Comparative evaluation of chromatographic analytical methods in untargeted screening of emerging contaminants in wastewater

Emerging contaminants include pharmaceuticals, personal care products, industrial additives, and other human-made chemicals that are constantly released into the environment; the effluent from wastewater treatment plants (WWTPs) is one of the main sources of ECs, which are usually released into surface waters and then they end up into sediment, soil, groundwater and seas. The term “emerging contaminant” refers to substances that have been recently identified or are currently under assessment and are largely not subject to environmental regulation. Their spread can cause known or suspected harm to the environment and human health, yet they have not been subjected to international regulation. This thesis is part of a larger project called "EXWASTER", funded by Ministry of University and Research, which involves three main universities: University of Modena and Reggio Emilia (lead partner), University of Genova and University of L'Aquila. The aim of this project is to develop novel analytical methods, both based on targeted and untargeted approaches, synergistically applied to monitor inorganic (metal ions, ionic species) and organic (emerging contaminants) compounds in different water samples. Multivariate data analysis and experimental design techniques are being applied to interpret acquired data, to verify quality of the reclaimed water and give wastewater a high added value. This thesis work involved the comparative evaluation of two untargeted analytical methods for the determination of the widest range of emerging contaminants in environmental samples of wastewater, based on UHPLC-MS/MS through the use of two different stationary phase columns: a biphenyl and a C18; the study has investigated spot samples collected from a wastewater treatment plant in Northern Italy. Specifically, wastewater was sampled both before and after 24-hour turnover, and then treated in laboratory via several steps, which included filtration, solid-phase extraction (SPE), elution, concentration and dilution. At the same time, multi-standard solutions containing 39 targeted analytes at two different concentrations were prepared, as well as several solvent and procedural blanks, recoveries, spikes, and QC samples. Analyses were performed at Centro Interdipartimentale Grandi Strumenti, UNIMORE, via UHPLC Ultimate 3000 coupled with Q-Exactive high resolution hybrid Quadrupole-Orbitrap mass spectrometer, which is served by a heated electrospray as ionization source (HESI), performing both positive and negative ionization modes. Due to high resolution, huge datasets were produced, specifically four ones: for each column, two acquisitions (one for positive and one for negative mode) were taken. Data analysis was carried out with chemometric approaches, including ROIMCR algorithm modeling and putative identification via Compound Discoverer. ROI performance was run by systematically setting different signal threshold parameters. Next, best threshold value was chosen, with which different MCR models were componentized, and compared through fit indicators as well as by components classification, in order to choose the best one for each of the four data-multisets. Putative identification was supported by five confidence levels, which allowed a more comprehensive understanding of every compound identification reliability; results, produced in the resolved features, showed the presence of various level 1 (targeted analytes) and level 2 compounds in environmental samples, including drugs and pharmaceuticals, pesticides, surfactants and industrial cleansers, personal-care derived products, polymers and plasticizers. An explorative principal component analysis (PCA) was performed. A first attempt at calculating matrix effect and method recovery was also made. Finally, a comparative evaluation, based on results obtained, was performed. Further studies will allow a better development of these novel techniques.