Reactivity of ferulic acid and its derivatives toward hydrogen peroxide and peracetic acid

Article abstract of DOI:10.1021/jf9902494

The reactions of ferulic acid and its derivatives with hydrogen peroxide and peracetic acid in lignin-retaining bleaching conditions have been investigated to determine their susceptibility to oxidative degradation. The conjugated side chain of ferulic acid and its etherified or esterified derivative was shown to be fairly stable, especially to hydrogen peroxide. The major reaction was trans-cis isomerization that possibly involved a radical mechanism but did not cause bond cleavage. The peracetic acid reaction increased the rate of trans-cis isomerization and was also accompanied by a minor cleavage of the side chain. Esterification did not have a substantial effect on the reactivity of ferulic acid, but 4-O- etherification significantly stabilized it against these two oxidants. By contrast, aldehyde substitution tremendously enhanced the susceptibility of the cinnamyl side chain to oxidative degradation, as evidenced by an intensive degradation of coniferaldehyde.

Full text of DOI:10.1021/jf9902494

J. Agric. Food Chem. 1999, 47, 3325−3331
3325
Rea ctivity of F er u lic Acid a n d Its Der iva tives tow a r d Hyd r ogen
P er oxid e a n d P er a cetic Acid
George X. Pan,* Liam Spencer, and Gordon J . Leary
Alberta Research Council, 250 Karl Clark Street, Edmonton, Alberta, Canada T6N 1E4
The reactions of ferulic acid and its derivatives with hydrogen peroxide and peracetic acid in lignin-
retaining bleaching conditions have been investigated to determine their susceptibility to oxidative
degradation. The conjugated side chain of ferulic acid and its etherified or esterified derivative was
shown to be fairly stable, especially to hydrogen peroxide. The major reaction was trans-cis
isomerization that possibly involved a radical mechanism but did not cause bond cleavage. The
peracetic acid reaction increased the rate of trans-cis isomerization and was also accompanied by
a minor cleavage of the side chain. Esterification did not have a substantial effect on the reactivity
of ferulic acid, but 4-O-etherification significantly stabilized it against these two oxidants. By contrast,
aldehyde substitution tremendously enhanced the susceptibility of the cinnamyl side chain to
oxidative degradation, as evidenced by an intensive degradation of coniferaldehyde.
Keyw or d s: Hydrogen peroxide; peracetic acid; oxidation; ferulic acid; ferulate; coniferaldehyde
INTRODUCTION
use of TAED-activated hydrogen peroxide allows the
attainment of enhanced bleaching for laundry (J ames
and MacKirdy, 1990; Grime and Clauss, 1990) and also
paper-making pulps (Turner and Mathews, 1998).
In lignin-retaining bleaching of paper-making pulps,
nucleophiles preferentially attack conjugated carbon-
carbon double bonds and carbonyl groups such as those
found in cinnamaldehyde and quinonoid structures but
not electron-rich aromatic rings (Dence, 1994a). Con-
sequently, chromophores are removed with little dis-
solution of lignin. The mechanism of hydrogen peroxide
bleaching involves nucleophilic attack by perhydroxyl
anion (HOO-). The pKa for its formation is 11.6 at 25
°C (Legrini et al., 1993). In practice, to promote hydro-
gen peroxide bleaching, the addition of sufficient alkali
is required for formation of an adequate concentration
of HOO- ions. Since the pKa of peracetic acid is 8.2
(Plesnicar, 1978), its bleaching reaction can be per-
formed at a lower pH. Under neutral or moderately
alkaline conditions, peracetate ion, CH3COOO-, like the
perhydroxyl ion (HOO-), is a strong nucleophile, leading
to selective removal of chromophores.
In the present investigation, ferulic acid and its
related model compounds were reacted with hydrogen
peroxide and peracetic acid under typical lignin-retain-
ing bleaching conditions. The reaction substrates were
selected to assess the effect of 4-O-etherification and
esterification on the reactivity of ferulic acid. The
reactions were followed spectrophotometrically, and
reaction products were determined quantitatively and
qualitatively. Analysis of the disappearance rate of the
substrates and identification of reaction products allow
the assessment of the reactivity of the side chain of
ferulic acid toward hydrogen peroxide and peracetic
acid.
