Dinuclear zinc complexes using pentadentate phenolate ligands

Article abstract of DOI:10.1021/ic8014173

The synthesis and characterization of 15 dinuclear zinc complexes are reported, including the X-ray crystal structures of 5 complexes. The ligand motif utilizing p-cresol as a bridging unit between the two zinc centers and a set of three related ligands has been synthesized; 2,6-bis(R)-p-cresol (where R is CH₂NMe(CH₂)₂NMe₂ = L¹, CH=N(CH₂)₂NMe₂ = L², and CH ₂NH(CH₂)₂NMe₂ = L³). Dizinc trihalide complexes [LⁿZn₂(μ-X)X₂] (where X = Cl, Br, I) have been prepared. The trihalide complexes were treated with potassium ethoxide, resulting in quantitative substitution of the bridging halide group to give [LⁿZn₂(μ-OEt)X₂]. The dizinc ethoxide complexes were tested in situ as initiators for lactide ring opening polymerization. Dizinc triacetate complexes have also been synthesized [LⁿZn₂(OAc)₃], as well as cationic diacetate species containing two bridging acetate groups, [LⁿZn ₂(μ-OAc)₂][BPh₄]. Structural differences between complexes of the three ligands are discussed.

Full text of DOI:10.1021/ic8014173

Inorg. Chem. 2008, 47, 11711-11719
Dinuclear Zinc Complexes Using Pentadentate Phenolate Ligands
Paul D. Knight, Andrew J. P. White, and Charlotte K. Williams*
Department of Chemistry, Imperial College London, London, SW7 2AZ, U.K.
Received July 29, 2008
The synthesis and characterization of 15 dinuclear zinc complexes are reported, including the X-ray crystal structures
of 5 complexes. The ligand motif utilizing p-cresol as a bridging unit between the two zinc centers and a set of
three related ligands has been synthesized; 2,6-bis(R)-p-cresol (where R is CH2NMe(CH2)2NMe2 ) L1,
CHdN(CH2)2NMe2 ) L2, and CH2NH(CH2)2NMe2 ) L3). Dizinc trihalide complexes [LnZn2(µ-X)X2] (where X ) Cl,
Br, I) have been prepared. The trihalide complexes were treated with potassium ethoxide, resulting in quantitative
substitution of the bridging halide group to give [LnZn2(µ-OEt)X2]. The dizinc ethoxide complexes were tested in
situ as initiators for lactide ring opening polymerization. Dizinc triacetate complexes have also been synthesized
[LnZn2(OAc)3], as well as cationic diacetate species containing two bridging acetate groups, [LnZn2(µ-OAc)2][BPh4].
Structural differences between complexes of the three ligands are discussed.
Introduction
dizinc complexes have allowed new catalytic processes
to be developed that are beyond the scope of natural
enzymes. These processes include Mannich-type reac-
tions,5 Aldol reactions,6 Friedel-Crafts alkylations,7 alky-
nylation reactions,8 as well as polymerization catalysis.9
Over the past few decades, there has been much research
into bimetallic complexes due to the interesting properties
arising from the close proximity of the two metal centers.1
These include studies of bimetallic complexes’ magnetic,
electronic, spectroscopic, and structural properties, as well
as their reactivity.1,2 Binuclear zinc complexes have
attracted particular interest as important structural mimics
of the active site of a range of metalloenzymes,3 such as
zinc-dependent aminopeptidases, metallo-ꢀ-lactamases,
and alkaline phosphatases.4 In addition, these synthetic
(4) (a) Yamami, M.; Tanaka, M.; Sakiyama, H.; Koga, T.; Kobayashi,
K.; Miyasaka, H.; Ohba, M.; Okawa, H. Dalton Trans. 1997, 4545.
(b) Sakiyama, H.; Mochizuki, R.; Sugawara, A.; Sakamoto, M.;
Nishida, Y.; Yamasaki, M. Dalton Trans. 1999, 997. (c) Kaminskaia,
N. V.; Spingler, B.; Lippard, S. J. J. Am. Chem. Soc. 2000, 122, 6411.
(d) Erxleben, A.; Hermann, J. Dalton Trans. 2000, 569. (e) Abe, K.-
J.; Izumi, J.; Ohba, M.; Yokoyama, T.; Okawa, H. Bull. Chem. Soc.
Jpn. 2001, 74, 85. (f) Sakiyama, H.; Igarashi, Y.; Nakayama; Hossain,
M. J.; Unoura, K.; Nishida, Y. Inorg. Chim. Acta 2003, 351, 256. (g)
Chen, J.; Wang, X.; Zhu, Y.; Lin, J.; Yang, X.; Li, Y.; Lu, Y.; Guo,
Z. Inorg. Chem. 2005, 44, 3424. (h) Jikido, R.; Shiraishi, H.; Matsufuji,
K.; Ohba, M.; Furutachi, H.; Suzuki, M.; Okawa, H Bull. Chem. Soc.
Jpn. 2005, 78, 1795. (i) Sakiyama, H.; Ono, K.; Suzuki, T.; Tone, K.;
Ueno, T.; Nishida, Y. Inorg. Chem. Commun. 2005, 8, 372. (j)
Tamilselvi, A; Nethaji, M.; Mugesh, G. Chem. Eur. J. 2006, 12, 7797.
(k) Mitiæ, N.; Smith, S. J.; Neves, A.; Guddat, L. W.; Gahan, L. R.;
Schenk, G. Chem. ReV. 2006, 106, 3338.
(5) (a) Trost, B. M.; Jaratjaroonphong, J.; Reutrakul, V. J. Am. Chem.
Soc. 2006, 128, 2778. (b) Trost, B. M.; Lupton, D. W. Org. Lett. 2007,
9, 2023.
(6) Trost, B. M.; Shin, S.; Sclafini, J. A. J. Am. Chem. Soc. 2005, 127,
8602.
(7) Trost, B. M.; Mu¨ller, C. J. Am. Chem. Soc. 2008, 130, 2438.
(8) Trost, B. M.; Weiss, A. H.; von Wangelin, A. J. J. Am. Chem. Soc.
2006, 128, 8.
(9) (a) Williams, C. K.; Brooks, N. R.; Hillmyer, M. A.; Tolman, W. B.
Chem. Commun. 2002, 2132. (b) Breyfogle, L. E.; Williams, C. K.,
Jr.; Hillmyer, M. A.; Tolman, W. B. Dalton Trans. 2006, 928. (c)
Garner, L. E.; Zhu, H.; Hlavinka, M. L.; Hagadorn, J. R.; Chen, E. Y.-
X. J. Am. Chem. Soc. 2006, 128, 14822. (d) Champouret, Y. D. M.;
Nodes, W. J.; Scrimshire, J. A.; Singh, K.; Solan, G. A.; Young, I.
Dalton Trans. 2007, 4565.
* c.k.williams@imperial.ac.uk.
(1) Gavrilova, A. L; Bosnich, B. Chem. ReV. 2004, 104, 349.
(2) (a) Paulucci, G.; Vigato, P. A.; Rossetto, G.; Casellato, U. Inorg. Chim.
Acta 1982, L71. (b) Xiaoxeng, L.; Genglin, W.; Daizheng, L.; Shiping,
Y.; Zonghui, J. Polyhedron 1995, 14, 511. (c) Cheng, P.; Liao, D.;
Yan, S.; Jiang, S.; Wang, G. Polyhedron 1995, 14, 2355. (d) Mikuriya,
M.; Tashima, S. Polyhedron 1998, 17, 207. (e) Erxleben, A. Inorg.
Chem. 2001, 40, 208. (f) Adams, H.; Bradshaw, D.; Fenton, D. E.
Polyhedron 2002, 21, 1957. (g) Adams, H.; Bradshaw, D.; Fenton,
D. E. Inorg. Chim. Acta 2002, 332, 195. (h) Ye, B.-H.; Li, X.-Y.;
Williams, I. D.; Chen, X.-M. Inorg. Chem. 2002, 41, 6426. (i) Adams,
H.; Cummings, L. R.; Fenton, D. E.; McHugh, P. E. Inorg. Chem.
