Oxyfunctionalization with CpIrIII(NHC)(Me)(Cl) with O2: Identification of a Rare Bimetallic IrIV μ-Oxo Intermediate

Article abstract of DOI:10.1021/ja512905t

Methanol formation from [CpIr^III(NHC)Me(CD₂Cl₂)]⁺ occurs quantitatively at room temperature with air (O₂) as the oxidant and ethanol as a proton source. A rare example of a diiridium bimetallic complex, [(CpIr(NHC)Me)₂(μ-O)][(BAr^F₄)₂], 3, was isolated and shown to be an intermediate in this reaction. The electronic absorption spectrum of 3 features a broad observation at ～660 nm, which is primarily responsible for its blue color. In addition, 3 is diamagnetic and can be characterized by NMR spectroscopy. Complex 3 was also characterized by X-ray crystallography and contains an Ir^IV-O-Ir^IV core in which two d⁵ Ir(IV) centers are bridged by an oxo ligand. DFT and MCSCF calculations reveal several important features of the electronic structure of 3, most notably, that the μ-oxo bridge facilitates communication between the two Ir centers, and σ/π mixing yields a nonlinear arrangement of the μ-oxo core (Ir-O-Ir ～ 150°) to facilitate oxygen atom transfer. The formation of 3 results from an Ir oxo/oxyl intermediate that may be described by two competing bonding models, which are close in energy and have formal Ir-O bond orders of 2 but differ markedly in their electronic structures. The radical traps TEMPO and 1,4-cyclohexadiene do not inhibit the formation of 3; however, methanol formation from 3 is inhibited by TEMPO. Isotope labeling studies confirmed the origin of the methyl group in the methanol product is the iridium-methyl bond in the [CpIr(NHC)Me(CD₂Cl₂)][BAr^F₄] starting material. Isolation of the diiridium-containing product [(CpIr(NHC)Cl)₂][(BAr^F₄)₂], 4, in high yields at the end of the reaction suggests that the Cp and NHC ligands remain bound to the iridium and are not significantly degraded under reaction conditions. (Chemical Presented).

Full text of DOI:10.1021/ja512905t

Article
pubs.acs.org/JACS
Oxyfunctionalization with Cp*IrIII(NHC)(Me)(Cl) with O2: Identification
of a Rare Bimetallic IrIV μ‑Oxo Intermediate
Matthew C. Lehman,‡ Dale R. Pahls,† Joseph M. Meredith,§ Roger D. Sommer,‡ D. Michael Heinekey,§
Thomas R. Cundari,† and Elon A. Ison*,‡
‡Department of Chemistry, North Carolina State University, 2620 Yarbrough Drive, Raleigh, North Carolina 27695-8204, United
States
†Department of Chemistry, Center for Advanced Scientific Computing and Modeling (CASCaM), University of North Texas,
Denton, Texas 76203, United States
§Department of Chemistry, University of Washington, Box 351700, Seattle, Washington 98195-1700, United States
S
* Supporting Information
ABSTRACT: Methanol formation from [Cp*IrIII(NHC)Me(CD2Cl2)]+
occurs quantitatively at room temperature with air (O2) as the oxidant
and ethanol as a proton source. A rare example of a diiridium bimetallic
complex, [(Cp*Ir(NHC)Me)2(μ-O)][(BArF )2], 3, was isolated and
4
shown to be an intermediate in this reaction. The electronic absorption
spectrum of 3 features a broad observation at ∼660 nm, which is
primarily responsible for its blue color. In addition, 3 is diamagnetic and
can be characterized by NMR spectroscopy. Complex 3 was also
characterized by X-ray crystallography and contains an IrIV−O−IrIV core
in which two d5 Ir(IV) centers are bridged by an oxo ligand. DFT and
MCSCF calculations reveal several important features of the electronic
structure of 3, most notably, that the μ-oxo bridge facilitates communication between the two Ir centers, and σ/π mixing yields a
nonlinear arrangement of the μ-oxo core (Ir−O−Ir ∼ 150°) to facilitate oxygen atom transfer. The formation of 3 results from
an Ir oxo/oxyl intermediate that may be described by two competing bonding models, which are close in energy and have formal
Ir−O bond orders of 2 but differ markedly in their electronic structures. The radical traps TEMPO and 1,4-cyclohexadiene do
not inhibit the formation of 3; however, methanol formation from 3 is inhibited by TEMPO. Isotope labeling studies confirmed
the origin of the methyl group in the methanol product is the iridium−methyl bond in the [Cp*Ir(NHC)Me(CD2Cl2)][BArF ]
4
starting material. Isolation of the diiridium-containing product [(Cp*Ir(NHC)Cl)2][(BArF )2], 4, in high yields at the end of the
4
reaction suggests that the Cp* and NHC ligands remain bound to the iridium and are not significantly degraded under reaction
conditions.
respect to the metal, were observed.3f This proposal was further
corroborated by computational studies, which indicated that H
INTRODUCTION
■
To develop catalysts that employ molecular oxygen/air as an
atom abstraction was the operative mechanism for O2
oxidant for the oxidation of hydrocarbons, it is important to
insertion.5 In the case of insertions into both PdII and PtII
understand the mechanism of the direct insertion of dioxygen
methyl complexes, radical chain pathways were proposed.3b,c,6
into transition metal carbon and hydrogen bonds.1 The
In contrast with the examples above with palladium and
mechanism for C−H activation has been well established
over the past 40 years.2 In contrast, mechanisms for oxygen
platinum, O2 insertion into iridium hydride and alkyl bonds has
rarely been observed.7 In a recent example, O2 insertion into an
insertion with O2 are not well developed. Further, little is
known about the nature of intermediates that may result from
IrIII−H bond was proposed to occur, but the exact nature of the
reaction intermediate remains elusive.7a C−H oxidation via
initial O2 insertion and the factors that lead to facile C−O bond
direct oxygen insertion was also proposed for Cp*Ir
formation from these intermediates.
The insertion of O2 into M−H and M−R (R = alkyl) bonds
precatalysts. However, in these reactions, the source of oxygen
is water, not O2, and cerium ammonium nitrate (CAN) and
sodium periodate were utilized as external oxidants.8 Last,
dioxygen-promoted reductive elimination of R−H bonds has
been observed.9 In these examples, iridium peroxo complexes
to form hydroperoxy and alkylperoxy species, respectively, has
been reported for Pt and Pd.1a,3 Mechanistic studies with PtIV−
H complexes revealed that O2 insertion occurs via a radical
chain pathway;4 however, O2 coordination and hydrogen/
methyl abstraction were postulated as possible mechanistic
pathways for PdII−H bonds because no rate dependence on
radical traps or initiators and clean first order kinetics with
Received: December 18, 2014
Published: February 20, 2015
© 2015 American Chemical Society
3574
DOI: 10.1021/ja512905t
J. Am. Chem. Soc. 2015, 137, 3574−3584

Journal of the American Chemical Society
Article
that resulted from the reductive elimination of R−H were
isolated.
two equivalents of NaBArF resulted in the quantitative
4
formation of 2, [Cp*Ir(NHC)Me(CD2Cl2)][BArF ].11
4
Recently, we described a rare example of C−O bond
formation (oxyfunctionalization) from an Ir−methyl bond in
[Cp*Ir(NHC)(Me)L][OTf] complexes where L = pyridine,
THF, and CO; NHC = 1,3,-dimethylimidazol-2-ylidene; and
OTf = trifluoromethanesulfonate.10 From kinetic studies and
isotope labeling experiments it was suggested that the source of
oxygen was O2 and the reaction proceeded by ligand
dissociation of the L-type ligand to form a 16-electron
intermediate, followed by O2 binding and functionalization.
We hypothesized that the utilization of an L-type ligand that
dissociates from the metal freely at or below room temperature,
could potentially allow the reaction to proceed at milder
conditions and allow for the identification and isolation of
relevant intermediates.
