Premature ageing of wine aromas Pr Denis Dubourdieu, Dr Alexandre Pons and Dr Valérie Lavigne 1
2
# CITATIONS Premox or not Premox When Premox phenomenon shake the web That is the question 4000 3500 3000 2500 2000 1500 1000 500 0 2005 2006 2007 2008 2009 2010 YEARS Estimation Sources : google, private blog and web site (R.Parker, J.Robinson) 3
loss of fruity aromas Manifestations of premox in dry white wines appearance of heavier aromatic nuances reminiscent of honey, beeswax, moth balls and pine resin increased of orangey-yellow color with bitterness on the aftertaste 4
Gas chromatographic analysis combined with olfactometer detection 5
Aromagrams of honey extracts and of white wine containing honey descriptors Sensory Retention time Organic extract Organic extract (min.) fromhoney from wine ZO1 ZO2 10.34 strawberry 13.25 overripe pineapple 14.5 hydrocarbons 15.53 herbaceous herbaceous, forest floor 18.18 burnt meat 19.15 tobacco, cigar 19.53 cep 22.22 mushroom 22.36 vinegar grapefruit, lemon 22.55 26.37 orange 32.43 bitter almond bitter almond 32.53 candied fruit 38.17 peaches in syrup 44.02 apricot 47.44 rose rose 50.17 honey honey 53.5 resin resin 54.22 mothballs mothballs 60.2 beeswax beeswax 62.3 vanilla vanilla 6
Aromatic markers for defective ageing of dry white wines S O Methional Boiled potatoes 0.5 µg/l Methionine (Ferreira et al., 2002) Carbonyl compounds O2 O Phenylacetaldehyde Wilted roses, honey 25 µg/l Phenylalanine (Ferreira et al., 2002) O CH 3 o-aminoacetophenone Moth balls, wax polish 0.7 µg/l Oxidation Indolacetic acid NH 2 7
Contribution of sotolon to the oxidised aroma of wine H 3 C O H 3-hydroxy-4.5-dimethyl-2(5H)-furanone C H 3 * O O (Curry, walnuts) vins jaunes from the Jura and Sherry (Guichard et al., 1993) fig and rancio aromas of vins doux naturels (Cutzach, 1999) walnut aromas of old Port (Ferreira, 2003) aromas of defective ageing in dry white wines (Lavigne, 2002) Odor detection threshold of the racemic mixture 7 µg/l 8
Sotolon: a chiral molecule HO CH 3 H 3 C OH O O H CH 3 H 3 C H O O R S Each of these enantiomer has different olfactories characteristics. 9
Odor detection threshold of each enantiomer of sotolon Odor detection thresholds (µg/l) Descriptors Model solution R 89 Walnuts, rancio S 0.8 (Curry, walnuts) Racemate 2 (Curry, walnuts) The S forme is far more odoriferous than the R forme. The (S)-sotolon is the one that contributes to the odor of 10 white wine affected by premox.
% Distribution of enantiomers of sotolon in dry white wines in bottle 50/50 30/70 70/30 100 80 60 40 20 0 Racemic mixture Excess (S)-sotolon Excess (R)-sotolon Sotolon s contribution to the premox characteristics of a wine depends on its level of S forme 11
How sotolon is formed in dry white wines. C O OH O aldocondensation HO O H O O α-ketobutyric acid Ethanal Sotolon 12
Glutathion (Glu Cys Gly) COO - H + 3N CH Glycine Acide glutamique CH 2 COO - CH 2 CO NH CH 2 CO NH CH Cystéine CH 2 SH 13
The relation between the levels of available nitrogen and glutathion in white wine must Must 1 Must 2 Must 3 Must 4 Must 5 Must 6 Must 7 Must 8 Available nitrogen (mg/l) 62 244 76 202 224 56 187 42 Glutathion (mg/l) 12 28 17 28 25 6 22 4 14
Reaction of adding a thiol (R-SH) to the catechin quinone H O O OH OH Oxidation H O O O O O H O Reduced catechin O H O Oxidised catechin (quinone) O H O H R-SH Thiol (aroma, glutathion,etc.) H O O O H S OH R adduct 15
Examples of glutathion content in various Sauvignon Blanc and Sémillon musts Must 1 Must 2 Must 3 Must 4 Must 5 Must 6 Must 7 Must 8 Glutathion (mg/l) 12 28 17 28 25 6 22 4 16
Changes in glutathion content in must during alcoholic fermentation concentration of glutathion (mg/l) 12 10 8 6 4 2 0 must T1 T3 T6 T9 sulphiting T20 T30 17
Relation between levels of glutathion in must and in young wine Glutathion in must (mg/l) 9 5 4 17 2 Glutathion in the corresponding wine (mg/l) 11 7 6 22 3 18
Relationship between the initial glutathion content of a model medium and the level at the end of alcoholic fermentation. 30 25 20 15 10 5 0 Level of glutathion in the medium before AF Level of glutathion in the medium after AF 19
Relationship between the level of available nitrogen and the level of glutathion at the end of alcoholic fermentation. 25 20 15 10 T0 End of AF 5 0 48 96 190 Level of available nitrogen in the model medium (mg/l) 20