Commun. 2003, 6, 19. (j) Dutta, B.; Bag, P.; Flo¨rke, U.; Nag, K. Inorg.
Chem. 2005, 44, 147. (k) Shanmuga Bharathi, K.; Kalilur Rahiman,
A.; Rajesh, K.; Sreedaran, S.; Aravindan, P. G.; Velmurugan, D.;
Narayanan, V. Polyhedron 2006, 25, 2859. (l) Biswas, P.; Ghosh, M.;
Dutta, S. K.; Flo¨rke, U.; Nag, K. Inorg. Chem. 2006, 45, 4830. (m)
Ambrosi, G.; Formica, M.; Fusi, V.; Giorgi, L.; Guerri, A.; Micheloni,
M.; Paoli, P.; Pontellini, R.; Rossi, P. Inorg. Chem. 2007, 46, 309.
(n) Doyle, D. J.; Gibson, V. C.; White, A. J. P. Dalton Trans. 2007,
358.
(3) (a) Parkin, G. Chem. ReV. 2004, 104, 699. (b) Vahrenkamp, H. Dalton
Trans. 2007, 4751.
10.1021/ic8014173 CCC: $40.75  2008 American Chemical Society
Inorganic Chemistry, Vol. 47, No. 24, 2008 11711
Published on Web 11/18/2008

Knight et al.
thesis reactions14 with the corresponding dizinc trihalide
complexes [[LnZn2(µ-X)X2]. The dizinc alkoxide complexes
were applied as initiators in the ring opening polymerization
of lactide. In addition, the neutral acetate complexes,
[LnZn2(OAc)3], were targeted for catalytic testing in CO2/
epoxide copolymerizations, as well as being model com-
pounds to probe structural features of the complexes.
Scheme 1. Structures of HL1-3 and synthesis of the dizinc complexes
Results and Discussion
The preligands HL1-3 (Scheme 1) were prepared in high
yield by modifications of literature procedures (see Support-
ing Information for details).4d,15
[L1Zn2(µ-X)X2]. The synthesis of the dizinc trihalide
complexes (i.e., [LZn2(µ-X)X2], where X ) Cl, Br, I) was
achieved in three ways; simply by mixing the preligand and
the zinc halide species in an appropriate solvent; by adding
a base to this mixture; or by preforming the ligand alkali
salt, followed by a salt elimination reaction with the zinc
halide. All the methods were successful, but the metathesis
reaction generally led to better yields of high-purity material.
The syntheses of [L1Zn2(µ-Cl)Cl2] and [L1Zn2(µ-Br)Br2]
yielded white solids that had low solubility in methanol or
THF. The 1H NMR spectrum of [L1Zn2(µ-Cl)Cl2], in CDCl3,
shows a single aromatic peak at 6.75 ppm. The terminal
NMe2 groups are also observed as a single peak, as are the
internal NMe groups at 2.61 ppm and 2.16 ppm, respectively.
However, the aromatic CH2 protons resonate as a pair of
AB doublets at 4.36 and 3.11 ppm; indicating that the
complex is chiral. The remaining CH2 protons in the ethylene
diamine units are also diastereotopic and are observed as
two pairs of multiplets between 2.86 and 3.00 ppm and 2.45
A common motif in the design of dinucleating ligands
involves an alkoxide or phenoxide group that bridges the
two metal centers, and the remainder of the ligand construct
usually requires further coordinating groups to retain the
metal atoms and prevent formation of multinuclear arrays.
Some examples of this type of ligand are shown in Scheme
1, and the zinc complexes derived from L1-3 will be the focus
of this study. Dinuclear zinc complexes of these ligands4d,e,h
and related systems2g,h,k,4b,f,i,9d,10 have been studied previ-
ously to generate cationic acetate complexes of the type
[LZn2(µ-OAc)2]+. Some neutral complexes, for example
using nitrate or isothiocyanate coligands, have also been
studied;2i,4c and in general the syntheses are undertaken in
aqueous environments. Although these complexes have been
tested in biologically relevant reactions (vide supra), we were
interested in studying anhydrous, neutral species for lactide
ring opening polymerization and CO2/epoxide copolymeri-
zation catalysis, which have both been demonstrated previ-
ously with zinc-based systems.9a,b,11,12 For such catalysis,
an alkoxide or carboxylate coligand is generally required for
the initiation step. Examples of monomeric dizinc alkoxide
complexes are rare,4c,13 but one such complex based on HL1
and synthesized by Williams et al.,9a,b has been shown to
have high catalytic reactivity in lactide ring-opening polym-
erization. However, the synthetic route to the reported
complex, [L1Zn2(µ-OEt)Cl2], was rather complex and of
limited generality. An important goal of this study was to
develop simpler and more efficient routes to dinuclear zinc
alkoxide complexes by utilizing potassium alkoxide meta-
1
and 2.56 ppm. The H NMR spectrum of the bromide
complex is very similar, and both complexes are symmetrical
on the NMR time scale, at 298 K. Using a similar method
to synthesize [L1Zn2(µ-I)I2] failed to yield a precipitate from
methanol or THF, and only an off-white/yellow material
could be obtained upon solvent removal. Recrystallization
(12) (a) Dechy-Cabaret, O.; Martin-Vaca, B.; Bourissou, D. Chem. ReV.
2004, 104, 6147. (b) Platel, R. H.; Hodgson, L. M.; Williams, C. K.
Polym. ReV. 2008, 48, 11. (c) Cheng, M.; Attygale, A. B.; Lobkovsky,
E. B.; Coates, G. W. J. Am. Chem. Soc. 1999, 121, 11583. (d)
Chamberlain, B. M.; Cheng, M.; Moore, D. R.; Ovitt, T. M.;
Lobkovsky, E. B.; Coates, G. W. J. Am. Chem. Soc. 2001, 123, 3229.
(e) Chisholm, M. H.; Eilerts, N. W.; Huffman, J. C.; Iyer, S. S.; Pacold,
M.; Phomphrai, K. J. Am. Chem. Soc. 2000, 122, 11845. (f) Chisholm,
M. H.; Huffman, J. C.; Phomphrai, K. J. Chem. Soc., Dalton Trans.
2001, 222. (g) Chisholm, M. H.; Gallucci, J. C.; Zhen, H. H.; Huffman,
J. C. Inorg. Chem. 2001, 40, 5051. (h) Chisholm, M. H.; Gallucci, J.;
Phomphrai, K. Inorg. Chem. 2002, 41, 2785. (i) Williams, C. K.;
Breyfogle, L. E.; Choi, S. K.; Nam, W. W., Jr.; Hillmyer, M. A.;
Tolman, W. B. J. Am. Chem. Soc. 2003, 123, 11350. (j) Chen, H. Y.;
Tang, H. Y.; Lin, C. C. Macromolecules 2006, 39, 3745. (k) Huang,
B. H.; Lin, C. N.; Hsueh, M. L.; Athar, T.; Lin, C. C. Polymer 2006,
47, 6622. (l) Lian, B.; Thomas, C. M., Jr.; Lehmann, C. W.; Roisnel,
T.; Carpentier, J. F. Inorg. Chem. 2007, 46, 328.
(13) (a) Drago, R. S.; Desmond, M. J.; Cordon, B. B.; Miller, K. A. J. Am.
Chem. Soc. 1983, 105, 2287. (b) Fondon, M; Garcia-Deibe, A. M.;
Ocampo, N.; Sanmartin, J.; Bermejo, M. R. Dalton Trans. 2004, 2135.
(c) Hlavinka, M. L.; McNevin, M. J.; Shoemaker, R.; Hagadorn, J. R.
Inorg. Chem. 2006, 45, 1815.
(14) Cheng, M.; Moore, D. R.; Reczek, J. J.; Chamberlain, B. M.;
Lobkovsky, E. B.; Coates, G. W. J. Am. Chem. Soc. 2001, 123, 8738.
(15) Higuchi, C.; Sakiyama, H.; Okawa, H.; Fenton, D. E. Dalton Trans.
