betaFeedback

WorksPhysics and Astronomy

Nature of particles azimuthal anisotropy at low and high transverse momenta in ultrarelativistic A+A collisions

Abstract

LHC data on the correlations of the elliptic flow v2v_2 of particles at low and high transverse momenta pTp_T from Pb+Pb collisions at center-of-mass energy per nucleon pair sNN=5.02\sqrt{s_{NN}} = 5.02 TeV are analyzed in the framework of the HYDJET++ model. This model includes soft and hard components which allows to describe the region of both low and high transverse momenta. The origin of v2v_2 values in different pTp_T regions is investigated at different centralities. It is shown that the experimentally observed correlations between v2v_2 at low and high pTp_T in peripheral lead-lead collisions is due to correlation of particles in jets.

Introduction

The properties of hot and dense matter produced in heavy ion collisions are thoroughly explored during several decades, especially after the start of heavy-ion programs at the Relativistic Heavy Ion Collider (RHIC) at BNL and the Large Hadron Collider (LHC) at CERN. The focus of the field is to investigate Quark Gluon Plasma (QGP) formation and its characteristics at different conditions. Numerous observables of final event serve to this purpose. Here we study one of the main observables in soft physics sector, namely the azimuthal momentum-space anisotropy of particles, and its manifestation in hard physics regime. The correlations between hard and soft contributions to azimuthal anisotropy of particles have attracted much attention, see [1]–[3].

At relatively low transverse momenta, the azimuthal anisotropy occurs due to the anisotropic expansion of the compressed matter, since particles are emitted preferably in the direction of the largest pressure gradient [4]. The anisotropy for particles with high transverse momenta is governed by the energy loss of hard partons traversing the hot and dense nuclear medium. Here more jet particles are emitted in the direction of shortest path length [5]. The sizable azimuthal anisotropy observed at RHIC energies was the main evidence for the nearly perfect liquid behavior of the created matter [6, 7, 8, 9]. The parton energy loss in hot and dense medium, so-called jet quenching, can be investigated in experiment with different observables, such as nuclear modification factor of particle, jets suppression. The azimuthal particle asymmetry at high transverse momenta is yet another observable revealing information on jet quenching process. At LHC, the ATLAS [10] and CMS [11] collaborations have performed measurements of the azimuthal anisotropy of charged particles produced in Pb+Pb collisions at sNN=5.02\sqrt{s_{NN}} = 5.02 TeV up to pT=60p_T = 60 and 100 GeV/c, respectively.

Anisotropic flow is quantitatively characterized by coefficients in the Fourier expansion of the azimuthal dependence of the invariant yield of particles in a form [12, 13]:

Ed3Ndp3=1πd2NdpT2dη×(1+∑n=1∞2vn(pT,η)cos⁡[n(φ−ΨRP)])(1)E\frac{d^3N}{dp^3} = \frac{1}{\pi}\frac{d^2N}{dp_T^2 d\eta} \times\left(1 + \sum_{n=1}^{\infty} 2v_n(p_T,\eta)\cos[n(\varphi-\Psi_{\mathrm{RP}})]\right) \tag*{(1)}

where φ\varphi is the azimuthal angle of particle, η\eta is the pseudorapidity and ΨRP\Psi_{\mathrm{RP}} is the reaction plane angle. The flow coefficients are:

vn=⟨⟨cos⁡[n(φ−ΨRP)]⟩⟩(2)v_n = \left\langle\left\langle\cos[n(\varphi-\Psi_{\mathrm{RP}})]\right\rangle\right\rangle \tag*{(2)}

where the averaging is performed over all particles in a single event and over all events. The initial reaction plane angle, which is defined by impact parameter vector, is not known, hence, the direct calculation of v2v_2 by (2) is not possible. The measurement of collective flow coefficients aims to measure global azimuthal anisotropy, i.e., particle correlations with respect to global event geometry, but the methods of measuring are sensitive to local particle correlations (such as resonance decays, jets, fluctuations and so on). The most advanced experimental techniques, implementing gaps in pseudorapidities between correlating particles, multiparticle cumulant methods and other, suggest suppression of local particle correlations.

