25 April 1996
PHYSICS ELSEXIER
LETTERS B
Physics Letters B 373 (1996) 334-338
A measurement of B( Do --+ K-&T~)/B(
DO -+ K-V+)
CLEO Collaboration B. Barish a, M. Chadhaa, S. Chan a, G. Eigen a, J.S. Miller a, C. O’Grady a, M. Schmidtler a, J. Urheima, A.J. Weinstein a, F. Wtirthweina, D.M. Asner b, M. Athanas c, D.W. Bliss c, W.S. Brower ‘, G. Masek ‘, HP Paar ‘, J. Gronberg c, C.M. Korte c, R. Kutschke c, S. MenaryC, R.J. MorrisonC, S. Nakanishi ‘, H.N. Nelson c, T.K. Nelson c, C. Qiao c, J.D. Richman ‘, D. Roberts c, A. Ryd ‘, H. Tajima c, M.S. Witherell c, R. Balest d, K. Cho d, W.T. Fordd, M. Lohnerd, H. Parkd, P. Rankind, J. Royd, J.G. Smithd, J.P. Alexandere, C. Bebeke, B.E. Bergere, K. Berkelmane, K. Bloome, D.G. Cassel e, H.A. Choe, D.M. Coffman e, D.S. Crowcroft e, M. Dickson e, P.S. Drell e, D.J. Dumas e, R. Ehrlich e, R. Elia e, P. Gaidarev e, B. Gittelman e, S.W. Gray e, D.L. Hartill e, B.K. Heltsley e, C.D. Jones e, S.L. Jones e, J. Kandaswamy e, N. Katayamae, PC. Kime, D.L. Kreinicke, T. Lee e, Y. Liu e, G.S. Ludwig e, J. Masui e, J. Mevissen e, N.B. Mistry e, C.R. Nge, E. Nordberg e, J.R. Patterson e, D. Peterson e, D. Riley e, A. Soffer e, C. Ward e, P. Avery f, C. Prescott f, S. Yangf, J. Yeltonf, G. Brandenburg s, R.A. Briere s, T. Liu s, M. Saulnier g, R. Wilsons, H. Yamamotos, T. E. Browderh, F. Li h, J.L. Rodriguez h, T. Bergfeld’, B.I. Eisenstein i, J. Ernst i, G.E. Gladding’, G.D. Gollin’, M. Palmer i, M. Selen i, J.J. Thaler’, K.W. Edwardsj, K.W. McLeanj, M. Oggj, A. Bellerivek, D.I. Britton k, R. Janicek k, D.B. MacFarlane k, PM. Pate1 k, B. Spaan k, A.J. Sadoff!, R. Anunarm, P. Baringer m, A. Beanm, D. Besson m, D. Coppagem, N. Coptym, R. Davis”, N. Hancockm, S. Kotov m, I. Kravchenkom, N. Kwakm, S.Anderson n, Y. Kubota”, M. Lattery n, J.K. Nelson n, S. Patton “, R. Poling n, T. Riehle n, V. Savinov “, M.S. Alarno, I.J. Kim O, Z. Ling O, A.H. Mahmood O, J.J. O’Neill O, H. Severini O, C.R. Sun O, S. Timm O, F. Wappler O, J.E. Duboscq P, R. Fulton r, D. Fujinor, K.K. Gan P, K. Honscheid P, H. Kagan P, R. Kass P, J. Lee P, M. Sung r, A. Undrus Py’, C. Whitep, R. Wanker, A. Wolfr, M.M. Zoeller P, X. Fu 9, B. Nemati q, S.J. Richichi 9, W.R. Ross 9, P. Skubic ‘J, M. Wood q, M. Bishai r, J. Fast r, E. Gerndt r, J.W. Hinson r, T. Miao r, D.H. Miller’, M. Modesitt r, E.I. Shibata’, I.P.J. Shipsey’, P.N. Wang r, M. Yurko r, L. Gibbons ‘, S.D. Johnson ‘, Y. Kwon ‘, S. Roberts ‘, E.H. Thorndike ‘, C.P. Jessop t, K. Lingel t, H. Marsiske t, M.L. Per1 t, S.F. Schaffnert, R. Wang t, T.E. Coan “, J. Dominick”, V. Fadeyev “, I. Korolkov”, M. Lambrecht “, S. Sanghera”, V. Shelkov”, R. Stroynowski”, I. Volobouev”, G. Wei “, M. Artuso”, A. Efimov “, M. Gao “, M. Goldberg”, R. Greene”, D. He “, 0370-2693/96/$12.00 @ 1996 Elsevier Science B.V. All rights reserved I’IISO370-2693(96)00159-l
CLEO Collaboration/ Physics Letters B 373 (1996) 334-338
335
N. Horwitz “, S. Kopp “, G.C. Moneti”, R. Mountain”, Y. Mukhin’, S. Playfer “, T. Skwamicki’, S. Stone”, X. Xing’, J. Bartelt w, S.E. Csorna w, V. Jain w, S. Marka w, A. Freyberger ‘, D. Gibaut ‘, K. Kinoshita ‘, P Pomianowski ‘, S. Schrenk x, D. Cinabro Y
