Contents lists available atScienceDirect
Physics Letters B
www.elsevier.com/locate/physletb
Observation of B
0smesons and measurement of the B
0s/B
+yield ratio in PbPb collisions at √
s
nn= 5 . 02 TeV
.The CMSCollaboration
CERN,Geneva,Switzerland
a rt i c l e i n f o a b s t r a c t
Articlehistory:
Received4September2021
Receivedinrevisedform21March2022 Accepted23March2022
Availableonline4April2022 Editor: M.Doser
Keywords:
CMS
Quarkgluonplasma Heavyflavor Bmesons Hadronization
The B0s and B+ production yields are measured in PbPb collisions at a center-of-mass energy per nucleonpairof5.02 TeV. The data sample, collectedwith theCMS detector atthe LHC, corresponds toanintegratedluminosityof1.7 nb−1.Themesonsarereconstructedintheexclusivedecaychannels B0s →J/ψ(μ+μ−)φ(K+K−)andB+→J/ψ(μ+μ−)K+,inthetransversemomentumrange7–50 GeV/c and absoluterapidity0–2.4.The B0s mesonis observedwithastatistical significance inexcessoffive standarddeviationsforthefirsttimeinnucleus-nucleuscollisions.Themeasurementsareperformedas functionsofthetransversemomentumoftheBmesonsand ofthePbPbcollisioncentrality.Theratio ofproductionyieldsofB0s andB+ ismeasuredand comparedtotheoreticalmodelsthatincludequark recombinationeffects.
©2022TheAuthor(s).PublishedbyElsevierB.V.ThisisanopenaccessarticleundertheCCBYlicense (http://creativecommons.org/licenses/by/4.0/).FundedbySCOAP3.
1. Introduction
Relativistic heavy ion collisions allow the study of quantum chromodynamics(QCD)athighenergydensitiesandtemperatures.
Undersuchconditions,astateinwhichquarksandgluonsarethe relevant degreesof freedom,the quark-gluonplasma (QGP) [1,2], is formed [3] aspredicted by lattice QCD calculations [4]. Multi- ple probes are necessary forcharacterizing the properties of the QGP medium. Among these, heavy quarks, whichare abundantly produced attheCERN LHC,havethe potentialof providingnovel insightsintoQCDcalculations, servingasprobesoftheQGP [5,6].
AstheytraversetheQGP,thesehard-scatteredpartonsloseenergy by means of elastic collisions andmedium-induced gluon radia- tion [7–10].The study ofpartonenergyloss canprovideinsights into the energydensity anddiffusion propertiesof theQGP. The fullreconstructionofbeautyandcharmgivesaccesstotheirfour- momentaandallowsthestudyoftheflavorandmassdependences ofsuchprocesses.Althoughheavyflavordatabecameavailableat BNL RHICandatthe LHC [5],there arestill large theoreticalun- certaintiesonthehadronizationofheavyquarksinthepresenceof theQGP [11].ToextractthetransportpropertiesoftheQGPusing heavyquarks,anunderstandingofthehadronizationmechanismis necessary.
E-mailaddress:cms-publication-committee-chair@cern.ch.
An enhancement in the strangeness content of the medium formed in heavy ion collisions [12–18] is expected when the medium temperature lies above the strange quark mass [19].
Throughaquark-recombinationmechanism [20–23],acorrespond- ing yield enhancement of hadrons containing strangeness is ex- pected relative to corresponding hadrons that do not contain strange quarks.Hints ofthepresence ofsuchrecombinationpro- cessesforheavymesons containingstrangeness were recentlyre- portedforopen-charm [24] andopen-beauty [25] mesonsinlead- lead(PbPb)collisionsatanucleon-nucleoncenter-of-massenergy of√
snn=5.02 TeVattheLHC.
TheproductionofB mesonsemployingfullyreconstructedde- cays was studied at the LHC in proton-proton (pp) collisions at center-of-mass energies, √
s, of 7, 8 and 13 TeV [26–40] over wide transverse momentum (pT) and rapidity ( y) intervals, and inproton-lead(pPb) collisions at√
snn of5.02and8.16 TeV [41, 42].In PbPb collisions, the CMSCollaboration reportedresults at
√snn=5.02 TeVfortheB+[43] andB0s [25] mesons.InthisLetter, weextendtheseresultsbyemploying thePbPbdatasetcollected attheendof2018, asample largerby aboutafactorofthree as compared to the first measurement with 2015 data,resulting in increasedstatisticalprecisionandsignificance.
