Self termination of ring opening reaction of p-substituted phenol-based benzoxazines: An obstructive effect via intramolecular hydrogen bond

Article abstract of DOI:10.1002/jhet.130

(Chemical Equation Presented) The ring opening polymerizations of p-substituted phenol-based benzoxazines are self-terminated as soon as dimers form. The polymerization of benzoxazine monomers does not proceed according to the theoretical mechanism even though the conditions, temperature, molar ratio, solvent polarity, and reactant ratio are varied. The speculated mechanism, involving the unique structure of a dimer with interand intramolecular hydrogen bonds, is applied to explain an obstructive effect on ring opening polymerization. In this article, we clarify an important case which the stereo structure of the compound controls the reaction and prevents the polymerization expected from the theoretical mechanism.

Full text of DOI:10.1002/jhet.130

714
Self Termination of Ring Opening Reaction of p-Substituted
Phenol-Based Benzoxazines: An Obstructive Effect via
Intramolecular Hydrogen Bond
Vol 46
Suwabun Chirachanchai,a,b* Apirat Laobuthee,c* and Suttinun Phongtamruga
aThe Petroleum and Petrochemical College, Chulalongkorn University, Bangkok 10330, Thailand
bCenter for Petroleum, Petrochemicals, and Advanced Materials, Chulalongkorn University,
Bangkok 10330, Thailand
cDepartment of Materials Engineering, Kasetsart University, Bangkok 10900, Thailand
*E-mail: csuwabun@chula.ac.th or fengapl@ku.ac.th
Received March 10, 2009
DOI 10.1002/jhet.130
Published online 13 July 2009 in Wiley InterScience (www.interscience.wiley.com).
The ring opening polymerizations of p-substituted phenol-based benzoxazines are self-terminated as
soon as dimers form. The polymerization of benzoxazine monomers does not proceed according to the
theoretical mechanism even though the conditions, temperature, molar ratio, solvent polarity, and reac-
tant ratio are varied. The speculated mechanism, involving the unique structure of a dimer with inter-
and intramolecular hydrogen bonds, is applied to explain an obstructive effect on ring opening polymer-
ization. In this article, we clarify an important case which the stereo structure of the compound controls
the reaction and prevents the polymerization expected from the theoretical mechanism.
J. Heterocyclic Chem., 46, 714 (2009).
INTRODUCTION
substituted phenol should produce a high-molecular-
weight linear polymer, since the aza-methylene linkage
is occurred at only the ortho positions in the structure
[eq. (3), Scheme 1].
3,4-Dihydro-1,3-2H-benzoxazines are known as heter-
ocyclic compounds obtained from p-substituted phenols,
formaldehyde, and primary amines in a molar ratio of
1:2:1 via the Mannich reaction [1]. Theoretically, ben-
zoxazines act as a monomer that undergoes the ring
opening reaction to obtain a polymer chain with methyl-
ene-amine-Mannich bridges at both the ortho and para
positions [2–6] [eq. (1), Scheme 1]. Kopf and Wagner
[7] proposed the oxazine-derived phenolics as a poten-
tial precursor in synthesis of novolacs. Benzoxazines are
reported as polybenzoxazines for the first time when
Ning and Ishida [6] demonstrated novel phenolic resins
obtained from a series of bisphenol A-based benzoxa-
zine monomers. The success in polymeric material of
bisphenol A-based polybenzoxazines might be related to
the crosslink structure generated from the two hydroxyl
groups belonging to a single bisphenol ring [eq. (2),
Scheme 1] [6].
Riess et al. [5] attempted the polymerization of phe-
nol by using various conditions, which are the type of
phenols, reaction temperatures, molar ratios, and phenol
initiator concentrations. The reactions were done in bulk
using NMR and vapor pressure methods to find that the
degree of polymerization of the main products were
about at the level of tetramer to hexamer. Although the
reason why polymerization terminated was not clarified,
the reaction was explained in terms of the kinetics and
mechanisms [5].
Until now, there is no report about the linear polyben-
zoxazines even the mechanism insists the stepwise reac-
tion to produce a chain of aza-methylene-phenol poly-
mer. For the past few years, our group has paid attention
onto the linear chain of polybenzoxazines, however, to
our surprise the polymerization of p-substituted phenol-
based benzoxazines does not proceed as expected. As
will be describe here, the many reactions we have car-
ried out allows us to conclude that the ring opening of
p-substituted phenols provide neither linear oligomer nor
polymer (Scheme 2), but only dimers.
