Hybrid monomer design for unifying conflicting polymerizability, recyclability, and performance properties

Article abstract of DOI:10.1016/j.chempr.2021.02.003

Synthetic polymers have become indispensable for modern life and the global economy. However, the manufacturing and disposal of most of today's polymers follow a linear economy model, which has caused accelerated depletion of finite natural resources, severe worldwide plastics pollution, and enormous materials value loss. The design of future circular polymers considers closed-loop lifecycles toward a circular economy. A key challenge of this promising design includes innovation in monomer structure that could enable not only efficient polymerization to polymers with properties comparable to today's polymers but also selective depolymerization to recover the monomers with high yield and purity. However, these contrasting properties are conflicting in a single monomer structure. This work introduces a hybrid monomer design concept that hybridizes contrasting parent monomer structures to an offspring monomer that can unify conflicting (de)polymerizability and performance properties. Intrinsically recyclable polymers represent a circular economy approach to address plastics problems. However, the design of such circular polymers is challenged by unyielding trade-offs between the monomer's polymerizability and the polymer's depolymerizability and performance properties. Here, we introduce a hybrid monomer design strategy that synergistically couples a high ceiling temperature (HCT) sub-structure for high polymerizability and performance properties with a low ceiling temperature (LCT) sub-structure for high depolymerizability and recyclability within the same monomer structure. Thus, structural hybridization between HCT ε-caprolactone and LCT γ-butyrolactone led to an offspring [3.2.1]bicyclic lactone, which exhibits both high polymerizability and depolymerizability, otherwise conflicting properties in a typical monomer. The resulting polymer becomes a high-performance material, and thermal transition temperatures are ～200°C higher and tensile modulus 10× higher than its parent polymers. These results demonstrate that the HCT/LCT hybrid monomer strategy is a powerful approach for designing circular polymers where conflicting properties must be exploited and unified. Hybrid monomer design is shown to be a powerful approach to develop robust circular plastics without yielding to common property trade-offs by hybridizing parent monomer structures to an offspring monomer that can unify conflicting (de)polymerizability and performance properties.

Full text of DOI:10.1016/j.chempr.2021.02.003

ll
Article
Hybrid monomer design for unifying
conflicting polymerizability, recyclability, and
performance properties
Changxia Shi, Zi-Chen Li, Lucia
Caporaso, Luigi Cavallo, Laura
Falivene, Eugene Y.-X. Chen
lafalivene@unisa.it (L.F.)
eugene.chen@colostate.edu (E.Y.-X.C.)
HIGHLIGHTS
Circular plastics can be designed
with high-performance properties
Hybrid monomer design radically
alters properties of the resulting
recyclable polymers
Property trade-offs can be
overcome by hybridizing low/high
ceiling temperature monomers
Hybrid monomer design is shown to be a powerful approach to develop robust
circular plastics without yielding to common property trade-offs by hybridizing
parent monomer structures to an offspring monomer that can unify conflicting (de)
polymerizability and performance properties.
Shi et al., Chem 7, 670–685
March 11, 2021 ª 2021 Elsevier Inc.
https://doi.org/10.1016/j.chempr.2021.02.003

ll
Article
Hybrid monomer design for unifying
conflicting polymerizability,
recyclability, and performance properties
Changxia Shi,1,2 Zi-Chen Li,2 Lucia Caporaso,3 Luigi Cavallo,4 Laura Falivene,3,
*
and Eugene Y.-X. Chen1,5,
*
SUMMARY
The bigger picture
Synthetic polymers have become
indispensable for modern life and
the global economy. However,
the manufacturing and disposal of
most of today’s polymers follow a
linear economy model, which has
caused accelerated depletion of
finite natural resources, severe
worldwide plastics pollution, and
enormous materials value loss.
The design of future circular
polymers considers closed-loop
lifecycles toward a circular
Intrinsically recyclable polymers represent a circular economy
approach to address plastics problems. However, the design of
such circular polymers is challenged by unyielding trade-offs be-
tween the monomer’s polymerizability and the polymer’s depoly-
merizability and performance properties. Here, we introduce a
hybrid monomer design strategy that synergistically couples a
high ceiling temperature (HCT) sub-structure for high polymerizabil-
ity and performance properties with a low ceiling temperature (LCT)
sub-structure for high depolymerizability and recyclability within
the same monomer structure. Thus, structural hybridization be-
tween HCT ε-caprolactone and LCT g-butyrolactone led to an
offspring [3.2.1]bicyclic lactone, which exhibits both high polymeriz-
ability and depolymerizability, otherwise conflicting properties in a
typical monomer. The resulting polymer becomes a high-perfor-
mance material, and thermal transition temperatures are ꢀ200ꢁC
higher and tensile modulus 103 higher than its parent polymers.
These results demonstrate that the HCT/LCT hybrid monomer strat-
egy is a powerful approach for designing circular polymers where
conflicting properties must be exploited and unified.
economy. A key challenge of this
promising design includes
innovation in monomer structure
that could enable not only
efficient polymerization to
polymers with properties
comparable to today’s polymers
but also selective
INTRODUCTION
depolymerization to recover the
monomers with high yield and
purity. However, these
The current largely unsustainable plastics manufacturing, consumption, and disposal
schemes have caused alarming plastics problems associated with their environ-
mental pollution as well as tremendous energy and materials value loss in the econ-
omy.1–4 Among several approaches being explored5–16 to combat such problems,
the development of next-generation, chemically recyclable polymers17–33 promises
a circular plastics economy approach to address these complex issues.34–36 Mono-
mer design37,38 is the cornerstone of discovering new circular polymers with intrinsic
chemical recyclability, the design of which requires to overcome typical trade-offs be-
tween monomer’s polymerizability and polymer’s depolymerizability, as well as be-
tween polymer’s recyclability and performance, until a practically useful balance is
struck. These stringent thermodynamic, kinetic, and real-world performance require-
ments present a formidable challenge for developing circular plastics that exhibit not
only full chemical recyclability but also high-performance properties. Nevertheless,
notable advances have been achieved toward that ultimate goal. For example, the
ring-opening polymerization (ROP) of low ceiling temperature (LCT), ‘‘nonstrained’’
five-membered g-butyrolactone (g-BL) led to poly(g-butyrolactone) (PGBL) that can
be completely depolymerized back to g-BL in high (>99%) purity and yield with a
low energy input.39,40 The limited stability and mechanical performance properties
contrasting properties are
conflicting in a single monomer
structure. This work introduces a
hybrid monomer design concept
that hybridizes contrasting parent
monomer structures to an
offspring monomer that can unify
conflicting (de)polymerizability
and performance properties.
670 Chem 7, 670–685, March 11, 2021 ª 2021 Elsevier Inc.

