Stereoisomeric effects in thermo-remendable polymer networks based on diels-alder crosslink reactions

Article abstract of DOI:10.1002/pola.24134

This study describes the synthesis, the spectroscopic, and the thermal characterization of linear and crosslinked polymers as well as a number of corresponding model compounds, containing Diels-Alder adducts derived from furan and maleimide groups. The thermal reversibility (rDA, DA) of structurally varied model compounds, polymeric and network structures were studied by differential scanning calorimetry, where possible in combination with ¹H NMR spectroscopy. It was established that the endo and exo DA stereoisomers show significantly different thermal responses: the rDA of the endo DA-adducts typically takes place at 20-40 K lower temperatures than that of the corresponding exo DA-adducts in all cases, with the exception of some aromatic maleimides. Although In situ isomerization was observed to a limited extent and only In some cases, this effect is not expected to influence the thermoremendability of DA-crosslinked networks being dependent on two separate stereoisomeric rDA steps.

Full text of DOI:10.1002/pola.24134

Stereoisomeric Effects in Thermo-Remendable Polymer Networks Based
on Diels–Alder Crosslink Reactions
JUDIT CANADELL,1 HARTMUT FISCHER,2 GIJSBERTUS DE WITH,1 ROLF A. T. M. VAN BENTHEM1
1Department of Materials and Interface Chemistry, Faculty of Chemical Engineering and Chemistry,
Eindhoven University of Technology, 5600 MB Eindhoven, The Netherlands
2Department of Materials Performance, TNO Science and Industry, 5612 AP Eindhoven, The Netherlands
Received 17 April 2010; accepted 17 May 2010
DOI: 10.1002/pola.24134
Published online in Wiley InterScience (www.interscience.wiley.com).
ABSTRACT: This study describes the synthesis, the spectro-
scopic, and the thermal characterization of linear and cross-
linked polymers as well as a number of corresponding model
compounds, containing Diels–Alder adducts derived from furan
and maleimide groups. The thermal reversibility (rDA, DA) of
structurally varied model compounds, polymeric and network
structures were studied by differential scanning calorimetry,
where possible in combination with 1H NMR spectroscopy. It
was established that the endo and exo DA stereoisomers show
significantly different thermal responses: the rDA of the endo
DA-adducts typically takes place at 20–40 K lower temperatures
than that of the corresponding exo DA-adducts in all cases,
with the exception of some aromatic maleimides. Although in
situ isomerization was observed to a limited extent and only in
some cases, this effect is not expected to influence the ther-
moremendability of DA-crosslinked networks being dependent
C
V
on two separate stereoisomeric rDA steps.
2010 Wiley Peri-
odicals, Inc. J Polym Sci Part A: Polym Chem 48: 3456–3467,
2010
KEYWORDS: crosslinking; Diels–Alder cycloaddition; Diels-Alder
polymers; differential scanning calorimetry; furans; isomer;
maleimides; NMR; stereoisomers; thermoremendable
INTRODUCTION The Diels–Alder (DA) reaction is one of the
most versatile reactions in organic chemistry, in which a
diene and a dienophile react to form a cyclohexene.1,2 One of
the interesting characteristics of this reaction from materials
science perspective is the thermal reversibility of some of
the DA adducts, repetitively regenerating the original reac-
tants. The furan-maleimide combination in DA reactions is
one of the most often used because this combination com-
monly exhibits fast kinetics and high yields. Incorporation of
this DA-based crosslinks in thermoset materials has been
used to introduce a thermoremendable3–7 and recyclable8–10
functionality.
Although the stereochemical effects are well documented for
the DA and rDA reactions in molecular and synthetic organic
chemistry studies,20 the consequences of these stereochemical
effects have not yet been taken into account in the field of
polymer chemistry and materials science. No distinction is
made between the endo and exo stereoisomers, assuming that
their thermal behavior is at least similar if not equal. However,
differences in thermal reversibility (rDA, DA) can have an im-
portant influence on the material properties, especially in the
case of thermoremendable polymers and networks. Although
the furan-maleimide DA-adducts have been reported to show
the onset of rDA at about 100 ꢀC,10 the practical use of de-
7
ꢀ
crosslinking a polymer network is described at 150–170 C.
Our goal was to establish whether stereochemical effects can
help explain this large temperature difference.
Furthermore, the furan chemistry has attracted a big interest
for its renewable character11 (furfural is commonly obtained
in the industry from agricultural and forestry sources). Using
DA reactions between furans and maleimides mainly two dif-
ferent synthetic approaches have been employed to create
macromolecular structures: (1) a step-growth polymerization
through multifunctional complementary monomers,9,12–15
e.g., bisfuran and bismaleimide monomers; and (2) a modifi-
cation of linear polymers with pendant furan or maleimide
groups with complementary DA reactants,16–19 e.g., the
crosslinking of an acrylic polymer with pendant furan groups
with a bismaleimide.
Furthermore, a partial thermal reversibility at lower temper-
atures can be beneficial for reducing internal stresses of
crosslinked thermosets, in analogy to reversible supramolec-
ular crosslinks.21,22 By breaking only the covalent bonds of
the endo DA-isomers inside the network under mild condi-
tions and reforming them in a less stressed state, stresses
can be relaxed. Important material properties such as dura-
bility, adhesion to other materials, appearance of micro-voids
and micro-cracks, etc. can be improved.
Correspondence to: R. A. T. M. van Benthem (E-mail: r.a.t.m.v.benthem@tue.nl)
C
V
Journal of Polymer Science: Part A: Polymer Chemistry, Vol. 48, 3456–3467 (2010) 2010 Wiley Periodicals, Inc.
3456
INTERSCIENCE.WILEY.COM/JOURNAL/JPOLA

