Ring-opening polymerization of rac-lactide and ε-caprolactone using zinc and calcium salicylaldiminato complexes

Article abstract of DOI:10.1139/v2012-012

Tridentate Schiff base complexes of zinc and calcium were prepared and tested in the ring-opening polymerization of ε-caprolactone and rac-lactide to generate biodegradable polymeric materials from biocompatible metals. Alteration of the pendant donor arm attached to the imine backbone provides some control over catalyst composition and polymerization activity. Complexes of the formula [ONN]ZnN(SiMe₃)₂, where [ONN] = 2-(N-donor arm-imine)[4,6-di(tert-butyl)phenoxide], were isolated with ethyldimethylamine, ethylpiperidine, and ethylmorpholine substituents, while disproportionation led to the isolation of [ONN]₂Zn complexes with methylpyridine, quinoline, and ethyldiisopropylamine derivatives, two of which were crystallographically characterized. Calcium complexes were more stable and novel [ONN]CaN(SiMe₃)₂ complexes with ethylpiperidine and ethyldiisopropylamine substituents were reported. Zinc and calcium catalysts coordinated to a single tridentate ligand were effective at initiating the polymerization of ε-caprolactone, but did not control the polymerizations, whereas the bis(ligand) complexes produced no polymer. These catalysts were effective at controlling the polymerization of rac-lactide. Coordinatively saturated complexes inhibit the polymerization, while initiation from either the amido or ligand alkoxide functionalities produces poly(lactic acid) with low polydispersities.

Full text of DOI:10.1139/v2012-012

419
Ring-opening polymerization of rac-lactide and
ε-caprolactone using zinc and calcium
salicylaldiminato complexes
Joshua B. L. Gallaway, Justin R. K. McRae, Andreas Decken, and Michael P. Shaver
Abstract: Tridentate Schiff base complexes of zinc and calcium were prepared and tested in the ring-opening polymeriza-
tion of 3-caprolactone and rac-lactide to generate biodegradable polymeric materials from biocompatible metals. Alteration
of the pendant donor arm attached to the imine backbone provides some control over catalyst composition and polymeriza-
tion activity. Complexes of the formula [ONN]ZnN(SiMe3)2, where [ONN] = 2-(N-donor arm-imine)[4,6-di(tert-butyl)phen-
oxide], were isolated with ethyldimethylamine, ethylpiperidine, and ethylmorpholine substituents, while disproportionation
led to the isolation of [ONN]2Zn complexes with methylpyridine, quinoline, and ethyldiisopropylamine derivatives, two of
which were crystallographically characterized. Calcium complexes were more stable and novel [ONN]CaN(SiMe3)2 com-
plexes with ethylpiperidine and ethyldiisopropylamine substituents were reported. Zinc and calcium catalysts coordinated to
a single tridentate ligand were effective at initiating the polymerization of 3-caprolactone, but did not control the polymeriza-
tions, whereas the bis(ligand) complexes produced no polymer. These catalysts were effective at controlling the polymeriza-
tion of rac-lactide. Coordinatively saturated complexes inhibit the polymerization, while initiation from either the amido or
ligand alkoxide functionalities produces poly(lactic acid) with low polydispersities.
Key words: lactide, caprolactone, zinc, calcium, ring opening polymerization.
Résumé : On a préparé des complexes de zinc et de calcium avec des bases de Schiff tridentates et on a évalué la possibi-
lité de les utiliser dans les polymérisations par ouverture de cycle de la 3-caprolactone et du rac-lactide conduisant à des
matériaux polymères biodégradables à partir de produits métalliques biodégradables. L’altération du bras donneur pendant
attaché au squelette de l’imine permet d’avoir un certain contrôle sur la composition du catalyseur et son activité dans la po-
lymérisation. On a pu isoler des complexes de formule [ONN]ZnN(SiMe3)2 (dans lesquels [ONN] = 2-(imine avec bras N-
donneur)[4,6-di(tert-butyl)phénolate] portant des substituants éthyldiméthylamine, éthylpipéridine et éthylmorpholine alors
qu’une disproportionation permet d’isoler les complexes [ONN]2Zn avec les dérivés de la méthylpyridine, la quinoléine et
l’éthyldiisopropylamine, dont deux ont été caractérisés par cristallographie. Les complexes du calcium sont plus stables et
on a isolé pour la première fois les nouveaux complexes [ONN]CaN(SiMe3)2 avec des substituants éthylpipéridine et éthyl-
diisopropylamine. Les catalyseurs du zinc et du calcium ne comportant qu’une coordination à un ligand tridentate sont effi-
caces pour initier la polymérisation de la 3-caprolactone, mais ils ne permettent pas de contrôler les polymérisations alors
que les complexes avec un bis(ligand) ne permettent pas de produire des polymères. Ces catalyseurs sont efficaces pour
contrôler la polymérisation du rac-lactide. Les complexes à coordination saturée inhibent la polymérisation alors qu’une ini-
tiation par des groupes fonctionnels amido- ou un ligand alcoolate conduisent à la production d’acide polylactique avec de
faibles polydispersités.
Mots‐clés : lactide, caprolactone, zinc, calcium, polymérisation par ouverture de cycle.
