10.1021/jp044257l.s001
R. Krishna
R.
Krishna
J. M. van Baten
J.
M. van Baten
Diffusion of Alkane Mixtures in Zeolites: Validating the Maxwell−Stefan Formulation
Using MD Simulations
American Chemical Society
2005
diffusivitie
ternary mixture data
MFI
zeolite
D i
component parameters Đ i
occupancy
MD Simulations Molecular dynamics
LTA
component sorption isotherms
saturation capacities ϑ i
exchange Đ ij
CBMC
exchange Đ ii
FAU
exchange coefficient Đ ij
species i
2005-04-07 00:00:00
Journal contribution
https://acs.figshare.com/articles/journal_contribution/Diffusion_of_Alkane_Mixtures_in_Zeolites_Validating_the_Maxwell_Stefan_Formulation_Using_MD_Simulations/3292276
Molecular dynamics (MD) simulations have been carried out for pure components, binary, ternary, and
quaternary mixtures containing methane, ethane, propane, and <i>n</i>-butane in FAU zeolite at 300 K for a range
of molecular loadings ϑ, approaching saturation limits. The <i>n</i>-dimensional matrix of Maxwell−Stefan
(M−S) diffusivities [Δ], defined by (<b>N</b>) = −ρ[Δ][Γ](∇ϑ), was determined along with the <i>self</i>-diffusivities,
<i>D</i><i><sub>i</sub></i><sub>,self</sub>. Additionally, configurational-bias Monte Carlo (CBMC) simulations were carried out to obtain the
pure component sorption isotherms and the saturation capacities ϑ<i><sub>i</sub></i><sub>,sat</sub>. From the information on Δ<i><sub>ij</sub></i>, <i>D</i><i><sub>i</sub></i><sub>,self</sub>,
and ϑ<i><sub>i</sub></i><sub>,sat</sub>, the various M−S diffusivities were determined: (1) component Đ<i><sub>i</sub></i>, reflecting the interactions of
the species <i>i</i> with the zeolite, <i>self</i>-exchange Đ<i><sub>ii</sub></i>, and (2) <i>binary</i> exchange Đ<i><sub>ij</sub></i>. The obtained data underline the
major advantage of the M−S formulation that at a given occupancy, θ =
ϑ<i><sub>i</sub></i>/ϑ<i><sub>i</sub></i><sub>,sat</sub> within the zeolite, the
Đ<i><sub>i</sub></i> has nearly the same value for species <i>i</i> whether this species is present on its own or in a mixture with
other species. The same advantage holds, too, for the self-exchange Đ<i><sub>ii</sub></i>; the value at a given occupancy, θ,
is the same whether determined from pure component, binary, or ternary mixture data. For all binary and
ternary mixtures studied, it was verified that the binary exchange coefficient Đ<i><sub>ij</sub></i> can be interpolated from the
corresponding values of the self-exchange parameters Đ<i><sub>ii</sub></i> and Đ<i><sub>jj</sub></i> using a generalization of the interpolation
formula developed earlier (Skoulidas et al., <i>Langmuir,</i> <b>2003, </b><i>19</i>, 7977). We also demonstrate that if the
occupancy dependence of the <i>pure</i> component parameters Đ<i><sub>i</sub></i> and Đ<i><sub>ii</sub></i> are modeled properly, this information
is sufficient to provide very good estimates of the matrix [Δ] for mixtures with 2, 3, or 4 components over
the entire range of loadings. Simulations of mixture diffusion of alkanes in MFI and LTA confirm that the
above-mentioned advantages of the M−S formulation also hold for these zeolite topologies.