Mechanistic study of rhodium/xantphos-catalyzed methanol carbonylation

Article abstract of DOI:10.1021/om2006968

Rhodium/iodide catalysts modified with the xantphos ligand are active for the homogeneous carbonylation of methanol to acetic acid using either pure CO or CO/H₂. Residues from catalytic reactions contain a Rh(III) acetyl complex, [Rh(xantphos)(COMe)I₂] (1), which was isolated and crystallographically characterized. The xantphos ligand in 1 adopts a "pincer" κ³-P,O,P coordination mode with the xanthene oxygen donor trans to the acetyl ligand. The same product was also synthesized under mild conditions from [Rh(CO)₂I]₂. Iodide abstraction from 1 in the presence of donor ligands (L = MeCN, CO) gives the cationic acetyl species [Rh(xantphos)(COMe)I(L)]⁺, whereas in CH ₂Cl₂ migratory CO deinsertion gives [Rh(xantphos)(Me)I(CO) ]⁺ (4), which reacts with H₂ to liberate methane, as observed in catalytic reactions using syngas. A number of Rh(I) xantphos complexes have been synthesized and characterized. Oxidative addition of methyl iodide to the cation [Rh(xantphos)(CO)]⁺ is very slow but can be catalyzed by addition of an iodide salt, via a mechanism involving neutral [Rh(xantphos)(CO)I] (6). IR spectroscopic data and DFT calculations for 6 suggest the existence in solution of conformers with different Rh-O distances. Kinetic data and activation parameters are reported for the reaction of 6 with MeI, which proceeds by methylation of the Rh center and subsequent migratory insertion to give 1. The enhancement of nucleophilicity arising from a Rh- - -O interaction is supported by DFT calculations for the S_N2 transition state. A mechanism for catalytic methanol carbonylation based on the observed stoichiometric reaction steps is proposed. A survey of ligand conformations in xantphos complexes reveals a correlation between P-M-P bite angle and M-O distance and division into two broad categories with bite angle <120° (cis) or >143° (trans).

