-
Quantum chromodynamics (QCD) is the theory of the strong interaction between quarks mediated by gluons, which are color-charged [1]. In principle, QCD allows complex quark and gluon compositions of hadrons, such as multiquark hadrons, hadronic molecules, hybrid hadrons, and glueballs, which are nonstandard hadronic particles. Most of these nonstandard hadrons have unusual masses, decay widths, etc., which cannot be given a satisfactory explanation by the traditional quark model. To date, more than thirty non-
qˉq state candidates in light and heavy sectors have been reported experimentally [2]. These resonances are crucial for the deep understanding of hadron spectroscopy and the nonperturbative nature and spontaneous chiral symmetry breaking of QCD.In the low-lying scalar meson sector,
f0(980) [3],a0(980) [4], andD∗s0(2317) [5] have nonexoticJP(=0+) quantum numbers. However, their masses are much lower than the quark model expectation for the corresponding P-waveqˉq states [6, 7]. Their natures are still under debate in spite of the efforts during the past several decades. Since their masses are near the threshold of the constituent particles and have spin-parity quantum numbers corresponding to the S-wave combinations of the constituent particles, one would naturally identify them as hadronic molecules, which are analogs of nuclei.f0(980) and/ora0(980) could beKˉK molecules [8–15] andD∗s0(2317) could be aDK molecule [16–26]. Such a picture leads to results consistent with the experiments. In addition to these particles, the possible S-wave bound states of theˉBK ,DˉD ,BˉB ,BD ,ˉBˉK ,DD ,ˉBˉB , andˉBD systems have not been observed experimentally.On the theoretical aspect, the authors in Ref. [26] systematically studied the possible S-wave bound state of two pseudoscalar mesons by the nonrelativistic Schr
¨o dinger (NRS) equation. Reference [18] predicted the existence of aBˉK bound stateB∗s0 with a mass of 5.725± 0.039 GeV based on the heavy chiral unitary approach. Subsequently, Refs. [27] and [28] confirmed the existence ofB∗s0 in theBˉK bound state scenario and further studied the decay widths of its possible decay channels. Recently, Kong et al. [24] systematically investigatedDK /ˉBK andˉDK /BK systems in a quasipotential Bethe-Salpter equation (qBSE) approach by considering the light meson exchange potential and found that only the isoscalar systems can exist as molecular states. However, the mass ofX(5568) reported by the D0 collaboration [29] is considerably below theBK threshold, and hence, it cannot be aBK molecule [30–35]. In Ref. [36], the authors reported their findings on the S-waveDˉD ,BD , andBˉB systems in the chiral SU(3) quark model (QM); their calculations favor the existence of the isoscalarBˉB molecule, but the existence of isovectorDˉD andBD molecules is disfavored. The authors of Ref. [37] studiedD(∗)D(∗) andB(∗)B(∗) molecular states by solving the coupled channel Schr¨o dinger (CCS) equations; onlyI(JP)=1(0+)BB can be a bound state in thePP (P=D,B ) system because the kinetic term is suppressed in the bottom sector, and the effect of channel couplings becomes more important. With the qBSE approach [38], the existence ofDˉD andBˉB molecular states withI(JP)=0(0+) were predicted, yet no bound state was produced from theDD andˉBˉB interaction. In Ref. [39], a new hidden charm resonance with a mass of 3.7 GeV was predicted within the coupled channel unitary approach. Later, theDˉD bound state was searched in several processes, such asB→DˉDK [40],ψ(3770)→γD0ˉD0 [41], andγγ→DˉD [42, 43]. There are some differences in the results of different methods. Therefore, more efforts are needed to investigate the possible S-wave bound state composed of two pseudoscalar mesons.In the present study, we systematically investigate whether the S-wave bound states of two-pseudoscalar meson systems exist in the Bethe-Salpeter (BS) approach (in the ladder approximation and the instantaneous approximation for the kernel). For doubly heavy pseudoscalar meson systems, we consider not only the interaction through exchanged light mesons (ρ, ω and σ) but also the contribution of heavy vector mesons (
J/ψ orΥ ). As reported in Refs. [44, 45], in spite of the large mass ofJ/ψ , which suppresses the propagator of the exchangedJ/ψ , it was found that the interaction could bind theD∗ˉD∗ andDˉD∗ systems. Similarly, Refs. [38, 46] also reported that the contribution from a heavy meson exchange (J/ψ orΥ ) is very important to form a molecular state, especially in systems in which the contributions from ρ and ω cancel each other out.The remainder of this paper is organized as follows. In Sec. II, we discuss and establish the BS equation for two-pseudoscalar meson systems. This equation is solved numerically, and the numerical results of the two-pseudoscalar meson systems are presented in Sec. III. In the last section, we summarize the study.
