"one-Pot" Aminolysis/Thiol-Maleimide End-Group Functionalization of RAFT Polymers: Identifying and Preventing Michael Addition Side Reactions

Article abstract of DOI:10.1021/acs.macromol.6b01512

We show that many of the nucleophiles (catalysts, reducing agents, amines, thiols) present during "one-pot" aminolysis/thiol-maleimide end-group functionalization of RAFT polymers can promote side reactions that substantially reduce polymer end-group functionalization efficiencies. The nucleophilic catalyst 1,8-diazabicyclo[5.4.0]undec-7-ene and the reducing agent tributylphosphine were shown to initiate anionic polymerization of N-methylmaleimide (NMM) in both polar and nonpolar solvents whereas hexylamine-initiated polymerization of NMM occurred only in high-polarity solvents. Furthermore, triethylamine-catalyzed Michael reactions of the representative thiol ethyl 2-mercaptopropionate (E2MP) and NMM in polar solvents resulted in anionic maleimide polymerization when [NMM]₀ > [E2MP]₀. Base-catalyzed enolate formation on the α-carbon of thiol-maleimide adducts was also shown as an alternative initiation pathway for maleimide polymerization in polar solvents. Ultimately, optimal "one-pot" reaction conditions were identified allowing for up to 99% maleimide end-group functionalization of dithiobenzoate-terminated poly(N,N-dimethylacrylamide). Much of the work described herein can also be used to ensure near-quantitative conversion of small molecule thiol-maleimide reactions while preventing previously unforeseen side reactions.

Full text of DOI:10.1021/acs.macromol.6b01512

Article
pubs.acs.org/Macromolecules
“One-Pot” Aminolysis/Thiol−Maleimide End-Group Functionalization
of RAFT Polymers: Identifying and Preventing Michael Addition Side
Reactions
Brooks A. Abel† and Charles L. McCormick*,†,‡
‡
†Department of Polymer Science and Engineering and Department of Chemistry and Biochemistry, The University of Southern
Mississippi, Hattiesburg, Mississippi 39406-5050, United States
S
* Supporting Information
ABSTRACT: We show that many of the nucleophiles (catalysts, reducing
agents, amines, thiols) present during “one-pot” aminolysis/thiol−maleimide
end-group functionalization of RAFT polymers can promote side reactions that
substantially reduce polymer end-group functionalization efficiencies. The
nucleophilic catalyst 1,8-diazabicyclo[5.4.0]undec-7-ene and the reducing agent
tributylphosphine were shown to initiate anionic polymerization of N-
methylmaleimide (NMM) in both polar and nonpolar solvents whereas
hexylamine-initiated polymerization of NMM occurred only in high-polarity
solvents. Furthermore, triethylamine-catalyzed Michael reactions of the
representative thiol ethyl 2-mercaptopropionate (E2MP) and NMM in polar
solvents resulted in anionic maleimide polymerization when [NMM]0
>
[E2MP]0. Base-catalyzed enolate formation on the α-carbon of thiol−
maleimide adducts was also shown as an alternative initiation pathway for
maleimide polymerization in polar solvents. Ultimately, optimal “one-pot” reaction conditions were identified allowing for up to
99% maleimide end-group functionalization of dithiobenzoate-terminated poly(N,N-dimethylacrylamide). Much of the work
described herein can also be used to ensure near-quantitative conversion of small molecule thiol−maleimide reactions while
preventing previously unforeseen side reactions.
conversion at room temperature in the presence of oxygen and
water and typically occurs much more rapidly than analogous
thiol−acrylate or thiol−acrylamide reactions.25−31 Further-
more, thiol−maleimide end-group modification of RAFT
polymers can be performed as “one-pot” reactions without
isolation of the intermediate polymeric thiol (Scheme 1).21
To ensure quantitative polymer end-group functionalization,
a molar excess of maleimide relative to polymeric thiol is
required. It is therefore desirable to optimize the reaction
conditions to favor rapid and efficient polymer conjugation
with minimal excess of functional maleimide, especially when
INTRODUCTION
■
Reversible addition−fragmentation chain transfer (RAFT)
polymerization has made possible the synthesis of functionally
diverse polymers with predetermined molecular weights and
low dispersities (Đ) using a wide variety of monomer types and
polymerization conditions (e.g., aqueous and organic
media).1−3 The versatility of RAFT in synthesizing tailor-
made polymers also stems from the fidelity by which polymer
end-functionality can be controlled. Such end-functionalized
polymers have been used to prepare advanced macromolecular
architectures including block copolymers,4 star copolymers,5
molecular brushes,6,7 and polymer bioconjugates.8,9 Telechelic
RAFT polymers can be synthesized directly by controlling the
RAFT agent R- and Z-group functionality10 or by postpolyme-
rization end-group modification.11−13 The latter approach often
exploits the inherent reactivity of the residual thiocarbonylthio
moiety present on RAFT polymers, allowing for facile removal
and replacement of the unstable RAFT agent with a benign or
functional end-group.
Scheme 1. “One-Pot” Aminolysis/Thiol−Maleimide End-
Group Functionalization of RAFT Polymers
In recent years, reduction or aminolysis of thiocarbonylthio-
terminated RAFT polymers to the corresponding polymeric
thiol has afforded a myriad of thiol “click” end-group
functionalization routes including thiol−isocyanate,14 thiol−
epoxy,15 thiol−halogen,16 thiol−disulfide,17,18 and thiol−ene
reactions.19−24 Particularly advantageous is the thiol−malei-
mide Michael reaction which proceeds to near-quantitative
Received: July 13, 2016
Revised: August 11, 2016
© XXXX American Chemical Society
A
DOI: 10.1021/acs.macromol.6b01512
Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
mercaptan (Fluka, >99%) hexylamine (Aldrich, 99%), 1,8-
diazabicyclo[5.4.0]undec-7-ene (DBU) (Aldrich, >99.0%), triethyl-
amine (Aldrich, >99.5%), tributylphosphine (Aldrich, 97%), trimethyl
phosphite (Aldrich, >99%), dimethyl sulfoxide-d6 (Cambridge
Isotopes, 99.9%), acetonitrile-d3 (Cambridge Isotopes, 99.8%),
methylene chloride-d2 (Cambridge Isotopes, 99.8%), ethanol-D
(Cambridge Isotopes, D 99%, <6% D2O), and deuterium oxide
(Aldrich, 99.9%) were used as received.
Characterization. NMR spectra for structural analysis and kinetic
studies were obtained using a Varian INOVA 300 MHz NMR
spectrometer. Polymer molecular weights and molecular weight
distributions (Mw/Mn) were determined by size exclusion chromatog-
raphy (SEC) using DMF with 20 mM LiBr as the eluent at a flow rate
of 1.0 mL/min in combination with two Agilent PolarGel-M columns
heated to 50 °C and connected in series with a Wyatt Optilab DSP
interferometric refractometer and Wyatt DAWN EOS multiangle laser
light scattering (MALLS) detector (λ = 633 nm). Absolute molecular
weights and Mw/Mn were calculated using a Wyatt ASTRA SEC/LS
software package. Polymer dn/dc in the above eluent at 35 °C was
determined offline using a Wyatt Optilab DSP interferometric
refractometer and Wyatt ASTRA dn/dc software.
using costly biologic, therapeutic, or diagnostic agents.
