Fate of damascenone in wine: The role of SO2

Article abstract of DOI:10.1021/jf048582h

Damascenone has been shown to undergo reaction with common wine components. Following the action of acid and heat alone, two bicyclic compounds, 4,9,9-trimethyl-8-methylenebicyclo[3.3.1]non-6-en-2-one (2) and 4,4,9-trimethyl-8-methylenebicyclo[3.3.1] non-6-en-2-one (3), were isolated. However, this conversion takes place only very slowly, if at all, under milder conditions (45 °C). When treated with a variety of nucleophiles at pH 3.0 and 5.5, the concentration of damascenone in buffered aqueous ethanol decreased by minor amounts (10-20%) except for cysteine and 2-mercaptoethanol addition at pH 5.5 (～40 and ～30%, respectively) and SO₂ (>90% at pH 3.0; 100% at pH 5.5). An adduct from this last combination was prepared and shown to be the C₉ sulfonic acid derivative of damascenone. A detailed investigation into the effect of SO₂ demonstrated that loss of damascenone in model wine was directly related to the concentration of added SO₂ but was essentially unaffected by small changes in pH.

Full text of DOI:10.1021/jf048582h

J. Agric. Food Chem. 2004, 52, 8127−8131 8127
Fate of Damascenone in Wine: The Role of SO2
MERRAN A. DANIEL,†,‡,§ GORDON M. ELSEY,*,†,‡,§ DIMITRA L. CAPONE,‡,§
‡,§
MICHAEL V. PERKINS,†,§ AND MARK A. SEFTON
School of Chemistry, Physics and Earth Sciences, Flinders University, P.O. Box 2100,
Adelaide, South Australia 5001, Australia; The Australian Wine Research Institute, P.O. Box 197,
Glen Osmond, South Australia 5064, Australia; and Cooperative Research Centre for Viticulture,
P.O. Box 154, Glen Osmond, South Australia 5064, Australia
Damascenone has been shown to undergo reaction with common wine components. Following the
action of acid and heat alone, two bicyclic compounds, 4,9,9-trimethyl-8-methylenebicyclo[3.3.1]non-
6-en-2-one (2) and 4,4,9-trimethyl-8-methylenebicyclo[3.3.1] non-6-en-2-one (3), were isolated.
However, this conversion takes place only very slowly, if at all, under milder conditions (45 °C). When
treated with a variety of nucleophiles at pH 3.0 and 5.5, the concentration of damascenone in buffered
aqueous ethanol decreased by minor amounts (10-20%) except for cysteine and 2-mercaptoethanol
addition at pH 5.5 (∼40 and ∼30%, respectively) and SO2 (>90% at pH 3.0; 100% at pH 5.5). An
adduct from this last combination was prepared and shown to be the C9 sulfonic acid derivative of
damascenone. A detailed investigation into the effect of SO2 demonstrated that loss of damascenone
in model wine was directly related to the concentration of added SO2 but was essentially unaffected
by small changes in pH.
KEYWORDS: Damascenone; sulfur dioxide; sulfonic acid; nucleophilic components; degradation;
bicyclodamascenone
INTRODUCTION
â-Damascenone (damascenone) (1) is one of the most
important aroma compounds known to science. It is thoroughly
ubiquitous in nature and is one of the staples of the international
perfume industry (1). It is characterized by a “stewed apple” or
“pear” aroma and has been shown to be a positive contributor
to the fruity aroma of red wine varieties (2). Its aroma threshold
has been measured as 50 ng L-1 in model wine (3) and as low
as 2 ng L-1 in water (4), making it among the most potent
odorants known. Compounds such as damascenone belong to a
class of aroma compounds known as the C13-norisoprenoids,
which are believed to be degradation products of higher
carotenoid species such as neoxanthin (5).
Damascenone is formed in wine, initially, in significant
amounts (microgram per liter quantities), mainly during fer-
mentation, but the concentration of this compound has been
shown to then fall off, by as much as 75% over the first few
months of wine maturation (6). Wine is a complex medium,
containing a myriad of compounds displaying a wide range of
chemical reactivities. Damascenone contains two cross-conju-
gated enone moieties, and so one might imagine that dama-
scenone would be reactive toward nucleophilic species, common
to wine, under suitable conditions. In addition, external additives,
for example, sulfur dioxide, might contribute to the reduction
in concentration of damascenone over time. Demole and Enggist
(7) have shown that under acidic conditions, damascenone will
undergo intramolecular cyclization to produce two bicyclic
compounds, assigned on the basis of low-field NMR spectra,
as 2 and 3. However, the conditions they employed (180 °C,
p-TsOH) are completely atypical for the handling of wine, and
it is unclear whether these particular compounds could form
under more realistic winelike conditions.
