WorksPhysics and Astronomy
ISR Reconstruction of Radiative Charmonia as the Origin of the Belle–BESIII Line-Shape Puzzle
Abstract
A longstanding disagreement persists between Belle and BESIII over the line shapes of , , and near 4.63 GeV. Belle observes enhancements through initial state radiation that are absent or much weaker at BESIII. We propose that the extra Belle yield arises from radiative production of a charmonium. Luminosity normalization increases this contribution to the extracted cross sections, so tens of fb can explain the enhancements. Belle II can test this mechanism through full reconstruction and searches for missing charmonia in open charm modes.
Introduction
The observation of charmonium-like states has opened a window onto hadron spectroscopy and deepened our understanding of the nonperturbative behavior of the strong interaction. Among these states, the states—charmonium-like states produced in electron–positron annihilation—constitute the so-called problem [1, 2], which has motivated a variety of theoretical efforts [3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]; see Refs. [14, 15, 16] for recent reviews.
A long-standing puzzle in this context concerns the invariant mass spectrum. In 2008, the Belle Collaboration reported a higher state, , in [17]. With the accumulation of experimental data, the BESIII Collaboration searched for through the same process [18, 19] but found no enhancement structure in this energy range. This discrepancy constitutes a persistent Belle–BESIII difference in the line shape. Despite serious cross-checks by both collaborations, no explanation has been found.
Remarkably, this phenomenon is not unique. With the BESIII and Belle measurements of and [20, 21, 22], the same Belle–BESIII difference emerges in the line shapes of the corresponding invariant mass spectra. This suggests an underlying dynamical mechanism common to these channels, the identification of which is a timely task in this area.
Motivated by the strong coupling of higher states to [23], we conjecture that the spectrum measured by Belle contains a higher signal. The key point is the reconstruction method used in initial state radiation (ISR) analyses. Full reconstruction requires reconstructing both hadrons from their decay products, as in Belle’s analysis [24]. Their measured momenta allow the photon momentum to be inferred and its angular distribution to be determined without photon detection.
ton detection. In the channel, Belle uses partial reconstruction: one and an energetic photon are reconstructed, and the other baryon is identified through recoil constraints [17]. Radiative events can satisfy the photon energy and recoil mass requirements and enter Belle’s sample.
In this Letter, we extend this line of reasoning to the and channels. We show that the inclusion of higher charmonium contributions naturally accounts for the BESIII–Belle difference in the line shapes of these invariant mass spectra: higher charmonia lying above the and thresholds can couple to these channels and produce additional resonant contributions that may appear in the Belle measurements but be absent in those of BESIII. Normalization by the small effective ISR luminosity increases their contribution to Belle’s extracted cross sections, allowing radiative products of tens of fb to generate visible enhancements and account for the observed line shape differences. The Supplemental Material (SM) [25] gives the fitting methods and results of the spectroscopy calculations.
Production mechanism.—Figure 1 compares the two topologies yielding a photon and two hadrons . Here is the collision energy and the hadron pair invariant mass. Panel (a) shows ISR at Belle; after photon emission, the virtual photon produces the hadron pair through the same transition measured by BESIII:

Figure 1. (a) ISR production of the hadron pair at Belle. The circle is the transition also measured by BESIII. (b) Radiative production of at Belle’s energy, followed by . The square is ; the filled dot is the decay. Here is the squared invariant mass of and the squared collision energy.
Panel (b) is radiative production of a charmonium resonance at the full Belle energy,
Both photons in panel (b) have , so charge conjugation fixes . Radiative production allows states with and states with , ; these labels include all states with the corresponding quantum numbers. At GeV, a resonance near 4.63 GeV recoils against a photon whose energy in the collision center of mass is
In a BESIII direct scan at the same hadron pair mass, the collision energy is used to produce the pair, leaving no energy for a hard photon.
The ISR and radiative processes have the same final state and can satisfy the same recoil requirements. In a fully reconstructed analysis, the squared mass recoiling against the two hadrons, , is required to be close to zero, as expected for one missing photon. The three Belle analyses considered here use partial reconstruction: one hadron and an energetic photon are reconstructed, and the other hadron is identified through recoil constraints. Requiring full reconstruction of only one charm hadron improves efficiency, especially when the accessible decay branching fractions are small [17, 29]. For , Belle reconstructs one , requires the photon energy to exceed 3.5 GeV, and identifies the recoiling by together with an antiproton from its decay [17]. The two analyses impose the same kind of recoil constraint and require a photon above 3 GeV in the collision frame [20, 21]. Equation (3) shows that an ISR photon and a radiative photon have the same energy at the same pair mass, so the cut on the photon energy and the recoil mass do not separate them.
