Elucidation of the 1,3-sulfanylalcohol oxidation mechanism: An unusual identification of the disulfide of 3-sulfanylhexanol in sauternes botrytized wines

Article abstract of DOI:10.1021/jf102022s

A four-step purification method was developed to isolate a citrus odorant detected by gas chromatography-olfactometry (GC-O), which was apparently specific to Sauternes botrytized wines. A fragmentation pattern of the odorant was obtained by multidimensional gas chromatography- mass spectrometry- olfactometry (MDGC-MS-O). The exact mass measurement was used to determine its elemental formula as C₆H₁₂OS. On the basis of these data, the unusual structure of 3-propyl-1,2-oxathiolane was synthesized and characterized for the first time. This confirmed its identification. Its occurrence in Sauternes wine extracts was demonstrated to result from the thermal oxidative degradation of 3-sulfanylhexanol disulfide (3,3'-disulfanediyldihexan-1-ol) in the GC injector. This disulfide was synthesized and then firmly identified for the first time in Sauternes wine. Although the presence of 3-sulfanylhexanol oxidation products had previously been reported in natural extracts (but not wine), the full oxidation pathway from 3-sulfanylhexanol to 3-propyl-γ-sultine via 3,3'- disulfanediyldihexan-1-ol was clearly established for the first time. Because the disulfide has mainly been detected in Sauternes botrytized wines, this finding suggested a singular reactivity of 3-sulfanylhexanol in botrytized wines, thus opening up a wide range of new opportunities in wine chemistry.

Full text of DOI:10.1021/jf102022s

10606 J. Agric. Food Chem. 2010, 58, 10606–10613
DOI:10.1021/jf102022s
Elucidation of the 1,3-Sulfanylalcohol Oxidation Mechanism:
An Unusual Identification of the Disulfide of 3-Sulfanylhexanol
in Sauternes Botrytized Wines
†
E
LISE
S
ARRAZIN,*,†,‡
ERNARD
S
VITLANA
S
HINKARUK,§ MONIQUE
P
ONS,† CECILE
†
T
HIBON
,
,^
B
B
ENNETAU
,
AND
P
HILIPPE
D
ARRIET
†UMR INRA 1219 0nologie, Institut des Sciences de la Vigne et du Vin, Universite
Bordeaux II, 210 Chemin de Leysotte, CS 5008, F-33882 Villenave d’Ornon, France, §ENITA de Bordeaux,
1 cours du General de Gaulle, CS 40201, F-33175 Gradignan Cedex, France, Universite Bordeaux 1,
Institut des Sciences Moleculaires, 351 cours de la Liberation, F-33405 Talence Cedex, France, and
^CNRS, UMR 5255 ISM, F-33405 Talence Cedex, France. ‡ Present address: Laboratoire de Chimie des
´
Victor Segalen
´
´
´
´
´
^
cules Bioactives et des Aromes, UMR CNRS 6001, Universite
Parc Valrose, 06108 Nice Cedex 1, France.
Mole
´
´
de Nice-Sophia Antipolis,
A four-step purification method was developed to isolate a citrus odorant detected by gas
chromatography-olfactometry (GC-O), which was apparently specific to Sauternes botrytized
wines. A fragmentation pattern of the odorant was obtained by multidimensional gas chromato-
graphy-mass spectrometry-olfactometry (MDGC-MS-O). The exact mass measurement was
used to determine its elemental formula as C6H12OS. On the basis of these data, the unusual
structure of 3-propyl-1,2-oxathiolane was synthesized and characterized for the first time. This
confirmed its identification. Its occurrence in Sauternes wine extracts was demonstrated to result
from the thermal oxidative degradation of 3-sulfanylhexanol disulfide (3,30-disulfanediyldihexan-1-ol)
in the GC injector. This disulfide was synthesized and then firmly identified for the first time in
Sauternes wine. Although the presence of 3-sulfanylhexanol oxidation products had previously been
reported in natural extracts (but not wine), the full oxidation pathway from 3-sulfanylhexanol to
3-propyl-γ-sultine via 3,30-disulfanediyldihexan-1-ol was clearly established for the first time. Because the
disulfide has mainly been detected in Sauternes botrytized wines, this finding suggested a singular
reactivity of 3-sulfanylhexanol in botrytized wines, thus opening up a wide range of new opportunities in
wine chemistry.
KEYWORDS: 3-Sulfanylhexanol; 3-propyl-1,2-oxathiolane; 3,30-disulfanediyldihexan-1-ol; 3-propyl-γ-sultine;
Sauternes wines
INTRODUCTION
Among naturally occurring 1,3-sulfanylalcohols, 3-sulfanyl-
hexanol (1) is one of the compounds most studied and analyzed.
It is a powerful odorant, reminiscent of citrus and tropical fruit. It
was initially isolated from passion fruit and contributes greatly to
its characteristic aroma (4). It has also been identified as one of
the key aroma compounds in Sauvignon blanc wines (5) and,
Naturally occurring 1,3-sulfanylalcohols are highly volatile
compounds with powerful, penetrating aromas, responsible for
the sensory characteristics and quality of various foods and
beverages. They exhibit both attractive (tropical fruit and box
tree) and unpleasant (cat urine, sweat, and onion) odors (1-3).
Indeed, the olfactory descriptors associated with 1,3-sulfanylal-
cohols depend upon the length and branching of their alkyl chain.
The sensory impact of these compounds also depends upon their
concentration, varying from favorable to unpleasant with in-
creasing concentrations in the matrix (1-3). Although 1,3-sulfa-
nylalcohols are generally present at trace levels, they contribute
strongly to the overall aroma of natural products, foods, and
beverages, especially wine. However, the mechanisms involved in
their formation, reactivity, and stability, as well as their particular
role and additive effects in complex sensory perception, have not
yet been fully elucidated and continue to be investigated.
€
more recently, in Riesling, Gewurztraminer (6), Petit and Gros
Manseng (6), Cabernet Sauvignon (7,8), Merlot (7,8), and Swiss
Petite Arvine (9). The major sensory impact of 3-sulfanylhexan-
1-ol makes its origin and reactivity of great interest in wine
chemistry.
Recent research has reported the major contribution of 3-sul-
fanylhexanol to the aroma of Sauternes botrytized wines (6, 10).
Botrytized sweet wines are produced by a particular wine-making
process only encountered in a few areas in the world, which
contributes to their uniqueness. Indeed, they are produced from
overripe grapes affected by the Botrytis cinerea fungus under
specific climatic conditions, with alternating foggy mornings and
sunny afternoons (11). Because of the unusual fungal develop-
ment, the composition of the grapes is deeply modified and
*To whom correspondence should be addressed. Telephone: þ33-4-
92-07-61-69. Fax: þ33-4-92-07-61-25. E-mail: elise.sarrazin@unice.fr.
pubs.acs.org/JAFC
Published on Web 09/21/2010
© 2010 American Chemical Society

Article
J. Agric. Food Chem., Vol. 58, No. 19, 2010 10607
requires a customized wine-making process. In addition, botry-
tized wines are generally aged longer in oak barrels than dry white
wines. All of these factors contribute to the development of a
complex aromatic profile different from that of dry white
wines. This exceptional range of aromas may evoke citrus and
dried fruits in young botrytized wines, orange peel in older
wines, and honey or waxy nuances in wines that undergo
oxidative aging (12).
Identification of 3-Propyl-1,2-oxathiolane (4). Liquid-Liquid
Extraction. Sauternes wine (4.5 L) was extracted with dichloromethane
(2 ꢀ 600 mL), with magnetic stirring, for 10 min each time. The combined
organic phases were centrifuged (4000 rpm) for 15 min and separated in a
separating funnel. They were then washed with sodium p-hydroxymercur-
ibenzoate solution [1 mM in Trizma base (2-amino-2-(hydroxymethyl)-
1,3-propanediol) at 0.2 M] (2 ꢀ 120 mL), for 10 min each time. The organic
phase obtained was dried over anhydrous sodium sulfate and concentrated
to 2 mL using a rotary evaporator.
