Organocatalytic conjugate-addition polymerization of linear and cyclic acrylic monomers by N-heterocyclic carbenes: Mechanisms of chain initiation, propagation, and termination

Article abstract of DOI:10.1021/ja4088677

This contribution presents a full account of experimental and theoretical/computational investigations into the mechanisms of chain initiation, propagation, and termination of the recently discovered N-heterocyclic carbene (NHC)-mediated organocatalytic conjugate-addition polymerization of acrylic monomers. The current study specifically focuses on three commonly used NHCs of vastly different nucleophilicity, 1,3-di-tert-butylimidazolin-2-ylidene (I^tBu), 1,3- dimesitylimidazolin-2-ylidene (IMes), and 1,3,4-triphenyl-4,5-dihydro-1H-1,2,4- triazol-5-ylidene (TPT), and two representative acrylic monomers, the linear methyl methacrylate (MMA) and its cyclic analog, biomass-derived renewable γ-methyl-α-methylene-γ-butyrolactone (MMBL). For MMA, there exhibits an exquisite selectivity of the NHC structure for the three types of reactions it promotes: enamine formation (single-monomer addition) by IMes, dimerization (tail-to-tail) by TPT, and polymerization by I^tBu. For MMBL, all three NHCs promote no dimerization but polymerization, with the polymerization activity being highly sensitive to the NHC structure and the solvent polarity. Thus, I^tBu is the most active catalyst of the series and converts quantitatively 1000-3000 equiv of MMBL in 1 min or 10 000 equiv in 5 min at room temperature to MMBL-based bioplastics with a narrow range of molecular weights of M_n = 70-85 kg/mol, regardless of the [MMBL]/[I^tBu] ratio employed. The I^tBu-catalyzed MMBL polymerization reaches an exceptionally high turnover frequency up to 122 s ^-1 and a high initiator efficiency value up to 1600%. Unique chain-termination mechanisms have been revealed, accounting for the production of relative high-molecular-weight linear polymers and the catalytic nature of this NHC-mediated conjugate-addition polymerization. Computational studies have provided mechanistic insights into reactivity and selectivity between two competing pathways for each NHC-monomer zwitterionic adduct, namely enamine formation/dimerization through proton transfer vs polymerization through conjugate addition, and mapped out extensive energy profiles for chain initiation, propagation, and termination steps, thereby satisfactorily explaining the experimental observations.

Full text of DOI:10.1021/ja4088677

Article
pubs.acs.org/JACS
Organocatalytic Conjugate-Addition Polymerization of Linear and
Cyclic Acrylic Monomers by N‑Heterocyclic Carbenes: Mechanisms of
Chain Initiation, Propagation, and Termination
Yuetao Zhang,† Meghan Schmitt,† Laura Falivene,‡ Lucia Caporaso,‡ Luigi Cavallo,*,§
and Eugene Y.-X. Chen*,†
†Department of Chemistry, Colorado State University, Fort Collins, Colorado 80523-1872, United States
‡Dipartimento di Chimica e Biologia, Universita
̀
di Salerno, I-84084, Fisciano, Italy
§Physical Sciences and Engineering Division, Kaust Catalysis Center, King Abdullah University of Science and Technology (KAUST),
Thuwal 23955-6900, Saudi Arabia
S
* Supporting Information
ABSTRACT: This contribution presents a full account of
experimental and theoretical/computational investigations into
the mechanisms of chain initiation, propagation, and termination
of the recently discovered N-heterocyclic carbene (NHC)-
mediated organocatalytic conjugate-addition polymerization of
acrylic monomers. The current study specifically focuses on three
commonly used NHCs of vastly different nucleophilicity, 1,3-di-
tert-butylimidazolin-2-ylidene (ItBu), 1,3-dimesitylimidazolin-2-yli-
dene (IMes), and 1,3,4-triphenyl-4,5-dihydro-1H-1,2,4-triazol-5-
ylidene (TPT), and two representative acrylic monomers, the
linear methyl methacrylate (MMA) and its cyclic analog, biomass-
derived renewable γ-methyl-α-methylene-γ-butyrolactone (MMBL). For MMA, there exhibits an exquisite selectivity of the NHC
structure for the three types of reactions it promotes: enamine formation (single-monomer addition) by IMes, dimerization (tail-
to-tail) by TPT, and polymerization by ItBu. For MMBL, all three NHCs promote no dimerization but polymerization, with the
polymerization activity being highly sensitive to the NHC structure and the solvent polarity. Thus, ItBu is the most active catalyst
of the series and converts quantitatively 1000−3000 equiv of MMBL in 1 min or 10 000 equiv in 5 min at room temperature to
MMBL-based bioplastics with a narrow range of molecular weights of Mn = 70−85 kg/mol, regardless of the [MMBL]/[ItBu]
ratio employed. The ItBu-catalyzed MMBL polymerization reaches an exceptionally high turnover frequency up to 122 s−1 and a
high initiator efficiency value up to 1600%. Unique chain-termination mechanisms have been revealed, accounting for the
production of relative high-molecular-weight linear polymers and the catalytic nature of this NHC-mediated conjugate-addition
polymerization. Computational studies have provided mechanistic insights into reactivity and selectivity between two competing
pathways for each NHC-monomer zwitterionic adduct, namely enamine formation/dimerization through proton transfer vs
polymerization through conjugate addition, and mapped out extensive energy profiles for chain initiation, propagation, and
termination steps, thereby satisfactorily explaining the experimental observations.
As an important class of organic catalysts, N-heterocyclic
carbenes (NHCs) have attracted increasing interest due to their
unique reactivity and selectivity observed in many different
types of organic reactions.2 Thanks to the pioneering work of
Hedrick, Waymouth, and their co-workers,3 the utility of the
NHC-mediated reactions has been expanded to polymer
synthesis,4 via predominantly the ring-opening polymerization
(ROP) of heterocyclic monomers, such as lactides,5 lactones,6
epoxides,7 cyclic carbonates,8 cyclic siloxanes,9 and N-carboxyl-
anhydrides.10 NHC-mediated step-growth polymerization has
been reported as well.11 Polymerization of α,β-unsaturated
esters (or acrylic monomers) such as methyl methacrylate
INTRODUCTION
■
Organocatalysis1 using small-molecule organic compounds as
catalysts has risen to prominence over the past decade in
organic synthesis, polymer synthesis, and other areas, thanks to
several advantages it can offer relative to other modes of
catalysis. First, many small-molecule organic catalysts are
commercially available, inexpensive, and air/moisture stable.
Second, small-molecule organocatalysts are readily available
from renewable resources and relatively nontoxic and thus
“greener” than other types of catalysts. Third, asymmetric
catalysis can be achieved by chiral organic reagents, many of
which are naturally available from biological sources as single
enantiomers. Overall, organocatalysis is especially advantageous
when metal-free products or processes are of primary concern.
Received: September 3, 2013
Published: November 18, 2013
© 2013 American Chemical Society
17925
dx.doi.org/10.1021/ja4088677 | J. Am. Chem. Soc. 2013, 135, 17925−17942

Journal of the American Chemical Society
Article
Figure 1. Outlined three distinctive reaction pathways involved in the reaction of NHCs and acrylic monomers.
(MMA) has also been recently realized through the classic
group-transfer polymerization (GTP) initiated by a silyl ketene
acetal (SKA),12 using NHCs as alternative nucleophilic catalysts
to the commonly used fluoride and oxygenated anionic
catalysts for activating the SKA initiator.13 In addition, such
acrylic monomers can be rapidly polymerized by frustrated
Lewis pairs (FLPs)14 consisting of bulky NHC bases, such as
the Arduengo carbenes 1,3-di-tert-butylimidazolin-2-ylidene
(ItBu) and 1,3-dimesitylimidazolin-2-ylidene (IMes),15 and
the strongly acidic, sterically encumbered pentafluorophenyl
alane Al(C6F5)3, via the proposed zwitterionic imidazolium
enolaluminate intermediates.16 Using such Arduengo NHCs
alone, no MMA conversion was observed in either toluene16 or
THF.13c On the other hand, Glorius and Matsuoka discovered
that the Enders triazolylidene carbene TPT (1,3,4-triphenyl-
4,5-dihydro-1H-1,2,4-triazol-5-ylidene),17 which was estimated
to be 103 times less nucleophilic than the imidazolylidene
carbene (IMes),18 catalyzes tail-to-tail dimerization (umpolung)
of MMA and other methacrylate substrates, while the common
imidazolylidene carbenes are ineffective.19
It is clear from the above overview, although the NHC-
mediated ROP of heterocyclic monomers has been highly
successful, the NHC-mediated conjugate-addition polymer-
ization of conjugated polar alkenes, such as acrylic monomers,
still required the use of a nucleophilic initiator (SKA) in the
case of GTP13 or a Lewis acid catalyst [Al(C6F5)3] in the case
of the FLP polymerization.16 When used alone (i.e., without
combining with either the SKA initiator or the alane catalyst),
NHCs such as ItBu and IMes in toluene16c or (1,3-di-isopropyl-
4,5-dimethylimidazolin-2-ylidene (IiPrMe2) in THF13c effected
no monomer conversion in the MMA polymerization at
ambient temperature, although the polymerization of tert-butyl
acrylate by IiPrMe2 achieved 25% monomer conversion after 16
h.13c We20 recently discovered that an NHC alone initiates
extremely rapid conjugate-addition polymerization of cyclic
acrylic monomersrenewable methylene butyrolactones,21
including the naturally occurring α-methylene-γ-butyrolactone
(MBL)22 and the biomass-derived γ-methyl-α-methylene-γ-
butyrolactone (MMBL).23 The polymerization achieves
quantitative monomer conversion in 1 min with a low ItBu
loading at room temperature (RT), affording medium- or high-
molecular-weight (MW) polymers. The rate of the polymer-
ization is strongly affected by the relative nucleophilicity of the
NHC catalyst employed, with the most nucleophilic ItBu in the
series exhibiting the highest activity, the less nucleophilic IMes
displaying a much lower activity, and the least nucleophilic TPT
often showing no activity (in DMF) at all. Intriguingly, there
exists a remarkable selectivity of the NHC for the substrate
structure, thus leading to three different modes of reaction
involving acrylic substrates (Figure 1):20 TPT promotes
dimerization of methacrylates such as MMA (pathway A);
IMes selectively forms the single-addition productthe
enamine or the deoxy-Breslow intermediate19,20,24,25with
methacrylates such as MMA (pathway B); and ItBu mediates
rapid polymerization of methylene butyrolactones such as
MMBL (pathway C). That preliminary study generated many
fascinating results on the NHC-mediated polymerization of
acrylic monomers, but at the same time it brought forth four
unaddressed important, fundamental questions: (a) What are
the underlying reasons for the observed exquisite selectivity of
the NHC structure for the three types of reactions it promotes:
enamine formation, dimerization, and polymerization? (b)
What are the mechanisms of chain initiation, propagation, and
termination or transfer? (c) Is it a catalytic polymerization in
the absence or presence of a suitable chain-transfer agent
(CTA)? And (d) what is the structure of the resulting polymer,
linear or cyclic? Accordingly, the central objective of this study
was to address these important questions through combined
experimental and theoretical/computational investigations, the
results of which are presented herein.
RESULTS
■
Reactivity of NHCs toward MMA and MMBL. For MMA
polymerization, NHCs such as ItBu and IMes in toluene16c or
IiPrMe2 in THF13c effected no monomer conversion at ambient
temperature. However, depending on the NHC structure and
the reaction medium, three types of interesting reactivity have
been observed. Specifically, TPT exhibits no or negligible
activity for polymerization of MMA in toluene or DMF, at RT
or 80 °C, up to 24 h.20 However, it catalyzes efficient tail-to-tail
dimerization of MMA in toluene or polar solvents (e.g., DME
or 1,4-dioxane) at 80 °C, affording dimethyl-2,5-dimethylhex-2-
enedioate (E/Z = 95/5) in good yields (cf., pathway A, Figure
17926
dx.doi.org/10.1021/ja4088677 | J. Am. Chem. Soc. 2013, 135, 17925−17942

