Anionic ring-opening homo- and copolymerization of lactams by latent, protected N-heterocyclic carbenes for the preparation of PA 12 and PA 6/12

Article abstract of DOI:10.1021/ma4018586

Laurolactam (LL) is polymerized in the bulk using strongly basic N-heterocyclic carbenes (NHCs) as initiators at temperatures of 180-200 C to prepare the corresponding polyamide (PA 12). In-situ rheology of the polymerization progress reveals that an anionic mechanism is active, which is supported by the strong dependence of initiator activity on the basicity of the NHCs. GPC data and kinetic investigations show the process to be moderately controlled and fast, allowing high or quantitative yields in short polymerization times. Fifteen different NHC-CO₂-adducts and NHC-metal complexes were used as thermally labile but room temperature stable NHC-precursors. Depending on the ring size and N-substituent, some of these protected NHCs allow forming a mixture of monomer and NHC-precursor that is suitable for long-term storage and readily polymerizable by simple heating. All polymerizations are executed without activator or other additives and thus represent a true one-component system for the production of PA 12. Finally, LL is copolymerized with ε-caprolactam (ε-CLA). It is found that a copolymer with a considerable gradient is formed, with ε-CLA being incorporated preferentially at the onset of the polymerization.

Full text of DOI:10.1021/ma4018586

Article
pubs.acs.org/Macromolecules
Anionic Ring-Opening Homo- and Copolymerization of Lactams by
Latent, Protected N‑Heterocyclic Carbenes for the Preparation of PA
12 and PA 6/12
Stefan Naumann,† Friedrich Georg Schmidt,⊥ Maria Speiser,† Margit Bohl,‡ Stefan Epple,§
̈
Christian Bonten,§ and Michael R. Buchmeiser†,‡,*
†Institute of Polymer Chemistry, University of Stuttgart, Pfaffenwaldring 55, D-70569 Stuttgart, Germany
⊥Evonik Industries AG, Paul-Baumann-Strasse 1, D-45764 Marl, Germany
‡Institute of Textile Chemistry and Chemical Fibers, Korschtalstrasse 26, D-73770 Denkendorf, Germany
̈
§Institut fur Kunststofftechnik, University of Stuttgart, Boblingerstrasse 70, D-70199 Stuttgart, Germany
̈
̈
S
* Supporting Information
ABSTRACT: Laurolactam (LL) is polymerized in the bulk
using strongly basic N-heterocyclic carbenes (NHCs) as
initiators at temperatures of 180−200 °C to prepare the
corresponding polyamide (PA 12). In-situ rheology of the
polymerization progress reveals that an anionic mechanism is
active, which is supported by the strong dependence of
initiator activity on the basicity of the NHCs. GPC data and
kinetic investigations show the process to be moderately
controlled and fast, allowing high or quantitative yields in short polymerization times. Fifteen different NHC−CO2−adducts and
NHC−metal complexes were used as thermally labile but room temperature stable NHC-precursors. Depending on the ring size
and N-substituent, some of these protected NHCs allow forming a mixture of monomer and NHC-precursor that is suitable for
long-term storage and readily polymerizable by simple heating. All polymerizations are executed without activator or other
additives and thus represent a true one-component system for the production of PA 12. Finally, LL is copolymerized with ε-
caprolactam (ε-CLA). It is found that a copolymer with a considerable gradient is formed, with ε-CLA being incorporated
preferentially at the onset of the polymerization.
However, disadvantages of free NHCs are the limited
stability and storability and also the awkward handling of free
carbenes since they are often of oily consistence. Consequently
much effort has been put into research concerning compounds
that exist in the form of adducts or complexes but can release
the active NHC after warming (“protected NHCs”) or by
solvent effects. Examples for this type of chemistry are azolium
hydrogenates,13 CO2-, COS- or CS2-adducts,14 Sn-, Zn- or Mg-
based complexes,15 and isothiocyanate-,16 alcohol-17 and
chloroform-adducts.18 A range of differently substituted CO2-
and metal-salt protected NHCs was therefore synthesized
(Figure 1). These initiators have been used in the polymer-
ization of methyl methacrylate19 or ε-caprolactone,20 where in
both cases no reaction at room temperature occurred, but could
be induced by heating. Compounds of this type were also used
as latent precatalysts for the synthesis of polyurethanes.15
Recently, we also extended this principle to ε-caprolactam (ε-
CLA).21 It was demonstrated for the first time that NHCs can
effectively deliver poly(ε-caprolactam) (PA 6), if they are basic
INTRODUCTION
■
Far from being mere ligands for transition metals, N-
heterocyclic carbenes (NHCs) are ever more recognized as
powerful organocatalysts on their own.1 Because of their
nucleophilicity2 they can bring about a wealth of organic
transformations, like umpolung reactions or transesterifica-
tions.3 NHCs are also bases, though there are much less
examples of their application as such,4 which is probably related
to the frequent use of imidazol-2-ylidenes and imidazolin-2-
ylidenes. These compounds are much weaker bases than for
example tetrahydropyrimidin-2-ylidenes.5 Both types of NHCs
are part of this work, and special emphasis is put on their use as
bases. Combined with the aforementioned properties, the
flexibility with regard to ring-size, N-substituents, and back-
bone, as well as the relative ease of preparation,6 have won
NHCs a prominent position in polymerization chemistry.7
Among others, methyl methacrylate was polymerized both
directly or via a group transfer mechanism.8 Epoxides,9
lactones,10 lactides,11 and cyclosiloxanes12 were successfully
ring-opened to give the corresponding polymers. In most cases
excellent yields and good control of polydispersity were
achieved.
