Full atom-efficiency transformation of wasted polycarbonates into epoxy thermosets and the catalyst-free degradation of the thermosets for environmental sustainability

Article abstract of DOI:10.1039/d0gc01318h

We report two sustainable features in this work. The first feature is that we successfully applied wasted polycarbonate (WPC) as an epoxy curing agent for epoxy resins to prepare WPC-cured epoxy thermosets. The curing of WPC and epoxy is based on the reaction of a carbonate and an epoxide, which is proven by a model reaction of diphenyl carbonate (DPC) and glycidyl phenyl ether (GPE) in the presence of a catalytic amount of 4-dimethylaminopyridine (DMAP). The WPC-cured epoxy thermosets show comparable thermal properties to commercial-based, phenol novolac (PN)-cured epoxy thermosets. Interestingly, the films of WPC-cured epoxy thermosets are foldable, while those of PN-cured epoxy thermosets are brittle, demonstrating the advantage of using WPC as a curing agent over PN. Since the WPC was used directly without any digestion or pyrolysis process, the atomic efficiency is 100percent, making this WPC recycling strategy economically attractive. The second feature is that we successfully degraded the WPC-cured epoxy thermosets to phenoxy resins through a catalyst-free aminolysis process, that is, the products based on the WPC-cured epoxy thermosets are degradable when their lifespans are expired. The transformation of WPC into epoxy thermosets, along with the degradation of the epoxy thermosets, makes this work attractive for sustainability.

Full text of DOI:10.1039/d0gc01318h

Green Chemistry
View Article Online
View Journal
PAPER
Full atom-efficiency transformation of wasted
polycarbonates into epoxy thermosets and the
catalyst-free degradation of the thermosets for
environmental sustainability†
Cite this: DOI: 10.1039/d0gc01318h
Shih Wei Shao,a Chien Han Chen,b Jian Ren Chan,a Tzong Yuan Juang,c
a,b
Mahdi M. Abu-Omar d and Ching Hsuan Lin
*
We report two sustainable features in this work. The first feature is that we successfully applied wasted poly-
carbonate (WPC) as an epoxy curing agent for epoxy resins to prepare WPC-cured epoxy thermosets. The
curing of WPC and epoxy is based on the reaction of a carbonate and an epoxide, which is proven by a
model reaction of diphenyl carbonate (DPC) and glycidyl phenyl ether (GPE) in the presence of a catalytic
amount of 4-dimethylaminopyridine (DMAP). The WPC-cured epoxy thermosets show comparable thermal
properties to commercial-based, phenol novolac (PN)-cured epoxy thermosets. Interestingly, the films of
WPC-cured epoxy thermosets are foldable, while those of PN-cured epoxy thermosets are brittle, demon-
strating the advantage of using WPC as a curing agent over PN. Since the WPC was used directly without any
digestion or pyrolysis process, the atomic efficiency is 100%, making this WPC recycling strategy economi-
cally attractive. The second feature is that we successfully degraded the WPC-cured epoxy thermosets to
phenoxy resins through a catalyst-free aminolysis process, that is, the products based on the WPC-cured
epoxy thermosets are degradable when their lifespans are expired. The transformation of WPC into epoxy
thermosets, along with the degradation of the epoxy thermosets, makes this work attractive for sustainability.
Received 15th April 2020,
Accepted 9th June 2020
DOI: 10.1039/d0gc01318h
rsc.li/greenchem
2026.10 Therefore, it is essential to recycle or reuse wasted poly-
carbonate (WPC) for environmental sustainability.
Introduction
Recycling of waste products is key to preventing environmental
At present, three approaches have been applied to recycling
pollution.1–9 Bisphenol A-based polycarbonate (PC), a polymer WPC. The first one is physical methods such as regrinding
produced from bisphenol A and phosgene through an inter- mechanically, followed by re-extrusion. Physical recycling leads
facial reaction or with diphenyl carbonate (DPC) through an to a detrimental influence on the performances of PC, such as
ester exchange reaction, is widely used in engineering due to its transparency and mechanical properties.11 The second one is
heat resistance, impact resistance (toughness) and optical trans- chemical methods such as alcoholysis and aminolysis to yield
parency. Data storage discs (CDs and DVDs), safety glasses, useful monomers. The monomers include bisphenol A,12–14
sheeting in a greenhouse, and some automotive components phenol,15 bis(hydroxyethyl)ether of BPA,12,16 dimethyl carbo-
are made of PC. Five million tons of PC have been produced per nate,17–19 diethyl carbonate,20 dialkyl carbonate,21 hydroxyl-N,
year. The market of global PC sheets was valued at US$ 1.62 N′-diphenylene-isopropylidenyl biscarbamates,22 carbonylation
billion in 2017 and is expected to reach US$ 2.32 billion by agents,23 and N,N′-dibenzylurea.23 Recently, the chemical re-
cycling of WPC has been reviewed by Kim.24 Generally, chemi-
cal recycling requires some separation processes and is on the
aDepartment of Chemical Engineering, National Chung Hsing University, Taichung,
research level, not on a practical level. The third one is
Taiwan. E-mail: linch@nchu.edu.tw
bAdvanced Research Center for Green Materials Science and Technology, National
thermal pyrolysis.25–27 Although thermal pyrolysis is an envir-
onmentally friendly solventless process, the low selectivity
Taiwan University, Taipei 10617, Taiwan
cDepartment of Cosmeceutics, China Medical University, Taichung, Taiwan
dDepartment of Chemistry and Biochemistry, University of California, Santa Barbara,
California, USA
†Electronic supplementary information (ESI) available: FTIR and 1H-NMR
spectra of PC, GPC curves of PC, 1H-NMR spectra of the model reactions, the
relation of 1/Tg and epoxy equivalent weight (EEW), the simplified curing chem-
istry of WPC and epoxy I. See DOI: 10.1039/d0gc01318h
hinders its development. Therefore, it would be attractive if
one could chemically recycle or reuse WPC without any diges-
tion and pyrolysis process.
PC has been used to toughen epoxy resins since the 1990s.
Hourston et al. blended 10 wt% PC into Epikote 828 and di-
ethylenetriamine.28 The results show that the K1c and G1c
This journal is © The Royal Society of Chemistry 2020
Green Chem.