ARTICLE
The purpose of this investigation is to study the influences
of the properties of the endo/exo stereoisomers of furan-
maleimide DA-adducts on the thermal reversibility of model
compounds, polymeric and network structures by using
differential scanning calorimetry (DSC), where possible in
combination with 1H NMR spectroscopy. In addition, the influ-
ence of structural variations in both the furan and maleimide
moieties of the model compounds is investigated with respect
to the (rDA) decomposition temperatures of the DA adducts.
(Aldrich)] or synthesized as described below.24 4-(N,N-dimethy-
lamino)pyridine (DMAP) and dicyclohexylcarbodiimide (DCC)
were purchased from Aldrich (Zwijndrecht, The Netherlands).
The solvents methanol, dichloromethane, chloroform, dimethyl-
formamide and acetone were obtained from Biosolve (Valkens-
waard, The Netherlands).
Synthesis of Furans
EXPERIMENTAL
Materials
Furfuryl methacrylate (FMA; Aldrich, Zwijndrecht, The Neth-
erlands) was purified by passing through a short alkaline alu-
mina column to remove the inhibitor, 2,20-Azobis(isobutyroni-
trile) (AIBN; Aldrich) was recrystallized from methanol before
use. Furan compounds were synthesized as described
below.23 Maleimides were obtained from commercial sources
[N-propylmaleimide (Aldrich), N-phenylmaleimide (Aldrich),
N-benzylmaleimide (Aldrich), N-tertbutylmaleimide (UKOrg-
Synthesis, Kiev, Ukraine), N-(1-methylpropyl)maleimide
A typical procedure is as follows: Under nitrogen atmos-
phere, isobutyric acid (1 mol) was added dropwise to a solu-
tion of furfuryl alcohol derivate (1.2 mol), DCC (1.2 mol) and
DMAP (0.25 mol) in dry CH2Cl2. The reaction mixture was
stirred overnight at room temperature. The by-product dicy-
clohexylurea (DCU) formed during the reaction was filtered
off and the solution was concentrated in vacuo. The purifica-
1
tion method, the yield, and the H-NMR spectroscopy data of
(UKOrgSynthesis),
N,N-4,4-(diphenylmethane)bismaleimide
each compound are listed below.
R1
H
R2
H
Purification method
Yield (%)
59
1H NMR [CDCl3; d (ppm)]
Vacuum distillation
(1.2 Pa, 26–28 ꢀC)
1.16(d, 6H, ACH3, J ¼ 7.2 Hz); 2.56(heptet, 1H, ACHA, J ¼
7.2 Hz); 5.05(s, 2H, ACH2AOA); 6.35(dd, 1H, ¼CHA, J ¼ 3.2,
1.6 Hz); 6.38(d, 1H, ¼CHA, J ¼ 3.2 Hz); 7.41(dd, 1H,
¼CHAO, J ¼ 1.6, 0.8 Hz)
1.16(d, 6H, ACH3, J ¼ 7.2 Hz); 2.07(s, 3H, ACH3); 2.56(hep-
tet, 1H, ACHA, 7.2 Hz); 5.03(s, 2H, ACH2AOA); 6.22(d, 1H,
¼CHAC, J ¼ 1.6 Hz); 7.32 (d, 1H, ¼CHAO, J ¼ 1.6 Hz)
Me
H
H
Flash chromatography
on silica gel column
(eluents: ethyl acetate/
n-pentane 1:4 v:v)
54
64
66
Me
Br
Flash chromatography
on silica gel column
1.16(d, 6H, ACH3, J ¼ 6.8 Hz); 2.29(d, 3H, ACH3, J ¼ 0.8 Hz);
2.57(heptet, 1H, ACHA, J ¼ 6.8 Hz), 5.93 (m, 1H, ¼CHA);
6.27 (dd, 1H, J ¼ 2.8 Hz, 0.8 Hz)
(eluents: ethyl acetate/
petroleum ether 1:1 v:v)
H
Flash chromatography
on silica gel column
(eluents: ethyl acetate/
n-pentane 1:9 v:v)
1.17(d, 6H, ACH3, J ¼ 6.8 Hz); 2.57(heptet, 1H, ACHA, J ¼
6.8 Hz); 5.00(s, 2H, ACH2AOA); 6.28(d, 1H, ¼CHAC, J ¼ 3.2
Hz); 6.37(d, 1H, ¼CHAC, J ¼ 3.2 Hz)
a yellowish oil. 1H NMR [CDCl3; d (ppm)]: 1.65(m, 4H,
ACH2); 2.34(m, 4H, ACOACH2A); 6.35(dd, 2H, - ¼¼CHA, J ¼
3.2, 2.0 Hz); 6.39(d, 2H, ¼¼CHA, J ¼ 3.2, 2.0 Hz); 7.41(dd,
2H, ¼¼CHAO, J ¼ 2.0, 0.8 Hz).
About 10.15 g (69.4 mmol) of adipic acid was added in small
proportions over a period of 2 h at a temperature below 5
ꢀC in a nitrogen atmosphere to a solution of 15.0 g (152.9
mmol) of furfuryl alchol, 31.55 g (152.9 mmol) of DCC, and
2.54 g (20.79 mmol) of DMAP. After the addition was com-
pleted, the mixture was stirred for 24 h at room tempera-
ture. DCU was removed by filtration and the filtrate was con-
centrated under reduced pressure. The brown liquid was
purified by column chromatography on silica gel and eluting
with ethyl acetate/n-pentane 1:4 v:v, giving 16.2 g (16%) of
Synthesis of Maleimides
Into a 500-mL, three-neck round-bottom flask equipped with
a nitrogen inlet, a condenser, and a magnetic stirrer, 27.45 g
(0.28 mol) of maleic anhydride was dissolved in 150 mL of
STEREOISOMERIC EFFECTS IN POLYMER NETWORKS, CANADELL ET AL.
3457

