Cyclic poly(α-peptoid)s and their block copolymers from N-heterocyclic carbene-mediated ring-opening polymerizations of N-substituted N-carboxylanhydrides

Article abstract of DOI:10.1021/ja907380d

(Chemical Equation Presented) N-Heterocyclic carbene (NHC)-mediated ring-opening polymerization (ROP) of N-substituted N-carboxylanhydride ( ^NR-NCA) yields cyclic poly(4-peptoid)s with controlled molecular weights (M_n = 3-30 kg mol^-1) and narrow molecular weight distributions (PDI = 1.04-1.12). The reactions exhibit characteristics of a living polymerization with minimal chain transfer. This enables the facile synthesis of cyclic diblock copoly(4-peptoid)s through sequential monomer addition. The cyclic polymer architectures were verified by MALDI-TOF mass spectrometry and intrinsic viscosity measurements. Mark-Houwink-Sakurada plot analyses revealed that cyclic poly(4-peptoid)s prepared from NHC-mediated polymerizations exhibit lower intrinsic viscosities than their linear analogues prepared from primary amine-initiated polymerizations. The ratio of their intrinsic viscosities is consistent with the former being mostly cyclic.

Full text of DOI:10.1021/ja907380d

Published on Web 12/01/2009
Cyclic Poly(r-peptoid)s and Their Block Copolymers from N-Heterocyclic
Carbene-Mediated Ring-Opening Polymerizations of N-Substituted
N-Carboxylanhydrides
Li Guo and Donghui Zhang*
Department of Chemistry and Macromolecular Studies Group, Louisiana State UniVersity, Baton Rouge, Louisiana 70803
Received August 31, 2009; E-mail: dhzhang@lsu.edu
Poly(R-peptoid)s, which are structural mimics of poly(R-pep-
tide)s, feature an N-substituted polyglycine backbone with protei-
nogenic or synthetic side chains on the nitrogen atom (Figure 1).
They have attracted much attention for their potential as peptido-
mimetic therapeutic agents or drug delivery carriers due to their
enhanced enzymatic and hydrolytic stability relative to poly(R-pep-
50 °C under a nitrogen atmosphere (Scheme 1). The polymers were
isolated by precipitation with hexane and analyzed for their
1
13
1
molecular structures and end groups by H and C{ H} NMR
spectroscopy and MALDI-TOF and electrospray ionization (ESI)
mass spectrometry (MS). The polymer absolute molecular weight
n w n
(M ) and molecular weight distribution (PDI ) M /M ) were
1
,2
tide)s. In contrast to the poly(R-peptide)s, whose secondary
structure (e.g., helix) is stabilized by hydrogen bonding, the
conformation of poly(R-peptoid)s, which are free of hydrogen
bonding interactions because of the N-substitution, is largely
determined by the backbone rigidity and the steric and electrostatic
determined by size-exclusion chromatography (SEC) coupled with
multiangle light scattering (MALS), viscometry (VISC), and
differential refractive index (DRI) triple detection (SEC-
MALS-VISC-DRI).
3
Scheme 1
characteristics of the side chains. For example, oligomeric R-pep-
toids with aliphatic or aromatic chiral side chains have been shown
4
to form a stable polyproline I helix.
Figure 1. Molecular structures of poly(R-peptoid)s and poly(R-peptide)s
(
R ) proteinogenic or synthetic side chain).
