Synthesis, structure, and reactivity of an anionic Zr-oxo relevant to CO2 reduction by a Zr/Co heterobimetallic complex

Article abstract of DOI:10.1021/ic302473j

Oxidative addition of CO₂ to the reduced Zr/Co complex (THF)Zr(MesNPⁱPr₂)₃Co (1) followed by one-electron reduction leads to formation of an unusual terminal Zr-oxo anion [2][Na(THF)₃] in low yield. To facilitate further study of this compound, an alternative high-yielding synthetic route has been devised. First, 1 is treated with CO to form (THF)Zr(MesNPⁱPr₂) ₃Co(CO) (3); then, addition of H₂O to 3 leads to the Zr-hydroxide complex (HO)Zr(MesNPⁱPr₂)₃Co(CO) (4). Deprotonation of 4 with Li(N(SiMe₃)₂) leads to the anionic Zr-oxo species [2][Li(THF)₃] or [2][Li(12-c-4)] in the absence or presence of 12-crown-4, respectively. The coordination sphere of the Li⁺ countercation is shown to lead to interesting structural differences between these two species. The anionic oxo fragment in complex [2][Li(12-c-4)] reacts with electrophiles such as MeOTf and Me₃SiOTf to generate (MeO)Zr(MesNPⁱPr₂)₃Co(CO) (5) and (Me₃SiO)Zr(MesNPⁱPr₂)₃Co(CO) (6), respectively, and addition of acetic anhydride generates (AcO)Zr(MesNP ⁱPr₂)₃Co(CO) (7). Complex [2][Li(12-c-4)] is also shown to bind CO₂ to form a monoanionic Zr-carbonate, [(12-crown-4)Li][(κ²-CO₃)Zr(MesNPⁱPr ₂)₃Co(CO)] ([8][Li(12-c-4)]). A more stable version of this compound [8][K(18-c-6)] is formed when a K⁺ counteranion and 18-crown-6 are used. Binding of CO₂ to [2][Li(12-c-4)] is shown to be reversible using isotopic labeling studies. In an effort to address methods by which these CO₂-derived products could be turned over in a catalytic cycle, it is shown that the Zr-OMe bond in 5 can be cleaved using H⁺ and the CO ligand can be released from Co under photolytic conditions in the presence of I₂.

Full text of DOI:10.1021/ic302473j

Article
pubs.acs.org/IC
Synthesis, Structure, and Reactivity of an Anionic Zr−Oxo Relevant
to CO2 Reduction by a Zr/Co Heterobimetallic Complex
Jeremy P. Krogman, Mark W. Bezpalko, Bruce M. Foxman, and Christine M. Thomas*
Department of Chemistry, Brandeis University, 415 South Street MS 015, Waltham, Massachusetts 02454, United States
S
* Supporting Information
ABSTRACT: Oxidative addition of CO2 to the reduced Zr/
Co complex (THF)Zr(MesNPiPr2)3Co (1) followed by one-
electron reduction leads to formation of an unusual terminal
Zr−oxo anion [2][Na(THF)3] in low yield. To facilitate
further study of this compound, an alternative high-yielding
synthetic route has been devised. First, 1 is treated with CO to
form (THF)Zr(MesNPiPr2)3Co(CO) (3); then, addition of
H2O to 3 leads to the Zr−hydroxide complex (HO)Zr-
(MesNPiPr2)3Co(CO) (4). Deprotonation of 4 with Li(N-
(SiMe3)2) leads to the anionic Zr−oxo species [2][Li(THF)3]
or [2][Li(12-c-4)] in the absence or presence of 12-crown-4,
respectively. The coordination sphere of the Li+ countercation
is shown to lead to interesting structural differences between these two species. The anionic oxo fragment in complex [2][Li(12-
c-4)] reacts with electrophiles such as MeOTf and Me3SiOTf to generate (MeO)Zr(MesNPiPr2)3Co(CO) (5) and
(Me3SiO)Zr(MesNPiPr2)3Co(CO) (6), respectively, and addition of acetic anhydride generates (AcO)Zr(MesNPiPr2)3Co(CO)
(7). Complex [2][Li(12-c-4)] is also shown to bind CO2 to form a monoanionic Zr−carbonate, [(12-crown-4)Li][(κ2-
CO3)Zr(MesNPiPr2)3Co(CO)] ([8][Li(12-c-4)]). A more stable version of this compound [8][K(18-c-6)] is formed when a
K+ counteranion and 18-crown-6 are used. Binding of CO2 to [2][Li(12-c-4)] is shown to be reversible using isotopic labeling
studies. In an effort to address methods by which these CO2-derived products could be turned over in a catalytic cycle, it is
shown that the Zr−OMe bond in 5 can be cleaved using H+ and the CO ligand can be released from Co under photolytic
conditions in the presence of I2.
Scheme 1
INTRODUCTION
■
Recent interest in the reduction of CO2 stems from the need
for a carbon-neutral alternative to fossil fuels.1 One CO2
activation strategy that has emerged is the use of bifunctional
transition metal complexes containing a Lewis-acidic and
Lewis-basic site to aid in CO2 binding.2−4 Early/late
heterobimetallic complexes are one particular design that has
sparked interest in this area;5−7 however, one of the potential
difficulties of employing an early metal in CO2 reduction is the
thermodynamic stability of the metal−oxygen bond. Often
cleavage of the M−O bond relies on ligand substitution,
forming another stable M−X bond (where X is a “hard”
ligand). In a recent publication, we reported the oxidative
addition of one of the CO bonds of CO2 across the metal−
metal multiple bond of the reduced Zr/Co heterobimetallic
complex (THF)Zr(MesNPiPr2)3Co (1) to form the oxo-
bridged complex (η2-iPr2PNMes)Zr(MesNPiPr2)2(μ-O)Co-
(CO) (Scheme 1).8 Use of an early/late heterobimetallic
system such as the Zr/Co complexes shown in Scheme 1 may
provide some advantages over a monometallic system for
electrocatalytic reduction of CO2. For example, the Zr/Co
interaction described in our previous reports may increase the
lability of a terminal Zr−oxo ligand produced from CO2
reduction via dative donation from Co into the Zr−O σ*
Received: November 12, 2012
Published: February 28, 2013
© 2013 American Chemical Society
3022
dx.doi.org/10.1021/ic302473j | Inorg. Chem. 2013, 52, 3022−3031

Inorganic Chemistry
Article
molecular orbital.9−11 In addition, the dative cobalt−zirconium
interaction has been shown to facilitate the two-electron
reduction taking place at cobalt by shifting the reduction
potentials anodically from −2.66 to −1.65 V.10
that this characteristic of the Zr/Co heterobimetallic system
may be responsible for the remarkably high-energy stretching
frequency of such highly reduced cobalt−carbonyls (Table 1).11
Upon investigating the subsequent reduction chemistry of
the CO2 activation product (η2-iPr2PNMes)Zr(MesNPiPr2)2(μ-
O)Co(CO), we found that while the predominant product of
Na/Hg amalgam reduction was the two-electron-reduced
carbonate species [(THF)2Na]2(CO3)Zr(MesNPiPr2)3Co-
(CO), careful treatment with one equivalent of reductant
afforded the Zr−oxo anion [(THF)3Na][OZr(MesNPiPr2)3Co-
(CO)] ([2][Na(THF)3]) in low yield (Scheme 1).8 Alternative
reducing agents such as Mg0 and Na(napthalenide) did not lead
to a cleaner, more selective reaction to form [2]−. There are
few other well-defined and structurally characterized Zr−oxo
anions in the literature, and the reactivity of these species has
not been reported.12−14 Cummins and co-workers reported a
similar C3-symmetric tris(anilido)Ti−oxo anion15 and later
reported reversible addition of CO2 to this species to form a Ti-
ligated carbonate.16 Such a process seems relevant to our
previously reported isolation of a Zr−carbonate species upon
reduction of the CO2 activation product (η2-iPr2PNMes)Zr-
(MesNPiPr2)2(μ-O)Co(CO).8 In this article, the oxo anion
[2]− is explored in more detail, looking at the effects of varying
the countercation and exploring the reactivity of this species
with electrophiles and additional CO2 in the context of
developing a potential catalytic cycle for CO2 activation/
functionalization involving the ZrCo tris(phosphinoamide)
framework.