Our previous article reported that only a small
amount of ferulic and p-coumaric acid was removed from
wheat straw by hydrogen peroxide treatment under
practical bleaching condition (Pan et al., 1998). The
present study was aimed at further assessing the
reactivity of these phenolic acids toward hydrogen
peroxide and a stronger oxidant, peracetic acid.
Hydrogen peroxide and peracetic acid as chlorine-free
bleaching agents have become increasingly important
in today’s bleaching technology for the pulp and paper
industry. Both of the oxidants are versatile and can be
used in electrophilic or nucleophilic reactions of organic
compounds depending upon the reaction media chosen
(Dence, 1994a). Peracetic acid may be regarded as an
activated derivative of hydrogen peroxide, HOOH, in
which one hydrogen atom is replaced by an acyl group.
The active oxygen of hydrogen peroxide, which is not
readily available for oxidation of organic compounds, can
be transformed to the active peroxyacid form by direct
oxidation of acetic acid with hydrogen peroxide in the
presence of mineral acid as a catalyst (Plesnicar, 1978),
as described in the equation shown below. As a result,
+
H
RCO2H + H2O2 y z RCO3H + H2O
the higher oxidizing power of peracetic acid allows it to
react with organic substrates to a greater extent or to
be used in oxidation reactions under more moderate
conditions than hydrogen peroxide. Peracetic acid has
been well documented as an epoxidizing/hydroxylating
agent for olefinic bonds in organic chemistry (Plesnicar,
1978). Alternatively, peracetic acid can be generated in
situ by adding an activator such as tetraacetylethyl-
enediamine (TAED) in alkaline hydrogen peroxide solu-
tion (Davies and Deary, 1991). It is reported that the
EXPERIMENTAL PROCEDURES
Ma ter ia ls. The model compounds studied in this investiga-
tion were predominantly the trans isomers and are shown in
Figure 1. Ferulic acid (I), ethyl ferulate (II), coniferaldehyde
* To whom correspondence should be addressed. E-mail:
Pang@arc.ab.ca. Phone: (780) 450-5027. Fax: (780) 450-5397.
10.1021/jf9902494 CCC: $18.00 © 1999 American Chemical Society
Published on Web 07/14/1999

3326 J. Agric. Food Chem., Vol. 47, No. 8, 1999
Pan et al.
F igu r e 1. Model compounds studied: ferulic acid (I), ethyl
ferulate (II), coniferaldehyde (III), and 3,4-dimethoxycinnamic
acid (IV).
F igu r e 2. UV spectral changes during the reaction of ferulic
acid (I) with hydrogen peroxide (P) after 4 h of reaction time.
(III), 3,4-dimethoxycinnamic acid (IV), and acetovanillone
were purchased from Aldrich Chemical Co. and used without
further purification. Other reagents, including 30% hydrogen
peroxide and 32% equilibrium peracetic acid solutions, were
commercial products and employed as received.
Rea ction of Mod el Com p ou n d s w ith Hyd r ogen P er -
oxid e a n d P er a cetic Acid . Each oxidation was performed
in a 50 mL three-necked flask capped with rubber septa at 50
°C in an oil bath. Each run employed approximately 0.5 mmol
of substrate and 2.0 mmol of oxidant. In the case of hydrogen
peroxide, the model compounds were dissolved in 30 mL of
pH 11.2 phosphate buffer solution containing 4 mg of metal
chelating agent (DTPMPA). After the materials were com-
pletely dissolved, the required amount of hydrogen peroxide
was added through a syringe. For the peracetic acid reactions,
the model compounds were dissolved in 15 mL of ethanol and
then mixed with 15 mL of aqueous peracetic acid solution
preadjusted to pH 7 with NaHCO3 solution. In the experiments
under nitrogen or oxygen, the gas was continuously passed
into the reaction mixture and vented through a needle. In the
case of nitrogen, the substrate solution had been purged with
nitrogen for 10 min before addition of hydrogen peroxide or
peracetic acid. During the 4 h reaction period, the reaction
mixture was shaken occasionally and aliquots (1 mL) of the
reaction solution were withdrawn with a syringe at chosen
intervals for UV-vis and GC analyses, respectively.
