Geoscience Reference
In-Depth Information
some common pitfalls and proper spectroscopic techniques for understanding EEMs and
applying indices.
9.4.1 Instrument-Specific Effects and Proper EEM Correction
Multiple studies have highlighted the effect of instrument variation on fluorescence DOM
analysis. Kalbitz et al. (
2001
) found that the HIX
SYN
values determined by two different
fluorometers correlated well but recommended using a correction curve from a reference
standard in order to compare the entire synchronous scan, suggesting that the index, as a
ratio of two points, may be less sensitive to instrument variation. Holbrook et al. (
2006
)
compared various EEM correction methods and their impact on peaks related to HIX
EM
and FI. They found that for these indices, because they both depend on a single excitation
wavelength, emission corrections are the most important to apply, but recommend always
collecting in ratio mode (collecting the signal as emission intensity normalized to the lamp
intensity) to account for any variations in temperature or lamp intensity. Murphy et al.
(
2010
) compared multiple standard samples measured in 20 different labs using 8 different
fluorometer models. They found that correction method generally affected the FI, but by
treating all samples identically the variation could be brought down to 8% between labs.
This identified a need for a standardized correction procedure to compare EEMs and indi-
ces between instruments, labs, and studies. Lawaetz and Stedmon (2009) recommended
a procedure for EEM correction and a Raman intensity calibration, resulting in fluores-
cence intensity in Raman units to better aid comparisons between labs. Finally, Cory et al.
(
2010
) offered a detailed analysis of correction procedures and data comparisons across
three common fluorometers. Cory found that although different instruments, with differ-
ent signal-to-noise ratios, will have variations, properly applied correction procedures will
greatly reduce variation among the FIs.
9.4.2 Concentration Issues and the Inner-Filter Effect
Inner-filter effects refer to the attenuation of light prior to detection by the fluorometer,
due either to absorption of excitation light before reaching the fluorescent molecule (the
primary inner-filter effect) or absorption of light emitted from the fluorescent molecule
before being detected by the fluorometer (the secondary inner-filter effect). To obtain an
accurate spectroscopic reading, inner-filter effects must be accounted for in highly absorb-
ing samples either by dilution or application of an inner-filter correction (Mobed et al.,
1996
; McKnight et al.,
2001
). The most commonly used inner filter correction is based on
a correction for a path length of 0.5 cm in a 1-cm cell (Lakowicz,
2006
: 56). However, the
application of this inner-filter correction is sufficient only for samples with low enough
absorbance for this assumption to adequately represent the attenuation of light, after which
dilution is required. Ohno (
2002a
) demonstrated that, for calculation of HIX
EM
values (see
Section 9.2.2
), inner-filter corrections were effective for removing inner-filter effects only
up to an absorbance at 254 nm of 0.3 cm
-1
. Kalbitz et al. (
2001
) recommended a DOC no
