Small-Molecule Investigation of Diels-Alder Complexes for Thermoreversible Crosslinking in Polymeric Applications

Article abstract of DOI:10.1021/acs.joc.1c00855

Combinations of dienes and dienophiles were examined in order to elicit possible combinations for thermoreversible crosslinking units. Comparison of experimental results and quantum calculations indicated that reaction kinetics and activation energy were much better prediction factors than change in enthalpy for the prediction of successful cycloaddition. Further testing on diene-dienophile pairs that underwent successful cycloaddition determined the feasibility of thermoreversibility/retro-reaction of each of the Diels-Alder compounds. Heating and testing of the compounds in the presence of a trapping agent allowed for experimental determination of reverse kinetics and activation energy for the retro-reaction. The experimental values were in good agreement with quantum calculations. The combination of chemical calculations with experimental results provided a strong insight into the structure-property relationships and how quantum calculations can be used to examine the feasibility of the thermoreversibility of new Diels-Alder complexes in potential polymer systems or to fine-tune thermoreversible Diels-Alder systems already in use.

Full text of DOI:10.1021/acs.joc.1c00855

pubs.acs.org/joc
Article
Small-Molecule Investigation of Diels−Alder Complexes for
Thermoreversible Crosslinking in Polymeric Applications
̈
Jarrett R. Rowlett,* Peter Deglmann, Johannes Sprafke, Nabarun Roy, Rolf Mulhaupt,
and Bernd Bruchmann
Cite This: J. Org. Chem. 2021, 86, 8933−8944
Read Online
sı
*
Metrics & More
Article Recommendations
Supporting Information
ACCESS
ABSTRACT: Combinations of dienes and dienophiles were examined in order to
elicit possible combinations for thermoreversible crosslinking units. Comparison of
experimental results and quantum calculations indicated that reaction kinetics and
activation energy were much better prediction factors than change in enthalpy for the
prediction of successful cycloaddition. Further testing on diene−dienophile pairs that
underwent successful cycloaddition determined the feasibility of thermoreversibility/
retro-reaction of each of the Diels−Alder compounds. Heating and testing of the
compounds in the presence of a trapping agent allowed for experimental
determination of reverse kinetics and activation energy for the retro-reaction. The
experimental values were in good agreement with quantum calculations. The
combination of chemical calculations with experimental results provided a strong
insight into the structure−property relationships and how quantum calculations can
be used to examine the feasibility of the thermoreversibility of new Diels−Alder
complexes in potential polymer systems or to fine-tune thermoreversible Diels−Alder
systems already in use.
normally be impossible to process due to high viscosity and
could be reprocessed as needed. With successful reversible
crosslinking, many other potential advantages could be
achieved as well depending on the polymer system it was
implemented into.
INTRODUCTION
■
The Diels−Alder reaction, first discovered in 1928 by Otto
Diels and Kurt Alder, is a well-studied and investigated
reaction mechanism within the chemistry field. This reaction
has been invaluable within the area of organic chemistry,
allowing for the formation of six-membered rings and other
complex ring structures while maintaining control over regio-
and stereoselectivity. However, the use of the Diels−Alder
reaction has not been limited to just organic chemistry. Many
groups have incorporated these units in a polymer structure in
order to adjust and influence polymer properties.1−14 These
applications include self-healing polymers, degradable/recycla-
ble polymers, crosslinking, and chain extension.15−23 All of
these applications rely on the reversibility of the Diels−Alder
complex utilizing different dienes and dienophiles to achieve
desired effects.
While many efforts have been focused on the self-healing
applications, reversible crosslinking/chain extension is also a
very interesting application for the Diels−Alder complexes as
functional groups incorporated into polymer species. By
utilizing these groups in the matrix of the polymer, it would
allow for a decrosslinked polymer with significantly lower
molecular weight upon heating and a high molecular weight or
crosslinked polymer upon cooling. This process is repeatable
since the conservation of the original diene and dienophile is
maintained in the retro Diels−Alder reaction. Thus, this would
allow access to polymers and molecular weights that would
Examples of attempts toward reversible crosslinking utilizing
cycloaddition/deaddition include dicyclopentadiene, furan−
maleimide, and anthracene−maleimide, among others.3−8,19−21
Typically, these consist of an electron-rich diene and an
electron-poor dienophile; however, in the case of dicyclopen-
tadiene, the low stability and high reactivity of the starting
cyclopentadiene overcome this barrier.3,19,24 Unfortunately, the
stability and reactivity of cyclopentadiene prevent the use of
this functional group in some polymers and high-temperature
applications.3,19,24 The anthracene−maleimide complex is not
thermally reversible and has been claimed to be reversible
when sonicated and in low concentrations in the polymer;
however, some of these publications have since been
retracted.17,25−27 The Craig group was able to take
sonication-activated “mechanophore” groups and show similar
Received: April 13, 2021
Published: June 21, 2021
https://doi.org/10.1021/acs.joc.1c00855
© 2021 American Chemical Society
J. Org. Chem. 2021, 86, 8933−8944
8933

The Journal of Organic Chemistry
pubs.acs.org/joc
Article
Figure 1. List of dienes and dienophiles tested for thermoreversibility.
reactivity with extrusion and mechanical force; however, this
has yet to be successfully applied to Diels−Alder systems.28,29
Thus, primarily, the focus of research for Diels−Alder systems
for reversible crosslinking has been the furan/maleimide
complex.13,14 While the furan/maleimide system has been
implemented successfully in many polymers, there is a ceiling
temperature limit of ∼100 °C for this diene and dienophile
pair.13,14,23 Thus, for higher-temperature systems, new or
improved Diels−Alder systems are necessary in order to reach
these higher ceiling temperature limits. Better understanding of
Diels−Alder systems is needed in order to extend the selection
of potential dienes and dienophiles that can be applied and
incorporated into polymer systems for thermoreversible
purposes.
Therefore, in this article, a series of dienes and dienophiles
are examined in order to determine how the structure and
functional groups affect the formation and reversibility of the
Diels−Alder cycloaddition. These experimental results in-
cluded both experimentally examining the addition and retro-
reaction for the cycloaddition. The kinetics and the energy
barrier of the retro-reactions were calculated by monitoring the
reverse reactions via NMR spectroscopy. These experimental
results are then used in parallel with quantum calculations of
the small molecules to illuminate which factors are the most
important in determining the formation and reversibility of the
complexes. Combining these results provides a strong basis for
how the quantum calculations can be used to filter or
investigate a wide range of new pairs/systems or alternatively
to fine-tune systems already in place.
in polymer applications, with the list of dienes and dienophiles
shown in Figure 1. Each pair of dienophile and diene were
reacted together to examine if Diels−Alder complexes would
form and, if one occurred, then examine if thermoreversibility
was possible. All of these materials were commercially available
with the exception of dienophiles C and D, which were
synthesized prior to testing. For the initial model reactions,
each combination was dissolved in DMSO-d6 and heated in an
increasing fashion until a reaction was observed. Reactions
were started at 45 °C, increased to 60 °C, and so forth until
145 °C. If no reaction was observed at 145 °C after 18 h, then
the diene and dienophile combination was declared unreactive
for the purposes of this article. While certainly some of the
unreactive diene and dienophile pairs could have formed given
higher temperature conditions, addition of catalysts or
stabilizers, longer reactions time, and so on for the purposes
of this investigation were excluded. Since this investigation is
focused on Diels−Alder systems for reversible chain extension
or crosslinking of polymers, systems that need temperatures
higher than 150 °C, long reaction times, or catalysts are likely
not applicable. An optimal system would undergo the retro-
Diels−Alder reaction and decrosslink when heated at a
temperature within a normal processing range and then return
to the original crosslinked state upon cooling of the polymer.
In a system that requires more than 150 °C to form, cycling of
the polymer could be difficult. Therefore, finding and
predicting which systems form readily in reasonable temper-
ature ranges was the first step of the investigation. The
summary of the initial experimental investigation results are
shown in Table 1, where it is denoted whether a reaction took
place. Additionally, for diene 4 (sorbic acid), it was found that
the molecule was unstable and underwent degradation at
RESULTS AND DISCUSSION
■
Five dienes and five dienophiles were chosen for the
investigation of thermoreversibility of Diels−Alder complexes
8934
https://doi.org/10.1021/acs.joc.1c00855
J. Org. Chem. 2021, 86, 8933−8944