-
The BS wave function for the bound state
|P⟩ composed of two pseudoscalar mesons have the following form:χ(x1,x2,P)=⟨0|TP1(x1)P2(x2)|P⟩,
(1) where
P1(x1) andP2(x2) are the field operators of the two constituent particles at space coordinatesx1 andx2 , respectively. The BS wave function in momentum space is defined asχP(x1,x2,P)=e−iPX∫d4p(2π)4e−ipxχP(p),
(2) where p represents the relative momentum of the two constituent particles, and
p=λ2p1−λ1p2 (p1=λ1P+p ,p2=λ2P−p ) withλ1=m1/(m1+m2) andλ2=m2/(m1+m2) .p1(2) andm1(2) represent the momentum and mass of the constituent particle, respectively.The BS wave function
χP(p) satisfies the following BS equation:χP(p)=SP1(p1)∫d4q(2π)4K(P,p,q)χP(q)SP2(p2),
(3) where
SP1(p1) andSP2(p2) are the propagators of constituent particles, andK(P,p,q) is the kernel, which is defined as the sum of all two-particle irreducible diagrams.In the following, we use the variables
pl(=p⋅v) andpt(=p−plv) as the longitudinal and transverse projections of the relative momentum (p) along the bound state velocity (v), respectively. Then, the propagators of the constituent mesons can be expressed asSP1(λ1P+p)=i(λ1M+pl)2−ω21+iϵ,
(4) and
SˉP2(λ2P−p)=i(λ2M−pl)2−ω22+iϵ,
(5) where
ω1(2)=√m21(2)+p2t (we have definedp2t=−pt⋅pt ).To obtain the interaction kernel of the two-pseudoscalar meson system through exchanging light and heavy vector mesons and the light scalar meson, the following effective Lagrangians, as in Refs. [38, 47, 48], are needed:
LKKV=igKKρ→ρμ⋅(K†→τ∂μK−∂μK†→τK)+i(gKKωωμ+gKKωϕμ)(K†∂μK−∂μK†K),LDDV=igDDV(Db∂αD†a−D†a∂αDb)Vαba+igDDJ/ψ(D∂αD†−D†∂αD)J/ψα,LBBV=igBBV(Bb∂αB†a−B†a∂αBb)Vαba+igBBΥ(B∂αB†−B†∂αB)Υα,LDDσ=gDDσDaD†aσ,LBBσ=gBBσBaB†aσ,
(6) where
J/ψα ,Υα , and σ represent theJ/ψ ,Υ , and σ field operators, respectively, and the nonet vector meson matrix reads asV=(ρ0√2+ω√2ρ+K∗+ρ−−ρ0√2+ω√2K∗0K∗−ˉK∗0ϕ).
