betaFeedback

WorksPhysics and Astronomy

χc1(4010)χ_{c1}(4010) as a D∗±D∓D^{*\pm}D^{\mp} Molecular State: Mass, Decay Constant, and Width

Abstract

The charmonium-like state χc1(4010)χ_{c1}(4010), recently observed by the LHCb Collaboration in B+→D∗±D∓K+B^+ \to D^{*\pm}D^{\mp}K^+ decays, is investigated as an axial-vector D∗±D∓D^{*\pm}D^{\mp} molecular state within the QCD sum rule approach. A two-point sum rule incorporating quark, gluon, and mixed vacuum condensates up to dimension 1010 yields the mass and decay constant, mχc1(4010)=(4005.89±47.15) MeVm_{χ_{c1}(4010)} = (4005.89 \pm 47.15)~\text{MeV} and fχc1(4010)=(2.87±0.55)×10−3 GeV4f_{χ_{c1}(4010)} = (2.87 \pm 0.55)\times 10^{-3}~\text{GeV}^{4}, in agreement with the LHCb measurement. Since this mass lies above the D∗±D∓D^{*\pm}D^{\mp} threshold, the strong coupling constants at the χc1(4010)D∗±D∓χ_{c1}(4010)D^{*\pm}D^{\mp} vertices are extracted from three-point sum rules and used to compute the partial widths of the D∗−D+D^{*-}D^{+} and D∗+D−D^{*+}D^{-} channels. The resulting total width, Γχc1(4010)=(57.64±9.64) MeVΓ_{χ_{c1}(4010)} = (57.64 \pm 9.64)~\text{MeV}, is compatible with experiment within uncertainties. These results support a D∗±D∓D^{*\pm}D^{\mp} molecular interpretation of χc1(4010)χ_{c1}(4010) and offer sum-rule predictions for comparison with future measurements at BESIII, LHCb, and Belle~II.

Introduction

States with quantum numbers or mass spectra that fall outside the standard classification of mesons (qqˉq\bar q) and baryons (qqqqqq), including multiquark configurations and gluon hybrids, have become a central topic in hadron spectroscopy. In the charmonium (ccˉc\bar c) and bottomonium (bbˉb\bar b) regions, several narrow resonances fail to match constituent quark model predictions for conventional heavy mesons [1, 2, 3, 4, 5, 6, 7, 8]. Their structure remains unresolved and is an open problem in nonperturbative QCD. Proposed interpretations include multiquark systems, hadronic molecules, and hybrid states [9, 10, 11]. Several theoretical models have also been extended to directly accommodate these states.

The charmonium-like sector with JPC=1++J^{PC}=1^{++} illustrates this problem clearly. The χc1(3872)\chi_{c1}(3872), usually treated as the first observed state in this channel, has been debated since its discovery [12, 13, 14, 15]. Its mass lies almost exactly at the D0Dˉ∗0D^{0}\bar D^{*0} threshold, complicating a simple χc1(2P)\chi_{c1}(2P) assignment, and its decay width is far narrower than decay models predict for that state. The higher mass χc1(4140)\chi_{c1}(4140) shows the opposite pattern: both its mass and decay width are higher than a χc1(2P)\chi_{c1}(2P) charmonium assignment predicts [16, 17, 18, 19, 20].

In 2024, an exotic candidate with JPC=1++J^{PC}=1^{++}, denoted χc1(4010)\chi_{c1}(4010), was identified by the LHCb collaboration via analysis of the D∗±D∓D^{*\pm}D^{\mp} invariant mass distribution in B+→D∗±D∓K+B^{+}\to D^{*\pm}D^{\mp}K^{+} decays [21]. The measured resonance parameters are

mχc1(4010)=(4012.5−3.9−3.7+3.6+4.1) MeV,m_{\chi_{c1}(4010)}=(4012.5^{+3.6+4.1}_{-3.9-3.7})\ \mathrm{MeV},
Γχc1(4010)=(62.7−6.4−6.6+7.0+6.4) MeV.(1)\Gamma_{\chi_{c1}(4010)}=(62.7^{+7.0+6.4}_{-6.4-6.6})\ \mathrm{MeV}. \tag*{(1)}

Ding et al. included χc1(4010)\chi_{c1}(4010) as an extra resonance in their amplitude analysis of B+→D∗+D−K+B^{+}\to D^{*+}D^{-}K^{+} [22], motivated by the fact that molecular contributions from χc1(3872)\chi_{c1}(3872) and Tcsˉ0∗(2870)0T^{*}_{c\bar s0}(2870)^{0} alone could not reproduce the LHCb spectrum. Adding χc1(4010)\chi_{c1}(4010) accounts for the excess near 4010 MeV and brings the amplitude model into agreement with the data.

The analysis uses CC-parity relations to constrain charmonium(-like) contributions across the two charge-conjugate decay channels. Four such states decaying to D∗±D∓D^{*\pm}D^{\mp} are observed: ηc(3945)\eta_{c}(3945) (0−+0^{-+}), hc(4000)h_{c}(4000) (1+−1^{+-}), χc1(4010)\chi_{c1}(4010) (1++1^{++}), and hc(4300)h_{c}(4300) (1+−1^{+-}). The amplitude model for B+→D∗+D−K+B^{+}\to D^{*+}D^{-}K^{+} also includes the exotic resonances Tcsˉ0∗0(2870)0T^{*0}_{c\bar s0}(2870)^{0} and Tcsˉ1∗0(2900)T^{*0}_{c\bar s1}(2900), first seen in B+→D+D−K+B^{+}\to D^{+}D^{-}K^{+} decays, and confirms their presence in this channel as well. If the Tcsˉ0,1∗0T^{*0}_{c\bar s0,1} states are eventually confirmed, this cross-channel difference could help clarify how they are produced [21]. These results establish the D∗±D∓D^{*\pm}D^{\mp} threshold region as populated by multiple resonances, motivating a dedicated structural study of χc1(4010)\chi_{c1}(4010) and others.

The 1++1^{++} charmonium-like sector already includes χc1(4140)\chi_{c1}(4140), χc1(4274)\chi_{c1}(4274), and χc1(4685)\chi_{c1}(4685) [23, 24], while χc1(3872)\chi_{c1}(3872) and χc2(3930)\chi_{c2}(3930) lie considerably lower [25]. In mass, χc1(4010)\chi_{c1}(4010) [21] sits between these two sets: above both of the latter and below χc1(4140)\chi_{c1}(4140) and χc1(4274)\chi_{c1}(4274). Its quantum numbers, JPC=1++J^{PC}=1^{++}, are what a conventional χc1\chi_{c1} charmonium state would have, but its mass does not match any established χc1\chi_{c1} state predicted by standard charmonium models [21]. Such considerations have opened the door to several alternative interpretations, including a compact tetraquark [26], a meson-meson molecule, or even a hybrid configuration that mixes conventional charmonium with exotic components [25]. QCD sum rules (QCDSR) have similarly been applied to compute the mass spectrum of the heavy-quark spin symmetry partners of X(3872)X(3872) within a molecular framework [27].

The χc1(4010)\chi_{c1}(4010) resonance plays a key role in coupled-channel analyses of the χc1(2P)\chi_{c1}(2P) sector. Its closeness to open-charm thresholds makes it sensitive to both the underlying ccˉc\bar c core and hadronic molecular components, and it serves as a reference point for the strength of ccˉc\bar c-molecule mixing in hidden-charm states [28, 29, 30, 31]. Its position above open-charm thresholds constrains the properties of the bare χc1(2P)\chi_{c1}(2P) core and its coupling strength to DDˉ∗D\bar D^{*} channels. These analyses find that X(3872)X(3872) does not arise from a simple mass shift of χc1(2P)\chi_{c1}(2P), but from coupled-channel dynamics involving both charmonium and molecular components, with the predicted mass of χc1(4010)\chi_{c1}(4010) acting as a key indicator of this mixing. Lattice determinations of the 2 3P12\,{}^{3}P_{1} charmonium mass cluster around 4010–4067 MeV [32, 33], so a conventional χc1(2P)\chi_{c1}(2P) assignment for χc1(4010)\chi_{c1}(4010) cannot be excluded on mass grounds alone. Our sum rule analysis shows that a D∗±D∓D^{*\pm}D^\mp molecular current also reproduces the observed mass and width, and the coupled-channel studies of Refs. [16, 17] suggest that the physical state carries both components.

Coupling the bare χc1(2P)\chi_{c1}(2P) state to the DDˉ∗D\bar{D}^{*} channels produces two distinct poles in the complex energy plane: one is a bound state near the DDˉ∗D\bar{D}^{*} threshold, and the other originates from the bare state and shifts into the complex continuum [34]. This structure yields a DDˉ∗D\bar{D}^{*} lineshape with a threshold enhancement and a broad resonance-like feature around 3990 MeV, consistent with the analysis in Ref. [17]. The two features correspond to the χc1(3872)\chi_{c1}(3872) and χc1(4010)\chi_{c1}(4010) states, respectively [17].

