Studies of φ meson radiative decays with KLOE

  • Published on

  • View

  • Download

Embed Size (px)


  • 1Studies of meson radiative decays with KLOE

    The KLOE Collaboration

    Presented by P.Gauzzia

    aDipartimento di Fisica, Universita` degli Studi di Roma La Sapienza and INFN Sezione di Roma,P.le A.Moro 2, 00185 - Rome (Italy)

    A sample of 5.3 107 mesons, produced at the Frascati -factory DANE, has been used by the KLOECollaboration to study the radiative decays. The decays 0 and 00 have been exploited tostudy the scalar mesons a0(980) and f0(980). Furthermore a new determination of the mixing angle hasbeen obtained from the measurement of the ratio of the decay rates of to .

    1. Introduction

    The Frascati -factory DANE is an e+e col-lider working at the peak of the (1020) reso-nance. The peak cross section is 3300 nb.The KLOE experiment has collected an inte-grated luminosity of 25 pb1, corresponding to7.5 107, during the 2000 data taking, and 190pb1, corresponding to 5.7 108, during the2001 one. The results reported in this paper havebeen obtained from the analysis of a sample of 16

    The KLOE Collaboration: A. Aloisio, F. Ambrosino,A. Antonelli, M. Antonelli, C. Bacci, G. Bencivenni,S. Bertolucci, C. Bini, C. Bloise, V. Bocci, F. Bossi,P. Branchini, S. A. Bulychjov, R. Caloi, P. Cam-pana, G. Capon, G. Carboni, M. Casarsa, V. Casavola,G. Cataldi, F. Ceradini, F. Cervelli, F. Cevenini, G. Chie-fari, P. Ciambrone, S. Conetti, E. De Lucia, G. De Rober-tis, P. De Simone, G. De Zorzi, S. DellAgnello, A. Denig,A. Di Domenico, C. Di Donato, S. Di Falco, A. Doria,M. Dreucci, O. Erriquez, A. Farilla, G. Felici, A. Fer-rari, M. L. Ferrer, G. Finocchiaro, C. Forti, A. Franceschi,P. Franzini, C. Gatti, P. Gauzzi, S. Giovannella, E. Gorini,F. Grancagnolo, E. Graziani, S. W. Han ,M. Incagli, L. In-grosso, W. Kluge, C. Kuo, V. Kulikov, F. Lacava, G. Lan-franchi, J. Lee-Franzini, D. Leone, F. Lu, M.Martemianov,M. Matsyuk, W. Mei, L. Merola, R. Messi, S. Miscetti,M. Moulson, S. Muller, F. Murtas, M. Napolitano, A. Ne-dosekin, F. Nguyen, M. Palutan, E. Pasqualucci, L. Pas-salacqua, A. Passeri, V. Patera, E. Petrolo, L. Pontecorvo,M. Primavera, F. Ruggieri, P. Santangelo, E. Santovetti,G. Saracino, R. D. Schamberger, B. Sciascia, A. Sciubba,F. Scuri, I. Sligoi, T. Spadaro, E. Spiriti, G. L. Tong,L. Tortora, E. Valente, P. Valente, B. Valeriani, G. Ve-nanzoni, S. Veneziano, A. Ventura, G. Xu, G. W. YuE-mail:

    pb1 from the year 2000 data. The analysis ofthe 2001 data is in progress. More details on theanalyses presented here can be found in refs.[1].

    2. S(0++) (S=f0, a0)The nature of the scalar mesons (JPC=0++)

    f0(980) and a0(980) is not well established. Theyare not easily interpreted as ordinary qq mesons,belonging to the 3P0 nonet, alternative hypothe-ses have been proposed: qqqq states[2] or KKbound states[3]. According to the theory thebranching ratios of f0 and a0, aswell as the f0 and a0 mass shapes, are sensitiveto the structure of these particles[4].

    Concerning the f0, the decay f0 00has been measured, by selecting events withoutcharged tracks and with exactly ve prompt pho-tons in the detector. The same nal state hasbeen used to study the decay a0, witha0 0 and . These two processeshave been previously measured by the experi-ments SND and CMD-2 at VEPP-2M in Novosi-birsk[5]. In addition another nal state, consist-ing of two charged pions and ve prompt photons,was also used to study the decay a0, cor-responding to a0 0 and +0. Thislast decay chain has been observed by KLOE forthe rst time.The processes that are expected to contribute tothe ve photon nal state, besides the signals,are: (i) (500) with 00 ((500) is

    ELSEVIERProceedings of ICHEP 2002, pp. 677680

    S. Bentvelsen, P. de Jong, J. Koch and E. Laenen (Editors) 31st INTERNATIONALCONFERENCEONHIGH ENERGY PHYSICS AMSTERDAM

    c 2003 Elsevier Science B.V. All rights reserved.doi:10.1016/S0920-5632(02)02189-8