Received: September 6, 2013
Revised: October 10, 2013
© XXXX American Chemical Society
A
dx.doi.org/10.1021/ma4018586 | Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
Figure 1. NHC-carboxylates and NHC-metal complexes used in the present work. Brackets: pKa-values of the parent NHCs (aqueous solution at 25
°C).31 Cy = cyclohexyl, Mes =2,4,6-trimethylphenyl, Dipp =2,6-diisopropyl-phenyl.
enough. This rendered especially N-alkyl bearing tetrahydro-
pryimidinium- based initiators the most suitable. The polymer-
izations displayed characteristics typical for anionic lactam
polymerization and yields were high, though not quantitative.
These promising results indicate that NHC-catalysis may be
applied to other lactams as well and has the potential to
compete successfully with other organic-based initiators like
phosphazenes.22 Accordingly, in the present work the NHC-
initiated homopolymerization of laurolactam (LL) to PA 12 as
well as the copolymerization of LL and ε-CLA are presented
(Scheme 1). Its many methylene linkages and accordingly low
amide density make PA 12 a much less polar polyamide than
PA 6. The consequence is that PA 12 is much less prone to
water uptake than PA 6, a property that is much desired since it
promotes dimensional stability in a polymer body.23 The
downside is the lower melting point (Tm = 180 °C) compared
to PA 6 (Tm = 225 °C). As a result, copolymers of LL and ε-
CLA have been investigated to combine high melting points
with lowest possible water uptake,24 resulting in various types
of commercially important PA 6/12.25
Mechanistically, LL can be expected to polymerize following
the well-known anionic ring-opening of ε-CLA,26 as depicted in
Scheme 1, where it is proposed that the N-heterocyclic carbene
is acting as base. Generally, anionic lactam polymerization
displays an induction time, but is in sum much faster than
hydrolytic polymerization of lactams.26 This allows the use of
reactive extrusion techniques, where usually the monomeric
melt is combined with initiator/catalyst (organic or inorganic
bases) and activator (such as N-acyl lactam or cyclohexyl
isocyanate) and polymerized within seconds or few minutes to
the polyamide. A reduction of necessary components
considerably simplifies the technical realization, and for PA
12 the reduction to a two-component system combined with
Scheme 1. Thermal deprotection of the initiator (top)
induces anionic ROP of LL
reactive extrusion has been successfully executed by Manson et
̊
al.27 The system consisting solely of a protected NHC and LL
discussed below is thought to be applied in this kind of
technique, the more so since it allows a further reduction of
polymerization components.
EXPERIMENTAL SECTION
■
The synthesis of the NHC-precursors was performed without special
precautions under air. All manipulations involving free carbenes, i.e.
the formation of NHC-carboxylates, NHC-metal salt complexes and
polymerizations, were conducted in a nitrogen-filled glovebox
(MBraun Lab Master 130, MBraun, Garching, Germany) or by
standard Schlenk techniques. CH2Cl2, diethyl ether, toluene, pentane
and THF were dried by a solvent purification system (SPS, MBraun).
KOtBu was purified by sublimation; all other starting materials and
chemicals were purchased from Sigma-Aldrich, TCI Europe, ABCR or
STREM and used without further purification. ε-Caprolactam was a
gift from L. Bruggemann KG (Heilbronn, Germany). The sample was
̈
a commercial grade monomer designated for anionic polymerization
(dried, packed in airtight foliage). Laurolactam was bought from
B
dx.doi.org/10.1021/ma4018586 | Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
Scheme 2. Preparation of the Initiators via Amidines (Top) or via Cyclization of Diamines (middle) and Synthesis of
a
Diazepinium-Derivative 1 (Bottom)
a
This compound decomposed when reacted with KHMDS.
Aldrich (98%). Both lactams were stored under nitrogen and used
without further purification. All noncommercial amidines, NHC-
precursors, NHC-carboxylates and NHC-metal complexes were
synthesized as described in the literature.19,20,28 NMR measurements
were recorded on a Bruker Avance III 400. IR spectra were measured
on an IFS 28 (Bruker) using KBr pellets or via ATR on a Bruker Alpha
instrument. GC-MS data were recorded on an Agilent Technologies
device consisting of a 7693 autosampler, a 7890A GC and a 5975C
quadrupole MS. An SPB-5 fused silica column (34.13 m × 0.25 mm ×
0.25 μm film thickness) was used. For GPC, a column set by PSS
Polymer Standard Service (Mainz, Germany) was used (precolumn
PSS SDV, 20 μm, 8 × 50 mm, column SDA 0830201E3, PSS SDV,
1000 Å, 20 μm, 8 × 300 mm, column SDA 0803201E5, PSS SDV, 10
000 Å, 20 μm, 8 × 300 mm). The GPC system was maintained at 70
°C and a flow rate of 1 mL/min using PS standards and m-cresol as a
solvent. DSC measurements were conducted using a DSC 4000 by
Perkin-Elmer (software: Pyris) and heating rates of 20 °C/min in a
range of −50 to +250 °C. Rheology was measured using an RMS 800
(Rheometric scientific). During the measurements, a nitrogen
atmosphere was applied. Prior to addition of the premixed initiator/
monomer-sample the plates were heated to the desired reaction
temperature. About 2 mL of melt was necessary for each measurement.
2,4-Dicyclohexyl-4,5-dihydro-1H-benzo-1,3-diazepinium
Bromide (1). N,N′-Di(cyclohexyl)formamidine19 (14 mmol, 2.92 g),
1,2-bis(bromomethyl)benzene (15.53 mmol, 4.10 g, 1.1 equiv), and
potassium carbonate (7.16 mmol, 0.99 g, 0.5 equiv) were suspended in
130 mL of acetonitrile. The mixture was refluxed for 5 h, while the
reaction progress was monitored via 1H NMR. Then all volatiles were
removed by evaporation, and the residue was dissolved in dichloro-
methane. After filtration, diethyl ether was added to the clear solution,
which resulted in the preciptation of a white solid. Yield: 5 g (12.8
1
mmol, 91%). H NMR (CDCl3, δ): 8.85 (s, 1H), 7.36 (m, 2H), 7.20
(m, 2H), 4.73 (s, 4H), 4.05 (m, 2H), 1.08−2.00 (m, 20H). 13C NMR
(CDCl3, δ): 154.5, 134.8, 129.9, 128.2, 67.6, 47.3, 30.9, 24.94, 24.86.
