Flavor Violating Higgs Decays

Flavor Violating Higgs Decays

Joachim Kopp

Galileo Galilei InstituteNovember 26, 2012

Based on work done in collaboration withRoni Harnik and Jure Zupan

arXiv:1209.1397

Joachim Kopp Flavor Violating Higgs Decays 1

Outline

1 Flavor mixing in the Higgs sector

2 Couplings to leptons

3 Couplings to quarks

4 Flavor-violating Higgs decays at the LHC

5 Summary


Flavor Mixingin the Higgs Sector

Motivation

Scenario 1: Several sources of EW symmetry breakingIf fermion masses have more than one origin, they do not need to bealigned with the Yukawa couplings

Simplest example: Type III 2-Higgs-Doublet Model

LY ⊃ −Y (1)ij Liej

RH(1) − Y (2)ij Liej

RH(2) + h.c.

−→ −mi eiLei

R − Y effij f i

Lf jRh + couplings to heavier Higgs bosons + h.c.

(h = Lightest neutral Higgs boson, mh ∼ 125 GeV)

Assume heavy Higgs boson are decoupled.see for instance Davidson Greiner, arXiv:1001.0434


Motivation (2)

Scenario 2: Extra Higgs couplingsAssume existence of heavy new particles, which induce effective operators ofthe form

∆LY = −λ′ijΛ2 (f i

Lf jR)H(H†H) + h.c.+ · · · ,

→ after EWSB, new (but misaligned) contributions to mass matrices andYukawa couplings

Effective Lagrangian is again

LY ⊃ −mi eiLei

R − Y effij f i

Lf jRh + h.c.

see for instance Giudice Lebedev, arXiv:0804.1753


Effective Yukawa Lagrangfian

Effective Yukawa Lagrangian

LY = −mi f iLf i

R − Y aij (f i

Lf jR)ha + h.c.+ · · ·

Previously studied by many authors:

Bjorken Weinberg, PRL 38 (1977) 622 McWilliams Li, Nucl. Phys. B 179 (1981) 62Shanker, Nucl. Phys. B 206 (1982) 253 Barr Zee, PRL 65 (1990) 21Babu Nandi, hep-ph/9907213 Diaz-Cruz Toscano, hep-ph/9910233Han Marfatia, hep-ph/0008141 Kanemura Ota Tsumura, hep-ph/0505191Blanke Buras Duling Gori Weiler, arXiv:0809.1073Casagrande Goertz Haisch Neubert Pfoh, arXiv:0807.4937Giudice Lebedev, arXiv:0804.1753 Aguilar-Saavedra, arXiv:0904.2387Albrecht Blanke Buras Duling Gemmler, arXiv:0903.2415Buras Duling Gori, arXiv:0905.2318 Azatov Toharia Zhu, arXiv:0906.1990Agashe Contino, arXiv:0906.1542 Davidson Greiner, arXiv:1001.0434Goudelis Lebedev Park, arXiv:1111.1715 Blankenburg Ellis Isidori, arXiv:1202.5704Arhrib Cheng Kong, arXiv:1208.4669 McKeen Pospelov Ritz, arXiv:1208.4597. . .


Effective Yukawa Lagrangfian

Effective Yukawa Lagrangian

LY = −mi f iLf i

R − Y aij (f i

Lf jR)ha + h.c.+ · · ·

New in this talk:Comprehensive list of up-to-date constraints (including subdominant ones)

Omit approximations where feasibleFirst LHC limitsStrategy for future LHC searches


Couplings to Leptons

Low-energy constraints on LFV in the Higgs sector

h

µ+

e−

e+

µ−

Y ∗eµPL + YµePR

Y ∗eµPL + YµePR

M–M oscillations

h

N

µ

N

eY ∗µePL + YeµPR

µ–e conversion

τ

h

τµ

γ

µY ∗µτPL + YτµPR Y ∗

τµPL + YµτPR

g − 2, EDMs

hτ

µ

µ

µ

Y ∗τµPL + YµτPR

Y ∗µµPL + YµµPR

τ → 3µ, µee, etc.

τ

h

ττ

γ

µY ∗ττPL + YττPR Y ∗

τµPL + YµτPR

µ

h γ, Z

tt

τ

γ

µ

µ

h γ, Z

WW

τ

γ

µµ

h γ, Z

W W

τ

γ

µ

τ → µγ, µ→ eγ, etc.


