12mm Exotic Higgs Decays at CMS 8mm · Introduction (1/2) I Higgs-like particle with mass ˘125 GeV was observed at LHC I Critical question: is it the SM Higgs boson? I (1) Precise

Exotic Higgs Decays at CMS

Aysen Tatarinov for CMS Collaboration

Texas A&M University

Dark Interactions 2014Brookhaven National Laboratory

June 11–13, 2014

Aysen Tatarinov (Texas A&M University) Exotic Higgs Decays at CMS Dark 2014 1 / 34

Introduction (1/2)

I Higgs-like particle with mass ∼125 GeV was observed at LHC

I Critical question: is it the SM Higgs boson?

I (1) Precise measurementsof its branching ratios:

I This may take many yearsI Current 95% CL limit:BBSM ≤ 0.52

I (2) Direct searches for non-SMdecays of SM-like Higgs:

I In case of observation: this isnon-SM Higgs!

I In case of no signal: restrictbroad class of scenariosbeyond the SM

BSMBR0 0.2 0.4 0.6 0.8 1

ln L

∆-

2

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

5.0Observed

Exp. for SM H

CMS Preliminary -1 19.6 fb≤ = 8 TeV, L s -1 5.1 fb≤ = 7 TeV, L s

BSM, BRgκ, γκ

CMS-PAS-HIG-13-005


https://cds.cern.ch/record/1542387

Introduction (2/2)

I Searches for invisible decays of the Higgs boson:I Vector Boson Fusion productionI ZH production with the Z boson decaying to a pair of charged leptonsI ZH production with the Z boson decaying to bottom quarksI Interpretation in terms of Higgs-portal models of DM interactions

I Search for non-SM Higgs decays to a pair of new light bosons, eachof which decays to boosted muon pairs: h→ 2a→ 4µ

I Interpretation in the context of SUSY with hidden dark sector


Invisible Higgs Decays in Vector Boson Fusion Channel

I Experimental signature:I Two jets with high transverse momentum, pT , well separated in

pseudorapidityI Large missing transverse evergy, Emiss

T

I CMS-PAS-HIG-13-013, arXiv:1404.1344v2


http://cds.cern.ch/record/1596283

http://arxiv.org/abs/arXiv:1404.1344v2

Vector Boson Fusion Channel: Results (1/2)

I EmissT and Mjj in data and simulation after the full selection:

I Signal contribution for mH = 125 GeV and B(H → inv) = 100%I Background estimates and signal predictions shown cumulatively

[GeV]missTE

150 200 250 300 350 400 450 500

Eve

nts

/ 20

GeV

-110

1

10

210

310

410 CMS

-1 = 8 TeV, L = 19.5 fbs

VBF H(inv)

Observed

inv) = 100%→B(H = 125 GeV,HVBF m

V+jets

, tW, DY(ll)+jets, VVtt

[GeV]jjM1500 2000 2500 3000 3500

Eve

nts

/ 100

GeV

-210

-110

1

10

210

310

410 CMS

-1 = 8 TeV, L = 19.5 fbs

VBF H(inv)

Observed

inv) = 100%→B(H = 125 GeV,HVBF m

V+jets

, tW, DY(ll)+jets, VVtt


Vector Boson Fusion Channel: Results (2/2)

I No significant excess in data over total background

I Set upper limit on invisible Higgs decays in VBF channel

I Assuming SM VBF production cross section, 95% CL observed(expected) upper limit on B(h→ inv) for mH = 125 GeV is 0.65 (0.49)

[GeV]Hm100 150 200 250 300 350 400

inv)

[pb]

→ x

B(H

σ

0

0.5

1

1.5

2

2.5 invisible→CMS VBF H

-1 = 8 TeV, L = 19.5 fbs95% CL limits

Observed limitExpected limit

)σExpected limit (1)σExpected limit (2

(SM)VBFσ

[GeV]Hm100 150 200 250 300 350 400

(SM

)V

BF

σ in

v)/

→ x

B(H

σ

0.5

1

1.5

2

2.5 invisible→CMS VBF H

-1 = 8 TeV, L = 19.5 fbs95% CL limits


)σExpected limit (1



Search for Invisible Higgs Decays in Z(ll)H Channel

I Experimental signature:

