Strategies to Search for Supersymmetry

John Ellis

Warsaw, May 18th, 2007

Outline

• Introduction to supersymmetry

• Standard search strategy:– Neutralino dark matter– Missing-energy signature– Hadronic sparticle decays?

• Gravitino dark matter– Stau NLSP

• Metastable charged particle

• Stau or stop?

Why Supersymmetry (Susy)?

• Hierarchy problem: why is mW << mP ?

(mP ~ 1019 GeV is scale of gravity)• Alternatively, why is

GF = 1/ mW2 >> GN = 1/mP

2 ?• Or, why is

VCoulomb >> VNewton ? e2 >> G m2 = m2 / mP2

• Set by hand? What about loop corrections?

δmH,W2 = O(α/π) Λ2

• Cancel boson loops fermions• Need | mB

2 – mF2| < 1 TeV2

Loop Corrections to Higgs Mass2

• Consider generic fermion and boson loops:

• Each is quadratically divergent: ∫Λd4k/k2

• Leading divergence cancelled if

Supersymmetry!

2

x 2

Astronomers tell us that most of the matter in the universe is invisible

We will look for it

with the LHC

Dark Matter in the Universe

Astronomers saythat most of thematter in theUniverse isinvisible Dark Matter

‘Supersymmetric’ particles ?

We shall look for them with the

LHC

Dark Matter in the Universe

Lightest Supersymmetric Particle

• Stable in many models because of conservation of R parity:

R = (-1) 2S –L + 3B

where S = spin, L = lepton #, B = baryon #

• Particles have R = +1, sparticles R = -1:Sparticles produced in pairs

Heavier sparticles lighter sparticles

• Lightest supersymmetric particle (LSP) stable

Fayet

Possible Nature of LSP

• No strong or electromagnetic interactionsOtherwise would bind to matterDetectable as anomalous heavy nucleus

• Possible weakly-interacting scandidatesSneutrino

(Excluded by LEP, direct searches)Lightest neutralino χ (partner of Z, H, γ)Gravitino

(nightmare for astrophysical detection)

Constraints on Supersymmetry

• Absence of sparticles at LEP, Tevatron

selectron, chargino > 100 GeV

squarks, gluino > 250 GeV

• Indirect constraints

Higgs > 114 GeV, b → s γ

• Density of dark matter

lightest sparticle χ:

0.094 < Ωχh2 < 0.124

3.3 σeffect ingμ – 2?

• Particles + spartners

• 2 Higgs doublets, coupling μ, ratio of v.e.v.’s = tan β• Unknown supersymmetry-breaking parameters:

Scalar masses m0, gaugino masses m1/2, trilinear soft couplings Aλ, bilinear soft coupling Bμ

• Often assume universality:Single m0, single m1/2, single Aλ, Bμ: not string?

• Called constrained MSSM = CMSSM• Gravitino mass? Minimal supergravity (mSUGRA)

Additional relations: m3/2 = m0, Bμ = Aλ – m0

Minimal Supersymmetric Extension of Standard Model (MSSM)

Current Constraints on CMSSM

WMAP constraint on relic density

Excluded because stau LSP

Excluded by b s gamma

Excluded (?) by latest g - 2

Assuming the lightest sparticleis a neutralino

JE + Olive + Santoso + Spanos

Effects in Specific Regions

• Co-annihilation:– Important when two or more sparticles nearly

degenerate– e.g., neutralino and stau

• Strip in (m1/2, m0) plane

• Rapid annihilation via direct-channel Higgs pole(s):– h important when m1/2 small, m0 large– H, A important when tan, m1/2 large

Current Constraints

on CMSSM

Impact ofHiggsconstraintreducedif larger mt, focus-pointregion far up

Differenttan βsign of μ

JE + Olive + Santoso + Spanos

Sparticles may not be very light

FullModel

samples

Detectable@ LHC

ProvideDark Matter

Dark MatterDetectable

Directly

Lightest visible sparticle →

← S

econd lightest visible sparticle

JE + Olive + Santoso + Spanos

How ‘Likely’ are Heavy Sparticles?

Fine-tuning of EW scale Fine-tuning of relic density

Larger masses require more fine-tuning: but how much is too much?