Constraints on h→ µe

10- 7 10-6 10- 510- 4 10-310- 2 10-1 100 10110- 7

10-6

10- 5

10- 4

10-3

10- 2

10-1

100

101

ÈYeΜÈ

ÈY ΜeÈ

Μ ® eΓ

M ® M

Μ ® 3e

Μ ® e conv.

Hg-2Le +ED

Me

Hg-2Le for ImHY

Μe YeΜ L=0

EDM

e for ReHYΜe Y

eΜ L=0

ÈYΜe Y

eΜ È=me m

Μ �v 2

BRHh

®ΜeL

=0.99

10-

910

-8

10-

710

-6

10-

510

-4

10-

310

-2

10-

10.5

see also: Blankenburg Ellis Isidori, arXiv:1202.5704Goudelis Lebedev Park, arXiv:1111.1715


Assumption here:Diagonal Yukawa coupling unchangedfrom their SM values.

Constraints on h→ τµ and h→ τe

10- 3 10- 2 10-1 100

10- 3

10- 2

10-1

100

ÈY ΜΤ È

ÈY ΤΜ

È

Τ®

ΜΓ

Τ ® 3 Μ

H g -2 L Μ

+ED

MΜ

Hg -2 L Μ

�Im

HY ΤΜY ΜΤ

L =0

ÈYΤΜ Y

ΜΤ È =mΜ m

Τ � v 2

BRH

h®

ΤΜL

=0.99

10-

3

10-

2

10-

1

0.50.75

10-5 10-4 10- 3 10- 2 10-1 10010-5

10-4

10- 3

10- 2

10-1

100

ÈYeΤ ÈÈY Τ

eÈ

Τ®

eΓ

Τ ® eΜΜ

Hg-2 Le +

EDMe

H g-2 L

e forIm H Y

Τe YeΤ L =0

ÈYΤe Y

eΤ È =me m

Τ � v 2

BRH

h®

ΤeL=

0.99

10-

6

10-

5

10-

3

10-

2

10-

1

0.5

EDMe �

Re HYΤe Y

eΤ L =0

Substantial flavor violation (BR(h→ τµ, τe) ∼ 0.01) perfectly viable.

see also: Blankenburg Ellis Isidori, arXiv:1202.5704Goudelis Lebedev Park, arXiv:1111.1715

Davidson Greiner, arXiv:1001.0434


Assumption here:Diagonal Yukawa coupling unchangedfrom their SM values.

Couplings to Quarks

Constraints on Higgs couplings to light quarksTight constraints from neutral meson oscillations

h

d

b

b

dY ∗bdPL + YdbPR

Y ∗bdPL + YdbPR

t

h

h

t

u

c

c

uY ∗ctPL + YtcPR

Y ∗tuPL + YutPR Y ∗

ctPL + YtcPR

Y ∗tuPL + YutPR

Work in Effective Field Theory:

Heff = Cdb2 (bRdL)2 + Cdb

2 (bLdR)2 + Cdb4 (bLdR)(bRdL) + . . .

Wilson coefficients constrained in UTfit (Bona et al.), arXiv:0707.0636see also Blankenburg Ellis Isidori, arXiv:1202.5704


Constraints on Higgs couplings to light quarksTight constraints from neutral meson oscillationsWork in Effective Field Theory:

Heff = Cdb2 (bRdL)2 + Cdb

2 (bLdR)2 + Cdb4 (bLdR)(bRdL) + . . .

Wilson coefficients constrained in UTfit (Bona et al.), arXiv:0707.0636see also Blankenburg Ellis Isidori, arXiv:1202.5704

