Measurement of medium–induced acoplanarity in central $\mathrm{Au}\text{--}\mathrm{Au}$ and $pp$ collisions at $\sqrt{s_{\mathrm{NN}}}=200$ GeV using direct–photon+jet and $\pi^{0}$+jet correlations

Introduction - Matter under extreme conditions of temperature and density forms a state of matter called Quark–Gluon Plasma (QGP) which consists of deconfined quarks and gluons [1, 2, 3, 4]. A QGP filled the early universe a few microseconds after the Big Bang, and is generated in collisions of heavy nuclei at the Relativistic Heavy–Ion Collider (RHIC) and the Large Hadron Collider (LHC). Experimental measurements at these facilities, and their comparison to theoretical model calculations, show that the QGP exhibits emergent collective behavior, flowing with the lowest possible specific shear viscosity [5].

Lattice Quantum Chromodynamics (QCD) calculations of high–temperature matter at zero net–baryon density indicate that the effective number of degrees of freedom in the QGP is about 15% less than the Stefan–Boltzmann limit for a non-interacting quark–gluon gas, even at temperatures several times the pseudo–critical temperature $T_{\mathrm{c}}\approx 155$ MeV [6, 7, 8, 9, 10]. This indicates that QGP quasiparticles in this temperature range are complex multi–particle states of quarks and gluons, which may drive its collective dynamics  [3]. However, the microscopic structure of the QGP remains largely unexplored experimentally.

In high–energy hadronic collisions, quarks and gluons (partons) in the projectiles can experience hard (high momentum–transfer ${Q}^{2}$) scattering. The scattered parton is initially virtual, decaying in a parton shower which hadronizes as an observable spray of hadrons (a “jet”) [11, 12, 13, 14, 15]. In high-energy nuclear collisions, hard scatterings occur before the formation of the QGP, and scattered partons subsequently interact with it (“jet quenching”) [16, 17, 18]. Jet quenching modifies observed jet production rates and substructure, providing unique probes of the QGP [17, 18].

The secondary scattering of hard partons in the QGP has been proposed as a probe of QGP quasiparticles [19, 20, 21, 22, 23, 24], in analogy to Rutherford scattering as a probe of the atomic nucleus [25]. Observation of secondary, in–medium partonic scattering requires a coincidence observable, in which a trigger particle associated with an initial hard scattering specifies a direction, and the azimuthal difference $\Delta\phi$ between a recoil jet and the trigger (acoplanarity with respect to the plane defined by the beam axis and trigger) is measured. Acoplanarity distributions have been measured at the LHC for $pp$ and central $\mathrm{Pb}\text{--}\mathrm{Pb}$ collisions at center of mass energy $\sqrt{s_{\mathrm{NN}}}=5.02$ TeV, by the CMS and ALICE collaborations [26, 27, 28, 29, 30, 31]. CMS utilizes a direct–photon ($\gamma_{\mathrm{dir}}$) trigger with transverse energy $E_{\mathrm{T}}>60$ GeV and recoil jets with transverse momentum $p_{\mathrm{T,jet}}>30$ $\mathrm{GeV/}c$, with jet resolution parameter $R=0.3$; no medium–induced modification is observed within uncertainties [30]. ALICE reports semi-inclusive distributions of charged–particle recoil jets with $p_{\mathrm{T,jet}}^{\mathrm{ch}}>10$ $\mathrm{GeV/}c$ recoiling from hadron triggers with $p_{\mathrm{T}}^{\mathrm{trig}}>20$ $\mathrm{GeV/}c$ (h+jet); significant medium–induced broadening of the acoplanarity distribution is observed in the range $10<p_{\mathrm{T,jet}}^{\mathrm{ch}}<20$ $\mathrm{GeV/}c$ for $R=0.4$ and 0.5, but not for $R=0.2$ or at higher $p_{\mathrm{T,jet}}^{\mathrm{ch}}$ [27, 28]. This marked dependence of the broadening on $R$ and $p_{\mathrm{T,jet}}^{\mathrm{ch}}$ suggests that it arises from response of the QGP medium to excitation by a jet (“wake”), rather than single hard Rutherford–like scattering [27, 28]. At RHIC, the STAR collaboration has also reported an acoplanarity measurement based on semi–inclusive h+jet correlations in central and peripheral $\mathrm{Au}\text{--}\mathrm{Au}$ collisions at $\sqrt{s_{\mathrm{NN}}}=200$ GeV, with no significant medium–induced broadening observed for $R=0.3$ [32].

