Polybenzoxazines from renewable diphenolic acid

Article abstract of DOI:10.1002/pola.24541

The novel benzoxazine monomers, DPA-Bz and MDP-Bz from renewable diphenolic acid (DPA), which mimics the structure of bisphenol A (BPA), were synthesized by traditional approaches. The structure and purity of the monomers was confirmed by FTIR, ¹

Full text of DOI:10.1002/pola.24541

Polybenzoxazines from Renewable Diphenolic Acid
´ ˜
`
´
C. ZUNIGA, M. S. LARRECHI, G. LLIGADAS, J. C. RONDA, M. GALIA, V. CADIZ
Departament de Qu´ımica Anal´ıtica i Qu´ımica Orga`nica, Universitat Rovira i Virgili. Campus Sescelades,
Marcel.l´ı Domingo s/n. 43007, Tarragona, Spain
Received 23 November 2010; accepted 6 December 2010
DOI: 10.1002/pola.24541
Published online 10 January 2011 in Wiley Online Library (wileyonlinelibrary.com).
ABSTRACT: The novel benzoxazine monomers, DPA-Bz and
MDP-Bz from renewable diphenolic acid (DPA), which mimics
the structure of bisphenol A (BPA), were synthesized by tradi-
tional approaches. The structure and purity of the monomers
was confirmed by FTIR, 1H NMR, and 13C NMR spectra. The
thermally activated polymerization of the MDP-Bz and DPA-Bz
afforded thermosetting polybenzoxazines with higher Tg’s,
270 ꢀC and 208 ꢀC respectively, and higher crosslinking density
compared to BPA-Bz, due to the transesterification or esterifica-
tion reactions occurred during curing process. These reactions
are in accordance with the number of independent reactions
determined analyzing by SVD the chemical rank of the IR
spectra data matrices recorded along the homopolymerization
reactions monitored at 200 ꢀC. Spectral and concentration
profiles of the active chemical species involved in these pro-
C
V
cesses were obtained by MCR-ALS.
2011 Wiley Periodicals,
Inc. J Polym Sci Part A: Polym Chem 49: 1219–1227, 2011
KEYWORDS: benzoxazine; crosslinking; diphenolic acid; infrared
spectroscopy; renewable resource
INTRODUCTION The development of environmentally com-
patible polymers is one of the current challenges in polymer
chemistry. The scarcity of nonrenewable resources encour-
aged the scientific community to develop and commercialize
new biobased products that can alleviate the wide-spread
dependence on fossil fuels and, enhance security, the envi-
ronment and the economy.1 Polymers from renewable
resources will play an increasing role because of economical
and environmental benefits and the new property profiles
that renewable polymers can exhibit, such as biocompati-
bility and biodegradability.2
complete curing and the brittleness of the cured materials
that can sometimes limit their potential applications. To
properly address these issues and overcome the associated
disadvantages, several strategies such as: preparation of
monomers with additional functionality, synthesis of novel
polymeric precursors, and blending with a high-performance
polymer or filler and fiber, have been attempted.7
The synthesis of bisphenol A (BPA) based benzoxazine has
been reported,8 and its polymerization yields thermosets
with high structural integrity and thus the materials would
possess very good properties. In recent decade, it has
attracted much attention from researchers to synthesize
new polyesters and polycarbonates using diphenolic acid
(DPA)9–12 instead of BPA, because DPA has a structure
similar to BPA. The research interest resulted from that DPA
is commercially available and much cheaper than BPA and it
can introduce functional carboxyl group into the polymer
structure. Moreover, DPA is a condensation product of
phenol and levulinic acid. Levulinic acid is believed to be a
cheap platform chemical and can be commercially produced
from cellulose-rich biomass (especially from waste biomass)
in large scale.13
Polybenzoxazine as a novel phenolic type thermoset has
been developed to overcome the short-comings associated
with traditional phenolics resins such as releasing condensa-
tion by-products and using strong acids as catalysts, while
retaining good thermal properties and flame retardancy of
phenolic resins.3 It can be prepared from benzoxazines via
thermally induced ring opening polymerisation.4 Benz-
oxazines are readily synthesized either in solution or by a
melt-state reaction using a combination of a phenolic deriva-
tive, formaldehyde, and a primary amine.5 Polybenzoxazines
have not only the advantageous properties of conventional
phenolic resins but also other interesting advantages such as
heat resistance, superior electronic properties, low water
absorption, low surface energy, and excellent dimensional
In this article, we report the synthesis of DPA based benz-
oxazine (DPA-Bz) and the DPA ester derivative, MDP, based
benzoxazine (MDP-Bz), (see Scheme 1), and compare the
polymerization and properties of both DPA and MDP based
benzoxazines with those of the standard BPA based benz-
oxazine (BPA-Bz). The presence of a carboxylic acid in DPA
monomer should lower the high temperature needed to
stability.6 Although polybenzoxazines offer
a variety of
advantages, pure polybenzoxazine-based polymers also suffer
number of disadvantages too. The disadvantages of the typi-
cal polybenzoxazines are the high temperature needed for
Correspondence to: V. Ca´diz (E-mail: virginia.cadiz@urv.cat)
C
V
Journal of Polymer Science Part A: Polymer Chemistry, Vol. 49, 1219–1227 (2011) 2011 Wiley Periodicals, Inc.
