Eu . Phys. J. C (2016) 76:664
DOI 10.1140/epjc/s10052-016-4489-7
Regula A icle - Expe imen al Physics
Sea ch o Higgs-like bosons decaying in o long-li ed exo ic
pa icles
LHCb Collabo a ion
CERN, 1211 Gene a 23, Swi ze land
Recei ed: 13 Sep embe 2016 / Accep ed: 7 No embe 2016 / Published online: 2 Decembe 2016
© CERN o he bene i o he LHCb collabo a ion 2016. This a icle is published wi h open access a Sp inge link.com
Abs ac A sea ch is p esen ed o massi e long-li ed
pa icles, in he 20–60 GeV/c2mass ange wi h li e imes
be ween 5 and 100ps. The da ase used co esponds o
0.62 b−1o p o on-p o on collision da a collec ed by he
LHCb de ec o a √s=7 TeV. The pa icles a e assumed
o be pai -p oduced by he decay o a Higgs-like boson wi h
mass be ween 80 and 140GeV/c2. No excess abo e he back-
g ound expec a ion is obse ed and limi s a e se on he p o-
duc ion c oss-sec ion as a unc ion o he long-li ed pa icle
mass and li e ime and o he Higgs-like boson mass.
1 In oduc ion
The s anda d model o pa icle physics (SM) has shown g ea
success in desc ibing physics p ocesses a e y sho dis-
ances. Ne e heless, open ques ions emain, such as he hie -
a chy p oblem, he imp ecise uni ica ion o gauge couplings,
and he absence o candida es o da k ma e . Conside able
e o s ha e been made o add ess hese issues, esul ing in
a la ge a ie y o models. Supe symme y (SUSY), in which
he s ong and elec oweak o ces a e uni ied a a eno mali-
sa ion scale nea he Planck scale, p o ides a possible solu-
ion o he hie a chy p oblem; he minimal supe symme ic
s anda d model (MSSM) is he simples , phenomenologi-
cally iable ealisa ion o SUSY [1,2].
The p esen s udy ocuses on a subse o models ea u -
ing massi e long-li ed pa icles (LLP) wi h a measu able
ligh dis ance. We concen a e on scena ios in which he
LLP decays had onically in he LHCb e ex de ec o , a -
elling dis ances which can be la ge han hose o ypical b
had ons.
A la ge numbe o LLP sea ches ha e been pe o med
by he expe imen s a he LHC and Te a on, mainly using
he Hidden Valley amewo k [3] as a benchma k model [4–
8]. Hidden Valley p ocesses ha e also been sough by
LHCb [9], which is able o explo e he o wa d apidi y egion
only pa ially co e ed by o he LHC expe imen s. In addi-
e-mail: [email p o ec ed]
ion, i is able o igge on pa icles wi h low ans e se
momen a, allowing he expe imen o p obe ela i ely small
LLP masses.
The e en opology conside ed in his s udy is qui e di e -
en om ha o Hidden Valley models. The minimal supe -
g a i y model (mSUGRA) ealisa ion o he MSSM is used
as a benchma k model wi h ba yon numbe iola ion [10],
as sugges ed in Re s. [11,12]. He e a Higgs-like boson p o-
duced in pp collisions decays in o wo LLPs (neu alinos),
subsequen ly decaying in o h ee qua ks each. The Higgs-
like pa icle mass anges om 80 up o 140GeV/c2, co e -
ing he mass o he scala boson disco e ed by he ATLAS
and CMS expe imen s [13,14]. The explo ed LLP li e ime
ange o 5–100ps is highe han he ypical bhad on li e ime,
and co esponds o an a e age ligh dis ance o up o 30cm,
which is inside he LHCb e ex de ec o egion. The LLP
mass ange conside ed is be ween 20 and 60GeV/c2.
2 De ec o desc ip ion
The LHCb de ec o [15,16] is a single-a m o wa d spec-
ome e co e ing he pseudo apidi y ange 2 <η<5,
designed o he s udy o pa icles con aining bo cqua ks.
The de ec o includes a high-p ecision acking sys em con-
sis ing o a silicon-s ip e ex de ec o su ounding he pp
in e ac ion egion (VELO), a la ge-a ea silicon-s ip de ec o
loca ed ups eam o a dipole magne wi h a bending powe
o abou 4 Tm, and h ee s a ions o silicon-s ip de ec o s
and s aw d i ubes, placed downs eam o he magne . The
acking sys em p o ides a measu emen o he momen um,
p, o cha ged pa icles wi h a ela i e unce ain y ha a ies
om 0.5% a low momen um o 1.0% a 200GeV/c. The min-
imum dis ance o a ack o a p ima y e ex (PV), he impac
pa ame e , is measu ed wi h a esolu ion o (15 +29/pT)
µm, whe e pTis he componen o he momen um ans e se
o he beam, inGeV/c. Di e en ypes o cha ged had ons
a e dis inguished using in o ma ion om wo ing-imaging
Che enko de ec o s. Pho ons, elec ons and had ons a e
123
664 Page 2 o 15 Eu . Phys. J. C (2016) 76 :664
iden i ied by a calo ime e sys em consis ing o scin illa ing-
pad and p eshowe de ec o s, an elec omagne ic calo ime e
and a had onic calo ime e . Muons a e iden i ied by a sys em
composed o al e na ing laye s o i on and mul iwi e p opo -
ional chambe s. The online e en selec ion is pe o med by
a igge [17], which consis s o a ha dwa e s age, L0, based
on in o ma ion om he calo ime e and muon sys ems, ol-
lowed by wo so wa e s ages, HLT1 and HLT2, which un a
simpli ied e sion o he o line e en econs uc ion.
3 E en gene a ion and de ec o simula ion
Va ious simula ed e en samples a e used in his analysis.