TheB+ (ub)¯ andB0s (sb)¯ mesons arereconstructedvia theex- clusivedecay channelsB0s→J/ψφ(1020) andB+→J/ψK+, with J/ψ→ μ+μ− andφ (1020)→K+K−. Throughout thisLetter, un- lessotherwise specified, the y and pT variables given are those of the B mesons, and charge-conjugate states are implied. The
https://doi.org/10.1016/j.physletb.2022.137062
0370-2693/©2022TheAuthor(s).PublishedbyElsevierB.V.ThisisanopenaccessarticleundertheCCBYlicense(http://creativecommons.org/licenses/by/4.0/).Fundedby SCOAP3.
production yields ofthe B0s andB+ mesonsin PbPb collisions at
√snn=5.02 TeV andtheir ratios are reported, using a data set correspondingtoanintegratedluminosityof1.7 nb−1collectedby the CMS experiment. The measurements are performed as func- tions ofthemeson pT andofthePbPbcollision centrality,where thelattercharacterizesthedegreeofoverlapofthetwoleadnuclei andisrelatedtothetemperatureofthecreatedmedium.
2. Experimentalapparatusanddatasample
The central feature of the CMS detector is a superconduct- ing solenoid, which provides a magnetic field of 3.8 T. Within the solenoid volume are a silicon trackerthat measures charged particles in the pseudorapidity range |η|<2.5, a lead tungstate crystal electromagnetic calorimeter, and a brass and scintillator hadroncalorimeter.Forchargedparticlesof1<pT<10 GeV/c and
|η|<1.4,thetrackresolutionsare typically1.5%in pT and25–90 (45–150) μm in the transverse (longitudinal) impact parameter [44]. Muons are measured inthe range |η|<2.4, withdetection planes made using three technologies: drift tubes, cathode strip chambers, andresistive-platechambers. The muonreconstruction algorithmstartsbyfindingtracksinthemuondetectors,whichare thenfittedtogetherwithtracksreconstructedinthesilicontracker to formglobalmuons. Matchingmuonsto tracksmeasured inthe silicon trackerresults in a relative pT resolutionfor muonswith 20<pT<100 GeV/c of1.3–2.0%inthebarrelandbetterthan6%
in the endcaps. The hadron forward (HF) calorimeter uses steel as an absorber and quartz fibers as the sensitive material. The two halves ofthe HF are located 11.2 m away from the interac- tion point, one on each end, providing together coverage in the range3.0<|η|<5.2.In thisanalysis, theHFinformationis used toselectPbPbcollisioneventsandtodefinetheir centralityclass.
Centrality, defined as the fraction of the total inelastic hadronic crosssectionwith0%representingcollisionswiththelargestover- lapofthetwonuclei,isdeterminedexperimentallyusingthetotal energy in both HF calorimeters [45]. Events are filtered using a two-tiered trigger system [46]. The first level (L1), composed of customhardware processors,usesinformationfromthecalorime- ters andmuon detectorsto selectevents within a fixed time in- tervaloflessthan4 μs.Thesecondlevel,knownasthehigh-level trigger(HLT),consistsofafarmofprocessorsrunningaversionof the fullevent reconstructionsoftware optimizedforfastprocess- ing.AdetaileddescriptionoftheCMSexperimentandcoordinate systemcanbefoundinRef. [47].
SeveralsamplesofMonteCarlo(MC)simulatedeventsareused toevaluatebackgroundcomponents,signalefficiencies anddetec- tor acceptance corrections. The simulations include signal sam- plescontainingonlytheBmesondecaychannelsbeingmeasured, andbackground samplesofother b hadron decaysalso involving J/ψ mesons in the final state. Simulated proton-protoncollisions are generated with pythia8 v212 [48] tune CUETP8M1 [49] and propagated through the CMS detector model using the Geant4 package [50]. The decay of the B mesons is modeled with evt- gen1.3.0 [51], and final-state photon radiation is simulated with photos2.0 [52].Each pythia8eventisembeddedintoasimulated PbPbcollisioneventgeneratedwith hydjet1.8 [53],whichistuned to reproduceglobaleventproperties,suchasthechargedhadron pT spectrumandparticlemultiplicity [54].
The PbPbcollision eventswere collectedwithan L1hardware trigger requiring the presence of two muon candidates, with no explicit momentum threshold, in coincidence withlead bunches crossingattheinteractionpoint.AttheHLToneofthemuoncan- didates is reconstructed using information both from the muon detectors andthe inner tracker,while only informationfrom the muondetectorsisrequiredfortheothermuoncandidate.Thetwo muon candidateswere requiredtohavean invariantmasswithin
1<mμμ<5 GeV/c2.Fortheofflineanalysis,eventshavetopassa setofselectioncriteriadesignedtorejecteventsfrombackground processes (beam-gas collisions), asdescribed in Ref. [54]. Events arerequiredtohaveatleastonereconstructedprimaryinteraction vertex, formed by two ormore tracks, with a distance fromthe centerofthenominalinteractionregion oflessthan 15 cm along the beam axis. The shapes of the clusters in the pixel detector mustbe compatiblewiththose expectedfromparticles produced byaPbPbcollision [55].Inordertoselectinelastichadroniccolli- sions,thePbPbeventsarealsorequiredtohaveatleasttwotowers in each of the HF detectors with energy deposits of more than 4 GeVpertower. Onlyeventswithcentrality0–90%areselected.
