Dijet-Hadron Correlations in QCD

• Dijet-hadron correlations are advanced QCD observables that bridge single-particle and full jet measurements by analyzing near- and away-side angular distributions.
• They employ triggered high-pT hadron or jet signals with mixed-event corrections, Fourier decompositions, and Gaussian fits to separate jet peaks from background flow.
• These measurements elucidate trigger bias, vacuum baselines, and medium-induced jet quenching across systems such as p+p, d+Au, and Pb–Pb, enabling extraction of transport coefficients.

Dijet-hadron correlations are correlation measurements in which a high-$p_T$ hadron or a reconstructed jet is used as a proxy for one leg of a hard scattering, and the distribution of associated hadrons is studied on the near side $(\Delta\phi \approx 0)$ and away side $(\Delta\phi \approx \pi)$ as a function of relative azimuth $\Delta\phi$, pseudorapidity $\Delta\eta$, and momentum. In practice, the subject spans triggered di-hadron correlations, jet-hadron correlations, heavy-flavor analogs such as $D^0$-hadron correlations, and large-rapidity dijet observables. Across RHIC and LHC measurements, these correlations serve as a bridge between single-inclusive hadron observables and fully reconstructed jets, constraining vacuum fragmentation, trigger bias, long-range ridge structure, collective harmonics, transverse-momentum broadening, and medium-induced softening and broadening of recoil jets (Renk et al., 2011).

1. Observable content and analysis methodology

The standard experimental basis is the per-trigger associated yield or an equivalent mixed-event-normalized correlation function in $(\Delta\eta,\Delta\phi)$, with

$$\Delta\eta=\eta_{\rm assoc}-\eta_{\rm trig},\qquad \Delta\phi=\phi_{\rm assoc}-\phi_{\rm trig}.$$

In STAR’s $d+$Au analysis, the raw correlation is corrected with mixed events, the background magnitude is obtained with the Zero-Yield-At-Minimum (ZYAM) method, and the near side and away side are integrated over $|\Delta\phi|<\pi/3$ and $(\Delta\phi \approx 0)$0, respectively; the same analysis also uses the Fourier coefficients

$(\Delta\phi \approx 0)$1

without background subtraction (Adamczyk et al., 2015). In ALICE high-$(\Delta\phi \approx 0)$2 di-hadron measurements, the per-trigger yield is integrated over $(\Delta\phi \approx 0)$3 on the near side and $(\Delta\phi \approx 0)$4 on the away side to define the jet-associated yields that enter $(\Delta\phi \approx 0)$5 and $(\Delta\phi \approx 0)$6 (Zhu, 2013).

Two-dimensional fitting is frequently used to separate localized jet peaks from $(\Delta\phi \approx 0)$7-independent harmonic structure. In the $(\Delta\phi \approx 0)$8-hadron analysis at STAR, the full correlation is fit with a constant pedestal, a quadrupole term $(\Delta\phi \approx 0)$9, and Gaussian near-side and away-side peaks in $(\Delta\phi \approx \pi)$0, with the reconstructed $(\Delta\phi \approx \pi)$1 serving as a proxy for a charm jet (Jentsch, 2018). In untriggered Pb–Pb analyses, ALICE likewise decomposes the correlation into a near-side 2D Gaussian plus harmonic cosine terms up to $(\Delta\phi \approx \pi)$2, and defines the Gaussian volume as $(\Delta\phi \approx \pi)$3 (Grosse-Oetringhaus, 2011).

At high $(\Delta\phi \approx \pi)$4, yield modification factors quantify medium effects on the correlated jet fragments. The near-side and away-side per-trigger associated yields $(\Delta\phi \approx \pi)$5 are converted into

$(\Delta\phi \approx \pi)$6

while at lower and intermediate $(\Delta\phi \approx \pi)$7 the same underlying correlation is often characterized by $(\Delta\phi \approx \pi)$8 or by a fit combining jet-like peaks with harmonic flow backgrounds (Adare, 2011).

2. Vacuum baselines, trigger bias, and minimum-bias dijets

In $(\Delta\phi \approx \pi)$9 and $\Delta\phi$0Au, dijet-hadron correlations are already nontrivially biased by the trigger. Triggered hard dihadron correlations sit between single-inclusive hadrons and fully reconstructed jets, but the trigger requirement introduces non-trivial complications: kinematic bias selects unusually hard fragmentation, intrinsic $\Delta\phi$1 biases the momentum balance of the away-side parton, and the trigger hadron enhances the quark-jet fraction because quark jets fragment harder than gluon jets (Renk et al., 2011). This is why even vacuum away-side $\Delta\phi$2 is not a simple fragmentation function.