Typically, thiol−Michael addition reactions employ a base
catalyst (e.g., tertiary amine) to generate the nucleophilic
thiolate species.31 However, nucleophilic catalysts such as
amidines, phosphines, and amines have been used to increase
the rates Michael addition reactions by several orders of
magnitude.25,32−38 The proposed mechanism of nucleophile
“catalyzed” thiol−maleimide Michael addition is illustrated in
Scheme 2.38 Rather than direct deprotonation of thiol,
Scheme 2. Mechanism of Nucleophile-Initiated Thiol−
Maleimide Michael Addition
Reactions of N- and P-Based Nucleophiles with N-
Methylmaleimide. A solution of N-methylmaleimide (25.0 mg,
1
2.23 × 10−4 mol, 10 equiv) and CH2Cl2 (10 μL, H NMR internal
standard) in DMSO-d6 (1.00 mL) was prepared in an NMR tube in
the presence of air. An initial 1H NMR spectrum was taken (t = 0 min)
followed by direct addition of the appropriate nucleophile (2.23 ×
10−5 mol, 1 equiv) to the NMR tube, and the solution was mixed by
inverting three times. Subsequent spectra were taken at timed intervals
and the fractional change in maleimide concentration ([Mal]/[Mal]0)
measured by comparing the relative integrated peak areas of the
maleimide olefin protons (DMSO-d6, 7.02 ppm, 2H) to the protons of
CH2Cl2 (DMSO-d6, 5.76 ppm, 2H).
conjugate addition of the nucleophile to the maleimide double
bond forms the zwitterionic enolate 1, which in turn functions
as a strong base (pKa ≈ 25) capable of generating the
nucleophilic thiolate species while also forming a nucleophile−
succinimide byproduct 2. In this regard, the nucleophile does
not function as a catalyst which is regenerated during each
catalytic cycle, but rather serves as an initiator that generates
the steady state enolate/thiolate concentration necessary for
the thiol−ene chain transfer mechanism to operate. Subsequent
propagation occurs by thiolate addition to maleimide forming
the corresponding enolate 3, which abstracts a proton from
thiol, regenerating the thiolate along with the desired thiol−
maleimide Michael addition product 4.
Recently, while investigating the potential of nucleophilic
catalysts to improve the efficiency of RAFT polymer end-group
functionalization with N-substituted maleimides, we discovered
that in certain instances these catalysts reduce the extent of end-
group functionalization compared to reactions performed using
only a base catalyst. Reagent order of addition and solvent
polarity were also determined to have marked effects on end-
group functionalization efficiency (vide infra). These observa-
tions have prompted this study aimed at understanding the
influences of nucleophile type, solvent, and reaction conditions
on the efficacy of “one-pot” aminolysis/thiol−maleimide end-
group functionalization of RAFT polymers. Furthermore, the
results discussed herein offer new mechanistic insights into
potentially detrimental side reactions that can occur during
thiol−maleimide Michael addition reactions.
Reaction of Ethyl 2-Mercaptopropionate with N-Methyl-
maleimide. A solution of N-methylmaleimide (25.0 mg, 2.23 × 10−4
mol, 10 equiv), triethylamine (3.13 μL, 2.23 × 10−5 mol 1.0 equiv),
1
and CH2Cl2 (10 μL, H NMR internal standard) in DMSO-d6 (1.00
mL) was prepared in an NMR tube in the presence of air. An initial 1H
NMR spectrum was taken (t = 0 min) upon which ethyl 2-
mercaptopropionate (2.90 μL, 2.23 × 10−5 mol 1.0 equiv) was added
to the NMR tube, and the solution was mixed by inverting three times.
Subsequent spectra were taken at timed invervals and the fractional
change in maleimide concentration measured by comparing the
relative integrated peak areas of the maleimide olefin protons (DMSO-
d6, 7.02 ppm, 2H) to the protons of CH2Cl2 (DMSO-d6, 5.76 ppm,
2H).
Synthesis of 3-Benzylsulfanyl-1-methylmaleimide (7). An
initially colorless solution of benzyl mercaptan (2.64 g, 21.2 mmol)
and N-methylmaleimide (2.36 g, 21.2 mmol) in MeCN (50 mL) was
first prepared at room temperature followed by the addition of TEA
(0.281 mL, 2.12 mmol) via syringe. The resulting red solution was
stirred at room temperature for 30 min and quenched with acetic acid
(1.0 mL) to give a colorless solution. The solvent was then removed
by rotary evaporation, and the crude reaction mixture was redissolved
in diethyl ether (100 mL) and washed with 0.1 M HCl (100 mL), H2O
(100 mL), and saturated NaCl (100 mL). The product was further
purified by column chromatography (65:35 hexanes:EtOAc, Rf =
0.35), yielding 7 (4.55 g, 91%) as a viscous oil that solidified into a
1
waxy solid after 7 days; mp 47−52 °C. H NMR (600 MHz, DMSO-
d6): δ 7.31 (m, 4H), 7.24 (m, 1H), 3.93 (dd, J = 62.3, 13.1 Hz, 2H),
3.76 (dd, J = 9.0, 3.9 Hz, 1H), 3.06 (dd, J = 18.5, 9.0 Hz, 1H), 2.78 (s,
3H), 2.45 (dd, J = 18.5, 3.9 Hz, 1H).
EXPERIMENTAL SECTION
■
Reaction of 7 with N-Methylmaleimide. A solution of N-
Materials. 2-Cyano-2-propyl benzodithioate was synthesized
according to a literature procedure.39 Azobis(isobutyronitrile)
(AIBN) (Aldrich, 98%) was recrystallized from anhydrous methanol
and stored at −10 °C. N,N-Dimethylacrylamide (Aldrich, 99%) and
benzylamine (Aldrich, 99%) were vacuum distilled immediately prior
to use. Maleic anhydride (Aldrich, 99%), acetic anhydride (Fisher,
99.2%), sodium acetate (Fisher, anhydrous), N-methylmaleimide
(Aldrich, 97%), ethyl 2-mercaptopropionate (Aldrich, >95%), benzyl
methylmaleimide (25.0 mg, 2.23 × 10−4 mol, 10 equiv), 7 (5.29 mg,
1
2.23 × 10−5 mol, 1.0 equiv), and CH2Cl2 (10 μL, H NMR internal
standard) in DMSO-d6 (1.00 mL) was prepared in an NMR tube in
1
the presence of air. An initial H NMR spectrum was taken upon
which TEA (3.13 μL, 2.23 × 10−5 mol, 1 equiv) was added directly to
the NMR tube, and the solution was mixed by inverting three times.
Subsequent spectra were acquired at timed intervals, and the fractional
B
DOI: 10.1021/acs.macromol.6b01512
Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
Figure 1. Effect of solvent on the time-dependent fractional change in [Mal]/[Mal]0 upon reaction of NMM with representative nucleophiles (a)
hexylamine (HexAM), (b) 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU), (c) tributylphosphine (TBP), and (d) trimethyl phosphite (TMP) as
1
measured by in situ H NMR analysis.
change in maleimide concentration was measured by comparing the
relative integrated peak areas of the maleimide olefin protons (DMSO-
d6, 7.02 ppm, 2H) to the protons of CH2Cl2 (DMSO-d6, 5.76 ppm,
2H).
mic acid intermediate was dried in vacuo and used without further
purification (40.70 g, 97%).
N-Benzylmaleamic acid (40.70 g, 198 mmol) was added as a solid to
a stirred solution of acetic anhydride (90.00 g, 881 mmol) and
anhydrous sodium acetate (13.00 g, 158 mmol), and the reaction was
heated at 100 °C for 30 min, resulting in the formation of a dark
brown homogeneous solution. The reaction mixture was then poured
into a vigorously stirred solution of ice cold water (600 mL) followed
by stirring for 30 min. The resulting brown precipitate was isolated by
vacuum filtration and washed with water (3 × 100 mL). The solids
were resuspended in water (500 mL) and stirred vigorously for 30 min
before isolation again by vacuum filtration. The crude compound was
further purified by recrystallization from ethanol:H2O (2:1, v:v) to
afford N-benzylmaleimide (27.02 g, 73%) as fine beige crystals; mp
67−69 °C. 1H NMR (300 MHz, CDCl3): δ 7.22 (b, 5H), 6.64 (s, 2H),
4.61 (s, 2H). 13C NMR (CDCl3): δ 170.41, 136.17, 134.19, 128.69,
128.38, 127.86, 41.42.
Hydrogen−Deuterium Exchange Kinetics of 7. 1H NMR
spectra were recorded with a Bruker Ascend 600 MHz spectrometer.
Briefly, a solution of 7 (10.0 mg, 4.25 × 10−5 mol, 1 equiv) and D2O
(100 μL, 5.54 mmol, 130 equiv) in DMSO-d6 (0.600 mL) was
1
prepared in an NMR tube in the presence of air, and an initial H
NMR spectrum was acquired (t = 0 min). TEA (5.93 μL, 4.25 × 10−5
mol, 1 equiv) was then added to the NMR tube, and the solution was
mixed by inverting three times. Subsequent spectra were acquired at
timed intervals and the fractional change in peak area (At/A0) of
protons Ha (3.80−3.75 ppm, 1H), Hb (3.10−3.00 ppm, 1H), and Hc
(2.50−2.40 ppm, 1H) were measured relative to the peak area of the
benzylsulfanyl aromatic protons (7.30−7.20 ppm, 5H).