ll
Article
of PGBL have been subsequently addressed by trans fusion of a cyclohexyl ring to the
g-BL core, affording a fused bicyclic lactone monomer that led to much-enhanced
monomer polymerizability as well as better polymer thermal stability and crystallinity
while maintaining the full chemical recyclability as a circular plastic.21,22 However, the
resulting highly crystalline polymer is a brittle material, and to increase its ductility as
a packaging material, synthesizing a copolymer with a sufficient amount of the flex-
ible PGBL incorporated was required.41 Another notable set of designer monomers
for circular plastics is a class of bridged bicyclic thiolactone monomers such as chiral
N-substituted cis-4-thia-L-proline thiolactones25 and 2-thiabicyclo[2.2.1]heptan-3-
one ([221]BTL).42 The bridged bicyclic framework locks the monomer structure in
the cis-configuration, thus rendering the chemical recyclability and depolymerization
selectivity. The classic Thorpe-Ingold effect through gem-disubstitution has also
been exploited in the b-butyrothiolactone monomer design to enable the chemical
recyclability.43 Here, it is worth noting that the chemistry of the analogous lactones
is notably different than the above thiolactones. For example, replacing the S atom
in [221]BTL with the O atom gives the bicyclic lactone congener, which leads to a poly-
ester that is atactic, amorphous, and not recyclable—essentially the opposite of the
properties of the polythioester.42
Thus, it remains a challenging task to combine several desired, but conflicting, prop-
erties into one lactone monomer. As copolymerization of two or more different
monomers is considered as an effective strategy to obtain the materials with tailored
properties by adjusting the chemical nature, composition, and sequence of the co-
monomers, often delivering improved or unattainable properties in relation to those
of the constituent homopolymers,44–51 it thus became an attractive strategy to
address the trade-off issues. In this context, the LCT g-BL was copolymerized with
high ceiling temperature (HCT) and strained seven-membered ε-caprolactone
(ε-CL)52,53 to produce copolymers aimed for enhanced polymerizability and thermal
stability for PGBL, and degradability and recyclability for poly(ε-caprolactone)
(PCL).54–57 However, over the entire copolymer composition range, the glass-transi-
tion temperature (Tg, from –48ꢁC to –65ꢁC) and melting-transition temperature (Tm,
from 11ꢁC to 63ꢁC) of the copolymers are still low and confined within the values of
their parent homopolymers (Figure 1A).58 Most recently, the comonomer strategy
has been utilized to render degradability and recyclability of thermosets via incorpo-
ration of a cleavable comonomer.59 Whereas copolymerization can be employed as
an effective strategy to mitigate the trade-offs, increasing speciation can complicate
the chemical recycling and the copolymer properties are often still confined within
those of constituent homopolymers.58
1Department of Chemistry, Colorado State
University, Fort Collins, CO 80523-1872, USA
2Beijing National Laboratory for Molecular
Sciences, Key Laboratory of Polymer Chemistry
and Physics of Ministry of Education, Center for
Soft Matter Science and Engineering, College of
Chemistry & Molecular Engineering, Peking
University, Beijing 100871, China
Hybridization is a common phenomenon found in chemistry and biology, the pro-
cess of which combines the qualities of different varieties or species of parents to
produce offspring that can be distinctively or radically different than the parents.
Applying this general hybridization principle, here we introduce a hybrid monomer
design strategy that synergistically couples a HCT sub-structure for high polymeriz-
ability and performance properties with a LCT sub-structure for high depolymeriz-
ability and recyclability within the same monomer structure. To demonstrate the
uniqueness and effectiveness of this strategy for circular polymer design where con-
flicting properties must be balanced or exploited, herein we describe that structural
hybridization between the HCT ε-CL and the LCT g-BL leads to an offspring [3.2.1]
bicyclic lactone (BiL), 6-oxabicyclo[3.2.1]octan-7-one, which can be synthesized
from 3-cyclohexene-1-carboxylic acid via a two-step reaction in a relatively large lab-
oratory scale (80 g) and high yield (ꢀ92% isolated overall yield) (Figure 1B). We hy-
pothesized that this hybrid monomer structure could bring about an intriguing
3Universita` di Salerno, Dipartimento di Chimica e
Biologia, Via Papa Paolo Giovanni II, 84100
Fisciano, SA, Italy
4King Abdullah University of Science and
Technology (KAUST), Physical Sciences and
Engineering Division, KAUST Catalysis Center,
Thuwal 23955-6900, Saudi Arabia
5Lead contact
*Correspondence: lafalivene@unisa.it (L.F.),
eugene.chen@colostate.edu (E.Y.-X.C.)
https://doi.org/10.1016/j.chempr.2021.02.003
Chem 7, 670–685, March 11, 2021 671

ll
Article
Figure 1. Hybrid monomer design versus conventional copolymerization approach
(A) Conventional comonomer copolymerization approach.
(B) Structural hybridization approach reported in this work.
divergent polymerization/depolymerization pathway scenario where the polymeri-
zation proceeds as if via the ring-opening of the HCT sub-structure and the depoly-
merization occurs as if through the ring-closing to form the LCT sub-structure (Fig-
ure 1B), thereby achieving both high polymerizability and depolymerizability,
otherwise conflicting properties in a typical monomer. We further reasoned that
the ROP of one ring in the bicyclic BiL will keep the other ring intact, thus leading
to an intriguing ring-containing (in-chain) polyester, and that incorporation of the
in-chain aliphatic rings could enhance the thermal properties of the resulting poly-
mer because of increased backbone rigidity.21,60,61 Furthermore, the resulting poly-
mer exhibits two stereogenic centers in a repeat unit, which presents a number of
possible stereo-microstructures or tacticities and, thus, the opportunity to use poly-
mer stereochemistry controllable by catalysts or initiators to tune materials
properties.
Indeed, BiL can be readily polymerized under ambient conditions to high-molecular-
weight poly(BiL) (PBiL) materials that exhibit both high-performance properties and
complete chemical recyclability. Most remarkably, PBiL displays a high Tg up to 135ꢁC
and Tm up to 263ꢁC, which are ꢀ200ꢁC higher than both Tg and Tm values of the parent
homopolymers (PGBL and PCL) and their copolymers (Figure 1B). In addition, mechan-
ical properties of PBiL, particularly Young’s modulus, are ꢀ103 higher than its parent
polymers, demonstrating that the HCT/LCT hybrid monomer strategy is a powerful
approach for designing circular polymers and delivering radically different or much-
enhanced properties over the parent homopolymers or copolymers.
RESULTS
Suppressing epimerization to achieve exclusive stereoretention ROP
The hybrid atom-bridged bicyclic BiL is locked in a cis-configuration whereas a trans-
configurational isomer is not possible in its monomer state. Upon the ROP, however,
the resulting polymer PBiL can adopt either a cis (sterereoretention) or trans (epimeriza-
tion)configurationalstructure,orboth. Hence,ourfirstobjective wastoexaminewhether
the ROP retains the cis-configuration or undergoes epimerization to form a trans struc-
tureandpossible effects ofthe catalystormechanism onthe stereorentention versusepi-
merization dynamics. To this end, we probed the ROP characteristics employing both
organocatalyzed62–64 and metal-catalyzed coordination-insertion mechanisms.65,66
As triazabicyclodecene (TBD) has been shown to be one of the most active and efficient
organic catalysts for the ROP of lactones,67,68 at the outset, we investigated character-
istics of the ROP of BiL withTBD. The TBD-catalyzed ROP in toluene(6.0M) at room tem-
perature (RT, ꢀ23ꢁC) with [BiL]/[TBD]/[BnOH] = 600/3/1 (BnOH = benzyl alcohol as the
672 Chem 7, 670–685, March 11, 2021

ll
Article
A
C
B
D
1
Figure 2. H NMR (23ꢁC, CDCl3) spectra and thermal properties of PBiL produced by organopolymerization
(A) PBiL produced by TBD ([BiL]/[TBD]/[BnOH] = 600/3/1) contaning 18% trans isomer (derived from epimerization).
(B) A reference PBiL produced by metal-catalyzed ROP (stereoretention polymer product without epimerization) with La-1 ([BiL]/[La-1] = 100/1).
(C) DSC (from a second heating scan at 10ꢁC minꢂ1) thermogram of PBiL (Mn = 56.7 kg molꢂ1) by TBD.
(D) TGA and DTG (at 10ꢁC minꢂ1 heating scan) thermograms of PBiL (Mn = 56.7 kg molꢂ1) by TBD.
initiator) achieved 95% conversion after 48 h. The resulting polymer PBiL had a medium
number-average molecular weight (Mn) of 56.7 kg molꢂ1 but the dispersity (Ð) was rela-
tively broad (Ð = 1.22). NMR studies of the PBiL produced by TBD showed that epime-
rization at the carbon stereogenic center adjacent to the carbonyl carbon took place
during the ROP, thusaffording thePBiLcontainingboth cis(82%) andtrans (18%) stereo-
configurations(Figure2A). Morespecifically, the1H NMRspectrumexhibited two setsof
peaks at both the alkoxy methine proton [–CHO–] region (labels a and a0 for cis and trans
configurations, respectively)andtheacylmethine proton[–CHC(C=O)–] region(labels b
and b0 for cis and trans configurations, respectively). This finding is consistent with re-
ports that strongly nucleophilic and basic TBD often causes side reactions during poly-
merization, suchas transesterification and epimerization.69 As a result, the resulting PBiL
is an amorphous material but displays a high glass-transition temperature (Tg) of 119ꢁC,
as determined by differential scanning calorimetry (DSC) (Figure 2C). This Tg value rep-
resents ꢀ170ꢁC–180ꢁC enhancements in Tg in relation to homopolymers PGBL and PCL
aswellastheircopolymers. Noteworthy also is thatitexhibits ahigh decompositiontem-
perature (Td,5%, defined by the temperature at 5% weight loss) of 318ꢁC and a high
maximum rate decomposition temperature (Tmax) of 354ꢁC, as measured by thermogra-
vimetric analysis(TGA)andderivativethermogravimetric analysis (DTG)(Figure2D). This
Td value is about 116ꢁC higher than that of PGBL and similar to that of PCL.
To suppress or eliminate epimerization during polymerization, next, we turned our
attentionto themetal-mediated coordination-insertion mechanismby metal complexes
bearing a vacant coordination site for binding thus activating the monomer and a nucle-
ophilic ligand for initiating the polymerization.65,66 As a start, we first employed La
[N(SiMe3)2]3 (La-1), as it is commercially available and has been shown to be highly effec-
tive for catalyzing coordination-insertion ROP of lactones39 and also leading to cyclic
Chem 7, 670–685, March 11, 2021 673