While linear poly(N-Me-glycine) (a.k.a., polysarcosin) can be
synthesized from an amine-initiated ring-opening polymerization
N
(
ROP) of the corresponding N-substituted N-carboxylanhydride ( R-
5
1
13
1
NCA, R ) Me), the method is limited with respect to the scope
The H and C{ H} NMR spectra of the polymers are consistent
with the proposed repeating unit structure formed from ROP of
of monomers. The reaction suffers from low efficiency and is
5
b
7
heavily monomer-dependent, which makes it impractical for the
general synthesis of poly(R-peptoid)s. Hence, the synthesis of
poly(R-peptoid)s in a controlled manner remains a challenge.
the corresponding substrates. End-group analysis of a low-MW
1
polymer sample by H NMR spectroscopy and ESI MS revealed a
1
major polymeric species (>95% by H NMR integration) whose
5f
A zwitterionic mechanism was proposed for the pyridine /tertiary
structure is consistent with an NHC-poly(N-Bu-glycine) spirocyclic
5
i
5j
N
5k
7
amine /solvent -initiated or thermal ROP of Me-NCA, but
control over the polymer molecular weight (MW) and molecular
weight distribution [as indicated by the polydispersity index (PDI)]
is limited. N-Heterocyclic carbenes (NHCs) have been shown to
adduct 3 (Scheme 1). An analogous NHC-polyester spirocycle
was reported previously in an NHC-mediated polymerization of
6
b
ꢀ-lactone. Surprisingly, MALDI TOF MS analysis of the same
sample revealed that the sample consists mainly of free cyclic
poly(N-Bu-glycine) (4) (95% by mass) with spirocycle 3 notably
6
mediate ROP of cyclic substrates (e.g., cyclic ester, ethylene
6
d
7
oxide) through a zwitterionic propagating species to yield
polymers with controlled MW and PDI. Inspired by these studies,
absent. To resolve the inconsistency in the ESI and MALDI TOF
MS results, spirocycle 3 was treated with 2 equiv of dithranol, the
N
we reasoned that NHCs should mediate the ROP of R-NCA in a
matrix material used in the MALDI TOF MS experiment, in THF
1
controlled and efficient manner because of the enhanced nucleo-
philicity of both the NHC and the zwitterionic propagating species.
Herein we report the first general and efficient synthetic route
at room temperature. H NMR analysis of the reaction product
revealed the formation of a major polymeric species whose structure
7
is consistent with 4.
N
toward cyclic poly(R-peptoid)s from the controlled ROP of R-
To verify the cyclic architecture of 3 in the high-MW range,
linear poly(N-Bu-glycine)s with different MWs were independently
NCA with an NHC initiator and demonstrate its utility in preparing
N
cyclic block copoly(R-peptoid)s (Scheme 1).
prepared by an amine-initiated polymerization of Bu-NCA. SEC
N
N
-1
Bu-NCA (M
1
) and Me-NCA (M
2
) were synthesized from the
n
analysis revealed that poly(N-Bu-glycine)s (M ) 10-30 kg mol )
N
corresponding N-substituted glycine precursors by the standard
obtained from NHC 1-mediated polymerizations of Bu-NCA
exhibit lower intrinsic viscosities than the linear analogues having
identical MWs (Figure 2). This indicates that the former polymers
have physically more compact architectures. The [η]cyclic/[η]linear ratio
7
N
Fuchs method. Polymerizations of Bu-NCA in the presence of
various loadings of bis(2,6-diisopropylphenyl)imidazol-2-ylidene
(
1) were allowed to proceed for 16 h in tetrahydrofuran (THF) at
1
8072 9 J. AM. CHEM. SOC. 2009, 131, 18072–18074
10.1021/ja907380d  2009 American Chemical Society

C O M M U N I C A T I O N S
[
(
0.83(2)] exceeds the theoretical prediction for cyclic polymers
0.662) based on the infinite freely jointed chain model. It is
monomer addition (Figure 3b) revealed an increase of polymer
molecular weight M from 5.5 to 8.3 kg mol and a change in
n
8
,9
-1
consistent with samples being mostly cyclic polymers. While
residual linear contaminants can cause the ratio to deviate, non-θ-
condition, finite size, and real chain stiffness also contribute to the
PDI from 1.11 to 1.05. The latter values agree with those for
polymers independently synthesized from a 75:1 [M /[1] reaction
) 8.7 kg mol , PDI ) 1.08). The MW distribution remains
]
1 0
0
-1
(M
n
10
discrepancy. Linear fits of the respective Mark-Houwink-Sakurada
plots (log [η] vs log M ) of the cyclic and linear poly(N-Bu-
narrow after the chain extension, and no low-MW polymer was
detected. Taken together, these results strongly indicate a living
polymerization with minimal inter- or intrachain transfer. In contrast
to other NHC-mediated ROPs of cyclic substrates, the activated
w
glycine)s (Figure 2a) yielded nearly identical Mark-Houwink
exponents [Rcyclic) 0.65(1), Rlinear) 0.66(2)]. They are consistent
with self-avoiding walks or slightly extended conformations for
N
functional group (i.e., the anhydride) of the R-NCA monomer is
11
both polymers, hence eliminating conformational effects as being
responsible for their intrinsic viscosity differences. The cyclic
polymers also elute later than their linear analogues with identical
different from those (i.e., amide) present in the polymer backbones.
