Square-Planar Cu(II) diketiminate complexes in lactide polymerization

Article abstract of DOI:10.1021/ic402133c

Cu(OiPr)₂ was reacted with several β-diketimine ligands, nacnac^RH. Sterically undemanding ligands with N-benzyl substituents afforded the dimeric heteroleptic complexes [nacnac^BnCu(μ-OiPr)] ₂ and [3-Cl-nacnac^BnCu(μ-OiPr)]₂ (Bn = benzyl). With sterically more demanding amines, dimerization was not possible, and the putative nacnacCuOiPr intermediate underwent ligand exchange to the homoleptic bisdiketiminate complexes Cu(nacnac^ipp)₂ and Cu(nacnac^Naph)₂ (ipp = 2-isopropylphenyl, Naph = 1-napthyl). Homoleptic complexes were also prepared with N-benzyl ligands to yield Cu(nacnac^Bn)₂ and Cu(3-succinimido-nacnac ^Bn)₂. All complexes were characterized by single-crystal X-ray diffraction. Even bulkier ligands with N-anthrylmethyl, N-mesitylmethyl, or N-methylbenzyl substituents failed to react with Cu(OiPr)₂. In the case of nacnac^dippCuOiPr, putative nacnac^dippCuOiPr decomposed by β-hydride elimination. Heteroleptic complexes [nacnac ^BnCu(μ-OiPr)]₂ and [3-Cl-nacnac^BnCu(μ- OiPr)]₂ are very highly active rac-lactide polymerization catalysts, with complete monomer conversion at ambient temperature in solution in 0.5-5 min. In the presence of free alcohol, the homoleptic complexes seem to be in equilibrium with small amounts of the respective heteroleptic complex, which are sufficient to complete polymerization in less than 60 min at room temperature. All catalysts show high control of the polymerization with polydispersities of 1.1 and below. The obtained polymers were essentially atactic, with a slight heterotactic bias at ambient temperature and at -17 C.

Full text of DOI:10.1021/ic402133c

Article
pubs.acs.org/IC
Square-Planar Cu(II) Diketiminate Complexes in Lactide
Polymerization
Todd J. J. Whitehorne and Frank Schaper*
́ ́ ́ ́
Departement de chimie, Universite de Montreal, 2900 Boul. E.-Montpetit, Montreal, Quebec, H3T 1J4, Canada
S
* Supporting Information
ABSTRACT: Cu(OiPr)2 was reacted with several β-diketi-
mine ligands, nacnacRH. Sterically undemanding ligands with
N-benzyl substituents afforded the dimeric heteroleptic
complexes [nacnacBnCu(μ-OiPr)]2 and [3-Cl-nacnacBnCu(μ-
OiPr)]2 (Bn = benzyl). With sterically more demanding
amines, dimerization was not possible, and the putative
nacnacCuOiPr intermediate underwent ligand exchange to
the homoleptic bisdiketiminate complexes Cu(nacnacipp)2 and
Cu(nacnacNaph)2 (ipp = 2-isopropylphenyl, Naph = 1-napthyl). Homoleptic complexes were also prepared with N-benzyl ligands
to yield Cu(nacnacBn)2 and Cu(3-succinimido-nacnacBn)2. All complexes were characterized by single-crystal X-ray diffraction.
Even bulkier ligands with N-anthrylmethyl, N-mesitylmethyl, or N-methylbenzyl substituents failed to react with Cu(OiPr)2. In
the case of nacnacdippCuOiPr, putative nacnacdippCuOiPr decomposed by β-hydride elimination. Heteroleptic complexes
[nacnacBnCu(μ-OiPr)]2 and [3-Cl-nacnacBnCu(μ-OiPr)]2 are very highly active rac-lactide polymerization catalysts, with
complete monomer conversion at ambient temperature in solution in 0.5−5 min. In the presence of free alcohol, the homoleptic
complexes seem to be in equilibrium with small amounts of the respective heteroleptic complex, which are sufficient to complete
polymerization in less than 60 min at room temperature. All catalysts show high control of the polymerization with
polydispersities of 1.1 and below. The obtained polymers were essentially atactic, with a slight heterotactic bias at ambient
temperature and at −17 °C.
reaction, the presence of two important transition states for
insertion and ring-opening of comparable energy,3 the
multitude of possible reaction pathways given the chirality of
the catalyst, the polymer chain and the monomer, and
interfering side reactions, such as chain transfer between
catalyst centers or transesterification. While the existing
literature is already too vast to be easily summarized,4 some
very general trends with regard to the central metal and typical
coordination geometries can be seen: Following Spassky’s
initial work, five- or six-coordinated aluminum complexes can
provide high stereocontrol toward either isotactic or hetero-
tactic monomer enchainment, but suffer from reduced
activities.5 Indium (or gallium) analogues show in general
higher activity. No highly isospecific and active catalyst has yet
been reported,6 but high selectivity for heterotactic enchain-
ment can be achieved.7 Group 3 and rare earth catalysts with a
variety of coordination geometries are among the most active
catalysts known and can yield highly heterotactic to moderately
isotactic polymer.8 Alkaline and earth alkaline compounds,
again with a variety of coordination geometries, show low to
high activities and in rare cases a moderate preference for
isotactic monomer enchainment.9 Group 4 metal catalysts,
mostly of octahedral geometries, can show moderate amounts
of isospecificity, but have in general low to moderate activity.10
INTRODUCTION
■
With the increasing use of plastics, concerns about the
accumulation of plastic debris in the environment favor the
use of polymers that are 100% biodegradable. One such
polymer is polylactic acid (PLA), the condensation polymer of
lactic acid.1 PLA is currently marketed, albeit on a small scale
for commodity polymers. It is biodegradable given the right
conditions, since hydrolysis yields easily metabolized lactic acid.
An additional advantage stems from the fact that lactic acid can
be obtained by fermentation of natural products and does not
dependdirectlyon fossil fuels.2
Industrial production of PLA currently relies on the
polymerization of lactide, the dimeric anhydride of lactic acid,
using a nonselective tin catalyst, in particular Sn(Oct)2, at
elevated temperatures. Although only isotactic PLA is currently
of commercial interest, the use of an unselective catalyst is
possible since lactic acid is obtained from fermentation of
natural products in enantiopure form. The polymerization of
lactide has attracted in the past two decades considerable
attention in the scientific community. This interest is fuelled
partly by the potential of commercial applications, and partly by
the lack of catalysts which combine properties such as correct
stereocontrol, stability, activity, and polymer molecular weight
control.
Establishing clear structure−reactivity relationships for
lactide polymerization catalysts, which would enable rational
catalyst design, is handicapped by the overall reversibility of the
Received: August 20, 2013
Published: November 18, 2013
© 2013 American Chemical Society
13612
dx.doi.org/10.1021/ic402133c | Inorg. Chem. 2013, 52, 13612−13622

Inorganic Chemistry
Article
Figure 1. A: 97%, 24 h, 130 °C, molten monomer, lactide/Cu = 200.14a B: 97%, 8 h, 145 °C, molten monomer, lactide/Cu = 200.14e C: 76%, 4 h,
160 °C, molten monomer, lactide/Cu = 200.14b D: 52%, 24 h, 130 °C, molten monomer, lactide/Cu = 1000.14d E: 80%, 35 h, 70 °C, toluene,
lactide/Cu = 50.14c F: >92%, 6 h, 110 °C, toluene, lactide/Cu = 400.14f G: no polymerization, 160 °C, molten monomer.14b H: 78%, 24 h, 130 °C,
molten monomer, lactide/Cu = 1000.15
Tetrahedral Mg and Zn complexes have been investigated
widely, following the seminal work of Coates et al.,11 and show
moderate to high activities.12 They tend to produce heterotactic
PLA by a chain-end control mechanism with high stereo-
control. Only in rare cases were low isospecificities observed.13
Despite the number of investigations into metal-catalyzed
lactide polymerization, the use of catalysts with square-planar
geometry was somewhat neglected. We are aware only of a
limited number of examples reported for lactide polymerization
based on copper(II)14 and nickel(II) (Figure 1),14b,15 and none
for Cr(II), Pd(II), Pt(II), Rh(I), or Ir(I). The initial literature
survey on copper(II) catalyzed lactide polymerization seems
less than promising. While several catalysts were able to provide
decent molecular weight control, they all required high
temperatures, typically in molten monomer and showed only
low activities even under those conditions.