’
a California Institute of Technology, Pasadena. CA 91125. USA b University of California, San Diego, Lo Jolla, CA 92093, USA c University of California, Santa Barbara, CA 93106, USA d University of Colorado, Boulder; CO 80309-0390, USA e Cornell University, Ithaca, NY 14853, USA f Universityof Florida, Gainesville, FL 32611, USA $ Harvard University. Cambridge, MA 02138, USA b University of Hawaii at Manoa, Honolulu, HI 96822, USA i University of Illinois, Champaign-Urbana, IL 61801, USA j Carleton University, Ottawa, Ont. KIS 586. Canada and the Institute of Particle Physics, Canada k McGill University. Montrbal, Que. H3A 2T8. Canada and the Institute of Particle Physics, Canada 1 Ithaca College, Ithaca, NY 14850, USA m University of Kansas, Lawrence, KS 66045, USA ” University of Minnesota, Minneapolis, MN 55455, USA ’ State University of New York at Albany, Albany, NY 12222, USA P Ohio State University, Columbus, OH 43210, USA 9 University of Oklahoma, Norman, OK 73019, USA T Purdue University, West Lafayette. IN 47907, USA ’ University of Rochester; Rochestes NY I4627. USA Stanford Linear Accelerator Center: Stanford University, Stanford, CA 94309, USA ’ Southern Methodist University, Dallas, TX 75275. USA v Syracuse University, Syracuse, NY 13244, USA w Vanderbilt University, Nashville, TN 37235, USA ’ Virginia Polytechnic Institute and State University, Blacksburg, VA 24061. USA Y Wayne State University, Detroit, MI 48202, USA Received 5 February 1996 Editor: L. Montanet
Abstract Using a sample of 3.1 fb-’ integrated luminosity accumulated with the CLEO II detector at the Cornell Electron Storage Ring, we measure the ratio of branching fractions B(D” + K-7r+?r”)/B(Do + K-r+) = 3.81 f 0.07 f 0.26, the most precise determination of this quantity to date. PACS: 13.2O.F~; 13.25.-k; 13.25.Ft; 14MLb
Precision studies of the hadronic decays of charmed mesons are important for a variety of reasons. In the case of simple, yet relatively common, decay modes such as Do -+ K-CT+ and Do -+ K-d&’ [ 11, precise knowledge of the total branching fractions is important since many B physics analyses depend on these to extract meaningful results. Any uncertainty in the ’ Permanent address: BINP, RU-630090 Novosibirsk,Russia.
charm decays will translate directly into larger systematic errors for the heavier quark analyses. These modes are also used as normalization for many charmed meson branching ratio measurements, in particular Do + K-&d’ is a useful reference for other channels involving a rr” in the final state. Past determinations of the ratio of branching fractions R = f3( Do -+ K-&d’) /B( Do -+ K-h) are shown in Table 1 [ 2-71. There are two hints that sug-
336
CLEO Collaboration/Physics
Letters B 373 (I 996) 334-338
Table I measunxnents of B(Dn -+ K-n+?ro)/t3(D” -.+ Mark II and Mark III results were reported as absolute branching fractions. The systematic error shown for Mark III is the sum in quadrature of the systematic errors of the two branching fractions Previous
K-T+).
