Innovative approach using aminomaleimide for unlocking phenolic diversity in high-performance maleimidobenzoxazine resins

Article abstract of DOI:10.1016/j.polymer.2017.05.061

A new innovative synthetic approach for the preparation of maleimidobenzoxazines was developed. For the first time the phenolic diversity in maleimidobenzoxazine compounds was enabled, opening many possibilities for tailoring properties of maleimidobenzoxazines through phenolic diversity. The key in a successful implementation of a novel synthetic approach was utilization of aminomaleimide compound prepared by modified synthetic path, without the need for chromatographic purification. Yields of the reaction between aminomaleimide and phenols reached up to 95%, which is a significant improvement prior to previous studies dealing with maleimidobenzoxazines. Aminomaleimide and all novel maleimidobenzoxazine substances were characterized by ¹H NMR, ¹³C NMR and FTIR spectroscopies. The curing behaviour of prepared maleimidobenzoxazines was investigated by DSC analysis and mechanical properties by DMA analysis. Novel maleimidobenzoxazines showed similar properties comparing to the already presented in the literature showing high glass transition temperature values over 240 °C. Thermogravimetric analysis proved improved thermal stability of maleimidobenzoxazines compared to anilinobenzoxazines. Finally, the use of prepared maleimidobenzoxazines was investigated by the means of Diels-Alder reaction as a possible benzoxazine bearing maleimide group precursor for self-healing purposes.

Full text of DOI:10.1016/j.polymer.2017.05.061

Accepted Manuscript
Innovative approach using aminomaleimide for unlocking phenolic diversity in high-
performance maleimidobenzoxazine resins
Žiga Štirn, Aleš Ručigaj, Matjaž Krajnc
PII:
S0032-3861(17)30538-4
10.1016/j.polymer.2017.05.061
JPOL 19721
DOI:
Reference:
To appear in: Polymer
Received Date: 8 March 2017
Revised Date: 24 May 2017
Accepted Date: 26 May 2017
Please cite this article as: Štirn Ž, Ručigaj Aleš, Krajnc Matjaž, Innovative approach using
aminomaleimide for unlocking phenolic diversity in high-performance maleimidobenzoxazine resins,
Polymer (2017), doi: 10.1016/j.polymer.2017.05.061.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT
1
2
Innovative approach using aminomaleimide for unlocking phenolic diversity in high-performance
maleimidobenzoxazine resins
3
Žiga Štirn, Aleš Ručigaj and Matjaž Krajnc*
4
University of Ljubljana, Faculty of Chemistry and Chemical Technology, Večna pot 113, 1000 Ljubljana,
5
Slovenia
6
Author information
7
*Corresponding Author
e-mail: matjaz.krajnc@fkkt.uni-lj.si
phone: +386 1 479 8600
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Abstract
A new innovative synthetic approach for the preparation of maleimidobenzoxazines was developed. For
the first time the phenolic diversity in maleimidobenzoxazine compounds was enabled, opening many
possibilities for tailoring properties of maleimidobenzoxazines through phenolic diversity. The key in a
successful implementation of a novel synthetic approach was utilization of aminomaleimide compound
prepared by modified synthetic path, without the need for chromatographic purification. Yields of the
reaction between aminomaleimide and phenols reached up to 95%, which is a significant improvement
prior to previous studies dealing with maleimidobenzoxazines. Aminomaleimide and all novel
maleimidobenzoxazine substances were characterized by H NMR, 13C NMR and FTIR spectroscopies.
1
The curing behaviour of prepared maleimidobenzoxazines was investigated by DSC analysis and
mechanical properties by DMA analysis. Novel maleimidobenzoxazines showed similar properties
comparing to the already presented in the literature showing high glass transition temperature values
over 240 °C. Thermogravimetric analysis proved improved thermal stability of maleimidobenzoxazines
compared to anilinobenzoxazines. Finally, the use of prepared maleimidobenzoxazines was investigated
1

ACCEPTED MANUSCRIPT
1
2
by the means of Diels-Alder reaction as a possible benzoxazine bearing maleimide group precursor for
self-healing purposes.
3
Keywords benzoxazine; maleimide; phenolic diversity
4
5
1. Introduction
6
High performance materials continue to receive considerable attention as an addition to a significant
amount of studies in the last several decades [1–3]. Among the advanced materials, polybenzoxazines
exhibit many attractive properties, including high glass transition temperatures, low moisture
absorption, good thermal stability, mechanical and electrical properties, excellent properties in chemical
resistance and flame retardancy. Besides, benzoxazine monomers could be obtained from various low-
cost raw materials leading towards the rich molecular design flexibility that opens exceptional possibility
in further improvements of material properties. Polybenzoxazine properties were additionally
effectively improved by copolymerization and hybrid systems, where enhancement of glass transition
temperature, thermal stability and processability was achieved [4–7]. One of the effective approaches in
achieving better polymer properties is using maleimide compounds in benzoxazine monomer
modifications.
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Maleimidobenzoxazines are recognized as high performance materials. Benzoxazines with incorporated
maleimide functionality are commonly formed from the reaction of only one phenolic derivative known
as 4-hydroxyphenylmaleimide (HPMI) accompanied with formaldehyde and various amines as reactants.
The mentioned is reflected in the lack of chemical versatility as in the synthesis of benzoxazine
monomer only the amine part contributes to the diversity of the molecule. According to the literature
[8–17] there is a huge lack of phenolic diversity for current synthesized maleimidobenzoxazine resins.
Nevertheless, it was reported that the combination of benzoxazine and maleimide generates the
material with the quality possessions of both of them [10,13]. Maleimidobenzoxazines were prepared
2