Scheme 2. Syntheses of Complexes 1 and 2
In this investigation, [Cp*Ir(NHC)Me(Cl)] has been
synthesized and examined for the formation of methanol in
dichloromethane with O2 as the oxidant. We demonstrate that
When 2 was exposed to air, an immediate color change from
a yellow-brown to a deep blue solution was observed. This new
complex was identified as [(Cp*Ir(NHC)Me)2(μ-O)]-
NaBArF can be utilized to abstract the chloride ligand and
[(BArF )2], 3 (Scheme 3). Formally, 3 is a rare example of
4
generate [Cp*Ir(NHC)Me(CD2Cl2)][BArF ] (BArF
=
4
4
4
tetrakis[3,5-(trifluoromethyl)phenyl]borate) in situ. This com-
plex was oxidized in the presence of air to form methanol in
quantitative yields when ethanol was utilized as a proton
source. A rare example of a diiridium(IV) species, [(Cp*Ir-
Scheme 3. Synthesis of [(Cp*Ir(NHC)Me)2(μ-
O)][(BArF ) ], 3
4 2
(NHC)Me)2(μ-O)][(BArF )2] has been identified and shown
4
to be a kinetically competent intermediate for methanol
formation. Computational studies suggest that the μ-oxo bridge
facilitates electronic communication between the iridium
centers. The results presented provide insight into the nature
of a critical intermediate that results from the activation of O2
by a transition metal complex. Thus, a framework may be
provided for the design of catalysts for the oxidation of
hydrocarbons.
an IrIV bimetallic complex; however, complex 3 is diamagnetic
and can be characterized by NMR spectroscopy (see
Supporting Information). Compound 3 is stable in the solid
state; however, in solution at room temperature, 3 decomposes
to an unidentified iridium species over the course of 10 min
(see Supporting Information).
The electronic absorption spectrum of 3 was obtained (see
Supporting Information). The main features of the absorption
spectrum are a broad absorption at ∼660 nm (this peak is
primarily responsible for the blue color) as well as absorptions
at ∼450, 360, and 309 nm. Strong, broad absorptions in the
visible range (400−700 nm) are often seen in complexes
containing a bridging μ-oxo ligand.12 For example, the well-
known compound [(bipy)2(H2O)RuIII(μ-O)RuIII(H2O)-
(bipy)2]4+, which also contains two d5 metal centers, has a
λmax of ∼640 nm.12a Furthermore, Crabtree and co-workers and
others have reported a series of μ-oxo Ir complexes with
absorptions in this range.12f−h,13
RESULTS AND DISCUSSION
■
Syntheses of Complexes. Previously, the formation of
methanol in water from [Cp*Ir(NHC)(Me)L][OTf] (L =
pyridine, THF, CO) was observed at 100 °C with O2 (60 psi)
as the external oxidant (Scheme 1).10 It was suggested from
Scheme 1. Formation of Methanol from
[Cp*Ir(NHC)(Me)L][OTf] in Water
Structure of 3. X-ray quality crystals of 3 were obtained
from the slow diffusion of pentane into a concentrated
methylene chloride solution of 3 at −40 °C (Figure 1).
Complex 3 consists of two half sandwich d5 Ir(IV) centers
bridged by an oxygen atom. These types of complexes are
exceedingly rare: (xylyl)3Ir−O−Ir(xylyl)3 (xylyl = 2,6-
Me2C6H3) is the only other example of which we are
aware.14 In this molecule, the compound crystallizes on a 3-
fold axis with the bridging oxygen on the inversion center; this
results in a strictly linear (180°) Ir−O−Ir bond. By comparison,
the Ir−O−Ir bond angle in 3 is approximately 150°; this
deviation from linearity most likely results from steric crowding
kinetic studies that the rate-determining step for this reaction
involved the dissociation of the L-type ligand to form the 16-
electron intermediate [Cp*Ir(NHC)(Me)][OTf]. To perform
this reaction under milder conditions, we hypothesized that it
may be necessary to synthesize complexes that allowed for the
facile generation of this intermediate.
Complex 1, Cp*Ir(NHC)MeCl, was synthesized from the
reaction of Cp*Ir(NHC)Cl2 (NHC = 1,3,-dimethylimidazol-2-
ylidene) with 0.6 equiv of dimethyl zinc and characterized by
1H and 13C NMR spectroscopy (Scheme 2). The addition of
3575
DOI: 10.1021/ja512905t
J. Am. Chem. Soc. 2015, 137, 3574−3584

Journal of the American Chemical Society
Article
Figure 2. a) Scan of the UV−vis spectrum for the conversion of 2 to 3
vs time. Spectra were obtained every 40 s. The reaction was performed
with 2 (1.0 mM), NaBArF4 (1.2 mM) in CH2Cl2 in an uncapped UV−
vis cell open to air at 273 K b) Time course for the conversion of 2 to
3 at 660 nm. Data are fit with nonlinear least-squares fitting to an
Figure 1. Thermal ellipsoid plot of [(Cp*Ir(NHC)Me)2(μ-O)]2+, 3,
(50% ellipsoids). Hydrogen atoms and the BArF counteranions were
t
equation describing exponential growth in Abs = a + b(1-e−kobs ).
4
removed for clarity. Selected bond lengths (Å) and angles (deg): Ir2−
O1, 1.935; Ir1−O1, 1.934; Ir1−C16, 2.091; Ir2−C32, 2.105; Ir1−O1−
Ir2, 149.60; C16−Ir1−Ir2−C32, −71.97.
compared. Methanol formation from 1, over the course of 30
min followed first order kinetics and a first order rate constant,
kobs = 1.1(2) × 10−3 s−1 was obtained (Figure 3a). In
around the Ir centers arising from the bulky Cp* and NHC
ligands (see Supporting Information).
The Ir−O distance in 3 (1.935 Å) is consistent with a
bridging oxo ligand, which range from 1.902 to 2.139 Å,15 as
opposed to Ir2(μ-hydroxo)15d,16 and Ir2(μ-alkoxo)17 ligands
which range from 2.059 to 2.145 Å and 2.078 to 2.160 Å
respectively.
Formation of Methanol. Methanol formation was not
observed during the reaction outlined in Scheme 3. However,
the addition of a proton source could facilitate the release of the
methanol product. Methanol formation with 1 was thus
attempted with a variety of acids and alcohols (acetic acid,
ethanol, and phenol). The use of ethanol as a proton source
resulted in quantitative yields of methanol in 1 h at room
temperature. No methanol product was observed with acetic
acid, and methanol yields were significantly attenuated (∼15%)
when phenol was used as a proton source.
Quantitative methanol yields were observed in 30 min when
the reaction conditions were optimized further and the reaction
was performed open to air. Thus, the optimized conditions for
methanol formation are depicted in Scheme 4.
Figure 3. a) Plot of methanol formation from 1 vs time. Conditions: 1
(15 mmol, 7.1 mg); NaBArF (30 mmol, 26.6 mg); C2H5OH (75
4
mmol, 4.4 μL) in 0.5 mL CD2Cl2 open to air at room temperature.
Reactions were monitored by 1H NMR spectroscopy with 1,3,5-
trimethoxybenzene as an internal standard; b) Plot of methanol
formation from 3 vs time. Conditions: 3 (7.5 mmol, 19.7 mg);
C2H5OH (75 mmol, 4.4 μL) in 0.5 mL CD2Cl2 open to air at room
temperature. Reactions were monitored by 1H NMR spectroscopy
with 1,3,5-trimethoxybenzene as an internal standard. Data are fit with
nonlinear least-squares fitting to an equation describing exponential
t
growth in [Methanol] = a + b(1-e−kobs ).
Scheme 4. Optimized Conditions for the Formation of
Methanol from 1
comparison, methanol formation from 3 resulted in a first order
rate constant, kobs = 1.7(2) × 10−3 s−1 (Figure 3b). The kobs
values suggest that 3 is a kinetically competent intermediate in
this reaction.
The intermediacy of 3 was investigated further by obtaining
activation parameters for the formation of methanol from both
1 and 3 according to the Eyring equation. Reactions were
performed at 0, 10, and 25 °C (Figure 4). Entropically, the
formation of methanol from both 1 and 3 is unfavorable (ΔS‡ =
−46(11) and −37(9) eu, respectively). The enthalpic
component of the activation energy is small and positive
(ΔH‡ = 8(2) and 10(3) kcal/mol respectively), which is
consistent with the observed room temperature reactivity for
both complexes. When combined with the observed rate
constants discussed previously, the relative activation energies
provide additional evidence for the intermediacy of complex 3
in the formation of methanol from 1.
Studies with Radical Traps. To determine if the formation
of methanol occurs via a radical pathway, reactions were
performed in the presence of radical traps. Reactions with 1
were performed under the conditions outlined in Scheme 4
with the radical trap TEMPO (2,2,6,6′-tetramethylpiperadin-1-
Kinetic Studies. The transformation of 2 to 3 was
monitored by UV−vis spectroscopy (Figure 2). The presence
of isosbestic points at 407 and 540 nm suggests that 2 is
converted directly to 3 without the appreciable accumulation of
any intermediates (Figure 2a). Further, when the absorbance at
660 nm was monitored over time an exponential growth for at
least five half-lives was observed suggesting that the reaction
exhibits a first order dependence on [2] (Figure 2b).
