Suppression of Upsilon Production in d+Au and Au+Au Collisions at sqrt(s_NN) = 200 GeV
L. Adamczyk, J. K. Adkins, G. Agakishiev, M. M. Aggarwal, Z. Ahammed, I. Alekseev, J. Alford, C. D. Anson, A. Aparin, D. Arkhipkin, E. C. Aschenauer, G. S. Averichev, A. Banerjee, D. R. Beavis, R. Bellwied, A. Bhasin, A. K. Bhati, P. Bhattarai, H. Bichsel, J. Bielcik, J. Bielcikova, L. C. Bland, I. G. Bordyuzhin, W. Borowski, J. Bouchet, A. V. Brandin, S. G. Brovko, S. Bültmann, I. Bunzarov, T. P. Burton, J. Butterworth, H. Caines, M. Calderón de la Barca Sánchez, D. Cebra, R. Cendejas, M. C. Cervantes, P. Chaloupka, Z. Chang, S. Chattopadhyay, H. F. Chen, J. H. Chen, L. Chen, J. Cheng, M. Cherney, A. Chikanian, W. Christie, J. Chwastowski, M. J. M. Codrington, G. Contin, J. G. Cramer, H. J. Crawford, X. Cui, S. Das, A. Davila Leyva, L. C. De Silva, R. R. Debbe, T. G. Dedovich, J. Deng, A. A. Derevschikov, R. Derradi de Souza, S. Dhamija, B. di Ruzza, L. Didenko, C. Dilks, F. Ding, P. Djawotho, X. Dong, J. L. Drachenberg, J. E. Draper, C. M. Du, L. E. Dunkelberger, J. C. Dunlop, L. G. Efimov, J. Engelage, K. S. Engle, G. Eppley, L. Eun, O. Evdokimov, R. Fatemi, S. Fazio, J. Fedorisin, P. Filip, E. Finch, Y. Fisyak, C. E. Flores, C. A. Gagliardi, D. R. Gangadharan, D. Garand, F. Geurts, A. Gibson, M. Girard, S. Gliske, L. Greiner, D. Grosnick, Y. Guo, A. Gupta, S. Gupta, W. Guryn, B. Haag, O. Hajkova, et al. (260 additional authors not shown)
SSuppression of Υ Production in d +Au and Au+Au Collisions at √ s NN = 200 GeV L. Adamczyk, J. K. Adkins, G. Agakishiev, M. M. Aggarwal, Z. Ahammed, I. Alekseev, J. Alford, C. D. Anson, A. Aparin, D. Arkhipkin, E. C. Aschenauer, G. S. Averichev, J. Balewski, A. Banerjee, Z. Barnovska, D. R. Beavis, R. Bellwied, A. Bhasin, A. K. Bhati, P. Bhattarai, H. Bichsel, J. Bielcik, J. Bielcikova, L. C. Bland, I. G. Bordyuzhin, W. Borowski, J. Bouchet, A. V. Brandin, S. G. Brovko, S. B¨ultmann, I. Bunzarov, T. P. Burton, J. Butterworth, H. Caines, M. Calder´on de la Barca S´anchez, D. Cebra, R. Cendejas, M. C. Cervantes, P. Chaloupka, Z. Chang, S. Chattopadhyay, H. F. Chen, J. H. Chen, L. Chen, J. Cheng, M. Cherney, A. Chikanian, W. Christie, J. Chwastowski, M. J. M. Codrington, R. Corliss, J. G. Cramer, H. J. Crawford, X. Cui, S. Das, A. Davila Leyva, L. C. De Silva, R. R. Debbe, T. G. Dedovich, J. Deng, A. A. Derevschikov, R. Derradi de Souza, S. Dhamija, B. di Ruzza, L. Didenko, C. Dilks, F. Ding, P. Djawotho, X. Dong, J. L. Drachenberg, J. E. Draper, C. M. Du, L. E. Dunkelberger, J. C. Dunlop, L. G. Efimov, J. Engelage, K. S. Engle, G. Eppley, L. Eun, O. Evdokimov, R. Fatemi, S. Fazio, J. Fedorisin, P. Filip, E. Finch, Y. Fisyak, C. E. Flores, C. A. Gagliardi, D. R. Gangadharan, D. Garand, F. Geurts, A. Gibson, M. Girard, S. Gliske, D. Grosnick, Y. Guo, A. Gupta, S. Gupta, W. Guryn, B. Haag, O. Hajkova, A. Hamed, L-X. Han, R. Haque, J. W. Harris, J. P. Hays-Wehle, S. Heppelmann, K. Hill, A. Hirsch, G. W. Hoffmann, D. J. Hofman, S. Horvat, B. Huang, H. Z. Huang, P. Huck, T. J. Humanic, G. Igo, W. W. Jacobs, H. Jang, E. G. Judd, S. Kabana, D. Kalinkin, K. Kang, K. Kauder, H. W. Ke, D. Keane, A. Kechechyan, A. Kesich, Z. H. Khan, D. P. Kikola, I. Kisel, A. Kisiel, D. D. Koetke, T. Kollegger, J. Konzer, I. Koralt, W. Korsch, L. Kotchenda, P. Kravtsov, K. Krueger, I. Kulakov, L. Kumar, R. A. Kycia, M. A. C. Lamont, J. M. Landgraf, K. D. Landry, J. Lauret, A. Lebedev, R. Lednicky, J. H. Lee, W. Leight, M. J. LeVine, C. Li, W. Li, X. Li, X. Li, Y. Li, Z. M. Li, L. M. Lima, M. A. Lisa, F. Liu, T. Ljubicic, W. J. Llope, R. S. Longacre, X. Luo, G. L. Ma, Y. G. Ma, D. M. M. D. Madagodagettige Don, D. P. Mahapatra, R. Majka, S. Margetis, C. Markert, H. Masui, H. S. Matis, D. McDonald, T. S. McShane, N. G. Minaev, S. Mioduszewski, B. Mohanty, M. M. Mondal, D. A. Morozov, M. G. Munhoz, M. K. Mustafa, B. K. Nandi, Md. Nasim, T. K. Nayak, J. M. Nelson, L. V. Nogach, S. Y. Noh, J. Novak, S. B. Nurushev, G. Odyniec, A. Ogawa, K. Oh, A. Ohlson, V. Okorokov, E. W. Oldag, R. A. N. Oliveira, M. Pachr, B. S. Page, S. K. Pal, Y. X. Pan, Y. Pandit, Y. Panebratsev, T. Pawlak, B. Pawlik, H. Pei, C. Perkins, W. Peryt, A. Peterson, P. Pile, M. Planinic, J. Pluta, D. Plyku, N. Poljak, J. Porter, A. M. Poskanzer, N. K. Pruthi, M. Przybycien, P. R. Pujahari, H. Qiu, A. Quintero, S. Ramachandran, R. Raniwala, S. Raniwala, R. L. Ray, C. K. Riley, H. G. Ritter, J. B. Roberts, O. V. Rogachevskiy, J. L. Romero, J. F. Ross, A. Roy, L. Ruan, J. Rusnak, N. R. Sahoo, P. K. Sahu, I. Sakrejda, S. Salur, A. Sandacz, J. Sandweiss, E. Sangaline, A. Sarkar, J. Schambach, R. P. Scharenberg, A. M. Schmah, W. B. Schmidke, N. Schmitz, J. Seger, P. Seyboth, N. Shah, E. Shahaliev, P. V. Shanmuganathan, M. Shao, B. Sharma, W. Q. Shen, S. S. Shi, Q. Y. Shou, E. P. Sichtermann, R. N. Singaraju, M. J. Skoby, D. Smirnov, N. Smirnov, D. Solanki, P. Sorensen, U. G. deSouza, H. M. Spinka, B. Srivastava, T. D. S. Stanislaus, J. R. Stevens, R. Stock, M. Strikhanov, B. Stringfellow, A. A. P. Suaide, M. Sumbera, X. Sun, X. M. Sun, Y. Sun, Z. Sun, B. Surrow, D. N. Svirida, T. J. M. Symons, A. Szanto de Toledo, J. Takahashi, A. H. Tang, Z. Tang, T. Tarnowsky, J. H. Thomas, A. R. Timmins, D. Tlusty, M. Tokarev, S. Trentalange, R. E. Tribble, P. Tribedy, B. A. Trzeciak, O. D. Tsai, J. Turnau, T. Ullrich, D. G. Underwood, G. Van Buren, G. van Nieuwenhuizen, J. A. Vanfossen, Jr., R. Varma, G. M. S. Vasconcelos, A. N. Vasiliev, R. Vertesi, F. Videbæk, Y. P. Viyogi, S. Vokal, A. Vossen, M. Wada, M. Walker, F. Wang, G. Wang, H. Wang, J. S. Wang, X. L. Wang, Y. Wang, Y. Wang, G. Webb, J. C. Webb, G. D. Westfall, H. Wieman, G. Wimsatt, S. W. Wissink, R. Witt, Y. F. Wu, Z. Xiao, W. Xie, K. Xin, H. Xu, N. Xu, Q. H. Xu, Y. Xu, Z. Xu, W. Yan, C. Yang, Y. Yang, Y. Yang, Z. Ye, P. Yepes, L. Yi, K. Yip, I-K. Yoo, Y. Zawisza, H. Zbroszczyk, W. Zha, Zhang, J. B. Zhang, S. Zhang, X. P. Zhang, Y. Zhang, Z. P. Zhang, F. Zhao, J. Zhao, C. Zhong, X. Zhu, Y. H. Zhu, Y. Zoulkarneeva, and M. Zyzak (STAR Collaboration) a r X i v : . [ nu c l - e x ] J a n AGH