Coupled-channel study of 4 ​ S 4S - 3 ​ D 3D mixing dynamics in ψ ⁡ ( 4220 ) \psi(4220) and ψ ⁡ ( 4380 ) \psi(4380) Zi-Long Man 1,2,3,4, Si-Qiang Luo 1,2,3,4, Zi-Yue Bai 1,2,3,4, Xiang Liu 1,2,3,4 Abstract Among charmoniumlike X ​ Y ​ Z XYZ states, the ψ ⁡ ( 4220 ) \psi(4220) and ψ ⁡ ( 4380 ) \psi(4380) states have emerged as key candidates for exploring the charmonium spectrum. In this work, we propose a 4 ​ S 4S - 3 ​ D 3D charmonium mixing scheme for the ψ ⁡ ( 4220 ) \psi(4220) and ψ ⁡ ( 4380 ) \psi(4380) , induced by coupled-channel effects. By constructing a coupled-channel model, we identify the dynamical mechanism responsible for the large mixing angle observed in previous studies, which cannot be explained by conventional potential models alone. Our analysis reveals that the D ​ D 1 DD_{1} channel significantly influences the lower state ( ψ ⁡ ( 4220 ) \psi(4220) ), while the D ∗ ​ D 1 D^{*}D_{1} channel primarily affects the higher state ( ψ ⁡ ( 4380 ) \psi(4380) ). Furthermore, we investigate the two-body Okubo-Zweig-Iizuka (OZI)-allowed strong decay behaviors of these states, providing insights into their total widths. This study not only supports the 4 ​ S 4S - 3 ​ D 3D mixing scheme but also offers a deeper understanding of the role of coupled channels in shaping the charmonium spectrum above 4 GeV. Our results align with experimental observations and provide a framework for interpreting future data on charmonium states. I Introduction The non-perturbative aspects of the strong interaction are closely linked to the study of hadron spectroscopy. Over the past two decades, a series of new hadronic states, including the charmoniumlike XYZ states, has sparked extensive discussions about exploring the exotic hadronic zoo and constructing the conventional hadron family [ 1 , 2 , 3 , 4 , 5 , 6 , 7 ] . Among these observed charmoniumlike XYZ states, the ones directly produced in e^{+}e^{-} annihilation are referred to as the Y states, which are distinguished by their masses. The challenge of understanding these Y states has come to be known as the “ Y Problem”, as highlighted in the BESIII White Paper [ 8 ] . The first Y state, the Y(4260) , was reported by the BaBar Collaboration in 2005, where an enhancement in the cross-section of e^{+}e^{-}\to\pi^{+}\pi^{-}J/\psi was observed. This structure, with a mass around 4.26 GeV and a width of approximately 90 MeV, was identified as a vector state [ 9 ] . Later, the CLEO [ 10 ] and Belle [ 11 ] Collaborations confirmed the existence of the Y(4260) . Several theoretical explanations have been proposed for its nature, including the hybrid state [ 12 , 13 ] , the compact tetraquark state [ 14 , 15 , 16 , 17 , 18 ] , and the molecular state [ 19 , 20 , 21 , 22 , 23 ] . However, these resonance assignments face significant experimental challenges. Notably, the Y(4260) structure is absent in open-charm decay channels [ 24 , 25 , 26 , 27 ] , and there is a dip structure observed around 4.26 GeV in R value measurements [ 28 , 29 , 30 , 31 , 32 , 33 ] . In 2011, the Lanzhou Group proposed a nonresonant explanation for the Y(4260) structure, attributing it to the interference between two well-established charmonia, \psi(4160) and \psi(4415) , along with background contributions from the continuum. They showed that this interference could effectively reproduce the asymmetric Y(4260) structure. This nonresonant explanation naturally accounts for the absence of evidence for the Y(4260) in open-charm decay channels and R value data [ 34 ] . Furthermore, this interference mechanism is considered universal, as it can also explain the Y(4360) structure, observed in the e^{+}e^{-}\to\psi(3686)\pi^{+}\pi^{-} process [ 35 ] , through a similar interference effect [ 36 , 37 ] . Clearly, this is not the final chapter of the story. In 2014, the Lanzhou Group observed a similarity in the mass gaps between the J/\psi and \Upsilon families [ 38 ] . Based on the mass gap between the \Upsilon(4S) and \Upsilon(3S) , and the charmonium state \psi(4040)\equiv\psi(3S) , they predicted the mass of the \psi(4S) to be 4264 MeV [ 38 ] . This differs from the \psi(4415) assignment as the \psi(4S) in the quenched potential model [ 39 ] , but is consistent with predictions from some unquenched potential models [ 40 , 41 , 42 ] , where the screening potential accounts for unquenched effects. Notably, they suggested that the predicted \psi(4S) , with a mass around 4264 MeV, would have a narrow width due to the node effect [ 38 ] . This prediction is yet to be tested experimentally. Meanwhile, experimentalists identified evidence for a narrow structure with a mass of 4216\pm 18 MeV and a width of 39\pm 32 MeV by fitting the experimental data of the e^{+}e^{-}\to\pi^{+}\pi^{-}h_{c} cross sections [ 43 ] . This narrow structure is considered a strong candidate for the predicted \psi(4S) [ 38 ] . In 2015, the BESIII Collaboration reported a narrow resonance in the e^{+}e^{-}\to\omega\chi_{c0} process, with a mass of 4230\pm 8\pm 6 MeV and a width of 38\pm 12\pm 2 MeV. The Lanzhou Group proposed the \psi(4S) to explain the enhancement observed in the e^{+}e^{-}\to\omega\chi_{c0} process [ 44 ] . Later, a combined fit to experimental data from e^{+}e^{-}\to\psi(2S)\pi^{+}\pi^{-} [ 45 ] , h_{c}\pi^{+}\pi^{-} [ 46 ] , and \omega\chi_{c0} [ 44 ] further refined the resonance