1995, 4015.
(10) (a) Asato, E.; Chinen, M.; Yoshitno, A.; Sakata, Y; Sigiura, K.-I. Chem.
Lett. 2000, 678. (b) Fenton, D. E. Inorg. Chem. Commun. 2002, 5,
537.
(11) (a) Sugimoto, H.; Inoue, S. J. Polym. Sci. Part A: Polym. Chem. 2004,
42, 5561. (b) Coates, G. W.; Moore, D. R. Angew. Chem., Int. Ed.
2004, 43, 6618. (c) Schenk, S.; Notni, J.; Ko¨hn, U.; Wermann, K.;
ers, E. Dalton Trans. 2006, 4191. (d) Darensbourg, D. J.; Lewis, S. J.;
Rodgers, J. L.; Yarbrough, J. C. Inorg. Chem. 2003, 42, 581. (e) van
Meerendonk, W. J.; Duchateau, R.; Koning, C. E.; Gruter, G.-J. M.
Macromolecuels 2005, 38, 7306. (f) Bok, T.; Yun, H.; Lee, B. Y.
Inorg. Chem. 2006, 45, 4228. (g) Xiao, Y.; Wang, Z.; Ding, K.
Macromolecules 2006, 39, 2006. (h) Liu, B.; Tian, C.; Zhang, L.; Yan,
W.; Zhang, W. J. Polym. Sci. Part A: Polym. Chem. 2006, 44, 6243.
(j) Pilz, M. F.; Limberg, C.; Lazarov, B. B.; Hultzsch, K. C.; Ziemer,
B. Organometallics 2007, 26, 3668. (i) Sugimoto, H.; Ogawa, A. React.
Funct. Polym. 2007, 67, 1277.
11712 Inorganic Chemistry, Vol. 47, No. 24, 2008

Dinuclear Zinc Complexes Using Pentadentate Phenolate
of the crude material yielded a microcrystalline solid.
However, the 1H NMR spectrum shows multiple broadened
peaks and indicates more than one species is present in
solution. The aromatic region contains one major and at least
two minor peaks, all of differing integrations. Three major
peaks are observed for the NMe2, NMe, and ArMe groups,
which are slightly broadened, but several smaller peaks are
also observed in similar regions, and at least two pairs of
doublets are assigned to the CH2 protons. Nevertheless,
elemental analysis was in agreement with the formula
[L1Zn2I3], and the FAB mass spectrum shows a group of
peaks at 721 amu, assigned to [M-I]+ (and is consistent
with the FAB mass spectrum of the chloride and bromide
1
complexes). The H NMR spectrum of [L1Zn2(µ-I)I2] is
rationalized by different structural isomers of the complex;
the two coordinated NMe groups can adopt different
conformations (i.e., RR, SS, RS, and SR); the five-membered
chelate rings may also adopt different conformations, and
further possible isomers may arise from the coordination
modes around the zinc centers. These conformations are
usually expected to rapidly interconvert, but may be slowed
in the iodide complex due to the increased size of the iodide
Figure 1. The molecular structure of [L1Zn2(µ-OEt)I2] (50% probability
ellipsoids).
1
groups compared with chloride and bromide. VT H NMR
spectroscopy (d2-TCE) showed convergence at 368 K, and
averaged broadened resonances were observed.
[L1Zn2(µ-OEt)X2]. Following the successful syntheses of
the dizinc trihalide complexes, the alkoxide species [L1Zn2(µ-
OEt)X2] were prepared in quantitative yield by a salt
metathesis reaction between the bridging halide group and
potassium ethoxide.14 The reactions were carried out under
nitrogen, in Youngs tap NMR tubes using CD3CN as solvent,
Figure 2. Projection of the molecular structure of [L1Zn2(µ-OEt)I2] (left)
and [L1Zn2(µ-OEt)Cl2]9a (right) along the phenolate axis.
Table 1. Selected Bond Lengths (Å) and Angles (°) for
[L1Zn2(µ-OEt)I2]
Zn(1)sN(8)
Zn(1)sO(1)
Zn(1)sO(30)
Zn(1)· · ·Zn(2)
2.194(2)
Zn(2)sN(17)
2.396(2)
1
and were monitored by H NMR spectroscopy. In the case
2.0885(17) Zn(2)sO(1)
1.9758(17) Zn(2)sO(30)
3.1700(6) O(1)sC(1)
2.0008(17)
2.0375(18)
1.344(3)
of [L1Zn2(µ-Cl)Cl2], a reaction was observed within 30 min
to generate a mixture of new species that had significant
broadening of all the peaks in the 1H NMR spectrum. Upon
comparing the integration of the aromatic protons with the
CH2, NMe, and ArMe integrals, as well as those of the new
ethoxide groups (at ca. 1.19 ppm and 3.92 ppm), we proposed
the formation of the desired species [L1Zn2(µ-OEt)Cl2]. This
was confirmed by removal of the solvent and addition of
Zn(1)sO(1)sZn(2) 101.62(7)
Zn(1)sO(30)sZn(2) 104.34(8)
Table 2. Selected Bond Lengths (Å) and Angles (°) for the Two
Independent Complexes ([L2Zn2(µ-I)I2]-I and [L2Zn2(µ-I)I2]-II) Present
in the Crystals of [L2Zn2(µ-I)I2]
isomer I
isomer II
Zn(1)sO(1)
Zn(2)sI(3)
Zn(2)sN(14)
Zn(1)· · ·Zn(2)
Zn(1)sI(3)
2.033(2)
2.6951(5)
2.064(3)
3.4248(5)
2.8006(5)
2.086(3)
2.071(3)
2.6841(5)
2.059(3)
3.4591(6)
2.7221(5)
2.074(4)
1
CD2Cl2, the resulting H NMR spectrum was in agreement
with the structure previously reported.9a The reactions of
[L1Zn2(µ-Br)Br2] and [L1Zn2(µ-I)I2] with potassium ethoxide
in CD3CN proceeded in a similar manner to that of [L1Zn2(µ-
Cl)Cl2]. After 30 min, several new species were present in
the NMR spectra and showed similar features to the spectrum
of [L1Zn2(µ-OEt)Cl2]. Upon leaving the NMR tube reaction
of [L1Zn2(µ-I)I2] and KOEt, crystals of [L1Zn2(µ-OEt)I2] were
obtained.
Zn(1)sN(7)
Zn(2)sO(1)
2.141(3)
2.130(3)
O(1)sC(1)
1.319(4)
1.320(4)
Zn(1)sI(3)sZn(2)
Zn(1)sO(1)sZn(2)
77.070(13)
110.24(12)
79.556(15)
110.86(12)
When viewed along the phenolate bond (Figure 2), the
C(7)sN(8) and C(16)sN(17) bonds of the pendant arms of
the molecule are essentially perpendicular to the plane of
the aromatic ring, giving the molecule a distinct chirality.
The zinc centers are both five coordinate with distorted
square pyramidal geometries. The bond distances (Table 1)
between the phenolate oxygen atom (O(1)) and the two zinc
centers are somewhat different, with a shorter phenolate bond
(2.0008(17) Å) to Zn(2). The bond distances between the
ethoxide oxygen (O(30)) and the two zinc centers are also
different, with a shorter ethoxide bond (1.9758(17) Å) to
X-rayCrystallography.Themolecularstructureof[L1Zn2(µ-
OEt)I2] is shown in Figure 1. With the exception of the
ethoxide bridging unit (based on O(30)), the complex shows
approximate C2 symmetry about an axis coincident with the
C(1)-O(1) vector. The two N,N′ five-membered chelate rings
have the same configuration (both λ in Figure 1). (The
complex crystallized in a racemic space group, so the crystal
contains equal numbers of the molecule shown in Figure 1
and its enantiomer).
Inorganic Chemistry, Vol. 47, No. 24, 2008 11713

Knight et al.
