Solvent dependence of the structure of ethylzinc acetate and its application in CO2/epoxide copolymerization

Article abstract of DOI:10.1021/om200004a

NMR and single-crystal X-ray diffraction experiments indicate that the ligand stoichiometry of the complexes formed from the reaction between zinc bis(acetate) and diethylzinc depends on the nature of the solvent (coordinating vs noncoordinating) and that the strength of the donor interaction of a coordinating solvent (THF vs pyridine) affects the nuclearity of the complex's repeat unit in the solid state. The complexes are active catalysts for the copolymerization of cyclohexene oxide and CO₂, under mild conditions.

Full text of DOI:10.1021/om200004a

ARTICLE
pubs.acs.org/Organometallics
Solvent Dependence of the Structure of Ethylzinc Acetate and Its
Application in CO2/Epoxide Copolymerization
Katherine L. Orchard, Jonathan E. Harris, Andrew J. P. White, Milo S. P. Shaffer,* and Charlotte K. Williams*
Department of Chemistry, Imperial College London, Exhibition Road, London SW7 2AZ, U.K.
S Supporting Information
b
ABSTRACT: NMR and single-crystal X-ray diffraction experiments indicate that the
ligand stoichiometry of the complexes formed from the reaction between zinc
bis(acetate) and diethylzinc depends on the nature of the solvent (coordinating vs
noncoordinating) and that the strength of the donor interaction of a coordinating
solvent (THF vs pyridine) affects the nuclearity of the complex’s repeat unit in the
solid state. The complexes are active catalysts for the copolymerization of cyclohexene
oxide and CO2, under mild conditions.
rganozinc compounds were one of the earliest classes of
organometallic compounds; indeed Frankland reported the
[RZn(OOCR0X)] (X = OH, NH2, SH);27,36ꢀ38 a variety of
structures have been observed, including tetrameric rings of
[RZn(μ2-OOCR0(μ2-X))] units (R = Et; R0X = CPh2(NH2)),
more complex structures such as [Zn6R3(μ3-OOCR0(μ2-X))3L3]
(R = Et; R0X = CPh2(OH); L = solvent),36 and extended polymers
O
preparation of diethylzinc as early as 1848.1,2 In the centuries that
have followed, heteroleptic alkyl zinc reagents (RZnX, where X
= halide, alkoxide, aryloxide, carboxylate, amide, etc.) have
become established as useful reagents in organic synthesis,3,4
perhaps most well-known in palladium/nickel-catalyzed Negishi
cross-coupling reactions or in addition reactions with carbonyl
compounds;5ꢀ10 they are also used as catalysts, for example in
such as [RZn(μ2-OOCR0(μ2-X))(py)]n (R
=
Et; R0X
= (CH2)2SH).27
Previously, we reported39 that the reaction between ZnEt2 and
Zn(OAc)2 in toluene gave an unusual pentameric complex with a
ligand stoichiometry of OAc:Et of 3:2 ([Zn5(μ3-OAc)6(Et)4], 1;
Figure 1). This structure is favored in the solid state and in
solution (benzene, toluene), regardless of the reagent excess and
synthetic route (ligand exchange between ZnEt2 and Zn(OAc)2
or reaction between ZnEt2 and acetic acid). As mentioned above,
the same core pentanuclear structure was subsequently also
reported by Redshaw et al. for various aryl zinc aryl carbo-
xylates.40
epoxide homopolymerization or the copolymerization of epox-
11ꢀ16
ides and CO2
and as precursors for the preparation of
ZnO.17ꢀ29 The most widely studied are alkylzinc alkoxides,
which form dimers ([RZn(μ-OR0)]2), tetramers ([RZn(μ3-
OR0)]4 “cubes”), and heptameric species ([Zn7(μ3-OR0)8(R)6])
depending on the nature of the ligand, the dialkyl zinc:alcohol
ratio, and the solution concentration.30ꢀ32
In comparison, the structures and solution behavior of alkyl-
zinc carboxylates (of the general form [RZn(OOCR0)]) have
received much less attention. In 1965, Coates reported the
synthesis of methylzinc acetate, a material that was insoluble in
benzene and proposed to have a polymeric structure.33 Methyl-
zinc acetate could be dissolved in excess pyridine, and a dimeric
[MeZn(μ-OOCCH3)(py)]2 species was proposed, although the
molecular weight, determined by cryoscopy, decreased with
increasing dilution. Since then, although the reactivity of ethyl-
zinc carboxylates has been reported,34 very few structural studies
have been conducted. Crystal structures are known for a small
number of aryl carboxylate derivatives:35 ethylzinc 2,6-bis-(2,4,6-
trimethylphenyl)benzoate was found to form [EtZn(μ2-
OOCR)]2 dimers, whereas ethylzinc benzoate was found to be
a cyclic hexamer of the form [EtZn(μ3-OOC(C6H5))]6. Re-
cently, a pentanuclear structure has also been found by Redshaw
One area in which zinc carboxylate reagents have shown
significant promise is as catalysts in the alternating copolymer-
ization of epoxides and carbon dioxide (Scheme 1). The copo-
lymerization of propylene oxide (PO) and CO2 was first reported
by Inoue et al. in 1969, using ZnEt2/water as the catalyst
mixture;13 since this early discovery, many advances in activity,
selectivity, and polymer quality have been made, and several
detailed and comprehensive reviews have been published.15,41ꢀ44
Zinc carboxylate motifs are common among catalysts for this
copolymerization: many early zinc catalysts were based on
mixtures of ZnEt2 and di- and trihydric ligands, including
dicarboxylic and hydroxycarboxylic acids.14 Indeed, the most
widely applied heterogeneous catalyst for the copolymerization
of PO and CO2 is zinc glutarate [Zn(O2C(CH2)3CO2)]n.15,45
Despite its broad utility, understanding and improving surface
active site structure(s) of zinc glutarate have remained extremely
challenging. Indeed, syntheses of this material are sensitive to the
et al. for complexes of the form [Zn5(C6F5)4(Ar0)6] x(toluene),
3
where Ar0 = 2-chlorobenzoate or 2,4,6-trimethylbenzoate and
x = 1.5 and 2, respectively.40 Of the aliphatic carboxylate deriva-
tives, several complexes have been prepared for which the
carboxylate group contains a second coordinating functionality,
Received: January 4, 2011
Published: March 23, 2011
r
2011 American Chemical Society
2223
dx.doi.org/10.1021/om200004a Organometallics 2011, 30, 2223–2229
|

Organometallics
ARTICLE
Figure 2. Selected regions of the 1H NMR spectrum (C6D6) of
[EtZn(OAc)(py)], 2 (see Figure S3, Supporting Information, for the
complete spectrum).