The Journal of Organic Chemistry
pubs.acs.org/joc
Article
a
Table 1. Summary of Results from the Model Reactions of the Chosen Dienes and Dienophiles
diene
dienophile
G (TS)
ΔG
k (80 °C)
k (150 °C)
reaction
compound
1
2
3
4
A
A
A
A
180.5
146.9
152.6
160.8
170.5
171.3
187.8
126.2
119.0
116.3
121.4
137.6
122.6
124.4
140.4
135.2
133.0
136.6
145.7
144.9
141.0
178.5
179.1
172.3
170.8
178.5
193.2
188.2
141.0
135.0
129.1
136.2
137.6
−21.4
−22.2
−32.4
−51.0
−71.4
+36.2
+48.4
−48.6
−49.8
−62.7
−73.9
−78.2
+15.2
+18.4
−29.4
−35.3
−45.1
−46.1
−44.9
+41.9
+37.6
−2.7
3.1×10−13
3.5×10−8
3.3×10−9
2.6×10−10
3.9 × 10−12
1.2 × 10−11
6.4 × 10−14
3.2×10−5
3.2×10−4
5.7×10−4
2.6×10−4
8.7 × 10−7
2.4 × 10−4
1.3 × 10−4
3.1×10−7
1.7×10−6
2.8×10−6
1.5×10−6
5.9 × 10−8
1.4 × 10−7
4.3×10−7
5.8×10−13
5.8×10−13
1.0×10−11
1.2×10−11
9.4 × 10−13
1.2 × 10−14
7.1×10−14
2.2×10−7
2.1×10−6
1.1×10−5
1.9×10−6
2.3×10−6
4.9×10−10
7.1×10−6
1.4×10−6
1.3×10−7
8.7 × 10−9
6.9 × 10−9
6.5 × 10−11
2.5×10−3
1.9×10−2
4.1×10−2
1.0×10−2
9.7 × 10−5
6.9 × 10−3
4.2 × 10−3
4.4×10−5
1.9×10−4
3.5×10−4
1.2×10−4
9.8 × 10−6
1.2 × 10−5
3.7×10−5
8.7×10−10
7.6×10−10
5.2×10−9
7.9×10−9
8.9 × 10−10
1.4 × 10−11
5.7×10−11
3.7×10−5
2.1×10−4
1.1×10−3
1.5×10−4
9.7×10−5
×
×
×
×
5
A
×
1
2
3
4
B
B
B
B
√
√
√
√
I
II
III
IV
5
B
√
V
1
2
3
4
C
C
C
C
×
√
√
×
VI
VII
5
C
×
1
2
3
4
D
D
D
D
×
×
×
×
−1.1
−12.6
−29.7
−35.9
+63.5
+55.5
−24.4
−26.1
−25.2
−59.4
+46.7
5
D
×
1
2
3
4
5
E
E
E
E
E
×
√
√
×
VIII
IX
b
×
a
In the case of diene/dienophile combinations where this type of regioisomerism is possible, the upper entry denotes the reaction toward the endo
b
and the lower value toward the exo product. The preferred pathway is marked in bold. The Diels−Alder product was formed, but an additional
intramolecular transesterification reaction also took place.
higher temperatures, so only reaction temperatures up to 90
°C were undertaken for the combinations utilizing diene 4.
In addition to the experimental model reactions, the
reactivity of each pair of diene and dienophiles was calculated
from the procedure described in the Experimental Section to
give the theoretical values for energy change, activation energy,
and reaction rate constant at both 80 and 150 °C. The results
of the calculations are also shown in Table 1, with the results of
the experimental model reactions for comparison. Differences
between exo and endo along with the cis and trans
conformations are distinguished in Table 1. When the success
of the model reactions was compared with the molecular
modeling, it was clear that the success of the reaction
depended highly on the reaction kinetics. This was in contrast
to the enthalpy change for the reaction, where no trend could
be seen. In fact, some combinations that had large changes in
ΔG formed no complexes, while others with little energy
difference readily reacted. This was especially clear for
dienophiles A and E which were predicted to form very stable
complexes; however, the reaction kinetics were not favorable,
and no complexes formed with any of the dienes. The only
complexes that formed in the model reactions were the same
complexes predicted to be favorable in the calculations. A
complex was determined to be favorable if the predicted
reaction rate was in the range of 10−4 or higher at 150 °C
(based on the results from Table 1), except for complexes
using diene 4, which used 80 °C because of the previously
mentioned stability issues.
Thus, it was found that the activation energy is the major
barrier and the best predictive measure to analyze whether a
pair of dienes and dienophiles will undergo successful
cycloaddition to form a Diels−Alder complex in a reasonable
temperature range and time. While this result is certainly not
surprising, it does confirm that the quantum calculations used
here were successful in the ability to predict which
combinations would react better than others. Based on these
results, we can see that the approximate cutoff for activation
energy was 135 kJ/mol for a significant amount of the
cycloaddition reaction to occur at 150 °C in a timely manner.
Reaction rates were significantly higher and activation
energies were significantly lower for the electron-deficient
dienophiles B, C, and E than those for the more electron-rich
A and D. The difference in the electron density strongly
affected the reactivity. This could also be one reason why the
methylated dienophile C was less reactive than B, although
here, steric effects by the larger methyl group if compared to a
8935
https://doi.org/10.1021/acs.joc.1c00855
J. Org. Chem. 2021, 86, 8933−8944

The Journal of Organic Chemistry
pubs.acs.org/joc
Article
Figure 2. Structure and reaction of the combination of reactants 3 and E to form a dihydrobenzobarrelene and fused lactone ring structure,
compound IX.32
Figure 3. Structures of the eight compounds (I−VIII) synthesized and purified for reversibility testing.
8936
https://doi.org/10.1021/acs.joc.1c00855
J. Org. Chem. 2021, 86, 8933−8944