We have conducted a study into the reactivity of damascenone
toward mild acid and toward various nucleophilic species in an
effort to determine and understand the fate of this important
aroma compound in wine.
* Corresponding author (telephone +61 8 8201 3071; fax +61 8 8201
2905; e-mail Gordon.Elsey@flinders.edu.au).
† Flinders University.
‡ The Australian Wine Research Institute.
§ Cooperative Research Centre for Viticulture.
10.1021/jf048582h CCC: $27.50
© 2004 American Chemical Society
Published on Web 12/04/2004

8128 J. Agric. Food Chem., Vol. 52, No. 26, 2004
Daniel et al.
MATERIALS AND METHODS
determination (r2) of 0.996 and the following linear regression
equation: y ) 2.59x + 0.122. To ensure that the accuracy of the analysis
was maintained, duplicate control wines, each with and without spiked
standard addition of 100 µg of damascenone per liter of wine, were
analyzed with every set of samples; the measured damascenone in these
samples was always in the range of 95 < [damascenone measured] <
100 µg/L.
Heating of Damascenone under Acidic Conditions. Method A.
Damascenone (500 mg, 2.6 mmol) and p-toluenesulfonic acid (5 mg)
were heated in a sealed tube in a microwave oven (400 W) for 10 min.
The two major products were isolated by column chromatography (5%
EtOAc/×4) as a mixture (195 mg, 39%). A second chromatography
provided pure samples of both 2 and 3 as colorless oils.
Method B. A solution of damascenone (100 mg, mmol) in model
wine (10% ethanol, pH 2.8, 50 mL) was heated in a sealed vessel at
100 °C for 11 days. The organic products were extracted with ether
and purified by column chromatography (5% EtOAc/×4) to give the
same two products as isolated from method A (26 mg, 26%) plus
unreacted damascenone.
Reagents and Instruments. Chemicals were purchased from Sigma-
Aldrich. All solvents used were of pesticide grade from OmniSolv
(Darmstadt, Germany). All organic solvent solutions were dried over
anhydrous sodium sulfate before being filtered. pH measurements were
made with an EcoScan pH 5/6 m (Eutech Instruments, Singapore),
which was calibrated before use. Column chromatography was per-
1
formed using silica gel 60 (230-400 mesh) from Merck. Routine H
and 13C NMR spectra were recorded with a Varian Gemini spectrometer
at operating frequencies of 300 and 75.5 MHz, respectively. HMBC
and CIGAR experiments were run on a Varian Unity Inova 600 MHz
spectrometer using Varian vnmr6.1c software. LC-MS spectra were
recorded on an API 300 electrospray mass spectrometer, and high-
resolution mass spectra were recorded on an Agilent ESI-TOF mass
spectrometer with accurate mass capabilities. Model wine buffer
solutions in the range pH 2.8-3.4 were prepared by saturating a 10%
ethanol solution with potassium hydrogen tartrate and adding 10%
tartaric acid solution until the required value was attained. Buffer
solutions at pH 5.5 were prepared by saturating a 10% ethanol solution
with potassium hydrogen phosphate and adding 10% citric acid solution
until the required value was attained.
4,9,9-Trimethyl-8-methylenebicyclo[3.3.1]non-6-en-2-one (2): δH
(CDCl3) 6.34 (1H, d J ) 9.8 Hz, H7), 5.93 (1H, m, H6), 5.00 (1H, s,
H11a), 4.99 (1H, s, H11b), 2.77 (1H, br s, H1), 2.45 (1H, m, H4), 2.13-
Procedures. Preparation of Samples for Analysis of Damascenone
(1). For analysis of the damascenone plus miscellaneous nucleophile
experiments, an aliquot of the reaction mixture (20 µL) was diluted to
10 mL with model wine in a 15 mL glass screw-cap vial with an
aluminum-lined cap (Supelco). To this was added an aliquot (100 µL)
of a solution of [2H4]damascenone (8) in ethanol (5 µg/mL) as internal
standard. Pentane/ethyl acetate (2:1, 3 mL) was added, and the mixture
was shaken briefly. A portion of the organic layer was then transferred
to a vial for GC-MS analysis. For analysis of the detailed damascenone
plus sulfur dioxide experiments, an aliquot (100 µL) of the solution of
[2H4]damascenone (8) in ethanol (5 µg/mL) was added directly as
internal standard to the hydrolysate sample (5 mL). Extraction was
performed as described above. For calculating the concentration of the
analytes in the wine samples, replicate standards were prepared at the
same time as the wine samples, by adding the internal standard solution
(100 µL, 5 µg/mL) and a solution of damascenone in ethanol (100 µL,
5 µg/mL) to dichloromethane (1.8 mL) and analyzing this mixture by
the GC-MS method (see below) to calculate the relative response
factors.