In the present investigation, we examine the resonance χc1(4010)\chi_{c1}(4010) as an axial vector configuration represented as D∗±D∓D^{*\pm}D^\mp, characterized by the spin-parity JPC=1++J^{PC}=1^{++}, and proceed to compute both the mass, decay constant, and the full decay width of this particular state. The analysis is carried out within the framework of QCDSR, which is a powerful nonperturbative approach in high-energy physics. We derive the mass and decay constant of χc1(4010)\chi_{c1}(4010) from the two-point sum rule. As shown later, its mass is large enough that the dominant decay channels are D∗+D−D^{*+}D^{-} and D∗−D+D^{*-}D^{+}; we compute the corresponding partial widths using the three-point sum rule technique. For the decay χc1(4010)→D∗±D∓\chi_{c1}(4010)\to D^{*\pm}D^\mp, this yields a sum rule for the strong coupling constant g1,2(q2)g_{1,2}(q^{2}) at the χc1(4010)D∗±D∓\chi_{c1}(4010)D^{*\pm}D^\mp vertex. Evaluating g1,2(q2)g_{1,2}(q^{2}) at the D±D^\pm mass shell, q2=mD±2q^{2}=m_{D^\pm}^{2}, fixes g1,2g_{1,2}, which is then used to compute the partial decay width.

The paper is organized as follows. Sec. II computes the mass and decay constant of the axial-vector molecular state χc1(4010)\chi_{c1}(4010). Sec. III analyzes the decay channels χc1(4010)→D∗±D∓\chi_{c1}(4010)\to D^{*\pm}D^\mp and determines the total decay width. Sec. IV presents our conclusions.

QCDSR Formalism: THE SPECTROSCOPIC PARAMETERS OF THE MOLECULE c1(4010)

The QCDSR method rests on the operator product expansion (OPE) [35, 36, 37], in which perturbative QCD at short distances expands a correlator into a power series in the squared momentum, with coefficients given by the Wilson coefficients. A Borel transformation improves the convergence of this series at low momentum, or equivalently long distance. The same correlator can be computed in two different ways: at the quark level, which yields the OPE (or QCD) side, and at the hadronic level,

which produces the phenomenological side. A dispersion relation connects the two, and this connection defines the QCDSR, from which hadronic properties are extracted.

The mass and decay constant of χc1\chi_{c1} follow from the two-point QCDSR approach [4, 38], which works well for both ordinary particles and multiquark hadrons. The central object in this framework is the hadron’s interpolating current. For χc1\chi_{c1}, the current Jμ(x)J_{\mu}(x) is constructed as [39]

Jμ(x)=i2(cˉa(x)γμda(x) dˉb(x)γ5cb(x)−dˉa(x)γμca(x) cˉb(x)γ5db(x)),(2)J_{\mu}(x)=\frac{i}{\sqrt{2}}\left(\bar{c}_{a}(x)\gamma_{\mu}d_{a}(x)\,\bar{d}_{b}(x)\gamma_{5}c_{b}(x)-\bar{d}_{a}(x)\gamma_{\mu}c_{a}(x)\,\bar{c}_{b}(x)\gamma_{5}d_{b}(x)\right), \tag*{(2)}

where a,ba,b are color indices, dd and cc are the down and charm quark fields. It describes a molecular state with JPC=1++J^{PC}=1^{++}.

To obtain the sum rules for the mass and decay constant of χc1\chi_{c1}, it is necessary to start by calculating the following correlation function [37]:

Πμν(p)=i∫d4xeip⋅x⟨0∣T(Jμ(x)Jν†(0))∣0⟩.(3)\Pi_{\mu\nu}(p)=i\int d^{4}x e^{ip\cdot x}\langle0|\mathcal{T}(J_{\mu}(x)J_{\nu}^{\dagger}(0))|0\rangle. \tag*{(3)}

On the phenomenological side, Πμν(p)\Pi_{\mu\nu}(p) can be expressed in terms of the spectral parameters of χc1\chi_{c1}. Inserting the full set of states with JPC=1++J^{PC}=1^{++} into Πμν(p)\Pi_{\mu\nu}(p) and integrating over xx gives ΠμνPhys(p)\Pi_{\mu\nu}^{\mathrm{Phys}}(p):

ΠμνPhys(p)=⟨0∣Jμ∣χc1(p,ε)⟩⟨χc1(p,ε)∣Jν†∣0⟩mχc12−p2+⋯ .(4)\Pi_{\mu\nu}^{\mathrm{Phys}}(p)=\frac{\langle0|J_{\mu}|\chi_{c1}(p,\varepsilon)\rangle\langle\chi_{c1}(p,\varepsilon)|J_{\nu}^{\dagger}|0\rangle}{m_{\chi_{c1}}^{2}-p^{2}}+\cdots. \tag*{(4)}

This is the hadronic representation of the correlator, ΠμνPhys(p)\Pi_{\mu\nu}^{\mathrm{Phys}}(p), the phenomenological side of the sum rule. The term written explicitly in Eq. (4) is the ground-state contribution from χc1\chi_{c1}; the ellipsis represents contributions from radially excited states and the continuum. ΠμνPhys(p)\Pi_{\mu\nu}^{\mathrm{Phys}}(p) can be reformulated in a more convenient form by means of the matrix element

⟨0∣Jμ∣χc1(p,ε)⟩=mχc1fχc1εμ(p),(5)\langle0|J_{\mu}|\chi_{c1}(p,\varepsilon)\rangle=m_{\chi_{c1}}f_{\chi_{c1}}\varepsilon_{\mu}(p), \tag*{(5)}

where εμ\varepsilon_{\mu} is the polarization vector of the axial-vector state. Accordingly, the expression ΠμνPhys(p)\Pi_{\mu\nu}^{\mathrm{Phys}}(p) in terms of the parameters mm and ff is easily expressed as:

ΠμνPhys(p)=mχc12fχc12mχc12−p2(gμν−pμpνmχc12)+⋯ .(6)\Pi_{\mu\nu}^{\mathrm{Phys}}(p)=\frac{m_{\chi_{c1}}^{2}f_{\chi_{c1}}^{2}}{m_{\chi_{c1}}^{2}-p^{2}}\left(g_{\mu\nu}-\frac{p_{\mu}p_{\nu}}{m_{\chi_{c1}}^{2}}\right)+\cdots. \tag*{(6)}

Two independent Lorentz structures parameterize the correlation function. We use the ∼gμν\sim g_{\mu\nu} term for the sum rule analysis since it only receives a contribution from the spin-1 particle. This leads to the invariant amplitude

ΠμνPhys(p)=mχc12fχc12mχc12−p2,(7)\Pi_{\mu\nu}^{\mathrm{Phys}}(p)=\frac{m_{\chi_{c1}}^{2}f_{\chi_{c1}}^{2}}{m_{\chi_{c1}}^{2}-p^{2}}, \tag*{(7)}

which is written as the r.h.s. of Eq. (6). The QCD side is obtained by substituting the explicit form of Jμ(x)J_\mu(x) into Eq. (3) and contracting the heavy and light quark fields, which expresses the result in terms of the charm- and light-quark propagators:

ΠμνOPE(p)=12∫d4x eip⋅x{Tr⁡[γμSdaa′(x)γνSca′a(−x)]Tr⁡[γ5Scbb′(x)γ5Sdb′b(−x)]−Tr⁡[γμSdab′(x)γ5Sdb′a(−x)]Tr⁡[γ5Scba′(x)γνSda′b(−x)]−Tr⁡[γμScab′(x)γ5Sdb′a(−x)]Tr⁡[γ5Sdba′(x)γνSca′b(−x)]+Tr⁡[γμScaa′(x)γνSda′a(−x)]Tr⁡[γ5Sdbb′(x)γ5Scb′b(−x)]}.(8)\Pi_{\mu\nu}^{\mathrm{OPE}}(p)=\frac{1}{2}\int d^{4}x\,e^{ip\cdot x}\left\{\operatorname{Tr}\left[\gamma_{\mu}S_{d}^{aa'}(x)\gamma_{\nu}S_{c}^{a'a}(-x)\right]\operatorname{Tr}\left[\gamma_{5}S_{c}^{bb'}(x)\gamma_{5}S_{d}^{b'b}(-x)\right]\right. \\ -\operatorname{Tr}\left[\gamma_{\mu}S_{d}^{ab'}(x)\gamma_{5}S_{d}^{b'a}(-x)\right]\operatorname{Tr}\left[\gamma_{5}S_{c}^{ba'}(x)\gamma_{\nu}S_{d}^{a'b}(-x)\right] \\ -\operatorname{Tr}\left[\gamma_{\mu}S_{c}^{ab'}(x)\gamma_{5}S_{d}^{b'a}(-x)\right]\operatorname{Tr}\left[\gamma_{5}S_{d}^{ba'}(x)\gamma_{\nu}S_{c}^{a'b}(-x)\right] \\ \left.+\operatorname{Tr}\left[\gamma_{\mu}S_{c}^{aa'}(x)\gamma_{\nu}S_{d}^{a'a}(-x)\right]\operatorname{Tr}\left[\gamma_{5}S_{d}^{bb'}(x)\gamma_{5}S_{c}^{b'b}(-x)\right]\right\}. \tag*{(8)}

For the light dd quark, we use the propagator in coordinate space, which takes the form