´ ˜
POLYBENZOXAZINES FROM RENEWABLE DPA, ZUNIGA ET AL.
1219

JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY DOI 10.1002/POLA
SCHEME 1 Synthesis of DPA-Bz and MDP-Bz.
complete curing of these benzoxazines. We have previously
found that the presence of carboxyl groups influences
thermal curing and the processing temperatures are lower.14
The curing behavior of these materials was studied by differ-
ential scanning calorimetry (DSC) and FTIR.
acid (0.06 mol) and 240 mL of toluene were placed into
500 mL two-necked round-bottom flask. The reaction mix-
ture was heated at 110 C for 6 h. The resulting light orange
solution was filtered and concentrated under vacuum to
obtain a syrup that was subsequently dried under high vacuum
giving a yellowish solid with a yield of 95% of DPA-Bz.
ꢀ
Conventional spectroscopy and chemometric analysis of the
FTIR spectra data obtained in monitoring both benzoxazine
homopolymerizations were carried out. The number of the
independent reactions that take place in the processes were
evaluated from the rank of the matrix data by singular value
decomposition (SVD).15 Spectra and concentration profiles of
the representative chemical species involved in the reactions
were obtained applying multivariate curve resolution-alter-
nating least square (MCR-LS).16
1H NMR (CDCl3/TMS, d ppm): 7.30–6.70 (16H, ArAH), 5.34
(4H, s, OACH2AN), 4.58 (4H, s, ArACH2AN), 2.35 (2H, t),
2.13 (2H, t), 1.53 (3H, s).13C NMR (CDCl3, d ppm): 179.9 (s),
152.5 (s), 148.5 (s), 141.1 (s), 129.4 (d), 126.9 (d), 125.3
(d), 121.4 (d), 120.4 (s), 118.1 (d), 116.7 (d), 79.2 (t), 50.7
(t), 44.6 (s), 36.4 (t), 30.2 (t), 27.8 (q). IR: mcmꢁ1) 1702
(C¼¼O st)
Synthesis of Methyl-4,40-bis(4-hydroxyphenyl)
Pentanoate
Finally, the properties of the materials were evaluated by
thermogravimetric analysis (TGA) and dynamic mechanical
analysis (DMTA).
A 500 mL round-bottom flask equipped with a condenser
was charged with 4,40-bis(4-hydroxyphenyl) pentanoic acid
(0.26 mol), p-toluenesulfonic acid monohydrate (2.6 ꢂ 10ꢁ4
mol), trimethyl orthoformate (0.38 mol) and 200 mL of
methanol. The mixture was refluxed for 16 h, the methanol
was removed under vacuum and the resulting syrup was
dissolved in ethyl ether and washed several times with a
NaHCO3 saturated solution and with water. Finally, the ethyl
ether was removed with a rotary evaporator and the product
was dried under vacuum at room temperature. The resulting
product was milled and used without further purification in
the next step. A yield of 98% was obtained.
EXPERIMENTAL
Materials
The following chemicals were obtained from the sources
indicated and used as received: ammonium sulfate (Probus),
paraformaldehyde (Probus), 4,40-bis(4-hydroxyphenyl) penta-
noic acid (Aldrich), trimethyl orthoformate (Fluka), stearic
acid (Aldrich), methylstearate (Aldrich), 2,4-dimethylphenol
(Fluka), p-toluenesulfonic acid monohydrate (Panreac), and
benzyldimethylamine (Fluka). The 1,3,5-triphenylhexahydro-
1,3,5 triazine and Bisphenol A based benzoxazine (BPA-Bz)
were synthesized according to the reported procedure.17,18
All solvents were purified by standard procedures.
1H NMR (CDCl3/TMS, d ppm): 7,05 (4H, d, Jo ¼ 6.8 Hz), 6.73
(4H, d, Jo ¼ 6.8), 4.68 (2H, s, OH) 3.62 (3H, s, OCH3), 2.39
(2H, t), 2.10 (2H, t), 1.56 (3H, s).
Synthesis of 4,40-Bis-[6-(3-phenyl-3,4-dihydro-2H-1,
3-benzoxazine)] Pentanoic Acid (DPA-Bz)
1,3,5-Triphenylhexahydro-1,3,5-triazine (0.04 mol), parafor-
maldehyde (0.12 mol), 4,40-bis(4-hydroxyphenyl) pentanoic
13C NMR (CDCl3, d ppm): 172.8 (s), 153.9 (s), 138.3 (s),
126.8 (d), 113.6 (d), 50.2 (q), 42.9 (s), 35.4 (t), 28.8 (t), 26.4
(q). IR: mcmꢁ1) 1700 (C¼¼O st).