Scheme 1
A serendipitous choice of the right spectator ligand, however,
showed that square-planar copper complexes can indeed be
interesting candidates for lactide polymerization, and we
recently reported preliminary results that [(nacnacBnCu(μ-
OiPr)]2 is highly active in lactide polymerization, reaching
complete conversion of monomer in less than 1 min at room
temperature (nacnacBn = N,N-dibenzyl-pentane-2,4-diiminato,
lactide/Cu = 300).16 In the present manuscript, we extend
these investigations to other N-alkyl and N-aryl copper(II)
diketiminate complexes.
prepare nacnacdippCuOiPr, 4a, by a protonation route. Upon
addition of the ligand, green Cu(OiPr)2 solubilized and the
color changed to brown, indicating that the ligand did indeed
react with Cu(OiPr)2. However, no isolable product could be
obtained. Analysis of the volatiles of the reaction showed the
presence of acetone. We speculate that 4a was indeed formed
as an intermediate, analogous to the reported three-coordinated
complexes 1 and 2. The latter do not contain β-hydrogen
atoms on the alkoxide ligand and could be isolated. 4a, on the
other hand, undergoes β-H elimination to yield acetone and a
putative Cu(II) hydride which, unsurprisingly, is unstable
(Scheme 1). Direct hydride transfer to the diketiminate ligand
following a Meerwein−Ponndorf−Verley mechanism would
likewise yield acetone,19 but seems less likely in this case.
RESULTS AND DISCUSSION
■
Complex Syntheses. Only two examples of Cu(II)
diketiminate alkoxide or aryloxide complexes have been
reported in the literature. Tolman and co-workers prepared
nacnacdippCuOAr, 1, (dipp = 2,6-diisopropylphenyl, Scheme 1)
from the reaction of nacnacdippCuCl with thallium aryl oxides.17
Warren and co-workers obtained nacnacArCuOtBu, 2, by
oxidation of the respective Cu(I) complex with tert-butyl
peroxide.18 Since aryl oxides, as well as tert-butoxide groups
tend to give slow polymerization initiation, our interest was
focused on more reactive alkoxides, such as isopropoxide. We
thus reacted nacnacdippH, 3a, with Cu(OiPr)2 in an attempt to
13613
dx.doi.org/10.1021/ic402133c | Inorg. Chem. 2013, 52, 13612−13622

Inorganic Chemistry
Article
Alternative routes to nacnacdippCuOR likewise failed: Probably
because of steric hindrance, nacnacdippH failed to react with
Cu(OtBu)2 even at 100 °C (toluene), and only starting
materials were observed. Surprisingly, Cu(OMe)2 was un-
reactive as well. Other bulky diketimine ligands, such as
nacnacAnH (An = 9-antrhylmethyl), nacnacMesH (Mes =
mesitylmethyl), or S,S-nacnacR*H (R* = α-methylbenzyl),
failed to react even with Cu(OiPr)2.
Given the lack of reactivity with sterically encumbered
diketimines, monosubstituted nacnacippH, 3b, (ipp = 2-
isopropylphenyl, Scheme 2) was reacted with Cu(OiPr)2.
a second protonation than insoluble Cu(OiPr)2 (Scheme 2,
path A) or it is thermodynamically labile and undergoes ligand
exchange to 5b and Cu(OiPr)2 without further participation of
3b (Scheme 2, path B). The fact that changing reaction
conditions, that is, order of addition, reagent concentrations, or
temperature, did not affect the reaction outcome already
indicated that 4b might rather be thermodynamically labile.
This was confirmed in the salt metathesis reaction of 7 with
KOtBu, since there is no possible pathway in which 5b can be
obtained as the kinetic product. Heteroleptic Cu(II) alkoxides
with monosubstituted N-aryl substituents thus seem to be
inherently labile toward ligand redistribution to form
homoleptic, four-coordinate complexes. With the disubstituted
N-aryl ligand 3a, formation of a homoleptic bis(diketiminate)
complex (nacnacdipp)2Cu was most likely not possible because
of the increased steric bulk of the disubstituted N-aryl, and
decomposition of the intermediate 4a via β-H elimination is
presumed. A reactivity similar to that of 3b was observed with
nacnacNaphH, 3c (Naph = 1-naphthyl, Scheme 2), which also
yielded the homoleptic complex (nacnacNaph)2Cu, 5c, upon
reaction with Cu(OiPr)2.
Scheme 2
Reaction of Cu(OiPr)2 with nacnacBnH, 3d, finally afforded
the heteroleptic complex (nacnacBnCuOiPr)2, 4d (Scheme 3).16
Scheme 3
However, instead of heteroleptic nacnacippCuOiPr, 4b, the
homoleptic bisdiketiminate complex (nacnacipp)2Cu, 5b, was
obtained, at a variety of reaction conditions. For comparison
purposes, 5b was prepared independently by reaction of 2 equiv
of 3b with Cu(OiPr)2. Attempts to obtain nacnacippCuOtBu by
a route similar to the one used by Warren and co-workers for 2
failed because of the instability of the respective Cu(I) complex
6, which disproportionated into Cu(0) and 5b. An increased
tendency for decomposition by disproportionation has been
observed previously for Cu(I) diketiminate complexes with N-
alkyl ligands, which required stabilization with an ancillary
ligand.20 Monosubstituted N-aryl diketimines seem to suffer
from similar stability problems. Reaction of the lithiated ligand
In Cu(I) chemistry, diketiminate ligands with primary N-alkyl
ligands have been shown to be sterically significantly less bulky
than those with N-aryl substituents.20b,c Given the propensity of
nacnacCuOR to undergo ligand exchange to the homoleptic
complexes, isolation of 4d with a sterically undemanding
spectator ligand was counterintuitive. It seems unlikely that
formation of the homoleptic complex Cu(nacnacBn)2, 5d,
should not be possible with an N-benzyl substituent. Indeed,
5d could not only be prepared from the reaction of 2 equiv of
3d with Cu(OiPr)2, but was also formed in the reaction of 4d
with a second equivalent of 3d (Scheme 3).21 Given the easy
access to 5d, an additional mechanism has thus to be
responsible for the stabilization of 4d. The solid-state structure
of 4d yielded further insights. Contrary to 1 and 2, complex 4d
crystallized as an alkoxide-bridged dimer in the solid state (vide
infra). A UV/vis study of 4d confirmed that the dimeric
structure is most likely retained in solution: UV/vis-spectra of
4d in toluene did not change notably and obeyed the Beer−
Lambert law over a concentration range of 0.050−6.5 mM
17
nacnacippLi with CuCl2 yielded the respective copper(II)
chloride complex, 7, but subsequent reaction with KOtBu did
not yield the heteroleptic complex. Instead homoleptic 5b was
obtained again. The latter reaction provided some indications
with regard to the inaccessibility of 4b. In protonations of
Cu(OiPr)2 with 3b, the putative intermediate 4b might either
be kinetically labile, given that it might be more reactive toward
13614
dx.doi.org/10.1021/ic402133c | Inorg. Chem. 2013, 52, 13612−13622

Inorganic Chemistry
Article
Figure 2. UV/vis spectra of 4d in different solvents or after addition of methyl lactate or lactide, respectively.
