Observation of Structure in the J/Ψ-Pair Mass Spectrum
Total Page:16
File Type:pdf, Size:1020Kb
EUROPEAN ORGANIZATION FOR NUCLEAR RESEARCH (CERN) CERN-EP-2020-115 LHCb-PAPER-2020-011 CERN-EP-2020-115June 30, 2020 22 June 2020 Observation of structure in the J= -pair mass spectrum LHCb collaborationy Abstract p Using proton-proton collision data at centre-of-mass energies of s = 7, 8 and 13 TeV recorded by the LHCb experiment at the Large Hadron Collider, corresponding to an integrated luminosity of 9 fb−1, the invariant mass spectrum of J= pairs is studied. A narrow structure around 6:9 GeV/c2 matching the lineshape of a resonance and a broad structure just above twice the J= mass are observed. The deviation of the data from nonresonant J= -pair production is above five standard deviations in the mass region between 6:2 and 7:4 GeV/c2, covering predicted masses of states composed of four charm quarks. The mass and natural width of the narrow X(6900) structure are measured assuming a Breit{Wigner lineshape. Submitted to Science Bulletin c 2020 CERN for the benefit of the LHCb collaboration. CC BY 4.0 licence. yAuthors are listed at the end of this paper. ii 1 1 Introduction 2 The strong interaction is one of the fundamental forces of nature and it governs the 3 dynamics of quarks and gluons. According to quantum chromodynamics (QCD), the 4 theory describing the strong interaction, quarks are confined into hadrons, in agreement 5 with experimental observations. The quark model [1,2] classifies hadrons into conventional 6 mesons (qq) and baryons (qqq or qqq), and also allows for the existence of exotic hadrons 7 such as tetraquarks (qqqq) and pentaquarks (qqqqq). Exotic states provide a unique 8 environment to study the strong interaction and the confinement mechanism [3]. The first 9 experimental evidence for an exotic hadron candidate was the χc1(3872) state observed 10 in 2003 by the Belle collaboration [4]. Since then a series of novel states consistent 11 with containing four quarks have been discovered. Recently, the LHCb collaboration 12 observed resonances interpreted to be pentaquark states [5{8]. All hadrons observed to 13 date, including those of exotic nature, contain at most two heavy charm (c) or bottom (b) 14 quarks, whereas many QCD-motivated phenomenological models also predict the existence 15 of states consisting of four heavy quarks, i.e. T , where Qi is a c or a b quark [9{33]. Q1Q2Q3Q4 16 Theoretically, the interpretation of the internal structure of such states usually assumes 17 the formation of a diquark (Q1Q2) and an antidiquark (Q3Q4) attracting each other. 18 Application of this diquark model successfully predicts the mass of the doubly charmed ++ 19 baryon Ξcc [34,35] and helps to explain the relative rates of bottom baryon decays [36]. 20 Tetraquark states comprising only bottom quarks, Tbbbb, have been searched for by the + − 21 LHCb and CMS collaborations in the Υ µ µ decay [37,38], with the Υ state consisting 22 of a bb pair. However, the four-charm states, Tcccc, have not yet been studied in detail 23 experimentally. A Tcccc state could disintegrate into a pair of charmonium states such 24 as J= mesons, with each consisting of a cc pair. Decays to a J= meson plus a heavier 25 charmonium state, or two heavier charmonium states, with the heavier states decaying 26 subsequently into a J= meson and accompanying particles, are also possible. Predictions 2 27 for the masses of Tcccc states vary from 5:8 to 7:4 GeV/c [9{24], which are above the masses 28 of known charmonia and charmonium-like exotic states and below those of bottomonium 29 hadrons. This mass range guarantees a clean experimental environment to identify possible 30 Tcccc states in the J= -pair (also referred to as di-J= ) invariant mass (Mdi-J= ) spectrum. 31 In proton-proton (pp) collisions, a pair of J= mesons can be produced in two separate 32 interactions of gluons or quarks, named double-parton scattering (DPS) [39{41], or in a 33 single interaction, named single-parton scattering (SPS) [42{49]. The SPS process includes 34 both resonant production via intermediate states, which could be Tcccc tetraquarks, and 35 nonresonant production. Within the DPS process, the two J= mesons are usually 36 assumed to be produced independently, thus the distribution of any di-J= observable can 37 be constructed using the kinematics from single J= production. Evidence of DPS in pp 38 collisions has been found in studies at the Large Hadron Collider (LHC) experiments [50{54]. 