Transversity and T-odd PDFs from Drell-Yan processes with $pp$, $pD$ and $DD$ collisions

TL;DR

The paper analyzes Drell-Yan processes with polarized protons and deuterons to access transversity and T-odd PDFs (Sivers and Boer-Mulders). It develops a weighted-SSA formalism that isolates PDFs in favorable kinematic regimes and provides quantitative estimates across RHIC, NICA, COMPASS, and J-PARC kinematics, including two transversity evolution models. A new polarized Drell-Yan event generator is introduced to realistically assess SSA feasibility and guide detector simulations, showing measurable asymmetries (order 5–10%) in collider modes and identifying when polarized beams are essential in fixed-target setups. Overall, the work argues that symmetric colliders offer the strongest pathway to simultaneously probe Sivers, Boer-Mulders, and transversity PDFs, with practical feasibility aided by the new generator for experimental planning.

Abstract

We estimate the single-spin asymmetries (SSA) which provide the access to transversity as well as to Boer-Mulders and Sivers PDFs via investigation of the single-polarized Drell-Yan (DY) processes with $ pp$, $pD$ and $DD$ collisions available to RHIC, NICA, COMPASS, and J-PARC. The feasibility of these SSA is studied with the new generator of polarized DY events. The performed estimations demonstrate that there exist the such kinematical regions where SSA are presumably measurable. The most useful for PDFs extraction are the limiting kinematical ranges, where one can neglect the sea PDFs contributions which occur at large values of Bjorken x. It is of interest that on the contrary to the Sivers PDF, the transversity PDF is presumably accessible only in the especial kinematical region. On the contrary to the option with the symmetric collider mode (RHIC, NICA), this is of importance for the COMPASS experiment and the future J-PARC facility where the fixed target mode is available.

Figures (10)

• Figure 1: Estimation of SSA $A_{UT}^{\sin(\phi-\phi_S)\frac{q_T}{M_N}}{\Bigl |}_{pp^\uparrow}$ for NICA, s=400$GeV^2$, with $Q^2=4\,GeV^2$ (left) and $Q^2=15\,GeV^2$ (right). Rome numbers $I,II$ denote respectively fits I and II from Ref. efremov_old and $III$ denotes the fit from Ref. efremov_new.
• Figure 2: Results of Ref. ansel_kotz on $h_{1u}$ in comparison with the results obtained with two versions of evolution model and with the Soffer bound. The bold solid line (top) corresponds to upper bound given by the Soffer inequality. Dashed line corresponds to the evolution model with the Soffer inequality saturation at the initial model scale $Q_0^2=0.23\,GeV^2$. The solid line corresponds to the upper boundary of the error band on $h_{1u}$. The dashed-dotted line corresponds to the evolution model, where $h_{1u,\bar{u}}=\Delta u(\Delta\bar{u})$ at initial scale $Q_0^2=0.23\,GeV^2$. The dotted line corresponds to the fit of Ref. ansel_kotz on $h_{1u}$. GRV94 grv94 parametrization for $q(x)$ and GRSV95 GRSV95 parametrization for $\Delta q(x)$ are used.
• Figure 3: Estimation of SSA $A_{UT}^{\sin(\phi+\phi_S)\frac{q_T}{M_N}}{\Bigl |}_{pp^\uparrow}$ for RHIC, $s=200^2\,GeV^2$, with $Q^2=4\,GeV^2$ (left) and $Q^2=20\,GeV^2$ (right). The solid and dotted curves correspond to the two different input ansatzes for $h_{1u}$ which are used in evolution model. These are $h_{1q,\bar{q}}=\Delta q,\bar{q}$ and $h_{1q}=(\Delta q+q)/2$$h_{1\bar{q}}=(\Delta \bar{q}+\bar{q})/2$, respectively. Here GRV94 grv94 parametrization for $q(x)$ and GRSV95 GRSV95 parametrization for $\Delta q(x)$ are used.
• Figure 4: Estimation of SSA $A_{UT}^{\sin(\phi+\phi_S)\frac{q_T}{M_N}}{\Bigl |}_{pp^\uparrow}$ for NICA, $s=400\,GeV^2$, with $Q^2=4\,GeV^2$ (left) and $Q^2=15\,GeV^2$ (right). The solid and dotted curves correspond to the two different input ansatzes for $h_{1u}$ which are used in evolution model. These are $h_{1q,\bar{q}}=\Delta q,\bar{q}$ and $h_{1q}=(\Delta q+q)/2$$h_{1\bar{q}}=(\Delta \bar{q}+\bar{q})/2$, respectively. Here GRV94 grv94 parametrization for $q(x)$ and GRSV95 GRSV95 parametrization for $\Delta q(x)$ are used.
• Figure 5: Estimation of ratio $R=A_{UT}^{\sin(\phi-\phi_S)\frac{q_T}{M_N}}{\Bigl |}_{Dp^\uparrow}/A_{UT}^{\sin(\phi-\phi_S)\frac{q_T}{M_N}}{\Bigl |}_{pp^\uparrow}$ for NICA kinematics with $Q^2=4\,GeV^2$ (left) and $Q^2=15\,GeV^2$ (right).