Sauternes, as well as Tokaji and late-harvest Riesling wines
from Alsace, Rheingau, and Mosel regions, are some of the most
prestigious botrytized wine appellations. Sauternes wines are
produced in the region of Bordeaux, France, from two grape
High-Performance Liquid Chromatography (HPLC) Fractionation.
The concentrate was purified by reverse-phase HPLC, using the method
developed by Pineau et al. (26). The column used was a 318 mm ꢀ 3.9 mm
inner diameter, 4 μm, Novapak C18 (Waters, Guyancourt, France). The
chromatographic conditions included a flow rate of 0.5 mL/min and an
injection volume of 250 μL. The linear program gradient involved phase A,
water, and phase B, ethanol, 0% B reaching 100% B in 50 min. A fraction
collector (Bio-Rad Laboratories, Marnes-la-Coquette, France) was con-
nected to the end of the column to collect 1 mL eluted solvent every 2 min.
The HPLC eluate was recovered in 25 separate fractions, evaluated for
their smell.
varieties (i.e., Semillon and Sauvignon blanc), also used to make
´
dry white wines. In recent years, the typical aromas of Sauternes
wines have been largely studied (10, 13-18). Recently, Sarrazin
et al. (14) demonstrated the contribution of 3(2H)-furanones
(namely, homofuraneol, furaneol, and norfuraneol) and pheny-
lacetaldehyde to caramel and honey nuances, respectively, as
well as the major role of volatile thiols in the citrus aromas of
Sauternes wines (10). Three new sulfanyl alcohols, 3-sulfanylpen-
tan-1-ol, 3-sulfanylheptan-1-ol, and 2-methyl-3-sulfanylbutan-
1-ol, in addition to the two well-known thiols, 3-sulfanylhexanol
and 4-methyl-4-sulfanylpentan-2-one, have been identified and
characterized. Nevertheless, one citrus odoriferous component
detected by gas chromatography-olfactometry (GC-O) was not
identified (14). This odorant, not detected by GC-O in dry white
Fraction Re-extraction. The 25 fractions were again extracted with
dichloromethane, as described by Pons et al. (23). The alcohol content of
the fractions eluted by HPLC was adjusted to 12% (v/v). Then, each
fraction was extracted with dichloromethane (3 ꢀ 0.5 mL). The solvent
extract was concentrated under nitrogen to 200 μL and used for GC-O
analysis, as well as MDGC-MS-O analysis.
Extraction and Separation of 3-Sulfanylhexanol Disulfide (6)
from Wine. Liquid-Liquid Extraction. Sauternes wine (4.5 L) was
extracted with dichloromethane (300, 150, and 150 mL), with magnetic
stirring, for 10 min each time. The combined organic phases were
centrifuged (4000 rpm) for 15 min and separated in a separating funnel.
The organic phase obtained was dried over anhydrous sodium sulfate and
concentrated to 2 mL using a rotary evaporator. The solvent extract was
then concentrated under nitrogen to 750 μL.
HPLC Fractionation. The wine extract was purified by normal-phase
HPLC. The column used was a 250 mm ꢀ 4.6 mm inner diameter, 5 μm,
SunFire silica column (Waters, Guyancourt, France). The chromato-
graphic conditions included a flow rate of 1.0 mL/min and an injection
volume of 250 μL. The linear program gradient involved phase A, pentane,
phase B, ether, and phase C, ethyl acetate: 0-5 min, 100% A; 5-12 min,
100% B; then reaching 80% B and 20% C in 3 min; maintained at 80% B
and 20% C for 5 min; then reaching 50% B and 50% C in 3 min;
maintained at 50% B and 50% C for 5 min; and finally, reaching 100% C
in 10 min. A fraction collector (Bio-Rad Laboratories, Marnes-la-
Coquette, France) was connected to the end of the column to collect
fractions of the eluted solvent every minute.
wines made from the same grape varieties (i.e., Semillon and
´
Sauvignon blanc), was thus apparently specific to botrytized
wines; therefore, an experiment was designed to identify it.
The detection and identification of trace compounds in complex
matrices, such as wine, generally require numerous purification
steps to suppress major volatile compounds (19-21). Multidimen-
sional gas chromatography, especially heart-cut bidimensional gas
chromatography, now offers an elegant alternative approach for
identifying trace odorants. This hyphenated technique, combined
with olfactometry, has recently been used to identify off-flavors
in wines (22, 23) and other alcoholic beverages (24, 25), thus
suggesting new solutions for trace analysis.
In this work, we explain the origin of the specific citrus odorant
detected by GC-O in Sauternes wine extracts. Sauternes wine
was purified via a four-step method and analyzed using multi-
dimensional gas chromatography combined with mass spectro-
metry and olfactory detection (MDGC-MS-O). The zone of
interest was identified as 3-propyl-1,2-oxathiolane, which was
synthesized and fully characterized for the first time. The chemi-
cal origin of 3-propyl-1,2-oxthiolane detected by GC-O was then
investigated.
Derivatization with BSTFA. Fractions 17-25 were combined, then
evaporated to dryness in a vial, and capped under nitrogen flow. The dry
residue was resuspended in BSTFA-Regisil solution (20 μL) with
anhydrous pyridine (50 μL). The mixture was incubated at 70 ꢀC for
15 min. A total of 2 μL of derivatized sample were analyzed by gas
chromatography-mass spectrometry (GC-MS).
Heart-Cut MDGC-MS-O. The MDGC separation was per-
formed on two GC instruments, as described by Pons et al. (23). Instru-
ment 1 was a Hewlett-Packard 5890 series II (Agilent Technologies,
Massy, France), and Instrument 2 was a 6890 GC coupled with a 5973
quadrupole mass spectrometer (Agilent Technologies, Massy, France).
The two instruments were connected with a West 4400 temperature-
controlled transfer line set at 230 ꢀC (ILS, Lyon, France). A cryotrap was
also placed at the start of the second column for cryofocusing. The column
used for pre-separation was a 30 m ꢀ 0.25 mm inner diameter, 0.25 μm,
SPB1 (Supelco, from Sigma-Aldrich, Saint Quentin Fallavier, France).
The second column was a 50 m ꢀ 0.25 mm inner diameter, 1.0 μm, BPX5,
BPX50, or BPX70 (SGE, Milton Keynes, U.K.). The MDGC system was
operated under constant pressure to maintain the balance between the two
columns throughout the oven temperature program. The end of the second
column was split (1:1) via a crosspiece (Gerstel, from RIC, Lille, France)
between the MS detector and the ODP II sniffing port (Gerstel, from RIC,
Lille, France). The MS detector functioned in electronic impact (70 eV) or
chemical ionization mode (CH4 reactant). Mass spectra were taken over
the m/z 40-300 range.
MATERIALS AND METHODS
Wine Samples. The botrytized wines analyzed in this study were all
from the Bordeaux region and the 2003 vintage. They were produced from
Se
procedures.
Standard Products. Dichloromethane (Chromasolv grade), diethyl
´
millon and Sauvignon blanc grapes, according to standard winemaking
ether, ethyl acetate, sodium p-hydroxymercuribenzoate, Trizma base
(99%), and anhydrous ethyl acetate were purchased from Sigma-Aldrich
Chemicals (Saint Quentin Fallavier, France). N,O-Bis(trimethylsilyl)-tri-
fluoroacetamide (BSTFA) in mixture with 1% trimethylchlorosilane
(BSTFA-Regisil) was from Regis Technologies (Morton Grove, IL).
Anhydrous pyridine was purchased from Interchim (Montluc-on, France).
Ethanol (Lichrosolv grade) was supplied by Merck (Martillac, France).
