search for neutrino mass generation mechanisms with atlas

Search for neutrino mass generation mechanisms with ATLAS experiment

Kenji Hamano (University of Melbourne/CoEPP) 01/03/2113 workshop at Melbourne

Plan of the talk §  Introduction

§  Neutrino mass generation mechanism. §  Search mode in ATLAS.

§  Search in ATLAS §  Doubly Charged Higgs search. §  Type III seesaw heavy lepton search.

01/03/2013 CoEPP CAASTRO Workshop, K. Hamano 2

Neutrino Mass §  Neutrinos have small mass. But the Standard Model does

not have the mechanism to generate neutrino masses. §  Two mechanisms (without SUSY) to explain small mass

of neutrinos : 1.   Seesaw mechanism (tree level).

§  Type I : Additional SU(2) singlet right-handed neutrino. §  Do not couple to W or Z. §  Not really testable in LHC.

§  Type II : Additional SU(2) triplet scalar (φ0,φ+,φ++). §  Type III : Additional SU(2) triplet lepton (N-,N0,N+).

2.   Radiative mass generation mechanism (loop level). §  Zee model : simple one loop mass generation.

§  Excluded. §  Zee-Babu model : Additional two scalars h+ and k++

§  currently simplest two loop model.


Type II Seesaw and Zee-Babu Models

§  Doubly Charged Scalar (φ++ or k++ )is often called Doubly Charged Higgs (H++).

§  H++ decays to same-sign charged lepton pair. §  Unique signature in LHC.

§  H++ can be pair produced in pp collisions.


γ∗, Z∗

q

q

k−−

k++

Figure 2: Pair production of k

gauge charges as well as depending only on one unknown parameter: the mass of the scalar.

The partonic cross section at LO reads

σ =πα2Q2β3

6

[2Q2

q

s−

2(gL + gR)Qq

c2w

s − M2Z

(s − M2Z)2 + Γ2

ZM2Z

+(g2

L + g2R)

c4w

s

(s − M2Z)2 + Γ2

ZM2Z

],

(30)

where s is the energy squared in the center of mass frame (CM) of the quarks, Q stands for

electric charges, gL and gR are given for the quarks by gL = T3 − s2wQq and gR = −s2

wQq

and β is the velocity of the produced scalars in this frame β =√

1 − 4m2/s.

Equation (30) shows that pair production is four times more efficient for k than for h due

to their charges (assuming equal masses), which translates into a better discovery potential

for k. The k pair production cross section, σkk, at NLO for the LHC and Tevatron is displayed

in fig. 3. To compute it, we have used CompHEP [62] with CTEQ6.1L libraries [63] to find

the LO cross section and afterwards we have included a K-factor of 1.25 for the LHC and

1.3 for Tevatron to take into account NLO corrections, see [64].

Single production might be also interesting when double production is not possible. Single

production can proceed with a k accompanied by two singly charged scalars, fig. 4, or by

two charged leptons replacing the scalar h’s. If the k is accompanied by two charged leptons

the amplitudes are proportional to the Yukawa couplings, whose exact values we ignore and

might be small.

It is important to note that the cross section will be dominated by the virtual particles

in the propagators if they could be on-shell. In the case of k being produced with two h, the

single production will be dominated by the first diagram if s > 2mk, because in this case

k∗ can be created on-shell. One might argue that the energy in the center of mass frame of

17

It is possible to reconstruct 2 same-sign pairs in one event.

Type III Seesaw Model §  Pair production cross-section :

§  Possible to search in LHC

§  Decay modes :


N�N�N�N0

N�N0

200 400 600 800 1000 1200 140010�1

1

10

102

103

M in GeV

Σinfb

M � 250 GeVM � 500 GeVM � 1000 GeV

0.0 0.5 1.0 1.5 2.0 2.5 3.010�4

10�3

10�2

10�1

1

pT �M

M�dN�d

p T��N

Figure 1: Left plot: total production cross section at LHC. Right plot: pT distribution.

σ =β(3− β2

)

48πNcs(V

2L + V 2

R) (7b)

where Nc = 3, β ≡�

1− 4M2/s is the N velocity (0 ≤ β ≤ 1) and

VA = 0 for qq → N0N0

VA =Qqe2

s+

gqAg2

2

s−M2Z

for qq → N+N−

VA =g22

s−M2W

δAL√2

for ud→ N+N0

(8)

where gqA = T3− s2

W Qq is the Z coupling of quark q for A = {L, R}. This result does not agree

with eq. (10) of [16]. SU(2)L invariance is restored in the limit M2 � M2Z , and the result

2σuu = 2σdd = σud = σdu agrees with [13]. The cross section e−e+ → N+N−, relevant for a

possible future collider, is found by replacing q → e in eq. (8).

Fig. 1a shows σ(pp → N0N±) and σ(pp → N+N−) as function of M at LHC, i.e. at√

s =

14 TeV. We integrated the parton distribution functions of [17], and we checked that the result

numerically agrees with the one obtained implementing the triplet model in MadGraph [18].

This would lead to about 3 · 103

(10) pairs created at LHC for M = 250 GeV (M = 1 TeV)

for an integrated luminosity of 3/fb which should be collected at LHC in less than one year.

These numbers have to be multiplied by about 2 orders of magnitude after 5 years of data

taking. Therefore LHC should be able to produce at least a few tens of events up to masses of

M ∼ 1 TeV, or even 1.5 TeV in the long term. Fig. 1b shows the distribution in the transverse

momentum pT , as computed by our MonteCarlo for three representative values of M : it is

peaked at pT ∼M/2.

Testing the production cross section would allow to identify the quantum numbers of the

particle and to test that/if the theory correctly predicts the gauge interactions of a fermionic

5

100 100030010�3

10�2

10�1

1

10

M in GeV

�in1�cm

N� � N0��Ν

N� � N0Π�

N0 �W��, N� �W�Ν

N0 � ZΝ, N� � Z��

N0 � Νh, N� � ��h

Figure 2: Triplet decay widths as function of the triplet mass for m1 = meV and mh = 115 GeV.

Notice that, while pp → N0N0 does not arise at tree level, this production channel is

effectively produced by the N± → N0π± decay because the π± are too soft to be observed.

The decay mode into pions is dominant for m <∼ 3 ·10−4 eV · (100 GeV/M)2, so that the effective

production rate pp→ N0N0 is given by the sum of all cross sections in fig. 1a.

5 Signals at LHC and displaced vertices

Production of N0N± and N±N∓ and their decays give rise to a variety of possible final states.

We focus on those involving jets, that have higher rates than purely leptonic final states,

and need a discussion of Standard Model backgrounds and how they can be suppressed. In

section 5.1 we study the signal with the higher rate; lepton flavour violation (LFV) is studied in

section 5.2, and lepton number violation (LNV) in section 5.3. For simplicity in what follows we

often leave implicit that events with each particle replaced by its anti-particle are also possible:

signal and background rates are similar but not equal.

