Evaluation of the matrix effect in the analysis of 74 pesticides present in golden gooseberry, purple passion fruit and Hass avocado by UPLC-MS/MS

Matrix Effect (ME) refers to the influence of all components within a matrix on the quantification of target analytes and occurs when detection techniques such as mass spectrometry (MS) are used to determine compound residues in food such as pesticides, compromising unpredictably the accuracy and precision of the results. Therefore, ME must be assessed during method validation to ensure reliable pesticide determination. This work evaluated ME in 74 pesticides present in golden gooseberry (GG), purple passion fruit (PPF) and Hass avocado (HA) by UPLC-MS/MS. The SANTE 11312/2021 guideline was followed to validate the methods, and the great majority of analytes satisfied the acceptance criteria recommended by SANTE. The calibration-graph and concentration-based methods were employed to assess ME. The calibration-graph analysis revealed similar pesticide responses in GG and PPF but differences in HA. Spearman correlation tests confirmed a stronger positive correlation between GG-PPF than GG-HA and PPF-HA, respectively. The compounds which exhibited contrasting behavior in GG and PPF (methomyl, fenhexamid, carbendazim, among others), contradict SANTE’s recommendation to validate at least a single matrix per commodity group. The concentration-based method demonstrated that lower levels are more affected than higher levels by the ME in the three fruits and exhibited to be the most proper way to assess ME due to the obtention of more accurate results for all analytes at each concentration level. • The statistical correlation was stronger for the pair GG-PPF than GG-HA and PPF-HA. • Minor nutrient variations in similar-content fruits may influence ME of pesticides. • Calibration-graph method may be used for general comprehension of ME of pesticides. • ME of pesticides is more precise when it is assessed by concentration level. • SANTE’s one-matrix validation hint is inaccurate; all matrices need validation.