View Article Online
Paper
Green Chemistry
Scheme 1 The reaction of (a) phenyl acetate with glycidyl phenyl ether (GPE)36 and (b) diphenyl carbonate (DPC) with GPE in the presence of a
catalytic amount of DMAP (this work).
increase somewhat on the addition of PC, but no reaction was dynamic exchange reactions,4,37–40 known as vitrimers, have
discussed in that work. Abbate et al. dissolved PC in an epoxy been developed. The epoxy thermoset prepared using the
resin and observed a single-phase homogeneous blend.29,30 epoxy-anhydride curing system is one of the most representa-
The molecular interaction among the carbonyl of PC and the tive vitrimers.41–44 The dynamic transesterification between
epoxide was evidenced by IR spectroscopy. Woo et al. reported the hydroxyl and aliphatic ester enables the repairing and
a single-phase morphology and a composition-dependent Tg reprocessing of the thermoset. Due to this sustainability issue,
in a thermoset made of PC, tetraglycidyl-4,4′-diaminodiphenyl the development of recyclable or degradable thermosets is
methane (TGDDM), and 4,4′-diaminodiphenyl sulfone needed. Ma et al. reviewed the methods of recyclable carbon
(DDS).31 They reported that transesterification and alcoholysis fiber reinforced composites (CFRPs) based on degradable ther-
were the main reactions in the TGDDM/PC blends upon mosets.45 Degradable thermosets with ester bonds,37,38,40,46–48
heating.32 Ma and Chang et al. studied the blend of the diglyci- acetal linkages,49 Schiff base bonds,44,50 disulfide bonds,51
a
dyl ether of bisphenol A (DGEBA) with 20 wt% PC in the pres- hexahydrotriazine structure,52 boronic ester linkages,53 and di-
ence of a tertiary amine.33,34 They observed that the carbonyl N-benzylaniline linkages54 have been reviewed.45
absorption shifted to a lower frequency initially and an
The WPC-cured thermosets prepared in this work do not
unusual IR absorption peak appeared at 1804 cm−1. They con- contain the dynamic exchangeable bonds in the network, so the
cluded that the transesterification reaction between the epoxide network is very stable at elevated temperatures. The network is
and carbonate explained the initial upshift of carbonyl absorp- rigid, so we expect that the degradation is not as easy as vitri-
tion,35 and the formation of 4-(phenoxymethyl)-1,3-dioxolan-2- mers with a covalent adaptable network. However, as shown in
one explained the IR absorption peak at 1804 cm−1
.
Scheme 1b, the reaction leads to the formation of aliphatic–ali-
In previous work, we have successfully analyzed the reac- phatic carbonate linkages. Therefore, the chains in the WPC-
tion of phenyl acetate and glycidyl phenyl ether (GPE) cured epoxy network are expected to be connected by aliphatic–
in the presence of
a catalytic amount of 4-dimethyl- aliphatic carbonate linkages. These unique connecting linkages
aminopyridine (DMAP).36 Through 1H NMR analyses, the have never been seen in traditional epoxy thermosets.
reaction occurred through the reaction of the aromatic ester Therefore, WPC-cured thermosets might be degradable by trans-
with the epoxide and produced 1,3-diphenoxy-2-acetoxypro- esterification or aminolysis of the aliphatic–aliphatic carbon-
pane exclusively (Scheme 1a). Due to the structural simi- ates. To the best of our knowledge, preparing carbonate-con-
larity between an aromatic ester and an aromatic–aromatic nected epoxy thermosets and degrading them by transesterifica-
carbonate, we speculated that an aromatic–aromatic carbon- tion or aminolysis has not been reported yet. The strategy for
ate might also react with epoxides. In this work, we report transforming WPC into epoxy thermosets and degrading the
the result of the model reaction of diphenyl carbonate WPC-cured thermosets is reported in this work.
(DPC) and GPE in the presence of a catalytic amount
of DMAP. The analyses show that the reaction product is
bis(1,3-diphenoxypropan-2-yl)carbonate (BDPC) exclusively
(Scheme 1b, will be discussed later). The reaction trans-
formed the carbonate linkage from an aromatic–aromatic
Experimental
Materials
one to an aliphatic–aliphatic one. By utilizing this chem-
istry, we directly used a WPC coming from a base layer of a
compact disc (CD) as a curing agent for epoxy resins.
Through this approach, we successfully converted WPC into
epoxy thermosets. No digestion or pyrolysis process was
applied, making this approach economically attractive. To
the best of our knowledge, the use of a WPC as a curing
agent for epoxy resins has not been reported yet.
The source of waste polycarbonate (WPC) came from a compact
disc (CD). The fresh PC was purchased from Aldrich. The
coating of the CD was removed by (a) soaking in 5 wt%
NaOH(aq) or (b) soaking in 5 wt% H2SO4(aq) at room temperature
for 6 h then washing three times with DI water or (c) simply
grinding using sandpaper. IR spectra (Fig. S1†) show that the
base-soaked CD, acid-soaked CD, sandpaper-ground CD, and
fresh PC have a plate form. All of the samples are measured in
an ATR mode and have the same IR carbonyl absorption at
1769 cm−1. NMR spectra (Fig. S2†) show that the base-soaked
Conventional thermosets are non-recyclable due to their
highly crosslinked structure. Recently, thermosets with
Green Chem.
This journal is © The Royal Society of Chemistry 2020

View Article Online
Green Chemistry
Paper
CD and fresh PC have the same NMR patterns, indicating that for 2 h, and 160 °C for 2 h. According to the relative height of
the coating-removal process does not change the structure of aromatic signals between GPE and DPC, we observed that a
PC. The GPC curves of base-soaked CD, acid-soaked CD, sand- small amount of GPE decreased gradually with reaction time.
paper-ground CD, and fresh PC are shown in Fig. S3.† The base- The decrease in GPE is due to the evaporation of GPE at high
soaked CD has a weight-average molecular weight of 19 500 g temperatures. However, the same signal positions suggest
mol−1, which is slightly smaller than that (21 000 g mol−1) of that the DPC does not react with GPE without DMAP.
the sandpaper-ground CD, probably due to a small degree of
DMAP-catalyzed reaction of DPC and GPE
alkaline hydrolysis. However, the acid-soaked CD has a relatively
higher weight-average molecular weight of 41 000 g mol−1 due
2.0 g of DPC (9.3 mmol), 2.8 g of GPE (18.6 mmol), and
to an unknown reason. For simplicity, only the base-soaked CD
was selected as a source of WPC for preparing epoxy thermosets
due to its similar weight-average molecular weight to those of
the sandpaper-ground CD and fresh PC. Diphenyl carbonate
(DPC, from Sigma-Aldrich), glycidyl phenyl ether (GPE, from
TCI), and 4-dimethylaminopyridine (DMAP, from Alfa) were
used as received. PN with an active-hydrogen equivalent weight
of 105 g eq−1 was kindly supplied by Chang Chun Plastics.
Bisphenol A-based epoxy (I–III, with an epoxy equivalent weight
(EEW) of 188, 482, and 901 g eq−1, respectively) were kindly sup-
plied by Chang Chun Plastics with the commercial names of
BE188, BE501, and BE504, respectively. Dicyclopentadiene
phenol-based epoxy (IV), with a commercial name of HP-7200
0.014 g of DMAP (0.5 wt% based on GPE) were introduced into
a flask equipped with a nitrogen inlet, a stirrer, and a thermo-
couple. The reaction mass was stirred at 120 °C for 24 h to
obtain a viscous liquid. During this, the reaction mass was
sampled for 1H NMR analysis (Fig. S5†). The reaction mass
that was stirred at 120 °C for 2 h was measured by 1H–1H
COSY NMR and H–13C HSQC NMR (Fig. 1). H NMR (DMSO-
d6), δ = 4.32 (8H, H5), 5.38 (2H, H6), 6.97 (8H, H1 and H3), 7.28
(4H, H2). 13C NMR (DMSO-d6), δ = 65.9 (C5), 74.6 (C6), 114.4
(C3), 120.9 (C1), 129.4 (C2), 153.7 (C4). FTIR (KBr): ν (cm−1) =
1750 (CvO stretching of the carbonyl group). High-resolution
1
1
LC-MS (ESI-MS) m/z: [M+] calcd for C31H31O7 515.20 g mol−1
anal., 515.2050 g mol−1
;
.
and an epoxy equivalent weight (EEW) of 260 g eq−1, was kindly Reaction of DPC and DMAP
supplied by Dainippon Ink, Japan. The structures of epoxy
resins and PN are shown in Scheme 2.
2.0 g of DPC and 0.014 g of DMAP were introduced into a flask
equipped with a nitrogen inlet, a stirrer, and a thermocouple.
The reaction mass was stirred at 120 °C for 2 h. After this, the
1
Uncatalyzed reaction of DPC and GPE
reaction mass was measured by H NMR (Fig. S6†). The same
spectra before and after stirring at 120 °C for 2 h clearly show
that DPC and DMAP do not react with each other.
2.0 g of DPC (9.3 mmol) and 2.8 g of GPE (18.6 mmol) were
introduced into a flask equipped with a nitrogen inlet, a stirrer,
and a thermocouple. The reaction mass was stirred at 120 °C for
2 h, 140 °C for 2 h, and 160 °C for 2 h. During this, the reac-
Preparation of the uncatalyzed WPC/epoxy blend
tion mass was sampled directly for NMR analysis. Fig. S4† In the preparation of the WPC/epoxy blend, the molar ratio of
shows the 1H NMR spectra of the reaction product of DPC the carbonate moiety to the epoxide moiety was controlled at
and GPE at different stages, including 120 °C for 2 h, 140 °C 1/2. Using the blend of WPC and epoxy IV as an example: 2.5 g
Scheme 2 Structures of epoxy resins (I–IV) and PN.
This journal is © The Royal Society of Chemistry 2020
Green Chem.