2.02 (2H, m, H3), 1.99 (1H, m, H5), 0.97 (3H, d J ) 6.8 Hz, H10),
0.97, 0.95 (6H, 2 × s, H12,13); δC (CDCl3) 210.8 (C2), 139.9 (C8), 129.7
(C6), 129.4 (C7), 116.8 (C11), 64.7 (C1), 45.9 (C5), 41.9 (C3), 36.3 (C9),
33.2 (C4), 26.5 (C12), 26.0 (C13), 19.3 (C10); MS, m/z (%) 190 (44),
175 (15), 148 (54), 146 (47), 133 (28), 120 (38), 119 (76), 105 (100),
91 (33), 79 (19), 77 (27), 69 (54), 41 (19).
4,4,9-Trimethyl-8-methylenebicyclo[3.3.1]non-6-en-2-one (3): δH
(CDCl3) 6.31 (1H, d J ) 9.9 Hz, H7), 5.96 (1H, m, H6), 5.03 (1H, s,
H12a), 4.99 (1H, s, H12b), 2.97 (1H, br s, H1), 2.40-2.28 (2H, m, H3a,9),
1.91-1.80 (2H, m, H3b,5), 1.06, 1.04 (6H, 2 × s, H10,11), 0.92 (3H, d J
) 6.7 Hz, H13); δC (CDCl3) 210.4 (C2), 137.8 (C8), 130.1 (C6), 128.8
(C7), 117.5 (C12), 59.7 (C1), 47.9 (C3), 46.8 (C5), 40.8 (C4), 31.2 (C9),
28.5 (C10), 27.7 (C11), 17.5 (C13); MS, m/z (%) 190 (11), 175 (12), 134
(7), 106 (85), 105 (24), 91 (100), 83 (15), 77 (11), 41 (11).
ConVersion of 2 and 3 into Their RespectiVe 2,4-Dinitrophenylhy-
drazones. Ketone 2 (14 mg, 0.07 mmol) in ethanol (0.6 mL) was treated
with 0.5 mL of a solution of 2,4-dinitrophenylhydrazine [prepared by
dissolving 2,4-dinitrophenylhydrazine (200 mg) in concentrated H2-
SO4 (1 mL) and diluted with water (1.4 mL) and 95% EtOH (5 mL)]
and left to stand for 2 days at 5 °C. The resultant precipitate was filtered
and washed successively with ethanol, saturated sodium bicarbonate
solution, and water before being recrystallized from 95% EtOH to give
orange needles (12 mg, 44%): mp 136-140 °C; δH (CDCl3) 11.10
(1H, s, NH), 9.05 (1H, d J ) 2.5 Hz, H16), 8.22 (1H, dd J ) 9.8 and
2.5 Hz, H18), 7.91 (1H, d J ) 9.8 Hz, H19), 6.28 (1H, d J ) 10.0 Hz,
H7), 5.80 (1H, dd J ) 10.0 and 6.6 Hz, H6), 4.94 (2H, br s, H11), 2.92
(1H, br s, H1), 2.47 (1H, dd J ) 15.7 and 5.2 Hz, H3a), 2.27 (1H, m,
H4), 1.94 (1H, m, H5), 1.74 (1H, dd J ) 15.7 and 11.7 Hz, H3b), 0.98
(3H, d J ) 6.7, H10), 0.94 (6H, 2 × s, H12,13); δC (CDCl3) 162.6 (C2),
142.2 (C8), 145.4, 137.5, 130.0, 130.0, 123.5, 116.5 (ArC), 129.9 (C7),
129.1 (C6), 116.2 (C11), 57.3 (C1), 46.3 (C5), 35.5 (C9), 31.4 (C4), 27.9
(C3), 26.7 (C12), 26.1 (C13), 19.6 (C10).
