5396 J. Agric. Food Chem., Vol. 50, No. 19, 2002
Engel and Schieberle
described for 2-acetyl-2-thiazoline (18). The calibration factors (CF)
were determined in mixtures of equal amounts of unlabeled odorants
and the corresponding labeled standards in ratios of 3:1 to 1:3 (by
weight) by means of mass chromatography and were calculated using
the equation CF ) (C
compound, C ) concentration of the labeled compound, I
of the ion of the labeled compound, and I ) intensity of the ion of the
u L
I /C
L
I
u
), with C
u
) concentration of the unlabeled
L
L
) intensity
u
unlabeled compound. For mass chromatography, the capillary column
was coupled to an ITD-800 ion trap detector (Finnigan, Bremen,
Germany), running in the chemical ionization mode with methanol as
reactant gas. Mass spectra were generated at 70 eV.
RESULTS AND DISCUSSION
Odorants Formed under Aqueous Reaction Conditions.
In a first experiment, fructose was reacted with cysteamine under
aqueous conditions (model A). The flavor extract prepared by
solvent extraction followed by a SAFE distillation elicited a
very intense roasty, popcorn-like aroma. By application of the
ADA on a 100:1 concentration (100 mL of reaction mixture f
Figure 4. Mass spectrum (MS/EI) of the deuterium-labeled 1,3-thiazolidine.
[2
4
H ]-2-Acetyl-2-thiazolidine was synthesized as previously reported
(18).
1
mL of extract), 10 odor-active areas were detected, among
Model Reaction. To simulate cooking conditions, either cysteamine
which compound 19 (Table 1) with an intense popcorn-like
aroma exhibited by far the highest flavor dilution (FD) factor.
On the basis of the data of the reference compound, 19 was
identified as 5-acetyl-3,4-dihydro-2H-1,4-thiazine, previously
identified by us for the first time in a processed glucose/cysteine
solution (3). Two additional odorants with high odor intensities
were identified as N-(2-mercapotethyl)-1,3-thiazolidine (16) and
the well-known caramel-like carbohydrate degradation product
(
3.3 mmol; model A) or 1,3-thiazolidine-2-carboxylic acid (3.3 mmol;
model B) was dissolved in phosphate buffer (100 mL, 0.1 mol/L, pH
.0) and reacted with fructose (10 mmol) in a laboratory autoclave
type II, 200 mL total volume; Roth, Karlsruhe, Germany) by raising
7
(
the temperature within 20 min from 20 to 145 °C.
To mimic dry-heating conditions, either cysteamine (3.3 mmol;
model C) or 1,3-thiazolidine-2-carboxylic acid (3.3 mmol; model D)
and fructose (10 mmol) were dissolved in a low volume of water (1.2
mL, pH 7.0) and mixed with silica gel (19.1 g). The mixture was
transferred into eight glass vials (1 cm i.d., total volume ) 10 mL),
and the material was heated for 10 min at 150 °C in a metal block. In
parallel, the reaction was also performed using wheat starch as the
matrix.
4
-hydroxy-2,5-dimethyl-3(2H)-furanone (15). Evidence for the
thiazolidine, which has not been reported prior to this work, is
given separately (15). Somewhat lower FD factors were shown
by 2-acetyl-2-thiazoline (11), 2-furfurylthiol (4), and ethane-
1
,2-dithiol (3). Compared to the previously studied glucose/
Isolation of the Volatiles and Aroma Extract Dilution Analysis
(
AEDA). The volatile reaction products formed were isolated by
cysteine mixture (3), in particular, ethanedithiol (3), N-mercap-
toethylpyrrole (10), and N-(2-mercaptoethyl)-1,3-thiazolidine
(16) were characterized as additional aroma contributors to the
fructose/cysteamine model studied here.