Sdij(x)=ix̸2π2x4δij−md4π2x2δij−⟨dˉd⟩12[1−imd4x̸]δij−x2192m02⟨dˉd⟩[1−imd6x̸]δij−igsGijαβ32π2x2(x̸σαβ+σαβx̸)δij−ix2x̸gs2⟨dˉd⟩27776δij−x4⟨dˉd⟩⟨gs2G2⟩27648+⋯ .(9)S_{d}^{ij}(x)=\frac{i\not{x}}{2\pi^{2}x^{4}}\delta_{ij}-\frac{m_{d}}{4\pi^{2}x^{2}}\delta_{ij}-\frac{\langle\bar{d}d\rangle}{12}\left[1-i\frac{m_{d}}{4}\not{x}\right]\delta_{ij}-\frac{x^{2}}{192}m_{0}^{2}\langle\bar{d}d\rangle\left[1-i\frac{m_{d}}{6}\not{x}\right]\delta_{ij} \\ -\frac{ig_{s}G_{ij}^{\alpha\beta}}{32\pi^{2}x^{2}}\left(\not{x}\sigma_{\alpha\beta}+\sigma_{\alpha\beta}\not{x}\right)\delta_{ij}-i\frac{x^{2}\not{x}g_{s}^{2}\langle\bar{d}d\rangle^{2}}{7776}\delta_{ij}-\frac{x^{4}\langle\bar{d}d\rangle\langle g_{s}^{2}G^{2}\rangle}{27648}+\cdots. \tag*{(9)}

The heavy quark propagator Scij(x)S_{c}^{ij}(x) reads:

Scij(x)=i∫d4k(2π)4e−ik⋅x{k̸+mck2−mc2δij−gsGijαβ4σαβ(k̸+mc)+(k̸+mc)σαβ(k2−mc2)2+gs2G212δijmck2+mck̸(k2−mc2)4+gs3G348δijk̸+mc(k2−mc2)6[k̸(k2−3mc2)+2mc(2k2−mc2)]+⋯ }.(10)S_{c}^{ij}(x)=i\int\frac{d^{4}k}{(2\pi)^{4}}e^{-ik\cdot x}\left\{\frac{\not{k}+m_{c}}{k^{2}-m_{c}^{2}}\delta_{ij}-\frac{g_{s}G_{ij}^{\alpha\beta}}{4}\frac{\sigma_{\alpha\beta}(\not{k}+m_{c})+(\not{k}+m_{c})\sigma_{\alpha\beta}}{(k^{2}-m_{c}^{2})^{2}} \right. \\ \left.+\frac{g_{s}^{2}G^{2}}{12}\delta_{ij}m_{c}\frac{k^{2}+m_{c}\not{k}}{(k^{2}-m_{c}^{2})^{4}}+\frac{g_{s}^{3}G^{3}}{48}\delta_{ij}\frac{\not{k}+m_{c}}{(k^{2}-m_{c}^{2})^{6}}\left[\not{k}(k^{2}-3m_{c}^{2})+2m_{c}(2k^{2}-m_{c}^{2})\right]+\cdots\right\}. \tag*{(10)}

The correlation function ΠμνOPE(p)\Pi_{\mu\nu}^{\mathrm{OPE}}(p) is computed from the quark propagators to a fixed order in the OPE. It has a Lorentz structure proportional to the term gμνg_{\mu\nu}, with invariant amplitude ΠOPE(p)\Pi^{\mathrm{OPE}}(p). Equating ΠOPE(p)\Pi^{\mathrm{OPE}}(p) to the phenomenological amplitude ΠPhys(p)\Pi^{\mathrm{Phys}}(p) gives the sum rule, on the phenomenological side of which the ground-state contribution appears together with those of higher resonances and the continuum; these latter contributions must be separated from the ground state. We first apply the Borel transformation to both sides, which suppresses the higher-resonance contributions exponentially and introduces explicit dependence on the Borel parameter M2M^{2}.

The continuum is then removed via quark-hadron duality, which introduces the threshold parameter s0s_{0}. The sum rule is constructed by equating the OPE and phenomenological sides of the correlator [40]:

BΠμνOPE(p2)≡BΠμνPhys(p2).(11)\mathcal{B}\Pi_{\mu\nu}^{\mathrm{OPE}}(p^{2})\equiv\mathcal{B}\Pi_{\mu\nu}^{\mathrm{Phys}}(p^{2}). \tag*{(11)}

By applying the Borel transformation, the main contribution to ΠPhys(p)\Pi^{\mathrm{Phys}}(p) takes a simple form

BΠPhys(M2)=mχc12fχc12e−mχc12/M2.(12)\mathcal{B}\Pi^{\mathrm{Phys}}(M^{2})=m_{\chi_{c1}}^{2}f_{\chi_{c1}}^{2}e^{-m_{\chi_{c1}}^{2}/M^{2}}. \tag*{(12)}

Following the Borel transformation and continuum subtraction, the amplitude ΠμνOPE(p2)\Pi_{\mu\nu}^{\mathrm{OPE}}(p^{2}) reads

Π(M2,s0)=∫4(mc+md)2s0ds ρOPE(s)e−s/M2+Π(M2).(13)\Pi(M^{2},s_{0})=\int_{4(m_{c}+m_{d})^{2}}^{s_{0}}ds\,\rho^{\mathrm{OPE}}(s)e^{-s/M^{2}}+\Pi(M^{2}). \tag*{(13)}

The first term in Eq. (13) accounts for the bulk of the contributions and is written in terms of the two-point spectral density ρOPE(s)\rho^{\mathrm{OPE}}(s), taken as the imaginary part of the correlator. The following term Π(M2)\Pi(M^{2}) includes nonperturbative contributions that are extracted directly from ΠμνOPE(p)\Pi_{\mu\nu}^{\mathrm{OPE}}(p).

The sum rules for mm and ff can be written as:

mχc12=Π′(M2,s0)Π(M2,s0)(14)m_{\chi_{c1}}^{2}=\frac{\Pi'(M^{2},s_{0})}{\Pi(M^{2},s_{0})} \tag*{(14)}

and

fχc12=emχc12/M2mχc12Π(M2,s0),(15)f_{\chi_{c1}}^{2}=\frac{e^{m_{\chi_{c1}}^{2}/M^{2}}}{m_{\chi_{c1}}^{2}}\Pi(M^{2},s_{0}), \tag*{(15)}

where Π′(M2,s0)=dΠ(M2,s0)/d(−1/M2)\Pi'(M^{2},s_{0})=d\Pi(M^{2},s_{0})/d(-1/M^{2}). The mass and decay constant sum rules given by Eqs. (14) and (15) also contain quark, gluon, and mixed condensates that are considered as universal parameters and well-known objects in the computational frameworks [35, 36, 41, 42, 43]. The corresponding numerical values for all of these parameters are given below. We have also included the charm and light quark masses in Eq. (14). The nonperturbative contributions to Π(M2,s0)\Pi(M^2,s_0) coming from vacuum condensates up to the dimension 10 are contained in it. For the numerical calculations, we set md2=0m_d^2=0 and retain only the terms proportional to mdm_d.

The extracted mass and decay constant depend on two auxiliary parameters, the Borel mass M2M^2 and the continuum threshold s0s_0. Physical observables should not depend on these in principle, but the procedure for separating the ground-state contribution from higher resonances and the continuum introduces a residual dependence on both. Standard criteria are used to restrict M2M^2 and s0s_0 to a working window that satisfies two conditions: the pole contribution must dominate, and the OPE must converge. Both conditions are needed for the sum rule predictions to be trustworthy. The pole contribution, PC, is defined by [44]

PC=Π(M2,s0)Π(M2,∞),(16)\mathrm{PC} = \frac{\Pi(M^2,s_0)}{\Pi(M^2,\infty)}, \tag*{(16)}

whereas to check the OPE convergence, we employ the ratio

R(M2)=ΠDimN(M2,s0)Π(M2,s0),(17)R(M^2) = \frac{\Pi^{\mathrm{DimN}}(M^2,s_0)}{\Pi(M^2,s_0)}, \tag*{(17)}

where ΠDimN(M2,s0)\Pi^{\mathrm{DimN}}(M^2,s_0) is the summation of the last few terms in the OPE, with DimN=Dim(8+9+10)\mathrm{DimN}=\mathrm{Dim}(8+9+10). The functions R(M2)R(M^2) and PC help to establish the lower and upper limits for the Borel parameter, respectively. More in detail at Mmin⁡2M^2_{\min} the R(Mmin⁡2)R(M^2_{\min}) must be smaller than a certain prefixed number and at Mmax⁡2M^2_{\max} the pole contribution PC must be larger than a minimal allowed value for this quantity. In the following, we apply the following requirements on PC and R(M2)R(M^2) in the current analysis:

PC≥0.5,R(Mmin⁡2)≤0.05.(18)\mathrm{PC} \ge0.5,\quad R(M^2_{\min}) \le0.05. \tag*{(18)}
ParameterValue [35, 36, 42, 43, 45, 46, 47, 48]
mdm_d4.67−0.17+0.48 MeV4.67^{+0.48}_{-0.17}\ \mathrm{MeV}
mcm_c1.28±0.03 GeV1.28 \pm0.03\ \mathrm{GeV}
⟨dˉd⟩\langle\bar{d}d\rangle(−0.24±0.01)3 GeV3(-0.24 \pm0.01)^3\ \mathrm{GeV}^3
⟨αsπG2⟩\left\langle\frac{\alpha_s}{\pi}G^2\right\rangle(0.012±0.004) GeV4(0.012 \pm0.004)\ \mathrm{GeV}^4
⟨gs3G3⟩\left\langle g_s^3G^3\right\rangle(0.57±0.29) GeV6(0.57 \pm0.29)\ \mathrm{GeV}^6
m02m_0^20.8±0.1 GeV20.8 \pm0.1\ \mathrm{GeV}^2

Table I. TABLE I. Input parameters.