Figure 3. X-ray structure of compounds 5b−d and 5f. Hydrogen atoms and disordered atoms in 5c are omitted for clarity. Thermal ellipsoids are
drawn at 50% probability.
(Supporting Information, Figure S1), indicating the absence of
any monomer−dimer equilibria in this concentration range.
Likewise, addition of 0.25−10 equiv of pyridine did not affect
the UV/vis spectra of 4d in toluene (Supporting Information,
Figure S2). Subtle, but more notable changes were observed
between UV/vis spectra of 4d in noncoordinating solvents,
13615
dx.doi.org/10.1021/ic402133c | Inorg. Chem. 2013, 52, 13612−13622

Inorganic Chemistry
Article
a
Table 1. Selected Geometric Data for Homoleptic Complexes 5b, 5c, 5d, and 5f
5b
5c
1.952(3),
5d
5f
nacnacMe2Cu23c,d
nacnac2Cu23
1.93−1.99
Cu−NA
Cu−NB
1.943(1),
2.016(1)
1.947(1),
2.019(1)
0.07
1.946(1),
1.951(1)
1.974(1),
1.976(1)
0.02
1.945(1),
1.958(1)
1.945(1);
1.958(1)
0.01
1.95
1.967(3)
1.909(5)−
1.975(3)
1.95
Δ(Cu−N)
0.02−0.06
92.21(12),
91.9(3),
<0.01
95
0−0.02
95−98
NA−Cu−NA,
NB−Cu−NB,
90.99(6),
91.69(6)
95.21(6),
95.58(6)
94.04(6),
93.31(6)
94.74(16)
98.78(13)−128.5(3)
73, 81
NA−Cu−NB
104.29(6)−147.23(6)
68
101.74(6)−135.41(6)
67
100.82(6)−134.23(6)
62
135
62
98−137
60−67
1−7
b
(NA)2Cu/(NB)2Cu
c
θM
29, 32
15/9/67
syn, syn
13, 25, 31
4/3/82, 8/9/72
syn, anti
7, 9
5, 9
3
Δθx/Δθy/Δθz
orientation N-R
3/5/64
syn, anti
3/0/59
anti, anti
0/1/60
d
a
Δθz given for deviations from square-planar geometry (Scheme 4). NA and NB denote the nitrogen atoms in the first and in the second diketiminate
b
c
ligand, respectively. Angle between the planes formed by the N atoms of each diketiminate and Cu, roughly comparable to θz. Bending of Cu out
of the ligand mean plane, described as the angle between the N2Cu plane and the plane formed by the two nitrogen and the two α-carbon atoms of
the diketiminate ligand. Relative orientation of either the N−CH2-R or the N−Ar substituents.
d
such as dichloromethane or toluene, and coordinating solvents,
such as tetrahydrofuran (THF) or pyridine (Figure 2). Spectra
similar to those obtained in coordinating solvents were
obtained in toluene after addition of 1 equiv of methyl lactate
or 10 equiv of lactide. Based on these UV/vis studies, it appears
that 4d is dimeric in solution in the absence of strong donors
without any noticeable dissociation. In the presence of
intermolecular or intramolecular Lewis bases, the dimer can
break apart to yield a tetracoordinated monomeric species
(Figure 2).
Scheme 4
Since N-benzyl diketimines were the only ligands which
yielded a heteroleptic copper(II) complex, we attempted to
slightly vary the electronic and steric nature of the ligand by
substitution in the 3-position. Reaction of 3d with N-
chlorosuccinimide afforded the expected 3-chlorosubstituted
ligand 3e (Scheme 3). Reaction of the lithiated ligand
nacnacBnLi(THF) with N-bromosuccinimide, on the other
hand, afforded the 3-succinimido substituted ligand 3f, probably
by reaction of the initially formed 3-bromosubstituted ligand
with lithium succinimide.20a Despite their minor differences,
both ligands reacted differently. Reaction of 3e with Cu(OiPr)2
yielded the heteroleptic complex 4e, while reaction of 3f with
Cu(OiPr)2 afforded homoleptic 5f, at least under the
conditions applied here.
Solid State Structures. Crystal structures were obtained
for the homoleptic complexes 5b−5d, and 5f (Figure 3, Table
1). The coordination geometry in (L∧L)2M complexes can be
described using the θ angles introduced by White and co-
workers,22 where θx = θy = θz = 90° represent ideal tetrahedral
coordination and θx = θy = 90°, θz =0° ideal square-planar
geometry (Scheme 4). All bisdiketiminate copper complexes,
Cu(nacnac)2, described here or elsewhere,23 show distorted
geometries with θz ≈ 60° (Table 1). Distortion from ideal
geometry might occur for electronic reasons in tetrahedral
copper complexes because of their d9 electron configuration
and for steric reasons in square-planar complexes because of the
in-plane interactions of the N-substituents. (Pseudo)tetrahedral
Cu(II) complexes are not uncommon and the fact that the
value of θz does not correlate with the steric demand of the N-
substituent would argue to assign the coordination geometries
as distorted tetrahedral. However, complexes electronically very
similar to Cu(nacnac)2, that is, Cu(acac)2 and Cu(acnac)2
(acnac = 4-imino-penta-2-nonate), show square-planar geo-
metries. Closer inspection of the structure of the sterically least
encumbered bisdiketiminate complex, Cu(nacnacMe)2,23c,d
shows that θz ≈ 60° actually represents the closest sterically
possible approach to square-planar symmetry, even for N-
substituents as small as methyl. Coordination geometries in
5b−d and 5f (Figure 3) are thus best described as distorted
square-planar, even though the θz-values are closer to a
tetrahedral geometry. In 5b, steric demands of the N-
isopropylphenyl substituent lead to further distortion from
ideal geometry. Although some of the steric strain is released
(as often encountered in sterically demanding diketiminate
complexes) by a strong bending of the Cu metal out of the
mean plane of the diketiminate ligand (θM ≈ 30°, Scheme 4,
Table 1), coordination around Cu in 5b is highly unsym-
metrical with differences in Cu−N bond lengths of 0.07 Å, Δθx
= 15° and Δθy = 9° (Table 1). The N-naphthyl substituent in
5c likewise introduces strong distortions. Interligand π-stacking
between naphthyl substituents and between naphthyl and
diketiminate is observed in 5c, which was absent in 5b probably
because of the isopropyl substituent. While θz values remain
≈60°, reduction of the steric impact of the N-substituent
reduces additional distortions caused by steric interactions.
Complexes 5d and 5f thus display only small differences in
̆
Cu−N bond lengths (Δ = 0.01−0.02 A), a much smaller
bending of the Cu atom out of the ligand mean plane (θM = 5−
9°), and a symmetry much closer to the (crystallographic) C2
symmetry observed for Cu(nacnacMe)2 (Δθx < 3°, Δθy < 5°).
The alkoxide complexes 4d and 4e crystallize as oxygen-
bridged dimers (Figure 4). Their coordination geometry is best
13616
dx.doi.org/10.1021/ic402133c | Inorg. Chem. 2013, 52, 13612−13622

Inorganic Chemistry
Article
Figure 4. X-ray structure of 4d16 (left) and 4e (right). Hydrogen atoms are omitted for clarity. Thermal ellipsoids are drawn at 50% probability.