JOURNAL OF POLYMER SCIENCE: PART A: POLYMER CHEMISTRY DOI 10.1002/POLA
DMF. After the complete dissolution, 30.0 g (0.28 mol) of o-
toluidiꢀne was added in one portion. The solution was heated
to 90 C for 30 min; then, 114.33 g (1.12 mol) of anhydride
acetic, 0.50 g (2.0 mmol) of Nickel (II) acetate tetrahydrate,
and 10.0 g (98.82 mmol) of triethylamine were added. The
reaction mixture was maintained at that temperature for 30
min more. After it cooled at room temperature, the solution
was poured into 2 L of ice-water and stirred for 1h. The
resulting brown precipitate was redissolved with CHCl3 and
washed with water three times. The product was purified by
chromatography on a silica gel column using ethyl acetate/n-
pentane 1:4 v:v, yielding 16 g of a white crystals (16.2 g,
31%). 1H NMR [CDCl3; d (ppm)]: 2.16(s, 3H, ACH3); 6.68(s,
2H, ¼ CHA); 7.11(d, 1H, arom); 7.35–7.27(m, 3H, arom).
A typical procedure of synthesis of bismaleimides is as fol-
lows: To a solution of maleic anhydride (2 mol) in DMF,
the bisamine (1 mol) was added at room temperature.
The solution was heated to 90 ꢀC for 30 min; then, acetic
anhydride (4 mol), 0.50 g (2.0 mmol) of Nickel (II) acetate
tetrahydrate (8 mmol), and triethylamine (0.4 mmol) were
added. The reaction mixture was maintained at that tem-
perature for 30 min more. After it cooled at room temper-
ature, the solution was poured into ice-water and stirred
for 1 h. The resulting brown precipitate was redissolved
with CHCl3 and washed with water three times. The puri-
fication method, the yield, M.P., and the 1H NMR spectros-
copy data of both compounds are listed below.
R
a
Purification method
M.P. (ꢀC)
140
Yield (%)
41
1H NMR [CDCl3; d (ppm)]
Flash chromatography on silica gel column
(eluents: ethyl acetate/n-pentane 7:3 v:v).
Recrystallization from acetone.
1.29(m, 4H, ACH2A); 1.56(m, 4H, ACH2A);
3.49(t, 4H, NACH2A); 6.67(s, 4H, ¼CHA)
b
Flash chromatography on silica gel column
(eluent: ethyl acetate). Recrystallization from
ethyl acetate/n-pentane 50:50
83–85
32
0.70(t, 3H, CH3); 2.20–1.50(4m, 4H, ACH2A);
3.56–3.3(m, 2H, NACH2A); 3.82(m, 1H, ACHA);
6.61(s, 4H, ¼CHA)
Product Characterization
added dropwise into n-pentane to precipitate the polymer as
a white powder. The polymer was further purified by repreci-
pitation of the dissolved polymer from dichloromethane into
methanol and dried under vacuum at room temperature. The
1H NMR and gCOSY spectra were obtained with a Varian Gem-
ini 400 spectrometer with Fourier transform. Calorimetric
studies (DSC) were carried out on a TA Instruments Q2000-
1037 DSC thermal analyzer in hermetic aluminum pans under
N2 at scanning rates of 2 K/min. In all cases, the sample
weight was approximately 5 mg. Size exclusion chromatogra-
phy (SEC) was performed using a Waters GPC equipped with
Waters model 510 pump and a model 410 differential refrac-
tometer. A set of two mixed bed columns (Mixed-C, Polymer
1
yield was 7.2 g (72%). The polymer was characterized by H
NMR spectroscopy, GPC (Mn ꢂ 10,300 g/mol and Mw
ꢂ
17,800 g/mol) and DSC (Tg ¼ 65 ꢀC).
Synthesis of DA Adducts From the Polyfurfuryl
Methacrylate
The following procedure describes a typical procedure of
modification via the DA reaction of the pendant furan rings
in P1 with monofunctional maleimides: stoichiometric
amounts of furan and maleimide were dissolved in chloro-
form. The solution was heated under nitrogen atmosphere at
55 ꢀC during 48 h. The polymers were precipitated into
methanol, washed two times with methanol and vacuum-
dried at room temperature.
ꢀ
Laboratories, 30 cm, 40 C) was used. Tetrahydrofuran stabi-
lized with BHT was used as the eluent, and the system was
calibrated using narrow molecular weight polystyrene stand-
ards ranging from 600 to 7 ꢁ 106 g/mol.
Synthesis of Model DA Adducts
A typical procedure is as follows: Stoichiometric amounts of
furan and maleimide were dissolved in chloroform. The solu-
ꢀ
tions were heated at 55 C from 24 to 48 h and the ensuing
Crosslinking via the DA Reaction
products purified by chromatography on a silica gel column
The following procedure describes a typical procedure of
network formation via the DA reaction of the pendant furan
using ethyl acetate/n-pentane as eluents.
Synthesis of Polyfurfuryl Methacrylate, P1
The polymer with pendant furan rings was prepared by free-
radical polymerization using the following procedure: A solu-
tion of 10 g (60.18 mmol) of FMA and 100 mg of AIBN (0.61
mmol) in toluene (100 mL) in a slench tube was degassed
and flushed with nitrogen to create an inert atmosphere. After
it was heated at 90 ꢀC for 20 h, the reaction mixture was
SCHEME 1 DA reaction between furans and maleimides.
3458
INTERSCIENCE.WILEY.COM/JOURNAL/JPOLA