Minimum-bias trigger-associated analyses make this bias structure explicit. In $\Delta\phi$3-$\Delta\phi$4, the trigger is defined as the hadron with the largest transverse rapidity $\Delta\phi$5, and the associated distribution is constructed as a conditional yield $\Delta\phi$6. A hard component extracted from the two-component model is consistent with measured jet fragment systematics derived from $\Delta\phi$7-$\Delta\phi$8 collisions, and the inferred kinematic limits show jet-related triggers down to $\Delta\phi$9 GeV/$\Delta\eta$0, away-side associated hadrons down to $\Delta\eta$1 GeV/$\Delta\eta$2, and same-side associated hadrons down to $\Delta\eta$3 GeV/$\Delta\eta$4 (Trainor et al., 2013). The same work argues that the conventional transverse region in underlying-event analyses is not jet-free, because the triggered dijet contributes substantially there (Trainor et al., 2013).

RHIC dihadron systematics further separate the near-side into a jet-like peak and a ridge. The jet-like component is narrow in both azimuth and pseudorapidity, its energy, system, and particle-composition dependence are consistent with vacuum fragmentation, and data indicate that the jet-like correlation is dominantly produced by vacuum fragmentation (Nattrass, 2010). A specialized vacuum counterpart is the Collins-driven back-to-back dihadron modulation in unpolarized $\Delta\eta$5: the product of two Collins fragmentation functions produces a $\Delta\eta$6 asymmetry predicted to be sizable in the mid-rapidity region for moderate jet transverse momentum (Kang et al., 2010).

3. Long-range structure and dijet coupling in small systems

Small-system measurements show that long-range structure need not be independent of dijet production. In $\Delta\eta$7Au at $\Delta\eta$8 GeV, STAR observed a finite correlated yield at large relative pseudorapidity on the near side, and this yield as a function of $\Delta\eta$9 appears to scale with the dominant, primarily jet-related, away-side yield (Adamczyk et al., 2015). In the same analysis, the ratio $D^0$0 at $D^0$1 was approximately independent of $D^0$2; the linear fit gave a slope

$D^0$3

with $D^0$4, consistent with a constant ratio (Adamczyk et al., 2015).

The harmonic coefficients in $D^0$5Au display a similarly nontrivial pattern. $D^0$6 is approximately inversely proportional to the mid-rapidity multiplicity, $D^0$7 is approximately independent of multiplicity, and $D^0$8 has similar magnitude in the forward and backward directions at large $D^0$9 (Adamczyk et al., 2015). This undermines the assumption that jet contributions are the same in high- and low-activity small-system events. A common subtraction strategy in “double-ridge” analyses treats jets as activity-independent; STAR’s $(\Delta\eta,\Delta\phi)$0Au result shows that the away-side jet component itself varies with event activity and rapidity direction (Adamczyk et al., 2015).

The broader RHIC dihadron program had already established that the near-side ridge is narrow in azimuth but broad in pseudorapidity and roughly independent of pseudorapidity, whereas the jet-like correlation remains narrow in both variables (Nattrass, 2010). Taken together, these measurements constrain any interpretation that treats long-range small-system structure as either purely collective or purely jet-fragmentation-driven. A plausible implication is that long-range pair-wise correlations in small systems are tightly entangled with the presence of dijets.

4. Medium modification in heavy-ion collisions

In nucleus-nucleus collisions, dijet-hadron correlations become a direct probe of jet quenching. ALICE showed that long-range two-particle Fourier coefficients $(\Delta\eta,\Delta\phi)$1 factorize approximately as $(\Delta\eta,\Delta\phi)$2 for $(\Delta\eta,\Delta\phi)$3 when $(\Delta\eta,\Delta\phi)$4 GeV in central Pb–Pb, but that the factorization breaks progressively at higher momenta, quantifying the onset of nonflow dominance due to the away-side jet (Adare, 2011). In the high-$(\Delta\eta,\Delta\phi)$5 regime $(\Delta\eta,\Delta\phi)$6 GeV and $(\Delta\eta,\Delta\phi)$7 GeV, the same analysis found a near-side enhancement $(\Delta\eta,\Delta\phi)$8 and an away-side suppression $(\Delta\eta,\Delta\phi)$9, with $$\Delta\eta=\eta_{\rm assoc}-\eta_{\rm trig},\qquad \Delta\phi=\phi_{\rm assoc}-\phi_{\rm trig}.$$0 showing the same qualitative pattern (Adare, 2011).