The Journal of Organic Chemistry
pubs.acs.org/joc
Article
hydrogen atom certainly play a role too. Conversely, for the
dienes, adding electron density to the structure seemed to
increase the reactivity. The carboxylic acid group deactivated
the complex, and only the most reactive dienophile B would
react with diene 1. While the unsubstituted anthracene and
methanol-substituted anthracene (dienes 2 and 3) each reacted
with three dienophiles (B, C, and E), methanol substitution on
the anthracene increased the reactivity in both the model
reactions and the quantum calculations. This substitution
effect was not the same as the addition of the methyl group on
N-phenylmaleimide though as substitution on the dienophile
decreased the reactivity and stability more than the addition of
the carboxylic acid to the anthracene group. One interesting
note was that anthracene substitution (dienes 1−3) did not
affect the complex stability with dienophile E but did greatly
affect the kinetics of dienophile E in the same manner as that
of the other dienophiles.
As anthracene is a highly reactive diene and would not react
with dienophiles A and E, finding other suitable dienes for
these dienophiles could be challenging. Thus, this limits
potential dienophiles to only very electron-deficient double
bonds. However, using the results from the quantum
calculations, one should be easily able to determine the
feasibility for new systems when provided.
For the investigation of the reversibility, the nine
combinations that were observed to react were synthesized
and isolated to obtain the pure Diels−Alder complexes. The
reactions to form the complexes were fairly straightforward,
with only compound V requiring the use of column
chromatography to acquire the pure compound. The other
compounds only required simple washing or recrystallization
to isolate the pure product. It was found during the
investigation that the ninth combination (IX) underwent a
further intramolecular reaction to form a dihydrobenzobarre-
lene and fused lactone ring structure (Figure 2). This reaction
is discussed and analyzed in a different publication and thus
was excluded from further investigation as it was no longer
possible to reform the starting diene and dienophile.32
The eight complexes that underwent further investigation
are shown in Figure 3. For the compounds V and VII, two
different possible isomers formed, which was observed to occur
in a 50:50 mixture between the isomers (see the Supporting
Information for NMR); however, for V, separation of the
isomers was possible via column chromatography to produce
gravimetric analysis (TGA), and experimental analysis were
performed. Unfortunately, the results from the DSC and TGA
were found to be of debatable value since degradation,
evaporation, and polymerization all contributed to the thermo-
analytical results, making a sound interpretation impracticable.
The experimental testing, however, provided a much better
medium and insight into the reactivity of the compounds.
Experimental testing consisted of heating all of the compounds
(I−VIII) in an increasing fashion (100, 150, 200 °C, etc.) up
to 250 °C in a flask with another highly reactive diene or
dienophile present in the flask, used as a trapping agent. After
heating, the mixtures were analyzed via NMR spectroscopy to
determine if an exchange reaction took place by looking to see
if a second Diels−Alder complex formation took place with the
trapping agent. If the degradation took place or if no exchange
reaction occurred by 250 °C, the reactions were considered
unsuccessful. The results for the experimental analysis are
shown in Table 2.
Table 2. Results from the Experimental Exchange Reaction
compounds
diene exchange
observed via
NMR upon
heating
I
II
III
IV
V
VI
VII
VIII
×
×
×
×
√
×
√
√
Of the eight starting compounds, three were observed to
successfully undergo the retro-reaction via the experimental
heating measurements. For compound IV, degradation was
found above 100 °C, and thus, it was heated no further.
Reversibility was found for compound V at 100 °C and at 220
°C for compounds VIII and VII. The remaining compounds
exhibited no degradation but showed no reversibility below
250 °C within the time frame or heating parameters used for
these experiments. The authors speculate that with further
heating or longer reaction times, VI might be found to be
reversible, but for the purposes of this article, only compounds
V, VII, and VIII were considered to be reversible. The reason
this is speculated is because based upon the quantum
calculations, the stability of VI was significantly reduced
(∼14 kJ/mol) compared to that of the non-methylated
analogue II. Furthermore, it was unexpected that VII would
be found reversible and not VI as VII was calculated to be
more stable (>10 kJ/mol). One possible explanation for this
effect might have been that only one of the isomers of VII is
reversible, while the other isomer is not in the tested
temperature range (Figure 4). Initially, with the heating
results, the reversibility of VII was unclear, so the compound
was heated with the trapping agent N-phenylmaleimide to
further investigate. After heating at 225 °C for 1 h, when the
1H NMR spectrum of VII was analyzed, it was found that the
original amount of the trans isomer was still present; however,
80% of the cis isomer had been exchanged (Figure 5). Whether
only the cis isomer is reacting or there is competing
isomerization could not be determined, though likely a
combination of both is applicable. Proof of the thermorever-
sibility of VIII can be found in the Supporting Information.
In order to better characterize the equilibrium temperature
of compounds V and VIII, further analysis was performed.
Each complex was heated with a trapping agent, and the
progress of the reactions was followed by NMR spectroscopy
Vexo and Vendo. The two isomers of VII are referred to as VIIcis
and VIItrans. The cis and trans indicate whether the methyl
group of the maleimide was added to the same side (cis) of the
complex as the methoxy group or to the opposite side (trans).
In this article, if the authors refer to VII, this is the 50:50
mixture of the isomers, otherwise the cis and trans versions will
be indicated. Additionally, it should be noted for the analysis
that compounds I−III based upon anthracene and N-
phenylmaleimide were not expected to be reversible, as has
been discussed in the literature; however, for the purposes of
the study, these were used as a standard to complete an overall
analysis.1,13 Furthermore, while the reversibility of the furan
and maleimide complex (V) is well-known, what was the most
interesting in this study was whether the addition of a methyl
group to the maleimide (VI and VII) or different alternative
dienes/dienophiles could produce a reversible species (IV and
VIII).
To assess the availability of the retro-reaction of the eight
compounds, differential scanning calorimetry (DSC), thermal
8937
https://doi.org/10.1021/acs.joc.1c00855
J. Org. Chem. 2021, 86, 8933−8944