GC-MS Analysis. Samples were analyzed with a Hewlett-Packard
(HP) 6890N gas chromatograph fitted with a Gerstel MPS2 autosampler
and coupled to an HP 5973N mass spectrometer. The liquid injector
was operated in fast liquid injection mode with a 10 µL syringe (SGE,
Australia) fitted. The gas chromatograph was fitted with an ap-
proximately 30 m × 0.25 mm i.d. J&W fused silica capillary column
DB-WAX, 0.25 µm film thickness. The carrier gas was helium (BOC
Gases, ultrahigh purity), and the flow rate was 1.2 mL/min. The oven
temperature, started at 50 °C, was held at this temperature for 1 min,
then increased to 220 °C at 10 °C/min, and held at this temperature for
10 min. The injector was held at 200 °C and the transfer line at 250
°C. The sample volume injected was 2 µL, and the splitter, at 42:1,
was opened after 36 s. Fast injection was done in pulsed splitless mode
with an inlet pressure of 25.0 psi maintained until splitting. The glass
liner (Agilent Technologies) was borosilicate glass with a plug of
resilanized glass wool (2-4 mm) at the tapered end to the column.
Positive ion electron impact spectra at 70 eV were recorded in the range
m/z 35-350 for scan runs. For quantification of damascenone, mass
spectra were recorded in the selective ion monitoring (SIM) mode. The
ions monitored in SIM runs were m/z 69, 175, and 190 for damascenone,
and 73, 179, and 194 for [2H4]damascenone. Selected fragment ions
were monitored for 20 ms each. The underlined ion for each compound
was the ion typically used for quantitation, having the best signal-to-
noise and the least interference from other components. The other ions
were used as qualifiers.
Ketone 3 (20 mg, 0.1 mmol) in ethanol (0.8 mL) was treated exactly
as above, with 0.7 mL of a solution of 2,4-dinitrophenylhydrazine. The
product was collected and recrystallized from 95% EtOH to give orange
needles (25 mg, 64%): mp 174-177 °C; δH (CDCl3) 11.10 (1H, s,
NH), 9.06 (1H, d J ) 2.5 Hz, H16), 8.22 (1H, dd J ) 9.6 and 2.5 Hz,
H18), 7.92 (1H, d J ) 9.6 Hz, H19), 6.27 (1H, d J ) 9.9 Hz, H7), 5.86
(1H, dd J ) 9.9 and 6.6 Hz, H6), 5.00 (1H, br s, H12a), 4.96 (1H, br s,
H12b), 3.14 (1H, br s, H1), 2.26 (1H, m, H9), 2.18 (1H, d J ) 14.6 Hz,
H3a), 2.02 (1H, d J ) 14.6 Hz, H3b), 1.85 (1H, m, H5), 1.05, 0.97 (6H,
2 × s, H10,11), 0.89 (3H, d J ) 6.7 Hz, H13); δC (CDCl3) 162.3 (C2),
140.4 (C8), 145.5, 137.5, 129.9, 129.1, 123.6, 116.4 (ArC), 129.9 (C6),
129.5 (C7), 117.2 (C12), 52.6 (C1), 47.4 (C5), 39.4 (C4), 34.4 (C3), 31.9
(C9), 28.7 (C10), 27.5 (C11), 17.7 (C13).
Validation. The method was validated by a series of duplicate
standard additions of unlabeled damascenone (1.0-200 µg/L, n ) 9
× 2 for compound) to a dry white wine (Australian Chenin Blanc,
11.5% alc/vol, pH 3.40). The standard addition curve obtained was
linear throughout the concentration range, with a coefficient of
General Procedure for Heating of Damascenone with Various
Nucleophiles. Damascenone (0.1 g/L) was added to buffered aqueous
ethanol at either pH 3.0 or pH 5.5. The required nucleophile was added
to give a concentration of 0.1 g/L, and aliquots (5 mL) were sealed in

Fate of Damascenone in Wine
J. Agric. Food Chem., Vol. 52, No. 26, 2004 8129
glass ampules and stored at 45 °C for 1, 2, or 3 months. After the
required time, the ampules were removed and stored at -20 °C prior
to analysis. The ampules were opened, and the damascenone was
quantified as detailed above. Nucleophilic species added in this
experiment were cysteine, sodium acetate, mercaptoethanol, ethylamine,
and SO2 (added as sodium metabisulfite).
(500 mL, pH 3.2) and stirred at room temperature for 4 weeks. The
mixture was extracted first with ethyl acetate to remove unreacted
damascenone, and then the products were extracted with acetonitrile
and ethyl acetate (5:2), and the solvent was removed. NMR analysis
of the crude mixture showed the sulfonic acid 7 to be the major product.