ll
Article
Table 1. Results of ROP of BiL by different metal catalyst/initiator systems
[BiL]/[Cat]
/[I]
Mn(calcd)b
Run no.
Catalyst no.
Time (h)
Conv.a (%)
(kg molꢂ1
)
Mnc (kg molꢂ1
)
Ðc (Mw/Mn)
I*d (%)
Tacticitye (Pm
)
1
La-1
300/1/3
200/1/1
1,000/1/1
1,500/1/1
100/1/1
1,000/1/1
100/1/1
100/1/1
100/1/1
100/1/1
100/1/1
100/1/1
8
92
91
93
89
93
52
86
89
87
65
52
46
11.7
23.1
117
13.1
27.5
94.3
153
1.10
1.03
1.06
1.06
1.07
1.05
1.05
1.05
1.06
1.06
1.05
1.09
90
0.35
0.34
0.38
0.36
0.46
0.52
0.34
0.42
0.44
0.74
0.85
0.87
2
rac-La-2
rac-La-2
rac-La-2
rac-Y-1
rac-Y-1
rac-Y-2
rac-Y-3
rac-Y-3
(S,S)-Y-3
(S,S)-Y-3
(S,S)-Y-3
3
84
3
12
12
12
16
12
72
72
48
48
48
124
110
98
4
168
5
11.8
65.6
11.0
11.3
11.1
8.43
6.67
5.92
12.0
59.2
13.4
10.4
9.90
8.90
6.70
5.60
6
111
82
7
8
84
9f
10
11f
12g
112
95
100
106
Conditions: [BiL]0 = 6.0 M, toluene, RT, initiator (I) = BnOH (except runs 9, 11, and 12).
aMonomer conversions measured by 1H NMR of the quenched solution in benzoic acid/chloroform.(+)-1-Phenylethanol as initiator.
bMn (calculated) = ([BiL]0/[I]0) 3 conv.% 3 (molecular weight of BiL) + (molecular weight of I).
cWeight-average molecular weights (Mw), number-average molecular weights (Mn), and dispersity indices (Ð = Mw/Mn) determined by GPC at 40ꢁC in CHCl3
coupled with a DAWN HELEOS II multi (18)-angle light scattering detector and an Optilab TrEX dRI detector for absolute molecular weights.
dInitiation efficiency (I*) = Mn(calcd)/Mn(exptl).
eMeasured by 13C NMR in the carbonyl region with Pm percentages relative to the cis/cis diisotactic peak at 174.0 ppm established by the enantiomerically pure
monomer.
f(R)-(+)-1-Phenylethanol as initiator.
g(S)-(-)-1-Phenylethanol as initiator.
polyesters when used in the absence of an alcohol co-initiator.21 The effectiveness of La-
1 toward the ROP of BiL was examined with a [BiL]/[La-1] ratio of 100/1 in toluene,
achieving 96% conversion at RT after 24 h. Importantly, the PBiL produced by the
metal-catalyzed coordination-insertion ROP completely returned the cis-configuration
without noticeable racemization on the basis of 1H NMR analysis (Figures 2B versus
2A). Specifically, the1H NMR spectrum of the PBiLproduced by La-1 displayed no peaks
associated with the trans configuration (i.e., a0- and b0-labeled peaks attributed to epi-
merization, Figure 2B). When combined with the co-initiator BnOH that converts instan-
taneously the silylamide precatalyst to the alkoxide catalyst in situ,39 the polymerization
at RT in toluene with [BiL]/[La-1]/[BnOH] = 300/1/3 achieved 92% conversion after 8 h
(run1, Table1). Theresultingpolymerhada narrower dispersityofÐ = 1.10ascompared
with the TBD-derived PBiL.
The ROP of BiL by the [La-1] + [BnOH] system is well-controlled, as evidenced by the
following four lines of evidence. First, the ROP followed strictly first-order kinetics
(R2 = 0.993) up to high conversions (Figure 3A). Second, the molecular weight
increased linearly with monomer conversion whereas the dispersity of the resulting
PBiL remained low (Ð < 1.1) (Figure 3B). Third, the absolute molecular weight of the
isolated PBiL at 92% conversion was measured by a gel-permeation chromatog-
raphy (GPC) instrument equipped with multi (18)-angle light scattering and dRI (dif-
ferential refractive index) detectors to have a Mn of 13.1 kg molꢂ1; this measured Mn
is close to the calculated Mn of 11.7 kg molꢂ1 based on the [BiL]/[I] ratio, thus giving a
high initiation efficiency of 90% (run 1, Table 1). Furthermore, this measured Mn is
essentially identical to the calculated Mn of 13.9 kg molꢂ1 based on the chain
ends of PBiL characterized by 1H NMR (Figure S5), which also revealed a linear chain
structure and showed high end-group fidelity. Fourth, the BnOH-initiated linear
structure with only the predicted end groups, BnO-[BiL]n–H, was confirmed by ma-
trix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-
TOF MS) by using a low molecular weight sample prepared with [BiL]/[La-1]/
674 Chem 7, 670–685, March 11, 2021

ll
Article
A
B
C
D
1
Figure 3. H NMR (23ꢁC, CDCl3) spectra and thermal properties of PBiL by metal-catalyzed stereoretention ROP
(A) Semilogarithmic kinetic plot of the ROP with [BiL]/[La-1]/[BnOH] = 300/1/3 ([BiL]0 = 6.0 M, toluene, RT).
(B) Plots of Mn and Ð values versus BiL conversion.
(C) MALDI-TOF MS spectrum of the PBiL produced with [BiL]/[La-1]/[BnOH] = 100/1/3.
(D) Plot of m/z values (y) versus the theoretical number of BiL repeat units (x).
[BnOH] = 100/1/3 (Figure 3C). Specifically, the MS spectrum consisted of only one
series of molecular ion peaks, and the spacing between the two neighboring molec-
ular ion peaks was that of the exact molar mass of the repeat unit (m/z = 126.15), as
shown by the slope (126.11) of the linear plot of m/z values (y) versus the number of
BiL repeat units (x) (Figure 3D). The intercept of the plot, 131.0, represents the total
mass of chain ends + that of Na+ [108.14 (BnO/H) + 23.0 (Na+) g molꢂ1].
Pure isotactic polymer from (1R, 5S)-BiL and isotactic polymer from rac-BiL
The above described exclusive stereoretention ROP enabled by the metal-catalyzed
coordination-insertion ROP offers the opportunity to produce perfectly isotactic PBiL
starting from a chiral monomer, (1R, 5S)-6-oxabicyclo[3.2.1]octan-7-one [(1R, 5S)-
BiL], synthesized from commercial reagent (R)-(+)-3-cyclohexenecarboxylic acid.
The ROP with [[(1R, 5S)-BiL]/[La-1]/[BnOH] = 500/1/3 yielded the corresponding chi-
22:4
ral polymer P[(1R, 5S)-BiL] with ½aꢃ589 = –70 deg dmꢂ1gꢂ1 cm3 in CHCl3, Mn = 19.3 kg
molꢂ1, and Ð = 1.08 at 94% conversion. The calculated theoretical Mn based on the
[BiL]/[I] ratio is 21.1 kg/mol, which is close to both the values measured by GPC and by
1H NMR (Mn = 22.5 kg molꢂ1-1) (Figure S7), indicating a near quantitative initiation ef-
ficiency and chain-end fidelity. The resulting polymer is a pure isotactic material, dis-
playing no stereoerrors on its 13C NMR (Pm = 1.00) (Figures 4A and S8). This chiral
Chem 7, 670–685, March 11, 2021 675

ll
Article
A
B
C
Figure 4. Pure isotactic polymer from (1R, 5S)-BiL
13
(A) C NMR (23ꢁC, CDCl3) spectrum of P[(1R, 5S)-BiL] prepared with [[(1R, 5S)-BiL]/[La-1]/[BnOH] = 500/1/3 (6.0 M, toluene, RT).
(B) DSC thermograms of P[(1R, 5S)-BiL] from first heat, cooling, and second heating scans at 10ꢁC minꢂ1
.
(C) pXRD profiles of semi-crystalline P[(1R, 5S)-BiL] (1) and reference amorphous PBiL derived from rac-BiL by La-1 + BnOH (2).
polymer exhibited a remarkably high Tm of 293ꢁC on the DSC first heating scan or
263ꢁC on the second heating scan with a crystallization temperature of 226ꢁC but
no observable Tg (Figure 4B). The semi-crystalline nature of this polymer was further
confirmed by its powder X-ray diffraction (pXRD) profile, featuring sharp diffraction
peaks at 2q values of 17.3ꢁ, 18.3ꢁ, 21.5ꢁ, and 22.5ꢁ, in relation to an atactic, amor-
phous reference PBiL sample produced from rac-BiL by La-1 + BnOH (Figure 4C).
By using La-1 alone without the alcohol co-initiator, a chiral cyclic polymer with an ul-
tra-high Mn of over 1 million Da (1,100 kg molꢂ1, Ð = 1.35) was prepared, which
accordingly exhibited no end groups (Figure S9) and stereoerrors (Figure S10).
When rac-BiL was polymerized by achiral catalyst La-1, only iso-enriched PBiL was pro-
duced with a low isotacticity of Pm = 0.35 (Figure S6). To improve the stereoselectivity of
the ROP of rac-BiL, we next employed discrete chiral yttrium and lanthanum complexes
supported by C2-symmetric salen ligands (Figure 5A) as such catalysts have been shown
to mediate highly stereoselective polymerization of eight-membered rac-dio-
lides.44,45,70 The use of rac-La-2 supported by the sterically demanding salcy ligand
with bulky CPh3 substituents at the 3-positions of the ligand rendered much more active
polymerization system but the isotacticity of PBiL remained essentially the same (runs 2–
4, Table 1). However, the high activity of rac-La-2/BnOH enabled the synthesis of high-
molecular-weight linear PBiL (Mn = 153 kg molꢂ1, Ð = 1.06, run 4, Table 1) by employing
a high ratio of [BiL]/[rac-La-2]/[BnOH] = 1,500/1/1. This PBiL is an amorphous material
exhibiting the highest Tg (135 ꢁC) of the series (Figure 5C).
Moving to a metal center with a smaller ionic radius, rac-Y-1 with 3,5-CMe3 substituents
onthesalcy ligandbroughtabouttheROPwithanearlyperfect initiation efficiency (98%)
and, more significantly, considerably enhanced isotacticity (Pm = 0.4, run 5, Table 1) with
[BiL]/[rac-Y-1]/[BnOH] = 100/1/1. Increasing this ratio to 1,000/1/1 afforded PBiL with
Mn = 59.2 kg molꢂ1, Ð = 1.05, and Pm = 0.52 (run 6, Table 1). As expected, this PBiL is
still anamorphousmaterialwithTg = 128ꢁC. Wealso examined effectsofsolventpolarity
[toluene, tetrahydrofuran (THF), dichloromethane], [BiL]/[rac-Y-1] ratio (50/1 to 300/1),
and [BiL] concentration (3.0 to 12 M) on polymerization characteristics (Table S1). Based
on the results summarized in Table S1, several trends can be observed. Although the
ROP was fastest in THF, the resulting PBiL exhibited the broadest dispersity (Ð up to
1.18). In contrast, the ROP carried out in toluene had a moderate polymerization rate,
but it produced PBiL with the narrowest dispersity. However, the tacticity of the polymer
was not noticeably affected by the above variations.
676 Chem 7, 670–685, March 11, 2021

