Probes of dark matter at the LHC

Transcription

1 Probes of dark matter at the LHC G. Bélanger LAPTH Annecy-le-Vieux DMAstroLHC Mass2017, Odense, 4 May 2017

2 Motivation Strong evidence for dark matter from astrophysical and cosmological observations at different scales Precise determination of amount of CDM (PLANCK) W cdm h 2 = / DM a new particle? 2

3 A wide variety of DM candidates WIMPs FIMPs SIMPs GIMPs Asymmetric SIDM

4 T. Tait.And a wide variety of models

5 WIMPs One class of candidates : weakly-interacting massive particles Lead to roughly correct amount of DM Thermal equilibrium in early Universe Typical weak interaction -> Wh 2 ~0.1 Also coannihilation when new particles nearly degenerate with DM - Boltzmann suppression exp(-dm/t) can be compensated by larger cross sections

6

7 Probing the nature of dark matter All determined by interactions of WIMPS with Standard Model Specified within given particle physics model

8 WIMP searches Were expecting lots of new particles at TeV scale as soon as LHC turned on, wanted to measure their properties but no excess!! Colliders -> constraints on DM and on newparticles Limitsfromastroparticle searches Direct detection (LUX, CDMS, Xenon, Cresst, DAMIC, DAMA.) Indirect detection (FermiLAT, HESS, Magic, AMS, Icecube ) with photons, positrons, antiprotons, neutrinos Hints in astroparticle searches (S. Profumo s talk) Fermi-LAT Galactic Center, PAMELA, AMS02

9 DM searches Sensitive enough to probe DM models Ongoing Xenon1T m<10gev more challenging Gamma rays from Dwarfs robust limits Probe generic annihilation cross section for DM below ~70GeV

10 Searches for dark matter at the LHC Can only check for a stable particle at the colllider scale not cosmological scale

11 Beyond the standard model For many years clear direction on how to explore BSM/DM Start with problems with SM: symmetry breaking, Higgs, unification, fermion masses One of the problems with SM SUSY XtraDim Little Higgs. Discrete symmetry WIMP Dark matter Interplay Collider, dark matter searches, cosmology More possibilities for WIMPs and for DM ( J. March-Russell s talk)

12 Another appproach, bottom-up Start with stable neutral particle, and build from there (mediator, other dark particles) exploiting hints from data at LHC or astroparticle searches

13 DM production at LHC The traditional searches - DM in decay chain of new particles preferably coloured or charged, e.g. neutralino in SUSY Signature : MET + jet, leptons model dependent

14 DM production at LHC The model independent approach Direct production of DM and Initial state radiation of gluon, photon.. serves as a trigger : monojet, monophoton, monox Signature : jet + large missing ET

15 DM production at LHC Exploiting the Higgs : search for invisible decays of the Higgs (relevant only if mdm<mh/2) Charged tracks and displaced vertices - for long-lived nextlightest dark sector particle : typically small mass splitting or very weak interactions Search for new particle (mediator) in SM final states

16 EFT/simplified model approach Direct production of pairs of DM + radiation : high ET miss + single jet/photon/boson Effective interaction operators For each operator : monojet limit compared to limit from direct detection Caveats : monojet limit valid assuming scale NP large, may not be valid at LHC energies-> simplified models

17 Looking for monojet within large SM background no excess > constraint on simplified model CMS Allowed D5 D8

18 Constraints from monojet for different mediators vs (in)direct detection CMS Assume M DM <M med /2 Allowed D5 D8

19 Simplified models Capture essential features with small number of parameters/assumptions SM + mediator +DM + some Z 2 symmetry Specific example : pseudoscalar mediator, fermion DM, also assume couplings proportional to Yukawas-> 3rd generation Loop coupling to two-gluons and two-photons

20 An example : pseudoscalar Requirement of relic density compatible with PLANCK and indirect searches for DM (photons from Dwarf Galaxies FermiLAT), antiprotons Narrow range of couplings allowed Most of favoured region will be probe in < 10 years by FermiLAT

21 At the LHC Several probes : monojet searches for mediator in visible (gg,tt,tt)or invisible decays, contribution of mediator to di-top cross section, associated production of mediator, tta, bba

22 At the LHC Several probes : monojet, searches for mediator in visible or invisible decays, contribution of mediator to di-top cross section Large couplings of A to quarks Banerjee, Barducci, GB, Fuks, Goudelis, Zaldivar, 1705.xxx

23 At the LHC LHC constraints strongly depend on mediator couplings to quarks Independent of coupling to DM in visible channels allow to cover the region m DM ~m A /2 with very small coupling hard for indirect detection

24 At the LHC With enhanced couplings to b quarks and taus importance of associated production Main constraint from visible channel, small invisible width, gga production suppressed no monojet