Two conditions determine the allowed windows for M2M^2 and s0s_0: the perturbative part of the correlation function Π(M2,s0)\Pi(M^2,s_0) must dominate over the nonperturbative contributions, as shown in Figure 2, and the extracted hadronic properties must remain stable against variation of the Borel parameter M2M^2. Applying these criteria in the numerical analysis gives the following acceptable windows for M2M^2 and s0s_0:

Figure 2. Normalized contributions to Π(M2,s0)\Pi(M^2,s_0) as functions of the Borel parameter. All curves are evaluated at s0=20 GeV2s_0=20\ \mathrm{GeV}^2.

M2∈[3,4] GeV2,s0∈[19.5,20.5] GeV2.(19)M^2 \in[3,4]\ \mathrm{GeV}^2,\quad s_0 \in[19.5,20.5]\ \mathrm{GeV}^2. \tag*{(19)}

These values satisfy the described restrictions. It can be seen that the pole contribution remains above 0.5 for all M2M^2 and s0s_0 values considered according to Eq. (19).

Predictions regarding the mass and the decay constant are derived by averaging the values computed across different selections of the Borel parameter M2M^2 and the continuum threshold s0s_0:

mχc1=(4005.89±47.15) MeV,m_{\chi_{c1}} = (4005.89 \pm47.15)\ \mathrm{MeV},
fχc1=(2.87±0.55)×10−3 GeV4.(20)f_{\chi_{c1}} = (2.87 \pm0.55) \times10^{-3}\ \mathrm{GeV}^4. \tag*{(20)}

Figure 1. Dependence of the pole contribution (PCPC) on the Borel parameter M2M^2 for several values of s0s_0. The horizontal line corresponds to PC = 0.5, establishing the dominance condition.

Figure 3 and Figure 4 show the mass and decay constant of the D∗DD^*D molecule state as functions of the Borel parameter M2M^2 and the continuum threshold s0s_0. For the sum rule analysis to be reliable, the physical quantities should stay stable as the Borel parameter changes. In practice, the extracted mass depends to some extent on the chosen working region: even within a Borel window that gives a quasi-stable prediction for mm, some residual dependence remains, and this residual dependence is the main source of theoretical error in the sum rule predictions. The errors on the mass are smaller than those on the decay constant, because the mass comes from a ratio of correlators, which cancels part of the M2M^2 and s0s_0 dependence. The decay constant, by contrast, depends explicitly on the correlation function Π(M2,s0)\Pi(M^2,s_0), so it is more sensitive to changes in these auxiliary parameters.

Figure 3. Variation of the molecular mass mχc1m_{\chi_{c1}} with respect to the Borel parameter M2M^2 (left) and the continuum threshold s0s_0 (right) for selected parameter values.

Figure 4. The dependence of the molecular decay constant fχc1f_{\chi_{c1}} on the Borel parameter M2M^2 (left panel) and the continuum threshold s0s_0 (right panel).

Therefore, the theoretical errors on the mass and decay constant are estimated to be around ±1%\pm1\% and ±20%\pm20\%, respectively. The mass of the resonance χc1\chi_{c1} considered as an axial vector molecular state is consistent with the experimental data of the LHCb collaboration. Based on this correspondence, we conclude that χc1\chi_{c1} can be interpreted as a molecular state with quantum numbers JPC=1++J^{PC}=1^{++}.

The spectroscopic parameters of the molecular state χc1\chi_{c1} derived here will be used in the next section to estimate the width of its decay channels.

STRONG COUPLINGS AND DECAY WIDTH OF χc1(4010)\chi_{c1}(4010)

Reliable values for the strong coupling constants are needed to understand strong interaction dynamics and hadronic processes in the nonperturbative regime of QCD. QCDSR is one of the most effective nonperturbative approaches for studying hadronic properties at low energies, and it has been applied to a range of mesonic vertices, including D∗+D−D^{*+}D^- and D∗−D+D^{*-}D^+, to determine their strong couplings more precisely. These couplings enter directly into descriptions of hadronic decays and hadron-charm interactions, and they constrain the internal structure of exotic states and the mechanisms behind charm-flavored hadronic system formation.

The spectroscopic properties of the molecular candidate χc1\chi_{c1} constrain its kinematically accessible decay modes. Since χc1\chi_{c1} has been observed in the D∗+D−D^{*+}D^- and D∗−D+D^{*-}D^+ invariant-mass spectra, B+→D∗±D∓K+B^+ \to D^{*\pm}D^\mp K^+ is likely its dominant production mechanism, with no alternative decay channels yet confirmed experimentally.

[21, 31]

Here we study the D∗+D−D^{*+}D^{-} and D∗−D+D^{*-}D^{+} decay channels of the axial-vector molecule χc1(4010)\chi_{c1}(4010) and predict the corresponding partial decay widths. These are fall-apart (OZI-superallowed) decays: all the valence quarks of the initial χc1\chi_{c1} go directly into the final-state mesons. By considering its mass and spin-parity quantum numbers JPC=1++J^{PC}=1^{++} of χc1\chi_{c1}, we identify its possible decay modes. The decay to the meson pair D∗±D∓D^{*\pm}D^{\mp} is an open channel for the transition of χc1\chi_{c1} into ordinary hadronic states. The strong coupling constant gg governs the decay width of χc1(4010)→D∗+D−/D∗−D+\chi_{c1}(4010)\to D^{*+}D^{-}/D^{*-}D^{+}. Apart from gg, this width depends on the masses of the initial- and final-state mesons and their decay constants. The mass and decay constant of χc1(4010)\chi_{c1}(4010) were determined in the present work, while the parameters of the final-state mesons D∗±D^{*\pm} and D±D^{\pm} are taken from Table II. In this section, we carry out a detailed study of the strong decay processes into D∗+D−D^{*+}D^{-} and D∗−D+D^{*-}D^{+} [49].

ParameterValue (MeV)
mD∗±m_{D^{*\pm}}2010.26±0.052010.26\pm0.05 [45]
mD±m_{D^\pm}1869.65±0.051869.65\pm0.05 [45]
fD±f_{D^\pm}212.7±0.6212.7\pm0.6 [50]
fD∗±f_{D^{*\pm}}263±21263\pm21 [51]

Table II. Masses and decay constants of the D∗±D^{*\pm} and D±D^\pm mesons used as input parameters in the numerical analysis.

χc1→D∗−D+\chi_{c1}\to D^{*-}D^{+}

We initiate our examination with the first decay channel, D∗−D+D^{*-}D^{+}. The first step is to calculate the strong coupling constant related to the vertex χc1D∗−D+\chi_{c1}D^{*-}D^{+}, which describes the strong interaction between the exotic molecular state and the two conventional mesons. To formulate the strong coupling gg, we employ the three-point QCDSR technique for this vertex, starting with an analysis of the following correlation function:

Π(p,p′)=i2∫d4x d4y ei(p′⋅y−p⋅x)×⟨0∣T{JμD∗−(y)JD+(0)J†(x)}∣0⟩.(21)\Pi(p,p')=i^2\int d^4x\,d^4y\,e^{i(p'\cdot y-p\cdot x)} \times\langle0|T\{J_\mu^{D^{*-}}(y)J^{D^+}(0)J^\dagger(x)\}|0\rangle. \tag*{(21)}

Here, J(x)J(x) represents the interpolating current associated with the molecular state χc1\chi_{c1}, whereas JμD∗−(y)J_\mu^{D^{*-}}(y) and JD+(0)J^{D^+}(0) denote the currents corresponding to the vector and pseudoscalar mesons D∗−D^{*-} and D+D^{+}, respectively. The four-momenta of χc1\chi_{c1} and D∗−D^{*-} are represented by the symbols pp and p′p', respectively, while the momentum of the D+D^{+} meson is expressed as q=p−p′q=p-p'. The definitions of the mesonic currents are as follows:

JμD∗−(x)=cˉj(x)iγμdj(x),(22)J_\mu^{D^{*-}}(x)=\bar{c}_j(x)i\gamma_\mu d_j(x), \tag*{(22)}
JD+(x)=dˉi(x)iγ5ci(x),(23)J^{D^+}(x)=\bar{d}_i(x)i\gamma_5 c_i(x), \tag*{(23)}

with ii and jj as color indices. To derive the sum rule, by considering the physical manifestation of the correlation function Π(p,p′)\Pi(p,p'), which corresponds to the phenomenological side of the sum rule and is written as ΠPhys(p,p′)\Pi^{\mathrm{Phys}}(p,p'). This portion is mainly saturated by the contributions from the ground-state mesons, and hence we have:

ΠPhys(p,p′)=⟨0∣JμD∗−∣D∗−(p′)⟩⟨0∣JD+∣D+(q)⟩(p′2−mD∗−2)(q2−mD+2)×⟨D+(q)D∗−(p′)∣χc1(p)⟩⟨χc1(p)∣J†∣0⟩(p2−mχc12)+⋯ .(24)\Pi^{\mathrm{Phys}}(p,p')= \frac{\langle0|J_\mu^{D^{*-}}|D^{*-}(p')\rangle\langle0|J^{D^+}|D^+(q)\rangle} {(p'^{2}-m_{D^{*-}}^{2})(q^{2}-m_{D^+}^{2})} \times \frac{\langle D^+(q)D^{*-}(p')|\chi_{c1}(p)\rangle\langle\chi_{c1}(p)|J^\dagger|0\rangle} {(p^{2}-m_{\chi_{c1}}^{2})} +\cdots. \tag*{(24)}

