Chemistry Reference
In-Depth Information
HOW TO CARRY OUT YOUR OWN IONIC LIQUID
SIMULATIONS
Can anybody perform an atomistic simulation of an ionic liquid? The
answer to this question is a qualified ''yes,'' but it calls to mind the old joke
that just because you can buy a knife, it does not mean you can be a surgeon.
Likewise, just because someone can install a MD or MC software program
does not mean that they will be able to obtain meaningful results by running
the code. Like any aspect of science, be it experimental or computational,
there is a learning curve that must be overcome before you get to the point
where you can obtain meaningful results. In the ''old days'' when this author
began molecular modeling research, there were only a few topics available and
most of the ''tricks of the trade'' were passed on by word of mouth. The
Reviews in Computational Chemistry
series had just begun in response to
this situation, and the Internet consisted of ARCHIE and Gopher searches.
Therefore, if you wanted to do molecular modeling you bought a topic such
as
Computer Simulation of Liquids
by Allen and Tildesley
99
and wrote your
own code on an expensive (and therefore, usually shared) Unix workstation. In
other words, the activation barrier for modeling was high, and those who paid
the energy penalty were usually pretty sophisticated programmers and mode-
lers by the time they had completed a simulation. Thanks to the prevalence of
commercial and open-source software, graphical user interfaces, and instant
downloads, a novice can set up and start running a simulation in an afternoon,
while 15 years ago it might have taken several months just to get started. This
development is a good thing, but it brings forth the warning that care must be
taken so that the results being generated are meaningful and interpreted cor-
rectly. In this section, some guidelines and tips are provided for those new to
modeling and who want to conduct classical simulations of ionic liquids.
The first question a new user needs to answer is: What do I want to get
out of a simulation? The answer to this question really dictates the type and
level of sophistication needed for the simulation. If you are interested in liquid
structure or basic volumetric properties, either MD or MC methods will do
just fine. If you need information on dynamics, then only MD will do. If
you are more interested in thermodynamic phase behavior, however, MC is
the best tool to use. These statements seem benign, but if you are new to mod-
eling you should also be aware that many modeling researchers are passionate
advocates of either MD or MC, but not both. They tend to use only one of
these methods and reject the other method, probably because they are so
heavily invested in the one method they use. The arguments about which
method is ''best'' brings to mind (to this author at least) the fervor with which
people argue over whether ''Windows'' or ''Mac'' is better. Our group is
agnostic about which simulation method is better, and so we choose the right
tool for the problem at hand. We think you should do the same.
