WorksPhysics and Astronomy
Gravitational form factors in the perturbative limit
Abstract
Generalized distribution amplitudes (GDAs) have attracted significant attention in recent years due to their connection with the energy-momentum tensor (EMT) form factors (FFs). The GDAs can be experimentally accessed through the study of amplitudes in and , where is a pseudoscalar meson pair such as and . In this work, we calculate these amplitudes in the perturbative limit and express the extracted GDAs in terms of meson distribution amplitudes that have been constrained by the previous experiments. Our explicit calculation verifies the existence of a new EMT FF that violates the conservation law of EMT when the hadronic matrix element of the EMT operator is considered separately for each quark flavor. In addition, our result shows that the GDAs are identical in and , which confirms the universality of GDAs in the perturbative limit. In the future, the GDAs and the EMT FFs studied in this paper can be probed at Belle II. Our study enhances the accessibility to the -wave GDAs in and , and provides a promising approach for searching for exotic hybrid mesons in future experiments.
Introduction
Hadronic matrix elements of the energy-momentum tensor (EMT) are commonly parameterized in terms of EMT form factors (FFs), which are also known as gravitational FFs. The EMT FFs have attracted considerable interest because they provide insight into the proton spin puzzle [1] and mechanical properties of hadrons [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15]. They also describe the interaction of hadrons with classical gravity and the manifestation of the equivalence principle which can be considered for quarks and gluons separately [16, 17].
Generalized parton distributions (GPDs) and generalized distribution amplitudes (GDAs) serve as indirect probes of the EMT FFs in the spacelike and timelike regions, respectively [18, 19, 20, 21, 22]. The GDAs have been extensively studied in hadron pair production processes such as [23, 24, 25] and [26, 27, 28, 29], where a perturbative treatment is valid at large photon virtuality and small invariant mass squared of the meson pair. These analyses can be extended to the GDAs accessed in the production of two different pseudoscalar mesons such as and . In contrast to the case, there could be the -wave component in the GDAs, which interestingly leads to the existence of a new EMT FF which may be called the shear viscosity term [30]. The term has not been commonly considered because it breaks the conservation law of hadronic matrix elements of EMT. However, in principle, it could exist for a single quark flavor or gluon as long as it vanishes when we take the sum for all the flavors and gluon. At current stage, there is no clear evidence for the existence of the term. The smallness of its value may indicate a connection [30] between the approximate validity of the equivalence principle for quarks and gluon individually and the low viscosity observed in the holographic framework [31], with the latter further supported by a recent calculation involving black hole gravity [32].
The -wave GDAs are related to the study of exotic hybrid mesons. If a resonance is observed from the -wave in [33] and [28], its quantum number cannot be described by the quark model. Recently, several candidates with have been reported by experiment such as [34, 35], [36, 37], [38, 39, 40, 41] and [42]. However, further investigation is still needed for the confirmation of the states [43, 44]. In future, one can investigate these exotic states in the production of , which is possible at Belle II and BESIII. Furthermore, the measurements of these reactions can be used to study the hadronic light-by-light contribution to the muon’s anomalous magnetic moment [45, 46, 47, 48].
Although the GDA [49] has been extracted from [50], there are no experimental measurements for the GDAs. In light of the current situation, we would like to focus on an interesting way to indirectly evaluate GDAs using meson distribution amplitudes (DAs). It is known that the GDA can be expressed in terms of pion DAs in the amplitude for in the kinematic region , which is referred to as the perturbative limit [51]. In this work, we apply the same technique to the helicity amplitudes for and in order to establish a relation between GDAs and meson DAs in the perturbative limit. The obtained relation will be used to confirm the presence of a nonzero term and, in addition, to verify the universality of the GDAs in and .
The remainder of this paper is organized as follows: In § 2, we provide a brief introduction to the GDAs and define the kinematical variables in . In Section III, we present a detailed calculation of the helicity amplitude for . The relation between GDAs and DAs is obtained in the perturbative limit. In Section IV, we extend this analysis to . The universality of the GDAs is discussed. In § 5, we evaluate the EMT FFs in terms of the meson DAs. § 6 concludes the paper with a summary.
GDAs OF
We define the following kinematic variables for the process of ,
We work in a frame where the virtual photon has only nonzero -component, . For convenience, we introduce two lightcone vectors and , and they are expressed in terms of and ,
The light-cone components of a Lorentz vector are defined as and .