I Two leptons (e+e− or µ+µ−) with invariant mass consistent with Zmass

I Large missing transverse evergy, EmissT



http://cdsweb.cern.ch/record/1561758


Results (1/2)I mT and ∆φ(ll) in data and simulation after the full selection:

I Signal contribution for mH = 125 GeV and B(H → inv) = 100%I Background estimates shown cumulativelyI Signal predictions shown separatelyI For limit setting at 7 TeV 1D distribution of mT is usedI For limit setting at 8 TeV 2D distribution of mT and ∆φ(ll) is usedI Limit on invisible Higgs decays in Z(ll)H channel to be shown later in

combination with limit Z(bb̄)H channel

[GeV]Tm200 300 400 500 600 700 800 900 1000

Eve

nts

0

2

4

6

8

10

12

14

16

18 CMS

-1 = 7 TeV, L = 4.9 fbs

Z(ll) H(inv)

Observed

inv)=100%→B(H

=125GeV),H

ZH(m

ZZ

WZ

DY(ll)+jets

,tW,WW,W+jetstt

[GeV]Tm200 300 400 500 600 700 800 900 1000

Eve

nts

0

10

20

30

40

50

60CMS

-1 = 8 TeV, L = 19.7 fbs

Z(ll) H(inv)

Observed

inv)=100%→B(H

=125GeV),H

ZH(m

ZZ

WZ

DY(ll)+jets

,tW,WW,W+jetstt

(l,l) [radians]φ∆0 0.5 1 1.5 2 2.5 3

Eve

nts

0

20

40

60

80

100CMS

-1 = 8 TeV, L = 19.7 fbs

Z(ll) H(inv)

Observed

inv)=100%→B(H

=125GeV),H

ZH(m

ZZ

WZ

DY(ll)+jets

,tW,WW,W+jetstt

7 TeV 8 TeV 8 TeVAysen Tatarinov (Texas A&M University) Exotic Higgs Decays at CMS Dark 2014 8 / 34

Search for Invisible Higgs Decays in Z(bb̄)H Channel

I Experimental signature:

I Two jets consistent with Z → bb̄I Large missing transverse evergy, Emiss

T





Z(bb̄)H Channel: Results (1/2)I Multivariate Analysis: Boosted Decision Trees (BDT)I BDT output in high-, intermediate-, low- dijet pT regions in data and

simulations after the full selectionI Signal contribution for mH = 125 GeV and B(H → inv) = 100%I Background estimates shown cumulativelyI Signal predictions shown separately

Eve

nts

/ 0.1

-210

-110

1

10

210

310

410

510CMS

-1 = 8 TeV, L = 18.9 fbs

T) H(inv) high pbZ(b

Observed

inv)=100%→B(H =125GeV,

HZH m

Z+jets

W+jets

, tWtt

QCD

VV

)bVH(b

BDT output-1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1

MC

Dat

a

0

1

2

Eve

nts

/ 0.1

-210

-110

1

10

210

310

410

510 CMS-1 = 8 TeV, L = 18.9 fbs

T) H(inv) intermediate pbZ(b

Observed

inv)=100%→B(H =125GeV,

HZH m

Z+jets

W+jets

, tWtt

QCD

VV

)bVH(b

BDT output-1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1

MC

Dat

a

0

1

2

Eve

nts

/ 0.1

-210

-110

1

10

210

310

410

510CMS

-1 = 8 TeV, L = 18.9 fbs

T) H(inv) low pbZ(b

Observed

inv)=100%→B(H =125GeV,

HZH m

Z+jets

W+jets

, tWtt

QCD

VV

)bVH(b

BDT output-1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1

MC

Dat

a0

1

2


ZH Channel: Results (2/2)

I Combine limits on invisible Higgs decays in Z(bb̄)H channel and Z(ll)Hchannel

I Assuming SM ZH production cross section, 95% CL observed(expected) upper limit on B(h→ inv) for mH = 125 GeV is 0.81 (0.83)

[GeV]Hm105 110 115 120 125 130 135 140 145

inv)

[pb]

→ x

B(H

σ

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

1.8CMS

)HbCombination of Z(b invisible→and Z(ll)H, H

= 8 TeV (Both ZH channels)s-1L = 18.9-19.7 fb

= 7 TeV (Z(ll)H only)s-1L = 4.9 fb

95% CL limitsObserved limitExpected limit

)σExpected limit (1)σExpected limit (2

(SM)ZHσ

[GeV]Hm105 110 115 120 125 130 135 140 145

(SM

)Z

Hσ

inv)