University of Science and Technology, Cracow, Poland Argonne National Laboratory, Argonne, Illinois 60439, USA University of Birmingham, Birmingham, United Kingdom Brookhaven National Laboratory, Upton, New York 11973, USA University of California, Berkeley, California 94720, USA University of California, Davis, California 95616, USA University of California, Los Angeles, California 90095, USA Universidade Estadual de Campinas, Sao Paulo, Brazil Central China Normal University (HZNU), Wuhan 430079, China University of Illinois at Chicago, Chicago, Illinois 60607, USA Cracow University of Technology, Cracow, Poland Creighton University, Omaha, Nebraska 68178, USA Czech Technical University in Prague, FNSPE, Prague, 115 19, Czech Republic Nuclear Physics Institute AS CR, 250 68 ˇReˇz/Prague, Czech Republic Frankfurt Institute for Advanced Studies FIAS, Germany Institute of Physics, Bhubaneswar 751005, India Indian Institute of Technology, Mumbai, India Indiana University, Bloomington, Indiana 47408, USA Alikhanov Institute for Theoretical and Experimental Physics, Moscow, Russia University of Jammu, Jammu 180001, India Joint Institute for Nuclear Research, Dubna, 141 980, Russia Kent State University, Kent, Ohio 44242, USA University of Kentucky, Lexington, Kentucky, 40506-0055, USA Korea Institute of Science and Technology Information, Daejeon, Korea Institute of Modern Physics, Lanzhou, China Lawrence Berkeley National Laboratory, Berkeley, California 94720, USA Massachusetts Institute of Technology, Cambridge, MA 02139-4307, USA Max-Planck-Institut f¨ur Physik, Munich, Germany Michigan State University, East Lansing, Michigan 48824, USA Moscow Engineering Physics Institute, Moscow Russia National Institute of Science Education and Research, Bhubaneswar 751005, India Ohio State University, Columbus, Ohio 43210, USA Old Dominion University, Norfolk, VA, 23529, USA Institute of Nuclear Physics PAN, Cracow, Poland Panjab University, Chandigarh 160014, India Pennsylvania State University, University Park, Pennsylvania 16802, USA Institute of High Energy Physics, Protvino, Russia Purdue University, West Lafayette, Indiana 47907, USA Pusan National University, Pusan, Republic of Korea University of Rajasthan, Jaipur 302004, India Rice University, Houston, Texas 77251, USA Universidade de Sao Paulo, Sao Paulo, Brazil University of Science & Technology of China, Hefei 230026, China Shandong University, Jinan, Shandong 250100, China Shanghai Institute of Applied Physics, Shanghai 201800, China SUBATECH, Nantes, France Temple University, Philadelphia, Pennsylvania, 19122, USA Texas A&M University, College Station, Texas 77843, USA University of Texas, Austin, Texas 78712, USA University of Houston, Houston, TX, 77204, USA Tsinghua University, Beijing 100084, China United States Naval Academy, Annapolis, MD 21402, USA Valparaiso University, Valparaiso, Indiana 46383, USA Variable Energy Cyclotron Centre, Kolkata 700064, India Warsaw University of Technology, Warsaw, Poland University of Washington, Seattle, Washington 98195, USA Yale University, New Haven, Connecticut 06520, USA University of Zagreb, Zagreb, HR-10002, Croatia
We report measurements of Υ meson production in p + p , d +Au, and Au+Au collisions using theSTAR detector at RHIC. We compare the Υ yield to the measured cross section in p + p collisions inorder to quantify any modifications of the yield in cold nuclear matter using d +Au data and in hotnuclear matter using Au+Au data separated into three centrality classes. Our p + p measurement isbased on three times the statistics of our previous result. We obtain a nuclear modification factor for Υ(1S+2S+3S) in the rapidity range | y | < d +Au collisions of R d Au = 0 . ± . . ) ± . . ) ± . p + p sys . ). A comparison with models including shadowing and initial state partonenergy loss indicates the presence of additional cold-nuclear matter suppression. Similarly, in thetop 10% most-central Au+Au collisions, we measure a nuclear modification factor of R AA = 0 . ± . . ) ± . . ) ± . p + p sys . ), which is a larger suppression factor than that seen in coldnuclear matter. Our results are consistent with complete suppression of excited-state Υ mesons inAu+Au collisions. The additional suppression in Au+Au is consistent with the level expected inmodel calculations that include the presence of a hot, deconfined Quark-Gluon Plasma. However,understanding the suppression seen in d +Au is still needed before any definitive statements aboutthe nature of the suppression in Au+Au can be made. PACS numbers: 25.75.Cj, 25.75.Nq, 14.40.Pq, 25.75.-q, 12.38.MhKeywords: Upsilon suppression, Quarkonium In-Medium, Relativistic Heavy-ion Collisions, STAR
INTRODUCTION
In the study of the properties of the Quark-GluonPlasma (QGP) an extensive effort has been devoted tomeasuring quarkonium yields since these have been pre-dicted to be sensitive to color deconfinement [1]. Studieshave mainly focused on charmonium, but with the highcollision energies available at the Relativistic Heavy IonCollider (RHIC) and the Large Hadron Collider (LHC)we can now study bottomonium in hot nuclear matterwith sufficient statistics. For a recent review of quarko-nium in-medium, see e.g. Ref. [2], Sec. 5. One predictionis that excited quarkonium states are expected to disso-ciate at or above temperatures near that of the crossoverto the deconfined QGP phase, T c ≈ −
190 MeV [3–5]. The more tightly bound ground states are expectedto dissociate at even higher temperatures. The detailsof the temperature dependence of the dissociation of theexcited states and of the feed-down pattern of the ex-cited states into the ground state lead to a sequentialsuppression pattern of the inclusive upsilon states withincreasing temperature [6]. The binding energy of theΥ(2S) state ( ∼
540 MeV) is about half that of the Υ(1S)state ( ∼ ∼
200 MeV. Recent studies take into account not only theDebye screening effect on the heavy quark potential butalso an imaginary part of the potential which modifiesthe widths of the various quarkonia states (e.g. [7–10]).In Ref. [8] it is estimated that the Υ(2S) state will meltat a temperature of T ≈
