Date on Master's Thesis/Doctoral Dissertation


Document Type

Doctoral Dissertation

Degree Name

Ph. D.


Chemical Engineering

Committee Chair

Fu, Xiao-An

Author's Keywords

Breath analysis; Lung cancer; MEMS; Biomarker; Preconcentration; Carbonyl compounds


Carbonyl compounds; Chemical detectors; Air sampling apparatus; Lungs--Cancer--Diagnosis


Detection of volatile organic compounds (VOCs) at trace level (parts per billion volume (ppbv) to parts per trillion (pptv)) has become an important research area because of demanding applications in homeland security, environmental monitoring, and noninvasive diagnosis of diseases. The analysis of trace VOCs challenges existing analytical instruments because their concentrations are beyond current instrument limits of detection. In this dissertation, we have investigated an innovative microreactor that is suitable for quantitative analysis of volatile carbonyl compounds in ambient air as well as in human exhaled breath. The approach is based on microreactor chips fabricated from four inch silicon wafers. The chips have thousands of micropillars in the microfluidic channels for uniformly distributing gaseous samples flowing through the microreactors. The surfaces of the micropillars are functionalized with a quaternary ammonium aminooxy salt 2-(aminooxy)-N,N,N-trimethylethanammonium (ATM) iodide for trapping trace ketones and aldehydes by means of oximation reactions. ATM adducts and unreacted ATM are eluted from the microreactor with less than 40µL of methanol and directly analyzed by nanoelectrospray Fourier transform ion cyclotron resonance (FTICR) mass spectrometry (MS). Capture efficiencies of above 98% have been achieved for ketones and aldehydes. Carbonyl compounds at levels of 1 ppbv have been detected using the microreactor for capture. Ambient air samples from indoor and outdoor areas were collected using Tedlar bags, and analyzed using the microreactor and FTICR-MS. The analysis of air samples provides the reference for analysis of exhaled breath samples. Finally, exhaled breath samples from patients with untreated lung cancer (LC) patients, patients with benign pulmonary nodules and healthy volunteers (smokers and non-smokers) were collected and analyzed using the same method. The FTICR-MS spectra showed that the exhaled breath samples from LC patients have a unique pattern of VOCs, in comparison with healthy controls and patients with benign pulmonary nodules. 2-Butanone, hydroxy-acetaldehyde, 3-hydroxy-2-butanone, and 4-hydroxy-hexenal (4-HHE) were found to have significant higher concentrations for LC patients. A simple diagnosis method based on four elevated VOCs could easily discriminate lung cancer patients from healthy controls and patients with pulmonary nodules with 90.6% diagnosis sensitivity and 81.3% specificity.