Events with Isolated Leptons and Missing Transverse Momentum in ep Collisions at HERA

0.35 MB·English

by  G. Brandt

Events with Isolated Leptons and
7
Missing Transverse Momentum in ep
0
0
2 Collisions at HERA

n
a
J ∗
Gerhard Brandt

8
2

Physi
al Institute, Philosophenweg 12, 69120 Heidelberg
1
v
0
5
0 Abstra
t
1
0 The analysis of events with isolated leptons and missing transverse momentum in
7
the H1 experiment is dis
ussed for the ele
tron, muon and tau
hannels. In the
0 W
/ Standard Model (SM) framework, produ
tion of real -bosons gives rise to su
h
x
e topologies. Contributions to the ba
kground are dominated by QCD pro
esses.
-
p An ex
ess of observed signal over ba
kground presents a
han
e of the dis
overy
he of new−p1hysi
s. The results usinget−hepHERA 1994-2006 data set
orresponding to
v: 341 pb are presented We+hpile the sample shows good agreement between datσa
i and SM expe
tation, in
ollisions an ex
ess over the SM expe
tation with 3.4
X
signi(cid:28)
an
e is observed at high hadroni
transverse momentum.
r
a
1 Introdu
tion
P
T
Events with high- isolated leptons and missing transverse momentum are
sensitive to physi
s beyond the Standard Model (BSM). Both H1 and ZEUS
have previously published sear
hes for events with isolated ele
trons and
τ
muons [1, 2, 4, 3, 5℄ and −leptons [6, 7℄. The observation of su
h events in
ep

ollisions at HERA has
reated
onsiderable interest in the high-energy
physi
s
ommunity (many referen
es
an be found in [2℄) and is
ontinously
updated as new data be
ome available [8, 9, 10℄. Se
tion 2 reviews the ex-
perimental
onditions (
ollider and dete
tor) of the sear
h, se
tion 3 presents
the most important physi
s pro
esses involved, se
tion 4 presents the sear
h
∗
email: gbrandtmail.desy.de
e/µ + P τ + P
Tmiss Tmiss
for events, se
tion 5 the sear
h for and se
tion 6
dis
usses the results and possible interpretations.
2 Experimental Conditions
ep
HERA operating at DESY in Hamburg is the worlds only
ollider. It
√s =
1

ollides ele
trons or positrons with protons at a
enter-of-mass energy
320
GeV for most of its operating time. HERA data taking is divided in two
mainrunningperiods, HERA-1intheyears 1994to2000,andHERA-2inthe
years 2000−1to 2007. During HERA-1, H1
olle
tede+apdata set
orresponding
to 118 pb of data, dominated by about 90% by data. For the HERA-
2 period, the luminosity has been upgraded and spin rotators installed to
provide a longitud−i1nally polarised ele
tron beam. For the results dis
ussed
hereabou−t250pb ofHERA-2datahavebeenavailabletoH1. Thisin
ludes
e p
the full −1 data sample whi
h
orresponed+spto a total integrated luminosity
of 182 pb . Data taking
ontinues with running until the end of HERA
operation, whi
h is s
heduled for June 2007.
Two
ollider experiments, H1 and ZEUS, operate at HERA. This arti
le
fo
usses on H1, des
ribed in detail in [11℄. A drawing of the H1 dete
tor
4π
is shown in (cid:28)gure 1a. It has almost angular
overage to dete
t missing
transverse momentum and the subdete
tors are arranged for ex
ellent lep-
ton identi(cid:28)
ation. Tra
k information is provided by
entral jet
hambers, a
forward tra
ker and sili
on vertexing dete
tors. The tra
kers in a solenoid
magnet (cid:28)eld allow momentum measurement of
harged parti
les. Energy
measurement for ele
tromagneti
and hadroni
showers is provided by a LAr
(liquidargon)
alorimeterinthe
entral and forward(proton beamdire
tion)
region, and a lead-(cid:28)bre spaghetti
alorimeter ((cid:16)SpaCal(cid:17)) in the ba
kward di-
re
tion. For muon dete
tion, the iron (cid:29)ux return yoke of the magnet is in-
strumented with limited streamer tubes to measure es
aping muons. In the
forward dire
tion, there is an additional forward muon dete
tor installed.
To
ompare re
orded data to SM predi
tions, the relevant physi
s pro-

