Environmentally-friendly technology for rapid identification and quantification of emerging pollutants from wastewater using infrared spectroscopy

• Published in: Environmental toxicology and pharmacology 2020-11, Vol.80, p.103458, Article 103458
• Main Authors: Quintelas, C., Melo, A., Costa, M., Mesquita, D.P., Ferreira, E.C., Amaral, A.L.
• Format: Article
• Language: English
• Subjects: 17β-Estradiol; Activated sludge; Atrazine; Carbamazepine; Emerging pollutants; Environmental monitoring; Environmental Monitoring - instrumentation; Environmental Monitoring - methods; FTIR spectroscopy; Herbicides; Ibuprofen; Infrared spectroscopy; kNN; Least squares; Least-Squares Analysis; Models, Theoretical; OLS; Paracetamol; PLS; Pollutants; Pollution monitoring; Prediction models; Reagents; Sex hormones; Sludge treatment; Spectroscopy, Fourier Transform Infrared - methods; Spectrum analysis; Sulfamethoxazole; Technology; Waste Water - chemistry; Wastewater pollution; Wastewater treatment; Water Pollutants, Chemical - analysis
• ISSN: 1382-6689; 1872-7077
• DOI: 10.1016/j.etap.2020.103458

•Quantification of emerging pollutants (EP) in wastewaters by IR spectroscopy.•A new approach to the use of chemometrics for model development.•Partial least squares and ordinary least squares modeling approaches were employed.•This developed model allows good prediction abilities regarding EP concentrations. The monitoring of emerging pollutants in wastewaters is nowadays an issue of special concern, with the classical quantification methods being time and reagent consuming. In this sense, a FTIR transmission spectroscopy based chemometric methodology was developed for the determination of eight of these pollutants. A total of 456 samples were, therefore, obtained, from an activated sludge wastewater treatment process spiked with the studied pollutants, and analysed in the range of 200 cm−1 to 14,000 cm−1. Then, a k-nearest neighbour (kNN) analysis aiming at identifying each sample pollutant was employed. Next, partial least squares (PLS) and ordinary least squares (OLS) modelling approaches were employed in order to obtain suitable prediction models. This procedure resulted in good prediction abilities regarding the estimation of atrazine, desloratadine, paracetamol, β-estradiol, ibuprofen, carbamazepine, sulfamethoxazole and ethynylestradiol concentrations in wastewaters. These promising results suggest this technology as a fast, eco-friendly and reagent free alternative methodology for the quantification of emerging pollutants in wastewaters.