betaFeedback

WorksPhysics and Astronomy

Origin of Tensor Polarized Structure of the Deuteron

Abstract

The microscopic origin of the deuteron tensor polarized structure function b1b_1 remains unresolved. Conventional nuclear models attribute a nonzero b1b_1 mainly to the DD-wave component of the deuteron wave function, but generally fail to reproduce the magnitude and momentum-dependence observed by the HERMES experiment. Using a QCD-based light-front framework with degrees of freedom beyond the conventional nucleonic picture, in particular including hidden-color components and relativistic spin–momentum correlations, we find that a small but nonzero estimated hidden-color admixture can be sufficient to generate a substantial tensor response (b1(x)b_1(x)), while it is essentially insensitive to the DD-wave component. The physical DD-wave probability of 0.85%0.85\%, fixed by the empirical quadrupole moment, changes b1b_1 by only 1.7%1.7\%, with the dominant contribution arising from the colored-cluster SS-wave sector. These results demonstrate that the tensor-polarized structure of the deuteron is driven primarily by hidden-color dynamics rather than by the DD-wave component.

Hidden-color dynamics

At hadronic resolution, the deuteron is conventionally described as a weakly bound system of two color-singlet nucleons. At the quark level, however, its six quarks can couple to an overall color singlet through both singlet–singlet and hidden-color channels. We write this structure schematically as

∣D⟩=c11∣ψ1⊗1⟩+c88∣ψ8⊗8⟩;∣c11∣2+∣c88∣2=1,(1)|D\rangle= c_{11}|\psi_{1\otimes1}\rangle+ c_{88}|\psi_{8\otimes8}\rangle;\qquad|c_{11}|^2 + |c_{88}|^2 = 1, \tag*{(1)}

where ∣ψ1⊗1⟩|\psi_{1\otimes1}\rangle represents the conventional color-singlet nucleon–nucleon channel. The state ∣ψ8⊗8⟩|\psi_{8\otimes8}\rangle collectively represents configurations in which two colored subclusters are coupled to an overall color singlet. These configurations are referred to as “hidden color” because the individual colored clusters cannot appear as asymptotic physical states but entangled in their color degrees of freedom with the deuteron. They are expected to become increasingly relevant when the deuteron is resolved at short distances or high resolutions [10, 11].

Orbital and color dynamics represent distinct aspects of the deuteron wave function. The DD-wave describes its nonspherical orbital structure and is essential for the quadrupole moment, whereas hidden color describes non-nucleonic couplings among its six quarks. Including both effects allows us to test whether b1b_1 is controlled by orbital deformation or by the internal color dynamics.

The interplay between hidden-color and orbital dynamics can be studied using different QCD-motivated descriptions. Here, we employ a LF realization in which the hidden-color sector is represented by two effective colored clusters coupled to an overall color-singlet state [21]. This construction provides one concrete way to incorporate non-nucleonic color degrees of freedom while retaining a relativistic description of the internal spin and orbital dynamics.

The corresponding light-front wave function (LFWF) depends on the longitudinal momentum fraction zz, relative transverse momentum k⊥\mathbf{k}_{\perp}, deuteron LF spin projection Λ\Lambda, and cluster helicities λ1\lambda_1 and λ2\lambda_2, and can be decomposed as

Ψ{λ1λ2}Λ(z,k⊥)=∑L,LzS,SzCLLz,SSz1Λ ψ~LLz(z,k⊥) Sλ1λ2SSz(z,k⊥)(2)\Psi^{\Lambda}_{\{\lambda_1\lambda_2\}}(z,\mathbf{k}_{\perp}) = \sum_{\substack{L,L_z\\S,S_z}} C^{1\Lambda}_{LL_z,SS_z}\,\widetilde{\psi}_{LL_z}(z,\mathbf{k}_{\perp})\,\mathcal{S}^{SS_z}_{\lambda_1\lambda_2}(z,\mathbf{k}_{\perp}) \tag*{(2)}

where CLLz,SSz1Λ=⟨LLz,SSz∣1Λ⟩C^{1\Lambda}_{LL_z,SS_z} = \langle LL_z,SS_z|1\Lambda\rangle is the Clebsch–Gordan which couples the orbital angular momentum (L,Lz)(L,L_z) and total cluster spin (S,Sz)(S,S_z) to the deuteron spin J=1J = 1. We retain the 3S1{}^{3}S_1 and 3D1{}^{3}D_1 channels, corresponding to (L,S)=(0,1)(L,S) = (0,1) and (2,1)(2,1), respectively. The function ψ~LLz\widetilde{\psi}_{LL_z} describes the relative cluster motion, while Sλ1λ2SSz\mathcal{S}^{SS_z}_{\lambda_1\lambda_2} is obtained by applying the Melosh rotations to the canonical cluster-spin state [22, 23].

