Photochemical and time resolved spectroscopic studies of intermediates relevant to iridium-catalyzed methanol carbonylation: Photoinduced CO migratory insertion

Article abstract of DOI:10.1021/ic0513033

Photoreaction, time-resolved infrared (TRIR), and DFT studies were utilized to probe transformations between iridium complexes with possible relevance to the mechanisms of the iridium/iodide-catalyzed methanol carbonylation to acetic acid. Solution-phase

Full text of DOI:10.1021/ic0513033

Inorg. Chem. 2006, 45, 1861−1870
Photochemical and Time Resolved Spectroscopic Studies of
Intermediates Relevant to Iridium-Catalyzed Methanol Carbonylation:
Photoinduced CO Migratory Insertion
Maurizio Volpe,† Guang Wu,† Alexei Iretskii,†,‡ and Peter C. Ford*,†
Department of Chemistry and Biochemistry, UniVersity of California,
Santa Barbara, California 93106-9510, and Department of Chemistry,
Lake Superior State UniVersity, 650 West Easterday AVenue, Sault Sainte Marie, Michigan 49783
Received October 20, 2005
Photoreaction, time-resolved infrared (TRIR), and DFT studies were utilized to probe transformations between
iridium complexes with possible relevance to the mechanisms of the iridium/iodide-catalyzed methanol carbonylation
to acetic acid. Solution-phase continuous and laser flash photolysis of the tetraphenylarsonium salt of the fac-
[CH3Ir(CO)2I3]- anion (1a) under excess carbon monoxide resulted in migratory insertion to give the acyl complex
+
ion mer,trans-[Ir(C(O)CH3)(CO)2I3]- (2a). The latter was isolated as its AsPh4 salt, and its X-ray crystal structure
was determined. TRIR spectra indicate that several transients are generated upon flash photolysis of 1a. The
principal photoreaction is CO dissociation, and this is proposed to generate the isomeric complexes fac-[CH3Ir-
(CO)(Sol)I3]- (ICO(fac), Sol
regenerate 1a with a second-order rate constant (kCO
(mer) is the apparent precursor to 2a. Kinetics studies indicate the photoinduced formation of a third intermediate
(IM), hypothesized to be the anionic acyl complex fac-[Ir(C(O)CH3)(CO)(Sol)I3]-. In the absence of added CO,
these intermediates undergo dimerization to form a mixture of isomers with the apparent formula [Ir(C(O)CH3)(CO)I3]2
One of these dimers was isolated as the AsPh4 salt, and the crystal structure was determined. Addition of excess
pyridine to a solution of the dimers gave the neutral complex mer,trans-[Ir(C(O)CH3)(CO)(py)2I2], which was
characterized by FTIR, NMR, and X-ray crystallography. These transformations, especially the unprecedented
photoinduced CO insertion reaction, are discussed and interpreted in terms of the factors favoring migratory insertion
dynamics.
)
solvent) and mer,trans-[CH3Ir(CO)(Sol)I3]- (ICO(mer)). ICO(fac) reacts with CO to
1
1
)
∼
2.5 × -
107 M- s- in ambient dichloroethane, while ICO
2-
.
+
Introduction
“anionic” cycles.3,4 Quantitative investigations point to the
“anionic” cycle illustrated by Scheme 1 as being predominant
under conditions of commercial catalysis.3,4
A key species in Scheme 1 is the methyl iridium complex
fac-[CH3Ir(CO)2I3]- (1a), the resting state of this cycle.
According to the accepted mechanism, the steps leading to
The carbonylation of methanol to acetic acid by a rhodium/
iodide catalyst (Monsanto process) is one of the most
successful industrial applications of homogeneous catalysis.1
More recently, a methanol carbonylation process based on
a promoted iridium/iodide catalyst (Cativa process) has been
commercialized.2 The catalytic mechanism that has been
outlined for this process involves participation of various
iodo iridium carbonyl species in interlocking “neutral” and
(3) Forster, D. J. Chem. Soc., Dalton Trans. 1979, 1639-1645.
(4) (a) Bassetti, M.; Monti, D.; Haynes, A.; Pearson, J. M.; Stanbridge, I.
A.; Maitlis, P. M. Gazz. Chim. Ital. 1992, 122, 391-393. (b) Pearson,
J. M.; Haynes, A.; Morris, G. E.; Sunley, G. J.; Maitlis, P. M. J. Chem.
Soc., Chem. Commun. 1995, 1045-1046. (c) Maitlis, P. M.; Haynes,
A.; Sunley, G. J.; Howard, M. J. Chem. Soc., Dalton Trans. 1996,
2187-2196. (d) Ghaffar, T.; Adams, H.; Maitlis, P. M.; Sunley, G.
J.; Baker, M. J.; Haynes, A.; Howard, M. Chem. Commun. 1998,
1023-1024. (e) 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. J. Am. Chem. Soc. 2004,
126, 2847-2861.
* To whom correspondence should be addressed. E-mail:
ford@chem.ucsb.edu.
† University of California, Santa Barbara.
‡ Lake Superior State University.
(1) Parshall, G. W.; Ittel, S. D. Homogeneous Catalysis; John Wiley and
Sons: New York, 1992.
(2) (a) Chem. Br. 1996, 32, 7. (b) Sunley, G. J.; Watson, D. J. Catal.
Today. 2000, 58, 293-307.
10.1021/ic0513033 CCC: $33.50
Published on Web 01/10/2006
© 2006 American Chemical Society
Inorganic Chemistry, Vol. 45, No. 4, 2006 1861

Volpe et al.
Scheme 1. Anionic Cycle of Proposed Mechanism for Ir/I-Catalyzed
Methanol Carbonylation
fused to a quartz cell, in a CaF2 IR cell sealed with serum caps, or
in the flow system described below. The sample concentration for
the measurement was typically 1-3 mM except for bulk photolysis
reactions where higher concentrations were used. The solutions were
deaerated by entrainment with argon or carbon monoxide with three
freeze-pump-thaw cycles then equilibrated under Ar or CO
(various PCO).
The time-resolved infrared spectrometer and its applications to
the reactions of metal carbonyl intermediates in migratory insertion
reactions have been described previously.7 In the current config-
uration, the IR probe is generated by a lead salt diode laser with a
single longitudinal mode selected by a CVI Digikrom Model 240
monochromator. The probe beam was overlapped at the plane of
the infrared cell with a 355 nm pump pulse from a Lumonics
HY600 Nd:YAG laser. The high-pressure/variable-temperature (HP/
VT) flow system assembled at UCSB7g allowed for the equilibration
of sample solutions for time-resolved infrared (TRIR) experiments
with various pressures of CO at the desired temperatures and
provided a steady flow of fresh sample solution to a CaF2 IR cell
for photolysis experiments.