Wor k u p P r oced u r e of Rea ction Solu tion s. Duplicate
samples of the reaction mixture were immediately transferred
to vials containing 1 mL of 5% sodium metabisulfite solution
in order to quench any remaining hydrogen peroxide or
peracetic acid and to bring the sample solution to pH < 7. To
prepare samples for GC analysis, the sample in the vial was
diluted with 20 mL of 50% aqueous methanol solution and
acidified to pH 2.5 with 10% HCl solution. After 2.0 mg of
acetovanillone (internal standard) was added, the mixture was
extracted with 3 × 20 mL ethyl acetate. The combined ethyl
acetate extracts were dried over anhydrous Na2SO4 and
evaporated to dryness under reduced pressure.
F igu r e 3. UV spectral changes during the reaction of ferulic
acid (I) with peracetic acid (PA) after 4 h of reaction time.
Ma ss Sp ectr om etr y. GC-MS analyses were conducted on
a Micromass AutoSpec instrument using the same column and
temperature program as in the GC analyses. The electron
energy used was 70 eV.
RESULTS AND DISCUSSION
Sp ectr a l Ch a n ges. In the present investigation,
reactions were periodically monitored by UV-vis spec-
troscopy. This allowed a quick assessment of the reac-
tivity of substrates toward hydrogen peroxide and
peracetic acid in different reaction media. Figure 2
shows that ferulic acid (I) was only slightly reactive to
hydrogen peroxide. The presence of oxygen during the
reactions did not affect the consumption of the starting
material. On the other hand, the peracetic acid reaction
resulted in a pronounced change in UV absorption, as
revealed by Figure 3. It appeared that peracetic acid
not only destroyed ferulic acid (I) but also produced
products that had absorption maxima at lower wave-
lengths. In contrast to the observation in the hydrogen
peroxide reaction, excluding oxygen reduced the rate of
consumption of the substrate by peracetic acid.
GC An a lysis. The dried extracts of oxidation products were
silylated with 0.5 mL of pyridine and 0.25 mL of bis(trimeth-
ylsilyl)trifluoroacetamide (BSTFA) at 70 °C for 15 min. Tri-
methylsilylated (TMS) derivatives of the products were sepa-
rated on a HP 5890 gas chromatograph (splitless injection)
equipped with a BPX70 column (25m × 0.32 mm × 0.25 µm)
and a flame ionization detector using helium as the carrier
gas (linear velocity 27 cm/s at 50 °C). Injector and detector
temperatures were 250 and 350 °C, respectively. The column
was held at 50 °C for 2 min and ramped to 270 °C at 10 °C/
min. The amounts of individual components were calculated
using response factors determined by running the authentic
standards.
To assess the effect of substitutes on the reactivity of
the cinnamyl side chain, three derivatives of ferulic acid
were investigated and the results are shown in Figures
4-6. Ethyl ferulate (II) reacted with hydrogen peroxide
UV-vis Sp ectr om etr y. UV-vis spectra were recorded on
a Varian Cary 1E UV-visible Spectrophotometer using metha-
nol-water (1:1, v/v) as the solvent.

Oxidation of Ferulic Acid
J. Agric. Food Chem., Vol. 47, No. 8, 1999 3327
F igu r e 4. UV spectral changes during the reaction of ethyl
ferulate (II) with hydrogen peroxide (P) and peracetic acid (PA)
after 4 h of reaction time.
F igu r e 5. UV spectral changes during the reaction of conifer-
aldehyde (III) with hydrogen peroxide (P) and peracetic acid
(PA) after 4 h of reaction time.
F igu r e 7. GC chromatograms of trimethylsilylated deriva-
tives from peracetic acid reaction products of ferulic acid (I),
ethyl ferulate (II), coniferaldehyde (III), and 3,4-dimethoxy-
cinnamic acid (IV). IS (internal standard): acetovanillone.
Asterisk (/) denotes impurities.
hydrogen peroxide and peracetic acid almost completely
destroyed this aldehyde derivative. However, as shown
in Figure 5, these two oxidants gave a different com-
position of reaction products. It has been found that the
peracetic acid produced more phenylcarboxylic-acid-
related compounds so that the absorption maxima
shifted to lower wavelengths (Pan et al., 1999). Unlike
compounds I-III, 3,4-dimethoxycinnamic acid (IV)
showed very little reactivity toward either hydrogen
peroxide or peracetic acid. This compound virtually did
not react with hydrogen peroxide. The peracetic acid
reaction caused a shift of the absorption curve, but it
did not considerably decrease the absorbance.