ARTICLE
SCHEME 2 Model DA adducts.
1
FIGURE 1 H NMR spectra of the DA adduct 1 kept at 100 ꢀC for (a) 0 h, (b) 3 h (solvent, 1,2-dichlorobenzene-d4).
STEREOISOMERIC EFFECTS IN POLYMER NETWORKS, CANADELL ET AL.
3459

JOURNAL OF POLYMER SCIENCE: PART A: POLYMER CHEMISTRY DOI 10.1002/POLA
dichloromethane, reprecipitated in methanol, and vacuum-
dried at room temperature.
RESULTS AND DISCUSSION
All the systems studied here are based on DA adducts derived
from furan-maleimide (Scheme 1). To facilitate the investiga-
tion of the structural influence of furans and maleimides and
the endo/exo behavior on the thermal reversibility of the DA
reaction, different model DA adducts were synthesized.
Scheme 2 illustrates the structures of the model compounds.
The structure of the DA adducts were determined by 1H
NMR spectroscopy. As an example, Figure 1a shows the 1H
NMR spectrum of 1. The characteristics peaks assigned to
FIGURE 2 gCOSY spectra of 1 in CDCl3.
rings in P1 with bifunctional maleimides: stoichiometric
amounts of furan and maleimide were dissolved in chloro-
form. The solution was heated under nitrogen atmosphere at
55 ꢀC during 48 h yielding a gel. Unreacted starting materi-
als were removed by Soxhlet extraction with dichlorome-
thane for 4 h. The final crosslinked polymers were vacuum
dried at room temperature.
Synthesis of Polymers with DA Functionality
in the Main Chain
The linear polymers with DA functionality in the main chain
were obtained by the step-growth polymerization of an stoi-
chiometric mixture of a bisfuran and a bismaleimide in chlo-
roform for 48 h at 55 ꢀC under nitrogen atmosphere. The
polymers were precipitated into methanol, redissolved in
TABLE 1 Composition and Thermal Behaviour
of the Model DA Adducts
DA Adduct Endo/Exo (%)a MP (ꢀC)
TrDA (ꢀC)
1
2
3
4
5
6
7
60/40
45/55
51/49
67/33
44/56
61/39
100/0
0/100
100/0
36/65
26/74
NDb
—
—
—
—
—
—
—
66
98
—
—
—
98 (endo), 138 (exo)
101 (endo), 143 (exo)
94 (endo), 124 (exo)
92 (endo), 126 (exo)
95 (endo), 135 (exo)
96 (endo), 140 (exo)
91
114
8
125
9
102 (endo), 134 (exo)
97 (endo), 130 (exo)
98 (endo), 126 (exo)
10
11
12
a
19/81
157 (exo) 110 (endo), 157 (exo)
The ratio endo/exo was determined after purification by column
chromatography.
FIGURE 3 Dynamic DSC thermograms of model DA adducts
b
Not determined due to the overlapping of the peaks.
(scanning rate, 2 K/min).
3460
INTERSCIENCE.WILEY.COM/JOURNAL/JPOLA

