Inferring the Surface Energy Distribution of Low Dimensional Materials

Abstract

Two dimensional nanomaterials exhibit novel and superlative properties, as such; the last decade has seen huge research interest in 2D nanomaterials. To fully exploit these properties it is essential to have an accurate description of them. One such proper ty is the surface energy of a material. It directly influences the interaction the material has with its environment. Since 2D nanomaterials are by definition dominated by their surface, the surface energy is an exceptionally important property. An example of this is the liquid phase exfoliation of 2D nanomaterials which relies on matching the surface energy of the solvent and the material. Despite its importance it remains poorly understood, often quoted as single value. What is needed is a complete descri ption of a material’s surface energy, one that takes into account the heterogeneity of a material. This thesis establishes an approach that can be used to infer the surface energy distribution and assign surface energies to the different aspects of a mate rial, in the case of layered materials these are the edges, basal plane and the defects in the basal plane. This work takes advantage of the ability of Finite - Dilution Inverse Gas Chromatography to measure surface energy over a wide range of surface covera ges. This allows it to probe the high and low surface energy sites of the material. This data is plotted in what is known as a surface energy profile. The first section of this work focussed on examining the relationship between the surface energy distrib ution and the surface energy profile. The profiles are fitted to a stretched exponential function in order to quantitatively describe them and objectively study alterations in their shape resulting from changes to the surface energy distribution. From the fits it was possible to extract the following descriptive parameters; surface energy at zero coverage ( γ d,φ=0 ) , surface energy at full coverage ( γ d,φ=1 ) , and the decay constant (φ 0 ). To systematically study the dependence of these parameters on the surface energy distribution this work employs a method developed by Smith et al. to simulate the surface energy profile produced by a given surface energy distribution. Starting with the simplest distributions consisting of only a single Gaussian curve two impor tant relations were discovered. First, the surface energy at full coverage was found to equal the mean value of the distribution. Second, the difference between γ d,φ=0 ii and γ d,φ=1 was found to depend mainly on the standard deviation of the Gaussian curve. T he complexity of distributions was increased by introducing a second and third Gaussian curve. Simulating the profiles for a huge array of distributions it was found that most distributions do not produce profiles of the exponential - like shape seen experim entally. Fortunately three types of distributions were identified as being able to produce exponential - like profiles. It is demonstrated that using the behaviour of the γ d,φ=0 and γ d,φ=1 it is possible to identify the type of distribution, infer the surfac e energy distribution and assign surface energies to each aspect of a material. The method was applied to molybdenum disulphide (MoS 2 ) and boron nitride. All aspects of MoS 2 were found to have a mean surface energy of ~ 40 mJ/m 2 . While the defects of MoS 2 have a low mean surface energy, they cover a wide range of energies up to 100 mJ/m 2 . On the other hand the edges were found to have a narrow range of energies and behave as low energy sites. Modelling the flake as rhombuses, two simple models were develop ed relating the specific surface area and the decay constant, φ 0 , to the mean flake length. The results from these models support the conclusions regarding the surface energy of the defects and the edges. The results also indicate that the defect density o f MoS 2 is dependent on flake length; this is supported by an example from the literature. A slightly different conclusion is drawn regarding boron nitride. In this case, the edges are found to have a mean value of ~ 70 mJ/m 2 and cover a wide range of surfac e energy. The defects and basal plane sites are found to be indistinguishable; both have a surface energy with a mean value of ~ 40 mJ/m 2 and a narrow range of energies. It is hoped that this work demonstrates the utility of the method described and that i t will be applied to materials outside of the field of nanoscience. Another potential avenue for further research could be the use of lasers to induce defects in the basal plane of 2D materials and study the change in surface energy.