A New Heavy Flavor Program for the Future Electron-Ion Collider
Xuan Li, Ivan Vitev, Melynda Brooks, Lukasz Cincio, J. Matthew Durham, Michael Graesser, Ming X. Liu, Astrid Morreale, Duff Neill, Cesar da Silva, Walter E. Sondheim, Boram Yoon
AA New Heavy Flavor Program for the Future Electron-Ion Collider
Xuan Li ,(cid:63) , Ivan
Vitev , Melynda
Brooks , Lukasz
Cincio , J. Matthew
Durham , Michael
Graesser , Ming X.
Liu , Astrid
Morreale , Duff
Neill , Cesar da Silva , Walter E.
Sondheim , Boram
Yoon . Los Alamos National Laboratory
Abstract.
The proposed high-energy and high-luminosity Electron-Ion Collider (EIC)will provide one of the cleanest environments to precisely determine the nuclear partondistribution functions (nPDFs) in a wide x - Q range. Heavy flavor production at the EICprovides access to nPDFs in the poorly constrained high Bjorken- x region, allows us tostudy the quark and gluon fragmentation processes, and constrains parton energy loss incold nuclear matter. Scientists at the Los Alamos National Laboratory are developinga new physics program to study heavy flavor production, flavor tagged jets, and heavyflavor hadron-jet correlations in the nucleon / nucleus going direction at the future EIC.The proposed measurements will provide a unique way to explore the flavor dependentfragmentation functions and energy loss in a heavy nucleus. They will constrain theinitial-state e ff ects that are critical for the interpretation of previous and ongoing heavyion measurements at the Relativistic Heavy Ion Collider and the Large Hadron Collider.We show an initial conceptual design of the proposed Forward Silicon Tracking (FST)detector at the EIC, which is essential to carry out the heavy flavor measurements. Wefurther present initial feasibility studies / simulations of heavy flavor hadron reconstructionusing the proposed FST. Heavy flavor production is an ideal probe to study the full evolution of the nuclear medium createdin heavy ion collisions at the Relativistic Heavy Ion Collider (RHIC) and the Large Hadron Col-lider (LHC). Heavy quarks are produced early on in the scattering process due to their high masses( m c / b (cid:29) Λ QCD ). The precise mechanisms by which heavy flavor interacts with the nuclear medium,however, are still not well understood in heavy ion collisions. The production process includes bothcold nuclear matter e ff ects such as modification of the nuclear Parton Distribution Functions (nPDF),and hot nuclear matter e ff ects such as parton energy loss within the Quark Gluon Plasma (QGP).The future Electron Ion Collider (EIC), recommended as top priority for new construction by the2015 Long Range Plan for nuclear science and strongly endorsed by the 2018 National Academyof Sciences committee, will utilize high-energy high-luminosity electron-nucleon (electron-nucleus)collisions to study several fundamental questions in the nuclear physics field. The deep inelasticscattering (DIS) processes of electrons on light / heavy nuclei at the future EIC will provide a clean (cid:63) e-mail: [email protected] a r X i v : . [ nu c l - e x ] F e b nvironment to directly explore the initial-state cold nuclear matter e ff ects and the final-state frag-mentation / hadronization in the nuclear environment. These studies will improve our understanding ofthe current and future RHIC and LHC measurements and shed light on the non-perturbative aspectsof quantum chromodynamics (QCD). The future EIC will provide electron + nucleus collisions with multiple nuclear species ranging from He to Pb at a series of center-of-mass energies from 20 GeV to 141 GeV [2]. Unlike heavyion experiments at RHIC and LHC, these asymmetric collisions will provide a clean environmentto precisely study the initial-state nPDFs in the large Bjorken-x region [4, 5] and the fragmen-tation / hadronization processes for heavy flavor production. Heavy flavor measurements in polar-ized electron + nucleon / nuclei collisions at the future EIC will provide opportunities to study the nu-cleon / nucleus spin structure such as the gluon Sivers e ff ect. Quarks and gluons lose energy when traversing a nuclear medium and their energy loss is mass / flavordependent ( ∆ E g > ∆ E u , d , s > ∆ E c > ∆ E b ) [1]. The quark / gluon fragmentation and hadronizationprocesses can be modified in the nuclear medium as shown in figure 1. The cross sections of final-state hadrons and jets to be measured in electron + proton and electron + nuclei collisions at the futureEIC are proportional to the initial quark / gluon PDFs, the quark / gluon hard scattering part which canbe calculated in perturbative QCD, and the quark / gluon fragmentation and hadronization processes.Comparison of di ff erent final-state cross sections can help separate the information about the initial-state nuclear PDFs and the final-state fragmentation and hadronization processes. Comparison ofcross sections measured in electron + proton and electron + nucleus collisions for the same final hadronstates can precisely determine the nuclear medium e ff ects such as the parton energy loss contributions. Figure 1.