ARTICLE
1
FIGURE 4 H NMR spectra of the DA adduct 6 in 1,2-dichlorobenzene-d4 at 25 ꢀC (a), 10 min at 95 ꢀC; (b), 1 h at 95 ꢀC; (c) and 1 h at
150 ꢀC. [The ratio endo/exo from spectrum a and b was determined by peak 7endo and 8exo and the rDA percentage by peak d and 3.]
the adduct appear at 6.27–6.20 ppm for the H4 and H5 of
the double bond, as well as the bridgehead H6 proton at
5.09 (endo) and 5.07 (exo). Additionally, the protons of the
fused rings appears at 3.42 ppm (H7-endo), 3.18 (H8-endo),
and 2.73-2.67(H7-exoþH8-exo). From the relative integral
values of the peaks H7-endo and H7-exoþH8-exo, the ratio of
endo and exo stereoisomers was determined. The stereo-
chemistry of the isomers could be assigned unequivocally by
STEREOISOMERIC EFFECTS IN POLYMER NETWORKS, CANADELL ET AL.
3461

JOURNAL OF POLYMER SCIENCE: PART A: POLYMER CHEMISTRY DOI 10.1002/POLA
SCHEME 3 Synthesis of polymer architectures containing DA adducts.
means of gCOSY experiments,25 because the dihedral angle
H6-H7 are close to 90ꢀ for the exo adduct, no correlation
was observed between these two protons (Fig. 2). Further-
more, the assignments are in agreement with previously
reported data of similar compounds.8,12 The stereochemistry
of the isomers of the DA adducts 7 and 8 was determined
differently due to the absence of H6. Therefore, it was
assumed that the H7-endo and H8-endo appear at higher
chemical shifts in the 1H NMR spectroscopy, because this
was the case of all the previous compounds.
firmed by 1H NMR spectroscopy by following the rDA reac-
tion at the maxima temperature of the endothermic peak
observed by DSC. A typical experiment is shown in Figure 1,
where the spectrum (a) shows the 1H NMR spectra of the
ꢀ
DA-adduct 1 and (b) after heating at 100 C for 3 h, showing
the appearance of the peaks associated to the starting furan
and maleimide, proving that the rDA occurs at this tempera-
ture. Small peaks associated to the exo adduct are still
observed in the 1H NMR spectrum b of Figure 1. This is an
indication that the exo adduct partly reverts at 90 ꢀC. Prob-
ably, the less pronounced peak at higher temperatures of the
DSC curves can be associated to the rDA of the exo adduct.
To determine the temperature of the rDA reaction, DSC was
carried out. Table 1 summarizes the composition (endo/exo
ratio) and the thermal transitions of the model DA adducts.
The DSC results of the mono-functional DA adducts obtained
from different maleimides are shown in Figure 3a. It can be
seen from these results, that all the compounds have a broad
endothermic peak with maxima ranging from 92 to 101 ꢀC,
which could be ascribed to the rDA reaction. This was con-
Analogous to the previously described DSC results, the
mono-functional DA adducts from the investigated furans
(Fig. 3b) show two broad endothermic peaks associated to
the rDA reaction of the endo and exo stereoisomers, which
1
were again confirmed and assigned by H NMR spectroscopy.
Figure 4a shows the 1H NMR spectrum of the endo/exo
3462
INTERSCIENCE.WILEY.COM/JOURNAL/JPOLA

ARTICLE
FIGURE 5 1H NMR spectrum of P1, P1-Ma, P1-Mb, and P1-Mc in CDCl3.
adduct 6 with a ratio of 60/40, respectively. After 10 min at
95 C (Fig. 4b), the yield of rDA was approximately 30% and
endo and exo stereoisomers as we observed previously with
the monofunctional adducts. Adduct 9 shows clearly two
peaks associated to the rDA of the endo (102 ꢀC) and exo
(134 ꢀC) stereoisomer, which are less distinguishable for 10
and 11. The rDA for the adduct 12 occurred at 110 ꢀC for
the endo stereoisomer, whereas the exo takes place at higher
temperature together with the melting point. As we previ-
ously mentioned, the study of these model compounds was
performed essentially to establish the structural influence of
furans and maleimides and the endo/exo behavior on the
rDA reaction. As we have shown, the DA adducts of various
furans and maleimides show similar thermal responses and
ꢀ
the ratio of endo/exo had changed to 33/37. This indicates
that at 95 ꢀC only the rDA of the endo stereoisomer takes
place. The rDA of the endo stereoisomer occurred completely
after 1 h at this temperature (Fig. 4c). By increasing the tem-
perature to 150 ꢀC, the rDA of the exo stereoisomer took place
(Fig. 4d). Therefore, the endothermic peak with the lowest
temperature observed on the DSC curve is assigned to the
rDA of the endo stereoisomer and the endothermic peak with
higher temperature to the rDA of the exo steroisomer. After
isolation of the stereoisomers of adduct 7 by column chroma-
tography a separate investigation of both isomers by DSC
could be performed. The curves of the DSC runs using the sep-
arate isomers have been plotted in Figure 3b and the rDA ꢀof
the endo adduct appears to take place at approximately 23 C
lower than the exo adduct. Note that the sharp endothermic
TABLE 2 Chemical Compositions of P1-Ma, P1-Mb, and P1-Mc
Polymer
DA Adduct (%)
Endo/Exo (%)
ꢀ
peaks of the curves 7exo and 8 at 68 and 98 C, respectively,
P1-Ma
P1-Mb
P1-Mc
75
60
85
45/55
46/54
19/81
are attributed to the melting of the DA adduct.
The bisfunctional DA adducts (Fig. 3c) also display in
dynamic DSC runs the two broad peaks associated to the
STEREOISOMERIC EFFECTS IN POLYMER NETWORKS, CANADELL ET AL.
3463

