Material Composition Analysis Using Time-Domain Terahertz Spectroscopy

In this work, we apply terahertz time-domain spectroscopy to measure the thickness and composition of dielectric samples consisting of multiple constituents. We accomplish this in two stages. First, a two-component model is introduced that allows us to estimate the volume fractions of two constituents in addition to the sample thickness. Here the sample is assumed to be a homogeneous mixture, with its permittivity given by the Bruggeman effective medium theory (EMT). We apply the procedure to paper samples to extract thickness and moisture content. We perform Monte-Carlo simulations with a realistic noise model to understand the parameter correlations and predict the expected uncertainties. The simulation results are consistent with the observed repeatability measurements performed on paper samples. We show that our measurements can compete with existing thickness and moisture sensors currently deployed in the industry. Second, a three-component model is presented to determine the volume fractions of three constituents, along with the thickness. We also develop two different scattering models to account for the effects of scattering from inhomogeneities. We validate our model by measuring pulp samples compressed to different thicknesses thus having variable densities. Again, our measurements match well with the independently measured parameter values. Monte-Carlo simulations suggest avenues of improvement of the performance of the system.