Structural changes of sinapic acid during alkali-induced air oxidation and the development of colored substances

Article abstract of DOI:10.1007/s11746-999-0172-6

Structural changes of sinapic acid were induced by air oxidation in aqueous solutions at pH 7-10 and followed by spectral and high-performance liquid chromatographic (HPLC) analysis. Color properties of the sinapic acid solutions were determined by taking

Full text of DOI:10.1007/s11746-999-0172-6

Structural Changes of Sinapic Acid
During Alkali-Induced Air Oxidation
and the Development of Colored Substances
R. Caia, S.D. Arntfielda,*, and J.L. Charltonb
Departments of aFood Science and bChemistry, University of Manitoba, Winnipeg, Manitoba, Canada R3T 2N2
ABSTRACT: Structural changes of sinapic acid were induced terified, and bound forms (11). Canola flour contains from 91
by air oxidation in aqueous solutions at pH 7–10 and followed
by spectral and high-performance liquid chromatographic
(HPLC) analysis. Color properties of the sinapic acid solutions
were determined by taking the transmittance spectra, calculat-
ing the Commission Internationale de l’Eclairage (CIE) 1931 tri-
stimulus values, and converting to Hunter L a b values. Reac-
tion rate constants for sinapic acid were determined by a kinetic
study based on the quantitative results from HPLC analysis.
These reactions were first order with respect to sinapic acid and
fit the appropriate equation with a coefficient of R2 > 0.97.
Sinapic acid was converted to thomasidioic acid with reaction
to 93.5% of their phenolic acid in the esterified form (9).
Sinapine, the choline ester of sinapic acid, is the major esteri-
fied phenolic acid (12–14), whereas sinapic acid represents a
high proportion of the free phenolic acid and 99% of the phe-
nolic acids released from hydrolysis of the esters in the flour
(15).
One of the major problems limiting the utilization of
canola protein is the dark-colored meal remaining after oil ex-
traction (16). During protein isolation, conditions of high pH
produce a darker protein isolate than do conditions of low pH
rate constants (k) of 8.54 × 10−6, 2.51 × 10−5, and 4.87 × 10−5 (17). Although the observations of a dark meal or protein iso-
s−1 in phosphate-boric acid buffers of pH 7, 8.5, and 10, respec-
late have been frequently reported (16,17) and more or less
tively. Similar reactions in ammonium bicarbonate buffers were
more than 10 times faster. With time, thomasidioic acid further
converted to 2,6-dimethoxy-p-benzoquinone and 6-hydroxy-
5,7-dimethoxy-2-naphthoic acid. Air oxidation of sinapic acid
aqueous solutions caused darkening of the color for the system,
with the 2,6-dimethoxy-p-benzoquinone as a major color con-
tributor.
associated with phenolics (8,9), detailed information regard-
ing the reactions responsible for the development of colored
substances is not available. Such information, however,
should indicate the mechanisms by which the colored sub-
stances develop and, therefore, is important to the oilseed
processors.
Although no relationship has ever been reported between
the structural changes of the phenolics and color properties of
canola/rapeseed protein, several researchers have observed
Paper no. J9024 in JAOCS 76, 757–764 (June 1999).
KEY WORDS: Canola phenolics, chromatographic analy-
sis, color, 2,6-dimethoxy-p-benzoquinone, 6-hydroxy-5,7- structural changes of sinapic acid in air-saturated basic con-
dimethoxy-2-naphthoic acid, reaction rate constant, sinapic
acid, thomasidioic acid.
ditions (18–21). Rubino et al. (19) reported the formation
of thomasidioic acid showing that the conversion was oxygen
dependent (20). Charlton and Lee (18) reported the for-
mation of 2,6-dimethoxy-p-benzoquinone and 6-hydroxy-
With an annual global production of more than 27 million 5,7-dimethoxy-2-naphthoic acid from sinapic acid via
metric tons (1), canola/rapeseed represents one of the most thomasidioic acid at pH 13 in the presence of oxygen. With
important oilseeds in the world (2). There is an interest in high-performance liquid chromatography (HPLC) analysis,
preparing a food-grade protein from canola meal (3–5), espe- Bouchereau et al. (21) detected new compounds derived from
cially because the amino acid content of canola meal is well rapeseed phenolics in the methanol extract of rapeseed flour,
balanced (6,7). However, the utilization of canola/rapeseed although the compounds were not identified. In a study look-
protein in human foods has been limited by the presence of ing at the effect of processing conditions on phenolic content,
antinutritional factors such as glucosinolates, phytates, and treatments, especially heating, have been shown to decrease
phenolics (8). Among them, phenolics have been the subject the sinapine content and increase the lignan amount (22). Al-
of many studies owing to their contributions to the dark color, though the oxidation of sinapic acid under alkaline conditions
bitter taste, and the astringency of rapeseed/canola meal (9) is known to lead to thomasidioic acid, 2,6-dimethoxy-p-ben-
as well as their detrimental effect on the gelation property of zoquinone and 6-hydroxy-5,7-dimethoxy-2-naphthoic acid,
canola protein (10). Canola phenolics are present in free, es- the color properties of these oxidation products have not been
determined. Reaction rate constants for these conversions
under different pH conditions also have not been determined,
*To whom correspondence should be addressed.