Heavy flavor hadron and jet production in electron + proton (left) and electron + nuclei (right) colli-sions. The heavy flavor fragmentation and hadronization process in electron + nuclei collisions is modified whencompared to the process in electron + proton collisions. There are competing theoretical models of nuclear modification in DIS reactions within nuclei.One is the absorption model, which describes hadronization that happens inside nuclear matter. Theother is quark / gluon energy loss, which is associated with hadronization outside the nuclear medium.Di ff erentiation between these two models is challenging with only light flavor hadron measurements.Figure 2 shows the predicted nuclear modification factor R eA for heavy flavor hadron measurements Hadron momentum fraction z h C r o ss s e c t i on m od i f i c a t i on R e A Pions, energy loss D mesons, energy lossB mesons, energy lossPions, absorptionD meson, absorptionB meson, absorption BD π Integrated luminosity in e+p = 10 fb -1 Integrated luminosity in e+A = 500 nb -1 FST at EIC
Figure 2.
Projected statistical uncertainties of the nuclear modification factor R eA for pions, D -mesons and B -mesons in electron + gold collisions at √ s =
69 GeV. The projected R eA mean value is based on assumptions.Theory calculations include parton energy loss (solid lines) and absorption within nuclei (dashed lines) [3]. Theintegrated luminosity in electron + proton (electron + gold) collisions is 10 f b − (500 nb − ). versus the hadron momentum fraction relative to their parent parton in e + Au collisions at √ s = ffi ciencies in the forward pseudorapidity region of 1 . < η < .
5, whichis the proposed Forward Silicon Tracking coverage. The integrated luminosity in electron + proton(electron + gold) collisions is 10 f b − (500 nb − ). Clear separation between these two models can onlybe achieved through heavy flavor hadron measurements, as they have very di ff erent fragmentationfunctions and formation time when compared to light flavor hadrons [3]. The future EIC with theproposed FST detector can provide su ffi cient statistical precision for these heavy flavor measurementswith enhanced sensitivity to nuclear transport properties. The initial design of the proposed Forward Silicon Tracking (FST) detector in fast simulation is shownin figure 3. The proposed FST consists of 2 barrel layers and 5 forward planes to cover the pseudora-pidity range of 1 . < η < . / nucleus going direction at the future EIC. This detectorcan precisely determine the transverse decay length of particles (see the schematics in figure 3), whichis critical for heavy flavor decay product reconstruction and tagging. Single track performance basedon a 3 layer mid-rapidity Monolithic Active Pixel Sensor (MAPS) [8] central vertex detector and theproposed FST with a hybrid design of MAPS plus other silicon technology is shown in figure 4. Auniform 3T magnetic filed is applied in the fast simulation and the magnetic field provided by the EICinteraction point design will be included in the full simulation later. With the help of the FST, betterthan 70 µ m spatial resolution can be obtained for forward tracks. Additionally, charged tracks withlow transverse momentum can get better than 2% momentum resolutions. igure 3. Initial design of the proposed Forward Silicon Tracker, which covers the pseudorapidity of 1 . < η < . / nuclei going direction at the future EIC, is highlighted inside the green box on the left. Thisdetector can precisely determine the transverse decay length of particles from the primary vertex ( b T ) as shownon the right. Figure 4.
