- Email: [email protected]
High pT Harmonics in PbPb Collisions at 5.02 TeV
Available online at www.sciencedirect.com
Nuclear Physics A 967 (2017) 397–400 www.elsevier.com/locate/nuclphysa
High pT Harmonics in PbPb Collisions at 5.02 TeV Quan Wang on behalf of the CMS Collaborationa a The
University of Kansas
Abstract Studies of azimuthal anisotropies for very high pT particles in relativistic heavy ion collisions provide crucial information on the path length dependence of the parton energy loss mechanism in the quark-gluon plasma. Final high-precision data on the elliptic (v2 ) and triangular (v3 ) anisotropy harmonics of charged particles, obtained with the scalar product √ method, are presented up to pT ∼ 100 GeV/c in PbPb collisions at sNN = 5.02 TeV, using data recorded during the LHC run 2 with the CMS detector. In particular, the v3 harmonic is explored to a very high pT regime for the first time, allowing for an improved understanding of the effect of initial-state fluctuations on the parton energy loss. The v2 values reaching up pT ∼ 100 GeV/c are also determined using 4-, 6- and 8-particle cumulants, shedding new light on the origin of the observed high-pT azimuthal anisotropies. These new results are compared to theoretical calculations and provide stringent constraints on the parton energy loss mechanisms and the influence of initial-state fluctuations. Keywords: heavy ion, flow, high pT , quark-gluon plasma, scalar product, multiparticle cumulant
1. Introduction It has been observed that energetic partons lose a significant fraction of their energies when traversing through the quark-gluon plasma (QGP), the hot and dense matter created in AuAu collisions at center-of√ √ mass energy per nucleon pair sNN = 200 GeV at RHIC [1, 2, 3, 4] and in PbPb collisions at sNN = 2.76 TeV and 5.02 TeV at the LHC [5, 6, 7, 8]. The nuclear modification factor (RAA ), a ratio of the produced particle spectrum between the heavy ion and pp collisions, has shown a large suppression of the charged particles with high transverse momentum (pT ) at both RHIC [9] and LHC [8] energies. In addition, the strong dijets asymmetry in PbPb collisions [5, 6] suggests that jets are quenched in the QGP. These observations trigger strong interests of studying the detailed high energy parton energy loss mechanisms and the interplay between the hard and soft sector of the QGP. The azimuthal anisotropy of the emitted charged particles can be described by the n-th Fourier expansion coefficient vn . The low-pT charged hadron v2 and v3 reflect the hydrodynamic conversion of the initial state geometry into a final momentum space anisotropy. Studying such coefficients of the high-pT particle can provide strigent constraints on the model calculations and reveal information that is complementary to previous measurements. In this talk, the Fourier coefficients of the azimuthal anisotropy are presented using the scalar product √ (for v2 and v3 ) and multiparticle cumulant (for v2 ) methods for PbPb sNN = 5.02 TeV collisions collected in 2015 with the CMS detector [10]. The pT region up to 14 GeV/c is covered by a minimum-bias trigger. To extend the measurement to higher pT , a dedicated trigger that selects events with at least one high-pT particle is used, which results in the pT reach up to 100 GeV/c. To ensure high tracking efficiency and http://dx.doi.org/10.1016/j.nuclphysa.2017.06.014 0375-9474/© 2017 The Author(s). Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
Q. Wang / Nuclear Physics A 967 (2017) 397–400
398