In addition, full reconstruction allows the photon momentum and direction to be inferred from the two hadron momenta through energy and momentum conservation, without requiring photon detection. Belle’s analysis used this method and observed the characteristic ISR peaks along the beams [24]. In the partially reconstructed samples considered here, the photon polar angle relative to the electron beam in the collision center of mass is measured directly from the detected photon.
Its angular coverage is limited by the calorimeter acceptance [30], excluding the directions closest to the beams where ISR is most strongly peaked. The predicted photon distribution for radiative production has no sharp peaks along the beams [31]. Within the central window used in several Belle ISR analyses [24, 32, 33, 34], the ISR [35] and radiative angular shapes are similar [25]. The restricted acceptance and present statistics therefore limit their separation. We add their yields because interference cancels when final states and their charge conjugates are combined with equal weight [35].
Belle normalizes the event yield in each hadron mass bin using the effective ISR luminosity,
where is the collected integrated luminosity, are the edges of mass bin , and is the ISR radiator, the probability density in the photon energy fraction [20, 36]. In ISR, an incoming lepton emits a photon before annihilation, lowering the effective annihilation energy to the invariant mass of the hadron pair. Selecting a narrow hadron mass bin restricts this photon to a narrow energy interval. The small emission probability in that interval is the integral in Eq. (4), so .
Radiative production in Eq. (2) proceeds through annihilation at the full collision energy, with the photon produced together with . No preceding ISR emission is needed to lower the annihilation energy. For channel and mass bin , is the radiative product cross section integrated over the bin and normalized with the full Belle luminosity . If these events enter the ISR sample, their contribution to Belle’s reported cross section is extracted using the smaller effective ISR luminosity ; we denote this contribution by .
The radiative product is the production cross section multiplied by and any daughter branching fractions retained in the measured observable. The published baryon cross section is corrected for the reconstruction and inclusive branching fractions [17], while the meson products retain the decay branching fractions [20, 21]. The corresponding reconstruction efficiencies are and . We multiply Belle’s dressed meson cross sections and uncertainties by the vacuum polarization correction factor 0.94 to express them in the Born convention used for the BESIII vector baselines [20, 21]. We quote and with the same Born normalization. Normalizing the radiative contribution with the two luminosities gives
We set . Counting radiative events as ISR therefore multiplies their cross section by .
Belle reports for and for each meson channel [17, 20, 21]. In , the bin has , efficiency including the branching fractions of the reconstructed decay modes, and a fitted signal yield of 21.4 events [20]. If these events were radiatively produced, their contribution in the Born normalization adopted for the meson fits would be
The same yield corresponds to a radiative product of
With , a radiative product of 16.5 fb in this bin would contribute 168.2 pb to the spectrum in the same Born normalization.
Fits and candidate analysis.—We describe the Belle invariant mass spectra as the sum of the vector and radiative contributions,
where is the total Born cross section fitted to Belle’s ISR spectrum, the vector contribution common to Belle ISR and BESIII direct scans, and the extra contribution extracted when radiative events are normalized with the ISR effective luminosity.
For we use the cross section of Ref. [37], including the , , , and contributions and their interference. Keeping its resonance masses, width parameters, and fitted form factor parameters at their published values, we compare the baseline with the sixteen BESIII points from 4.575 to 4.951 GeV and obtain [18, 19].
For the meson channels, we adopt BESIII’s published total fits containing two interfering vector Breit–Wigner amplitudes near 4.6 GeV and 4.7–4.8 GeV. BESIII reports for and for [22]. These vector baselines are fixed as background in the Belle fits.
