⚛️ lattice

Tensor-polarized parton distribution functions of the deuteron by a convolution model

This paper calculates tensor-polarized parton distribution functions for the deuteron using a convolution model, revealing significant discrepancies with existing HERMES data that suggest the need for new theoretical mechanisms beyond a simple proton-neutron bound system, while providing predictions for upcoming experimental tests at Jefferson Lab and Fermilab.

Original authors: S. Kumano, Kenshi Kuroki

Published 2026-07-13

📖 3 min read🧠 Deep dive

CC BY 4.0

Original authors: S. Kumano, Kenshi Kuroki

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine the deuteron, the simplest atomic nucleus after hydrogen, not as a static marble, but as a lively dance duo: a proton and a neutron holding hands, spinning and wobbling together. For decades, physicists have tried to understand the "spin" of these particles—essentially, how their internal gears (quarks and gluons) are turning. While we've studied single dancers (spin-1/2 nucleons) for years, this paper focuses on the unique, two-person spin-1 dance of the deuteron.

The authors set out to predict what happens when you look at this dancing pair through a high-energy microscope. They used a "convolution model," which is like a recipe: take the known, un-spinning ingredients of a single proton or neutron, mix them with a map of how the proton and neutron move inside the deuteron, and stir to get the final flavor. This is the "standard model" of how we usually calculate these nuclear recipes.

They cooked up their numbers at a specific energy level, $Q^2 = 2.5 \text{ GeV}^2$ , and compared their dish to a previous tasting by the HERMES collaboration back in 2005. Here is where the plot twists: the authors' "standard" recipe produced a result that looked completely different from what HERMES found.

In the authors' simulation, the distribution of the "valence" quarks (the main ingredients) flips its sign at a specific point, $x \approx 0.49$ . It's negative for smaller values and positive for larger ones. However, the HERMES data suggested the opposite behavior, flipping much earlier at $x \approx 0.22$ . It's as if the authors predicted a cake that tastes sweet at the bottom and sour at the top, while the HERMES tasters insisted it was sour at the bottom and sweet at the top.

The paper is very clear about what this means: the standard recipe of just a proton and a neutron bound together does not seem to match the HERMES data. The authors explicitly state that this discrepancy suggests we might be missing a "new mechanism" or an "exotic" ingredient—perhaps something like a hidden-color contribution—that isn't part of the simple dance duo model. They do not claim to have solved the mystery; instead, they highlight that the current standard theory is inconsistent with the existing experimental data.

The authors also calculated some "twist-3" effects, which are like subtle, higher-order wobbles in the dance that usually get ignored because they are tiny. However, because the upcoming experiments at the Thomas Jefferson National Accelerator Facility (JLab) will operate at energies where these wobbles might be noticeable, the authors provide these calculated values as a prediction. They also estimated the distribution of "antiquarks" (the anti-ingredients), suggesting that if the sum of the data doesn't add up to zero, it could mean these anti-ingredients are playing a bigger role than we thought.

So, what's the verdict? The authors have provided a fresh, standard-model calculation that acts as a control group. They found that their "standard" prediction clashes with the HERMES data, suggesting that the deuteron might be more complex than a simple proton-neutron pair. They aren't declaring a new discovery yet; rather, they are pointing a finger at the discrepancy and saying, "Something is off here, and we need new data from JLab and other facilities to figure out if we need a whole new theory of how these particles dance." The paper concludes that while the standard model is the best tool we have right now, it might need a serious upgrade to explain what nature is actually doing.

Technical Summary: Tensor-polarized parton distribution functions of the deuteron by a convolution model

Problem Statement
The paper addresses the theoretical description of tensor-polarized parton distribution functions (PDFs) for the spin-1 deuteron. While high-energy spin physics has extensively investigated spin-1/2 nucleons, spin-1 hadrons like the deuteron offer unique access to structure functions (specifically $b_1$ ) that do not exist in spin-1/2 systems. A significant discrepancy exists between theoretical predictions based on the "standard" deuteron model (a bound proton-neutron system) and experimental data from the HERMES collaboration (2005). The HERMES data suggests a finite tensor-polarized antiquark distribution, which challenges the standard convolution model interpretation. With new experimental preparations underway at the Thomas Jefferson National Accelerator Facility (JLab) and potential Drell-Yan experiments at facilities like Fermilab, there is an urgent need to establish robust theoretical predictions for tensor-polarized observables to interpret upcoming data and clarify the nature of the existing discrepancy.

