Searches for Hidden Sectors Using Lepton Signatures at CMS

Alexei Safonov

Texas A&M University

for the CMS Collaboration LHC Workshop, UChicago, November 2012

Why Searching for Hidden Sectors?

• • • Dark matter is one big reason ▫ If satellite experiments excesses have something to do with the dark matter, these could signify presence of dark sectors Higgs “problems”: ▫ While SUSY can be the answer to the hierarchy problem, MSSM isn’t that great of a candidate  Fine tuning, the m -problem etc.  …and in all likelihood it’s wrong anyway  ▫ NMSSM can help some of these problems  Yields a more complex higgs sector with new fields weakly coupling to SM particles Finally, they just might be there… A. Safonov, LHC Workshop, UChicago, November 2012 2

•

TeV Scale Dark Matter

PAMELA and Fermi observe rising positron fraction towards higher energy: ▫ Unknown pulsar? Cosmic rays interacting with giant molecular clouds?

▫ Or heavy dark matter annihilation in the galactic halo with a large x-section:  Light dark photon long-distance force between slow WIMPs 𝛾 𝐷 : an attractive    Sommerfeld enhancement 𝛾 𝐷 can weakly couple to SM via kinetic mixing with photon As no antiproton excess observed, M( 𝛾 𝐷 ) ≲ O(1 GeV) X X 

arXiv:1109.0521v1

  m m m m + + , e , e , e , e + + A. Safonov, LHC Workshop, UChicago, November 2012 3

NMMSM Phenomenology

• • Modified superpotential: ▫ MSSM: 𝜇𝐻 𝑢 𝐻 𝑑 ▫ ▫ NMSSM: λ𝑆𝐻 𝑢 𝐻 𝑑 + 1 κ𝑆 3 3 NMSSM less fine tuning and solves m -problem: m is generated by singlet field VEV and naturally has EW scale • • More complex Higgs sector: ▫ 3 CP-even higgses h 1,2,3 , 2 CP-odd higgses a 1,2 ▫ a 1 is hidden as it is mostly singlet and weakly couples to SM particles except through h 1 Experimentally relevant decays:  ℎ 1,2 → 𝑎 1 𝑎 1 (Branchings depend on mixing)  𝑎 1  → 𝑓𝑓 (standard higgs hierarchy) Couplings are weak but it has to decay somewhere A. Safonov, LHC Workshop, UChicago, November 2012 4

A “Long Living” Example

• A separate hidden strongly interacting sector coupling to SM only through a heavy Z’ ▫ Visible higgs(es) can naturally mix with the hidden higgs  If Z’ is heavy, “hidden pions” can easily have decay lengths O(0-100 cm) ▫ Z-like decay hierarchy for new hidden bosons Strassler, Zurek, PLB 661 (2008) •  One can have models with higgs-like decays too Striking signatures, relatively easy to look at A. Safonov, LHC Workshop, UChicago, November 2012 5

Hidden Sectors Search Strategies

• • Aim to produce something that links visible and hidden sectors and look for evidence of new hidden states: ▫ In the dark SUSY the “stable” visible LSP has no choice but to decay to hidden states even if small couplings   If we can make the LSP either through squark/gluino production or Higgs, we can see its decay products ▫ In the NMSSM new higgs states can have very weak coupling to SM, but appreciable coupling to the SM like higgs due to mixing – look for exotic higgs decays Similar story for the “long living” example model Brute force: make hidden sector particles ▫ Because of typically small couplings, need high luminosity and clean final states A. Safonov, LHC Workshop, UChicago, November 2012 6

Dark Photons in SUSY Cascades

• SUSY with squarks/gluinos accessible by LHC: • ▫ MSSM LSP is a neutralino decaying to dark neutralino and light g dark /h dark ▫ MSSM LSP is a squark decaying to q and light dark fermion and g dark /h dark Dark photons decay as SM g A. Safonov, LHC Workshop, UChicago, November 2012

Branching fraction of

𝜸 𝑫

arxiv:1002.2952

7

Selections

• • • Data: ▫ 35 pb -1 of 2010 LHC data ▫ Inclusive muon trigger p T >15 GeV Offline: ▫ Require at least 1 muon with p T >15 GeV, | h |<0.9

▫ Identify all other muons with p T >5 GeV, | h |<2.4

▫ Reconstruct muon jets and categorize ▫ No isolations, cluster using pairwise mass of muons Assume new bosons produced on-shell: A. Safonov, LHC Workshop, UChicago, November 2012 8

Topologies: Data and Backgrounds

name

R 1 2 R 1 4 R 2 22 R N 5+

description

Single dimuon+X

Lead

m

-Jet p T

>80 GeV/c Single quadmuon+X no explicit cut Two dimuons+X All other categories no explicit cut no explicit cut

Backgrounds

2 m ’s from a b-jet, Drell Yan 2 m ’s from a b-jet + 2 fakes bb-bar+X, 2 m ’s from each b Rare, from bb-bar+X/fakes • No events with consistent masses of dimuons in higher order categories A. Safonov, LHC Workshop, UChicago, November 2012 9

Model-Independent Interpretation

• • Use three simplest topologies to set “conservative” model independent limits: ▫ Dimuon+X ▫ Two-dimuon+X ▫ Quadmuon+X Limits of applicability: ▫ Mean p T ( m -jet)≤250GeV Easy to apply to other models: • ▫ Follow analysis steps to calculate branching and acceptance for a specific final state assuming an ideal detector ▫ Compare with the limit plot  Complex topologies can be reduced to one of these three A. Safonov, LHC Workshop, UChicago, November 2012 10

Models with TeV Scale Dark Matter

Model from JHEP 04 (2009) 014.