JOURNAL OF POLYMER SCIENCE: PART A: POLYMER CHEMISTRY DOI 10.1002/POLA
the endo stereoisomer always reverts at lower temperatures
than the exo.
The next step was to examine the behavior on macromolecu-
lar architectures. Scheme
3 summarizes our synthetic
approach: (i) DA reactions between the pendant furan
moieties attached to a polymeric chain and mono and bisma-
leimides; and (ii) DA reactions from bisfurans and bismalei-
mides to form linear polymers.
The methacrylic homopolymer containing furan groups (P1)
was obtained by conventional free-radical polymerization.
Although some authors reported the formation of insoluble
gels due to excessive chain transfer involving the furfuryl
groups,26,27 by working in diluted conditions, we obtained a
soluble polymer with Mn ¼ 10,300 g/mol and Mw ¼ 17,800
g/mol (averaged over two synthesis). The relatively low
value Mw/Mn ¼ 1.7 suggests that no side reactions took
place. Furthermore, the polydispersity is similar to other
synthetic methods described that are using controlled radical
polymerization (atom transfer radical polymerization,
1
ATRP).26 Figure 5 shows the H NMR spectrum with the cor-
responding assignments.
In a first step, P1 was modified with monofuntional malei-
mides to obtain soluble polymers (P1-Ma, P1-Mb, and P1-
Mc), which simplifies a correct structural characterization. As
1
can be seen from H NMR spectra (Fig. 5), all three polymers
show signals of protons of the furan ring together with pro-
tons of the adduct, indicating that the addition did not pro-
ceed fully. Furthermore, a mixture of both DA stereoisomers is
observed. From the relative integral values of corresponding
signals, both percentages of substitution and the endo/exo ra-
tio was calculated, which are summarized in Table 2.
The thermal reversibility of the DA adducts from the grafted
polymers were studied by DSC in combination of 1H NMR
spectroscopy to assign the observed thermal transitions. The
DSC thermograms of P1-Ma and P1-Mb (Fig. 6a) show two
distinguishable broad endothermic peaks at ꢃ100 ꢀC and
1
ꢀ
135 C. By making use of H NMR spectroscopy, we assigned
the lower temperature peak to the rDA of the endo adduct
and the higher temperature peak to the rDA of the exo
adduct (Fig. 7A).
FIGURE 6 Dynamic DSC thermograms of polymers. (a) P1-Ma,
P1-Mb, and P1-Mc; (b) P1-BisMa, P1-BisMb, and P1-BisMc; (c)
F-BisMa, F-BisMb, and F-BisMc (scanning rate, 2 K/min).
Although we did not observe thermal endo-exo isomerization
of the DA adducts, this is a well known process.13,28 To ana-
lyze the thermal isomerization, a polymer sample P1-Ma
was heated in solid state (which resembles more the DSC
conditions) at 100 ꢀC for a short period of time and by re-
cording a 1H NMR spectrum afterward, we confirmed that
the endo adduct was isomerized partially to the exo. Figure
In terms of applicability, for instance for a coating material,
internal stresses can be reduced by heating to a temperature
at which the rDA of the endo adduct takes place. This pro-
cess will continue until all the endo estereoisomer has been
converted to exo. However, this effect is not expected to
influence the thermoremendability of DA-polymers, because
this takes place at higher temperatures where both ster-
oisomers revert to the starting furan and maleimide. After
reformation of the DA adducts both stereoisomers will be
present again.
1
ꢀ
7B shows both H NMR spectra of P1-Ma at 25 C and after
heating for 30 min at 100 ꢀC in solid state. From spectrum
a, the ratio endo/exo was estimated to be 44/56, respec-
ꢀ
tively. After heating at 100 C (spectrum b), the yield of the
rDA was approximately 13% and the ratio endo/exo had
changed to 26/74, respectively. This result indicates that at
100 ꢀC in solid state, the endo isomer partly isomerizes to
the exo isomer (cal. 38%) and partly decomposes to the
starting maleimide and furan containing polymer (cal. 62%).
The thermal behavior of the polymer P1-Mc is different
from those of the previous polymers. Only one endothermic
3464
INTERSCIENCE.WILEY.COM/JOURNAL/JPOLA

