Search for New Physics with Rare Heavy Flavour Decays at LHCb 1 1 0 2 n Giampiero Mancinelli∗† a J CPPM,Aix-MarseilleUniversité,CNRS/IN2P3,Marseille,France 5 E-mail: [email protected] 2 ] x TheLHCbexperimenthasthepotential,duringthe2010-11run,toobservetheraredecayB0→ s e µ+µ−orimprovesignificantlyitsexclusionlimits. Thisstudywillprovideverysensitiveprobes - p ofNewPhysics(NP)effects. HighsensitivitytoNPcontributionsisalsoachievedbymeasuring e h photonpolarizationbyperformingatimedependentanalysisofB0→φγ,andbyanangularstudy s [ ofthedecayB0d→K∗0µ+µ−. Preparationsfortheseanalysesarepresentedandstudiesshownof 1 howexistingdata,forexamplepromptJ/ψ events,canbeusedtovalidatetheanalysisstrategy. v 8 3 8 4 . 1 0 1 1 : v i X r a 35thInternationalConferenceofHighEnergyPhysics-ICHEP2010, July22-28,2010 ParisFrance ∗Speaker. †onbehalfoftheLHCbCollaboration (cid:13)c Copyrightownedbytheauthor(s)underthetermsoftheCreativeCommonsAttribution-NonCommercial-ShareAlikeLicence. http://pos.sissa.it/ RareDecaysatLHCb GiampieroMancinelli 1. Introduction The study of rare B decays is a completely complementary approach with respect to the one from direct searches being carried out at general purpose detectors at the LHC. In fact, yet undis- covered particles can be produced and observed as real particles, or they can appear as virtual particles in loop processes, whose effects can be observed as deviations from the Standard Model (SM) expectation. This can be done via CP violation tests or with the help of very rare decays which are the topic of this contribution. In particular we give an overview of the main discovery channels: B0→µ+µ−,B0→K∗µ+µ−andB0→φγ. TheseareprocessesinvolvingFlavorChang- s d s ingNeutralCurrents(FCNC).Theyconstitutetheidealenvironmentforindirectsearches, asthey are suppressed in the SM and only realized via boxes or penguin diagrams. Here the "pollution" from the SM is typically very small and the supposedly small effects from NP can show up at the same,orevenhigher,levelwithrespecttotheSMones. Furthermore,wereNPtobediscoveredby CMSorATLAS,itsrealstructuremightonlybeunderstoodviaindirectmeasurements,whichcan giveaccesstothephasesofthenewcouplings,hencetotheflavourstructureofNP. To study these very rare processes high statistics are needed. This year’s goal for the LHC is toreachaninstantaneousluminosityof1032 cm−2 s−1,whichisonlyafactor2lessthantheLHCb nominalworkingvalue. Theexpectedintegratedluminositybytheendof2011is1fb−1. Mostof theresultspresentedhereuse15nb−1 ofdata. TheLHCbdetectorisaonearmspectrometerabout 20 m long covering about 300 mrad in polar angle around the beams. It has all the elements of a typicalmultipurposedetector. ItisadetectoroptimizedforBphysics,withexcellentspatialresolu- tion and vertex separation, needed for time dependent measurements, good particle identification, crucialfortriggerandflavor-tagging,andremarkablemomentum,hencemass,resolution[1]. 2. RareDecays In the following we present the preparation for some of the key measurements involving rare B decays ongoing with the very first data collected at the LHC. More details can be found in the LHCbroadmapdocument[2]. 2.1 SearchforB0→µ+µ− s Within the SM the dominant contribution for this mode stems from the Z−penguin diagram, while the box diagram is suppressed by a factor of |M /m |2. This is a FCNC mode and is also W t helicitysuppressed,hencetheSMexpectationisonly3.210−9 [3]. Theuncertaintyonthetheoret- ical expectation is small as well, which makes this mode very attractive as a SM test bench. This mode is very sensitive to NP, especially in models with an extended Higgs sector and high tanβ. For example in the MSSM the B0 → µ+µ− branching ratio (BR) goes as (tanβ)6. The current s 90%CLlimitfromtheTevatronis36×10−9 [4],hencethereislargemarginfordiscovery. With2 fb−1 weexpecttocoveralltheinterestingparameterspaceintheNUHMmodel[5]. InLHCb,the