The Journal of Organic Chemistry
pubs.acs.org/joc
Article
Figure 4. Reversibility of compound VII in thermal testing up to 225 °C.
by taking aliquots over time. An overview of the reactions can
be seen in Figure 6. In the case of V, since the isomers could be
separated, one pure isomer was chosen for analysis as it would
be easier to compare to the quantum calculations and because
the difference in activation energies for the retro-reaction of
the endo and exo isomers is minimal.
For each experiment, the reaction was conducted at three
different temperatures over a range of times to elicit the
kinetics of the exchange reaction. Since the N-phenyl-
maleimide and 9-anthracenemethanol/anthracene reactions
are significantly quicker than the reactions to reform the
compounds and are thermally irreversible, first-order kinetics
prevailed. Thus, the exchange of the reaction was plotted
versus time to give the reaction constant for each temperature,
shown in Table 3.
With the reaction constants determined for three different
temperatures, an Arrhenius plot was graphed to determine the
activation energy required for the opening of V and VIII
(Figure 7). The activation energy was determined for both
compounds by rearranging the Arrhenius equation (eq 1),
where k is equal to the reaction constant, A is the pre-
exponential factor, R is the gas constant, T is the temperature,
and Ea is the activation energy.
that the calculations used provide a good model for the
prediction of the reversibility of the Diels−Alder systems.
Since the trapping reaction of VII would likely affect the
isomerization, compound VII was heated both with and
without the presence of the N-phenylmaleimide trapping
agent. There was concern over whether isomerization is
happening in addition to the exchange reaction and how the
isomerization compared to the exchange reaction. The reaction
was done in the same fashion as that of the trapping reaction of
compound VIII, in the melt at three different temperatures.
These three temperatures were 215, 225, and 235 °C. The
results from the testing are shown in Table 5, which lists the
reaction constants found at each temperature with and without
the trapping agent. The rate was determined as the
disappearance of the cis peak. In order to determine the rate,
it was assumed that exclusively the N-phenylmaleimide
trapping reaction occurred before the reconversion of the
VIIcis compound could occur. This was deemed as a fair
assumption as the N-phenylmaleimide trapping reagent was in
large excess and has a much faster reaction rate than 2-methyl-
N-phenylmaleimide. However, it was more complicated in the
case of the reaction without the trapping agent. In this case, it
could not be determined whether the isomerization to the
VIItrans compound is occurring or the reactants from the retro-
reaction are evaporating out of the flask and leaving the
unreacted VIItrans compound behind. Therefore, the calcu-
lation was made using two different rates, one assuming that
the reactants evaporated out and the other assuming solely the
isomerization taking place.
Ea
RT
k = A eE /RT
ln k = ln(A) −
a
(1)
By multiplying the slope of the equation with the gas
constant R, the activation energy Ea was found for the opening
of the complexes, with values ∼104 kJ/mol for Vexo and ∼148
for VIII. As can be seen in Table 4, the experimental values
match well with the calculated values for the complexes. While
the experimental values are a little lower, they are still close to
the values predicted by the quantum calculations and confirm
While what was exactly happening in the reactions for the
measurement of kinetics could not be determined, all of the
different calculations gave similar values for the activation
energy of the reversal of the cis complex of VII (Table 6). The
8938
https://doi.org/10.1021/acs.joc.1c00855
J. Org. Chem. 2021, 86, 8933−8944

The Journal of Organic Chemistry
pubs.acs.org/joc
Article
Figure 5. 1H (500 MHz, DMSO-d6) NMR spectrum after 1 h in the melt at 225 °C with VII and N-phenylmaleimide used as the trapping agent.
The peak in yellow is from the newly formed complex with 9-anthracenemethanol and the trapping agent, and the blue peaks are from unreacted
VII. The doublet peak at 2.90 pm comes from the cis isomer and the singlet at 3.00 ppm from the trans isomer. See the Supporting Information for
the NMR of pure compounds III and VII.
calculated values were obtained in the same fashion as that for
V and VIII, and the Arrhenius plot is shown in Figure 8. All of
the plots showed great agreement, and the experimental values
were very close to the theoretical values from the quantum
calculations, mirroring the results of the experimental values
for VIII. Even though we cannot be sure what the actual
experimental value is for the activation energy, all of the values
are within 10% of each other, so a reasonable range can be
8939
https://doi.org/10.1021/acs.joc.1c00855
J. Org. Chem. 2021, 86, 8933−8944

The Journal of Organic Chemistry
pubs.acs.org/joc
Article
Figure 6. (A) Scheme of the exchange reaction conducted with 9-anthracenemethanol and compound V. (B) Scheme of the exchange reaction
conducted with N-phenylmaleimide and compound VIII.
Table 3. Reaction Constants Obtained from Kinetic
Investigation of Exchange Reaction of Compounds V and
VIII
Table 5. Comparison of Reactions Constants for VII with
and without a Trapping Agent
a
temperature reaction constant
reaction constant k (s−1)
R2
compound
(°C)
k (s−1
)
R2
compound
temperature (°C)
1.74 × 10−5
1.43 × 10−4
8.67 × 10−4
5.72 × 10−5
2.75 × 10−4
1.21 × 10−3
0.982
0.998
0.977
0.999
0.972
0.992
with the trapping agent
215
225
235
215
2.66 × 10−4
5.84 × 10−4
1.08 × 10−3
1.73 × 10−4
0.999
0.975
0.942
0.995
Vexo
65
85
105
190
210
230
w/o the trapping agent,
evaporation
VIII
225
235
215
4.03 × 10−4
7.52 × 10−4
1.15 × 10−4
0.997
0.991
0.991
w/o the trapping agent,
isomerization
225
235
2.35 × 10−4
4.52 × 10−4
0.996
0.996
a
Rate of the reaction was calculated for the heating without the
trapping agent by assuming both evaporation and isomerization.
Table 6. Comparison of Experimental and Quantum
Calculated Values for the Activation Energy of the Retro-
Diels−Alder Reaction of Compounds V and VIII
Ea for retro Diels−Alder reaction
(kJ/mol)
VII from reaction with the trapping agent
VII w/o the trapping agent assuming evaporation
VII w/o the trapping agent assuming isomerization
VIIcis theoretical value
∼145
∼152
∼141
∼158
Figure 7. Arrhenius plot of the diene exchange reaction of compound
V.
Table 4. Comparison of Experimental and Quantum
Calculated Values for the Activation Energy of the Retro-
Diels−Alder Reaction of Compounds V and VIII
that calculations can provide in predicting or fine-tuning
Diels−Alder complexes for thermoreversible systems. Fur-
thermore, similar to Diels−Alder formation, a maximum of
∼150−160 kJ/mol was seen in order for the Diels−Alder-retro
reaction to occur within a reasonable temperature window.
Ea for retro-Diels−Alder
experimental
(kJ/mol)
theoretical
(kJ/mol)
reaction
compound Vexo
compound VIII
∼104
∼148
∼106
∼161
CONCLUSIONS
assumed for these systems. It was also noted that similar to
VIII, the calculated experimental activation energy was higher
than the experimental results. Possible reasons for this could be
that since VII and VIII reactions were conducted in the melt,
the temperature of the solution could be less homogeneous
than that in the reaction of V due to the lack of the solvent and
stirring. Additionally, with the high temperatures used in VII
and VIII, some of the reactants and intermediates could have
been lost to evaporation, which would have been hard to see
and account for in the NMR spectra. Nevertheless, these
results once again confirm the validity of the quantum
calculations used in the beginning and show the usefulness
■
Model reactions of Diels−Alder pairs were successfully
conducted to provide insights into the factors which best
determine the validity of functional groups for thermorever-
sibility. By combining the experimental results with quantum
chemical calculations, which showed excellent agreement, this
provided a very powerful result. With the confirmation of the
accuracy, it allows us to make use of quantum calculations to
successfully pre-screen large amounts of systems without the
need to run actual experiments on each system or
combinations. While keeping in mind that polymer analogous
reactions are sometimes more difficult to achieve than their
8940
https://doi.org/10.1021/acs.joc.1c00855
J. Org. Chem. 2021, 86, 8933−8944

