Composite heavy vector triples in the ATLAS di-boson excess and at - - PowerPoint PPT Presentation

Composite heavy vector triples in the ATLAS di-boson excess and at future colliders

Gearing up for LHC 13, GGI, 18 September 2015

Andrea Thamm
 JGU Mainz

based on arXiv: 1506.08688 and 1502.01701 in collaboration with R. Torre and A. Wulzer

CMS, arXiv:1405.1994

W and Z tagged dijets

CMS, arXiv:1405.3447

W and Z semi-leptonic

CMS, arXiv:1405.3447

W and Z semi-leptonic W and Z tagged dijets

CMS, arXiv:1506.01443

HV

ATLAS, arXiv:1506.00962

W and Z tagged dijets

ATLAS, arXiv:1506.00962

Di-boson excess?

3.4σ local significance 2.5σ global significance

[ATLAS, arXiv:1506.00962]

Di-boson excess?

[Allanach, Gripaios, Sutherland: arXiv:1507.01638]

• W-fat jet: 69.4 GeV < m < 95.4 GeV
• Z-fat jet: 79.8 GeV < m < 105.8 GeV

Tagging efficiencies

[ATLAS, arXiv:1506.00962]

• efficiency of jet invariant mass cuts

nobs = 20 nobs = 15 nexp = 13.0 nexp = 10.8 nexc = 4.2 nobs = 10 nexp = 3.6 nexc = 6.4

Big statistical uncertainties:

SW Z = 7.0+3.8

−2.6

SW W = 4.2+3.2

−2.0

SZZ = 6.4+3.6

−2.4

combined fit only by ATLAS 
 lack information on the correlation of the big systematic uncertainties We extract the signal CS from a single channel and compare with the others

nexc = 7.0

[ATLAS, arXiv:1506.00962]

Excess events

nobs = 20 nexp = 13.0 BRW Z→had ≈ 0.47 (σ × BR)ATLAS BRW Z→had = 3.17 fb 3.4 events σW 0 × BRW 0→W Z = 6.5+5.1

−4.1 fb

nexc = 7.0 SW Z = 7.0+3.8

−2.6

Signal cross section

Heavy vector triples

Heavy vector triples

• among the most well motivated particles
• appear in composite Higgs models but also in weakly coupled theories
• associated to the EW gauge symmetry
• consider a 3 of SU(2)L

∼ gV cH Vµ

WL , ZL, h WL , ZL, h

Coupling to SM Vectors

LV = −1 4D[µV a

ν]D[µV ν] a + m2 V

2 V a

µ V µ a

+ i gV cHV a

µ H†⌧ a ↔

D

µ

H + g2 gV cF V a

µ Jµ a F

+ gV 2 cV V V ✏abcV a

µ V b ν D[µV ν] c + g2 V cV V HHV a µ V µ aH†H − g

2cV V W ✏abcW µ ν aV b

µV c ν

V =

• V +, V −, V 0

Vµ f ¯ f

cF V · JF → clV · Jl + cqV · Jq + c3V · J3 Jµ a

F

= X

f

f Lγµτ afL

Coupling to SM fermions

Wµ ∼ g gV × g cF

Phenomenological Lagrangian

∼ gV cH Vµ

WL , ZL, h WL , ZL, h

Coupling to SM Vectors

Vµ f ¯ f ∼ g2 gV cF

cF V · JF → clV · Jl + cqV · Jq + c3V · J3 Jµ a

F

= X

f

f Lγµτ afL

Coupling to SM fermions

LV = −1 4D[µV a

ν]D[µV ν] a + m2 V

2 V a

µ V µ a

+ i gV cHV a

µ H†⌧ a ↔

D

µ

H + g2 gV cF V a

µ Jµ a F

+ gV 2 cV V V ✏abcV a

µ V b ν D[µV ν] c + g2 V cV V HHV a µ V µ aH†H − g

2cV V W ✏abcW µ ν aV b

µV c ν

V =

• V +, V −, V 0

Phenomenological Lagrangian

LV = −1 4D[µV a

ν]D[µV ν] a + m2 V

2 V a

µ V µ a

+ i gV cHV a

µ H†⌧ a ↔

D

µ

H + g2 gV cF V a

µ Jµ a F

+ gV 2 cV V V ✏abcV a

µ V b ν D[µV ν] c + g2 V cV V HHV a µ V µ aH†H − g

2cV V W ✏abcW µ ν aV b

µV c ν

V =

• V +, V −, V 0
• Couplings among vectors
• do not contribute to V decays
• do not contribute to single production
• only effects through (usually small) VW mixing
• irrelevant for phenomenology only need (cH, cF )

