3.1. Topping Ingredient Analyses
Table 2 shows the results relating to the qualitative characteristics of the oils used for the pizza topping. The detected free acidity was significantly (
p < 0.01) different, with the lowest values in oil A (0.17 ± 0.05%), followed by oil B (0.30 ± 0.03%) and oil C (0.49 ± 0.01%). These results showed for sample A, as expected, the lowest acidity, due to the deacidification and deodorization processes of the rectified oils that constitute it. The other two samples of oils, B and C, despite having higher percentages of oleic acid, fell well below the maximum limit for extra virgin olive oils provided for by [
12], with better results for sample B. This could be explained by the fact that sample B, being a commercial oil, is obtained from a mixture of extra virgin olive oils with different acidity, while oil C, being monovarietal, is strictly influenced by the characteristics of the exclusive cultivar of origin. From this, the free acidity cannot be the only parameter used to establish the quality of an oil, because it is an index that can be easily manipulated by grinding or mixing processes. To evaluate the level of lipid oxidation following cooking of the various ingredients, the peroxides,
p-anisidine values, spectrophotometric analysis, and induction time were determined. The main products of lipid peroxidation are hydroperoxides, generally referred to as peroxides; therefore, the results the PV parameter give a clear indication of lipid autoxidation [
22]. As for the free acidity, sample A showed the lowest results (5.94 ± 0.04 mEq O
2/kg), due to the submitted treatments, and the other samples (B and C) fell within the limit of 20 mEq O
2/kg fixed by the European Commission for extra virgin olive oil category.
For further confirmation and deepening of these results, other oxidation parameters were evaluated, such as conjugated dienes and conjugated trienes. Although dienoic acids are less in quantity than monoenoic acids in olive oils, early oxidation occurs mainly in dienoic acids. Consequently, it was interesting to examine the relationship between the value of peroxides (PV) and conjugated dienes (% CDA): the ratio between PV and% CDA normally tends to increase in a system in which oxidation is essentially the result of the oxidation of singlet oxygen with the formation of hydroperoxides in conjugated and unconjugated form [
23]. The results on olive oils used in the topping confirmed the differences discussed above denoting the peculiar characteristics of the three samples. These results were reflected in the PV/%CDA of oils, respectively, of 53.05 ± 1.09, 193.62 ± 6.89 and 142.08 ± 14.15. High values of peroxides (primary oxidation) are always an index of low-quality oils, while low values of peroxides do not always indicate good quality [
24]: for this reason, the analyses were deepened with the evaluation of
p-anisidine and the value of TOTOX. The
p-anisidine value (
p-AV) is a more reliable and meaningful test, because it measures the secondary oxidation products, which are more stable during the heating process [
25]. For the first one, parameter of the secondary oxidation, the obtained results in the three samples were found to be consistent with what was found in the CDA% results and, specifically, by the PV/CDA% ratio in the oils used for the garnish of the pizzas. Relative to the determination of
p-anisidine, oil C showed the highest value (10.53 ± 0.00), differentiating itself from oil A and B (respectively, 9.10 ± 0.06 and 9.67 ± 0.07). The results of the TOTOX in the oils revealed the highest results in oil B (45.38 ± 1.55), followed by oil C (33.11 ± 0.26) and oil A (20.98 ± 0.14).
The investigation on the total oxidation index revealed the higher quality of the monovarietal extra virgin olive oil then the commercial extra virgin one: this can be linked to the detected qualitative differences in terms of bioactive compounds, such as phenols. Among the extra virgin olive oils used in the present study, the lower polyphenol values of oil B compared to C may be related to the different species of olives used for the production of oils, whereas oil C is monovarietal, with peculiar and exclusive chemical and sensory characteristics. The ANOVA showed significant differences among the three oils: oil C was characterized by the highest concentration of hydroxytyrosol, tyrosol, pinoresinol and apigenin, and in general total phenol content. Having been subjected to thermal rectification processes that led to an almost total degradation of the polyphenolic compounds of the oil, oil A showed the lowest values, followed by oil B (around 110 mg/kg) and oil C (180 mg/kg). The literature reports different levels of polyphenol content in Ottobratica olive oils: our results are similar to those reported by Sicari et al. [
26]. Only the concentration of
p-coumaric acid did not vary among the three oil samples (
Table 3).
The volatile compounds showed significant quantitative differences among oil types (
Table 4). The olive oil sample, in particular, showed the lowest quantity of volatiles, whereas the extra virgin olive oil Ottobratica cv. showed tenfold the quantity of volatiles of the olive oil. This result was expected, because olive oil is a mix of virgin olive oil (likely in little amounts, hence contributing with little amounts of volatiles), and refined oil, which is submitted, among other processing steps, to a deodorization phase to remove flavors, which unavoidably affects the other volatiles also. The ordinary (multicultivar blend) commercial EVOO had an intermediate content of volatile compounds. A varietal effect on the quali-quantitative profile of volatiles of extra virgin olive oils has been reported in several studies [
27,
28,
29].
