Solution and solid-state structural studies of epoxide adducts of cadmium phenoxides. Chemistry relevant to epoxide activation for ring-opening reactions

Article abstract of DOI:10.1021/ja020184c

The reaction of Cd[N(SiMe₃)₂]₂ with 2 equiv of the corresponding phenol in toluene has led to the isolation of [Cd(O-2,6-R₂C₆H₃)₂]₂ derivatives, where R represents the sterically bulky ^tBu and Ph substituents. The dimeric nature of these complexes in the solid state has been established via X-ray crystallography, i.e., trigonal geometry around cadmium is observed in 1 (R = ^tBu) where the two cadmium centers are bridged by two phenoxides with each metal containing a terminal phenoxide. Complex 2 (R = Ph) contains an additional interaction of the metal centers with carbon atoms of the aromatic substituents on the phenoxide ligands. These dimeric structures are maintained in weakly coordinating solvents as revealed by ¹¹³Cd NMR in d₂-methylene chloride, which displays ¹¹¹Cd-¹¹³Cd coupling. Nevertheless, because of the excessive steric requirements of these phenoxide ligands, these dimers are easily disrupted in solution by weak donor ligands such as epoxides. Three bisepoxide adducts have been isolated as crystalline solids and characterized by X-ray crystallography. As previously observed in other Cd(O-2,6- ^tBu₂C₆H₃)₂ ·L₂ complexes, these epoxide adducts adopt a crystallographically imposed square-planar geometry about the cadmium centers, with the exception of the exo-2,3-epoxynorbornane derivative, which displays a distorted tetrahedral geometry. Temperature-dependent ¹¹³Cd NMR studies have established that there is little difference in the binding abilities of these epoxides with either complex 1 or complex 2. Importantly, it is concluded from these studies that the lack of reactivity of α-pinene oxide and exo-2,3-epoxynorbornane toward copolymerization reactions with carbon dioxide, in the presence of zinc bisphenoxide catalysts, is not due to differences in epoxide metal binding. This is further affirmed by the isolation and crystallographic characterization of the very stable Zn(O-2,6-^tBu₂C₆H₃)₂ ·(exo-2,3-epoxynorbornane)₂ derivative.

Full text of DOI:10.1021/ja020184c

Published on Web 05/25/2002
Solution and Solid-State Structural Studies of Epoxide
Adducts of Cadmium Phenoxides. Chemistry Relevant to
Epoxide Activation for Ring-Opening Reactions
Donald J. Darensbourg,* Jacob R. Wildeson, Samuel J. Lewis, and
Jason C. Yarbrough
Contribution from the Department of Chemistry, Texas A&M UniVersity,
College Station, Texas 77843
Received February 5, 2002
Abstract: The reaction of Cd[N(SiMe3)2]2 with 2 equiv of the corresponding phenol in toluene has led to
t
the isolation of [Cd(O-2,6-R2C6H3)2]2 derivatives, where R represents the sterically bulky Bu and Ph
substituents. The dimeric nature of these complexes in the solid state has been established via X-ray
crystallography, i.e., trigonal geometry around cadmium is observed in 1 (R ) tBu) where the two cadmium
centers are bridged by two phenoxides with each metal containing a terminal phenoxide. Complex 2 (R )
Ph) contains an additional interaction of the metal centers with carbon atoms of the aromatic substituents
on the phenoxide ligands. These dimeric structures are maintained in weakly coordinating solvents as
revealed by 113Cd NMR in d2-methylene chloride, which displays 111Cd-113Cd coupling. Nevertheless,
because of the excessive steric requirements of these phenoxide ligands, these dimers are easily disrupted
in solution by weak donor ligands such as epoxides. Three bisepoxide adducts have been isolated as
crystalline solids and characterized by X-ray crystallography. As previously observed in other Cd(O-2,6-
tBu2C6H3)2‚L2 complexes, these epoxide adducts adopt a crystallographically imposed square-planar
geometry about the cadmium centers, with the exception of the exo-2,3-epoxynorbornane derivative, which
displays a distorted tetrahedral geometry. Temperature-dependent 113Cd NMR studies have established
that there is little difference in the binding abilities of these epoxides with either complex 1 or complex 2.
Importantly, it is concluded from these studies that the lack of reactivity of R-pinene oxide and exo-2,3-
epoxynorbornane toward copolymerization reactions with carbon dioxide, in the presence of zinc
bisphenoxide catalysts, is not due to differences in epoxide metal binding. This is further affirmed by the
isolation and crystallographic characterization of the very stable Zn(O-2,6-tBu2C6H3)2‚(exo-2,3-epoxynor-
bornane)2 derivative.
oxide.4 Nevertheless, these advances are somewhat overshad-
owed by the current lack of applications for these particular
Introduction
Presently, most of the commercially produced polycarbonates
are formed from the polycondensation of phosgene and diols.1
Notable exceptions are polypropylene carbonate and polyeth-
ylene carbonate which can also be synthesized from CO2 and
the corresponding epoxide in the presence of rather ineffective
heterogeneous zinc catalysts.2 Hence, it is desirable to produce
polycarbonates via this latter environmentally benign route
utilizing more effective homogeneous catalysts. These efforts
are intensified by the growing demand for these biodegradable
thermoplastics for a variety of important applications.3 In this
regard there have been tremendous gains made recently in the
area of homogeneously catalyzed production of polycarbonates
from carbon dioxide and alicyclic epoxides, e.g., cyclohexene
polymers.5 On the other hand, these catalyst systems have been
rather ineffective at coupling carbon dioxide and aliphatic
epoxides to afford polycarbonates, instead yielding mostly cyclic
carbonates.
We have recently been interested in finding alternative
alicyclic epoxides that react with carbon dioxide to provide
copolymers with Tg values more closely resembling that of the
widely applicable bisphenol-A polycarbonate.6 To this end we
report herein comparative metal binding studies of Group 12
(4) (a) Darensbourg, D. J.; Holtcamp, M. W. Macromolecules 1995, 28, 7577.
(b) Super, M.; Berluche, E.; Costello, C.; Beckman, E. Macromolecules
1997, 30, 368. (c) Super, M.; Beckman, E. J. Macromol. Symp. 1998, 127,
89. (d) Cheng, M.; Lobkovsky, E. B.; Coates, G. W. J. Am. Chem. Soc.