250 MeV, whereas the groundstate Υ(1S) will melt at temperatures near T ≈
450 MeV.We focus here on the measurement of bottomoniummesons in collisions at √ s NN =200 GeV. An observationof suppression in the bottomonium sector in hot nuclearmatter is important for two reasons. First, it would beevidence for color deconfinement in the produced mattersince the aforementioned effects are all ultimately basedon studies of the high temperature phase of QuantumChromodynamics (QCD) done on the lattice, where coloris an active degree of freedom. Second, bottomoniumsuppression provides a way to estimate model-dependentbounds on the temperature with the bounds dependingon the particular suppression pattern seen. The cross section for bottomonium production issmaller than that of charmonium [11–13] making the ex-perimental study of Υ production challenging. However,the theoretical interpretation of bottomonium suppres-sion is less complicated than that of charmonium for sev-eral reasons. While charmonium production at RHICand higher energies can be affected by the statistical re-combination of charm quarks that are produced in dif-ferent nucleon-nucleon collisions within the same nuclearinteraction event, this effect is negligible for bottomo-nium due to the much smaller b ¯ b production cross sec-tion ( σ b ¯ b is measured to be in the range 1.34 – 1.84 µ b[14] and calculated to be 1 . +0 . − . µ b [15], compared to σ c ¯ c ≈
550 – 1400 µ b [16, 17]). Another complication inthe charmonium case is that even in a purely hadronicscenario, charmonium mesons can be suppressed due totheir interaction with hadronic co-movers [18, 19]. Thecross section for inelastic collisions of Υ(1S) with hadronsis small [20]. Hence, absorption in the medium by theabundantly produced co-moving hadrons is predicted tobe minimal. The cold-nuclear-matter (CNM) effects onΥ production, which are those seen in p +A collisions andcan be due to shadowing of the parton distribution func-tions in the nucleus or energy-loss in the nucleus, can stillbe important. There is evidence of some Υ suppressionin fixed target experiments at 800 GeV/c lab momentumfrom E772 [21]. However, the CNM suppression observedfor Υ is smaller than that for J/ψ reported by NA50 [22].For all these reasons, the Υ family is expected to be acleaner and more direct probe of hot QCD, and of thecorresponding color deconfinement effects.In this letter we present measurements of Υ produc-tion in p + p , d +Au, and Au+Au collisions at √ s NN =200GeV via the e + e − decay channel obtained by the STARexperiment. We extract invariant cross sections for allthree collision systems studied. Using this p + p measure-ment as a baseline we obtain the nuclear modificationfactor ( R d Au and R AA ) of the three states combined:Υ(1S+2S+3S). The ratio R AA is used to quantify de-viations of the yields in d +Au and Au+Au compared tothose expected from a superposition of elementary p + p collisions. The data were taken during 2008 ( d +Au),2009 ( p + p ) and 2010 (Au+Au) at RHIC, and correspondto integrated luminosities of 28.2 nb − , 20.0 pb − , and1.08 nb − , respectively. All three datasets were takenwith the same detector configuration. For this reasonthe data from our previous p+p result (2006) was notincluded in this analysis; the amount of material in thedetector at that point was substantially larger than itwas in the three datasets discussed here. We compareour data to model calculations of the cross section basedon perturbative QCD (pQCD) [23], and to recent modelsof Υ production in d +Au and Au+Au collisions [24–29]. EXPERIMENTAL METHODS
The main detectors used are the STAR Time Projec-tion Chamber (TPC) [30] for tracking and the STAR Bar-rel Electro-Magnetic Calorimeter (BEMC) [31] for trig-gering. Both the TPC and BEMC are used for particleidentification. The starting point is the STAR Υ triggerwhose main components are a fast hardware Level-0 (L0)trigger, which fires when a tower in the BEMC has en-ergy E L − BEMC ≥ . > > ◦ and an invariant mass above 5 GeV /c (6.5 GeV /c )in p + p ( d +Au). Note that energy measured at the trig-gering level is partially calibrated leading to small butrandom biases. Hence, triggering thresholds are not pre-cise in energy. The Υ trigger is required to be in coin-cidence with the STAR minimum bias trigger. For p + p collisions the minimum bias trigger is based on the STARBeam-Beam Counters, while for d +Au and Au+Au it isbased on the STAR Zero-Degree Calorimeters (ZDC) andthe Vertex-Position Detectors (VPD). The L0-L2 combi-nation was used for the d +Au data in 2008 and for p + p data in 2009. In the Au+Au 2010 run, an upgrade tothe STAR data acquisition system allowed the processingof all the L0 triggers above the E L − BEMC = 4 . N part ) and number of bi-nary collisions ( N coll ). The trigger, tracking, and electronidentification efficiencies in the Au+Au case were studiedas a function of centrality (see Tab. I). The presence ofthe underlying Au+Au event background increases theenergy measured in the calorimeter towers and results ina slight increase in the trigger efficiency with increasing Centrality N part Rapidity Efficiency0-60% 162 ± | y | < . . < | y | < . | y | < . ± | y | < . . < | y | < . | y | < . ± | y | < . . < | y | < . | y | < . ± | y | < . . < | y | < . | y | < . N part (more central collisions). Similarly, the increasein the track density in the TPC results in a decrease inthe tracking efficiency which is especially noticeable athigh N part . We used the specific ionization of the tracksin the TPC gas ( dE/dx ) for electron identification. Inaddition, the projection of the track onto the locationof the BEMC shower maximum position was requiredto match the measured BEMC cluster position. Once atrack was matched to a calorimeter cluster, the ratio ofthe energy of the cluster to the TPC momentum ( E/p )was also used for electron identification. The combinedacceptance times efficiency for detecting an Υ at mid-rapidity ( | y | < .