The addition of five equivalents of ethanol to 3 led to near
quantitative formation of methanol (>90%) in 30 min as
1
observed by H NMR spectroscopy. To further confirm that 3
is an intermediate in the formation of methanol from 1, the
kinetics for the formation of methanol from 1 and 3 were
3576
DOI: 10.1021/ja512905t
J. Am. Chem. Soc. 2015, 137, 3574−3584

Journal of the American Chemical Society
Article
Isotope Labeling Experiments and Identification of
the Iridium Product. Recent reports involving complexes
with Cp*IrIII frameworks have suggested that the ancillary
ligands can undergo oxidative decomposition under strongly
oxidizing conditions.8c,e,12f−h,18 To determine whether the
ligand framework was maintained during the formation of
methanol, the origin of the methyl fragment in the methanol
product was determined as well as the identity of the iridium
product.
Isotope labeling studies were performed in which the methyl
group bound to the iridium was labeled with deuterium. The
complex 1-CD3 was synthesized from Cp*Ir(NHC)Cl2 and
CD3MgBr. The presence of an Ir−CD3 bond in 1-CD3 was
Figure 4. a) Eyring plot for the formation of methanol from 1 at 273,
283, and 298 K. Conditions: 1 (15 mmol, 7.1 mg); NaBArF (30
4
mmol, 26.6 mg); C2H5OH (75 mmol, 4.4 μL) in 0.5 mL CD2Cl2 open
1
to air at room temperature. Reactions were monitored by H NMR
2
confirmed by H NMR spectroscopy. Exposure of 1-CD3 to
spectroscopy with 1,3,5-trimethoxybenzene as an internal standard; b)
Eyring plot for the formation of methanol from 3 at 273, 283, and 298
K. Conditions: 3 (7.5 mmol, 19.7 mg); C2H5OH (75 mmol, 4.4 μL) in
0.5 mL CD2Cl2 open to air at room temperature. Reactions were
monitored by 1H NMR spectroscopy with 1,3,5-trimethoxybenzene as
an internal standard.
NaBArF4 and air resulted in the formation of 3-CD3, confirmed
by 1H, 2H, and 13C NMR spectroscopy (see Supporting
Information).
The reaction of 1-CD3 with NaBArF and ethanol in the
4
presence of air (Scheme 6) resulted in the formation of
yl)oxy. When 1 equiv of TEMPO was utilized, no methanol was
observed after 1 h. Methanol formation from 1 in the presence
of 0.1 equiv of TEMPO reached completion at ∼40% yield in 2
h. The data suggest that a radical pathway is operative for
methanol formation from 1. Similar reactivity was observed for
3. Methanol yields were inhibited when 1 equiv of TEMPO was
added to the reaction (Scheme 5). 1,4-Cyclohexadiene was also
Scheme 6. Isotope Labeling Studies
Scheme 5. Summary of Radical Trap Experiments
CD3OH, as observed by 2H NMR spectroscopy (see
Supporting Information). These data are consistent with the
direct oxidation of the Ir−Me bond to form methanol.
To determine whether the Cp* ligand was oxidized during
the course of the reaction, the iridium product was identified.
The reaction of 3 with 5 equiv of ethanol in air resulted in the
identification of [(Cp*Ir(NHC)Cl)2][(BArF )2], 4, in near
4
quantitative yield (89(3)%) by 1H NMR spectroscopy (Scheme
7).19 X-ray quality crystals were obtained and are in agreement
Scheme 7. Isolation of 4
examined as a radical trap for methanol formation from 3.
When 0.5 equiv of 1,4-cyclohexadiene was added, the observed
rate constant for methanol formation was reduced from kobs
=
1.7(2) × 10−3 s−1 to kobs = 7(2) × 10−5 s−1. This result, along
with the reactions described previously in the presence of
TEMPO, suggest that a radical pathway is operative for
methanol formation.
In contrast, the formation of 3 from 2 was not affected by
TEMPO. The same observed rate constant kobs = 1.2 × 10−2 s−1
was observed when 1 equiv of TEMPO was added (see
Supporting Information). Similar results were obtained when
1,4-cyclohexadiene was utilized as a radical trap. These data
suggest that the formation of 3 is not inhibited by the radical
trap TEMPO and imply that the formation of 3 from 2 occurs
via a nonradical pathway. The results of the experiments with
radical traps are summarized in Scheme 5.
with the spectroscopic assignment of 4 as the iridium product
for this reaction (see Supporting Information). Importantly,
both the Cp* and NHC ligands remain intact in 4 after the
oxidative conditions employed.
Role of Ethanol as a Proton Source. The role of ethanol
as an additive was explored. Ethanol could potentially act as a
proton source, resulting in an Ir−OEt complex, 6, from a
proposed Ir−OMe intermediate, 5. When 3 was treated with 5
3577
DOI: 10.1021/ja512905t
J. Am. Chem. Soc. 2015, 137, 3574−3584

Journal of the American Chemical Society
Article
equiv of ethanol in air as described in Scheme 7, acetaldehyde
was observed by H NMR spectroscopy in quantitative yields,
Scheme 9. Proposed Mechanism for the Formation of 3 from
2.
a
1
suggesting an alkoxide exchange followed by β-hydride
elimination (Scheme 8).
Scheme 8. Proposed β-hydride Elimination from an Ir−OMe
Complex
The disappearance of 3 in the presence of 5 equiv of ethanol
could also be monitored by UV−vis spectroscopy at 660 nm,
where 3 exhibits a strong absorbance but the product, 4, does
not. The disappearance of 3 at 660 nm follows clean first-order
kinetics for at least five half-lives (see Supporting Information).
Importantly, the observed rate constant, kobs = 5.2 × 10−3 s−1, is
independent of the ethanol concentration. This suggests that
alkoxide exchange and β-hydride elimination occurs after the
rate-determining step (see Supporting Information).
Computational Studies. Computational studies were
performed on some of the important species described above.
For DFT (PBE0 functional) calculations, to ameliorate
problems associated with calculating ionic complexes in the
gas phase, calculations were performed with implicit modeling
of the reaction solvent, CH2Cl2 (ε = 8.93), as modeled by the
SMD solvation method. Addition of dispersion effects was
found to be important, especially for modeling the thermody-
namics of the formation of bimetallic intermediates, so these
were included. Further details are given below in Computa-
tional Methods.
Activation of O2. To understand the mechanism for the
activation of O2 by 2, DFT calculations were performed. The
proposed mechanism for the conversion for 2 to 3 is depicted
in Scheme 9.
Complex 2 initially reacts with O2 to generate the iridium
dioxygen adduct, 8. Dioxygen binding to 2 is calculated to be
endergonic (ΔG° = 15.6 kcal/mol on the singlet surface and
17.4 kcal/mol on the triplet surface). Our experimental
observations suggest that this step must be rate-determining
because under constant O2 pressure, the rate law for the
conversion of 2 to 3 is given by (Figure 2):
a
Calculations were performed at the DFT(PBE0) level with the
solvent free 16-electron intermediate, cation, [Cp*Ir(NHC)Me+], 2′.
See Computational Methods for details. Reported energies are in kcal/
mol. Singlet energies are in red font.
structure of 10 is described below. Complex 10 subsequently
reacts with another molecule of complex 2 to generate 3. Thus,
for the overall chemical reaction, four molecules of 2 are
required for each dioxygen molecule.
As noted below, the electronic structure for complexes such
as 3 and 10 are not accurately described by single determinant
DFT methods. Further, several of the transition states have not
been located. As a result, the proposed mechanism in Scheme 9
is tentative at this time. However, the mechanism highlights the
existence of potentially important intermediates such as the μ-
peroxo complex 9 that may precede the formation of iridium
oxo/oxyl intermediates.
Electronic Structure of 3. Insights into the unique
reactivity of 3 were obtained by an analysis of the bonding in
this complex. As noted earlier, 3 was assigned on the basis of its
spectroscopic signature as an unusual IrIV−O−IrIV complex
whereby two d5 Ir(IV) centers are strongly coupled by the
bridging oxo ligand. In a previous report on (xylyl)3Ir−O−
Ir(xylyl)3, it was suggested from DFT studies that the
diamagnetic nature of the complex is a result of the interaction
of the dz2 orbitals on both metals with the s and p orbitals on
the bridging oxygen.14a This was evident in the frontier
molecular orbitals, which revealed a significant HOMO−
LUMO gap (2.5 eV) and was proposed to account for the
diamagnetism of the complex. The structure of 3 can be
similarly described by the formulation (X3)IrIV−O−IrIV(X3),
although the bonding pattern may differ due to the nonlinear
arrangement about the μ-oxo.
d[3]
dt
= kobs[2];
kobs = k[O2]
Furthermore, the presence of isosbestic points at 407 and
540 nm suggests that intermediates do not accumulate in
appreciable concentrations for this reaction.