Figure 1. Pentanuclear structure of the discrete complex [Zn5(OAc)6-
(Et)4], 1, formed by reaction between ZnEt2 and Zn(OAc)2, in toluene
(white bonds correspond to ethyl groups; black bonds correspond to
acetate groups).
of pyridine per Zn to an equimolar mixture of ZnEt2 and
Zn(OAc)2 enabled isolation of compound 2 [EtZn(OAc)(py)]
in 72% yield. The 1H NMR spectrum of 2 showed ligand ratios of
OAc:Et:py of 1:1:1 (Figure 2). As shown in Table 1, the 1H NMR
chemical shifts of the ethyl and acetate resonances are shifted
Scheme 1. Copolymerization of CO2 and an Epoxidea
1
much further downfield compared to 1. Both the H and 13C-
{1H} NMR spectra show that the py resonances are shifted
relative to free py in d6-benzene, demonstrating that the py is
strongly bound to the complex in solution.
a X = initiating group provided by the catalyst, such as halide or
carboxylate; EG = polymer end-group, usually H; R, R' = H, alkyl, aryl
group.
Crystals of 2 were obtained directly from the saturated
reaction mixture (toluene), and X-ray diffraction experiments
confirmed the 1:1 ligand ratio, showing 2 to be constructed of
EtZn(OAc)(py) monomeric units (Figure 3) that link together
to form an extended chain (Figure S1, Supporting Information).
In contrast to 1, the acetate groups of 2 are bound in a μ2-
coordination mode, with pyridine occupying the fourth coordi-
nation site of each tetrahedrally coordinated zinc. The ZnꢀO
bond lengths of 2 (2.014(3) and 2.046(2) Å) are comparable to
the ZnꢀO bond lengths of 1 (1.997(2)ꢀ2.1695(17) Å), and the
ZnꢀN bond length (2.111(3) Å) is comparable to ZnꢀN bond
lengths in related carbamato species (2.0786(17) Å)36 and to Zn-
bound py species (2.0339(8) Å).46
nature of the zinc precursor, stirring method, temperature, and
additives. Recently, Eberhardt et al. prepared zinc glutarate
derivatives from diethyl zinc, glutaric acid, and SO2, which
showed excellent productivities.11 In contrast to catalysts pre-
pared using dicarboxylic acids (such as glutaric acid), those
prepared from monocarboxylic acids have been much less
explored. An early study by Inoue et al.14 showed only low
activity of ZnEt2/mono-carboxylic acid mixtures (acetic acid,
benzoic acid) toward the copolymerization of PO and CO2 in
dioxane at 35 °C and 40 atm of CO2; however, to our knowledge,
detailed studies using more forcing conditions or with other
epoxides have not been carried out for these catalyst systems.
Herein, we present our findings that both the ligand stoichi-
ometry and nuclearity of the complex formed from the reaction
between ZnEt2 and Zn(OAc)2 depend on the nature of the
solvent. In addition, we have carried out preliminary testing of
the activity of the well-defined ethylzinc acetate complexes
toward copolymerization of CO2 and cyclohexene oxide (CHO).
Pulsed gradient spinꢀecho (PGSE) NMR spectroscopy
allows determination of molecular diffusion coefficients, from
which hydrodynamic radii can be estimated.47 The diffusion
coefficient of 2 in C6D6 was estimated to be 1.0 ꢁ 10ꢀ9 m2 sꢀ1
,
which corresponds to a hydrodynamic radius of 3.5 Å; this value
matches that estimated from the crystal data for the monomeric
[EtZn(OAc)(py)] unit, based on four monomers per unit cell
and a unit cell occupancy of 68%. On the basis of the measured
and calculated values, it is likely that 2 exists as a monomer in
solution.
’ RESULTS AND DISCUSSION
The coordination of the py groups in the solid state is also
supported by IR data; the absorption bands associated with the
vibrations of the pyridine ring are shifted to higher frequency
(free pyridine vibrations at 1581, 1030, and 604 cmꢀ1; adduct
vibrations at 1605, 1041, and 614 cmꢀ1, respectively).48 The
difference, Δ, between the asymmetric and symmetric carbox-
ylate stretches is 113 cmꢀ1, within the range for bridging acetate
groups,49 but slightly lower than that of 1 (165 cmꢀ1), which may
be an indication of the difference in the acetate bridging mode
(μ2 for 2 compared to μ3 for 1).
