Heavy Ion Physics Lecture 3: Particle Production

Transcription

1 Heavy Ion Physics Lecture 3: Particle Production HUGS 2015 Bolek Wyslouch

2 echniques to study the plasma Radiation of hadrons Azimuthal asymmetry and radial expansion Energy loss by quarks, gluons and other particles Suppression of quarkonia 2

3 Radiation of hadrons and photons Effects dependent on energy density Charged multiplicity Energy distribution Measuring the chemistry of collision, quark content of the plasma, temperature, speed of expansion Momentum spectrum Particle composition: π, K, p, γ Comparison of particle content in nuclear and protonproton collisions 3

4 How do we measure particle yields? Identify the particle (by its mass and charge) Measure the transverse momentum spectrum Integrate it to get the total number of particles In fixed target experiment everything goes forward ( due to cm motion) easy to measure total ( 4π) yield In collider experiment: measure the yield in a slice of rapidity : dn/dy Apply corrections for acceptance and decays 4

5 ime of flight measurements:phenix 5

6 Combine multiple detectors: PHENIX 6

7 PHOBOS 7

8 ALICE particle ID in ime Projection Chamber 8

9 Remote emperature Sensing Red Hot White Hot Hot Objects produce thermal spectrum of EM radiation. Red clothes are NO red hot, reflected light is not thermal. Photon measurements must distinguish thermal radiation from other sources: HADRONS!!! 9 Not Red Hot!

10 CMS Sub-detectors: ECAL 10

11 Direct and virtual photons Direct photon and virtual photon: Created throughout evolution of system. Very low cross-section with QCD medium. Kinetic range reveals source of productions High p (>5 GeV/c) --- from initial hard scattering. Low p (1-5 GeV/c) --- from QGP. 11

12 Real vs. Virtual Photons Direct photons γ direct /γ decay ~ 0.1 at low p, and thus systematics dominate. Number of virtual photons per real photon: Hadron decay: 1/Nγ dn ee /dm ee (MeV -1 ) form factor Direct photon Point-like process: About virtual photons with m ee > M π for every real photon π 0 m ee (MeV) Avoid the π 0 background at the expense of a factor 1000 in statistics 12

13 Direct (pqcd) Radiation Measuring direct photons via virtual photons: any process that radiates γ will also radiate γ* for m<<p γ* is almost real extrapolate γ* e+e- yield to m = 0 à direct γ yield m > m π removes 90% of hadron decay background S/B improves by factor 10: 10% direct γ à 100% direct γ* arxiv: < p < 2 GeV 2 < p < 3 GeV 3 < p < 4 GeV 4 < p < 5 GeV pqcd g q access above cocktail q γ fraction or direct photons: hadron decay cocktail r γ = = γ dir incl γ γ dir incl Small excess for m<< p consistent with pqcd direct photons 13 13

14 Fit Mass Distribution to Extract the Direct Yield: Example: one p bin for Au+Au collisions Yield truncated at parent mass PHENIX f c ( m ee dir ( m ) normalized to data for m < 30 MeV ee ) and f ee Direct γ* yield fitted in range 120 to 300 MeV Insensitive to π 0 yield c cocktail dir direct

15 Interpretation as Direct Photon Relation between real and virtual photons: dσ d ee α 1 L( M ) dm dp dy 3π M dp Extrapolate real γ yield from dileptons: dn dn for M 0 dm dm M ee γ Virtual Photon excess At small mass and high p Can be interpreted as real photon excess no change in shape can be extrapolated to m=0 σ γ 2 dy

16 hermal Radiation at RHIC Direct photons from real photons: Measure inclusive photons Subtract π 0 and η decay photons at S/B < 1:10 for p <3 GeV γ Direct photons from virtual photons: Measure e + e - pairs at m π < m << p Subtract η decays at S/B ~ 1:1 Extrapolate to mass 0 γ* (e + e - ) m=0 ALICE pqcd

17 Calculation of hermal Photons D.d Enterria, D.Peressounko, Eur.Phys.J.C 46 (2006) ini = 300 to 600 MeV τ 0 = 0.15 to 0.5 fm/c Initial temperatures and times from theoretical model fits to data: 0.15 fm/c, 590 MeV (d Enterria et al.) 0.2 fm/c, MeV (Srivastava et al.) 0.5 fm/c, 300 MeV (Alam et al.) 0.17 fm/c, 580 MeV (Rasanen et al.) 0.33 fm/c, 370 MeV (urbide et al. 1 7

18 Now with Real γ! Real gamma from photon conversions. 18

19 Chemical Equilibrium In a HI collision there is cornucopia of produced particles, seemingly a nightmare. However, if the system has exhibited thermalization, then the particle production might be understood through simple considerations. We ll consider two aspects of thermal predictions: Chemical Equilibrium Are all the various particle species produced at the right relative rates and abundances? Kinetic Equilibrium Is the particle production consistent with a single underlying temperature plus common flow velocities? 19

20 Statistical Ensemble We must choose an appropriate statistical ensemble. his choice in itself is instructive to the physics: Grand Canonical Ensemble: In a large system with many produced particles we can implement conservation laws in an averaged sense via appropriate chemical potentials. Canonical Ensemble: in a small system, conservation laws must be implemented on an EVEN-BY-EVEN basis. his makes for a severe restriction of available phase space resulting in the so-called Canonical Suppression. Where is canonical required: low energy HI collisions. high energy e+e- or hh collisions Peripheral high energy HI collisions. 20

