Stato della preparazione dellanalisi D Kp p

1
Stato della preparazione dellanalisi D ? K-pp

• Elena Bruna, Massimo Masera, Francesco Prino
• Università di Torino e INFN

Secondo convegno nazionale sulla fisica di ALICE,
30 Maggio 1 Giugno 06, Vietri sul Mare
2
Outline

• Physics motivations for D analysis
• Selection of the signal
• Single track cuts
• Secondary vertex finder
• How to combine the triplets
• Cuts on the triplets
• First preliminary results of the D feasibility
  study on a sample of events
• Perfect PID (given by the simulation)
• Experimental PID (combined)
• NO PID
• Time-consumption problems, first ideas to avoid
  these problems ? new analysis strategy

3
Why also D (and Ds)?
Accurate measurement of the charm production
cross section

• D0/D ratio puzzle
• Expected to be 3.08 from spin degeneracy of D and
  D and decay BR
• Measured by ALEPH_at_LEP to be 2.32
• Different selection strategies due to
• D has a longer mean life (ct 311mm compared
  to 123 mm of the D0)
• D fully reconstructable from a 3-charged body
  decay instead of the 2 (or 4) body decay of D0

4
D ? K-pp vs D0 ? K-p
Advantages

• D has a long mean life (ct 311mm compared to
  123 mm of the D0)
• D ? K-pp has a relatively large branching
  ratio (BR9.2 compared to 3.8 for D0 ? K-p).

drawbacks

• Combinatorial background for this 3-body channel
  is larger than for D0 ? K-p.
• The average PT of the decay product is softer (
  0.7 GeV/c compared to 1 GeV/c for the D0)

5
D statistics

• Nevents for 2107 MB triggers
• Ncc number of c-cbar pairs
• Includes shadowing EKS98
• Shadowing centrality dependence from Emelyakov et
  al., PRC 61, 044904
• D yield calculated from Ncc
• Fraction ND/Ncc (0.38) from tab. 6.7 in chapt.
  6.5 of PPR
• Geometrical acceptance and reconstruction
  efficiency
• Extracted from 1 event with 20000 D in full
  phase space
• B. R. D? Kpp 9.2

6
Simulation and analysis strategy
Too large statistics (108 events) would be
required to study the signal!!
Central Pb-Pb event (blt3.5 fm, dN/dy 6000,
vs5.5 TeV)
9 D/D- in ylt1
Simulation
Signal and background events separately generated
with the Italian GRID

• PT of the decay tracks is soft 0.7-0.8 GeV/c
• The magnetic field is low - 0.5 T - to allow the
  reconstruction of soft particles ( 7000 in
  ylt1) ? huge combinatorial background
• A dedicated trigger for D
  ? K-pp seems not possible.

Analysis

• Good secondary vertex reconstruction capability
  (c? (D) 300mm ? resolution of 200mm would be
  bad, 50mm would be a dream)
• Efficient system of cuts to discriminate the
  signal from the background

7
1st step Single track cuts
Cuts on PTp, PTK, track impact parameter (d0) on
all the tracks for both signal and background
events
Perfect PID is assumed
Choices of cuts which maximise the Significance
Not optimized cut we want to preserve D down to
PT1 GeV/c
8
2nd step Combining K-p pairs
IDEA start from BKG pairs of Kp tracks (once the
single track cut have been performed) and cut on
the distance between the 2-track vertex and the
primary one
p1
p2
Ki

pj

K
p
Kj
both Kp pairs are required to pass the cut
9
3rd step Combining the Triplets

• Single track cuts applied
• Cut on the 2-tr vertex applied

Kpp are combined together according to the sign
of their impact parameter
Background
Signal
d0K x d0p2
d0K x d0p2
Cut on d0
empty
d0K x d0p1
d0K x d0p1
When (d0K x d0p1)lt0 (d0K x d0p2)lt0. Due to the
kinematics of the decay
10
Secondary Vertex Finder on the Triplets
Cut on the quality of the Vertex. Sigma
BLACK signal RED BKG Kpp Triplets
P1 (x1,y1,z1)
track
SecondaryVertex (x0,y0,z0)
d1
ZOOM
BLACK signal RED BKG Kpp Triplets
BLACK signal RED BKG Kpp Triplets
11
Cuts on the Triplets

• Quality of the vertex Sigma (prev. Slide)
• Distance between primary and secondary vertices

BLACK signal RED BKG Kpp Triplets
cos?point

• Cut on the cosine of the pointing angle defined
  by the PT of the D and the line connecting
  primary and secondary vertices

The signal is peaked at 1
12
First strategy adopted for the analysis

• For each variable (in the order Sigma, distance,
  cos?point)
• loop on all the triplets (both signal and
  background)
• Keep the value of the cut with Max Significance
  S/v(SB) in the range MINVKpp-MDlt1s
• Use this value to maximize the Significance for
  the next cut variable

• Different ranges of the reconstructed pT of the
  triplet
• Perfect PID, PID and without PID

• Statistics used in this preliminary analysis
• BKG 500 HIJING events
• SIG 8.5 X 105 reconstructed signal triplets

