D Meson Decays from the Mark IIIt

SLA C-PUB-5052 July 1989 (T/ E) Recent Results on Hadronic D, and D Meson Decays from the Mark IIIt J. .Adler, Z. Bai, G.T. Blaylock, T. Bolton, J.-C. Brient, T.E. Browder, J.S. Brown, K.O. Bunuell, M. Burchell, T.H. Burnett, R.E. C&sell, D. Coffman, V. Cook, D.H. Coward, - F. DeJongh, D.E. Dorfan, J. Drinkard, G.P. Dubois, G. Eigen, K.F. Einsweiler, B.I. Eisenstein, T. Reese, C. Gatto, G. Gladding, C. Grab, J. Hauser, C.A. Heusch, D.G. Hitlin, J.M. Izen, P.C. Kim, L. Kopke, J. Labs, A. Li, W.S. Lockman, U. Mallik, C.G. Matthews, A.I. Mincer, R. Mir, P.M. Mockett, B. Nemati, A. Odian, L. Parrish, R. Partridge, D. Pitman, S.A. Plaetzer, J.D. Richman, M. Roco, H.F.W. Sadrozinski, M. Scarlatella, T.L. Schalk, R.H. Schindler, A. Seiden, C. Simopoulos, A.L. Spadafora, I.E. Stockdale, W. Stockhausen, W. Toki, B. Tripsas, F. Villa, M.Z. Wang, S. Wasserbaech, A. Wattenberg, A.J. Weinstein, S. Weseler, H.J. Willutzki, D. Wisinski, W.J. Wisniewski, R. Xu, Y. Zhu The Mark III Collaboration California Institute of Technology, Pasadena, CA 91125 University of California at Santa Crq Santa Crur, CA 95064 University of Illinois ht Urbana-Champaign, Urbana, IL 61801 University of Iowa, Iowa City, LA 52242 Stanford Linear Accelerator Center, Stanford, CA 94309 University of Washington, Seattle, WA 98195 . Abstract Recent results on hadronic D, and D decays from the Mark III collaboration are presented. The absolute branching ratio B(O$ + &r+) is studied by searching for fully reconstructed e+e- + oJ*oT events using seven hadronic decay modes of the 0:. A limit of B(ot 4 &r+) < 4.1% at 90% C.L. is obtained. Evidence is presented for the decay D$ + f0(975)z+ which agrees with a recent experimental observation. Upper limits are set for the relative branching ratios B(D,+ + qir+)/B(D,+ --) +r+) < 2.5 and B(Df + q’r+)/B(D$ -+ qkr+) < 1.9, where the 7 is studied in both the y-y and the x+?r-+’ decay modes and the ‘7’ in the &‘IT-, 5) + yy decay chain. The resonant substructure of D” -+ K-n+*-lr+ and D+ -+ K” *+?T-?T+ is studied. The branching ratio of D” -+ K*‘p’ is found to be smaller than the theoretically expected whereas the decay modes Do + af K- and D+ -+ .:I?’ are found to be large and account for 50% of these final states. Contributed to the XIV International Symposium on Lepton and Photon Interactions, Stanford, CA., August 7-12, 1989 t Th’is work was supported in part by the U.S. Department of Energy and the U.S. National Science Foundation under contracts DEAC03-76SF00515, DE-AC02-76ER01195, DEAC02-87ER40318, DEAC03-81ER40050, and DE-AM03-76SF0034. We report a study of hadronic D$ and D decays by the Mark III experiment. The o$ absolute branching ratios are investigated through a search for fully exclusive events. - The fan, VT, and q’n decay modes of the o$ are studied in inclusive analyses. Finally, the resonant substructure of two D + Kmm final states are measured. The data used in the Do and D+ analyses was recorded near the peak of the $(3770) during 1982, 1983 and 1984. The total integrated luminosity was J Ldt = 9.3 pb-I. This corresponds to roughly 50K produced DD events. The D$ data was taken at fi = 4.14 GeV in 1986 and corresponds to a total integrated luminosity of s Ldt = 6.3 pb-‘. The Mark III detector’ is a large acceptance magnetic solenoidal spectrometer optimized for the SPEA R energy range. In these analyses, data from the drift chamber, time-offlight system (TOF), and the electromagnetic calorimeter are used. The typical momentum resolution is Up/ p = (.015 x p)2 + (.015)2. For the $(3770) data set, the TOF resolution averages 175 ps, while in the 4 = 4.14GeV data, the resolution is degraded to 200 ps. The shower counter has a resolution of 0 (E) / E = IS%/ & and the efficiency is 100% for ._ E, > 100 MeV. ABSOLUTE HADRONIC BRANCHING FRACTIONS OF THE Dz MESON An upper limit on the absolute branching fraction B(Dz + &r+) is determined by searching for fully reconstructed e+e- + Dg*Dz events in the data sample collected at . & = 4.14 GeV. The following D$ modes are used in the analysis: &r+, Ji’ Ii’ +, fo(975)r+, 1(*(892)‘K+, I(*‘ K*+, qh+~+r-, and @r+rO. The twenty-eight possible final states are considered. For a particular final state, all combinations of photons and particle identification assignments are formed. Backgrounds are reduced by requirements on the resonant substructure of the D$ decay modes. Kinematic fits are employed to select combinations which are consistent with the e+e- + DH*Dz --+ yD,+ 0; hypothesis. The D$ and 0, candidates are constrained to have equal but unspecified mass M(X). Double tag events would produce a signal at the D$ mass in, the resulting M(X) distribution. For each double tag mode, a signal region in the M(X) distribution is selected which contains 95% of a Monte Carlo-generated signal. No candidate signal events are observed. 2 An upper limit on the D$ + &r+ branching fraction B4,+ is obtained by computing the likelihood of observing zero candidate events as a function of B4,t. The measured - - quantities used as input are: aLB4,t = 156f38: bKohrt = 0.92f0.38: bfO,t = 0.28 fO.lO$ bKeOKt = 0.93 f 0.125-’ bRloK*+ = 2.3 f 1.47 b4,+,+,- = 0.41 f 0.10~~7~g and b+t,p = 2.4fl.l:’ where 0 = a(e+e- + DQ* D$), L is the integrated luminosity, and b; = B(D$ -+ Mode i)/B(D$ + qh’). The likelihood function L(B++, gLBti,t, b;) is constructed by using Poisson statistics for the number of observed events, and Gaussian errors (including .. correlations) for the measured quantities. The likelihood is set to zero if B$,t C b; > 100%. The marginal likelihood LC(Bd,t ) is computed by integrating fZ(Bd,t , aLB4,t , bi) with respect to aLB4,t and the b;. The upper limit Bg0 of a 90% likelihood interval is the value of B below which 90% of the integral of ,C(B) is found. The result is Bgo = 3.8%. The uncertainty in the average detection efficiency is 8%. The upper limit is therefore increased to B(D; + c#m+) < 4.1% at 90% CL. The combined mass distribution for aii final states is shown in figure 1 and the marginal likelihood is shown in figure 2. In summary, we have obtained the first model-independent limit on the absolute branching fraction for D$ + &r+. Theoretical predictions for Bd,t are approximately 3%f3s24s25 consistent with the upper limit presented, and with previous estimates. The experimental results-imply that a large fraction of D$ decays have not yet been observed. EVIDENCE FOR 0,s + fo(975)r+ The D$ + f0(975)n+ analysis proceeds as follows. Any charged track which is not a well identified kaon by TOF is assumed to be a pion. All three pion combinations which have recoil masses close to that of the DE mass (2.075-2.125 GeV/ c2) are retained. A 1-C kinematic fit of the tracks to the hypothesis e+e- + s+7r-nfD,+F is performed, where the three-momentum of the Dz is not measured. Events with a fit probability greater than 2% are retained. There is an excess of events with ~+7r- invariant mass near that of the f,(975) and r+r--lr* mass near that of the D$. The rr+r- mass distribution is shown in figure 3 where the r+rr-rr* mass is required to lie