Simulation of self-diffusion of point-like and finite-size tracers in stochastically reconstructed Vycor porous glasses

1999 ◽  
Vol 111 (6) ◽  
pp. 2735-2743 ◽  
Author(s):  
M. E. Kainourgiakis ◽  
E. S. Kikkinides ◽  
A. K. Stubos ◽  
N. K. Kanellopoulos
2019 ◽  
Author(s):  
Thiago José Pinheiro dos Santos ◽  
Charlles Abreu ◽  
Bruno Horta ◽  
Frederico W. Tavares

Mass transport coefficients play an important role in process design and in compositional grading of oil reservoirs. As experimental measurements of these properties can be costly and hazardous, Molecular Dynamics simulations emerge as an alternative approach. In this work, we used Molecular Dynamics to calculate the self-diffusion coefficients of methane/n-hexane mixtures at different conditions, in both liquid and supercritical phases. We evaluated how the finite box size and the choice of the force field affect the calculated properties at high pressures. Results show a strong dependency between self-diffusion and the simulation box size. The Yeh-Hummer analytical correction [J. Phys. Chem. B, 108, 15873 (2004)] can attenuate this effect, but sometimes makes the results depart from experimental data due to issues concerning the force fields. We have also found that different all-atom and united-atom models can produce biased results due to caging effects and to different dihedral configurations of the n-alkane.


2019 ◽  
Author(s):  
Thiago José Pinheiro dos Santos ◽  
Charlles Abreu ◽  
Bruno Horta ◽  
Frederico W. Tavares

Mass transport coefficients play an important role in process design and in compositional grading of oil reservoirs. As experimental measurements of these properties can be costly and hazardous, Molecular Dynamics simulations emerge as an alternative approach. In this work, we used Molecular Dynamics to calculate the self-diffusion coefficients of methane/n-hexane mixtures at different conditions, in both liquid and supercritical phases. We evaluated how the finite box size and the choice of the force field affect the calculated properties at high pressures. Results show a strong dependency between self-diffusion and the simulation box size. The Yeh-Hummer analytical correction [J. Phys. Chem. B, 108, 15873 (2004)] can attenuate this effect, but sometimes makes the results depart from experimental data due to issues concerning the force fields. We have also found that different all-atom and united-atom models can produce biased results due to caging effects and to different dihedral configurations of the n-alkane.


2013 ◽  
Vol 44 (7) ◽  
pp. 827-836 ◽  
Author(s):  
S. Beckert ◽  
M. Gratz ◽  
J. Kullmann ◽  
D. Enke ◽  
F. Stallmach

1983 ◽  
Vol 66 (1) ◽  
pp. 69-72 ◽  
Author(s):  
JORG KARGER ◽  
JORG LENZNER ◽  
HARRY PFEIFER ◽  
HARTMUT SCHWABE ◽  
WOLFGANG HEYER ◽  
...  

1971 ◽  
Vol 26 (1) ◽  
pp. 85-93 ◽  
Author(s):  
J. S. Murday ◽  
R. M. Cotts

Abstract The liquid state self-diffusion coefficients of Li6 and Li7 in isotopically enriched Li6 and Li7 metal and in several isotopic alloys have been measured by the nuclear magnetic resonance spin echo, pulsed magnetic gradient technique. At the melting point 180.5 °C, the self diffusion co­efficients are measured as Dm6= (6.8 + 0.7) · 10-5 cm2/sec for Li6 in 99% Li6, and Dm7 = (5.8 ± 0.6) -10-5 cm2/sec for Li7 in 99.9% Li7. The ratio of measured values (Dm6/Dm7) = 1.18 ± 0.07, which is greater than the square root of the mass ratio, (m7/m6) 1/2 = 1.08. The isotopic ratio of self diffusion coefficients is observed to be less than the ratio of the mutual diffusion coefficients in almost pure Li6 and almost pure Li7, (Dᴍm6/Dᴍm7) = 1.35, and the viscosity ratio, (η7/η6) = 1.44. The dependence of D upon isotopic alloy concentration appears to be relatively weak and linear in all but the very low concentrations. As a test of the experimental method, D was measured in liquid Na, H2O , and D2O . In liquid sodium at the melting point, D = (3.7 + 0.3) • 10-5 cm2/sec. The Na measurement and the values of D in water are about 5% below other published data, but they are in agreement within experimental uncertainties. Results of this experiment contain syste­matic corrections amounting to 15 - 20% caused principally by the finite size of the droplets of the liquid metal samples. The corrections as well as the limitations of the technique in measurement of self diffusion coefficients in metals are discussed


Sign in / Sign up

Export Citation Format

Share Document