Phenomenological Lagrangian

LV = −1 4D[µV a

ν]D[µV ν] a + m2 V

2 V a

µ V µ a

+ i gV cHV a

µ H†⌧ a ↔

D

µ

H + g2 gV cF V a

µ Jµ a F

+ gV 2 cV V V ✏abcV a

µ V b ν D[µV ν] c + g2 V cV V HHV a µ V µ aH†H − g

2cV V W ✏abcW µ ν aV b

µV c ν

V =

• V +, V −, V 0

typical strength of V interactions

gV gV ∼ g ∼ 1

Weakly coupled model Strongly coupled model dimensionless coefficients

ci cH ∼ cF ∼ 1 cH ∼ −g2/g2

V

and cF ∼ 1

1 < gV ≤ 4π

Phenomenological Lagrangian

1 2 3 4 5 10-13 10-12 10-11 10-10 10-9 10-8 10-7 10-6 10-5 10-4 10-3 10-2 10-1 100 s ` = MV @TeVD dLêds ` @pbD WL

+ZL HV+L

WL

+WL

• HV0L

WL

• ZL HV-L

8 TeV

CTEQ6L1 Hm2 = MW

2 L

1 2 3 4 5 10-7 10-6 10-5 10-4 10-3 10-2 10-1 100 101 102 103 104 s ` = MV @TeVD dLêds ` @pbD uid j HV+L uiuj HV0L did j HV0L diuj HV-L

8 TeV

CTEQ6L1 Hm2 = s `L

• DY and VBF production

model
 independent model
 dependent

• can compute production rates analytically!
• easily rescale to different points in parameter space

σDY = X

i,j ∈ p

ΓV → ij MV 4π2 3 dLij dˆ s

• ˆ

s=M 2

V

σV BF = X

i,j ∈ p

ΓV → WL iWL j MV 48π2 dLWL iWL j dˆ s

• ˆ

s=M 2

V

quark initial state vector boson initial state

Production rates

ΓV±→ff

0 ' 2 ΓV0→ff ' Nc[f]

✓g2cF gV ◆2 MV 96π , ΓV0→W +

L W − L

' ΓV±→W ±

L ZL

' g2

V c2 HMV

192π ⇥ 1 + O(ζ2) ⇤ ΓV0→ZLh ' ΓV±→W ±

L h

' g2

V c2 HMV

192π ⇥ 1 + O(ζ2) ⇤

500 1000 1500 2000 2500 3000 3500 4000 10-3 10-2 10-1

M0 @GeVD BRHV0 Æ 2 XL

W+W- Zh uu dd ll è nn bb tt è gV = 3 Model B

• relevant decay channels: di-lepton, di-quark, di-boson

gV cH ' gV , g2cF /gV ' g2/gV

Decay widths

similar bounds for ATLAS

• excluded for masses < 1.5 TeV

, unconstrained for larger

• di-boson most stringent
• in excluded region , not reproduced

gV GF mZ

LHC bounds

1000 2000 3000 4000 10-4 10-3 10-2 10-1 100 101 102 103 104 MV @GeVD sHpp Æ VL @pbD

theoretically excluded

CMS BgV=3

pp Æ V0 pp Æ V+ V0 Æ tt V0 Æ WW Æ jj V0 Æ WW Æ lnqq

_'

V±Æ W±Z Æ 3l±n V±Æ W±Z Æ jj V0 Æ ll V±Æ l±n V0 Æ tt V± Æ tb

Heavy vector triples in the di-boson excess

500 1000 1500 2000 2500 3000 3500 1 2 3 4 5

MV @GeVD gV

Model B

theoretically excluded

• experimental limits converted into plane

(MV , gV )

• similar exclusions at low , leptonic final state dominates
• very different for larger coupling
• weaker limits if decay to top partners open

gV

[Pappadopulo, Thamm, Torre, Wulzer, arXiv:1402.4431]

New Physics?

yellow: CMS analysis dark blue: CMS light blue: CMS black: bounds from EWPT

WZ → jj WZ → 3lν l+ν

[Greco, Liu: arXiv:1410.2883] [Chala, Juknevich, Perez, Santiago :arXiv:1411.1771]

LHC bounds

500 1000 1500 2000 2500 3000 3500 1 2 3 4 5

MV @GeVD gV

Model A 500 1000 1500 2000 2500 3000 3500 1 2 3 4 5

MV @GeVD gV

Model B

theoretically excluded

• compare with weakly coupled vectors

[Pappadopulo, Thamm, Torre, Wulzer, arXiv:1402.4431]