Precision Observables in Susy

mW

sin2θW

Present & possiblefuture errors

Sensitivity to m1/2 in CMSSM along WMAP linesfor different A

Can one estimate the scale of supersymmetry?

tan β = 50tan β = 10

JE + Heinemeyer + Olive + Weber + Weiglein: 2007

MoreObservables

b → sγ

tan β = 10 tan β = 50

gμ - 2

JE + Heinemeyer + Olive + Weber + Weiglein: 2007

Bs → μμ

MoreObservables

tan β = 10 tan β = 50

JE + Heinemeyer + Olive + Weber + Weiglein: 2007

Bu →

Likelihoodfor m1/2

Global Fitto all

Observables

tan β = 10 tan β = 50

JE + Heinemeyer + Olive + Weber + Weiglein: 2007

Likelihoodfor Mh

Classic Supersymmetric Signature

Missing transverse energy

carried away by dark matter particles

Erice. Sept. 2, 2003 L. Maiani: LHC Status 14

m (l l ) spectrumend-point : 109 GeVprecision~ 0.3%

m (l l j)min spectrumend-point: 552 GeVprecision ~1 %

m (l±j) spectrumend-point: 479 GeVexp. precision ~1 %

m (l l j)max spectrumthreshold: 272 GeVexp. precision ~2 %

Reconstruction of ̀Typical’Sparticle Decay Chain

Msquark = 690M÷’ = 232

Mslepton= 157M÷= 121(GeV)

ATLAS

Lq~ → q χ02

R

~l

l χ01

l

Supersymmetric Benchmark Studies

Specific

benchmark

Points along

WMAP lines

Lines in

susy space

allowed by

accelerators,

WMAP data

Sparticle

detectability

Along one

WMAP line

Calculation

of relic

density at a

benchmark

point

Battaglia, De Roeck, Gianotti, JE, Olive, Pape

Summary of LHCScapabilities … and OtherAccelerators

LHC almost

`guaranteed’

to discover

supersymmetry

if it is relevant

to the mass problem

Battaglia, De Roeck, Gianotti, JE, Olive, Pape

Non-Universal Higgs Masses (NUHM)

• Generalize CMSSM (+)

mHi2 = m0

2(1 + δi)

• Free Higgs mixing μ,

pseudoscalar mass mA

• Larger parameter space

• Constrained by vacuum

stability

Regions Allowed

in Different Scenarios for

SupersymmetryBreaking

CMSSM

Benchmarks

NUHM

Benchmarks

GDM

Benchmarks

with stau NLSP

with neutralino NLSP

De Roeck, JE, Gianotti, Moortgat, Olive + Pape

• Trapezoidal shape

for quark-dijet

combinations• Endpoints related to

squark mass

Spectra in Squark W,Z,H Hadron Decays

Butterworth + JE + Raklev: 2007

Search for Squark W Hadron Decays

• Use kT algorithm to define jets

• Cut on W mass

• W and QCD jets have different subjet splitting scales

• Corresponding to y cut

Butterworth + JE + Raklev: 2007

Signals for Squark W,Z Decays

Butterworth + JE + Raklev: 2007

qW with

subjet cuts

qW without

subjet cuts

q + leptonic W qZ with subjet cuts

• Background-subtracted qW mass combinations in benchmark scenarios

• Constrain sparticle mass spectra

Search for Hadronic W, Z Decays

Butterworth + JE + Raklev: 2007

Information on Sparticle Spectra

Reconstructed sparticle masses as functions of LSP mass in scenarios and

Butterworth + JE + Raklev: 2007

charginoChargino/neutralino

squarksquark

A Light Heavy SUSY Higgs @ CDF?

Excess seen in spectrum

Apparently also in bb

CDF unable to exclude all sensitive region

BUT: not see by D0

Is this possible within NUHM?

YES: for limited ranges of tan , m1/2, m0, and A0

JE + Heinemeyer + Olive + Weiglein

PREDICT: Mh, b s , Bs , B , g - 2

all close to experimental limits

Possible Nature of SUSY Dark Matter

• No strong or electromagnetic interactionsOtherwise would bind to matterDetectable as anomalous heavy nucleus

• Possible weakly-interacting scandidatesSneutrino

(Excluded by LEP, direct searches)Lightest neutralino χ (partner of Z, H, γ)Gravitino

(nightmare for astrophysical detection)GDM: a bonanza for the LHC!

Possible Nature of NLSP if GDM

• NLSP = next-to-lightest sparticle• Very long lifetime due to gravitational

decay, e.g.:

• Could be hours, days, weeks, months or years!