It is important to note that, although the reaction and
mechanism of ring-opening benzoxazines [2] have been
studied since 1952, p-substituted phenol-based benzoxa-
zines and azamethylene phenols have rarely been
reported. Based on the reverse Mannich reaction [8], p-
C
V
2009 HeteroCorporation

July 2009
Self Termination of Ring Opening Reaction of p-Substituted Phenol-Based Benzoxazines:
An Obstructive Effect via Intramolecular Hydrogen Bond
715
Scheme 1
This article, thus, are to clarify (i) if indeed we never
chromatograms exhibit the results corresponding to
get the linear or cyclic compounds from the p-substi-
tuted based benzoxazine monomers as written in the for-
mula, (ii) why the p-substituted based benzoxazine
monomers give only dimer, and (iii) what obstructs the
polymerization. By solving these questions, we are able
not only to point out a rare example of a polymerization
that could not proceed according to theory but also to
design a synthesis pathway considering the factors
involved at the monomer level.
TLC. As shown in Figure 1, a new single sharp peak at
(tR) 3.430 min is observed for the white product,
whereas those for 1a and p-cresol are found at tR 3.344
and 3.418 min, respectively.
From FTIR spectra (Fig. 2), the white product gives
rise not only to the band at 3226 cmꢀ1 but also to a
broad band at 3100–2600 cmꢀ1. Lin-Vien et al. [10]
reported a broad band at 3200–2600 cmꢀ1 correspond-
ing to the strong OHꢁꢁꢁN bond. Thus, we speculate that
the white product with a free hydroxyl group in an open
ring benzoxazine forms an intermolecular hydrogen
bond with another free hydroxyl group and an intramo-
lecular hydrogen bond with an aza group. The band at
RESULTS AND DISCUSSION
Ring opening polymerization. Carboxylic acids and
phenol derivatives are known as acid catalysts for the
ring opening reaction of benzoxazines [5,9]. Ning and
Ishida [6] reported that the ring opening is preferable in
high-dielectric constant solvents such as methanol.
Scheme 2
After the mixture of 1a and p-cresol (molar ratio of
20:1) was refluxed in MeOH for 8 h, a white precipitate
appeared. The TLC of the MeOH solution shows two
spots at Rf for 0.48 and 0.56, referring to p-cresol and
1a, respectively. This confirmed that only starting mate-
rials were present. However, the white precipitate gives
an Rf for 0.30, implying a new product. The HPLC
Journal of Heterocyclic Chemistry
DOI 10.1002/jhet

716
S. Chirachanchai, A. Laobuthee, and S. Phongtamrug
Vol 46
Figure 3. 1H NMR spectrum of 4a.
Figure 1. HPLC chromatograms of (a) p-cresol, (b) 1a, and (c) 4a.
1502 cmꢀ1 in both 1a and the product corresponds to a
vibrational mode of a trisubstituted benzene. It is impor-
tant that no tetrasubstituted benzene band at 1485 cmꢀ1
lyzed by dissociation of phenol. To overcome the termi-
nation at dimer, we concentrated on the optimal amount
of p-cresol and solvent as well as reaction time and
temperature.
1
was observed. The H NMR spectrum of 1a shows two
Regardless of the variation of the molar ratio and sol-
vents, the white precipitate was obtained. The character-
ization by TLC, HPLC, FTIR, 1H NMR, and EA
showed that the product was the dimer 4a. In addition,
it was found that the generation of 4a is largely depend-
ent on the amount of p-cresol (Fig. 4). When the molar
ratio of the p-cresol increases, the yield of the dimer is
increased. Unexpectedly, at 1:1, the yield is highest for
all solvents. This implies that the reaction between 1a
and p-cresol might possibly be stoichiometric.
singlet peaks at dH ¼ 4.07 and 4.95 ppm, whereas that
of the white precipitate (Fig. 3) shows a singlet reso-
nance at dH ¼ 3.75 ppm belonging to methylene groups.
This suggests the existence of the aza-methylene linkage
in the product as a result of a ring opening reaction. If
the product is a polymer, four species of aromatic pro-
tons could be observed; however, only three protons at
dH 6.65 (d), 6.8 (s), and 6.90 (d) ppm resulted. Thus, we
speculated that the product was not a polymer but rather
a dimer as shown in 4a.
If p-cresol acts as an initiator in a reverse Mannich
reaction, a generated open-ring intermediate should
attack at the ortho position of either 1a or p-cresol (see
speculated mechanism). Thus, the product obtained
should be a mixture of dimer and mono-oxazine com-
pound as detailed in Scheme 3. However, the compound
Elemental analysis (EA) supports our conclusion. Ele-
mental analysis is C 77.90, H 8.56, and N 4.16, which
reflects exactly the dimer unit of 4a.
Effect of solvent and temperature. Bruke et al. [4]
reported the ring opening of benzoxazine monomer initi-
ated by an intermolecular hydrogen bond with phenol
derivatives. Riess et al. [5] proposed a mechanism cata-
Figure 4. Yields of the product obtained from the reaction of 1a and
p-cresol carried out at the boiling point of each solvent; MeOH (
iso-PrOH (n), iso-BuOH (~), cyclohexane (l), and xylene (h).
