Stepwise Iodide-Free Methanol Carbonylation via Methyl Acetate Activation by Pincer Iridium Complexes

Article abstract of DOI:10.1021/jacs.1c05185

Iodide is an essential promoter in the industrial production of acetic acid via methanol carbonylation, but it also contributes to reactor corrosion and catalyst deactivation. Here we report that iridium pincer complexes mediate the individual steps of methanol carbonylation to methyl acetate in the absence of methyl iodide or iodide salts. Iodide-free methylation is achieved under mild conditions by an aminophenylphosphinite pincer iridium(I) dinitrogen complex through net C-O oxidative addition of methyl acetate to produce an isolable methyliridium(III) acetate complex. Experimental and computational studies provide evidence for methylation via initial C-H bond activation followed by acetate migration, facilitated by amine hemilability. Subsequent CO insertion and reductive elimination in methanol solution produced methyl acetate and acetic acid. The net reaction is methanol carbonylation to acetic acid using methyl acetate as a promoter alongside conversion of an iridium dinitrogen complex to an iridium carbonyl complex. Kinetic studies of migratory insertion and reductive elimination reveal essential roles of the solvent methanol and distinct features of acetate and iodide anions that are relevant to the design of future catalysts for iodide-free carbonylation.

Full text of DOI:10.1021/jacs.1c05185

pubs.acs.org/JACS
Article
Stepwise Iodide-Free Methanol Carbonylation via Methyl Acetate
Activation by Pincer Iridium Complexes
Changho Yoo and Alexander J. M. Miller*
Cite This: https://doi.org/10.1021/jacs.1c05185
Read Online
sı
*
Metrics & More
Article Recommendations
Supporting Information
ACCESS
ABSTRACT: Iodide is an essential promoter in the industrial
production of acetic acid via methanol carbonylation, but it also
contributes to reactor corrosion and catalyst deactivation. Here we
report that iridium pincer complexes mediate the individual steps
of methanol carbonylation to methyl acetate in the absence of
methyl iodide or iodide salts. Iodide-free methylation is achieved
under mild conditions by an aminophenylphosphinite pincer
iridium(I) dinitrogen complex through net C−O oxidative
addition of methyl acetate to produce an isolable methyliridium-
(III) acetate complex. Experimental and computational studies
provide evidence for methylation via initial C−H bond activation
followed by acetate migration, facilitated by amine hemilability. Subsequent CO insertion and reductive elimination in methanol
solution produced methyl acetate and acetic acid. The net reaction is methanol carbonylation to acetic acid using methyl acetate as a
promoter alongside conversion of an iridium dinitrogen complex to an iridium carbonyl complex. Kinetic studies of migratory
insertion and reductive elimination reveal essential roles of the solvent methanol and distinct features of acetate and iodide anions
that are relevant to the design of future catalysts for iodide-free carbonylation.
The primary challenge in halide-free carbonylation with
molecular catalysts is accessing the organometallic methyl
complexes that mediate C−C bond formation. Most catalysts
must rely on potent alkylating agents such as methyl iodide to
form the M−CH3 unit (Scheme 1A). One strategy to minimize
methyl iodide concentrations utilizes quaternary ammonium
iodide or phosphonium iodide salts that produce only
equilibrium amounts of methyl iodide under the reaction
conditions of carbonylation.17 Iodide is still required, however,
and many of the concerns noted above remain in these
systems.
An alternative approach involves designing organometallic
catalysts capable of C−O oxidative addition (Scheme 1B),
avoiding the need for H3C−I oxidative addition altogether.
The most obvious and ideal reactant would be CH3OH, but
we are not aware of any well-defined C−O oxidative addition
reactions of methanol. We hypothesized that methyl acetate
(MeOAc) could be a more promising candidate than methanol
for halide-free alkylation because acetate is a better leaving
group than hydroxide. Methyl acetate is also a coproduct and
INTRODUCTION
■
The production of acetic acid by methanol carbonylation is
currently one of the largest scale industrial processes based on
homogeneous catalysis, with annual capacity exceeding 13
million tons.1,2 Iodide salts or methyl iodide are essential
promoters in the catalytic reaction. Hydroiodic acid and
methyl iodide are continually generated in situ during the
process, with the latter initiating the organometallic catalytic
cycle by forming a metal−methyl species (Scheme 1).3−5
However, the need for iodide leads to serious disadvantages.
Methyl iodide and hydroiodic acid are toxic and corrosive,
necessitating expensive safety and engineering controls.4,6
Iodide also complicates the chemical pathways. In the
rhodium-catalyzed (“Monsanto”) process, the precipitation of
RhI3 is a significant catalyst deactivation pathway.4−6 In the
iridium-catalyzed (“Cativa”) process, iodide inhibits CO
migratory insertion by generating inactive iodide-bound
species, necessitating the use of halide abstractors such as
ruthenium(II) salts to achieve high activity.4−7 Therefore,
development of low-iodide or iodide-free processes has
attracted the attention of both academic and industrial
scientists. Advances in iodide-free carbonylation have been
driven by heterogeneous Lewis acid catalysts.8−16 While
promising, these systems typically require temperatures of ca.
200 °C and produce dimethyl ether as an undesired byproduct.
Homogeneous molecular catalysts for methanol carbonylation
without iodide promoters remain elusive.
Received: May 19, 2021
https://doi.org/10.1021/jacs.1c05185
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX
© XXXX American Chemical Society
A

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
complexes in the presence of methyl iodide and metal salt
promoters.22 Although partial dissociation of the ligand was
observed under catalytic conditions, stoichiometric studies
established the viability of each reaction step, including Lewis
acid promotion of the C−C bond-forming migratory insertion
step.23 We recently isolated a NCOP iridium(I)−dinitrogen
compound and found that it facilitates decarbonylative C−O
bond cleavage of ethers, initiated by C−H bond activation.24
These results led us to a stepwise study of iodide-free methanol
carbonylation using methyl acetate as the methylating reagent.
Here, the aminophenylphosphinite ligand supports clean
formal oxidative addition of methyl acetate to generate an
isolable methyliridium acetate complex. The subsequent
migratory insertion and reductive elimination steps could
then be studied individually, enabling a detailed understanding
of how iodide-free conditions with acetate ions compared to
conditions with iodide ions (helping to address the questions
in Scheme 1B). The elimination process, in particular, has
previously eluded careful interrogation in iridium-catalyzed
carbonylation, leading to conflicting views on whether C−O or
C−I bond formation occurs from the acetyl intermediate in the
Cativa process.3,25−27 The present study provides a rare
opportunity to directly compare acetyl complex reactivity by
either methanolysis to generate methyl acetate directly or
reductive elimination with iodide to generate acetyl iodide as
an intermediate.
