DNP/JPS 1

The azimuthal anisotropy of electrons from heavy flavor decays in √SNN = 200 GeV Au-Au collisions by PHENIX

Shingo Sakai for the PHENIX collaborations(Univ. of Tsukuba)

DNP/JPS 2

Why heavy flavor v2 ?

(1)π,K,K0s,p,Λ, Ξ,d ( light quark->u,d,s )

v2 scales approximately with

the number of valence quark of hadrons

-> indicate partonic level flow

if charm quark flow - partonic level thermalization - high density @ the early stage of collision

(2) Heavy flavor energy loss - v2@high pT reflects energy loss

- if heavy flavor energy loss is small, smaller v2 @ high pT might be expected compared with light quarks v2

M. Djordjevic, M. Gyulassy, R. Vogt, S. Wicks. nucl-th/0507019

DNP/JPS 3

Heavy flavor study @ PHENIX

Elecron sources

charm decay beauty decay

Dalitz decays Di-electron decays Photon conversions Kaon decays Thermal dileptons

Subtract photonic electrons following methods

“Cocktail subtraction” – calculation of “photonic” electron background from all known sources “Converter subtraction”– extraction of “photonic” electron background by special run with additional converter (brass, X = 1.7%)

photonic

non-photonic

DNP/JPS 4

Electron v2 measurement @ PHENIX

e-

Electron v2 is measured by R.P. method

R.P. --- determined with BBCTracking (pT,φ) 　 --- DC + PCelectron ID --- RICH & EMCal

dN/d(-) = N (1 + 2v2obscos(2(-)))

(E-p/p/sigma) distribution

eID @ RICH

B.G.

After subtract B.G.

Fig : Energy (EMcal) & momentum matching

of electrons identified by RICH.

Clear electron signals around E-p/p = 0

DNP/JPS 5

Non-photonic electron v2 measurementconverter methodSeparate non-photonic & photonic e v2 by usingNon-converter run & converter run

Non-converter ; Nnc = Nγ+Nnon-γ => (1+RNP)v2nc = v2γ + RNPv2non-γ Converter ; Nnc = R *Nγ+Nnon-γ => (R +RNP) v2c = R v2γ + RNPv2non-γ

* R -- ratio of electrons with & without converter 　 v2nc --- inclusive e v2 measured with non-converter run　 v2c --- inclusive e v2 measured with converter run v2γ --- photonic e v2, v2non-γ --- non photonic e v2

cocktail methodDetermined photonic electron v2 with simulation Then subtract it from electron v2 measured with non-converter run

v2non-γ = {(1+RNP)v2 - v2γ} }/RNP

Run04: X=0.4%

Run02: X=1.3%

Non-pho./pho.

(QM05 F. Kajihara)

DNP/JPS 6

Inclusive electron & photonic electron v2

inclusive e v2

photonic e v2

Non photonicsignal

Photonic b.g.

- inclusive electron v2 (pho. + non-pho.) &

photonic electron v2

pT < 1.0 GeV/c --- converter method

pT > 1.0 GeV/c --- cocktail method

- inclusive electron v2 is smaller than

photonic electron v2

- Above 1.0 GeV/c, non-photonic electron

contribution is more than 50 % !

DNP/JPS 7

Charm quark flow ?

-pT dependence of non-photonic electron v2 after subtracting photonic electron from inclusive electron v2

-Compared with quark coalescence model prediction. D -> e v2 with/without charm quark flow

(Greco, Ko, Rapp: PLB 595 (2004) 202)

*Due to light quark has finite v2, D meson has

non-zero v2 though charm quark v2 = 0

Below 2.0 GeV/c ; consistent with charm quark flow calculation

Greco, Ko, Rapp: PLB 595 (2004) 202

DNP/JPS 8

D v2 from non-photonic electron v2 (1)

access D meson v2 from non-photonic electron v2 below 2 GeV/c (due to large error & uncertainty B meson contribution above 2 GeV/c)

(1) Assume D meson v2 as; v2 D = a * fv2(pT) fv2(pT) = pi v2 fv2(pT) = kaon v2 fv2(pT) = proton v2 fv2(pT) = D v2 (y_T scailing)

(2) Calculate D -> e (PYTHIA)(3) D v2 is given as weight to get e v2 (4) chi-squared test with “measured” non-photonic electron v2 below 2 GeV/c χ2 = Σ{(v2 non-γ- v2 D->e)/σv2 }2

(3) Find chi-squared minimum of a

experiment simulation

fv2(pT) = pi v2 fv2(pT) = kaon v2

fv2(pT) = proton v2 fv2(pT) = D v2 (y_T scal.)ch

i2

chi2

chi2

chi2

a(%) a(%)

a(%) a(%)

DNP/JPS 9

D v2 from non-photonic electron v2 (2)

-D meson v2 after scaling with chi2 minimum 　 value (achi2_mim) for each D meson shape (fv2(pT))

v2 D = achi2_mim * fv2(pT)

fv2(pT) = pi v2 fv2(pT) = kaon v2 fv2(pT) = proton v2 fv2(pT) = D v2 (y_T scailing)

-D (= achi2_mim * fv2(pT)) -> e v2

compared with “measured” non-photonic electron v2

D meson v2 after scaling with

chi2 minimum value (achi2_mim) for each fv2(pT)

DNP/JPS 10

Simulation of B->e v2

D & B v2 (assume D & B v2 same)

D -> e

B -> e (B v2 flat @ high pT)

B -> e (B v2 decrease @ high pT )

- B meson v2 same as D meson v2

decrease @ high pT

flat @ high pT

- Clear difference electron v2

from B & D

-If B meson v2 is

(1) Smaller v2 than D meson

(2) decreasing @ high pT

(3) B->e overcome D->e

around 3~4 GeV/c

Non-photonic e v2 might be

reduced.

pT

v 2

DNP/JPS 11

Conclusion

Non-photonic electron v2 from heavy flavor decays has been 　　 measured with RHIC-PHENIX

Compare with model calculations assuming charm flow or not Our result consistent with charm flow model below 2.0 GeV/c

Access D meson v2 from non-photonic electron v2 assuming D meson v2 shape as pion, Kaon, proton & D meson from y_T scailing.

DNP/JPS 12

Back up slides

DNP/JPS 13

y_T scailing

0 1 2 3 4 5

v 2

0.00

0.05

0.10

0.15

0.20

0.25

0.30 s 200 GeVNNAu Au

0SKp

fsTy

5 < Centrality < 30 %

K

(STAR)

(PHENIX)

(STAR)

(PHENIX)

(PHENIX)

DNP/JPS 14

Photon Converter (Brass: 1.7% XPhoton Converter (Brass: 1.7% X00))

Yield of conversion electron can be determined by the radiation length (X0) of material amount.

We know precise X0 of each detector material, but don’t the total effective value (+ air etc.).

However, we can measure the yield of conversion electron by inserting of converter.

Then, the photonic electron is subtracted from inclusive.

Advantage is small systematic error even in low pT region.

Photonic Subtraction-Converter Method

Ne Inclusive electron yield

Material amounts: 0

1.1% 1.7%

Dalitz : 0.8% X0 equivalent

0

With converterConversion in converter

W/O converter

0.8%

Non-photonic

Conversion from known material

? %

Photonic

F. Kajihara (session CC 7)