Premium
Two‐dimensional fluorescence as a fingerprinting tool for monitoring wastewater treatment systems
Author(s) -
Galinha Claudia F.,
Carvalho Gilda,
Portugal Carla A. M.,
Guglielmi Giuseppe,
Reis Maria A. M.,
Crespo João G.
Publication year - 2011
Publication title -
journal of chemical technology and biotechnology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.64
H-Index - 117
eISSN - 1097-4660
pISSN - 0268-2575
DOI - 10.1002/jctb.2613
Subject(s) - fluorescence , biological system , effluent , wastewater , multivariate statistics , fluorescence spectroscopy , quenching (fluorescence) , chemistry , environmental science , computer science , analytical chemistry (journal) , environmental chemistry , environmental engineering , machine learning , optics , biology , physics
Abstract BACKGROUND: The use of two‐dimensional (2D) fluorescence for monitoring complex biological systems requires careful assessment of the effect of chemical species present, which may be fluorescent and/or may interfere with the fluorescence response of target fluorophores. Given the complexity of fluorescence data (excitation emission matrices—EEMs), the challenge is how to recover the information embedded into those EEMs that can be related quantitatively with the observed performance of the biological processes under study. RESULTS: This work shows clearly that interference effects (such as quenching and inner filter effects) occur due to the presence of multiple species in complex biological media, such as natural water matrices, wastewaters and activated sludge. A statistical multivariate analysis is proposed to recover quantitative information from 2D fluorescence data, correlating EEMs with the observed performance. A selected case study is discussed, where 2D fluorescence spectra obtained from the effluent of a membrane bioreactor were compressed using PARAFAC and successfully correlated with the effluent chemical oxygen demand, using projection to latent structures modelling. CONCLUSION: This study demonstrates the potential of using 2D fluorescence spectroscopy as a status fingerprint. Additionally, it is shown how statistical multivariate data analysis can be used to correlate EEMs with selected performance parameters for monitoring of biological systems. Copyright © 2011 Society of Chemical Industry