Parton propagation and hadronization at the EIC

Transcription

1 Parton propagation and hadronization at the EIC Alberto Accardi Hampton U. & Jlab ENC/EIC workshop

2 Outline Physics motivations Parton propagation and fragmentation How long is parton lifetime? selected HERMES data Perspectives EIC, COMPASS, ENC Conclusions Review soon to appear: Accardi, Arleo, Brooks, d'enterria, Muccifora Parton propagation and hadronization in QCD matter

3 Physics motivations e Nuclei as space time analyzers nucleons as femto detectors medium rather well known low final state multiplicity * h q e 3

4 Physics motivations e Nuclei as space time analyzers nucleons as femto detectors medium rather well known low final state multiplicity * h q e Non perturbative aspects of hadronization approaching microscopic understanding of Fragmentation Functions how do partons dress up? Space time evolution of hadronization color confinement dynamics 4

5 Physics motivations e Nuclei as space time analyzers nucleons as femto detectors medium rather well known low final state multiplicity * h q e Non perturbative aspects of hadronization approaching microscopic understanding of Fragmentation Functions how do partons dress up? Space time evolution of hadronization color confinement dynamics Parton propagation in perturbative QCD QCD energy loss: basic pqcd, only indirectly tested so far DGLAP parton showers 5

6 Physics motivations e Nuclei as space time analyzers nucleons as femto detectors medium rather well known low final state multiplicity * h q e Non perturbative aspects of hadronization approaching microscopic understanding of Fragmentation Functions how do partons dress up? Space time evolution of hadronization color confinement dynamics Parton propagation in perturbative QCD QCD energy loss: basic pqcd, only indirectly tested so far DGLAP parton showers Connection to other fields Calibration of jet quenching in A+A properties of QGP Tuning of parton showers in Monte Carlo generators Hadron attenuation corrections for oscillation experiments 6

7 Nuclear collisions 1 - ndis e * HERMES PLB 577(03)37 NPB 780(07)1] h q Elab = 27.5 GeV? He Cold QCD matter e h dna (zh ) DIS dzh NA DA (z) h RM (zh ) = ¼ h DD (z) 1 dnd (zh ) DIS dzh ND Ne 1 Nuclear effects on PDF cancel in ratio Exposes modifications of hadronization Kr zh RM < 1 hadron attenuation in cold nuclear matter 7

8 Nuclear collisions 2 Heavy ion collisions h A h A? p q q Hot QCD matter h RAB (pt ) = A? PHENIX, PRL91(02) h 2 dn =d pt A+B TAB (b) (d¾ h =d2 pt )p+p Medium modifications of hadronization isolated by comparison of h+a and A+A d+au Au+Au RAuAu<1 & RdAu>1 hadron attenuation in hot nuclear matter 8

9 ndis vs. h e * e A q * e e ndis is a clean environment for (1) space time evolution of hadronization nucleons as femto detectors medium rather well known (2) Cold nuclear matter effects quark energy loss nuclear modifications of FF parton showers h A h q q A+A collisions Jet quenching in A+A properties of hot nuclear matter 9

10 ndis vs. h A+A collisions h A q A q * e e Eq = = Ee Ee' 2 25 GeV E q = pth / z Eh = zh 2 20 GeV E h = pth 2 20 GeV at HERMES/Jlab Q2 Eq2 (pt/z)2 is not Q2 = q2 is measured HERMES/JLAB is relevant to RHIC mid rapidity...but beware the virtuality: 2 GeV2 vs. O(100) GeV2!! 10

11 Parton propagation and fragmentation see: A.A., EPJC 2007, mini review A.A. et al, full review soon to appear

12 The (naïve) framework : quark, prehadron, hadron Hadronization is non perturbative (many) models General features: Colored prehadron : large inelastic cross sect. Caveats: Color neutralization gluon radiation stops production time tp col. neutr. time tcn prehadron collapses on hadron's h wavefunction formation time th It's tricky to define tp, tcn, th: working tools simplify: tp = tcn Leading order pqcd mindset ( *+q q), but NLO may be large 12