Scheme 1. Comparison of a Carbonylation Process Using
Iodide and a Possible Iodide-Free Process via C−O
Activation
common solvent for methanol carbonylation.1−4 However,
3
examples of methyl acetate Csp −O bond cleavage are
RESULTS AND DISCUSSION
■
extremely limited and generally produce product mix-
tures.18−21 In an encouraging example, the Goldman group
showed that initial C−H bond activation at an iridium center
MeOAc Activation. Initial studies investigating the
activation of MeOAc by previously reported (NCOP)Ir(CO)
complexes did not show promising reactivity, despite the
ability of these carbonyl complexes to carry out the individual
steps and overall catalytic reaction of methanol carbonylation
in the presence of methyl iodide.22,23 Inspired by reports from
the Goldman group demonstrating net C−O bond activation
via initial C−H bond activation,21,28−30 and our own recent
observation of ether decarbonylation via C−H bond
21
3
initiates net Csp −O oxidative addition of methyl acetate.
However, the reaction required prolonged heating at 125 °C
and suffered from unwanted C−H activation of tert-butyl
substituents of the tertiary phosphine donor, precluding
isolation of the methyliridium acetate product.
Our group previously reported catalytic methanol carbon-
ylation using aminophenylphosphinite (NCOP) pincer iridium
Figure 1. (A) MeOAc activation by 1 via C−H activation. (B) Structural representation of 2. (C) 31P{1H} NMR spectrum after reaction of 1 and
1
MeOAc at room temperature for 24 h. (D) Partial H−13C HSQC spectrum of 3 showing correlation of geminal protons with the carbon in an
acetoxymethyl (−CH2OAc) group.
B
https://doi.org/10.1021/jacs.1c05185
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
Figure 2. Calculated Gibbs free energies (kcal/mol) for reaction of 1 with MeOAc via C−H activation (blue, black, and green) and direct C−O
activation (red). Values of G are given relative to 1/2(1 − N2) + MeOAc. The free energies correspond to a reference state of 1 M concentration for
each species participating in the reaction and T = 298.15 K.
activation,24 we turned to the dinitrogen complex
[(MeO‑EtNCOP)Ir]2(μ-N2) (1).
ure 1A and Figures S4−S8). The CH2OAc group was
identified by the two diastereotopic geminal protons at δ
1
6.23 and δ 5.57 (d, JHH = 11.3 Hz) and 13C resonance at δ
A red solution of 1 and 1 equiv of MeOAc in benzene
became colorless after heating at 80 °C for 1 h. NMR
spectroscopy revealed 75% yield of a new species with a
31P{1H} NMR signal (δ 133.36) upfield shifted from that of 1
(δ 163.68). A diagnostic upfield methyl resonance in the
13C{1H} NMR spectrum (δ − 29.47, d, JPC = 6.7 Hz) supports
the formation of the iridium methyl acetate complex
89.49 (Figure 1D). After 24 h at room temperature, a mixture
of 1 (10%), 3 (70%), and 2 (20%) was obtained (Figure 1C).
Heating this reaction mixture at 80 °C for 1 h resulted in
complete consumption of 1 and 3 to produce 2 in 75% yield.
If, instead of heating, a similarly obtained mixture containing 1,
2, and 3 was exposed to vacuum to remove the MeOAc, only 1
and 2 remained after the solids were redissolved in C6D6 under
N2. The formation of 3 from 1 is therefore a reversible process,
while the formation of 2 is not.
The net C−O oxidative addition of methyl acetate mediated
by the (MeO‑EtNCOP)Ir center is noteworthy for proceeding in
high yield under relatively mild conditions. For comparison,
MeOAc activation by a diphosphine-based iridium pincer
complex utilized by Goldman et al. required heating at 125 °C
due to formation of a stable intermediate in which methylene
inserts into the iridium−aryl bond of the pincer backbone.21
Under the reaction conditions required to finally reach the
methyliridium acetate complex, there is a competing side
reaction involving C−H bond activation of a tert-butylphos-
phine substituent, preventing isolation of (tBu4PCP)Ir(Me)-
(OAc) (tBu4PCP is 2,6-(tBu2PCH2)2-C6H3). In contrast, the
present complex 2 is generated even at room temperature, is
produced in high yield at 80 °C according to NMR
spectroscopy, and can be isolated in a thermally stable
crystalline form.
(
MeO‑EtNCOP)Ir(CH3)(OAc) (2) (Figure 1A). Compound 2
was isolated in 42% yield after crystallization from diethyl ether
at −35 °C. An X-ray diffraction (XRD) study revealed
apseudo-octahedral Ir coordination with the methyl group cis
to the pincer phenyl ligand and the acetate ligand in a
bidentate (κ2) binding mode (Figure 1B).
Monitoring the conversion of 1 to 2 by NMR spectroscopy
at room temperature revealed an intermediate. In C6D6, the
reaction of 1 with 1 equiv of MeOAc was slow (∼60%
conversion of 1 in 10 h at room temperature) and resulted in
multiple species along with trace amounts of 2. When the
reaction was instead conducted in a 1:1 mixture of
MeOAc:C6D6 (∼110 equiv of MeOAc relative to 1) at room
temperature, however, a prominent new 31P resonance grew in
at δ 143.52 (Figure 1C). The intermediate 3 features a hydride
1
resonance at δ −26.11 (d, JPH = 28.6 Hz) in the H NMR
spectrum, indicating a weak donor trans to the hydride.
1H−13C HSQC and HMBC NMR experiments enabled
assignment as (MeO‑EtNCOP)Ir(H)(κ2C,O-CH2OAc) (3, Fig-
C
https://doi.org/10.1021/jacs.1c05185
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
Figure 3. (A) Synthesis of carbonyl species. (B) Structural representation of 5-cis from X-ray diffraction analysis, with ellipsoids drawn at 50%
probability level. Hydrogen atoms are omitted for clarity.
To understand why the aminophenylphosphinite pincer
ligand supports cleaner reactivity under milder conditions, the
detailed mechanism of net C−O oxidative addition was
examined by using density functional theory (DFT). Four
pathways were considered: direct C−O oxidative addition and
three pathways that start with C−H bond activation (Figure
2). The experimental observation of alkyl hydride intermediate
3 provides strong evidence against a direct C−O oxidative
addition pathway, instead supporting initial oxidative addition
of a C−H bond of MeOAc (Figure 1A). Accordingly, the
barrier to C−H activation was computed to be ca. 30 kcal/mol
while the barrier to direct C−O oxidative addition was
computed to be 52.1 kcal/mol.
Possible C−H activation routes were evaluated in detail
through DFT calculations (Figure 2). Paths A, B, and C all
start with N2 dissociation and C−H activation by the
coordinatively unsaturated intermediate INT-1. The resulting
hydride species 3 is higher in energy than 1 (ΔG = +14.0 kcal/
mol) due to unfavorable N2 dissociation (ΔG = +20.3 kcal/
mol). However, in experimental practice the reaction may be
driven by low N2 solubility at the elevated temperature and/or
the excess amount of MeOAc, consistent with our
experimental observation of equilibrium formation of 3 with
∼110 equiv of MeOAc at room temperature.