Complex 8 then reacts with another molecule of 2 to
generate the μ-peroxo complex 9. This reaction is endergonic
on the singlet surface (ΔG° = 32.7 kcal/mol) but much less so
on the triplet surface (ΔG° = 22.3 kcal/mol). Cleavage of the
O−O bond in 9 results in two oxo iridium complexes 10 (1ΔG°
= 18.8 kcal/mol, 3ΔG° = 8.7 kcal/mol)). The electronic
3578
DOI: 10.1021/ja512905t
J. Am. Chem. Soc. 2015, 137, 3574−3584

Journal of the American Chemical Society
Article
At the DFT level of theory, 3 is calculated to have a triplet
ground state, but a singlet state is calculated to be only 0.7 kcal/
mol higher in energy. Note that several conformations
generated by twisting of the C−Ir−Ir−C dihedral were
evaluated for singlet and triplet 3, and the results below are
for the lowest energy conformations computed. Triplet 3 has
average computed Ir−Me bonds of 2.10 Å, Ir−O bonds of 1.93
Å, an Ir−O−Ir angle of 160°, and a C−Ir−Ir−C dihedral angle
of −67°; the singlet analog has average Ir−Me bonds of 2.09 Å,
Ir−O bonds of 1.91 Å, an Ir−O−Ir angle of 150°, and a C−Ir−
Ir−C dihedral angle of −72°. The DFT computed metric values
are thus in excellent agreement with the solid-state structure of
3 (see Figure 2), particularly for the singlet with respect to the
Ir−O−Ir bond angle and the C−Ir−Ir−C dihedral angle.
Although the structural differences between singlet and
triplet models of 3 are not large, the differences in the
calculated UV−vis spectra are larger (see Supporting
Information). The calculated UV−vis spectrum further
supports the contention that the singlet structure is the true
minimum. Furthermore, these computational results are
consistent with the diamagnetic nature of complex 3 as
observed by its well-defined NMR spectrum.
significant multiconfiguration character for the diamagnetic,
singlet μ-oxo complex 3.
The complex (xylyl)3Ir−O−Ir(xylyl)3 was synthesized from
the partial oxygen atom transfer from the IrV-oxo species,
(xylyl)3IrVO to (xylyl)3Ir.14a The formation of 3 can be
envisioned from a similar reaction between [Cp*Ir(NHC)Me]+
and the putative metal oxo [Cp*IrO(NHC)Me]+ (Scheme
10). Similar to the bonding model formulated by Wolczanski
Scheme 10. Proposed Mechanism for the Formation of 3
from a Putative IrV oxo/oxyl Intermediate
Although calculations at the DFT level of theory suggest that
the singlet and triplet spin states lie within 1 kcal/mol, with the
triplet marginally lower, it was expected that coupling between
the formally d5 Ir(IV) centers may be better described with
MCSCF (multiconfiguration SCF) than single-determinant
DFT methods, so MCSCF computations on a slightly simpler
model (where methyl substituents on the Cp* and NHC
ligands were replaced with hydrogen, 3′) were performed. The
MCSCF computations (10-orbital, 10-electron CASSCF
(complete active space SCF) show the singlet state to be
significantly lower in energy by ∼17 kcal/mol than triplet 3′,
consistent with a diamagnetic ground state as inferred from the
spectroscopy of 3.20
and co-workers21 for oxygen-atom transfer to/from nitrosyl
complexes, orbital symmetry requirements yield a nonlinear
arrangement for the oxygen transfer and σ/π mixing, as can be
seen in the frontier orbitals that are orthogonal to those
depicted in Figure 5 (see Figure 6).21
In addition, the important MCSCF natural orbitals for
complex 3′ are informative with respect to the nature of the
bonding across the Ir−O−Ir moiety and are depicted in Figure
5. These frontier orbitals show significant delocalization among
Figure 6. MCSCF frontier orbitals (top) and their cartoon
representation (bottom) of 3′ showing σ/π mixing along the IrOIr
core. These orbitals lie in the plane perpendicular to that plotted in
Figure 5
Figure 5. (a) Bonding, (b) nonbonding, and (c) antibonding Ir dπ −
O pπ − Ir dπ natural orbitals of 3′ obtained from MCSCF calculations.
Natural orbital occupation numbers are 1.93, 1.87, and 0.21 e-,
respectively.
Electronic Structure of Oxo Iridium Intermediates. It
has been suggested in previous studies that d5 IrIV−O−IrIV d5
species such as 3 may exist in equilibrium with putative Ir oxo/
oxyl intermediates.14a In addition, Cp*Ir(V) oxo species have
been postulated recently in catalytic systems for the
hydroxylation of hydrocarbons and water oxidation.8d,f−h
Although the Cp*Ir oxo species were not isolated, under-
standing the electronic structure of proposed Cp*Ir oxo
intermediates is important for the design of new catalysts.
Singlet, [Cp*IrVO(NHC)Me]+, and triplet, [Cp*IrIV
O(NHC)Me]+, species were calculated. As shown in Scheme
11, the triplet species is 10.1 kcal/mol more stable than the
corresponding singlet species by DFT.
the three atoms of the Ir−O−Ir core. The orbital arrangement
is quite different from the orbital model put forth for the linear
(xylyl)3Ir−O−Ir(xylyl)3 by Brown et al.,14a likely due in part to
the nonlinearity about the bridging oxo of 3. The frontier
orbitals indicate strong coupling between iridium centers,
supporting the formulation of 3 as IrIV−O−IrIV. Indeed, the
frontier orbitals are reminiscent of bonding in a 3-center, 4-
electron allyl anion with bonding (πb, Figure 5a), nonbonding
(πn, Figure 5b) and antibonding (πa, Figure 5c) components.
The significant departure of natural orbital occupation numbers
from ideal values of 0 and 2 e− (πb1.93πn1.87πa0.21) highlights
3579
DOI: 10.1021/ja512905t
J. Am. Chem. Soc. 2015, 137, 3574−3584

Journal of the American Chemical Society
Article
1
Scheme 11. Relative Energies of [Cp*IrVO(NHC)Me]+
3
and [Cp*IrIVO(NHC)Me]+
3
The location of spin density for [Cp*IrIVO(NHC)Me]+
was calculated with unrestricted DFT methods to understand
the nature of the Ir−O multiple bond in this species. The spin
density is localized primarily on the Ir (0.76 e-) and oxygen
(1.02 e−) atoms with a little spin density on one carbon atom of
the Cp* ligand (0.13 e−) (see Supporting Information).
The Ir−O bond length in [Cp*IrIVO(NHC)Me+] was
calculated to be 1.800 Å. In contrast, the Ir−O bond length in
[Cp*IrV(O)(NHC)Me+] was calculated to be 1.794 Å.
These data suggest that in both species, there is Ir−O multiple
bonding because the bond lengths in both species are shorter
than typical Ir−O single bonds (∼2.11(8) Å).22
Given the multireference character of 3′ and the delocaliza-
tion of the spin density in the plotted DFT orbitals, an MCSCF
treatment of the singlet and triplet states of the monometallic
oxo/oxyl complex was deemed prudent. A similar truncation of
the methyl groups was performed giving [CpIr(O)(NHC′)-
Me]+. Interestingly, a larger active space was necessary (14-
orbital, 14-electron CASSCF) for the monomeric Ir complexes
as compared with that for 3′. The calculations also predict a
triplet ground state for the monometallic species, but the
CASSCF calculations predict only a 4.3 kcal/mol difference
between the singlet and triplet states. Inclusion of second-order
perturbation at this level of theory (MCQDPT2) showed the
triplet to be the ground state by 5.6 kcal/mol.
Figure 8. (a) σ-bonding and (b,c) π-bonding and (d) σ-antibonding
3
and (e,f) π-antibonding natural orbitals of [CpIrIVO(NHC’)Me]+.
Natural orbital occupation numbers are 1.94, 1.95, 1.95, 0.07, 1.04, and
1.04 e−, respectively.
The triplet state has two Ir−O π-bonding natural orbitals,
one primarily Ird−Op (Figure 8b), and the second, Ird−Op with
contributions from Ir−Me and Ir−NHC bonding orbitals
(Figure 8c). The Ir−O σ-bonding orbital now also has
contributions from the methyl and NHC ligands. The two
unpaired electrons occupy two π-antibonding natural orbitals,
yielding a pair of orthogonal 3-electron/2-center bonds (Figure
8e and 8f). One of the π-antibonding natural orbitals also has a
small contribution from the Cp ring, which corresponds to the
spin density calculated by the DFT methods (see Supporting
Information).
Proposed Mechanism for Methanol Formation. A
mechanism consistent with all of the data is shown in Scheme
12). Chloride abstraction from 1 by NaBArF leads to 2.