IR (ATR, cm−1): 2928 (s), 2856 (m), 1660 (s), 1449 (m), 1072 (m),
752 (s), 610 (m). HRMS (ESI) m/z calcd for C21H31N2Br, 390.1671;
found, 311.2489 [M − Br−]+.
Homopolymerization of LL. For polymerization, pellets of LL
(500−2000 mg) were combined with the desired amount of initiator
(5−15 mg) and transferred into a Schlenk vessel, which was sealed,
removed from the glovebox, and submerged in a preheated oil bath.
After the required reaction time, the polymerization was quenched by
rapid cooling. The solid polymeric mass was then dissolved in hot m-
cresol and precipitated from acetone. After the product was washed
three to four times with acetone, an off-white polymer was isolated and
dried in vacuo. All yields given in this work are isolated yields as
calculated from the mass of polymer and mass of monomer.
For kinetic experiments, a master batch was created by mixing
monomer and initiator using 5−15 g of LL. Both were finely
homogenized using mortar and pestle to give a white powder. From
this master batch, smaller potions (0.75 g) were taken to conduct
polymerizations of varying reaction times. To ensure identical
conditions, for each experiment, the same type of vessel and stirrer
was used, as well as the same stirring rate.
Copolymerization of LL and ε-CLA. For copolymerization,
initiator, pellets of LL, and flakes of ε-CLA were combined in a
Schlenk vessel. This was sealed, removed from the glovebox and
submerged in a preheated oil bath. Again the reaction was stopped by
C
dx.doi.org/10.1021/ma4018586 | Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
rapid cooling. Dissolving in hot m-cresol was possible and work-up
proceeded as described above for the homopolymerizations.
To identify the copolymer composition, the polymer yield was
calculated. After that, the residual monomer ratio of LL and ε-CLA in
the washing solution (m-cresol enriched acetone) was investigated
using GC−MS analysis. The found ratios were compared to a
calibration curve using samples containing 25, 50, 67, and 75 mol % of
LL. Knowledge of the residual amounts of monomer then allowed for
calculating the copolymer composition.
Table 1. Bulk Polymerizations of LL Using Protected NHCs
at T = 180 °C
a
b
c
no.
NHC
NHC:LL
yield [%]
Mn(PDI) [g/mol]
1
−
pure LL
1:100
1:300
1:500
1:100
1:100
1:100
1:100
1:100
1:100
1:100
1:100
1:100
1:100
1:100
1:100
1:100
1:100
−
100
100
77
−
2
6-Cy-CO2
7100 (2.6)
3
6-Cy-CO2
13900 (2.4)
4
6-Cy-CO2
13200 (2.4)
5
6-Hep-CO2
6-iPr-CO2
93
6700 (2.3)
6
82
8500 (2.9)
RESULTS AND DISCUSSION
■
7
6-Mes-CO2
6-Dipp-CO2
6-OMe-CO2
5s-tBu-CO2
5-tBu-CO2
5-Cy-CO2
<1
<1
<1
96
−
−
−
Synthesis of Initiators. To provide a fast and simple
nomenclature, the protected NHCs used in this work are
named according to the sequence “ring size-substituent-
protecting group”. So, 5s-tBu-CO2 is representing 1,3-di(tert-
butyl)-imidazolinium-2-carboxylate. “5s” refers to a saturated
five-membered ring and thus differentiates imidazolinium- from
imidazolium derivatives. Free carbenes like 1,3-di(cyclochexyl)-
imidazol-2-ylidene would be abbreviated as 5-Cy.
8
9
10
11
12
13
14
15
16
17
18
7000 (2.6)
71
7600 (2.8)
23
11600 (3.1)
5-iPr-CO2
<5
<1
<1
<1
<1
−
−
−
−
−
−
7-Mes-CO2
7-Dipp-CO2
5s-Mes-MgCl2
5s-Mes-ZnCl2
5s-Mes-SnCl2
The synthetic routes (Scheme 2) that were used to prepare
the target compounds (Scheme 1) are well-known in the
literature and combine a simple synthesis with a minimal
amount of steps while maintaining a high flexibility with regard
to N-substituents and ring size.19,20,28
<1
a
b
c
Heating time =45 min. Isolated yield. GPC (m-cresol) relative to
PS standards.
Readily available primary amines served as starting material
for amidines, which were reacted with dihaloalkanes to yield the
cyclic NHC-precursors (6-Mes-CO2, 6-Dipp-CO2, 6-Hep-
CO2, 6-Cy-CO2, 6-OMe-CO2, 7-Mes-CO2, and 7-Dipp-
CO2). Alternatively, commercially accessible diamines were
used for cyclization (5s-tBu-CO2, 6-iPr-CO2). With exception
of the generation of the free carbenes and concomitant reaction
with either CO2 or the metal salt, no protective atmosphere was
necessary. In all cases the initiators were synthesized in a
maximum of three steps. In case the NHC-precursors were
commercial products, only one step delivered the desired
compounds (e.g., 5s-Mes-MCl2 (M = Mg, Zn, Sn)). Though
the methods shown in Scheme 2 generally delivered a broad
range of protected NHCs, it was not possible to prepare the
CO2-adduct of compound 1 this way. Though 1 can easily be
prepared in high yields, upon addition of a base (potassium
hexamethyldisilazide or KHMDS) to a suspension of 1 in THF,
decomposition was observed. Similar behavior has been
described for comparable benzannulated diazepinium-derived
compounds.29
yields even at low initiator loadings sharply contrast with
reactions delivering absolutely no PA 12. The distinctly best
results were obtained using 6-Cy-CO2, and generally N-alkyl
bearing initiators performed well. As was shown for the NHC-
mediated synthesis of PA 6,21 the activity of the protected
NHC can clearly be related to the basicity of the corresponding
free carbene. The correlation of ring size and basicity for NHCs
is 7 ≈ 6 ≫ 5s > 5, reflecting the fact that basic properties
mainly depend on the N−C−N angle.30 Furthermore, N-aryl-
instead of N-alkyl substitution considerably diminishes the pKa
values, which span a range of roughly 21 (5s-Mes: 21.3) to 28
(6-iPr: 28.2) in H2O at 25 °C.31 Accordingly, for the NHCs
presented here, an order of basicity of 6-Cy ≈ 6-Hep ≈ 6-iPr
≫ 5s-tBu > 5-tBu > 5-Cy ≈ 5-iPr ≫ 5s-Mes can be expected.
