High p T Parton Energy Loss from Jet Pair Correlation Observables in the PHENIX Experiment
aa r X i v : . [ nu c l - e x ] S e p High p T Parton Energy Loss from Jet Pair CorrelationObservables in the PHENIX Experiment
Chin-Hao Chen a for the PHENIX collaboration a Department of Physics and Astronomy, Stony Brook University, Stony Brook, NY 11794-3800, USA
Abstract
By controlling several experimental observables, such as momentum and reaction plane, thePHENIX experiment has performed a systematic study of high p T parton energy loss in hotnuclear matter, as well as jet-induced medium modifications, such as the ridge and shoulder.Baseline d + Au measurements are also studied to provide the comparison as a control for coldnuclear matter e ff ect.One of the most striking results made at RHIC is jet quenching in the hot dense nuclearmatter. By using two particle correlations, we are able to study how high p T partons lose energyin the medium. When triggering on high p T particles, the opposing or away-side jet shape varieswith associated p T . At high associated p T , an away-side punch through jet is observed, whichmeans that some part of the jet survived the medium. At intermediate and low associated p T , theaway-side displays a double peak structure, instead of a single peak at ∆ φ ≈ π . Measurementsin p + p and d + Au are used as benchmarks: p + p measures hard scattering with no medium, whiled + Au determines cold nuclear matter e ff ects. An important question is where the lost energygoes. When we look at low p T correlations, we find that not only is the away-side modified, butthe near-side is modified as well. The near-side has an enhancement along ∆ η , or the “ridge”.The away-side has a double peak structure, which is the “shoulder” [1].In d + Au collisions, we use non-identified charged hadrons as both the trigger and associatedparticles and study the two particle azimuthal correlations [2]. The result is compared withp + p collisions. The angular widths of jets in d + Au collisions are consistent with these in p + pcollisions. Comparison of the yields in p + p and d + Au is characterized by the ratio of the pertrigger yield (PTY) in d + Au to the per trigger yield in p + p, or I dAu = ( PT Y dAu ) / ( PT Y pp ). Athigh trigger p T ( > / c), I dAu approaches 1 indicates that when both trigger and associatedparticles are from hard scattering, the yields are consistent with p + p. In d + Au collisions, thereis more enhancement in baryon than in meson [3]. So further correlation study with particleidentification can help to quantify the initial state e ff ect on the particle production mechanism.The observations in d + Au show that the modification of the jet is due to the medium. Toquantify this, we use high p T π triggers and compare central Au + Au collisions (centrality 0-20%) with p + p results. There are several advantages to using π as trigger particles. First, π arewell reconstructed in PHENIX with less background at high p T than charged hadrons. Second, itmay be easier to interpret the results with identified than non identified hadron triggers becauseof the baryon anomaly at intermediate p T [4].We study π -h correlations in central Au + Au and p + p by examing the yields in the away-side [5]. We use I AA = ( PT Y AA ) / ( PT Y pp ) to describe the away-side yield. Fig. 1 shows I AA vs Preprint submitted to Nuclear Physics A November 19, 2018 artner p T in three di ff erent trigger p T bins. The away-side I AA are measured in two di ff erentangular ranges, | ∆ φ − π | < π/ | ∆ φ − π | < π/
6, corresponding to the entire away-side andthe punch-through jet region respectively. We observe a similar suppression pattern at associatedparticle p T > / c in both the head and total away-side region. Also, the away-side width inAu + Au is broader than in p + p at π p T = / c but not as broad at higher π p T . [5]. Sothe away-side shoulder structure is not obvious within the large error bar in present data at high π p T . This may indicate that the whole away-side is suppressed as much as the punch throughjet with high p T triggers. Figure 1: Away-side I AA as a function of partner p T in three trigger p T bins in π − h ± . In order to further explore the away-side suppression shown above, we vary the angle of the π with respect to the reaction plane. In mid-central collisions, we are then able to vary theaverage path length of the away-side jet through the matter. We define the angle between thetrigger particle and the reaction plane as φ S , with φ S = φ S = π out-of-plane.Two particle correlations as a function of φ S provides a sensitive probe of the medium lengthdependence of the parton medium interaction.We use 4-7 GeV / c π triggers, and 3-4 GeV / c non-identified charged hadron as associatedparticles to measure the correlations as a function of φ s [6]. In the 20-60% centrality bin, the pathlengh variation is stronger than central collisions from in to out-of-plane. Fig. 2 shows the away-side per trigger yields (PTY) are measured as a function of φ s . The away-side yield decreaseswith increasing φ s , indicating that the away- side yield decreases from in to out-of-plane. Asimilar trend is observed at higher associated p T (4-5 GeV / c).We fit the φ S dependence with a line, and plot the χ of the fit as a function of the ratio of( PT Y | φ S = π ) / ( PT Y | φ S = ). The results are in Fig. 3. The minimum χ is at ( PT Y | φ S = π ) / ( PT Y | φ S = ) ≈ p T are lowered, we begin probing the medium re-sponse to the jets. In heavy ion collisions, there is an enhancement, known as the ridge, along ∆ η at ∆ φ ≈ ∆ η - ∆ φ correlations. We use inclusive photons at 2-3 GeV / c as triggerparticles