1220
WILEYONLINELIBRARY.COM/JOURNAL/JPOLA

ARTICLE
TABLE 1 DSC Data and Curing Conditions of the Benzoxazine Monomers
Tonset
(ꢀC)
Tmax
(ꢀC)
DHo
Sample
(kJ/mol)
Curing Conditions
BPA-Bz
DPA-Bz
MDP-Bz
210
150
210
256
192
261
122
154
129
190 ꢀC (2 h), 215 ꢀC (2 h), 235 ꢀC (8 h), 260 ꢀC (2 h)
140 ꢀC (2 h), 160 ꢀC (2 h), 180 ꢀC (8 h), 190 ꢀC (2 h)
190 ꢀC (2 h), 215 ꢀC (2 h), 235 ꢀC (8 h), 260 ꢀC (2 h)
Synthesis of Methyl 4,40-bis-[6-(3-phenyl-3,
Calorimetric studies were carried out on a Mettler DSC821e
thermal analyzer using N2 as a purge gas (100 mL/min) at
scanning rate of 20 ꢀC/min. Thermal stability studies were
carried out on a Mettler TGA/SDTA851e/LF/1100 with N2
4-dihydro-2H-1,3-benzoxazine)] Pentanoate (MDP-Bz)
A 1 L two-necked round-bottom flask equipped with a mag-
netic stirrer and a condenser was charged with 1,3,5-tri-
phenylhexahydro-1,3,5-triazine (0.08 mol), paraformaldehyde
(0.24 mol), methyl 4,40-bis(4-hydroxyphenyl)pentanoate
(methyl diphenolate) (0.12 mol) and 480 mL of toluene. The
mixture was refluxed at 110 ꢀC for 20 h. After cooling to
room temperature, the benzoxazine monomer was filtered,
dissolved in ethyl ether, and washed five times with a 3N
NaOH aqueous solution and with water. After washing, the
organic phase was dried over sodium sulfate and the solvent
was evaporated under vacuum at room temperature. A pale
yellowish solid was obtained with a yield of 93%.
ꢀ
or air as a purge gas at scan rates of 10 C/min. Mechanical
properties were measured using a dynamic mechanical
thermal analysis (DMTA) apparatus (TA DMA 2928). Speci-
mens (10 ꢂ 6 ꢂ 1.9 mm) were tested in a three point bend-
ing configuration. The thermal transitions were studied in
the 35–320 ꢀC range at a heating rate of 3 ꢀC/min and at a
fixed frequency of 1 Hz.
Data Acquisition of FTIR Spectra
The data obtained correspond to the FTIR/ATR spectra
recorded between 600 and 2000 cmꢁ1 every 0.482 cmꢁ1 for
DPA-Bz and MDP-Bz homopolymerizations. The reactions
were monitored along 4.5 h for DPA-Bz and 15 h for MDP-
Bz. The bands at 1490, 1230, and 940 cmꢁ1 associated to
the benzoxazine ring, were used to follow the progress of
the reaction. In the DPA-Bz homopolymerization, 71 spectra
were obtained. In the MDP-Bz homopolymerization, 180
spectra were obtained, and 70 spectra were selected for
analysis, corresponding to the obtained each 5 min during
the first hour, each 10 min during the second hour and the
rest each 15 min. The spectra recorded in the FTIR/ATR
were exported and converted into MATLAB binary files,19 to
carry out the chemometric analysis.
1H NMR (CDCl3/TMS, d ppm): 7.32–6.74 (16H, ArAH), 5.37
(4H, s, OACH2AN), 4.62 (4H, s, ArACH2AN), 3.63 (3H, s,
OCH3), 2.40 (2H, t), 2.14 (2H, t), 1.56 (3H, s).
13C NMR (CDCl3, d ppm): 174.4 (s, C¼¼O), 152.5 (s), 148.5
(s), 141.2 (s), 129.3 (d), 126.9 (d), 125.3 (d), 121.3 (s),
120.3 (s), 118.0 (d), 116.6 (d), 79.2 (t, O-CH2-N), 51.7 (q,
OCH3), 50.7 (t, Ar-CH2-N), 44.6 (s), 36.6 (t), 30.1 (t), 27.8
(q). IR: mcmꢁ1) 1735 (C¼¼O st)
Crosslinking Reaction
Polybenzoxazine samples were prepared with a manual hy-
draulic press 15-ton sample pressing (SPECAC) equipped
with a water cooled heated platens. Previously, each mono-
mer was degassed in a glass oven for 30 min at 140 ꢀC.