Journal of the American Chemical Society
Article
Figure 2. Formation of single-addition product enamine I (the deoxy-Breslow intermediate) from the reaction of MMA + IMes and its X-ray crystal
structure. Hydrogen atoms were omitted for clarity and ellipsoids drawn at 50% probability. Selected bond lengths [Å]: C1−C4 = 1.353(2), C2−C3
= 1.316(2), C4−C5 = 1.508(2).
a
Table 1. Selected MMBL Polymerization Results by ItBu in DMF at 25 °C
b
c
c
d
run no.
[ItBu] (mmol/L)
[MMBL]0/ [ItBu]0
time (s)
conv. (%)
Mn (kg/mol)
PDI (Mw/Mn)
I* (%)
1
4.68
2.34
200
400
60
60
100
100
38.5
69.0
84.7
71.9
88.7
81.2
52.6
67.1
72.9
76.3
78.1
77.7
79.0
78.8
79.9
44.3
64.6
73.4
76.8
79.5
78.8
81.0
78.6
79.3
45.9
53.1
53.5
55.7
59.8
64.0
65.7
69.1
69.3
71.3
70.9
74.1
72.6
72.9
74.2
1.68
1.93
2.11
1.80
1.45
1.48
2.10
2.00
2.00
1.99
1.97
1.98
1.97
1.97
1.93
1.98
2.03
1.96
1.96
1.92
1.94
1.92
1.98
1.96
2.33
2.07
1.98
2.00
2.05
1.97
2.00
2.00
2.01
1.96
1.99
1.95
1.96
1.96
1.95
58.1
64.9
106
2
3
1.17
800
60
100
4
0.936
0.312
0.187
0.117
1000
3000
5000
8000
60
>99
>99
>99
33.4
58.3
69.6
81.8
88.6
93.5
95.5
97.2
99.0
18.3
55.2
69.4
78.2
81.7
86.3
94.6
97.3
98.5
19.5
30.8
33.8
36.0
47.2
50.1
56.8
61.5
68.9
71.1
75.1
79.3
81.5
85.3
88.3
154
5
60
375
6
120
15
683
7
569
8
30
778
9
45
855
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
39
30
31
32
33
34
35
36
37
38
39
60
961
80
1016
1078
1083
1105
1110
463
100
120
150
180
20
0.0936
10 000
40
957
60
1059
1140
1151
1226
1308
1386
1391
571
80
100
120
180
250
300
20
0.0780
12 000
40
780
60
849
80
869
100
120
150
180
210
240
300
360
420
600
900
1061
1052
1162
1196
1336
1340
1424
1438
1509
1573
1599
a
b
Carried out at ambient temperature (∼25 °C) in 4.5 mL DMF and 0.5 mL MMBL solution with fixed [MMBL]0 = 0.936 M. Monomer
c
1
conversions measured by H NMR. Mn and PDI determined by gel-permeation chromatography (GPC) in DMF relative to PMMA standards. A
d
small (0.1−2%) high-molecular-weight peak (>106 g/mol), attributable to the polymer aggregation, was present in most of the samples. Initiator
efficiency (I*) = Mn (calcd)/Mn (exptl), where Mn (calcd) = MW(M) × [M]/[I] × conversion (%) + MW (chain-end groups).
1), while the common imidazolylidene carbenes are ineffec-
tive.19 This intermolecular umpolung of MMA was proposed to
proceed through the initial conjugate addition of TPT to MMA
to form the corresponding ester enolate that undergoes proton
transfer (or tautomerization) affording an enamine intermedi-
ate, the deoxy-Breslow intermediate analogous to the Breslow
17927
dx.doi.org/10.1021/ja4088677 | J. Am. Chem. Soc. 2013, 135, 17925−17942

Journal of the American Chemical Society
Article
intermediate involved in the benzoin reaction;26 addition of the
enamine to another molecule of MMA leads to the tail-to-tail
dimerization product, via again the ester enolate intermediate
that undergoes proton transfer and release of TPT.19 This
mechanism bears a close resemblance to the one proposed for
the intramolecular umpolung of Michael acceptors catalyzed by
NHCs.27 We found that, despite the fact that ItBu did not react
with MMA in toluene at RT (up to 3 days), IMes reacted
smoothly with MMA at RT to form a single-addition product,
enamine, or deoxy-Breslow intermediate I (Figure 2). Unlike
the TPT-derived enamine intermediate involved in the
dimerization of MMA, the enamine derived from IMes is
stable, isolable, and nonreactive toward additional MMA,
presumably due to its strong, irreversible binding of IMes
with MMA and high-energy barrier for conjugate addition to
another MMA molecule (cf., pathway B, Figure 1). Noteworthy
is that the reaction of TPT and MMA at RT also forms the
enamine as two isomers (enamine II, see the Supporting
Information for details); but unlike the IMes-based enamine I,
upon heating to 80 °C enamine II further reacts with MMA to
form the dimer in a catalytic fashion. The molecular structure of
enamine I (Figure 2) clearly displays a CC double bond
formed between C1 and C4 carbons, with a bond distance of
1.353(2) Å. This single-addition product resembles those
enamines derived from the 1:1 addition of TPT to more highly
(doubly) activated Michael acceptors.17,28 Although the
structurally characterized aza-Breslow intermediate29 and the
methylated enaminol Breslow intermediate30 have been
reported previously, enamine I20 and other analogous
intermediates derived from NHC31 represent the structurally
characterized deoxy-Breslow intermediate.
attributable to the presence of both the nearly planar five-
membered lactone ring (which provides resonance stabilization
for the active species) and the higher energy exocyclic CC
double bond (as a result of the ring strain and the fixed s-cis
conformation).21 As a result, MMBL behaves rather differently
toward NHCs than MMA. First, TPT (0.5 mol %) exhibited no
activity toward MMBL in either THF or DMF at RT up to 24
h, but in toluene it showed some polymerization activity,
achieving 79% monomer conversion in 24 h (TOF = 7 h−1)
and producing PMMBL with Mn = 38.9 kg/mol. Second, the
more nucleophilic IMes is more active in MMBL polymer-
ization; for instance, the MMBL polymerization by IMes (0.5
mol %) at RT achieved 98% and 100% monomer conversion
after 24 h, affording PMMBL with Mn = 51.2 and 17.8 kg/mol,
in toluene and DMF, respectively. IMes also polymerizes
MMBL at 80 °C in toluene with no enamine formation. Third,
the most nucleophilic NHC of the series, ItBu, is the most
active catalyst. In toluene, the MMBL polymerization by ItBu
(0.5 mol %) achieved 93% conversion in only 30 min, affording
the relatively high-molecular-weight PMMBL (Mn = 58.4 kg/
mol). Switching to DMF, the polymerization was complete in
less than 1 min. In fact, quantitative monomer conversion was
achieved in <1 min with a [MMBL]/[ItBu] ratio as high as
3000 (i.e., 0.03 mol % ItBu), giving rise to a very high TOF of
>1.8 × 105 h−1 (51 s−1).
Table 1 summarizes selected MMBL polymerization results
by ItBu in DMF at RT with varied [MMBL]/[ItBu] ratios from
200 to 12 000. As in the polymerization of MMBL mediated by
metal- and metalloid-based catalysts or initiators,21a−h no ring-
opening of the butyrolactone ring was observed; control runs
using NHCs or NHC/Lewis acid pairs for the ROP of five-
membered lactones, γ-butyrolactone and γ-valerolactone, also
showed no polymerization activity. Overall, these results
established ItBu as the best and most active catalyst of the
current NHC series for rapid polymerization of acrylic
monomers, especially the biorenewable MMBL (cf., pathway
C, Figure 1).
Intriguingly, although ItBu exhibits no reactivity toward
MMA in toluene or THF (neither enamine formation nor
dimerization) and neither TPT nor IMes polymerizes MMA at
RT, 0 °C, or −40 °C in toluene or DMF, ItBu polymerizes
MMA in DMF at RT with an appreciable activity. With a 0.5
mol % loading of ItBu, the MMA polymerization in DMF (8.0
mL) at RT achieved 68% conversion in 1 h, corresponding to a
good turnover frequency (TOF) of 136 h−1. The PMMA
produced is a syndio-rich atactic material, with syndiotacticity
(rr) = 55.7%, heterotacticity (mr) = 39.6%, and isotacticity
(mm) = 4.7%, typical tacticity values of PMMA produced by
anionic polymerization in polar donor solvents at RT.32 The
number-average molecular weight (Mn) of the PMMA was 33.2
kg/mol (at 87% monomer conversion), with a polydispersity
index (PDI) of Mw/Mn = 1.99. As this measured Mn was much
higher than the calculated Mn (17.4 kg/mol) according to the
[MMA]0/[ItBu]0 ratio employed, the resulting initiator
efficiency (I* = Mn (calcd)/Mn (exptl), where Mn (calcd) =
MW(M) × [M]/[I] × conversion (%) + MW (chain-end
groups)) was only 52%. We also carried out a series of MMA
polymerizations at RT with a fixed monomer concentration of
[MMA]o = 0.936 (in 4.5 mL DMF) but varying the [MMA]0/
[ItBu]0 ratio from 200, 400, 800, to 1000, and found that the
resulting initiator efficiencies for these runs, calculated on the
basis of the measured monomer conversion and polymer Mn
values, were in the range of 12−30%. Overall, these results
Characteristics and Kinetics of MMBL Polymerization
by ItBu. Having established ItBu as the best catalyst system in
all aspects for the MMBL polymerization in DMF, we
subsequently examined this polymerization in more detail,
specifically concerning its degree of control and kinetics that
varied the [MMBL]/[ItBu] ratios from 200 to 15 000 (Table
1). As can be seen from this table, the MMBL polymerization
by ItBu is exceedingly rapid, achieving quantitative monomer
conversion in <1 min with the [MMBL]/[ItBu] ratio as high as
3000 (runs 1−5). At higher [MMBL]/[ItBu] ratios, the
polymerization can still achieve quantitative or near quantitative
monomer conversion at progressively longer reaction times:
5000 in 2 min (run 6), 8000 (99%) in 3 min (run 15), and 10
000 (99%) in 5 min (run 24). At an even higher [MMBL]/
[ItBu] ratio of 12 000, the polymerization achieved 88%
conversion in 15 min (run 39). The TOF for the polymer-
ization of MMBL by 0.01 mol % ItBu, calculated from the
maximum slope of the conversion vs time plot (Figure 3), was
high, 4.4 × 105 h−1 (122 s−1). This high rate of polymerization
is compared to a kapp of 26.5 s−1 for the MMA polymerization
by anionic polymerization using tetrakis[tris(dimethylamino)-
t
indicate that the MMA polymerization by Bu in DMF has a
typical initiator efficiency value of only ∼50% or lower, thus a
noncatalytic process.
The bioderived MMBL can be described as the cyclic analog
of MMA; however, it exhibits greater reactivity in chain-growth
polymerization than typical alkyl methacrylates such as MMA,
+
phosphoranylidenamino]phosphonium (P5 ) as the counterion
in THF33 and to a TOF value of 26 h−1 for a typical anionic
polymerization of MMBL by BuLi in THF ([MMBL]/[BuLi] =
60, 2 h, 86% conversion).34
17928
dx.doi.org/10.1021/ja4088677 | J. Am. Chem. Soc. 2013, 135, 17925−17942