Table 2. Selected Geometric Data for Heteroleptic Alkoxide Complexes 4d,16 and 4e
4d16
4e
nacnacdiipCuOAr17
1.86−1.90
nacnacArCuOtBu18a
1.88−1.89
Cu−N
Cu−O
1.921(3), 1.951(3);
1.923(3), 1.946(3)
1.951(2), 1.979(2);
1.954(2), 1.982(2)
95.19(12); 94.92(11)
96.14(11)−152.75(11)
41, 43
1.917(3), 1.932(2)
1.948(2), 1.977(2)
1.75−1.82
1.78−1.79
N−Cu−N
94.63(10)
96−97
129−135
96
N−Cu−O
N2Cu/Cu2O2
98.46(9)−149.23(9)
118−146
a
45
b
θM
13, 14
15
0−6
1−10
Δθx/Δθy/Δθz
orientation N-R
2/4/35
0/2/40
anti
anti
a
b
Angle between the planes formed by the N atoms of the diketiminate and Cu and by the Cu2(μ-O)2 core. Bending of the Cu out of the ligand
mean plane, described as the angle between the N2Cu plane and the plane formed by the two nitrogen and two α-carbon atoms of the diketiminate
c
ligand. Relative orientation of the N−CH2-R substituents.
described as distorted square-planar, with a distortion from
planarity (Δθz = 50−55) because of unfavorable steric
interactions of the alkoxide ligands with the N-substituents.
Cu−N distances are slightly longer than those in monomeric
nacnacdiipCuOAr17 or nacnacArCuOtBu18a (Table 2) and the
bending of Cu out of the ligand mean plane (θM, Table 2) is
slightly more pronounced. Introduction of a chloride
substituent in the 3-position of the ligand has barely noticeable
steric consequences. Interaction between the chloride sub-
stituent and the ligand methyl groups, leads to a slight (1°)
widening of the Cβ-Cα-CMe angles in the ligand (when
compared to 4d), which in turn very slightly increases the
steric pressure in the front of the complex. Even this minor
effect seems to be absent in the case of a succinimido
substituent since structures 5d and 5f are practically identical.
Lactide Polymerization. Performance in rac-lactide
polymerization was first investigated using heteroleptic 4d
and 4e (Table 3). Preliminary results on the surprisingly high
activity of 4d (Table 3), given the generally low activity
reported for copper(II) complexes in lactide polymerizatio-
n,14a−e have been reported recently.16 Structurally very similar
4e was also highly active in lactide polymerization. With 1 mM
4d or 4e in dichloromethane at room temperature, polymer-
izations reached completion in <1 or 3 min, respectively.
Polymerizations with both catalysts were highly controlled and
narrow polydispersities below 1.1 were obtained, indicating fast
1
activation and the absence of side reactions. H NMR spectra
showed signals at the position and with the intensities expected
for an isopropoxy end group. Intensities were too low, however,
to provide a reliable polymer molecular weight estimate from
NMR spectroscopy. The obtained polymers were atactic, with a
slight heterotactic bias (Pr = 0.56 for 4d, 0.53 for 4e. Pr =
probability of alternating monomer insertion). Tetrahedral
magnesium alkoxide complexes, carrying the same ligands as 4d
and 4e, showed a slight isotactic preference for lactide
polymerization at low temperatures.13c,24 In the case of
square-planar 4d and 4e, polymerizations at −17 °C led only
to a slight increase of the heterotactic bias. Despite very narrow
polydispersities, the observed polymer molecular weight
sometimes differed from expectations (e.g., Table 3, #1).
Although experimental errors, such as weighing errors and
catalyst decomposition, are more likely to be responsible for the
13617
dx.doi.org/10.1021/ic402133c | Inorg. Chem. 2013, 52, 13612−13622

Inorganic Chemistry
Article
Table 3. Polymerization of rac-Lactide with Heteroleptic and Homoleptic Copper Complexes
Cu: lactide (:
[Cu]/
a
Mn (exp.)/
a
b
#
catalyst
4d
BnOH)
mM
conversion time (min)
Pr
Mn mol/g
Mn
Mw/Mn
kobs min
d
1
1:300
2
90−98%
1−3
0.56−
27 400−62 300
0.7−1.5
1.04−
3.6(4)−4.4(5)
0.57
1.07
b
2
4d, −17 °C
1:300
2
80−98%
30
0.60−
37 400
1.1
1.06
0.61
3
4
4d
4e
1:300:1 (3d)
2
2
36%
90
0.64
53 300
0.5
1.4
1.06
1.04
5.1(1) × 10−3
0.8(1)−1.3(2)
b
1:300
95−98%
3−11
0.53−
29 400
0.54
b
5
4e, −17 °C
1:300
1:300
2
95−98%
60−175
0.53−
30 300
1.4
1.04
0.54
6
7
8
9
5d
5d
5d
5d
5d
5b
5b
5b
5b
5c
5c
5f
2
2
2
2
2
2
2
2
2
2
2
2
2
10%
95%
95%
75%
97%
80%
98%
95%
98%
7%
10
45
40
40
125
85
60
60
60
60
60
60
60
≈10−2
c
1:300:1
0.60
0.59
0.60
0.58
0.63
0.59
0.60
0.57
26 400
29 900
41 600
46 400
487 700
30 500
35 300
1400
1.6
1.04
1.03
1.04
1.03
1.12
1.48
1.07
1.12
1.06
1.08
1.14
1.53
11.2(1) × 10−2
9.6(2) × 10−2
3.8(1) × 10−2
4.4(1) × 10−2
1:300:1
1.4
1:300:1 (iPrOH)
1:300:1 (iPrOH)
1:300
0.8
10
11
12
13
14
15
16
17
18
0.9
0.07
1.4
c
1:300:1
12.0(4) × 10−2
6.7(3) × 10−2
17.2(9) × 10−2
1:300:1
1:300:25
1:300
1.2
1.3
56 000
4300
0.06
4.8
1:300:1
1:300
95%
10%
95%
0.52
6.4(3) × 10−2
79 300
4300
0.06
4.8
5f
1:300:1
0.57
12.7(5) × 10−2
If a range is provided: minimum and maximum values of three experiments. Conditions: CH2Cl2, ambient temperature. Pr determined from
1
decoupled H NMR by Pr = 2·I1/(I1+I2), with I1 = 5.20−5.25 ppm (rmr, mmr/rmm), I2 = 5.13−5.20 ppm (mmr/rmm, mmm, mrm). Mn and Mw
a
determined by size exclusion chromatography vs polystyrene standards, with a Mark−Houwink correction factor of 0.58. Concentrations and ratios
b
provided per Cu atom, that is, for (putative) monomeric complexes. GPC analysis was performed only on selected examples. Mn(expected) =
c
d
[lactide]/([Cu] + [ROH])·conversion·Mlactide + MROH. Alcohol added after 10 min polymerization time. Values taken from ref 16.
observed discrepancies than transesterification reactions,
changes in Pr over time were investigated. (Because of the
overlap of rr-triads, formed during transesterification, with mm-
triads, transesterification results in an artificially low Pr value if
the latter is determined from homonuclear decoupled NMR.)
Thus, rac-lactide was polymerized with 4d (>95% conversion
after 1 min) and kept for 12 h under polymerization conditions
to allow transesterification of the formed polymer. Analysis of a
polymer sample showed that the Pr value of the polymer
decreased slightly, indicating a very low amount of trans-
esterification over a time period 1000× longer than the
polymerization time. In a similar experiment, S,S-lactide was
polymerized by 4d to isotactic PLA and kept for 12 h without
quenching under polymerization conditions. No change in the
Pr value was observed. 4d thus does not catalyze polymer
epimerization. In both cases, addition of additional monomer
confirmed that 4d remained active.16
polymerization occurs under immortal polymerization con-
ditions catalyzed by about 2% of heteroleptic complex present
in the equilibrium under these conditions (Scheme 5). The
latter mechanism is supported by the fact that 4d is stable
under immortal polymerization conditions,16 that UV/vis
titrations of 4d with ligand 3d yield the homoleptic complex
5d, and that addition of 1 equiv of 3d to rac-lactide
polymerizations with 4d reduced activities drastically by 3
orders of magnitude (Table 3, #3). If isopropanol instead of
benzyl alcohol is employed as cocatalyst, a notably longer
induction period and an overall lower activity was observed
(Table 3, Supporting Information, Figure S3). While the
induction period can be explained by the lower acidity of
isopropanol compared to benzyl alcohol, activity is governed by
the acidity of the polymeryl alcohol and should be independent
of the starting alcohol, once all catalyst initiated. We have at the
moment no explanation for this behavior.