Samples were compressed in rectangular steel mold (80 mm
ꢂ 51,5 mm ꢂ 3 mm) under a pressure of 9.2 MPa. Because
of every monomer has different thermal polymerization
behavior, the curing temperatures were adjusted according
to eachꢀDSC thermogram. BPA-Bz and MDP-Bz were heated
The recorded spectra were arranged in two matrices M (71
ꢂ 2905) and N (70 ꢂ 2905) whose rows were the number
of recorded spectra and whose columns were the wave-
lengths of DPA-Bz and MDP-Bz reactions, respectively.
RESULTS AND DISCUSSION
1,3,5-Triphenylhexahydro-1,3,5-triazine is an active inter-
mediate and precursor in the synthesis of aromatic amine-
based benzoxazine structures.17,18 It is therefore possible
to obtain this triazine from aniline and paraformaldehyde,
which reacts with BPA and paraformaldehyde leading to
the bis(3,4-dihydro-2H-3-phenyl-1,3-benzoxazinyl) isopropane
(BPA-Bz) in high yield and purity. In a similar way, the syn-
thetic route for DPA-Bz (Scheme 1) uses triazine and DPA to
obtain methyl 4,40-bis-[6-(3-phenyl-3,4-dihydro-2H-1,3-ben-
zoxazine)] pentanoic acid (DPA-Bz) also with high yield. It is
worth to note that the resulting benzoxazine was obtained in
high purity, thus, no further purification procedure was
necessary. Probably in this case, byproducts and oligomeric
species form the insoluble fraction that was removed by
filtration. Likewise, previous esterification of DPA with
ꢀ
ꢀ
ꢀ
at 190 C for 2h, 215 C for 2h, 235 C for 8 h, and 260 C
for 2h, while DPA-Bz was heated at 140 ꢀC for 2h, 160 ꢀC
ꢀ
ꢀ
for 2h, 180 C for 8 h, and 190 C for 2h. (Table 1)
Instrumentation
The FTIR espectra were recorded on a JASCO 680 FTIR spec-
trophotometer with a resolution of 4 cmꢁ1 in the absorbance
and transmittance modes. An attenuated total reflection
(ATR) accessory with thermal control and a diamond crystal
(Golden Gate heated single-reflection diamond ATR, Specac.
Teknokroma) was used to determine FTIR spectra. 1H
(400 MHz) and 13C (100.5 MHz) NMR spectra were obtained
using a Varian Gemini 400 spectrometer with Fourier trans-
form and CDCl3 as solvent.
´ ˜
POLYBENZOXAZINES FROM RENEWABLE DPA, ZUNIGA ET AL.
1221

JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY DOI 10.1002/POLA
resonances of ANACH2AOA and ANACH2APh of the
oxazine ring, respectively. No detectable resonances due to
impurities are seen in the spectrum.
Moreover, the expected structures of the monomers were
further evidenced by FTIR. For both monomers, the presence
of cyclic ether of benzoxazine structure is confirmed by
the absorbance peaks at about 1230 and 1035 cmꢁ1 due
to the CAOAC symmetric stretching and asymmetric stretch-
ing modes respectively. The characteristic band at 952–
942 cmꢁ1 due to the CAH out-of-plane deformation mode
indicated the formation of oxazine rings.20
It is well-known that 1,3-benzoxazines exhibit exothermic
ring opening reaction around 200–250 ꢀC, which can be
monitored by DSC. Figure 2 shows the DSC plots of the
synthesized systems DPA-Bz, MDP-Bz, and the BPA-Bz for
comparison. Table 1 shows the collected onset and maximum
temperatures of the crosslinking exotherms. As can be seen,
the ring opening of benzoxazines of MDP-Bz and BPA-Bz
occur at very similar temperature, about 260 ꢀC. However,
the polymerization temperature observed for DPA-Bz was
quite low compared to those of the ordinary benzoxazines.
The thermogram reveals a broad exotherm with an onset at
150 ꢀC and a maximum at 190 ꢀC with 154 kJ/mol as the
exothermic energy. As NMR data show high purity mono-
mers free of oligomers, the low polymerization temperature
of this monomer can directly be attributed to monomer
characteristic, that is, to the presence of carboxylic group.
FIGURE 1 1H and 13C NMR spectra of DPA-Bz.
methanol, the resulting pentanoate reacts with triazine and
paraformaldehyde leading to the methyl 4,40-bis-[6-(3-phenyl-
3,4-dihydro-2H-1,3-benzoxazine)] pentanoate (MDP-Bz).
The structure of the benzoxazine monomers was confirmed
by spectral analysis. As can be seen from Figure 1 the
nuclear magnetic resonance spectroscopy (1H NMR) spec-
trum of the DPA-Bz exhibits not only the specific signals of
the benzoxazine ring, but also chemical shifts that belong to
the alkyl chain and the aromatic signals. Notably, while the
two signals at 5.3 and 4.6 ppm correspond to ACH2 protons
of benzoxazine ring, methyl protons of the pentanoic acid
appears at 1.5 ppm (singlet, 3H) and methyl protons of the
pentanoate at 3.6 ppm (in the case of MDP-Bz). Alkyl pro-
tons of the propionic acid or propionate moiety resonate at
2.1 ppm (triplet, CACH2, 2H) and 2.3 ppm (triplet,
ACH2ACOOH, 2H) or 2.4 ppm (triplet, ACH2ACOOCH3, 2H).