Pentane (Prolabo Normapur grade) was purchased from VWR
(Fontenay-sous-Bois, France) and distilled to improve its quality. 3-Sul-
fanylhexanol (>95%) and the chemical reagents used for chemical
synthesis were obtained from Alfa Aesar (Bischheim, France). Anhydrous
sodium sulfate was supplied by VWR (Fontenay-sous-Bois, France).

10608 J. Agric. Food Chem., Vol. 58, No. 19, 2010
Sarrazin et al.
MDGC Conditions for Wine Extract Analysis. The injector tempera-
ture was set at 230 ꢀC. Oven 1 was programmed from 45 ꢀC (1 min) to
250 ꢀC at 3 ꢀC/min, followed by a 40 min isotherm to ensure that all of
compounds were eluted. The cut time (28.8-29.8 min) was selected.
During this period, the cryogenic trap was maintained at -50 ꢀC. The
temperature of oven 2 was initially set at 45 ꢀC for 29.8 min and then
raised to 240 ꢀC at 3 ꢀC/min, followed by a 30 min isotherm. Helium 5.3
(Linde Gas, Bassens, France) was used as the carrier gas, with a flow
rate of 2 mL/min.
GC-MS Analysis with Exact Mass Measurement Used To
Identify 3-Propyl-1,2-oxathiolane (4). Exact mass measurement
was kindly performed by Dr. Estelle Delort (Firmenich SA, Geneva,
Switzerland). Analyses were carried out using a 7890 GC (Agilent
Technologies) coupled to a GCT Premier time of flight mass spectrometer
(Waters, Milford, MA). The column used was 30 m ꢀ 0.25 mm inner
diameter, 0.25 μm, SPB1 (Supelco from Sigma-Aldrich, Saint Quentin
Fallavier, France). A total of 1 μL extract was injected using a split injector
(250 ꢀC, split ratio of 1:5). The carrier gas was helium, with a flow rate of
1.0 mL/min. For all analyses, the oven temperature program was as
follows: 60 ꢀC for 5 min and then a 5 ꢀC/min gradient to 250 ꢀC, followed
by a 60 min isotherm. The acquisition time was set to 0.50 s, with an
interscan delay of 0.01 s, over a mass range of 1-350 Da. Spectra were
recorded in electron impact (EI) mode, using an electron energy of 70 eV,
emission current of 608 μA, trap current of 200 μA, and source tempera-
ture of 200 ꢀC. Calibration was performed using heptacosa (perfluoro-
tributylamine, mass spectrometry grade, Apollo Scientific Ltd., Bradbury,
U.K.). Calibration data were collected for 1 min in centroid mode. A total
of 60 spectra were summed to generate a 24-point calibration curve from
m/z 69 to 614. The curve was fitted to a second-order polynomial such that
the standard deviation of the residuals was 0.001 amu or lower. Heptacosa
was continuously introduced into the ion source, and the ion m/z 218.9856
was used as a lock mass. Mass spectra and molecular formula were
obtained using MassLynx software (Waters). The difference, d, between
the exact mass calculated from the molecular formula and that measured
is calculated by the software and expressed in parts per million (ppm)
(d = (Mmeas. - Mcalc.)/Mcalc. ꢀ 106).
Figure 1. Synthesis of of 3-propyl-1,2-oxathiolane (4), 3-propyl-γ-sultine
(5), and 3,30-dithiobishexan-1-ol (6). Reagents and conditions: (a) Me3SiCl,
Et3N, and THF; (b) N-bromophthalimide and Py; and (c) pyrolysis in vacuo.
Mass spectra were recorded over the m/z 30-550 range, with 5 ppm
sensitivity.
Synthesis. Experimental Equipment. 1H and 13C nuclear magnetic
resonance (NMR) spectra, as well as 2D NMR experiments, were recorded
with a Bruker AC-300 FT (1H, 300 MHz; 13C, 75 MHz). Chemical shifts
(δ) and coupling constants (J) are expressed in ppm and Hz, respectively.
No internal standard was present in NMR samples. Spectra were
referenced using the frequency of the deuterated solvent (the lock
frequency). Standard Bruker pulse sequences were used for homo- and
heteronuclear correlation experiments [correlation spectroscopy (COSY),
heteronuclear multiple-quantum coherence (HMQC), and heteronuclear
multiple-bond correlation (HMBC)]. Infrared (IR) spectra were recorded
with a Perkin-Elmer Paragon 1000 FTIR spectrophotometer. Thin-layer
chromatography (TLC) was performed on TLC plates: thickness, 0.25
˚
mm; particle size, 15 μm; pore size, 60 A (SDS, Peypin, France). Merck
silica gel 60 (70-230 mesh and 0.063-0.200 mm) was used for flash
chromatography. Spots were revealed with ultraviolet (UV) as well as
KMnO4 (0.05% in water). Tetrahydrofuran (THF) was dried by refluxing
a solution containing sodium wires and benzophenone under nitrogen and
distilled immediately before use. All moisture-sensitive reactions were
carried out under an argon atmosphere in oven- or flame-dried glassware.
Synthesis of 3-Propyl-1,2-oxathiolane, 3-Propyl-γ-sultine, and 3,30-
Dithiobishexan-1-ol. The compounds were synthesized from 3-sulfanyl-
hexanol (1) using an optimized procedure, similar to that proposed by
Davis and Whitham (27) (Figure 1). Freshly distilled chlorotrimethylsilane
(6.32 mL, 49.5 mmol, 1.1 equiv) was added dropwise to a solution of
3-sulfanylhexanol (6.03 g, 6.2 mL, 45 mmol) and freshly distilled triethy-
lamine (5 g, 6.9 mL, 49.5 mmol, 1.1 equiv) in anhydrous THF (50 mL), at
room temperature, under an argon atmosphere. The mixture was stirred
for 16 h. The white precipitate that formed was filtered, and the solvent
evaporated under reduced pressure. The colorless, viscous liquid contain-
ing 1-O-trimethylsilyl-3-hexanthiol (2) (9.2 g, 99% purity by GC-MS and
NMR) was immediately dissolved in anhydrous pyridine (10 mL, 8 equiv)
under an argon atmosphere and cooled to 0 ꢀC. N-Bromophthalimide
(12.21 g, 54 mmol, 1.2 equiv) was added in one portion. The resulting
mixture was stirred at 0 ꢀC for 1 h, then allowed to warm to room
temperature overnight, and filtered. The filtrate was washed with 30%
citric acid aqueous solution (180 mL). The aqueous layer was extracted
with diethyl ether (2 ꢀ 100 mL). Combined organic layers were washed
with brine (100 mL), dried over magnesium sulfate, filtered, and concen-
trated under reduced pressure to give 2-(1-hydroxyhexan-3-ylthio)isoindo-
line-1,3-dione (3) as an orange oil (9 g, 62%). A Kugelrohr short-path
vacuum distillation apparatus was used for the thermal decomposition of 3
and the fractionation of the resulting compounds, by smooth heating
in vacuo as follows: at 60 ꢀC and 9 mmHg for 5 min, 80 ꢀC and 9 mmHg for
10 min, 80 ꢀC and 1 mmHg for 1 min, and finally, 100 ꢀC and 1 mmHg for
2 h. Three fractions were collected and characterized by NMR (1H, 13C,
HMBC, HMQC, and COSY), IR, and GC-MS. The first fraction,
collected in a cold trap (liquid nitrogen), gave 3-propyl-1,2-oxathiolane
(4) as a colorless liquid (40 mg, 0.7%, ∼85% purity by GC-MS and
NMR). Compound 4 was highly unstable and oxidized rapidly to
3-propyl-γ-sultine (3-propyl-1,2-oxathiolane-2-oxide) (5). The distillate
collected in the first receiving bulb contained compound 5. This fraction
GC-O. Analyses were carried out using a Hewlett-Packard 5890 GC,
equipped with a flame ionization detector (from Agilent Technologies,
Massy, France) and a ODO-1 sniffing port (SGE, Milton Keynes, U.K.).