For the inactivity of 7-Mes-CO2 and 7-Dipp-CO2, see below.
This is nicely reflected by the polymerization yields in Table 1.
Under identical conditions 6-Cy-CO2 allows quantitative
isolation of polymer; very high yields can also be achieved by
the action of 6-Hep-CO2, 6-iPr-CO2 and 5s-tBu-CO2. 5-tBu-
CO2 already delivers somewhat worse results (entry 11) and 5-
Cy-CO2 seems to represent a case where a limiting minimal
basic strength is reached and only 23% of PA 12 can be
isolated. At a pKa value lower than that (approximately 24 in
H2O at 25 °C),31 the initiators lose their activity and practically
no product is formed. A mild exception is the high activity of
5s-tBu-CO2, which performs somewhat better than expected.
Interestingly, no PA 12 was formed when N-aryl bearing
initiators were used, irrespective of the ring size. This explains
the inactivity of both 7-Mes-CO2 and 7-Dipp-CO2 and was
even found true for the more electron-rich 6-OMe-CO2 (entry
9). Also the NHC-metal complexes 5s-Mes-MgCl2, 5s-Mes−
Zn-Cl2, and 5s-Mes-SnCl2 did not deliver PA 12 (entries 16−
18). All these data together suggest that (i) steric factors play a
role, too, and (ii) the polymerization process is not susceptible
to Lewis acid activation or at least not susceptible enough to
overcome the relatively weak basicity of the NHC in action (5s-
Mes). The same metal complexes performed very well in the
polymerization of ε-caprolactone,20 though in the presence of
NHC-carboxylates and NHC-metal complexes were obtained
as amorphous powders or crystalline materials. After prolonged
storage (up to two years) under nitrogen atmosphere at room
temperature, NMR analysis showed no visible signs of
decomposition or decarboxylation/deprotection.
Homopolymerization of LL. Polymerization experiments
were effortlessly possible by mixing initiator and LL
(commercial grade monomer, see also the Experimental
Section) and subjection of this mixture to heating under
nitrogen. Because of the high melting point of LL (152 °C) and
the experience from ε-CLA homopolymerization,21 a reaction
temperature of at least 180 °C seemed reasonable. Under these
conditions, a monomeric, stirrable melt was formed. In case
polymer formed, a rapid solidification after few minutes was
observed. When no polymerization took place, the reaction
mass stayed liquid and the residues were soluble in acetone
after cooling. Table 1 lists the polymerization results with every
single initiator.
It is immediately obvious that the structure of the initiator is
decisive for the outcome of the polymerization. Quantitative
D
dx.doi.org/10.1021/ma4018586 | Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
benzylic alcohol, but there a nucleophilic pathway (attack by the
NHC on the carbonyl and ring-opening) is accessible for this
kind of NHC. In contrast to that, the observations discussed
above and the rheology measurements (see below) support the
assumption of strictly basic activity of the free NHCs in lactam
polymerization (deprotonation of the lactam). In the absence
of NHCs control reactions did not deliver any PA 12 (entry 1).
The molecular weights given in Table 1 are relatively uniform
when a ratio of 1:100 (NHC:LL) is applied, ranging from 6700
to 8500 g/mol in all examples with the exception of 5-Cy-CO2.
There, the molecular weight is higher (11 600 g/mol, entry 12)
although the conversion is low, which can be related to
ineffective initiation due to the relatively low basicity of the free
NHC. Impressively, the use of 6-Cy-CO2 allows raising the
ratio of initiator to monomer to 1:300 (Table 1, entry 3), and
still quantitative yields are achieved. The tripling of the ratio
roughly is accompanied by a doubling of the molecular weight,
which emphasizes the nonliving character of anionic lactam
polymerization (compare entries 2 and 3; see also discussion
below). Finally, at an initiator loading of 0.2 mol % the
conversion of monomer arrives at 77% after 45 min (entry 4).
Compared to the homopolymerization of ε-CLA,21 the
polymerization of LL results in better yields and considerably
lower molecular weights. Keeping the different melting points
in mind (180 °C, PA 12; 225 °C, PA 6), it seems reasonable to
assume that chain mobility is an important factor. A reaction
temperature below the melting point of the polymer can cause
precipitation of the polyamide from the melt of its own
monomer and thus “bury” the growing chains, the more so
since polyamides can display high degrees of crystallinity.25
Consequently, in the case of LL polymerization, a more
complete conversion alongside a higher concentration of
growing chains can be expected in the melt, which would
result in lower molecular weights. In addition to this, anionic
ROP of lactams is usually accompanied by side reactions and
far from being a living process.32 Because of that molecular
weights cannot be calculated in the same way as it can be done
for truly living anionic polymerizations. Furthermore, discrep-
ancies in molecular weight can be induced by residual water
content of the monomer (see below).