esses are modelled by Monte Carlo generators. The dete
tor response is
simulated by H1 software based on GEANT 3 [12℄. The simulated events
are then pro
essed through the
omplete re
onstru
tion and analysis
hain,
identi
al to the data treatment.
1
Inthisarti
le,thetermele
tron referstobothele
tronsandpositronsunlessexpli
itly
stated.
HERA II
2)V
e H1 e+p NC 03-04 (prel.)
pb/G 10 ZHE1U eS-p e N+pC N2C00 250 (0p4rel.)
s2 (/dQd10-1 SZSEMMU eeS+-p pe N-NpCC N ((CCC T0TE4E-QQ0656M M(p))rel.)
10-3 H1 e+p CC 03-04 (prel.)
H1 e-p CC 2005 (prel.)
ZEUS e+p CC 2004
ZEUS e-p CC 04-05 (prel.)
10-5 SM e+p CC (CTEQ6M)
SM e-p CC (CTEQ6M)
y < 0.9
Pe = 0
10-7
103 104
Q2 (GeV2)
a) b)
Fig. 1: a) The H1 dete
tor at HERA [11℄. Ele
trons enter from the left side,
andprotonsfromtherightside; b)AHERA(cid:16)textbook(cid:17) measurement:
In
lusive unpolarised
ross se
tions of NC and CC DIS pro
esses in
ep

ollisions measured by the H1 and ZEUS experiments (taken from
Q2 > 1000
[13℄). The uni(cid:28)
ation of ele
troweak for
es is visible for
GeV.
3 Physi
s in ep Collisions
In the Standard Model events with isolated leptons and missing energy are
W
mainly expe
ted from real produ
tion with subsequent leptoni
de
ay,
±
ep eW (֒ lν)X,
→ →
where X denotes the hadroni
(cid:28)nal state (HFS). This is
omposed of the
s
attered quark jet and the proton remnant. The main
ontribution to this
W
pro
ess is radiation of a boson o(cid:27) the s
attered quark line, shown in
(cid:28)gure2a. Thisandother
ontributingpro
esses aresimulatedbytheEPVEC
1
generator[14℄. Thetotal
rossse
tionisabout pb. Withatypi
aldete
tion
e(cid:30)
ien
y of 40% for the teoptal leptoni
bran
hing r−a1tio this meWans about 100
events are expe
ted inan datasample of 250pb . ForSM produ
tion,
PX
T
the transverse momentum of the HFS is predi
ted to be well below 25
PX PX
T T
GeV. At high the SM predi
tion is low. Looking at its provides a
PX
T
measure for how untypi
al, or interesting, an event is. An event with a
of 29 GeV is shown in (cid:28)gure 3.
Several physi
s pro
esses
ontribute to the ba
kground. The most promi-
ep eX
nent physi
s pro
esses measuredatHERA areneutral
urrent (NC) →
ep νX
and
harged
urrent (CC) → in deep inelasti
s
attering (DIS), shown
in (cid:28)gures 2b and
respe
tively. These pro
esses are simulated by the RAP-
GAP [15℄ and DJANGO [16℄ generatorsQ. 2The regime of DIS is de(cid:28)ned by
the negative momentum transfer square of the boson ex
hanged bQe2tw>een1
the indi
ent ele
tron and the quark s
attered from the proton, at
Q2 < 1
GeV. The regime GeV known as photoprodu
tion is simulated in the
PYTHIA [17℄ framework. Here, two jets are produ
ed by boson-gluon-fusion
as shown in (cid:28)gure 2d. For the analysis presented in this paper NC and CC in
DIS are important ba
kground pro
esses due to their high
ross se
tions and
event topologies. They are also tools, be
ause they provide high statisti
s
samples of (cid:28)nal state obje
ts su
h as leptons and QCD jets whose properties
need to be well understood to make a

urate predi
tions. Figure 1b shows
measurements of NC and CC
ross se
tions done at HERA.
As a template for physi
s beyond the Standard Model produ
ing isolated
lepton and missing transverse momentum topologies, the anomalous produ
-
tionoftopquarksvia(cid:29)avor
hangingneutral
urrents(FCNC)isused(shown
in Fig. 2e). This pro
ess was generated by the ANOTOP [18℄ programme
and used to optimise a

eptan
e for a possible BSM signal.
4 Sear
h for Events with Isolated Ele
trons and Muons
Ele
trons are dete
ted as isolated,
ompa
t ele
tromagneti