The pp collisions a e gene a ed wi h Py hia 6[18]. The p o-
cess simula ed is h0→˜χ0
1˜χ0
1, whe e he Higgs-like boson o
mass mh0is p oduced ia gluon-gluon usion, wi h he pa on
densi y unc ion aken om CTEQ6L [19]. The neu alino
˜χ0
1is an LLP o mass mLLP and li e ime τLLP, which decays
in o h ee qua ks ia he mSUGRA ba yon numbe iola -
ing p ocess a ailable in Py hia. The co esponding decay
la ou s uc u e o he neu alino wi h a mass o 48 GeV/c2
is 18.5% o each o he combina ions wi h a bqua k (udb,
usb,cdb,csb), and 13% o each udq and cdq, whe e qis
no a bqua k, i.e. abou 75% o LLPs ha e a bqua k in he
decay. This ac ion becomes 70% o mLLP =20 GeV/c2.
Two sepa a e de ec o simula ions a e used, a ull simula-
ion whe e he in e ac ion o he gene a ed pa icles wi h he
de ec o is based on Gean 4 [20,21], and a as simula ion.
In Gean 4, he de ec o and i s esponse a e implemen ed
as desc ibed in Re . [22]. Signal models o a ep esen a i e
se o heo e ical pa ame e s ha e been gene a ed and ully
simula ed (Appendix A, Table 5). In he emainde o his
pape , he ollowing nomencla u e is chosen: a p e ix “BV”,
indica ing ba yon numbe iola ion, is ollowed by he LLP
mass in GeV/c2and li e ime, and he p e ix “mH” ollowed
by he mh0 alue in GeV/c2. Mos o he ully simula ed
models ha e mh0=114GeV/c2, which is in he middle o he
chosen Higgs-like pa icle mass ange. Only e en s wi h a
leas one ˜χ0
1in he pseudo apidi y egion 1.8<η<5.0a e
p ocessed by Gean 4, co esponding o abou 30% o he
gene a ed e en s.
The as simula ion is used o co e a b oade pa ame-
e space o he heo e ical models. He e he cha ged pa i-
cles om he h0→˜χ0
1˜χ0
1p ocess alling in he geome ical
accep ance o he de ec o a e p ocessed by he e ex econ-
s uc ion algo i hm. The as simula ion is alida ed by com-
pa ison wi h he ull simula ion. The de ec ion e iciencies
p edic ed by he ull and he as simula ion di e by less
han 5% o all he signal models. The dis ibu ions o mass,
momen um and ans e se momen um o he econs uc ed
LLP, and o he econs uc ed e ex posi ion coincide.
E en s wi h di ec p oduc ion o cha m, bo om and op
qua ks a e conside ed as sou ces o backg ound. Samples o
such e en s we e p oduced and ully simula ed. In pa icula ,
17 ×106inclusi e bb e en s (9 ×106inclusi e cc e en s)
we e p oduced wi h a leas wo bhad ons (chad ons) in
1.5<η<5.0, and hal a million e en s wi h a leas one
muon in he accep ance.
4 E en selec ion and signal de e mina ion
This analysis sea ches o e en s wi h pai s o displaced high-
mul iplici y e ices. The main backg ound is due o sec-
onda y in e ac ions o pa icles wi h he de ec o ma e ial.
These e en s a e disca ded by a ma e ial e o, which ejec s
e ices in egions occupied by de ec o ma e ial [23]. The
emaining candida es a e ound o be compa ible wi h bb
e en s.
F om simula ion, LLP candida es wi hin he de ec o
accep ance a e selec ed by he L0 and HLT1 igge s wi h
an e iciency o mo e han 85%. The simula ion indica es
ha he igge ac i i y is domina ed by he had onic com-
ponen o he signal expec ed om high mul iplici y e en s.
In HLT2, p ima y e ices and displaced e ices a e econ-
s uc ed om cha ged acks [24]. Genuine PVs a e iden i ied
by a small adial dis ance om he beam axis, Rxy <0.3 mm,
and mus ha e a leas 10 acks, including a leas one o -
wa d ack (i.e. in he di ec ion o he spec ome e ) and one
backwa d ack. Once he se o PVs is iden i ied, all o he
econs uc ed e ices a e candida es o he decay posi ion o
LLPs. The p eselec ion equi es a leas one PV in he e en
and wo LLP candida es. The LLP candida es mus ha e a
leas ou o wa d acks, no backwa d acks, and a minimum
in a ian mass econs uc ed om cha ged acks la ge han
3.5GeV/c2 o one candida e, and la ge han 4.5GeV/c2 o
he o he . In addi ion, he wo seconda y e ices mus ha e
Rxy >0.4 mm and pass he ma e ial e o.
The p eselec ion c i e ia d as ically supp ess he had onic
backg ound. Only 37 e en s (74 LLP candida es) su i e
om he simula ed se o 17.1×106bb e en s gene a ed in he
LHCb accep ance, co esponding o an in eg a ed luminos-
i y o 0.3 pb−1. Th ee simula ed cc e en s pass he selec ion.