The complete deuteron LFWF is expressed as

ΨΛ=aSΨSΛ+aDΨDΛ;∣aS∣2+∣aD∣2=1(3)\Psi^{\Lambda} = a_S\Psi^{\Lambda}_S + a_D\Psi^{\Lambda}_D;\qquad|a_S|^2 + |a_D|^2 = 1 \tag*{(3)}

where ΨSΛ\Psi^{\Lambda}_S and ΨDΛ\Psi^{\Lambda}_D denote the 3S1{}^{3}S_1 and 3D1{}^{3}D_1 components with the probabilities ∣aS∣2|a_S|^2 and ∣aD∣2|a_D|^2, respectively.

In the mixed longitudinal-momentum and transverse-coordinate representation, the wave function factorizes as [24]

ψLLz(z,b⊥)=ϕL(ζ)2πζeiLzθX(z),ζ=z(1−z) b⊥(4)\psi_{LL_z}(z,\mathbf{b}_{\perp}) = \frac{\phi_L(\zeta)}{\sqrt{2\pi\zeta}}e^{iL_z\theta}X(z),\qquad\zeta= \sqrt{z(1-z)}\,b_{\perp} \tag*{(4)}

where L=∣Lz∣L = |L_z|, b⊥\mathbf{b}_{\perp} is the transverse separation between the clusters, b⊥=∣b⊥∣b_{\perp} = |\mathbf{b}_{\perp}|, and θ\theta is the azimuthal angle of b⊥\mathbf{b}_{\perp}. The transverse radial mode ϕL(ζ)\phi_L(\zeta) is obtained by solving the LF holographic Schrödinger equation [24], while the longitudinal mode X(z)X(z) is determined from the ’t Hooft equation [25]. The factor eiLzθe^{iL_z\theta} carries the azimuthal dependence associated with the orbital-angular-momentum projection LzL_z.

The momentum-space wave function follows from the transverse Fourier transform

ψ~LLz(z,k⊥)=∫d2b⊥(2π)2e−ik⊥⋅b⊥ψLLz(z,b⊥),(5)\widetilde{\psi}_{LL_z}(z,\mathbf{k}_{\perp}) = \int\frac{d^2\mathbf{b}_{\perp}}{(2\pi)^2}e^{-i\mathbf{k}_{\perp}\cdot\mathbf{b}_{\perp}}\psi_{LL_z}(z,\mathbf{b}_{\perp}), \tag*{(5)}

where k⊥\mathbf{k}_{\perp} is conjugate to b⊥\mathbf{b}_{\perp}. The LFWF is obtained by solving the LF Schrödinger-like equations for the transverse and longitudinal dynamics, as detailed in Ref. [21]. The deuteron mass eigenvalue and the corresponding LFWF are determined by three parameters: the effective cluster mass mCm_C, the longitudinal confinement strength gg, and the transverse confinement scale κ\kappa. The parameter gg enters the ’t Hooft equation for the longitudinal mode, whereas κ\kappa determines the LF holographic potential for the transverse mode. Requiring the physical deuteron mass correlates these three parameters and fixes the corresponding LFWF.

The resolution of the effective two-cluster configuration is characterized by the probing energy Ep2=MD−2mCE_p^2 = M_D - 2m_C, where MDM_D is the deuteron mass. Small EpE_p corresponds to the weakly bound nucleonic regime, whereas increasing EpE_p probes shorter-distance dynamics for which colored-cluster configurations can contribute. Although EpE_p does not directly quantify the hidden-color content of the deuteron wave function, it serves as an effective probe of these short-distance color dynamics.