• MSSM LSP is a squark decaying to a quark, light dark fermion and either g dark (left) or h dark (right) More details in CMS-EXO-11-013 and JHEP07 (2011) 098 A. Safonov, LHC Workshop, UChicago, November 2012 11

Search for Displaced Lepton Pairs

• • Generic search for H

0

decays X  ▫ 2011 data: L=4 fb -1  XX, using leptonic ll, with X having substantial lifetime for l=e, L=5 fb -1 for l= m Selections: ▫ Displaced e/ | h |<2 with p T >41/33 GeV and d 0 / s d >3/2 ▫ Require at least one X-candidate per event:    m candidate defined as a track within A common vertex with c 2 /ndf>4, displaced more than 8(5) s vtx-fit from the beamline for e( m) channel M(ee/ mm )>15 GeV, D R( mm )>0.2, p T (e) from ECAL Isolation: S p T <4 GeV counting tracks w/ p T >1 GeV in D R(trk,e/ m )<0.4 around each lepton (but not counting the other lepton in the X-candidate) A. Safonov, LHC Workshop, UChicago, November 2012 12

Search for Displaced Lepton Pairs

• • Avoid standard lepton ID: ▫ Inefficient for displaced tracks Efficiency driven by tracking reconstruction efficiency ▫ Cross-checked with cosmic muon data A. Safonov, LHC Workshop, UChicago, November 2012 • • Backgrounds dominated by Drell Yan events ▫ Shape from simulation cross checked with data Normalization from the fit of the vertex L xy / s distribution ▫ B= 1.4

+1.8

−1.2

channel and 0.02

+0.09

−0.02

for e/ m 13

Search for Displaced Lepton Pairs

• • Signal region (significant L xy ): no excess ▫ N0 events with X mm candidates ▫ 4 events X  ee candidates (expected 1.4

+1.8

−1.2

Limits as a function of the new boson mass ) A. Safonov, LHC Workshop, UChicago, November 2012 14

Search for Displaced Lepton Pairs

• Limits vs lifetime for M(H)=200 GeV and 1 TeV ▫ Reflect track reconstruction efficiency dependence on decay path length ▫ More details in CMS-EXO-11-101 (public note) A. Safonov, LHC Workshop, UChicago, November 2012 15

Light Dark Sectors and Higgs

• NMSSM: 𝑝𝑝 → ℎ 1,2 → 𝑎 1 𝑎 1 → 4𝜇 ▫ Either h1 or h2 (or both) can decay to a component dependent 1 a 1 , BR depends on the singlet ▫ Production cross-section for h and BR highly model • Dark SUSY with light dark photons: • 𝑝𝑝 → ℎ → 2𝑛 1 → 2𝑛 𝐷 + 2𝛾 𝐷 → 2𝑛 𝐷 + 4𝜇 ▫ Similar signature, but softer dimuons and missing energy A. Safonov, LHC Workshop, UChicago, November 2012 16

Light Dark Sector Higgs Limits

• • Spin-off of the 35 pb -1 ▫ ~5fb -1 of 2011 data Following Phys. Rev. D 81 (2010) 075021.

muon jet analysis Focus on the topology with two muon pairs of consistent mass ▫ Di-muon trigger with p T >17 and p T >8 GeV ▫ Same selections, but apply loose track-based isolation  High signal efficiency, strong suppression of bb backgrounds  Insensitive to pile-up ▫ No pairs with consistent mass found  Expect ~ 1 event A. Safonov, LHC Workshop, UChicago, November 2012 17

Exotic Higgs Limits

• • Express in terms of limits on production ▫ NMSSM: pp  h 1 or h 2  a 1 a 1  4 m ▫ Dark SUSY: pp  h c 1 c 1  g d g d c 1 c 1  4 m + MET Plots use SM higgs production cross-section ▫ Most of time not true, but convenient benchmarking A. Safonov, LHC Workshop, UChicago, November 2012 18

NMSSM Parameter Space

• • To gauge what it does to the NMSSM parameter space, scan NMSSM parameter space ▫ Focus on m a <2m cross-sections t ▫ Use actual NMSSM  Not SM ▫ See EXO-12-012 for details Make deep inroads into the allowed space A. Safonov, LHC Workshop, UChicago, November 2012 19

Search for Direct NMSSM a

1

Production

• • Production via gluon fusion ▫ Large cross-section if mixing with MSSM A is large  Suppressed as a constraints (cos 1 2 q has to be highly singlet to abide experimental A <1) ▫ Usual enhancement with tan b Search for resonances in dimuon spectrum ▫ 5.5

Search for Direct NMSSM a

1

Production

• • Analysis selections: ▫ Muon p T >5.5 GeV and | h |<2.4

▫ Isolation (per muon): In the two signal regions, fit for the sum of Crystal Ball (signal) and a 1 st order polynomial (background) ▫ Plus the radiative tail of Upsilon for the low mass region A. Safonov, LHC Workshop, UChicago, November 2012 21

Search for Direct NMSSM a

1

Production

• • No significant excesses in data Limits vs m(a) on the production rate ▫ Further interpretation in terms of cos q A and tan b A. Safonov, LHC Workshop, UChicago, November 2012 • LHC limits start superseding those from BaBar 22

Summary

• • • Several CMS analyses aiming at searches for hidden sectors ▫ Different scenarios for production mechanisms and the lifetime of the new hidden bosons ▫ Electron channels starting gaining ground ▫ When possible, results presented in a quasi model independent fashion to allow future interpretations No discoveries, but the new ground in sensitivity ▫ Important complementarity to the SM Higgs searches as the searches for exotic higgs decays can rule out many non-SM scenarios The data keeps coming in, so stay tuned A. Safonov, LHC Workshop, UChicago, November 2012 23