In path A, alkyl hydride 3 undergoes acetate migration to
produce an intermediate with a methylidene ligand cis to the
hydride (INT-A2), likely by retro-electrocyclization (TS-A2,
ΔG‡ = +36.6 kcal/mol). The methyl species 2 can be
generated by 1,1-hydride migration (TS-A3, ΔG‡ = +43.9
kcal/mol), which has the highest activation barrier in the
overall reaction.
In Path B, C−H bond activation produces 3-trans, in which
the acetoxymethyl carbon (−CH2OAc) is trans to hydride (ΔG
= +3.3 kcal/mol relative to 3). Both 3-trans and 3 are accessed
by the same early C−H activation transition state TS-1, with
barrierless coordination of oxygen either cis or trans to the
hydride producing the respective isomers of 3 (Figures S46
and S47). Complex 3-trans could also form by isomerization
of 3, either via TS-1 (ΔG = +16.2 kcal/mol relative to 3) or via
oxygen dissociation, bending of the Ir−CH2OAc bond, and
recoordination of oxygen (ΔG < 14 kcal/mol based on the
potential energy surface scan, Figure S48). The subsequently
formed trans-methylidene (INT-B2) cannot undergo direct
C−H reductive elimination with hydride, so it is first inserted
into the Ir−Caryl bond to give INT-B3. This process has the
highest energy TS in path B (TS-B3, ΔG‡ = +40.3 kcal/mol).
The subsequent C−H reductive elimination yields INT-B4, in
which the methyl from methyl acetate is added on the aromatic
backbone of the NCOP ligand. In the transition state (TS-B4),
the amine donor is dissociated, enabling a geometric distortion
that facilitates concerted reductive elimination between
adjacent hydride and ArCH2 ligands. Finally, the C−C
oxidative addition of the Ar−CH3 bond in INT-B4 produces
the methyl species 2.
In path C, the amine arm acts as a proton relay. From trans-
methylidene intermediate INT-B2, Ir−N bond cleavage and
N−H bond formation occur (i.e., amine deprotonation of the
hydride ligand) to give INT-C3. The ammonium group then
transfers the proton to the methylidene ligand to form the
methyl species 2. The highest activation barrier is deprotona-
tion of hydride by the amine (TS-C3, ΔG‡ = +40.2 kcal/mol).
All pathways involving initial C−H activation (paths A, B,
and C, Figure 2) are plausible based on the computational data
and have significantly lower computed activation barriers than
the direct C−O bond activation (path D). The slightly lower
overall free energy spans for paths B and C relative to path A
are attributed to hemilability of the amine donor of the pincer
ligand. A mechanism akin to path B was proposed in C−O and
C−C bond activation by diphosphine-based pincer rhodium
and iridium complexes;21,31−33 however, a higher barrier is
encountered in the (tBu4PCP)Ir analogue to TS-B4.21 The
lower barrier of TS-B4 for the (NCOP)Ir system is ascribed to
amine hemilability. Whereas no phosphine dissociation was
apparent in calculations of the (tBu4PCP)Ir system, thus
requiring an isomerization sequence before reductive elimi-
nation, the (NCOP)Ir system does not require any geometric
isomerization, and instead amine dissociation is apparent in
TS-B4 (Figure 2), which would provide increased flexibility for
low-barrier C−H reductive elimination. Path C also features a
relatively low-barrier pathway enabled by a labile amine donor
acting as a hydride-to-methylidene proton shuttle. Hemilability
of the amine donor may therefore be responsible for the
relatively low barriers and thus the clean reactivity of the
(NCOP)Ir complex under mild conditions compared to the
(
tBu4PCP)Ir system.
Having identified a clean methylation reaction involving
methyl acetate, we individually examined CO migratory
insertion and acetyl reductive elimination steps.
CO Migratory Insertion. Addition of 1 atm of CO to a
solution of methyliridium acetate complex 2 at room
D
https://doi.org/10.1021/jacs.1c05185
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
Scheme 2. (A) Carbonylation of 4-cis in CD3OD; (B) Possible Routes for Formation of 6 and Acetyl Products
Scheme 3. (A) Isotopic Labeling Experiment of 4-cis-13CH3 Carbonylation; (B) The Overall Net Reaction of Methanol
Carbonylation by Pincer Complexes
temperature in C6D6 immediately produced the carbonyl
complex (MeO‑EtNCOP)Ir(CH3)(OAc)(CO) (4-trans, where
trans indicates the relative orientation of the carbonyl and
methyl ligands). The methyl resonance was found at δ 0.71 in
the 1H NMR spectrum. Heating a solution of 4-trans at 80 °C
in C6D6 under CO produced a new species with a methyl
resonance shifted to δ 0.47. The new species was assigned as
the isomer (MeO‑EtNCOP)Ir(CH3)(CO)(OAc) (4-cis, Figure
3A). The 13C{1H} NMR spectrum of 4-cis confirms the
presence of a CO ligand (δ 185.36) and a methyl ligand (δ
−30.92, d, JPC = 7.2 Hz). The infrared (IR) spectrum of 4-cis
is consistent with a single carbonyl ligand (νCO = 2023 cm−1)
under 1 atm of CO overnight resulted in no C−C bond
formation. Trace amounts of (MeO‑EtNCOP)Ir(CO) (6) were
the only new product observed.
Reactions in methanol, however, tell a different story
(Scheme 2). Addition of 1 atm of CO to a solution of 4-cis
in CD3OD at ambient temperature resulted in formation of a
new methyl species (31P{1H} NMR δ 141.93) in ∼70% yield
within 5 h. New methyl resonances (1H NMR δ 0.60, d, JPH
=
2.0 Hz and 13C{1H} NMR δ −9.21, d, JPC = 6.2 Hz) were
found slightly downfield of those in 4-cis. Two carbonyl
carbon resonances were found at δ 173.17 and 168.75 in the
13C NMR spectrum, indicating that the new product is a
methyl dicarbonyl complex with an outer-sphere acetate
counteranion, [(MeO‑EtNCOP)Ir(CH3)(CO)2][OAc] ([7]-
[OAc], Scheme 2). Monitoring the reaction over 18−22 h
did not lead significant changes in the ratio of products,
and a monodentate acetate ligand (νC=O = 1622 cm−1, νC−O
=
1316 cm−1). After 36 h heating at 80 °C, the ratio of 4-trans:4-
cis was 1:10. Higher purity samples of 4-cis could be obtained
from the reaction of (MeO‑EtNCOP)Ir(CO) (6) with CH3I to
produce (MeO‑EtNCOP)Ir(CH3)(CO)(I) (5-cis), followed by
iodide abstraction with AgOAc (Figure 3). Heating a solution
of pure 4-cis under N2 in C6D6 at 80 °C produced a mixture of
4-trans and 4-cis, with a similar ratio as observed when after
heating 4-trans, confirming that the two isomers are in
equilibrium (ΔG = −1.6(1) kcal/mol favoring formation of 4-
cis). DFT calculations also predict that 4-cis is thermodynami-
cally favored over 4-trans (ΔG = −2.9 kcal/mol in the gas
phase).
suggesting that the reaction achieved equilibrium (Keq
24.0(3), ΔG = −1.88(1) kcal/mol).