The COOH proton is not observed when CDCl3 is used as
solvent. The lack of any resonance around 3.7 ppm, typical
frequency for ring opened oligomers due to the Mannich
bridge protons of open oxazine rings: ACH2ANPhACH2A
structure, indicates that all the monomers are essentially
free of ring-opened oligomers, and the purification proce-
dure, when it was needed, was good enough to obtain
samples with high purity. Moreover, integration analysis of
the oxazine and aromatic proton peaks pointed out that the
closed-ring structured monomer is about 99%. Specially, it is
important to have high purity monomers free of phenolic
impurities and ring-opened oligomeric species for the evalu-
ation of the polymerization rate as these species lower the
polymerization temperature of the monomers acting as acid
cationic initiators. The corresponding 13C NMR spectrum
also supports the structure of these compounds. The two
singlets at 79.2 ppm and 50.7 ppm are typical of the carbon
According to DSC data the curing benzoxazine resins were
carried out in a mold with compression by heating samples
at different temperatures and times collected in Table 1.
The dynamomechanical properties of the polybenzoxazines
were investigated by DMTA. The dynamic mechanical behav-
ior of the cured benzoxazine resins was obtained as a
function of the temperature beginning in the glassy state of
each composition to the rubbery plateau of each material
[Fig. 3(a)]. For the thermoset polymerized from benzoxazine
FIGURE 2 DSC plots of DPA-Bz, MDP-Bz and BPA-Bz.
1222
WILEYONLINELIBRARY.COM/JOURNAL/JPOLA

ARTICLE
DMTA enables Tg’s of the crosslinked materials to be deter-
mined. It is detected as the maximum of the loss modulus
(E00), which corresponds to the initial drop from the glassy
state into the transition. Moreover, the a-relaxation peak of
the loss factor, tan d, is associated with Tg and corresponds
to the transition midpoint of the log of the E0curve. As
shown in Figure 3(b), the Tg, determinedꢀ as the peak
ꢀ
temperature of tan d curve is as high as 270 C and 208 C
for MDP-Bz and DPA-Bz, respectively. The Tg values from
both measurements are shown in Table 2. As expected, Tg as
tan d peak is higher than E’’ peak. In this table is also
collected the Tg value, as 1/2DCp, determined by DSC, which
is in accordance with DMTA values. The analysis of the
height of the tan d peak values indicates a higher crosslink-
ing density for MDP-Bz and DPA-Bz compared to BPA-Bz.
This could be explained by the transesterification or esterifi-
cation of the acidic group and the phenolic OH, which is
resulted from ring opening of oxazine, respectively. Similar
condensation was previously observed in the poly(amic
acid)-phenolic OH condensation22 and in benzoxazines with
p-carboxyphenyl linkage.23
To gain more insight about the curing processes, the most
common method used to monitor the benzoxazine ring open-
ing, DSC, do not provide direct information about the mecha-
nism of the reaction. The number of species involved in the
curing reaction can be obtained by separation techniques
such as high performance liquid chromatography.24 This is
essentially an offline technique and it is often used for the
kinetic investigation of reactions that do not evolve towards
solid products. Spectroscopic techniques such as FTIR/ATR
and chemometrics analysis of the recorded data through the
reaction make it possible to monitor the process on line
and provided useful information about the number of the
chemical structures evolved in the curing reaction. We stud-
ied the behavior of DPA-Bz and MDP-Bz polymerizations
by monitoring by FTIR/ATR spectroscopy in isothermal
FIGURE 3 (a) Tan delta and (b) Storage modulus versus tem-
perature of DPA-Bz, MDP-Bz, and BPA-Bz.
monomers, the storage modulus is maintained approximately
the same value for a wide temperature range up to 150 ꢀC.
The crosslinking density of a polymer can be estimated from
the plateau of the elastic modulus in the rubbery state.21
However, this theory is strictly valid only for lightly cross-
linked materials, and was therefore used only to make quali-
tative comparisons of the level of crosslinking among the
various polymers. BPA-Bz shows the lowest crosslinking
density while DPA-Bz and MDP-Bz show higher crosslinking
density.
ꢀ
experiments at 200 C.
This technique allowed us to follow the evolution of the
functional groups involved in the process by means of the
variation in the corresponding absorptions. Figure 4 shows
FTIR spectra of DPA-Bz and MDP-Bz polymerizations at
different times. These spectra show the characteristics
benzoxazine bands at about 1230 cmꢁ1, 1030 cmꢁ1, and
940 cmꢁ1, which follow a progressive diminution with time.