In Eq. (24), mD+m_{D^+} is the mass of the D+D^+ meson, and the ellipses denote contributions from higher resonances and continuum states. We simplify this expression using the relevant matrix elements, defined in terms of the physical parameters of the mesons. Inserting these matrix elements

⟨0∣JμD∗−∣D∗−(p′,ϵ)⟩=fD∗−mD∗−ϵμ(p′),(25)\langle0|J_\mu^{D^{*-}}|D^{*-}(p',\epsilon)\rangle=f_{D^{*-}}m_{D^{*-}}\epsilon_\mu(p'), \tag*{(25)}
⟨0∣JD+∣D+(q)⟩=fD+mD+2mc+md.(26)\langle0|J^{D^+}|D^+(q)\rangle= \frac{f_{D^+}m_{D^+}^2}{m_c+m_d}. \tag*{(26)}

We describe the vertex χc1D∗−D+\chi_{c1}D^{*-}D^+ by the matrix element

⟨D+(q)D∗−(p,ε)∣χc1(p,ϵ)⟩=g1(q2)[p⋅q ϵ⋅ε∗−p⋅ε∗ q⋅ϵ],(27)\langle D^+(q)D^{*-}(p,\varepsilon)|\chi_{c1}(p,\epsilon)\rangle= g_1(q^2)\left[p\cdot q\,\epsilon\cdot\varepsilon^* - p\cdot\varepsilon^*\,q\cdot\epsilon\right], \tag*{(27)}

here mD∗−m_{D^{*-}}, mD+m_{D^+}, mχc1m_{\chi_{c1}}, fD∗−f_{D^{*-}}, fD+f_{D^+}, fχc1f_{\chi_{c1}} are the masses and decay constants of the mesons D∗−D^{*-}, D+D^+ and χc1\chi_{c1}, respectively, where we denote the strong coupling for χc1→D∗−D+\chi_{c1}\to D^{*-}D^+ by g1(q2)g_1(q^2).

The expression for the strong coupling constant g1(q2)g_1(q^2) depends on the spectroscopic parameters of the molecular state χc1\chi_{c1}, as well as on the masses and decay constants of the D∗±D^{*\pm} and D∓D^\mp mesons, which are used as input parameters in the numerical calculations. These parameters are listed in Table II. Using the hadronic matrix elements, we construct the correlation function ΠPhys(p,p′)\Pi^{\mathrm{Phys}}(p,p') in terms of hadronic parameters; the phenomenological part takes the following form:

Π1μνPhys(p,p′)=g1(q2)fχc1mχc1fD+mD+2fD∗−mD∗−(mc+md)(p2−mχc12)(p′2−mD∗−2)×1q2−mD+2[mD+2−mD∗−2+q22gμν−pμpν+pμ′pν−mD+2mD∗−2pμ′pν′+mχc12+mD∗−2−q22mD∗−2pμpν′]+⋯ .(28)\begin{aligned} \Pi^{\mathrm{Phys}}_{1\mu\nu}(p,p')={}&g_1(q^2)\frac{f_{\chi_{c1}}m_{\chi_{c1}}f_{D^+}m_{D^+}^2f_{D^{*-}}m_{D^{*-}}}{(m_c+m_d)(p^2-m_{\chi_{c1}}^2)(p'^2-m_{D^{*-}}^2)}\\ &\times\frac{1}{q^2-m_{D^+}^2}\left[\frac{m_{D^+}^2-m_{D^{*-}}^2+q^2}{2}g_{\mu\nu}-p_\mu p_\nu\right.\\ &\left.\quad+p'_\mu p_\nu-\frac{m_{D^+}^2}{m_{D^{*-}}^2}p'_\mu p'_\nu+\frac{m_{\chi_{c1}}^2+m_{D^{*-}}^2-q^2}{2m_{D^{*-}}^2}p_\mu p'_\nu\right]+\cdots. \tag*{(28)} \end{aligned}

Applying the double Borel transformation with respect to p2p^2 and p′2p'^2 will efficiently suppress the higher states and the continuum, leading to better convergence of the sum rules.

To calculate the QCD side of the three-point sum rule, we write the correlation function Π(p,p′)\Pi(p,p') in terms of the heavy- and light-quark propagators, and obtain the following expression:

Π1μνOPE(p,p′)=−12∫d4x d4y eip′⋅ye−ip⋅x×{Tr⁡[γμSdia(y−x)γνScai(x−y)]×Tr⁡[γ5Scjb(−x)γ5Sdbj(x)]−Tr⁡[γμSdib(y−x)γ5Scbi(x−y)]×Tr⁡[γ5Scja(−x)γνSdaj(x)]}.(29)\begin{aligned} \Pi^{\mathrm{OPE}}_{1\mu\nu}(p,p')={}&-\frac{1}{\sqrt{2}}\int d^4x\,d^4y\,e^{ip'\cdot y}e^{-ip\cdot x}\\ &\times\left\{\operatorname{Tr}\left[\gamma_\mu S_d^{ia}(y-x)\gamma_\nu S_c^{ai}(x-y)\right]\right.\\ &\quad\times\operatorname{Tr}\left[\gamma_5S_c^{jb}(-x)\gamma_5S_d^{bj}(x)\right]\\ &\quad-\operatorname{Tr}\left[\gamma_\mu S_d^{ib}(y-x)\gamma_5S_c^{bi}(x-y)\right]\\ &\left.\quad\times\operatorname{Tr}\left[\gamma_5S_c^{ja}(-x)\gamma_\nu S_d^{aj}(x)\right]\right\}. \tag*{(29)} \end{aligned}

The correlator ΠOPE(p,p′)\Pi^{\mathrm{OPE}}(p,p') is calculated by taking into account the nonperturbative contributions up to dimension 5 and, similarly to ΠPhys(p,p′)\Pi^{\mathrm{Phys}}(p,p'), it contains the Lorentz structure in the form of the term gμνg_{\mu\nu}. It is shown that the contributions from operators of dimension higher than 5 are small and can easily be neglected without significantly changing the results. With the relevant invariant amplitude denoted by ΠOPE(p2,p′2,q2)\Pi^{\mathrm{OPE}}(p^2,p'^2,q^2), we match its double Borel transformation BΠOPE(p2,p′2,q2)\mathcal{B}\Pi^{\mathrm{OPE}}(p^2,p'^2,q^2) with that of BΠPhys(p2,p′2,q2)\mathcal{B}\Pi^{\mathrm{Phys}}(p^2,p'^2,q^2), then we apply continuum subtraction and obtain the sum rule for the strong coupling g1(q2)g_1(q^2).

After carrying out the double Borel transformation and subtracting the continuum, the QCD side takes the following form:

Π1(M2,s0,q2)=∫4(mc+md)2s0ds∫(mc+md)2s0′ds′ ρ1(s,s′,q2)e−s/M12e−s′/M22,(30)\Pi_1(M^2,s_0,q^2)=\int_{4(m_c+m_d)^2}^{s_0}ds\int_{(m_c+m_d)^2}^{s'_0}ds'\,\rho_1(s,s',q^2)e^{-s/M_1^2}e^{-s'/M_2^2}, \tag*{(30)}

where ρ1(s,s′,q2)\rho_1(s,s',q^2) is the spectral density and M2=(M12,M22)M^2=(M_1^2,M_2^2) and s0=(s0,s0′)s_0=(s_0,s'_0) are the Borel and continuum threshold parameters. The pairs (M12,s0)(M_1^2,s_0) and (M22,s0′)(M_2^2,s'_0) correspond to the channels of the molecule χc1\chi_{c1} and the D∗−D^{*-} meson, respectively. Matching both sides of the sum rule, the strong coupling g1(q2)g_1(q^2) is extracted as in Eq. (31). Hereafter we abbreviate m≡mχc1m\equiv m_{\chi_{c1}} and f≡fχc1f\equiv f_{\chi_{c1}}:

g1(q2)=2(mc+md)fmD+2fD∗−mD∗−q2−mD+2m2−mD∗−2+q2em2/M12emD∗−2/M22Π1(M2,s0,q2).(31)g_1(q^2)=\frac{2(m_c+m_d)}{fm_{D^+}^2f_{D^{*-}}m_{D^{*-}}}\frac{q^2-m_{D^+}^2}{m^2-m_{D^{*-}}^2+q^2}e^{m^2/M_1^2}e^{m_{D^{*-}}^2/M_2^2}\Pi_1(M^2,s_0,q^2). \tag*{(31)}

Given that the sum rules are assumed reliable solely within the spacelike region (Q2=−q2>0Q^2=-q^2>0), we compute g1(Q2)g_1(Q^2) at fixed Q2Q^2 values that range 11–5 GeV25\ \mathrm{GeV}^2.