Continuous photolysis experiments were conducted using stan-
dard procedures8 in a darkroom. The excitation source was a 200
W high-pressure mercury lamp mounted on an Oriel optical train
(UVP) with intensity regulation by an internal feedback loop. The
light from this source was first passed through an IR filter, followed
by neutral density filters to reduce intensity, and an interference
filter to isolate the mercury line at 366 nm. A lens was used to
focus the light onto the sample, and the solution was stirred
continuously. Ferrioxalate actinometry was used to evaluate light
intensities.8a,b The sample was irradiated for the specific period,
and the reaction progress was monitored by UV-visible and FTIR
spectroscopy. Quantum yields were determined by fitting plots of
incremental quantum yields, Φi (first 10-30% of the reaction), for
absorbance changes vs percent reaction with a linear equation.8c,d
The y intercept is the photoreaction quantum yield, Φ. Experiments
were carried out multiple times to ensure reproducibility.
Bulk photolysis reactions were carried out by using standard
gastight Schlenk flasks or gastight NMR tubes for experiments in
deuterated solvent. The apparatus used is an ACE Glass photo-
chemical safety UV cabinet model 7836-20 and a 200 W mercury
lamp immersed in a quartz well with flowing water for cooling
and for IR filtering.
CO migratory insertion to give an acyl intermediate are rate
limiting.4a,c In nonpolar media, this transformation is dra-
matically accelerated by added methanol and Lewis acids,
is strongly inhibited by iodide salts such as [Bu4N]I,4b and
is promoted by neutral ruthenium iodo carbonyls and main-
group iodides (e.g., InI3, GaI3, SnI2, and ZnI2).4e These
observations were interpreted in terms of 1a being activated
by iodide scavengers that promote transformation under CO
to the neutral fac-tricarbonyl complex [CH3Ir(CO)3I2], which
undergoes methyl migration and I- addition to form the acyl
complex ion fac-[Ir(C(O)CH3)(CO)2I3]- (2b). This reactivity
has been attributed to enhanced electrophilicity of the cis
carbonyl, owing to competition for metal d electrons from
the other COs.4d,e,5 Reductive elimination from 2b and
hydrolysis of the resulting acetyl iodide leads to acetic acid
formation.
The goal of the present study was to utilize photochemical
techniques to generate and time-resolved spectroscopy to
interrogate intermediates possibly relevant to methanol
carbonylation promoted by iridium salts. In the course of
this investigation, we have demonstrated that irradiation of
1a promotes an unprecedented photoinduced CO migratory
insertion to give the stable acetyl complex ion mer,trans-
[Ir(C(O)CH3)(CO)2I3]- (2a), which was isolated as the
AsPh4+ salt and structurally characterized. Here we describe
the characterization of various transient and stable species
formed under various conditions and discuss the possible
relevance of these with regard to catalytic pathways involving
iridium iodo carbonyl complexes.
(5) (a) Similar enhanced electrophilicity of coordinated CO was used to
interpret [CO] effects on the water gas shift catalytic activity of
rhodium halo carbonyl complexes in aqueous amine solutions.5b,c (b)
Lima Neto, B. S.; Ford, K. H.; Pardey, A. J.; Rinker, R. J.; Ford, P.
C. Inorg. Chem. 1991, 30, 3837-3842. (c) Ford, P. C.; Rockicki, A.
AdV. Organomet. Chem. 1988, 28, 139-217.
(6) Riddick, J. A.; Bunger, W. B.; Sakano, T. K. Organic SolVents Physical
Properties and Methods of Purification, 4th ed.; John Wiley and
Sons: New York, 1986; Vol. II.
(7) (a) Di Benedetto, J. A.; Ryba, D. W.; Ford, P. C. Inorg. Chem. 1989,
28, 3503-3507. (b) Boese, W. T.; Ford, P. C. J. Am. Chem. Soc.
1995, 117, 8381-8391. (c) Massick, S.; Rabor, J.; Elbers, S.;
Marhenke, J.; Bernhard, S.; Schoonover, J.; Ford, P. C. Inorg. Chem.
2000, 39, 3098-3106. (d) Massick, S. M.; Mertens, V.; Jon Marhenke,
J.; Ford, P. C. Inorg. Chem., 2002, 41, 3553-3559. (e) Massick, S.
M.; Bu¨ttner, T.; Ford, P. C. Inorg. Chem. 2003, 42, 575-580. (f) Ford,
P. C.; Massick, S. M. Coord. Chem. ReV. 2002, 226, 39-46. (g)
Massick, S. M.; Ford, P. C. Organometallics 1999, 18, 4362-4366.
(8) (a) Calvert, J. G.; Pitts, J. N. Photochemistry; John Wiley and Sons:
New York, 1967; pp 783-786. (b) Malouf, G.; Ford, P. C. J. Am.
Chem. Soc. 1977, 99, 7213-7221. (c) Chaisson, D. A.; Hintze, R. E.;
Stuemer, D. H.; Petersen, J. D. Mcdonald, D. P.; Ford, P. C. J. Am.
Chem. Soc. 1972, 94, 6665-6673. (d) Works, C. F.; Jocher, C. J.;
Bart, G. D.; Bu, X.; Ford, P. C. Inorg. Chem. 2002, 41, 3728-3736.
(9) (a) Forster, D. Inorg. Nucl. Chem. Lett. 1969, 5, 433-436. (b) Forster,
D. Inorg. Chem,. 1972, 11, 473-475.
Experimental Section
Materials. Iridium trichloride hydrate, tetraphenylarsonium
chloride, and iodomethane were purchased from Aldrich and used
without further purification. The solvents used in these experiments
(1,2-dichoroethane (DCE), dichloromethane, acetonitrile, and tet-
rahydrofuran (THF)) were purified and dried following standard
distillation procedures.6 Deuterated acetonitrile and dichloromethane
were purchased from Cambridge Isotope Laboratories and used
without any further purification. Gases were obtained from Praxair.
Photolysis Procedures. The solutions studied by continuous and
laser flash photolysis techniques were contained in a gastight flask
1862 Inorganic Chemistry, Vol. 45, No. 4, 2006

Intermediates in Iridium-Catalyzed Methanol Carbonylation
Table 1. Parameters of X-ray Diffraction Data Collections and
Instrumentation. UV-visible spectra were recorded using a
Hewlett-Packard diode array spectrophotometer or a Shimadzu
model 2401PC spectrophotometer. Infrared spectra were obtained
using a BioRad model FTS 60 SPC 3200 FTIR spectrometer. ESI-
MS were acquired on a VG Fision Platform II single-quadrupole
mass spectrometer with an electrospray ionization source and a
Fision Masslink data system. All samples were diluted in HPLC
grade acetonitrile immediately prior to measurement. NMR spectra
were registered on Varian 200 and 400 MHz spectrometers in CD2-
Cl2 (5.32 ppm) or CD3CN (1.94 ppm). Solid-state MAS NMR
spectra were recorded at room temperature on a Bruker 300 MHz
Avance Spectrometer with a 4 mm broadband MAS probe double
tuned to 1H(300.1 MHz) and 13C(75.5 MHz). A spinning speed of
1 kHz was used in the experiments.