Important natural orbitals for Ir−O bonding in the singlet
(Figure 7) and triplet states (Figure 8) are depicted. For the
singlet, there is one σ-bonding orbital and one π-bonding
orbital, as would be expected in a typical double bond. There is
also significant deviation of the natural orbital occupancy for
t h e s i n g l e t f r o m s i n g l e - d e t e r m i n a n t v a l u e s
(σ1.94π1.86π*0.15σ*0.07) indicating multireference character.
4
Exposure of this complex to O2 results in the formation of 10,
which subsequently reacts with another equivalent of 2 to form
3. Kinetic studies suggest that for the activation of dioxygen, the
initial reaction of 2 with O2 is rate-determining because the rate
law for this reaction shows a first-order dependence on 2 and
the presence of isosbestic points in the scan of the UV−vis
spectrum suggests that intermediates do not accumulate in
appreciable concentrations. The activation of O2 by 2 (1st step
Scheme 9) is nonradical in nature and should not be inhibited
1
by radical traps. Indeed, the H NMR spectrum of 2 was
unaffected by the presence of 1 equiv of TEMPO or 1,4-
cyclohexadiene. Further, the same observed rate constant (kobs
= 1.2 × 10−2 s−1) for the formation of 3 from 2 in the presence
and absence of TEMPO was observed.
Methanol formation occurs after the formation of 3. As
suggested by previous reactions with (xylyl)3Ir−O−Ir(xylyl)3
by Brown and co-workers, 3, 2, and 10 are expected to exist in
equilibrium. The formation of the Ir−OMe intermediate, 5,
results from migration of the methyl ligand to the oxo ligand in
10. Treatment of this species with ethanol results in alkoxide
exchange to generate the Ir−OEt species, 6.23 β-hydride
elimination results in the formation of the hydride, [Cp*Ir-
(NHC)H][BArF ], 7 and acetaldehyde. This complex under-
4
goes metathesis with the methylene chloride solvent to form
4.24
Figure 7. (a) σ-Bonding and (b) π-bonding natural orbitals and (c) σ-
antibonding and (d) π-antibonding natural orbitals of [CpIrV
O(NHC’)Me]+. Natural orbital occupation numbers are 1.94, 1.86,
0.07, and 0.15 e−, respectively.
The mechanism presented above includes the observed
iridium intermediate 3. Kinetic studies suggest that 3 is an
important intermediate in this reaction. The presence of 7 is
3580
DOI: 10.1021/ja512905t
J. Am. Chem. Soc. 2015, 137, 3574−3584

Journal of the American Chemical Society
Article
Scheme 12. Proposed Mechanism for Methanol Formation
1
implied by the quantitative observation of acetaldehyde by H
Complex 3 likely results from an oxygen atom transfer
reaction of 2 with a Cp*Ir oxo intermediate. The electronic
structure of this monometallic oxo intermediate was also
investigated computationally. MCSCF calculations suggest that
the Ir-oxo intermediate may be described by two species that
are close in energy and therefore accessible under reaction
conditions: a singlet IrV−oxo intermediate and a triplet IrIV oxyl
intermediate. Both species have formal Ir−O bond orders of 2;
however, their electronic structures are quite different. Bonding
in the IrV-oxo is reminiscent of the VB description of
NMR spectroscopy and the isolation and characterization of 4
in quantitative yields. Isotope labeling experiments confirm that
the Ir-Me bond is oxidized to form the methanol product.
Importantly, the isolation of 4 suggests that the Cp* and NHC
ligands are not oxidized significantly under reaction conditions.
CONCLUSIONS
■
Methanol formation from Cp*IrIII(NHC)Me complexes was
shown to occur quantitatively at room temperature with air
(O2) as the oxidant and ethanol as a proton source. A rare
example of a diiridium bimetallic complex, 3, has been isolated
1
formaldehyde by Harding and Goddard and the Δg state of
O2.25 The triplet IrIV−oxyl is more akin to the double bond in
the triplet ground state (3∑g−) of dioxygen.26 As one might
expect, these differences in electronic structure imply different
manifolds of reactivity.
and characterized by H and 13C NMR spectroscopy, UV−vis
1
spectroscopy, X-ray crystallography, and both DFT and
MCSCF computations. Complex 3 contains a IrIV−O−IrIV
core whereby two d5 Ir(IV) centers are bridged by an oxo
ligand. Insights into the electronic structure of this intermediate
were provided by computational studies, which suggest that the
μ-oxo bridge facilitates electronic communication between the
iridium centers. Evidence for 3-center, 4-electron bonding
reminiscent of an allyl anion was obtained from MCSCF
calculations for the μ-oxo intermediate, which experiments
show to be a kinetically competent intermediate in the
production of methanol by the reaction of dioxygen with
these iridium-methyl complexes. Further, these studies suggest
that similar to the bonding model formulated by Wolczanski
and co-workers,21 orbital symmetry requirements yield a
nonlinear arrangement for the oxygen transfer and σ/π mixing.
Obviously, obtaining the desired nonlinear M−O−M arrange-
ment for oxygen atom transfer can be thwarted by utilizing
ligands that are less sterically demanding, suggesting that
identification of ligand sets that permit nonlinear M−O−M
arrangements is a key design element in this oxyfunctionaliza-
tion chemistry and hence related alkane oxidation catalysis.
Ir-oxo/oxyl complexes have emerged from this research, as
well as recent papers as important intermediates in C−O bond
formation, C−H bond activation, and water splitting. To our
knowledge, apart from Wilkinson’s IrMes3O,27 such entities
have eluded crystallographic characterization, and more
research to elucidate their reactivity with different organo-
metallic and coordination chemistry supporting ligands is
clearly needed. One can infer from the present work that
careful attention to their electronic structure will need to go
beyond single-determinant DFT methods.
The formation of 3 from 2 is not inhibited by the radical
traps TEMPO and 1,4-cyclohexadiene; however, methanol
formation from 3 is inhibited by TEMPO. Isotope labeling
studies confirmed the origin of the methyl group in the
methanol product is the iridium−methyl bond. Isolation of the
diiridium-containing product 4 in high yields at the end of the
reaction shows that the Cp* and NHC ligands remain bound to
the iridium and are not significantly degraded under reaction
conditions. Complexes with the Cp*Ir(NHC) motif may be
3581
DOI: 10.1021/ja512905t
J. Am. Chem. Soc. 2015, 137, 3574−3584

Journal of the American Chemical Society
Article
further developed into a catalytic cycle for the partial oxidation
of hydrocarbons. More importantly, the identification of 3 as an
intermediate in formation of methanol from dioxygen can
provide insight into the activation of transition metal alkyl
complexes with O2 and the subsequent functionalization that
results from C−O bond formation. Studies are currently
ongoing to understand how the unique electronic properties of
3 influence the oxyfunctionalization of metal−carbon bonds.
Synthesis of 3-CD3. In the glovebox, 1-CD3 (25 mg, 0.052
mmol) and NaBArF (94 mg, 0.11 mmol) were added to an
4
∼20 mL scintillation vial with 5 mL of CH2Cl2. The solution
was allowed to stand for 30 min. Ten milliliters of pentane was
then added to the vial and the sample was sealed and removed
from the glovebox. The vial was then opened to the air,
immediately sealed, and left in the freezer at −30 °C for at least
5 h. The blue crystalline product was removed from the vial,
washed with excess pentane, and dried in vacuo. The remaining
crystals were collected as [(Cp*Ir(NHC)CD3)2(μ-O)]-
EXPERIMENTAL SECTION
■
[(BArF )2] (85 mg, 62% yield). 1H NMR (400 MHz,
4
General Considerations. Reagents were purchased from
CD2Cl2, δ): 2.10 (s, 15H, Cp*), 2.95 (s, 3H, N−Me), 3.62
(s, 3H, N−Me), 7.91 (d, JHH = 4 Hz, 1H, CH), 7.88 (d, JHH = 4
Hz, 1H, CH). 13C NMR (100 MHz, CD2Cl2, δ): 135, 126.2,
123.4, 117.7, 110.8, 8.3.
commercial sources and used as received. H and 13C NMR
1
spectra were recorded on a Varian Mercury 400 MHz or a
Varian Mercury 300 MHz spectrometer at room temperature.
1H and 13C NMR chemical shifts are listed in parts per million
(ppm) and are referenced to residual protons or carbons of the
deuterated solvents, respectively. Elemental analyses were
performed by Atlantic Microlabs, Inc. X-ray crystallography
was performed at the X-ray Structural Facility of North
Carolina State University by Dr. Roger Sommer.
