Jet-related 2-particle ( - PowerPoint Presentation

Jet-related 2-particle (g-h,h-h) Correlations Recent Developments Justin Frantz BNL 4/14/16 Justin Frantz Ohio University HiPt LhcRhic 16 1

Using 2-p Correlations for Eloss As a tool for studying jet Eloss, “high” (mid) pT 2PC remain importantAllows for study of lowest Q2 possible (a la sPHENIX ) Ejet= 5-20 GeV (Jet Reco systematics large there )Justin Frantz Ohio University HiPt LhcRhic 162 QM‘14 2 p 0 -h QM‘15 STAR CMS 2014

Life At The Hard/Soft Interface Flow subtraction necessary: many 2ppractictioners current focus Great progress at softest end: understanding flow bkg contributionshigher orders v2-v9, corr btw orders, etc…Latest: considering .vs rapidity dependences Still to finish: v1 !!!!At med/high pT , how to applyknowledge still under developmentJustin Frantz Ohio University HiPt LhcRhic 16 3

v1: Unknown Known?ATLAS/ J.Jia: 2012 : Demonstrated successful description of dipolar (h symmetric ) v1: as fluctuations/HYDRO + Global MOMENTUM CONSERVATION 2 component model: GMC + HYDRONeeds applied more widely—RHIC ( h symmetric) v1 measurements!Subtract at least HYDRO part from 2p correlations ATLAS data

Hydro v1 resultsThis should be included in subtractions!NOTE: PROCEDURE RELIES ON HYDRO “FACTORIZATION” HOLDING– NEEDS REEVALUATED!STAR result similar with different method (robust?) Justin Frantz Ohio University HiPt LhcRhic 16 5

Study GMC further : “Medium Recoil vs. Jet Recoil? Borghini nucl-th/0612093v1 Project certain Df 13 regions in 2+1 type Df12 projection ? Df 13 >pi < pi ? Subtract GMC ? Isn’t that part of quenching effect? Yes but not all of GMC is quenching Constraints from other observables? HBT Natural Asymm in 3-p correlation from GMC:

Direct Photon vs Di-hadronsAmongst other “Golden” advantages: Underlying event subtraction easier for Direct g !v_n of (iso) prompt g should be ~ 0 However, explicit iso cut introduces non-zero v2! And v_n mostly cancels in g subtraction 7 8/14/2012 Justin Frantz - PHENIX PHENIX 2013 Result

Outline:Direct Photon - Hadron Correlations OutlookDi-Hadron CorrelationsUsing case of small system Eloss as exampleJustin Frantz Ohio University HiPt LhcRhic 16 8

Enhancement at Low zWe know there’s enhancement at low z  g-h before h-h!LHC FF’s (note Normilization ! high z/pT = ~1 )Interpreted as recovered Eloss E2pc IAA Enhancement should be measured relative to suppressed high z Justin Frantz Ohio University HiPt LhcRhic 16 9 PLB 739 (2014) 320–342 PHENIX PRL 111, 32301 0-20% Au+Au g -h I AA

Example: No Significant Enhancement in Small Cones? Judging enhancement only by IAA > 1 is misleadingNo enhancement in the core? Justin Frantz Ohio University HiPt LhcRhic 16 10 Integration Region (~Cone size)

Example: No Significant Enhancement in Small Cones? Judging enhancement only by IAA > 1 is misleadingMust look at size of IAA relative to suppressed jet level—enhancement in all parts, “jet core” included Justin Frantz Ohio University HiPt LhcRhic 1611 Integration Region (~Cone size)

My take: last talk: STAR g-h resultsQM 15! Now on arXivVery nice even higher statistics gamma and pi0Rise/Enhancement above high z value present: Finally! Justin Frantz Ohio University HiPt LhcRhic 1612 Qualitative Agreement!

PHENIX PRL 111, 323010-20% Au+Au g-h I AA Disagreement w/ PHENIX?? Qualitatively similar (rise)But sizable disagreement @ low z if expect FF AuAu/pp universal?Very different jet E & Min pt Difference 1.2 STAR vs 0.5 PHENIXSignificant CONCLUSION: First evidence of eitherNon FF mod (IAA) universality: pT matters not z (medium scale?) OR: Eloss for low Ejet different! Justin Frantz Ohio University HiPt LhcRhic 16 13 13 PHENIX: g p T 5-9 GeV/c p T assoc ~ 0.8 GeV/c

LHCJustin Frantz Ohio University HiPt LhcRhic 1614 p T jet  150 GeV pT jet  100 GeV Universal as a fn of pT assoc?but z only changes slightly