Id en tifica tion of Rea ction P r od u cts. Figure 7
shows the GC separation of the trimethylsilylated
derivatives of reaction mixtures obtained from peracetic
acid reactions of ferulic acid (I), ethyl ferulate (II),
coniferaldehyde (III), and 3,4-dimethoxycinnamic acid
(IV). Overall, hydrogen peroxide reacted with these
substrates to a smaller extent or to give fewer products.
The products were identified by comparison of their
retention times with those of authentic compounds and
F igu r e 6. UV spectral changes during the reaction of 3,4-
dimethoxycinnamic acid (IV) with hydrogen peroxide (P) and
peracetic acid (PA) after 4 h of reaction time.
to some extent. Again, peracetic acid resulted in a more
pronounced reaction than did hydrogen peroxide. Co-
niferaldehyde (III) showed a great reactivity. Both

3328 J. Agric. Food Chem., Vol. 47, No. 8, 1999
Pan et al.
Ta ble 1. Ma ss Sp ectr om etr ic Da ta of Tr im eth ylsilyla ted Der iva tives fr om Hyd r ogen P er oxid e a n d P er a cetic Acid
Rea ction P r od u cts
compound
MS (m/z)
I
338 (M, 100), 323 (65), 308 (60), 293 (26), 279 (9), 249 (48), 219 (21), 191 (9), 175 (7), 146 (14), 73 (45), 59 (4), 45 (5)
338 (M, 100), 323 (45), 308 (44), 293 (21), 279 (7), 249 (35), 219 (14), 191 (7), 175 (5), 146 (6), 73 (30), 59 (3), 45 (5)
294 (M, 88), 279 (20), 264 (100), 249 (19), 236 (8), 219 (18), 192 (20), 73 (21), 59 (3), 45 (4)
Icis
II
IIcis
III
V
294 (M, 86), 279 (27), 264 (100), 249 (30), 236 (11), 219 (25), 192 (27), 73 (34), 102 (7), 59 (5), 45 (5)
250 (M, 100), 235 (19), 220 (97), 207 (6), 192 (55), 177 (13), 166 (14), 151 (7), 117 (7), 102 (8), 89 (15), 73 (58), 59 (13), 45 (16)
224 (M, 40), 209 (48), 194 (100), 165 (9), 151 (5), 137 (10), 121 (3), 104 (8), 89 (9), 73 (27), 59 (6), 45 (9)
312 (M, 75), 297 (100), 282 (31), 267 (53), 253 (37), 223 (64), 193 (35), 165 (10), 126 (16), 89 (10), 73 (82), 59 (8), 45 (17)
326 (M, 77), 311 (38), 296 (9), 267 (28), 238 (4), 223 (7), 209 (58), 193 (12), 179 (32), 163 (6), 149 (8), 73 (100), 59 (5), 45 (10)
284 (M, 80), 269 (14), 254 (100), 239 (10), 209 (4), 179 (5), 112 (5), 73 (27), 59 (3), 45 (5)
VI
VII
VIII
IX
282 (M, 80), 267 (32), 252 (71), 238 (6), 223 (10), 209 (100), 193 (15), 179 (73), 163 (5), 149 (15), 73 (42), 59 (5), 45 (6)
stedt and Agnemo, 1980). The double-bond rupture
leads to the formation of vanillin (V). Subsequently, this
benzaldehyde can be readily oxidized to the correspond-
ing benzoic acid, i.e., vanillic acid (VI). Also, vanillin
(V) as a p-hydroxybenzaldehyde may be further oxidized
to methoxyhydroquinone (VIII) by the Dakin reaction
with hydrogen peroxide (Dence, 1994b) or the Baeyer-
Villiger reaction with peracetic acid (Strumila and
Rapson, 1975; Nimz and Schwind, 1979). On the other
hand, the formation of homovanillic acid (VII) would
suggest that the side-chain cleavage could involve
mechanisms other than the well-known R,â double-bond
scission. In addition, ethyl homovanillate (IX) was found
only in the peracetic acid reactions. This product prob-
ably resulted from the esterification of homovanillic acid
(VII) by ethanol used as solvent in the peracetic acid
reaction. More detailed results concerning the reaction
of coniferaldehyde-type structures will be published
elsewhere (Pan et al., 1999).