If the QCD factorization () holds, the amplitude for can be factorized into two parts, the subprocess as the hard scattering part and the GDA as the nonperturbative soft part. The factorized formula is illustrated in Figure 1. The GDA is defined by [33, 52, 53, 54],

Figure 1. The leading-twist GDA is accessed in .
where the Wilson line reduces to unity in the lightcone gauge and indicates the momentum fraction carried by the quark hadronizing into pair. The dependence on the factorization scale , which is conventionally set as , is omitted in Eq. (3) for simplicity. The charge conjugation leads to
with . Unlike the GDA, the GDA does not satisfy followed by the interchange of identical particles. In the asymptotic limit , the GDA takes the form
where represents the orbital angular momentum of the meson pair, and is the polar angle of the meson pair in the center of mass frame which can be expressed in terms of the parameter ,
where is the Kallen function.
At the lowest order with respect to , the leading-twist amplitude for survives only in the case that the incoming photons have the same helicity, and it is expressed as [33],
where the helicities of the virtual and real photons are denoted as and , respectively.
GDAs in perturbative limit
In the perturbative limit , the amplitude for can be expressed in terms of the hard scattering amplitude and soft pion DAs [51, 55]. This approach can be extended to the process , and the helicity amplitude is given by
where is the parton level amplitude for , and the sum runs over the quark flavors and due to isospin conservation. The function denotes the DA for a neutral pseudoscalar meson , such as and , and is defined as
where is the momentum fraction carried by the quark hadronizing into . can be written in terms of its eigenfunctions, Gegenbauer polynomials,
where , and is the -meson decay constant for the quark flavor . The isospin symmetry implies for . The -quark DA exists for meson. The Gegenbauer coefficients are different between and , and are denoted by and , respectively.
In addition, we introduce an alternative definition of the DAs,
where the flavor-singlet current and the flavor-octet current are given by
Similarly to Eq. (10), these DAs can be expanded in terms of Gegenbauer polynomials,
Following the convention in [56, 57], we assume that the Gegenbauer coefficients and are the same for the and mesons, respectively. The decay constants and and the Gegenbauer coefficients and are different from those defined for each quark flavor in (10). and describe the deviation of the and mesons from their naive quark content, respectively, and they are responsible for the mixing of and mesons (see e.g. [58]).
From Eqs. (9), (11) and (12), one can show the following relations,
Taking the first moments of these equations with respect to , we obtain the relations among the decay constants [59, 60, 61, 62, 63],
As a result, the relations among the Gegenbauer coefficients are given by
We will evaluate the shear viscosity term in terms of and because we are interested in the existence of a nonzero for each quark flavor and its cancellation among all the relevant quark flavors. However, the expression in terms of and is also important when we take into account the scale evolution with respect to as we will discuss below, and Eq. (16) enables us to switch these two expressions.
In Eq. (8), denotes the amplitude of . The hadronization of the gluon pair into is described by the gluon DA, which is defined as [56, 61, 62, 63, 64]
where the symbol denotes antisymmetrization of a tensor, , and is given by
The gluon DA is also expressed in terms of Gegenbauer polynomials,
where we assume that the Gegenbauer coefficients are the same for and [56, 57]. The gluon DA of vanishes in the asymptotic limit . For convenience, we adopt the convention . Charge conjugation leads to and
The meson DAs depend on the factorization scale , whose evolution is governed by the ERBL equation [65, 66]. The evolution equations for the Gegenbauer coefficients and are given by the same form,
where , , and is the anomalous dimension,
The Gegenbauer coefficients and mix under the evolution [56, 67],
where
The anomalous dimensions are given by
The initial conditions of are simply written in terms of and by setting in Eq. (23). Note that we do not take into account the scale dependence of the singlet decay constant , because it is the order of and ignored in this study.
