New insights into the formation of aroma-active strecker aldehydes from 3-oxazolines as transient intermediates

Article abstract of DOI:10.1021/jf301489j

2-Substituted-5-methyl-3-oxazolines, a novel class of aroma precursors that are able to release the respective Strecker aldehydes by hydrolysis, were identified. Hydrolysis can take place after the addition of water or with human saliva during mastication, respectively. 2-Isobutyl-, 2-sec-isobutyl-, 2-isopropyl, and 2-benzyl-5-methyl-3-oxazolines were synthesized and structurally identified by means of gas chromatography-mass spectrometry (GC-MS) in the electron impact mode and in the chemical ionization mode as well as by one- and two-dimensional NMR experiments. With these compounds at hand, a variety of stability experiments were performed using headspace-GC-MS or proton transfer reaction-MS techniques on the basis of stable isotope dilution assays, proving the ability to release the respective Strecker aldehydes was dependent on the pH value as well as on the hydrolysis time. After the addition of water at 37 °C, for example, >70 mol % of 3-methylbutanal or >40 mol % of phenylacetaldehyde was liberated from a solution of 2-isobutyl-5-methyl-3- oxazoline or 2-benzyl-5-methyl-3-oxazoline, respectively, after 5 min. Furthermore, the presence of 2-isobutyl-5-methyl-3-oxazoline in dark chocolate containing 70% cocoa was proven by GC-MS.

Full text of DOI:10.1021/jf301489j

Article
pubs.acs.org/JAFC
New Insights into the Formation of Aroma-Active Strecker Aldehydes
from 3-Oxazolines as Transient Intermediates
Michael Granvogl,† Ersan Beksan,‡ and Peter Schieberle*,†,‡
†Lehrstuhl fur Lebensmittelchemie, Technische Universitat Munchen, and
̈
̈
̈
‡Deutsche Forschungsanstalt fur Lebensmittelchemie, Lise-Meitner-Straße 34, D-85354 Freising, Germany
̈
S
* Supporting Information
ABSTRACT: 2-Substituted-5-methyl-3-oxazolines, a novel class of aroma precursors that are able to release the respective
Strecker aldehydes by hydrolysis, were identified. Hydrolysis can take place after the addition of water or with human saliva
during mastication, respectively. 2-Isobutyl-, 2-sec-isobutyl-, 2-isopropyl, and 2-benzyl-5-methyl-3-oxazolines were synthesized
and structurally identified by means of gas chromatography−mass spectrometry (GC-MS) in the electron impact mode and in
the chemical ionization mode as well as by one- and two-dimensional NMR experiments. With these compounds at hand, a
variety of stability experiments were performed using headspace-GC-MS or proton transfer reaction−MS techniques on the basis
of stable isotope dilution assays, proving the ability to release the respective Strecker aldehydes was dependent on the pH value
as well as on the hydrolysis time. After the addition of water at 37 °C, for example, >70 mol % of 3-methylbutanal or >40 mol %
of phenylacetaldehyde was liberated from a solution of 2-isobutyl-5-methyl-3-oxazoline or 2-benzyl-5-methyl-3-oxazoline,
respectively, after 5 min. Furthermore, the presence of 2-isobutyl-5-methyl-3-oxazoline in dark chocolate containing 70% cocoa
was proven by GC-MS.
KEYWORDS: Strecker aldehydes, aroma precursors, aroma release, 5-methyl-3-oxazolines, PTR-MS
In a previous study on the isolation of aroma compounds
from milk chocolate,10 it was observed that in comparison to a
distillation performed by solvent-assisted flavor evaporation
(SAFE) technique11 phenylacetaldehyde and 3-methylbutanal
were increased by factors of ∼120 and ∼13, respectively, when
steam distillation was applied in volatile isolation. Although a
formation of both Strecker aldehydes from the amino acid/
carbohydrate reaction might have occurred during the thermal
treatment, these data suggested for the first time the presence
of unknown precursors for Strecker aldehydes showing a
distinct instability in the presence of hot water. This
assumption was later corroborated by the results of Schuh
and Schieberle,13 who reported on the generation of Strecker
aldehydes simply by the addition of hot water to black tea
leaves. Furthermore, in a recent publication,14 the significant
generation of Strecker aldehydes from dry processed foods,
such as malt, bread crust, crackers, and, in particular, cocoa and
chocolate, upon heating in water was confirmed by quantitative
studies.
INTRODUCTION
■
Nearly 150 years ago, Strecker1 reported on the formation of 3-
methylbutanal and carbon dioxide when leucine was reacted
with the tricarbonyl compound alloxan. However, the
important impact of such aldehydes, later designated “Strecker”
aldehydes,2 was not reported until Keeney and Day3 were able
to generate a toasted cheese aroma by reacting a protein
hydrolysate with ninhydrine. Shortly after this finding, the
thermally induced formation of methylpropanal and 3-
methylbutanal was also suggested as an important step in
cocoa flavor formation during roasting.4
Today, it is well accepted that the formation of Strecker
aldehydes from their parent amino acids is initiated by various
α-dicarbonyl compounds formed by carbohydrate degradation,
such as 2,3-butanedione, 2-oxopropanal, or deoxyosones as well
as quinones enzymatically formed by polyphenol oxidation.5
The reaction pathway is exemplified in Figure 1 for the
formation of phenylacetaldehyde (E) from phenylalanine,
starting from intermediate A. Besides phenylalanine, also
valine, leucine, isoleucine, methionine, and alanine are known
to serve as precursors for odor-active aldehydes following the
same pathway (Table 1). However, aldehydes are not the only
compounds formed by a Strecker degradation of α-amino acids.
As previously shown,7,8 also phenylacetic acid (F from
intermediate D; Figure 1) and 2-phenylethylamine (B from
intermediate A; Figure 1), showing the partial structure of the
parent amino acid phenylalanine, were confirmed to be formed
via a Strecker degradation. In addition, besides the free amino
acids, also the Amadori reaction products have been confirmed
as direct precursors of Strecker aldehydes without the need of
an α-dicarbonyl compound as catalyst of the reaction.9,10
More than 40 years ago,15 it was observed that, as expected,
the reaction of valine with 2,3-butanedione led to the formation
of methylpropanal. However, probably because the reaction was
performed in diglyme in the absence of water, 2-isopropyl-4,5-
dimethyl-3-oxazoline was shown to be additionally formed, but
because no synthesis of the latter compound was performed,
the role of the oxazoline as a possible intermediate in aldehyde
formation was not elucidated. The aim of this study was,
Received: April 8, 2012
Revised: June 1, 2012
Accepted: June 2, 2012
Published: June 3, 2012
© 2012 American Chemical Society
6312
dx.doi.org/10.1021/jf301489j | J. Agric. Food Chem. 2012, 60, 6312−6322

Journal of Agricultural and Food Chemistry
Article
Synthesis of 2-Substituted-5-methyl-3-oxazolines. The gen-
eral reaction pathway, which was developed as a retrosynthetic
approach, involved the formation of imines from Strecker aldehydes
and 1-amino-2-propanol, which immediately cyclized into the 5-
methyl-3-oxazolidines. These were finally oxidized by Dess−Martin
periodinane to yield the respective 5-methyl-3-oxazolines (Figure 2A−
D), which are identical with a possible cyclic tautomer of intermediate
A in the Strecker reaction (Figure 1).
General Reaction Conditions. (R)-1-Amino-2-propanol (20 mmol)
was dissolved in dichloromethane (80 mL; dried over anhydrous
sodium sulfate) and was singly reacted with either 3-methylbutanal, 2-
methylbutanal, methylpropanal, or phenylacetaldehyde (each 20
mmoL; phenylacetaldehyde freshly prepared by distillation) (Figure
2A−D). The reaction mixture was stirred at room temperature for 30
min (phenylacetaldehyde) or for 12 h (for all other aldehydes) to
obtain the respective 5-methyl-3-oxazolidine in yields of ∼15%
(phenylacetaldehyde) and ∼50% (all other aldehydes), respectively.