NEXT FOR Gamma-h STAR/PHENIX: same pt binning, check(Centrality was different too: 0-40 vs 0-12)What’s next? LHC  Moderate pT trig (~5-15 Gev ) esp w/ low pT (assoc)ISOLATION CUT (already at higher p T gamma-jet)Important for Centrality Dep Justin Frantz Ohio University HiPt LhcRhic 16 15 JF/ Student Nowo Riveli THESIS RESULT ERROR SIZE DEMO (NOT DATA!) g iso –h

INTERESTING Complication ISO vnIso Cut ON vn Underlying Event induces “NEGATIVE vn”vn needed for gam-h UE subtraction (n>2 small though?)So this must be accounted for (STAR include iso?): Derivations for n=2 by JF and Student Tyler Danley: on arXiv soon Justin Frantz Ohio University HiPt LhcRhic 16 16 v2 Underlying Event (e.g. g iso ) No RXNP Resolution Results! v 2E can -v 2T !!!!   “v 2E ”

INTERESTING Complication ISO vnw/ RXP Resolution Included Distributions Even more interesting Wide range Cone sizes / v2’s (Central Periph) weird cases possible Justin Frantz Ohio University HiPt LhcRhic 16 17 v 2E can -v 2T !!!!   NOW WITH Reaction Plane ( Resolution   TA ( g -h) distribution unaffected by , of course  

Di-Hadronsstarting in d+AuJustin Frantz Ohio University HiPt LhcRhic 1618

Switch to p+A h-h Medium? Eloss?We’ve seen the soft observables: The obvious associated question: is there Jet Eloss? LOOK @ INTERMEDIATE PT E jet = [5-15] h-hSingle particle/jet R pA / RdA suffer from “ Ncoll” systematicsWarning: can’t conclude there are no jet Eloss effects in p+A, without knowing the size we would expect given A+A Eloss!Justin Frantz Ohio University HiPt LhcRhic 1619 Zakarov 1311.1159 Lots of things to be resolved: Cold Effects +Hot Can we see more studies like this?

How small could Eloss in pA be?…in 2PC: Suppression AND EnhancementSimple (upper bound) emprical estimate: assume suppression and enhancement scale with ~ L N part1/3Suppression= 1- Npart 1/3/3501/3 * R AA (high pt jet) = 0.4 Enhancement: 1+ Npart 1/3/3501/3 * (low z) = 0.5In Central d+Au: Supp = 0.9 Enh = 1.1   Justin Frantz Ohio University HiPt LhcRhic 16 20

IAA in DAU2014 re-analysis:Statistics allow a view of centrality dependenceBut efficiency systematics are still too big to resolveJustin Frantz Ohio University HiPt LhcRhic 16 21

Qualitatively Similar Features to AARun 8 high precision data allows us to see delicate featuresEnhancement/SuppressionEfficiency Systematics not shown  same (const ) for all   Justin Frantz Ohio University HiPt LhcRhic 16 22

NS/AS Ratios: A Nice Observable for searching for Eloss in small systems?Assume well-known surface bias picture for Au+Au should apply to a smaller d+AuLook for Differences in Awayside Modification compared to Nearside Justin Frantz Ohio University HiPt LhcRhic 16 23 p+A ? trigger A+A plasma Black Red

Not just small systems:As we go towards very peripheral in A+A this is a good observable (traditional vn subtraction becomes hard) I am convinced this is an excellent observable to pursue for heavy ion energy loss studies in the medium pt regionSeveral cancellations of systematic errorsAs long as surface bias (little NS modification), main energy loss features are still very present .And/OR more sensitive to v1 Justin Frantz Ohio University HiPt LhcRhic 16 24

PLB 739 (2014) 320–342 IAA  RI AuAu : h-h Enhance/SuppressExample Au+Au IAA before v3 subtracted (but v3 does not change the (large!) qualitative trendFull subtraction results soon from PHENIXJustin Frantz Ohio University HiPt LhcRhic 1625 Awayside Yields PHENIX PRL ppg83 Trigger h : OLD DATA NOT VN>2 SUBTRACTED!

CMS RESULT PbPbForgotten (?) CMS result did apply full vn subtractionREALLY HIGH PT TRIG vs LO&HI PTA GIANT enhancement Low z; Suppression = JET RAAJustin Frantz Ohio University HiPt LhcRhic 16 26

RI in Central dAu EventsReminiscent of Au+Au behavior but much smaller.RI uncertainty cancellations allows statistical significance for small effectChi2 of constant fit is large: many sigma significance Justin Frantz Ohio University HiPt LhcRhic 16 27 Remember Look here

“Trivial” Causes?Maybe it’s not Eloss, but in any case what is it?!Strategy: address all “trivial” explanations we can test:“Hydro” v3, v1Enhanced Nuclear kTInitial State nPDF effectsTrivial Rapidity Distributions Mismatching p+p vs d+Au? HIJING show anything like this?If none of above  INTERESTING Looking for other ideas? Justin Frantz Ohio University HiPt LhcRhic 16 28