investigation by NMR or GC-MS, so each point given in
Figure 2 represents an independent polymerization. These were
taken from one batch of mixed initiator and monomer and
processed under identical conditions, except for varying
reaction times (see also Experimental). At 180 °C, 5s-tBu-
CO2 reaches isolated polymer yields of >90% after 600 s. An
increase of the polymerization temperature to 200 °C results in
the same isolated yields after only 400 s. 6-Cy-CO2 promotes
an even faster reaction, especially in the early stages of the
polymerization. It is not uncommon to polymerize lactams
anionically via batch or extrusion techniques at temperatures of
250−270 °C.27 This underlines that the reaction conditions
applied so far are relatively mild and a further increase in the
polymerization rates at higher temperatures can well be
expected. An equally interesting point is the correlation of
conversion and molecular weight during polymerization. To
this end, the polyamides received after different reaction times
were subjected to GPC analysis (see also Figure 2 for one set of
examples). Importantly, the growing monomer conversion is
not accompanied by according chain growth but by a certain
decrease of molecular weight. This behavior is most probably
effected by “chain multiplication”, a side reaction that is caused
by the lactam anion attacking the amide bonds of the
polymer.33 This process gains importance as the concentration
of polymer increases and leads to a degradation of the
polyamide parallel to further incorporation of monomer, with
the overall result of slowly decreasing molecular weight.
Though there is a paucity on reaction kinetics and especially
reaction kinetics coupled with molecular weight analysis in the
literature, it has been described that in anionic ROP of lactams
high molecular weights are built up very fast to degrade
thereafter, which contrasts the hydrolytic lactam polymer-
ization.34 We observed this effect in all investigated polymer-
izations (see Supporting Information for further examples),
which again strongly supports the proposed basic action of the
NHC.
Recording of the polymerization rheology (Figure 3) allows a
complementary view on the reaction process. In a rheometer
To gain further insight into the polymerization process, the
conversion of monomer over time was studied (Figure 2). The
resulting mass of solidified PA 12 during reaction and its
limited solubility do not allow removing samples for
Figure 3. In-situ rheological measurement of LL homopolymerization.
[NHC]:[LL] = 1:100.
equipped with plate geometry and preheated plates the
homogenized mixtures of LL and initiator were inserted and
the increase in viscosity (up to 106 Pa•s) over time was
measured. An induction period with unchanged viscosity
(monomeric melt) is followed by a rapid increase of viscosity.
Induction times like these are typically encountered in anionic
Figure 2. Polymer yield vs reaction time and Mn (PDI) for one set of
experiments. [NHC]:[LL] = 1:100. Formation of homogeneous,
stirred melt = t0.
E
dx.doi.org/10.1021/ma4018586 | Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
ROP of lactams,26 because the first polymerization step consists
of the ring-opening of nonactivated lactam (see Scheme 1).
Industrially, a preactivated monomer (“activators”, see also
introduction) is added to suppress these induction times. As
can be seen from Figure 3, in the case of 5s-tBu-CO2 and an
initiator loading of 1 mol %, the polymerization starts after
about 160 s at 180 °C. At 170 °C, the induction period is
somewhat longer (2oo s). The growing viscosities reflect
formation of polymeric material, but this can only indirectly be
compared to monomer consumption, as viscosity itself depends
on temperature, molecular weight, molecular weight distribu-
tion and crystallinity. However, the accelerating effect of
increased temperature observed in the kinetic experiments
(Figure 2) is mirrored by the results from rheology. Comparing
the curves at 170 and 180 °C (Figure 3), the former displays a
considerably lower viscosity after the same reaction time than
the latter (103 vs 104 Pa·s after 400 s). This allows safely
assuming a lower degree of conversion, the more so since the
lower temperature should increase the viscosity. From an
engineering point of view and for future extrusion experiments,
a correlation of kinetics and rheology is of course highly
desirable.27,35
Figure 4. Bulk copolymerization of LL and ε-CLA using 6-Cy-CO2.
[NHC]:[LL]:[ε-CLA] = 1:50:50 at T = 180 °C. With growing
conversion the incorporation of LL increases.
decrease in the melting point of the corresponding copolymer
(Table 2), which is lower than that of either homopolymer.
The effect is quite strong and in good agreement with results
from the literature, where PA 6/12 of varying constitution was
prepared.23 There, a minimum melting point was found at a
molar ratio of 50% LL in the copolymer, which fits the data
given in Table 2 (see Supporting Information for examples of
DSC analyses). The melting points listed in Table 2 refer to
polyamide that was stored under ambient conditions prior to
DSC measurements, which means that the melting points are
roughly 10 °C lower than expected from dry polyamide due to
uptake of atmospheric water. It should be noted that after
prolonged reaction times, again “chain multiplication” was
observed, the molecular weight decreased from 13 200 g/mol
after a polymerization time of 140 s to 9400 g/mol after 400 s
(Table 2). The effect is somewhat stronger than in LL
homopolymerization (see above) and accordingly seems to
continue. Thus, after 45 min of polymerization, it was not
possible to observe a peak via GPC analysis of the
corresponding sample. This is most presumably due to both
the formation of lower molecular weights and a very broad
molecular weight distribution, as judged from GPC analysis. It
should be emphasized that the decrease in molecular weight is
not responsible for the lowering of the melting point (compare
entries 1, 4, and 5 in Table 2). However, it can be expected that
the nonliving polymerization process in combination with chain
multiplication will entail a “scrambling” of the gradient,
particularly in case longer reaction times are applied.