• Generic possibilities:lightest neutralino χlightest slepton, probably lighter staulighter stop

• Constrained by astrophysics/cosmology

Density belowWMAP limit

Decays do not affectBBN/CMB agreement

DifferentRegions of

SparticleParameterSpace if

Gravitino LSP

JE + Olive + Santoso + Spanos

χ NLSP

stau NLSP

Minimal Supergravity Model (mSUGRA)

Excluded by b s γ

LEP constraintsOn mh, chargino

Neutralino LSPregion

stau LSP(excluded)

Gravitino LSPregion

JE + Olive + Santoso + Spanos

More constrained than CMSSM: m3/2 = m0, Bλ = Aλ – 1

Regions Allowed

in Different Scenarios for

SupersymmetryBreaking

CMSSM

Benchmarks

NUHM

Benchmarks

GDM

Benchmarks

with stau NLSP

with neutralino NLSP

De Roeck, JE, Gianotti, Moortgat, Olive + Pape

Spectra inNUHM and GDM

BenchmarkScenarios

Typical example of

non-universal Higgs masses:

Models with stau NLSP

Models with gravitino LSP

De Roeck, JE, Gianotti, Moortgat, Olive + Pape

Properties of NUHM and GDM Models

• Relic density ~ WMAP in NUHM models

• Generally < WMAP in GDM models

Need extra source of gravitinos at high temperatures, after inflation?

• NLSP lifetime: 104s < τ < few X 106s De Roeck, JE, Gianotti, Moortgat, Olive + Pape

Final States in GDM Models with Stau NLSP

• All decay chains

end with lighter stau

• Generally via χ

• Often via heavier

sleptons

• Final states contain

2 staus, 2 τ,

often other leptons

De Roeck, JE, Gianotti, Moortgat, Olive + Pape

Triggering on GDM Events

Will be selected by many separate triggers

via combinations of μ, E energy, jets, τJE, Raklev, Øye: 2007

Efficiency for Detecting Metastable Staus

Good efficiency for reconstructing stau tracks

JE + Raklev + Oye

ATLAS Momentum resolution

Good momentum resolution

JE + Raklev + Oye

Stau Mass Determination

Good mass resolution

JE + Raklev + Oye

Reconstructing Sparticle Masses

JE + Raklev + Oye

Neutralino stau + tau SquarkR q +

Reconstructing GDM Events

JE, Raklev, Øye: 2006

Gluino → qq χ

Slepton → l χ χ2 → slepton l

Sneutrino → stau WChargino

Sparticle Mass Spectra

JE + Raklev + Oye

Numbers of Visible Sparticle Species

At different

colliders

De Roeck, JE, Gianotti, Moortgat, Olive + Pape

Slepton Trapping at LHC?

• βγ typically peaked ~ 2

• Staus with βγ < 1 leave central tracker

after next beam crossing

• Staus with βγ < ¼ trapped inside calorimeter

• Staus with βγ < ½ stopped within 10m

• Can they be dug out?

De Roeck, JE, Gianotti, Moortgat, Olive + Pape

Feng + Smith

Hamaguchi + Kuno + Nakaya + Nojiri

Extract Cores from Surrounding Rock?• Use muon system to locate impact point on

cavern wall with uncertainty < 1cm

• Fix impact angle with accuracy 10-3

• Bore into cavern wall and remove core of size 1cm × 1cm × 10m = 10-3m3 ~ 100 times/year

• Can this be done before staus decay?

Caveat radioactivity induced by collisions!

2-day technical stop ~ 1/month

• Not possible if lifetime ~104s, possible if ~106s?

Very little room for water tank in LHC caverns,only in forward directions where few staus

De Roeck, JE, Gianotti, Moortgat, Olive + Pape

Potential Measurement Accuracies

Measure stau mass to 1%

Measure m½ to 1%

via cross section, other masses?

Distinguish points ζ, η

De Roeck, JE, Gianotti, Moortgat, Olive + Pape

Gravitino Dark Matter even more interesting

than Neutralino Dark Matter!

Stop NLSP in GDM Scenario?

Not possible

within CMSSM

Diaz-Cruz, JE, Olive + Santoso: 2007

Stop NLSP possible within NUHM

Tightly-constrained

scenario with

distinctive signature

Diaz-Cruz, JE, Olive + Santoso: 2007

Summary

• Supersymmetry the most ‘expected’ surprise at the LHC

• ‘Expected’ signature missing energy

• Sparticle masses not necessarily universal

• Not the only possibility– Metastable charged particle?– Strongly interacting?

• Expect the unexpected!