=
Heating a solution of [7][OAc] in CD3OD under 1 atm of
CO at 65 °C led to complete conversion to the carbonyl
complex 6 (84% yield) after 2 days, with concomitant
production of partially deuterated methyl acetate
CH3COOCD3 (1H NMR δ 2.02, 106% yield) and acetic
acid CH3COOD (1H NMR δ 1.92, 42% yield) (Scheme 2A).
The partially deuterated methyl acetate could form upon
reductive elimination of acetyl with CD3OD solvent or upon
reductive elimination of methyl and acetate groups from
[7][OAc] (without formation of the acetyl intermediate)
No CO migratory insertion to form an acetyl product was
observed during thermolysis of 4-cis under CO in C6D6.
Similarly, refluxing solutions of 4-cis in CD2Cl2 or CD3CN
E
https://doi.org/10.1021/jacs.1c05185
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
Figure 4. Calculated transition state energy (free energies in kcal/mol) for CO insertion in 4-cis and [7]+ and acetate dissociation energy (A).
Calculated transition state structures for migratory insertion in 4-cis (B) and [7]+ (C).
followed by transesterification of methyl acetate with CD3OD
solvent (Scheme 2B). In the former case, a total of 2 equiv of
acetyl products (CH3COOCD3 + CH3COOD) would be
formed, whereas only 1 equiv is expected in the latter case
(Scheme 2B). The formation of ∼1.5 equiv of acetyl products
indicates acetyl formation via carbonylation. To confirm the
origin of the acetic acid and methyl acetate products, an
isotopic labeling experiment was performed.
cationic species is expected to possess a more electrophilic CO
ligand, facilitating nucleophilic attack by the methyl ligand.7,35
In addition, the methyl group in [7]+ may be more
nucleophilic due to the strong trans influence of the carbonyl
ligand relative to acetate. Support for this notion comes from
comparisons (Table S13) of calculated Ir−CH3 bond distances
in [7]+ (2.14 Å) vs 4-cis (2.11 Å) and comparisons of NBO
charges on the methyl carbon in [7]+ (−0.80) vs 4-cis
(−0.77). The NBO charge on the carbonyl carbon is similar in
magnitude but opposite in sign, consistent with a more
electrophilic CO ligand in [7]+ compared to 4-cis.
The calculations suggest methanol solvation does not
significantly impact the CO insertion barrier for either 4-cis
or [7]+ (Figure 4A). Instead, we propose that the primary role
of the methanol solvent is to promote pre-equilibrium acetate
dissociation through dipole and hydrogen-bonding interac-
tions. The calculations agree that substitution of acetate by CO
is more accessible in methanol (ΔGdiss = +6.6 kcal/mol) than
in CH2Cl2 (+14.6 kcal/mol) (Figure 4A). The experimental
data show that the acetate dissociation and CO binding is
slightly exergonic (ΔGdiss = −1.88(1) kcal/mol), which is in
reasonable agreement with DFT when considering that the
calculations do not account for explicit solvent interactions
such as hydrogen-bonding interactions between the methanol
solvent and acetate anion.
Reaction of 6 with 13CH3I, followed by iodide abstraction
with AgOAc, afforded the 13C-labeled methyl complex
(
MeO‑EtNCOP)Ir(13CH3)(CO)(OAc) (4-cis-13C). After heat-
ing a CD3OD solution of 4-cis-13C at 65 °C under 1 atm of
CO for 18 h, a 13C-enriched signal was detected at δ 20.48 in
the 13C NMR spectrum, with a corresponding doublet (δ 2.02,
1JCH = 129.4 Hz) in the H NMR spectrum indicating the
1
formation of labeled methyl acetate 13CH3COOCD3 by
carbonylation and elimination of acetyl with CD3OD (Scheme
3A). The reaction is balanced by proton transfer from
methanol to acetate, forming acetic acid without 13C
enrichment, CH3COOD (1H NMR δ 1.95, s), as a coproduct.
Because 4-cis can be produced from the activation of methyl
acetate followed by CO addition, the overall reaction is
methanol carbonylation to acetic acid using methyl acetate as a
methylating promoter (Scheme 3B).
Solvent Effects on CO Insertion: Facilitating Acetate
Dissociation. The acceleration of migratory insertion by
methanol solvent has been observed with the Cativa catalyst.34
Similarly, the carbonylation of methyl complex 4-cis was only
observed in methanol solution. Given that methanol was the
only solvent in which CO substituted the acetate ligand in 4-
cis, we hypothesized that formation of cationic dicarbonyl
species [7]+ was key for CO migratory insertion. The DFT-
calculated transition state energies for CO migratory insertion
in 4-cis and [7]+ are compared in Figure 4A. Both transition
state structures are consistent with the usual mechanism of
methyl migration to the CO ligand (Figure 4B,C). The barrier
for neutral acetate species 4-cis is high (GTS,4‑cis = +48.1 kcal/
mol); the migratory insertion barrier for cationic dicarbonyl
The DFT calculations suggests that generation of a cationic
species is important for CO migratory insertion. This is
consistent with a body of experimental evidence that relatively
electron-deficient cationic alkyl carbonyl complexes undergo
fast migratory insertion.7,35 For example, iodide inhibits CO
migratory insertion in the Cativa process, and halide
abstractors can be used to achieve high activity.4−7 In our
prior work with pincer−crown ether ligands, we also observed
CO insertion in cationic species [κ4-(15c5NCOP)Ir(13CH3)-
(CO)]+ was ∼11-fold faster than in neutral iodide species κ3-
(
15c5NCOP)Ir(13CH3)(CO)(I).23
To directly assess migratory insertion at a cationic species,
we generated a cationic dicarbonyl complex and examined
acetyl formation reactivity. The cationic bis(carbonyl) species
[7]+ is ca. 20 kcal/mol lower (GTS,7 = +27.8 kcal/mol). The
+
F
https://doi.org/10.1021/jacs.1c05185
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
with BArF − anion (ArF = 3,5-bis(trifluoromethyl)phenyl),
processes relevant to methanol carbonylation,3,25−27 and
some reports point to methanolysis while others propose C−
I reductive elimination to produce acetyl iodide as an
intermediate.3,25−27,36,37
4
[(MeO‑EtNCOP)Ir(CH3)(CO)2][BArF ] ([7][BArF ]), was
4
synthesized from the reaction of 4-cis with NaBArF 4 under a
4