To express the twist-2 GDAs in terms of meson DAs, we need to calculate the helicity amplitudes of Eq. (8), where the helicities of the incoming photons are identical due to Eq. (7). There are four types of Feynman diagrams contributing to as shown in Figure 2, and the additional diagrams can be found by particle interchange from these four diagrams [51]. We choose the light-cone gauge for the calculation in which the gluon propagator takes the form

Figure 2. Feynman diagrams for in .
where is the momentum of the gluon. If we expand the hard amplitudes in powers of , the diagrams and , the diagram , and the diagram are order of , , and , respectively. Thus we only keep the contribution from in the perturbative limit of [51]. The first term in Eq. (8) is then evaluated as
In Eq. (27), the first (second) part originates from the diagrams where the gluon propagator connects the quark-antiquark pair hadronizing into () meson. Neglecting the terms of order , we have and . Thus, Eq. (27) can be re-expressed as
where is the step function, and the symmetry is taken into account. Comparing Eq. (28) with Eq. (7), the contribution from the first term in Eq. (8) to the quark GDAs is extracted,
We next consider in Eq. (8) associated with the gluon DA. There are also four types of Feynman diagrams contributing to as shown in Figure 3. Only Feynman diagrams of and contribute to the amplitude at the order of , however, the and diagrams are suppressed by , and are therefore neglected in the perturbative limit. The second term of Eq. (8) is then evaluated as

Figure 3. Feynman diagrams for in
The first part in Eq. (30) originates from the diagram . The second (third) part originates from the diagram in which the gluon pair couples to the antiquark (quark) hadronizing into , and they give identical contributions due to the symmetry . We then extract the corresponding GDAs from the amplitude of Eq. (30),
where the condition Eq. (20) is used to simplify the expression. The three parts in Eq. (31) correspond directly to those in Eq. (30). In Eq. (31), we need to regard as a function of and , , and represents the integration boundary of , defined by the condition .
The GDA is just the sum of Eqs. (29) and (31),
Note that is not exactly same as because it contains the components that violate the charge conjugation symmetry given in Eq. (4). We can re-express as
where only the first part satisfies the symmetry of Eq. (4), and the second part does not contribute to the amplitude of Eq. (7) due to the existence of the prefactor . Therefore, in the perturbative limit the GDA is given by
The formula (34) could give logarithmic singularities around and for a specific form of the DAs like the asymptotic form [51], which means that the relation between GDAs and DAs is not well-defined in the whole region of . Fortunately, the singularities are integrable and canceled in the moment of the GDAs as we will see in section V. We regard this as a fact that the moment of the GDAs is insensitive to the soft contribution which spoils the factorized formula (34) in the perturbative limit. More careful treatment was discussed in [51] by introducing a cutoff to regularize the singularities. It was shown that the cutoff dependence turns to a power correction after taking the moment and, therefore, the moment of the GDAs which is expressed in terms of DAs is trustworthy as long as the power correction is negligible.
Universality of GDAs
The GDAs can be probed by a spacelike photon in . Currently, there are no experimental facilities capable of testing the GDAs of Eq. (34) in the perturbative limit. However, the GDAs can also be accessed in , which can be investigated at Belle II in the perturbative limit. The virtual photon is timelike in this process and the helicity amplitudes are also given by Eq. (7) [26, 28]. Therefore, the universality of GDAs can be tested by comparing the results from and , analogous to the universality of hadron GPDs in timelike and spacelike hard exclusive processes [68].
We define the following variables to describe the amplitude for ,
where and are the squared invariant masses for the final meson-photon pairs, and they are positive unlike the variables and in Eq. (1). The process is related to through the crossing symmetry. Our explicit calculation shows that the amplitudes for can also be described by Eqs. (27) and (30) after the following replacements,
and are respectively given by and . The GDAs extracted from the helicity amplitudes of the timelike process are consistent with Eqs. (29), (31) and (34) which are derived from the spacelike process in the perturbative limit. Thus, our calculation verifies the universality of GDAs in this limit.
At Belle II, the kinematic variables can reach and in , and the perturbative limit is sufficiently satisfied. Therefore, our formulas for GDAs can be tested experimentally in the near future.
Gravitational Form Factors
The EMT for a single quark flavor is defined as
where . The first moment of the GDA corresponds to the timelike matrix element of EMT operator [22, 23, 24],
which is expressed in terms of the transition EMT FFs [30, 69],
The term does not appear in the case and the positive corresponds to the stability condition of pion [70, 71]. It is also notable that has been obtained for by the model calculations [22, 72, 73]. The new term is associated with the P-wave component of GDA, which violates the conservation law [30]. However, this violation vanishes when summing over all quark flavors and the gluon. Now these EMT FFs can be expressed in terms of the meson DAs using the extracted GDAs in Eq. (34),
where and . Substituting Eqs. (10) and (19) into Eq. (40), the timelike EMT FFs can be written as
where or denotes the quark flavor, and the Gegenbauer polynomials with are neglected for and due to lengthy expressions. These formulas can simply be extend to by replacing the indices with . If the asymptotic limit is taken, we obtain and find that the term vanishes. However, if the general expressions of meson DAs are adopted, the first term of arises when the quark DA of the meson differs from that of the meson, and the second term will be nonzero provided that the gluon DA does not vanish (), as illustrated by Eq. (41).
