1. Why study Diffractive W Boson?

2. Data Samples Z boson sample: Start with Run1b Z ee candidate sample Central and forward electron W boson sample: Start with Run1b W e candidate sample hep-ex/0308032;Accepted by Phys. Lett B

3. Multiplicity in W Boson Events  -2.5 -1.5 0 1.1 3.0 5.2  Minimum side Peak at (0,0) indicates diffractive W boson signal (91 events) Plot multiplicity in 3<|  |<5.2

4. W Boson Event Characteristics M T =70.4 E T =36.9 E T =35.2 Standard W Events Diffractive W Candidates E T =35.1 E T =37.1 M T =72.5

5. Observation of Diffractive W/Z  Observed clear Diffractively produced W and Z boson signals  Background from fake W/Z gives negligible change in gap fractions Sample Diffractive Probability Background All Fluctuates to Data Central W(1.08 + 0.19 - 0.17)% 7.7  Forward W(0.64 + 0.18 - 0.16)% 5.3  All W(0.89 + 0.19 – 0.17)% 7.5  All Z (1.44 + 0.61 - 0.52)% 4.4  n cal n L0 Diffractive W and Z Boson Signals Central electron WForward electron W All Z n cal n L0 n cal n L0 DØ Preliminary

6. DØ/CDF Comparison CDF {PRL 78 2698 (1997)} measured R W = (1.15 ± 0.55)% for |  |<1.1 where R W = Ratio of diffractive/non-diffractive W (a significance of 3.8  ) This number is corrected for gap acceptance using MC giving 0.81 correction, so uncorrected value is (0.93 ± 0.44)%, consistent with our uncorrected data value: We measured (1.08 +0.19 –0.17)% for |  |<1.1 Uncorrected measurements agree, but corrections derived from MC do not… Our measured(*) gap acceptance is (21 ± 4)%, so our corrected value is 5.1% ! (*) : derived from POMPYT Monte Carlo Comparison of other gap acceptances for central objects from CDF and DØ using 2-D methods adopted by both collaborations: DØ central jets 18% (q) 40%(g) CDF central B 22%(q) 37% overall CDF J/  29% It will be interesting to see Run II diffractive W boson results!

7. Run I Gaps Pioneered central gaps between jets, 3 papers, 3 Ph. D’s Observed and measured forward gaps in jet events at  s = 630 and 1800 GeV. Rates much smaller than expected from naïve Ingelman-Schlein model. Require a different normalization and significant soft component to describe data. Large fraction of proton momentum frequently involved in collision. Observed jet events with forward/backward gaps at  s = 630 and 1800 GeV Observed W and Z boson events with gaps

8. Run II Improvements Larger luminosity allows search for rare processes Integrated FPD allows accumulation of large hard diffractive data samples Measure , t over large kinematic range Higher E T jets allow smaller systematic errors Comparing measurements of HSD with track tag vs. gap tag yields new insight into process

9.   E Soft Diffraction and Elastic Scattering: Inclusive Single Diffraction Elastic scattering (t dependence) Total Cross Section Centauro Search Inclusive double pomeron Search for glueballs/exotics Hard Diffraction: Diffractive jet Diffractive b,c,t, Higgs Diffractive W/Z Diffractive photon Other hard diffractive topics Double Pomeron + jets Other Hard Double Pomeron topics Rapidity Gaps: Central gaps+jets Double pomeron with gaps Gap tags vs. proton tags Topics in RED were studied with gaps only in Run I 1000 tagged events expected in Run II DØ Run II Diffractive Topics

10. Run II Rapidity Gap System  Use signals from Luminosity Monitor and Veto Counters (designed at UTA) to trigger on rapidity gaps with calorimeter towers for gap signal  Work in progress (Mike Strang UTA, Tamsin Edwards U. Manchester); no time to present in this talk VC: 5.2 <  < 5.9 LM: 2.5 <  < 4.4

11. Forward Proton Detector Layout  9 momentum spectrometers comprised of 18 Roman Pots  Scintillating fiber detectors can be brought close (~6 mm) to the beam to track scattered protons and anti-protons  Reconstructed track is used to calculate momentum fraction and scattering angle –Much better resolution than available with gaps alone  Cover a t region (0 < t < 3.0 GeV 2 ) never before explored at Tevatron energies  Allows combination of tracks with high-p T scattering in the central detector D S Q2Q2 Q3Q3 Q4Q4 S A1A1 A2A2 P 1U P 2I P 2O P 1D p p Z(m) D2 D1 23 33 59 3323 0 57 Veto Q4Q4 Q3Q3 Q2Q2

12. Castle Design 50 l/s ion pump Beam Worm gear assembly Step motor Detector Constructed from 316L Stainless Steel Parts are degreased and vacuum degassed Vacuum bettter than 10 -10 Torr 150 micron vacuum window Bakeout castle, THEN insert fiber detectors Thin vaccum window

13. All 6 castles with 18 Roman pots comprising the FPD were constructed in Brazil, installed in the Tevatron in fall of 2000, and have been functioning as designed. A2 Quadrupole castle installed in the beam line. Castle Status

