Tandem catalysis for the preparation of cylindrical polypeptide brushes

Article abstract of DOI:10.1021/ja308620h

Here, we report a method for synthesis of cylindrical copolypeptide brushes via N-carboxyanhydride (NCA) polymerization utilizing a new tandem catalysis approach that allows preparation of brushes with controlled segment lengths in a straightforward, one-pot procedure requiring no intermediate isolation or purification steps. To obtain high-density brush copolypeptides, we used a "grafting from" approach where alloc-α-aminoamide groups were installed onto the side chains of NCAs to serve as masked initiators. These groups were inert during cobalt-initiated NCA polymerization and gave allyloxycarbonyl-α-aminoamide-substituted polypeptide main chains. The alloc-α-aminoamide groups were then activated in situ using nickel to generate initiators for growth of side-chain brush segments. This use of stepwise tandem cobalt and nickel catalysis was found to be an efficient method for preparation of high-chain-density, cylindrical copolypeptide brushes, where both the main chains and side chains can be prepared with controlled segment lengths.

Full text of DOI:10.1021/ja308620h

Article
pubs.acs.org/JACS
Tandem Catalysis for the Preparation of Cylindrical Polypeptide
Brushes
Allison J. Rhodes† and Timothy J. Deming*,†,‡
‡
†Department of Chemistry and Biochemistry and Department of Bioengineering, University of California, Los Angeles, California
90095, United States
S
* Supporting Information
ABSTRACT: Here, we report a method for synthesis of cylindrical
copolypeptide brushes via N-carboxyanhydride (NCA) polymerization
utilizing a new tandem catalysis approach that allows preparation of
brushes with controlled segment lengths in a straightforward, one-pot
procedure requiring no intermediate isolation or purification steps. To
obtain high-density brush copolypeptides, we used a “grafting from”
approach where alloc-α-aminoamide groups were installed onto the
side chains of NCAs to serve as masked initiators. These groups were
inert during cobalt-initiated NCA polymerization and gave allyloxycarbonyl-α-aminoamide-substituted polypeptide main chains.
The alloc-α-aminoamide groups were then activated in situ using nickel to generate initiators for growth of side-chain brush
segments. This use of stepwise tandem cobalt and nickel catalysis was found to be an efficient method for preparation of high-
chain-density, cylindrical copolypeptide brushes, where both the main chains and side chains can be prepared with controlled
segment lengths.
cylindrical copolymer brushes has not been achieved. Here, we
report a new method for synthesis of cylindrical copolypeptide
brushes via NCA polymerization utilizing a tandem catalysis
approach that allows preparation of brushes with controlled
segment lengths in a straightforward, one-pot procedure that
requires no intermediate isolation or purification steps.
INTRODUCTION
■
Branched chain copolypeptides have intrigued scientists for
many years. Sela performed pioneering studies on what he
termed “multi-chain polypeptides” that were evaluated for their
immunostimulating properties.1 More recently, hyperbranched
and dendritic polypeptides, with their abundance of functional
groups and three-dimensional globule-like presentation of
functionality, have been found valuable for multiple presenta-
tion of antigens in vaccines, as imaging agents, and for drug and
oligonucleotide delivery.2−7 Although the properties of
branched polypeptides show great promise, controlled synthesis
of these materials remains challenging. Stepwise peptide
dendrimer synthesis provides excellent control over polypep-
tide branching, but is a tedious process that can be difficult to
scale. An alternative is the preparation of hyperbranched or
dendrigraft polypeptides via polymerization of α-amino acid N-
carboxyanhydrides (NCAs).8−12 These materials have been
prepared via a stepwise sequence of polymerization, side-chain
deprotection, and polymerization,8−11 as well as by simulta-
neous polymerization and deprotection.12 Although these
methods are more efficient and scalable compared to dendrimer
preparation, they tend to give limited control over branch
architecture, or provide only low branch density in polymer
brushes.8−12 High-density cylindrical copolypeptide brushes are
desirable synthetic targets since the potential to control their
three-dimensional shape makes them intriguing as components
in block copolymers, which can then be used for preparation of
self-assembled materials with complex morphologies.13−15