Zn(1). The distance between the two zinc atoms is 3.1700(6)
Å, and both coordinated NMe groups (N(8) and N(17)) have
the same configuration (S in Figure 1). The structure is
closely related to that of [L1Zn2(µ-OEt)Cl2], reported by
Williams et al.9a (Figure 2). However, for the chloride
complex, the five-membered chelate rings have differing
configurations (i.e., one ring is δ and the other λ), and this
removes the rotational symmetry of the molecule. Although
this configurational difference between the two structures
may only be a solid-state packing effect (the chloride
structure also contains a molecule of THF in the unit cell),
it serves to highlight the flexibility of the ligand and the
resulting complexity of the NMR spectra. The distance
between the two zinc atoms in the chloride complex (3.176
Å) is similar to that in the iodide complex (3.1700(6) Å).
1
Although the H NMR spectra and the X-ray crystal
structures indicated formation of the desired ethoxide
1
complexes, [L1Zn2(µ-OEt)X2], continued monitoring by H
NMR spectroscopy showed that these species were not stable
in solution. Although the rate of change was slow at room
temperature (several days to see formation of new peaks),
at elevated temperatures the process occurred within several
hours to give a mixture of products, including some of the
starting trihalide species [L2Zn2(µ-X)X2] (particularly in the
case of X ) Cl). Nonetheless, the methathesis route to this
type of bridging alkoxide dizinc complex is more straight-
forward and higher yielding than the alternative route
described by Williams et al.9a
[L2Zn2(µ-X)X2]. Because the NMe groups in the pendant
arms in L1 allow for the formation of various diastereoiso-
mers, it was envisaged that removal of this group would
simplify the NMR spectra of the resulting complexes. Thus,
HL2 (Scheme 1) was targeted for further study.
The halide complexes, [L2Zn2(µ-X)X2], were prepared in
a similar manner to those of L1 and were obtained in good
yields (74-96%). The chloride and bromide complexes were
both bright yellow solids, whereas the iodide complex was
orange/yellow, and all three complexes had reduced solubility
compared to the L1 analogues. Unlike the complexes of L1,
all three halide complexes had very similar 1H NMR spectra.
In CD2Cl2, a single imine peak is observed at 8.38-8.40
ppm, and a single aromatic peak is found at ca. 7.26 ppm.
The CH2 groups are observed as broad peaks at 3.74-3.80
ppm and 2.86-2.94 ppm, although the latter peak is
extremely broad. Two sharp singlets occur at 2.56-2.58 ppm
and 2.24-2.32 ppm and are assigned to the NMe2 groups
and ArMe group, respectively. The 1H NMR spectrum
indicates that the complexes exist as a single symmetric
species in solution, even in the case of the iodide complex.
However, the broad nature of the CH2 groups suggests that
there may be significant fluxionality in the configuration of
Figure 3. The molecular structures of (a) [L2Zn2(µ-I)I2]-I and (b) [L2Zn2(µ-
I)I2]-II, the two crystallographically independent complexes present in the
crystals of [L2Zn2(µ-I)I2] (both 50% probability ellipsoids).
X-ray Crystallography. The crystal structure of [L2Zn2(µ-
I)I2] contains two independent complexes ([L2Zn2(µ-I)I2]-I
and [L2Zn2(µ-I)I2]-II) with different configurations of the
N(14)/N(17) N,N′ five membered chelate ring. In complex
[L2Zn2(µ-I)I2]-I, the chelate rings have opposite configura-
tions (λ and δ respectively, Figure 3a), whereas in complex
[L2Zn2(µ-I)I2]-II the two rings have the same configuration
(both λ, Figure 3b). (It is important to note that the complex
crystallized in a racemic space group, so the crystal contains
equal numbers of the complexes shown in Figure 3a and b,
and their enantiomers). The isomer [L2Zn2(µ-I)I2]-I is not
1
observed in solution, as shown by the H NMR spectra,
which show a single symmetrical species, even at 193 K.
As expected, in both independent complexes the C(7)sN(7)
and C(14)sN(14) bonds lie much closer to the plane of the
phenolate unit than was seen for their counterparts in the
structure of [L1Zn2(µ-OEt)I2, due to the constraints of the
imine groups. The distances between the zinc centers
(3.4248(5) and 3.4591(6) Å in complexes [L2Zn2(µ-I)I2]-I
and [L2Zn2(µ-I)I2]-II respectively) are significantly longer
than those observed for [L1Zn2(µ-OEt)I2]. This effect prob-
1
the N,N′ chelate rings. Indeed, low temperature H NMR
studies of [L2Zn2(µ-I)I2], in CD2Cl2, show that the broad
signals associated with the CH2 groups sharpen and split at
223 K, to give four multiplets at 4.02, 3.53, 3.18, and 2.60
ppm. In addition, the NMe2 groups also split to give two
sharp peaks, of 6H each, at 2.53 and 2.47 ppm, which is
consistent with a chiral, symmetrical species.
11714 Inorganic Chemistry, Vol. 47, No. 24, 2008

Dinuclear Zinc Complexes Using Pentadentate Phenolate
1
ably relates to the improved donation from the imine versus
amine groups, supported by the reduction in the length of
the ZnsN(imine) bonds (between 2.059(3) and 2.086(3) Å)
c.f. the ZnsN(amine) counterparts in [L1Zn2(µ-OEt)I2]
(2.194(2) and 2.396(2) Å), and the associated ca. 9° widening
of the Zn(1)sO(1)sZn(2) angle.
of various isomers. VT H NMR (d2-TCE) shows conver-
gence of the peaks at 383 K, to give an averaged set of
broadened signals.
[L3Zn2(µ-OEt)Br2]. [L3Zn2(µ-Br)Br2] was treated with
KOEt in a Youngs tap NMR tube, using CD3CN as solvent.
1
After 30 min the H NMR spectra showed the complete
[L2Zn2(µ-OEt)X2]. In a similar manner to the L1 com-
disappearance of the starting material and the formation of
several isomeric forms of [L3Zn2(µ-OEt)Br2]. The spectrum
is similar to those observed for [L1Zn2(µ-OEt)X2], although
more peaks are observable in the aromatic region, which may
be due to an increase in the number of isomers allowed by
this less sterically hindered ligand. The integral of the
aromatic region compared to the aliphatic and ethoxide
regions is consistent with the formation of the desired species
in quantitative yield.
[L1-3Zn2(OAc)3]. The bimetallic zinc acetate complexes
were prepared by treatment of the alkali salt of the ligands
with Zn(OAc)2. Using HL1, a white solid was obtained. The
1H NMR spectrum of this material is very broad, but shows
a singlet at 6.76 ppm for the aromatic protons and a singlet
at 2.18 ppm for the aromatic methyl group. There is a sharp
singlet at 1.94 ppm that integrates to nine protons, and this
has been assigned to the methyl protons for the three acetate
groups. There are several very broad peaks between 2.55
and 4.40 ppm, which are attributed to the CH2 protons and
four broad peaks between 2.19 and 2.47 ppm, which have
been assigned to the NMe groups. This compound was
fluxional at room temperature, and the singlet observed for
the acetate methyl protons suggests the differing acetate
groups are in rapid exchange. VT-NMR spectroscopy showed
that as the temperature was reduced some of the peaks began
to sharpen, whereas others broadened and shifted. At 213
K, at least three peaks were seen in the aromatic region, and
at least four peaks were observed for the acetate methyl
plexes, the reactions of [L2Zn2(µ-X)X2] with potassium
1
ethoxide were followed by H NMR spectroscopy, using
CD3CN as the solvent and the compounds produced in
quantitative yield. For [L2Zn2(µ-Cl)Cl2], a reaction occurred
within an hour, and two new species were observed in the
NMR spectrum, both of which were due to symmetrical
products. The major species, ca. 85%, shows a new quartet
at 3.97 ppm and a triplet at 1.36 ppm, assigned to a zinc-
bound ethoxide group. The minor species also has zinc-
ethoxide resonances at 3.53 and 1.11 ppm. The remaining
peaks, that is, two new imine, aromatic, NMe2, and ArCH3
peaks, are all pairs of singlets in the ratio of 85:15. The CH2
groups remain broad and overlap for the two species. The
1H NMR spectra of the products of the reactions of [L2Zn2(µ-
Br)Br2] and [L2Zn2(µ-I)I2] with potassium ethoxide were
similar to those of [L2Zn2(µ-Cl)Cl2], although the ratios of
the two symmetrical species were approximately 1:1 with
the bromide complex and 3:7 with the iodide complex.
