Cationic zinc complexes: A new class of catalyst for living lactide polymerization at ambient temperature

Article abstract of DOI:10.1039/c0cc03463k

Cationic zinc complexes of a bis(phosphinimine) pincer ligand have been prepared. Methylzinc and zinc-lactate complexes have been structurally characterized, and the latter is the first cationic metal complex to promote coordination-insertion polymerizati

Full text of DOI:10.1039/c0cc03463k

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
Cationic zinc complexes: a new class of catalyst for living lactide
polymerization at ambient temperaturew
Craig A. Wheaton and Paul G. Hayes*
Received 25th August 2010, Accepted 17th September 2010
DOI: 10.1039/c0cc03463k
Cationic zinc complexes of a bis(phosphinimine) pincer ligand
have been prepared. Methylzinc and zinc–lactate complexes
have been structurally characterized, and the latter is the first
cationic metal complex to promote coordination–insertion
polymerization of lactide at ambient temperature. This novel
catalyst system is remarkably active and also exhibits living
character. A detailed investigation of the kinetics and mechanism of
the polymerization process has been undertaken.
Polylactide (PLA), produced primarily by the ring-opening
polymerization of lactide (LA), is one of the most
commercially relevant biodegradable polymers.1 A number
Scheme 1 Synthesis of cationic complexes 2a and 2b, and the cationic
of well-defined metal alkoxide systems are known to induce
zinc–lactate complex 3 by reaction of (i) 1a (Ar = m-(CF3)2–C6H3) or
the ring-opening polymerization of LA via a coordination–
insertion mechanism.2 Specifically, zinc-based systems have
1b (Ar = C6H5) with ZnMe2 and (ii) 1a with EtZn(methyl-L-lactate).
been among the most thoroughly studied due to high activity
with p-isopropylphenyl (Pipp) at the phosphinimine N-aryl site.
coupled with superior molecular weight control. Of the
Cationic zinc complexes of this ligand have been prepared,
numerous zinc alkoxide catalysts reported, the vast majority
including a zinc–lactate complex that exhibits excellent LA
are supported by monoanionic ancillary ligands.3 The use of
polymerization characteristics (Scheme 1). This new generation
neutral ligands is much less common,2a,4 and there has been
catalyst promotes the rapid and living ring-opening poly-
virtually no exploration of the LA polymerization properties
of cationic systems.5 This is surprising given the high activity
of formally zwitterionic tris(pyrazolyl)borate catalysts.3b
catalyst.
Cationic, coordinatively unsaturated species should be expected
merization of rac-LA at ambient temperature, which to our
knowledge, has never before been achieved with a cationic metal
The bis(phosphinimine) ligand L was easily prepared from
to promote facile coordination of LA, and for this reason
4,6-bis(diphenylphosphino)dbf and Pipp-azide under standard
Staudinger conditions,7 and isolated as an analytically pure
we have targeted cationic zinc complexes as homogeneous
catalysts for the polymerization of LA.6
pale yellow solid in 85% yield. A single resonance appears
in the 31P{1H} NMR spectrum of L at d ꢀ5.10 (C6D6).
Our ongoing research has explored neutral mono- and
Diagnostic 1H NMR peaks include the isopropyl methine,
bis-phosphinimine ligands constructed from a dibenzofuran
(dbf) backbone, and cationic heteroleptic zinc complexes
which resonates as a septet at d 2.73, and the corresponding
isopropyl methyls, which resonate as a doublet at d 1.16.
thereof. It has been demonstrated that cationic alkylzinc
Using previously described methods,6a,8 protonation of the
complexes can be readily prepared, but these have displayed
poor catalytic properties.6a More recently, we reported the
ligand was easily carried out generating ion-pairs 1a and 1b.
preparation and crystallographic characterization of a series
of coordinatively unsaturated zinc complexes.6b Alkylzinc
Reaction of 1a with dimethylzinc in toluene afforded the cationic
alkylzinc complex 2a with concomitant loss of methane. This
complexes were again mediocre catalysts; however, a zinc–
reaction proved exceedingly facile, proceeding to completion
lactate species was found to polymerize rac-LA at 60 1C. This
represented the first direct observation of coordination–insertion
polymerization of LA by a cationic zinc complex.
