Synthesis, Microstructure, and Properties of High-Molar-Mass Polyglycolide Copolymers with Isolated Methyl Defects

Article abstract of DOI:10.1021/acs.biomac.1c00269

An efficient, fast, and reliable method for the synthesis of high-molar-mass polyglycolide (PGA) in bulk using bismuth (III) subsalicylate through ring-opening transesterification polymerization is described. The difference between the crystallization (Tc ≈ 180 °C)/degradation (Td ≈ 245 °C) temperatures and the melting temperature (Tm ≈ 222 °C) significantly affects the ability to melt-process PGA homopolymer. To expand these windows, the effect of copolymer microstructure differences through incorporation of methyl groups in pairs using lactide or isolated using methyl glycolide (≤10% methyl) as comonomers on the thermal, mechanical, and barrier properties were studied. Structures of copolymers were characterized by nuclear magnetic resonance (1H and 13C NMR) spectroscopies. Films of copolymers were obtained, and the microstructural and physical properties were analyzed. PGA homopolymers exhibited an approximately 30 °C difference between Tm and Tc, which increased to 68 °C by incorporating up to 10% methyl groups in the chain while maintaining overall thermal stability. Oxygen and water vapor permeation values of solvent-cast nonoriented films of PGA homopolymers were found to be 4.6 cc·mil·m-2·d-1·atm-1 and 2.6 g·mil·m-2·d-1·atm-1, respectively. Different methyl distributions in the copolymer sequence, provided through either lactide or methyl glycolide, affected the resulting gas barrier properties. At 10% methyl insertion, using lactide as a comonomer significantly increased both O2 (32 cc·mil·m-2·d-1·atm-1) and water vapor (12 g·mil·m-2·d-1·atm-1) permeation. However, when methyl glycolide was utilized for methyl insertion at 10% Me content, excellent barrier properties for both O2 (2.9 cc·mil·m-2·d-1·atm-1) and water vapor (1.0 g·mil·m-2·d-1·atm-1) were achieved.

Full text of DOI:10.1021/acs.biomac.1c00269

pubs.acs.org/Biomac
Article
Synthesis, Microstructure, and Properties of High-Molar-Mass
Polyglycolide Copolymers with Isolated Methyl Defects
Esra Altay,† Yoon-Jung Jang,† Xiang Qi Kua, and Marc A. Hillmyer*
Cite This: Biomacromolecules 2021, 22, 2532−2543
Read Online
sı
*
Metrics & More
Article Recommendations
Supporting Information
ACCESS
ABSTRACT: An efficient, fast, and reliable method for the
synthesis of high-molar-mass polyglycolide (PGA) in bulk using
bismuth (III) subsalicylate through ring-opening transesterification
polymerization is described. The difference between the crystal-
lization (Tc ≈ 180 °C)/degradation (Td ≈ 245 °C) temperatures
and the melting temperature (Tm ≈ 222 °C) significantly affects
the ability to melt-process PGA homopolymer. To expand these
windows, the effect of copolymer microstructure differences
through incorporation of methyl groups in pairs using lactide or
isolated using methyl glycolide (≤10% methyl) as comonomers on
the thermal, mechanical, and barrier properties were studied.
Structures of copolymers were characterized by nuclear magnetic
resonance (1H and 13C NMR) spectroscopies. Films of copolymers
were obtained, and the microstructural and physical properties were analyzed. PGA homopolymers exhibited an approximately 30
°C difference between Tm and Tc, which increased to 68 °C by incorporating up to 10% methyl groups in the chain while
maintaining overall thermal stability. Oxygen and water vapor permeation values of solvent-cast nonoriented films of PGA
homopolymers were found to be 4.6 cc·mil·m−2·d−1·atm−1 and 2.6 g·mil·m−2·d−1·atm−1, respectively. Different methyl distributions
in the copolymer sequence, provided through either lactide or methyl glycolide, affected the resulting gas barrier properties. At 10%
methyl insertion, using lactide as a comonomer significantly increased both O2 (32 cc·mil·m−2·d−1·atm−1) and water vapor (12 g·mil·
m−2·d−1·atm−1) permeation. However, when methyl glycolide was utilized for methyl insertion at 10% Me content, excellent barrier
properties for both O2 (2.9 cc·mil·m−2·d−1·atm−1) and water vapor (1.0 g·mil·m−2·d−1·atm−1) were achieved.
Glycolic acid (GA) is produced from several feedstocks such
as formaldehyde and acetic acid, and its synthesis from oxalic
acid has also been reported.8 GA derived from renewable
biomass feedstocks, i.e., glyoxal and dihydroxyacetone, can
render it a biorenewable and sustainable monomer.9 With the
simplest chemical structure among all of the polyesters, PGAs
are rigid and highly crystalline (45−55%) thermoplastics. The
high gas barrier properties attributed to low free volume,
mechanical strength associated with high crystallinity and
density, and ease of hydrolytic breakdown and degradation
into CO2 and water in a compost within about a month are all
attributes that contribute to the high interest in this
material.10−19
INTRODUCTION
■
The increase in manufacturing and disposal of synthetic
polymers has led to severe polymer pollution, which is an ever-
growing concern. One solution to combat global plastic waste
is to emphasize, support, and carry out basic research on the
development of biobased sustainable polymers as next-
generation degradable and compostable materials. Develop-
ment of biodegradable and biorenewable polymers, such as
polyglycolide (PGA) and polylactide (PLA), for use over a
wider range of applications and as replacements for petroleum-
based synthetic materials has been at the forefront of research
efforts.1−4 In 2018, a Global Market Insights, Inc. report
predicts the value of polyglycolic acid market to be over $9
billion by 2024. In particular, the medical industry is the largest
contributor to this because of the increased use for biomedical
applications such as resorbable sutures (e.g., Dexon) and other
materials.5 Additionally, a significant demand from the
packaging industry is expected owing to PGAs low
permeability for gases, such as O2 and CO2 or water vapor,
that is comparable to high-barrier polymers such as ethyl/vinyl
alcohol copolymers (EVOH) and poly(vinylidene dichloride)
(PVDC).6,7
A wide variety of synthetic approaches have been reported
for the synthesis of PGA homopolymers, including solid-state
Received: March 2, 2021
Revised: April 20, 2021
Published: May 10, 2021
https://doi.org/10.1021/acs.biomac.1c00269
Biomacromolecules 2021, 22, 2532−2543
© 2021 American Chemical Society
2532