If we set GeV, recent studies predicts on average [74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89], where the Eq. (21) is used with the input at GeV. The authors of Ref. [57] obtain , , and from a combined analysis of the experimental measurements by the CLEO Collaboration [90] and BABAR Collaboration [91]. After incorporating the evolution effects described in Eqs. (21) and (23), we have , , and for the quark octet, singlet and gluon DAs, respectively. We take the and decay constants from Ref. [60], and finally obtain the coefficients and using Eq. (16). One can also infer from Eq. (41) that for the term, the contribution of the gluon DA is much larger than that of the gluon DA due to . Thus, the existence of the term seems quite plausible for the and pairs.
Moreover, with some model assumptions [30] one can express the famous ratio of viscosity to entropy density in terms of which appears to be about 0.055. It is slightly smaller than the bound [31] equal to , but mentioned model assumptions cannot pretend for high accuracy and require further studies, especially in the timelike channel. Using the isospin symmetry relations for meson DAs, we find for the and pairs. Consequently, term vanishes when summing over quark flavors, and the conserved hadronic matrix elements of EMT is recovered.
For the pair, Eq. (40) will be slightly modified,
where the quark flavor can be , , or . We obtain using the isospin symmetry relation. The term should vanish when we sum over the quark flavors and the gluon,
The gluon GDA will appear in the amplitudes of [92] and when the higher-order corrections are included, and one of the typical Feynman diagrams is depicted in Figure 4(a). In this work, the gluonic contribution is identically zero, and the existence of a nonzero and its cancellation can be also investigated through the higher-order corrections to and in the perturbative limit, and we also show one of its typical Feynman diagrams in Figure 4(b), which could be addressed in a future study.

Figure 4. (a) Gluon GDA is accessed in the higher order corrections to the amplitudes of [92], and the factorization condition , is satisfied; (b) Gluon GDA is extracted from the higher order corrections to the amplitudes of in the perturbative limit.
In the future, the process can be measured at Belle II in the perturbative limit, which allows us to test the GDAs of Eq. (34) and the EMT FFs of Eqs. (40) and (42) experimentally. At Belle II, the GDAs can be also accessed in , although the kinematics in this case do not satisfy the perturbative limit. Given that the maximum values of and in recent measurements [50] are approximately and , respectively, the kinematic boundary is close to satisfying the perturbative limit. Therefore, Eqs. (34), (40) and (42) can still serve as a boundary constraint in the extraction of GDAs.
The existence of also indicates that there are -wave GDAs for and , enabling the search for exotic resonances through the -wave production of these meson pairs in [33, 52] and [28], which is possible at Belle II and BESIII. Recently, the exotic resonance was observed by BESIII through the -wave analysis of in [34, 35]. Given the similarity between and , it is promising to search for in and .
Summary
In the perturbative limit , the helicity amplitudes of and can be factorized into the hard scattering amplitudes and soft meson DAs even when is a pair of different mesons like and . This fact suggests a possible connection between the GDAs and the meson DAs, particularly the gluon DA associated with production. We have derived the formulas for GDAs in terms of the quark and the gluon DAs, and confirmed the universality of GDAs between and in the perturbative limit. The derived formulas allow us to express the timelike transition EMT FFs in terms of the meson DAs whose parameters have been constrained by the previous experiments. We have verified the existence of a new EMT FF which does not exist for the case, which confirms the anticipated appearance of exotic quantum numbers, being the counterpart of naive T-violation in spacelike channel making, in turn, the contact with a dissipative nature of viscosity. Although this new EMT FF violates the conservation law of EMT when its hadronic matrix element is considered for a single quark flavor, our result ensures that the conservation law is restored for after summing over all the relevant quark flavors. At Belle II, the measurement of satisfies the condition of the perturbative limit, making it possible to test the GDAs and EMT FFs obtained in this work experimentally. Furthermore, the obtained GDAs and EMT FFs can serve as boundary constraints for extracting GDAs from at Belle II and at BESIII. Since the term originates from the -wave components of the GDAs, our study suggests that it is feasible to search for exotic resonances through the -wave production of in and .