General Procedure for Methanol Formation in CD2Cl2
from 1. In an ∼3 mL Schlenk tube, 1 (7.1 mg, 0.015 mmol)
was added to a solution containing 0.5 mL of 0.03 M 1,3,5-
trimethoxybenzene in CD2Cl2 and 0.15 M ethanol. NaBArF
4
(26.6 mg, 0.03 mmol) was added to initiate the reaction. The
solution was allowed to stir open to air for the allotted amount
of time. The reaction was then removed from the stir plate, and
the crude reaction mixture was then transferred to an NMR
Cp*Ir(NHC)MeCl, 1. In a 25 mL Schlenk flask, Cp*Ir-
(NHC)Cl2 (NHC = 1,3,-dimethylimidazol-2-ylidene) (250 mg,
0.505 mmol) was allowed to stir in THF for 10 min. This
solution was frozen in a liquid N2 bath under a stream of N2.
Zn(Me)2 (252 uL, 1.2 M in toluene) was added to the flask
under a stream of N2. The flask was sealed, and the solution was
allowed to stir for 20 min. Solvent was removed in vacuo to
leave a yellow oil. Subsequent addition of pentane afforded a
yellow powder of Cp*Ir(NHC)MeCl (184 mg, 79% yield). 1H
NMR (400 MHz, CD2Cl2, δ): 0.81 (s, 3H, Ir−Me), 1.61 (s,
15H, Cp*), 3.53 (bs, 3H, N−Me), 3.94 (bs, 3H, N−Me), 6.98
(bs, 2H, CH). 13C NMR (100 MHz, CD2Cl2, δ): 161.2, 122.3,
87.9, 8.8, −16.2.
1
tube for H NMR analysis.
General Procedures for Methanol Formation in
CD2Cl2 from 3. In an ∼3 mL Schlenk tube, 3 (7.1 mg,
0.015 mmol) was added to a solution that contained 0.5 mL of
0.03 M 1,3,5-trimethoxybenzene in CD2Cl2 and 0.15 M
ethanol. The solution was allowed to stir open to air for the
allotted amount of time. The reaction was then removed from
the stir plate, and the crude reaction mixture was then
1
transferred to an NMR tube for H NMR analysis.
Kinetic Studies. General Conditions. An ∼3 mL Schlenk
flask was charged with the appropriate amount of 1 or 3 and a
stir bar. A standard solution (0.5 mL) of CD2Cl2 with 0.03 M
1,3,5-trimethoxybenzene and 0.15 M ethanol was added to the
[(Cp*Ir(NHC)Me)2(μ-O)][(BArF ) ], 3. In the glovebox, 1
4 2
(25 mg, 0.053 mmol) and NaBArF (94 mg, 0.11 mmol) were
4
added to an ∼20 mL scintillation vial with 5 mL of CH2Cl2.
The solution was allowed to stand for 30 min. Ten milliliters of
pentane was then added to the vial, and the sample was sealed
and removed from the glovebox. The vial was then opened to
the air, immediately sealed, and left in the freezer at −30 °C for
at least 5 h. The blue crystalline product was removed from the
vial, washed with excess pentane, and dried in vacuo. The
remaining crystals were collected as [(Cp*Ir(NHC)Me)2(μ-
flask. NaBArF (26.6 mg, 0.03 mmol) was added to initiate the
4
reaction with 1. The solution was allowed to stir open to air for
the allotted amount of time. The reaction was then removed
from the stir plate, and the crude reaction mixture was then
1
transferred to an NMR tube for H NMR analysis. Each data
point is the average of at least two runs. Error bars represent
the standard deviation.
O)][(BArF )2] (98 mg, 71% yield). 1H NMR (400 MHz,
Formation of 1. In an NMR tube, 1 (7.1 mg, 0.015 mmol)
was added to 0.5 mL of a 0.03 M solution of 1,3,5-
trimethoxybenzene in CD2Cl2 in the glovebox. The reaction
was allowed to stand for 30 min before it was removed from the
glovebox. The NMR tube was then exposed to air and
immediately placed in the instrument. The sample was analyzed
utilizing an array to acquire spectra at continual points over the
course of an hour. The yields for 1, 3, and iridium total (1 + 3)
were calculated vs the 1,3,5-trimethoxybenzene internal stand-
ard.
Formation of 1 in the Presence of TEMPO and 1,4-
Cyclohexadiene. In an NMR tube 1 (7.1 mg, 0.015 mmol)
was added to a 0.5 mL solution of 0.03 M 1,3,5-
trimethoxybenzene and 0.03 M TEMPO in CD2Cl2 in the
glovebox. The reaction was allowed to stand for 30 min before
it was removed from the glovebox. The NMR tube was then
exposed to air and immediately placed in the instrument. The
sample was analyzed utilizing an array to acquire spectra at
continual points over the course of an hour. The yields for 1, 3,
4
CD2Cl2, δ): 2.11 (s, 15H, Cp*), 3.13 (s, 3H, Ir−Me), 2.96 (s,
3H, N−Me), 3.63 (s, 3H, N−Me), 7.91 (d, JHH = 4 Hz, 1H,
CH), 7.88 (d, JHH = 4 Hz, 1H, CH). 13C NMR (100 MHz,
CD2Cl2, δ): 135, 126.2, 123.4, 117.7, 110.8, 8.3. Anal. Calcd for
C96.45H76.90Cl0.9B2F48Ir2N4O: C, 43.59; H, 2.92; N, 2.11. Found:
C, 43.20; H, 2.90; N, 2.15.
Cp*Ir(NHC)CD3Cl, 1-CD3. In a Schlenk tube, CD3MgI
(0.535 mL, 0.535 mmol, 1 M in ether) was added to a 10 mL
THF solution of Cp*Ir(NHC)Cl2 (NHC = 1,3,-dimethylimi-
dazol-2-ylidene) (250 mg, 0.505 mmol). The solution was
allowed to stir for 1 h. The reaction was quenched with water,
and the product was extracted with methylene chloride. Solvent
was removed in vacuo to leave a yellow oil. Subsequent
addition of pentane afforded a yellow powder of Cp*Ir(NHC)-
CD3Cl (165 mg, 69% yield). 1H NMR (400 MHz, CD2Cl2, δ):
1.72 (s, 15H, Cp*), 3.61 (s, 3H, N−Me), 3.89 (s, 3H, N−Me),
7.00 (bs, 1H, CH), 6.84 (bs, 1H, CH). 13C NMR (100 MHz,
CD2Cl2, δ): 158.7, 122.2, 88.6, 9.3.
3582
DOI: 10.1021/ja512905t
J. Am. Chem. Soc. 2015, 137, 3574−3584

Journal of the American Chemical Society
Article
(h) Wu, W.; Jiang, H. Acc. Chem. Res. 2012, 45 (10), 1736−1748.
(i) Olah, G. A. Angew. Chem., Int. Ed. 2005, 44, 2636−2639.
and iridium total (1 + 3) were calculated vs the 1,3,5-
trimethoxybenzene internal standard.
(2) (a) Bergman, R. G. Nature 2007, 446 (7134), 391−3.
(b) Goldman, A. S.; Goldberg, K. I. ACS Symp. Ser. 2004, 885, 1−43.
(3) (a) Scheuermann, M. L.; Boyce, D. W.; Grice, K. A.; Kaminsky,
W.; Stoll, S.; Tolman, W. B.; Swang, O.; Goldberg, K. I. Angew. Chem.,
Int. Ed. 2014, 53 (25), 6492−6495. (b) Prantner, J. D.; Kaminsky, W.;
Goldberg, K. I. Organometallics 2014, 33 (13), 3227−3230.
Eyring Plots. To obtain Eyring plots, reactions were run
under the general conditions outlined above at 0, 10, and 25
°C. At 0 °C, the temperature was maintained by running the
reactions in an ice bath. For the data at 10 °C a 1,4-dioxane/
liquid N2 bath was used to maintain a constant temperature. At
the lower temperature, the standard solutions were cooled to
the appropriate temperature before the reaction was initiated.
Computational Methods. DFT calculations were per-
formed using the Gaussian 09 suite of programs.28 Structures
were optimized with the PBE029 hybrid/exchange functional.
For iridium, the Stuttgart30 pseudopotential with a triple-ζ level
valence and f polarization function (exponent = 0.685). For all
other elements, a 6-311G** basis set was used. Optimizations
were performed with implicit solvation of methylene chloride
(ε = 8.93) by the SMD31 method, and dispersion effects were
included using Grimme’s D3 dispersion correction with
Becke−Johnson damping.32 Structures were identified as
minimum or transition states by the presence of zero or one
imaginary frequency in the energy Hessian.
́
(c) Petersen, A. R.; Taylor, R. A.; Vicente-Hernandez, I.; Mallender,
P. R.; Olley, H.; White, A. J. P.; Britovsek, G. J. P. J. Am. Chem. Soc.
2014, 136 (40), 14089−14099. (d) Oloo, W. N.; Zavalij, P. Y.;
Vedernikov, A. N. Organometallics 2013, 32 (19), 5601−5614.