Collisioncentrality bins are givenin percentageranges, withthe 0–90% bincorresponding to the90% ofthe collisions having the largestoverlapofthetwonuclei.ThePbPbsamplecorrespondsto anintegratedluminosity ofapproximately1.7 nb−1.Thisvalueis indicativeonly, asthePbPbyieldisnormalizedbythetotalnum- berofminimumbiaseventssampled,NMB=11.8×109.
3. Candidateselectionandsignalextraction
Muoncandidatesareselectedwithinkinematicconstraintsthat ensurethat theirreconstruction efficiencystaysabove 10%.These limitsarepμT >3.5 GeV/c for|ημ|<1.2,pμT >1.5 GeV/c for2.1<
|ημ|<2.4,andpμT > (5.47−1.89|ημ|) GeV/c inthe1.2<|ημ|<
2.1 region.Themuonsarealsorequiredtomatchthetrigger-level muon candidates, and softmuon selection criteriaare applied to globalmuons (i.e. muons reconstructedusingthe combinedinfor- mationofthetrackerandmuondetectors),asdefinedinRef. [56].
Two muons of opposite charge, with an invariant mass within
±150 MeV/c2 of theworld-averageJ/ψmesonmass [57],are se- lected to reconstructa J/ψ candidate, with a mass resolution of typically15–30 MeV/c2,dependingonthedimuonrapidityandpT. The muon pairs are fitted witha common vertexconstraint and arekeptifthep-valueofthe χ2 ofthefitisgreaterthan1%,thus loweringthebackgroundfromcharmandbeautyhadronsemilep- tonicdecays.Similarly,the φ(1020) mesoncandidatesare formed withacommonvertexconstraintbetweentwooppositelycharged particletracks with pT>300 MeV/c, both requiredto passstan- dard selection criteriadescribed in Ref. [54]. The invariant mass, assuming theworld-average charged kaonmass [57] forthe two tracks,witharesolutionof∼3.9 MeV/c2,isrequiredtobewithin 15 MeV/c2oftheworld-averageφ(1020)mesonmass [57].
TheB0s (B+)mesoncandidatesareconstructedbycombininga J/ψcandidatewithaφ(1020)(singletrack)candidate,andrequir- ing that they originate from a common vertex. In the kinematic vertexfittothedimuon plustwo-track(single-track)system, the tracks are assigned the mass of the charged kaon, and the in- variantmassof themuon pairis constrainedtothe nominalJ/ψ mesonmass [57].TheBcandidateselectionisperformedviamul- tivariate discriminators, based on a boosted decision tree (BDT) method [58]. The selection is optimized separately for each me- son,aswell aseachindividualbinof pT.Thediscriminatingvari- ables employed include: the χ2 probability of the decay vertex (the probability for the muon tracks fromthe J/ψ meson decay andthe other chargedparticle track(s) to originate froma com- monvertex),thedecaylength(normalizedbyitsuncertainty),the pointing angle (the angle between the line segment connecting theprimary anddecayverticesandthemomentum vectorofthe B meson), the cosine value of the angle between the B meson displacement and momentum in the transverse plane, the two- dimensional(2D)distancebetweentheprimaryanddecayvertices of their daughter tracks (normalized by their uncertainties), and the pT ofthe daughtercharged particle track(s). For theB0s me- son,twoadditionalselectionvariablesareused:thedistancealong the z-directionbetweenthe primaryvertexandthe decayvertex
(normalizedby itsuncertainty),aswellastheabsolutedifference betweenthereconstructed φ(1020)mesoninvariant massandits nominalvalue.
The BDT training is performed employing simulated B signal samples,andbackgroundsamplestakenfromdatasidebands(can- didates withthe invariant mass0.15–0.35 GeV/c2 away fromthe B meson nominal mass [57]). The signal samples are scaled to the number of B candidates predicted by fixed-order plus next- to-leadingorderlogarithmic(FONLL)perturbativeQCDcalculations [59–61] fortheintegratedluminosityoftheanalyzeddatasample.
ThebackgroundrejectionachievedwiththeBDTselectionislarger than97%.