A further step for the numerical analysis consists of determining the working intervals for the Borel and continuum subtraction parameters, namely, M2M^2 and s0s_0. The bounds on M2M^2 and s0s_0 are the usual ones in the sum rule analysis, and they were discussed in the previous section. The intervals for M12M_1^2 and s0s_0 corresponding to the χc1→D∗−D+\chi_{c1}\to D^{*-}D^+ channel are determined by Eq. (19). The parameters (M22M_2^2, s0′s'_0) for the channel of the D∗−D^{*-} meson are changed within bounds

M22∈[2.5,3.5] GeV2,s0′∈[5.7,5.8] GeV2.(32)M_2^2 \in[2.5, 3.5]~\mathrm{GeV}^2,\quad s'_0 \in[5.7, 5.8]~\mathrm{GeV}^2. \tag*{(32)}

Regions of the M2M^2 and s0s_0 are chosen to suppress their contributions to the coupling g1(Q2)g_1(Q^2). An appropriate analytic function is used to fit the extracted values and extrapolate to the physical plane Q2=−mD+2Q^2=-m_{D^+}^2:

Gi(Q2)=Gi0exp⁡(ci1Q2m2+ci2(Q2m2)2).(33)\mathcal{G}_i(Q^2)=\mathcal{G}_i^0\exp\left(c_i^1\frac{Q^2}{m^2}+c_i^2\left(\frac{Q^2}{m^2}\right)^2\right). \tag*{(33)}

The parameters Gi0\mathcal{G}_i^0, ci1c_i^1, ci2c_i^2 are obtained from a fit to the QCDSR results. Numerical calculations indicate that G10=0.626 GeV−1\mathcal{G}_1^0=0.626~\mathrm{GeV}^{-1}, c11=0.032c_1^1=0.032, and c12=−0.00134c_1^2=-0.00134 lead to a good fit to the sum rule results.

The strong coupling at the physical point is finally given by:

g1≡G1(−mD+2).(34)g_1\equiv\mathcal{G}_1(-m_{D^+}^2). \tag*{(34)}

The strong coupling constants of the heavy-light mesons may be useful in understanding the final-state interactions in the heavy quarkonium (or meson) decays [52]. Calculation of this strong coupling constant in the spacelike region and subsequent extrapolation to Q2=−m2Q^2=-m^2, with mm the mass of the off-shell meson, determines the coupling constant.

To reduce uncertainties from the extrapolation procedure, we perform the calculation twice, with each meson off-shell, usually the ones with the largest mass differences, yielding two different strong coupling constants. These functions must then take the same values at their poles, so we obtain a single coupling constant for the vertex. The calculation of the coupling constant is a primary result, whereas the physically strong coupling constants are derived quantities with values determined in experimental high-energy physics [52, 53]. This can be done within the QCDSR technique, provided that the sum rule is sufficiently stable [54, 55] as illustrated in Figure 5.

Figure 5. The sum-rule predictions together with the fitted curve for the strong coupling g1,2(Q2)g_{1,2}(Q^2) are shown. The red star marks the physical point Q2=−mD±2Q^2=-m^2_{D^\pm}.

At the mass shell q2=mD+2q^2=m_{D^+}^2, this function is equal to

g1≡G1(−mD+2)=(0.62±0.08) GeV−1,(35)g_1\equiv\mathcal{G}_1(-m_{D^+}^2)=(0.62\pm0.08)~\mathrm{GeV}^{-1}, \tag*{(35)}

and the corresponding partial width is computed as:

Γ[χc1→D∗−D+]=g12λ124πmχc12∣M1∣2,(36)\Gamma[\chi_{c1}\to D^{*-}D^+]=g_1^2\frac{\lambda_1}{24\pi m_{\chi_{c1}}^2}|M_1|^2, \tag*{(36)}

where

∣M1∣2=14mD∗−2[m6−2m4mD+2+2(mD∗−3−mD+2mD∗−)2+m2(mD+4+6mD+2mD∗−2−3mD∗−4)].(37)|M_1|^2=\frac{1}{4m_{D^{*-}}^2}\left[m^6-2m^4m_{D^+}^2+2\left(m_{D^{*-}}^3-m_{D^+}^2m_{D^{*-}}\right)^2+m^2\left(m_{D^+}^4+6m_{D^+}^2m_{D^{*-}}^2-3m_{D^{*-}}^4\right)\right]. \tag*{(37)}

Here, λ1=λ(m,mD+,mD∗−)\lambda_1=\lambda(m,m_{D^+},m_{D^{*-}}) is the Källén (triangle) function, which gives the momentum of the final-state particles in the rest frame of the decaying particle. With xx the mass of the decaying particle and y,zy,z the masses of the two final-state particles,

λ(x,y,z)=x4+y4+z4−2(x2y2+x2z2+y2z2)2x.(38)\lambda(x,y,z)=\frac{\sqrt{x^4+y^4+z^4-2(x^2y^2+x^2z^2+y^2z^2)}}{2x}. \tag*{(38)}

The coupling constant obtained above in Eq. (35) allows for a straightforward calculation of the decay width for the process χc1→D∗−D+\chi_{c1}\to D^{*-}D^+:

Γ[χc1→D∗−D+]=(30.53±7.30) MeV.(39)\Gamma[\chi_{c1}\to D^{*-}D^+]=(30.53\pm7.30)~\mathrm{MeV}. \tag*{(39)}

c1 D*+ D-

The second decay channel, χc1→D∗+D−\chi_{c1}\to D^{*+}D^-, may be analyzed in a manner analogous to the initial process. Nonetheless, in this scenario, we utilize distinct interpolating currents that are aligned with the quantum numbers corresponding to the final-state mesons D−D^- and D∗+D^{*+}. In particular, the following formulations for the interpolating currents are employed:

JμD∗+(x)=dˉj(x)iγμcj(x),(40)J_\mu^{D^{*+}}(x)=\bar d_j(x)i\gamma_\mu c_j(x), \tag*{(40)}
JD−(x)=cˉi(x)iγ5di(x),(41)J^{D^-}(x)=\bar c_i(x)i\gamma_5d_i(x), \tag*{(41)}

where the indices ii and jj are color indices.

To proceed with the sum rule analysis, the relevant matrix elements for D−D^- and D∗+D^{*+} mesons need to be defined, as will be discussed in the following steps. These matrix elements are essential to form the phenomenological side of the correlation function connected to the χc1D∗+D−\chi_{c1}D^{*+}D^- vertex

⟨0∣JμD∗+∣D∗+(p′,ε)⟩=fD∗+mD∗+εμ(p′),(42)\langle0|J_\mu^{D^{*+}}|D^{*+}(p',\varepsilon)\rangle=f_{D^{*+}}m_{D^{*+}}\varepsilon_\mu(p'), \tag*{(42)}
⟨0∣JD−∣D−(q)⟩=fD−mD−2mc+md.(43)\langle0|J^{D^-}|D^-(q)\rangle=\frac{f_{D^-}m_{D^-}^2}{m_c+m_d}. \tag*{(43)}

We express the vertex χc1D∗+D−\chi_{c1}D^{*+}D^- by the matrix element

⟨D−(q)D∗+(p′,ε)∣χc1(p,ε)⟩=g2(q2)[p⋅q ε⋅ε∗−p⋅ε∗ q⋅ε].(44)\langle D^-(q)D^{*+}(p',\varepsilon)|\chi_{c1}(p,\varepsilon)\rangle=g_2(q^2)\left[p\cdot q\,\varepsilon\cdot\varepsilon^* - p\cdot\varepsilon^*\,q\cdot\varepsilon\right]. \tag*{(44)}

The strong coupling associated with the process χc1→D∗+D−\chi_{c1}\to D^{*+}D^- is denoted as g2(q2)g_2(q^2). Employing these expressions, the novel representation of the correlation function ΠPhys(p,p′)\Pi^{\mathrm{Phys}}(p,p') can be readily derived, leading to the phenomenological side as follows:

Π2μνPhys(p,p′)=g2(q2)fMfD−mD−2fD∗+mD∗+(mc+md)(p2−m2)(p′2−mD∗+2)×1(q2−mD−2)[m2−mD∗+2+q22gμν−pμpν+pμ′pν−m2mD∗+2pμ′pν′+m2+mD∗+2−q22mD∗+2pμpν′]+⋯ .(45)\Pi^{\mathrm{Phys}}_{2\mu\nu}(p,p') = g_{2}(q^{2})\frac{f_{M}f_{D^{-}}m^{2}_{D^{-}}f_{D^{*+}}m_{D^{*+}}}{(m_{c}+m_{d})(p^{2}-m^{2})(p'^{2}-m^{2}_{D^{*+}})} \times\frac{1}{(q^{2}-m^{2}_{D^{-}})}\left[\frac{m^{2}-m^{2}_{D^{*+}}+q^{2}}{2}g_{\mu\nu}-p_{\mu}p_{\nu} +p'_{\mu}p_{\nu}-\frac{m^{2}}{m^{2}_{D^{*+}}}p'_{\mu}p'_{\nu}+\frac{m^{2}+m^{2}_{D^{*+}}-q^{2}}{2m^{2}_{D^{*+}}}p_{\mu}p'_{\nu}\right] +\cdots. \tag*{(45)}

Applying a double Borel transformation to both p2p^{2} and p′2p'^{2} suppresses contributions from higher states and the continuum, leading to improved convergence of the sum rules. To be able to study the QCD side of the three-point sum rule, we write Π(p,p′)\Pi(p,p') in terms of quark propagators as follows:

Π2μνOPE(p,p′)=−12∫d4x d4y eip′⋅ye−ip⋅x{×Tr⁡[γμScib(y−x)γ5Sdbi(x−y)]×Tr⁡[γ5Sdja(−x)γνScaj(x)]−Tr⁡[γμScia(y−x)γνScai(x−y)]×Tr⁡[γ5Sdjb(−x)γ5Scbj(x)]}.(46)\Pi^{\mathrm{OPE}}_{2\mu\nu}(p,p')=-\frac{1}{\sqrt{2}}\int d^{4}x\,d^{4}y\,e^{ip'\cdot y}e^{-ip\cdot x} \left\{ \times\operatorname{Tr}\left[\gamma_{\mu}S^{ib}_{c}(y-x)\gamma_{5}S^{bi}_{d}(x-y)\right] \times\operatorname{Tr}\left[\gamma_{5}S^{ja}_{d}(-x)\gamma_{\nu}S^{aj}_{c}(x)\right] -\operatorname{Tr}\left[\gamma_{\mu}S^{ia}_{c}(y-x)\gamma_{\nu}S^{ai}_{c}(x-y)\right] \times\operatorname{Tr}\left[\gamma_{5}S^{jb}_{d}(-x)\gamma_{5}S^{bj}_{c}(x)\right] \right\}. \tag*{(46)}

The correlator ΠOPE(p,p′)\Pi^{\mathrm{OPE}}(p,p') is found using nonperturbative contributions up to dimension 5, and like the correlator ΠPhys(p,p′)\Pi^{\mathrm{Phys}}(p,p'), it contains the Lorentz structure proportional to the term gμνg_{\mu\nu}. Our analysis demonstrates that the contributions of operators with dimensions higher than 5 are sufficiently suppressed and can therefore be neglected without introducing any appreciable changes to the final results. We denote the appropriate invariant amplitude by ΠOPE(p2,p′2,q2)\Pi^{\mathrm{OPE}}(p^{2},p'^{2},q^{2}), and we match its double Borel transformation BΠOPE(p2,p′2,q2)\mathcal{B}\Pi^{\mathrm{OPE}}(p^{2},p'^{2},q^{2}) to the Borel-transformed and continuum-subtracted phenomenological side BΠPhys(p2,p′2,q2)\mathcal{B}\Pi^{\mathrm{Phys}}(p^{2},p'^{2},q^{2}), and again after performing the continuum subtraction, we obtain the sum rule for the coupling g2(q2)g_{2}(q^{2}).

After double Borel transformation and continuum subtraction, the QCD side can be written as:

Π2(M2,s0,q2)=∫4(mc+md)2s0ds∫(mc+md)2s0′ds′ ρ2(s,s′,q2)e−s/M12e−s′/M22,(47)\Pi_{2}(M^{2},s_{0},q^{2})=\int_{4(m_{c}+m_{d})^{2}}^{s_{0}}ds\int_{(m_{c}+m_{d})^{2}}^{s'_{0}}ds'\,\rho_{2}(s,s',q^{2}) e^{-s/M_{1}^{2}}e^{-s'/M_{2}^{2}}, \tag*{(47)}

where ρ2(s,s′,q2)\rho_{2}(s,s',q^{2}) is the spectral density and M2=(M12,M22)M^{2}=(M_{1}^{2},M_{2}^{2}) and s0=(s0,s0′)s_{0}=(s_{0},s'_{0}) denote the Borel and continuum threshold parameters. The pairs (M12,s0)(M_{1}^{2},s_{0}) and (M22,s0′)(M_{2}^{2},s'_{0}) correspond to the channels of the molecule χc1\chi_{c1} and the D∗+D^{*+} meson, respectively. Matching both sides of the sum rule, we obtain the strong coupling g2(q2)g_{2}(q^{2}) as:

g2(q2)=2(mc+md)fMfD−mD−2fD∗+mD∗+q2−mD−2m2−mD∗+2+q2em2/M12emD∗+2/M22Π2(M2,s0,q2).(48)g_{2}(q^{2})=\frac{2(m_{c}+m_{d})}{f_{M}f_{D^{-}}m^{2}_{D^{-}}f_{D^{*+}}m_{D^{*+}}} \frac{q^{2}-m^{2}_{D^{-}}}{m^{2}-m^{2}_{D^{*+}}+q^{2}} e^{m^{2}/M_{1}^{2}}e^{m^{2}_{D^{*+}}/M_{2}^{2}}\Pi_{2}(M^{2},s_{0},q^{2}). \tag*{(48)}

Since the sum rules are reliable only in the spacelike region (Q2=−q2>0Q^{2}=-q^{2}>0), we compute g2(Q2)g_{2}(Q^{2}) at fixed values of Q2Q^{2} over the range 1–5 GeV2^{2}. Employing the constraints from Eqs. (19) and (32), we determine the working windows of the Borel parameter M2M^{2} and the continuum threshold s0s_{0}. The resultant values are fitted using an appropriate analytic function that facilitates extrapolation to the physical region Q2=−mD−2Q^{2}=-m^{2}_{D^{-}}:

Gi(Q2)=Gi0exp⁡(ci1Q2m2+ci2(Q2m2)2).(49)\mathcal{G}_{i}(Q^{2})=\mathcal{G}^{0}_{i}\exp\left(c^{1}_{i}\frac{Q^{2}}{m^{2}}+c^{2}_{i}\left(\frac{Q^{2}}{m^{2}}\right)^{2}\right). \tag*{(49)}

The parameters, Gi0\mathcal{G}^{0}_{i}, ci1c^{1}_{i} and ci2c^{2}_{i} are derived from a fit to the QCDSR results. Numerical calculations indicate that G20=0.592GeV−1\mathcal{G}^{0}_{2}=0.592\mathrm{GeV}^{-1}, c21=0.053c^{1}_{2}=0.053 and c22=−0.0035c^{2}_{2}=-0.0035 lead to a very good agreement with the sum rule predictions. At the end, the strong coupling at the physical point is:

g2≡G2(−mD−2),(50)g_{2}\equiv\mathcal{G}_{2}(-m^{2}_{D^{-}}), \tag*{(50)}

As a result, at the mass shell q2=mD−2q^{2}=m^{2}_{D^{-}}, this function takes the value

g2≡G2(−mD−2)=(0.59±0.07) GeV−1.(51)g_{2}\equiv\mathcal{G}_{2}(-m^{2}_{D^{-}})=(0.59\pm0.07)\ \mathrm{GeV}^{-1}. \tag*{(51)}

and the corresponding partial width takes the form:

Γ[χc1→D∗+D−]=g22λ224πmχc12∣M2∣2,(52)\Gamma[\chi_{c1}\to D^{*+}D^{-}]=g_{2}^{2}\frac{\lambda_{2}}{24\pi m^{2}_{\chi_{c1}}}|M_{2}|^{2}, \tag*{(52)}

with λ2=λ(m,mD−,mD∗+)\lambda_{2}=\lambda(m,m_{D^{-}},m_{D^{*+}}) taken from Eq. (38), where

∣M2∣2=14mD∗+2[m6−2m4mD−2+2(mD∗+3−mD−2×mD∗+)2+m2(mD−4+6mD−2mD∗+2−3mD∗+4)].(53)|M_{2}|^{2}=\frac{1}{4m^{2}_{D^{*+}}}\left[m^{6}-2m^{4}m^{2}_{D^{-}}+2\left(m^{3}_{D^{*+}}-m^{2}_{D^{-}}\times m_{D^{*+}}\right)^{2}+m^{2}\left(m^{4}_{D^{-}}+6m^{2}_{D^{-}}m^{2}_{D^{*+}}-3m^{4}_{D^{*+}}\right)\right]. \tag*{(53)}

The coupling constant obtained from Eq. (51) allows for a direct calculation of the decay width of the channel χc1→D∗+D−\chi_{c1}\to D^{*+}D^{-}:

Γ[χc1→D∗+D−]=(27.11±6.30) MeV.(54)\Gamma[\chi_{c1}\to D^{*+}D^{-}]=(27.11\pm6.30)\ \mathrm{MeV}. \tag*{(54)}

Using data concerning the partial decay widths of the principal channels associated with the hadronic molecule χc1D∗±D∓\chi_{c1}D^{*\pm}D^{\mp}, one can readily approximate its total decay width

Γ[χc1→D∗±D∓]=(57.64±9.64) MeV,(55)\Gamma[\chi_{c1}\to D^{*\pm}D^{\mp}]=(57.64\pm9.64)\ \mathrm{MeV}, \tag*{(55)}

which indicates that χc1\chi_{c1} constitutes a relatively broad resonance. Within theoretical and experimental uncertainties, our prediction for the width of the hadronic molecule χc1\chi_{c1} is compatible with the LHCb value in Eq. (1).

Conclusions

We have explored the axial-vector state χc1\chi_{c1} as a hadronic molecule using the QCDSR approach. The predicted mass, mχc1=(4005.89±47.15)m_{\chi_{c1}} = (4005.89 \pm47.15) MeV, agrees with the LHCb measurement [21], consistent with a molecular interpretation of this state. Although the state is unstable against strong dissociation into D∗±D∓D^{*\pm}D^\mp, we calculated the corresponding partial decay widths using the three-point QCD sum rule technique, with quark-hadron duality used to constrain the strong couplings at the relevant meson-meson vertices. The resulting total width, Γχc1=(57.64±9.64)\Gamma_{\chi_{c1}} = (57.64 \pm9.64) MeV, is consistent with the LHCb result as well.