5.1 The signal with the higher rate

In view of the small fb-scale cross sections for N0, N± production, we first discuss the channel

with the relatively higher rate (for m1 >∼ 10−4 eV):

pp→ N+N0 → νW+W±�∓ → 4 jets + missing energy + a charged lepton, all hard. (11)

7

R. Franceschini, T. Hambye and A. Stumia, arXiv:0805.1613[hep-ph]

Type III Seesaw Model §  Possible search modes :

§  Same-sign dilepton (l- from N0 and l- from N+) §  3 leptons (Reconstruct N+ through Z(ll)+l+) §  4 leptons (Reconstruct N+ and another lepton from N0)


February 26, 2013 – 15 : 40 DRAFT 3

of the heavy singlet of the Type I Seesaw model, which is drastically suppressed by the generally small91

Yukawa couplings.92

If one requires O(1) Yukawa couplings, the mass of the heavy leptons MN should be of the order93

of the grand unification scale, in order to account for neutrino masses smaller than a few eV. However,94

in principle the scale can be as low as hundreds of GeV, in which case either the Yukawa couplings95

are smaller or an alternative method, such as for instance an inverse seesaw should be at work [1]. In96

this case the heavy field responsible for neutrino masses could be discovered at the LHC. This implies97

that there is a very large range in which to search for neutrino mass generating seesaw mechanisms.98

However, the signatures of the triplets are very characteristic, and allow for the construction of a new99

physics Lagrangian if accessible. The coupling of the N triplet to the Higgs generates a mass-mixing100

term between N0, N±, and ν", "L respectively. The N0/νL and N−/"L mixings are negligibly small.101

There is a small mass splitting as a result between N0 and N+, where in the limit M " MZ , this gives102

∆M = MN+ − MN0 166 MeV [2].103

1.2 Type III Seesaw phenomenology104

The Type III Seesaw model manifests with distinct experimental signatures at the LHC through direct105

production of N0 and N± . Although the both the N0 and the N± have many decay modes, by far the106

cleanest search is in fully leptonic channel. One such decay is shown in Figure 1. By identifying and107

reconstructing all leptons associated with the decay of the N± it is possible to perform a search for a108

narrow resonance on virtually no background. Below is a complete list of decay modes relevant for such109

a study at the LHC.110

• pp→ N0 + N± → "±W∓ + "±Z111

• pp→ N+ + N− → "−Z + "+Z112

• pp→ N+ + N− → "+Z + ν"W−

113

• pp→ N+ + N− → ν"W+ + "−Z114

• pp→ N0 + N± → ν"Z + ν"W±115

u

d

W+

N0

N+ Z

!+

!−

!+

!−

W+

Figure 1: One of the cleanest decay channels for the discovery of the Type III Seesaw model.

LHC and ATLAS detector §  LHC (Large Hadron Collider at CERN).

§  Proton-proton collider. §  Center of mass energy 7TeV (2011), 8TeV (2012)

§  ATLAS detector from inside to outside : §  Inner detector : reconstruct charged tracks. §  Calorimeters : Measure energies of electron, photon and

hadrons. §  Muon spectrometer


ATLAS experiment §  Look at transvers components.

§  Only part of protons (partons=quarks or gluons) collide.

§  Longitudinal component of initial energy or momentum of partons are unknown.

§  Transverse component of initial energy and momentum are zero.

§  Transverse momentum pT, transverse energy ET

§  Pseudo rapidity η

§  This is a Lorentz invariant way of expressing azimuthal angle θ.

§  Small |η| means θ~90 deg.


e. Non-relativistic limit Observe that for a non-relativistic particle, rapidity is the sameas velocity along the z-axis, for then

y =1

2lnE + pz

E − pz�1

2lnm+mvz

m−mvz= vz. (10)

Non-relativistic velocities transform additively under boosts, and the non-linear change of

variable from velocity to rapidity allows this additive rule Eq. (9) to apply to relativistic

particles (but only in one direction of boost).

One way of seeing this is as follows: The relativistic law for addition of velocities in one

dimension is

β13 =β12 + β231 + β12β23

, (11)

where β12 is the velocity of some object 1 measured in the rest-frame of object 2, etc. Thisformula is reminiscent of the following property of hyperbolic tangents:

tanh(A+B) =tanhA+ tanhB

1 + tanhA tanhB. (12)

So to obtain a linear addition law, we should write β12 = tanhA12, and then the rule Eq.(11) for the addition of velocities becomes simply A13 = A12 + A23. The A variables areexactly relative rapidities, since

vz =pz

E=p+ − p−

p+ + p−= tanh y. (13)

f. Relative velocity Rapidity is the natural relativistic velocity variable. Suppose wehave a proton and a pion with the same rapidity at pT = 0. Then they have no relativevelocity; to see this, one just boosts to the rest frame of one of the particles. But these same

particles have very different energies: Ep =mpmπEπ.

D. Pseudo-rapidity

As I will now explain, the rapidity of a particle can easily be measured in a situation

where its mass is negligible, for then it is simply related to the polar angle of the particle.

First let us define the pseudo-rapidity of a particle by

η = − ln tanθ

2, (14)

where θ is the angle of the 3-momentum of the particle relative to the +z axis. It is easy toderive an expression for rapidity in terms of pseudo-rapidity and transverse momentum:

y = ln

�m2 + p2T cosh

2 η + pT sinh η�m2 + p2T

. (15)

4

Doubly Charged Higgs Search §  H++→l+l+

§  Same-sign dilepton search. §  ee, eµ, µµ channels.

§  2011, 7TeV, 4.7fb-1 data (arXiV:1210.5070). §  Event Selection

§  Select well isolated electrons and muons with pT>20GeV and |η|<2.5.

§  Pair up highest and second highest pT leptons with same-sign.

§  Invariant mass of a lepton pair > 15GeV §  Remove Z mass region for ee pairs to reduce

background from Z.


Invariant mass §  Invariant mass

distributions after event selection.

§  If H++ exists, we should see a peak.


ee-channel µµ-channel

eµ-channel

3

set [28] added in quadrature to the difference between thecentral value of this set and the CTEQ6L PDF set.

The dilepton mass distribution observed in data is shownfor the e

±e±, µ±µ±, and e

±µ± channels in Fig. 1 and iscompared to the background expectation and four hypothet-ical H

±± signals normalised to their respective cross sec-tions (assuming a branching ratio to the given lepton flavourof 100%). The data show no clear peak structure and agreewell with the background estimate in all three channels.

A limit on the number of lepton pairs originating fromH

±± bosons (Nrec) in each mass window is derived using aCLs technique [29]. It is converted to a limit on the crosssection times branching ratio for doubly-charged Higgs pro-duction using the acceptance times efficiency values derivedfrom MC simulation. Since this analysis counts lepton pairsand each event contains two H

±± bosons, the cross sectiontimes branching ratio for pair production is given by

σ(pp → H±±

H∓∓)×BR(H±± → �±��±) =

Nrec(�±��±)

2×A× ε ×L, (2)

where A× ε is the acceptance times efficiency to detect alepton pair from H

±± decay within a given mass window.The integrated luminosity L is 4.7 fb−1.

The 95% CL expected and observed upper limits onthe cross section times branching ratio as a function of theH

±± boson mass are shown in Fig. 2. The expected limitis determined as the median outcome of simulated pseudo-experiments in the absence of any signal. Also shown are thetheoretical cross sections calculated at next-to-leading order(NLO) for H

±± production with left- and right-handed cou-plings [16]. The uncertainty on these cross sections is ±10%due to scale dependence in the NLO calculation, parton dis-tribution function uncertainties, and neglecting higher-orderelectroweak corrections.

At low mass, the expected cross-section limits are moststringent for the µ±µ± channel due to the low backgroundlevels in this channel. At high mass, the expected e

±e± and

µ±µ± limits are comparable while the e±µ± limit is about

30% worse due to the larger background from WZ produc-tion. In general the observed and expected limits agree wellwith each other. The largest deviations of the observed limitfrom the expected limit are within the 2σ uncertainty on theexpected limit. The cross-section limits range from 25 fb (inthe e

±e± channel at low mass) to 0.6 fb (in all channels at

high mass).Comparison of the cross-section limits with the theoret-

ical production cross section places constraints on m(H±±).The lower limits on the H

±± mass at 95% CL are listed inTable 1 for the three final states when BR(H±± → �±��±) =100%, as well as branching ratios of 33%, 22%, and 11%.For a democratic scenario where the BR to each pair of lep-ton flavours is the same, the branching ratio is 22% for the

) [GeV]±e±m(e

0 100 200 300 400 500 600

Elec

tron

pairs

/ 10

GeV

0

10

20

30

40

50

60Data 2011Non-promptCharge flipsPrompt

250 GeV±±LH

300 GeV±±LH

350 GeV±±LH

400 GeV±±LH

ATLAS

! -1Ldt = 4.7 fb = 7 TeVs

±e±e

(a)

