Login / Signup

Finite-size excess-entropy scaling for simple liquids.

Mauricio SevillaAtreyee BanerjeeRobinson Cortes-Huerto
Published in: The Journal of chemical physics (2023)
Explicit and implicit size effects in computer simulations result from considering systems with a fixed number of particles and periodic boundary conditions, respectively. We investigate these effects in the relation D*(L) = A(L) exp(α(L)s2(L)) between reduced self-diffusion coefficient D*(L) and two-body excess entropy s2(L) for prototypical simple-liquid systems of linear size L. To this aim, we introduce and validate a finite-size two-body excess entropy integral equation. Our analytical arguments and simulation results show that s2(L) exhibits a linear scaling with 1/L. Since D*(L) displays a similar behavior, we show that the parameters A(L) and α(L) are also linearly proportional to 1/L. By extrapolating to the thermodynamic limit, we report the coefficients A∞ = 0.048 ± 0.001 and α∞ = 1.000 ± 0.013 that agree well with the universal values available in the literature [M. Dzugutov, Nature 381, 137-139 (1996)]. Finally, we find a power law relation between the scaling coefficients for D*(L) and s2(L), suggesting a constant viscosity-to-entropy ratio.
Keyphrases
  • systematic review
  • magnetic resonance imaging
  • magnetic resonance
  • machine learning
  • diffusion weighted imaging
  • ionic liquid
  • mass spectrometry
  • neural network