Synthesis of polyaramids in γ-valerolactone-based organic electrolyte solutions

Article abstract of DOI:10.1039/d0gc03470c

The current synthetic procedures for polyaramids mainly involve the use of amide solvents such asN-methylpyrrolidone andN,N-dimethylacetamide. However, these solvents are suspected to be teratogenic and are considered ‘Substances of Very High Concern’ by the European Commission. Here we propose a benign alternative solvent system: an Organic Electrolyte Solution (OES) consisting of γ-valerolactone (GVL) and a small amount of the ionic liquid 1-methyl-3-octylimidazolium chloride, [C₈MIm][Cl]. Three commercially relevant polyaramids were synthesized: poly-p-phenylene terephthalamide (PPTA), poly-m-phenylene isophthalamide (PMIA) and copoly(p-phenylene/3,4′-diphenylether terephthalamide) (ODA/PPTA). PMIA was successfully synthesized in the OES containing [C₈MIm][Cl] in a molar fraction ofx_IL= 0.043, achieving an inherent viscosity ofη_inh= 1.94 ± 0.064 dL g^?1, which is on par with the current industrial standard and the benchmark lab scale synthesis. The reaction mixture could also be directly used for the wet spinning of polyaramid fibers, and all components of the solvent could be recycled in good yields by a series of evaporation and distillation steps. ODA/PPTA could be synthesized, but only rather low inherent viscosities were achieved. The reaction mixture was too viscoelastic to be spun by our small-scale spinning setup. PPTA always instantly precipitated and could not be synthesized from a [C₈MIm][Cl]/GVL OES. α-Picoline, the organic base which was added to capture the released HCl during the reaction, was found to play a pivotal role in the polymerization reaction. By undergoing an acid-base reaction with HCl, it forms a protic ionic liquidin situwhich increases the solubility of the polymer.

Full text of DOI:10.1039/d0gc03470c

Green Chemistry
View Article Online
View Journal | View Issue
PAPER
Synthesis of polyaramids in γ-valerolactone-based
organic electrolyte solutions†
Cite this: Green Chem., 2021, 23,
1228
Jonas Winters,
Raheed Bolia,
Wim Dehaen
and Koen Binnemans
*
The current synthetic procedures for polyaramids mainly involve the use of amide solvents such as
N-methylpyrrolidone and N,N-dimethylacetamide. However, these solvents are suspected to be terato-
genic and are considered ‘Substances of Very High Concern’ by the European Commission. Here we
propose a benign alternative solvent system: an Organic Electrolyte Solution (OES) consisting of
γ-valerolactone (GVL) and a small amount of the ionic liquid 1-methyl-3-octylimidazolium chloride,
[C8MIm][Cl]. Three commercially relevant polyaramids were synthesized: poly-p-phenylene terephthala-
mide (PPTA), poly-m-phenylene isophthalamide (PMIA) and copoly(p-phenylene/3,4’-diphenylether ter-
ephthalamide) (ODA/PPTA). PMIA was successfully synthesized in the OES containing [C8MIm][Cl] in a
molar fraction of xIL = 0.043, achieving an inherent viscosity of ηinh = 1.94 0.064 dL g−1, which is on par
with the current industrial standard and the benchmark lab scale synthesis. The reaction mixture could
also be directly used for the wet spinning of polyaramid fibers, and all components of the solvent could
be recycled in good yields by a series of evaporation and distillation steps. ODA/PPTA could be syn-
thesized, but only rather low inherent viscosities were achieved. The reaction mixture was too viscoelastic
to be spun by our small-scale spinning setup. PPTA always instantly precipitated and could not be syn-
thesized from a [C8MIm][Cl]/GVL OES. α-Picoline, the organic base which was added to capture the
released HCl during the reaction, was found to play a pivotal role in the polymerization reaction. By
undergoing an acid–base reaction with HCl, it forms a protic ionic liquid in situ which increases the solu-
bility of the polymer.
Received 15th October 2020,
Accepted 18th January 2021
DOI: 10.1039/d0gc03470c
rsc.li/greenchem
process these polymers from the melt.5 Polyaramids thus need
to be synthesized and processed from solution. This is again
Introduction
Since their development in the 1960s, polyaramids (aromatic complicated by the fact that they are exceptionally difficult to
polyamides) have become a staple amongst the high perform- dissolve in most conventional organic solvents. The large
ance polymers.1–4 They display several desirable properties number of amide functionalities can form intermolecular
including a high chemical and thermal resistance, a high hydrogen bonds between the amino and carbonyl groups.
strength/mass ratio and non-flammability. These properties Furthermore, the rigid aromatic backbone limits the mobility
make them suitable to serve in a wide range of applications of the amide functionalities, preorganising them for neatly
such as protective clothing, material reinforcement, ballistic packed and very efficient hydrogen bond networks.
protection, fiber optics and in aviation/aerospace technology.
To dissolve polyaramids, the intermolecular hydrogen bond
Currently, three types of polyaramids are produced commer- network needs to be broken up. For PPTA, this is done by
cially on a large scale: (1) poly-p-phenylene terephthalamide using a mixture of N-methyl-2-pyrrolidone (NMP) and CaCl2 as
(PPTA), (2) poly-m-phenylene isophthalamide (PMIA) and (3) the solvent for synthesis.6 The chloride anions provided by
copoly(p-phenylene/3,4′-diphenylether terephthalamide) (ODA/ CaCl2 can interact with the amino groups, thus interrupting
PPTA). However, the production of these polyaramids entails the intermolecular hydrogen bond network. Although this
some difficulties. Their melting points are well above their binary solvent system cannot make a thermodynamically
degradation temperatures, so it is impossible to synthesize or stable solution of PPTA, it can delay precipitation long enough
to let the polymerization proceed until a high average molar
mass is achieved. A thermodynamically stable solution is
required for processing. Thus, PPTA is dissolved in concen-
trated sulfuric acid, which protonates the amide functional-
KU Leuven, Department of Chemistry, Celestijnenlaan 200F, P.O. box 2404, B-3001
Leuven, Belgium. E-mail: Koen.Binnemans@kuleuven.be
†Electronic supplementary information (ESI) available: Synthetic procedures for
ities, creating a net positive charge on the chains and making
polyaramid oligomers, purification methods for the diamine monomers and
them repel each other.7 PMIA and ODA/PPTA are both more
details on the 3D-printed spinning setup. See DOI: 10.1039/d0gc03470c
1228 | Green Chem., 2021, 23, 1228–1239
This journal is © The Royal Society of Chemistry 2021

