Dimensionality dependence of diffusion-entropy scaling: Sensitivity to the diffusion mechanism

Abstract: While entropy quantifies the volume of the accessible phase space, diffusion characterizes the rate of its exploration, capturing distinct yet interconnected aspects of a system's dynamics. In this Letter, we employ computer simulations to independently compute D and S for Lennard-Jones (LJ) liquid and water in two and three dimensions, and for water also in one dimension, across a broad range of thermodynamic states. We observe that the ratio of diffusion coefficients between two states exhibits a nearly perfect exponential dependence on their entropy difference. For LJ liquids, the prefactor of the exponential shows a strong dimensionality dependence, consistent in trend but quantitatively different from theoretical predictions. In contrast, water displays a remarkably weak dimensionality dependence, deviating from theoretical expectations, which we attribute to the dominant role of jump diffusion. Surprisingly, the exponential diffusion-entropy relationship persists even when translational and rotational contributions to entropy are considered separately, underscoring the robustness of the D-S relation across different degrees of freedom.