Nuclear modification factors, directed and elliptic flow of electrons from open heavy flavor decays in Au+Au collisions from STAR
NNuclear Physics A 00 (2020) 1–4
NuclearPhysics A / locate / procedia XXVIIIth International Conference on Ultrarelativistic Nucleus-Nucleus Collisions(Quark Matter 2019)
Nuclear modification factors, directed and elliptic flow ofelectrons from open heavy flavor decays in Au + Au collisionsfrom STAR
Matthew Kelsey for the STAR Collaboration a a Lawrence Berkeley National Laboratory, Berkeley, California 94720, USA
Abstract
We present the analyses of single electrons from semileptonic bottom and charm hadron decays at mid-rapidity in √ s NN =
200 and 54.4 GeV Au + Au collisions (talk in Heavy-Flavor session III). The data at √ s NN =
200 GeV incorporateinformation from the Heavy Flavor Tracker which enables the topological separation of electrons originating frombottom and charm hadron decays. We report the first STAR measurements at √ s NN =
200 GeV of v for bottom decayelectrons as a function of p T and v for charm decay electrons as a function of electron rapidity. Additionally, wepresent the improved measurements of heavy-flavor decay electron R AA and a new measurement of the ratio of R CP between bottom and charm decay electrons. Finally, we also report the measurement of non-photonic electron v in √ s NN = Keywords:
Heavy-Flavor, Electron, Quark, Anisotropic Flow, Parton Energy Loss, Nuclear Modification Factors
1. Introduction
Measurements of hadrons containing Heavy-Flavor (HF) quarks are one of the key probes of the QuarkGluon Plasma (QGP) produced in heavy-ion collisions since HF quark production is restricted to the initialhard scatterings before the formation of the QGP. Parton energy loss in the QGP medium is expected tofollow a hierarchy ordered by parton color charge and mass, i.e., ∆ E ( g ) > ∆ E ( u , d ) > ∆ E ( c ) > ∆ E ( b ).Measurements of charm hadron nuclear modification factors ( R AA ) at the RHIC ( √ s NN =
200 GeV) andLHC show values that are comparable to light-flavor hadrons at high transverse momentum ( p T ). Thesimilar values of R AA can be explained by models with parton energy loss in the QGP taking into accountdi ff erences in fragmentation and spectra between light and charm quarks. A systematic comparison ofbottom and charm hadron R AA can elucidate the mass dependence of parton energy loss since models predictsignificantly di ff erent values for both. These proceedings present measurements of single electron R AA frombottom and charm semileptonic decays using the Heavy Flavor Tracker (HFT) at STAR. Additionally, decayleptons preserve the HF hadron momentum direction and are therefore excellent proxies for anisotropic flowmeasurements which provide information of the HF quark transport properties of the QGP, and the initialtilt of the QGP bulk in the case of HF directed flow ( v ). a r X i v : . [ nu c l - e x ] M a r Matthew Kelsey / Nuclear Physics A 00 (2020) 1–4
2. Bottom and Charm Decay Electron Nuclear Modification Factors
The fraction of bottom decay electrons to the sum of all non-photonic electrons (NPE) in Au + Au colli-sions is extracted with a template fit to the log(DCA / cm) distribution, where the DCA is the 3D distance ofclosest approach of a candidate electron to the primary vertex, as shown in Fig. 1 left panel. The fractionsof N ( b → e ) / N ( b + c → e ) are shown in Fig. 1 middle and right panels for minimum bias (MB) and in binsof centrality, respectively, for Au + Au collisions at √ s NN =
200 GeV. The ratios show a clear enhancementin MB and (mid-)central collisions with respect to the p + p data ([1] and preliminary 2012 STAR data) andthe FONLL predictions [2, 3]; The ratios in peripheral collisions are consistent with p + p data and FONLLpredictions. log(DCA/cm) - - - - E v e n t s / ( . ) c [3.5,4.5] GeV/ ˛ T p STAR Preliminary = 200 GeV NN sAu+Au 0-80% Combined template e Photonic Hadrons e fi c e fi b log(DCA/cm) - - - - P u ll - ) c (GeV/ T p ) e → c + b ) / N ( e → b N ( STAR Preliminary 0-80% = 200 GeV NN sAu+Au = 200 GeVs p + p = 200 GeVs p + p FONLL ) c (GeV/ T p ) e → b + c ) / N ( e → b N ( p+p FONLL
STAR preliminary = 200 GeV NN sAu+Au Fig. 1. Left: Example fit to the log(DCA / cm) distribution in MB √ s NN =
200 GeV Au + Au collisions. Middle: The measured fractionof bottom decay electrons to all NPE electrons in MB √ s NN =
200 GeV Au + Au collisions. Right: The bottom decay electron fractionas a function of p T in di ff erent collision centrality categories. In the middle and right panels the p + p data ([1] and preliminary 2012STAR data) and the FONLL predictions [2, 3] (black dashed line) are also shown. Statistical uncertainties are shown as error bars andsystematic ones as brackets. AA R STAR Preliminary =200 GeV NN sAu+Au 0-80% e → b e → c e → b DUKE e → c DUKE ) c (GeV/ T p e → c AA / R e → b AA R DUKENull hyp.
