Characterising Charm Jet Properties with Azimuthal Correlations of D Mesons and Charged Particles with ALICE at the LHC †

Charm quarks are produced via hard parton scattering in ultra-relativistic heavy-ion collisions, hence are ideal probes to study a possible de-confined state of matter, known as Quark Gluon Plasma (QGP). The angular correlation of a meson containing a charm quark with other charged particles in heavy-ion collisions can help in studying the properties of QGP. Similar studies in pp collisions can give insight about the charm production mechanism while in p-Pb collisions could provide essential information to disentangle final-state QGP-induced modifications from effects caused by cold nuclear matter. In this proceedings, the results are presented for p-Pb collisions at √ sNN = 5.02 TeV and pp collisions at √ s = 13 TeV, so far the highest available energy at the LHC. The results are compared with Monte Carlo (MC) simulations using PYTHIA and POWHEG event generators and with pp collision results at √ s = 7 TeV.


Introduction
The azimuthal correlation of a D-meson with charged particles in pp collisions can give insight into the charm production mechanisms.There are three main charm production mechanisms: pair production, flavor excitation, and gluon splitting [1].The pair production is a leading order [LO] process while flavor excitation and gluon splitting are next to leading order [NLO] processes as shown in Figure 1.The correlation of D-meson in pp collisions as shown in the center of Figure 2 can constrain the perturbative QCD inspired Monte-Carlo (MC) models, as well as serve as a reference for p-Pb and Pb-Pb collisions.In p-Pb collisions, we have initial state effects like shadowing, k T broadening (Cronin effect), gluon saturation, and energy loss in cold nuclear matter while in Pb-Pb collisions we have final state effects due to the medium formation together with initial state effects [2,3].The study of azimuthal correlation in p-Pb collisions is necessary to disentangle final state effects (energy loss, fragmentation, and recombination) from the initial state effects.In addition, it also allows searching for long-range double-ridge structure in the heavy-flavor sector as shown in Figure 2 [3].ALICE has performed measurements of D-hadron correlation in pp collisions and p-Pb collisions at √ s = 13 TeV and √ s NN = 5.02 TeV, respectively with ALICE at the LHC.We have compared the near-side (NS) and away-side (AS) jet properties for pp and p-Pb collisions and also compared the data with perturbative QCD inspired MC models, PYTHIA [LO] [4,5] and POWHEG [NLO] [6].The study of the correlation in pp collisions at √ s NN = 5.02 TeV is ongoing and it will be of primary importance to assess possible modifications of the correlation peak due to the presence of cold nuclear matter effects.

Analysis Details
The study is performed on pp 2017 and p-Pb 2016 Run 2 data at √ s = 13 TeV and √ s NN = 5.02 TeV, respectively, collected by the ALICE detector at the LHC [8].In the analysis, we used following sub-detectors of ALICE, the inner tracking system (ITS) for tracking and reconstruction of primary and secondary vertices, time projection chamber (TPC) for tracking and particle identification, and time of flight detector (TOF) for particle identification.The D-meson (D + , D 0 , and D * + ) reconstruction from its daughters as given in Table 1 is done using topological cuts and particle identification information to reduce the combinatorial background.An invariant mass analysis is performed in order to separate the signal (Gaussian fit) and background (exponential fit) as shown in the left panel of Figure 2. D meson candidates are correlated with primary charged particles (e, µ, π, K, and p) by calculating ∆η = η trigger − η assoc and ∆φ = φ trigger − φ assoc in the same event.The correlation of signal D-meson is extracted by using the method of side-band subtraction as given by the expression below.
where signal correlation is the correlation of the true D-meson with the charged particles.SF is the scale factor calculated from the ratio of the yield in the background region and side-band (left LSB, right RSB) region and is given by B 0 /(B L + B R ).The distribution is corrected for the detector acceptance and inhomogeneity by using mixed-event techniques, in which we mix the event with the same z-vertex and multiplicity.The majority of secondary particles is removed by using distance of closest approach cuts, DCA xy and DCA z .The distribution is also corrected for secondary track contaminations from strange decays and conversion inside the detector material by evaluating the fraction of accepted secondary tracks from MC simulation.The distribution is further corrected for the D-meson reconstruction efficiency, tracking efficiency, bias induced due to B-hadron to D-hadron decay topology.The final correction is the subtraction of contribution from the D-meson coming from the B-hadron decay.Finally, we study the per trigger normalized correlation distributions of D-mesons with primary charged particles.These fully corrected correlation distributions are projected on ∆φ axis for the interval |∆η| < 0.9.Next, the weighted average of all three D-mesons is done.In order to reduce the statistical uncertainty the reflection from [− π 2 , 3 π 2 ] to [0, π] is done as shown in the left panel of Figure 3. Finally, we fit the ∆φ distribution by the following equation: Equation ( 2), which include the near-side (NS) peak, away-side (AS) peak, a baseline and together with a double Gaussian [9].

Results
The

Conclusions
In p-Pb collisions, the yield of associated charged particles increases with the p T of D-mesons while the Gaussian width of the near-side peak decreases.The comparison of p-Pb data with pQCD inspired MC models like PYTHIA [LO] and POWHEG [NLO] shows that the data is compatible with models within uncertainties.Finally, the measurement in p-Pb collisions is also compared with that in pp collisions and within uncertainties both measurements are consistent.

Figure 1 .Figure 2 .
Figure 1.Charm production mechanisms: (a) gluon fusion, (b) quark annihilation, in the pair production (LO), the heavy quarks are back to back at 180 • , (c) flavor excitation (NLO), the heavy quarks have a large opening angle, (d) gluon splitting (NLO), the heavy quarks have a very small opening angle [1].

Figure 3 .
Figure 3. (Left) Example of reflected correlation distribution fitted with a double Gaussian and a constant baseline and (Right) comparison of NS yield and Gaussian width (σ NS ) with the MC models.

Figure 4 .
Figure 4. Comparison of NS yield and Gaussian width (σ NS ) for pp data at √ s = 7, 13 TeV and p-Pb data at √ s NN = 5.02 TeV.