The diastereoisomers of the 3-oxazolidines were separated neither for
further synthesis nor for NMR experiments. Aliquots of the
oxazolidines (2.5 mmol) were then dissolved in dichloromethane
(60 mL), and Dess−Martin periodinane (3.5 mmol) was added. The
oxidation into the corresponding 5-methyl-3-oxazolines was finished
within 30 min. After the addition of pentane (25 mL), the mixture was
filtered and submitted to high-vacuum distillation at ∼50 °C.11 The
distillate was evaporated to obtain ∼1 mL of an oily residue. After the
addition of a pentane/diethyl ether mixture (3 mL; 70:30 by vol), the
respective target compound was isolated by column chromatography
(22 × 2 cm) as detailed below using a diol phase (40 μm particle size;
Bakerbond, Mallinckrodt Baker, Griesheim, Germany) suspended in
pentane.
Analytical Data for 2-Substituted-5-methyl-3-oxazolidines
and 2-Substituted-5-methyl-3-oxazolines. 2-Isobutyl-5-methyl-
3-oxazolidine (1; Figure 2A). MS-EI, m/z (%) 86 (100), 42 (80), 56
(69), 41 (31), 58 (31), 84 (19), 44 (18), 43 (15), 98 (15), 101 (14),
128 (10), 45 (9), 57 (9), 39 (7), 59 (7), 40 (6), 55 (6), 87 (6), 54 (5),
85 (5), 100 (5), 99 (4), 142 (4), 82 (3), 142 ([M − H]+; 3), 140 (M+;
1
tr); MS-CI, m/z (%) 144 ([M + H]+; 100), 145 (9); H NMR (400
Figure 1. Formation pathway leading to the Strecker aldehyde (E), the
Strecker acid (F), and the Strecker amine (B) as exemplified for the
reaction of phenylalanine with an α-dicarbonyl compound.
MHz; CD2Cl2) δ 0.91−0.95 [m, 12 H, H−C(9, 9′, 10, 10′)], 1.14 [d, J
= 6.1 Hz, 3H, H−C(6 or 6′)], 1.15 [d, J = 6.1 Hz, 3H, H−C(6′ or 6)],
1.37−1.53 [m, 4H, H−C(7, 7′)], 1.73−1.82 [m, 2H, H−C(8, 8′)],
2.47 [dd, J = 11.9, 7.5 Hz, 1H, H−C(4a′)], 2.66 [dd, J = 11.5, 4.9 Hz,
1H, H−C(4a)], 3.05 [dd, J = 11.6, 7.2 Hz, 1H, H−C(4b)], 3.31 [dd, J
= 11.9, 6.2 Hz, 1H, H−C(4b′)], 3.87−3.95 [m, 2H, H−C(5, 5′)], 4.34
[t, J = 5.9 Hz, 1H, H−C(2)], 4.46 [t, J = 5.9 Hz, 1H, H−C(2′)]; 13C
NMR (100 MHz, CD2Cl2) δ 20.21 [C(6 or 6′)], 21.99 [C(6′ or 6)],
22.91 [C(9, 9′ or 10, 10′)], 23.19 [C(10, 10′ or 9, 9′)], 25.84 [C(8,
8′)], 44.08 [C(7 or 7′], 53.21 [C(4)], 54.53 [C(4′)], 72.30 [C(5 or
5′)], 72.43 [C(5′ or 5)], 90.56 [C(2′)], 91.69 [C(2)].
Purification of 2-Isobutyl-5-methyl-3-oxazoline (2; Figure 2A).
Stepwise elution was performed as follows: pentane (100 mL; fraction
A), pentane/diethyl ether (50 mL; 95:5 by vol; fraction B), and
pentane/diethyl ether (200 mL; 95:5 by vol; fraction C). Because two
peaks were detected in fraction C (Figure 3), these were separated by
rechromatography using the same diol phase: pentane (100 mL;
fraction A), pentane/diethyl ether (50 mL; 95:5 by vol; fraction B),
pentane/diethyl ether (50 mL; 95:5 by vol; fraction C); pentane/
therefore, to get an insight into the possible role of such
oxazolines as precursors of Strecker aldehydes, able to liberate
the aroma compounds in the presence of water.
MATERIALS AND METHODS
■
Food Samples. Chocolate was purchased at a local supermarket.
Chemicals. (R)-1-Amino-2-propanol, (S)-1-amino-2-propanol,
methylpropanal, 2-methylbutanal, 3-methylbutanal, phenylacetalde-
hyde, 2-methyl-2-propen-1-ol, 2-methyl-3-buten-1-ol, 3-methyl-3-
buten-1-ol, [13C2]-phenylethanol, Dess−Martin periodinane, and
tris(triphenylphosphine) rhodium(I) chloride were from Sigma-
Aldrich (Taufkirchen, Germany). Deuterium gas was from Messer
Griesheim (Krefeld, Germany). All other chemicals were of analytical
grade.
Table 1. Six of the 20 Proteinogenic Amino Acids Acting as Precursors in the Formation of Odor-Active Strecker Aldehydes
amino acid
Strecker aldehyde
odor quality
odor threshold (μg/kg of water)
a
methionine
leucine
→ 3-(methylthio)propanal
→ 3-methylbutanal
→ methylpropanal
→ 2-methylbutanal
→ phenylacetaldehyde
→ acetaldehyde
cooked potato-like
malty
0.4
a
0.5
a
valine
malty
0.5
a
isoleucine
phenylalanine
alanine
malty
1.5
b
honey-like
fresh, green
5.2
a
25
a
b
According to ref 6. Threshold was determined following the procedure described in ref 6.
6313
dx.doi.org/10.1021/jf301489j | J. Agric. Food Chem. 2012, 60, 6312−6322

Journal of Agricultural and Food Chemistry
Article
Figure 2. Synthetic routes used in the preparation of 2-substituted-5-methyl-3-oxazolines: (A) 2-isobutyl- (2); (B) 2-sec-isobutyl- (4); (C) 2-
isopropyl- (6); (D) 2-benzyl-5-methyl-3-oxazoline (8); (E) 2-isobutyl-3-oxazoline (10).
absolute configuration was not determined] was present in fraction C,
a mixture of isomers 2a and 2b was found in fraction D, and isomer 2b
[(R,S) or (R,R)] was isolated in fraction E. MS-EI, m/z (%) 84 (100),
57 (26), 54 (15), 41 (14), 56 (11), 85 (11), 70 (10), 82 (9), 71 (8), 99
(8), 39 (7), 83 (7), 97 (6), 42 (5), 55 (5), 126 (4), 44 (3), 53 (3), 140
([M − H]+; 1), 141 (M+; tr); MS-CI, m/z (%) 142 ([M + H]+; 100),
143 (8); 1H NMR (400 MHz; CD2Cl2) δ 0.97 [d, J = 6.7 Hz, 6H, H−
C(9, 10)], 1.30 [d, J = 6.7 Hz, 3H, H−C(6)], 1.42−1.48 [m, 1H, H−
C(7a)], 1.57−1.63 [m, 1H, H−C(7b)], 1.83−1.92 [m, 1H, H−C(8)],
4.69−4.74 [m, 1H, H−C(5)], 5.55−5.59 [m, 1H, H−C(2)], 7.36 [d, J
= 2.6 Hz, 1H, H−C(4)]; 13C NMR (100 MHz, CD2Cl2) δ 19.75
[C(6)], 22.78 [C(9)], 23.30 [C(10)], 25.27 [C(8)], 46.80 [C(7)],
82.07 [C(5)], 105.61 [C(2)], 163.94 [C(4)].