They con ain bhad ons and hence belong o he ca ego y o
inclusi e bb, which is also he case o he wo su i ing
e en s. F om he 0.62 b−1da a sample, 42.9×103e en s
a e selec ed. The bbc oss-sec ion alue measu ed by LHCb,
288 ±4±48 µb[25,26], p edic s (76 ±22)×103e en s,
1.8±0.5 imes he yield obse ed in da a. The es ima e uses
he nex - o-leading-o de POWHEG calcula ion [27] o co -
ec Py hia, and he de ec ion e iciency ob ained om he
simula ed e en s. The measu ed yield has also been com-
pa ed o he a e obse ed in LHCb by a dedica ed inclusi e
bbanalysis, based on a opological igge [28]. The consis-
123
Eu . Phys. J. C (2016) 76 :664 Page 3 o 15 664
Fig. 1 Da a (black do s)and
simula ed dis ibu ions a e
p eselec ion no malised o uni
in eg al. The e a e wo LLP
candida es pe e en . The
simula ed bbbackg ound is
shown by he illed ed
his og ams wi h e o ba s.The
dashed (blue), do ed (pu ple)
and solid (g een)lines a e
dis ibu ions o ully simula ed
signal models. The subplo s
show anumbe o acks used o
econs uc he LLP candida es,
bLLP ans e se momen um, c
LLP in a ian mass, d adial
dis ance, Rxy,eunce ain y o
he adial posi ion, σR,and
unce ain y o he longi udinal
posi ion, σZ, o he LLP e ex
LLP numbe o acks
No malised dis ibu ion
Da a
Backg ound
BV48 10ps mH114
BV35 10ps mH114
BV48 10ps mH125
0
0.1
0.2
0.3
0.4
0.5 (a)LHCb
LLP pT
No malised dis ibu ion
[GeV/c]
Da a
Backg ound
BV48 10ps mH114
BV35 10ps mH114
BV48 10ps mH125
0
0.1
0.2
0.3
0.4 (b) LHCb
LLP mass
No malised dis ibu ion
[GeV/c2]
Da a
Backg ound
BV48 10ps mH114
BV35 10ps mH114
BV48 10ps mH125
0
0.1
0.2
0.3
0.4 (c) LHCb
10-1
10-2
10-3
LLP adial dis ance
No malised dis ibu ion
[mm]
Da a
Backg ound
BV48 10ps mH114
BV35 10ps mH114
BV48 10ps mH125
0102030
010203040
010203040 0510152025
(d)LHCb
LLP σR
No malised dis ibu ion
[mm]
Da a
Backg ound
BV48 10ps mH114
BV35 10ps mH114
BV48 10ps mH125
0
0.1
0.2
0.3 (e) LHCb
LLP σZ
No malised dis ibu ion
[mm]
Da a
Backg ound
BV48 10ps mH114
BV35 10ps mH114
BV48 10ps mH125
0
0.1
0.2
0.3
0.4
0 0.05 0.1 0.15 0 0.2 0.4 0.6 0.8 1
( ) LHCb
Fig. 2 Dis ibu ions o a he
LLP dis ance o ligh om he
PV, and, b he adial dis ance o
he LLP e ex, Rxy. The ully
simula ed signal models a e
chosen wi h LLP li e imes o 5,
10, and 50ps. Symbols a e
de ined as in Fig. 1
LLP dis ance o ligh
No malised dis ibu ion
[mm]
Da a
Backg ound
BV48 10ps mH114
BV48 5ps mH114
BV48 50ps mH114
0
0.1
0.2
0.3
0.4
0 50 100 150 200
(a)LHCb
10-1
10-2
10-3
LLP adial dis ance
No malised dis ibu ion
[mm]
Da a
Backg ound
BV48 10ps mH114
BV48 5ps mH114
BV48 50ps mH114
0 5 10 15 20 25
(b) LHCb
ency wi h he bb backg ound is e i ied wi hin a s a is ical
p ecision o 10%.
The shapes o he dis ibu ions o he ele an obse ables
a e compa ible wi h he bbbackg ound. Figu e 1compa es
he dis ibu ions o he LLP candida es aken om da a and
om simula ed bbe en s. The dis ibu ions o h ee ully
simula ed signal models a e also shown. The mass and he
pT alues a e calcula ed assuming he pion mass o each
cha ged ack. Figu e 1d p esen s he adial dis ibu ion o
he displaced e ices; he d op in he numbe o candida es
wi h a e ex abo e Rxy ∼5 mm is due o he ma e ial e o.
The a iables σRand σZshowninFig.1e, a e he posi-
ion unce ain ies p o ided by he e ex i in he ans e se
dis ance Rxy and along he zaxis, pa allel o he beam. The
alues o σRand σZa e la ge o he candida es om bb
backg ound han o he signal because ligh boos ed pa i-
cles p oduce close pa allel acks, wi h he consequence ha
he e ex i has la ge unce ain ies han o he decay o
hea ie pa icles p oducing mo e di e ging acks. Figu e 2
p esen s he LLP dis ance o ligh and Rxy dis ibu ions com-
pa ed o h ee ully simula ed signal models, co esponding
o τLLP alues o 5, 10, and 50ps.
The econs uc ed ou - ec o s o he wo LLPs in he
e en a e added o o m he Higgs-like candida e (di-LLP),
123
664 Page 4 o 15 Eu . Phys. J. C (2016) 76 :664
Fig. 3 Dis ibu ions o a he
pTo he Higgs-like candida e,
and bi s in a ian mass.
Symbols a e de ined as in Fig. 1
di-LLP pT
No malised dis ibu ion
[GeV/c]
Da a
Backg ound
BV48 10ps mH114
BV35 10ps mH114
BV48 10ps mH125
0
0.1
0.2
0.3
0.4
0.5 (a)LHCb
di-LLP mass
No malised dis ibu ion
[GeV/c2]
Da a
Backg ound
BV48 10ps mH114
BV35 10ps mH114
BV48 10ps mH125
0
0.1
0.2
0.3
0.4
0.5
0204060
020406080
(b) LHCb
Table 1 De ini ion o he c i e ia used o he signal de e mina ion.
Selec ions Sel1and Bkg1a e he baseline selec ions used in he i ,
Sel2and Bkg2a e used o he de e mina ion o sys ema ic e ec s.
