The interaction of electromagnetic radiation with matter has led to a large number of interesting applications. The propagation of electromagnetic waves within materials is described by Maxwell's equations. However, the fundamental understanding of the causes of the response of the material, defined by constitutive relations for its complex, frequency-dependent dielectric constant, can only be achieved through the study of processes occurring at the molecular scale. The fluctuation-dissipation theorem relates the frequency-dependent dielectric constant of a material to equilibrium fluctuations in its dipole moment. This fact can be used to determine dielectric properties from equilibrium molecular dynamics simulations for frequencies covering the microwave region of the electromagnetic spectrum (300 MHz - 300 GHz). In this work, the ability of current force fields to predict dielectric spectra of one component systems and mixtures is examined, showing accurate results when compared with experimental data for the systems under consideration. Additionally, the influence of temperature on the dielectric spectra is analysed, yielding equally satisfactory results. In the particular case of ethanol/water mixtures, the estimation of dielectric spectra at intermediate concentrations using molecular dynamics simulations outperforms the traditional use of mixing rules. The simulations of these systems reveal the importance of collaborative processes between groups of molecules, such as hydrogen bond networks, in the overall dielectric response. The reduction of the contribution of these processes as temperature increases confirms the weakening of these networks at high temperatures.The predicted dielectric properties are used in a heating model to estimate temperature profiles in microwave heating processes. Unexpected results are obtained which reveal the need for accurate determination of the electric field distribution within the workload in order to obtain representative heating profiles. In contrast, penetration depths are accurately determined from dielectric properties generated through molecular simulations.
|Date of Award||17 Dec 2016|
- University Of Strathclyde
|Supervisor||Leo Lue (Supervisor) & Martin Sweatman (Supervisor)|