Investigating Hard Splittings via Jet Substructure in pp and Pb-Pb Collisions at s NN − − − √ =5.02 TeV with ALICE
IInvestigating Hard Splittings via Jet Substructure in ppand Pb–Pb Collisions at √ s NN = . TeV with ALICE
Raymond Ehlers ∗ on behalf of the ALICE Collaboration Oak Ridge National Laboratory
E-mail: [email protected]
Jets lose energy as they propagate through the Quark-Gluon Plasma, modifying their partonshower. Jet substructure, which provides access to the evolution of jet splittings, is expected tobe sensitive to interactions between the medium and the jet, providing the opportunity to furtherconstrain both jet and medium properties. By utilizing grooming techniques, we can focus on themost pertinent hard splittings. Of particular interest is the search for large transverse momentumkicks which may indicate the presence of point-like scatters within the Quark-Gluon Plasma. Weexplore the jet substructure of inclusive jets in pp and Pb–Pb collisions at √ s NN = .
02 TeV,utilizing Soft Drop and other grooming methods, as well as the Lund Plane, in order to access thehardest jet splitting, with a particular focus on the hardest k T splitting. HardProbes20201-6 June 2020Austin, Texas ∗ Speaker © Copyright owned by the author(s) under the terms of the Creative CommonsAttribution-NonCommercial-NoDerivatives 4.0 International License (CC BY-NC-ND 4.0). https://pos.sissa.it/ a r X i v : . [ nu c l - e x ] S e p nvestigating Hard Splittings via Jet Substructure with ALICE Raymond Ehlers
1. Introduction
As partons from high momentum transfer processes propagate through the medium, theyinteract with it, losing energy and modifying their parton shower. These interactions between thejet and the hot and dense QCD medium known as the Quark-Gluon Plasma (QGP) are expected tomodify the internal jet structure. Jet substructure measurements provide access to jet splittings, andconsequently may be sensitive to these modifications.To perform such measurements, selections are often made on the jet splitting properties viagrooming techniques [1–3]. In pp collisions, grooming limits contamination of the jet shower by softQCD processes, while in Pb–Pb collisions, grooming helps select the hard component of quenchedjets. Utilizing these techniques, substructure may provide direct access to medium properties suchas color coherence [4], or quasi-particle structure which can be searched for indirectly by lookingfor large angle Moliere scattering [5].ALICE [6] is well suited for performing jet substructure measurements due to the precisiontracking provided by the Inner Tracking System and Time Projection Chamber in the central barrel.For these measurements, charged-particle R = . k T jets were reconstructed using FastJet 3.2.1[7] in pp and Pb–Pb collisions at √ s NN = .
02 TeV that were collected in 2017 and 2018 respectively.Jets were required to be contained entirely within the ALICE central barrel acceptance. In Pb–Pbcollisions, background subtraction is of particular importance. For substructure analysis, ALICEperforms background subtraction via Constituent Subtraction [8]. Performance was optimizedwith the goal of reducing the background contribution while minimizing any possible bias on thesubstructure variables. These studies determined an optimal value of ∆ R max = .
2. Groomed Jet Substructure in 30–50% Pb–Pb Collisions
To characterize jet substructure in 30–50% semi-central Pb–Pb collisions, the Soft Dropgrooming algorithm [1] was utilized to select the first sufficiently hard splitting. In particular,we measured the shared momentum fraction, z g , the angular separation between the subjets fromthe selected splitting, R g , and the number of splittings until finding the hard splitting, n SD [1, 3].In order to avoid background contaminated splittings, splittings were considered sufficiently hardwhen they passed the requirement of z cut = . z cut = .
4. For each variable, Bayesian iterative2D unfolding was utilized to correct for background fluctuations and detector effects [9].The results of this analysis for jets measured within 60 < p Tch,jet <
80 GeV / c are shown inFig. 1 and Fig. 2. In the left panel of Fig. 1, z g measured in Pb–Pb collisions is compared tothe same measurement in pp collisions for z cut = .