Biomacromolecules
pubs.acs.org/Biomac
Article
polycondensation of halogenoacetates;11,12 synthesis from
haloacetic acids,13 solution,14 or melt/solid polycondensa-
tion;15 and alternating copolymerization of C1 feedstocks
formaldehyde and carbon monoxide.16 Ring-opening trans-
esterification polymerization (ROTEP) has been the most
commonly used technique for obtaining high-molar-mass
PGA-based polyesters. Tin alkoxides and carboxylates are the
most widely used catalysts for the synthesis of polyesters and
have been employed to obtain PGA molar masses ranging from
about 2 to 42 kg·mol−1 in melt polymerizations using 1-
dodecanol and 1,4-butanediol as initiators, for example.20,21
Recent research has focused on finding new nontoxic metal-
based catalysts22 or effective organocatalysts23−27 for the
polyester synthesis. For example, bismuth salts28 have been
employed as catalysts for the copolymerization of glycolide
(G) with ε-caprolactone (CL)29 and lactides (L)30 but exhibit
reduced reactivity compared to typical tin catalysts (e.g.,
Sn(Oct)2). Lu et al. reported the synthesis of high-molar-mass
PGAs (20−250 kg·mol−1) in bulk at 130−150 °C using
diphenyl bismuth bromide.31 Chen and co-workers studied the
synthesis and catalytic activities of bis(triazol) derivatives of
Bi3+ complexes with fluorinated and nonfluorinated ligands in
the bulk ROP of G and attained high-molar-mass (52 kg·
mol−1) PGA homopolymers.32 The melt ring-opening trans-
esterification polymerization of PLAs has been studied using
bismuth (III) subsalicylate (BiSS), an effective, relatively low-
cost, and commercially available catalyst.33
achieving expansion of the thermal processing window.
Incorporation of methyl groups in the PGA backbone using
lactides has recently been reported, and the thermal,
mechanical, and barrier properties of the resultant copolymers
were investigated.36 However, the barrier performance of the
copolymers was quite diminished compared to PGA. Inspired
by the alternating copolymers of lactic acid and glycolic acid
synthesized from the ring-opening polymerization of methyl
,37
glycolide (MG),17
we envisioned that insertion of
glycolate−lactate−glycolate units into PGA could be carried
out using MG and yield samples having improved process-
ability with gas barrier properties similar to PGA. While the use
of MG to prepare PLGA block copolymers with ε-caprolactone
and lactides,38 and star-shaped poly(( )-lactic acid-alt-glycolic
acid), has been investigated,39 the incorporation of MG as a
comonomer into PGAs has not been reported. To test our
hypothesis, we synthesized PGA homopolymers and its
copolymers with either L or MG at low-methyl-group
incorporation (≤10%) to sufficiently expand the thermal
operating window without detrimentally affecting the notable
barrier properties of PGAs. Based on nuclear magnetic
resonance (NMR) spectra, we demonstrate that copolymers
of G with either L or MG reveal different microstructures, and
the resultant polymers have been investigated to explore
microstructure−property relationships in these new copoly-
mers.
Low-molar-mass PGA samples display lower crystallinity,
relatively poor thermal and mechanical behavior, and higher
water permeability. While high-molar-mass PGA is desirable,
high crystallinity leads to low solubility in common solvents,
and PGA is only practically soluble in hexafluoroisopropanol
(HFIP) and trifluoroacetic acid (TFA). Therefore, PGA
homopolymers or copolymers with high PGA content are
typically prepared using high-temperature bulk or melt
polymerization methods. PGAs can depolymerize at relatively
low temperatures (Td ≈ 245 °C) following the first-order
reaction through intramolecular ester interchange process and
release glycolide as the major degradation product.34 PGAs
also possess high melting temperatures (Tm ≈ 222 °C) and
high crystallization temperatures (Tc ≈ 180 °C). The small
difference between typical decomposition and melting temper-
atures can lead to difficulties during the polymerization. Owing
to these challenges, studies are limited to some extent
regarding the synthesis of PGA homopolymers or copolymers
with a high PGA content. Simple and reproducible synthetic
methods are desirable for PGA synthesis to produce high-
molar-mass samples. Forming robust films also depends on the
thermal properties of polymeric samples. As a relatively small
difference between Td and Tm limits the polymer processing
due to degradation, the narrow operating window between Tm
and Tc can limit melt extrusion processes due to rapid
crystallization. Therefore, it is important to expand the window
between Tm and Tc and between Td and Tm by manipulating
the chemical structure without compromising the favorable
properties of PGA. One way to do this is to introduce a
comonomer at a low concentration, which is expected to yield
lower Tm and Tc by moderate disruption to the crystal
structure.35
EXPERIMENTAL DETAILS
■
Materials. Glycolide monomer was purchased from Sigma-
Aldrich. It was first recrystallized with activated carbon from ethyl
acetate and then recrystallized from ethyl acetate. The glycolide was
recrystallized again from dichloromethane and dried in a vacuum oven
for 2 days. ( )-Lactide was purchased from Sigma-Aldrich. It was
recrystallized from toluene three times before polymerization.
Bismuth subsalicylate, benzyl alcohol, diphenyl phosphate, and
poly(ethylene glycol) (PEG) 600 were purchased from Sigma-Aldrich
and used as received. PEG 600 was purged with nitrogen gas in a
three-neck round-bottom flask at 60 °C for 3 days to remove any
residual water. Hexafluoro-2-isopropanol was obtained from Oak-
wood Chemical and used as received. Solvents like ethyl acetate and
methanol were purchased from Sigma-Aldrich while anhydrous ethyl
ether was purchased from Fisher Chemical. They were used as
received. Synthesis of O-(2-bromopropionyl) glycolic acid and methyl
glycolide were synthesized according to literature procedures.40
1
Characterization. H NMR spectra were recorded on a Bruker
Advance III HD 400 MHz spectrometer equipped with an
autosampler. Chemical shifts are reported in δ units, expressed in
ppm using the residual trifluoroacetic acid-d1 signal (11.50 ppm) as an
internal standard. Differential scanning calorimetry (DSC) analyses
were performed on a TA Instruments Discovery DSC using
hermetically sealed aluminum Tzero pans. Scans were conducted
under a nitrogen atmosphere at a heating and cooling rate of 10 °C·
min−1. Thermogravimetric analysis (TGA) was performed on a TA
Instruments Q500 under nitrogen atmosphere at a heating rate of 10
°C·min−1. Size exclusion chromatography (SEC) was performed using
a solution of 0.05 M potassium trifluoroacetate in HFIP at a flow rate
of 0.35 mL·min−1 and 40 °C on an EcoSEC SEC system HLC-
8240GPC series liquid chromatograph fitted with a refractive index
detector and two Tosoh TSKgel SuperAWM-H. Molecular masses
were determined by conventional calibration using poly(methyl
methacrylate) (PMMA) standards. Before SEC analysis, the dissolved
polymer was filtered through a 0.2 μm filter (Whatman) before
injection into the column. A polarized optical microscopy (Olympus
BX53) equipped with a hot stage (Linkam LTS420E) was used to
study the crystallinity and morphologies of neat polymers. Samples
were heated to 240 °C and hold for 3 min, then cooled to 25 °C to
We posit that the desirable physical, thermal, and
mechanical properties of PGAs can be tuned through low
levels of comonomer incorporation while preserving the
desirable crystallization and barrier properties of PGAs while
2533
https://doi.org/10.1021/acs.biomac.1c00269
Biomacromolecules 2021, 22, 2532−2543

Biomacromolecules
pubs.acs.org/Biomac
Article
crystallize at either 10 or 5 °C·min−1 depending on the composition
of the copolymer.
conversion, Mn,NMR = 215 kg·mol−1, Mn,SEC = 91 kg·mol−1, and Đ =
2.0.
PGA films to test gas permeability were fabricated using a 200 °C
Max. Compact Tape Casting Coater with heated vacuum bed (10″W
x 16″L) and Doctor Blade-MSK-AFA-H200A. The film-making
process was conducted at room temperature and the vacuum was
applied. Tensile test for the PGA film was performed using a
Shimadzu Autograph AGS-X Tensile Tester. Before the tensile test,
the PGA film was made with the thickness ranging from 100 to 200
μm by casting the polymer solution onto glass Petri dishes, followed
by drying under ambient conditions and application of vacuum. Water
vapor transmission rate (WVTR) studies for the films (having a
masked test area of 5 cm2) were carried out with Permatran 3/33
Model G (Mocon) according to ASTM F1249-01 at 90% humidity
level, atmospheric pressure, and a temperature of 38 °C. Oxygen
transmission rates (OTR) of the films (having a test area of 5 cm2
masked) were measured with OxTran 2/21 Model H (Mocon)
according to ASTM D3985-02 standard at 0% humidity level with
oxygen gas at atmospheric pressure and 23 °C test temperature.
Nitrogen gas was used as the carrier gas.
Example of PGA Homopolymer Synthesis Using Diphenyl
Phosphate (DPP) Catalyst (4). Glycolide (1.5 g, 13.0 mmol),
diphenyl phosphate (8.08 mg, 0.032 mmol), and poly(ethylene
glycol) (PEG 600) (6.46 mg, 5.72 μL) with a mole ratio of 1200:3:1
were added into an oven-dried vial containing a magnetic stir bar.
PEG 600 (5.72 μL) was measured using a microliter syringe and
injected into the vial. The preparation of the reaction was done in a
glovebox and any apparatus used was dried in an oven beforehand.
The vial was capped tightly before being brought out of the glovebox.
It was sealed with electrical tape and wrapped with aluminum foil to
ensure a constant heat transfer. The vial was heated in a heating block
at 200 °C for 2 h. After 2 h, the reaction was stopped and allowed to
cool to room temperature. The vial was then unsealed and an
appropriate amount of HFIP was added to dissolve the reaction
mixture. After the reaction mixture was fully dissolved in HFIP, it was
precipitated into diethyl ether (100 mL). The solution was decanted,
and the polymer was dried in a vacuum oven overnight. The polymer
was refluxed twice using ethyl acetate at 88 °C and each reflux
reaction time was 2 h. Compound 4 had a 25% conv. Mn,NMR = 29 kg·
mol−1, Mn,SEC = 19 kg·mol−1 and Đ = 1.7.
Example of Preparation of Thin Films Using Lower-Molar-
Mass PGAs (30k−40k). In a 20 mL scintillation vial, 650 mg of PGA
beads was first dissolved in 7.31 mL of HFIP to make a solution with
a concentration of 5 wt % in HFIP. The vial was heated in a heating
block at 50 °C until the polymer was completely dissolved in the
HFIP. A magnetic stir bar and 2.44 mL of HFIP were then added into
the vial to decrease the concentration of the solution to 4.0 wt % in
HFIP. The vial was placed on a magnetic stirrer without heating it and
allowed for a homogeneous solution to form. The solution was filtered
through a 0.45 μm polypropylene filter and ready for film making. On
a tape casting coater, a piece of aluminum foil was placed underneath
a fluorinated ethylene propylene (FEP) film on the surface of the
coater. The coater was set at room temperature, the height of the
applicator was set at 0.345 mm, and the speed of the applicator was
set at 50 mm/s. A small amount of isopropanol was sprayed on the
FEP surface and wiped using a piece of Kimwipe before pouring the
polymer solution on it. The solution was poured evenly on the FEP
surface and an inch off from the edges of the applicator to avoid
accumulation of solution at the sides of the film. Once the vacuum
was applied, the applicator was started to push the polymer solution
and spread it to form a uniform surface and thickness. Tweezers were
used to peel the PGA film off from the FEP film as it would stick on
the surface even when it dried out. A micrometer screw gauge was
used to measure the thickness of the PGA film.
Synthesis of 2-((2-bromopropanoyl)oxy)acetic acid.19
A
round-bottom flask was charged with glycolic acid (23 g, 0.3 mol)
in dioxane (110 mL) under an argon atmosphere. 2-Bromopropionyl
bromide (32 mL, 0.31 mol) in dioxane (115 mL) was added slowly at
15 °C for 2 h under Ar. The outlet of the reaction flask was bubbled
through the NaOH solution. The reaction mixture was stirred at 15
°C for 2 h, then dioxane was removed in vacuo. The remaining oil was
distilled, and the product was collected at 125 °C under 1000 mTorr.
The solid was recrystallized from toluene to obtain a white colorless
1
solid. Yield: 20 g (31%). H NMR (CDCl3, δ, ppm) = 1.88 (3H,
−CH3), 4.47 (1H, −CH); 4.75 (2H, −CH2) and 13C NMR (CDCl3,
δ, ppm) = 21.73 (−CHCH3), 38.98 (−CHCH3), 61.13
(−CH2COOH), 169.78 (−CHCOO−), 172.72 (−COOH).
Synthesis of Methyl Glycolide.19 A round-bottom flask was
charged with NaHCO3 (2.2 g, 26 mmol) in dimethylformamide
(DMF) (150 mL) under an argon atmosphere. 2-((2-
Bromopropanoyl)oxy)acetic acid (5 g, 24 mmol) in DMF (150
mL) was added slowly for 2 h at 80 °C under Ar. The reaction
mixture was stirred at 80 °C for an additional 2 h, then DMF was
removed in vacuo. After adding brine, the organic layers were
extracted with ethyl acetate. The combined organic layer was dried
over MgSO4, filtered, and concentrated. The crude was recrystallized
1
from ethyl acetate to obtain a colorless solid. Yield: 1.7 g (81%). H
NMR (TFA-d1, δ, ppm) 5.24 (1H, −CHCH3), 5.09 (2H,
−CH2COO−), 1.69 (3H, −CHCH3) and 13C NMR (TFA-d1, δ,
ppm) = 172.73 (−CHCOO−), 170.75 (−CH2COO−), 77.39
(−CHCH3), 68.36 (−CH2COO−); 17.19 (−CH3).
Example of PGA Homopolymer Synthesis Using Bismuth
Subsalicylate (BiSS) Catalyst (14). In an oven-dried 100 mL
round-bottom flask containing a magnetic stir bar, glycolide (10 g, 86
mmol), bismuth subsalicylate (3.9 mg, 0.011 mmol), and benzyl
alcohol (3.74 μL, 0.036 mmol) with a mole ratio of 2400:0.3:1 were
added. Benzyl alcohol (0.187 mL) and toluene (0.813 mL) were
added into a small oven-dried vial to make a 1 mL stock solution. The
stock solution (20 μL) was taken out and injected into the 100 mL
round-bottom flask. The preparation of the reaction was done in a
glovebox, and any apparatus used was dried in an oven beforehand to
get rid of water vapors. The 100 mL round-bottom flask was sealed
with a rubber septum before bringing it out of the glovebox. It was
insulated with glass wool and aluminum foil and heated at 230 °C in a
heating block for 10 min. After 10 min, the reaction was stopped and
allowed to cool to room temperature. The round-bottom flask was
then unsealed and an appropriate amount of HFIP was added to
dissolve the reaction mixture. When the reaction mixture was fully
dissolved in HFIP, it was precipitated into anhydrous ethyl ether (400
mL). The solution was decanted, and the polymer was dried in a
vacuum oven overnight. The polymer was refluxed twice using ethyl
acetate to remove excess glycolide monomer from the polymer
product. Each reflux reaction time was 2 h. Compound 14 had an 82%
Example for Preparation of Thin Film Using Higher-Molar-
Mass PGAs (>40k). In a 20 mL scintillation vial, 802 mg of PGA
beads was first dissolved in 9.02 mL of HFIP to create a polymer
solution with a concentration of 5.3 wt % in HFIP. The vial was
heated in a heating block at 50 °C until the polymer beads were
completely dissolved in HFIP. HFIP (1.03 mL) and isopropanol
(IPA, 0.26 mL) were then added into the vial. A polymer solution
with a concentration of 4.5 wt % in HFIP and 2.5 vol % of IPA was
made. A magnetic stir bar was also put into the vial. The vial was
placed on the magnetic stirrer without heating it until a homogeneous
solution was formed. The polymer solution was filtered through a 0.45
μm polypropylene filter and ready for film making. On a tape casting
coater, a piece of aluminum foil was placed underneath a FEP film on
the surface of the coater. The coater was set at room temperature, the
height of the applicator was set at 0.25 mm, and the speed of the
applicator was set at 50 mm·s−1. The polymer solution was poured
evenly. The applicator was started to push and spread the polymer
solution on the FEP film surface after the vacuum was applied. When
the PGA film was dried out, it would peel off by itself from the FEP
film. The thickness of the PGA film was measured using a micrometer
screw gauge.
2534
https://doi.org/10.1021/acs.biomac.1c00269
Biomacromolecules 2021, 22, 2532−2543