Table 1. Zr−Co Interatomic Distances and Infrared ν(CO)
of XZr(MesNPiPr2)3Co(CO) Complexes of the ZrIVCo0
Oxidation State
compound
Zr−Co (Å)
ν(CO) (cm−1
)
a
b
[2][Li(THF)3]
2.5375(3)
2.9236(3)
2.5333(5)
2.7486(6)
2.7499(14)
2.7616(6)
2.7110(5)
2.8468(5)
2.5965(4)
1848, 1855
a
b
[2][Li(12-c-4)]
1843, 1859
b
[2][Na(THF)3]
1853
a
c
4
1893, 1888
c
5
1888
c
6
1885
c
7
1895
c
[8][K(18-c-6)]
9
1875
c
1904
a
b
IR spectrum measured in the solid state (KBr pellet). IR spectrum
c
measured in solution (THF). IR spectrum measured in C6H6
The solid state structure of compound 3 was obtained via
single-crystal X-ray diffraction (Figure S32, Supporting
Information); however, repeated attempts to crystallize this
compound resulted in a structure with THF/Cl disorder at the
axial Zr position (a result of contamination with CH2Cl2 in the
glovebox atmosphere over time), so rigorous conclusions about
the interatomic distances in 3 cannot be drawn.
Addition of 1 equiv of H2O to 3 at room temperature
resulted in immediate evolution of gas and a color change of
the reaction mixture from red to yellow (Scheme 2). The new
compound had a ν(CO) of 1888 cm−1 (Table 1) and
RESULTS AND DISCUSSION
■
The low yield (24%) of the anionic Zr−oxo species
[2][Na(THF)3]8 when generated by the reduction route
shown in Scheme 1 necessitated that a new synthetic route be
developed prior to studying its reactivity. An alternative
synthesis began by adding excess carbon monoxide to 1 to
give the diamagnetic complex (THF)Zr(MesNPiPr2)3Co(CO)
(3) in 94% isolated yield (Scheme 2). The 31P NMR spectrum
1
paramagnetically shifted resonances in its H NMR spectrum
(C6D6). In addition to the five resonances attributable to the
tris(phosphinoamide) ligand framework, the 1H NMR
spectrum features a singlet at δ −41 ppm. A single-crystal X-
ray diffraction study identified the new compound as the Zr−
hydroxide compound (HO)Zr(MesNPiPr2)3Co(CO) (4, Fig-
ure 1). The resonance at δ −41 ppm in the 1H NMR spectrum
was assigned to the hydroxide proton. While this may seem like
an unusual shift for a hydroxide proton, it is important to bear
in mind that the hydroxide ligand is not directly bound to a
paramagnetic metal center but is still influenced by the pendant
paramagnetic Co ion. This gives rise to an unusual chemical
shift, which can nonetheless be unambiguously assigned by
Scheme 2
1
1
comparison of the H NMR spectrum of 4 with the H NMR
of a species obtained upon deprotonation (vide infra): These
spectra are nearly identical, except for the absence of the δ −41
ppm resonance in the deprotonated complex. Further support
for the assignment of 4 as a Zr−hydroxide complex is the O−H
stretching modes present in the IR spectra in solution (C6H6,
3691 cm−1) and solid state (3625 cm−1). Compound 4 is
formed in 78% yield and has an S = 1/2 ground state with μeff =
2.11 μB and can be assigned a ZrIVCo0 redox state. Compound
4 has a Zr−Co distance of 2.7486(6) Å and Zr−O bond
distance of 1.909(2) Å. The Zr−Co distance in 4 is relatively
long compared to both fully reduced ZrIIICo0 complex 1 and
the ZrIVCoI dihalide complex ClZr(MesNPiPr2)3CoI.10,11 The
Zr−O bond distance in 4 is shorter than that in other reported
terminal Zr−hydroxides, such as structures reported by Parkin
and co-workers with Zr−O bond distances of 2.010(2),
2.061(7), and 2.001 (6) Å.20,21 The Zr-bound hydroxyl group
is not engaged in any intermolecular hydrogen-bonding
of 3 revealed a downfield shift to δ 60 ppm relative to the
resonance of complex 1 at δ 35 ppm. The characteristic ν(CO)
of 3 at 1890 cm−1 is similar to the carbonate species of similar
ZrIIICo0 oxidation state that we reported in an earlier
publication (ν(CO) = 1884 cm−1).8 We note that assignment
of formal oxidation states to each metal in complexes featuring
metal−metal bonds is essentially meaningless, but this
formalism is used herein to keep track of overall oxidation
state. The one-electron reactivity that we previously encoun-
tered at Zr in two-electron-reduced species is consistent with
the ZrIII oxidation state assignment.17−19 We previously
observed that backbonding from cobalt to a dinitrogen ligand
is affected by the competitive Lewis acidity of zirconium and
3023
dx.doi.org/10.1021/ic302473j | Inorg. Chem. 2013, 52, 3022−3031

Inorganic Chemistry
Article
Figure 1. Displacement ellipsoid (50%) representations of 4 (left), [2][Li(THF)3] (middle), and [2][Li(12-c-4)] (right). Hydrogen atoms except
for H21 of 4 were omitted for clarity. Relevant interatomic distances (Angstroms) and angles (degrees). 4: Zr−Co, 2.7486(6); Zr−O2, 1.909(2);
Co−C46, 1.758(4); Co−Zr−O2, 179.32(8); Co−C46−O1, 178.6(4). [2][Li(THF)3]: Zr−Co, 2.5375(3); Zr−O20, 2.130(6); Li−O20, 1.854(6);
Co−C46, 1.7445(19); Co−Zr−O20, 169.0(2); Zr−O20−Li, 157.9(5); Co−C46−O1, 179.54(17). [2][Li(12-c-4)]: Zr−Co, 2.9236(3); Zr−O2,
1.8328(11); Li−O2, 1.806(3); Co−C460, 1.765(5); Co−Zr−O2, 178.24(4); Zr−O2−Li, 176.02(12); Co−C460−O10, 179.1(4).
interactions, presumably due to the steric constraints imposed
by the bulky mesityl substituents. Structurally characterized
terminal hydroxides are relatively uncommon, likely due to the
ability of the hydroxide ligand to bridge two or more metal
centers or the propensity of early metal hydroxides to undergo
further reactivity leading to dimeric bridging oxo species.22 The
stability of the terminal hydroxide of compound 4 may be
attributed to the steric bulk of the trisphosphinoamide ligand
framework.
hydroxide proton to no longer appear in the resulting 1H NMR
spectrum (δ −41 ppm in C6D6 for 4); however, after addition
of DBU to 4 in C6D6, the resonance corresponding to the
hydroxide proton remains but is shifted downfield to δ −33
ppm. This result led us to conclude that DBU is not a strong
enough base to deprotonate the hydroxide ligand of 4 but that
an intermolecular hydrogen-bonding interaction between DBU
1
and the hydroxide proton leads to a shift in the latter’s H
NMR signal. This prediction was supported by the solid state
structure of 4·DBU (Figure S36, Supporting Information), in
which the proton associated with the hydroxide ligand was
located in the difference map and found to be ∼1.8 Å from the
basic DBU nitrogen atom. In light of this observation, a
stronger base was chosen to deprotonate 4.
