Two grapevine varieties, a red (Aglianico, AG) and a white (Greco di Tufo, GR) one, were used to investigate the effect of adding different copper (II) levels (high: 20 mg/L, low: 10 mg/L) to musts before fermentation. In order to avoid possible copper (II) residues on grape surfaces coming from previous in-field treatments, harvested grapes were accurately washed and dried before crushing. This procedure did not remove possible copper (II) naturally occurring inside the berry and likely depending upon the soil condition. The effects on fermentation progress, acetaldehyde content, color characteristics, vanillin-reactive flavans (VRF), BSA-precipitable tannins, polymeric pigments, and total phenols were evaluated during fermentation, at the end of alcoholic fermentation, and after 12 months of aging.
2.1. Fermentation Evolutions and Copper (II) Evolution during Fermentations
The alcoholic fermentation process differed in duration between the red and white grapes. As shown in
Supplementary Figure S1A, the red Aglianico must completed fermentation in 19 days, while the white Greco di Tufo must required a longer fermentation period up to 34 days (
Supplementary Figure S1B). This difference is likely attributable to the varying nitrogen content in the musts. Red grape must, containing both skin and seeds in addition to pulp, usually has a higher concentration of nitrogen sources compared to white must, which typically consists of pulp alone [
20]. Since nitrogen plays a crucial role in yeast metabolism during fermentation, the lower nitrogen content in the white grape must likely delayed the onset of the exponential fermentation phase, where yeast growth and sugar conversion rapidly increase.
During alcoholic fermentation, the copper content was monitored, and a comparison between the samples is shown in
Table 1 and
Table 2. It is important to underline that time zero does not represent the baseline for copper (II) content evaluation, as the copper (II) analyses were carried out approximately four hours after sampling, as detailed in Materials and Methods. This delay may have resulted in an apparently non-cumulative effect, since some of the copper (II) may have started precipitating and/or being absorbed by yeasts before the analyses [
12,
21,
22]. The analysis at time zero for red grape Aglianico (
Table 1) showed an average copper (II) concentration of 5.35 ± 0.64 mg/L. These concentrations are lower than those reported in the literature by Fregoni and Corallo [
23], where the total copper (II) residue concentration in grapes ranged from 10 to 56 mg/L. As in our study, grapes were washed before winemaking to eliminate each possible residual copper (II), this first datum confirms the efficacy of the washing step carried out on grapes before crashing especially if copper (II) is used in vineyards. The levels found at zero time for AG10 and AG20 are consistent with the amount of copper (II) added at the beginning of the experiment. As expected, significant differences (
p > 0.05) were found between control and treated wines at time zero as well as at day 1, but no significant differences were found between control and treated samples from day 5 to the end of alcoholic fermentation. Our results showed a range of residual copper (II) at the EAF from 0 to 1 mg/L, in agreement with results previously observed for Cabernet Sauvignon [
18]. These results highlighted that, even if there are high amounts of copper (II) in red grapes, after alcoholic fermentation, the copper (II) values fall below the threshold limit of 1 mg/L, in agreement also with current EU regulation for wines [
24].
