Geoscience Reference
In-Depth Information
EEMs provide an estimate of how nearly every measureable fluorophore responds to
metal additions, enabling the study of the metal-ligand fluorophores of DOM in greater
detail. Luster et al. (
1996
) used EEMs to study fluorescence changes on Cu
2+
addition
to DOM leached from leaf litter at copper to carbon ratios, Cu/C less than 1000. Up to
Cu/C ~1:250, quenching occurred in protein and (to a lesser extent) phenolic fluorescence
regions. Highly conjugated fluorophores were quenched at Cu/C greater than 250. Thus,
the amount of quencher added produced substantially different degrees of fluorescence loss
in DOM discernible in “identified” EEM peak regions. Luster et al. (
1996
) also determined
stability constants for Cu(II) and Al(III) in the leaf litter-extracted DOM. They were able
to model three binding sites on this DOM using the major terrestrial humic peaks found
in their EEM spectra and relating these fluorophores to functional group ligands such as
phenolics and carboxylic acids (both properties of salicyclic acid moiety). Metal addi-
tions, especially Al(III), can cause flocculation of material from solution. Sharpless and
McGown (
1999
) found that aggregation was caused by adding Al(III) reducing short and
long wavelength fluorescence in aquatic HA, but reducing only longwave fluorescence in
terrestrial (peat and soil) HA. They concluded that a decrease in fluorescence intensity of
HA is caused by precipitation of Al(III)-HA complexes from solution rather than fluores-
cence quenching.
Ohno et al. (
2008
) expanded on the use of EEM fluorescence to investigate metal binding
parameters on soil DOM leached separately from conifer and from deciduous tree litters.
They used a parallel factor model (PARAFAC) model to decompose three terrestrial humic
components that responded differently to additions of Fe(III) and Al(III). Their component
1 (peak at ex325/em450, similar to C peak) in the deciduous DOM was strongly quenched
both by Al and Fe, while their component 1 for the coniferous DOM was quenched only
by Al. Their component 3 (ex240/em400, similar to A peak) was not quenched by addition
of Fe(III) yet slightly increased in signal by addition of Al(III). Yamashita and Jaffe (
2008
)
have also used EEM-PARAFAC to model Cu(II) and Hg(II) binding to mangrove DOM in
the Florida Coastal Everglades. They found that Cu(II) actually decreases then enhances
protein fluorescence, while Hg(II) quenches the protein component fluorescence.
Metals have variable affinity for binding sites on DOM, which produces a range of
effects on DOM fluorescence quenching - effects that depend on both the metal and the
fluorophore. The competition for binding sites between highly ionic Al(III) and Fe(III) is
a key example. The work by Zhao and Nelson (
2005
) suggested that either Al(III) cannot
replace complexed Fe(III), or that Fe(III) displaced into solution dynamically quenches
FA fluorescence. Apparently, then, not all fluorophores complex equally; from the exam-
ination of four different divalent metals, Antunes et al. (2007) found distinct changes in
freshwater FA and a commercial HA EEMs fluorescence that varied with metal addition.
This result is similar to that found by Yamashita and Jaffe (
2008
) but their work specifi-
cally addresses the fact that often-overlapping broad emission spectra must be deconvolved
(they used principal component analysis [PCA] and multivariate curve resolution) into
discrete fluorescent components, which can be chemically meaningful (Ohno et al.,
2008
).
Thus, the application of multivariate statistical techniques such as PARAFAC to EEM data