However, as can be seen, the evolution of benzoxazine bands
TABLE 2 Thermal and Thermogravimetric Properties of Cured Polybenzoxazines
Tg (ꢀC)
TGA (N2)
TGA (O2)
Sample
1= DC
E00
Tan dmax
T10% (ꢀC)a
Tmax (ꢀC)b
Y800 (%)c
T10% (ꢀC)a
Tmax (ꢀC)b
Y800 (%)c
2
p
max
BPA-Bz
DPA-Bz
MDP-Bz
a
170
207
265
154
185
235
172
208
270
366
348
379
427
425
420
c
34
32
31
386
353
388
418
423
418
18
7
12
Temperature of 10% of weight loss.
Char yield at 800 ꢀC.
b
Temperature of the maximum weight loss rate.
´ ˜
POLYBENZOXAZINES FROM RENEWABLE DPA, ZUNIGA ET AL.
1223

JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY DOI 10.1002/POLA
FIGURE 4 FTIR-ATR spectra of DPA-Bz and MDP-Bz homopolymerizations at 200 ꢀC.
is different in both processes, being much faster in DPA-Bz.
Moreover, the trisubstituted benzene ring bands in the
benzoxazine structures at about 1490 cmꢁ1, 820 cmꢁ1, and
690 cmꢁ1 follow the same trend as crosslinking proceeded,
decreasing as a tetrasubstituted benzene rings appear, and
consequently much faster in DPA-Bz. Both figures show the
different evolution in the bands corresponding to carbonyl
groups, in the zone around 1700 cmꢁ1, of DPA-Bz and MDP-
Bz. From the first stages of the reaction of DPA-Bz, two
carbonyl bands can be distinguished corresponding to the
phenolic ester (about 1728 cmꢁ1) and the free carboxylic
group (about 1710 cmꢁ1), however at the beginning of
the MDP-Bz homopolymerization, only a strong band at
1735 cmꢁ1 attributable to initial methyl ester carbonyl
group is observed, which decreases with time at once other
absorptions appear, as can be seen in Figure 5 where the
expanded carbonylic zone is shown. So, at 135 min (spec-
trum 3), two different absorptions are clearly detected: one
the aforementioned at about 1735 cmꢁ1 and a second at
about 1725 cmꢁ1 attributable to phenolic ester carbonyl
group as a consequence of the transesterification process
that presumably has taken place. When the reaction makes
progress up to 435 min (spectrum 4), a third absorption, at
1712 cmꢁ1, is detected that could be due to the carbonyl
group of a carboxylic acid from a partial hydrolysis of the
ester group.
Additional information about the number of independent
reactions involved in the homopolymerization processes can
be obtained from the rank analysis of the spectra data matri-
ces M and N evaluated by singular value decomposition
(SVD).15 As example in Table 3, the values of the 8 first
singular values are showed. Considering as threshold the sin-
gular values associated to the noise (k ¼ 0.5), in both M
(DPA-Bz) and N (MDP-Bz) matrices, four factors are signifi-
cantly different from noise, thus, the chemical rank is four.
Analysis of the rank associated with matrices of data
obtained in the study of a chemical reaction has been widely
discussed by Amrhein et al.25 who showed that in this type
of matrix data the rank is equal to minimum of (r þ 1, c)
where r is the number of independent reactions in the chem-
ical system and c is the number of absorbent species. This
TABLE 3 Rank Analysis of Matrices M and N
Eigen Value (k)
DPA-Bz
MDP-Bz
1
2
3
4
5
6
7
8
68.4914
3.6771
1.0860
0.5678
0.1816
0.1431
0.1160
0.1108
57.1656
4.9428
1.6592
0.3945
0.2459
0.1629
0.1294
0.1252
FIGURE 5 Expanded carbonylic zone of FTIR spectra of MDP-Bz
polymerization at 5 min (1), 45 min (2), 135 min (3), 435 min
(4), 885 min (5). [Color figure can be viewed in the online issue,
which is available at wileyonlinelibrary.com.]
1224
WILEYONLINELIBRARY.COM/JOURNAL/JPOLA

ARTICLE
SCHEME 2 Postulated chemical
structures involved in the cross-
linking process.
result allows us to propose that for the DPA-Bz and MDP-Bz
thermal curing, three processes are involved: benzoxazine
ring opening to form Mannich bridges, further crosslinking
and esterification or transesterification processes in DPA-Bz
or MDP-Bz respectively. Four chemical species representa-
tives of the structures involved in these reactions are postu-
lated: A, B, C, and D (Scheme 2).
where D represent the experimental spectra data matrix (M and
N in our case), C contain the concentration profiles of the chem-
ical species A, B, C, and D, S their pure spectra and E accounts
for the experimental error. This resolution was carried out by
multivariate curve resolution-alternating least squares method
(MCR-ALS).16 This method analyzes the data based on very few
fundamental restrictions, implicit in Beer-Lambert law.