21 Canonical Suppression 21

22 K/p ratios vs. Centrality Simple expectation: Particles carrying conserved quantum numbers (strangeness, baryon number) should exhibit loss of canonical suppression with centrality. K is strangeness 1 p is baryon number 1 Normalized compared to pion, both curves rise rapidly with centrality and then saturate. 22

23 hermal yields Begin with the formula for the number density of all species: n 0 i = g 2 p dp 3 ( E µ BBi µ ssi µ 3I ) e ± 1 i 2 / 2 π here g i is the degeneracy E 2 =p 2 +m 2 µ B, µ S, µ 3 are baryon, strangeness, and isospin chemical potentials respectively + for fermions and for bosons Given the temperature and all m, on determines the equilibrium number densities of all various species. he ratios of produced particle yields between various species can be fitted to determine, µ. 23

24 Reality check: Approximate µ B assuming a temperature of 170 MeV and that the anti-proton/proton ratio is 0.7 and independent of momentum n 0 i = g 2 p dp 3 ( E µ BBi µ ssi µ 3I ) e ± 1 i 2 / 2 π All factors in the above equation cancel except the Baryon number (proton=+1, p bar =-1). So p p e e µ / µ B B / = e 2µ B Question: Which has large µ B, high energy or low energy collisions? 24 / ; µ B 30MeV

25 Conservation Constraints 25

26 Dependence of µ s on Τ,µ Β At any given and µ B, there exists only a single µ s that makes the final state strangeness neutral. Same for I 3. Entire model has two free parameters and then makes a prediction for all particle ratios. 26

27 Feed-down via decay 27

28 Controversy: γ s Many authors modify the pure thermal ansatz by introducing a strangeness fugacity γ s as: n 0 i gi ( strange) = γ s 2 2π 2 p dp 3 ( E µ BBi µ ssi µ 3I )/ e ± 1 his factor in the range 0-1 determines the level at which strangeness has reached the Grand Canonical level. Some authors feel that such a factor violates the thermal ansatz, whereas others like having a measure of the level of equilibrium. 28

29 29

30 30

31 Radial Flow For any interacting system of particles expanding into vacuum, flow is a natural consequence. During the cascade process, one naturally develops an ordering of particles with the highest common underlying velocity at the outer edge. his motion complicates the interpretation of the momentum of particles as compared to their temperature and should be subtracted. Hadrons are released in the final stages of the collision and therefore measure FREEZE-OU 31

32 Singles Spectra Peripheral: Pions are concave due to feeddown. K,p are exponential. Yields are MASS ORDERED. Central: Pions still concave. K exponential. p flattened at left Mass ordered wrong (p passes pi!!!) Central Peripheral Underlying collective VELOCIIES impart more momentum to heavier species consistent with the basic trends 32

33 Blast Wave-I Let s consider a hermal Boltzmann Source: 3 d N 3 dp e 3 d N E = 3 dp E m 3 d N dm dφdy Ee E = m cosh( y) If this source is boosted radially with a velocity β boost, the resulting distribution, evaluated at y=0 and integrated over φ is: 1 m dn dm m I 0 p cosh( y) e sinh( ρ) K 1 m m cosh( ρ) where ρ = tanh 1 ( ) β boost 33

34 Blast Wave-II he entire source from the collision may be considered as a superposition of many sources each with a different strength and boost velocity. he simplest assumption (and non-physical ) is that the source is a uniform sphere of radius R and that the boost velocity varies linearly to some maximum value. hen: 1 m ρ( r) dn dm = tanh 0 R 1 r 2 β drm MAX I r R 0 p sinh( ρ) K 1 m cosh( ρ) 34

35 Blast Wave Fits Fit AuAu spectra to blast wave model: β S (surface velocity) drops with dn/dη (temperature) almost constant. centrality 35 Centrality p (GeV/c)

36 Beam Energy Scan shows Systematics emperature ch (MeV) Au+Au Collisions at RHIC SAR Preliminary Data: 5% Au+Au collisions A. Adronic, et al. J. Cleymans, et al. LG: Z. Fodor, et al. LG: S. Gupta, et al. ( C = 170 MeV) Baryonic Chemical Potential µ B (MeV) //SAR//11 BESII//References//03/plot2_cfo_March2014.kumac// SAR Preliminary Collective velocity <β> (c) Chemical Freeze-out: (GCE) - Central collisions. - Centrality dependence, not shown, of ch and µ B! Kinetic Freeze-out: - Central collisions => lower value of kin and larger collectivity β - Stronger collectivity at higher energy

37 emperature/flow Summary Clear break in behavior ~20 GeV. 37

38 Recast Data vs µ B Lowering RHIC energy requires accelerator upgrades. BES-II planned 2018/2019

39 Summary of particle production measurements Statistical treatment of particle production in heavy ion collision describes data pretty well he produced particles come from a hot medium with temperatures consistent with expectations for a quark-gluon plasma he measurements have to account for rapid expansion of the hot region he main action is at low energies where one hopes to see effects at the phase boundary 39