The ma e ial e o and he equi emen Rxy >0.4 mm a e applied o
bo h LLP candida es. The las column gi es he numbe o da a e en s
selec ed, o a di-LLP econs uc ed mass abo e 19 GeV/c2
Selec ion N ack
min mLLP
min
(GeV/c2)
σR
max (mm) σZ
max Nd(mm)
Sel16 6 0.05 0.25 157
Sel25 5 0.05 0.25 387
Bkg14 4 – – 23.2k
Bkg25 5 – – 10.1k
he co esponding in a ian mass and pTdis ibu ions a e
gi en in Fig. 3.
Fu he cu s a e applied o he p eselec ed da a, o inc ease
he s a is ical sensi i i y. The igu e o me i used is gi en by
/√Nd+1, whe e is he signal e iciency om simula-
ion o a gi en selec ion, and Nd he co esponding num-
be o candida es ound in he da a. The baseline selec ion
(Sel1) is de ined by a minimum numbe o cha ged acks
on each e ex N ack
min =6, a minimum econs uc ed mass
mLLP
min =6GeV/c2, and maximum unce ain ies om he
e ex i σR
max =0.05 mm, and σZ
max =0.25 mm. All he
selec ions used in his analysis a e desc ibed in Table 1, wi h
he indica ion o he numbe o da a e en s selec ed o a di-
LLP econs uc ed mass abo e 19 GeV/c2. Selec ion Bkg1is
used o model he backg ound in he i p ocedu e desc ibed
in Sec . 5, selec ions Sel2and Bkg2a eused os udysys-
ema ic e ec s.
5 De e mina ion o he di-LLP signal
The signal yield is de e mined by a i o he di-LLP in a ian
mass, assuming ha he wo LLPs a e he decay p oduc s
o a na ow esonance. This echnique is hampe ed by he
di icul y in p oducing a eliable backg ound model om
simula ion, despi e he ac ha i is easonable o belie e ha
only bbe en s a e he su i ing SM componen . The e o e, in
his analysis he al e na i e is chosen o in e he backg ound
model om da a by elaxing he selec ion equi emen s, as
gi en by lines Bkg1and Bkg2o Table 1. The compa ison o
he esul s ob ained wi h he di e en signal and backg ound
selec ions is subsequen ly used o es ima e he sys ema ic
e ec s.
The signal empla e is he his og am buil om BV sim-
ula ed e en s selec ed unde he same condi ions as da a,
i.e. Sel1. The backg ound empla e is he his og am ob ained
om da a e en s selec ed by he Bkg1condi ions. The num-
be o signal (backg ound) candida es Ns(Nb) is de e mined
by an ex ended maximum likelihood i . The esul s a e gi en
in Fig. 4 o he BV48 10 ps mH114 signal. The i χ2/nd is
0.6. No e ha only he po ion o he di-LLP mass spec um
abo e 19 GeV/c2is used, in o de o be su icien ly abo e he
mass h eshold se by he selec ions. Al e na i ely, Sel2and
Bkg2a e used o assess sys ema ic e ec s. The i esul s o
he selec ions (Sel1,Bkg2), (Sel2,Bkg1) a e shown in Fig. 5.
The co esponding i χ2/nd alues a e 0.6 and 1.0. The
esul s a e gi en in Table 2 o all ully simula ed signal
models. All i s gi e a nega i e numbe o signal candida es,
compa ible wi h ze o. These esul s a e co ela ed because
he da a sample is in common and he di-LLP mass shapes
a e almos iden ical o he di e en ully simula ed models
as depic ed in Fig. 3. A check is pe o med on 142 di-LLP
candida es selec ed om simula ed bbbackg ound wi hou
he equi emen on Rxy and wi h mLLP
min =4GeV/c2 o bo h
LLPs. The i ed numbe o signal e en s is −0.8±3.5.
The beha iou and sensi i i y o he p ocedu e is u he
s udied by adding a small numbe o signal e en s o he da a
acco ding o a gi en signal model. Figu e 6shows he esul s
o wo models wi h 10 signal e en s added o he da a. The
i ed Nsco esponds well o he numbe o injec ed signal
e en s.
An al e na i e i p ocedu e has been applied, using
pa ame e ised signal and backg ound empla es. The sum
o wo exponen ial unc ions is used o he backg ound, and
an exponen ial con ol ed wi h a Gaussian unc ion o he
signal. The esul s a e consis en wi h a null signal o all he
models.
123
Eu . Phys. J. C (2016) 76 :664 Page 5 o 15 664
di-LLP mass [GeV/c2]
02040 60 80 100
)
2
cEn ies/(2 GeV/
2−
10
1−
10
1
10
2
10 Da a
To al
Backg ound
Signal
(a)LHCb
]
2
c
di-LLP mass [GeV/
20 40 60 80
)
2
cEn ies/(2 GeV/
0
10
20
30
40
50
60
di-LLP mass [GeV/c2]
020 40 60 80 100
−2
0
2
(b) LHCb
Fig. 4 Resul s o he i based on he model BV48 10 ps mH114. In a
log dis ibu ion and blinea scale wi h pull dis ibu ion. Do s wi h e o
ba s a e he da a, he do ed ( ed)and hedashed (g een)his og ams
show he i ed backg ound and signal con ibu ions, espec i ely. The
pu ple his og am is he o al i ed dis ibu ion
As a inal check a wo-dimensional sideband sub ac ion
me hod (“ABCD me hod” [29]) has been applied in he econ-
s uc ed mass o one LLP and he numbe o acks o he
o he LLP, also gi ing esul s consis en wi h ze o signal.