We describe with a relativistic Breit–Wigner and the phase space for ,
Here, is the phase space density for as a function of the hadron pair mass . Radiative production of such highly excited charmonia may involve complex unquenched effects [16], making the production strengths difficult to calculate. We therefore determine the overall normalization from a fit to each Belle spectrum. In the fitted line shape, we treat the smooth luminosity conversion factor as constant near the resonance and absorb it into . For resonance mass and width at , the relativistic Breit–Wigner function is . Its width depends on the hadron momentum in the rest frame and relative orbital angular momentum as
with and Blatt–Weisskopf barrier factors [38]. We use a barrier radius [39]. The decay factor is [39], with . For a massless photon, standard kinematics [40] gives
Below threshold, . We use for . For the allowed decays of states, we use ; for states decaying into , we use . The bin averages in the Born convention used in the fits give the radiative product over the mass window :
For the luminosity conversion, we treat all Belle data as taken at .
The fit uses sixteen Belle bins of 20 MeV from 4.58 to 4.90 GeV [17]. The meson fits cover 4.51–4.95 GeV for and 4.54–4.94 GeV for [20, 21]; their uncertainty treatment is given in the SM [25]. For , the error bars in Belle’s published figure show statistical uncertainties only. We include the 10% experimental systematic of Ref. [17] in each bin, and an estimated common 4.9% uncertainty based on recent branching fraction measurements [41, 42]. The data term with common normalization is
Here, labels the fitted mass bins, is the measured cross section, and is the total prediction averaged over bin . The factor accounts for the common branching fraction uncertainty. The uncertainty combines the statistical and 10% experimental systematic errors in quadrature. The full also includes the normalization constraint given in Ref. [25]. Uncertainties on fitted parameters are the profile intervals [40].
We denote the additional resonances in , , and by , , and , respectively. Figure 2 shows the fits to these three spectra. Our previous analysis [23] studied an contribution to the spectrum through an -wave decay. We therefore consider this state as a candidate for and adopt the fixed input (4629, 24.2) MeV from that study. As shown in Figure 2(a), this radiative contribution can describe Belle’s enhancement near 4.63 GeV above the fixed vector baseline. The fit reduces /dof from 17.12/16 for the vector baseline alone to 6.81/15, with a local significance of . We also allow the mass and width to vary, obtaining MeV and MeV, with /dof = 4.67/13. Both fixed inputs lie within the corresponding individual intervals of this best fit [25].
Figure 2. Fits to the (a) , (b) , and (c) spectra. Black circles and green open squares show Belle and BESIII data. Grey dotted curves show the fixed vector baselines . Red solid and purple dash-dotted curves show the total fits and radiative contributions , including the fitted normalization. The labels , , and correspond to (a)–(c). The mass and width of are fixed; those of and are fitted. Error bars include statistical and systematic uncertainties.
For , the fit gives MeV and MeV, lowering /dof from 35.72/22 for the vector baseline to 21.61/19. For , the fit gives MeV and MeV, with /dof = 8.87/17, compared with 17.12/20 for the baseline alone. With their masses and widths fixed at the fitted values, the local significances are for and for . The fitted radiative products are fb, fb, and fb [25]. These contributions of tens of fb can produce the Belle enhancements through the ISR luminosity normalization. The inputs used for lie within the corresponding individual intervals of the free fit. Fixing MeV gives /dof = 9.69/19 [25], so an contribution can also describe this spectrum. Since is forbidden, we consider , , and states with allowed -wave decays to this channel as candidates for .
To examine possible charmonium assignments, we calculate the spectrum with a screened potential and the strong decays with the quark pair-creation (QPC) model [1, 2, 10, 43, 44, 45], retaining the parameters of our previous study [23]. The central calculated masses of the , , and states are [25]
The fitted mass lies between the central predictions for these two multiplets. Orbital mixing has been considered in studies of vector charmonia [1, 10]. Among the states, only the has an -wave partner with the same . Possible mixing of with , whose calculated mass is MeV, could lower the mass into the fitted range. We discuss this mixing possibility for a interpretation of in the SM [25]. In addition, the possibly mixed candidate may also contribute to the channel. However, the calculated partial width into this channel is larger than those of the mixed tensor candidates, supporting as a candidate for .
Summary.— We have proposed radiative production of higher charmonia as a possible common origin of the differences between the Belle and BESIII line shapes. Radiatively produced charmonia can pass Belle’s partial reconstruction requirements, and normalization by the effective ISR luminosity increases their contribution to the extracted cross sections. The baryonic and enhancements can be described by contributions, while a candidate can describe . Possible – mixing offers an interpretation of the tensor candidate in the latter channel. The spectrum and decay properties of highly excited charmonia remain incompletely understood, and further studies are needed to identify the contributing states.