Spectral and concentration profiles of their evolution along of
the reaction can be obtained when the experimental spectra
data matrix were fitted to the model D ¼ C.ST þ E equation,
The variance associated with the solutions, found by MCR-LS
for the DPA-Bz and MDP-Bz homopolymerizations was
99.99% and 99.98%, respectively.
FIGURE 6 Spectra Profiles recovered by MCR-LS for DPA-Bz and MDP-Bz homopolymerizations. [Color figure can be viewed in the
online issue, which is available at wileyonlinelibrary.com.]
´ ˜
POLYBENZOXAZINES FROM RENEWABLE DPA, ZUNIGA ET AL.
1225

JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY DOI 10.1002/POLA
FIGURE 7 Concentration profiles recovered by MCR-LS for DPA-Bz and MDP-Bz homopolymerizations. [Color figure can be viewed
in the online issue, which is available at wileyonlinelibrary.com.]
Figure 6 shows the normalized spectra for each of the above
postulated four chemical structures. In both cases, the spec-
trum (1) contains characteristic absorption bands of the
initial benzoxazines (Fig. 4). Their similarity with the experi-
mental spectra was evaluated through the correlation coeffi-
cient, founding values of 0.9990 and 0.9999, respectively.
The spectrum (2), is in agreement with the disappearance
of characteristic bands at 1230 cmꢁ1, 1030 cmꢁ1, and
940 cmꢁ1, of benzoxazine ring, thus, these spectra can be
considered representative of Mannich bridge structures from
ring opening. In the spectra 2, 3 and 4 in DPA-Bz and spec-
tra 3 and 4 in MDP-Bz is possible to observe double absorp-
tion peaks around 1600 cmꢁ1, characteristic vibrations of
C¼¼C aromatic ring, which could be associated to a new
crosslinked species (C and D in scheme 2). Moreover, differ-
ences in the carbonylic absorptions around 1700 cmꢁ1, can
be observed in spectra 3 and 4, associated to the esterifica-
tion processes (D in scheme 2). After this discussion it is
worth to point out that spectra 2, 3, and 4 can not be
considered as pure spectra of each species, but they are a
good estimation of them.
Information about the kinetic of these reactions can be
obtained by analizing the evolution of the concentration of
each chemical species, A, B, C, and D along the homopoly-
merizations (Fig. 7). Curves 1 that undergo a fast decrease
from the beginning are associated with the starting
FIGURE 8 TGA plots in nitrogen
and air of polybenzoxazines from
DPA-Bz, MDP-Bz and BPA-Bz.
1226
WILEYONLINELIBRARY.COM/JOURNAL/JPOLA

ARTICLE
benzoxazine monomers (spectra 1 in Fig. 6 and A in Sch-
eme 2), which rapidly suffer the ring opening. Curves 2,
associated to Mannich bridge structures (spectra 2 in Fig. 6
and B in scheme 2), increase at the same rate as curves 1
decrease at the beginning of reaction and decrease progres-
sively as the curing proceed and new crosslinked sites are
formed including the reactive positions in the aniline ring
(C in scheme 2). This new chemical situations should be
depicted by curves 3. Finally, curves 4 would be associated
to the esterification process (D in scheme 2) that in the DPA-
Bz reaction is observed from the beginning of the reaction
while in the MDP-Bz reaction needs an induction time of
about 50 min before transesterification started. This fact
could be related with the more acidic medium in the first
reaction that facilitates the esterification. To unravel the
transesterification process a model experiment in which
methylestearate and 2,4-dimethylphenol were heated at
The authors express their thanks to CICYT (Comisio´n Intermi-
nisterial de Ciencia y Tecnologı´a) (MAT2008-01412) for finan-
cial support for this work.
REFERENCES AND NOTES
1 Feedstocks for the Future: Renewable for the Production of Chem-
icals and Materials; ACS Symposium Series 921; Bozell, J. J.; Patel,
M., Eds.; American Chemical Society: Washington, DC, 2006
2 (a) Belgacem M. N.; Gandini, A. Monomers, Polymers and
Composites from Renewable Resources; Elsevier: Oxford 2008;
(b) Gandini A. Macromolecules 2008, 41, 9491–9504.
3 Ghosh, N. N.; Kiskan, B.; Yagci, Y. Prog Polym Sci 2007, 32,
1344–1391.
4 Endo, T.; Sudo, A. J Polym Sci Part A: Polym Chem 2009, 47,
4847–4858.
5 Ning, X.; Ishida, H. J Polym Sci Part A: Polym Chem 1994,
ꢀ
200 C, in presence of dimethylbenzylamine, was carried out.
32, 1121–1129.
The reaction was followed by IR showing the disappearance
of the absorption at 1739 cmꢁ1 due to the carbonyl group of
methylestearate and the appearance of a band at 1725 cmꢁ1
of the corresponding carbonyl group of phenolic ester.
