Date of Award
Spring 2023
Document Type
Dissertation
Degree Name
Doctor of Philosophy (PhD)
Department
Chemistry
First Advisor
Brudvig, Gary
Abstract
Terahertz (THz) spectroscopy probes frequencies between the infrared and microwave regions of the electromagnetic spectrum. A wide range of phenomena leave signatures in this region, including excitons, phonons, rotations of gas molecules, and free carriers. The last of these is of particular interest, as it allows THz spectroscopy to serve as a non-contact probe of conductivity. This can provide particularly rich information when THz spectroscopy is combined with ultrafast visible photoexcitation, to measure transient photoconductivity with sub-picosecond resolution. This is particularly useful in characterizing materials for solar energy applications. Data analysis is often a key obstacle in obtaining the insights promised by THz spectroscopy. In some cases, it is sufficient to report the absorption coefficient or similar quantities. However, more often we would like to use the THz measurements to access properties like the dielectic or photoconductivity spectrum. Such quantitative spectral treatments require more involved modeling of the sample. The rigors of this modeling depend primarily on the sample geometry. For simple sample geometries like pellets, the desired results are accessible by straightforward and established techniques. However, increasing the complexity of the sample geometry quickly places it outside the limits of these straightforward approaches---it is challenging, for example, to perform the desired data extraction on samples with even just two adjacent optically thin layers. The work discussed in this dissertation seeks to expand the range of sample geometries for which we can readily process THz spectroscopy data. We first focus on sample geometries which consist of a series of planar layers. While the general principles involved in modeling such planar samples are well established, applying them to multilayered structures can often lead to unwieldy expressions which we are unlikely to be able to handle without significant effort and/or error. In order address this, we developed a software package which builds these expressions automatically given any arbitrary planar sample geometry and uses these expressions to extract the refractive index spectrum of the layer of interest. While the class of planar samples includes most common samples, non-planar sample geometries have shown various attractive features as well. In particular, wire grid electrodes have shown promise as signal-enhancing substrates and THz transparent electrodes for operando THz spectroscopy measurements. We implemented and tested a 1D rigorous coupled wave approach for modeling these samples. Although it performed well when tested with finite-element simulations, it failed to accurately process experimental data collected on nanoparticulate films. We assessed a number of possible explanations for this discrepancy and suggest a number of possible strategies for addressing it.
Recommended Citation
Tayvah, Uriel, "Data Analysis Tools for Terahertz Spectroscopy" (2023). Yale Graduate School of Arts and Sciences Dissertations. 1084.
https://elischolar.library.yale.edu/gsas_dissertations/1084