ARGUS NA14 CLEO 1.5 Mark Ill
121 131 [4J 151
E516 Mark II
161 [7]
PDG Ave. PDG Fit
1992 1991 1991
3.04 f 0.16 f 0.34 4.0 f 0.9 f I .o 2.8 f 0.14 zt 0.52
1988 1984 1981
3.17f0.42f0.43 4.2 f 1.4 2.85 f 1.13 3.07 f 0.29 3.51 f 0.28
4000 N0 f5
3000
% z E 2000 5 L 1000
0
gest the current world average value of this ratio may be too low. First, the PDG average and fitted values
are not in good agreement [ 81. Second, in studies [ 91 of the decay of B mesons to charmed final states, the measured B branching ratios are consistently higher when the Do is reconstructed in the K-T’@ decay than when the K-T+ mode is used. In this analysis we use an integrated luminosity of 3.1 fb-’ of e+e- collisions accumulated with the CLEO II detector at the Cornell Electron Storage Ring (CESR) , running at center of mass energies at or just below the Y (4s) resonance. Details of the CLEO II detector are described elsewhere [ IO]. To reconstruct Do -+ K-T+ decays we consider all pairs of well-fitted oppositely-charged tracks. When reconstructing Do -+ K-$# decays we include two electromagnetic showers each with energy above 100 MeV to form the ?ro candidate. To reduce the rate of fake ?ro’s from random shower combinations and to increase the resolution, we require that both showers be in the central region of our detector [ 11I, and neither shower be near any charged tracks entering the calorimeter. The invariant mass of the two photons is required to be between 115 MeVlc and 155 MeV/c ( z 3.5~). To improve the measurement of the ?r” 4vector, the two photons are kinematically fitted to the known Z-Omass. The Do mesons are required to be produced via the decay Def -+ D’?r+. Reconstructing this decay sequence provides additional kinematic information which allows us to reduce the combinatoric background significantly. In addition, for the dominant Cabibbo favored Do -+ K-T+( $) mode, the “slow”
0 Mass
10 Difference
20
(MeV I c*)
Fig. 1. The mass difference Ah4 = M(D*+)M(D”) -145 MeV/c* for the Do -+ K-&T’ mode. The shaded area represents the signal region and the filled area is the sideband region.
pion from the D*+ decay is known to have the same charge as the pion from the decaying Do, allowing us to assign masses to the Do daughter tracks without the use of extra particle identification information. We reconstruct the D*+ by combining “slow” pion candidates, having momentum between between 225 MeVlc and 425 MeVlc, with all Do candidates. We require the measured mass difference, AM = M(D*+) - M(D’) - 145MeV/c*, to be within 2.5 u of the accepted value (- 1.92 MeV/c* and 1.92 MeV/c*) as shown in Fig. 1. Data in a mass difference sideband (6.0 to 9.5 MeV/c*) are used to estimate the rate of fake D*+ + Do& decays. We compute the scaled momentum xol of the D*, defined as the measured momentum of the D* divided by the maximum possible D’ momentum, pz, = EL - M2,., and require that xg* > 0.6. Since most of the combinatoric background is at low values of no*, and the D” mesons are produced with a hard fragmentation spectrum, using this cut dramatically reduces the background as well as restricting our analysis to continuum e+e- + ci? reactions. Combinations of particles which pass the above cuts are shown in Figs. 2 and 3. In both figures we see a large number of events in the Do signal region. We fit the Do mass peak with the sum of two bifurcated Gaussians [ 123. The background is fit between 1.7 and 2.0 GeV with a straight line. After making the side-
CLEO Collaboration/Physics
331
Letters B 373 (1996) 334-338
band subtraction the fitted yields were 15,013 f 204 events for Da --) Km&n-a and 9,808 f 127 events for Do ---) K-IT+. Large samples of GEANT [ 131 based Monte Carlo D*+
1.70
1.90
1.80
2.00
(GeVI c*)
MK-7T+?T0 Fig. 2. The K-?r+?ra invariant mass spectrum for selected events. The fitted points ate the mass difference signal region and the solid histogram is from the mass difference sideband. The failure of the sideband data to saturate the background is due to fake combinations of slow pions from real D*+ decays with fake D” -+ K-p+?rO candidates.
i
3000Piu 3 f p
zooo-
L? 5 t lOOO-
1.70
-1.80
MK_r+ (GaV I
1.90
2.00
c*)
Fig. 3. The K-rr+ invariant mass spectrum for selected events. The fitted points am the mass difference signal region and the solid histogram is from the mass difference sideband. The failure of the sideband data to saturate the background is due to fake combinations of slow pions from real D*+ decays with fake D” - K-n+ candidates.