Synthesis of PDMA-CPDB. N,N-Dimethylacrylamide (28.0 g, 282
mmol), 2-cyano-2-propyl benzodithioate (298.0 mg, 1.34 mmol),
AIBN (44.1 mg, 0.27 mmol), and benzene (100 mL) were combined
in a 250 mL round-bottomed flask equipped with magnetic stir bar and
sealed with a rubber septum before purging with N2 for 45 min. The
reaction vessel was then heated in an oil bath at 60 °C for 5 h, upon
which the reaction was quenched via exposure to air and freezing in
liquid nitrogen. The solvent was removed by rotary evaporation, and
the polymer precipitated four times into pentane, redissolving in a
minimal amount of CH2Cl2 between precipitations. The final product
was dried overnight in vacuo before characterizing via 1H NMR (D2O)
and SEC-MALLS (DMF 20 mM LiBr). Mn(NMR) = 3220 g/mol,
Mn(SEC) = 3360 g/mol, Mw/Mn = 1.06.
N-Benzylmaleimide. A solution of maleic anhydride (20.00 g, 204
mmoL) in anhydrous diethyl ether (250 mL) was first prepared at
room temperature in a three-necked 1 L round-bottom flask equipped
with magnetic stir bar, condenser, and addition funnel. A solution of
benzylamine (21.86 g, 204 mmol) in anhydrous diethyl ether (100
mL) was added dropwise via addition funnel over 30 min such that the
exothermic reaction produced a mild reflux of the solvent. The
reaction mixture was stirred for 1 h at room temperature before
isolating the resulting solids by vacuum filtration followed by washing
with anhydrous diethyl ether (100 mL). The isolated N-benzylmalea-
Simultaneous “One-Pot” Aminolysis/Thiol−Maleimide End-
Group Modification of PDMA-CPDB (Method 1). A representative
procedure is as follows: pDMA-CPDB (100.0 mg, 3.10 × 10−5 mol, 1
equiv) and N-benzylmaleimide (29.0 mg, 1.55 × 10−4 mol, 5 equiv)
were measured into a 5 mL test tube equipped with rubber septum
and dissolved in 1.00 mL of DMSO. The reaction mixture was
degassed via three freeze−pump−thaw cycles and backfilled with
argon. 100 μL of a solution of hexylamine in DMSO (102 μL/mL,
7.75 × 10−5 mol, 2.5 equiv) and 100 μL of a solution of DBU in
DMSO (40.0 μL/mL, 3.10 × 10−5 mol, 1 equiv) were then added
sequentially via gastight syringe and the reaction stirred for 12 h at
room temperature (23 °C). End-modified pDMA was purified by
precipitation three times into diethyl ether (50 mL) and dried
1
overnight in vacuo. End-group analysis was performed using H NMR
(D2O) by comparing the integrated peak area of the benzyl aromatic
protons (7.50−7.15 ppm, 5H) to the integrated peak area of the
pDMA N,N-dimethyl side chain and methyne backbone protons
(3.30−2.20 ppm, 213.22H). NMR samples were filtered through a
0.20 μm Millex PTFE filter prior to analysis.
Sequential “One-Pot” Aminolysis/Thiol−Maleimide End-
Group Modification of PDMA-CPDB (Method 2). A representative
procedure is as follows: pDMA-CPDB (100.0 mg, 3.10 × 10−5 mol, 1
equiv) and trimethyl phosphite (18.3 μL, 1.55 × 10−4 mol, 5 equiv)
were measured into a 5 mL test tube equipped with rubber septum
C
DOI: 10.1021/acs.macromol.6b01512
Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
and dissolved in 1.00 mL of DMSO. The reaction mixture was
degassed via three freeze−pump−thaw cycles and backfilled with
argon. 100 μL of a solution of hexylamine in DMSO (102 μL/mL,
7.75 × 10−5 mol, 2.5 equiv) was then added via gastight syringe, and
the reaction was stirred for 30 min at room temperature (23 °C) upon
which a previously degassed solution of N-benzylmaleimide (29.0 mg,
1.55 × 10−4 mol, 5 equiv) in DMSO (0.5 mL) was added and the
reaction stirred for 12 h at room temperature. End-modified pDMA
was purified by precipitation three times into diethyl ether (3 × 50
mL) and dried overnight in vacuo. End-group analysis was performed
show decreases in [Mal]/[Mal]0 to a value of 0.9, beyond
which no change is observed up to 12 h, indicating exclusive
aza-Michael addition takes place in these solvents. Conversely,
reaction of HexAM with NMM in DMSO results in rapid
maleimide consumption, corresponding to an average of 7.5
maleimides per amine within the first 3 min. Subsequently, no
change in [Mal]/[Mal]0 is observed up to 12 h. These results
are consistent with those reported by Azechi et al.40 and
confirm that HexAM can initiate the anionic polymerization of
NMM in polar solvents such as DMSO whereas exclusive aza-
Michael addition takes place in less polar solvents. The effect of
solvent polarity on the reaction of HexAM with NMM is also
readily observed in Figure 1a by noting the increase in reaction
rate with increasing solvent polarity in the order of CH2Cl2 (ε =
8.93) < EtOH (ε = 24.5) < MeCN (ε = 37.5) < DMSO (ε =
46.7).
The time-dependent [Mal]/[Mal]0 plots for the reactions of
DBU and NMM are shown in Figure 1b. Again, reaction rate
increases with increasing solvent polarity with 100% NMM
conversion reached in <10 min in DMSO. Quantitative
consumption of NMM was also observed for reactions
performed in EtOH and MeCN while 71% maleimide
conversion was measured in CH2Cl2 after 12 h. Noteworthy
is that precipitation was observed during the reaction of DBU
and NMM in CH2Cl2 after ∼30 min. Limited solubility of the
propagating macro zwitterionic enolate in CH2Cl2 likely
prevents the propagating chain-end from reacting with NMM
in solution, accounting for the nonquantitative maleimide
conversion observed. Nonetheless, these results indicate that
DBU-initiated polymerization of NMM occurs rapidly and
extensively in both polar and nonpolar solvents.
Interestingly, the kinetic plots in Figure 1b do not rapidly
reach a constant value of [Mal]/[Mal]0 as was observed during
the reaction of HexAM and NMM in Figure 1a. Unlike protic
nucleophiles (e.g., 1°, 2° amines, and thiols), aprotic
nucleophiles such as DBU cannot undergo complete Michael
addition due to the lack of a transferable hydrogen from the
nucleophile. Therefore, if an active hydrogen-containing
compound (e.g., thiol) is in low concentration or completely
absent, the nucleophile-derived zwitterionic enolate 1 can
initiate maleimide propagation without enolate termination by
proton transfer, thus allowing for a living-like polymerization
process to occur. Furthermore, termination by nucleophilic
displacement in macrozwitterionic polymerizations can regen-
erate the nucleophilic initiator and increase polymer molecular
weight even after complete monomer consumption (Scheme
3).44 Evidence of this occurring during the reaction of DBU
with NMM in DMSO (Figure 1b) was garnered by noting the
increased viscosity of the reaction after 12 h and a high
molecular weight polymer peak eluting at the exclusion limit of
our DMF SEC-MALLS system (Figure S1).
Tributylphosphine (TBP) also initiates polymerization of
NMM in all solvents tested (Figure 1c). Reaction rates increase
with solvent polarity with 95% maleimide conversion reached
in 2.5 min in DMSO. Precipitation is also observed during
reactions performed in EtOH and CH2Cl2 after ∼5 min
following the addition of TBP. Poor solubility of the
propagating phosphonium zwitterionic enolate is again a likely
reason for the nonquantitative maleimide conversions achieved
in EtOH and CH2Cl2 during the time frame of these kinetic
experiments.