between 1.94 and 1.98 GeV/ c2. The r+rr-r* 3 ;: mass is shown in figure 4 where the n+a- mass is required to lie between 0.94 and 0.98 GeV/ c2. The background shape is obtained by requiring the r+n- mass to lie within the - - sideband region (0.86 to 0.90 GeV/ c2). If a11 events with 7r+7r- combinations with masses between 0.94 and 0.98 are assumed to be from the f,(975) decay, we obtain the preliminary measurement: a(e+e- + D$DiT)B(Ds -+ for+) = (14.9 f 4.2 f 6.5) pb Using our measured rate for o(e+e- + D$D~~)B(Ds + &r+), we obtain the preliminary result: B (0s’ + for’ ) = .58 f .21 f .28 B (0s’ + #T+) For the above measurement we assumed that the entire f0 is contained in the mass interval, .94 - .98 GeV/ c2. This result will increase by a factor of two if we use the 50 MeV/ c2 width of the f,(975) as measured in the reaction J/T/I + $fO by the Mark III. Our measurement for the ratio is in agreement with the value of .28 % .lO f .03 obtained by E691.26 ‘* SEA RCH FOR D$ t VT+ The decay D$ + q7rr+ was investigated in two final states, 77 + X+X-~’ and q + yy, A combined upper limit is obtained. + 0 q + 7 r 7r % - In the analysis of the 7 + 7rT+7rlT-7ro decay mode a 1-C kinematic fit of all yy candidate pairs in the event is made to the r” mass . Pairs for which P (x2) > 0.05 are then used in a 2-C fit to the hypothesis e+e- + Dz*r+7r-7r07r*, rr” + yy, where all combinations of the three pion candidate tracks are tried. The constraints are the 7r” mass and the mass of the unobserved DQ+*. After imposing P (x2) > 0.05 for the 2-C fit, @it > 70 MeV/ c2 for the photons from the r” and the requirement that the three pion mass be within 534 < M (T’T+?~-) < 564 MeV/ c2, we obtain the plot shown in figure 5. The curve includes a background whose shape is determined from the r+r-7r07r* mass distribution obtained when an r+7r-rITo sideband is selected outside the 17 region. From a maximum likelihood fit 4 we find an excess of 16.6 f 6.1 events in the signal region. Correcting for the reconstruction - - efficiency of 12.7% and for the 7 branching ratio, this excess corresponds to c - B (0; + 1/r+) = 44 f 16 f 12 pb. In the second final state that is discussed, we are unable to confirm this excess. In the 7 -+ yy analysis all pairs of candidate photons in the event are fit to the hypothesis rl + YY- If the yy pair satisfies a 1-C fit with P (x2) > 0.20, it is used in a 2-C fit to e+e- --+ D,**+q, 7 + yy , where the 77 mass is fixed and the missing D,‘* mass is constrained but not measured. In order to reduce combinatorial background, further cuts are used; P (x2) > 0.10 for the 2-C fit, Eyh’ > 0.5 GeV/ c2 and E$’ > 0.2 GeV/ c2. The resulting q7r+ mass distributionis shown in figure 6. There is no evidence for a D$ signal. The resulting limit, adjusting for the reconstruct-ion efficiency of 23.6%, for the 17 --Y yy branching ratio, and allowing for systematic error is: CT - B (Ds+ -+ VT+) < 42.5 pb. (90% C.L.) _ Joint Upper Limit To properly combine these results we calculate a joint likelihood as a function of the number of produced events and we conservatively set a 90% C.L. limit of N,, < 825 produced events. When this is combined with the integrated luminosity we obtain: CT. B (DYj- ---f yr+) < 66 pb (90% C.L.) Using our measured D$ + &r+ cross section, we obtain the preliminary result: B (E 7 VT+) B(D$ + &r+) < 2.5 (90% C.L.) : SEARCH FOR O,+ ---f q’r+ The r]‘r+ analysis is performed using the decay chain, D$ + q?r+, q’ --+ 7pr+7r-, 77 --+ yy. Photon candidates are selected on the basis of a 1-C fit to to the 77 mass. We