Isothiaproline can be regarded as a condensation product of
cysteamine and glyoxylic acid and is known to be formed from
both intermediates in synthetic experiments. Reacting this
extraction with diethyl ether and distillation under high vacuum as
recently described (19). The distillates were concentrated at 40 °C by
means of a Vigreux column (60 × 1 cm i.d.) and a microdistillation
apparatus to exactly 1 mL. The odor-active compounds were evaluated
by AEDA as previously described (2, 3).
High-Resolution Gas Chromatography (HRGC))Mass Spec-
trometry (MS). HRGC was performed with a type 5160 gas chro-
matograph (Fisons Instruments, Mainz, Germany) using the following
capillaries: (30 m × 0.32 mm fused silica capillary, free fatty acid
phase, 0.25 µm; J&W Scientific, Fisons Instruments) ) FFAP; and
“synthetic” amino acid with fructose (model B) led to an extract
with a roasty, pretzel-like aroma. By applying AEDA, 12 odor-
active areas were detected (Table 1; model B). The results of
the identification experiments in combination with the FD
factors revealed 2-acetyl-2-thiazoline (11) and 5-acetyl-3,4-
dihydro-2H-1,4-thiazine (19) as the most important aroma
compounds in this mixture. Lower FD factors were found for
(30 m × 0.32 mm fused silica capillary DB-5, 0.25 µm; J&W Scientific,
Fisons Instruments) ) DB 5. The samples were applied by the cold
on-column injection technique at 40 °C. After 2 min, the temperature
of the oven was raised at 40 °C/min to 60 °C, held for 1 min
isothermally, then raised at 6 °C/min to 230 °C, and held for 15 min.
The flow of the carrier gas helium was 1.5 mL/min. The FID was held
at 220 °C. Linear retention indices (LRI) of the compounds were
calculated from the retention times of n-alkanes. MS analysis was
performed by means of an MAT 95S (Finnigan, Bremen, Germany) in
tandem with the capillaries described above. Mass spectra in the electron
impact mode (MS/EI) were generated at 70 eV and in the chemical
ionization mode (MS/CI) at 115 eV with isobutane as reactant gas.
For high-resolution MS (HRMS) the masses measured were corrected
using perfluorokerosene as the reference.
4
-hydroxy-2,5-dimethyl-3(2H)-furanone (15), 2-propionyl-2-
thiazoline (12), and N-(2-mercaptoethyl)-1,3-thiazolidine (16).
A comparison of these results to those obtained for cysteamine
(model A) clearly showed a reduction in 19, whereas 11 was
enhanced.
Assuming that isothiaproline is cleaved during the thermal
reaction as postulated in Figure 5 and that the reaction pathways
generating the same odorants in both systems are identical, the
slow release of cysteamine from isothiaproline, which in turn
might react with 2-oxopropanal to yield 11, should favor the
formation of this thiazoline, which is known to be unstable at
higher temperatures (18).
Odorants Formed under Roasting Conditions. To simulate
roasting conditions, both reactions were repeated, but water was
replaced by silica gel/water (9 + 1 w/w) as the matrix.
Furthermore, a higher temperature was immediately applied.
Although silica is known to catalyze certain reactions, prelimi-
nary experiments revealed the same odorants when the reaction
Stable Isotope Dilution Assays/Quantification. After cooling,
aliquots of the respective reaction mixture were spiked with known
amounts of the three labeled internal standards [ H
dihydro-2H-1,4-thiazine, [ H
and [ H
depending on the amounts of the analytes present). The samples were
extracted with diethyl ether (five times; total volume ) 120 mL), the
volatiles were isolated by a SAFE distillation (19), and the resulting
distillate was separated by HRGC. The ion intensities of the respective
molecular ions obtained by MS/CI were monitored as previously
2
4
]-5-acetyl-3,4-
2
4
]-N-(2-mercaptoethyl)-1,4-thiazolidine,
2
4
]-2-acetyl-2-thiazoline (5-10 µg, dissolved in ethyl ether