The reduced byproduct of the reaction to produce 4 from
H2O has been tentatively assigned as 0.5 equiv of H2, as this
assignment would be consistent with the gas that was evolved
during reaction. A similar coupling of two methyl radicals
extruded from a one-electron oxidation reaction of a Hf/Co
complex with MeI to generate ethane has been reported and
verified spectroscopically.23 Structurally characterized Zr−
hydroxide complexes have been synthesized by substitution
reactions or addition of E−H bonds to a Zr−oxo
species.20−22,24−26 In a particularly relevant example, Rosenthal
and co-workers report a reaction involving a Ti(III) starting
material and water to produce 0.5 equiv of H2 and a TiIV−
hydroxide species.27 In contrast, Zr(II) starting materials have
been shown to react with 2 equiv of water to form a Zr−
dihydroxide complex and 1 equiv of H2.21 The mechanisms of
H2 formation from the Ti(III) and Zr(II) starting materials
differ in that the product from the Ti(III) starting material
results from a hydrogen-atom transfer (HAT) mechanism,27,28
whereas the Zr(II) starting material first reacts with 1 equiv of
water to form a dimeric μ-O Zr−hydride species which
subsequently reacts with a second equivalent of water to form
the final Zr−dihydroxide species and 1 equiv of H2.26 At first
glance, reaction of 3 with H2O appears to be most closely
related to reaction of the Ti(III) complex reported by
Rosenthal, but we concede that an intermolecular mechanism
involving a Zr−hydride or Co−hydride species cannot be ruled
out at this time. The mechanism for this reaction and other E−
H bond activations of this type by Zr/Co heterobimetallic
complexes are discussed in more detail elsewhere.17,18
To generate the desired anionic oxo species, complex 4 was
treated with 1 equiv of lithium hexamethyldisilazide (LiN-
(SiMe2)2) (Scheme 3). Monitoring the progress of the reaction
1
in THF-d8 by H NMR reveals loss of the resonance assigned
to the proton for the hydroxide ligand at δ −33.7 ppm, while
the remainder of the NMR resonances shift only slightly upon
Scheme 3
In an effort to generate a Zr−oxo anion from 4 via
deprotonation, complex 4 was treated with 1 equiv of 1,8-
diazabicycloundec-7-ene (DBU) in C6D6. If deprotonation had
occurred we would expect the peak corresponding to the
3024
dx.doi.org/10.1021/ic302473j | Inorg. Chem. 2013, 52, 3022−3031

Inorganic Chemistry
Article
Figure 2. Calculated SOMO of [2][Li(12-c-4)] (left) and β-orbital-derived O to Zr donor−acceptor interactions obtained via NBO calculations
(Edel = 16.3 and 14.7 kcal/mol, respectively). α-Orbital-derived O to Zr donor−acceptor interactions are included in the Supporting Information
(BP86/LANL2TZ(f)/6-311G(d)/D95 V).
deprotonation. The structure of the new species was
determined unambiguously by single-crystal X-ray diffraction
to be [(THF)3Li][OZr(MesNPiPr2)3Co(CO)] [2][Li(THF)3]
(Figure 1). The lithium countercation is associated with the
disordered oxo atom, with distances ranging from 1.85 to 1.90
Å, and has three THF solvate molecules. In general, the
structure bears close resemblance to that previously reported
for [2][Na(THF)3], except that the sodium ion also interacts
with the π system of a mesityl substituent in the latter
complex.8 The oxygen atom bound to zirconium is disordered
over three positions with bond distances ranging from 2.130(6)
to 2.108(2) Å. In all three disordered positions, the Zr−O bond
distance in [2][Li(THF)3] is slightly longer than in [2][Na-
(THF)3] (Zr−O bond distance = 2.071(2) Å)8 and is also
significantly longer than that in the Zr−oxo anion structure
reported by Stephan (1.847(9) Å) or the neutral Zr−oxo
structure reported by Parkin (1.804(4) Å).12,24 In this context,
it is also important to note that the Zr−Co distance (2.5375(3)
Å) in [2][Li(THF)3] is relatively short and indicative of a Zr−
Co interaction. The relatively long Zr−O bond in [2][Li-
(THF)3] appears to coincide with increased dative donation
from Co to Zr.
In an attempt sequester the lithium cation, deprotonation of
4 with Li(N(SiMe3)2) was performed in the presence of 12-
crown-4 in toluene (Scheme 3). The solid state structure
obtained via single-crystal X-ray diffraction revealed the
formulation of the resulting complex to be [(12-crown-
4)Li][OZr(MesNPiPr2)3Co(CO)] [2][Li(12-c-4)] (Figure
1). The crown ether did not sequester the lithium cation
(even when excess crown ether was added); however,
substantial structural changes occurred. The Zr−O distance
of [2][Li(12-c-4)] is 1.8328(11) Å, which is ∼0.3 Å shorter
than in [2][Li(THF)3], and the Zr−Co distance is elongated
to 2.9236(3) Å (compared to 2.5375(3) Å in [2][Li(THF)3]).
Another notable structural change is the contraction of the N−
Zr−N bond angles to an average of 115.48(5)° for compound
[2][Li(12-c-4)] from 119.05(6)° in [2][Li(THF)3] as the Zr
atom pushes out of the plane of the nitrogen donors to adopt a
geometry closer to tetrahedral. The short Zr−O bond distance
of [2][Li(12-c-4)] indicates an increase in the Zr−O bond
order, and the local 3-fold symmetry at zirconium allows
formation of one σ and two π bonds for a formal Zr−O bond
order of three. Attempting to exchange the cation in [2][Li(12-
c-4)] with a noncoordinating cation such as tetraethylammo-
nium consistently resulted in a mixture of products, most
notably reformation of compound 4. Since tetraalkylammo-
nium cations are known to undergo elimination in the presence
of strong base, we attempted cation exchange with bis-
(triphenylphosphine)iminium ([PPN]+); however, the PPN+
cation did not appear to exchange with Li+, indicating that the
coordinating cation is necessary to stabilize the anionic Zr−oxo
fragment.
Solution infrared spectra for compounds [2][Li(THF)3] and
[2][Li(12-c-4)] showed ν(CO) of 1855 (THF) and 1859 cm−1
(benzene), respectively (Table 1). These values are consistent
with the previously reported ν(CO) for [2][Na(THF)3] (1853
cm−1).8 The similarity of ν(CO) for these three compounds
was unexpected based on the drastically different solid state
structures, particularly since the stronger Zr−Co interaction in
[2][Li(THF)3] and [2][Na(THF)3] would be expected to
decrease the electron density on Co, leaving less electron
density for backbonding with CO (leading to a higher energy
CO stretch). To further investigate this phenomenon, solid
state infrared spectra (KBr pellet) were collected for
compounds [2][Li(THF)3] and [2][Li(12-c-4)] and revealed
ν(CO) of 1848 and 1843 cm−1, respectively. The difference of
only 5 cm−1 in the ν(CO) of compounds [2][Li(THF)3] and
[2][Li(12-c-4)] in the solid state suggests similar electronic
structures for the two compounds and is not consistent with
our observations that longer Zr−Co distances tend to
correspond to lower ν(CO) (Table 1).
A computational investigation using density functional theory
(DFT) was undertaken on the Li-capped anions [2][Li(12-c-
4)] and [2][Li(THF)3] to further probe their electronic
structure and, in particular, to evaluate the Zr−O bond order.
Using crystallographic coordinates as a starting point, geometry
optimizations of both compounds using Gaussian09 (BP86/
LANL2TZ(f)/6-311G(d)/D95 V)29 resulted in nearly identical
structures in stark contrast to the differences observed between
these two compounds experimentally via X-ray crystallography
(see Supporting Information). Both optimized geometries
feature a short Zr−O distance (1.86 Å) and a long Zr−Co
interatomic distance (2.95 Å), quite similar to the solid state
Zr−O and Zr−Co distances in [2][Li(12-c-4)] (1.84 and 2.92
Å, respectively). The computed SOMOs of the ZrIVCo0 [2]−
2
complexes clearly correspond to the Co dz orbital and show
essentially no interaction with the Zr center (since the two
computed structures were nearly identical, we will limit our
discussion to [2][Li(12-c-4)], whose SOMO is shown in
Figure 2), as expected based on the long Zr−Co interatomic
distance. Natural bond orbital (NBO)30 analysis of [2][Li(12-
c-4)] revealed two O→Zr donor−acceptor interactions
corresponding to O−Zr π bonds (Figure 2), verifying the
assignment as a triple-bonded Zr−oxo fragment. Moreover, the
computed Zr−O Wiberg bond index (WBI) is 1.26,
3025
dx.doi.org/10.1021/ic302473j | Inorg. Chem. 2013, 52, 3022−3031

Inorganic Chemistry
Article
corresponding to ∼2.5 times the computed Zr−N single-bond
WBI (∼0.49).
As shown in Scheme 4, reactivity of [2][Li(12-c-4)] with
acetic anhydride (Ac2O) also proceeds via electrophilic attack
of the oxo anion to produce the corresponding lithium acetate
salt and AcOZr(MesNPiPr2)3Co(CO) (7). The 1H NMR
spectrum of 7 contains a new resonance at δ 3.1 ppm assigned
to the methyl group of the acetate ligand. Single-crystal X-ray
diffraction studies show that the acetate ligand of 7 is bound in
a κ2-fashion (Figure 3). ν(CO) of 7 is at higher energy than 5
and 6 at 1895 cm−1 corresponding to a stronger Zr−Co
interaction in 7, which is further illustrated by a shorter Zr−Co
distance of 2.7110(5) Å (Table 1).