Acknowledgments
We acknowledge useful discussions with Markus Diehl, Yang Li, Cédric Lorcé, Sergey Mikhailov, Bernard Pire, Georgy Prokhorov and Valentin Zakharov. Qin-Tao Song was supported by the National Natural Science Foundation of China under Grant Number 12005191. O.V. Teryaev is thankful to PiFi program of CAS for support and Prof. Xurong Chen for hospitality at IMP (Huizhou).
References
- [1]X. D. Ji, Phys. Rev. Lett. 78 (1997), 610-613.
- [2]M. V. Polyakov and P. Schweitzer, Int. J. Mod. Phys. A 33 (2018) no.26, 1830025.
- [3]V. D. Burkert, L. Elouadrhiri and F. X. Girod, Nature 557 (2018) no.7705, 396-399.
- [4]V. D. Burkert, L. Elouadrhiri, F. X. Girod, C. Lorcé, P. Schweitzer and P. E. Shanahan, Rev. Mod. Phys. 95 (2023) no.4, 041002.
- [5]C. Lorcé, H. Moutarde and A. P. Trawiński, Eur. Phys. J. C 79 (2019) no.1, 89.
- [6]H. Dutrieux, C. Lorcé, H. Moutarde, P. Sznajder, A. Trawiński and J. Wagner, Eur. Phys. J. C 81 (2021) no.4, 300.
- [7]K. Kumerički, Nature 570 (2019) no.7759, E1-E2.
- [8]P. E. Shanahan and W. Detmold, Phys. Rev. Lett. 122 (2019) no.7, 072003.
- [9]A. Freese and G. A. Miller, Phys. Rev. D 103 (2021), 094023.
- [10]Y. Li and J. P. Vary, Phys. Rev. D 109 (2024) no.5, L051501.
- [11]A. García Martín-Caro, M. Huidobro and Y. Hatta, Phys. Rev. D 110 (2024) no.3, 034002.
- [12]S. Nair et al. [BLFQ], Phys. Rev. D 110 (2024) no.5, 056027.
- [13]C. Lorcé and Q. T. Song, Phys. Lett. B 864 (2025), 139433.
- [14]W. Broniowski and E. Ruiz Arriola, arXiv:2503.09297 [hep-ph].arxiv.org/abs/2503.09297
- [15]M. Goharipour et al. [MMGPDs], [arXiv:2501.16257 [hep-ph]].arxiv.org/abs/2501.16257
- [16]O. V. Teryaev, arXiv:hep-ph/9904376 [hep-ph].DOI
- [17]O. V. Teryaev, Front. Phys. (Beijing) 11 (2016) no.5, 111207.
- [18]M. Diehl, Phys. Rept. 388 (2003), 41-277.
- [19]A. V. Belitsky and A. V. Radyushkin, Phys. Rept. 418 (2005), 1-387.
- [20]S. Boffi and B. Pasquini, Riv. Nuovo Cim. 30 (2007) no.9, 387-448.
- [21]K. Goeke, M. V. Polyakov and M. Vanderhaeghen, Prog. Part. Nucl. Phys. 47 (2001), 401-515.
- [22]M. V. Polyakov, Nucl. Phys. B 555 (1999), 231.
- [23]M. Diehl, T. Gousset, B. Pire and O. Teryaev, Phys. Rev. Lett. 81 (1998), 1782-1785.
- [24]M. Diehl, T. Gousset and B. Pire, Phys. Rev. D 62 (2000), 073014.
- [25]C. Lorcé, B. Pire and Qin-Tao Song, Phys. Rev. D 106 (2022), 094030.
- [26]Z. Lu and I. Schmidt, Phys. Rev. D 73 (2006), 094021 [erratum: Phys. Rev. D 75 (2007), 099902].
- [27]J. P. Lees et al. [BaBar], Phys. Rev. D 92 (2015) no.7, 072015.
- [28]B. Pire and Q. T. Song, Phys. Rev. D 107 (2023) no.11, 114014.
- [29]B. Pire and Q. T. Song, Phys. Rev. D 109 (2024) no.7, 074016.
- [30]O. Teryaev, JPS Conf. Proc. 37, 020406 (2022).
- [31]P. Kovtun, D. T. Son and A. O. Starinets, Phys. Rev. Lett. 94 (2005), 111601.