1
using H NMR (D2O) by comparing the integrated peak area of the
benzyl aromatic protons (7.50−7.15 ppm, 5H) to the integrated peak
area of the pDMA N,N-dimethyl side chain and methyne backbone
protons (3.30−2.20 ppm, 213.22H). NMR samples were filtered
through a 0.20 μm Millex PTFE filter prior to analysis.
RESULTS AND DISCUSSION
■
Nucleophile-Promoted Michael Addition Side Reac-
tions. Preliminary efforts in our lab to catalyze the “one-pot”
aminolysis/thiol−maleimide end-group modification of acryl-
amido RAFT polymers in DMSO using the amidine 1,8-
diazabicyclo[5.4.0]undec-7-ene (DBU) resulted in (i) low end-
group functionalization efficiencies as compared to reactions
performed in the absence of DBU, (ii) quantitative maleimide
consumption, and (iii) the formation of high molecular weight
impurities observable by SEC-MALLS. From these observa-
tions we hypothesized that the presence of a strong nucleophile
(DBU) results in zwitterionic polymerization of maleimide
occurring faster than the desired thiol−maleimide Michael
addition. Accordingly, it is also possible that other nucleophiles
(e.g., amines and trialkylphosphines) commonly used during
“one-pot” aminolysis/thiol−maleimide modification of RAFT
polymers can promote similar side reactions. Recently, Azechi
et al. showed that 1°, 2°, and 3° amines initiate the anionic
polymerization of N-substituted maleimides in highly polar
aprotic solvents (DMSO and DMF), whereas no polymer-
ization is observed in less polar solvents (THF).40 Additionally,
trialkylphosphines have been used as initiators for the
zwitterionic polymerizations of methylene malonic esters,41
cyanoacrylates,42 and N-substituted maleimides.43 We therefore
chose to first investigate the effect of solvent on the reactions of
N-substituted maleimides with nucleophiles commonly em-
ployed during “one-pot” aminolysis/thiol−maleimide RAFT
polymer reactions.
Reaction of N- and P-Based Nucleophiles with N-
Methylmaleimide. Initially, we examined the effect of solvent
on the reactions of representative N- and P-based nucleophiles
with N-methylmaleimide (NMM) in the absence of thiol as
outlined in Figure 1. A stoichiometric excess of maleimide
relative to nucleophile ([Nu]0:[Mal]0 = 1.0:10) was chosen to
reflect the relative concentrations of these reagents used during
RAFT polymer end-group thiol−maleimide reactions and to
elucidate whether or not maleimide polymerization was
occurring. The time-dependent fractional change in maleimide
1
concentration ([Mal]/[Mal]0) was monitored using in situ H
NMR by following the peak area of the maleimide olefin
protons relative to the peak area of an internal standard
(CH2Cl2).
Figure 1a shows the time-dependent [Mal]/[Mal]0 plots for
the reactions of hexylamine (HexAM) and NMM in different
solvents with the dashed line ([Mal]/[Mal]0 = 0.9)
representing the theoretical decrease in [Mal]/[Mal]0 predicted
for the reaction of HexAM and NMM via single aza-Michael
addition. Reactions conducted in MeCN, EtOH, and CH2Cl2
From these results it is apparent that DBU- and TBP-
initiated maleimide polymerization is unavoidable, even in low
D
DOI: 10.1021/acs.macromol.6b01512
Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
similarity to the polymeric thiol that would be produced upon
aminolysis of a polyacrylate RAFT polymer.
As shown in Figure 2, only the reaction of E2MP and NMM
in CH2Cl2 gives the change in [Mal]/[Mal]0 expected for
Scheme 3. Proposed Mechanisms of Initiation, Propagation,
and Termination for the Nucleophile-Initiated Anionic
Polymerization of N-Substituted Maleimides
Figure 2. Effect of solvent on the time-dependent fractional change in
polarity solvents. While DBU is optional during “one-pot”
aminolysis/thiol−maleimide modification of RAFT polymers,
mild reducing agents such as TBP are generally required to
prevent disulfide formation from occurring between polymeric
thiols during the RAFT agent aminolysis step.20,26 Recently, Ho
et al. reported the use of trialkyl phosphites as cheaper and less
toxic alternatives to trialkylphosphines as reducing agents
during “one-pot” RAFT polymer aminolysis/thiol−ene reac-
tions.45 While trialkyl phosphites can undergo conjugate
addition to electron-deficient olefins, they are less nucleophilic
than phosphines and typically require elevated temperatures
(100 °C) for such reactions to occur.46,47 As shown in Figure
1d, the reaction of trimethyl phosphite (TMP) with NMM
results in no measurable change in [Mal]/[Mal]0 in MeCN,
EtOH, and CH2Cl2. Only after prolonged reaction times (12 h)
in DMSO is a 65% decrease in [Mal]/[Mal]0 observed.
Consequently, TMP is a suitable alternative to phosphines as a
reducing agent during thiol−maleimide reactions when used in
less polar solvents. It should also be noted that trace amounts
of water must be present for trialkyl phosphite (and phosphine)
reduction of disulfides to occur.48 Therefore, rigorous
anhydrous conditions should be avoided during RAFT polymer
aminolysis when using trialkyl phosphites as reducing agents.
Reaction of Thiols with N-Methylmaleimide. It is
evident from the kinetic plots shown in Figure 1 that strong
N- and P-based nucleophiles can react with maleimides to form
either the Michael addition product or polymaleimide depend-
ing upon the nucleophile type (protic or aprotic) and solvent
polarity. Therefore, it is plausible that other strong nucleophiles
such as thiolates can initiate polymerization of maleimides in
polar solvents. To this end, we chose to investigate the reaction
of ethyl 2-mercaptopropionate (E2MP) with NMM under
conditions identical to those outlined in Figure 1, except with
the addition of TEA to generate the nucleophilic thiolate
species. E2MP was chosen as a model thiol due to its structural
[Mal]/[Mal]0 during the TEA-catalyzed reaction of E2MP with NMM
1
as measured by in situ H NMR analysis.
exclusive thiol−maleimide Michael addition ([Mal]/[Mal]0 =
0.9, dashed line). Meanwhile, the reactions conducted in more
polar solvents (DMSO, EtOH, and MeCN) show a continued
decrease in [Mal]/[Mal]0 up to 12 h, with 100% maleimide
conversion reached in 90 min in DMSO. Initially, we
anticipated the kinetic profiles for the reactions of E2MP and
NMM (Figure 2) to rapidly reach a constant [Mal]/[Mal]0
value due to rapid enolate termination by means of proton
transfer, as previously observed for the reactions of HexAM and
NMM (Figure 1a). However, a continued decrease in [Mal]/
[Mal]0 up to 12 h is evidence of either slow reaction of thiol or
an additional reaction pathway.
Closer inspection of the kinetic profiles for reactions
conducted in EtOH and MeCN reveals an inflection point at
the first time point (2.5 min) when [Mal]/[Mal]0 ≈ 0.9,
indicating that thiol−maleimide Michael addition likely occurs
faster than subsequent maleimide propagation in these solvents.
1
Furthermore, in situ H NMR analysis was used to show 100%
thiol conversion was reached by the first time point (2.5 min)
for reactions conducted in DMSO, MeCN, and CH2Cl2
(Figures S2−S4, respectively). Thiol conversion could not be
measured for the reaction performed in EtOH-d6 due to
deuterium exchange between the solvent and sulfhydryl proton.
Quantitative thiol consumption early in each reaction requires
that proton transfer from thiol to enolate also occur rapidly.
Therefore, the continued decrease in [Mal]/[Mal]0 observed at
longer reaction times in more polar solvents is not due to slow
reaction of thiol and is indicative of an alternate maleimide
reaction pathway.
One possible explanation for the continued decrease in
[Mal]/[Mal]0 at longer reaction times is TEA-initiated
E
DOI: 10.1021/acs.macromol.6b01512
Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
maleimide polymerization. The kinetic plots for the reactions of
TEA with NMM in DMSO and EtOH (Figure S5) show a
decrease in [Mal]/[Mal]0 of 0.07 and 0.04, respectively, after
12 h. Meanwhile, no change in [Mal]/[Mal]0 was observed in
MeCN and CH2Cl2 after 12 h. While this confirms TEA can
initiate polymerization of NMM in polar solvents, the minimal
change in [Mal]/[Mal]0 observed after 12 h in DMSO and
EtOH, and the lack of reaction in MeCN, is insufficient to
account for the decrease in [Mal]/[Mal]0 that occurs after the
same time period during the reactions of E2MP with NMM in
these solvents.
