3.1. Ultrasound Assisted Extraction (UAE)
The performance of the ethanol–water–hydrochloric acid system in the extraction of polyphenols from olive pomace (sample O1) and lees (sample W4) samples was evaluated by UAE. Both water and ethanol are well-suited solvents in terms of applications involving food, cosmetic, and nutraceutical industries, and water–ethanol mixtures are widely proposed in the literature for the extraction of polyphenols from different matrices. It has been also reported that, in general, the acidity of the medium has a positive effect on the polyphenol extraction yield [41
], so hydrochloric acid was included as a variable to be assessed.
In this study, ethanol percentage was varied 40–80%, while the concentration of hydrochloric acid was varied 0–0.5%, and experiments were performed according to the procedure described in Section 2.4.4
Results are given in Table S1 (Supplementary Material)
, and response surfaces for TPC extraction, expressed in terms of mg GAE g−1
sample, are shown in Figure 1
. In the case of olive pomace (Figure 1
a), a quite flat surface was obtained, indicating that the composition of the solvent had a minor influence on the extraction within the explored range. ANOVA study of the results (Table S5, Supplementary Material
) confirmed that, under the conditions tested, there was no significant effect of the concentration of either ethanol (p
= 0.06) or HCl (p
= 0.90) on the extraction recovery.
In the literature, the extraction of polyphenols from olive pomace by UAE has been carried out using diverse percentages of ethanol, up to 90% [25
]. Results presented here show that, in the explored ranges, the percentage of ethanol was not a critical issue, and hydrochloric acid was not necessary to enhance extraction.
Conversely, in the case of lees filters (Figure 1
b), there was a clear influence of ethanol concentration on TPC recovery. Thus, extraction improved significantly (p
= 1.2 × 10−10
) when the concentration of ethanol was above 40%. This is in accordance with the 50% EtOH proposed by other authors for polyphenol extraction from grape pomace [39
] and grape skin [40
]. On the other hand, there was no influence of the HCl concentration (p
= 0.64), which agrees with results reported by Bachtler and Bart [41
], about the effect of HCl concentration on polyphenol extraction from vine leaves.
Taking into account that HCl did not show a significant effect on polyphenol extraction by UAE, neither from olive pomace nor lees filters, subsequent studies on MAE and PLE focused on the water–ethanol system, without considering the addition of HCl.
The different behavior of the two matrices was attributed to the differences in polyphenol composition, which is connected to the chromatographic profiles of both kinds of samples. Chromatograms of the olive pomace extracts (Figure 2
) were more complex than those of lees extracts (Figure 3
). In olive pomace, extracts were rich in compounds with a wide range of polarities, so that increasing the ethanol percentage in the extraction solvent may have contributed to improvement in the recovery of less polar compounds while decreasing the recovery of polar ones. The presence of oleuropein and luteolin, characteristic compounds in olives, was confirmed by UHPLC–HRMS, as well as 3-hydroxytyrosol, caffeic acid, p
-coumaric acid, and rutin. In contrast, lees extracts were abundant in medium polarity compounds, which were better extracted when the percentage of organic solvent was increased. The presence of gallic acid, caffeic acid, hesperidin, resveratrol, and quercetin was confirmed by UHPLC–HRMS.
3.2. Microwave Assisted Extraction (MAE)
Polyphenols extraction by MAE from olive pomace and lees filters was investigated using water–ethanol mixtures. Conditions reported in the literature for the extraction of polyphenols by MAE are diverse. For instance, Habibi et al. [26
] proposed a mixture of EtOH:water 60:40 (v/v
) as extraction solvent, 220 W of microwave power, and 12 min of extraction time, whereas Jurmanović et al. [27
], established as optimum MAE conditions EtOH:water 20:80 (v/v
), 700 W of microwave power, and 10 min of extraction time.