LHC bounds

yellow: CMS analysis dark blue: CMS light blue: CMS black: bounds from EWPT

WZ → jj WZ → 3lν l+ν

Strongly coupled model Weakly coupled model

• strongly coupled vectors have weaker bounds

• neutral and charged components contribute to the various selection regions

SW Z = L × A × [( × BR)V ± BRW Z→had✏W Z→W Z + ( × BR)V 0 BRW W →had✏W W →W Z]

• Once we fix the mass there is only one parameter gV

nobs = 20 nexp = 13.0 nexp = 7.0

[Thamm, Torre, Wulzer, arXiv:1506.08688]

SW Z = 7.0+3.8

−2.6

SW W ∈ [2.2, 10.3] SZZ ∈ [1.4, 6.6] SW W = 4.2+3.2

−2.0

SZZ = 6.4+3.6

−2.4

HVT signal cross section

Thamm, Torre, Wulzer, arXiv:1506.08688

σ x BR(W'→WZ) [pb] Resonance mass [TeV] σ → σ →

σ × BReff(WZ) [pb]

a)

σ x BR(W'→WZ) [pb] Resonance mass [GeV] σ → σ →

σ × BReff(WZ) [pb]

b)

σ x BR(G→ZZ) [pb] Resonance mass [GeV] σ → σ →

σ × BReff(ZZ) [pb]

c)

σ x BR(W'→WZ) [pb] Resonance mass [GeV] σ → σ →

σ × BReff(WZ) [pb]

d)

σ x BR(V'→HV) [pb] Resonance mass [TeV] σ → σ →

σ × BR(V H) [pb]

f)

Resonance mass [GeV] σ → σ x BR(Z'→WW) [pb] Resonance mass [TeV] σ → σ →

e)

σ × BReff(WW) [pb]

CMS Fully hadronic ATLAS leptonic Z CMS leptonic Z ATLAS leptonic W CMS leptonic W CMS HV combination

Compatibility with other searches

Conclusion I

• perfectly agrees with some channels
• could maybe even explain some small excesses
• maybe slight tension in other channels
• maybe this is exactly what we expect?

Heavy vector triples at future colliders

Composite Higgs models at future colliders

1 2 3 4 5 10-7 10-6 10-5 10-4 10-3 10-2 10-1 100 101 102 103 104 s ` = MV @TeVD dLêds ` @pbD uid j HV+L uiuj HV0L did j HV0L diuj HV-L

8 TeV

CTEQ6L1 Hm2 = s `L

2 4 6 8 10 10-9 10-8 10-7 10-6 10-5 10-4 10-3 10-2 10-1 100 101 102 103 104 s ` = MV @TeVD dLêds ` @pbD uid j HV+L uiuj HV0L did j HV0L diuj HV-L

14 TeV

CTEQ6L1 Hm2 = s `L

10 20 30 40 50 10-7 10-6 10-5 10-4 10-3 10-2 10-1 100 101 102 103 104 105 s ` = MV @TeVD dLêds ` @pbD uid j HV+L uiuj HV0L did j HV0L diuj HV-L

100 TeV

CTEQ6.6M Hm2 = s `L

assume: excluded signal is only a function of number of background events

L0 L1 L0 L1

Limit extrapolation

B(s, L, mρ) ∝ L · X

{i,j}

Z dˆ s1 ˆ s dLij dˆ s ( √ ˆ s; √s) [ˆ sˆ σij (ˆ s)]

[Thamm, Torre, Wulzer: 1502.01701]

identify relevant background process background rescales with parton luminosities

Limit extrapolation - assumptions

• limit only driven by background for a cut-and-count experiment of

events within narrow window

• shape analyses depend on background and signal kinematical distributions
• however, no large deviations expected

2 4 6 8 10-7 10-6 10-5 10-4 10-3 10-2 10-1 mρ [TeV] σ(pp → ρ)x BR [pb]

σ (pp→ρ) BR (ρ→ll) σ (pp→ρ) BR (ρ→WZ) LHC HL-LHC LHC8 LHC HL-LHC LHC8

10 20 30 40 10-7 10-6 10-5 10-4 10-3 10-2 10-1 mρ [TeV] σ(pp → ρ)x BR [pb]

σ (pp→ρ) BR (ρ→ll) σ (pp→ρ) BR (ρ→WZ) FCC-1 ab-1 FCC-10 ab-1 FCC-1 ab-1 FCC-10 ab-1