parameters of the narrow structure around 4.2 GeV, which is consistent with the predicted \psi(4S) . Experimental precision must match theoretical precision, often leading to surprising results. In 2017, the BESIII Collaboration revealed that the Y(4260) structure consists of two substructures, the Y(4220) and Y(4320) , based on more precise data from e^{+}e^{-}\to J/\psi\pi^{+}\pi^{-} [ 47 ] . The first substructure, the Y(4220) , has a mass of 4222.0\pm 3.1\pm 1.4 MeV and a width of 44.1\pm 4.3\pm 2.0 MeV. Additionally, the narrow Y(4220) was also observed in the e^{+}e^{-}\to h_{c}\pi^{+}\pi^{-} [ 48 ] and e^{+}e^{-}\to\psi(3686)\pi^{+}\pi^{-} [ 49 ] processes. These combined theoretical and experimental efforts suggest that the narrow Y(4220) is a strong candidate for the predicted \psi(4S) . When carrying out a quantitative study of charmonium spectroscopy by a quenched potential model by introducing a screening potential, as did in Ref. [ 38 ] , the calculated mass of the \psi(4S) of charmonium is about 50 MeV larger than the measured value of the Y(4220) [ 50 ] . Figure 1: Schematic diagram illustrating the 4S - 3D mixing in the potential model and the coupled-channel mechanism. For the coupled-channel contribution, hadronic loops involving B and C mesons are considered in the calculation, which will be presented in Sec. III . Obviously, this discrepancy cannot be overlooked. Given that the 2S - 1D mixing scheme has been successfully applied to study the spectroscopic behaviors of the two low-lying charmonia \psi(3686) and \psi(3770) , it is reasonable to believe that a similar S - D mixing scheme could be relevant in understanding the nature of the Y(4220) . In this context, the Lanzhou Group introduced a 4S - 3D mixing scheme for the Y(4220) [ 50 ] , showing that the mass of the mixed state can be lowered to match that of the Y(4220) when the mixing angle is in the range of \pm(30^{\circ}\sim 36^{\circ}) . Additionally, a second mixed state, \psi(4380) , was predicted. In fact, the Y(4220) serves as a scaling point for constructing higher charmonium states above 4.2 GeV. Subsequently, a 5S - 4D mixing scheme for the \psi(4415) was proposed, predicting its partner, \psi(4500) . This approach results in an unquenched mass spectrum of vector charmonia above 4 GeV, comprising states such as \psi(4040) , \psi(4160) , \psi(4220) , \psi(4380) , \psi(4415) , and \psi(4500) . These states could be observed [ 50 , 51 , 52 , 53 ] in cross-section data from processes such as e^{+}e^{-}\to\psi(3686)\pi^{+}\pi^{-} [ 49 ] , e^{+}e^{-}\to\pi^{+}D^{0}D^{*-} [ 54 ] , e^{+}e^{-}\to J/\psi\eta [ 55 ] , and e^{+}e^{-}\to J/\psi K^{+}K^{-} [ 56 ] . While the 4S - 3D mixing scheme has been successfully applied to explain the Y(4220) phenomenologically [ 50 ] , it is essential to identify a mechanism that can explain the large mixing angle observed in the 4S - 3D charmonium mixing scheme. This angle is sufficiently large and cannot be reproduced by the tensor term in the potential model [ 57 ] (see the second term in Fig. 1 ). Therefore, the main motivation for this investigation is to uncover the underlying dynamical mechanism responsible for this large mixing angle, which will further elucidate the 4S - 3D mixing scheme for the \psi(4220) . In this work, we propose a coupled-channel induced charmonium mixing scheme for the \psi(4220) as illustrated in the schematic diagram in Fig. 1 . By constructing a coupled-channel model for the \psi(4220) and its partner, the \psi(4380) , we successfully reproduce the large mixing angle observed in Ref. [ 50 ] . However, in contrast to previous studies that employed an unquenched potential model with a screening potential [ 50 ] , our approach demonstrates how the coupled channels influence the masses of the bare states. Our calculations reveal that the DD_{1} channel plays a significant role in the lower state ( \psi(4220) ), while the nearby D^{*}D_{1} channel contributes primarily to the higher state ( \psi(4380) ). Additionally, we examine the two-body Okubo-Zweig-Iizuka (OZI)-allowed strong decay behaviors of the physical states \psi(4220) and \psi(4380) , which help determine the total width of these states. The paper is organized as follows: In Sec. II , we introduce a potential model to study the bare masses of \psi(4S) and \psi(3D) . Then, we employ a comprehensive coupled-channel model to discusses 4S-3D mixing scheme in III . The final section provides a brief summary. II Bare mass of the discussed charmonia from a potential and the corresponding mixing scheme induced by the tensor term We begin by calculating the bare masses of the discussed charmonia using a potential model, as these results serve as essential inputs for the subsequent analysis within the coupled-channel framework. Various versions of potential models have been employed to describe the hadron mass spectrum, most of which are derived from the Cornell potential [ 57 , 58 , 59 , 60 , 61 , 62 ] . Among these, the Godfrey-Isgur (GI) model stands out as it meets the precision requirements necessary for the study of the current hadron mass spectrum. The GI model, a typical quenched potential model, is a semi-relativistic framework with the Hamiltonian given by \displaystyle H_{0}=(\bm{p}^{2}+{m_{c}}^{2})^{\frac{1}{2}}+(\bm{p}^{2}+{m_{\bar{c}}}^{2})^{\frac{1}{2}}+\tilde{V}_{\rm{eff}}(\bm{p},\bm{r}), (1) where