13 Hadron attenuation in ndis h RM (z) = 1 h dna (z) dz 1 h (z) dnd dz DIS NA DIS ND Energy loss (gluon bremsstrahlung) [Arleo; Wang et al.; Accardi] hadronization outside the medium gluon radiation off struck quark q h Prehadron absorption [Accardi et al.; Falter et al.; Kopeliovich, et al.] color neutralization inside the medium prehadron nucleon scatterings q [A.A., et al., NPA 761(05)67] [A.A., Acta.Phys.Hung. '06 & PRC '07] h 13

14 Parton lifetime 1 pqcd estimate pqcd estimate [see Vitev, Adil '07] assume, more probably, K p D B HERMES ( ~13 GeV, z~0.5) 28 fm 9 fm 7 fm 3 fm 0.8 fm 0.1 fm RHIC (pth ~ 7 GeV z~0.7) 18 fm 7 fm 6 fm 3 fm 0.8 fm 0.1 fm ~ inside the medium!! Large formation time, used in en. loss models to justify assumptions 14

15 Parton lifetime 2 Lund model Inside out cascade [Bialas Gyulassy '87] creation (string breaking) C Prehadron formed at qq i Hadron hi formed when q and q meet Pi t Average formation times are computable 8 <htp i = f (zh )(1 zh ) zh º z º :hth i = htp i + h boost string tension (non pert. scale) energy conservation At large z 1 Eh string breaks early to leave nucleus remnant X all energy to the hadron: tp 0 At small z 0 hadron created at high rank after many string breakings: tp tp t (see also GiBUU model later on in the talk) 15

16 Formation time estimates 2 Lund model < th > t \ 8 <htp i = f (zh )(1 zh ) zh º z º :hth i = htp i + h [Accardi et al., NPA 761(05)67] < t*> For a =14 GeV pion at Hermes, htp i. 5 fm hth i» 10 fm > RA shorter time scales than in pqcd estimates Rh Full fledged hadrons behave as expected: hth i» m zh º h (see also GiBUU model later on in the talk) 16

17 How long is the quark lifetime?

18 1) Hadron quenching vs. zh Red: absorption model (¾p = 0:65 ¾h) [A.A., et al., NPA 761(05)67] Blue: energy loss model with SW quenching weights (^ q = 0:6 GeV2 =fm) NPB 780(07)1 [A.A., Acta.Phys.Hung. '06 & PRC '07] Both describe the data: no info on parton lifetime 18

19 1) Hadron quenching vs. zh Red: absorption model (¾p = 0:65 ¾h) [A.A., et al., NPA 761(05)67] Blue: energy loss model with SW quenching weights (^ q = 0:6 GeV2 =fm) NPB 780(07)1 [A.A., Acta.Phys.Hung. '06 & PRC '07] Both describe the data: no info on parton lifetime 19

20 2) Scaling of RM [A.A., PLB B649 (07) 384] RM should scale with zh, not with z and separately RM = RM (zh ; º) with = Czh (1 zh)º Scaling exponent can distinguish absorption and energy loss absorption models: > 0 [short hadron formation time] energy loss models: 0 [from energy conservation] Hadronization starts inside the nucleus! 20

21 3) pt broadening [A.A., nucl th/ ] Assume very long production times: pure energy loss models h p2t i 3 ¼ q^l ¼ q^ RA 4 hp2t i = hp2t ia hp2t id e h * pt q Use the extracted e q^ = 0:6 GeV2 /fm tp>ra Obtain too large broadening!! Ne Kr Xe h p2t i 1.4 GeV2 2.2 GeV2 2.6 GeV2 21