ACCEPTED MANUSCRIPT
1
2
from low-viscosity monomers and have shown excellent thermal stability, increased char yield and glass-
transition temperature [9,10]. On the other hand, the synthesis of maleimidobenzoxazines usually
showed lower yields in comparison to the typical benzoxazines (with no maleimide functional group) in
the range from 30–80% [8,9,12,15,17]. The mentioned could be assigned to the lower reactivity of the
HMPI comparing to the typical phenols used in benzoxazine synthesis, and the competitive side
reactions [12].
3
4
5
6
7
Furthermore, Jin et al. [17] developed a new synthetic route in order to produce bifunctional
maleimidobenzoxazines since the conventional methods were inappropriate. The main reason was poor
solubility of 4-hydroxyphenylmaleimide (HPMI) in common organic solvents (chloroform or dioxane)
used in typical benzoxazine synthesis. Among that, the reaction of diamines with formaldehyde formed
a stable triaza gel structure making difficult to react with phenols in the next step of the reaction. Jin et
al. solved the problem using high boiling point nonpolar solvents achieving simultaneous dissolution of
triaza gel and increased solubility of HPMI. Nevertheless, yields remained lower comparing to typical
benzoxazines reaching up to 65%. Over the years the synthesis of maleimidobenzoxazines were entirely
based only on HPMI as maleimide precursor, which is why we were interested in introducing
aminomaleimides as a starting molecule for maleimidobenzoxazine synthesis.
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
The aim of our work was oriented towards the synthesis of versatile maleimidobenzoxazines by
incorporating the maleimide molecule directly on the amine reactive compound. A new synthetic
approach towards greater phenolic diversity of resins by preparing aminomaleimide which could be
easily dissolved in chloroform was developed (Scheme 1). Scheme 1A represents currently used
synthetic approach for preparation of maleimidobenzoxazines with only one place of possible diversity
(red marked R1 on the amine part). On the other hand, newly developed synthetic approach proposed in
this study (Scheme 1B) shows the utilization of aminomaleimides with two possible positions in the
molecule for further molecule diversity of final products. The diversity can be applied by choosing the
3

ACCEPTED MANUSCRIPT
1
2
phenolic derivative from the vast library of phenols that are applicable in benzoxazine synthesis
(phenolic derivative needs to contain one of the ortho position to the hydroxyl group unsubstituted),
which was thoroughly investigated in this study (red marked R2). The second possible diversity is the
consequence of aminomaleimide part marked with X in Scheme 1, although it was keeping constant in
this research study. Further synthesis of maleimidobenzoxazine monomer was performed under normal
conditions that are already extensively used and described in the literature [18]. Yields of the reaction
between aminomaleimide and phenols up to 95% were achieved in the preparation of
maleimidobenzoxazines which is a considerable improvement prior to previous researches dealing with
the maleimidobenzoxazine synthesis [8,9,12,15,17]. From the best of our knowledge this work utilizes
the first application of aminomaleimides in preparation of maleimidobenzoxazines. The proposed
procedure opens promising and novel way in the synthesis of diverse maleimidobenzoxazine monomers
leading to versatile high performance thermosets.
3
4
5
6
7
8
9
10
11
12
13
14
Scheme 1. Current approach in the preparation of maleimidobenzoxazines (A) and newly developed
synthetic approach presented in this study (B).
15
16
17
2. Experimental section
4

ACCEPTED MANUSCRIPT
1
2
2. 1. Materials
Maleic anhydride (MA), 4,4’-diaminodiphenylmethane (DDM), triethylamine, di-tert-butyl dicarbonate
(Boc2O), ammonium chloride (NH4Cl), hydrochloric acid (HCl), acetic anhydride (Ac2O), sodium acetate
(NaOAc), trifluoroacetic acid (TFA), NaHCO3, sodium sulfate (Na2SO4), sodium hydroxide (NaOH),
paraformaldehyde, phenol, 4-tert-butylphenol (TBP), bisphenol A (BA), aniline, activated charcoal,
solvents and deuterated solvents were purchased from Aldrich (Chemie GmbH, Steinheim, Germany),
Merck (Darmstadt, Germany) or Euriso-top (Saint-Aubin, France). All of the chemicals were used as
received.
3
4
5
6
7
8
9
Already known benzoxazine bearing furan group 3-(furan-2-ylmethyl)-8-methoxy-3,4-dihydro-2H-
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
benzo[e][1,3]oxazine
(G-f)
and
several
anilinobenzoxazines
3-phenyl-3,4-dihydro-2H-
benzo[e][1,3]oxazine (P-a), 6-(tert-butyl)-3-phenyl-3,4-dihydro-2H-benzo[e][1,3]oxazine (TBP-a), 6,6'-
(propane-2,2-diyl)bis(3-phenyl-3,4-dihydro-2H-benzo[e][1,3]oxazine) (BA-a) were prepared according to
the literature [19].
The following novel maleimidobenzoxazines were prepared from aminomaleimide (AM) as a starting
molecule: 1-(4-(4-(2H-benzo[e][1,3]oxazin-3(4H)-yl)benzyl)phenyl)-1H-pyrrole-2,5-dione (P-m), 1-(4-(4-
(6-(tert-butyl)-2H-benzo[e][1,3]oxazin-3(4H)-yl)benzyl)phenyl)-1H-pyrrole-2,5-dione
((((6,6'-(propane-2,2-diyl)bis(2H-benzo[e][1,3]oxazine-6,3(4H)-diyl))bis(4,1-
(TBP-m),
1,1'-
phenylene))bis(methylene))bis(4,1-phenylene))bis(1H-pyrrole-2,5-dione) (BA-bm) and 1-(4-(4-(6-(2-(4-
hydroxyphenyl)propan-2-yl)-2H-benzo[e][1,3]oxazin-3(4H)-yl)benzyl)phenyl)-1H-pyrrole-2,5-dione (BA-
mm).
Detailed preparation procedure of aminomaleimide and various maleimidobenzoxazines and its
characterization can be found in Supplementary Material.
2. 2. Polymerization of prepared anilinobenzoxazines and maleimidobenzoxazines
5