The column used was a 50 m ꢀ 0.25 mm inner diameter, 1.0 μm, BPX5
(SGE, Milton Keynes, U.K.). A total of 2 μL of extract was injected using
a splitless injector(230 ꢀC; purge time, 1 min;purge flow, 50 mL/min) or an
on-column injector, at oven temperature (50 ꢀC). For all analyses, the oven
temperature program was as follows: 50 ꢀC for 1 min and then a 3 ꢀC/min
gradient to 250 ꢀC, followed by a 15 min isotherm. The carrier gas was
hydrogen (Linde Gas, Bassens, France), with a flow rate of 1.0 mL/min.
GC-MS. Analyses were carried out using a 6890-5973 GC-MS
(Agilent Technologies, Massy, France). The column used was a 50 m ꢀ
0.25 mm inner diameter, 1.0 μm, BPX5 (SGE, Milton Keynes, U.K.). A
total of 2 μL of extract was injected using a splitless injector (230 ꢀC; purge
time, 1 min; purge flow, 50 mL/min) or an on-column injector, at oven
temperature (50 ꢀC). For all analyses, the oven temperature program was
as follows: 50 ꢀC for 1 min and then a 3 ꢀC/min gradient to 250 ꢀC,
followed by a 15 min isotherm. The carrier gas was hydrogen (Linde Gas,
Bassens, France), with a flow rate of 1.0 mL/min. The 5973 MS detector,
functioning in EI mode (70 eV), was connected to the GC via a transfer line
heated to 250 ꢀC. Mass spectra were taken in the SCAN mode over the
m/z 40-300 range.
GC-MS Analysis with Exact Mass Measurement Was Used To
Identify 3-Sulfanylhexanol Disulfide (6) in Wine. Analyses were
carried out using a 7890 GC (Agilent Technologies, Massy, France)
coupled to an AccuTOF JMS-T100GC time of flight mass spectrometer
(JEOL Europe, Croissy sur Seine, France). The column used was 60 m ꢀ
0.25 mm inner diameter, 0.25 μm, DB1 (J&W, from Agilent Technologies,
Massy, France). A total of 2 μL extract was injected using a splitless
injector (240 ꢀC; purge time, 1 min; purge flow, 50 mL/min). The carrier
gas was helium, with a flow rate of 1.0 mL/min. For all analyses, the
temperature program was as follows: 45 ꢀC for 1 min, then a 5 ꢀC/min
gradient to 200 ꢀC, 3 ꢀC/min to 240 ꢀC, and then 5 ꢀC/min to 280 ꢀC,
followed by a 15 min isotherm. The JEOL detector, functioning in EI
mode (70 eV), was connected to the GC via a transfer line heated to 250 ꢀC.

Article
J. Agric. Food Chem., Vol. 58, No. 19, 2010 10609
was further purified by silica gel chromatography, using petroleum ether/
diethyl ether (50:50), to afford compound 5 as a colorless liquid (0.2 g, 3%,
þ95% purity by GC-MS and NMR). NMR and IR data were similar to
those already reported (28, 29), indicating that 3-propyl-γ-sultine con-
sisted of a mixture of the two pairs of diastereoisomers [form trans
corresponds to the unlike (2S3R þ 2R3S) mixture of stereoisomers, and
form cis corresponds to the like (2R3R þ 2S3S) mixture of stereoisomers].
In our case, the compound 5 was obtained in 75:25 trans/cis ratio and
cannot be separated by standard silica gel chromatography. The aniso-
tropic effect of the sulfinyl functional group on nearby nuclei produced the
deshielding of H-3 and H-6 signals in trans and cis pairs, respectively.
Together with 2D NMR analysis, it allowed the full 1H and 13C chemical-
shift assignment for both trans and cis products.
The last fraction, obtained from the distilling flask (pot flask) and the
first collecting flask, contained the disulfide (6). It was purified by silica gel
chromatography, using petroleum ether/diethyl ether (50:50), to give
compound 6 as a colorless liquid (6.7 g, 56%, >99% purity by GC-MS
and NMR). The MS data were identical to those previously published (30).
1-O-Trimethylsilyl-3-hexanthiol (2). 1H NMR (300 MHz, CDCl3) δ:
0.13 (s, 9H, Si(CH3)3), 0.91 (t, J = 6.6 Hz, 3H, CH3-6), 1.34 (d, J = 7.15,
1H, SH), 1.37-1.69 (m, 5H, CH2-2a, CH2-4, CH2-5), 1.91 (m, 1H, CH2-
2b), 2.95 (m, 1H, CH-3), 3.75 (m, 2H, CH2-1). MS (70 eV), m/z (%): 206
(1) [Mþ], 191 (4), 157 (39), 143 (8), 129 (55), 121 (29), 116 (67), 103 (34), 91
(56), 88 (68), 83 (55), 73 (100), 55 (43), 47 (9), 45 (14).
Figure 2. Partial GCchromatogram of winefractions14-15 obtained with
three selected ions (m/z 55, 89, and 132). The retention time of the peak
indicated by an arrow corresponds to 3-propyl-1,2-oxathiolane.
CH2-10). 13C NMR (75 MHz, CDCl3) δ: 13.7 and 13.9 (CH3-6 and CH3-
60), 20.0 and 20.05 (CH2-5 and CH2-50), 36.9, 37.0, 37.1, and 37.15 (CH2-2,
CH2-20, CH2-4, and CH2-40), 59.9 and 60.0 (CH2-1 and CH2-10), 48.7 and
49.6 (CH-3 and CH-30). COSY, HMQC and HMBC data (not shown)
confirmed the signal attribution. IR, ν (cm-1): 549, 734, 1054, 1465, 1727,
3373. MS (70 eV), m/z (%): 266 (16) [Mþ], 148 (23), 115 (9), 101 (10), 84
(5), 83 (83), 57 (6), 55 (100), 45 (6), 43 (14).
2-(1-Hydroxyhexan-3-ylthio)isoindoline-1,3-dione (3). 1H NMR (300
MHz, CDCl3) δ: 0.91 (t, J = 7.2 Hz, 3H, CH3-6), 1.33-1.90 (m, 6H, CH2-
2, CH2-4, CH2-5), 2.86 (m, 1H, CH-3), 2.92 (br s, 1H, OH), 3.65-3.85 (m,
2H, CH2-1), 7.78 (d, J = 8.6, 2H, CHar), 7.91 (d, J = 8.6, 2H, CHar).
3-Propyl-1,2-oxathiolane (4). 1H NMR (300 MHz, CDCl3) δ: 0.93
(t, J = 7.2 Hz, 3H, CH3-8), 1.32-1.60 (m, 2H, CH2-7), 1.35-1.78 (m, 2H,
CH2-6), 1.81-2.05 (m, 1H, CH-4b), 2.15-2.33 (m, 1H, CH-4a), 3.16-3.29
(m, 1H, CH-3), 3.33-3.85 (m, 2H, CH2-5). 13C NMR (75 MHz, CDCl3) δ:
14.0 (CH3-8), 22.5 (CH2-7), 37.0 (CH2-4), 38.5 (CH2-6), 53.5 (CH-3),
75.1 (CH2-5). COSY data: H-8 f H-7; H-7 f H-6, H-8; H-6 f H-3, H-7;
H-4b f H-3, H-4a, H-5; H-4a f H-3, H-4b, H-5; H-3 f H-4a, H-4b, H-6;
H-5 f H-4a, H-4b. IR, ν (cm-1): 480, 733, 1055, 1464, 3500. MS (70 eV),
m/z (%): 132 (40) [Mþ], 103 (4), 89 (65), 83 (19), 77 (14), 67 (7), 61 (15), 55
(100), 45 (16), 43 (12). HR-MS (EI, 70 eV), m/z: 132.0644 for [M]þ (calcd
for [C6H12OS]þ, 132.0609), 89.0054 for [C3H5OS]þ (calcd, 89.0061).