) [GeV]±µ±µm(0 100 200 300 400 500 600

Muo

n pa

irs /

10 G

eV

0

5

10

15

20

25

30

35

40Data 2011Non-promptPrompt

250 GeV±±LH

300 GeV±±LH

350 GeV±±LH

400 GeV±±LH

ATLAS

! -1Ldt = 4.7 fb = 7 TeVs

±µ±µ

(b)

) [GeV]±µ±m(e

0 100 200 300 400 500 600

Lept

on p

airs

/ 10

GeV

0

10

20

30

40

50

60

70


250 GeV±±LH

300 GeV±±LH

350 GeV±±LH

400 GeV±±LH

ATLAS

! -1Ldt = 4.7 fb = 7 TeVs

±µ±e

(c)

Fig. 1 Invariant mass distributions for (a) e±

e±, (b) µ±µ±, and (c)

e±µ± pairs passing the full event selection. The data are shown as filled

circles. The stacked histograms represent the backgrounds composedof pairs of prompt leptons from SM processes, pairs with at least onenon-prompt lepton, and for the electron channels, backgrounds arisingfrom charge misidentification and conversions. The open histogramsshow the expected signal from simulated H

±±L

samples, assuming a100% branching ratio to the decay channel considered and coupling toleft-handed fermions. Lepton pairs in the e

±e± channel with an invari-

ant mass between 70 GeV and 110 GeV are excluded because of thelarger background from charge misidentification in Z → e

±e∓ decays.

The last bin is an overflow bin.

3









σ(pp → H±±

H∓∓)×BR(H±± → �±��±) =

Nrec(�±��±)

2×A× ε ×L, (2)







±e± and







) [GeV]±e±m(e

0 100 200 300 400 500 600

Elec

tron

pairs

/ 10

GeV

0

10

20

30

40

50


250 GeV±±LH

300 GeV±±LH

350 GeV±±LH

400 GeV±±LH

ATLAS

! -1Ldt = 4.7 fb = 7 TeVs

±e±e

(a)

) [GeV]±µ±µm(0 100 200 300 400 500 600

Muo

n pa

irs /

10 G

eV

0

5

10

15

20

25

30

35


250 GeV±±LH

300 GeV±±LH

350 GeV±±LH

400 GeV±±LH

ATLAS

! -1Ldt = 4.7 fb = 7 TeVs

±µ±µ

(b)

) [GeV]±µ±m(e

0 100 200 300 400 500 600

Lept

on p

airs

/ 10

GeV

0

10

20

30

40

50

60

70


250 GeV±±LH

300 GeV±±LH

350 GeV±±LH

400 GeV±±LH

ATLAS

! -1Ldt = 4.7 fb = 7 TeVs

±µ±e

(c)


e±, (b) µ±µ±, and (c)



±±L




±e∓ decays.


3









σ(pp → H±±

H∓∓)×BR(H±± → �±��±) =

Nrec(�±��±)

2×A× ε ×L, (2)







±e± and







) [GeV]±e±m(e

0 100 200 300 400 500 600

Elec

tron

pairs

/ 10

GeV

0

10

20

30

40

50


250 GeV±±LH

300 GeV±±LH

350 GeV±±LH

400 GeV±±LH

ATLAS

! -1Ldt = 4.7 fb = 7 TeVs

±e±e

(a)

) [GeV]±µ±µm(0 100 200 300 400 500 600

Muo

n pa

irs /

10 G

eV

0

5

10

15

20

25

30

35


250 GeV±±LH

300 GeV±±LH

350 GeV±±LH

400 GeV±±LH

ATLAS

! -1Ldt = 4.7 fb = 7 TeVs

±µ±µ

(b)

) [GeV]±µ±m(e

0 100 200 300 400 500 600

Lept

on p

airs

/ 10

GeV

0

10

20

30

40

50

60

70


250 GeV±±LH

300 GeV±±LH

350 GeV±±LH

400 GeV±±LH

ATLAS

! -1Ldt = 4.7 fb = 7 TeVs

±µ±e

(c)


e±, (b) µ±µ±, and (c)



±±L




±e∓ decays.


Backgrounds §  Prompt [yellow]

§  Dibosons (WZ, ZZ), ttbarW/Z, WWjj §  Estimated using MC simulation

§  Charge-flip (ee-channel) [dark blue] §  Opposite-sign pairs where one charge is

misidentified. §  Derive scale factor from data using same-sign Z

peak, and estimate from MC. §  Non-prompt (“fakes”) [light blue]

§  Semi-leptonic b/c-hadron decay, jets reconstructed as electrons, pion/kaon decay in flight.

§  Use data-driven technique. §  Measure fake factor using sideband regions.


Systematic Uncertainties §  Non-prompt background (Fake)

§  Limited data statistics, fake estimate : 15-100%

§  Charge-flip background §  Scale factor, MC cross section etc. : ~10%

§  Prompt background §  Lepton ID efficiency : ~3% §  MC cross section : ~10%

§  Luminosity : 3.9% §  PDF (Parton Distribution Function) uncertainty

on signal : ~2%


Data and Backgrounds


July 25, 2012 – 12 : 32 DRAFT 8

Table 7: Expected and observed numbers of pairs of isolated like-sign muon for various bins in dimuon

invariant mass, m(µµ). The uncertainties shown include statistical and systematic.

m(µµ) Bin Center [GeV] 50 150 250 360 460 560

Non-prompt 2.3 ± 0.9 1.7 ± 0.9 0.4 ± 0.4 0.1 ± 0.1 0.0+0.3−0.0

0.0+0.3−0.0

Charge flips 0+0.0−0

0+0.0−0

0+0.0−0

0+0.0−0

0+0.1−0

0+0.1−0

Prompt 9.1 ± 1.2 12.5 ± 1.6 5.7 ± 0.8 1.9 ± 0.3 0.8 ± 0.2 0.4 ± 0.1

Sum of backgrounds 11.4+1.5−1.5

14.2+1.9−1.9

6.1+0.9−0.9

1.9+0.3−0.3

0.8+0.4−0.2

0.4+0.3−0.1

Data 13 16 8 1 1 1

) [GeV]±e±m(e

0 50 100 150 200 250 300 350 400

Ele

ctro

n p

airs

/ 2

0 G

eV

0

20

40

60

80

100

120Data 2011

Non-prompt

Charge flips

Prompt

250 GeV±±H

300 GeV±±H

350 GeV±±H

400 GeV±±H

ATLAS Internal

∫ -1Ldt = 4.7 fb

= 7 TeVs

±e±e

(a)

Figure 5: Invariant mass distributions for e±e± (a), µ±µ± (b), and e±µ± (c) pairs passing the full event

selection. The data are shown as closed circles. The stacked histograms represent the backgrounds

composed of pairs of prompt leptons from SM processes, pairs with at least one non-prompt lepton, and

for the electron channels, backgrounds arising from charge misidentification and conversions. The open

histograms show the expected signal from simulated H±± samples, assuming 100% branching ratio to

the channel considered. Events in the e±e± channel with invariant masses between 70 GeV and 110 GeV

are excluded because of the large background from charge misidentification in Z → e±e∓ decays. The

last bin is an overflow bin.

Mass points used for limit setting : 50, 60, 70, 80, 90, 100, 110, 120, 130, 140, 150, 160, 170, 180, 190, 200, 210, 220, 230, 240, 250, 260, 270, 280, 290, 300, 320, 340, 360, 380, 400, 420, 440, 460, 480, 500, 520, 540, 560, 580, 600

August 13, 2012 – 10 : 18 DRAFT 11

) [GeV]±±m(H

0 200 400 600 800 1000

Effic

ien

cy×

Acc

ep

tan

ce

0

0.1

0.2

0.3

0.4

0.5

0.6

±e±e

±µ±e

±µ±µ

ATLAS Internal

∫ -1Ldt = 4.7 fb

= 7 TeVs

Figure 3: Total acceptance times efficiency (εtot in the text) vs simulated H±± mass for the e±e±, e±µ±,

and µ±µ± channels, fitted with piecewise empirical functions described in the text.

are shown in Figure 4, along with the expected contributions for H±± bosons at various masses with147

100% branching ratio to the channel in question.148

Table 10: Expected and observed numbers of like-sign electron pairs for various bins in invariant mass,

m(e±e±). The uncertainties shown include the statistical and systematic components.