/→

x B

(Hσ

0.5

1

1.5

2

2.5

3CMS

)HbCombination of Z(b invisible→and Z(ll)H, H

= 8 TeV (Both ZH channels)s-1L = 18.9-19.7 fb


95% CL limits





Combination of VBF, Z(ll)H, Z(bb̄)H channelsI Combine limits on invisible Higgs decays of all three searches

I Assuming SM ZH production cross section, 95% CL observed(expected) upper limit on B(h→ inv) for mH = 125 GeV is 0.58 (0.44)

I The best direct measurement of B(h→ inv) to date

[GeV]Hm115 120 125 130 135 140 145

SM

σ in

v)/

→ x

B(H

σ

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

1.8

2CMS

Combination of VBF and invisible→ZH, H

= 8 TeV (VBF + ZH)s-1L = 18.9-19.7 fb


95% CL limits





Interpretation (1/2)

I Interpret the limit on B(h→ inv), assuming SM production cross section, inthe context of a Higgs-portal model of dark matter interactions

I If Mχ < mH/2, the width of invisible Higgs decays, Γinv relates tospin-independent DM-nucleon elastic cross section as follows for scalar(S),vector (V), fermionic (F) DM:


Interpretation (2/2)

I Upper limits on the DM-nucleon cross section for mH = 125 GeV andB(h→ inv) < 0.51 at 90%CL as a function of the DM mass

I Severeley restrict the DM-nucleon cross section for light DM

[GeV]χMDM Mass 10 210 310

[pb]

SI -N

χσ

DM

-nuc

leon

cro

ss s

ectio

n

-1310

-1210

-1110

-1010

-910

-810

-710

-610

-510

-410

-310

-210

-110

σCRESST 1σCRESST 2

XENON100(2012)XENON10(2011)DAMA/LIBRACoGeNT(2013)/90%CLCoGeNT(2013)/99%CLCDMS(2013)/95%CLCOUPP(2012)LUX(90%CL)

CMS invisible→ZH, H Combination of VBF and

= 125 GeVHm inv) < 0.51 @ 90% CL→B(H

(VBF+ZH)-1 = 8.0 TeV, L = 18.9-19.7 fbs (ZH)-1 = 7.0 TeV, L = 4.9 fbs

vector

scalar

fermion

Min

Lattice

Max


Search for muon (leptin) jets

I Search for non-SM Higgs decays to a pair of new light bosons, eachof which decays to boosted and isolated muon pairs (dimuons):

h→ 2a→ 4µ (CMS-PAS-HIG-13-010)I ma within the range 0.25–3.55 GeV (roughly between 2mµ and 2mτ )

p

p

ha

a

µµ

µµ

+

_+

_

I Analysis is designed to remain model independentI Allows easy reinterpretation in the context of any scenario

with the same signature

I Example interpretation: SUSY + hidden sector



SUSY + hidden sector

I Recent observation of rising positron fraction at high energies by satellite experiments

I Dark matter annihilation: new light γD as an attractive long-distance force between slowmoving WIMPs

I Simplified implementation of dark sector (for simulation only):I dark neutralino nD (new LSP) + dark photon γD

I if mn1 <mh2

: h → 2n1

I n1 decays into dark sector particles: n1 → nDγDI γD weakly couples to SM via kinetic mixing with photon

p

p

hγ

γ

µµ

µµ D

D

+

_+

_

nD

nD

n1

n1

p

p

hγ

γ

µµ

µµ D

D

+

_+

_

nD

nD

n1

n1

Phys. Rev. Lett. 110, 141102 (2013)


Analysis Selection

p

p

ha

a

µµ

µµ

+

_+

_

I At least four muons: pT > 8 GeV/c, |η| < 2.4, good track quality

I At least one good quality muon with pT > 17 GeV/c, |η| < 0.9

I Assign two opposite-sign muons to a dimuon

I mµµ < 5 GeV/c2 and (good common vertex or ∆Rµµ < 0.01)

I Further consider events with exactly two dimuons

I Apply isolation requirement to dimuons: supresses backgroundby a factor of 50, reject about 20% of signal


Signal Region

I Target events where dimuons are produced in decaysof new light bosons with the same mass