As can be seen in Figure 7, 3,4-dimethoxycinnamic
acid (IV) gave small amounts of one product, homover-
atric acid (X). Similar to homovanillic acid (VII), this
product was not found in the hydrogen peroxide reac-
tion. 3,4-Dimethoxycinnamic acid (IV) was virtually
unreactive to hydrogen peroxide. This result explained
the change in the UV absorption of this substrate due
to the peracetic acid reaction, as shown in Figure 6.
Ra tes of Disa p p ea r a n ce of Su bstr a tes a n d Yield s
of Rea ction P r od u cts. To further analyze the course
of the consumption of the substrates and the course of
the formation of products, quantitative data are given
in Tables 2-4. The amounts of starting materials
remaining upon hydrogen peroxide or peracetic acid
reaction (Table 2) were consistent with those observed
by UV absorption measurements (Figures 2-6). Overall,
ferulic acid (I) and ethyl ferulate (II) showed a similar
reactivity to either hydrogen peroxide or peracetic acid.
Both were partially consumed by the reactions. Perace-
tic acid resulted in a greater extent of consumption of
substrates than did hydrogen peroxide. The effect of
nitrogen atmosphere on the rate of reaction was not
clearly observed. Excluding oxygen from the reaction
system slowed the reaction of ferulic acid (I) with
peracetic acid and the reaction of ethyl ferulate (II) with
hydrogen peroxide but appeared to have little effect on
the other degradation reactions. Moreover, addition of
oxygen to the hydrogen peroxide reaction of ferulic acid
(I) did not seem to increase the rate of reaction.
F igu r e 8. Structures of hydrogen peroxide and peracetic acid
reaction products: vanillin (V), vanillic acid (VI), homovanillic
acid (VII), methoxyhydroquinone (VIII), ethyl homovanillate
(IX), and homoveratric acid (X).
confirmed by mass spectrometric data given in Table
1. Figure 8 illustrates the structures of identified
products.
As shown in Figure 7, ferulic acid (I) gave two
products with the starting material remaining in a
considerably large amount. Compound Icis was as-
signed to cis-ferulic acid because Compounds I and Icis
had a nearly identical mass spectrum, as shown in Table
1. This observation indicated that the reaction substrate
underwent isomerization to its cis form under our
experimental conditions. The second product was homo-
vanillic acid (VII), which was found only in the peracetic
acid reaction. The formation of this compound would
suggest that the peracetic acid attack on the olefinic
double bond of ferulic acid could cause the cleavage of
the side chain.
By contrast, ethyl ferulate (II) only gave one product.
Similarly, this product was the cis form of II due to
isomerization. Moreover, only trace amounts of ferulic
acid (I) were observed when ethyl ferulate (II) was
reacted with hydrogen peroxide. This indicated that the
occurrence of saponification of II was very limited under
our experimental conditions (pH ≈ 11).
Figure 7 reveals that the reaction of coniferaldehyde
(III) was far more pronounced than that of the above
two substrates. It formed five products and only a very
small amount of the starting material remained. It is
well-known that the reaction of coniferaldehyde struc-
tures with hydrogen peroxide involves oxidative cleav-
age of the R,â double bond, which is initiated by
nucleophilic attack on CR followed by formation of
epoxide intermediate (Reeves and Pearl, 1965; Geller-
Table 2 shows that either hydrogen peroxide or
peracetic acid reacted nearly completely with conifer-
aldehyde (III). The aldehyde substituent makes the side
chain of cinnamyl structures highly susceptible to
oxidative degradation. By contrast, 3,4-dimethoxycin-
namic acid (IV) showed no reactivity to hydrogen
peroxide and was recovered in almost 100% yield.

