Analysis Note: Study of Exclusive $\pi^{0}$ Production Measurement at ePIC of the Future Electron-Ion Collider

epic-AN-AC-2025-002 Study of Exclusive π0Production Measurement at ePIC of the Future Electron-Ion Collider Principal Author List: Jihee Kim (Brookhaven National Laboratory, [email protected]v) (December 16, 2025) Abstract The Electron-Ion Collider (EIC) is a next-generation experimental facility designed to investigate the fundamental structure of matter through Deep Inelastic Scattering. Its primary goal is to explore the properties of quarks and gluons in nucleons and nuclei, thereby advancing our understanding of the building blocks of visible matter in the universe. The EIC community outlined the physics program of the EIC in White Paper, and the demanding detector requirements and potential technologies to deploy at an EIC detector were published in a comprehensive Yellow Report. The generalpurpose detector resulting from this effort, ePIC, is designed to perform a broad physics program. A key physics channel accessible at the EIC is hard exclusive π0production, which plays a central role in probing Generalized Parton Distributions. This process provides critical insights into the three-dimensional structure of nucleons and nuclei and offers a unique opportunity to study quark orbital angular momentum—a key component in resolving the origin of nucleon mass and spin, a long-standing question in nuclear physics. Additionally, exclusive π0production can mimic the final state of Deeply Virtual Compton Scattering (DVCS), making it a potential background to DVCS measurements. In this analysis note, I present a simulation study of the exclusive process e+p→ e′+p′+π0using the ePIC detector. I evaluate the differential cross-section with respect to the momentum transfer tbetween the initial and final-state proton, assess the sensitivity of asymmetry measurements in this channel, and estimate the background contribution to DVCS measurement. Contents 1 Introduction 6 2 Analysis details 7 2.1 Data and Monte Carlo Samples . . . . . . . . . . . . . . . . . . . . . . . . . 7 2.2 Event, Particle and PID Selections . . . . . . . . . . . . . . . . . . . . . . . 9 2.2.1 ScatteredElectron ............................ 9 2.2.2 ScatteredProton ............................. 11 2.2.3 NeutralPion ............................... 13 2.3 Systematic Uncertainties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 2.4 Results....................................... 15 2.5 Discussions - DIS Background . . . . . . . . . . . . . . . . . . . . . . . . . . 19 2.6 Discussions - Physics Background to DVCS using 5×41 GeV2........ 22 2.7 Discussions - Energy threshold in the forward calorimeter . . . . . . . . . . . 24 3 Summary 27 iii List of Figures 1 Hard exclusive π0productiondiagram ..................... 6 2 MC-level final-state particle distribution . . . . . . . . . . . . . . . . . . . . 8 3 Scattered electron pseudorapidity and energy distributions . . . . . . . . . . 10 4 DISkinematicvariables ............................. 11 5 Scattered proton acceptance and t-distribution . . . . . . . . . . . . . . . . . 12 6 Scattered proton momentum and t-distribution . . . . . . . . . . . . . . . . 12 7 Neutral pion pseudorapidity and momentum distributions . . . . . . . . . . . 13 8 Neutral pion pseudorapidity and momentum distributions with potential improvement..................................... 14 9 Invariant mass distribution of neutral pion . . . . . . . . . . . . . . . . . . . 14 10 t-distribution with tBABE method ........................ 16 11 t-distribution with teXBE method ........................ 16 12 Combinedt-distribution ............................. 17 13 Combinedt-distribution ............................. 18 14 Combined t-distribution with realistic fluctuations . . . . . . . . . . . . . . . 19 15 DIS background distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 16 Kinematics of neutral pions and two decay photons . . . . . . . . . . . . . . 22 17 2D pseudorapidity distribution of two decay photons . . . . . . . . . . . . . 23 18 Example of hit map of merged event in forward calorimeter . . . . . . . . . . 