The implications of χc1(4010)\chi_{c1}(4010) for the nature of X(3872)X(3872) are significant. Within the D‾D∗\overline{D}D^* scattering framework, both states emerge simultaneously as poles of the scattering amplitude, forming a hadronic-molecular/charmonium mixed doublet straddling the D‾D∗\overline{D}D^* threshold [16, 17, 31]. The agreement between our derived parameters and the LHCb result [21] supports the molecular interpretation of χc1\chi_{c1} [28, 29].

The coupling constants extracted in this work are important inputs for describing hadronic dynamics, including the treatment of final-state interactions in heavy quarkonium decay and the calculation of absorption cross sections relevant to heavy quarkonium suppression in hadronic matter. While the present results suggest χc1\chi_{c1} as a viable axial-vector molecular state, a fully conclusive picture of its internal structure would benefit from a complementary study of its spectroscopic properties within a diquark-antidiquark (compact tetraquark) configuration, allowing a direct comparison between the two pictures alongside further experimental data. We hope that upcoming high-precision measurements at BESIII, LHCb, Belle II, CEPC, FCC, and ILC will provide further experimental constraints and deepen our understanding of the dynamics of this exotic state.

DECLARATION OF AI-ASSISTED LANGUAGE EDITING

AI assistance was limited to language editing and improving readability. The authors reviewed and revised all such edits and remain fully responsible for the manuscript’s scientific content.

References

References

  1. [1]N. Brambilla et al., Phys. Rept. 873, 1-154 (2020)
  2. [2]R. F. Lebed, R. E. Mitchell and E. S. Swanson, Prog. Part. Nucl. Phys. 93, 143-194 (2017)
  3. [3]S. S. Agaev, K. Azizi and H. Sundu, Phys. Rev. D 93 no.11, 114036 (2016)
  4. [4]S. Agaev, K. Azizi and H. Sundu, Turk. J. Phys. 44 no.2, 95-173 (2020)
  5. [5]S. S. Agaev, K. Azizi, B. Barsbay and H. Sundu, Nucl. Phys. B 939, 130-144 (2019)
  6. [6]S. S. Agaev, K. Azizi and H. Sundu, Nucl. Phys. B 975 (2022), 115650
  7. [7]S. S. Agaev, K. Azizi and H. Sundu, arXiv:2607.09354 .arxiv.org/abs/2607.09354
  8. [8]S. S. Agaev, K. Azizi and H. Sundu, Phys. Lett. B 871 (2025), 140014
  9. [9]F. K. Guo, C. Hanhart, U. G. Meißner, Q. Wang, Q. Zhao and B. S. Zou, Rev. Mod. Phys. 90 no.1, 015004 (2018) [erratum: Rev. Mod. Phys. 94 no.2, 029901 (2022)]
  10. [10]H. X. Chen, W. Chen, X. Liu and S. L. Zhu, Phys. Rept. 639, 1-121 (2016)
  11. [11]G. D. Esmer, B. Barsbay, K. Azizi, H. Sundu and S. Türkmen, arXiv:2603.14368
  12. [12]S. K. Choi et al. [Belle], Phys. Rev. Lett. 91, 262001 (2003)
  13. [13]G. J. Wang, Z. Yang, J. J. Wu, M. Oka and S. L. Zhu, Sci. Bull. 69, 3036-3041 (2024)
  14. [14]T. Aushev et al. [Belle], Phys. Rev. D 81, 031103 (2010)
  15. [15]H. Sundu, SDU J. Nat. Appl. Sci. 20 no.3, 448 (2016)
  16. [16]P. P. Shi, M. Albajadeo, M. L. Du, F. K. Guo and J. Nieves, Phys. Rev. D 111 no.7, 074043 (2025)
  17. [17]X. L. Gao, J. X. Cui, Y. H. Zhou and Z. Y. Zhou, arXiv:2504.14575arxiv.org/abs/2504.14575
  18. [18]S. S. Agaev, K. Azizi and H. Sundu, Phys. Rev. D 95 no.11, 114003 (2017)
  19. [19]S. S. Agaev, K. Azizi and H. Sundu, Phys. Rev. D 107 no.9, 094018 (2023)
  20. [20]S. S. Agaev, K. Azizi and H. Sundu, Phys. Rev. D 107 no.5, 054017 (2023)
  21. [21]R. Aaij et al. [LHCb], Phys. Rev. Lett. 133 no.13, 131902 (2024)
  22. [22]Z. M. Ding, Q. Huang and J. He, Eur. Phys. J. C 85 no.10, 1133 (2025)
  23. [23]R. Aaij et al. [LHCb], JHEP 12, 139 (2020)
  24. [24]S. L. Olsen, T. Skwarnicki and D. Zieminska, Rev. Mod. Phys. 90, no.1, 015003 (2018)
  25. [25]X. S. Dai, S. Jia, A. Nefediev, J. Nieves, C. Shen and L. Zhang, arXiv:2603.09315
  26. [26]Z. G. Wang, Front. Phys. (Beijing) 21, no.1, 016300 (2026)
  27. [27]H. Mutuk, Y. Sarac, H. Gümüş and A. Ozpineci, Eur. Phys. J. C 78 no. 11, 904 (2018)
  28. [28]K. Miyake and Y. Yamaguchi, Phys. Rev. D 112 no.3, 036004 (2025)
  29. [29]K. Miyake and Y. Yamaguchi, PoS HADRON2025, 060 (2026)
  30. [30]J. Z. Wang, Z. Y. Lin, Y. K. Chen, L. Meng and S. L. Zhu, Phys. Rev. D 111 no.11, L111502 (2025)
  31. [31]H. Li, C. Shi, Y. Chen, M. Gong, J. Liang, Z. Liu and W. Sun, arXiv:2402.14541arxiv.org/abs/2402.14541
  32. [32]M. Okamoto et al. [CP-PACS], Phys. Rev. D 65, 094508 (2002)
  33. [33]P. Chen, Phys. Rev. D 64, 034509 (2001)
  34. [34]Q. Deng, R. H. Ni, Q. Li and X. H. Zhong, Phys. Rev. D 110 no.5, 056034 (2024)
  35. [35]M. A. Shifman, A. I. Vainshtein and V. I. Zakharov, Nucl. Phys. B 147, 385 (1979)
  36. [36]M. A. Shifman, A. I. Vainshtein and V. I. Zakharov, Nucl. Phys. B 147, 448 (1979)
  37. [37]L. J. Reinders, H. Rubinstein and S. Yazaki, Phys. Rept. 127, 1 (1985)
  38. [38]R. M. Albuquerque, J. M. Dias, K. P. Khemchandani, A. Martínéz Torres, F. S. Navarra, M. Nielsen and C. M. Zanetti, J. Phys. G 46, no.9, 093002 (2019)
  39. [39]Z. S. Chen, Z. R. Huang, H. Y. Jin, T. G. Steele and Z. F. Zhang, Chin. Phys. C 46, no.6, 063102 (2022)
  40. [40]P. Colangelo and A. Khodjamirian, At The Frontier of Particle Physics, 1495-1576 (2001)
  41. [41]B. L. Ioffe, Nucl. Phys. B 188, 317 (1981), B 191, 591(E) (1981)
  42. [42]B. L. Ioffe, Prog. Part. Nucl. Phys. 56, 232-277 (2006)
  43. [43]S. Narison, Nucl. Part. Phys. Proc. 270-272, 143 (2016)
  44. [44]S. S. Agaev, K. Azizi and H. Sundu, Phys. Rev. D 99, no.11, 114016 (2019)
  45. [45]S. Navas et al. [Particle Data Group], Phys. Rev. D 110, 030001 (2024) and 2025 update
  46. [46]P. Gubler and D. Satow, Prog. Part. Nucl. Phys. 106, 1-67 (2019)
  47. [47]S. Narison, Phys. Lett. B 693, 559-566 (2010) [erratum: 705, 544-544 (2011)]
  48. [48]Q. Yu, H. Zhou and X. G. Wu, arXiv:2605.07618arxiv.org/abs/2605.07618
  49. [49]S. S. Agaev, K. Azizi and H. Sundu, Phys. Lett. B 870, 139885 (2025)
  50. [50]A. Bazavov et al. [Fermilab Lattice and MILC Collaborations], Phys. Rev. D 98, 074512 (2018)
  51. [51]Z. G. Wang, Eur. Phys. J. C 75, 427 (2015)
  52. [52]B. Osório Rodrigues, M. E. Bracco and M. Chiapparini, Nucl. Phys. A 929, 143-156 (2014)
  53. [53]G. L. Yu, Z. Y. Li and Z. G. Wang, Eur. Phys. J. C 75 no.6, 243 (2015)
  54. [54]A. Cerqueira, Jr, B. O. Rodrigues, M. E. Bracco and C. M. Zanetti, Braz. J. Phys. 52 no.1, 12 (2022)
  55. [55]M. E. Bracco, M. Chiapparini, F. S. Navarra and M. Nielsen, Prog. Part. Nucl. Phys. 67, 1019-1052 (2012)

Paper details

vlnc:W_a2qz5a2u
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

AI assistance

View statement
Looking for related papers…
Contents