Thermal Degradation of Allyl Isothiocyanate in Aqueous Solution

Article abstract of DOI:10.1021/jf970488w

Allyl isothiocyanate in an aqueous solution was heated and refluxed at 100 °C for 1 h. The reaction mixtures were then extracted with methylene chloride and analyzed using gas chromatography (GC) and gas chromatography-mass spectrometry (GC-MS). The mixtures in aqueous phase were analyzed by high-performance liquid chromatography (HPLC) and liquid chromatography-mass spectrometry (LC-MS) equipped with an atmospheric-pressure chemical ionization (APCI) interface. The compounds identified in the methylene chloride extracts included diallyl sulfide, diallyl disulfide, diallyl trisulfide, diallyl tetrasulfide, allyl thiocyanate, 3H-1,2-dithiolene, 2-vinyl-4H-1,3-dithiin, 4H-1,2,3-trithiin, and 5-methyl-1,2,3,4-tetrathiane. N,N′-Diallylthiourea, which was the major degradation product in the aqueous phase from the thermal reaction of allyl isothiocyanate, was identified by using LC-MS (APCI+), direct-probe EI-MS, and ¹H-NMR. The possible mechanism for the formation of these products was proposed.

Full text of DOI:10.1021/jf970488w

220
J. Agric. Food Chem. 1998, 46, 220−223
Th er m a l Degr a d a tion of Allyl Isoth iocya n a te in Aqu eou s Solu tion
Chung-Wen Chen and Chi-Tang Ho*
Department of Food Science, Cook College, New J ersey Agricultural Experiment Station, Rutgers,
The State University of New J ersey, New Brunswick, New J ersey 08901-8520
Allyl isothiocyanate in an aqueous solution was heated and refluxed at 100 °C for 1 h. The reaction
mixtures were then extracted with methylene chloride and analyzed using gas chromatography
(GC) and gas chromatography-mass spectrometry (GC-MS). The mixtures in aqueous phase were
analyzed by high-performance liquid chromatography (HPLC) and liquid chromatography-mass
spectrometry (LC-MS) equipped with an atmospheric-pressure chemical ionization (APCI) interface.
The compounds identified in the methylene chloride extracts included diallyl sulfide, diallyl disulfide,
diallyl trisulfide, diallyl tetrasulfide, allyl thiocyanate, 3H-1,2-dithiolene, 2-vinyl-4H-1,3-dithiin, 4H-
1
,2,3-trithiin, and 5-methyl-1,2,3,4-tetrathiane. N,N′-Diallylthiourea, which was the major degrada-
tion product in the aqueous phase from the thermal reaction of allyl isothiocyanate, was identified
by using LC-MS (APCI+), direct-probe EI-MS, and H-NMR. The possible mechanism for the
formation of these products was proposed.
1
Keyw or d s: Allyl isothiocyanate; thermal degradation; N,N′-diallylthiourea; LC-MS; atmospheric-
pressure chemical ionization
INTRODUCTION
EXPERIMENTAL PROCEDURES
Allyl isothiocyanate (AITC), which is a major pungent
flavor compound in mustard and Wasabi, is generated
from its precursor, allyl glucosinolate, namely, sinigrin
Th er m a l Rea ction of Allyl isoth iocya n a te (AITC). A
sample of 1 g of allyl isothiocyanate (Aldrich Chemical Co.,
Milwaukee, WI) in 200 mL of distilled water was adjusted to
pH 2.7, 7.0, or 9.0 with 0.1 N HCl or 0.1 N NaOH. The reaction
solution was heated under reflux at 100 °C for 1 h. The
reaction mixtures were then spiked with 0.7 mL of dodecane
(
Kawakishi and Namiki, 1969; Masuda et al., 1996).
Sinigrin, which widely exists in mustard and other
members of the Brassica species of the cruciferous
vegetables such as cabbage and cauliflower, breaks
down and releases AITC by the action of myrosinase
(10.9 mg/mL) as internal standard and extracted with 25 mL
of methylene chloride. The methylene chloride phase was
dried over anhydrous sodium sulfate and concentrated to 5
mL by a stream of nitrogen gas prior to gas chromatography
(
thioglucoside glucohydrolase) when the plant tissue is
disrupted (Chin and Lindsay, 1993; Shahidi and Gabon,
990; VanEtten et al., 1976; Zrybko and Rosen, 1997).