^),
Figure 2. FTIR spectra of (a) 1a and (b) 4a.
Journal of Heterocyclic Chemistry
DOI 10.1002/jhet

July 2009
Self Termination of Ring Opening Reaction of p-Substituted Phenol-Based Benzoxazines:
An Obstructive Effect via Intramolecular Hydrogen Bond
717
Scheme 3
Figure 6. HPLC chromatograms of the product obtained from 1a and
p-cresol at 65ꢂC in neat condition with various ratios; (a) 0.5:1, (b)
1:1, (c) 2:1, (d) 3:1, (e) 4:1, (f) 10:1, and (g) 20:1.
Neat liquid state reaction. It can be expected that
the polymerization of p-substituted phenol-based ben-
zoxazine might be favorable in neat condition. After
heating the mixture of 1a and p-cresol in various ratios
to the molten state, the white powder precipitated soon
after 30 min. Figure 6 shows that each product obtained
exhibits only a single component with tR 3.430 min.
The NMR spectra of all products were same as shown
in Figure 3, hence, we conclude that all products are 4a.
Figure 7 shows the comparative studies between neat
and solvent (xylene) conditions with a fixed molar ratio
of 1a and p-cresol of 1:1. Both reactions give the dimer
in high yield (ꢃ90%) at 65ꢂC. It is obvious that the
neat condition also provides a stoichiometric reaction
between 1a and p-cresol. When the reactions are carried
out at temperatures either lower or higher than 65ꢂC,
the yield is drastically decreased. This supports our
speculation about the effect of temperature.
obtained is a single component and its characterization
does not correspond to the mono-oxazine product [11].
Figure 4 also implies that the yields of 4a are
increased for all molar ratios when the reactions have
been carried out at low temperature. The yields from
lower reaction temperatures, such as MeOH (bp 65ꢂC)
and cyclohexane (bp 80ꢂC) are higher than those from
higher reaction temperatures, such as iso-BuOH (bp
110ꢂC) and mixed xylenes (bp 135ꢂC).
Figure 5 implies two important results related to the
optimum temperature and polarity effect. As expected,
the highest yields are obtained at 65ꢂC for every sol-
vent. Considering the solvent polarity, the nonpolar ones
(yield 40–60%) give higher yields than the polar ones
(18–35%). This implies that the reaction is preferably
carried out with nonpolar solvents (see speculated
mechanism).
It is important to point out that when the reaction was
carried out to satisfy the thermal initiation (above
Figure 5. Yields of the product obtained from 1a and p-cresol under
various temperatures: 65, 80, 110, and 135ꢂC in various solvents;
Figure 7. Yields of the product obtained from 1a and p-cresol at vari-
ous temperatures; 25, 40, 65, 80, 110, 135, 160, and 180ꢂC in
mixed xylenes and neat condition.
MeOH , iso-PrOH , iso-BuOH , xylene , and cyclohexane
.
Journal of Heterocyclic Chemistry
DOI 10.1002/jhet

718
S. Chirachanchai, A. Laobuthee, and S. Phongtamrug
Vol 46
Figure 8. Mass spectrum of the mixture obtained from 1a and p-cresol
at 180ꢂC.
Figure 9. Yields of the product obtained from 1a and p-cresol in vari-
ous temperatures; 25 (^), 40 (n), 65 (~), 80 (~), 110 (^), 135 (*),
160 (l), and 180ꢂC (ꢄ) in neat condition.
140ꢂC) condition as reported by Riess et al. [5], the pre-
cipitation of dimer 4a is decreased to 2–3% (Fig. 7).
Figure 7 also suggests that the reaction either in solution
(mixed xylenes) or neat give maximum yield at the
reaction temperature 65ꢂC. At the same time, the neat
liquid becomes a dark-brownish highly viscous liquid. A
further study by LC-MS indicates the various fragments
belonging to the dimer, appearing at m/z ¼ 340, as well
as other incomplete structures associated with the start-
ing materials (Fig. 8). The trace fragments (m/z ¼ 570
and 678) of higher molecular weight than dimer is also
observed. These fragments could result from dimer
aggregation or from trimer and tetramer. This implies
that high temperature does not favor the polymerization
but brings into play the competition between thermal
dissociation of benzoxazine and chain propagation as
reported by Riess et al. [5]
Reactivity of 1a and p-cresol. Figure 9 gives impor-
tant information about the reactivity of 1a and p-cresol
as related to the reaction temperature. Generally, stoichi-
ometric balance gives the highest yield with an equiva-
lent molar ratio between the two reactants.