Judging from the polymerization mechanism (Scheme 1), the
most important factors for copolymer composition should be
(a) the pKa-value of the lactams, (b) ring tension and heat of
polymerization of the respective lactams, and (c) effect of the
counterion (protonated NHC). For both LL and ε-CLA, a pKa-
value of 27.2 (DMSO, 25 °C) is reported.37 So there should be
no preference in the formation of “activated monomer”
(deprotonated lactam) for either monomer. The higher ring
tension and concomitant higher heat of polymerization of ε-
CLA,33 however, are obviously decisive for the outcome of the
copolymerization and cause the preference for ε-CLA
incorporation at the onset of the polymerization. For
conventionally activated and catalyzed copolymerizations of ε-
CLA and LL a similar tendency has been observed.38 There, the
distribution of the monomeric units within the copolymer,
whether in a more random or a more “blocky” way, is
influenced by the nature of the catalysts.39 Whether the same is
In the solid state, NHC−carboxylates are known to be stable
up to high temperatures,14a and consequently, we observed no
diminishing activity of the monomer/initiator mixtures over a
period of 1 month. The samples were hereby stored at room
temperature under exclusion of water. However, exposition of
the powdered mixtures to atmosphere for 1 h resulted in
complete quenching of any polymerization activity. This may
be attributed to the uptake of water, either by the finely
powdered monomer or by the initiator itself. The reactivity of
NHC−CO2 adducts toward water ranges from inertness36 over
13
formation of hydrogen carbonates
to decomposition,14a
depending on structure and conditions. Generally, this
observed sensitivity toward atmospheric humidity is in
accordance with the assumption of an anionic polymerization
mechanism.
Copolymerization of LL and ε-CLA. Having established
that the NHC-mediated homopolymerization of LL and ε-
CLA21 is possible, the copolymerization of both monomers and
the composition of the resulting copolymer was investigated.
To this end, the same reaction conditions (bulk, 180 °C) and
the best initiator (6-Cy-CO2) were used as described above.
Flakes of ε-CLA and pellets of LL were combined with the
required amount of initiator (molar ratio of [NHC]:[LL]:[ε-
CLA] = 1:50:50, which corresponds to a LL-content of 64% by
weight). Several batches were reacted this way and stopped
after different polymerization times. The precipitation and
washing solutions were collected, concentrated and analyzed by
GC−MS to calculate the ratios of residual monomer. This
allowed to determine the composition of the PA 6/12
copolymer (see also Experimental). Figure 4 depicts the results
for this series of experiments.
As expected, isolated yields increased over time from 17%
after 140 s to 96% after 45 min. More interesting, a clear
preference for the incorporation of ε-CLA at the beginning of
the polymerization is apparent. In the early stages of the
reaction, only 25% of the polymer constitute of LL. Only at
higher conversions, where the monomeric melt is enriched with
LL, it is consumed more frequently. In other words, a
copolymer with a considerable gradient is formed. Strong
support for the accuracy of these data is provided by thermal
analysis. A growing amount of incorporated LL is mirrored by a
F
dx.doi.org/10.1021/ma4018586 | Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
Table 2. Bulk Copolymerizations of LL and ε-CLA Using 6-Cy-CO2 at T = 180 °C
a
b
c
NHC:LL:ε-CLA
t
mol % LL /yield [%]
Tm [°C]
Mn(PDI) [g/mol]
1:50:50
1:50:50
1:50:50
1:50:50
1:150:150
1:100:0
140 s
25/17
26/33
39/68
51/96
47/86
100/100
171
163
138
120
120
13200 (2.4)
9900 (2.4)
9400 (2.1)
n.d.
200 s
400 s
45 min
45 min
45 min
13200 (2.5)
169
7100 (2.6)
a
b
c
The mol % of LL in the copolymer, calculated by GC−MS analysis of residual monomer content. Isolated overall yield of polymer. GPC (m-
cresol) relative to PS standards.
true for NHC-mediated anionic copolymerizations has still to
be investigated, but a truly innocent role of the protonated
NHC as counterion seems doubtful. Size as counterion and
formation of ion pairs can well be expected to influence the
copolymerization preferences.
Decreasing the initiator loading (entry 5, Table 2) to
1:150:150 still allows isolating the PA 6/12 copolymer in good
yields of 86%. In accordance to the high conversion, a high
degree of LL incorporation and a low melting point is found. It
is interesting to note that LL homopolymerization with an
applied ratio of NHC:LL of 1:300 still produces quantitative
yields (see above), while the copolymerization shows somewhat
reduced activity. Since chain mobility should not be a problem
here (low melting point of the copolymer) it seems most
probable that residual water content in the commercial ε-CLA
is responsible for the loss of activity.
CO2/monomer mixtures constitute a completely metal-free
approach to polyamides.
ASSOCIATED CONTENT
■
S
* Supporting Information
13C NMR of compound 1, further details on molecular weight
development during polymerization, a figure showing chain
multiplication, and DSC graphs. This material is available free
of charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
*(M.R.B.) E-Mail: michael.buchmeiser@ipoc.uni-stuttgart.de.
■
Funding
Notes
Financial support by Evonik Industries and the University of
Stuttgart is gratefully acknowledged.
CONCLUSION
■
ACKNOWLEDGMENTS
■
It has been demonstrated that protected NHCs are well suited
for the application as thermally latent initiators in the additive-
free bulk polymerization of laurolactam. Mixtures of initiator
and monomer are storable for weeks without any loss of activity
and allow effortless polymerization by simple heating. After
thermal activation, polymerizations were shown to proceed
swiftly and moderately controlled. Success in the production of
the polyamide is dependent on the structure of the initiating
NHC, where 1,3-di(cyclohexyl)-tetrahydropyrimidinium-2-car-
boxylate (6-Cy-CO2) turned out to be the most efficient pre-
initiator in this study. With this protected NHC, quantitative
yields of PA 12 are accessible in relatively short polymerization
times with acceptable molecular weight distribution. The
correlation of NHC-basicity and NHC-activity in polymer-
ization, the sensitivity toward atmospheric humidity, the
occurrence of “chain multiplication” and the observation of
induction times all support the presence of lactam anions and
thus a basic action of the NHC.
B.Sc. Laura Widmann is gratefully acknowledged for support in
synthesis. Manfred Simmet (IKT Stuttgart) and Dr. Rainer
Gutmann (ITCF Denkendorf) are thanked for help with
polymer characterization. Finally, L. Bruggemann KG (Heil-
̈
bronn, Germany) is acknowledged for a gift of ε-caprolactam.