The aim of the present paper is to study the correlations between the low-pTp_T and high-pTp_T regions of the elliptic flow of charged hadrons in Pb+Pb collisions at sNN=5.02\sqrt{s_{NN}} = 5.02 TeV at different centralities. These correlations were observed experimentally in [10] and [11]. The origin of the correlations is not understood yet, mainly because of not too many models, such as EPOS [14], QGSJet [15] and HYDJET++ [16] (for recent developments, see also [17] and [18]), which are describing the soft and the hard processes simultaneously. Here we employ the phenomenological model HYDJET++ to explore the behavior of azimuthal particle anisotropies at different transverse momenta and its connection to common global event geometry, and also the feasibility to establish such connection by experimental techniques.

HYDJET++ model

We use the HYDJET++ model [16] for nucleus-nucleus collision simulation which combines soft and hard physics as two components of resulting final heavy ion event. The soft physics is based on relativistic hydrodynamics and is represented by the thermal hadronic state generated on the chemical and thermal freeze-out hypersurfaces with preset freeze-out conditions. It includes the longitudinal, radial and elliptic flow effects and the decays of hadronic resonances. The code of it is based on the adapted version of the event generator FASTMC [19, 20]. The particle multiplicity distributions are Poissonians and their mean multiplicities are estimated within the effective thermal volume. The parameters for the soft part need to be tuned to describe experimental data. For instance, the elliptic flow of the produced particles is governed by the spatial and momentum anisotropy of the fireball. The parameter of the spatial anisotropy, ϵ2(b)\epsilon_{2}(b), regulates the elliptic profile of the final freeze-out hypersurface at a given impact parameter bb. The momentum anisotropy parameter, δ(b)\delta(b), deals with the modulation of the flow velocity profile. Both parameters are linked via the hydro-inspired parametrization [20]. Parameters responsible for spatial and momentum triangular anisotropy are implemented in the model as well, thus giving rise to triangular and other odd harmonics flow.

The basis for hard physics of nucleus-nucleus collision is elementary QCD parton-parton scatterings as it is realized in PYTHIA [21] with additional simulation of parton energy loss in a dense medium by PYQUEN model [22] with subsequent hadronization. Both collisional loss due to parton rescattering and gluon radiation loss are taken into account when propagating hard parton through a medium. The medium is treated as a boost-invariant longitudinally expanding medium at some temperature with the transverse asymmetric geometry given by impact parameter of a collision (initial elliptic shape is considered). The initial temperature of the medium for central collisions is one of the parameters of the model. The PYQUEN routine is used to generate a single hard nucleon-nucleon (NN) collision. To calculate the mean number of mini-jets produced in A+A collision at a certain impact parameter bb one has to (i) determine the number of binary NN collisions in the event and (ii) multiply it to the integral cross section of the hard process. Hardness of the process depends solely on the minimum transverse momentum transfer, pTmin⁡p_{T}^{\min}. Therefore, the parameter pTmin⁡p_{T}^{\min} is one of the major parameters which regulates the contributions of soft and hard particles to total multiplicity. If the transverse momentum of initial hard scattering does not exceed pTmin⁡p_{T}^{\min}, the partons produced in the scattering are excluded from the hard processes. The products of their hadronization are then automatically added to spectrum of hadrons produced in soft processes. Figure 1(a) demonstrates the contribution of the soft and hard components to the formation of elliptic flow in the model. Here the coefficients v2RPv_{2}^{\mathrm{RP}} are calculated with respect to the reaction plane. Hydrodynamic nature of elliptic flow of the soft component (shown by red triangles) provides the increasing v2v_{2} with rising pTp_{T}. For the hard component, non-zero elliptic flow arises at pT>10p_{T} > 10 GeV/cc due to jet quenching effect in asymmetric medium. Figure 1(b) shows the spectra dN/dpTdN/dp_T of both components together with total spectrum. At low transverse momenta the multiplicity of soft particles dominates and one can see that the resulting flow in this region is determined mainly by the elliptic flow of soft component, while at high pTp_T the hard component is dominant and determines total elliptic flow value. In the intermediate pTp_T region the result is obtained by a simple superposition of two independent contributions.

