A relativistic wave function replaces the S-D picture
The deuteron is a spin-1 bound state of a proton and a neutron, and its low-energy properties are usually described with nonrelativistic S- and D-wave components. That description works for bulk properties, but high-energy scattering probes short-distance configurations and the high-momentum parts of the wave function, where relativistic effects become significant. This paper instead uses a relativistic deuteron light-front wave function derived from a realistic nucleon-nucleon interaction within the light-front framework.
In this formulation the deuteron light-front wave function is expressed through six invariant scalar functions f1 to f6 rather than the nonrelativistic S- and D-wave components. The underlying interaction is a one-boson-exchange kernel that includes pseudoscalar (pi, eta), scalar (sigma, delta), and vector (rho, omega) channels, and is a reparametrization of the full Bonn interaction. The same light-front wave functions have previously described deuteron electromagnetic form factors up to Q2 of about 3 GeV2.
Figure 1 of the paper shows that at low momentum q below roughly 0.5 GeV the wave function is dominated by f1 and f2, while f5 becomes comparable in magnitude for q above about 0.5 GeV. In the nonrelativistic limit q much less than the nucleon mass, f3 to f6 become negligible and f1 and f2 reduce to the conventional S- and D-wave components. This is the sense in which the relativistic treatment extends, rather than discards, the familiar picture.
Three LMDs encode three different polarizations
The paper computes longitudinal momentum distributions (LMDs) of the deuteron, which describe the probability of finding a constituent carrying a fractional longitudinal momentum z inside the deuteron. These are extracted from a light-front correlator, with the gauge link reducing to unity in the light-cone gauge A+ = 0.
Three distinct LMDs are parametrized. The unpolarized LMD f1(z) describes the probability of finding an unpolarized constituent carrying momentum fraction z inside an unpolarized deuteron. The longitudinally polarized LMD g1L(z) characterizes a longitudinally polarized constituent inside a longitudinally polarized deuteron. The tensor-polarized LMD f1LL(z) describes unpolarized constituents inside a tensor-polarized deuteron.
These are not interchangeable quantities. They come from different Dirac structures of the correlator, and they are built from different combinations of the helicity-dependent number densities. The unpolarized and longitudinally polarized LMDs are compared with AV18 potential model deuteron wave functions, while the tensor-polarized LMD f1LL(z) is shown separately and takes negative values over the displayed range of z.
Convolution with NNPDF links LMDs to structure functions
The deuteron structure functions are evaluated within the standard convolution formalism. The deuteron parton distribution is factorized as a convolution of the deuteron LMD with the quark distributions inside a nucleon, and the structure function is then written as x times the sum over quark flavors of the squared electric charge times that deuteron parton distribution.
The nucleon parton distribution functions used as input come from the neural-network-based NNPDF collaboration. The paper states that the calculated unpolarized and longitudinally polarized structure functions provide a good description of the experimental data.
This is a model calculation, not a new measurement. The agreement with data is a statement about how well the convolution of these particular LMDs with NNPDF nucleon PDFs reproduces existing measurements, and it inherits whatever constraints and uncertainties the NNPDF input carries.
The tensor structure function b1 stays too small
The most striking result concerns the tensor-polarized structure function b1. If the deuteron were simply an S-wave bound state of a proton and a neutron, b1 would vanish. Even after including the conventional D-state admixture in the standard convolution model, theoretical calculations predict a b1 that differs significantly from the measurements reported by the HERMES collaboration.
In this work, the predicted tensor-polarized structure function b1 remains significantly smaller than the HERMES measurements. The authors interpret this as suggesting that the standard convolution formalism may not be sufficient to fully interpret the data. They also note that the existing HERMES data for the tensor-polarized structure function are subject to large uncertainties.
The paper discusses one possible explanation raised in the literature: hidden-color configurations arising from six-quark components in the deuteron wave function, which could contribute significantly to the measured tensor structure function b1. The authors do not claim to have demonstrated this; they present it as a motivation for new hadronic mechanisms and for more precise measurements.