REFERENCES
■
(1) (a) Enders, D.; Niemeier, O.; Henseler, A. Chem. Rev. 2007, 107,
5606−5655. (b) Marion, N.; Díez-Gonzal
́
ez, S.; Nolan, S. P. Angew.
Chem., Int. Ed. 2007, 46, 2988−3000. (c) Nair, V.; Vellalath, S.; Babu,
B. P. Chem. Soc. Rev. 2008, 37, 2691−2698.
(2) Maji, B.; Breugst, M.; Mayr, H. Angew. Chem., Int. Ed. 2011, 50,
6915−6919.
(3) (a) Biju, A. T.; Padmanaban, M.; Wurz, N. E.; Glorius, F. Angew.
Chem., Int. Ed. 2011, 50, 8412−8415. (b) Vora, H. U.; Wheeler, P.;
Rovis, T. Adv. Synth. Catal. 2012, 354, 1617−1639. (c) Grossmann, A.;
Enders, D. Angew. Chem., Int. Ed. 2012, 51, 314−325.
(4) Ryan, S. J.; Candish, L.; Lupton, D. W. Chem. Soc. Rev. 2013, 42,
4906−4917.
(5) Magill, A. M.; Cavell, K. J.; Yates, B. F. J. Am. Chem. Soc. 2004,
126, 8717−8724.
(6) Benhamou, L.; Chardon, E.; Lavigne, G.; Bellemin-Laponnaz, S.;
Finally, the copolymerization of LL and ε-CLA was
investigated. It was found that a gradient copolymer is formed,
with ε-CLA being incorporated preferentially at the onset of the
polymerization. The considerable discrimination between both
monomers by the polymerization system in spite of their very
similar physical properties promises the possibility of producing
more “blocky” polyamides in simple bulk polymerization by
appropriate choice of monomers and reaction conditions. Work
on these topics seems certainly worthwhile and is currently
carried out in our group.
Overall, the simple and cheap preparation of the protected
NHCs and the application of an easy-to-handle one-component
system certainly make this process an interesting choice for
large scale extrusion polymerization, especially since the NHC−
Ces
(7) (a) Fev
́
ar, V. Chem. Rev. 2011, 111, 2705−2733.
̀
re, M.; Pinaud, J.; Gnanou, Y.; Vignolle, J.; Taton, D.
Chem. Soc. Rev. 2013, 42, 2142−2172. (b) Dove, A. P. ACS Macro Lett.
2012, 1, 1409−1412.
(8) (a) Zhang, Y.; Chen, E. Y.-X. Angew. Chem. 2012, 124, 2515−
2519. (b) Scholten, M. D.; Hedrick, J. L.; Waymouth, R. M.
Macromolecules 2008, 41, 7399−7404.
(9) (a) Raynaud, J.; Absalon, C.; Gnanou, Y.; Taton, D. J. Am. Chem.
Soc. 2009, 131, 3201−3209. (b) Raynaud, J.; Ottou, W. N.; Gnanou,
Y.; Taton, D. Chem. Commun. 2010, 46, 3203−3205.
(10) (a) Kamber, N. E.; Jeong, W.; Gonzalez, S.; Hedrick, J. L.;
Waymouth, R. M. Macromolecules 2009, 42, 1634−1639. (b) Shin, E.
G
dx.doi.org/10.1021/ma4018586 | Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
J.; Jeong, W.; Brown, H. A.; Koo, B. J.; Hedrick, J. L.; Waymouth, R.
318. (e) Costa, G.; Pedemonte, E.; Russo, S.; Sava,
̀
E. Polymer 1979,
M. Macromolecules 2011, 44, 2773−2779.
20, 713−718.
(11) Nyce, G. W.; Glauser, T.; Connor, E. F.; Mock, A.; Waymouth,
(33) Sekiguchi, H. in: Ivin, K. J.; Saegusa, T. Ring-opening
polymerization; Elsevier Applied Science: London, 1984, 809−918.
̈
R. M.; Hedrick, J. L. J. Am. Chem. Soc. 2003, 125, 3046−3056.
(12) Rodriguez, M.; Marrot, S.; Kato, T.; Ster
(34) (a) Vieweg, R.; Muller, A. Kunststoff-Handbuch Band VI:
́
in, S.; Fleury, E.;
̈
Polyamide; Hanser: Munich, 1966; (b) Griehl, W. Faserforsch.
Baceiredo, A. J. Organomet. Chem. 2007, 692, 705−708.
(13) (a) Fevre, M.; Pinaud, J.; Leteneur, A.; Gnanou, Y.; Vignolle, J.;
Taton, D.; Miqueu, K.; Sotiropoulos, J.-M. J. Am. Chem. Soc. 2012,
re, M.; Coupillaud, P.; Miqueu, K.;
Textiltechn. 1956, 7, 207.
̀
́
(35) Dave, R. S.; Kruse, R. L.; Udipi, K.; Williams, D. E. Polymer
1997, 38, 949−954.
134, 6776−6784. (b) Fev
̀
(36) Smiglak, M.; Holbrey, J. D.; Griffin, S. T.; Reichert, W. M.;
Swatloski, R. P.; Katritzky, A. R.; Yang, H.; Zhang, D.; Kirichenko, K.;
Rogers, R. D. Green Chem. 2007, 9, 90−98.
Sotiropoulos, J.-M.; Vignolle, J.; Taton, D. J. Org. Chem. 2012, 77,
10135−10144.
(14) (a) van Ausdall, B. R.; Glass, J. L.; Wiggins, K. M.; Aarif, A. M.;
(37) Hashimoto, K. Progr. Polym. Sci. 2000, 25, 1411−1462.
(38) (a) Frunze, T.; Kotel’nikov, V.; Kurashev, V.; Ivanova, S.;
Komarova, L.; Korshak, V. Polym. Sci. USSR 1976, 18, 348−353.