E-mail: arntfie@cc.umanitoba.ca
Copyright © 1999 by AOCS Press
757
JAOCS, Vol. 76, no. 6 (1999)

758
R. CAI ET AL.
which should provide useful information when considering acid (20) and the conversion of thomasidioic acid to 2,6-
protein isolation under alkaline conditions. dimethoxy-p-benzoquinone and 6-hydroxy-5,7-dimethoxy-
As part of a series of investigations looking at the contri- 2-naphthoic acid (18). Water vapor loss was 1% per day
butions of phenolics to the color property of canola protein, under the test condition. This loss was ignored during the
structural changes of sinapic acid induced by air oxidation at kinetic study. The pH values of these solutions were
pH 7–10 were followed by spectral and HPLC analysis for checked before each analysis was made. Only the pH value
10 d. Reaction rate constants were determined at pH values of the phosphate-boric acid solution at pH 10 had a slight
of 7, 8.5, and 10. Color properties of solutions were also de- deviation (decreased) and was readjusted by adding sodium
termined in order to evaluate the color contributions of the hydroxide. Spectral and HPLC analyses were carried out for
new substances to the system.
10 d. For HPLC, a sample size of 1 µL was used for the ki-
netic study whereas 25 µL was injected during the identifi-
cation of the p-benzoquinone and the 2-naphthoic acid using
a 25-µL sample loop. Three samples of pH 7, 8.5, and 10 for
MATERIALS AND METHODS
Sources of materials. Sinapic acid (3,5-dimethoxy-4-hydroxy- ultraviolet (UV) spectral analysis were prepared by a 20-
cinnamic acid) was purchased from Aldrich Chemical Co. fold dilution of the sinapic acid solution (200 µg/mL) with
(Milwaukee, WI). Acetic acid and sodium hydroxide used for phosphate-boric acid buffers of different pH values and
HPLC were verified American Chemical Society (A.C.S.) stirred in air at room temperature (22°C). For the spectral
grade and purchased from Fisher Scientific Co. (Nepean, On- and HPLC analyses of the the standard thomasidioic
tario, Canada). Other chemicals used for HPLC were HPLC- acid, 2,6-dimethoxy-p-benzoquinone and 6-hydroxy-5,7-
grade. All other chemicals, unless stated otherwise, were veri- dimethoxy-2-naphthoic acid, these compounds were dis-
fied A.C.S grade and purchased from Fisher Scientific Co.
solved directly in the appropriate buffers just before the
Thomasidioic acid was prepared according to the proce- measurement.
dure outlined by Lee (23) and Rubino et al. (19), which in-
HPLC analysis. Chromatographic equipment consisted of
volved the oxidation of sinapic acid in an ammonium bicar- two Waters (Milford, MA) pumps (models 501 and 510) and
bonate buffer (pH 8.5). The crude product was recrystallized an automated gradient controller model 680, a Shimadzu
from acetone several times to yield thomasidioic acid as a col- (Kyoto, Japan) SPD-6A UV spectrophotometric detector, and
orless solid (24). This purified thomasidioic acid was used as a Hewlett-Packard (Avondale, PA) model HP3396II integra-
a standard for identifying thomasidioic acid formed from tor. A reverse-phase C18 column (Supelcosyl, 3-µm particle
sinapic acid. 2,6-Dimethoxy-p-benzoquinone was prepared size, 33 × 4.6 mm i.d.; Supelco, Bellefonte, PA) was used.
according to the method outlined by Lee (23), which involved Buffer A was a 0.05 M acetate buffer prepared by a 1:100 di-
air oxidation of 2,6-dimethoxyphenol in aqueous acetic acid lution of a stock pH 4.7 acetate buffer. The stock buffer was
solution in the presence of chromium trioxide. 6-Hydroxy- prepared by adjusting 5 M acetic acid to pH 4.7 with solid
5,7-dimethoxy-2-naphthoic acid was also prepared using the sodium hydroxide (27). Buffer A was filtered through a 0.45
method of Lee (23) by air oxidation of sinapic acid in strongly µm filter. Buffer B was 100% HPLC-grade methanol. The
basic solution (pH 13). 2,6-Dimethoxy-p-benzoquinone is column was maintained at 37°C and run at a constant flow
reported to be yellow, whereas 6-hydroxy-5,7-dimethoxy- rate of 1.4 mL/min.
2-naphthoic acid crystallizes as pale tan needles from
MeOH/H2O (25).