5) in Au+Au taking into account allaforementioned effects was found to vary from ∼
12% inperipheral collisions to ∼
9% in central collisions.The cuts used in these analyses were chosen such thatthe tracking and electron identification efficiencies wouldbe similar across the three datasets, allowing the system-atic uncertainties to approximately cancel in the mea-surement of R AA . For further detail, the techniques usedin these Υ measurements were described extensively forour previous p + p measurement [13] based on a 7.9 pb − dataset. All evaluated sources of systematic uncertaintyare summarized in Tab. II. An important effect in addi-tion to those discussed in [13] is the change in trackingand mass resolution with increasing detector occupancy.Simulated Υ events were embedded in real data and theirreconstructed line shapes were studied as a function ofcollision system and detector occupancy. In p + p colli-sions, we find a mass resolution of 1.3% for reconstructedΥ(1S). Due to additional TPC alignment errors for the d +Au and Au+Au data the mass resolutions of the Υ(1S)increased to 2.7% in d +Au and peripheral Au+Au colli-sions and 2.9% in central collisions. This decreased massresolution was accounted for in the binary scaling esti-mates of Υ(1S+2S+3S) yields (see gray bands in Figs. 1and 4). Systematic uncertainties in those scaling esti-mates (line shapes) are included in the errors in Table II. Source Relative uncertaintyLuminosity, Vernier scan ( p + p ) ± ± d +Au) ± p + p ) ± d +Au) ± . ± . d +Au min. bias σ ± . , − . . − . . − . E cluster +1 . − . θ cut ∼ ∼ ± × . E/p cut efficiency ± dE/dx cut efficiency ± × . d +Au excited state ratio +0%, -2%Au+Au excited state ratio +1% , − p + p line shape +6 . − . d +Au line shape +1 . − . . − . p ⊥ shape ± . d +Au and Au+Au +0%, -7.5%Total syst., σ pp +21 . − . σ d Au +17 . − . σ AuAu +16 . − . . , − . R d Au , syst. +3 . − . R d Au (1S), syst. +3 . − . R AA , syst. +3 . − . R AA (1S), syst. +3 . − . For all results we quote, the Υ data are integrated overall transverse momenta.
RESULTS AND DISCUSSION
Figure 1 shows the invariant mass distributions of elec-tron pairs for p + p (top) and d +Au (bottom) in the kine-matic region | y Υ | < .
5. Unlike-sign pairs are shown asred filled circles and like-sign pairs as hollow blue circles.The data are fit with a parameterization consistingof the sum of various contributions to the electron-pairinvariant-mass spectrum. The fit is performed simulta-neously with the like-sign and unlike-sign spectra using amaximum-likelihood method. The lines in Fig. 1 showthe yield from the combinatorial background (dashedblue line), the result of adding the physics backgroundfrom Drell-Yan and b ¯ b pairs (dot-dashed green line), andfinally the inclusion of the Υ contribution (solid red line).The shape of the Drell-Yan continuum is obtained viaa next-to-leading order (NLO) pQCD calculation fromVogt [32]. PYTHIA 8 was used to calculate the shape of ) (GeV/c ee m8 9 10 11 12 13 C oun t s STAR p+p = 200 GeVs |<0.5 ee |y - - +N + + N Comb. Background (CB)bCB + Drell-Yan + b(1S+2S+3S) ¡ + bCB + DY + b + - N ) (GeV/c ee m8 9 10 11 12 13 C oun t s - - +N + + N Comb. Background (CB)bCB + Drell-Yan + b(1S+2S+3S) ¡ + bCB + DY + b> coll 5. (a): p + p . (b): d +Au. Unlike-sign pairs are shown as red filled circles and like-sign pairs ashollow blue circles. The gray band shows the expected yield if R d Au =1 including resolution effects. See text for descriptionof yield extraction. the b ¯ b contribution [33]. We model each of the Υ stateswith a Crystal Ball function [34], which incorporates de-tector resolution and losses from bremsstrahlung in thedetector material.The fit is done to the unlike and like-sign data si-multaneously. The fit to the combinatorial backgroundcomponent extracted from the like-sign data is shared bythe functional form used to parameterize the unlike-signdata. In the usual like-sign subtraction procedure someinformation would be lost. In contrast, by performing asimultaneous fit to both the like-sign and unlike-sign sig-nals we optimize the statistical power of our data. TheL2 trigger condition has the effect of cutting off the lowerinvariant masses. This cut-off shape is parameterized inthe fits using an error function.We integrate the unlike-sign invariant mass distribu-tion in the region 8 . − 11 GeV /c and subtract fromthe data the fit to the combinatorial, Drell-Yan, and b ¯ b background components in order to obtain the yieldof Υ(1S+2S+3S). After accounting for efficiencies andsampled luminosity, we calculate a production cross sec-tion in p + p collisions of: B ee × dσ/dy | | y | < . = 64 ± . + fit) +14 − pb. Our previous result of 114 ± +23 − pb [13] is consistent with our new measurement. Thegreater sampled luminosity and decreased detector ma-terial in 2009 led to improved statistics and lower sys-tematic uncertainties in the present measurement.In Fig. 1(b), the gray band shows the expected sig-nal from the p + p data scaled by the number of binarycollisions. Due to differences in detector occupancy anddetector calibrations the width of the Υ signal differs be-tween collision systems and centralities. As discussed inthe previous section, a misalignment in the TPC in the d +Au and Au+Au datasets led to a broadening of theΥ line shapes compared to the p + p dataset. This canbe seen by examining the line shapes for the Υ states inFig. 