A spin analysis of the 4 p channels produced in central pp interactions at 450 GeV r c WA

Ž 0. 0 0 0 0 0 q y q y q y 0 0 The reactions ppTMp X p , where X is observed decaying to p p p p , p p p p and p p p p , have f s Ž . Ž . Ž . been studied at 450 GeVrc. There is evidence for an a 1320 p decay mode of the h 1645 and h 1870 in the 2 2 2 q y 0 0 q y q y Ž . p p p p and p p p p final states. The f 1950 is consistent with being a single resonance with a dominant 2 Ž . Ž . Ž . f 1270 pp decay mode. The f 1370 is found to decay dominantly to rr while the f 1500 is found to decay to rr and 2 0 0 ss . q 2000 Published by Elsevier Science B.V. All rights reserved. 0370-2693r00r$ see front matter q 2000 Published by Elsevier Science B.V. All rights reserved. Ž . PII: S0370-2693 99 01413-6 ( ) D. Barberis et al.rPhysics Letters B 471 2000 440–448 441 The WA76, WA91 and WA102 Collaborations have studied the centrally produced pppp final state in the reaction ppTMp pppp p 1 Ž . Ž . f s w x w x w x at 85 1 , 300 2 , and 450 GeVrc 3,4 . The subscripts f and s indicate the fastest and slowest particles in the Ž . laboratory respectively. In addition to the f 1285 , which was observed at all energies, peaks were observed at 1 1.45 and 1.9 GeV in the 300 and 450 GeVrc data. In contrast, no clear evidence was seen for these states in the w x 85 GeVrc data of the same experiment 1 . The increased prominence of these states with increased incident w x energy 2 is consistent with the formation of these states via a double Pomeron exchange mechanism, which is w x predicted to be a source of gluonic states 5 . Ž . G P C q qq Ž . The peak at 1.9 GeV, called the f 1950 , was found to have I J s0 2 and decay to f 1270 pp and 2 2 Ž . w x a 1320 p 3,4 . It was not possible to determine whether this was one resonance with two decay modes, or two 2 resonances. However, in a recent analysis of the centrally produced hpp final state by the WA102 w x P C qq Ž . experiment 6 , no evidence was found for a J s2 a 1320 p decay mode. The peak at 1.45 GeV was 2 shown to have I J P C s00 and decay to rr. It has been described as being due to the interference between Ž . Ž . w x the f 1370 and the f 1500 3,4 which implies that both states should have a substantial rr decay mode. 0 0 w x Ž . However, the initial analysis of the 4p channel in pp by Crystal Barrel 7 showed that the f 1500 decays via 0 w x ss with a very weak coupling to rr. A recent preliminary analysis 8 of the Crystal Barrel data concludes that Ž . Ž . ) ) the f 1370 decays strongly to ss and rr while the f 1500 decays to ss and p p where the p p decays 0 0 to rp . Ž . Ž . Ž . Therefore, there are still several unanswered questions on the decays of the f 1950 , f 1370 and f 1500 . 2 0 0 In order to address these questions, in this paper an analysis of several different 4p decay modes will be Ž . Ž . presented. In particular, to investigate the ss contribution in the decays of the f 1370 and f 1500 . One of 0 0 the major problems of studying the pppp final state is the number of possible isobar decay modes that are present. Therefore a study will be made firstly of the p p p p 0 final state which has the nice feature that it Ž . Ž . Ž . is free from any contribution from r 770 , a 1260 or a 1320 isobars. Next an analysis will be presented of 1 2 the ppp p 0 decay mode which has different systematic effects compared to the pppp decay. Finally a reanalysis of the pppp decay mode will be presented. The data come from the WA102 experiment which has been performed using the CERN Omega Spectromew x ter, the layout of which is described in Ref. 9 . The reaction ppTMp p p p p 0 p 2 Ž . Ž . f s Ž . has been studied at 450 GeVrc. Reaction 2 has been isolated from the sample of events having two outgoing charged tracks and eight g s reconstructed in the GAMS-4000 calorimeter, by first imposing the following cuts < < < < on the components of missing momentum: missing P -14.0 GeVrc, missing P -0.20 GeVrc and x y < < missing P -0.16 GeVrc, where the x axis is along the beam direction. A correlation between pulse-height z and momentum obtained from a system of scintillation counters was used to ensure that the slow particle was a proton. Fig. 1a shows the two photon mass spectrum for 8g-events when the mass of the other three 2g-pairs lies 0 Ž . 0 within a band around the p mass 100–170 MeV . A clear p signal is observed with a small background. Ž . Ž Events belonging to reaction 2 have been selected using a kinematical fit 8C fit, four-momentum conservation 0 . 0 being used and the masses of four p s being fixed . Events corresponding to the reaction hp , where the h 0 Ž 0. 0 0 0 0 decays to 3p , have been removed by requiring M 3p G0.6 GeV. The major background to the p p p p X Ž . 