Figure 1 displays the correlated combinations of mCm_C, gg, and κ\kappa that satisfy the deuteron mass eigenvalue condition. Along the solutions shown, increasing mCm_C reduces the confinement strengths required to reproduce MDM_D, whereas decreasing mCm_C requires larger values of both the effective confinement scales. This trend is consistent with the expected color structure of the deuteron at different resolutions. In the formal chiral limit mC→0m_C \to0, the system becomes more compact and strongly interacting; color recoupling then implies an octet–octet probability of about 80% [10, 11]. Conversely, as mC→mNm_C \to m_N, the required interactions weaken and the system approaches the conventional weakly bound singlet–singlet nucleonic configuration, for which P88=∣c88∣2→0P_{88} = |c_{88}|^2 \to0.

Combinations of the effective cluster mass $m_C$ and the longitudinal and transverse confinement strengths $g$ and $\kappa$

Figure 1. Combinations of the effective cluster mass mCm_C and the longitudinal and transverse confinement strengths gg and κ\kappa, respectively that reproduce the deuteron mass. The highlighted point denotes the parameter set used in this work, {mC,g,κ}={0.838,0.50,0.13}\{m_C,g,\kappa\} = \{0.838,0.50,0.13\} GeV, corresponding to Ep≃200E_p \simeq200 MeV.

For the consistency of our analysis approximating the cluster quark PDFs as the unpolarized nucleon PDFs from the NNPDF fits, we use mC=0.838m_C = 0.838 GeV, g=g =. 0.500.50 GeV, and κ=0.13\kappa= 0.13 GeV. This set corresponds to the highlighted point and to Ep≃200E_p \simeq200 MeV [21]. The selected point, corresponding to Ep≃200E_p \simeq200 MeV, stays close to the nucleonic limit and is well separated from the formal chiral limit (as shown in Figure 1). Although the mass eigenvalue condition does not directly determine the color amplitudes c11c_{11} and c88c_{88}, the hidden-color content may be estimated indirectly by interpolating between the two limiting regimes. Assuming a linear interpolation in mCm_C between the compact six-quark limit, P88=80%P_{88}=80\% at mC=0m_C=0, and the nucleonic limit, P88=0P_{88}=0 at mC=mNm_C=m_N, we obtain P88est=0.8(1−mCmN)≃8.5%P_{88}^{\mathrm{est}}=0.8\left(1-\frac{m_C}{m_N}\right)\simeq8.5\%. This provides a qualitative estimate of the hidden-color admixture rather than a probability obtained through an explicit color projection. Within this qualitative interpolation, we estimate that the deuteron contains a hidden-color admixture of less than 10%.

To fix the orbital admixture, we require the full S+DS+D wave function to reproduce the empirical deuteron quadrupole moment, obtaining PD=∣aD∣2=0.85%P_D=|a_D|^2=0.85\%. This relatively small DD-wave probability should be contrasted with the (∼4−6%\sim4-6\%) DD-wave admixture [26, 27, 28] commonly found in conventional descriptions of the deuteron as a weakly bound nucleon–nucleon (NNNN) system. In the present approach, however, the deuteron is described in terms of strongly interacting color clusters rather than as a shallow NNNN bound state. In particular, the inclusion of hidden-color components, even though their overall probability is less than 10%, modifies the internal dynamics and leads to a substantially smaller DD-wave admixture. Thus, the small value of PDP_D is a consequence of the underlying cluster structure of the deuteron with the entangled color degrees of freedom rather than an additional phenomenological assumption.

Despite its small probability, the DD-wave remains important for observables that are particularly sensitive to the orbital structure. In particular, the quadrupole moment receives a strong contribution from the interference between the SS- and DD-wave components. Consequently, the quadrupole moment is highly sensitive to the DD wave: the SS-wave result, QD(S)=0.079 GeV−2Q_D^{(S)}=0.079\ \mathrm{GeV}^{-2}, is brought to the empirical value QDexp=7.34 GeV−2Q_D^{\mathrm{exp}}=7.34\ \mathrm{GeV}^{-2} [29] by the 0.85%0.85\% DD-wave admixture. We therefore use this independently determined value without further adjustment when calculating b1b_1.

