*2.4. HPLC Analysis of Total Phenolic Compounds Free OHTyr and Tyr, and Total OHTyr and Tyr*

The analysis of free OHTyr and Tyr and of total phenolic compounds was performed according to the IOC official method [20]. Phenolic compounds were extracted from the EVOO sample (2.0 g) with 5 mL MeOH:H2O 80:20 *v*/*v* by shaking for 1 min, extracting in an ultrasonic bath for 15 min, and centrifuging at 5000 rpm for 25 min. The supernatant was filtered on PVDC filters (0.45 μm) and analyzed by HPLC. An HP 1100 system coupled to a diode array detector (Agilent Technologies, Santa Clara, CA, USA) was used. The column was a SphereClone ODS-2, 5 μm, 250 × 4.6 mm id kept at room temperature during chromatographic separation. The eluents were formic acid solution (pH 3.2), acetonitrile, and methanol. The analyses were carried out at a flow rate of 1 mL min<sup>−</sup>1, with an injection volume of 20 μL and a total analysis time of 82 min, applying the gradient reported in the IOC method. The areas were registered at 280 nm, with syringic acid as the internal standard. The content of phenolic compounds was expressed as mg of Tyr per kg of oil.

Total OHTyr and Tyr were determined by HPLC after acid hydrolysis (with sulfuric acid at 80 ◦C for 2 h) of the hydroalcoholic extract obtained according to the IOC method [21]. For HPLC analysis of the hydrolyzed extracts, formic acid solution at pH 3.2 (solvent A) and acetonitrile (solvent B) were used as eluents. An HP1200 liquid chromatograph coupled to a diode array detector (Agilent Technologies, Santa Clara, CA, USA) was used with a 150 × 3 mm (5 μm) Gemini RP18 column (Phenomenex, Torrance, CA, USA). The flow rate and the injection volume were 0.4 mL min−<sup>1</sup> and 20 μL, respectively. A linear gradient was applied starting from 95% A to 70% A in 5 min, to 50% A in 5 min, to 2% A in 5 min with a final plateau of 5 min. Total analysis time was 22 min. For UV detection, the wavelength of 280 nm was used. The results were expressed as mg of Tyr per kg of oil for tyrosol, and as mg of OHTyr per kg of oil for hydroxytyrosol after application of the formula reported by Bellumori et al. [22], for keeping into account the 35% overestimation of OHTyr when the calibration curve of tyrosol is used.

#### *2.5. NADES-UV Method Validation*

The NADES-UV method was validated by measuring the linearity, the limits of detection (LOD) and quantitation (LOQ), as well as the precision (repeatability), and accuracy (recovery). Linearity of the response in the analytical range was assessed by regression analysis and R<sup>2</sup> values. LOD was expressed as 3.3 × S.D. of the blank (*<sup>n</sup>* = 3), while LOQ was expressed as 10 × S.D. of the blank (*n* = 3). Sunflower oil was used as

blank sample. Repeatability was assessed with repeated intra-day (*n* = 3 × 5 samples) and inter-day (*n* = 3 × 1 sample × 3 days) trials. Recovery was determined as apparent recovery compared to the reference method [21,22] and expressed in percent.

#### *2.6. Statistical Analysis*

Spectra, after subtraction of the pure solvent spectrum, were pre-processed using Solo 8.6.2 (Eigenvector Research, Inc., Manson, WA, USA), by applying the second derivative Savitzky–Golay algorithm (polynomial order: 2, window: 11 pt). Analysis of correlation, least squares regression, and analysis of variance (ANOVA) were performed with Origin 2021 (OriginLab Corporation, Northampton, MA, USA).

#### **3. Results and Discussion**