In this Letter, the STAR experiment reports the first measurement of jet acoplanarity of charged–particle jets recoiling from direct photon ($\gamma_{\mathrm{dir}}$) and $\pi^{0}$ triggers in $pp$ and central $\mathrm{Au}\text{--}\mathrm{Au}$ collisions at $\sqrt{s_{\mathrm{NN}}}=200$ GeV. The trigger particles have $11<E_{\mathrm{T}}^{\mathrm{trig}}<15$ GeV, with the semi–inclusive distribution of recoil jets reported in $10<p_{\mathrm{T,jet}}^{\mathrm{ch}}<20$ $\mathrm{GeV/}c$ for $R=0.2$ and 0.5. Uncorrelated jet background yield is corrected using event mixing. This letter extends the analysis reported in Refs. [33, 34], to measure acoplanarity.

The measurement of acoplanarity distributions with $\gamma_{\mathrm{dir}}$ and $\pi^{0}$ triggers in the same analysis provides systematic variation in the recoil–jet color charge and pathlength distributions [33, 34], elucidating their influence on jet quenching effects. This measurement complements that reported by ALICE in Ref. [27], exploring acoplanarity broadening using the same observable but with a collision system at markedly different $\sqrt{s_{\mathrm{NN}}}$, thereby probing sensitivity to variation in the QGP temperature and expansion dynamics [35, 36]. Theoretical model calculations incorporating jet quenching are also compared to the measurements.

Detector, dataset, and analysis - Data for $pp$ and $\mathrm{Au}\text{--}\mathrm{Au}$ collisions at $\sqrt{s_{\mathrm{NN}}}=200$ GeV were recorded during the 2009 and 2014 RHIC runs, respectively. The detector, datasets, triggering, offline event selection, and track reconstruction are described in Ref. [33]. Online event selection is based on high–energy single showers measured in the Barrel Electromagnetic Calorimeter (BEMC) [37]. Centrality of $\mathrm{Au}\text{--}\mathrm{Au}$ collisions is determined offline using the uncorrected charged–particle multiplicity within pseudo–rapidity $|\eta|<0.5$; the 15% highest–multiplicity (“central”) $\mathrm{Au}\text{--}\mathrm{Au}$ collisions are selected for analysis. Events are further selected in both the $pp$ and central $\mathrm{Au}\text{--}\mathrm{Au}$ datasets by requiring the presence of a $\gamma_{\mathrm{dir}}$ or $\pi^{0}$ candidate with $11<E_{\mathrm{T}}^{\mathrm{trig}}<15$ GeV. The integrated luminosity for the analysis is 23 pb^-1 and 3.9 nb^-1 for $pp$ and $\mathrm{Au}\text{--}\mathrm{Au}$ collisions, respectively.

High-$p_{\mathrm{T}}$ photon production in RHIC collisions arises from several sources [38, 39]: direct ($2\rightarrow 2$) production (Compton, annihilation), fragmentation, and hadronic decays. Discrimination of $\gamma$ and $\pi^{0}$–induced showers utilizes EM shower shape measured in the BEMC and its Shower Maximum detector (BSMD) [40, 33]. The purity of the resulting $\pi^{0}$–tagged population is estimated from simulation to be greater than 95% [34], while the photon–tagged population contains an admixture of $\pi^{0}$ and is labeled “$\gamma$–rich” ($\gamma_{\mathrm{rich}}$). The $\gamma_{\mathrm{dir}}$ fraction of the $\gamma_{\mathrm{rich}}$ population, which is determined using the measured rate of nearby correlated charged hadrons, depends on collision system and $E_{\mathrm{T}}^{\mathrm{trig}}$ and is in the range $\approx 40-80\%$ [34]. Correction based on this rate accounts for the hadronic decay component, and much but not all of the fragmentation photon contribution [40, 34]. This population is labeled $\gamma_{\mathrm{dir}}$.