ll
Article
A
C
B
D
Figure 5. Iso-rich and isotactic polymers from rac-BiL
(A) Chemical structures of the catalysts employed in this study.
(B) 13C NMR (25ꢁC, CDCl3) spectra in the carbonyl region of iso-rich (1, 2) and isotactic (3, 4) PBiL samples by racemic and enantiopure Y-3 catalysts in
combination with achiral and chiral alcohols: (1) Pm = 0.42 (run 8, Table 1); (2) Pm = 0.74 (run 10, Table 1); (3) Pm = 0.85 (run 11, Table 1); and (4) Pm = 0.87
(run 12, Table 1).
(C) DSC thermogram of the iso-rich, amorphous PBiL (Mn = 153 kg molꢂ1, Ð = 1.06) by rac-La-2.
(D) DSC thermograms of PBiL materials with varied isotacticity by racemic and enantiopure Y-3 catalysts: (1) Pm = 0.42; (2) Pm = 0.74; (3) Pm = 0.85; (4) Pm
=
0.87.
Further increasing the ligand steric bulk by introducing bulkier 3,5-CMe2Ph substituents
(rac-Y-2) and 3-CPh3-5-Me substituents (rac-Y-3) to the salcy ligand did not further in-
crease the isotacticity (runs 7 and 8, Table 1). Replacing the BnOH co-initiator with chiral
alcohol, (R)-(+)-1-phenylethanol, as initiator also did not noticeably affect the tacticity
(run 9, Table 1). These results suggest the large number of stereoerrors generated by
these racemic catalysts could be originated from polymer chain exchange71 between
two antipodal catalyst centers (vide infra). To test this hypothesis, we employed enantio-
meric pure catalyst (S,S)-Y-3 and indeed observed a substantial increase in the polymer
isotacticity to Pm = 0.74 (run 10, Table 1). When this enantiomeric catalyst was combined
with a chiral alcohol co-initiator, (R)-(+)-1-phenylethanol or (S)-(ꢂ)-1-phenylethanol, the
isotacticity was further increased to Pm = 0.85 (run 11, Table 1) or 0.87 (run 12, Table 1).
Noteworthy alsois that this high tacticityachievedwas accompanied by a perfect (100%)
or near-perfect (106%, indicative of some chain transfer events) initiation efficiency. With
this high level of isotacticity, the PBiL materials derived from the stereoselective poly-
merization of rac-BiL by the chiral catalyst/initiator system finally became semi-crystal-
line, as evidenced by their endothermic first-order melting transitions with high Tm
values at 150ꢁC and 154ꢁC (Figure 5D).
Mechanisms of stereoselection and stereoerror generation
To shed light on the mechanism of stereoselection by the enantiopure catalyst and the
generation of stereoerrors by the racemic catalyst, density functional theory (DFT)
calculations were carried out first for the polymerization of BiL as a racemic mixture
Chem 7, 670–685, March 11, 2021 677

ll
Article
A
C
B
Figure 6. Summary of computational results
(A) Polymerization pathways and energies for the polymerization of rac-BiL with enantiopure (S,S)-Y-3.
(B) Polymeric chain exchange pathway proposed for the polymerization of rac-BiL with rac-Y-3 (TS = transition state).
(C) Optimized geometry of the transition state for the dimer formation along the polymer chain exchange pathway. Atoms reported as follows: Y, cyan;
O, red; N, blue; C, gray; and H, white. All free energies reported in toluene and kcal/mol.
with (S,S)-Y-3. At first, we investigated the relative stability of (S,S)-Y-3 bearing a (S,R)-
chain or a (R,S)-chain, namely (S,S)-Y-3/(S,R)-chain or (S,S)-Y-3/(R,S)-chain propagating
sites, and found 1.5 kcal/mol in favor of the (S,S)-Y-3/(R,S)-chain site, indicating a mod-
erate thermodynamic preference for the reactivity of (R,S)-BiL enantiomer with (S,S)-Y-3
catalyst. Next, we investigated the kinetics of BiL insertion into the Y–O chain bond to
examine the effect of the monomer chirality on the interaction with the chain end (Fig-
ure 6A). Starting from the more stable (S,S)-Y-3/(R,S)-chain propagating sites, insertion
of the same enantiomer (R,S)-BiL is preferred over the insertion of the opposite enan-
tiomer (S,R)-BiL by a DDGz of 1.5 kcal/mol—activation energies for (R,S)-BiL and (S,R)-
BiL insertions are 20.5 and 22.0 kcal/mol, respectively—in good agreement with the
experimental Pm of 0.85. When a stereoerror occurs and a (S,R)-chain is formed, the re-
sulting (S,S)-Y-3/(S,R)-chain site then inserts preferentially the (S,R)-BiL monomer over
the (R,S)-BiL one by a DDGz of 1.2 kcal/mol: activation energies for (R,S)-BiL and (S,R)-
BiL insertions are 19.1 and 17.9 kcal/mol, respectively. Overall, these results show
that, with an enantiopure catalyst, the polymerization of rac-BiL is moderately stereose-
lective through a combined catalyst-site and chain-end control mechanism, where the
catalyst chirality selects the chirality of the chain and, in turn, the chain chirality selects
the chirality of the incoming monomer, resulting in a stereoblock-like polymer structure.
Any possible polymer chain exchange (vide infra) between the (S,S)-Y-3/(R,S)-chain and
(S,S)-Y-3/(S,R)-chain sites would have no additional consequences on the polymer mi-
crostructures. As switching the chirality of the initiating alcohol, (R or S)-Me(Ph)CHOH,
was also found to not affect, noticeably, the tacticity of the resulting PBiL (run 11 versus
12, Table 1), we computed the transition state for insertion of both BiL enantiomers into
the (S,S)-Y–OCH(Ph)Me bond. The reaction with (S,S)-Y-(S)-OCH(Ph)Me was favored by
ꢀ3.5 kcal/more over the reaction with (S,S)-Y-(R)-OCH(Ph)Me, but this difference is the
same for both BiL enantiomers, thus confirming that the initiation step does not select
preferentially either enantiomer of the monomer.
678 Chem 7, 670–685, March 11, 2021

