Return to search

Tracing Anthropogenic Wastes: Detection of Fluorescent Optical Brighteners in a Gradient of Natural Organic Matter Fluorescence

A dual wavelength method was developed for the field detection of optical brighteners
(OBs), fluorescent laundry additives used as indicators of anthropogenic wastes. The
method was quantitative for OBs under variable levels of fluorescent colored dissolved
organic matter (CDOM). Based on excitation at 300-400 nm and 440 and 550 nm
emission, the method assumed a constant ratio of fluorescence due to CDOM alone, even
if absolute amplitude varied. Concentrations of OBs were computed as the difference
between the observed 440 nm emission and the expected CDOM fluorescence at 440 nm,
as extrapolated from the 550 nm fluorescence and established CDOM fluorescence ratio.
Real-time inner filter corrections were based on absorption modeled from 550 nm
fluorescence and from exponential relationships at alternate wavelengths. The effects of
temperature and dissolved oxygen on CDOM fluorescence and computed OB were
quantified but were minimal because effects were comparable between the two
fluorescence regions. Assumptions on the locally conservative behavior of CDOM were
supported in field surveys of sewered and non-sewered areas. Varying water masses were
detected, but OB quantities were detected that did not co-vary with fluorescence alone.
Eleven geographic regions of peninsular Florida and sources of OBs were sampled to
evaluate the method under a broader range of CDOM and to conduct an extensive
detergent spike analysis. Fluorescence data were collected as EEMs and subjected to
PARAFAC modeling, isolating eight spectral factors that could sufficiently describe all
samples. There were no visible regions of the spectra that were unique to detergents or
OBs, but a previously unreported peak in the UV (<230 / 284 ex / em) was tentatively
identified as a detergent surfactant and should be pursued as a potential complementary
indicator of anthropogenic wastes. Limits on EEM fluorescence measurements were
identified: maximum linear range, maximum turbidity, and sensitivity to assumptions. A
sub-sampling technique of EEM data approximated the filter fluorometer readings, was
used to optimize the dual wavelength method, validated the method with spike
recoveries, and presented alternative approaches.

Identiferoai:union.ndltd.org:USF/oai:scholarcommons.usf.edu:etd-5011
Date09 November 2009
CreatorsDixon, Laura Kellie
PublisherScholar Commons
Source SetsUniversity of South Flordia
Detected LanguageEnglish
Typetext
Formatapplication/pdf
SourceGraduate Theses and Dissertations
Rightsdefault

Page generated in 0.0022 seconds