Oxidation of Ferulic Acid
J. Agric. Food Chem., Vol. 47, No. 8, 1999 3329
Ta ble 2. Am ou n ts (%) of Rem a in in g Sta r tin g Ma ter ia ls u p on Hyd r ogen P er oxid e a n d P er a cetic Acid Rea ction s
ferulic acid
ethyl ferulate
coniferaldehyde
3,4-dimethoxycinnamic acid
reaction conditions
(I)
(II)
(III)
(IV)
H2O2
4 h
87.1
87.5
89.0
59.4
48.7
76.8
74.0
92.6
2.4
1.3
99.2
99.9
N2, 4 h
O2, 4 h
1 h
4 h
N2, 4 h
CH3CO3H
84.2
58.0
58.3
6.3
5.3
1.5
98.4
96.8
101
Ta ble 3. Yield (m ol %) of Ma in P r od u cts Obta in ed fr om th e Hyd r ogen P er oxid e a n d P er a cetic Acid Rea ction s of F er u lic
Acid (I), Eth yl F er u la te (II), a n d 3,4-Dim eth oxycin n a m ic Acid (IV)
ferulic acid (I)
homovanillic
acid (VII)
ethyl ferulate (II)
IIcis
3,4-dimethoxycinnamic acid (IV)
homoveratric acid (X)
Icis
H2O2
4 h
N2, 4 h
1 h
4 h
N2, 4 h
13.9
1.8
32.6
35.9
19.4
17.2
7.5
19.4
22.9
13.9
CH3CO3H
4.5
6.0
3.9
3.4
5.8
2.1
Ta ble 4. Yield (m ol %) of P r od u cts Obta in ed fr om th e Hyd r ogen P er oxid e a n d P er a cetic Acid Rea ction s of
Con ifer a ld eh yd e (III)
vanillin
(V)
vanillic acid
homovanillic acid
methoxyhydroquinone
reaction conditions
(VI)
(VII)
(VIII)
H2O2
CH3CO3H
4 h
N2, 4 h
1 h
4 h
N2, 4 h
0.8
trace
7.1
5.2
2.1
13.5
8.9
13.6
15.8
14.9
7.3
6.6
10.9
11.6
12.7
20.4
21.6
10.1
4.8
6.4
However, there was a slight consumption of this sub-
strate by peracetic acid.
peroxide and peracetic acid. In general, similar products
were obtained but in varying yield, with the exception
of ethyl homovanillate (IX) that was only present in the
peracetic acid reaction. As its GC response factor was
not determined, the yield of this product was not
calculated but its relative abundance did not signifi-
cantly vary during the peracetic acid reaction. Among
the other products, there seemed to be a correlation
between the consumption of vanillin (V) and the forma-
tion of methoxyhydroquinone (VIII). As discussed ear-
lier, vanillin can be converted to methoxyhydroquinone
(V) by the Dakin reaction, in hydrogen peroxide bleach-
ing or the Baeyer-Villiger oxidation by peracetic acid.
It appeared that the hydrogen peroxide reaction condi-
tions might favor the occurrence of the Dakin reaction
whereas the peracetic acid tended to oxidize vanillin (V)
to vanillic acid (VI). As a consequence, the hydrogen
peroxide reaction yielded more methoxyhydroquinone
(VIII) and less vanillin (V) and the peracetic acid
reaction produced more vanillic acid (VI) and less
methoxyhydroquinone (VIII). In addition, peracetic acid
gave more homovanillic acid (VII) than hydrogen per-
oxide.
P ossible Rea ction Mech a n ism s. In the present
investigation, ferulic acid (I) was shown to undergo two
reactions: isomerization and side-chain oxidative cleav-
age. The transformation of trans-ferulic acid to its cis
isomer is reported to occur upon hydrogen peroxide-
UV photolysis (Sun et al., 1996). In our case, the
reaction mechanisms for isomerization are not clear.
