Significant lability of guaiacylglycerol β-phenacyl ether under alkaline conditions

Article abstract of DOI:10.1021/jf071147d

It was observed that the β-O-4 bond cleavage of a dimeric phenolic lignin model compound with an α-carbonyl group at the B-ring, 2-(2-ethoxy-4-formylphenoxy)-1-(4-hydroxy-3-methoxyphenyl)propane-1,3-diol (I), is extremely fast in a mild anaerobic alkaline treatment (0.45 mol/L NaOH, 95°C, 0.8 MPa of N₂). This phenomenon significantly contrasts with the case of a common dimeric phenolic lignin model compound without any specific functional group, 1-(4-hydroxy-3-methoxyphenyl)-2-(2-methoxyphenoxy) propane-1,3-diol (II). The most plausible mechanism is the migration of the B-ring from the β- to the α-position following the S_NAr mechanism. Because this migration affords the alkaline labile phenolic α-O-4-type compound (XI), the formation of the quinone methide as well as the cleavage of the originally alkaline very stable alkyl-aryl ether bond is promoted. This promotion of the quinone methide formation explains why a relatively large amount of 4-hydroxy-3-methoxybenzaldehyde (IV) is produced from I in an oxygen-alkali treatment.

Full text of DOI:10.1021/jf071147d

J. Agric. Food Chem. 2007, 55, 9043–9046 9043
Significant Lability of Guaiacylglycerol ꢀ-Phenacyl
Ether under Alkaline Conditions
†
AIKO IMAI, TOMOYA YOKOYAMA,* YUJI MATSUMOTO, AND
‡
GYOSUKE MESHITSUKA
Laboratory of Wood Chemistry, Department of Biomaterial Sciences, Graduate School of Agricultural
and Life Sciences, The University of Tokyo, 1-1-1 Yayoi, Bunkyo-ku, Tokyo 113-8657, Japan
It was observed that the ꢀ-O-4 bond cleavage of a dimeric phenolic lignin model compound with an
R-carbonyl group at the B-ring, 2-(2-ethoxy-4-formylphenoxy)-1-(4-hydroxy-3-methoxyphenyl)propane-
1,3-diol (I), is extremely fast in a mild anaerobic alkaline treatment (0.45 mol/L NaOH, 95 °C, 0.8
MPa of N ). This phenomenon significantly contrasts with the case of a common dimeric phenolic
2
lignin model compound without any specific functional group, 1-(4-hydroxy-3-methoxyphenyl)-2-(2-
methoxyphenoxy)propane-1,3-diol (II). The most plausible mechanism is the migration of the B-ring
from the ꢀ- to the R-position following the S Ar mechanism. Because this migration affords the alkaline
N
labile phenolic R-O-4-type compound (XI), the formation of the quinone methide as well as the cleavage
of the originally alkaline very stable alkyl-aryl ether bond is promoted. This promotion of the quinone
methide formation explains why a relatively large amount of 4-hydroxy-3-methoxybenzaldehyde (IV)
is produced from I in an oxygen–alkali treatment.
KEYWORDS: Alkyl-aryl ether; ꢀ-O-4; carbonyl; lignin; migration; quinone methide
INTRODUCTION
resulting in a kind of intramolecular S 2 reaction (back-side
N
attack) (1). The contribution of the γ-position to this reaction
The most important utilization of wood has still been
producing pulp for papermaking, although the use for bioethanol
production and others has recently expanded. Therefore, the
investigation of the pulping process is still the most interesting
and necessary topic. Chemical pulp for papermaking is generally
produced by the Kraft process in which NaOH and Na2S are
utilized as reagents. More than 90% of lignin is removed in
this process, and the residual several percent is eliminated in
subsequent bleaching processes.
is small (2). To advance this back-side attack sufficiently as a
delignification reaction, conditions required are at least more
severe than 1 mol/L NaOH and 130 °C (3).
It is well known that functional groups affect the rate of the
back-side attack (4–6). Among these, the effect of the carbonyl
group is the most interesting. Gierer and Ljunggren reported
that the alkyl-aryl ether bond of a ꢀ-O-4 substructure with an
R-carbonyl group at the B-ring (see Figure 1 for the definition
of A- and B-rings) cleaves 40 times faster than that of another
common ꢀ-O-4 substructure without any specific functional
group, following the back-side attack mechanism (7).