It is well known that coordinating ligands can lower the
nuclearity of heteroleptic zinc species; for example, for alkylzinc
carbamato ([RZn(OOCNR02)]) and alkylzinc alkoxide
([RZn(OR0)]) species, the addition of pyridine breaks up the
discrete tetrameric complexes ([RZn(μ3-XR0)]4) to form dimers
([RZn(μ2-XR0)(py)]2).32,37 Previously, we showed that the
reaction between equimolar quantities of ZnEt2 and Zn(OAc)2
in toluene or benzene yielded the pentanuclear complex, 1, and
excess ZnEt2.39 On the other hand, the addition of one equivalent
2224
dx.doi.org/10.1021/om200004a |Organometallics 2011, 30, 2223–2229

Organometallics
Table 1. Comparison of the Chemical Shifts (in C6D6) of [EtZn(OAc)(py)], 2, with [Zn5(OAc)6(Et)4], 1, and Free Pyridine (py)
ARTICLE
1H NMR chemical shift (ppm)
13C{1H} NMR chemical shift (ppm)
group
1
3
free py50
1
3
free py50
ꢀOOCH3
ꢀCH2CH3
py ortho
1.89
1.56
2.07
1.71
8.60
6.81
6.55
181.2
12.9
179.7
14.0
8.53
6.98
6.66
149.5
138.5
124.9
150.3
135.3
123.6
py para
py meta
Figure 3. Structure of the asymmetric unit of [EtZn(OAc)(py)], 2.
Figure 5. Molecular structure of the asymmetric unit of 3, demonstrat-
ing the tetrahedrally coordinated zinc atoms and the acetate bridging
arrangement that leads to the extended sheet structure (Figure S2,
Supporting Information).
diffraction showed that the structure consists of dimeric
[EtZn(OOCCH3)]2 units (Figure 5). The asymmetric units
are connected to one another via bridging of neighboring μ3-
acetate groups to form two-dimensional sheets of 16-membered
[Zn(OAc)]4 rings joined by Zn2O2 nodes (herringbone struc-
ture; Figure S2, Supporting Information). Interestingly, no
molecules of THF remained bound in the solid state, despite
the great excess present during crystallization. The ZnꢀO bond
lengths in the structure are comparable to those for 1 and 2, lying
within the range 1.977(7)ꢀ2.137(7) Å. The OꢀZnꢀO bond
angles of the dimeric repeat unit are 75.1(3)° and 77.0(3)°,
which are comparable to the OꢀZnꢀO angles of the four-
membered rings of 1; the remaining OꢀZnꢀO angles range
between 92.5(3)° and 107.9(3)°.
The 1:1 ligand ratio and absence of bound THF are supported
by elemental analysis (calculated C 31.30, H 5.25, found C 31.27,
H 5.14) and 1H NMR spectroscopy (Figure 4): when redissolved
in d8-THF, the only THF signals observed were due to residual
solvent; when dissolved in d6-benzene (vide infra), no THF
signals were observed.
Figure 4. 1H NMR spectrum (THF-d8) of 3. Peak marked with “*” is
residual solvent; peak marked with “Δ” is due to small amounts of
(dissolved) ethane.
The analogous reaction, adding one equivalent of tetrahydro-
furan (THF) to a toluene solution of ZnEt2 and Zn(OAc)2,
yielded an opaque suspension that could not be fully character-
ized; however, the reaction between equimolar quantities of
Zn(OAc)2 and ZnEt2 directly in THF yielded a product, 3, with
the molecular formula EtZnOAc (1H NMR spectroscopy,
Figure 4). The use of an excess of Zn(OAc)2 (3:2 OAc:Et, as
found for 1) yielded the same product, 3, contaminated with
excess Zn(OAc)2 (NMR, IR; not shown). Crystals of 3 were
obtained by slow evaporation from a THF solution, and X-ray
The PGSE NMR spectrum of 3 in d8-THF gave an estimate of
the diffusion coefficient of 1.7 ꢁ 10ꢀ9 m2 sꢀ1, which corresponds
to a hydrodynamic radius of 2.7 Å. The radius of the dimeric unit
estimated from the crystal data was 3.7 Å, based on four dimers
2225
dx.doi.org/10.1021/om200004a |Organometallics 2011, 30, 2223–2229

Organometallics
ARTICLE
per unit cell and a unit cell occupancy of 75.2%. The values are
comparable, indicating that the sheet structure is likely to be
disrupted and that 2 exists as a discrete dimer when solvated by
excess THF. It seems likely that the acetate groups would favor a
μ2-bridging coordination in an isolated (solvated) dimer, as
solvent molecules could occupy the fourth coordination site of
the zinc atoms; however, no evidence for such a change in
coordination has been obtained.
more protected toward the center of the structure. It is not clear,
however, whether the observed structure “appears” less polar
than a hypothetical 1:1 cubane structure similar to that seen for
the ethylzinc alkoxides. It is possible that the pentanuclear
arrangement is structurally more compact than other, higher
structures or any attainable structures with 1:1 ligand ratios.
The tendency for the complexes to revert to the ligand ratio
and structure dictated by the preference of the solvent
(regardless of the structure that the solid possesses prior to
dissolution) indicates the lability of the complexes and, impor-
tantly, the great influence the solvent parameters exert on the
structures of organometallic complexes.