3-Propyl-γ-sultine (5). IR, ν (cm-1): 732, 903, 1125, 1465.
1,2-Bis(1-O-trimethylsilylhexan-3-yl)disulfide (7). The compound 6
was silylated as already described for the synthesis of 2.
1H NMR (300 MHz, CDCl3) δ: 0.13 (s, 18H, Si(CH3)3), 0.93 (t, J = 7.3
Hz, 6H, CH3-6, CH3-60), 1.39-1.51 (m, 4H, CH2-5, CH2-50), 1.54-1.68
(m, 4H, CH2-4, CH2-40), 1.83 (apparent q, J = 6.4 Hz, 4H, CH2-2, CH2-
20), 2.79 (apparent p, J = 6.4 Hz, 2H, CH-3, CH-30), 3.66-3.80 (m, 4H,
CH2-1, CH2-10). 13C NMR (75 MHz, CDCl3) δ: -0.47 (Si-CH3), 13.9
(CH3-6 and CH3-60), 19.9 and 20.0 (CH2-5 and CH2-50), 36.5, 36.6, 37.0,
and 37.1 (CH2-2, CH2-20, CH2-4, and CH2-40), 48.3 and 48.4 (CH-3 and
CH-30), 59.9 and 60.0 (CH2-1 and CH2-10). COSY, HMQC and HMBC
data (not shown) confirmed the signal attribution. MS (70 eV), m/z (%):
410 (10) [Mþ], 310 (3), 173 (14), 157 (3), 148 (12), 129 (3), 122 (5), 115 (8),
103 (61), 91 (9), 83 (100), 75 (14), 73 (63), 55 (28). HR-MS (EI, 70 eV), m/z:
410.2183 [M]þ (calcd for [C18H42O2S2Si2]þ, 410.2165).
MS (70 eV), m/z (%): 148 (3) [Mþ], 83 (21), 78 (5), 69 (5), 56 (10),
55 (100), 43 (10), 42 (10).
RESULTS AND DISCUSSION
Identification of 3-Propyl-1,2-oxathiolane. The volatile com-
pounds from a total of 4.5 L Sauternes wine were isolated by
liquid-liquid extraction with dichloromethane. Acidic com-
pounds, as well as volatile thiols, were removed by washing the
organic extract with an alkaline mercury solution. Because the
citrus compound could not be accurately identified from this
extract by MDGC-MS-O, the extract was further purified to
avoid coelution, using C18 reverse-phase HPLC, according to the
method developed by Ferreira et al. (31) and adapted by Pineau et
al. (26). This separation afforded 25 fractions that were extracted
again and analyzed by GC-O on a BPX5 capillary column. A
citrus odorant was detected in fractions 14-15 at the target linear
retention index (1105) (14). The two fractions were pooled and
analyzed by MDGC-MS-O, using a SPB1 column in the first
oven and a BPX5 column in the second oven. Mass spectrometry
combined with olfactory detection produced a fragmentation
pattern corresponding to the odorant (Figures 2 and 3A). Repeat
MDGC-MS-O analysis, using the same fractions but changing
the column in the second oven (using either BPX70 or BPX50
columns), confirmed this result. Linear retention indices on these
two capillaries were 1592 and 1236, respectively.
cis Isomer. 1H NMR (300 MHz, CDCl3) δ: 0.99 (t, J = 7.2 Hz, 3H,
CH3-8), 1.45-1.60 (m, 2H, CH2-7), 1.66-1.95 (m, 2H, CH2-6), 2.15-2.32
(m, 1H, CH2-4), 2.90-3.04 (m, 1H, CH-3), 4.19-4.38 (m, 1H, CH2-5a),
4.73-4.88 (m, 1H, CH2-5b). 13C NMR (75 MHz, CDCl3) δ: 14.0 (CH3-8),
21.7 (CH2-7), 28.0 (CH2-4), 28.8 (CH2-6), 68.7 (CH-3), 75.1 (CH2-5).
COSY data: H-8 f H-7; H-7 f H-6, H-8; H-6 f H-3, H-7; H-4 f H-3,
H-5a, H-5b; H-3 f H-4, H-6; H-5b f H-4, H-5a; H-5a f H-4, H-5b;
HMQC data: H-8 f C-8, H-7 fC-7, H-6 f C-6, H-4 f C-4, H-3 f C-3,
H-5b f C-5, H-5a f C-5; HMBC data: H-8 f C-6, C-7; H-7 f C-3, C-6,
C-8; H-6 f C-3, C-4, C-7, C-8; H-4 f C-3, C-6; H-3 f C-4, C-6; H-5b f
C-4; H-5a f C-3, C-4.
trans Isomer. 1H NMR (300 MHz, CDCl3) δ: 0.97 (t, J = 7.2 Hz, 3H,
CH3-8), 1.32-1.45 (m, 1H, CH2-6a), 1.45-1.60 (m, 2H, CH2-7),
1.60-1.70 (m, 1H, CH2-6b), 1.78-1.95 (m, 2H, CH2-4a), 2.58-2.70 (m,
1H, CH2-4b), 3.15-3.28 (m, 1H, CH-3), 4.45-4.52 (m, 1H, CH2-5a),
4.73-4.86 (m, 1H, CH2-5b). 13C NMR (75 MHz, CDCl3) δ: 13.8 (CH3-8),
21.2 (CH2-7), 29.1 (CH2-4), 31.2 (CH2-6), 73.3 (CH-3), 74.7 (CH2-5). COSY
data: H-8 f H-7; H-6b f H-3, H-6a, H-7; H-7 f H-6a, H-6b, H-8; H-6a f
H-3, H-6b, H-7; H-4b f H-3, H-4a, H-5a, H-5b; H-4a f H-3, H-4b, H-5a,
H-5b; H-3 f H-4a, H-4b, H-6a, H-6b; H-5b f H-4a, H-4b, H-5a; H-5a f
H-4a, H-4b, H-5b; HMQC data: H-8 f C-8, H-7 fC-7, H-6b f C-6,
H-6a fC-6, H-4b fC-4, H-4a fC-4, H-3 fC-3, H-5b fC-5, H-5a fC-5.
HMBC data: H-8 f C-6, C-7; H-6b f C-3, C-4, C-7, C-8; H-7 f C-3, C-6,
C-8; H-6a f C-3, C-4, C-7, C-8; H-4b f C-3, C-6; H-4a f C-3, C-6; H-3 f
C-5, C-6; H-5b f C-3, C-4; H-5a f C-3, C-4.
Chemical ionization mode was performed to obtain the molec-
ular weight of the citrus compound (M, 132; [M þ H]þ, 133) (data
not shown). Subsequently, the exact mass measurement was used
to determine the elemental formula as C6H12OS, revealing one
degree of unsaturation in the structure. The MS pattern, presenting
only a few fragmentations, was indicative of a cyclic structure.
3,30-Disulfanediyldihexan-1-ol (6). 1H NMR (300 MHz, CDCl3) δ: 0.92
(t, J = 7.3 Hz, 6H, CH3-6, CH3-60), 1.35-1.65 (m, 8H, CH2-4, CH2-40,
CH2-5, CH2-50), 1.83-1.93 (m, 4H, CH2-2, CH2-20), 2.85 (apparent p, J =
6.4 Hz, 2H, CH-3, CH-30), 3.10 (br s, 2H, OH), 3.70-3.90 (m, 4H, CH2-1,

10610 J. Agric. Food Chem., Vol. 58, No. 19, 2010
Sarrazin et al.