Single track performance for the proposed Forward Silicon Tracker in di ff erent pseudorapidity regionsfrom η = . η = .
0. A uniform 3T magnetic filed is applied for the track performance evaluation. Theresolution of the transverse decay length b T versus the track p T is shown on the left and the track p T dependentmomentum resolution ∆ p T / p T is shown on the right. To demonstrate the proposed FST capability, the track performance discussed in Sect. 2.2 is embeddedinto the PYTHIA8 [6] simulation for hadron reconstruction studies. One e + p collision is embeddedwith on average of 0.02 p + p collisions, which account for the beam interaction background. As theEIC background evaluation is underway, we will include other backgrounds such as the synchrotronradiation in future studies. Charm mesons including D , D ± and D ± s can decay into K ± , π ± or lighter D -mesons with a relatively large branching ratio. The reconstruction algorithm in simulation is devel-oped to cluster charged tracks by matching their transverse decay length ( b T ) ( ∆ b T < µ m ) measuredby the FST (see the geometry in figure 3). To enhance the probability of finding charm mesons, thecharged track clusters are required to contain at least one K ± track with 100% PID.Figure 5 shows the invariant mass distributions of charged track clusters which includes at leastone K ± track in 18 GeV electron and 100 GeV proton collisions in di ff erent mass regions. Twodi ff erent FST pixel pitch values are used to compare the reconstructed D -meson mass distributions.With 10 f b − integrated luminosity, clear signals of D ( ¯ D ), D ± and D ± s are found with the help of .83 1.84 1.85 1.86 1.87 1.88 1.89 1.9 1.91 GeV/c ± Reconstructed cluster mass with K w/ LANL FSTMCTotal fit ) D ( Rec. D ± Rec. D
SIG/BKG: 1.510 ± D ) SIG/BKG: 1.360 D ( D ± Reconstructed cluster mass with K GeV/c ± Reconstructed cluster mass with K ) SIG/BKG: 0.670 ± s D ( ± s D w/ LANL FSTMC s ± Rec. D ± Reconstructed cluster mass with K GeV/c ± Reconstructed cluster mass with K w/ LANL FSTMCTotal fit ) D ( Rec. D ± Rec. D
SIG/BKG: 1.028 ± D ) SIG/BKG: 0.984 D ( D ± Reconstructed cluster mass with K GeV/c ± Reconstructed cluster mass with K ) SIG/BKG: 0.262 ± s D ( ± s D w/ LANL FSTMC s ± Rec. D ± Reconstructed cluster mass with K
Figure 5.
Invariant mass of charged track clusters with at least one K ± track in 18 GeV electron on 100 GeVproton collisions. No charge separation is applied. The integrated luminosity is 10 f b − . Clear D ( ¯ D ), D ± and D ± s signals are found with the help of the proposed FST. The performance in the top panel is based on theproposed FST has 30 µ m pixel pitch, 500 kHZ trigger rate and 0 . X radiation length per detector layer. Thedistributions in the bottom panel are based on the proposed FST has 200 µ m pixel pitch, 500 kHZ trigger rate and0 . X radiation length per detector layer. the proposed FST, which has 30 µ m pixel pitch, 500 kHZ trigger rate and 0 . X radiation lengthper detector layer. The signal to background ratio for D -meson reconstruction is determined froma fit which includes the signal and combinatorial background contributions. When the pixel pitchincreases to 200 µ m , the mass resolution for the reconstructed D -mesons gets broader and the signalto background ratio is reduced.To study the physics results, such as the D -meson reconstruction, dependence on the detectorperformance, we scan the 3D parameter space which includes the pixel pitch, the trigger integrationtime, and the average material budgets per layer. Table 1 summarizes the parameter values in thesimulation studies. Table 1.