With larger samples, comparing photon angular distributions in the enhancement region and neighboring mass intervals would help test whether the excess contains a radiative component. At Belle II, full reconstruction of both hadrons would allow the recoil photon direction to be inferred from their momenta, helping distinguish ISR from radiative production. Hadron decay angles would also provide information on the resonance spin. For example, angular correlations involving in the channel could help distinguish the and assignments.
Similar differences in the line shapes may also occur in other open charm channels. Many Belle points in between 4.2 and 4.4 GeV lie above the BESIII measurements, although the two data sets remain consistent within uncertainties [15]. In this region, well above the threshold, the large vector baseline means that an additional radiative contribution may produce only a modest relative increase in the total cross section. Channels opening closer to the resonance region, such as , may therefore be more promising for observing a similar effect. Close collaboration between theorists and experimentalists will help test this mechanism and search for similar resonance structures in other channels.
Acknowledgments
This work is supported by the Natural Science Foundation of Gansu Province (No. 26RCKA012 and No. 25JRRA799), the National Natural Science Foundation of China under Grants No. 12335001, No. 12247101, and No. 12135005, the 111 Center under Grant No. B20063, the National Key R&D Program of China under Contract No. 2024YFA1610503, the Fundamental Research Funds for the Central Universities (lzujbky-2023-stlt01), and Lanzhou City High-Level Talent Funding. T.-L. G. is supported by the Gansu Province Postgraduate Innovation Star Program No. 2026CXZX-042. The authors used ChatGPT for language editing and for assistance with the numerical code, reviewed the output, and take responsibility for the content.
Data availability. The cross sections come from Refs. [17, 18, 19, 20, 21, 22] and their supplemental material. The Belle points are digitized from the published figure.
Fitting methods
Photon angular distributions
The photon polar angle is measured relative to the electron beam in the collision center of mass. The radiative distributions follow the calculation at leading order in Ref. [31], and the ISR distribution follows Ref. [35]. Each curve is averaged over bins of width 0.05 in and normalized to unit area within .
The full angular range is shown to display the ISR peaks along the beams. When both hadrons are reconstructed, their momenta determine the recoil photon direction, allowing the ISR peaks along the beams to be studied without detecting the photon [24]. In the partially reconstructed samples considered here, the photon must be detected, and the calorimeter acceptance excludes the directions closest to the beams.
We use as the central comparison interval, following several Belle ISR analyses [24, 32, 33, 34]. Figure S1 shows that the ISR and radiative distributions have similar shapes in this interval, where the strongest ISR peaks are excluded. The restricted angular coverage and present statistics therefore limit separation of the two processes.
Figure S1. Theoretical photon angular distributions at GeV and hadron pair mass GeV. Steps show averages over 0.05 bins. Each distribution has unit area within , marked by the vertical gray dashed lines.
Data and fitting method
The fits use 20 MeV bins: 16 bins over 4.58–4.90 GeV for , 22 over 4.51–4.95 GeV for , and 20 over 4.54–4.94 GeV for [17, 20, 21]. Belle’s collected integrated luminosity is 695 fb for the baryonic channel and 921.9 fb for each meson channel [17, 20, 21]. The central values of the published cross sections are retained. Channels related by charge conjugation are included throughout. The points are digitized from the published figure [21].
For , corresponding to , , and , respectively, denotes the radiative product including the decay branching fractions . Here is the additional resonance and the hadron pair in channel . These products use the same Born normalization as the fitted spectra.
For , , because Belle’s published cross section has been corrected for the reconstruction and inclusive branching fractions [17]. The measured products in the meson channels include for and for , while the branching fractions of the reconstructed decay modes are included in the efficiencies [20, 21].
We average the fitted Born contribution over each experimental mass bin and denote the result by . The radiative product over the mass window is then
The ISR luminosity is defined in the Letter [20, 21, 36]. We set , so each bin contribution is multiplied by before summing over the mass window. For the luminosity conversion, we treat all Belle data as taken at GeV.