Traditional ageing on the lees Retains fruity aromas of the young wine Maintains wine in an oxidation-reduction state that encourages a reductive «bouquet» : truffle, burnt and mineral. Avoids or delays premox 21
Changes in the level of 3-MH in a Sauvignon Blanc wine aged different ways in barrel Concentration of 3-M (ng/l) 1600 1400 1200 1000 800 600 400 End of AF November April 200 0 Used Barrel Used barrel racked Used barrel New barrel racked 22
Amount of sotolon in the wines at the end of barrel ageing Concentration of sotolon (µg/l) 9 8 7 6 5 4 3 2 1 0 Barrel used on the lees Used barrel without lees Barrel new on the lees New barrel without lees Detection threshold (white wine) 23
Effect of barrel ageing techniques on changes in the glutathion level of wines concentration of glutathion (mg/l) 18 16 14 12 10 8 6 4 2 0 End of AF December January February May Samples new barrel on the lees new barrel without lees used barrel on the lees used barrel without the lees 24
Glutathion, sotolon and 3-mercapto-hexanol at the end of barrel ageing Glutathion, a natural component of grapes makes it possible to prevent the defective ageing of white wines. Used barrel on the lees Glutathion Sotolon 3-MH (mg/l) (µg/l) (ng/l) 5.8 1.3 1400 New barrel without the lees 0.5 9.7 420 Therefore, the ageing conditions most favourable to retaining the aromatic qualities of dry white wines are once that limit the decrease in the concentration of glutathion. 25
Oxygen consumption (µg/l/h) of a white wine aged for 6 months entirely its lees Wine aged on its lees 611 Filtered wine 0.01 Lees alone 542 Heat-treated lees 19 (Fornairon et al., 1999) 26
Identification of the random nature of premature ageing in two dry white wines (tasting in 2005 of 12 bottles of each wine) % % 70 60 50 40 30 20 10 0 70 60 50 40 30 20 10 0 Little or no signs of age Little or no signs of age Showing average signs of age Showing average signs of age Looking very aged Looking very aged Graves 2001 1997 Pessac Léognan 27
Correlation between the oxygen dissolved in bottled wines and prematurely-aged aromas Analyses of 20 samples of a Graves white wine (1997 vintage) after 7 years in bottle. Average 6.00 5.00 4.00 3.00 2.00 1.00 0.00 R20.7084 0 20 40 60 80 100 120 140 Dissolved oxygen (µg/l) 28
DO420 nm Effect of dissolved oxygen content on colour 0.4 R 2 = 0.8315 0.3 0.2 0.1 0 0 10 20 30 40 50 60 70 80 90 Level of dissolved oxygen (µg/l) 29
Free SO 2 (mg/l) Correlation between free SO 2 and dissolved oxygen 25 20 15 10 R 2 = 0.7909 5 0 0 20 40 60 80 100 120 140 Level of dissolved O2 (µg/l) 30
Choice of closure Concentration of sotolon (µg/l) 3.5 3 2.5 2 1.5 1 0.5 0 R 2 = 0.7191 0 20 40 60 80 100 120 140 Dissolved oxygen in bottle (µg/l) 31
Level of oxygen (µg/l) Dissolved oxygen measured in white wine six months after bottling Effect of closure 120 100 80 60 40 20 0 1 2 3 4 5 Types of closure used 32
Free SO 2 (mg/l) Changes in the level of free SO 2 in bottle Effect of closure 30 25 20 15 3 1 2 4 5 Types of closure used 33
Preventing the defective ageing of white wines, requires maintening their level of glutathion: Doing so, calls for 8 relatively simple rules: 1. having vines with sufficient vigour, thanks to a nitrogen intake in keeping with their needs 2. limiting the extraction of phenolic compounds during pressing 3. protecting both the must and the wine from oxidation (inert gas, SO 2 ) 4. making sure that alcoholic fermentation take place rapidly and is completely finished 5. reducing the time lag before malolactic fermentation (stirring lees, inocculation) 6. ageing the wines in reductive conditions (SO 2, lees, moderate new oak) 7. limiting the dissolution of oxygen when preparing the wine for bottling 34 8. choosing a closure that is suited to the wine
Manifestations of defective ageing of red wines In must and wine from overripe grapes In wine (ageing in barrel, ageing in bottle) Olfactory descriptors Prune, cooked fruit and fig 35
