*3.4. HS-SPME-GC-MS Analysis*

For the VOC analysis, a slightly modified method optimized by Brizzolara et al. [40,41] was used. A total of 15 g of pericarp tissue (without seeds) were homogenized with 15 mL 1M NaCl water solution in a 50 mL plastic tube using T25 Ultra-Turrax (IKA, Königswinter, Germany) homogenizer. The obtained puree was frozen in liquid nitrogen and stored at −80 ◦C. Three replicates per treatment were analyzed for both the red ripe stage with a different level of Se accumulation in fruit.

VOCs were identified using a gas-chromatograph (Perkin Elmer Clarus 680) coupled with a mass spectrometer (Perkin Elmer Clarus 600 S). A total of 5 g of tomato puree was transferred into a 20 mL crimp vial (Sigma Aldrich, Milano, Italy) and was thawed at 15 ◦C for 15 min. Samples were incubated for 60 min at 40 ◦C, and VOCs were extracted for 45 min at the same temperature using an SPME Fiber (50/30 µm, DVB/CAR/PDMS, Sigma Aldrich, Italy).

We used the following GC temperature program: the initial temperature was 40 ◦C, which was maintained for 1 min; the temperature was then increased to 250 ◦C at a rate of 4 ◦C min−<sup>1</sup> and held for 1 min; finally, the temperature was increased to 280 ◦C at a rate of 15 ◦C min−<sup>1</sup> and held for 2 min.

Desorption was performed in spitless mode using a PSSI injector at 260 ◦C. A fused silica 30 m × 0.25 mm × 0.25 µm film thickness SLB-5MS column was used for the analysis, and helium was used as a carrier gas at a constant flow of 1 mL min−<sup>1</sup> . Chromatograms were analyzed using AMDIS (National Institute of Standards and Technology, Gaithersburg, MD, United States), and compound identification was carried out by comparing detected spectra with NIST v. 2 (National Institute of Standards and Technology, United States). Only compounds with 80%, or higher, levels of matching were considered. The retention index (RI) information was used to increase the identification efficiency. VOCs were also compared with molecules in the literature on tomato fruit [35,42]. In order to reduce

variability due to the SPME fiber decay, the raw peak area was normalized on the sum of the areas for each specific chromatogram.