2-sec-Butyl-5-methyl-3-oxazolidine (3; Figure 2B; Mixture of Four
Diastereoisomers Due to Three Chiral Carbon Atoms). MS-EI, m/z
(%) 86 (100), 41 (55), 70 (49), 69 (44), 43 (42), 98 (41), 58 (30), 42
(28), 44 (25), 84 (23), 56 (21), 39 (20), 59 (19), 115 (18), 55 (11),
57 (10), 68 (10), 53 (8), 128 (8), 67 (7), 82 (7), 99 (7), 54 (6), 78
(6), 97 (6), 100 (4), 142 ([M − H]+; 3), 143 (M+; tr); MS-CI, m/z
1
(%) 144 ([M + H]+; 100), 145 (9); H NMR (400 MHz; CDCl3) δ
0.90−0.98 [m, 24H, H−C(9, 10)], 1.11−1.27 [m, 16H, H−C(6, 8a)],
1.50−1.68 [m, 8H, H−C(8b, 7)], 2.51−2.56 [m, 2H, H−C(4a, 4a′)],
2.67−2.71 [m, 2H, H−C(4a″, 4a″′)], 3.09−3.14 [m, 2H, H−C(4b″,
4b″′)], 3.33−3.38 [m, 2H, H−C(4b, 4b″)], 3.87−4.05 [m, 4H, H−
C(5)], 4.21−4.23 [m, 2H, H−C(2″, 2″′)], 4.32−4.35 [m, 2H, H−C(2,
2′)]; 13C NMR (100 MHz, CDCl3) δ 11.36, 11.38, 11.50, 11.51, 13.87,
13.96, 14.29, 14.38 [C(9, 10)], 19.77, 19.78, 20.66, 20.67 [C(6)],
Figure 3. GC-FID chromatograms showing the formation of two
isomers of 2-isobutyl-5-methyl-3-oxazoline (2a and 2b) from 2-
isobutyl-5-methyl-3-oxazolidine (1a and 1b) over time.
diethyl ether (50 mL; 95:5 by vol; fraction D), and pentane/diethyl
ether (50 mL; 95:5 by vol; fraction E). Isomer 2a [(R,R) or (R,S); the
6314
dx.doi.org/10.1021/jf301489j | J. Agric. Food Chem. 2012, 60, 6312−6322

Journal of Agricultural and Food Chemistry
Article
25.09, 25.19, 25.49, 25.60 [C(8)], 38.55, 38.57, 38.62, 38.64 [C(7)],
52.69 [C(4″, 4″′)], 54.00, 54.01 [C(4, 4′)], 71.99, 72.04, 72.30, 72.32
[C(5)], 94.91, 94.99 [C(2, 2′)], 95.95, 95.98 [C(2″, 2″′)].
Purification of 2-Benzyl-5-methyl-3-oxazoline (8; Figure 2D).
Elution was performed as follows: pentane (100 mL; fraction A),
pentane/diethyl ether (200 mL; 90:10 by vol; fraction B), pentane/
diethyl ether (150 mL; 90:10 by vol; fraction C), and pentane/diethyl
ether (50 mL; 90:10 by vol; fraction D). Fraction C containing 8 was
used for rechromatography as follows: pentane (100 mL; fraction A),
pentane/diethyl ether (250 mL; 90:10 by vol; fraction B), pentane/
diethyl ether (125 mL; 90:10 by vol; fraction C), pentane/diethyl
ether (100 mL; 80:20 by vol; fraction D), and pentane/diethyl ether
(100 mL; 80:20 by vol; fraction E). Isomer 8a was present in fraction
C, a mixture of compounds 8a and 8b was found in fraction D, and 8b
was isolated in fraction E. MS-EI, m/z (%) 84 (100), 91 (71), 57 (39),
92 (38), 65 (17), 77 (10), 41 (9), 104 (9), 131 (9), 103 (7), 78 (6),
130 (6), 39 (5), 44 (5), 105 (5), 40 (4), 42 (4), 43 (4), 54 (4), 55 (4),
56 (4), 85 (4), 173 (3), 175 (M+; 3), 174 ([M − H]+; 2); MS-CI, m/z
Purification of 2-sec-Butyl-5-methyl-3-oxazoline (4; Figure 2B).
Elution was performed as follows: pentane (100 mL; fraction A),
pentane/diethyl ether (150 mL; 95:5 by vol; fraction B), pentane/
diethyl ether (20 mL; 95:5 by vol; fraction C), pentane/diethyl ether
(50 mL; 95:5 by vol; fraction D), and pentane/diethyl ether (100 mL;
90:10 by vol; fraction E). Fraction B contained isomer 4a, and
fractions D and E contained isomer 4b. No further rechromatography
was needed. MS-EI, m/z (%) 84 (100), 56 (56), 57 (42), 112 (34), 85
(28), 41 (27), 70 (21), 68 (15), 86 (15), 55 (13), 83 (13), 126 (11),
42 (9), 43 (8), 97 (7), 39 (6), 58 (5), 140 (4), 140 ([M − H]+; 3),
1
141 (M+; tr); MS-CI, m/z (%) 142 ([M + H]+; 100), 143 (10); H
NMR (400 MHz; CD2Cl2) δ 0.89 [d, J = 6.8 Hz, 3H, H−C(10)], 0.95
[t, J = 7.4 Hz, 3H, H−C(9)], 1.16−1.29 [m, 1H, H−C(8a)], 1.29 [d, J
= 6.7 Hz, 3H, H−C(6)], 1.51−1.63 [m, 1H, H−C(8b)], 1.66−1.78
[m, 1H, H−C(7)], 4.79−4.87 [m, 1H, H−C(5)], 5.63−5.69 [m, 1H,
H−C(2)], 7.48 [d, J = 2.6 Hz, 1H, H−C(4)]; 13C NMR (100 MHz,
CD2Cl2) δ 11.88 [C(9)], 13.83 [C(10)], 18.60 [C(6)], 24.97 [C(8)],
40.74 [C(7)], 82.35 [C(5)], 109.80 [C(2)], 164.41 [C(4)].
1
(%) 176 ([M + H]+; 100), 177 (8); H NMR (400 MHz; CD2Cl2) δ
1.25 [d, J = 6.7 Hz, 3H, H−C(6)], 2.92 [dd, J = 13.8 Hz, 5.7 Hz, 1H,
H−C(7a)], 3.08 [dd, J = 13.8 Hz, 5.0 Hz, H−C(7b)], 4.67−4.70 [m,
1H, H−C(5)], 5.95−5.98 [m, 1H, H−C(2)], 7.22−7.33 [m, 5H, H−
C(9, 10, 11, 12, 13)], 7.40 [d, J = 2.4 Hz, 1H, H−C(4)]; 13C NMR
(100 MHz, CD2Cl2) δ 18.43 [C(6)], 42.11 [C(7)], 82.39 [C(5)],
106.58 [C(2)], 126.75 [C(11)], 128.46 [C(9, 13)], 130.40 [C(10,
12)], 106.58 [C(8)], 164.70 [C(4)].
Synthesis of 2-Isobutyl-3-oxazoline. 2-Isobutyl-3-oxazoline
(10; Figure 2E) was synthesized following the same approach
described above for 2-isobutyl-5-methyl-3-oxazoline (2) but using 2-
aminoethanol instead of (R)-1-amino-2-propanol in the reaction with
3-methylbutanal.
2-Isopropyl-5-methyl-3-oxazolidine (5; Figure 2C). MS-EI, m/z
(%) 84 (100), 86 (94), 41 (48), 70 (41), 58 (32), 55 (28), 56 (27), 42
(20), 59 (20), 43 (19), 85 (11), 39 (9), 45 (9), 57 (8), 68 (5), 83 (5),
40 (4), 54 (4), 69 (4), 82 (4), 87 (4), 114 (4), 128 ([M − H]+; 3),
1
129 (M+; tr); MS-CI, m/z (%) 130 ([M + H]+; 100), 131 (8); H
NMR (400 MHz; CDCl3) δ 0.96−1.01 [m, 12H, H−C(8, 8′, 9, 9′)],
1.19 [d, J = 6.2 Hz, 3H, H−C(6 or 6′)], 1.21 [d, J = 6.2 Hz, 3H, H−
C(6′ or 6)], 1.72−1.81 [m, 2H, H−C(7, 7′)], 2.54 [dd, J = 11.9, 7.7
Hz, 1H, H−C(4a′)], 2.68 [dd, J = 11.5, 5.0 Hz, 1H, H−C(4a)], 3.12
[dd, J = 11.5, 7.0 Hz, 1H, H−C(4b)], 3.35 [dd, J = 11.8, 6.1 Hz, 1H,
H−C(4b′)], 3.89−4.04 [m, 2H, H−C(5, 5′)], 4.15 [d, J = 5.8 Hz, 1H,
H−C(2)], 4.26 [d, J = 5.8 Hz, 1H, H−C(2′)]; 13C NMR (100 MHz,
CDCl3) δ 17.88 [C(8 or 8′ or 9 or 9′)], 17.94 [C(8 or 8′ or 9 or 9′)],
18.12 [C(8 or 8′ or 9 or 9′)], 18.17 [C(8 or 8′ or 9 or 9′)], 19.76 [C(6
or 6′)], 20.64 [C(6′ or 6)], 32.12 [C(7 or 7′)], 32.24 [C(7′ or 7)],
52.74 [C(4)], 54.03 [C(4′)], 72.22 [C(5 or 5′)], 72.46 [C(5′ or 5)],
96.05 [C(2′)], 97.04 [C(2)].