ACCEPTED MANUSCRIPT
1
2
Maleimidobenzoxazine samples P-m and TBP-m were melted at elevated temperature ranging between
120 and 150 °C. However, since BA-mm and BA-bm melted at much higher temperatures, the addition
of solvent (DMF) was required for smooth processing of a sample. Afterwards, the liquid samples were
poured into a silicon mould (10 × 80 × 10 mm in dimensions) and vacuumed at elevated temperatures
for approximately 30 min. After the completion of the process the samples were stepwise heated for
one hour at 180, 190, 200, 210, 220, 230, 240, and 250 °C. Anilinobenzoxazines samples P-a, BA-a and
TBP-a were cured in a similar manner as previous, however they were heated at different temperatures,
namely for one hour at 180, 190 and 200 °C. Unfortunately, due to the extreme brittleness of the TBP-a
sample after the curing process, we were unable to produce mechanically stable sample for further
mechanical processing to perform dynamical mechanical analysis. In order to provide representative
sample, the mixture of TBP-a and P-a in a mass ratio of 0.80 and 0.20, respectively, was prepared
instead.
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
2. 3. Evaluation of Diels–Alder (DA) and retro Diels–Alder (rDA) reaction
For DA studies maleimidobenzoxazine samples (P-m, TBP-m, BA-bm and BA-mm : BA-bm) were mixed
with G-f in an equimolar amount and homogenized in a mortar. DSC experiments were performed by
heating the prepared mixtures from 25 to 350 °C by a heating rate of 10 K min–1. Additionally, rDA
reaction was investigated by heating mixtures first at 90 °C for 1 h, followed by DSC analysis.
2. 4. Characterization
1H NMR analyses were performed on a Bruker Avance III 500 MHz NMR spectrometer at the
temperature 296 K and frequency 500 MHz and on Bruker DPX 300 MHz NMR at the temperature 296 K
and frequency 300 MHz. 13C NMR analyses were performed on a Bruker Avance III 500 MHz NMR
6

ACCEPTED MANUSCRIPT
1
2
spectrometer at the temperature 296 K and frequency 125.721 MHz. Deuterated chloroform (CDCl3) and
deuterated dimethyl sulfoxide ((CD3)2SO) both containing trimethylsilane (TMS) as internal standard
were used as solvents.
3
4
Thermal properties and curing behaviour of the samples were studied by thermogravimetric analysis
(TGA) and differential scanning calorimetry (DSC), respectively. The DSC measurements were performed
on Mettler Toledo DSC1 instrument with intra-cooler using STAR software. In and Zn standards were
used for the temperature calibration and for the determination of the instrument time constant.
Samples of around 10 mg were weighed in standard 40 µL alumina pans. A heating rate of 10 K min–1
was used and the purge nitrogen gas flow was maintained to be constant at 30 mL min–1 for all
experiments. TGA measurements of samples were performed in an argon 5.0 atmosphere with a flow
rate of 60 ml min-1 using a Netzsch STA 449 F3 Jupiter apparatus in the temperature range from 35 up to
900 °C with the heating rate of 10 K min–1. For all measurement standard 70 µL Al2O3 crucibles with
perforated lids were used and the baseline was automatically subtracted. The sample masses were
around 15 mg.
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Thermomechanical properties of cured polymers were examined by dynamical mechanical analysis
(DMA). The DMA measurements were performed on Mettler Toledo DMA 861e instrument.
Measurements were taken on cured rectangular specimens in dual cantilever bending mode at a
frequency 1 Hz, heating rate 3 K min–1 and at constant displacement 10 μm. The temperature range
between 25 and 300 °C was used.
Room temperature XRD spectra of the P-m compounds were recorded by using a wide angle powder X-
ray diffractometer (Siemens D5000 diffractometer) with Cu Kα (λ = 0.15405 nm) radiation. The XRD data
were collected in the 2θ range from 5 to 70° 2θ in steps of 0.034° 2θ with a measurement time 100 s per
step.
IR spectra were obtained with Bruker Alpha Platinum ATR spectrometer on a solid support.
7

ACCEPTED MANUSCRIPT
1
2
3. Results and discussion
3
3. 1. Preparation of aminomaleimide
4
A key step in the preparation of maleimidobenzoxazines is utilization of aminomaleimide (AM)
derivatives. The conventional methods for the preparation of N-substituted maleimides [20–24] are
5
6
generally
not
applicable
in
the
preparation
of
aminomaleimides.
Moreover,
7
maleimides/aminomaleimides produced by conventional methods were often mischaracterized. Due to
the lack of appropriate analytical methods authors frequently misinterpreted aminoizomaleimides for
aminomaleimides [25]. Interestingly, in the recently published literature [26–28], authors are suggesting
preparation of 1-(2-aminoethyl)-1H-pyrrole-2,5-dione (representative of aminomaleimides) via
environmentally friendly synthetic route without prior need for protection of amine group. The detailed
investigation showed that synthesized product was mischaracterized, similar as in previous studies due
to the use of inappropriate characterization technique. More details can be found in a Supplementary
Material. In fact, the preparation of aminomaleimides is successful when the protection of the amino
group is applied [25,29–32]. However, when applying the protection, once simple synthesis leads to a
several steps synthesis procedure, which includes chromatography separation suitable only for a small
scale processes. To overcome mentioned difficulties a modified method was developed, which still
needs the protection of the amino group but avoids the need of chromatographic separation and can be
consequently used also in the multigram batches (Scheme 2). Detailed synthetic procedure is available
in the Supplementary Material.
8
9
10
11
12
13
14
15
16
17
18
19
20
8