within several minutes at ambient temperature. Complex 2a has
apparent C2v symmetry in solution, resonating sharply at d 25.4 in
1
the 31P{1H} NMR spectrum (C6D5Br solvent). In the H NMR
In this communication we report a variation of the ligand
system employed in the preceding study,6b whereby the steric
demands are reduced by replacing 2,4,6-trimethylphenyl (Mes)
spectrum, the Zn–Me appears at d ꢀ0.79, while the diagnostic
isopropyl methine and methyl groups shift upfield to d 2.53 and
d 1.00, respectively. Despite repeated attempts, single crystals
suitable for an X-ray diffraction experiment remained elusive.
However, upon utilization of the more crystalline tetraphenyl-
borate counter-ion, single crystalsz of 2b were readily obtained
(Fig. 1).9 No significant interaction exists between the cation and
the anion in the solid state, and thus, this can be considered an
excellent representation of the cation of 2a. The geometry at zinc
is distorted trigonal pyramidal, with the phosphinimines
(Zn(1)–N(1) = 2.044(2) A, Zn(1)–N(2) = 2.051(3) A) and the
Department of Chemistry and Biochemistry, University of Lethbridge,
4401 University Drive, Lethbridge, Alberta, Canada.
E-mail: p.hayes@uleth.ca; Fax: +1 403-329-2057;
Tel: +1 403-329-2313
w Electronic supplementary information (ESI) available: Additional
synthetic and crystallographic details, as well as further information
about kinetic experiments and polymer characterization. CCDC
793414–793416. For ESI and crystallographic data in CIF format
see DOI: 10.1039/c0cc03463k
c
8404 Chem. Commun., 2010, 46, 8404–8406
This journal is The Royal Society of Chemistry 2010

View Article Online
Zn(1)–N(2)
=
2.000(5) A), and weaker, but significant,
coordination of the dbf oxygen (Zn(1)–O(1) = 2.336(5) A). The
geometry at zinc is trigonal bipyramidal, with the phosphinimine
nitrogen atoms and the lactate O(2) in the equatorial positions
(Zn(1)–O(2) = 1.909(6) A). The sum of angles about the
equatorial sites is 357.3(6)1, with all angles slightly less than
1201. The lactate carbonyl occupies the apical site (Zn(1)–O(3) =
2.120(6) A) opposite the dbf oxygen, which was vacant in
complex 2b. The angle between these apical sites is 169.8(2)1.
The lactate moiety exists as a 50 : 50 mixture of L and
D
isomers, which was modelled as a two-site positional disorder
of the C(55) atom. This observation suggests racemization
caused by the harsh reaction conditions used for its synthesis.
Notably, however, preliminary experiments using L-LA argue
against epimerization during polymerization (vide infra).
Complex 3 is a highly active catalyst for the polymerization of
LA at ambient temperature, giving 90% conversion of 200 equiv.
of rac-LA in 50 min (25 1C, CD2Cl2 solvent). The resulting PLA
has a slightly hetero-enriched microstructure (Pr = 0.63), likely
due to a chain end control process.11 Such activity is comparable
to some of the most highly active zinc-based catalyst systems
known.3 For example, neutral complexes LZnOEt and LMgOEt
(L = tris(3-tert-butylpyrazolyl)borate) polymerize 500 equiv. of
L-LA to 90% in 6 d and 1 h, respectively.3b Measurement
of the rate under these conditions confirmed the reaction
was first order in [monomer], with an observed rate constant of
8.65(4) ꢁ 10ꢀ4 sꢀ1. In addition, determination of kobs at various
concentrations of 3 established the polymerization to be first
order in [catalyst]. Thus, this process has an overall second-order
rate law (rate = k[3][LA]; k = 0.17(1) Mꢀ1 sꢀ1). After complete
consumption of monomer, addition of another 200 equiv. of LA
demonstrated the living character of the system, which continued
polymerization at a similar rate (k = 0.16(2) Mꢀ1 sꢀ1).
