PowerPoint-Pr

1
13th International Workshop on Deep Inelastic
Scattering Madison, April 27th May 1st , 2005
Analysis of the anti-charmed baryon state at H1
Karin Daum Wuppertal
on behalf of

• Outline
• Observation of the Dp(3100) 1) resonance at H1
• Summary from searches for Dp(3100)
• Model assumptions for the analysis
• Acceptance corrected ratios ?(Dp(3100))/?(D)
• Conclusions

1) Since the spin is unknown Dp(3100) rather
than ?c will be used
2
Observation of the Dp(3100) resonance _at_ H1
A.Atkas et al., Phys. Lett. B588(2004)17.
HERA-I, 75 pb-1
DIS 1 GeV2 lt Q2 lt 100 GeV2
Photoproduction Q2 lt 1 GeV2
Confirmed by independent photoproduction sample
Background fluctuation probability 4 x 10-8
(Poisson) ? 5.4 ? (Gauss)
Preliminary _at_ ICHEP2004 R(Dp(3100)/D)
1.46?0.32
Bare rate uncorrected
3
Results of Dp(3100) searches
H1 observation in ep ? cc X
Negative results for qc from ALEPH ee- ? Z0
? cc FOCUS ?N ? cc X CDF pp ? cc
X BELLE ee- ? ?(4s) ? B0B0 ZEUS ep ?
cc X
Different physics processes investigated (except
ZEUS) Detailed analysis of Dp(3100) from H1
needed
4
Model assumptions for the analysis
Basic production process of charmed hadrons BGF

• Assumption
• c-quarks from the hard sub-process
• interacts with QCD vacuum to create
• charmed hadrons
• e.g. D, ?c, Dp(3100)
• (ordinary fragmentation process)
• Technical procedure for
• correcting data
• Use RAPGAP 3.1
• Mimic Dp(3100) by appropriate
• modification of mass and decay
• modes of D1 and D2
• No spin assignment done, i.e.
• isotropic decay

LO QCD
The model will be normalised to the total
Dp(3100)/D yield when comparing with data
pseudo-rapidity ? - log(tan(?/2)) D-inelastici
ty z (PpD)/(Pq)
5
Acceptance corrected Rcor(Dp(3100)/D)
Kinematic region 1ltQ2lt100 GeV2 0.05ltyelt0.7

• 1. In the visible D range as given in our
  publication
• Visible Dp range Pt(Dp)gt1.5 GeV,
  -1.5lt?(Dp)lt1
• Visible D range Pt(D)gt1.5 GeV, -1.5lt?(D)lt1,
  z(D)gt0.2
• (applied to inclusive D and to Ds from
  Dp(3100) decay)

preliminary
95 Upper limit from ZEUS for DIS lt0.59 in
different phase space Q2gt1 GeV2
yelt0.95 pt(D)gt1.35 GeV, ?(Dp)lt1.6,
pt(D)/?Et?gt10gt0.12
Systematic errors include uncertainties due to
D, Dp selection, veto for D1D2, background
shape, dE/dx-measurement, Variation of Dp(3100)
fragmentation and pseudo-rapidity ?
6
Acceptance corrected Rcor(Dp(3100)/D)
Kinematic region 1ltQ2lt100 GeV2 0.05ltyelt0.7

• 1. In the visible D range as given in our
  publication
• Visible Dp range Pt(Dp)gt1.5 GeV,
  -1.5lt?(Dp)lt1
• Visible D range Pt(D)gt1.5 GeV, -1.5lt?(D)lt1,
  z(D)gt0.2
• (applied to inclusive D and to Ds from
  Dp(3100) decay)

preliminary

• 2. Extrapolated to the full D phase space in
  Dp(3100) decay
• Visible Dp/D range Ptgt1.5 GeV, -1.5lt?lt1
• (applied to D for inclusive D and to Dp for
  Dp(3100))