We fit the bin averages of the total cross section described in the Letter, with the vector baselines fixed as background [18, 19, 22, 26, 37]. For each channel , the data contribution to is
Here is the measured cross section, is the average total prediction in bin , and denotes the resonance parameters varied in the fit. The uncertainty combines the statistical error and systematic errors that are independent between bins. A single factor multiplies the total prediction in every bin and is fitted together with the resonance parameters. The total also includes the Gaussian constraint , where is the relative common normalization uncertainty. The upper and lower limits for each parameter are the values at which increases by 1 from its minimum. The other free parameters, including , are refitted when determining these limits [40].
Experimental uncertainties
Belle’s original figure gives statistical errors only [17]. Its total systematic uncertainty of about 42% is dominated by branching fraction uncertainties. We retain the 10% experimental contribution in the error for each point and estimate the common branching fraction uncertainty from recent BESIII measurements.
BESIII measures [41] and [42]. Including subsequent antihyperon decays, gives . Using the precision of as a reference for the reconstructed modes, we conservatively add the two relative errors, , to estimate the common normalization uncertainty.
For , we estimate the common normalization uncertainty as 4.9% based on Belle’s efficiency variation, retaining the remaining systematic errors for each point [20].
| Fixed | Fixed | Free | Free | Free |
| [MeV] | 4629 | |||
| [MeV] | 24.2 | |||
| [(pb/GeV)] | ||||
| [pb] | ||||
| [fb] | ||||
| 6.81/15 | 4.67/13 | 21.61/19 | 8.87/17 |
Table SM-I. Fit results for the three Belle spectra with fixed vector baselines. The vector baselines alone give , , and in these channels, respectively. Radiative products use the same Born normalization as the fitted spectra and the mass intervals given above. The coefficient is the fitted amplitude normalization for channel . Errors correspond to .
For , we treat the full 10.9% systematic uncertainty as a common normalization uncertainty and use the statistical error for each point [21].
Fit results
Panels (a) and (b) of Figure S2 compare the fits with fixed and free mass and width. We adopt the fixed input MeV from the previous analysis of the line shape [23]. Both fixed inputs lie within the corresponding individual intervals of the free fit. The local significance with fixed mass and width is .
Figure S2. Fits to Belle data (black circles) for in (a,b), in (c), and in (d); green open squares show BESIII data. The resonance mass and width are fixed in (a) and fitted in (b)–(d). The resonance terms use in (a,b), in (c), and in (d). Grey dotted, red solid, and purple dash-dotted curves show the fixed vector baseline, total fit, and radiative contribution. The total and radiative curves include the fitted normalization. Error bars include statistical and systematic uncertainties.
Panels (c) and (d) of Figure S2 show the meson fits. With the masses and widths fixed at their fitted values, the local significances are for and for . Fixing the mass and width in to MeV gives and a local significance of ; both inputs lie within the individual intervals of the free fit.
Spectroscopy calculations
We calculate the charmonium masses and wave functions in the modified Godfrey–Isgur (MGI) model used in our previous study [23]. The linear scalar confinement is replaced by a screened form, written before smearing as
Here is the quark–antiquark separation, and is the screening parameter. The color Coulomb and spin dependent interactions, Gaussian smearing, and momentum dependent relativistic corrections follow the GI formalism [43] as used in Refs. [2, 23].
We adopt the central parameters GeV,
GeV, GeV, and GeV from Ref. [23]. Following the same study, we estimate the mass uncertainties by varying GeV. The values and are used at and GeV, respectively, with all other parameters unchanged. The quoted errors give the shifts from the central masses to the largest and smallest masses obtained with these three parameter sets.
Table SM-II gives the calculated masses and model uncertainties of the nearby pseudoscalar and states. The central masses of the and multiplets lie below and above the fitted meson masses, while lies close to the enhancement region.
The strong decays are calculated in the quark pair-creation (QPC) model [44, 45], in which a light pair is created with . The strengths for creating and pairs are and , respectively, and the constituent quark masses are GeV and GeV. The wave functions of the final mesons are taken from [27, 28]. The final meson masses are GeV.
| State | Mass | State | Mass |
Table SM-II. Calculated masses in MeV. The errors show the model variation under the three parameter sets specified in the text.
