Measurement of the Ω lifetime - Infn

Physics Letters B 561 (2003) 41–48www.elsevier.com/locate/npeMeasurement of the Ω 0 c lifetimeFOCUS CollaborationJ.M. Link, M. Reyes, P.M. YagerUniversity of California, Davis, CA 95616, USAJ.C. Anjos, I. Bediaga, C. Göbel, J. Magnin, A. Massafferri, J.M. de Miranda,I.M. Pepe, A.C. dos ReisCentro Brasileiro de Pesquisas Físicas, Rio de Janeiro, RJ, BrazilS. Carrillo, E. Casimiro, E. Cuautle, A. Sánchez-Hernández, C. Uribe, F. VázquezCINVESTAV, 07000 México City, DF, MexicoL. Agostino, L. Cinquini, J.P. Cumalat, B. O’Reilly, J.E. Ramirez, I. Segoni, M. WahlUniversity of Colorado, Boulder, CO 80309, USAJ.N. Butler, H.W.K. Cheung, G. Chiodini, I. Gaines, P.H. Garbincius, L.A. Garren,E. Gottschalk, P.H. Kasper, A.E. Kreymer, R. KutschkeFermi National Accelerator Laboratory, Batavia, IL 60510, USAL. Benussi, S. Bianco, F.L. Fabbri, A. ZalloLaboratori Nazionali di Frascati dell’INFN, I-00044 Frascati, ItalyC.Cawlfield,D.Y.Kim,A.Rahimi,J.WissUniversity of Illinois, Urbana-Champaign, IL 61801, USAR. Gardner, A. KryemadhiIndiana University, Bloomington, IN 47405, USAC.H. Chang, Y.S. Chung, J.S. Kang, B.R. Ko, J.W. Kwak, K.B. LeeKorea University, Seoul 136-701, South Korea0370-2693/03/$ – see front matter © 2003 Published by Elsevier Science B.V.doi:10.1016/S0370-2693(03)00388-5

42 FOCUS Collaboration / Physics Letters B 561 (2003) 41–48K. Cho, H. ParkKyungpook National University, Taegu 702-701, South KoreaG. Alimonti, S. Barberis, M. Boschini, A. Cerutti, P. D’Angelo, M. DiCorato,P. Dini, L. Edera, S. Erba, M. Giammarchi, P. Inzani, F. Leveraro,S. Malvezzi, D. Menasce, M. Mezzadri, L. Moroni, D. Pedrini,C.Pontoglio,F.Prelz,M.Rovere,S.SalaINFN and University of Milano, Milano, ItalyT.F. Davenport IIIUniversity of North Carolina, Asheville, NC 28804, USAV. Arena, G. Boca, G. Bonomi, G. Gianini, G. Liguori, M.M. Merlo, D. Pantea,D. Lopes Pegna, S.P. Ratti, C. Riccardi, P. VituloDipartimento di Fisica Nucleare e Teorica and INFN, Pavia, ItalyH. Hernandez, A.M. Lopez, E. Luiggi, H. Mendez, A. Paris, J. Quinones,W. Xiong, Y. ZhangUniversity of Puerto Rico, Mayaguez, PR 00681, USAJ.R. WilsonUniversity of South Carolina, Columbia, SC 29208, USAT. Handler, R. MitchellUniversity of Tennessee, Knoxville, TN 37996, USAD. Engh, M. Hosack, W.E. Johns, M. Nehring, P.D. Sheldon, K. Stenson,E.W. Vaandering, M. WebsterVanderbilt University, Nashville, TN 37235, USAM. SheaffUniversity of Wisconsin, Madison, WI 53706, USAReceived 25 February 2003; accepted 10 March 2003Editor: L. Montanet