Biomacromolecules
pubs.acs.org/Biomac
Article
conversion (99%) (Mn,SEC = 12 kg·mol−1, 2) when PEG 400
was employed as an initiator (Figure S2), and homopolymers
with molar masses reaching up to 15 kg·mol−1 (3) and
dispersity values below 2 were obtained within a few hours.
However, at higher [M]0/[I]0 ratios, it proved difficult to
achieve high conversion, and increasing the catalyst loading
and reaction times might result in degradation (Figure S3C).
Sn(Oct)2 is a thermally stable catalyst, which is widely used
for bulk polymerization of cyclic esters at high temper-
atures.20,43,44 At 230 °C, Sn(Oct)2 mediated the rapid
polymerization of G (3 min) using benzyl alcohol as an
initiator at [M]0/[I]0 = 100, thereby generating polyglycolide
of 14 kg·mol−1 with 92% conversion (Table 1; 5). Increasing
[M]0/[I]0 ratio to 300 resulted in higher-molar-mass (23 kg·
mol−1) polymer (Table 1; 6). However, even higher molar
masses could not be achieved, even with an [M]0/[I]0 ratio of
600 (Table 1; 7). Additionally, Mn,SEC (16 kg·mol−1) deviated
from Mn,NMR (70 kg·mol−1), suggesting fortuitous initiation by
water. Due to the lack of success in preparing polymer samples
with reproducible and high molar masses, the resulting large
dispersities, and pale brown products, Sn(Oct)2 was not a
practical catalyst in our hands for the synthesis of high-molar-
mass PGA samples under the conditions we explored.
RESULTS AND DISCUSSION
■
ROTEP Synthesis of PGA Homopolymers. Bulk
polymerization of G was studied using various catalysts to
identify the optimal conditions for the preparation of high-
molar-mass PGAs through ROTEP (Scheme 1). Acid-
catalyzed (diphenyl phosphate) and metal-catalyzed (Sn(Oct)2
and BiSS) bulk homopolymerizations of G were investigated
(Table 1).
Scheme 1. ROTEP Synthesis of Polyglycolide Using ROH
Initiators
ROTEP synthesis of PGA in bulk was also performed using
BiSS as a catalyst and BnOH as an initiator at 230 °C under an
inert atmosphere (Table 1). Using a 0.03 catalyst-to-initiator
ratio ([C]0/[I]0) resulted in a low conversion (13%) and a low
Diphenyl phosphate (DPP) is an effective organocatalyst for
ROTEP of cyclic esters.26,41 Liu et al. reported polymerization
of ε-caprolactone, δ-valerolactone, and (−)-lactide with high
conversion at 180 °C in bulk using DPP at an equimolar ratio
of catalyst to initiator.42 However, this method has not been
reported previously for the synthesis of PGAs. Initially, G was
polymerized in bulk using DPP as a catalyst and benzyl alcohol
as an initiator at 200 °C, which resulted in 99% conversion
after 2 h (Mn,SEC = 7.8 kg·mol−1) (Table 1; 1). The relative
molar mass by SEC analysis (vs PMMA standards) was similar
M
n,SEC (9 kg·mol−1) (Table S1; s1). Increasing [C]0/[I]0 to 0.1
provided higher conversion (99%) and higher Mn,SEC (20 kg·
mol−1), but high dispersity values (Đ = 3.3) (Table S1; s2). As
we increased the [C]0/[I]0 ratio from 0.1 to 0.3, the
dispersities decreased from 3.3 to 1.9 and molar masses
increased to 49 kg·mol−1 (Table S1; 9). At even higher [C]0/
[I]0 ratios (0.6−1.8), however, a slight increase in dispersity
was observed in conjunction with a decrease in molar mass
(Table S1; s5−s7). Considering the conversion, molar mass,
and dispersities, the optimum [C]0/[I]0 was determined to be
0.3 (Figure 1A), or about 3 initiator molecules for every
catalyst molecule. The optimal reaction time for BiSS-catalyzed
1
to the molar mass by H NMR spectroscopic analysis with
Mn,NMR obtained by comparing the peak from the BnOH end
group at 7.3 ppm to the methylene peak of PGA at 5.0 ppm
(Figure S1). Bulk polymerizations at higher monomer-to-
initiator ([M]0/[I]0) ratios gave insoluble products. ROTEP
synthesis of PGA at [M]0/[I]0 = 100 afforded quantitative
Table 1. ROTEP Synthesis of Polyglycolide
mole ratios
a
a
b
b
sample ID
C
M
C
I
T (°C) time (min) conv (%) Mn,theo (kg·mol−1
)
Mn,NMR (kg·mol−1
)
Mn,SEC (kg·mol−1
)
Đ
1
DPP
DPP
DPP
DPP
Sn(Oct)2
Sn(Oct)2
Sn(Oct)2
BiSS
BiSS
BiSS
BiSS
BiSS
BiSS
BiSS
BiSS
BiSS
67
100
300
1200
100
300
600
100
300
600
900
900
1200
2400
3600
4800
0.3
1
1
3
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
200
200
200
200
230
230
230
230
230
230
230
230
230
230
230
230
120
120
180
480
3
30
60
10
10
10
10
10
10
10
10
10
99
99
53
25
92
85
85
97
94
98
98
98
94
82
89
96
7.7
12
18
34
11
30
59
11
33
9.1
11
14
29
9.9
40
70
14
40
71
121
63
134
215
369
497
7.8
12
15
19
14
23
16
13
49
74
86
96
69
91
92
108
2.2
2.0
2.0
1.7
2.8
3.6
3.4
2.7
1.9
1.9
2.0
1.9
2.0
2.0
2.0
1.9
c
2
d
3
d
4
e
5
0.006
0.006
0.006
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.6
1.0
e
6
7
e
e
8
e
9
e
10
11
12
13
14
15
16
68
e
102
102
131
230
371
532
f
e
e
e
e
a
b
1
Determined by the internal standard method with alcohol based on H NMR spectroscopy. Obtained by HFIP SEC using PMMA calibration.
c
d
e
f
Initiator: PEG 400. Initiator: PEG 600. Initiator: BnOH. Initiator: 1-dodecanol.
2535
https://doi.org/10.1021/acs.biomac.1c00269
Biomacromolecules 2021, 22, 2532−2543