[Traduit par la Rédaction]
poly(esters) to pharmaceuticals, microelectronics, and other
fields.5,6
Aliphatic poly(esters) such as poly(lactic acid) (PLA) and
poly(3-caprolactone) (PCL) have gained considerable interest
in recent years because of their biodegradability and renew-
ability.1–4 These aliphatic polymers are also bioassimilable,
as hydrolysis produces nontoxic components eliminated via
the Krebs cycle as CO2 and H2O.3 These poly(esters) have
physical properties that can be tuned and varied during poly-
mer processing by orientation, blending, branching, cross-
linking, or plasticization, enabling the application of aliphatic
PLA is an important biodegradable polymer of high indus-
trial potential because the lactide monomer can be acquired
through the fermentation of renewable resources.3 The most
common application of PCL as a biodegradable polymer in-
volves the development of pharmaceutical drugs such as Cap-
ronor, an implanted contraceptive delivery system.7 PCL is
synthesized from the monomer 3-caprolactone, which is
Received 23 November 2011. Accepted 21 January 2012. Published at www.nrcresearchpress.com/cjc on 4 April 2012.
J.B.L. Gallaway, J.R.K. McRae, and M.P. Shaver. Department of Chemistry, University of Prince Edward Island, 550 University
Avenue, Charlottetown, PEI C1A 4P3, Canada.
A. Decken. Department of Chemistry, University of New Brunswick, P.O. Box 4400, Fredericton, NB E3B 5A3, Canada.
Corresponding author: Michael P. Shaver (email: mshaver@upei.ca).
Can. J. Chem. 90: 419–426 (2012)
doi:10.1139/V2012-012
Published by NRC Research Press

420
Can. J. Chem. Vol. 90, 2012
traditionally produced by the Bayer–Villiger oxidation of cy-
clohexanone with peracids or hydrogen peroxide8 but can
also be produced from the fermentation of starch, suggesting
renewability.9
Fig. 1. Calcium and zinc Schiff base complexes previously devel-
oped for the polymerization of cyclic esters.13–16
Aliphatic poly(esters) are commonly synthesized through
ring-opening polymerization (ROP) of the corresponding cy-
clic ester as it offers greater control over the resulting poly-
mers.1–4 While ROP of lactide has been accomplished with
many metals, calcium and zinc are particularly promising for
industrial applications, because of their low cost, high activ-
ity, and low toxicity.10 Ligand sets supporting active zinc and
calcium lactide catalysts include b-diketiminates, tris(pyrazol)
hydroborates, anilido-oxazolinates, amino-bis(pyrazol)s, N-
heterocyclic carbenes, heteroscorpionates, and others.10 The
best systems offer excellent conversions over a broad temper-
ature range and polydispersity indexes (PDIs) as low as 1.1.10
ROP of 3-caprolactone with calcium and zinc catalysts is
comparatively understudied, although simple organic amido cal-
cium and zinc complexes, H2NCaOCHMe2 and Zn(N(SiMe3)2)2,
show high activity and poor control.10,11 PCL synthesis is
also efficiently mediated by heterobimetallic aluminum and
zinc complexes with anilido-aldimine ligands12 and some
scorpionate-derived monometallic frameworks.10
Inspiration for this study, however, was derived from the
work of Darensbourg et al.13–16 who showed that a series of
tridentate Schiff base complexes, shown in Fig. 1, of benign
metals were effective in the ROP of cyclic esters. Complexes
were variably effective, with pendant dimethylamine donors
coordinated to calcium providing the best activity for lactide
and trimethylene carbonate, whereas zinc complexes were
less effective unless substituted with bulky amino acid de-
rived ligands and used for lactide and 3-caprolactone poly-
merizations.
were performed under an N2 atmosphere in the glovebox or
on a Schlenk line. The Schlenk reactions utilized a dual
manifold Schlenk line with Kontes valves. All samples were
1
characterized via H and 13C NMR spectroscopy. These spec-
tra were obtained by running samples on a 300 MHz Bruker
Avance spectrometer. Elemental analyses were performed at
Guelph Analytical Laboratories. A Bruker AXS P4/SMART
1000 diffractometer was used to obtain crystallographic data.
Polymerization data were obtained using a PolymerLabs
GPC50 equipped with two Jordi Gel DVB mixed bed col-
umns (300 mm × 7.8 mm). The mobile phase used in the
system was HPLC grade THF and the polymer samples were
dissolved in THF to create a solution for GPC analysis, at
50 °C, at a 1 mg/mL concentration, and at a 1 mL/min flow
rate. Molecular weights were measured and corrected relative
to styrene standards.17
Herein we report an extension of this series of complexes,
further examining the structure/activity relationships and the
role the ligand plays in controlling the ROP of rac-lactide
and 3-caprolactone. New substitutents include diisopropyl-
amine (N(i-Pr)2), piperidine (Pip), and morpholine (Mor)
functionalities. Complexes substituted with a single tridentate
ligand are effective catalysts for the ROP of rac-lactide.
Synthesis and characterization
Ligand frameworks
NMe2
Ligands
½ONNꢀME2 (1), Q[ONN] (2), and Py[ONN]
Experimental
(3), where R[ONN] is 2-(N-donor arm-imine)(4,6-di(tert-
butyl)phenoxide, were synthesized according to a literature
procedure.14
General experimental procedures
Chemicals and solvents used for the syntheses were pur-
chased from Sigma-Aldrich, Fisher Chemicals, and Acros
Chemicals. 3,5-Di-tert-butyl-2-hydroxybenzaldehyde (99%
pure) and starting amine reagents were all used as received.