View Article Online
Paper
Green Chemistry
Fig. 1 (a) 1H–1H COSY and (b) 1H–13C HSQC NMR spectra of the DMAP-catalyzed reaction of GPE and DPC after reacting at 120 °C for 2 h.
of WPC (carbonate 9.9 × 10−3 mol) and 5.158 g of IV (epoxide property comparison. Using the curing system of PN and epoxy
1.98 × 10−2 mol) were introduced into a flask equipped with a IV as an example: 0.48 g of PN (phenolic hydroxyl 4.6 × 10−3
nitrogen inlet, a stirrer, and a thermocouple. The reaction mol), 1.20 g of IV (epoxide 4.6 × 10−3 mol), and 0.0036 g of
mass was stirred at 200 °C for 1 h to obtain a homogeneous 2-MI (0.3 wt% of IV) were stirred at 140 °C to obtain a homo-
viscous liquid. Fig. S7† shows the 1H NMR spectra of WPC, geneous mixture, and then cured at 160 °C (1 h), 180 °C (2 h),
epoxy I, and the WPC/I blend (without DMAP) that was stirred and 200 °C (2 h) under a nitrogen atmosphere. The PN/epoxy
at 200 °C for 1 h. The pattern in Fig. S7(c)† is the combination thermoset is named PN/I, PN/II, PN/III, and PN/IV if the epoxy
of those in Fig. S7(a) and S7(b),† clearly indicating that WPC resin is I, II, III, and IV, respectively.
do not react with epoxy I without DMAP. Fig. S8† shows the 1H
Degradation of DPC by 1-hexylamine
NMR spectra of WPC, epoxy IV, and the WPC/IV blend
1.0 g of DPC (4.67 mmol) and 0.9447 g of 1-hexylamine
(9.34 mmol) were introduced into a flask equipped with a
nitrogen inlet, a stirrer, and a thermocouple. The solution was
stirred at 100 °C for 4 h. After this, the solution was distilled
under vacuum at 125 °C to remove 1-hexylamine. The remain-
ing solution was then analyzed by 1H NMR. The result is
shown in Fig. 11.
(without DMAP) that was stirred at 200 °C for 1 h. The pattern
in Fig. S8(c)† is the combination of those in Fig. S8(a) and
S8(b),† also clearly indicating that the WPC does not react with
epoxy IV without DMAP. These results in Fig. S7 and S8† are
consistent with the result in Fig. S4,† in which DPC and GPE
do not react with each other without DMAP. We took advan-
tage of the stability of WPC and epoxy resin at 200 °C, and dis-
solved WPC into epoxy resin at 200 °C for 1 h in a solvent-free
process to obtain a viscous WPC/epoxy blend.
Degradation of BDPC by 1-hexylamine
1.0 g of BDPC (1.94 mmol) and 0.3933 g of 1-hexylamine
(3.88 mmol) were introduced into a flask equipped with a
nitrogen inlet, a stirrer, and a thermocouple. The solution was
stirred at 125 °C for 6 h. After this, the solution was distilled
under vacuum at 125 °C to remove 1-hexylamine. The remain-
ing solution was then analyzed by 1H NMR. The result is
shown in Fig. 12.
Preparation of DMAP-catalyzed WPC/epoxy thermosets in a
solvent-free process
To a WPC/epoxy blend, 0.1 wt% of DMAP was added and
stirred at 140 °C to obtain a homogeneous mixture, and then
cured at 160 °C (1 h), 180 °C (2 h), and 200 °C (2 h) under a
nitrogen atmosphere to obtain a PC/epoxy thermoset. The
WPC/epoxy thermoset is named WPC/I, WPC/II, WPC/III, and
WPC/IV if the epoxy resin is I, II, III, and IV, respectively.
Degradation of WPC/I by 1-hexylamine
WPC/I (0.1715 g) and 1-hexylamine (5 mL) were introduced
into a flask equipped with a nitrogen inlet, a stirrer, and a
thermocouple. The mixture was stirred at 125 °C until WPC/I
was fully degraded (about 1.75 h). After this, the solution was
distilled under vacuum at 125 °C to remove 1-hexylamine. The
Preparation of DMAP-catalyzed WPC/epoxy thermosetting
films via solution casting
We used a solution casting method to make a thermosetting
film for the measurement of tensile properties. The prepa-
ration procedure is the same as that in the solvent-free process
except that N-methyl-2-pyrrolidone (NMP) was added to make
a 40 wt% solution. The solution was cast on glass and dried
following 80 °C for 12 h, and then cured at the same program
as that in the solvent-free process.
1
remaining solution was then analyzed by H NMR. The result
is shown in Fig. 13.
Characterization
1H NMR, 13C NMR, 1H–1H COSY NMR, and 1H–13C HSQC
NMR analyses were performed using a Varian Inova 600 NMR
in DMSO-d6. Thermal gravimetric analysis (TGA) was per-
Curing procedure for PN/epoxy thermosets
Phenol novolac (PN) is a commercial curing agent for epoxy formed with a PerkinElmer Pyris 1 at a heating rate of 20 °C
resins. The PN-cured epoxy thermosets were also prepared for min−1 in a nitrogen or air atmosphere. Dynamic mechanical
Green Chem.
This journal is © The Royal Society of Chemistry 2020