di-LLP mass [GeV/c2]
020406080100
)
2
cEn ies/(2 GeV/
2−
10
1−
10
1
10
2
10 Da a
To al
Backg ound
Signal
(a)LHCb
di-LLP mass [GeV/c2]
020 40 60 80 100
)
2
cEn ies/(2 GeV/
2−
10
1−
10
1
10
2
10 Da a
To al
Backg ound
Signal
(b) LHCb
Fig. 5 Resul s o he i based on he model BV48 10ps mH114, o
di e en combina ions o signal and backg ound selec ions, asignal
om Sel1and backg ound om Bkg2,bsignal om Sel2and back-
g ound om Bkg1.Do s wi h e o ba s a e da a, he dashed (g een)
line is he i ed signal and he do ed ( ed)line he backg ound. In bo h
cases he i ed signal is nega i e. The his og am (blue) is he o al i ed
unc ion
6 De ec ion e iciency and sys ema ic unce ain ies
The de e mina ion o he de ec ion e iciency is based on sim-
ula ed e en s. The geome ical accep ance o he de ec ion
Table 2 Values o he i ed
signal and backg ound e en s
o he di e en ully simula ed
signal models. The
signal/backg ound combina ions
a e de ined in he i s ow
Model (Sel1,Bkg
1)(Sel
1,Bkg
2)(Sel
2,Bkg
1)
NsNbNsNs
BV48 5ps mH114 −2.6±4.4 163.6 ±13.6−4.8±3.9−1.7±3.9
BV48 10ps mH114 −3.3±3.5 164.3 ±13.4−4.6±3.1−3.1±3.6
BV48 15ps mH114 −3.5±3.6 164.5 ±13.5−4.4±3.1−2.0±3.6
BV48 50ps mH114 −1.4±3.6 162.4 ±13.3−2.7±3.4−2.1±4.2
BV48 100ps mH114 −0.7±4.1 161.7 ±13.4−3.5±3.9−3.2±4.2
BV35 10ps mH114 −4.3±3.3 165.3 ±13.4−5.9±3.1−4.6±3.5
BV20 10ps mH114 −1.9±1.6 162.8 ±12.9−2.7±1.7−2.0±2.4
BV48 10ps mH100 −1.7±4.7 162.7 ±13.7−4.4±4.4−5.2±4.7
BV48 10ps mH125 −2.8±3.5 163.8 ±13.4−4.1±3.2−3.2±3.6
BV55 10ps mH114 −3.1±3.7 164.1 ±13.5−4.6±3.4−1.1±3.7
BV55 10ps mH125 −2.6±3.5 163.6 ±13.4−4.0±3.2−3.9±3.8
123
664 Page 6 o 15 Eu . Phys. J. C (2016) 76 :664
di-LLP mass [GeV/c2]
020406080100
)
2
cEn ies/(2 GeV/
2−
10
1
−
10
1
10
2
10 Da a + 10 signal
To al
Backg ound
Signal
(a)LHCb
di-LLP mass [GeV/c2]
020 40 60 80 100
)
2
cEn ies/(2 GeV/
2−
10
1−
10
1
10
2
10 Da a + 10 signal
To al
Backg ound
Signal
(b)
LHCb
Fig. 6 Resul s o he i o he da a o which 10 signal e en s ha e been
added andomly chosen ollowing he signal model. Fo he heo e ical
model BV48 10 ps mH100, in a, he i ed signal is 11.1±7.0e en s;
o BV48 10 ps mH125, in b, he esul is 9.3±5.6e en s
o one ˜χ0
1in LHCb is, depending on he model, be ween
20 and 30%. A e selec ion Sel1 he p edic ed o al di-LLP
de ec ion e iciency is in he ange 0.1–1% o mos o he
models. Po en ial disc epancies be ween simula ion and da a
a e conside ed as sou ces o sys ema ic unce ain ies. Table 3
summa ises he con ibu ions o he sys ema ic unce ain ies,
which a e alid o all ully simula ed models, domina ed by
he 15% con ibu ion om he igge .
The consis ency be ween he igge e iciency in da a and
simula ion is checked by selec ing LLP e en s wi h an inde-
penden igge , designed o he de ec ion o J/ψ e en s.
Compa ing he ac ion o he da a ha also passes he double-
LLP selec ion wi h he co esponding ac ion in simula ed
inclusi e J/ψ e en s, consis en e iciencies a e ound wi hin
a s a is ical unce ain y o 30%. A mo e p ecise esul is
ob ained when equi ing only a single LLP candida e [9]
and assuming unco ela ed con ibu ions om he wo LLPs
o de e mine he e iciency o de ec ing wo LLPs in coinci-
dence. A maximum disc epancy be ween da a and simula ion
o 15% is in e ed, which is he alue adop ed.
The consis ency be ween he ack econs uc ion e i-
ciency in da a and simula ion is s udied by a compa ison
Table 3 Con ibu ions o he sys ema ic unce ain y o ully simula ed
models. Fo he analysis based on he as simula ion he same o al
sys ema ic unce ain y is adop ed augmen ed by 5% o accoun o he
ela i e imp ecision o he as and ull simula ions. The con ibu ions
om he signal and he da a-d i en backg ound models used in he
di-LLP mass i a e discussed in he ex
Sou ce %
T igge 15
T ack econs uc ion 5
Ve ex econs uc ion 4
pTand mass calib a ion 6
Ma e ial e o 4
PV mul iplici y 0.1
Beam line posi ion 0.7
Theo e ical model 9.9
In eg a ed luminosi y 1.7
To al 20.5
o he numbe o acks selec ed in displaced e ices om
bbe en s. The a e age numbe o acks pe LLP in da a is
highe han in simula ed e en s by abou 0.07 acks. Assum-
ing ha his small e ec is en i ely due o a di e ence in
acking e iciency, he o e all di-LLP de ec ion e iciency
changes by a mos 5%.