Given that only homoleptic diketiminate complexes were
obtained with ligands 3b, 3c, and 3f, their activity in rac-lactide
polymerization was verified. Polymerization with 5d under
conditions identical to 4d yielded, unsurprisingly, a very low
activity because of the absence of a suitable initiator group, and
only 10% conversion was observed after 10 min (Table 4,
Supporting Information, Figure S3). Addition of benzyl alcohol
increased the activity by an order of magnitude. The
polymerization was completely controlled with polydispersities
below 1.05 and showed the molecular weight expected from the
amount of added alcohol. The pseudo-first order rate constant
kobs = 9 × 10−2 min−1 was, however, 50× lower than the rate
observed for the heteroleptic complex 4d, although the same
active species should have been obtained after the first
monomer insertion. We thus assume that protonation of a
diketiminate ligand by benzyl alcohol is unfavorable and that
Polymerizations with homoleptic 5b in the absence of any
cocatalyst again proceeded sluggishly. A positive curvature of
the conversion/time plot up to 60 min (Supporting
Information, Figure S4) and a polymer weight of 500 000 g/
mol indicate slow activation of small amounts of 5b, most likely
by impurities in the monomer. It should be noted that although
only 5% of 5b were active in polymerization with an extremely
slow initiation, and although polymerization had to be
quenched at 80% when the solution became too viscous to
stir, the obtained polymer still showed a very narrow
polydispersity of 1.12. Addition of benzyl alcohol as a cocatalyst
reduced the induction period to about 10 min and yielded an
activity approximately half as high as that of 5d. In the presence
of 25 equiv of benzyl alcohol, the induction period disappeared
and activity increased by a factor of 2−3, in agreement with the
mechanism proposed in Scheme 5.
13618
dx.doi.org/10.1021/ic402133c | Inorg. Chem. 2013, 52, 13612−13622

Inorganic Chemistry
Article
Table 4. Details of X-ray Diffraction Studies
4e
5b
5c
5d
5f
formula
C44H54Cl2Cu2N4O2
868.89; 1.400
150; 908
C46H58CuN4
730.50; 1.194
100; 1564
monoclinic
P21/c
C50H42CuN4
762.41; 1.291
100; 798
C38H42CuN4
618.29; 1.302
100; 654
C46H48CuN6O4·CH2Cl2
897.37;
Mw (g/mol); dcalcd. (g/cm3)
T (K); F(000)
crystal system
space group
150; 1876
monoclinic
P21/n
triclinic
triclinic
monoclinic
P21/c
P1
̅
P1
̅
unit cell:
a (Å)
13.6092(14)
10.2310(10)
14.8126(16)
16.6129(5)
10.4872(3)
23.5375(7)
12.0601(13)
12.1002(13)
16.0178(18)
78.080(4)
75.677(4)
60.513(5)
1961.5(4); Z
1.089; multiscan
3−72; 0.97
54014; 0.026
7430; 0.048
0.076
9.5454(6)
11.3169(7)
15.0776(10)
93.248(3)
103.821(3)
92.335(3)
1576.57(18); 2
1.219; multiscan
3−70; 0.98
35571; 0.028
5844; 0.043
0.037
11.2594(4)
17.7065(7)
21.5097(8)
b (Å)
c (Å)
α (deg)
β (deg)
91.846(5)
97.799(2)
91.886(2)
γ (deg)
V (Å3); Z
2061.4(4); 2
2.786; multiscan
4−70; 0.99
39607; 0.023
3895; 0.059
0.047
4062.8(2); 4
1.018; multiscan
3−71; 0.99
58938; 0.026
7830; 0.046
0.037
4285.9(3); 4
2.283; multiscan
3−70; 1.00
100806; 0.063
8180;
μ (mm−1); abs. corr.
θ range (deg); completeness
collected reflections; Rσ
unique reflections; Rint
R1(F) (I > 2σ(I))
wR(F2) (all data)
GoF(F2)
0.036
0.145
0.103
0.1948
0.102
0.101
1.04
1.03
1.04
1.06
1.04
residual electron density
0.35; −0.82
0.44; −0.40
1.25; −1.18
0.71; −0.36
0.45; −0.39
Scheme 5
Figure 5. Conversion−time plots for rac-lactide polymerization with
homoleptic 5b−d and 5f.
acidity of N-aryl diketimines compared to N-alkyl diketimines.
All polymers obtained with 5b−d, and 5f were essentially
atactic with a slight heterotactic bias (Pr = 0.52−0.60). The Pr-
value obtained for 5d/BnOH (Pr = 0.59−0.60) is 3% higher
than for 4d, although the active species should be identical in
both polymerizations. This might simply be due to
experimental error. However, we have noted a similar behavior
in diketiminate magnesium complexes, which showed a slightly
decreased Pr value (change <5%) in the presence of excess
alcohol.24
Homoleptic 5c and 5f likewise showed very low polymer-
ization activity from a small number of catalyst centers in the
absence of benzyl alcohol (Table 4), but became moderately
active in its presence (Table 4, Figure 5). Both catalysts show
unexpectedly low polymer molecular weight in the presence of
benzyl alcohol, and 5f was the first Cu catalyst presented here
which gave polydispersities above 1.1. Comparison of the
conversion/time data in Figure 5 reveals that the structurally
similar N-benzyl complexes 5d and 5f also have very similar
activities. Apart from the loss of polymer molecular weight
control, there seems thus to be no noticeable impact of the
succinimido substituent on polymerization. Complexes 5b and
5c with mono-ortho-substituted N-aryl substituents also show
very similar activity, about half as high as the N-alkyl
derivatives. Contrary to 5d and 5f, they show the presence of
an induction period. Both, delayed initiation as well as lower
activity, would be expected in 5b and 5c because of the higher
CONCLUSIONS
■
The chemistry of Cu(II) diketiminate alkoxide complexes is
governed by their strong tendency to achieve four-coordination
in square-planar geometry on one hand and the steric
constraints of the diketiminate ligand on the other. In the
case of sterically undemanding N-alkyl diketimines, the
monomeric intermediate nacnacCuOiPr, obtained upon re-
13619
dx.doi.org/10.1021/ic402133c | Inorg. Chem. 2013, 52, 13612−13622

Inorganic Chemistry
Article
N,N′-dibenzyl-2-amino-3-chloro-4-imino-2-pentene, Clnac-
nacBnH, 3e. To a solution of 3d (5.48 g, 19.7 mmol) in dry THF
(150 mL) was added N-chlorosuccinimide (3.00 g, 22.4 mmol). After
stirring at room temperature for 45 min, a white precipitate formed
which was removed by filtration. H2O (500 mL) was then added. The
product was extracted using hexanes (2 × 600 mL). After drying over
Na2SO4 the solvent was evaporated. The obtained yellow oil was
crystallized from dry ethanol at −80 °C, washed with cold dry ethanol
and recrystallized from refluxing ethanol. The eluate yielded a second
fraction at −80 °C (colorless crystals, 3.41 g, 55%).
action of Cu(OiPr)2 with diketimine, stabilizes via simple
dimerization. Sterically more demanding mono-ortho-substi-
tuted N-aryl substituents, such as 9-naphthyl nor 2-
isopropylphenyl, do not yield a dimeric heteroleptic complex.