ACCEPTED MANUSCRIPT
1
2
Scheme 2. Synthetic procedure of 1-(4-(4-aminobenzyl)phenyl)-1H-pyrrole-2,5-dione preparation
3
4
(aminomaleimide AM).
5
3. 2. Preparation of novel maleimidobenzoxazines
6
The novel maleimidobenzoxazine monomers were prepared by the reaction of AM and
paraformaldehyde with the set of different phenols according to the Scheme 3. Presented approach
holds several advantages over the conventional one, which is based on the reaction between HPMI,
paraformaldehyde and amine/aniline derivative. Due to the unfavourable chemical composition, HPMI
has a high melting point and poor solubility in a common organic solvents used in the ordinary
benzoxazine synthesis [17]. This phenomenon results in a higher degree of impurities and also
contributes toward inferior reaction yields in a solvent synthesis. Nevertheless, the major advantage of
our approach is the variation of the phenolic derivatives. In such a way the diversity of benzoxazine is
greatly enhanced compared to the method where HPMI is used. Even more, according to our knowledge
this is the first synthetic approach in the introduction of maleimide functionality through amine
molecule and not over phenolic derivative (HPMI) as it was formerly done. Another benefit of
aminomaleimide is, in contrast to HMPI, its solubility in low polar solvents. It is known that in the
solution process of benzoxazine synthesis, a nonpolar solvent is preferable, since polar solvents allow
7
8
9
10
11
12
13
14
15
16
17
18
9

ACCEPTED MANUSCRIPT
1
2
3
4
side reactions leading to the poor yields of benzoxazine monomers [33]. The successful appliance of AM
in the solution process were synthesized maleimidobenzoxazines presented in Scheme 3 using
chloroform as solvent. All novel benzoxazine monomers were obtained in high yield and high purity.
Detailed synthetic procedure is available in the Supplementary Material.
5
6
Scheme 3. Synthetic utilization of a new approach in the preparation of maleimidobenzoxazines.
7
8
The structure of novel maleimidobenzoxazines was confirmed by 1H NMR (Fig. 1), 13C NMR (Fig. 2) and IR
spectroscopy (Fig. 3). Characteristic signals for the oxazine ring appeared in the range of 5.33–5.27 (a)
and 4.60–4.52 ppm (b) corresponding to –O–CH2–N– and –Ar–CH2–N–, respectively. The singlet signal
located in the range of 6.74–6.81 ppm (c) corresponds to –CH=CH– in the maleimide functional group.
Since DDM was used as starting diamine for preparation of AM classical singlet shape resonance of –Ar–
CH2–Ar– protons can be observed in the range of 3.91–3.89 ppm (d). Most of the remaining signals are
the consequence of the diversity of phenolic structure, where resonance of (CH3)3– was observed at 1.27
ppm (e) for TBP-m. The resonance (CH3)2– for BA-bm and BA-mm was observed at 1.56 and 1.58 ppm (f),
9
10
11
12
13
14
15
10

ACCEPTED MANUSCRIPT
1
2
respectively. Signals from aromatic protons could be located in the range of 7.23–6.67 (g) for all of the
prepared compounds (Fig. 1).
3
4
Fig. 1. 1H NMR spectra of prepared maleimidobenzoxazines.
5
6
According to the 13C NMR analysis (Fig. 2) characteristic peaks for oxazine ring in the range of 79.7–79.3
(a) and 50.5–50.9 ppm (b) for –O–CH2–N– and –Ar–CH2–N–, respectively. Resonance observed in the
range of 134.1–134.3 ppm (c) was assigned to –CH=CH– carbon in the maleimide structure, also
resonance in the range of 169.6-169.8 ppm (h) was recognized as C=O carbon in maleimide functional
group. Also in all of the derivatives there was observed resonance for carbon –Ar–CH2–Ar– in the range
7
8
9
10
11
12
13
of 40.7–40.8 ppm (d). Other typical resonances for C NMR were observed at 31.5 ppm (e) for (CH3)3–
as a part of TBP-m, and 31 ppm (f) for (CH3)2– as a part of BA-bm and BA-mm (Fig. 2).
11

ACCEPTED MANUSCRIPT
1
2
Fig. 2. 13C NMR spectra of prepared maleimidobenzoxazines.
3
4
The IR spectra of all synthesized maleimidobenzoxazines (Fig. 3) include typical C=O absorbance at about
1709–1708 cm–1 (I) due to the carbonyl group presented in maleimide functionality. Signal at around
1148 cm–1 (II) is assigned to vibration of C–N–C of oxazine and imide functionalities. Absorbance around
824 cm–1 (III) is the consequence of C–H wagging of vinylene group, while absorbance at around 690 cm–
5
6
7
1
8
(IV) is characteristic for C–H twisting of vinylene group. Additionally, for benzoxazine functionality
9
antisymmetric stretching of C–O–C was observed around 1230 cm–1 (V). Although IR spectra of
compounds are very similar, one can observe additional absorbance around 759 cm-1 for the P-m
sample. The absorbance is characteristic for out of phase bending of C–H in disubstituted aromatic ring
which is present in P-m sample.
10
11
12
12