(e) Chuang, G. J.; Wang, W.; Lee, E.; Ritter, T. J. Am. Chem. Soc.
2011, 133 (6), 1760−1762. (f) Denney, M. C.; Smythe, N. A.; Cetto,
K. L.; Kemp, R. A.; Goldberg, K. I. J. Am. Chem. Soc. 2006, 128 (8),
2508−2509. (g) Dura-Vila, V.; Mingos, D. M. P.; Vilar, R.; White, A. J.
P.; Williams, D. J. Chem. Commun. 2000, 16, 1525−1526. (h) Jintoku,
T.; Nishimura, K.; Takaki, K.; Fujiwara, Y. Chem. Lett. 1990, No. 9,
1687−8. (i) Jintoku, T.; Takaki, K.; Fujiwara, Y.; Fuchita, Y.; Hiraki, K.
Bull. Chem. Soc. Jpn. 1990, 63 (2), 438−41. (j) Jintoku, T.; Taniguchi,
H.; Fujiwara, Y. Chem. Lett. 1987, No. 9, 1865−8. (k) Konnick, M. M.;
Guzei, I. A.; Stahl, S. S. J. Am. Chem. Soc. 2004, 126 (33), 10212−
10213. (l) Landis, C. R.; Morales, C. M.; Stahl, S. S. J. Am. Chem. Soc.
2004, 126 (50), 16302−16303. (m) Lyons, T. W.; Sanford, M. S.
Chem. Rev. 2010, 110 (2), 1147−1169. (n) Zhang, Y.-H.; Yu, J.-Q. J.
Am. Chem. Soc. 2009, 131 (41), 14654−14655.
MCSCF33 and MCQDPT234 were performed on truncated
systems on the optimized DFT structures using the GAMESS35
program. For these multireference calculations, no dispersion
or solvent corrections were considered. For the bimetallic
systems a (10,10) active space was chosen and for
monometallic complexes a (14,14) active space with second-
order perturbation theory was chosen on the basis of
convergence of the singlet−triplet energy gaps.
(4) Look, J. L.; Wick, D. D.; Mayer, J. M.; Goldberg, K. I. Inorg.
Chem. 2009, 48 (4), 1356−1369.
(5) (a) Keith, J. M.; Goddard, W. A. J. Am. Chem. Soc. 2009, 131 (4),
1416−1425. (b) Popp, B. V.; Morales, C. M.; Landis, C. R.; Stahl, S. S.
Inorg. Chem. 2010, 49 (18), 8200−8207. (c) Popp, B. V.; Stahl, S. S.
Chem.Eur. J. 2009, 15 (12), 2915−2922.
(6) Rostovtsev, V. V.; Henling, L. M.; Labinger, J. A.; Bercaw, J. E.
Inorg. Chem. 2002, 41 (14), 3608−3619.
ASSOCIATED CONTENT
* Supporting Information
■
(7) (a) Allen, K. E.; Heinekey, D. M.; Goldman, A. S.; Goldberg, K. I.
Organometallics 2014, 33 (6), 1337−1340. (b) Ishikawa, A.; Nakao, Y.;
Sato, H.; Sakaki, S. Inorg. Chem. 2009, 48 (17), 8154−8163.
(8) (a) Blakemore, J. D.; Schley, N. D.; Balcells, D.; Hull, J. F.; Olack,
G. W.; Incarvito, C. D.; Eisenstein, O.; Brudvig, G. W.; Crabtree, R. H.
J. Am. Chem. Soc. 2010, 132 (45), 16017−16029. (b) Brewster, T. P.;
Blakemore, J. D.; Schley, N. D.; Incarvito, C. D.; Hazari, N.; Brudvig,
G. W.; Crabtree, R. H. Organometallics 2011, 30 (5), 965−973.
(c) Codola, Z.; Cardoso, J. M. S.; Royo, B.; Costas, M.; Lloret-Fillol. J.
Chem.Eur. J. 2013, 19 (22), 7203−7213. (d) Hull, J. F.; Balcells, D.;
Blakemore, J. D.; Incarvito, C. D.; Eisenstein, O.; Brudvig, G. W.;
Crabtree, R. H. J. Am. Chem. Soc. 2009, 131 (25), 8730−8731.
(e) Ingram, A. J.; Wolk, A. B.; Flender, C.; Zhang, J.; Johnson, C. J.;
Hintermair, U.; Crabtree, R. H.; Johnson, M. A.; Zare, R. N. Inorg.
Chem. 2014, 53 (1), 423−433. (f) Zhou, M.; Balcells, D.; Parent, A. R.;
Crabtree, R. H.; Eisenstein, O. ACS Catal. 2012, 2 (2), 208−218.
(g) Zhou, M.; Hintermair, U.; Hashiguchi, B. G.; Parent, A. R.;
Hashmi, S. M.; Elimelech, M.; Periana, R. A.; Brudvig, G. W.; Crabtree,
R. H. Organometallics 2013, 32 (4), 957−965. (h) Zhou, M.; Schley,
N. D.; Crabtree, R. H. J. Am. Chem. Soc. 2010, 132 (36), 12550−
12551.
S
Additional experimental procedures for methanol formation,
isotopic labeling and kinetic studies. X-ray experimental for 3
and 4. Full Gaussian reference and coordinates for key
intermediates. This material is available free of charge via the
Internet at http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
*E-mail: eaison@ncsu.edu
■
Notes
The authors declare no competing financial interests.
ACKNOWLEDGMENTS
■
This work was supported by the NSF as part of the Center for
Enabling New Technologies through Catalysis (CENTC),
CHE-0650456 and CHE-1205189. We thank Karen I. Gold-
berg, William D. Jones, and Melanie S. Sanford for helpful
discussions.
(9) Williams, D. B.; Kaminsky, W.; Mayer, J. M.; Goldberg, K. I.
Chem. Commun. 2008, 35, 4195−4197.
REFERENCES
(10) Lehman, M. C.; Boyle, P. D.; Sommer, R. D.; Ison, E. A.
Organometallics 2014, 33 (19), 5081−5084.
■
(1) (a) Boisvert, L.; Goldberg, K. I. Acc. Chem. Res. 2012, 45 (6),
899−910. (b) Company, A.; Lloret, J.; Gomez, L.; Costas, M. Catal.
Met. Complexes 2012, 38, 143−228. (c) Golisz, S. R.; Gunnoe, T. B.;
Goddard, W. A., III; Groves, J. T.; Periana, R. A. Catal. Lett. 2011, 141
(2), 213−221. (d) Cai, X.-C.; Majumdar, S.; Fortman, G. C.; Cazin, C.
S. J.; Slawin, A. M. Z.; Lhermitte, C.; Prabhakar, R.; Germain, M. E.;
Palluccio, T.; Nolan, S. P.; Rybak-Akimova, E. V.; Temprado, M.;
Captain, B.; Hoff, C. D. J. Am. Chem. Soc. 2011, 133 (5), 1290−1293.
(e) Campbell, A. N.; Stahl, S. S. Acc. Chem. Res. 2012, 45 (6), 851−
863. (f) Shi, Z.; Zhang, C.; Tang, C.; Jiao, N. Chem. Soc. Rev. 2012, 41
(8), 3381−3430. (g) Stahl, S. S. Science 2005, 309 (5742), 1824−1826.
(11) Similar to the complex [Cp*Ir(PMe3)(Me)(Ch2Cl2)][BArF ]
4
reported by Bergman and coworkers, the cationic complexes reported
here are shown as methylene chloride adducts as the solvent free
complexes were shown to be unstable. See: Arndtsen, B. A.; Bergman,
R. G. Science 1995, 270, 1970.
(12) (a) Liu, F.; Concepcion, J. J.; Jurss, J. W.; Cardolaccia, T.;
Templeton, J. L.; Meyer, T. J. Inorg. Chem. 2008, 47 (6), 1727−1752.
(b) Uemura, S.; Spencer, A.; Wilkinson, G. Dalton Trans. 1973, 23,
2565−71. (c) Castillo-Blum, S. E.; Richens, D. T.; Sykes, A. G. Chem.
Commun. 1986, 14, 1120−1. (d) Castillo-Blum, S. E.; Richens, D. T.;
3583
DOI: 10.1021/ja512905t
J. Am. Chem. Soc. 2015, 137, 3574−3584

Journal of the American Chemical Society
Article
Sykes, A. G. Inorg. Chem. 1989, 28 (5), 954−60. (e) Burford, R. J.;
Piers, W. E.; Ess, D. H.; Parvez, M. J. Am. Chem. Soc. 2014, 136 (8),
3256−3263. (f) Hintermair, U.; Sheehan, S. W.; Parent, A. R.; Ess, D.