As the amide bonds are highly inert toward transamidation, inter-
and intrachain transfer is significantly reduced.
7
MWs. This is consistent with the former having a smaller
Table 1. NHC 1-Mediated ROP of NBu-NCA (M
)
a
1
hydrodynamic volume, assuming the absence of non-size-exclusion
separation mechanisms.
-1
M (kg mol
n
)
b
c
d
d
PDI
entry
[M ] /[1]
1 0 0
[M ]
1 0
conv. (%)
calcd
SEC
1
2
3
4
5
6
7
8
9
25:1
50:1
100:1
200:1
400:1
25:1
0.4
0.4
0.4
0.4
0.4
0.8
0.8
0.8
0.8
0.8
100
100
96
95
89
100
100
96
88
79
2.8
5.6
10.9
21.5
40.3
2.8
3.3
5.6
10.1
14.5
16.7
3.0
1.07
1.08
1.10
1.09
1.12
1.06
1.06
1.05
1.06
1.04
50:1
5.6
5.6
100:1
200:1
400:1
10.9
19.9
35.8
10.7
18.9
26.7
1
0
a
All reactions were conducted at 50 °C in THF with [M
1
]
0
) 0.4 or
b
1
0
.8 M and stopped after 16 h. Conversion was calculated from the H
NMR spectrum of an aliquot of the reaction mixture. Theoretical MW
c
Figure 2. Mark-Houwink-Sakurada plot of cyclic (0) and linear (O)
poly(N-Bu-glycine)s and their linearly fit curves (dashed line, cyclic; solid
line, linear).
was calculated from the conversion and the [M ] /[1] ratio.
1
0
0
d
Experimental MW and PDI were obtained from a tandem SEC-
MALS-VISC-DRI system in 0.1 M LiBr/DMF solution at 50 °C using
-
1
To assess the NHC control over the polymerization, reactions
a measured dn/dc of 0.0714(13) mL g .
with different initial monomer to initiator ratios ([M
1 0 0
] /[1] ) were
conducted. Pauci-disperse (PDI ) 1.04-1.12) poly(N-Bu-glycine)s
-1
with predictable MWs ranging from ∼3 to 30 kg mol can be
readily synthesized by controlling the [M /[1] ratio and the
1
]
0
0
reaction time (Table 1). The experimental polymer MWs generally
concur with the theoretical values calculated from the conversion
and [M
of the experimental M
pronounced at the high-MW end (Table 1, entries 5 and 10),
possibly as a result of the presence of impurities (e.g., H O) that
The good
1
0
] /[1]
0
ratio for single-site initiation by NHC. Deviations
n
from the theoretical values are more
2
5
a-e
1
can initiate the nucleophilic polymerization of M .
control over the cyclic polymer MW and PDI is maintained even
at elevated initial monomer concentrations, in contrast to the case
of conventional synthetic strategies for cyclic polymers, where high
dilution conditions are required to inhibit linear polymer formation.
