Measurement%20of%20charm%20and%20bottom%20production%20in%20pp%20collisions%20at%20vs%20=%20200%20GeV%20at%20RHIC-PHENIX

1
Measurement of charm and bottom production in pp
collisions at vs 200 GeV at RHIC-PHENIX

• Yuhei Morino for the PHENIX collaboration
• CNS, University of Tokyo
• JSPS

2
1.Introduction
Phys. Rev. Lett. 98, 172301 (2007)
Behavior of heavy quarks in hotdense matter

• Observation via lepton measurement
• Large energy loss
• Large V2
• strongly interacting matter for charm!

_at_hotdense matter

• next open question
• bottom flow?
• bottom energy loss?

charm and bottom are not separated. Need to
separate charm/bottom to get more information.
b contribution?
3
b contribution to non-photonic electron
Phys.Rev.Lett 95 122001

• FONLL Fixed Order plus Next to Leading Log pQCD
  calculation
• Large uncertainty on c/b crossing 3 to 9 GeV/c

Measurement of b?e/c?e is key issue. This talk
will show the latest result of measurement of
c,b in pp collisions at mid-rapidity AuAu
results were reported at D.Hornbacks talk
4
Heavy quark measurement at PHENIX

• lepton from semileptonic decay
• large branching ratio
• c and b mixture

K
p-
direct measurement

• direct ID (invariant mass)
• large combinatorial background

5
2 Measurement of non-photonic electron
measurement
Inclusive electron ( g conversion, p Dalitz,etc
and heavy quark )
e-

• Background subtraction
• cocktail method
• converter method

Semileptonic decay
Non-photonic electron (charm and bottom)
.
S/Ngt1 _at_ptgt2GeV/c
b?e/(c?eb?e)?
PRL, 97, 252002 (2006)
6
c,b separation in non-photonic electron
D0?e K-(NO PID) reconstruction
Ntag Nunlike - N like

• background subtraction?(unlike-like)
• photonic component
• jet component

tagging efficiency? when trigger electron is
detected, conditional probability of associate
hadron detectionin PHENIX acc
From data
From simulation (PYTHIA and EvtGen)

decay component (85)?kinematics
e
jet component (15)
7
reconstruction signal and simulation
8
bottom fraction in non-photonic electron

• first result of b fraction measurement at PHENIX
• The result is consistent with FONLL

9
electron spectra from charm and bottom
b?e (non-photonic) X (b?e/(c?eb?e))
PRL, 97, 252002 (2006)
charm
bottom
sdata/sFONLL 2 reasonable value
10
cross section of bottom
pt extrapolation of b?e spectra by pQCD
First measurement of bottom cross section at
mid-rapidity in pp collisions at
PHENIX. rapidity extrapolation by NLO pQCD
(ylt0.35?y integrated) sbb(data)/sbb(FONLL)2
11
3 Heavy quark measurement via di-electron
ee- pair
A.Toiatalk
arXiv0802.0050
heavy quark is dominant source _at_mee gt1.1GeV
12
Di-electron from heavy quark
cocktail calculations are subtracted from data

• bottom, DY,subtraction
• ? charm signal !!
• mass extrapolation (pQCD)
• rapidity extrapolation (pQCD)

c dominant
b dominant
After Drell-Yan subtracted, fit
(acharmbbottom) to the data.
charm and bottom cross sections from ee- and
c,b?e agree!
13
total cross section of charm and bottom
vs dependence of cross section with NLO
pQCD agrees with data
14
4 Direct measurement of D meson
15
D0?K-pp0 reconstruction
S.Butsykposter
large branching ratio(14.1)
Clear peak of D0 (5ltptlt15GeV/c) meson observed
in D0?K- p p0 decay channel
16
D0?K-p reconstruction with electron tag
electron tag reduce combinatorial background
P.Shukla poster

• observe D0 peak
• cross section of D is coming up

17
5 Summary and Outlook

• b?e/(c?e b?e) has been studied in pp
  collisions at vs 200GeV via e-h correlation.
  Cross section of bottom was obtained from
  electron spectra and b?e ratio.
• Consistent with FONLL calculation
  (data/fonll 2)
• This is baseline measurement for
  understanding heavy quark energy loss and
  v2 observed in AuAu collisions and further
  discussion on heavy quark energy loss will be
  done.
• Cross sections of charm and bottom were obtained
  from di-electron in pp collisions at vs 200GeV.
• Clear peak of D0 meson observed in pp collisions
  at vs 200GeV in D0-gtK p- p0 and D0-gtK p-
  channels.
• Analysis to determine cross section is on
  going.
• Silicon Vertex Tracker will be installed for more
  precise study.