Analytical Data for 2-Isobutyl-3-oxazolidine and 2-Isobutyl-
3-oxazoline. 2-Isobutyl-3-oxazolidine (9; Figure 2E). MS-EI, m/z
(%) 72 (100), 56 (60), 42 (48), 44 (39), 87 (34), 41 (28), 45 (24),
114 (18), 39 (16), 43 (16), 55 (8), 84 (8), 98 (8), 40 (7), 54 (7), 57
(7), 88 (6), 69 (6), 89 (6), 53 (5), 68 (5), 96 (5), 128 ([M − H]+; 3),
1
129 (M+; tr); MS-CI, m/z (%) 130 ([M + H]+; 100), 131 (7); H
NMR (400 MHz; CDCl3) δ 0.91 [d, J = 6.7 Hz, 6H, H−C(8, 9)],
1.38−1.51 [m, 2H, H−C(6)], 1.69−1.78 [m, 1H, H−C(7)], 2.89−
2.96 [m, 1H, H−C(4a)], 3.13−3.29 [m, 1H, H−C(4b)], 3.59−3.63
[m, 2H, H−C(5)], 4.25−4.28 [m, 1H, H−C(2)]; 13C NMR (100
MHz, CDCl3) δ 22.61 [C(8 or 9)], 22.99 [C(9 or 8)], 25.45 [C(7)],
43.17 [C(6)], 46.22 [C(4)], 64.58 [C(5)], 91.11 [C(2)].
Purification of 2-Isopropyl-5-methyl-3-oxazoline (6; Figure 2C).
Elution was performed as follows: pentane (100 mL; fraction A),
pentane/diethyl ether (150 mL; 95:5 by vol; fraction B), pentane/
diethyl ether (100 mL; 90:10 by vol; fraction C), and pentane/diethyl
ether (50 mL; 90:10 by vol; fraction D). Fractions C and D contained
both isomers, but these could not be separated further. MS-EI, m/z
(%) 84 (100), 56 (36), 57 (35), 112 (26), 83 (20), 41 (15), 70 (12),
68 (11), 43 (9), 85 (9), 55 (5), 39 (4), 126 ([M − H]+; 3), 127 (M+;
Purification of 2-Isobutyl-3-oxazoline (10; Figure 2E). Stepwise
elution was performed as follows: pentane (100 mL; fraction A),
pentane/diethyl ether (50 mL; 95:5 by vol; fraction B), and pentane/
diethyl ether (100 mL; 95:5 by vol; fraction C). Due to the fact that
only one chiral center is present in the molecule, only one peak was
obtained in fraction C and, thus, no rechromatography was necessary.
MS-EI, m/z (%) 70 (100), 42 (21), 69 (19), 71 (19), 41 (18), 85 (14),
43 (12), 54 (12), 56 (9), 57 (8), 39 (8), 55 (4), 72 (4), 82 (4), 126
([M − H]+; 1), 127 (M+; tr); MS-CI, m/z (%) 128 ([M + H]+; 100),
129 (7); 1H NMR (400 MHz; CD2Cl2) δ 0.97 [d, J = 6.7 Hz, 6H, H−
C(8, 9)], 1.42−1.47 [m, 1H, H−C(6a)], 1.55−1.60 [m, 1H, H−
C(6b)], 1.82−1.90 [m, 1H, H−C(8)], 4.46−4.50 [m, 1H, H−C(5a)],
4.57−4.61 [m, 1H, H−C(5b)], 5.64−5.69 [m, 1H, H−C(2)], 7.55 [d,
J = 2.5 Hz, H−C(4)]; 13C NMR (100 MHz, CD2Cl2]) δ 22.73 [C(8)],
23.32 [C(9)], 25.29 [C(7)], 45.13 [C(6)], 75.16 [C(5)], 106.05
[C(2)], 160.31 [C(4)].
1
tr); MS-CI, m/z (%) 128 ([M + H]+; 100), 129 (8); H NMR (400
MHz; CDCl3) δ 0.93 [d, J = 6.8 Hz, 3H, H−C(8)], 0.96 [d, J = 6.8
Hz, 3H, H−C(9)], 1.29 [d, J = 6.7 Hz, 3H, H−C(6)], 1.91−1.99 [m,
1H, H−C(7)], 4.80−4.86 [m, 1H, H−C(5)], 5.56−5.59 [m, 1H, H−
C(2)], 7.48 [d, J = 2.5 Hz, 1H, H−C(4)]; 13C NMR (100 MHz,
CDCl3) δ 16.61 [C(8 or 9)], 17.26 [C(9 or 8)], 18.26 [C(6)], 33.52
[C(7)], 82.10 [C(5)], 110.25 [C(2)], 163.87 [C(4)].
2-Benzyl-5-methyl-3-oxazolidine (7; Figure 2D). MS-EI, m/z (%)
86 (100), 132 (54), 91 (37), 59 (22), 41 (17), 58 (17), 105 (16), 130
(12), 133 (12), 77 (10), 92 (10), 65 (9), 103 (9), 117 (9), 104 (8), 39
(6), 87 (6), 177 (M+; 6), 53 (4), 176 ([M − H]+; 2); MS-CI, m/z (%)
178 ([M + H]+; 100), 179 (12); 1H NMR (400 MHz; CD2Cl2) δ 0.89
[d, J = 6.1 Hz, 3H, H−C(6)], 0.96 [d, J = 6.1 Hz, 3H, H−C(6′)],
2.17−2.24 [m, 2H, H−C(4a, 4a′)], 2.69−2.76 [m, 2H, H−C(4b, 4b′)],
2.79−2.90 [m, 4H, H−C(7, 7′)], 3.59−3.69 [m, 2H, H−C(5, 5′)],
4.55 [t, J = 4.9 Hz, 1H, H−C(2)], 4.69 [t, J = 4.8 Hz, 1H, H−C(2′)],
7.05−7.09 [m, 2H, H−C(11, 11′)], 7.16−7.17 [m, 4H, H−C(10, 10′,
12, 12′)], 7.25−7.28 [m, 4H, H−C(9, 9′, 13, 13′)]; 13C NMR (100
MHz, CDCl3) δ 19.76 [C(6′)], 20.56 [C(6)], 40.45 [C(7′)], 40.91
[C(7)], 52.61 [C(4)], 53.94 [C(4′)], 72.27 [C(5′)], 72.33 [C(5)],
91.36 [C(2′)], 92.41 [C(2)], 126.40 [C(11)], 126.44 [C(11′)], 128.10
[C(10, 12)], 128.23 [C(10′, 12′)], 130.10 [C(9′, 13′)], 130.16 [C(9,
13)].
Synthesis of Isotopically Labeled Standards. [2H2]-3-Methyl-
butanal. 3-Methyl-3-buten-1-ol was deuterated using the Wilkinson
catalyst to obtain [3,4-2H2]-3-methylbutanol, which was subsequently
oxidized into [3,4-2H2]-3-methylbutanal by Dess−Martin period-
inane,16,17 following the approach reported previously for the synthesis
of [5,5,6,6-2H4]-hexanal.18
[2H2]-2-Methylbutanal. Starting from 2-methyl-3-buten-1-ol, the
target compound was synthesized following the two-step procedure
described for [2H2]-3-methylbutanal.
[2H2]-Methylpropanal. Starting from 2-methyl-2-propen-1-ol, the
target compound was synthesized following the two-step procedure
described for [2H2]-3-methylbutanal.
6315
dx.doi.org/10.1021/jf301489j | J. Agric. Food Chem. 2012, 60, 6312−6322

Journal of Agricultural and Food Chemistry
Article
Figure 4. Release of 3-methylbutanal after the addition of water (black line, 3-methylbutanal; red line, [2H2]-3-methylbutanal as internal standard).