reduce the misreconstructed tracks, the particles to be studied are restricted to the pseudorapidity coverage of |η| < 1. 2. Method The scalar product method [11] is used to measure the flow harmonic v2 {SP} and v3 {SP}. We take advantage of the long-range behavior by establishing an event-plane angle in the forward range of pseudorapidity and then measuring the correlations of particles in the midrapidity with respect to the reference angle. The vn coefficients can be expressed in terms of event-averaged Q-vectors, M Qn Q∗nA vn {SP} ≡ ωk · einφk , (1) , with Qn ≡ QnA Q∗nB QnA Q∗nC k=1 QnB Q∗nC where M represents the number of tracks or hadronic forward (HF) calorimeter towers (3 < |η| < 5) with energy above a certain threshold in each event, φk is the azimuthal angle of each k track (QnB ) or HF tower (QnA and QnC ), and ωk is a weighting factor equal to pT for the tracks and ET for the HF towers. Benefit from the large pseudorapidity coverage, a gap of at least 3 unit in η is guaranteed between the HF detector and the particles being studied. Such large η gap is sufficient to suppress non-flow effects, such as back-to-back jets and resonance decay etc., which have been simulated using event generators. The multiparticle correlations are measured using the Q-cumulant method [12]. The reference v2 {m} values, m = 4, 6, 8, are firstly evaluated by correlating m particles within the reference phase space of |η| < 2.4 and pT range of 1 < pT < 5 GeV/c. The pT dependent v2 (pT ) are then calculated with respect to the reference by replacing one of the m particles with a particle from certain pT range within the midrapidity, |η| < 1. 3. Results 404 μ b-1 (5.02 TeV PbPb)
CMS
vn
0.2
Centrality: 0-5%
5-10%
10-20%
30-40%
40-50%
50-60%
0.1
0.0
20-30%
vn
0.2 0.1
0.0 20
40 60 p T (GeV/c)
80
20
40 60 p T (GeV/c)
80
20
40 60 p T (GeV/c)
80
20
40 60 p T (GeV/c)
80
Fig. 1. The v2 and v3 results [13] from the SP method as a function of pT , in seven collision centrality ranges from 0–5% to 50–60%. The vertical bars (shaded boxes) represent the statistical (systematic) uncertainties. The curves represent calculations made with the CUJET3.0 [14] and the SHEE models [15] (see text).
The pT -dependent v2 (pT ) and v3 (pT ) using the SP method are shown in Figure 1 for seven collision centrality ranges from the most central (0–5%) to peripheral (50–60%) [13]. For the first time, we measure
Q. Wang / Nuclear Physics A 967 (2017) 397–400
399
the v2 and v3 up to 100 GeV/c. At low pT , both the v2 and v3 increase with increasing pT , and maximize at pT ≈ 3 GeV/c, then decreasing toward higher pT . The v2 (pT ) shows strong centrality dependence while the v3 (pT ) shows little. The centrality dependencies of v2 and v3 as expected for the conversion of the initial spatial eccentricity and its fluctuation into momentum space as described in hydrodynamic models. The v2 values remain positive up to pT ≈ 70 GeV/c, and become consistent with zero at higher pT . However, the v3 are found to to be positive up to pT ≈ 20 GeV/c over 0–40% centrality ranges. Above 20 GeV/c, the v3 values are consistent with zero within the experimental uncertainties. The v2 and v3 results are compared to CUJET3.0 [14] and SHEE [15] model calculations for various centrality bins. The CUJET3.0 model uses perturbative quantum chromodynamics (pQCD) calculations to describe the hard parton energy loss in the QGP, assuming a smooth hydrodynamic background. The CUJET3.0 curves qualitatively describe the v2 trend, but fail to quantitatively reproduce the centrality dependence, as shown in Figure. 1. The SHEE model uses viscous hydrodynamics and linear path-length dependence of the energy loss inspired inspired by pQCD. It also takes into account of the event-by-event fluctuations in the soft sector. The SHEE curves agree with both v2 and v3 data very well over the entire centrality coverage, which suggests that the modeling of the initial-state event-by-event fluctuations might be a crucial ingredient to describe the experimental data. To investigate the multiparticle nature of the high-pT particle correlations, the 4-, 6-, and 8-particle cumulant v2 results are shown in Figure. 2 with the SP v2 results included for comparison [13]. For the first time, the multiparticle cumulant v2 is measured up to 100 GeV/c. The results are expected from hydrodynamics models where v2 {SP} > v2 {4} ≈ v2 {6} ≈ v2 {8} for pT < 3 GeV/c. Over all centrality ranges and pT > 10 GeV/c, the multiparticle cumulant results remain similar, which are again similar to v2 {SP}, v2 {SP} ≈ v2 {4} ≈ v2 {6} ≈ v2 {8}. This further suggests that the azimuthal anisotropy is strongly affected by the initial-state geometrical configuration and its event-by-event fluctuations.