ARTICLE
FIGURE 7 (A) 1H NMR spectra of the
P1-Ma in 1,2-dichlorobenzene-d4 at
(a) RT, (b) 1 h at 100 ꢀC, and (c) 1 h
at 130 ꢀC. (B) 1H NMR spectra of the
P1-Ma in CDCl3 at a) RT, and (b) af-
ter heating 30 min at 100 ꢀC in solid
state. [The ratio endo/exo from spec-
trum a and b of Figure B was deter-
mined by peak gendo and gexoþhexo
and the rDA percentage by peak A
and f.]
ꢀ
peak can be observed around 123 C (Fig. 6a). The fact that
the endothermic peak is broader than for P1-Ma and P1-Mb
suggest that the rDA of the endo and exo stereoisomer take
places at close temperatures, and not possible to separate by
DSC. Previously, for the model DA compounds containing ar-
omatic structure, such as adduct 12, two peaks were
observed. Because, one of the isomers was liquid and the
other was solid, the rDA of the exo stereoisomer took place
together with the melting, due to lack of mobility. Addition-
ally, note that compound 4 contains a methyl group in the
aromatic ring to avoid crystallinity, and therefore cannot be
directly compared to the polymer P1-Mc, which does not
contain this methyl group.
In a second step, P1 was modified with bisfunctional malei-
mides obtaining crosslinked polymers (P1-BisMa, P1-BisMb,
and P1-BisMc). Again the thermal behavior was evaluated
by dynamic DSC experiments. As can be seen from the ther-
mograms plotted in Figure 6b, two endotermic peaks are
STEREOISOMERIC EFFECTS IN POLYMER NETWORKS, CANADELL ET AL.
3465

JOURNAL OF POLYMER SCIENCE: PART A: POLYMER CHEMISTRY DOI 10.1002/POLA
TABLE 3 Characterization of Polymers F-BisMa, F-BisMb, and
observed for P1-BisMa and P1-BisMb. By comparison with
the homologue linear polymers previously described, the
peak at lower temperatures was assigned to the rDA of the
endo adduct and the one at higher temperatures to the exo.
The shift of the endothermic peak at lowest temperature
compared to the linear homologues can be explained by the
restrictions of mobility of the crosslinked system. Finally, the
rDA of P1-BisMc was observed at 158 ꢀC as a single
transition.
F-BisMc
Mn
Mw
Endo/
Furan/
Maleimide (%)a
Polymer
(g/mol)
(g/mol)
Exo (%)
F-BisMa
F-BisMb
F-BisMc
a
5,920
3,700
1,900
8,900
5,300
2,600
50/50
NDb
63/37
60/40
50/50
33/67
Composition end-groups.
The linear copolymers with DA functionality in the main
chain were obtained by step-growth polymerization of bis-
furans and bismaleimides. The resulting polycondensates (F-
BisMa, F-BisMb and F-BisMc) had low molecular weights
ranging from 5920 g/mol to 1900 g/mol with polydisper-
sities around 1.4. The 1H NMR spectrum of F-BisMa is
shown in Figure 8, suggesting that both furan and maleimide
are present as end-groups, with an excess of furan end-
groups. This observation is in agreement with previously
reported data of similar polymers.12 The molecular weights,
the ratio endo/exo, and the percentage of furan and malei-
mide end-groups of the three polymers are listed in Table 3.
b
Not determined due to overlapping of the peaks.
effect on the (rDA) decomposition temperatures. The DSC
thermograms typically show endothermic reactions occurring
in the range 90–140 ꢀC in all cases. Two subsequent endo-
thermic processes can be discerned within this range that
can be ascribed to stereochemical effects.
With the exception of adducts derived from aromatic malei-
mides, quite similar observations were made for the poly-
meric and network architectures. It was established with the
additional use of NMR that both, the model compounds and
the polymeric structures, the endo and exo DA stereoisomers
show significantly different thermal responses: the rDA of
the endo DA-adducts typically takes place at 20–40 K lower
temperatures than that of the corresponding the exo DA-
adducts.
The thermal reversibility of these polymers was evaluated by
DSC experiments. F-BisMa and F-BisMb present a similar
behavior. As can be seen from Figure 6c, both polymers
show glass transition temperature (Tg) followed by two clear
endothermic peaks of the rDA of both the endo and exo ste-
reoisomer, respectively. The higher Tg of F-BisMc is related
to the aromatic structure of the bismaleimide, which confers
greater rigidity to the polymer. Furthermore, only one broad
In general, the rDA of endo adducts takes place around 100
ꢀC and the exo around 135 ꢀC. For the crosslinked systems,
the temperatures were around 120 ꢀC and 140 ꢀC, respec-
tively. For the aromatic structures, the endo and the exo rDA
temperature were close together and therefore difficult to
separate by DSC.
ꢀ
endothermic peak is observed around 124 C, where the rDA
reaction of both the endo and exo stereoisomers occur.
CONCLUSIONS
Although in situ isomerization was observed to a limited
extent in some cases, this effect is not expected to influence
the thermoremendability of DA-crosslinked networks being
dependent on two separate stereoisomeric rDA steps.
In this study, several structurally varied model compounds,
polymer and network architectures containing DA adducts
from furan and maleimide groups have been synthesized,
obtaining in all the cases a mixture of endo and exo stereo-
isomers as recorded by 1H NMR. Their thermal reversibility
(rDA, DA) was studied by DSC.
We acknowledge Innovatiegerichte Onderzoekprogramma’s
(IOP Self-healing materials) for support through grant SHM
08740.
With respect to the structural variations in the DA-adduct
model compounds it was concluded that they had limited
REFERENCES AND NOTES
1 Carruthers, W. Cycloaddition Reactions in Organic Synthesis;
Pergamon Press: Oxford, 1990.
2 Nicolaou, K. C.; Snyder, S. A.; Montagnon, T.; Vassilikogian-
nakis, G. Angew Chem Int Ed Engl 2002, 41, 1668–1698.
3 Chen, X.; Dam, M. A.; Ono, K.; Mal, A.; Shen, H.; Nutt, S.;
Sheran, K.; Wudl, F. Science 2002, 295, 1698–1702.
4 Chen, X.; Wudl, F.; Mal, A. K.; Shen, H.; Nutt, S. R. Macromo-
lecules 2003, 36, 1802–1807.
5 Liu, Y.-L.; Hsieh, C.-Y. J Polym Sci Part A: Polym Chem 2006,
44, 905–913.
FIGURE 8 1H NMR spectra of F-BisMA in CDCl3 (F are the peaks
of therminal furan end-groups and MA is the peak of therminal
maleimide end-groups).
6 Tian, Q.; Yuan, Y. C.; Rong, M. Z.; Zhang, M. Q. J Mat Chem
2009, 19, 1289–1296.
3466
INTERSCIENCE.WILEY.COM/JOURNAL/JPOLA