23 iv List of Tables 1 DIS Background Exclusivity Cut Selection . . . . . . . . . . . . . . . . . . . 20 2 DIS Background Full Exclusivity Cut Selection . . . . . . . . . . . . . . . . 21 v 1 Introduction A central objective of the EIC physics program is to explore the three-dimensional structure of nucleons and nuclei. This structure can be accessed through Generalized Parton Distributions (GPDs), which are experimentally probed via exclusive processes such as Deeply Virtual Compton Scattering (DVCS) and Deeply Virtual Meson Production (DVMP). DVCS offers the cleanest access to GPDs and is particularly sensitive to chiral-even GPDs through its cross-section. In parallel, hard exclusive π0production (DVπ0P), illustrated in Figure 1, plays a complementary role by providing sensitivity to chiral-odd GPDs, which are related to quark transversity distributions. Moreover, DVπ0P provides a unique opportunity to investigate quark orbital angular momentum (OAM). A recent theoretical study [1] indicates that single-target spin asymmetries in this channel are sensitive to quark OAM. To date, no direct experimental measurement of quark OAM has been performed. The EIC, with its capability for high proton polarization, provides an ideal environment to explore this process. These measurements are expected to contribute significantly to resolving the origin of nucleon mass and spin—a fundamental and long-standing question in nuclear physics. While DVπ0P is of strong intrinsic interest, it also constitutes a potential background to DVCS measurements. If one of the decay photons from the π0is not detected—due to limited geometric acceptance, energy thresholds, or detector granularity—the event may be misidentified as a DVCS signal. Therefore, DVπ0P must be thoroughly understood and accurately modeled to control background contributions in DVCS analyses. Previous studies of DVπ0P have been conducted at fixed-target experiments such as JLab and COMPASS, where the accessible kinematic region is dominated by valence quarks. The EIC, in contrast, will provide high proton polarization and broad phase-space coverage, extending into the low-xand high-Q2regions, thereby enabling a comprehensive investigation of this process. Figure 1: A diagram illustrating the hard exclusive π0production. Taken from [1]. 6 2 Analysis details To study the feasibility of the DVπ0P process at the EIC for ep collisions at 10×130 GeV2 and 5×41 GeV2, exclusive π0events (e+p→e′+p′+π0) were generated privately using the EpIC Monte Carlo generator (version 1.1.6). A total of 1 ×106events were processed through the EIC afterburner [2], which incorporates realistic beam effects such as the crossing angle, angular divergence, and momentum spread. These effects are particularly important for accurate modeling of forward particle acceptance and momentum resolution. The simulated events were then passed through the ePIC detector simulation using the craterlake configuration, followed by reconstruction with the ePIC software framework. The resulting data were used for the final analysis. 2.1 Data and Monte Carlo Samples The EpIC [3,4] is a state-of-the-art Monte Carlo event generator developed for the study of exclusive processes. It is built on the PARTONS platform [5,6], which provides a modular and extensible software architecture, allowing for the integration of a wide range of theoretical models describing the partonic structure of the nucleon. EpIC currently supports a range of exclusive reactions, including Deeply Virtual Compton Scattering (DVCS), Timelike Compton Scattering (TCS), Deeply Virtual Meson Production (DVMP) with π0, and Double Deeply Virtual Compton Scattering (DDVCS). It also features the implementation of radiative corrections. EpIC is well-suited for both the analysis of existing experimental data and for impact studies, particularly in the context of future experiments at EIC [7]. DVπ0P events were generated within the following kinematic ranges: •10−5< xB<0.95, •0.01 < y < 0.95, •1 GeV2< Q2<1000 GeV2, •0.01 GeV2<|t|<1.6 GeV2, Figure 2 presents the two-dimensional distributions of momentum versus pseudorapidity for the scattered electron, recoil proton, neutral pion, and the two decay photons produced in DVπ0P events from ep collisions. 