(GC) and gas chromatography-mass spectrometry (GC-MS)
1
analysis. The concentration of each individual compound was
calculated by the equation of A/IS × R; where A is the area
count of the compound, IS is the area count of the internal
standard, and R ()2.1) is the GC-FID response factor of
dodecane to AITC.
The stability of sinigrin in the methanol/ammonia/water
solution, which was used for the extraction of mustard
seed, over a 24-h period at various temperatures was
reported by Shahidi and Gabon (1990) and Shahidi
The aqueous phase was concentrated to 20 mL by a vacuum
rotary evaporator and analyzed using high-performance liquid
chromatography (HPLC) and liquid chromatography-mass
spectrometry (LC-MS). The compound of interest was col-
lected by HPLC and identified by using direct-probe electron-
(
1993). Sinigrin was degraded by 16-21% at 45 °C but
only degraded by 1-2% at 25 °C, and its decomposition
products including allyl cyanide, thioglucose and its
dimer, glucose, furfuryl alcohol, but-3-enylnitrile, allyl
isothiocyanate, and 3,4-epithiobutanenitrile were also
reported (Shahidi, 1993; Shahidi and Gabon, 1990).
It has been shown that AITC is unstable and is
gradually decomposed to other compounds having a
garliclike odor in the presence of water at both room
temperature and 37 °C (Kawakishi and Namiki, 1969).
AITC is also reported to be sensitive to temperature and
pH. The high temperature of 37 °C and alkaline
conditions accelerate the decomposition of AITC (Ina et
al., 1981).
Because no information exists about the stability and
chemical reaction of AITC under normal cooking condi-
tions, the purpose of this study was to investigate the
thermal stability of AITC in boiling water and to
identify thermal degradation products from AITC. This
paper also reports the pH effect on the thermal stability
of AITC.
1
impact MS and H NMR.
GC a n d GC-MS An a lysis. A Varian model 3400 gas
chromatograph equipped with a flame ionization detector and
a fused silica capillary column (DB-Wax, J &W 30 m × 0.25
mm i.d.) was used. The GC oven was temperature-pro-
grammed from 50 to 220 °C at a rate of 4 °C/min. The carrier-
gas (He) flow rate was 2.0 mL/min. A split ratio of 25:1 was
used. The temperatures of the injection port and detector were
2
40 and 250 °C, respectively. The GC-MS analysis was
performed on an HP model 5890 coupled to an HP 5971 mass-
selective detector. The same column and temperature pro-
gram were used.
HP LC a n d LC-MS An a lysis. HPLC analysis was per-
formed using a Varian 5000 liquid chromatograph equipped
with a Varian 2050 detector. A TSK-GEL ODS-80TS reverse-
phase column (250 mm × 4.6 mm i.d., TOSOH Co., J apan)
was used. Elution was carried out at a flow rate of 1.0 mL/
min using a solvent system of acetonitrile/water (40/60) in 10
min isocratically. The program was then linearly increased
to 100% acetonitrile in 2 min and kept at 100% acetonitrile
for another 15 min. The eluate was monitored at 254 nm. MS
*
Author to whom correspondence should be addressed [fax
(732) 932-8004; e-mail ho@aesop.rutgers.edu].