Structure Determinationsa
structure
formula
1
2
5
C
28H23O3AsI3Ir
C54H46O4As2I6Ir2 C13H13N2O2I2Ir
a (Å)
b (Å)
c (Å)
R(°)
â(°)
γ(°)
V(Å3)
Z
10.885(5)
13.069(6)
13.452(7)
63.893(7)
66.249(7)
85.134(8)
1563(1)
2
10.528(2)
12.052(2)
12.423(2)
71.549(2)
85.826(2)
70.565(2)
1409.2(3)
1
10.0999(4)
17.2639(7)
90
106.762(1)
90
1678.8(1)
4
fw
1055.28
2.242
80.311
triclinic P1h
dark brown
2054.54
2.421
92.11
triclinic P1h
orange
675.25
2.672
116.33
monoclinic P21/n
dark yellow
D
calc (g/cm3)
µ(cm-1
)
space group
color
size (mm3)
temp (K)
0.3 × 0.25 × 0.1 0.2 × 0.15 × 0.1 0.15 × 0.1 × 0.05
Syntheses. fac-[CH3Ir(CO)2I3][AsPh4] (AsPh4+ salt of 1a). This
compound was prepared from cis-[Ir(CO)2I2][Ph4As] according to
literature methods.4e,9 ESI-MS: 644.69 m/z (M-). IR (CH2Cl2) ν-
(CO) in cm-1: 2098 (ꢀ ) 1.9 × 103 M-1 cm-1), 2046 (1.3 × 103).
1H NMR (CD2Cl2): δ 2.1 (s, 3H, CH3), δ 7.6-7.9 (m, 20H, Ph4-
As). 13C NMR (CD2Cl2): δ 156.4 (CO), 135.5, 133.5, 131.9 (Ph,
CH), 121.0 (Ph, AsC), -16.0 (CH3).
293
110
117
Mo KR λ(Å)
0.71073
0.587
0.71073
0.670
0.71073
0.648
b
T
min/Tmax
θ range (°)
1.75 e θ e 28.25 2.29 e θ e 28.02 2.11 e θ e 28.71
-13 e h e 12
-16 e k e 17
-16 e l e 16
13 130/6788
-13 e h e 12
-15 e k e 15
-16 e l e 15
10 838/5980
-13 e h e 12
-12 e k e 13
-20 e l e 21
14 505/3981
total reflns/
unique reflms
Rint ) 0.0291
327
Rint ) 0.0306
309
Rint ) 0.0310
183
+
mer,trans-[Ir(C(O)CH3)(CO)2I3][AsPh4] (AsPh4 salt of 2a).
variables
R [I > 2σ(I)]
R1 ) 0.0316
wR2 ) 0.0772
1.043
R1 ) 0.0478
wR2 ) 0.1270
1.060
R1 ) 0.0237
wR2 ) 0.0559
1.028
fac-[CH3Ir(CO)2I3][AsPh4] (205 mg, 0.2 mM) was dissolved in 50
mL of deaerated CH2Cl2, and the solution equilibrated with CO at
atmospheric pressure for 1 h. The solution was then subjected to
bulk photolysis for 2 h until complete reaction of 1a was achieved
(the course of reaction was followed by FTIR and ESI-MS). The
volume of the solution was reduced to about 5 mL in vacuo, and
GOF
a Structure 1: The AsPh4+ salt of 2a: mer,trans-[Ir(C(O)CH3)(CO)2I3]-
[AsPh4]. Structure 2: The AsPh4+ salt of 4: [Ir(C(O)CH3)(CO)I3]2[AsPh4]2.
Structure 5: mer,trans-[Ir(C(O)CH3)(CO)(py)2I2]. b The transmission ratio
was from absorption correction using program SADABS.
+
crystallization of the AsPh4 salt of 2a achieved by layering the
Found: C, 23.44; H, 1.92; N, 3.99. IR (CH2Cl2) ν(CO) cm-1
:
CH2Cl2 solution with diethyl ether. The orange crystals were separ-
ated, washed with diethyl ether, and slowly dried under flowing
nitrogen. Yield: 140 mg (68%). Anal. Calcd for (C28H23O3IrI3As):
C, 31.87; H, 2.20. Found: C, 31.93; H, 2.10. ESI-MS: 672.70 m/z
(M-). IR (CH2Cl2) ν(CO) in cm-1: 2064 (ꢀ ) 2.3 × 103 M-1 cm-1),
1655 (7 × 102). 1H NMR (CD2Cl2): δ 2.78 (s, 3H, C(O)CH3), 7.6-
7.9 (m, 20H, AsPh4+). 13C NMR (CD2Cl2): δ 161.1 (CO), 199.3
(C(O)CH3), 135.5, 133.5, 131.9 (Ph, CH), 51.0 (C(O)CH3). A crys-
tal suitable for single-crystal X-ray diffraction was selected and
the structure determined. At convergence, R1 ) 0.0316 and GOF
) 1.043 for 327 variables refined against 6788 reflections with I
> 2σ.
1
2053vs, 1662s. H NMR (CD2Cl2): δ 2.9 (s, 3H, C(O)CH3); 7.3
(q), 7.6-7.9 (m), 8.8 (d) and 9.1 (d). 13C NMR (CD2Cl2): δ 197.0
(C(O)CH3), 157.0 (CO); 139.5, 126.2, 125.1 (Py, CH); 52.0 (C(O)-
CH3). A crystal suitable for single-crystal X-ray diffraction was
selected and the structure determined. At convergence, R1 ) 0.0237
and GOF ) 1.028 for 183 variables refined against 3981 reflections
with I > 2σ.
Crystal Structures. Crystallographic data were collected using
a Bruker CCD platform diffractometer. The SMART program
(Bruker AXS, Inc.) was used to determine the unit cell parameters
and data collection (30 s/frame, 0.3°/frame for a sphere of diffraction
data). The data were collected at room and low temperature. The
raw frame data were processed using the SAINT program (Bruker
AXS, Inc.). The absorption correction was applied using the
SADABS program (Bruker AXS, Inc). Subsequent calculations
were carried out using the SHELXTL program (Bruker AXS, Inc.).
The structure was solved by direct methods and refined on F2 by
full-matrix least-squares techniques. All hydrogen atoms were
located theoretically. The detailed information of data collections
and structure determinations are listed in Table 1. Other crystal
data and cell packing for the structures of mer,trans-[Ir(C(O)CH3)-
(CO)2I3][AsPh4], the dimer [Ir(C(O)CH3)(CO)I3]2][AsPh4]2, and
mer,trans-[Ir(C(O)CH3)(CO)(py)2I2] are presented in the Supporting
Information Figures S1-S3 and Tables S1-S3.
+
[Ir(C(O)CH3)(µ-I)(CO)I2]2[AsPh4]2 (AsPh4 salt of 4). fac-
[CH3Ir(CO)2I3][AsPh4] (205 mg, 0.2 mM) was dissolved in 25 mL
of acetonitrile under argon atmosphere. The solution was then
subjected to bulk photolysis until species 1a was no longer detected
by IR spectroscopy (about 2 h). The solvent was removed in vacuo,
and then the residue was dissolved in 10 mL of CH2Cl2 under argon.