When complexes [2][Li(THF)3] and [2][Li(12-c-4)] were
isolated and stored as solids at −35 °C for later use as a starting
material, 1H NMR spectra of the solids often contained
mixtures of [2]− and 4 as a result of an unknown
decomposition pathway. Due to the sensitivity of the Zr−oxo
anion compounds, reactivity studies were performed by
generating the Zr−oxo anion [2][Li(12-c-4)] in situ. Reactions
of [2][Li(12-c-4)] with methyl trifluoromethanesulfonate
(MeOTf) and timethylsilyl trifluoromethanesulfonate
(TMSOTf) resulted in formation of the corresponding
methoxide or trimethylsiloxide complexes (Scheme 4).
Similar to the previously reported tris(anilide)Ti−oxo
anion,15,16 the Zr−oxo anion [2][Li(12-c-4)] also showed
reactivity with 1 equiv of CO2 (Scheme 5). The resulting
product crystallizes from benzene as yellow prismatic crystals.
Scheme 4
1
The H NMR spectrum of the crystals showed resonances for
bound 12-crown-4, and the paramagnetically shifted peaks
assigned to the trisphosphinoamide ligand resembled the
pattern of 7 more closely than 5 or 6, so the structure was
tentatively assigned to be [(12-crown-4)Li][(κ2-CO3)Zr-
(MesNPiPr2)3Co(CO)] ([8][Li(12-c-4)]).
Compound [8][Li(12-c-4)] is soluble in pentane, making it
difficult to isolate from any byproducts formed during the
reaction, so the analogous complex with a potassium counter-
cation and 18-crown-6 solvate, [8][K(18-c-6)], was synthe-
sized via deprotonation of 4 with K(N(SiMe3)2) in the
presence of 18-crown-6, followed by addition of 1 equiv of
CO2. Compound [8][K(18-c-6)] has ν(CO) = 1875 cm−1 and
a correspondingly long Zr−Co distance of 2.8468(5) Å (Table
1, Figure 4). The IR spectrum of [8][K(18-c-6)] also contains
an absorbance at 1642 cm−1 assigned to the carbonate ligand.
When 13CO2 is used, the stretch assigned to the carbonate
appears at 1600 cm−1 (calculated = 1605 cm−1, see Supporting
Information). Despite its paramagnetism (S = 1/2, μeff = 1.75
μB), the 13C NMR spectrum of the 13C-labeled carbonate
complex [8][K(18-c-6)] shows a resonance at δ 158 ppm
which is consistent with the value reported for the diamagnetic
Ti−carbonate species at 160 ppm.16
Compounds (MeO)Zr(MesNPiPr2)3Co(CO) (5) and
(Me3SiO)Zr(MesNPiPr2)3Co(CO) (6) have paramagnetically
shifted 1H NMR spectra with five resonances indicative of a Zr/
Co system linked by three equivalent phosphinoamide ligands.
The silyl methyl groups of 6 can be assigned to a resonance at
1
1.0 ppm in the H NMR spectrum; however, the methyl
protons of the methoxide ligand could not be located in the 1H
NMR spectrum of compound 5 and may be overlapping with a
broad resonance belonging to a ligand signal. Solid state
structures for compounds 5 and 6 were determined by single-
crystal X-ray diffraction studies (Figure 3 and Figure S39,
Supporting Information). The Zr−Co distances in structures 5
and 6 are 2.7499(14) and 2.7616(6) Å, respectively, with many
of the structural features of 5 and 6 similar to those of
hydroxide complex 4. ν(CO) stretches of 5 and 6 are 1888 and
1885 cm−1, respectively, in agreement with their closely related
solid state structures (Table 1).
Addition of CO2 to the tris(anilide)Ti−oxo anion is
reportedly reversible, with CO2 binding being disfavored in
the presence of strongly coordinating ethereal solvents.16 While
carbonate compound [8][K(18-c-6)] is stable as a solid at −35
Figure 3. Displacement ellipsoid (50%) representations of 5 (left) and 7 (right). Hydrogen atoms were omitted for clarity. Relevant interatomic
distances (Angstroms) and angles (degrees). 5: Zr−Co, 2.7499(14); Zr−O2, 1.912(5); Co−C46, 1.815(12); Co−Zr−O2, 174.43(16); Zr−O2−
C47, 168.5(6); Co−C46−O1, 178.1(9). 7: Zr−Co, 2.7110(5); Zr−O1, 2.244(2); Zr−O2, 2.253(2); Co−C46−O3, 178.8(3), O1−C47−O2,
117.7(3).
3026
dx.doi.org/10.1021/ic302473j | Inorg. Chem. 2013, 52, 3022−3031

Inorganic Chemistry
Article
Scheme 5
remains unclear, and addition of CO to oxo anion [2]− results
in no apparent reaction (although a reversible reaction to
generate CO2 cannot be ruled out at this time). However, the
reactivity of oxo anion [2]− with CO2 to form carbonate [8]−
lends insight into carbonate formation under reductive
conditions in the presence of exogenous CO2. Once a one-
electron reduction of (η2-iPr2PNMes)Zr(MesNPiPr2)2(μ-O)-
Co(CO) occurs to form an anionic oxo such as [2]−, CO2 can
bind to form an S = 1/2 carbonate species such as [8]−.
Additional one-electron reduction of this complex would lead
to formation of the diamagnetic carbonate complex
[(THF)2Na]2(CO3)Zr(MesNPiPr2)3Co(CO). Indeed, cyclic
voltammetry of [8][K(18-c-6)] shows a reversible reductive
feature around −2.0 V (see Supporting Information).
With some reactivity for [2][Li(12-c-4)] established we set
out to complete a hypothetical synthetic cycle for reduction of
CO2. One such synthetic cycle was reported by Donahue and
co-workers regarding a W(II/IV) cycle for CO2 reduction to
CO.31 As shown in Scheme 6a, addition of 1 equiv of HCl to
Figure 4. Displacement ellipsoid (50%) representations of [8][K(18-
c-6)]. Hydrogen atoms were omitted for clarity. Relevant interatomic
distances (Angstroms) and angles (degrees): Zr−Co, 2.8468(5); Zr−
O2, 2.130(2); Zr−O3, 2.136(2); K−O4, 2.548(2); Co−C46,
1.750(4); O2−C47−O3, 110.6(3); K−O4−C47, 160.3(2); Co−
C46−O1, 176.5(4).
Scheme 6
°C for more than 1 month with no signs of decomposition, we
observed decomposition of approximately 50% of [8][K(18-c-
6)] to 4 and free phosphinoamine ligand in solution after 5
days by 1H NMR spectroscopy. Formation of 4 from
[8][K(18-c-6)] indicates that carbonate formation may also
be reversible in this system. When excess CO2 was added to
13C-labeled carbonate product [8][K(18-c-6)] in a J-Young
NMR tube, the resonance assigned to the carbonate moiety at
158 ppm in the 13C NMR spectrum disappeared and was
replaced by a new resonance at 125 ppm assigned to free 13CO2
(see Supporting Information). A solution IR spectrum of the
isolated solid from this reaction confirmed that unlabeled CO2
displaced the labeled 13CO2, as a strong absorbance reappeared
at 1642 cm−1. Notably, prolonged exposure of [8][K(18-c-6)]
to dynamic vacuum did not lead to detectable CO2 loss to
regenerate the oxo species or hydroxo complex 4.
Previously, we reported that the CO2 activation product
(η2-iPr2PNMes)Zr(MesNPiPr2)2(μ-O)Co(CO) can be reduced
with excess Na/Hg to afford the two-electron-reduced Zr−
carbonate species [(THF)2Na]2(CO3)Zr(MesNPiPr2)3Co-
(CO). This product is formed in high yield in the presence
of an additional equivalent of CO2; however, it also forms in
∼40% yield when no external CO2 is added. The mechanism by
which carbonate formation occurs in the absence of added CO2
either the hydroxide complex 4 or the methoxide complex 5
affords ClZr(MesNPiPr2)3Co(CO) (9) with concomitant
release of H2O or MeOH. In the case of 5, methanol was
1
detected in the H NMR spectrum of the volatiles from this
reaction as a singlet at δ 3.07 ppm (see Supporting
Information). Compound 9 was isolated in 78% yield and
crystallized from diethyl ether to provide red-orange X-ray-
quality crystals (Figure S45, Supporting Information). ν(CO)
for 9 of 1905 cm−1 is consistent with weaker back-bonding
from Co to CO and thus a shorter Zr−Co distance (2.5965(4)
Å). Compound 9 can be rereduced to 3 using Na/Hg amalgam;
however, 3 does not react in the same manner as 1 with CO2 to
3027
dx.doi.org/10.1021/ic302473j | Inorg. Chem. 2013, 52, 3022−3031

Inorganic Chemistry
Article
afford (η2-iPr2PNMes)Zr(MesNPiPr2)2(μ-O)Co(CO), falling
short of a completed cycle for CO2 reduction. Thus, to reform
1 and complete a hypothetical cycle, compound 9 was
photolyzed in the presence of I2 to regenerate ClZr-
(MesNPiPr2)3CoI (10) in 67% isolated yield (Scheme 6b).