22 3) pt broadening [A.A., nucl th/ ] Assume short tp and no energy loss: In prehadron stage, no broadening: elastic scattering very small Incoherent partonic scattering: pt2 linear in quark in medium path hp2t i / htp i ¼ C zh0:5 (1 zh ) º hp2t i = hp2t ia hp2t id e h * q pt e pt broadening should: 1) rise with A1/3 until tp RA, then level off tp< RA 2) decrease as zh 1 3) rise with, then level off 4) possibly, decrease with Q2 if tp /Q2 as claimed by Kopeliovich et al., NPA 740(04)211 22

23 3) pt broadening [A.A., nucl th/ ] Assume short tp and no energy loss: hp2t i / htp i ¼ C zh0:5 (1 zh ) º 4 Let's assume: htp i ¼ RXe 3 at zh = 0:4 º = 14 GeV hp2t i / htp i ¼ (0:8 GeV) zh0:5 (1 zh ) º 23

24 3) pt broadening [A.A., nucl th/ ] Assume short tp and no energy loss: hp2t i / htp i ¼ 0:8 zh0:5 (1 zh ) º It should: 1) rise with A1/3, then level off 2) decrease as zh 1 3) rise with then level off 4) possibly, decrease with Q2 (if pt2 1/Q2)?? at strong variance with Kopeliovich et al. Partonic dynamics beyond production time & multi scattering modified DGLAP evolution? NLO effects? colored dipoles? Role of virtuality? 24

25 Perspectives

26 Present and future e+a facilities = future = present Elab=6 GeV Elab=12 GeV Elab=53 GeV (EIC Elab = GeV) Elab = 160 GeV Elab = 12 & 27 GeV 26

27 EIC

28 The EIC high luminosity HERMES small x, large, large Q2 reach Test /extend HERMES cross check results multi differential observables 2 particle correlation (h h, h,...) Unique: tests of parton dynamics heavy flavors, jets,... EIC ENC 28

29 2 large Q, The EIC large 2 W Q,W Large range: 10 < < 1600 GeV hadrons formed well outside of the nuclear medium in medium parton dynamics can be experimentally isolated Large Q2: role of virtuality in hadron attenuation New access to pt broadening studies fundamental tests of pqcd energy loss study medium modification of DGLAP evolution test parton shower algorithms in Monte Carlo generators (!!) Heavy flavors: radiative vs. collisional parton energy loss J/psi normal absorption in clean environment,... Plus: Jet shape modifications, dijets, +jet, dihadron correlations,... vs., baryon fragmentation, small x & CGC,... 29

30 EIC vs. HERMES [Accardi, Dupré, Hafidi, EIC e+a note] EIC HERMES Simulation with PYTHIA no nuclear effect yet 10 weeks of beam at erhic High statistics: from 2D to 5D distributions Large reach in pt and Q2 (xb) Large range in small hadronization inside A large precision tests of QCD en. loss, DGLAP evolution, parton showers (Simulations by R.Dupré) charged pions erhic GeV 30

31 0 vs. in hot nuclear matter Why is as much suppressed as in Au+Au? points towards long lived quark but ndis analysis suggests formed on short time scales nucl ex/ Q2 (pt/z)2 Is it so also in ndis? [ is heavier hadronizes earlier, smaller x sect.] measurement possible at CLAS (low Q2), EIC (high Q2) 31

32 0 vs. at EIC attenuation [Accardi, Dupré, Hafidi, EIC e+a note] Can precisely test if RM (¼ 0 ) ¼ RM ( ) as seen in the QGP at RHIC h RM (zh ) = 1 DIS NA h h dna (zh ). 1 dnd (zh ) DIS dzh dzh ND EIC ¼ 0 h RM EIC hºi = 14 GeV hzh i = 0:4 HERMES ¼0 zh (Simulations by R.Dupré) 32