The initial elution solvent was 15% methanol and 85%
buffer A. After a 12-min isocratic flow, a 2-min linear gradi-
Sample preparation. Phosphate-boric acid buffers were ent was used to change the solvent composition to 100%
prepared according to Britton and Robinson (26). Ammonium methanol. This composition was maintained for 2 min, after
bicarbonate buffers (0.12 M) were prepared according to Lee which another 2-min linear gradient returned the solvent to
(23) and adjusted to pH 8.5 and 10 using ammonium hydrox- its original composition.
ide. Experiments in this buffer system were conducted to
Peaks were identified with standard sinapic acid, thoma-
complement the phosphate-boric acid system. Results in sidioic acid, 2,6-dimethoxy-p-benzoquinone, and 6-hydroxy-
phosphate-boric acid buffers constituted the main body of 5,7-dimethoxy-2-naphthoic acid; and the HPLC response
work in this paper.
(peak area) was calibrated using standardized solutions of the
A sinapic acid solution (200 µg/mL) was prepared with above compounds.
deionized water. Three samples of 4 mL each in phosphate-
Kinetic study. Reaction rate constants (k) were obtained by
boric acid buffer of pH 7, 8.5, and 10 (two samples at pH a kinetic study based on the time-dependent change in con-
8.5 and 10 in ammonium bicarbonate buffers), respectively, centration of sinapic acid as determined from the peak areas
were prepared by combining 2 mL of the sinapic acid solu- of the HPLC chromatograms.
tion with 2 mL of each buffer solution of different pH val-
For each reaction at pH 7, 8.5, and 10, the concentration
ues so that 100 µg/mL (0.446 mmol) sinapic acid solutions of sinapic acid was determined about 10 times by HPLC dur-
with different pH values were obtained. These solutions ing the course of the reaction. Each of the three groups of
were stirred in air at room temperature (22°C). Oxygen is time-dependent reactant concentrations at different pH values
needed for the conversion of sinapic acid to thomasidioic was fitted to a first-order reaction equation,
JAOCS, Vol. 76, no. 6 (1999)

STRUCTURAL CHANGES AND COLORATION OF SINAPIC ACID
759
co
c
TABLE 1
ln =
= kt
[1]
Concentration of Sinapic Acid, Thomasidioic Acid,
2,6-Dimethoxy-p-benzoquinone, and 6-Hydroxy-5,7-dimethoxy-2-
naphthoic Acid as a Function of Time During Alkaline Oxidation
of a 0.446 mmol/L Sinapic Acid Solution in Phosphate-Boric Acid
Buffers of pH 7, 8.5, and 10
and a second-order equation,
1
1
−
= kt
[2]
c
co
pH 7
pH 8.5
pH 10
where co and c were the initial concentration (mol/L) and the
concentration at time t, t is the reaction time (s), and k is the
reaction rate constant; s−1 in the first-order equation, and
L mol−1 s−1 in the second-order equation.
Conc. (mmol/L)
Conc. (mmol/L)
Conc. (mmol/L)
Time
(h)
Time
(h)
Time
(h)
SAa
TA
SAa
TA
SAa
TA
0
5
10
16
23
27
34
51
55
0.446
0.349
0
N.D.
0
1
3
5
7
0.446
0
0
0.446
0
0.351 0.014
0.310 0.033
0.251 0.076
0.229 0.077
0.5 0.374
0.026
0.069
0.113
0.140
0.165
0.174
0.180
0.189
0.196
0.041
0.003
If the reaction is first order, the plot of ln (1/c) vs. t should
give a straight line. Similarly, if the reaction is second order,
the plot of 1/c vs. t should give a straight line. A linear regres-
sion was conducted for each group of sinapic acid concentra-
tions. R2 (R = correlation coefficient) was used to judge the
linear relationship between the independent value t and the
dependent values ln (1/c) or 1/c and therefore to determine
which equation better fit the data (28). The equation better
able to fit the data was used to calculate the reaction rate con-
stant k.
0.316 0.017
0.259 0.037
0.225 0.060
0.198 0.066
0.142 0.074
0.085 0.090
0.090 0.094
0.007 0.143
2
4
5
7
0.284
0.216
0.176
0.113
11 0.179 0.088
16 0.092 0.112
20 0.063 0.145
24 0.045 0.168
26 0.041 0.171
10 0.045
15 0.030
20 0.012
132
169
240
24 0.007
0.001 0.184 169 N.D.
N.D. 0.141 241 N.D.
0.136 168 N.D.
0.078 241 N.D.
BQ
NA
BQ
NA
BQ
NA
Spectral analysis and color determination. Spectral analy-
sis and color determination have been described elsewhere
(29).
133
169
240
N.D. N.D.
26
N.D.
N.D.
24 N.D.
N.D.
0.132
0.144
0.008 0.017 169 0.011 0.072 168 0.044
0.013 0.030 240 0.044 0.125 241 0.043
aReaction rate constant k value as in Table 2. SA, sinapic acid; TA, thoma-
sidioic acid; BQ, 2,6-dimethoxy-p-benzoquinone; NA, 6-hydroxy-5,7-
dimethoxy-2-naphthoic acid; N.D., not detected.
