282
J. Agric. Food Chem. 1996, 44, 282−289
On th e Role of 2,3-Dih yd r o-3,5-d ih yd r oxy-6-m eth yl-4(H)-p yr a n -4-on e
in th e Ma illa r d Rea ction
Myong-Ock Kim and Werner Baltes*
Institute of Food Chemistry, Technical University Berlin, Gustav-Meyer-Allee 25, D-13355 Berlin, Germany
To investigate the thermal degradation pathways of 2,3-dihydro-3,5-dihydroxy-6-methyl-4(H)-pyran-
-one (1) in the Maillard reaction, the 1 C-labeled and unlabeled 1 were synthesized and heated in
3
4
model systems of food processing. The extent and position of the labeling of the reaction products
were interpreted by the mass spectroscopy data. The volatiles identified were, among others, 2,4-
dihydroxy-2,5-dimethyl-3(2H)-furanone (2), 2,5-dimethyl-4-hydroxy-3(2H)-furanone, cyclotene, mal-
tol, 5-hydroxymaltol, and some acyclic carbonyls. Under roasting conditions, 2 was formed as a
major product. It was concluded that 1 might be transferred to highly reactive open-chain
intermediates like the enolic forms of 1-deoxyosone. The further reaction pathways varied with
the reaction conditions. Possible degradation pathways of 1 that resulted from the labeling
experiments as well as the formation of the described products are discussed.
Keyw or d s: Maillard reaction; isotopic labeling experiments; 2,3-dihydro-3,5-dihydroxy-6-methyl-
4
(H)-pyran-4-one; acetylformoin; 2,5-dimethyl-4-hydroxy-3(2H)-furanone; cyclotene
INTRODUCTION
was dissolved in 40 mL of water and extracted with ethylac-
etate for 3 h, and the organic layer was evaporated. The yellow
oil of 1 was distilled at 120-140 °C and 0.1 torr. The oil was
recrystallized twice from ether:pentane. The purified sample
was stored in a refrigerator, and its purity was always
examined by gas chromatography (GC) prior to use: yield, 0.9
The thermal generation of aromas is primarily influ-
enced by the Maillard reaction. By reaction of sugars
with amino acids, a great many decomposition reactions
take place that are responsible for the large number of
aroma compounds in thermal aromas. To gain knowl-
edge about the reaction mechanisms, we carried out
some model reactions of sugars with selected amino
acids, which were heated under the conditions of cook-
ing, roasting, and autoclaving of food (Baltes and Knoch,
g (3% of theory); mp 73-74 °C (67-70 °C; Shaw et al., 1971);
1
3
H NMR (in CDCl ) 2.10 (3H, s), 4.45 (1H, q), 4.38 (1H, q),
and 4.04 (1H, q) ppm.
13
Syn th esis of 2,3-Dih yd r o-3,5-d ih yd r oxy-6-( C)m eth yl-
13
4
(H)-p yr a n -4-on e. C-Labeled 1 was synthesized from 27
13
mmol of (1- C)-R-D-glucose and 27 mmol of piperidine as
described for 1.
1
993; Kunert-Kirchhoff and Baltes, 1990).
In the course of some model reactions, we obtained
relatively large amounts of 2,3-dihydro-3,5-dihydroxy-
-methyl-4(H)-pyran-4-one (1; Reese and Baltes, 1992),
Syn th esis of 2,4-Dih yd r oxy-2,5-d im eth yl-3(2H)-fu r a -
n on e (2). Compound 2 was synthesized according to the
procedures of Goto et al. (1963).
6
which was formed via the 1-deoxyosone pathway (Mills
and Hodge, 1976), has been identified by Mills et al.
Degr a d a t ion of 2,3-Dih yd r o-3,5-d ih yd r oxy-6-m et h yl-
(H)-p yr a n -4-on e (1). First, 0.7 mmol of 1 was dissolved in
mL of sodium phosphate buffer solution (0.4 M, pH 5.8) or
in 1 mL of distilled water in an ampule and heated either
under reflux for 1-5 h or at 150 °C in the drying oven for 1 h.
The reaction mass was then extracted with diethylether and
4
1
(
(
1970), and is found in many heated and stored foods
Tatum et al., 1967; Ledl et al., 1976). Therefore, we
supposed that 1 might be a relatively stable degradation
compound of hexoses. To investigate the role of 1 during
the Maillard reaction, we synthesized it from glucose
and used it in model reactions as described. To confirm
the degradation pathways as well as the formation
mechanisms of products formed (Nyhammar et al.,
treated with NaHCO
3
solution. After drying the diethylether
by freezing and removal of the water at -18 °C, the extract
was concentrated by careful distillation on a Vigreux column
and analyzed by capillary GC-MS. Another sample was
prepared in a similar manner, except the initial pH of solution
1
983), we also carried out isotopic labeling experiments
1
3
was adjusted to 7.5 with 10% Na
2 3
CO solution.
with C-labeled 1.
Degr a d a t ion of 2,3-Dih yd r o-3,5-d ih yd r oxy-6-(13C)-
1
3
m eth yl-4(H)-p yr a n -4-on e. First, 0.3 mmol of C-labeled 1
was dissolved in 0.3 mL of sodium phosphate buffer solution
EXPERIMENTAL PROCEDURES
(0.4 M, pH 5.8) in an ampule and heated at 150 °C in the
drying oven for 1 h. The reaction mass was then extracted
with diethylether and treated as described for 1.
Syn th esis of 2,3-Dih yd r o-3,5-d ih yd r oxy-6-m eth yl-4(H)-
p yr a n -4-on e (1). Compound 1 was synthesized according to
the procedures of F. Ledl (personal communication) with the
following slight modifications. A mixture of 0.2 mol of R-D-
glucose, 0.2 mol of piperidine, and 150 mL of ethanol was
refluxed for 1.5 h. Then, 0.2 mol of acetic acid in 30 mL of
ethanol was added slowly, and the mixture was heated at 90
Degr a d a tion of Acetylfor m oin (2), 1-Hyd r oxy-2-p r o-
p a n on e (22), a n d Glycer a ld eh yd e (39). First, 0.7 mmol of
2
was dissolved in 1 mL of sodium phosphate buffer solution
(0.4 M, pH 5.8) in an ampule and heated at 150 °C in the
drying oven for 1 h. In another experiment, 0.04 mol of
1-hydroxy-2-propanone (22) or glyceraldehyde (39) in 60 mL
of sodium phosphate buffer solution (0.4 M, pH 5.8) was heated
in a laboratory autoclave (Berghof, Germany) at 150 °C for 1
h. The reaction mixtures were treated as already described.
°
C for 22 h. Ethanol was evaporated under reduced pressure
to one-third the original volume. The insoluble piperidinore-
ductone was filtered and washed with isopropanol, and the
solvent was evaporated under reduced pressure. The residue
0021-8561/96/1444-0282$12.00/0
© 1996 American Chemical Society