Journal of the American Chemical Society
Article
Figure 4. Plots of Mn and PDI of PMMBL vs MMBL conversion in
DMF at 25 °C: [MMBL]0 = 0.935 M, [ItBu] = 0.0936 mM with a
[MMBL]0/[ItBu]0 ratio of 10 000/1.
Figure 3. MMBL conversion (%) vs time (s) plot. Conditions:
[MMBL]/[ItBu] = 10 000, DMF, 25 °C.
Another interesting feature of this polymerization, observed
through analysis of the resulting polymer molecular weights vs
the calculated values, is large variations of the initiator efficiency
I* as a function of the [MMBL]/[ItBu] ratio. When this ratio
was ≤400, the calculated I* was lower than 100% (58−65%),
suggesting a slower rate of initiation relative to the rate of
propagation and/or decomposition of the initiator or the active
species. An experiment was then designed to determine which
scenario is applicable to the present system. Specifically, a small
scale MMBL polymerization by ItBu was carried out in toluene-
d8. The resulting polymer was removed by filtration after all the
monomer had been converted, and the filtrate was analyzed by
1H NMR. Most of ItBu still remained in the filtrate, implying
that only a small amount of ItBu was involved in the initiation
step (which is still attached to the polymer chain) due to the
lower initiation rate, relative to the rate of propagation. When
the [MMBL]/[ItBu] ratio was 800, the measured Mn of 84.7
kg/mol matched almost perfectly with the calculated Mn, thus
affording a near quantitative I* value of 106% (run 3). A further
increase in the [MMBL]/[ItBu] ratio to 1000, 3000, and 5000
resulted in a steady increase in the I* value to 154% (run 4),
375% (run 5), and 683% (run 6), respectively. The I*
continued to go up to 1110% (run 15) in the 8000 ratio and
1391% (run 24) in the 10 000 ratio and reached the highest
1600% (run 39) in the 12 000 ratio. These results indicate
substantial internal chain transfer occurred, even in the absence
of any added suitable CTAs, thus producing multiple polymer
chains (up to 16 chains) per initiator molecule, indicative of a
catalytic polymerization system through chain transfer to
monomer.
Figure 4 depicts the profile of Mn of PMMBL vs MMBL
conversion in a high [MMBL]/[ItBu] ratio of 10 000. A linear
increase of Mn with MMBL conversion was observed at early
and mid stages of the polymerization up to 80% conversion.
However, at later stages of polymerization the polymer MW
remained nearly constant (79−81 kg/mol, runs 20−24, Table
1) with monomer conversion up to completion, indicative of
chain termination or transfer (i.e., chain termination followed
by chain reinitiation) effectively competing with chain
propagation. The MW distribution corresponds to what is
expected for chain growth with chain transfer, that is Mw/Mn ≈
2.0.35 Comparing the Mn values of the final isolated polymers
with different [MMBL]/[ItBu] ratios, we found the Mn
increased with the increasing of the [MMBL]/[ItBu] ratio
initially and then remained nearly constant when this ratio was
higher than 800 (Figure 5). Possible physical reasons due to
Figure 5. Plot of Mn of PMMBL vs [MMBL]/[ItBu] ratio.
viscosity/solubility issues are unlikely in this case as the
polymer at this MW range is very soluble in DMF.
Furthermore, carrying out the 1000:1 ratio polymerization in
more dilute conditions (using 9.5 mL DMF) produced
PMMBL with Mn = 67.8 kDa (PDI = 1.96), which is rather
similar to that produced under current standard conditions (4.5
mL DMF). However, switching the solvent to toluene enabled
the polymerization to produce PMMBL with a much higher
MW of Mn = 107 kDa (PDI = 2.99) at 93% conversion after 4
h; the polymerization was much slower, and the resulting
polymer crashed out of the solution, but this much higher MW
yielded a nearly quantitative initiator efficiency of I* = 97%
(thus no chain transfer). The above observations, together with
the high I* values, demonstrate that this polymerization in
DMF is an effective catalytic polymerization system through
internal chain transfer when the [MMBL]/[ItBu] ratio is higher
than 800. This scenario can be examined by sequential addition
of monomer. Specifically, two chain-extension polymerizations
with a [MMBL]/[ItBu] ratio of 1000 were carried out to verify
the catalytic property of the polymerization. Upon complete
conversion of the initial monomer pool, an additional 1000 and
500 equiv of the monomer was added to a total monomer
17929
dx.doi.org/10.1021/ja4088677 | J. Am. Chem. Soc. 2013, 135, 17925−17942

Journal of the American Chemical Society
Article
concentration of 1.87 and 1.40 M, respectively. The reaction
was allowed to completion. GPC analysis of the resulting
polymers showed that both final polymers exhibited unimodal
MW distribution, and the Mn of the polymers were identical
(Figure 6), even though the total combined [MMBL]/[ItBu]
ratios were different (i.e., 2000 vs 1500).
Figure 8. Plot of ln(kapp) vs ln[ItBu] for the MMBL polymerization by
ItBu in DMF at 25 °C.
(0.02 mol % ItBu), 8000 (0.0125 mol % ItBu), and 10 000 (0.01
mol % ItBu), as a function of ln[ItBu]0, was fit to a straight line
(R2 = 0.999) with a slope of 2.099. Thus, the kinetic order with
respect to [ItBu], given by the slope of ∼2, reveals that the
overall polymerization is second order in ItBu concentration.
The mechanistic scenario revealed by our computational
studies (vide infra) indicates the slowest step of the ItBu-
mediated MMBL polymerization process is the formation of
the ItBu-MMBL adduct (i.e., chain initiation to generate the
zwitterionic active propagating specie), thus in agreement with
the observed first-order dependence on monomer concen-
tration. However, it is not obvious why there was a second-
order dependence on [ItBu], but factors contributing to this
experimental observation could include inefficient chain
initiation or reinitiation by ItBu and, especially under such
low catalyst loading conditions, catalyst decomposition and
poisoning by the adventitious impurities brought into the
system from the solvent and/or monomer.
Chain Termination and End Groups. A possible
mechanistic scenario for the conjugate-addition polymerization
of acrylic monomers by NHCs was proposed to proceed
through reiterative 1,4-conjugate addition of the propagating
ester enolate to the incoming monomer (cf., pathway C, Figure
1),20 following the well-established conjugate-addition mecha-
nism for the controlled anionic polymerization of acrylics using
discrete or in situ generated metal ester enolates.32 Determi-
nation of the polymer chain structure, including chain-end
groups, typically provides critical insight into chain termination
and/or transfer mechanisms. To this end, analysis of low-
molecular-weight MMBL oligomers produced with NHCs
(TPT or ItBu) in toluene by MALDI-TOF MS showed that
neither initiating nor terminating end-groups were present in
the major fraction of the oligomers (Figure 9 by TPT; Figure
S7 by ItBu); this result may indicate a possible cyclic chain
structure due to the absence of any end groups seen by the MS
method, and the fact that it is known that the NHC-initiated
zwitterionic ROP of lactide produces cyclic poly(lactide).4,5
The origin of such structure for the current conjugate-addition
polymers could be explained by a mode of chain termination
via backbiting of the growing ester enolate to the electrophilic
carbon directly bonded to the NHC, accompanied by
elimination of the NHC (Scheme 1).
Figure 6. Overlay GPC traces of the two sequential polymerizations
with a different combined [MMBL]/[ItBu] ratio of 2000 (red) and
1500 (blue).
Kinetic experiments employed the [MMBL]0/[ItBu]0 ratios
ranging from 5000 to 15 000 (i.e., with 0.02−0.0067 mol % ItBu
loadings). Even with such low catalyst loadings, a considerable
amount of monomer had already been converted into polymer
by the first data point could be collected (e.g., 18.3% monomer
conversion was observed in 20 s of the reaction with 0.02 mol
% ItBu). Although the kinetic profile of the polymerization at
the earlier stage of the reaction was not captured, the kinetic
plots with the available data clearly showed a first-order
dependence on monomer concentration for all the ratios
investigated (Figure 7). Furthermore, a double logarithm plot
(Figure 8) of the apparent rate constants (kapp), obtained from
the slopes of the best-fit lines to the plots of ln([MMBL]0/
[MMBL]t) vs time for the three [MMBL]/[ItBu] ratios of 5000
Two alternative hypotheses based on the linear polymer
structure could also explain the MALDI-TOF MS result. The
first structure could be formed through a sequence of events
involving abstraction of a proton α to the NHC by the enolate
Figure 7. Plots of the first-order kinetics of ln([MMBL]0/[MMBL]t)
vs time (sec) for the MMBL polymerization by ItBu in DMF at 25 °C.
t
■
●
Conditions: [MMBL]0 = 0.935 M; [I Bu]0 = 0.187 ( ), 0.117 ( ),
□
○
▲
0.0936 ( ), 0.0780 ( ), and 0.0624 ( ) mM.
17930
dx.doi.org/10.1021/ja4088677 | J. Am. Chem. Soc. 2013, 135, 17925−17942

Journal of the American Chemical Society
Article
Figure 9. MALDI-TOF MS spectrum of low-molecular-weight MMBL oligomers produced by TPT in toluene at RT. Inset: plot of m/z values of the
major mass series vs the number of MMBL repeat units (n). The slope of 112.07 corresponds to the mass of the MMBL repeat unit and the intercept
of 22.23 is a sum of the masses of Na+ (from the added NaI) and end groups (none).
MALDI-TOF MS. The second structure, in the case of MMBL
polymerization, could be formed through abstraction of an H
from the β-C of the monomer by the reactive enolate chain end
to generate the saturated chain with the cation [NHC]+ still
attached to it and the monomer anion that reinitiates a new
chain, also effecting a catalytic polymerization (Scheme 1).
To provide experimental evidence to distinguish these two
possible chain structures (cyclic vs linear), we utilized GPC
with a light scattering detector coupled to a viscometer to
analyze and compare the PMMBL samples of similar MW
prepared with ItBu and the GTP method initiated by a SKA,
Me2CC(OMe)OSiMe3, the latter method of which is known
to produce linear MMBL polymers.21a,f A Mark−Houwink plot
of the two types of the PMMBL samples is depicted in Figure
10, featuring indistinguishable intrinsic viscosities of the
polymers; in essence, if the PMMBL produced by ItBu had a
cyclic structure, then one would observe lower intrinsic
viscosity for the cyclic polymer than the linear one, with a
[η]cyclic/[η]linear ratio of ∼0.7.36 The two classes of polymers
also showed nearly superimposable linearly fit plots of log(Mw)
vs elution volume for the same hydrodynamic volumes. Overall,
Scheme 1. Possible Pathways Leading to Cyclic and Linear
Polymers
chain end to form an enamine intermediate, CE1, which reacts
with another monomer to form a linear polymer with chain end
structure CE2 (Scheme 1), accompanied by release of the
NHC (cf., Discussion, vide infra). The released NHC initiates a
new chain, thereby effecting a catalytic polymerization. The last
step of the reaction sequence, CE1 + M → CE2 + NHC, is
analogous to the NHC-catalyzed tail-to-tail dimerization
(umpolung) of methacrylates.19 In the chain end structure of
CE2, one MMBL unit on one chain end loses an H atom, while
the other MMBL unit on the other chain end gains an H atom
and, overall, as if the whole chain had no end groups seen by
Figure 10. Double logarithm plot of intrinsic viscosity [η] vs weight-
average molecular weight (Mw) for PMMBL by NHC and SKA
initiators.
17931
dx.doi.org/10.1021/ja4088677 | J. Am. Chem. Soc. 2013, 135, 17925−17942