33 0 vs. at EIC pt broadening [Accardi, Dupré, Hafidi, EIC e+a note] Can detect different dynamics, if any h p2t i [GeV2 ] Tests role of virtuality in detail hp2t i = hp2t ia hp2t id hºi = 14 GeV hzh i = 0:4 HERMES ¼ medium modified DGLAP (Domdey et al.) 0 EIC ¼ 0 pqcd scaling of pions (slide 14) EIC Q2 [ GeV2 ] (Simulations by R.Dupré) 33

34 Large light vs heavy [Accardi, Dupré, Hafidi, EIC e+a note] At large formation times are very long attenuation disappears pt broadening little affected ideal test of pqcd energy loss h p2t i [GeV2 ] ¼0 D B EIC hºi = 1000 GeV hzh i = 0:4 Q2 [GeV2 ] NOTE: pt broadening values are arbitrary 2% acceptance for heavy flavors (Simulations by R.Dupré) 34

35 Large light vs heavy [Accardi, Dupré, Hafidi, EIC e+a note] At large formation times are very long attenuation disappears pt broadening little affected ideal test of pqcd energy loss h p2t i [GeV2 ] ¼0 D B EIC hºi = 1000 GeV hzh i = 0:4 Q2 [GeV2 ] NOTE: pt broadening values are arbitrary 2% acceptance for heavy flavors Q2 range relevant to A+A (Simulations by R.Dupré) 35

36 Large light vs heavy flavors [Accardi, Dupré, Hafidi, EIC e+a note] At large formation times are very long attenuation disappears pt broadening little affected ideal test of pqcd energy loss Wicks et al., nucl th/ D B EIC hºi = 1000 GeV hzh i = 0:4 electrons RAA h p2t i [GeV2 ] ¼ Can solve RHIC's heavy flavor puzzle 0 pt [GeV/c] Q2 [GeV2 ] NOTE: pt broadening values are arbitrary 2% acceptance for heavy flavors Q2 range relevant to A+A (Simulations by R.Dupré) 36

37 Heavy quarks identification D and B particles have a non negligible life time (cτ ~ 100 μm) High precision vertex tracking used or planned in a lot of experiments ( ALICE, CMS STAR, PHENIX ZEUS... ): Silicon strip detector Pixel detector... Needs spatial resolution [A.Bruell, EIC MC Group Boulder, Nov. 04] x. 100 ¹m hadrons but more studies are needed D 100 µm 37

38 COMPASS

39 Perspectives at COMPASS Nuclear targets are possible [Abbon et al., Nucl. Instrum. Meth. A577 (2007) 455] Larger than HERMES <150 GeV) pure parton energy loss studies DGLAP evolution before the EIC! vs. e beams energy loss in DY compared to DIS 39

40 ENC

41 Perspectives at the ENC Energy between HERMES and COMPASS <50 GeV hadronization not completely outside the nucleus?? Higher luminosity multi differential observables heavy flavors (?) Collider mode target vs. current fragmentation 41

42 Conclusions Quark lifetime: rather short in cold matter seems, O(RA) but long in hot matter at RHIC a QGP signal??? In ndis, pt broadening is next theoretical challenge test of space time evolution of hadronization best place to test: pqcd energy loss, DGLAP, parton showers role of virtuality in hadron attenuation At the EIC: many new channels (jets, heavy flavors,...) long parton lifetimes: precision study of en.loss, DGLAP opportunities for theoretical developments Open question: perspectives at the ENC, COMPASS 42

43 Conclusions Quark lifetime: rather short in cold matter seems, O(RA) but long in hot matter at RHIC a QGP signal??? In ndis, pt broadening is next theoretical challenge test of space time evolution of hadronization best place to test: pqcd energy loss, DGLAP, parton showers role of virtuality in hadron attenuation At the EIC: many new channels (jets, heavy flavors,...) long parton lifetimes: precision study of en.loss, DGLAP opportunities for theoretical developments Open question: perspectives at the ENC, COMPASS need good PID, vertex detector 43