2. Within experimental uncertainties, z g isconsistent with no modification. n SD is shown in the right panel of Fig. 1, and is also consistentwith no modification relative to pp collisions. The left and right panels of Fig. 2 show R g measuredin Pb–Pb and pp collisions for z cut values of 0.2 and 0.4, respectively. Both panels show similarbehavior, with small angle splittings enhanced in Pb–Pb collisions, while large angle splittings aresuppressed. The measurements were tested for consistency with no modification of the ratio fromunity by adding the statistical and systematic uncertainties in quadrature. Within the context of thissimple metric, both measurements were found to be inconsistent with no modification ( p = . nvestigating Hard Splittings via Jet Substructure with ALICE Raymond Ehlers g z g z d s d j e t, i n c s pp 50% - Pb 30 - PbSys. uncertainty
ALICE Preliminary = 5.02 TeV NN s T k Charged jets anti-| < 0.5 jet h = 0.4, | R c < 80 GeV/
T, ch jet p
60 < = 0 b = 0.2, cut z Soft Drop = 0.88 tagged AA f = 0.89, tagged pp f g z pp P b - P b ALI-PREL-352054 SD n S D n d s d j e t, i n c s pp 50% - Pb 30 - PbSys. uncertainty
ALICE Preliminary = 5.02 TeV NN s T k Charged jets anti-| < 0.5 jet h = 0.4, | R c < 80 GeV/
T, ch jet p
60 < = 0 b = 0.2, cut z Soft Drop SD n pp P b - P b ALI-PREL-352059
Figure 1:
Measurement of z g (left) and n SD (right) for R = . < p Tch,jet <
80 GeV / c in pp and Pb–Pb collisions. Both measurements are consistent with no modification withinexperimental uncertainties. et al [4]. Given the experimental uncertainties, these measurements may have the potential toprovide differentiation between model settings and insight into color coherence.
3. Hardest k T in pp and Pb–Pb Collisions Beyond measurements of the substructure variables themselves, can jet substructure be usedas a tool to isolate the effects of jet-medium interactions? To address this question, we considerthe search for the presence of medium scattering centers via the measurement of rare, wide anglescattering relative to the trigger jet axis, known as Moliere Scattering [5]. Searches by ALICEusing large-angle hadron-jet decorrelation at √ s NN = .
76 are consistent with no medium-induced g R g R d s d j e t, i n c s g q pp 50% - Pb 30 - PbSys. uncertainty
ALICE Preliminary = 5.02 TeV NN s T k Charged jets anti-| < 0.5 jet h = 0.4, | R c < 80 GeV/
T, ch jet p
60 < = 0 b = 0.2, cut z Soft Drop = 0.88 tagged AA f = 0.89, tagged pp f g R pp P b - P b JETSCAPE, MATTER+LBT (Prel.) = 0 res L Pablos et al., T p = 2/ res L Pablos et al., ¥ = res L Pablos et al.,
ALI-PREL-353300 g R g R d s d j e t, i n c s g q pp 50% - Pb 30 - PbSys. uncertainty
ALICE Preliminary = 5.02 TeV NN s T k Charged jets anti-| < 0.5 jet h = 0.4, | R c < 80 GeV/
T, ch jet p
60 < = 0 b = 0.4, cut z Soft Drop = 0.56 tagged AA f = 0.58, tagged pp f g R pp P b - P b JETSCAPE, MATTER+LBT (Prel.) = 0 res L Pablos et al., T p = 2/ res L Pablos et al., ¥ = res L Pablos et al.,
ALI-PREL-353397
Figure 2:
Measurement of R g for z cut = . z cut = . R = . < p Tch,jet <
80 GeV / c in pp and Pb–Pb collisions. Both values of z cut show enhancement for small anglesplittings, as well as suppression for large angle splittings. The models are described in the text. nvestigating Hard Splittings via Jet Substructure with ALICE Raymond Ehlers n split / N j e t s d N / d n s p li t ALICE SimulationPYTHIA8 √ s = 5.02 TeVAnti- k T charged jets R = 0.4, | η jet | < < p partT,ch jet <
80 GeV / c Leading k T Leading k T z > k T DroptimeDrop
ALI-SIMUL-352437 n split / N j e t s d N / d n s p li t ALICE SimulationPYTHIA8 √ s = 5.02 TeVAnti- k T charged jets R = 0.4, | η jet | < < p partT,ch jet <
80 GeV / ck partT > / c Leading k T Leading k T z > k T DroptimeDrop
ALI-SIMUL-352442
Figure 3:
Measurement of the number of splittings until the hardest splitting is identified n split for inclusive k T (left) and k T > / c (right) for R = . < p Tch,jet <
80 GeV / c in PYTHIA8 Monash 2013. The splittings selected by the different grooming methods converge at high- k T . acoplanarity of recoil jets within measurement uncertainties [11]. As an alternative, we investigatethe possibility of using jet substructure as a tool to search for these medium scattering centers. Assubjets propagate through the medium, they may be deflected by a scattering center, which shouldlead to an increase in the yield of high- k T splittings in Pb–Pb collisions relative to pp collisions.In order to identify the hardest k T splitting, we investigated four methods: leading k T , leading k T for all z > . a = k T Drop) and a = k T selects the maximum k T splitting from all availablesplittings, while the z > . k T out of all splittings with z > . κ ( a ) = z i ( − z i ) p T,i ( ∆ RR ) a , where i identifiesa particular splitting, to determine the hardest splitting [2]. All methods consider all iterativesplittings, following the leading subjet to the next splitting.To initially study these grooming methods, they were applied to PYTHIA 8 Monash 2013[12] at particle level. The performance was characterized through properties such as the number ofsplittings until the hardest splitting is identified, n split , as shown in Fig. 3. These studies demonstratedthat although the grooming methods perform differently at low k T , for sufficiently high k T splittings(here, k T > / c ), all grooming methods converge, selecting the same splittings.With these comparisons in mind, the four grooming methods were applied to measure thehardest k T in pp collisions at √ s = .