7 1 10 2 10 3 10 4 10 3.5−3−2.5−2−1.5−1−0.5−0 MC e' η 0 2 4 6 8 10 12 [GeV/c] MC e' p 1 10 2 10 3 10 4 10 4 5 6 7 8 9 10 11 MC p' η 0 20 40 60 80 100 120 140 [GeV/c] MC p' p 1 10 2 10 3 10 4 10 3−2−1−0 1 2 3 4 MC 0 π η 0 10 20 30 40 50 60 70 80 90 100 [GeV/c] MC 0 π p 1 10 2 10 3 10 4 10 4−2−0 2 4 6 8 MC 2 γ, 1 γ η 0 10 20 30 40 50 60 70 80 90 100 [GeV/c] MC 2 γ, 1 γ p Figure 2: Generator-level momentum versus pseudo-rapidity distributions for the scattered electron, recoil proton, neutral pion, and its two decay photons. Results are obtained using the ePIC October Simulation Campaign (version 25.10.2). EpIC DVπ0P events can be generated from the following directory: /gpfs/mnt/gpfs02/eic/jkim/run-EpIC-Jihee/ To set up the environment, use: source set_env_EpIC EpIC can then be executed on the RHIC Computing Facility (RCF) using an XML input card. For example: run_epic -n 1 -i 10_130_plus_PI0.xml run_epic -n 1 -i 5_41_plus_PI0.xml Here, -n 1 specifies the number of jobs to run, and 10 130 plus PI0.xml and 5 41 plus PI0.xml are the configuration file defining the kinematics and process settings for DVπ0P event generation. 8 All results in this analysis are obtained using datasets from the ePIC October simulation campaign: •10×130 GeV2sample (version 25.10.2) •5×41 GeV2sample (version 25.10.3) Version 25.10.3 was released following the correction of a memory leak, which does not affect any physics-related results. The 10 ×130 GeV2sample was produced using version 25.10.2, which includes improvements to the Roman Pot reconstruction. The 5 ×41 GeV2 sample was generated specifically for background studies in the pre-TDR phase and was therefore released using version 25.10.3. The analysis code is currently under development and is located at: /gpfs02/eic/jkim/DVpi0P/analysis_scripts/ep_10on130_exclusive_pi0_analysis.C 2.2 Event, Particle and PID Selections For the DVπ0P analysis, event exclusivity is required, meaning that all final-state particles must be reconstructed: the scattered electron, the two photons from the π0decay, and the recoil proton. The scattered electron and the decay photons are detected in the central detector, while the recoil proton is measured using the far-forward detection systems—specifically the B0 spectrometer and the Roman Pot (RP) detectors. To assess the detector’s capability to support this physics program, detailed simulations were carried out. The detector acceptance, performance, and reconstruction efficiency were evaluated by studying the reconstructed distributions of key observables associated with exclusive π0production. The following sections describe the acceptance characteristics and the selection criteria applied to each category of final-state particles. 2.2.1 Scattered Electron Accurate identification and reconstruction of the scattered electron are essential, as they directly determine the kinematics of the Deep Inelastic Scattering (DIS) event. Electron selection in this study was performed using truth-level association. If a cluster was found in the backward electromagnetic calorimeter (EMCal), its energy was used to reconstruct the electron’s four-momentum, while the position (pseudorapidity and azimuthal angle) was 9 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 ] 2 Momentum Transfer, -t [GeV 1 10 2 10 3 10 4 10 5 10 6 10 7 10 8 10 ] -2 dN/dt [GeV 0 π+p'+e'→ep: 2 130 GeV× 10epEpIC <0.95 B <x -5 , 10 2 <100 GeV 2 0.01<y<0.85, 1<Q Corrected for acceptance and efficiency effects Figure 10: Number of events as a function of the momentum transfer t, integrated over x and Q2and corrected for detector acceptance and efficiency. The momentum transfer is reconstructed using the scattered proton via the tBABE method. The black solid line represents the true distribution, while the blue open circles indicate the reconstructed distribution after only detector efficiency and acceptance. Results are obtained using the ePIC October Simulation Campaign (version 25.10.2). 