Figure 4. Oxidationpathwayof3-sulfanylhexanol(1) demonstratedbythe
thermal degradation of compound 6 in the GC port and chemical oxidation
experiments.
reported to undergo chemical oxidation using sodium metaper-
iodate (32). Dependent upon the oxidizing agent amount and
reaction time, the nature and ratio of final products varied and led
to eitherthe disulfide (6) or3-propyl-γ-sultine (5) (32). Afterward,
the compound 5, the final oxidation product of 3-sulfanylhex-
anol, was identified in yellow passion fruit (28). Therefore, the
disulfide (6) was hypothesized to oxidize to the corresponding
γ-sultine (5) via a dismutation mechanism, with simultaneous
formation of 3-sulfanylhexanol (1) (28). Because the new com-
pound, 3-propyl-1,2-oxathiolane (4), was identified, a more
complete mechanism was suggested on the basis of its sulfur
oxidation state (Figure 4). This mechanism is in agreement with
the oxidation pathways already proposed for sulfur-containing
organic compounds and based on a high affinity of neighboring
hydroxyl groups for electrophilic sulfur atom (28, 33, 34). 3-Pro-
pyl-1,2-oxathiolane (4) was hypothesized to be an intermediate
oxidizing product between compounds 5 and 6, resulting from the
dismutation of the disulfide (6). To validate this hypothesis,
synthetic compound 6 and compound 5 were analyzed by
GC-MS, using either a 50 or 230 ꢀC injector. GC-MS analysis
indicated that compound 4 was detected when synthetic com-
pound 6 was injected at 230 ꢀC but not at 50 ꢀC. In addition,
compound 4 was not formed when compound 5 was injected at
either temperature. Consequently, the formation of compound 4
was demonstrated to result from the thermal degradation of the
disulfide (6). To confirm these results, synthetic compound 6 was
injected at varying GC injector temperatures and the products
obtained were analyzed. The total ion current (TIC) area percen-
tage of compound 6 was shown to decrease as the GC injector
temperature increased (Table 1). At the same time, the TIC area
percentages of compounds 1, 4, and 5 increased, confirming that
they were formed in the GC injector. The disulfide (6) undergoes
oxidative cleavage to compounds 1 and 4, in accordance with
internal Cannizzaro-like rearrangement mechanism. Because the
parts of the GC contain only oxygen-free carrier gas, the oxygen
needed for further oxidation of compound 4 to compound 5 must
already be present in trace concentrations in the sample prepara-
tion before injection. Therefore, because compound 4 was formed
via the thermal dismutation of compound 6 and oxidized quickly
to compound 5, it was revealed to be the intermediate oxidizing
product between compounds 6 and 5.
Figure 3. Massspectrum of3-propyl-1,2-oxathiolane detected inthewine:
(A) fractions 14-15 and (B) synthesized.
Moreover, the loss of m/z 43 corresponded to the loss of a propyl
radical. An initial hypothesis concerning the unusual structure of
3-propyl-1,2-oxathiolane was formulated on the basis of these
data. To our knowledge, this compound had never previously been
reported. To confirm its structure, it was synthesized by a pathway
similar to that proposed by Davis and Whitham (27) for synthesiz-
ing 1,2-oxathiolane. The synthesis consisted of three steps
(Figure 1). First, the hydroxy group of 3-sulfanylhexanol (1) was
protected via silylation, and then N-bromophtalimide derivative
(3) was formed and pyrolyzed to give three oxidation products:
3-propyl-1,2-oxathiolane (4), 3-propyl-γ-sultine (5), and disulfide
(6). Synthetic 3-propyl-1,2-oxathiolane was reminiscent of citrus
and grapefruit, in agreement with the olfactory descriptors used for
the target odorant. Its mass spectrum and linear retention indices,
identical to data obtained from wine fractions, confirmed its
identification (Figure 3B). To the best of our knowledge, this was
the first time that 3-propyl-1,2-oxathiolane was synthesized and
fully characterized (1H and 13C NMR, 2D NMR, and IR).
Unfortunately, 3-propyl-1,2-oxathiolane proved to be very
unstable and quickly oxidized to 3-propyl-γ-sultine. Further-
more, the citrus odor was detected by GC-O using a high-
temperature injection mode but not when the Sauternes wine
extract was injected at room temperature, using an on-column
injector. The origin of 3-propyl-1,2-oxathiolane thus required
further clarification.
Origin of 3-Propyl-1,2-oxathiolane. One key odorant in Sau-
ternes wine is 3-sulfanylhexanol (1) (6, 10). This compound, as
well as its oxidative dimerization product, the disulfide (6), are
also found in yellow passion fruit (30). 3-Sulfanylhexanol was

Article
J. Agric. Food Chem., Vol. 58, No. 19, 2010 10611
Table 1. Influence of the GC Injector Temperature on the Degradation of
3-Sulfanylhexan-1-ol Disulfide (6)a
GC injector temperature (ꢀC)
retention
time (min) 200
230
250
300
3-propyl-1,2-oxathiolane (4)
3-sulfanylhexanol (1)
35.3
38.7
45.8
46.6
83.9
<0.1%
0.1%
0.3%
0.3%
0.3%
0.4%
0.1%
0.1%
0.1%
0.7%
0.2%
0.1%
3-propyl-γ-sultine (isomer 1) (5)
3-propyl-γ-sultine (isomer 2) (5)
disulfide (6)
<0.1% <0.1%
0.1% 0.1%
99.7% 99.3% 99.1% 98.9%
a TIC area percentage of the products detected (analysis performed on a BPX5
capillary column).
In addition, GC-MS analysis of the mixture obtained follow-
ing thermolysis of compound 6 at 200 ꢀC in a sealed ampule for
17 h revealed the quasi-total decomposition of compound 6, with
the formation of compounds 1, 4, 5, and other degradation
products (data not shown). This oxidation pathway was further
confirmed by the reaction between compound 6 and N-bromo-
succinimide (NBS), a soft oxidative reagent (35). Treating com-
pound 6 with 1 equiv of NBS in dichloromethane at room
temperature for 1 h gave compound 5 as the principal compound
(90% yield). GC-MS analysis of the crude revealed the presence
of compounds 1 (3%) and 4 (2%), but the latter disappeared
rapidly as it oxidized to compound 5. All of these experiments
clearly confirmed that the oxidation of the disulfide (6) led to the
formation of both compound 1 and the corresponding γ-sultine
(5) via 3-propyl-1,2-oxathiolane (4), the molecule identified and
characterized in this work. To our knowledge, this was the first
time that this oxidation pathway was demonstrated experimen-
tally. Consequently, the identification of compound 4 in Sau-
ternes wines by GC-MS analysis provided indirect evidence of
the presence of the disulfide (6) and not of compound 5.
Figure 5. Partial GC chromatogram of the BSTFA-derivatized wine frac-
tion (a) supplemented or (b) not with synthetic disulfide (6). Chromato-
gram obtained with the selected ion m/z 410. The retention time of the peak
indicated by an arrow corresponds to BSTFA-derivatized disulfide (7).
Identifying 3-Sulfanylhexanol Disulfide (6). To validate this
observation, we developed a purification procedure for isolating
the disulfide (6) from Sauternes wine, using normal-phase HPLC.
First, a botrytized wine extract supplemented with synthetic
compound 6 was used to optimize the elution conditions as well
as to determine the retention time of the disulfide, which was
found in fractions 17-25. The method was then applied to a
botrytized wine extract. Fractions 17-25 were assembled, con-
centrated, and analyzed by GC-MS. Using an apolar column,
the retention time of the disulfide (6) was 83.9 min, whereas the
retention times of 3-sulfanylhexanol and 3-propyl-γ-sultine were
38.7 and 45.8-46.6 min, respectively. Compound 6 eluted at the
maximal oven temperature (250 ꢀC) and was thus poorly volatile.