The proposed FST parameter table
Parameter name valuesPixel pitch 30 µ m µ m µ m µ m Material budgets per layer 0.3% X X X X Trigger integration time 2 µ s µ s µ s By varying the pixel pitch, the material budget per layer and the trigger integration time, theextracted signal over background ratios for reconstructed D ( ¯ D ), D ± and D ± s in 18 GeV electron on100 GeV proton collisions are shown in figure 6. The reconstruction statistical uncertainties, whichre determined with 10 f b − integrated luminosity and include 95% tracking e ffi ciency, are better than0.1%. Signal/Background VS pixel pitch
Pixel pitch (micron) S i g / B k g Sig/Bkg ratio ± D ) D ( D s ± D Material per layer: 0.4%XTrigger rate: 500 kHZ
Signal/Background VS pixel pitch Signal/Background VS Material budget per layer ) Material budget per layer (% X S i g / B k g Sig/Bkg ratio ± D ) D ( D s ± D Pixel pitch: 30 micronTrigger rate: 500 kHZ
Signal/Background VS Material budget per layer Signal/Background VS trigger integration time trigger integration time (micro-sec) S i g / B k g Sig/Bkg ratio ± D ) D ( D s ± D Pixel pitch: 30 micron Material per layer: 0.4%X
Signal/Background VS trigger integration time
Figure 6.
Signal to background ratio for reconstructed D ( ¯ D ), D ± and D ± s versus pixel pitch (left), materialsbudgets per detector layer (middle) and the trigger integration time (right) for di ff erent detector performancecombinations. These reconstruction values are achieved with 10 f b − integrated luminosity. The D -meson reconstruction signal to background ratio decreases as pixel pitch, material budgetsper detector layer and trigger integration time increases. It has a strong dependence on the materialbudgets per detector layer with fixed pixel pitch at 30 µ m and fixed trigger rate at 500 kHZ. Furtherstudies include a wider detector parameter phase space scan, higher mass hadron (e.g. B -mesons)and jet reconstruction, more realistic interaction point design and backgrounds in full simulation areunderway. The future EIC will provide unique opportunities to utilize heavy flavor and jet probes to preciselydetermine the nuclear PDFs, explore the gluon Sivers function and study parton energy loss mech-anism in nuclear medium within a wide Bjorken-x and Q kinematic region. A new heavy flavorand jet program has started at Los Alamos National Laboratory (LANL) to develop new observablesand initialize a conceptual detector design for a proposed Forward Silicon Tracking detector for thefuture EIC. Heavy flavor production in e + A collisions at the future EIC will provide strong discrim-inating power between competing theoretical models for the parton transport coe ffi cients in nuclearmedium. A fast simulation version of the FST design is completed and its tracking performance isconsistent with or better than the EIC detector R&D handbook requirements [7]. Initial studies in thefast simulation demonstrate the capability for D -meson reconstruction by the proposed FST. A silicondetector R&D lab is being set up to verify the performance of the proposed silicon techniques, suchas MAPS. Ongoing theoretical and experimental studies at LANL, which include new heavy flavorand jet observables determination, physics projections in full simulation, detector design optimiza-tion and system integration, will bring about key advances in theory and experimentation and provideimportant contributions to the EIC physics and detector developments. References [1] Azfar Adil, Ivan Vitev, Phys. Lett. B : (2007)139-146.[2] A. Accardi et al., Eur. Phy. J. A., (2016), 9 .[3] Ivan Vitev, Radiative process and jet modification at the EIC (2019), arXiv: 1912.10965.[4] E. Chudakov et al., JLAB-THY-16-2354, arXiv: 1610.08536.5] E. C. Aschenauer et al., Phys. Rev. D (2017), 114005.[6] T. Sjostrand, S. Mrenna and P. Skands, Comput. Phys. Comm. (2008), 852,arXiv:0710.3820.[7] E. C. Aschenauer et al., "Electron-Ion Collider Detector Requirements and R&D Handbook".[8] M. Mager, the ALICE collaboration, Nuclear Instruments and Methods in Physics ResearchSection A: Accelerators, Spectrometers, Detectors and Associated Equipment,824