CCentral Exclusive Production at LHCb
R. McNulty ∗ School of Physics, University College Dublin,Dublin 4, Ireland,(on behalf of the LHCb collaboration.)E-mail: [email protected]
Central Exclusive Production (CEP) is a unique process at hadron machines in which particlesare produced via colourless propagators. LHCb have measured the cross-sections for the CEP ofvector mesons, J / ψ , ψ ( S ) , ϒ ( S ) , ϒ ( S ) and ϒ ( S ) , which are photo-produced. In the doublepomeron exchange process, preliminary measurements have been made of χ c , χ c , χ c mesonproduction while the first observations of the CEP of pairs of charmonia, J / ψ J / ψ and J / ψψ ( S ) ,have been made and limits obtained on the pair production of other charmonia. XXIV International Workshop on Deep-Inelastic Scattering and Related Subjects11-15 April, 2016DESY Hamburg, Germany ∗ Speaker. c (cid:13) Copyright owned by the author(s) under the terms of the Creative CommonsAttribution-NonCommercial-NoDerivatives 4.0 International License (CC BY-NC-ND 4.0). http://pos.sissa.it/ a r X i v : . [ h e p - e x ] A ug EP at LHCb
R. McNulty
1. Introduction
Central exclusive production (CEP) at the LHC is characterised by an isolated system of par-ticles surrounded by two rapidity gaps that extend down to the intact colliding protons; see [1] andthe papers referenced in [2] for a recent survey of the field. The lack of additional activity signalsthe presence of colourless propagators: two photons; two pomerons; or a photon and a pomeron.Measurements of CEP test QCD, investigate the nature of the pomeron, and can be used to con-strain the gluon PDF. At √ s =
13 TeV, measurements of J / ψ production will probe the gluon PDFdown to a fractional momentum of the proton x = × − , a scale at which saturation effects maybecome visible.The LHCb detector [3] is suited to studying CEP as it is fully instrumented with tracking,calorimetry and particle identification in the pseudorapidity, η , range between 2 and 5. In addition,charged activity in the backward region, in the approximate range − . < η < − .
5, can be vetoeddue to the presence of a silicon strip detector that surrounds the interaction point. LHCb is designedto trigger on particles produced at low transverse momentum. CEP events are selected by triggeringon muons with transverse momentum, p T , above 400 MeV/c or electromagnetic or hadronic energyabove 1000 MeV, in coincidence with a total event charged multiplicity of less than 10 as recordedby a scintillating pad detector. A further advantage of LHCb for CEP is the low number of proton-proton interactions (typically 1.5) per beam crossing.The LHCb measurements of CEP to date have concentrated on final states with muons. Inthis report, I focus on measurements of the photoproduction of single charmonium and bottomiumas well as the production of double charmonia, which is principally produced by double pomeronexchange. The single charmonium measurements use almost 1 fb − of data taken at √ s = − at √ s = − of data, are also available on χ c production and dimuons produced by the QED diphotonexchange process [4]. New forward scintillators have been installed for the √ s =