1(a) ( p + p ) and Fig. 1(b) ( d +Au). The average de-tector occupancy is comparable between the two systems,however the d +Au dataset has a noticeably broader lineshape due to the aforementioned differences in calibra-tion. The effects of the broadening of the line shapes aretaken into account in systematic uncertainties (Tab. II).The comparison of the gray band with the d +Au data inpanel (b) indicates a suppression of Υ production withrespect to binary-collision scaling.A similar procedure is followed for the region 0 . < | y Υ | < p + p collisions. We combine the results toobtain the differential cross section: B ee × dσ/dy | | y | < =58 ± . + fit) +12 − pb. In d +Au collisions, we ana-lyze the yields separately in the regions − < y Υ < − . . < y Υ < d +Au system is notsymmetric about y = 0. Hence, averaging between for-ward and reverse rapidities is not warranted as it is in p + p . Throughout this paper, the positive rapidity re-gion is the deuteron-going direction, and the negativerapidity region is the Au-going direction. Integratingover our measured range ( | y Υ | < d +Au collisions is found to be B ee × dσ/dy | | y | < =19 ± . + fit) ± . ) nb.We extract the Υ(1S) yield directly by integratingover a narrower mass window (8.8-9.8 GeV/c ). Thismass window was chosen due to its high acceptance ratefor Υ(1S) and its high rejection rate for the excitedstates. To account for sensitivity to the shape of the Υsignal, we varied the parameters of the line shape ob-tained from simulations and data-driven methods dis-cussed previously by their measured uncertainties andvaried the excited states from unsuppressed to completelysuppressed. We then recalculated both efficiency and pu- rity (see Tab. II, Υ(1S) purity). Those variations weretaken into account as additional systematics when quot-ing Υ(1S) results.Figure 2(a) shows the extracted Υ(1S+2S+3S) crosssection for p + p and d +Au as a function of rapidity. The p + p measurements are shown as blue stars and the d +Aumeasurements as red circles. The p + p result in the re-gion 0 . < | y Υ | < . y = 0 . y = − . 75 to illustrate thatthe latter is not an independent measurement. The datafrom PHENIX at forward rapidity for p + p (filled blue di-amonds) and d +Au (hollow red diamonds) are also shown[35].The cross sections in p + p are compared to an NLOpQCD calculation of Υ production in the Color Evapo-ration Model (CEM) [23], which is consistent with ourdata within the statistical and systematic uncertainties.The same calculation is performed for d +Au includingshadowing effects [24]. The EPS09 set of nuclear Par-ton Distribution Functions (nPDF) [36] were used. Themodel is in agreement with our data except for the mid-rapidity point which is lower than the prediction. Tostudy this observation for d +Au further, we make a closercomparison to models and to previous measurements ofΥ production in p+A collisions.To focus on expected shadowing effects, we obtain thenuclear modification factor R d Au as a function of rapid-ity. The nuclear modification factor is defined in nucleus-nucleus collisions as R AA = 1 σ AA σ pp × < N coll > × B ee × (cid:16) dσ AA dy (cid:17) Υ B ee × (cid:16) dσ pp dy (cid:17) Υ where the first factor accounts for the difference in in-elastic cross section in p + p to d +Au or Au+Au colli-sions. The second factor accounts for the average num-ber of nucleon collisions in a d +Au or Au+Au collisionas calculated by a Glauber model. The third factor ac-counts for the measured Υ production in p + p , d +Au orAu+Au. We used the following total inelastic cross sec-tions: σ pp = 42 mb, σ d Au = 2 . σ AuAu = 6 b.Our results for R d Au are shown in Fig. 2(b) and sum-marized in Tab. III. Our data (red stars) are compared toCEM calculations with the uncertainty from the EPS09nPDF shown as the shaded region. Note that this predic-tion for R d Au , which includes modification of the nuclearPDFs but does not include absorption, implies a modi-fication factor of R d Au ≈ . 1. A calculation in Ref. [25]explored various nPDFs (EKS98, EPS08, and nDSg) andalso gave R d Au values above 1 with enhancements in therange of 5-20%. The models are in agreement with thedata except in the y ∼ System Centrality States Rapidity R AA,dA d +Au Min. bias 1S+2S+3S -1.0 < y Υ < -0.5 0.84 ± ± ± ± | y Υ | < ± ± ± ± . < y Υ < . ± ± ± ± | y Υ | < ± ± ± ± < y Υ < -0.5 0.74 ± ± . +0 . − . ± | y Υ | < ± ± . +0 . − . ± < y Υ < ± ± . +0 . − . ± | y Υ | < ± ± . +0 . − . ± | y Υ | < ± ± ± ± | y Υ | < ± ± ± ± | y Υ | < ± ± . +0 . − . ± | y Υ | < ± ± . +0 . − . ± | y Υ | < ± ± ± ± | y Υ | < ± ± ± ± | y Υ | < ± ± . +0 . − . ± | y Υ | < ± ± . +0 . − . ± | y Υ | < ± ± ± ± | y Υ | < ± ± ± ± | y Υ | < ± ± . +0 . − . ± | y Υ | < ± ± . +0 . − . ± | y Υ | < ± ± ± ± | y Υ | < ± ± ± ± | y Υ | < ± ± . +0 . − . ± | y Υ | < ± ± . +0 . − . ± R d Au and R AA results. The results are listed in the form a ± b ± c ± d ± e where a is R d Au or R AA , b is the d +Au or Au+Au statistical uncertainty, c is the p + p statistical uncertainty, d is the d +Au or Au+Au systematic uncertainty,and e is the p + p systematic uncertainty. calculation for a combination of energy loss and shad-owing using EPS09 is shown as the dashed-dotted line.The energy-loss model is also in agreement with the dataexcept for the mid-rapidity point. The model from [26]does not include absorption from interactions with spec-tator nucleons. However, those effects only play a role inthe rapidity region y (cid:46) . 2, where the Υ mesons would becloser to the frame of the Au spectators. Therefore, thesuppression at mid-rapidity is indicative of effects beyondshadowing, initial-state parton energy loss, or absorptionby spectator nucleons.We compare our measurements with results from E772at √ s NN = 40 GeV, where suppression of the Υ states in p + A was observed. This is illustrated in Fig. 3(a), whichshows the ratio of the cross section in d +Au collisionsfor STAR ( p + A for E772) to that of p + p collisions nor-malized by the mass number A . E772 plotted a ratio ofextracted cross sections normalized by the data where theproton beam hit a liquid deuterium target ( A = 2). As-suming that the cross section scales as σ pA = A α σ pp , andusing their p + d result as the baseline, the solid line showsthat the ratio should scale as ( A/ α − . Our measure-ment in d +Au for the Υ(1S) state (red star) is consistentwith the fit to the E772 data, shown as hollow blue circlesfor Υ(1S) and hollow green squares for Υ(2S+3S). Ourresults cover the rapidity range | y | < < y < . α value as a function of Feynman-x ( x F ) in Fig. 3(b).The larger suppression we observe at mid-rapidity is alsoconsistent with the larger suppression seen in E772 for x F ∼ b ¯ b pairs (dot-dashed green line), and the Υ contribu-tion (solid red line). The absence of the L2 trigger inthe Au+Au dataset removes the cut-off effect. One cantherefore see the background (modeled as the sum of twoexponentials), dominated by the combinatorial compo-nent, rising at lower invariant mass. Measured cross sec-tions are summarized in Tab. IV. The gray bands in theAu+Au figure illustrate the expected