44 FOCUS Collaboration / Physics Letters B 561 (2003) 41–48Fig. 1. Invariant mass distributions for Ω 0 c candidates: (a) reconstructed mass of Ξ − K − π + π + .Thereare38± 9 events at a mass of2696.5 ± 1.9 MeV/c 2 . (b) Reconstructed mass of Ω − π + .Thereare23± 7 events at a mass of 2699.4 ± 3.4 MeV/c 2 . (c) Combined invariantmass distribution. There are 64 ± 14 events at a mass of 2697.5 ± 2.2 MeV/c 2 . We define the signal region (hatched area) to be within 2σ ofthe fitted mass value and the two sideband regions (dotted area) are 4–12σ from the fitted mass value.three charged tracks of the right charge combinationfor the Ξ − K − π + π + mode, and an Ω − and an oppositelycharged track for the Ω − π + mode. The confidencelevel of the decay vertex of the Ωc 0 candidate isrequired to be greater than 10%. The combined momentumvector located at the decay vertex forms theΩc 0 track. A candidate driven vertexing algorithm [12]uses the Ωc 0 track as a seed track to find a productionvertex with a confidence level greater than 1%. Theprimary multiplicity, including the seed track, mustbe at least 3 tracks and the production vertex mustbe inside a target. The significance of separation betweenthe production and the decay vertices (L/σ L )must be greater than 2 for the Ξ − K − π + π + modeand greater than 0 for the Ω − π + mode. Different valuesare chosen for the L/σ L cut due to a differencein the secondary vertex resolution. Čerenkov particleidentification (PID) is performed by constructing a loglikelihood value W i for the particle hypotheses (i =e,π,K,p).Theπ consistency of a track is defined byW π = W min − W π ,whereW min is the minimum Wvalue of the other three hypotheses. Similarly, we defineW K,π = W π − W K for kaon identification. Werequire W K,π > 3 for kaons and W π > −6 forpions. In the Ξ − K − π + π + mode, we add additionalcombination PID cuts based on Monte Carlo simulationstudies. We define ∑ W π to be the positivesum over pion candidates and the negative sumover other particle candidates. Also W K,π +W K,pis used for separation between the protons and thekaons. ∑ W π and W K,π + W K,p are requiredto be greater than 7. In the Ω − π + mode the momentumasymmetry, (P Ω − − P π +)/(P Ω − + P π +), isrequired to be greater than −0.2 and less than 0.7.Also the π + transverse momentum must be larger than0.2 GeV/c and the momentum of Ωc 0 must be greaterthan 50 GeV/c. The resulting mass spectra are shownin Fig. 1. The mass spectra are fit with a Gaussianfunction for the signal distribution and a first orderpolynomial function for the background. From thecombined sample, we find a fitted mass of 2697.5 ±2.2 MeV/c 2 (systematic uncertainty not evaluated),consistent with the results of other experiments.4. Lifetime measurementTo measure a lifetime in fixed target experiments,we use a binned maximum likelihood technique [13].We fit the reduced proper time distribution, definedas t ′ = (L − Nσ L )/βγ c = t − Nσ t ,whereN is theseparation cut value between the production and thedecay vertex, βc is the particle velocity, and γ is theLorentz boost factor to the Ωc 0 center of mass frame.The signal region is defined to lie within 2σ of thefitted Ωc 0 mass. The background is assumed to have

FOCUS Collaboration / Physics Letters B 561 (2003) 41–48 45Fig. 2. The f(t ′ ) correction function is displayed for each mode.Fig. 3. (a) The corrected t ′ distribution with the lifetime fit function for the combined signal. (b) The t ′ distributions of expected backgroundsin the signal band for each mode; the dark region is for Ξ − K − π + π + and the light one is for Ω − π + . Lines show the lifetime fitting functionsfor signal and background distributions. The lifetime fit finds 59 ± 12 signal events rather than 64 ± 14 due to the 2σ mass window used.the same lifetime behavior in the signal region as inthe sidebands, 4–12σ away from the peak. Taking S asthe number of signal events in the signal region and Bas the total number of background events in the sameregion, the expected number of events n i in the ithreduced proper time bin centered at ti ′ is given by:n i = S f(t′ i )e−t i ′/τ∑(1)i f(t′ i )e−t i ′/τ + B b i∑i b ,iwhere b i describes the background reduced propertime as estimated from sidebands and f(ti ′ ) is a correctionfunction which takes into account the effectsof spectrometer acceptance and efficiency, analysis cutefficiencies, and particle absorption. The f(t ′ ) distributionare shown in Fig. 2 for each decay mode.The likelihood is constructed from the productof the Poisson probability of observing s i eventswhen n i are expected with the Poisson probability ofobserving N b = ∑ b i events in the sidebands when

