Environmental Engineering Reference
In-Depth Information
constraints, UV and fluorescence approaches to characterizing DOM have
exponentially increased in recent years (Cory and McKnight 2005 ; Fellman
et al. 2009 ; Helms et al. 2008 ; Inamdar et al. 2011 , 2012 ; Jaff ยด et al. 2008 ; Miller
and McKnight 2010 ; Wilson and Xenopoulos 2009 ) including some excellent
reviews on the subject (Cory et al. 2011 ; Fellman et al. 2010 ).
The UV absorption spectra for DOM is generally obtained using a standard
spectrophotometer equipped with a 1 cm path-length quartz cuvette (volume of
4 ml) over the 190-1,100 nm wavelength range at 1-nm intervals. Prior to the
sample spectra, the instrument is set up and corrected for scattering and baseline
fluctuations by using deionized (DI) water. The absorption spectrum for DOM
follows an exponential pattern with a decrease in absorption with increasing
wavelength. Some of the key UV metrics that are derived from this spectrum and
which have been used to characterize DOM are reported in Table 7.1 . The UV
metric that has been most commonly reported is the specific UV absorbance
(SUVA) which is computed by dividing the decadic UV absorbance at 254 nm by
the concentration of DOC (mg C/L) (Weishaar et al. 2003 ). SUVA has been found
to be strongly and positively correlated with aromatic content of DOM as deter-
mined by 13 C-NMR (Weishaar et al. 2003 ). SUVA values can however be
influenced by the pH, nitrate and dissolved iron (Fe) content of the sample and
appropriate screening and corrections need to be applied (Weishaar et al. 2003 ).
Since Fe absorbs light at 254 nm, elevated concentrations of Fe (
0.5 mg/L) in the
DOM sample can lead to incorrect (high) SUVA values (Weishaar et al. 2003 ).
A metric similar to SUVA, the absorption coefficient at 254 nm (a 254 in m) is also
calculated by using the naperian UV absorption coefficient (Green and Blough
1994 ). The a 254 also provides a measure of aromaticity but without normalization to
DOC (Helms et al. 2008 ). Another UV index, the spectral slope ratio, S R ,is
calculated as the ratio of the slope of the shorter UV wavelength region
(275-295 nm) to that of the longer UV wavelength region (350-400 nm) (Helms
et al. 2008 ) and is obtained using linear regression on the log-transformed spectral
ranges (Yamashita et al. 2010 ). The spectral slope ratio, S R has been found to be
inversely related to the molecular weight of DOM (Helms et al. 2008 ).
In fluorescence spectroscopy, DOM samples are exposed to light in a fluorome-
ter for a range of excitation wavelengths and the corresponding emitted wavelength
and light intensity is recorded (Lakowicz 1999 ). The matrix of the fluorescence
intensities that is generated is referred to as the excitation-emission matrix (EEMs),
an example of which is illustrated in Fig. 7.1 . Prior to generating fluorescence scans
and deriving meaningful indices from the EEMs a number of important steps need
to be performed such as correcting for instrument bias, diluting samples with
high absorbance values (e.g., A254
>
0.2) and applying corrections to account
for inner-filter effects (McKnight et al. 2003 ). Once the EEMs are generated a
variety of fluorescence indices can be generated by using the ratios of fluorescence
intensities from specific regions (wavelengths) of the EEM matrix. In addition,
EEMs can be further analyzed using rigorous multivariate statistical tools such as
parallel factor analysis (PARAFAC, Stedmon et al. 2003 ) that decomposes the
EEMs matrix into chemically and mathematically distinct components with the
Search WWH ::




Custom Search