Examples of aromagrams for red wines affected or unaffected by defective ageing Time from retention IRL Extract from red wine control Extract from red wine with an aroma of ageing defectiv e 21.42 1419 vinegar vinegar 21.80 1430 vegetables, soup sou p 24.04 1493 paper 27.52 1592 fatty acid vegetable bouillon 29.00 1634 dirty feet dirty feet 30.25 1668 floral 30.89 1688 slightly smoky rancid 32.63 1736 rancid vegetables 32.88 1742 prune 33.07 1747 vegetables rancid 34.06 1776 sweat floral 34.55 1790 rancid 36.42 1841 fruit compote apple cooked fruitfr uit compote 40.26 1951 rose rose 43.07 2031 meaty rancid 43.44 2041 cooked fruit ZO1 ZO2 Three distinct sections are reminiscent of prune or cooked fruit 36
Aromagrams of fruit extracts reminiscent of prematurely aged red wines Time retention IRL Fig extract Prune extract 21.42 1419 vinegar 21.80 1430 soup 24.09 1492 dust 24.65 1509 geranium 28.44 1617 fig 29.06 1634 dirty feet 29.35 1644 wilted rose 31.73 1711 almonds, flowers 30.89 1688 rancid 32.84 1740 prune 33.09 1748 rancid 36.46 1841 cooked fruit cooked fruit 39.44 1929 floral figs, dried fruit 40.14 1950 almond 43.00 2031 burnt sugar 43.44 2041 smoky cooked fruit ZO1 ZO2 Relevance of the "prune" descriptor to describe the odour of wines 37
Example of aromagrams of wines made from raisined grapes Time retention IK Extract from red wine control Extract from red wine with an aroma of ageing defective Extract of wine from grapes " raisined" 21.42 1419 vinegar vinegar vinegar 21.80 1430 vegetables, soup soup soup 24.04 1493 paper bread without crust 27.52 1592 fatty acid vegetable bouillon 29.00 1634 dirty feet dirty feet dirty feet 30.25 1668 floral 30.89 1688 slightly smoky rancid rancid 32.63 1736 rancid vegetables 32.88 1742 prune prune 33.07 1747 vegetables rancid 34.06 1776 sweat floral floral, chemical 34.55 1790 rancid 36.42 1841 cooked fruit, fruit compote cooked fruit cooked fruit, from apple compote of apple compote apple 40.26 1951 rose rose 43.07 2031 meaty rancid 43.44 2041 cooked fruit cooked fruit ZO1 ZO2 The degree of ripeness of the grapes determines the ageing of wines in bottle 38
Incidence of oxygen on aroma sectors Time from retention IRL Extract from red wine control Extract from aged red wine 15 days in the presence of oxygen 21.42 1419 vinegar vinegar 21.80 1430 vegetables, soup sou p 27.52 1592 fatty acid vegetable bouillon 29.00 1634 dirty feet dirty feet 30.89 1688 slightly smoky rancid 32.63 1736 rancid vegetables 32.88 1742 prune 33.07 1747 vegetables rancid 34.06 1776 sweat floral 36.42 1841 fruit compote apple cooked fruit compote apple 40.26 1951 rose rose 43.07 2031 meaty rancid 43.44 2041 cooked fruit ZO1 ZO2 ZO1 and ZO2 are formed by oxidative mechanisms 39
Identification of the comopund associated with ZO1 O O Concentration (µg/l) Descriptors 0.1 mentholated 3-methyl-2,4-nonanedione Aniseed, dry grass 1 aniseed, fruit pits, prune 10 aniseed 3-methyl-2,4-nonanedione (MND) identified for the first time in red wines Naturally present in soy oil (Guth, 1989) and green tea (Guth, 1993) Comes from the oxidation of certain furanic fatty acids (Guth, 1991) Detection threshold in a hydroalcoholic solution 16 ng/l 40
Perception Intensity Average Effect of increasing concentration of MND in red wine on the detection intensity of some typical descriptors 4,5 4 3,5 3 2,5 2 1,5 1 Fresh Fruit Anise Prune Kernel Rancio 0,5 0 0 50 90 170 250 330 MND [ng/l] The premature aging of red wines Pons A., Lavigne V., Dubourdieu D. 41
MND distribution in wines Non-Premox Premox White The premature aging of red wines Pons A., Lavigne V., Dubourdieu D. Rosé Red White Rosé Red 42
Concentration of 3-methyl-2,4-nonanedione in wines from several vintages of the same estate in Bordeaux Vintages (ng/l) Aroma index 1982 155.1 9 1983 260.2 16 1985 140.2 9 1986 380.1 22 1991 109.2 7 1996 133.4 5 2000 145.4 9 [MND](ng/L) 400 350 300 R 2 = 0.820 250 200 150 100 50 0 0 1 2 3 4 5 Average Levels vary according to vintages The intensity of the prune character of wines is correlated with their concentration of MND 3-methyl-2,4-nonanedione is a marker of the appearance of this aroma in red wines 43
MND ng/l Ageing potential of red Bordeaux wines and 3- methyl-2,4-nonanedione 250 200 150 100 50 Premier Vin Second Vin 0 2006 2005 2004 2003 2002 Millésimes There is little difference in the first few years after the wine is bottled The first wine seems less likely to develop prune nuances during ageing in bottle Greater ageing potential, increased resistance to oxidative phenomena Analytical validation of empirical attribution during tasting of the first and 44 second wine from the same estate