lusters in the

alorimeter and a high quality tra
k pointing to that
luster. Muons are
dete
ted by signals in the muon dete
tors, mat
hing tra
ks in the tra
kers
and energy deposits in the LAr
alorimeter typi
al for minimally ionising
parti
les. Ele
trons anPdlM>uo1n0s are dete
ted in the phase spa
5e◦θof<lep1t4o0n◦
T l
transverse momentum GeV and polar angular range
[19, 21℄. Leptons are required to be isolated against other tra
ks and jets in
the event. This redu
es leptons inor
lose to QCD jets whi
h are expe
ted to
Pl
T
be at lower . In DIS the s
attered ele
tron is mostly seen in the ba
kward
◦
θ > 155 P
l Tmiss
region of the dete
tor. The missing transverse momentum
is re
onstru
ted from the four-ve
tor sum of all re
onstru
ted (cid:28)nal state par-
ti
les (hadrons and leptons) and required to be greater than 12 GeV. This
greatly redu
es NC and photoprodu
tion
ontributions to the ba
kground,
P
Tmiss
where only appears due to (cid:29)u
tuations in the energy measurement.
Additionally SM pro
esses are suppressed topologi
ally, for example by re-
quiring a
oplanarity of the lepton and the hadroni
(cid:28)nal state four-ve
tor
in the transverse plane. For neutral
urrent and photoprodu
tion events a
ba
k-to-ba
k topology of the (cid:28)nal state obje
ts is expe
ted from momentum

onservation.
e+ e+ e+ e+
q’
g , Z0
q’ W+ q
q q
n
l
p X l+ p X
p p
a) b)
e+ n– e+ e+
e
W+
q
q’ –
q q
p Xp p Xp

) d)
e+ e+
b
t
u,c k g W+ n–, q–
l
l+,q’
p X
p
e)
l + P
Tmiss
Fig. 2: Physi
s pro
esses at HERA
ontributing to topologies as
signal (a and e) or ba
kground due to mismeasurements (b,
,d): a)
W
Produ
tion of real -Bosons in the Standard Model; b) Neutral
ur-
rent;
) Charged
urrent; d) Photoprodu
tion of jets by boson-gluon-
fusion; e) Anomalous produ
tion of single top quarks via FCNC, the
template used for possible BSM
ontributions.
Pe = 37
T
Fig. 3: Event display for an event with an isolated ele
tron of GeV,
Pmiss = 44
T
missingPtrXan=sv2e9rse momentum of eG+epV and a hadroni
sys-
T
tem of GeV, observed in HERA-2 data. The Standard
Model expe
tation for events with this topology is low.
Figure 4 shows theer+espults of the sear
h for the
ombined ele−
1tron and
me−upon
hannels for the data sample
o−rr1esponding to 158 pb and the
data sample
orresponding to 184 pb . A total of 46 events are ob-
served in the
omplete data, with 43.0 ± 6.0 expe
ted from the SM. At large
PX > 25
T
hadroni
transverse momentum GeV a total of 18 events are ob-
see+rpved, but only 11.5 ± 1.8 predi
ted. 15 of these events are observed in the
s−ample, where only 4.6 ± 0.8 are expe
ted. Only 3 events are seen in
e p
the data sample, where 6.9 ± 1.0 are expe
ted. The signi(cid:28)
an
e of the
e+p σ
o−bserved ex
ess in the data sample is 3.4 . No ex
ess is observed in the
e p
data sample. This observation represents a
han
e for dis
overy of new
physi
s at HERA.
5 Sear
h for Events with Isolated Tau Leptons
DuetoleptonuniversalityintheStandardModel,thetau
hannelisexpe
ted
τ
to behave identi
altothe ele
tronand muon
hannels. Leptoni
-de
ays are
τ
in
luded in the ele
tron and muon
hannels. This sear
h adds hadroni
-
de
ays with one
harged hadron, leadingto a 1-prongsignature, whi
h
overs
τ
about 50% of the total -lepton bran
hing ratio [20, 21℄. 3-prong de
ays are
not in
luded be
ause the ba
kground is too large to yield a signi(cid:28)
ant signal-
to-ba
kground ratio. jet
P > 7
T
Tau-je2t0s◦a<reθidjeetn<ti(cid:28)1e2d0◦based on hadroni
jets in the phase spa
e
GeV and . The
andidate jets are expe
ted to be
ollimated
τ
due to the boost of the -lepton. This is ensured by requiring a jet radius
l+Pmiss events at HERA 1998-2006 (e-p, 184 pb-1) l+Pmiss events at HERA 1994-2004 (e+p, 158 pb-1)
T T
ss ss
entent110022 HA1ll DSMata (prelim.) NNSDMat a = = 1284.4 – 3.4 entent110022 HA1ll DSMata (prelim.) NNSDMat a = = 2188.5 – 2.6
vv Signal vv Signal
EE EE
1100 1100
11 11
1100--11 1100--11
00 1100 2200 3300 4400 5500 6600 7700 8800 00 1100 2200 3300 4400 5500 6600 7700 8800
PPXX ((GGeeVV)) ee aanndd mm cchhaannnneellss PPXX ((GGeeVV)) ee aanndd mm cchhaannnneellss
TT TT
a) b)
PX e
T
Fig. 4: Transverse momentumofthe hadroni
system forthe
ombined -
µ e/µ+P
and −
hannels in the Tmiss sear
h. The points show the data
withstatisti
alerrors. TheopenhistogramshowstheSMexpe
tation.
Systemati
and statisti
al errors added in quadrature are shown as
the shaded band. The signal
ontribution of the SM expe
tation
W
dominated by real produ
tion is show−n as a hat
hed histogram
e p
and amounts to about 70%e+.pWhile the data (a) agree well with
the SM expe
tation , the data (b) show an ex
ess of events over
σ
the SM predi
tion amounting to 3.4 signi(cid:28)
an
e.
R < 0.12 η φ
jet
in the − −plane. Exa
tly one tra
k is required in a
one
D = 1.0 η φ
in the − −plane around the jet axis. Neutrinos from the de
ay
W ντ ֒ lν ν¯ P > 12

