Measurement of Non-photonic Electrons in p + p Collisions at s NN − − − − √ = 200 GeV with reduced detector material in STAR
aa r X i v : . [ h e p - e x ] J a n Measurement of Non-photonic Electrons in p + pCollisions at √ s N N = 200 GeV with reduceddetector material in STAR
F.Jin for the STAR Collaboration a , b a Shanghai Institute of Applied Physics, Chinese Academy of Sciences,P.O. Box800-204, Shanghai 201800, China b Brookhaven National Laboratory, Upton, NY 11973, USAEmail: [email protected]
Abstract.
In this paper, we present our analysis of mid-rapidity non-photonic electron (NPE)production at p T > c in p+p collisions at √ s NN = 200 GeV. The datasetis ∼
78M TOF-triggered events taken from RHIC year 2008 runs. Through themeasurement of e/π ratio, we find that the photonic background electrons from γ conversions are reduced by about a factor of 10 compared with those in STAR previousruns due to the absence of inner tracking detectors and the supporting materials. Thedramatic increase of signal-to-background ratio will allow us to improve the precisionon extracting the charm cross-section via its semi-leptonic decays to electrons. Submitted to:
J. Phys. G: Nucl. Phys.
1. Introduction
Ultra-relativistic heavy ion collisions can provide sufficient conditions for the formationof a deconfined plasma of quarks and gluons. Heavy-flavor quarks(charm and bottom)are produced dominantly through high- Q partonic interactions [1]. Because of the largemass, it’s expected that the cross-section of heavy flavor production can be calculatedin perturbative quantum chromodynamics (pQCD) [2]. Precise measurements of charmtotal cross-section and transverse momentum spectrum in p+p collisions will provide abaseline to understand the charm production and in-medium mechanism in heavy ioncollisions [3].To date, one way to study heavy flavor production is to measure NPE productionfrom their semi-leptonic decay. Although the systematic uncertainties are quite large,the charm cross-section measured by STAR is different from that measured by PHENIXby a factor ∼ σ [1, 4]. STAR has large and uniform acceptance, but the materialclose to beam pipe in previous run is ∼ X ). In run8, STARremoved inner tracking detectors, SVT (Silicon Vertex Tracker) and SSD (Silicon StripDetector). The material budget integrating from interaction point to TPC inner fieldcage is ∼ X . There are wraps around the beam pipe to bake out the beam pipeand glues at inner field cage, which are estimated. The exact material in terms ofradiation length is mapped from the data.During the 2008 RHIC runs, TPC has upgraded the electronics of one of its 24sectors to a factor of 10 faster with negligible dead time using a pipeline buffer (TPXin DAQ1000) [5]. Fully instrumentation of the 24 sectors will been completed afterrun8. Five trays of Time-of-Flight (TOF) was placed behind the TPX sector, and eachtray covers -1 < | η | < φ<π/
30 in azimuth. Two pVPDs wereinstalled to provide a starting time for TOF detectors, each staying 5.4m away from theTPC center along the beam line. The starting time resolution is ∼ ∼ ∼ ∼ p T inan identical procedure by a combination of ionization energy loss dE/dx in TPC andvelocity β from TOF [6], many of the systematic uncertainties associated with individualcharge pion and electron cancel. In the analyses presented here, we’ll focus on the e/π ratio and compare the ratio from previous measurements and background. Afterselection of good runs and a vertex cut of | z vtx | < < y < e/π ratio. In order to get good primary track, we have a vertex Z difference cut, | z vtx (pVPD)- z vtx (TPC) | <
2. Particle (electron and pion) Identification