Dry solvents including tetrahydrofuran (THF), toluene, pen-
tane, and ether were obtained from an Innovative Technology
glovebox system, which was fitted with a solvent purification
system. The inline purification system was made up of alu-
mina and a copper catalyst. These anhydrous solvents were
degassed by three or four freeze–pump–thaw cycles, before
taking them into an MBraun LABmaster sp. glovebox. Other
solvents, including CDCl3, C6D6, and hexamethyldisiloxane,
were dried over CaH2 for at least 24 h, before vacuum trans-
ferring or vacuum distilling. The MBraun glovebox was
connected to a –35 °C freezer as well as [H2O] and [O2]
analyzers. Air-sensitive reactions involving zinc and calcium
Nði-PrÞ2
Nði-PrÞ2
Synthesis of
½ONNꢀði-PrÞ2
½ONNꢀ (4):
To a stirred solution of 3,5-di-tert-butyl-2-hydroxybenzal-
dehyde (2.51 g, 10.6 mmol) in methanol (25 mL) was added
dropwise 2-(diisopropylamino) ethylamine (1.54 g,
10.6 mmol). The bright yellow solution was refluxed at 65 °C
for approximately 10 h and then allowed to cool to room
temperature. All volatiles were removed in vacuo to yield a
bright yellow oil. The oil was dried under vacuum on the
Schlenk line for 2 h and then purified by dissolving in cold
pentane. The resulting solution was stored at –20 °C over-
night, and the same purification process was repeated three
times to give a dense yellow solid (1.37 g, 3.8 mmol, 36%).
1H NMR (CDCl3, ppm) d:14.0 (1H, s, OH), 8.4 (1H, s,
CH=N), 7.4–7.1 (2H, m, Ar–H), 3.6 (2H, t, J = 7 Hz, CH2),
Published by NRC Research Press

Gallaway et al.
421
3.0 (2H, septet, J = 7 Hz, CHMe2), 2.7 (2H, t, J = 7 Hz,
CH2), 1.5 (9H, s, C(CH3)3), 1.3 (9H, s, C(CH3)3), 1.0 (12H,
d, J = 7 Hz, i-Pr CH3). 13C NMR (CDCl3, ppm) d: 166.3
(CH=N), 158.6, 140.0, 136.9, 126.9, 125.9, 118.1 (Ar),
61.6, 49.3, 35.2, 35.0, 34.7, 31.9, 31.8, 29.6. Elemental anal.
calcd for C23H40N2O (%): C 76.61, H 11.18, N 7.77; found:
C 76.81, H 11.11, N 7.77.
Synthesis of Mor[ONN]ZnN(SiMe3)2 (9):
Complex 9 was prepared analogously to 8 using Mor[ONN]
(1.01 g, 2.91 mmol), THF (40 mL), and Zn(N(SiMe3)2)2
(1.12 g, 2.91 mmol). In contrast to 8, 9 was mixed
with 10 mL of THF and added to a stirring mixture of
Zn(N(SiMe3)2)2 in 30 mL THF. This synthesis yielded a
light yellow solid (1.10 g, 1.92 mmol, 67%). 1H NMR
(CDCl3, ppm) d: 8.2 (1H, s, CH=N), 7.5 (1H, d, J =
3 Hz, Ar–H), 6.9 (1H, d, J = 3 Hz, Ar–H), 3.7 (4H, t, J =
6 Hz, O–CH2), 2.9 (4H, t, J = 6 Hz, N–CH2), 2.6 (2H, t,
J = 7 Hz, CH2), 1.9 (2H, t, J = 7 Hz, CH2), 1.5 (9H, s,
C(CH3)3), 1.3 (9H, s, C(CH3)3), 0.1 (18H, s, SiMe3). 13C
NMR (CDCl3, ppm) d: 172.9 (C=N), 168.8, 141.7, 135.4,
130.0, 129.5, 117.2 (Ar), 68.2, 67.0 (O–CH2), 59.0, 57.7
(CH2), 35.8, 34.0 (C(CH3)3), 31.6, 29.7 (C(CH3)3), 5.1
(SiMe3). Elemental anal. calcd for C24H42N3O2SiZn (%): C
57.87, H 8.50, N 8.44; found: C 57.81, H 8.30, N 8.39.
Synthesis of Pip[ONN] (5):
Ligand 5 was prepared analogously to 4, using 3,5-di-tert-
butyl-2-hydroxybenzaldehyde (2.56 g, 10.6 mmol), methanol
(45 mL), and 4-(2-aminoethyl)piperidine (1.36 g,
10.6 mmol). This yielded a yellow powder (2.42 g,
1
10.6 mmol, 70%). H NMR (CDCl3, ppm) d: 13.8 (1H, s,
OH), 8.4 (1H, s, CH=N), 7.4 (1H, d, J = 2 Hz, Ar–H), 7.1
(1H, d, J = 2 Hz, Ar–H), 3.8 (2H, t, J = 7 Hz, CH2), 2.7
(2H, t, J = 7 Hz, CH2), 2.5 (4H, t, J = 5 Hz, CH2), 1.6 (6H,
m, CH2), 1.5 (9H, s, C(CH3)3, 1.3 (9H, s, C(CH3)3). 13C
NMR (CDCl3, ppm) d: 166.75 (C=N), 158.3, 140.1, 136.9,
127.0, 126.0, 118.1 (Ar), 60.0, 57.3, 55.1, 35.4, 35.2, 34.3,
31.7, 26.2, 24.4 (CH2). Elemental anal. calcd for C22H36N2O
(%): C 76.69, H 10.53, N 8.13; found: C 76.90, H 10.45, N
8.10.