The precise structures of the two symmetrical alkoxide
species remain unclear. The fluxionality of [L1Zn2(µ-OEt)X2]
complexes has been noted, but the increased rigidity of L2
is expected to increase the energetic barrier for interconver-
sion of different conformations in the analogous complexes.
If the [L2Zn2(µ-OEt)X2] complexes show related structures
to L1, then the two halide atoms could be on opposite sides
of the plane of the molecule or on the same face. However,
the X-ray crystal structures observed for this type of complex
only show the halide groups on opposite sides of the plane
of the molecule. Alternatively, different N,N′ chelate ring
conformations could account for the two symmetrical species
observed, although the barrier to conversion is expected to
be significantly lower and is also likely to be the source of
the broadening of the CH2 peaks in the 1H NMR spectra. In
any event, a significant barrier to conversion exists since
prolonged heating is required to cause changes in the relative
ratios of the two species. Nevertheless, these complexes do
not suffer from the decomposition or reformation of the
trihalide species that were observed for the L1 complexes.
[L3Zn2(µ-X)X2]. The ligand flexibility was increased
compared to L1 and L2 by synthesizing the related ligand,
L3 (Scheme 1). The zinc halide complexes, [L3Zn2(µ-X)X2],
were synthesized by the same methods previously described
and were white solids. The 1H NMR spectra were similar to
[L1Zn2(µ-X)X2]; single peaks are observed for the aromatic,
NMe2, and the aromatic methyl protons, where X is chloride
or bromide, and indicate that a single, symmetric species
exists in solution, in each case. The peaks associated with
the various differing CH2 protons have increased multiplicity
due to the presence of the NH protons. In the case of the
1
groups. High temperature H NMR studies (d2-TCE) did
simplify the spectrum, with coalescence of the peaks
occurring to generate averaged resonances, although these
peaks remained broad to 383 K. Nevertheless, elemental
analysis and FAB mass spectrometry were in agreement with
the structural formula of [L1Zn2(OAc)3].
Using L2, gave a yellow/orange solid and simplified the
analysis of the H NMR spectrum. The peaks are all sharp,
including the CH2 groups that occur as two triplets at 3.70
and 2.80 ppm. This is in contrast to the halide complexes of
L2, where the CH2 resonances are very broad in all cases.
Sharp singlets are observed for the imine, aromatic, NMe2,
and ArMe protons, and only a single peak is observed for
the three acetate methyl groups. The complex was sym-
metrical, and the single peak for the acetate groups indicates
a rapid exchange process for this functional group. Reduced
1
1
temperature H NMR studies (CD2Cl2) did not alter the
spectrum, apart from some broadening of the peaks, and only
one acetate methyl group (integrating to nine protons) was
observable down to 183 K.
The reaction of HL3 with zinc acetate yielded a white solid.
As with [L1Zn2(OAc)3], the H NMR spectrum contained
1
1
iodide complex, and as with L1, the H NMR spectrum of
broad resonances. A singlet is observed for the aromatic
protons, as well as the aromatic methyl group and the three
[L3Zn2(µ-I)I2] is significantly broadened and shows evidence
Inorganic Chemistry, Vol. 47, No. 24, 2008 11715

Knight et al.
acetate groups. Two broad peaks are observed between 2.18
and 2.45 ppm, which integrate to 14 protons and are assigned
to the NMe2 groups and a CH2 group. The series of at least
seven broad peaks between 2.55 and 4.08 ppm are assigned
to the remaining CH2 and NH protons. Reduced-temperature
1H NMR studies caused the peaks to sharpen, but several
isomers were observed, although the acetate groups were
1
differentiated at 193 K. High-temperature H NMR studies
showed coalescence of the peaks at 333 K to give a spectrum
consistent with the formula [L3Zn2(OAc)3]. Continued heat-
ing to 383 K caused a significant sharpening of the peaks.
In all three acetate complexes only a single peak for the
1
acetate groups was observed in the H NMR spectra, at
room temperature. At least two different chemical environ-
ments are expected, depending on the number of bridging
acetate units, but the NMR data indicates that the acetate
groups are in rapid exchange around the zinc centers.
Previously, zinc acetate complexes using this class of ligand
have resulted in the formation of cationic species with a
noncoordinating counterion (e.g., [ClO4]-, [PF6]-, or
[BF4]-).2g,h,k,4b,d-f,h,i,9d,10 Indeed, to our knowledge there is
only one example of a complex containing two zinc centers
and three coordinated acetate groups, only one of which is
bridging.4j Therefore, it seems likely that the acetate com-
plexes using L1-3 are cationic and contain two zinc centers
bridged by two acetate groups. The third acetate group would
be acting as a counterion (either weakly bound or noncoor-
dinating), and is in rapid exchange with the coordinated
acetate groups.
Figure 4. The molecular structure of [L1Zn2(µ-OAc)2][BPh4] (50%
probability ellipsoids).
NMe2 protons are observed as two singlets of six protons
each, highlighting the chiral but conformationally stable
nature of this complex.
X-ray Crystallography. The solid-state structures of
[L1-3Zn2(µ-OAc)2][BPh4] show that the complex cations have
approximate C2 symmetry about an axis coincident with the
C(1)sO(1) vector. In all the complexes, the two five-
membered N,N′ chelate rings have the same chirality within
each complex.16 In Figure 4 both of the N,N′ chelate rings
have λ twist geometries, whereas in Figures 5 and 6 the rings
each have a δ twist geometry. However, as all the complexes
crystallized in a centrosymmetric space group, there are equal
numbers of the opposite enantiomers present in the crystals.
As expected, the rigid imine groups in the complex of ligand
L2 (Figure 5) lie essentially coplanar with the phenolate ring.
In the complexes of L1 and L3, however, the equivalent CsN
single bonds are approximately orthogonal to this plane and
on opposite sides, giving the complexes a chirality. It is
notable that the specific enantiomers illustrated in Figures 4
and 6 have the same configuration for this chirality while
having opposite chirality for the N,N′-five membered chelate
rings (vide supra). It is also interesting to note, however,
that the chirality at the amine nitrogens (N(8)/N(17) in
[L1Zn2(µ-OAc)2]+, and N(8)/N(16) in [L3Zn2(µ-OAc)2]+),
while the same within each complex, differ between the two
ligands, being S,S in Figure 4 and R,R in Figure 6. Only one
diastereoisomer of [L1Zn2(OAc)2][BPh4] and [L3Zn2(OAc)2]
[BPh4] are observed in the NMR spectra, thus highlighting
[L1-3Zn2(µ-OAc)2][BPh4]. It was relevant to determine
whether an acetate group could be exchanged with a
noncoordinating counterion, and for these studies the tet-
raphenylborate anion was selected as it allows monitoring
1
of the complexation by H NMR spectroscopy. The com-
plexes were readily synthesized by addition of sodium
tetraphenylborate to a solution of HLn and zinc acetate, in
methanol, and the pure complexes were obtained in good
yields (88-94%). The reaction using HL1 yielded a white
solid, and the 1H NMR spectrum displayed very sharp peaks,
which is in stark contrast to the spectrum of [L1Zn2(OAc)3].
The tetraphenyl borate counterion signals are observed at
7.39, 7.01, and 6.87 ppm and integrate to 8, 8, and 4 protons,
respectively. Single peaks are observed for the aromatic
ligand protons, the NMe groups, the aromatic methyl group,
as well as the acetate methyl protons, which integrate to six
protons. Interestingly, the NMe2 protons are observed as two
1
singlets of six protons each, at 2.31 and 1.90 ppm. This H
NMR data indicates that the complex is chiral, symmetrical,
and conformationally rigid.
