Geoscience Reference
In-Depth Information
“component” refers to inferred fluorophores or fluorophore groups that may have one or
more peaks. Thus, the humic-like component has two peaks, C and A
C
.
One of the difficulties in reconciling results is that early EEM studies did not collect
data below 250 nm excitation, and most PARAFAC models require that data in this range
be omitted owing to the presence of large molecular scattering peaks. Another difficulty
is that previous tables of CDOM fluorescence components organized by peak name or
number fail to represent adequately the double peak characteristics of both humic-like and
amino acid-like components. In
Table 3.1
, we have therefore imposed a solution acknowl-
edging that UVC peaks exist for peaks C, M, T, and B, and have designated them as A
C
,
A
M
, A
T
, and A
B,
respectively. In so doing, PARAFAC and bulk EEM components are much
more readily matched.
3.1.2 Humic-like EEM Components
Humic-like fluorescence is the dominant signal in most AOM and is due to the presence
of humic substances that arise from remineralization of organic matter occurring in soils
on land and in the water column and sediments in aquatic environments, both freshwa-
ter and marine. The chemical nature of humic substances varies across environments,
and this changing composition is reflected in the EEM, making it extremely difficult to
generalize about humic-like fluorescence in AOM. In general, two types of humic-like
fluorescence have been described - type C and type M. Because type C is the one most
commonly found in both freshwater and seawater, the discussion that follows begins with
its description.
The EEM for a typical AOM sample is not characterized by the symmetrical peaks of
pure fluorescent compounds (fluorophores), which appear round in the EEM contour view,
but rather by what may be compared to an elephant under a blanket in EEM 3-D view.
Figure 3.1
shows two views of EEMs for quinine sulfate dihydrate and water from the
Columbia River in Oregon. On the left, quinine sulfate fluorescence in contour view shows
a peak in the UVC region at
λ
ex
= 250 nm and a peak in the UVA (315-400 nm) region at
λ
ex
= 350, with a shoulder at 300 nm. The emission maximum occurs at 450 nm and, as
for all pure fluorophores, is independent of excitation energy. The variability in emission
is thus due solely to the width of the peak. For pure compounds, the emission intensity at
any given wavelength pair has a constant ratio to the intensity of emission at the fluores-
cence maximum. In contrast, the contour plot of the UVA peak from natural organic matter
in the Columbia River sample is not round, but rather oval, with the long axis of the oval
at an angle to both excitation and emission axes. This indicates that emission wavelength
is dependent on excitation wavelength, and therefore the fluorescence of the sample is not
due to a pure compound. Peak width is not constant, but rather is dependent on the rela-
tive amounts of fluorophores comprising the mixture in the sample. Data for fluorescence
intensity of natural organic matter samples thus require inclusion of the wavelength pair at
which the measurement was recorded, and there is no a priori relationship between differ-
ent points of the matrix.