The e ex econs uc ion e iciency is a ec ed by he
acking e iciency and esolu ion. A s udy o e ices om
B0→J/ψ K∗0wi h J/ψ →μ+μ−and K∗0→K+π−
has shown ha he da a and simula ion de ec ion e iciencies
o his ou -p ong p ocess ag ee wi hin 7.5% [9]. This has
been e alua ed o co espond a mos o a 4% disc epancy
be ween he di-LLP e iciency in da a and simula ion.
A maximum misma ch o 10% on bo h he ans e se
momen um and mass scales is in e ed om he compa i-
son o da a and simula ed bbdis ibu ions, which p opaga es
o a 6% con ibu ion o he sys ema ic unce ain y.
The e ec o he ma e ial e o co esponds o a educ ion
o he geome ical accep ance and depends mainly on he
LLP li e ime. An analysis wi h he equi emen o Rxy <
4 mm allows o in e a maximum sys ema ic unce ain y o
4%.
A small con ibu ion o he sys ema ic unce ain y o 0.1%
is de e mined by eweigh ing he simula ed e en s o ma ch
he PV mul iplici y in he da a.
The unce ain y on he posi ion o he beam line is less
han 20 µm[30]. I can a ec he seconda y e ex selec ion,
mainly ia he equi emen on Rxy. By al e ing he PV posi-
ion in simula ed signal e en s, he maximum e ec on he
di-LLP selec ion e iciency is 0.7%.
The Higgs-like pa icle p oduc ion model is mainly
a ec ed by he unce ain y on he pa on luminosi y. A
maximum a ia ion o he de ec ion e iciency o 9.5% is
123
Eu . Phys. J. C (2016) 76 :664 Page 7 o 15 664
Table 4 De ec ion e iciency
wi h o al unce ain y, and uppe
limi s a 95% CL on he
c oss-sec ion imes b anching
a io o he p ocess pp →h0X,
h0→˜χ0
1˜χ0
1→6q o he ully
simula ed models
Model E iciency (%) Expec ed uppe
limi (pb)
Obse ed uppe
limi (pb)
BV48 5ps mH114 0.528 ±0.114 3.2+2.1
−1.13.5
BV48 10ps mH114 0.925 ±0.194 1.8+1.2
−0.61.7
BV48 15ps mH114 0.966 ±0.208 1.8+1.2
−0.61.6
BV48 50ps mH114 0.419 ±0.090 4.6+2.9
−1.64.4
BV48 100ps mH114 0.171 ±0.037 11.9+7.1
−3.812.3
BV35 10ps mH114 0.268 ±0.058 5.6+3.8
−2.04.9
BV20 10ps mH114 0.016 ±0.003 52+38
−20 54
BV48 10ps mH100 0.864 ±0.186 2.5+1.6
−0.82.6
BV48 10ps mH125 0.771 ±0.166 2.0+1.4
−0.72.0
BV55 10ps mH114 0.851 ±0.183 1.9+1.3
−0.71.9
BV55 10ps mH125 0.937 ±0.201 1.7+1.1
−0.61.7
ob ained ollowing he p esc ip ions gi en in [31]. A sec-
ond con ibu ion o 3% is ob ained by eweigh ing he
Py hia gene a ed e en s o ma ch a ecen calcula ion o
he pTdis ibu ions [32]. The o al heo e ical unce ain y is
9.9%, ob ained by summing in quad a u e he men ioned con-
ibu ions.
In addi ion o he sys ema ic unce ain y on he de ec ion
e iciency, he ollowing con ibu ions ha e been conside ed.
The unce ain y on he in eg a ed luminosi y is 1.7% [33]. As
p e iously s a ed, he unce ain y on he momen um scale and
he in a ian mass scale is smalle han 10%. This alue is
also assumed o he di-LLP mass calib a ion. To assess he
impac on he signal measu emen , pseudoexpe imen s a e
p oduced wi h 10 e en s o simula ed signal added o he
backg ound ollowing he nominal signal dis ibu ion bu
wi h he di-LLP mass alue scaled by ±10%. The subse-
quen maximum a ia ion o he i ed numbe o e en s is
±1.6, o all he signal hypo heses. The unce ain y due o
he shape o he backg ound empla e is ob ained by com-
pa ing he numbe o i ed e en s ob ained wi h he Bkg1
and Bkg2selec ions. The change is less han one e en , o
all he signal models. The di e ence in da a and simula ion
in he di-LLP mass esolu ion and he s a is ical p ecision o
he signal empla es used in he i ha e a negligible e ec .
Hence, a i unce ain y o ±2 e en s is conside ed in he
calcula ion o he c oss-sec ion uppe limi s.
Fo he analysis based on he as simula ion, a 5% unce -
ain y is added o accoun o he ela i e imp ecision o
he as simula ion wi h espec o he ull simula ion, as
explained in Sec . 3.
7 Resul s
The 95% con idence le el (CL) uppe limi s on he p o-
duc ion c oss-sec ion imes b anching a io a e p esen ed in
Table 4, o he ully simula ed models, based on he CLs
app oach [34]. The as simula ion allows he explo a ion o
a la ge egion o pa ame e space. The c oss-sec ion imes
b anching ac ion uppe limi s a 95% CL o benchma k
heo e ical models a e shown in Fig. 7( he co esponding
ables a e gi en in Appendix C).
The es ima ed de ec ion e iciencies can be ound in
Appendix B, Tables 6and 7. The e iciency inc eases wi h
mLLP because mo e pa icles a e p oduced in he decay o
hea ie LLPs. This e ec is only pa ially coun e ac ed by
he loss o pa icles ou side o he spec ome e accep ance,
which is especially he case wi h hea ie Higgs-like pa i-
cles. Ano he compe ing phenomenon is ha he lowe boos
o hea ie LLPs esul s in a sho e a e age ligh leng h, i.e.
he equi emen o a minimum Rxy dis a ou s hea y LLPs.
