Date of Award

5-2018

Degree Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Chemistry

Committee Chair

Garth Simpson

Committee Member 1

Lyudmila Slipchenko

Committee Member 2

Shelly Claridge

Committee Member 3

Hilkka Kenttämaa

Abstract

Drugs are an essential element protecting human lives from many diseases such as cancer, diabetes, and cardiovascular disorders. One of the highlights in drug development in recent years is the establishment of rational drug design: a collection of various multi-disciplinary approaches that at the core, focus on designing molecules with specific properties for identified targets and biomolecules with known functional roles and structural information. The candidate molecules will then go through a series of examinations to characterize their physiochemical properties, and an iterative process is used to improve the design of the drug to achieve desirable attributes. The time consuming and highly expensive nature of drug development constantly calls for new analytical techniques that have increasingly higher throughput, faster analysis speed, richer chemical and structural information, and lower risk and cost. Conventional analytical methods for pharmaceutical materials, such as X-ray diffraction analysis and Raman spectroscopy, often suffer from prolonged measurement time. In many cases, the identification of regions of interest within the sample is non-trivial in itself. Nonlinear optical imaging techniques, including second harmonic generation (SHG) microscopy and two-photon excited ultraviolet fluorescence (TPE-UVF) microscopy were developed as fast, real-time, and non-destructive methods for selective identification and characterization of crystalline materials present in pharmaceutical samples. These techniques were integrated with synchrotron X-ray diffraction analysis and Raman spectroscopy to significantly reduce the overall measurement time of these structure characterization techniques. In the meanwhile, with the now increased speed of measurement, the amount of experimental data acquired per unit time has also drastically increased. The rate at which data are analyzed, digested, and interpreted is becoming the bottleneck in data-driving decision-making. Novel electronics that only collect data at the most information-rich time points were employed to significantly increase the signal-to-noise ratio (SNR) during data acquisition, reducing the total amount of data needed for material characterization. Advanced sampling algorithms to reduce the total amount of measurements required for perfect data space reconstruction, automated programs for data acquisition and analysis, and efficient data analysis algorithms based on machine learning were developed for accelerated data processing for nonlinear optical imaging analysis, Raman spectra processing, and X-ray diffraction indexing.

Download

Share

COinS