Biomacromolecules
pubs.acs.org/Biomac
Article
temperature, and reaction time were performed well to give a
linear increase in Mn,SEC with increasing [M]0/[I]0 from 100 to
900 at a constant [C]0/[I]0 and near-constant conversion
(Table 1; 8−11 and Figure 1B). Monomodal distributions
were observed for the polymers based on SEC analysis (Figure
S4A). Furthermore, Mn,SEC (vs PMMA standards) was in good
agreement with Mn,NMR, consistent with the polymerizations
being largely initiated by BnOH. However, a higher molar
mass of polymer was not obtained upon further increasing
[M]0/[I]0 to 1200 and 2400, presumably due to the presence
of fortuitous initiator (e.g., low levels of water) (Table 1; 13,
14). To achieve higher-molar-mass polymers, polymerization
at an even higher [C]0/[I]0 ratio was performed and provided
Mn,SEC = 108 kg·mol−1 (Table 1; 16). ROTEP of G was also
performed in the absence of an alcohol initiator (Table S1;
s9); however, the polymerization in the presence of benzyl
alcohol was more controlled and yielded lower-dispersity
product (Table S1; s8). When BnOH was replaced with 1-
dodecanol (Table 1; 12) as a higher-boiling point alcohol to
initiate the polymerization, a similar Mn,SEC value was obtained.
Regardless of the reaction conditions, all polymers prepared at
230 °C in 10 min using BiSS as a catalyst yielded white
powders upon precipitation from HFIP into cold diethyl ether.
BiSS proved to be a superior catalyst for bulk polymerization of
G and was employed to synthesize the target copolymers.
Copolymerization of Glycolide with Lactide or
Methyl Glycolide Using BiSS. Racemic ( )-lactide (L) or
( )-methyl glycolide (MG) was copolymerized with glycolide
(G) using BiSS as a catalyst (Scheme 2). To keep the methyl
content low, a maximum of 10−20% mole fraction of L or MG
in the feed was employed (Table 2). Copolymerizations of G
and L in bulk for 10 min using BnOH as an initiator resulted in
poly(glycolide-stat-(( )-lactide)) (PGL) molar masses from
22 to 102 kg·mol−1 at high conversions and dispersities less
than 2. Copolymerization of L and G can be conducted at
temperatures lower than 230 °C due to the reduced Tm values
of copolymers. Therefore, PGL syntheses were performed at
180−220 °C. The level of incorporated methyl groups was
Figure 1. Plots for the ROTEP synthesis of PGA homopolymers
prepared in bulk at 230 °C in 10 min using BiSS as a catalyst. (A)
Mn,SEC and dispersity as a function of [C]0/[I]0 at [M]0/[I]0 = 300.
(B) Mn,SEC and dispersity against [M]0/[I]0 at [C]0/[I]0 = 0.3.
1
calculated from H NMR analysis based on methylene peaks
(5.00 ppm) of PGA units and methyl peaks of PLA units (1.64
ppm) in the resulting polymers (Figure S5) and found within
the range of 4−10% (Table 2; PGL1−PGL6).
Likewise, poly(glycolide-stat-(( )-methyl glycolide))
(PGMG) was prepared by varying the [M]0/[I]0 ratio using
10 and 20% mole fractions of MG in the feed (Table 2). The
bulk copolymerization with [M]0/[I]0 of 500:1 with 10% feed
polymerizations was found to be 10 min at 230 °C to obtain
high conversions. As expected, shorter reaction times resulted
in lower conversion (25% after 5 min; Table S1, s4) and a
slight discoloration of products was observed at 30 min.
Polymerizations using the optimized [C]0/[I]0 ratio, reaction
Scheme 2. Copolymerization of Glycolide with ( )-Lactide or ( )-Methyl Glycolide Using BiSS
2536
https://doi.org/10.1021/acs.biomac.1c00269
Biomacromolecules 2021, 22, 2532−2543

Biomacromolecules
pubs.acs.org/Biomac
Article
a
Table 2. Polymerization of Glycolide with ( )-Lactide or ( )-Methyl Glycolide Using BiSS
mole ratios
b
c
d
e
f
[G]0 to [L]0 or
[MG]0
T
time
(min)
conv.
(%)
Mn,theo
Mn,NMR
Mn,SEC
Me
e
sample ID
M
C
BnOH (°C)
(kg/mol)
(kg/mol)
(kg/mol)
Đ
(%)
PGL1
PGL2
PGL3
PGL4
PGL5
PGL6
PGMG1
PGMG2
PGMG3
PGMG4
PGMG5
PGMG6
PGMG7
PGMG8
PGMG9
PGMG10
90:10
90:10
90:10
85:15
85:15
85:15
90:10
90:10
90:10
90:10
90:10
80:20
80:20
80:20
80:20
80:20
100
500
1000
900
1200
900
500
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
210
210
210
220
220
180
210
210
210
210
210
210
210
180
180
180
10
10
10
10
10
10
20
30
20
20
20
20
20
20
30
20
97
94
85
90
77
92
97
98
87
85
72
97
86
84
93
80
11
55
101
92
111
98
57
30
57
22
73
102
86
81
108
34
17
55
81
39
27
46
33
46
1.8
1.8
1.7
1.9
1.8
2.0
2.3
3.0
2.6
1.9
2.3
2.4
2.5
2.3
2.3
2.1
8
8
4
10
6
10
5
4
5
4
4
10
9
8
9
8
139
100
119
116
10
14
30
27
13
21
31
19
19
500
59
1000
1500
2000
500
1000
1000
1000
2000
102
150
169
58
102
99
111
188
0.3
10
34
a
b
1
Polymerizations were performed in bulk. Normalized average with mole ratio of convG and convCM determined by H NMR spectroscopy.
c
d
e
Calculated by ([MG]/[I]) × convG × MWG + ([MCM]/[I]) × convCM × MWCM. Determined by 1H NMR spectroscopy. Determined by HFIP
f
SEC calibrated using PMMA standard. Methyl incorporation in the polymer backbone calculated by the number of methyl group divided by the
1
sum of the number of methylene and methyl group based on H NMR spectroscopy. CM: ( )-methyl glycolide or ( )-lactide.
Figure 2. (A) 1H NMR analysis of PGMG (2.5 kg·mol−1, 9% methyl group in the polymer backbone) and PGL (2.6 kg·mol−1, 9% of methyl group
in the polymer backbone) in TFA-d1 and CDCl3. (B) 13C NMR analysis of PGMG, PGL, and PLA (4.9 kg·mol−1) in TFA-d1 and CDCl3. (C) End
group analysis of the P(MG) (2.5 kg·mol−1), PGMG, and PGL in TFA-d1 and CDCl3.
composition of MG at 210 °C consumed monomer rapidly
(97% conversion within 20 min), producing PGMG of Mn =
34 kg·mol−1 (Table 2; PGMG1). While 10 mol % L feed was
expected to result in 10% methyl group content in the
backbone for PGL, 5% methyl group content was expected
PGL1 and PGMG1) during propagation. For longer reaction
times (30 min), the polymerizations resulted in the PGMG
with decreased molar mass (Mn = 17 kg·mol−1) possibly
because of intramolecular transesterification (Table 2;
PGMG2). Higher-molar-masses of PGMG were achieved by
increasing the [M]0/[I]0 ratio with 4−5% of methyl group
incorporation (Table 2; PGMG3−PGMG5). For high-molar-
mass samples (Mn,SEC = 81 kg·mol−1), however, the molar mass
obtained by 1H NMR analysis deviated from the value
obtained by the SEC because of the technical difficulty of
the end group analysis by NMR spectroscopy. Using a 20%
feed ratio of MG, a variety of molar masses were obtained by
1
when 10 mol % MG feed was used. Based on H NMR
analysis, 10% feed mole fraction of MG resulted in 5% methyl
incorporation in the backbone as expected (Table 2;
PGMG1). However, unlike MG, L incorporation deviated
from the L feed ratio, resulting only in 8% incorporation
(Table 2; PGL1). This observation may be due to the lower
overall steric hindrance in MG as compared to L (Table 2;
2537
https://doi.org/10.1021/acs.biomac.1c00269
Biomacromolecules 2021, 22, 2532−2543