S0021-8561(97)00488-3 CCC: $15.00 © 1998 American Chemical Society
Published on Web 01/19/1998

Allyl Isothiocyanate Degradation
J. Agric. Food Chem., Vol. 46, No. 1, 1998 221
Ta ble 1. Ma ss Sp ectr a l Da ta of Vola tile Com p ou n d s fr om Th er m a l Degr a d a tion of Allyl Isoth iocya n a te
compound
diallyl sulfide
RIa
MS spectral data m/z (relative intensity)
identification
+
+
1
2
3
4
5
6
7
8
9
1
1118
116 ((M + 2) , 1), 114 (M , 27), 99 (31), 73 (57), 72 (34),
1 (28), 47 (21), 45 (100), 41 (47)
b, c
7
allyl isothiocyanatef
allyl thiocyanate
1353
1440
1468
1513
1787
1833
1841
2005
101 ((M + 2) , 3), 99 (M , 74), 72 (33), 71 (9), 70 (6),
+
+
5
9 (5), 58 (5), 45 (12), 44 (8), 41 (100)
+
+
101 ((M + 2) , 3), 99 (M , 56), 98 (10), 72 (32), 71 (11),
e
5
8 (23), 46 (6), 45 (23), 41 (100)
+
+
diallyl disulfide
148 ((M + 2) , 1), 146 (M , 10), 113 (12), 105 (12),
b, c
d
8
1 (19), 79 (11), 71 (12), 47 (12), 45 (45), 41 (100)
+
+
3H-1,2-dithiolene
106 ((M + 2) , 5), 105 (13), 104 (M , 56), 103 (100),
7
1 (14), 69 (12), 64 (10), 58 (9), 45 (33)
+
+
diallyl trisulfide
180 ((M + 2) , 1), 178 (M , 7), 113 (96), 103 (9), 73 (100),
c
6
9 (4), 64 (13), 47 (15), 45 (64), 41 (80)
+
+
2-vinyl-4H-1,3-dithiin
4H-1,2,3-trithiin
146 ((M + 2) , 7), 144 (M , 57), 111 (45), 103 (15),
b, c
e
9
7 (20), 73 (13), 72 (86), 71 (100), 45 (45)
+
+
138 ((M + 2) , 16), 136 (M , 100), 103 (15), 72 (54),
7
1 (93), 69 (10), 64 (19), 45 (35)
+
+
5-methyl-1,2,3,4-tetrathiane
0 diallyl tetrasulfide
172 ((M + 2) , 18), 170 (M , 100), 128 (43), 106 (30),
c
6
4 (34), 59 (10), 45 (13), 41 (18)
+
210 (M , 12), 146 (55), 105 (37), 104 (16), 81 (24),
3 (78), 64 (52), 45 (45), 41 (100)
c
7
a
RI ) retention index, CW-20M. Mass spectral database of Wiley 138. c Mass spectral data from Block et al. (1988). d Mass spectral
b
data from Yu et al. (1994). Tentative identification according to mass spectral data of EI. f Starting material.
e
analysis was performed on a Micromass Platform II system
(Micromass Co., MA) equipped with a Digital DECPc XL560
computer for data analysis. Positive-ion mass spectra were
obtained using the heated nebulizer atmospheric-pressure
chemical-ionization (APCI) interface. The ion-source temper-
ature was set at 150 °C, and the sample cone voltage was 10
V.
Dir ect P r obe EI-MS An a lysis. Direct probe EI-MS was
conducted on a Finnigan MAT 8230 high-resolution mass
spectrometer. The temperature of the probe was increased
from 35 to 350 °C at 5 °C/s.
1
1
H-NMR An a lysis. The H-NMR (360 MHz) spectrum was
obtained in pyridine-d solution on a modified Nicolet NT 360
5
instrument. Chemical shifts are reported as parts per million
relative to tetramethylsilane (TMS), and coupling constants
are in hertz.
Syn th esis of N,N′-Dia llylth iou r ea . AITC (0.1 mol) was
diluted with 30 mL of toluene, and allylamine (0.1 mol) was
added dropwise. The reaction solution was then stirred at 60
°
C for 2 h. The reaction mixture was concentrated and kept
in a -20 °C freezer for overnight. The resulting solid was
washed with hexane three times. The purity of this product
was 99% based on HPLC analysis. This method was modified
from the method of Pascual and Rindlisbacher (1994).
F igu r e 1. Proposed mass fragmentation for the compound 8.
atoms of sulfur. This compound was proposed as 4H-
RESULTS AND DISCUSSION
1
,2,3-trithiin based on the mass fragmentation as shown
Vola tile Com p ou n d s Gen er a ted fr om th e Th er -
m a l Degr a d a tion of AITC. After thermal reaction in
boiling water for 1 h, the odor of AITC was totally
changed from a pungent, mustardlike flavor note to a
strong garliclike note. The degraded solution of AITC
was extracted and then analyzed by GC and GC-MS.