Although there have been many successes in controlled
synthesis of cylindrical hybrid−polypeptide copolymer
brushes,16−20 the preparation of entirely polypeptide-based
The synthesis of branched polypeptides via NCA polymer-
ization is challenging since protecting groups typically need to
be removed from side-chain functionalities (e.g., primary
amines of lysine) to generate initiators for branch points.8−12
Although this can be done in situ via hydrogenation to generate
hyperbranched polypeptides,12 this approach is difficult to
control, and more regular branched structures are typically
obtained only when polymers with deprotected side chains are
isolated and purified before resuming polymerization.8−11 To
avoid this need for intermediate purification steps and to obtain
high-density brush copolypeptides, we pursued a “grafting
from” approach where alloc-α-aminoamide groups (alloc =
allyloxycarbonyl) were installed onto the side chains of NCAs
to serve as masked initiators. These groups were envisioned to
be inert during NCA polymerization to give alloc-α-amino-
amide-substituted polypeptide main chains, but then would be
activated in situ to generate NCA polymerization initiators for
side-chain brush growth (Figure 1). We have shown previously
that alloc-α-aminoamides react quantitatively with L2Ni(COD)
(L = donor ligand, COD = cyclooctadiene) in dimethylforma-
mide (DMF) to generate amido-amidate nickelacycles, which
are efficient initiators for living polymerization of NCAs.21 We
Received: August 30, 2012
Published: November 8, 2012
© 2012 American Chemical Society
19463
dx.doi.org/10.1021/ja308620h | J. Am. Chem. Soc. 2012, 134, 19463−19467

Journal of the American Chemical Society
Article
Figure 1. Schematic diagram showing two-stage, one-pot synthesis of cylindrical brush copolypeptides. The N-carboxyanhydride (NCA) component
(blue) of Nε-(alloc-L-methionyl)-L-lysine NCA is first polymerized using (PMe3)4Co initiator to give a linear polypeptide that bears pendant initiator
precursors (green). Side-chain initiators are then activated using dmpeNi(COD), followed by addition of a second NCA monomer (red) to give the
brush copolymers.
have also found (vide infra) that this reaction is selective for
Ni(0), since Co(0) complexes (i.e., (PMe3)4Co) do not react
with alloc-α-aminoamides under similar conditions. Since
(PMe3)4Co reacts directly with NCAs to generate amido-
amidate cobaltacycles, which also initiate the living polymer-
ization of NCAs,22 use of stepwise tandem cobalt and nickel
catalysis should enable the facile synthesis of brush
copolypeptides from alloc-α-aminoamide-substituted NCAs
(Figure 1). A key feature of this process is the envisioned use
of different initiator formation mechanisms for main-chain and
side-chain growth,23 so that these polymerizations can be
sequentially controlled in a single-pot procedure.
L-methionyl)-L-lysine-N-carboxyanhydride (KAM NCA), to take
advantage of the side-chain segment cleavability at this residue.
Polymerization of KAM NCA using (PMe3)4Co in tetrahy-
drofuran (THF) proceeded readily at ambient temperature to
give poly(Nε-(alloc-L-methionyl)-L-lysine), poly(KAM), with
complete monomer conversion and no reaction at the side-
chain alloc groups. To determine chain lengths, KAM NCA was
polymerized at different monomer-to-initiator ratios, and after
complete monomer consumption, active chains were end-
capped with isocyanate-terminated poly(ethylene glycol) (PEG,
Mn = 2000 Da).29 Compositional analysis of purified, end-
capped polymers by 1H NMR gave number-average poly(KAM
)
chain lengths that increased linearly with stoichiometry (Figure
2). Chain length distributions of these poly(KAM) samples were
RESULTS AND DISCUSSION
■
For the synthesis of alloc-α-aminoamide-substituted NCAs, we
chose to use L-lysine as the main-chain-forming NCA, as
substituted lysine NCAs are readily polymerized and the side-
chain amine group is easily functionalized.24−26 We used the
hydrophobic amino acids L-isoleucine and L-methionine for
construction of the alloc-α-aminoamide side-chain groups
(Scheme 1) since these required no protecting groups and
Scheme 1. Synthesis of Allyloxycarbonyl-Aminoamide-
a
Containing NCA Monomers
Figure 2. (a) Molecular weight (Mn, filled circles) and polydispersity
index (Mw/Mn, open circles) of poly(KAM) as a function of monomer-
to-initiator ratio ([M]:[I]) after 100% monomer conversion. Mn and
Mw/Mn were determined by 1H NMR and gel permeation
chromatography (GPC/LS). (b) GPC chromatogram (normalized
LS intensity in arbitrary units versus elution time) of a poly(KAM
sample (Table S1, entry 2).