(b) Kricheldorf, H. R.; Hull, W. E. J. Polym. Sci., Polym. Chem. 1978,
16, 2253−2264.
̀
Louie, J. J. Org. Chem. 2009, 74, 7935−7942. (b) Fevre, M.; Vignolle,
J.; Taton, D. Polym. Chem. 2013, 4, 1995−2003. (c) Hans, M.;
Wouters, J.; Demonceau, A.; Delaude, L. Eur. J. Org. Chem. 2011,
2011, 7083−7091. (d) Delaude, L.; Demonceau, A.; Wouters, J. Eur. J.
Inorg. Chem. 2009, 2009, 1882−1891.
(39) For an overview see: Dubois, P.; Coulembier, O.; Raquez, J.-M.
Handbook of ring-opening polymerization; Wiley-VCH: Weinheim,
Germany, 2009, Chapter 7.8.
(15) (a) Bantu, B.; Pawar, G. M.; Decker, U.; Wurst, K.; Schmidt, A.
M.; Buchmeiser, M. R. Chem.Eur. J. 2009, 15, 3103−3109.
(b) Bantu, B.; Pawar, G. M.; Wurst, K.; Decker, U.; Schmidt, A.
M.; Buchmeiser, M. R. Eur. J. Inorg. Chem. 2009, 2009, 1970−1976.
(16) Norris, B. C.; Sheppard, D. G.; Henkelman, G.; Bielawski, C. W.
J. Org. Chem. 2011, 76, 301−304.
(17) (a) Coulembier, O.; Dove, A. P.; Pratt, R. C.; Sentman, A. C.;
Culkin, D. A.; Mespouille, L.; Dubois, P.; Waymouth, R. M.; Hedrick,
J. L. Angew. Chem., Int. Ed. 2005, 44, 4964−4968. (b) Coulembier, O.;
Lohmeijer, B. G. G.; Dove, A. P.; Pratt, R. C.; Mespouille, L.; Culkin,
D. A.; Benight, S. J.; Dubois, P.; Waymouth, R. M.; Hedrick, J. L.
Macromolecules 2006, 39, 5617−5628.
(18) Arduengo, A. J., III; Calabrese, J. C.; Davidson, F.; Rasika Dias,
H. V.; Goerlich, J. R.; Krafczyk, R.; Marshall, W. J.; Tamm, M.;
Schmutzler, R. Helv. Chim. Act. 1999, 82, 2348−2364.
(19) Naumann, S.; Schmidt, F. G.; Schowner, R.; Frey, W.;
Buchmeiser, M. R. Polym. Chem. 2013, 4, 2731−2740.
(20) Naumann, S.; Schmidt, F. G.; Frey, W.; Buchmeiser, M. R.
Polym. Chem. 2013, 4, 4172−4181.
(21) Naumann, S.; Epple, S.; Bonten, C.; Buchmeiser, M. R. ACS
Macro Lett. 2013, 2, 609−612.
(22) (a) Memeger, W.; Campbell, G. C.; Davidson, F. Macromolecules
1996, 29, 6475−6480. (b) Yang, H.; Zhao, J.; Yan, M.; Pispas, S.;
Zhang, G. Polym. Chem. 2011, 2, 2888−2892.
(23) Garner, D. P.; Fasulo, P. D. J. Appl. Polym. Sci. 1988, 36, 495−
509.
(24) Udipi, K.; Dave,
1997, 38, 927−938.
(25) Marchildon, K. Macromol. React. Eng. 2011, 5, 22−54.
(26) Odian, G. G. Principles of polymerization; Wiley: Hoboken, NJ,
2004.
́
R. S.; Kruse, R. L.; Stebbins, L. R. Polymer
(27) (a) Luisier, A.; Bourban, P.-E.; Manson, J.-A. E. J. Polym. Sci.,
̊
Part A: Polym. Chem. 2002, 40, 3406−3415. (b) Luisier, A.; Bourban,
P.-E.; Manson, J.-A. Composites Part A: Appl. Sci. and Manuf. 2003, 34,
̊
583−595.
(28) Iglesias, M.; Beetstra, D. J.; Knight, J. C.; Ooi, L.-L.; Stasch, A.;
Coles, S.; Male, L.; Hursthouse, M. B.; Kingsley, J. C.; Dervisi, A.;
Fallis, I. A. Organometallics 2008, 27, 3279−3289.
(29) (a) Kolychev, E. L.; Portnyagin, I. A.; Shuntikov, V. V.;
Khrustalev, V. N.; Nechaev, M. S. J. Organomet. Chem. 2009, 694,
2454−2462. (b) Scarborough, C. C.; Popp, B. V.; Guzei, I. A.; Stahl, S.
S. J. Organomet. Chem. 2005, 690, 6143−6155.
(30) (a) Droge, T.; Glorius, F. Angew. Chem., Int. Ed. 2010, 49,
̈
6940−6952. (b) Nelson, D. J.; Nolan, S. P. Chem. Soc. Rev. 2013, 42,
6723−6753.
(31) Higgins, E. M.; Sherwood, J. A.; Lindsay, A. G.; Armstrong, J.;
Massey, R. S.; Alder, R. W.; O’Donoghue, A. C. Chem. Commun. 2011,
47, 1559−1561.
(32) (a) Kusk
179, 337−344. (b) Roda, J.; Kral
̌
ova,
́
M.; Roda, J.; Kral
ícek, J.; Karel, S. Eur. Polym. J. 1976,
12, 729−732. (c) Sebenda, J.; Kouril, V. Eur. Pol. J. 1971, 7, 1637−
1648. (d) Tani, H.; Konomi, T. J. Polym. Sci., Part A: 1 1966, 4, 301−
́ ̌
ícek, J. Makromol. Chem. 1978,
̌
́
̌
̌
̌
H
dx.doi.org/10.1021/ma4018586 | Macromolecules XXXX, XXX, XXX−XXX