Table 1 lists the volatile compounds identified from the
thermal degradation of AITC. These compounds were
tentatively identified by comparing their mass spectral
data with references or by interpreting the CI mass
spectra (compounds 3 and 8). The mass spectral data
of compound 3 is very similar to that of AITC (compound
in Figure 1.
It has been reported that AITC can gradually be
decomposed to allyl dithiocarbamate (compound A in
Figure 2), which is then degraded to diallyl polysulfides
when AITC is kept in an aqueous solution at 37 °C for
several days (Kawakishi and Namiki, 1969). However,
compound A was not detected when AITC was heated
at 100 °C for 1 h according to our study. It may be
possible that compound A is thermally labile and easily
decomposed under such a vigorous heating condition or
that generation of this compound is not favorable in this
cooking condition.
The mechanisms for the formation of these volatile
compounds from AITC are proposed as shown in Figure
2. Allyl mercaptan (compound B), which could be a
decomposition product from compounds 3 or A through
a hydrolysis process or reduction reaction (Kawakishi
and Namiki, 1969), is a key intermediate for generating
aliphatic sulfides and cyclic sulfides. The dimerization
of allyl mercaptan leads to the formation of diallyl
disulfide (compound 4) as proposed by Kawakishi and
2
). This compound was tentatively identified as allyl
thiocyanate, an isomer of AITC. Allyl thiocyanate can
also be thermally generated from AITC when pure AITC
was introduced into the GC. The thermal isomerization
of AITC to allyl thiocyanate is due to the high temper-
ature of the injection port based on our study. According
to the mass spectral data of compound 8 (Table 1), the
+
•
fragment intensity of (M + 2) is 16% of the molecular
ion. This suggests that compound 8 contains three

222 J. Agric. Food Chem., Vol. 46, No. 1, 1998
Chen and Ho
Ta ble 2. p H Effect on th e Gen er a tion of Vola tile
Com op u n d s fr om Th er m a l Degr a d a tion of Allyl
Isoth iocya n a te (AITC)
amount (mg/g of AITC)
pH 2.7 pH 7.0 pH 9.0
compound
diallyl sulfide
allyl isothiocyanate
allyl thiocyanate
diallyl disulfide
3H-1,2-dithiolene
diallyl trisulfide
2-vinyl-4H-1,3-dithiin
4H-1,2,3-trithiin
1
2
3
4
5
6
7
8
9
1
13.09
379.06
9.21
8.09
0.13
22.16
290.27
13.84
11.47
0.17
13.63
tr
26.40
270.46
12.32
12.71
0.19
14.11
tr
10.31
a
tr
tr
tr
tr
5-methyl-1,2,3,4-tetrathiane
0 diallyl tetrasulfide
0.01
tr
0.02
tr
0.03
tr
a
tr ) trace.
F igu r e 2. Possible pathways for the formation of volatile
compounds generated from thermally degraded allyl isothio-
cyanate.
Namiki (1969). It has been shown that heating pure
diallyl disulfide leads to significant amounts of diallyl
sulfide, diallyl trisulfide, and diallyl tetrasulfide (Block
et al., 1988). Diallyl sulfide (compound 1) and diallyl
trisulfide (compound 6) can be generated from diallyl
disulfide by a disproportionation or an allyldithio-radical
addition reaction (Block et al., 1988). Diallyl tetrasul-
fide probably occurs from the interaction of two al-
lyldithio radicals. 2-Vinyl-4H-1,3-dithiin (compound 7)
is a thermal degradation product from garlic oil and can
be generated from two molecules of thioacrolein
F igu r e 3. HPLC chromatograph of the nonvolatile products
generated from thermally degraded allyl isothiocyanate: col-
umn, TSK-GEL ODS-80TS reverse-phase column (250 mm ×
4
.6 mm i.d.); flow rate, 1.0 mL/min; solvent system, acetoni-
(
CH2dCHsCHdS) (Block et al., 1988). It is possible
trile/water (40/60) for 10 min, 40%-100% acetonitrile in 2 min,
that thioacrolein results from the oxidation of allyl
mercaptan. Compounds 9 and 5 were reported to be
the thermal degradation products from alliin and deoxy-
alliin, respectively, the flavor precursors of garlic (Yu
et al., 1994). Yu et al. (1994) proposed that compound
1
00% acetonitrile for 15 min; detector wavelength, 254 nm.