Technique Coupling Constraint

D0 oscillations |Yuc |2, |Ycu|2 < 5.0× 10−9

|YucYcu| < 7.5× 10−10

B0d oscillations |Ydb|2, |Ybd |2 < 2.3× 10−8

|YdbYbd | < 3.3× 10−9

B0s oscillations |Ysb|2, |Ybs|2 < 1.8× 10−6

|YsbYbs| < 2.5× 10−7

K 0 oscillations

<(Y 2ds), <(Y 2

sd ) [−5.9 . . . 5.6]× 10−10

=(Y 2ds), =(Y 2

sd ) [−2.9 . . . 1.6]× 10−12

<(Y ∗dsYsd ) [−5.6 . . . 5.6]× 10−11

=(Y ∗dsYsd ) [−1.4 . . . 2.8]× 10−13


Couplings involving top quarks

10- 2 10-1 100 10110- 2

10-1

100

101

ÈYqt È Hq = c, uL

ÈY tqÈHq=

c,u

L

0.5

10-1

10- 2

10- 3

10- 4

10-5

0.5

10-1

10- 2

10- 3

10- 4

BRHh® t * q LBRHt ® hq L

single

top

bo

un

do

nÈYct È,ÈY

tc Èsin

gleto

pb

ou

nd

onÈY

ut È,ÈY

tu È

t®h

qlim

itHCraig

etal.L

Constraints from

Single top production

t

h

tt

g

qY ∗ttPL + YttPR Y ∗

tqPL + YqtPR

CDF 0812.3400, DØ 1006.3575ATLAS 1203.0529

t → hqCraig et al. 1207.6794

based on CMS multilepton search1204.5341

Not sensitive

t → ZqCMS 1208.0957


Flavor-Violating Higgs Decaysat the Large Hadron Collider

h→ τµ and h→ τe at the LHC

Basic idea:

h→ τ` has the same final stateas h→ ττ`(but is enhanced by 1/BR(τ → `))

Recast h→ ττ searchhere: ATLAS, arXiv:1206.5971

We consider only 2-leptonfinal statesUse VBF cuts(much lower BG than gg fusion)

see howeverDavidson Verdier, arXiv:1211.1248

based on ATLAS, arXiv:1206.5971

æ

æ

æ æ æ

æ æ

0 100 200 300 4000

5

10

15

20

25

ΤΤ collinear mass mΤΤ @GeV DE

ven

ts�1

0G

eV

ATLAS 4.7 fb-1

ATLAS MCææ ATLAS data

5 ´ H ® Τ{

Τ{

5 ´ H ® Τ{

Μ

H ÈY ΜΤ2

+ÈY ΤΜ2 L 1 � 2

=m Τ

v


h→ τµ and h→ τe at the LHC

Most important cuts2 forward jets (pT ,j1 > 40 GeV,pT ,j2 > 25 GeV, |∆η| > 3.0,minv

j1,j2 > 350 GeV)

no hard jet activity in betweenno b tags2 opposite sign leptons `1, `2 withpT ,` & 10–20 GeV (depending onflavors)

τ momentum fraction x carried by`1, `2 satisfies 0.1 < x1,2 < 1.0(computed in collinear approximation)

30 GeV < m`` < 75 (100) GeVfor same (opposite) flavor leptons

/ET > 20 (40) GeVfor same (opposite) flavor leptons


æ

æ

æ æ æ

æ æ

0 100 200 300 4000

5

10

15

20

25

ΤΤ collinear mass mΤΤ @GeV DE

ven

ts�1

0G

eV

ATLAS 4.7 fb-1


5 ´ H ® Τ{

Τ{

5 ´ H ® Τ{

Μ

H ÈY ΜΤ2

+ÈY ΤΜ2 L 1 � 2

=m Τ

v


Mass reconstruction for h→ τ`τ`

Problem: 4 neutrinos in final state

Solution: Assume all τ decay products collinearEllis Hinchliffe Soldate van der Bij, NPB 1987

Per τ :2 unknown (|pντ |, |pν` |)2 constraints: /ET ,x , /ET ,y


Limits on h→ τµ and h→ τe from the LHC

Technicalities

Use MadGraph 5, v1.4.6,Pythia 6.4, PGSUse only 120–160 GeV binDerive one-sided 95% CL limit