1998, 120, 11018. (e) 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. (f) Darensbourg,
D. J.; Wildeson, J. R.; Yarbrough, J. C.; Reibenspies, J. H. J. Am. Chem.
Soc. 2000, 122, 12487. (g) Cheng, M.; Moore, D. R.; Reczek, J. J.;
Chamberlain, B. M.; Lobkovsky, B. E.; Coates, G. W. J. Am. Chem. Soc.
2001, 123, 8738.
* Corresponding author. E-mail: djdarens@mail.chem.tamu.edu.
(1) Engineering Thermoplastics: Polycarbonates, Polyacetals, Polyesters,
Cellulose Esters; Bottenbruch, L., Ed.; Hanser Publishing: New York, 1996;
p 112.
(2) PAC Polymers Inc., a subsidiary of Axess Corporation, 100 Interchange
Boulevard, Newark, DE 19711.
(3) Cowie, J. M. G. Polymers: Chemistry and Physics of Modern Materials,
2nd ed.; Blackie Academic and Professional: London, UK, 1991.
(5) Beckman, E. J. Science 1999, 283, 946.
(6) Koning, C.; Wildeson, J.; Parton R.; Plum, B.; Steeman, P.; Darensbourg,
D. J. Polymer 2001, 42, 3995.
9
10.1021/ja020184c CCC: $22.00 © 2002 American Chemical Society
J. AM. CHEM. SOC. 2002, 124, 7075-7083
7075

A R T I C L E S
Darensbourg et al.
Synthesis of [Cd(O-2,6-Ph2C6H3)2]2, (2). A 5-mL toluene solution
(0.228 g, 0.92 mmol) of 2,6-diphenyl phenol was added to a 5-mL
toluene solution of Cd[N(SiMe3)2]2 (0.200 g, 0.46 mmol), and the clear
pale yellow solution was allowed to stir for 1 h at ambient temperature.
The solution was dried under vacuum and the light yellow solid was
recrystallized from a minimum amount of methylene chloride at -20
°C. The supernate was removed using a cannula and the crystals were
dried under vaccum to yield 0.191 g of product (68.6%). Anal. Calcd
metals with cyclohexene oxide, exo-2,3-epoxynorbornane, and
R-pinene oxide, as well as the relative activities of these
epoxides to afford polycarbonates via the homogeneously
catalyzed carbon dioxide coupling route. Specifically, we have
investigated the binding of these alicyclic epoxides with
cadmium phenoxides both in solution by 113Cd NMR and in
the solid state by X-ray crystallography. In addition, the binding
of exo-2,3-epoxynorbornane with [(2,6-tBu2C6H3O)2Zn]2, one
of the known active catalysts for the coupling of cyclohexene
oxide and carbon dioxide to produce polycarbonate,4a,e is
examined by X-ray crystallography.
1
for C72H52O4Cd2: C, 71.69; H, 4.35. Found: C, 69.66; H, 4.61. H
NMR (CD2Cl2) for bridging and terminal ligands respectively at
-40 °C: δ 6.87-7.79 [br, 13H, {2, 6-Ph2C6H3}]. 13C{H} NMR
(CD2Cl2): δ 125.9-131.6 {Ph2C6H3}{3,4,5-C6H3}; 140.4, 140.7 {2,6-
C6H3}; 155.1, 160.4 {ipso-C6H3}. 113Cd{H} NMR (CD2Cl2): δ 77.43
113
111
(J
) 95 Hz).
High-Pressure Copolymerization of CO2 with r-Pinene Oxide.
Experimental Section
Cd- Cd
Methods and Materials. Unless otherwise specified, all syntheses
and manipulations were carried out on a double manifold Schlenk
vacuum line under an atmosphere of argon or in an argon-filled
glovebox. Glassware was flamed dried thoroughly prior to use. Solvents
were freshly distilled from sodium benzophenone before use. 2,6-Di-
tert-butylphenol and 2,6-diphenylphenol were purchased from Aldrich
Chemical Co. and were sublimed and stored in a glovebox. Both
cyclohexene oxide and R-pinene oxide were purified by distillation over
calcium hydride. exo-2,3-epoxynorbornane was purchased from Aldrich
Chemical Co. and stored in a glovebox prior to use. Zn[N(SiMe3)2]2
and Cd[N(SiMe3)2]2 were prepared according to the published literature,7
stored in the glovebox, and used immediately after removal from the
box. Infrared spectra were recorded on a Mattson 6081 spectrometer
with DTGS and mercury cadmium telluride (MCT) detectors. All
isotopically labeled solvents for NMR experiments were purchased from
Cambridge Isotope Laboratories. 1H and 13C NMR spectra were
recorded on Varian XL-200E, Unity +300 MHz, and VXR 300 MHz
superconducting high-resolution spectrometers. Solution-state 113Cd
spectra were recorded on a Varian XL-400 superconducting high-
resolution spectrometer operating at 88 MHz using an external 0.1 M
Cd(ClO4)2/D2O reference. Crystals suitable for X-ray analysis of
complexes 1-5 and 7 were obtained, and X-ray data were collected at
110 K on a Bruker Smart 1000 CCD diffractometer.8-12 Elemental
analyses were carried out by Galbraith Laboratories Inc.
A sample of the catalyst [Zn(O-2,6-F2C6H3)2THF]2 (0.100 g) was
dissolved in 20.0 mL of R-pinene oxide. The solution was loaded via
a syringe port into a 150 mL stainless steel Parr autoclave that had
previously been dried overnight under vacuum at 80 °C. Two sets of
experiments were performed. The first involved a reaction pressurized
to 42 bar with CO2 and then heated to 80 °C for a 24-48 h reaction
period. The second varied from the first by raising the reaction
temperature from 80 °C to 120 °C. In addition, R-pinene oxide was
also tested for its ability to terpolymerize with CO2 and cyclohexene
oxide by using a mixture of monomers (70 mol % cyclohexene oxide
and 30 mol % R-pinene oxide). After allowing the reactor to cool to
ambient temperature, the reaction mixture was diluted with CH2Cl2 and
analyzed by infrared spectroscopy in the ν(CO2) region.