Image not extracted

Figure 1. Color online. (a) The elliptic flow as a function of pTp_T, v2RP(pT)v_2^{\mathrm{RP}}(p_T), for soft (triangles) and hard (squares) components, and resulting total flow (circles) of charged hadrons with ∣η∣<2.4|\eta| < 2.4 in 20–30% centrality bin of Pb+Pb collisions calculated within HYDJET++ at sNN=5.02\sqrt{s_{NN}} = 5.02 TeV. (b) The same as (a) but for the transverse momentum spectra dN/dpTdN/dp_T of charged hadrons for soft and hard components and also the total one.

HYDJET++ model has managed to describe many experimental features and phenomena measured in heavy ion collisions at RHIC and LHC energies. These signals include pseudorapidity and centrality dependence of the multiplicity of charged particles, their transverse momentum spectra, radii of ππ\pi\pi correlations in central Pb+Pb collisions [23], centrality and momentum dependencies of elliptic and higher-order components of the anisotropic flow [24]–[27], flow fluctuations [28], angular dihadron correlations [29], multiplicity correlations of charged particles in forward-backward hemispheres [30], as well as effects of jet quenching [31, 32] and, finally, production of heavy mesons [34, 35] [33].

Calculation of azimuthal anisotropy coefficients

For the correct comparison with CMS data on v2v_2 of charged particles [11] we apply the scalar product method (SP) and four-particle cumulant method [36]. We do not simulate the detector response, i.e., we use charged particles from the model, but apply the same η\eta-cuts as in experimental data. The two- and four-particle correlations for the second order flow harmonic are defined as

⟨⟨2⟩⟩=⟨⟨ei2(φ1−φ2)⟩⟩,⟨⟨4⟩⟩=⟨⟨ei2(φ1+φ2−φ3−φ4)⟩⟩,(3)\begin{aligned} \langle\langle2\rangle\rangle&= \langle\langle e^{i2(\varphi_1-\varphi_2)}\rangle\rangle,\\ \langle\langle4\rangle\rangle&= \langle\langle e^{i2(\varphi_1+\varphi_2-\varphi_3-\varphi_4)}\rangle\rangle, \tag*{(3)} \end{aligned}

where double average means averaging over all particle combinations in an event and over all events in a data sample. The estimator of the reference 4-particle cumulant, c2{4}c_2\{4\}, is defined as

c2{4}=⟨⟨4⟩⟩−⟨⟨2⟩⟩2(4)c_2\{4\} = \langle\langle4\rangle\rangle- \langle\langle2\rangle\rangle^2 \tag*{(4)}

For the differential flow calculation one of the particle in (3) is restricted to belong to a certain pTp_T bin. We denote it by ⟨⟨2′⟩⟩\langle\langle2'\rangle\rangle and ⟨⟨4′⟩⟩\langle\langle4'\rangle\rangle, respectively. The differential 4-particle cumulant reads

d2{4}=⟨⟨4′⟩⟩−2⟨⟨2′⟩⟩⟨⟨2⟩⟩.(5)d_2\{4\} = \langle\langle4'\rangle\rangle- 2\langle\langle2'\rangle\rangle\langle\langle2\rangle\rangle. \tag*{(5)}

