A measurement of the branching fractions ž / ž / of the f 1285 and f 14201 1 produced in central pp interactions at 450 GeV r c WA

Ž . Ž . A study of the f 1285 and f 1420 produced in central pp interactions has been performed. For the first time in a 1 1 Ž . single experiment the branching fractions of both mesons in all major decay modes have been determined. Both the f 1285 1 Ž . and f 1420 are consistent with being produced by double Pomeron exchange. q 1998 Elsevier Science B.V. All rights 1 reserved. P C qq Ž . In central production the J s1 f 1285 and 1 Ž . w x f 1420 are clearly observed 1,2 . In contrast, there 1 is no evidence for any 0 contribution in the 1.28 w x yq and 1.4 GeV regions 1–3 . This suppression of 0 states in central production was first observed by experiment WA76 at the CERN Omega Spectromew x ter 1 and has been independently confirmed by the w x E690 experiment at Fermilab 4 . Although the reason for this suppression of 0 states is not known it has a very important implication since it can help us determine more precisely the characteristics of the Ž . Ž . f 1285 and f 1420 than is possible in other exper1 1 iments which see both 0 and 1 states. Experiment WA102 is designed to study centrally produced exclusive final states formed in the reaction ppTMp X 0 p 1 Ž . Ž . f s at 450 GeVrc. The subscripts f and s indicate the fastest and slowest particles in the laboratory respectively and X 0 represents the central system that is presumed to be produced by double exchange processes. The experiment has been performed using the CERN Omega Spectrometer, the layout of which is w x described in Ref. 5 . Analyses of the K K p . and pppp S w x channels have been published previously 6,7 . This Ž . Ž . paper presents a study of the f 1285 and f 1420 1 1 through an analysis of the hpp, r g and fg channels. The reaction ppTMp hpp p 2 Ž . Ž . f s where the h has been observed decaying to gg and ppp , has been isolated from the sample of events having four and six outgoing charged tracks respectively plus two g s each with energy greater than 0.5 GeV reconstructed in the electromagnetic calorimeter , by first imposing the following cuts on < the components of the missing momentum: missing < < < P -17.0 GeVrc, missing P -0.16 GeVrc and x y < < missing P -0.12 GeVrc, where the x-axis is along z the beam direction. A correlation between pulseheight and momentum obtained from a system of scintillation counters was used to ensure that the slow particle was a proton. For the case hTMgg the effective mass of the Ž . two g s is shown in Fig. 1 a for the events contributŽ . q y ing to the f 1285 mass region of the hp p 1 spectrum. The gg mass spectrum shows a clear h Ž . signal s s 31 MeV with little background, which Ž . was selected by requiring 0.45 m gg 0.65 GeV. For the case hTMppp 0 the effective mass q y 0 Ž . of the p p p is shown in Fig. 1 b for the events Ž . contributing to the f 1285 mass region of the 1 hpp spectrum. The ppp 0 mass spectrum Ž . shows a clear h signal s s 11 MeV with effectively no background, which was selected by requirŽ q y 0. ing 0.5 m p p p 0.6 GeV. The quantity 2Ž . 2Ž q y. D, defined as D s MM p p y M hp p , f s 2Ž . where MM p p is the missing mass squared of f s the two outgoing protons, was then calculated for < < 2 each event and a cut of D F 3.0 GeV was used to select the hpp channel. Events containing a qqŽ . Ž q. fast D 1232 were removed if M p p -1.3 f GeV, which left 90 393 centrally produced hpp events for hTMgg and 24 585 for hTMppp . After the selection of the hpp channel a kinematical fit was performed in order to apply overall energy and momentum balance. 1 The showers associated with the impact of the charged tracks on the calorimeter have been removed from the event before the requirement of only two g s was made. ( ) D. Barberis et al.rPhysics Letters B 44