High-Pressure Copolymerization of CO2 with exo-2,3-Epoxynor-
bornane. A sample of the catalyst [Zn(O-2,6-F2C6H3)2THF]2 (0.100
g) was dissolved in 16.0 mL of a toluene/epoxynorbornane solution,
which was prepared by dissolving 10.00 g of epoxynorbornane in 12.0
mL of toluene. The resulting solution was loaded via a syringe port
into a 150 mL stainless steel Parr autoclave that had previously been
dried overnight under vacuum at 80 °C. Similar to the experiments
performed with R-pinene oxide monomer, two sets of reactions were
performed. The first involved a reaction pressurized to 42 bar with
CO2 and heated to 80 °C for a 24-48 h reaction period. The second
varied from the first by raising the reaction temperature from 80 °C to
120 °C. As with the R-pinene oxide, epoxynorbornane was also tested
for its ability to terpolymerize with CO2 and cyclohexene oxide by
using a mixture of monomers (70 mol % cyclohexene oxide and 30
mol % epoxynorbornane). Analysis of the reaction mixture was carried
out as described above using infrared spectroscopy in the ν(CO2) region.
Note! Cadmium compounds and their wastes are extremely toxic
and must be handled carefully. Cadmium waste products should be
stored in a separate, clearly marked container.
Synthesis of [Cd(O-2,6-t-Bu2C6H3)2]2, (1). A 5-mL toluene solution
(0.190 g, 0.92 mmol) of 2,6-di-tert-butyl phenol was added to a 5-mL
toluene solution of Cd[N(SiMe3)2]2 (0.20 g, 0.46 mmol). After being
stirred for 1 h at ambient temperature the clear light yellow/orange
solution was vacuum-dried and the resulting yellow solid was recrystal-
lized from a minimum amount of methylene chloride at -20 °C. The
supernate was transferred using a cannula and the crystals dried under
vacuum to yield 0.175 g of product (72.5%). Anal. Calcd for C56H84O4-
Cd2: C, 64.28; H, 8.11. Found: C, 63.37; H, 8.31. 1H NMR (CD2Cl2)
for bridging and terminal ligands respectively: δ 1.13, 1.59 [s, 18H,
{-CMe3}]; 6.46, 6.82 [t, 1H, {4-H}]; 6.92, 7.21 [d, 2H, {3,5-H}].
13C{H} NMR(CD2Cl2): δ 31.1, 33.6 {-CMe3}; 35.6, 36.5 {-CMe3};
116.7, 121.2 {4-C6H3}; 125.0, 127.4 {3, 5-C6H3}; 136.5, 139.9
{2,6-C6H3}; 165.5{ipso-C6H3}. 113Cd{H} NMR (CD2Cl2): δ 7.17
Results and Discussion
The dimeric cadmium phenoxide derivatives were synthesized
by the pathway described below and isolated in purified yields
of greater than 68%. That is, Cd[N(SiMe3)2]2 was reacted with
2 equiv of the corresponding phenol in a toluene medium
(Scheme 1). The resulting light yellow solutions were allowed
to stir at ambient temperature for 2 h, and the product was
isolated by vacuum removal of the solvent. Crystals suitable
for X-ray crystallography were obtained by dissolving the solid
product in a minimum quantity of methylene chloride and
cooling the solution at -20 °C for 3-4 days.
113
(J
111
) 117 Hz).
Cd- Cd
Scheme 1
(7) Burger, H.; Sawodny, W.; Wannaget, V. J. Organomet. Chem. 1965, 3,
113.
(8) SMART 1000 CCD; Bruker Analytical X-ray Systems: Madison, WI, 1999.
(9) Bruker, SAINT-Plus, version 6.02; Madison, WI, 1999.
(10) Sheldrick, G. SHELXS-86, Program for Crystal Structure Solution; Institut
fur Anorganische Chemie der Universitat: Tammanstrasse 4, D-3400
Gottingen, Germany, 1986.
(11) Sheldrick, G. SHELXL-97, Program for Crystal Structure Refinement;
Institut fur Anorganische Chemie der Universitat: Tammanstrasse 4,
D-3400 Gottingen, Germany, 1997.
(12) Bruker, SHELXTL, version 5.0; Madison, WI, 1999.
9
7076 J. AM. CHEM. SOC. VOL. 124, NO. 24, 2002

Epoxide Adducts of Cadmium Phenoxides
A R T I C L E S
Table 1. Selected Bond Distances (Å) and Bond Angles (deg) for
Complexes 1 and 2a
Complex 1
Cd(1)-O(1)
Cd(1)-O(2A)
O(2)-C(15)
2.0422(15) Cd(1)-O(2)
2.1954(17) O(1)-C(1)
1.374(2)
2.1773(17)
1.356(2)
O(1)-Cd(1)-O(2)
O(2)-Cd(1)-O(2A)
Cd(1)-O(2)-C(15)
142.01(6)
76.67(6)
133.50(12)
O(1)-Cd(1)-O(2A) 136.30(6)
Cd(1)-O(2)-Cd(1A) 103.33(6)
Cd(1)-O(1)-C(1)
136.05(12)
Complex 2
Cd(1)-O(1A)
Cd(1)-O(1D)
Cd(1)-O(1C)
Cd(1)-C(8A)
O(1A)-C(1A)
O(1C)-C(1C)
2.063(3)
2.231(3)
2.158(3)
2.680(5)
1.325(5)
1.357(5)
Cd(2)-O(1B)
Cd(2)-O(1D)
Cd(2)-O(1C)
Cd(2)-C(8B)
O(1B)-C(1B)
O(1D)-C(1D)
2.080(3)
2.202(3)
2.186(3)
2.622(5)
1.331(6)
1.351(5)
O(1C)-Cd(1)-O(1D) 77.03(11)
O(1C)-Cd(2)-O(1D) 77.06(11)
Cd(1)-O(1C)-Cd(2) 104.35(12)
Cd(1)-O(1D)-Cd(2) 101.42(12)
O(1A)-Cd(1)-O(1D) 103.32(11)
O(1A)-Cd(1)-O(1C) 178.62(13)
C(8A)-Cd(1)-O(1A) 78.61(13)
C(8A)-Cd(1)-O(1C) 102.64(13)
C(8A)-Cd(1)-O(1D) 102.34(13)
O(1B)-Cd(2)-O(1C) 162.07(11)
O(1B)-Cd(2)-O(1D) 119.69(12)
O(1B)-Cd(2)-C(8B) 82.70(13)
C(8B)-Cd(2)-O(1C) 102.16(13)
C(8B)-Cd(2)-O(1D) 98.14(13)
Cd(2)-O(1B)-C(1B) 127.3(3)
Cd(2)-O(1D)-C(1D) 141.2(3)
Cd(2)-O(1C)-C(1C) 125.2(3)
Cd(1)-O(1A)-C(1A) 127.6(3)
Cd(1)-O(1D)-C(1D) 116.1(3)
Cd(1)-O(1C)-C(1C) 127.9(3)
Figure 2. Thermal ellipsoid representation of complex 2, [Cd(O-2,6-
Ph2C6H3)2]2.
