ABSTRACT
Over the past decades, the associated research on metal-organic
frameworks (MOFs) has been rapidly developed into one of the
most prolific areas in chemistry, chemical engineering, and material
science [1-5]. The intrinsic nature of MOFs allows the ability
of systematically modulating the pore dimensions, surface areas,
topologies and surface chemistry within them in an extremely
broad range. This remarkable feature of MOFs is largely absent in
conventional porous materials such as zeolites [6], which permits
MOFs to serve as an ideal platform for various specific targets
and thus puts them in a unique position. For this reason, one of
significant efforts has been devoted in recent years to explore and
improve the performance of MOFs for gas separation [5]. From
an experimental point of view, it is not only very time consuming
to explore the thousands of MOFs reported in the literature, but
also the direct measurements on the related behaviors of mixtures
confined in MOFs remain challenging. In addition, highly detailed
information that leads to the macroscopic properties observed
in MOFs is as well not easy to be addressed using experimental
methods. In contrast, molecular modeling provides a very valuable
complement to experimental studies of MOFs, which can give
deep insights into the mechanisms that control the gas separation
capability of MOFs at amolecular level. Comparedwith experiments,
this method can be utilized to isolate the key influencing factors and
quantify their separate contributions to the separation behaviors of
a large population of MOFs [7], as well as the cooperative effects
among a set of selected factors [8]. Furthermore, due to a variety
of inorganic and organic moieties, it offers a theoretically unlimited
number of possible structures of MOFs. Molecular modeling is then
a particularly attractive tool for large-scale virtual screening of the
existing MOFs and new hypothetical MOFs designed by computer [7,
9-11], allowing experimental endeavors to only concentrate on the
candidates with the most promising separation performance.
