Multiscale Approaches and Perspectives to Modeling Aqueous Electrolytes and Polyelectrolytes - Multiscale Molecular Methods in Applied Chemistry - page 260

3 Classical Models: Detailed-Atomistic and Coarse-Grained
Force Fields for Multiscale Modeling of Electrolyte
and Polyelectrolyte Solutions
In aqueous electrolyte and polyelectrolyte solutions, electrostatic interactions are
strongly screened, owing to the large dielectric constant of the aqueous medium. In
effect, the interaction between two ions in water at room temperature only exceeds
k B T when the distance between them is smaller than 0.7 nm. At this distance the
ions are separated by approximately three water molecules; hence the effective
ionic interaction can be fully understood only when solvent molecules are explicitly
considered. Monovalent and divalent ions polarize molecular dipole orientations of
surrounding water molecules only within their first hydration shells, as indicated
in Fig. 2 . Hence, the electrostatic fields emanating from ions in water are to a large
extent screened beyond the first hydration shell and, therefore, electrostatic inter-
actions between ions in water are expected not to be strong enough to justify that
water-water (hydrogen bonding) interactions can be left out in the description of
inter-ionic interactions and interactions between ions and polyelectrolytes.
Global thermodynamic properties of complex systems are often affected by the
local-scale chemical details. In recent years, experimental and computational
studies of aqueous systems containing charged and uncharged polymers, polypep-
tides, and proteins have provided ample evidence that effects of ions are local,
ion-specific, and involve direct interactions with macromolecules and their first
hydration shells [ 50 - 54 ]. These systems, however, are usually too large to be
described with theoretical models that include of all those details. Local chemical
processes in these systems may moreover depend on global system properties
determined by structural organization or dynamical processes at large time and length
scales, i.e., processes on local and global scales are interdependent. Computer
simulations that use only one type of Hamiltonian (quantum-mechanical, classical
atomistic, or classical coarse-grained) cannot provide a complete understanding of
these complex systems and multiscale simulation approaches are needed instead.
Here, we limit the discussion to hierarchical multiscale simulations. In hierarchical
simulations, quantum-mechanical, classical-atomistic, or classical coarse-grained
Hamiltonians are used consecutively. The different Hamiltonians each describe the
same system with different resolution (different number of degrees of freedom),
which is achieved by “parameter-inheritance” and exchange of configurations by
means of forward and backward mapping procedures [ 55 , 56 ]. Systematic coarse
graining refers to deriving a coarse-grained (CG) model based on molecular simula-
tions with a higher resolution, quantum-mechanical or classical atomistic model [ 56 ].
Linking chemistry and properties in computer simulations has made huge progress
in recent years. However, significant challenges remain which are the subjects of
many studies performed today, where systematic coarse-graining methods are
being further explored. Some recent examples of our own research will be reviewed
here. Section 3.1 presents a short review of ion pairing, a phenomenon which
is determined by chemistry-specific, local details, but which potentially affects