View Article Online
Green Chemistry
Paper
soluble than PPTA as a result of their respective structures. and cosolvents has successfully been applied to the field of
PMIA is m-oriented, which results in a less efficient orientation cellulose dissolution, where it has resulted in higher solubili-
of the amide groups, whereas ODA/PPTA has had some flexi- ties and faster dissolution kinetics.26–29 OESs were also proven
bility introduced to the chain by adding ether linkages. Both to be suitable solvents for the synthesis of PPTA and for pro-
polymers can be synthesized in NMP without the addition of cessing PMIA into porous films.30,31 However, these appli-
CaCl2, and the synthetic solutions can be used directly for the cations still relied on the addition of amide solvents such as
spinning of fibers.8,9
NMP or DMA. Recently, we have shown that OESs based on
The production of polyaramids is heavily dependent on the γ-valerolactone (GVL) are excellent solvents for PMIA and ODA/
use of NMP. However, this solvent has been placed on a PPTA, and that they can be used for wet spinning polyaramid
‘Substance of Very High Concern’ (SVHC) list by the European fibers.32 GVL was found to work synergistically with ILs, even
Commission, because of its suspected teratogenic properties.10 surpassing the solubility of polyaramids in the pure ILs. More
This makes NMP subject to potential restrictions for use in importantly, GVL is a very attractive solvent for large scale
industry, and so a new benign solvent is needed. PMIA can industrial applications because of its green character, as it is a
alternatively be produced via an interphase polymerization, renewable biomass-derived solvent that has a low toxicity and
using water and THF as a biphasic solvent system.11 However, is readily biodegradable.33–35 The addition of GVL to ILs for
this type of polymerization offers poor control of local the dissolution of polyaramids makes the whole process more
monomer stoichiometry, leading to significantly broader economically and ecologically feasible at an industrial scale.
molar mass distributions. Also, because PMIA is precipitated
This paper explores the potential of these GVL-based OESs
during the polymerization process, it is impossible to directly as solvents for the synthesis of polyaramids. All three of the
spin it from the synthetic solution, and it nevertheless needs above-mentioned polyaramids are synthesized in OESs, and
to be separated and redissolved in NMP. Another strategy has the inherent viscosities obtained from the polymerizations are
been to make the polyaramids more soluble in other solvents compared to the benchmark values from polymerizations per-
by altering their chemical structure.12,13 However, these altered formed in NMP. The influence of several parameters such as
polyaramids are made from more complex (and thus more the molar composition of the OES, the organic base added to
expensive) monomers, and their properties do differ from the scavenge HCl, the type of IL and the polymer concentration are
original polyaramids.
investigated. Finally, the possibility to directly spin fibers from
In recent years, efforts have been made by our laboratory the reaction mixtures is assessed, along with a method for
and others to explore the potential of ionic liquids (ILs) as recyclability of the components of the OES solvent.
alternative solvents for polyaramids, both for dissolution14–17
and synthesis.18–21 ILs are solvents which consist entirely of
ions, which means they can act in a similar fashion to the
Experimental
Products
NMP/CaCl2 solvent system. Indeed, it was found that only ILs
with strongly coordinating anions such as Cl−, AcO− or PO4R−
are suitable solvents for polyaramids, as these are required to The ionic liquids were purchased from IoLiTec (Heilbronn,
interrupt the intermolecular hydrogen bonds. For synthesis, Germany). Isophthaloyl chloride (≥99%), α-picoline (lonza
only ILs with a chloride anion can be used, as the other anions quality, ≥99%), Pd/C 5% loading, N-methylpyrrolidone extra
will cause side reactions. Although ILs do seem to have poten- dry (99.5%), 2,6-dichloro-4-nitropyridine (97%) and terephtha-
tial as alternative solvents, they do have some severe disadvan- loyl chloride (+99%) were purchased from Sigma-Aldrich
tages which can make them less suitable for large scale (Diegem, Belgium). 3-Nitroaniline (98%), extra dry N,N-di-
polymerizations.22,23 Firstly, ILs are generally very hygroscopic methylformamide (99%), sulfuric acid (96%), aniline (99.8%),
and they readily absorb moisture from the atmosphere.24 This benzoyl chloride (99%), triethylamine (99%), 1-methyl-
means that the ILs must be thoroughly dried prior to their use imidazole (99%), p-phenylenediamine (>99%), 4-dimethyl-
in polyaramid synthesis, as moisture will precipitate the aminopyridine (99%) and γ-valerolactone (98%) were pur-
polymer and react with the acyl chloride monomers and termi- chased
from
Acros
Organics
(Geel,
Belgium).
nate the reaction. Secondly, ILs also tend to be very viscous, m-Phenylenediamine (99%) was purchased from Chem-Lab
which can cause problems such as slow reaction kinetics, poor (Zedelgem, Belgium) and 3,4-oxydianiline (95%) was pur-
homogeneity and engineering issues. Thirdly, ILs are currently chased from Fluorochem Limited (Hadfield, UK). [2-Me-
much more expensive than conventional molecular organic HPyr][Cl] used as IL in polymer synthesis was recovered from
solvents. This is mostly due to the small scale at which they the recycling experiments. p-Phenylenediamine (99+%), 3,4-
are currently being produced, but it is a problem nevertheless.
oxydianiline (95%) and m-phenylenediamine (99%) were puri-
All of the above-mentioned issues can be resolved by dilut- fied extensively prior to use. γ-Valerolactone and α-picoline
ing the IL with a cosolvent, creating what is generally referred were dried using molecular sieves (4 Å) until the water content
to as an organic electrolyte solution (OES). Adding a cosolvent was below 50 ppm. The ionic liquids were dried on a vacuum
can lead to a drastic reduction of the viscosity,25 and the lower line operating at 0.6–1.0 mbar until the water content was less
relative IL content makes the overall solvent less hygroscopic, than 200 ppm. All other chemicals were used as received,
easier to dry, and less expensive. This strategy of mixing ILs without any further purification. The synthesis and character-
This journal is © The Royal Society of Chemistry 2021
Green Chem., 2021, 23, 1228–1239 | 1229