Heating of Damascenone with SO2. Damascenone (1 mg/L) was
added to buffered aqueous ethanol at either pH 3.2 or pH 3.4. Sodium
metabisulfite was added to each to give sulfur dioxide concentrations
of either 80 ppm (120 mg/L Na2S2O5) or 200 ppm (300 mg/L). It was
noted that addition of these quantities of sodium metabisulfite had only
a negligible effect (∼0.01 pH unit for 300 mg added) on the pH of the
solutions. Aliquots of each were sealed in glass ampules and stored at
either room temperature (for 2, 4, 6, and 8 days and 2, 3,w and 4 weeks)
or 45 °C (for 12 h and 1, 2, 3, 4, 5, 6, 7, 10, and 14 days). After the
required time had elapsed, the ampules were opened, and the dama-
scenone was quantified as described above.
9-Hydroxymegastigma-3,5-dien-7-one (4). To a solution of 2,5,5-
trimethyl-1-acetylcylcohexa-1,3-diene (6) (9) (50 mg, 0.31 mmol) in
dry THF (2 mL) at -78 °C was added dropwise LiHMDS (0.34 mL,
1 M in THF, 0.34 mmol). After 1 h of stirring at this temperature,
acetaldehyde (0.017 mL, 0.31 mmol) was added and the reaction
allowed to stir for 90 min, during which time the solution was allowed
to warm to room temperature. The reaction was quenched with NH4Cl
solution, the solvent removed, and the product extracted with ether.
The ether extracts were washed with water and brine and dried (Na2-
SO4), and the solvent was evaporated. The alcohol was purified by
column chromatography (30% EtOAc/×4) to yield 4 (34 mg, 54%) as
a colorless oil: δH (CDCl3) 5.89-5.75 (2H, m, H3,4), 4.29 (1H, m,
H9), 2.77 (1H, dd J ) 18.3 and 2.7 Hz, H8a), 2.63 (1H, dd J ) 18.3
and 8.9 Hz, H8b), 2.09 (2H, dd J ) 4.1 and 1.5 Hz, H2), 1.72 (3H, s,
H13), 1.22 (3H, d J ) 6.5 Hz, H10), 1.08, 1.07 (6H, 2 × s, H11,12); δC
(CDCl3) 211.8, 141.3, 128.0, 127.9, 127.8, 63.8, 53.2, 39.7, 33.9, 26.2,
26.1, 22.3, 19.1; MS, m/z (%) 208 (4), 149 (39), 133 (31), 122 (31),
121 (100), 107 (20), 105 (42), 91 (25), 77 (15), 43 (38).
9-Ethoxymegastigma-3,5-dien-7-one (5). To metallic sodium (50 mg,
2.2 mmol) was added dry ethanol (1 mL). After the sodium had
completely dissolved, a portion of the sodium ethoxide solution (0.065
mL, 0.14 mmol) was added to a stirred solution of damascenone (27
mg, 0.14 mmol) in dry EtOH (2 mL). The reaction was stirred for 90
min before being quenched with NH4Cl solution and concentrated under
reduced pressure. The product was extracted with ether, washed (H2O,
brine), and dried (Na2SO4) and the solvent evaporated to give a 3:2
mixture of product plus damascenone. The product was purified by
column chromatography (CH2Cl2) to yield 5 (14 mg, 42%) as a pale
yellow oil: δH (CDCl3) 5.87-5.75 (2H, m, H3,4), 4.02 (1H, m, H9),
3.58 (1H, dq J ) 9.1 and 7.0 Hz, H14a), 3.44 (1H, dq J ) 9.1 and 7.0
Hz, H14b), 2.95 (1H, dd J ) 17.6 and 6.9 Hz, H8a), 2.50 (1H, dd J )
17.6 and 5.5 Hz, H8b), 2.08 (2H, dd J ) 3.8 and 1.2 Hz, H2), 1.74 (3H,
s, H13), 1.20 (3H, d J ) 6.2 Hz, H10), 1.15 (3H, t J ) 7.0, H15), 1.09,
1.07 (6H, 2 × s, H11,12); δC (CDCl3) 208.4, 142.0, 128.3, 127.6, 127.6,
70.9, 64.1, 52.7, 39.8, 33.9, 26.2, 26.1, 20.0, 19.0, 15.5; MS, m/z (%)
236 (5), 221 (7), 192 (3), 175 (6), 149 (73), 133 (49), 122 (100), 121
(85), 107 (60), 105 (54), 91 (40), 73 (76), 45 (72). HRMS (ESI): found
237.1857 (M + H+), C15H25O2 requires 237.1855; found 259.1673 (M
+ Na+), C15H24O2Na requires 259.1674.