Full text of DOI:10.1021/om2006968

Article
pubs.acs.org/Organometallics
Mechanistic Study of Rhodium/xantphos-Catalyzed Methanol
Carbonylation
Gary L. Williams,† Christopher M. Parks,† C. Robert Smith,† Harry‡Adams,† Anthony Haynes,*,†
‡
Anthony J. H. M. Meijer,† Glenn J. Sunley, and Sander Gaemers
†Department of Chemistry, University of Sheffield, Sheffield S3 7HF, U.K.
‡BP Chemicals Ltd, Hull Research and Technology Centre, Saltend, Hull HU12 8DS, U.K.
S
* Supporting Information
ABSTRACT: Rhodium/iodide catalysts modified with the
xantphos ligand are active for the homogeneous carbonylation
of methanol to acetic acid using either pure CO or CO/H2.
Residues from catalytic reactions contain a Rh(III) acetyl
complex, [Rh(xantphos)(COMe)I2] (1), which was isolated
and crystallographically characterized. The xantphos ligand in
1 adopts a “pincer” κ3-P,O,P coordination mode with the xanthene oxygen donor trans to the acetyl ligand. The same product
was also synthesized under mild conditions from [Rh(CO)2I]2. Iodide abstraction from 1 in the presence of donor ligands (L =
MeCN, CO) gives the cationic acetyl species [Rh(xantphos)(COMe)I(L)]+, whereas in CH2Cl2 migratory CO deinsertion gives
[Rh(xantphos)(Me)I(CO)]+ (4), which reacts with H2 to liberate methane, as observed in catalytic reactions using syngas. A
number of Rh(I) xantphos complexes have been synthesized and characterized. Oxidative addition of methyl iodide to the cation
[Rh(xantphos)(CO)]+ is very slow but can be catalyzed by addition of an iodide salt, via a mechanism involving neutral
[Rh(xantphos)(CO)I] (6). IR spectroscopic data and DFT calculations for 6 suggest the existence in solution of conformers
with different Rh−O distances. Kinetic data and activation parameters are reported for the reaction of 6 with MeI, which
proceeds by methylation of the Rh center and subsequent migratory insertion to give 1. The enhancement of nucleophilicity
arising from a Rh- - -O interaction is supported by DFT calculations for the SN2 transition state. A mechanism for catalytic
methanol carbonylation based on the observed stoichiometric reaction steps is proposed. A survey of ligand conformations in
xantphos complexes reveals a correlation between P−M−P bite angle and M−O distance and division into two broad categories
with bite angle <120° (cis) or >143° (trans).
As well as improving catalytic activity, diphosphine ligands
have also been used to influence product selectivity. Moloy and
Wegman found that a dppp-modified rhodium/iodide catalyst
was effective for the reductive carbonylation of methanol to
INTRODUCTION
■
The large-scale commercial production of acetic acid is based
on carbonylation of methanol, catalyzed by rhodium or iridium
iodocarbonyl complexes in homogeneous solution.1−10 For the
rhodium-based process, originally developed by Monsanto, it is
well established that the rate-determining step is the oxidative
addition of methyl iodide to a square-planar Rh(I) complex,
[Rh(CO)2I2]−. The reaction proceeds via nucleophilic attack by
Rh on MeI, and considerably faster oxidative addition rates can
be achieved for Rh(I) complexes containing strongly electron
donating ligands. Numerous studies have aimed to take
advantage of this effect by using phosphine ligands and
derivatives to enhance the activity of a rhodium-based
carbonylation catalyst.11−17 For example, Cole-Hamilton and
co-workers reported high initial activity for a PEt3-modified
catalyst, but the beneficial effect was short-lived due to loss of
phosphine ligand and reversion of the rhodium catalyst to
[Rh(CO)2I2]−.16,17 A number of investigations have focused on
diphosphine ligands and derivatives, in the hope that a more
robust catalyst would result from the chelate effect.18−35
Although some notable rate enhancements have been achieved,
long-term catalyst stability is generally an issue and the authors
are not aware of any commercial processes that are based on
phosphine-modified methanol carbonylation.
acetaldehyde using syngas instead of pure CO.36,37
A
ruthenium cocatalyst allowed in situ hydrogenation of the
aldeyde, resulting in an overall homologation of methanol to
ethanol. The Rh/dppp system gave >80% selectivity for
acetaldehyde/ethanol, whereas other diphosphines (e.g.,
dppb) gave acetic acid as the major product. The behavior
was rationalized on the basis of competitive reactions of a
Rh(III) acetyl intermediate, [Rh(diphosphine)(COMe)I2], as
illustrated in Scheme 1. Coordination of CO and subsequent
hydrolysis (as in the conventional mechanism for the Monsanto
process) leads to acetic acid, whereas reaction with H2 gives
acetaldehyde.
Other polydentate phosphines have also been tested for their
effects on product selectivity. In a patent arising from research
at BP Chemicals Limited, Gaemers and Sunley reported
increased selectivity to acetic acid, under syngas, for a range of
ligands with relatively rigid backbones and/or wide bite angles
Received: July 28, 2011
Published: October 25, 2011
© 2011 American Chemical Society
6166
dx.doi.org/10.1021/om2006968|Organometallics 2011, 30, 6166−6179

Organometallics
Article
S c h e m e 1 . C o m p e t i t i v e R e a c t i o n s o f
[Rh(diphosphine)(COMe)I2] with CO and H2
RESULTS AND DISCUSSION
■
Rhodium(III) Acetyl Complexes. On cooling of the
solution after a Rh/xantphos catalyzed carbonylation reaction,
a substantial quantity of crystalline solid was found to
precipitate.43 The IR spectrum of this product displayed an
absorption at 1683 cm−1, consistent with ν(CO), but there
were no bands in the region associated with terminal carbonyl
ligands. The 31P NMR spectrum showed a doublet at δ 9.73
(1JRh−P = 109.1 Hz) consistent with a rhodium xantphos
complex containing equivalent PPh2 moieties.
The same compound, 1, was formed under mild conditions
by the reaction sequence illustrated in Scheme 2. The reaction
Scheme 2. Synthetic Route to [Rh(xantphos)(COMe)I2] (1)
(in comparison with dppp).38 For example, using the wide bite
angle ligand xantphos (Chart 1, first reported by Kranenburg et
Chart 1. Diphosphine Ligands Applied in Rhodium-
Catalyzed Carbonylation Using Syngas
al.39) very little acetaldehyde or ethanol was produced from
methanol under reductive carbonylation conditions (2:1
H2:CO, 68 bar, 140 °C) and the major liquid product was
acetic acid. However, a large amount of methane was also
formed (accounting for ca. 60% of the methanol consumed).
Another notable example was binap, which increased the
selectivity to acetic acid but did not give so much methane.
Further to this, Lamb et al. assessed a number of C4-bridged
diphosphines and found that the selectivity to acetic acid under
reductive carbonylation conditions was higher for the more
rigid dppx and binap ligands than for dppb.30 Such systems may
have potential for a “hydrogen-tolerant” carbonylation process,
whereby high selectivity to acetic acid is maintained using CO
of lower purity, thus alleviating some of the costs of syngas
separation.
of methyl iodide with [Rh(CO)2(μ-I)]2 in acetonitrile gives an
iodide-bridged dimer that is a convenient labile precursor to
other Rh(III) acetyl complexes.44 Reaction of the dimer with
xantphos occurs readily with the loss of CO and MeCN ligands,
giving 1 in good yield, with spectroscopic properties identical
with those of the material isolated from catalytic experiments.
The structure of complex 1 was confirmed by X-ray
crystallography. The molecular structure is illustrated in Figure
1, and selected geometrical parameters are given in Table 1.
Xantphos was also tested as a ligand for rhodium-catalyzed
methanol carbonylation with pure CO as the feed gas.40 The
carbonylation rate measured for a xantphos-modified rhodium
catalyst was marginally faster than that for the unmodified
system (14.8 vs 12.9 mol dm−3 h−1 at 28 bar and 190 °C). In
contrast, when [Rh(CO)(dppe)I3] was used as catalyst
precursor, the activity was substantially lower (4.5 mol dm−3
h−1), consistent with data reported by Carraz et al.29,41
Enhanced carbonylation activity for a Rh/xantphos catalyst
has also been reported recently by Deb and Dutta.42
In this paper we report a study of the organometallic
chemistry of rhodium xantphos complexes underlying the
catalytic reactions summarized above. The starting point for
this investigation was the isolation from a catalyst residue of a
Rh(III) acetyl complex containing the xantphos ligand. A
mechanism for the catalytic carbonylation reaction is proposed
on the basis of structural, spectroscopic, and reactivity studies
of Rh(I) and Rh(III) species. Related chemistry of the bulkier
o-tolyl xantphos ligand is also reported, and the ability of
xantphos to adopt different coordination modes is considered.
Figure 1. ORTEP diagram showing the molecular structure of 1.
Thermal ellipsoids are shown at the 50% probability level. Hydrogen
atoms are omitted for clarity.
6167
dx.doi.org/10.1021/om2006968|Organometallics 2011, 30, 6166−6179