(7) The coupling constants involved in Eq. (6) are taken as
gKKρ=gKKω=gKKϕ=3 ,gDDV=gBBV=βgv/√2 withgv=5.8 ,β=0.9 ,gDDJ/ϕ=mJ/ψ/fJ/ψ withfJ/ψ=405 MeV, andgBBΥ=mΥ/fΥ withfΥ=715.2 MeV.In the so-called ladder approximation, the interaction kernel
K(P,p,q) can be derived in the lowest-order form as follows:K(p1,p2;q1,q1,mV)=−(2π)2δ4(q1+q2−p1−p2)CIgPPVgP′P′V(p1+q1)μ(p2+q2)νΔμν(k,mV),K(p1,p2;q1,q1,mσ)=−(2π)2δ4(q1+q2−p1−p2)CIg2PPσΔσ(k,mσ), (8) where
mV represent the masses of the exchanged light and heavy vector mesons (ρ, ω, ψ,J/ψ , andΥ ).Δμν(k,mV) andΔσ(k,mσ) represent the propagators for the vector and scalar mesons, respectively, and have the following forms:Δμν(k,mV)=−ik2−m2V(gμν−kμkνm2V),Δσ(k,mσ)=ik2−m2σ.
(9) CI in Eq. (8) is the isospin coefficient forI=0 andI=1 . For theKˉK ,DK ,ˉBK ,DˉD ,BˉB , andBD systems,C0={3/2forρ1/2forω1forϕ1forJ/ψ1forΥ1forσ,C1={−1/2forρ1/2forω1forϕ1forJ/ψ1forΥ1forσ.
(10) For the
ˉKˉK ,DˉK ,ˉBˉK ,DD ,ˉBˉB , andˉBD systems,C0={3/2forρ−1/2forω−1forϕ−1forJ/ψ−1forΥ1forσ,C1={−1/2forρ−1/2forω−1forϕ−1forJ/ψ−1forΥ1forσ.
(11) In Eqs. (10) and (11), the exchanged mesons of ϕ,
J/ψ , and Υ only appear for theKˉK /ˉKˉK ,DˉD /DD , andBˉB /ˉBˉB systems, and σ is only considered in the doubly heavy pseudoscalar meson systems.In order to manipulate the off shell effect of the exchanged mesons and finite size effect of the interacting hadrons, we introduce a form factor
F(k2) at each vertex. Generally, the form factor has the following form:FM(k2)=Λ2−m2Λ2−k2,
(12) where Λ, m, and k represent the cutoff parameter, mass, and momentum of the exchanged meson, respectively. This form factor is normalized at the on shell momentum of
k2=m2 . In contrast, ifk2 is taken to be infinitely large (−∞ ), the form factor, which can be expressed as the overlap integral of the wave functions of the hadrons at the vertex, would approach zero. Considering the difference in the wave functions and masses of the light and heavy mesons, and ensuring a positive form factor, different magnitudes of cutoff Λ will be chosen for the heavy and light mesons.The propagators (4) and (5), interaction kernel (8), and form factor (12) are substituted into the BS equation, i.e., Eq. (3), and the instantaneous approximation (
pl=ql , where the energy exchanged between the constituent particles of the binding system is neglected) in the kernel is considered. Consequently, when a vector meson and a scalar meson in the center-of-mass frame of the bound state (→P=0 ) are exchanged, Eq. (3) becomesχP(pl,→pt)=i[(λ1M+pl)2−ω21+iϵ][(λ2M−pl)2−ω22+iϵ]∫dql2πd3→qt(2π)3 ×{CIgPPVgP′P′V4(λ1M+pl)(λ2−pl)+(→pt+→qt)2+(→p2t−→q2t)/m2V−(→pt−→qt)−m2V(Λ2−m2V)2[Λ2+(→pt+→qt)2]2+CIgPPσgP′P′σ1−(→pt−→qt)−m2σ(Λ2−m2σ)2[Λ2+(→pt+→qt)2]2}χP(ql,→qt).