hain → → l are taken into a

ount by requiring Tmiss
τ
GeV. Be
ause of the similarity of -jets to narrow QCD jets with a low
tra
k mulitpli
ity, the signal to ba
kground ratio is less favoQra2b<le t1han in
the ele
tron and muon
hannels. Jets in photoprodu
tion ( GeV)
P
Tmiss
may enter the sample due to fake from (cid:29)u
tuations in the energy
τ
measurement and QCD jets faking -jets. CC ba
kground is mu
h higher
than in the ele
tron and muon
hannel. The
orre
t des
ription of dijet CC
events be
omes important sin
e their signature is very similar to the events
in the interesting region at high PTX. The se
ond jet, whi
h
ould fake a
τ
-jet, may
ome from next to leading order pro
essτes or (cid:28)nal statee+rpadiation.
Figure 5 shows the preliminary results for the analysis. In data, 8

andidaPteXs are observed for 10.8 ± 2.5 expe
ted. All
andidatees−apre in the
T
lowest bin where most of the ba
kground is expe
ted. In data 17
events are observed, with 13.5 ± 2.8 expe
ted. Three events are observed
PX
at T > 25 GeV, while only 0.74 ±0.18 are expe
ted. At (cid:28)rst glan
e this
seems to suggest a trend opposite to the ele
tron and muon
hannels, but
a) b)
PX τ+P
T Tmiss
Fig. 5: Transverse momentumofthehadroni
PsyXstem in e+p events.
T
While there are no
andidates at highPX observed in thee−p sample
T
(a), therearethree
andidatesathigh observed inthe sample.
this
on
lusion is premature. The statisti
al signi(cid:28)
an
e of the result is low.
Also at
loser inspe
tion, some of the three observed
andidates seem quite
τ
untypi
al for -jets and are more likely ba
kground that
an possibly be
removed in an improved analysis.
6 Interpretation
While the low number of observed events in these analyses does not yet show
a
lear pattern pointing to a strong hypothesis for a BSM pro
ess to explain
the ex
ess, it is already intriguing to think about possible me
hanisms.
Upon the (cid:28)rst observation of these events at HERA-1, one of the most
top
ex
iting interpretations was the produ
tion of single -quarks. While this
is allowed kinemati
ally and possible in CC pro
esses at HERA, the
ross
se
tion is too low to see a signal in the available data samples. If observed,
top
-quark events
ould therefore only be produ
ed anomalously via (cid:29)avor

hanging neutral
urrents (FCNC). The pro
ess is illustrated in (cid:28)Zgu0re 2e
where the s
attered quark
ouples to the ex
hanged photon (or , not
top
shown) via a FCNC
oupling and produ
es a -quark. The subsequent
t bW W lν W qq¯
de
ay → followed by → or → produ
es either the sear
hed
l + PMiss b
T
signature or a 3 jet topology. The -jet gives rise to a highly en-
PX
T
ergeti
hadroni
system X, giving rise to the high . These leptoni
and
hadroni
signatures have been analysed and limits have been set whi
h are