404 μb-1 (5.02 TeV PbPb)
CMS
Centrality: 5-10%
10-20%
20-30%
30-40%
40-50%
50-60%
v2
0.2 0.1
0.0
v2
0.2 0.1
0.0 20
40
60
pT (GeV/c)
80
20
40
60
pT (GeV/c)
80
20
40
60
pT (GeV/c)
80
Fig. 2. Comparison between the v2 results from the SP and the 4-, 6-, and 8-particle cumulant methods, as a function of pT , in six centrality ranges from 5-10% to 50-60% [13]. The vertical bars (shaded boxes) represent the statistical (systematic) uncertainties.
Figure. 3 shows the correlations between the v2 values of high-pT (left: 14–20, middle: 20–26, and right:
Q. Wang / Nuclear Physics A 967 (2017) 397–400
400
26–35 GeV/c) and low-pT (1–1.25 GeV/c) for SP and multiparticle cumulant methods [13]. A strong linear correlation is observed between the high-pT and low-pT v2 values. This suggests the initial-state geometry and its fluctuations are likely to be the common origin of the high- and low-pT particle azimuthal anisotropy, which corresponds to the path-length dependent energetic parton energy loss and the hydrodynamic flow. 404 µb-1 (5.02 TeV PbPb)
CMS v2{SP}
High-pT v2
0.10
20 < pT < 26 GeV/c
14 < pT < 20 GeV/c
26 < pT < 35 GeV/c
v2{4}
0.05
0.05
0.10
0.15
0.05
0.10
0.15
0.05
0.10
0.15
Low-pT v2 (1.0 < pT < 1.25 GeV/c)
Fig. 3. Correlation between the high-pT v2 measured in the 14–20 (left), 20–26 (middle), and 26–35 GeV/c (right) pT ranges and the low-pT v2 measured in the 1 < pT < 1.25 GeV/c range, with the SP (closed circles) and cumulant (open squares) methods [13].
4. Conclusion √ The azimuthal anisotropy of charged particle produced in PbPb collisions at sNN = 5.02 TeV has been studied with the CMS detector. The v2 and v3 harmonics are measured over the widest pT range to date from 1 up to 100 GeV/c. For the first time, the 4-, 6-, and 8-particle cumulant method is used for pT > 20 GeV/c. The v2 values are found to be positive up to pT ≈ 70 GeV/c over the measured centrality ranges, while the v3 values are consistent with zero for pT > 20 GeV/c. For pT < 3 GeV/c, the ordering of v2 {SP} > v2 {4} ≈ v2 {6} ≈ v2 {8} is consistent with a collective behavior arising from the hydrodynamics expansion of a QGP. The results are compared to model comparisons and favor a linear path-length dependence of parton energy loss and strong initial-state fluctuations. A strong correlation between the low-pT and high-pT v2 values is observed, further suggesting the common connection to the initial-state geometry. References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15]
K. Adcox, et al., Nucl. Phys. A757 (2005) 184–283. arXiv:nucl-ex/0410003. J. Adams, et al., Nucl. Phys. A757 (2005) 102–183. arXiv:nucl-ex/0501009. I. Arsene, et al., Nucl. Phys. A757 (2005) 1–27. arXiv:nucl-ex/0410020. B. B. Back, et al., Nucl. Phys. A757 (2005) 28–101. arXiv:nucl-ex/0410022. G. Aad, et al., Phys. Rev. Lett. 105 (2010) 252303. arXiv:1011.6182. S. Chatrchyan, et al., Phys. Rev. C84 (2011) 024906. arXiv:1102.1957. K. Aamodt, et al., Phys. Lett. B696 (2011) 30–39. arXiv:1012.1004. V. Khachatryan, et al., Submitted to: JHEParXiv:1611.01664. I. Arsene, et al., Phys. Rev. Lett. 91 (2003) 072305. arXiv:nucl-ex/0307003. S. Chatrchyan, et al., JINST 3 (2008) S08004. M. Luzum, J.-Y. Ollitrault, Phys. Rev. C87 (4) (2013) 044907. arXiv:1209.2323. A. Bilandzic, C. H. Christensen, K. Gulbrandsen, A. Hansen, Y. Zhou, Phys. Rev. C89 (6) (2014) 064904. arXiv:1312.3572. A. M. Sirunyan, et al.arXiv:1702.00630. J. Xu, J. Liao, M. Gyulassy, JHEP 02 (2016) 169. arXiv:1508.00552. J. Noronha-Hostler, B. Betz, J. Noronha, M. Gyulassy, Phys. Rev. Lett. 116 (25) (2016) 252301. arXiv:1602.03788.