signal from the p + p data scaled by the number of binary collisions. There isa clear suppression of the expected yield in Au+Au col-lisions.This suppression is quantified in Fig. 5, which displaysthe nuclear modification factor, R AA , plotted as a func-tion of N part with the 0-10% most-central collisions cor-responding to (cid:104) N part (cid:105) = 326 ± 4. Figure 5(a) shows thedata for all three states in the rapidity range | y | < | y | < . R AA and R d Au for the ground state Υ(1S) ¡ y-2.5 -2 -1.5 -1 -0.5 0 0.5 1 1.5 2 2.5 / d y ( pb ) s d · - l + l B ( S + S + S ) , ¡ NN s - l + l fi (1S+2S+3S) ¡ (a) STAR ppdAu / 1000 PHENIX ppdAu / 1000 R. Vogt NLO pQCD CEMppdAu / 1000 ¡ y-2.5 -2 -1.5 -1 -0.5 0 0.5 1 1.5 2 2.5 d A u ( S + S + S ) R ¡ dAu STAR R STAR p+p Syst. UncertaintyPHENIXShadowing, EPS09 (Vogt) )eEnergy Loss (Arleo, PeignEnergy Loss + EPS09 (b) FIG. 2: (Color online) (a) B ee × dσ/dy vs. y for p + p colli-sions (blue stars) and for d +Au collisions (red, filled circles;scaled down by 10 ). Note that the hollow star at y = − . y = 0 . 75 since these are notindependent measurements. Results obtained by PHENIXare shown as filled diamonds. Systematic errors are shownas boxes around the data. The shaded bands are from next-to-leading-order pQCD color evaporation model calculations.The d +Au prediction uses the EPS09 nPDF which includesshadowing [24]. (b) R d Au vs. y for STAR (red stars) andPHENIX (green diamonds) results. The band on the rightshows the overall normalization uncertainty for the STAR re-sults due to systematic uncertainties in the p + p measurement.The shaded band shows the prediction for R d Au from EPS09and its uncertainty. The dashed curve shows suppression dueto initial-state parton energy loss and the dot-dashed curveshows the same model with EPS09 incorporated [26]. Mass Number (A) , / ) pd s A ( p A s pp s A / d A s scaling A C Ca Fe WAu (a) = 40 GeV, 0 0. The data confirm that bottomo-nia are indeed suppressed in d +Au and in Au+Au col-lisions. For d +Au collisions, we find R d Au (1 S + 2 S +3 S ) = 0 . ± . 14 ( d +Au stat . ) ± . 10 ( p + p stat . ) ± . 03 ( d +Au syst . ) ± . p + p syst . ) in the range | y | < p + p collisions of42 mb, for d +Au collisions of 2.2 b, and (cid:104) N coll (cid:105) = 7 . ± . ) (GeV/c ee m C oun t s = 200 GeV NN sAu+Au |<0.5, Run 10 ee |y - - +N + + N Comb. Background (CB)bCB + Drell-Yan + b(1S+2S+3S) ¡ + bCB + DY + b> coll 13 (Au+Au stat . ) ± . 07 ( p + p stat . ) ± . 02 (Au+Au syst . ) ± . 06 ( p + p syst . ),which is ≈ . σ away from unity. The results are sum-marized in Tab. III.In the narrower rapidity range (Fig. 5(b)), we see anindication of a lower R d Au as discussed earlier. Our dataand the E772 data show a larger suppression at y ∼ x F ∼ | y | < . d +Au up to central Au+Aucollisions. This suggests that suppression in d +Au in thiskinematic range needs to be understood before interpret-ing the suppression in Au+Au.For d +Au collisions we find R d Au (1S) = 0 . ± . 15 ( d +Au stat . ) ± . 11 ( p + p stat . ) +0 . − . ( d +Au syst . ) ± . p + p syst . ) in the range | y | < . 0. Forthe 0-10% most-central collisions we find R AA (1 S ) = 0 . ± . 13 (Au+Au stat . ) ± . 10 ( p + p stat . ) +0 . − . (Au+Au syst . ) ± . 08 ( p + p syst . ).Similar suppression is found by CMS in Pb+Pb collisions( R AA (1S) ≈ . 45 at similar N part ) [37–39]. We observethe nuclear modification factor for the Υ(1S) as afunction of N part to be consistent with unity in d +Authrough mid-central Au+Au collisions (see Fig. 5c). Inthe most central Au+Au collisions, we see an indicationof suppression of the Υ(1S) at the 2 . σ level. In the con-text of suppression of the excited states, if the feed-downfraction remains ∼ 49% as measured at higher energiesand high- p ⊥ it is possible that an R AA (1S) as low as0.51 could be due solely to suppression of the excitedstates [43].One can relate the R AA of the combined states to thatof the ground state via the equation R AA (1 S +2 S +3 S ) = R AA (1 S ) × (1 + N AA (2 S + 3 S ) /N AA (1 S )) / (1 + N pp (2 S +3 S ) /N pp (1 S )). The ratio of the excited states to theground state can be obtained from measurements byCMS and Fermilab experiments [40, 41] and alterna-tively from combining theoretical calculations [23] withmeasured branching ratios from the PDG [42]. In thecase where N AA (2 S + 3 S ) = 0, R AA (1S+2S+3S) ≈ R AA (1 S ) × . 7. This is consistent with our observed R AA values, and can also be inferred by examining the massrange 10–11 GeV /c in Fig. 4, where no significant 2S or3S signals are seen.By applying the methods described in [44], we can cal- part N AA , d A ( S + S + S ) R ¡ (1S+2S+3S) ¡ STAR = 200 GeV NN s |<1.0 ¡ |y STAR Au+AuSTAR d+Aup+p Stat.+Fit UncertaintyCommon Normalization Syst. STAR Au+Au, Centrality IntegratedStrickland-Bazow ModelEmerick-Zhao-Rapp Model (a) part N AA , d A ( S + S + S ) R ¡ (1S+2S+3S) ¡ STAR = 200 GeV NN s |<0.5 ¡ |y STAR Au+AuSTAR d+Aup+p Stat.+Fit UncertaintyCommon Normalization Syst. STAR Au+Au, Centrality IntegratedStrickland-Bazow ModelEmerick-Zhao-Rapp Model (b) part N AA , d A ( S ) R ¡ STAR Au+AuSTAR d+Aup+p Stat.+Fit UncertaintyCommon Normalization Syst. STAR Au+Au, Centrality IntegratedStrickland-Bazow ModelLiu-Chen Model (1S) ¡ STAR = 200 GeV NN s |<1.0 ¡ |y (c) FIG. 5: (Color online) Nuclear modification factor forΥ(1S+2S+3S), in | y | < . | y | < . | y | < . d +Au (green square) and Au+Au (blackcircles) collisions as a function of N part . The boxes aroundunity show the statistical (shaded) and systematic (filled) un-certainty from the p + p measurement. The gray bands aroundthe data points are the systematic uncertainties. The data arecompared to calculations from Refs. [27–29]. Binding Energy (GeV)0 0.2 0.4 0.6 0.8 1 1.2 1.4 AA R (1S) ¡ ¡ T , p y J/0-10% CentralitySTAR Inclusive Quarkonium Measurements = 200 GeV, |y|<1 NN sAu+Au, FIG. 6: (Color online) Nuclear modification factor of quarko-nium states as a function of binding energy as measured bySTAR. The horizontal line of the Υ(2S+3S) upper limit spansthe range from the 3S to 2S binding energy; the arrow isplaced at the weighted average of the binding energies. Thehigh- p T J/ ψ results are from Ref. [45]. culate an upper limit on the R AA of the combined 2Sand 3S states. Using the fit to Drell-Yan and b ¯ b (dashed,green curve) as the background, we find an upper limit of29 signal counts with 95% confidence in the mass range10–11 GeV /c for 0–60% centrality collisions. To trans-form this upper limit into an upper limit on R AA (2S+3S),we assumed that the purity of excited states in this massrange is the same as in the p + p case. While the excitedstates are likely more suppressed than the ground statein the Au+Au case, using the p + p purity gives us anupper limit in the Au+Au purity which can be used tocalculate an upper limit on the R AA . The 2S+3S crosssection in p + p was extracted from the full cross section,assuming the purity can be obtained based on the PDGbranching ratios [42] and the relative production crosssections of the three states. In the centrality range of0–60%, we thus obtain a 95%-confidence upper limit of R AA (2S+3S) < . 