13 TeV running,and their impact on future measurements will be briefly discussed.
2. Photoproduction of J / ψ and ψ ( S ) mesons Candidates for J / ψ mesons produced through CEP are selected [5] by requiring two identifiedmuons inside the LHCb acceptance and no photons or additional tracks in either forward or back-ward directions. The p T of the dimuon is required to be below 0.8 GeV /c , and its invariant massto be within 65 MeV/c of the known J / ψ or ψ ( S ) masses. The invariant mass of all candidates(with the mass requirement removed) is shown in Fig.1 (left). The non-resonant background, dueto the QED production of dimuons via photon propagators, is modelled with an exponential func-tion and is estimated to account for ( . ± . ) % of the J / ψ and ( ± ) % of the ψ ( S ) sample.Feed-down backgrounds inside the J / ψ mass window, amounting to ( . ± . ) %, are due to χ c or ψ ( S ) mesons decaying to J / ψ and photons that are undetected due to being very soft or goingoutside the detector acceptance. Inelastic J / ψ or ψ ( S ) production, in which the proton dissoci-ates but does not produce activity inside the LHCb acceptance, is assessed by fitting the t ≈ p T distribution and assuming that d σ / dt can be modelled by two exponentials for signal and back-ground, as assumed in Regge theory and observed at HERA [6, 7]. The fitted parameters for the1 EP at LHCb
R. McNulty exponentials are consistent with those found at HERA, having corrected for kinematic differences.In total, ( . ± . ) % of the J / ψ sample and ( ± ) % of the ψ ( S ) sample is estimated to beexclusively produced. ] Invariant mass [MeV/c E v en t s pe r M e V / c LHCb ) c ) (MeV/ - m + m m( ) c E v e n t s / ( M e V / Total 4.5 £ y £ (nS) signal, 2 ¡ Non-resonant background
LHCb
Figure 1:
Invariant mass of dimuons in (left) the charmonium analysis [5] with the J / ψ and ψ ( S ) windowsindicated and (right) the bottomonium analysis [15]. The cross-section for CEP charmonia is determined from the estimated number of CEP events,correcting for the detector efficiency and acceptance: the former is determined with tag-and-probetechniques in data while the latter is found from simulated events. The total cross-sections aregiven in Table 1 while differential cross-sections as a function of rapidity are shown in Fig. 2 com-pared to LO and approximate NLO predictions from [8]. Results on the total cross-section havebeen compared to other predictions [9, 10, 11, 12, 13] and all agree with the data. A photopro-duction cross-section can be derived from these results once rapidity gap corrections and photonflux factors are included. A two-fold ambiguity is present due to not knowing which of the protonsthe photon was radiated from. This can be resolved in a model-dependent way by assuming theH1 derived power-law for one of the solutions. The (model-dependent) cross-section derived isshown in Fig. 2, compared to results from HERA, fixed target collisions, and proton-lead collisionsat ALICE [14], in which the aforementioned ambiguity can be resolved.
3. Photoproduction of ϒ mesons Candidates for ϒ mesons produced through CEP are selected [15] by requiring two muons inLHCb and no other charged tracks. The muon pair must have p T < /c and an invariantmass between 9 and 20 GeV/c , which allows the shape of the continuum to be determined at thesame time as the fit to the ϒ ( S ) , ϒ ( S ) and ϒ ( S ) . The distribution of candidates is shown in Fig.1(right).Feed-down backgrounds coming from the various χ b states are estimated to contribute 39 ± ϒ production is assessed by fittingthe t ≈ p T distribution, but with the signal shape given by the SUPERCHIC generator[13]. Aftersubtraction of the χ b component, ( ± ) % is assessed to be exclusively produced; thus of thetotal ϒ yield, one third is CEP. The power-law for J / ψ is taken from [6] while for ψ ( S ) , I use R ( W ) = σ ( ψ ( S )) / σ ( J / ψ ) = .
166 from [7]. EP at LHCb
R. McNulty
Quantity measured Kinematic region Measurement (pb) σ ( pp → pJ / ψ p ) · BR ( J / ψ → µ µ ) < η µ , y J / ψ < . ± ±
19 [5] σ ( pp → p ψ ( S ) p ) · BR ( ψ ( S ) → µ µ ) < η µ , y ψ ( S ) < . . ± . ± . σ ( pp → p ϒ ( S ) p ) < η µ , y ϒ ( S ) < . . ± . ± . σ ( pp → p ϒ ( S ) p ) < η µ , y ϒ ( S ) < . . ± . ± . σ ( pp → p ϒ ( S ) p ) < η µ , y ϒ ( S ) < . < . σ ( J / ψ J / ψ ) < y J / ψ J / ψ < . ± ± σ ( J / ψψ ( S )) < y J / ψψ ( S ) < . + − ±
10 [20] σ ( ψ ( S ) ψ ( S ) < y ψ ( S ) ψ ( S ) < . <
237 at 90% c.l. [20] σ ( χ c χ c ) < y χ c χ c < . < σ ( χ c χ c ) < y χ c χ c < . <
45 at 90% c.l. [20] σ ( χ c χ c ) < y χ c χ c < . <
141 at 90% c.l. [20]
Table 1:
Total cross-section results for charmonia and bottomonia states. The double charmonia cross-sections include events where proton dissociation occurs.