By layering the CH2Cl2 solution with Et2O, dark orange-red crystals
were obtained. Anal. Calcd for (C54H46O4Ir2I6As2): C, 31.57; H,
2.26. Found: C, 31.50; H, 2.33. IR (KBr) ν(CO) cm-1: 2046sh,
2034vs, 2009s, 1670sh, 1650s, 1637sh; IR (CH2Cl2) ν(CO) cm-1
:
2094w, 2054sh, 2041vs, 1692wb, 1664wb. 13C NMR (solid state):
δ 159.7 (CO), 190.6 (C(O)CH3), 137.1, 133.8, (Ph, CH), 120.7,
(CAs), 46.5 (C(O)CH3). A crystal suitable for single-crystal X-ray
diffraction was selected and the structure determined. At conver-
gence, R1) 0.0478 and GOF ) 1.060 for 309 variables refined
against 5980 reflections with I > 2σ.
Computations. DFT calculations were performed at B3LYP
theory level using the LACVP* polarized basis set. Geometries
were optimized inside the Titan software suite (Wavefunction, Inc.).
Except where noted, optimizations were calculated as singlet ground
states due to significant HOMO-LUMO gaps (>3 eV), without
any symmetry constraints and represent structures in the gas phase.
Ir(C(O)CH3)(CO)(py)2I2] (5). Excess pyridine was added to a
freshly photolyzed acetonitrile solution of 1a leading to the
preparation of species 4 under argon. By layering the reaction
solution with Et2O, AsPh4I soon separated as white crystals. From
the remaining solution, bright orange crystals were obtained (few
days). Anal. Calcd for (C13H13O2IrN2I2): C, 23.12; H, 1.94; N, 4.15.
Results and Discussion
Photolysis Studies. Figure 1 displays the changes in the
IR spectrum of a DCE solution of the fac-[CH3Ir(CO)2I3]-
Inorganic Chemistry, Vol. 45, No. 4, 2006 1863

Volpe et al.
Figure 1. FTIR spectral changes (νCO region) for a DCE solution of the
+
AsPh4 salt of 1a (1.3 mM) under 1.0 atm of CO (298 K) as the result of
continuous photolysis at 366 nm. Successive spectra over the first 150 s
were recorded after 5 s intervals of irradiation in a CaF2 IR cell (0.1 cm
path length). (The band at 1969 cm-1 denotes the appearance of the cis-
[Ir(CO)2I2]- anion in the later stages.)
Figure 2. Structure of 2a (counterion and hydrogens omitted for clarity,
ellipsoid probability at 50%). Selected bond lengths (Å) are Ir-C(1) )
1.951(6), Ir-C(2) ) 1.911(7), Ir-C(3) ) 2.125(6), C(1)-O(1) ) 1.099(7),
C(2)-O(2) ) 1.126(7), C(3)-O(3) ) 1.146(7), C(3)-C(4) ) 1.440(8),
Ir-I(1) ) 2.8144(11), Ir-I(2) ) 2.6792(90), and Ir-I(3) ) 2.6902(10).
Selected angles (deg) are O(1)-C(1)-Ir ) 178.7(5), O(2)-C(2)-Ir )
175.0(5), O(3)-C(3)-Ir ) 118.5(4), and C(4)-C(3)-Ir ) 117.1(4).
+
anion (1a) (1.3 mM as the AsPh4 salt) during photolysis
under CO (1 atm) at ambient temperature using the excitation
wavelength (λirr) 366 nm. A 3D representation of these data
appears in the Supporting Information as Figure S4. In these
figures, the absorption bands of 1a at 2097 and 2043 cm-1
(ꢀ ) 1900 and 1300 M-1 cm-1, respectively) decrease
smoothly with concomitant growth of two new absorption
bands at 2062 and 1657 cm-1 (ꢀ ) 2300 and 700 M-1 cm-1,
respectively, determined from the isolated and well-
characterized product). Isosbestic points are clearly evident
at 2080 and 2051 cm-1.
involving the slow formation of the cis-[Ir(CO)2I2]- anion
(ν(CO) bands at 2044 and 1969 cm-1 in ambient DCE, eq
2). The net reaction is acetyl iodide reductive elimination,
although it is uncertain whether this photoinduced process
occurs directly from 2a, which is thermally stable under these
conditions, or from its isomer fac-[Ir(C(O)CH3)(CO)2I3]-
(2b) (see below).
On the basis of FTIR, NMR, and ESI-MS spectroscopic
analysis, the photoproduct was identified as the mer,trans-
[Ir(C(O)CH3)(CO)2I3]- anion (2a) (eq 1). The FTIR bands
at 1657 and 2062 cm-1 can be assigned, respectively, to the
stretching mode of the acyl carbonyl and antisymmetric
stretching modes of the trans terminal carbonyls. The
+
Photolysis of 1a as the AsPh4+ salt in ambient THF under
CO showed similar results but with a higher degree of
reductive elimination of acetyl iodide to yield cis-[Ir(CO)2I2]-
[IR (THF) ν(CO): 2037vs, 1959vs]. When 1a was photo-
lyzed in ambient acetonitrile under CO, additional species
showing νCO bands were observed at 2083w and 2045s cm-1
together with the formation of 2a (2055vs, 1648w,b cm-1)
structure of 2a was confirmed by isolating the AsPh4 salt
in a bulk photolysis experiment and determining the X-ray
crystal structure (Figure 2). The structure of the correspond-
ing fac-isomer 2b was recently reported.4e
-
and cis-Ir(CO)2I2 (2043s and 1969s cm-1).
The course of the photochemical reaction of 1a was also
1
studied by H NMR in CD3CN. After 15 min of photolysis
The quantum yield for appearance of the product 2a was
calculated by monitoring the strongest absorption changes
at 2062 cm-1 and plotting the incremental quantum yield,
Φi, versus time over the early stages of the reaction in DCE
solutions (see Experimental Section).8c,d Extrapolation of this
plot to the y intercept accounted for inner-filter effects (absor-
bance at 366 nm changed during experiment) and gave a
in an argon-filled NMR tube, some unreacted 1a (δ 2.100,
s, CH3), remained but about three times as much of an acetyl
species (δ 2.575, s, C(O)CH3) was present. When excess
13CO was added at ambient temperature, the mono-labeled
2a mer,trans-[Ir(C(O)CH3)(CO)(13CO)I3]- was observed and
characterized by NMR (1H NMR: δ 2.78 (s, 3H, C(O)CH3),
13C NMR: δ 161.1 (CO)) and by ESI-MS (673.70 m/z (M-)).
Thus, the spectroscopic data strongly suggest that, in the
presence of acetonitrile, a solvent-stabilized acyl anionic
complex [Ir(C(O)CH3)(CO)(AN)I3]- is generated photo-
chemically.
Φ
2a of 0.045 ( 0.002, reproducible over multiple measure-
ments. The quantum yield for disappearance of 1a (Φd) is
(within experimental uncertainty) the same as that for the
appearance of 2a, as is evident from the spectral changes in
Figure 1.