However, it required a free phenolic hydroxyl group, and
the presence of oxygen promoted this process. The
isomerization occurred only with ferulic acid (I) and
ethyl ferulate (II) but not with 3,4-dimethoxycinnamic
acid (IV). Moreover, the reactions with hydrogen per-
oxide or peracetic acid under nitrogen significantly
reduced the formation of cis isomers. The experimental
evidence, therefore, suggests the involvement of phe-
As revealed by Table 3, the major products obtained
from the reactions of ferulic acid (I) and ethyl ferulate
(II) were the cis isomers. Both the reactions produced
trace amounts of vanillin (V) and vanillic acid (VI). In
addition, the peracetic acid reaction of ferulic acid (I)
produced up to 6% of homovanillic acid (VII). It ap-
peared that trans-cis isomerization occurred to a
greater extent with peracetic acid than with hydrogen
peroxide. Furthermore, excluding oxygen from the reac-
tions with either of the oxidants significantly reduced
the formation of cis isomers. It should be pointed out
that the yields of the cis products given in Table 3 were
obtained by using the GC response factors of their trans
counterparts (assuming the same value for both the
trans and cis isomers of ferulic acid and ethyl ferulate,
respectively). Combination of the results in Tables 2 and
3 indicated that in most cases the sum of the yields of
the trans and cis isomers was around 90% or higher for
both ferulic acid (I) and ethyl ferulate (II). In a few cases
the peracetic acid reaction gave a total recovered mate-
rial yield of 70-80%. In addition to the observed
homovanillic acid (VII) and traces of vanillin (V) and
vanillic acid (VI), the discrepancy of mass balance could
be due to other reactions by peracetic acid, e.g., ring-
opening reaction, forming more intensively degraded
products.
Unlike the free phenolic substrates, the reaction of
3,4-dimethoxycinnamic acid (IV) had a good mass
balance. The 4-O-methylated compound was more stable
to oxidation by either hydrogen peroxide or peracetic
acid. The only reaction seen was the slight side-chain
cleavage by peracetic acid to produce homoveratric acid
(X). The magnitude of this reaction appeared to be
similar to that observed for ferulic acid (I).
Table 4 reports the yields of products obtained from
the reaction of coniferaldehyde (III) with hydrogen

3330 J. Agric. Food Chem., Vol. 47, No. 8, 1999
Pan et al.
dimethoxyphenyl)pyruvic acid with hydrogen peroxide.
They attributed this reaction to the decarboxylation of
R-keto acids (Bunton, 1962). On the other hand, in-
volvement of radicals in the side-chain rupture cannot
be ruled out.
CONCLUSIONS
In lignin-retaining bleaching conditions using hydro-
gen peroxide or peracetic acid, ferulic acid (I) undergoes
two reactions: isomerization and side-chain oxidative
cleavage. Esterification does not have a substantial
effect on the reactivity of ferulic acid, but 4-O-etherifi-
cation fully stabilizes it against isomerization. By
contrast, aldehyde substitution tremendously enhances
the susceptibility of the cinnamyl side chain to oxidative
degradation, as shown by the intensive degradation of
coniferaldehyde (III) under the same reaction condi-
tions, leading to the formation of a number of products.
On the basis of the results presented here, it would be
reasonable to conclude that the conjugated side chain
of ferulic acid and its etherified or esterified derivative
is fairly stable to the two oxidants studied, especially
hydrogen peroxide. The major reaction observed is
trans-cis isomerization that possibly involves a radical
mechanism but does not cause bond cleavage. The
peracetic acid reaction increases the rate of trans-cis
isomerization and is also accompanied by a minor
cleavage of the side chain.
F igu r e 9. Proposed mechanism for trans-cis isomerization
of ferulic acid (I) and ethyl ferulate (II) during hydrogen
peroxide or peracetic acid reaction.
LITERATURE CITED
Bunton, C. A. Nucleophilic Reactions of Peroxides. In Peroxide
Reaction Mechanisms; Edwards, J . O., Ed.; Interscience:
New York, London, 1962; pp 11-28.
Davies, D. M.; Deary, M. E. Kinetics of the hydrolysis and
perhydrolysis of tetraacetylethylenediamine, a peroxide
bleach activator. J . Chem. Soc., Perkin Trans. 2 1991, 1549-
1552.
Dence, C. W. Reaction Principles in Pulp Bleaching. In Pulp
BleachingsPrinciples and Practice; Dence, C. W., Reeve, D.
W., Eds.; TAPPI Press: Atlanta, GA, 1994a; pp 113-124.
Dence, C. W. Chemistry of Mechanical Pulp Bleaching. In Pulp
BleachingsPrinciples and Practice; Dence, C. W., Reeve, D.
W., Eds.; TAPPI Press: Atlanta, GA, 1994b; pp 161-181.
Gellerstedt, G.; Agnemo, R. The reactions of lignin with
alkaline hydrogen peroxide. Part III. The oxidation of
conjugated carbonyl structures. Acta Chem. Scand. 1980,
B34, 275-280.
Grime, K.; Clauss, A. Laundry bleaches and activators. Chem.