(13) In the above equation, there are poles in the
pl plane at−λ1M−ω1+iϵ ,−λ1M+ω1−iϵ ,λ2M+ω2−iϵ , andλ2M−ω2+iϵ . After integratingpl on both sides of Eq. (13) by selecting the proper contour, we can obtain the three-dimensional integral equation for˜χP(→pt) (˜χP(→pt)=∫dplχP(pl,→pt) ), which only depends on the the three momentum,→pt . By completing the azimuthal integration, the three-dimensional BS equation becomes a one-dimensional integral equation as˜χP(|→pt|)=∫d|→pt|A(|→pt|,|→qt|)˜χP(|→qt|),
(14) where the propagators and kernels after one-dimensional simplification are included in
A(|→pt|,|→qt|) . The numerical solutions for˜χP(|→pt|) can be obtained by discretizing the integration region into n pieces (with n sufficiently large). In this way, the integral equation becomes a matrix equation, and the BS scalar function˜χP(|→pt|) becomes an n dimensional vector. -
In this section, we will solve the BS equation numerically and study whether the S-wave bound states composed of two pseudoscalar mesons exist. In our model, there is only one parameter, the cutoff Λ, which comes from the form factor. The binding energy
Eb is defined asEb=M−m1−m2 in the rest frame of the bound state. We take the averaged masses of the pseudoscalar mesons and the exchanged light and heavy mesons from PDG [2],mK = 494.988 MeV,mD = 1868.04 MeV,mB = 5279.44 MeV,mρ = 775.26 MeV,mω = 782.65 MeV,mϕ = 1019.461 MeV,mJ/ψ = 3096.9 MeV, andmΥ = 9460.3 MeV.In Fig. 1, we present some possible bound states composed of two pseudoscalar mesons when only the light meson (ρ, ω, ϕ, and σ) exchange contributions are considered. Here, we vary the binding energy from 0 to –50 MeV and the cutoff in a wide range of 0.8–5 GeV. We find that only the
KˉK ,DK ,ˉBK ,DˉD ,BˉB ,BD ,DˉK ,ˉBˉK , andˉBD systems withI=0 can exist as bound states. TheˉKˉK ,DD , andˉBˉB systems withI=0 are forbidden because of Bose symmetry, and the interactions inI=1 systems are repulsive; hence, no bound states exist in theˉKˉK ,DD , andˉBˉB systems. Furthermore, we cannot predict with certainty the masses of bound states that will be measured experimentally because our results are dependent on the cutoff Λ. The contribution of the σ exchange is included in our work, despite the large uncertainties in its mass and structure. In our previous works [49, 50] and Ref. [51], it was found that the contribution of the σ exchange is too small to form bound states, and the same result is found in our current work.Figure 1. (color online) The numerical results for
KˉK ,DK ,ˉBK ,DˉD ,BˉB , andBD systems withI=0 (a) andDˉK ,ˉBˉK , andˉBD systems withI=0 (b).The systems that may exist as bound states are presented in Fig. 1. It is noted that, in the hidden bottom system, the cutoff is the smallest in Fig. 1(a). This is because the B meson has the largest mass, which requires a smaller cutoff value as compared with those in the other systems, and the cutoff is determined by the overlap integrals of the wave functions and hadrons at the vertices. The size of the B meson is the smallest among the constituent particles. There is no bound state for the system with