This result was in agreement with the work of Werkhoff et al.,
who only described a linear retention index above 2600 on a DB-
Wax column for compound 6 (30). Because compound 1 is
detected in wine at trace concentrations (generally under 10 ppb),
the levels of the disulfide (6) were supposed to be in the same
concentration range in wine. Consequently, we derivatized com-
pound 6 to increase its volatility and improve its detection. The
Sauternes-concentrated fraction pool was derivatized using
BSTFA and analyzed immediately by GC-MS. We obtained a
signal corresponding to the MS pattern of the silylated disulfide
(7) at the same retention timeas the extract supplemented withthe
synthetic disulfide (Figure 5). Identification was then confirmed
by analyzing the same concentrated sample using a time of flight
high-resolution mass spectrometer (Figure 6). As far as we know,
this was the first time that the disulfide (6) had been identified in
wine.
Figure 6. High-resolution mass spectrum of silylated 3-sulfanylhexan-1-ol
disulfide (7) detected in the wine: (A) fractions 17-25 and (B) synthe-
sized.
tized Sauternes wines but not in dry white wines made from the
same grape varieties (14). This indicated that the disulfide (6)
was preferentially present in botrytized wines. To validate this
observation, a dry white Bordeaux wine was extracted, deriva-
tized using BSTFA, and then analyzed by GC-MS TOF.
A signal corresponding to the silylated disulfide (7) was obtained,
but its intensity was near the detection limit. This proved the
formation of the disulfide (6) during wine aging thanks to the
unique composition of the botrytized wine. Various studies have
demonstrated a decrease in volatile thiol levels in the presence of
oxygen during wine aging and after bottling (36-38). Indeed,
volatile thiols have been established to react with polymeric
phenolic compounds in wine in the presence of oxygen, via
nucleophilic-acid-catalyzed substitutions. More recently, com-
pound 1 and other wine volatile thiols have also been shown
As previously reported, the odoriferous zone corresponding to
3-propyl-1,2-oxathiolane (4) was detected by GC-O in botry-

10612 J. Agric. Food Chem., Vol. 58, No. 19, 2010
Sarrazin et al.
to react readily with two main flavonoids [(-)-epicatechin and
(þ)-catechin] via aniron-catalyzedmechanism (39). However, the
broad enzymatic pool of B. cinerea induces significant changes in
the chemical composition of the grapes used in botrytized
wines (11). Among the fungal activities of B. cinerea, the
laccase (benzendiol oxygen oxidoreductase) oxidizes a wide
range of grape phenolic substrates, leading to the formation of
quinones, followed by their polymerization and precipitation
(40-42). The phenolic composition of botrytized grapes and
wines is thus deeply modified. This may contribute to the
explanation of the singular reactivity of compound 1 in
botrytized wines in comparison to dry white wines, thus
bringing new insights into the evolution of the wine key aroma
compound 1.
(7) Bouchilloux, P.; Darriet, P.; Henry, R.; Lavigne-Cruege, V.;
Dubourdieu, D. Identification of volatile and powerful odorous
thiols in Bordeaux red wine varieties. J. Agric. Food Chem. 1998, 46,
3095-3099.
(8) Murat, M. L.; Tominaga, T.; Dubourdieu, D. Mise en e
´
vidence de
^
compose
´
s cle
´
fs dans l’arome des vins rose
´
s et clairets de Bordeaux.
J. Int. Sci. Vigne Vin 2001, 35, 99-105.
(9) Fretz, C. B.; Luisier, J. L.; Tominaga, T.; Amado, R. 3-Mercapto-
hexanol: An aroma impact compound of Petite Arvine wine. Am. J.
Enol. Vitic. 2005, 56, 407-410.
(10) Sarrazin, E.; Shinkaruk, S.; Tominaga, T.; Bennetau, B.; Frerot, E.;
Dubourdieu, D. Odorous impact of volatile thiols on the aroma
of young botrytized sweet wines: Identification and quantification
of new sulfanyl alcohols. J. Agric. Food Chem. 2007, 55, 1437-
1444.
^
(11) Ribereau-Gayon, P.; Dubourdieu, D.; Doneche, B.; Lonvaud, A.
´
1,3-Sulfanylalcohols, especially compound 1, play a major role
in the aroma of various food products. It is thus challenging to
extend our knowledge of their reactivity and stability in natural
matrices. Our work focused on Sauternes wine aroma and
identified a new citrus odorant, compound 4, detected by GC-O
in Sauternes wine extracts. This unusual compound was shown to
result from the thermal oxidative degradation of the disulfide (6)
in the GC injector, which was then firmly identified in Sauternes
wine extracts. Although the presence of 3-sulfanylhexanol oxida-
tion products had previously been reported in natural extracts
(but not wine), the full oxidation pathway from compounds 1 to 5
via a disulfide intermediate (6) was clearly established in this
research for the first time. This finding may be generalized to
other 1,3-sulfanylalcohols, thus bringing new insights into the
sensory stability of natural substances. In particular, because
compound 1 has a high olfactory impact in various wines, further
work is now required to investigate the presence of the disulfide
(6) in other botrytized and non-botrytized wines, as well as to
explore the presence of the sultine (5) in Sauternes botrytized
wines.
Botrytized sweet wines (Sauternes and Tokay). In Handbook of
Enology; Wiley: New York, 1999; Vol. 1: The Microbiology of Wine
and Vinifications, pp 410-419.
(12) Peynaud, E. Le vin et les jours; Dunod: Paris, France, 1985.
(13) Bailly, S.; Jerkovic, V.; Meuree, A.; Timmermans, A.; Collin, S. Fate
of key odorants in Sauternes wines through aging. J. Agric. Food
Chem. 2009, 57, 8557-8563.
(14) Sarrazin, E.; Dubourdieu, D.; Darriet, P. Characterization of key-
aroma compounds of botrytized wines, influence of grape botrytiza-
tion. Food Chem. 2007, 103, 536-545.
(15) Thibon, C.; Shinkaruk, S.; Tominaga, T.; Bennetau, B.; Dubourdieu,
D. Analysis of the diastereoisomers of the cysteinylated aroma pre-
cursor of 3-sulfanylhexanol in Vitis vinifera grape must by gas chro-
matography coupled with ion trap tandem mass spectrometry.
J. Chromatogr., A 2008, 1183, 150-157.
(16) Thibon, C.; Dubourdieu, D.; Darriet, P.; Tominaga, T. Impact of
noble rot on the aroma precursor of 3-sulfanylhexanol content in
Vitis vinifera L. cv Sauvignon blanc and Semillon grape juice. Food
Chem. 2009, 114, 1359-1364.
(17) Bailly, S.; Jerkovic, V.; Marchand-Brynaert, J.; Collin, S. Aroma
extraction dilution analysis of Sauternes wines. Key role of poly-
functional thiols. J. Agric. Food Chem. 2006, 54, 7227-7234.
(18) Campo, E.; Cacho, J.; Ferreira, V. The chemical characterization of
the aroma of dessert and sparkling white wines (Pedro Ximenez,
Fino, Sauternes, and Cava) by gas chromatography-olfactometry
and chemical quantitative analysis. J. Agric. Food Chem. 2008, 56,
2477-2484.
(19) Tominaga, T.; Murat, M. L.; Dubourdieu, D. Development of a
method for analyzing the volatile thiols involved in the characteristic
aroma of wines made from Vitis vinifera L. cv. Sauvignon blanc.
J. Agric. Food Chem. 1998, 46, 1044-1048.
(20) Ferreira, V.; Jarauta, I.; Ortega, L.; Cacho, J. Simple strategy for the
optimization of solid-phase extraction procedures through the use of
solid-liquid distribution coefficients application to the determina-
tion of aliphatic lactones in wine. J. Chromatogr., A 2004, 1025,
147-156.
(21) Allen, M. S.; Lacey, M. J.; Boyd, S. Determination of methoxy-
pyrazines in red wines by stable isotope dilution gas chromato-
graphy mass spectrometry. J. Agric. Food Chem. 1994, 42, 1734-
1738.
