine tuning < 10% possible for m( g˜ ) < 2.5 TeV and m( t˜ ) < 1.5 TeV [4] • Next analysis: 1L + track, softer leptons, HL-LHC! (2026+)

200 400

[4] arXiv:1611.05873

25 October 2018 J. Gonski !19 What’s Next?

• Natural SUSY is still well-motivated! (& experimental constraints are weakest in the electroweak sector) - Fine tuning < 10% possible for m( g˜ ) < 2.5 TeV and m( t˜ ) < 1.5 TeV [4] • Up next: softer leptons, 1L + track, HL-LHC! (2026+)

200 400

[4] arXiv:1611.05873

25 October 2018 J. Gonski !20 Backup

25 October 2018 J. Gonski !21 can be reduced to a one-dimensional problem as in the Standard Model,

V = m2 H 2 + H 4 (2) H | | | |

2 Precision Corrections to Fine Tuning where mH will be in general a linear combination of the various masses of the Higgs ﬁelds with Buckley, Monteux, Shih arXiv:1611.05873 2 2 coecients that depend on mixing angles, e.g. in the MSSM. Each contribution, mH ,

Λ= 2 to the Higgs mass should be less�� than �� orm2 of= them2 order+ ofm2mH,m otherwise2 various contributions H H,bare 2 2 need to be finely tuned to cancel each other. Therefore mH /mH should not be large. By 2 2 (at tree level) using m2 = 2m2 one can define as a measureΔAndm the of= last| miracle fine-tuningμ| is… [26], h H In order for this (very naive) picture to hold, the SUSY particle spectrum must start at the TeV scale or lower

So we should be producing2 sparticles at the LHC13 2mH 2 . ≤ 10 (3) ⌘ mh !X μ ≤ 400GeV 2 Λ= � Here, mh reduces to the physical�� Higgs�� boson mass in the MSSM in the decoupling regime. In 2 fully mixed MSSM scenarios, or in more general potentials, mh will be a (model-dependent) 25 October 2018 J. Gonski !22 linear combination of the physical neutral CP-even Higgs boson masses. As is well known, increasing the physical Higgs boson mass (i.e. the quartic coupling) alleviates the fine- tuning [34, 35]. If we specialize to the decoupling limit of the MSSM and approximate the quartic coupling Figure 6: Top: 10% fine tuned regions with ⇤ = 20 TeV (left) and ⇤ = 100 TeV (right), with respect to the gluino mass (green)2 and the2 stop mass2 (blue, delimited by blue and purple 2 2 2 by its tree level value (g + g0 )cos 2, then we find that mh =cos 2 mZ . We then lines), for 1st/2nd generation/ squarks degenerate with the third generation (solid lines), at recover the5 TeV usual (dashed formula lines) and at 10 for TeV fine (dotted tuning lines). The in wedge-shaped the MSSM, intersection Eq. (delimited1,inthelargetan limit. by red lines) is the fully natural 10 region. Bottom: Left: same as the other plots, but  for ⇤ = 107 GeV. Right: same as the other plots, but2 for ⇤ = 1016 GeV and 1% fine-tuning, 3 In a SUSYthat is, the theory lines and shaded at treeregions correspond level, m to H 100.will include the µ term . Given the size of the  top quark mass, m2 also includes the soft mass of the Higgs field coupled to the up-type (top, left), 100 TeVH (top, right), 107 GeV (bottom, left) and 1016 GeV (bottom, right). The shaded areas mark the regions where the contribution to the Higgs fine-tuning is quarks, mH . Whether the soft mass for the down-type Higgs, mH , or other soft terms in lessu than 10% (except for ⇤ =1016 GeV, where we show 1% tuned regions instead): d in green is the natural region for the gluino, while in blue we see the stop natural an extended Higgs sector, should be as light as µ and mH is instead a model-dependent parameter space. The dashed and dotted lines indicate how the natural regions evolveu as the 1st/2nd generation squark are taken to 5 and 10 TeV, respectively. Red lines question, and a heavier mH can even lead to improvements [48]. The key observation that delimit the intersection of gluinod and stop regions, i.e. the region in which both gluinos is relevant for SUSY collider phenomenology is that higgsinos must be light because their mass is directly controlled by µ, 15

1/2 m 1 µ < 200 GeV h (4) ⇠ ✓120 GeV ◆ 20%!

