Marta Bevilacqua

Fluorescence spectroscopy has been applied for analysis of complex samples, such as food and beverages. Parallel factor analysis (PARAFAC) is a well‐known decomposition method for fluorescence excitation–emission matrices (EEMs). When the complexity of the system increases, it becomes considerably more difficult to determine the optimal number of PARAFAC components, especially when the fluorophores of the system are unknown. The two commonly applied diagnostics, core consistency and split‐half analysis, appear to underestimate the model complexity due to covarying components and local minima, respectively. As a more robust alternative, we propose a resampling approach with multiple initializations and submodel comparisons for estimating the optimal number of PARAFAC components in complex data.

Fluorescence spectroscopy combined with parallel factor analysis (PARAFAC) has successfully been applied for the analysis of food and beverages containing numerous autofluorescent compounds. For the decomposition of such data, it is crucial to establish the PARAFAC model complexity. This is not a trivial matter, especially when the sample complexity increases. Diagnostics are available for assisting the choice of the number of PARAFAC components, such as the core consistency. In this short communication, we show that when it comes to real (complex) data, the core consistency is too conservative and other diagnostic tools must be taken into account. We emphasize that it is imperative to inspect the PARAFAC excitation and emission loadings and assess whether these are chemically meaningful.

The use of front‐face fluorescence spectroscopy, and three‐way chemometric analysis (like PARAllel FACtor analysis, PARAFAC), has been successfully applied for the analysis of unpretreated samples in the fields of food and environmental analysis. It would be desirable to evaluate the potential of this approach for the analysis of any real sample in any field of research. Even in the simplest of the real samples acquired with front‐face, the presence of scattering and inner filter effects will occur, potentially hampering the subsequent data analysis due to deviations from Beer‐Lambert's law. This paper addresses these concerns in practice, proposing a strategy of spectral preprocessing to mitigate these effects. This is done by measuring several data sets with different levels of scattering and inner filter effects by fluorescence in excitation‐emission mode. The results show that the occurrence of these interferents (sometimes neglected with front‐face mode) affects the fluorescence signal and interferes with any traditional analysis on these data, as much as they hamper the successful use of methods like PARAFAC. The proposed preprocessing strategy is based on one of the most traditional correction for the inner filter effect with right‐angle mode. However, we suggest applying a tunable factor, b, that will account for the degree of deviation from linearity between concentration of a given analyte and its fluorescence signal. It is demonstrated that by choosing a proper b‐value, this correction helps in finding an acceptable solution for the PARAFAC algorithm, in line with Beer‐Lambert's law.

In this paper, we discuss the validity of using score plots of component models such as partial least squares regression, especially when these models are used for building classification models, and models derived from partial least squares regression for discriminant analysis (PLS-DA). Using examples and simulations, it is shown that the currently accepted practice of showing score plots from calibration models may give misleading interpretations. It is suggested and shown that the problem can be solved by replacing the currently used calibrated score plots with cross-validated score plots.

Fermentation processes are still compromised by a lack of monitoring strategies providing integrated process data online, ensuring process understanding, c

Most plastics need to incorporate flame retardants to meet fire safety standards requirements. The amount and the type of flame retardants can differ, so that in waste plastics a large variety of polymers and flame retardants can be found. The recycling of plastics containing flame retardants is increasing. However, only plastics of the same polymer type and the same additive content can be recycled together. Three models based on different chemometrics techniques applied to hyperspectral imaging in the near infrared range were developed [partial least square-discriminant analysis, decision tree (DT) and hierarchical model (HM)]. Optimal results were obtained for all classification techniques. HM shows the highest error at all levels due to the noisy spectra of the black plastics. However, DT classification gave outstanding results, considering that the sensitivity was higher than 0.9 in all cases. Thus, the application of DT with hyperspectral imaging could be used to sort plastic samples with respect to the type of polymer and the flame retardant used with a high degree of accuracy in an automated way. These findings are highly valuable for the plastic and waste management industries.