[Clascomment] OPT-IN:Di erential cross sections and spin density matrix elements for the reaction gamma p --> phi p

[Jean-Marc Laget]

Dear lead authors,

I do agree that your data represent a fantastic set that supersedes (not "suppresses", in caption of Fig. 18!) previous data. However, I have a few reservations on the way the previous CLAS data (Anciant et al., and McCormick et al.) are presented:

1- The primary motivation of the measurement of the t-distribution was not the study of u-channel contribution (page 19), but the study of the two gluon description of the Pomeron at large t. It turned out that u-channel contribution appeared also at the most backward angle. Not only my two gluon exchange model led to a better representation than the Pomeron at intermediate angles, but the way I treated u-channel exchange resulted in a phi-NN coupling in the range of the values found in other OZI violating processes. I do not understand why this approach is not cited. I agree that it is more relevant at higher energies, but it seems to work also around 3.5 GeV.   

2- The study of these OZI violating processes was one of the motivations, besides SCHC, of the extraction of the decay function rho_00 (K. McCormick et al. PRC69 (2004) 032203R) in the helicity frame. The u-channel clearly appeared at the most backward angles, while SCHC was satisfied to a good extent at forward and intermediate angles.

I do not understand why this work is not cited. Either your highest energy data agree with the old CLAS data and it would be worth to show the comparison, or they don't agree and one would have to understand why. In any case a Figure of the t distribution of rho_00 in the helicity frame at E_gamma=3.6 GeV would be interesting.

3- In Fig. 18, the two CLAS data sets agree around -t=2 GeV2 and above and differ by about 30% in the range -2<t<-1 GeV2. Besides the progresses in the undertanding and mastering of CLAS, as well as of in the analysis tools, this disagreement may come from the different energy binning. Have you tried to integrate your data over a larger bin? It may also come from a different treatment of the physical background under the phi peak and/or the Lambda_star cuts. Have you tried to process your data with the same procedure as in the Anciant's analysis?

Finally, you way know that I have updated my Pomeron/2gluons model by taking into account box diagrams (channel coupling) and compared it to the analysis of D. Tedeschi. Could you provide me with your experimental cross sections (dsigma/dt vs t), for updating my plot. Thanks. Did you succeed to understand the origin of the differences between your analysis and Dave's analysis (same experiment!)?.

All the best,

JM