The remaining AITC (compound 2) was less than 38%
after thermal reaction for 1 h. The amount of allyl
thiocyanate (compound 3) was calculated after deducing
the thermally isomerized allyl thiocyanate (about 10%
from AITC) caused by the high temperature at the GC
injection port. The AITC was more stable when the
reaction was conducted in an acidic solution than in
neutral or alkaline solution. This result is consistent
with the results of Ina et al. (1981). The amounts of
volatile compounds increased when the pH increased
(except compound 3). Table 2 shows that the amount
of compound 3 generated at pH 7.0 is more than pH
2.7 and 9.0, suggesting the isomerization of AITC is
more favorable in neutral conditions than in acidic and
basic conditions.
9
could be from the interaction of one molecule of allyl
mecaptan and three molecules of hydrogen sulfide, while
compound 5 could be from interaction of one molecule
of allyl mecaptan and one molecule of hydrogen sulfide
followed by an oxidation reaction. Compound 8 has not
been reported to be the thermal degradation product of
garlic oil. However, its analogous compound, 1,2,3-
trithiane, has been found in the thermally degraded
alliin aqueous solution and its formation mechanism is
proposed as the interaction of one molecule of allyl
mecaptan and two molecules of hydrogen sulfide (Yu et
al., 1994). 1,2,3-Trithiane can probably be further
oxidized to generate compound 8.
Non vola tile Com p ou n d s fr om Th er m a l Degr a -
d a tion of AITC. After thermal reaction, the nonvola-
tile compounds in the aqueous phase were analyzed by
HPLC. There is one major peak (retention time ) 7.6
p H Effect on th e Gen er a tion of Vola tile Com -
p ou n d s fr om AITC. The pH effect on the generation
of volatile compounds from AITC is shown in Table 2.

Allyl Isothiocyanate Degradation
J. Agric. Food Chem., Vol. 46, No. 1, 1998 223
Con clu sion . Our study demonstrate that AITC, a
pungent mustardlike flavor compound, is a very un-
stable compound. It can be decomposed to aliphatic and
cyclic sulfur-containing compounds with a garliclike
flavor note under cooking condition. AITC can also be
decomposed to nonvolatile compound N,N′-diallylthio-
urea. The disubstituted alkylthioureas have been re-
ported to possess the activity against insect pests and
spider mites (Pascual and Rindlisbacher, 1994). The
biological activity of N,N′-diallylthiourea needs to be
addressed. The generation of N,N′-diallylthiourea from
foods containing AITC under cooking or storage condi-
tions also need to be further studied.
F igu r e 4. Direct-probe EI-MS of a major nonvolatile com-
pound from the thermal degradation of allyl isothiocyanate.
LITERATURE CITED
Block, E.; Iyer, R.; Grisoni, S.; Saha, C.; Belman, S.; Lossing,
F. P. Lipoxygenase inhibitors from the essential oil of garlic.
Markovnikov addition of the allyldithio radical to olefins.
J . Am. Chem. Soc. 1988, 110, 7813-7827.
Chin, H.-W.; Lindsay, R. C. Volatile sulfur compounds formed
in disrupted tissues of different cabbage cultivars. J . Food
Sci. 1993, 58, 835-839, 841.
Ina, K.; Nobukuni, M.; Sano, A.; Kishima, I. Stability of allyl
isothiocyanate. Nippon Shokuhin Kogyo Gakkaishi 1981, 12,
6
27-631.
Kawakishi, S.; Namiki, M. Decomposition of allyl isothiocy-
anate in aqueous solution. Agric. Biol. Chem. 1969, 33, 452-
4
59.
Masuda, H.; Harada, Y.; Tanaka, K.; Nakajima, M.; Tateba,
H. Characteristic odorants of Wasabi (Wasabia japonica
Matum), J apanese horseradish, in comparison with those
of horseradish (Armoracia rusticana). In Biotechnolgy for
Improved Foods and Flavors; Takeoka, G. R., Teranishi, R.,
Williams, P. J ., Kobayashi, A., Eds.; ACS Symposium Series
637; American Chemical Society: Washington, DC, 1996;
pp 67-78.