13
Perfect PID pT integrated
Distance
Sigma
Significance
Significance
Chosen cuts Sigma0.018 cm Dist1900
mm cosqpoint0.995 Signif 4411 S/ev
0.001 B/ev 0.004
cos?point
The Significance is normalized to 107 central
events
Significance
14
SIGNAL BKG at different cut levels
Number of SIGNAL triplets per event (full inv
mass range)
Number of BACKGROUND triplets per event (full inv
mass range)
NO cuts
Dist 2tr-vert
sigma
Dist 2tr-vert
cos?point
sigma
NO cuts
cos?point
Single track cuts
Distance prim-sec
Single track cuts
Distance prim-sec
15
PID

• Combined Bayesian PID (ITSTPCTOFTRDHMPID) is
  used
• Prior probabilities used in input
• P(p)0.055
• P(K)0.072
• P(p)0.864
• P(e)0.006
• P(m)0.003
• PID in the ITS done with
• Clusters from all the 4 layers (2 SDD2 SSD)
• Convoluted Landau-Gaussian fits to the response
  functions in each layer
• See AliITSpidESD2 implemented in AliRoot
• Track tagged as type i when the corresponding
  combined Bayesian probability is P(itrack)gt0.85

16
PID pT integrated
Distance
Sigma
Significance
Significance
Chosen cuts Sigma0.019 cm Dist2000
mm cosqpoint0.995 Signif 4015 S/ev
0.0007 B/ev 0.002
cos?point
Significance
17
Without PID pT integrated

• No selection of tracks based on the particle
  identity

Sigma
Significance
Significance
Distance
Chosen cuts Sigma0.019 cm Dist1800
mm cosqpoint0.999 Signif 3912 S/ev 8 X
10-4 B/ev 0.004
cos?point
Significance
18
Preliminary global results

• Low pT under study
• Rebinning
• Additional cuts

Significance
Analysis feasible also without PID, but more time
consuming
19
D elliptic flow measurement perspectives
2107 Minimum Bias events

• Error bars quite large
• Would be larger in a scenario with worse event
  plane resolution (lower dNch/dy or v2)
• May prevent to draw conclusions in case of small
  anisotropy of D mesons
• v2 vs pT requires a semi-peripheral trigger

20
Conclusions

• The reconstruction of D ? K-pp is a promising
  study.
• The preliminary results show that the analysis is
  feasible with a pretty good Significance.
• More statistics is mandatory for a more accurate
  optimization of the cuts.
• There still is room for optimization work in
  progress.

21
Outlook

• New cut strategy
• For each event fill 2 multi-dimesional matrices
  (one for Signal and one for BKG), each cell
  containing the number of triplets corresponding
  to all the possible combinations of cut
  variables. Es. with 5 cut variables and 30 steps
  for each variable ? 305 cells (200MB per ev for
  SignalBKG)
• Sum all the multi-dimesional matrices (Signal and
  BKG) on all the events
• Maximize the multi-dim matrix of the Significance
• Apply LDA
• pp studies to come on the PDC06 events and on
  italian production with parametrized TPC

22
Backup slides
23

Hadronic 3-charge-body decays of D
D?K-?? BR 9.2
24
Kinematics (1)
K
PT distributions of the generated particles (ONLY
PYTHIA generation, NO propagation and
reconstruction in the detector) (nonresonant
events)
Mean 0.87 GeV/c
D
Mean 1.66 GeV/c
?
Mean 0.67 GeV/c
Knowledge of the PT shapes of the decay products
important at the level of the selection strategy
25
Dalitz Plots Kinematics (2)
Non resonant
Resonant
26
Combining K-p pairs /2

• Best Significance
• cut distance of the 2-tr vertex 700 mm
• 63 Signal triplets pass the cut
• S/ev0.006
• 105 remaining BKG triplets per event.

Pt reconstructed D Mean2.66 GeV/c
S/B 6X10-8 still low
Full invariant mass range
Low PT D still kept
27
Display D decay (made with the kinematics)
Primary vertex

• Impact parameter
• Definition segment of minimum distance of the
  (prolonged) track from the primary vertex
• Sign
• primary vertex in the track circle
• - primary vertex out of the track circle

K
p1-
p2-
Secondary vertex
K d0gt0 p1- d0gt0 p2- d0gt0
Points on the 3 prolonged tracks defining the
impact parameter d0
28
Measurement of v2

• Calculate the 2nd order coefficient of Fourier
  expansion of particle azimuthal distribution
  relative to the reaction plane
• The reaction plane is unknown.
• Estimate the reaction plane from particle
  azimuthal anisotropy
• Yn Event plane estimator of the unknown
  reaction plane
• Calculate particle distribution relative to the
  event plane
• Correct for event plane resolution
• Resolution contains the unknown YRP
• Can be extracted from sub-events

29
Worse resolution scenario

• Low multiplicity and low v2

Large contribution to error bar on v2 from event
plane resolution
30
Semi-peripheral trigger

• v2 vs. pT that would be obtained from 2107
  semi-peripheral events ( 6ltblt9 )