Biomacromolecules
pubs.acs.org/Biomac
Article
Figure 3. DSC analysis of 16 (108k, 0% Me), PGL3 (102k, 4% Me), PGMG4 (81k, 4% Me), PGL2 (73k, 8% Me), and PGMG7 (46k, 9% Me)
under N2(g) atmosphere. (A) Determined at a heating rate of 10 °C·min−1 based on the second heating cycle. (B) Determined at a cooling rate of
10 °C·min−1 based on the first cooling cycle.
Table 3. Thermal Properties of PGA and Copolymers Prepared Using BiSS
a
c
d
TGA
DSC
a
b
e
f
g
sample ID
% Me
Mn,SEC (kg·mol−1
)
Td,5 wt % (°C)
Tg (°C)
Tm (°C)
ΔHm (J/g)
Tc (°C)
χc (%)
s1
13
16
PGL3
PGMG4
PGL2
PGMG7
0
0
0
4
4
8
9
9
69
108
102
81
275
302
298
272
294
246
260
broad
37
37
46
43
212
222
222
208
205
203
182
85
90
93
51
65
57
23
185
195
194
156
137
143
41
44
45
25
31
28
11
73
46
38
36
h
129
a
Methyl incorporation in the polymer backbone calculated by the number of methyl group divided by the sum of the number of methylene and
b
c
methyl group based on 1H NMR spectroscopy. Determined by HFIP SEC calibrated using PMMA standards. Determined by TGA at a heating
rate of 10 °C/min under nitrogen atmosphere. Determined by DSC analysis under N2 atmosphere at a heating and cooling rate of 10 °C/min
d
e
f
based on the second heating cycle. ΔHm was determined based on the second heating cycle. Determined by DSC analysis under N2 atmosphere at
g
a cooling rate of 10 °C/min based on the first cooling cycle. Percent crystallization (%) was calculated by χc (%) = [(ΔH/ΔH0) × 100], where
h
ΔH0 = 206.4 J/g. Determined by DSC analysis under N2 atmosphere at a heating rate of 10 °C/min based on the second heating cycle because Tc
was not observed in the first cooling cycle.
varying [M]0/[I]0 ratio from 500 to 2000 with 8−10% methyl
incorporation in the polymer backbone (Table 2; PGMG6−
PGMG10). Notably, 180 °C was a high enough temperature
for melt polymerization due to the low melting temperature of
the resulting copolymers. These experiments highlighted that
molar masses and methyl incorporation of polymers can be
controlled by [M]0/[I]0 ratio and mole fraction of MG,
respectively.
and S6) peaks. For 13C NMR spectra, two and one methine
carbon peaks appeared for PGL and PGMG, respectively,
while PLA had more than three methine carbon peaks. This
data suggested that PGL and PGMG had glycolate−lactate−
lactate−glycolate and glycolate−lactate−glycolate sequences,
respectively, without significant levels of blocky PLA sequences
(Figure 2B). The 1H NMR spectra showed methine protons in
PGMG (Ha) and PGL (Hb) that exhibited slightly different
chemical shifts at 5.32 and 5.30 ppm, respectively (Figure 2A).
1H NMR and 13C NMR spectroscopies indicated that PGL
and PGMG have distinct methyl group distributions, as
expected.
To analyze end groups and elucidate the reactive site of MG,
an oligomer of P(MG) (2.5 kg·mol−1) was synthesized. The
end groups of methine protons (Hc) and methylene protons
(Hd) adjacent to the hydroxyl group appeared at 4.61 and
4.40−4.56 ppm, respectively, supported by adding TFAA
(Figure 2C). The integration of Hc was 71% compared to sum
of Hc and Hd, suggesting that sterically less hindered carbonyl
of MG is more likely to react with the alcohol propagating
group than more hindered carbonyl site.17−19,45 In this regard,
when 20% of MG was incorporated to PGA, we anticipated
that PGMG would have 14% (20% × 71%) of Hc as an end
group. In fact, the integration of Hc in PGMG was 20%,
suggesting that methyl groups were likely distributed
1H NMR, 13C NMR, and heteronuclear multiple bond
correlation (HMBC) spectroscopies were used to investigate
detailed chemical structures and microstructures of polymers.
To analyze end groups of the polymers, low-molar-mass PGA
1
and copolymers were synthesized. In the H NMR spectrum,
PGA backbone peak appeared at 5.09 ppm in TFA-d1 (Figure
S7). The methylene protons adjacent to the terminal hydroxyl
group were clearly identified as the peak at 4.66 ppm was
shifted to 5.05 ppm (Figure S1) by the addition of
trifluoroacetic anhydride (TFAA) resulting in the putative
formation of the trifluoroacetate end group. The other end
group peaks at 5.01 and 5.39 ppm correlated with benzyloxy
carbonyl and phenyl groups by HMBC analysis, respectively,
suggesting that they were from the end group as a result of
1
BnOH initiation (Figure S7). H NMR spectra of PGL and
PGMG showed methyl (1.6 ppm) and methine protons (5.3
ppm) in addition to PGA backbone methylene (Figures 2, S5,
2538
https://doi.org/10.1021/acs.biomac.1c00269
Biomacromolecules 2021, 22, 2532−2543

Biomacromolecules
pubs.acs.org/Biomac
Article
essentially randomly with slight enrichment at the end of
polymer chains. The consecutive lactate-lactate sequence was
also incorporated slightly more at the end of the PGL chains,
supported by the integration of methylene and methine
protons in the end groups.
thermal properties and crystallization behavior, even with a
similar level of methyl incorporation.
Crystalline Structures of PGA Homopolymers and
Copolymers. Crystalline structures of PGA, PGL, and PGMG
were investigated using wide-angle X-ray diffractometry
(WAXD). WAXD data of solvent-cast films were recorded at
25 °C using a Bruker Discover D8 instrument equipped with a
Co-Kα source (λ = 1.79 Å). PGA showed diffraction peaks at
25.7 and 33.4°, which are attributed to the (110) and (020)
crystalline planes, respectively (Figure 4).49 Interplanar
Thermal Properties of PGA Homopolymers and
Copolymers. Thermogravimetric analysis (TGA) and differ-
ential scanning calorimetry (DSC) were used to investigate the
thermal properties of PGA homopolymers and copolymers
(Figure 3 and Table 3). The thermal decomposition of PGAs
as characterized by the 5% mass loss (Td,5%) values measured
at a rate of 10 °C·min−1 under nitrogen was in the range of
275−302 °C. PGA prepared using DPP (Table 1; 3) displayed
Td,5% at 321 °C, which was higher than the Td,5% values of
homopolymer samples obtained using either BiSS or Sn(Oct)2.
The residual DPP catalyst (0.001−5 wt %) possibly acts as a
thermal stabilizer in these cases.46 The glass-transition
temperature (Tg) for PGA was observed at 37 °C (Table 3;
13 and 16) but was not observed for low-molar-mass PGA
sample (9 kg·mol−1) (Table 3; s1). The melting (Tm) and the
crystallization (Tc) temperatures showed a slight variation
ranging from 212 to 222 °C and 185 to 195 °C, respectively
(Figure 3, black line and Table 3). The thermal properties
obtained for PGA samples were in good agreement with
previous studies.47
Derivative thermogravimetry results showed one-step
decomposition for PGA and PGL, but two-step decomposition
was observed for PGMG. The copolymers thermally degraded
at similar temperatures as the PGA homopolymers. The
resulting PGMG and PGL polymers exhibited a 5% mass loss
at 246−294 °C (Table 3 and Figure S8). Notably, the
difference between the thermal decomposition and melting
temperature (Td,5%−Tm) is an important property for melt
processing applications, as it forms a range of suitable
temperatures at which the polymer can be processed without
substantive degradation due to mass loss. Thus, it is worth
noting that Td,5%−Tm increased up to 89 °C (Table 3;
PGMG4) when methyl groups were incorporated, compared
to high-molar-mass PGA homopolymer samples (76 °C, Table
3, 16). At a low level of Me group incorporation, Tm reduced
from 222 to 205 °C and χc reduced from 45 to 25% consistent
with the fact that even a small amount of methyl groups can act
as defects thwarting crystallization. Interestingly, PGL and
PGMG had different thermal properties even with similar
methyl group contents. PGL with 8% methyl group exhibited
Tm at 203 °C with 28% crystallinity (Table 3; PGL2) and no
cold crystallization peak, while PGMG with 9% Me showed Tm
at 182 °C with 11% crystallinity and Tcc at 129 °C (Table 3;
PGMG7) when measured at a cooling rate of 10 °C·min−1.
The difference between Tm and Tc is also an important value
for extrusion of films and sheets because fast crystallization
upon cooling can limit processing protocols. PGA has a
relatively small Tm−Tc difference (28 °C); thus, it can be
challenging to prepare films and sheets through extrusion.48
Incorporation of methyl groups is advantageous in this regard,
as the decreased crystallization temperatures (Tc), allowed for
larger Tm−Tc in the range from 52 to 68 °C (Figure 3).
Surprisingly, PGMG with 9% methyl incorporation did not
show a crystallization peak in the cooling cycle, presumably
due to slow crystallization, while PGL with 8% methyl
incorporation revealed a crystallization exotherm centered at
143 °C. The thermal analysis indicates that different methyl
distributions from lactide or methyl glycolide induced different
Figure 4. WAXD profiles of PGA (11, 86k), PGL4% (PGL3, 102k),
PGL10% (PGL4, 86k), PGMG4% (PGMG4, 81k), and PGMG8%
(PGMG10, 34k). Gray dash lines are the main diffraction angles.
distances (d) of PGA and PGL were obtained from WAXD
profile, resulting in 4.02 and 3.11 Å for (110) and (020)
planes, respectively. These values agree with previous studies.49
PGL had similar diffraction peaks compared to PGA at 25.8
and 33.6° for (110) and (020), respectively, suggesting that
( )-lactide units act as defects that were not incorporated into
the crystal lattice. This trend was different from earlier studies
that concluded (−)-lactic acid co-crystallized with glycolic acid
in poly(glycolic acid-stat-((−)-lactic acid)), presumably due to
the stereo-irregularity and different microstructure of PGL in
our case.49 Interestingly, diffraction peaks of PGMG appeared
at 25.5 and 32.8° for (110) and (020) planes, respectively,
which were slightly shifted to lower angles compared to PGA
and PGL. The d values of PGMG were 4.06 and 3.17 Å for
(110) and (020) planes, respectively. Similar results have been
found in the previous studies of poly(glycolic acid-stat-
((−)-lactic acid)). It was stated that higher d values of
copolymers than PGA suggested co-crystallization of glycolic
acid and (−)-lactic acid units in the crystalline lattices.
The effects of L or MG insertion on the spherulite growth
behavior were also investigated through nonisothermal
crystallization using polarized optical microscopy. Figure 5
shows the photomicrographs of neat PGL and PGMG taken
during the cooling from molten state at 240 °C. Copolymer
film samples were ramped to 240 °C at a rate of 30 °C·min−1
and held at this temperature for 3 min. Then, the temperature
was reduced to 25 °C at a cooling rate of 10 °C·min−1 and the
crystal formation of PGL and PGMG copolymers was recorded
during the cooling process. The spherulites for both PGL4%
and PGMG4% began to appear at 169 °C. When the
2539
https://doi.org/10.1021/acs.biomac.1c00269
Biomacromolecules 2021, 22, 2532−2543