Synthesis of Py[ONN]2Zn (10):
Complex 10 was prepared analogously to 8, leading to an
impure reaction mixture. Recrystallization led to the isolation
of the bis(ligand) complex in 20% yield. Altering reaction
conditions to use 2 equiv of Py[ONN] (1.11 g, 3.42 mmol),
THF (40 mL), and 1 equiv of Zn(N(SiMe3)2)2 (0.66 g,
1.71 mmol) yielded a brown powder (10) in high yield
Synthesis of Mor[ONN] (6):
Ligand 6 was prepared analogously to 4, using 3,5-di-tert-
butyl-2-hydroxybenzaldehyde (2.56 g, 10.6 mmol), methanol
(45 mL), and 4-(2-aminoethyl)morpholine (1.39 g,
10.6 mmol) to yield a pale yellow solid (3.82 g, 10.6 mmol,
1
(0.85 g, 1.54 mmol, 90%). H NMR (CDCl3, ppm) d: 8.5
(2H, m, Ar–H), 8.2 (2H, s, CH=N), 7.3–6.8 (10H, m, Ar–
H), 4.7 (4H, s, CH2), 1.4 (18H, s, C(CH3)3), 1.3 (18H, s,
C(CH3)3). 13C NMR (CDCl3, ppm) d: 171.6 (C=N), 170.6,
157.1, 149.0, 141.8, 137.5, 133.5, 129.2, 129.1, 122.7,
122.3, 117.9 (Ar), 63.0 (CH2), 35.8, 34.5 (C(CH3)3), 31.9,
29.9 (C(CH3)3). Elemental anal. calcd for C42H34N4O2Zn
(%): C 70.82, H 7.64, N 7.87; found: C 71.00, H 7.81, N
7.66.
1
85%). H NMR (CDCl3, ppm) d: 13.7 (1H, s, OH), 8.4 (1H,
s, CH=N), 7.4 (1H, d, J = 2 Hz, Ar–H), 7.1 (1H, d, J =
2 Hz), 3.73 (8H, t, J = 4 Hz, CH2), 2.7 (2H, t, J = 7 Hz,
CH2), 2.5 (2H, t, J = 7 Hz, CH2) 1.4 (9H, s, C(CH3)3), 1.3
(9H, s, C(CH3)3). 13C NMR (CDCl3, ppm) d: 167.0 (C=N),
158.3, 140.2, 136.9, 127.1, 126.0, 118.1 (Ar), 67.2 (CH2),
59.5, 57.0 (CH2), 54.1 (CH2), 35.2, 34.3 (C(CH3)3), 31.7,
29.6 (C(CH3)3). Elemental anal. calcd for C21H34N2O2 (%):
C 72.79, H 9.89, N 8.08; found: C 72.84, H 9.91, N 8.15.
Q
Synthesis of [ONN]2Zn (11):
Complex 11 was prepared analogously to 10 using Q[ONN]
(3.00 g, 7.76 mmol), THF (40 mL), and Zn(N(SiMe3)2)2
(1.40 g, 3.88 mmol). Complex 11 was recrystallized from a
mixture of hexamethyldisiloxane and THF prior to use. The
synthesis yielded a bright red solid (2.72 g, 34.6 mmol,
88%). 1H NMR (CDCl3, ppm) d: 9.0 (2H, d, J = 5 Hz,
Ar–H), 8.7 (2H, s, CH=N), 8.4 (2H, d, J = 5 Hz, Ar–H),
8.4–7.0 (12H, m, Ar–H), 1.6 (18H, s, C(CH3)3), 1.3 (18H,
s, C(CH3)3). Elemental anal. calcd for C48H54N4O2Zn (%):
C 73.50, H 6.94, N 7.14; found: C 73.71, H 7.00, N 7.05.
Zinc complexes
NMe2
The complex
½ONNꢀZnNðSiMe3Þ2Me2 (7) was synthe-
sized according to a literature procedure.14
Synthesis of Pip[ONN]ZnN(SiMe3)2 (8):
To a stirring solution of Pip[ONN] (2.00 g, 6.16 mmol) in
THF (40 mL), a 1:1 stoichiometric ratio of Zn(N(SiMe3)2)2
(2.38 g, 6.16 mmol) was added as a THF solution. After
2 h, the THF was removed in vacuo and the compound was
isolated. To ensure complete dryness, the resulting solid was
left under vacuum for several hours. To purify the solid, a
pentane/toluene mixture was used to recrystallize the product.
Nði-PrÞ2
Synthesis of
½ONNꢀ2Znði-PrÞ2 (12):
Complex 12 was prepared analogously to 10 using
Nði-PrÞ2
½ONNꢀði-PrÞ2 (1.50 g, 4.15 mmol), THF (50 mL), and
Zn(N(SiMe3)2)2 (0.80 g, 2.07 mmol). The resulting product
was a yellow solid (1.79 g, 3.02 mmol, 86%). 1H NMR
(CDCl3, ppm) d: 8.2 (2H, s, CH=N), 7.4 (2H, d, J = 3 Hz,
Ar–H), 7.0 (2H, d, J = 3 Hz, Ar–H) 3.5 (4H, t, J = 9 Hz,
CH2), 2.9 (4H, septet, J = 4 Hz, CH(CH3)2), 2.7 (4H, t, J =
9 Hz, CH2), 1.5 (18H, s, (C(CH3)3), 1.3 (18H, s, C(CH3)3),
0.9 (12H, d, J = 4 Hz, CH(CH3)2), 0.8 (12H, d, J = 4 Hz,
CH(CH3)2). Elemental anal. calcd for C46H78N4O2Zn (%): C
70.42, H 10.02, N 7.14; found: C 71.55, H 10.50, N 7.50.
While repeated elemental analysis tests of isolated crystals
were off, suggesting decomposition in shipping, crystals ob-
tained from the same batch on which an X-ray diffraction
1
This yielded a yellow solid (1.97 g, 3.42 mmol, 58%). H
NMR (CDCl3, ppm) d: 8.2 (1H, s, CH=N), 7.4 (1H, d, J =
3 Hz, Ar–H), 6.9 (1H, d, J = 3 Hz, Ar–H), 3.7 (2H, t, J =
6 Hz, CH2), 2.9 (2H, t, J = 6 Hz, CH2), 2.7 (4H, t, J = 5 Hz,
N–CH2), 1.7 (6H, pentet, J = 5 Hz, CH2), 1.5 (9H, s, C(CH3)3),
1.3 (9H, s, C(CH3)3), 0.1 (18H, s, SiMe3). 13C NMR (CDCl3,
ppm) d: 170.1 (C=N), 167.6, 141.6, 134.7, 129.4, 128.0,
117.7 (Ar), 55.7, 55.5 (CH2), 53.8, 52.6 (N–CH2), 35.6, 33.8
(C(CH3)3), 31.3, 29.4 (C(CH3)3), 25.7, 23.7, 23.0 (CH2), 5.6
(Si(CH3)3). Elemental anal. calcd for C25H44N3OSiZn (%): C
60.52, H 8.94, N 8.49; found: C 60.40, H 9.01, N 8.29.