The same behavior during fermentation was observed for the white Greco di Tufo variety (
Table 2). At time zero, the average copper (II) concentration in control wine was 3.94 ± 0.27 mg/L. Unlike the Aglianico variety, we could observe (
Table 2) a significant difference between treated and control samples throughout the alcoholic fermentation with a constant higher content of copper (II) in GR20 and GR10 with respect to the control samples. This trend was not observed for the red variety, which differs from the white one by the presence of pomace during fermentation. Probably, the pomace induces mineral adsorption phenomena by subtracting copper (II) from the aqueous medium [
25]. At the end of alcoholic fermentation (EAF), unlike Aglianico samples, GR10 and GR20 showed the highest residual amount of copper (II), exceeding the EU legal threshold limit value of 1.0 mg/L (until 1.66 ± 0.04). For both fermentations, a greater loss of copper (II) was detected in the last phases of fermentation, probably because in these steps, the yeast cells acted as adsorbent material for this element. Previous researchers also showed that during fermentation, a large proportion of copper (II) is precipitated with yeast cells [
22,
26,
27]. These results are of great interest because the control of the copper (II) level is of critical importance for wine, because copper (II) acts as catalyst of chemical oxidation reactions, and its catalyst activity in wines is dependent upon concentrations, as shown by Cacho et al. [
28] in an experiment in which they evaluated the levels of copper (II) ranging from 0.58 to 8.95 mg/L. Concerning microbial evolution during the fermentation of Greco di Tufo and Aglianico grapes (
Supplementary Figure S3), a significant inhibitory effect of copper (II) on non-
Saccharomyces yeasts was observed for Greco grapes. This effect was proportional to the amount of copper (II) added. Albeit not relevant to our experiment, in which selected yeasts were used, these findings could either have interesting implications for non-inoculated musts or for fermentation in which non-
Saccharomyces yeasts are used. As a consequence of the copper (II) effect,
Saccharomyces yeasts may gain a greater competitive advantage over non-
Saccharomyces ones in all fermentations, potentially establishing dominance over a shorter period of time.
2.2. Acetaldehyde Production during Fermentations
In this study, the monitoring of acetaldehyde was carried out with a double purpose: aldehyde is an important secondary metabolite from microbial activity, and it is also the main product of the chemical oxidation of wine [
29].
Analyses of acetaldehyde were carried out during wine fermentation. In red wines (
Figure 1A), acetaldehyde increased (day 1) and then decreased (day 5) until the end of alcoholic fermentation. This trend is well known and common to all fermentations: acetaldehyde increases during the first days of fermentation [
30] and successively decreases due to its reduction to ethanol by NADH [
31] as well as to its involvement in reactions with wine components including phenolic compounds [
32,
33]. In regard to the possible effect of copper (II), previous research suggests that higher copper (II) concentrations can stress yeasts, leading to increased acetaldehyde production [
34]. Therefore, we would have expected the treated wines (AG20 having the highest copper (II)) to have the highest concentrations of acetaldehyde, but no statistical differences between treated and untreated wines were detected during the first phases of the fermentation process. This is probably due to the low level of copper (II) in treated wines with respect to the study of Liang and Zhou [
34]. At the EAF, however, the control wine showed a higher content of acetaldehyde compared to the treated samples (AG10 and AG20). This finding is quite surprising, but it can be explained by taking into account the higher reactivity of acetaldehyde towards phenolic compounds [
35]. In addition, when the role of acetaldehyde in phenolic compound solubilization and precipitation during winemaking was evaluated, Teng et al. [
36] demonstrated that the acetaldehyde-mediated pathway predominated in pigment formation, thus suggesting that all conditions during winemaking which give rise to acetaldehyde appear to be important factors involved in the formation of stable, soluble non-bleachable pigments. Since all these reactions occurred during wine oxidation, well-known evidence that copper (II) and iron accelerate the reaction with oxygen [
37] explains the observed trend for copper (II)-treated wines.
Regarding the Greco di Tufo wines, during the first fermentation phase (
Figure 1B), all the experimental samples did not show any statistical difference in terms of acetaldehyde content up to day 15. However, starting from day 15, fermentations appeared to struggle, and the rate of sugar consumption quickly slowed down (
Supplementary Figure S1). This suggests that the yeasts underwent stress around that time. Several factors could have contributed to this stress response. After day 15, the absence of oxygen and essential survival factors for yeasts [
38], combined with the increasing ethanol concentration in the fermenting must [
39], likely limited yeast viability and slowed alcoholic fermentation. It is important to note that this effect is more pronounced in white grape musts compared to red ones. The presence of skins and pulps in red wines provides yeasts with additional nutrients and precursors for essential survival factors, potentially delaying or mitigating the stress response observed in white grape fermentations. This slowdown could explain the observed higher acetaldehyde content in control wine with respect to GR10 and GR20.