0 0 0 0 final state comes from the decay of the h and f 1285 to hp p where the h decays to 3p and one p is 1 undetected. This produces events in the threshold to 1.3 GeV region of the p p p p 0 mass spectrum. A kinematical fit has been used to remove these events. Using real hpp events a simulation has been performed to estimate the residual contamination in the p p p p 0 final state which is found to be less than 5 %. ( ) D. Barberis et al.rPhysics Letters B 471 2000 440–448 442 0 0 0 0 . Ž . 0 . Ž 0 0 0 0 . . Ž 0 0 . Fig. 1. The p p p p channel. a M gg when the other three gg pairs lie in the p region, b M p p p p and c M p p for Ž 0 0 0 0 . . P C qq . P C qq Ž . 1.3F M p p p p F1.5 GeV. Results of the spin analysis d the J s0 ss wave and e the J s2 f 1270 pp wave with 2 fits described in the text. The resulting p p p p 0 effective mass spectrum is shown in Fig. 1b and consists of 1438 events. There is a Ž . broad enhancement in the 2 GeV region, compatible with the f 1950 , but there is no clear peak in the 1.45 2 GeV region. One of the aims of this paper is to investigate the role of the s in the decays of the 4p final state. In order to Ž . determine the amount of ss a parameterization of the pp S-wave s is required. Parameters for the s may be process dependent. The nature of the s is not clearly established and parameterizations of it may be process dependent. It may be a resonant state or just two pions produced in relative S-wave. However, the parameterization of the pp S-wave is crucial in all these analyses in order to determine the amount of ss observed. Assuming that the 1.3–1.5 GeV p p p p 0 mass region is dominated by the decay to ss , then in order to determine the best parameters for the s , a study has been made of the p p 0 mass spectrum for the Ž 0 0 0 0 . region 1.3FM p p p p F1.5 GeV, shown in Fig. 1c. A Monte Carlo simulation has been performed using the decay X 0 TMss with sTMp p . Formally, we describe the shape of the s using a spin 0 relativistic Breit–Wigner amplitude of the form q G m0 A M s . Ž . pp 2 2 ž / q m ym y im G 0 0 pp 0 < Ž . < 2 In order to describe the centrally produced mass spectrum the resulting function A M has been multiplied pp Ž 2 .1r2 3 w x by the kinematical factor M y4m rM 10 . The mass and width of the s have been varied and the pp p pp ( ) D. Barberis et al.rPhysics Letters B 471 2000 440–448 443 Monte Carlo data passed through the detector simulation and the resulting mass distributions were compared to the real data. The parameters that best describe the data are m s450"50 MeV and Gs600"50 MeV and 0 Ž . 0 0 the resulting mass spectrum solid histogram is superimposed on the real p p mass spectrum shown in Fig. 1c. A spin-parity analysis of the p p p p 0 channel has been performed in 120 MeV p p p p 0 mass bins w x Ž . using an isobar model and the method described in Ref. 3 , assuming that only the ss and f 1270 pp 2 intermediate states with J P C s0 or 2 contribute. The data can be fitted using only the J P C s0 ss P C qq Ž . wave and the J s2 f 1270 pp wave with J s0 which are shown in Fig. 1d and Fig. 1e respectively. 2 Z Ž . Superimposed on Fig. 1d is the distribution that would be expected for the f 1500 ; as can be seen there is little 0 Ž . Ž . evidence for any f 1370 contribution. Superimposed on Fig. 1e is what would be expected from the f 1950 ; 0 2 P C qq Ž . as can be seen it well describes the J s2 f 1270 pp wave, showing that the peak at 1.95 GeV is 2 consistent with being a single resonance. An analysis has next been performed on the centrally produced ppp p 0 system. The reaction ppTMp ppp p 0 p 3 Ž . Ž . f s has been isolated from the sample of events having four outgoing charged tracks and four g s reconstructed in the GAMS-4000 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 missing P -0.12 GeVrc. The two photon mass x y z Ž . 0 Ž spectrum not shown when the mass of the other 2g-pair lies within a band around the p mass 100–170 . 0 Ž . MeV shows a clear p signal with small background. Events belonging to reaction 3 have been selected Ž 0 using a kinematical fit 6C fit, four-momentum conservation being used and the masses of two p s being . qqŽ . Ž q. fixed . Events containing a fast D 1232 were removed if M p p -1.3 GeV, which left 283 408 centrally f produced ppp p 0 events. Fig. 2a shows the ppp p 0 effective mass spectrum. The mass spectrum is very similar to that observed q y q y w x Ž . previously in the p p p p channel 4 , namely, a clear peak at 1.28 GeV associated with the f 1285 , a 1 peak at 1.45 GeV and a broad enhancement around 2.0 GeV. A spin-parity analysis of the ppp p 0 channel has been performed using an isobar model and the method w x described in Ref. 3 . Assuming that only angular momenta up to 2 contribute, the intermediate states considered are