and correlate with non-identified charged particles. At this photon p T range, the inclu-sive photons are mostly from meson decay. In heavy ion collisions, the multiplicity increaseswith the number of participants, N part . If the ridge is directly related only to the combinatoricbackground, then the yield of ridge should scale with the underlying events. Fig. 4 shows theratio between the ridge yield and the yield of the underlying event, as a function of partner p T and N part . The ridge yield is a few percent of the underlying events, but for N part > igure 2: Away-side per trigger yield vs φ S in π − h ± Figure 3: χ value vs in-plane to out-of-plane per trig-ger yield ratio. decreases with increasing N part . This means that the ridge yield does not increase as fast as thenumber of particles in the underlying event. The ratio increases with p T indicating that the ridgeis harder than the underlying event. Consequently, the ridge can not be due to combinatorialbackground.The away-side has a double peak structure in central Au + Au collisions. We use three Gaus-sians to fit the away-side and separate the di ff erent components. The punch-through jet is de-scribed by a Gaussian peak at ∆ φ = π . Two additional Gaussian fits, symmetric about π , are usedto describe the jet modification, known as the shoulders. The shoulder yield is the sum of the twoGaussian fits. The spectra of ridge and shoulder in di ff erent centralities are used to compare themedium modification in both near and away-side. Fig. 5 shows the inverse slope extracted fromthe spectra of the ridge and shoulder as a function of N part . This similarity may indicate that boththe ridge and shoulder come from a similar mechanism. The inverse slopes of ridge and shoulderare compared with hard scattering, or p + p, and inclusive charged hadron, from the medium. Bothare softer than hard scattering and are slightly harder than the medium. Consequently, neitherthe ridge nor the shoulder come directly from hard processes, because they are much softer thanp + p. This will give us information on how semi-hard particles interact with the matter. If bothridge and shoulder are from the medium, there must be some mechanism to excite them, to causethe spectra to be harder than that of the medium. part N0 100 200 300 400 n ea r s i d e B ack g r ound ) hD ( n ea r s i d e Je t NN S |<0.7 hD p fD g inc. <3.0 GeV/c T g Th Th Th Figure 4: Ridge to background ratio vs N part part N0 100 200 300 400 [ G e V / c ] s l op e1 ridgeshoulderpp, nearpp, awayinc. hadronAu+Au, p+p = 200 GeV NN S Figure 5: Inverse slope of ridge and shoulder vs N part . If the away-side jet is quenched in the medium, conservation of transverse momentum re-3uires that the momentum must come out somewhere. The near and away-side yield is weightedwith transverse momentum along the direction of the trigger particle to investigate how the p T iscarried by the associated particles in the near and away-sides. Here we did not include the p T ofthe trigger particle. We take the ratio between away and near-side weighted yields as a functionof N part in di ff erent ∆ η regions and plot the result in Fig. 6. At 0 . < | ∆ η | < .
1, the near-side ismostly jet. Both near and away-side yields increase with N part , but the away to near-side ratio oftotal p T is approximately constant with N part This value of 0.55 is consistent with p + p, indicatingthat the p T observed in the near and away-side jets increase together with collision centrality. Asthe collisions become more central, the fraction of the transverse momentum carried by the headregion of the away-side jet decreases, while that in the shoulder increases. This seems to suggestthat transverse momentum lost from the scattered parton is transferred to the shoulder region.At 0 . < | ∆ η | < .
7, the ratios between the away and near-sides are large in p + p and pe-ripheral collisions because of very little near-side yield. At more central collisions, the near-sideincreases because the ridge is the dominant source of particles in this ∆ η range. The associatedparticle p T ratio between shoulder and ridge hold constant along N part at about 0.5. This saysthat the ridge carries about twice the p T of the shoulder. part N0 100 200 300 400 n ea r hD / T , n ea r ass o c p S a w ay hD / T , a w ayass o c p S -h g inc. [2.0,3.0] GeV/c ˛ T g p [1.0,2.0] GeV/c ˛ Th p | < 0.1 near hD away hD part N0 100 200 300 400 n ea r hD / T , n ea r ass o c p S a w ay hD / T , a w ayass o c p S -h g inc. [2.0,3.0] GeV/c ˛ T g p [1.0,2.0] GeV/c ˛ Th p | < 0.7 near hD away hD Figure 6: Ratio of away-side p T over near-side p T vs N part at 0 . < | ∆ η | < . . < | ∆ η | < . In summary, in central Au + Au collision, at high p T , the high p T parton loses energy in themedium and has a strong reaction plane dependence. At low p T , the ridge and shoulder havesimilar inverse slope, which may indicate both have the same origin. The ratio of the p T carriedby the near-side and away-side associated particles at ∆ η ≈ p T loss in themedium comes out from the shoulder. The new high-statistics data sets will allow us to performmore detailed analysis, such as high p T particle correlation with particle identification, to studythe parton energy loss in the medium. References [1] A. Adare et al.,
Phys. Rev. C (2008) 014901.[2] J. Jia [PHENIX collaboration] [arXiv:0906.3776][3] S. S. Adler et al., Phys. Rev. Lett. (2003) 072303[4] S. S. Adler et al., Phys. Rev. Lett. (2003) 172301.[5] A. Adare [PHENIX collaboration] Euro. Phys. J. C (2009) 127.[6] M. McCumber [PHENIX collaboration] [arXiv:0905.0429][7] V. Pantuev [arxiv:nucl-ex / Phys. Rev. C (2008) 034904.(2008) 034904.