In MAE, apart from issues dealing with the affinity among solvent and analytes, the polarity of the solvent plays an important role, since the absorption of microwave radiation is more efficient in polar media, and thus the extraction recovery might be improved. In this study, ethanol was varied between 20 and 80%, a wider range than in UAE. For temperature, the range was 60–120 °C, and for extraction time, 5 and 15 min were considered. Experiments were performed following the procedure described in Section 2.4.5
Results are shown in Figure 4
and in Table S2 (Supplementary Material)
. In the case of olive pomace (Figure 4
a,c), TPC improved by increasing the ethanol concentration from 20 to 50%, and temperature from 60 to 90 °C. However, when increasing ethanol concentration from 50 to 80%, some decrease in TPC values was observed; at 120 °C TPC decayed, which was attributed to degradation of polyphenols in the samples at higher temperatures. ANOVA results confirmed that there was a significant effect of ethanol concentration (p
= 1.1 × 10−9
) and temperature (p
= 8.5 × 10−5
) on the extraction of polyphenols. Conversely, regarding extraction time, no significant improvement was observed when increasing time from 5 to 15 min (p
In the case of lees filters (Figure 4
b,d), a similar trend with respect to olive pomace samples was observed. The TPC improved with the increase of ethanol concentration, from 20 to 50%, and with temperature, from 60 to 90 °C. Nevertheless, by increasing the ethanol concentration from 50 to 80%, TPC remained almost the same, and at 120 °C, TPC decreased because of thermal degradation. Concerning extraction time, no significant effect was observed when increasing time from 5 to 15 min. ANOVA (Table S5, Supplementary Material
) confirmed that both ethanol concentration and temperature had a significant influence on the polyphenol extraction (p
= 2.4 × 10−8
= 1.0 × 10−4
, respectively), but not extraction time (p
= 0.18). Garrido et al. [32
] proposed the use of 48% ethanol in water and 10 min of extraction time at 25 °C for MAE extraction of phenolic compounds from Chardonnay grape marc; they also reported degradation at high temperatures, which is consistent with our results.
3.3. Pressurized Liquid Extraction (PLE)
Finally, polyphenol extraction by PLE from olive pomace and lees filters samples using water–ethanol mixtures was explored.
EtOH percentage and temperature were varied at three levels (ethanol: 40, 60, and 80%; temperature: 80, 100, and 120 °C), with an extraction static time of 5 min and 1 cycle, with a total of 9 experiments for each matrix, which were performed in triplicate, following the procedure described in Section 2.4.6
Results are shown in Figure 5
and in Table S3 (Supplementary Material)
. In the case of olive pomace (Figure 5
a), results showed that from 40 to 60% ethanol, there was some improvement of the extraction, but at 80% EtOH, there was a decrease in TPC. Regarding extraction temperature, no relevant influence was observed. ANOVA results confirmed that under the conditions tested, the effect of the ethanol concentration on TPC extraction was significant (p
= 9.2 × 10−9
), while the influence of temperature was irrelevant (p
These results agree with extraction conditions reported by other authors for PLE extraction of polyphenols from olive pomace [19
] or olive leaves [20
], which proposed ethanol:water 50:50 (v/v
) as extraction solvent, and different temperature conditions (120 °C for pomace and 80 °C for leaves).
A range of conditions for the PLE extraction time or the number of cycles can be found in the literature. For instance, Putnik et al. [20
] proposed 2 cycles of 5 min of extraction time (ethanol:water 50:50 (v/v
); 80 °C) for polyphenol extraction from olive leaves, whereas Xynos et al. [14
] applied 3 extraction cycles of 5 min (EtOH at 190 °C). In this study, PLE experiments with 1 and 2 cycles, and 5, 10, and 15 min extraction times were performed (ethanol:water 50:50 (v/v
), T = 100 °C).
Results are shown in Table S4 (Supplementary Material)
. For olive pomace, it was concluded that one extraction cycle was enough, since no advantages were observed when adding an extra cycle (p
= 0.57). Regarding extraction time, there was an effect on TPC (p
= 0.02); the highest TPC values were obtained at 5 min. The decrease of TPC at longer extraction time was probably due to a degradation of polyphenols. In this sense, Putnik et al. [20
] also observed a TPC decrease with increasing extraction time.
In the case of lees filters (Figure 5
b), TPC increased with ethanol concentration from 40 to 60% but decreased at 80%. Regarding extraction temperature, TPC increased from 80 to 100 °C and decreased at 120 °C, except for 80% EtOH. ANOVA tests confirmed that ethanol concentration and temperature had significant influence in the polyphenol extraction (p
= 1.7 × 10−7
= 7.4 × 10−3
, respectively). No significant effects either of the number of cycles or extraction time (p
= 0.07 and p
= 0.39, respectively) were found.