44 The end

45 Backup slides

46 Parton lifetime in hot QCD matter

47 0 vs. Why is as much suppressed as in Au+Au? points towards long lived quark but ndis analysis suggests formed on short time scales nucl ex/ Q2 (pt/z)2 Is it so also in ndis? [ is heavier hadronizes earlier, smaller x sect.] measurement possible at CLAS (low Q2), EIC (high Q2) 47

48 From ndis to heavy-ion collisions GiBUU absorption model: tp th [Falter et al. PRC '04, Gallmeister, Mosel NPA '08] Formation times: tp from PITHYA (Lund model) Cross sections: leading h: =f(t) h 8 subleading h: * = 0 < const. t (linear, or quantum di usion) f (t) / : 2 t (color transparency) Consistency of HERMES + EMC data selects linear f(t)~t [Gallmeister, Mosel NPA'08] 48

49 Parton lifetime in hot QCD matter 2 Apply it to Au+Au collisions at RHIC: [Cassing, Gallmeister,Greiner NPA '04; Cassing, Gallmeister NPA '05] Way too little suppression: long lived partons in QGP? Why do partons seem short lived in cold QCD matter, but long lived in hot QCD matter? 49

50 Short review of e+a data For the latest data see: 1) HERMES, NPB 780 (2007) 1 2) Trento Fragmentation Workshop, Feb

51 Measurements at HERA HERMES: fixed target, Elab = 27.5 GeV and 12 GeV ¼+ RM Hadron attenuation versus = virtual energy zh = Eh/ (hadron's fractional energy) Q2 = photon virtuality pt = hadron transv. momentum hadron flavor =, K, p, p A = target mass number Hadron pt broadening zh Dihadron correlations 2 dimensional binning (!) [HERMES, NPB 780 (2007) 1] zh º [GeV] pt [GeV] 51

52 Preliminary results at Jefferson Lab 6 (12) GeV beam huge luminosity multi differential binning!! +h correlations and much more... RM Elab = 5 GeV K.Hafidi, nucl ex/ Brooks, Hicks, talks at Trento Workshop pt broadening de s = N c pt2 dx Elab = 5 GeV 52

53 Proton anomaly also in cold matter! HERMES Elab=27 GeV Phys.Lett.B577(03)37 Kr Kr proton anomaly!...but anti protons are normal... analogous to baryon/meson anomaly in p+p, p+a and A+A what do they have in common, if anything? What about Higher Twist baryon production? 53

54 Hadronization in elementary collisions perturbative QCD factorization of short and long distance physics X ij d¾hadron = Ái ¾ ^parton Djjh j ij Fragmentation Fns (from e++e h+x) Parton Distribution Fns (from inclusive DIS) Universality: Fragm. Fns. from e++e h+x describe hadronization in DIS and p+p h+x j j i i 54

55 The collider point of view p xe= Ee = s [A.A., PRC 76 (07) ] h (pth, y1) A A e+, xe h (pth, y1) A q, x A q2, x2 e+ (pt,y2) q, x q (pt,y2) In LO kinematics: Q2 = p2t (1 + ey1 y2 ) p pt s y1 º= e 2M 1 y= 1 + ey2 y1 zh = z 55

56 Cold quenching in p+a and A+A collisions A.A., PRC76 (07) At low s, or negative, a parton travels slowly through the target nucleus, when seen in the nucleus rest frame. h (pth, y1) p A q2, x2 q1, x1 q1 same as h (,zh) q2 At LO: q (pt,y2) Q2 = p2t (1 + ey1 y2 ) p pt s y1 º= e 2M zh = z If parton slow enough (small ) the hadron will be quenched in the cold nucleus by the same mechanism that quenches it in ndis. 56

57 Cold quenching in p+a and A+A collisions Use A.A., PRC76 (07) p pt s y1 Q2 = p2t (1 + ey1 y2 ) º= e zh = z 2M in any hadron quenching model validated by HERMES RM data e.g., energy loss with SW quenching weights [AA,...] For A+A at suppression comes from both nuclei ~ squared: Cold quenching in target nuclei not negligible in A+A at SPS: easily competes with quenching from hot medium 57