Limit extrapolation

current 8 TeV LHC limits and extrapolated bounds CMS search for

• opposite sign di-leptons
• fully leptonic WZ

[CMS-PAS-EXO-12-061] [ATLAS 1405.4123] [CMS 1407.3476] [ATLAS 1406.4456]

• constant at large masses


(zero background events)

• too conservative bounds at low

masses

Composite Higgs Model

• predicts direct and indirect effects
• modification of Higgs couplings


(predictable in a fairly model- independent way)

• EWPT 


(sensitive to effects only computable in specific models)

• Flavour

a = gW W h = √1 − ξ

• production of EW vector resonances


(here consider 3 of )

• production of top partners 


(mass controls generation of Higgs potential and fine-tuning,
 very model dependent)

SU(2)L

[Pappadopulo, Thamm, Torre, Wulzer: 1402.4431] [Matsedonskyi, Panico, Wulzer: 1409.0100]

• parameter space:

mρ gρ ξ = g2

ρ

m2

ρ

v2

• for illustration focus on minimal composite Higgs model

Minimal Composite Higgs

assume global symmetry: SO(5)/SO(4)

SO(5) × U(1)X SO(4) × U(1)X SU(2)L × U(1)Y U(1)EM

breaking scale f > v

a = p 1 − ξ b = 1 − 2ξ b3 = −4 3ξ p 1 − ξ d(4)

3

= p 1 − ξ

where

ξ = v2 f 2

[Contino, Nomura, Pomarol: hep-ph/0306259] [Agashe, Contino, Pomarol: hep-ph/0412089] [Agashe, Contino: hep-ph/0510164 ] [Contino, Da Rold, Pomarol: hep-ph/0612048] [Barbieri, Bellazzini, Rychkov, Varagnolo: hep-ph/ 0706.0432]

Higgs emerges as a pseudo-NG boson

L = 1 2 (∂µh)2 − V (h) + v2 4 Tr

• DµΣ†DµΣ

✓ 1 + 2a h v + b h2 v2 + b3 h3 v3 + . . . ◆ V (h) = 1 2m2

hh2 + d3

✓m2

h

2v ◆ h3 + d4 ✓ m2

h

8v2 ◆ h4 + . . .

Higgs couplings receive corrections of order ξ

Indirect measurements

= √1 − ξ kF = √1 − ξ kF = 1 − 2ξ √1 − ξ

MCHM4: MCHM5: expected LHC reach:

ξ = 0.1

Indirect measurements

[CMS-NOTE-2012-006] [ATL-PHYS-PUB-2013-014] [Dawson et. al.1310.8361] [CLIC 1307.5288]

2 4 6 8 10 2 4 6 8 10 12 mρ [TeV] gρ

ξ=1 L H C H L

• L

H C I L C T L E P / C L I C L H C 8 L H C H L

• L

H C

(mρ, gρ)

• theoretically excluded
• LHC8 at 8 TeV with 20


LHC at 14 TeV with 300
 HL-LHC at 14 TeV with 3

• di-leptons more sensitive for small
• di-boson more sensitive for large
• increase in : improves mass reach
• increase in L: improves reach
• resonances too broad for large

ξ ≤ 1 fb−1 fb−1 ab−1 gρ gρ √s gρ gρ

Results in

95% C.L.

[Thamm, Torre, Wulzer: 1502.01701]

(mρ, gρ)

• theoretically excluded
• LHC8 at 8 TeV with 20


HL-LHC at 14 TeV with 3

ξ ≤ 1 fb−1 ab−1

10 20 30 40 2 4 6 8 10 12 mρ [TeV] gρ

ξ=1 LHC HL-LHC HL-LHC FCC-1ab-1 FCC-10ab-1 I L C TLEP / CLIC

• direct: more effective for small 


ineffective for large

• indirect: more effective for large

gρ gρ gρ

Results in

95% C.L.

[Thamm, Torre, Wulzer: 1502.01701]

Results in

• theoretically excluded
• LHC8 at 8 TeV with 20


LHC at 14 TeV with 300
 HL-LHC at 14 TeV with 3

fb−1 ab−1

2 4 6 8 10 10-3 10-2 10-1 100 mρ [TeV] ξ

TLEP / CLIC LHC L H C 8 L H C H L

• L

H C HL-LHC ILC gρ=4π gρ = 1

1 ≤ gρ ≤ 4π

95% C.L.

fb−1

[Thamm, Torre, Wulzer: 1502.01701]

(mρ, ξ)

Results in

• theoretically excluded
• LHC8 at 8 TeV with 20


HL-LHC at 14 TeV with 3

fb−1 ab−1 1 ≤ gρ ≤ 4π

95% C.L.