CO atmosphere (Scheme 4). CD2Cl2 solutions of [7][BArF ]
4
The reactivity of iodide species 5-cis was examined under
CO to compare with the previously described reactivity of
acetate complex 4-cis. Because of poor solubility of 5-cis in
CD3OD, 5-cis was dissolved in a mixture of 90% CD3OD and
10% 1,2-dichloroethane (DCE) and charged with 1 atm of
CO. At ambient temperature, relatively little iodide dissocia-
tion was observed. After 24 h the mixture comprised unreacted
5-cis, the isomer where the CO is trans to methyl
Scheme 4. Generation of Acetyl via Cationic Species
Formation
(
MeO‑EtNCOP)Ir(CH3)(I)(CO) (5-trans), and ∼40% yield of
[7]+. This contrasts the behavior of 4-cis, which generated
70% yield of [7]+ after only 5 h (vide supra), indicating that
iodide dissociation is less favorable than acetate dissociation in
methanol solvent. Heating this mixture for 3 h at 65 °C led to
∼40% conversion to two iridium carbonyl products, 6 and
iridium(III) hydridoiodide species (MeO‑EtNCOP)Ir(H)(CO)-
1
(I) (9), identified by a hydride resonance in the H NMR
spectrum (δ −16.64, d, JPH = 19.4 Hz), in an ∼1:8 ratio. The
formation of hydridoiodide 9 is similar to our previous study of
a crown-ether-containing iodide complex22 but contrasts the
reactivity of 4-cis to produce only iridium(I) carbonyl 6. This
raises the possibility that one role of iodide is to shift
speciation away from iridium(I) carbonyl, which could have
important implications in catalysis. For example, hydride
complexes are proposed to be responsible for catalyzing the
undesired water-gas shift reaction as a side-reaction during the
Cativa process.1,3,26,38
To better compare the influence of acetate and iodide
ligands on acetyl formation, the kinetics of CO insertion of
iodide (5-cis) and acetate (4-cis) complexes were studied. The
kinetics were first compared in CD3OD/DCE (8:2) solution,
since 5-cis is insoluble in pure methanol (Table 1). Samples
display a methyl resonance at δ −8.57 (d, JPC = 6.8 Hz) in 1H
NMR spectra and two carbonyl resonances at δ 171.54 (s) and
167.41 (d, JPC = 5.4 Hz) in 13C NMR spectra. The CO
stretching frequencies of [7][BArF ] observed by IR spectros-
4
copy (νCO = 2105, 2064 cm−1) are higher energy than those of
4-cis (2023 cm−1) and 5-cis (2015 cm−1), confirming that the
carbonyl ligands are more electrophilic in [7]+.
The cationic species [7][BArF ] underwent CO insertion in
4
acetonitrile, as predicted.23 Thermolysis of [7][BArF ] in
4
CD3CN at 80 °C under 1 atm of CO for 10 h resulted in
∼60% yield of a new species (31P{1H} NMR δ 141.61) with a
diagnostic acetyl peak (1H NMR δ 1.82, s) indicative of
[(MeO‑EtNCOP)Ir(COCH3)(CO)2][BArF ] ([8][BArF ]). Un-
Table 1. Half-Lives (t1/2, min) for Conversion of 4-cis, 5-cis,
4
4
a
and [7][BArF ] under 1 atm of CO in Methanol
fortunately, we were unable to isolate [8][BArF ] because
4
4
removal of the CO atmosphere resulted in reversion to
half-life for conversion (min)
[7][BArF ] (Scheme 4).
4
b
b
solvent
5-cis
4-cis
[7][BArF ]
4
The combined results are consistent with acetyl formation
requiring acetate dissociation to reach a cationic intermediate
capable of CO migratory insertion. Accordingly, only 25%
conversion of [7][BArF ] to [8][BArF ] was observed in the
CD3OD/DCE (8:2)
CD3OD/DCE (9:1)
CD3OD only
250(10)
120(10)
72
140(30)
120
c
270
c
−
69
95
4
4
a
1
Reactions were heated at 65 °C and monitored by H and 31P{1H}
presence of (mostly insoluble) LiOAc in CD3CN under 1 atm
of CO over 50 h at 80 °C. Complete inhibition of migratory
insertion is observed in the presence of tetrabutylammonium
acetate, with immediate formation of 4-cis and 4-trans and no
NMR (25 °C); based on standard deviation of two trials in CD3OD/
DCE (8:2), the expected uncertainty for other conditions is 10%.
Half-life (t1/2) is the time to 50% conversion based on an exponential
fit of the decaying signal for the Ir−methyl complex (first ∼35%
detectable [8][BArF ]. Whereas these data show that acetate
b
4
conversion; see the Experimental Section for details). The inverse-
binds strongly to iridium in acetonitrile (acetate dissociation is
unfavorable), acetate dissociation to produce cationic iridium
species is much more facile in methanol. In fact, [7][OAc]
formed in situ in methanol under CO has almost identical
gated 31P{1H} NMR integrals for each methyl species disappearing
was summed to a single integral and plotted to obtain a weighted
c
average half-life. 5-cis is insoluble in MeOH only.
spectral features to [7][BArF ] in methanol. Formation of the
4
cationic dicarbonyl complex enables rapid migratory insertion.
Comparing Acetate and Iodide Ligands in CO
Insertion and Methyl Acetate Formation. Little is
known about how migratory insertion and organic acetyl
liberation will change based on the presence of iodide or
acetate ligands, but differences in reactivity in these later steps
of the proposed catalytic cycle could be important in iodide-
free carbonylation processes. In fact, there is relatively little
mechanistic information about any reductive elimination
containing 16 mM Ir were prepared in the glovebox, charged
with 1 atm of CO, and heated at 65 °C. The reaction progress
1
was followed by H and 31P{1H} NMR spectroscopy. Because
the Ir iodide and acetate complexes establish an equilibrium
mixture of cis/trans isomers and the dicarbonyl cation [7]+
under CO in MeOH, the total amount of methyl species was
used to evaluate the half-life under pseudo-first-order
conditions (see the Supporting Information for details).
G
https://doi.org/10.1021/jacs.1c05185
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
Figure 5. Kinetics of CO insertion and reductive elimination with (A) 5-cis, (B) 4-cis, and (C) [7][BArF ] in CD3OD/DCE (8:2) solution.
4
Consumption of sum of the methyl species (blue circles) and yields of the final carbonyl species (red squares) and the acetyl intermediate (green
triangles) is shown.
Figure 6. Free energy landscape of carbonylation of iridium−methyl compounds comparing effect of anion and solvent.