Continued photolysis (λirr ) 366 nm) of such solutions
under CO for >10 h revealed a secondary photolysis process
Continuous photolysis (λirr ) 366 nm) of 1a (2.5 mM) in
DCE solution under an inert atmosphere (Ar) led to different
IR spectral changes (Figure 3) than those seen under CO.
1864 Inorganic Chemistry, Vol. 45, No. 4, 2006

Intermediates in Iridium-Catalyzed Methanol Carbonylation
sharp, positive absorption at 2112 cm-1 that can be attributed
to formation of the fac,cis isomer 2b (eq 3). Similar reactions
with CO and other ligands have been reported for analogous
rhodium dimers.10
An analogous photolysis of 1a was carried out under argon
in a NMR tube with CD2Cl2 as the reaction solvent. The
1
product solution demonstrated two signals in the H NMR
Figure 3. FTIR spectra (νCO region) tracking the continuous photolysis
+
spectrum at 2.63 and 2.54 ppm in an approximate 1:1 ratio
consistent with the formation of acyl complexes, presumably
the dimers. After addition of excess 13CO to the photolyzed
solution, singly labeled 2a mer,trans-[Ir(C(O)CH3)(CO)(13CO)-
I3]- proved to be the only detectable isotopically enriched
species in the 13C NMR spectrum.
When excess pyridine was added to a CH2Cl2 solution of
the dimers generated under argon, the product was the neutral
bis(pyridine) complex 5 (eq 4). This was isolated in almost
(366 nm) of 1a as a AsPh4 salt (2.5 mM) in DCE solution under Ar (1
atm). (298 K, 0.2 cm path length, IR cell).
Within a few minutes, new IR absorption bands appeared
that overlapped the 2043 cm-1 band characteristic of 1a. In
addition, three broad and weaker bands between 1665 and
1708 cm-1 were detected. The quantum yield for the disap-
pearance of 1a under these conditions (0.072 ( 0.004) was
evaluated by monitoring the IR absorption changes at 2097
cm-1 and plotting Φi versus time to correct for inner-filter
effects.
Longer-term photolysis of acetonitrile solutions of the
AsPh4+ salt of 1a (2 h) under argon gave a solution for which
the IR spectrum displayed ν(CO) absorptions at 2100w,
2059sh, 2041s, 1708w,b, 1690w,b, and 1665w,b cm-1. These
bands were assigned to a mixture of the anionic acetyl
iridium dimers, [Ir(C(O)CH3)(µ-I)(CO)I2]22-, one of which,
quantitative yield, and the X-ray crystal structure was
determined (Figure 5). It is likely that the formation of 5
proceeds by breaking the weak bridging iodide bond trans
to acetyl group followed by substitution of the iodide trans
to CO by a second pyridine ligand. The rhodium analogue
of 5 was recently reported, but the structure was not
determined.10b
+
4, was isolated as a AsPh4 salt and the crystal structure
determined (Figure 4). The electron density map indicated
the carbonyl (C3, O2) and iodo (I1) ligands to have disorder,
which can be treated as two isomers (Figure 4a and b)
regarding the orientations of carbonyl and iodo in the axial
sites. By constraining the equal thermal parameters of C3,
O2, and I1 in two isomers, the refinement derived occupan-
cies of 81.8:18.2 for the two isomers.
When a solution of the dimers generated by photolysis
under argon in ambient dichloromethane was exposed to
excess CO, the resulting IR spectral changes indicated the
formation of the mononuclear mer,trans-[Ir(C(O)CH3)-
(CO)2I3]- anion 2a. Careful IR subtraction spectra analysis
during the reaction also revealed appearance of a weak, but
-
As noted above, the iridium(I) anion, cis-Ir(CO)2I2 , was
observed as a product in the long-term photolysis of DCE
solutions of 1a under CO, although the mer,trans-[Ir(C(O)-
CH3)(CO)2I3]- anion (2a) is the principal product formed
initially. In this context, DCE solutions were prepared from
+
the isolated AsPh4 salt of 2a (1.5 mM) in order to
investigate the photochemistry of this species directly and
-
to evaluate whether cis-Ir(CO)2I2 might result from the
Figure 4. (a) and (b). Structure of 4 (counterion and hydrogens omitted for clarity; ellipsoid probability at 50%). The observed disorder between the iodo
and carbonyl ligands in the axial sites is shown in (a) and (b). Selected bond lengths (Å) are Ir-C(1) ) 1.995(16), Ir-C(3A) ) 1.854(10), Ir-C(3B) )
2.07(5), C(1)-O(1) ) 1.183(10), C(3A)-O(2A) ) 1.061(9), C(3B)-O(2B) ) 1.089(10), C(1)-C(2) ) 1.364(10), Ir-I(1A) ) 2.642(4), Ir-I(1B)) 2.6949(10),
Ir-I(2) ) 2.6769(9), Ir-I(3) ) 2.6905(8), and Ir(1)-I(3) ) 2.8639(10). Selected angles (deg) are Ir-I(3)-Ir(1) ) 93.33(2), I(2)-Ir-I(3) ) 175.77(3),
C(3A)-Ir-I(1B) ) 177.7(4), C(1)-Ir-I(3) ) 94.5(4), C(1)-Ir-I(2) ) 89.1(4), O(1)-C(1)-Ir ) 119.1(15), and C(2)-C(1)-Ir ) 130.7(17).
Inorganic Chemistry, Vol. 45, No. 4, 2006 1865

Volpe et al.
Figure 7. Absorbance change at 2097 (DCE) and 2100 cm-1 (CH3CN)
following 355 nm laser flash photolysis of a 3 mM sample solution of 1a
as the AsPh4+ salt in the presence of CO 60 psig at 298 K. (The absorbance
changes shown are in mO.D. units.)
Figure 5. Structure of Ir(C(O)CH3)(CO)(py)2I2 (5). (Hydrogen atoms
omitted for clarity; ellipsoid probability at 50%) Selected bond lengths (Å)
are Ir-C(1) ) 2.048(4), Ir-C(3) ) 1.861(4), Ir-I(1) ) 2.6739(4), Ir-I(2)
) 2.6774(4), C(1)-O(1) ) 1.210(5), C(3)-O(2) ) 1.122(5), C(1)-C(2)
) 1.285(19), Ir-N(1) ) 2.255(3), and Ir-N(2) ) 2.146(3). Selected angles
(deg) are O(1)-C(1)-Ir ) 121.4(3), C(2)-C(1)-Ir ) 120.3(3), O(2)-
C(3)-Ir ) 178.5(4), I(1)-Ir-N(1) ) 91.00(9), I(1)-Ir-N(2) ) 91.71(9),
and N(2)-Ir-N(1) ) 85.28(12).
Figure 2) can be compared to that of the fac,cis isomer 2b,
recently reported by Haynes et al.4e The influence of the acyl
group is reflected in the longer Ir-I bond trans to acetyl
group. For each structure, this bond is longer (by ∼0.1 Å)
than the Ir-I bond cis to the acetyl group, which in 2a is
trans to another iodide but in 2b is trans to a carbonyl. The
trans influence of the acyl is also evident in the structure of
mer,trans-[Ir(C(O)CH3)(CO)(py)2I2] (5, Figure 5), which
shows the Ir-N bond trans to the acyl to be 0.1 Å longer
than that trans to the carbonyl.