Bin center [GeV] 50 150 250 360 460 560

Non-prompt 5.6 ± 3.4 2.3 ± 1.4 1.1 ± 0.7 0.3 ± 0.3 0.0+0.2−0.0

0.0+0.2−0.0

Charge flips and conversions 9.5 ± 4.3 7.6 ± 1.8 2.7+1.4−0.7

1.5 ± 0.6 0.5+1.2−0.1

0.5 ± 0.2

Prompt 4.0 ± 0.6 5.2 ± 0.8 2.2 ± 0.4 0.8 ± 0.2 0.2 ± 0.1 0.2 ± 0.1

Sum of Backgrounds 19.1 ± 5.6 15.1 ± 2.5 6.0+1.6−1.1

2.5 ± 0.7 0.7+1.2−0.1

0.6+0.3−0.2

Data 18 17 7 3 1 0

3.2 Limits on doubly charged Higgs boson production149

Since no significant discrepancy is observed between data and background estimate, the data are used

to derive upper cross-section limits on pair production of doubly charged Higgs bosons. Using the mass

bins described in Section 2.2, 95% C.L. limits are placed as function of the hypothesized H±± mass.

We here aim to constrain the H±±H∓∓ process, however, the analysis counts lepton pairs and two pairs

per event can contribute. The translation between number of events with pair produced H±± bosons and

number of lepton pairs is done as follows. The cross section for pair production of H±± bosons, σHH , is

given by:

σHH =NHH

L,

where NHH is the true number of events containing a pair of H±± bosons and L is the integrated lumi-

nosity. NHH is then related to the number of H±± bosons decaying to a certain decay channel with a

August 13, 2012 – 10 : 18 DRAFT 12

Table 11: Expected and observed numbers of like-sign muon pairs for various bins in invariant mass,

m(µ±µ±). The uncertainties shown include the statistical and systematic components.

Bin center [GeV] 50 150 250 360 460 560

Non-prompt 2.3 ± 0.9 1.7 ± 0.9 0.4 ± 0.4 0.1 ± 0.1 0.0+0.3−0.0

0.0+0.3−0.0

Charge flips 0+0.0−0

0+0.0−0

0+0.0−0

0+0.0−0

0+0.1−0

0+0.1−0

Prompt 9.1 ± 1.3 12.5 ± 1.7 5.7 ± 0.8 1.9 ± 0.3 0.8 ± 0.2 0.4 ± 0.1

Sum of backgrounds 11.4 ± 1.6 14.2 ± 1.9 6.1 ± 0.9 1.9 ± 0.3 0.8+0.4−0.2

0.4+0.3−0.1

Data 13 16 8 1 1 1

Table 12: Expected and observed numbers of like-sign eµ pairs for various bins in invariant mass,

m(e±µ±). The uncertainties shown include the statistical and systematic components.

Bin center [GeV] 50 150 250 360 460 560

Non-prompt 9.6 ± 5.0 7.44 ± 2.4 3.6 ± 3.8 1.0 ± 0.5 0.6 ± 0.5 0.1 ± 0.1

Charge flips and conversions 15.4 ± 6.8 5.1 ± 2.1 3.1 ± 1.4 0.2 ± 0.3 0.7 ± 0.5 1.2 ± 0.9

Prompt 13.2 ± 2.3 22.6 ± 3.0 9.0 ± 1.3 2.9 ± 0.5 1.5 ± 0.3 1.0 ± 0.3

Sum of Backgrounds 38.2 ± 8.6 35.1 ± 4.2 15.7 ± 4.2 4.0 ± 0.7 2.8 ± 0.8 2.3 ± 1.0

Data 31 35 14 4 5 0

ee-channel

eµ-channel

µµ-channel

Cross section §  Cross section can be calculated from the

number of events (Nrec) :

§  Product Sigma*BR(H++→l+l+) can be determined.

§  (geometrical acceptance A) *(selection efficiency ε) depends on mass and bin size.


Not

revi

ewed

,for

inte

rnal

circ

ulat

ion

only

3

The data mass distribution is shown for the e±e±,

µ±µ±, and e±µ± channels in Fig. 1 and is compared to the

background expectation and a few hypothetical H±± sig-

nals normalized to their respective cross sections (assuming

a branching ratio to the given lepton flavor of 100%). The

data show no clear peak structure and agree well with the

background estimate in all three channels.

A limit on the number of lepton pairs originating from

H±± bosons (Nrec) in each mass window is derived using

a CLs technique [28]. It is converted to a limit on the cross

section times branching ratio for doubly charged Higgs pro-

duction using the acceptance times efficiency values derived

from MC simulation. Since this analysis counts lepton pairs

and each event contains two H±± bosons, the cross section

times branching ratio for pair production is given by

σ(pp→H±±H∓∓)×BR(H±±→ !±!′±)=Nrec(!±!′±)

2×A× ε ×L, (2)

where A× ε is the acceptance times efficiency to detect a

lepton pair from H±± decay within a given mass window.

The integrated luminosity is 4.7 fb−1.

The 95% C.L. expected and observed upper limits on

the cross section times branching ratio as a function of the

H±± boson mass are shown in Figure 2. The expected limit

is determined as the median outcome of pseudo-experiments

in the absence of any signal. Also shown are the theoretical

cross sections calculated at next-to-leading order (NLO) for

H±± production for left- and right-handed couplings [16].

The uncertainty on these cross sections is ±10% due to

scale dependence in the NLO calculation, parton distribu-

tion function uncertainties, and neglecting higher order elec-

troweak corrections [29].

At low mass, the expected cross-section limits are most

stringent for the µ±µ± channel due to the low background

levels in this channel. At high mass, the e±e± and µ±µ±

limits are comparable while the e±µ± limit is about 30%

worse due to the larger background from WZ production.

In general the observed and expected limits agree well with

each other. The largest deviations of the observed limit with

respect to the expected limit are within 2σ of the uncertainty

on the expected limit.

Based on the comparison with the theoretical produc-

tion cross section, constraints can be placed on m(H±±).The lower limits on the H±± mass at 95% C.L. are listed in

Table 1 for the three final states when BR(H±± → !±!′±) =100%, as well as branching ratios of 33, 22, and 11%. For

a democratic scenario where the BR to each pair of lepton

flavors is the same, the branching ratio is 22% for the e±e±

and µ±µ± final states and 11% for the e±µ± final state. In

addition, the same mass limit can be placed on the singlet

H±± from Refs [9–11] as its production cross sections and

decay kinematics are the same as the left-handed H±±L . Fig-

ure 3 shows the mass limits as a function of the branching

ratio into the given final state.

) [GeV]±e±m(e

0 100 200 300 400 500 600

Elec

tron

pairs

/ 10

GeV

0

10

20

30

40

50


250 GeV±±LH

300 GeV±±LH

350 GeV±±LH

400 GeV±±LH

ATLAS Internal

! -1Ldt = 4.7 fb = 7 TeVs

±e±e

(a)

) [GeV]±µ±µm(0 100 200 300 400 500 600

Muo

n pa

irs /

10 G

eV

0

5

10

15

20

25

30

35


250 GeV±±LH

300 GeV±±LH

350 GeV±±LH

400 GeV±±LH

ATLAS Internal

! -1Ldt = 4.7 fb = 7 TeVs

±µ±µ

(b)

) [GeV]±µ±m(e

0 100 200 300 400 500 600

Lept

on p

airs

/ 10

GeV

0

10

20

30

40

50

60

70


250 GeV±±LH

300 GeV±±LH

350 GeV±±LH

400 GeV±±LH

ATLAS Internal

! -1Ldt = 4.7 fb = 7 TeVs

±µ±e

(c)

Fig. 1 Invariant mass distributions for e±e± (a), µ±µ± (b), and e±µ±

(c) pairs passing the full event selection. The data are shown as closedcircles. The stacked histograms represent the backgrounds composedof pairs of prompt leptons from SM processes, pairs with at least onenon-prompt lepton, and for the electron channels, backgrounds arisingfrom charge misidentification and conversions. The open histogramsshow the expected signal from simulated H±±

L samples, assuming a100% branching ratio to the decay channel considered and coupling toleft-handed fermions. Lepton pairs in the e±e± channel with an invari-ant mass between 70 GeV and 110 GeV are excluded because of thelarger background from charge misidentification in Z → e±e∓ decays.The last bin is an overflow bin.