Nuclear Physics A 967 (2017) 397–400 www.elsevier.com/locate/nuclphysa
High pT Harmonics in PbPb Collisions at 5.02 TeV Quan Wang on behalf of the CMS Collaborationa a The
University of Kansas
Abstract Studies of azimuthal anisotropies for very high pT particles in relativistic heavy ion collisions provide crucial information on the path length dependence of the parton energy loss mechanism in the quark-gluon plasma. Final high-precision data on the elliptic (v2 ) and triangular (v3 ) anisotropy harmonics of charged particles, obtained with the scalar product √ method, are presented up to pT ∼ 100 GeV/c in PbPb collisions at sNN = 5.02 TeV, using data recorded during the LHC run 2 with the CMS detector. In particular, the v3 harmonic is explored to a very high pT regime for the first time, allowing for an improved understanding of the effect of initial-state fluctuations on the parton energy loss. The v2 values reaching up pT ∼ 100 GeV/c are also determined using 4-, 6- and 8-particle cumulants, shedding new light on the origin of the observed high-pT azimuthal anisotropies. These new results are compared to theoretical calculations and provide stringent constraints on the parton energy loss mechanisms and the influence of initial-state fluctuations. Keywords: heavy ion, flow, high pT , quark-gluon plasma, scalar product, multiparticle cumulant
1. Introduction It has been observed that energetic partons lose a significant fraction of their energies when traversing through the quark-gluon plasma (QGP), the hot and dense matter created in AuAu collisions at center-of√ √ mass energy per nucleon pair sNN = 200 GeV at RHIC [1, 2, 3, 4] and in PbPb collisions at sNN = 2.76 TeV and 5.02 TeV at the LHC [5, 6, 7, 8]. The nuclear modification factor (RAA ), a ratio of the produced particle spectrum between the heavy ion and pp collisions, has shown a large suppression of the charged particles with high transverse momentum (pT ) at both RHIC [9] and LHC [8] energies. In addition, the strong dijets asymmetry in PbPb collisions [5, 6] suggests that jets are quenched in the QGP. These observations trigger strong interests of studying the detailed high energy parton energy loss mechanisms and the interplay between the hard and soft sector of the QGP. The azimuthal anisotropy of the emitted charged particles can be described by the n-th Fourier expansion coefficient vn . The low-pT charged hadron v2 and v3 reflect the hydrodynamic conversion of the initial state geometry into a final momentum space anisotropy. Studying such coefficients of the high-pT particle can provide strigent constraints on the model calculations and reveal information that is complementary to previous measurements. In this talk, the Fourier coefficients of the azimuthal anisotropy are presented using the scalar product √ (for v2 and v3 ) and multiparticle cumulant (for v2 ) methods for PbPb sNN = 5.02 TeV collisions collected in 2015 with the CMS detector [10]. The pT region up to 14 GeV/c is covered by a minimum-bias trigger. To extend the measurement to higher pT , a dedicated trigger that selects events with at least one high-pT particle is used, which results in the pT reach up to 100 GeV/c. To ensure high tracking efficiency and http://dx.doi.org/10.1016/j.nuclphysa.2017.06.014 0375-9474/© 2017 The Author(s). Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
Q. Wang / Nuclear Physics A 967 (2017) 397–400
398
reduce the misreconstructed tracks, the particles to be studied are restricted to the pseudorapidity coverage of |η| < 1. 2. Method The scalar product method [11] is used to measure the flow harmonic v2 {SP} and v3 {SP}. We take advantage of the long-range behavior by establishing an event-plane angle in the forward range of pseudorapidity and then measuring the correlations of particles in the midrapidity with respect to the reference angle. The vn coefficients can be expressed in terms of event-averaged Q-vectors, M Qn Q∗nA vn {SP} ≡ ωk · einφk , (1) , with Qn ≡ QnA Q∗nB QnA Q∗nC k=1 QnB Q∗nC where M represents the number of tracks or hadronic forward (HF) calorimeter towers (3 < |η| < 5) with energy above a certain threshold in each event, φk is the azimuthal angle of each k track (QnB ) or HF tower (QnA and QnC ), and ωk is a weighting factor equal to pT for the tracks