58 8 <htp i = f (zh )(1 zh ) zh º z º :hth i = htp i + h t \ Formation time estimates 2 Lund model < th > < t*> For a =14 GeV pion at Hermes, htp i. 5 fm hth i» 10 fm > RA Prehadron absorption with this estimate [A.A. et al., NPA 761(05)67] see also: Falter, Gallmeister, nucl th/ for similar ideas in a transport model Monte Carlo simulation RM z 58

59 Formation time estimates 3 Dipole model Leading hadron formation (z > 0.5) [Kopeliovich et al., NPA 740(04)211] + Prehadron production time tp = time at which gluon becomes decoherent with parent quark At large z 1, Eh quark must be short lived (or radiates too much energy) zh º htp i / (1 zh) 2 Q energy conservation boost virtuality (perturbative scale) Evolution to hadron by path integral formalism usually htp i < RA hth i À RA 59

60 3) pt broadening [A.A., nucl th/ ] Let's assume no energy loss for a moment: hp2t i / htp i ¼ C zh0:5 (1 zh ) º It should: 1) rise with A1/3, then level off JLAB preliminary, Will Brooks, Trento workshop 60

61 3) pt broadening [A.A., nucl th/ ] Let's assume no energy loss for a moment: hp2t i / htp i ¼ C zh0:5 (1 zh ) º It should: 1) rise with A1/3, then level off 2) decrease as zh 1 3) rise with then level off de s = N c pt2 dx?? pt broadening Elab = 5 GeV HERMES prelim. 61

62 3) pt broadening [A.A., nucl th/ ] Medium enhanced DGLAP evolution? (see talk by K.Zapp) [Ceccopieri et al. PLB'08; Armesto et al. JHEP'08; Domdey et al. arxiv: ] [Domdey et al., arxiv: ] NLO effects? struck q at large xb (large Q2) struck qq¹ at small xb (small Q2) Colored dipoles with tp~0 tcn~(1 z)?? 62

63 4) Correlations between xb and Q2 [A.A., nucl th/ ] For example: if Lund string model for tp is valid, Deviations from such scaling are going to expose the underlying physics: 63

64 Two-particle correlations at HERMES Double hadron attenuation R2 in A+A = same side correlations, (akin to jet yield on top of ridge) e * q e h1? R2 (z2 ) = N2 (z2 ) N1 A N2 (z2 ) N1 D z2 z1 ; z1 0:5 h2 HERMES PRL96(06)

65 Two-particle correlations at HERMES Small A depence: surface bias? e * q e h1? h2 E.g., NLO hard scattering LO NLO N2 (z2 ) N1 A R2 (z2 ) = N2 (z2 ) N1 D z2 z1 ; z1 0:5 zilo = Eih =º zinlo = Eih =ºi > zilo more absorption: hard scattering for observed h is close to surface 65

66 1) The A2/3 power law Old thinking: the A2/3 law Energy loss (LPM effect in QCD): 1 RM ~ z> ~ L2 ~ A2/3 Hadron absorption: 1 RM ~ < no. of nucleons seen > ~ L ~ A1/3 WRONG! RM=cA (He,N,Ne,Kr) A2/3 also for absorption models! 1.0 [A.A., et al., NPA 761(2005)67] additional dimensionful scale: prehadron production length htp i neutralize it additional power of A + HERMES 1 RM / A1=3 (RA =htp i)n = A(1+n)=3 typically A2/3!! [A.A.,EPJC 2007] A dependence of RM does not test dominance of partonic or prehadronic physics: no info on parton lifetime 66