The synthesis of [L2Zn2(µ-OAc)2][BPh4] yielded a yellow/
1
orange solid. The H NMR spectrum was also sharp and
indicated that the expected complex was present. The NMe2
groups, in this instance, were observed as a broad singlet,
and this may be due to the increased rigidity of the ligand
generating a planar structure. The reaction using HL3 yielded
a white solid and, as for the previous cationic complexes,
(16) In the structure of [L2Zn2(µ-OAc)2][BPh4], two orientations were found
for the N(16)/N(19) five-membered chelate ring, one with the same
configuration as the N(8)/N(11) chelate ring (ca. 66%) and one with
the opposite configuration (ca. 34%). Complexes with the minor
occupancy configuration of the N(16)/N(19) chelate ring clearly do
not possess the approximate C2 symmetry seen for complexes with
the major occupancy configuration. The two configurations have
slightly different positions for the N(19) donor atom, and only the
major occupancy position has been considered in the discussion.
1
the peaks in the H NMR spectrum were sharp and in
agreement with the formula [L3Zn2(µ-OAc)2][BPh4]. The
11716 Inorganic Chemistry, Vol. 47, No. 24, 2008

Dinuclear Zinc Complexes Using Pentadentate Phenolate
Table 3. Selected Bond Lengths (Å) and Angles (°) for
[L1Zn2(µ-OAc)2][BPh4]
Zn(1)sO(1)
1.9666(10)
Zn(1)sN(8)
2.2176(13)
2.1538(12)
1.9670(11)
3.1822(2)
1.9669(10)
2.1868(13)
1.9626(11)
2.1435(11)
1.3535(17)
Zn(1)sO(25)
Zn(1)sO(30)
Zn(1)· · ·Zn(2)
Zn(2)sO(1)
Zn(2)sN(17)
Zn(2)sO(27)
Zn(2)sO(32)
O(1)sC(1)
Zn(1)sO(1)sZn(2)
107.99(5)
Table 4. Selected Bond Lengths (Å) and Angles (°) for
[L2Zn2(µ-OAc)2][BPh4]
Zn(1)sO(1)
2.1035(13)
Zn(1)sN(8)
2.0215(16)
1.9727(14)
1.9882(15)
3.2620(4)
2.1028(13)
2.0096(17)
1.9690(15)
1.9741(16)
1.315(2)
Zn(1)sO(25)
Zn(1)sO(30)
Zn(1)· · ·Zn(2)
Zn(2)sO(1)
Zn(2)sN(16)
Zn(2)sO(27)
Zn(2)sO(32)
O(1)sC(1)
Figure 5. The molecular structure of [L2Zn2(µ-OAc)2][BPh4] (30%
probability ellipsoids).
Zn(2)sO(1)sZn(1)
101.70(6)
Table 5. Selected Bond Lengths (Å) and Angles (°) for
[L3Zn2(µ-OAc)2][BPh4]
Zn(1)sO(1)
2.035(2)
Zn(1)sN(8)
2.088(3)
1.978(3)
1.996(3)
3.2580(5)
2.031(2)
2.089(3)
1.987(3)
1.976(3)
1.342(4)
106.51(11)
Zn(1)sO(25)
Zn(1)sO(30)
Zn(1)· · ·Zn(2)
Zn(2)sO(1)
Zn(2)sN(16)
Zn(2)sO(27)
Zn(2)sO(32)
O(1)sC(1)
Zn(1)sO(1)sZn(2)
which are in turn longer than those in [L1Zn2(µ-OAc)2][BPh4]
(1.9666(10) and 1.9669(10) Å). This is accompanied by a
much longer Zn(1)-N(8) (2.2176(13) Å) distance in L1,
compared to the Zn(1)-N(8) (2.0215(16) Å) distance of L2,
and may be due to the increased constraint and coordinating
strength of the imine units. The bond distances for the similar
bonds of the L3 complex are almost midway for those of L1
and L2. The angles around the zinc centers in the L2 and L3
complexes are very similar, whereas those for the L1 complex
are substantially different (see the Supporting Information),
presumably due to the steric pressure from the N(8) and
N(17) methyl groups.
Lactide Ring Opening Polymerization. Zinc alkoxide
complexes have good precedent as initiators for the controlled
polymerization of lactide.12 In fact, zinc initiators have shown
some of the fastest rates as well as excellent stereocontrol
in lactide polymerization.12 Previously, [L1Zn2(µ-OEt)Cl2]
was shown to be a rapid and controlled initiator,9a and the
Mg(II) and Co(II) analogues were less effective.9b With the
improved synthesis of [LnZn2(µ-OEt)X2] complexes in hand,
it was of interest to test the new complexes as initiators for
lactide polymerization (Table 6).
Figure 6. The molecular structure of [L3Zn2(µ-OAc)2][BPh4] (50%
probability ellipsoids).
the coordinative flexibility of this system. The most obvious
difference between the three complexes is the “gape” angle
between the two acetate ligands, with the vectors from C(26)
and C(31) subtending angles of ca. 93, 127, and 124° at the
Zn(1) · · · Zn(2) centroid in complexes [L1Zn2(µ-
OAc)2][BPh4], [L2Zn2(µ-OAc)2][BPh4], and [L3Zn2(µ-
OAc)2][BPh4], respectively. Interestingly, this pattern extends
to the metal · · · metal separation [3.1822(2), 3.2620(4), and
3.2580(5) Å in the complexes of L1, L2, and L3, respectively].
The planar, rigid structure of coordinated L2 results in the
greatest intermetallic separation. The closer zinc-zinc
distance in the case of L1 versus L3 is tentatively assigned
to steric effects resulting from the methyl substituents.
Unlike the situation in the structures [L1Zn2(µ-OEt)I2] and
[L2Zn2(µ-I)I2], the [L1-3Zn2(µ-OAc)2][BPh4] complexes have
symmetric ZnsO(1) bonds (Tables 3-5), although they
differ between the three complexes with those in [L2Zn2(µ-
OAc)2][BPh4] (2.1035(13) and 2.1028(13) Å) longer than
those in [L3Zn2(µ-OAc)2][BPh4] (2.035(2) and 2.031(2) Å),
The polymerizations were all conducted in methylene
chloride at room temperature and at a loading initiator/lactide
ratio of 1:200. The initiators were all generated in situ by
reaction between the [LnZn2(µ-OEt)X2] and KOEt, followed
Inorganic Chemistry, Vol. 47, No. 24, 2008 11717

Knight et al.
Table 6. Polymerization of rac-Lactidea
have been developed, especially noteworthy are zinc phe-
noxide,19 zinc ꢀ-diiminate,14,20 and chromium(III) or cobal-
t(II) salen complexes.21 The zinc ꢀ-diiminate complexes
show very high turn-over frequencies (TOF), as well as
excellent control.13,18 Recent mechanistic studies, suggest
that the most effective ꢀ-diiminate complexes are loosely
associated dimers under the polymerization conditions.