18
back up
19
Electron Signal and Background
Photonic electron Background

• Conversion of photons in material
• Main photon source p0??? ? gg
• In material g ? ee- (Major contribution of
  photonic electron)
• Dalitz decay of light neutral mesons
• p0??? ? g ee- (Large contribution of photonic)
• The other Dalitz decays are small contributions
• Direct Photon (is estimated as very small
  contribution)
• Heavy flavor electrons (the most of all
  non-photonic)
• Weak Kaon decays
• Ke3 K ? p0 e ?e (lt 3 of non-photonic in pT gt
  1.0 GeV/c)
• Vector Meson Decays
• w, ?, f??J?? ? ee- (lt 2-3 of non-photonic in
  all pT.)

Non-photonic electron Signal and minor
background
20
Background Subtraction Cocktail Method

• Most sources of background
• have been measured in PHENIX
• Decay kinematics and
• photon conversions can be reconstructed by
  detector simulation
• Then, subtract cocktail of all background
  electrons from the inclusive spectrum
• Advantage is small statistical error.

21
Background Subtraction Converter Method
We know precise radiation length (X0) of each
detector material The photonic electron yield can
be measured by increase of additional material
(photon converter was installed) Advantage is
small systematic error in low pT
region Background in non-photonic is subtracted
by cocktail method
Photon Converter (Brass 1.7 X0)
22
Consistency Check of Two Methods
Accepted by PRL (hep-ex/0609010)
Both methods were always checked each other Ex.
Run-5 pp in left
Left top figure shows Converter/Cocktail ratio of
photonic electrons Left bottom figure shows
non-photon/photonic ratio
Accepted by PRL (hep-ex/0609010)
23

• 2.25pb-1 of triggered pp data as reference
• Material conversion pairs removed by analysis cut
• Combinatorial background removed by mixed events
• additional correlated background
• cross pairs from decays with four electrons in
  the final state
• particles in same jet (low mass)
• or back-to-back jet (high mass)
• well understood from MC

24
Method I

• Tune cocktail to PHENIX measured hadrons
• Subtract cocktail
• Extract cross section in multi steps as in ppg065
• A. dsigma_ee/dy(1.1ltMeelt2.5 in ideal PHENIX
  acceptance) This is what directly measured. Only
  systematic error in the data and Statistical data
  present.
• A1. Extrapolate to 0ltMlt5 GeV However, since
  ds/dy(1.1ltMlt2.5) is a very tiny fraction of
  dsigma/dy(0ltM), I would rather not mention about
  it.
• B. dsimga_ee/dy(1.1ltMeelt2.5 yelt0.35)This is
  when two arm acceptance of PHENIX is corrected.
  Since the two arm nature is corrected, this is
  something a theorist can easily calculate.(now
  acceptance error is involved)
• C. dsimga/dy of ccbar (now PYTHIA error is
  involved kt, pdfs and branching ratio because
  we go from electrons to charm)
• D. sigma(ccbar) total (now add error for rapidity
  distribution)
• In the paper we mention only A., C. and D. for
  simplicity
• A. is calculated from the data, C. and D. are
  derived in the procedure explained in the next
  slide

25
Method II

• Tune cocktail to PHENIX measured hadrons
• Subtract cocktail
• Fit p0charm p1bottom drell yan
• Charm cross section 567 mb (ppg065)
• Beauty cross section 3.77 mb (Claus Jaroceck
  and commonly used in single electron analysis)
• Drell Yan 0.040 mb and scaled to NLO
  calculations from Werner Vogelsang
• DY (from PYTHIA) p0charm p1bottom
•     p0           9.13960e-01 8.24258e-02
• p1           1.06418e00 7.13970e-01
• DY (scaling Pythia to Werners calculations for
  Mgt4GeV) p0charm p1bottomQ/2
•     p0           9.08741e-01 8.25467e-02
•     p1           1.14892e00 7.17499e-01
• Q
•     p0           8.97103e-01 8.25275e-02
•     p1           1.24826e00 7.17928e-01
• Q2
•     p0           9.09590e-01 8.25467e-02
•     p1           1.13538e00 7.17499e-01