Fig. 1 X-Ray crystal structure of the cation of 2b, with ellipsoids
shown at the 30% level. Hydrogen atoms, disordered atomic positions,
and a phenyl group of P(1) have been omitted for clarity.
methyl group (Zn(1)–C(55) = 2.002 A) occupying the equatorial
positions and the more weakly coordinated dbf oxygen in the
apical site (Zn(1)–O(1) = 2.284(2) A). The phosphinimine bite
angle is slightly reduced from ideal (N(1)–Zn–N(2) = 112.5(1)1),
and the sum of angles about the equatorial positions is 359.5(5)1.
The coordination of the dbf oxygen is of significant interest given
that such an interaction was not observed in analogous complexes
of the bulkier Mes-substituted ligand.6b Presumably, the reduced
bulk or donating ability of the Pipp groups renders this bonding
mode possible.
Alkylzinc compounds 2a and 2b are inactive for poly-
merization of LA. Thus, complex 3 was prepared by reaction
of 1a with EtZn(methyl-L-lactate) in bromobenzene at 100 1C
for 2 h. Characteristic signals for the lactate group of complex
3 were observed in the 1H NMR spectrum at d 3.19 (O–CH3),
3.76 (CHCH3), and 0.94 (CHCH3).10 Despite the presence of
the asymmetric methyl-lactate moiety, 3 exhibits apparent C2v
symmetry in solution with a single resonance in the 31P{1H}
NMR spectrum at d 28.9, suggesting rapid fluxional behaviour
in solution on the NMR timescale.
Upon exposure of complex 3 to excess rac-LA, immediate
conversion to a new species with a 31P{1H} NMR chemical shift
(at d 29.4) slightly downfield of the initial catalyst (d 28.9) was
observed. More notably, the lactate end group gives rise to an
O–CH3 singlet at d 3.64 in the 1H NMR spectrum. The methine
signal, however, was not observed and is likely obscured beneath
the LA/PLA signals. Given these spectroscopic observations, this
new species is presumed to be the product of insertion of one
or more monomer units, providing strong evidence for a
coordination–insertion mechanism. In addition, a low molecular
weight polymer sample ([LA]0/[3] = 50) was analysed by
MALDI-ToF mass spectrometry, and the observed masses of
all oligomer fragments were consistent with the presence of a
methyl-L-lactate end-group. Unfortunately, the mass peaks
are separated by m/z 72, which suggests intermolecular trans-
esterification occurs to a significant degree during polymerization.
Determination of the activation parameters for the
polymerization of rac-LA by 3 was accomplished by measuring
the rate of reaction at temperatures ranging from 37 1C to ꢀ7 1C.
An Eyring plot of the data revealed DHz = 47(1) kJ molꢀ1 and
DSz = ꢀ147(4) J Kꢀ1 molꢀ1. These values closely match
activation parameters for other known coordination–insertion
lactide polymerization catalysts and are indicative of a
well-controlled polymerization process.12
The solid-state structure of 3, shown in Fig. 2, reveals
a similar bonding mode to 2b, with strong coordination
of both phosphinimine donors (Zn(1)–N(1) = 2.014(6) A;
Fig. 2 X-Ray crystal structure of the cation of 3, with ellipsoids
drawn at the 30% probability level. Hydrogen atoms and disordered
atomic positions have been omitted for clarity.
Several polymer samples have been prepared using a range
of catalyst concentrations, with the monomer to catalyst ratio
c
This journal is The Royal Society of Chemistry 2010
Chem. Commun., 2010, 46, 8404–8406 8405

View Article Online
functionalities, to inhibit transesterification side reactions and
enhance stereocontrol, respectively.
P.G.H. thanks NSERC, CFI, Canada School of Energy and
Environment and GreenCentre Canada for financial support.
C.A.W. acknowledges NSERC and the Alberta Ingenuity
Fund (Alberta Innovates) for student awards. Thanks to Dr.