View Article Online
Paper
Green Chemistry
ization of the aramid oligomers, as well as the purification transferred to a beaker filled with water and broken up into a
methods for the diamines are described in the ESI.†
powder using an immersion blender. The polymer was fil-
trated, washed with water, further purified via 12 h of Soxhlet
extraction with water and finally dried overnight in a vacuum
Instrumentation
The water content was measured on a Karl Fischer coulometer oven at 50 °C.
(Mettler-Toledo DL39). The inherent viscosity of the polyara-
Poly-p-phenyleneterephthalamide (PPTA). After extensive
mids was determined using a certified calibrated Ubbelohde drying, the ionic liquid and cosolvent were mixed in appropri-
10–100 mm2 s−1 (0.1 K) capillary viscometer from ESI† ana- ate ratios to obtain an OES with the desired IL molar fraction.
lytics in a Lauda thermostated water bath. The short path evap- The OES (40 g) was then added to a custom-made cylindrical
orator used in the recycling experiments was the glass oven reaction vessel fitted with an anchor stirrer and flushed with
1
B-585 Kugelrohr (Büchi). The H NMR spectra (300 MHz) were N2. This type of reaction vessel was used because it was more
measured on a Bruker Avance 300 MHz spectrometer. The suitable for stirring and breaking the reaction mixture as it
NMR spectra were analyzed using TopSpin 4.0.2 software.
turned into
a
crumble.36 p-Phenylenediamine (1 eq.,
13.5 mmol, 1.46 g) was added immediately after purification to
reaction vessel, and the reaction mixture was stirred at 200
Polymerization procedures
Poly-m-phenyleneisophthalamide (PMIA). After extensive rpm. When the diamine monomer was dissolved, the solution
drying, the ionic liquid and cosolvent were mixed in appropri- was put in an ice bath and α-picoline (2.1 eq., 28.4 mmol,
ate ratios to obtain an OES with the desired IL molar fraction. 2.64 g) was added using a syringe. The stirring rate was
The OES (40 g) was then added to a dried 3-neck flask and increased to 500 rpm and terephthaloyl chloride (1.0 eq.,
flushed with N2. m-Phenylenediamine (1.0 eq., 19.4 mmol, 13.5 mmol, 2.74 g) was added. The stirring rate was further
2.09 g) was added immediately after purification to the reac- increased to 1000 rpm, and stirring continued for 20 more
tion mixture under a gentle N2 flow. An overhead stirrer (4 cm minutes. The reaction mixture was transferred to a beaker
collapsible two-blade impeller fitted in a Cowie universal filled with water and broken up into a powder using an immer-
stirrer guide) was then placed on the flask under N2 flow, and sion blender. The polymer powder was thoroughly washed
the reaction mixture was stirred at a rate of 200 rpm. When the with water, further purified via 12 h of Soxhlet extraction with
m-phenylenediamine was dissolved, the solution was then put water and finally dried overnight in a vacuum oven at 50 °C.
on an ice bath and α-picoline (2.1 eq., 40.7 mmol, 3.79 g) was
Determination of the inherent viscosity
added using a syringe. The stirring rate was increased to 500
rpm and isophthaloyl chloride (1.0 eq., 19.4 mmol, 3.94 g) was
added. The viscosity of the reaction mixture rapidly increased,
Dried polyaramid (0.1 g) was dissolved in 20 mL of concen-
trated sulfuric acid (96%) to obtain a solution with a concen-
creating a thick white/yellow gel. The reaction was allowed to tration of 0.5 g dL−1. An Ubbelohde capillary viscometer,
proceed for 2 more hours. The reaction mixture was trans-
ferred to a beaker filled with water and broken up into a
placed in a water bath at 25 °C, was used to measure the flow
times of pure sulfuric acid (tsolvent) and of the polyaramid/sul-
powder using an immersion blender. The polymer was fil- furic acid solution (tsolution). From these flow times, the relative
trated, washed with water, further purified via 12 h of Soxhlet
extraction with water and finally dried overnight in a vacuum
oven at 50 °C.
viscosity of the solution (ηrel), and subsequently the inherent
viscosity (ηinh) were calculated as an estimate measure for the
average molar mass of the polyaramid (eqn (1) and (2)).
Copoly(p-phenylene/3,4′-diphenylether
terephthalamide)
ln ηrel
c
(ODA/PPTA). After extensive drying, the ionic liquid and cosol-
vent were mixed in appropriate ratios to obtain an OES with
the desired IL molar fraction. The OES (40 g) was then added
ηinh
¼
ð1Þ
ð2Þ
tsolution
ηrel
¼
tsolvent
to
a
dried 3-neck flask and flushed with N2.
p-Phenylenediamine (0.5 eq., 4.25 mmol, 0.46 g) and 3,4-oxy-
dianiline (0.5 eq., 4.25 mmol, 0.85 g) were added immediately
after purification to the reaction mixture under a gentle N2
flow. An overhead stirrer (4 cm collapsible two-blade impeller
fitted in a Cowie universal stirrer guide) was then placed on
the flask under N2 flow, and the reaction mixture was stirred at
where c is the concentration, and tsolution and tsolvent are the
flow times of the solution and solvent, respectively.
Results and discussion
a rate of 200 rpm. When the diamine monomers were dis- The general reaction scheme for the synthesis of polyaramids
solved, the solution was put on an ice bath and α-picoline (2.1 is shown in Fig. 1. The polymerization proceeds via a polycon-
eq., 17.8 mmol, 1.66 g) was added using a syringe. The stirring densation reaction between aromatic diamines and aromatic
rate was increased to 500 rpm and terephthaloyl chloride (1.0 diacyl chlorides. The released HCl is captured by an organic
eq., 8.50 mmol, 1.73 g) was added under a N2 flow. The vis- base such as α-picoline. Because this is a polycondensation
cosity of the reaction mixture increased over the course of reaction, high molecular weights can only be obtained if the
15–30 minutes, creating a thick yellow gel. The reaction was monomers are present in near perfect stoichiometric balance
allowed to proceed for 4 more hours. The reaction mixture was and of very high purity.37 Therefore, weighing and adding the
1230 | Green Chem., 2021, 23, 1228–1239
This journal is © The Royal Society of Chemistry 2021

View Article Online
Green Chemistry
Paper
Fig. 1 General reaction scheme for the polycondensation between aromatic diamines and aromatic diacyl chloride to obtain polyaramids. The
released HCl is preferably scavenged by adding an organic base such as α-picoline.
monomers must be done with great care. Furthermore, the and 2.45 dL g−1 respectively. The reason for this large error
monomers are chemically very reactive. The diacyl chlorides margin is probably twofold. Firstly, during this synthesis there
can react with water present in the reaction mixture and dia- was a certain room for human error in weighing and addition
mines slowly oxidize upon contact with atmospheric oxygen. It of the monomers. This reaction uses three different monomers
is therefore important ensure that the solvent is thoroughly rather than two, and the smaller amount of monomers used in
dried, and the diamines are purified extensively directly before terms of mass amplifies any weighing inaccuracies. Secondly,
use. The purification procedures for the diamines are provided over the course of the reaction, the reaction mixture had the
in the ESI.† Specific synthetic procedures for each of the poly- tendency to become very viscous, turning into a viscoelastic
aramids are described above. All polymerizations followed gel that crept up the shaft of the stirrer (Fig. 2). This phenom-
these exact procedures unless mentioned otherwise. enon is known as the Weissenberg effect and it makes efficient
Procedures for the benchmark syntheses can be found in our stirring impossible.39 These factors make it difficult to achieve
previous work.32
high average molecular weights for ODA/PPTA on our small
lab-scale setup. Nevertheless, most of the inherent viscosities
Benchmark polymerization reactions in NMP
The industrial standards for the inherent viscosities for PMIA,
ODA/PPTA and PPTA are ηinh = 2 dL g−1, ηinh = 2–3 dL g−1 and
ηinh = 5–6 dL g−1, respectively.4,38 However, these polymeriz-
ation reactions are very sensitive to several parameters such as
stirring efficiency, temperature control, water uptake,
monomer purity, weighing errors, etc. All these parameters are
difficult to control on a small lab-scale reaction setup,
meaning that there might be a difference in solvent perform-
ance compared to industrial scale setups. Therefore, before
using OESs as synthetic solvents, benchmark reactions were
performed to more accurately compare the performance of the
OESs versus NMP under the same reaction conditions. PMIA
could reliably be synthesized using our setup, with an average
inherent viscosities of ηinh = 2.06 0.08 dL g−1, which matches
the industrial standard. However, the industrial processes gen-
erally add aniline as a chain stopper to have more control over
the reaction. The fact that we ‘only’ reach a benchmark value
of about 2 dL g−1 without the use of chain stoppers does mean
that our lab scale reaction setup underperforms with respect
to the specialized industrial setups. The synthesis of ODA/
PPTA proved far less reliable than the synthesis of PMIA. A
total of 14 polymerizations was performed, out of which the
average inherent viscosity was only ηinh = 1.53 0.33 dL g−1
with the lowest and highest obtained values being 1.08 dL g−1 Fig. 2 Weissenberg effect in a polymerization reaction of ODA/PPTA.
,
This journal is © The Royal Society of Chemistry 2021
Green Chem., 2021, 23, 1228–1239 | 1231