m_{c} and m_{\bar{c}} represent the masses of the c and \bar{c} quarks, respectively. The effective potential \tilde{V}_{\rm eff}(\bm{p},\bm{r}) incorporates a short-range one-gluon-exchange term \gamma^{\mu}\otimes\gamma_{\mu} and a long-range confinement term 1\otimes 1 . In the nonrelativistic limit, \tilde{V}_{\rm eff}(\bm{p},\bm{r}) reduces to: \displaystyle V_{\rm eff}(r)=H^{\rm conf}+H^{\rm hyp}+H^{\rm so}, (2) where the spin-independent term H^{\rm conf} combines the linear confinement potential and the Coulomb potential: \displaystyle H^{\rm conf}=br-\frac{4\alpha_{s}(r)}{3r}+c. (3) The color-hyperfine interaction H^{\text{hyp }} in Eq. ( 2 ) consists of two components: \displaystyle H^{\text{hyp }}=H_{S}+H_{T}, (4) where the spin-spin interaction H_{S} is given by \displaystyle H_{S}= \displaystyle\frac{32\pi\alpha_{S}(r)}{9m_{c}m_{\bar{c}}}\bm{S}_{c}\cdot\bm{S}_{\bar{c}}\delta^{3}(\bm{r}), (5) and the tensor term H_{T} is expressed as \displaystyle H_{T}=\frac{4\alpha_{s}}{3m_{c}m_{\bar{c}}r^{3}}\left(\frac{3(\bm{S}_{c}\cdot\bm{r})(\bm{S}_{\bar{c}}\cdot\bm{r})}{r^{2}}-\bm{S}_{c}\cdot\bm{S}_{\bar{c}}\right), (6) which may result in S - D mixing of charmonium. Here, \bm{S}_{c} and \bm{S}_{\bar{c}} denote the spins of the c -quark and \bar{c} -quark, respectively. The spin-orbit interaction H^{\rm so} in Eq. ( 2 ) is further divided into two terms \displaystyle H^{\rm so}=H^{\rm so(cm)}+H^{\rm so(tp)}, (7) where the color-magnetic term H^{\rm so(cm)} and the Thomas precession term H^{\rm so(tp)} are given by \displaystyle H^{\rm so(cm)}=\frac{4\alpha_{s}(r)}{3r^{3}}\left(\frac{\bm{S}_{c}}{m^{2}_{c}}+\frac{\bm{S}_{\bar{c}}}{m^{2}_{{\bar{c}}}}+\frac{\bm{S}_{c}+\bm{S}_{\bar{c}}}{m_{c}m_{\bar{c}}}\right)\cdot\bm{L}, (8) and \displaystyle H^{\rm so(tp)}=-\frac{1}{2r}\frac{\partial H^{\rm conf}}{\partial r}\left(\frac{\bm{S}_{c}}{m_{c}^{2}}+\frac{\bm{S}_{\bar{c}}}{m_{{\bar{c}}}^{2}}\right)\cdot\bm{L}, (9) respectively. Here, \bm{L} represents the relative orbital angular momentum between the quark and antiquark. A detailed description of these equations can be found in Ref. [ 57 ] . The parameters of the GI model are fitted from the masses of charmonia. In Ref. [ 63 ] , the parameter values were determined, and the masses of the low-lying charmonium states were well reproduced. Therefore, we adopt the same parameters in this study, which are listed in Table 1 . For further details on the fitted parameters, refer to Ref. [ 63 ] . Table 1: The parameters of the GI model. Parameters Values Parameters Values m_{q} 0.220 GeV b 0.175 \text{GeV}^{2} m_{c} 1.628 GeV c -0.245 GeV m_{s} 0.419 GeV \epsilon_{\text{cont}} -0.103 s 0.821 GeV \epsilon_{\text{so(v)}} -0.279 \sigma_{0} 2.33 GeV \epsilon_{\text{so(s)}} -0.3 \alpha_{s} 0.6 \epsilon_{\text{tens}} -0.114 \Lambda 0.2 GeV With the above preparation, we may utilize the GI model to give the bare masses of the involved charmonia \psi(4S) and \psi(3D) , which are 4433.0 MeV and 4491.3 MeV, respectively. The obtained mass of \psi(4S) is 211 MeV higher than that of the \psi(4220) , but close to that of the \psi(4415) . This is why the charmonium state \psi(4415) is identified as a \psi(4S) state within the quenched potential model [ 39 ] . However, with the advent of high-precision hadron spectroscopy in the unquenched model, we find that this scenario changes, as demonstrated in Ref. [ 50 ] , where the \psi(4S) state is lowered to 4274 MeV , closer to the observed \psi(4220) . In the next section, a detailed coupled-channel analysis further confirms this, offering a different perspective. When introducing the 4S - 3D mixing scheme [ 50 ] , the following relation holds: \displaystyle\begin{pmatrix}|\psi^{\prime}_{4S-3D}\rangle\\ |\psi^{\prime\prime}_{4S-3D}\rangle\end{pmatrix}=\begin{pmatrix}\cos\theta&\sin\theta\\ -\sin\theta&\cos\theta\end{pmatrix}\begin{pmatrix}|4^{3}S_{1}\rangle\\ |3^{3}D_{1}\rangle\end{pmatrix}, (10) where \theta represents the mixing angle. The lower state |\psi^{\prime}_{4S-3D}\rangle corresponds to the \psi(4220) , while the higher state |\psi^{\prime\prime}_{4S-3D}\rangle can be identified as the \psi(4380) [ 50 ] . Using the masses of \psi(4S) and \psi(3D) as inputs, the mixing angle can be determined to be \pm(30^{\circ}\sim 36^{\circ}) . This approach is phenomenological in nature. Starting from the GI model, the tensor term can contribute to the mixing of S - and D -wave charmonium states. A similar situation arises in the case of the deuteron, where the bound state properties are closely related to the S - D mixing induced by the tensor term in the effective potential that describes the proton-neutron interaction [ 64 ] . In the following, we examine the contribution of the tensor term in Eq. ( 6 ) to the mixing of S - and D -wave charmonium states by solving the equation: \displaystyle\begin{pmatrix}M_{S}^{0}&\langle\psi_{S}|H_{\text{T}}|\psi_{D}\rangle\\ \langle\psi_{D}|H_{\text{T}}|\psi_{S}\rangle&M_{D}^{0}\end{pmatrix}\begin{pmatrix}C_{S}\\ C_{D}\\ \end{pmatrix}=M\begin{pmatrix}C_{S}\\ C_{D}\\ \end{pmatrix}. (11) The diagonal terms, M_{S}^{0} and M_{D}^{0} , represent the bare masses of \psi(4S) and \psi(3D) , respectively. According to our calculations, the 4S - 3D mixing angle induced by the tensor term is only 0.5^{\circ} , which is insufficient to account for the large mixing