and ET for the HF towers. Benefit from the large pseudorapidity coverage, a gap of at least 3 unit in η is guaranteed between the HF detector and the particles being studied. Such large η gap is sufficient to suppress non-flow effects, such as back-to-back jets and resonance decay etc., which have been simulated using event generators. The multiparticle correlations are measured using the Q-cumulant method [12]. The reference v2 {m} values, m = 4, 6, 8, are firstly evaluated by correlating m particles within the reference phase space of |η| < 2.4 and pT range of 1 < pT < 5 GeV/c. The pT dependent v2 (pT ) are then calculated with respect to the reference by replacing one of the m particles with a particle from certain pT range within the midrapidity, |η| < 1. 3. Results 404 μ b-1 (5.02 TeV PbPb)
CMS
vn
0.2
Centrality: 0-5%
5-10%
10-20%
30-40%
40-50%
50-60%
0.1
0.0
20-30%
vn
0.2 0.1
0.0 20
40 60 p T (GeV/c)
80
20
40 60 p T (GeV/c)
80
20
40 60 p T (GeV/c)
80
20
40 60 p T (GeV/c)
80
Fig. 1. The v2 and v3 results [13] from the SP method as a function of pT , in seven collision centrality ranges from 0–5% to 50–60%. The vertical bars (shaded boxes) represent the statistical (systematic) uncertainties. The curves represent calculations made with the CUJET3.0 [14] and the SHEE models [15] (see text).
The pT -dependent v2 (pT ) and v3 (pT ) using the SP method are shown in Figure 1 for seven collision centrality ranges from the most central (0–5%) to peripheral (50–60%) [13]. For the first time, we measure
Q. Wang / Nuclear Physics A 967 (2017) 397–400
399
the v2 and v3 up to 100 GeV/c. At low pT , both the v2 and v3 increase with increasing pT , and maximize at pT ≈ 3 GeV/c, then decreasing toward higher pT . The v2 (pT ) shows strong centrality dependence while the v3 (pT ) shows little. The centrality dependencies of v2 and v3 as expected for the conversion of the initial spatial eccentricity and its fluctuation into momentum space as described in hydrodynamic models. The v2 values remain positive up to pT ≈ 70 GeV/c, and become consistent with zero at higher pT . However, the v3 are found to to be positive up to pT ≈ 20 GeV/c over 0–40% centrality ranges. Above 20 GeV/c, the v3 values are consistent with zero within the experimental uncertainties. The v2 and v3 results are compared to CUJET3.0 [14] and SHEE [15] model calculations for various centrality bins. The CUJET3.0 model uses perturbative quantum chromodynamics (pQCD) calculations to describe the hard parton energy loss in the QGP, assuming a smooth hydrodynamic background. The CUJET3.0 curves qualitatively describe the v2 trend, but fail to quantitatively reproduce the centrality dependence, as shown in Figure. 1. The SHEE model uses viscous hydrodynamics and linear path-length dependence of the energy loss inspired inspired by pQCD. It also takes into account of the event-by-event fluctuations in the soft sector. The SHEE curves agree with both v2 and v3 data very well over the entire centrality coverage, which suggests that the modeling of the initial-state event-by-event fluctuations might be a crucial ingredient to describe the experimental data. To investigate the multiparticle nature of the high-pT particle correlations, the 4-, 6-, and 8-particle cumulant v2 results are shown in Figure. 2 with the SP v2 results included for comparison [13]. For the first time, the multiparticle cumulant v2 is measured up to 100 GeV/c. The results are expected from hydrodynamics models where v2 {SP} > v2 {4} ≈ v2 {6} ≈ v2 {8} for pT < 3 GeV/c. Over all centrality ranges and pT > 10 GeV/c, the multiparticle cumulant results remain similar, which are again similar to v2 {SP}, v2 {SP} ≈ v2 {4} ≈ v2 {6} ≈ v2 {8}. This further suggests that the azimuthal anisotropy is strongly affected by the initial-state geometrical configuration and its event-by-event fluctuations.
404 μb-1 (5.02 TeV PbPb)
CMS
Centrality: 5-10%
10-20%
20-30%
30-40%
40-50%
50-60%
v2
0.2 0.1
0.0
v2
0.2 0.1
0.0 20
40
60
pT (GeV/c)
80
20
40
60
pT (GeV/c)
80
20
40
60
pT (GeV/c)
80
Fig. 2. Comparison between the v2 results from the SP and the 4-, 6-, and 8-particle cumulant methods, as a function of pT , in six centrality ranges from 5-10% to 50-60% [13]. The vertical bars (shaded boxes) represent the statistical (systematic) uncertainties.