There has been much research dealing with reactions of lignin
(
delignification) under alkaline conditions, because the pulping
process is based on these reactions as described above. The
delignification in the process is mainly attained by the cleavage
of the ꢀ-O-4 bond (see Figure 1 for the definition of ꢀ-O-4),
which is the most abundant substructure in wood lignin. Figure
In this paper, a dimeric phenolic lignin model compound with
an R-carbonyl group at the B-ring, 2-(2-ethoxy-4-formylphe-
noxy)-1-(4-hydroxy-3-methoxyphenyl)propane-1,3-diol (I in Fig-
ure 2), was subjected to mild anaerobic alkaline and oxygen–
alkali treatments, and the effect of the carbonyl group is
reevaluated by comparing the results with those obtained from
the same treatments of a common dimeric phenolic lignin model
compound without any specific functional group, 1-(4-hydroxy-
1
illustrates the general mechanism for the cleavage of the
nonphenolic ꢀ-O-4 bond under alkaline pulping conditions
generally ca. 2 mol/L NaOH, >150 °C). An alcoholate anion
produced at an R-position attacks the neighboring ꢀ-carbon,
(
*
To whom correspondence should be addressed. Telephone:
3-methoxyphenyl)-2-(2-methoxyphenoxy)propane-1,3-diol (II).
+
81-3-5841-5264. Fax: +81-3-5802-8862. E-mail: yokoyama@
woodchem.fp.a.u-tokyo.ac.jp.
A reaction product obtained from the B-ring of I in a mild
anaerobic alkaline treatment, 3-ethoxy-4-hydroxybenzaldehyde
†
Current address: Technical Research Institute, Toppan Printing Co.,
Ltd., 4-2-3 Takanodaiminami, Sugito-cho, Kitakatsushika-gun, Saitama
(
III), and another reaction product derived from the A-ring of
3
45-0046, Japan.
‡
I and II in oxygen–alkali treatments, 4-hydroxy-3-methoxy-
benzaldehyde (IV), were quantified. Rational mechanisms are
proposed for the ꢀ-O-4 bond cleavage in a mild anaerobic
Current address: Department of Clothing Science and Fine Arts,
Faculty of Home Economics, Tokyo-Kasei University, 1-18-1 Kaga,
Itabashi-ku, Tokyo 173-8602, Japan.
1
0.1021/jf071147d CCC: $37.00  2007 American Chemical Society
Published on Web 10/04/2007

9044 J. Agric. Food Chem., Vol. 55, No. 22, 2007
Imai et al.
Figure 1. General neighboring group participation mechanism for the
cleavage of the nonphenolic ꢀ-O-4 bond under alkaline pulping
conditions.
Figure 2. Chemical structure of the compounds used or referred to in
this study.
Figure 3. Yield of III and IV in the mild anaerobic alkaline and the
oxygen–alkali treatments of I or II.
alkaline treatment of I and for the oxidation in an oxygen–alkali
treatment of I.
the injector and detector were 280 °C. Separations were achieved on a
capillary column of NB-1 (30 m × 0.25 mm × 0.4 µm film thickness;
GL Sciences Inc., Tokyo, Japan). The temperature program was from
MATERIALS AND METHODS
1
50 to 280 °C at a rate of 5 °C/min while maintaining the initial and
Materials. Figure 2 illustrates the chemical structure of the
compounds used or referred to in this study. Dimeric lignin model
compounds, I and II, were synthesized according to the methods of
Katayama et al. (8) and Li et al. (9). These compounds were purified
using a preparative medium-pressure liquid chromatograph (YFLC-
final temperatures for 10 and 20 min, respectively.
RESULTS AND DISCUSSION
Anaerobic Alkaline Treatment. Gierer and Ljunggren
reported lots of kinetic data concerning the ꢀ-O-4 bond cleavage
of several dimeric lignin model compounds (7). We can estimate
the extent of the ꢀ-O-4 bond cleavage of these dimeric model
compounds in a reaction for 100 min under the conditions
employed in this study (0.45 mol/L NaOH, 95 °C) assuming
the back-side attack mechanism, when an Arrhenis plot is
created for each compound using the kinetic data presented in
the literature. When 2-(4-acetyl-2-methoxyphenoxy)ethanol (V)
is assumed to be subjected to the mild anaerobic alkaline
treatment for 100 min under the conditions employed in this
study, the mole percent yield of 4-acetyl-2-methoxyphenol (VI)
produced from the ꢀ-O-4 bond cleavage of V (based on the
initial amount of V) can be estimated at 0.3%. Gierer and
Ljunggren also suggested that a dimeric model compound, 2-(4-
acetyl-2-methoxyphenoxy)-1-(3,4-dimethoxyphenyl)propane-
5
40; Yamazen Co., Osaka, Japan) on a silica gel column (Ultrapack
SI-40C; Yamazen Co.). The eluent was a mixture of chloroform/ethanol
1
(
10/1). The structures were confirmed by H NMR. A semiconductor
grade of NaOH and FeCl (Sigma-Aldrich Japan K. K., Tokyo, Japan)
and ultrapure water were used in all of the experiments.