The differential v2{4}(pT)v_2\{4\}(p_T) coefficient is derived as

v2{4}(pT)=−d2{4}(−c2{4})−3/4.(6)v_2\{4\}(p_T) = -d_2\{4\}(-c_2\{4\})^{-3/4}. \tag*{(6)}

The cumulant calculations are based on QQ-cumulant method [37], where cumulants are expressed in terms of the corresponding QnQ_n vectors. The v2{4}(pT)v_2\{4\}(p_T) is calculated in midrapidity region ∣η∣<2.4|\eta| < 2.4. Methods using many-particle correlations suppose to reflect global collective flow and to suppress local few-particle correlations. CMS measurement of v2v_2 at high pTp_T [11] has found similarity of results with 4-, 6-, 8-particle cumulants and also with SP method. The SP method also uses QnQ_n vectors framework. The Q2Q_2 vector for the second harmonic is defined as

Q2=∑k=1Mωkei2φk,(7)Q_2 = \sum_{k=1}^{M} \omega_k e^{i2\varphi_k}, \tag*{(7)}

where MM is the multiplicity of used particles and ωk\omega_k is a weight for a given particle kk. Here we use SP method with three subevents, similarly to CMS calculation [11]. Elliptic flow can be obtained as follows

v2{SP}=⟨Q2Q2A∗⟩⟨Q2AQ2B∗⟩⟨Q2AQ2C∗⟩⟨Q2BQ2C∗⟩.(8)v_2\{\mathrm{SP}\} = \langle Q_2Q_{2A}^{*}\rangle\sqrt{\frac{\langle Q_{2A}Q_{2B}^{*}\rangle\langle Q_{2A}Q_{2C}^{*}\rangle}{\langle Q_{2B}Q_{2C}^{*}\rangle}}. \tag*{(8)}

The Q2Q_2 vector for sub-events AA and BB is determined in −5<η<−3-5 < \eta< -3 and 3<η<53 < \eta< 5 pseudorapidity regions, respectively, and for sub-event CC in ∣η∣<0.75|\eta| < 0.75 region. The vectors Q2AQ_{2A}, Q2BQ_{2B} and Q2CQ_{2C} are calculated with weight ω\omega equal to pTp_T of particle. The Q2Q_2 vector of particles of interest is calculated in ∣η∣<1|\eta| < 1 pseudorapidity region with the unit weight. If the particle of interest comes with the positive η\eta, then Q2AQ_{2A} is calculated using the negative η\eta region, and vice versa. By its nature, the SP method resembles two-particle correlation methods, but using several sub-events with large η\eta gap suppresses few-particle correlations.

In the model we also calculate elliptic flow v2RPv_{2}^{\mathrm{RP}} with respect to reaction plane angle directly with (2). v2RPv_{2}^{\mathrm{RP}} is only associated with global event geometry. Figure 2 shows the elliptic flow v2{4}v_{2}\{4\}, v2{SP}v_{2}\{\mathrm{SP}\} and v2RPv_{2}^{\mathrm{RP}} calculated from the HYDJET++ simulated events with centrality 20–30% and compared to the CMS data [11]. At relatively low transverse momenta, pT<4p_{T}<4 GeV/cc, the v2{4}(pT)v_{2}\{4\}(p_{T}) and v2{SP}v_{2}\{\mathrm{SP}\} in HYDJET++ are similar to the generated original elliptic coefficient v2RPv_{2}^{\mathrm{RP}}. It is not surprising, because in this momentum region the bulk of the produced particles mainly correlates only with the reaction plane, whereas the non-flow correlations are quite small. The model calculations are also in good agreement with the experimental results. Note, that in the data the difference between v2{SP}(pT)v_{2}\{\mathrm{SP}\}(p_{T}) and v2{4}(pT)v_{2}\{4\}(p_{T}) is more pronounced. The model-generated v2{SP}v_{2}\{\mathrm{SP}\} is closer to experimentally restored v2{4}(pT)v_{2}\{4\}(p_{T}) rather than to v2{SP}v_{2}\{\mathrm{SP}\}. The description of elliptic flow at other centralities can be found in [38]. As was shown in [28], a better quantitative agreement of HYDJET++ results with the data can be reached if both anisotropy parameters, ϵ2(b)\epsilon_{2}(b) and δ(b)\delta(b), are treated as independent ones. In the region of high transverse momenta with pT>10p_{T}>10 GeV/cc the model calculations fit to the data fairly well. In the region of intermediate transverse momenta with 4<pT<104<p_{T}<10 GeV/cc,