2 It is straightforward to extend this discussion to include SM singlets that receive vevs, see for example [35]. 3 In theories where the µ-term is generated by the vev of some other ﬁeld, its e↵ective size is generically bound to be of the order of the electroweak scale by naturalness arguments. For a proof in the NMSSM see, e.g.,[35].

8 4 decoupled. Once electroweak symmetry is broken, the Higgsinos mix with the neutral bino, and both the charged and neutral components of the wino multiplet. This mixing splits the Higgsino multiplets, giving 4 0 0 slightly di↵erent masses to the two neutral Higgsino states 1, 2 and the charged state 1± and endowing these three states with adecoupled. small wino Once or bino electroweak component. symmetry The size is of broken, the splitting the Higgsinos and the hierarchy mix with among the neutral bino, and both the the three states dependscharged on the size and of neutralM1 and componentsM2 relative of to the each wino other multiplet. and to µ. This Note mixing that, had splits we chosen the Higgsino multiplets, giving the bino or wino to be the lightest electroweakino, the number of light states would be0 di↵erent;0 one neutral slightly di↵erent masses to the two neutral Higgsino states , and the charged state ± and endowing state for a light bino, or one neutral and one charged state for a light wino. 1 2 1 these three states with a small wino or bino component. The size of the splitting and the hierarchy among To get some idea for the typical splitting size and parametric dependences of the mass splitting, we first the three states depends on the size of M1 and M2 relative to each other and to µ. Note that, had we chosen proceed analytically and look in two simple limits, M M > µ and M M > µ . Throughout this the bino or wino to be the lightest1 electroweakino,2 | | the2 number1 | of| light states would be di↵erent; one neutral study we will take M and M to be strictly positive. The neutralino mass matrix in the MSSM in the 4 1 state for2 a light bino, or one neutral and one charged4 state for a light wino. (B0, W 0, 0, 0) basis is d u To get some idea for the typical splitting sizedecoupled. and parametric Once electroweak dependences symmetry of is the broken, mass the splitting, Higgsinos mix we with first the neutral bino, and both the decoupled. Once electroweak symmetry is broken, the Higgsinos mix with the neutral bino, and both the charged and neutral components of the wino multiplet. This mixing splits the Higgsino multiplets, giving charged and neutral components of the winoproceed multiplet. analytically ThisM mixing1 and splits look0 the in Higgsino twomW simple multiplets,t✓W c limits,m givingW t✓MW s1 M2 > µ and M2 M1 > µ . Throughout0 0 this e f 0 0 slightly di↵erent masses| to| the two neutral Higgsino| | states 1, 2 and the charged state 1± and endowing slightly di↵erent masses to the two neutral Higgsino states , and the charged state ± and endowing study we will take10 2 M1 andM2M2 tom beW strictly1c these positive.mW threes states The with neutralino a small wino mass or bino matrix component. in the The MSSM size of in the the splitting and the hierarchy among these three states with a small wino or binoM component.N0˜ 0 =00 0 The0 size of the splitting and the hierarchy among 1 (2) (B , W , d,m uW)t basis✓W c ismW c 0 the threeµ states depends on the size of M1 and M2 relative to each other and to µ. Note that, had we chosen the three states depends on the size of M and M relative to each other and to µ. Note that, had we chosen 1 2 B m t s m s µ the bino0 or winoC to be the lightest electroweakino, the number of light states would be di↵erent; one neutral the bino or wino to be the lightest electroweakino, theB numberW ✓ ofW light statesW would be di↵erent; one neutral C M1 state for0 a lightm bino,W t or✓W onec neutralmW t✓ andW s one charged state for a light wino. state for a light bino, or one neutral and onee chargedf @ state for a light wino. A while the