Cross section limit


4

) [GeV]±±m(H100 200 300 400 500

) [fb

]± e

± e

! ±±

BR(

H")

##

H++

H!

(pp

$ -110

1

10

210Observed 95% CL upper limitExpected 95% CL upper limit

$ 1±Expected limit $ 2±Expected limit

)=1±e± e!±±L

), BR(H##L H++

L H!(pp $

)=1±e± e!±±R

), BR(H##R H++

R H!(pp $

ATLAS

% -1Ldt = 4.7 fb = 7 TeVs

(a)

) [GeV]±±m(H100 200 300 400 500 600

) [fb

]±µ±µ

! ±±

BR(

H")

##

H++

H!

(pp

$ -110

1

10



)=1±µ±µ !±±L

), BR(H##L H++

L H!(pp $

)=1±µ±µ !±±R

), BR(H##R H++

R H!(pp $

ATLAS

% -1Ldt = 4.7 fb = 7 TeVs

(b)

) [GeV]±±m(H100 200 300 400 500 600

) [fb

]±µ±

e!

±±

BR(

H")

##

H++

H!

(pp

$ -110

1

10



)=1±µ± e!±±L

), BR(H##L H++

L H!(pp $

)=1±µ± e!±±R

), BR(H##R H++

R H!(pp $

ATLAS

% -1Ldt = 4.7 fb = 7 TeVs

(c)

Fig. 2 Upper limit at 95% CL on the cross section times branchingratio for pair production of H

±± bosons decaying to (a) e±

e±, (b)

µ±µ±, and (c) e±µ± pairs. The observed and median expected limits

are shown along with the 1σ and 2σ variations in the expected limits.In the range 70<m(H±±)< 110 GeV, no limit is set in the e

±e± chan-

nel. Also shown are the theoretical predictions at next-to-leading orderfor the pp → H

±±H

∓∓ cross section for H±±L

and H±±R

bosons. Thevariation from bin to bin in the expected limits is due to fluctuations inthe background yields derived from small MC samples.

Table 1 Lower mass limits at 95% CL on H±± bosons decaying to

e±

e±, µ±µ±, or e

±µ± pairs. Mass limits are derived assuming branch-ing ratios to a given decay mode of 100%, 33%, 22%, or 11%. Bothexpected and observed limits are given.

BR(H±±L

→ �±��±) 95% CL lower limit on m(H±±L

) [GeV]

e±

e± µ±µ±

e±µ±

exp. obs. exp. obs. exp. obs.100% 407 409 401 398 392 37533% 318 317 317 290 279 27622% 274 258 282 282 250 25311% 228 212 234 216 206 190

BR(H±±R

→ �±��±) 95% CL lower limit on m(H±±R

) [GeV]

e±

e± µ±µ±

e±µ±


e±

e± and µ±µ± final states and 11% for the e

±µ± finalstate. In addition, the same mass limits can be placed on thesinglet H

±± in the Zee-Babu model as its production crosssections and decay kinematics are the same as for H

±±L

. Fig-ure 3 shows the mass limits as a function of the branchingratio into each of the three final states.

In conclusion, a search for doubly-charged Higgs bosonsdecaying to e

±e±, e

±µ±, or µ±µ± has been performed bysearching for a narrow resonance peak in the dilepton massdistribution. No such peak was observed in a data samplecorresponding to an integrated luminosity of 4.7 fb−1 of pp

collisions at√

s = 7 TeV recorded by the ATLAS detectorat the LHC in 2011. Cross-section limits between 17 fb and0.6 fb are set depending on the mass of the H

±± boson andthe final state. Assuming pair production, couplings to left-handed fermions, and a branching ratio of 100% for eachfinal state, masses below 409 GeV, 398 GeV, and 375 GeVare excluded at 95% CL for e


±µ± finalstates, respectively. Lower mass limits are also set for sce-narios with right-handed couplings or smaller branching ra-tios. The limits on H

±±L

bosons also apply to the singlet inthe Zee-Babu model.

Acknowledgements We thank CERN for the very successful oper-ation of the LHC, as well as the support staff from our institutionswithout whom ATLAS could not be operated efficiently.

We acknowledge the support of ANPCyT, Argentina; YerPhI, Ar-menia; ARC, Australia; BMWF and FWF, Austria; ANAS, Azerbai-jan; SSTC, Belarus; CNPq and FAPESP, Brazil; NSERC, NRC andCFI, Canada; CERN; CONICYT, Chile; CAS, MOST and NSFC,China; COLCIENCIAS, Colombia; MSMT CR, MPO CR and VSCCR, Czech Republic; DNRF, DNSRC and Lundbeck Foundation, Den-mark; EPLANET and ERC, European Union; IN2P3-CNRS, CEA-DSM/IRFU, France; GNSF, Georgia; BMBF, DFG, HGF, MPG and

4

) [GeV]±±m(H100 200 300 400 500

) [fb

]± e

± e

! ±±

BR(

H")

##

H++

H!

(pp

$ -110

1

10



)=1±e± e!±±L

), BR(H##L H++

L H!(pp $

)=1±e± e!±±R

), BR(H##R H++

R H!(pp $

ATLAS

% -1Ldt = 4.7 fb = 7 TeVs

(a)

) [GeV]±±m(H100 200 300 400 500 600

) [fb

]±µ±µ

! ±±

BR(

H")

##

H++

H!

(pp

$ -110

1

10



)=1±µ±µ !±±L

), BR(H##L H++

L H!(pp $

)=1±µ±µ !±±R

), BR(H##R H++

R H!(pp $

ATLAS

% -1Ldt = 4.7 fb = 7 TeVs

(b)

) [GeV]±±m(H100 200 300 400 500 600

) [fb

]±µ±

e!

±±

BR(

H")

##

H++

H!

(pp

$ -110

1

10



)=1±µ± e!±±L

), BR(H##L H++

L H!(pp $

)=1±µ± e!±±R

), BR(H##R H++

R H!(pp $

ATLAS

% -1Ldt = 4.7 fb = 7 TeVs

(c)



e±, (b)



±e± chan-


±±H


and H±±R



e±

e±, µ±µ±, or e


BR(H±±L


) [GeV]

e±

e± µ±µ±

e±µ±


BR(H±±R


) [GeV]

e±

e± µ±µ±

e±µ±


e±




±±L



±e±, e


collisions at√





±±L




4

) [GeV]±±m(H100 200 300 400 500

) [fb

]± e

± e

! ±±

BR(

H")

##

H++

H!

(pp

$ -110

1

10



)=1±e± e!±±L

), BR(H##L H++

L H!(pp $

)=1±e± e!±±R

), BR(H##R H++

R H!(pp $

ATLAS

% -1Ldt = 4.7 fb = 7 TeVs

(a)

) [GeV]±±m(H100 200 300 400 500 600

) [fb

]±µ±µ

! ±±

BR(

H")

##

H++

H!

(pp

$ -110

1

10



)=1±µ±µ !±±L

), BR(H##L H++

L H!(pp $

)=1±µ±µ !±±R

), BR(H##R H++

R H!(pp $

ATLAS

% -1Ldt = 4.7 fb = 7 TeVs

(b)

) [GeV]±±m(H100 200 300 400 500 600

) [fb

]±µ±

e!

±±

BR(

H")

##

H++

H!