)
a
Reagents and conditions: (a) allylchloroformate, Na2CO3, β-cyclo-
obtained by gel permeation chromatography (GPC)/LS
analysis and were found to possess low polydispersity indices
(Mw/Mn) between 1.12 and 1.28, indicating the formation of
well-defined polypeptides (Figure 2). Poly(KAM) was prepared
in high yield with precisely controlled chain lengths up to
nearly 300 residues long, and could also be prepared as diblock
copolymers with other amino acids (Table 1). Overall, these
data show that KAM NCA, similar to other NCAs, is able to
undergo living polymerization when initiated with (PMe3)4Co.
Circular dichroism spectroscopy of poly(KAM) in THF revealed
that this polymer, similar to other poly(L-lysine) deriva-
tives,24−26 is predominantly α-helical (see Supporting Informa-
tion (SI), Figure S1), which imparts poly(KAM) with good
solubility in organic solvents and may provide an exposed
presentation of the side-chain alloc-α-aminoamide groups.19
The key feature of poly(KAM) is the reactivity of its side-
chain alloc-L-methionyl groups that will be utilized for
dextrin, H2O, 3.5 h (88% yield). (b) DCC, NHS, THF, 0−21 °C, 1 h
(69% yield). (c) Nα-Cbz-L-Lys-OH, Na2CO3, 1:1 THF:H2O, 21 °C, 48
h (72% yield). (d) DCMME, DCM, 40 °C, 36 h (70% yield). 4a = Nε-
(allyloxycarbonyl-L-methionyl)-L-lysine-N-carboxyanhydride (R =
−CH2CH2SCH3), 4b = Nε-(allyloxycarbonyl-L-isoleucyl)-L-lysine-N-
carboxyanhydride (R = −CH(CH3)CH2CH3).
similar residues had been found to form good initiators in
earlier work.20 Methionine was also chosen since it provides a
means for chemoselective, post-polymerization cleavage of side-
chain segments by reaction with CNBr.27 The methionine- and
isoleucine-derivatized lysine NCA monomers were prepared
using standard methods (Scheme 1) and were obtained in
reasonable yields after purification using flash column
chromatography.28 Although both monomers were found to
be efficiently polymerized using (PMe3)4Co, we have focused
the studies here on the methionine-based monomer, Nε-(alloc-
19464
dx.doi.org/10.1021/ja308620h | J. Am. Chem. Soc. 2012, 134, 19463−19467

Journal of the American Chemical Society
Article
Table 1. Synthesis of Diblock Copolypeptides Using (PMe3)4Co in THF at 21 °C
b
c
first segment
diblock copolymer
a
a
d
entry
1st monomer
2nd monomer
Mn
Mw/Mn
DP
Mn
Mw/Mn
DP
yield (%)
1
2
3
4
17 K NCA
17 K NCA
50 KAM NCA
50 KAM NCA
17 KAM NCA
6 KAM NCA
50 K NCA
25 K NCA
18 700
18 700
31 200
31 200
1.18
1.18
1.11
1.11
71
71
91
91
43 000
29 300
55 000
45 300
1.12
1.29
1.12
1.25
142
102
184
145
99
97
100
100
a
First and second monomers added stepwise to the initiator; number indicates equivalents of monomer per (PMe3)4Co. K NCA = Nε-Cbz-L-lysine-
b
N-carboxyanhydride. KAM NCA = Nε-(alloc-L-methionyl)-L-lysine-N-carboxyanhydride. Molecular weight and polydispersity index after
c
polymerization of the first monomer (determined by GPC/LS for poly(K); determined by GPC/LS and H NMR for poly(KAM)). Molecular
1
d
1
weight and polydispersity index after polymerization of the second monomer (as determined by GPC/LS and H NMR). Total isolated yield of
diblock copolypeptide. DP = degree of polymerization.
cylindrical polypeptide brush growth. To evaluate this
chemistry, the side chains in freshly prepared, unpurified
poly(KAM) (Scheme 2a) were reacted with stoichiometric
a
Scheme 2. Synthesis and Reactivity of Poly(KAM
)
Figure 3. Activation and quenching of alloc side chains in poly(KAM).