[13C2]-Phenylacetaldehyde. [13C2]-Phenylacetaldehyde was synthe-
sized by oxidation of [1,2-13C2]-2-phenylethanol by Dess−Martin
periodinane.
respectively, was added through the septum via a syringe. Preliminary
tests with acetone/air mixtures proved a leak-free connection. Air was
sampled from the flask via a continuous flow of 150 mL/min and
analyzed as described by Lindinger.19 The following parameters were
used: inlet temperature, 120 °C; chamber temperature, 80 °C; inlet
flow, 70 mL/min. With the scan mode applied, samples were scanned
for the full mass range from m/z 20 to 180 at specific dwell times. For
multiple ion detection, the following mass to charge ratios and
individual dwell times were selected: m/z 59 for acetone of breath
(100 ms), m/z 73 for methylpropanal (200 ms), m/z 87 for 2- and 3-
methylbutanal (200 ms), m/z 89 for [2H2]-3-methylbutanal (200 ms),
m/z 121 for phenylacetaldehyde (500 ms), and m/z 123 for [13C2]-
phenylacetaldehyde (500 ms). Signal intensities were normalized to a
theoretical intensity of the primary ion H3O+ of 1 × 107 counts per
second.
Static Headspace High-Resolution Gas Chromatography−
Mass Spectrometry (HS-HRGC-MS). Analysis was performed on a
Trace GC Ultra (Thermo Scientific, Dreieich, Germany) connected to
a mass spectrometer Saturn 2100T ion trap (Varian, Darmstadt,
Germany). Aliquots (500 μL) of the headspace volume were injected
with a CombiPal autosampler (CTC Analytics, Zwingen, Switzerland)
equipped with a 1 mL gastight syringe (SGE Analytic Science,
Darmstadt, Germany) into a hot PPKD injector (Thermo Scientific).
After each injection, a possible carry-over into the syringe was
eliminated by an automatic syringe flush with helium. Volatiles were
trapped in a Cold Trap 915 (Thermo Scientific) at −150 °C and were
subsequently transferred by thermal desorption (initial temperature,
−150 °C for 0.1 min; heating rate, 12 °C/s; final temperature, 240 °C
for 3 min) onto a DB-5 capillary column (30 m × 0.25 mm i.d., 0.25
μm film thickness) (J&W Scientific; Agilent, Waldbronn, Germany)
starting with a temperature of 0 °C for 2 min and subsequent heating
at 6 °C/min to 240 °C (3 min). The oven was heated from −5 °C at 4
°C/min to 50 °C and then to 220 °C at 10 °C/min. The effluent was
monitored by a mass spectrometer running in the chemical ionization
(CI) mode (ionization energy, 70 eV) with methanol as the reagent
gas.
Generation of Strecker Aldehydes from 2-Substituted-5-
methyl-3-oxazolines. The respective 5-methyl-3-oxazoline (2, 4, 6,
or 8; Figure 2A−D) was singly reacted as follows: To the compound
dissolved in pentane/diethyl ether (2 mL) was added ethanol (0.5
mL). The pentane/diethyl ether mixture was carefully evaporated, and
the residue was finally made up to 5 mL with ethanol. Aliquots (0.5
mL) were then added to water (5 mL) containing the respective
isotopically labeled standard, and the solutions were stirred in closed
glass vials at 37 °C for 5, 15, or 30 min, respectively. A control
experiment was performed without the addition of water.
For quantitation of the respective aldehyde by stable isotope
dilution analysis (SIDA), the samples were cooled and either directly
subjected to HS-HRGC-MS or extracted with diethyl ether (total
volume = 45 mL). Then, the organic phases were combined, dried
over anhydrous sodium sulfate, and concentrated to ∼4 mL, and an
aliquot (2 μL) was used for HRGC-MS.
Influence of the pH on the Generation of Aldehydes from 2-
Substituted-5-methyl-3-oxazolines. Buffers were prepared by
mixing solution A (disodium hydrogen phosphate dihydrate; 11.876
g/L of water) and solution B (potassium dihydrogen phosphate; 9.078
g/L of water): pH 5, A + B (0.95 + 99.05 by vol); pH 7, A+B (61.2 +
38.8 by vol); pH 9, an aqueous solution of sodium hydroxide (5 mol/
L) was added to A until a pH of 9 was reached. An aliquot of the
oxazoline solution in ethanol (0.5 mL) was added to the respective
buffer solution (5 mL; either pH 5.0, pH 7.0, or pH 9.0). Then, the
respective isotopically labeled standard was added, and the solutions
were stirred in closed glass vials at 37 °C for 5, 15, or 30 min,
respectively. A control experiment was performed without the addition
of the respective buffer solution.
Quantitation of Strecker Aldehydes Generated from
Chocolate upon Addition of Water. Chocolate (70% cocoa
content) was frozen in liquid nitrogen and ground to a powder by
means of a mill. An aliquot (50 g) was melted at 37 °C, and [2H2]-3-
methylbutanal (53 μg, dissolved in 500 μL of dichloromethane) and
[13C2]-phenylacetaldehyde (18 μg, dissolved in 100 μL of dichloro-
methane) were added. After stirring for 5 min, the chocolate (“spiked
chocolate”) was cooled to room temperature and then used for PTR-
MS measurements.
Two Dimensional Gas Chromatography−Mass Spectrome-
try (TDGC-MS). Analysis was accomplished by means of a GC/GC-
MS system with a heart-cut interface. The system consisted of a Trace
GC Ultra (Thermo Scientific) equipped with a DB-FFAP column (30
m × 0.32 mm i.d., 0.25 μm film thickness) (J&W Scientific), a CP-
3800 gas chromatograph (Varian) equipped with a DB-1701 column
(30 m × 0.32 mm i.d., 0.25 μm film thickness) (J&W Scientific), and a
CombiPal autosampler. The first GC housed a moving column stream
switching system (MCSS) dividing the effluent in equal parts to a
flame ionization detector and a sniffing outlet or, alternatively, leading
the effluent in total to the second GC column. The second column
was connected to a mass spectrometer Saturn 2200 (Varian) running
in the CI mode with methanol as reactant gas (ionization energy, 115
eV). Mass chromatograms were recorded, and concentrations of the
A reference mixture in sunflower oil (10 g) containing
methylpropanal (2.5 μg), 3-methylbutanal (6.25 μg), phenylacetalde-
hyde (25 μg), [2H2]-3-methylbutanal (4.3 μg), and [13C2]-phenyl-
acetaldehyde (20 μg) served as the control.
Proton Transfer Reaction−Mass Spectrometry (PTR-MS).
Spiked chocolate (5.0 g) was weighed into a conical glass flask (200
mL) and sealed with a gastight septum. The chocolate was melted at
37 °C, and then the flask was connected to a PTR-MS (Ionicon
Analytik, Innsbruck, Austria) via a peek capillary. After defined times at
37 °C, either acidified water (5 mL, pH 4) or human saliva (5 mL),
6316
dx.doi.org/10.1021/jf301489j | J. Agric. Food Chem. 2012, 60, 6312−6322

Journal of Agricultural and Food Chemistry
Article
analytes were calculated from the area counts of specific ions of the
analytes and the labeled standards as reported before.7
Characterization of Precursors for Strecker Aldehyde
Generation. One intermediate proposed in the Strecker
reaction following decarboxylation is intermediate A (Figure
5), which upon hydrolysis generates the respective Strecker
HRGC-MS. HRGC-MS analyses were performed by means of a type
5890 series II gas chromatograph (Hewlett-Packard, Waldbronn,
Germany) coupled to a sector field mass spectrometer MAT 95 S
(Finnigan MAT, Bremen, Germany) running either in the electron
impact (EI) mode (ionization energy, 70 eV) or in the CI mode using
isobutane as the reagent gas (ionization energy, 115 eV). The samples
were injected by the cold on-column technique at 40 °C onto a DB-
FFAP fused silica capillary column (30 m × 0.25 mm i.d., 0.25 μm film
thickness) (J&W Scientific). The oven temperature was held for 2 min
isothermally, then raised at 40 °C/min to 60 °C, held for 1 min, and
finally raised at 6 °C/min to 230 °C (3 min).
NMR Spectroscopy. 1H, 13C, and HSQC NMR spectra were
recorded using a Bruker 400 MHz DRX spectrometer (Bruker,
Rheinstetten, Germany). Samples were dissolved either in CD2Cl2 or
in CDCl3 with 0.03% of tetramethylsilane (TMS). Chemical shifts
were determined using TMS as the internal standard (1H NMR) or
the carbon signal of either CD2Cl2 (53.9 ppm; 13C NMR) or CDCl3
(77.0 ppm; 13C NMR), respectively.