The relativistic rise of the dE/dx in TPC from electrons provides a possible separationof electrons from the rest of the hadrons except dE/dx from slow hadrons impingingthe electron dE/dx band at several crossing points as function of momentum. Electronidentification requires TOF PID cut, | β -1 | < dE/dx distribution of electron and fast hadrons as a function of p T , shown in Fig 1(b). Projecting this plot in different p T bins and using suitable function fit to dE/dx distribution, we obtain the raw yields of electron.We used two function forms to estimate the background dE/dx distribution. One isa Gaussian function and the other is an exponential function. We found the 2-Gaussianfunction cannot describe the left shoulder region of electron dE/dx due to the tail ofthe dE/dx from fast hadron in lower p T region( p T < c ). Instead, a functionof exponential+Gaussian was used in the fit. We also use two methods to producethe background tail shape of fast hadrons and evaluate the uncertainty due to hadroncontamination: 1)inverse velocity difference between measurement and calculation Figure 1. (Color online)(a) particle inverse velocity 1/ β as a function of momentum p . The zone between two red solid lines stands for TOF PID cut, | β -1 | < π sample to make sure the tail shape of fast hadrons. (b) dE/dx ofelectron and fast hadrons versus transverse momentum p T . < β (measured)-1/ β ( π ) < dE/dx distribution; 2)energydeposited in EMC E< dE/dx distribution together with a background distribution frommethod 1) in 0.5
Figure 2. (Color online)Red dashed line stands for electron dE/dx distribution; bluedotted-dashed line stands for fast hadrons (In our analysis, it’s π below p T < c );pink histogram stands for pure π .(a) the black solid line stands for 2-Gaussian fit. Itcan’t describe the overlap well. (b) the black solid line stands for exponential+Gaussianfit. It fits the tail shape of fast hadrons, the peak and width of electron and their overlapwell. In 1.6
0. (b) Example of the m distribution in a given p T bin. Redsolid line is fit for π ; blue dashed line for K; pink dotted-dashed line for p. Counting the entries at the range -0.1 5% in low p T rangeand was used as part of the systematic uncertainty. 3. Non-photonic and photonic background electrons The inclusive electron raw yields have three components [7]: (1)electrons from heavy-flavor decay (charm and bottom), (2)photonic background electrons from Dalitz decaysof light mesons ( π , η etc.) and gamma conversion. (3)other background electrons from K e decays and dielectron decays of vector mesons. Photonic background 2) is muchlarger than other background, so we will use signed DCA (sDCA) (distance of closestapproach of a track in TPC to the interaction point) to reject the electron backgroundfrom gamma conversion at high radius and cocktail method to remove background fromDalitz decays of light mesons.Figure. 4 (a) shows radial distance (r) distribution of gamma decay vertex to theprimary vertex from a GEANT simulation. There are two major background sources ofgamma conversion, material around the beam pipe ( Be beam pipe ∼ X + wrapsfor the beam pipe bake-out) and TPC Inner Field Cage (IFC ∼ X ). Here, weused sDCA cut to remove gamma conversion at high radius ( < C oun t s rDecayVertex (cm) TPC gasSTAR Preliminary Beam pipe: 0.29%X Wrap: 0.14%X Air: 0.1%X IFC: 0.45%X (a) B ea m p i pe T P C I nne r F i e l d C age (GeV/c) T p s DCA ( c m ) -2024 IFC sDCA cut STAR Preliminary(b) Figure 4. (Color online)(a) radial distance of gamma decay vertex to the primaryvertex from simulation. (b) sDCA as a function of p T . Lines are the sDCA forconversion at TPC inner field cage (IFC) and the sDCA cuts. shows the sDCA as a function of p T from run8 data. A hand calculation of wherethe sDCA should be from conversions at the IFC agrees nicely with the band in thedata. sDCA = q ( p T / . + r − ( p T / . r is the γ conversion radiusin a uniform solenoidal magnetic field of 0.5Tesla. We can use this expression to getthe sDCA value of sDCA1 when r =30cm. With -0.5 < sDCA < sDCA1 cut, we rejectedelectron from gamma conversion in the air with r> π Dalitz decays. Through fit tothe charge π spectra in non-singly diffractive (NSD) p+p collisions, we get a function B/ (1 + ( m T − m ) /nT ) n (in this expression, m T and m stand for the particle transversemass and rest mass separately and it has three parameters: B , n and T ). With fixedparameter n = 9.7, this expression fit not only charge π spectra but also charge kaon, K ⋆ , ρ and φ well, so we use this expression as input to the generator. With cocktailmethod, we get the electron background from Dalitz decays of light mesons. e/π ratio Figure. 5 shows the e/π ratio from run8 data, compared to various background cocktails,NPE from previous results, and run3 inclusive electron to pion ratio. We also check theconsistence of the e/π ratio from run8 and the results from run3. We take the materialbudget from which γ conversion in detector is ∼ 10 in run3 than that in run8, and weinclude the e/π from π and e/π from η Dalitz decays from cocktail method and NPE/ π measured in run3 d+Au data scaled by the binary collisions together, then we find thetotal sum of e/π from run8 is consistent with STAR TOF inclusive e/π in run3, and the p T dependence can be well reproduced as well. In addition, γ conversion is equivalent to (GeV/c) T p0 0.5 1 1.5 2 2.5 3 3.5 p e / -4 -3 -2 -1 p Dalitz/ p STAR Simulation PRL98 p STAR EMC NPE/ PRL98 p STAR TOF inclusive/ PRL94 p STAR TOF NPE/ PRL94 p / bin STAR d+Au NPE/N predicted p STAR inclusive/ p Dalitz/ h STAR Simulation in run3 p+p p e/ STAR Preliminary (a) (GeV/c) T p0 0.5 1 1.5 2 2.5 3 3.5 p e / -4 -3 -2 -1 p Dalitz/ p STAR Simulation run8 p+p p STAR inclusive/ PRL98 p STAR EMC NPE/ PRL94 p STAR TOF NPE/ PRL94 p / bin STAR d+Au NPE/N predicted p STAR inclusive/ p Dalitz/ h STAR Simulation in run8 p+p p e/ STAR Preliminary (b) Figure 5. (Color online)(a) e/π ratio as a function of p T . The dashed line is thesum of various background e/π ratio ( π / π , η/π γ / π and NPE/ π ) for run3 with anestimate of material around the beam pipe to be a factor of x 10 higher than that inrun8. (b) Full circles represent the inclusive e/π ratio in run8 and the dashed line issimilar as (a) by requiring 0.69% X for gamma conversions. π Dalitz decay in run8, comparable with the estimatedmaterial budget in this run. 5. Conclusions In summary, we present our analysis of mid-rapidity NPE production at p T > c in p+p collisions at √ s NN = 200 GeV. Through the measurement of e/π ratio, we findthat the photonic background electrons from gamma conversions are reduced by abouta factor of 10 compared with those in STAR previous runs due to the absence of innertracking detectors and the supporting materials. and preliminary results from run8dataset agree with the results from run3. 6. Acknowledgements This work was supported in part by the National Natural Science Foundation of Chinaunder grant No. 10610285 and No.-10875159, the Knowledge Innovation Project of theChinese Academy of Science under grant nos. KJCX2-YW-A14 and KJCX3-SYW-N2. References [1] B.I. Abelev, et al. ,[STAR Collaboration], Phys. Rev. Lett. , 192301 (2007);[2] M. Cacciari et al. , Phys. Rev. Lett. , 122001 (2005);[3] B. W. Zhang et al. , Phys. Rev. Lett. , 072301 (2004);[4] A. Adare, et al. ,[PHENIX Collaboration], arXiv:0802.0050v2 [hep-ex];[5] R. Esteve Bosch, A. Jimenez de Parga, B. Mota, and L. Musa, IEEE Trans. Nucl. Sci. , 50:2460–2469, 2003.[6] M. Shao, O. Barannikova, X. Dong, Y. Fisyak, L. Ruan, P. Sorensen, Z. Xu, arXiv:nucl-ex/0505026;[7] A. Adare, et al. ,[PHENIX Collaboration], Phys. Rev. Lett.97