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 ] 2 Momentum Transfer, -t [GeV 1 10 2 10 3 10 4 10 5 10 6 10 7 10 8 10 ] -2 dN/dt [GeV 0 π+p'+e'→ep: 2 130 GeV× 10epEpIC <0.95 B <x -5 , 10 2 <100 GeV 2 0.01<y<0.85, 1<Q Corrected for acceptance and efficiency effects Figure 11: Number of events as a function of the momentum transfer t, integrated over xand Q2and corrected for detector acceptance and efficiency. The momentum transfer is reconstructed using the neutral pion and scattered electron via the teXBE method. The black solid line represents the true distribution, while the blue open circles indicate the reconstructed distribution after only detector efficiency and acceptance. Results are obtained using the ePIC October Simulation Campaign (version 25.10.2). Finally, the two t-distributions are combined to produce the complete spectrum, as shown 16 in Figure 12, ensuring coverage across the full kinematic range. 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 ] 2 Momentum Transfer, -t [GeV 1 10 2 10 3 10 4 10 5 10 6 10 7 10 8 10 ] -2 dN/dt [GeV 0 π+p'+e'→ep: 2 130 GeV× 10epEpIC <0.95 B <x -5 , 10 2 <100 GeV 2 0.01<y<0.85, 1<Q Corrected for acceptance and efficiency effects Figure 12: Number of events as a function of the momentum transfer t, integrated over x and Q2and corrected for detector acceptance and efficiency. The momentum transfer is reconstructed using both the tBABE and teXBE methods. The black solid line represents the true distribution, while the blue open circles indicate the reconstructed distribution after only detector efficiency and acceptance. Results are obtained using the ePIC October Simulation Campaign (version 25.10.2). Figure 13 presents a projected measurement for an integrated luminosity of 5 fb−1, representative of the EIC early science. The left panel shows the differential cross section as a function of the momentum transfer t, integrated over xand Q2and corrected for detector acceptance and efficiency. These t-distributions form the basis for spatial imaging of the quark distribution within the nucleon via Fourier transformation. The resulting quark spatial distribution is presented in the right panel of Figure 13. 17 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 ] 2 Momentum Transfer, -t [GeV 1 10 2 10 3 10 4 10 5 10 6 10 ] -2 /dt [fb GeVσd ePIC Performnace 2 130 GeV×P 10 0 πe+p DV 72 GeV≈ s, -1 = 5 fb proj L 10−8−6−4−2−0 2 4 6 8 10 b [fm] 0 0.005 0.01 0.015 0.02 0.025 0.03 0.035 F(b) db ∫ F(b) / Figure 13: (Left) Differential cross section as a function of momentum transfer t, integrated over xand Q2, and corrected for detector acceptance and efficiency, shown for a 5 fb−1 projection. The black solid line represents the true distribution, while the blue open circles indicate the reconstructed distribution after only detector efficiency and acceptance. (Right) Fourier transform of the differential cross-section, representing the quark spatial distribution. The black solid line represents the true distribution, while the blue open circles indicate the reconstructed distribution after only detector efficiency and acceptance. Results are obtained using the ePIC October Simulation Campaign (version 25.10.2). The distribution of the squared momentum transfer, t, shown in the left panel of Fig. 14, can be normalized by the integrated luminosity to obtain the differential cross-section, dσ/dt. It illustrates the same distribution after applying a smearing procedure, based on the mean values and associated uncertainties, to emulate realistic fluctuations in the projected measurement. The resulting quark spatial distribution is presented in the right panel of Figure 14. 18 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 ] 2 Momentum Transfer, -t [GeV 1 10 2 10 3 10 4 10 5 10 6 10 ] -2 /dt [fb GeVσd ePIC Performnace 2 130 GeV×P 10 0 πe+p DV 72 GeV≈ s, -1 = 5 fb proj L 10−8−6−4−2−0246810 b [fm] 0 0.005 0.01 0.015 0.02 0.025 0.03 0.035 F(b) db ∫ F(b) / Figure 14: (Left) Differential cross-section as a function of momentum transfer t, integrated over xand Q2, corrected for detector acceptance and efficiency, and normalized by the integrated luminosity. The momentum transfer is reconstructed using both the tBABE and teXBE methods. The black solid line represents the true distribution, while the blue open circles indicate the reconstructed distribution after only detector efficiency and acceptance. (Right) Fourier transform of the differential cross-section, representing the quark spatial distribution. The black solid line represents the true distribution, while the blue open circles indicate the reconstructed distribution after only detector efficiency and acceptance. Results are obtained using the ePIC October Simulation Campaign (version 25.10.2). 