The iron(III) tartrate photochemistry of wine: impacts of bottle colour and weight


THIS ARTICLE SUMMARISES the latest findings on the iron(III) tartrate photochemistry of relevance to wines. This specific photochemical process can cause tartaric acid to fragment and induce colour changes in wine. It demonstrates the wavelengths of light relevant to such photochemistry, as well as the compositional impacts of the wine matrix. Finally, under the conditions adopted, more protection was offered by darker bottles of heavier weight.

Effects of exposure to light

Most wine will have little exposure to light while it is in tank or barrel. The majority of exposure to light will occur after bottling, and in retail outlets or in domestic situations where artificial fluorescent lighting generates energy at low visible and near UV wavelengths. For wine to undergo spoilage by light it needs a component that is able to absorb the energy of the light and consequently drive spoilage reactions. It is in this manner that photochemical agents can lead to the conversion of sulfur-containing amino acids to off-smelling compounds (Maujean and Seguin 1983), or the production of coloured compounds from wine phenolic compounds (Dias et al. 2010, 2012, Maury et al. 2010). The components of wine that have been implicated in this photochemical role are riboflavin (or vitamin B2) (Majean and Seguin 1983, Mattivi et al. 2000) and metal ions in combination with wine organic acids (Clark et al. 2007).
The focus in this article will be on the second photochemical system comprising of metal/organic acid complexes and follows on from our past research in this area (Clark et al. 2007, Maury et al. 2010). Such metal organic complexes have been actually utilised in photography in the past, including the use of iron(III) citrate in the production of blue prints such as that shown in Figure 1 on page 119. (i.e. see the reference list for a link on how to make your own blue prints). In blue prints, the iron(III) citrate absorbs light and converts the iron(III) to iron(II), and the iron(II) is then able to form a blue colour, or prussian blue to be exact, with another component in the photographic mix (that is, potassium ferricyanide - not to be confused with the related potassium ferrocyanide salt used in blue fining). Other photography emulsions have actually utilised iron(III) tartrate systems (Ware 1999).
Metal ions and organic acids are abundant in all wines. The metal ion most implicated in wine photochemical processes is iron, which can exist in two different redox forms of iron(II) and iron(III), whereby iron(III) contains one less electron than iron(II). It is generally accepted that it is the iron(III) form of the iron that is photoactive. The source of iron in wine is varied and typically ranges from 2-20mg/L, with an average around 5mg/L. Modern winemaking equipment (i.e. utilising stainless steel) means that iron contamination in wines is much lower than it often was in the past when less inert iron containing alloys came into contact with wine.
The main organic acids in white wines consist of tartaric acid, malic acid and to a lesser extent lactic acid, succinic acid and citric acid. The organic acid is important in the photochemical process as the iron cannot undergo efficient light induced redox transitions (that is, from iron(III) to iron(II)) in the absence of the organic acid. Given that all wines will contain at least some iron and organic acids regardless of their variety, source and/or age, all have the potential to undergo photochemical spoilage processes.
Of interest in our current research was the identification of the components of the wine matrix that were most critical to the photochemical process, as well as identifying the wavelengths of light that allowed the most efficient photoactivity. Finally, the impact of wine bottle colour and bottle weight on the photochemical process was to be ascertained. Studies were initially conducted in a model wine system to simplify the number of wine components able to interact with the light. The model system consisted of 12%(v/v) ethanol at pH3.2 with tartaric acid (~3-4g/L) and also with 5mg/L of iron present. The light source used in this study was a xenon-arc lamp (that simulates the solar light spectrum) and irradiation was performed for 30 minutes and temperature held at 45°C.
Upon exposure of the model wine system to the light from the 150W xenon-arc lamp (at a distance of 140mm), a small amount of the tartaric acid was converted to glyoxylic acid, which essentially involves the tartaric acid molecule breaking in half. No glyoxylic acid production occurred without added iron and/or without exposure of light to the model wine system. Therefore, it was evident that iron(III) tartrate is indeed photoactive in wine conditions. Given that both the original iron(III) tartrate and glyoxylic acid are largely colourless, no change in colour of the model wine was evident after initial light exposure (Figure 2).
Despite the chemical name, the 'glyoxylic acid' actually contains an aldehyde group, as well as an acid group, and consequently can bind sulfur dioxide, the main preservative of wine. Its action in binding sulfur dioxide is of similar efficiency as acetaldehyde, whereby it will be essentially completely bound in the presence of some free sulfur dioxide. This means that the light exposure of the wine has the potential to lower the main preservative and thereby limit the shelf-life of the wine.
In the absence of sulfur dioxide, glyoxylic acid can react with grape skin-derived phenolic compounds (i.e., catechins) and produce yellow coloured compounds in the wine (Es-Safi et al. 2000). These pigments have been identified as xanthylium cation pigments and are responsible for the yellow colour in the lower sample shown in Figure 2. When formed at higher concentrations or in the presence of pulp-derived wine phenolic compounds, they adopt a more brown appearance rather than yellow. This reaction will be more pronounced in white wines produced from heavier pressed grapes than in wine produced from light pressings because the skin-derived phenolic compounds are extracted to a higher extent in heavier pressings.
Although the yield of glyoxylic acid from a solution of iron(III) tartrate is low, it does not take the production of much glyoxylic acid to cause significant binding of sulfur dioxide or produce significant quantities of the coloured compounds. For example, a wine with 5g/L tartaric acid would only require 1.5% conversion of the total tartaric acid to glyoxylic acid in order to fully deplete 30mg/L free sulfur dioxide. For most wines, this would be the near complete removal of the wine's preservative. These results demonstate that iron(III) tartrate photochemistry has the potential to impact on white wine quality, either through the consumption of the main preservative, and/or through the production of undesirable brown colouration.