Figure 2. Color online. The comparison of the model elliptic flow with CMS data [11] for v2{4}(pT)v_{2}\{4\}(p_{T}) (data: circles; model: open squares) and v2{SP}(pT)v_{2}\{\mathrm{SP}\}(p_{T}) (data: blue down triangles; model: open triangles) in Pb+Pb collisions at sNN=5.02\sqrt{s_{NN}}=5.02 TeV for the centrality 20–30%. The elliptic flow calculated with respect to reaction plane angle in the model, v2RPv_{2}^{\mathrm{RP}}, is also shown (black down triangles). Lines are drawn to guide the eye.

where we have transition between soft and hard physics, the model calculations are lower than the experimental data. However, this region is out of scope of our study and, therefore, it does not affect the results of the present work.

Elliptic flow correlation at low and high transverse momenta

The lead-lead collisions at sNN=5.02\sqrt{s_{NN}} = 5.02 TeV were generated within the HYDJET++ for the following centrality intervals: σ/σgeo=5\sigma/\sigma_{geo} = 5–10%, 10–15%, 15–20%, 20–30%, 30–40%, 40–50% and 50–60%. The number of generated events for each centrality bin gradually increases from 2×1062 \times10^{6} events for σ/σgeo=5\sigma/\sigma_{geo} = 5–10% to 7×1067 \times10^{6} events for σ/σgeo=50\sigma/\sigma_{geo} = 50–60%. Figure 3 shows the correlations between the elliptic flow of hadrons with transverse momenta 1.0<pT<1.251.0 < p_{T} < 1.25 GeV/cc and that of hadrons with quite high transverse momenta, 14<pT<2014 < p_{T} < 20 GeV/cc. One can see that the HYDJET++ calculated flows, v2{4}(pT)v_{2}\{4\}(p_{T}) shown in Figure 3(a) and v2{SP}v_{2}\{\mathrm{SP}\} displayed in Figure 3 (b), demonstrate the same the centrality dependence of the elliptic flow correlations as observed in the experiment. However, elliptic flow values calculated with respect to reaction plane

Figure 3. Color online. The centrality dependence of the correlation between elliptic flows of hadrons with low and high transverse momenta in Pb+Pb collisions at sNN=5.02\sqrt{s_{NN}} = 5.02 TeV. The results are shown for seven centrality bins 5–10%, 10–15%, 15–20%, 20–30%, 30–40%, 40–50%, and 50–60%. Model calculations are presented for v2RPv_{2}^{\mathrm{RP}} (black triangles), v2{4}v_{2}\{4\} (blue circles) and v2{SP}v_{2}\{\mathrm{SP}\} (blue squares). CMS data for v2{4}v_{2}\{4\} (red down triangles) and v2{SP}v_{2}\{\mathrm{SP}\} (red stars) are taken from [11]. HYDJET++ calculations are performed also with jet quenching, j. q., (full symbols in (a) and (b)), and without jet quenching, (open symbols in (c) and (d)).