RESULTS AND DISCUSSION
Structural changes. (i) HPLC analysis. The concentrations
for sinapic acid, thomasidioic acid, 2,6-dimethoxy-p-benzo-
quinone, and 6-hydroxy-5,7-dimethoxy-2-naphthoic acid as a naphthoic acid via thomasidioic acid at high pH has been pre-
function of time during air oxidation of three 0.466 mmol/L viously reported (18). In this study, these two oxidation prod-
sinapic acid in phosphate-boric acid buffer solutions of pH 7, ucts were found in samples at all three pH values (7, 8.5, and
8.5, and 10 are given in Table 1. Virtually all the sinapic acid 10) after 169 h. The fact that the concentrations of 2,6-
was converted to thomasidioic acid during the alkali-induced dimethoxy-p-benzoquinone and 6-hydroxy-5,7-dimethoxy-2-
air oxidation at pH 7, 8.5, or 10. With the reaction at pH 7, naphthoic acid increased while that of thomasidioic acid de-
approximately 50% conversion to thomasidioic acid occurred creased was consistent with the known oxidation of thoma-
after 23 h and 99.8% after 169 h, resulting in the production sidioic acid to 6-hydroxy-5,7-dimethoxy-2-naphthoic acid
of 0.184 mmol/L of thomasidioic acid. However, between (18).
169 and 240 h the thomasidioic acid concentration decreased
The amounts of 2,6-dimethoxy-p-benzoquinone and 6-hy-
to 0.141 mmol/L while those of the 2,6-dimethoxy-p-benzo- droxy-5,7-dimethoxy-2-naphthoic acid were determined to
quinone and 6-hydroxy-5,7-dimethoxy-2-naphthoic acid in- be 0.013 and 0.030, 0.044 and 0.125, and 0.043 and 0.144
creased from 0.008 and 0.017 mmol/L to 0.013 and 0.030 mmol/L, a mole ratio of 1:2.3, 1:2.8, and 1:3.3 for pH 7, 8.5,
mmol/L, respectively. At pH 8.5, there was 91% conversion and 10, respectively, after air oxidation for approximately
to thomasidioic acid after 26 h, while at pH 10, there was 90% 240 h. This ratio is less than the theoretical mole ratio of 1:1
conversion to thomasidioic acid after only 10 h. The conver- that would result if equal number of moles of the two com-
sion was nearly complete (98.4%) within 24 h. The concen- pounds were produced as the only reaction products from
trations of thomasidioic acid peaked at 0.171 mmol/L after thomasidioic acid.
26 h at pH 8.5 and at 0.196 mmol/L after 24 h at pH 10. At
Similar reactions were found in ammonium bicarbonate
both pH levels, increased oxidation (up to 241 h) resulted in buffers at pH 8.5 and pH 10, but the reaction rates were more
decreases in thomasidioic acid concentration that were ac- than 10 times faster.
companied by increases in the levels of 2,6-dimethoxy-p-ben-
(ii) UV spectrum. The UV spectra of 10 µg/mL sinapic
zoquinone and 6-hydroxy-5,7-dimethoxy-2-naphthoic acid. acid solutions after air oxidation for 0 and 24 h are shown in
Structural changes of sinapic acid were apparently similar for Figure 1A and B. The UV spectra of 10 µg/mL standard
all three pH conditions, but the conversions were slower at thomasidioic acid solutions of different pH values are given
lower pH.
in Figure 1C. At zero time, increasing pH caused a strong
The air oxidative conversion of sinapic acid to 2,6- bathochromic shift, the shift of the maximum absorbance to-
dimethoxy-p-benzoquinone and 6-hydroxy-5,7-dimethoxy-2- ward higher wavelengths, relative to the spectrum recorded at
JAOCS, Vol. 76, no. 6 (1999)

760
R. CAI ET AL.
boxylic and the hydroxy groups are ionized (anion II)
(Scheme 1). The strongest bathochromic shifts occur on ion-
ization of the phenolic hydroxyl. After 24 h (Fig. 1B), inter-
pretation of the UV spectrum found that the absorbances of
sinapic acid of different anionic forms (anion I, 305 nm, and
anion II, 355 nm) had decreased as expected. The UV ab-
sorbances of the solutions at pH 8.5 and 10 in fact were simi-
lar to the spectra of the oxidized product, thomasidioic acid
(Fig. 1B,C). Unlike the situations at pH 8.5 and 10, the spec-
trum obtained after air oxidation at pH 7 for 24 h did not re-
semble that of the standard thomasidioic acid, since only 50%
sinapic acid was converted to thomasidioic acid.
Kinetic study of the disappearance of sinapic acid. The
correlation coefficients (R2) for the first-order equations were
0.98, 0.99, and 0.99, respectively, for the reactions in phos-
FIG. 1. Ultraviolet spectra of a 10 µg/mL phosphate-boric acid buffer solu-
tion of pH 7, 8.5, and 10 containing (A) sinapic acid after 0 h, (B) sinapic
acid after 24 h, and (C) standard thomasidioic acid.
pH 7 (30). The effect of alkaline conditions on the absorption
properties of phenolic compounds has been reported by sev-
eral researchers (31,32). Large bathochromic shifts were
noted in most cases under basic conditions (32). The pKa val-
ues of the carboxylic and hydroxy groups of sinapic acid are
4.47 and 9.21, respectively (33). Therefore, at pH 7 only the
carboxylic group is ionized (anion I). At pH 10, both the car-
SCHEME 1
JAOCS, Vol. 76, no. 6 (1999)

STRUCTURAL CHANGES AND COLORATION OF SINAPIC ACID
761
TABLE 2
First-Order Reaction Equations, Reaction Rate Constants (k), Half-Lives (t1/2),
and Correlation Coefficients (R2) for the Reactions of Sinapic Acid in Solutions
of pH 7, 8.5, and 10
pH
Equation
k (s−1
)
t1/2 (h)a
R2
Phosphate-boric acid buffer
7
8.5
ln (co/c) = 0.0375 + 8.54 × 10−6
t
t
8.54 × 10−6
2.51 × 10−5
4.87 × 10−5
22.50
7.66
3.95
0.98
0.99
0.99
ln (co/c) = 0.075 + 2.51 × 10−5
ln (co/c) = 0.0828 + 4.87 × 10−5
t
10
Ammonium bicarbonate buffer
8.5
10
ln (co/c) = 0.0878 + 4.30 × 10−4
t
4.30 × 10−4
6.18 × 10−4
0.45
0.31
0.99
0.97
ln (co/c) = −0.1765 + 6.18 × 10−4
t
aThe half-life (t1/2) for the reactions was found as t1/2 = 0.693/k, where k is the reaction rate constant.
phate-boric acid buffers of pH 7, 8.5, and 10, whereas R2 for (pH 10) of sinapic acid was converted to thomasidioic acid
the second-order equations were 0.93, 0.65, and 0.85, respec- within 28 h. Therefore, yellow intensity at this time should re-
tively (Table 2). The first-order equations were therefore used flect the color of the phenolate anions of thomasidioic acid.
to calculate the reaction rate constant (k) values.
Moreover, the conversion to thomasidioic acid from sinapic
The reaction equations, reaction constants (k), and half- acid may involve the formation of some intermediates, which
lives (t1/2) for both the phosphate-boric acid buffer and the may affect the color. It seemed apparent that at least one inter-
ammonium bicarbonate buffer are also summarized in mediate between sinapic acid and thomasidioic acid must ac-
Table 2. In phosphate-boric acid buffers, the reaction rate count for the color increase at 18 h. Since 2,6-dimethoxy-p-
constants are 8.54 × 10−6, 2.51 × 10−5, and 4.87 × 10−5 s−1 for benzoquinone and 6-hydroxy-5,7-dimethoxy-2-naphthoic acid
reactions at pH 7, 8.5, and 10, respectively. The reaction at were present when the color darkening was greatest at the end
pH 10 was about twice as fast as the reaction at pH 8.5 and of the reaction, these substances could be major contributors
about five times as fast as the reaction at pH 7. These differ- to color of the system.
ences can also be seen from the half-lives (t1/2) of the reac-
Color profiles in ammonium bicarbonate buffers at pH 8.5
tions, which were 22.50, 7.66, and 3.95 h for pH 7, 8.5, and and 10 were found similar to those found in phosphate-boric
10, respectively. The reactions in ammonium bicarbonate acid buffers
buffers were more than 10 times faster than those in the phos-
(ii) Color property of sinapic acid and thomasidioic acid
phate-boric acid buffer, with reaction rate constants (k) of in basic conditions. The color properties of sinapic acid and
4.30 × 10−4 and 6.18 × 10−4 s−1 at pH 8.5 and 10, respectively. thomasidioic acid in phosphate-boric buffers were determined
This demonstrates the possible effect of different media on
the reaction. Reaction orders in ammonium bicarbonate
buffers, however, were the same as those in phosphate-boric
acid buffers.
TABLE 3
Effect of Time on the Hunter L a b Values of 0.446 mmol/L Sinapic
Acid Solutions During Oxidations in Phosphate-Boric Acid Buffers
Color changes in relation to structural changes. (i) Color
intensity, Hunter L a b values. The effects of time on the
Hunter L a b values of the reaction mixtures (phosphate-boric
acid buffers) at pH 7, 8.5, and 10 are shown in Table 3. Ini-
tially the color of the reaction mixture at pH 7 was almost un-
changed compared to the control. However, this color inten-
sity steadily increased with time as seen by a decrease in the L
value and an increase in the magnitude of both a and b values
within 169 h. For reactions at pH 8.5 and 10, yellow intensity
(b) was higher than that of the control at the beginning of the
reaction. This yellow color increased during the initial stage
of the reaction (18 h), decreased slightly at 28 h, but became
most intense by 169 h. At this time, both 2,6-dimethoxy-p-
benzoquinone and 6-hydroxy-5,7-dimethoxy-2-naphthoic acid
were found to be present. The initial high-intensity yellow of
the solution could be attributed to the absorption of the sinapic
acid phenolate anion (anion II), as seen from the UV spectra
(Fig.1A). This color increased and then faded slightly during
the course of the conversion of sinapic acid to thomasidioic
acid. According to the HPLC results, 92% (pH 8.5) and 99%
of pH 7, 8.5, and 10a
Hunter color value
a
Time (h)
L
b
Controlb
pH 7
0
18
28
169
pH 8.5
0
18
28
169
pH 10
0
18
28
169
99.77 0.15
−5.55 0.31
5.47 0.25
99.18 0.71
96.84 0.54
96.33 0.38
88.68 0.91
−5.22 0.17
−5.92 0.37
−4.47 0.29
−7.72 0.39
5.85 0.48
12.52 0.52
15.53 0.06
19.46 0.58
98.40 0.53
92.32 1.16
91.00 2.38
88.40 1.28
−5.80 0.17
−6.46 0.36
−8.27 1.45
−5.69 0.45
7.70 0.53
14.23 0.57
12.84 1.16
19.08 0.39
99.43 0.81
93.49 1.32
97.00 1.1
86.97 0.91
−7.90 0.61
−6.49 0.36
−6.25 0.27
−2.42 0.33
11.47 0.59
12.49 0.5
11.69 1.1
17.9 0.98
aMean of three replicates SD.