An experiment was conducted to ascertain the important wine components that could impact on this iron(III) tartrate photochemical process. As a result of the light exposure, 0.43mM of glyoxylic acid was produced in the control sample, which equates to 2% conversion of total tartaric acid to glyoxylic acid and would have the potential to bind 28mg/L free sulfur dioxide. Interestingly, lowering the dissolved oxygen concentration in the sample, by purging with argon, also lowered the yield of glyoxylic acid to 0.18mM, inferring that oxygen is important in the photochemical process for the cleavage of tartaric acid.
The impact of oxygen on the light exposure effect can be further illustrated by Figure 3. Chardonnay wine was bottled in Flint bottles (i.e., three bottles with 'high oxygen' and three bottles with 'low oxygen') with different levels of total packaged oxygen (i.e, 'low oxygen' = ~4mg/L, and 'high oxygen' = ~ 7mg/L), and the head-space of the 'high oxygen' sample was aerated daily. Both wines were exposed to 16 hours of irradiation (160W at a distance of 400mm) and eight hours of darkness at identical temperatures (30°C) over 18 days. As is evident in the picture the wines with higher oxygen at bottling were much browner in colour than the equivalent wines without oxygen. In fact, the wines with low oxygen had little development in colour over the irradiation period and hence oxygen would appear to be a critical component for the photochemical colour development in white wine.
Further studies were conducted to assess which particular wavelengths of light were critical for the formation of glyoxylic acid from the iron(III) tartrate solution. The results showed that the wavelengths of light between 200-520nm could generate glyoxylic acid from iron(III) tartrate, and that the wavelengths between 300-520nm were more efficient. The latter wavelength range included near UV light, that is 300-380nm, and also some visible light, at 380-520nm which corresponds to violet/blue light.
Figure 4 highlights the photoactive region of the spectrum whilst also illustrating the transmission spectra for different coloured wine bottles. On this graph a transmittance of 0% means that the actual bottle is absorbing all the light and that none will reach the wine inside the bottle. A transmittance above 0%, means that the wine bottle will allow at least some of the incident light through to the wine. In effect, the bottles themselves act as filters of incident light and thereby modify the composition of light falling upon the wine.
From the result in Figure 4 we can see that the Flint and Arctic Blue bottles allow most light through to the wine at wavelengths above 300nm. The remaining bottles have differing intensities of light transmission but with regard to the critical wavelength region of 300-520nm, the lowest light transmission is by the Amber bottles, followed by Antique Green and then finally by the French Green bottles. Interestingly, all the bottles allow at least some light of wavelengths below 520nm into the wine. Therefore, complete protection of the wine from iron(III) tartrate photochemistry is not possible regardless of which colour bottle is utilised, however, the darker bottles shown above would appear to be the most effective at limiting the photochemistry.
To assess the impact of bottle weight on the transmission spectra, the Flint and Antique Green bottles were investigated. Both these bottle colours are readily available in both light and heavy weight variants. Although classified by weight, these bottles also differ by the thickness of the glass depending at which position the glass is measured on the bottle. However, for the mid-section of the bottle, the glass thickness is around 2mm and 3mm for the light and heavy weight bottles, respectively and regardless of colour. Figure 4 shows that there was negligible difference in the transmission spectra of the Flint bottle based on the weight of the bottle. Whilst for Antique Green there was a significant difference in the transmission spectra. In the critical wavelengths of 300-520nm, the heavy weight bottle had at least 30% less light transmitted than for the light weight bottle, which was consistent with the relative thickness of the bottles.
Using the Flint and Antique Green bottles, an iron(III) tartrate system was irradiated and the amount of glyoxylic acid produced was monitored. The results in Figure 5 show negligible difference between the glyoxylic acid generated with the Flint heavy and Flint light weight bottles. Alternately, the heavy weight Antique Green bottle produced around a third of the level of glyoxylic acid as in the light weight Antique Green bottle. Such a result is consistent with the transmission spectra in Figure 4 and the thickness of the light and heavy weight glass.
The comparative amount of glyoxylic acid generated in the Flint and Antique Green bottles is more difficult to reconcile based merely on the transmission spectra, which would suggest much higher levels of glyoxylic acid in the Flint bottles compared to the Antique Green bottles. However, it would appear that the production of glyoxylic acid by the iron(III) tartrate photochemistry may be governed by a limiting reagent when produced at high rates, such as in the Flint bottles. The limiting reagent may be molecular oxygen that impacts on glyoxylic acid photochemical formation but becomes limited in the Flint bottle samples due to high rates of tartaric acid degradation. Further work is required to enable prediction of the amount of glyoxylic acid produced for a given bottle colour, thickness and experimental conditions (i.e., dissolved oxygen concentration, light intensity). In any case, these results clearly demonstrate lowered glyoxylic acid production in the darker coloured bottle and also lowered glyoxylic acid produced in the thicker variant of the darker coloured bottle.
In conclusion, we have shown that iron(III) tartrate is a photoactive system in wine conditions. As a consequence, a product can be generated that can shorten the shelf-life of the wine, and also contribute to detrimental colour changes. The photochemical process for glyoxylic acid production from iron(III) tartrate is reliant on the presence of dissolved oxygen. It has also been shown that UV light as well as short wavelength visible light (300-520nm) can induce the photochemical process, and wine would be given more protection from the photochemical process from darker coloured bottles with a thicker wall of glass. Current research is investigating the impacts of light intensity, temperature during irradiation and sulfur dioxide depletion during irradiation.
Further details on these experiments and results can be found within Clark et al. (2011).