chargino mass matrix is 0 ToM get2 some ideamW forc the typicalm splittingW s size and parametric dependences of the mass splitting, we first To get some idea for the typical splittingWhy size and parametricQuasi-Degenerate? dependencesMN˜ 0 of= the0 mass splitting,proceed we first analytically and look in two simple1 limits, M M > µ and (2)M M > µ . Throughout this mW t✓ c mW c 0 µ 1 2 2 1 proceed analytically and look in two simple limits, M1 M2 > µ and M2 M1 p> µ . ThroughoutW this | | | | | | M2 2|s|mW study we will take M1 and M2 to be strictly positive. The neutralino mass matrix in the MSSM in the M ˜ = B m t s m s µ 0 (3)C study we will take M1 and M2 to be strictly positive. TheC neutralinop mass matrix inW the✓W MSSM (B in0, W theW0, 0, 0) basis is 0 0 0 0 Depends on neutrino mass matrix2cm inW MSSM!B µ Using d u C (B , W , d, u) basis is ✓ @ ◆ A while the chargino mass matrix is M 0 m t c m t s where for simplicity we neglect all CP phases. e f 1 W ✓W W ✓W M1 0 mW t✓ c mW t✓ s e f W W 0 M2 mW c mW s M ˜ 0 = 0 1 (2) 0 M2 mW c mW s M2 p2smW N m t c m c 0 µ M ˜ 0 = (2) W ✓W W N 0 1 MC˜ = (3) Case I: M1 M2 >mWµt✓W c mW c 0 µ p B m t s m s µ 0 C 2cmW µ W ✓W W B m |t | s m s µ 0 C ✓ ◆ B C In this case, theB heavyW ✓W bino canW be integrated out. DependingC on the sign of µ, the mixing angle@ between A @ A while the chargino mass matrix is while the charginoW 0 and mass matrix0 = is1 0 where 0 , or forW simplicity0 and 0 = we1 neglect 0 + all 0 CP, can phases. be enhanced as M approaches µ.When p2 u d + p2 u d 2 ± M2 p2smW mW M2 µ, the splittingsM betweenp2s them mostly Higgsino states are MC˜ = (3) ⇥ ⇤ 2 W ⇥ ⇤ p2cmW µ ⌧ ⌥ MC˜ Case= I: M1 M2 > µ (3) ✓ ◆ f p2cmfW µ | | 2 In✓ this case, the heavy◆ binom canW (1 be integrateds2) where out. for Depending simplicity we on neglect the sign all CP of phases.µ, the mixing angle between m m 0 ⌥ (4) where for simplicity we neglect all CP phases. ± 1 W 0 and 0 = 1 1 0 ⇡0 ,2( orMW2 0+andµ ) 0 = 1 0 + 0 , can be enhanced as M approaches µ.When p2 u d + p2 u d 2 2 | | Case2 I: M1 M2 > µ ± Case I: M M > µ mW M2mWµ,(1 thes splittings2) between the mostlymW ( Higgsinoµ s2 + statesM|2)| are 1 2 0 0 0 In this case, the heavy bino can be integrated out. Depending on the sign of µ, the mixing angle between | | m m⌧± ⌥ ±⇥ , ⇤ m m ⇥±| | ⇤ (5) In this case, the heavy bino can be integrated2 f out.1 Depending⇡ 2(M onµ the) sign of µ, the2f mixing1 angle⇡ between0 (M 2 0 µ12) 0 0 0 0 1 0 0 2 W and2 2 = p u d , or W and + = p u + d , can be enhanced as M2 approaches µ.When 0 0 1 0 0 0 0 1 0 0 | | mW |(1|2 s2) 2 ± W and = p u d , or W and + = p u + d , can be enhanced as M2 approachesm mµm.When0W M2 µ, the splittings between the mostly Higgsino states are (4) 2 2 ± ± ⌥ where the index corresponds to when µ is positive or negative,1 respectively.1 ⌧ 2(M⌥ As+M⇥ µ2 )approaches⇤ µ the ⇥ ⇤ mW M2 µ, the splittings between the mostly Higgsino states are f ⇡ 2 f 2 ⌧ ⌥ ⇥ ±⇤ ⇥ ⇤ | | | | mW (1 s2) f wino fraction inf the lightestWino/bino neutralino as2 lightest increases, electroweakino, while the # light wino2 states fraction is different! in the next-to lightest2 neutralinom m 0 ⌥ (4) m (1 s ) m ( ±µ s +1 M ) mW (1 s2) W 2 W 1 2 ⇡ 22(M2 + µ ) m m•0 one neutral⌥ statem for0 a lightm bino; ± , (4) m 0 m 0 ±| | (5) remains approximately1± constant.1 From Eq.2 (4) and± Eq. (5) it is clear that the2 lightest1 neutralino 2 and 2 | | ⇡ 2(M2 + µ ) 1 ⇡ 2(M µ ) ⇡ (M2 µ ) 2 • one neutral |and| one charged state for a2 light wino. mW2(1 s2) mW ( µ s2 + M2) 2 2 | | m0 m ±| | , m0 m0 ±| | (5) chargino are morem degenerate(1 s ) than the second m neutralino( µ s and+ M the) lightest chargino. 2 1± 2 1 2 2 W 2 W 2 2 ⇡ 2(M2 µ ) ⇡ (M2 µ ) m0 m ± , m0 m0 ±| | (5) 2 1± where the 2 index1 corresponds2 to2 when µ is positive or negative, respectively. As| | M2 approaches µ the | | ⇡ 2(M2 µ ) ⇡ (M2 µ ) | | ± | | where the indexarXiv:1401.1235 corresponds to when µ is positive or negative, respectively.