Jet reconstruction likewise follows Ref. [33]. Jets are reconstructed from charged–particle tracks in $|\eta_{\mathrm{track}}|<1$ and $p_{\mathrm{T,track}}>0.2$ $\mathrm{GeV/}c$ using the $\mathrm{anti-}k_{\mathrm{T}}$ algorithm with $R=0.2$ and 0.5, with E–scheme recombination and active ghost area 0.01 [41]. Jets whose centroid has $|\eta_{\mathrm{jet}}|<1-R$ are accepted for analysis, and measured distributions are normalized to unit $\eta_{\mathrm{jet}}$. The raw jet transverse momentum, $p_{\mathrm{T,jet}}^{\mathrm{raw}}$, is adjusted according to $p_{\mathrm{T,jet}}^{\mathrm{reco,ch}}=p_{\mathrm{T,jet}}^{\mathrm{raw}}-\rho A_{\mathrm{jet}}$, where $\rho$ is the event–wise estimated background $p_{\mathrm{T}}$-density and $A_{\mathrm{jet}}$ is the jet area  [42, 33]. This event–wise approximate correction is refined by the deconvolution of detector effects (“unfolding”).

The number of jet candidates as a function of $p_{\mathrm{T,jet}}$ and $\Delta\phi$ is normalized by $N_{\mathrm{trig}}$, the number of triggers,

The measured recoil–jet yield in central $\mathrm{Au}\text{--}\mathrm{Au}$ collisions has multiple contributions: correlated recoil jets from the same hard (high–${Q}^{2}$) scattering process which generates the trigger, corresponding to $Y(p_{\mathrm{T,jet}}^{\mathrm{ch}},\Delta\phi)$; jets arising from other partonic scattering processes, which are uncorrelated with the trigger (multiple partonic interactions, or MPI); and combinatorial jets arising from the random combination of tracks generated by soft (low–${Q}^{2}$) processes. The jet yield in central $\mathrm{Au}\text{--}\mathrm{Au}$ collisions at $\sqrt{s_{\mathrm{NN}}}=200$ GeV due to MPI is negligible [43]. However, the combinatorial yield in such collisions can be significant relative to the correlated signal, especially for large $R$ at low $p_{\mathrm{T,jet}}^{\mathrm{ch}}$.

Correction of the raw recoil-jet yield to measure the $Y(p_{\mathrm{T,jet}}^{\mathrm{ch}},\Delta\phi)$ distribution is first carried out on 1-D $Y(p_{\mathrm{T,jet}}^{\mathrm{reco,ch}})$ raw distributions binned in raw $\Delta\phi$, following the procedures described in Ref. [43]. The corrected 1-D distributions as a function of $p_{\mathrm{T,jet}}^{\mathrm{ch}}$ are then combined to form a corrected 2-D distribution in ($p_{\mathrm{T,jet}}^{\mathrm{ch}}$,$\Delta\phi$) using weights that account for $\Delta\phi$-smearing due to residual background. The correction steps are outlined below, with detail provided for the $\Delta\phi$–smearing correction.