The EMC ratio reproduces BONuS only qualitatively
The paper also studies the EMC effect in the weakly bound deuteron through the structure function ratio F2D/(F2p+F2n). This ratio deviates from unity because of nuclear binding and Fermi motion effects. The deuteron, with its weak binding energy and simple two-nucleon structure, shows only minimal nuclear modifications compared with heavier nuclei, which makes it a useful reference for isolating genuine medium effects.
The free proton structure function is calculated using the NNPDF data set, while the neutron contribution is constructed assuming charge and isospin symmetry. The analysis reproduces the qualitative behavior observed in the BONuS data, but noticeable differences remain in the position of the EMC suppression region.
This distinction matters for how the result should be read. Reproducing the qualitative shape of the ratio is a weaker statement than reproducing the location of the suppression. The paper is explicit that the position of the suppression region is where the disagreement persists.
The two-nucleon picture cannot fit everything at once
The authors state that within the current two-nucleon picture, the theoretical calculations are unable to simultaneously describe all experimental data. The remaining discrepancies may point to dynamics beyond the conventional hadronic description and could provide sensitivity to the underlying QCD structure of the deuteron.
Several effects are still absent from the present framework. The paper notes the possibility of reducing the disagreement by incorporating higher-twist effects as well as non-nucleonic degrees of freedom into the present model. A natural extension described by the authors is to incorporate higher Fock-state components, such as pionic contributions, that go beyond the two-nucleon picture of the deuteron.
These are stated as possibilities and extensions, not as demonstrated results. The paper does not show that adding these effects resolves the discrepancies; it identifies them as directions that could reduce the disagreement.
Where these predictions apply and where they stop
The predictions are specific to the deuteron as described by this relativistic light-front wave function, built from a one-boson-exchange kernel within the weak-binding approximation, and convoluted with NNPDF nucleon parton distributions. The unpolarized and longitudinally polarized structure functions are the quantities that describe the data well in this setup.
The tensor-polarized structure function b1 is the quantity that does not. Because the HERMES measurements themselves carry large uncertainties, the gap between prediction and measurement should be read with that caveat in mind, and the paper frames more precise measurements as the way to clarify the situation.
The EMC ratio result is likewise bounded: the qualitative behavior matches BONuS, but the position of the suppression region does not. The framework also omits higher-twist effects and non-nucleonic degrees of freedom, which the authors identify as possible sources of the remaining disagreement.
The tensor channel is the decisive test
The paper's own comparison splits its results into two groups. The unpolarized and longitudinally polarized structure functions are consistent with existing data, while the tensor-polarized structure function b1 is not. That split, not the overall agreement, is the most informative outcome of the study.
The practical consequence is a priority ordering. If the standard convolution of a relativistic deuteron wave function with nucleon PDFs already describes the unpolarized and longitudinally polarized channels, then the tensor channel is where the two-nucleon picture is most clearly stressed. The authors connect this to planned measurements: improved studies of the tensor-polarized spin-1 deuteron at Jefferson Lab, complementary information from proton-deuteron Drell-Yan processes at Fermilab, and polarized deuteron beams at NICA and future Electron-Ion Collider facilities.
For a reader deciding what to take from this paper, the tensor-polarized structure function b1 is the quantity to watch. It is where the model and the existing HERMES data diverge most, where the data uncertainties are largest, and where the paper expects the next round of measurements to be most decisive.
Sources used in this answer
Nuclear structure functions and EMC effect in a relativistic deuteron
A relativistic deuteron light-front wave function convoluted with NNPDF nucleon PDFs describes the unpolarized and longitudinally polarized structure functions well, but predicts a tensor-polarized structure function b1 significantly smaller than HERMES measurements, and reproduces the BONuS EMC ratio only qualitatively with differences in the suppression region.