To investigate the solvent dependence and lability of the
structures of the complexes, isolated 1 and 3 were dissolved in d8-
THF and d6-benzene, respectively, and their structures studied
1
by H NMR spectroscopy. Isolated 3 was largely insoluble in
benzene-d6 at room temperature but, when heated, yielded a
clear solution, which remained on cooling to 25 °C. The H
CO2/CHO Copolymerizations. The results of CO2/CHO
copolymerization studies using 1 and 3 are summarized in
Table 2. Preliminary experiments demonstrated no activity for
either complex toward PO copolymerization at 15 atm CO2;
however, both were active toward cyclohexene oxide (CHO) at
this pressure (in neat epoxide). It is worth noting that copoly-
merizations using zinc glutarate and other zinc carboxylate-based
catalysts are frequently carried out at much higher pressures of
CO2 (40 atm and above).
The donating abilities of epoxides toward zinc complexes have
been reported to be lower than that of THF;52 however, it is
plausible that the same preference for the 1:1 ligand ratio occurs
for CHO as for THF. It may be expected, therefore, that 1 would
segregate into the 1:1 complex (3) and excess Zn(OAc)2; as
such, 1, 3, and Zn(OAc)2 alone were tested in the copolymer-
ization reactions to compare activity. Complex 2 showed only
trace quantities of polymer formation; it is proposed that the
strongly bound pyridine coligand could hinder epoxide coordi-
nation and thus retard copolymerization.
In the absence of CO2 (1 atm N2), both 1 and 3 were active
catalysts for the homopolymerization of neat CHO, yielding a
highly viscous liquid (polycyclohexene oxide) after 10 h (Table 2,
rows 1 and 2). The yields of the isolated poly(cyclohexene oxide)
were similar for both catalysts, and although the conversion was
low (7% and 5% for 1 and 3, respectively), the Mn values were
high. The low conversion and high molecular weight are com-
parable to the results found for epoxide homopolymerization
catalyzed by alkylzinc alkoxides,43 indicating that only a small
proportion of the metal sites were active for catalysis. The low
PDIs of the polymers (Mw/Mn = 1.6 and 1.5 for 1 and 3,
respectively) demonstrate that, for the active zinc sites, the rate of
propagation is somewhat higher than the rate of initiation.
At 15 atm CO2 and in neat CHO, both 1 and 3 were
moderately active catalysts for the copolymerization. The TOFs
decreased with increasing reaction time (3 to 7.5 h), which
corresponded to a decrease in reaction rate due to an increase in
the viscosity of the reaction mixture. The conversion was
marginally higher for 3 than 1 at the 3 h time point (Table 2;
rows 4 and 6), and 1 had a much higher proportion of ether
linkages after 3 h than 3. By 7.5 h, the discrepancy had
disappeared and both catalysts had approximately the same
conversion and the same proportion of ether linkages. The
molecular weight of the copolymer catalyzed by 1 was consider-
ably higher after 3 h than both the equivalent copolymer
catalyzed by 3 and the polymer catalyzed by 1 after 7.5 h (Mn
44 800, 12 900, and 12 900 g molꢀ1, respectively). The reduction
in Mn with time for 1 is thought to be caused by initial
homopolymerization to generate high molecular weight poly-
(cyclohexene oxide) chains (which represent a higher proportion
of the total chains) followed by copolymerization and chain
1
NMR spectrum of the cooled solution showed resonances of the
3:2 pentanuclear complex (1) and those of free ZnEt2; the sum of
the integration of the ethyl peaks gave a ratio of [OAc]:[total Et]
of 1:1, as expected. Similarly, attempted dissolution of isolated 1
in THF-d8 at room temperature yielded a suspension with a large
amount of undissolved solid, but, on heating the mixture to reflux
(70 °C), the solid dissolved to give a clear solution, which
1
remained on cooling to 25 °C. The H NMR spectrum of the
cooled solution showed total ligand ratios of 3:2 [OAc]:[Et], but
the acetate resonance was shifted downfield with respect to that
of 3, toward that of free Zn(OAc)2. PGSE of the clear solution at
room temperature gave an estimate of the hydrodynamic radius
of the species in solution of 2.5 Å, comparable to 3; the expected
radius of 1 was 5.1 Å, based on the crystal structure data.39 From
the 1H NMR spectrum, it is believed that the solution contained
the dimeric, 1:1 species (3) and excess Zn(OAc)2 in dynamic
exchange, rather than the pentanuclear complex. Addition of one
equivalent of ZnEt2 per mole of [Zn5(OAc)6(Et)4] to the
solution gave a 1H NMR spectrum equivalent to that of 3.
It can be concluded that the pentanuclear complex, 1, is
strongly favored in aromatic solvents and that ethylzinc carbox-
ylates dissolved in aromatic solvents will revert to the penta-
nuclear complex regardless of their ligand stoichiometry prior to
dissolution. The process of interchanging between the 1:1 dimer
and the 3:2 pentanuclear structure is accelerated by heating,
presumably acting to disrupt the extended sheet and discrete
complex structures, respectively.
In line with the results for derivatives containing carboxylates
with a second coordinating functionality,27,36ꢀ38 the two polar,
donating solvents impose a preference for a 1:1 ligand ratio in the
ethylzinc carboxylate complexes; however, whereas py acts as an
additional, permanent ligand, THF acts as a labile ligand that
does not remain bound in the solid state; as such, the pyridine
adduct forms as monomers of [EtZn(OAc)(py)], whereas the
product formed in THF exists as [EtZn(OAc)]2 dimers. The
difference between the behavior of pyridine and THF can be
explained by the higher donor number of py compared to THF
but also by the higher polarity of py (greater driving force to bind
to the metal center) and the greater acceptor ability:51 the
enhanced electrophilic nature of py compared to THF is likely
to introduce a significant back-donation (π-accepting character)
to the Znꢀpy bond.