(pp

$ -110

1

10



)=1±µ± e!±±L

), BR(H##L H++

L H!(pp $

)=1±µ± e!±±R

), BR(H##R H++

R H!(pp $

ATLAS

% -1Ldt = 4.7 fb = 7 TeVs

(c)



e±, (b)



±e± chan-


±±H


and H±±R



e±

e±, µ±µ±, or e


BR(H±±L


) [GeV]

e±

e± µ±µ±

e±µ±


BR(H±±R


) [GeV]

e±

e± µ±µ±

e±µ±


e±




±±L



±e±, e


collisions at√





±±L




§  Calculate 95% CL cross section limit.

→ set mass limit.

ee channel emu channel

mumu channel

Mass Limit vs BF


5

) [GeV]±±Lm(H

100 150 200 250 300 350 400 450

)± l'

± l!

±± L

BR(H

00.10.20.30.40.50.60.70.80.9

1±e±Observed limit: e±e±Expected limit: e±µ±µObserved limit: ±µ±µExpected limit: ±µ±Observed limit: e±µ±Expected limit: e

ATLAS

" -1Ldt = 4.7 fb = 7 TeVs

(a)

) [GeV]±±Rm(H

100 150 200 250 300 350

)± l'

± l!

±± R

BR(H

00.10.20.30.40.50.60.70.80.9

1±e±Observed limit: e±e±Expected limit: e±µ±µObserved limit: ±µ±µExpected limit: ±µ±Observed limit: e±µ±Expected limit: e

ATLAS

" -1Ldt = 4.7 fb = 7 TeVs

(b)

Fig. 3 The mass limits as a function of the branching ratio for the H±±

decaying to e±

e±, e

±µ±, and µ±µ± for (a) H±±L

and (b) H±±R

bosons.Shown are both the observed limits (solid lines) and the expected limits(dashed lines). The stepping behaviour, where the same mass limit isvalid for a range of branching ratios, results from fluctuations in theobserved cross-section limits shown in Fig. 2.

AvH Foundation, Germany; GSRT, Greece; ISF, MINERVA, GIF,DIP and Benoziyo Center, Israel; INFN, Italy; MEXT and JSPS,Japan; CNRST, Morocco; FOM and NWO, Netherlands; BRF andRCN, Norway; MNiSW, Poland; GRICES and FCT, Portugal; MERYS(MECTS), Romania; MES of Russia and ROSATOM, Russian Federa-tion; JINR; MSTD, Serbia; MSSR, Slovakia; ARRS and MVZT, Slove-nia; DST/NRF, South Africa; MICINN, Spain; SRC and WallenbergFoundation, Sweden; SER, SNSF and Cantons of Bern and Geneva,Switzerland; NSC, Taiwan; TAEK, Turkey; STFC, the Royal Societyand Leverhulme Trust, United Kingdom; DOE and NSF, United Statesof America.

The crucial computing support from all WLCG partners is ac-knowledged gratefully, in particular from CERN and the ATLAS Tier-1facilities at TRIUMF (Canada), NDGF (Denmark, Norway, Sweden),CC-IN2P3 (France), KIT/GridKA (Germany), INFN-CNAF (Italy),NL-T1 (Netherlands), PIC (Spain), ASGC (Taiwan), RAL (UK) andBNL (USA) and in the Tier-2 facilities worldwide.

References

1. H. Georgi, M. Machacek, Nucl. Phys. B 262, 463(1985)

2. J.F. Gunion, R. Vega, J. Wudka, Phys. Rev. D 42, 1673(1990)

3. J.E. Cieza Montalvo, N.V. Cortez, J. Sa Borges, M.D.Tonasse, Nucl. Phys. B 756, 1 (2006); erratum ibid. 796422 (2008)

4. N. Arkani-Hamed, A.G. Cohen, E. Katz, A.E. Nelson,T. Gregoire, J.G. Wacker, JHEP 0208, 021 (2002)

5. M. Magg, C. Wetterich, Phys. Lett. B 94, 61 (1980)6. J. Schechter, J.W.F. Valle, Phys. Rev. D 22, 2227 (1980)7. G. Lazarides, Q. Shafi, C. Wetterich, Nucl. Phys. B 181,

287 (1981)8. R.N. Mohapatra, G. Senjanovic, Phys. Rev. D 23, 165

(1981)9. A. Zee, Nucl. Phys. B 264, 99 (1986)

10. K.S. Babu, Phys. Lett. B 203, 132 (1988)11. M. Nebot, J.F. Oliver, D. Palao, A. Santamaria, Phys.

Rev. D 77, 093013 (2008)12. J.C. Pati, A. Salam, Phys. Rev. D 10, 275 (1974)13. R.N. Mohapatra, J.C. Pati, Phys. Rev. D 11, 566 (1975)14. G. Senjanovic, R.N. Mohapatra, Phys. Rev. D 12, 1502

(1975)15. T. Rizzo, Phys. Rev. D 25, 1355; addendum ibid. 27,

(2983) 657 (1982)16. M. Mühlleitner and M. Spira, Phys. Rev. D 68, 117701

(2003)17. J. Beringer et al. (Particle Data Group), Phys. Rev. D

86, 010001 (2012)18. ATLAS Collaboration, Phys. Rev. D 85, 032004 (2012)19. CMS Collaboration (2012). arXiv:1207.2666 (submit-

ted to Eur. Phys. J. C)20. SINDRUM Collaboration, Nucl. Phys. B 299, 1 (1988)21. V. Rentala, W. Shepherd, S. Su, Phys. Rev. D 84,

035004 (2011)22. ATLAS Collaboration, Eur. Phys. J C 71, 1630 (2011)23. ATLAS Collaboration, (2011). ATLAS-CONF-2011-

11624. ATLAS Collaboration, JINST 3, S08003 (2008)25. ATLAS Collaboration. arXiv:1210.4538 (submitted to

JHEP) (2012)26. ATLAS Collaboration, Eur. Phys. J C 72, 1909 (2012)27. T. Sjöstrand, S. Mrenna, P. Skands, JHEP 05, 026

(2006)28. A. Martin, W. Stirling, R.S. Thorne, G. Watt, Eur. Phys.

J. C 63, 189 (2009)29. A. L. Read, J. Phys. G 28, 2693 (2002)

4

) [GeV]±±m(H100 200 300 400 500

) [fb

]± e

± e

! ±±

BR(

H")

##

H++

H!

(pp

$ -110

1

10



)=1±e± e!±±L

), BR(H##L H++

L H!(pp $

)=1±e± e!±±R

), BR(H##R H++

R H!(pp $

ATLAS

% -1Ldt = 4.7 fb = 7 TeVs

(a)

) [GeV]±±m(H100 200 300 400 500 600

) [fb

]±µ±µ

! ±±

BR(

H")

##

H++

H!

(pp

$ -110

1

10



)=1±µ±µ !±±L

), BR(H##L H++

L H!(pp $

)=1±µ±µ !±±R

), BR(H##R H++

R H!(pp $

ATLAS

% -1Ldt = 4.7 fb = 7 TeVs

(b)

) [GeV]±±m(H100 200 300 400 500 600

) [fb

]±µ±

e!

±±

BR(

H")

##

H++

H!

(pp

$ -110

1

10



)=1±µ± e!±±L

), BR(H##L H++

L H!(pp $

)=1±µ± e!±±R

), BR(H##R H++

R H!(pp $

ATLAS

% -1Ldt = 4.7 fb = 7 TeVs

(c)



e±, (b)



±e± chan-


±±H


and H±±R



e±

e±, µ±µ±, or e


BR(H±±L


) [GeV]

e±

e± µ±µ±

e±µ±


BR(H±±R


) [GeV]

e±

e± µ±µ±

e±µ±


e±




±±L



±e±, e


collisions at√





±±L




Mass > 409 GeV

Type III : Triplet (N+) Search §  We explicitly reconstruct N+ (Z(ll)+l). §  Require one more lepton to reduce

background. 　→ 4-lepton final states.