3
Alkaline Treatment under Anaerobic and Aerobic Conditions.
3
An aqueous solution containing NaOH and FeCl (4 mL) was bubbled
with nitrogen and added to a stainless steel reactor equipped with a
Teflon inscribed tube and lid. To the reactor was added 0.4 mL of an
ethanol solution containing I or II. After the air in the reactor was
replaced with nitrogen and the reactor was pressured with 0.8 MPa of
nitrogen, the reactor was soaked in a water bath (95 °C) at which the
reaction time was defined as 0. Four reactions were run for 0, 15, 50,
and 100 min. Oxygen–alkali treatments (0.8 MPa) were also carried
out following the same procedure. The concentrations of the additives
3
and I (or II) were as follows: NaOH, 0.45 mol/L; FeCl , 0.33 mmol/
L; I, 2.6 mmol/L (or II, 3.1 mmol/L). Because I or II was primarily
dissolved into ethanol, the reaction solution was 9.1% aqueous
ethanol.
1
,3-diol (VII) with an A-ring, affords VI about 8 times faster
than V due to the presence of the A-ring (7). On the basis of
this suggestion, the mole percent yield of VI produced from
the ꢀ-O-4 bond cleavage of VII (based on the initial amount of
VII) can be estimated at 4.1%, when VII is assumed to be
subjected to the mild anaerobic alkaline treatment for 100 min
under the conditions employed in this study. The mole percent
yield of III from I (based on the initial I) in the mild anaerobic
alkaline treatment in this study is shown by the b in Figure 3.
As can be seen, the mole percent yield at 100 min was about
Quantification of Degradation Products. After the reactor was
cooled in an ice–water for 1 min, the lid was opened, and an alkaline
solution of an internal standard (4-hydroxy-3,5-dimethoxybenzaldehyde)
was added. The lid was closed again, and the resultant solution was
shaken well. The solution was transferred into a glass beaker. Then,
the solution was acidified with hydrochloric acid and extracted with
dichloromethane three times. The dichloromethane layers were collected
and dried over anhydrous sodium sulfate. A portion of the dichlo-
romethane solution was transferred into a micro tube and dried by a
current of nitrogen. The content was trimethylsilylated with 75 µL of
N,O-bis(trimethylsilyl)acetamide at 100 °C for 5 min and injected into
a gas chromatograph.
8
3%. This is extremely larger than the estimated value, 4.1%,
for VII. Furthermore, it is presumed that the ꢀ-O-4 bond
cleavage of I by the back-side attack mechanism is slower than
VII due to the presence of the already ionized phenolic hydroxyl
group in I. These yield estimations suggest that the ꢀ-O-4 bond
Quantitative analyses for III and IV were performed on a GC-17A
instrument (Shimadzu Co., Kyoto, Japan) equipped with a flame
ionization detector using helium as a carrier gas. The temperatures of

Significant Lability of Guaiacylglycerol ꢀ-Phenacyl Ether
J. Agric. Food Chem., Vol. 55, No. 22, 2007 9045
that the alkyl-aryl ether in the substructure cleaves is brought
about. This phenomenon can develop a practical application
described in the following section.
Oxygen–Alkali Treatment. The mole percent yield of III
from I (based on the initial I) in the oxygen–alkali treatment is
shown by the O in Figure 3. The yield at 15 min was almost
the same as in the mild anaerobic alkaline treatment, and the
formation rate gradually decreased, resulting in the approxi-
mately 70% yield at 100 min, which was lower than the yield
at 100 min in the mild anaerobic alkaline treatment. The reaction
product III, which carries an R-carbonyl group, is quite stable
toward molecular oxygen under conditions of oxygen–alkali
(
11). However, because III can react with peroxides produced
in the aerobic reaction system, it is anticipated that even if the
amount of III produced in the oxygen–alkali treatment is the
same as in the mild anaerobic alkaline treatment, some of them
are probably transformed into other compounds.