The formation of a discrete complex with unequal numbers of
ligands (3:2 OAc:Et for the acetate derivative) in toluene and
benzene solutions is explained by the need to minimize the
polarity of the complex with respect to its outer coordination
shell: in 1, the ethyl and acetate methyl groups are directed
outward toward the solvent, keeping the polar ZnꢀO bonds
2226
dx.doi.org/10.1021/om200004a |Organometallics 2011, 30, 2223–2229

Organometallics
ARTICLE
Table 2. Copolymerization Results with Cyclohexene Oxide (CHO) and CO2 (CHO:catalyst ratio 300:1)
d
catalyst
T/°C p(CO2)/atm time/h % conversiona % carbonateb % cyclic carbonatec Mn ꢁ 103/gmolꢀ1 PDId TOFe TONf
1
3
25
25
0
0
0
10
10
24
7
5
0
0
0
464
234
1.6
1.5
2
1
1
35
16
36
[(cC5H9)7Si7O11(OSiMePh2)]2Zn4Me4 120
ref 12
13
1
80
80
80
80
80
15
15
15
15
15
80
3
24
39
32
41
31
34
39
66
66
66
67
92
3
4
2
2
2
2
44.8
2.5 24
10.2 16
10.2 31
9.9 16
128
232
184
240
185
95
1
7.5
12.9
10.8
12.9
3
3
3
7.5
7.5
Zn(OAc)2
5.33
10.6
14.2 13
[(cC5H9)7Si7O11(OSiMePh2)]2Zn4Me4 80
ref 12
24
10.8
5.3
3
1
9
SiO2-tethered ZnEt2
120
80
20
24
18
6
93
9.00
15.0
ref 12
Al(OiPr)3
ref 53
90
51
9.1
2.0
579
a Conversion was calculated as mol polymer/mol initial CHO. b Determined by comparison of the integration of the resonances at δ = 3.45 (ether
linkage) and 4.65 ppm (carbonate linkage) in the H NMR spectrum (CDCl3). Determined by comparison of the integrals of the resonances at
δ =3.45(polycarbonate) and4ꢀ4.1 (cycliccarbonate) ppm in the1H NMRspectrum(CDCl3). d Determinedbygel-permeation chromatography in THF,
using narrow Mw poly(styrene) standards. e Calculated as mol polymer/mol Zn per h. f Productivity, calculated as grams of polymer per gram of Zn used.
1
c
transfer reactions to give shorter chains. The MALDI-ToF
spectra of the copolymers catalyzed by 1 and 3 showed that,
for both catalysts, the only observable polymer end-groups were
hydroxyl groups, which is indicative of a chain transfer mechan-
ism (Figures S4 and S5, Supporting Information). In line with the
1H NMR spectroscopy results, the MALDI-ToF spectrum of the
polymer catalyzed by 1 showed much higher quantities of ether
linkages after 3 h compared to the equivalent copolymer
catalyzed by 3.
stoichiometries of 1:1 and 3:2 are interchangeable by heating in
the appropriate solvent. Copolymerization studies show promis-
ing activity toward the formation of poly(cyclohexene
carbonate), but the conditions require optimization; if the
polymer quality can be improved, the well-defined complex 3
could represent a low-cost, low-toxicity alternative to other,
small-molecule, homogeneous catalysts.
’ EXPERIMENTAL SECTION
Despite the reported14 lack of activity toward PO, Zn(OAc)2
alone was surprisingly active under the current reaction conditions
and yielded a copolymer with an equivalent proportion of ether
and carbonate linkages after 7.5 h to that produced by 1 and 3. In
general, however, both 1 and 3 showed a higher activity and higher
performance in terms of productivity (per gram Zn); in compar-
ison to the Zn(OAc)2, the Mn values of the copolymers formed by
1 and 3 were higher and the PDIs narrower, demonstrating the
improved behavior of these catalysts. In light of the expected
segregation of 1 to 3 plus Zn(OAc)2, the slightly lower activity of 1
compared to pure 3 after 3 h is explained by the lower activity of
Zn(OAc)2; the difference in reactivities means that the rates
become comparable only after 7.5 h (16 Znꢀ1 hꢀ1).
The results for 1 and 3 are comparable to recently reported
anchored alkylzinc species (Table 2, rows 10 and 11) and the
related metal alkoxide Al(iOPr)3 (Table 2, row 12) and are likely to
be similar to the zinc bis(phenoxides) reported by Darensbourg
(although those systems require ∼80 bar CO2 pressures).12,52,53
However, although the weight-average molecular weights (Mw) of
the polymers catalyzed by 1 and 3 are high and the catalyst activities
are moderate, the overall quality of the polymers is low. The
catalytic performance of 1 and 3 may be improved by rigorous
optimization; for example, it is expected that CO2 pressures of
40 atm and above would improve the carboxylate content of the
copolymers.52
General Considerations. Unless otherwise stated, all reactions
were conducted under a nitrogen atmosphere either using standard
Schlenk techniques or in a nitrogen-filled glovebox. Solvents were
distilled from sodium and stored under nitrogen. Unless otherwise
stated, solvents were degassed prior to use by performing three free-
zeꢀpumpꢀthaw cycles. Deuterated solvents were dried by placing over
calcium hydride, performing three freezeꢀthaw cycles under vacuum,
refluxing for at least 48 h, distilling under vacuum, and storing under
nitrogen. Diethylzinc was purchased from Aldrich, vacuum distilled, and
stored in an ampule, under nitrogen, at ꢀ38 °C. Infrared (IR)
spectroscopy was carried out using a Perkin-Elmer Spectrum 100
Fourier transform IR spectrometer, using dried Nujol (sodium). Ele-
mental analysis was carried out using a Carlo Erba CE1108 elemental
analyzer, and samples were manipulated under inert atmosphere
(helium glovebag); analysis was performed by Mr. S. Boyer at
London Metropolitan University, North Campus, Holloway Road,
London, N7.