The iodide species (a mixture of 5-cis, 5-trans, and [7][I])
were consumed with t1/2 = 250 min, as the two Ir carbonyl
products 6 and 9 appeared (Figure 5A). Only a small amount
of an Ir acetyl intermediate (31P{1H} NMR δ 140.46) was
present during the reaction. This suggests a two-step sequence
in which the initial migratory insertion is the rate-determining
step.
120 min (Figure 5B). The reaction forms large amounts of an
1
Ir acetyl intermediate (31P{1H} NMR δ 141.92, H NMR δ
1.81), before giving way to the product 6, acetic acid, and
methyl acetate after prolonged heating (vide supra). Here,
reductive elimination of methyl acetate is the rate-determining
step, but the distinct rates of each step enable independent
kinetic analysis of the initial migratory insertion step.
The acetate species (a mixture of 4-cis, 4-trans, and
The slower rate of migratory insertion of iodide complex 5-
[7][OAc]) were consumed at a significantly faster rate, t1/2
=
cis relative to acetate complex 4-cis is consistent with the lesser
H
https://doi.org/10.1021/jacs.1c05185
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
degree of iodide dissociation relative to acetate dissociation
observed for this complex, which limits access to the needed
cationic intermediate [7]+ for CO insertion. The trend is
opposite for the reductive elimination step, however, with the
iodide complex supporting faster methyl acetate formation.
The difference could be due to a lower barrier kinetic pathway
for formation of acetyl iodide as an intermediate that reacts
with methanol to produce methyl acetate, as is typically
proposed in the Cativa process.1−5 The labeling study above
(see Scheme 3) established that reactions starting from acetate
complex 4-cis produce methyl acetate directly via coupling of
methanol and the acetyl. Consistent with this hypothesis,25 the
reaction of 4-cis proceeded faster as the methanol content was
increased (Table 1), entering a regime where the second step
starts to influence the rate of methyl−iridium complex
conversion, reaching a maximum in pure CD3OD, t1/2 = 69
min. No such methanol promotion is observed for the iodide
complex, which is more consistent with an iodide/acetyl
reductive elimination pathway.
CONCLUSIONS
■
An iodide-free carbonylation reaction sequence is reported,
based on net C−O bond activation of methyl acetate by a
pincer iridium(I) complex followed by CO insertion and
formation of acetic acid and another equivalent of methyl
acetate (Scheme 5). The net reaction is methanol carbon-
Scheme 5. Summary of the Carbonylation Reaction via
Methyl Acetate C−O Activation
If the faster rate of migratory insertion of acetate complex 4-
cis is due to accessing a cationic dicarbonyl intermediate, the
cationic species with a BArF4 counteranion, [7][BArF ], should
4
exhibit similar kinetics. As shown in Figure 5C, [7][BArF ] was
4
consumed at almost the same rate as 4-cis, consistent with our
prediction (t1/2 of 140 min). This is consistent with migratory
insertion from similar cationic species [7]+ in both cases.
Surprisingly, however, no acetyl intermediate was observed,
and the reaction promptly generated Ir carbonyl product 6.
Methanol is again indicated to be the nucleophile, based on
faster rates with higher methanol content (Table 1). The faster
rate of reductive elimination in [7][BArF ] may be rationalized
4
by the lack of coordinating anion: DFT calculations suggest
the acetyl intermediate derived from 4-cis rebinds acetate,
which would slow down reductive elimination relative to the
ylation to acetic acid, with methyl acetate acting as a
methylating promoter (and with additional conversion of an
iridium(I)−dinitrogen complex to an iridium(I)−carbonyl
complex).
cationic acetyl derived from [7][BArF ] (Figure 6). Acetate
4
binding trans to the acetyl ligand is ca. 8 kcal/mol more
favorable than acetate binding trans to the methyl ligand
(Tables S12 and S14), suggesting that while acetate
dissociation to form a cationic methyl complex is accessible,
rebinding of acetate after migratory insertion may inhibit
methyl acetate formation.
The carbonylation sequence provides a unique opportunity
to understand how iodide and acetate influence various
individual steps relevant to carbonylation catalysis. The CO
migratory insertion is strongly solvent dependent. In methanol,
acetate complexes undergo fast migratory insertion, attributed
to facile acetate substitution by CO to form a cationic
dicarbonyl intermediate that facilitates CO insertion. In
dichloromethane or acetonitrile, however, CO migratory
insertion was not observed, presumably due to unfavorable
acetate substitution by CO. The acetate complex undergoes
migratory insertion more than twice as fast as the iodide
complex in methanol. The faster rate of C−C bond formation
with an acetate ligand is attributed to more favorable formation
of a cationic methyl dicarbonyl intermediate by acetate
substitution relative to iodide substitution in methanol.
Experimental data on acetyl reductive elimination are
particularly lacking for iridium-catalyzed carbonylation.3,25−27
Here, we find evidence for distinct reductive elimination
pathways in the presence of acetate and iodide. In methanol
solvent, iodide complexes undergo faster elimination than
acetate complexes, and the rate does not increase with higher
concentrations of methanol. This suggests a direct reductive
elimination of acetyl iodide occurs first, followed by reaction
with the solvent methanol to generate methyl acetate. In
iodide-free conditions, the solvent methanol is the reductive
elimination partner, directly generating methyl acetate in a
Figure 6 summarizes the anion and solvent effects on CO
migratory insertion and reductive elimination. Methanol
solvation promotes acetate dissociation from 4-cis to produce
a cationic methyl species [7]+ that has a much lower activation
barrier for CO migratory insertion than the neutral pathway.
Iodide ions inhibit migratory insertion according to the same
principles due to preferential halide association to the iridium
center that inhibits access to key intermediate [7]+. Methanol
solvent is also essential for methyl acetate reductive
elimination, via either a concerted inner sphere mechanism
producing acetyl iodide that reacts with methanol or an outer-
sphere nucleophilic addition mechanism. The acetate complex
is proposed to undergo outer-sphere reductive elimination by
methanol addition to a cationic acetyl intermediate, as
indicated by labeling studies and the observation that methyl
acetate is produced more rapidly with [7][BArF ]. In the
4
presence of iodide, methyl acetate formation is faster, which we
attribute to accessing an inner sphere C−I reductive
elimination mechanism.
I
https://doi.org/10.1021/jacs.1c05185
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
(2) Kalck, P.; Le Berre, C.; Serp, P. Recent Advances in the
Methanol Carbonylation Reaction into Acetic Acid. Coord. Chem. Rev.
2020, 402, 213078.
(3) Morris, G. Carbonylation of Methanol to Acetic Acid and
Methyl Acetate to Acetic Anhydride. In Mechanisms in Homogeneous
Catalysis; Wiley-VCH Verlag GmbH & Co. KGaA: Weinheim, FRG,
2005; Vol. 54, pp 195−230.