The crystal structure determined for the dimeric anion [Ir-
2-
(C(O)CH3)(µ-I)(CO)I2]2 (4, Figure 4) shows disorder in
the axial sites for both Ir(III) centers. It is not clear whether
this indicates the presence of two closely related dimers that
have co-crystallized or is simply the disorder in the cellular
packing of a single species. (The refined structure included
∼18% disorder for the centro-symmetric isomer.) The former
is consistent with the observation that the IR spectrum of
the isolated solid displays two acetyl carbonyl νCO bands,
indicating the presence of at least two isomers in the isolated
solid. The disorder suggests that co-crystallization of species
with the axial CO oriented on the same side of the diiodo
bridges with another species having the COs on opposite
sides. In other respects, the structure of 4 resembles that of
an analogous dinuclear rhodium(III) compound.11 The iri-
dium atoms show a distorted octahedral arrangement, and
the two octahedra are linked together by a double iodide
bridge across an apparent center of symmetry. The iridium-
bridging iodide distances are 2.6905(9) and 2.8639(10) Å,
in line with the trans influence of the acyl group. The
variation in the Ir-iodide bridge distances is similar to that
observed for the Rh-iodide bridges in [Rh(C(O)CH3)(µ-I)-
(CO)I2]2[Me3PhN]2.11,12
Figure 6. FTIR spectra (νCO region) tracking the continuous photolysis
+
(366 nm) of 2a as a AsPh4 salt of (1.5 mM) under CO (1 atm) in DCE
(298 K): 0 and 2 h, solid lines; 16 h, dotted line. The bands attributed to
formation of 2b and to cis-[Ir(CO)2I2]- are indicated.
secondary photolysis of 2a. Irradiation of these solutions at
366 nm under CO (1.0 atm) at ambient temperature led to
the IR and UV-vis spectral changes reported in Figure 6
and in Supporting Information Figure S5, respectively. These
-
indicate that both cis-Ir(CO)2I2 and the fac-[Ir(C(O)CH3)-
(CO)2I3]- anionic isomer 2b, as well as acetyl iodide, are
formed; however, the net photoreactivity of 2a (Φ e 0.007)
is about an order of magnitude less than that of 1a. Although
the spectral changes are somewhat ambiguous, we speculate
-
that 2b is the likely predecessor of cis-Ir(CO)2I2 . Concerted
thermal or photochemical reductive elimination from 2a
-
should give the less-stable trans-Ir(CO)2I2 , although one
might speculate on other photoreaction scenarios.
Laser Flash Photolysis Studies. Figure 7 displays the
Comments on the X-ray Crystal Structures. The
structure of the mer,trans-[Ir(C(O)CH3)(CO)2I3]- anion (2a,
temporal absorbance of the 2097 cm-1
νCO band of fac-[CH3-
Ir(CO)2I3]- (3 mM in DCE solution, 60 psig CO at 25 °C)
(10) (a) Adams, H.; Bailey, N. A.; Mann, B. E.; Manuel, C. P.; Spencer,
C. M.; Kent, A. G. J. Chem. Soc., Dalton Trans. 1988, 489-496. (b)
Haynes A.; Maitlis, P. M.; Stanbridge, I. A.; Haak, S.; Pearson, J.
M.; Adams, H.; Bailey, N. A. Inorg. Chim. Acta 2004, 357, 3027-
3037.
(11) Adamson, G. W.; Daly, J. J.; Forster, D. J. Organomet. Chem. 1974,
71, C17-C19.
(12) Evans, J. A.; Russell, A. B.; Shaw, B. L. Chem. Commun. 1971, 841-
842.
1866 Inorganic Chemistry, Vol. 45, No. 4, 2006

Intermediates in Iridium-Catalyzed Methanol Carbonylation
absorbance bleach (∆Ao) was independent of [CO], the
experimental ∆A∞/∆Ao value being 0.64 (( 0.02). This
indicates that other intermediates besides ICO are formed in
the photoreaction or that competing processes deplete ICO
by pathways that are also first order in [CO] but lead to
different products. The first alternative, namely that several
intermediates are formed, receives support from the TRIR
spectrum recorded (point by point) immediately following
flash photolysis of a DCE solution of 1a under argon
(Supporting Information Figure S6). A mono-carbonyl
complex such as ICO should display a single νCO band;
however, weak absorptions were seen at 2156, 2116-2110,
2075, 2059, and 1665 cm-1 and a stronger one at 2032 cm-1.
Since the bleach of the 2043 cm-1 band of 1a proved to be
about half that expected given the magnitude of the bleach
at 2097 cm-1, it appears likely that another transient band
of moderate strength appears at ∼2040 cm-1. From these
data, it is clear that several intermediates are formed in the
flash photolysis of 1a.
Figure 8. Plot of kobs vs [CO] for re-formation of 1a, monitored at 2097
cm-1 (298 K, DCE).
following 355 nm pulse laser excitation. This IR band
undergoes an immediate bleach followed by partial recovery
that can be fit to an exponential function to give the first-
order rate constant kobs ) 8.7 ((0.2) × 105 s-1. Varying
[CO] over the range 6-40 mM13 leads to increases in the
reformation rate for 1a but does not change the magnitude
of the initial bleaching nor the quantitatiVe amount of
recoVery. The same figure also displays the temporal
behavior at 2100 cm-1 (νCO max) resulting from flash
photolysis of 1a in acetonitrile under analogous conditions.
In the latter case, there is little recovery on the microsecond
time scale of the experiment. Such solvent-dependent
behavior is consistent with initial formation of an unsaturated
intermediate via ligand photodissociation followed by trap-
ping of that species by coordination to a solvent molecule.
Acetonitrile would bind more strongly than would DCE,
hence any reaction with CO, or another ligand, would be
much more sluggish in acetonitrile.
One such intermediate would be an acyl complex formed
by methyl migration to a cis carbonyl as a reactive pathway
from the initial excited state (ES), in competition with
deactivation to 1a and with CO dissociation to give ICO
.
Trapping by solvent would give an acyl species, IM, which
in the absence of CO would be expected to decay back to
1a (eq 6).14 In the presence of CO, formation of 2b is likely.
The possible role of reversible iodide dissociation in the
generation and decay of the transient bleaching of 1a in DCE
was probed by conducting further flash photolysis experi-
ments in the presence of added [Bu4N]I (15-30 mM). The
excess I- had no effect on the rate or on the extent of 1a
regeneration, so it was concluded that iodide photodissocia-
tion is not playing a role in this case.