39 The LHCb experimentp has measured the di-J= production in pp collisions at centre-of- 40 mass energies of s = 7 [55] and 13 TeV [56]. The DPS contribution is found to dominate 41 the high Mdi-J= region, in agreement with expectation. 42 In this paper, fully charmed tetraquark states Tcccc are searched for in the di-J= 1 p 43 invariant mass spectrum, using pp collision data collected by LHCb at s = 7; 8 and −1 44 13 TeV, corresponding to an integrated luminosity of 9 fb . The two J= candidates in + − 45 a pair are reconstructed through the J= ! µ µ decay, and are labelled randomly as 46 either J= 1 or J= 2 . 47 2 Detector and data set 48 The LHCb detector is designed to study particles containing b or c quarks at the LHC. It is 49 a single-arm forward spectrometer covering the pseudorapidity range 2 < η < 5, described 50 in detail in Refs. [57, 58]. The online event selection is performed by a trigger, which 51 consists of a hardware stage, based on information from the calorimeter and muon systems, 52 followed by a software stage, which applies a full event reconstruction. At the hardware 53 stage, events are required to have at least one muon with high momentum transverse to 54 the beamline, pT. At the software stage, two oppositely charged muon candidates are 55 required to have high pT and to form a common vertex. Events are retained if there is 56 at least one J= candidate selected by both the hardware and software trigger stages. 57 Imperfections in the description of the magnetic field and misalignment of subdetectors 58 lead to a bias in the momentum measurement of charged particles, which is calibrated + 59 using reconstructed J= and B mesons [59], with well-known masses. 60 Simulated samples are used to model the signal properties, including the invariant 61 mass resolution and the reconstruction efficiency. In the simulation, pp collisions are 62 generated using Pythia [60] with a specific LHCb configuration [61]. Decays of unstable 63 particles are described by EvtGen [62], in which final-state radiation is generated using 64 Photos [63]. The interaction of the generated particles with the detector and its response 65 are implemented using the Geant4 toolkit [64], as described in Ref. [65]. 66 3 Candidate selection 67 In the offline selection, two pairs of oppositely charged muon candidate tracks are recon- 68 structed, with each pair forming a vertex of a J= candidate. Each muon track must have 69 pT > 0:65 GeV/c and momentum p > 6 GeV/c. The J= candidates are required to have a 2 70 dimuon invariant mass in the range 3:0 < Mµµ < 3:2 GeV/c . A kinematic fit is performed 71 for each J= candidate constraining its vertex to coincide with a primary pp collision 72 vertex (PV) [66]. The requirement of a good kinematic fit quality strongly suppresses 73 the contamination of di-J= candidates stemming from feed-down of b-hadrons, which 74 decay at displaced vertices. The four muon tracks in a J= -pair candidate are required to 75 originate from the same PV, reducing to a negligible level the number of pile-up candidates 76 with the two J= candidates produced in separated pp collisions. Fake di-J= candidates, 77 comprising two muon-track candidates reconstructed from the same real particle, are 78 rejected by requiring muons of the same charge to have trajectories separated by an angle 79 inconsistent with zero. For events with more than one reconstructed di-J= candidate, 80 accounting for about 0.8% of the total sample, only one pair is randomly chosen. 2 ) 2 c 3000 Data LHCb Total fit 2500 J/ψ +J/ψ 1 2 bkg +J/ψ 1 2 2000 J/ψ +bkg 1 2 bkg +bkg 1500 1 2 Candidates / (2 MeV/ 1000 500 0 ] 3200 2 2 ) 2 250 [MeV/c µ 3150 (1) µ 200 M 150 3100 Candidates/(5 MeV/c 100 3050 50 3000 0 3000 3050 3100 3150 3200 500 3000 2500 2000 1500 1000 (2) 2 Candidates / (2 MeV/c2) M µµ [MeV/c ] (1) (2) Figure 1: (Bottom right) Two-dimensional (Mµµ ;Mµµ ) distribution of di-J= candidates and its (1) (2) projections on (bottom left) Mµµ and (top) Mµµ . Four components are present as each projection consists of signal and background J= candidates. The labels J= 1;2 and bkg1;2 represent the (1);(2) signal and background contributions, respectively, in the Mµµ distribution. 81 The di-J= signal yield is extracted by performing an extended unbinned maximum- 82 likelihood fit to the two-dimensional distribution of J= 1 and J= 2 invariant masses, (1) (2) 83 (Mµµ ;Mµµ ), as displayed in Fig. 1, where projections of the data and the fit result are 84 shown. For both J= candidates, the signal mass shape is modelled by a Gaussian kernel 85 with power-law tails [67]. Each component of combinatorial background, consisting of 86 random combinations of muon tracks, is described by an exponential function.