(a) Partial 1H NMR spectrum in TFA-d showing alloc proton
resonances of poly(KAM). Resonances: −CH2CHCH2 (5.87 ppm),
−CH2CHCH2 (5.27 ppm), and −CH2CHCH2 and α-carbon
protons (4.60 ppm). (b) Partial 1H NMR spectrum in TFA-d of
poly(KAM) after activation with dmpeNi(COD) followed by
quenching with 4.0 M HCl to remove nickel complexes. By NMR
integrations, 91− 100% of alloc side chains were activated (see SI).
active nucleophilic initiators, we reacted them with different
lengths of isocyanate-terminated PEG (Mn = 350 or 1000 Da),
which have been shown to react quantitatively with active
nickelacycle chain ends (Scheme 2d).29 Analysis of these
products (see SI, Table S2) revealed that the Ni-activated side
chains of poly(KAM) were capped by PEG chains with 91−
100% efficiency, showing that the formation of active
nickelacycle initiators at each side chain in poly(KAM) proceeds
with high efficiency.
a
Reagents and conditions: (a) (PMe3)4Co, THF, 21 °C, 1 h. (b)
dmpeNi(COD), DMF, 80 °C, 16 h. (c) 4.0 M HCl, 21 °C, 2 h. (d) α-
Having verified that the side-chain groups of poly(KAM) can
be converted in situ to active nickelacycle initiators, we next
explored the grafting of cylindrical polypeptide brushes from
these activated polymers (Scheme 2e). To obtain high-chain-
density cylindrical brushes, the initiation efficiency for side-
chain initiating groups needs to be very high to ensure that
polypeptides grow from each side chain. We had previously
found that initiation efficiency for NCA polymerization using
small-molecule nickelacycle initiators was low in THF, due to
nickelacycle aggregation.23 However, initiation efficiency was
quantitative in DMF, which better solubilizes these com-
plexes.23 Here, we also found that growth of side-chain brush
segments was optimal in DMF, and that nickelacycle
aggregation on activated poly(KAM) was further suppressed
by use of 2 equiv of dmpe per nickel center. For the purpose of
being able to adequately characterize side-chain segment
growth, we used a block copolymer main chain, poly(Nε-Cbz-
L-lysine)-block-poly(KAM), poly(K)-b-poly(KAM), where the
poly(K) segment served as a high-molecular-weight end-
group for determination of average side-chain segment lengths
Methoxy-ω-isocyanoethyl-poly(ethylene glycol), PEG-NCO (Mw
350 Da), DMF, 21 °C, 16 h. (e) Bn-Glu NCA, DMF, 21 °C, 16 h. (f)
0.25 M cyanogen bromide in 70% formic acid in water, 4 h.
=
Ni(COD)2 and bis(dimethylphosphino)ethane (dmpe) at 80
°C to generate amido-amidate nickelacycle initiating groups
(Scheme 2b). This reaction is known to proceed quantitatively
on small molecules,21 yet we needed to confirm that active
nickelacycle initiators were also formed in high yield on the
poly(KAM) side chains. As a first step, to determine if all alloc-L-
methionyl groups react with Ni(0), the product of this reaction
was quenched by addition of 4.0 M HCl. While alloc-L-
methionyl groups are stable to these conditions, the activated
nickel complexes are hydrolyzed to give poly(Nε-(L-methionyl)-
1
L-lysine) (Scheme 2c). Analysis of the polymer product by H
NMR showed that at least 91% of the alloc groups had reacted
when stoichiometric Ni(0) was used (Figure 3). It is likely that
higher conversion of alloc groups could be obtained by using
excess Ni(0), but this was not pursued since free Ni(0) will also
react with NCAs and would need to be removed in a
subsequent purification step.23 To show that the alloc-L-
methionyl groups not only react with Ni(0) but also form
1
by H NMR analysis (see SI, Table S3). For initial proof of
concept, γ-benzyl-L-glutamate NCA (E NCA) was used for
19465
dx.doi.org/10.1021/ja308620h | J. Am. Chem. Soc. 2012, 134, 19463−19467

Journal of the American Chemical Society
Article
Figure 4. Lengths and chain length distributions of poly(E) segments grown from activated poly(KAM) side chains. (a) Number-average molecular
weight after 100% monomer conversion of poly(E) segments as a function of monomer-to-activated KAM initiator ratio ([M]:[I]) in poly(K)-b-
poly(KAM) block copolymers. Values were determined using 1H NMR integrations (filled circles), and each data point represents the average of four
repeat experiments (error bars show the range of data obtained). Dotted line represents the expected calculated values. (b) Visualization of PGA
chain length distributions using SDS−PAGE. Lanes 1 and 5: protein molecular weight standards (from top: 25, 20, 15, 10, 5, and 2 kDa). Lanes 2
and 3: synthetic PGA standards (lane 2, Mn = 5120, Mw/Mn = 1.05; lane 3, Mn = 18 900, Mw/Mn = 1.13). Lane 4: PGA sample (Mn = 3820) cleaved
from a brush copolymer using CNBr. (c) GPC chromatogram (normalized LS intensity in arbitrary units versus elution time) of a PGA sample (Mn
= 3820) cleaved from a brush copolymer using CNBr (same as in lane 4 of panel b) that was re-benzylated and found to have Mw/Mn = 1.13.