then perform a 2-C fit to the hypothesis e+e- + D,+Fqn+~-?r* , 77 --+ yy, where the masses of the 17 and the missing DQ+F are fixed. In addition, we require; P (x2) > 0.10 for . the 2-C -- fit, _ @it > 0.15 GeV/ c2, and lrn(q7r+r-) - mq, 1 < .015 GeV/ c2. No significant signal is observed as shown in figure 7. An upper limit is obtained. Using the measured 0 x B(D, + &r), we obtain a preliminary ratio of : B (0s’ + q/r+) B(D,+ -+ qhr+) < 1.9 (90% C.L.) B( Dt -+9x+) The 77~ results are consistent with the measurement B(Dt+dxt) N 3 from the Mark 1127, an early preliminary result -g g = 2.6 .-f 0.6 f 0.8 from the Mark 11128, and.q the B(D$&) < 1.5 (90% C . L . ) set by the E691.r’ The 77’~ limit is, however, much limit B(Df+qh+) lower than the ratio g g g N 4.8 reported by the Mark II.27 as well as the ratio B(Df-++) _ 6.9 f 2.4 f 1.4 reported by NA14 ’ . 2g These results suggest that branching B(D,+-+c#d) - ratios of the qn and 7’~ decays of the D, may be much smaller than earlier indications. RESONANT SUBSTRUCTURE OF D + Kmwr DECAYS There are several D + Kerr channels with very large branching ratios. To obtain a complete picture of charm decays, it is important to understand the amplitudes which contribute to this final state. In this analysis, we measure the resonant substructure of D” --+ I<-~~+T+T- and D+ + K’T+T+T-. Branching ratios to vector-vector and axial vector-pseudoscalar final states are obtained. The data was collected with the Mark III detector at the SLAC e+e- storage ring SPEAR near the peak of the $(3776), which decays predominantly to 00. All candidate D + Kmm events are kinematically constrained to the D mass, with the recoil mass allowed to vary. The signal can then be identified as a peak in the recoil mass spectrum at the D 6 - mass. With this type of constraint, all events have the same amount of phase space for the decay throughout the entire recoil mass plot. The recoil mass plots for the two channels - - discussed here are shown in figure 8. The contribution of each decay mode into each final state is determined using a maximum likelihood fit. The likelihood function L is a function in the five dimensional phase space defined by the 4-momenta of the decay products of the D candidate. It consists of a signal-term L, and a background term Lb: L= &s/BLs+LB h/B +l The ratio of signal to background, RS,B, is calculated for each event as a function of recoil mass. The function LB is determined from a fit to the sidebands of the recoil mass plot and then fixed while the total likelihood function is fit from the signal region. For each decay chain, we include a complex amplitude which models the physical process. These amplitudes consist of a relativistic Breit-Wigner for each resonance in the decay chain, multiplied by a matrix element which depends on the spin and parity of the intermediate resonances and final decay products. These matrix elements are derived using either the Lorentz invariant amplitude formalism or the helicity formalism. L, is a coherent sum of these amplitudes, in which the relative fractions and phases are allowed to vary. The advantage of this approach is that the amplitudes provide a complete description of the processes in the five-dimensional phase space, and all the information available in the event is used in the fit. _ The results are presented in the tables below. Projections of the fit function are shown in figure 9. The fractions have been scaled so that the likelihood function is properly normalized; due to interference, they do not sum to one. In the model of BSW, the branching fractions are B(D” + I(*‘ ,‘ ) = 6.1% and B(D” t K-u:) = 5.0%. Our result for Do -+ K-a; is consistent with the theoretical expectation whereas the rate for Do + I?