Project 3.1.1

Identification and origin of volatile compounds responsible for important sensory attributes

Project summary

The flavour of wine is determined to a large extent by volatile compounds that are perceived during consumption by the sense of smell. The overall flavour of a wine, as well as specific flavour notes, is provided by dozens of chemical compounds of widely varying potency and concentration. The key volatile chemical compounds causing some flavour properties of wines such as ‘stone fruit’ are not known, and this project aims to better understand ‘tropical fruit’ flavour in Chardonnay, ‘green’ flavour in red wines, and the viticultural and winemaking influences on the concentration of the compounds, including the ‘black pepper’ compound rotundone. The sensory properties and consumer preferences for wines with different levels of these flavours is also an important aspect of this project.

Latest information

‘Stone fruit’ and ‘tropical fruit’ flavour in white wines
A sensory-compositional study of Viognier and Chardonnay wines has highlighted two lactones that are associated with ‘apricot’ or ‘peach’ flavour, and especially the ‘apricot’ character of Viognier. For ‘peach’ flavour, a number of yeast-derived acetate esters were indicated as important. A further lactone compound which has previously been little studied in wine has been indicated as a potent white wine flavour compound.

A consumer preference test with 90 consumers evaluating six wines (four Viognier, two Chardonnay) showed that ‘peach’ flavour was not an important factor driving liking in this study, either positively or negatively. The well-liked wines for most consumers were less viscous, had stronger ‘citrus’ and ‘tropical fruit’ flavour and had higher acidity, with the preferences of a smaller group of consumers being related to ‘overall fruit flavour’ intensity and aftertaste rather than specific flavour types. In this study, wines with stronger ‘apricot’ flavour were not well liked by two-thirds of the consumers. However, these wines also had relatively high ratings for ‘honey’ and ‘viscosity’ which may have been the main negative drivers rather than ‘apricot’.

A study of wines made from a set of Chardonnay juices sourced from across Australia, as well as a survey of commercially produced Chardonnay wines, found an unexpectedly high concentration of the potent thiols that are known to contribute to ‘tropical fruit’ flavour, especially in Sauvignon Blanc. In the survey of commercial wines, all had a concentration of the thiol 3-mercaptohexanol (3-MH) above the reported sensory detection threshold of 60 ng/L, with some up to 2,600 ng/L, a level comparable to that found in some highly aromatic Sauvignon Blanc wines. The juice precursors to these compounds were similarly high in many samples. A sensory study of unwooded Chardonnay wines made under controlled winemaking conditions from juices sourced from vineyards all across Australia showed that juices from the Adelaide Hills, Margaret River and Mount Barker in Western Australia gave wines with relatively high ‘tropical fruit’/‘passionfruit’ flavour, with a juice from the South Australian Riverland also producing a wine with strong ‘tropical fruit’ flavour. The ‘passionfruit’ attribute was most strongly related to the compounds 3-MH and 3-MHA (3-mercaptohexyl acetate), while ‘box hedge’ aroma was also linked to 4-MMP (4-mercapto-4-methyl pentan-2-one). The aroma attribute ‘flint’ was related to the concentration of another sulfur compound, benzyl mercaptan, and this is one of the first reports directly linking the ‘struck flint’ character to this compound. Results to date indicate that substantial flavour in Chardonnay can be contributed by thiols. Consumer acceptance data was obtained from 156 consumers, assessing six Chardonnay wines, selected on the basis of major differences in ‘tropical fruit’ aroma/flavour. The liking of the consumers was strongly related to the intensity of ‘passionfruit’ aroma and ‘fruit flavour’. Two clusters of consumers (60%) gave higher liking scores to the wines with higher ‘passionfruit’ aroma, with one cluster (22%) not appreciating the wine with stronger ‘box hedge’ aroma. Only a small minority of consumers (18%) preferred wines with low ‘tropical fruit’ attributes.

‘Pepper’ aroma in Shiraz

Several studies in model systems have provided good insight into the key mechanisms for the formation of the pepper compound rotundone in the grape berry. Results have shown a clear effect of light and air on formation of rotundone from a precursor, with a faster formation but lower final concentration of rotundone in the presence of light. With full sunlight, rotundone formed quickly but then started to decrease after three days. With 50% shade, rotundone formed more slowly for six days, plateaued for seven days, and then decreased slowly over three weeks. With no light, the rotundone formed slowly and then remained stable. This supports observations that the more shaded bunches or berries in a Shiraz vineyard have higher levels of rotundone, most likely due to reduced degradation in the lower light conditions.