Journal of the American Chemical Society
Article
Figure 11. MALDI-TOF MS spectrum of low-molecular-weight MMA oligomers produced by ItBu in DMF at RT. Inset: plot of m/z values of the
major mass series vs the number of MMA repeat units (n).
these results clearly indicate the PMMBL produced by ItBu has
a linear structure.
masses of the chain-end groups corresponding to the ItBu
moiety. Corroborating with this result, the H NMR exhibited
1
To identify the termination chain-end groups of the linear
polymer, two low-molecular-weight PMMBL samples produced
similar peaks to those depicted in Figure S8 for the enamine
intermediate, namely chain end structure CE1. Hence, in the
case of MMA polymerization by ItBu, the chain end generated
is CE1, not CE2. In other words, the chain termination stops at
CE1; hence, the MMA polymerization by ItBu appears to be
noncatalytic as the NHC or other nucleophiles are not
regenerated for chain reinitiation during the chain termination
step.
Chain Transfer and Effects of Chain Transfer Agents.
We hypothesized that catalytic polymerization of acrylics by
NHCs can be achieved by an internal or external chain-transfer
pathway outlined in Scheme 2. As described above, the internal
1
by ItBu were characterized by H NMR. One of them was
produced by ItBu in a low [MMBL]0/[ItBu]0 ratio of 5:1 in
DMF at −40 °C. The solvent was removed, and the product
1
was washed by hexanes. The H NMR (Figure S8) exhibits
peaks at δ 7.10 ppm (d, 1H, J = 3.2 Hz, N−CH=CH−N) and
6.46 ppm (d, 1H, J = 3.6 Hz, N−CH=CH−N), characteristics
of the protons on the imidazole ring of the enamine
intermediate, specifically chain end structure type CE1
(Scheme 1). The other low-molecular-weight PMMBL was
1
produced by ItBu in toluene at RT. The H NMR (Figure S8)
exhibits peaks at δ 6.44 ppm (br, 1H, −CH=), 6.36 ppm (br,
1H, −CH=), 4.46 ppm (br, 1H, OCHMe), and 2.61 ppm (br,
2H, CH2CH=), consistent with chain end structure type CE2
(Scheme 1). On the other hand, for the polymers produced at
high [MMBL]/[ItBu] ratios (e.g., 10 000 or 12 000) in DMF at
RT, several small peaks in the olefinic region were observed at
6.21, 5.60, and 6.70 ppm, attributable to the protons of the
exocyclic double bond (6.21 and 5.60, which are similar, but
not identical, to those the MMBL monomer peaks at 6.14 and
5.58 ppm), and the internal double bond (6.70 ppm),
respectively, as a result of chain transfer to monomer and
reinitiation processes (Scheme 1). Overall, the above combined
studies by MALDI-TOF MS, GPC, NMR, and computational
methods (vide infra) concluded the PMMBL produced by
NHCs at RT is a linear polymer with different chain end
groups, depending on polymerization conditions. The origin of
such end groups and the accompanying chain termination and
transfer mechanisms are described in the Discussion section.
Interestingly for MMA polymerization by ItBu in DMF at
RT, the resulting PMMA had different chain end characteristics
than those of PMMBL. Figure 11 depicts the MALDI-TOF MS
of low-molecular-weight MMA oligomers and the plot of m/z
values vs the number of MMA repeat units (n), the latter of
which yielded a straight line with a slope of 100.1 and an
intercept of 181.4. The slope corresponds to the mass of the
MMA monomer, whereas the intercept is the sum of the
Scheme 2. Proposed Chain Transfer Pathway via Addition of
an External CTA
chain transfer relies on either a chain-termination process that
liberates NHC to reinitiate multiple chains or a chain transfer
to monomer that generates the monomer anion to reinitiate
multiple chains; the former process would be effective at later
stages of a polymerization reaction, while the latter process
would be favored at a high [monomer] to [NHC] ratio (vide
supra). The external chain transfer could proceed in the
presence of suitable CTAs such as enolizable organic acids
which terminate the growing ester enolate chain via
protonation, followed by nucleophilic attack of the resulting
enolate anion to the electrophilic carbon directly bonded to Nu,
releasing the catalyst (Scheme 2). Alternatively, the resulting
enolate can directly initiate new chains.
17932
dx.doi.org/10.1021/ja4088677 | J. Am. Chem. Soc. 2013, 135, 17925−17942

Journal of the American Chemical Society
Article
To achieve effective external chain transfer, we examined four
organic acids, methyl isobutyrate (MIB), 3-methyl butanone
(MBO), dimethyl succinate (DMS), and dimethyl malonate
(DMM), with a decreasing order of their pKa’s (DMSO)
values: MIB (∼29) > MBO (∼27) ≈ DMS ≫ DMM (15.9). In
comparison, the conjugate acid pKa (DMSO) of ItBu was
reported to be ∼23 (23.237 or 22.738). Based on these pKa
values, we reasoned that the acidity of MIB will be too low to
protonate the growing ester enolate chain end and thus not be
an effective CTA, but MBO and DMS should be effective CTAs
as their acidity is higher than the ester, the resulting product of
protonation of the growing ester enolate chain. Moreover,
DMM should be able to protonate ItBu and generate the
corresponding imidazolium carbanion ion pair. The relative
acidity of different CTAs vs ItBu can be confirmed by their
ketone enolate nucleophile Nu′, rather than by the NHC Nu
(cf., Scheme 2).
The behavior of DMS as a CTA was similar to MBO (runs
13, 14 vs 8, 12), due to their similar acidity. In sharp contrast,
the MMBL polymerization with DMM as a CTA showed a
drastic decrease in the polymerization rate as the amount of
DMM added increased (runs 15−21). This observation is
indicative of ineffective reinitiation (i.e., kri < kp), which is due
to the fact that the nucleophilicity of the enolate [CH-
(COOMe)2]− derived from DMM is much lower than that of
the growing ester enolate. Nevertheless, DMM is the most
efficient CTA of the series in terms of the ability to control the
polymer MW. This conclusion was drawn by comparing chain-
transfer constants (Ctr), defined as the ratio of rate constants of
chain transfer to propagation, Ctr = ktr/kp.39 Specifically, Ctr
values were obtained from the slopes of linear plots of 1/Mn vs
the [CTA]/[M] ratio (Figure 13). Accordingly, Ctr (× 104)
values for the MMBL polymerization by ItBu with MBO and
DMM as CTAs were estimated to be 1.99 and 30.1,
respectively. To confirm that the reinitiation is predominately
through the enolate as a result of the chain transfer rather than
by the NHC, low-molecular-weight MMBL oligomers
produced by ItBu with DMM as the CTA were analyzed by
MALDI-TOF MS (Figure 14). A plot of m/z values of the
major mass series vs the number of MMBL repeat units (n)
yielded a straight line with a slope of 112.06 and an intercept of
154.95. The slope corresponds to the mass of the MMBL
monomer, whereas the intercept is the sum of the masses of the
masses of Na+ (from the added NaI, 23) and end groups
(DMM, 132). This result clearly shows that a large majority of
the chains were produced by the CTA-based enolate.
1
reactions with ItBu in toluene-d8. Thus, the H NMR showed
that no reaction occurred between ItBu and MIB, MBO, or
DMS. However, a white precipitate was formed immediately
upon mixing DMM with ItBu. The NMR (CD2Cl2) of the solid
collected by filtration showed the formation of the correspond-
ing ion pair [ItBuH]+[CH(COOMe)2]− (as two isomers). The
structure of this ion pair (E isomer) was further confirmed by
X-ray diffraction analysis (Figure 12).
We also examined the behavior of catalytic chain-transfer
polymerization of the parent methylene butyrolactone, MBL,
using MBO as a CTA, the results of which are summarized in
Table 3. As can be seen from the table, all MBL polymer-
izations in a [MBL]/ItBu] ratio of 1000 achieved quantitative
monomer conversion in 72 h, regardless of the amount of the
MBO added (from 0 to 100 equiv, relative to the catalyst ItBu).
However, the resulting polymer MW decreased gradually from
25.2 kg/mol (run 1) to 16.3 kg/mol (run 6) with increasing the
equiv of MBO from 0 to 100. Accordingly, the I* values
increased from 385% to 595%, featuring again a catalytic
polymerization process.
Figure 12. X-ray crystal structure of [ItBuH]+[CH(COOMe)2]− (E
isomer) with thermal ellipsoids drawn at the 50% probability.
DISCUSSION
Indeed, MIB was showed to be ineffective as a CTA because
the Mn of the resulting PMMBL was not significantly affected
by varying the amount of MIB (from 1 to 100 equiv relative to
ItBu, runs 2−6 vs run 1, Table 2). On the other hand, a gradual
increase in the amount of MBO added from 1 to 100 equiv
relative to ItBu resulted in a corresponding decrease in Mn
68.3−29.2 kg/mol (runs 7−11); a further increase in the MBO
added to 1000 equiv lowered the Mn to only 15.4 kg/mol (run
12). Moreover, addition of MBO did not noticeably affect the
polymerization rate, achieving quantitative monomer con-
version in 1 min for all runs), which implies that the relative
rate constants for propagation (kp) and reinitiation (kri) by the
resulting ketone enolate are about the same (kp ≈ kri); thus, this
chain-transfer polymerization with observed decrease in Mn
while maintaining the more or less same polymerization rate
upon addition of the CTA can be characterized as a “normal
chain-transfer” polymerization.39 As the chain initiation by the
NHC nucleophile is slow relative to propagation, the above
results suggest the chain reinitiation is predominately by the
■
The behavior of NHCs (IMes, ItBu, and TPT) as initiators or
catalysts in the polymerization of MMA and MMBL was further
investigated by density functional theory (DFT) calculations.40
All geometries were localized in the gas phase. However,
because polymerizations were performed in toluene and/or
DMF, we performed single point energy calculations on the
final geometries to take into account solvent effects. For
monomers 1 (MMA and MMBL) we investigated the general
reactivity shown in Scheme 3. Upon addition of an NHC
molecule to monomer 1, leading to zwitterionic adduct 2, two
possible scenarios were considered. In the first scenario,
indicated in Scheme 3 as the enamine/dimerization pathway,
there is an overall H-transfer from the Cα to the Cβ atom of 2,
leading to enamine 3, followed by reaction of 3 with another
monomer molecule, affording the dimerization product 4 with
release of the NHC. In the second scenario, indicated in
Scheme 3 as the polymerization pathway, the NHC adduct 2
directly reacts with another monomer molecule via classical 1,4-
17933
dx.doi.org/10.1021/ja4088677 | J. Am. Chem. Soc. 2013, 135, 17925−17942

Journal of the American Chemical Society
Article
a
Table 2. Results of MMBL Polymerization by ItBu in the Presence of CTAs
run no.
CTA
[CTA]/[ItBu]
[CTA]/[MMBL]
time (min)
conv. (%)
Mn (kg/mol)
PDI (Mw/Mn)
I* (%)
1
2
−
MIB
0
1
0
1
>99
>99
>99
>99
>99
>99
>99
>99
>99
>99
>99
>99
>99
89
71.9
82.5
77.5
80.0
68.0
68.7
68.3
47.7
43.7
31.5
29.2
15.4
50.4
13.5
72.1
25.6
14.9
8.91
6.13
6.41
−
1.80
2.17
2.04
2.10
2.02
1.94
1.56
1.47
1.53
1.33
1.52
1.30
1.95
1.64
1.99
1.81
1.54
1.34
1.26
1.28
−
154
134
143
139
163
161
162
232
254
352
380
720
220
738
155
438
744
1244
1736
1398
0.001
0.01
0.02
0.05
0.1
1
3
MIB
10
1
4
MIB
20
1
5
MIB
50
1
6
MIB
100
1
1
7
MBO
MBO
MBO
MBO
MBO
MBO
DMS
DMS
DMM
DMM
DMM
DMM
DMM
DMM
DMM
0.001
0.01
0.02
0.05
0.1
1
8
10
1
1
9
20
10
11
12
13
14
15
16
17
18
19
20
21
50
1
100
1000
10
1
1
1
0.01
1
1
1000
1
1
0.001
0.005
0.01
0.02
0.05
0.1
5
100
100
99
5
15
10
60
20
1080
2880
2880
11520
>99
b
50
95
c
100
1000
80
d
1
98
a
b
Carried out at ∼25 °C in 4.5 mL DMF and 0.5 mL MMBL, where [MMBL]0 = 0.936 M, [MMBL]/[ItBu] = 1000. 21% collected by precipitation
in methanol. 6% collected by precipitation in methanol. No precipitation in methanol.
c
d
rather large %VBur of TPT is consequence of the rotation of the
phenyl rings next to the carbene C atom that assume a
conformation nearly coplanar with the NHC ring. In IMes, the
ortho Me groups prevent this rotation. On the other hand, the
in-plane orientation of the phenyl rings in TPT allows
extending the conjugation of the NHC ring, further stabilizing
the free TPT molecule. A more polar solvent, DMF, appears to
have a rather small impact on the energy values of the
enamine/dimerization manifold reported in Table 4.
Zwitterionic adduct 2 is the branching point for dimerization
vs polymerization. In the dimerization branch the first step is
the H-transfer reaction to form enamine 3, which reacts with a
second MMA monomer to form the dimerization product 4
accompanied by NHC elimination (Scheme 3). Regarding the
H-transfer step, owing to the expected high energy of the direct
Cα to Cβ 1,2-shift, we focused on the two-step, bimolecular
mechanism proposed in Scheme 4. In the first step, one H atom
(H1 in Scheme 4) is transferred from the Cα of a first NHC-
MMA adduct to the Cβ of a second one, leading to the
formation of the tightly bound ion pair 2′. In the second step,
the resulting intermediate (IM) 2′ undergoes a further H-
transfer reaction where another H atom (H2 in Scheme 4) is
transferred from the Cα of the second moiety to the Cβ of the
first moiety, generating two molecules of enamine product 3. A
similar two-step bimolecular proton-transfer mechanism was
proposed for the Stetter reaction in nonprotic media involving
1,4-addition of aldehydes and Michael acceptors.42
The energetics of IM and TS species in Scheme 4 is reported
in Table 4. Starting with TPT, the TPT-MMA adduct 2
undergoes a first H-transfer through TS 2−2′ with a barrier of
∼11−12 kcal/mol, leading to the charged intermediate 2′,
roughly 14 kcal/mol below 2. The second H-transfer, through
TS 2′−3, also has a barrier of ∼13 kcal/mol and leads to the
final enamine 3, which is clearly more stable than the initial
adduct 2 (by ∼16 kcal/mol) and also of IM 2′ (by ∼2 kcal/
mol). The enamine product 3 can react with a further MMA
Figure 13. Plots of 1/Mn vs [CTA]/[M] (M = MMBL) for chain-
transfer polymerization of MMBL by ItBu. CTA = DMM (Δ); MBO
▲
(
).
conjugate addition to start the polymerization process. The
mechanisms of the two pathways were investigated computa-
tionally both in toluene and DMF, considering all three NHCs,
TPT, IMes, and ItBu, as depicted in Figure 1, and two
monomers, the linear MMA and its cyclic analog MMBL.
Reactivity of NHCs toward MMA. Energetics relevant to
the formation of the zwitterionic species 2 through NHC
addition to monomer is summarized in Table 4. According to
calculations, in either toluene or DMF the lowest energy NHC-
MMA zwitterionic adduct 2 is that formed with IMes. Based on
the data summarized in Table 4, the relative stability of the
NHC-MMA adduct is IMes > TPT ≈ ItBu in toluene. This
trend is not that expected based on the relative nucleophilicity
of the NHCs (ItBu > IMes > TPT), but rather it correlates with
the relative steric demand of the NHCs, as quantified by the
buried volume (%VBur) of the NHC moiety41 in the MMA
adduct, ItBu = 36.1% ≈ TPT = 36.8% > IMes =32.3%. The
17934
dx.doi.org/10.1021/ja4088677 | J. Am. Chem. Soc. 2013, 135, 17925−17942