Another potential explanation for the continued decrease in
[Mal]/[Mal]0 observed in the kinetic plots of Figure 2 involves
deprotonation of the thiol−maleimide Michael adduct by TEA
to regenerate the nucleophilic enolate species. Initially, this
seems unlikely since the pKa of TEA in DMSO (10.75) is much
lower than that of most succinimide-derived enolates (∼25).
However, enolate formation of 2-aminosuccinimide residues in
peptides has been shown to occur under mildly basic conditions
in aqueous media (pH = 7.4), indicating that heteroatom
substitution of the succinimide α-carbon can reduce enolate
pKa compared to unsubstituted succinimides.49
Thiol−Maleimide Adducts as Nucleophiles. To test
whether thiol−maleimide Michael addition products are
capable of reinitiating maleimide polymerization in the
presence of TEA, the Michael adduct of benzyl mercaptan
and NMM (7) was synthesized and purified by column
chromatography. It should be noted that the Michael adduct of
E2MP and NMM was not synthesized due to potential
complications arising from the presence of chemically distinct
Figure 3. Effect of solvent on the time-dependent fractional change in
[Mal]/[Mal]0 during the TEA-catalyzed reaction of 7 with NMM as
measured by in situ H NMR analysis.
1
TEA. Figure 4a shows the time-dependent fractional change in
peak area (At/A0) for the three chemically distinct protons of 7
which could be abstracted by TEA to form an enolate on either
the α-carbon (Ha) or the β-carbon (Hb and Hc) relative to the
benzylsulfanyl group. For simplicity, only the structure of the S
enantiomer is shown in Figure 4a. Interestingly, H−D exchange
was only observed for proton Ha with a 97% decrease in At/A0
occurring by 34 min. Sigma withdrawing effects by the adjacent
thioether are most likely responsible for the increased acidity of
Ha, leading to exclusive enolate formation at the α-carbon.
1
diastereomers and the complexity in identifying the H NMR
chemical shifts of the resulting four stereoisomers. Meanwhile,
the reaction of benzyl mercaptan and NMM affords a racemic
mixture of chemically indistinguishable enantiomers.
As shown in Figure 3, [Mal]/[Mal]0 decreases with time
during the TEA-catalyzed reactions of 7 and NMM in DMSO,
EtOH, and MeCN whereas no change in [Mal]/[Mal]0 is
observed in CH2Cl2 up to 12 h. The kinetic profiles obtained in
each solvent for the reactions of 7 and NMM in Figure 3
accurately reflect the kinetic profiles observed below [Mal]/
[Mal]0 = 0.9 (dashed line) for the analogous reactions of E2MP
and NMM shown in Figure 2. This is not a surprising result
since the dashed line in Figure 2 represents the change in
[Mal]/[Mal]0 predicted for formation of the thiol−maleimide
Michael adduct. Any decrease in [Mal]/[Ma]0 below a value of
0.9 in Figure 2 would presumably occur as a result of TEA-
promoted regeneration of the nucleophilic enolate species and
therefore resemble the kinetic profiles in Figure 3.
Also worth noting is that TEA has little influence on the
time-dependent [Mal]/[Mal]0 plots for the reaction of HexAM
and NMM in DMSO as seen in Figure S6 of the Supporting
Information. Only after 12 h was a minimal decrease in [Mal]/
[Mal]0 observed for the TEA-catalyzed reaction of HexAM and
NMM relative to the [Mal]/[Mal]0 values measured for the
same reaction performed in the absence of TEA. This indicates
that TEA-catalyzed reactions of thiol−maleimide adducts with
additional maleimide may be unique among heteroatom−
maleimide Michael addition products.
1
Figure 4b shows the H NMR spectral overlay of select time
points during the H−D exchange experiments with 7. The peak
corresponding to Ha decreases in area with time while
maintaining the doublet of doublets (dd) splitting pattern
that arises from spin−spin coupling with Hb and Hc.
Meanwhile, the peaks corresponding to the geminal protons
Hb and Hc do not change in area, but rather show a change in
splitting pattern from dd to d as spin−spin coupling with Ha is
diminished through deuterium exchange. Also apparent in
Figure 4b is the significant downfield chemical shift of Ha (3.78
ppm) relative to Hb (3.07 ppm) and Hc (2.46 ppm) that arises
from the deshielding (sigma withdrawing) effects of the
benzylsulfanyl group.
Scheme 4 shows the proposed reaction pathways for the
TEA-catalyzed thiol−maleimide reaction when a stoichiometric
excess of maleimide relative to thiol is employed in polar
solvents. Thiolate addition to maleimide forms the β-enolate 3
+
which can either abstract a proton from thiol or NH(Et)3 to
give the thiol−maleimide adduct 4 or react directly with
maleimide to form the propagating species 8 when conditions
favor propagation over termination (i.e., [Mal] ≫ [thiol]).
Deprotonation of 4 by TEA forms the α-enolate 9 which can
reversibly terminate by proton transfer or react with maleimide
to form the propagating species 10. The propagating species 8
and 10 derived from β- and α-enolates, respectively, can
continue to react with maleimide until termination occurs by
We next sought to provide direct evidence that TEA is a
strong enough base to generate the enolate of 7 in polar
1
solvents. To this end, we used in situ H NMR analysis to
measure the relative rates of hydrogen−deuterium (H−D)
exchange of 7 in a DMSO-d6/D2O mixture in the presence of
F
DOI: 10.1021/acs.macromol.6b01512
Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
proton transfer. Also, the termination products of 8 can in
theory reinitiate maleimide polymerization by α-enolate
formation. These additional reaction pathways can account
for the continued decrease in [Mal]/[Mal]0 following initial
formation of the thiol−maleimide Michael adduct for the TEA-
catalyzed reactions of E2MP with NMM shown in Figure 2. It
should also be noted that under these conditions exclusive
enolate formation at the α-carbon rules out the possibility of
TEA-catalyzed β-elimination (retro Michael addition) as a
means of regenerating the nucleophilic thiolate species after
formation of the thiol−maleimide Michael adduct.
“One-Pot” Aminolysis/Thiol−Maleimide End-Group
Functionalization of RAFT Polymers. It is apparent that
many of the reagents (e.g., amines and phosphines) commonly
used during “one-pot” RAFT polymer aminolysis/thiol−
maleimide reactions can react with maleimides to form
polymaleimide or Michael addition byproducts and potentially
outcompete the desired polymeric thiol−maleimide Michael
reaction. We have also shown that intermediate strength bases
such as TEA can deprotonate thiol−maleimide Michael adducts
in polar solvents to form a nucleophilic α-enolate capable of
subsequent reaction with maleimide. However, the effects of
these side reactions on RAFT polymer end-group functional-
ization efficiency are not yet well understood. Also worth
investigating is the influence of aminolysis method on end-
group functionalization efficiency. The simplest method
involves simultaneous aminolysis of the RAFT polymer in the
presence of maleimide and does not require the use of a
reducing agent to prevent disulfide formation. However, the
competing amine−maleimide aza-Michael addition could
prevent quantitative RAFT agent aminolysis and subsequently
reduce the degree of end-group functionalization. Alternatively,
a sequential method allows for complete aminolysis of the
RAFT agent to occur prior to the addition of maleimide. This
route minimizes side reactions between the amine and
maleimide but necessitates the use of a reducing agent to
prevent polymeric disulfide formation from occurring during
the aminolysis stage.
Figure 4. (a) Time-dependent fractional change in peak area (At/A0)
for protons Ha, Hb, and Hc during TEA-catalyzed H−D exchange of 7
1
in DMSO. (b) H NMR spectral overlay of select time points during
H−D exchange experiments with 7.
