Transcript Event Shape Variables and DIS at HERA

Deuterons in Photoproduction at
HERA
Erik Strahler, University of Wisconsin
2004 Summer Project Report
August 5, 2004
Outline:
Introduction
Cuts
dE/dx
Mean Particle Curves
Sigmas
Results
Conclusion
E.Strahler, U.Wisconsin, Deuterons in Photoproduction, Summer 2004 Report
August 5, 2004 - 1
Introduction
Inclusive Photoproduction sample
1998 + part of 2000 data (about 1/3)
~15.4 pb-1
Search for a deuteron (anti-deuteron) signal
Form of signal tells us about whether deuterons
are produced by direct hadronization, or
whether by a ‘gluing together’ process
afterwards.
E.Strahler, U.Wisconsin, Deuterons in Photoproduction, Summer 2004 Report
August 5, 2004 - 2
Cuts
Online:
Offline:
• Ntrkvtx r 5
• PT r 150 MeV
• -1.75 < h < 1.75
• Reject real electrons
probability > 90% AND
energy > 5 GeV AND
yel < 0.8
• |Zvtx| < 30 cm
• 165 < Wjb < 252 GeV
(from ZUFOS)
• r 5 primary vertex tracks
• > 40 CTD hits (stereo + axial)
• > 5 hit superlayers
E.Strahler, U.Wisconsin, Deuterons in Photoproduction, Summer 2004 Report
August 5, 2004 - 3
dE/dx
dE/dx plotted against track
momentum
Clear bands corresponding
to p, K, p, and d
Create preliminary curves
to roughly separate the
particle bands
E.Strahler, U.Wisconsin, Deuterons in Photoproduction, Summer 2004 Report
August 5, 2004 - 4
Mean Particle Curves
•Want to fit a curve to each
particle band
•Split dE/dx vs. p into 15
momentum slices
•Plot dE/dx in each bin, and
fit gaussians.
•Large peak belongs to p
(K are buried). Smaller
peak in is protons.
•In log scale you can see
the small deuteron
contribution.
E.Strahler, U.Wisconsin, Deuterons in Photoproduction, Summer 2004 Report
August 5, 2004 - 5
Mean Particle Curves …
•Plot Gaussian fit means
vs. center of momentum
bins
•Fit with parameterization
of the Bethe-Bloch formula
dE/dx = C0 / p2 +C1
•deuteron statistics are not
high enough for real
particle identification
•proton statistics are quite
good, so we’ll use them to
get the cut
E.Strahler, U.Wisconsin, Deuterons in Photoproduction, Summer 2004 Report
August 5, 2004 - 6
# of sigmas cut
Vary number of sigmas
away from mean particle
curve
y axis shows total number
of tracks above mean curve
minus number lying
between mean curve and
sigma curve n.
A place with little change
between adjacent sigma
curves is thus ideal for a
cut
E.Strahler, U.Wisconsin, Deuterons in Photoproduction, Summer 2004 Report
August 5, 2004 - 7
Final Cuts
•Above method used to
determine upper cuts.
•Start with 3 s curve and
vary by trial and error to get
a rough cut for the lower
edge of proton band
•Make universal cut of
dE/dx > 1.8 mips to
eliminate p, K background.
E.Strahler, U.Wisconsin, Deuterons in Photoproduction, Summer 2004 Report
August 5, 2004 - 8
Results
Deuterons:
1998:
d/p = 0.023
p/total = 0.0057
1998+2000:
d/p = 0.034
p/total = 0.0055
Possible deuteron
excess in 2000 data
Circled area: remnants from the
proton band that escaped the
cut. A larger remnant could
explain the 2000 excess.
E.Strahler, U.Wisconsin, Deuterons in Photoproduction, Summer 2004 Report
August 5, 2004 - 9
Results…
Antideuterons:
1998:
dbar/pbar = 0.0017
pbar/total = 0.0055
1998+2000:
dbar/pbar = 0.0017
pbar/total = 0.0054
Once again, uncut proton
contamination.
Good agreement
E.Strahler, U.Wisconsin, Deuterons in Photoproduction, Summer 2004 Report
August 5, 2004 - 10
Conclusion
• Only first step towards deuteron cross-section
• Need to consider whether background is
present, and how to remove
• Measured value of dbar/pbar = 0.0017 is about
twice that measured by H1. This could be due
to lack of full background studies.
• Further work is called for!
E.Strahler, U.Wisconsin, Deuterons in Photoproduction, Summer 2004 Report
August 5, 2004 - 11