--P DOT+, Do +
K-T+
(Do -+ K-r+?ro)
events were analyzed to determine the reconstruction efficiency of these modes to be 20.4% and 8.2% respectively. The Do --f K-&d Monte Carlo included the resonant substructure of the three body decay using published amplitudes [ 141. Using the yields found from data and the above efficiencies, we find the ratio of branching fractions to be R = 3.81 f 0.07 where the error shown is only statistical. To obtain an estimate of the systematic errors associated with the above ratio, several potential sources were explored. Using Monte Carlo, we looked at the variation in the efficiency for Do -+ K-‘rr+# as a function of the details of the resonant substructure. Using other measurements of the resonant substructure [ 15,161, we assign a 3.4% systematic error. To investigate the fitting procedure, we start by varying the sideband in AM and find a change of 0.3% in R. We also fit the signal with a larger background window and a second order polynomial and find a similar change. Changing the signal fitting function to a single Gaussian, a bifurcated Gaussian, a double Gaussian, a bifurcated double Gaussian or a bifurcated double Gaussian with the shape constrained by the detector simulation Monte Carlo changes R by less than 1.5%. Extracting the signal yields from the AM distribution instead from the Do mass plot also produces less than 1.5% variation in the final ratio. We have also varied the cuts used in the analysis. When the mass difference requirements are tightened to be within 1.25 MeV Jc* of the nominal mass difference value (about 1.5~)) we observe a 0.8% change in the ratio. From a study of the decays 77 -+ yy and v + n-aTo?ro?ro, we assign a 5.5% systematic error for uncertainty in the overall 9 finding efficiency. Table 2 summarizes the investigated sources of systematic error and the contribution of each to the ratio of branching fractions. We add these in quadrature to arrive at an estimate of the total systematic error of 6.9%. Summarizing the above results we obtain a measurement of the ratio of branching fractions
338
CLEO Collaboration/Physics
Letters B 373 (1996) 333-338
Table 2 Sources and estimates of systematic uncertainty as described in the text
111Charge conjugation is implied throughout.
[21 ARGUS Collaboration, H. Albrecht et al., Z. Phys. C 56
Sources of systematic error
(1992) 7. Monte Carlo Statistics ?ro finding efficiency K-afrro resonant substructure Fitting M(D*+) - M(DO) cut
1.7 % 5.5 8
Total
6.9 8
3.4 %
1.5 % 0.8 %
131 NA14 Collaboration, MP
Alvarez et al., Z, Phys. C 50 (1991) II. [41 CLEO Collaboration, K. Kinoshita et al., Phys. Rev. D 43 (1991) 2836. I51 Mark III Collaboration, J. Adler et al., Phys. Rev. Lett. 60 (1988) 89. I61 ES16 Collaboration, D.J. Summers et al., Phys. Rev. Lett. 52 (1984) 410.
R=
f?(D’ -+ K-TT+T') B(DO+K-T+)
= 3.81f 0.07f 0.26.
This measurement, with about 15,000 ( 10,000) events in the K-T+&' (K-b) peak,is the most precise determination of the ratio to date. We combine this result with CLEO’s recent measurement of the absoiute branching fraction B( Do -+ K-d) = 0.0391f 0.0008 f 0.0017 [ 171 to extract the branching ratio f3( Do ---f K-T+T') = 0.149f 0.004 f 0.012, which is comparable in both magnitude and uncertainty to the current Particle Data Group [ 8 1 best fit value for this mode of B( Do + K-b@) = 0.135 f 0.011.
We gratefully acknowledge the effort of the CESR staff in providing us with excellent luminosity and running conditions. This work was supported by the National Science Foundation, the U.S. Department of Energy, the Heisenberg Foundation, the Alexander von Humboldt Stiftung, the Natural Sciences and Engineering Research Council of Canada, and the AI? Sloan Foundation.
171 Mark 11Collaboration, R.H. Schindler et al., Phys. Rev. D 24 (1981) 78. [81 L. Montanet et al., Phys. Rev. D 50 (1994) 1173; and 1995 off-year update for the 1996 edition (URL: http://pdg.lbl.gov/). [91 CLEO Collaboration, S. Alam et al., Phys. Rev. D 50 ( 1994) 43. Seven of the eight exclusive charmed hadronic decay modes studied are larger when the final state Do is reconstructed as K-rr+?rO as compared to K-n+. [lOI CLEO Collaboration, Y. Kubota et al., Nucl. Instrum. Methods A 320 (1992) 66. 1111 1cosO\ < 0.71 where 0 is the angle between the beam direction and the direction of the shower from the interaction point. iI21 A bifurcated Gaussian has two (T’S,one for the width above the mean and the other for the width below the mean. In the case of the double bifurcated Gaussian we constrain the mean and ratio of the left and right u’s to be identical for both Gaussians. [I31 R. Brun et al., CERN Report No. CERN-DD/EE/ICI, 1987 (unpublished). I141 E691 Collaboration, J.C. Anjos et al., Phys. Rev. D 48 (1993) 56. (151 E687 Collaboration, P Frabetti et al., Phys. Lett. B 331 (1994) 217. 1161 Mark III Collaboration, J. Adler et al., Phys. L&t. B 196 (1987) 107. L17l CLEO Collaboration, D. Akerib et al., Phys. Rev. Lett. 71 (1993) 3070.