2.0422(15) Å is about 0.14 Å shorter than the bridging Cd-O
distances. The remaining bond angles about the cadmium centers
O(1)-Cd(1)-O(2) and O(1)-Cd(1)-O(2A) were found to be
142.01(6)° and 136.30(6)°.
a Estimated standard deviations are given in parentheses.
Figure 2 shows a thermal ellipsoid representation of complex
2, along with a partial atom-labeling scheme. By way of contrast,
2, although dimeric, contains four-coordinate cadmium centers
separated by 3.4314(8) Å which are each bonded to two bridging
phenoxide, a terminal phenoxide, and an aromatic carbon atom
from one of the phenyl substituents of the terminal phenoxide
ligands. The parallelogram formed by the two cadmium atoms
and bridging phenoxide oxygen atoms possesses average Cd-
O-Cd and O-Cd-O bond angles of 102.9° and 77.0°,
respectively. The average Cd-O bridging bond distance is
2.194[3] Å, whereas the average Cd-O terminal bond distance
is 2.072[3] Å. The six-membered ring formed via interaction
of cadmium with the terminal phenoxide ligand and one of the
carbon atoms of its phenyl substituents is in a distorted chair
conformation. The Cd(1)-C(8A) and Cd(2)-C(8B) bond
distances at 2.680(5) and 2.622(5) Å, respectively, are slightly
longer than the sum of the covalent radii of cadmium and carbon
(2.25 Å) but considerably shorter than the sum of their van der
Waal radii (3.28 Å). As a consequence of the interaction of
cadmium with the aromatic carbon the angle formed by the
bridging phenoxide-cadmium-terminal phenoxide, O(1C)-
Cd(1)-O(1A), is nearly linear at 178.62(14)°. A similar less
dramatic effect is seen at the Cd(2) center, where the O(1B)-
Cd(2)-O(1C) angle was found to be 162.07(11)°. This ad-
ditional interaction of the cadmium centers in 2 with the terminal
phenoxide ligands by way of its phenyl substituents may account
for the added stability of the dimer noted in solution (vide infra).
Complexes 1 and 2 remain dimeric upon dissolution in d2-
methylene chloride as is evident from their 113Cd or 111Cd NMR
spectra. Figure 3 displays the 113Cd NMR spectrum of complex
2 at ambient temperature, where coupling is readily seen
Figure 1. Thermal ellipsoid representation of complex 1, [Cd(O-2,6-t-
Bu2C6H3)2]2.
Complexes 1 and 2 have been characterized in the solid state
by X-ray crystallography, and a list of selected bond lengths
and angles is provided in Table 1. Figure 1 displays a thermal
ellipsoid rendering of 1, along with a partial atom-labeling
scheme. The structure of complex 1 consists of a planar
arrangement of the two cadmium atoms and the two oxygen
atoms of the bridging phenoxide ligands which form a paral-
lelogram with an O(2)-Cd(1)-O(2A) bond angle of 76.67(6)°
and a Cd(1)-O(2)-Cd(1A) bond angle of 103.33(6)°. The
cadmium centers in 1, located 3.4298(13) Å from one another,
are coordinated in a distorted trigonal planar geometry as seen
in its zinc analogue,13 with two bridging Cd-O bond distances
(Cd(1)-O(2) and Cd(1)-O(2A)) of 2.1773(17) and 2.1954(17)
Å, respectively. The terminal Cd(1)-O(1) bond distance at
1
between the two cadmium spin states of I ) /2 (i.e., 111Cd
113
113Cd
(12.75%) and Cd (12.26%)) with δ
J
) 77.4 ppm and
113Cd-111Cd
) 95 Hz. As expected from our previous studies,
the 113Cd NMR resonances are very dependent on the substit-
uents on the phenoxide ligands.14 In this instance, δ
for
113Cd
(13) Kunert, M.; Bra¨uer, M.; Klobes, O.; Go¨rls, H.; Dinjus, E.; Anders, E. Eur.
J. Inorg. Chem. 2000, 1803.
(14) Darensbourg, D. J.; Niezgoda, S. A.; Draper, J. D.; Reibenspies, J. H. J.
Am. Chem. Soc. 1998, 120, 4690.
9
J. AM. CHEM. SOC. VOL. 124, NO. 24, 2002 7077

A R T I C L E S
Darensbourg et al.
Figure 4. Thermal ellipsoid representation of complex 3, [Cd(O-2,6-t-
Bu2C6H3)2](cyclohexene oxide)2.
Figure 3. 113Cd NMR spectrum of [Cd(O-2,6-Ph2C6H3)2]2 in d2-methylene
chloride at ambient temperature.
Scheme 2
complex 1 is shifted significantly upfield at 7.17 ppm with a
113Cd-111
larger J
Cd value of 117 Hz. Upon adding various epoxides
to methylene chloride solutions of complex 1, dimer disruption
occurs with sequential formation of mononuclear mono- and
bisepoxide adducts (Scheme 2). Several of the latter derivatives
(3-5) have been isolated as single crystals and characterized
by X-ray crystallography.
Figure 5. Thermal ellipsoid representation of complex 4, [Cd(O-2,6-t-
Bu2C6H3)2](R-pinene oxide)2.