ACCEPTED MANUSCRIPT
1
2
Fig. 3. IR spectra of prepared maleimidobenzoxazines.
3
4
3. 3. DSC analysis of novel maleimidobenzoxazines
5
Thermal properties and curing of prepared maleimidobenzoxazines were assessed by DSC analysis (Fig.
4). Endothermic peak at 164 °C was observed in the case of P-m sample, whereas in the case of other
maleimidobenzoxazines no endothermic peak was detected indicating the melting of the substance.
Mentioned results are presumably the consequence of impurities and amorphicity of the material,
which will be furtherly discussed on P-m case study. However, exothermic thermal behaviour of all
presented maleimidobenzoxazines were similar. It was observed that exothermic contributions
contained at least two overlapped peaks representing the complex mechanism of a curing process.
According to the literature [11,12] the curing of maleimidobenzoxazines has at least two curing stages.
The polymerization at lower temperatures could be associated to the autocatalytic nature of
benzoxazine resin, when phenol groups are generated while the benzoxazine ring starts to open [12].
The radical polymerization of maleimide C=C bonds could explain the second exothermic peak at higher
temperature [11] (Table 1). Results of exothermic peaks temperatures presented in Table 1 are in the
range of previously synthesized maleimidobenzoxazines [8,9,12,17]. Additional exothermic contribution
was observed in the case of BA-bm with exothermic peak at 307 °C, which can be associated to the
6
7
8
9
10
11
12
13
14
15
16
17
18
13

ACCEPTED MANUSCRIPT
1
2
3
4
5
6
7
minor degree of material degradation [17]. The determined enthalpy for BA-bm (144 kJ/mol), which
contained two maleimide and two benzoxazine groups per molecule, was in accordance to what Jin et
al. [17] had observed for their bismaleimide type benzoxazines (144–195 kJ/mol). It can be observed
that P-m had the highest total exothermic peak enthalpy, which could indicate that small content (small
1
since H and 13C were not showing any significant amount of impurity) of impurity was greatly affecting
the DSC behaviour. The effect of impurities was further studied by the preparation of three different
sample grades of P-m.
8
9
Fig. 4. DSC thermograms of maleimidobenzoxazines.
10
11
Table 1: Maximum peak temperatures and enthalpies of prepared maleimidobenzoxazines.
Sample 1st exo [°C] 2nd exo [°C] 3rd exo [°C] Total
∆
Hexo [J/g] Total
∆Hexo [kJ/mol]
237
228
201
212
248
/
339
183
130
177
145
P-m
252
/
/
218
274
212
TBP-m
BA-mm
BA-bm
/
/
307
12
13
3. 4. Impurities effects on the curing behaviour of maleimidobenzoxazines
14

ACCEPTED MANUSCRIPT
1
2
The P-m was chosen as a case study for the investigation of the impurity effects on the curing behaviour
of maleimidobenzoxazines. Three different P-m samples were prepared from raw P-m benzoxazine each
having different purity level. P-m_CC was obtained by purification of raw P-m with column
chromatography, P-m_AP was obtained when raw P-m was triturated in acetone and filtered, and lastly
P-m_AR was obtained by evaporation of acetone from filtrate solution of P-m_AP. Due to the different
3
4
5
1
6
purification treatment, H NMR analysis showed different purity results. The purity of P-m_CC reached
7
over 99%, P-m_AP 96% and P-m_AR 70%. As expected, further DSC experiments showed that the
presence of impurities in the samples effects their curing behaviour. Peak shift can be observed in the
case of benzoxazine polymerization to the lower temperatures by impurity increase (Fig. 5), which is in
accordance with the literature. The greater the content of impurities are presented in the sample the
greater the temperature shift in benzoxazine polymerization should be expected [18,34,35]. It has to be
noted that impurities behave as a catalyst in this polymerization process [35,36]. The observed peak
temperatures for P-m_CC, P-m_AP, and P-m_AR were 237, 228 and 210 °C, respectively (Table 1).
Interestingly, varying content of impurities had no effect on temperature shift of second exothermic
signal present at approximately 248 °C. It can be concluded that maleimide polymerization was not
affected by this type of impurities, probably due to the chemical structure of impurities. The mentioned
results were in accordance with previously reported literature on impurity investigation of N-(3-phenyl-
3,4-dihydro-2H-benzo[e][1,3]oxazin-8-yl)benzamide, where ring-opening polymerization at 179 °C for
non-purified (crude) product was observed. Furthermore, partially purified product by TLC had same
exothermic signal as crude, whereas by column chromatography purified sample had higher exothermic
signal at 210 °C. Additional washing of powder furnished needle crystals that had exothermic signal even
higher at 215 °C [35]. Nevertheless, on the basis of the results we obtained, the impurities could be of
ionic type or pro-ionic type. During the heating the impurities were transformed to ionic type species,
which would explain the decrease in polymerization temperature of benzoxazine and not maleimide
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
15