  • 2the other scalar meson expected in the mass re-gion below 1 GeV, recently observed at FNAL bythe E791 Collaboration[6]), (ii) 00 with 0, , (iii) e+e 0 with 0.Some contamination is expected from events withthree ((iv) , and 0) andseven ((v) , 000) prompt pho-tons, wrongly reconstructed as ve photon ones.A kinematic t is performed, by requiring the4-momentum conservation to reject events from(iv) and (v), then the best photon pairing issearched for, by associating pairs of photons to0 and , to discriminate between 00and 0. The 0 background is reducedby rejecting events with 0 invariant mass com-patible with the mass.

    The nal sample for 00 consists of

    m (MeV)


    dm x



    eV-1 )







    300 400 500 600 700 800 900 1000

    Figure 1. Dierential decay rate for 00;the solid line is the result of Fit B, individualcontributions are also shown.

    3102 events with a 20% contamination (evaluatedby Monte Carlo (MC)). The total detection e-ciency for the signal is 40%. By normalizing tothe number of produced the following branch-ing ratio can be obtained: Br( 00) =(1.09 0.03 0.05) 104 in agreement with theNovosibirsk measurements[5], with much smalleruncertainties.



    700 750 800 850 900 950 1000M (MeV)








    700 750 800 850 900 950 1000M (MeV)








    700 750 800 850 900 950 1000M (MeV)



    (x10-7 MeV-1)


    Figure 2. 0 invariant mass spectra: (a) vephoton sample; (b) + plus ve photon sam-ple. (c) Theoretical curve obtained from the com-bined t.

    For 0, 916 events have been selectedwith 30% contamination (MC), and 32% over-all eciency. By normalizing to the number of produced, KLOE obtains Br( 0) =(8.510.510.57)105, also in agreement withthe Novosibirsk measurements[5].The nal state of 0 and +0consists of two charged tracks coming from the in-teraction point and ve prompt photons; this sig-nature is unique among the possible nal states,so that the background comes from events withtwo tracks and three ( , +0),four (e+e 0, +0; KLKS ,KS 00 and KL l) or six ( KLKS +000, with prompt KL decay) promptphotons. The analysis scheme is similar to thatof the ve photon sample: 197 events have been

    678 P. Gauzzi

  • 3selected with 4 4 background events and with19% eciency. From this decay chain the Br( 0) = (7.960.600.47)105 is extracted, ingood agreement with the result of the ve photonsample.

    In order to disentangle the various contribu-

    Table 1Fit results.f0 Fit A Fit B2/ndf 109.5/34 43.2/33Mf0 (MeV) 962 4 973 1g2f0KK/(4) (GeV

    2) 1.29 0.14 2.79 0.12g2f0KK/g


    3.22 0.29 4.00 0.14g - 0.060 0.008a02/ndf 27.2/25g2a0KK/(4) (GeV

    2) 0.40 0.04ga0/ga0KK 1.35 0.09Br( 00 0) (0.5 0.5) 105

    tions, the data have been tted to theoreticalmass spectra obtained from the following model:(a) the coupling of to S(0++) is assumed to oc-cur through a charged kaon loop[4], (b) the scalarpropagator with nite width corrections is used,(c) the 00 parametrization is taken fromVDM calculations[7].Two dierent ts have been tried on the 00

    mass spectrum, after background subtraction:without (Fit A) and with (Fit B) (500);the (500) has been parametrized as a xed widthBreit-Wigner (M = 478 MeV, = 324 MeV[6])and a point-like coupling has been as-sumed. The free parameters of the ts are:the f0 mass, the coupling g2f0KK/(4), the ratiog2f0KK/g


    , the Br( 00 00) andg (only for Fit B).The parameter values are listed in tab.1, the 00 contribution turns out to be negligi-ble for both ts. Fit A gives a large 2, while FitB shows a good agreement with the experimentaldata. In g.1 the dierential decay rate is plotted,together with the various contributions from FitB; the depletion of the mass spectrum below 700MeV is produced by a large negative interferencebetween f0 and .For 0, a combined t of the two mass

    spectra, after background subtraction, has beenperformed, by xing their relative normalizationto the ratio Br( )/Br( +0). Thefree parameters are: the coupling g2a0KK/(4),the ratio ga0/ga0KK and Br( 00 0)(tab.1). The mass of the a0 is not free, it hasbeen xed to the PDG value Ma0=984.8 MeV.The results are shown in g.2; by integrating thetheoretical curve the a0 contribution can be eval-uated: Br( a0 0) = (7.40.7)105.In tab.2 the t results are compared to the predic-tions of the K+K loop model for some possiblestructures of the scalar mesons. The f0 param-eters are compatible with the qqqq model, whilethe a0 ones disagree with that model.