Journal of the American Chemical Society
Article
Figure 14. MALDI-TOF MS spectrum of low-molecular-weight MMBL oligomers produced by ItBu with DMM as the CTA. Inset: plot of m/z
values of the major mass series vs the number of MMBL repeat units (n).
a
Table 3. Results of MBL Polymerization by ItBu in the Presence of MBO
run no.
[MBL]/[ItBu]
[CTA]/[ItBu]
[CTA]/[MBL]
time (h)
conv. (%)
Mn (kg/mol)
PDI (Mw/Mn)
I* (%)
1
2
3
4
5
6
1000
1000
1000
1000
1000
1000
0
1
0
72
72
72
72
72
72
>99
>99
>99
>99
>99
>99
25.2
24.0
22.7
21.7
18.5
16.3
1.50
1.52
1.50
1.45
1.46
1.77
385
404
427
447
524
595
0.001
0.01
0.02
0.05
0.1
10
20
50
100
a
Carried out at ∼25 °C in 4.5 mL DMF and 0.5 mL MBL, [MBL]0 = 1.14 M.
Scheme 3. Outlined Two Competing Reaction Pathways Involving Acrylic Monomers (MMA and MMBL) and NHCs
monomer via a classical monomer addition step (Scheme 5),
through TS 3−3′ with a barrier of ∼21 kcal/mol, leading to the
high energy IM 3′. A final H-transfer step converts 3′ into
dimerization product 4 with release of TPT and an energy gain
of only 0.4 kcal/mol in toluene relative to 3 and basically
thermo-neutral in DMF. These yielded mechanistic features for
the TPT-catalyzed MMA dimerization, including bimolecular
(or intermolecular) proton transfer and the rate-limiting step
being the addition of the enamine intermediate to the second
MMA (which exhibits the highest activation barrier of ∼21
kcal/mol), agreeing well with the most recent results of
experimental mechanistic studies by Matsuoka and co-work-
ers.43
reasonably close energy. However, there is a relevant difference
in the relative stability of enamine 3 with respect to product 4.
In fact, according to calculations, enamine 3 is the most stable
species in the IMes promoted reactivity of MMA, whereas for
TPT the most stable species is dimer 4. This difference suggests
that in the presence of IMes MMA dimerization is unfavored
because dimer 4 is 2.3 kcal/mol higher in energy than enamine
3. Finally, ItBu is unable to promote MMA dimerization, due to
the high-energy barrier for the first H-transfer, 27 kcal/mol, as a
result of the greater steric hindrance of ItBu in the ItBu-MMA
adduct. To reaffirm this reasoning, we also performed the
calculations with 1,3-di-isopropylimidazolin-2-ylidene (IiPr)
and found the energy barrier for its H-transfer process is ∼11
kcal/mol lower than that with ItBu, in agreement with the steric
difference between the two NHC-MMA adducts. Overall, all
Moving to IMes, we found that the energy profiles of IMes
are rather similar to those of TPT, with all TSs being of
17935
dx.doi.org/10.1021/ja4088677 | J. Am. Chem. Soc. 2013, 135, 17925−17942

Journal of the American Chemical Society
Article
Lastly for ItBu, the computational data are rather different
from those for IMes and TPT and further are different in
toluene and in DMF. In fact, in toluene the H-transfer barrier
leading to enamine 3 is too high (27 kcal/mol), preventing the
reaction to move along this pathway, while the MMA addition
barrier starting polymerization is low, but the addition product
5 is not stable. As consequence, ItBu is practically inactive
toward MMA in toluene. Moving to DMF, the H-transfer
reaction along the enamine/dimerization pathway is still
inaccessible, but the ItBu-MMA adduct can add another
monomer with a negligible energy barrier leading to stable
addition product 5, −4.0 kcal/mol below MMA + ItBu.
Obviously, the higher stability of the addition product (the
zwitterionic propagating species) along the polymerization
pathway in the more polar DMF is due to the electrostatic
stabilization. Again, these computational data are in agreement
with the experiments that showed that ItBu does not react with
MMA in toluene, but it polymerizes MMA in DMF (vide
supra).
Table 4. Energy of the Species Shown in Schemes 3−5, with
a
MMA as the Monomer
IMes
ItBu
TPT
IMes
ItBu
TPT
toluene
DMF
1
2
0
0
0
0
0
0
−7.3
0.4
−2.9
−8.1
−1.4
−1.2
Enamine/Dimerization
2−2′
2′
2′−3
3
4.3
27.4
9.4
−17.2
−3.8
−19.2
2.1
6.6
27.7
15.3
29.9
−4.4
9.7
10.0
−15.4
−1.7
−18.8
1.9
−12.4
−5.9
−21.9
−2.5
−6.1
−19.6
16.3
29.0
−5.3
9.9
−10.3
−1.4
−19.9
−2.0
−7.7
−18.8
3−3′
3′
6.0
0.9
5.4
0.1
4
−19.6
−19.6
Polymerization
4.5
−18.8
−18.8
2−5
5
−3.7
−9.4
3.6
0.4
−5.5
−12.4
0.6
2.8
−3.6
−4.0
−6.1
a
Energy is reported in kcal/mol and relative to NHC + free MMA. X−
Y denotes transition state (TS) from species X to species Y.
Termination Reaction in MMA Polymerization. In this
section we discuss a possible mechanism of termination
operative in the MMA polymerization promoted by ItBu in
DMF. We investigated the pathway proposed in Scheme 6 and
used two methyl groups to model the chain segment
connecting the cationic and anionic moieties of the growing
chain in the starting zwitterionic species (P = CH3 in Scheme
6). This termination reaction corresponds to a H-transfer from
the −CH2− group bound to the ItBu to the enolate C atom of
the chain end, through a TS with a barrier of 8.8 kcal/mol,
leading to the termination product (an enamine) with chain
ends shown in the Scheme 6. The resulting termination
product is 13.1 kcal/mol lower in energy relative to the starting
zwitterionic propagation species considered. However, the
barrier for the propagation species to add another MMA (∼2
kcal/mol) is much lower than that for this termination step,
explaining why a medium-high MW polymer (33.2 kg/mol) is
achieved by ItBu in DMF. We also considered the possibility of
the enamine termination product to add another MMA
molecule, but the energy barrier for this addition is calculated
to be 26.8 kcal/mol. Overall, these computational results are in
agreement with experimental evidence for the observed
these computational results are consistent with our exper-
imental findings described in the Results section.
Next we examined the behavior of the NHC-MMA
zwitterionic adduct 2 in promoting polymerization, through
addition of a second MMA to 2 (Scheme 3). Analysis of the
data reported in Table 4 indicates that addition of a second
MMA molecule to any of the NHC-MMA adducts, through TS
2−5, is a rather feasible process with a barrier smaller than 10
kcal/mol, whatever NHC and solvent is considered. The
product of the first propagation step (i.e., the second monomer
addition), 5, is only slightly more stable than the starting
species, making addition of MMA to the NHC-MMA adduct
reversible. Comparison of the H-transfer with the polymer-
ization pathway, that is TS 2−2′ vs 2−5, indicates that for TPT
the barrier for the second monomer addition, roughly 4−7
kcal/mol (depending on solvent), is about one-half of that for
the H-transfer converting 2 into enamine species 3 and finally
the dimer product 4. However, the overall product of the H-
transfer, dimer 4, is the thermodynamically more stable species,
in agreement with the experimental evidence that the TPT−
MMA adduct undergoes H-transfer and than dimerizes. Moving
to IMes, the energy profiles of TPT and IMes promoted
polymerization are again rather similar, with the difference
being that IMes, in agreement with the experimental results,
stops at the more stable enamine product 3 without reaching to
dimer 4. These conclusions are independent of the solvent
polarity.
1
enamine chain ends by H NMR and MALDI-TOF MS (vide
supra), and the MMA polymerization is noncatalytic (i.e., the
terminated enamine species is not longer active for further
MMA additions).
Reactivity of NHCs toward MMBL. The same competing
enamine/dimerization and polymerization pathways were
explored for MMBL, the results of which are reported in
Table 5. The overall scenario emerging from Table 5 is rather
Scheme 4. Proposed Two-Step Bimolecular H-Transfer Mechanism Involving the Ion Pair Intermediate
17936
dx.doi.org/10.1021/ja4088677 | J. Am. Chem. Soc. 2013, 135, 17925−17942