32 (see Fig. 6).Our data are also compared to model calculations in-corporating hot-nuclear-matter effects for Au+Au [27–29]. These aim to incorporate lattice-QCD results perti-nent to screening and broadening of bottomonium and tomodel the dynamical propagation of the Υ meson in thecolored medium. Both models are in agreement with thelevel of suppression seen in Au+Au. The model proposedby Emerick, Zhao, and Rapp (EZR), Ref. [28], includespossible CNM effects, modeled as an absorption cross sec-tion of up to 3 mb which can account for a value of R AA as low as 0.7. In this model the additional suppressionto bring R AA down to ≈ . R AA , using the internal energy asthe heavy-quark potential and an initial temperature of1the fireball of T = 340 MeV, which given the input fromlattice QCD results, is not hot enough to melt the di-rectly produced Υ(1S). Hence, the suppression is mostlydriven by the dissociation of the excited states (both theS-states and the P-states). The initial temperature usedin the EZR model is 330 MeV (with a formation timeof 0.6 fm/c). The temperatures of the QGP needed inStrickland’s model, Ref. [27], are in the range 428 – 442MeV. However, it should be noted that neither the Strick-land model, nor the calculation from Liu et al. includeany CNM effects.Considering two possible sources of suppression, CNMand QGP effects, we used a Monte Carlo pseudoexperi-ment to compare our results to different possible sourcesof suppression. We investigated four possible scenarios:(1) No suppression compared to p + p ; (2) Suppressiondue to CNM effects only; (3) QGP suppression only; (4)Suppression from both CNM and QGP effects. We simu-lated Υ production in p + p , d +Au, and Au+Au collisionsvia a Poisson generator. CNM effects were included viathe suppression parametrization used by E772 [21] andpresented in Fig. 3(a). We used the predictions fromthe Strickland model [27] to estimate suppression fromQGP effects. For scenario (4), the expected suppres-sion is simply taken to be the product of the suppressionfrom scenario (2) and scenario (3). For this pseudoexper-iment we assumed a flat prior based on the allowed R AA given in Strickland-Bazow [27], depicted as the band forthis calculation in Fig. 5, stemming from the choice of1 < πη/S < R d Au in the range | y | < . R AA forthe most-central Au+Au bin in the range | y | < . 0. Bycomparing the results of the pseudoexperiments with ourmeasurements, we are able to exclude scenario (1) at a ∼ σ confidence level. Finally, we see that hypothesis (4)(dot-dashed curve), including both hot and cold nucleareffects, is consistent with our measurements when boththe d +Au and Au+Au results are taken into account.We repeated this procedure for the rapidity range | y | < . 5. The results are shown in Figs. 7 (c) and (d). Inthe mid-rapidity range we find a larger amount of sup-pression in d +Au than what we observe in the range | y | < . 0. Furthermore, R d Au is comparable to R AA in0-10% for this rapidity range. This could indicate thatsuppression of bottomonium already occurs in d +Au col-lisions. However, given the uncertainties in our currentresults, no particular model of Υ suppression in d +Auis favored. Hence, further investigation of cold-nuclear-matter effects on Υ production is highly warranted. Thesuppression effects seen in d +Au, which are not explained P r ob a b ili t y D e n s i t y (1S+2S+3S) ¡ STAR dAu No Suppression-scaling a A (a) |<1.0 ¡ |y AA R P r ob a b ili t y D e n s i t y STAR AuAu 0-10% No Suppression-scaling a A QGP Effects only and QGP Effects a A (b) P r ob a b ili t y D e n s i t y (1S+2S+3S) ¡ STAR dAu No Suppression-scaling a A (c) |<0.5 ¡ |y AA R P r ob a b ili t y D e n s i t y STAR AuAu 0-10% No Suppression-scaling a A QGP Effects only and QGP Effects a A (d) FIG. 7: (Color online) Summary of the results of four differ-ent pseudoexperiments: No Suppression (solid gold), CMNeffects only (dashed red), QGP effects only (dotted green),and both CMN and QGP effects (dot-dashed blue). Weshow our results for two systems and two rapidity ranges:(a) d +Au | y | < . 0, (b) Au+Au | y | < . 0, (c) d +Au | y | < . | y | < . 5. Our data is shown as a red vertical linewith systematics shown by the pink box. The QGP effectsare modeled in [27]. by the models discussed here, still need to be understoodbefore the Au+Au results can be fully interpreted.2 CONCLUSIONS In conclusion we studied Υ(1S+2S+3S) production in p + p , d +Au, and Au+Au collisions at √ s =200 GeV.We measured the cross section in p + p collisions to be B ee × dσ/dy | | y | < = 61 ± . + fit) +13 − (syst . ) pb andfind it to be consistent within errors with NLO calcu-lations. The cross section in d +Au collisions is foundto be B ee × dσ/dy | | y | < = 19 ± . + fit) ± . )nb. We obtain a nuclear modification factor in this ra-pidity region ( | y | < 1) of R d Au (1 S + 2 S + 3 S ) = 0 . ± . 14 ( d +Au stat . ) ± . 10 ( p + p stat . ) ± . 03 ( d +Au syst . ) ± . p + p syst . ). Models of Υ production in cold nuclearmatter, which include shadowing and initial-state par-tonic energy loss, are consistent with the cross-sectionswe observe in our d +Au data. Higher statistics d +Audata are required to further investigate the 3 σ devi-ation we observe at | y | < . 5. We measured theΥ(1S+2S+3S) nuclear modification factor in Au+Au col-lisions at √ s NN =200 GeV as a function of centrality. Inthe range | y | < R AA (1 S + 2 S + 3 S ) = 0 . ± . 