reported in Ref. [ 50 ] . This mixing angle has negligible impact on both the masses and decay properties of \psi(4S) and \psi(3D) . This is why we need to find an alternative source for the 4S - 3D mixing scheme, which will be the main focus of the following section. III 4S - 3D mixing scheme induced by the coupled-channel mechanism Among the newly observed hadronic states, a universal phenomenon known as the “low mass puzzle” has been identified in the X(3872) [ 65 ] , D_{s0}(2317) [ 66 ] , D_{s1}(2460) [ 67 ] , and \Lambda_{c}(2940) [ 68 ] . Specifically, the measured masses of these states are consistently lower than the masses predicted by quenched models when these states are treated as conventional hadrons (i.e., q\bar{q} mesons or qqq baryons). To address this low mass puzzle, a direct approach has been to consider the possibility of exotic hadronic configurations for these states. This approach has opened a new window into the study of hadron spectroscopy over the past few decades [ 2 , 4 , 5 , 7 , 69 ] . In fact, the emergence of this low mass puzzle has highlighted the limitations of the quenched model, which was once the cornerstone of hadron spectroscopy. The influence of this type of model has persisted to the present day, but the observed discrepancies have prompted a reevaluation of its applicability. When unquenched effects are taken into account, the low mass puzzle observed in the X(3872) , D_{s0}(2317) , D_{s1}(2460) , and \Lambda_{c}(2940) can be significantly alleviated [ 70 , 71 , 72 , 42 , 73 , 74 ] . This demonstrates that unquenched effects cannot be ignored, and in fact, should be emphasized in modern hadron spectroscopy. It is for this reason that the current era of hadron spectroscopy is increasingly defined by the unquenched model, marking a significant shift in our understanding of hadronic states. In the study of hadron spectroscopy, there are different approaches to account for the unquenched effect. As demonstrated in previous studies [ 50 , 75 , 76 ] , one method involves introducing a screening potential within the framework of the potential model. Although this is a phenomenological approach, it effectively captures the realistic unquenched effect, as evidenced by recent applications in the study of higher-lying charmonium states [ 75 , 77 , 76 ] . Another approach is to perform a comprehensive coupled-channel analysis, which explicitly incorporates the coupling between the bare state and its allowed hadronic channels [ 78 , 79 ] . The latter method is employed in the present work. It is worth noting that the equivalence of these two treatments has been approximately established in Refs. [ 80 , 81 ] . Next, we perform a coupled-channel analysis for the 4^{3}S_{1} and 3^{3}D_{1} charmonium states and their mixing. This analysis is based on the following expression \displaystyle\begin{pmatrix}M^{0}_{S}&\langle\psi_{S}|H_{\text{T}}|\psi_{D}\rangle&\sum_{BC}\int\langle\psi_{S}|H_{I}|BC,\mathbf{P}\rangle d^{3}\mathbf{P}\\ \langle\psi_{D}|H_{\text{T}}|\psi_{S}\rangle&M^{0}_{D}&\sum_{BC}\int\langle\psi_{D}|H_{I}|BC,\mathbf{P}\rangle d^{3}\mathbf{P}\\ \sum_{BC}\int\langle BC,\mathbf{P}|H_{I}|\psi_{S}\rangle d^{3}\mathbf{P}&\sum_{BC}\int\langle BC,\mathbf{P}|H_{I}|\psi_{D}\rangle d^{3}\mathbf{P}&E_{BC}\\ \end{pmatrix}\begin{pmatrix}C_{S}\\ C_{D}\\ C_{BC}\end{pmatrix}=M\begin{pmatrix}C_{S}\\ C_{D}\\ C_{BC}\end{pmatrix}. (12) Here, E_{BC}=\sqrt{M_{B}^{2}+\mathbf{P}^{2}}+\sqrt{M_{C}^{2}+\mathbf{P}^{2}} represents the energy of the meson continuum states, assuming that the interaction between B and C is negligible. The off-diagonal term H_{I} describes the interactions between the meson-antimeson continuum states and the bare charmonium states. The mixing angle \theta can be determined by solving the reduced coupled-channel equation \displaystyle\begin{pmatrix}M_{S}^{0}+\Delta M_{S}(M)&\langle\psi_{S}|H_{\text{T}}|\psi_{D}\rangle+\Delta M_{SD}(M)\\ \langle\psi_{D}|H_{\text{T}}|\psi_{S}\rangle+\Delta M_{SD}(M)&M_{D}^{0}+\Delta M_{D}(M)\end{pmatrix}\begin{pmatrix}C_{S}\\ C_{D}\\ \end{pmatrix}=M\begin{pmatrix}C_{S}\\ C_{D}\\ \end{pmatrix}. (13) The forms of \Delta M_{S}(M) and \Delta M_{D}(M) are given by \displaystyle\Delta M_{S(D)}(M) \displaystyle= \displaystyle\text{Re}\sum_{BC}\int_{0}^{\infty}\frac{\mid\langle\psi_{S(D)}|H_{I}|BC,P\rangle\mid^{2}P^{2}dP}{M-E_{B}-E_{C}}, and \Delta M_{SD}(M) is expressed as \Delta M_{SD}(M)=\text{Re}\sum_{BC}\int_{0}^{\infty}\frac{\langle\psi_{S}|H_{I}|BC,P\rangle\langle BC,P\mid H_{I}\mid\psi_{D}\rangle}{(M-E_{B}-E_{C})}P^{2}dP. (15) In principle, all possible open-charm meson loops should be included in the self-energy function. However, a challenge arises in calculating the infinite number of hadron loops in \Delta M_{S(D)}(M) and \Delta M_{SD}(M) . This issue was addressed in Ref. [ 82 ] , where the authors proposed the once-subtracted dispersion relation, which effectively limits the number of loops and resolves this problem. Here, we choose same number of hadronic loops to calculate mass shifts of mixing charmomium states, and these allowed loops with mass thresholds below the bare masses of \psi(4S) and \psi(3D) . We now employ the once-subtracted method to rewrite \Delta M_{S(D)}(M) and \Delta M_{SD}(M) as \displaystyle\Delta