I=1 . This is because the isospin coefficients of ρ and ω are –1/2 and 1/2, respectively, as shown in Eq. (10), and the masses of ρ and ω are almost equal. Therefore, contributions from the ρ and ω exchanges almost cancel each other out. Among these possible bound states, theKˉK andDK bound states can be related to the experimentally observedf0(980) andD∗s0(2317) , respectively [52]. Based on the heavy chiral unitary approach [18] and linear chiral symmetry [53], the authors predicted the existence of a b-partner stateB∗s0 ofD∗s0(2317) as theBˉK bound state, which can also be confirmed in our model with the cutoff Λ = 2436 MeV. The experimentally observedX(5568) cannot be aBˉK bound state in our model [35]. In Ref. [39], a new hidden charm resonance with a mass of 3.7 GeV (named asX(3700) ) was predicted corresponding mostly to aDˉD state. Later, it has been searched inB→DˉDK [40],e+e−→J/ψDˉD [54],ψ(3770)→X(3700)γ [55],γγ→DˉD [42, 43],Λb→ΛDˉD [56], etc. Recently, lattice QCD also found aDˉD bound state just below the threshold with the binding energyEb=−4.0+3.7−5.0 MeV [57]. The existence of theBˉB bound state was also confirmed by the effective potential model [58], the heavy quark effective theory [59], the chiral SU(3) QM [36], and the qBSE [38]. TheBD system, analogous to theDK system, can also be a bound state in the local hidden gauge symmetry (HGS) approach [60].Recently, the
Tcc with the quantum numberI(JP)=0(1+) and quark contentccˉuˉd was reported by the LHCb Collaboration [61], which is the first experimentally discovered open charmed tetraquark state. TheTcc has a mass just below the threshold ofDD∗ and can be an ideal candidate for theDD∗ bound state. In fact, this inspired us to investigate the possibility of two-pseudoscalar meson systems as bound states with open flavor. In Ref. [62], the authors systematically investigated possible deuteron-like molecular states with two heavy quarks using the one-boson-exchange (OBE) model. According to their results, theI=1 DD system might not be a molecule; theI=1 ˉBˉB ,I=0 , andI=1 DˉB systems might be molecule candidates, but the results are sensitive to the cutoff. Based on the Heavy-Meson Effective Theory, the systems ofDD withI=1 ,ˉBˉB withI=1 , andDˉB withI=0 andI=1 can exist as shallow bound states [63].According to the results of the two-pseudoscalar meson systems, which are presented in Fig. 1(b), only the isoscalar system can exist as bound states. This is because, for the isovector systems, the isospin coefficients corresponding to the ρ and ω exchanges are –1/2, as shown in Eq. (11); hence, the total interaction is repulsive in the isovector systems. Comparing the results in Fig. 1 (a) and Fig. 1(b), we can see that the cutoff Λ is larger in Fig. 1(b), which is caused by the difference in the isospin coefficients, i.e., 1/2 and –1/2 in the systems with
I=0 due to the ω exchange in Eq. (10) and Eq. (11), respectively. From Fig. 1(a) and Fig. 1(b), we can also find that, for the constituent particles with same masses, the larger the mass of the constituent particle, the smaller the cutoff Λ; moreover, the larger the difference in masses of the constituent particles, the larger the cutoff Λ. To easily compare the results of different theoretical models, we list the results of some models and our results in Table 1.KˉK DK ˉBK DˉD BˉB BD ˉKˉK DˉK ˉBˉK DD ˉBˉB ˉBD I 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 qBSE [24, 38] − − √ × √ × √ × √ × − − − − √ × √ × × × × × − − HGS [60, 64] − − − − − − − − − − √ × − − − − − − − − × × √ × OBE [58, 62] − − − − − − × √ × √ √ × − − − − − − × × × √ √ √ CCS [37, 65] − − − − − − − − × √ − − − − − − − − × × × √ − − QM [36] − − − − − − √ × √ × √ × − − − − − − − − − − − − NRS [26] × × √ × √ × √ × √ × √ × × × √ × × × × × × × √ × Our results √ × √ × √ × √ × √ × √ × × × √ × √ × × × × × √ × Table 1. The results for different theoretical models.
