**4. Discussion**

The previously reported assignments for EVOOs 13C-NMR definitely accounted for five fatty esters in the following quantitative order: oleic (O), palmitic (P), linoleic (L), stearic (S) and linolenic (Ln) [27,28]. Some other tentative assignments concerned mono-unsaturated fatty esters like palmitoleic (PO) and 11-eicosenoic (E) constructs [29]. Despite the wide availability of NMR reports [30], none of these clearly explained the systematic presence of unknown resonances (in our processing batch at 129.92, 129.82, 31.82, 22.69 ppm and others) which account for a relevant quantitative contribution (around 3%, Figure 3). Scientific hesitancy probably owes to the general opinion that the total amount of other fatty esters is limited to less than 1% of EVOOs. This idea was questioning the NMR technique itself as possible analytical method but the serendipitous extraction of the *Capparis spinosa* oil (CSO) allowed us to solve this inconsistency because of the remarkable presence of *cis*-vaccenic (V) esters. The comparison between NMR and GC-FID analyses of CSO consistently confirmed the main presence of the V ester with a minor contribution of the O ester. The analogous analytical approach executed over several EVOO samples made us realize that the detected mono-unsaturated fatty esters were again O and V but in a reversed quantitative proportion respect to the CSO. Against this background, the main NMR resonances attributed to V in other peculiar food matter [31–33] (as also the reported CSO sample) were matching the EVOO signals as reported in Figure 3. It definitely gave us the chance to include the V component in the EVOO quantitative panel according to the 13C-NMR resonances afore mentioned. The V remarkable presence is not just a production side product as we did not observe the presence of *trans* isomers (resonances downfield respect 5.40 ppm in the 1H-NMR and relative other singlets in the 13C-NMR). Once again these results confirm the stability and sound presence of the *cis* form of unsaturated esters in spite of the minor thermodynamic stability. In order to perform the updated comprehensive quantitative NMR analysis of EVOO we have adopted an accustomed procedure based on MARA-NMR. Although it is not the first analytical comparison between GC and NMR [34,35], the novel MARA-NMR strategy suitably refined according to the new information led to a very good fitting (Figure 4). The whole outcome is reported in Tables (Tables 1 and 2, and Tables S1 and S2 in Supplementary Data). In order to ge<sup>t</sup> consistent data, we have chosen to compare the percent presence of L and Ln as detected, whereas the saturated fatty esters (SFA%) were considered as the sum S+P and the mono-unsaturated fatty esters (MUFA) were

considered as O+V+PO. We think it is actually an important parallel evaluation whose general trend shows a very good fitting also kept with samples showing sensibly different proportions. Finally, by properly considering all the mono-unsaturated fatty esters, the MUFA% estimation reached an unprecedented very good matching. On the other hand, the slight systematic overestimation of GC-FID respect to the NMR for L% and Ln% and underestimation of SFA% deserves to be elucidated with further studies requiring standard mixtures similar to EVOO, which is a tri-acyl-glycerol mixture. At the moment, these substrates are not available but work is in progress to develop further information. Although it is not the first case of V detection and also quantification [36], the EVOOs routine quantifications barely evidence the resolution for O-V peaks; this is clearly shown in the GC picture of the European Regulation 2013 [9]. Our observations also demonstrated that new GC-FID columns keep a better (affordable) resolution, whereas routine instruments adopted for serial records easily present the V peak as O shoulder. Fortunately, recorded 13C-NMR provide the missing information about the V fraction (not really taken into account so far) for any EVOO sample (Figure 3). According to our opinion, future studies could take advantage from a "powered" MARA-NMR working over sensitivity-enhanced 13C-NMR profile (optimized scans); these could push further the frontiers of quick qNMR in EVOOs by enabling the independent quantification of fatty esters in the 2- internal position of glycerides but also the improved quantification of other minor components (see Supplementary Materials). This contribution also opens the way toward new studies concerning sensory attributes, geographical origin and beneficial effects [37] of EVOO as fundamental functional food with the major presence of glycerol esters [38].