Instead, homoleptic Cu(nacnac)2 is obtained, which seems to
be the thermodynamic rather than the kinetic product. Further
increase in steric bulk, that is, the use of 2,6-disubstituted N-aryl
substituents such as the ubiquitous 2,6-diisopropylphenyl,
prevents the formation of the homoleptic complex for steric
reasons. In the absence of β-hydrogen atoms on the alkoxide
ligand, the three-coordinated intermediate can then be isolated
as the reaction product. If β-hydrogen atoms are present,
decomposition, presumably via β-H-elimination, will yield the
respective ketone or aldehyde and unknown Cu-containing
products.
1H NMR (CDCl3, 400 MHz, 298 K): δ 12.22 (bs, 1H, NH), 7.25−
7.19 (m, 10H, Ph), 4.49 (s, 4H, NCH2), 2.18 (s, 6H, Me). 13C{1H}
NMR (CDCl3, 75 MHz, 298 K): δ 160.3 (CN), 140.5 (ipso Ph),
128.6 (ortho or meta Ph), 127.3 (ortho or meta Ph), 126.8 (para Ph),
127.7 (ClC), 51.5 (NCH2), 17.3 (Me). Anal. Calcd. for C19H21ClN2:
C, 72.95; H, 6.77; N, 8.95. Found: C, 72.89; H, 6.66; N, 9.17.
N,N′-dibenzyl-2-amino-3-succinimido-4-imino-2-pentene,
3-succinimido-nacnacBnH, 3f. N-Bromosuccinimide (224 mg, 1.26
mmol) and nacnacBnLi(THF) (440 mg, 1.24 mmol) were suspended in
THF (30 mL) and heated at 60 °C for 24 h to afford a brown solution.
1,4-Dioxane (1 mL) was added to precipitate LiCl. After additional
stirring for 15 min at room temperature, the mixture was filtered and
the filtrate evaporated. The resulting brown residue was dissolved in
dichloromethane (3 mL). Hexane (12 mL) was added, the mixture
filtered, and the filtrate allowed to slowly evaporate, yielding dark-
Heteroleptic (XnacnacBnCuOiPr)2 showed activities in lactide
polymerization comparable to the best catalysts reported. The
catalysts show high molecular weight control with polydisper-
sities typically below 1.1 even under immortal polymerization
conditions and do not suffer from side reactions such as
transesterification, epimerization, or (undesired) chain transfer.
Combined with their high activity, they are promising
candidates for the production of block-copolymers, and we
are currently investigating the scope of these catalysts with
regards to different monomers.
Homoleptic Cu(nacnac)2 can be activated by alcohol as a
cocatalyst to yield moderately active polymerization catalysts (1
h to completion at room temperature) which retain the high
polymer molecular weight control of the heteroleptic
complexes. The easy accessibility of Cu bisdiketiminate
complexes drastically simplifies synthetic requirements. Since
only a small part of the homoleptic complex is activated for
polymerization, polymer molecular weight is determined from
the monomer:alcohol ratio, and small amounts of catalyst
decomposition by impurities should be tolerated without
impact on the obtained polymer molecular weight. These
catalysts do not contain an initiating group and are thus suitable
catalysts for the polymerization of macroinitiators.
1
yellow crystals (250 mg, 54%). H NMR (CD3CN), 400 MHz, 298
K): δ 12.57 (bs, 1H, NH), 7.28−7.19 (m, 10H, Ph), 4.47 (s, 4H,
CH2), 2.76 (s, 4H, NCH2), 1.76 (s, 6H, CH3). 13C{1H} NMR
(CD3CN, 101 MHz, 298 K): δ 179.3 (CO), 161.6 (CN), 141.5
(ipso Ph), 129.5 (ortho or meta Ph), 128.4 (ortho or meta Ph), 127.7
(para Ph), 98.9 (NC), 51.5 (NCH2), 28.8 (CH2), 14.8 (Me). Anal.
Calcd. for C23H25N3O2: C, 73.57; H, 6.71; N, 11.19. Found: C, 73.28;
H, 6.69; N 10.84.
[nacnacBnCu(μ-O)iPr]2, 4d.16 Cu(OiPr)2 (500 mg, 2.75 mmol)
was suspended in toluene (20 mL). 3d (610 mg, 2.20 mmol) was
added slowly to the mixture and allowed to react at room temperature
for 18 h. The reaction mixture changed color from deep green to deep
blue. Solvent and isopropanol were removed under reduced pressure.
The blue solid was taken up in toluene (15 mL) and filtered through a
fine frit. The filtrate was placed at −25 °C yielding purple-red crystals
(550 mg, 63%). Anal. Calcd. for C22H28CuN2O: C, 66.06; H, 7.05; N,
7.00. Found: C, 66.33; H, 7.22; N 6.82. UV/vis (toluene): λ/nm (ε·M·
cm) 357 (10000), 431 (570), 523 (560), 720 (sh, 100). UV/vis
(THF): λ/nm (ε·M·cm) 421 (sh 470), 533 (400).
EXPERIMENTAL SECTION
[nacnacBnCu(μ-O)iPr]2, 4e. Using the same procedure as for 4d.
Cu(OiPr)2 (500 mg, 2.75 mmol), hexanes or toluene (20 mL), 3e
(690 mg, 220 mmol) yielded 350 mg (0.80 mmol, 36%) of purple-red
crystals.
Anal. Calcd. for C22H27ClCuN2O: C, 60.82; H, 6.26; N, 6.45.
Found: C 60.76, H 6.17, N 6.34.
■
General Considerations. All reactions were carried out using
Schlenk or glovebox techniques under nitrogen atmosphere. Cu-
(OiPr)2,25 nacnacdippH (3a),26 nacnacippH (3b),27 nacnacNaphH (3c),28
and nacnacBnH (3d)29 were prepared according to literature. Solvents
were dried by passage through activated aluminum oxide (MBraun
SPS), deoxygenated by repeated extraction with nitrogen, and stored
over molecular sieves. C6D6 was dried over sodium and degassed by
three freeze−pump−thaw cycles. CDCl3 and CD2Cl2 were dried over
3 Å molecular sieves. rac-Lactide (98%) was purchased from Sigma−
Aldrich, purified by 3× recrystallization from dry ethyl acetate and kept
at −30 °C. All other chemicals were purchased from common
Cu(nacnacipp)2, 5b. Diketimine 3b (2.00 g, 5.98 mmol) was added
to a heterogeneous solution of Cu(OiPr)2 (0.54g, 2.98 mmol) in
toluene (20 mL) and allowed to stir overnight. The solvent was then
removed in vacuo from the dark green solution, giving a dark green
waxy solid. The mixture was dissolved in a minimum of hexanes and
filtered over Celite. Pure product was isolated by crystallization at −35
°C to yield purple crystals (0.68 g, 0.93 mmol, 63%). Mp: 154 °C.
Anal. Calcd. for C46H58CuN4: C, 75.63; H, 8.00; N, 7.67. Found: C,
75.47; H, 8.02; N, 7.47. UV/vis (toluene): λ/nm (ε·M·cm): 440 (sh,
1400), 460 (1300).
1
commercial suppliers and used without further purification. H and
13C NMR spectra were acquired on a Bruker AVX 400 spectrometer.
The chemical shifts were referenced to the residual signals of the
deuterated solvents (C6D6: 1H: δ 7.16 ppm, 13C: δ 128.38 ppm,
CDCl3: 1H: δ 7.26 ppm, CD2Cl2: 1H: δ 5.32 ppm, CD3CN: 1H: δ 1.94
ppm, 13C: δ 1.32 ppm). Elemental analyses were performed by the
Cu(nacnacNaph)2, 5c. Following the same procedure as for 5b, 3c
(0.82 g, 1.1 mmol), Cu(OiPr)2 (0.10 g, 0.54 mmol), toluene (10 mL)
gave a dark green waxy solid. The mixture was dissolved in a minimum
of dichloromethane and filtered over Celite. Crystallization by slow
evaporation yielded green crystals (0.21 g, 0.28 mmol, 51%). Mp: 149
°C. Anal. Calcd. for C50H42CuN4: C, 78.76; H, 5.55; N, 7.35. Found:
C, 78.79; H, 5.63; N, 7.33. UV/vis (toluene): λ/nm (ε·M·cm) 537
(1000), 670 (1200).