Formation of complex 1 from compound 10 was previously
shown to proceed readily via reduction (Scheme 6c).10,11
tometer using Cary WinUV software. Elemental analyses were
performed at Complete Analysis Laboratory Inc., Parsippany, NJ.
Solution magnetic moments were measured using Evans’ method and
are reported without taking into account any diamagnetic contribu-
tions (Pascal’s constants were not used).32,33
X-ray Crystallography Procedures. All operations were
performed on a Bruker-Nonius Kappa Apex2 diffractometer using
graphite-monochromated Mo Kα radiation. All diffractometer
manipulations, including data collection, integration, scaling, and
absorption corrections, were carried out using the Bruker Apex2
software.34 Preliminary cell constants were obtained from three sets of
12 frames. Fully labeled diagrams and data collection and refinement
details are included in Tables S1−S3 and on pages S23−S46 of the
Supporting Information file. Further crystallographic details may be
found in the accompanying CIF files.
Computational Details. All calculations were performed using
Gaussian09, Revision A.02 for the Linux operating system.29 Density
functional theory calculations were carried out using a combination of
Becke’s 1988 gradient-corrected exchange functional35 and Perdew’s
1986 electron correlation functional36 (BP86). A mixed basis set was
employed using the LANL2TZ(f) triple-ζ basis set with effective core
potentials for cobalt and zirconium,37 Gaussian09’s internal 6-
311+G(d) for heteroatoms (nitrogen, oxygen, phosphorus), and
Gaussian09’s internal LANL2DZ basis set (equivalent to D95V38) for
carbon and hydrogen. Using crystallographically determined geo-
metries as a starting point, geometries were optimized to a minimum,
followed by analytical frequency calculations to confirm that no
imaginary frequencies were present.
CONCLUSIONS
■
In summary, the reactivity of an unusual anionic Zr−oxo
species toward electrophiles and CO2 has been investigated. In
order to study this reactivity, an alternative high-yielding
synthesis of the Zr−oxo anion species has been devised. This
synthetic route involves formation of a terminal Zr−hydroxide
via a one-electron oxidation of 3 by H2O, extruding H2.
Deprotonation of the Zr−hydroxide species using Li(N-
(SiMe3)2) provides an anionic Zr−oxo species similar to that
previously reported. Interestingly, we found that many of the
structural features, including Zr−Co distance and Zr−O
distance, of these heterobimetallic species in the solid state
were highly dependent on the solvation of the alkali
countercation. Nonetheless, the solution behavior (spectrosco-
py and reactivity) of these complexes did not seem to vary with
counterion. In general, multiple bonding between the oxo
ligand and Zr appears to weaken the dative interaction between
Zr and Co, as evident by elongated metal−metal separations
and increased Co−CO back-bonding.
Electrochemistry. Cyclic voltammetry measurements were carried
out in a glovebox under a dinitrogen atmosphere in a one-
compartment cell using a CH Instruments electrochemical analyzer.
A glassy carbon electrode and platinum wire were used as the working
and auxiliary electrodes, respectively. The reference electrode was Ag/
AgNO3 in THF. Solutions (THF) of electrolyte (0.40 M [nBu4N]-
[PF6]) and analyte (2 mM) were also prepared in the glovebox.
(THF)Zr(MesNPiPr2)3Co(CO) (3). A solution containing 0.6283 g
(0.6275 mmol) of (N2)Co(iPr2PNMes)3ZrTHF in THF (40 mL) was
charged to a Schlenk tube containing a stir bar and sealed with a
Teflon valve. Approximately one-half of the volatiles were removed in
vacuo, resulting in a color change from the red N2-ligated species to
the blue/green 1. The reaction vessel was resealed, frozen, and
backfilled with excess CO(g). As the solution thawed and began
stirring it turned to a red/orange color. The solution was allowed to
stir for 10 min before being refrozen and evacuated. After thawing the
reaction mixture, volatiles were removed in vacuo until about 10 mL of
the reaction mixture remained. The THF solution was transferred to a
scintillation vial, and volatiles were removed, yielding an analytically
pure red/orange solid (0.5875 g, 93.57%). X-ray-quality red crystals
were grown from concentrated diethyl ether at −35 °C; however, in all
cases these crystals were contaminated with 9 (Cl− likely resulted from
contamination with CH2Cl2 in the glovebox atmosphere over time).
Repeated careful attempts to crystallize 3 in the absence of any Cl−
contamination did not result in any crystals suitable for X-ray
diffraction. 1H NMR (400 MHz, C6D6): δ 6.77 (s, 6H, Mes), 2.73 (m,
6H, CH(CH3)2), 2.57 (m, 4H, THF), 2.47 (s, 18H, Mes-Me), 2.13 (s,
9H, Mes-Me), 1.85 (m, 18H, CH(CH3)2), 1.53 (m, 18H, CH(CH3)2),
0.56 (m, 4H, THF). 31P NMR (162 MHz, C6D6): δ 60 (very broad).
13C NMR (100.5 MHz, C6D6): δ 149.0 (ipso-Mes), 134.5 (Mes), 131.0
(Mes), 129.1(Mes), 69.1 (THF), 44.2 (PC(CH3)2), 24.9 (THF), 24.2
(PC(CH3)2), 23.0 (PC(CH3)2, 23.2 Mes-Me, 20.7 (Mes-Me). IR
(THF): 1890 cm−1. UV−vis (C6H6, λ(nm) (ε, M−1 cm−1): 364 (sh),
456 (640), 510 (270), 1021 (110). Anal. Calcd for
C50H83CoN3O2P3Zr: C, 59.98; H, 8.36; N, 4.20. Found: C, 59.82;
H, 8.19; N, 4.14.
The Zr−oxo anion reacts readily with electrophiles such as
Me+, Me3Si+, and Ac2O to form zirconium methoxide, siloxide,
and acetate products, respectively. An additional equivalent of
CO2 also reacts with the anionic oxo fragment to generate a
Zr−carbonate anion. Carbonate formation is apparently
reversible, as elucidated through exchange between isotopically
labeled 13CO2 and CO2 in this carbonate species. Lastly, we
demonstrated that the oxoanion can be converted, albeit
through a series of stoichiometric steps, back to a complex
capable of activating additional CO2. The fundamental reaction
steps examined herein are of relevance to the electrochemical
reduction of CO2, namely, the steps required to functionalize
and cleave strong early metal−oxygen linkages. Future studies
will focus on investigations into realizing a true catalytic cycle
for CO2 reduction using these heterobimetallic Zr/Co
complexes and other heterobimetallic combinations.
EXPERIMENTAL SECTION
■
General Considerations. Unless specified otherwise, all manip-
ulations were performed under an inert atmosphere using standard
Schlenk or glovebox techniques. Glassware was oven dried before use.
Benzene, pentane, diethyl ether, tetrahydrofuran, and toluene were
dried using a Glass Contours drying column. All solvents were stored
over 3 Å molecular sieves. Benzene-d6 and toluene-d8 (Cambridge
Isotopes) were degassed via repeated freeze−pump−thaw cycles and
dried over 3 Å molecular sieves. THF-d8 was dried over CaH2, vacuum
transferred, and degassed via repeated freeze−pump−thaw cycles.
(THF)Zr(MesNPiPr2)3CoN2 was synthesized using literature proce-
dures.11 Carbon dioxide (bone dry grade 3.0) and carbon monoxide
(CP grade 2.5) were purchased from Airgas and used without further
purification. All other chemicals were purchased from commercial
(HO)Zr(MesNPiPr2)3Co(CO) (4). A solution containing 0.5777 g
(0.5769 mmol) of 3 in THF (10 mL) was charged to a Schlenk tube
containing a stir bar and sealed with a Teflon valve. A 5 mL (0.5950
mmol) amount of a 0.119 M solution of H2O in THF, prepared from
107 μL of N2-sparged deionized water and 50 mL of THF, was added
dropwise via a syringe to the stirring solution of 3 in THF. The
vendors and used without further purification. For H and 13C NMR
1
spectra the solvent resonance was referenced as an internal standard,
and for 31P{1H} NMR spectra the 85% H3PO4 resonance was
referenced as an external standard. IR spectra were recorded on a
Varian 640-IR spectrometer controlled by Resolutions Pro software.