The Journal of Organic Chemistry
pubs.acs.org/joc
Article
Figure 8. Arrhenius plot of the diene exchange and reverse reactions of compound VII.
to give compounds I−VIII. The compounds were synthesized via two
different methods, with method 1 for compounds I−V based upon
the N-phenylmaleimide dienophile using chloroform and method 2
for compounds VI−VIII, which used p-xylene as the solvent.
Method 1 is as follows: the products were dissolved in 100 mL of
chloroform with the ratio depending on the diene and dienophile
being used. The reaction was refluxed in an oil bath for 18 h and then
allowed to cool to room temperature. The crude product was
obtained by concentrating the solvent and then stirring in 100 mL of
cyclohexane for 1 h. Products were obtained via vacuum filtration.
Compounds IV and V required one step more for purification.
Compound 5 was stirred for 1 h in 100 mL of isopropanol.
Compound IV was purified via flash chromatography with a column
packed with silica gel and using a mobile phase of ethyl acetate.
Method 2 is as follows: the products were dissolved in 100 mL of p-
xylene with the ratio depending on the diene and dienophile being
used. The reaction was refluxed in an oil bath for 36 h and then
allowed to cool to room temperature. p-Xylene was removed via heat
and vacuum filtration to give a crude residue. The crude product was
precipitated by stirring the residue in 100 mL of cyclohexane. The
products were collected via vacuum filtration and if necessary further
purified via recrystallization in methanol.
low molecular counterparts, these results will not only be
applicable within the small context of this study but will be able
to be used to predict the reaction mechanisms of new Diels−
Alder systems or to fine-tune systems already in place.
EXPERIMENTAL SECTION
■
Materials. Dimethyl fumarate (97%), p-xylene (>99%), anthra-
cene (98%), 9-anthracenemethanol (97%), sorbic acid (99%), 9-
anthracenecarboxylic acid (99%), N-phenylmaleimide (98%), aniline
(98%), cis-2-butene-diol (98%), citraconic anhydride (98%), and
tetrahydrophthalic anhydride (98%) were obtained from Sigma-
Aldrich and used as obtained. DMSO-d6, CDCl3, acetic anhydride,
chloroform, dichloromethane, and sodium acetate were also used as
obtained.
Synthesis of Reactants. Dienophiles A, B, and E along with
dienes 1, 2, 3, and 4 (Figure 1) were commercially available and
required no additional purification or synthesis. Diene 5 referred to in
this article as “BAMF-diacetate” [BAMF = 2,5-bis(aminomethyl)-
furane] was kindly provided by BASF for this research project.
Dienophile C and dienophile D were synthesized using a previously
reported procedure.33
N-Phenyl-10-hydro-9,10-ethano-anthracene-11,12-dicar-
boximide-9-carboxylic Acid I. Compound I synthesized from 9-
anthracenecarboxylic acid (1.09 g, 4.9 mmol) and N-phenylmaleimide
(8.40 g, 48.5 mmol) gave 0.55 g of a pure white powdery product
Dienophile C: 2-Methyl-N-phenylmaleimide. Dienophile C
was synthesized via a two-step reaction as follows, which was adapted
from a previously reported procedure.33 Citraconic anhydride (41.7 g,
370 mmol) was dissolved in diethyl ether (400 mL), and aniline (34.6
g, 370 mmol) in 35 mL of diethyl ether was added dropwise. The
solution was stirred for 1 h at room temperature, then cooled in an ice
bath, and filtered. A slightly off-white product was obtained with a
nearly quantitative yield (71.8 g, 95% yield), which was the
intermediate salt product. Next, the intermediate salt (16 g, 78
mmol) was placed in a flask with acetic anhydride (160 mL) and
sodium acetate (3.5 g, 43 mmol) and refluxed in an oil bath for 3 h.
The reaction was allowed to cool to room temperature and quenched
with water. Upon standing, the product precipitated out of the
solution to give a pure white powdery product (6 g, 41% yield).
Dienophile D: 2-Phenyl-3a,4,7,7a-tetrahydro-1H-isoindole-
1,3(2H)-dione. Dienophile D was synthesized in the exactly same
procedure as that of dienophile C. Tetrahydrophthalic anhydride was
used instead of citraconic anhydride for a yield of approximately 10%.
Model Reactions. Model reactions were run in deuterated
solvents with aliquots taken over time for investigation of each diene
and dienophile combination. The reactions were conducted in
DMSO-d6 and ran at temperatures ranges from 45 to 145 °C with
a 1:1 molar ratio of reactants to investigate complex formation.
Reactions were typically run between 4 and 18 h for the investigation
of Diels−Alder formation.
1
after stirring in cyclohexane for a yield of 28%. H NMR (DMSO-d6,
500 MHz): δ 14.00−13.80 (s, 1H), 8.05−8.00 (m, 1H), 7.63−7.58
(m, 1H), 7.24−7.20 (m, 9H), 6.44 (m, 2H), 4.89 (d, 1 H, J = 3.1 Hz),
4.00 (d, 1H, J = 8.6 Hz), 3.48 (dd, 1H, J = 8.6, 3.1 Hz). 13C{1H}
NMR (DMSO-d6, 125 MHz): δ 176.1, 175.5, 171.5, 141.9, 140.7,
139.4, 137.9, 132.2, 129.5, 127.5, 127.0, 125.9, 125.4, 124.9, 123.5,
56.3, 49.2, 47.8, 45.9. HRMS (ESI) m/z: [M + Na]+ calcd for
compound NaC25H17O4N, 418.1055; found, 418.10498.
N-Phenyl-9,10-dihydro-9,10-ethano-anthracene-11,12-di-
carboximide II. Compound II synthesized from anthracene (1.19 g,
6.7 mmol) and N-phenylmaleimide (2.62 g, 15.1 mmol) gave 2.00 g
of a pure white powdery product after stirring in cyclohexane for a
1
yield of 85%. H NMR (CDCl3, 500 MHz): δ 7.48−7.43 (m, 2H)
7.41−7.36 (m, 2H), 7.35−7.30 (m, 3H), 7.28−7.23 (m, 4H), 6.55−
6.50 (m, 2H), 4.93 (t, 2H, J = 1.6 Hz), 3.43 (dd, 2H, J = 2.0, 1.4 Hz).
13C{1H} NMR (CDCl3, 125 MHz): δ 176.1, 141.3, 138.7, 129.2,
128.8, 127.2, 126.8, 126.5, 125.2, 124.3, 47.1, 45.9. HRMS (ESI) m/z:
[M + Na]+ calcd for compound NaC24H17O2N, 374.1157; found,
374.11514.
N-Phenyl-9-methoxy-10-hydro-9,10-ethano-anthracene-
11,12-dicarboximide III. Compound III synthesized from 9-
anthracenemethanol (2.90 g, 13.9 mmol) and N-phenylmaleimide
(4.60 g, 25.8 mmol) gave 1.1 g of a pure white powdery product after
Synthesis of Diels−Alder Complexes. The complexes that were
found possible by the model reactions were synthesized and isolated
8941
https://doi.org/10.1021/acs.joc.1c00855
J. Org. Chem. 2021, 86, 8933−8944