side-chain segment growth since it forms soluble, α-helical
chains that can be readily distinguished from the lysine-based
main chain. For copolypeptide brush preparation, poly(K)-b-
poly(KAM) copolymers of different segment lengths (Table S3)
were prepared in THF using (PMe3)4Co as described above.
using SDS−PAGE and compared to narrow molecular weight
distribution PGA standards (Mw/Mn = 1.05−1.13) (Figure 4b).
We found that the cleaved PGA segments gave bands similar in
width and location to the standards, indicating narrow chain
length distributions and molecular weights close to expected
values. These results were further supported by rebenzylation
of a PGA sample to give a poly(E) for analysis using GPC/LS
(Figure 4c) that was also found to possess a narrow molecular
weight distribution (Mw/Mn = 1.13), which confirmed the
growth of a high density of uniform brush segments.
Brush copolymers could also be prepared using poly(KAM
)
homopolymers, if desired. After concentration of the crude
reaction mixtures under vacuum followed by dilution with
DMF, the KAM side chains in the copolymers were activated by
reaction with Ni(COD)2 and dmpe at 80 °C to give
macroinitiators that were used directly with no further isolation
or purification steps.
CONCLUSION
Different amounts of E NCA in DMF were added directly to
the block copolymer macroinitiator solutions, resulting in
growth of the brush segments. The polymerizations of E NCA
were found to go to completion and the copolypeptide brushes
were obtained in high yields (Table S3). Compositional
■
The use of tandem cobalt and nickel catalysis has been shown
to be an efficient method for preparation of high chain density,
cylindrical copolypeptide brushes, where both the main chains
and side chains can be prepared with controlled segment
lengths. This new methodology avoids the need for
intermediate deprotection and purification steps, yields well-
defined copolymers, and should be valuable for the
straightforward preparation of new copolypeptide architectures.
We plan to use this chemistry to create three-dimensional
copolymers, with block segments either along the main chain or
in the side chains, which should give rise to new self-assemblies
that can incorporate the useful properties of branched
polypeptides.
1
analysis of the copolypeptide brushes by H NMR showed
that average poly(γ-benzyl-L-glutamate), poly(E), segment
lengths increased linearly with E NCA monomer-to-activated
K
AM initiator ratios and were close to values expected for 100%
brush initiation efficiency, indicating controlled polymerization
of the side-chain segments (Figure 4a). Poly(E) segments were
grown to degrees of polymerization of <25 for ease of
characterization relative to the main chains. However, on the
basis of previous results using similar initiators,21,29 we expect
longer side-chain segments could be grown. To measure chain
length distributions of the poly(E) segments, we utilized the
methionine linker to cleave the brush segments from the main
chain. A model reaction was performed by mixing Nε-(alloc-L-
methionyl)-Nα-Cbz-L-lysine with CNBr, resulting in complete
cleavage of the alloc-L-methionyl group from the lysine residue,
which confirmed the utility of this reaction for chemoselective
ASSOCIATED CONTENT
■
S
* Supporting Information
Experimental procedures and spectral data for all new
compounds; polymerization data; Mn vs [M]/[I] plots; and
CD spectra. This material is available free of charge via the
Internet at http://pubs.acs.org.
peptide cleavage (see SI). A similar reaction on poly(KAM
)
showed that CNBr cleavage at the methionine linkers works
equally as well on this polypeptide, and gave the expected alloc-
homoserine lactone byproduct (Scheme 2f). To perform this
reaction on the copolypeptide brushes, the poly(E) segments
were first deprotected to poly(L-glutamic acid), PGA, to
improve their solubility in polar media, and the brushes were
then incubated with CNBr in aqueous formic acid (see SI). The
resulting cleaved polypeptide segments were then analyzed
AUTHOR INFORMATION
■
Corresponding Author
demingt@seas.ucla.edu
Notes
The authors declare no competing financial interest.
19466
dx.doi.org/10.1021/ja308620h | J. Am. Chem. Soc. 2012, 134, 19463−19467

Journal of the American Chemical Society
Article
ACKNOWLEDGMENTS
■
The authors acknowledge the valuable contributions of Dr.