The most abundant compound among all the detected ones was (E)-2-hexenal, particularly concentrated in the EVOO from single cultivar Ottobratica Calabrese. This compound has a positive association with olive ripeness and, together with (Z)-3-hexen-1-ol (also abundant in the single cultivar Ottobratica cv. EVOO), is related to a green, grassy note [
30].
1-Hexanol, characterized by a green, herbal flavor, was another compound present in noticeable amounts in the Ottobratica cv. extra virgin oil sample. Together with hexanal (which, however, was present in low and similar amounts in the three oils), it originates via the lipoxygenase degradation pathway of linolenic acid [
31]. Another lipoxygenase derived volatile was 2,4-hexadienal, having a fatty, sweet, green odor. Nonanal, a typical oxidation marker with a waxy odor and associated with sensory defects, was almost absent in all the oil samples [
32,
33]. Its absence in the two EVOO oils was due to their good quality, while the absence in olive oil was probably imputable to the deodorization carried out during refining.
Another compound present in relevant amounts in the Ottobratica cv. EVOO was 6-methyl-5-hepten-2-one. Together with benzaldehyde, this compound has been reported in unfiltered EVOO, where it tends to increase with storage, and has been related to the activity of exogenous microorganisms and eventually associated to sensory defects such as musty [
34,
35]. The presence of a rich residual microflora has been reported in freshly extracted EVOO, being mainly formed by yeasts entrapped in the solid particles and the microdroplets of vegetation water dispersed in the oil phase [
36]. While filtration at the end of the extraction process largely removes this suspended material, when the olive oil is not subjected to filtration, the residual microflora may remain partly active in the oil or in the sediment that gradually forms at the bottom of the container, and may significantly contribute to chemical and sensory changes of the oil throughout storage [
36]. The Ottobratica cv. EVOO oil, indeed, was not immediately filtered, hence the contact with olive cell residues transferred this volatile compound to the oil. Hexyl acetate was also present, as a derivative of 3-hexen-1-ol.
Table 5 shows the main qualitative analyses carried out on the sample of S. Marzano tomatoes used for the garnish of the Neapolitan pizza. The total acidity was 0.43 ± 0.02% citric acid, total soluble solid content 7.75 ± 0.21° Bx and pH 4.20. The total polyphenol content was quantified in 48.62 ± 0.10 mg/kg of gallic acid with chlorogenic (12.01 ± 0.06 mg/kg), protocatechuic (10.89 ± 0.04 mg/kg) and ferulic (2.74 ± 0.08 mg/kg) acids and rutin (9.09 ± 0.08 mg/kg) as the main phenolic compounds identified on UPLC analysis. The results obtained from the various physicochemical analyses carried out on the tomato showed a good quality and level of antioxidant compounds. In fact, the detected total acidity was completely average for a quality tomato. The total soluble solid content showed a value higher than the minimum required (5° Bx) by the canning industries, and the pH was an index of safety. Flavor is generally related to the relative concentrations of sugars and acids in fruit, especially fructose and citric acid. The best and tastiest combination is a high sugar content and a high acid content. A normal pH range in tomatoes is between 4.0 and 4.5, and the lower the pH, the sweeter or bitterer the fruit will be. A good concentration of acids and sugars correlates with an optimal pH, and can suggest good organoleptic properties of the product as well as analytical ones. The chroma analysis indicates the fullness of color and gave an average result of 9, with positive values also for parameter a* (color variation from green to red), probably given by the presence of carotenoids. The content of chlorogenic acid showed a good level of maturation and processing following a harvest that took place over adequate time [
37].
3.2. Pizza Topping Mix Analyses
Figure 1 shows the results of the total content of polyphenols in pizza topping mix (tomato + the three different oils) before and after cooking. The three toppings reflected the previously mentioned analytical results for polyphenol content: the topping with oil C in fact was the richest on bioactive compounds (55.24 ± 0.04 and 72.01 ± 0.90 respectively). All the samples showed significant differences from one another.
Regarding the qualitative analysis, the most present compound in the fresh and cooked topping was the protocatechuic acid which was around the value of 9.90 mg/kg for the three toppings before cooking, while it was found in higher concentrations (10.51, 10.79 and 10.51 mg/kg) for the three toppings after the cooking of the pizza, with a slightly higher value in the topping with oil B (
Table 6). An increase in post-cooking of rutin is also observed in both topping A and C. In general, there are no substantial differences between the different toppings, but some minimal differences following cooking.
The qualitative and quantitative determination of phenolics in pizza topping mix before and after cooking showed a great influence of the phenolic compounds from tomato with respect to those of the oil: this is mainly due to the interaction between the various compounds of the tomato and the oil promoted precisely by the cooking process. Pernice et al. [
38] reported that the heating of a mixture of tomato and olive oil (5%) determined a protective effect by the oil on the antioxidants of the tomato, precisely due to the interaction between the bioactive compounds of the one and the other ingredient. Cooking also increases the disposal of polyphenols [
39], which were detected at higher content, and so increased the functional properties of the three pizzas tested. Significant differences were observed in particular for the content of rutin, which is included in dietary adjuvants such as vitamin
p and is indicated for various health-promoting effects [
40].
With regard to the analysis of the lipid fraction extracted from the pizza topping after cooking, the resulting peroxide value indicated the topping with oil C of better quality than those composed by oil B. From the
p-anisidine value, these results were regarding also the topping A after cooking (
Figure 2).
The results relating to the lower oxidation state of topping C after cooking led to the consideration that this is probably due to the greater presence and better quality of bioactive compounds (polyphenols, carotenoids and chlorophylls) which, as is known, help prevent and slow down such phenomena. Oxidants, including secondary ones, are reported in
Figure 2 for
p-anisidine values. On the other hand, considering the lipid fraction extracted from the cooked topping, the
p-anisidine value, as well as the primary oxidation (PV/% CDA) is much lower in the fraction composed of oil C, while it is higher in A and B. This leads us to understand how, despite higher initial values as regards
p-anisidine, peroxides, % CDA and the PV/CDA% ratio in extra virgin olive oils B and C, after cooking, topping C (composed of tomato and oil extra virgin olive oil C with good levels of bioactive compounds to protect the oxidative advancement of lipids) was found to be qualitatively better than the other two oils (
Figure 3). Therefore, by analyzing the TOTOX value, which considers primary and secondary oxidation (peroxides and
p-anisidine), it is clear that the results of the analysis carried out on crude oil are in line with what was previously observed, i.e., the best result in sample C in terms of lipid oxidation level.
3.3. Volatile Compounds of Pizza Samples
Table 7 reports the volatile compounds of pizza seasoned with tomato sauce and the three different types of oil considered in the typical marinara-style pizza. The volatile profile of pizza was much richer compared to the initial oils due to the presence of additional ingredients (flour, yeast and tomato sauce), and due to fermentation activities, thermal reactions and lipid oxidation occurring during pizza processing. However, the three pizza types still showed some significant differences as a function of the type of oil, especially in the levels of 1-hexanol, (Z)-3-hexen-1-ol and (E)-2-hexenal, which decreased due to oxidation (in favor of the formation of other aldehydes and hexanoic acid) [
41] but whose levels maintained the same trend observed in the starting oils. In addition, hexanal, nonanal and hexanoic acid, deriving from lipid oxidation, increased compared to the starting oil but were less concentrated in the Ottobratica cv. EVOO, which contained more phenolic compounds with known antioxidant activity.
As for the compounds originating via fermentative activities, ethanol, 2-methyl-1-propanol and 3-methyl-1-butanol are typical yeast products, produced during dough leavening [
42]. Relevant amounts of 5-hepten-2-one-6-methyl were detected in all pizza samples, irrespective of the type of oil added. This compound, already present in the starting oil, is present also in tomato products, where derives from the degradation of lycopene; therefore, its increase compared to the starting oil could be due to the tomato sauce used to season the pizza samples [
43]. 2-Butanone3-hydroxy, instead, absent in the starting oil, was another fermentation-originated compound [
44]. Neither 5-hepten-2-one-6-methyl nor 2-butanone3-hydroxy showed a significant difference among the three pizza types.
As for the compounds related to thermal reactions, due to the typically high temperature of baking adopted to prepare pizza (approximatively 300 °C), this class of compounds comprised 2-methylbutanal, 3-methylbutanal, 2-furancarboxaldehyde (furfural), 5-methyl-2-furancarboxaldehyde (5-methylfurfural), methylpyrazine, ethylpyrazine, 2-pentylfuran, 2-furanmethanol, 2-furanylethanone. All these compounds, absent or present at very low amounts in the three types of oils considered, derived from the Maillard reaction and are commonly reported in the volatile profile of bakery products, being associated with caramel, bread-like flavor [
45,
46,
47,
48].
3.4. Polar Compounds of the Oil and Pizza
Table 8 shows the content of polar compounds of the oils and of the corresponding pizza samples. These compounds arise from the oxidation and hydrolysis of lipids and their analysis is an effective means to evaluate the quality of any lipid [
49,
50]. The oxidation products, in particular—namely, triacylglycerol oligopolymers (TAGP) and oxidized triacylglycerols (Ox-TAG)—are implicated with the alteration of the nutritional properties of foods and may cause adverse physiological effects [
51]. The diacylglycerols (DAG), instead, originate from lipid hydrolysis.
The oils showed different levels of polar compounds, with the highest values of TAGP and DAG in olive oil, confirming its lower-quality categorization. Between the two EVOOs, the single cultivar Ottobratica cv. oil was of higher quality. The multicultivar EVOO showed the highest content of Ox-TAG.
The results of pizza samples show that a relevant increase of each class of polar compounds occurred during processing, but the difference among oils was maintained, e.g., the less degraded lipid fraction was observed in pizza seasoned with Ottobratica cv. olive oil, which contained higher levels of phenolic antioxidants. Indeed, the preparation of bakery products has been reported to induce the oxidative degradation of lipids [
52,
53,
54].
The observed results were similar to those reported in focaccia, an Italian bakery product similar to pizza [
55].