M_{S(D)}(M) \displaystyle= \displaystyle\text{Re}\sum_{BC}\int_{0}^{\infty}\frac{(M_{J/\psi}-M)|\langle\psi_{{S(D)}}^{0}|H_{I}|BC,P\rangle|^{2}P^{2}dP}{(M-E_{B}-E_{C})(M_{J/\psi}-E_{B}-E_{C})}, (16) and \Delta M_{SD}(M)=\text{Re}\sum_{BC}\int_{0}^{\infty}\frac{(M_{J/\psi}-M)\langle\psi_{S}\mid H_{I}\mid BC,P\rangle\langle BC,P\mid H_{I}\mid\psi_{D}\rangle}{(M-E_{B}-E_{C})(M_{J/\psi}-E_{B}-E_{C})}P^{2}dP, (17) respectively. Here, M_{J/\psi} denotes the subtraction point, which we take as the mass of the J/\psi meson in the charmonium system. Studies using this method have successfully reproduced reasonable meson masses, as demonstrated in Refs. [ 63 , 81 , 83 , 84 ] . To describe the coupling between the bare state and the hadronic channel composed of B and C , we express the matrix element as \langle\psi|H_{I}|BC\rangle=\mathcal{M}_{JL}(A\rightarrow B+C) . To calculate the amplitude \mathcal{M}_{JL}(A\rightarrow B+C) , we employ the quark pair creation (QPC) model [ 85 , 86 ] . The corresponding Hamiltonian is given by \displaystyle H_{I} \displaystyle= \displaystyle-3\gamma\sum_{m}\left<1m1-m\mid 00\right>\int\limits d\mathbf{p}_{3}d\mathbf{p}_{4}\delta^{3}(\mathbf{p}_{3}+\mathbf{p}_{4}) (18) \displaystyle\times{\cal Y}_{1}^{m}\left(\frac{\mathbf{p}_{3}-\mathbf{p}_{4}}{2}\right)\chi_{1-m}^{34}\varphi_{0}^{34}\omega_{0}^{34}b_{3}^{\dagger}(\mathbf{p}_{3})d_{4}^{\dagger}(\mathbf{p}_{4}), where the dimensionless parameter \gamma represents the strength of the q\bar{q} pair creation from the vacuum. In our calculations, we adopt \gamma=0.44 as the strength parameter for the quark pair creation. This value is chosen because it reproduces the decay widths well of the two P -wave charmonium states, \psi(3915)\equiv\psi(2^{3}P_{0}) and \psi(3930)\equiv\psi(2^{3}P_{2}) . If a strange quark pair is created, the effective strength is adjusted to \gamma_{s}=\frac{m_{n}}{m_{s}}\gamma , where m_{n} ( n=u,d ) and m_{s} denote the constituent masses of the nonstrange and strange quarks, respectively. The momenta of the created quark q and antiquark \bar{q} are denoted by \mathbf{p}_{3} and \mathbf{p}_{4} , respectively. The solid harmonic function is given by {\cal Y}_{1}^{m}(\mathbf{p})=|\mathbf{p}|Y_{1}^{m}(\theta,\phi) , and \chi_{1-m}^{34} , \varphi_{0}^{34} , and \omega_{0}^{34} represent the spin, flavor, and color wave functions of the created q\bar{q} pair, respectively. Using the amplitude \mathcal{M}_{JL}(A\rightarrow B+C) , the OZI-allowed two-body strong decay width for the charmonium state can be calculated as \displaystyle\Gamma_{\text{total}}=\sum_{BC}\frac{2\pi PE_{B}E_{C}}{M}\sum_{JL}\mid\mathcal{M}_{JL}(A\rightarrow B+C)\mid^{2}, (19) where P is the momentum of the final-state mesons, and M is the mass of the physical state under consideration. In our calculations, the spatial wave functions of the involved mesons serve as crucial inputs. These wave functions are obtained by solving a potential model, as described in Sec. II . Specifically, we adopt the following form for depicting the obtained numerical spatial wave function \displaystyle\Psi_{nLM_{L}}(\mathbf{P}) \displaystyle=\sum\limits_{n_{1}}^{n_{\text{max}}}C_{n}R_{nL}(P)Y_{LM_{L}}(\Omega_{\mathbf{P}}), (20) where R_{nL}(P) represents the radial wave function in momentum space, and n_{\text{max}} is the maximum number of basis states. To ensure an accurate representation of the numerical spatial wave functions, we set n_{\text{max}}=20 . This approach provides the necessary framework to determine the OZI-allowed two-body strong decay widths of charmonium states. III.1 The properties of \psi(4S) and \psi(3D) within coupled-channel analysis Figure 2: The M dependence of function M_{0} and \Delta M(M) for \psi(4S) and \psi(3D) , respectively. There are three (two) discontinuities in the line shape of \psi(4S) ( \psi(3D) ) corresponding to the DD_{1}(2430)^{0} , D^{*}D_{1}(2430)^{0} and D^{*}D_{0}(2300) ( DD_{1}(2420) and D^{*}D_{1}(2420) ) channels, respectively. Table 2: The mass shift \Delta M_{i} from various coupled channels, the corresponding ratio \Delta M_{i}/\sum_{i}\Delta M_{i} , the partial decay widths, and branching ratios for \psi(4S) and \psi(3D) within the coupled-channel framework. In the last two rows, we provide the average value from PDG [ 87 ] for \psi(4220) and the calculated resonance parameters for \psi(4S) . The masses, mass shifts, and decay widths are in units of MeV. Channels \Delta{M}_{i} \Delta{M}_{i}/\sum_{i}\Delta{M}_{i} \Gamma_{i} \Gamma_{i}/\sum_{i}\Gamma_{i} \Delta{M}_{i} \Delta{M}_{i}/\sum_{i}\Delta{M}_{i} \Gamma_{i} \Gamma_{i}/\sum_{i}\Gamma_{i} DD - 5.8 3.8% 0.3 0.6% - 10.7 7.0% 1.8 7.8% DD^{*} - 12.3 8.0% 9.7 18.3% - 3.1 2.0% 3.4 14.4% D_{s}D_{s} - 1.0 0.7% 0.2 0.3% - 1.1 0.7% 0.4 1.9% D^{*}D^{*} - 35.1 22.8% 41.0 77.8% - 36.9 24.2% 8.0 34.0% D_{s}D_{s}^{*} - 4.4 2.8% 0.4 0.8% - 1.3 0.9% 0.9 3.6% DD_{0}^{*}(2300) - - - - - - - - D_{s}^{*}D_{s}^{*} - 7.1 4.6% 1.1 2.1% - 10.1 6.6% 0.7 3.1% DD_{1}(2430)^{0} - 16.1 10.4% - - - 16.4 10.7% 2.4 10.2% DD_{1}(2420) - 8.7 5.7% - - - 14.5 9.5% 5.7 24.0% D_{s}D_{s0}(2317) - - - - - - - - DD_{2}^{*}(2460) - 8.5 5.5% - - - 5.1 3.3% 0.2 0.9% D^{*}D_{0}^{*}(2300) - 12.1 7.8% - - - 8.4 5.5% - - DD_{0}(2550)^{0} - 2.2 1.4% - - - 6.2 4.1% - - D^{*}D_{1}(2430)^{0} - 22.8 14.8% - - - 21.4 14.0% - - D_{s}D_{s1}(2460) - 2.3 1.5% - - - 3.3 2.2% - - D_{s}^{*}D_{s0}(2317) - 2.2 1.4% - - - 1.1 0.7% - - D^{*}D_{1}(2420) - 13.2 8.6% - - - 13.0 