UV−vis spectra were recorded on a Cary 50 UV−vis spectropho-
3028
dx.doi.org/10.1021/ic302473j | Inorg. Chem. 2013, 52, 3022−3031

Inorganic Chemistry
Article
reaction mixture was allowed to stir for 15 min, during which time gas
was evolved and the color of the solution changed from red to yellow.
The reaction mixture was transferred to a scintillation vial, and
volatiles were removed. The remaining yellow solid was washed with
cold pentane to yield an analytically pure solid (0.4251 g, 77.9%).
Yellow X-ray-quality crystals were grown in concentrated diethyl ether
at room temperature. 1H NMR (400 MHz, C6D6): δ 8.2 (s, 6H, Mes-
Ar), 5.5 (br s, 18H, CH(CH3)2), 3.1 (br s, 18H, CH(CH3)2), 2.3 (s,
C54H91LiN3O6P3ZrCo: C, 57.48; H, 8.13; N, 3.72. Found: C, 57.43;
H, 8.23; N, 3.70.
(MeO)Zr(MesNPiPr2)3Co(CO) (5). A solution containing 8.7 μL
(0.054 mmol) of 12-crown-4 and 0.0090 g (0.054 mmol) of lithium
hexamethyldisilazide in toluene (2 mL) was cooled in a scintillation
vial along with a separate solution of 0.0460 g (0.0486 mmol) of 4 in
toluene (2 mL) in a scintillation vial containing a stir bar. The cold
solution containing LiHMDS/12-crown-4 was added dropwise to the
stirring, cold solution of 4. The reaction mixture was stirred 10 min
before another cold solution of 5.5 μL (0.050 mmol) of MeOTf in
toluene (1 mL) was added dropwise to the stirring reaction mixture.
Volatiles were removed from the reaction mixture in vacuo. Remaining
solid was extracted into ether, and again, volatiles were removed in
vacuo, giving an analytically pure yellow solid (0.0372 g, 79.7%).
Yellow X-ray-quality crystals were grown in concentrated diethyl ether
at room temperature. 1H NMR (400 MHz, C6D6): δ 8.3 (s, 6H, Mes-
Ar), 5.5 (br s, 18H, CH(CH3)2), 3.1 (br s, 18H, CH(CH3)2), 2.3 (s,
9H, Mes-Me), −2.8 (br s, 18H, Mes-Me). IR (C6H6): 1888 cm−1.
UV−vis (C6H6, λ(nm) (ε, M−1 cm−1): 366 (sh), 505 (240), 1027
(280). Evans’ method (μeff, C6D6): 1.96 μB Anal. Calcd for
C47H78N3P3ZrCoO2: C, 58.79; H, 8.19; N, 4.38. Found: C, 58.68;
H, 8.06; N, 4.29.
1
18H, Mes-Me), −2.8 (br s, 9H, Mes-Me), −40.6 (s, 1H, OH). H
NMR (400 MHz, THF-d8): 8.0 (s, 6H, Mes-Ar), 5.3 (br s, 18H,
CH(CH3)2), 3.0 (br s, 18H, CH(CH3)2), 2.2 (s, 18H, Mes-Me), −2.9
(br s, 9H, Mes-Me), −33.7 (s, 1H, OH). IR (C6H6): 3690 cm−1(O−
H), 1888 cm−1(CO). KBr pellet: 3624 cm−1 (O−H), 1893 cm−1
(CO). UV−vis (C6H6, λ(nm) (ε, M−1 cm−1): 498 (520), 689 (220),
1045 (300). Evans’ method (μeff, C6D6): 2.11 μB. Anal. Calcd for
C46H76N3O2P3ZrCo: C, 58.39; H, 8.10; N, 4.44. Found: C, 58.30; H,
8.14; N, 4.39.
DBU-HOZr(MesNPiPr2)3Co(CO) (4·DBU). A solution containing
0.0389 g (0.0411 mmol) of 4 in C6H6 (2 mL) was prepared in a
scintillation vial containing a stir bar. A solution containing 6.4 μL
(0.043 mmol) of 1,8-diazabicycloundec-7-ene (DBU) in C6H6 was
added to the solution of 4 dropwise. The reaction mixture was allowed
to stir for 15 min, and volatiles were removed in vacuo. Pentane was
added to the remaining yellow-orange solid and placed in the freezer at
−35 °C. After 24 h, the supernatant was decanted and the resulting
solid was dried in vacuo to give 4·DBU as an analytically pure yellow-
orange solid (0.0435 g, 96.3%). Yellow-orange X-ray-quality crystals
were grown from a concentrated solution of diethyl ether at room
(Me3SiO)Zr(MesNPiPr2)3Co(CO) (6). A solution containing 4.0 μL
(0.025 mmol) of 12-crown-4 and 0.0041 g (0.025 mmol) of lithium
hexamethyldisilazide in toluene (2 mL) was cooled in a scintillation
vial along with a separate solution of 0.0211 g (0.0223 mmol) of 4 in
toluene (2 mL) in a scintillation vial containing a stir bar. The cold
solution containing LiHMDS/12-crown-4 was added dropwise to the
stirring, cold solution of 4. The reaction mixture was stirred 10 min
before another cold solution of 4.2 μL (0.023 mmol) of TMSOTf in
toluene (1 mL) was added dropwise to the stirring reaction mixture.
Volatiles were removed from the reaction mixture in vacuo. The
remaining solid was extracted into ether, and again the volatiles were
removed in vacuo, giving an analytically pure yellow solid (0.0187 g,
82.5%). Yellow, rhombohedral X-ray-quality crystals were grown from
1
temperature. H NMR (400 MHz, C6D6): δ 8.2 (s, 6H, Mes-Ar), 5.4
(br s, 18H, CH(CH3)2), 3.6 (DBU), 3.1 (br s, 18H, CH(CH3)2), 2.9
(DBU), 2.8 (DBU), 2.7 (DBU), 2.3 (s, 18H, Mes-Me),1.8 (DBU), 1.5
(DBU), 1.3 (DBU), −2.8 (br s, 9H, Mes-Me), −33.0 (s, 1H, OH). IR
(C6H6): 1886 cm−1. UV−vis (C6H6, λ(nm) (ε, M−1 cm−1): 367 (sh),
509 (200), 1027 (240). Evans’ method (μeff, C6D6): 2.11 μB Anal.
Calcd for C55H92N5O2P3CoZr: C, 60.14; H, 8.44; N, 6.38. Found: C,
59.97; H, 8.62; N, 6.42.
1
diethyl ether at −35 °C. H NMR (400 MHz, C6D6): δ 8.2 (s, 6H,
Mes-Ar), 5.6 (br s, 18H, CH(CH3)2), 3.1 (br s, 18H, CH(CH3)2), 2.3
(s, 9H, Mes-Me), 1.0 (s, 9H, Si(CH3)3), −2.7 (br s, 18H, Mes-Me). IR
(C6H6): 1885 cm−1. UV−vis λ(nm) (ε, M−1 cm−1): 504 (550), 687
(200), 1040 (230). Evans’ method (μeff, C6D6): 2.37 μB Anal. Calcd for
C49H84N3P3ZrCoSiO2: C, 57.79; H, 8.31; N, 4.13. Found: C, 57.84; H,
8.25; N, 4.02.
