Thiols in red wine (Part 4): The matrix

by | Sep 1, 2020 | Oenology research, Winetech Technical

It is all about the (wine) matrix. The work presented in this article forms part of a series of experiments on the interaction between thiols and red wine matrices as described in Part 1.



The key word for this latest aspect of thiols in red wine which we are going to talk about is surely ‘matrix’. By definition, in biology: “matrix is the fine material used to bind together the coarser particles of a composite substance”; in terms of this project: the essence of a wine. Of course the great majority of it is composed of water (about 86%) and ethanol (about 12%) but, despite the great progress made in wine chemistry, there is still a very large number of unexplored and unidentified compounds in the matrix that appear to play an important sensory role.


Problem statement

The identification and quantification of this large amount of unrevealed molecules still elude us and might be some time before we manage to do so. As a consequence, we have set to shed some light on the matrix by rather looking at its impact on the sensory properties of wine when thiols come into play.



The experiment evaluated the thiol interactions 4MMP x 3MH and 4MMP x 3MHA at high concentrations (based on reported literature values) in de-aromatised wine from five different matrices. Wines were selected and consisted of 13 samples in total (three different Cabernet Sauvignon (plus a blind duplicate), three Pinotage, three Merlot and three Shiraz), all commercially available (Figure 1). To create the de-aromatised base, the wines were evaporated under vacuum to avoid oxidation. Spiking with thiols was carried out 24 hours before organoleptic evaluation. The wines were evaluated blind by 15 experienced judges using Projective Mapping (PM) as sensory method (for more details see Part 1).


FIGURE 1. Layout of Experiment 3.



The results can be analysed from two separate points of view, the first one by looking at the attributes generated by the panel, the second one looking at the spatial configuration of the samples on the 2D plane.

Considering the top 10 most used descriptors generated, eight were common to both combinations of thiols for all cultivars and ‘berry jam’ clearly took the crown. Some unique characteristics can also be observed: while the 4MMP x 3MH mixture is described as ‘fruity’ and ‘cooked vegetables’, 4MMP x 3MHA displayed ‘caramel’ and ‘blackcurrant’ as typical features. An interesting observation comes from focusing on ‘blackcurrant’. It was expected for this attribute to play a central role: it is often associated with 4MMP and here it could be clearly seen for both sets. The intriguing aspect is how the perception changes: in the 4MMP x 3MH combination, the ‘blackcurrant’ word is cited 32 times versus the 52 for the 4MMP x 3MHA set. An assumption made often in sensory science when looking at the frequency of cited attributes, is that the more a term is used, the more intense it must be. In this case, it would lead us to the conclusion that 3MHA increases the perception of ‘blackcurrant’ much more compared to 3MH (as already seen in Part 3 for only one matrix).

When observing the MFA maps (Figure 2a and 2b), the configurations of the samples have no obvious pattern related to cultivars. For example, Pinotage samples group together regardless of the thiol combination. Cabernet Sauvignon samples clustered together for 3MH x 4MMP, but not for 3MHA x 4MMP. Two of the Merlot samples are closed to each other, but not in the same combination across the two maps. Shiraz samples are scattered in both cases.


FIGURE 2. Multifactorial analysis (MFA) maps for the two interactions studied. 1, 2 and 3 – three different wine bases, a and b – two CS samples independently processed, Dup – blind duplicate in the sensory test.


Take home message

The journey that took us into the fascinating world of thiols in red wine started few magazine issues ago and it is now coming to an end. So what have we learnt so far?

In Part 1 we talked about the bigger picture, the complexity of defining thiols-associated aromas, the interactions between thiols and about the approach followed by our group of researchers in expanding the knowledge on the subject.

In the following issue (Part 2 – Experiment 1) we discovered that when thiols are present in red wine, the aroma attributes that come to mind are very different from the ones for white wine. Instead of ‘guava’, ‘grapefruit’ or ‘granadilla’, we get ‘caramel’, ‘savoury’ and other spicy notes, among others. In addition, the unmistakable role of the matrix became evident. With Experiment 1, we concluded that it is difficult to ascribe specific attributes to one of the three thiols tested. Exception is 4MMP for which the association with ‘blackcurrant’ has been confirmed as for the case of its presence in white wines. On the contrary, 3MH and 3MHA were associated with the presence of ‘savoury/soy’ aromas.

Part 3 (Experiment 2) elucidated the impact of thiols in combination with each other in one base wine. It demonstrated that, when looking at single compounds, the perception of the attributes is non-linear, especially when a second thiol is introduced. Synergistic and masking effects are noted and those results confirmed the important role the matrix plays in the perception of wine aromas. It also laid the foundation for investigating pairwise interaction of thiols in five different cultivars.

Finally, in this issue (Experiment 3) we highlighted the importance of studies including various matrices. Experiments carried out in model wine are necessary for fundamental research but, to better understand the complex role of wine as a system, it is essential to incorporate different scenarios.



We discuss the results of Experiment 3 in which we assessed the importance of thiols interaction (4MMP x 3MH and 4MMP x 3MHA) in different wine matrices. From the study, it is clear that results cannot be extrapolated from one cultivar to another and the important role the matrix plays in the perception of wine aroma is confirmed.


– For more information, contact Valeria Panzeri at or Astrid Buica at


Article Archives

Search for more articles

Generic selectors
Exact matches only
Search in title
Search in content
Post Type Selectors

Stay current with our monthly editions

Share This
Shopping cart
There are no products in the cart!
Continue shopping