ACCEPTED MANUSCRIPT
1
2
moieties, since benzoxazines follow the ionic type polymerization [37–39], while maleimides typically
follow radical type polymerization [15,17,40,41].
3
4
Fig. 5. DSC experiments of P-m at a heating rate of 10 K/min.
5
6
The second important observable change was a temperature shift of endothermic peak as a result of the
melting process. It can be observed that melting point depression represented by endothermic peak was
presented only in the high purity samples of P-m_CC and P-m_AP, whereas in the case of the P-m_AR
there was no indication of the melting process. Moreover, comparing the P-m_CC and P-m_AP samples,
the endothermic peak broadened and shifted to the lower temperature (from 164 to 159 °C) as a result
of higher impurity of P-m_AP over P-m_CC. The endothermic peak was not observed in the case of P-
m_AR, due to the amorphous structure of a substance confirmed by powder XRD analysis (Fig. 6).
Identical XRD spectra of P-m_CC and P-m_AP can be observed showing crystalline structure of the
substance, whereas broad peak of P-m_AR sample indicate its amorphous structure. The mentioned is in
accordance with the DSC results, where no endothermic behaviour of P-m_AR was detected. Similar
results were observed also by Ohashi et al. [35], where the melting enthalpy for crude sample was not
7
8
9
10
11
12
13
14
15
16
16

ACCEPTED MANUSCRIPT
1
2
observed, while samples with higher purity showed distinct endothermic signal indicating the melting
process of the sample.
3
4
Fig. 6. Powder XRD analysis of P-m compounds.
5
6
As expected there were also enthalpy differences observed by comparing each sample. The higher the
1
7
impurity content the lower endothermic and exothermic contributions were observed. According to H
8
NMR spectroscopy results, which indicated minimal difference in the impurity content for P-m_CC and
P-m_AP, the same explanation could be recognized by DSC thermograms. Even though there were some
visible changes between P-m_CC and P-m_AP samples on DSC thermograms, the small difference in
endothermic and exothermic enthalpy value indicated similar purity of both samples. In contrast, high
difference of P-m_AR purity regarding to P-m_CC and P-m_AP samples reflected in significantly lower
enthalpy of polymerization.
9
10
11
12
13
14
Table 2: Maximum peak temperatures and enthalpies of different P-m samples.
Sample
1st endo [°C]
164
∆
Hendo [J/g] 1st exo [°C] 2nd exo [°C] Total
∆
Hexo [J/g]
339
336
–86
–83
237
228
248
248
P-m_CC
P-m_AP
159
17

ACCEPTED MANUSCRIPT
/
210
248
241
P-m_AR
1
2
3
4
5
6
Additionally, substances were investigated by IR analysis (Fig. 7). It was discovered that P-m_CC and P-
m_AP were practically identical, since both had crystalline structure. In contrary, IR spectrum of the P-
m_AR sample displayed broader lines and poorer line resolution, which can be attributed to the
amorphous structure of the substance. Nevertheless, signals were otherwise similar to that of P-m_CC
and P-m_AP. Results showed that IR analysis is not conclusive method for purity evaluation (Fig. 7).
7
8
Fig. 7: IR spectra of different P-m substances.
9
10
11
12
13
14
15
16
For the purpose of this paper the impurity investigation was an important examination of the
polymerization behaviour. In the case of benzoxazine polymerization controlled amounts of impurities
can be usually advantageous since impurities decrease melting temperature of the substance and lower
the onset of the polymerization process. The mentioned is essential for further processing of a given
material into other applications, since pure maleimidobenzoxazines have high polymerization
temperature. Impurities present in benzoxazine monomers are usually residual phenols and amine
complexes that are subsequently also incorporated into the polymer matrix. Alternative for benzoxazine
18

ACCEPTED MANUSCRIPT
1
2
processing is solvent addition to the benzoxazine monomers, which at higher temperatures dissolves
monomers and after evaporation keeps monomer in a melted state. The downside of this process is
possible residue of solvent in the final material, which could affects mechanical and other properties of
the material.
3
4
5
All further studies were conducted on newly prepared samples: P-m_AP, P-m_AR, TBP-m and BA-mm :
BA-bm, where BA-mm : BA-bm is a sample mixture between BA-mm and BA-bm in a molar ratio of 1:1.5.
The reason why the mixture of BA-mm and BA-bm was used instead of the pure BA-bm was in the
difficulties of DMA sample preparation. Pure BA-bm had high melting point leading to the simultaneous
melting and polymerization process making it very inconvenient for processing a sample into a relevant
DMA specimen. All attempts of processing BA-bm sample, even with the help of a DMF as a solvent,
ended being futile. Either large quantities of solvent were entrapped in the sample, or upon aggressive
vacuuming, the sample was full of cavities. To bypass mentioned difficulties a mixture of BA-mm and BA-
bm was used instead, where due to the lower melting point the formation of the specimen was
manageable. Additionally, BA-mm and BA-bm both have 1:1 molar ratio of maleimide to benzoxazine
groups and are constructed from the same initial phenol and amine compounds. For that reason we
speculate that the mixture of BA-mm and BA-bm was a representative sample to describe both BA-mm
and BA-bm properties.
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
3. 5. DMA analysis of novel maleimidobenzoxazines
The viscoelastic properties of cured samples were evaluated by DMA analysis. Storage and loss modulus
for the cured maleimidobenzoxazines and anilinobenzoxazines are presented in Fig. 8. Storage modulus
for maleimidobenzoxazines (Fig. 8a) were slowly decreasing with increasing temperature up to 190 °C,
where steep decline in storage modulus for the TBP-m sample was observed followed by steep declines
in modulus at 225, 236 and 264 °C for P-m_AR, P-m_AP and BA-mm : BA-bm, respectively. Glass
19

