Classifying micropollutants by environmental risk in groundwater using screening analysis associated to a hybrid multicriteria method combining (Q)SAR tools, fuzzy AHP and ELECTRE

• Published in: The Science of the total environment 2023-09, Vol.892, p.164588-164588, Article 164588
• Main Authors: Becker, Raquel W., Araújo, Débora S., Jachstet, Letícia A., Ruiz-Padillo, Alejandro, do Amaral, Bianca, de Souza, Jussara E., Müller Athayde, Camila de V., Athayde, Gustavo B., Sirtori, Carla
• Format: Article
• Language: English
• Subjects: LC-QTOF MS; Multicriteria decision-making; Pesticides; Qualitative analysis; Quantitative structure-activity relationships
• ISSN: 0048-9697; 1879-1026
• DOI: 10.1016/j.scitotenv.2023.164588

This study aimed to evaluate and monitor pesticides in groundwater of the Serra Geral aquifer, located in the Paraná Basin 3 (southern Brazil), using Liquid Chromatography coupled with a Quadrupole-Time-of-Flight Mass Spectrometer (LC-QTOF MS). A total of 117 samples, collected in three different moments, were analyzed over 36 months. Groundwater samples from 35 wells and four surface water points were monitored in each sampling campaign. A pesticide screening methodology was proposed with the tentative identification of 1607 pesticides and pesticide metabolites. The application of the proposed methodology resulted in the verification of 29 pesticides and pesticide metabolites, 7 as confirmed analytes and 22 as suspect compounds. (Q)SAR in silico predictions and GUS index calculation provided data on the potential environmental risk of the identified compounds, with eight endpoints considered. After in silico predictions, an alternative hybrid multicriteria method was applied, combining the weighting of endpoints of fuzzy AHP and micropollutants classification by environmental risk using ELECTRE. The fuzzy AHP results indicated the greatest importance of mutagenicity among the eight evaluated indicators, while the scarce influence of the physicochemical properties on the environmental risk suggested their exclusion from the model. Accordingly, the ELECTRE results highlighted the prevalence of thiamethoxam and carbendazim as the most dangerous for the environment. The application of the proposed method enabled the selection of the compounds that must be monitored, considering mutagenicity and toxicity predictions for environmental risk analysis. [Display omitted] •Micropollutants were first evaluated by screening analysis using LC-QTOF MS in BP3.•(Q)SAR tools were used to predict eight endpoints for the identified compounds.•An alternative hybrid multicriteria method was employed using fuzzy AHP and ELECTRE.•Fuzzy AHP was helpful in weighting the relative importance of selected endpoints.•ELECTRE allowed classifying compounds based on their environmental risk assessment.