View Article Online
Paper
Green Chemistry
were in the range of 1.45–1.55 dL g−1. This is significantly less
than the industrial standard, but it is the most appropriate
reference to which we can compare the performance of the
OESs.
The final polymer PPTA could be synthesized more reliably
and matched the industrial standard with an average inherent
viscosity of 5.17 0.12 dL g−1. Again, improving the reaction
conditions by upscaling and using more specialized equip-
ment can likely increase the inherent viscosities obtained for
most of the performed polymerizations. However, larger scale
experiments were beyond the scope of this work.
Polyaramid synthesis in [C8MIm][Cl]/GVL
Next, the syntheses of these polyaramids (PMIA, ODA/PPTA
and PPTA) were performed using OESs as solvents. The OESs
used in these initial experiments were mixtures of the IL
[C8MIm][Cl] and organic cosolvent GVL in varying mixing
ratios (Fig. 3). This combination of compounds was selected
because [C8MIm][Cl] has already shown potential in the syn-
thesis of PPTA,18,30 and because GVL is well-known for its
green properties. Furthermore, our previous work has shown
that [C8MIm][Cl]/GVL OESs have among the highest solubili-
ties for polyaramids. The inherent viscosities of the obtained
polymers are plotted versus molar fraction of [C8MIm][Cl] in
the OESs (Fig. 4). The inherent viscosities for the industrial
standards and the obtained benchmark values are displayed as
horizontal lines. For PMIA, only the reaction with the
maximum inherent viscosity was repeated in triplicate, yield-
ing a standard deviation of 0.064 dL g−1. For ODA/PPTA, mul-
tiple reactions around the maximum were repeated and the
variation in the results is displayed. For PPTA, no triplicates
were performed because of the poor performance of the OESs
in these polymerization reactions.
Pure GVL (xIL = 0.0) turned out to be a mediocre solvent for
PMIA synthesis, with the polymer having an inherent viscosity
of only ηinh = 0.78 dL g−1. After a short period of time, the
polymer precipitated from the reaction mixture, preventing the
polymerization reaction from proceeding. However, addition
of small amounts of IL rapidly increased the inherent viscos-
ities of the polymers. A maximum was reached at xIL = 0.043,
achieving an inherent viscosity of ηinh = 1.94 0.06 dL g−1
,
which is on par with the benchmark and the industrial stan-
dard. Further increasing the IL content of the OESs resulted in
declining inherent viscosities, eventually reaching ηinh = 0.35
dL g−1 for the pure IL. Even though pure [C8MIm][Cl] is an
excellent solvent for PMIA, it turned out to be a poor medium
for synthesis, because ILs have some drawbacks that can sub-
Fig. 4 Inherent viscosities of the polyaramids obtained from polymeriz-
ations performed in [C8MIm][Cl]/GVL as a function of the IL molar frac-
tions of the OES.
stantially impact the polymerization process. Firstly, there is
the typical high viscosity of ILs, which causes slow monomer
dissolution, poor homogeneity of the reaction mixture and
slow reaction kinetics. Secondly, ILs (especially imidazolium
Fig. 3 Chemical structures of GVL and [C8MIm][Cl]. The chemical
structures of other ILs used in this work are provided in the ESI.†
1232 | Green Chem., 2021, 23, 1228–1239
This journal is © The Royal Society of Chemistry 2021