Biomacromolecules
pubs.acs.org/Biomac
Article
Figure 6. Representative uniaxial extension data of 14 (PGA, 91k, χc
= 46%); PGMG4 (4% Me, 81k, χc = 39%) and PGMG9 (9% Me, 46k,
χc = 25%); and PGL2 (8% Me, 73k, χc = 27%) and PGL3 (4% Me,
102k, χc = 45%), extended at 5 mm·min−1, with the break point
indicated by X.
ultimate tensile strength even though it possessed a lower
degree of crystallinity. Furthermore, PGMG with 9% methyl
incorporation showed higher toughness than PGL. These
experiments suggested that distinct methyl distribution in the
polymer backbone resulted in discernable changes in
mechanical properties.
Barrier Properties of PGA Homopolymer, and PGL
and PGMG Copolymers. Thin films of PGA-based polymer
samples were cast from HFIP solution for barrier testing. Prior
to casting the films, polymer samples were dried in a vacuum
oven for at least 4 days to remove any moisture and dissolved
in HFIP completely at 60 °C at the initial polymer
concentrations of 4−6 wt %. Polymer sheets were cast using
a casting blade onto fluorinated ethylene propylene (FEP)
thin-film surface. Then, they were air-dried at room temper-
ature for 16 h and vacuum-dried at room temperature for 24 h.
The resulting thin-film samples with an average thickness of 25
μm were opaque due to the high levels of crystallinity (30−
42%) (Figure 7A−E).
Figure 5. Polarized optical micrographs of copolymers recorded at
various temperatures: PGL3 (4% Me) at: (A) 169 °C, (B) 140 °C,
and (C) 100 °C; PGL2 (8% Me) at: (D) 160 °C, (E) 140 °C, and
(F) 100 °C; PGMG4 (4% Me) at: (G) 169 °C, (H) 140 °C, and (I)
100 °C; and PGMG9 (9% Me) at: (J) 106 °C, (K) 96 °C, and (L) 61
°C.
temperature was reduced to 140 °C, the area covered with
crystals was larger for PGL4% than PGMG4%, but their
spherulite sizes were similar. An increase in spherulite size
was observed for PGMG4% as the temperature reduced from
140 to 100 °C. The same temperature decrease did not
profoundly affect the size of PGL4% spherulites. The crystal
formation for PGMG9% was too slow at 10 °C·min−1;
therefore, we utilized a slower cooling rate of 5 °C·min−1.
Increasing the amount of % Me to 9% using MG reduced both
the growth rate and Tc. PGMG9% started to form crystals at
106 °C. The data support the DSC results, where
incorporating 9% Me using MG significantly reduced the
rate of crystal formation. Also, the size and the area covered
with crystals significantly reduced.
Mechanical Properties. Tensile tests were performed to
investigate the mechanical properties of PGA homopolymers
and copolymers. Samples were solvent-cast from HFIP, dried
for more than 1 day, and cut into dog-bone test shapes.
Representative data sets are shown in Figure 6 and Table S3.
PGA and copolymers had Young’s moduli in the range from
0.46 to 1.0 GPa. PGA exhibited brittle fracture with an
ultimate tensile strength of 12 MPa and an elongation at break
of 4.5% due to the high crystallinity of the material. These
values are lower than commercial PGA (Kuredux) prepared by
injection molding with a high tensile strength of 117 MPa and
an elongation at break of 13%, as reported by Kureha,
presumably due to the higher degree of chain alignment and
crystallinity from different processing conditions.50 The
copolymers showed ultimate tensile strengths up to 12 MPa
and 8% strains at break. Interestingly, PGMG with 9% Me
slightly outperformed the other polymers with respect to
Figure 7. Tape-casted PGA film samples: (A) 14 (PGA, 0% Me); (B)
PGMG4 (4% Me); (C) PGL3 (4% Me); (D) PGMG10 (8% Me);
and (E) PGL2 (8% Me).
Oxygen permeation (OP) and water vapor permeation
(WP) were measured for nonoriented solvent-cast films of
PGA, PGL, and PGMG samples (Figure 8 and Table S4). OP
at 23 °C and 0% relative humidity (RH) of PGA film was 4.6
cc·mil·m−2·d−1·atm−1, which was higher than that of biaxially
oriented PGA (Kuredux, 0.6 cc·mil·m−2·d−1·atm−1) reported
by Kureha according to ISO-15105-2 at 50% RH.51
Presumably, lower OP of Kuredux is likely a result of biaxially
oriented films. Stretching films in one direction (uniaxially
oriented) or two directions (biaxially oriented) reduces
2540
https://doi.org/10.1021/acs.biomac.1c00269
Biomacromolecules 2021, 22, 2532−2543