bControl: sinapic acid in deionized water with a natural pH of 4.3 at zero
time.
JAOCS, Vol. 76, no. 6 (1999)

762
R. CAI ET AL.
based on visible light transmittance spectra. A comparison
(iii) Spectral properties of 6-hydroxy-5,7-dimethoxy-2-
between transmittance spectra of equal weight concentrations naphthoic acid and 2,6-dimethoxy-p-benzoquinone. UV and
of sinapic acid and thomasidioic acid at different pH values is light transmittance spectra of the 6-hydroxy-5,7-dimethoxy-
shown in Figure 2. Both compounds showed little color at pH 2-naphthoic acid at different pH values are shown in Figure 3.
7 and a remarkable decrease in percentage transmittance at At pH 7, the 2-naphthoic acid showed two major peaks at
low wavelengths as pH increased from 7 to 10. These agreed around 250 and 300 nm (Fig. 3A). Higher pH caused a sig-
with the UV spectra where a large absorption band was found nificant shift of the peaks toward higher wavelengths (around
for both compounds in the visible region. However, the trans- 255 and 330 nm at pH 10). However, these shifts did not
mittance decreases for thomasidioic acid at pH 8.5 and 10 cause a large change in the visual transmittance spectra (Fig.
were less pronounced than those for sinapic acid, indicating 3B). At pH 7 the 2-naphthoic acid showed little color. The
that thomasidioic acid was less colored than sinapic acid at color intensity only had a slight increase as pH increased from
pH 8.5 and 10.
7 to 10, indicating it was not a significant color contributor
As the color properties of sinapic acid and thomasidioic even under basic conditions. 2-Naphthoic acid occurs as pale
acid are pH dependent, the color determined using basic con- tan needles (25).
ditions will be different from those determined at neutral or
A comparison of UV and light transmittance spectra of
acidic conditions. Similarly, during protein isolation from the 2,6-dimethoxy-p-benzoquinone is shown in Figure 4. In
canola meal with basic extraction and acidic precipitation, the the UV spectra (Fig. 4A), 2,6-dimethoxy-p-benzoquinone
colored sinapic or thomasidioic anions at basic conditions can showed a maximal absorbance at 280 nm in contrast to the
become colorless during acidic precipitation provided no two maxima of sinapic acid at 230 and 325 nm. Although
other interactions between protein and phenolics occur. How- sinapic acid showed no absorbance at wavelengths above 380
ever, the secondary oxidation products such as the 2,6- nm, p-benzoquinone showed some absorbance over this re-
dimethoxy-p-benzoquinone should remain as colored sub- gion, indicative of a colored substance. In the light transmit-
stances if these substances are present.
tance spectra (Fig. 4B), 2,6-dimethoxy-p-benzoquinone
FIG. 2. Transmittance spectra of a 100 µg/mL phosphate-boric buffer solu-
tion of pH 7, 8.5, and 10 containing (A) sinapic acid and (B) thomasidioic
acid.
FIG. 3. Ultraviolet absorbance (A) (solution diluted 10 times) and visual light
transmittance (B) spectra of a 100 µg/mL 6-hydroxy-5,7-dimethoxy-2-naph-
thoic acid.
JAOCS, Vol. 76, no. 6 (1999)

STRUCTURAL CHANGES AND COLORATION OF SINAPIC ACID
763
quinone and the 6-hydroxy-5,7-dimethoxy-2-naphthoic acid
was determined to involve the secondary oxidation of the first
formed thomasidioic acid (18). If this is the case, then the two
intermediates, the bisquinone methide and the monoquinone
methide, may also affect the color profile of the system. These
intermediates may account for the color development during
the conversion of sinapic acid to thomasidioic acid.