TherawBmesonsignalyieldsareextractedusinganextended unbinned maximum likelihood fit to the invariant mass spectra.
Thesignal shapeismodeledby asumoftwo Gaussianfunctions, with parameters determined from MC simulation, except for the commonmeanandtheoverallsignalyield,whicharefreeparam- eters ofthe fit.The combinatorialbackground,fromuncorrelated combinationsofJ/ψcandidateswithextraparticles,givesrisetoa fallingcontributionintheinvariantmassspectrumthatismodeled by an exponential function.An additional backgroundsource can arisefrompossiblecontaminationfromotherbhadrondecays.For the B+ spectrum,partially reconstructedB decays,such asB0→ J/ψK*0→ μ+μ−K+π−,leadtoaheightenedbackgroundinthein- variant mass region below 5.2 GeV/c2. The decay B+→J/ψπ+, wherethepionismisidentifiedasakaon,resultsinasmallpeak- ingstructureinthesignalregion.Suchpartiallyreconstructedand misreconstructedbhadroncomponentsaremodeledfromsimula- tion, via an error function and a Gaussian function, respectively;
bothshapesandthenormalizationoftheGaussianfunction(rela- tivetothesignal)arefixedinthefittothedata.FortheB0s meson case,such backgroundcontributionsarefoundtobe negligible,as a consequence ofthe tightselection on the massof theφ(1020) candidate. The results of the fits to the invariant mass distribu- tionsareshowninFig.1.
The statisticalsignificanceofthe Bmesonsignalsisestimated fromtheratiooflikelihoodsobtainedbyfittingthedatawiththe full model and the background-only model. Using the standard asymptoticapproximationforthelikelihood [62],theestimateob- tainedfortheB0s mesoncorrespondstoapproximately16standard deviations.The estimate isobtainedwith thenominalsignal and background model described above, andverified with the model variations presentedin Section 5. This is thefirst observation of theB0s signalinheavyioncollisions.
4. Productionyields
For each B meson, the production yield scaled by the over- lapnuclear function, TAA [63],andthenumberofminimum bias events,NMB,iscomputedineach pTinterval,accordingto
1 TAA
dN
dpT = 1 2BNMBTAA
Nobs(pT)
pT 1
α(pT,y)(pT,y), (1) where Nobs denotestherawsignalyield,extractedineach pT in- terval ofwidthpT.Thefactor1/2accountsforthefactthat the raw yield is measured for particles and antiparticles added to- gether while the production yield is given for one species only, andBstandsforthebranchingfractionofthecorrespondingdecay chain.The factor TAA isequal to thenumberof nucleon-nucleon (NN) binary collisionsdividedby theNNtotalinelasticcrosssec- tion. The value of TAA can be interpreted as the NN-equivalent integrated luminosity per heavy ion collision. The TAA values for √
snn=5.02 TeV PbPb collisions for the centrality intervals 0–30%, 30–90%, and 0–90% are, respectively, (15.4±0.3) mb−1, (1.7±0.1) mb−1,and(6.3±0.1)mb−1,asestimatedfromaMC
Fig. 1. InvariantmassdistributionsofB+ (upper)andB0s (lower)candidates,after theBDTselection,foreventcentralityintherange0–90%.Thelowerpanelsshow thepulls,obtainedasthedifference betweenthedatapoints andthefit result, dividedbytheuncertaintyindata.
Glaubermodel [63],implementedinthe tglaubermc v3.2software package [64]. Themeasurement isperformedinasimilar fashion asa functionof thePbPb collision centrality. The centralityclass is represented as the average number of participating nucleons,
Npart.Thisnumber,determined usingtheabove Glaubermodel, ishighlycorrelatedwiththeimpactparameterofthecollision.
The acceptance and the efficiency factors, the last term in Eq. (1), are obtainedemploying 2D fine-grainedmaps ofthe ac- ceptance α(pT,y)andtheefficiency (pT,y).Thesemapsarede- termined,for each centrality region, fromMC-simulated samples of B meson signal events, generated within the analysis fiducial regiongivenby: |y|<1.5 and pT>10 GeV/c;1.5<|y|<2.4 and 7<pT<50 GeV/c.The acceptancecorresponds tothefractionof generatedeventspassingthemuon andkaonselectionthresholds specified in Section 3, while the efficiency is determined as the fractionofacceptedeventsthatpassthefullanalysisselectioncri- teria, including the trigger. The maps are used to determine the 1/(α)valueforeachBcandidateindata,basedonthekinematics (pT,y)ofthecandidates,andthecorrespondingaverage1/(α) obtained for the candidates within a 160 MeV/c2 window cen- teredon the Bmeson’s world-average mass.The efficiencymaps