Correction for the combinatorial jet yield in central $\mathrm{Au}\text{--}\mathrm{Au}$ collisions is carried out using mixed events (ME) [43, 33], which are constructed by combining single tracks from multiple real events (“same events,” or SE) in each of 540 distinct bins in multiplicity, $z_{\mathrm{vtx}}$ (primary vertex beamline position), event plane orientation, and run–averaged luminosity [33]. Figure 1 shows examples of the SE and ME (normalized, see below) distributions, and their ratio, for central $\mathrm{Au}\text{--}\mathrm{Au}$ collisions with $\gamma_{\mathrm{rich}}$ triggers and recoil jets with $R=0.2$ and 0.5, in the ranges $2.2<\Delta\phi<2.5$ and $2.8<\Delta\phi<3.0$. A negative value of $p_{\mathrm{T,jet}}^{\mathrm{reco,ch}}$ arises if $p_{\mathrm{T,jet}}^{\mathrm{raw}}$ is less than $\rho A_{\mathrm{jet}}$; for large negative values of $p_{\mathrm{T,jet}}^{\mathrm{reco,ch}}$ this occurs predominantly due to uncorrelated background, resulting in the same $p_{\mathrm{T,jet}}^{\mathrm{reco,ch}}$–dependent shape of the $Y_{\mathrm{SE}}$ and $Y_{\mathrm{ME}}$ distributions in that region [26, 43, 27, 28, 33, 34].

The ME distribution is normalized to the SE distribution in the negative $p_{\mathrm{T,jet}}^{\mathrm{reco,ch}}$ region (MEnorm), which corrects the effective acceptance difference of the SE and ME populations due to displacement of uncorrelated jet candidates at low and negative $p_{\mathrm{T,jet}}^{\mathrm{reco,ch}}$ by hard, correlated jet candidates [26, 43]. The SE/ME yield ratio (lower panel insets) is independent of $p_{\mathrm{T,jet}}^{\mathrm{reco,ch}}$ within a few percent over a negative–$p_{\mathrm{T,jet}}^{\mathrm{reco,ch}}$ range in which the yield itself varies by several orders of magnitude; for $R=0.5$, the range in $p_{\mathrm{T,jet}}^{\mathrm{reco,ch}}$ over which the ratio is flat within statistical uncertainty is limited. The extracted normalization factors have values between 0.9 and unity [33, 34]. The systematic dependence of the normalization factor on the upper bound of the normalization region is negligible [44].

The distribution of recoil–jet yield correlated with the trigger corresponds to the difference distribution [26, 43]

Figure 1 shows $\widetilde{Y}_{\Delta\phi}$($p_{\mathrm{T,jet}}^{\mathrm{reco,ch}}$) distributions, which vary smoothly even for small signal/background yield, $\widetilde{Y}\ll Y_{\mathrm{SE}}$. Negative values from the subtraction are not displayed due to the logarithmic vertical scale, but all such points have central values consistent with zero within statistical uncertainty. These features indicate that the ME distribution reproduces accurately the uncorrelated jet distribution in SE population, which can therefore be corrected with high precision [43].

The $\widetilde{Y}_{\Delta\phi}(p_{\mathrm{T,jet}}^{\mathrm{reco,ch}})$ distributions are then corrected via unfolding for instrumental effects in both $pp$ and central $\mathrm{Au}\text{--}\mathrm{Au}$ collisions, and for residual uncorrelated background fluctuations in central $\mathrm{Au}\text{--}\mathrm{Au}$ collisions [33, 43]. Finally, the unfolded distributions are corrected for jet–finding efficiency [33]. The dominant systematic uncertainty is due to unfolding [33].

Instrumental effects generate negligible $\Delta\phi$–smearing relative to the analysis binning. The only significant $\Delta\phi$–smearing is due to spatial variation of uncorrelated background in central $\mathrm{Au}\text{--}\mathrm{Au}$ collisions, which can modify the jet centroid direction. Correction for $\Delta\phi$–smearing is implemented bin–wise by a weight matrix $w(\Delta\phi_{\mathrm{true}},\Delta\phi_{\mathrm{meas}})$ that scales the measured $\widetilde{Y}_{\Delta\phi}(p_{\mathrm{T,jet}}^{\mathrm{ch}})$ distributions. The weights are determined by embedding detector–level PYTHIA-generated events for $pp$ collisions into real central $\mathrm{Au}\text{--}\mathrm{Au}$ events, with systematic uncertainty determined by varying the jet fragmentation model to mimic jet quenching effects. The systematic uncertainty of this correction is negligible [44].