RESULTS AND DISCUSSION
Loss of Damascenone under Acidic Conditions. Demole
and Enggist (7) have reported that damascenone undergoes acid-
catalyzed intramolecular cyclization to give two products, which
they assigned as 2 and 3 on the basis of low-field NMR and
mass spectral data. We repeated the original experiment under
modified conditions (400 W microwave, with added p-TsOH)
and obtained two compounds that were isolated by silica
chromatography. For both compounds, NMR and mass spectral
data obtained matched those reported for 2 and 3. In an attempt
to obtain crystals for X-ray crystallographic analysis, each of
the compounds was converted into the corresponding 2,4-
dinitrophenylhydrazone derivative. Unfortunately, the crystals
produced proved unsuitable for such analysis. However, long-
range coupling experiments (HMBC, CIGAR) on the hydrazone
derivatives were in complete agreement with the reported
structures.
In contrast to the initial experiment, which was conducted in
the absence of solvent, the reaction was repeated in 10% aqueous
ethanol solution (pH 2.8). After being heated at 100 °C for 11
days, the same two major products were observed in the product
mixture.
Two other (very minor) components of this reaction mixture
were the two C9 adducts of damascenonesthe hydrate 4 and
the ethanol adduct 5. Both have been observed previously as
components in the acid hydrolysis of suspected damascenone
precursors (10), although their structures were only tentatively
assigned. Authentic samples of each were therefore prepared,
and the assignment confirmed. The hydrate 4 was prepared by
addition of acetaldehyde across the enolate of 2,5,5-trimethyl-
1-acetylcylcohexa-1,3-diene (6) (9). The ethanol adduct 5 was
prepared by treating damascenone with sodium ethoxide in
ethanol. From analysis of the NMR spectra, it was clear that
the reaction had proceeded as expected. In both compounds the
two protons on C8 presented as the AB portion of an ABX
system, with the X portion (the lone hydrogen on C9) further
coupled to the protons on C10. Also, in both cases, the signals
for H3 and H4 were essentially unchanged from those of
damascenone itself, confirming the regiochemistry of the
products.
Loss of Damascenone Due to Nucleophilic Attack. There
are many species present in wine that could conceivably undergo
reaction with damascenone and account for its loss in aging
wines. These include various carboxylic acids, amines, and
thiols. An initial experiment was set up in which damascenone
was sealed with one of the added nucleophilic species, indicated
in Figure 1, in buffered aqueous ethanol solution (pH 3.0). As
all of the solutions (including the control) were prepared from
aqueous ethanol buffered with tartrate/tartaric acid, they all
contained these species as potential nucleophiles, with the
capacity to form adducts such as 4 and 5 and conjugates with
tartrate, as well as with the added nucleophiles listed in Figure
1.
Sulfonic Acid DeriVatiVe (7). A mixture of damascenone (95 mg,
0.5 mmol) and Na2S2O5 (150 mg) in DMF (5 mL) was heated at 45 °C
for 3 days. The DMF was removed, and the product was dissolved in
H2O, washed with EtOAc, and extracted with acetonitrile/EtOAc (5:1)
to yield, after drying and concentration, a yellow solid (51 mg, 35%):
δH (D2O) 6.03-5.87 (2H, m, H3,4), 3.47 (1H, ddq J ) 9.3, 3.0, and 6.7
Hz, H9), 3.35 (1H, dd J ) 19.1 and 3.0 Hz, H8a), 2.94 (1H, dd J )
19.1 and 9.3 Hz, H8b), 2.14 (2H, dd J ) 4.1 and 1.5 Hz, H2), 1.72 (3H,
s, H13), 1.37 (3H, d J ) 6.7 Hz, H10), 1.09 (6H, 2 × s, H11,12); δC
(D2O) 212.2 (C7), 140.8 (C6), 129.3 (C3), 129.0 (C5), 127.7 (C4), 50.7
(C9), 47.3 (C8), 39.3 (C2), 33.3 (C1), 25.6 (C11), 25.6 (C12), 18.4 (C13),
15.2 (C10); LC-MS (negative ion mode), m/z 271 (M - Na).
The solutions were heated at 45 °C for 60 days, and the
remaining damascenone was then quantified. Figure 1 shows
the damascenone remaining after 60 days. In most cases, the
falloff in damascenone content was minor (10-20%). Therefore,
under these conditions, the rate of formation of compounds 2-5,
Formation of the Sulfonic Acid DeriVatiVe (7) under Wine Condi-
tions. Damascenone (100 mg, 0.5 mmol) and sodium metabisulfite (100
mg, 0.5 mmol, equivalent to 1.0 mmol SO2) were added to model wine

8130 J. Agric. Food Chem., Vol. 52, No. 26, 2004
Figure 1. Damascenone remaining after 60 days at 45
Figure 2. Damascenone remaining after 90 days at 45
Daniel et al.