67 Initial state parton energy loss in p+a / A+A Initial & final state cold energy loss in p+a / A+A : p+a (A+A) Vitev PRC'07 DY ndis Initial state energy loss can be large [Vitev PRC'07] test models against DY data (but beware nuclear effects in target w.f.) Needs unified energy loss formalism for DY, ndis, h+a 67

68 Formation time estimates 1 energy loss models Twist 4 modified Fragmentation Fns. [Wang&Guo '00, Wang & Wang '02] h h h Quark energy loss à la BDMPS [Arleo '02] E loss stops at t* pure E loss z 68

69 Effect of medium geometry approximations Example: energy loss model with SW quenching weights Fixed vs. variable path length: change in slope variable p.l. describes data, fixed L too steep Thick vs. finite size medium: changes q^ 2 thick: q^ ¼ 0:2 GeV =fm f.s.: q^ ¼ 0:6 GeV2=fm Correct geometry needed at qualitative & quantitative levels 69

70 The EIC jet physics First time for jet physics in e+a map out observables as a function of parton energy Tests of energy loss models: e.g., modification of jet shapes in cold nuclear matter [Borghini, Wiedemann, '06] 1/ΝJet dn/d [from N. Bianchi] EJet=100 GeV LHC ξ=ln(ejet/phadron) light quark jets vs. heavy quark jets vs. gluon jets dijets, jet correlations,... 70

71 The EIC small x Increased production of heavy flavors heavy quarks D, B mesons heavy quark puzzle at RHIC back to back partons away side correlations: hadron hadron, hadron Wicks et al., nucl th/ electrons RAA J/ normal suppression J/ ' suppression pattern theoretically and experimentally cleaner in e+a pt [GeV/c] 71

72 2 Q range The EIC large Q range Access to true perturbative QCD regime Color transparency Q2 dependence of mentioned observables is pt broadening going to plateau at large Q2? Super fast quarks in nuclei: DIS regime at xb 1 exotic mechanisms short range nucleon correlations 6, 9,..., n quark bags... CLAS, PRL96(2006) Quasi elastic regime 72

73 2 W The EIC large W Heavy mesons from fragmentation, large rate vs. attenuation (no difference at RHIC) low is heavier, hadronizes earlier high same valence, same partonic effects? role of Q2 (at RHIC, Q2 ~ GeV2) extend to strange / charm sector PHENIX, PRC75(2007) Baryons from fragmentation study baryon transport investigate baryon anomaly seen in fixed target e+a, in p+p through A+A needs a good variety of baryons p,, strange and charmed,... 73

74 2 W The EIC large W Heavy quark energy loss and hadronization of D, B mesons in the spotlight at RHIC, big deal at LHC measure D, B in e+a! At HERMES luminosity is too low for D meson At Jlab 12 GeV high luminosity may compensate for low and large x (and PID) chances for D meson measurement close to but not zero electrons RAA Wicks et al., nucl th/ pt [GeV/c] Needs an Electron Ion Collider! 74

75 Results Elab = 27 GeV HERMES HERMES Kr ( )>0 : Formation time scaling for pions! Why best(h ) ~ 0 on Kr? proton anomaly! 75

76 2) Scaling analysis - example A.A., PLB B649 (07)

77 Results Elab = 12 GeV HERMES pions are still positive! confirms results at 27 GeV best(h+) ~ 0 but best(h ) > 0 proton anomaly hypothesis confirmed! 77

78 3) pt broadening q h* h q h* h A A low z high z <pt2> broadening [Kopeliovich et al., NPA 740(04)211] 1) Directly proportional to quark's in medium path 2) Can measure prehadron formation time t* 3) Detect hadronization inside or outside the nucleus dipole x sect. where: Can be cross checked by the scaling analysis of RM 78

79 3) pt broadening Model independent measurement of <l*>=lp [Kopeliovich,Nemchik,Schmidt, hep ph/ ] where dipole cross section 1) fit lp to data for each nucleus 2) determine C by minimizing differences of lp among nuclei C=