Journal of the American Chemical Society
Article
Scheme 5. Energetics (kcal/mol) of MMA Dimerization Catalyzed by TPT
a
Scheme 6. Energetics for the Species Involved in the Termination Step of MMA Polymerization by ItBu in DMF
a
Note the cationic and anionic moieties of the growing chain are connected through two P sites.
similar to that discussed above for MMA; hence, the details are
enamine/dimerization pathway for MMBL either. Conversely,
IMes and TPT result in a low energy barrier for the first H-
transfer step, ∼9−15 kcal/mol, and in stable products, ∼18−23
kcal/mol below NHC + free MMBL, in both toluene and
DMF. However, for both IMes and TPT, enamine 3 is more
stable than dimer 4, which is in agreement with the
experimental evidence that no dimer of MMBL is formed
whatever NHC is considered.
not discussed here, illuminating instead the differences.
Table 5. Energy of the Species Shown in Schemes 3 and 4,
a
with MMBL as the Monomer
IMes
ItBu
TPT
IMes
ItBu
TPT
toluene
DMF
Next we focused on the first step of the polymerization
pathway, in which the NHC-MMBL zwitterionic adduct 2
attacks the exocyclic CH2 group of a MMBL molecule via 1,4-
conjugate addition. As for MMA, the most stable NHC-
monomer adduct is the one with IMes in both solvents,
followed by the adduct with TPT in toluene and the one with
ItBu in DMF. Again, the energy barrier for the NHC-MMBL
adduct 2 to add another monomer molecule, through TS 2−5,
is quite small with all the three NHCs considered herein, 2.4
kcal/mol for ItBu, 3.4 kcal/mol for TPT and 4.0 kcal/mol for
IMes in toluene. A similar trend is seen in DMF. These
computational data would suggest that all the three NHCs
considered could promote MMBL polymerization. In fact,
experimentally ItBu promotes extremely rapid polymerization
of MMBL in both solvents, especially in DMF, and IMes also
initiates MMBL polymerization in either solvent, albeit with
much lower activity than ItBu, while TPT is inactive for MMBL
polymerization in DMF, but in toluene it is active with even
lower activity than IMes (vide supra).
1
1−2
2
0
0
0
0
0
0
6.1
14.4
7.1
7.8
12.9
−3.1
8.7
−8.8
−0.6
−2.7
−9.9
−2.6
Enamine/Dimerization
2−2′
2′
2′−3
3
4.1
24.5
6.7
−19.3
−6.8
−19.3
−2.0
5.3
24.8
4.8
8.7
−18.6
−10.8
−23.2
−7.0
−10.9
−18.7
4.7
25.0
−5.0
9.6
−15.6
−6.3
−20.2
−6.8
−13.7
−18.4
−15.6
−3.0
−18.6
7.6
27.7
−4.3
−2.9
−1.2
−18.4
3−3′
3′
3.8
−5.1
−8.1
−18.4
4
−18.7
−18.7
Polymerization
0.7
2−5
5
−4.8
−10.0
1.8
−4.9
−10.4
0.7
2.1
−4.5
−7.0
−6.7
−7.2
a
Energy is reported in kcal/mol and relative to NHC + free MMBL.
In short, the energetic numbers reported in Table 5 indicate
again that ItBu behaves differently with respect to IMes and
TPT also in case of MMBL. The energy of TS 1−2 for the
formation of the zwitterionic adduct 2 with ItBu is roughly 7−8
kcal/mol higher in energy than that with IMes and TPT, and
the energy barrier for the H-transfer step along the enamine/
dimerization pathway is roughly 20 kcal/mol higher than that
calculated for IMes and TPT. This barrier is large enough to
conclude that ItBu is unable to promote reactivity along the
The different reactivity trend toward MMBL polymerization
among the three NHCs can be rationalized by relative aptitude
in competing H-transfer and monomer-addition reactions
involving the zwitterionic adduct 2. Specifically, in the case of
ItBu, the H-transfer reaction barrier along the enamine/
dimerization pathway is much larger than the barrier along
17937
dx.doi.org/10.1021/ja4088677 | J. Am. Chem. Soc. 2013, 135, 17925−17942

Journal of the American Chemical Society
Article
Scheme 7. A Possible Alternative Initiation and Propagation Pathway for the MMBL Polymerization by NHCs, Where the NHC
Serves as a Base
the polymerization pathway, permitting the reactivity only
along the polymerization pathway and shutting down the
enamine/dimerization pathway. In case of TPT and IMes, both
the H-transfer reaction along the enamine/dimerization
pathway and the MMBL addition along the polymerization
pathway can occur, in principle, since they both show relatively
low-energy barriers. Although the barrier for the polymerization
step is clearly lower than the barrier for the H-transfer, the most
remarkable difference between the two pathways is in the
stability of the products. In fact, the MMBL addition product 5
along the polymerization pathway is roughly 10 kcal/mol less
stable than the H-transfer reaction product enamine 3. Since
MMBL addition from 2 to 5 can be reversed, the system can
accumulate into the most stable enamine 3, reducing the
polymerization activity. However, the system with IMes is more
biased toward the polymerization pathway, since the preference
for the polymerization pathway, measured by the energy
difference between TSs 2−2′ and 2−5, increases by roughly 2
kcal/mol on going from TPT to IMes. This analysis explains
the experimental evidence that the MMBL polymerization
activity with IMes is somewhat between ItBu and TPT
activities.
to another MMBL molecule. The energetics of the alternative
mechanism shown in Scheme 7 in DMF is reported in Table 6.
a
Table 6. Energy of the Species Shown in Scheme 7
IMes
ItBu
TPT
DMF
1
1−6
6
0
0
0
17.6
17.7
−
16.7
14.3
−
20.1
32.0
−
b
6−7
7
−4.4
−7.8
+9.9
a
Energy is reported in kcal/mol and relative to NHC + free MMBL.
No energy barrier.
b
The rate-determining step for all the three NHCs is the
formation of the active species 6 through abstraction of the β-
proton from MMBL by an NHC base. In fact, once the anion 6
is formed, it can add a monomer molecule without any energy
barrier. Thus, the highest energy barrier in the reaction pathway
of Scheme 7, through TS 1−6 is roughly 17−20 kcal/mol for
the three NHC systems. Comparing the energy of TS 1−6 in
Table 6, and TS 1−2 in Table 5, it is clear that the mechanism
of Scheme 7 is highly unfavored relative to the mechanism of
Scheme 3 for both IMes and TPT. In fact, the energy barrier of
TS 1−2 is 9.8 and 11.4 kcal/mol lower than TS 1−6 for IMes
and TPT, respectively. For ItBu, the difference is lower, but
with TS 1−2 being 3.8 kcal/mol lower in energy than TS 1−6,
the proton abstraction pathway outlined in Scheme 7 is still less
favorable than the nucleophilic addition pathway outlined in
Scheme 3. This conclusion is further reaffirmed by the
experiment that showed no reaction took place between ItBu
and γ-valerolactone (1−5 equiv), a suitable nonpolymerizable
monomer model possessing even more acidic protons than the
monomer MMBL.
Termination Reaction in MMBL Polymerization. To
shed light on the termination mechanism operating in the
MMBL polymerization promoted by ItBu, we explored three
possible termination pathways. The first possible mechanism,
consisting of SN2 nucleophilic substitution reaction where the
NHC is the leaving group, would lead to formation of a cyclic
polymer, according to the two possible pathways schematized
in Scheme 8. In the mechanism of Scheme 8a the nucleophile
center is the Cα atom of the last MMBL unit and a new C−C
bond is formed in 10, while in that of Scheme 8b the
nucleophile center is the enolate oxygen, where the major
negative charge is localized, and an O−C bond is formed in 11.
As the other general trend for ItBu and IMes, comparison of
the energetics in toluene with that in DMF indicates that in the
more polar DMF solvent the NHC-monomer adduct 2 and the
product 5 tend to be more stable, and the energy of the TS
involved in the monomer addition becomes slightly lower; this
observation is in agreement with the greater experimental
activity of these two NHCs in DMF. However, the trend for
TPT is opposite in the two solvents, in agreement with the
experimental results showing that TPT is inactive for MMBL
polymerization in DMF.
Considering that the data reported in Table 5 indicate that
the TS 1−2 for the formation of ItBu-MMBL adduct is the
highest in energy, and yet the experimental data showed that
ItBu is the most active in the MMBL polymerization, we
explored the alternative initiation pathway outlined in Scheme
7. In this pathway, an NHC serves as a strong base to abstract a
proton of MMBL, forming a high-energy, highly reactive anion
that initiates rapid polymerization. We examined abstraction of
both β- and γ-protons of MMBL and found that abstraction of
β-H costs 14.3 kcal/mol, while abstraction of γ-H costs 49.5
kcal/mol. Accordingly, the much preferred β-H abstraction
generates an initiating species, anionic monomer 6 stabilized by
electronic delocalization over the extend π-conjugated orbitals
(Scheme 7). The first propagation step consists of addition of 6
17938
dx.doi.org/10.1021/ja4088677 | J. Am. Chem. Soc. 2013, 135, 17925−17942

Journal of the American Chemical Society
Article
Scheme 8. Examined Possible Termination Pathways to a Cyclic Polymer
Scheme 9. Possible Termination Pathways Through H-Transfer and Enamine Intermediate 13
a
In both cases two methyl groups were used to model the chain
segment connecting the cationic and anionic moieties of the
growing chain in the starting zwitterionic species 9 (P = CH3 in
Scheme 8). This treatment avoids the complication of modeling
a long polymer chain joining the two reacting moieties.
Nevertheless, both pathways showed very high energy, with the
mechanism of Scheme 8a resulting in a barrier higher than 40
kcal/mol, while for the mechanism in Scheme 8b the calculated
barrier is ∼50 kcal/mol, thus effectively ruling out these
termination pathways. This analysis is consistent with our
experimental result that showed no cyclic polymer formation.
Keeping in mind the feasibility of H-transfer reactions with
these systems, we investigated the second possible chain
termination pathways outlined in Scheme 9 starting from the
active zwitterionic propagating species 12. The relative
energetics of all species involved are summarized in Table 7.
Table 7. Energy of the Species Shown in Scheme 9
ItBu toluene
ItBu DMF
12
12−13
13 + CE1
13−14
CE2
0
0
9.9
10.8
−10.7
−0.2
−30.7
9.7
−12.8
2.1
−31.2
8.5
13−15
15
6.2
6.7
15−CE3
15−CE4
CE3
>30
39.6
−18.2
−9.8
>30
42.1
−17.4
−7.6
CE4
a
Energy is in kcal/mol.
17939
dx.doi.org/10.1021/ja4088677 | J. Am. Chem. Soc. 2013, 135, 17925−17942

