Date on Master's Thesis/Doctoral Dissertation
Chemical Engineering, PhD
Committee Co-Chair (if applicable)
Breath analysis; microfabrication; volatile organic compounds; microreactors; simulations; machine learning; classification
Volatile Organic Compounds (VOC) generated endogenously in the human body can be used to detect diseases that induce oxidative stress and inflammation. Breath analysis has been used for the detection of diseases such as COPD, Depression, Lung Cancer and most recently COVID-19. Methods such as Exhaled Breath Condensate (EBC), Sorbent Tubes, Solid Phase Microextraction (SPME), and silicon microreactors have shown considerable capability in extracting and concentrating trace VOC’s present in human breath. Silicon microreactors functionalized with capture agents to derivatize carbonyl compounds are effective but the overall maximum flow rate of breath sample possible during analysis is low, making direct breath collection difficult. In this dissertation, we investigated both an improvement to our current design of microreactor and proposed an innovative radial flow microreactor whose flow resistance was low enough to facilitate direct breath collection through exhalation. First, CFD was used to simulate the base design of our standard microreactor (3L) and compared with the velocity gradients of several new designs with reduced micropillar array width. From simulation, 3 new designs were fabricated (3LN) and evaluated in
terms of pressure drop and Capture Efficiency (CE). A new design was found that performed similarly to the base design but with a smaller footprint. However, the velocity gradient in the simulation was a poor overall predictor of performance. To address the challenge of low max flow rates in silicon microreactors, an innovative radial flow microreactor was designed and fabricated using MEMS fabrication techniques. New procedures were developed to etch the inlet and outlet holes on the top and bottom of the wafer. The top inlet hole on the glass was created using laser processing or HF acid etching while the bottom side outlet holes were created with Deep Reactive Ion Etching (DRIE). Using the developed fabrication scheme, several designs of radial flow microreactors were fabricated and tested for flow resistance using a 3D printed Breath Collection Tool (BCT). The design with the least flow resistance was functionalized with silica nanoparticles and loaded with O-(2,3,4,5,6-pentafluorobenzyl)-hydroxylamine hydrochloride (PFBHA) to evaluate its performance in capturing aldehydes and ketones in exhaled breath. The radial flow microreactor and BCT achieved a direct exhalation flow rate of 2.75 L/min and maintained a capture efficiency greater than 20% for Propionaldehyde, 2-Butanone, and 2-Pentanone at an evacuation flow rate of 2.3 L/min. The radial flow microreactor design was used in a clinical study of 36 patients, 21 with COVID-19 and 15 without. Eleven calculated features obtained from GC-MS analysis were identified as potential indicators for COVID-19 using One-Way ANOVA. The features were used in a Machine Learning (ML) model and predicted a validation test with above 90% for both sensitivity and specificity.
Morris, James, "Novel breath collection techniques for detection of COVID-19." (2023). Electronic Theses and Dissertations. Paper 4131.
Available for download on Saturday, February 10, 2024