“Soft” (Hydro?) v1, v3?Because we divide full AS (Dphi pi/2) by full NS even harmonics such as v2 cancel perfectly after ZYAMin AS/NS yields themselves ZYAM level systematic evaluated, given separatelyFor odd harmonics, like v3 there is a residual affecting RI=AS/NSJustin Frantz Ohio University HiPt LhcRhic 16 29 v3a= 0.5, v3b = 0.1 AS/NS = 1.00425

Estimate for v3[Remember other than the recent PHENIX He, there are no RHIC d+Au v3 measurements]Thus we estimate it as being the same size as at the LHC. (Judging by 200 GeV He+Au probably overestimate! )Justin Frantz Ohio University HiPt LhcRhic 16 30 0.8 0.6 0.4 0.2 0.0 d+Au Pi0-h Correlation Fn’s 5-7 x 0.5-1 5-7 x 1-2 ~ 1% Effect   Still this is still d+Au & high pt  jet still has good S/B (included in RI systematics)

Positive v 1v 1 is Stickier We don’t need to consider higher odds because “residuals” become small due to oscillations and the v n’s get even smallerHowever, if we use LHC as estimate, v1 may not be negligible: Big enough positive v1 would “subtract” lowering AS/NS Low pt would make enhancement bigger than measured!But big enough negative v1 at low pt would go in right direction for enhancement! Justin Frantz Ohio University HiPt LhcRhic 1631 Negative v 1

Negative v1 @ Low pT, You Ask?In fact a negative v1 is indeed observed in this pt region at the LHC by ATLAS Justin Frantz Ohio University HiPt LhcRhic 16 32

Implication for dAu HYDRO v1 effect Using the ATLAS estimate of the p+Pb hydro part (similar size to v3 and v2, same for Pb+Pb) about half of the current total systematics i.e. fairly negligibleCurrently no extra systematic or correction for thisGlobalMomConsv and/or jets responsible? if jets  Eloss?GMC: hard to disentangle from the jets, if GMC is modified separately it would be interesting!GMC would still not explain the suppression at high zT REMOVAL OF GMC MAKES SUPPRESSION LARGERJustin Frantz Ohio University HiPt LhcRhic 16 33 ~2-3 % RI v1 (low p T )   HYDRO Vn d+Au :

“Trivial” Causes?Maybe it’s not Eloss, but in any case what is it?!Strategy: address all “trivial” explanations we can test:“Hydro” v3, v1Enhanced Nuclear kTInitial State nPDF effectsTrivial Rapidity Distributions Mismatching p+p vs d+Au? HIJING show anything like this?If none of above  INTERESTING Looking for other ideas? Justin Frantz Ohio University HiPt LhcRhic 16 34 GLOBAL MOMENTUM CONTRIBUTIONPOSSIBLE

Enhanced Nuclear kT ?kT == Acoplanarity of di-jetsSmears Awayside  Known part of the 2pc AS width If d+Au had long sought after enhanced nuclear kT, could this cause this effect in RI? Justin Frantz Ohio University HiPt LhcRhic 16 35

PYTHIA Nucl kT TestUsing kT constraints from STAR jet measurements  No effect for 0-100% MinbiasHowever, kT smear larger in Central? Justin Frantz Ohio University HiPt LhcRhic 16 36 Using STAR k T Increase ( Minbias )

“Trivial” Causes? Strategy: address all “trivial” explanations we can test:“Hydro” v3, v1Enhanced Nuclear kTInitial State nPDF effectsTrivial Rapidity Distributions Mismatching p+p vs d+Au?HIJING show anything like this?If none of above  INTERESTING Looking for other ideas? Justin Frantz Ohio University HiPt LhcRhic 16 37

Initial State Nuclear PDF’s?nPDF effects would seem unlikely to cause this, since they probably often affect both jets in a di-jetStudies with EPS09 (and 09s) confirm this expectationNOTE UNITS: << 1% negligible effectJustin Frantz Ohio University HiPt LhcRhic 16 38 RI Extracted from EPS09 Study nPDF extracted from EPS09 code

“Trivial” Causes? Strategy: address all “trivial” explanations we can test:“Hydro” v3, v1Enhanced Nuclear kTInitial State nPDF effectsTrivial Rapidity Distributions Mismatching p+p vs d+Au?HIJING show anything like this?If none of above  INTERESTING Looking for other ideas? Justin Frantz Ohio University HiPt LhcRhic 16 39

HIJINGWe ran HIJING with default settings First, this can test for very trivial effects e.g. due to the 2p method and to the mismatch in rapidity distributionsMore importantly any other “cold” nuclear physics embedded in HIJING (mini-jets, momentum conservation (?), etc…)Justin Frantz Ohio University HiPt LhcRhic 16 40