°
C, pH 3.0.
°
C, pH 5.5.
or of adducts with most of the nucleophiles present, is
insufficient to account for the decrease in damascenone during
bottle storage. However, when SO2 was added, the concentration
of damascenone decreased to <10% of the original value. It
was anticipated that increasing the pH to higher values might
result in an increase in concentration of the conjugate bases of
the added nucleophiles and thus an increase in the nucleophi-
licity of the medium. A further experiment was conducted at
pH 5.5, and the results are collected in Figure 2. Other than
reaction with sulfur dioxide, the higher pH (and longer reaction
time) had little effect except for the levels of damascenone with
cysteine and mercaptoethanol as the added nucleophiles, for
which the concentrations of damascenone fell to approximately
60 and 70% of its original value, respectively. Although these
results indicate a propensity for damascenone to react with thiols
under certain conditions, it is clear that the reductions observed
in the damascenone content are still insufficient to account for
the reported losses (6) of damascenone (up to 75% over 3
months at cellar temperature).
Figure 3. Effect of pH and added SO2 on the level of damascenone over
time, at 25 °C (upper) and 45 °C (lower).
nucleophile (introduced as sodium metabisulfite). After 60 days
at pH 3.0 and 45 °C, the level of damascenone remaining was
<10% of the original value. After 90 days at pH 5.5 and 45
°C, there was no detectable damascenone (LOD ∼ 0.5-1% of
starting concentration); that is, it had been completely consumed.
Sulfur dioxide (or its aqueous equivalent, bisulfite) is known
to react with carbonyl compounds to produce bisulfite adducts.
Accordingly, sodium metabisulfate was reacted with dama-
scenone in DMF and the product extracted with acetonitrile/
ethyl acetate. Spectral analysis revealed that the product isolated
was in fact the sulfonic acid derivative 7, where the sulfonate
functionality had been formed by 1,4-addition to the R,â-
unsaturated enone, rather than 1,2-addition across the ketone.
The regiochemistry of addition was established from the proton
Loss of Damascenone Due to SO2. The most striking entries
in Figures 1 and 2 are those corresponding to SO2 as the added

Fate of Damascenone in Wine
J. Agric. Food Chem., Vol. 52, No. 26, 2004 8131
(2) Aznar, M.; Lo´pez, R.; Cacho, J.; Ferreira, V. Prediction of aged
red wine aroma properties from aroma chemical composition.
Partial least squares regression models. J. Agric. Food Chem.
2003, 51, 2700-2707.
(3) Guth, H. Quantitation and sensory studies of character impact
odorants of different white wine varieties. J. Agric. Food Chem.
1997, 45, 3027-3032.
NMR spectrum where the two protons on C8 and the proton on
C9 form an ABX system (J ) 19.1, 9.3, and 3.0 Hz). The
spectrum is in close agreement with those for 4 and 5. Examples
from the literature reporting the observation of this unexpected
regiochemistry are extremely uncommon. Pfoertner (11) ob-
served conjugate addition when he treated various substituted
chalcones with sodium sulfite and sodium pyrosulfite in aqueous
methanol. Bisulfite has also been observed to add across a
double bond in certain heteroaromatic systems; Pitman and
Sternson (12) have measured equilibrium constants for such
processes.
To investigate whether 7 could form under more typical wine
conditions, damascenone was treated with sodium metabisulfite
in a model wine (pH 3.2, room temperature, 4 weeks). The
product was extracted with acetonitrile/ethyl acetate and ana-
lyzed by NMR, which showed that the major component was
indeed 7.
(4) Buttery, R. G.; Teranishi, R.; Ling, L. Identification of dama-
scenone in tomato volatiles. Chem. Ind. 1988, 238.
(5) Winterhalter, P.; Rouseff, R. Carotenoid-derived aroma com-
pounds: an introduction. Carotenoid-DeriVed Aroma Com-
pounds; Winterhalter, P., Rouseff, R., Eds.; ACS Symposium
Series 802; American Chemical Society: Washington, DC, 2002;
pp 1-17.
(6) Guth, H. Comparison of different white wine varieties in odor
profiles by instrumental analysis and sensory studies. In The
Chemistry of Wine FlaVor; Waterhouse, A. L., Ebeler, S. E.,
Eds.; ACS Symposium Series 714; American Chemical Soci-
ety: Washington, DC, 1999; pp 39-54.