The cyclohexene oxide and R-pinene oxide bisadducts,
complexes 3 and 4, both adopt a crystallographically imposed
square-planar geometry about the metal center. This is akin to
what is observed in the THF, THT(tetrahydrothiophene), and
PC(propylene carbonate) analogues.14 On the other hand, the
bis-exo-2,3,-epoxynorbornane adduct, complex 5, exhibits a
highly distorted tetrahedral arrangement of oxygen donor ligands
about the cadmium center. Thermal ellipsoid representations of
these cadmium derivatives, which are among the few structurally
characterized metal epoxide complexes, are displayed in Figures
4-6.15-20 Table 2 contains selected bond distances and bond
angles for these epoxide complexes. Comparative bond distances
in other published Cd(O-2,6-tBuC6H3)2 adducts are provided
in Table 3. The Cd-O(epoxide) bond distances in complexes
3-5 span a very narrow range, 2.357(2) Å and 2.307[5] Å in
the cyclohexene oxide and exo-2,3-epoxynorbornane derivatives,
respectively. Unexpectedly, the Cd-O bond distance of the THF
analog was found to be significantly longer at 2.465(3) Å.14,21
The only notable difference in the solid-state structures of
complexes 3 and 4 is the average of the two carbon-O-
cadmium angles defining the epoxide cadmium interaction.
Alternatively, this angle may be defined by the Cd-O vector
and the midpoint of the OC2 plane. In complex 3 this angle is
found to be 127.52°, whereas in 4 it is more obtuse at 142.1°,
i.e., it is more pyramidal in 3. On the other hand, complex 5,
which has two crystallographically independently located ep-
oxide ligands, displays analogous angles of 129.4° and 138.3°.
The corresponding Cd-O(epoxide) bond distances in structur-
ally characterized six-coordinate cyclic ether adducts of [TpPh]-
Cd(acetate), where TpPh ) hydrotris(3-phenylpyrazol-1-yl)-
(15) Groves, J. T.; Han, Y.; van Engen, D. J. Chem. Soc., Chem. Commun.
1990, 436.
(16) Harder, S.; Boersma, J.; Brandsma, L.; Kanters, J. A.; Duisenberg, A. J.
M.; van Lenthe, J. H. Organometallics 1990, 9, 511.
(17) Darensbourg, D. J.; Holtcamp, M. W.; Khandelwal, B.; Klausmeyer, K.
K.; Reibenspies, J. H. J. Am. Chem. Soc. 1995, 117, 538.
(18) Darensbourg, D. J.; Niezgoda, S. A.; Holtcamp, M. W.; Draper, J. D.;
Reibenspies, J. H. Inorg. Chem. 1997, 36, 2424.
(19) Dias, H. V. R.; Wang, Z. Y. Inorg. Chem. 2000, 39, 3724.
(20) Beckwith, J. D.; Tschinkl, M.; Picot, A.; Tsunoda, M.; Bachman, R.; Gabba¨ı,
F. P. Organometallics 2001, 20, 3169.
(21) Goel, S. C.; Chiang, M. Y.; Buhro, W. E. J. Am. Chem. Soc. 1990, 112,
6724.
9
7078 J. AM. CHEM. SOC. VOL. 124, NO. 24, 2002

Epoxide Adducts of Cadmium Phenoxides
A R T I C L E S
Table 3. Comparative Bond Distances in Square-Planar
Cd(O-2,6-tBu2C6H3)2L2 Derivativesa
L
Cd−L (Å)
Cd−O (Å)
ref
tetrahydrofuran
2.498(5)
2.465(3)
2.768(2)
2.393(8)
2.357(2)
2.058(4)
2.068(3)
2.102(6)
2.038(8)
2.0858(19)
2.063(4)
2.063[5]
16b
9c
9
tetrahydrothiophene
propylene carbonate
cyclohexene oxide
R-pinene oxide
9
this work
this work
this work
2.336(4)
exo-2,3-epoxynorbornaned
av 2.307[5]
a All complexes have crystallographically imposed square-planar geom-
etry about Cd(II) except the last entry. b Determined at ambient temperature.
c Determined at 193 K. d Complex has a distorted tetrahedral geometry about
the Cd(II) center.
Figure 6. Thermal ellipsoid representation of complex 5, [Cd(O-2,6-t-
Bu2C6H3)2](exo-2,3-epoxynorbornane)2.
Table 2. Selected Bond Distances (Å) and Bond Angles (deg) for
Complexes 3-5a
Complex 3
Cd(1)-O(1)
O(1)-C(1A)
O(2)-C(1B)
2.357(2) Cd(1)-O(2)
1.460(4) O(1)-C(2A)
1.347(3)
2.0858(19)
1.459(4)
O(1)-Cd(1)-O(2)
C(2A)-O(1)-Cd(1)
C(1B)-O(2)-Cd(1)
86.50(8) O(2)-Cd(1)-O(1A)
128.15(17) C(1A)-O(1)-Cd(1)
122.56(17)
93.50(8)
126.89(18)
Complex 4
2.336(4) Cd(1)-O(2)
1.485(9) O(1)-C(2A)
1.349(7)
Cd(1)-O(1)
O(1)-C(1A)
O(2)-C(1B)
2.063(4)
1.455(8)
Figure 7. Temperature-dependent 113Cd NMR spectra of complex 1 in
the presence of 2 equiv of cyclohexene oxide in d2-methylene chloride.
O(1)-Cd(1)-O(2)
C(2A)-O(1)-Cd(1)
C(1B)-O(2)-Cd(1)
96.43(15) O(2)-Cd(1)-O(1A)
138.5(4)
121.2(3)
83.57(15)
145.6(4)
complex 3-5 and demonstrate a close similarity in Cd-
O(epoxide) bond distances strongly suggests that there is nothing
particularly unique about cyclohexene oxide as compared to
R-pinene oxide and exo-2,3-epoxynorbornane with regard to
metal binding and thereby activation.
Notwithstanding, it is of importance to establish the relative
metal binding abilities of these particular epoxides since their
reactivity toward homopolymerization to provide polyethers or
copolymerization with CO2 to afford polycarbonates are so
divergent. That is cyclohexene oxide readily homopolymerizes
or copolymerizes with CO2 in the presence of catalysts such as
Zn(O-2,6-R2C6H3)2‚(THF)24a,e or [Zn(O-2,6-F2C6H3)2‚THF]2,4f
whereas R-pinene and exo-2,3-epoxynorbornane exhibit no
reactivity under similar conditions (vide infra).