For , we use the spin mixture , with . The decay calculations use the fit mass, 4625.1 MeV, for . For the candidate formed by mixing and
, we use the fit mass of 4627.3 MeV. The partial widths are calculated with the amplitude normalization of Ref. [23],
Here is the QPC partial-wave decay amplitude for channel with final total spin and relative orbital angular momentum . The two final mesons have momentum magnitude in the resonance rest frame. Charge conjugates are included.
To illustrate possible mixing of and , we denote the lower and higher mass states by and , respectively:
Here and denote the unmixed and states, both with , and is the mixing angle. Using the central 6P and 5F masses, we choose to place at the fitted mass of 4627.3 MeV. This angle illustrates possible additional mixing.
| State | [MeV] | ||
| 4625.1 | 0 | 1.4 | |
| 4627.3 | 4.2 | 0.1 | |
| 4627.3 | 5.1 | 0.1 |
Table SM-III. Partial widths in keV, using the fitted mass for and the fitted mass for .
Table SM-III shows that the calculated partial width into is larger than those of the mixed candidates. We therefore use the candidate for this channel and a tensor candidate for , where the decay is forbidden. Additional tensor contributions to remain possible; their relative yields depend on radiative production rates, full resonance widths, and reconstruction efficiencies. Further angular measurements can help separate the contributions.
References
References
- [1]J.-Z. Wang, D.-Y. Chen, X. Liu, and T. Matsuki, Constructing J/ψ family with updated data of charmoniumlike Y states, Phys. Rev. D 99, 114003 (2019), arXiv:1903.07115 [hep-ph].DOI
- [2]J.-Z. Wang, R.-Q. Qian, X. Liu, and T. Matsuki, Are the Y states around 4.6 GeV from e⁺e⁻ annihilation higher charmonia?, Phys. Rev. D 101, 034001 (2020), arXiv:2001.00175 [hep-ph].DOI
- [3]D.-Y. Chen, J. He, and X. Liu, Nonresonant explanation for the Y(4260) structure observed in the e⁺e⁻ → J/ψπ⁺π⁻ process, Phys. Rev. D 83, 054021 (2011), arXiv:1012.5362 [hep-ph].DOI
- [4]D.-Y. Chen, J. He, and X. Liu, Novel explanation of charmoniumlike structure in e⁺e⁻ → ψ(2S)π⁺π⁻, Phys. Rev. D 83, 074012 (2011), arXiv:1101.2474 [hep-ph].
- [5]D.-Y. Chen, X. Liu, X.-Q. Li, and H.-W. Ke, Unified Fano-like interference picture for charmoniumlike states Y(4008), Y(4260) and Y(4360), Phys. Rev. D 93, 014011 (2016), arXiv:1512.04157 [hep-ph].DOI
- [6]D.-Y. Chen, X. Liu, and T. Matsuki, Interference effect as resonance killer of newly observed charmoniumlike states Y(4320) and Y(4390), Eur. Phys. J. C 78, 136 (2018), arXiv:1708.01954 [hep-ph].DOI
- [7]L.-P. He, D.-Y. Chen, X. Liu, and T. Matsuki, Prediction of a missing higher charmonium around 4.26 GeV in J/ψ family, Eur. Phys. J. C 74, 3208 (2014), arXiv:1405.3831 [hep-ph].DOI
- [8]D.-Y. Chen, X. Liu, and T. Matsuki, Observation of e⁺e⁻ → χc0ω and missing higher charmonium ψ(4S), Phys. Rev. D 91, 094023 (2015), arXiv:1411.5136 [hep-ph].
