Wetting and Heat Transfer in Graphene-Based Composites - Multiscale Molecular Simulations

Interfaces are ubiquitous at the nanoscale in a range of materials and typically play a key role in determining the composite material properties. Understanding the thermodynamics and heat transfer across interfaces is a crucial challenge in composite material design and engineering. For instance, graphene-based polymer composites have been hyped to render orders of magnitude enhancement in thermal properties but have not yet lived up to the promise, primarily due to the thermal resistance arising at the interface. Simulating these systems certainly provides a way to gain valuable insights into physics at the interface and can answer some outstanding questions. However, simulating these systems is non-trivial and offers two major hurdles. The first issue is that the interface must be properly represented using suitable models to perform a realistic comparison with the experiments. The second problem is that of the scale, i.e., if one aims to model these complex systems at the highest resolution with all the chemical details inserted, the relevant physical problems often span length and time scales that are out of reach of the state of the art computational power. We address the above two problems in this thesis, i.e., how to represent interactions between various species faithfully and two, how to derive coarse-grain models accordingly. Furthermore, we also obtain a detailed understanding of what changes coarse-graining brings to the solid-liquid interfaces. The current thesis aims to obtain a fundamental understanding of the solid-liquid interfaces and to propose a framework that allows simulation of large systems of particle networks surrounded by a fluid component. To this end, we first address the nature of interfacial thermodynamics by studying solid-liquid interfaces of graphene-based systems. Then we extend this study to coarse-grained models of liquids to study in detail how the interfacial behavior is modified on coarse-graining. We primarily observe that the entropic part of the free energy is not accurately reproduced up on coarse-graining. We have learnt that, to accurately represent the interfaces, one needs to reproduce both the interaction energy and its fluctuations. Using these insights, we propose an approach to simulate graphene particles at the water-oil interface. By relying on the macroscopic wetting coefficient as the relevant wetting parameter, a simple approach to derive solid-liquid interaction parameters was proposed and tested. By simulating at various values of the wetting coefficient, we were able to confirm that the regime, for which the particles adsorb, is consistent with the macroscopic predictions. It is further shown that this approach works equally well for atomistic and coarse-grained models of liquids. In the next step, a detailed study on the influence of water coarse-graining on the interfacial heat transfer across water-graphitic interfaces has been conducted. By computing the interfacial thermal conductance (G), we evaluated the differences in heat transfer for atomistic and coarse-grained water at different strengths of wetting ranging from hydrophobic to mildly hydrophilic. Quite surprisingly, we observe that a coarse-grained model of water is sufficient to compute G. This important insight means that, heat transfer across solid-liquid interface is mainly in the low frequency regime and removing high frequency modes (or degrees of freedom) by coarse-graining has only a nominal influence. The above insights will enable us to model large scale systems (up to 100nm) of graphene at water-oil interfaces and further will allow us to characterize heat transfer across networks of particles. The computational studies performed here will further the efforts to develop approaches in multiscale modeling and will assist in addressing outstanding questions in the areas of heat transfer and selfassembly in polymer nanocomposites