View Article Online
Green Chemistry
Paper
analysis (DMA) was performed using a PerkinElmer Pyris the reaction product is BDPC. The high-resolution mass
Diamond DMA with a sample size of 5.0 cm × 1.0 cm × spectrum shows that the exact mass of the reaction product is
100 μm. The storage modulus and tan δ were determined by 515.2050 g mol−1. The interpreted molecular formula is
subjecting the sample to the temperature scan mode at a pro- C31H31O7 (M + 1), which is consistent with the formula of BDPC.
grammed heating rate of 5 °C min−1 with a frequency of 1 Hz.
Thermal mechanical analysis (TMA) was performed using an
SII TMA/SS6100 at a heating rate of 5 °C min−1. The coefficient
of thermal expansion (CTE) was recorded in the temperature
range of 50 to Tg − 20 °C. The stress–strain curve of epoxy ther-
mosetting films was obtained by EZ-SX at room temperature
with a sample size of 5 cm in length, 1 cm in width, and
40–80 μm in thickness, depending on the sample we cast.
Proposed mechanism for the DMAP-catalyzed model reaction
Fig. S4† shows the 1H NMR spectra of the uncatalyzed reaction
product of GPE and DPC after stirring at 120–160 °C for 2 h.
They all show the same patterns, indicating that DMAP is a
1
required element for the reaction. Fig. S5† shows the H NMR
spectra of DPC with 0.5 wt% DMAP before and after stirring at
120 °C for 2 h. The same spectra were observed, indicating
that the nucleophilicity of DMAP is not high enough to carry
out the nucleophilic substitution with DPC. According to the
1H NMR analysis of the model reaction (Fig. S5† and 1), we
propose a reaction mechanism for the model reaction in
Results and discussion
Model reaction
Fig. S5† shows the 1H NMR analysis of the DMAP-catalyzed Scheme 3 to explain the formation of BDPC. The first step is
reaction of GPE and DPC after reacting at 120 °C for different the attack of DMAP on the epoxide of GPE, forming intermedi-
periods of time. The characteristic signals of epoxide at 3.8 ate (I) with an alkoxy anion. The release of angle strain in an
and 4.2 ppm (Hc), 3.3 ppm (Hb), and 2.6 and 2.8 ppm (Ha) epoxide should be the driving force of this reaction. The
decreased gradually with the reaction time and disappeared second step is the nucleophilic substitution of the alkoxy
completely after 2 h. The Ar–H signals became more compli- anion of (I), which is highly nucleophilic, on the carbonate of
cated after 1 h but gradually simplified into two signals after DPC (marked in blue in step 2), forming intermediate (II) and
2 h. The simple Ar–H signals indicated that the aromatic sec- a phenoxy anion. In this reaction, the alkoxy anion that exhi-
tions of the product are structurally symmetric. After careful bits stronger basicity changed to a phenoxy anion that exhibits
evaluation of the other peaks, such as the signals of methine relatively weaker basicity. The change in basicity is a driving
(H6) at 5.6 ppm and methylene (H5) at 4.3 ppm, we believe that force for the nucleophilic substitution. In contrast, the nucleo-
the reaction product is bis(1,3-diphenoxypropan-2-yl)carbonate philic addition of the alkoxy anion of (I) on the epoxide of GPE
(BDPC, Scheme 1b). The signal at 7.05 comes from the Ar–H (the homopolymerization of epoxides, marked in red in step 2)
that are ortho and para to the ether linkage (H1 and H3), and would result in an alkoxy anion that has the same basicity.
the signal at 7.35 ppm comes from the Ar–H that are meta to Therefore, the thermodynamic factor leads to the exclusive
the ether linkage (H2). According to the homopolymerization selectivity of the nucleophilic substitution over the nucleophi-
of GPE in the presence of a catalytic amount of DMAP,36 the lic addition. The third step is the nucleophilic substitution of
signals of methine at 5.2 ppm and methylene at 3.6–3.8 ppm the phenoxy anion on the carbon of the C–N bond of (II),
were observed. However, we did not find these signals in forming 1,3-diphenoxypropan-2-yl phenyl carbonate (III) and
Fig. S5,† suggesting that we can exclude the homopolymeriza- releasing DMAP. The fourth step is the nucleophilic substi-
tion of GPE in the reaction of GPE and DPC. Fig. 1 shows the tution of the anion of (I) on the carbonate of (III), forming
(a) 1H–1H COSY and (b) 1H–13C HSQC NMR analyses of the intermediate (IV) and a phenoxy anion. The fifth step is the
DMAP-catalyzed reaction of GPE and DPC at 120 °C for 2 h. nucleophilic substitution of the phenoxy anion on the carbon
The assignment in the spectra supports the finding that of the C–N bond of (IV), forming BDPC (V) and releasing
Scheme 3 Proposed mechanism for the DMAP-catalyzed model reaction.
This journal is © The Royal Society of Chemistry 2020
Green Chem.

View Article Online
Paper
Green Chemistry
Scheme 4 Proposed intermolecular interaction of the alkoxy anion (I) with the carbonyl group of (a) ethyl benzoate and (b) acetophenone.
DMAP. This mechanism clearly presents the reaction between
the epoxide and aromatic–aromatic carbonate in the presence
of DMAP.
To further support the assumption that no homopolymeri-
zation of epoxides occurred in step 2 of Scheme 3 (marked in
red), we designed the following two model reactions. Fig. S9†
shows the NMR spectra of ethyl benzoate and GPE with
0.5 wt% DMAP before and after stirring at 120 °C for 2 h. They
all show the same patterns, indicating that no homopolymeri-
zation of epoxides occurred. Fig. S10† shows the NMR spectra
of acetophenone and GPE with 0.5 wt% DMAP before and
after stirring at 120 °C for 2 h. They all show the same pat-
terns, also indicating that no homopolymerization of the
epoxide occurred. However, we have confirmed the homopoly-
merization of GPE at 120 °C with 0.5 wt% of DMAP in the pre-
vious work.36 To explain the fact that no homopolymerization
of GPE occurred in Fig. S5, S9 and S10,† we proposed that the
carbonyl groups in DPC (Fig. S5†), ethyl benzoate (Fig. S9†)
and acetophenone (Fig. S10†) have some kind of inter-
molecular interaction with the alkoxy anion (I), limiting the
nucleophilicity of (I) toward epoxy, as shown by the dashed
line in Scheme 4.
Fig. 2 FTIR analysis of the DMAP-catalyzed reaction of GPE and DPC
after reacting at 120 °C. The reaction time is marked on the figure.
number of 3300–3500 cm−1. This result is consistent with the
proposed mechanism in Scheme 3, in which no homopolymer-
ization of epoxides occurred.
FTIR analysis of the DMAP-catalyzed model reaction and WPC/
epoxy blend
Fig. 2 shows the FTIR analysis of the DMAP-catalyzed reaction
of GPE and DPC after reacting at 120 °C for different periods of
DSC thermograms of the WPC/epoxy blend
time. The carbonyl absorption gradually shifted from 1780 cm−1 Fig. 4 shows the DSC thermograms of epoxy I, epoxy IV, WPC/I
to 1760 cm−1 after 0.5 h, and to 1750 cm−1 after 2 h. In the IR blend, and WPC/IV blend with 0.5 wt% DMAP in each sample.
principle, an aliphatic–aliphatic carbonate has a lower absorp- The WPC/epoxy blends for DSC measurement were prepared
tion wavenumber than an aromatic–aromatic carbonate due to by dissolving WPC, epoxy, and DMAP in dichloromethane, and
the resonance effect.55 The up-shift suggests that the carbonate drying in a vacuum oven at 30 °C overnight. Since the blend
changed from an aromatic–aromatic one to an aromatic–ali- was dried at 30 °C, the pre-reaction is negligible. As shown in
phatic one, and finally to an aliphatic–aliphatic one. No further Fig. 4, we observed a respective exotherm at 95 °C and 109 °C
carbonyl shift was observed after reacting for 2 h, indicating for the neat epoxy I and IV. According to the DSC and NMR
that the reaction was complete in 2 h. The result is consistent analyses in the previous work,56,57 the exotherm corresponds
with the previous NMR analyses in Fig. S5† and Fig. 1.
to the DMAP-catalyzed homopolymerization of an epoxide. For
Fig. 3 shows the FTIR analysis of the DMAP-catalyzed WPC/ the WPC/I and WPC/IV blends, we observed a respective
IV blend after accumulative curing at each temperature (80, exothermic peak at 157 °C and 162 °C. According to the NMR
140, and 180 °C) for 20 min. The carbonyl absorption gradu- analyses in Fig. S5† and 1, along with the carbonyl up-shift in
ally shifted from 1780 cm−1 at room temperature to 1750 cm−1 Fig. 2 and 3, we believe that the exotherm corresponds to the
at 180 °C. This result indicates that the carbonate changed reaction of the epoxide and carbonate linkages. Also, we
from an aromatic–aromatic one to an aromatic–aliphatic one, observed an interesting phenomenon in the DSC thermogram
and finally to an aliphatic–aliphatic one. Also, Fig. 3 shows for WPC/I and WPC/IV blends: no exotherm at around
that no obvious hydroxyl absorption was observed in the wave- 95–110 °C corresponding to the homopolymerization of epox-
Green Chem.
This journal is © The Royal Society of Chemistry 2020