- [9]D.-Y. Chen, X. Liu, and T. Matsuki, Search for missing ψ(4S) in the e⁺e⁻ → π⁺π⁻ψ(2S) process, Phys. Rev. D 93, 034028 (2016), arXiv:1509.00736 [hep-ph].DOI
- [10]J.-Z. Wang and X. Liu, Confirming the existence of a new higher charmonium ψ(4500) by the newly released data of e⁺e⁻ → K⁺K⁻J/ψ, Phys. Rev. D 107, 054016 (2023), arXiv:2212.13512 [hep-ph].DOI
- [11]J.-Z. Wang and X. Liu, Identifying a characterized energy level structure of higher charmonium well matched to the peak structures in e⁺e⁻ → π⁺π⁻D⁰D̄⁰, Phys. Lett. B 849, 138456 (2024), arXiv:2306.14695 [hep-ph].arxiv.org/abs/2306.14695
- [12]T.-C. Peng, Z.-Y. Bai, J.-Z. Wang, and X. Liu, How higher charmonia shape the puzzling data of the e⁺e⁻ → ηJ/ψ cross section, Phys. Rev. D 109, 094048 (2024), arXiv:2403.03705 [hep-ph].DOI
- [13]Z.-L. Man, S.-Q. Luo, Z.-Y. Bai, and X. Liu, Coupled-channel study of 4S-3D mixing dynamics in ψ(4220) and ψ(4380), Phys. Lett. B 868, 139644 (2025), arXiv:2502.08072 [hep-ph].DOI
- [14]H.-X. Chen, W. Chen, X. Liu, Y.-R. Liu, and S.-L. Zhu, An updated review of the new hadron states, Rept. Prog. Phys. 86, 026201 (2023), arXiv:2204.02649 [hep-ph].DOI
- [15]X. Wang, X. Liu, and Y. Gao, Colloquium: Hadron production in open-charm meson pairs at e⁺e⁻ colliders, Rev. Mod. Phys. 98, 021001 (2026), arXiv:2502.15117 [hep-ex].DOI
- [16]Z.-Y. Bai, D.-Y. Chen, Q. Huang, X. Liu, S.-Q. Luo, and J.-Z. Wang, Unquenched charmonium and beyond, Phys. Rept. 1204, 1 (2027), arXiv:2602.19887 [hep-ph].DOI
- [17]G. Pakhlova et al. (Belle), Observation of a near-threshold enhancement in the e⁺e⁻ → Λc⁺Λ̄c⁻ cross section using initial-state radiation, Phys. Rev. Lett. 101, 172001 (2008), arXiv:0807.4458 [hep-ex].DOI
- [18]M. Ablikim et al. (BESIII), Precision measurement of the e⁺e⁻ → Λc⁺Λ̄c⁻ cross section near threshold, Phys. Rev. Lett. 120, 132001 (2018), arXiv:1710.00150 [hep-ex].DOI
- [19]M. Ablikim et al. (BESIII), Measurement of Energy-Dependent Pair-Production Cross Section and Electromagnetic Form Factors of a Charmed Baryon, Phys. Rev. Lett. 131, 191901 (2023), arXiv:2307.07316 [hep-ex].DOI
- [20]S. Jia et al. (Belle), Observation of a vector charmonium-like state in e⁺e⁻ → D⁺sDs1(2536) + c.c., Phys. Rev. D 100, 111103 (2019), arXiv:1911.00671 [hep-ex].
- [21]S. Jia et al. (Belle), Evidence for a vector charmoniumlike state in e⁺e⁻ → D⁺sD∗−s(2573) + c.c., Phys. Rev. D 101, 091101 (2020), arXiv:2004.02404 [hep-ex].
- [22]M. Ablikim et al. (BESIII), Study of the decay and production properties of Ds1(2536) and Ds2(2573), Phys. Rev. Lett. 133, 171903 (2024), arXiv:2407.07651 [hep-ex].arxiv.org/abs/2407.07651
- [23]C.-X. Liu, J.-Z. Wang, and X. Liu, Exploring charmonia ηc(6S, 7S) in the Λ⁺cΛ̄⁻c invariant mass spectrum of B⁰ → K⁻sΛ⁺cΛ̄⁻c, Phys. Rev. D 114, 074050 (2026), arXiv:2608.15524 [hep-ph].
- [24]G. Pakhlova et al. (Belle), Measurement of the near-threshold e⁺e⁻ → D̄D cross section using initial-state radiation, Phys. Rev. D 77, 011103 (2008), arXiv:0708.0082 [hep-ex].DOI
- [25]See Supplemental Material for the fitting methods and results of the spectroscopy calculations. The Supplemental Material cites Refs. [26–28].
- [26]C. Chen, B. Yan, and J.-J. Xie, The electromagnetic form factors and spin polarization of Λ⁺c in the process e⁺e⁻ → Λ⁺cΛ̄⁻c, Chin. Phys. C 49, 023102 (2025), arXiv:2407.19445 [hep-ph].DOI
- [27]Q.-T. Song, D.-Y. Chen, X. Liu, and T. Matsuki, Higher radial and orbital excitations in the charmed meson fam ily, Phys. Rev. D 92, 074011 (2015), arXiv:1503.05728 [hep-ph].
