Status of study for beauty detection with ALICE

1
Status of study for beauty detection with
ALICE CERN, 7 / 7 / 2004
2
Contents - review on the inclusive
semi-electronic channel for beauty total cross
section measurement - how can we measure the
beauty Pt distribution? - fast simulation
details - first evaluation of B ? e D0 ( ?
Kp) X - conclusions
People Federico Antinori, Andrea Dainese,
Marcello Lunardon, Sandra Moretto, Rosario Turrisi
3
Beauty production beseline
X-section from NLO calculations ?qq b N
qq in Pb-Pb 5.5 TeV (5 ?tot) c-cbar 45 115
b-bar 1.8 4.6
high uncertainty 1.8-7.3
in p-p 14 TeV c-cbar 1.5 0.16 b-
bbar 0.06 0.007
Semi-electronic channel 10
Statistics 107 central Pb-Pb events 9 M
beauty electrons 109 p-p events 1.5 M
beauty electrons
4
Semi-electronic beauty detection strategy 1)
electron identification in TRDTPC 2) cuts on
high Pt and high track impact parameter d0
Pt distribution
d0 distribution
5
Electron tagging
TRD 90 electron efficiency 1 misidentified
pions (constant in 1-6 GeV/c Pt
range)
pion contamination 1
de/dx PID in TPC brings low momentum pion
contamination to less than 10-4
6
Beauty semi-electronic simulation details
Signal Pythia6 with MSEL5, CTEQ4L, forced
semi-electronic decay Background 20000 HIJING
cent2 (blt 2fm) events tracking - AliRoot
version 3.07.04 - TPC parameterization - Kalman
Filter in ITS (request for SPD layers) - Pt
independent efficiency 75 for TPC-TRD
matching PID - assuming complete rejection of
K,p and heavier particles - 80 electron
reduction factor for efficiencies of TRD (0.9)
and TPC (0.9) - pion contamination of 1 TRD
(constant with P) times eff(P) TPC (see
previous slide)
7
Beauty semi-electronic simulation results
8
Beauty semi-electronic Pt(e) Pt(b) correlation
Pt(b) - Pt(e) correlation poor
9
Beauty semi-electronic conclusions

• - good expected statistics for total cross
  section measurement (cross section error estimate
  in progress by R. Turrisi), but
• difficult to extract a good Pt differential
  quark cross section information, then
• we want to try to associate more tracks to the
  electron
• some examples
• - B ? e D0 ( ? Kp) X
• - secondary vertexes analysis (N tracks vertex
  with electron and selection on electron PtRel)

10
Example B ? e D0 ( ? Kp) X
Pt(b)-Pt(rec) simple correlation function
11
Simulation strategies

• 1st step explore possibilities
• very fast simulation no detector
• event generator
• pt-dependent smearing of momenta and d0 (rf and
  z)
• pt-dependent tracking and PID efficiencies
• fast secondary vertexing (straight-line
  approximation)
• the tool is being prepared in Padova and will
  join AliRoot
• 2nd step assess performance
• wise fast simulation ITS simulation and
  reconstruction, TPC,TRD,TOF parameterized (as
  done for D0 study)
• TPC parameterization now outdated
• we are evaluating the possibility to use PDC04
  Pb-Pb and pp events to update it, including TRD
  (tracking PID) and TOF (PID)

12
Fast simulation details
- Basic tracks information from generator put
in the FastTrack structure - Main detector
effects included with gaussian smearing according
with available resolution studies - Pt
(Pt) - Tg? (Pt) - ? (Pt) - d0
(Pt,PID) - z0 (Pt,PID) - ITSTPC tracking
efficiency (Pt,PID) - TRD behavior as for
inclusive semi-electronic simulation
13
Fast simulation details
typedef struct Int_t evnum Int_t
lab Int_t pdg Int_t pdgmum Int_t
proc Int_t charge Int_t flag
Float_t d0 Float_t px Float_t py Float_t
pz Float_t vx Float_t vy Float_t vz
Float_t v1x Float_t v1y Float_t v1z
Float_t tmp FastTrack
generator information
d0 calc. from smeared vertex
smeared momentum
smeared vertex
14
Fast simulation details
Pt resolution
15
Fast simulation details
Impact parameter resolution
Tracking efficiency
16
First test of B ? e D0 (? Kp) X
Expected statistics for 107 min. bias Pb-Pb 5.5
TeV events 8107 B mesons 10
semi-electronic ? 8106 65 with a D0 ?
5106 3.8 with D0 ? Kp ? 2105 ? total B.R
0.22
Example preselection on electron tagging with
Pt(e) gt 1 GeV/c and d0(e) gt 150 ?m ? 1.8
105 beauty SE electrons ? 4400 accK? ?det
K? ?selcut ? a few hundreds of detectable decays
17
First test of B ? e D0 (? Kp) X

• Signal
• Pythia6 tuned MNR, forced SemiElectronic
  D0?K?
• Background
• 50000 HIJING cent2 events
• Code written to be included in AliRoot
• FastTrack ? AliD0toKpiAnalysis ? selection of
  AliD0toKpi candidates with classical cuts (D0
  mass, DCA, CosPointingAngle, d0Kd0?
• FastTrack AliD0toKpi ? AliBtoD0eAnalysis ?
  selection of AliBtoD0e candidates
• Selection on B mass (D0 e)
• B DCA (D0 flight line with el. track)
• dist(V0,VB)
• dist(V0,VD)
• CosVBVD0
• electron PtRel

18
First test of B ? e D0 (? Kp) X

• Minimum selection
• - Pt(e,K,?) gt 0.5 GeV/c
• d0(e,K,?) gt 50 ?m
• minimal D0 selection cuts
• Initial S/N lt 10-6 (out of laptop memory)

19
First test of B ? e D0 (? Kp) X
20
First test of B ? e D0 (? Kp) X
Some trial Minimum selection - Pt(e,K,?) gt
0.5 GeV/c - d0(e,K,?) gt 50 ?m ? equivalent of
700 B candidates A little more realistic
selection - Pt(e) gt 1 GeV/c, Pt(K,?) gt 0.5
GeV/c - d0(e) gt 150 ?m, d0(K,?) gt 50 ?m -
narrow cuts on D0 and B variables ? expected
about 180k electrons from beauty over a
total of about 300k selected events ? cut
efficiency on signal 14 ? 90 candidates ?
cut efficiency on bck 0.23 ? 700 false
candidates S/B 13
21
What next

• 1) More detailed analysis of the B ? e D0 (?
  K?) X kinematical variables
• 2) Low Pt electron recovery using a better PID
  analysis in ITSTPC
• 3) Relaxation of electron tagging just a tracked
  particle with high d0 and Pt
• (e,?,?) ? more signal, more background
• 4) Evaluation of B ? e D0 (? K?) X in p-p in
  case of an electron trigger in TRD
• 5) Secondary vertexes analysis (N tracks vertex
  with electron and electron
• PtRel) preliminary estimate with 3 track s
  ? low statistics
• 2 tracks ? to be studied
• 6) Second stage analysis for best channel