(4) Beller, M.; Steinhoff, B. A.; Zoeller, J. R.; Cole-Hamilton, D. J.;
Drent, E.; Wu, X.; Neumann, H.; Ito, S.; Nozaki, K. Carbonylation. In
Applied Homogeneous Catalysis with Organometallic Compounds; Wiley:
Cambridge, 2017; pp 91−190.
(5) Maitlis, P. M.; Haynes, A.; James, B. R.; Catellani, M.; Chiusoli,
G. P. Iodide Effects in Transition Metal Catalyzed Reactions. Dalton
Trans. 2004, 3409−3419.
(6) Haynes, A. Acetic Acid Synthesis by Catalytic Carbonylation of
Methanol. In Catalytic Carbonylation Reactions; Springer: Berlin, 2006;
Vol. 18, pp 179−205.
(7) Haynes, A.; Maitlis, P. M.; Morris, G. E.; Sunley, G. J.; Adams,
H.; Badger, P. W.; Bowers, C. M.; Cook, D. B.; Elliott, P. I. P.;
Ghaffar, T.; Green, H.; Griffin, T. R.; Payne, M.; Pearson, J. M.;
Taylor, M. J.; Vickers, P. W.; Watt, R. J. Promotion of Iridium-
Catalyzed Methanol Carbonylation: Mechanistic Studies of the Cativa
Process. J. Am. Chem. Soc. 2004, 126 (9), 2847−2861.
(8) Bagno, A.; Bukala, J.; Olah, G. A. Chemistry in Superacids. 8.
Superacid-Catalyzed Carbonylation of Methane, Methyl Halides,
Methyl Alcohol, and Dimethyl Ether to Methyl Acetate and Acetic
Acid. J. Org. Chem. 1990, 55 (14), 4284−4289.
(9) Wegman, R. W. Vapour Phase Carbonylation of Methanol or
Dimethyl Ether with Metal-Ion Exchanged Heteropoly Acid Catalysts.
J. Chem. Soc., Chem. Commun. 1994, 947.
(10) Ellis, B.; Howard, M. J.; Joyner, R. W.; Reddy, K. N.; Padley, M.
B.; Smith, W. J. Heterogeneous Catalysts for the Direct, Halide-Free
Carbonylation of Methanol. Stud. Surf. Sci. Catal. 1996, 101, 771−
779.
(11) Blasco, T.; Boronat, M.; Concepción, P.; Corma, A.; Law, D.;
Vidal-Moya, J. A. Carbonylation of Methanol on Metal-Acid Zeolites:
Evidence for a Mechanism Involving a Multisite Active Center. Angew.
Chem., Int. Ed. 2007, 46 (21), 3938−3941.
(12) Boronat, M.; Martínez-Sánchez, C.; Law, D.; Corma, A.
Enzyme-like Specificity in Zeolites: A Unique Site Position in
Mordenite for Selective Carbonylation of Methanol and Dimethyl
Ether with CO. J. Am. Chem. Soc. 2008, 130 (48), 16316−16323.
(13) Ni, Y.; Shi, L.; Liu, H.; Zhang, W.; Liu, Y.; Zhu, W.; Liu, Z. A
Green Route for Methanol Carbonylation. Catal. Sci. Technol. 2017, 7
(20), 4818−4822.
reaction that is proceeds faster when the concentration of
methanol increases.
The major current limitation is that the reaction is not
catalytic because (a) the resulting iridium(I)−carbonyl
complex does not readily react with MeOAc to re-form a
methyl−iridium complex and (b) the solvents for methylation
and acetylation are incompatible (Scheme 5). Further work is
needed to find a system capable of facile activation of methyl
acetate in the presence of CO and reaction conditions that can
support all the elementary steps of the carbonylation process.
Our observations of individual steps in an iodide-free
carbonylation scheme may aid in future development of
iodide-free carbonylation catalysts and provide rare insight into
reductive elimination processes that furnish organic acetyls.
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge at
https://pubs.acs.org/doi/10.1021/jacs.1c05185.
■
sı
Experimental details and characterization data (PDF)
DFT coordinates (XYZ)
Accession Codes
CCDC 2084756−2084757 contain the supplementary crys-
tallographic data for this paper. These data can be obtained
free of charge via www.ccdc.cam.ac.uk/data_request/cif, or by
emailing data_request@ccdc.cam.ac.uk, or by contacting The
Cambridge Crystallographic Data Centre, 12 Union Road,
Cambridge CB2 1EZ, UK; fax: +44 1223 336033.
AUTHOR INFORMATION
Corresponding Author
■
Alexander J. M. Miller − Department of Chemistry, University
of North Carolina at Chapel Hill, Chapel Hill, North
Carolina 27599-3290, United States; Email: ajmm@
email.unc.edu
Author
Changho Yoo − Department of Chemistry, University of
North Carolina at Chapel Hill, Chapel Hill, North Carolina
27599-3290, United States; Green Carbon Research Center,
Korea Research Institute of Chemical Technology (KRICT),
Daejeon 34114, Republic of Korea
(14) Meng, X.; Yuan, L.; Guo, H.; Hou, B.; Chen, C.; Sun, D.;
Wang, J.; Li, D. Carbonylation of Methanol to Methyl Acetate over
Cu/TiO2-SiO2 Catalysts: Influence of Copper Precursors. Mol. Catal.
2018, 456, 1−9.
Complete contact information is available at:
https://pubs.acs.org/10.1021/jacs.1c05185
(15) Qi, J.; Christopher, P. Atomically Dispersed Rh Active Sites on
Oxide Supports with Controlled Acidity for Gas-Phase Halide-Free
Methanol Carbonylation to Acetic Acid. Ind. Eng. Chem. Res. 2019, 58
(28), 12632−12641.
(16) Tong, C.; Zhang, J.; Chen, W.; Liu, X.; Ye, L.; Yuan, Y.
Combined Halide-Free Cu-Based Catalysts with Triple Functions for
Heterogeneous Conversion of Methanol into Methyl Acetate. Catal.
Sci. Technol. 2019, 9 (21), 6136−6144.
(17) Zoeller, J. R.; Moore, M. K.; Vetter, A. J. Carbonylation
Process. US7582792, 2009.
(18) Ittel, S. D.; Tolman, C. A.; English, A. D.; Jesson, J. P. The
Chemistry of 2-Naphthyl Bis[Bis(Dimethylphosphino)Ethane] Hy-
dride Complexes of Fe, Ru, and Os. 2. Cleavage of sp and sp3 C-H, C-
O, and C-X Bonds. Coupling of Carbon Dioxide and Acetonitrile. J.
Am. Chem. Soc. 1978, 100 (24), 7577−7585.
(19) Trovitch, R. J.; Lobkovsky, E.; Bouwkamp, M. W.; Chirik, P. J.