[(THF)3Li][OZr(MesNPiPr2)3Co(CO)] [2][Li(THF)3]. Lithium hex-
amethyldisilazide 0.0064 g (0.038 mmol) was dissolved in THF (1
mL) and frozen in a scintillation vial containing a stir bar. A separate
solution containing 0.0350 g (0.0370 mmol) of 4 in THF (2 mL) was
cooled to near freezing and pipetted onto the frozen solution of
LiHMDS. The reaction mixture was allowed to stir for 5 min while
warming to room temperature. The reaction mixture was filtered,
concentrated, and layered with pentane for crystallization at room
temperature. After 24 h, the orange X-ray-quality crystals (0.0219 g,
(κ2-AcO)Zr(MesNPiPr2)3Co(CO) (7). A solution containing 10.8
μL (0.0667 mmol) of 12-crown-4 and 0.0111 g (0.0667 mmol) of
lithium hexamethyldisilazide in toluene (2 mL) was cooled in a
scintillation vial along with a separate solution of 0.0573 g (0.0606
mmol) of 4 in toluene (2 mL) in a scintillation vial containing a stir
bar. The cold solution containing LiHMDS/12-crown-4 was added
dropwise to the stirring, cold solution of 4. The reaction mixture was
stirred 10 min before another cold solution of 5.9 μL (0.062 mmol) of
acetic anhydride in toluene (1 mL) was added dropwise to the stirring
reaction mixture. Volatiles were removed from the reaction mixture in
vacuo. The remaining solid was extracted into pentane, concentrated,
and placed in a −35 °C freezer. After 24 h X-ray-quality orange crystals
were isolated (0.0393 g, 65.6%) by decanting the remaining pentane
solution. 1H NMR (400 MHz, C6D6): δ 8.4 (s, 6H, Mes-Ar), 5.9 (br s,
18H, CH(CH3)2), 3.1 (s, 3H, C(O)CH3), 2.8 (br s, 18H, CH(CH3)2),
2.1 (s, 9H, Mes-Me), −2.0 (br s, 18H, Mes-Me). IR (C6H6): 1895
cm−1. UV−vis (C6H6, λ(nm) (ε, M−1 cm−1): 448 (800), 511 (230),
1011 (200). Evans’ method (μeff, C6D6): 1.83 μB Anal. Calcd for
C48H78N3O3P3CoZr: C, 58.34; H, 7.96; N, 4.25. Found: C, 58.29; H,
8.03; N, 4.16.
1
50.7%) were isolated from the mother liquor. H NMR (400 MHz,
THF-d8): δ 7.6 (s, 6H, Mes-Ar), 5.0 (br s, 18H, CH(CH3)2), 3.6
(THF), 3.2 (br s, 18H, CH(CH3)2), 2.2 (s, 9H, Mes-Me), 1.7 (THF)
−3.5 (br s, 18H, Mes-Me). IR (THF-d8): 1855 cm−1. KBr pellet: 1848
cm−1. UV−vis (THF, λ(nm) (ε, M−1 cm−1): 366 (sh), 506 (500),
1042 (190). Anal. Calcd for C58H99LiN3O5P3ZrCo: C, 59.62; H, 8.54;
N, 3.60. Found: C, 59.51; H, 8.59; N, 3.70.
[(12-crown-4)Li][OZr(MesNPiPr2)3Co(CO)] [2][Li(12-c-4)]. A
solution containing 20.6 μL (0.1271 mmol) of 12-crown-4 and
0.0217 g (0.1297 mmol) of lithium hexamethyldisilazide in toluene (3
mL) was cooled in a scintillation vial along with a separate solution of
0.1167 g (0.1235 mmol) of 4 in toluene (3 mL) in a scintillation vial
containing a stir bar. The cold solution containing LiHMDS/12-
crown-4 was added dropwise to the stirring, cold solution of 4. The
reaction mixture was stirred 10 min. Volatiles were removed in vacuo.
The remaining yellow solid was washed with cold pentane, giving the
product (0.1073 g, 77.04%). Yellow X-ray-quality crystals were grown
[(12-Crown-4)Li][(CO3-κ2)Zr(MesNPiPr2)3Co(CO)] [8][Li(12-c-
4)]. A solution containing 8.4 μL (0.052 mmol) of 12-crown-4 and
0.0087 g (0.052 mmol) of lithium hexamethyldisilazide in benzene (2
mL) was cooled in a scintillation vial along with a separate solution of
0.0470 g (0.0497 mmol) of 4 in benzene (2 mL) in a scintillation vial
containing a stir bar. The cold solution containing LiHMDS/12-
crown-4 was added dropwise to the stirring, cold solution of 4. The
reaction mixture and stir bar were transferred to a Schlenk tube and
1
in concentrated diethyl ether at room temperature. H NMR (400
MHz, C6D6): δ 7.9 (s, 6H, Mes-Ar), 5.3 (br s, 18H, CH(CH3)2), 4.3
(8H, 12-Crown-4), 3.5 (overlap, 12-Crown-4), 3.4 (br s, overlap,
CH(CH3)2), 2.3 (s, 9H, Mes-Me), 0.50 (br, 6H, CH(CH3)2), −3.3 (br
s, 18H, Mes-Me). IR (C6H6): 1859 cm−1. KBr pellet: 1843 cm−1. UV−
vis (C6H6, λ(nm) (ε, M−1 cm−1): 367 (sh), 466 (sh), 505 (320).
Evans’ method (μeff
, C6D6): 1.92 μB Anal. Calcd for
3029
dx.doi.org/10.1021/ic302473j | Inorg. Chem. 2013, 52, 3022−3031

Inorganic Chemistry
Article
sealed with a Teflon valve. After the reaction stirred for 10 min, the
solution was frozen and then the headspace of the flask was evacuated
and backfilled with 1.25 equiv of CO2 using a known-volume gas bulb
and partial pressure methods. After stirring for 15 min, the reaction
mixture was filtered, concentrated, and allowed to crystallize at room
temperature. After 24 h the mixture of crystals containing the product
and 4 were washed with ether to leave an analytically pure orange solid
(0.0108 g, 18.6%). 1H NMR (400 MHz, C6D6): δ 8.3 (s, 6H, Mes-Ar),
5.6 (br s, 18H, CH(CH3)2), 3.8 (s, 16H, 12-crown-4), 3.1 (br s, 18H,
CH(CH3)2), 2.3 (s, 9H, Mes-Me), −2.0 (br s, 18H, Mes-Me). IR
(C6H6): 1885 cm−1. UV−vis (C6H6, λ(nm) (ε, M−1 cm−1): 508 (90),
1018 (120). Evans’ method (μeff, C6D6): 2.04 μB Anal. Calcd for
C55H91O8N3P3ZrCoLi: C, 56.35; H, 7.82; N, 3.58. Found: C, 56.41 H,
7.92; N, 3.54.
material is available free of charge via the Internet at http://
pubs.acs.org.
AUTHOR INFORMATION
■
Corresponding Author
*E-mail: thomasc@brandeis.edu.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We acknowledge the financial support of the U.S. Department
of Energy under Award No. DE-SC0004019. C.M.T. is grateful
for a 2011 Sloan Research Fellowship.
[(18-Crown-6)K][(κ2-CO3)Zr(MesNPiPr2)3Co(CO) [8][K(18-c-
6)]. A solution containing 0.0174 g (0.0658 mmol) of 18-crown-6
and 0.0131 g (0.0658 mmol) of potassium hexamethyldisilazide in
toluene (2 mL) was cooled in a scintillation vial along with a separate
solution of 0.0593 g (0.0627 mmol) of 4 in toluene (3 mL) in a
scintillation vial containing a stir bar. The cold solution containing
KHMDS/18-crown-6 was added dropwise to the stirring, cold solution
of 4. The reaction mixture and stir bar were transferred to a Schlenk
tube and sealed with a Teflon valve. After the reaction stirred for 10
min, the solution was frozen and then the headspace of the flask was
evacuated and backfilled with 1.25 equiv of CO2 using a known-
volume gas bulb and partial pressure methods. After stirring for 15
min, the reaction mixture was filtered and volatiles were removed in
vacuo. The remaining solid was washed with pentane, leaving a yellow
analytically pure solid (0.0422 g, 52.1%). 1H NMR (400 MHz, C6D6):
δ 8.0 (s, 6H, Mes-Ar), 5.9 (br s, 18H, CH(CH3)2), 3.6 (s, 24H, 18-
crown-6), 2.9 (br s, 18H, CH(CH3)2), 2.1 (s, 9H, Mes-Me), −2.2 (br
s, 18H, Mes-Me). IR (C6H6): 1875 cm−1. UV−vis (C6H6, λ(nm) (ε,
M−1 cm−1): 510 (290), 1001 (200). Evans’ method (μeff, C6D6): 1.75
μB Anal. Calcd for C59H99N3O10KP3CoZr: C, 54.82; H, 7.72; N, 3.25.
Found: C, 54.77; H, 7.64; N, 3.25.
REFERENCES
■
(1) Lewis, N. S.; Nocera, D. G. Proc. Nat. Acad. Sci. U.S.A. 2006, 103,
15729−15735.