October 23, 2012 – 15 : 31 DRAFT 3

1.2 Type III Seesaw phenomenology84

The Type III Seesaw model manifests with distinct experimental signatures at the LHC through direct85

production of N0 and N± . Although the both the N0 and the N± have many decay modes, by far the86

cleanest search is in fully leptonic channel. One such decay is shown in Figure 1. By identifying and87

reconstructing all leptons associated with the decay of the N± it is possible to perform a search for a88

narrow resonance on virtually no background. Below is a complete list of decay modes relevant for such89

a study at the LHC.90

• pp→ N0 + N± → l±W∓ + l±Z91

• pp→ N+ + N− → l−Z + l+Z92

• pp→ N+ + N− → l+Z + νlW−

93

• pp→ N+ + N− → νlW+ + l−Z94

• pp→ N0 + N± → νlZ + νlW±95

p

p

W+

N0

N+Z

!+

!−

!+

!−

W+

Figure 1: One of the cleanest decay channels for the discovery of the Type III Seesaw model.

In the analysis of the Type III seesaw model at the LHC we consider maximal mixing angles of the

new heavy fermions to electrons and muons; the mixing angle for tau leptons is taken to be zero and will

be the subject of future studies. The presence of multiple mixing angles will alter the branching ratios of

the N0/± and have minimal effect of the production cross section of the new heavy states. Thus the limits

produced within this analysis may be interpreted for a range of mixing angles and Yukawa couplings.

References [2–4] derive in detail the following bounds on the allowed combinations of mixing angles for

the Type III Seesaw model.

|Ve| < 5.5 × 10−2 (1)

|Vµ| < 6.3 × 10−2 (2)

|Vτ| < 6.3 × 10−2 (3)

|VeVµ| < 1.7 × 10−7 (4)

|VeVτ| < 4.2 × 10−4 (5)

|VµVτ| < 4.9 × 10−4 (6)

Event selection §  Select isolated electrons and muons. §  Event level selections:

25/02/2013 CONF note approval, K. Hamano 18

4 or more leptons 4 or more isolated leptons Z candidate Opposite sign same flavor lepton pair.

|mZ(ll) – mZ(PDG)| < 10 GeV

3rd lepton Closest in φ to the reconstructed Z 4th lepton Highest pT lepton remaining in the event Second Z veto Reject events with a second Z candidate

Backgrounds §  Main background is ZZ §  Small backgrounds:

§  VVV (WWW*, ZWW* and ZZZ*) §  ttbarV (ttbarW, ttbarW*, ttbarZ, ttbarZ*) §  Z+jets

§  All backgrounds are estimated from MC.


Systematic uncertainties


Not

revi

ewed

,for

inte

rnal

circ

ulat

ion

only

February 16, 2013 – 10 : 43 DRAFT 6

Table 3: Summary of the systematic uncertainties on the normalisation of the signal (mN = 120 GeV)and background contributions, given in percentages.. The entries filled with a hyphen correspond to un-certainties that are consistent with zero or to systematic variations that are not applicable to the respectivesample.

Wγ ZZ VVV Z+jets ttV Signal [120 GeV]Ee Resolution − 0.2 < 0.1 − < 0.1 0.3Ee Scale − 0.1 0.3 − 0.6 0.6e Identification 2.9 2.7 2.8 2.8 2.7 2.7µ Res. ID − 0.1 < 0.1 − 1.7 0.1µ Res. Spectr. − 0.1 < 0.1 − 1.7 0.1Eµ Scale − < 0.1 < 0.1 − 5.8 0.2Shape − − − 100 − −Scale Factor − − − 370 − −Fast sim. − − − − − 6.8Signal PDF − − − − − 0.9Cross Section 30 6.4 100 11 50 −Total 30 7.0 100 390 50 7.4

independent of MZ(��)�. All systematic uncertainties are added in quadrature.198

5 Results199

This analysis is sensitive to events with four isolated leptons, arising from pp→ N±N0, with subsequent200

decays N± → Z�±, Z → �+�− and N0 → W±�∓. This decay mode dominates the contribution to201

this topology in the Type III Seesaw model. The total expected background in the signal region is202

19.8 ± 1.3(stat.) ± 5.4(sys.) events. The distribution of mZ(��)� in the signal region, using the sideband203

estimated Z+jets background is shown in Figure 2. The shape and normalisation are in agreement with204

the SM expectation.205

Limits are then placed on the product of the cross section and branching fractions:206

σ(pp→ N±N0)B(N± → Z�±)B(N0 → W±�∓) =Nevt

Aε�Ldt

(1)

where Nevt is the number of events, A is the acceptance, ε is the selection efficiency, which includes207

B(Z → �+�−) in the mass window, and�Ldt is the integrated luminosity of the data sample. Values of208

A × ε for different values of mN are listed in Table 4.209

The cross section times branching fraction limit is calculated as a function of the heavy lepton mass210

mN , within mass windows. The optimal width of these windows depends on the mass of the heavy lepton211

under study, increasing in width for higher masses. Limits are determined at 16 mass points between 110212

GeV to 260 GeV with ± 10 GeV mass windows to account for detector resolution, and 16 mass points213

between 280 GeV and 540 GeV with ±20 GeV mass windows. The number of expected and observed214

events after all selection requirements are given for a few representative mass windows in Table 5.215

No significant excess of events is observed relative to the SM expectations. Therefore an upper limit216

at the 95% confidence level is set on σ × B, on the production cross section of N0N+ and its branching217

fraction, B(N± → Z�±)B(N0 → W±�∓), to four lepton final states, via N± → Z�± and N0 → W±�∓,218

•  Main sources •  MC cross section uncertainties. •  Z+jets background shape = 100% •  Z+jets background rate (scale factor):

•  One event survived the selection cuts in the signal region. •  Poisson 95% upper bound of 1 event is 4.74 events -> 370%.

Results §  Estimated background in the signal region:

§  Observed events : 19


Not

revi

ewed

,for

inte

rnal

circ

ulat

ion

only

February 16, 2013 – 10 : 43 DRAFT 6




5 Results199









Aε�Ldt

(1)













0.2

Results for limit calculation §  Separated into mass bins for limit setting:

§  10 GeV (20 GeV) bins for 110-260 (280-540) GeV.

§  Signal efficiency for each mass bin :


February 20, 2013 – 12 : 05 DRAFT 50

Mass Mass window

(GeV) (GeV)

110 [100,120]

120 [110,130]

130 [120,140]

140 [130,150]

150 [140,160]

160 [150,170]

170 [160,180]

180 [170,190]

190 [180,200]

200 [190,210]

210 [200,220]

220 [210,230]

230 [220,240]

240 [230,250]

250 [240,260]

260 [250,270]

280 [260,300]

300 [280,320]

320 [300,340]

340 [320,360]

360 [340,380]

380 [360,400]

400 [380,420]

420 [400,440]

440 [420,460]

460 [440,480]

480 [460,500]

500 [480,520]

520 [500,540]

540 [520,560]

Table 23: Mass points and windows/bins.

mN bin Expected Events Observed Signal Outside of mass bin

(GeV) ZZ Z+jets VVV ttV Total Events Events Expected Observed Signal

120 5.0 ± 0.4 0.4 ± 1.7 0.05 ± 0.05 0.16 ± 0.08 5.7 ± 1.7 4 17.8 ± 1.2 14.5 ± 1.3 15 2.8 ± 0.5

160 2.2 ± 0.2 0.1 ± 0.4 0.05 ± 0.05 0.10 ± 0.05 2.5 ± 0.5 3 7.6 ± 0.5 16.1 ± 0.2 16 2.9 ± 0.3

200 0.85 ± 0.07 0.04 ± 0.15 0.03 ± 0.03 0.05 ± 0.03 1.0 ± 0.2 1 1.0 ± 0.1 18.9 ± 1.3 18 0.6 ± 0.1

300 0.28 ± 0.03 0.04 ± 0.15 0.02 ± 0.02 0.04 ± 0.02 0.4 ± 0.2 0 0.25 ± 0.02 19.4 ± 1.3 19 0.19 ± 0.02

500 0.03 ± 0.01 0.00 ± 0.04 0.004 ± 0.004 0.006 ± 0.003 0.04 ± 0.04 0 0.015 ± 0.002 19.7 ± 1.3 19 0.059 ± 0.005

Table 24: Expected and observed events in and outside of each mass bin.