RESULTS AND DISCUSSION
■
Measurement of Aroma Release. As reported previ-
ously,14 treatment of chocolate with water at 100 °C led to the
formation of Strecker aldehydes in significant amounts. Because
it can be assumed that the formation of these potent aroma
compounds may also occur in the mouth during mastication,
the release of Strecker aldehydes upon addition of either water
or saliva to chocolate at 37 °C was measured by PTR-MS.
To simulate this process, chocolate (70% cocoa content) was
frozen in liquid nitrogen and ground to a powder. An aliquot
(50 g) was melted at 37 °C and [2H2]-3-methylbutanal as well
as [13C2]-phenylacetaldehyde were added as internal standards.
The “spiked chocolate” was then submitted to PTR-MS
measurements. As exemplified for 3-methylbutanal in Figure
4, the signals of the labeled compound (red line) and the
analyte (black line) were nearly constant between cycles 52
(start of measurement) and 120 without water added,
indicating a constant headspace concentration of standard
and analyte. Upon the addition of water (cycles 120−125), 3-
methylbutanal increased significantly while the internal stand-
ard remained constant (cycles 125−240). This can be seen as
proof that 3-methylbutanal is generated from precursors in
chocolate, which are susceptible to water. The same results
were obtained after the addition of human saliva (data not
shown).
To confirm the results, a blank sample consisting of
methylpropanal, 3-methylbutanal, phenylacetaldehyde, [2H2]-
3-methylbutanal, and [13C2]-phenylacetaldehyde in sunflower
oil was measured by PTR-MS. All monitored compounds
showed only a small increase (∼10%) after the addition of
water due to the change in matrix (data not shown).
To get an initial idea on the amounts of the aldehydes
released, headspace-GC-MS was performed. For this purpose,
the spiked chocolate was stirred for 5 min at 37 °C. Then, the
first injection was done, which was used as blank sample, and
the chocolate sample was cooled to approximately 16 °C. After
the blank run was finished, human saliva was added through the
septum via a syringe and the sample was heated again to 37 °C
under stirring for 5 min. Then, the second injection (sample)
was done. Addition of saliva caused an increase of
methylpropanal by a factor of 43 (from 25.4 to 1103 μg/kg),
of 2-methylbutanal by a factor of 8.5 (from 354 to 3009 μg/kg),
and of 3-methylbutanal by a factor of 8.9 (502 to 4456 μg/kg).
Figure 5. Strecker degradation of leucine catalyzed by an α-dicarbonyl
compound; alternative route leading to 3-methylbutanal via a 4-
oxazoline (B) or a 3-oxazoline (C), respectively.
aldehyde (left side). However, it can be assumed that the 4-
oxazoline (B; Figure 5) and/or the tautomeric 3-oxazoline (C;
Figure 5) might be formed as additional intermediates.
Assuming that these cyclic intermediates are stable in the
absence of water, such oxazolines could be the precursors
suggested to be present in dry foods12 and leading to the
formation of Strecker aldehydes upon the addition of water. To
test this assumption, a common strategy for the synthesis of the
3-oxazolines to be expected from different amino acids was
developed. Because numerous α-dicarbonyl compounds may
occur in foods, which influence the substituents at positions 4
and 5 in the oxazolines, the synthesis was focused on products
to be expected from a reaction of α-amino acids with 2-
oxopropanal. The common synthetic route was planned as a
retrosynthetic approach. (R)-1-Amino-2-propanal was reacted
with either 3-methylbutanal, 2-methylbutanal, methylpropanal,
or phenylacetaldehyde to first yield the corresponding 5-
methyl-3-oxazolidines. These were then oxidized into the
corresponding 5-methyl-3-oxazolines (Figure 2A−D). By this
approach, four 5-methyl-3-oxazolidines and four 5-methyl-3-
oxazolines with different substituents in the 2-position were
newly synthesized. In the following, the application of this
synthetic strategy is exemplified for the synthesis of 2-isobutyl-
5-methyl-3-oxazoline.
The reaction of 3-methylbutanal with (R)-1-amino-2-
propanol directly led to the formation of 2-isobutyl-5-methyl-
3-oxazolidine (1, Figure 2A). Due to the two chiral carbon
atoms, a pair of diastereoisomers should be expected for 1
[(R,R) and (R,S)], but using an FFAP column as stationary GC
phase, no separation was achieved (1a and 1b; Figure 3). The
6317
dx.doi.org/10.1021/jf301489j | J. Agric. Food Chem. 2012, 60, 6312−6322

Journal of Agricultural and Food Chemistry
Article
mass spectra (MS-EI; Figure 6A) and MS-CI (data not shown)
confirmed the assumed molecular mass of 143. The cyclic
group in the oxazolidine (Figure 7) into a methine group in the
oxazoline (Figure 8) confirmed the suggested structure. To
prove that 2a and 2b (Figure 3) are diastereoisomers, a
procedure for their separation by column chromatography was
developed. By application of 1H NMR and 13C NMR
measurements on both isomers, the same spectra showing
only a marginal shift of the signals were obtained, thereby
confirming that both are diastereoisomers (data not shown).
To further corroborate the presence of diastereoisomers, 3-
methylbutanal was reacted with (S)-1-amino-2-propanol,
yielding again two isomers of 2-isobutyl-5-methyl-3-oxazoline
[(S,R) and (S,S)]. These molecules revealed the same mass
spectrometric data as well as the corresponding NMR data,
proving the existence of a second pair of diastereoisomers
bearing the (S) configuration at carbon atom 5 in the ring
system (2c and 2d; data not shown). To finally prove the
stereochemistry, 1-amino-2-propanol was replaced by 2-amino-
ethanol, yielding 2-isobutyl-3-oxazoline (10; Figure 2E). As 10
contained only one chiral atom, this compound showed only
one peak after separation on the achiral GC stationary phase.
The same procedure was applied to synthesize three
additional oxazolines, 2-sec-butyl-5-methyl-3-oxazoline (4;
Figure 2B), 2-isopropyl-5-methyl-3-oxazoline (6; Figure 2C),
and 2-benzyl-5-methyl-3-oxazoline (8; Figure 2D).
Generation of Strecker Aldehydes by Treatment of 2-
Substituted-5-methyl-3-oxazolines with Water. With the
possible precursors of the respective Strecker aldehydes at
hand, the generation of the odorants was studied in model
systems. The first series of experiments was performed with the
leucine-related precursor 2-isobutyl-5-methyl-3-oxazoline (2;
Figure 2A). In a closed vessel, an aliquot of the oxazoline was
dissolved in ethanol, and any contact with water was avoided.
Then, an aliquot of the solution was added to water containing
the isotopically labeled standard [2H2]-3-methylbutanal. The
solution was stirred in closed glass vials for 5, 15, or 30 min,
respectively, at 37 °C.
To ensure the reliability of the data, quantitation was
performed using two different techniques. In the first approach,
the headspace above the solution was withdrawn with a gastight
syringe and directly analyzed by GC-MS. The second approach
used solvent extraction prior to GC-MS analysis.
The results showed that only 5 min after addition of water,
nearly 70 mol % of 2a or 2b, respectively, was converted into 3-
methylbutanal when reacted singly (Table 2). After 30 min,
>90 mol % of the aroma compound was generated. No
differences were found between the data obtained with both
methods of quantitation, suggesting 2 as a very powerful
precursor in the formation of the aroma compound.
The next series of experiments was aimed at getting
information on the influence of the pH value on the generation
of 3-methylbutanal from 2-isobutyl-5-methyl-3-oxazoline. After
10 min at pH 7, >50 mol % of 2a was converted into the aroma
compound (Table 3), corroborating the key role of the 3-
oxazoline in Strecker aldehyde formation. However, the
reaction was even faster at pH 5, leading to a nearly complete
conversion of 2a into 3-methylbutanal after only 10 min. On
the other hand, alkaline conditions (pH 9; Table 3) partly
stabilized the precursor, because only 9 mol % of 3-
methylbutanal was formed from 2a after 10 min.