C(1A)-O(1)-Cd(1)
Complex 5
Cd(1)-O(1A)
Cd(1)-O(1C)
O(1A)-C(1A)
O(1C)-C(1C)
O(1D)-C(2D)
2.060(5) Cd(1)-O(1B)
2.301(5) Cd(1)-O(1D)
1.345(7) O(1B)-C(1B)
1.471(8) O(1C)-C(2C)
1.464(9) O(1D)-C(1D)
2.065(5)
2.313(5)
1.347(7)
1.472(9)
1.464(8)
O(1A)-Cd(1)-O(1B) 147.6(2)
O(1A)-Cd(1)-O(1C) 100.57(18) O(1B)-Cd(1)-O(1D) 97.25(19)
O(1B)-Cd(1)-O(1C) 95.56(19) O(1C)-Cd(1)-O(1D) 114.37(19)
C(1A)-O(1A)-Cd(1) 124.5(4)
C(1C)-O(1C)-Cd(1) 133.9(4)
C(2D)-O(1D)-Cd(1) 139.7(5)
O(1A)-Cd(1)-O(1D) 101.2(2)
C(1B)-O(1B)-Cd(1) 124.9(4)
C(2C)-O(1C)-Cd(1) 124.9(4)
C(1D)-O(1D)-Cd(1) 136.8(5)
a Estimated standard deviations are given in parentheses.
borate, similarly exhibit little variance,17,18 nevertheless, the bond
distance of the THF adduct is quite similar to the others. That
is, these Cd-O bond distances increase along the following
series: THF (2.388(8) Å) < cyclohexene oxide (2.395(4) Å)
< propylene oxide (2.414(4) Å) < dioxane (2.448(7) Å).18
Consistent with the similarities in Cd-O bond distances for
the cyclic ether adducts in this latter series, the ∆H° parameters
for binding, as determined by temperature-dependent 113Cd
NMR spectroscopy, did not vary notably.
Figures 7 and 8 contain several of the temperature-dependent
113Cd NMR traces for the addition of 2 and 4 equiv of
cyclohexene oxide(CHO) to the dimeric complex 1 in d2-
methylene chloride. As is evident from the spectra in Figure 7,
where 2 equiv of CHO were employed, initially there is a slight
excess (about 10% more) of cadmium in the form of the
monomer 1‚(CHO), (3-), relative to 1 + CHO at ambient
temperature. That is the 113Cd resonance at 7.17 ppm with
113Cd-111Cd
J
) 117 Hz is due to the dimer, complex 1, whereas
Herein, we have in like fashion measured the epoxide
interactions with the cadmium dimer complex, [Cd(O-2,6-t-
Bu2C6H3)2]2 (1), in the weakly interacting methylene chloride
solvent (Scheme 2). Our ability to isolate single crystals of
the signal at ca. -0.90 ppm is assigned to the mono-epoxide
adduct of the monomer of 1, complex 3-. Upon lowering the
temperature the 113Cd resonance due to 3- shifts downfield with
binding to the second equivalent of CHO to afford complex 3.
9
J. AM. CHEM. SOC. VOL. 124, NO. 24, 2002 7079

A R T I C L E S
Darensbourg et al.
Figure 10. Plots of the lnKeq vs 1/T for the equilibrium process defined in
eq 1 for the two sets of conditions described in Figure 9: A (2) and B (9).
line broadening of the signal due to the predominant species in
solution, complex 3. Because of solvent limitations, it was not
possible to reduce the temperature further and reach the very
slow exchange limit.
Figure 8. Temperature-dependent 113Cd NMR spectra of complex 1 in
the presence of 4 equiv of cyclohexene oxide in d2-methylene chloride.
From data such as those provided in Figure 9, coupled with
the initial concentrations of the dimeric cadmium complex (1)
and cyclohexene oxide, it is possible to calculate the equilibrium
constant between complex 3- plus cyclohexene oxide and
complex 3 (eq 1) as a function of temperature.22 That is, it can
monomer 1‚(epoxide) + epoxide h
monomer 1‚(epoxide)2 (1)
be assumed that the average 113Cd chemical shift position is
directly proportional to the percentage of each monomeric
cadmium species present in solution. Hence, temperature-
dependent Keq values for the process defined in eq 1 were
computed for two different sets of initial epoxide concentrations.
Representive plots of lnKeq vs T-1 for two different experiments
involving the binding of cyclohexene oxide to the monomer of
complex 1 are given in Figure 10. Calculation of the enthalpy,
entropy, and free energy of the equilibrium defined in eq 1 are
summarized in Table 4, where the epoxide is varied from
cyclohexene oxide to exo-2,3-epoxynorbornane.
As is obvious from the data listed in Table 4 the binding
abilities of cyclohexene oxide and exo-2,3-epoxynorbornane do
not differ significantly. Indeed, this observation is consistent
with the similarity in the Cd-Oepoxide bonding distances found
in the solid-state structures of complexes 3 and 5, despite the
difference in the geometries of these two derivatives. Qualita-
tively, the binding of the epoxides, R-pinene oxide and
propylene oxide, to the monomer of complex 1 is similar to
that quantitatively assessed above. Assuming the relatiVe binding
ability of these epoxides to the metal center of Cd(O-2,6-
tBu2C6H3)2 is not considerably different from that in the zinc
analogues, this observation is in stark contrast to the relative
ease of copolymerization of these epoxides with CO2 to afford
polycarbonates (vide infra).
Finally, it was of interest to examine the effect on metal-
epoxide binding upon changing the substituents on the phenox-
ide ligands of cadmium from tert-butyl to phenyl, for we have
previously noted that solutions of complex 2 in noncoordinating
or weakly coordinating solvents are less receptive to adding
neutral bases than are those of complex 1.4e,14,23 This trend in
Figure 9. Plot of the 113Cd NMR signal for epoxide adducts of complex
1 as a function of temperature in the presence of (A) 2 equiv of cyclohexene
oxide and (B) 4 equiv of cyclohexene oxide.