Fig. 2. The measured R AA for bottom and charm decay electronsas a function of electron p T . Statistical uncertainties on the dataare shown as error bars and systematic ones as brackets; the shadedgray boxes show the common uncertainty due to the inclusive NPE R AA measurement. The blue shaded region at R AA = coll . . The bottom panel shows the bottom to charmdecay electron R AA ratio. The blue shaded curve shows the nullhypothesis described in the text. In both panels Duke model [4]predictions are shown as the dotted lines. Using the bottom decay electron fractions inAu + Au and p + p collisions, and inclusive NPE R AA from preliminary STAR Au + Au data, thebottom and charm decay electron R AA are cal-culated and shown in Fig. 2 along with their ra-tio. The data show the bottom decay electron R AA is larger than that of charm decay electrons,and from a constant fit to the ratio this sepa-ration is 1.92 ± ± ff erent from unity at roughly a3 σ level. A null hypothesis for the ratio (blueshaded curve) is constructed by applying D me-son R AA [5] to the c / b → e simulation, whichtakes into account the di ff erent decay kinemat-ics in the R AA ratio. The p-value of the datato this curve is found to be .014, disfavoringthe hypothesis of identical charm and bottomhadron R AA .We compare to the Duke Langevin transportmodel [4] shown as the dotted lines in both pan-els, which contains the mass dependence of en-ergy loss and other e ff ects which may influencethe measured R AA (e.g., hadronization and ini-tial HF hadron p T spectra). Within uncertaintiesthe model is able to describe both the absolutevalues of R AA and their ratio. This compatibility between the data and model shows a good indication thatbottom quarks lose less energy in the QGP compared to charm quarks. atthew Kelsey / Nuclear Physics A 00 (2020) 1–4 An additional measurement of the ratios of bottom to charm decay electron R CP is performed and thedata show no strong p T dependence. Performing the same constant fits described in the above text, we find R CP (0-20% / = ± ± R CP (0-20% / = ± ± σ and 4.4 σ , respectively.