In contrast, the small admixture has only a minor effect on the charge and magnetic observables. The charge radius changes from 2.17±0.202.17\pm0.20 fm to 2.21±0.212.21\pm0.21 fm, to be compared with the experimental value rCexp=2.130±0.003±0.009r_C^{\mathrm{exp}}=2.130\pm0.003\pm0.009 fm [30]. The quoted theoretical uncertainties are obtained by varying the parameters mCm_C, gg, and κ\kappa by 10% around their central values. The magnetic radius remains 1.98±0.191.98\pm0.19 fm, consistent with the extracted value rMexp=1.90±0.14r_M^{\mathrm{exp}}=1.90\pm0.14 fm [31], while the magnetic moment changes only from 0.840.84 to 0.820.82, close to the empirical value μDexp=0.85\mu_D^{\mathrm{exp}}=0.85 [29]. The unpolarized and helicity-dependent distributions are likewise essentially unchanged [21].

*Tensor polarized structure function.—*Using the well-established convolution framework, implemented for effective clusters in Ref. [21], we evaluate the quark contribution to the tensor structure function as

b1(x,Q2)=12∑q,Neq2∫x1dzzf1LL(z)qN(xz,Q2),(6)b_1(x,Q^2)=\frac{1}{2}\sum_{q,N}e_q^2\int_x^1\frac{dz}{z}f_{1LL}(z)q^N\left(\frac{x}{z},Q^2\right), \tag*{(6)}

where NN labels the two effective clusters, zz is their longitudinal momentum fraction, and eqe_q is the quark charge in units of the positron charge. The tensor-polarized longitudinal momentum distribution is f1LL(z)=2PΛ0(z)−[PΛ+1(z)+PΛ−1(z)]f_{1LL}(z)=2P^0_{\Lambda}(z)-[P^{+1}_{\Lambda}(z)+P^{-1}_{\Lambda}(z)], where PΛA(z)P^A_{\Lambda}(z) denotes the density of a cluster with helicity +1/2+1/2 in a deuteron with LF helicity Λ\Lambda, obtained from the LFWF by integrating over transverse momentum and summing over the spectator helicity. The cluster quark PDFs qNq^N are approximated by unpolarized nucleon PDFs from the NNPDF fits at Q2≃5 GeV2Q^2\simeq5\ \mathrm{GeV}^2, as in Ref. [21].

As shown in Figure 2, the SS- and S+DS+D-wave results remain close throughout the xx range considered. Their maximum absolute difference, normalized to the maximum magnitude of the SS-wave result, is only 1.7%. The description of the HERMES data [1] is therefore essentially unchanged by the inclusion of the DD wave.

Figure 2. Deuteron tensor structure function xb1(x)xb_1(x) at Q2=5 GeV2Q^2 = 5\ \mathrm{GeV}^2. The orange-shaded band shows the full S+DS + D result with PD=0.85%P_D = 0.85\%, while the purple-shaded band denotes the pure SS-wave result. The bands reflect the 10% variation of mCm_C, gg, and κ\kappa around their central values. The dashed and dotted curves show the conventional nucleonic calculations of Ref. [4]: basic convolution and the virtual-nucleon approximation, respectively. Open circles with error bars represent the HERMES data [1].

Figure Figure 2 also compares our result with the two conventional nucleonic calculations of Ref. [4], obtained using the basic convolution approach and the virtual-nucleon approximation. In the ordinary nucleonic picture, a pure SS-wave deuteron gives b1=0b_1 = 0; a nonzero tensor response arises from SS–DD interference and the pure DD-wave contribution. Their predictions nevertheless differ from the HERMES data in magnitude and xx dependence. By contrast, the colored-cluster SS-wave sector already produces a nonzero result consistent with the data. The DD wave then provides only a small correction, despite being required by the quadrupole moment. This comparison identifies hidden-color dynamics as the dominant source of the tensor response.

**Conclusion and outlook.—**Our results identify a source of the deuteron tensor structure function beyond the conventional mechanism. Within the present relativistic light-front description, hidden-color degrees of freedom produce a nonzero b1b_1 already in the SS-wave sector, consistent with the HERMES data. These results indicate that a small but nonzero estimated hidden-color admixture, below 10%, can be sufficient to generate a substantial tensor response. Adding a DD-wave component with probability PD=0.85%P_D = 0.85\%, fixed independently by the empirical quadrupole moment, changes b1b_1 only slightly: the maximum absolute difference is 1.7% of the maximum magnitude of the SS-wave result. The same DD-wave admixture brings the quadrupole moment into agreement with its empirical value, while leaving the charge and magnetic observables and the unpolarized and helicity-dependent distributions essentially unchanged.