2.5 Discussions - DIS Background To assess the feasibility of exclusive π0production measurements, it is essential to understand the contribution of inclusive DIS processes that can mimic the exclusive topology. The goal is to identify inclusive DIS events that reproduce the same final-state signature as exclusive π0production. A sample of 500,000 inclusive DIS events was used for this study. The main selection strategy requires events to contain only four reconstructed final-state particles and to satisfy the quality cuts defined in the previous section. The applied selection criteria are listed below: 1. Central Detector: Exactly three reconstructed particles 2. Central Detector: One negatively charged particle and two neutrals 3. Far-Forward Detector: One reconstructed positively charged particle, with an angular acceptance requirement 4. Electron candidate: E/p selection 5. Neutral-pion candidate: Mγγ invariant-mass selection 19 The resulting selection efficiencies for two representative Q2ranges are summarized in Table 1. Selection 1 < Q2<10 10 < Q2<100 1 11.3232 % 3.3658 % 2 2.1324 % 0.486 % 3 0.5494 % 0.1132 % 4 0.479 % 0.0838 % 5 0.14 % 0.0114 % Table 1: Remaining fractions of inclusive DIS events after each step of the selection procedure. Neutral pions can be produced through various inclusive DIS processes, and given the large DIS cross section compared to the extremely small exclusive π0cross section, the background can easily dominate. The selections above represent basic exclusivity requirements, but additional cuts are necessary to further suppress the DIS background. Two additional selections were introduced: •Missing mass squared: Calculated from the incoming beams and reconstructed finalstate particles. For true exclusive π0events, the missing mass squared should be close to zero. Deviations from zero indicate the presence of additional undetected particles. Detector resolution and smearing are naturally included, as reconstructed quantities are used. •Far-forward detector veto: Events with hits in the Zero Degree Calorimeter (ZDC) are rejected. The full set of selections is: 1. Central Detector: Exactly three reconstructed particles 2. Central Detector: One negative and two neutral charges 3. Far-Forward Detector: One reconstructed positive-charge particle and angular acceptance cut 4. Electron candidate: E/p cut 5. Neutral-pion candidate: Mγγ invariant mass cut 6. Missing mass squared: |M2 missing|<1 GeV2 7. Far-forward detector veto (ZDC) 20 The selection efficiencies after including these additional cuts are shown in Table 2. Selection 1 < Q2<10 10 < Q2<100 1 11.3232 % 3.3658 % 2 2.1324 % 0.486 % 3 0.5494 % 0.1132 % 4 0.479 % 0.0838 % 5 0.14 % 0.0114 % 6 0.0032 % 0.0004 % 7 0.0016 % 0.0002 % Table 2: Remaining fractions of inclusive DIS events after each step of the selection procedure. The remaining inclusive DIS background events, overlaid on the corrected t-distribution, are shown in Figure 15. All surviving background events contain multiple neutral pions. Even after applying the appropriate cross-section scaling, the residual background level remains significant, amounting to approximately 60% at low Q2and 27.6% at higher Q2. The results further indicate a clear t-dependence of the background contribution. In several bins, only a single background event remains, corresponding to a statistical uncertainty of 100%. These observations underscore the difficulty of fully suppressing inclusive DIS contamination in the analysis. 