Why study Diffractive W Boson?
Data Samples
Central and
forward electron
W boson sample:
Start with Run1b
W en candidate
sample
Z boson sample:
Start with Run1b
Z ee candidate
sample
hep-ex/0308032;Accepted by Phys. Lett B
Multiplicity in W Boson Events
Minimum side

Plot multiplicity in 3<||<5.2

-2.5 -1.5 0 1.1
3.0
5.2
Peak at (0,0) indicates diffractive W boson signal (91 events)
W Boson Event Characteristics
Standard W Events
Diffractive W Candidates
ET=35.2
ET=35.1
ET=36.9
ET=37.1
MT=70.4
MT=72.5
Observation of Diffractive W/Z
Diffractive W and Z Boson Signals


nL0
ncal nL0
Central electron W
ncal
Observed clear Diffractively
produced W and Z boson
signals
Background from fake W/Z
gives negligible change in gap
fractions
Forward electron W
DØ Preliminary
Sample
nL0
All Z
ncal
Central W
Forward W
All W
All Z
Diffractive
Probability Background
All
Fluctuates to Data
(1.08 + 0.19 - 0.17)%
7.7s
(0.64 + 0.18 - 0.16)%
5.3s
(0.89 + 0.19 – 0.17)%
7.5s
(1.44 + 0.61 - 0.52)%
4.4s
DØ/CDF Comparison
CDF {PRL 78 2698 (1997)} measured RW = (1.15 ± 0.55)% for ||<1.1 where
RW = Ratio of diffractive/non-diffractive W (a significance of 3.8s)
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!
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
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 ET jets allow smaller systematic errors
•Comparing measurements of HSD with track tag vs.
gap tag yields new insight into process
DØ Run II Diffractive Topics
Soft Diffraction and Elastic Scattering:
Rapidity Gaps:
Inclusive Single Diffraction
Elastic scattering (t dependence)
Total Cross Section
Centauro Search
Inclusive double pomeron
Search for glueballs/exotics
Central gaps+jets
Double pomeron with gaps
Gap tags vs. proton tags
Topics in RED were studied
with gaps only in Run I
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
E


<100 W boson events in Run I, >1000
tagged events expected in Run II
Run II Rapidity Gap System
VC:
5.2 <  < 5.9
LM:
2.5 <  < 4.4
 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
Forward Proton Detector Layout
p
p
D2
D1
59
57
D
A2
33
S
A1
23
Q4 Q3 Q2
Q2 Q3 Q4
Veto
0
P1U
S
P1D
23
P2O
P2I
33
Z(m)
 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 GeV2) never before
explored at Tevatron energies
 Allows combination of tracks with high-pT scattering
in the central detector
Castle Design
50 l/s ion pump
Worm gear assembly
Thin vaccum
window
Beam
Step motor
• 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
Castle Status
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.
Acceptance
Quadrupole ( p or p )
450
400
350
280
MX(GeV)
200

Dipole ( p only)
GeV2
450
400
350
20
200
GeV2
MX(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
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
V’
V
Trigger
X
17.39 mm
U’
U
X’
 Planes in a frame offset by ~2/3
fiber
 Each channel filled with four fibers
 2 detectors in a spectrometer
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.
Detector Status
• 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
M.Strang
Pot Motion Software
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.
FPD Trigger and Readout
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.
Elastic Trigger
A1U A2U
LM
VC
Pbar
Halo Early Hits
P
In-time hits in AU-PD
detectors, no early time hits,
or LM or veto counter hits
pb
p f  x  pb
p    pb
P1D
P2D
Approximately 3 million elastic triggers
taken with stand-alone DAQ
About 1% (30,000) pass multiplicity cuts
 1 x
t  (pb  pf )2
–Multiplicity cuts used for ease of
reconstruction and to remove halo spray
background
Segments to Hits
Segments
x
 Combination of fibers in a frame
determine a segment
(270 m)
y
 Need two out of three possible
segments to get a hit
10s
– U/V, U/X, V/X
• Can reconstruct an x and y
 Can also get an x directly from the
x segment
v
x
u
 Require a hit in both detectors of
spectrometer
Initial Reconstruction
P1D
beam
Y
Reconstructed 
X
P2D
beam
Y
Dead Fibers due to cables that have since been fixed
=p/p should peak
at 0 for elastic
events!!
Spectrometer Alignment
P1D x vs. P1D x (mm)
P1D y vs. P2D y (mm)
 Good correlation in hits between detectors of the same
spectrometer but shifted from kinematic expectations
– 3mm in x and 1 mm in y
Elastic Data Distributions
After alignment and multiplicity cuts (to remove background from halo spray):
=p/p
dN/dt
Acceptance
loss
Residual halo
contamination

Events are peaked at zero, as
expected, with a resolution of
s = 0.019
The fit shows the bins that will be
considered for corrected dN/dt
“Final” dN/dt Results
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
b   4.02  0.19 GeV 2
Dr. Jorge Molina,
(LAFEX, Brazil)
defended his thesis on
these results 10/31/03
(first FPD thesis!)
Dipole TDC Resolution
 Can see bunch structure of both
proton and antiproton beam
D2 TDC
p
p halo from
previous
bunch
 Can reject proton halo at dipoles
using TDC timing
 CDF does not have this capability
D1 TDC
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
Dipole Diffraction Results
Geometrical Acceptance
14σ (Monte Carlo)

Flat-t distribution
Data

0.08
0.06
0.04
0.02
0.
|t| (GeV2 )
|t| (GeV2 )
Run II Diffractive Z → µµ
Event Display
MZ all
MZ gap
Tagged Z event in FPD
gap
muons
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