13 (Au+Au stat . ) ± . 07 ( p + p stat . ) ± . 02 (Au+Au syst . ) ± . 06 ( p + p syst . ),indicating additional Υ suppression in hot nuclear mattercompared to cold nuclear matter. In 0-60% centrality wefind a 95%-confidence upper limit on the nuclear modifi-cation of the excited states of R AA (2S+3S) < . 32. Cal-culations of the centrality dependence of Υ R AA usingmodels based on lattice QCD calculations of bottomo-nium melting in a hot medium are found to be consistentwith our data. Therefore, the suppression seen in centralAu+Au collisions is indicative of the presence of decon-fined matter in heavy-ion collisions. It would be desir-able to have a higher statistics d +Au dataset in order tostrengthen the conclusions regarding cold-nuclear mod-ifications to Υ production before a stronger connectionbetween parton deconfinement, Debye screening, and theobserved Υ suppression in Au+Au can be made. ACKNOWLEDGEMENTS We thank R. Vogt, M. Strickland, R. Rapp, F. Arleo,and J.P. Lansberg for providing us calculations in theSTAR kinematic regions. We thank the RHIC Opera-tions Group and RCF at BNL, the NERSC Center atLBNL, the KISTI Center in Korea, and the Open Sci-ence Grid consortium for providing resources and sup-port. This work was supported in part by the Offices ofNP and HEP within the U.S. DOE Office of Science, theU.S. NSF, CNRS/IN2P3, FAPESP-CNPq of Brazil, theMinistry of Education and Science of the Russian Feder-ation, NNSFC, CAS, MoST and MoE of China, the Ko-rean Research Foundation, GA and MSMT of the CzechRepublic, FIAS of Germany, DAE, DST, and CSIR ofIndia, the National Science Centre of Poland, National Research Foundation (NRF-2012004024), the Ministry ofScience, Education and Sports of the Republic of Croa-tia, and RosAtom of Russia. [1] T. Matsui and H. Satz, Phys. Lett. B , 416 (1986).[2] N. Brambilla, S. Eidelman, B. K. Heltsley, R. Vogt,G. T. Bodwin, E. Eichten, A. D. Frawley and A. B. Meyer et al. , Eur. Phys. J. C , 1534 (2011) [arXiv:1010.5827[hep-ph]].[3] F. Karsch and E. Laermann, In: R.C. Hwa et al.(Eds.): Quark Gluon Plasma 3, World Scientific, pp. 1-59[arXiv:hep-lat/0305025].[4] M. Cheng, S. Ejiri, P. Hegde, F. Karsch, O. Kaczmarek,E. Laermann, R. D. Mawhinney and C. Miao et al. , Phys.Rev. D , 054504 (2010) [arXiv:0911.2215 [hep-lat]].[5] S. Borsanyi et al. [Wuppertal-Budapest Collaboration],JHEP , 073 (2010) [arXiv:1005.3508 [hep-lat]].[6] S. Digal, P. Petreczky and H. Satz, Phys. Rev. D ,094015 (2001) [hep-ph/0106017].[7] M. Laine, O. Philipsen, P. Romatschke and M. Tassler,JHEP , 054 (2007) [hep-ph/0611300].[8] P. Petreczky, C. Miao and A. Mocsy, Nucl. Phys. A ,125 (2011) [arXiv:1012.4433 [hep-ph]].[9] A. Dumitru, Y. Guo and M. Strickland, Phys. Rev. D , 114003 (2009) [arXiv:0903.4703 [hep-ph]].[10] A. Rothkopf, T. Hatsuda and S. Sasaki, Phys. Rev. Lett. , 162001 (2012) [arXiv:1108.1579 [hep-lat]].[11] A. Adare et al. [PHENIX Collaboration], Phys. Rev.Lett. , 232002 (2007) [hep-ex/0611020].[12] A. Adare et al. [PHENIX Collaboration], Phys. Rev. D , 012001 (2010) [arXiv:0912.2082 [hep-ex]].[13] B. I. Abelev et al. [STAR Collaboration], Phys. Rev. D , 012004 (2010) [arXiv:1001.2745 [nucl-ex]].[14] H. Agakishiev et al. [STAR Collaboration], Phys. Rev. D , 052006 (2011) [arXiv:1102.2611 [nucl-ex]].[15] M. Cacciari, P. Nason and R. Vogt, Phys. Rev. Lett. ,122001 (2005) [hep-ph/0502203].[16] J. Adams et al. [STAR Collaboration], Phys. Rev. Lett. , 062301 (2005) [nucl-ex/0407006].[17] A. Adare et al. [PHENIX Collaboration], Phys. Rev. C , 044905 (2011) [arXiv:1005.1627 [nucl-ex]].[18] C. Gerschel and J. Hufner, Phys. Lett. B , 253 (1988).[19] C. Gerschel and J. Hufner, Ann. Rev. Nucl. Part. Sci. , 255 (1999) [hep-ph/9802245].[20] Z. -w. Lin and C. M. Ko, Phys. Lett. B , 104 (2001)[nucl-th/0007027].[21] D. M. Alde, H. W. Baer, T. A. Carey, G. T. Garvey,A. Klein, C. Lee, M. J. Leitch and J. W. Lillberg et al. ,Phys. Rev. Lett. , 2285 (1991).[22] B. Alessandro et al. [NA50 Collaboration], Eur. Phys. J.C , 329 (2006) [nucl-ex/0612012].[23] A. D. Frawley, T. Ullrich and R. Vogt, Phys. Rept. ,125 (2008). [arXiv:0806.1013 [nucl-ex]].[24] Calculations for charmonium with shadowing done inR. Vogt, R. E. Nelson and A. D. Frawley, PoS Confine-mentX, (2012). Bottomonium calculations based ona similar procedure (private communication).[25] E. G. Ferreiro, F. Fleuret, J. P. Lansberg, N. Matagneand A. Rakotozafindrabe, Eur. Phys. J. C , 2427[arXiv:1110.5047 [hep-ph]]. [26] F. Arleo and S. Peign´e, JHEP , 122 (2013)[arXiv:1212.0434 [hep-ph]], and private communication.[27] M. Strickland and D. Bazow, Nucl. Phys. A , 25(2012). [arXiv:1112.2761 [nucl-th]].[28] A. Emerick, X. Zhao and R. Rapp, Eur. Phys. J. A ,72 (2012) [arXiv:1111.6537 [hep-ph]].[29] Y. Liu, B. Chen, N. Xu and P. Zhuang, Phys. Lett. B , 32 (2011) [arXiv:1009.2585 [nucl-th]].[30] M. Anderson et al. , Nucl. Instr. Meth. A , 659 (2003).[31] M. Beddo et al. , Nucl. Instr. Meth. A , 725 (2003).[32] R. Vogt, RHIC-II Science Work-shop April 28-30, 2005. http://rhicii-science.bnl.gov/heavy/doc/April05Meeting/[33] T. Sjostrand, arXiv:0809.0303 [hep-ph].[34] J. Gaiser, PhD thesis, SLAC-R-255, Appendix F., p. 178.[35] A. Adare et al. [PHENIX Collaboration], Phys. Rev.Lett. , 242301 (2012) [arXiv:1211.4017 [nucl-ex]].[36] K. J. Eskola, H. Paukkunen and C. A. Salgado, JHEP , 065 (2009) [arXiv:0902.4154 [hep-ph]].[37] S. Chatrchyan et al. [CMS Collaboration], Phys. Rev. Lett. , 052302 (2011).[38] S. Chatrchyan et al. [CMS Collaboration], JHEP ,063 (2012) [arXiv:1201.5069 [nucl-ex]].[39] S. Chatrchyan et al. [CMS Collaboration], Phys. Rev.Lett. , 222301 (2012) [arXiv:1208.2826 [nucl-ex]].[40] V. Khachatryan et al. [CMS Collaboration], Phys. Rev.D , 112004 (2011) [arXiv:1012.5545 [hep-ex]].[41] G. Moreno, C. N. Brown, W. E. Cooper, D. Fin-ley, Y. B. Hsiung, A. M. Jonckheere, H. Jostlein andD. M. Kaplan et al. , Phys. Rev. D (1991) 2815.[42] J. Beringer et al. (Particle Data Group), Phys. Rev. D ,010001 (2012)[43] T. Affolder et al. [CDF Collaboration], Phys. Rev. Lett. , 2094 (2000) [hep-ex/9910025].[44] W. A. Rolke, A. M. L´opez, & J. Conrad, Nucl. Instr.Meth. A, , 493 (2005)[45] L. Adamczyk et al. [STAR Collaboration], Phys. Lett. B722