38
A. Jayaraman et al.
1 Introduction
This chapter presents our recent development in coarse-grained models for investigating structure and dynamics in synthetic and biological polymers with chemistries
that have specific and directional molecular interactions. There is significant motivation to study synthetic polymers with specific and directional interactions due to
the many desirable features these interactions provide when designing novel materials. Soft materials with specific and directional interactions such as hydrogen bonds
or h-bonds, in short, can have (a) thermo-reversible phase behavior (e.g., enthalpically driven morphologies with intact h-bonds at low temperatures and entropically
driven morphologies with loss of h-bonds at high temperatures), (b) precisely tuned
nanostructure with desirable geometries (e.g., programmable coordination numbers
in the assembled state depending on the valency of the directional interaction) and
(c) well-mixed/blended morphologies that are not possible without the stabilization
provided by h-bonds within the material [1–5]. Similarly, we can find many examples
of biological polymers/macromolecules (e.g., proteins and DNA) with specific and
directional interactions, in many cases based on h-bonds, that stabilize secondary and
tertiary structures of these macromolecules with defined function and properties.
While computational studies have been tremendously useful in understanding
molecular phenomena and/or guiding synthesis of new polymers for a wide variety
of applications, the inability to capture small scale, specific and directional interactions alongside macromolecular length and time scales represents a key limitation of
most computational studies to date. Theory and simulations of small and/or macromolecular materials with h-bond interactions have been largely limited to atomistic simulations. Atomistic simulations represent the donor and acceptor atoms
and their partial charges explicitly to reproduce the electrostatic nature of the hbond at the atomic level. In particular, in the case of biomacromolecules (e.g.,
proteins, DNA, RNA), these atomistic approaches have been useful in simulating
stable secondary and tertiary structures stabilized by h-bonds, among other interactions (e.g., hydrophobic interactions). Additionally, these atomistic simulations with
explicit solvents also capture the h-bonds between biomacromolecules and water.
Despite the chemically realistic representation of the h-bonds, due to the large computational time associated with these atomistic simulations, one cannot expect to see
large time scale or length scale rearrangements or ordering (from disordered state)
in the macromolecular system simulated atomistically. In fact, in most biomolecular atomistic simulations, the initial configuration for the simulation is an ordered
secondary/tertiary structure that is obtained from the data bank of crystal/solution
structures, and during the course of the (unbiased) simulation one can observe only
small structural rearrangements (e.g., in the order of 10–50 atoms). In soft materials where we intend to use computational tools to predict unknown assembled
structures from an initially unassembled state, disorder-order transition and/or large
cluster formation from small building blocks, all driven through dynamic h-bond
formation and atomistic simulations alone are infeasible.
A. Jayaraman et al.
1 Introduction
This chapter presents our recent development in coarse-grained models for investigating structure and dynamics in synthetic and biological polymers with chemistries
that have specific and directional molecular interactions. There is significant motivation to study synthetic polymers with specific and directional interactions due to
the many desirable features these interactions provide when designing novel materials. Soft materials with specific and directional interactions such as hydrogen bonds
or h-bonds, in short, can have (a) thermo-reversible phase behavior (e.g., enthalpically driven morphologies with intact h-bonds at low temperatures and entropically
driven morphologies with loss of h-bonds at high temperatures), (b) precisely tuned
nanostructure with desirable geometries (e.g., programmable coordination numbers
in the assembled state depending on the valency of the directional interaction) and
(c) well-mixed/blended morphologies that are not possible without the stabilization
provided by h-bonds within the material [1–5]. Similarly, we can find many examples
of biological polymers/macromolecules (e.g., proteins and DNA) with specific and
directional interactions, in many cases based on h-bonds, that stabilize secondary and
tertiary structures of these macromolecules with defined function and properties.
While computational studies have been tremendously useful in understanding
molecular phenomena and/or guiding synthesis of new polymers for a wide variety
of applications, the inability to capture small scale, specific and directional interactions alongside macromolecular length and time scales represents a key limitation of
most computational studies to date. Theory and simulations of small and/or macromolecular materials with h-bond interactions have been largely limited to atomistic simulations. Atomistic simulations represent the donor and acceptor atoms
and their partial charges explicitly to reproduce the electrostatic nature of the hbond at the atomic level. In particular, in the case of biomacromolecules (e.g.,
proteins, DNA, RNA), these atomistic approaches have been useful in simulating
stable secondary and tertiary structures stabilized by h-bonds, among other interactions (e.g., hydrophobic interactions). Additionally, these atomistic simulations with
explicit solvents also capture the h-bonds between biomacromolecules and water.
Despite the chemically realistic representation of the h-bonds, due to the large computational time associated with these atomistic simulations, one cannot expect to see
large time scale or length scale rearrangements or ordering (from disordered state)
in the macromolecular system simulated atomistically. In fact, in most biomolecular atomistic simulations, the initial configuration for the simulation is an ordered
secondary/tertiary structure that is obtained from the data bank of crystal/solution
structures, and during the course of the (unbiased) simulation one can observe only
small structural rearrangements (e.g., in the order of 10–50 atoms). In soft materials where we intend to use computational tools to predict unknown assembled
structures from an initially unassembled state, disorder-order transition and/or large
cluster formation from small building blocks, all driven through dynamic h-bond
formation and atomistic simulations alone are infeasible.