ACKNOWLEDGMENT
This paper is dedicated to the memory of Dr. Takatoshi
Tominaga, who suddenlypassed away on June 8, 2008. Takatoshi
devoted his life to understand the mysteries of wine aroma. His
passion, knowledge, and friendship are deeply missed.
´
We thank Dr. Estelle Delort and Dr. Eric Frerot from
Firmenich SA (Geneva, Switzerland) for their kind help in the
GCT Premier analysis, as well as Dr. Sandrine Garbay from
^
the Chateau d’Yquem (Sauternes, France) and Dr. Jean-Jacques
Filippi (Universite de Nice-Sophia Antipolis, France) for fruitful
discussion.
´
LITERATURE CITED
(1) Vermeulen, C.; Gijs, L.; Collin, S. Sensorial contribution and
formation pathways of thiols in foods: A review. Food Rev. Int.
2005, 21, 69-137.
(2) Goeke, A. Sulfur-containing odorants in fragrance chemistry. Sulfur
(22) Darriet, P.; Pons, M.; Henry, R.; Dumont, O.; Findeling, V.;
Cartolaro, P.; Calonnec, A.; Dubourdieu, D. Impact odorants
contributing to the fungus type aroma from grape berries contami-
nated by powdery mildew (Uncinula necator); incidence of enzymatic
activities of the yeast Saccharomyces cerevisiae. J. Agric. Food Chem.
2002, 50, 3277-3282.
(23) Pons, A.; Lavigne, V.; Eric, F.; Darriet, P.; Dubourdieu, D.
Identification of volatile compounds responsible for prune aroma
in prematurely aged red wines. J. Agric. Food Chem. 2008, 56,
5285-5290.
(24) Sakuma, S.; Amano, H.; Ohkochi, M. Identification of off-flavor
compounds in beer. J. Am. Soc. Brew. Chem. 2000, 58, 26-29.
(25) Wanikawa, A.; Hosoi, K.; Kato, T.; Nakagawa, K. Identification of
green note compounds in malt whisky using multidimensional gas
chromatography. Flavour Frag. J. 2002, 17, 207-211.
Rep. 2002, 23, 243.
(3) Starkenmann, C.; Troccaz, M.; Howell, K. The role of cysteine and
cysteine-S conjugates as odour precursors in the flavour and
fragrance industry. Flavour Frag. J. 2008, 23, 369-381.
(4) Engel, K. H.; Tressl, R. Identification of new sulfur-containing
volatiles in yellow passionfruit (Passiflora edulis f. flavicarpa).
J. Agric. Food Chem. 1991, 39, 2249-2252.
(5) Tominaga, T.; Furrer, A.; Henry, R.; Dubourdieu, D. Identification
of new volatile thiols in the aroma of Vitis vinifera L. var. Sauvignon
blanc wines. Flavour Frag. J. 1998, 13, 159-162.
(6) Tominaga, T.; Baltenweck-Guyot, R.; Peyrot des Gachons, C.;
Dubourdieu, D. Contribution of volatile thiols to the aromas of
white wines made from several Vitis vinifera grape varieties. Am. J.
Enol. Vitic. 2000, 51, 178-181.

Article
J. Agric. Food Chem., Vol. 58, No. 19, 2010 10613
(26) Pineau, B.; Barbe, J. C.; Van Leeuwen, C.; Dubourdieu, D. Examples of
perceptive interactions involved in specific “red-” and “black-berry”
aromas in red wines. J. Agric. Food Chem. 2009, 57, 3702-3708.
(27) Davis, A. P.; Whitham, G. H. 1,2-Oxathiolan, a simple sultene.
J. Chem. Soc., Chem. Commun. 1981, 741-742.
(36) Brajkovich, M.; Tibbits, N.; Peron, G.; Lund, C. M.; Dykes, S. I.;
Kilmartin, P. A.; Nicolau, L. Effect of screwcap and cork closures on
SO2 levels and aromas in a Sauvignon Blanc wine. J. Agric. Food
Chem. 2005, 53, 10006-10011.
(37) Blanchard, L.; Darriet, P.; Dubourdieu, D. Reactivity of 3-mercap-
tohexanol in red wine: Impact of oxygen, phenolic fractions, and
sulfur dioxide. Am. J. Enol. Vitic. 2004, 55, 115-120.
(28) Yolka, S.; Dunach, E.; Loiseau, M.; Lizzani-Cuvelier, L.; Fellous,
R.; Rochard, S.; Schippa, C.; George, G. γ-Sultines: A new class of
flavour compounds. Flavour Frag. J. 2002, 17, 425-431.
(38) Lopes, P.; Silva, M. A.; Pons, A.; Tominaga, T.; Lavigne, V.;
Saucier, C.; Darriet, P.; Teissedre, P. L.; Dubourdieu, D. Impact
of oxygen dissolved at bottling and transmitted through closures on
the composition and sensory properties of a Sauvignon Blanc wine
during bottle storage. J. Agric. Food Chem. 2009, 57, 10261-10270.
(39) Nikolantonaki, M.; Chichuc, I.; Teissedre, P. L.; Darriet, P. Reactivity
of volatile thiols with polyphenols in a wine-model medium: Impact of
oxygen, iron, and sulfur dioxide. Anal. Chim. Acta 2010, 660, 102-109.
(40) Dubourdieu, D. La vinification des vins liquoreux de pourriture
noble. Rev. Fr. Oenol. 1999, 176, 32-35.
^
(29) Coulomb, J.; Certal, V.; Fensterbank, L.; Lacote, E.; Malacria, M.
Formation of cyclic sulfinates and sulfinamides through homolytic
substitution at the sulfur atom. Angew. Chem., Int. Ed. 2006, 45, 633-637.
(30) Werkhoff, P.; Guntert, M.; Krammer, G.; Sommer, H.; Kaulen,
J. Vacuum headspace method in aroma research: Flavor chemistry
of yellow passion fruits. J. Agric. Food Chem. 1998, 46, 1076-1093.
(31) Ferreira, V.; Hernandez-Orte, P.; Escudero, A.; Lopez, R.; Cacho, J.
Semipreparative reversed-phase liquid chromatographic fractiona-
tion of aroma extracts from wine and other alcoholic beverages.
J. Chromatogr., A 1999, 864, 77-88.
€
(41) Salgues, M.; Cheynier, V.; Gunata, Z.; Wylde, R. Oxidation of grape
(32) Yolka, S.; Fellous, R.; Lizzani-Cuvelier, L.; Loiseau, M. Oxidation
of 3-sulfanyl-alcohols with sodium metaperiodate: New synthesis of
sultines. Tetrahedron Lett. 1998, 39, 991-992.
(33) Givens, E. N.; Hamilton, L. A. A cyclic sulfinate ester. Preparation
and reactions of 1,2-benzoxathian-2-oxide. J. Org. Chem. 1967, 32,
2857-2861.
juice 2-S-glutathionyl caffeol tartaric acid by Botris cinerea laccase
and characterization of a new substance 2,5-di-S-glutathionyl caf-
feoyl tartatic acid. J. Food Sci. 1986, 55, 1191-1194.
(42) Dubernet, M.; Ribereau-Gayon, P.; Lerner, H. R.; Harel, E.; Mayer,
´
M. A. Purification and properties of laccase from Botrytis cinerea.
Phytochemistry 1977, 16, 191-193.
(34) Doi, J. T.; Luehr, G. W.; Musker, W. K. Oxidative cleavage and
cyclization of disulfide carboxylic-acids and alcohols by aqueous
iodine;A facile route to 5-membered ring sultines. J. Org. Chem.
1985, 50, 5716-5719.
Received for review May 27, 2010. Revised manuscript received
September 7, 2010. Accepted September 8, 2010. This work was
funded by the CIVB (Bordeaux Wine Council) and the Conseil
(35) Harville, R.; Reed, S. F. Sulfide oxidation with N-halosuccinimides.
´
Regional d’Aquitaine.
J. Org. Chem. 1968, 33, 3976-3977.