View Article Online
Green Chemistry
Paper
chlorides) are very hygroscopic. Even though all compounds growing PPTA oligomers are hardly soluble in ILs.14 However,
were thoroughly dried prior to the polymerization reaction, this low solubility is not necessarily a big problem for this
some contact with the atmosphere is unavoidable using this reaction, as the current industrial solvent system NMP/CaCl2
small lab-scale reaction setup. The solvent can readily absorb cannot dissolve PPTA either. The synthesis of PPTA proceeds
moisture during these short periods of contact, which can de- differently compared to that of PMIA or ODA/PPTA. The
activate the acyl chloride monomers and reduce the solubility solvent does not function by forming stable PPTA solution, but
of the polymer. This explains why the increase in IL content in rather by delaying precipitation as long as possible to prolong
the synthesis solvents led to a decrease in the inherent viscos- the potential reaction time, leading to higher molecular
ities obtained. However, some IL is still needed because PMIA weight polymers. In this respect, the high viscosity of ILs is
is not soluble in pure GVL. The chloride anions of the ionic actually a benefit, slowing down precipitation and forming a
liquid are required to interact with the amide NH groups so as gel-state in which the polymer can still react if stirred
to prevent intermolecular hydrogen bonding between the properly.18,30 Because of this polymerization mechanism, the
polymer chains. It seems that a delicate balance must be main- reaction has to proceed very rapidly, which is not possible due
tained with respect to the amount of IL that has to be added to to the poor solubility of TDC in GVL. This poor solubility
the OES. A small amount of IL is required to keep the polymer caused an even bigger problem at lower IL molar fractions.
in solution, while at the same time not adding too much so as The unreacted TDC flakes would get coated in insoluble PPTA
to minimize the above-mentioned drawbacks, resulting in an oligomers, effectively terminating the reaction. Overall, it is
optimal IL molar fraction of xIL = 0.043.
evident that GVL is not a suitable co-solvent for the synthesis
For ODA/PPTA the plot has a similar shape. Again we see an of PPTA. Again, an alternative co-solvent should have sufficient
initial increase of the obtained inherent viscosities with solubility of TDC at low temperatures. Furthermore, a good
increased IL content, reaching a maximum at very low xIL, fol- distribution of TDC needs to be achieved by rapid stirring and
lowed by a sharp decline. The reactions around the maximum perhaps adding the monomer in a molten state. This ought to
were performed in triplicate, owing to the variability in the prevent the coating of unreacted TDC flakes, and allow for a
results from the benchmark reaction. The results are again rapid polymerization reaction in a gel-state.
fairly scattered, with the maximal inherent viscosity being ηinh
= 1.22 dL g−1. However, most reactions yielded inherent viscos-
In situ ionic liquid formation
ities much below the benchmark value of ηinh = 1.53 dL g−1
.
The highest inherent viscosity for PMIA was obtained using an
OES with an IL molar fraction of xIL = 0.043. This is actually a
very low IL content, one at which PMIA should not be soluble
in the concentration we used for the synthesis.32 Furthermore,
despite the fact that pure GVL is completely unable to dissolve
any PMIA, it was still able to keep the growing polymer in solu-
tion for a while. This is because [C8MIm][Cl] is not the only IL
present in the reaction mixture. During the polycondensation
reaction between the amine and acyl chloride groups, HCl is
released. This HCl is scavenged by the base α-picoline, creating
a protic IL, 2-methylpyridin-1-ium chloride ([2-Me-HPyr][Cl]),
in situ as is displayed in Fig. 5. This protic IL is always present
in a one-to-one ratio with the amide bonds of the polymers
and should definitely have an impact on the solubility. Liu
et al. made a similar discovery in their work on the synthesis
of the polyaramid poly–(benzimidazole–terephthalamide).44,45
They found that the addition of imidazole or sodium acetate
to scavenge in situ the released HCl greatly increased the solu-
bility of their polyaramids.
Several reasons can be given for these poorer results. Firstly,
the diacyl chloride monomer terephthaloyl chloride (TDC) is
poorly soluble in GVL. In the initial stages of the reaction, only
the small amount of TDC that is dissolved can partake in the
polymerization. The heat released from this exothermic reac-
tion ultimately causes the rest of the TDC to dissolve, but this
means the reaction mostly occurs well above the optimal temp-
erature range. This is important because at higher tempera-
tures some terminating side reactions can occur.40 Secondly,
as was the case for PMIA, there is the delicate balance between
the pros and cons of adding ILs to the OES. ODA/PPTA is
simply not soluble enough for synthesis in the lower xIL
ranges. Our previous work has shown that ODA/PPTA is highly
soluble in the xIL range of 0.2–0.6,32 but in this range all the
above-mentioned drawbacks of ILs are highly pertinent. Using
more specialized equipment that allows for more efficient heat
dissipation and the addition of TDC in a molten state can alle-
viate the first issue of poor solubility of TDC. However, solving
the issue of poor ODA/PPTA solubility will probably require a
different IL/co-solvent combination. The co-solvent should
have sufficient solubility of TDC at low temperatures, and the
overall OES needs to have an improved solubility of ODA/PPTA
compared to GVL/[C8MIm][Cl].
To study the impact of the in situ IL formation, three meta-
oriented oligomers were synthesized to emulate a growing
PMIA polymer chain. The oligomers used were: (1) a diacyl
The graph for PPTA did not display a maximum at low IL
molar fractions. The reaction mixture formed a precipitate
nearly instantly in all solvents except in the pure IL. PPTA is
notoriously difficult to dissolve and is only known to form
thermodynamically stable solutions in highly acidic or basic
Fig. 5 In situ formation of the protic ionic liquid [2-Me-HPyr][Cl] by
solvents, which are unsuitable for synthesis.41–43 Even the scavenging HCl by α-picoline.
This journal is © The Royal Society of Chemistry 2021
Green Chem., 2021, 23, 1228–1239 | 1233

View Article Online
Paper
Green Chemistry
Fig. 6 Chemical structures of polyaramid oligomers.
chloride-centred trimer (MTO), (2) a diamine-centred trimer much more basic pyridine derivative than α-picoline, and it is
(MTN), and (3) a diamine-centred pentamer (PTN) (Fig. 6). The often used as a catalyst for amidation reactions.46 2,6-Dichloro-
solubility of 10 wt% of these oligomers was tested in pure 4-nitropyridine on the other hand is a much less basic pyri-
GVL, the OES [C8MIm][Cl]/GVL with IL molar fraction xIL
=
dine derivative because of the electron-withdrawing nitro and
0.043 and in the same OES also containing [2-Me-HPyr][Cl] in chloro functionalities. No measured value for its pKa could be
a molar fraction of x[2-Me-HPyr][Cl] = 0.092, similar to the reaction found in the literature, but we can assume that it is less than
mixture after full conversion of the monomers. The results of 1.6, as that is the pKa value of 4-nitropyridine which does not
these solubility tests are shown in Fig. 7. Pure GVL was unable have the chloro substituents. Finally, triethylamine and
to dissolve any of the oligomers. The OES was able to dissolve 1-methylimidazole were used as they are both more basic than
MTO, but MTN formed a cloudy solution, and the pentamer α-picoline, but will form an ammonium chloride and imidazo-
PTN was found to be insoluble. These findings already indi- lium chloride in situ, respectively, rather than a pyridinium
cate that neither GVL nor the OES with xIL = 0.043 can keep a chloride. Results of the polymerizations are summarized in
growing polymer chain in solution. However when [2-Me- Table 1.
HPyr][Cl] was added, all the oligomers were completely
When no base was added, early precipitation of the reaction
soluble, indicating that this in situ formed IL provides the mixture occurred, terminating the polymerization prematurely.
majority of the dissolving capacity of the solvent. This explains No base means no in situ IL formation, and a reduced solubi-
why only a small amount of [C8MIm][Cl] is required for a suc- lity. Furthermore, if no base is present the HCl can protonate
cessful polymerization reaction. This initial small amount of
[C8MIm][Cl] is enough to keep some dimers and trimers in
solution during the early stages of the reaction. This allows for
the in situ formation of [2-Me-HPyr][Cl], which gradually
increases the dissolving capacity of the reaction mixture,
keeping the growing polymer chain in solution.
Table 1 pKa values and inherent viscosities of PMIA synthesized in
[C8MIm][Cl]/GVL xIL = 0.043 with different bases
Base
pKa
ηinh
Further evidence of the essential role of the in situ formed
IL was found by performing polymerization reactions by using
different organic bases or no added base at all. The bases used
were 4-dimethylaminopyridine (DMAP), 2,6-dichloro-4-nitro-
pyridine and 1-methylimidazole. DMAP was used as it is a
α-Picoline
6.0
7.0
9.7
10.8
—
<1.6
1.94
0.80
0.79
0.66
0.14
0.12
1-Methylimidazole
4-Dimethylaminopyridine
Triethylamine
None
2,6-Dichloro-4-nitropyridine
Fig. 7 Solubility tests of polyaramid oligomers. Left: solubility in only GVL, centre: solubility in [C8MIm][Cl]/GVL OES with xIL = 0.043, right: OES
also containing [2-Me-HPyr][Cl]. Solubility of the oligomers increases from left to right.
1234 | Green Chem., 2021, 23, 1228–1239
This journal is © The Royal Society of Chemistry 2021