æ

æ

æ æ æ

æ æ

0 100 200 300 4000

5

10

15

20

25

ΤΤ collinear mass mΤΤ @GeV D

Eve

nts

�10

GeV

ATLAS 4.7 fb-1


5 ´ H ® Τ{

Τ{

5 ´ H ® Τ{

Μ

H ÈY ΜΤ2

+ÈY ΤΜ2 L 1 � 2

=m Τ

v

Result

BR(h→ τµ) < 0.13√

Y 2τµ + Y 2

µτ < 0.011

BR(h→ τe) < 0.13√

Y 2τe + Y 2

eτ < 0.011


LHC constraints on h→ τµ and h→ τe

10- 3 10- 2 10-1 100

10- 3

10- 2

10-1

100

ÈY ΜΤ È

ÈY ΤΜ

È

Τ®

ΜΓ

Τ ® 3 Μ

H g -2 L Μ

+ED

MΜ

Hg -2 L Μ

�Im

HY ΤΜY ΜΤ

L =0

ÈYΤΜ Y

ΜΤ È =mΜ m

Τ � v 2

BRH

h®

ΤΜL

=0.99

10-

3

10-

2

10-

1

0.50.75

10-5 10-4 10- 3 10- 2 10-1 10010-5

10-4

10- 3

10- 2

10-1

100

ÈYeΤ ÈÈY Τ

eÈ

Τ®

eΓ

Τ ® eΜΜ

Hg-2 Le +

EDMe

H g-2 L

e forIm H Y

Τe YeΤ L =0

ÈYΤe Y

eΤ È =me m

Τ � v 2

BRH

h®

ΤeL=

0.99

10-

6

10-

5

10-

3

10-

2

10-

1

0.5

EDMe �

Re HYΤe Y

eΤ L =0



10- 3 10- 2 10-1 100

10- 3

10- 2

10-1

100

ÈY ΜΤ È

ÈY ΤΜ

È

Τ®

ΜΓ

Τ ® 3 Μ

H g -2 L Μ

+ED

MΜ

Hg -2 L Μ

�Im

HY ΤΜY ΜΤ

L =0

ÈYΤΜ Y

ΜΤ È =mΜ m

Τ � v 2

Our LHC limitH ATLAS 7 TeV , 4.7 fb - 1 L B

RHh

®Τ

ΜL=

0.99

10-

3

10-

2

10-

1

0.50.75

10-5 10-4 10- 3 10- 2 10-1 10010-5

10-4

10- 3

10- 2

10-1

100

ÈYeΤ ÈÈY Τ

eÈ

Τ®

eΓ

Τ ® eΜΜ

Hg-2 Le +

EDMe

H g-2 L

e forIm H Y

Τe YeΤ L =0

ÈYΤe Y

eΤ È =me m

Τ � v 2

Our LHC limitH ATLAS 7 TeV , 4.7 fb - 1 L

BRH

h®

ΤeL=

0.99

10-

6

10-

5

10-

3

10-

2

10-

1

0.5

EDMe �

Re HYΤe Y

eΤ L =0



10- 3 10- 2 10-1 100

10- 3

10- 2

10-1

100

ÈY ΜΤ È

ÈY ΤΜ

È

Τ®

ΜΓ

Τ ® 3 Μ

H g -2 L Μ

+ED

MΜ

Hg -2 L Μ

�Im

HY ΤΜY ΜΤ

L =0

ÈYΤΜ Y

ΜΤ È =mΜ m

Τ � v 2

Our LHC limitH ATLAS 7 TeV , 4.7 fb - 1 L B

RHh

®Τ

ΜL=

0.99

10-

3

10-

2

10-

1

0.50.75

10-5 10-4 10- 3 10- 2 10-1 10010-5

10-4

10- 3

10- 2

10-1

100

ÈYeΤ ÈÈY Τ

eÈ

Τ®

eΓ

Τ ® eΜΜ

Hg-2 Le +

EDMe

H g-2 L

e forIm H Y

Τe YeΤ L =0

ÈYΤe Y

eΤ È =me m

Τ � v 2

Our LHC limitH ATLAS 7 TeV , 4.7 fb - 1 L

BRH

h®

ΤeL=

0.99

10-

6

10-

5

10-

3

10-

2

10-

1

0.5

EDMe �

Re HYΤe Y

eΤ L =0


WORLD’S BEST LIMIT!

Strategy for a dedicated h→ τµ and h→ τe search

Possible improvementsDifferent invariant mass formula(assuming 1 neutrino rather than 3)

I Avoids smearing of signalI Shifts Z → ττ peak to

lower invariant mass

Consider hadronic τ ’s(especially for CMS)

Modified cutsI CMS h→ τhadτ` search requires

mT (`, /pT ) < 40 GeV to suppressW + jets

I In h→ τhadµ, neutrino and muontypically not collinear→ large mT (`, /pT )

50 100 150 2000

2

4

6

8

10

12

Transverse mass of the Μ-p�T system @GeV D

Eve

nts

�10G

eV

5 fb-1, 7 TeVMG �Pythia �Delphes

BG rescaled toCMS-HIG -11-029

W+jets

Z+jets

h ® Τ had Τ Μ

h ® Τ had Μ

HÈY ΜΤ2+ÈY ΤΜ

2L1� 2= mΤ

v

QCD BG neglected



Possible improvementsDifferent invariant mass formula(assuming 1 neutrino rather than 3)

I Avoids smearing of signalI Shifts Z → ττ peak to

lower invariant mass

Consider hadronic τ ’s(especially for CMS)

Modified cutsI CMS h→ τhadτ` search requires

mT (`, /pT ) < 40 GeV to suppressW + jets

I In h→ τhadµ, neutrino and muontypically not collinear→ large mT (`, /pT )

50 100 150 2000

2

4

6

8

10

12

Transverse mass of the Μ-p�T system @GeV DE

ven

ts�10

GeV



W+jets

Z+jets

h ® Τ had Τ Μ

h ® Τ had Μ

HÈY ΜΤ2+ÈY ΤΜ

2L1� 2= mΤ

v

QCD BG neglected



50 100 150 200 250 3000

5

10

15

20

ΤΜ invariant mass mΤΜ @GeV D

Eve

nts

�10G

eV



W+jets

Z+jets

h ® Τ had Τ Μ

h ® Τ had Μ

HÈY ΜΤ2+ÈYΤΜ

2L1� 2= mΤ

v

QCD BG neglected

50 100 150 200 250 3000

5

10

15

ΤΜ invariant mass mΤΜ @GeV DE

ven

ts�10

GeV



W+jets

Z+jets

h ® Τ had Τ Μ

h ® Τ had Μ

HÈY ΜΤ2+ÈYΤΜ

2L1� 2= mΤ

v

QCD BG neglected

For Yµτ , Yτµ close to the current upper limits, spectacular signals possible.


Exploiting Higgs production in gluon-gluon fusionDavidson Verdier, arXiv:1211.1248

ObservationsComputed pT ,ν (using collinear approximation) is ' /ET

Muon in h→ τµ is much harder than in h→ τ`τ`.

TEδ

10 8 6 4 2 0 2 4 6 8 10

Even

ts / 0

.2

110

1

10

210 + jets

l

+ l→Z

tt

WW,WZ,ZZ

singlet

SM Higgs

Signal

(GeV)T

Muon p0 20 40 60 80 100 120 140 160 180 200

Even

ts / 2

GeV

110

1

10

210

+ jets

l+

l→Z

tt

WW,WZ,ZZ

singlet

SM Higgs

Signal

Projected sensitivity Davidson Verdier, arXiv:1211.1248

BR(h→ τµ), BR(h→ τe) < 4.5× 10−3


Exploiting Higgs production in gluon-gluon fusionDavidson Verdier, arXiv:1211.1248

ObservationsComputed pT ,ν (using collinear approximation) is ' /ET

Muon in h→ τµ is much harder than in h→ τ`τ`.

TEδ

10 8 6 4 2 0 2 4 6 8 10

(G

eV

)T

Mu

on

p

0

20

40

60

80

100

120

140

160

180

200

210

110

1

TEδ

10 8 6 4 2 0 2 4 6 8 10

(G

eV

)T

Mu

on

p

0

20

40

60

80

100

120

140

160

180

200

310

210

110

Background Signal

Projected sensitivity Davidson Verdier, arXiv:1211.1248

BR(h→ τµ), BR(h→ τe) < 4.5× 10−3


SummaryFlavor-violating Higgs couplings arise in

I Models with several sources of electroweak symmetry breakingI Models with heavy fields coupled to the Higgs

In the lepton sector:I Constraints from `1 → `2 + γ, `1 → `2 + X , µ–e conversion in nuclei, g − 2,

EDMs, M–M oscillationsI Strong constraints in the µ–e sectorI Very weak constraints in the τ–e and τ–µ sectors

In the quark sector:I Strong constraints on couplings to light quarksI Very weak constraints on couplings to top quarks

At the LHCI Constraints on anomalous top–Higgs couplings from single top productionI A recast ATLAS h→ τ`τ` search already provides strongest limits on

h→ τµ and h→ τeI A dedicated search would be much more sensitive


Thank you!

Flavor Violating Higgs Decays

Documents

Transcript of Flavor Violating Higgs Decays