Proton NMR monitoring of the epoxide when bound and free
is consistent with this interpretation. The addition of 4 equiv of
CHO to 1 at ambient temperature initially results in >90% of
cadmium existing in the form of complex 3-; again upon
lowering the temperature the 113Cd resonance derived from a
rapid exchange of epoxide ligand between complex 3 and 3-
exhibits a linear dependence on temperature (see Figure 9),
ultimately ending up at approximately 25 ppm at -80 °C. This
latter signal is assigned to complex 3. Similar observations were
noted when complex 1 is reacted with 200 equiv of CHO. As
seen in Figure 7, there is a small downfield shift (∼2 ppm) of
the 113Cd resonance for complex 1 as the temperature is lowered
from ambient to -60/-80 °C. Also noted in Figure 7, where
the [epoxide] is deficient (mol Cd monomer/mol epoxide )
0.54), ligand exchange between 3 and 3- is slowed as the
temperature is lowered, where at -80 °C there is significant
(22) Roach, E. T.; Hardy, P. R.; Popov, A. I. Inorg. Nucl. Chem. Lett. 1973, 9,
359.
(23) (a) Darensbourg, D. J.; Zimmer, M. S.; Rainey, P.; Larkins, D. L. Inorg.
Chem. 1998, 37, 2852. (b) Darensbourg, D. J.; Zimmer, M. S.; Rainey, P.;
Larkins, D. L. Inorg. Chem. 2000, 39, 1578.
9
7080 J. AM. CHEM. SOC. VOL. 124, NO. 24, 2002

Epoxide Adducts of Cadmium Phenoxides
A R T I C L E S
Table 4. Calculated Thermodynamic Parameters for the Reaction Defined in Eq 1a
epoxide
∆H° (kJ/mol)
∆S° (J/(K mol))
∆G° (kJ/mol)
K (298K)
eq
cyclohexene oxide (2 equiv)
cyclohexene oxide (4 equiv)
exo-2,3-epoxynorbornane oxide (4.4 equiv)
-37.2 ( 1.5
-37.8 ( 2.1
-33.9 ( 0.8
-123.2 ( 6.4
-125.1 ( 8.9
-112.0 ( 3.4
-0.50 ( 0.37
-0.47 ( 0.52
-0.49 ( 0.19
1.23
1.21
1.22
a
Measurements carried out in methylene chloride.
Figure 12. Temperature-dependent 113Cd NMR spectra of complex 2 in
the presence of 20 equiv of propylene oxide.
Figure 11. Temperature-dependent 113Cd NMR spectra of complex 2 in
the presence of 20 equiv of cyclohexene oxide.
metal-ligand binding was though to reflect greater steric
requirements for the 2,6-diphenyl ligands. Figures 11 and 12
depict the 113Cd NMR spectra of two identically performed
experiments involving the addition of 20 equiv of cyclohexene
oxide and propylene oxide(PO), respectively, to complex 2 in
d2-methylene chloride. As seen for other Lewis bases, complex
2 is less reactive toward epoxides when compared with complex
1. For example, complex 1 undergoes almost complete formation
of monomer 1‚(CHO) in the presence of 4 equiv of epoxide at
ambient temperature. By way of contrast, complex 2 in the
presence of 20 equiv of CHO or PO at ambient temperature
exists in rapid exchange with the monomer 2‚(epoxide) mostly
in the form of the dimer (recall that the 113Cd resonance for the
dimer is found at 77.4 ppm). There is a slight preference of
complex 2 reacting with CHO as compared to PO. Upon further
lowering of the temperature the equilibrium defined in eq 1 for
monomer 2 is established with Keq values of 1.27 and 0.547 at
-40 °C for CHO and PO, respectively. From the temperature-
dependent equilibrium constants, ∆H° values of -27.6 and
-29.9 kJ/mol were determined. Hence, there is not a great deal
of difference in the enthalpy values for reaction 1 as the
phenoxide ligands are varied from 2,6-di-tert-butyl phenolate
(see Table 4) to 2,6-diphenylphenolate. It is therefore possible
to conclude from these studies that the reduction in reactivity
of the dimeric complex 2 with Lewis bases as compared to
complex 1 is the result of dimer 2 being inherently more stable.
This added stability of complex 2 in solution is most likely due
to additional metal interactions with the phenolate ligands via
their aromatic substituents as is observed in the solid-state
structure of 2 (vide supra).
As mentioned earlier, we are intensely interested in synthesiz-
ing 1:1 alternating copolymers of carbon dioxide and alicyclic
epoxides other than cyclohexene oxide. This is in an effort to
prepare polycarbonates with Tg values closer to that of bisphe-
nol-A polycarbonate. That is, the Tg of poly(cyclohexane
carbonate) at 115 °C is some 35 °C lower than that of
bisphenol-A polycarbonate at 150 °C, which dramatically limits
the high-temperature applications of the former polycarbonate.6
Therefore, it was desirable to employ bulkier, readily available
epoxides such as R-pinene oxide or exo-2,3-epoxynorbornane
epoxides (Scheme 3) in hopes of preparing polycarbonate with
higher Tg values than 115 °C. It was felt that these bulky
epoxides should show a decrease in internal motion of the
polymer chains upon applying thermal energy. As a result, the
restriction of intermolecular motion should raise the Tg values
of the copolymer with respect to that of the cyclohexene oxide
9
J. AM. CHEM. SOC. VOL. 124, NO. 24, 2002 7081

A R T I C L E S
Darensbourg et al.
Scheme 3
derived copolymer. Indeed, the pinene oxide derived copolymer
might be semicrystalline.
Unfortunately, reactions of R-pinene oxide or exo-2,3-
epoxynorbornane under 65 bar of carbon dioxide pressure in
the presence of [Zn(O-2,6-F2C6H3)2‚THF]2 at 80 and 120 °C
resulted in no copolymer production or other products resulting
from CO2/epoxide coupling reactions, such as cyclic carbonates.
Under similar or milder reaction conditions, zinc bisphenoxide
derivatives have proven to be extremely effective catalysts for
the copolymerization of cyclohexene oxide and carbon dioxide
to polycarbonates. Furthermore, mixtures of 70 mol % cyclo-
hexene oxide and 30 mol % R-pinene oxide or exo-2,3-
epoxynorbornane were equally unsuccessful at producing ter-
polymers with carbon dioxide under these conditions. In these
instances only poly(cyclohexene carbonate) was produced in
greatly reduced yields due to competitive binding at the zinc
center of the unreactive epoxides, R-pinene oxide and exo-2,3-
epoxynorbornane. Consistent with these negative observations,
the zinc analogue of complex 1, [Zn(O-2,6-tBu2C6H3)2]2, rapidly
catalyzes the homopolymerization of cyclohexene oxide to
polyethers at ambient temperature, making it extremely difficult
to obtain single crystals of Zn(O-2,6-tert-butylC6H3)2‚(CHO)2.