The $\gamma_{\mathrm{dir}}$-triggered recoil–jet distributions for each $\Delta\phi$ bin are then determined from the corrected $\gamma_{\mathrm{rich}}$–triggered and $\pi^{0}$-triggered distributions [33].

Results - Figure 2 shows corrected $\widetilde{Y}_{p_{\mathrm{T}}}$($\Delta\phi$) distributions for $\gamma_{\mathrm{dir}}$ and $\pi^{0}$ triggers in $pp$ and central $\mathrm{Au}\text{--}\mathrm{Au}$ collisions, for $R=0.2$ and 0.5. The distributions fall steeply away from $\Delta\phi=\pi$, with greater yield for $R=0.5$ than for $R=0.2$. Figure 1, lower panels, shows greater yield for $R=0.5$ than for $R=0.2$ for $p_{\mathrm{T,jet}}^{\mathrm{reco,ch}}>0$ at large angles relative to $\Delta\phi=\pi$; this effect is therefore not generated predominantly by corrections. A similar effect is reported in Ref. [27].

Figure 2 shows a calculation for $pp$ collisions at $\sqrt{s}=200$ GeV using PYTHIA–6 STAR tune [45], with the $E_{\mathrm{T}}^{\mathrm{trig}}$ distribution smeared to account for the BEMC detector response [33]. This calculation describes the measurements well for both $\gamma_{\mathrm{dir}}$ and $\pi^{0}$ triggers. The figure also shows an analytic QCD calculation at Next–to–Leading–Log (NLL) accuracy with Sudakov resummation [46, 23] for $\pi^{0}$ triggers in $pp$ collisions, in the range $2.5<\Delta\phi<\pi$ rad. This calculation, which is not smeared by the $E_{\mathrm{T}}^{\mathrm{trig}}$ resolution [33], reproduces the $pp$ data well for $R=0.5$, but not for $R=0.2$.

Figure 3 shows ${I}_{\mathrm{AA}}(\Delta\phi)$, the ratio of $\widetilde{Y}_{p_{\mathrm{T}}}$($\Delta\phi$) distributions measured in central $\mathrm{Au}\text{--}\mathrm{Au}$ and $pp$ collisions, for $\gamma_{\mathrm{dir}}$ and $\pi^{0}$ triggers and recoil jets with $R=0.2$ and 0.5. For $\gamma_{\mathrm{dir}}$ triggers the binning is different for central $\mathrm{Au}\text{--}\mathrm{Au}$ and $pp$ collisions, due to different dataset sizes. The denominator of ${I}_{\mathrm{AA}}(\Delta\phi)$ is therefore determined by fitting an exponential function to the $pp$ spectrum and interpolating. The smoothly–varying systematic uncertainty is likewise interpolated. For the $pp$ data points in Fig. 2 which show a limit, Figure 3 utilizes the lower limit of the systematic uncertainty for the $\mathrm{Au}\text{--}\mathrm{Au}$ data for the numerator in the ratio. The uncertainty boxes are the quadrature sum of uncorrelated uncertainties in numerator and denominator; these residual uncertainties in the ratio are nevertheless correlated between different $\Delta\phi$ bins. Fig. 3, right panels, show ${I}_{\mathrm{AA}}(\Delta\phi)$ integrated over $3\pi/4<\Delta\phi<\pi$, reported as ${I}_{\mathrm{AA}}$ in Ref. [34]; these measurements are consistent.

