Current status of Upsilon background and RAA Sasha
Current status of Upsilon background and RAA Sasha Lebedev (ISU) 8 th s. PHENIX Collaboration Meeting 7 -9 July 2019 – Lehigh University 1
Introduction Technical note with all the details: https: //www. phenix. bnl. gov/WWW/p/draft/lebedev/s. PHENIX/anote_bg/newnote. pdf What goes into the calculation: 1) Hadron rejection factor (combinatorial background). 2) Upsilon mass resolution (signal/background ratio). 3) Correlated (physics) background (mostly Drell-Yan and correlated bottom). Cuts and assumptions: Electron ID efficiency 90% (using E/p in CEMC only) Electron p. T > 2 Ge. V Tracking efficiency 100% Mass resolution in p+p 89 Me. V, in Au+Au 100 Me. V Only correlated background in p+p 2
Expected Upsilon yields Number of events from s. PHENIX five-year running scenario s. PH-TRG-000 - 240 B events in min. bias Au+Au and 8300 B events in p+p collisions. - Upsilon numbers scaled up from s. PHENIX proposal. p+p 3
Hadron rejection EHCALIN/ECEMC Full GEANT-4 simulation. Single hadrons embedded in central Hijing Au+Au events. We used only ECEMC/P cut with 90% electron ID efficiency pions electrons Energy in 3 x 3 block around the track projection was used as ECEMC No clustering ECEMC/P 4
Pions and Kaons 5
Protons and anti-protons 6
Fake electron yields Hadron yields from PHENIX measurement (ppg 014 and ppg 030) Hadron suppression in Au+Au is taken into account (ppg 146). blue: pions magenta: kaons black: anti-protons green: protons red: all fake positives brown: all fake negatives 7
Correlated background in Au+Au Drell-Yan: tuned PYTHIA. Charm and bottom: PHENIX measurement using measured charm/bottom ratio and R AA. Black: total correlated background Green: Drell-Yan Blue: correlated bottom Red: correlated charm Cross-terms like fake-e plus charm/bottom/DY or bottom-DY combination is negligible. Dominated by Drell-Yan. Invariant mass, Ge. V 8
Correlated background in p+p No beauty suppression in p+p. 8300 B p+p events. 9
Generating background Once we know p. T distributions and mean multiplicities above 2 Ge. V p. T cut, we can generate background events. Multiplicity is generated according to a Poisson distribution with additional multiplicity smearing with a Gaussian (mean = 1 and sigma = 0. 25) to reproduce NCOLL fluctuations within centrality bin. p. T generated for each particle. After enough events is generated to understand background shape around 10 Ge. V, background is fitted and the fit scaled to appropriate number of events. Cross-terms (like fake electron combined with DY electron) turned out to be negligible. 10
Signal plus background (all p. T) in Au+Au 24 B 0 -10% central Au+Au events, integrated over p. T. For combinatorial BG subtraction we assume that we know the shape and normalization. Blue: combinatorial background Red: correlated background 11
Inv. mass after combinatorial bg subtraction 12
Invariant mass distributions vs. p. T 13
RAA calculation • Suppression from Xiaojun Yao, Berndt Müller, ar. Xiv: 1811. 09644, similar to M. Strickland D. Bazow, Nucl. Phys. A 879 25, (2012) used in s. PHENIX proposal. • 24 B 0 -10% central Au+Au. Uncertainty from 8300 B p+p events is included. • Direct counting in mass windows: 9. 25 - 9. 65 Ge. V; 9. 80 - 10. 20 Ge. V; 10. 20 - 10. 55 Ge. V. • Fit using fixed Upsilon line shape to improve statistical uncertainty. - fixing shape is important (well-tuned simulation? ) 14
RAA Direct counting Fit 15
- Slides: 15