ll
Article
To verify the computational result that the chiral catalyst (S,S)-Y-3 selectively poly-
merizes (R,S)-BiL over its enantiomer (S,R)-BiL, we examined the kinetic profile of
the ROP of rac-BiL catalyzed by (S,S)-Y-3 until it reached to around 50% monomer
conversion. The polymerization strictly follows the first-order kinetics (Figure S53).
The specific rotation of the residual monomer when the ROP was quenched at
49% conversion was measured to be +19 deg dmꢂ1gꢂ1cm3 in CHCl3. When
22:4
compared with the specific rotation of (R,S)-BiL, ½aꢃ589 = –31 deg dmꢂ1gꢂ1cm3 in
CHCl3, it can be concluded that the residual monomer pool was (S,R)-BiL enriched.
Overall, the experimental result agrees with that of the DFT calculation.
Moving to the polymerization by a racemic catalyst mixture, rac-Y-3, possible poly-
mer chain exchange between two antipodal catalyst centers carrying opposite
chains must now be taken into account. For the exchange to happen without free
alcohol, we propose here a mechanism that involves an intermolecular transesterifi-
cation step between the two enantiomers of the catalyst exchanging chains of oppo-
site chiralities, through the formation of a dimeric intermediate (Figure 6B). The two
anionic chain ends act as bridges between the two propagating species while main-
taining the favored seven-coordinate ligand environment typical of such Y catalysts.
Out of all the possible dimeric species that could be formed, we focused on the one
having opposite enantiomers of Y-3 and opposite chiral chains, as it is the one that
imposes stereomicrostructure consequences in the stereoselective polymerization
as a result of polymer chain exchange. Interestingly, this proposed dimeric interme-
diate (Figure 6B) was found to be more stable than the relative monomeric
propagating species, by 2.8 kcal/mol if it dissociates as (S,S)-Y-3/(R,S)-chain and
(R,R)-Y-3/(S,R)-chain, or by 5.8 kcal/mol if it dissociates as (S,S)-Y-3/(S,R)-chain and
(R,R)-Y-3/(R,S)-chain. The stability of the dinuclear species could be attributed to
˚
possible stabilizing Y–Y interaction (3.8 A), as compared with the reported Y–Y dis-
tances of 3.6–3.7 A,72–74 and m-alkoxy bridging (Figure S54), which is consistent with
˚
the observations that yttrium salen complexes are favored to form dinuclear struc-
tures in non-coordinating solvent (e.g., toluene) via alkoxy bridging.71,75,76 This anal-
ysis suggests that the active site environment is mostly populated as dimers when
the majority of the monomer is consumed. In this context, we have summarized
the DFT results in Figure 6B by setting this dimeric species as a reference structure
at 0 kcal/mol; the transition state for the dimer formation (Figure 6C) was located at
22 kcal/mol above that reference point.
Dissociation of the dimer toward the formation of higher energy monomeric (S,S)-Y-
3/(S,R)-chain and (R,R)-Y-3/(R,S)-chain sites leads to more reactive but less stereose-
lective sites than the dissociation toward the lower energy monomeric (S,S)-Y-3/
(R,S)-chain and (R,R)-Y-3/(S,R)-chain sites, which are less reactive but more stereose-
lective. Specifically, the transition states for the former combination are lower in en-
ergy (17.9 and 19.1 kcal/mol) and less stereoselective (DDGz = 1.2 kcal/mol) than the
latter combination, with the transition states located at higher energies (22.0 and
20.5 kcal/mol) and being more stereoselective (DDGz = 1.5 kcal/mol). Thus, in the
case of polymerization by rac-Y-3, polymer chain exchange via dimeric intermedi-
ates represents an additional source of stereoerrors, as the process converts two
slower and more selective propagating sites into two faster and less selective sites,
thus reducing the overall stereoselectivity.
Overall, in the presence of rac-Y-3, access to the less stereoselective propagating
species can occur either by polymer chain exchange or by inherent addition/inser-
tion errors, accounting for the additional generation of stereoerrors to produce poly-
mers with a low Pm of ꢀ0.45. Hence, understanding of this mechanism provides a
Chem 7, 670–685, March 11, 2021 679

ll
Article
straightforward way of enhancing the stereoselectivity by employing an enantio-
meric pure catalyst, under which conditions the additional stereoerrors originated
from the polymer exchange pathway are eliminated, thereby producing polymers
with much higher stereoregularity (vide supra).
Thermal and mechanical properties and chemical recyclability
The most marked effect of structural hybridization of the LCT g-BL with the HCT ε-CL
that led to the design of the hybrid monomer BiL has been observed on thermal
property enhancements. Specifically, the PBiL material derived from the [5+7] hybrid
monomer BiL displays a high Tg ranging from 119ꢁC to 135ꢁC (Figures 2C and 5C),
thus representing up to 200ꢁC enhancements in Tg in relation to PGBL and PCL ho-
mopolymers as well as their copolymers. The Tm of the pure isotactic P[(1R, 5S)-BiL]
was determined to be 263ꢁC (Figure 4B), also reflecting a ꢀ200ꢁC increase in rela-
tion to the Tm values of PGBL and PCL as well as their copolymers. The remarkable
enhancement trend can be extended to the thermal stability with PBiL (with 18%
trans) exhibiting a high Td,5% of 318ꢁC and a high Tmax of 354ꢁC (Figure 2D), which
are about 116ꢁC and 136ꢁC higher than the Td,5% and Tmax values of PGBL. The
Td,5% of all cis-PBiL with Mn = 153 kg molꢂ1 prepared by rac-La-2 (run 4, Table 1)
showed a further increased Td,5% to 336ꢁC (Figure 7A), approaching to that of PCL.
Mechanical properties of two PBiL samples, one with all cis-chains prepared by rac-
La-2 (Mn = 94.3 kg molꢂ1) and the other with cis (82%) and trans (18%) prepared by
TBD (Mn = 56.7 kg molꢂ1), were examined by tensile testing of their dog-bone-
shaped specimens. The all-cis-PBiL material is best described as a hard, strong,
and brittle glass, as characterized by a high Young’s modulus (E) = 3.05 G 0.18
GPa, a high ultimate tensile strength (sB) = 53.6 G 0.3 MPa, and a low elongation
(εB) = 2.2% (Figure 7B). As compared with PGBL (Mn = 42.5 kg molꢂ1),77 cis-PBiL
enhanced Young’s modulus by ꢀ103 and tensile strength by ꢀ33 but with signifi-
cantly reduced ductility. A similar trend was found when compared with PCL,78
observing enhanced Young’s modulus by >103 and tensile strength by ꢀ1.53. In
contrast, the PBiL material with 18% trans structure is a plastic, characterized by a
yield point, a lower (but still high) E (2.00 G 0.10 GPa), a lower sB (18.9 G 3.3
MPa), and a higher ductility (εB = 85.8% G 23.2%). Overall, these results showed
that all-cis-PBiL is drastically different than its hybridizing parents PGBL and PCL
with much higher modulus but much lower ductility, and the results also demon-
strated the mechanical properties of PBiL materials can be tuned by introducing
some trans chains via epimerization.
Another hypothesized unique feature of the HCT/LCT hybrid monomer is that it could
exhibit both high polymerizability by virtue of the HCT structural motif and high de-
polymerizability via the LCT structural fragment, thereby unifying otherwise conflict-
ing properties in a single monomer. To quantify the polymerizability of BiL and the
depolymerizability of PBiL as a function of reaction conditions, the thermodynamics
of the BiL polymerization were probed by using [M]/[La-1]/[BnOH] = 100/1/3 (M = BiL)
and [M]0 = 0.5 mol Lꢂ1 in toluene-d8 via a variable-temperature NMR study. The equi-
librium monomer concentration, [M]eq, obtained by plotting [M]t as a function time
until [M] became constant, was measured to be 0.293, 0.216, 0.171, 0.127, and
0.0935 mol Lꢂ1 for 50, 40, 30, 20ꢁC, and 10ꢁC, respectively. The van’t Hoff plot of
ln[M]eq versus 1/T gave a straight line (R2 = 0.999, Figures S45 and S46), from which
the thermodynamic parameters were calculated to be DHop = –21.1 kJ∙molꢂ1 and
DSpo = –55.8 J∙molꢂ1∙Kꢂ1, based on the equation ln[M]eq = DHpo/RT – DSpo/R.53 The
Tc value was calculated to be 575 K (302ꢁC) at [M]0 = 10.0 mol Lꢂ1 or 379 K (106ꢁC)
at [M]0 = 1.0 mol Lꢂ1, based on the equation Tc =DHop /{DSpo{+ Rln[M]0}.53 These results
680 Chem 7, 670–685, March 11, 2021

ll
Article
A
C
B
Figure 7. Thermal and mechanical properties and chemical recyclability of PBiL
(A) TGA and DTG thermograms of PBiL (Mn = 153 kg molꢂ1, Ð = 1.06) prepared by rac-La-2.
(B) Stress-strain curves for the PBiL prepared by rac-La-2 (Mn = 94.3 kg molꢂ1, all cis structure, black curve) and the PBiL prepared by TBD (Mn = 56.7 kg
molꢂ1, cis (82%) and trans (18%) structures, blue curve). Inset: blowup region from 0% to 10% elongation.
(C) 1H NMR (toluene-d8) spectra of PBiL (Mn = 13.1 kg molꢂ1) prepared by La-1 at RT (1); the recovered monomer BiL obtained after depolymerization by
La-1 at 120ꢁC (2); and the starting monomer BiL for comparison (3).
indicate both good polymerizability of BiL and depolymerizability of PBiL by adjust-
ing the reaction conditions (temperature and concentration).
Consistent with the above derived thermodynamic parameters, the polymerization
results summarized in Table 1 showed the relatively high polymerizability of BiL,
achieving a high monomer conversion up to 93% at RT and 6.0 M of initial BiL concen-
tration. Next, we probed the depolymerizability of the preformed PBiL by La-1 at RT
by using the same catalyst at higher temperatures. Specifically, the depolymerization
of PBiL in toluene-d8 was monitored in situ by 1H NMR, revealing that the depolymer-
ization achieved near completion (>95%) after heating at 120 ꢁC for 36 h to regen-
erate the monomer BiL (Figure 7C). It is worth noting that, when the depolymerization
of PBiL prepared by the TBD-catalyzed ROP at RT (Mn = 56.7 kg molꢂ1, 82% cis + 18%
trans) was performed by using the same basic TBD catalyst at 120ꢁC, the PBiL under-
went further epimerization, whereas only a small amount of BiL was regenerated after
1 h. Nevertheless, when the depolymerization was extended to 12 h, all PBiL,
including both cis and trans chains, was completely depolymerized to cleanly recover
BiL (Figure S47). This exclusive selectivity for the BiL monomer recovery is possible,
despite the fact that base-mediated epimerization can inter-convert the cis and trans
chains, because the bridged bicyclic monomer BiL exists only in the cis-configuration,
which is fixed by the spatially atom-bridged structure, and the monomer is stable un-
der such base condition (in the absence of initiator). Therefore, PBiL can be
completely depolymerized and converted into the most thermodynamically favored
state—the monomer state under suitable conditions.
Chem 7, 670–685, March 11, 2021 681