Not

revi

ewed

,for

inte

rnal

circ

ulat

ion

only

February 16, 2013 – 10 : 43 DRAFT 7

[GeV]Z(ll)lm

100 150 200 250 300 350 400 450 500

Eve

nts

/ 2

0 G

eV

-210

-110

1

10

210 Data stat)⊕ Bkg (sys WW ZZ

γ W VVV

Vt t Single Top Z+jets W+jets N [200 GeV] N [300 GeV]

ATLAS

Internal-1 L dt = 5.8 fb∫ = 8 TeV, s

Figure 2: Invariant mass, Z(��)� of thee N± candidates in the signal region, for data (black points), and

the expected total background (solid histograms). The rightmost bins in the histograms include overflow

events.

Table 4: The product of the signal acceptance and efficiency (including B(Z → ��)) for each simulated

mass point.

mN [ GeV] Aε100 0.0006

120 0.0091

160 0.012

200 0.011

300 0.010

500 0.006

Table 5: The expected and observed event yields inside and outside of each mass bin, where the errors

include both statistical systematic uncertainties added in quadrature.

mN Expected Events (Inside signal mass bin) Observed Signal Outside signal mass bin

(GeV) ZZ Z+jets VVV ttV Total Events Events Expected Observed Signal only

120 5.0 ± 0.4 0.4 ± 1.7 0.05 ± 0.02 0.16 ± 0.08 5.7 ± 1.7 4 17.8 ± 1.2 14.5 ± 1.3 15 2.8 ± 0.5

160 2.2 ± 0.2 0.1 ± 0.4 0.05 ± 0.02 0.10 ± 0.05 2.5 ± 0.5 3 7.6 ± 0.5 16.1 ± 0.2 16 2.9 ± 0.3

200 0.85 ± 0.07 0.04 ± 0.15 0.03 ± 0.02 0.05 ± 0.03 1.0 ± 0.2 1 1.0 ± 0.1 18.9 ± 1.3 18 0.6 ± 0.1

300 0.28 ± 0.03 0.04 ± 0.15 0.02 ± 0.01 0.04 ± 0.02 0.4 ± 0.2 0 0.25 ± 0.02 19.4 ± 1.3 19 0.19 ± 0.02

500 0.03 ± 0.01 0.00 ± 0.04 0.002 ± 0.001 0.005 ± 0.005 0.04 ± 0.04 0 0.015 ± 0.002 19.7 ± 1.3 19 0.059 ± 0.005

Cross section and Mass limits §  95% CL limit on number of events were

determined by frequentist CLs method. §  Cross section is


Not

revi

ewed

,for

inte

rnal

circ

ulat

ion

only

February 16, 2013 – 10 : 43 DRAFT 6




5 Results199









Aε�Ldt

(1)













BF from theory

Constant BF=1

(BF(Z->ll) is a part of Ae)

•  Expected limit : 243+/-5 GeV •  Observed limit : 245 GeV

Conclusion §  There are two main neutrino mass generation

mechanisms : seesaw and loop (Zee-Babu model). §  Type II seesaw and Z-B can be probed by doubly

charged scalar search (4.7fb-1, 7TeV).

§  Type III seesaw can be probed by heavy lepton (N) search (5.8fb-1, 8TeV).


Mass limit ee µµ eµ ATLAS 409 GeV 398 GeV 375 GeV CMS 382 GeV 395 GeV 391 GeV

Mass limit ATLAS 245 GeV (4-lepton search) CMS 211 GeV (3-lepton search)

(ArXiv:1207.2666)

(7TeV, 4.9fb-1, ArXiV:1210.1797)

ZZ background §  ZZ MC is used in the signal region. §  ZZ enhanced region to compare with data.

§  3rd and 4th lepton satisfy Z candidate condition (Same-sign, opposite-flavor and in Z mass window).


•  Both rate and shape are well modeled

Z+jets background §  Rate was given by MC in the signal region §  Shape was taken from Z+jets enhanced region.

§  3 or more lepton, looser IP and isolation cuts, isolation of 3rd lepton was reversed.

§  Good data and MC agreement: §  Z+jets background is well modeled by MC.


Mass limits vs BF


Not

revi

ewed

,for

inte

rnal

circ

ulat

ion

only

February 16, 2013 – 10 : 43 DRAFT 8

Table 6: The expected and observed limits on mN obtained using the frequentist method, specified at a

95% confidence level, for six assumed sets of branching fractions.

Mass limit [GeV]

B(N± → Z�±)B(N0 → W±�∓) Expected Observed

1.00 350 ± 5 350

0.66 326 ± 5 330

0.33 284 ± 6 290

0.25 266 ± 7 280

0.15 240 ± 6 230

where the lepton can be an electron or a muon. The branching fractions of N0 → W±�∓ and N± → Z�±219

depend on mN [6].220

The frequentist CLs method is used to determine 95% CL upper limits on the number of events for221

each mass bin [41]. The upper limits on σB(N± → Z�±)B(N0 → W±�∓) are determined as a function of222

the fermion mass mN , shown in Figure 3. Two theoretical cross section curves are overlaid, corresponding223

to maximal and nominal values of the product of branching fractions, B(N± → Z�±)B(N0 → W±�∓), 1224

and 0.15 respectively. The nominal value of 0.15 is the median predicted branching frction in the mass225

range covered by this search [6]. The mass limit on mN is also determined as a function of the product226

of branching fractions B(N± → Z�±)B(N0 → W±�∓), shown in Figure 4. In Table 6, lower limits on the227

mass are given for several values of B(N± → Z�±)B(N0 → W±�∓).228

AssumingB(N± → Z�±)B(N0 → W±�∓) = 1, a limit of mN > 350±5 GeV is expected: the observed229

limit is mN > 350 GeV. When B(N± → Z�±)B(N0 → W±�∓) is set to 0.15 the expected limit is reduced230

to mN > 240 ± 6 GeV, and the observed limit becomes mN > 230 GeV.231

This analysis inclusively searches for heavy leptons decaying into electron and muon final states.232

As the electron and muon reconstruction efficiencies are similar, the limit therefore holds true for cases233

where the electron and muon couplings are similar in size. In the case where the tau coupling is non-234

negligible, the mass limit from this measurement will be reduced.235

[GeV]Nm

150 200 250 300 350 400 450 500

W l)

[fb

]→

0 B

F(N

×)± Z

l→ ±

BF

(N×)±

N0

N→

(pp

σ

210

310Observed 95% C.L. upper limit

Expected 95% C.L. upper limit

σ 1±Expected limit

σ 2±Expected limit

W l)=1→0 BF(N×)± Z l→±),BF(N±N

0 N→(pp σ

W l)=0.15→0 BF(N×)± Z l→±),BF(N±N

0 N→(pp σ

ATLAS Internal

∫ ,-1Ldt = 5.8 fb = 8 TeVs

|=0.055e

|V

|=0.063µ|V

|=0τ|V

Figure 3: The expected (dashed line) and observed (solid line) exclusion limits at 95% confidence level

on σB as a function of the fermion mass mN assuming |Ve| = 0.055, |Vµ| = 0.063 and |Vτ| = 0 and

for B=1 and 0.15 (theory LO). The dark(green) and light (yellow) shaded areas represent the 1 standard

deviation (68% C.L.) and 2 standard deviations (95% CL) limits on the expected, respectively.

•  Mass limits are calculated assuming constant BF values.

•  Other models may give different decay BF and different efficiencies.

•  Corresponding mass limit can be determined from the plot.

Possible application to other models

search for neutrino mass generation mechanisms with atlas

Documents