Chemistry Reference
In-Depth Information
around 1.1. This represents a difference in free energy of retention of 0.2 kJ/mol, an
extremely small quantity to measure in either experiment or simulation. Addition-
ally, this agreement with experiment allows for much greater confidence in the
molecular details from simulation results.
Analysis of
K
(
z
) profiles and of the orientational distribution for the four-ring
PAHs shows that the analyte molecules penetrate deeply into the bonded-phase
region (the maxima in
K
(
z
) are typically found at
z
10
˚
) where the analytes
exhibit an increasing preference to align perpendicular to the surface with increas-
ing ODS coverage [
40
]. The orientational preference is most pronounced for NAP,
the most retained compound. Analysis of the lateral distribution of the analytes at
the highest ODS coverage (where the largest selectivities are observed) indicates
that NAP is preferentially found in locations that are relatively more crowded
with ODS chains, whereas BcP exhibits a preference for relatively less crowded
regions [
40
].
From a calculation of the density of ODS methylene segments for configurations
without an analyte molecule being present, it is evident that the upper part of the
bonded phase resembles a liquid phase without pre-existing cavities [
40
]. Further-
more, it can be shown that the chain structure is modified by the presence of the
PAH analytes. For all coverages, values of cos
y
ete
and
S
increase by about 0.1 for
ODS chains near to any of the four-ring PAH analytes compared ODS chains
without any analyte being present, i.e., the conformational flexibility of the ODS
chains allows them to respond to the presence of the analyte and the favorable
regions do not correspond to static cavities.
5 Conclusions
The GEMC/CBMC simulation methodology applied here has proved very useful
for studying structure and retention in complex GLC and RPLC systems. The
methodology affords the computation of retention data with sufficient precision
(and accuracy for the TraPPE force field) for complex analytes to validate the
predictions against experimental data. This validation, in turn, builds confidence in
the molecular-level details on structure and retention mechanism that can be
obtained from the simulations.
In closing, we would like to mention some applications of the GEMC/CBMC
approach and very much related combination of CBMC and the grand canonical
Monte Carlo technique to other complex systems: prediction of structure and
transfer free energies into dry and water-saturated 1-octanol [
72
], prediction of
the solubility of polymers in supercritical carbon dioxide [
73
], prediction of the
upper critical solution pressure for gas-expanded liquids [
74
], investigation of the
formation of multiple hydrates for a pharmaceutical compound [
75
], exploration
of multicomponent vapor-to-particle nucleation pathways [
76
], and investigations
of the adsorption of articulated molecules in zeolites and metal organic frameworks
[
77
,
78
].
Search WWH ::
Custom Search