I Signal region: reconstructed dimuon masses consistent with each other:I |m1 −m2| ≤ 5 · σ(m1+m2

2 ) (where σ(m) — dimuon mass resolution)

I Study of dimuon mass resolution: σ(m) ∼ 0.026 + 0.013 ·mI Use narrow SM resonances in data: ω, φ, J/ψ, ψ′


SM Background (1/2): bb̄

I 2D background template obtained from bb̄ enriched data: events with onedimuon and one muon (no isolation requirement)

I Off-diagonal part of bb̄ 2D shape is normalized to 8 events observed inoff-diagonal region in data

I 1.8± 0.6 of bb̄ events expected in the signal region

)2 c x

0.0

25 G

eV/

2 cE

vent

s / (

0.02

5 G

eV/

-510

-410

-310

-210

)2 c x

0.0

25 G

eV/

2 cE

vent

s / (

0.02

5 G

eV/

-510

-410

-310

-210

0.5 1 1.5 2 2.5 3 3.5

0.5

1

1.5

2

2.5

3

3.5

]2c [GeV/1

µµm0.5 1 1.5 2 2.5 3 3.5

]2 c [G

eV/

2µµm

0.5

1

1.5

2

2.5

3

3.5

-1 = 20.65 fbint = 8 TeV LsCMS Prelim. 2012


Background Estimation (2/2): Prompt Double J/ψ

I Prompt double J/ψ productionI 2D Crystal Ball template normalized to dataI 2.0± 2.0 prompt double J/ψ events expected in the signal region

)2 c x

0.0

25 G

eV/

2 cE

vent

s / (

0.02

5 G

eV/

-310

-210

-110

)2 c x

0.0

25 G

eV/

2 cE

vent

s / (

0.02

5 G

eV/

-310

-210

-110

]2c [GeV/1

µµm0.5 1 1.5 2 2.5 3 3.5

]2 c [G

eV/

2µµm

0.5

1

1.5

2

2.5

3

3.5

0.5 1 1.5 2 2.5 3 3.5

0.5

1

1.5

2

2.5

3

3.5



Signal Region Yields

I Unblind the signal region(diagonal region)

I One event is observed in thesignal region

I 3.8± 2.1 background eventsexpected in the signal region(1.8 ± 0.6 bb̄,2.0 ± 2.0 double J/ψ)

)2 c x

0.0

25 G

eV/

2 cE

vent

s / (

0.02

5 G

eV/

-510

-410

-310

-210

-110

)2 c x

0.0

25 G

eV/

2 cE

vent

s / (

0.02

5 G

eV/

-510

-410

-310

-210

-110

]2c [GeV/1

µµm0.5 1 1.5 2 2.5 3 3.5

]2 c [G

eV/

2µµm

0.5

1

1.5

2

2.5

3

3.5

]2c [GeV/1

µµm0.5 1 1.5 2 2.5 3 3.5

]2 c [G

eV/

2µµm

0.5

1

1.5

2

2.5

3

3.5



Model Independent Limit

95% CL limit on σ(pp → h→ 2a)× B2(a→ 2µ)× αgen

I αgen — kinematic andgeometric acceptance ongenerator level

I We use flat εfullαgen

= 0.63 ±0.05observed in all MC samples weused

I Applicable to models with 4µcoming from new light bosonswith mass in range 0.25–3.55GeV, where new light bosonstypically isolated and spatiallyseparated

]2c [GeV/1am

0 0.5 1 1.5 2 2.5 3 3.5 4

[fb]

gen

α) µ

2

→(a

2

2a)

B→

(pp

σ 0

0.1

0.2

0.3

0.4

0.5

Model independent 95% CL upper limit



Results: SUSY with Hidden Dark Sector Scenario

I 95% CL limit on σ(pp → h→ 2a)B2(a→ 2µ) vs mh

]2c [GeV/hm90 100 110 120 130 140 150

) [fb

]µ

2

→ Dγ(2

) B

D+

2nDγ

2→

h

→(p

p σ

0

1

2

3

4

5

6

7

8

9

10Dark SUSY 95% CL Limit:

2c = 1 GeV/Dn

, m2c = 10 GeV/1nm

2c = 0.4 GeV/D

γand m

),h

(mSM

σ h)=→(pp σPrediction with

) = 0.25%,1 2n→B(h

) = 50%,D+ nD

γ → 1

B(n

) = 45%µ 2→ D

γand B(



Conclusions

Searches for invisible Higgs decays in VFB and ZH channels with 7 and 8 TeVdata collected at CMS experiment