After correcting for the detection efficiency, found using simulated events, and taking accountof the luminosity, the cross-sections are determined and reported in Table 1. The low statisticsand sizeable background impact on the significance of the 2S and 3S states. However, there aresufficient statistics to divide the 1S state into three bins of rapidity, which are plotted in Fig.2.Compared to the charmonium analysis, the difference between the LO and NLO predictions islarge and the data show a clear preference for the latter. A photoproduction cross-section can bederived from these results. Here, the smaller of the two ambiguous solutions (that contributesbetween 5% and 20%) is ignored. The result is shown in Fig.2 showing good consistency withHERA results [16, 17].
4. Production of pairs of charmonia
The pair production of charmonia has been observed inclusively in a previous LHCb analy-sis [18], where tetraquark contributions or double parton scattering (DPS) may play an importantrole. The contribution of DPS in CEP is minimal and the process proceeds dominantly throughdouble pomeron exchange. The cross-section for CEP of J / ψ J / ψ inside the LHCb acceptance ispredicted to lie in the range 2-20 pb, depending on the model used for the soft survival factor andthe gluon PDF that enters with the fourth power [19].The selection of the CEP of charmonia pairs [20] requires exactly four tracks, at least threeof which are identified as muons. The invariant masses of two unlike-sign pairs of muons is re-quired to be consistent with that of the J / ψ or ψ ( S ) mesons. The pairwise combinations areshown in Fig.3(a); 37 candidates are consistent with J / ψ J / ψ production and 5 are consistent with J / ψψ ( S ) . The invariant mass of the four tracks is shown in Fig.3(b) and agrees qualitatively withthe spectrum observed inclusively [18]. The signal is only seen when there are precisely 4 tracksin the event; it is not present when additional tracks are present in the detector. After accountingfor efficiency and luminosity, the cross-section for pairs of charmonia produced in the absence ofother charged or neutral activity in LHCb is determined and given in Table 1.3 EP at LHCb
R. McNulty
Rapidity / d y [ nb ] s d Data with uncorrelated uncertaintyData total uncertaintyLO prediction (JMRT)NLO prediction (JMRT)
LHCb (a) W (GeV) ( nb ) s LHCb (W+ solutions)LHCb (W- solutions)ALICEH1ZEUSFixed target experimentsPower law fit to H1 data
LHCb
Rapidity / d y [ nb ] s d Data with uncorrelated uncertaintyData total uncertaintyLO prediction (JMRT)NLO prediction (JMRT)
LHCb (b) W (GeV) ( nb ) s LHCb (W+ solutions)LHCb (W- solutions)H1 =0.166) y (J/ s (2S)) y ( s Power Law with R(W)=
LHCb
Υ(1 S ) rapidity D i ff e r e n ti a l c r o ss - s ec ti on ( pb ) LHCb (a)
Υ(1 S ) dataLO (7TeV)LO (8TeV)NLO (7TeV)NLO (8TeV) σ ( γ p ) ( pb ) W (GeV)
LHCb sensitivity
LHCb(b)
LONLOB.G. bCGCGauss LC bCGCH1 2000LHCb run 1ZEUS 1998/2009 Figure 2:
Differential cross-sections for exclusively produced (top left) J / ψ , (centre left) ψ ( S ) , and (bot-tom left) ϒ , compared to LO and NLO predictions [8]. Compilation of photoproduction cross-sections fromvarious experiments for (top right) J / ψ , (centre right) ψ ( S ) , (bottom right) ϒ . The LHCb results aremodel-dependent as explained in the text. Because of the low statistics, it is difficult to assess how often proton dissociation occurs. A fitto the t distribution suggests ( ± ) % of the sample is CEP implying an exclusive cross-sectionfor J / ψ J / ψ inside the LHCb acceptance of ( ± ) pb, in broad agreement with the predictions.The analysis is also sensitive to the production of pairs of χ c mesons, which would be visiblethrough their decays to a J / ψ meson and a photon. Searches are made for additional photons inevents with two J / ψ candidates. A single event is consistent with χ c χ c production but is alsoconsistent with partially reconstructed J / ψψ ( S ) production. Limits are set on pairs of tensormeson charmonia and are given in Table 1. 4 EP at LHCb
R. McNulty
Higher di-muon invariant mass [MeV] Lo w e r d i - m uon i n v a r i an t m a ss [ M e V ] LHCb
Mass [MeV] E v en t s pe r M e V LHCb
Figure 3:
Invariant mass of (a, left) dimuon pairs (b, right) the J / ψ J / ψ system [20].