3.4. Extraction of Polyphenols from Olive Oil Mill and Winery Wastes
Taking into account the results of the previous extraction experiments, the ethanol:water 50:50 (v/v
) mixture was selected as the extraction solvent for both kinds of residues and the three techniques. Table 2
summarizes the proposed experimental conditions for TPC extraction. Finally, it was decided to apply the three techniques, under the selected conditions, to a set of diverse residues from olive oil and wine companies.
a shows the results obtained for samples related to the olive oil sector. In a global sense, there were not large differences between the results of the extraction techniques, although it can be observed that MAE was the most efficient. ANOVA analysis of the results (Table S6, Supplementary Material
) confirmed the significant differences between the three extraction techniques (p
= 9.3 × 10−8
) under the studied conditions and also pointed out that there was interaction between samples and techniques (p
= 2.3 × 10−4
), i.e., the performance of the technique depended on the sample. With regard to winery wastes, again there were not large differences between the performance of the extraction techniques (Figure 6
b), but PLE provided the highest TPC values, except for the W4 sample, for which MAE was the most efficient. ANOVA (Table S7, Supplementary Material
) confirmed the significant differences between the extraction techniques (p
= 5.7 × 10−7
) and that there was sample–technique interaction (p
= 6.2 × 10−10
). Compared with similar cases using ethanol:water 50:50 (v/v
), Drosou et al. [39
] and Caldas et al. [40
] found that UAE was more efficient than MAE for the extraction of polyphenols from grape pomace and grape skin samples, respectively.
Overall, in this study MAE provided higher efficiency for olive oil wastes and PLE for winery wastes, but results from UAE were also satisfactory. In this context, Talmaciu et al. [45
], in a comparative investment costs study considering different extraction techniques, such as MAE and UAE and supercritical fluid extraction, concluded that UAE is the one that requires lower capital and operational costs. Instead, in a study about polyphenols extraction from red wine pomace, Vega et al. [46
], carried out a techno-economic and life cycle assessment, and concluded that PLE not only had higher capital expenses, but also higher operational expenses and environmental concerns. Hence, considering extraction performance and simplicity, but also investment and operational costs, we propose UAE for further scaling up.
Focusing on the phenolic yield of the different samples, it can be noticed that the phenolic content in olive pomace samples was, in general, higher than that of winery residues. Additionally, there were significant differences between the yields of olive pomaces (p
= 1.1 × 10−22
) and also between the yields of winery residues (p
= 1.7 × 10−23
). Concerning olive pomace, since all the samples were processed by using similar treatment stages, the differences in yields were mainly attributed to geographical and varietal issues. Olive oils from southern Spanish regions are richer in phenolic compounds due to the higher levels of hydric stress induced by the more severe climatic conditions [47
]. In addition, oils from varieties such as picual are noticeably richer in polyphenols than arbequina counterparts, with concentrations ca. two-fold higher [48
]. The behavior found in olive oils can be reasonably extrapolated to their corresponding residues. This agrees with the results found here, i.e., samples from northern Spanish areas (O3, O4) and/or those produced with arbequina olives contain lower phenolic concentrations. Regarding winery residues, the yield for sample W3 (lees filters, white wine, chardonnay, sauvignon blanc, and xarel·lo varieties) was clearly lower than those of the rest of the winery samples, which is related to the lower phenolic contents of lees from white wines compared to lees of reed wines [49
]. Conversely, the phenolic yield of sample W4 (lees filters, red wine, garnacha, tempranillo, cabernet sauvignon, and cariñena) was the highest, in the range of winery residue samples, and thus the filters of lees of red wine production appeared to be a good source for the recovery of polyphenols.
Finally, we evaluated the antioxidant activity of extracts from the samples using Folin–Ciocalteu and ABTS assays [50
]. The results, as well as those from HPLC, are collected in Table 3
There was no clear correlation between the results of the different methods. The main reason is that the three methods were based on different approaches. The results of HPLC–UV were obtained from the absorbance measured at 280 nm, which in these types of samples was mainly due to polyphenols. The Folin–Ciocalteu method is based on a redox reaction, while the ABTS method is based on a radical reaction, and in both assays, different polyphenols show different sensitivity [42
]. In any case, all the assays indicated that extracts of residue samples from olive oil production had higher antioxidant activity levels than the extracts of the winery residue samples, with the exception of sample W4.