To investigate the effects of solvent, catalyst, reducing agent,
and aminolysis method on RAFT polymer end-group
functionalization efficiency, we first synthesized the water-
soluble polymer poly(N,N-dimethylacrylamide) (pDMA-
CPDB) using the RAFT agent 2-cyano-2-propyl benzo-
dithioate. Low dispersities (Đ = 1.06) and excellent agreement
between the number-average molecular weights determined by
size exclusion chromatography (Mn(SEC) = 3360 g/mol) and
Scheme 4. Proposed Reaction Pathways for the TEA-
Catalyzed Thiol−Maleimide Reaction in Which Reversible
Enolate Formation Is Operational
1
by H NMR (Mn(NMR) = 3220 g/mol) are evidence of a
controlled polymerization and high dithiobenzoate chain-end
fidelity. “One-pot” reactions of pDMA-CPDB with N-benzyl-
maleimide (BnM) were conducted at room temperature for 12
h using the initial molar ratios of [pDMA-CPDB]0:[HexAM]0:
1
[BnM]0 = 1.0:2.5:5.0 (Scheme 5). End-group analysis by H
Scheme 5. “One-Pot” Aminolysis/Thiol−Maleimide
Modification of PDMA-CPDB with BnM
G
DOI: 10.1021/acs.macromol.6b01512
Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
NMR was performed in D2O by comparing the integrated peak
area of the benzyl aromatic protons (7.50−7.15 ppm) to the
known integrated peak area of the pDMA N,N-dimethyl side
chain and methyne backbone protons (3.30−2.20 ppm)
(Figures S7−S27). The poor solubility of BnM and its
nucleophile-initiated byproducts (i.e., poly(BnM)) in D2O
allows for accurate quantification of only BnM that is covalently
attached to pDMA.
our preliminary DBU-catalyzed RAFT polymer functionaliza-
tion reactions was DBU-initiated maleimide polymerization
outcompeting the desired polymeric thiol−maleimide reaction.
Also worth noting is the incomplete aminolysis observed for
the reaction performed in EtOH (entry 1c) with 15% of
dithibenzoate end-groups remaining after 12 h and can be
attributed to the amine−maleimide aza-Michael addition
reaction competing with CTA aminolysis.
Table 1 summarizes the results of the “one-pot” aminolysis/
thiol−maleimide reactions of pDMA-CPDB with BnMA with
Entries 2a−2d in Table 1 show the effect of simultaneous
aminolysis/thiol−maleimide Michael addition in the absence of
DBU on end-group functionalization efficiency. Incomplete
aminolysis was observed in all solvents except for CH2Cl2 (2a)
with reactions performed in DMSO (2d) and EtOH (2c)
retaining 46% and 90%, respectively, of the original
dithiobenzoate functionality. Accordingly, only 31% and 8%
end-group functionalization was observed in DMSO and EtOH,
respectively. These results show that the amine−maleimide aza-
Michael addition can occur faster than CTA aminolysis in more
polar solvents and are consistent with the effect of solvent
polarity on the reaction rates of HexAM with NMM shown in
Figure 1a.
Table 1. Summary of “One-Pot” Aminolysis/Thiol−
a
Maleimide Reactions of PDMA-CPDB with BnM
e
f
g
catalyst/red
agent
funct
(%)
CTA
(%)
color
b
entry method
solvent
(12 h)
c
1a
1b
1c
1d
2a
2b
2c
2d
3a
3b
3c
3d
4a
4b
4c
4d
5a
5b
5c
5d
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
2
2
CH2Cl2
MeCN
EtOH
DBU
93
37
25
0
0
0
R
R
R
R
Y
O
O
R
Y
R
Y
R
R
R
R
R
Y
Y
Y
R
DBU
DBU
DBU
−
−
−
−
−
−
−
15
0
DMSO
CH2Cl2
MeCN
EtOH
89
81
8
0
3
The reactions performed in entries 3a−3d of Table 1 were
identical to those performed in entries 2a−2d except the
HexAM was allowed to react with pDMA-CPDB for 30 min
prior to the addition of BnM (method 2). CTA aminolysis was
qualitatively confirmed to occur within 30 min by noting the
change in color that takes place as the dithiobenzoate end-
groups (orange) are aminolyzed to the corresponding N-
hexylthiobenzamide (yellow). As seen in entries 3a−3d, 100%
CTA aminolysis was accomplished in all solvents while end-
group functionalization efficiencies were significantly improved
compared to reactions conducted using the simultaneous
aminolysis method (entries 2a−2d). High end-group function-
alization (95%) was achieved in the less polar aprotic solvents
CH2Cl2 (3a) and MeCN (3b) while only moderate degrees of
functionalization were obtained in EtOH (84%, 3c) and DMSO
(72%, 3d). While these results are promising, it is well-known
that disulfide coupling of polymeric thiols can occur during
CTA aminolysis, resulting in both reduced end-group
functionalization efficiencies and high molecular weight
impurities, thus necessitating the use of a reducing agent.20
Entries 4a−4d of Table 1 summarize the effects of TBP as a
reducing agent on end-group functionalization efficiency when
using the sequential aminolysis method. The use of TBP as a
reducing agent results in decreased end-group functionalization
efficiencies for reactions performed in CH2Cl2 (4a) and MeCN
(4b) compared to analogous reactions conducted without TBP
(entries 3a and 3b, respectively). Meanwhile, no effect of TBP
was observed on the functionalization efficiencies of reactions
performed in EtOH and DMSO (entries 4c and 4d,
respectively). From these results and the kinetic plots in
Figure 1c, we conclude that trialkylphosphines are not suitable
for use as reducing agents during thiol−maleimide end-group
modification of RAFT polymers due to competing phosphine-
initiated maleimide polymerization. Alternatively, using the less
nucleophilic TMP as a reducing agent affords substantially
increased degrees of end-group functionalization in all solvents
as seen in Table 1 entries 5a−5d with 98% and 99% end-group
functionalization achieved in CH2Cl2 (5a) and MeCN (5b),
respectively.
90
46
0
DMSO
CH2Cl2
MeCN
EtOH
31
95
95
84
72
89
86
84
72
98
99
86
89
0
0
DMSO
CH2Cl2
MeCN
EtOH
−
TBP
0
d
0
TBP
TBP
TBP
0
0
DMSO
CH2Cl2
MeCN
EtOH
0
d
TMP
0
TMP
TMP
TMP
0
0
DMSO
0
a
b
[pDMA-CPDB]0:[HexAM]0:[BnM]0 = 1.0:2.5:5.0. Method 1:
simultaneous aminolysis/thiol−maleimide; method 2: sequential
c
aminolysis/thiol−maleimide. [pDMA-CPDB]0:[DBU]0 = 1.0:1.0.
d
e
[pDMA-CPDB]0:[TBP or TMP]0 = 1.0:5.0. Percent end-group
functionalization of pDMA-CPDB with BnM as measured by 1H
NMR. Percent CTA remaining after reaction of pDMA-CPDB with
BnMA as measured by H NMR. Reaction color after 12 h: red (R),
orange (O), or yellow (Y).
f
g
1
the last column indicating the color exhibited by each reaction
after 12 h. It should be noted that all reactions performed in
this work that resulted in maleimide polymerization also
became red in color. Addition of trifluoroacetic acid to these
reactions gave near colorless solutions, thus indicating that the
red coloration is likely the result of a persistent enolate
concentration and allows for reaction color to be used as a
qualitative indicator of zwitterionic/anionic maleimide poly-
merization.
Entries 1a−1d reflect our initial attempts to catalyze thiol−
maleimide end-group functionalization reactions with DBU as a
catalyst and were performed using the simultaneous aminolysis
method (method 1). A general trend of decreasing end-group
functionalization efficiency with increasing solvent polarity was
observed with 0% functionalization achieved in DMSO. These
results are relatable to the trend observed for the reactions of
DBU and NMM shown in Figure 1b, where the rate of DBU-
initiated maleimide polymerization increases with solvent
polarity. These results confirm that the cause of failure for
Stoichiometric Considerations. Efficient and quantitative
end-group functionalization of RAFT polymers using ami-
nolysis/thiol−maleimide chemistry also requires consideration
H
DOI: 10.1021/acs.macromol.6b01512
Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
of the initial reactant stoichiometry. Ideally, minimal excess of
maleimide should be used relative to polymeric thiol to limit
the waste of potentially costly N-substituted maleimide
compounds. However, inevitable side reactions such as
amine−maleimide aza-Michael addition must be taken into
account when choosing reactant stoichiometry such that [Mal]0
> [PnSH] + [RNH2] where [PnSH] and [RNH2] are the
polymeric thiol and unreacted amine concentrations, respec-
tively, after complete RAFT agent aminolysis has occurred. In
this work, we found that aminolysis of pDMA-CPDB with
HexAM using a molar ratio of [pDMA-CPDB]0:[HexAM]0 =
1.0:2.5 results in complete loss of dithiobenzoate end-groups
within 30 min. However, other work conducted by our group
(not reported herein) has shown that dithiobenzoate-functional
polystyrene synthesized by RAFT requires several hours for
complete aminolysis to occur using the same dithiobenzoate to
amine ratio. Therefore, the reactant feed ratios reported herein
should be considered a starting point for stoichiometric
optimization of different RAFT polymer systems.