In general, NMR spectra were collected on a Bruker AV-400
instrument. The 1H PGSE (DOSY) experiments were performed on a
Bruker AV-500 spectrometer, equipped with a z-gradient bbo/5 mm
tunable probe and a BSMS GAB 10 A gradient amplifier providing a
maximum gradient output of 5.35 G/cmA. All experiments were
measured using the stebpgp1s pulse program (TopSpin 2.1.3 software)
at a constant temperature of 300 K and a gas flow of 400 L per hour. The
spectra were collected at a frequency of 500.13 MHz with a spectral
width of 4000 Hz (centered on 4 ppm) and 32 768 data points. A
relaxation delay of 10 s was employed along with a diffusion time (Δ) of
70 ms. Bipolar gradient pulses (δ/2) of 2.2 ms and homospoil gradient
pulses of 1.1 ms were used. The gradient strength of the homospoil pulse
In summary, we have shown that the structure of “ethylzinc
acetate” in solution and the solid state is strongly affected by the
nature of the solvent and that the observed structures with ligand
2227
dx.doi.org/10.1021/om200004a |Organometallics 2011, 30, 2223–2229

Organometallics
ARTICLE
was ꢀ17.13%. A total of 32 experiments were collected with the bipolar
gradient strength, initially at 2% (first experiment), linearly increased to
95% (32nd experiment). All gradient pulses were sine shaped, and after
each application a recovery delay of 200 μs was used. The spectra were
processed using an exponential function with a line broadening of 2 Hz.
Further processing was achieved using the Bruker dosy software or
DOSYm software (NMRtec).
F2 refinement, R1(obs) = 0.0396, wR2(all) = 0.1342, 1232 independent
observed absorption-corrected reflections [|Fo| > 4σ(|Fo|), 2θmax
= 145°], 130 parameters. The absolute structure of 3ꢀwas determined
þ
by a combination of R-factor tests [R1 = 0.0396, R1 = 0.0409] and
by use of the Flack parameter [xþ = þ0.03(15), xꢀ = þ0.97(15)].
CCDC 799384.
Synthesis of 1, [Zn5(OAc)6(Et)4]. This is described elsewhere.39
Synthesis of 2, [EtZn(OOCCH3)((C5H5)N)]. Diethylzinc (1.0 M
in toluene, 1.90 mL, 1.90 mmol) was added to a suspension of anhydrous
zinc bis(acetate) (0.35 g, 1.89 mmol) in toluene (3 mL). Pyridine
(0.32 mL, 3.97 mmol) was added, forming a yellow solution. The
mixture was stirred for 2 h, giving a clear, colorless solution with no
particulates. Volatiles were removed in vacuo to yield a sticky, white
solid. The solid was washed with hexane (5 mL) to yield a dry, white
powder (0.32 g, 72%): 1H NMR (C6D6) δ 8.60 (m, 2 H, pyridine ortho),
6.81 (tt, 1 H, J1 = 1.6 Hz, J2 = 7.7 Hz, pyridine para), 6.55 (m, 2 H,
pyridine meta), 2.07 (s, 3 H, ꢀOOCCH3), 1.71 (t, 3 H, J = 8.0 Hz,
ꢀCH2CH3), 0.81 (q, 2 H, J = 8.0 Hz, ꢀCH2CH3) ppm; 13C{1H} NMR
(C6D6) δ 179.7 (ꢀCdO), 149.5 (pyridine ortho), 138.5 (pyridine
para), 124.9 (pyridine meta), 24.3 (ꢀOOCCH3), 14.0 (ꢀCH2CH3),
ꢀ0.8 (ꢀCH2CH3) ppm; IR (Nujol mull) 1605 (sharp, py ν(CdC)),25
1551 (ν(CdO)asymm.), 1438 (ν(CdO)symm), 1259, 1217 (sharp),
1156, 1071, 1040, 1013, 974, 941, 910, 796, 761, 703, 681, 635, 614,
590 cmꢀ1. Anal. Calcd for C9H13O2NZn: C 46.47, H 5.63, N 6.02.
Found: C 46.38, H 5.62, N 5.84.
’ ASSOCIATED CONTENT
S
Supporting Information.
Further experimental and
b
X-ray crystallographic data are available in pdf format and as a
CIF file free of charge via the Internet at http://pubs.acs.org.
’ AUTHOR INFORMATION
Corresponding Author
*E-mail: m.shaffer@imperial.ac.uk, c.k.williams@imperial.ac.uk.
’ ACKNOWLEDGMENT
This work was funded by the Engineering and Physical
Sciences Research Council (EP/C544846/1 and EP/
C544838/1). We thank Mr. P. Haycock and Mr. R. N. Sheppard
for the PGSE experiments and helpful discussions.
Synthesis of 3, [EtZn(OOCCH3)]. Diethylzinc (0.40 g,
3.24 mmol) was added to a suspension of zinc bis(acetate) (0.593 g,
3.23 mmol) in tetrahydrofuran (5 mL). The mixture was stirred for 4 h
to yield a clear, colorless solution. Volatiles were removed in vacuo to
yield a fine, white powder (0.84 g, 5.47 mmol, 85%): 1H NMR (THF-d8)
δ 1.97 (s, 3 H, ꢀOOCCH3), 1.13 (t, 3 H, J = 8.0 Hz, ꢀCH2CH3), 0.09
(q, 2 H, J = 8.0 Hz, ꢀCH2CH3) ppm; 13C{1H} NMR (THF-d8)
δ 180.2 (ꢀCdO), 23.7 (ꢀOOCCH3), 13.3 (ꢀCH2CH3), 1.7
(ꢀCH2CH3) ppm; IR (Nujol mull) 1584 (br) (ν(CdO)asymm.), 1559
(w), 1421 (ν(CdO)symm), 1362, 1319, 1232, 1177, 1165, 1039, 1022,
995, 957, 928, 900(w), 857(w), 723, 696, 623, 524 cmꢀ1. Anal. Calcd for
C4H7O2Zn: C 31.30, H 5.25. Found: C 31.27, H 5.14.