At the EAF, the control sample still showed a slightly but significantly higher acetaldehyde content compared to the GR20 treatment, probably because copper (II) may have a positive effect on yeast metabolism [
40]. Further investigation is needed to elucidate the observed results.
2.3. Effect of Copper (II) on Phenolic Composition and Chromatic Characteristics of Red and White Wine during Fermentation
To understand the impact of copper (II) on the red wine fermentation process, color intensity (CI), hue, monomeric native anthocyanins (
Supplementary Table S4), total native anthocyanins, and polymeric pigments were evaluated through HPLC and spectrophotometric analysis, as shown in
Table 3.
During the fermentation process, the evolution of CI and of polymeric pigments was monitored in all samples (
Table 3). A general increase in CI was observed over time. This is due to the release of phenolic compounds by the skins and seeds, especially anthocyanins absorbing at 520 nm, and to the chemical reactions leading to the formation of polymeric pigments [
33]. These latter are formed during maceration and wine aging [
41] thanks to the reactions of condensation between flavan-3-ols (including catechin, epicatechin, proanthocyanidins) and anthocyanins either mediated or not by acetaldehyde [
33,
42]. At the EAF, significant differences among samples were detected. A decrease in color intensity and the total native anthocyanins with an increase in the hue was observed in samples with major amounts of copper (II) (AG20). Probably, the effect of copper (II) as a catalyst for oxidation reactions led to the loss of these compounds. This behavior was not seen in sample AG10. Interestingly, the analysis of polymeric pigments at the EAF revealed a statistically significant difference in the sample with the most copper (II) (AG20) compared to the control. The lower acetaldehyde content in the AG20 sample (
Figure 1A) might have played a role in influencing the overall color component profile, potentially affecting the formation or stability of certain pigments besides the observed increase in polymeric pigments. Ćurko et al. [
43] found comparable outcomes regarding anthocyanins in the initial months of aging following the micro-oxygenation of wines with added metals, which supports the evidence that metals would promote increased ethyl-bridged polymerization reactions, as suggested by Danilewicz [
44].
The analysis of chromatic characteristics was completed by the determination of ΔE which allows us to measure the color perception by human eyes (
Figure 2). ΔE represents the color variation between the control and treated wines, and values greater than two (>2) CIELab units indicate that wines show a difference detectable to the human eye [
45]. ΔE between the control and each sample was higher than 2 only at the beginning of AF. In the last phases of AF, we observed a decrease in ΔE to values lower than 2.
During AF, all the parameters related to phenolics increased until the EAF (
Table 4). Three different groups of phenolic substances were determined by evaluating their specific reactivity: (i) Vanillin-reactive flavans (VRFs); this parameter is basically correlated with low-molecular mass proanthocyanidins, including dimers, trimers, and tetramers of flavan-3-ols that contribute to wine bitterness [
46,
47]; (ii) BSA–tannin interactions involve high-molecular weight proanthocyanidins. These proanthocyanidins can be precipitated by BSA, and the degree of precipitation increases with increasing polymerization (or size) from trimers to octamers. This suggests that any changes in tannin composition and size can affect their ability to bind with BSA [
48]. Finally, in order to have an overview of the phenolic fraction, we also evaluated (iii) the total phenolic fraction by resorting to the reaction with iron. In the AG20 wines, VRFs and BSA-reactive tannins were significantly lower than in both AG10 and the control. In regard to the total phenol amount, a clear trend as a function of the copper (II) amount was not detected. This could be due to the fact that phenols can undergo hydrolytic cleavage in the presence of oxygen with consequential formations of new intra- and inter-molecular bonds that can alter the reactivity towards the reagents used for their determination [
49,
50].