View Article Online
Green Chemistry
Paper
the diamine monomers, which can slow down or perhaps even
The other ILs all yielded the same shape of curve as
terminate the polymerization reaction. Precipitation of the [C8MIm][Cl], showing a maximum at low xIL. The OES
reaction mixture was again observed when 2,6-dichloro-4-nitro- [C6Pyr][Cl]/GVL achieved a maximum inherent viscosity of 1.92
pyridine was added as a base. This pyridine derivative is not dL g−1
,
reaching the benchmark value and rivaling
only a weaker base than α-picoline, but it is also a weaker base [C8MIm][Cl]/GVL. The OESs made from [C4MIm][Cl] and [2-
than the diamine monomer, which has a pKa of 5.1. This Me-HPyr][Cl] achieved lower maximal inherent viscosities, at
means that the monomer is protonated preferentially over the 1.38 dL g−1 and 1.41 dL g−1, respectively. Interestingly, a
base, giving a situation which is similar to using no base at all. longer alkyl chain length in the cation led to higher inherent
4-Dimethylaminopyridine is a stronger base than α-picoline, viscosities. A similar trend was observed in our earlier work on
but it was not completely soluble in the reaction mixture, the synthesis of PPTA in IL/NMP mixtures.30 Cations with a
resulting in a lower inherent viscosity. Triethylamine and longer alkyl chain length can more effectively shield different
1-methylimidazole are also stronger bases than α-picoline, but polymer chains via steric hindrance. Furthermore, longer alkyl
also these bases yielded poorer results. Evidently, the pKa of chains may also reduce the hygroscopicity and the viscosity of
the base is not directly correlated to the inherent viscosity of the solvent. However, although the ILs with longer alkyl chains
the polyaramids obtained, as long as the basicity is high did give better results, these other ILs should not be dis-
enough to effectively scavenge all the HCl. The inherent viscos- carded. As mentioned above, upscaling and using more
ities obtained for the different bases follow the trend: specialized equipment can lead an improvement of the results.
α-picoline > 1-methylimidazole > triethylamine. This trend is Thus, although the OES [2-Me-HPyr][Cl]/GVL underperforms
in accordance with that of the solubility of polyaramids in pyri- compared to NMP, it might still be worthwhile to optimize the
dinium chlorides > imidazolium chlorides > ammonium reaction for this solvent at an industrial scale, as this would
chlorides.
greatly simplify the purification and recycling of the solvent.
In all our syntheses of PMIA, the monomers were added in
an amount such that the final polymer concentration would be
10 wt%. However, current synthetic and spinning procedures
use a polymer concentration of 20 wt%, because faster pro-
duction can be achieved at a lower cost in this way. We there-
fore performed several polymerization reactions in the
[C8MIm][Cl]/GVL OES around the optimal IL molar fraction,
using a polymer concentration of 20 wt%. We expected that
this would lead to decreased inherent viscosities, as the solubi-
lity limit of the OES might be surpassed. However, we actually
observed a gradual increase in inherent viscosity for almost all
IL molar fractions, and the peak was much broader making
the IL/cosolvent balance less delicate compared to 10 wt%
(Fig. 9). The fact that we did not observe any polymer precipi-
tation is a result of the in situ IL formation mentioned above.
Influence of altering the ILs and the polymer concentration
for PMIA synthesis
Considering the fact that the in situ formed IL [2-Me-HPyr][Cl]
seems to be chief driving force for dissolution of the polymers,
OESs were made from ILs other than [C8MIm][Cl] to investi-
gate how this would influence the obtained inherent viscos-
ities: [C4MIm][Cl] which has a shorter side chain compared to
[C8MIm][Cl], rendering it cheaper and less energy intensive to
produce; [C6Pyr][Cl] which has a pyridinium cation instead of
an imidazolium; and finally the in situ formed IL [2-Me-
HPyr][Cl] itself was tested, as using this as an IL would greatly
simplify any post-reaction recycling steps. Several PMIA
polymerization reactions were performed in the lower xIL range
around the optimum for [C8MIm][Cl]/GVL, the results of these
reactions are displayed in Fig. 8.
Fig. 9 Inherent viscosities of PMIA polymerization reaction performed
at 20 wt% vs. 10 wt% polymer concentration.
Fig. 8 PMIA polymerization reactions in several IL/GVL OESs.
This journal is © The Royal Society of Chemistry 2021
Green Chem., 2021, 23, 1228–1239 | 1235

View Article Online
Paper
Green Chemistry
As the polymer concentration was increased, the amount of
α-picoline was also increased by an equal amount. This means
more IL was generated in situ and the dissolution capacity of
the solvent increased. Moreover, increasing the polymer con-
centration has the advantage that some of the issues related to
small-scale polycondensation reactions are mitigated. As the
monomers were added in larger amounts, the magnitude of
the effect of weighing/addition inaccuracies is decreased.
Furthermore, the amount of water absorbed from the atmo-
sphere remains the same, while the amount of acyl chloride
has been doubled. As a result there is less water present rela-
tive to the amount of acyl chloride. This means that the nega-
tive impact of monomer impurity, as a result of the side reac-
tion between water and the acyl chloride, on the inherent vis-
cosity is reduced. So it is not only possible but beneficial to
perform the synthesis at the same concentration as in the
current industrial procedures. Furthermore, this experiment
may serve as a demonstration that upscaling can indeed lead
to better results for polycondensation reactions.
Direct spinning and solvent regeneration
In our previous work we have shown that it is possible to
create PMIA and ODA/PPTA fibers from GVL-based OESs
through wet spinning.32 From an economical point of view, it
would be beneficial to be able to spin the polymer directly
from the synthetic solution, and to be able to purify and
recycle the solvent after spinning. Therefore, polymerizations
of PMIA were performed using an [C8MIm][Cl]/GVL OESs with
a molar fraction of xIL = 0.043 and a polymer concentration of
10 wt% and 20 wt%. After the polymerization reaction was fin-
ished, the reaction mixtures were allowed to settle for about
30 minutes to allow air bubbles to escape. The reaction mix-
tures were then transferred to a syringe, from which a fiber
could be spun using a syringe pump and a 3D-printed coagu-
lation bath and winding apparatus (Fig. 10). The spinning
occurred at room temperature and the coagulation bath con-
tained distilled water. Both reaction mixtures were readily spin-
nable, giving uniform white fibers that could easily be pulled
from the coagulation bath with the winding setup.
Characterization of the mechanical properties of the fibers was
beyond the scope of this work, but this experiment gives the
proof-of-principle that this environmentally friendly OES can
be used for both synthesis and the direct spinning from the
reaction mixture. This is a great improvement from an environ-
mental and economical viewpoint compared to the current
industrial syntheses using harmful solvents such as NMP or
DMA, or the interphase type polymerizations which require
precipitation and redissolving in NMP at a later stage.
ODA/PPTA could not be spun directly after synthesis. The
reaction mixture was not a liquid, but rather a viscoelastic gel
that was very difficult to handle. When trying to transfer this
gel to a syringe it would start to break up as it absorbed moist-
ure from the atmosphere, and the polymer would start to pre-
cipitate. Even if upscaling and optimizing would result in a
Fig. 10 Direct spinning of PMIA from the reaction mixture. Details on
the spinning setup and spinning parameters can be found in the ESI S6.†
After spinning, the content of the coagulation bath was
higher inherent viscosity for ODA/PPTA, it is unlikely that this transferred to a round-bottom flask. Water and GVL were evap-
reaction mixture could be used for direct spinning.
orated together using a rotary evaporator operating at a temp-
1236 | Green Chem., 2021, 23, 1228–1239
This journal is © The Royal Society of Chemistry 2021