3. Bottom and Charm Decay Electron Anisotropic Flow
The charm decay electron v data are shown in Fig. 3 left and right panels for both charge averaged andsplit by charge, respectively. The average electron d v / d y is in agreement with the STAR D data [6], and issignificant at a 5 σ level. To check for decorrelation in the c → e decay we folded the measured D v intosimulation, and find no significant v information is lost in the decay electron within the measured p T range.The better precision of the electron measurement will be able to further constrain the charm quark dragcoe ffi cient and tilt of the QGP bulk. The di ff erence between e + and e − v is compared to a hydrodynamicmodel including initial electromagnetic e ff ects [7], and is consistent with zero within statistical uncertainties. Rapidity − v − − − − − e → c D STAR c > = 1.5 GeV/ eT p , < c > 1.2 GeV/ eT p STAR Preliminary ± ± /dy = -0.051 dv e ± ± /dy = -0.080 dv D v − − − − + e → c - e → c c > = 1.5 GeV/ eT p > 1.2 GeV, < eT p STAR Preliminary ± ± /dy = -0.059 dv + e ± ± /dy = -0.044 dv - e Rapidity0.5 − v ∆ − − ± ± /dy = -0.015 v ∆ d D Hydro+EM
Fig. 3. Left: Measured data for the charm decay electron v compared with the STAR D measurement [6]. Linear fits to both theelectron (black dot-dash) and D (blue dot-dot-dot-dash) data are shown. Right: Measured v split by electron charge are shown inthe top panel with linear fits to the data shown as the blue dotted and red dot-dashed lines, respectively. The bottom panel shows thedi ff erence between e + v and e − v and a fit to the the di ff erence as the black dot-dashed line, and a hydrodynamic model includinginitial electromagnetic e ff ects [7] as the solid magenta curve. Statistical uncertainties are shown as error bars and systematic ones asbrackets. The charm and bottom decay electron v data are shown in Fig. 4 left and right panels, respectively. Thecontributions from non-flow e ff ects (shown as the shaded gray boxes) are estimated using electron-hadroncorrelations in semileptonic charm and bottom decays in PYTHIA. The measured charm decay electron v is in agreement with the measured STAR D v [8] folded to the decay electron in simulation shown as theshaded magenta band. For the bottom decay electron v two event plane reconstruction methods are usedwith tracks reconstructed in the Time Projection Chamber (TPC, -1 < η <
1) or hits in the Forward MesonSpectrometer (FMS, -2.5 < η < -4). In both panels the Duke model is also shown as the dotted black line,and within uncertainties is able to describe the c → e v data well. After subtracting non-flow contributions,the model is also able to describe the b → e v data within uncertainties. A null hypothesis ( v =
0) for thebottom decay electron v using the TPC event plane with the non-flow subtracted from the central valuegives a p-value of .00067 (3.4 σ ), indicating evidence for non-zero bottom decay electron v .
4. Inclusive NPE v in √ s NN = + Au Collisions
The inclusive NPE v is measured in √ s NN = + Au collisions in a similar way to previousSTAR measurements at √ s NN = v that issimilar in magnitude to the data at 200 GeV. Matthew Kelsey / Nuclear Physics A 00 (2020) 1–4 ) c (GeV/ T p v =200 GeV NN sAu+Au e fi c v e fi D Folded 0-80% e fi c Duke ) c (GeV/ T p ) e fi b ( v - =200 GeV NN sAu+Au TPC Event PlaneFMS Event Plane e fi b Duke
Fig. 4. Left: Measured data for the charm decay electron v . The folded STAR D data [8] are shown as the shaded magenta band(described in the text). Right: Measured bottom decay electron v using the TPC (closed black circles) and FMS (open blue squares)event plane methods. Statistical uncertainties are shown as error bars and systematic ones as brackets; the gray boxes show theestimated non-flow contributions. In both panels the Duke model [4] is shown as the dotted black line.
5. Conclusion (GeV/c) T p H F e l e c t r on v Au+Au, 0-60%
200 GeV62.4 GeV54.4 GeV
STAR Preliminary
Fig. 5. Measured inclusive NPE v in √ s NN = + Aucollisions. Also shown are the previous STAR measurements of in-clusive NPE v in √ s NN =
200 and 62.4 GeV Au + Au collisions [9].Statistical uncertainties are shown as error bars and systematic onesas brackets.
We measure the charm and bottom decayelectron R AA using the full STAR HFT data setand find that the bottom decay electron R AA is larger than the charm decay electron R AA by roughly a factor of two with a significanceof about 3 σ . The measured ratios of bottomand charm decay electron R CP show the samehierarchy with significance greater than 3.5 σ .These observations, combined with the agree-ment with the Duke model including mass de-pendence of quark energy loss, are consistentwith the mass hierarchy of parton energy loss.The data for charm decay electron v and v are consistent with existing measurements of D by STAR and provide additional constraints onQGP transport properties, and in the former casefurther information on the initial tilt of the QGP.The first measurement of significant non-zero bottom decay electron v (3.4 σ from zero)at RHIC is presented. The measured v values with the estimated non-flow contributions subtracted areconsistent with Duke model calculations incorporating bottom quark transport in the QGP.Finally, with the new data in √ s NN = + Au collisions, we measure a significant inclusiveNPE v that is consistent with the previous STAR measurement at √ s NN =
200 GeV. This measurementindicates that charm quarks interact strongly with the QGP medium in √ s NN = + Au collisions.