*‘,’ is much smaller than predicted. Table I Results for Do + IC-r+a+r- .031 f .008 f .Oll Table II Results for D+ + -0 I( r+ r+ T Amplitude Phase Fraction 4-Body Nonresonant .184 & .052 f .lO Ii--a; Branching Ratio33 1.37 f .17 .012 f .004 f .007 .612 f .053 f .15 - .O Kr( 1270)‘7r+ .OlO f .013 f .02 1.30 f .90 & (14oo)o~+ .163 f .048 f .08 .24 f .26 .081 f .OiO f .027 .q < .Oll .024 f .009 f .013 SUM M A R Y A search for fully reconstructed D,Dz decays is performed aad none are observed. We thus obtain the upper limit, B(D$ + &r+) < 4.1% at 90% C.L. A search is performed for the for, q7r and q’r decay modes of the D,. We confirm the existence of the first decay mode and set upper limits on the last two decay modes. An analysis of the resonant substructure of the four-body decay D -+ Knnn is performed. We observe a large K-a: rate and a small I?*‘ ,’ rate in the Do + I(-7r+,-,+ channel. We gratefully acknowledge the dedicated efforts of the SPEA R staff. One of us (C. Gatto) wishes to thank the Fondazione Angelo Della Riccia for their support. This work was supported by the Department .of Energy, under contracts DE-AC03-76SF00515, DEAC02-76ER01195, DE-AC03-81ER40050, DE-AM03-76SF00034 and by the National Science Foundation. 8 References 1. D. Bernstein et al., Nucl. Instrum. Methods 226, 301(1984) 2. From the Mark III measurements of aB(D$ ---f @r+) and aB(D$ + I(*‘K+) (Reference 3) and the world average value of BReoKt / Bd,t . 3. J. Adler et al., SLAC preprint SLAC-PUB-4952, 1989 (unpublished). 4. JC.-Anjos et al., Phys. Rev. Lett. 62, 125 (1989). 5. H. Albrecht et al., Phys. Lett. B 179, 398 (1986). 6. J.C. Anjos et al., Phys. Rev. Lett. 60, 897 (1988). 7. S. Barlag et al., CERN preprint CERN-EP/ 88-103, 1988 (unpublished). 8. M.P. Alvarez et al., CERN preprint CERN-EP/ 88-148, 1988 (unpublished). 9. J.A. McKenna, Ph.D. thesis, University of Toronto, report RX-1191, 1987. 10. J.C. Anjos et al., Phys. Lett. 223, 267 (1989). .m 11. The D$ lifetime, 4.36 28::; x lo-r3 s, is taken from Reference 12. 12. G.P. Yost et al. (Particle Data Group), Phys. Lett. 204B, 1 (1988). 13. M. Bauer, B. Stech and M. Wirbel, Z. Phys. C 34, 103 (1987). Revised values of al = 1.2 and a2 = -0.5 are taken from B. Stech, Heidelberg report HD-THEP87-18, 1987 (unpublished). The prediction for B(D” t K-a:) is from a private communication with B. Stech. 14. B.Yu. Blok and M.A. Shifman, [Sov. J. Nucl. Phys. 45, 135 (1987); 45, 301 (1987); _ 45, 522 (1987); 46, 767 (1987)]. 15. A. Chen et al., Phys. Rev. Lett. 51, 634 (1983). 16. G. Moneti, in Proceedings of the XXIII International Conference on High Energy Physics, Berkeley, 1986, edited by S. Loken (World Publishing, Singapore, 1987). 17. M. Althoff et al., Phys. Lett. B 136, 130 (1984). 18. W. Braunschweig et al., Z. Phys. C 35, 317 (1987). 19. H. Albrecht et al., Phys. Lett. B 146, 111 (1984). 9 20. H. Albrecht et al., Phys. Lett. B 187, 425 (1987). -. - 21. M. Derrick et al., Phys. Rev. Lett. 54, 2568 (1985). 22. S. Abachi e t al., Argonne National Laboratory preprint ANL-HEP-CP-86-71, 1986 (unpublished). 23. The Particle Data Group’s estimate is B4,t = (8 f 5)% (Reference 12 24. The .D$ lifetime, 4.36 ‘ i:$ x lo-l3 s, is taken from Reference 12. 25. B.Yu. Blok and M.A. Shifman, Yad. Fiz. 45, 841 (1987) [Sov. J. Nucl. Phys. 45, 522 (1987)]. 26. J.C. Anjos et al., Phys. Rev. Lett. 62, 125 (1989). 