46 FOCUS Collaboration / Physics Letters B 561 (2003) 41–48Fig. 4. The predicted events (histogram) are superimposed on the observed events (points) while the shaded distribution displays the t ′distribution of the background for (a) the Ξ − K − π + π + mode and (b) the Ω − π + mode.4B background events are expected. The factor of 4accounts for the fact that the sideband region is fourtimes wider than the signal region. The likelihoodtakes the form:( ∏ n s i) (iL =e−n i (4B)N be −4B )×.s i !N b !iThe combined likelihood function is given by theproduct of the likelihoods:L Ω 0 c= L Ξ − K − π + π + × L Ω − π +.(2)There are 3 fit parameters; one parameter for thelifetime τ and two parameters, B Ξ − K − π + π + andB Ω − π +, for the backgrounds from each mode. Ourmeasurement of the Ωc 0 lifetime is 72 ±11 fs as shownin Fig. 3. In Fig. 4, the t ′ distributions from the dataand from the fit are compared with each other.5. Studies of systematic errorsWe have studied various systematic uncertaintiesassociated with the Monte Carlo modeling by computingthe lifetimes of independent data samples splitby particle/antiparticle, primary vertex position (upstreamand downstream target region), Ω 0 c momentum(greater than 70 GeV/c and less than 70 GeV/c) andproduction vertex multiplicity (> 5and 5). All lifetimesfrom these samples are consistent within the statisticalerror as shown in Fig. 5 points 1–8, indicatinga negligible systematic error due to the Monte Carlosimulation.Since the binned likelihood method has been usedto measure the lifetime, we have investigated theuncertainty from the fit range and binning effects byexamining the variance in lifetime for different t ′ binsizes (Fig. 5, points 10–11) and for different fittingranges (Fig. 5, point 9).The proper time resolution of our fully simulatedMonte Carlo for the Ωc0 data is about 40–50 fs.We have tested the accuracy of the fitting procedurewhen the lifetime is comparable to the proper timeresolution. We used a toy Monte Carlo study to testthe fitting procedure using the proper time from whichwe extracted a systematic uncertainty of 4 fs. Thetoy Monte Carlo test was also used to validate thestatistical error determination.The systematic uncertainty due to the backgroundcontamination is examined by investigating the reflectionsfrom other charm baryon decays and by varyingthe sideband and signal regions (Fig. 5, points 12–17).The studies of the systematic uncertainties are summarizedin Table 1. The total systematic uncertainty ofthe Ωc 0 lifetime measurement is determined to be 11 fsby adding all of the systematic uncertainties in quadrature.

FOCUS Collaboration / Physics Letters B 561 (2003) 41–48 47Fig. 5. Lifetime measurements for systematic studies. The solid line represents the best determined value for the Ωc 0 lifetime and the two dottedlines show the extent of the statistical error.Table 1The itemized list of the systematic uncertainties. The numbers in the items column refer to the entry numbers in Fig. 5Systematic source Items Uncertainty (fs)Split sample method Particle and antiparticle (1,2) ∼ 0Upstream and downstream target (3,4) ∼ 0High and low momentum (5,6) ∼ 0Primary vertex multiplicity (7,8) ∼ 0Fit variant Bin size (10,11) and fitting region (9) ±9t ′ resolution Toy Monte Carlo studies ±4Background Sideband (12,13,14) and signal band (15,16,17) ±5Total ±11Table 2The Ωc 0 lifetime measurementsExperiment Lifetime Decay modesE687 86 +27−20 ± 28 fs + K − K − π +WA89 55 +13−11 +18−23 fs Ξ − K − π + π + , Ω − π + π + π −PDG2002 64 ± 20 fsFOCUS 72 ± 11 ± 11 fs Ξ − K − π + π + , Ω − π +6. ConclusionWe measure an Ωc 0 lifetime of 72±11±11 fs using64 ± 14 events in the two decay modes, Ω − π + andΞ − K − π + π + . We compare our result with previousmeasurements in Table 2. Our lifetime result is consistentwith previous Ωc 0 lifetime results. This 20% measurementof the lifetime substantially improves uponthe current (30%) world average.AcknowledgementsWe wish to acknowledge the assistance of the staffsof Fermi National Accelerator Laboratory, the INFNof Italy, and the physics departments of the collaboratinginstitutions. This research was supported in part bythe US National Science Foundation, the US Depart-

48 FOCUS Collaboration / Physics Letters B 561 (2003) 41–48ment of Energy, the Italian Istituto Nazionale di FisicaNucleare and Ministero della Istruzione, Universitàe Ricerca, the Brazilian Conselho Nacional de DesenvolvimentoCientífico e Tecnológico, CONACyT-México, and the Korea Research Foundation of theKorean Ministry of Education.References[1] S.F. Biagi, et al., Z. Phys. C 28 (1985) 175.[2] ARGUS Collaboration, J. Stiewe, in: J.R. Sanford (Ed.),Proceedings of 26th International Conference in High EnergyPhysics, Dallas 1992, Vol. 1, American Institute of Physics,New York, 1993, p. 1076.[3] P.L. Frabetti, et al., Phys. Lett. B 338 (1994) 106.[4] P.L. Frabetti, et al., Phys. Lett. B 357 (1995) 678.[5] P.L. Frabetti, et al., Phys. Lett. B 300 (1993) 190.[6] M.I. Adamovich, et al., Phys. Lett. B 358 (1995) 151.[7] D. Cronin-Hennessy, et al., Phys. Rev. Lett. 86 (2001) 3730.[8] H.-Y. Cheng, Phys. Lett. B 289 (1992) 455.[9] Particle Data Group, K. Hagiwara, et al., Phys. Rev. D 66(2002) 010001.[10] J.M. Link, et al., Nucl. Instrum. Methods A 484 (2002) 270.[11] J.M. Link, et al., Nucl. Instrum. Methods A 484 (2002) 174.[12] E687 Collaboration, P.L. Frabetti, et al., Nucl. Instum. MethodsA 320 (1992) 519.[13] P.L. Frabetti, et al., Phys. Lett. B 268 (1991) 584.