Coarse grained models of graphene and graphene oxide for use in aqueous solution

Abstract

Obtaining stable aqueous dispersions of graphene-based materials is a major obstacle in the development and widespread use of graphene in nanotechnology. The efficacy of atomistic simulations in obtaining a molecular-level insight into aggregation and exfoliation of graphene/graphene oxide (GO) is hindered by length and time scale limitations. In this work, we developed coarse-grained (CG) models of graphene/GO sheets, compatible with the polarizable Martini water model, using molecular dynamics, iterative Boltzmann inversion and umbrella sampling simulations. The new CG models accurately reproduce graphene/GO–water radial distribution functions and sheet–sheet aggregation free energies for small graphene (−316 kJ mol−1) and GO (−108 kJ mol−1) reference sheets. Deprotonation of carboxylic acid functionalities stabilize the exfoliated state by electrostatic repulsion, providing they are present at sufficiently high surface concentration. The simulations also highlight the pivotal role played by entropy in controlling the propensity for aggregation or exfoliation. The CG models improve the computational efficiency of simulations by an order of magnitude and the framework presented is transferrable to sheets of different sizes and oxygen contents. They can now be used to provide fundamental physical insights into the stability of dispersions and controlled self-assembly, underpinning the computational design of graphene-containing nanomaterials.

The Martini Model in Materials Science

“…Martini models for fullerene, [ 67–69 ] carbon nanotubes (CNTs), [ 70–73 ] graphene [ 74–76 ] or MXene [ 77 ] flakes, clay nanoparticles, [ 78 ] and functionalized nanoparticles [ 79–81 ] have also been developed to study their interaction with both other synthetic or biomolecular systems. The C 60 fullerene model developed by Monticelli represents a good parametrization strategy for nanoparticles.…”

“…Apart from their importance to modeling organic electronics, which are the subject of the next section, they have been used to simulate polymer nanoparticle composites, [ 78,80,95,96,155–165 ] self‐assembly of nanoparticles, [ 79,166–168 ] and solution processing of nanoparticles. [ 76,77,85,86 ]…”

“…For example, Martini as been used to simulate single‐layer coated nanoparticle membranes, [ 166 ] or nanoparticle superlattices for energy applications. [ 168 ] Additionally, the dispersion of CNTs and 2D materials, such as graphene and MXene, in surfactant aqueous solutions [ 76,77,85,86 ] has been investigated. Understanding and optimizing such dispersion is of paramount importance for the processing of these materials; in the case of the 2D material MXene, Li et al.…”

The Martini Model in Materials Science

“…Martini models for fullerene [58][59][60], carbon nanotubes (CNTs) [61][62][63][64], graphene [65][66][67] or MXene [68] flakes, clay nanoparticles [69], and functionalized nanoparticles [70][71][72] have also been developed to study their interaction with both other synthetic or biomolecular systems. The C 60 fullerene model developed by Monticelli represents a good parametrization strategy for nanoparticles.…”

“…Martini parameters are available for a wide variety of nanoparticles including fullerenes [58][59][60], carbonnanotubes [61][62][63][64], and gold [70,71] among others. Apart from their importance to modeling organic electronics, which are subject to the next section, they have been used to simulate polymer nanoparticle composites [69,71,85,86,[130][131][132][133][134][135][136][137][138], self-assembly of nanoparticles [70,[139][140][141], and solution processing of nanoparticles [67,68,75,142].…”