Coates and others have proposed that the two metal centers
may be cooperating with one another, leading to enhanced
rates. According to such a mechanism, one metal center
activates the epoxide, which is subsequently attacked by a
nucleophilic group bound to the other metal center, and this
proposal has led to the deliberate preparation of various
bimetallic zinc catalysts.11c,f,h,22 The bimetallic complexes
of this study were tested for copolymerisation activity, using
cyclohexene oxide under 1 atm pressure of CO2, at 50 °C
for a period of 24 h. [L1-3Zn2(µ-X)X2] were activated with
potassium ethoxide, PPN-Cl, and also DMAP, as previous
studies using Cr(III) or Co(II) complexes have established
that halide complexes and cocatalysts yield very active
catalysts.23 [L1-3Zn2(µ-OAc)3] and [L1-3Zn2(µ-OAc)2][BPh4]
were tested without addition of cocatalyst. In all cases, the
new complexes failed to enable isolation of any significant
quantity of polymer. Several of the complexes were also
tested at higher pressures of CO2 (7 bar) and higher
temperature (80 °C), but again no activity was observed. The
lack of catalytic activity may be due to the metal centers
being sterically congested. The putative propagating species
is a zinc carbonate, yet we have observed that this ligand
system favors formation of cationic acetate complexes
[LnZn2(µ-OAc)2]+.
c
complex
% conversionb
time/min
Mn
PDIc
[L1Zn2(µ-OEt)Cl2]
[L1Zn2(µ-OEt)Br2]
[L1Zn2(µ-OEt)I2]
[L2Zn2(µ-OEt)Br2]
[L3Zn2(µ-OEt)Br2]
90
91
98
91
92
183
22
1453
4299
1
18 500
18 500
18 200
14 200
23 800
1.15
1.15
1.15
1.06
1.06
a Polymerization conditions: [LA]0) 1 M, [LnZn2(µ-OEt)X2] ) 5 mM,
b
1
CH2Cl2, 298 K. Determined from the H NMR spectrum by integration
of the methylene resonances from polylactide (5.20 ppm) and lactide (5.00
ppm). c Determined from the SEC in THF using MALS detection to obtain
the absolute molecular weights.
by removal of the KX byproduct and excess KOEt by
filtration. Although an excess of KOEt was applied, the
possibility of its interference on the polymerization can be
discounted as KOEt has previously been shown to be a very
poor initiator at room temperature.12b The polymerizations
were compared by the time taken to reach a conversion of
greater than 90%. A full kinetic study has already been
undertaken with [L1Zn2(µ-OEt)Cl2] and established that it
behaves in the usual manner, giving a second-order rate law
with first-order dependencies on lactide and initiator
concentration.9a Within the series of complexes using L1,
the relative order of rates is [L1Zn2(µ-OEt)Br2] > [L1Zn2(µ-
OEt)Cl2] > [L1Zn2(µ-OEt)I2]. The order does not correlate
directly with the electronegativity of the halide element, with
the bromide complexes showing the best rates. Thus, the
bromide complexes of L2 and L3 were also tested: [L2Zn2(µ-
OEt)Br2] was extremely slow, but [L3Zn2(µ-OEt)Br2] was
very rapid, giving complete conversion in 1 min. The nature
of the ancillary ligand has a significant influence on the rate:
L3 . L1 . L2; the slowest polymerization results from using
the rigid, electron-rich imine ligand, whereas the fastest
polymerization results from the flexible amine ligand. The
polymerization control exerted by all initiators was reason-
able; the polylactide produced had a narrow polydispersity
index. The control over molecular weight was rather variable;
the molecular weight expected for a 5 mM concentration of
initiator and at >90% conversion is 26 000. The fit between
experimental and theoretical molecular weight appears to
correlate with the activity of the initiator, with the fastest
initiator showing the best fit and the slowest initiator showing
significantly lower molecular weights than the theoretical
value. It is also notable that all the polymers show narrow
polydispersity indices. A likely explanation for this behavior
is that chain transfer is occurring; however, the source of
the protic chain transfer agent (usually water or an alcohol
group) is not currently clear. Chisholm has established that
chain transfer occurs significantly more rapidly than propa-
gation in lactide polymerization initiated by tin complexes.17
It is clear that [L3Zn2(µ-OEt)Br2] holds the greatest potential
for future development in terms of both its rate and
polymerization control.
Conclusions
We report the synthesis and characterization of a range
of new bimetallic zinc complexes of the formula [LnZn2(µ-
X)X2] (where X ) Cl, Br, I, OEt, OAc), as well as the
cationic acetate complexes, [LnZn2(µ-OAc)2][BPh4]. [L1Zn2(µ-
X)X2] (where X ) Cl, Br) showed NMR spectra consistent
with a single symmetrical solution structure, but increasing
the steric bulk of the X group to iodide and acetate groups
led to multiple configurations in solution. The X-ray crystal
(19) (a) Darensbourg, D. J.; Holtcamp, M. W.; Struck, G. E.; Zimmer,
M. S.; Niezgoda, S. A.; Rainey, P.; Robertson, J. B.; Draper, J. D.;
Reibenspies, J. H. J. Am. Chem. Soc. 1999, 121, 107. (b) Darensbourg,
D. J.; Wildeson, J. R.; Yarbrough, J. C.; Reibenspies, J. H. J. Am.
Chem. Soc. 2000, 122, 12487. (c) Darensbourg, D. J.; Zimmer, M. S.;
Rainey, P.; Larkins, D. L. Inorg. Chem. 2000, 39, 1578.
(20) (a) Cheng, M.; Lobkovsky, E. B.; Coates, G. W. J. Am. Chem. Soc.
1998, 120, 11018. (b) Allen, S. D.; Moore, D. R.; Lobkovsky, E. B.;
Coates, G. W. J. Am. Chem. Soc. 2002, 124, 14284. (c) Moore, D. R.;
Cheng, M.; Lobkovsky, E. B.; Coates, G. W. J. Am. Chem. Soc. 2003,
125, 11911. (d) Eberhardt, R.; Allmendinger, M.; Luinstra, G. A.;
Rieger, B. Organometallics 2003, 22, 211. (e) van Meerendonk, W. J.;
Duchateau, R.; Koning, C. E.; Gruter, G. J. M. Macromol. Rapid
Commun. 2004, 25, 382.
(21) (a) Paddock, R. L.; Nguyen, S. T. J. Am. Chem. Soc. 2001, 123, 11498.
(b) Qin, Z. Q.; Thomas, C. M.; Lee, S.; Coates, G. W. Angew. Chem.,
Int. Ed. 2003, 42, 5484. (c) Darensbourg, D. J.; Yarbrough, J. C. J. Am.
Chem. Soc. 2002, 124, 6335.
(22) (a) Xiao, Y. L.; Wang, Z.; Ding, K. L. Chem. Eur. J. 2005, 11, 3668.
(b) Lee, B. Y.; Kwon, H. Y.; Lee, S. Y.; Na, S. J.; Han, S. I.; Yun,
H. S.; LeeH.; Park, Y. W J. Am. Chem. Soc. 2005, 127, 3031.
(23) Darensbourg, D. J. Chem. ReV. 2007, 107, 2388.
CO2-Cyclohexene Oxide Copolymerisation. The co-
polymerisation of epoxides and CO2 was first reported in
1969.18 Recently, several more active and controlled catalysts
(17) Chisholm, M. H.; Delbridge, E. E.; Gallucci, J. C. New J. Chem. 2004,
28, 145.
(18) Inoue, S.; Koinuma, H.; Tsuruta, R. J. Poly. Sci., Part B:Polym. Lett.
1969, 7, 287.
11718 Inorganic Chemistry, Vol. 47, No. 24, 2008

Dinuclear Zinc Complexes Using Pentadentate Phenolate
In Situ Generation of [L1Zn2(µ-OEt)Br2]. A Youngs tap NMR
tube was charged with [L1Zn2Br3] (17.5 mg, 0.025 mmol) and
potassium ethoxide (2.5 mg, 0.030 mmol). Deuterated acetonitrile
was then added (ca. 0.5 mL), and the mixture was shaken. The
structures show several potential sources of asymmetry,
including the orientation of the ArCH2 groups, the config-
uration of the N-R (R ) Me or H) groups, as well as the
ring configurations formed by the coordinated pendant arms.
Further asymmetry may also arise from the configurations
of the zinc centers themselves and, in particular, the positions
of the X substituents. It is clear that complexes with L1 and
L3 have flexible geometries,24 whereas the imine unit in L2
generates essentially planar complex geometries. Substituting
the bridging halide group for an ethoxide group was achieved
in quantitative yield using a salt metathesis reaction with
potassium ethoxide. These alkoxide complexes were viable
initiators for lactide ring opening polymerization, with the
bromide complex using L3 being extremely fast and well-
controlled. The new complexes are not suitable catalysts for
the copolymerization of carbon dioxide and epoxides, and
this may be a result of steric crowding around the metal
centers.
1
resulting mixture was analyzed by H NMR spectra after 30 min
and showed formation of the title compound in quantitative yield.
1H NMR (400 MHz, CD3CN) δ ppm: 6.89-6.72 (m, 2H, ArH),
4.30-2.99 (br m, 6 H, ArCH2 and OCH2CH3), 2.90-2.00 (br m,
26 H, NCH2, NMe, NMe2, ArMe), 1.49-1.02 (br m, 3H,
OCH2CH3).
[L1Zn2(OAc)3]. This complex has been synthesized in air and
under nitrogen in a similar manner to the previous complexes,
although the final product is soluble in most organic solvents, which
can make purification difficult. Alternatively, a round-bottom flask
was charged with HL1 (228 mg, 0.68 mol), dichloromethane (2
mL), and hexane (10 mL). Sodium hydride [60% in mineral oil]
(24 mg, 0.61 mmol) was then added, and the mixture was stirred
for 2 h. After addition of zinc acetate (268 mg, 1.22 mmol), the
mixture was stirred overnight, during which time a white precipitate
formed that was filtered and dried (672 mg, 86%). Synthesis was
also performed under dry nitrogen conditions. 1H NMR (400 MHz,
d2-TCE, 303 K) δ ppm: 6.82 (s, 2H, ArH), 4.33 (d, 2H, ArCH2,
Experimental Section
2
2JHH ) 11 Hz), 3.13 (d, 2H, ArCH2, JHH ) 11 Hz), 2.93 (m, 4H,
Full experimental details for ligand and complex syntheses can
be found in the Supporting Information; however, some examples
of our standard complex syntheses are given here.
NCH2), 2.62 (s, 12H, NMe2), 2.50 (m, 4H, NCH2), 2.23 (s, 3H,
ArMe), 2.16 (br s, 6H, NMe), 2.10 (s, 9H, OAc). 13C{1H} NMR
(100 MHz, d2-TCE, 303 K) δ ppm: 176.9 (OAc), 157.7 (Ar), 132.2
(Ar),127.3 (Ar), 123.4 (Ar), 61.7 (ArCH2), 55.9 (NCH2), 52.2 (br,
NCH2), 46.2 (NMe2), 43.6 (NMe), 22.1 (OAc), 20.2 (ArMe). MS
(FAB) m/z 583 [M-OAc]+ (100%), 357 (14%). Elem. anal. found
(calculated) for C25H44N4O7Zn2 (%) C, 46.72 (46.67); H, 6.96
(6.89); N, 8.63 (8.71).
[L1Zn2(µ-Cl)Cl2]. In Air. A 25 mL round-bottomed flask with
stir bar was charged with HL1 (101 mg, 0.30 mmol), methanol (10
mL), and sodium methoxide (17 mg, 0.31 mmol). After the pale
yellow/green solution was stirred for 15 min, zinc chloride (82 mg,
0.60 mmol) was added, and the solution rapidly decolorized. A
white solid precipitated, and after stirring for a further hour, the
mixture was filtered. The white solid was dissolved in dichlo-
romethane and filtered through Celite. Removal of volatiles and
drying under vacuum yielded a white solid (136 mg, 79%).
Under Dry Nitrogen. A Schlenk tube with stir bar was charged
with HL1 (200 mg, 0.59 mmol) and was placed under nitrogen.
Dry THF (15 mL) was added, and the solution was cooled to -78
°C before transferring the solution to a Schlenk tube containing
potassium hydride (35 mg, 0.89 mmol). A gas was evolved and
the pale green solution turned yellow. The mixture was allowed to
warm to room temperature and stirred overnight before filtering
the solution into a Schlenk tube containing zinc chloride (162 mg,
1.19 mmol). The solution rapidly decolorized, and a large amount
of white precipitate formed. Solvent was removed in vacuo, and
the white solid was dissolved in dichloromethane, and then filtered.
After removing the solvent and drying in vacuo, a white solid was
obtained (275 mg, 81%). 1H NMR (400 MHz, CDCl3) δ ppm: 6.75
[L1Zn2(µ-OAc)2][BPh4]. To a solution of HL1 (55 mg, 0.16 mmol)
in methanol (5 mL) was added Zn(OAc)2 ·2H2O (72 mg, 0.33 mmol).
Upon stirring for several minutes, the zinc acetate fully dissolved, after
which time sodium tetraphenylborate (59 mg, 0.17 mmol) was added.
A white precipitate formed rapidly, and the mixture was stirred for 10
min. The white solid was filtered and dried (137 mg, 93%). Crystals
suitable for X-ray crystallography were grown from ethanol. 1H NMR
(400 MHz, CDCl3) δ ppm: 7.38 (br m, 8H, B-ArH), 7.01 (t, 8H,
B-ArH, 3JHH ) 7 Hz), 6.87 (t, 4H, B-ArH, 3JHH ) 7 Hz), 6.83 (s,
2H, ArH), 3.92 (d, 2H, ArCH2, 2JHH ) 13 Hz), 3.11 (d, 2H, ArCH2,
2JHH ) 13 Hz), 2.69 (m, 2H, NCH2), 2.56 (s, 6H, NMe), 2.55 (m, 2H,
NCH2), 2.31 (s, 6H. NMe2), 2.24 (s, 3H, ArMe), 2.09 (m, 2H, NCH2),
2.02 (s, 6H, OAc), 1.90 (s, 6H, NMe2), 1.87 (m, 2H, NCH2). 13C{1H}
NMR (100 MHz, CDCl3) δ ppm: 179.7 (OAc), 164.3 (q, B-Ar, 1JBC
) 49 Hz), 157.1 (Ar), 136.2 (B-Ar), 132.9 (Ar), 128.3 (Ar), 125.4
(B-Ar), 122.9 (Ar), 121.6 (B-Ar), 61.6 (ArCH2), 55.9 (NCH2), 49.6
(NCH2), 48.6 (NMe), 44.0 (NMe2), 43.6 (NMe2), 23.5 (OAc), 20.3
(ArMe). MS (FAB) m/z 583 [M-BPh4]+ (77%), 357 (16%). Elem.
anal. found (calculated) for C47H61BN4O5Zn2 (%) C, 62.62 (62.47);
H, 6.76 (6.80); N, 6.11 (6.20).
2
(s, 2H, ArH), 4.36 (d, 2H, ArCH2, JHH ) 12 Hz), 3.11 (d, 2H,
2
ArCH2, JHH ) 12 Hz), 2.86-3.00 (m, 4H, NCH2), 2.61 (s, 12H,
NMe2), 2.54 (m, 2H, NCH2), 2.45 (m, 2H, NCH2), 2.16 (s, 9H,
NMe and ArMe). 13C{1H} NMR (100 MHz, CDCl3) δ ppm: 158.3
(Ar), 131.8 (ArH), 125.8 (Ar), 123.3 (Ar), 62.9 (ArCH2), 56.2
(NCH2), 55.8 (NCH2), 47.2 (br, NMe2), 43.8 (NMe), 20.1 (ArMe).
MS (FAB) m/z 572 [M]+ (3%), 537 [M-Cl]+ (56%), 464 (6%),
435 [M-ZnCl2]+ (9%), 376 (15%), 333 (72%). Elem. anal. found
(calculated) for C19H35Cl3N4OZn2 (%) C, 39.79 (39.85); H, 6.08
(6.16); N, 9.73 (9.78).
Acknowledgment. The EPSRC are acknowledged for
funding the research (EP/C544846/1 and EPC544838/1).
PURAC are thanked for the donation of rac-lactide.
Supporting Information Available: The full experimental
section for all ligands and complexes prepared as well as details
of the X-ray crystallography are available in the Supporting
Information. This material is available free of charge via the Internet
at http://pubs.acs.org.
(24) In a related system, R-methybenzyl groups were used as substituents
on the inner nitrogen atom of the pendant arms in an attempt to control
stereoselective substrate recognition for aminopeptidase models. Two
preferable conformers were found by calculation from six possible
conformers, assuming C2 symmetry with a homochiral ligand. See
ref 4I.
IC8014173
Inorganic Chemistry, Vol. 47, No. 24, 2008 11719