Systematic errors include uncertainties due to
D, Dp selection, veto for D1D2, background
shape, dE/dx-measurement, Variation of Dp(3100)
fragmentation and pseudo-rapidity ?
7
?(Dp(3100))/?(D) vs. event kinematics
Kinematic region 1ltQ2lt100 GeV2 0.05ltyelt0.7
Q2

-
8
?(Dp(3100))/?(D) vs. event kinematics
Statistical errors only
Kinematic region 1ltQ2lt100 GeV2 0.05ltyelt0.7
W
Q2
Q2


s
-
9
?(Dp(3100))/?(D) for D observables
Statistical errors only
Kinematic region 1ltQ2lt100 GeV2 0.05ltyelt0.7
Lab. frame
zD-cut replaced by xobs(Dp)-cut

• Compared to normal D production
• D from Dp(3100) are
• Suppressed for central ? in the lab.
• Significantly softer in pt(D) and z(D)
• Closer to photon direction in ?p
• The simple MC approach does not
• describe the data

?p- frame
10
?(Dp(3100)) for Dp observables
Statistical errors only
Kinematic region 1ltQ2lt100 GeV2 0.05ltyelt0.7
Lab. frame
zD-cut replaced by xobs(Dp)-cut
?p- frame

• Dp(3100) production is
• Suppressed for central ? in the lab.
• Close to photon direction in ?p
• (These features are not described by
• the simple MC approach)
• MC approach in reasonable agreement
• with pt- and z-distributions of Dp(3100)

11
Fragmentation function of Dp(3100), D

• Analysis1 performed in ?p
• Project all particles in the ? hemisphere
• into plane perpendicular to ? direction
• Divide event into 2 hemispheres
• defined by the D direction
• Sum up all particles in D hemisphere
• ? c-quark (including QCD-effects)

1)Similar to analysis of Zuzana Rurikova, HFWG,
Session 2
zD-cut replaced by xobs(Dp)-cut
Statistical errors only

• D from Dp(3100) gets very
• little energy from c-quark
• Dp(3100) fragmentation is hard
• (as expected from its mass)

12
Conclusions

• Preliminary results on acceptance corrected ratio
  Dp/D
• in DIS in the visible D region is
• Rcor(Dp(3100)/D) 1.590.33
• Ds from Dp(3100) decay are significantly
  softer
• than normal Ds
• Dp(3100) production in central ?lab suppressed
• Dp(3100) produced close to the photon direction
• Dp(3100) fragmentation is hard
• The simple fragmentation approach with isotropic
  decay
• does describe W and Q2 of Dp(3100) production
• does not describe D properties from Dp(3100)
  decay
• does reasonably well for properties of Dp(3100),
  
• except for ?lab and ?

0.33 -0.45
13
Backup slides
Physics related slides
14
Remarks on Dp search by ZEUS

• We observe
• D from Dp(3100) decay take only little energy
• of the event
• Production of Dp(3100)is different in ? from
• inclusive D production
• The charged and neutral multiplicity in Dp(3100)
  events
• tends to be higher than in ordinary D events