Journal of the American Chemical Society
Article
The first step of the termination sequence corresponds to a H-
transfer from the −CH2− group bound to the ItBu to the
enolate C atom of the chain end in 12, through TS 12−13 with
a barrier of roughly 10 kcal/mol, leading to enamine type
species 13 and chain end CE1 (Scheme 9). Chain end CE1
could not be identified by NMR, due to its overlap with the
main polymer peaks, but the enamine moiety in 13 can be
identifiable by NMR and MALDI-TOF MS (vide supra).
Interestingly, this reaction also allows for formation of an
enamine-type structure with ItBu, a step not possible by direct
H-transfer between two ItBu-MMBL adducts (cf., Table 5).
Enamine species 13 can further react with another MMBL
molecule through a classical 1,4-addition to give IM 14,
through TS 13−14 with an energy barrier of also roughly 10
kcal/mol (in DMF), and finally chain end, CE2, through a
second H-transfer step. Alternatively, the reaction of enamine
species 13 with another MMBL could proceed via a less favored
1,2-addition pathway to give IM 15, through TS 13−15 with a
barrier of roughly 15 kcal/mol. Subsequently, IM 15 could
undergo another H-transfer or nucleophilic substitution by the
oxygen atom, leading respectively to chain end CE3 or CE4.
However, the TSs for the final nucleophilic attack or H-transfer
have been calculated to be more than 40 kcal/mol higher in
energy, making this termination pathways unlikely.
chain end CE2 and release of the NHC catalyst (green box,
Scheme 9), has an activation energy barrier of slightly higher
than 10 kcal/mol, while the chain termination through chain
transfer to monomer (Scheme 10) has an energy barrier of <7
kcal/mol. On the basis of experimental results (vide supra), the
H-transfer/enamine addition termination is effective only with
a high NHC loading (e.g., the NMR signals relative to CE1 and
CE2 were found with a [MMBL]0/[ItBu]0 ratio of 5:1 or
similar). In other instances (e.g., high [MMBL]0/[ItBu]0 ratios)
the chain-transfer termination to monomer, which is more
favored from an energetic point of view (6.8 vs 10 kcal/mol),
takes place. Finally, these activation energies associated with the
chain termination events are compared to a barrier of only ∼3
kcal/mol for the active propagation species to add another
MMBL, thereby explaining why a rather high MW polymer (up
to ∼90 kg/mol) can be achieved by ItBu.
CONCLUSIONS
■
In summary, this contribution presented a full account of our
combined experimental and theoretical/computational inves-
tigations into the recently discovered NHC-mediated organo-
catalytic conjugate-addition polymerization of acrylic mono-
mers, represented by the most common linear monomer MMA
and its cyclic analog MMBL. Remarkably, there exhibits an
exquisite selectivity of the NHC structure for the three types of
reactions with the substrate it promotes: enamine formation,
dimerization, and polymerization. This organopolymerization is
especially effective for the biomass-derived renewable methyl-
ene butyrolactones such as MMBL, thus quantitatively
converting a large excess of monomer (e.g., 10 000 equiv or
0.01 mol % NHC loading) into the corresponding bioplastics
within 5 min and achieving an exceptionally high maximum
turnover frequency up to 122 s−1. Unique chain termination
mechanisms have been revealed, which account for the
production of relative high-molecular-weight linear polymers
and the catalytic nature of this polymerization. Computational
studies have provided mechanistic insights into reactivity and
selectivity between two competing pathways for each NHC-
monomer zwitterionic adduct: enamine formation/dimerization
through proton transfer vs polymerization through conjugate
addition.
We further investigated the third possible chain termination
pathway through chain transfer to monomer as outlined in
Scheme 10. In this case the reactive enolate chain end abstracts
Scheme 10. Proposed Chain Termination Pathway via Chain
Transfer to Monomer
Key conclusions drawn from the herein described study are
listed as follows:
(1) IMes reacts with MMA to form exclusively the single-
monomer addition product, enamine, or deoxy-Breslow
intermediate 3. This reaction proceeds through a
proposed two-step, bimolecular mechanism involving
initial facile formation of the zwitterionic IMes-MMA
adduct 2, followed by proton transfer between the two
molecules of 2 to generate a tightly bound ion pair
intermediate which leads to the final enamine 3 after an
additional proton-transfer reaction within the ion pair.
The IMes-based enamine is the most stable enamine of
the current NHC series, and it does not react further
with monomer to form the dimerization product 4, due
to the fact that dimer 4 is calculated to be 2.3 kcal/mol
higher in energy than enamine 3.
a proton from the β-C of the monomer, with a barrier of only
6.8 kcal/mol. As result, the old chain P still has the [NHC]+
attached to it as the polymeric counterion, paired with the
monomer anion that re-initiates a new chain, effecting a
catalytic polymerization (cf., Schemes 1 and 10). Although the
overall reaction seems to be not favored (i.e., the products are
1.6 kcal/mol higher in energy vs the reactants), it is worth
noting that the generated strong nucleophile (anionic
monomer) can readily add a monomer without any energy
barrier (Scheme 10), leading to formation of more favored
products (with an energy gain of 20.5 kcal/mol for the addition
of the second monomer).
Overall, after having explored several different termination
pathways, we found that only H-transfer/enamine addition and
transfer to monomer pathways are energetically feasible.
Between these two pathways, the chain termination, through
an H-transfer reaction to generate the enamine type
intermediate 13, followed by an 1,4-addition to another
MMBL molecule and subsequent second H-transfer to finally
(2) TPT catalyzes efficient dimerization of MMA to form
dimer 4 in good yields. In contrast to IMes, TPT-derived
enamine 3 can react further with MMA through a
conjugate-addition step followed by a proton-transfer
step, affording dimer 4. Energetically, the addition of
17940
dx.doi.org/10.1021/ja4088677 | J. Am. Chem. Soc. 2013, 135, 17925−17942

Journal of the American Chemical Society
Article
enamine 3 to the second MMA has the highest energy
barrier of ∼20 kcal/mol, thus a rate-limiting step of the
dimerization process, and for TPT the most stable
species is dimer 4, thus providing an overall thermody-
namic driving force for the dimerization.
organic acid such as MBO can also promote external
catalytic chain-transfer polymerization.
ASSOCIATED CONTENT
* Supporting Information
Full experimental and computational details. This material is
available free of charge via the Internet at http://pubs.acs.org.
■
S
(3) ItBu does not react with MMA in toluene to form the
corresponding zwitterionic adduct 2, which is unstable
due to the large buried volume %VBur of ItBu, and is also
unable to promote MMA dimerization, due to the high-
energy barrier (27 kcal/mol) for the first proton transfer
on going from zwitterionic adduct 2 to enamine
intermediate 3.
AUTHOR INFORMATION
■
Corresponding Authors
luigi.cavallo@kaust.edu.sa
eugene.chen@colostate.edu
(4) ItBu polymerizes MMA to PMMA with a medium-high-
molecular-weight (Mn = 33.2 kg/mol) in the polar
solvent DMF, due to the increased stability of the first
zwitterionic adduct and the subsequent zwitterionic
active propagating species in DMF as well as a relatively
low-energy barrier (∼2 kcal/mol) for subsequent
conjugate additions of propagating species to monomer.
The chain termination reaction, through proton transfer
from the −CH2− group bound to the ItBu to the enolate
C atom of the growing chain end, generates an enamine
chain end, which is incapable of reacting further with
monomer. Hence, the MMA polymerization promoted
by ItBu in DMF is noncatalytic.
(5) For MMBL, the NHC promotes no dimerization but
polymerization, with the polymerization activity being
highly sensitive to the NHC structure and the solvent
polarity. Again, ItBu is unable to promote MMBL
reactivity along the enamine/dimerization pathway, due
to the high-energy barrier (25 kcal/mol in toluene or 28
kcal/mol in DMF) for the first proton transfer on going
from zwitterionic adduct 2 to enamine intermediate 3.
For IMes and TPT, the energy barrier for proton transfer
is much lower, but enamine 3 is more stable than dimer
4, thus no dimer formation. On the other hand, the
energy barrier for the polymerization pathway via
conjugate addition of the NHC-MMBL zwitterionic
adduct 2 to monomer is quite small with all the three
NHCs considered, 2.4 kcal/mol for ItBu, 3.4 kcal/mol for
TPT, and 4.0 kcal/mol for IMes in toluene (a similar
trend is also seen in DMF). The different reactivity trend
toward MMBL polymerization among the three NHCs,
ItBu > IMes > TPT, can be rationalized by the relative
aptitude in competing proton-transfer and conjugate-
addition reactions involving zwitterionic adduct 2, the
branching point of the two reactivity pathways.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This work was supported by the National Science Foundation
(NSF-1012326 and NSF-1300267) for the study carried out at
Colorado State University. L.C. thanks the HPC team of Enea
(www.enea.it) for using the ENEA-GRID and the HPC
facilities CRESCO (www.cresco.enea.it) in Portici, Italy.
REFERENCES
■
(1) Selected recent reviews: (a) Wende, R. C.; Schreiner, P. R. Green
Chem. 2012, 14, 1821−1849. (b) Jacobsen, E. N.; MacMillan, D. W. C.
Proc. Natl. Acad. Sci. U.S.A. 2010, 107, 20618−20619. (c) Grondal, C.;
Jeanty, M.; Enders, D. Nat. Chem. 2010, 2, 167−178. (d) MacMillan,
D. W. C. Nature 2008, 455, 304−308. (e) Houk, K. N.; List, B. Eds.
Asymmetric Organocatalysis, special edition issue, Acc. Chem. Res.
2004, 37, 487−631.
(2) Selected recent reviews: (a) Ryan, S. J.; Candish, L.; Lupton, D.
W. Chem. Soc. Rev. 2013, 42, 4906−4917. (b) Grossmann, A.; Enders,
D. Angew. Chem., Int. Ed. 2012, 51, 314−325. (c) Bugaut, X.; Glorius,
F. Chem. Soc. Rev. 2012, 41, 3511−3522. (d) Biju, A. T.; Kuhl, N.;
Glorius, F. Acc. Chem. Res. 2011, 44, 1182−1195. (e) Droge, T.;
̈
Glorius, F. Angew. Chem., Int. Ed. 2010, 50, 6940−6952. (f) Chiang, P.-
C.; Bode, J. W. In RSC Catalysis Series; Royal Society of Chemistry:
Cambridge, 2010, pp 399−435. (g) Moore, J. L.; Rovis, T. Top. Curr.
Chem. 2009, 291, 77−144. (h) Hahn, F. E.; Jahnke, M. C. Angew.
Chem., Int. Ed. 2008, 47, 3122−3172. (i) Nair, V.; Vellalath, S.; Babu,
B. P. Chem. Soc. Rev. 2008, 37, 2691−2698. (j) Marion, N.; Díez-
́
Gonzalez, S.; Nolan, S. P. Angew. Chem., Int. Ed. 2007, 46, 2988−3000.
(k) Enders, D.; Niemeier, O.; Henseler, A. Chem. Rev. 2007, 107,
5606−5655. (l) Bourissou, D.; Guerret, O.; Gabbaï, F. P.; Bertrand, G.
Chem. Rev. 2000, 100, 39−91.
(3) (a) Nyce, G. W.; Glauser, T.; Connor, E. F.; Mock, A.;
Waymouth, R. M.; Hedrick, J. L. J. Am. Chem. Soc. 2003, 125, 3046−
3056. (b) Connor, E. F.; Nyce, G. W.; Myers, M.; Mock, A.; Hedrick,
J. L. J. Am. Chem. Soc. 2002, 124, 914−915. (c) Nederberg, F.;
Connor, E. F.; Moller, M.; Glauser, T.; Hedrick, J. L. Angew. Chem., Int.
Ed. 2001, 40, 2712−2715.
(4) Selected reviews: (a) Brown, H. A.; Waymouth, R. M. Acc. Chem.
̈
̈
(6) ItBu rapidly polymerizes MMBL in DMF to PMMBL,
thus quantitatively converting 1000−3000 equiv of
monomer in 1 min or 10 000 equiv in 5 min. The
molecular weight of the resulting polymer remains nearly
constant in a narrow range of Mn = 70−85 kg/mol
regardless of the [MMBL]/[ItBu] ratio employed as long
as it is higher than 800, due to substantial internal chain
transfer that also results in high initiator efficiency values
up to 1600%. Two chain-transfer termination pathways
have been identified, with the H-transfer/enamine
addition termination being present only with a high
catalyst loading and the more energetically favored chain
transfer to monomer being effective in other instances,
but both pathways provide internal chain-transfer routes
for catalytic polymerization. Addition of a suitable
̈
̀
Res. 2013, DOI: 10.1021/ar400072z. (b) Fevre, M.; Pinaud, J.;
Gnanou, Y.; Vignolle, J.; Taton, D. Chem. Soc. Rev. 2013, 42, 2142−
2172. (c) Kiesewetter, M. K.; Shin, E. J.; Hedrick, J. L.; Waymouth, R.
M. Macromolecules 2010, 43, 2093−2107. (d) Kamber, N. E.; Jeong,
W.; Waymouth, R. M.; Pratt, R. C.; Lohmeijer, B. G. G.; Hedrick, J. L.
Chem. Re. 2007, 107, 5813−5840.
(5) (a) Jeong, W.; Shin, E. J.; Culkin, D. A.; Hedrick, J. L.;
Waymouth, R. M. J. Am. Chem. Soc. 2009, 131, 4884−4891. (b) Culkin,
D. A.; Jeong, W.; Csihony, S.; Gomez, E. D.; Balsara, N. P.; Hedrick, J.
L.; Waymouth, R. M. Angew. Chem., Int. Ed. 2007, 46, 2627−2630.
(c) Dove, A. P.; Li, H.; Pratt, R. C.; Lohmeijer, B. G. G.; Culkin, D. A.;
Waymouth, R. M.; Hedrick, J. L. Chem. Commun. 2006, 2881−2883.
(d) Coulembier, O.; Dove, A. P.; Pratt, R. C.; Sentman, A. C.; Culkin,
D. A.; Mespouille, L.; Dubois, P.; Waymouth, R. M.; Hedrick, J. L.
17941
dx.doi.org/10.1021/ja4088677 | J. Am. Chem. Soc. 2013, 135, 17925−17942