H.; Richens, D. T.; Vaccaro, P. H.; Brudvig, G. W.; Crabtree, R. H. J.
Am. Chem. Soc. 2013, 135 (29), 10837−10851. (g) Hintermair, U.;
Hashmi, S. M.; Elimelech, M.; Crabtree, R. H. J. Am. Chem. Soc. 2012,
134 (23), 9785−9795. (h) Graeupner, J.; Hintermair, U.; Huang, D.
L.; Thomsen, J. M.; Takase, M.; Campos, J.; Hashmi, S. M.; Elimelech,
M.; Brudvig, G. W.; Crabtree, R. H. Organometallics 2013, 32 (19),
5384−5390.
position of the complex. See, for example: Alben
́
iz, A. C.; Espinet, P.;
Lopez-Fernandez, R.; Sen, A. J. Am. Chem. Soc. 2002, 124 (38),
́
́
11278−11279.
(24) Halide exchange with halogenated solvents is a typical reaction
of transition metal hydrides. See, for example: Matano, Y.; Brown, S.
N.; Northcutt, T. O.; Mayer, J. M. Organometallics 1998, 17, 2939−
2941. In addition, for a proposed mechanism for the formation of 4
from 7, see Scheme S2 in the Supporting Information.
(25) Harding, L. B.; Goddard, W. A. J. Am. Chem. Soc. 1975, 97 (22),
6293−6299.
(26) Moss, B. J.; Bobrowicz, F. W.; Goddard, W. A. J. Chem. Phys.
1975, 63 (11), 4632−4639.
(13) McGhee, W. D.; Foo, T.; Hollander, F. J.; Bergman, R. G. J. Am.
Chem. Soc. 1988, 110 (25), 8543−8545.
(27) Hay-Motherwell, R. S.; Wilkinson, G.; Hussain-Bates, B.;
Hursthouse, M. B. Polyhedron 1993, 12 (16), 2009−12.
(28) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; et al. Gaussian 09,
Revisions, D.01; Gaussian, Inc.: Wallingford, CT, 2009.
(29) Adamo, C.; Barone, V. J. Chem. Phys. 1999, 110 (13), 6158−
6170.
(14) (a) Fortner, K. C.; Laitar, D. S.; Muldoon, J.; Pu, L.; Braun-
Sand, S. B.; Wiest, O.; Brown, S. N. J. Am. Chem. Soc. 2007, 129 (3),
588−600. (b) Fortner, K. C.; Laitar, D. S.; Muldoon, J.; Pu, L.; Braun-
Sand, S. B.; Wiest, O.; Brown, S. N. J. Am. Chem. Soc. 2007, 129 (3),
588−600.
(15) (a) Cheng, P. T.; Nyburg, S. C. Inorg. Chem. 1975, 14 (2), 327−
329. (b) Carty, P.; Walker, A.; Mathew, M.; Palenik, G. J. J. Chem. Soc.
D 1969, 23, 1374−5. (c) Baranovskii, I. B.; Golubnichaya, M. A.;
Koz’min, P. A.; Surazhskaya, M. D. Zh. Neorg. Khim. 1995, 40 (10),
1634−40. (d) Ahmed, T. S.; Tonks, I. A.; Labinger, J. A.; Bercaw, J. E.
Organometallics 2013, 32 (11), 3322−3326.
(16) (a) Sutherland, B. R.; Cowie, M. Organometallics 1985, 4 (9),
1637−48. (b) Chandrasekhar, V.; Mahanti, B.; Bandipalli, P.;
Bhanuprakash, K. Inorg. Chem. 2012, 51 (20), 10536−10547.
(c) Panichakul, D.; Su, Y.; Li, Y.; Deng, W.; Zhao, J.; Li, X.
Organometallics 2008, 27 (24), 6390−6392. (d) Slaney, M. E.;
Anderson, D. J.; Ferguson, M. J.; McDonald, R.; Cowie, M. J. Am.
Chem. Soc. 2010, 132 (46), 16544−16558. (e) Ionkin, A. S.; Wang, Y.;
Marshall, W. J.; Petrov, V. A. J. Organomet. Chem. 2007, 692 (22),
4809−4827. (f) Dorta, R.; Rozenberg, H.; Shimon, L. J. W.; Milstein,
D. J. Am. Chem. Soc. 2002, 124 (2), 188−189. (g) Dorta, R.;
Rozenberg, H.; Shimon, L. J. W.; Milstein, D. Chem.Eur. J. 2003, 9
(21), 5237−5249. (h) Dadci, L.; Elias, H.; Frey, U.; Hoernig, A.;
Koelle, U.; Merbach, A. E.; Paulus, H.; Schneider, J. S. Inorg. Chem.
1995, 34 (1), 306−15. (i) Park, J. T.; Nishioka, T.; Suzuki, T.; Isobe,
K. Bull. Chem. Soc. Jpn. 1994, 67 (7), 1968−71.
(30) Andrae, D.; Haeussermann, U.; Dolg, M.; Stoll, H.; Preuss, H.
Theor. Chim. Acta 1990, 77 (2), 123−41.
(31) Marenich, A. V.; Cramer, C. J.; Truhlar, D. G. J. Phys. Chem. B
2009, 113 (18), 6378−6396.
(32) Grimme, S.; Ehrlich, S.; Goerigk, L. J. Comput. Chem. 2011, 32
(7), 1456−1465.
(33) Schmidt, M. W.; Gordon, M. S. Annu. Rev. Phys. Chem. 1998, 49,
233−266.
(34) Chaban, G. M.; Gordon, M. S. J. Phys. Chem. A 1999, 103 (1),
185−189.
(35) Schmidt, M. W.; Baldridge, K. K.; Boatz, J. A.; Elbert, S. T.;
Gordon, M. S.; Jensen, J. H.; Koseki, S.; Matsunaga, N.; Nguyen, K. A.;
et al. J. Comput. Chem. 1993, 14 (11), 1347−63.
(17) (a) Kownacki, I.; Kubicki, M.; Blazejewska-Chadyniak, P.
Polyhedron 2013, 53, 26−31. (b) Yuan, X.; Zhang, S.; Ding, Y. Inorg.
Chem. Commun. 2012, 17, 26−29. (c) Kownacki, I.; Marciniec, B.;
Kubicki, M. Chem. Commun. 2003, 1, 76−77.
(18) (a) Crabtree, R. H. 2012, 112 (3), 1536−1554;. (b) Park-
Gehrke, L. S.; Freudenthal, J.; Kaminsky, W.; DiPasquale, A. G.;
Mayer, J. M. Dalton Trans. 2009, 11, 1972−1983. (c) Thomsen, J. M.;
Sheehan, S. W.; Hashmi, S. M.; Campos, J.; Hintermair, U.; Crabtree,
R. H.; Brudvig, G. W. J. Am. Chem. Soc. 2014, 136 (39), 13826−13834.
(d) Turlington, C. R.; Harrison, D. P.; White, P. S.; Brookhart, M.;
Templeton, J. L. Inorg. Chem. 2013, 52 (19), 11351−11360.
(19) Complex 4 with MeB(C6F5)3− counterions have been previously
reported by Heinekey and coworkers: Meredith, J. M.; Robinson, R.,
Jr.; Goldberg, K. I.; Kaminsky, W.; Heinekey, D. M. Organometallics
2012, 31 (5), 1879−1887.
(20) Calculations on the full model 3 with smaller MCSCF active
space and 3′ with larger active spaces all show that the energy of the
singlet is less than the triplet.
(21) (a) Veige, A. S.; Slaughter, L. M.; Lobkovsky, E. B.; Wolczanski,
P. T.; Matsunaga, N.; Decker, S. A.; Cundari, T. R. Inorg. Chem. 2003,
42 (20), 6204−6224. (b) Veige, A. S.; Slaughter, L. M.; Wolczanski, P.
T.; Matsunaga, N.; Decker, S. A.; Cundari, T. R. J. Am. Chem. Soc.
2001, 123 (26), 6419−6420.
(22) The average Ir−O bond length was obtained from 1205 hits in a
search of the Cambridge Structural Database
(23) In the presence of proton sources, the species [Cp*IrIV
O(NHC)Me][BArF ] may readily couple with radical traps such as
4
TEMPO and 1,4-cyclohexadiene and prevent formation of 5 and 6,
which are necessary for the formation of the hydride 7. Alternatively, 7
may react rapidly with the radical traps, resulting in the unfavorable
formation of an unknown iridium species and subsequent decom-
3584
DOI: 10.1021/ja512905t
J. Am. Chem. Soc. 2015, 137, 3574−3584