Figure 6. (A) Mass spectrum (MS-EI) of 2-isobutyl-5-methyl-3-
oxazolidine (1); (B) mass spectrum (MS-EI) of 2-isobutyl-5-methyl-3-
oxazoline (2a).
1
structure was derived from NMR measurements. Both the H
NMR and the two-dimensional HSQC experiment (Figure 7)
proved the absence of an olefinic proton, which should be
present in an acyclic structure (Figure 2A). For data
interpretation, the numbering of the molecules is illustrated
in Figure 2.
The oxazolidine was then oxidized, yielding 2-isobutyl-5-
methyl-3-oxazoline (2; Figure 2A), the respective mass
spectrum (MS-EI) is shown in Figure 6B. Due to the two
chiral carbon atoms [(R,R) and (R,S)], again a pair of
diastereoisomers is to be expected for 2, and two peaks (2a
and 2b; Figure 3) showing identical mass spectra in the MS-EI
as well as in the MS-CI were obtained. GC-MS proved the
expected loss of 2 mass units during the oxidation of 2-isobutyl-
5-methyl-3-oxazolidine into 2-isobutyl-5-methyl-3-oxazoline
(M+, m/z 141). The position of the double bond in 2 was
determined by NMR. By applying one- and two-dimensional
experiments, the double bond between the nitrogen and carbon
atom 4 in the ring system could be proven due to the chemical
shift of the hydrogen atom at carbon atom 4 into the higher
parts per million area. In addition, by means of a HSQC
experiment, the change at carbon atom 4 from a methylene
To confirm that the degradation of 2a and 2b was
quantitatively correlated with 3-methylbutanal formation, the
decrease in their concentrations were measured in parallel with
aldehyde formation (Table 4). As expected, for both isomers a
6318
dx.doi.org/10.1021/jf301489j | J. Agric. Food Chem. 2012, 60, 6312−6322

Journal of Agricultural and Food Chemistry
Article
Figure 7. HSQC NMR spectrum of 2-isobutyl-5-methyl-3-oxazolidine (1).
Figure 8. HSQC NMR spectrum of 2-isobutyl-5-methyl-3-oxazoline (2a).
quite good correlation of the time course of degradation with
aldehyde formation was found (cf. Tables 4 and 2). However,
interestingly, both isomers 2a and 2b are also converted into
each other: 2b is formed to a certain extent during the
degradation of 2a (expt A; Table 4), and 2a is formed during
the degradation of 2b (expt B; Table 4). This result suggested
the possibility of ring opening/closure during the degradation
process.
In the next experiment, the formation of phenylacetaldehyde
from both isomers of 2-benzyl-5-methyl-3-oxazoline (8; Figure
2D) was studied. After a 5 min treatment with water at 37 °C,
∼17 mol % of phenylacetaldehyde was formed from both
6319
dx.doi.org/10.1021/jf301489j | J. Agric. Food Chem. 2012, 60, 6312−6322

Journal of Agricultural and Food Chemistry
Article
respectively (Table 6). However, compared to 2 (cf. Tables 2
and 3), the degradation of 8 was somewhat slower.
Table 2. Time Course of the Generation of 3-Methylbutanal
from 2-Isobutyl-5-methyl-3-oxazoline (2a and 2b; Figure
2A) after the Addition of Water
a
Table 6. Influence of the pH Value on the Generation of
3-methylbutanal (mol %)
Phenylacetaldehyde from 2-Benzyl-5-methyl-3-oxazoline
time (min)
isomer 2a
isomer 2b
a
(8a/8b; Figure 2D)
b
control
0.3
72.7
82.4
91.1
0.4
72.5
81.4
92.2
phenylacetaldehyde (mol %) generated at
5
15
30
time (min)
pH 5
pH 7
pH 9
b
control
0.4/0.5
0.3/0.4
7.7/9.4
0.3/0.4
4.9/5.4
a
b
The precursor was stirred in water at 37 °C. Control experiment
15
30
60
41.2/63.3
72.1/90.3
87.7/93.1
without addition of water.
18.0/19.5
30.2/23.4
7.1/7.6
14.3/10.8
a
Table 3. Influence of the pH Value on the Time Course of
The precursor was stirred in the respective buffer solution at 37 °C.
Control experiment without addition of the respective buffer
b
the Generation of 3-Methylbutanal from 2-Isobutyl-5-
a
methyl-3-oxazoline (2a; Figure 2A)
solution.
3-methylbutanal (mol %) formed at
In a last series of experiments, also the degradation of 4
time (min)
pH 5
pH 7
pH 9
(Figure 2B) and 6 (Figure 2C) at 37 °C in the presence of
water was studied. Due to the very similar behavior of 2a and
2b as well as of 8a and 8b, compounds 4 and 6 were not
separated into the isomers for these experiments. As found for
2 and 8, both oxazolines underwent degradation into the
respective Strecker aldehydes 2-methylbutanal and methylpro-
panal. After the addition of water to 4, 15 mol % of 2-
methylbutanal was quantitated after 5 min, 33 mol % after 15
min, and 56 mol % after 30 min, respectively (Table 7). For
compound 6, after the addition of water, 46 mol % of
methylpropanal was released after 5 min, 54 mol % after 15
min, and 56 mol % after 30 min, respectively (Table 7).
b
control
0.5
95.1
98.9
0.3
54.8
97.9
0.3
9.0
10
60
35.7
a
The precursor was stirred in the respective buffer solution at 37 °C.
Control experiment without addition of the respective buffer
solution.
b
Table 4. Degradation/Isomerization of 2-Isobutyl-5-methyl-
3-oxazoline (2; Figure 2A) in Water at 37 °C
a
expt A
expt B
compd 2a
unreacted
(mol %)
compd 2b
formed
(mol %)
compd 2b
unreacted
(mol %)
compd 2a
formed
(mol %)
time
(min)
Table 7. Generation of 2-Methylbutanal from 2-sec-Butyl-5-
b
methyl-3-oxazoline (4; Figure 2B) and Methylpropanal from
control
98.1
16.2
8.4
1.9
5.1
98.3
30.2
16.7
1.7
3.5
3.7
3.8
a
2-Isopropyl-5-methyl-3-oxazoline (6; Figure 2C)
5
15
30
10.2
4.3
time
(min)
2-methylbutanal (mol %) from methylpropanal (mol %) from
2.9
14.4
isomers 4
isomers 6
a
b
b
The precursor was stirred in water at 37 °C. Control experiment
without addition of water.
control
1.1
14.6
33.3
55.6
0.1
46.0
53.5
56.1
5
15
30
isomers (Table 5) and the yields passed 40 mol % after 30 min.
Thus, also this oxazoline can be regarded as an important
a
b
The precursor was stirred in water at 37 °C. Control experiment
without addition of water.
Table 5. Generation of Phenyacetaldehyde from 2-Benzyl-5-
methyl-3-oxazoline (8a and 8b; Figure 2D) after the
Addition of Water
The next aim was to get a first insight into the presence of
such 3-oxazolines in real foods. On the basis of the recent
results,14 a dark chocolate was extracted with dichloromethane,
and the extract was subjected to SAFE distillation.11 The
distillate was concentrated and analyzed by two-dimensional
GC-MS. Using 2 as the reference, 2-isobutyl-5-methyl-3-
oxazoline was identified on the basis of the same retention
time as well as the same mass spectrum in the chocolate extract
(Figure 9). The concentration of 2-isobutyl-5-methyl-3-oxazo-
line was quite low, but it has to be taken into account that
various α-dicarbonyl compounds, such as glyoxal, 1-deoxy-
osone, 3-deoxyosone, glucosone, or 2,3-pentandione, are
formed by carbohydrate degradation.20 All of these inter-
mediates would be able to generate the respective 3-oxazolines
but having different substituents at positions 4 and 5. Because
glucosone and the 3-deoxyosone are formed in the earlier
stages of the carbohydrate degradation,20 more polar 3-
oxazolines with polyhydroxy functions at positions 4 and 5
may be the major precursors in chocolate or cocoa, respectively.
a
phenylacetaldehyde (mol %) generated from
time (min)
isomer 8a
isomer 8b
b
control
0.2
16.6
23.1
42.5
0.3
17.2
32.2
46.5
5
15
30
a
b
The precursor was stirred in water at 37 °C. Control experiment
without addition of water.
precursor of the Strecker aldehyde. As found for 2, the pH had
also a clear influence on the degradation rate of 8: whereas at
pH 5 after 30 min 72 mol % of 8a and 90 mol % of 8b,
respectively, were converted into phenylacetaldehyde, the yields
were much lower at pH 7 (18 mol % of 8a and 20 mol % of 8b)
as well as at pH 9 (7 mol % of 8a and 8 mol % of 8b),
6320
dx.doi.org/10.1021/jf301489j | J. Agric. Food Chem. 2012, 60, 6312−6322

Journal of Agricultural and Food Chemistry
Article
Figure 9. GC-MS chromatogram of an extract obtained from dark chocolate. Arrows indicate the presence of 2-isobutyl-5-methyl-3-oxazoline.