Analysis of a subset of Grampians Shiraz grape samples from the 2015 season has been completed as a continuation of investigations into within-vineyard variation of rotundone using precision sampling. The concentration range for rotundone was found to be 9 to 101 ng/kg, which compares to 95 to 885 ng/kg in the same vines in 2012 and 3 to 19 ng/kg in 2013, again demonstrating the large vintage variation in rotundone concentration. The results confirm that the specific vines targeted across the seasons in low, medium and high rotundone zones of a single vineyard gave similar relative amounts of rotundone.

‘Green’ flavour in red wines
The contribution of specific aroma compounds to ‘green’ flavour in Shiraz and Cabernet Sauvignon is an important practical issue for wine producers. GC-olfactometry investigations have shown that the major ‘green’ compounds in Shiraz and Cabernet Sauvignon are those that have been previously identified, with isobutylmethoxypyrazine (IBMP) being of greatest importance. Further chemical analyses have revealed that dimethylsulfide can also be an important contributor to the ‘green’ aroma. C6 alcohols and aldehydes, on the other hand, appear to contribute less to ‘green’ flavour than had been previously believed. In Shiraz wines with ‘green’ sensory characteristics, IBMP was found at moderate concentrations, and was absent in the wines that did not exhibit ‘green’ aroma. This was a surprising result because Shiraz is believed to be a variety that cannot biosynthesise methoxypyrazines. It is possible, however, that the stalks may contain this compound. In a related study, fermentation of Adelaide Hills Shiraz grapes with stalks led to elevated concentrations of IBMP in wine, well above the sensory threshold. It may be that commercial whole bunch fermentation with stalks, or pre-fermentation cold soak, can cause extraction of this compound and may result in ‘green’ characters in Shiraz. C6 compounds were also found in much greater quantities in a treatment where grape leaves were added to the must compared to a minimal skin contact treatment.

To gain a better understanding of ‘green’ flavour a set of Cabernet Sauvignon wines was subjected to descriptive sensory profiling and volatile and non-volatile compositional analysis. The wines were found to have a range of ‘vegetal’ flavour scores, with variation in ‘eucalypt’, ‘mint’, ‘herbal’ and ‘cooked tomato’ characters, as well as in ‘astringency’ and ‘acid taste’. To gain a winemaker perspective on ‘green’ flavours a projective mapping (‘napping’) study was completed using a panel of 20 experienced winemakers. The winemakers grouped the wines on the basis of ‘green’ aroma characters as well as tannin/mouth-feel attributes.

A consumer liking study on six of the wines, selected to encompass the range of ‘green’ attributes, was completed with 113 consumers. The largest group of consumers preferred wines with lower astringency and bitterness, with the ‘green stalks’ or ‘vegetal’ attributes of the wines not so important in driving preference. ‘Red fruit’ and ‘fruit aftertaste’ were positive drivers for liking for these consumers. A minority of consumers very much appreciated wines with ‘eucalypt’/‘mint’ flavour (24% of the sample tested). The results indicate that controlling what might be considered ‘green tannin’ may be more important than minimising ‘green capsicum’ or ‘grassy’ attributes. The positive liking scores for an Adelaide Hills wine which was low in astringent/bitter attributes and dominated by ‘red fruit’ flavour mirrors results obtained in other red wine consumer studies by the AWRI, including a major study conducted in China in 2008, showing the attraction to many consumers of ‘red berry’ flavour in a wine with a soft mouth-feel.

Studies were conducted during the 2015 vintage on volatiles arising from vegetation growing close to vineyards. Pine or cypress trees were found to deposit needles into the grapevine canopy and even within bunches, which meant some were incorporated into the must. Clear evidence was found of needles entering the fermenter in the winery, and winemakers indicated a ‘pine’ flavour could be perceived in the finished wine. Known pine volatiles such as alpha- and beta-pinene were found in volatile traps placed close to the trees in the vineyards. Whether the direct aerial transfer of volatiles is an important mechanism is not yet known; to address this question final analysis of samples, including tree leaf matter, grape leaves, stalks and berries, is underway.

Project Team

Markus Herderich
Leigh Francis
Alice Barker
Sheridan Barter
Dimitra Capone
Wes Pearson
Tracey Siebert
Patricia Williamson