HIJING RIWith default settings, HIJING does not reproduce the effectSmall enhancement at low zT appears to be due to default nuclear kT in HIJING consistent with PYTHIA studyJustin Frantz Ohio University HiPt LhcRhic 16 41

“Trivial” Causes? Strategy: address all “trivial” explanations we can test:“Hydro” v3, v1Enhanced Nuclear kTInitial State nPDF effectsTrivial Rapidity Distributions Mismatching p+p vs d+Au?HIJING show anything like this? Looking for other ideas?Justin Frantz Ohio University HiPt LhcRhic 16 42 Yes this looks INTERESTING!

Current Status: He+Au?Tried to take a look at He+Au see how effect evolves… STATISTICS GOODIs He+Au much more affected by v1 hydro (GMC?)?Consider “Geometric Eccentricities” forget v3: v1 will be why HeAu/CuAu are most important! Justin Frantz Ohio University HiPt LhcRhic 16 43 "Geo" Epsilon_1 : ~0 "Geo" Epsilon_2 : 0.781028 "Geo" Epsilon_3 : ~0"Geo" Epsilon_4 : 0.446895 "Geo" Epsilon_1 : 0.11299 "Geo" Epsilon_2 : 0.307606 "Geo" Epsilon_3 : 0.0938522 "Geo" Epsilon_4 : 0.0844464He+Aud+AuFluctuations will smear this HeAu /dAu CentralActual Error Size (ONLY) NOT DATA

Glauber v1,v2,v3 He vs dEffect almost completely smeared out in Phobos Glauber?Still higher v1/v3 in HeAu But Depends on details of calc method (gaussian smear,etc…) / Response Justin Frantz Ohio University HiPt LhcRhic 16 44 e2 e3 e1 Impact Parameter ( fm ) Nagle et. al. HeAu : SuperSONIC where’s v 1 ?

PHENIX: v1 He+Au ? New Cu+Au Recent HeAu v2,v3 …. v1?We do have Cu+Au v1: similar ideaJustin Frantz Ohio University HiPt LhcRhic 16 45 ATLAS Unique opportunity to study large v1 systems HeAu pAu CuAu

PHENIX Cu+Au v1 PaperConfusion about AMPT: but obviously has many contributions that can be studied…Be braver Justin Frantz Ohio University HiPt LhcRhic 1646 looks reasonable

Further: dAu New CorrelationsPHENIX QM 15: Takao Sakaguchi  direct v1 input to dAuJustin Frantz Ohio University HiPt LhcRhic 16 47

Further: dAu New Correlations PHENIX QM 15: Takao Sakaguchi Justin Frantz Ohio University HiPt LhcRhic 16 48 v2: cancels in RI completely v1: can check small effect on dAuv1: repeat study for HeAu!!! STAY TUNED!

Conclusions2-p Alive and Well -- Promising future for exploring Eloss for E [5,15]v1 needs studied more, but should start being subtracted…New STAR AuAu gamma-jet confirm PHENIX qualitative results,Interesting CONCLUSIONS from comparison: low Ejet Eloss diff or FragFn Mod function of pt not z-universalCheck if there’s quantitative tension? Iso cut for gamma-h at 5-15 GeV important for centrality dep, introduces complications—but proof of principle By canceling systematics, the RI ratio reveals a modification in the traditionally jet interpreted pT regions that is qualitatively similar to A+A E loss but much smallerNo “easily available” initial state or other “trivial” explanation appears to explain this modificationv1 laboratory RHIC: Cu+Au He+Au Justin Frantz Ohio University HiPt LhcRhic 1649

BackupJustin Frantz Ohio University HiPt LhcRhic 1650

Another PYTHIA study: q vs g jetsGluon jets known to be softer than quark jets Demonstrated below for a couple different processesClass A: All gg final states to qq(qbar) final states _ Class B: qg->qg / qg->gq (Compton with glue on awayside divided by w/ quark on AS) -First think of these curves as g to q Fragmentation Function Ratios-(They are actually RI-like Away/Near double ratios: A:N for gluon AS final states divided by A:N for quark AS jets) PYTHIA Justin Frantz Ohio University HiPt LhcRhic 16 51

Couple ScenariosChanges to the mix of final state jets: even extreme scenarios don’t reproduce the effectShapes also not consistent with effect we see RI ratio dAu:pp assuming 100 % conversion of all final q jets into gluon jets in dAu. RIHowever we note that bigger effects can be generated by changing the mixture in different ways, e.g. of putting higher weight on Compton processes PYTHIA PYTHIA Justin Frantz Ohio University HiPt LhcRhic 16 52