ARTICLE
7 Wouters, M.; Craenmehr, E.; Tempelaars, K.; Fischer, H.;
19 Goiti, E.; Heatley, F.; Huglin, M. B.; Rego, J. M. Eur Polym
Stroeks, N.; van Zanten, J. Prog Org Coat 2009, 64, 156–162.
Mater 2004, 40, 1451–1460.
8 Watanabe, M.; Yoshie, N. Polymer 2006, 47, 4946–4952.
20 Lee, M. W.; Herndon, W. C. J Org Chem 1978, 43, 518.
9 Kuramoto, N.; Hayashi, K.; Nagai, K. J Polym Sci Part A:
21 Wietor, J.-L.; Dimopoulos, A.; Govaert, L. E.; van Benthem,
R. A. T. M.; de With, G.; Sijbesma, R. P. Macromolecules 2009,
42, 6640–6646.
Polym Chem 1994, 32, 2501–2504.
10 Zhang, Y.; Broekhuis, A. A.; Picchioni, F. Macromolecules
2009, 42, 1906–1912.
22 Dimopoulos, A.; Wietor, J.-L.; Wubbenhorst, M.; Napolitano,
¨
S.; van Benthem, R. A. T. M.; de With, G.; Sijbesma, R. Macro-
molecules, in revision.
11 Gandini, A. Pol´ımeros: Ciencia e Tecnologia 2005, 15, 95–101.
12 Gandini, A.; Coelho, D.; Silvestre, A. J. D. Eur Polym J 2008,
23 Kose, M. M.; Yesilbag, G.; Sanyal, A. Org Lett 2008, 10,
44, 4029–4036.
2353–2356.
13 Gousse´, C.; Gandini, A. Polym Int 1999, 48, 723–731.
24 Kossmehl, G.; Nagel, H.-I.; Pahl, A. Angew Makromol Chem
14 Adzima, B. J.; Aguirre, H. A.; Kloxin, C. J.; Scott, T. F.; Bow-
1995, 227, 139–137.
man, C. N. Macromolecules 2008, 41, 9112–9117.
25 Gomes-Constantino, M.; da Silva-Filho, L. C.; Cunha-Neto,
A.; Gomes-Heleno, V. C.; da Silva, G. V. J.; Callegari-Lopes, J. J
Spect Acta: Part A 2005, 61, 171–176.
15 Yoshie, N.; Watanabe, M.; Araki, H.; Ishida, K. Polym Degrad
Stab 2010, 95, 826–829.
16 Gousse´, C.; Gandini, A.; Hodge, P. Macromolecules 1998,
26 Kavitha, A.; Choudhury, A.; Singha, N. K. Macromol Symp
31, 314–321.
2006, 240, 232–237.
17 Imai, Y.; Itoh, H.; Naka, K.; Chujo, Y. Macromolecules 2000,
27 Goiti, E.; Huglin, M. B.; Rego, J. M. Polymer 2001, 42,
33, 4343–4346.
10187–10193.
18 Gheneim, R.; Perez-Berumen, C.; Gandini, A. Macromole-
cules 2002, 35, 7246–7253.
28 Jegat, C.; Mignard, N. Polym Int 2008, 60, 799–808.
STEREOISOMERIC EFFECTS IN POLYMER NETWORKS, CANADELL ET AL.
3467