- [32]D. D. Lapygin, G. Y. Prokhorov, O. V. Teryaev and V. I. Zakharov, arXiv:2502.18199 [hep-th].
- [33]I. V. Anikin, B. Pire, L. Szymanowski, O. V. Teryaev and S. Wallon, Eur. Phys. J. C 47 (2006), 71-79.
- [34]M. Ablikim et al. [BESIII], Phys. Rev. Lett. 129 (2022) no.19, 192002 [erratum: Phys. Rev. Lett. 130 (2023) no.15, 159901].
- [35]M. Ablikim et al. [BESIII], Phys. Rev. D 106 (2022) no.7, 072012 [erratum: Phys. Rev. D 107 (2023) no.7, 079901].
- [36]D. Alde et al. [IHEP-Brussels-Los Alamos-Annecy (LAPP)], Phys. Lett. B 205 (1988), 397.
- [37]G. S. Adams et al. [E862], Phys. Lett. B 657 (2007), 27-31.
- [38]G. S. Adams et al. [E852], Phys. Rev. Lett. 81 (1998), 5760-5763.
- [39]M. Aghasyan et al. [COMPASS], Phys. Rev. D 98 (2018) no.9, 092003
- [40]A. Rodas et al. [JPAC], Phys. Rev. Lett. 122 (2019), 042002.
- [41]M. G. Alexeev et al. [COMPASS], Phys. Rev. D 105 (2022), 012005.
- [42]J. Kuhn et al. [E852], Phys. Lett. B 595 (2004), 109-117.
- [43]R. L. Workman et al. [Particle Data Group], PTEP 2022 (2022), 083C01.
- [44]H. X. Chen, W. Chen, X. Liu, Y. R. Liu and S. L. Zhu, Rept. Prog. Phys. 86 (2023) no.2, 026201.
- [45]C. Aubin, M. Benayoun, J. Bijnens, T. Blum, I. Caprini, G. Colangelo, H. Czyz, A. Denig, C. A. Dominguez and S. Eidelman, et al. arXiv:1407.4021 [hep-ph].
- [46]I. Danilkin and M. Vanderhaeghen, Phys. Lett. B 789 (2019), 366-372.
- [47]I. Danilkin, C. F. Redmer and M. Vanderhaeghen, Prog. Part. Nucl. Phys. 107 (2019), 20-68.
- [48]B. Moussallam, Eur. Phys. J. C 81 (2021) no.11, 993.
- [49]S. Kumano, Qin-Tao Song and O. V. Teryaev, Phys. Rev. D 97 (2018), 014020.
- [50]M. Masuda et al. [Belle], Phys. Rev. D 93 (2016) no.3, 032003.
- [51]M. Diehl, T. Feldmann, P. Kroll and C. Vogt, Phys. Rev. D 61 (2000), 074029.
- [52]I. V. Anikin, B. Pire, L. Szymanowski, O. V. Teryaev and S. Wallon, Phys. Rev. D 71 (2005), 034021.
- [53]I. V. Anikin, B. Pire, L. Szymanowski, O. V. Teryaev and S. Wallon, Phys. Rev. D 70 (2004), 011501.
- [54]I. V. Anikin, B. Pire, L. Szymanowski, O. V. Teryaev and S. Wallon, Nucl. Phys. A 755 (2005), 561-564.
- [55]S. J. Brodsky and G. P. Lepage, Phys. Rev. D 24 (1981), 1808.
- [56]P. Kroll and K. Passek-Kumericki, Phys. Rev. D 67 (2003), 054017.
- [57]P. Kroll and K. Passek-Kumericki, J. Phys. G 40 (2013), 075005.
- [58]Y. Klopot, A. Oganesian and O. Teryaev, Phys. Rev. D 87 (2013) no.3, 036013 [erratum: Phys. Rev. D 88 (2013) no.5, 059902].
- [59]T. Feldmann, P. Kroll and B. Stech, Phys. Lett. B 449 (1999), 339-346.
- [60]G. S. Bali et al. [RQCD], JHEP 08 (2021), 137.
- [61]Y. Y. Charng, T. Kurimoto and H. n. Li, Phys. Rev. D 74 (2006), 074024 [erratum: Phys. Rev. D 78 (2008), 059901].
- [62]P. Ball and G. W. Jones, JHEP 08 (2007), 025.