View Article Online
Green Chemistry
Paper
Fig. 3 FTIR spectra of the DMAP-catalyzed WPC/IV blend after accumulative thermal curing at each temperature for 20 min. Left: 2000–800 cm−1
right: 3800–1800 cm−1
;
.
carbonate linkage, changing the carbonate from an aromatic–
aromatic one to an aliphatic–aliphatic one.
Thermal properties
Fig. 5 shows the DMA thermograms of the WPC/epoxy thermo-
sets. Only one tan δ peak was observed for each sample, sup-
porting the co-reaction between WPC and epoxy resins. The
Fig. 4 DSC thermograms of (a) I, (b) IV, (c) WPC/I blend, and (d) WPC/
IV blend. All samples contain 0.5 wt% DMAP.
ides was observed. The DSC data support the exclusive selecti-
vity of the nucleophilic substitution (in blue arrow) over the
nucleophilic addition (in red arrow) in step 2 of Scheme 3 for
the model reaction.
The proposed structure of WPC/epoxy thermosets
The detailed curing mechanism of WPC and an epoxy resin is
similar to that of GPE and DPC in Scheme 3. For simplicity,
the simplified curing mechanism of WPC and epoxy I is
shown in four steps (expressed in black arrows) in Scheme S1.†
The first step is the attack of DMAP on the epoxide of I,
forming an alkoxy anion. The second step is the attack of the
resulting alkoxy anion on the electron-deficient carbonate of
WPC, forming a phenoxy anion. The third step is the attack of
the electron-rich phenoxy anion on the electron-deficient
carbon of the C–N bond, which was formed in the first step.
DMAP was then released in the fourth step. The four-step reac-
tion repeated at the other side of carbonate (expressed in pink
arrows). The overall reaction is that two epoxy molecules
Fig. 5 DMA thermograms of the WPC/epoxy and PN/epoxy
(expressed in red) were inserted into the aromatic–aromatic thermosets.
This journal is © The Royal Society of Chemistry 2020
Green Chem.

View Article Online
Paper
Green Chemistry
Table 1 Thermal and mechanical properties of WPC/epoxy thermosets
CTEc
Crosslinking densityf
Modulus
(GPa)
Sample ID
Tga (°C)
Tgb (°C)
(ppm °C−1
)
Td 5%d (°C)
CRe (%)
(mmol cm−3
)
σg (MPa)
εh (%)
WPC/I
169 (172)i
146 (137)i
125 (126)i
184 (187)i
150 (150)i
122 (107)i
98 (103)i
46 (44)i
47 (49)i
48 (46)i
41 (42)i
382 (415)i
391 (419)i
407 (415)i
385 (422)i
7 (12)i
4 (10)i
2 (1)i
2.53
1.37
0.63
0.95
2.0
1.6
1.2
1.1
79
51
48
57
7.4
10.8
25.3
5.9
WPC/II
WPC/III
WPC/IV
160 (166)i
6 (9)i
a Measured by DMA at a heating rate of 5 °C min−1. The peak temperature of the tan δ curve. b Measured by TMA at a heating rate of 5 °C min−1
c Coefficient of thermal expansion, recorded at 50 °C to 130 °C. d Measured by TGA at a heating rate of 20 °C min−1. The temperature corresponds
.
to a 5% weight loss. e Residual weight % at 800 °C. f Calculated by the equation: υe = E′/3RT, where E′ is the storage modulus from DMA at Tg
+
30 °C; T is the temperature in K corresponding to the storage modulus value; R is the gas constant. g Tensile strength. h Elongation at break. i The
detected value of the PN-cured epoxy thermoset.
modulus after Tg maintains a constant value with the increase WPC/II, WPC/III, and WPC/IV, respectively (Table 1). The
of temperature for each sample, indicating that the samples respective coefficient of thermal expansion (CTE) from 50 °C
are thermosets. A solubility test shows that the four thermo- to (Tg − 20) °C is 46, 34, 48, and 41 ppm °C−1. TMA data also
sets are insoluble in THF, chloroform, DMSO, DMAc, and show that the WPC/epoxy and PN/epoxy thermosets exhibit
NMP, also supporting that the samples are thermosets. For similar thermal–mechanical properties.
thermosets WPC/I, WPC/II, and WPC/III, the values of Tg and
Fig. 7 shows the TGA thermograms of the WPC/epoxy ther-
modulus after Tg are in the order of WPC/I > WPC/II > WPC/ mosets. For the WPC/I, WPC/II, and WPC/III thermosets, the
III, decreasing with the increase of epoxy equivalent weight order of 5 wt% decomposition temperature (Td 5%) is WPC/I
(EEW) of epoxy for the WPC/epoxy thermoset (Table 1). (382 °C) < WPC/II (391 °C) < WPC/III (407 °C). The aliphatic–
Fig. S11† shows a linear relationship between 1/Tg and EEW of aliphatic carbonate linkage in the WPC/epoxy thermoset is
epoxy resins for the WPC/epoxy thermoset. The height of tan thought to be the weakest linkage in thermal stability, so the
(delta) in a DMA thermogram, which is inversely related to the thermal stability is related to the weight ratio of the carbonate
rigidity of the network, is in the order of WPC/I < WPC/II < in a thermoset. The weight ratio of the carbonate linkage is
WPC/III. The thermoset derived from epoxy resin with high WPC/I > WPC/II > WPC/III, and it explains the order of
EEW leads to a lower crosslinking density, explaining the Tg thermal stability. TGA data in Table 1 show that the thermal
value and the height of tan(delta) data in the DMA thermo- stability of WPC/epoxy thermosets is lower than that of PN/
grams. The WPC/IV thermoset, due to the multi-functionality epoxy thermosets, probably due to the weaker carbonate bond.
of epoxy IV, has the highest Tg value of 184 °C. Phenol novolac
Tensile properties
(PN) is a commercial curing agent for epoxy resins. Four PN-
cured epoxy thermosets (PN/I, PN/II, PN/III, and PN/IV) were
also prepared and measured for property comparison. Upon
comparing the data of WPC/epoxy and PN/epoxy thermosets,
we found that the WPC-cured epoxy thermosets show compar-
able Tg values to the PN-cured epoxy thermosets. This result
makes these WPC/epoxy thermosets attractive for applications.
Fig. 6 shows the TMA thermograms of the WPC/epoxy and
PN/epoxy thermosets. The respective Tg value obtained from
the onset temperature is 150, 118, 98, and 160 °C for WPC/I,
Fig. 8 shows the photographs of the WPC/I–WPC/IV thermoset-
ting films. The films are homogeneous, light yellow, and fold-
able (remain intact after a 180° folding). In contrast, PN/I–/IV
films are brittle upon folding. This result demonstrates the
advantage of WPC as a curing agent over PN in toughness. To
the best of our knowledge, epoxy thermosetting films with
foldability are rarely seen. Fig. 9 shows the stress–strain curves
of the WPC/I–WPCIV thermosetting films. The respective
tensile strengths (σ) are 79, 51, 46, and 57 MPa, and the
Fig. 6 TMA thermograms of the WPC/epoxy and PN/epoxy thermosets.
Green Chem.
This journal is © The Royal Society of Chemistry 2020

View Article Online
Green Chemistry
Paper
Fig. 7 TGA thermograms of the WPC/epoxy and PN/epoxy thermosets.
Fig. 8 Photographs of (a) WPC/I, (b) WPC/II, (c) WPC/III and (d) WPC/IV thermosetting films.
equivalent weight of epoxy III is 901 g eq−1, which corresponds
to a molecular weight of 1802 g mol−1. Therefore, WPC/III
shows the highest value of elongation at break, probably due
to the highest molecular weight of epoxy III that led to the
lowest crosslinking density. We also calculated the cross-
linking density based on the modulus at Tg + 30 °C in the
DMA thermograms,58 as listed in Table 1. WPC/III indeed
showed the lowest crosslinking density.
Degradation of the WPC/epoxy thermosets
As shown in the DMA thermograms, the WPC/epoxy thermo-
sets are rigid at elevated temperatures (with the modulus of
107–108 Pa at a temperature above 200 °C). The network is ther-
mally stable (with Td 5% >382 °C), so the degradation is not as
convenient as a vitrimer with a covalent adaptable network.45
As shown in Scheme S1,† the chains in the network are linked
by aliphatic–aliphatic carbonate linkages. We thought that the
linkages could probably be degraded by alkaline hydrolysis,
Fig. 9 The stress–strain curve of WPC/I–WPC/IV thermosetting films.
respective elongations at break (ε) are 7.4, 10.8, 25.3, and transesterification, or aminolysis. Therefore, we applied three
5.9%. WPC/I shows the highest value of tensile strength, prob- methods to degrade the WPC/epoxy thermosets. The first
ably due to the lowest molecular weight of epoxy I. The epoxy method to degrade the thermoset is through alkaline
This journal is © The Royal Society of Chemistry 2020
Green Chem.