A plot of the kobs vs [CO] (298 K) proved to be linear
with a nonzero intercept (Figure 8). This behavior strongly
suggests that at least one of the intermediates initially formed
upon flash photolysis (ICO) is the result of CO photodisso-
ciation. The second-order rate constant for the reaction of
CO with ICO (eq 5)14 would be derived from the equation
Since only about one-third of the 1a depleted by the flash
is regenerated, it is necessary to account for the remaining
material and to evaluate pathways leading to the formation
of 2a. As noted above, continuous photolysis under argon
leads primarily to formation of the dimers of the acyl
carbonyl triiodide anions, [Ir(C(O)CH3)(µ-I)(CO)I2]22-. The
AsPh4+ salt of 4 was isolated and characterized structurally,
but the IR spectra suggest that other isomers are also present.
These could be formed via dimerization of IM (with ap-
propriate loss of solvent) or by trapping of IM by other
monomers. However, since CO addition to IM would give
2b rather than 2a,14 another intermediate, such as IM-mer is
a likely precursor to the latter (eq 7). Reaction of IM-mer
with CO would give 2a, while reverse migration of CH3-
to the metal would give the mer isomer 1b. Although 1b
kobs ) kCO[CO] + kM used to fit the data in Figure 8, which
gave the values kCO ) 2.50 (( 0.05) × 107 M-1 s-1 and kM
) 1.6 ((0.1) × 105 s-1. The nonzero intercept (kM) suggests
that there is also a CO-independent pathway for the re-
formation of 1a. This will be discussed further.
As noted above, the ratio of the overall absorbance bleach
(2097 cm-1 band) at long time (∆A∞) relative to the prompt
(14) (a) In discussing the reactions of the thermal substitution reactions of
the hexacoordinate iridium(III) organometallic compounds, it will be
assumed that these occur with retention of configuration, as has
generally been seen in the ligand substitution reactions of other 4d6
and 5d6 complexes. (b) It should be noted that, in earlier flash
photolysis studies of acyl carbonyl complexes of manganese and
cobalt,7b-f intermediates seen in weakly coordinated solvents were
concluded to have η2-acyl coordination rather than forming the
solvento species. In the present case, one should consider the possibility
that, for species such as IM and IM-mer, the vacated coordination site
may be lightly stabilized not by solvent coordination but by such η2-
acyl interaction.
(13) Cargill R. W. Carbon Monoxide, 1st ed.; IUPAC Solubility Data Series
Vol. 23; Pergamon Press: Elmsford, NY, 1990.
Inorganic Chemistry, Vol. 45, No. 4, 2006 1867

Volpe et al.
Scheme 2. Hypothetical Species Formed upon Photochemical
Excitation of 1a
was not detected, this is consistent with the view that 1b is
reactive toward migratory insertion.17
Computational Studies. The complexity of the photore-
action results led us to employ DFT computations to seek
insight into potential intermediates and prospective trans-
formations. A key feature of the above discussion is that
ligand photodissociation from a hexacoordinate species such
as 1a is accompanied by stereochemical lability of the
coordination sphere. Although thermally induced ligand
substitutions of 4d6 and 5d6 complexes generally occur by
stereoretentive pathways, analogous photochemical reactions
are often coupled to isomerizations.15 The energy introduced
by absorption of a UV-vis photon generally exceeds the
metal-ligand bond energies, so dissociative ligand labili-
zation is a likely electronic ES mechanism. The species
formed by the CO dissociation from the ES of 1a would be
pentacoordinate and is likely to carry vibrational excitation
in excess of any activation energy required for pseudo-
rotation to other pentacoordinate isomers. This is illustrated
in Scheme 2. Photodissociation of CO would first give a
square pyramidal anion, [CH3Ir(CO)I3]- (3a), with the methyl
group in the equatorial position, which might either be
trapped by solvent to give ICO or isomerize to the pentaco-
ordinate anion 3b which would react with solvent to give
mer-[CH3Ir(CO)I3(Sol)]- (ICO-mer), a predecessor of 1b. In
the condensed phase, solvent trapping of the pentacoordinate
intermediates is likely to be competitive with isomerizations,
so a distribution of products is probable.
Supporting Information Table S4) clearly show that 3a has
a strong tendency to rearrange to 3b. Pseudorotation to the
square pyramidal isomer with an axial methyl is calculated
to be 22 kcal mol-1 downhill if both are singlet states. The
triplet state of [CH3Ir(CO)I3]- optimizes to a trigonal
bipyramidal geometry with the methyl group and two iodides
in the equatorial positions (3c) with an energy very close to
that calculated for 3a (Table S4) Since that trigonal bipy-
ramidal species would be a likely intermediate in the 3a f
3b transformation, the respective energies of 3a, 3c, and 3b
portends a ready pathway for such an isomerization.16
Although the calculated energy difference between 3a and
3b seems surprisingly large, the trend is consistent with the
model proposed previously for photoisomerization of rhod-
ium(III) and iridium(III) complexes.15 Trapping of 3b by
solvent would give ICO-mer, and reaction of the latter with
CO gives 1b with one carbonyl trans to the methyl group
and one cis. It is this type of coordination that characterizes
the species argued to be precursor to the C-C bond
formation step of the Cativa cycle (Scheme 1).
In this context, DFT computations were initiated to
examine whether isomerization from 3a to 3b would be a
reasonable expectation. It should be emphasized that these
were carried out for gas-phase species and ignore both
general and specific solvation effects. Given these qualifica-
tions, it is notable that the DFT calculations (summarized in
Solvent-assisted migratory insertion of 1b to give IM-mer
followed by rapid reaction with CO leads to 2a, the principal
product of the photolysis of 1a under CO. DFT calculations
of 1b and of a likely transition state for the migratory
insertion reactions (Table S4) would suggest an activation
energy of ∼16 kcal mol-1 for the first step in this sequence
(eq 7). However, it should be reemphasized that the
computations do not account for specific or general solvation
effects. Attempts to account for such by coordinating a CH2-
Cl2 molecule to the iridium center in the transition state of
the migration pathway only lowered the Ea by 2 kcal mol-1,
although a pathway involving the direct reaction between
CO and ICO-mer to give 2a demonstrated a smaller Ea (∼10
kcal mol-1). Notably, the calculated Ea in the absence of
such interactions is in reasonable agreement with that (∼16
kcal mol-1) computed by Kinnunen and Laasonen using DFT
techniques for the same reaction.17 These workers, further-
more, calculated the Ea for the analogous reaction of 1a and
demonstrated that the barrier to thermally induced migratory
(15) (a) Skibsted, L. H.; Strauss D.; Ford, P. C. Inorg. Chem. 1979, 18,
3171. (b) Skibsted, L. H.; Ford, P. C., Inorg. Chem. 1980, 19, 1828.
(c) Talebinasab-Sarvari, M.: Ford, P. C. Inorg. Chem. 1980, 19, 2640,
(d) Ford, P. C.; Wink, D.; DiBenedetto, J. Prog. Inorg. Chem. 1983.
30, 213-269. (e) In these studies, it was concluded that the
pentacoordinated ES intermediate formed by ligand dissociation from
the ligand-field ES of a d6 hexacoordinate complex undergoes facile
isomerization among various square pyramidal geometries before
nonradiative deactivation and trapping by solvent to give substituted
products. To a first approximation, the most favored configuration
among these square pyramidal ES intermediates has the strongest
π-donor ligand in the axial site (see ref 15c).