Regarding the Greco di Tufo wine, a decrease in HCAs was observed during alcoholic fermentation in all samples (
Table 5). HCAs are the major phenols in grape juice and the major class of phenolics in white must. These compounds are also the first to be oxidized by polyphenol oxidase during the first phase of winemaking, and, subsequently, they initiate browning reactions. These reactions can be highly deleterious for the color of white wine [
51]. A parameter that can be correlated to the phenolic compounds and oxidation of white wines is the absorbance at 420 nm (Abs420 nm). This is considered an indicator of the degree of browning of musts and white wines during alcoholic fermentation and storage [
52]. The lowest value in Abs 420 nm was registered in all the treated and untreated samples at the end of alcoholic fermentation. This trend may be related to the possible polymerization and precipitation of quinones leading to brown pigments [
53], as well as their potential absorption by the yeast cell wall [
54]. At the EAF, GR20 showed a higher amount of these compounds. Probably, the higher content of copper (II) induced a higher level of the oxidative process and hence an increased production of xanthylium cation pigments [
55,
56]. Regarding the total phenol content, no significant differences were observed between zero time and the EAF.
2.4. Effect of Copper (II) on One-Year Aged Wines
After 12 months of aging, all samples were analyzed in order to quantify the residual copper (II) in wine and lees as well as to identify changes affected by copper (II) on acetaldehyde, total phenols, BSA-reactive tannins, and vanillin-reactive flavans.
Regardless of the initial copper (II) content, after 12 months of aging, the content of copper (II) is below the EU threshold limit of 1 mg/L (from 0.17 mg/L to 0.20 mg/L for Greco and from 0.01 mg/L to 0.06 mg/L for Aglianico). No significant differences among treatments were observed. The only difference between the white and red samples was the copper (II) content. Aglianico showed a lower copper (II) content compared to Greco di Tufo. Of particular interest is the amount of copper (II) found in the lees: the amount was different for white and red wines. It ranged from 124 mg/kg to 182 mg/kg for Greco and from 226 mg/kg and 361 mg/kg for Aglianico. Probably, the higher content of lees in red wines and the presence of more solid residues in them are responsible for the higher level of copper (II). Our data are consistent with a review concerning the possible re-use of wine lees, which reported a wide range for the copper (II) content in wine lees (13–1187 mg/kg) [
57]. It is therefore important to consider the content of this metal in lees, especially when used in extraction protocols to extract valuable components from them.
In
Figure 3, the concentration of acetaldehyde (
Figure 3A), the sum of polymeric pigments (
Figure 3B), BSA-reactive tannins (
Figure 3C), and vanillin-reactive flavans (
Figure 3D) at the EAF and after 12 months of aging are shown. The decline in acetaldehyde, which possesses an electrophilic carbonyl group, is due to its numerous reactions with a number of nucleophiles naturally occurring in wine, including alcohols, thiols, sulfur dioxide (SO
2), and flavonoids [
32]. Regarding polyphenols, acetaldehyde can react with anthocyanins and flavanols to form methylmethine-bridged compounds [
58], which can react with additional acetaldehyde, anthocyanins, and flavanols to generate polymeric-type structures that can alter wine sensory attributes [
59]. Additionally, the reaction between anthocyanins and acetaldehyde can lead to the formation of stable pyranoanthocyanins [
60]. In
Figure 3B, the concomitant increase in polymeric pigments can be easily correlated to acetaldehyde, VRF, and BSA-reactive tannin loss. VRFs, which are mainly proanthocyanidins with low molecular masses, Ref. [
61], decreased in all of the samples after 12 months of aging.