Figure. 3 shows the correlations between the v2 values of high-pT (left: 14–20, middle: 20–26, and right:
Q. Wang / Nuclear Physics A 967 (2017) 397–400
400
26–35 GeV/c) and low-pT (1–1.25 GeV/c) for SP and multiparticle cumulant methods [13]. A strong linear correlation is observed between the high-pT and low-pT v2 values. This suggests the initial-state geometry and its fluctuations are likely to be the common origin of the high- and low-pT particle azimuthal anisotropy, which corresponds to the path-length dependent energetic parton energy loss and the hydrodynamic flow. 404 µb-1 (5.02 TeV PbPb)
CMS v2{SP}
High-pT v2
0.10
20 < pT < 26 GeV/c
14 < pT < 20 GeV/c
26 < pT < 35 GeV/c
v2{4}
0.05
0.05
0.10
0.15
0.05
0.10
0.15
0.05
0.10
0.15
Low-pT v2 (1.0 < pT < 1.25 GeV/c)
Fig. 3. Correlation between the high-pT v2 measured in the 14–20 (left), 20–26 (middle), and 26–35 GeV/c (right) pT ranges and the low-pT v2 measured in the 1 < pT < 1.25 GeV/c range, with the SP (closed circles) and cumulant (open squares) methods [13].
4. Conclusion √ The azimuthal anisotropy of charged particle produced in PbPb collisions at sNN = 5.02 TeV has been studied with the CMS detector. The v2 and v3 harmonics are measured over the widest pT range to date from 1 up to 100 GeV/c. For the first time, the 4-, 6-, and 8-particle cumulant method is used for pT > 20 GeV/c. The v2 values are found to be positive up to pT ≈ 70 GeV/c over the measured centrality ranges, while the v3 values are consistent with zero for pT > 20 GeV/c. For pT < 3 GeV/c, the ordering of v2 {SP} > v2 {4} ≈ v2 {6} ≈ v2 {8} is consistent with a collective behavior arising from the hydrodynamics expansion of a QGP. The results are compared to model comparisons and favor a linear path-length dependence of parton energy loss and strong initial-state fluctuations. A strong correlation between the low-pT and high-pT v2 values is observed, further suggesting the common connection to the initial-state geometry. References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15]
K. Adcox, et al., Nucl. Phys. A757 (2005) 184–283. arXiv:nucl-ex/0410003. J. Adams, et al., Nucl. Phys. A757 (2005) 102–183. arXiv:nucl-ex/0501009. I. Arsene, et al., Nucl. Phys. A757 (2005) 1–27. arXiv:nucl-ex/0410020. B. B. Back, et al., Nucl. Phys. A757 (2005) 28–101. arXiv:nucl-ex/0410022. G. Aad, et al., Phys. Rev. Lett. 105 (2010) 252303. arXiv:1011.6182. S. Chatrchyan, et al., Phys. Rev. C84 (2011) 024906. arXiv:1102.1957. K. Aamodt, et al., Phys. Lett. B696 (2011) 30–39. arXiv:1012.1004. V. Khachatryan, et al., Submitted to: JHEParXiv:1611.01664. I. Arsene, et al., Phys. Rev. Lett. 91 (2003) 072305. arXiv:nucl-ex/0307003. S. Chatrchyan, et al., JINST 3 (2008) S08004. M. Luzum, J.-Y. Ollitrault, Phys. Rev. C87 (4) (2013) 044907. arXiv:1209.2323. A. Bilandzic, C. H. Christensen, K. Gulbrandsen, A. Hansen, Y. Zhou, Phys. Rev. C89 (6) (2014) 064904. arXiv:1312.3572. A. M. Sirunyan, et al.arXiv:1702.00630. J. Xu, J. Liao, M. Gyulassy, JHEP 02 (2016) 169. arXiv:1508.00552. J. Noronha-Hostler, B. Betz, J. Noronha, M. Gyulassy, Phys. Rev. Lett. 116 (25) (2016) 252301. arXiv:1602.03788.