| | As M2 approaches µ the wino fraction in the lightest neutralino increases, while± the wino fraction in the next-to lightest neutralino | | where the Caseindex corresponds II: M 1 M to2 when>µµscenariois positive or negative, respectively. As M approacheswinoµ the fraction in the lightest neutralino increases, while the wino fraction in the next-to lightest neutralino ± 2 | | wino fractionIn in thisthe lightest case,25 October theneutralino heavy 2018 increases, winoremains component while approximately the winoJ. can Gonski fraction be integrated constant. in the next-to out. From lightest Similar Eq. neutralino (4)remains to theand approximately previous Eq. (5) scenario, it constant. is!23 clear depending thatFrom Eq. the (4) lightest and Eq. neutralino (5) it is clear and that the lightest neutralino and remains approximatelyon the sign constant. of µ, the From splittings Eq.chargino (4) and between are Eq. more (5) the it degenerate is lightest clear that neutralinos than the lightest the andsecond neutralino thechargino neutralino chargino and are moreare and degenerate the lightest than chargino. the second neutralino and the lightest chargino. chargino are more degenerate than the second neutralino and the lightest chargino. m2 t2 (1 s ) W ✓W 2 Case II: M1 M2 >µscenario Case II: M1m M2 >µm0scenario ± (6) Case II: M1 M2 >µscenario 1± 1 In this case, the heavy wino component can be integrated out. Similar to the previous scenario, depending In this case, the heavy wino⇡ component2(M1 µ can) be integrated out. Similar to the previous scenario, depending In this case, the heavy wino component can be integrated2 2 out. Similar to the previous scenario,| | dependingon2 the2 sign of µ, the splittings between the lightest neutralinos and the chargino are m t (1 s2) m t ( µ s2 + M1) on the sign of µ, the splittings between theon lightest the sign neutralinosW of✓Wµ, the and thesplittings chargino between are the lightestW ✓W neutralinos and the chargino are 2 2 m 0 m ⌥ , m 0 m 0 ±| | (7) m t (1 s2) ± 2 W ✓W 2 1 2 1 2 0 ± ⇡ 2(2 M2 1 + µ ) ⇡ (M µ ) m± m (6) m t (1 s2) 12 2 1 1 W ✓W | | m t | (1| s2) ⇡ 2(M1 µ ) m m0 ± (6) W ✓W | | 1± 1 m m 0 ± 2 2 2(6)2 ⇡ 2(M1 µ ) ± m t (1 s2) m t ( µ s2 + M1) Since t✓W 0.5, the mixing between the (heavy) bino and the1 Higgsino1 is smaller than the mixingW ✓W of a W ✓W | | ⇡ 2(M10 µ ) ⌥ 0 0 ±| | 2 2 2 2 m m± , m m (7) ' m t (1 s ) m t ( µ s + M ) 2 1 2 1 2 2 W ✓W 2 W ✓W 2 1 | | ⇡ 2(M1 + µ ) ⇡ (M1 µ ) (heavy)0 wino and the Higgsino ⌥ of the0 first case,0 leading±| to|2 a smaller2 overall splitting between the neutralinos2 2 m m± , m m 2 m t (1 s ) (7) m t ( | µ| s + M ) | | 2 1 ⇡ 2(M + µ ) 2 1 ⇡ (M Wµ 2)✓W 2 0 W ✓W 2 1 1 0 1 0 0 and the chargino. Furthermore,| | unlikem the firstm case, the splitting| | between⌥ Since,1t✓andWm01±.5,is them greater mixing than between between the±| (heavy)| bino and the Higgsino(7) is smaller than the mixing of a 2 1± '2 1 2 2 0 ⇡ 2(M1 + µ ) (heavy) wino and the Higgsino⇡ of the( firstM1 case,µ leading) to a smaller overall splitting between the neutralinos Since t✓W 0.5,and the mixing±. between the (heavy) bino and the Higgsino is smaller than the mixing of a ' 2 1 | | | | 0 (heavy) wino and the Higgsino of the first case, leading to a smaller overall splitting between the neutralinosand the chargino. Furthermore, unlike the first case, the splitting between 1 and 1± is greater than between Since t✓W 0.5, the mixing between the (heavy)0 bino and the Higgsino is smaller than the mixing of a 0 and ±. and the chargino. Furthermore, unlike the first case, the' splitting between 1 and 1± is greater than between2 1 0 Numerical Scan of the bino-wino(heavy) wino Parameter and the Space Higgsino of the first case, leading to a smaller overall splitting between the neutralinos and ±. 2 1 0 and the chargino. Furthermore, unlike the firstNumerical case, the Scan splitting of the bino-wino between Parameter1 and Space1± is greater than between 0 Numerical Scan of the bino-wino Parameter2 Spaceand 1±.