10 20 30 40 10-4 10-3 10-2 10-1 100 mρ [TeV] ξ

TLEP / CLIC LHC HL-LHC F C C

• 1

a b

• 1

F C C

• 1

a b

• 1

HL-LHC ILC gρ=4π gρ=1

[Thamm, Torre, Wulzer: 1502.01701]

(mρ, ξ)

Conclusions

• CH is a very compelling framework
• many ways to look for it:


direct: vector resonance and top partners
 indirect: coupling modifications

• excess: maybe exactly what a resonance at the edge of discovery should

look like?

• learn a lot from LHC RunII
• … and if not, then at a future collider!

Backup

Limit extrapolation

Input: experimental bounds on at with for various search channels

σ×BR √s0 = 8 TeV L0 ' 20 fb−1

• extrapolate limits to different proton-proton collider at and L

√s

• driven by number of background events in a small invariant mass

window around the resonance peak

B(s, L, mρ) = B(s0, L0, m0

ρ)

• utput

same limit on number of signal events

• excluded cross section at the equivalent mass

[σ×BR](s, L; mρ) = L0 L · [σ×BR](s0, L0; m0

ρ)

∆ˆ s m2

ρ

= 10%

Limit extrapolation - equivalent mass

• extraction of equivalent mass

B(s, L, mρ) = B(s0, L0, m0

ρ)

• number of background events within window

B(s, L, mρ) ∝ L · X

{i,j}

Z dˆ s1 ˆ s dLij dˆ s ( √ ˆ s; √s) [ˆ sˆ σij (ˆ s)]

ˆ s ∈ [m2

ρ − ∆ˆ

s/2, m2

ρ + ∆ˆ

s/2]

partonic cross-section contributing to background

• partonic cross section: SM process much above SM masses

[ˆ sˆ σij (ˆ s)] ' cij

constant

• parton luminosities constant within small integration limit

B(s, L, mρ) ∝ ∆ˆ s m2

ρ

· L · X

{i,j}

cij dLij dˆ s (mρ; √s) X

{i,j}

cij dLij dˆ s (mρ; √s) = L0 L X

{i,j}

cij dLij dˆ s (m0

ρ; √s0)

• equating backgrounds

[Thamm, Torre, Wulzer: 1502.01701]

Limit extrapolation - equivalent mass

X

{i,j}

cij dLij dˆ s (mρ; √s) = L0 L X

{i,j}

cij dLij dˆ s (m0

ρ; √s0)

• need relevant background process and parton luminosities
• sum drops for single partonic initial state
• otherwise linear combination of parton luminosities weighted by cij

Limit extrapolation - equivalent mass

• Subtlety at low masses:


lowest mass point of 8 TeV limit determined by sensitivity of specific analysis

• arbitrary lowest equivalent mass depending on luminosity
• smoothly raise luminosity of future collider
• extrapolated limit is the strongest at each mass
• low-mass limit conservative, not optimal

2 4 6 8 10-6 10-5 10-4 10-3 mρ [TeV] σ(pp → ρ)x BR [pb]

cut and count analysis extrapolation from 8 TeV LHC8 LHC HL-LHC

10 20 30 40 10-6 10-5 10-4 10-3 mρ [TeV] σ(pp → ρ)x BR [pb]

cut and count analysis extrapolation from 8 TeV FCC-1 ab-1 FCC-10 ab-1

EWPT

• set some of strongest constraints on CH models
• incalculable UV contributions can relax constraints

∆ ˆ S = g2 96π2 ξ log ✓ Λ mh ◆ + m2

W

m2

ρ

+ α g2 16π2 ξ , ∆ ˆ T = − 3g0 2 32π2 ξ log ✓ Λ mh ◆ + β 3y2

t

16π2 ξ

IR contribution due to Higgs coupling modifications tree level exchange of vector resonances

short distance effects

• and constants of order 1
• define and marginalise

α β χ2(ξ, mρ, α, β)

[Grojean, Matsedonskyi, Panico: 1306.4655]

EWPT

EWPT

• define and marginalise
• to avoid unnatural cancellations

χ2(ξ, mρ, α, β)

2 4 6 8 10 10-4 10-3 10-2 10-1 100 mρ ξ

Present ILC TLEP α = β = 0 δχ2 < 5

10 20 30 40 10-4 10-3 10-2 10-1 100 mρ ξ

Present ILC TLEP α = β = 0 δχ2 < 5

δχ2 = χ2(ξ, mρ, α = 0, β = 0) χ2(ξ, mρ, α, β)

[Baak et al: 1209.2716] [Baak et al: 1407.3792]