14. Acceptance Quadrupole ( p or ) Dipole ( only) M X (GeV) 450 400 350 280 200 GeV 2 450 400 350 20 200 M X (GeV)   Geometric  Acceptance Dipole acceptance better at low |t|, large  Cross section dominated by low |t| Combination of Q+D gives double tagged events, elastics, better alignment, complementary acceptance

15. FPD Detector Design  6 planes per detector in 3 frames and a trigger scintillator  U and V at 45 degrees to X, 90 degrees to each other  U and V planes have 20 fibers, X planes have 16 fibers  Planes in a frame offset by ~2/3 fiber  Each channel filled with four fibers  2 detectors in a spectrometer 0.8 mm 3.2 mm 1 mm 17.39 mm U U’ X X’ V V’ Trigger

16. Detector Construction At the University of Texas, Arlington (UTA), scintillating and optical fibers were spliced and inserted into the detector frames. The cartridge bottom containing the detector is installed in the Roman pot and then the cartridge top with PMT’s is attached.

17. 20 detectors built over a 2+ year period at UTA. In 2001-2002, 10 of the 18 Roman pots were instrumented with detectors. Funds to add detectors to the remainder of the pots have recently been obtained from NSF (should acknowledge funding from UTA REP, Texas ARP, DOE, and Fermilab as well). During the shutdown (Sep-Nov. 2003), the final eight detectors and associated readout electronics have been installed. P2 Quadrupole castle with up and down detectors installed Detector Status M.Strang

18. Pot motion is controlled by an FPD shifter in the DØ Control Room via a Python program that uses the DØ online system to send commands to the step motors in the tunnel. The software is reliable and has been tested extensively. It has many safeguards to protect against accidental insertion of the pots into the beam. Pot Motion Software

19. FPD Trigger and Readout

20. Stand-alone DAQ Due to delays in DØ trigger electronics, we have maintained our stand-alone DAQ first used in the fall 2000 engineering run. We build the trigger with NIM logic using signals given by our trigger PMT’s, veto counters, DØ clock, and the luminosity monitor. If the event satisfies the trigger requirements, the CAMAC module will process the signal given by the MAPMT’s. With this configuration we can read the fiber information of only two detectors, although all the trigger scintillators are available for triggering.

21. In-time hits in AU-PD detectors, no early time hits, or LM or veto counter hits A1UA2U P2D P1D P Pbar Halo Early Hits LM VC  Approximately 3 million elastic triggers taken with stand-alone DAQ  About 1% (30,000) pass multiplicity cuts –Multiplicity cuts used for ease of reconstruction and to remove halo spray background Elastic Trigger

22. Segments to Hits y x 10  u x v Segments (270  m)  Combination of fibers in a frame determine a segment  Need two out of three possible segments to get a hit –U/V, U/X, V/X Can reconstruct an x and y  Can also get an x directly from the x segment  Require a hit in both detectors of spectrometer

23.  =  p/p should peak at 0 for elastic events!! Dead Fibers due to cables that have since been fixed P1D P2D beam Y X Y Reconstructed  beam Initial Reconstruction DØ Preliminary

24. Spectrometer Alignment  Good correlation in hits between detectors of the same spectrometer but shifted from kinematic expectations –3mm in x and 1 mm in y P1D x vs. P1D x (mm) P1D y vs. P2D y (mm) DØ Preliminary

25. After alignment and multiplicity cuts (to remove background from halo spray):  Events are peaked at zero, as expected, with a resolution of   = 0.019 Elastic Data Distributions dN/dt  =  p/p Acceptance loss Residual halo contamination The fit shows the bins that will be considered for corrected dN/dt

26. After unsmearing and acceptance corrections,the data points were normalized to the points obtained by the E710 experiment The results are in excellent agreement with the model of M. Bloch showed in the figure Warning: error bars need the contribution of the unsmearing errors  in progress “Final” dN/dt Results Dr. Jorge Molina, (LAFEX, Brazil) defended his thesis on these results 10/31/03 (first FPD thesis!)

27. Dipole TDC Resolution  Can see bunch structure of both proton and antiproton beam  Can reject proton halo at dipoles using TDC timing  CDF does not have this capability D1 TDC D2 TDC p p halo from previous bunch

28. Standalone Readout vs. AFE Readout Standalone Readout  Uses a trigger based on particles passing through trigger scintillators at detector locations  No TDC cut –this cut removes lower correlation (halo) in y plot  Diffracted pbars fall in upper correlation of y plot AFE readout  Similar correlations  Uses trigger: –one jet with 25GeV and North luminosity counters not firing  This trigger suppresses the halo band

29. Dipole Diffraction Results  Geometrical Acceptance 14σ (Monte Carlo) Data |t| (GeV 2 ) Flat-t distribution 0. 0.02 0.06 0.04 0.08  |t| (GeV 2 )

30. M Z all M Z gap Tagged Z event in FPD Event Display gap muons Run II Diffractive Z → µµ

31. Summary and Future Plans  Early FPD stand-alone analysis shows that detectors work, will result in elastic dN/dt publication (already 1 Ph.D.)  FPD now integrated into DØ readout (detectors still work)  Commissioning of FPD and trigger in progress  Full 18 pot FPD will start taking data after shutdown (12/03)  Tune in next year for first integrated FPD physics results