Figure 10. Generation of a 3-oxazoline by an oxidative decarboxylation of the Amadori compound of phenylalanine.
Alternatively, the oxazolines could directly be formed from the
Amadori compounds as exemplified for phenylalanine degra-
dation in Figure 10. The key step is the oxidation of the
enaminol function in tautomer A2 to initiate the decarbox-
ylation from B into C. Cyclization of this tautomer then leads
to the formation of the oxazoline (D).
A possible pathway for the hydrolytic cleavage of the 3-
oxazolines into the respective Strecker aldehydes is suggested in
Figure 11. Because the degradation was favored under slightly
acidic conditions (cf. Tables 3 and 6), it can be assumed that
protonation at the ring nitrogen allows a nucleophilic attack of
water at ring carbon atom 4. Ring opening generates an amino
acetal, followed by a substitution of the amino group by a
hydroxyl group. The semiacetal formed is finally cleaved into
phenylacetaldehyde and 2-hydroxypropanal.
Figure 11. Hypothetical pathway leading to the formation of
phenylacetaldehyde from 2-benzyl-5-methyl-3-oxazoline.
In conclusion, 3-oxazolines are a novel class of Strecker
aldehyde precursors, which are stable in dry form but easily
hydrolyze and liberate the aldehydes in the presence of water.
The hydrolysis may take place either by the addition of water to
6321
dx.doi.org/10.1021/jf301489j | J. Agric. Food Chem. 2012, 60, 6312−6322

Journal of Agricultural and Food Chemistry
Article
Wick, E. L., Hornstein, J., Eds.; Kluwer Academic/Plenum Publishers:
New York, 1999; pp 147−153.
(13) Schuh, C.; Schieberle, P. Characterization of the key aroma
compounds in the beverage prepared from Darjeeling black tea:
quantitative differences between tea leaves. J. Agric. Food Chem. 2006,
54, 916−924.
(14) Buhr, K.; Pammer, C.; Schieberle, P. Influence of water on the
generation of Strecker aldehydes from dry processed foods. Eur. Food
Res. Technol. 2010, 230, 375−381.
(15) Rizzi, G. The formation of tetramethylpyrazine and 2-isopropyl-
4,5-dimethyl-3-oxazoline in the Strecker reaction of DL-valine with 2,3-
butanedione. J. Org. Chem. 1969, 34, 2002−2004.
(16) Young, J. F.; Osborn, J. A.; Jardine, F. H.; Wilkinson, G. Hydride
intermediates in homogeneous hydrogenation reactions of olefins and
acetylenes using rhodium catalysts. Chem. Commun. 1965, 73, 131−
132.
(17) Dess, D. B.; Martin, J. C. Readily accessible 12-I-5 oxidant for
the conversion of primary and secondary alcohols to aldehydes and
ketones. J. Org. Chem. 1983, 48, 4155−4156.
a food product or during mastication. The hydrolysis efficacy
depends on the pH value of the solution, but lower hydrolysis
speed can be compensated by longer hydrolysis time, for
example, during chewing.
ASSOCIATED CONTENT
■
S
* Supporting Information
Figures S1−S18. This material is available free of charge via the
Internet at http://pubs.acs.org.
AUTHOR INFORMATION
■
Corresponding Author
*Phone: +49 8161 712932. Fax: +49 8161 712970. E-mail:
Peter.Schieberle@Lrz.tum.de.
Notes
The authors declare no competing financial interest.
(18) Steinhaus, M.; Sinuco, D.; Polster, J.; Osorio, C.; Schieberle, P.
Characterization of the key aroma compounds in pink guava (Psidium
guajava L.) by means of aroma re-engineering experiments and
omission tests. J. Agric. Food Chem. 2009, 57, 2882−2888.
(19) Lindinger, W.; Hansel, A.; Jordan, A. On-line monitoring of
volatile organic compounds at pptv levels by means of proton-transfer-
reaction mass spectrometry (PTR-MS) medical applications, food
control and environmental research. Int. J. Mass Spectrom. Ion Processes
1998, 173, 191−241.
ACKNOWLEDGMENTS
■
We gratefully acknowledge the skillful assistance of Joerg Stein.
We thank Johannes Polster, Dr. Oliver Frank, and Johanna
Kreissl for help in performing and interpreting the NMR
experiments. We also thank Ines Otte and Sami Kaviani-Nejad
for performing the GC-MS measurements.
(20) Hofmann, T.; Muench, P.; Schieberle, P. Quantitative model
studies on the formation of aroma-active aldehydes and acids by
Strecker-type reactions. J. Agric. Food Chem. 2000, 48, 434−440.
REFERENCES
■
(1) Strecker, A. Note on a peculiar oxidation by alloxan. Justus Liebigs
Ann. Chem. 1862, 123, 363−365.
(2) Schonberg, A.; Moubacher, R. The Strecker degradation of α-
̈
amino acids. Chem. Rev. 1952, 50, 261−277.
(3) Keeney, M.; Day, E. A. Probable role of the Strecker degradation
of amino acids in development of cheese flavor. J. Dairy Sci. 1957, 40,
874−876.
(4) Bailey, S. D.; Mitchell, D. G.; Bazinet, M. L.; Weurman, C.
Studies on the volatile components of different varieties of cocoa
beans. J. Food Sci. 1962, 27, 165−170.
(5) Rizzi, G. The Strecker degradation of amino acids: newer avenues
for flavor formation. Food Rev. Int. 2008, 24, 416−435.
(6) Czerny, M.; Christlbauer, M.; Christlbauer, M.; Fischer, A.;
Granvogl, M.; Hammer, M.; Hartl, C.; Hernandez, N.; Schieberle, P.
Re-investigation on odour thresholds of key food aroma compounds
and development of an aroma language based on odour qualities of
defined aqueous odorant solutions. Eur. Food Res. Technol. 2008, 228,
265−273.
(7) Hofmann, T.; Schieberle, P. New results on the formation
pathways of aroma-active compounds by Strecker-type reaction. In
Frontiers of Flavor Science; Schieberle, P., Engel, K.-H., Eds.; Deutsche
Forschungsanstalt fur Lebensmittelchemie: Garching, Germany, 2000;
̈
pp 481−486, ISBN 3-00-005556-8.
(8) Granvogl, M.; Bugan, S.; Schieberle, P. Formation of amines and
aldehydes from parent amino acids during thermal processing of cocoa
and model systems: new insights into pathways of the Strecker
reaction. J. Agric. Food Chem. 2006, 54, 1730−1739.
(9) Hofmann, T.; Schieberle, P. Formation of aroma-active Strecker-
aldehydes by a direct oxidative degradation of Amadori compounds. J.
Agric. Food Chem. 2000, 48, 4301−4305.
(10) Cremer, D. R.; Vollenbroeker, M.; Eichner, K. Inverstigation of
the formation of Strecker aldehydes from the reaction of Amadori
rearrangement products with α-amino acids in low moisture model
systems. Eur. Food Res. Technol. 2000, 211, 400−403.
(11) Engel, W.; Bahr, W.; Schieberle, P. Solvent assisted flavour
evaporation − a new and versatile technique for the careful and direct
isolation of aroma compounds from complex food matrices. Eur. Food
Res. Technol. 1999, 209, 237−241.
(12) Schieberle, P.; Pfnuer, P. Characterization of key odorants in
chocolate. In Flavor Chemistry: Thirty Years of Progress; Teranishi, R.,
6322
dx.doi.org/10.1021/jf301489j | J. Agric. Food Chem. 2012, 60, 6312−6322