Copolymerization Reactions. In a typical experiment, a Schlenk
tube was charged with the ethylzinc carboxylate catalyst (0.66 mmol Zn)
and cyclohexene oxide (20 mL, 198 mmol) and stirred for 15 min. The
solution was transferred to an oven-dried Parr reaction vessel under
nitrogen, degassed, and placed under CO2 atmosphere. The reactor was
brought to 80 °C under 3 atm of CO2, before increasing the pressure to
15 bar and stirring for 7.5 h. The resulting polymer mixture was removed
from the reactor using dichloromethane, and the solvent removed under
reduced pressure. The polymeric product was obtained as a clear solid.
The crude material was analyzed by 1H NMR spectroscopy (to
determine the % conversion and the % ether linkages) and by GPC
(to determine the Mn, Mw, and PDI).
’ REFERENCES
(1) Seyferth, D. Organometallics 2001, 20, 2940.
(2) Frankland, E. J. Chem. Soc. 1850, 2, 263.
(3) Knochel, P.; Almena Perea, J. J.; Jones, P. Tetrahedron 1998,
54, 8275.
(4) Boudier, A.; Bromm, L. O.; Lotz, M.; Knochel, P. Angew. Chem.,
Int. Ed. 2000, 39, 4414.
(5) Negishi, E.-i.; Matsushita, H.; Okukado, N. Tetrahedron Lett.
1981, 22, 2715.
(6) Matsushita, H.; Negishi, E. J. Am. Chem. Soc. 1981, 103, 2882.
(7) Son, S.; Fu, G. C. J. Am. Chem. Soc. 2008, 130, 2756.
(8) Knochel, P.; Singer, R. D. Chem. Rev. 1993, 93, 2117.
(9) Melzig, L.; Metzger, A.; Knochel, P. J. Org. Chem. 2010, 75, 2131.
(10) Schade, M. A.; Metzger, A.; Hug, S.; Knochel, P. Chem.
Commun. 2008, 3046.
(11) Eberhardt, R.; Allmendinger, M.; Zintl, M.; Troll, C.; Luinstra,
G. A.; Reiger, B. Macromol. Chem. Phys. 2004, 205, 42.
(12) Duchateau, R.; van Meerendonk, W. J.; Huijser, S.; Staal,
B. B. P.; van Schildt, M. A.; Gerritsen, G.; Meetsma, A.; Koning, C. E.;
Kemmere, M. F.; Keurentjes, J. T. F. Organometallics 2007, 26, 4204.
(13) Inoue, S.; Koinuma, H.; Tsuruta, T. J. Polym. Sci. Part C: Polym.
Lett. 1969, 7, 287.
(14) Kobayashi, M.; Inoue, S.; Tsuruta, T. J. Polym. Sci. Polym. Chem.
Ed. 1973, 11, 2383.
(15) Luinstra, G. Polym. Rev. 2008, 48, 192.
(16) Sarazin, Y.; Schormann, M.; Bochmann, M. Organometallics
2004, 23, 3296.
(17) Driess, M.; Merz, K.; Rell, S. Eur. J. Inorg. Chem. 2000, 2517.
(18) Kim, C. G.; Sung, K.; Chung, T.; Jung, D. J.; Kim, Y. Chem.
Commun. 2003, 2068.
(19) Hambrock, J.; Rabe, S.; Merz, K.; Birkner, A.; Wohlfart, A.;
Fischer, R. A.; Driess, M. J. Mater. Chem. 2003, 13, 1731.
(20) Kahn, M.; Monge, M.; Snoeck, E.; Maisonnat, A.; Chaudret, B.
Small 2005, 1, 221.
(21) Kahn, M. L.; Cardinal, T.; Bosquet, B.; Monge, M.; Jubera, V.;
Chaudret, B. ChemPhysChem. 2006, 7, 2392.
X-ray Crystallography. Crystal data for 2: C9H13NO2Zn, M
= 232.57, monoclinic, P21/c (no. 14), a = 7.6301(3) Å, b = 16.0551(6) Å,
c = 9.2431(3) Å, β = 109.296(4)°, V = 1068.69(7) Å3, Z = 4, Dc
= 1.445 g cmꢀ3, μ(Cu KR) = 2.953 mmꢀ1, T = 273 K, colorless blocks,
Oxford Diffraction Xcalibur PX Ultra diffractometer; 2080 independent
measured reflections (Rint = 0.0287), F2 refinement, R1(obs) = 0.0399,
wR2(all) = 0.1251, 1728 independent observed absorption-corrected
reflections [|Fo| > 4σ(|F |), 2θmax = 145°], 119 parameters. CCDC
o
799383.