Published by NRC Research Press

422
Can. J. Chem. Vol. 90, 2012
Fig. 2. Zinc complexes of Schiff base ligands.
Fig. 4. ORTEP drawing (spheroids at 50% probability) of Q[ONN]2Zn
(11). Protons and tert-butyl carbons are omitted for clarity.
Nði-PrÞ2
Synthesis of
½ONNꢀCaNðSiMe3Þ2ði-PrÞ2 (15):
Complex 15 was prepared analogously to 14 using 1 equiv
Nði-PrÞ2
of
½ONNꢀði-PrÞ2 (1.01 g, 2.80 mmol), 2 equiv of NaN
(SiMe3)2 (1.03 g, 5.60 mmol), THF (30 mL), and CaI2
(0.82 g, 2.80 mmol), isolating 15 as a bright yellow solid
1
Fig. 3. Calcium complexes of Schiff base ligands.
(0.80 g, 1.43 mmol, 51%). H NMR (CDCl3) d: 8.3 (1H, s,
CH=N), 7.7 (1H, d, J = 3 Hz, Ar–H), 7.2 (1H, d, J = 3 Hz,
Ar–H), 3.0 (2H, septet, J = 6 Hz, CH(CH3)2), 2.6 (4H, m,
J = 9 Hz, CH2), 1.5 (9H, s, C(CH3)3), 1.4 (9H, s, C(CH3)3),
1.0 (12H, d, J = 6 Hz, CH(CH3)3), 0.3 (18H, s, SiMe3). Ele-
mental anal. calcd for C26H48N3CaOSi (%): C 64.14, H 9.94,
N 8.63; found: C 64.07, H 10.01, N 8.47.
Crystal structure analyses
Crystals of 11 and 12 were grown by vapour diffusion us-
ing pentane and toluene at –35 °C. Single crystals were
coated with Paratone-N oil, mounted using a polyimide Mi-
croMount, and frozen in the cold nitrogen stream of the goni-
ometer. A hemisphere of data was collected on a Bruker
AXS P4/SMART 1000 diffractometer using u and q scans
with a scan width of 0.3° and 10 s (11) or 30 s (12) exposure
times. The detector distance was 5 cm. The data were re-
duced using SAINT18 and corrected for absorption with
SADABS.19 The structure was solved by direct methods and
refined by full-matrix least-squares on F2(SHELXTL).20 For
11, one of the tert-butyl groups, attached at C(4), was disor-
dered and the site occupancies determined using an isotropic
model at 0.7 (C(9)–C(11)) and 0.3 (C(9′)–C(11′) and fixed in
subsequent refinement cycles. All non-hydrogen atoms were
refined using anisotropic displacement parameters. Hydrogen
atoms were included in calculated positions and refined using
a riding model. Graphics were created using the ORTEP pro-
gram.21
study was performed were used in the reported polymeriza-
tion reactions.
Calcium complexes
NMe2
½ONNꢀCaNMe2ðSiMe3Þ2 (13) was synthesized accord-
ing to a literature procedure.14
Synthesis of Pip[ONN]CaN(SiMe3)2 (14):
Pip[ONN] (0.95 g, 2.76 mmol) and NaN(SiMe3)2 (1.01 g,
5.5 mmol) were dissolved in THF (10 mL) and stirred at
room temperature for 5 h. A stirring suspension of CaI2
(0.79 g, 2.68 mmol) in THF was added dropwise to this sol-
ution and allowed to stir for 12 h. The solvent was removed
in vacuo and the product was dissolved in pentane and fil-
tered through celite. Removal of the volatiles in vacuo af-
1
forded 13 as a yellow solid (0.50 g, 9.23 mmol, 33%). H
NMR (CDCl3, ppm) d: 8.1 (1H, s, CH=N), 7.7 (1H, d, J =
3 Hz, Ar–H), 7.2 (1H, d, J = 3 Hz, Ar–H), 3.3 (2H, t, J =
6 Hz, CH2), 2.1 (2H, t, J = 6 Hz, CH2), 1.9 (4H, t, J = 5 Hz,
CH2), 1.8 (9H, s, C(CH3)3), 1.5 (9H, s, C(CH3)3), 1.2 (6H,
pentet, J = 5 Hz, CH2), 0.5 (18H, s, SiMe3). Elemental anal.
calcd for C25H44N3CaOSi (%): C 63.78, H 9.42, N 8.93;
found: C 63.70, H 9.50, N 8.62.
Results and discussion
Catalyst synthesis and characterization
Tridentate Schiff base ligands were readily synthesized via
Published by NRC Research Press

Gallaway et al.
423
Nði-PrÞ2
Fig. 5. ORTEP drawing (spheroids at 50% probability) of
½ONNꢀ2Znði-PrÞ2 (12). Protons and tert-butyl carbons are omitted for clarity.
1
Table 1. Selected bond distances (Å) and angles (°)
imine condensation reactions and characterized by H and
N(i‐Pr)2
for Q[ONN]2Zn (11) and
[ONN]2Zn (12).
13C NMR spectroscopy (eq. [1]). Six ligands were used in
this study, substituted with dimethylamine, diisopropylamine,
pyridine, quinoline, piperidine, and morpholine units.
