higgs portal at future colliders · higgs portal dark matter model p. ko, j. li. arxiv:1610.03997 l...

Characterizing DM properties with (non)minimalHiggs portal at future colliders

Pyungwon Ko

Korea Institute for Advanced Study

Based on B. Dutta, T. Kamon, P. Ko, Jinmian Li, arXiv: 1705.02149, 1712.05123

Aug. 28, 2018

COSMO 2018

IBS, Daejon, Korea

Pyungwon Ko (KIAS) Higgs portal DMs at colliders ICHEP2018 1 / 24

Outline

1 Generalities

2 Status of LHC searches for Higgs portal DM models

3 Prospects of searches at future colliders500 GeV ILC100 TeV FCC-pp/SppC

4 Conclusion


Crossing Relations and WIMP Detections

Crossing sym is an exact sym of a single Lagrangian in QFT

It should be applied to a model Lagrangian which is valid in allthe relevant kinematic region

One can not apply it to EFT or Simplified Models


Effective scalar operator

Limits after LHC run-1, red: mono-jet; green:one-lepton; yellow: two-leptons; blue: multi-jets.

σEFT/FT(pp→ χχ)@LHC

Region I: EFT limit is valid.

Region II: Resonant enhancementin FT.

Region III: σEFT ∝ 1m2med

VS

phase space factor in FT


Scalar mediated simplified model

J. Brooke, M. Buckley, et. al. arXiv:1603.07739

LH 3 −gχHχ̄χ−∑f

gvyf√2Hf̄f

LA 3 −igχAχ̄γ5χ−∑f

igvyf√2Af̄γ5f

Projected CMS invisible Higgs search in VBF

Constraint on scalar is weaker thanpseudoscalar, due to smaller cross section and

a softer EmissT .

Requires sizeable coupling for detection(gv = gχ = g), may not realistic, e.g.Γ > mH/A.

SM gauge symmetry is not conserved if H/Ais an singlet.

φ · f̄f = φ · (f̄LfR + f̄RfL)


Higgs portal dark matter model

P. Ko, J. Li. arXiv:1610.03997

LFDM 3 −gχSχ̄χ− λHSH†HS2 − µ30S

− µ1SH†H − µ2

3!S3 − λS

4!S4

LintFDM 3 −(H1 cosα+H2 sinα)

∑f

mf

vhf̄f

+ gχ(H1 sinα−H2 cosα) χ̄χ

gχ = 3, sinα = 0.3, mχ = 80 GeV(Throughout the talk. )

Production limited by SM Higgs-fermioncouplings.

Recast of the CMS invisible Higgs search inMono-jet.

Interference between two mediators (reducethe sensitivity at low MH2

).

All points way below the current LHCsensitivity.


Simplified vs. UV complete modelsS. Baek, P. Ko, M.H. Park, W.I. Park, C. Yu, arXiv:1506.06556


Outline

1 Generalities



4 Conclusion


Minimal gauge invariant models: FDM, VDM and SDM

Lagrangians: Ko et al, arXiv:1112.1847 (SFDM), 1212.2131 (VDM)

LFDM = LSM + χ̄(i/∂ −mχ − gχS)χ + 12∂µS∂

µS − 12m2

0S2 − λHSH†HS2

−µ30S − µ1SH

†H − µ23!S3 − λS

4!S4

LVDM = − 14VµνV

µν +DµΦ†DµΦ− λΦ(Φ†Φ−v2φ2

)2 − λHΦ(H†H −v2h2

)(Φ†Φ−v2φ2

)

LSDM = 12∂µS∂

µS − 12m2

0S2 − λHSH†HS2 − λS

4!S4

Interaction Lagrangians:

LintFDM = −(H1 cosα +H2 sinα)

[∑f

mfvh

f̄f −2m2

Wvh

W+µ W

−µ −m2Zvh

ZµZµ

]+gχ(H1 sinα−H2 cosα) χ̄χ

LintVDM = −(H1 cosα +H2 sinα)

[∑f

mfvh

f̄f −2m2

Wvh

W+µ W

−µ −m2Zvh

ZµZµ

]− 1

2gVmV (H1 sinα−H2 cosα) VµV

µ

LintSDM = h [

∑f

mfvh

f̄f −2m2

Wvh

W+µ W

−µ −m2Zvh

ZµZµ] + λHSvh hS

2


Higgs portal DM within EFT

The models for the fermion and vector DM are not good, since

they are not renormalizablenot gauge invariant for vector DM casethus they violate unitarity

One cannot apply crossing relations for fermion or veector DM

Need to work in (the simplest possible) UV completions

In fact they are not even well-defined EFTs


General feature

HiDM

,t ≡ m2

DD

dσSDM

dt∝ σh

∗SDM × |

1

t−m2h

+ imhΓh|2,

dσFDM

dt∝ σh

∗FDM × |

1

t−m2H1

+ imH1ΓH1

−1

t−m2H2

+ imH2ΓH2

|2 ·(2t− 8m

2χ

),

dσVDM

dt∝ σh

∗VDM × |

1

t−m2H1

+ imH1ΓH1

−1

t−m2H2

+ imH2ΓH2

|2 ·(

2 +(t− 2m2

D)2

4m4V

).

[GeV]DDm0 50 100 150 200 250 300 350 400 450 500

Eve

nt F

ract

ion

0

0.02

0.04

0.06

0.08

0.1

0.12

0.14

0.16

0.18

0.2

0.22

0.24

SM

SDM

VDMs

FDM200

FDM300

FDM400

FDM500

=3χParton, g

200 300 400 500 600 700 800

DDm

0.01

0.1

1

Eve

nt F

ract

ion

=3χ

Parton, g

SDM

FDM 200 300 500

VDM 200 300 500

=3χ

Parton, g

200 300 400 500 600 700 800 [GeV]DDm

1−10

1

10

(V/F

) ra

tio


DM production at the ILC

The dominant DM production process:

e+e− → Z(→ ff) H(∗)1,2 (→ DD)

DM pair four-momentum:PµDD = Pµ

e++ Pµ

e− − PµZ = (

√s− EZ ,−~pZ)

DM pair invariant mass:m2DD = s+m2

Z − 2EZ√s


Features of DM production at the ILC

[GeV]DDm0 50 100 150 200 250 300 350 400 450 500

Eve

nt F

ract

ion

0

0.02

0.04

0.06

0.08

0.1

0.12

0.14

0.16

0.18

0.2

0.22

0.24

SM

SDM

VDMs

FDM200

FDM300

FDM400

FDM500

=3χParton, g

[GeV]DDm0 50 100 150 200 250 300 350 400 450 500

Eve

nt F

ract

ion

0

0.01

0.02

0.03

0.04

0.05

0.06

0.07

0.08

SM

SDM

VDMs

FDM200

FDM300

FDM400

FDM500

=3χ

Delphes, g

Dominant background processes:

e−

e+

νe

ν̄e

Zj

j

e−

e+

Z

Z

ν̄e

νe

j

j

e−

e+

W

νe

ν̄e

Z

ν̄e

j

j


Discovery prospects of the hadronic channel (FDM)

Preselection cuts:

Lepton veto

Exactly two jets

EmissT > 50 GeV

Boosted decision tree analysis with inputs:

mDD, pT (j1), pT (Z), EmissT , ∆φmin, pT (j2), mjj

10−3

10−2

10−1

100

−0.8 −0.6 −0.4 −0.2 0 0.2 0.410−3

10−2

10−1

100

−0.8 −0.6 −0.4 −0.2 0 0.2 0.4

NS/N

B

BDT Cut

FDM200FDM300FDM400FDM500

NS/N

B

BDT Cut


0

2

4

6

8

10

12

14

−0.8 −0.6 −0.4 −0.2 0 0.2 0.40

2

4

6

8

10

12

14

−0.8 −0.6 −0.4 −0.2 0 0.2 0.4

Significance

BDT Cut

L =1000 fb−1


Significance

BDT Cut

L =1000 fb−1



Discovery prospects of the hadronic channel (FDM)

FDM200 FDM300 FDM400 FDM500

σ0 [fb] 1.643 0.9214 0.4221 0.2526

εpre 0.796 0.717 0.655 0.698

BDT 0.3615 0.2132 0.1929 0.2129

NS/1000 fb−1 697.8 410.5 148 102

NB/1000 fb−1 2248.5 11453.5 12736 10898

NS/√NS +NB 12.85 3.769 1.31 0.97


Outline

1 Generalities



4 Conclusion


Signal: mono-jet?