For large acoplanarity, all panels show suppression of ${I}_{\mathrm{AA}}(\Delta\phi)$ for recoil jets with $R=0.2$ and significant enhancement for $R=0.5$. The value of ${I}_{\mathrm{AA}}(\Delta\phi)$ at $\Delta\phi\approx 2.65$ differs for $R=0.2$ and 0.5 by a factor $20\pm 2$ (sys) for $\pi^{0}$ triggers and recoil jet $15<p_{\mathrm{T,jet}}^{\mathrm{ch}}<20$ $\mathrm{GeV/}c$ (bottom panel), and a factor $3.8\pm 1.7$ for $\gamma_{\mathrm{dir}}$ triggers and recoil jet $10<p_{\mathrm{T,jet}}^{\mathrm{ch}}<15$ $\mathrm{GeV/}c$ (top panel); statistical error is negligible. Significant differences are likewise observed for $\Delta\phi\approx 2.35$. A similar, marked $R$–dependent broadening of ${I}_{\mathrm{AA}}(\Delta\phi)$ was observed in the same $p_{\mathrm{T,jet}}^{\mathrm{ch}}$-range for h+jet correlations in central $\mathrm{Pb}\text{--}\mathrm{Pb}$ collisions at $\sqrt{s_{\mathrm{NN}}}=5.02$ TeV [27].

Medium–induced yield enhancement at large acoplanarity may arise from secondary partonic scattering with QGP quasiparticles [19, 20, 21, 22, 23, 24]. However, such scattering effects should be evident for all $R$–values, which are used to probe the population of hard-scattering processes with different apertures. In contrast, Fig. 3 and Ref. [27] show selective enhancement for $R=0.5$ but not $R=0.2$, which is not consistent with jet scattering as the predominant underlying mechanism.

Another potential source of acoplanarity broadening is MPIs, as discussed above. While the MPI contribution to ${I}_{\mathrm{AA}}({p_{\mathrm{T,jet}}^{\mathrm{ch}}})$  is negligible at RHIC energies [43], MPI may still affect the tail of the ${I}_{\mathrm{AA}}(\Delta\phi)$ distribution. However, the strong $R$-dependence of ${I}_{\mathrm{AA}}(\Delta\phi)$ in Fig. 3 likewise disfavors that scenario, since MPI effects should also broaden the acoplanarity distribution for all $R$ values.

A consistent picture accommodating these observations is that selective acoplanarity broadening for large $R$ and low $p_{\mathrm{T,jet}}^{\mathrm{ch}}$ arises from medium response, whereby the trigger–correlated “jets” observed at large angular deviation from $\Delta\phi\approx\pi$ represent the diffuse wake or medium response to a recoil jet propagating in the QGP [27, 28]. Figure 3 therefore shows evidence of the medium response to the passage of an energetic jet, which has not been identified previously at RHIC.

The $\gamma_{\mathrm{dir}}$–triggered and $\pi^{0}$–triggered distributions are expected to differ in recoil jet relative quark/gluon fraction and in average in-medium path length [33, 34]. The medium–induced azimuthal broadening shown in Fig. 3 is qualitatively similar for $\gamma_{\mathrm{dir}}$ and $\pi^{0}$ triggers, though the distributions differ in detail. Systematic comparison of these two distributions with model calculations may provide new insight into the color charge and pathlength dependence of jet quenching effects.

Figure 3 shows comparisons of several theoretical calculations incorporating jet quenching with the data: JEWEL [47, 48], with medium–recoil effects; the analytic QCD calculation shown in Fig. 2, with Gaussian distributed in–medium broadening [23]; and the Hybrid Monte Carlo model [49, 50, 51] with hydrodynamic wake implemented. These calculations are not smeared by the $E_{\mathrm{T}}^{\mathrm{trig}}$ resolution, whose effects are similar in $pp$ and $\mathrm{Au}\text{--}\mathrm{Au}$ and largely cancel in the ${I}_{\mathrm{AA}}(\Delta\phi)$ ratio [33].

JEWEL is based on PYTHIA, which describes the $pp$ measurements well (Fig. 2). JEWEL also describes well the $R$–dependent medium–induced acoplanarity broadening seen at the LHC [27]. However, at RHIC energies JEWEL does not exhibit significant medium–induced broadening for either $R=0.2$ and 0.5, for both $\gamma_{\mathrm{dir}}$ and $\pi^{0}$ triggers (Fig. 3).