Complex
11
12
Bond distance (Å)
Zn(1)–O(1)
2.0260(15)
2.0016(15)
2.1047(16)
2.2730(19)
2.1025(17)
2.2365(18)
1.9081(12)
1.9106(13)
2.0099(16)
—
2.0018(16)
—
Zn(1)–O(2)
Zn(1)–N(1)
Zn(1)–N(2)
½1ꢀ
Zn(1)–N(3)
Zn(1)–N(4)
Bond angle (°)
O(1)–Zn(1)–N(1)
O(1)–Zn(1)–N(2)
N(1)–Zn(1)–N(2)
O(2)–Zn(1)–N(3)
O(2)–Zn(1)–N(4)
N(3)–Zn(1)–N(4)
(O1)–Zn(1)–O(2)
86.20(6)
158.42(6)
75.36(7)
88.52(6)
163.54(6)
76.20(7)
91.00(6)
95.08(6)
—
—
96.15(6)
—
Zinc catalysts were prepared by the dropwise addition of
the proligand of choice to a stirring solution of bis(bis(tri-
methylsilyl)amide)zinc in THF. Filtration, drying, and recrys-
tallization from a 1:3 toluene–pentane mixture accessed the
desired products. Long recrystallizations promoted a transfor-
—
113.69(6)
R
mation from the desired [ONN]ZnN(SiMe3)2 complexes, ob-
R
Calcium complexes were prepared via the [ONN] sodium
salt with 1 equiv of ligand and 2 equiv of NaN(SiMe3)2 re-
acted for 5 h in THF followed by dropwise addition of CaI2
to the reaction mixture. Removal of solvent after 12 h fol-
lowed by filtration and recrystallization from pentane afforded
three desired complexes of the form R[ONN]CaN(SiMe3)2
characterized by 1H and 13C NMR spectroscopy (13–15,
Fig. 3). Yields were moderate compared with the corre-
sponding Zn complexes, potentially because of the relative
stability of the CaI2. Attempts to improve yields with longer
reaction times and higher temperatures were unsuccessful
and led to complex decomposition. Of note, pyridine and
quinoline functionalities were not employed as the complexes
had been previously tested, whereas morpholine substituents
promoted the formation of an intractable mixture of solids.
Again, the complex formation was noted in spectroscopic
R
served in crude NMR spectra, to the bis(ligand) [ONN]2Zn
species. This disproportionation reaction significantly low-
ered isolated yields of the desired products and in some
cases prevented the isolation of a pure sample of the de-
sired R[ONN]ZnN(SiMe3)2 catalyst. Improved yields of the
R[ONN]2Zn species could be obtained from purposefully
using a 2:1 ligand–metal ratio. Ligands 1–6 and isolated
complexes 7–12 are shown in Fig. 2. Complex formation
was readily identified by a diagnostic shift in the imine res-
onance to d 8.2 for alkyl-substituted nitrogen donor arms
and d 8.5 and 8.7 for pyridyl and quinolynyl donor arms,
respectively, as well as the disappearance of the hydroxy
signals. The integration and presence of N(SiMe3)2 protons
at d 0.1 was diagnostic for the formation of mono- or bis-
ligated complexes.
Published by NRC Research Press

424
Can. J. Chem. Vol. 90, 2012
[ONN]2Zn (12).
N(i‐Pr)2
Table 2. Selected crystallographic data and refinement details for Q[ONN]2Zn (11) and
Complex
Crystallographic parameters
Empirical formula
Formula mass
Colour, habit
Dimensions (mm)
Crystal system
Space group
11
12
C48H54N4O2Zn
784.32
Orange, rod
0.60×0.10×0.05
Triclinic
P-1
C46H78N4O2Zn
784.49
Colourless, rod
0.60×0.30×0.25
Monoclinic
P2(1)/c
Z
2
4
a, b, c (Å)
10.1627(17), 13.790(2), 16.511(3)
11.0836(13), 14.8067(18), 28.774(3)
a, b, g (°)
104.350(2), 104.649(2), 99.769(2)
2101.1(6)
188(1)
1.240
Mo Ka
90, 93.411(2), 90
4713.7(10)
188(1)
1.105
Mo Ka
V (Å3)
Temperature (K)
Dcalcd (Mg m–3)
Radiation
m (mm–1)
0.627
0.558
F(000)
832
1712
Reflections (obs.)
Reflections (ind.)
Data/restr./param.
Theta range (°)
Goodness-of-fit
R1, wR2
13846
8955
32145
10552
8955 / 15 / 520
1.57–27.49
1.052
0.0413, 0.0822
0.330, –0.518
10552 / 0 / 498
1.42–27.50
1.027
0.0369, 0.0872
0.284, –0.275
Largest peak/hole (e A–3)
studies by a shift in the imine resonance, the disappearance
of hydroxyl resonances, and the presence and integration of
N(SiMe3)2 signals.
esterification and broadened molecular weight distributions,
contrasting previous reports,15,21 but supporting the poor ac-
tivity of previously reported chiral variants.22 None of the
catalysts were effective at controlling the polymerization,
however, certain trends can be highlighted. Complexes with
a single ligand and bis(trimethylsilyl)amido initiating group
initiate a productive PCL polymerization, however, polydis-
persities (PDIs) are high and observed molecular weights are
higher than expected from monomer conversion. Complexes
with two ligands, lacking the amido functionality, are unable
to initiate the polymerization. The calcium complexes are sig-
nificantly faster at polymerizing the 3-caprolactone monomer.
Chromatographic analysis of these polymers still suggests a
lack of control over this reaction, with multimodal molecular
weight distributions. While disproportionation is possible for
the monoligated complexes under polymerization conditions,
which may explain the lack of control, no disproportionation
was observed for 7–9 or 13–15 in coordinating solvents.
