Assessing the lattice QCD space diffusion coefficient and the thermalization time of charm quark by mean of D meson observables at LHC

Abstract

A central goal in the study of heavy-flavour production is to determine the interaction strength between Heavy Quarks (HQs) and the Quark-Gluon Plasma (QGP), quantified by the spatial diffusion coefficient $D_s(T)$. Recent lattice QCD (lQCD) results with dynamical fermions suggest a remarkably low value of $2πT D_s \approx 1$ at $T=T_c$ for charm quarks - significantly lower than both quenched QCD estimates and most phenomenological models - which typically yield $2πT D_s \approx 3.5 - 5$. This discrepancy raises the question of whether such a small $D_s(T)$, corresponding to a thermalization time $τ_{th} \approx 1 - 1.5$ fm/c, is compatible with experimental measurements of key observables like the nuclear modification factor $R_{AA}$, the elliptic and triangular flow coefficients $v_2$ and $v_3$ for D mesons. Using an event-by-event Langevin transport framework, we analyze several scenarios and highlight the pivotal role played by the momentum dependence of the drag coefficient $A(p) = τ_{th}^{-1}(p)$. Our findings show that a small $2πT D_s (p\rightarrow 0)\approx 1 - 2$ values can align with experimental data \emph{only} if a significant momentum dependence in $τ_{th}(p)=1/A(p)$ is included, as predicted by T-matrix approaches, or by the extended Quasi-Particle Model (QPMp). In contrast, assuming a momentum-independent $τ_{th} = M_c D_s^{\text{lQCD}} / T$, it fails to reproduce the observed phenomenology. Furthermore, a short thermalization time of $τ_{th} \approx 1.5$ fm/c implies a loss of sensitivity of the final-state observables to the initial charm-quark momentum distribution up $p_T \approx M_c$, suggesting a possible universal behavior driven by a dynamical attractor.