The ZEUS cut pt(D)/?Et?gt10 for background
suppression is designed just to remove high
multiplicity events with little energy for the
D Furthermore The kinematic visible D
regions are not directly comparable
15
Remarks on Dp search by ALEPH
Dsignal form PL B599(2004)1
Distributions suggest that Ds with large XE are
favoured
Furthermore Decay length cut used in PL B599 is
a veto for small XE according to EPJ C16.
Limit ?(?c?Dp)/?(D)lt0.3
Dsignals vs. XE from EPJ C16(2000)597
16
Remarks on Dp search by ALEPH
Rb ?22, Rc ?17
D _at_ LEP are produced predominantly by beauty
ltxEgtcc ?0.488
Similar shapes
17
Remarks on Dp search by ALEPH
Ds from charm
At LEP D fragmentation function significant
softer than at HERA due to QCD evolution Ds
from ?c should lead to a shift in XE by about 0.3
ALEPH, EPJ C16(2000)597
ALEPH Limit ?(?c?Dp)/?(D)lt0.3 Likely to be
NOT in disagreement
For xobs(D)gt0.7 ?(?c?Dp)/?(D)0.170.13
18
Remarks on Dp search by Belle
Exclusive channel in B0 decay B(B0 ? ?cp?)?B(?c
?Dp)/ B(?c ?Dpp?)lt11 _at_ 90 C.L. They
indirectly conclude from their limit on B(B0 ?
?cp?)?B(?c ?D-p)/ B(?c ?D-pp?)lt1.2 _at_ 90
C.L. by arguing the ?c decay into pseudoscalar
plus proton should be favoured that they are not
in agreement with H1 BUT it is not clear which
decay mode is favoured, Depends on spin of the
Dp(3100)
Not in contradiction with H1 result
19
Remarks on Dp search by CDF
Charm production via gluon gluon fusion Similar
to BGF at HERA CDF charm trigger sensitive to
central rapidity in c.m.s. ?lt0.7 with 2 svtx
tracks with ptgt2 GeV H1 sees Soft Ds from
Dp(3100) not central in c.m.s.
20
Remarks on Dp search by FOCUS
Fixed target experiment 180 GeV photons on
9Be ?hadronic mass W18 GeV
Hera 60ltWlt280 GeV
Hera
D
?c threshold
M(DD)
D
nucleon
Large phase space suppression for ?c in FOCUS No
Monte Carlo used by FOCUS
21
D fragmentation
Here Dp(3100) Is contributing
22
Backup slides
Analysis related slides
23
Typical Dp candidates
HERA-I
All events scanned No anomalies observed e.g.
split tracks, wrong reconstruction Signal not due
to kinematic reflections
24
Systematic error for ?(?c)/?(D) in visible D
region
Relative systematic errors ?dm window 1.5 MeV
instead of 2.5 MeV - 9 Fit with our
background model instead of (M(Dp)-M(D))?
- 12 z(D)gt0.1 instead of z(D)gt0.2 - 21
Exclude D1,D2 signal region by m(D?)-2.45gt50
MeV 18 Uncertainty in dE/dx 10
Re-weighting of ?c fragmentation function -
5 Re-weighting of ?(?c) distribution - 3
Total - 28 21
Total systematic error -0.450.33
25
Systematic error for ?(?c)/?(D) for full D
region
Relative systematic errors ?dm window 1.5 MeV
instead of 2.5 MeV - 10 Fit with our
background model instead of (M(Dp)-M(D))?
- 14 z(D)gt0.1 instead of z(D)gt0.2 - 8
Exclude D1,D2 signal region by m(D?)-2.45gt50
MeV 17 Selection with xobs(?c) instead of
z(D) - 15 Uncertainty in dE/dx 10
Re-weighting of ?c fragmentation function
28 Re-weighting of ?(?c) distribution - 4
Total - 26 34
Total systematic error -0.640.85
If the xobs(?c) cut is used instead of the z(D)
cut the systematic uncertainty due to
fragmentation reduces to 11
26
Acceptance corrected ?c/D yield ratio-III shat
?g system
pt(c)

-
But we observe charmed hadrons instead of
quarks Normal procedure Replace quantities of
c-quark by those of D We measure also
fragmentation variable xobs ? we can do better
27
Remarks on ?(?c?Dp)/ ?(D)(xobs)
xobs(D) very soft ! For xobs(D)gt0.5 ?(?c?Dp)
/?(D)1.080.31 For xobs(D)gt0.7 ?(?c?Dp)/?(
D)0.170.13
28
Reconstruction of shat
Bins in shat 9-40-100-1000 GeV2
?c MC
D MC
91.7 71.2 64.5
83.1 73.6 81.9
purity
purity
80.8 73.4 81.6
64.8 67.1 93.2
stability
stability
29
Systematics dE/dx
Check of dE/dx selection efficiencies for protons
using ?0 in data and SPQ MC
SPQ MC
efficiency
? Data
Agreement between data and MC 5
pt and ? distributions for protons from ?0 may be
different for those from ?c
? use
systematic error of 10
30
D Signal
Golden channel D? D0 ? s ?(K?)?s (low BR but
clean signal) M(D)-M(D0) 145.4 MeV
96-00 data 75 pb-1 DIS Q²gt1 GeV²
Mass difference technique ?MDM(K??s) - M (K?)
Good Signal/Background 3400 Ds in DIS to start
with
Non charm induced background wrong charge D
fake D0 (Kp/ K-p-)pS
31
Proton selection
Most probable dE/dx