ll
Article
DISCUSSION
The design of useful circular polymers requires building-block monomer structures
that can balance conflicting monomer polymerizability, polymer depolymerizability,
and polymer performance properties. The herein described HCT/LCT hybrid mono-
mer design synergistically utilizes both the high polymerizability and performance
properties of the HCT sub-structural unit and the high depolymerizability of the
LCT sub-structural unit, thereby unifying the conflicting properties in a single
monomer structure. The mechanism of such hybrid monomers to solve the property
trade-offs can be thought of as the intriguing two-pathway scenario wherein the
polymerization proceeded via the ring-opening of the HCT sub-structure and the
depolymerization occurred through the ring-closing to form the LCT sub-structure.
Even more intriguingly, the thermal and mechanical properties of the polymers
derived from the hybridized monomers can be far superior to those homo- or copol-
ymers of the parent monomers. This marked property enhancement has been
demonstrated by the prototype example of the hybrid monomer BiL consisting of
both HCT ε-CL and LCT g-BL. Specifically on thermal properties, PBiL exhibits a
high Tg up to 135ꢁC and Tm up to 263ꢁC, which are ꢀ200ꢁC above both the Tg
and Tm values of the parent PGBL and PCL as well as their copolymers. On mechan-
ical properties, PBiL also showed a marked enhancement in Young’s modulus and
tensile strength, displaying ꢀ103 higher Young’s modulus and ꢀ33 higher tensile
strength relative to PGBL and >103 higher modulus and ꢀ1.53 higher tensile
strength relative to PCL. The HCT/LCT hybrid monomer strategy, as a powerful
approach for circular polymer design, which has been demonstrated by the results
reported here, could be broadly applicable to designing other monomers that can
address property trade-offs within a single monomer structure.
The (de)polymerization chemistry of analogous bicyclic lactones and thiolactones, as
well as the properties of their derived corresponding polyesters and polythioesters,
can be significantly different. For example, bicyclic thioester [221]BTL produces crys-
talline and fully recyclable polythioester PBTL with the unique tacticity-independent
crystallinity,42 whereas the bicyclic lactone congener gives amorphous and non-
recyclable polyester—basically the opposite of PBTL. Three factors count for these
large differences. First, with the larger S atom (versus O), the thioester bond is longer
than the ester bond, which, together with differences in bond angles, imparts differ-
ences in ring strain. Second, the polarity of the carbonyl group in the thioester bond
is higher than that in the ester bond, effectively increasing the interactions between
polymer chains and, thus, leading to higher Tm values of the polythioester. Third, the
conformation of the in-chain cyclopentane ring is also important for PBTL’s tolerance
of crystallinity toward stereodefects. The current polyester derived from BiL pos-
sesses neither the thioester bond nor the in-chain cyclopentane ring, explaining
why no unusual tacticity-independent crystallinity was observed.
Computational studies have traced the origin of stereoerrors generated from the stereo-
selective polymerization of racemic BiL by both enantiopure and racemic catalysts. The
derived polymer chain exchange mechanism involves the formation of the surprisingly
stable dimeric propagating intermediate between the two enantiomers of the catalyst
sites, which provides a pathway for exchanging the chains of opposite chiralities. Disso-
ciation of the dimeric intermediate leads to monomeric propagating sites with more
reactive but less stereoselective and less reactive but more stereoselective site combina-
tions. With this mechanism understood, an effective strategy to eliminate such polymer-
exchange-induced stereoerrors has been realized by employing a single enantiomer of
the catalyst, which significantly enhanced the stereoselectivity of the polymerization.
682 Chem 7, 670–685, March 11, 2021

ll
Article
EXPERIMENTAL PROCEDURES
Resource availability
Lead contact
Requests for further information should be directed to and will be fulfilled by the
Lead Contact, Eugene Y.-X. Chen (eugene.chen@colostate.edu).
Materials availability
This study did not generate new unique materials.
Data and code availability
All data needed to evaluate the conclusions in the paper are present in the paper
and/or the Supplemental information.
SUPPLEMENTAL INFORMATION
Supplemental information can be found online at https://doi.org/10.1016/j.chempr.
2021.02.003.
ACKNOWLEDGMENTS
This work was supported in part by Colorado State University (CSU) and by the U.S.
Department of Energy, Office of Energy Efficiency and Renewable Energy,
Advanced Manufacturing Office (AMO), and Bioenergy Technologies Office
(BETO). This work was performed as part of the BOTTLE Consortium, which includes
members from CSU, and funded under contract number DE-AC36-08GO28308 with
the National Renewable Energy Laboratory, operated by Alliance for Sustainable En-
ergy. The computational study used the resources of the King Abdullah University of
Science and Technology Super-Computing Laboratory (KSL).
AUTHOR CONTRIBUTIONS
C.S. and E.Y.-X.C. conceived the idea and designed the experiments. C.S. carried
out the experiments, and L.F. and L. Cavallo performed and analyzed the DFT calcu-
lations. C.S., L.F., and E.Y.-X.C. co-wrote the manuscript, and all authors partici-
pated in data analysis and discussions and read and edited the manuscript. E.Y.-
X.C. directed the project.
DECLARATION OF INTERESTS
E.Y.-X.C. and C.S. are inventors on a U.S. provisional application submitted by Col-
orado State University Research Foundation, which covers the herein described
polymers. All other authors declare no competing interests.
Received: October 23, 2020
Revised: December 23, 2020
Accepted: February 1, 2021
Published: March 11, 2021
REFERENCES
1. Jambeck, J.R., Geyer, R., Wilcox, C., Siegler,
T.R., Perryman, M., Andrady, A., Narayan, R.,
and Law, K.L. (2015). Marine pollution. Plastic
waste inputs from land into the ocean. Science
347, 768–771.
3. Lamb, J.B., Willis, B.L., Fiorenza, E.A., Couch,
C.S., Howard, R., Rader, D.N., True, J.D., Kelly,
L.A., Ahmad, A., Jompa, J., and Harvell, C.D.
(2018). Plastic waste associated with disease on
coral reefs. Science 359, 460–462.
McKinsey & company). https://www.
ellenmacarthurfoundation.org/publications/
the-new-plastics-economy-rethinking-the-
future-of-plastics.
5. US Department of Energy (2020). Report of the
basic energy sciences roundtable on chemical
upcycling of polymers, April 30–May 1, 2019.
https://science.osti.gov/-/media/bes/pdf/
2. Geyer, R., Jambeck, J.R., and Law, K.L. (2017).
Production, use, and fate of all plastics ever
made. Sci. Adv. 3, e1700782.
4. World Economic Forum (2016). The new.
plastics economy: rethinking the future of
plastics (Ellen MacArthur foundation and
Chem 7, 670–685, March 11, 2021 683