With time, thomasidioic acid gradually oxidized to form
the p-benzoquinone and the 2-naphthoic acid, which in turn
increased the yellow intensity, with the p-benzoquinone being
the major color contributor. This is also the case for the reac-
tion at pH 7. The steady increase in color intensity is consis-
tent with air oxidation of sinapic acid to form thomasidioic
acid and further oxidation to the p-benzoquinone and the 2-
naphthoic acid. The darkening of color at pH 7 is particularly
noteworthy since the solution had a neutral pH value and there
was no significant sign of color changes at the onset of the re-
action, unlike the more basic pH conditions, which produced
an immediate yellow color. A reaction scheme showing the
conversion of sinapic acid to thomasidioic acid, and further
oxidation to 2,6-dimethoxy-p-benzoquinone and 6-hydroxy-
5,7-dimethoxy-2-naphthoic acid is given in Scheme 1.
ACKNOWLEDGMENT
Financial support of this research by the Natural Sciences and Engi-
neering Research Council of Canada and the award of the J.W. Bar-
low Graduate Fellowship to R. Cai are gratefully acknowledged.
REFERENCES
FIG. 4. Ultraviolet absorbance (A) (solution diluted 10–20 times) and visual
light transmittance (B) spectra of a 100 µg/mL 2,6-dimethoxy-p-benzo-
quinone and a 100 µg/mL sinapic acid (all in methanol).
1. Eskin, N.A.M., B.E. McDonald, R. Przybylski, L.J. Malcolm-
son, R. Scarth, T. Mag, K. Ward, and D. Adolph, Canola Oil,
offprint from Bailey’s Industrial Oil and Fat Products, Vol. 2,
edited by Y.H. Hui, John Wiley & Sons, Inc., New York, 1996,
pp. 1–95.
2. Shahidi, F., Canola and Rapeseed: Production, Chemistry, Nu-
trition and Processing Technology, Van Nostrand Reinhold,
New York, 1990, pp. 3–13.
3. Ismond, M.A.H., and W.D. Welsh, Application of New Method-
ology to Canola Protein Isolation, Food Chem. 45:125–127
(1992).
4. Gillberg, L., and B. Tornell, Preparation of Rapeseed Protein
Isolate. Dissolution and Precipitation Behavior of Rapeseed Pro-
teins, J. Food Sci. 41:1063–1069 (1976).
5. Owen, D.F., C.O. Chichester, J. Granadino, and F. Monckeberg,
A Process for Producing Nontoxic Rapeseed Protein Isolate and
an Acceptable Feed By-product, Cereal Chem. 48:91–96
(1971).
6. Liu, R.F.K., L.U. Thompson, and J.D. Jones, Yield and Nutri-
tive Value of Rapeseed Protein Concentrate, J. Food Sci.
47:977–981 (1982).
7. Ohlson, R., and K. Anjou, Rapeseed Protein Products, J. Am.
Oil Chem. Soc. 56:431–437 (1976).
8. Shahidi, F., and M. Naczk, Food Phenolics: Sources, Chemistry,
Effects and Applications, Technomic Publishing Company, Inc.,
Lancaster, 1995, pp. 53–73.
showed a noticeable low transmittance in the wavelength
range from 380 to 480 nm. This is consistent with the yellow
appearance of that substance (25). This also indicated that p-
benzoquinone is a strong contributor to the color of the sys-
tem. HPLC analysis found other undetermined colored sub-
stances were produced during the reaction.
Mechanisms of the reactions. Results from this research
showed that after 28 h there was 92% (pH 8.5) and 99% (pH
10) disappearance of sinapic acid in phosphate-boric acid
buffers (0.446 mmol/L). Bathochromic shifts at the beginning
of the reaction were caused by the formation of phenolate an-
ions, which could be responsible for the initial colors of the
mixtures (31–33). The color increased and then slightly de-
creased in the first 28 h at pH 8.5 and pH 10 during the course
of the conversion of sinapic acid to thomasidioic acid. Simi-
lar color changes were found in reactions in ammonium bi-
carbonate buffers.
Lee (23) determined that the dimerization of sinapyl radi-
cals was the most likely pathway for oxidation of sinapic acid.
The dimerization product, a bisquinone methide, would then
partially tautomerize to form a monoquinone methide and
then cyclize to form thomasidioic acid. The mechanism for
the conversion of sinapic acid to the 2,6-dimethoxy-p-benzo-
9. Shahidi, F., and M. Naczk, An Overview of Phenolics of Canola
and Rapeseed: Chemical, Sensory and Nutritional Significance,
J. Am. Oil Chem. Soc. 69:917–924 (1992).
10. Rubino, M.I., S.D. Arntfield, C.A. Nadon, and A. Bernatsky,
Phenolic Protein Interactions in Relation to the Gelation Proper-
ties of Canola Protein, Food Res. Int. 29:653–659 (1996).
JAOCS, Vol. 76, no. 6 (1999)

764
R. CAI ET AL.
11. Krygier, K., F. Sosulski, and L. Hogge, Free, Esterified and In-
soluble-Bound Phenolic Acid. I. Extraction and Purification Pro-
cedure, J. Agric. Food Chem. 30:330–333 (1982).