́ ́ ́ ́
Laboratoire d’analyze elementaire (Universite de Montreal). Molecular
weight analyses were performed on a Waters 1525 gel permeation
chromatograph equipped with three Phenomenex columns and a
refractive index detector at 35 °C. THF was used as the eluent at a
flow rate of 1.0 mL·min−1 and polystyrene standards (Sigma−Aldrich,
1.5 mg·mL−1, prepared and filtered (0.2 mm) directly prior to
injection) were used for calibration. Obtained molecular weights were
corrected by a Mark−Houwink factor of 0.58.30
Cu(nacnacBn)2, 5d. Following the same procedure as for 5b, 3d
(0.70 g, 2.5 mmol), Cu(OiPr)2 (0.25 g, 1.2 mmol), toluene (20 mL)
13620
dx.doi.org/10.1021/ic402133c | Inorg. Chem. 2013, 52, 13612−13622

Inorganic Chemistry
Article
gave a purple wax-like solid. The mixture was dissolved in a minimum
of hexanes and filtered over Celite. 5d was isolated by crystallization at
−35 °C as purple crystals (0.98 g, 0.158 mmol, 64%). Mp: 62 °C.
Anal. Calcd. for C38H42CuN4: C, 73.81; H, 6.85; N, 9.06. Found: C,
73.74; H, 6.81; N, 9.02. UV/vis (toluene): λ/nm (ε·M·cm) 446 (550),
546 (1500).
R. E. Prud’homme for access to GPC and Elena Nadezhina for
elemental analyses.
REFERENCES
■
(1) (a) Luckachan, G. E.; Pillai, C. K. S. J. Polym. Environ. 2011, 19,
637. (b) Ahmed, J.; Varshney, S. K. Int. J. Food Prop. 2011, 14, 37.
Cu(3-succinimido-nacnacBn)2, 5f. Following the same procedure
as for 5b, 3f (0.32 g, 0.74 mmol), Cu(OiPr)2 (65 mg, 0.36 mmol),
toluene (10 mL) gave a dark-purple waxy solid. The solid was
dissolved in a minimum of dichloromethane and filtered over Celite.
Crystallization by slow evaporation yielded 0.30 g (0.74 mmol, 87%)
of purple crystals. Mp: 174 °C. Anal. Calcd. for C46H48CuN6O4·
CH2Cl2: C, 62.90; H, 5.62; N, 9.36. Found: C, 62.06; H, 5.71; N, 8.96.
(One equivalent of dichloromethane was found in the X-ray structure.)
UV/vis (toluene): λ/nm (ε·M·cm) 432 (620), 531 (1300), 726 (sh,
400).
(c) Inkinen, S.; Hakkarainen, M.; Albertsson, A.-C.; Sodergard, A.
̈
̊
Biomacromolecules 2011, 12, 523.
(2) Hottle, T. A.; Bilec, M. M.; Landis, A. E. Polym. Degrad. Stab.
2013, 98, 1898.
(3) Marshall, E. L.; Gibson, V. C.; Rzepa, H. S. J. Am. Chem. Soc.
2005, 127, 6048.
(4) (a) Dutta, S.; Hung, W.-C.; Huang, B.-H.; Lin, C.-C. In Synthetic
Biodegradable Polymers; Rieger, B., Kunkel, A., Coates, G. W.,
̈
Reichardt, R., Dinjus, E., Zevaco, T. A., Eds.; Springer-Verlag: Berlin,
Germany, 2011; pp 219. (b) Dijkstra, P. J.; Du, H.; Feijen, J. Polym.
Chem. 2011, 2, 520. (c) Ajellal, N.; Carpentier, J.-F.; Guillaume, C.;
Guillaume, S. M.; Helou, M.; Poirier, V.; Sarazin, Y.; Trifonov, A.
Dalton Trans. 2010, 39, 8363. (d) Stanford, M. J.; Dove, A. P. Chem.
Soc. Rev. 2010, 39, 486. (e) Williams, C. K.; Hillmyer, M. A. Polym.
Rev. 2008, 48, 1. (f) O’Keefe, B. J.; Hillmyer, M. A.; Tolman, W. B. J.
Chem. Soc., Dalton Trans. 2001, 2215.
rac-Lactide Polymerization. In the glovebox a stock solution of
the catalyst (100 μL, 5.0 × 10−2 M in CH2Cl2, 5.0 μmol) was added to
lactide (220 mg, 1.5 mmol) in dichloromethane (2.5 mL). If desired,
benzyl alcohol (5.0 × 10−2 M in CH2Cl2) was added to the reaction
mixture. Samples for kinetic investigations were taken at the desired
intervals and added to vials already containing a dichloromethane
solution of acetic acid (5 mM). Reaction mixtures were quenched at
the desired polymerization time by addition of a dichloromethane
solution of acetic acid (5 mM). For samples as well as the bulk
reaction, volatiles were immediately evaporated. Solid polymer samples
were stored at −80 °C. Conversion was determined from 1H NMR in
CDCl3 by comparison to remaining lactide. Pr values were determined
(5) (a) Spassky, N.; Wisniewski, M.; Pluta, C.; Le Borgne, A.
Macromol. Chem. Phys. 1996, 197, 2627. (b) Zhong, Z.; Dijkstra, P. J.;
Feijen, J. Angew. Chem., Int. Ed. 2002, 41, 4510. (c) Nomura, N.; Ishii,
R.; Akakura, M.; Aoi, K. J. Am. Chem. Soc. 2002, 124, 5938.
(d) Hormnirun, P.; Marshall, E. L.; Gibson, V. C.; White, A. J. P.;
Williams, D. J. J. Am. Chem. Soc. 2004, 126, 2688. (e) Chisholm, M.
H.; Patmore, N. J.; Zhou, Z. Chem. Commun. (Cambridge, U. K.) 2005,
127. (f) Chisholm, M. H.; Gallucci, J. C.; Quisenberry, K. T.; Zhou, Z.
Inorg. Chem. 2008, 47, 2613.
1
from homodecoupled H NMR spectra.
X-ray Diffraction. Single crystals were obtained directly from
isolation of the products as described above. Diffraction data were
collected with Cu Kα radiation on Bruker Microstar/Proteum,
equipped with Helios mirror optics and rotating anode source or on
a Bruker APEXII with a Cu microsource/Quazar MX optics using the
APEX2 software package.31 Data reduction was performed with
SAINT,32 absorption corrections with SADABS.33 Structures were
solved with direct methods (SHELXS97). All non-hydrogen atoms
were refined anisotropic using full-matrix least-squares on F2 and
hydrogen atoms refined with fixed isotropic U using a riding model
(SHELXL97).34 In 5f, cocrystallized dichloromethane was found to be
disordered and refined with appropriate restraints (0.7:0.3 occupancy).
In 5c, one diketiminate ligand was found disordered with N-naphtyl
orientations inverted by 180 °C. The disorder was resolved using
appropriate restraints (SIMU/SADI) and refined to 0.7:0.3
occupation. Additional fluxionality exists and electron density indicates
even further N-napthyl rotamers, which were not resolved. Further
experimental details can be found in Table 4 and in the Supporting
Information (CIF).
(6) Dagorne, S.; Normand, M.; Kirillov, E.; Carpentier, J.-F. Coord.
Chem. Rev. 2013, 257, 1869.
(7) Pietrangelo, A.; Hillmyer, M. A.; Tolman, W. B. Chem. Commun.
(Cambridge, U. K.) 2009, 2736.