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 ] 2 Momentum Transfer, -t [GeV 1 10 2 10 3 10 4 10 5 10 6 10 ] -2 /dt [fb GeVσd 0 π+p'+e'→ep: 2 130 GeV× 10epEpIC <0.95 B <x -5 , 10 2 <100 GeV 2 0.01<y<0.85, 1<Q Corrected for acceptance and efficiency effects Figure 15: Number of events as a function of the momentum transfer t, integrated over xand Q2, and corrected for detector acceptance and efficiency. The remaining inclusive DIS background is overlaid as a solid red line. The black solid line represents the true distribution, while the blue open circles indicate the reconstructed distribution after only detector efficiency and acceptance. Results are obtained using the ePIC October Simulation Campaign (version 25.10.2). 21 2.6 Discussions - Physics Background to DVCS using 5×41 GeV2 To estimate the physics background for DVCS measurements, the single-photon misidentification rate is evaluated using the exclusive π0process. A background event occurs when one of the two photons from a neutral pion is misidentified as a DVCS photon, which can result from geometrical acceptance, energy thresholds, or detector granularity. The procedure is as follows: Electromagnetic calorimeters in the backward, mid-rapidity, and forward regions are used. First, the scattered electron is excluded, and a minimum cluster energy cut of 0.1 GeV is applied. Events with only a single reconstructed photon cluster are then counted. The kinematic distributions of neutral pions and their two decay photons are shown in Figure 16. Most neutral pions, particularly those with higher momenta, are boosted into the forward region, along with their decay photons. The two photons are concentrated within similar pseudorapidities, with an opening-angle threshold of approximately 0.01 radians. As a result, very few events lose a photon purely due to geometric acceptance, as illustrated in Figure 17. 1 10 2 10 3 10 6−4−2−0 2 4 6 MC 0 π η 0 5 10 15 20 25 30 [GeV/c] MC 0 π p 1 10 2 10 3 10 6−4−2−0 2 4 6 MC 1,2 γ η 0 5 10 15 20 25 30 [GeV/c] MC 1,2 γ p Figure 16: Two-dimensional distributions of momentum versus pseudorapidity for neutral pions (left) and their two decay photons (right). Results are obtained using the ePIC October Simulation Campaign (version 25.10.3). 22 1 10 2 10 3 10 4 10 6−4−2−0 2 4 6 MC 1 γ η 6− 4− 2− 0 2 4 6 MC 2 γ η 0 0.02 0.04 0.06 0.08 0.1 [rad] MC 2 γ 1 γ θ 1 10 2 10 3 10 4 10 Events Figure 17: (Left) Two-dimensional pseudorapidity distribution of the two decay photons. (Right) Opening-angle distribution of the two decay photons. Results are obtained using the ePIC October Simulation Campaign (version 25.10.3). The physics background is first estimated using the current EICrecon reconstruction pipeline, which gives a misidentification rate of 33.6%. However, there is significant potential for improvement. For example, forward EMCAL reconstruction parameters can be optimized based on its granularity (0.015 rad). A simple algorithm was implemented for the forward calorimeter to identify separable events: if the distance between the first and second highestenergy hits exceeds twice the tower size, the event is considered separable. An example is shown in Figure 18, where the reconstructed hit map in XY coordinates is colored by energy deposition. Roughly one pixel corresponds to one tower. Although this event is currently reconstructed as a single cluster, the MC-level opening angle and distance between the two highest-energy towers indicate that it can be resolved as two clusters. 400−300−200−100−0 100 200 X [mm] 100 200 300 400 500 600 700 Y [mm] hHitPosXY Entries 36 Mean x 96.76− Mean y 336.5 Std Dev x 16.92 Std Dev y 27.02 0 1 2 3 4 5 6 7 8 9 hHitPosXY Entries 36 Mean x 96.76− Mean y 336.5 Std Dev x 16.92 Std Dev y 27.02 Figure 18: Reconstructed hit map of a forward calorimeter event from the 5×41 GeV2sample. Color represents energy deposition. This event is currently reconstructed as one cluster, but hit-level information shows it can be resolved into two clusters. Results are obtained using the ePIC October Simulation Campaign (version 25.10.3). 