As a noticeable result, the mole percent yield of IV (based
on the initial I) in the oxygen–alkali treatment is shown by the
]
in Figure 3. Almost no IV was produced in the oxygen–alkali
Figure 4. Mechanism proposed for the reaction route from I to III and IV
in the mild anaerobic alkaline and the oxygen–alkali treatments.
treatment of II (3 in Figure 3). In general, ring-opening
degradation products are obtained when common phenolic lignin
model compounds without a specific functional group, such as
II, are treated under conditions of oxygen–alkali (12, 13). By
considering the reaction mechanism, it is presumed that IV is
produced by the oxidation of the quinone methide-type (XII)
or the styrene-type structure (XIII), which is derived from XII
by the abstraction of the ꢀ-proton. Therefore, the reaction route
of I in the oxygen–alkali process can be the same as in the
anaerobic alkaline process until the formation of XII, and IV
is produced by the oxidation of XII or XIII. The difficulty of
the quinine methide formation from II under the conditions
employed reflects to the result that almost no IV is obtained
from II in the oxygen–alkali process. Although the negligible
amount of IV was the only product detected in the oxidation of
II, much research ensures that a fairly large amount of II was
oxidized and the A-ring of II must have been converted into
ring-opening degradation products by the oxidation (12–17).
Even in the oxygen–alkali treatment of I, the formation of IV
and ring-opening degradation products should compete with
each other. The yield of IV from I (] in Figure 3) in the
oxygen–alkali treatment is about half of III (O in Figure 3).
This result must indicate that the oxidation of XIII results not
only in the formation of IV but also in ring-opening degradation
products.
cleavage of I under the conditions employed in this study does
not follow the back-side attack mechanism shown in Figure 1.
It was reported by Criss et al. that in a quite mild alkaline
treatment (0.34 mol/L NaOH, room temperature) of a dimeric
nonphenolic lignin model compound with an R-carbonyl group
at the B-ring, 2-(4-formyl-2-methoxyphenoxy)-1-(3,4-dimethoxy-
phenyl)ethanol (VIII), the B-ring migrates to the R-position and
an equilibrium is established between VIII and 2-(4-formyl-2-
methoxyphenoxy)-2-(3,4-dimethoxyphenyl)ethanol (IX) (10).
The equilibrium is also established in the case of using IX as
a starting compound. The mechanism presented and proved in
the literature is an aromatic nucleophilic substitution reaction
(
SNAr; cf. Figure 4), in which the alcoholate oxygen at the
R-position attacks the C4 carbon of the B-ring and an intermedi-
ate with a five-membered ring (a kind of X) forms. This
knowledge makes it rational to apply this mechanism to the
phenomenon observed in this study, although there are the
structural differences between I and VIII, the phenolic and
nonphenolic A-rings, the presence and absence of the γ-position,
and the formyl and acetyl functional groups at the B-rings,
respectively.
Possible Practical Application. The results of Criss et al.
10) and those obtained in this study suggest the following
The mechanism proposed for the ꢀ-O-4 bond cleavage of I
under the conditions employed in this study is illustrated in
Figure 4. Once the R-O-4-type compound, 3-(2-ethoxy-4-
formylphenoxy)-3-(4-hydroxy-3-methoxyphenyl)propane-1,2-
diol (XI), forms by the migration of the B-ring, this compound
easily converts into the corresponding quinone methide (XII)
and the alkyl-aryl ether cleaves. To advance the quinone methide
formation frequently, conditions required are only 1 mol/L
NaOH and 10 °C for an R-O-4-type structure and, on the other
hand, 1 mol/L NaOH and 140 °C for a ꢀ-O-4-type structure
(
phenomenon. Nonphenolic ꢀ-O-4 substructures with R-carbonyl
groups at the B-rings are converted into corresponding non-
phenolic R-O-4 substructures by the migration of the B-rings.
When the phenolic hydroxyl group of an A-ring of one of these
R-O-4 substructures is liberated during pulping reactions, the
alkyl-aryl ether cleaves promptly. Therefore, if R-carbonyl
groups can be introduced sufficiently into lignin by some
oxidation methods, it is expected that portions of the ꢀ-O-4
polymer, such as endwise types, peel off a unit at a time from
phenolic end residues under relatively mild alkaline conditions.
This mechanism seems to give some idea for the development
of oxidation-based cookings as sulfur-free methods.
(3). The difficulty of the quinone methide formation from a ꢀ-O-
4
-type structure like II is due to the low leaving ability of the
hydroxide anion from the R-position and the effect of an aqueous
alkaline solvent.
The migration mechanism presented by Criss et al. is for
nonphenolic ꢀ-O-4 substructures with R-carbonyl groups at the
B-rings (10). When this mechanism operates on a phenolic
LITERATURE CITED
(
1) Gierer, J. Chemistry of delignification part 1: General concept
and reactions during pulping. Wood Sci. Technol. 1985, 19, 289–
312.