The reaction exhibits characteristics of a living polymerization
with minimal inter- or intrachain transfer. For example, the polymer
MW increases linearly with conversion while the PDI (1.04-1.17)
remains low even at high conversions (up to ∼97%) (Figure 3a).
This suggests that the rate of cyclization of 2 to yield cyclic poly(N-
Bu-glycine) 4 is much slower than that of propagation. Consistent
with a living polymerization, preliminary kinetic studies ([M
.4 M, [M /[1] ) 50:1, 50 °C, THF) revealed that the rate of
polymerization has a first-order dependence on the monomer
1
] )
n
Figure 3. (a) Representative plots of experimental M (9) and PDI (2) vs
0
1
]
0
0
conversion ([M
chromatograms of poly(N-Bu-glycine) prior to (dashed curve, [M
50:1) and after (solid curve, [M /[1] ) 25:1) the chain-extension
experiment and of an independently synthesized poly(N-Bu-glycine) sample
dot-dashed curve, [M /[1] ) 75:1).
1
]
0
/[1]
0
) 150:1, [M
]
1 0
) 0.4 M, 50 °C, THF). (b) SEC
1
]
0
/[1]
0
)
1
]
0
0
-1
concentration (d[M
A chain-extension experiment with Bu-NCA ([M
) was also conducted to verify the living nature of the polymer-
1
]/dt ) kobs[M
1
], kobs) 0.41(2) h ).
N
1 0 0
] /[1] ) 50:
(
1
]
0
0
1
ization. Upon complete conversion of the initial monomers, an
additional 25 equiv of the identical monomers were added, and
the reaction was allowed to continue to completion. SEC chro-
matograms of polymers obtained prior to and after the second
It is evident that polymeric spirocycles 3 (Scheme 1) are the
catalytic resting states during polymerization. By analogy to
controlled radical polymerizations and the NHC-mediated zwitte-
6
b
rionic polymerizations of ꢀ-lactones, the effective concentration
J. AM. CHEM. SOC. 9 VOL. 131, NO. 50, 2009 18073

C O M M U N I C A T I O N S
of active propagating species 2 is lowered via equilibrium with the
dormant propagating species 3, thereby reducing side reactions (e.g.,
step-growth polymerization, deprotonation) and enhancing control
of polymer MW and PDI. In support of this hypothesis, when
polymerizations were conducted in a solvent that can better stabilize
the zwitterionic species [e.g., dimethyl formamide (DMF)], control
of the polymer MW was significantly compromised, and only low-
MW polymers were obtained regardless of the initial monomer/
initiator ratio.
from this study will guide the optimization of reaction conditions
to ultimately produce high-MW cyclic poly(R-peptoid)s with 99+%
purity.
Acknowledgment. D.Z. thanks Dr. Rafael Cueto for assisting
with the SEC studies and Dr. Dan Pu for conducting the MALDI-
TOF MS experiments. This work was supported by LSU and the
Louisiana Board of Regents (LEQSF-RD-A-11).
12
1
Supporting Information Available: H NMR and ESI MS spectra
1
of a low-MW spirocyclic NHC-poly(N-Bu-glycine) adduct; H NMR
and MALDI TOF MS spectra of a low-MW cyclic poly(N-Bu-glycine);
MALDI TOF MS spectrum of a low-MW poly(N-Bu-glycine) sample
1
obtained from NHC 1-mediated polymerization of M
C{ H} spectra of M
1
in DMF; H and
1
3
1
1 2
and M monomers, high-MW cyclic poly-
(
N-Bu-glycine), cyclic poly(N-Me-glycine), and cyclic poly(N-Me-
glycine)-b-poly(N-Bu-glycine) block copolymer; SEC chromatograms
of polymer samples prior to and after heterochain-extension experiments
for cyclic block copoly(R-peptoid) synthesis; cyclic poly(N-Bu-glycine)
MW and PDI data using calibration curves constructed with polystyrene
standards; poly(N-Bu-glycine) MW and PDI data for polymerizations
conducted in DMF; and Mark-Houwink-Sakurada plots for cyclic
and linear poly(N-Bu-glycine) and poly(N-Me-glycine)-b-poly(N-Bu-
glycine), with and without band-broadening corrections for MWs, and
the corresponding logarithmic plots of MW versus elution volume. This
material is available free of charge via the Internet at http://pubs.acs.org.