Zhibo Li and Ilya Yakovlev to the early stages of this project,
and Steevens Alconcel for assistance in obtaining the SDS−
PAGE data. This work was supported by the NSF under award
no. MSN 0956481.
REFERENCES
■
(1) Sela, M. Adv. Immunol. 1966, 5, 29−129.
(2) Kim, Y.; Zeng, F.; Zimmerman, S. C. Chem.Eur. J. 1999, 5,
2133−2138.
(3) Crespo, L.; Sanclimens, G.; Pons, M.; Giralt, E.; Royo, M.;
Albericio, F. Chem. Rev. 2005, 105, 1663−1681.
̌
(4) Niederhafner, P.; Sebestík, J.; Jezek, J. J. Pept. Sci. 2005, 11, 757−
̌
788.
(5) Lubbert, A.; Nguyen, T. Q.; Sun, F.; Sheiko, S. S.; Klok, H.-A.
̈
Macromolecules 2005, 38, 2064−2071.
(6) Vlasov, G. P. Russ. J. Bioorg. Chem. 2006, 32, 227−242.
(7) Hartwig, S.; Nguyen, M. M.; Hecht, S. Polym. Chem. 2010, 1, 69−
71.
(8) North, M.; Birchall, A. C. Chem. Commun. 1998, 1335−1336.
(9) Klok, H.-A.; Rodriguez-Hernan
́
dez, J. Macromolecules 2002, 35,
8718−8723.
(10) Rodriguez-Hernan
́
dez, J.; Gatti, M.; Klok, H.-A. Biomacromo-
lecules 2003, 4, 249−258.
(11) Tsogas, I.; Theodossiou, T.; Sideratou, Z.; Paleos, C. M.; Collet,
H.; Rossi, J. C.; Romestand, B.; Commeyras, A. Biomacromolecules
2007, 8, 3263−3270.
(12) Vlasov, G. P.; Tarasenko, I. I.; Valueva, S. V.; Kipper, A. I.;
Tarabukina, E. B.; Filippov, A. P.; Avdeeva, E. V.; Vorob’ev, V. I.
Polym. Sci. Ser. A 2005, 475, 731−739.
(13) Kim, K. T.; Winnik, M. A.; Manners, I. Soft Matter 2006, 2,
957−965.
(14) Kim, K. T.; Park, C.; Kim, C.; Winnik, M. A.; Manners, I. Chem.
Commun. 2006, 1372−1374.
(15) Wang, J.; Lu, H.; Kamat, R.; Pingali, S. V.; Urban, V. S.; Cheng,
J.; Lin, Y. J. Am. Chem. Soc. 2011, 133, 12906−12909.
(16) Zhang, B.; Fischer, K.; Schmidt, M. Macromol. Chem. Phys.
2005, 206, 157−162.
(17) Breitenkamp, R. B.; Emrick, T. Biomacromolecules 2008, 94,
2495−2500.
(18) Lu, H.; Wang, J.; Lin, Y.; Cheng, J. J. Am. Chem. Soc. 2009, 131,
13582−13583.
(19) Engler, A. C.; Lee, H.; Hammond, P. T. Angew. Chem., Int. Ed.
2009, 48, 9334−9338.
(20) Liu, Y.; Chen, P.; Li, Z. Macromol. Rapid Commun. 2012, 33,
287−295.
(21) Curtin, S. A.; Deming, T. J. J. Am. Chem. Soc. 1999, 121, 7427−
7428.
(22) Deming, T. J. Macromolecules 1999, 32, 4500−4502.
(23) Deming, T. J.; Curtin, S. A. J. Am. Chem. Soc. 2000, 122, 5710−
5717.
(24) Yu, M.; Nowak, A. P.; Pochan, D. J.; Deming, T. J. J. Am. Chem.
Soc. 1999, 121, 12210−12211.
(25) Schaefer, K. E.; Keller, P.; Deming, T. J. Macromolecules 2006,
39, 19−22.
(26) Kramer, J. R.; Deming, T. J. J. Am. Chem. Soc. 2010, 132,
15068−15071.
(27) Lawson, W. B.; Gross, E.; Foltz, C. M.; Witkop, B. J. Am. Chem.
Soc. 1961, 83, 1509−1510.
(28) Kramer, J. R.; Deming, T. J. Biomacromolecules 2010, 11, 3668−
3672.
(29) Brzezinska, K. R.; Curtin, S. A.; Deming, T. J. Macromolecules
2002, 35, 2970−2976.
19467
dx.doi.org/10.1021/ja308620h | J. Am. Chem. Soc. 2012, 134, 19463−19467