25 Other simplified models Importance of visible channels also for other simplified model: spin 1 or 2 Kraml, Laa, Mawatari,Yamashita,

26 Top- down approach Supersymmetry &extra dimensions

27 Supersymmetry Motivation: unifying matter (fermions) and interactions (mediated by bosons) Prediction: new particles supersymmetric partners of all known fermions and bosons : differ spin 1/2 Not discovered yet Hierarchy problem SUSY particles can stabilize Higgs mass against radiative corrections Quadratic divergences in Higgs mass corrrections cancelled when SUSY broken softly, TeV scale àshould be within reach of LHC R-parity to prevent proton decay -> LSP stable ->dark matter MSSM : Minimal field content : partner of SM particles and two higgs doublets (for fermion masses) Neutralinos : neutral spin ½ partners of gauge bosons (bino,wino) and Higgs scalars (higgsinos)

28 The neutralino mass matrix Mass and nature of neutralino LSP : determined by smallest mass parameter M 1 < M 2, µ bino µ < M 1, M 2 Higgsino ( in this case mχ 1 ~mχ 2 ~mχ + ) M 2 < µ, M 1 wino Determine couplings of neutralino to vector bosons, scalars

29 Neutralino DM Neutralino is mixed state exact nature will determine its annihilation properties Vary µ, M 1, M 2 to change nature of LSP, tanb = 10, all other SUSY parameters set to 4TeV Wino higgsino bino 1000GeV GeV f B In general neutralino LSP can only be subdominant DM component unless TeV scale Exception : bino overdominant Higgsino and wino mean degenerate particles µ at TeV scale is not natural from Higgs points of view

30 Direct detection Correct relic Xenon1T will probe large regions of parameter space Coupling of LSP to Higgs maximal for mixed gaugino/higgsino Constraints from DD (LUX) on neutralinos (mixed higgsino-bino) that naturally reproduce measured relic density Bino-wino escape detection also TeV scale DM Natural SUSY : µ small (higgsino content LSP)

31 SUSY dark matter at LHC pp collider 7-8TeV (Run 1) and 13TeV (Run2) some results for 36fb -1 Exploit many DM signatures Production of coloured/charged particles: DM in decay chain (MET+..) DM in Higgs decays Monojet- type signatures : not relevant for DM production but for compressed spectra Charged tracks and displaced vertices (for quasi stable NLSP )

32 LHC -SUSY search channels Largestcross sections : coloured particles Signatures : jets +Missing transverse energy +. Best limits ~ 2 TeV (gluinos) No stops -> tension for the Higgs sector (fine-tuning) Is SUSY hiding? What aboutdark matter?

33 Relevance for DM Stop important for DM if contribute to coannihilation - typical mass splitting ~40 GeV monojet ~Dm required for correct relic for bino+stop coann ATLAS,

34 Electroweak-inos Direct connection with dark matter (neutralino sector) Reach dependent on search channel (here simplified model)

35 Electroweak-inos Weak constraints on charginos which decay into gauge bosons

36 Electroweak-inos : χ 0 1 rches ] LSP Weak constraints on charginos which decay into gauge bosons Even more so in the framework of full model (here pmssm) MSSM with 19 parameters Fraction of Models Excluded ) [GeV] 0 1 m(χ ATLAS 1 s=8 TeV, 20.3 fb χ ± χ W (*) 0 (*) 0 χ Z χ 1 1 pmssm: χ Electroweak searches [ ] 0 1 LSP Fraction of Models Excluded χ ) [GeV] ± m( χ ) [GeV] (b) Chargino neutralino 1 hes (as listed in Table 1) (a) on the plane and (b) on the ±1 0 1 plane. from Ref. [53] is for a simplified model that assumes pure-wino ±