Methodology
The authors employ a convolution formalism to calculate tensor-polarized PDFs for the deuteron. The core assumption is that the tensor-polarized PDFs of the deuteron are derived by convoluting the unpolarized PDFs of the constituent nucleons with the tensor-polarized momentum distribution of those nucleons within the deuteron.

Key methodological components include:

Convolution Integral: The structure function $b_1$ and the tensor-polarized PDFs ( $\delta_T q$ and $\delta_T \bar{q}$ ) are calculated using an integral over the lightcone momentum fraction $y$ of the nucleon in the deuteron. The nucleon momentum distribution, $\delta_T f(y)$ , is derived from the deuteron wave function, specifically utilizing the CD-Bonn potential which includes S- and D-state admixtures.
Flavor Separation: Unlike previous applications of this model, the authors perform the calculation separately for flavor-dependent valence-quark and antiquark distributions. This allows for the isolation of specific quark flavors ( $u, d, s$ ) and their corresponding antiquarks.
Input Parameters: The calculations are performed at $Q^2 = 2.5 \text{ GeV}^2$ . The nucleon structure functions are derived using MSTW2008 parametrizations for PDFs and SLAC-R1998 for the longitudinal-transverse ratio $R$ . Nuclear modifications to the PDFs and the ratio $R$ are neglected, as they are considered small effects in the deuteron.
Higher-Twist Estimation: The calculated twist-2 distributions are converted to the Trento convention ( $f_{1LL}$ ) and subsequently used to estimate twist-3 distributions ( $f_{LT}$ ) via a Wandzura-Wilczek-like relation, neglecting the dynamical twist-3 terms.

Key Contributions and Results

Flavor-Dependent Predictions: The paper provides the first convolution-model calculations of tensor-polarized PDFs separated by flavor for both valence quarks and antiquarks in the deuteron.
- Valence Quarks: The calculated distributions $\delta_T u^D_v$ and $\delta_T d^D_v$ are large at high $x$ . They exhibit a node at $x \approx 0.49$ , being negative for $x < 0.49$ and positive for $x > 0.49$ .
- Antiquarks: The distributions $\delta_T \bar{u}^D$ , $\delta_T \bar{d}^D$ , and $\delta_T \bar{s}^D$ are found at relatively small $x$ and are significantly smaller in magnitude compared to the valence distributions.
Comparison with HERMES Data: The authors compare their "standard" convolution-model results with PDFs determined by a $\chi^2$ $χ^{2}$ fit to the HERMES $b_1$ $b_{1}$ data.
- Discrepancy: A significant difference is observed. The HERMES-fit distributions show a node at $x \approx 0.22$ with opposite oscillatory behavior (positive at small $x$ , negative at large $x$ ) compared to the convolution model.
- Antiquark Implications: The convolution model predicts very small tensor-polarized antiquark distributions, whereas the HERMES data implies a finite sum for $b_1$ , suggesting a non-negligible tensor-polarized antiquark component.
Twist-3 Estimation: Using the Wandzura-Wilczek-like relation, the authors estimate the twist-3 distributions $f_{LT}$ . They note that while these are suppressed by a $1/Q$ factor in the cross-section, the relatively low $Q^2$ values of planned JLab experiments could make these effects sizable and experimentally accessible via semi-inclusive deep inelastic scattering (SIDIS) or Drell-Yan processes.

Significance and Claims
The paper claims that the convolution model represents the "standard" theoretical expectation for the deuteron structure. The substantial disagreement between these standard predictions and the HERMES data indicates that the simple bound system of a proton and a neutron may be insufficient to explain the observed tensor polarization.

The authors suggest that this discrepancy could point to exotic mechanisms, such as hidden-color contributions, or other new hadronic physics beyond the standard model. They emphasize that the upcoming JLab experiments, which will measure the leading-twist structure function $b_1$ with higher precision, are critical for resolving this issue. The provided theoretical predictions serve as a baseline to test against future global analyses of world data. If the new data confirms the HERMES results, it would necessitate a re-evaluation of the deuteron's internal structure; if the data aligns with the convolution model, it would suggest the HERMES extraction requires re-evaluation. The paper concludes that these studies could lead to a new field in hadron physics by clarifying the origin of tensor polarization in spin-1 nuclei.

More like this