Automated, downscaled QuEChERS-based sample preparation with performic acid epoxidation for GC–MS/MS analysis of twelve phthalates in edible oils

• Automated miniaturized QuEChERS for 12 priority phthalates in edible oils. • Temperature-controlled double extraction improves robustness for high-melting oils. • In-vial performic acid epoxidation removes olefin interferences without reconcentration. • GC–MS/MS (MRM) with column backflush improves selectivity and robustness. • Method verified using edible-oil proficiency-test samples. An automated, miniaturized QuEChERS-type sample-preparation workflow was developed for the determination of twelve priority phthalates in edible oils by gas chromatography–tandem mass spectrometry (GC–MS/MS): dimethyl phthalate (DMP), diethyl phthalate (DEP), di-n-propyl phthalate (DNPP), diisobutyl phthalate (DIBP), dibutyl phthalate (DBP), di-n-hexyl phthalate (DNHP), benzyl butyl phthalate (BBP), dicyclohexyl phthalate (DCHP), di(2-ethylhexyl) phthalate (DEHP), di-n-octyl phthalate (DNOP), diisononyl phthalate (DINP), and diisodecyl phthalate (DIDP). The workflow integrates temperature-controlled double extraction (2 × 3 mL acetonitrile at 80°C) to improve robustness for high-melting matrices, PSA/C18 dispersive clean-up, and in-vial performic acid epoxidation to mitigate olefinic interferences (e.g., squalene), followed by GC–MS/MS (MRM) with column backflush. Epoxidation was implemented in the final isooctane extract without an additional post-epoxidation concentration step, enabling simultaneous quantification across the complete target analyte set, including lower-molecular-weight phthalates. Performance evaluation in squalene-rich amaranth oil showed linear calibration (R² > 0.99) and good precision (RSD ≤ 10% at 0.1 and 1 mg/kg; DINP and DIDP at 1 and 10 mg/kg). LOQs were 2.7–41 µg/kg for DMP–DNOP, while DINP and DIDP showed higher LOQs (93 and 272 µg/kg). Two edible-oil proficiency-test samples agreed with assigned values within acceptable performance limits. AGREEprep benchmarking improved from 0.12 (ISO reference workflow) to 0.23, mainly driven by miniaturization and automation. Future work will address higher LOQs for DINP and DIDP, optimize clean-up for medium-chain triglycerides (MCT)-rich matrices, and expand validation to a broader range of edible oils.