The Journal of Organic Chemistry
pubs.acs.org/joc
Article
1
stirring in cyclohexane for a yield of 21%. H NMR (DMSO-d6, 500
MHz): δ 7.75−7.70 (m, 1H), 7.52−7.47 (m, 1H), 7.37−7.37 (m,
5H), 7.25−7.15 (m, 4H), 6.45−6.40 (m, 2H), 5.38 (t, 1H, J = 3.9
Hz), 4.87 (s, 2H), 4.80 (s, 1H), 3.43 (s, 2H). 13C{1H} NMR
(DMSO-d6, 125 MHz): δ 176.4, 175.6, 142.9, 140.5, 140.1, 132.4,
129.4, 128.9, 127.1, 126.9, 126.7, 126.4, 125.4, 125.0, 124.2, 123.0,
58.6, 49.9, 48.2, 46.2, 45.7. HRMS (ESI) m/z: [M + Na]+ calcd for
compound NaC25H19O3N, 404.1263; found, 404.12573.
(3.03 g, 17 mmol) and dimethyl fumarate (2.54 g, 17.6 mmol) gave
1.75 g of a pure white product after stirring in cyclohexane and
recrystallization from methanol for a yield of 32%. 1H NMR (CDCl3,
500 MHz): δ 7.40−7.35 (m, 2H), 7.30−7.25 (m, 2H), 7.20−7.10 (m,
4H), 4.75 (s, 2H), 3.65 (s, 6H), 3.45 (s, 2H). 13C{1H} NMR (CDCl3,
125 MHz): δ 172.8, 142.1, 140.3, 126.4, 124.6, 123.8, 52.2, 47.9, 46.7.
HRMS (CI) m/z [M + NH4]+ calcd for compound C20H22O4N,
340.1549; found, 340.1771.
Characterization and Thermal Analysis of Compounds.
NMR spectra were run in either DMSO-d6 or CDCl3, depending on
the solubility, on a Bruker DSX NMR-spectrometer operating at 500
1H-Isoindole-4-carboxylic Acid, 2,3,3a,4,7,7a-Hexahydro-7-
methyl-1,3-dioxo-2-phenyl-, Methyl Ester IV. Compound IV
synthesized from sorbic acid (2.32 g, 20.7 mmol) and N-phenyl-
maleimide (3.58 g, 20.1 mmol) gave 3.08 g of a pure white fine
powdery product after stirring in cyclohexane and then isopropanol
for a yield of 78%. 1H NMR (DMSO-d6, 500 MHz): δ 12.70 (s, 1H),
7.51−7.36 (m, 3H), 7.14 (m, 2H), 6.28 (dt, 1H, J = 9.3, 3.2 Hz), 5.80
(dt, 1H, J = 9.3, 3.2 Hz), 3.89 (dd, 1H, J = 8.8, 5.3 Hz), 3.35−3.30
(m, 2H), 2.59 (m, 1H), 1.31 (d, 3H, J = 7.5 Hz). 13C{1H} NMR
(DMSO-d6, 125 MHz): δ 177.1, 176.6, 172.5, 134.6, 132.7, 129.4,
128.8, 127.4, 44.3, 31.0, 17.0 HRMS (ESI) m/z: [M + Na]+ calcd for
compound NaC16H15O4N, 308.0899; found, 308.08942.
1
MHz for H NMR and at 125 MHz for 13C NMR. GC/HRMS was
run on a Thermo TSQ 700 coupled with gas chromatography with
either chemical ionization or electrospray ionization with the adduct
ion ammonium or sodium. DSC was run on a Seiko 6200 instrument
with a heating rate of 5 °C/min under a N2 atmosphere with the
temperature range 50−250 °C. TGA was run on a STA 409
instrument with a heating rate of 5 °C/min under a N2 atmosphere
with a temperature range of 50−300 °C. DSC data were collected on
the first run of the experiment.
Experimental Investigation of Reversibility. Each compound,
except compound V, was placed into a 10 mL glass vial with an
equimolar mixture of either unreacted anthracene or unreacted 9-
anthracenemethanol. 9-Anthracenemethanol was used in the case of
compounds II, VI, and VIII as anthracene was the starting diene for
these compounds; for the rest of the compounds, anthracene was
used. The mixtures were between 50 and 60 mg in total weight, and
each vial was slowly heated over 30 min to 225 °C. The mixture was
kept at this temperature for 15 min and then slowly allowed to cool
down to room temperature. The reaction mixture was dissolved in a
suitable deuterated solvent and analyzed via NMR spectroscopy.
Kinetics Investigation. For compound V, kinetics of the retro-
Diels−Alder reaction was measured by following the reaction of
aliquots in solution via 1H NMR in DMSO-d6. The procedure was as
follows: compound V (150 mg, 0.39 mmol) and 9-anthracenemetha-
nol (81 mg, 0.39 mmol) were placed in 4.5 mL of DMSO-d6 with
stirring. Aliquots were taken in time intervals, placed into an NMR
tube, and submitted for analysis. Three different temperatures (65, 85,
and 105 °C) were used for tracing the reaction process.
N,N′-((1,3-Dioxo-2-phenyl-2,3,3a,7a-tetrahydro-1H-4,7-ep-
oxyisoindole-4,7-diyl)bis(methylene))diacetamide V. Com-
pound V synthesized from BAMF-diacetate (1.89 g, 9.0 mmol) and
N-phenylmaleimide (16.75 g, 96.7 mmol) gave 1.7 g of a mixture of
two isomers in a 50:50 mixture for a yield of 49%. Purification from
flash column chromatography with ethyl acetate as the mobile phase
1
gave the pure isomers. H NMR (DMSO-d6, 500 MHz): isomer A
(exo): δ 7.96 (t, 2H, J = 5.8 Hz) 7.56−7.39 (m, 3H), 7.23 (m, 1H),