View Article Online
Paper
Green Chemistry
Table 2 Degradation of WPC/epoxy thermosets
Thermosets
Solvent
Catalyst
Temp. (°C)
Time (h)
Residual weight (%)
Run 1
Run 2
Run 3
Run 4
Run 5
Run 6
Run 7
Run 8
Run 9
Run 10
Run 11
Run 12
Run 13
WPC/I, 0.055 g
WPC/I, 0.051 g
WPC, 0.3 g
WPC, 0.3 g
WPC/I, 0.3 g
WPC/I, 0.3276 g
WPC/IV, 0.3378 g
WPC/I, 0.124 g
WPC/I, 0.9860 g
WPC/I, 0.1715 g
WPC/II, 0.1870 g
WPC/III, 0.1598 g
WPC/IV, 0.1607 g
H2O (15 mL)
H2O (15 mL)
H2O (10 mL)
H2O (10 mL)
2-EH (15 mL)
2-EH (15 mL)
2-EH (15 mL)
2-EH (10 g)
2-EH (35 g)
HA (5 mL)
5 M NaOH(aq)
5 M KOH(aq)
5 M NaOH(aq)
5 M KOH(aq)
Ti(OBu)4 (0.01 g)
Zn(Ac)2 (0.83 g)a
Zn(Ac)2 (0.83 g)a
[HDBU][IM] (0.270 g)b
[HDBU][IM] (0.947 g)b
—
80
80
80
80
170
170
170
170
8
8
8
8
24
8
48
12
18
1.75
3.5
4.0
1.5
96
95
Slurry state
0
87
35
99
46
0
0
0
0
0
200 (In a Parr reactor)
125
125
125
125
HA (10 mL)
HA (10 mL)
HA (5 mL)
—
—
—
a The concentration of Zn(Ac)2 corresponds to 0.21 M. b The amount of [HDBU][IM] is 1.6 mol% of 2-EH.
hydrolysis. We used 5 M NaOH(aq) and KOH(aq) to degrade
WPC/I, but the thermoset showed little response to the alka-
line condition (runs 1 and 2, Table 2). The thermoset
remained 95–96% intact at 80 °C for 8 h. We are surprised at
the stability of the aliphatic–aliphatic carbonates, compared
with aromatic–aromatic carbonates under the alkaline con-
ditions (runs 3 and 4). The second method to degrade the
thermoset is through transesterification. We used a tra-
ditional catalyst, Ti(OBu)4, for the transesterification of the
aliphatic–aliphatic carbonates and 2-ethylhexanol (2-EH), but
87 wt% of the thermoset remained (run 5). We then used
Zn(Ac)2 as the catalyst, a catalyst recommended by Qi et al.48
The thermoset degraded slowly, with 35 wt% remaining for
WPC/I (run 6) and 99 wt% remaining for WPC/IV (run 7). We
also used the ionic liquid [HDBU][IM], prepared from the
Fig. 10 The curve of residual weight % versus degradation time for the
aminolysis of WPC/epoxy thermosets in runs 10–13, Table 2.
reaction of DBU and imidazole at 25 °C for 8 h, as the cata-
lyst.59 The degradation occurred gradually at 170 °C for 12 h
(46 wt% remained, run 8), and a complete degradation can
be achieved at 200 °C at around 3 atm for 18 h in a Parr
reactor (run 9). However, the harsh condition of high temp-
erature and high pressure limits this process. The low reactiv-
ity of hydroxyl in 2-EH might be the reason for the slow degra-
dation. The third method to degrade the thermoset is
through aminolysis. We used 1-hexylamine, with a highly
reactive amine, to carry out the aminolysis. Due to the high
reactivity of the amine, the aminolysis occurred under cata-
lyst-free conditions (runs 10–13). Fig. 10 shows the curve of
residual weight % versus degradation time for WPC/epoxy
thermosets in runs 10–13, Table 2. We observed an initial
increase in weight for WPC/I and WPC/II, probably due to the
swelling of thermosets by 1-hexylamine. The experimental
results show that the respective time for full degradation is
approximately 105, 210, 240, and 90 min for WPC/I, WPC/II,
WPC/III, and WPC/IV, respectively.
product after reacting at 100 °C for 4 h. No residual DPC was
found in Fig. 11(c), indicating that DPC was degraded comple-
tely. In Fig. 11(c), we observed the signals for phenol, includ-
ing phenolic hydroxyl at 9.4 ppm, and Ar–H at 7.2 and
6.8 ppm. We also found the signals for 1,3-dihexylurea, includ-
ing urea (NH–CO–NH) at 5.7 ppm, methylene (N–CH2–, Hf) at
3.0 ppm, methylene (Hb–e) at 1.2–1.4 ppm, and methyl (Ha) at
0.8 ppm. The signals of phenolic hydroxyl and urea decreased
after D2O was added into the aminolysis product (Fig. 11(d)),
confirming the assignment of phenolic hydroxyl and urea.
From the analysis of Fig. 11, we established that the aminolysis
of DPC by 1-hexylamine led to phenol and 1,3-dihexylurea
(Scheme 5).
Fig. 12 shows the 1H NMR analysis of the aminolysis of
BDPC by 1-hexylamine. Fig. 12(b) shows the 1H NMR spectrum
of BDPC. Fig. 12(c) shows the 1H NMR spectrum of the amino-
lysis product after reacting at 100 °C for 4 h. Comparing
Fig. 12(b) and (c), the signal of methine (H8) at 5.4 ppm of
BDPC remained, indicating that the degradation was not com-
plete. Fig. 12(d) shows the 1H NMR spectrum of the aminolysis
Products of the degradation
1
Fig. 11 shows the H NMR analysis of the aminolysis of DPC
by 1-hexylamine. Fig. 11(b) shows the 1H NMR spectrum of
1
DPC. Fig. 11(c) shows the H NMR spectrum of the aminolysis product after reacting at 125 °C for 6 h. From Fig. 12(b)
Green Chem.
This journal is © The Royal Society of Chemistry 2020

View Article Online
Green Chemistry
Paper
Fig. 11 1H NMR analysis for the aminolysis of DPC by 1-hexylamine. (a) The spectrum of 1-hexylamine, (b) the spectrum of DPC, (c) the spectrum of
the aminolysis product after reacting at 100 °C for 4 h, and (d) the spectrum of the aminolysis product of (c) with D2O added.
Scheme 5 Aminolysis of DPC by 1-hexylamine.
Fig. 12 1H NMR analysis for the aminolysis of BDPC by 1-hexylamine. (a) The spectrum of 1-hexylamine, (b) the spectrum of BDPC, (c) the spectrum
of the aminolysis product after reacting at 100 °C for 4 h, (d) the spectrum of the aminolysis product after reacting at 125 °C for 6 h, and (e) the spec-
trum of the aminolysis product of (d) with D2O added.
This journal is © The Royal Society of Chemistry 2020
Green Chem.