(2) Fachinetti, G.; Floriani, C.; Zanazzi, P. F. J. Am. Chem. Soc. 1978,
100, 7405−7407.
(3) Gambarotta, S.; Arena, F.; Floriani, C.; Zanazzi, P. F. J. Am. Chem.
Soc. 1982, 104, 5082−5092.
(4) Yin, X.; Moss, J. R. Coord. Chem. Rev. 1999, 181, 27−59.
(5) Gade, L. H. Angew. Chem., Int. Ed. 2000, 39, 2658−2678.
(6) Wheatley, N.; Kalck, P. Chem. Rev. 1999, 99, 3379−3420.
(7) Cooper, B. G.; Napoline, J. W.; Thomas, C. M. Catal. Rev. Sci.
Eng. 2012, 54, 1−40.
(8) Krogman, J. P.; Foxman, B. M.; Thomas, C. M. J. Am. Chem. Soc.
2011, 133, 14582−14585.
(9) Thomas, C. M. Comments Inorg. Chem. 2011, 32, 14−38.
(10) Greenwood, B. P.; Forman, S. I.; Rowe, G. T.; Chen, C.-H.;
Foxman, B. M.; Thomas, C. M. Inorg. Chem. 2009, 48, 6251−6260.
(11) Greenwood, B. P.; Rowe, G. T.; Chen, C.-H.; Foxman, B. M.;
Thomas, C. M. J. Am. Chem. Soc. 2010, 132, 44−45.
(12) Hoskin, A. J.; Stephan, D. W. Organometallics 1999, 18, 2479−
2483.
ClZr(MesNPiPr2)3Co(CO) (9) from 5. A solution containing
0.0089 g (0.0093 mmol) of 5 in C6D6 was added to a J-Young
NMR tube. A 9.9 μL (0.0099 mmol) amount of a 1 M HCl/ether
solution was added to the C6D6 solution of 5. After mixing, the
solution changed from yellow to orange. Volatiles of the reaction
mixture were transferred to a separate J-Young NMR tube. The orange
solid was extracted into diethyl ether. Volatiles were removed in vacuo,
leaving an analytically pure yellow/orange solid (0.0070 g, 78%). Red-
orange X-ray-quality crystals were grown from a concentrated diethyl
ether solution at −35 °C over 24 h. 1H NMR (400 MHz, C6D6): δ 8.4
(s, 6H, Mes-Ar), 5.3 (br s, 18H, CH(CH3)2), 3.9 (br, 6H, CH(CH3)2),
3.0 (br s, 18H, CH(CH3)2), 2.3 (s, 9H, Mes-Me), −2.3 (br s, 18H,
Mes-Me). IR (C6H6): 1904 cm−1. UV−vis (C6H6, λ(nm) (ε, M−1
cm−1): 403 (sh), 1025 (310). Evans’ method (μeff, C6D6): 1.81 μB
Anal. Calcd for C46H75N3OCl P3CoZr: C, 57.27; H, 7.84; N, 4.36.
Found: C, 56.98; H, 7.91; N, 4.11. Formation of MeOH was
confirmed by the presence of a singlet at 3.07 ppm in the proton NMR
of the volatiles from the reaction mixture.
(13) Jacoby, D.; Isoz, S.; Floriani, C.; Chiesi-Villa, A.; Rizzoli, C. J.
Am. Chem. Soc. 1995, 117, 2805−2816.
(14) Jacoby, D.; Floriani, C.; Chiesi-Villa, A.; Rizzoli, C. J. Am. Chem.
Soc. 1993, 115, 7025−7026.
(15) Mendiratta, A.; Figueroa, J. S.; Cummins, C. C. Chem. Commun.
2005, 3403−3405.
(16) Silvia, J. S.; Cummins, C. C. Chem. Sci. 2011, 2, 1474−1479.
(17) Napoline, J. W.; Bezpalko, M. W.; Foxman, B. M.; Thomas, C.
M. Chem. Commun. 2013, DOI: 10.1039/C2CC35594A.
(18) Napoline, J. W.; Krogman, J. P.; Shi, R.; Kuppuswamy, S.;
Bezpalko, M. W.; Foxman, B. M.; Thomas, C. M. Submitted for
publication, 2013.
(19) Zhou, W.; Marquard, S. L.; Bezpalko, M. W.; Foxman, B. M.;
Thomas, C. M. Organometallics 2013, DOI: 10.1021/om301194g.
(20) Howard, W. A.; Parkin, G. J. Am. Chem. Soc. 1994, 116, 606−
615.
(21) Howard, W. A.; Trnka, T. M.; Waters, M.; Parkin, G. J.
Organomet. Chem. 1997, 528, 95−121.
ClZr(MesNPiPr2)3CoI (10). A solution containing 0.0152 g (0.0158
mmol) of 9 in C6D6 was added to a J. Young NMR tube. Another
solution containing 0.0022 g (0.0087 mmol) of I2 in C6D6 was added
to the J. Young tube. The J. Young tube was irradiated with 300 nm
light (Rayonet RPR 3000 Å bulbs) for 3 h 45 min. The reaction
mixture was a brown-green solution. The reaction mixture was filtered,
and the volatiles were removed in vacuo. Remaining solids were
washed with pentane, leaving a green solid (0.0113 g, 67.1%), which
(22) Roesky, H. W.; Singh, S.; Yusuff, K. K. M.; Maguire, J. A.;
Hosmane, N. S. Chem. Rev. 2006, 106, 3813−3843.
(23) Setty, V. N.; Zhou, W.; Foxman, B. M.; Thomas, C. M. Inorg.
Chem. 2011, 50, 4647−4655.
(24) Howard, W. A.; Waters, M.; Parkin, G. J. Am. Chem. Soc. 1993,
115, 4917−4918.
1
was identified by comparison of its H NMR spectrum (Figure S12,
Supporting Information) to that previously reported for 10.11
(25) Jany, G.; Fawzi, R.; Steimann, M.; Rieger, B. Organometallics
1997, 16, 544−550.
ASSOCIATED CONTENT
(26) Hillhouse, G. L.; Bercaw, J. E. J. Am. Chem. Soc. 1984, 106,
5472−5478.
■
S
* Supporting Information
(27) Kessler, M.; Hansen, S.; Hollmann, D.; Klahn, M.; Beweries, T.;
Spectroscopic data for complexes 2−9, crystallographic details
and data in CIF format, and computational details. This
Spannenberg, A.; Bruckner, A.; Rosenthal, U. Eur. J. Inorg. Chem. 2011,
̈
2011, 627−631.
3030
dx.doi.org/10.1021/ic302473j | Inorg. Chem. 2013, 52, 3022−3031

Inorganic Chemistry
Article
(28) Cuerva, J. M.; Campana, A. G.; Justicia, J.; Rosales, A.; Oller-
̃
Lopez, J. L.; Robles, R.; Cardenas, D. J.; Bunuel, E.; Oltra, J. E. Angew.
́
́
̃
Chem., Int. Ed. 2006, 45, 5522−5526.
(29) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;
Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Braone, V.; Mennucci,
B.; Petersson, G. A.; et al. Gaussian 09, Revision A.1; Gaussian, Inc.:
Wallingford, CT, 2009. See Supporting Information for full reference.
(30) Glendening, E. D.; Reed, A. E.; Carpenter, J. E. NBO, Version
3.1.
(31) Jayarathne, U.; Chandrasekaran, P.; Jacobsen, H.; Mague, J. T.;
Donahue, J. P. Dalton Trans. 2010, 39, 9662−9671.
(32) Sur, S. K. J. Magn. Reson. 1989, 82, 169−173.
(33) Evans, D. F. J. Chem. Soc. 1959, 2003−2005.
(34) Apex2, Version 2 User Manual, M86-E01078; Bruker Analytical
X-ray Systems: Madison, WI, 2006.
(35) Becke, A. D. Phys. Rev. A 1988, 38, 3098−3100.
(36) Perdew, J. P. Phys. Rev. B 1986, 33, 8822−8824.
(37) Hay, P. J.; Wadt, W. R. J. Chem. Phys. 1985, 82, 299−310.
(38) Dunning, T. H., Hay, P. J. In Modern Theoretical Chemistry;
Schaefer, H. F., Ed.; Plenum: New York, 1976; pp 1−28.
3031
dx.doi.org/10.1021/ic302473j | Inorg. Chem. 2013, 52, 3022−3031