View Article Online
Green Chemistry
Paper
erature of 60 °C and a pressure of 65 mbar. It was necessary to
[2-Me-HPyr][Cl] was thus removed using a short path evap-
remove water and GVL together because of the presence of ILs. orator (Kugelrohr apparatus), operating at a temperature of
These can increase the boiling point of water closer to that of 100 °C and a pressure of 1 mbar. The progress of the distilla-
GVL and making selective removal more challenging.47 Once tion process was monitored by periodically sampling and
1
removed from the ILs, the GVL and water can easily be separ- recording H NMR spectra (Fig. 12). After 12 hours, all traces
ated via distillation.32,48 Finally, in order to regenerate the of [2-Me-HPyr][Cl] were removed and [C8MIm][Cl] was recov-
expensive [C8MIm][Cl], it needs to be removed from the ered with a yield of 89%. If wanted, the [2-Me-HPyr][Cl] can be
mixture with [2-Me-HPyr][Cl]. A useful difference between converted back into α-picoline by the addition of a stronger
protic ILs and other ILs is the fact that protic ILs can be base. However, an upscaled version of this short path distilla-
removed with short path distillation in a Kugelrohr appar- tion step will consume a large amount of energy. An ideal situ-
atus.49 The acid–base reaction that forms these protic ILs is ation would be if the polymerization process in [2-Me-
reversible, meaning that the protic cation (a conjugate acid) HPyr][Cl]/GVL were to be optimized on an industrial scale.
and the anion (a conjugate base) can revert back to the original This would eliminate the need to separate [2-Me-HPyr][Cl]
base (α-picoline) and acid (HCl). These are neutral compounds from another IL. After evaporating water and GVL, the in situ
which can be distilled off. Once condensed, the base and acid formed [2-Me-HPyr][Cl] can be directly used to prepare a new
react again to form [2-Me-HPyr][Cl] (Fig. 11).
OES.
Fig. 11 Separation of [C8MIm][Cl] and [2-Me-HPyr][Cl] via short path distillation in a Kugelrohr apparatus.
Fig. 12 Separation of [C8MIm][Cl] and [2-Me-HPyr][Cl] monitored by 1H NMR. After 12h, all the [2-Me-HPyr][Cl] was found to be removed.
This journal is © The Royal Society of Chemistry 2021
Green Chem., 2021, 23, 1228–1239 | 1237