5. Future prospects
Experimentally, LHCb is sensitive to about 5.5 units in pseudorapidity allowing modest ra-pidity gap sizes to be identified. The identification of CEP events would be significantly aided ifthe veto region could be increased. To this end, for the √ s =
13 TeV data-taking, forward showercounters, consisting of five planes of scintillators, have been installed perpendicular to the beam at −
114 m, − . − . +
20 m and +
114 m from the interaction point. With these, LHCb hassensitivity to particles in the regions − < η < − , − . < η < − . , . < η <
10. The effectof the additional veto on the composition of a sample of J / ψ candidates gathered at √ s =
13 TeVcan be seen in Fig. 4. The estimated proton dissociation contribution is roughly halved in the signalregion below p T = . , with an even greater suppression at higher values. ) transverse momentum squared (GeV y J/ N u m b e r o f E v e n t s p e r . G e V ) transverse momentum squared (GeV y J/ N u m b e r o f E v e n t s p e r . G e V Figure 4:
Transverse momentum squared of J / ψ candidates at √ s =
13 TeV with (left) and without (right)the use of newly installed scintillators close to the beamline [21].
Additional improvements for the √ s =
13 TeV running include a reduction in the hadronictrigger thresholds, which will allow LHCb to extend current studies to investigate CEP of lightvector mesons, a region of spectroscopic interest where the presumed dominance of the doublepomeron exchange mechanism provides a gluon-rich laboratory to search for glueballs and satura-tion effects. 5
EP at LHCb
R. McNulty
References [1] M.G. Albrow, T.D. Coughlin, J.R. Forshaw,
Prog.Part.Nucl.Phys.
149 (2010), arXiv:1006.1289.[2] M.G. Albrow,
Int.J.Mod. Phys.
A29
JINST S08005 (2008).[4] LHCb collaboration, CERN-LHCb-CONF-2011-022 (2011).[5] LHCb collaboration,
J.Phys.
G41
Eur.Phys.J.
C73
Phys.Lett.
B541
251 (2002), hep-ex/0205107.[8] S.P. Jones et al.,
JHEP
085 (2013), arXiv:1307.7099;
J.Phys.
G41
Phys. Rev.
C84
Phys. Rev.
D78
Phys.Rev.
D76
Phys.Rev.Lett. Eur.Phys.J.
C76
Phys.Rev.Lett.
JHEP
084 (2015), arXiv:1505.08139.[16] H1 collaboration,
Phys.Lett.
B483 (2000) 23, hep-ex/0003020.[17] ZEUS collaboration,
Phys.Lett.
B680
Phys.Lett.
B707
52 (2012), arXiv:1109.0963.[19] L.A. Harland-Lang et al.,
J.Phys.
G42
J.Phys.
G41115002 (2014), arXiv:1407.5973.[21] LHCb collaboration, LHCb-CONF-2016-007 (2016).