Carbon-Oxygen Bond Cleavage by Bis(Imino)Pyridine Iron Com-
pounds: Catalyst Deactivation Pathways and Observation of Acyl C-O
Bond Cleavage in Esters. Organometallics 2008, 27 (23), 6264−6278.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
Eastman Chemical Co. is gratefully acknowledged for financial
support. The authors thank Faraj Hasanayn (American
University Beirut) for assistance with computational studies
and Josh Chen (UNC-CH) for assistance with X-ray
crystallography. The NMR spectroscopy was supported by
the National Science Foundation under Grants CHE-1828183
and CHE-0922858. The mass spectrometry was supported by
the National Science Foundation under Grant CHE-1726291.
REFERENCES
■
(1) Le Berre, C.; Serp, P.; Kalck, P.; Torrence, G. P. Acetic Acid. In
Ullmann’s Encyclopedia of Industrial Chemistry; Wiley-VCH Verlag
GmbH & Co. KGaA: Weinheim, Germany, 2014; Vol. 10, pp 1−34.
J
https://doi.org/10.1021/jacs.1c05185
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
(20) Manbeck, K. A.; Kundu, S.; Walsh, A. P.; Brennessel, W. W.;
Jones, W. D. Carbon-Oxygen Bond Activation in Esters by
Platinum(0): Cleavage of the Less Reactive Bond. Organometallics
2012, 31 (14), 5018−5024.
Alternative by-Product Formation Pathways in Catalytic Methanol
Carbonylation. Dalton Trans. 2013, 42 (47), 16538−16546.
(21) Kundu, S.; Choi, J.; Wang, D. Y.; Choliy, Y.; Emge, T. J.;
Krogh-Jespersen, K.; Goldman, A. S. Cleavage of Ether, Ester, and
Tosylate C(sp3)-O Bonds by an Iridium Complex, Initiated by
Oxidative Addition of C-H Bonds. Experimental and Computational
Studies. J. Am. Chem. Soc. 2013, 135 (13), 5127−5143.
(22) Gregor, L. C.; Grajeda, J.; White, P. S.; Vetter, A. J.; Miller, A. J.
M. M. Salt-Promoted Catalytic Methanol Carbonylation Using
Iridium Pincer-Crown Ether Complexes. Catal. Sci. Technol. 2018, 8
(12), 3133−3143.
(23) Gregor, L. C.; Grajeda, J.; Kita, M. R.; White, P. S.; Vetter, A. J.;
Miller, A. J. M. Modulating the Elementary Steps of Methanol
Carbonylation by Bridging the Primary and Secondary Coordination
Spheres. Organometallics 2016, 35 (17), 3074−3086.
(24) Yoo, C.; Dodge, H. M.; Farquhar, A. H.; Gardner, K. E.; Miller,
A. J. M. Decarbonylative Ether Dissection by Iridium Pincer
Complexes. Chem. Sci. 2020, 11 (44), 12130−12138.
(25) Matsumoto, T.; Mizoroki, T.; Ozaki, A. Mechanistic Study of
Methanol Carbonylation Catalyzed by an Iridium Complex in the
Presence of Methyl Iodide. J. Catal. 1978, 51 (1), 96−100.
(26) Forster, D. Kinetic and Spectroscopic Studies of the
Carbonylation of Methanol with an Iodide-Promoted Iridium
Catalyst. J. Chem. Soc., Dalton Trans. 1979, 1639−1645.
(27) Kinnunen, T.; Laasonen, K. Reaction Mechanism of the
Reductive Elimination in the Catalytic Carbonylation of Methanol. A
Density Functional Study. J. Organomet. Chem. 2001, 628 (2), 222−
232.
(28) Choi, J.; Choliy, Y.; Zhang, X.; Emge, T. J.; Krogh-Jespersen,
K.; Goldman, A. S. Cleavage of sp3 C-O Bonds via Oxidative Addition
of C-H Bonds. J. Am. Chem. Soc. 2009, 131 (43), 15627−15629.
(29) Haibach, M. C.; Lease, N.; Goldman, A. S. Catalytic Cleavage
of Ether C-O Bonds by Pincer Iridium Complexes. Angew. Chem., Int.
Ed. 2014, 53 (38), 10160−10163.
(30) Chapp, S. M.; Schley, N. D. Evidence for Reversible
Cyclometalation in Alkane Dehydrogenation and C-O Bond Cleavage
at Iridium Bis(Phosphine) Complexes. Organometallics 2017, 36 (22),
4355−4358.
(31) Rybtchinski, B.; Vigalok, A.; Ben-David, Y.; Milstein, D. A
Room Temperature Direct Metal Insertion into a Nonstrained
Carbon-Carbon Bond in Solution. C-C vs C-H Bond Activation. J.
Am. Chem. Soc. 1996, 118 (49), 12406−12415.
(32) van der Boom, M. E.; Liou, S.; Ben-David, Y.; Gozin, M.;
Milstein, D. Carbon-Carbon Bond Activation by Rhodium(I) in
Solution. Comparison of sp2-sp3 vs sp3-sp3 C-C, C-H vs C-C, and Ar-
CH3 vs Ar-CH2CH3 Activation. J. Am. Chem. Soc. 1998, 120 (51),
13415−13421.
(33) Rybtchinski, B.; Milstein, D. Metal Insertion into C-C Bonds in
Solution. Angew. Chem., Int. Ed. 1999, 38 (7), 870−883.
(34) Pearson, J. M.; Haynes, A.; Morris, G. E.; Sunley, G. J.; Maitlis,
P. M. Dramatic Acceleration of Migratory Insertion in [MeIr-
(CO)2I3]− By Methanol and by Tin(II) Iodide. J. Chem. Soc., Chem.
Commun. 1995, 1045.
(35) West, N. M.; Miller, A. J. M.; Labinger, J. A.; Bercaw, J. E.
Homogeneous Syngas Conversion. Coord. Chem. Rev. 2011, 255 (7−
8), 881−898.
(36) Lassauque, N.; Davin, T.; Nguyen, D. H.; Adcock, R. J.;
Coppel, Y.; Le Berre, C.; Serp, P.; Maron, L.; Kalck, P. Direct
Involvement of the Acetato Ligand in the Reductive Elimination Step
of Rhodium-Catalyzed Methanol Carbonylation. Inorg. Chem. 2012,
51 (1), 4−6.
(37) Hanh Nguyen, D.; Lassauque, N.; Vendier, L.; Mallet-Ladeira,
S.; Le Berre, C.; Serp, P.; Kalck, P. Reductive Elimination of
Anhydrides from Anionic Iodo Acetyl Carboxylato Rhodium
Complexes. Eur. J. Inorg. Chem. 2014, 2014 (2), 326−336.
(38) Elliott, P. I. P.; Haak, S.; Meijer, A. J. H. M.; Sunley, G. J.;
Haynes, A. Reactivity of Ir(III) Carbonyl Complexes with Water:
K
https://doi.org/10.1021/jacs.1c05185
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX