ULTRAVIOLET LIGHT APPLICATIONS FOR THE DEGRADATION OF NATURAL ORGANIC MATTER IN WATER MATRICES

Abstract

Ultraviolet light, in combination with chemical oxidants has been hypothesized as a universal treatment technology for the degradation of harmful organic compounds in water matrices. Advanced oxidation processes (AOPs) non-selectively degrades organic contaminants from water matrices via the production of hydroxyl radicals. A thorough understanding of the nature of the specific natural organic matter (NOM) profile for a surface water is necessary to mitigate negative side effects of AOP use, such as the formation of disinfection byproducts (DBPs). The goal of this research was to examine the changes in natural organic matter that is partially oxidized during AOP exposure. This goal was achieved via the following objectives: 1. Assess the AOP removal efficiency for a seasonal taste and odor compound found in surface waters in Atlantic Canada (geosmin) using a natural water matrix (Pockwock Lake, Halifax, Nova Scotia). 2. Examine the impact that AOPs have on DBPs using a source water that historically has elevated disinfection byproduct concentrations in their finished water (Pottle Lake, Sydney, Nova Scotia). 3. Model the primary changes in oxidized NOM from a fluorescent excitation/emission perspective. 4. Development of a high throughput approach for more efficient UV exposure experimentation using common NOM constituents as challenge compounds.

Geosmin was effectively removed below human detection levels in 1000 mJ cm-2 fluence, 10 mg L-1 hydrogen peroxide samples, suggesting that peroxide-based AOPs are better suited for taste and odor compound removal when compared UV and UV/ozone-based AOPs. Peroxide-based AOPs caused increases in DBP formation potential for both THM and HAAs, which exceeded the respective Health Canada guidelines of 100 µgL-1 and 80 µgL-1. Parallel factor (PARAFAC) fluorescent excitation emission matrix (FEEM) models revealed that hydrophobic acidic and basic fractions are susceptible to ozone AOPs whereas peroxide-based UV-AOPs initially oxidize hydrophilic acidic, hydrophobic basic and hydrophobic basic fractions. A high throughput approach for rapid UV exposure of samples was also developed and validated through proof-of-concept testing using tryptophan and tyrosine as model NOM compounds. The high throughput approach allows for faster, and more efficient bench scale AO experiment design while also allowing for tunable UV emittance via a microplate reader.