Timothy Clark and Yun Yang of GreenCentre Canada for
GPC measurements.
Notes and references
z Crystal data for 2b and 3: 2b: C85H76.5BBr0.5N2OP2Zn,
M = 1320.09, triclinic, a = 9.878(7), b = 17.43(1), c = 21.93(1) A,
a = 80.235(8)1, b = 82.751(8)1, g = 80.160(8)1, U = 3647(4) A3,
ꢀ
T = 173(2) K, space group P(1) (no. 2), Z = 2, 44 021 reflections,
12 848 unique (Rint = 0.0270) were used in all calculations after
removing electrons from solvent accessible voids. The final wR(F2)
was 0.1284 (all data); 3: C90H67BF24N2O4P2ZnꢂC6H6, M = 1912.68,
monoclinic, a = 19.087(2), b = 26.603(2), c = 19.538(2) A, b =
116.133(1)1, U = 8906.6(1) A3, T = 173(2) K, space group P2(1)/c
(no. 14), Z = 4, 72 717 reflections, 9322 unique (Rint = 0.0803) were
used in all calculations. The final wR(F2) was 0.1665 (all data).
Fig. 3 Plot of observed Mn
( ) and PDI ( ) as a function of the
monomer to catalyst ratio. The straight line represents calculated Mn
values (Mn = [LA]0 : [3] ꢁ 144.13), while the curved line shows the
general trend for experimental Mn values.
ranging from 100 to 1000, and their molecular weights have
been analyzed by gel permeation chromatography (GPC)
(Fig. 3). At high concentrations of catalyst 3 ([LA]0/[3] =
100 and 200), the molecular weights of the resulting polymer
samples closely approximate the calculated values. However,
at lower concentrations of 3, the molecular weights drop off
1 (a) R. Auras, B. Harte and S. Selke, Macromol. Biosci., 2004, 4,
835–864; (b) J.-C. Bogaert and P. Coszach, Macromol. Symp., 2000,
153, 287–303; (c) K. Fukushima and Y. Kimura, Polym. Int., 2006,
55, 626–642; (d) H. R. Kricheldorf, Chemosphere, 2001, 43, 49–54.
2 (a) C. A. Wheaton, P. G. Hayes and B. J. Ireland, Dalton Trans.,
2009, 4832–4846; (b) R. H. Platel, L. M. Hodgson and
C. K. Williams, Polym. Rev., 2008, 48, 11–63; (c) J. Wu,
T.-L. Yu, C.-T. Chen and C.-C. Lin, Coord. Chem. Rev., 2006,
250, 602–626; (d) O. Dechy-Cabaret, B. Martin-Vaca and
D. Bourissou, Chem. Rev., 2004, 104, 6147–6176.
3 For representative examples see: (a) M. Cheng, A. B. Attygalle,
E. B. Lobkovsky and G. W. Coates, J. Am. Chem. Soc., 1999, 121,
11583–11584; (b) M. H. Chisholm, N. W. Eilerts, J. C. Huffman,
S. S. Iyer, M. Pacold and K. Phomphrai, J. Am. Chem. Soc., 2000,
122, 11845–11854; (c) L. E. Breyfogle, C. K. Williams,
V. G. Young, Jr., M. A. Hillmyer and W. B. Tolman, Dalton
Trans., 2006, 928–936; (d) J. Ejfler, S. Szafert, K. Mierzwicki,
L. B. Jerzykiewicz and P. Sobota, Dalton Trans., 2008, 6556–6562;
(e) M. S. Hill and P. B. Hitchcock, J. Chem. Soc., Dalton Trans.,
2002, 4694–4702; (f) H.-Y. Chen, H.-Y. Tang and C.-C. Lin,
Macromolecules, 2006, 39, 3745–3752.
significantly. For example,
a number-average molecular
weight slightly greater than 50 000 Da, which is only B35%
of the calculated value, was achieved when [LA]0/[3] = 1000.