At high-reaction temperatures such as 110 and 135ꢂC,
our results indicate that even a lesser amount (stoichi-
metric imbalance) of p-cresol in the system, such as a
molar ratio of 1a and p-cresol of 2:1, 3:1, and 4:1, pro-
vides a yield of 4a higher than that of equivalent molar
quantities (stoichiometric balance).
gives the maximum yield when compared with other
ratios. This shows that there is no deactivation of 1a at
the low temperature, and the high yield is obtained as
expected.
Taking the above discussion into consideration, we
speculated that without thermal degradation, the reactiv-
ity of p-cresol is ꢃ0.8–0.9. This is strongly supported
by the results that 4b–4c, 5a–5c, and 6a–6c (Scheme 4)
give similar yields (80–90%).
When we consider that there is no deactivation at
65ꢂC, it is natural to expect that all molar ratios should
give a yield of 4a of 90%. However, Figure 9 shows
that the yield of 4a decreases significantly with an
excess amount of 1a. This implies that the reaction of
1a and p-cresol could not proceed effectively. In other
words, the ring opening reaction occurs only when the
intermediate between 1a and p-cresol is effectively
formed.
Since only the dimer is obtained in every case, we
conclude that the p-substituted phenol in the reaction
does not provide a reactive site for another step in ring
Scheme 4
It should be noted that the stoichiometric ratio 1:1
and the effectiveness of this stoichiometric ratio are dif-
ferent due to the reactivity of the reactive species. In
this case, the reactive species are deactivated by heat
[5]. Thus, the high yield of 4a might result in the case
of stoichiometric imbalance because the reactive species,
1a, were always present in high concentration in the
system.
In the case of low temperatures (65ꢂC and 80ꢂC), the
yield from the reactant ratios in stoichiometric balance
Journal of Heterocyclic Chemistry
DOI 10.1002/jhet

July 2009
Self Termination of Ring Opening Reaction of p-Substituted Phenol-Based Benzoxazines:
An Obstructive Effect via Intramolecular Hydrogen Bond
719
Scheme 5
intermolecular hydrogen bonding between 1a and p-cre-
sol needs lesser thermal motion to maintain stability.
Thus, the effect of temperature to drive the reaction pro-
vides a dilemma. This might be the reason why we
found that the reaction can best proceed at a defined
temperature (at 65ꢂC) rather than at a higher or lower
temperature (below or higher than 65ꢂC).
In the case of solvent polarity, proton dissociation of
phenol results when a polar solvent is used. However, at
the same time, the hydrogen bond with the polar solvent
will stabilize that proton. In contrast, nonpolar solvents
promote the intermolecular hydrogen bonding between
the benzoxazine and phenol derivatives. The significant
yield of dimer in the nonpolar solvent might result from
phenol dissociation rather than intermolecular hydrogen
bonding. The effect of intermolecular hydrogen bonding
is much enhanced as evidenced from experiments in the
neat liquid state.
opening polymerization. In fact, Riess et al. [5] reported
that the in situ ring opening polymerization of p-substi-
tuted phenol initiated by disubstituted phenol (10%)
gave a range of dimers to octamers as evaluated by
NMR and vapor pressure methods. Although the struc-
tural characterization of the purified product in those
cases was not reported in detail, the results support
our speculation about some obstructive effects in
polymerization.
Stereo structure of dimer: A key factor for ob-
structive effect in polymerization. Previously, we
detailed the unique stereo structure of benzoxazine
dimer with strong inter- and intramolecular hydrogen
bonds. The stabilization of a symmetrical compound
through an intramolecular hydrogen bond inevitably
gives us an asymmetric compound [11]. The X-ray
structural analyses of dimers (4a–4c, 5a–5c, and 6a–6c)
shows clearly that p-substituted based benzoxazines
have inter- and intramolecular hydrogen bonds that sta-
bilize the compounds [11,13,14]. Thus, after the single
step of ring opening polymerization that produces the
dimer, the stability of the network of inter- and intramo-
lecular hydrogen bonding brings about the self termina-
tion. It can be concluded that stereo structure of dimer
is the key factor to terminate the polymerization.
It is known that the cyclization might be successful in
dilute conditions [12]. Thus, an attempt was made to
prepare cyclic compounds by using a mixture of 1a and
p-cresol (20:1) diluted in mixed xylenes to 5 ꢄ 10ꢀ3M.
These systems did not even give a white precipitate as
in the previous reactions and the reactants remained in
the solution.
Speculated mechanism. To explain why self termi-
nation occurs to prevent the polymerization, we combine
our results with the mechanisms proposed by Burke
et al. [4] and Riess et al. [5] We speculated that the
reaction proceeded as follows. As shown in Scheme 5,
intermolecular hydrogen bonding takes place between
benzoxazine and the free ortho position in the phenol.