ll
Article
reports/2020/Chemical_Upcycling_Polymers.
pdf.
22. Zhu, J.B., and Chen, E.Y.-X. (2019). Catalyst-
sidearm-induced stereoselectivity switching in
polymerization of a racemic lactone for
stereocomplexed crystalline polymer with a
circular life cycle. Angew. Chem. Int. Ed. Engl.
58, 1178–1182.
approach to sustainability. Green Chem 19,
3692–3706.
6. National Academies of Sciences, Engineering,
and Medicine (2020). Closing the loop on the
plastics dilemma proceedings of the a
workshop in brief (The National Academies
Press). https://doi.org/10.17226/25647.
35. Hong, M., and Chen, E.Y.-X. (2019). Future
directions for sustainable polymers. Trends in
Chemistry 1, 148–151.
36. Coates, G.W., and Getzler, Y.D.Y.L. (2020).
Chemical recycling to monomer for an ideal,
circular polymer economy. Nat. Rev. Mater. 5,
501–516.
23. Shi, C.-X., Guo, Y.-T., Wu, Y.-H., Li, Z.-Y., Wang,
Y.-Z., Du, F.-S., and Li, Z.-C. (2019). Synthesis
and controlled organobase-catalyzed ring-
opening polymerization of morpholine-2,5-
dione derivatives and monomer recovery by
acid-catalyzed degradation of the polymers.
Macromolecules 52, 4260–4269.
7. Scheutz, G.M., Lessard, J.J., Sims, M.B., and
Sumerlin, B.S. (2019). Adaptable crosslinks in
polymeric materials: resolving the intersection
of thermoplastics and thermosets. J. Am.
Chem. Soc. 141, 16181–16196.
37. Olse´ n, P., Odelius, K., and Albertsson, A.C.
(2016). Thermodynamic presynthetic
considerations for ring-opening
polymerization. Biomacromolecules 17,
699–709.
8. Garcia, J.M., and Robertson, M.L. (2017). The
future of plastics recycling. Science 358,
870–872.
24. MacDonald, J.P., and Shaver, M.P. (2016). An
aromatic/aliphatic polyester prepared via ring-
opening polymerisation and its remarkably
selective and cyclable depolymerisation to
monomer. Polym. Chem. 7, 553–559.
38. Schneiderman, D.K., and Hillmyer, M.A. (2016).
Aliphatic polyester block polymer design.
Macromolecules 49, 2419–2428.
9. Hillmyer, M.A. (2017). The promise of plastics
from plants. Science 358, 868–870.
25. Yuan, J., Xiong, W., Zhou, X., Zhang, Y., Shi, D.,
Li, Z., and Lu, H. (2019). 4-Hydroxyproline-
derived sustainable polythioesters: controlled
ring-opening polymerization, complete
recyclability, and facile functionalization. J. Am.
Chem. Soc. 141, 4928–4935.
10. Zhu, Y., Romain, C., and Williams, C.K. (2016).
Sustainable polymers from renewable
resources. Nature 540, 354–362.
39. Hong, M., and Chen, E.Y.X. (2016). Completely
recyclable biopolymers with linear and cyclic
topologies via ring-opening polymerization of
g-butyrolactone. Nat. Chem. 8, 42–49.
11. Zhang, X., Fevre, M., Jones, G.O., and
Waymouth, R.M. (2018). Catalysis as an
enabling science for sustainable polymers.
Chem. Rev. 118, 839–885.
40. Hong, M., and Chen, E.Y.-X. (2016). Towards
truly sustainable polymers: a metal-free
recyclable polyester from biorenewable non-
strained g-butyrolactone. Angew. Chem. Int.
Ed. Engl. 55, 4188–4193.
26. Lloyd, E.M., Lopez Hernandez, H., Feinberg,
E.C., Yourdkhani, M., Zen, E.K., Mejia, E.B.,
Sottos, N.R., Moore, J.S., and White, S.R.
(2019). Fully recyclable metastable polymers
and composites. Chem. Mater. 31, 398–406.
12. Rorrer, N.A., Nicholson, S., Carpenter, A.,
Biddy, M.J., Grundl, N.J., and Beckham, G.T.
(2019). Combining reclaimed PET with bio-
based monomers enables plastics upcycling.
Joule 3, 1006–1027.
41. Sangroniz, A., Zhu, J.B., Tang, X., Etxeberria,
A., Chen, E.Y.X., and Sardon, H. (2019).
Packaging materials with desired mechanical
and barrier properties and full chemical
recyclability. Nat. Commun. 10, 3559.
27. Kaitz, J.A., Lee, O.P., and Moore, J.S. (2015).
Depolymerizable polymers: preparation,
applications, and future outlook. MRS
Commun 5, 191–204.
13. Yoshida, S., Hiraga, K., Takehana, T., Taniguchi,
I., Yamaji, H., Maeda, Y., Toyohara, K.,
Miyamoto, K., Kimura, Y., and Oda, K. (2016). A
bacterium that degrades and assimilates
poly(ethylene terephthalate). Science 351,
1196–1199.
42. Shi, C., McGraw, M.L., Li, Z.C., Cavallo, L.,
Falivene, L., and Chen, E.Y.-X. (2020). High-
performance pan-tactic polythioesters with
intrinsic crystallinity and chemical recyclability.
Sci. Adv. 6, eabc0495.
28. Christensen, P.R., Scheuermann, A.M., Loeffler,
K.E., and Helms, B.A. (2019). Closed-loop
recycling of plastics enabled by dynamic
covalent diketoenamine bonds. Nat. Chem. 11,
442–448.
14. Tournier, V., Topham, C.M., Gilles, A., David,
B., Folgoas, C., Moya-Leclair, E., Kamionka, E.,
Desrousseaux, M.L., Texier, H., Gavalda, S.,
et al. (2020). An engineered PET depolymerase
to break down and recycle plastic bottles.
Nature 580, 216–219.
43. Xiong, W., Chang, W., Shi, D., Yang, L., Tian, Z.,
Wang, H., Zhang, Z., Zhou, X., Chen, E.-Q., and
Lu, H. (2020). Geminal dimethyl substitution
enables controlled polymerization of
penicillamine-derived b-thiolactones and
reversed depolymerization. Chem 6, 1831–
1843.
29. Garcıa, J.M., Jones, G.O., Virwani, K.,
´
McCloskey, B.D., Boday, D.J., ter Huurne,
G.M., Horn, H.W., Coady, D.J., Bintaleb, A.M.,
Alabdulrahman, A.M.S., et al. (2014).
Recyclable, strong thermosets and organogels
via paraformaldehyde condensation with
diamines. Science 344, 732–735.
15. Jia, X., Qin, C., Friedberger, T., Guan, Z., and
Huang, Z. (2016). Efficient and selective
degradation of polyethylenes into liquid fuels
and waxes under mild conditions. Sci. Adv. 2,
e1501591.
44. Tang, X., Westlie, A.H., Watson, E.M., and
Chen, E.Y.-X. (2019). Stereosequenced
crystalline polyhydroxyalkanoates from
diastereomeric monomer mixtures. Science
366, 754–758.
30. Liu, Y., Zhou, H., Guo, J.Z., Ren, W.M., and Lu,
X.B. (2017). Completely recyclable monomers
and polycarbonate: approach to sustainable
polymers. Angew. Chem. Int. Ed. Engl. 56,
4862–4866.
16. Rahimi, A., and Garcıa, J.M. (2017). Chemical
´
recycling of waste plastics for new materials
production. Nat. Rev. Chem. 1, 0046.
45. Tang, X., Westlie, A.H., Caporaso, L., Cavallo,
L., Falivene, L., and Chen, E.Y.-X. (2020).
Biodegradable polyhydroxyalkanoates by
stereoselective copolymerization of racemic
diolides: stereocontrol and polyolefin-like
properties. Angew. Chem. Int. Ed. Engl. 59,
7881–7890.
31. Lizundia, E., Makwana, V.A., Larran˜ aga, A.,
Vilas, J.L., and Shaver, M.P. (2017). Thermal,
structural and degradation properties of an
aromatic–aliphatic polyester built through ring-
opening polymerisation. Polym. Chem. 8,
3530–3538.
17. Schneiderman, D.K., and Hillmyer, M.A. (2017).
50th anniversary perspective: there is a great
future in sustainable polymers.
Macromolecules 50, 3733–3749.
18. Lu, X.B., Liu, Y., and Zhou, H. (2018). Learning
nature: recyclable monomers and polymers.
Chemistry 24, 11255–11266.
32. Schneiderman, D.K., Vanderlaan, M.E.,
Mannion, A.M., Panthani, T.R., Batiste, D.C.,
Wang, J.Z., Bates, F.S., Macosko, C.W., and
Hillmyer, M.A. (2016). Chemically recyclable
biobased polyurethanes. ACS Macro Lett 5,
515–518.
46. Xu, Y.-C., Ren, W.-M., Zhou, H., Gu, G.-G., and
Lu, X.-B. (2017). Functionalized polyesters with
tunable degradability prepared by controlled
ring-opening (co)polymerization of lactones.
Macromolecules 50, 3131–3142.
19. Jehanno, C., Pe´ rez-Madrigal, M.M.,
Demarteau, J., Sardon, H., and Dove, A.P.
(2019). Organocatalysis for depolymerisation.
Polym. Chem. 10, 172–186.
47. Wilson, J.A., Hopkins, S.A., Wright, P.M., and
Dove, A.P. (2015). Synthesis of u-
pentadecalactone copolymers with
independently tunable thermal and
degradation behavior. Macromolecules 48,
950–958.
33. Diesendruck, C.E., Peterson, G.I., Kulik, H.J.,
Kaitz, J.A., Mar, B.D., May, P.A., White, S.R.,
Martı´nez, T.J., Boydston, A.J., and Moore, J.S.
(2014). Mechanically triggered heterolytic
unzipping of a low-ceiling-temperature
polymer. Nat. Chem. 6, 623–628.
20. Tang, X., and Chen, E.Y.-X. (2019). Toward
infinitely recyclable plastics derived from
renewable cyclic esters. Chem 5, 284–312.
21. Zhu, J.B., Watson, E.M., Tang, J., and Chen,
E.Y.-X. (2018). A synthetic polymer system with
repeatable chemical recyclability. Science 360, 34. Hong, M., and Chen, E.Y.-X. (2017). Chemically
48. Hakkarainen, M., Hoglund, A., Odelius, K., and
¨
398–403.
recyclable polymers: a circular economy
Albertsson, A.C. (2007). Tuning the release rate
684 Chem 7, 670–685, March 11, 2021