Figure 4. Mark-Houwink-Sakurada plot of cyclic (0) and linear (O)
poly(N-Me-glycine)-b-poly(N-Bu-glycine) block copolymers with identical
compositions and their linearly fit curves (dashed line, cyclic; solid line,
linear).
References
The utility of this reaction for accessing cyclic block copoly-
R-peptoid)s was demonstrated through the synthesis of a poly(N-
(
1) Patch, J. A.; Barron, A. E. Curr. Opin. Chem. Biol. 2002, 6, 872–877.
2) Barron, A. E.; Zuckermann, R. N. Curr. Opin. Chem. Biol. 1999, 3, 681–
(
(
6
87.
Me-glycine)-b-poly(N-Bu-glycine) block copolymer by sequential
(
3) Baldauf, C.; G u¨ nther, R.; Hofmann, H.-J. Phys. Biol. 2006, 3, S1–S9.
N
N
addition of Me-NCA ([M
2
]
0
/[1]
0
) 50:1) and Bu-NCA ([M
1
]
0
/
(4) (a) Wu, C. W.; Kirshenbaum, K.; Sanborn, T. J.; Patch, J. A.; Huang, K.;
Dill, K. A.; Zuckermann, R. N.; Barron, A. E. J. Am. Chem. Soc. 2003,
[
1] )25:1). Both steps were allowed to reach completion. SEC
0
1
25, 13525–13530. (b) Sanborn, T. J.; Wu, C. W.; Zuckermann, R. N.;
analyses of polymers obtained prior to and after the second
monomer addition revealed an increase of polymer MW from 4.4
Barron, A. E. Biopolymers 2002, 63, 12–20. (c) Wu, C. W.; Sanborn, T. J.;
Zuckermann, R. N.; Barron, A. E. J. Am. Chem. Soc. 2001, 123, 2958–
2
963. (d) Wu, C. W.; Sanborn, T. J.; Huang, K.; Zuckermann, R. N.; Barron,
kg mol- (PDI ) 1.04) to 6.5 kg mol (PDI ) 1.03), in agreement
1
-1
A. E. J. Am. Chem. Soc. 2001, 123, 6778–6784.
-1
7
(5) (a) Waley, S. G.; Watson, J. J. Am. Chem. Soc. 1948, 70, 2299–2300. (b)
Ballard, D. G. H.; Bamford, C. H. J. Chem. Soc. 1958, 355–360. (c) Waley,
S. G.; Watson, J. Proc. R. Soc. London, Ser. A 1949, 199, 499–517. (d)
Bamford, C. H.; Block, H.; Mason, D.; Openshaw, A. W. SCI Monogr.
with the theoretically predicted value (M ) 6.4 kg mol ).
n
Extraction of the copolymer product with water did not yield any
poly(N-Me-glycine) homopolymers, suggesting 100% living chains
1966, 20, 304–313. (e) Sisido, M.; Imanishi, Y.; Higashimura, T. Makromol.
1
13
1
in the first polymerization step. H and C{ H} NMR spectra of
Chem. 1977, 178, 3107–3114. (f) Kricheldorf, H. R.; B o¨ singer, K.
Makromol. Chem. 1976, 177, 1243–1258. (g) Kricheldorf, H. R.; von
Lossow, C.; Schwarz, G. Macromol. Chem. Phys. 2004, 205, 918–924. (h)
Kricheldorf, H. R.; von Lossow, C.; Schwarz, G.; Fritsch, D. Macromol.