The DD wave needed to reproduce the quadrupole moment therefore does not drive the tensor partonic response. Instead, the persistence of b1b_1 in the colored-cluster SS-wave sector points to non-nucleonic color dynamics as an important ingredient in its microscopic origin. Precision measurements of b1b_1 can test this interpretation and help distinguish the roles of orbital motion and color dynamics in the deuteron.

We are aware of ongoing efforts to measure the transverse-momentum structure of the deuteron through tensor-polarized SIDIS measurements at CLAS12 [20]. The available projections for these observables remain approximate. In our forthcoming study, we will explore how the hidden-color dynamics essential for the inclusive tensor response manifest themselves in the transverse structure of the deuteron. Our preliminary results indicate that hidden-color dynamics are likewise essential for describing the transverse tensor-polarized distributions.

**Acknowledgements.—**We thank Siqi Xu, Gerald A. Miller and James P. Vary for insightful discussions. XZ is supported by National Natural Science Foundation of China, Grant No. 12375143, by Key Research Program of Frontier Sciences, Chinese Academy of Sciences, Grant No. ZDBS-LY-7020, by international partnership program of the Chinese Academy of Sciences, Grant No. 016GJHZ2022103FN, by the Strategic Priority Research Program of the Chinese Academy of Sciences, Grant No. XDB34000000. CRJ is supported in part by the U.S. Department of Energy (Grant No. DE-FG02-03ER41260) and acknowledges the National Energy Research Scientific Computing Center (NERSC) supported by the Office of Science of the U.S. Department of Energy under Contract No. DE-AC02-05CH11231.

References

References

  1. [1]H. A. Airapetian et al. (HERMES), Phys. Rev. Lett. 95, 242001 (2005), arXiv:hep-ex/0506018.DOI
  2. [4]W. Cosyn, Y.-B. Dong, S. Kumano, and M. Sargsian, Phys. Rev. D 95, 074036 (2017), arXiv:1702.05337 [hep-ph].DOI
  3. [10]C. R. Ji, J. Phys. Conf. Ser. 543, 012004 (2014).
  4. [11]B. L. G. Bakker and C.-R. Ji, Prog. Part. Nucl. Phys. 74, 1 (2014).
  5. [20]J. Poudel, A. Bacchetta, J.-P. Chen, D. Keller, I. Fernando, E. Long, D. Ruth, N. Santiesteban, and K. Slifer, (2025), arXiv:2502.20044 [hep-ph].arxiv.org/abs/2502.20044
  6. [21]S. Kaur, C. Mondal, X. Zhao, and C.-R. Ji, Phys. Rev. D 113, 054008 (2026), arXiv:2507.09886 [hep-ph].arxiv.org/abs/2507.09886
  7. [22]H. J. Melosh, Phys. Rev. D 9, 1095 (1974).
  8. [23]C. R. Ji, P. L. Chung, and S. R. Cotanch, Phys. Rev. D 45, 4214 (1992).
  9. [24]S. J. Brodsky, G. F. de Teramond, H. G. Dosch, and J. Erlich, Phys. Rept. 584, 1 (2015), arXiv:1407.8131 [hep-ph].DOI
  10. [25]G. ’t Hooft, Nucl. Phys. B 75, 461 (1974).
  11. [26]R. Machleidt, Phys. Rev. C 63, 024001 (2001), arXiv:nucl-th/0006014.DOI
  12. [27]R. B. Wiringa, V. G. J. Stoks, and R. Schiavilla, Phys. Rev. C 51, 38 (1995), arXiv:nucl-th/9408016.
  13. [28]V. G. J. Stoks, R. A. M. Klomp, C. P. F. Terheggen, and J. J. de Swart, Phys. Rev. C 49, 2950 (1994), arXiv:nucl-th/9406039.DOI
  14. [29]M. Garcon and J. W. Van Orden, Adv. Nucl. Phys. 26, 293 (2001), arXiv:nucl-th/0102049.DOI
  15. [30]I. Sick and D. Trautmann, Nucl. Phys. A 637, 559 (1998).
  16. [31]A. Afanasev, V. D. Afanasev, and S. V. Trubnikov, (1998), arXiv:nucl-th/9808047.

Paper details

vlnc:W_mug2n2zn
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

Not recorded for this paper

View statement
Looking for related papers…
Contents