For the analytic QCD calculation the Gaussian broadening width is $\langle\hat{q}{L}\rangle$, where $\hat{q}$ is the jet transport coefficient [16], $L$ is the in–medium path length, and $\langle{\ldots}\rangle$ indicates averaging over collisions. The band in Figure 3 corresponds to $3<\langle\hat{q}{L}\rangle<13$ GeV², reproducing the measured ${I}_{\mathrm{AA}}(\Delta\phi)$ for $R=0.5$, but not for $R=0.2$. This feature is intrinsic to the simple Gaussian azimuthal broadening employed, which models jet–QGP multiple scattering without $R$ dependence. This disagreement with data provides additional evidence that the observed marked $R$–dependence does not arise predominantly from in–medium soft scattering.

The Hybrid Model predicts medium–induced narrowing for both $R=0.2$ and $R=0.5$ (Fig. 3), in disagreement with the data, with similar predictions at LHC energies [27]. The End Matter presents Hybrid Model calculations for recoil jet $5<p_{\mathrm{T,jet}}^{\mathrm{ch}}<10$ $\mathrm{GeV/}c$ which exhibit $R$–dependent broadening whose qualitative features are similar to those seen in data for larger values of $p_{\mathrm{T,jet}}^{\mathrm{ch}}$, but only with wake implemented. This provides additional insight into the physical origin of the broadening, and its modeling.

Summary- This Letter reports measurements of the semi–inclusive acoplanarity distribution of charged-particle jets recoiling from $\gamma_{\mathrm{dir}}$ and $\pi^{0}$ triggers in $pp$ and central $\mathrm{Au}\text{--}\mathrm{Au}$ collisions at $\sqrt{s_{\mathrm{NN}}}=200$ GeV. Significant $R$–dependent medium–induced acoplanarity broadening is observed, corresponding to a yield enhancement for large $R$ compared to small $R$ jets of up to a factor 20. A picture that accommodated these observations and a corresponding measurement at the LHC is that the broadening arises predominantly from diffuse QGP medium response to the passage of an energetic parton, i.e. the jet wake, rather than single hard Rutherford–like scattering off of QGP quasiparticles. Theoretical calculations incorporating jet quenching and QGP wake effects exhibit significant differences with the measurements, requiring modification of their underlying physics description to improve the agreement. These measurements provide new insight into the nature of the interaction between jets and the QGP, and the application of jets to probe QGP dynamics.

Acknowledgments- We thank Jaime Norman, Yu Shi, Shu-Yi Wei, Bowen Xiao, Feng Yuan, Danny Pablos, and Krishna Rajagopal for providing calculations. We thank the RHIC Operations Group and SDCC at BNL, the NERSC Center at LBNL, and the Open Science Grid consortium for providing resources and support. This work was supported in part by the Office of Nuclear Physics within the U.S. DOE Office of Science, the U.S. National Science Foundation, National Natural Science Foundation of China, Chinese Academy of Science, the Ministry of Science and Technology of China and the Chinese Ministry of Education, NSTC Taipei, the National Research Foundation of Korea, Czech Science Foundation and Ministry of Education, Youth and Sports of the Czech Republic, Hungarian National Research, Development and Innovation Office, New National Excellency Programme of the Hungarian Ministry of Human Capacities, Department of Atomic Energy and Department of Science and Technology of the Government of India, the National Science Centre and WUTID-UB of Poland, the Ministry of Science, Education and Sports of the Republic of Croatia, German Bundesministerium für Bildung, Wissenschaft, Forschung and Technologie (BMBF), Helmholtz Association, Ministry of Education, Culture, Sports, Science, and Technology (MEXT), Japan Society for the Promotion of Science (JSPS) and Agencia Nacional de Investigación y Desarrollo (ANID) of Chile.