√ and× indicate whether the corresponding system is a bound state or not, respectively.− indicates the corresponding system without any study.In Ref. [45], the authors systematically studied the interaction of
DˉD∗ in the isospinI=0 channel. In their work, it is shown that the exchange of a lightqˉq is OZI forbidden in theI=1 channel. As a consequence, only theJ/ψ exchange is allowed in the case ofI=1 , and the simultaneous two pion exchange, which was evaluated in Ref. [45,44], was found to be weaker than the exchange of the vector meson. In spite of the large mass of theJ/ψ , which suppresses the propagator of the exchangedJ/ψ , it was found in Refs. [45,44] that the interaction could bind theDˉD∗ andD∗ˉD∗ systems withI=1 weakly. Subsequently, in Ref. [66], it was also found that the bound stateDˉD∗ withIG(JP)=1+(1+) would disappear if theJ/ψ exchange was removed, which means that theJ/ψ exchange is important to provide an attractive interaction and produce the pole in the isovector system. In the present work, we also consider the effects of exchanged heavy mesons. Considering the differences in the wave functions and masses of the light and heavy mesons, and ensuring a positive from factor, we choose different magnitudes of cutoffsΛL andΛH for the exchanged light and heavy mesons, respectively.We vary the cutoff
ΛL of the exchanged light mesons in the range of 800–1500 MeV to find the cutoffΛH of the exchanged heavy mesons that can form bound states. The results for some possible bound states ofDˉD andBˉB are presented in Figs. 2 and 3. From these results in Figs. 2 and 3, we can see that the effect of the exchange of a heavy meson cannot be ignored. As the contributions of the exchanged ρ and ω almost cancel each other out in theI=1 DˉD andBˉB systems, the main contribution comes from the heavy meson exchange. It can be seen that the results forΛH are almost twice the mass of the exchanged heavy meson in Figs. 2(b) and 3(b). This is similar to the case of considering only the light meson exchange, in which the value of the cutoff Λ is also approximately twice the mass of the exchange meson, as in Fig. 1(a). The possibility of bound states being formed when only considering the contribution of the heavy meson exchange, along with the existence ofˉDD andBˉB bound states withI=1 , is yet to be confirmed experimentally. -
In this study, we derived the BS equation for the S-wave
KˉK ,DK ,BˉK ,DˉD ,BˉB ,BD ,KK ,DˉK ,ˉBˉK ,DD ,ˉBˉB , andˉBD systems and systematically investigated the possible bound states of these systems with the ladder approximation and instantaneous approximation for the kernel. In our model, the kernel containing one-particle-exchange diagrams was induced by the light meson (ρ, ω, ϕ, and σ) and heavy meson (J/ψ andΥ ) exchanges. To investigate the bound states, we have numerically solved the BS equations for S-wave systems composed of two pseudoscalar mesons. The possible S-wave bound states investigated in our study are helpful in explaining the structures of experimentally discovered exotic states and predicting unobserved exotic states.We found that the
KˉK ,DK ,ˉBK ,DˉD ,BˉB ,BD ,DˉK ,ˉBˉK , andˉBD systems withI=0 can exist as bound states. TheˉKˉK ,DD , andˉBˉB systems withI=0 are forbidden because of Bose symmetry, and the interactions in theI=1 systems are repulsive; hence, no bound states exist in theˉKˉK ,DD , andˉBˉB systems. We also found that, for the constituent particles with the same mass, the larger the mass of the constituent particle, the smaller the cutoff Λ; moreover, the larger the difference in the masses of the constituent particles, the larger the cutoff Λ. The contribution of the σ exchange is too small to form bound states.In the calculation, we considered the heavy meson exchanges in the kernel. We found that the effect of the heavy meson exchange cannot be neglected for the
DˉD andBˉB systems. Since the contributions from the ρ and ω exchanges almost cancel each other out in theI=1 DˉD andBˉB systems, the main contribution comes from the heavy meson exchanges, and theI=1 DˉD andBˉB systems can exist as bound states. However, since the cutoffΛH for the heavy meson exchanges is very large, the possibility of bound states being formed when only considering the contribution of the heavy meson exchange, along with the existence ofˉDD andBˉB bound states withI=1 , is yet to be confirmed experimentally.With the restarted LHC and other experiments, more experimental studies of exotic hadrons will be performed in the near future. Recently, the LHCb collaboration observed three never-before-seen particles: a new kind of pentaquark and the first-ever pair of tetraquarks, which includes a new type of tetraquark [67]. These will help physicists better understand how quarks bind together into exotic particles. The theoretical explanation of the structures of experimentally observed exotic hadrons and the existence of possible molecular states predicted theoretically remain controversial. Therefore, more precise experimental studies of the exotic states will be needed to test the results of theoretical studies and improve theoretical models.
-
One of the authors (Z.-Y. Wang) thanks Professor Jia-Jun Wu and Dr. Rui-Cheng Li for helpful discussions and useful suggestions.