Pascual, A.; Rindlisbacher, A. N-(Pyrid-3-yl)thioureas and
derivatives as acaricides. I. Synthesis and biological proper-
ties. Pestic. Sci. 1994, 42, 253-263.
Shahidi, F. Thiglucosides of Brassica oilseeds and their
process-induced chemical transformations. In Sulfur Com-
pounds in Foods; Mussinan, C. J ., Keelan, M. E., Eds.; ACS
Symposium Series 564; American Chemical Society: Wash-
ington, DC, 1993; pp 106-126.
F igu r e 5. Possible pathways for the formation of N,N′-
diallylthiourea from allyl isothiocyanate (Kawakishi and
Namiki, 1969).
min) present in the HPLC chromatograph as shown in
Figure 3. The molecular weight of this compound was
determined as 156 using LC-MS (APCI+). The com-
pound of interest was collected by HPLC and further
1
identified by H NMR and direct-probe EI-MS. The UV
1
and H-NMR data of this compound are as follows. UV
1
max (in methanol): 211 and 243 nm. H NMR (pyridine-
d5 vs TMS): 4.48 (2H, br), 5.08 (1H, dd, J ) 10.3, 1.3
Hz), 5.27 (1H, dd, J ) 17.2, 1.4 Hz), 6.05 (1H, m), 8.46
(
1H, br). The mass spectrum of this compound is
present in Figure 4. The strong fragment of m/z 41
Shahidi, F.; Gabon. J .-E. Fate of singrin in methanol/ammo-
nial/water-hexan extraction of B. J uncea mustard seed. J .
Food Sci. 1990, 55, 793-795.
suggests the presence of the allyl group in this structure.
+
The fragment of m/z 56 could be CH2dCHsCHdNH2 ,
and the fragment of m/z 141 is obtained after the double
bond on the allyl group is shifted to the adjacent carbon
followed by a loss of a methyl group. Combining the
VanEtten, C. H.; Daxenbichler, M. E.; Williams, P. H.; Kwolek,
W. F. Glucosinolates and derived products in Crusiferous
vegetables. Analysis of the edible part from twenty-two
varieties of cabbage. J . Agric. Food Chem. 1976, 24, 452-
1
mass-spectral and H-NMR data, this compound was
4
55.
proposed as N,N′-diallylthiourea. This structure was
Yu, T.-H.; Wu, C.-M.; Rosen, R. T.; Hartman, T. G.; Ho, C.-T.
Volatile compounds generated from thermal degradation of
alliin and deoxyalliin in an aqueous solution. J . Agric. Food
Chem. 1994, 42, 146-153.
Zrybko, C. L.; Rosen, R. T. Determination of glucosinolates in
mustard by high-performance liquid chromatography-elec-
trospray mass spectrometry. In Spices: Flavor Chemistry
and Antioxidant Properties; Risch, S. J ., Ho, C.-T., Eds.; ACS
Symposium Series 660; American Chemical Society: Wash-
ington, DC, 1997; pp 125-137.
1
also confirmed by the mass-spectral and H-NMR data
of the synthesized compound.
N,N′-diallylthiourea was reported to be one of the
major degradation products from AITC when AITC was
incubated at 37 °C for a few days (Kawakishi and
Namiki, 1969). However, there was insufficient data
in the report of Kawakishi and Namiki (1969) to support
this identification. Our study confirmed that N,N′-
diallylthiourea was the major nonvolatile compound in
the thermal degradation of AITC.
The mechanisms for the formation of N,N′-diallyl-
thiourea were proposed by Kawakishi and Namiki
Received for review J une 9, 1997. Revised manuscript received
October 20, 1997. Accepted November 5, 1997.
X
(
1969) as shown in Figure 5. AITC is hydrolyzed to
J F970488W
allylamine, which then reacts with AITC to generate
N,N′-diallylthiourea. This pathway was further con-
firmed by the reaction of allylamine and AITC for the
synthesis of N,N′-diallylthiourea.
X
Abstract published in Advance ACS Abstracts, December
15, 1997.