Here, the ring opening of benzoxazine requires some
protonation from the phenol derivatives at the nitrogen
atom of the oxazine ring to further react with the phenol
derivative at ortho position. If the phenol derivative was
an initiator, after generating the dimer, the hydroxyl
group of one unit of the dimer would further attack
another benzoxazine molecule. As a result, a linear
polymer chain would be obtained. However, in our case,
it is important to note that the reaction terminates as
soon as the dimer is formed. This implies that phenol
derivatives act as a reactant, not as an initiator, since
the reaction does not give products other than dimer.
The effect of temperature and polarity might play an
important role (Scheme 5). At this step, an effective
ring opening requires thermal degradation, whereas the
CONCLUSIONS
Although, in theory, a linear polymer can be obtained
from the ring opening polymerization of p-substituted
phenol-based benzoxazines, this work shows that in
practice we will probably never achieve the polymer.
Even when the reaction conditions were varied in terms
of solvent, neat liquid state, reaction temperature, and
concentration, the product was inevitably the dimer.
Considering the factors involved in the ring opening
reaction, our results show that, in the initial step, the
hydrogen bonding between phenol derivatives and ben-
zoxazine is primary when compare with phenol dissocia-
tion. Combining this with the previous X-ray structure
analyses [11,13,14], we conclude that the obstructive
effect of dimer in polymerization might result from the
strong intramolecular hydrogen bond between the
Journal of Heterocyclic Chemistry
DOI 10.1002/jhet

720
S. Chirachanchai, A. Laobuthee, and S. Phongtamrug
Vol 46
Reaction in neat liquid state under various
temperatures. Mixtures of 1a and p-cresol (0.5:1, 1:1, 2:1,
3:1, 4:1, 10:1, and 20:1) were prepared and stirred at room
temperature (ꢃ25ꢂC), 40, 65, 80, 110, 135, 160, and 180ꢂC.
The mixtures were allowed to react until viscous. The precipi-
tates obtained from the reaction were collected, washed with
diethyl ether before drying at 60ꢂC for 6 h.
hydroxyl group of the phenol ring and the aza methyl-
ene group in the dimer.
Thus, the mechanism and the reaction assumed from
the formula are not always practical. As shown here, the
unique stereo structure of benzoxazine dimer leads to
the self termination and obstructs the polymerization as
clarified in Scheme 5.
Similarly, 1b–1c, 2a–2c, and 3a–3c were reacted with p-cre-
sol, 2,4-dimethylphenol, and 4-ethylphenol, respectively. The
compounds obtained were qualitatively analyzed by FTIR, 1H
NMR, HPLC, LC-MS, and EA. From structural analyses, the
compounds obtained are proposed as shown in Scheme 3.
2,20-[(Cyclohexylimino)di(methylene)]bis(4-methylphenol)
(4a). 80% yield; Rf ¼ 0.30 (5% MeOH in CHCl3); clear and
colorless solid; mp ¼ 181ꢂC; FTIR (KBr, cmꢀ1): 3226 (br,
OH), 1500 (vs, CAC), 1449 (m, NACH), 1249 (s, CAN),
1210 (m, CANAC), 819 (s, CANAC); 1H NMR (200 MHz,
CDCl3, ppm): dH 1.1 (m, 2H, CH2), 1.45 (m, 4H, CH2), 1.82
(m, 4H, CH2), 2.22 (s, 6H, CH3AAr), 2.70 (m, 1H, CH), 3.72
(s, 4H, ArACH2AN), 6.68 (d, 2H, ArAH), 6.85 (s, 2H,
ArAH), 6.90 (d, 2H, ArAH). Anal. calcd. for C22H29NO2: C,
77.88; H, 8.55; N, 4.13. Found: C, 77.90; H, 8.56; N, 4.16.
2,20-[(Propylimino)di(methylene)]bis(4-methylphenol) (4b).
80% yield; Rf ¼ 0.22 (5% MeOH in CHCl3); clear and color-
less solid; mp ¼ 149ꢂC; FTIR (KBr, cmꢀ1): 3251 (br, OH),
1501 (vs, CAC), 1467 (m, NACH2), 1276 (s, CAN), 1210 (s,
CANAC), 819 (s, CANAC); 1H NMR (200 MHz, CDCl3,
ppm): dH 0.87 (t, 3H, CH3ACH2ACH2AN), 1.65 (m, 2H,
CH3ACH2ACH2AN), 2.22 (s, 6H, CH3AAr), 2.50 (t, 2H,
CH3ACH2ACH2AN), 3.70 (s, 4H, ArACH2AN), 6.68 (d, 2H,
ArAH), 6.85 (s, 2H, Ar-H), 6.90 (d, 2H, ArAH). Anal. calcd.
for C19H25NO2: C, 76.25; H, 8.36; N, 4.69. Found: C, 76.28;
H, 8.31; N, 4.70.
EXPERIMENTAL
Chemicals. Paraformaldehyde was purchased from Sigma
(St. Louis, MO). p-Cresol, 2,4-dimethylphenol, 4-ethylphenol,
methylamine (40% w/v in water), cyclohexylamine, and pro-
pylamine, deuterated chloroform (CDCl3), and anhydrous so-
dium sulfate were purchased from Fluka Chemicals (Buchs,
Switzerland). HPLC grade tetrahydrofuran (THF), methanol,
propan-2-ol, mixed xylenes, 2-methylpropan-1-ol, cyclohexane,
sodium hydroxide, and diethyl ether were the products of Ajax
chemicals (Australia). All chemicals were analytical grade and
used as received.
Procedures. Benzoxazines, 1a–3c, were prepared by using
phenol, formaldehyde, and amine derivatives in the ratio of
1:2:1, respectively. p-Cresol was added into solution of cyclo-
hexylamine and p-formaldehyde, then refluxed for 6 h. The
crude product was washed and solvent was removed to obtain
1a. Compounds 1b and 1c were prepared similarly but using
propylamine and methylamine, respectively, instead of cyclo-
hexylamine. In the preparation of 2a–2c and 3a–3c, 2,4-dime-
thylphenol and 4-ethylphenol were used as phenol derivatives,
1
respectively. The H nuclear magnetic resonance (NMR) spec-
trometers were a Bruker ACF with a proton frequency of 200
MHz. Fourier transform infrared spectra were measured at a
resolution of 4 cmꢀ1 by a Bruker Equinox55/S spectrophotom-
eter equipped with deuterated triglycine (DTGS) detector
under constant purge with dry air. High-performance liquid
chromatography (HPLC) was done with a Hewlett Packard
HP1100 HPLC and a diode array detector model G1315A
#DE72002547 fixed at 254 nm. The samples were eluted
through a Whatman Partisil 5, a silica gel column with an av-
erage pore diameter of 8.5 nm, and a surface area > 350 m2/g
by maintaining the flow rate at 1 mL/min throughout the
experiment. Liquid chromatography mass spectrometer (LC-
MS) was a Bruker Esquire-LC using methanol as a mobile
phase. Elemental analysis (EA) was performed by a Perkin
Elmer 2400 Series II CHNS/O analyzer with a combustion
temperature of 975ꢂC and a reduction temperature of 500ꢂC.
Reaction of 1a and p-cresol in various solvents and
temperatures. Benzoxazine 1a (1.6 mmol) and p-cresol were
reacted in molar ratios of 1:1, 4:1, 10:1, and 20:1 in various
solvents (5 mL), methanol (MeOH), propan-2-ol (iso-PrOH),
2-methylpropan-1-ol (iso-BuOH), cyclohexane, and xylene.
The mixtures of monomer 1a and p-cresol in each solvent
were reacted at room temperature (ꢃ25ꢂC), 40, 65, 80, 110,
135, 160, and 180ꢂC. The completion of the reaction was fol-
lowed by thin layer chromatography (TLC) and the reaction
was stopped after 8 h. The solvent was removed and the crude
product was washed with diethyl ether several times before
drying at 60ꢂC for 6 h.
2,20-[(Methylimino)di(methylene)]bis(4-methylphenol) (4c).
90% yield; Rf ¼ 0.24 (5% MeOH in CHCl3); clear and color-
less solid; mp ¼ 163ꢂC; FTIR (KBr, cmꢀ1): 3271 (br, OH),
1499 (vs, CAC), 1456 (m, NACH3), 1249 (s, CAN), 1209 (m,
CANAC), 815 (vs, CANAC); 1H NMR (200 MHz, CDCl3,
ppm): dH 2.23 (s, 6H, ArACH3), 2.23 (s, 3H, NACH3), 3.69
(s, 4H, ArACH2AN), 6.70 (d, 2H, ArAH), 6.83 (s, 2H,
ArAH), 6.86 (d, 2H, ArAH). Anal. calcd. for C17H21NO2: C,
75.28; H, 7.75; N, 5.17. Found: C, 75.31; H, 7.77; N, 5.19.
2,20-[(Cyclohexylimino)di(methylene)]bis(4,6-dimethylphenol)
(5a). 90% yield; Rf ¼ 0.38 (5% MeOH in CHCl3); clear and
colorless solid; mp ¼ 152ꢂC; FTIR (KBr, cmꢀ1): 3384 (br,
OH), 1484 (vs, CAC), 1451 (m, NACH), 1245 (m, CAN),
1199 (m, CANAC), 858 (m, CANAC); 1H NMR (200 MHz,
CDCl3, ppm): dH 1.1 (m, 2H, CH2), 1.45 (m, 4H, CH2), 1.82
(m, 4H, CH2), 2.20 (s, 6H, CH3AAr), 2.22 (s, 6H, CH3AAr),
2.70 (m, 1H, CH), 3.72 (s, 4H, ArACH2AN), 6.70 (s, 2H,
ArAH), 6.85 (s, 2H, ArAH). Anal. calcd. for C24H33NO2: C,
78.47; H, 8.99; N, 3.82. Found: C, 78.49; H, 8.97; N, 3.85.