6 Low, H.Y.; Ishida, H. Macromolecules 1997, 30, 1099–1106.
7 Yagci, Y.; Kiskan, B.; Ghosh, N. N. J Polym Sci Part A: Polym
Chem 2009, 47, 5565–5576.
8 Ishida, H.; Rodr´ıguez, Y. Polymer 1995, 36, 3151–3158.
To determine the thermal stability and the decomposition
behavior, thermogravimetric analysis were carried out under
nitrogen and air atmospheres. Figure 8 shows TGA curves
and the first derivative curves and Table 2 summarizes TGA
data. Under air or nitrogen atmosphere, samples were
thermally stable below 250 ꢀC, and the degradation takes
place in one step. However a higher temperature of 10%
weight loss is observed for MDP-Bz derivative under nitro-
gen indicating a slightly higher resistance than DPA-Bz and
BPA-Bz derivatives to degrade. No significant differences can
be seen in temperature of the maximum weight loss rate
and in char yields at 800 ꢀC indicating that all systems
exhibit quite a similar thermal behavior.
9 Zhang, R. F.; Moore, J. A. Macromol Symp 2003, 199, 375–390.
10 Ping, Z.; Linbo, W.; Bo-Geng, L. Polym Degrad Stab 2009, 94,
1261–1266.
11 Moore, J. A.; Tannahill, T. High Perform Polym 2001, 13,
S305–S316.
12 Chu, F. K.; Hawker, C. J.; Pomery, P. J.; Hill, D. J. T. J Polym
Sci Part A: Polym Chem 1997, 35, 1627–1633.
13 Bozell, J. J.; Moens, L.; Elliot D. C.; Wang, Y.; Neuenscwan-
der, G. G.; Fitzpatrick, S. W.; Bilski, R. J.; Jarnefeld, J. Resour-
ces, Conserv Recycling 2000, 28, 227–239.
14 Andreu, R.; Reina, J. A.; Ronda, J. C. J Polym Sci Part A:
Polym Chem 2008, 46, 6091–6101.
CONCLUSIONS
15 Vandeginste, M. G.; Massart, D. L.; Buydens, L. M. C.; de
Jong, S.; Lewi, P. J.; Smeyers Verbeke J. Handbook of Chemo-
metrics and Qualimetrics: Part B; Elsevier: Amsterdam,1998.
We have synthesized two novel benzoxazine monomers,
DPA-Bz and MDP-Bz from renewable DPA, with similar struc-
ture of BPA, by traditional approaches. The thermally
activated polymerization of the monomers afforded thermo-
setting polybenzoxazines, being the low polymerization tem-
perature of DPA-Bz attributed to the presence of carboxylic
group. Polybenzoxazines from MDP-Bz and DPA-Bz showed
higher Tg’s, 270 C and 208 C respectively, and higher cross-
linking density compared to BPA-Bz, due to the transesterifi-
cation or esterification of the acidic groups and the phenolic
OHs, formed during ring opening of oxazines. These reac-
tions are in accordance with the number of independent
reactions determined analyzing by SVD the chemical rank of
the IR spectra data matrices recorded along the homo-
polymerization reactions monitored at 200 ꢀC. Spectral and
concentration profiles of the active chemical species involved
in these processes were obtained by MCR-ALS. DPA-Bz
esterification is observed from the beginning of the reaction
while the MDP-Bz transesterification needs an induction
time of about 50 min under the studied conditions.
16 De Juan, A.; Tauler, R. Anal Chim Acta 2003, 500, 195–210
17 Brunovska, Z.; Liu, J. P.; Ishida, H. Macromol Chem Phys
1999, 200, 1745–1752.
18 Espinosa, M. A.; Ca´diz, V.; Galia`, M. J Appl Polym Sci 2003,
90, 470–481.
ꢀ
ꢀ
19 The Mathworks, MATLAB Version 7.0, Natick, MA, 2004.
20 Dunkers, J.; Ishida, H. Spectrochim Acta Part A 1995, 51, 855–867.
21 Tobolsky, A. V.; Carlson, D.W.; Indictor, N. J. J Polym Sci
1961, 54, 175–192.
22 Takeichi, T.; Guo, Y.; Rimdusit, S. Polymer 2005, 46, 4909–4916.
23 Lin, C. H.; Lin, H. T.; Chang, S. L.; Hwang, H. J.; Hu, Y. M.;
Taso, Y. R.; Su, W. C. Polymer 2009, 50, 2264–2272
24 Mijovic, J.; Fishbain, A.; Wijaya, J. Macromolecules 1992,
25, 979–985.
25 Amrhein, M.; Srinivasan, B.; Bonvin, D.; Schumacher, M. M.
Chem Intell Lab Syst 1996, 33, 17–33.
´ ˜
POLYBENZOXAZINES FROM RENEWABLE DPA, ZUNIGA ET AL.
1227