ACCEPTED MANUSCRIPT
1
2
transition temperatures were determined from the tan δ graph (Fig. 9). It was observed that sample BA-
mm : BA-bm had the highest glass transition temperature 274 °C and TBP-m had the lowest, namely 210
°C. As it was expected, P-m_AP had higher Tg than P-m_AR, being 252 and 243 °C, respectively, due to
the higher purity of P-m_AP sample. However, small difference of 11 °C could be explained by the fact
that majority of impurities like bifunctional amine and imine intermediates are actively involved in the
polymerization reactions [37,42]. The lowest glass transition of TBP-m in comparison to other samples
was surprising, since t-butyl group should pose additional anchor point in the polymer matrix in
comparison to hydrogen atom in P-m sample. However, due to the large bulkiness the t-butyl group
presumably obstructed the benzoxazine polymerization, even though it was substituted on para position
to the hydroxyl group in phenol derivative. The mentioned could lead to a short chain polymer matrix
with high brittleness. However, the brittleness of TBP-m was largely reduced compared to the TBP-a
sample due to the maleimide polymerization. When aminomaleimide was substituted with aniline (TBP-
a), the cured material displayed extreme brittleness compared to the TBP-m. For that reason we were
unable to obtain mechanical properties of pure TBP-a, since the sample collapsed before the analysis
was successfully performed. Consequently the sample with less brittle nature was prepared as a mixture
of TBP-a and P-a to evaluate the mechanical properties. As expected the glass transition temperatures
of anilinobenzoxazines had the same trend as maleimidobenzoxazines. The highest being BA-a at 184 °C
followed by P-a at 144 °C and lastly mixture of TBP-a and P-a at around 126 °C.
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
20

ACCEPTED MANUSCRIPT
1
2
3
Fig. 8. Storage and loss modulus for maleimidobenzoxazines (a) and anilinobenzoxazines (b).
4
21

ACCEPTED MANUSCRIPT
1
2
3
Fig. 9. Tan δ for maleimidobenzoxazine (a) and anilinobenzoxazines (b).
4
5
DMA analysis results of newly prepared maleimidobenzoxazines were compared to the similar
anilinobenzoxazines (with the same phenolic structures) (Table 3). Aminomaleimide molecule was
substituted with aniline in the preparation of benzoxazine monomers. Results were in accordance with
the previously published observations for maleimidobenzoxazine systems, where a large increase (80–
100 °C) in Tg compared to anilinobenzoxazines was observed [12,17].
6
7
8
9
10
11
22

ACCEPTED MANUSCRIPT
1
2
3. 6. Thermal stability of novel maleimidobenzoxazines
TGA results showed 5% and 10% weight loss for novel maleimidobenzoxazines in the range of 351–353
°C (Td5) and 372–379 °C (Td10), respectively (Table 3), indicating the improvement against typical
anilinobenzoxazines with Td5 and Td10 in the range of 282–304 °C. Additionally, Td5 and Td10 temperatures
of maleimidobenzoxazines presented in this work were in accordance with degradation temperatures
(330–383 °C and 366–410 °C, respectively) of already published maleimidobenzoxazines. Furthermore,
general difference between anilinobenzoxazines and maleimidobenzoxazines is the possibility of
maleimidobenzoxazines to have additional anchor into the existing polymer matrix by polymerization via
maleimide double bond, which is why char yields of maleimidobenzoxazines are higher compared to
anilinobenzoxazines [8,9,12,16,17].
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Char yields at 800 °C of a newly synthesized maleimidobenzoxazine compounds were between 40 and
47% (Table 3) which is a significant improvement compared to anilinobenzoxazines (24–36%) (Fig. 10),
while char yields of already known maleimidobenzoxazines from the literature [8,9,12,16,17] are usually
higher (45–62%). It can be presumed that the difference in char yields between presented
maleimidobenzoxazines and already reported in the literature is due to the benzoxazine structure. In
our case the majority of the benzoxazine mass is contributed by the amine part of the molecule, while in
the literature the majority of the mass is contributed by the phenolic part. Amine part is usually the
weakest point in the molecule, if monoamine is used in the synthesis of benzoxazine [43], however
when bifunctional phenols and amines are used, the end group is the weakest point of the molecule and
the starting point of the degradation. In this work, prepared maleimidobenzoxazines behave similar to
bifunctional benzoxazines, where early degradation could not be attributed to amine degradation as in
the case of benzoxazines based on monoamines. Interestingly, as it was already stated above, values of
char yields for maleimidobenzoxazines in this paper compared to maleimidobenzoxazines in the
literature was lower. For that reason we presume that AM has different degradation mechanism than
23

ACCEPTED MANUSCRIPT
1
2
HMPI part of the benzoxazine molecule, consequently contributing to lower char yields of
maleimidobenzoxazines based on AM compounds compared to those based on HMPI.
3
4
5
6
Fig. 10. TGA analysis of novel maleimidobenzoxazines (a) and similar anilinobenzoxazines (b).
Table 3: Thermal properties of studied benzoxazines based on DMA and TGA analysis.
7
8
Weight-loss
temperature [°C] at 800 °C [%]
5% 10%
Char yield
Sample name
Tg [°C] Tg [°C]
tan
δ
G’’ max
24