µ+µ− pairsareselectedontriggeredeventswithcriteriaveryclosetotheones used for the control samples in order to keep systematic uncertainties as low as possible. Each B0 →µ+µ− candidateisgivenalikelihoodtobesignalorbackground-likeina3Dspaceformed s by the ensemble of geometrical event variables, the invariant mass, and the particle identification 2 RareDecaysatLHCb GiampieroMancinelli (PID) information. The three likelihoods are largely uncorrelated. To translate the number of observedcandidatesintoaBRmeasurement,anormalizationchannelisneeded(asB0 →h+h(cid:48)− d/s or B− → J/ψK−), as well as the knowledge of the relative efficiencies. The largest systematic uncertainty(13%)comestotheuncertaintyontheB andB hadronizationfractions, f and f . s d/u s d/u A new method to measure f /f with B→DK and B →D π decays is suggested [6] and should s d s s lowerthissystematicaswellasallowameasurementof f /f atLHCenergies. s d y y nc 1 nc 1 e e ci ci effi 0.9 effi 0.9 0.8 0.8 0.7 0.7 0.6 corrected data 0.6 PLHreClibminary ∈trg(Bs→ µ µ)MC, direct MC s = 7 TeV Data ∈trg(Bs→ µ µ)data, via J/ψ 0.5 0.5 0 5000 10000 15000 0 5000 10000 15000 p(µ)+p(µ) p(µ)+p(µ) T 1 T 2 T 1 T 2 Figure1: TriggerefficiencyasafunctionoftheP Figure2: TriggerefficiencyasafunctionoftheP T T of the J/ψ candidates for data (black) and simu- oftheBs forsimulation(green)andfordata,based latedsignal(blue). ontheJ/ψ controlsample(blue). Afirstexampleoftheuseofcontrolsamplesforthedeterminationofthetriggerefficiencyis shown in figures 1 and 2. The first shows the trigger efficiency for J/ψ →µ+µ− decays in data and Monte Carlo (MC) simulation, which are in reasonable agreement. With this knowledge we canconvolutetheefficiencieswiththespectrumofthetransversemomentum(P )oftheB signal T s fromsimulationtoobtaintheexpectedefficienciesshowninfigure2. Accordingtothesimulation, this procedure is accurate to 1.3% and the efficiency we can expect from data for B0 → µ+µ− s is around 99% with the current trigger settings. J/ψ, K , and Λ decays are used to calculate s PID performances. Data and simulation are already in very good agreement. The expected mass resolution is about 20 MeV/c2, though currently the performances on data are not yet as good as what is expected from the simulation in the J/ψ control sample; ultimately B0 →K+K− will be s usedtoextracttheresolutionfromthedata. Figure3:GLdistributionfortheK0controlsample Figure4:GLdistributionfortheD0→K−π+con- s fordata(red)andsimulation(blue). trolsamplefordata(red)andsimulation(blue). 3 RareDecaysatLHCb GiampieroMancinelli Thegeometricallikelihood(GL)combinesinformationfromobservableslikeimpactparame- ters,vertexχ2,andmuonisolation. Developedwiththehelpofsimulateddata,thesignalresponse will eventually be calibrated with B → hh(cid:48) events. For the moment, the machinery is tested on twobodydecaysoflonglivedlowermassresonances. ThedistributionsoftheGLforsignaldata and simulation for K →π+π− and D0 →K−π+ are shown in figures 3 and 4: they are in good s but, at the moment, not perfect agreement. All studies done on data so far confirm the sensitivity calculated on simulated data. Furthermore the background level (see figure 5) appears to be well simulated,thoughthedimuondataathighmassesisatthemomentverylimited. Figure5: µ+µ−pairscandidatesmassdistribution Figure6: Expectedexclusionlimitsat90%CLfor fordata(dots)andsimulation(blue). theB0→µ+µ−BRasafunctionoftheluminosity. s In conclusion, in absence of signal, the current 90% CL limit from CDF will be improved withlessthan100pb−1 ofdataandwithlessthat200pb−1 LHCbshouldreachthefinalexpected Tevatronlimit(seefigure6). AttheendofnextyearLHCbshouldbeabletoexcludeanysignificant enhancementfromtheSMat90%CL.InpresenceofasignalattheendofnextyearLHCbshould beabletoobserve, at5σ, asignalaslargeas3.5theSMexpectation, whiletoreachtheSMBR, 10 fb−1 at 14 TeV are needed. NP discovery is expected with just 1 fb−1 for