View Article Online
Paper
Green Chemistry
Conclusion
References
The three most commonly produced polyaramids (PMIA,
ODA/PPTA and PPTA) were synthesized using GVL-based
organic electrolyte solutions (OESs) as benign alternative sol-
vents. The results from these polymerizations were compared
to those of benchmark reactions which were performed
using the established solvent NMP. The OESs were not suit-
able solvents for PPTA synthesis, as oligomers instantly pre-
cipitated owing to the low solubility of PPTA in these solvent
systems. ODA/PPTA could be synthesized in OESs, but the
results were mediocre compared to those obtained in NMP.
The poor results can be attributed to the low solubility of
the diacyl chloride monomer and of ODA/PPTA itself.
However, the performance of the OESs was similar to that of
NMP in the synthesis of the m-aramid PMIA if: (1) the molar
1 L. S. Kwolek, M. P. Winthrop and S. W. Richard, Process of
making wholly aromatic polyamides, US Patent, 3063966,
1962.
2 L. S. Kwolek, Optically Anisotropic Aromatic Polyamide
Dopes, US Patent, 3671542, 1972.
3 S. Ozawa, Y. Nakagawa, K. Matsuda, T. Nishihara and
H. Yunoki, Novel Aromatic Copolyamides Prepared From
3,4′-Diphenylene Type Diamines, And Shaped Articles
Therefrom, US Patent, 4075172, 1975.
4 M. Trigo-López, J. M. García, J. A. R. Ruiz, F. C. García and
R. Ferrer, Aromatic Polyamides, in Encyclopedia of Polymer
Science and Technology, 2018, pp. 1–51, DOI: 10.1002/
0471440264.pst249.pub2.
5 J. R. Brown and B. C. Ennis, Text. Res. J., 1977, 47(1), 62–66.
6 L. Vollbracht and T. J. Veerman, Process for the preparation
of poly-p-phenyleneterephthalamide, US Patent, 4308374,
1981.
7 M. Arpin and C. Strazielle, Polymer, 1977, 18(6), 591–598.
8 H. S. Morgan, Process for spinning wholly aromatic polya-
mide fibers, US Patent, 3414645, 1968.
fraction of IL in the OES was small, but not zero (xIL
=
0.043), (2) α-picoline was added to scavenge HCl and (3) the
IL used was either [C8MIm][Cl] or [C6Pyr][Cl]. The polymeriz-
ation reaction was found to be very sensitive to the amount
of IL added to the OES. The IL has properties that can be
both beneficial (increased solubility) and disadvantageous
(high viscosity and hygroscopicity) for the polymerization
9 S. Ozawa, Polym. J., 1987, 19(1), 119–125.
reaction and a very delicate balance exists between these 10 Scientific Committee on Consumer Safety (SCCS), “Opinion
factors. This balance became less delicate when the polymer
concentration in the reaction was increased from 10 wt% to
20 wt%, resulting in a broader tolerance for IL content in
on N-Methyl-2-pyrrolidone (NMP)”, European Union. 2011,
https://ec.Europa.eu/health/sites/health/files/scientific_
committees/consumer_safety/docs/sccs_o_050.pdf.
the OES. The organic base added to the reaction mixture to 11 P. W. Morgan, Condensation polymers: by interfacial and
scavenge HCl is also an important factor in the polymeriz- solution methods, Interscience Publishers, New York, 1965.
ation reactions. This base forms a chloride IL in situ which 12 I. In and S. Y. Kim, Polymer, 2006, 47(2), 547–552.
partakes in the dissolution process of the polymer. 13 Z. Yu, R. Deng, G. Rao, Y. Lu, Y. Wei, C. Fu, X. Yan, F. Cao
α-Picoline thus outperformed other bases, as this forms a
pyridinium chloride IL ([2-Me-HPyr][Cl]), which is known to 14 S. Dewilde, W. Dehaen and K. Binnemans, Green Chem.,
be a strong solvent for polyaramids. The reaction mixtures 2016, 18(6), 1639–1652.
could directly be used for wet spinning fibers without the 15 J. Winters, W. Dehaen and K. Binnemans, Phys. Chem.
need for isolating and redissolving the PMIA polymer. Chem. Phys., 2019, 21(7), 4053–4062.
Furthermore, it was possible to separate and recycle the 16 W. Yang, H. Yu, M. Zhu, H. Bai and Y. Chen, J. Macromol.
components of the OES from the coagulation bath, making Sci., Part B: Phys., 2006, 45(4), 573–579.
these solvent systems promising candidates to replace NMP 17 T. Zhao, H. Wang, Y. Zhang, B. Wang and J. Jiang,
and D. Zhang, J. Polym. Sci., 2020, 58(15), 2140–2150.
in PMIA synthesis and processing.
Int. J. Mol. Sci., 2007, 8(7), 680–685.
18 S. Dewilde, T. Vander Hoogerstraete, W. Dehaen and
K. Binnemans, ACS Sustainable Chem. Eng., 2018, 6(1),
1362–1369.
19 Y. S. Vygodskii, E. I. Lozinskaya and A. S. Shaplov,
Macromol. Rapid Commun., 2002, 23(12), 676–680.
20 S. Zahmatkesh, Int. Scholarly Res. Not., 2012, 2012, 1–7.
21 S. Mallakpour and H. Seyedjamali, Polym. Bull., 2009, 62(5),
605–614.
Conflicts of interest
There are no conflicts of interest to declare.
22 M. Kar, N. V. Plechkova, K. R. Seddon, J. M. Pringle and
D. R. MacFarlane, Aust. J. Chem., 2019, 72(2), 3–10.
23 P. Kubisa, Prog. Polym. Sci., 2009, 34(12), 1333–1347.
Acknowledgements
This research was funded by a doctoral (PhD) strategic basic 24 A. Maiti, A. Kumar and R. D. Rogers, Phys. Chem. Chem.
research grant of the Research Foundation Flanders (FWO) to Phys., 2012, 14(15), 5139–5146.
Jonas Winters (1S12717N). The help of Jakob Bussé in the 25 R. K. Seddon, A. Stark and M.-J. Torres, Pure Appl. Chem.,
design and building of the 3D-printed spinning setup is
greatly acknowledged.
2000, 72(12), 2275–2287.
26 R. Rinaldi, Chem. Commun., 2011, 47(1), 511–513.
1238 | Green Chem., 2021, 23, 1228–1239
This journal is © The Royal Society of Chemistry 2021

View Article Online
Green Chemistry
Paper
27 E. Gale, R. H. Wirawan, R. L. Silveira, C. S. Pereira,
M. A. Johns, M. S. Skaf and J. L. Scott, ACS Sustainable
Chem. Eng., 2016, 4(11), 6200–6207.
diamine and molten terephthaloyl chloride, US Patent,
3850888, 1974.
39 J. M. Dealy and T. K. P. Vu, J. Nonnewton. Fluid Mech., 1977,
3(2), 127–140.
28 D. L. Minnick, R. A. Flores, M. R. DeStefano and
A. M. Scurto, J. Phys. Chem. B, 2016, 120(32), 7906– 40 M. Jassal and S. Ghosh, Indian J. Fibre Text. Res., 2002, 27,
7919.
290–306.
29 M. T. Clough, Green Chem., 2017, 19(20), 4754–4768.
30 S. Dewilde, J. Winters, W. Dehaen and K. Binnemans,
Macromolecules, 2017, 50(8), 3089–3100.
41 N. Ogata, K. Sanui and S. Kitayama, J. Polym. Sci., Part A:
Polym. Chem., 1984, 22(3), 865–867.
42 R. R. Burch, W. Sweeny, H. W. Schmidt and Y. H. Kim,
Macromolecules, 1990, 23(4), 1065–1072.
31 B. S. Pascual, M. Trigo-López, J. A. Reglero Ruiz,
J. L. Pablos, J. C. Bertolín, C. Represa, J. V. Cuevas, 43 S. J. Picken, S. van der Zwaag and M. G. Northolt, Polymer,
F. C. García and J. M. García, Eur. Polym. J., 2019, 116, 91–
98.
32 J. Winters, W. Dehaen and K. Binnemans, Green Chem.,
2020, 22(18), 6127–6136.
33 W. R. H. Wright and R. Palkovits, ChemSusChem, 2012,
5(9), 1657–1667.
34 D. M. Alonso, S. G. Wettstein and J. A. Dumesic, Green
Chem., 2013, 15(3), 584–595.
1992, 33(14), 2998–3006.
44 Y. Dai, Z. Cheng, Y. Yuan, C. Meng, J. Qin and X. Liu,
ChemPhysChem, 2018, 19(19), 2468–2471.
45 Y. Dai, Z. Cheng, C. Meng, Y. Yuan, L. Luo and X. Liu,
ChemistrySelect, 2019, 4(1), 123–129.
46 D. J. Berry, C. V. DiGiovanna, S. S. Metrick and
R. Murugan, ARKIVOC, 2001, 2, 944–964.
47 Z. Lei, C. Dai, J. Zhu and B. Chen, AIChE J., 2014, 60(9),
3312–3329.
35 R. A. Sheldon, Green Chem., 2014, 16(3), 950–963.
36 S. M. C. Peixoto, J. R. Nunhez and C. G. Duarte, 48 D. Havasi, P. Mizsey and L. T. Mika, J. Chem. Eng. Data,
Braz. J. Chem. Eng., 2000, 17, 925–936.
2016, 61(4), 1502–1508.
37 W. H. Carothers, Trans. Faraday Soc., 1936, 32, 39–49.
38 J. Fitzgerald and K. Likhyani, Process for producing poly-p-
phenylene terephthalamide from solution of p-phenylene
49 A. W. T. King, J. Asikkala, I. Mutikainen, P. Järvi and
I. Kilpeläinen, Angew. Chem., Int. Ed., 2011, 50(28), 6301–
6305.
This journal is © The Royal Society of Chemistry 2021
Green Chem., 2021, 23, 1228–1239 | 1239