ACCEPTED MANUSCRIPT
274
252
243
210
184
144
1261
/
263
238
223
194
168
123
110
/
353
352
351
351
300
300
282
288
379
379
375
372
336
346
333
337
44
47
47
40
33
36
24
24
BA-mm : BA-bm
P-m_AP
P-m_AR
TBP-m
BA-a
P-a
TBP-a : P-a
TBP-a
1
2
There were only minor differences of Td5 and Td10 temperatures for maleimidobenzoxazine samples
leading to the conclusion that they likely follow the same degradation mechanism. Significant
differences of maleimidobenzoxazine materials were observed with respect to char yields. As expected,
the lowest char yield of 40% was observed for the TBP-m material, followed by 45% for the BA-mm : BA-
bm, and lastly 47% in the case of P-m_AP and P-m_AR. Interestingly, P-m_AP and P-m_AR behaved
exactly the same according to the TGA measurement and had the highest char yield, which additionally
confirms our assumptions from DMA results that impurities in P-m_AR actually polymerize into the bulk
of the material. Additional surprise was that BA-mm : BA-bm had lower char yield compared to P-m_AP
and P-m_AR, although the same trend was also observed in anilinobenzoxazines, where P-a had greater
char yield than BA-a. The biggest improvement was observed in the case of TBP-m, where maleimide
moiety contributed 16% towards char yield, and also greatly improved Td5 and Td10 (for 64 and 42 °C,
respectively) in comparison to TBP-a sample. Otherwise, in the case of anilinobenzoxazine results of P-a
and BA-a were consistent with the literature [9]. Thermal behaviour of TBP-a and TBP-a : P-a samples
3
4
5
6
7
8
9
10
11
12
13
14
25

ACCEPTED MANUSCRIPT
1
2
3
had inferior properties compared to other two anilinobenzoxazines, the biggest difference being
observed in the char yield which was from 12 to 10% lower than in the case of other two. The
mentioned behaviour was already expected from the DMA result.
4
5
3. 7. Diels-Alder reaction with novel maleimidobenzoxazines
6
One of the interesting aspects of maleimide based systems is that they are generally known as good
dienophiles in Diels–Alder (DA) reactions. Therefore maleimide based systems are extensively used in
DA reactions for different applications, some of which were already transferred and used in benzoxazine
chemistry. One of the example is the usage in the decreasing brittleness of benzoxazines by increasing
chain length of benzoxazines before curing [44]. Furthermore, Fe3O4 nanoparticles were successfully
introduced into benzoxazine molecules through maleimide/furan DA chemistry for the purposes of
achieving high performance magnetic polybenzoxazines [45]. DA reaction was used also in the
production of membrane for the direct methanol fuel cells [46]. Another attractive idea in the usage of
maleimidobenzoxazine is in the self-healing applications of polymer networks triggered by Diels-Alder
(DA) reaction. The product of the Diels-Alder reaction (DA adduct) is the mixture of two diastereomers
called endo and exo adducts. The endo compound is kinetically favoured, whereas exo is
thermodynamically favoured and thus more stable. It is known that DA adducts at appropriate
temperature undergo a reverse reaction called retro-DA (rDA) reaction that regenerates starting
materials. DA reaction was conducted between maleimidobenzoxazines and G-f according to the
Scheme 4. Samples for evaluation of DA reactions were prepared by homogenization of equimolar
amount of G-f and desired maleimidobenzoxazines in a mortar. DA reaction between G-f and
maleimidobenzoxazines was investigated by DSC analysis, where exothermic peak presented DA
reaction and endothermic peak rDA reaction.
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
26

ACCEPTED MANUSCRIPT
1
2
Scheme 4. Diels-Alder reaction between G-f and maleimidobenzoxazines.
3
4
As expected DA reaction between the furan and maleimide functional groups appeared in all of the
prepared mixtures (Fig. 11). The DA reaction was triggered by melting of the prepared mixture due to
the mobility of functional groups as it was already observed in our previous studies [47]. The highest
endothermic peak attributed to the melting of the sample can be observed in the case of P-m_AP
sample, which is why also the exothermic DA peak was less evident comparing to the other used
maleimidobenzoxazines. Nevertheless, DA reaction peak was detected in the range of 70 to 100 °C with
the peak maximum between 84–87 °C followed by the endothermic contribution of rDA reaction of the
endo adduct at around 106 °C. At higher temperatures the second DA exothermic peak at around 117 °C
was visible associated to the exo adduct formation. In the last step the endothermic behaviour at
approximately 134 °C corresponding to the rDA reaction into the starting compounds from the exo
adduct could be perceived (Fig. 11). Similar to previous findings there was no clear boundary between
the end of the rDA reaction and start of polymerization of benzoxazine and maleimide.
5
6
7
8
9
10
11
12
13
14
15
16
17
18
Additionally, G-f mixtures with P-m_AP, P-m_AR, TBP-m, BA-bm and BA-mm : BA-bm were heated for 1
h at 90 °C in order to perform DA reaction and study only the rDA effect during the DSC experiment (Fig.
12). For that reason there was no exothermic peak corresponding to the DA reaction, which is why
27

ACCEPTED MANUSCRIPT
1
2
endothermic response was only observable indicating the rDA reaction of previously obtained DA
adduct. However, in the case of BA-bm and BA-mm : BA-bm mixture, exothermic peak at around 117 °C
after rDA reaction was visible. The mentioned could be explained by the fact that the DA reaction
conversion of BA-bm and BA-mm : BA-bm is lower in comparison to the other examples and
consequently also the exo/endo ratio. We presumed that the reason for lower conversion was in the
partially complete melting process that reflected in lower molecular mobility. Consequently, the
reaction was not completely performed in the molten state leading to lower yield of DA reaction. For
that reason the transformation of endo adduct to exo adduct during the DSC experiment can be
observable. The effect of the heating time and temperature of the heating was discussed and reported
in our previous research [47].
3
4
5
6
7
8
9
10
11
12
Fig. 11: Thermal behaviour between G-f and novel maleimidobenzoxazines.
28