26
c?e, b?e spectra
of non-photnic electron in b/(bc) ? PPG65
spectra sys error of of non-photnic electron
?100correlation? sys error of PPG65 ?enlarge sys
error of bottom
non-photonic electron (totalgtb)
90 C.L
27
Acceptance filter and symmetrical fiducial cut
Symmetrical fiducial cut
Fiducial cut is also applied for
phi. (symmetrical fiducial cut) This cut will
make phase space symmetrical.
28
4. Analysis(RUN5)
From simulation (PYTHIA and EvtGen)
charm ec 0.0364 - 0.0034(sys) bottom eb
0.0145 - 0.0014(sys)
Electron pt 25GeV/c Hadron pt 0.45.0GeV/c
unlike pair like pair
(unlike-like) / of ele
29
Electron-hadron invariant mass(RUN5)
Mass of hadron is assigned 494MeV, hadron 0.4 lt
pt lt5 GeV/c
Electron pt 23 GeV/c
Electron pt 34 GeV/c
Unlike pair Like pair
Electron pt 45 GeV/c
Electron pt 25 GeV/c
30
Estimation of systematic error for signal
counting
Electron pt 25 GeV/c
real unlike / real like mixing unlike/ mixing like
Mixing unlike pair Mixing like pair
mixing unlike / mixing like 1, there are no
effect of phase space
RMS is 2. I will assign this 2 as
systematic err about signal counting.
mixing unlike/ mixing like
31
unlike/like ratios Vs invariant mass
Real Mixing event
Electron pt 23 GeV/c
Electron pt 34 GeV/c
Electron pt 45 GeV/c
Electron pt 25 GeV/c
32
Electron-hadron (unlike like) invariant
mass(RUN5)
0.5 lt invariant mass lt1.9 GeV pairs are counted
as signals.
Electron pt 23 GeV/c
Electron pt 34 GeV/c
Electron pt 45 GeV/c
Electron pt 25 GeV/c
33
Remaining electron electron pair rejection(RUN5)
Electron veto cut for hadron (n0lt0) cannot all
electron due to RICH acceptance. pair mass (mass
of hadron is assigned 0.511MeV)gt0.08GeV cut was
used for e-e pairs rejection. But there are
remaing electron pairs
Mass of associate particle is assigned 494MeV
Unlike pair Like pair
e h pair Estimated remaining e-e pair
remaing e-e pairs are estimated by invariant mass
distribution when mass of associated electron is
0.494Mev. Normalization factor is ( of e-h pair
in invariant mass lt0.08) /
( of e-e pair in invariant mass
lt0.08)
Unlike pair Like pair
Estimated e-e pairs are subtracted. Systematic
error of this subtraction is estimated by
statistics of ( of e-h pair in invariant mass
lt0.08) / ( of e-e pair in invariant mass
lt0.08)
34
spectra of FONLLPTYHIA(1.5ltktlt10GeV/c) 4.21 -
0.4
PYTHIA EvtGen combination0.88
PDG value changing B hadron ratio10-1
PYTHIA HVQMNR(NLO QCD) 3.44-0.25
35
Electron-hadron (unlike like) invariant
mass(RUN5) after remaining e-e pair rejection
Electron pt 23 GeV/c
Electron pt 34 GeV/c
Electron pt 45 GeV/c
Electron pt 25 GeV/c
36
High Pt extension
eID tight cut hadron background was estimated
from e/p distribution
the effect of h-h correlation hadron iD
standard eID cut problt0.01
0.6lte/plt0.8 ?99 hadron
ehadron 0.087- 0.043 (50 systematic error)
37
EvtGen
38
EvtGen and PYTHIA products
charm (bottom)
gluon
string
string
D
pi0
D0

gamma
e
K-
nu
gamma
EvtGen only
EvtGenPYTHIA
Fast monte carlo calculation for EvtGen ony and
EvtGenPYTHIA electron pt 25GeV/c
EvtGenPYTHIA ec 0.0342 EvtGen only ec 0.0301
15 of ec are from PYTHIA. This part may be
changed by PYTHIA fragmentation,etc I assign
PYTHIA 20 systematic error. This error
corresponds 3 error for ec
39
photonic electron unlike/like
40
Hadron contamination (ptgt5 GeV/c)
tight eid cut (normal n1gt4 probgt0.1) estimate
d hadron (at previous page)
electron peak is clearly seen at ptlt9GeV/c
3
5
11
fit estimated hadron distribution (back
ground) ?Fix ?Fit e/p distribution at tight eid
cut
12
17
hadron contamination (e/pgt0.9) was estimated
these fit functions.