Environmental Engineering Reference
In-Depth Information
2004). FTIRS was used by Wirrmann et al. (2001)
to quantify the mineral abundance in dated lake sedi-
ments, with the aim of investigating hydrologic
records during the Late Holocene. The sediment
samples were ground and diluted in KBr pellets
before measurement of the mid-infrared spectra. The
infrared absorbance was linearly correlated to the
composition and mass of constituents in the KBr
pellet. The spectra of sediment samples were com-
pared with spectra of pure mineral phases similar to
the ones found in the sediments, to perform a cali-
bration. The authors observed a good correlation
between their quantitative results from FTIR spec-
troscopy and chemical analysis.
However, the technique usually used in the mid-
infrared region by those studying sediments is diffuse
refl ectance infrared Fourier transform spectroscopy
(DRIFTS or DRIFT). DRIFT is a surface characteri-
zation method, involving a refl ection experiment
where the typical depths of penetration of the infra-
red beam into the surface are 1-10
organic matter in industrial and agricultural materi-
als and for process control (Hsu 1997; Skoog et al.
1998). The application of NIRS to the analysis of
environmental samples started in the 1990s and since
then has been increasing. NIRS has been used for
prediction of heavy metal concentration in freshwa-
ter sediments (Malley 1997), analysis of spatial vari-
ability in surface lake sediments (Korsman et al.
1999), analysis of C, CO 3 -2 , N and P in freshwater
sediments (Malley 1997), and the determination of
carbon in marine sediments (Chang et al. 2005).
In the infrared spectra, the detected changes in
transmittance (or absorption) intensity are presented
as a function of frequency. The separation and meas-
urement of infrared radiation in most commercial
instruments is performed using dispersive spectrom-
eters (based on diffraction gratings) or Fourier trans-
form spectrometers (based on interferometer fi lters).
The photometers and spectrophotometers used to
perform measurements in the near-infrared region
are dispersive spectrometers, similar in design and
components to those used in ultraviolet/visible
absorption spectrometry (Skoog et al. 1998). In
Fourier transform infrared spectroscopy, all frequen-
cies are analyzed simultaneously, rather than exam-
ining each component frequency sequentially, as in
the dispersive infrared spectrometer. Interferometric
instruments have high resolutions and are very accu-
rate with reproducible frequency determinations.
Moreover, their signal-to-noise ratios are better than
those of a good-quality dispersive instrument by
more than an order of magnitude (Hsu 1997; Skoog
et al. 1998).
Until the early 1980s, the use of the mid-infrared
region was limited to qualitative organic analysis and
structure determination based on absorption spectra,
because the only available instruments were of the
dispersive type. The multiple layers of information
featured in a mid-infrared spectrum were a major
challenge for the interpretation and quantifi cation of
the data. Since then, however, the appearance of
Fourier transform spectrometers, based on interfer-
ence fi lters, has brought a dramatic increase in the
number and type of applications of mid-infrared
radiation. This technique is known as Fourier trans-
form infrared spectroscopy (FTIRS). It offers some
important advantages in sediment analysis, such as
the ability to assess both mineral and organic struc-
tures in particles, and good sensitivity (Gallé et al.
m, suffi cient
depth to characterize the organic layer on mineral
surfaces. However, obtaining reproducible quantita-
tive DRIFT measurements requires strict attention to
experimental details, especially to particle size distri-
bution and packing density of the sample (Belton &
Wilson 1990; Gallé et al. 2004).
To address the question of whether the TOC
content of lake water follows changes in climate and
vegetation on a millennia timescale, Rosén & Persson
(2006) tested the hypothesis that DRIFTS of lake
sediments can be used to infer past changes in tree-
line position and TOC content of lake water. The
statistical method of principal component analysis
was used to get an overview of the spectral variabil-
ity of the lakes. Partial least square regression was
used to develop a transfer function between DRIFT
spectra of surface sediment (0-1 cm) and TOC. Both
quality and quantity of organic material can be
measured by DRIFTS. The relation between FTIR
spectra of sediment and the TOC content in the lake
water was probably because the sediments in lakes
with high and low TOC levels, respectively, have
quantitatively and qualitatively different composi-
tion, owing to different types of vegetation, algae,
input and degradation of organic material in the
water column. The authors succeeded in using the
transfer function developed between FTIR data of
the sediments and TOC to obtain information about
μ
Search WWH ::




Custom Search