However, it is unreactive at homopolymerizing exo-2,3-ep-
oxynorbornane, instead affording stable crystals of Zn(O-2,6-
tert-butylC6H3)2(exo-2,3-epoxynorbornane)2 (7). The X-ray
structure of complex 7 was found to be isostructural to that
observed for its cadmium analogue, complex 5. Indeed, the
metal oxygen bond distance differences of 0.18 and 0.20 Å are
very close to the difference in covalent radii of zinc and
cadmium (0.17 Å). A thermal ellipsoid representation of
complex 7 is shown in Figure 13, with comparative selected
bond distances and bond angles to those of its cadmium
analogue being listed in Table 5. Hence, it is obvious that the
lack of activity displayed by zinc bisphenoxide derivatives for
the homopolymerization of R-pinene oxide or exo-2,3-epoxy-
norbornane to polyethers or the copolymerization of these
epoxides and carbon dioxide to polycarbonates is not the result
of the epoxides not binding to the metal center, but instead due
to a higher reaction barrier for the epoxide ring-opening process.
Figure 13. Thermal ellipsoid representation of complex 7, [Zn(O-2,6-t-
Bu2C6H3)2](exo-2,3-epoxynorbornane)2.
Table 5. Comparative Selected Bond Distances (Å) and Bond
Angles (deg) in M(O-2,6-tBu2C6H3)2‚(exo-2,3-epoxynorbornane)2
Complexes
Zn
Cd
av M-O(phenoxide)
av M-O(epoxide)
Oep-M-Oep
Ophen-M-Ophen
av Oep-M-Ophen
av C-O-M
1.880[3]
2.108[3]
112.69(12)
143.64(13)
99.95[12]
134.4[2]
2.063[5]
2.307[5]
114.37(19)
147.6(2)
98.65[19]
130.85[5]
a Avereage standard derivatives in bond angles and distances are given
in brackets.
dimers and epoxides to be sequential, with initial formation of
the mono adduct at ambient temperature followed by a rapid
equilibrium process leading to bisadduct formation as the
temperature is lowered. From the 113Cd NMR spectra, equilib-
rium constants as a function of temperature for epoxide binding
were determined which revealed little difference in binding
between [Cd(O-2,6-tBu2C6H3)2]2 and cyclohexene oxide or exo-
2,3-epoxynorbornane, as well as the other epoxides such as
propylene oxide and R-pinene. Although the relative affinity
for epoxide binding of these epoxides to the [Cd(O-2,6-
Ph2C6H3)2]2 derivative was comparable, there was a slight
diminution in the binding to cadmium. The similarity of the
binding abilities of the epoxides, cyclohexene oxide and exo-
2,3-epoxynorbornane, to group 13 metal centers is in stark
contrast to the reactivity of these systems for affecting epoxide
ring-opening reactions. For example, whereas [Zn(O-2,6-tBu-
C6H3)2]2 rapidly catalyzes cyclohexene oxide homopolymeri-
zation to polyether or copolymerization with carbon dioxide to
polycarbonate, it only reacts with exo-2,3-epoxynorbornane to
afford the stable, crystallographically characterized Zn(O-2,6-
tBu2C6H3)2‚(exo-2,3-epoxynorbornane)2 complex. In part this
difference in behavior of cyclohexene oxide and exo-2,3-
epoxynorbornane is due to the reduction in epoxide ring strain
energy for the latter epoxide, which we estimate to be about 8
kcal/mol.
Summary
This study has demonstrated that phenoxides containing
sterically demanding substituents (tBu or Ph) in the 2,6-positions
can nevertheless serve as ligands capable of forming bisphe-
noxide derivatives of cadmium which exist as dimers both in
the solid state and in weakly coordinating solvents. These
dimeric complexes are readily disrupted in methylene chloride
solution by a variety of epoxides leading to crystalline bisep-
oxide adducts of [Cd(O-2,6-R2C6H3)2]. Indeed, these complexes
represent some of the few metal derivatives containing bound
epoxides as ligands which have been characterized by X-ray
crystallography. In methylene chloride solution 113Cd NMR
spectroscopy has shown the reaction between the cadmium
Finally, a few words are warranted with regard to the
unanticipated geometry noted herein and elsewhere for Cd(O-
2,6-tBu2C6H3)2L2 (L ) neutral donor ligand) derivatives.9,16 That
9
7082 J. AM. CHEM. SOC. VOL. 124, NO. 24, 2002

Epoxide Adducts of Cadmium Phenoxides
A R T I C L E S
is, in all instances with the exception of the one case reported
upon here (L ) exo-2,3-epoxynorbornane) the cadmium center
displays square-planar geometry in the presence of weakly
coordinating ligands such as THF, THT, propylene carbonate,
and epoxides. On the other hand, when L is a strongly donating
ligand such as PMe3, the expected distorted tetrahedral geometry
is observed.24 Previously, Buhro and co-workers have addressed
this issue in a qualitative manner for the first observed THF
adduct in terms of a linear bisphenoxide cadmium model with
two weakly ligating neutral bases.21 The distorted tetrahedral
structure determined for complex 5 would not be accommodated
by this explanation alone. Presently, we are carrying out rigorous
computational studies on the zinc and cadmium complexes, and
our preliminary results indicate that although there is a
significant difference in energy for the tetrahedral and square-
planar geometries in Zn(O-2,6-tBu2C6H3)2L2 derivatives, the
cadmium analogues show only a small difference in the stability
of the two geometries. Nevertheless, the tetrahedral form is
favored in both instances. These computational investigations
will be reported in detail upon their completion.
Acknowledgment. Financial support from the National Sci-
ence Foundation (CHE 99-10342 and CHE 98-07975 for the
purchase of X-ray equipment) and the Robert A. Welch
Foundation is greatly appreciated.
Supporting Information Available: Complete experimental
details of the X-ray diffraction studies on complexes 1-5 and
7 (PDF). This material is available free of charge via the Internet
at http://pubs.acs.org.
(24) Darensbourg, D. J.; Rainey, P.; Larkins, D. L.; Reibenspies, J. H. Inorg.
Chem. 2000, 39, 473.
JA020184C
9
J. AM. CHEM. SOC. VOL. 124, NO. 24, 2002 7083