The type of RAFT agent being aminolyzed must also be
considered when choosing reactant stoichiometry. Dithioben-
zoate-terminated polymers react with 1 equiv of amine to yield
polymeric thiol and thiobenzamide byproducts in equimolar
amounts. Conversely, trithiocarbonate-terminated polymers can
react with 2 equiv of amine to give the polymeric thiol, Z-group
derived thiol, and thiourea byproduct in equimolar amounts. In
this case, the reactant stoichiometry must allow for [Mal]0 >
[PnSH] + [RNH2] + [ZSH], where [ZSH] is the concentration
of small molecule Z-group derived thiol.
AUTHOR INFORMATION
■
Corresponding Author
*E-mail: charles.mccormick@usm.edu (C.L.M.).
Notes
The authors declare no competing financial interest.
Paper 160 in a series entitled “Water-Soluble Polymers.”
ACKNOWLEDGMENTS
■
The authors gratefully acknowledge the financial support
provided in part by the National Science Foundation Graduate
Research Fellowship Program under Grant GM004636/
GR04355 and the National Science Foundation’s EPSCoR
under Agreement IIA1430364.0.
REFERENCES
■
(1) Moad, G.; Rizzardo, E.; Thang, S. H. Living Radical
Polymerization by the RAFT Process. Aust. J. Chem. 2005, 58, 379−
410.
(2) Lowe, A. B.; McCormick, C. L. Reversible addition-fragmentation
chain transfer (RAFT) radical polymerization and the synthesis of
water-soluble (co)polymers under homogeneous conditions in organic
and aqueous media. Prog. Polym. Sci. 2007, 32, 283−351.
(3) Hill, M. R.; Carmean, R. N.; Sumerlin, B. S. Expanding the Scope
of RAFT Polymerization: Recent Advances and New Horizons.
Macromolecules 2015, 48, 5459−5469.
(4) Inglis, A. J.; Sinnwell, S.; Stenzel, M. H.; Barner-Kowollik, C.
Ultrafast Click Conjugation of Macromolecular Building Blocks at
Ambient Temperature. Angew. Chem., Int. Ed. 2009, 48, 2411−2414.
(5) Chan, J. W.; Yu, B.; Hoyle, C. E.; Lowe, A. B. The nucleophilic,
phosphine-catalyzed thiol-ene click reaction and convergent star
synthesis with RAFT-prepared homopolymers. Polymer 2009, 50,
3158−3168.
(6) Patton, D. L.; Advincula, R. C. A Versatile Synthetic Route to
Macromonomers via RAFT Polymerization. Macromolecules 2006, 39,
8674−8683.
(7) Li, Z.; Zhang, K.; Ma, J.; Cheng, C.; Wooley, K. L. Facile
Synthesis of Cylindrical Molecular Brushes by a Sequential RAFT and
ROMP “Grafting-Through” Methodology. J. Polym. Sci., Part A: Polym.
Chem. 2009, 47, 5557−5563.
(8) Bulmus, V. RAFT polymerization mediated bioconjugation
strategies. Polym. Chem. 2011, 2, 1463−1472.
(9) Li, M.; De, P.; Gondi, S. R.; Sumerlin, B. S. Responsive Polymer-
Protein Bioconjugates Prepared by RAFT Polymerization and Copper-
Catalyzed Azide-Alkyne Click Chemistry. Macromol. Rapid Commun.
2008, 29, 1172−1176.
(10) Tasdelen, M. A.; Kahveci, M. U.; Yagci, Y. Telechelic polymers
by living and controlled/living polymerization methods. Prog. Polym.
Sci. 2010, 36, 455−567.
CONCLUSIONS
■
New insights into nucleophile-promoted thiol−maleimide side
reactions have been provided. Nonquantitative end-group
functionalization of RAFT polymers using “one-pot” aminol-
ysis/thiol−maleimide reactions was determined to be the result
of nucleophile-initiated maleimide polymerization occurring
faster than the desired thiol−maleimide Michael reaction.
Furthermore, previously unreported base-catalyzed α-enolate
formation of thiol−maleimide Michael adducts in polar solvents
was shown to promote multiple maleimide additions per thiol.
Ultimately, such side reactions can be prevented by conducting
RAFT polymer thiol−maleimide end-group modification
reactions in less polar solvents while avoiding the use of strong
aprotic nucleophiles such as amidine catalysts and phosphine
reducing agents.
(11) Moad, G.; Rizzardo, E.; Thang, S. H. End-functional polymers,
thiocarbonylthio group removal/transformation and reversible addi-
tion-fragmentation-chain transfer (RAFT) polymerization. Polym. Int.
2011, 60, 9−25.
(12) Willcock, H.; O’Reilly, R. K. End group removal and
modification of RAFT polymers. Polym. Chem. 2010, 1, 149−157.
(13) Harvison, M. A.; Roth, P. J.; Davis, T. P.; Lowe, A. B. End
Group Reactions of RAFT-Prepared (Co)polymers. Aust. J. Chem.
2011, 64, 992−1006.
ASSOCIATED CONTENT
■
S
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.macro-
mol.6b01512.
SEC RI chromatogram of DBU-initiated poly(N-methyl-
maleimide), 1H NMR spectra of select time points
during reactions of E2MP and NMM in different
solvents, kinetic plots for reaction of TEA with NMM
in different solvents, kinetic plots for reaction of HexAM
with NMM in different solvents, and 1H NMR spectra of
N-benzylmaleimide end-functionalized poly(N,N-
dimethylacrylamide) (PDF)
(14) Li, H.; Yu, B.; Matsushima, H.; Hoyle, C. E.; Lowe, A. B. The
Thiol-Isocyanate Click Reaction: Facile and Quantitative Access to ω-
End-Functional Poly(N,N-diethylacrylamide) Synthesized by RAFT
Radical Polymerization. Macromolecules 2009, 42, 6537−6542.
(15) Harvison, M. A.; Davis, T. P.; Lowe, A. B. Macromolecular
thiolysis of oxiranes: end-group modification of RAFT prepared
homopolymers. Polym. Chem. 2011, 2, 1347−1354.
(16) Segui, F.; Qiu, X.-P.; Winnik, F. M. An Efficient Synthesis of
Telechelic Poly(N-isopropylacrylamides) and its Application to the
I
DOI: 10.1021/acs.macromol.6b01512
Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
Preparation of α,ω-Dicholesteryl and α,ω-Dipyrenyl Polymers. J.
Polym. Sci., Part A: Polym. Chem. 2008, 46, 314−326.
(17) York, A. W.; Scales, C. W.; Huang, F.; McCormick, C. L. Facile
Synthetic Procedure for ω, Primary Amine Functionalization Directly
in Water for Subsequent Fluorescent Labeling and Potential
Bioconjugation of RAFT-Synthesized (Co)Polymers. Biomacromole-
cules 2007, 8, 2337−2341.
(18) Roth, P. J.; Kessler, D.; Zentel, R.; Theato, P. A Method for
Obtaining Defined End Groups of Polymethacrylates Prepared by the
RAFT Process during Aminolysis. Macromolecules 2008, 41, 8316−
8319.
(19) Qiu, X. P.; Winnik, F. M. Facile and Efficient One-Pot
Transformation of RAFT Polymer End Groups via a Mild Aminolysis/
Michael Addition Sequence. Macromol. Rapid Commun. 2006, 27,
1648−1653.
(20) Spruell, J. M.; Levy, B. A.; Sutherland, A.; Dichtel, W. R.; Cheng,
J. Y.; Stoddart, F.; Nelson, A. Facile Postpolymerization End-
Modification of RAFT Polymers. J. Polym. Sci., Part A: Polym. Chem.