37 Long-lived charged particles Relevant for wino-lsp (chargino lifetime ns) T. Kaji, Moriond 2017

38 What s left after LHC (Run 1) production of DM + jet from ISR and/or compressed spectra ATLAS Analysis All LSPs Bino-like Wino-like Higgsino-like 0-lepton jets + ET miss 32.1% 35.8% 29.7% 33.5% 0-lepton jets + ET miss 7.8% 5.5% 7.6% 8.0% 0/1-lepton + 3b-jets + ET miss 8.8% 5.4% 7.1% 10.1% 1-lepton + jets + ET miss 8.0% 5.4% 7.5% 8.4% Monojet 9.9% 16.7% 9.1% 10.1% SS/3-leptons + jets + ET miss 2.4% 1.6% 2.4% 2.5% ( /`) + jets + ET miss 3.0% 1.3% 2.9% 3.1% 0-lepton stop 9.4% 7.8% 8.2% 10.2% 1-lepton stop 6.2% 2.9% 5.4% 6.8% 2b-jets + ET miss 3.1% 3.3% 2.3% 3.6% 2-leptons stop 0.8% 1.1% 0.8% 0.7% Monojet stop 3.5% 11.3% 2.8% 3.6% Stop with Z boson 0.4% 1.0% 0.4% 0.5% tb+et miss, stop 4.2% 1.9% 3.1% 5.0% `h, electroweak leptons, electroweak 1.3% 2.2% 0.7% 1.6% 2-, electroweak 0.2% 0.3% 0.2% 0.2% 3-leptons, electroweak 0.8% 3.8% 1.1% 0.6% 4-leptons 0.5% 1.1% 0.6% 0.5% Disappearing Track 11.4% 0.4% 29.9% 0.1% Long-lived particle 0.1% 0.1% 0.0% 0.1% H/A! + 1.8% 2.2% 0.9% 2.4% Total 40.9% 40.2% 45.4% 38.1%

39 What s left after LHC (a) Before ATLAS Run 1 (b) After ATLAS Run 1 ATLAS

40 Higgs invisible At LHC Measurement of Higgs in various production and decay modes Global fit to Higgs couplings and comparison with SM à Upper limit on invisible/not detected BR Can be combined with direct searches for invisible Higgs Current limits Brinv < 28% (CMS) Brinv < 24% (ATLAS) Future LHC : with 3000fb -1 can reach 5% At ILC : reach 0.4% 4 0

41 Generally in Higgs portal type model, both invisible width and SI cross section depend on h coupling to DM Light DM model are constrained, Djouadi et al ] 2 WIMP-nucleon cross section [cm DAMA/LIBRA (99.7% CL) CRESST II (95% CL) CDMS SI (95% CL) CoGeNT (99% CL) CRESST II (90% CL) SuperCDMS (90% CL) XENON100 (90% CL) LUX (90% CL) 1 10 ATLAS s = 7 TeV, fb -1 s = 8 TeV, 20.3 fb Vis. & inv. Higgs boson decay channels [κ W, κ Z, κ t, κ b, κ τ, κ µ, κ g, κ γ, κ Zγ, BR ] inv No κ W,Z assumption: BR <0.22 at 90% CL inv ATLAS 90% CL in Higgs portal model: Scalar WIMP Majorana WIMP Vector WIMP WIMP mass [GeV] 3 ATLAS Collaboration, arxiv:

42 Invisible Higgs If neutralino DM is light (<62GeV) contributes to invisible width of the Higgs Framework : MSSM with heavy sfermions After applying constraints from relic density (upper limit), Higgs at LHC, searches for chargino/neutralino, flavour +LEP

43 Invisible Higgs If neutralino DM is light (<62GeV) contributes to invisible width of the Higgs Will be completely probe in ongoing direct detection searches (Xenon1T) Barman, GB, Bhattacherjee, Godbole, Mendiratta, Sengupta,

44 Relaxing relic density constraint Beyond standard cosmological scenario Some higgsino component of LSP no longer required to ensure efficient annihilation Higgs invisible width can be very small

45 Relaxing relic density constraint Combination of direct detection (SI and SD), LHC searches, precise Higgs invisible (ILC) can cover large portion of parameter space neutralino below 10 GeV most difficult Specific LHC signatures in particular production of heavier electroweakinos (need high luminosity)

46 Projections Much to gain with higher luminosity since small cross section for electroweakinos L. Shchutska Far from covering the DM preferred region (TeV for higgsino)

47 Projections bino - wino : fairly unconstrained direct detection insensitive If nearly pure wino : mass splitting small, chargino long lifetime - >charged probe 2TeV wino (DM favoured) If mixed : compressed spectra, electroweakino production 100TeV collider 15ab -1, Bramante et al,

48 Summary MSSM+DM Higgs mass à fine-tuning issue with MSSM, heavily mixed stops One solution : add a singlet (NMSSM) Coloured sector under pressure by LHC if below TeV unless small mass difference with LSP Electroweak sector still wide open Still plenty of room for neutralino DM, altough as a single DM component somewhat under pressure by relic + direct search Higgsino case particularly difficult to probe at LHC Complementarity with indirect detection Other susy candidates possible : gravitino, sneutrino, axino

49 Another example : mued Theory of quantum gravity and unification of all interactions Universal extra dimensions : flat compact dim of size 10-18m, all fields propagate in the bulk mued: one extra dim size R compactified on a circle Higgs mass : free parameter After compactification : SM as effective thery valid until scale L Each SM particle has infinite tower of partner particles KK particles : same spin as their SM component Conserved momentum in 5th dimension leads to conserved KK number KK parity implies lightest KK particle is stable KK=(-1) n