2,20-[(Propylimino)di(methylene)]bis(4,6-dimethylphenol)
(5b). 90% yield; Rf ¼ 0.43 (5% MeOH in CHCl3); clear and
colorless solid; mp ¼ 116ꢂC; FTIR (KBr, cmꢀ1): 3298 (br,
OH), 1483 (vs, CAC), 1450 (m, NACH2), 1250 (m, CAN),
1
1199 (vs, CANAC), 852 (m, CANAC); H NMR (200 MHz,
CDCl3, ppm): dH 0.85 (t, 3H, CH3ACH2ACH2AN), 1.65 (m,
2H, CH3ACH2ACH2AN), 2.20 (s, 6H, CH3AAr), 2.22 (s, 6H,
CH3AAr), 2.50 (t, 2H, CH3ACH2ACH2AN), 3.65 (s, 4H,
Journal of Heterocyclic Chemistry
DOI 10.1002/jhet

July 2009
Self Termination of Ring Opening Reaction of p-Substituted Phenol-Based Benzoxazines:
An Obstructive Effect via Intramolecular Hydrogen Bond
721
ArACH2AN), 6.70 (s, 2H, ArAH), 6.85 (s, 2H, ArAH). Anal.
calcd. for C21H29NO2: C, 77.06; H, 8.87; N, 4.28. Found: C,
77.05; H, 8.86; N, 4.27.
Acknowledgments. The authors thank Prof. Hatsuo Ishida (Case
Western Reserve University, USA) and Prof. Kohji Tashiro
(Toyota Technological Institute, Japan) for inspiration and fruit-
ful discussion. The authors thank Dr. Buncha Pulpoka (Chula-
longkorn University, Thailand) for NMR measurement, Dr.
Vanida Bhavakul, Dr. Nittaya Ketkaew, and Ms. Nunticha Tek-
chien (King Mongkut’s University of Technology Thonburi,
Thailand) for HPLC measurement and valuable comment. Appre-
ciation is expressed to Mrs. Dorothy B. Wittcoff and Dr. Harold
A. Wittcoff for their suggestions and comments. One of the
authors (S.C.) gratefully acknowledges the partial financial sup-
port of the Hitachi Scholarship Foundation. A.L. thanks the Thai-
land Research Fund (Grant No. MRG4880097).
2,20-[(Methylimino)di(methylene)]bis(4,6-dimethylphenol)
(5c). 80% yield; Rf ¼ 0.39 (5% MeOH in CHCl3); clear and
colorless solid; mp ¼ 123ꢂC; FTIR (KBr, cmꢀ1): 3399 (br,
OH), 1484 (vs, CAC), 1427 (m, NACH3), 1243 (m, CAN),
1214 and 1201 (m, CANAC), 847 (m, CANAC); 1H NMR
(200 MHz, CDCl3, ppm): dH 2.22 (s, 12H, ArACH3), 2.25 (s,
3H, NACH3), 3.68 (s, 4H, ArACH2AN), 6.72 (s, 2H, ArAH),
6.81 (s, 2H, ArAH). Anal. calcd. for C19H25NO2: C, 76.26; H,
8.36; N, 4.68. Found: C, 76.27; H, 8.34; N, 4.69.
2,20-[(Cyclohexylimino)di(methylene)]bis(4-ethylphenol) (6a).
80% yield; Rf ¼ 0.21 (5% MeOH in CHCl3); clear and color-
less solid; mp ¼ 170ꢂC; FTIR (KBr, cmꢀ1): 3251 (br, OH),
1499 (vs, CAC), 1450 (m, NACH), 1250 (s, CAN), 1207 (m,
CANAC), 818 (m, CANAC); 1H NMR (200 MHz, CDCl3,
ppm): dH 1.15 (t, 6H, CH3ACH2AAr), 1.15 (m, 2H, CH2),
1.45 (m, 4H, CH2), 1.82 (m, 4H, CH2), 2.52 (q, 2H,
CH3ACH2AAr), 2.70 (m, 1H, CH), 3.72 (s, 4H,
ArACH2AN), 6.72 (d, 2H, ArAH), 6.87 (s, 2H, ArAH), 6.94
(d, 2H, ArAH). Anal. calcd. for C24H33NO2: C, 78.47; H,
8.99; N, 3.82. Found: C, 78.51; H, 8.97; N, 3.79.