A detailed quantitative assessment of the effect of SO2 on
the concentration of damascenone over time was conducted.
Two concentrations of SO2 (80 and 200 mg/L), two pH values
(3.2 and 3.4), and two temperatures (25 and 45 °C) were chosen,
with the final results displayed in Figure 3. As expected,
temperature played a key role in the rate of consumption of
damascenone, but this rate appeared to be effectively indepen-
dent of pH, within the narrow range employed in this experi-
ment. The time zero points show reduced levels of damascenone,
relative to the spiked amount (1000 ppb). The discrepancies
appear to closely parallel the amount of added sulfur dioxide;
when added at 80 mg/L the discrepancies are ∼10%, but when
added at 200 mg/L, the discrepancies are ∼25%. They are most
likely to be an artifact of the storage process. The individual
time zero samples were stored at -20 °C for ∼1 month prior
to analysis, and under these conditions some slow but significant
reaction could take place. All of the samples were analyzed at
the end of the experiment. However, the discrepancies in the
time zero points are substantially less than the changes in
damascenone concentration over the course of the experiment
and do not alter in any way the conclusions drawn.
(7) Demole, E.; Enggist, P. A chemical study of Virginia tobacco
flavor (Nicotiana tabacum L.). I. Isolation and synthesis of two
bicyclodamascenones. HelV. Chim. Acta 1976, 59, 1938-1943.
(8) Kotseridis, K.; Baumes, R.; Skouroumounis, G. K. Synthesis of
labelled [2H4] â-damascenone, [2H2] 2-methoxy-3-isobutylpyra-
zine, [2H3] R-ionone, and [2H3] â-ionone for quantification in
grapes, juices and wines. J. Chromatogr. A 1998, 824, 71-78.
(9) Phaff, R.; Bischofberger, N.; Mathies, P.; Petter, W.; Frei, B.;
Jeger, O. Carbonyl vs. epoxyketone photochemistry: Photolysis
of 1,2;3,4-diepoxy-2,6,6-trimethyl-1-cyclohexyl methyl ketone.
HelV. Chim. Acta 1985, 68, 1204-1216.
(10) Skouroumounis, G. K.; Sefton, M. A. Acid-catalyzed hydrolysis
of alcohols and their â-D-glucopyranosides. J. Agric. Food Chem.
2000, 48, 2033-2039.
(11) Pfoertner, v. K.-H. Substituted alkyl sulfonates by addition of
sodium hydrogen sulfide to chalcones. HelV. Chim. Acta 1980,
63, 664-667.
(12) Pitman, I. H.; Sternson, L. A. Method for estimating the
equilibrium constants for covalent addition of nucleophilic
reagents to heteroaromatic molecules. J. Am. Chem. Soc. 1976,
98, 5234-5238.
(13) Capone, D. L.; Simpson, R. F.; Cox, A.; Duncan, B.; Skourou-
mounis, G. K.; Sefton M. A. New insights into wine bottle
closure performancesflavour ‘scalping’ and cork taint. In
Proceedings of the 12th Australian Wine Industry Technical
Conference; Blair, R. J., Pretorius, I. S., Eds.; Australian Wine
Industry Technical Conference: Adelaide, SA, 2004; in press.
Decreases in damascenone content during wine maturation
can be attributed to several factors, but it is the interaction with
sulfur dioxide that is likely to be the most important process
accounting for such losses. These results also explain recent
observations (13) in our laboratories that the loss of dama-
scenone in white wine over time is less when such wines are
stored under an air headspace (with resultant consumption of
SO2) rather than in an anaerobic environment.
Received for review August 26, 2004. Revised manuscript received
October 24, 2004. Accepted October 26, 2004. This project is supported
by Australia’s grapegrowers and winemakers through their investment
body, the Grape and Wine Research and Development Corporation,
with matching funds from the Australian government, and by the
Commonwealth Cooperative Research Centres Program. The work was
conducted by The Australian Wine Research Institute and Flinders
University as part of the research program of the Cooperative Research
Centre for Viticulture.
ACKNOWLEDGMENT
We thank Profs. P. B. Høj and I. S. Pretorius, as well as Dr. E.
J. Waters, for their support and feedback.
LITERATURE CITED
(1) Pickenhagen, W. In FlaVor ChemistrysThirty Years of Progress;
Teranishi, E. L. W., Hornstein, I., Eds.; Kluwer Academic/
Plenum Publishers: New York, 1999; pp75-87.
JF048582H