This work was supported by the Grape and Wine Research and Development Corporation (GWRDC, project UM0902). The National Wine and Grape Industry Centre is a partnership between Charles Sturt University, DPI NSW and the NSW Wine Industry Association.
Dr Andrew C. Clark, National Wine and Grape Industry Centre, Charles Sturt University, Locked Bag 588, Wagga Wagga, NSW, 2650. [email protected]
Dr. Daniel A. Dias, Metabolomics Australia, School of Botany, The University of Melbourne, Parkville, VIC, 3010, Australia. [email protected]


Blue print instructions: (12 September 2012)
Clark, A. C., Dias, D. A., Smith, T. A., Ghiggino, K. P. and Scollary, G. R. (2011) Iron(III) tartrate as a potential precursor of light-induced oxidative degradation of white wine: studies in a model wine system. Journal of Agricultural and Food Chemistry 59: 3575-3581.
Clark, A. C., Prenzler, P. D. and Scollary, G. R. (2007) Impact of the conditions of storage of tartaric acid solutions on the production and stability of glyoxylic acid. Food Chemistry 102: 905-916.
Dias, D. A., Smith, T. A., Ghiggino, K. P. and Scollary, G. R. (2010) Ultraviolet light - A contributing factor to pigment development in white wine. The Australian and New Zealand Wine Industry Journal 25(3): 52-61.
Dias, D. A., Smith, T. A., Ghiggino, K. P. and Scollary, G. R. (2012) The role of light, temperature and wine bottle colour on pigment enhancement in white wine. Food Chemistry 135: 2934-2941.
Es-Safi, N.-E., Le Guerneve, C., Fulcrand, H., Cheynier, V. and Moutounet, M. (2000) Xanthylium salts formation involved in wine colour changes. International Journal of Food Science and Technology 35:63-74.
Mattivi, F., Monetti, A., Vrhovsek, U., Tonon, D. and Andrès-Lacueva, C. (2000) High-performance liquid chromatographic determination of the riboflavin concentration in white wines for predicting their resistance to light. Journal of Chromatography A 888: 121-127.
Maury, C., Clark, A. C. and Scollary, G. R. (2010) Determination of the impact of bottle colour and phenolic concentrations on pigment development in white wines stored under external conditions. Analytica Chimica Acta 660: 81-86.
Maujean, A. and Seguin, N. (1983) Contribution a l'étude des 'gouts de lumière' dans les vins de Champagne. 3. Les reactions photochimiques responables des 'gouts de lumière' dans le vinde Champagne. (Sunlight flavours in the wines of Champagne. 3 - Photochemical reactions responsible for sunlight flavours in Champagne wine). Sci. Alim. 3: 589-601.
Ware, M. (1999) Cyanotype. National Museum Photography, Film and Television. West Yorkshire, UK. Pg. 27, 153.

Other feature articles

AB Mauri



WID 2017