2,20-[(Propylimino)di(methylene)]bis(4-ethylphenol) (6b). 80%
yield; Rf ¼ 0.28 (5% MeOH in CHCl3); clear and colorless
solid; mp ¼ 132ꢂC; FTIR (KBr, cmꢀ1): 3265 (br, OH), 1499
(vs, CAC), 1447 (m, NACH2), 1247 (s, CAN), 1205 (m,
CANAC), 819 (s, CANAC); 1H NMR (200 MHz, CDCl3,
ppm): dH 0.87 (t, 3H, CH3ACH2ACH2AN), 1.18 (t, 3H,
CH3ACH2AAr), 1.65 (m, 2H, CH3ACH2ACH2AN), 2.52 (q,
2H, CH3ACH2AAr), 2.52 (t, 2H, CH3ACH2ACH2AN), 3.70
(s, 4H, ArACH2AN), 6.72 (d, 2H, ArAH), 6.87 (s, 2H,
ArAH), 6.94 (d, 2H, ArAH). Anal. calcd. for C21H29NO2: C,
77.06; H, 8.87; N, 4.28. Found: C, 77.08; H, 8.89; N, 4.31.
2,20-[(Methylimino)di(methylene)]bis(4-ethylphenol) (6c). 90%
yield; Rf ¼ 0.34 (5% MeOH in CHCl3); clear and colorless
solid; mp ¼ 130ꢂC; FTIR (KBr, cmꢀ1): 3301 (br, OH), 1499
(vs, CAC), 1460 (m, NACH3), 1251 (s, CAN), 1207 (m,
CANAC), 821 (s, CANAC); 1H NMR (200 MHz, CDCl3,
ppm): dH 1.17 (t, 6H, ArACH2ACH3), 2.25 (s, 3H, NACH3),
2.54 (q, 4H, ArACH2ACH3), 3.72 (s, 4H, ArACH2AN), 6.73
(d, 2H, ArAH), 6.87 (s, 2H, ArAH), 6.94 (d, 2H, ArAH).
Anal. calcd. for C19H25NO2: C, 76.26; H, 8.36; N, 4.68.
Found: C, 76.24; H, 8.35; N, 4.65.
REFERENCES AND NOTES
[1] Burke, W. J. J Am Chem Soc 1949, 71, 609.
[2] Burke, W. J.; Smith, R. P.; Weatherbee, C. J Am Chem
Soc 1952, 74, 602.
[3] Burke, W. J.; Mortenson Glennie, E. L.; Weatherbee, C. J
Org Chem 1964, 29, 909.
[4] Burke, W. J.; Bishop, J. L.; Mortenson Glennie, E. L.;
Bauer, W. N., Jr. J Org Chem 1965, 30, 3423.
[5] Riess, G.; Schwob, J. M.; Guth, G.; Roche, M.; Laude, B.
In Advances in Polymer Synthesis; Culbertson, B. M., McGrath, J. E.,
Eds.; Plenum: New York, NY, 1985; pp 27–49.
[6] (a) Ning, X.; Ishida, H. J Polym Sci Part A: Polym Chem
1994, 32, 1121; (b) Ning, X.; Ishida, H. J Polym Sci Part B: Polym
Phys 1994, 32, 921.
[7] Kopf, P. W.; Wagner, E. R. J Polym Sci Polym Chem Ed
1973, 11, 939.
[8] (a) Tramontini, M.; Angiolini, L. Mannich Bases: Chemis-
try and Uses; CRC Press: Tokyo, Japan, 1994; (b) Fryhle, C; Solo-
mons, G. Organic Chemistry; Wiley: New York, NY, 2000; pp 900.
[9] Dunkers, J.; Ishida, H. Spectrochim Acta [A] 1995, 51, 855.
[10] Lin-Vien, D.; Colthup, N. B.; Fateley, G. W.; Grassel, G. J.
In The Hanbook of Infrared and Raman Characteristic Frequencies of
Organic Molecules; Academic Press: San Diego, CA, 1991; pp 296.
[11] Laobuthee, A.; Chirachanchai, S.; Ishida, H.; Tashiro, K. J
Am Chem Soc 2001, 123, 9947.
[12] Ebdon, J. R.; Eastmond, G. C. New Methods of Polymer
Synthesis; Chapman & Hall: UK, 1995; pp 197.
[13] Dunkers, J.; Zarate, E. A.; Ishida, H. J Phys Chem 1996,
100, 13514.
[14] Phongtamrug, S.; Tashiro, K.; Miyata, M.; Chirachanchai,
S. J Phys Chem B 2006, 110, 21365.
Journal of Heterocyclic Chemistry
DOI 10.1002/jhet