This observation is consistent with the presence of monomer
impurities acting as chain-transfer agents. The molecular
weight distribution is narrow for all samples (PDI = 1.08–1.34),
with the most narrow distribution occurring at intermediate
catalyst loadings ([LA]0/[3] = 300 and 400). A plausible reason
for higher PDI’s at high catalyst loading is slow initiation relative
to propagation, which becomes statistically less relevant at lower
catalyst concentrations. Higher PDI values at lower catalyst
loading are most likely a result of transesterification, which was
observed to be occurring by MALDI-ToF mass spectrometry
(vide supra). The broadening effect of transesterification on
molecular weight distribution should necessarily depend on
reaction times for the polymerization experiment, which are
longer at lower catalyst loading.
4 (a) J. Boerner, S. Herres-Pawlis, U. Floerke and K. Huber, Eur. J.
¨
¨
Inorg. Chem., 2007, 5645–5651; (b) J. Boerner, U. Floerke, K. Huber,
¨
A. Doering, D. Kuckling and S. Herres-Pawlis, Chem.–Eur. J., 2009,
¨
¨
15, 2362–2376; (c) T. R. Jensen, L. E. Breyfogle, M. A. Hillmyer and
W. B. Tolman, Chem. Commun., 2004, 2504–2505.
5 Y. Sarazin, V. Poirier, T. Roisnel and J.-F. Carpentier, Eur. J.
Inorg. Chem., 2010, 3423–3428.
As a further test of the livingness of 3 for LA polymerization, a
polymer sample was prepared by sequentially polymerizing
two batches of 200 equiv. of rac-LA. Analysis of the resulting
polymer by GPC showed molecular weights similar to those
determined for the single-step polymerization of 400 equiv. of
rac-LA (Mn = 31.5 ꢁ 103 g molꢀ1 vs. 33.4 ꢁ 103 g molꢀ1). The
molecular weight distribution broadened slightly (1.21 vs.
1.09), likely as a result of the longer duration of the poly-
merization experiment.
6 (a) C. A. Wheaton, B. J. Ireland and P. G. Hayes, Organometallics,
2009, 28, 1282–1285; (b) C. A. Wheaton and P. G. Hayes, Dalton
Trans., 2010, 39, 3861–3869.
7 (a) J. Meyer and H. Staudinger, Helv. Chim. Acta, 1919, 2,
635–646; (b) M. Alajarin, C. L. Lopez-Leonardo, P. L. Llamas-
Lorente and D. Bautista, Synthesis, 2000, 2085–2091.
8 B. J. Ireland, C. A. Wheaton and P. G. Hayes, Organometallics,
2010, 29, 1079–1084.
9 Neither the bulk sample of 2b nor the single crystals were obtained
in 100% purity. See the ESIw for more details.
10 These resonances are from the major of two isomers of 3 which have
different 1H and 13C NMR signatures. See the ESIw for more details.
11 (a) B. M. Chamberlain, M. Cheng, D. R. Moore, T. M. Ovitt,
E. B. Lobkovsky and G. W. Coates, J. Am. Chem. Soc., 2001, 123,
3229–3238; (b) M. H. Chisholm, N. J. Patmore and Z. P. Zhou,
Chem. Commun., 2005, 127–129.
12 (a) A. F. Douglas, B. O. Patrick and P. Mehrkhodavandi, Angew.
Chem., Int. Ed., 2008, 47, 2290–2293; (b) M. H. Chisholm,
J. C. Gallucci and C. Krempner, Polyhedron, 2007, 26,
4436–4444; (c) M. H. Chisholm and E. E. Delbridge, New J.
Chem., 2003, 27, 1177–1183.
In summary, we have prepared the first cationic system
capable of catalyzing the ring-opening polymerization
of lactide at ambient temperature. The mechanism of
polymerization has been unambiguously established not to
occur via a cationic process, but rather, by a well-controlled
coordination–insertion mechanism. Future work will involve
modification of the steric bulk and electron donating capacity
of the phosphinimine donors, as well as installation of chiral
c
8406 Chem. Commun., 2010, 46, 8404–8406
This journal is The Royal Society of Chemistry 2010