    3. / The mass eigenstates and are related to

    the SU(3) octet-singlet states through the pseu-doscalar mixing angle. Even in the case for twomixing angles which appears in extended chiralperturbation theory, as well as from phenomeno-logical analyses, it has been argued that the twomixing parameters in the quark avour basis areequal, apart from small terms which violate theOZI rule[8]. It is then possible to parametrize themixing with only one angle P , which can be ex-tracted from the ratio R=Br( )/Br( )[9].Moreover the large Br(BK) value ob-served[10], also raises interest about the gluo-nium contents of the [11]. This gluonium con-tents can be tested with a precise measurementof Br( ).The same nal state, + and three promptphotons, has been choosen for both decays. Thisnal state proceeds through the chains: , +0, 0 (Br 3 103) and , +, 00 (Br 2105). decays are easily selected with smallbackground; 50210 events have been counted with37% eciency. Since this decay chain is about 100times more probable than the decay, it is alsothe main source of background for . Af-ter all background removal, 120 12 5 events have been selected, with 23% eciency.Thus R = (4.700.470.31)103 is obtained.

    Studies of meson radiative decays with KLOE 679

  • 4Table 2KLOE results compared to K+K loop predictions[4].f0 KLOE ss(uu+dd)/

    2 (uu+dd)/

    2 ss

    g2f0KK/(4) (GeV2) 2.79 0.12 2.3 0.15 0.3


    0.50 0.01 0.3 0.5 2 0.5Br( 00)104 1.08 0.07 1 0.15 0.2a0 ss(uu-dd)/

    2 (uu-dd)/


    g2a0KK/(4) (GeV2) 0.40 0.04 2.3 0.15

    ga0/ga0KK 1.35 0.09 0.91 1.53Br( a0)104 0.74 0.07 2 0.2












    0 0.2 0.4 0.6 0.8 1


    Figure 3. Bounds on X and Y from SU(3) cal-culations and experimental branching fractions.

    From this ratio the mixing angle in the quarkavour basis can be extracted: P = (41.8+1.91.6)

    ,that is equivalent to P = (12.9+1.91.6) in theoctet-singlet basis, in the picture with only onemixing angle. By using the current PDG valueBr( ) = (1.297 0.003)%, it followsBr( ) = (6.10 0.61 0.43) 105; thisis the most accurate determination up to now.The above values have been obtained neglectingany gluonium contents of and . If one allowsfor gluonium:

    | = Xuu + dd

    2 + Y |ss + Z |glue

    | = Xuu + dd

    2 + Y |ss + Z |glue

    A consistency check can be performed: if Z = 0,then |Y | = cosP ; other constraints on X andY can be obtained from (1) ( )/( 0) and (2) ( )/(0 ). Then in

    the X , Y plane (g.3) the overlap region ofthe three allowed bands is consistent within onestandard deviation with no gluonium contents inthe , in fact: X2 + Y

    2 = 0.94



    1. A. Aloisio et al.(KLOE Collaboration), Phys.Lett. B536 (2002) 209; Phys. Lett. B537(2002) 21; Phys. Lett. B541 (2002) 45.

    2. R.L. Jae, Phys. Rev. D15 (1977) 267.3. J. Weinstein, N. Isgur, Phys. Rev. Lett. 48

    (1982) 659.4. N.N. Achasov, V.I. Ivanchenko, Nucl. Phys.

    B315 (1989) 465.5. M.N. Achasov et al., Phys. Lett. B485 (2000)

    349; Phys. Lett. B479 (2000) 53; R.R.Akhmetshin et al., Phys. Lett. B462 (1999)380.

    6. E.M. Aitala et al., Phys. Rev. Lett. 86 (2001)770.

    7. N.N. Achasov, V.V. Gubin, Phys. Rev. D63(2001) 094007.

    8. H. Luetwyler, Nucl. Phys. Proc. Suppl. 64(1998) 223, R. Escribano, J.M.Fre`re, Phys.Lett. B459 (1999), 288.

    9. A.Bramon et al., Eur. Phys. J. C7 (1999) 271;T.Feldmann, Int. J. Mod. Phys. A15 (2000)159.

    10. S.J. Richichi et al., Phys. Rev. Lett. 85 (2000)520; B. Aubert et al., Phys. Rev. Lett. 87(2001) 221802; K. Abe et al., Phys. Lett. B517(2001) 309.

    11. E. Kou, Phys. Rev. D63 (2001) 054027.

    680 P. Gauzzi