Tmisp

0 100 200 300 400 500 600 700 800

Eve

nt F

ract

ion

0

0.05

0.1

0.15

0.2

0.25

0.3

Scalar

minΓFermion

minΓ×Fermion 20

minΓVector

minΓ×Vector 20

CMS mono-jet search,

arXiv:1703.01651

CMS tt+DM search in dileptonic

channel,

arXiv:1711.00752

M. R Buckley et.al.;

U. Haisch et.al.;

F.Boudjema et.al.;

J. Ellis et.al. ...


ttH channel is promising channel

The observed (expected) significance of 5.8 (4.9) stan-dard deviations

ttH cross section is enhanced more significantly withincreasing collision energy.


Signal and background

The dominant DM production process:

q

q̄

g

t̄

t

D

Dhi

g

g

hi

t

t̄

D

D

g

g

t̄

t

D

Dhi

Dominant background processes:

Cross section (NLO)

t̄t 1316.5 pbt̄tW 20.5 pbt̄tZ 64.2 pb

W`W`b 128.4 pb

Signal benchmark points:

t`t`+ DM

FDM200 34.2 fbFDM300 18.7 fbFDM400 14.8 fbFDM500 12.5 fb


Features of DM spin

200 300 400 500 600 700 800

DDm

0.01

0.1

1

Eve

nt F

ract

ion

=3χ

Parton, g

SDM

FDM 200 300 500

VDM 200 300 500

=3χ

Parton, g

200 300 400 500 600 700 800 [GeV]DDm

1−10

1

10

(V/F

) ra

tio

1− 0.8− 0.6− 0.4− 0.2− 0 0.2 0.4 0.6 0.8 1)llθcos(

0.1

Eve

nt F

ract

ion

=3χ

Parton, g

SDM

FDM200

VDM200

DM300

DM400

DM500

=3χ

Parton, g


Analysis strategy

Preselection: Exactly two opposite sign lepton and at least one bjet in the final state.

m`` /∈ [85, 95] GeV.

0 100 200 300 400 500 600 700 [GeV]T

missE

0

10

20

30

40

50

60

70

80

90

Cro

ss s

ectio

n [p

b]

=3χ

Delphes(After preselection), g

SM

SDM

FDM 200 300 400 500

VDM 200 300 400 500

=3χ


0 50 100 150 200 250 300(l,l) [GeV]T2m

0

10

20

30

40

50

60

70

80

Cro

ss s

ectio

n [p

b]

=3χ


SM

SDM

FDM 200 300 400 500

VDM 200 300 400 500

=3χ


EmissT > 150 GeV.

mT2(l, l) > 150 GeV.


Cuts flow for SM processes and Signals

Backgrounds t̄t t̄tW t̄tZ W`W`bCross section 1316.5 pb 20.5 pb 64.2 pb 128.4 pb

Presections 63.76 pb 351.8 fb 1.9 pb 25.4 pb

m`` /∈ [85, 95] GeV 59.8 pb 330.4 fb 1.05 pb 23.9 pbEmissT > 150 GeV 17.76 pb 69.61 fb 261.14 fb 3.5 pbmT2(l, l) > 150 GeV 23.83 fb 1.92 fb 32.1 fb 38.0 fb

Signals FDM200 FDM300 FDM400 FDM500Cross section 34.2 fb 18.7 fb 14.8 fb 12.5 fb

Presections 7.86 fb 3.99 fb 3.05 fb 2.55 fb

m`` /∈ [85, 95] GeV 7.47 fb 3.82 fb 2.92 fb 2.44 fbEmissT > 150 GeV 4.17 fb 2.44 fb 1.93 fb 1.63 fbmT2(l, l) > 150 GeV 0.87 fb 0.62 fb 0.54 fb 0.47 fb

L95% [fb−1] 305 602 793 1047


Spin characterization (EmissT and cos(θ``))

Two dimensional binned log-likelihood test: L(data|Hα) =∏i,j

tnijij e−tij

nij !

σsys. = 0.5%


Conclusion

The gauge invariant Higgs portal DM models for FDM and VDMrequire at least two mediators, while that of SDM only need one.

At the ILC, mH2 . 300 GeV can be probed at more than 3-σ level.

At the 100 TeV p-p collider, discovery/exclusion can be made withan integrated luminosity less than 1 ab−1 given a 1% systematicuncertainty, while the spin discrimination require integratedluminosity of a few O(10) ab−1 given a 0.5% systematicuncertainty.


Backup.


