Flavour asymmetry of antiquarks in nucleon and nucleus

TL;DR

The paper investigates whether the observed SU(2) flavour asymmetry in the proton sea, historically inferred from deuteron-target data, reflects intrinsic antiquark differences or nuclear effects in the deuteron. It adopts a CT18-like global PDF framework, parameterizing nonperturbative inputs at $Q_0=1.3$ GeV and evolving via the DGLAP equations to NNLO, incorporating pure-proton measurements including the $R$ parameter from Drell–Yan and DIS constraints, and contrasting with a parallel deuteron-based fit. The main finding is that $ar{d}(x)/ar{u}(x) \\approx 1$ around $x\sim 0.1$ for the proton data, while the deuteron-based analysis preserves the traditional $ar{d}(x)/ar{u}(x)>1$, implying a possible high-energy nuclear effect in the deuteron and challenging the conventional proton structure picture. This motivates proton-only global analyses and new high-energy measurements to extract antiquark flavour without nuclear bias and highlights the need to understand nuclear effects at high energy scales.

Abstract

Over the years, comprehensive experiments have shown a fact that the nucleons, such as the proton and neutron, are formed by not only the "valence" up and down quarks which were thought to comprise the nucleons in a simple constituent picture, but also "sea" quarks which can be any other flavour. However, it is still unknown how sea quarks are generated inside the nucleons. Since 1990s, measurements on high energy deuterons (formed by a proton and a neutron) indicated that the anti-down quark contribution was higher than the anti-up quark in the proton, based on the assumptions of the proton-neutron isospin symmetry and a small nuclear effect of the deuteron. Henceforth, sea quarks are considered to be generated via some flavour-asymmetrical mechanisms. Here we report an analysis on a series of new measurements from pure proton interactions which are free from those assumptions, unexpectedly showing that the anti-down quark component is rather consistent with the anti-up quark. It appears to be evidence that the previously observed asymmetry was caused by an unknown nuclear effect in the deuteron, rather than by a difference between antiquarks. We anticipate this work to be an essential new discovery and a motivation for studying nuclear structure, both experimentally and theoretically, at high energy scales, as it now appears fundamentally different from our understanding established in the past.

Figures (3)

• Figure 1: Distributions of $\bar{d}(x)/\bar{u}(x)$. Distributions of $\bar{d}(x)/\bar{u}(x)$, obtained from two separated analyses, one using pure proton data (blue curve with uncertainty band in same colour) and another using deuteron data (red curve with uncertainty band in same colour). The proton data-based analysis incorporates experimental results from HERA, BCDMS, LHC, Tevatron, and Fermilab's E866 (specifically, only the proton-hydrogen measurement). Additionally, measurements used to determine the $R$-parameter are included to constrain the antiquark distributions. In contrast, the deuteron-based analysis replaces the $R$-parameter measurements with experimental results from the NuSea and SeaQuest collaborations. For consistency, all distributions are calculated at the energy scale of 100 GeV. The uncertainties of these distributions are estimated by taking into account both the statistical errors and experimental systematics of the measurements.
• Figure 2: Comparison between the data and predictions. The measured $R$ parameters from the CMS (a) and D0 (b) data, together with the measured $\sigma(pD)/[2\sigma(pH)]$ from the NuSea (c) and SeaQuest (d) data, are shown. The vertical error bars represent the total uncertainties, including both statistical fluctuations and experimental systematics. The corresponding theoretical predictions (in solid or dashed curves) for these observables are provided for comparison. The lower panel in each subfigure displays the uncertainty-normalized difference between the predictions and the measurements, i.e., the difference divided by the standard deviation. The $R$ parameters are presented as a function of $|Y_Z|$, the rapidity of the Drell-Yan pair, while the ratio of $\sigma(pD)/[2\sigma(pH)]$ is presented as a function of $x$.
• Figure 3: Relative distributions of quarks and gluon. The quark and gluon distributions derived from the analysis of the pure proton dataset and that of the deuteron dataset, demonstrated as the ratio with respect to the corresponding predictions from CT18NNLO PDF. For the distributions of $u(x)$, $d(x)$, $\bar{u}$ and $\bar{d}$, the ratios are given in (a) and (b) for the proton data and the deuteron data analysis, respectively. For the distributions of $s(x)=\bar{s}(x)$, $g(x)$ and $c(x)$, the ratios are given in (c) and (d) for the proton data and the deuteron data analysis, respectively.