Abstract:We investigate by Molecular-Dynamics simulations the fast mobility - the rattling amplitude of the particles temporarily trapped by the cage of the neighbors - in mildly supercooled states of dense molecular (linear trimers) and atomic (binary mixtures) liquids. The mixture particles interact by the Lennard-Jones potential. The non-bonded particles of the molecular system are coupled by the more general Mie potential with variable We investigate by Molecular-Dynamics simulations the fast mobility - the rattling amplitude of the particles temporarily trapped by the cage of the neighbors - in mildly supercooled states of dense molecular (linear trimers) and atomic (binary mixtures) liquids. The mixture particles interact by the Lennard-Jones potential. The non-bonded particles of the molecular system are coupled by the more general Mie potential with variable repulsive and attractive exponents in a range which is characteristic of small $n$-alkanes and $n$-alcohols. Possible links between the fast mobility and the geometry of the cage (size and shape) are searched. The correlations on a per-particle basis are rather weak. Instead, if one groups either the particles in fast-mobility subsets or the cages in geometric subsets, the increase of the fast mobility with both the size and the asphericity of the cage is revealed. The observed correlations are weak and differ in states with equal relaxation time. Local forces between a tagged particle and the first-neighbor shell do not correlate with the fast mobility in the molecular liquid. It is concluded that the cage geometry alone is unable to provide a microscopic interpretation of the known, universal link between the fast mobility and the slow structural relaxation. We suggest that the particle fast dynamics is affected by regions beyond the first neighbours, thus supporting the presence of collective, extended fast modes.

A mapping space Odyssey: characterising the statistical and metric properties of reduced representations of macromolecules

Learning Effective Molecular Models from Experimental Observables.

Analysis of mapping atomic models to coarse-grained resolution

Surveying the energy landscape of coarse-grained mappings

Mapping transiently formed and sparsely populated conformations on a complex energy landscape

Reducing Entanglement With Physically-Inspired Fermion-To-Qubit Mappings

Comparing molecules and solids across structural and alchemical space

A Deep Graph Network–Enhanced Sampling Approach to Efficiently Explore the Space of Reduced Representations of Proteins

Persistent Dirac for molecular representation

EXCOGITO, an extensible coarse-graining toolbox for the investigation of biomolecules by means of low-resolution representation

Condensed-phase molecular representation to link structure and thermodynamics in molecular dynamics

A review of mathematical representations of biomolecules

Configurational mapping significantly increases the efficiency of solid-solid phase coexistence calculations via molecular dynamics: Determining the FCC-HCP coexistence line of Lennard-Jones particles

Inherent Structure Versus Geometric Metric For State Space Discretization

Unbiased Comparison between Theoretical and Experimental Molecular Structures and Properties: Toward an Accurate Reduced-Cost Evaluation of Vibrational Contributions

Weak links between fast mobility and local structure in molecular and atomic liquids

Local invertibility and sensitivity of atomic structure-feature mappings

Mapping quantum geometry and quantum phase transitions to real space by a fidelity marker

Computing committors via Mahalanobis diffusion maps with enhanced sampling data

A physics-inspired approach to the understanding of molecular representations and models

A global view of reactive coordinate manifolds from nonlinear dimensionality reduction