Chem. Phys. 2005, 206, 1165–1170. (i) Kricheldorf, H. R.; von Lossow,
C.; Schwarz, G. J. Polym. Sci., Part A: Polym. Chem. 2006, 44, 4680–
the purified polymers were consistent with the diblock copoly-
7
(
R-peptoid) backbone structure. SEC analysis of these block
copolymer samples also revealed that the polymers exhibit lower
7
intrinsic viscosities than their linear analogues (Figure 4). The
4
695. (j) Kricheldorf, H. R.; von Lossow, C.; Schwarz, G. Macromolecules
2
005, 38, 5513–5518. (k) Kricheldorf, H. R.; von Lossow, C.; Lomadze,
Mark-Houwink exponents for the cyclic and linear block copoly-
N.; Schwarz, G. J. Polym. Sci., Part A: Polym. Chem. 2008, 46, 4012–
4
mers are 0.51(4) and 0.57(9), implying self-avoiding-walk confor-
020.
1
1
mations for both polymers. The [η]cyclic/[η]linear ratio [0.84(5)]
suggests that the former polymers are mostly cyclic, although the
presence of residual linear contaminants cannot be completely ruled
out. Furthermore, the cyclic polymers also eluted later than their
(6) (a) Culkin, D. A.; Jeong, W.; Csihony, S.; Gomez, E. D.; Palsara, N. P.;
Hedrick, J. L.; Waymouth, R. M. Angew. Chem., Int. Ed. 2007, 46, 2627–
2
630. (b) Jeong, W.; Hedrick, J. L.; Waymouth, R. M. J. Am. Chem. Soc.
2007, 129, 8414–8415. (c) Jeong, W.; Shin, E. J.; Culkin, D. A.; Hedrick,
J. L.; Waymouth, R. M. J. Am. Chem. Soc. 2009, 131, 4884–4891. (d)
Raynaud, J.; Absalon, C.; Gnanou, Y.; Taton, D. J. Am. Chem. Soc. 2009,
131, 3201–3209.
7
linear analogues with identical MWs. This implies a smaller
(
(
7) Refer to the Supporting Information for full characterization details.
8) Bloomfield, V.; Zimm, B. H. J. Chem. Phys. 1966, 44, 315–323.
hydrodynamic volume for the former, assuming the absence of non-
size-exclusion separation mechanisms.
(9) (a) Roovers, J. In Cyclic Polymers, 2nd ed.; Semlyen, J. A., Ed.; Kluwer
Academic: Dordrecht, The Netherlands, 2000; pp 347-384. (b) Endo, K.
AdV. Polym. Sci. 2008, 217, 121–183.
In summary, we have demonstrated that cyclic poly(R-peptoid)s
can be synthesized with controlled polymer MWs and narrow MW
distributions in high purity from an NHC-mediated ROP of
N-substituted N-carboxylanhydride. The reaction exhibits charac-
teristics of a living polymerization with minimal chain transfer and
hence allows for the facile synthesis of cyclic diblock copoly-
(10) Jang, S. S.; Cagin, T.; Goddard, W. A. J. Chem. Phys. 2003, 19, 1843–
1
854.
(
11) (a) Hiemenz, P. C.; Lodge, T. P. Polymer Chemistry, 2nd ed.; CRC Press:
Boca Raton, FL 2007; pp 241-242, 334-341. (b) Rubinstein, M.; Colby,
R. H. Polymer Physics; Oxford University Press: New York, 2003; pp 102-
1
04.
(12) Reactions in DMF were conducted at room temperature to inhibit the
solvent-induced polymerization pathway discussed in ref 5j.
(
R-peptoid)s through sequential monomer addition. Elucidation of
the reaction mechanism is currently in progress. Insights obtained
JA907380D
1
8074 J. AM. CHEM. SOC. 9 VOL. 131, NO. 50, 2009