8.5% - - Total - 153.9 100% 52.7 100% - 152.8 100% 23.6 100% M=4279.1 M=4338.5 M_{\psi(4220)}=4222.1\pm 2.3 [ 87 ] \Gamma_{\psi(4220)}=49\pm 7 [ 87 ] Before discussing the 4S - 3D mixing scheme induced by the coupled-channel effect, it is essential to quantitatively investigate the impact of the coupled channels on the bare charmonium states \psi(4S) and \psi(3D) . This analysis will help us better understand the fundamental features of the mass shifts for \psi(4S) and \psi(3D) . Using the parameters outlined above, the numerical results for the masses, mass shifts, and decay widths of the \psi(4S) and \psi(3D) states are presented in Table 2 . The mass shifts \Delta M and the corresponding function M_{0}-M as a function of M for the \psi(4S) and \psi(3D) states are shown in Fig. 2 . To quantitatively study the coupled-channel corrections to the bare \psi(4S) and \psi(3D) states, Eq. ( 13 ) can be decomposed into two separate equations, one for \psi(4S) and the other for \psi(3D) . For \psi(4S) , the coupled-channel equation is expressed as: \displaystyle M-M_{S}^{0}-\Delta M_{S}(M)=0. (21) By solving this equation, the mass of \psi(4S) is estimated to be 4279.1 MeV, which is consistent with calculations of \psi(4S) within screened potential models [ 50 , 42 ] , but larger than the PDG value of 4222.1\pm 2.3 MeV [ 87 ] . From Table. 2 , our results indicate that the total mass shift of the \psi(4S) state is \Delta M=-153.9 MeV. The \psi(4S) state strongly couples with the DD^{*} , D^{*}D^{*} , DD_{1}(2430)^{0} , D^{*}D_{1}(2430)^{0} , D^{*}D_{0}(2300) , and D^{*}D_{1}(2420) channels. 1 1 1 Since the bare masses of \psi(4S) and \psi(3D) are approximately 4.5 GeV, the hadron loops involving DD_{1}(2430)^{0} , DD_{1}(2420) , D^{*}D_{1}(2430)^{0} , and D^{*}D_{1}(2420) can contribute to the mass shifts of both \psi(4S) and \psi(3D) . Here, the D_{1}(2420) and D_{1}(2430)^{0} share the same quantum numbers J^{P}=1^{+} . In the framework of heavy quark symmetry, the J^{P}=1^{+} state can be expressed as a linear combination of the 1^{1}P_{1} and 1^{3}P_{1} states. The mixing scheme is described as follows: \displaystyle\begin{pmatrix}\mid D_{1}(2430)\rangle\\ \mid D_{1}(2420)\rangle\end{pmatrix}=\begin{pmatrix}\text{cos}\phi&\text{sin}\phi\\ -\text{sin}\phi&\text{cos}\phi\end{pmatrix}\begin{pmatrix}\mid D(1^{1}P_{1})\rangle\\ \mid D(1^{3}P_{1})\rangle\end{pmatrix}, (22) where the mixing angle \phi=-54.7^{\circ} can be determined in the heavy quark limit [ 88 ] . For convenience, we use DD to denote the sum of the D\bar{D} and \bar{D}D decay channels in the present work, and so forth. Here, the mass shift, mass and width are in units of MeV. Based on heavy quark spin symmetry (HQSS), the D and D^{*} mesons are degenerate in the heavy quark limit. This explains why the mass shift of \psi(4S) receives similar contributions from the DD_{1}(2430)^{0} and D^{*}D_{1}(2430)^{0} channels. Similarly, the D_{1}(2420) and D^{*}_{0}(2300) mesons are also degenerate in the heavy quark limit. Therefore, the D^{*}D_{1}(2420) and D^{*}D^{*}_{0}(2300) channels also provide significant contributions to the mass shifts of the \psi(4S) state. When \psi(4S) couples to the DD_{1}(2430)^{0} , D^{*}D_{1}(2430)^{0} , and D^{*}D_{0}(2300) channels through S -wave interactions, the line shape of \Delta M exhibits three discontinuities. Additionally, the DD^{*} and D^{*}D^{*} channels also have significant branching ratios for mass shifts involving \psi(4S) , with their coupling primarily occurring via P -wave interactions. However, the coupling of \psi(4S) with the DD channel is relatively weak. The D_{1} state includes the D_{1}(2420) with a narrow width of 31.3\pm 1.9 MeV and the D_{1}(2430)^{0} with a broad width of 314\pm 29 MeV, as both D_{1}(2420) and D_{1}(2430)^{0} can decay into the D^{*}\pi channel. Within the framework of HQSS, the D_{1}(2420) decays to the D^{*}\pi channel mainly via the D -wave, while the D_{1}(2430)^{0} decays to the D^{*}\pi channel primarily through the S -wave [ 89 ] . Consequently, the coupling of \psi(4S) with the DD_{1}(2420) channel may be suppressed. On the other hand, the D_{1}(2430)^{0} and D_{1}(2420) are treated as mixed states in this study. Therefore, the coupling of \psi(4S) with the DD_{1}(2430)^{0} channel is stronger than that with the DD_{1}(2420) channel. The predicted strong decay width of \psi(4S) is 52.7 MeV, with dominant decays into the DD^{*} and D^{*}D^{*} channels. The partial width ratio is predicted to be \Gamma(DD^{*}):\Gamma(D^{*}D^{*})=0.24 . Contributions to the OZI-allowed two-body strong decay width from the DD , D_{s}D_{s} , D_{s}D_{s}^{*} , and D_{s}^{*}D_{s}^{*} channels are relatively suppressed. From the above analysis, the predicted mass of \psi(4S) is approximately 60 MeV larger than the measured mass of \psi(4220) , although the estimated two-body OZI-allowed strong decay width of \psi(4S) is close to the measured width of \psi(4220) . Therefore, further study is needed to better understand the nature of \psi(4220) . By decomposing Eq. ( 13 ), the coupled-channel equation for \psi(3D) can be written as \displaystyle M-M_{D}^{0}-\Delta M_{D}(M)=0. (23) The mass of \psi(3D) is predicted to be 4338.5 MeV based on the calculation in Eq. ( 23 ). Our result is slightly smaller than the average experimental value of 4373\pm 7 MeV for Y(4360) [ 87 ] , but it is consistent with the predictions of screened potential models [ 50 , 42 ] . From Fig. 2 , we find that the total mass shift of \psi(3D) is estimated to be -152.8 MeV. The D^{*}D^{*} , DD_{1}(2430)^{0} , DD_{1}(2420) , D^{*}D_{1}(2430)^{0} and D^{*}D_{1}(2420) channels provide the dominant contributions to the mass shift of \psi(3D) . Among the coupled channels, D^{*}D^{*} contributes the largest mass shift for \psi(3D) . In the present work, the estimated width of \psi(3D) is 23.6 MeV, which is also smaller than the measured width of 118\pm 12 MeV for Y(4360) [ 87 ] . A similar conclusion is reached in the modified GI model of \psi(3D) [ 50 ] . Therefore, based on its decay properties, the \psi(3D) charmonium state may be a narrow structure. III.2 The 4S - 3D mixing scheme induced by coupled-channel effect According to our results, the calculated mass of the \psi(4S) state is higher than the measured mass of the \psi(4220) . To resolve this mass discrepancy, we employ a comprehensive coupled-channel scheme, where the allowed hadronic loops composed of charmed mesons act as a bridge connecting the \psi(4S) and \psi(3D) states. This mechanism provides a source for the 4S - 3D mixing. In Fig. 3 , we illustrate the dependence of \Delta M_{SD} on M . The result demonstrates that \Delta M_{SD} is comparable in magnitude to the calculated values of \Delta M_{S} and \Delta M_{D} . This observation strongly suggests the possibility of significant 4S - 3D mixing. Our study explicitly confirms this behavior, highlighting the non-negligible contribution of such mixing effects. The masses of the mixed states {\psi}_{4S-3D}^{\prime} and {\psi}_{4S-3D}^{\prime\prime} , as well as the 4S - 3D mixing angle, are determined by solving the reexpressed equation of Eq. ( 13 ): \displaystyle\text{det}\begin{vmatrix}M_{S}^{0}+\Delta M_{S}(M)-M&\langle\psi_{S}|H_{\text{T}}|\psi_{D}\rangle+\Delta M_{SD}(M)\\ \langle\psi_{D}|H_{\text{T}}|\psi_{S}\rangle+\Delta M_{SD}(M)&M_{D}^{0}+\Delta M_{D}(M)-M\end{vmatrix}=0. (24) In Table 3 , we present the mass of the mixed state {\psi}_{4S-3D}^{\prime} , determined to be 4235.9 MeV. This result is consistent with the PDG mass value of 4222.1\pm 2.3 MeV [ 87 ] for the Y(4220) . Using Eq. ( 13 ), the mixing angle between \psi(4S) and \psi(3D) is found to be \theta=35^{\circ}. This value accounts for the large 4S - 3D mixing in the \psi(4220) , as directly obtained from experimental data fitting in Ref. [ 50 ] . Figure 3: The dependence of \Delta M_{SD} on M . In addition to the mass, we investigate the decay properties of {\psi}_{4S-3D}^{\prime} . The estimated total decay width for the OZI-allowed two-body strong decay of {\psi}_{4S-3D}^{\prime} is 32.0 MeV, which aligns with the measured width of the \psi(4220) . We identify D^{*}D^{*} as the dominant decay mode, with a branching fraction as high as 91.4\% . These findings demonstrate that a comprehensive coupled-channel approach can successfully explain a large 4S - 3D mixing angle and accurately reproduce both the mass and decay width of the \psi(4220) . Thus, interpreting the \psi(4220) as a mixed state {\psi}_{4S-3D}^{\prime} is well justified. The estimated mass of another mixed state {\psi}_{4S-3D}^{\prime\prime} is 4387.1 MeV, consistent with the conclusion from a screening potential model [ 50 ] . As shown in Table 3 , the mixing angle \theta=35^{\circ} results in a predicted total width of two-body OZI-allowed strong decays to be 44.1 MeV for {\psi}_{4S-3D}^{\prime\prime} , which indicates that it is possible to find this state in experiment. In fact, several former theoretical analyses of the processes e^{+}e^{-}\to\psi(2S)\pi^{+}\pi^{-} [ 50 ] , e^{+}e^{-}\to K^{+}K^{-}J/\psi [ 51 ] , e^{+}e^{-}\to\pi^{+}D^{0}D^{*-} [ 52 ] , and e^{+}e^{-}\to\eta J/\psi [ 53 ] provide strong evidence for the existence of this predicted {\psi}_{4S-3D}^{\prime\prime} , which may be referred to as the \psi(4380) . In Fig. 4 , we compare the resonance parameters of the mixed state {\psi}_{4S-3D}^{\prime\prime} with those of the reported vector charmonium-like structures around 4.36 GeV in different processes. Currently, the measured resonance parameters for vector charmonium-like structures in this energy region exhibit significant discrepancies. We find that the predicted resonance parameters of the mixed state {\psi}_{4S-3D}^{\prime\prime} are consistent with the results from the e^{+}e^{-}\to\chi_{c2}\gamma process, where the observed vector charmonium-like structure has a mass of 4371.7\pm 7.5\pm 1.8 MeV and a width of 51.1\pm 17.6\pm 1.9 MeV [ 90 ] . However, more precise experimental data are required to conclusively determine its properties. Our study indicates that the 4S - 3D mixing scheme, induced by coupled-channel effects, can lead to significant differences in the dominant decay modes of {\psi}_{4S-3D}^{\prime\prime} compared to those of a pure \psi(3D) state. For the mixed state {\psi}_{4S-3D}^{\prime\prime} , the DD^{*}_{2}(2460) channel plays a crucial role. Among the partial decay widths of the two-body OZI-allowed decays, the largest partial width originates from the DD^{*}_{2}(2460) channel. Furthermore, our results show that the D^{*}D^{*} , DD_{1}(2430)^{0} , DD_{1}(2420) , and DD^{*}_{2}(2460) channels are the dominant strong decay modes for {\psi}_{4S-3D}^{\prime\prime} , with the ratio \Gamma(D^{*}D^{*}):\Gamma(DD_{1}(2430)^{0}):\Gamma(DD_{1}(2420)):\Gamma(DD^{*}_{2}(2460))=1.4:1:1.4:2.9 . These findings are also consistent with those reported in Ref. [ 50 ] . Therefore, we encourage future experiments to search for {\psi}_{4S-3D}^{\prime\prime} through the open-charm decay channels D^{*}D^{*} , DD_{1}(2430)^{0} , DD_{1}(2420) , and DD^{*}_{2}(2460) and their subordinate decays. As shown in Tables II and III, if the \psi(4220) corresponds to the \psi^{\prime}_{4S-3D} state in the 4S-3D mixing scheme, its decay width is narrower than that of \psi(4S) because of smaller phase space. The dominant decay channel of \psi^{\prime}_{4S-3D} remains D^{*}D^{*} , with a branching ratio of 91.4%, while the other decay modes such as DD^{*} are significantly suppressed. If we treat \psi(4380) as the higher mass state \psi^{\prime\prime}_{4S-3D} , its total decay width is significantly enhanced compared to that of \psi(3D) state, primarily due to the larger phase space. Notably, the largest partial width of \psi(4380) is DD_{2}^{*}(2460) with a branching fraction reaching up to 37.3\% , in sharp contrast to the prediction for the \psi(3D) , which has only about 0.9\% . This difference arises from the modified wave function induced by the S-D mixing scheme. Additionally, \psi(4380) also has dominant decay rates into the D^{*}D^{*} , DD_{1}(2430)^{0} , and DD_{1}(2420) channels. The predicted dominant decay channels can provide critical information to test the assignment of \psi(4220) and support future experimental searches for \psi(4380) . Table 3: Two-body strong decay widths of {\psi}_{4S-3D}^{\prime} and {\psi}_{4S-3D}^{\prime\prime} with mixing angle 35^{\circ} . The decay widths are in units of MeV. {\psi}_{4S-3D}^{\prime} {\psi}_{4S-3D}^{\prime\prime} m=4235.9 MeV m=4387.1 MeV Channels \Gamma_{i} \Gamma_{i}/\sum_{i}\Gamma_{i} \Gamma_{{i}} \Gamma_{i}/\sum_{i}\Gamma_{i} DD 0.2 0.5% 0.5 1.1% DD^{*} 1.7 5.4% 0.7 1.7% D_{s}D_{s} 0.6 1.9% 0.1 0.2% D^{*}D^{*} 29.2 91.4% 8.0 18.1% D_{s}D_{s}^{*} 0.0 0.0% 2.5 5.7% DD_{0}^{*}(2300) - - - - D_{s}^{*}D_{s}^{*} 0.3 0.8% 0.8 1.9% DD_{1}(2430)^{0} - - 5.7 13.0% DD_{1}(2420) - - 7.7 17.4% D_{s}D_{s0}(2317) - - - - DD_{2}^{*}(2460) - - 16.5 37.3% D^{*}D_{0}^{*}(2300) - - 1.6 3.7% Total 32.0 100% 44.1 100% \Gamma_{\psi(4220)}=49\pm 7 [ 87 ] Figure 4: The red pentagram denotes the calculated resonance parameters of {\psi}_{4S-3D}^{\prime\prime} in our coupled-channel model. The other symbols represent the measured resonance parameters of vector charmonium-like structures around 4.36 GeV from various processes: e^{+}e^{-}\to J/\psi\pi^{+}\pi^{-} [ 47 ] , e^{+}e^{-}\to\psi(3686)\pi^{+}\pi^{-} [ 49 ] , e^{+}e^{-}\to h_{c}\pi^{+}\pi^{-} [ 48 ] , e^{+}e^{-}\to\psi(3686)\pi^{+}\pi^{-} [ 91 ] , e^{+}e^{-}\to\chi_{c2}\gamma [ 90 ] , e^{+}e^{-}\to J/\psi\pi^{+}\pi^{-} [ 92 ] , e^{+}e^{-}\to D^{+}D^{-}\pi^{+}\pi^{-} [ 93 ] , and e^{+}e^{-}\to J/\psi\eta [ 55 ] . IV Summary In this work, we investigate the nature of the \psi(4220) and its partner state \psi(4380) within the framework of a coupled-channel model, focusing on the 4S - 3D mixing scheme. By incorporating hadronic loop effects, we quantitatively analyze the mass shifts and decay properties of the \psi(4S) and \psi(3D) states. Our calculations reveal that the physical mass of the \psi(4S) state is lowered to 4279.1 MeV due to coupled-channel corrections, while the \psi(3D) state is predicted to have a mass of 4338.5 MeV. These results are consistent with predictions from screened potential models [ 42 , 50 ] but highlight discrepancies with experimental measurements [ 87 ] , particularly for the \psi(4220) . To resolve the mass discrepancy, we propose a comprehensive 4S - 3D mixing scheme induced by coupled-channel effects. The mixed state {\psi}_{4S-3D}^{\prime} , identified as the \psi(4220) , is found to have a mass of 4235.9 MeV and a total decay width of 32.0 MeV, in good agreement with experimental data [ 87 ] . The mixing angle between \psi(4S) and \psi(3D) is determined to be \theta=35^{\circ} , which confirms the large 4S - 3D mixing suggested in Ref. [ 50 ] . The dominant decay mode of {\psi}_{4S-3D}^{\prime} is D^{*}D^{*} , with a branching fraction of 91.4\% . We also predict the existence of a second mixed state, {\psi}_{4S-3D}^{\prime\prime} , with a mass of 4387.1 MeV and a total decay width of 44.1 MeV. This state is suggested to correspond to the narrow \psi(4380) , which can be searched for in open-charm decay channels such as D^{*}D^{*} , DD_{1}(2430)^{0} , DD_{1}(2420) , and DD^{*}_{2}(2460) . Our results demonstrate that the large 4S - 3D mixing angle is primarily driven by coupled-channel effects, providing a robust explanation for the observed properties of the \psi(4220) and \psi(4380) . This study underscores the importance of coupled-channel mechanisms in understanding the charmonium spectrum above 4 GeV and offers a new perspective on the role of meson loops in shaping the masses and decay properties of higher-lying charmonium states. Future experimental searches for the predicted \psi(4380) state in the suggested decay channels will provide crucial tests of our theoretical framework. Note added : In preparing the present work, we observed that a mixing scheme induced by the coupled-channel effect was applied to study the bottomonium system in Ref. [ 94 ] recently.