Polymerizations of rac-lactide mediated by complexes 7–
15 are summarized in Table 4. Reactions were carried out at
70 °C in toluene for 30 or 240 min depending upon the ob-
served reactivity of the complexes. Clear trends in catalyst
performance are observed. Complexes 7–9, ligated by a sin-
gle tridentate ligand, produce PLAs of predictable molecular
weights and low polydispersities. The fastest catalyst, capable
of polymerizing nearly 100 equiv of rac-lactide in under
30 min, is the previously reported 7, substituted with a steri-
cally unencumbered dimethylamine donor arm. Increasing the
steric bulk slows the polymerization, as expected. Both the
rate of polymerization and the control of the polymerization
are inhibited by the morpholine ligand, suggesting the pendant
oxygen of the morpholine ring may block lactide coordina-
tion and ROP. Bis(ligand) complexes also initate and con-
trol the polymerization of lactide, likely inserting into one
Crystals suitable for X-ray diffraction studies were grown
for 11 and 12 by vapour diffusion of pentane into toluene.
The crystal structure of 11 (Fig. 4) indicates the complex
has a distorted octahedral geometry with coordination of the
quinoline nitrogens to afford a formally 22-electron complex.
This six-coordinate complex, with both tridentate ligands ar-
ranged in a meridional geometry, contrasts with similar four-
coordinate diphenolato complexes synthesized by Zhang et
al.23 In 11, Zn–N bonds ranging from 2.10 to 2.27 Å contrast
with the stronger Zn–O bonds of 2.00 and 2.03 Å. The struc-
ture of 12, however, shows the Zn centre in a distorted Td
geometry (Fig. 5), similar to a previously published bidentate
salicylaldimine structure.23 A four-coordinate structure is fav-
oured, correlating with the removal of the electron-withdrawing
quinoline ring structure and the increased steric hindrance
of the isopropyl substituents. As expected, the complex has
shorter Zn–O (1.91 Å) and Zn–N (2.00, 2.01 Å) bond
lengths compared with 11. Selected bond distances, bond
angles, and crystallographic parameters are provided in Ta-
bles 1 and 2.
Ring-opening polymerizations of rac-lactide and
3-caprolactone — Screening reactions
Polymerizations of 3-caprolactone mediated by complexes
7–15 are summarized in Table 3. Polymerizations with
monomer–catalyst ratios of 100:1 were carried out in toluene
(3 mL) at room temperature. No additional alcohol initiator
was added to the reactions, so initiation from latent trimethyl-
silylamide or ligand alkoxide functionalities was expected.
Temperatures were maintained at room temperature as higher
temperatures and bulk polymerization led to significant trans-
Published by NRC Research Press

Gallaway et al.
425
Table 3. Polymerization of 3-caprolactone (3-CL)
mediated by complexes 7–15.
times to achieve higher conversions. Due to the inferior qual-
ity of the catalysts compared with industry standards, full re-
action kinetics were not pursued. However, further evidence
of polymerization control is offered by variation of [M]:[I]
(monomer to initiator) ratios, as shown in Table 4 for catalyst
15. When polymerizations are quenched at the same conver-
sion, molecular weights increase linearly and correlate well
with predicted values. A slight increase in PDI at longer re-
action times suggest the catalysts may have limited thermal
stability. All catalysts produced atactic PLA and showed no
stereospecificity. Interestingly, the calcium complexes show
lower activity, potentially because of their larger size and
lower Lewis acidity, contributing to a lower overall reactivity.
Time
(min)
Conv.
(%)
54
72
28
0
0
0
68
81
72
a
[I]
7
8
Mn
Mn,th
PDI
1.71
1.72
1.79
na
na
na
1.73
1.80
1.72
120
120
120
360
360
360
15
37898
25150
51714
na
na
na
32841
21544
32583
6324
8378
3356
na
9
10
11
12
13
14
15
na
na
7922
9406
8218
15
15
Note: Reactions carried out at 25 °C in 3 mL of toluene
with monomer:catalyst ratios of 100:1. Molecular weights cor-
rected for changes in relative retention times versus styrene
standards. [I], initiator; Mn, number average molecular weight;
Conclusions
Nine zinc and calcium complexes supported by phenoxy-
imine ligands with various pendant donors were prepared,
characterized, and screened as catalysts in the ROP of
3-caprolactone and rac-lactide. Seven of these catalysts are
novel and include pendant piperidine, morpholine, diisopro-
pylamine, pyridine, and quinoline donors to compare with
first-generation catalysts substituted with dimethylamine do-
nors. Zinc complexes have a tendency to disproportionate to
form bis(ligand) species, which are ineffective for the polymer-
ization of 3-caprolactone but initiate slowly to produce PLAs
of controlled PDIs and molecular weights. Improved control
was achieved through the use of zinc and calcium complexes
coordinated to a single ligand framework with low PDI.
Mn,th, theoretical molecular weight; PDI, polydispersity index;
na, not available.
aMn,th = % conversion × MW 3-CL × 100 + MW N(SiMe3)2.
Table 4. Polymerization of rac-lactide mediated by com-
plexes 7–15.
Time
(min)
30
30
30
240
240
240
240
240
240
400
620
Conv.
(%)
a
[I]
7
8
Mn
Mn,th
PDI
1.25
1.28
1.61
1.39
na
1.31
1.20
1.20
1.12
1.25
1.34
86
61
48
94
0
98
54
47
58
58
58
10079
7374
6338
20508
na
19878
8321
5987
8986
18007
31998
12556
8952
7079
13872
na
14484
7943
6934
8520
17040
34080
9
10
11
12
13
14
15
15b
15c
Acknowledgement
We thank the Natural Sciences and Engineering Research
Council of Canada (NSERC) and the University of Prince
Edward Island, Charlottetown, for financial support, and the
Canada Foundation for Innovation and the Atlantic Canada
Opportunities Agency for infrastructure. We would also like
to thank NSERC for Undergraduate Student Research Awards
to Justin McRae and Josh Gallaway.
Note: Reactions carried out at 70 °C in 3 mL of toluene with
monomer–catalyst ratios of 100:1. Molecular weights corrected
for changes in relative retention times versus styrene standards.
[I], initiator; Mn, number average molecular weight; Mn,th, theo-
retical molecular weight; PDI, polydispersity index; na, not
available.
References
(1) Stanford, M. J.; Dove, A. P. Chem. Soc. Rev. 2010, 39 (2), 486.
doi:10.1039/b815104k.
(2) Cameron, D. J. A.; Shaver, M. P. Chem. Soc. Rev. 2011, 40 (3),
1761. doi:10.1039/c0cs00091d.
(3) Dove, A. P. Chem. Commun. (Camb.) 2008, (48): 6446.
doi:10.1039/b813059k.
aMn,th = % conversion × MW rac-LA × 200 + MW initiating
group.
bRatio of 200:1 monomer–catalyst used.
cRatio of 400:1 monomer–catalyst used.
(4) Dechy-Cabaret, O.; Martin-Vaca, B.; Bourissou, D. Chem. Rev.
2004, 104 (12), 6147. doi:10.1021/cr040002s.
(5) Sinclair, R. G. J. Macromol. Sci. Pure Appl. Chem. 1996, A33
(5), 585. doi:10.1080/10601329608010880.
(6) Davis, K. A.; Burdick, J. A.; Anseth, K. S. Biomaterials 2003,
24 (14), 2485. doi:10.1016/S0142-9612(02)00582-3.
(7) Dumitriu, S. Polymeric Biomaterials; Marcel Dekker: New
York, 2002; Chapter 4.
of the zinc alkoxide bonds to the ligand.23,24 Molecular
weights were higher than theoretical, suggesting not all li-
gands may initiate the polymerization. Interestingly, the
quinoline-substituted complex, which in the solid state had
no open coordination sites, is completely unreactive towards
lactide even at higher temperatures and longer reaction
times, suggesting this octahedral coordination mode may be
maintained in solution. No shift in lactide or catalyst reso-
nances are noted in H NMR spectra of 1:1 mixtures of 11
and rac-lactide, supporting this statement.
Finally, the calcium complexes prepared were all effective
in the polymerization of rac-lactide, with similar levels of
control regardless of the nature of the pendant donor arm.
Reactions were slow, however, requiring longer reaction
(8) Arbaoui, A.; Redshaw, C. Polym. Chem. 2010, 1 (6), 801.
1
doi:10.1039/b9py00334g.
(9) Minami, M.; Kozaki, S. Method for Producing Aliphatic
Polyester. U.S. Patent Appl. Publ. 20030023026 A1 20030130,
2003.
(10) Hayes, P. G.; Wheaton, C. A.; Ireland, B. J. Dalton Trans.
2009, (25), 4832. doi:10.1039/b819107g.
Published by NRC Research Press

426
(11) Tang, H.-Y.; Chen, H.-Y.; Huang, J.-H.; Lin, C.-C. Macro-
molecules 2007, 40 (25), 8855. doi:10.1021/ma071540k.
(12) Gao, A.-H.; Yao, W.; Mu, Y.; Gao, W.; Sun, M.-T.; Su, Q.
Polyhedron 2009, 28 (13), 2605. doi:10.1016/j.poly.2009.05.
037.
Can. J. Chem. Vol. 90, 2012
(18) SAINT 7.23A. Bruker AXS, Inc., Madison, WI, 2006.
(19) Sheldrick, G. SADABS. Bruker AXS, Inc., Madison, WI,
2008.
(20) Sheldrick, G. M. Acta Crystallogr. 2008, A64 (1), 112. doi:10.
1107/s0108767307043930.
(13) Darensbourg, D. J.; Choi, W.; Richers, C. P. Macromolecules
2007, 40 (10), 3521. doi:10.1021/ma062780n.
(14) Darensbourg, D. J.; Choi, W.; Karroonnirun, O.; Bhuvanesh,
N. Macromolecules 2008, 41 (10), 3493. doi:10.1021/
ma800078t.
(21) Farrugia, L. J. J. Appl. Cryst. 1997, 30 (5), 565. doi:10.1107/
S0021889897003117.
(22) Labourdette, G.; Lee, J.-C. D.; Patrick, B. O.; Ezhova, M. B.;
Mehrkhodavandi, P. Organometallics 2009, 28 (5), 1309.
doi:10.1021/om800818v.
(15) Darensbourg, D. J.; Karroonnirun, O. Inorg. Chem. 2010, 49
(23) Zhang, C.; Wang, Z.-X. J. Organomet. Chem. 2008, 693 (19),
(5), 2360. doi:10.1021/ic902271x.
(16) Darensbourg, D. J.; Karroonnirun, O. Macromolecules 2010,
43 (21), 8880. doi:10.1021/ma101784y.
3151. doi:10.1016/j.jorganchem.2008.07.002.
(24) Bhunora, S.; Mugo, J.; Bhaw-Luximon, A.; Mapolie, S.; Van
Wyk, J.; Darkwa, J.; Nordlander, E. Appl. Organomet. Chem.
2011, 25 (2), 133. doi:10.1002/aoc.1728.
(17) Kowalski, A.; Duda, A.; Penczek, S. Macromolecules 1998, 31
(7), 2114. doi:10.1021/ma971737k.
Published by NRC Research Press