The cu on Rxy is mo e e icien a selec ing LLPs wi h la ge
li e imes, bu o li e imes la ge han ∼50 ps a po ion o he
decays alls in o he ma e ial egion and is disca ded. Finally,
a d op o sensi i i y is expec ed o LLPs wi h a li e ime close
o he bhad on li e imes, whe e he con amina ion om bb
e en s becomes impo an , especially o low mass LLPs.
8 Conclusion
A sea ch o Higgs-like bosons decaying in o wo long-li ed
pa icles decaying had onically has been ca ied ou using
da a om pp collisions a 7 TeV collec ed wi h he LHCb
de ec o , co esponding o a o al in eg a ed luminosi y o
0.62 b−1.
The model used o desc ibe he LLP decay is an mSUGRA
p ocess in which he ligh es neu alino ˜χ0
1decays h ough
a ba yon numbe iola ing coupling o h ee qua ks. Uppe
limi s ha e been placed on he p oduc ion c oss-sec ion o
Higgs-like boson masses om 80 o 140GeV/c2, LLP masses
123
664 Page 8 o 15 Eu . Phys. J. C (2016) 76 :664
Fig. 7 Expec ed (open do s
wi h 1σand 2σbands) and
obse ed ( ull do s) uppe limi s
a 95% con idence le el, a–c
shown o di e en masses o
he Higgs-like pa icle, d, o
di e en LLP li e imes, and eas
a unc ion o he LLP mass. The
alues o he o he pa ame e s
a e indica ed on he plo s.
Resul s in e ed om he as
simula ion
C oss-sec ion [pb]
[GeV/c2]
1
10
(a)mLLP =35GeV/c2
τLLP =10ps
LHCb
mh0
120
C oss-sec ion [pb]
[GeV/c2]
1
10
(b) mLLP =40GeV/c2
τLLP =10ps
LHCb
mh0
120
C oss-sec ion [pb]
[GeV/c2]
1
10
(c)
mLLP =48GeV/c2
τLLP =10ps
LHCb
mh0
50
C oss-sec ion [pb]
[ps]
1
10
80 100 120 140 100 140
100 140 025 75100
(d)mLLP =40GeV/c2
mh0= 100 GeV/c2
LHCb
τLLP
40
C oss-sec ion [pb]
[GeV/c2]
1
10
(e)
τLLP =10ps
mh0= 125 GeV/c2
LHCb
mLLP
50
C oss-sec ion [pb]
[ps]
1
10
20 30 50 60 025 75100
( ) mLLP =48GeV/c2
mh0= 125 GeV/c2
LHCb
τLLP
in he ange 20–60GeV/c2, and LLP li e imes in he ange
o 5–100ps. The numbe o candida es is de e mined by
he di-LLP in a ian mass i wi h signal empla es in e ed
om simula ion, and backg ound es ima es om da a. Fo
he explo ed pa ame e space o he heo y all esul s, which
a e co ela ed, a e consis en wi h ze o. Uppe limi s a 95%
CL o c oss-sec ion imes b anching a io o 1 o 5pb a e
in e ed o mos o he conside ed pa ame e ange. They
a e below 2pb o he decay o a 125GeV/c2Higgs-like pa -
icle in wo LLPs wi h mass in he 48–60GeV/c2 ange and
10ps li e ime.
Acknowledgemen s We exp ess ou g a i ude o ou colleagues in he
CERN accele a o depa men s o he excellen pe o mance o he
LHC. We hank he echnical and adminis a i e s a a he LHCb
ins i u es. We acknowledge suppo om CERN and om he na ional
agencies: CAPES, CNPq, FAPERJ and FINEP (B azil); NSFC (China);
CNRS/IN2P3 (F ance); BMBF, DFG and MPG (Ge many); INFN
(I aly); FOM and NWO (The Ne he lands); MNiSW and NCN (Poland);
MEN/IFA (Romania); MinES and FANO (Russia); MinECo (Spain);
SNSF and SER (Swi ze land); NASU (Uk aine); STFC (Uni ed King-
dom); NSF (USA). We acknowledge he compu ing esou ces ha a e
p o ided by CERN, IN2P3 (F ance), KIT and DESY (Ge many), INFN
(I aly), SURF (The Ne he lands), PIC (Spain), G idPP (Uni ed King-
dom), RRCKI and Yandex LLC (Russia), CSCS (Swi ze land), IFIN-
HH (Romania), CBPF (B azil), PL-GRID (Poland) and OSC (USA).
We a e indeb ed o he communi ies behind he mul iple open sou ce
so wa e packages on which we depend. Indi idual g oups o membe s
ha e ecei ed suppo om A H Founda ion (Ge many), EPLANET,
Ma ie Skłodowska-Cu ie Ac ions and ERC (Eu opean Union), Conseil
Géné al de Hau e-Sa oie, Labex ENIGMASS and OCEVU, Région
Au e gne (F ance), RFBR and Yandex LLC (Russia), GVA, Xun a-
Gal and GENCAT (Spain), He chel Smi h Fund, The Royal Socie y,
Royal Commission o he Exhibi ion o 1851 and he Le e hulme T us
(Uni ed Kingdom).
Open Access This a icle is dis ibu ed unde he e ms o he C ea i e
Commons A ibu ion 4.0 In e na ional License (h p://c ea i ecomm
ons.o g/licenses/by/4.0/), which pe mi s un es ic ed use, dis ibu ion,
and ep oduc ion in any medium, p o ided you gi e app op ia e c edi
o he o iginal au ho (s) and he sou ce, p o ide a link o he C ea i e
Commons license, and indica e i changes we e made.
Funded by SCOAP3.