02 TeV, as shown in Fig. 4. Bayesian iterative 2D unfoldingwas again employed. For k T > / c splittings, the k T spectra converge for all of the groomingmethods. This behavior is consistent with the earlier PYTHIA studies. Each measurement was alsodirectly compared to PYTHIA 8 Monash 2013 by applying the same grooming methods. PYTHIAis broadly consistent with the data within the statistical and systematic uncertainties, although thereis a hint of a shape difference that is consistent between all grooming methods.In order to assess the prospects for measuring the hardest k T in Pb–Pb collisions, we studiedthe correlation between the hardest k T in the PYTHIA splitting graph vs that which is found viadeclustering R = . k T , but a strong4 nvestigating Hard Splittings via Jet Substructure with ALICE Raymond Ehlers − − / N j e t s d N / d k T ( G e V / c ) − ALICE Preliminarypp √ s = 5.02 TeVAnti- k T charged jets R = 0.4, | η jet | < < p chT,ch jet <
80 GeV / ck T DroptimeDropLeading k T Leading k T z > k T (GeV / c ) D a t a L e a d i n g k T ALI-PREL-352215
Figure 4:
Measurement of the hardest k T splitting for four grooming methods for R = . < p Tch,jet <
80 GeV / c in pp collisions and PYTHIA 8 Monash 2013. The splittings selected by thedifferent grooming methods converge at high- k T . PYTHIA is broadly consistent with the data. correlation is observed at high- k T , encouraging the possibility of such a measurement in Pb–Pb.
4. Conclusions and Outlook
We presented fully unfolded z g , n SD , and R g measurements in 30–50% semi-central Pb–Pb andpp collisions at √ s NN = .
02 TeV. z g and n SD are consistent with no modification in Pb–Pb collisionsrelative to pp collisions, while R g shows enhancement for small angle splittings and suppression forlarge angle splittings. These modifications are consistent for both z cut = . k T splittings were measured in pp collisions for a variety of grooming methods. The groomingmethods selected a consistent set of splittings for k T > / c . The prospects for measuring thehardest k T splittings in Pb–Pb were also explored as a step towards applying jet substructure as toolto search for point-like scattering centers in the medium via large angle scattering. References [1] A.J. Larkoski, S. Marzani, G. Soyez and J. Thaler
JHEP (2014) 146 [ arxiv:1402.2657 ].[2] Y. Mehtar-Tani, A. Soto-Ontoso and K. Tywoniuk Phys. Rev. D (2020) 034004 [ arxiv:1911.00375 ].[3] S. Acharya et al.
Physics Letters B (2020) 135227 [ arxiv:1905.02512 ].[4] J. Casalderrey-Solana, G. Milhano, D. Pablos and K. Rajagopal
JHEP (2020) 44 [ arxiv:1907.11248 ].[5] F. D’Eramo, M. Lekaveckas, H. Liu and K. Rajagopal
JHEP (2013) 31 [ arxiv:1211.1922 ].[6] ALICE collaboration
Int. J. Mod. Phys. A (2014) 1430044 [ arxiv:1402.4476 ].[7] M. Cacciari, G.P. Salam and G. Soyez Eur. Phys. J. C (2012) 1896 [ arxiv:1111.6097 ].[8] P. Berta, L. Masetti, D.W. Miller and M. Spousta JHEP (2019) 175 [ arxiv:1905.03470 ].[9] T. Adye arXiv arxiv:1105.1160 ].[10] JETSCAPE collaboration arXiv arxiv:1903.07706 ].[11] ALICE collaboration
JHEP (2015) 170 [ arxiv:1506.03984 ].[12] T. Sjöstrand et al. Comput. Phys. Commun (2015) 159 [ arxiv:1410.3012 ].[13] Y. Chen
EMMI RRTF Workshop (2019) .(2019) .