Numerical Scan of the bino-wino Parameter Space But we haven’t found anything yet! The Stop

1500

Limits up-to ~1 TeV in stop mass … but no limits if m(LSP) ≳ 400 GeV

25 October 2018 J. Gonski !24 But we haven’t found anything yet!

The Gluino

2500

25 October 2018 J. Gonski !25 ATLAS SUSY Summary

25 October 2018 J. Gonski !26 Pre-selection

Starting with some2L basicHiggsino cuts following Search the suggestions from theoretical papers and from the Run-2 analysis:

• Met > 200 GeV → trigger • pT(jet1)>200 GeV → ISR • |!"(jet,Met)| > 1.0 → mis-measured Met • nJet(50 GeV)=1,2 → no jets expected from the signal • bJet veto → ttbar background • upper cut on lepton pT → soft leptons in the ﬁnal state; • upper cut on mll → small invariant masses expected; • Met/Ht → good discrimination as seen in the Run-2 analysis; • m## → Z## background;

25 October 2018 J. Gonski 6 !27 Peter Tornambè (Uni-Freiburg) Higgsino Working Group 02-10-’17 Making ETmiss with ISR

• With no other ﬁnal state objects, LSPs are back to back, no ETmiss in event

• Require ISR jet: collimate LSPs, generate measurable ETmiss for trigger

LSP1 pTLSP1 = 100 GeV pT = 100 GeV

pTLSP2 = 100 GeV

Total ETmiss = 0 GeV Total ETmiss = 200 GeV

25 October 2018 J. Gonski !28