23 In addition, standalone machine-learning (ML)–based studies were incorporated as a potential improvement for future reconstruction, particularly targeting mid-rapidity in the imaging barrel calorimeter. Detector performance shows that up to a neutral pion momentum of 35 GeV, two photons can be separated near pseudorapidity η= 0 (Fig. 8.148(b) in the ePIC pre-TDR v2.2). After implementing both hit-level improvements in the forward calorimeter and the barrel ML method, the misidentification rate is reduced to approximately 11.9%. Further improvements using forward ML, based on the ePIC pre-TDR documentation (Fig. 8.162 in v2.2), reduce the misidentification rate at 60 GeV to about 0.12%. Considering the cross sections of DVCS and exclusive π0, the conservative estimate for single-photon misidentification as a background contribution is 1.14%, consistent with recent EIC DVCS studies [7]. 2.7 Discussions - Energy threshold in the forward calorimeter In the EIC Yellow Report, the minimum energy threshold for the forward electromagnetic calorimeter (EMCal) was set to 100 MeV. At the request of the calorimeter group, a study was performed to evaluate the impact of increasing these thresholds to 200 MeV in the pseudorapidity range 1.4< η < 3.0, and to 400 MeV in the range 3.0< η < 3.5. To assess the impact of these raised energy thresholds in the forward region, DVπ0P and DVCS event samples were generated, and a simplified study was conducted at the generator level. The effects were evaluated in terms of π0reconstruction efficiency and the resulting background contribution to DVCS measurements. Figure 19 shows the kinematic distributions of the two decay photons from π0in DVπ0P events and the single photon from DVCS events. The π0decay photons are predominantly produced in the forward region, while the DVCS photon is primarily distributed in the midrapidity region. Regarding π0reconstruction efficiency, increasing the EMCal energy threshold from 100 MeV to 200 MeV in the range 1.4<η<3.0 results in a modest overall impact, as shown in Figure 20. The reduction in efficiency is primarily observed for low-momentum π0s, where the efficiency drops by approximately 10 %. High-momentum π0s are less affected due to their more energetic decay photons. Although low-momentum π0s dominate the cross-section, the total loss in reconstruction efficiency is partially offset by the preserved detection of higherenergy decay products. In the range 3.0<η<3.5, increasing the threshold to 400 MeV has negligible effect, as π0s in this region are typically highly boosted and their decay photons are well above the new threshold. To quantify the background contribution to DVCS measurements, Figure 21 compares photon momentum distributions from DVπ0P and DVCS events. The left panel corresponds 24 to the 100 MeV threshold, and the right to the 200 MeV case. The DVCS histogram (blue solid line) is scaled to match the integrated luminosity of the DVπ0P sample (black solid line). The black open circles represent DVπ0P events where one decay photon is undetected due to threshold cuts, potentially mimicking a DVCS event. To estimate the background contribution, we integrate the distribution of these singlephoton events (where one photon is missed and the other is within detector acceptance) and compare it to the DVCS distribution. In the pseudo-rapidity range 1.4<η<3.0, the background contribution is below 1 % at the 100 MeV threshold and approximately doubles when the threshold is raised to 200 MeV. In the 3.0<η<3.5 range, the background remains negligible, even with the increased threshold. 1 10 2 10 3 10 4 10 8−6−4−2−02468 MC 2 γ η or MC 1 γ η 0 10 20 30 40 50 60 70 80 90 100 [GeV] MC 2 γ or p MC 1 γ p hGammaMomVsEtaMC Entries 1999998 Mean x 1.376 Mean y 2.15 Std Dev x 0.6507 Std Dev y 2.632 8−6−4−2−0 2 4 6 8 MC γ η 0 10 20 30 40 50 60 70 80 90 100 [GeV] MC γ p hGammaMomVsEtaMC Entries 999999 Mean x 0.3969− Mean y 2.844 Std Dev x 1.403 Std Dev y 2.309 1 10 2 10 3 10 4 10 hGammaMomVsEtaMC Entries 999999 Mean x 0.3969− Mean y 2.844 Std Dev x 1.403 Std Dev y 2.309 Figure 19: Kinematic distributions of the two decay photons from π0in DVπ0P events (left) and the single photon from DVCS events (right). 0 5 10 15 20 25 30 [GeV/c] MC 0 π p 0 0.2 0.4 0.6 0.8 1 Efficiency 0 5 10 15 20 25 30 [GeV/c] MC 0 π p 0 0.2 0.4 0.6 0.8 1 Efficiency Figure 20: π0reconstruction efficiency as a function of momentum at minimal energy threshold 100 MeV (left) and 200 MeV (right) in the pseudo-rapidity range 1.4<η<3.0. 25