Biomacromolecules
pubs.acs.org/Biomac
Article
nucleation density in addition to the degree of crystallinity.52
On the contrary, PGL samples have higher gas permeability
than PGA because of lower degrees of crystallinity, suggesting
that two methyl groups between PGA units substantially
altered and deteriorated the barrier performance of PGA. In
conclusion, these different results between PGL and PGMG
indicated that distribution of methyl groups in the PGA
backbone was important for gas permeability.
CONCLUSIONS
■
Bismuth (III) subsalicylate was found to be the most effective
catalyst to prepare high-molar-mass PGA samples with Mn,SEC
up to 108 kg·mol−1 within 10 min through ring-opening
transesterification polymerization. Using the same approach, a
series of well-defined PGL and PGMG copolymers with high-
molar-masses and low methyl composition (4−10%) were
synthesized. Different methyl distributions were imparted in
the polymer backbone based on the comonomer used and
resulted in a change in copolymer microstructure, crystal-
lization behavior, and thermal, mechanical, and barrier
properties. As we increased the methyl group content in the
copolymers, the degradation temperature of the copolymers
did not significantly change, but the melting temperature
decreased. This resulted in an increase of Td,5%−Tm from 76 to
89 °C, therefore widening the processing window. A
substantial decrease in Tm from 222 to 182 °C and in χc
from 45 to 11% was observed with 9% Me inclusion.
Interestingly, PGL8% and PGMG9% have different melting
temperatures, enthalpies of melting, and crystallization
behaviors. Results from the mechanical tests suggested that
the resulting polymers were brittle due to the crystallinity with
ultimate tensile strengths up to 12 MPa and 8% strain at break.
Nonoriented solvent-cast PGA films provided a slightly higher
oxygen (4.6 cc·mil·m−2·d−1·atm−1) and water vapor (2.6 g·mil·
m−2·d−1·atm−1) permeations comparable to commercial PGA
materials. Incorporation of methyl group using lactide
significantly increased OP and WP, especially at 10% Me
insertion (OP = 32 cc·mil·m−2·d−1·atm−1 and WP = 12 g·mil·
m−2·d−1·atm−1). In contrast, the copolymer materials manufac-
tured using MG possessed improved barrier properties (2.9 cc·
mil·m−2·d−1·atm−1 of OP and 1.0 g·mil·m−2·d−1·atm−1 of WP)
compared to those using L. These data suggested that the
distribution of methyl groups along the backbone was
important for the thermal, mechanical, and gas barrier
properties in these sustainable materials. To better understand
these behaviors, more studies on the effect of morphological
parameters on mechanical and gas barrier properties could be
an interesting direction for further research.
Figure 8. (A) Oxygen and (B) water vapor permeability of PGA-
based polymers as a function of methyl content. PGA (11, 86k, χc =
45%), PGMG4% (PGMG4, 81k, χc = 35%), PGL4% (PGL3, 102k, χc =
31%), PGMG10% (PGMG6, 27k, χc = 31%), PGL10% (PGL4, 86k, χc =
30%).
molecular orientation and polymer fractional free volume and
thus enhances crystallinity. In general, biaxially oriented films
have lower permeability compared to nonoriented films due to
strain-induced crystallization. However, WP (90% RH and 38
°C) of PGA homopolymer film (11, 2.6 g·mil·m−2·d−1) was
lower than that of PGA reported by Kureha (7.9 g·mil·m−2·
d−1).
O2 permeation and water vapor permeation studies were
conducted for PGL4%, PGL10%, PGMG4%, and PGMG10%
copolymer samples using the same conditions. At low lactide
loading (4% Me), O2 permeation for PGL4% was found to be
11 cc·mil·m−2·d−1·atm−1, which significantly increased to 32 cc·
mil·m−2·d−1·atm−1 when methyl content was increased to 10%.
At a low methyl loading (4%) using L, WP was found similar
to pure PGA sample (2.6 g·mil·m−2·d−1·atm−1); however, a
significant increase in water vapor permeation (12 g·mil·m−2·
d−1·atm−1) was observed when % Me was increased to 10%. In
comparison to PGL copolymers, O2 and water vapor
permeation values were preserved when 4% Me was
incorporated using MG (OP = 7.8 cc·mil·m−2·d−1·atm−1 and
WP = 2.5 g·mil·m−2·d−1·atm−1). More impressively, when the
% Me was increased to 10% using MG, OP and WP both
decreased by about 50% to 2.9 cc·mil·m−2·d−1·atm−1 and to 1.0
g·mil·m−2·d−1·atm−1, respectively, compared to PGA homo-
polymer. However, further increasing the methyl group
content to 25% (PGMG25%) results in an amorphous polymer
(Table S2).
A combination of factors, such as spherulite size,
morphology, and polymer crystallinity, has an effect on the
barrier performance. PGMG samples showed comparable gas
permeability to PGA and lower gas permeability than PGL
even with the similar levels of % methyl groups in the
backbone. Presumably, this was because methyl glycolide and
glycolide units co-crystallized in PGMG. In addition, an
increase in hydrophobicity with insertion of methyl group
likely contributes to having superior performance in terms of
water vapor permeation. The gas permeability of PGMG
samples was not well correlated to the degree of crystallinity.
Therefore, barrier properties of PGMG samples seem to be
controlled by other factors, such as size of the crystallites/
ASSOCIATED CONTENT
■
sı
* Supporting Information
The Supporting Information is available free of charge at
https://pubs.acs.org/doi/10.1021/acs.biomac.1c00269.
1D and 2D nuclear magnetic resonance spectra, size
exclusion chromatography data, thermogravimetric
analysis data, mechanical properties data, and oxygen
permeability and water vapor permeability data (PDF)
Accession Codes Primary data files (NMR spectroscopy,
DSC, TGA, SEC, WAXD, and tensile testing data) are
available free of charge at https://doi.org/10.13020/
y5cb-bg84
2541
https://doi.org/10.1021/acs.biomac.1c00269
Biomacromolecules 2021, 22, 2532−2543

Biomacromolecules
pubs.acs.org/Biomac
Article
(9) Iglesias, J.; Martínez-Salazar, I.; Maireles-Torres, P.; Martin
Alonso, D.; Mariscal, R.; López Granados, M. Advances in Catalytic
Routes for the Production of Carboxylic Acids from Biomass: A Step
Forward for Sustainable Polymers. Chem. Soc. Rev. 2020, 49, 5704−
5771.
(10) Yamane, K.; Sato, H.; Ichikawa, Y.; Sunagawa, K.; Shigaki, Y.
Development of an Industrial Production Technology for High-
Molecular-Weight Polyglycolic Acid. Polym. J. 2014, 46, 769−775.
(11) Schwarz, K.; Epple, M. A Detailed Characterization of
Polyglycolide Prepared by Solid-State Polycondensation Reaction.
Macromol. Chem. Phys. 1999, 200, 2221−2229.
(12) Epple, M.; Herzberg, O. Polyglycolide with Controlled
Porosity: An Improved Biomaterial. J. Mater. Chem. 1997, 7, 1037−
1042.
AUTHOR INFORMATION
Corresponding Author
Marc A. Hillmyer − Department of Chemistry, University of
Minnesota, Minneapolis, Minnesota 55455-0431, United
States; orcid.org/0000-0001-8255-3853;
Email: hillmyer@umn.edu
■
Authors
Esra Altay − Department of Chemistry, University of
Minnesota, Minneapolis, Minnesota 55455-0431, United
States
Yoon-Jung Jang − Department of Chemistry, University of
Minnesota, Minneapolis, Minnesota 55455-0431, United
States
Xiang Qi Kua − Department of Chemistry, University of
Minnesota, Minneapolis, Minnesota 55455-0431, United
States
(13) Pinkus, A. G.; Subramanyam, R. New High-Yield, One-Step
Synthesis of Polyglycolide from Haloacetic Acids. J. Polym. Sci., Polym.
Chem. Ed. 1984, 22, 1131−1140.
(14) Ayyoob, M.; Lee, D. H.; Kim, J. H.; Nam, S. W.; Kim, Y. J.
Synthesis of Poly(Glycolic Acids) via Solution Polycondensation and
Investigation of Their Thermal Degradation Behaviors. Fibers Polym.
2017, 18, 407−415.
Complete contact information is available at:
https://pubs.acs.org/10.1021/acs.biomac.1c00269
(15) Takahashi, K.; Taniguchi, I.; Miyamoto, M.; Kimura, Y. Melt/
Solid Polycondensation of Glycolic Acid to Obtain High-Molecular-
Weight Poly(Glycolic Acid). Polymer 2000, 41, 8725−8728.
Author Contributions
†E.A. and Y.-J.J. contributed equally to this work.
(16) Gökturk, E.; Pemba, A. G.; Miller, S. A. Polyglycolic Acid from
̈
Notes
the Direct Polymerization of Renewable C1 Feedstocks. Polym. Chem.
2015, 6, 3918−3925.
The authors declare no competing financial interest.
(17) Lu, Y.; Swisher, J. H.; Meyer, T. Y.; Coates, G. W. Chirality-
Directed Regioselectivity: An Approach for the Synthesis of
Alternating Poly(Lactic-Co-Glycolic Acid). J. Am. Chem. Soc. 2021,
143, 4119−4124.
ACKNOWLEDGMENTS
■
The authors thank Braskem for financial support of this work.
They also thank the National Science Foundation Center for
Sustainable Polymers at the University of Minnesota, which is
a National Science Foundation-supported Center for Chemical
Innovation (CHE-1901635). The authors thank Dr. Jason
Clark at Braskem for helpful discussions. They thank Dr.
Christopher DeRosa at the University of Minnesota for helpful
discussions. They thank Donald Massey at Clemson University
for assistance with gas barrier measurements. WAXD data were
obtained in the Characterization Facility, University of
Minnesota, which receives partial support from NSF through
MRSEC program.
(18) Gilding, D. K.; Reed, A. M. Biodegradable Polymers for Use in
SurgeryPolyglycolic/Poly(Actic Acid) Homo- and Copolymers: 1.
Polymer 1979, 20, 1459−1464.
(19) Dong, C.-M.; Qiu, K.-Y.; Gu, Z.-W.; Feng, X.-D. Synthesis of
Poly(D,L-Lactic Acid-Alt-Glycolic Acid) from D,L-3-Methylglycolide.
J. Polym. Sci., Part A: Polym. Chem. 2000, 38, 4179−4184.
(20) Kaihara, S.; Matsumura, S.; Mikos, A. G.; Fisher, J. P. Synthesis
of Poly(L-Lactide) and Polyglycolide by Ring-Opening Polymer-
ization. Nat. Protoc. 2007, 2, 2767−2771.
(21) Gautier, E.; Fuertes, P.; Cassagnau, P.; Pascault, J.-P.; Fleury, E.
Synthesis and Rheology of Biodegradable Poly(Glycolic Acid)
Prepared by Melt Ring-Opening Polymerization of Glycolide. J.
Polym. Sci., Part A: Polym. Chem. 2009, 47, 1440−1449.
(22) Duval, C.; Nouvel, C.; Six, J.-L. Is Bismuth Subsalicylate an
Effective Nontoxic Catalyst for Plga Synthesis? J. Polym. Sci., Part A:
Polym. Chem. 2014, 52, 1130−1138.
(23) Pascual, A.; Leiza, J. R.; Mecerreyes, D. Acid Catalyzed
Polymerization of Macrolactones in Bulk and Aqueous Miniemulsion:
Ring Opening vs. Condensation. Eur. Polym. J. 2013, 49, 1601−1609.
(24) Zhou, X.; Hong, L. Controlled Ring-Opening Polymerization of
Cyclic Esters with Phosphoric Acid as Catalysts. Colloid Polym. Sci.
2013, 291, 2155−2162.
(25) Makiguchi, K.; Kikuchi, S.; Yanai, K.; Ogasawara, Y.; Sato, S.;
Satoh, T.; Kakuchi, T. Diphenyl Phosphate/4-Dimethylaminopyridine
as an Efficient Binary Organocatalyst System for Controlled/Living
Ring-Opening Polymerization of L-Lactide Leading to Diblock and
End-Functionalized Poly(L-Lactide)S. J. Polym. Sci., Part A: Polym.
Chem. 2014, 52, 1047−1054.
REFERENCES
■
(1) Dechy-Cabaret, O.; Martin-Vaca, B.; Bourissou, D. Controlled
Ring-Opening Polymerization of Lactide and Glycolide. Chem. Rev.
2004, 104, 6147−6176.
(2) Platel, R. H.; Hodgson, L. M.; Williams, C. K. Biocompatible
Initiators for Lactide Polymerization. Polym. Rev. 2008, 48, 11−63.
(3) Williams, C. K.; Hillmyer, M. A. Polymers from Renewable
Resources: A Perspective for a Special Issue of Polymer Reviews.
Polym. Rev. 2008, 48, 1−10.
(4) Yang, Z.-Y.; Chen, C.-W.; Rwei, S.-P. Influence of Asymmetric
Substituent Group 2-Methyl-1{,}3-Propanediol on Bio-Based Poly-
(Propylene Furandicarboxylate) Copolyesters. Soft Matter 2020, 16,
402−410.
(5) Frazza, E. J.; Schmitt, E. E. A New Absorbable Suture. J. Biomed.
Mater. Res. 1971, 5, 43−58.
(6) Kuredux Polyglycolic Acid (PGA) Technical Guidebook. http://
www.Kureha.Com/Pdfs/Kuredux_technical.Pdf (accessed September
6, 2020).
(7) Samantaray, P. K.; Little, A.; Haddleton, D. M.; McNally, T.;
Tan, B.; Sun, Z.; Huang, W.; Ji, Y.; Wan, C. Poly(Glycolic Acid)
(PGA): A Versatile Building Block Expanding High Performance and
Sustainable Bioplastic Applications. Green Chem. 2020, 22, 4055−
4081.
(8) Miltenberger, K. Ullmann’s Encyclopedia of Industrial Chemistry;
Wiley-VCH Verlag GmbH & Co. KGaA, 2000.
(26) Makiguchi, K.; Satoh, T.; Kakuchi, T. Diphenyl Phosphate as
an Efficient Cationic Organocatalyst for Controlled/Living Ring-
Opening Polymerization of δ-Valerolactone and ε-Caprolactone.
Macromolecules 2011, 44, 1999−2005.
(27) Dove, A. P. Organic Catalysis for Ring-Opening Polymer-
ization. ACS Macro Lett. 2012, 1, 1409−1412.
(28) Kricheldorf, H. R. Syntheses of Biodegradable and Biocompat-
ible Polymers by Means of Bismuth Catalysts. Chem. Rev. 2009, 109,
5579−5594.
2542
https://doi.org/10.1021/acs.biomac.1c00269
Biomacromolecules 2021, 22, 2532−2543