- [28]Q.-T. Song, D.-Y. Chen, X. Liu, and T. Matsuki, Charmed-strange mesons revisited: Mass spectra and strong decays, Phys. Rev. D 91, 054031 (2015), arXiv:1501.03575 [hep-ph].DOI
- [29]V. Zhukova et al. (Belle), Angular analysis of the e+e− → D(*)±D(*)∓ process near the open charm threshold using initial-state radiation, Phys. Rev. D 97, 012002 (2018), arXiv:1707.09167 [hep-ex].DOI
- [30]A. Bashirian et al. (Belle), The Belle detector, Nucl. Instrum. Methods Phys. Res. A 479, 117 (2002).
- [31]H. S. Chung, J. Lee, and C. Yu, Exclusive heavy quarkonium +γ production from e+e− annihilation into a virtual photon, Phys. Rev. D 78, 074022 (2008), arXiv:0808.1625 [hep-ph].DOI
- [32]G. Pakhlova et al. (Belle), Observation of the ψ(4415) → D̄D*2(2460) decay using initial-state radiation, Phys. Rev. Lett. 100, 062001 (2008), arXiv:0708.3313 [hep-ex].
- [33]G. Pakhlova et al. (Belle), Measurement of the e+e− → D0D̄*0 + π± cross section using initial-state radiation, Phys. Rev. D 80, 091101 (2009), arXiv:0908.0231 [hep-ex].DOI
- [34]G. Pakhlova et al. (Belle), Measurement of e+e− → D(*)s+D(*)s− cross sections near threshold using initial-state radiation, Phys. Rev. D 83, 011101 (2011), arXiv:1011.4397 [hep-ex].
- [35]S. Binner, J. H. Kühn, and K. Melnikov, Measuring σ(e+e− → hadrons) using tagged photons, Phys. Lett. B 459, 279 (1999), arXiv:hep-ph/9902399.
- [36]G. Pakhlova et al. (Belle), Measurement of the near-threshold e+e− → D(*)±D∓ cross section using initial-state radiation, Phys. Rev. Lett. 98, 092001 (2007), arXiv:hep-ex/0608018.
- [37]C. Chen, B. Yan, and J.-J. Xie, e+e− → Λ+cΛ̄−c cross sections and the Λ+c electromagnetic form factors within the extended vector meson dominance model, Chin. Phys. Lett. 41, 021302 (2024), arXiv:2312.16753 [hep-ph].DOI
- [38]F. von Hippel and C. Quigg, Centrifugal-barrier effects in resonance partial decay widths, shapes, and production amplitudes, Phys. Rev. D 5, 624 (1972).
- [39]R. Aaij et al. (LHCb), Observation of J/ψp resonances consistent with pentaquark states in Λ0b → J/ψK−p decays, Phys. Rev. Lett. 115, 072001 (2015), arXiv:1507.03414 [hep-ex].
- [40]F. Takahashi et al. (Particle Data Group), Review of Particle Physics, Int. J. Mod. Phys. A 41, 2630011 (2026).
- [41]M. Ablikim et al. (BESIII), Measurements of the absolute branching fractions of the Λ+c hadronic decays, JHEP 06 (2026), 269, arXiv:2601.01503 [hep-ex].DOI
- [42]M. Ablikim et al. (BESIII), Measurement of the absolute branching fraction of the inclusive decay Λ̄−c → n̄ + X, Phys. Rev. D 108, L031101 (2023), arXiv:2210.09561 [hep-ex].DOI
- [43]S. Godfrey and N. Isgur, Mesons in a Relativized Quark Model with Chromodynamics, Phys. Rev. D 32, 189 (1985).pubmed.ncbi.nlm.nih.gov/9955999
- [44]L. Micu, Decay rates of meson resonances in a quark model, Nucl. Phys. B 10, 521 (1969).DOI
- [45]A. Le Yaouanc, L. Oliver, O. Pène, and J.-C. Raynal, “Naive” quark-pair-creation model of strong-interaction vertices, Phys. Rev. D 8, 2223 (1973).