- [63]S. S. Agaev, V. M. Braun, N. Offen, F. A. Porkert and A. Schäfer, Phys. Rev. D 90 (2014) no.7, 074019.
- [64]A. Ali and A. Y. Parkhomenko, Eur. Phys. J. C 30 (2003), 367-380.
- [65]G. P. Lepage and S. J. Brodsky, Phys. Lett. B 87 (1979), 359-365.
- [66]A. V. Efremov and A. V. Radyushkin, Phys. Lett. B 94 (1980), 245-250.
- [67]V. N. Baier and A. G. Grozin, Nucl. Phys. B 192 (1981), 476-488.
- [68]D. Mueller, B. Pire, L. Szymanowski and J. Wagner, Phys. Rev. D 86 (2012), 031502.
- [69]H. Pagels, Phys. Rev. 144 (1966), 1250-1260.
- [70]K. Goeke, J. Grabis, J. Ossmann, M. V. Polyakov, P. Schweitzer, A. Silva and D. Urbano, Phys. Rev. D 75 (2007), 094021.
- [71]J. Hudson and P. Schweitzer, Phys. Rev. D 96 (2017) no.11, 114013.
- [72]M. V. Polyakov and C. Weiss, Phys. Rev. D 60 (1999), 114017.
- [73]J. F. Donoghue and H. Leutwyler, Z. Phys. C 52 (1991), 343-351.
- [74]V. L. Chernyak and A. R. Zhitnitsky, Nucl. Phys. B 201 (1982), 492 [erratum: Nucl. Phys. B 214 (1983), 547].
- [75]P. Ball, V. M. Braun and A. Lenz, JHEP 05 (2006), 004.
- [76]A. V. Radyushkin, Phys. Rev. D 80 (2009), 094009.
- [77]R. Arthur, P. A. Boyle, D. Brommel, M. A. Donnellan, J. M. Flynn, A. Juttner, T. D. Rae and C. T. C. Sachrajda, Phys. Rev. D 83 (2011), 074505
- [78]S. S. Agaev, V. M. Braun, N. Offen and F. A. Porkert, Phys. Rev. D 83 (2011), 054020.
- [79]A. P. Bakulev, S. V. Mikhailov, A. V. Pimikov and N. G. Stefanis, Phys. Rev. D 86 (2012), 031501
- [80]S. V. Mikhailov, A. V. Pimikov and N. G. Stefanis, Phys. Rev. D 103 (2021) no.9, 096003
- [81]I. C. Cloët, L. Chang, C. D. Roberts, S. M. Schmidt and P. C. Tandy, Phys. Rev. Lett. 111 (2013), 092001.
- [82]P. Kroll, H. Moutarde and F. Sabatie, Eur. Phys. J. C 73 (2013) no.1, 2278.
- [83]K. Raya, L. Chang, A. Bashir, J. J. Cobos-Martinez, L. X. Gutiérrez-Guerrero, C. D. Roberts and P. C. Tandy, Phys. Rev. D 93 (2016) no.7, 074017.
- [84]G. S. Bali, V. M. Braun, G. Blägle, M. Göckeler, M. Gruber, F. Hutzler, P. Korcyl, A. Schäfer, P. Wein and J. H. Zhang, Phys. Rev. D 98 (2018) no.9, 094507.
- [85]G. S. Bali et al. [R QCD], JHEP 08 (2019), 065; JHEP 11 (2020) 037.
- [86]J. Hua et al. [Lattice Parton], Phys. Rev. Lett. 129 (2022) no.13, 132001.
- [87]T. Zhong, Z. H. Zhu, H. B. Fu, X. G. Wu and T. Huang, Phys. Rev. D 104 (2021) no.1, 016021.
- [88]C. Shi, M. Li, X. Chen and W. Jia, Phys. Rev. D 104 (2021) no.9, 094016.
- [89]X. Gao, A. D. Hanlon, N. Karthik, S. Mukherjee, P. Petreczky, P. Scior, S. Syritsyn and Y. Zhao, Phys. Rev. D 106 (2022) no.7, 074505.
- [90]J. Gronberg et al. [CLEO], Phys. Rev. D 57 (1998), 33-54.
- [91]P. del Moro Sanchez et al. [BaBar], Phys. Rev. D 84 (2011), 052001.
- [92]N. Kivel, L. Mankiewicz and M. V. Polyakov, Phys. Lett. B 467 (1999), 263-270.