• Particle identification
• via dE/dx
• 3-5 accuracy
• 8 MIP resolution

Use dE/dx for background suppression
32
Opposite sign Dp mass distribution
Apply mass difference technique M(Dp)m(K??
p)-m(K??)MPDG(D)
Background well described by D MC and wrong
charge D from data
narrow resonance at M3099? 3(stat.) ? 5 (syst.)
MeV

• signal visible in different data taking periods

33
Signal in both D-p and in Dp
M(Dp) m(Kppp)-m(Kpp)m(D)
M(Dp)3.102 0.003 GeV
M(Dp)3.096 0.006 GeV
Signal of similar strength observed for both
charge combinations at compatible M(Dp)
34
Signal in like sign Dp combinations?
No significant peak in like sign Dp
Reasonably described by D MC and wrong charge D
from data
35
Does the resonance come from Ds?
Look at the correlation of ?M(D) vs. M(Dp)
36
Is the D-p1) signal due to protons?
Use proton in this region
Well identified protons
1) Charge conjugate always implied
37
Physics changes on-resonance ?

• ? Single particle momentum spectra
• are steeply falling
• This feature is preserved in the
• combinatorial background of
• invariant mass analyses

Harder spectrum for particles from decay due to
mass release Harder spectrum for particles from
decay of charmed hadrons due to hard charm
fragmentation
38
Physics changes on-resonance ?
Fit slope with ??exp -?p(p)
Look at momentum distribution of proton
candidates w/o dE/dx
No dE/dx cuts !
39
Physics changes on-resonance ?
No dE/dx cuts !
40
Kinematic tests
Monte Carlo expectation
wrong mass assignment
CPQ MC
Mass M independent of decay angle ? only for
correct mass assignment
correct mass assignment
M(Dp) GeV
CPQ MC
CPQ MC
M(Dp) GeV
Band like structure visible in the M(Dp)-M(Dx)
plane in data?
M(Dp) GeV
M(Dp) GeV
41
Kinematic test Dp vs. Dp
p-mass hypothesis excluded from the shape and
range of Dp mass distribution !
42
Significance estimate
signalbackground fit mass 3099 3(stat)
5(syst.) MeV width 12 3 MeV (cons. with
exp. Resolution) Numbers of signal and bgr
Nb45.0 2.8 (within 2s
24MeV) Ns50.6 11.2 (1.46 0.32
of D yield, uncorrected in acceptance)
For significance estimate Fit background only
hypothesis Nb51.7 - 2.7 Events in signal
region 95
Background fluctuation probability (52 ? 95)
4 x 10-8 (Poisson) 5.4 s (Gauss)
43
Search for charmed PQ, qc? Dp, in ZEUS
1995-2000 data, 127 pb-1 Selection of D, p close
to H1 cuts
DIS (Q2gt1 GeV) 5920?90 Ds ?p (Q2lt1
GeV)11670?140 Ds
No signal seen in Dp
Limits on ?c/D for DIS R(?c?Dp/D)lt0.51 _at_95
C.L. Including some systematic uncertainties
But selection different from H1 Production
mechanism of ?c same as for D
44
Details of fit
Charges MMeV sMeV
NS D-p Dp 3099 3 12 3 50.6
11.2 D-p 3102 3 9 3
25.8 7.1 Dp 3096 6 13 6
23.4 8.6
45
D signal in DIS and photoproduction