ALICE’s broader Pb–Pb di-hadron program reached the same qualitative conclusion. For $$\Delta\eta=\eta_{\rm assoc}-\eta_{\rm trig},\qquad \Delta\phi=\phi_{\rm assoc}-\phi_{\rm trig}.$$1 GeV/$$\Delta\eta=\eta_{\rm assoc}-\eta_{\rm trig},\qquad \Delta\phi=\phi_{\rm assoc}-\phi_{\rm trig}.$$2 and $$\Delta\eta=\eta_{\rm assoc}-\eta_{\rm trig},\qquad \Delta\phi=\phi_{\rm assoc}-\phi_{\rm trig}.$$3 GeV/$$\Delta\eta=\eta_{\rm assoc}-\eta_{\rm trig},\qquad \Delta\phi=\phi_{\rm assoc}-\phi_{\rm trig}.$$4, central Pb–Pb collisions show $$\Delta\eta=\eta_{\rm assoc}-\eta_{\rm trig},\qquad \Delta\phi=\phi_{\rm assoc}-\phi_{\rm trig}.$$5 on the near side and $$\Delta\eta=\eta_{\rm assoc}-\eta_{\rm trig},\qquad \Delta\phi=\phi_{\rm assoc}-\phi_{\rm trig}.$$6 on the away side, while peripheral collisions are consistent with unity on both sides (Zhu, 2013). At lower trigger momentum, the near-side jet shape broadens in $$\Delta\eta=\eta_{\rm assoc}-\eta_{\rm trig},\qquad \Delta\phi=\phi_{\rm assoc}-\phi_{\rm trig}.$$7 from $$\Delta\eta=\eta_{\rm assoc}-\eta_{\rm trig},\qquad \Delta\phi=\phi_{\rm assoc}-\phi_{\rm trig}.$$8 to central Pb–Pb, whereas $$\Delta\eta=\eta_{\rm assoc}-\eta_{\rm trig},\qquad \Delta\phi=\phi_{\rm assoc}-\phi_{\rm trig}.$$9 is approximately independent of centrality within uncertainties (Zhu, 2013).

At RHIC, reconstructed-jet triggers sharpen the picture. STAR jet-hadron correlations in 200 GeV Au–Au show broadening and softening of the away-side jet: the recoil peak is significantly broader in Au+Au than in $d+$0, high-$d+$1 associated yield is suppressed, low-$d+$2 associated yield is enhanced, and the energy-balance observable $d+$3 indicates that much of the lost high-$d+$4 energy reappears at lower $d+$5 (Ohlson, 2011). The same analysis demonstrated that 2+1 correlations preferentially select relatively unmodified jets, because requiring a high-$d+$6 recoil hadron biases the sample toward jets that suffered little modification (Ohlson, 2011).

Theoretical systematics clarify why $d+$7 is not a monotonic proxy for suppression. In hard dihadron modeling, the trigger produces a surface bias, but medium-induced kinematic bias can raise the average parent-parton energy of surviving trigger jets, while parton-type bias filters gluons more strongly than quarks; the net result is that $d+$8 need not be smaller than unity even when $d+$9 (Renk et al., 2011). At RHIC, away-side data favor models with substantial pathlength dependence, whereas at the LHC the harder spectra and larger gluon fraction modify the same balance of geometric and kinematic biases (Renk et al., 2011).

5. Harmonics, medium response, and broadening mechanisms

The decomposition of dijet-hadron correlations into jet peaks and collective structure is model dependent, but several robust patterns recur. In untriggered Pb+Pb hydrodynamics with flux-tube initial conditions, a very broad near-side ridge from asymmetric flow is largest in central collisions and disappears toward peripheral collisions, whereas a peak-like near-side structure associated with very low momentum components of jets is pronounced in peripheral events and vanishes toward central collisions (Werner et al., 2011). This identifies two qualitatively distinct sources of correlation: long-range flow and short-range soft jet remnants.

Transport studies sharpen the same distinction. Within AMPT, harmonic flow, hot spots, and dijets can all lead to double-peaked away-side dihadron azimuthal correlations, but $|\Delta\phi|<\pi/3$0-hadron correlations show a similar double-peak feature while remaining free of the contributions from harmonic flow and hot spots; this makes $|\Delta\phi|<\pi/3$1-hadron correlations a cleaner probe of jet-induced medium excitation in an expanding medium (Li et al., 2010). A common misconception is therefore that a double-hump away side is uniquely a Mach-cone signal; the AMPT decomposition shows that it can also arise from triangular flow and hot-spot expansion (Li et al., 2010).