ll
Article
of acidic degradation products through
based copolyesters: copolymerization
characteristics and composition-dependent
(thermal, eutectic, cocrystallization, and
degradation) properties. Macromolecules 50,
8469–8479.
69. Zhang, X., Jones, G.O., Hedrick, J.L., and
Waymouth, R.M. (2016). Fast and selective ring-
opening polymerizations by alkoxides and
thioureas. Nat. Chem. 8, 1047–1053.
macromolecular design of caprolactone-based
copolymers. J. Am. Chem. Soc. 129, 6308–
6312.
49. Tabata, Y., and Abe, H. (2014). Synthesis and
Properties of alternating copolymers of 3-
hydroxybutyrate and lactate units with different
stereocompositions. Macromolecules 47,
7354–7361.
70. Tang, X., and Chen, E.Y.X. (2018). Chemical
synthesis of perfectly isotactic and high melting
bacterial poly(3-hydroxybutyrate) from bio-
sourced racemic cyclic diolide. Nat. Commun.
9, 2345.
59. Shieh, P., Zhang, W., Husted, K.E.L., Kristufek,
S.L., Xiong, B., Lundberg, D.J., Lem, J., Veysset,
D., Sun, Y., Nelson, K.A., et al. (2020). Cleavable
comonomers enable degradable, recyclable
thermoset plastics. Nature 583, 542–547.
50. Wilson, J.A., Hopkins, S.A., Wright, P.M., and
Dove, A.P. (2016). Dependence of copolymer
sequencing based on lactone ring size and
ε-substitution. ACS Macro Lett 5, 346–350.
71. Ovitt, T.M., and Coates, G.W. (2002).
Stereochemistry of lactide polymerization with
chiral catalysts: new opportunities for
60. Yu, X., Jia, J., Xu, S., Lao, K.U., Sanford, M.J.,
Ramakrishnan, R.K., Nazarenko, S.I., Hoye, T.R.,
Coates, G.W., and DiStasio, R.A. (2018).
Unraveling substituent effects on the glass
transition temperatures of biorenewable
polyesters. Nat. Commun. 9, 2880.
stereocontrol using polymer exchange
mechanisms. J. Am. Chem. Soc. 124, 1316–1326.
51. Wilson, J.A., Hopkins, S.A., Wright, P.M., and
Dove, A.P. (2015). Synthesis and
72. Pan, C., Shen, W., Yang, L., Bao, L., Wei, Z., Jin,
P., Fang, H., Xie, Y., Akasaka, T., and Lu, X.
(2019). Crystallographic characterization of
Y2C2n (2n = 82, 88–94): direct Y–Y bonding and
cage-dependent cluster evolution. Chem. Sci.
10, 4707–4713.
postpolymerization modification of one-pot u-
pentadecalactone block-like copolymers.
Biomacromolecules 16, 3191–3200.
61. Xu, Y., ucu, T., Perry, M.R., and Shaver, M.P.
(2020). Alicyclic polyesters from a bicyclic 1,3-
dioxane-4-one. Polym. Chem. 11, 4928–4932.
52. Labet, M., and Thielemans, W. (2009). Synthesis
of polycaprolactone: a review. Chem. Soc. Rev.
38, 3484–3504.
62. Kamber, N.E., Jeong, W., Waymouth, R.M.,
Pratt, R.C., Lohmeijer, B.G.G., and Hedrick, J.L.
(2007). Organocatalytic ring-opening
73. Popov, A.A., Avdoshenko, S.M., Penda´s, A.M.,
and Dunsch, L. (2012). Bonding between
strongly repulsive metal atoms: an oxymoron
made real in a confined space of endohedral
metallofullerenes. Chem. Commun. 48, 8031–
8050.
53. Duda, A., and Kowalski, A. (2009).
Thermodynamics and kinetics of ring-opening
polymerization. In Handbook of Ring-Opening
Polymerization, P. Dubois, O. Coulembier, and
J.-M. Raquez, eds. (Wiley-VCH Verlag),
pp. 1–51.
polymerization. Chem. Rev. 107, 5813–5840.
63. Kiesewetter, M.K., Shin, E.J., Hedrick, J.L., and
Waymouth, R.M. (2010). Organocatalysis:
opportunities and challenges for polymer
synthesis. Macromolecules 43, 2093–2107.
74. Xu, L., Xia, C.-J., Wang, L.-F., Xie, L., Wang, B.,
Zhang, Y.-F., and Huang, X. (2014). Structural
evolution, sequential oxidation and chemical
bonding in tri-yttrium oxide clusters: Y3Oxꢂ
and Y3Ox (x = 0–6). RSC Adv 4, 60270–60279.
54. He, F., Li, S., Garreau, H., Vert, M., and Zhuo, R.
(2005). Enzyme-catalyzed polymerization and
degradation of copolyesters of ε-caprolactone
and g-butyrolactone. Polymer 46, 12682–
12688.
64. Dove, A.P. (2012). Organic catalysis for ring-
opening polymerization. ACS Macro Lett 1,
1409–1412.
65. Chen, E.Y.X. (2009). Coordination
polymerization of polar vinyl monomers by
single-site metal catalysts. Chem. Rev. 109,
5157–5214.
75. Wu, B., Gallucci, J.C., Parquette, J.R., and
RajanBabu, T.V. (2014). Bimetallic catalysis in the
highly enantioselective ring–opening reactions
of aziridines. Chem. Sci. 5, 1102–1117.
55. Agarwal, S., and Xie, X. (2003). SmI2/Sm-based
g-buyrolactoneꢂε-caprolactone copolymers:
microstructural characterization using one- and
two-dimensional NMR spectroscopy.
66. O’Keefe, B.J., Hillmyer, M.A., and Tolman, W.B.
(2001). Polymerization of lactide and related
cyclic esters by discrete metal complexes.
J. Chem. Soc. Dalton Trans. 2001, 2215–2224.
Macromolecules 36, 3545–3549.
76. Gu, W., Xu, P., Wang, Y., Yao, Y., Yuan, D., and
Shen, Q. (2015). Synthesis and characterization
of yttrium and ytterbium complexes supported
by salen ligands and their catalytic properties
for rac-lactide polymerization.
56. Du, G., Wei, Y., Zhang, W., Dong, Y., Lin, Z., He,
H., Zhang, S., and Li, X. (2013). Bis(imino)
diphenylamido rare-earth metal dialkyl
67. Pratt, R.C., Lohmeijer, B.G.G., Long, D.A.,
Waymouth, R.M., and Hedrick, J.L. (2006).
Triazabicyclodecene: a simple bifunctional
organocatalyst for acyl transfer and ring-
opening polymerization of cyclic esters. J. Am.
Chem. Soc. 128, 4556–4557.
complexes: synthesis, structure, and catalytic
activity in living ring-opening ε-caprolactone
polymerization and copolymerization with g-
butyrolactone. Dalton Trans 42, 1278–1286.
Organometallics 34, 2907–2916.
77. Zhang, C.-J., Hu, L.-F., Wu, H.-L., Cao, X.-H.,
and Zhang, X.-H. (2018). Dual organocatalysts
for highly active and selective synthesis of
linear poly(g-butyrolactone)s with high
molecular weights. Macromolecules 51, 8705–
8711.
57. Duda, A., Penczek, S., Dubois, P., Mecerreyes,
D., and Je´ roˆ me, R. (1996). Oligomerization and
copolymerization of g-butyrolactone — a
monomer known as unable to
68. Chuma, A., Horn, H.W., Swope, W.C., Pratt,
R.C., Zhang, L., Lohmeijer, B.G.G., Wade, C.G.,
Waymouth, R.M., Hedrick, J.L., and Rice, J.E.
(2008). The reaction mechanism for the
organocatalytic ring-opening polymerization
of l-lactide using a guanidine-based catalyst:
hydrogen-bonded or covalently bound? J. Am.
Chem. Soc. 130, 6749–6754.
homopolymerize, 1. Copolymerization with
ε-caprolactone. Macromol. Chem. Phys. 197,
1273–1283.
78. Moore, T., Adhikari, R., and Gunatillake, P.
(2005). Chemosynthesis of bioresorbable
poly(g-butyrolactone) by ring-opening
polymerisation: a review. Biomaterials 26,
3771–3782.
58. Hong, M., Tang, X., Newell, B.S., and Chen,
E.Y.-X. (2017). ‘‘Nonstrained’’ g-butyrolactone-
Chem 7, 670–685, March 11, 2021 685