Journal of the American Chemical Society
Article
Angew. Chem., Int. Ed. 2005, 44, 4964−4968. (e) Csihony, S.; Culkin,
D. A.; Sentman, A. C.; Dove, A. P.; Waymouth, R. M.; Hedrick, J. L. J.
Am. Chem. Soc. 2005, 127, 9079−9084.
(21) Recent reviews and examples: (a) Gowda, R. R.; Chen, E. Y.-X.
Encycl. Polym. Sci. Technol. 2013, DOI: 10.1002/0471440264.pst606.
(b) Chen, X.; Caporaso, L.; Cavallo, L.; Chen, E. Y.-X. J. Am. Chem.
Soc. 2012, 134, 7278−7281. (c) Shin, J.; Lee, Y.; Tolman, W. B.;
Hillmyer, M. A. Biomacromolecules 2012, 13, 3833−3840. (d) Zhang,
Y.; Gustafson, L. O.; Chen, E. Y.-X. J. Am. Chem. Soc. 2011, 133,
13674−13684. (e) Cockburn, R. A.; Siegmann, R.; Payne, K. A.;
Beuermann, S.; McKenna, T. F. L.; Hutchinson, R. A. Biomacromo-
lecules 2011, 12, 2319−2326. (f) Miyake, G. M.; Zhang, Y.; Chen, E.
Y.-X. Macromolecules 2010, 43, 4902−4908. (g) Hu, Y.; Xu, X.; Zhang,
Y.; Chen, Y.; Chen, E. Y.-X. Macromolecules 2010, 43, 9328−9336.
(h) Miyake, G. M.; Newton, S. E.; Mariott, W. R.; Chen, E. Y.-X.
(6) (a) Shin, E. J.; Brown, H. A.; Gonzalez, S.; Jeong, W.; Hedrick, J.
L.; Waymouth, R. M. Angew. Chem., Int. Ed. 2011, 50, 6388−6391.
(b) Sen, T. K.; Sau, S. Ch.; Mukherjee, A.; Modak, A.; Mandal, S. K.;
Koley, D. Chem. Commun. 2011, 47, 11972−11974. (c) Kamber, N. E.;
Jeong, W.; Gonzalez, S.; Hedrick, J. L.; Waymouth, R. M.
Macromolecules 2009, 42, 1634−1639. (d) Jeong, W.; Hedrick, J. L.;
Waymouth, R. M. J. Am. Chem. Soc. 2007, 129, 8414−8415.
(7) (a) Raynaud, J.; Absalon, C.; Gnanou, Y.; Taton, D.
Macromolecules 2010, 43, 2814−2823. (b) Raynaud, J.; Ottou, W.
N.; Gnanou, Y.; Taton, D. Chem. Commun. 2010, 46, 3203−3205.
(c) Raynaud, J.; Absalon, C.; Gnanou, Y.; Taton, D. J. Am. Chem. Soc.
2009, 131, 3201−3209.
́ ̌
Dalton Trans. 2010, 39, 6710−6718. (i) Juhari, A.; Mosnacek, J.; Yoon,
J. A.; Nese, A.; Koynov, K.; Kowalewski, T.; Matyjaszewski, K. Polymer
2010, 51, 4806−4813.
(22) Hoffman, H. M. R.; Rabe, J. Angew. Chem., Int. Ed. Engl. 1985,
24, 94−110.
(8) Nederberg, F.; Lohmeijer, B. G. G.; Leibfarth, F.; Pratt, R. C.;
Choi, J.; Dove, A. P.; Waymouth, R. M.; Hedrick, J. L.
Biomacromolecules 2007, 8, 153−160.
(23) (a) Manzer, L. E. ACS Symp. Ser. 2006, 921, 40−51. (b) Manzer,
L. E. Appl. Catal. A: Gen. 2004, 272, 249−256.
(24) Maji, B.; Horn, M.; Mayr, H. Angew. Chem., Int. Ed. 2012, 51,
6231−6235.
́
(9) (a) Rodriguez, M.; Marrot, S.; Kato, T.; Sterin, S.; Fleury, E.;
Baceiredo, A. J. Organomet. Chem. 2007, 692, 705−708. (b) Lohmeijer,
B. G. G.; Dubois, G.; Leibfarth, F.; Pratt, R. C.; Nederberg, F.; Nelson,
A.; Waymouth, R. M.; Wade, C.; Hedrick, J. L. Org. Lett. 2006, 8,
4683−4686.
(25) Knappke, C. E. I.; Arduengo, A. J., III; Jiao, H.; Neudorfl, J.-M.;
̈
Jacobi von Wangelin, A. Synthesis 2011, 3784−3795.
(26) Breslow, R. J. Am. Chem. Soc. 1958, 80, 3719−3726.
(27) Fisher, C.; Smith, S. W.; Powell, D. A.; Fu, G. C. J. Am. Chem.
Soc. 2006, 128, 1472−1473.
(10) (a) Guo, L.; Lahasky, S. H.; Ghale, K.; Zhang, D. J. Am. Chem.
Soc. 2012, 134, 9163−9171. (b) Guo, L.; Zhang, D. J. Am. Chem. Soc.
2009, 131, 18072−18074.
(11) (a) Coutelier, O.; El Ezzi, M.; Destarac, M.; Bonnette, F.; Kato,
T.; Baceiredo, A.; Sivasankarapillai, G.; Gnanou, Y.; Taton, D. Polym.
Chem. 2012, 3, 605−608. (b) Pinaud, J.; Vijayakrishna, K.; Taton, D.;
Gnanou, Y. Macromolecules 2009, 42, 4932−4936. (c) Nyce, G. W.;
Lamboy, J. A.; Connor, E. F.; Waymouth, R. M.; Hedrick, J. L. Org.
Lett. 2002, 4, 3587−3590.
(12) (a) Webster, O. W. Adv. Polym. Sci. 2004, 167, 1−34. (b) Sogah,
D. Y.; Hertler, W. R.; Webster, O. W.; Cohen, G. M. Macromolecules
1987, 20, 1473−1488. (c) Webster, O. W.; Hertler, W. R.; Sogah, D.
Y.; Farnham, W. B.; RajanBabu, T. V. J. Am. Chem. Soc. 1983, 105,
5706−5708.
(28) Enders, D.; Breuer, K.; Teles, J. H.; Ebel, K. J. Prakt. Chem.
1997, 339, 397−399.
(29) DiRocco, D. A.; Oberg, K. M.; Rovis, T. J. Am. Chem. Soc. 2012,
134, 6143−6145.
(30) Berkessel, A.; Elfert, S.; Yatham, V. R.; Neudorfl, M.-M.;
̈
Schlorer, N. E.; Teles, J. H. Angew. Chem., Int. Ed. 2012, 51, 12370−
̈
12374.
(31) Such deoxy-Brewlow intermediates can also be formed from the
reaction of NHC with alkyl halides,25 two of which have also been
structurally characterized later.24
(32) (a) Chen, E. Y.-X. Chem. Rev. 2009, 109, 5157−5214.
(b) Baskaran, D.; Muller, A. H. E. Prog. Polym. Sci. 2007, 32, 173−
̈
(13) Selected recent reviews and examples: (a) Fuchise, K.; Chen, Y.;
Satoh, T.; Kakuchi, T. Polym. Chem. 2013, 4, 4278−4291. (b) Raynaud,
J.; Liu, N.; Gnanou, Y.; Taton, D. Macromolecules 2010, 43, 8853−
8861. (c) Raynaud, J.; Gnanou, Y.; Taton, D. Macromolecules 2009, 42,
5996−6005. (d) Raynaud, J.; Ciolino, A.; Baceiredo, A.; Destarac, M.;
Bonnette, F.; Kato, T.; Gnanou, Y.; Taton, D. Angew. Chem., Int. Ed.
2008, 47, 5390−5393. (e) Scholten, M. D.; Hedrick, J. L.; Waymouth,
R. M. Macromolecules 2008, 41, 7399−7404.
219. (c) Rodriguez-Delgado, A.; Chen, E. Y.-X. J. Am. Chem. Soc. 2005,
̌
127, 961−974. (d) Vlcek, P.; Lochmann, L. Prog. Polym. Sci. 1999, 24,
793−873.
(33) Baskaran, D.; Muller, A. H. E. Macromol. Rapid Commun. 2000,
̈
21, 390−395.
(34) Suenaga, J.; Sutherlin, D. M.; Stille, J. K. Macromolecules 1984,
17, 2913−2916.
(35) Flory, P. J. Principles of Polymer Chemistry, Cornell University
Press: Ithaca, 1953.
(14) Stephan, D. W.; Erker, G. Angew. Chem., Int. Ed. 2010, 49, 46−
76.
(36) (a) Endo, K. Adv. Polym. Sci. 2008, 217, 121−183. (b) Roovers,
J. In Cyclic Polymers, 2nd ed.; Semlyen, J. A., Ed.; Kluwer Academic:
Dordrecht, The Netherlands, 2000; pp 347−384.
(15) (a) Arduengo, A. J., III; Bock, H.; Chen, H.; Denk, M.; Dixon,
D. A.; Green, J. C.; Herrmann, W. A.; Jones, N. L.; Wagner, M.; West,
R. J. Am. Chem. Soc. 1994, 116, 6641−6649. (b) Arduengo, A. J., III;
Dias, H. V. R.; Harlow, R. L.; Kline, M. J. Am. Chem. Soc. 1992, 114,
5530−5534.
(16) (a) Chen, E. Y.-X. Top. Curr. Chem. 2013, 334, 239−260.
(b) Zhang, Y.; Miyake, G. M.; John, M. G.; Falivene, L.; Caporaso, L.;
Cavallo, L.; Chen, E. Y.-X. Dalton Trans. 2012, 41, 9119−9134.
(c) Zhang, Y.; Miyake, G. M.; Chen, E. Y.-X. Angew. Chem., Int. Ed.
2010, 49, 10158−10162.
(17) (a) Enders, D.; Breuer, K.; Kallfass, U.; Balensiefer, T. Synthesis
2003, 1292−1295. (b) Enders, D.; Breuer, K.; Raabe, G.; Runsink, J.;
Teles, J. H.; Melder, J.-P.; Ebel, K.; Brode, S. Angew. Chem., Int. Ed.
1995, 34, 1021−1023.
(18) Maji, B.; Breugst, M.; Mayr, H. Angew. Chem., Int. Ed. 2011, 50,
6915−6919.
(37) Chu, Y.; Deng, H.; Cheng, J.-P. J. Org. Chem. 2007, 72, 7790−
7793.
(38) Kim, Y.-J.; Streitwieser, A. J. Am. Chem. Soc. 2002, 124, 5757−
5761.
(39) Odian, G. Principles of Polymerization, 4th ed.; John Wiley &
Sons, Inc.: Hoboken, NJ, 2004; pp 238−255.
(40) All DFT calculations were performed using the Gaussian09
package. The Supporting Information contains details and references.
(41) (a) Poater, A.; Cosenza, B.; Correa, A.; Giudice, S.; Ragone, F.;
Scarano, V.; Cavallo, L. Eur. J. Inorg. Chem. 2009, 1759−1766.
(b) Poater, A.; Ragone, F.; Giudice, S.; Costabile, C.; Dorta, R.; Nolan,
S. P.; Cavallo, L. Organometallics 2008, 27, 2679−2681. (c) Cavallo,
L.; Correa, A.; Costabile, C.; Jacobsen, H. J. Organomet. Chem. 2005,
690, 5407−5413.
(19) (a) Biju, A. T.; Padmanaban, M.; Wurz, N. E.; Glorius, F. Angew.
Chem., Int. Ed. 2011, 50, 8412−8415. (b) Matsuoka, S.-I.; Ota, Y.;
Washio, A.; Katada, A.; Ichioka, K.; Takagi, K.; Suzuki, M. Org. Lett.
2011, 13, 3722−3725.
(20) Zhang, Y.; Miyake, G. M.; Chen, E. Y.-X. Angew. Chem., Int. Ed.
2012, 51, 2465−2469.
(42) Hawkes, K. J.; Yates, B. F. Eur. J. Org. Chem. 2008, 5563−5570.
(43) During the revision process of the present manuscript, the
experimental mechanistic study of the TPT-catalyzed MMA
dimerization just appeared: Kato, T.; Ota, Y.; Matsuoka, S.-I.;
Takagi, K.; Suzuki, M. J. Org. Chem. 2013, 78, 8739−8747.
17942
dx.doi.org/10.1021/ja4088677 | J. Am. Chem. Soc. 2013, 135, 17925−17942