• DIS cleaner signal
• photoproduction supporting evidence

46
Acceptance effects?
Pion survival probability
Proton efficiency
Smooth variation with M(Dp) Shape reflects
opening of phase space
M(Dp) GeV
M(Dp) m(Kppp)-m(Kpp)MPDG(D)
47
Reflections from decays to Dp ?
D10, D20 ? Dp
loose D cuts p selection

D cuts of Dp p selection
D cuts of Dp proton selection
Expect 3.5 decays (D10, D20 ? Dp) in Dp signal
48
Could signal be due to decay D0 ? D0 g ?

• D0 ? D0 g? D0 ee-
• electrons from g-conversion
• asymmetric in energy
• misidentified as proton and ps ?

V0 with one common track with Dp
mee
No accumulation at small mee in Dp signal region
or elsewhere
49
Lots of further kinematic test

• Reflections from a possible signal in DK mass
  distribution ruled out
• Possible contributions from D ?D ? with
  ?-conversion ruled out
• Possible contributions from D /D ?D K
  ruled out
• Possible peak structures in all possible mass
  correlations with all possible mass hypotheses of
  the particles making the D and the Dp system to
  search for real or fake resonances, e.g ?, ? , ?
  , K ,?, f no enhancements found
• Possible peak structures in all possible mass
  correlations among the proton candidate the
  remaining charged particles of the event with all
  possible mass assignments to search for real or
  fake peaks, no enhancements found

0
0
0
S1 S2
0
0 S
2
50
Investigation of Dp and associated K0s
1.Selection of D DIS-events (dmlt170MeV,rw
charge) with V0 candidates
M(??)
Dp signal region
No indication for a K0 signal
51
Investigation of Dp and associated ?0s
Dp signal region
Wrong charge
For M(Dp) ? 3100 MeV M(?slowp) close to the ?0
mass due to kinematics. Was studied for
publication using primary tracks Conclusion No
problem Check with tracks from secondary vertices
D MC
CPQ MC
52
Investigation of Dp and associated ?0s
No cut in M(K-??p) applied
p-selection as for Dp paper
No ?0 signal left in dm-window !
53
Investigation of Dp and associated ?0s
p-selection as for Dp paper
Selection m(?p)-m(?0)lt9 MeV
All possible combinations
M(D0?)-M(D0) lt170 MeV
?(?0) ?slow(D)
M(D0?)-M(D0)-.1454 lt2.5 MeV
Signal in M(Dp) NOT faked by ?0s !
54
Hera kinematics in ep collisions
vs 300-318 GeV (energy c.m.) DIS
kinematics Photon virtuality Q2-q2 Electron
inelasticity y Scaling variable x Hadronic
mass W
Kinematic regimes
Scattered e detected Q² gt 1 GeV
Electroproduction (DIS) Scattered e not
detected Q² 0 GeV Photoproduction
55
Systematics variation of ?c fragmentation
function
Re-weighting of fragmentation function
Standard RAPGAP
Most sensitive to xobs z(D)
xobs(?c)
?2/NDF1.5
?2/NDF2.8
?2/NDF3.9
56
(No Transcript)
57
(No Transcript)
58
Cut in (Dp) fragmentation variable
New selection for Dp yield estimate Xobs(Qc) gt
min (0.5 P (proton), 0.5) (use for acceptance
corrected yields of Qc vs Zd , Xobs)
Cut xobsgt min(0.5p(p),0.5)
CPQ MC
Wrong charge D
? wrong charge D
? CPQ MC
59
Remarks on Dp search by ALEPH
xobs(D) very soft ! For xobs(D)gt0.5 ?(?c?Dp)
/?(D)1.080.31 For xobs(D)gt0.7 ?(?c?Dp)/?(
D)0.170.13
CPQ MC
DMC
Expected shift in peak position -0.3