I The best direct measurement of B(h→ inv) to date

I Severe restriction on DM-nucleon cross section for low mass DM(Mχ < mH/2) in the context of Higgs-portal model of dark matterinteractions

Search for non-SM Higgs decays to a pair of new light bosons, which decay toboosted and isolated muon pairs: h→ 2a→ 4µ (2mµ . ma . 2mτ )

I Model independent limit is set:

I Can be applied to any model with the same signatureI Interpreted in the context of SUSY with hidden dark sector


BACKUP SLIDES


Search for Invisible Higgs Decays in VBF Channel

Event selection:

I Two jets with pj1T , pj2T > 50 GeV

I Forward/backward: ηj1 · ηj2 < 0

I Well separated in pseudorapidity: ∆ηjj > 4.2

I High invariant mass: Mjj > 1100 GeV

I Missing transverse energy: EmissT > 130 GeV

I Veto events with e, µ with pT > 10 GeV —supress Z and W backgrounds

I ∆φjj < 1.0 — supress multijet backgrounds

I Central jet veto (CJV): veto events withadditional jet with pT > 30 GeV andηj1 < η < ηj2


VBF Channel: Background Estimation

I Main backgrounds: Z(νν)+jets and W(lν)+jets (not identified lepton)

I Data-driven estimation with Z and W decays to well identified leptons

I Z(νν)+jets control region: require oppositely charged pair of muons,veto additional leptons, relax some selections

I Number of Z(νν) events in the signal region:

I Ncµµobs — observed yield in the control region

I Ncbkg — background estimation from tt̄, single-top, diboson MC

I W(eν)+jets and W(µν)+jets control region: require electron or muon,veto additional leptons

I W(τν)+jets control region: require hadronic tau, veto additionalleptons, do not apply CJV

I QCD multijet backgrounds — data-driven (ABCD method)

I Minor backgrounds (tt̄, signle-top, diboson, DY(ll)) — MC simulation


Vector Boson Fusion Channel: Results


Search for Invisible Higgs Decays in Z(ll)H ChannelEvent selection:

I e+e− or µ+µ−, each lepton with pT > 20 GeV,invariant mass mll within ±15 GeV from Zboson mass

I Reject events with two or more jets withpT > 30 GeV — reduce DY(ll)+jetsbackground

I Veto events with additional e, µ with pT > 10GeV — supress Z and W backgrounds

I Reject events identified to have b quark —supress top-quark background

I Final requirements optimized for best expectedexclusion limit for mH = 125 GeV:

I EmissT > 130 GeV, ∆φ(ll ,Emiss

T ) > 2.7,|Emiss

T − pllT |/pllT < 0.25


Z(ll)H Channel: Background estimation

I Main backgrounds: WZ and ZZ — estimate using MC simulation

I DY(ll)+jets — use orthogonal control sample of γ+jets eventsI Minor backgrounds — tt̄, Wt, WW, W+jets

I Control sample with opposite-charge different-flavor dileptons (e±µ∓)I Multiply number of events in the control region, Neµ, by scale factorsαee and αµµ to estimate backgrounds in e+e− and µ+µ− final states

I Measure αee and αµµ in the sidebands (SB) of the Z peak(40 < mll < 70 GeV and 110 < mll < 200 GeV):

I NSBee , NSB

µµ , NSBeµ — number of events in e+e−, µ+µ−, e±µ∓ final

states in top-quark enriched sample


Z(ll)H Channel: Results


Search for Invisible Higgs Decays in Z(bb)H Channel

I Events with two jets consistent with Z → bb̄,large Emiss

T due to invisible Higgs decay

I Summary of selection criteria:


Z(bb)H Channel: Background Estimation & BDT Training

I All backgrounds are estimated from MC simulation

I Train BDT using simulated samples for signal and backgrounds after allselections

I Set of BDT input variables is the subset of the following variables chosen initerative optimization:


Z(bb̄)H Channel: Results


Documents

12mm Exotic Higgs Decays at CMS 8mm · Introduction (1/2) I Higgs-like particle with mass ˘125 GeV was observed at LHC I Critical question: is it the SM Higgs boson? I (1) Precise