7.12 (m, 1H), 6.44 (s, 2H), 3.95−3.85 (dd, 2H, J = 14.6, 5.8 Hz),
3.60−3.51 (dd, 2H, J = 14.6, 6.0 Hz), 3.27, (s, 2H), 1.87 (d, 6H, J =
2.3 Hz), isomer B (endo): δ 8.12 (t, 2H, J = 5.8 Hz), 7.56−7.39 (m,
3H), 7.23 (m, 1H), 7.12 (m, 1H), 6.53 (s, 2H), 6.44 (s, 2H), 4.05−
3.95 (dd, 2H, J = 14.6, 6.8 Hz), 3.63 (s, 2H), 3.63−3.54 (dd, 2H, J =
14.6, 5.2 Hz), 1.87 (d, 6H, J = 2.3 Hz). 13C{1H} NMR (DMSO-d6,
125 MHz): δ 174.4, 170.1, 139.2, 137.1, 132.5, 129.5, 129.0, 127.4,
90.9, 51.6, 50.0, 38.4, 23.1. HRMS (ESI) m/z: [M + Na]+ calcd for
compound NaC20H21O5N3, 406.1379; found, 406.13739.
N-Phenyl-9,10-dihydro-9,10-ethano-anthracene-11-meth-
yl-12-hydro-dicarboximide VI. Compound VI synthesized from
anthracene (1.62 g, 9.0 mmol) and 2-methyl-N-phenylmaleimide
(1.70 g, 9.0 mmol) gave 0.9 g of a slightly off-white powdery product
after stirring in cyclohexane and recrystallization from methanol for a
yield of 28%. 1H NMR (DMSO-d6, 500 MHz): δ 7.57−7.48 (m, 2H),
7.35−7.20 (m, 9H), 6.48−6.43 (m, 2H), 4.82 (d, 1H, J = 3.5 Hz),
4.50 (s, 1H), 2.90 (d, 1H, J = 3.5 Hz), 1.19 (s, 3H). 13C{1H} NMR
(DMSO-d6, 125 MHz): δ 179.6, 175.6, 141.0, 140.6, 139.5, 132.3,
129.4, 129.0, 127.1, 126.6, 125.4, 124.9, 53.3, 51.5, 50.9, 46.1, 22.0.
HRMS (ESI) m/z: [M + Na]+ calcd for compound NaC25H19O2N,
388.1313; found, 388.13062.
N-Phenyl-9-methoxy-10-hydro-9,10-ethano-anthracene-
11-methyl-12-hydro-dicarboximide and N-Phenyl-9-methoxy-
10-hydro-9,10-ethano-anthracene-11-hydro-12-methyl-dicar-
boximide VII. Compound VII synthesized from 9-anthraceneme-
thanol (2.29 g, 11.0 mmol) and 2-methyl-N-phenylmaleimide (3.04 g,
16.3 mmol) gave 2.3 g of a slightly off-white powdery product with an
equal molar mixture of two isomers after stirring in cyclohexane for a
yield of 53%. 1H NMR (DMSO-d6, 500 MHz): isomer A (cis): δ 7.75
(m, 1H), 7.59 (m, 1H), 7.36−7.16 (m, 9H), 6.45 (m, 2H), 5.16 (t,
1H, J = 4.0 Hz), 4.84 (t, 1H, J = 3.9 Hz), 4.76 (d, 1H, J = 3.3 Hz),
4.70 (dd, 1H, J = 11.5, 3.8 Hz), 2.90 (d, 1H, J = 3.4 Hz), 1.11 (s, 3H),
isomer B (trans): δ 7.75 (m, 1H), 7.51 (m, 2H), 7.36−7.16 (m, 8H),
6.45 (m, 2H), 5.39 (t, 1H, J = 4.0 Hz), 5.05 (dd, 1H, J = 11.5, 4.2
Hz), 4.84 (t, 1H, J = 3.9 Hz), 4.44 (s, 1H), 2.99 (s, 1H), 1.17 (s, 3H).
13C{1H} NMR (DMSO-d6, 125 MHz): δ 179.6, 179.0, 175.4, 174.9,
For compound VIII, kinetics of the retro-Diels−Alder reaction was
1
measured by following the reaction via aliquots in H NMR. The
procedure was as follows: compound VIII (87 mg, 0.27 mmol) and
N-phenylmaleimide (180 mg, 1.04 mmol) were placed into a 4.5 mL
vial and heated. Aliquots were taken in time intervals, quenched in
DMSO-d6, placed into an NMR tube, and submitted for analysis.
Three different temperatures (190, 210, and 230 °C) were used for
tracing the reaction process.
Quantum Chemical Calculations. For all quantum chemical
calculations, the program package TURBOMOLE was used.30 All
structure optimizations were performed at the BP86/def-TZVP level
of theory using the solvation model COSMO and assuming a medium
with a dielectric constant of infinity. Refined single-point energy
calculations were carried out with the M06-2X/def2-TZVP method. A
solvation treatment was performed via the solvation model COSMO-
RS assuming a DMSO environment.31 Statistic thermodynamics was
applied involving standard approximations such as a rigid rotor or a
harmonic oscillator to compute thermodynamic functions at 80 and
150 °C.
ASSOCIATED CONTENT
■
sı
* Supporting Information
The Supporting Information is available free of charge at
https://pubs.acs.org/doi/10.1021/acs.joc.1c00855.
1H NMR, 13C NMR, and GC/HRMS data for
compounds I−VIII, confirmation of the thermal
reversibility of VIII, compound V exchange reaction
kinetics, compound VIII exchange reaction kinetics, and
computational results calculation information (PDF)
141.0, 132.4, 129.5, 129.0, 127.2, 126.7, 126.1, 125.2, 124.4, 123.0,
58.5, 57.1, 56.8, 53.2, 52.5, 51.6, 50.6, 46.0, 21.9, 19.7. HRMS (ESI)
m/z: [M + Na]+ calcd for compound NaC26H21O3N, 418.1419;
found, 418.14127.
Dimethyl 9,10-Dihydro-9,10-ethanoanthracene-11,12-di-
carboxylate VIII. Compound VIII synthesized from anthracene
8942
https://doi.org/10.1021/acs.joc.1c00855
J. Org. Chem. 2021, 86, 8933−8944