2008, 47, 346−356.
(21) Boyer, C.; Granville, A.; Davis, T. P.; Bulmus, V. Modification of
RAFT-Polymers via Thiol-Ene Reactions: A General Route to
Functional Polymers and New Architectures. J. Polym. Sci., Part A:
Polym. Chem. 2009, 47, 3773−3794.
(22) Boyer, C.; Bulmus, V.; Davis, T. P. Efficient Usage of
Thiocarbonates for Both the Production and Biofunctionalization of
Polymers. Macromol. Rapid Commun. 2009, 30, 493−497.
(23) Lowe, A. B. Thiol-ene “click” reactions and recent applications
in polymer and materials synthesis. Polym. Chem. 2010, 1, 17−36.
(24) Lowe, A. B. Thiol-ene “click” reactions and recent applications
in polymer and materials synthesis: A first update. Polym. Chem. 2014,
5, 4820−4870.
(25) Chan, J. W.; Hoyle, C. E.; Lowe, A. B.; Bowman, M.
Nucleophile-Initiated Thiol-Michael Reactions: Effect of Organo-
catalyst, Thiol, and Ene. Macromolecules 2010, 43, 6381−6388.
(26) Scales, C. W.; Convertine, A. J.; McCormick, C. L. Fluorescent
Labeling of RAFT-Generated Poly(N-isopropylacrylamide) via a
Facile Maleimide-Thiol Coupling Reaction. Biomacromolecules 2006,
7, 1389−1392.
(27) Li, M.; De, P.; Gondi, S. R.; Sumerlin, B. S. End Group
Transformations of RAFT-Generated Polymers with Bismaleimides:
Functional Telechelics and Modular Block Copolymers. J. Polym. Sci.,
Part A: Polym. Chem. 2008, 46, 5093−5100.
(28) Li, M.; De, P.; Li, H.; Sumerlin, B. S. Conjugation of RAFT-
generated polymers to proteins by two consecutive thiol-ene reactions.
Polym. Chem. 2010, 1, 854−859.
(29) Huang, X.; Boyer, C.; Davis, T. P.; Bulmus, V. Synthesis of
heterotelechelic polymers with affinity to glutathione-S-transferase and
biotin-tagged proteins by RAFT polymerization and thiol-ene
reactions. Polym. Chem. 2011, 2, 1505−1512.
(30) Delaittre, G.; Pauloehrl, T.; Bastmeyer, M.; Barner-Kowollik, C.
Acrylamide-Based Copolymers Bearing Photoreleasable Thiols for
Subsequent Thiol-Ene Functionalization. Macromolecules 2012, 45,
1792−1802.
(31) Mather, B. D.; Viswanathan, K.; Miller, K. M.; Long, T. E.
Michael addition reactions in macromolecular design for emerging
technologies. Prog. Polym. Sci. 2006, 31, 487−531.
of a Strong Nucleophile to Generate a Strong Base. J. Am. Chem. Soc.
2003, 125, 8696−8697.
(36) Xi, W.; Wang, C.; Kloxin, C. J.; Bowman, C. N. Nitrogen-
Centered Nucleophile Catalyzed Thiol-Vinylsulfone Addition, Anoth-
er Thiol-ene “Click” Reaction. ACS Macro Lett. 2012, 1, 811−814.
(37) Taylor, J. E.; Bull, S. D.; Williams, J. M. J. Amidines,
isothioureas, and guanidines as nucleophilic catalysts. Chem. Soc. Rev.
2012, 41, 2109−2121.
(38) Northrop, B. H.; Frayne, S. H.; Choudhary, U. Thiol-maleimide
“click” chemistry: evaluating the influence of solvent, initiator, and
thiol on the reaction mechanism, kinetics, and selectivity. Polym. Chem.
2015, 6, 3415−3430.
(39) Thang, S. H.; Chong, B. Y. K.; Mayadunne, R. T. A.; Moad, G.;
Rizzardo, E. A Novel Synthesis of Functional Dithioesters,
Dithiocarbamates, Xanthates and Trithiocarbonates. Tetrahedron Lett.
1999, 40, 2435−2438.
(40) Azechi, M.; Toyota, N.; Yamabuki, K.; Onimura, K.; Oishi, T.
Anionic polymerization of N-substituted maleimide with achiral and
chiral amines as an initiator. Polym. Bull. 2011, 67, 631−640.
(41) Jaacks, V.; Franzmann, G. Formation of Macrozwitterions in
Anionic Vinyl Polymerization Initiated by Tertiary Phosphines.
Makromol. Chem. 1971, 143, 283−288.
(42) Pepper, D. C. Kinetics and Mechanisms of Zwitterionic
Polymerizations of Alkyl Cyanoacrylates. Polym. J. 1980, 12, 629−637.
(43) Sahu, U. S.; Bhadani, S. Homopolymerization of Maleimide.
Makromol. Chem., Rapid Commun. 1982, 3, 103−107.
(44) Brown, H. A.; De Crisci, A. G.; Hedrick, J. L.; Waymouth, R. M.
Amidine-Mediated Zwitterionic Polymerization of Lactide. ACS Macro
Lett. 2012, 1, 1113−1115.
(45) Ho, H. T.; Levere, M. E.; Pascual, S.; Montembault, V.; Soutif,
J.-C.; Fontaine, L. Phosphites as Alternative Coreagents for the One-
Pot Aminolysis/Thiol-Ene Synthesis of Maleimide-Functionalized
RAFT Polymers. J. Polym. Sci., Part A: Polym. Chem. 2012, 50,
1657−1661.
(46) Nordqvist, A.; Bjorkelid, C.; Andaloussi, M.; Jansson, A. M.;
̈
́
Mowbray, S. L.; Karlen, A.; Larhed, M. Synthesis of Functionalized
Cinnamaldehyde Derivatives by an Oxidative Heck Reaction and Their
Use as Starting Materials for Preparation of Mycobacterium
tuberculosis 1-Deoxy-D-xylulose-5-phosphate Reductoisomerase In-
hibitors. J. Org. Chem. 2011, 76, 8986−8998.
(47) Kim, S. H.; Kim, S. H.; Kim, H. J.; Kim, J. N. An Efficient
Conjugate Addition of Dialkyl Phosphite to Electron-Deficient
Olefins: The Use of a Nucleophilic Organocatalyst to Form a Strong
Base. Bull. Korean Chem. Soc. 2013, 34, 989−992.
(48) Humphrey, R. E.; Oleson, C. L.; Matula, G. M.; Vaught, A. C.
Reduction of Disulfides in the Milligram Range with Hypophos-
phorous Acid and Organic Phosphites. Microchem. J. 1971, 16, 429−
436.
(49) Radkiewicz, J. L.; Zipse, H.; Clarke, S.; Houk, K. N. Accelerated
Racemization of Aspartic Acid and Asparagine Residues via
Succinimide Intermediates: An ab Initio Theoretical Exploration of
Mechanism. J. Am. Chem. Soc. 1996, 118, 9148−9155.
(32) Li, G.-Z.; Randev, R. K.; Soeriyadi, A. H.; Rees, G.; Boyer, C.;
Tong, Z.; Davis, T. P.; Becer, C. R.; Haddleton, D. M. Investigation
into thiol-(meth)acrylate Michael addition reactions using amine and
phosphine catalysts. Polym. Chem. 2010, 1, 1196.
(33) Gimbert, C.; Lumbierres, M.; Marchi, C.; Moreno-Manas, M.;
̃
́
Sebastian, R. M.; Vallribera, A. Michael additions catalyzed by
phosphines. An overlooked synthetic method. Tetrahedron 2005, 61,
8598−8605.
(34) Methot, J. L.; Roush, W. R. Nucleophilic Phosphine
Organocatalysis. Adv. Synth. Catal. 2004, 346, 1035−1050.
(35) Stewart, I. C.; Bergman, R. G.; Toste, F. D. Phosphine-
Catalyzed Hydration and Hydroalkoxylation of Activated Olefins: Use
J
DOI: 10.1021/acs.macromol.6b01512
Macromolecules XXXX, XXX, XXX−XXX