12. Blair, R., and R.D. Reichert, Carbohydrate and Phenolic Con-
stituents in a Comprehensive Range of Rapeseed and Canola
tion and Processing Technology, edited by F. Shahidi, Van Nos-
trand Reinhold, New York, 1990, pp. 331–343.
23. Lee, K.S., Oxidative Coupling of Sinapic Acid, MSc. Thesis,
University of Manitoba, Winnipeg, Manitoba, Canada, 1997, pp.
94–118.
Fractions: Nutritional Significance for Animals, J. Sci. Food 24. Ahmed, R., M. Leherer, and R. Stevenson, Synthesis of Thoma-
Agric. 35:29–35 (1984). sic Acid, Tetrahedron 29:3753–3759 (1973).
13. Fenton, T.W., T. Leung, and D.R. Clandinin, Phenolic Compo- 25. Hostettler, F.D., and M.K. Seikel, Lignans of Ulmus thomasii
nents of Rapeseed Meal, J. Food Sci. 45:1702–1705 (1980).
14. Austin, F.L., and I.A. Wolff, Sinapine and Related Esters in
Heartwood. II. Lignans Related to Thomasic Acid, Ibid.
25:2325–2337 (1969).
Seed Meal of Crambe abyssinica, J. Agric. Food Chem. 26. Britton, H.T.S., and R.A. Robinson, The Use of Antimony-An-
16:132–135 (1968).
timonous Oxide Electrode in the Determination of the Concen-
tration of the Hydrogen Ions and in Potentiometric Titration.
The Prodeaux-Ward Universal Buffer Mixture, J. Chem. Soc.
132:458–473 (1931).
15. Krygier, K., F. Sosulski, and L. Hogge, Free, Esterified and In-
soluble Phenolic Acids. II. Composition of Phenolic Acids in
Rapeseed Flour and Hulls, Ibid. 30:334–336 (1982).
16. Youngs, C.G., Technical Status Assessment of Food Protein 27. Hagerman, A.E., and R.L. Nicholson, High Performance Liquid
from Canola, in Research on Canola Seed, Oil and Meal, 9th
Project Report, Canola Council of Canada, Winnipeg, Canada,
1991, pp. 309–341.
Chromatographic Determination of Hydroxycinnamic Acids in
the Maize Mesocotyl, J. Agric. Food Chem. 30:1098–1102
(1982).
17. Keshavarz, E., R.K.M. Cheung, R.C.M. Liu, and S. Nakai, 28. Barrow, G.M., Physical Chemistry, McGraw-Hill, New York,
Adaptation of the Three Stage Extraction Process to Rapeseed 4th edn., 1979, pp. 657–666.
Meal for Preparation of Colorless Protein Extracts, Can. Inst. 29. Cai, R., S.D. Arntfield, and J.L. Charlton, Structural Changes of
Food Sci. Technol. J. 10:73–77 (1977).
Sinapic Acid and Sinapine During Autoclaving with Respect to
the Development of Colored Substances, J. Am. Oil Chem. Soc.
76:433–441 (1999).
18. Charlton, J.L., and K.-A. Lee, Thomasidioic Acid and 6-Hy-
droxy-5,7-dimethoxy-2-naphthoic Acid: Are They Really Nat-
ural Products? Tetrahedron Lett. 38:7311–7312 (1997).
19. Rubino, M.I., S.D. Arntfield, and J.L. Charlton, Conversion of
Phenolics to Lignans: Sinapic Acid to Thomasidioic Acid, J.
Am. Oil Chem. Soc. 72:1465–1470 (1995).
30. Hendrickson, J.B., D.J. Cram, and G.S. Hammond, Organic
Chemistry, 3rd edn., McGraw-Hill, New York, 1970, pp.
246–256.
31. Ribereau-Gayon, P., Plant Phenolics, Oliver & Boyd, Edin-
burgh, 1972, pp. 97–99.
20. Rubino, M.I., S.D. Arntfield, and J.L. Charlton, Evaluation of
Alkaline Conversion of Sinapic Acid to Thomasidioic Acid, J. 32. Harborne, J.B., Ultraviolet Spectroscopy of Polyphenols, Meth-
Agric. Food Chem. 44:1399–1402 (1996).
21. Bouchereau, A., J. Hamelin, M. Renard, and F. Larher, Struc-
ods in Polyphenolics, edited by J.B. Pridham, The Macmillan
Company, New York, 1964, pp. 13–36.
tural Changes in Sinapic Acid Conjugates During Seedling De- 33. Smyk, B., and R. Drabent, Spectroscopic Investigation of the
velopment of Rape, Plant Physiol. Biochem. 30:467–475
(1992).
Equilibria of the Ionic Forms of Sinapic Acid, Analyst
114:723–726 (1989).
22. Jensen, S.K., H.S. Olsen, and H. Sorensen, Aqueous Enzymatic
Processing of Rapeseed for Production of High Quality Prod-
ucts, in Canola and Rapeseed: Production, Chemistry, Nutri-
[Received October 2, 1998; accepted March 7, 1999]
JAOCS, Vol. 76, no. 6 (1999)