Species-tagged measurements constrain the long-range sector further. In STAR’s identified-trigger study, the correlated yield in the ridge region is significantly higher for leading non-pions than for pions, the baryon/meson ratio for $|\Delta\phi|<\pi/3$2 is consistent with $|\Delta\phi|<\pi/3$3, but the corresponding ratio for $|\Delta\phi|<\pi/3$4 is $|\Delta\phi|<\pi/3$5, larger than simple constituent-quark scaling would suggest (Collaboration et al., 2014). The same data can be fit either by an azimuthal harmonic model or by a mini-jet modification model, but the latter returns a significantly negative quadrupole term for non-pion triggers, which is difficult to reconcile with a standard elliptic-flow interpretation (Collaboration et al., 2014).

Angular decorrelation measurements provide a more direct path to medium transport coefficients. A Sudakov-resummed treatment of dihadron and hadron-jet angular correlations established a vacuum baseline in $|\Delta\phi|<\pi/3$6 and peripheral $|\Delta\phi|<\pi/3$7, then used the excess decorrelation in central $|\Delta\phi|<\pi/3$8 to extract

$|\Delta\phi|<\pi/3$9

for a quark jet at RHIC top energy, corresponding to

$(\Delta\phi \approx 0)$00

at $(\Delta\phi \approx 0)$01 fm/$(\Delta\phi \approx 0)$02 in central Au–Au (Chen et al., 2016). This shows that dijet-hadron correlations are not only topological probes of ridge and recoil structure but also quantitative probes of transverse-momentum broadening.

6. Extensions: heavy flavor, large rapidity, and small-$(\Delta\phi \approx 0)$03 formalisms

Heavy-flavor trigger proxies extend the same logic to charm jets. In Au+Au at 200 GeV, $(\Delta\phi \approx 0)$04-hadron correlations reveal a jet-like near-side peak at $(\Delta\phi \approx 0)$05, an away-side structure near $(\Delta\phi \approx 0)$06, and a $(\Delta\phi \approx 0)$07-independent quadrupole modulation; the near-side widths in both $(\Delta\phi \approx 0)$08 and $(\Delta\phi \approx 0)$09 broaden from peripheral to central collisions, and the near-side associated yield per $(\Delta\phi \approx 0)$10 increases by about an order of magnitude from 50–80% to 0–20% centrality (Jentsch, 2018). Peripheral $(\Delta\phi \approx 0)$11-hadron correlations are consistent with PYTHIA 8.23, while central collisions show substantial broadening and yield enhancement comparable in trend to light-flavor dihadron correlations (Jentsch, 2018).

At very large rapidity separation, decorrelation is controlled by both BFKL and Sudakov dynamics. In Mueller–Navelet dijet production, Sudakov double logarithms appear when the produced dijets are almost back-to-back, and the resulting resummation must be combined with BFKL evolution: Sudakov suppression is important when the rapidity separation $(\Delta\phi \approx 0)$12 is not too large, while BFKL pomeron exchange dominates when $(\Delta\phi \approx 0)$13 is asymptotically large (Mueller et al., 2015). This suggests that large-$(\Delta\phi \approx 0)$14 dijet-hadron correlations should not be interpreted with a single resummation framework.

Exclusive small-$(\Delta\phi \approx 0)$15 channels provide a complementary limit. In coherent diffractive dijet production in $(\Delta\phi \approx 0)$16-hadron collisions, the cross section depends on the dipole amplitude $(\Delta\phi \approx 0)$17, making the observable sensitive to the color-dipole orientation and to saturation. Unlike the inclusive case, saturation in this diffractive channel leads to stronger azimuthal correlations between the jets, and the $(\Delta\phi \approx 0)$18-distribution exhibits a dip-type structure in the saturation region (Altinoluk et al., 2015). While this is not a standard heavy-ion dijet-hadron observable, it isolates the same small-$(\Delta\phi \approx 0)$19 transverse dynamics that feed inclusive angular correlations.

Across these extensions, the core structure of the field persists. Dijet-hadron correlations remain a hybrid observable: closer to full jets than single-particle suppression, but still governed by trigger bias, fragmentation bias, and background modeling. Their value lies precisely in that intermediate status. They expose how hard scatterings, shower evolution, collective expansion, and hadronization become entangled in QCD matter, and the current literature shows that no single mechanism—vacuum fragmentation, hydrodynamic flow, recombination, or jet-induced broadening—accounts for all observed near-side, away-side, and long-range correlation structures simultaneously (Renk et al., 2011).