Spectroscopic signals
Spectroscopic measurements such as UV–visible absorption
and luminescence emission should have a notorious place the
environmental analysis because they allow for operational simplicity and low cost. In addition, methods based on luminescence
signals have high sensitivity. However, the resulting spectra are
often affected by spectral interferences, which hinder their application in the analysis of complex matrices unless careful, but not
necessarily green, cleaning and extraction procedures are applied. In such situations, as will be discussed below,
the use of multivariate tools plays a fundamental role.
In a recent work, the determination of the herbicide at razin, a
priority pollutant with carcinogenic and endocrine disruptor activities, as well as its two main metabolites, was carried out in
natural waters through first-order UV–vis absorption spectra data
and partial least squares (PLS) regression.
Although the measurement stage was simple and the first-order
calibration allowed the simultaneous determination of analytes
with overlapping spectra, the analysis of natural waters was interfered by matrix effects and preliminary cleanup to the removal
organic matter was needed.
Methods based on fluorescence and phosphorescence signals
are more sensitive and selective than those based on UV–visible
absorption and, therefore, they are more favourable for the de-
termination of pollutants in the aquatic environment. Unavoidably, however, luminescence methods are also prone to be affected by spectral interferences present in the investigated matrices. This is where the
higher-order multivariate calibration is crucial in the development
of green methods.