Crystal data for 3: C8H16O4Zn2, M = 306.95, orthorhombic, Pca21
(no. 29), a = 13.4868(5) Å, b = 9.2244(4) Å, c = 9.2241(3) Å,
V = 1147.55(8) Å3, Z = 4, Dc = 1.777 gcmꢀ3, μ(Cu KR) = 5.063
mmꢀ1, T = 173 K, colorless plates, Oxford Diffraction Xcalibur PX Ultra
diffractometer; 1596 independent measured reflections (Rint = 0.0448),
(22) Kahn, M. L.; Monge, M.; Colliere, V.; Senocq, F.; Maisonnat,
A.; Chaudret, B. Adv. Funct. Mater. 2005, 15, 458.
(23) Monge, M.; Kahn, M. L.; Maisonnat, A.; Chaudret, B. Angew.
Chem., Int. Ed. 2003, 41, 5321.
2228
dx.doi.org/10.1021/om200004a |Organometallics 2011, 30, 2223–2229

Organometallics
ARTICLE
(24) Rataboul, F.; Nayral, C.; Casanove, M.-J.; Maisonnat, A.;
Chaudret, B. J. Organomet. Chem. 2002, 643ꢀ644, 307.
(25) Lizandara-Pueyo, C.; van den Berg, M. W. E.; De Toni, A.;
Goes, T.; Polarz, S. J. Am. Chem. Soc. 2008, 130, 16601.
(26) Polarz, S.; Roy, A.; Merz, M.; Halm, S.; Schr€oder, D.; Schneider,
L.; Bacher, G.; Kruis, F. E.; Driess, M. Small 2005, 1, 540.
(27) Boyle, T. J.; Pratt, H. D. I.; Alam, T. M.; Headley, T.; Rodriguez,
M. A. Eur. J. Inorg. Chem. 2009, 855.
(28) Gonzalez-Campo, A.; Orchard, K. L.; Sato, N.; Shaffer, M. S. P.;
Williams, C. K. Chem. Commun. 2009, 27, 4034.
(29) Beek, W. J. E.; Slooff, L. H.; Wienk, M. M.; Kroon, J. M.;
Janssen, R. A. J. Adv. Funct. Mater. 2005, 15, 1703.
(30) Bruce, J. M.; Cutsforth, B. C.; Farren, D. W.; Hutchinson, F. G.;
Rabagliati, F. M.; Reed, D. R. J. Chem. Soc. (B) 1966, 1020.
(31) Allen, G.; Bruce, J. M.; Farren, D. W.; Hutchinson, F. G. J.
Chem. Soc. (B) 1966, 799.
(32) Boyle, T. J.; Bunge, S. D.; Andrews, N. L.; Matzan, L. E.; Sieg,
K.; Rodriguez, M. A.; Headley, T. J. Chem. Mater. 2004, 16, 3279.
(33) Coates, G. E.; Ridley, D. J. Chem. Soc. 1965, 1870.
(34) Inoue, S.; Kobayashi, M.; Tozuka, T. J. Organomet. Chem. 1974,
81, 17.
(35) Lewinski, J.; Bury, W.; Dutkiewicz, M.; Maurin, M.; Justyniak,
I.; Lipkowski, J. Angew. Chem., Int. Ed. 2008, 47, 573.
(36) Redshaw, C.; Elsegood, M. R. J. Angew. Chem., Int. Ed. 2007,
46, 7453.
(37) Tang, Y.; Kassel, W. S.; Zakharov, L. N.; Rheingold, A. L.;
Kemp, R. A. Inorg. Chem. 2005, 44, 359.
(38) Hursthouse, M. B.; Malik, M. A.; Motevalli, M.; O’Brien, P. J.
Chem. Soc., Chem. Commun. 1991, 1690.
(39) Orchard, K. L.; White, A. J. P.; Shaffer, M. S. P.; Williams, C. K.
Organometallics 2009, 58, 5828.
(40) Redshaw, C.; Jana, S.; Shang, C.; Elsegood, M. R. J.; Lu, X.;
Guo, Z. X. Organometallics 2010, 29, 6129ꢀ6132.
(41) Coates, G. W.; Moore, D. R. Angew. Chem., Int. Ed. 2004,
43, 6618.
(42) Darensbourg, D. J.; Mackiewicz, R. M.; Phelps, A. L.; Billodeaux,
D. R. Acc. Chem. Res. 2004, 37, 836.
(43) Kuran, W. Prog. Polym. Sci. 1998, 23, 919.
(44) Kember, M. R.; Buchard, A.; Williams, C. K. Chem. Commun.
2011, 212.
(45) Chisholm, M. H.; Navarro-Llobet, D.; Zhou, Z. Macromolecules
2002, 35, 6494.
(46) Zhu, L. Y.; Xu, D.; Wand, X. Q.; Yu, G. J. Chem. Crystallogr.
2008, 38, 609.
(47) Price, W. S. Concepts Magn. Reson. 1997, 9, 299.
(48) Perelygin, I. S.; Klimchuk, M. A. J. Appl. Spectrosc. 1974, 24, 43.
(49) Nakamoto, K. In Infrared and Raman Spectroscopy of Inorganic
and Coordination Compounds Part B: Applications on Coordination,
Organometallic, and Bioinorganic Chemistry; Wiley: New York, 1997;
pp 59ꢀ60.
(50) Fulmer, G. R.; Miller, A. J. M.; Sherden, N. H.; Gottlieb, H. E.;
Nudelman, A.; Stoltz, B. M.; Bercaw, J. E.; Goldberg, K. I., Organome-
tallics 2010, 29, 2176ꢀ2179.
(51) Gutmann, V. Coord. Chem. Rev. 1976, 18, 225.
(52) 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.
(53) Zevaco, T. A.; Janssen, A.; Sypien, J.; Dinjus, E. Green Chem.
2005, 7, 659–666.
2229
dx.doi.org/10.1021/om200004a |Organometallics 2011, 30, 2223–2229