50 Vector boson DM UED At tree level masses at each KK level are degenerate Radiative corrections are crucial in determining exact mass splitting and LKP Minimal UED: LKP is B (1), partner of hypercharge gauge boson (spin 1) s-channel annihilation of LKP (gauge boson) typically more efficient than that of neutralino LSP > TeV scale DM Many degenerate particles -> coannihilations and annihilation enhancement by resonances natural Parameters : cut-off scale L, R -1, m h

51 Scale for UED DM Relic density strongly depends on coannihilation and contribution of level 2 particles in s-channel and in final state (since decay into SM particles) M. KakizakiGB, Kakizaki, Pukhov JCAP(2011)

52 mued at LHC Reinterpretation of SUSY searches with missing ET and jets, leptons Large production cross section for coloured partners, g1g1, Q1Q1,g1Q1 Stop search 1l+jets+MET Deutchsmann, Flacke, Kim, The region compatible with DM relic density under tension with LHC data

53 If DM is not a WIMP : signals at colliders?

54 FIMPS (Feebly interacting MP) Freeze-in (Hall et al ): in early Universe, DM so feebly interacting that decoupled from plasma Assume that after inflation abundance DM very small, interactions are very weak but lead to production of DM T~M, DM freezes-in - yield increase with interaction strength Several possibilities for FIMP DM Production by annihilation or by decay Signatures: indirect detection from X decay into DM+SM particles ->boost factor. Relic abundance and DM annihilation cross section no longer related

55 FIMP from decay Case where FIMP is DM, next to lightest odd particle has long lifetime freeze-out as usual then decay to FIMP e.g. in SUSY : gravitino or RH sneutrino Partner of LH neutrino NOT a good DM candidate Very large contribution to direct detection- through Z exchange (Falk,Olive, Srednicki, PLB354 (1995) 99) Neutrino have masses RH neutrino + supersymmetric partner wellmotivated if LSP then can be dark matter Thermalized : non-negligible L-R mixing (Arina et al, ) or new gauge interactions MSSM+U(1) (GB, DaSilva, Laa, Pukhov ), both are viable with respect to LHC constraints and feature new signatures Or not thermalized abundance from decay of other particles

56 MSSM+RH (s)neutrino The framework : MSSM + three generations (n R + sneutrinor). Assume pure Dirac neutrino mass Superpotential Small Yukawa couplings O(10-13 ) depending on assumption : neutrino mass saturates atmospheric neutrino or cosmological bound with degenerate neutrino Sneutrino mass same order as other sfermions can be LSP

57 Sneutrino not thermalized in early universe - produced from decay of MSSM-LSP after freeze-out Relic density obtained from that of the NLSP can be charged Consider stau as the NLSP - live from sec to min : decay outside detector Collider constraints Higgs; flavour constraints; susy searches (mostly not valid because stau is collider stable and charged); charged stable particles Constraints from BBN : lifetime of stau can be long enough for decay around or after BBNà impact on abundance of light elements Decay of particle with lifetime > 0.1s can cause non-thermal nuclear reaction during or after BBN spoiling predictions in particular if new particle has hadronic decay modes Kawasaki, Kohri, Moroi, PRD71, (2005) After all constraints, room for sneutrinor DM

58 LHC signatures Characteristic signature : stable charged particle NOT MET Searches Cascades : coloured sparticles decay into jets + SUSYà N jets + stau Pair production of two stable staus (model independent but lower cross section) Passive search for stable particles Stable stau behaves like «slow» muons b=p/e<1 Use ionisation properties and time of flight measurement to distinguish from muon kinematic distribution Banerjee, GB, Mukhopadyhyay, Serpico,

59 Charged tracks from cascades Long lived Dominant contribution from squark pairs (heavy gluinos) Background : tt,µµ+jets, WW,WZ strongly suppressed with cuts (approach suggested in Gupta et al PRD (2007)) Luminosity 1ab -1 can probe mass ~580GeV Dependence on mass of squarks Stau pair production has lower reach Banerjee, GB, Mukhopadyhyay, Serpico,

60 MoEDAL detector Passive detector Array of nuclear track detector stacks Surrounds intersection region point 8 Sensitive to highly ionising particles Does not require trigger, one detected event is enough Major condition : ionizing particle has velocity b<0.2 B. Acharya et al, Stau velocity distribution

61 Conclusion LHC put strong constraints on new physics in channels with missing ET and SM final states Impact on DM in simplified or UV complete models in conjonction with direct detection Still many channels to explore for Run2 with 2016 data (35fb -1 ) and in 2018 with 120fb -1 With searches for long-lived and stable particles signatures from whole classes of DM candidates/models