View Article Online
Paper
Green Chemistry
and (d), the signal of methine (H8) at 5.4 ppm of BDPC dis- bonate in BDPC is chemically more stable than the aromatic–
aromatic carbonate in DPC. This result is consistent with the
fact that WPC/I is stable in the alkaline hydrolysis (runs 1 and
2, Table 2), while WPC is unstable under alkaline conditions
(run 4, Table 2).
appeared, confirming that the degradation was complete.
In Fig. 12(d), we observed the signals for 1,3-diphenoxypro-
pan-2-ol, including Ar–H at 7.2 and 6.8 ppm, hydroxyl at
5.3 ppm, methine (Hb) at 4.25 ppm, and methylene (O–CH2–,
Ha) at 4.1 ppm. We also found the signals for 1,3-dihexylurea,
including urea (NH–CO–NH) at 7.4 ppm, methylene (N–CH2–,
Hh) at 3.0 ppm, methylene (Hd–g) at 1.2–1.4 ppm, and methyl
(Hc) at 0.8 ppm. The signals of hydroxyl and urea decreased
after D2O was added into the aminolysis product (Fig. 12(e)),
confirming the assignment of hydroxyl and urea. Note that the
position of the urea signal (NH–CO–NH) in Fig. 12(d) is
different from that in Fig. 11(c). The downshift is probably due
to the hydrogen bonding between urea and 1,3-diphenoxypro-
pan-2-ol. From the analysis of Fig. 12, we confirmed that the
aminolysis of the compound BDPC by 1-hexylamine leads to
1,3-diphenoxypropan-2-ol and 1,3-dihexylurea (Scheme 6).
Interestingly, Fig. 11(c) shows that the aminolysis of DPC can
be completed at 100 °C for 4 h, while Fig. 12(c) shows that the
aminolysis of BDPC at 100 °C for 4 h only led to an incomplete
reaction. This result indicates that the aliphatic–aliphatic car-
Fig. 13 shows the 1H NMR analysis of the aminolysis of
WPC/I by 1-hexylamine. Fig. 13(a) shows the 1H NMR spectrum
of the aminolysis product after reacting at 120 °C for 4 h. In
Fig. 13(a), we observed the signals for a phenoxy resin, a poly-
hydroxyether which is usually derived from the reaction of
bisphenol A and epichlorohydrin, including Ar–H at 7.2 and
6.8 ppm, hydroxyl at 5.3 ppm, methine (Hb) at 4.25 ppm,
methylene (O–CH2–, Ha) at 4.1 ppm, and methyl at 1.6 ppm.
We also found the signals for 1,3-dihexylurea, including urea
(NH–CO–NH) at 7.4 ppm, methylene (N–CH2–, Hi) at 3.0 ppm,
methylene (He–h) at 1.2–1.4 ppm, and methyl (Hd) at 0.8 ppm.
We also poured the aminolysis product into methanol.
1
Fig. 13(b) shows the H NMR spectrum of the methanol-preci-
pitated aminolysis product, in which most of the 1,3-dihexyl-
urea is soluble in methanol. Therefore, the precipitate should
be the phenoxy resin. The signals (Ar–H, OH, and Ha–Hc)
Scheme 6 Aminolysis of bis(1,3-diphenoxypropan-2-yl)carbonate (BDPC) by 1-hexylamine.
Fig. 13 1H NMR analysis of the aminolysis of WPC/I by 1-hexylamine. (a) The spectrum of the aminolysis product of WPC/I by 1-hexylamine after
reacting at 120 °C for 4 h and (b) the spectrum of the methanol-precipitated aminolysis product of (a).
Green Chem.
This journal is © The Royal Society of Chemistry 2020

View Article Online
Green Chemistry
Paper
Scheme 7 Aminolysis of WPC/I by 1-hexylamine.
support that the precipitate is the phenoxy resin. We tried to
analyze the molecular weight of the phenoxy resin, but its
Acknowledgements
insolubility in NMP and THF, probably due to the high mole- This work was financially supported by the “Advanced
cular weight, hinders the measurement. Based on the chem- Research Center For Green Materials Science and Technology”
istry of the model reaction in Scheme 6 and the NMR analysis from The Featured Area Research Center Program within the
in Fig. 13, we established that the aminolysis of WPC/I by framework of the Higher Education Sprout Project by the
1-hexylamine led to a phenoxy resin and 1,3-dihexylurea Ministry of Education (107L9006) and the Ministry of Science
(Scheme 7). 1,3-Dihexylurea can be an additive as a biodiesel and Technology in Taiwan (MOST 109-2634-F-002-042 and 107-
or biolubricant. The phenoxy resin is widely used in coatings, 2221-E-005-026).
adhesives, composites, prepregs, and molded parts.
Notes and references
Conclusions
1 M. Hong and E. Y.-X. Chen, Green Chem., 2017, 19, 3692–3706.
2 C. Yang, Q. Tan, L. Liu, Q. Dong and J. Li, ACS Sustainable
Chem. Eng., 2017, 5, 9586–9598.
3 A. Kazama and Y. Kohsaka, Polym. Chem., 2019, 10, 2764–
2768.
4 X. Kuang, Y. Zhou, Q. Shi, T. Wang and H. J. Qi, ACS
Sustainable Chem. Eng., 2018, 6, 9189–9197.
5 I. A. Ignatyev, W. Thielemans and B. Vander Beke,
ChemSusChem, 2014, 7, 1579–1593.
6 J. Song, H. Fan, J. Ma and B. Han, Green Chem., 2013, 15,
2619–2635.
7 S. Hu, Z. Zhang, Y. Zhou, J. Song, H. Fan and B. Han, Green
Chem., 2009, 11, 873–877.
8 Z. Xie, B. Chen, H. Wu, M. Liu, H. Liu, J. Zhang, G. Yang
and B. Han, Green Chem., 2019, 21, 606–613.
9 J. Esteban, A. J. Vorholt and W. Leitner, Green Chem., 2020,
22, 2097–2128.
Through FTIR and NMR analysis of the DMAP-catalyzed reac-
tion of DPC and GPE (Fig. 1 and 2), we confirm that an aro-
matic–aromatic carbonate linkage will react with two epoxides,
producing an aliphatic–aliphatic carbonate linkage in a quite
efficient and selective way (Scheme 1b). Inspired by this chem-
istry, we exemplified a practical approach to recycling wasted
polycarbonates (WPC). We recycled the after-use PC from a
compact disc and used it as a curing agent for four epoxy
resins (Scheme 3). The WPC-cured epoxy thermosets show
foldability (Fig. 9) and similar thermal properties to those of
PN-cured epoxy thermosets (Fig. 5–7). Since the WPC was used
directly without any digestion or pyrolysis, the atomic
efficiency is 100%, making this WPC recycling strategy econ-
omically attractive. To make the WPC/epoxy thermosets
degradable when their lifespans are expired, we explored
methods to degrade the thermosets. Among the methods we
have examined, aminolysis by 1-hexylamine under catalyst-free
1
10 Polycarbonate sheet market size worth $4.91 billion by
conditions shows the best degrading efficiency. H NMR ana-
2025,
https://www.grandviewresearch.com/press-release/
lysis shows that the aminolysis of the model compound,
BDPC, by 1-hexylamine leads to 1,3-diphenoxypropan-2-ol and
1,3-dihexylurea (Fig. 12 and Scheme 6). The degradation pro-
ducts of WPC/I, according to the NMR analysis (Fig. 13), are
the phenoxy resin and 1,3-dihexylurea (Scheme 7). The trans-
formation of WPC into epoxy thermosets and the catalyst-free
degradation of the thermosets could open a door for the re-
cycling (or re-use) of PC-made products for environmental
sustainability.
global-polycarbonate-sheet-market.
11 E. V. Antonakou and D. S. Achilias, Waste Biomass
Valorization, 2013, 4, 9–21.
12 A. Oku, S. Tanaka and S. Hata, Polymer, 2000, 41, 6749–
6753.
13 L.-C. Hu, A. Oku and E. Yamada, Polymer, 1998, 39, 3841–3845.
14 H. Jie, H. Ke, Z. Qing, C. Lei, W. Yongqiang and Z. Zibin,
Polym. Degrad. Stab., 2006, 91, 2307–2314.
15 M. Watanabe, Y. Matsuo, T. Matsushita, H. Inomata,
T. Miyake and K. Hironaka, Polym. Degrad. Stab., 2009, 94,
2157–2162.
Conflicts of interest
16 C.-H. Lin, H.-Y. Lin, W.-Z. Liao and S. A. Dai, Green Chem.,
2007, 9, 38–43.
The authors declare no competing financial interest.
This journal is © The Royal Society of Chemistry 2020
Green Chem.