27. G. Wormser et al., Phys. Rev. Lett. 61, 1057 (1988). The measured cross section at fi = 29 GeV/ c2 was 0 . B(Dt + VT+) = 5.2 f 2.2 pb and 0 . B(Dt + v’.R+) = 8.4 f 3.7 pb 28. J. C. Brient, SLAC-PUB-4607, March 1988 -- .q 29. G. Wormser , preprint LAL 89-10, May 1989 30. The transverse matrix element is proportional to cos 4 sin 0: sin SE’, where $ is the angle in the Do frame formed between the decay planes of the K* and the p, 0: is the angle in the p frame between the r and the Do and @* is the angle in the Ii” frame between the I< and the Do, 31. The longitudinal matrix element is proportional to cos 0: cosBg*. 32. J. Adler et al., Phys. Rev. Lett. 60, 89 (1988); 33. Branching ratios were calculated using B(D” + K-T+T+K-) = .091 f .008 f .008 and B(D+ -+ 1?‘7r+7r?r+r-) = .066 f .015 f .015 from ref 32, and the branching ratios of the intermediate resonances to the final state being studied. 10 FIGURE CAPTIONS -. - 1) Mass distribution for Ds events. The arrows indicate the widest signal region (f20 MeV/ c2) among the double-tag final states which yield entries between 1.85 and 2.05 GeV/ c 2. The shaded histogram shows the expected signal for B4,t = 4.1%. 2) Marginal likelihood L(B+t ). Th e value B+t = 3.8% is indicated by the arrow. The vertical scale is arbitrary. 3) r+r- mass distribution for (a) data and (b) Monte Carlo events 4) 7r+7nr* mass distribution 5) v+ mass distribution, q + 3a decay mode 6) v+ mass distribution, 7 + yy decay mode 7) 7’~ mass distribution, in the mode 7’ + q7r+7rn8) Recoil mass distributions of (a) K-K+~~+T- and (b) Ii-“r+r+~- .w. 9) Projections of the fit to the Do + 1<-~+7r+n- final state for six submasses. Plot (a) shows the K-T- mass distribution which has an enhancement near threshold due to the longitudinal polarization of the al. Plot (b) shows the 3~ mass distribution which contains a large al contribution. Plots (d) and (f) show the (rRSr-)hi and the (7r+7rr-)10 mass distributions which are the higher and lower mass ~+7r- combinations. Plots (c) and (e) show the K-r+ mass distributions formed from the recoiling tracks not used to form the mass combinations shown in plots (d) and (f), respectively. Figure I I I I I I I I CT 2 wO 1.85 2.05 7-69 6419Al Figure I 0 7-69 2 4 6 Bqn+ (%I I .8 I IO 6419A2 Figure 3 &C mass distribution about _the DS mass DATA .- mass (x+7f) . r”“l”“l -“l”“j (b) eo Monte Carlo 00 40 1 1 mass (n+K) Figure 4 Fit to dGc* Mass Distribution with Ds+f&+ 0 T 5.0 2.5 _ .* D. 1.7 ’ A’ ’ ’ L9’ ’ ’ ’ L9 ’ ’ Mass ( dCck ) ’ ’ 2 ’ - ZL I Figure 5 Ds --> 7) TT, 7) --> 3 n t”“I i “1”11”“1”“““’ -I 12.5 t 7.5 5.0 2.5 0.0 1.5 ” a d 1.6” ” 1.7L ” ” 1.6” 1 ”1.9 ” ” 2I ,n, ,I2.1 Fitied 4 x Mass GeV .* Figure 6 Ds --> 77 ?T, 7 ---> 7 y 125 100 > s ;; 75 \ z ‘L 50 8 25 - ;: 0 1.5 J”““ll”“l’ll’ 1.6 1.7 1.8 Fitted q n Mass GeV 1.9 2 2.1 I . - Figure 7 Ds --> 77’ TT, 7’ --> 7) n l-r, 7) --> y 7 20 Il(‘l” “~ “‘~ / I “’ : i 15 1 %u i r in t Mi k i I 1.6 1.7 I 1.8 Fitted q TT TT TI Mass GeV I I9 _- 2 -i 2.1 - - Figure 8 300 ( I a > 200 100 1 1.82 1.84 1 86 1.88 19 K-nTTf~+nTT- Recoil Mass 50 (b) 40 30 20 rln -- 10 - u o 18 n l,,,li,,llllidlil” 1.82 1 84 1 86 K”nTT+nT+~- Recoil Mass I ,“, ? Figure 9 100 5 0 60 g 60 ‘;; .z 40 2 20 100 > $ 2 \ .i 0 0.6 0.6 1 1.2 60 60 40 -0.4 14 0.6 06 1 1.2 1 4 1 6 TT+lT+ll- M a s s (GeV) K - n - M a s s (GeV) I I I I 125 > $ 100 ? 0.6 0.6 1 1.2 75 .2 1 4 60 > $ 60 2f < 40 0 4 0 6 0.6 1 1.2 .m (Ti+Ti-)hl -Mass (GeV) ( K - m + ) , M a s s (GeV) 1 5 5 20 : 0 0.6 0.6 (K-TT+)~ 1 M a s s (GeV) 1.2 1.4 0.2 0 4 (n+n-),, 0.6 0 6 M a s s (GeV) 1 12