The impact of wine components on fractionation of Cu and Fe in model wine systems: Macromolecules, phenolic and sulfur compounds

Highlights

• Assessment of a range of important wine components on techniques for the fractionation of Cu and Fe in wine.

• Use of SPE for fractionation of Cu and Fe in wine and model wines and their detection via ICP-OES.

• Labile Cu measurement in wine and model wine wines utilizing constant current medium exchange stripping potentiometry.

• Hydrogen sulfide: the only wine component that binds Cu strongly.

Abstract

A variety of techniques have been developed with the ability to measure different forms of metals in wine with the ultimate aim of providing a more accurate indicator of metal induced spoilage of wine. This study was conducted in order to identify which wine components influence the measurement of Cu and Fe in their fractionated and/or electrochemically active forms. The measurement techniques involved detection of labile Cu by stripping potentiometry and fractionation of Cu and Fe by sequential solid phase extraction, with detection by inductively coupled plasma-optical emission spectroscopy. The wine components assessed included those extracted from wine (red wine tannin, white wine protein, white wine polysaccharide, red wine polyphenol, white wine polyphenol), and commercially available monomeric compounds, including phenolic compounds and sulfur-containing compounds. For Cu, only hydrogen sulfide, which is known to induce the formation of Cu(I) sulfide, showed any appreciable influence on the fractionation and electrochemical detection of Cu. This form of Cu was also identified as the major component of red and white wines. For Fe, the fractionation was different for red versus white wine, and influenced significantly by extracted red wine polyphenol, (−)-epicatechin, gallic acid and tartaric acid. The wine components showed more influence on Fe at pH 4.00 compared to pH 3.25. These results enable a targeted use of these techniques in the assessment of metal-induced spoilage of wine.

Introduction

Copper (Cu) and iron (Fe) in wine are linked to both the reductive and oxidative development of wine (Clark et al., 2015b, Ribéreau-Gayon et al., 2006, Ugliano, 2013, Viviers et al., 2013). Their contribution to oxidation is primarily through the catalysis of reactions induced by molecular oxygen, ultimately contributing to the oxidation of wine components such as sulfur dioxide, phenolic compounds and ethanol (Clark et al., 2015b, Danilewicz and Wallbridge, 2010, Li et al., 2008). In some cases, the oxidation reactions can be desirable when they occur in a controlled manner, such as during the barrel aging of red wine, but often they are detrimental to the sensory features of the wine (Avizcuri et al., 2016, Culleré et al., 2007, Escudero et al., 2000). In terms of promoting reduction in wine, Cu and Fe are linked to the accumulation of certain sulfur compounds (hydrogen sulfide, methanethiol), when wine is aged in low oxygen conditions (Ugliano, 2013, Viviers et al., 2013). These compounds can be responsible for undesirable aromas and the repression of desirable characters of the wine.

Given the potential influence that Cu and Fe can exert on wine development, a variety of instrumental techniques have been developed to allow measurement of both the total concentrations and different forms of the metals in wine (Clark et al., 2015). The latter types of techniques either classify the metal in broad terms (i.e., fractionation) or as specific ligand metal chelates (i.e., speciation) (Clark et al., 2016, Danilewicz, 2014, Karadjova et al., 2002, Pohl and Prusisz, 2009, Pohl and Sergiel, 2009, Wiese and Schwedt, 1997). The fractionation and speciation measurements have been developed with the aim of measuring a form of the total metal concentrations that will be a better measure of metal activity in wine than the total concentration. However, in the scientific literature there has been more emphasis on the development of methods rather than studies linking the measurement to the role of Cu and Fe in oxidative and reductive development of wine. An example of the latter was the use of spectrometric measures of Fe(III) tartrate in model wine systems to follow the redox equilibrium reached between Fe(II) and Fe(III) in oxidizing model wine systems (Danilewicz, 2014). Further work established a measure of Fe(II) in white wine utilising a colorimetric ferrozine measure that accounted for bias towards Fe(II) during the measurement (Danilewicz, 2016). The results showed wine exposed to oxygen exhibited a decrease in the Fe(II):Fe(III) ratio and that the final ratio reached was dependent on the composition of wine. The fractionation of Cu and Fe by solid phase extraction and analysis by inductively coupled plasma with optical emission spectroscopy detection (SPE ICP-OES) has also been used to investigate the fractions of Cu and Fe more relevant to oxygen consumption rates in wine with ascorbic acid added (Rousseva, Kontoudakis, Schmidtke, Scollary, & Clark, 2016). In this study, it was found that the fractions of Cu that were able to elute through a C18 and cation exchange SPE cartridge were correlated with the highest oxygen consumption rates. The Cu that was trapped on a cation exchange column also showed reasonable correlation with oxygen consumption rates whilst the Cu retained on the C18 cartridge showed little relationship to oxygen consumption rates. The total Fe concentration and corresponding fractions had poor correlation with the oxygen consumption rates in the wine. Finally, the electrochemical measurement of labile and non-labile Cu, meaning Cu able to be detected and not detected by the technique, respectively, showed an ability to monitor the conversion of Cu(II) to Cu(I) sulfide in wine (Clark et al., 2016). This technique showed a significant drop in the labile Cu measured in a wine when hydrogen sulfide was added to a wine (in the form of sodium sulfide) and Cu(I) sulfide was generated. Conversely, in model wine with Cu(I) sulfide present, the loss of volatile hydrogen sulfide induced an increase in labile Cu that was attributed to disassociation of Cu(I) sulfide.