Appendices
A Fully simula ed signal da ase s
Table 5shows he pa ame e s used o gene a e he 11
ully simula ed signal models wi h Py hia 6. The Higgs-
123
Eu . Phys. J. C (2016) 76 :664 Page 9 o 15 664
Table 5 Pa ame e s o he signal models gene a ed by Py hia and ully
simula ed
Model M1
(GeV/c2)
an βmh0
(GeV/c2)
mLLP
(GeV/c2)
τLLP
(ps)
BV48 5ps mH114 62 5 114 48 5
BV48 10ps mH114 62 5 114 48 10
BV48 15ps mH114 62 5 114 48 15
BV48 50ps mH114 62 5 114 48 50
BV48 100ps mH114 62 5 114 48 100
BV35 10ps mH114 46 5 114 35 10
BV20 10ps mH114 28 5 114 20 10
BV48 10ps mH100 71 2.4 100 48 10
BV48 10ps mH125 60 8 125 48 10
BV55 10ps mH114 71 5.1 114 55 10
BV55 10ps mH125 69 6.2 125 55 10
like boson is p oduced by gluon-gluon usion. In he able
M1co esponds o he Py hia pa ame e RMSS(1), and
an β o RMSS(5). In addi ion, M2(RMSS(2)) is se a
250GeV/c2and μ(RMSS(4)) has he alue 140. A mh0 alue
o 125GeV/c2 equi es RMSS(16) = 2300.
B De ec ion e iciencies
Table 6gi es he de ec ion e iciency as a unc ion o mh0and
mLLP, he LLP li e ime is 10ps. Table 7gi es he e iciency as
a unc iono mLLP and τLLP, assuming mh0=114 GeV/c2.
Table 6 De ec ion e iciency alues in pe cen es ima ed by he as
simula ion as a unc ion o mh0and mLLP. The LLP li e ime is 10 ps.
The s a is ical unce ain y is 10% o ∼0.02%, 5 % o ∼0.1%,
3% o ∼0.5%, and 2% o ∼1%
mh0
(GeV/c2)
mLLP (GeV/c2)
20 25 30 35 40 48 55 60
80 0.0350.1260.2760.514––––
90 0.027 0.084 0.213 0.456 0.699 – – –
95 0.023 0.077 0.203 0.414 0.689 – – –
100 0.025 0.073 0.184 0.368 0.647 0.858 – –
105 0.018 0.066 0.139 0.324 0.574 1.018 – –
110 0.017 0.053 0.146 0.291 0.525 1.016 – –
114 0.014 0.048 0.134 0.259 0.472 0.963 0.817 –
120 0.016 0.047 0.107 0.222 0.402 0.836 1.013 –
125 0.009 0.042 0.097 0.225 0.377 0.765 0.997 –
130 0.014 0.037 0.085 0.191 0.325 0.708 0.914 0.991
140 0.002 0.031 0.075 0.163 0.277 0.566 0.782 0.881
Table 7 De ec ion e iciency in pe cen es ima ed by he as simula ion
as a unc ion o he mLLP and τLLP, o mh0= 114 GeV/c2. The s a is ical
unce ain y is 10% o ∼0.02%, 5 % o ∼0.1%, 3% o ∼0.5%,
and 2% o ∼1%
τLLP (ps)mLLP (GeV/c2)
20 25 30 35 40 48 55
5 0.021 0.053 0.129 0.234 0.366 0.545 0.289
10 0.014 0.048 0.134 0.259 0.472 0.963 0.817
15 0.013 0.042 0.113 0.198 0.389 0.932 1.052
20 0.007 0.035 0.083 0.174 0.338 0.834 1.150
25 0.006 0.034 0.073 0.148 0.289 0.731 1.126
30 0.005 0.026 0.066 0.128 0.241 0.643 1.091
40 0.003 0.017 0.044 0.114 0.193 0.490 0.960
50 0.004 0.015 0.035 0.082 0.157 0.397 0.806
70 0.002 0.009 0.021 0.062 0.104 0.280 0.596
100 0.001 0.005 0.015 0.033 0.071 0.178 0.383
C C oss-sec ion uppe limi s ables
Expec ed and obse ed 95% CL c oss-sec ion imes b anch-
ing a io uppe limi s o benchma k models, om he as
simula ion. Tables 8and 9gi e he limi s as a unc ion o
mh0, co e ing LLP masses om 35 o 60 GeV/c2,τLLP =
10 ps. Table 10: limi s as a unc ion o he LLP li e ime
o mh0=100 GeV/c2and mLLP =40 GeV/c2, and o
mh0=125 GeV/c2and mLLP =48 GeV/c2. Table 11: lim-
i s as a unc ion o he LLP mass, o mh0=125 GeV/c2,
τLLP =10 ps.
Table 8 Expec ed and obse ed 95% CL c oss-sec ion imes b anching
a io uppe limi s as a unc ion o mh0, wi h mLLP =35 GeV/c2,and
τLLP =10 ps, es ima ed by he as simula ion
Model Expec ed uppe
limi (pb)
Obse ed uppe
limi (pb)
BV35 10ps mH80 6.49+3.94
−2.16 6.20
BV35 10ps mH90 5.50+3.42
−1.89 4.56
BV35 10ps mH95 5.42+3.41
−1.88 4.06
BV35 10ps mH100 5.55+3.52
−1.92 4.45
BV35 10ps mH105 5.92+3.79
−2.06 4.78
BV35 10ps mH110 5.94+3.79
−2.06 4.56
BV35 10ps mH114 6.07+3.92
−2.11 4.77
BV35 10ps mH120 6.79+4.42
−2.39 5.47
BV35 10ps mH125 7.21+4.70
−2.54 6.03
BV35 10ps mH130 7.28+4.83
−2.59 7.08
BV35 10ps mH140 7.95+5.32
−2.85 6.35
123