(8) (a) Cao, T.-P.-A.; Buchard, A.; Le Goff, X. F.; Auffrant, A.;
Williams, C. K. Inorg. Chem. 2012, 51, 2157. (b) Thomas, C. M. Chem.
Soc. Rev. 2010, 39, 165.
(9) (a) Cushion, M. G.; Mountford, P. Chem. Commun. (Cambridge,
U. K.) 2011, 47, 2276. (b) Calvo, B.; Davidson, M. G.; Garcia-Vivo, D.
Inorg. Chem. 2011, 50, 3589.
(10) Sauer, A.; Kapelski, A.; Fliedel, C.; Dagorne, S.; Kol, M.; Okuda,
J. Dalton Trans. 2013, 42, 9007.
(11) Cheng, M.; Attygalle, A. B.; Lobkovsky, E. B.; Coates, G. W. J.
Am. Chem. Soc. 1999, 121, 11583.
(12) (a) Wheaton, C. A.; Hayes, P. G. Comments Inorg. Chem. 2011,
32, 127. (b) Wheaton, C. A.; Hayes, P. G.; Ireland, B. J. Dalton Trans.
2009, 4832.
(13) (a) Wang, L.; Ma, H. Macromolecules 2010, 43, 6535. (b) Buffet,
J.-C.; Davin, J. P.; Spaniol, T. P.; Okuda, J. New J. Chem. 2011, 35,
2253. (c) Drouin, F.; Whitehorne, T. J. J.; Schaper, F. Dalton Trans.
2011, 40, 1396. (d) Sung, C.-Y.; Li, C.-Y.; Su, J.-K.; Chen, T.-Y.; Lin,
C.-H.; Ko, B.-T. Dalton Trans. 2012, 41, 953.
(14) (a) Sun, J.; Shi, W.; Chen, D.; Liang, C. J. Appl. Polym. Sci. 2002,
86, 3312. (b) John, A.; Katiyar, V.; Pang, K.; Shaikh, M. M.; Nanavati,
H.; Ghosh, P. Polyhedron 2007, 26, 4033. (c) Bhunora, S.; Mugo, J.;
Bhaw-Luximon, A.; Mapolie, S.; Van Wyk, J.; Darkwa, J.; Nordlander,
E. Appl. Organomet. Chem. 2011, 25, 133. (d) Chen, L.-L.; Ding, L.-Q.;
ASSOCIATED CONTENT
* Supporting Information
Figures S1−S4. Details of the crystal structure determinations
(CIF). This material is available free of charge via the Internet
at http://pubs.acs.org.
■
S
AUTHOR INFORMATION
Corresponding Author
*E-mail: frank.schaper@umontreal.ca.
Notes
■
Zeng, C.; Long, Y.; Lu, X.-Q.; Song, J.-R.; Fan, D.-D.; Jin, W.-J. Appl.
̈
Organomet. Chem. 2011, 25, 310. (e) Gowda, R. R.; Chakraborty, D. J.
Mol. Catal. A: Chem. 2011, 349, 86. (f) Li, C.-Y.; Hsu, S.-J.; Lin, C.-l.;
Tsai, C.-Y.; Wang, J.-H.; Ko, B.-T.; Lin, C.-H.; Huang, H.-Y. J. Polym.
Sci., Part A: Polym. Chem. 2013, 51, 3840.
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
(15) Ding, L.; Jin, W.; Chu, Z.; Chen, L.; Lu, X.; Yuan, G.; Song, J.;
̈
We thank Lylia Dif-Yaiche for her contributions to the
synthesis of 5d during her internship. This work was supported
by the Natural Sciences and Engineering Research Council of
Canada (NSERC) and the Centre in Green Chemistry and
Fan, D.; Bao, F. Inorg. Chem. Commun. 2011, 14, 1274.
(16) Whitehorne, T. J. J.; Schaper, F. Chem. Commun. (Cambridge, U.
K.) 2012, 48, 10334.
(17) Jazdzewski, B. A.; Holland, P. L.; Pink, M.; Young, V. G.;
́
Catalysis (CGCC). We thank Pierre Menard-Tremblay and Dr.
Spencer, D. J. E.; Tolman, W. B. Inorg. Chem. 2001, 40, 6097.
13621
dx.doi.org/10.1021/ic402133c | Inorg. Chem. 2013, 52, 13612−13622

Inorganic Chemistry
Article
(18) (a) Wiese, S.; Badiei, Y. M.; Gephart, R. T.; Mossin, S.; Varonka,
M. S.; Melzer, M. M.; Meyer, K.; Cundari, T. R.; Warren, T. H. Angew.
Chem., Int. Ed. 2010, 49, 8850. (b) Melzer, M. M.; Mossin, S.;
Cardenas, A. J. P.; Williams, K. D.; Zhang, S.; Meyer, K.; Warren, T. H.
Inorg. Chem. 2012, 51, 8658.
(19) Normand, M.; Kirillov, E.; Roisnel, T.; Carpentier, J.-F.
Organometallics 2012, 31, 5511.
(20) (a) Oguadinma, P. O.; Schaper, F. Organometallics 2009, 28,
4089. (b) Oguadinma, P. O.; Schaper, F. Organometallics 2009, 28,
6721. (c) Oguadinma, P. O.; Schaper, F. Can. J. Chem. 2010, 88, 472.
(21) In several instances, 5d was obtained even when only 1 equiv of
diketimine was used, complicating the preparation of 4d. The reason
for the differences in reaction outcome remains unclear and they do
not correlate with the source of starting material, the operator, reaction
times, or changes in reaction temperature.
(22) Dobson, J.; Green, B.; Healy, P.; Kennard, C.; Pakawatchai, C.;
White, A. Aust. J. Chem. 1984, 37, 649.
(23) (a) Dessy, G.; Fares, V. Acta Crystallogr., Sect. C: Cryst. Struct.
Commun. 1979, 8, 101. (b) Park, K.-H.; Marshall, W. J. J. Am. Chem.
Soc. 2005, 127, 9330. (c) Morozova, N. B.; Stabnikov, P. A.; Baidina, I.
A.; Semyannikov, P. P.; Trubin, S. V.; Igumenov, I. K. J. Struct. Chem.
2007, 48, 889. (d) Marsh, R. E. Acta Crystallogr., Sect.B: Struct. Sci.
2009, 65, 782.
(24) Whitehorne, T. J. J.; Vabre, B.; Schaper, F. unpublished results
2013
(25) Singh, J. V.; Baranwal, B. P.; Mehrotra, R. C. Z. Anorg. Allg.
Chem. 1981, 477, 235.
(26) Clegg, W.; Cope, E. K.; Edwards, A. J.; Mair, F. S. Inorg. Chem.
1998, 37, 2317.
(27) Carey, D. T.; Cope-Eatough, E. K.; Vilaplana-Mafe, E.; Mair, F.
S.; Pritchard, R. G.; Warren, J. E.; Woods, R. J. Dalton Trans. 2003,
1083.
(28) Monillas, W.; Bazzoli, T.; Yap, G.; Theopold, K. J. Chem.
Crystallogr. 2010, 40, 67.
(29) El-Zoghbi, I.; Ased, A.; Oguadinma, P. O.; Tchirioua, E.;
Schaper, F. Can. J. Chem. 2010, 88, 1040.
(30) Save, M.; Schappacher, M.; Soum, A. Macromol. Chem. Phys.
2002, 203, 889.
(31) APEX2, Release 2.1-0; Bruker AXS Inc.: Madison, WI, 2006.
(32) SAINT, Release 7.34A; Bruker AXS Inc.: Madison, WI, 2006.
(33) Sheldrick, G. M. SADABS; Bruker AXS Inc.: Madison, WI, 1996,
2004.
(34) Sheldrick, G. M. Acta Crystallogr. 2008, A64, 112.
13622
dx.doi.org/10.1021/ic402133c | Inorg. Chem. 2013, 52, 13612−13622