View Article Online
Paper
Green Chemistry
17 R. Piñero, J. García and M. J. Cocero, Green Chem., 2005, 7, 40 T. Liu, C. Hao, L. Wang, Y. Li, W. Liu, J. Xin and J. Zhang,
380–387. Macromolecules, 2017, 50, 8588–8597.
18 M. Liu, J. Guo, Y. Gu, J. Gao, F. Liu and S. Yu, ACS 41 T. Liu, X. Guo, W. Liu, C. Hao, L. Wang, W. C. Hiscox,
Sustainable Chem. Eng., 2018, 6, 13114–13121.
19 L. Chen, Y. Wu, Y. Ni, K. Huang and Z. ZHU, J. Ind. Eng.
Chem., 2004, 11, 1787–1792.
20 L. Chen, Y. Wu, Y. Ni, K. Huang and Z. Zhu, Acta Sci.
Circumstantiae, 2004, 24, 604.
21 E. Quaranta, D. Sgherza and G. Tartaro, Green Chem., 2017,
19, 5422–5434.
22 C.-H. Wu, L.-Y. Chen, R.-J. Jeng and S. A. Dai, ACS
Sustainable Chem. Eng., 2018, 6, 8964–8975.
C. Liu, C. Jin, J. Xin and J. Zhang, Green Chem., 2017, 19,
4364–4372.
42 S. Zhang, T. Liu, C. Hao, L. Wang, J. Han, H. Liu and
J. Zhang, Green Chem., 2018, 20, 2995–3000.
43 Y. Li, T. Liu, S. Zhang, L. Shao, M. Fei, H. Yu and J. Zhang,
Green Chem., 2020, 22, 870–881.
44 S. Wang, S. Ma, Q. Li, X. Xu, B. Wang, W. Yuan,
S. Zhou, S. You and J. Zhu, Green Chem., 2019, 21, 1484–
1497.
23 S. Singh, Y. Lei and A. Schober, RSC Adv., 2015, 5, 3454– 45 B. Wang, S. Ma, S. Yan and J. Zhu, Green Chem., 2019, 21,
3460.
5781–5796.
24 J. G. Kim, Polym. Chem., 2020, DOI: 10.1039/C9PY01927H.
25 G. Montaudo, S. Carroccio and C. Puglisi, J. Anal. Appl.
Pyrolysis, 2002, 64, 229–247.
26 B. N. Jang and C. A. Wilkie, Thermochim. Acta, 2005, 426,
73–84.
27 H. Tagaya, K. Katoh, J.-I. Kadokawa and K. Chiba, Polym.
Degrad. Stab., 1999, 64, 289–292.
28 M. Chen, D. Hourston and W. Sun, Eur. Polym. J., 1992, 28,
1471–1475.
46 J. Han, T. Liu, C. Hao, S. Zhang, B. Guo and J. Zhang,
Macromolecules, 2018, 51, 6789–6799.
47 P. Yang, Q. Zhou, X. Y. Li, K. K. Yang and Y.-Z. Wang,
J. Reinf. Plast. Compos., 2014, 33, 2106–2114.
48 X. Kuang, E. Guo, K. Chen and H. J. Qi, ACS Sustainable
Chem. Eng., 2019, 7, 6880–6888.
49 S. L. Buchwalter and L. L. Kosbar, J. Polym. Sci., Part A:
Polym. Chem., 1996, 34, 249–260.
50 P. Taynton, H. Ni, C. Zhu, K. Yu, S. Loob, Y. Jin, H. J. Qi
and W. Zhang, Adv. Mater., 2016, 28, 2904–2909.
29 M. Abbate, E. Martuscelli, P. Musto, G. Ragosta and
G. Scarinzi, J. Polym. Sci., Part B: Polym. Phys., 1994, 32, 51 A. Ruiz de Luzuriaga, R. Martin, N. Markaide, A. Rekondo,
395–408.
G. Cabañero, J. Rodríguez and I. Odriozola, Mater. Horiz.,
2016, 3, 241–247.
52 Y. Yuan, Y. Sun, S. Yan, J. Zhao, S. Liu, M. Zhang, X. Zheng
and L. Jia, Nat. Commun., 2017, 8, 14657.
53 S. Wang, X. Xing, X. Zhang, X. Wang and X. Jing, J. Mater.
Chem. A, 2018, 6, 10868–10878.
54 J. N. Lo, S. R. Nutt and T. J. Williams, ACS Sustainable
Chem. Eng., 2018, 6, 7227–7231.
55 D. L. Pavia, G. M. Lampman, G. S. Kriz and J. A. Vyvyan,
Introduction to Spectroscopy, Brooks/Cole:Pacific Grove, 4th
edn, 2010, p. 68.
56 C.-H. Chen, K.-W. Lee, C.-H. Lin, M.-J. Ho, M.-F. Hsu,
S.-J. Hsiang, N.-K. Huang and T.-Y. Juang, Polymers, 2017,
10, 27.
30 T. M. Don and J. Bell, J. Appl. Polym. Sci., 1998, 69, 2395–
2407.
31 C. C. Su, J.-F. Kuo and E. M. Woo, J. Polym. Sci., Part B:
Polym. Phys., 1995, 33, 2235–2244.
32 C. C. Su and E. M. Woo, Macromolecules, 1995, 28, 6779–
6786.
33 M.-S. Li, C.-C. M. Ma, J.-L. Chen, M.-L. Lin and
F.-C. Chang, Macromolecules, 1996, 29, 499–506.
34 M.-S. Li, Y.-F. Su, C.-C. M. Ma, J.-L. Chen, M.-S. Lu and
F.-C. Chang, Polymer, 1996, 37, 3899–3905.
35 K. Funahashi, Bull. Chem. Soc. Jpn., 1979, 52, 1488–1492.
36 C. H. Chen, Z. C. Gu, Y. L. Tsai, R. J. Jeng and C. H. Lin,
Polymer, 2018, 140, 225–232.
37 D. Montarnal, M. Capelot, F. Tournilhac and L. Leibler, 57 C.-H. Chen, S.-H. Tung, R.-J. Jeng, M. M. Abu-Omar and
Science, 2011, 334, 965–968. C.-H. Lin, Green Chem., 2019, 21, 4475–4488.
38 K. Yu, Q. Shi, M. L. Dunn, T. Wang and H. J. Qi, Adv. Funct. 58 M. Ariraman, R. Sasikumar and M. Alagar, RSC Adv., 2015,
Mater., 2016, 26, 6098–6106. 5, 69720–69727.
39 W. Liu, D. F. Schmidt and E. Reynaud, Ind. Eng. Chem. Res., 59 M. Liu, J. Guo, Y. Gu, J. Gao and F. Liu, ACS Sustainable
2017, 56, 2667–2672. Chem. Eng., 2018, 6, 15127–15134.
Green Chem.
This journal is © The Royal Society of Chemistry 2020