After 12 months of aging, the loss of acetaldehyde was detected in all samples except AG20 with respect to the EAF (
Figure 3A). This could be ascribed to the presence of copper (II), as shown by Kontoudakis et al. [
62] who demonstrated that oxidative storage increases the most oxidative catalytic form of the metal, and as a consequence, the production of acetaldehyde and its further reactions during wine aging should be favored. This possible explanation seems to be in apparent contrast with the fact that the formation of polymeric pigments over time is higher in the control than AG20. Several factors, such as the anthocyanins–tannins ratio, the presence of acetaldehyde and other reactive carbonyls, and phenolic acids [
63,
64], may determine a different reactivity of monomeric anthocyanins in the formation of polymeric pigments over time. The lower content of polymeric pigments in AG20 (
Figure 3B) could be explained considering the degradation of ethyl-linked flavanols over time [
65]. It is well known that ethyl-bridged compounds are rapidly formed in oxidative conditions [
66,
67], but they also rapidly broken down. Zhang et al. [
68] showed that different classes of anthocyanin derivatives had different distribution patterns as a function of red wine age, and pyranoanthocyanins (except for vitisin B) were deemed to be more stable than the flavanol-related anthocyanin derivatives. In general, pinotins were viewed as the most stable pigments, followed by Flavanyl-pyranoanthocyanins and vitisin A. In a more recent paper, Zhang et al. [
69] also showed that oxidation facilitated the formation of pyranoanthocyanins. Since a higher concentration of copper (II) accelerated all reactions with oxygen [
37], this hypothesis can be considered as a possible explanation for the lower levels of polymeric pigments (with a molecular size ranging from dimers to oligomers) in AG20. Also, Ćurko et al. [
43] observed that in a Plavac mali red wine that was micro-oxygenated and had copper (II) and iron added, the concentrations of tannins and flavan-3-ols were lower than in not treated wines, indicating that the addition of iron and copper (II) can slightly accelerate a decrease in these compounds during aging probably due to the formation of ethyl-bridged compounds and their further reactions [
68,
70].
During the 12 months of aging, all wines exhibited a decrease in monomeric anthocyanins (
Figure 4A), which was accompanied by a rise in polymeric pigments (
Figure 3B). These changes resulted in a shift in the wines’ color characteristics in terms of CI and hue (
Figure 4B,C). More specifically, all wines showed an increase in both color intensity and hue compared to the EAF stage.
Interestingly, the total amount of native anthocyanins turned out to be basically the same in all the samples after 12 months of aging (
Figure 4A).
Among the wines, AG20 displayed statistically significant changes in hue and in the sum of polymeric pigments with respect to both the control and AG10. This might be explained by taking into account the higher degree of anthocyanin degradation in AG20 with the consequent formation of orange pigments (
Figure 4C) [
68]. This hypothesis is supported by the fact that in AG20, after 12 months, the sum of polymeric red pigments is (
Figure 3B) lower for AG20 compared with the control and AG10.
Data on the vanillin-reactive flavans (VRFs), hydroxycinnamic acids (HCAs), acetaldehyde, and absorbance at 420 nm (Abs420 nm) of white wine at the EAF and after 12 months of aging are shown in
Figure 5.
Regarding HCAs, a loss in all samples after 12 months of aging was detected with respect to the EAF. Acetaldehyde, which is the primary wine oxidation product, increased in all samples regardless of the added copper (II) amount. Such a strong increase can be related to the low amount of free sulfur dioxide in wines (around 4 mg/L in all the samples) that cannot limit the production of acetaldehyde and other oxidation products by the Fenton reaction [
71]. This parameter could be associated with the increase in the absorbance at 420 nm (
Figure 5D) that measures the amount of brown oxidated pigments in musts and wines. It is likely that in the absence of SO
2, the formation of glyoxylic acid from tartaric acid and the consequent formation of xanthilium salts [
72] were also favored. No effect of copper (II) on the absorbance at 420 nm was detected over time. This is in agreement with the results described by George et al. [
55] that showed that the 440 nm absorbance intensity of the copper (II) (II)-containing samples was in the same order as the samples without copper (II) (II), thus highlighting that iron, more than copper (II), is efficient in the production of xanthylium cation pigments in wine-like solutions containing tartaric acid and (+)-catechin. VRFs as well as HCAs decreased during aging (
Figure 5A,B). These results are in agreement with a previous study in which, during bottle storage, a decrease in the concentration of catechin and
trans-caftaric acid was observed [
73]. The authors also found that the browning index was more pronounced as a direct function of catechin and trans-caftaric acid degradation. The observed decrease in flavanols can be attributed to their reactivity [
74]. No significative differences were detected in all of the samples after 12 months of aging with respect to the control, thus confirming the obtained data on the browning index (
Figure 5C).