The Journal of Organic Chemistry
pubs.acs.org/joc
Article
(10) Zhong, Y.; Wang, X.; Zheng, Z.; Du, P. Polyether-maleimide-
based crosslinked self-healing polyurethane with Diels-Alder bonds. J.
Appl. Polym. Sci. 2015, 132, 41944.
AUTHOR INFORMATION
Corresponding Author
Jarrett R. Rowlett − Joint Research Network on Advanced
Materials and Systems (JONAS), Freiburg Materials
Research Center, University of Freiburg, D-79104 Freiburg,
Germany; orcid.org/0000-0003-2775-696X;
Email: Jarrett.rowlett@gmail.com
■
(11) Li, J.; Zhang, G.; Deng, L.; Zhao, S.; Gao, Y.; Jiang, K.; Sun, R.;
Wong, C. In situ polymerization of mechanically reinforced, thermally
healable graphene oxide/polyurethane composites based on Diels−
Alder chemistry. J. Mater. Chem. A 2014, 2, 20642−20649.
(12) Pratama, P. A.; Sharifi, M.; Peterson, A. M.; Palmese, G. R.
Room Temperature Self-Healing Thermoset Based on the Diels−
Alder Reaction. ACS Appl. Mater. Interfaces 2013, 5, 12425−12431.
(13) Tasdelen, M. A. Diels−Alder “click” reactions: recent
applications in polymer and material science. Polym. Chem. 2011, 2,
2133.
(14) Sanyal, A. Diels-Alder Cycloaddition-Cycloreversion: A Power-
ful Combo in Materials Design. Macromol. Chem. Phys. 2010, 211,
1417−1425.
(15) Min, Y.; Huang, S.; Wang, Y.; Zhang, Z.; Du, B.; Zhang, X.;
Fan, Z. Sonochemical Transformation of Epoxy−Amine Thermoset
into Soluble and Reusable Polymers. Macromolecules 2015, 48, 316−
322.
Authors
Peter Deglmann − Advanced Materials and Systems Research,
BASF SE, D-67056 Ludwigshafen, Germany
Johannes Sprafke − Advanced Materials and Systems
Research, BASF SE, D-67056 Ludwigshafen, Germany
Nabarun Roy − BASF Polyurethanes GmbH, D-49448
Lemfoerde, Germany
Rolf Mulhaupt − Institute for Macromolecular Chemistry,
̈
University of Freiburg, D-79104 Freiburg, Germany
Bernd Bruchmann − Joint Research Network on Advanced
Materials and Systems (JONAS), Freiburg Materials
Research Center, University of Freiburg, D-79104 Freiburg,
Germany; Advanced Materials and Systems Research, BASF
SE, D-67056 Ludwigshafen, Germany
(16) Rabjohns, M. A.; Hodge, P.; Lovell, P. A. Synthesis of aromatic
polyamides containing anthracene units via a precursor polymer
approach. Polymer 1997, 38, 3395−3407.
(17) Yoshie, N.; Saito, S.; Oya, N. A thermally-stable self-mending
polymer networked by Diels−Alder cycloaddition. Polymer 2011, 52,
6074−6079.
Complete contact information is available at:
https://pubs.acs.org/10.1021/acs.joc.1c00855
(18) Wang, J.; Piskun, I.; Craig, S. L. Mechanochemical
Strengthening of a Multi-mechanophore Benzocyclobutene Polymer.
ACS Macro Lett. 2015, 4, 834−837.
Notes
(19) Salamone, J. C.; Chung, Y.; Clough, S. B.; Watterson, A. C.
Thermally reversible, covalently crosslinked polyphosphazenes. J.
Polym. Sci., Part A: Polym. Chem. 1988, 26, 2923−2939.
(20) Peterson, A. M.; Jensen, R. E.; Palmese, G. R. Reversibly cross-
linked polymer gels as healing agents for epoxy-amine thermosets.
ACS Appl. Mater. Interfaces 2009, 1, 992−995.
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We kindly acknowledge BASF SE and especially the JONAS
program for the funding of this research.
(21) Bergman, S. D.; Wudl, F. Mendable polymers. J. Mater. Chem.
2008, 18, 41−62.
REFERENCES
■
(22) Voit, B. The potential of cycloaddition reactions in the
synthesis of dendritic polymers. New J. Chem. 2007, 31, 1139−1151.
(23) Liu, Y.-L.; Chuo, T.-W.; Wool, R. P.; Bergman, S. D.; Wudl, F.;
Syrett, J. A.; Becer, R.; Haddleton, D. M.; Burattini, S.; Greenland, B.
W.; et al. Self-healing polymers based on thermally reversible Diels−
Alder chemistry. Polym. Chem. 2013, 4, 2194.
(1) Kriegel, R. M.; Saliba, K. L.; Jones, G.; Schiraldi, D. A.; Collard,
D. M. The Rapid Chain Extension of Anthracene-Functional
Polyesters by the Diels-Alder Reaction with Bismaleimides. Macromol.
Chem. Phys. 2005, 206, 1479−1487.
(2) Yoshie, N.; Watanabe, M.; Araki, H.; Ishida, K. Thermo-
responsive mending of polymers crosslinked by thermally reversible
covalent bond: Polymers from bisfuranic terminated poly(ethylene
adipate) and tris-maleimide. Polym. Degrad. Stab. 2010, 95, 826−829.
(3) Ruckenstein, E.; Chen, X. Crosslinking of chlorine-containing
polymers by dicyclopentadiene dicarboxylic salts. J. Polym. Sci., Part A:
Polym. Chem. 2000, 38, 818−825.
(4) Brand, T.; Klapper, M. Control of viscosity through reversible
addition of telechelics via repetitive Diels-Alder reaction in bulk. Des.
Monomers Polym. 1999, 2, 287−309.
(5) Ilhan, F.; Rotello, V. M. Thermoreversible Polymerization.
Formation of Fullerene-Diene Oligomers and Copolymers. J. Org.
Chem. 1999, 64, 1455−1458.
(6) Goussé, C.; Gandini, A.; Hodge, P. Application of the Diels−
Alder Reaction to Polymers Bearing Furan Moieties. 2. Diels−Alder
and Retro-Diels−Alder Reactions Involving Furan Rings in Some
Styrene Copolymers. Macromolecules 1998, 31, 314−321.
(7) Nie, B.; Rotello, V. M. Thermally Controlled Formation of
Fullerene−Diene Oligomers and Copolymers. Macromolecules 1997,
30, 3949−3951.
(8) Zhang, G.; Zhao, Q.; Yang, L.; Zou, W.; Xi, X.; Xie, T. Exploring
Dynamic Equilibrium of Diels−Alder Reaction for Solid State
Plasticity in Remoldable Shape Memory Polymer Network. ACS
Macro Lett. 2016, 5, 805−808.
(24) Murphy, E. B.; Bolanos, E.; Schaffner-Hamann, C.; Wudl, F.;
Nutt, S. R.; Auad, M. L. Synthesis and Characterization of a Single-
Component Thermally Remendable Polymer Network: Staudinger
and Stille Revisited. Macromolecules 2008, 41, 5203−5209.
(25) Wiggins, K. M.; Syrett, J. A.; Haddleton, D. M.; Bielawski, C.
W. Mechanically facilitated retro [4+2] cycloadditions. J. Am. Chem.
Soc. 2011, 133, 7180−7189.
(26) Wiggins, K. M.; Brantley, J. N.; Bielawski, C. W. Methods for
activating and characterizing mechanically responsive polymers. Chem.
Soc. Rev. 2013, 42, 7130−7147.
(27) Syrett, J. A.; Mantovani, G.; Barton, W. R. S.; Price, D.;
Haddleton, D. M. Self-healing polymers prepared via living radical
polymerisation. Polym. Chem. 2010, 1, 102.
(28) Ramirez, A. L. B.; Kean, Z. S.; Orlicki, J. A.; Champhekar, M.;
Elsakr, S. M.; Krause, W. E.; Craig, S. L. Mechanochemical
strengthening of a synthetic polymer in response to typically
destructive shear forces. Nat. Chem. 2013, 5, 757−761.
(29) Black, A. L.; Lenhardt, J. M.; Craig, S. L. From molecular
mechanochemistry to stress-responsive materials. J. Mater. Chem.
2011, 21, 1655.
(30) Ahlrichs, R.; Bär, M.; Häser, M.; Horn, H.; Kölmel, C.
Electronic structure calculations on workstation computers: The
program system turbomole. Chem. Phys. Lett. 1989, 162, 165−169.
(9) Kuang, X.; Liu, G.; Dong, X.; Liu, X.; Xu, J.; Wang, D. Facile
fabrication of fast recyclable and multiple self-healing epoxy materials
through diels-alder adduct cross-linker. J. Polym. Sci., Part A: Polym.
Chem. 2015, 53, 2094−2103.
(31) Deglmann, P.; Muller, I.; Becker, F.; Schäfer, A.; Hungenberg,
̈
K.-D.; Weiß, H. Prediction of Propagation Rate Coefficients in Free
Radical Solution Polymerization Based on Accurate Quantum
8943
https://doi.org/10.1021/acs.joc.1c00855
J. Org. Chem. 2021, 86, 8933−8944

The Journal of Organic Chemistry
pubs.acs.org/joc
Article
Chemical Methods: Vinylic and Related Monomers, Including
Acrylates and Acrylic Acid. Macromol. React. Eng. 2009, 3, 496−515.
(32) Rowlett, J. R.; Plenk, C.; Deglmann, P.; Sprafke, J.; Mulhaupt,
̈
R.; Bruchmann, B. Fused Dihydrodibenzobarrelene
(Dibenzobicylco[2.2.2]octadiene) and Lactone Rings via Tandem
Diels-Alder and Condensation Reactions of Dialkyl Fumarates and 9-
Anthracenemethanol. ChemistrySelect 2016, 1, 4935−4939.
(33) Matuszak, N.; Muccioli, G. G.; Labar, G.; Lambert, D. M.
Synthesis and in Vitro Evaluation of N-Substituted Maleimide
Derivatives as Selective Monoglyceride Lipase Inhibitors. J. Med.
Chem. 2009, 52, 7410−7420.
8944
https://doi.org/10.1021/acs.joc.1c00855
J. Org. Chem. 2021, 86, 8933−8944