Benchmark points

The relevant parameters in FDM for collider search:gχ = 3, sinα = 0.3, mχ = 80 GeV and mH2

= (200, 300, 400, 500) GeV.

ΓFDMmin (H2) = Γ(H2 → χχ) + Γ(H2 → WW/ZZ) + Γ(H2 → ff)

= cos2α · g2

χ

mH2

8π(1−

4m2χ

m2H2

)3/2

+ sin2α ·

Gµm3H2

16√

2πδV

√√√√1− 4m2V

m2H2

(1− 4m2V

m2H2

+ 12m4V

m4H2

)

+ sin2α · (

mf

v)2 3mH2

8π(1−

4m2f

m2H2

)3/2

,

where f is the SM fermion, V = Z,W and δV = 1(2) for Z(W±).

Parameters for the vector DM production are chosen accordingly:sinα = 0.3, mV = 80 GeV and gV is chosen such that the total decaywidth of H2 is the same as benchmark points of FDM.

mH2[GeV] 200 300 400 500

Γmin(H2) [GeV] 14.2 60.1 103.0 144.5gV 3.53 3.07 2.37 1.91

Fix mS = 80 GeV and take appropriate λHS such that the productioncross section of the signal process is the same with that in the FDM.


Discovery prospects of the hadronic channel

Kinematic distributions:

(j1) [GeV]T

p0 50 100 150 200 250

Eve

nt F

ract

ion

0

0.005

0.01

0.015

0.02

0.025

0.03

0.035

0.04

0.045

0.05

SM

SDM

VDMs

FDM200

FDM300

FDM400

FDM500

(Z) [GeV]T

p0 50 100 150 200 250

Eve

nt F

ract

ion

0

0.005

0.01

0.015

0.02

0.025

0.03

0.035

0.04

0.045

0.05

SM

SDM

VDMs

FDM200

FDM300

FDM400

FDM500

[GeV]missTE

0 50 100 150 200 250

Eve

nt F

ract

ion

0

0.005

0.01

0.015

0.02

0.025

0.03

0.035

0.04

0.045

0.05SM

SDM

VDMs

FDM200

FDM300

FDM400

FDM500

minφ ∆0 0.5 1 1.5 2 2.5 3

Eve

nt F

ract

ion

0

0.01

0.02

0.03

0.04

0.05

0.06

0.07

SM

SDM

VDMs

FDM200

FDM300

FDM400

FDM500

∆φmin

= mini=1,2

∆φ(pmissT , p(ji))


Spin characterization

The same preselection and BDT cuts as used for FDM the benchmark pointFDM200 (FDM300) are applied to the corresponding benchmark point SDM200(SDM300) and VDM200 (VDM300).

SDM200 SDM300 VDM200 VDM300

σ0 [fb] 1.643 0.9214 1.734 0.8674

εpre 0.7875 0.7875 0.801 0.711

NS/1000 fb−1 447 322.3 726 363.5

S 4.4 3.3 0.59 0.44

[GeV]χχm160 170 180 190 200 210 220 230 240 250

Eve

nt N

umbe

r

200

400

600

800

1000

1200

SDM+SM

VDM+SM

FDM+SM

=200 GeV)2H

After BDT cut (m

[GeV]χχm160 180 200 220 240 260 280 300 320 340 360

Eve

nt N

umbe

r

1000

2000

3000

4000

5000

6000

SDM+SM

VDM+SM

FDM+SM

=300 GeV)2H

After BDT cut (m

SDM: δχ2

=

5∑i=1

(NFDM+SMi −NSDM+SM

i√NFDM+SMi

)2

VDM: S = |NFDMS −NVDM

S |/√NB


Spin characterization

Using only the distribution of EmissT :H0 is the FDM + SM, Hα can be VDM/SDM + SM


Discovery prospects

200 300 400 500 600 700 [GeV]T

missE

0

5

10

15

20

25

30

35

Cro

ss s

ectio

n [fb

]

=3χ

Delphes(After cuts), g

SM

SDM

FDM 200 300 400 500

VDM 200 300 400 500

=3χ

Delphes(After cuts), g

Binned log-likehood analysis:

L(data|Hα) =∏i

tnii e−ti

ni!, Q = −2 log

(L(data|Hα)

L(data|H0)

).

σsys. = 1%


higgs portal at future colliders · higgs portal dark matter model p. ko, j. li. arxiv:1610.03997 l...

Documents