(16) (a) In an earlier study,16b ab initio calculations at the restricted Hartree-
Fock level were used to probe the isomerizations between pentaco-
ordinate iridium(III) complexes with two phosphine ligands, two
hydrides and a π-donor ligand such as a halide. For such systems,
facile isomerizations among the pentacoordinate singlet states appar-
ently proceed via distorted trigonal bipyramidal structures that avoid
the pseudo-D3h structure, which is a higher-energy triplet species. This
would appear not to be a problem in the present case. (b) Riehl, J.-F.;
Jean, Y.; Eisenstein, O.; Pe´lissier, M. Organometallics 1992, 11, 729-
737.
(17) Kinnunen, T.; Laasonen, K. J. Mol. Struct. (THEOCHEM) 2001, 542,
273-288.
1868 Inorganic Chemistry, Vol. 45, No. 4, 2006

Intermediates in Iridium-Catalyzed Methanol Carbonylation
Scheme 3. Proposed Pathways for Photoinduced Migratory Insertion
we may conclude that such solvento acetyl complexes are
the precursors to the dimeric acyl complexes formed in less-
coordinating media. Flash photolysis studies also established
that long-lived transient species are formed in acetonitrile
solutions even under a CO atmosphere, consistent with the
proposed formation of intermediates ICO and ICO-mer as the
result of CO photodissociation.
A logical question is whether the photochemical investiga-
tions described here have relevance to the catalysis mech-
anisms for methanol carbonylation by related iridium iodo
carbonyl complexes, given the markedly different conditions
of the present studies. In detailed and very convincing
mechanistic characterizations of the catalytic systems,4e it
was demonstrated that, in contrast to the analogous rhodium
catalyst for methanol carbonylation, migratory insertion
pathways appeared to be rate limiting. Furthermore, catalysis
was promoted by species that may serve to labilize I- from
Ir(III) iodo complexes and to sequester the iodide and by
solvents that promote such iodide labilization. One conclu-
sion of these studies is that replacing I- by CO in the Ir(III)
activates those COs positioned appropriately to participate
in migratory insertion to form the acetyl intermediates that
are precursors to CH3C(O)I elimination and hydrolysis to
acetic acid. Nothing in the present study disputes those
arguments. However, our results suggest that I- labilization
may not be a requirement for facile migratory insertion to
form acyl Ir(III) complexes. Instead, the facility of acetyl
formation appears to be greatly enhanced by photoinduced-
diisomerization of 1a to 1b; thus, it may be largely dependent
on the orientation of other ligands relative to the methyl and
carbonyl groups participating directly in formation of the
acyl group.
insertion of that species was ∼12 kcal mol-1 higher. In the
present study, it is shown that the photochemical labilization
of CO, as well as the accompanying isomerization of the
resulting five-coordinate intermediate, provides a pathway
to coordination environments much more favorable to
migratory insertion than the direct reaction of 1a.
Summary. Scheme 3 draws together many of the obser-
vations and interpretations described above in the form of a
proposed sequence of events following the ambient-temper-
ature photoexcitation of the fac-[CH3Ir(CO)2I3]- anion 1a.
In this scenario, the ES formed (probably a ligand-field (d
f d) ES) is envisioned as having two reactive channels to
deactivation, dissociation of CO from the ES to form a
Acknowledgment. This research was supported by a
grant (DE-FG02-04ER15506) from the Division of Chemical
Sciences, Office of Basic Energy Research, U.S. Department
of Energy.
-
pentacoordinate species {CH3Ir(CO)I3 }* or migratory inser-
tion to carbonylate the Ir-CH3 bond and give an acyl
intermediate, IM, in what appears to be a lesser pathway.18
In the absence of added CO, the latter undergoes the
unimolecular reverse reaction to regenerate 1a (the kM
pathway) and may be a contributor to the formation of the
mixtures of dimers observed. In the presence of CO, IM
would be trapped to form 2b, a possible precursor to the
(18) (a) A referee has suggested that intramolecular migratory insertion
reactions for the ESs generated upon photolysis of 1a may find some
analogy in the oxidatively induced migratory insertion reactions
described some years ago by Giering et al.18b In this context, one might
consider whether the reactive ES(s) somehow mimic an intermediate
induced by a redox reaction, or is the migratory insertion chemistry
induced by radical pathways? We discount the latter on the basis of
the high reproducibility of quantum yield measurements for the
formation of the acyl species 2a, which argues against any type of
chain process. With regard to the former, we interpret the net
photochemical transformations as being largely the result of ligand
photolabilization, owing to the formation of ligand-field ESs of these
d6 complexes. Nonetheless, even ligand-field excitation involves
formation of a hole in a previously filled π-bonding MO (as well as
electronic population of a σ-antibonding MO),15d so a LF ES has (to
this extent) some analogy to the product of a one-electron oxidation
(or reduction) of the species in question. In this context, any migratory
insertion occurring directly from such a ligand-field ES may find some
analogy in the Giering experiment. (b) Prock, A.; Giering, W. P.;
Greene, J. E.; Meirowitz, R. E.; Hoffman, S. L.; Woska, D. C.; Wilson,
M.; Chang, R.; Chen, J. Organometallics 1991, 10, 3479-3485;
Woska, D. C.; Wilson, M.; Bartholomew, J.; Eriks, K.; Prock, A.;
Giering, W. P. Organometallics 1992, 11, 3343-3352; Woska, D.
C.; Bartholomew, J.; Greene, J. E.; Eriks, K.; Prock, A.; Giering, W.
P. Organometallics 1993, 12, 304-309.
-
small amount of cis-Ir(CO)2I2 seen in Figure 1. The
-
{CH3Ir(CO)I3 }* intermediate is apparently labile toward
isomerization between various geometric configurations;
however, such isomerization is competitive with deactivation
pathways concomitant with trapping by the solvent to form
the lightly stabilized six-coordinate solvento species {CH3Ir-
-
(CO)(Sol)I3 }, ICO and ICO-mer. Of these, the former would
react with CO to regenerate 1a (the kCO pathway), while the
latter would react with CO to form 1b. Under either Ar or
CO atmospheres, the stable photoproducts generated are
principally acyl complexes, dimers under inert atmosphere,
2a under CO.
Dimers were not formed under argon in acetonitrile
solutions. In this medium, the product was apparently the
mononuclear species [Ir(C(O)CH3)(CO)(CH3CN)I3]-. Thus,
Inorganic Chemistry, Vol. 45, No. 4, 2006 1869

Volpe et al.
Supporting Information Available: Cell packing diagrams for
three crystal structures plus several relevant figures showing spectra;
tables with listings of complete structure refinement details, bond
lengths, and angles, anisotropic displacement parameters for non-
hydrogen atoms, and hydrogen coordinates and isotropic displace-
ment parameters; and a table of results of computational studies.
This material is available free of charge via the Internet at
http://pubs.acs.org.
IC0513033
1870 Inorganic Chemistry, Vol. 45, No. 4, 2006