Biomacromolecules
pubs.acs.org/Biomac
Article
(29) Kricheldorf, H. R.; Rost, S. Copolymerizations of ε-
Caprolactone and GlycolideA Comparison of Tin(II)Octanoate and
Bismuth(III)Subsalicylate as Initiators. Biomacromolecules 2005, 6,
1345−1352.
(30) Kricheldorf, H. R.; Behnken, G. Copolymerizations of
Glycolide and L-Lactide Initiated with Bismuth(III)N-Hexanoate or
Bismuth Subsalicylate. J. Macromol. Sci., Part A 2007, 44, 795−800.
(31) Lu, Y.; Schmidt, C.; Beuermann, S. Fast Synthesis of High-
Molecular-Weight Polyglycolide Using Diphenyl Bismuth Bromide as
Catalyst. Macromol. Chem. Phys. 2015, 216, 395−399.
(32) Gao, J.; Yuan, Y.; Cui, A.-J.; Tian, F.; Chen, S.-C.; He, M.-Y.;
Chen, Q. Bismuth(III) Complexes Based on Bis(Triazol) Ligands:
Effect of Fluorine Substitution on the Structure and Catalysis for the
Polymerization of Glycolide. Z. Anorg. Allg. Chem. 2016, 642, 698−
703.
(33) Balasanthiran, V.; Beilke, T. L.; Chisholm, M. H. Use of over
the Counter Oral Relief Aids or Dietary Supplements for the Ring-
Opening Polymerization of Lactide. Dalton Trans. 2013, 42, 9274−
9278.
(48) Abe, S. Sequentially Biaxially-Oriented Polyglycolic Acid Film,
Production Process Thereof and Multi-Layer Film. U.S. Patent
US2011/0027590 A12011.
(49) Tsuji, H.; Kikkawa, K.; Arakawa, Y. Cocrystallization of
Monomer Units of Biobased and Biodegradable Poly(l-Lactic Acid-
Co-Glycolic Acid) Random Copolymers. Polym. J. 2018, 50, 1079−
1088.
(50) Kuredux Mechanical Properties. http://www.kuredux.com/en/
about/properties.html (accessed April 05, 2021).
(51) Ito, D.; Takatsuji, K. In Barrier Properties and Characteristics of
Polyglycolic Acid for Un-Oriented and Oriented Films, Annual Technical
ConferenceANTEC, 2011; pp 2608−2614.
(52) Drieskens, M.; Peeters, R.; Mullens, J.; Franco, D.; Lemstra, P.
J.; Hristova-Bogaerds, D. G. Structure versus Properties Relationship
of Poly(Lactic Acid). I. Effect of Crystallinity on Barrier Properties. J.
Polym. Sci., Part B: Polym. Phys. 2009, 47, 2247−2258.
(34) Cooper, D. R.; Sutton, G. J.; Tighe, B. J. Poly α-Ester
Degradation Studies. V. Thermal Degradation of Polyglycollide. J.
Polym. Sci., Polym. Chem. Ed. 1973, 11, 2045−2056.
(35) Yamane, K.; Miura, H.; Ono, T.; Nakajima, J.; Itoh, D.
Crystalline Polyglycolic Acid, Polyglycolic Acid Composition and
Production Process Thereof. European Patent EP1 914 258 B12008.
(36) Murcia Valderrama, M. A.; van Putten, R.-J.; Gruter, G.-J. M.
PLGA Barrier Materials from CO2. The Influence of Lactide Co-
Monomer on Glycolic Acid Polyesters. ACS Appl. Polym. Mater. 2020,
2, 2706−2718.
(37) Hosoya, K.; Maruyama, T.; Fujiki, T.; Tanaka, N.; Araki, M. A
Convenient Synthetic Method of Unsymmetrically Substituted Chiral
Lactone Using Liquid-Solid Biphase System. Chem. Express 1990, 21,
149−152.
(38) Dong, C.-M.; Qiu, K.-Y.; Gu, Z.-W.; Feng, X.-D. Living
Polymerization of D,L-3-Methylglycolide Initiated with Bimetallic
(Al/Zn) μ-Oxo Alkoxide and Copolymers Thereof. J. Polym. Sci., Part
A: Polym. Chem. 2001, 39, 357−367.
(39) Dong, C.-M.; Qiu, K.-Y.; Gu, Z.-W.; Feng, X.-D. Synthesis of
Star-Shaped Poly(d,l-Lactic Acid-Alt-Glycolic Acid) with Multifunc-
tional Initiator and SnOct2 Catalyst. Polymer 2001, 42, 6891−6896.
(40) Zhang, X.; He, Q.; Chen, Q.; Nealey, P. F.; Ji, S. Directed Self-
Assembly of High χ Poly(Styrene-b-(Lactic Acid-Alt-Glycolic Acid))
Block Copolymers on Chemical Patterns via Thermal Annealing. ACS
Macro Lett. 2018, 7, 751−756.
(41) Schneiderman, D. K.; Hillmyer, M. A. Aliphatic Polyester Block
Polymer Design. Macromolecules 2016, 49, 2419−2428.
(42) Liu, J.; Zhang, C.; Li, Z.; Zhang, L.; Xu, J.; Wang, H.; Xu, S.;
Guo, T.; Yang, K.; Guo, K. Dibutyl Phosphate Catalyzed Commercial
Relevant Ring-Opening Polymerizations to Bio-Based Polyesters. Eur.
Polym. J. 2019, 113, 197−207.
(43) Kricheldorf, H. R.; Kreiser-Saunders, I.; Stricker, A.
Polylactones 48. SnOct2-Initiated Polymerizations of Lactide: A
Mechanistic Study. Macromolecules 2000, 33, 702−709.
(44) Witzke, D. R.; Narayan, R.; Kolstad, J. J. Reversible Kinetics
and Thermodynamics of the Homopolymerization of L-Lactide with
2-Ethylhexanoic Acid Tin(II) Salt. Macromolecules 1997, 30, 7075−
7085.
(45) Takojima, K.; Makino, H.; Saito, T.; Yamamoto, T.; Tajima, K.;
Isono, T.; Satoh, T. An Organocatalytic Ring-Opening Polymerization
Approach to Highly Alternating Copolymers of Lactic Acid and
Glycolic Acid. Polym. Chem. 2020, 11, 6365−6373.
(46) Yamane, K.; Sato, H.; Kawakami, Y. Polyglycolic Acid-Based
Resin Composition and Shaped Product Thereof. U.S. Patent
US7,179,868 B22007.
(47) Yu, C.; Bao, J.; Xie, Q.; Shan, G.; Bao, Y.; Pan, P.
Crystallization Behavior and Crystalline Structural Changes of
Poly(Glycolic Acid) Investigated via Temperature-Variable WAXD
and FTIR Analysis. CrystEngComm 2016, 18, 7894−7902.
2543
https://doi.org/10.1021/acs.biomac.1c00269
Biomacromolecules 2021, 22, 2532−2543