Events with a rapidity gap between jets in pbar-p collisions at sqrt{s}=630 GeV
We report a measurement of the fraction of dijet events with a rapidity gap between jets produced by color-singlet exchange in ppbar collisions at sqrt{s}=630 GeV. In events with two jets of E_T > 8 GeV, pseudorapidity in the range 1.8 < |eta| < 3.5, and (eta_1)(eta_2) < 0, the color-singlet-exchange fraction is found to be
We compare events with dijets on opposite sides and on the same side of the central rapidity region. Same side (SS) dijet events should not have central rapidity gaps (no tracks or hit towers in the central region) except due to multiplicity fluctuations. Opposite side (OS) dijet events should contain a sample of events with central rapidity gaps due to color-singlet exchange (CSE). We see a clear excess of OS dijets events over SS with no tracks or hit towers in the central region.
Figure 1: Multiplicity distributions (a) for tracks and (b) for calorimeter towers in the regions eta<0.9 for opposite-side [OS, (eta_1)(eta_2) < 0] dijet events (solid lines), and eta<1.05 (eta<1.2) for tracks (towers) for same-side [SS, (eta_1)(eta_2) > 0] dijet events (dashed lines); (c,d) the bin-by-bin difference between OS and SS events normalized to the number of SS events. The SS distribution is scaled to the OS one by the ratio of OS/SS events for N_track>0 in (a) and for N_tower>2 in (b).
fig1.eps, fig1a.eps, fig1b.eps, fig1c.eps, fig1d.eps
OS events with 0 tracks and 0-2 hit towers are taken as the gap sample in the following plots. Based on the number of SS events with 0 tracks, 0-2 towers, we estimate the background in this gap sample to be 45%. The control sample in the following plots contains the OS events with 1-3 tracks and 0-6 hit towers, and is normalized to 45% of the gap sample to represent the background. The CSE fraction is then taken to be the (gap-background)/total.
Figure 2: Ratios at 630 GeV of gap events (solid points) and background events (open circles: control sample events normalized to the estimated 45% background) to all events as a function of (a) half the pseudorapidity separation between the jets and (b) the average transverse energy of the jets.
fig2.eps, fig2a.eps, fig2b.eps
The CSE fraction appears to fall slightly as the jets approach the edges of the acceptance. We see no strong dependence of CSE fraction on mean dijet E_T.
Figure 3: Ratio of color-singlet exchange to total number of events at 630 and 1800 GeV as a function of (a) half the pseudorapidity separation between the jets and (b) the average transverse energy of the jets. The solid (dashed) lines represent the average ratio R_1800 (R_630).
fig3.eps, fig3a.eps, fig3b.eps
We see no strong dependence of CSE fraction on x_T = 2 E_T/sqrt{s} or x = e^{eta} x_T/2. If the CSE coupled to quarks but not to gluons, the CSE fraction would be expected to follow the HERWIG result. The flat x-dependence indicates that the relative strength of the CSE effective coupling to quarks and gluons is comparable in magnitude to that of the color-exchange coupling.
Figure 4: Ratio of color-singlet exchange to total number of events at 630 (open circles) and 1800 (black circles) GeV as a function of (a,c) x, the ratio of the jet momentum along the beam to the momentum of the beam (two entries per event, one for each of the two leading jets) and (b,d) x_T*, the average scaled transverse energy of the two jets. The solid (dashed) lines in (a,b) represent the average ratio R_1800 (R_630). The solid lines in (c,d) represent the distributions of the fraction of CE dijet events due to quark-(anti)quark scattering to all CE dijet events produced by a HERWIG Monte Carlo simulation, including a simulation of the CDF detector. The normalization of the Monte Carlo result was adjusted to yield the best fit to the data.
fig4.eps, fig4a.eps, fig4b.eps, fig4c.eps, fig4d.eps
Mary Convery,
Dino Goulianos
Rockefeller University
Date blessed: 12/18/97, 5/7/98
last updated 7/31/00
convery@rock16.rockefeller.edu