BR already as small as17−20×10−9 dependingonhowwell f /f canbemeasured. d s 2.2 AsymmetriesinB0 →K∗0µ+µ− decays d Thisdecay,muchlessrare(ithasaBRabout1.0×10−6 intheSM),issensitivetomagnetic, vector,andaxialsemileptonicpenguinoperators,whosecoefficientsappearinthedifferentialdecay rate formula. Depending on the NP hypotheses, the differential decay rate as a function of the dimuonmasstakesadifferentform,hence,eventhoughthetotalBRmeasurementsagreewiththe SMpredictionwithin20%,NPcontributionscanaffectothercompositevariableandasymmetries. Inparticular,theForward-BackwardAsymmetry(A )intheµ+µ−restframeasafunctionofthe FB q2 ofthemuonswillbethefocusofthefirstanalyseswiththismodeinLHCb. WhatisenticinginthismeasurementisthatatthezeropointofA ,thedominanttheoretical FB uncertainties due to form factors calculations cancel out. Assuming that LHCb finds the same central value of A at low q2 found by the Belle collaboration [7], with 100 pb−1 LHCb would FB achieve an uncertainty similar to those of current best measurements and by the end of next year, havingcollectedabout1400events,itmightreacha4σ discrepancywithrespecttotheSM. 4 RareDecaysatLHCb GiampieroMancinelli 2.3 PhotonpolarizationwithB0→φγ andB0 →K∗0l+l− decays s d Already observed by the Belle Collaboration [8] this process B0 →φγ is even less rare than s previousones. TheinclusiveresultsfromtheBRofb→sγ,ingoodagreementwiththeSMexpec- tations,implystringentconstraintsonNPcontributions. TheexclusiveBRarenotveryinteresting as they suffer from very large theoretical uncertainties. The Q operator, which is the main oper- 7 ator responsible for b→sγ processes, produces almost exclusively left-polarized photons in the SM. This implies that φγ is not a CP eigenstate. One can in any case consider the fraction of wronglypolarizedphotonsina"semi-standard"timedependentCPanalysis. InB ,where∆Γ∼0, d a flavour-tagged time-dependent analysis is required to obtain sensitivity to the photon polariza- tion. In B instead, where ∆Γ/Γ is significantly different from 0, the flavour averaged decay rate s also has sensitivity to the polarization, making an analysis without flavour tagging possible. The energy calibration at LHCb is already promising. Currently it is based on low mass resonances, untilasampleofK∗0γ willbeobtained. Weexpectinoneyearabout4800B0→φγ eventswhich s shouldgivesensitivitiesoftheorderorbetterthanthecurrentworldaverages. AnotherwaytomeasurephotonpolarizationisviaB0 →K∗0e+e− decaysforverylowe+e− d invariant masses. The available statistics will be smaller and large backgrounds are expected. NonethelessthesystematicsareeasieranddifferentfromtheonesofB0→φγ,hencethismodeis s expected to be equally powerful. Finally B0 →K∗0µ+µ− decays for low µ+µ− masses can also d be used. Because of the higher masses, the previous analysis can not be just duplicated and the terms involved are more complex. On the other hand it will be possible to study other interesting observablesfromtheinterference. 3. Conclusions First B decays have been recorded in the LHCb detector, which is well on track for its heavy flavour program. The first validation work with 2010 data is very promising: trigger, particle identification, and tracking efficiencies are reasonably close to expectations and the prospects are alreadyinterestingwith100pb−1 ofdata. References [1] A.A.Alvesetal.[LHCbCollaboration],JINST3(2009)S08005. [2] B.Adevaetal.[LHCbCollaboration],LHCb-PUB-2009-029[arXiv:0912.4179v2]. [3] A.J.Burasetal.,JHEP1010:009(2010)[arXiv:1005.5310]. [4] [CDFCollaboration],CDFPublicNote9892. [5] O.Buchmuelleretal.,Eur.Phys.J.C64:391-415(2009)[arxiv:0907.5568v1]. [6] R.Fleischer,N.Serra,N.Tuning,Phys.Rev.D82,034038(2010)[arXiv:1004.3982]. [7] J.-T.Wei,P.Changetal.[BelleCollaboration],Phys.Rev.Lett.103,171801(2009) [arXiv:0904.0770v2]. [8] J.Wichtetal.[BelleCollaboration],Phys.Rev.Lett.100,121801(2008)[arXiv:0712.2659v2]. 5