angle, v2RPv_2^{\mathrm{RP}}, at low and high pTp_T remain correlated only for semicentral collisions. For centralities larger than 30–40% v2RPv_2^{\mathrm{RP}} at high pTp_T begins to die out. Elliptic flow v2RPv_2^{\mathrm{RP}} at high pTp_T appears solely due to jet quenching effect. To confirm this one can see the model simulations without jet quenching effect in Figure 3(c), where v2RPv_2^{\mathrm{RP}} at high pTp_T is consistent with zero. The jet quenching depends not only on path length of parton in a medium, but also on medium density and temperature. For peripheral collisions the asymmetry of path lengths increases but medium density decreases which leads to decreasing of v2RPv_2^{\mathrm{RP}} at high pTp_T, see Figure 3(a). The coefficients v2{4}(pT)v_2\{4\}(p_T) and v2{SP}v_2\{\mathrm{SP}\} have additional contribution to anisotropy rather than jet quenching effect, as can be seen from Figure 3(c), (d). Note that only jets contribute to the spectrum dN/dpTdN/dp_T in this interval of transverse momentum in our approach and, therefore, the anisotropy can be caused by dijet topology. Thus, we can distinguish at least two directions, namely, the jet axis and the reaction plane, relative to which the particles are correlated. The first one becomes more pronounced for peripheral collisions, which can be explained as the follows: in central collisions more jets are produced with random directions, for peripheral collisions the probability of dijet topology increases. The model simulations indicate that in collisions with centrality below 30–40% azimuthal anisotropy occurs due to jet quenching, while in more peripheral collisions, jet particle correlations make a significant contribution to the fourth order cumulant and scalar product methods.

Conclusions

The analysis of azimuthal anisotropy coefficients at low and high transverse momenta has been performed for Pb+Pb collisions generated within two-component model HYDJET++ at center-of-mass energy sNN=5.02\sqrt{s_{NN}} = 5.02 TeV per nucleon pair. Both phenomena, responsible for the origin of azimuthal anisotropy at low and high pTp_T, namely hydrodynamic expansion of the created matter and jet quenching, are related and sensitive to initial anisotropy of overlap region of nuclei. We have found the correlation of elliptic flow values measured with respect to reaction plane angle, v2RPv_2^{\mathrm{RP}}, at low and high pTp_T in central and mid central collisions (up to 40%). For more peripheral collisions the v2RPv_2^{\mathrm{RP}} at high pTp_T decreases. Nevertheless, for coefficients measured by cumulant and scalar product methods additional source of azimuthal anisotropy emerges for peripheral collisions, i.e., anisotropy connected to dijet topology. If the centrality of the collisions is 30–40% or less, the four-cumulant and the scalar product methods are sensitive mainly to the azimuthal anisotropy of initial overlapping region. As the collisions become more peripheral, the azimuthal anisotropy begins to be determined primarily by the correlation of particles inside the jets. Combination of these two origins of anisotropy tends to reproduce experimentally observed correlations of the elliptic flow values at different transverse momenta in all centralities.

Acknowledgments

We would like to thank A.I. Demyanov, L.V. Malinina and A.V. Belyaev for fruitful discussions. This work was supported in parts by Russian Foundation for Basic Research (RFBR) under Grants No. 18-02-00155, No. 18-02-40084 and No. 18-02-40085. LVB and EEZ acknowledge support of the Norwegian Research Council (NFR) under grant No. 255253/F50 “CERN Heavy Ion Theory.”

References

  1. [34]Lokhtin I P, Belyaev A V, Ponimatkin G, Pronina E Yu and Eyyubova G Kh 2017 J. Exp. Theor. Phys. 124 244
  2. [35]Lokhtin I P and Sidorova A A 2019 J. Exp. Theor. Phys. 128 586
  3. [36]Bilandzic A, Snellings R and Voloshin S 2011 Phys. Rev. C 83 044913
  4. [37]Bilandzic A et al 2014 Phys. Rev. C 89 064904
  5. [38]Bravina L V, Eyyubova G Kh, Korotkikh V L, Lokhtin I P, Petrushanko S V, Snigirev A M and Zabrodin E E 2021 Phys. Rev. C 103 034905

Paper details

vlnc:W_fcakwasj
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 8

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents