Reflectivity measurements of water and dioxane mixtures using a 100 GHz Gunn diode source

Terahertz (THz) sensing has shown potential as a novel imaging modality in medical applications due to its high water sensitivity. The design of medical THz sensing systems and their successful application to in vivo settings has attracted recent interest to the field, and highlighted the need for improved understanding of the interaction of THz waves with biological tissues. This paper explores the modeling of composite materials which combine strongly-interacting water with weakly-interacting species such as those that are common to biological tissues. The Bruggeman, Maxwell-Garnett, and power law effective media models are introduced and discussed. A reflection-mode 100 GHz Gunn diode sensing system was used to measure the reflectivity of solutions of water and dioxane as a function of relative concentration, and the results were compared with the predictions of the Maxwell-Garnett, power law, and Bruggeman mixing theories. The Maxwell-Garnett model fit poorly to experimental data on near-equal mixtures of water and dioxane and improved when the concentration of water exceeded ∼55% or was below ∼15%. The first-order power law model fit poorly to experimental data across the entire range except at nearpure solutions. Power law models employing 1/2 and 1/3 terms improved goodness of fit, but did not match the accuracy of the Bruggeman model. The Bruggeman provided the best fit to experimental data model as compared to Maxwell-Garnett and the power models and accurately predicted the solution reflectivity through the whole range of concentrations. This analysis suggests that the Bruggeman model may offer improved accuracy over more conventional dielectric mixing models when developing simulation tools for THz reflectometry of hydrated biological tissues.