ꢀ
-O-4 substructure with an R-carbonyl group at the B-ring as
in the reaction system of this study, the significant phenomenon

9046 J. Agric. Food Chem., Vol. 55, No. 22, 2007
Imai et al.
(
(
2) Gierer, J.; Noren, I. On the course of delignification during kraft
pulping. Holzforschung 1980, 34, 197–200.
3) Miksche, G. E. Lignin reactions in alkaline pulping processes. In
Chemistry of Delignification with Oxygen, Ozone, and Peroxides;
Gratzl, J. S., Nakano, J., Singh, R. P., Eds.; Uni Publishers Co.:
Tokyo, 1980; p 107.
4) Hubbard, T. F.; Schultz, T. P.; Fisher, T. H. Alkaline-hydrolysis
of nonophenolic ꢀ-O-4 lignin model dimers: Substituent effects
on the leaving phenoxide in neighboring group vs direct nucleo-
philic attack. Holzforschung 1992, 46, 315–320.
delignification conditions. J. Agric. Food Chem. 2007, 55, 1307–
1307.
(12) Eckert, R. C.; Chang, H.-m.; Tucker, W. P. Oxidative degradation
of phenolic lignin model compounds with oxygen and alkali. Tappi
1
973, 56, 134–138.
(
13) Gierer, J.; Imsgard, F. Studies on the autoxidation of t-butyl-
substituted phenols in alkaline media 1: Reaction of 4-t-butylguai-
acol. Acta Chem. Scand. 1977, B31, 537–545.
14) Gierer, J. Chemistry of delignification part 2: Reactions of lignins
during bleaching. Wood Sci. Technol. 1986, 20, 1–33.
15) Ljunggren, S.; Johansson, E. The kinetics of lignin reactions during
oxygen bleaching 3: The reactivity of 4-n-propylguaiacol and 4,4′-
di-n-propyl-6,6′-biguaiacol. Holzforschung 1990, 44, 291–296.
16) Johansson, E.; Ljunggren, S. The kinetics of lignin reactions during
oxygen bleaching part 4: The reactivities of different lignin model
compounds and the influence of metal ions on the rate of
degradation. J. Wood Chem. Technol. 1994, 14, 507–525.
17) Yasumoto, M.; Matsumoto, Y.; Ishizu, A. The role of peroxide
species in the carbohydrate degradation during oxygen bleaching
(
(
(
5) Collier, W. E.; Gisher, T. H.; Harris, L. L.; Schultz, T. H. Alkaline
hydrolysis of nonphenolic ꢀ-O-4 lignin model dimers: Further
studies of the substituent effect on the leaving. Holzforschung
(
1
996, 50, 420–424.
(
(
(
(
6) Schultz, T. P.; Fisher, T. H. Phenoxyl substituent effect on the
alkaline hydrolysis rates of ꢀ-O-4 lignin models: A review. J.
Pulp Pap. Sci. 1998, 24, 242–246.
7) Gierer, J.; Ljunggren, S. The reactions of lignins during sulfate
pulping part 16: The kinetics of the cleavage of ꢀ-aryl ether
linkages in structures containing carbonyl groups. SVen. Papper-
stidn. 1979, 82, 71–81.
1
: Factors influencing the reaction selectivity between carbohy-
drate and lignin model compounds. J. Wood Chem. Technol. 1996,
16, 95–107.
(
8) Katayama, T.; Nakatsubo, F.; Higuchi, T. Syntheses of arylglyc-
erol-ꢀ-aryl ethers. Mokuzai Gakkaishi 1981, 27, 223–230.
9) Li, S.; Lundquist, K.; Paulsson, M. Synthesis of guaiacylglycerol-
(
ꢀ
-guaiacyl ether. Acta Chem. Scand. 1995, 49, 623–624.
Received for review April 20, 2007. Revised manuscript received August
1, 2007. Accepted September 5, 2007. The authors gratefully acknowl-
edge financial support from the Ministry of Education, Culture, Sports,
Science, and Technology of Japan [Grant-in-Aid for Young Scientists
(
10) Criss, D. L.; Ingram, L. L.; Schultz, T. P.; Fisher, T. H.; Saebo,
D. B. A low-temperature internal nucleophilic aromatic substitu-
tion reaction on a ꢀ-O-4 lignin model dimer. J. Org. Chem. 1997,
3
6
2, 7885–7887.
(
B), No. 17780136].
(
11) Yokoyama, T.; Matsumoto, Y.; Meshitsuka, G. Detailed examina-
tion of the degradation of phenol derivatives under oxygen
JF071147D