ompetitive to limitson anomalous top produ
tion set at other
olliders [22℄.
Another possibility to explain the observation is to expli
itly
onstru
t
N
a model. This
ould be based on a new parti
le with fermion number
e q m
N
F=0
oupling to − . If sux
h a=paMrt2i
/lse had a high mass it would
Bj N
have to be produ
ed at large . In this region the parton den-
sity of valen
e quarks in the proton is mu
h larger than the density of sea
quaer+kps, hen
e the produe
−tpion
ross se
tion for su
h a partiR
le is mu
h larger
in running than in running. In the framework of -parity violating
stop
supersymmetry, su
h parti
les
ould be resonantly produ
ed -squarks,
sbottom
or -squarks produ
ed in the t-
hannel [23℄. These preliminary ideas
have yet to be intensively studied.
A major issue with the observation is that ZEUS
annot
on(cid:28)rm the ex-

ess observed by H1, although the analysis is performed in a more restri
ted
phase spa
e. Studies to properly
ompare the results and test the
ompati-
bility of both experiments are ongoing [8℄.
e+p
For the remainder of HERA data taking data will be re
orded. If the
datatak15in0g
o−n1tinues atthe
urrent rate, an additionalintegratedluminosity
ofL ∼ pb
anbeexpe
ted. Atthetimeofwritingtheanalysisofevents
with isolated leptons and missing transverse momentum is followed with full
e(cid:27)ort to extra
t the maximumof informationfrom this puzzling observation.
Referen
es
[1℄ T. Ahmed et al. [H1 Collaboration℄, DESY 94-248(1994).
[2℄ C. Adlo(cid:27) et al. [H1 Collaboration℄,Eur. Phys. J. C 5 (1998) 575 [hep-ex/9806009℄.
[3℄ V.Andreevetal.[H1Collaboration℄,Phys.Lett.B561(2003)241[hep-ex/0301030℄.
[4℄ J. Breitweg et al. [ZEUS Collaboration℄, Phys. Lett. B 471 (2000) 411 [hep-
ex/9907023℄.
[5℄ S. Chekanov et al. [ZEUS Collaboration℄, Phys. Lett. B 559 (2003) 153 [hep-
ex/0302010℄.
[6℄ S. Chekanov et al. [ZEUS Collaboration℄, Phys. Lett. B 583 (2004) 41 [hep-
ex/0311028℄.
[7℄ A. Aktas et al. [H1 Collaboration℄, submitted to Eur. Phys. J. C [hep-ex/0604022℄.
3
[8℄ H1 Collaboration, H1prelim-06-162 .
4
[9℄ ZEUS Collaboration, ZEUS-prel-06-012 .
3.
[10℄ H1Collaboration, H1prelim-06-064
[11℄ H1 Collaboration, The H1-Dete
tor at HERA, Nu
l. Instr.and Meth. A 386 (1997)
310 and 348.
[12℄ GEANT 3, R. Brun et al., CERN-DD/EE/84-1.
3
[13℄ H1 Collaboration, H1prelim-06-041 ; 5
B. Portheault, Dissertation, Université Paris-Sud XI (2005) .
[14℄ U. Baur, J. A. M. Vermaseren and D. Zeppenfeld, Nu
l. Phys. B 375 (1992) 3.
[15℄ H.Jung,Comput.Phys.Commun.86(1995)147,RAPGAPprogrammanual(1998)
http://www-h1.desy.de/~jung/RAPGAP.html.
[16℄ G. A. S
hülerand H. Spiesberger, DJANGO 2.1, Pro
. of the Workshop"Physi
sat
HERA" (1991), Eds. W. Bu
hmüller and G. Ingelman., Vol. 3, p. 1419.
[17℄ T. Sjöstrand, PYTHIA 5.7, Comput. Phys. Commun. 82 (1994) 74.
[18℄ E. Perez, ANOTOP, unpublished.
5
[19℄ D. South, Dissertation, University of Man
hester (2003) .
5
[20℄ C. Veelken, Dissertation, University of Liverpool (2006) .
5
[21℄ G. Brandt, Dissertation, University of Heidelberg (2007, in preparation) .
[22℄ A. Aktas et al. [H1Collaboration℄,Eur. Phys. J C 33 (2004) 9 [hep-ex/0310032℄.
[23℄ E. Perez, private
ommuni
ation
For more information on (R-parity violating) SUSY try5:
J. Haller, Dissertation, University of Heidelberg (2003)
S. Martin, A Supersymmetry Primer, [hep-ph/9709356℄.
3
H1 preliminary results available at:
4http://www-h1.desy.de/publi
ations/H1preliminary.short_list.html
ZEUS preliminary results available at:
5http://www-zeus.desy.de
H1 theses available at:
http://www-h1.desy.de/publi
ations/theses_list.html