Ind. 1990, 647-653.
F igu r e 10. Proposed pathway for formation of homovanillic
acid (VII) or homoveratric acid (X) from ferulic acid (I) or 3,4-
dimethoxycinnamic acid (IV) on peracetic acid reaction.
noxy radicals. We note that reactions using peroxygen
reagents including hydrogen peroxide and peracetic acid
are inevitably accompanied by their decomposition to
more reactive species such as hydroxy radical (HO•) and
•
superoxide radical (O2•-/HO2 ) (Dence, 1994a). It is
proposed in Figure 9 that a hydroxy or hydroperoxy
radical initiates hydrogen abstraction from a free phe-
nolic substrate to form a phenoxy radical that can
rearrange to a quinone methide radical intermediate.
In our study, the reaction temperature of 50 °C,
although not very high, is probably sufficient to cause
the cis-trans isomerization to some extent. On the other
hand, because the experiments were not carried out in
the dark, the presence of light may promote the isomer-
ization.
J ames, A. P.; MacKirdy, I. S. The chemistry of peroxygen
bleaching. Chem. Ind. 1990, 641-645.
On the other hand, appreciable amounts of homovan-
illic acid (VII) and homoveratric acid (X) were found
only in the reactions of ferulic acid (I) and 3,4-
dimethoxycinnamic acid (IV), respectively, with perace-
tic acid. As peracetic acid has been well documented as
an epoxidizing agent for olefinic bonds (Plesnicar, 1978),
reaction mechanisms for the formation of these substi-
tuted phenylacetic acids might involve epoxide inter-
mediate, as shown in Figure 10. However, it is not clear
how the epoxidized side chain of the proposed interme-
diate splits to give homovanillic acid (VII) or homover-
atric acid (X). One possibility could be via decarboxy-
lation followed by further oxidation of the rest of the
side chain. One example relevant to this mechanism is
the work of Reeves and Pearl (1965) who obtained
homoveratric acid in 80% yield from the reaction of (3,4-
Legrini, O.; Oliveros, E.; Braun, A. M. Photochemical processes
for water treatment. Chem. Rev. 1993, 93, 671-698.
Nimz, H. H.; Schwind, H. Oxidation of monomeric lignin model
compounds with peracetic acid. Cell. Chem. Technol. 1979,
13, 35-46.
Pan, G. X.; Bolton, J . L.; Leary, G. J . Determination of ferulic
and p-coumaric acids in wheat straw and the amounts
released by mild acid and alkaline peroxide treatment. J .
Agric. Food Chem. 1998, 46, 5283-5288.
Pan, G. X.; Spencer, L.; Leary, G. J . A comparative study on
reactions of hydrogen peroxide and peracetic acid with lignin
chromophores. Part 1. The reaction of coniferaldehyde model
compounds. Holzforschung 1999, submitted for publication.
Plesnicar, B. Oxidation with Peroxy Acids and Other Perox-
ides. In Oxidation in Organic Chemistry, Part C; Traha-
novsky, W. C., Ed.; Academic Press: New York, San
Francisco, London, 1978; pp 211-294.

Oxidation of Ferulic Acid
J. Agric. Food Chem., Vol. 47, No. 8, 1999 3331
Reeves, R. H.; Pearl, I. A. Reaction products formed upon the
alkaline peroxide oxidation of lignin-related model com-
pounds. Tappi 1965, 48 (2), 121-125.
Strumila, G.; Rapson, W. H. Reaction products of neutral
peracetic acid oxidation of model lignin phenols. Pulp Pap.
Can. 1975, 76 (9), T276-T280.
Turner, N. A.; Mathews, A. J . Enhanced delignification and
bleaching using TAED activated peroxide. Proceedings of
TAPPI Pulping Conference, Montreal, Quebec, Canada,
October 25-29, 1998, Book 3, 1269-1276.
Sun, Y.-P.; Yates, B.; Abbot, J .; Chen, C.-L. Oxidation of lignin
model compounds containing an R-carbonyl group and a
ring-conjugated double-bond by hydrogen peroxide-UV pho-
tolysis. Holzforschung 1996, 50 (3), 226-232.
Received for review March 8, 1999. Revised manuscript
received May 18, 1999. Accepted May 22, 1999.
J F9902494