However, it is knowledge of the wine components that influence fractionation of the metal within wine that is most lacking from fractionation methods. For example, in the case of the electrochemical measurement of labile and non-labile Cu (Clark et al., 2016), it is known that Cu(II) tartrate is measured as labile and Cu(I) sulfide is measured as non-labile, but the classification of other Cu complexes is not known. Similarly, for the fractionation of Cu and Fe by SPE ICP-OES, little is known about the influence of major wine components on the fractionation (Pohl and Prusisz, 2009, Pohl and Sergiel, 2009, Rousseva et al., 2016). Although the residual and cationic forms of Cu were shown to correlate best with oxygen consumption in a group of Chardonnay wines (Rousseva et al., 2016), not much is known about the identity of these fractions of Cu other than they can constitute Cu(II) organic acid complexes (Pohl and Sergiel, 2009, Rousseva et al., 2016).

The components of wine that can interact with Cu(II), Fe(II) and Fe(III) are numerous and vary widely in general type. They can consist of the wine macromolecules, including proteins, polysaccharides, tannins and their aggregate combinations, as well as monomeric species, including organic acids, phenolic compounds, and sulfur-containing compounds (Danilewicz, 2016, Karadjova et al., 2002, Kreitman et al., 2016a, Kreitman et al., 2016b). Although many studies have been conducted on the complexation of metal ions to these components in non-wine-like conditions (higher pH, ethanol-free or without organic acids), the specific conditions of wine are likely to be critical in indicating the dominant binders of metals. For example, although phenolic compounds are known to be significant binders of Fe in aqueous conditions (typically at pH > 4), Danilewicz showed with spectroscopic evidence that little chelation occurred between Fe and monomeric phenolic compounds at pH 3.6 when 5 g/L tartaric acid was present (Danilewicz, 2014, Danilewicz, 2016). Consequently, although Fe is known to interact with phenolic compounds in wine conditions via the transfer of electrons, the interaction was not sufficient to displace the predominant binding between tartaric acid and Fe. Alternatively, Pohl and Prusisz (2009) showed that the presence of 500 mg/L of tannic acid with Fe(III) could induce different SPE ICP-OES fractionation of Fe compared to a control (Fe(III) without tannic acid present). Other studies, have shown interactions between wine macromolecules and the metal ions. Karadjova et al. (2002) utilised general precipitation procedures for polysaccharides and proteins in wine to show significant concentrations of Cu and Fe in both types of precipitate, inferring association of metals by the macromolecules. Furthermore, complexation capacity studies have been conducted in red wine, using electrochemical techniques, and have shown complexing ligand concentrations and stability constants of ligands (Vasconcelos, Azenha, & Freitas, 1999).

This study was conducted in order to identify which wine components influence the measurement of Cu and Fe in their fractionated/electrochemical forms and to enable an improved understanding for the use of such measurements in following metal activity in wine. The measurement techniques involved SPE ICP-OES fractionation for both Cu and Fe, and stripping potentiometry for Cu. The impact of pH and different metal ratios was also to be assessed for various wine components.