Organometallics
Article
ca. 2.33 Å.46 The range of coordination modes adopted by the
xantphos ligand is surveyed later in this paper.
Table 1. Selected Bond Distances (Å) and Angles (deg) for
Complexes 1 and 2[BF4]
The xanthene backbone of the xantphos ligand in 1 is almost
planar (mean deviation from plane 0.052 Å), and the two
backbone methyl substituents give rise to a singlet at δ 1.65 in
1
2[BF4]
Rh−P(1)
Rh−P(2)
Rh−I(1)
Rh−I(2)
Rh−N
Rh−O(1)
Rh−Cacetyl
P(1)−Rh−P(2)
I(1)−Rh−I(2)
I(1)−Rh−N(2)
Cacetyl−Rh(1)−O(1)
2.3755(10)
2.3276(10)
2.6685(4)
2.6530(4)
2.3223(12)
2.3166(12)
2.6694(4)
1
the H NMR spectrum due to their equivalence. The acetyl
4
methyl group appears as a triplet at δ 2.75, with a small JP−H
coupling of 1.7 Hz. Selective decoupling by irradiation of the
31P resonance of 1 caused the triplet to collapse into a singlet,
confirming that the splitting is due to H−P coupling. This
observation contrasts with [Rh(diphosphine)(COMe)I2] com-
plexes containing cis-chelating phosphines, for which singlets
2.032(3)
2.374(3)
2.009(5)
160.12(4)
2.324(2)
1.980(4)
160.06(4)
175.082(15)
are reported for the acetyl protons.25,30,37,45 However, a JP−H
4
172.43(10)
177.49(12)
coupling of 1.4 Hz was reported for [Rh(PEt3)2(COMe)-
(C5F4N)I].47 Inspection of the X-ray crystal structures reveals
that, in the solid state, the complexes that display detectable
4JP−H coupling have mean P−Rh−Cacetyl angles >95°, whereas
the splitting is absent when this angle is closer to 90°. In the X-
ray structure of 1, the acetyl ligand adopts a conformation that
is almost eclipsed with respect to the xantphos P-donor atoms
(P1−Rh−C14−C15 dihedral angle ca. −18°). Although this
makes the P atoms inequivalent in the solid state, only one
doublet is observed in the 31P NMR spectrum of 1 (even at
−90 °C), indicating that rotation of the acetyl ligand about the
Rh−C bond is fast on the NMR time scale in solution.
Restricted rotation for the Rh−acetyl unit in [Rh-
(CO)2(COMe)I3]− at −130 °C has been reported by Mann
and co-workers.48
176.53(12)
Several other complexes of the type [Rh(diphosphine)-
(COMe)I2] have previously been structurally characterized
(e.g., for diphosphine = dppm,45 dppe,25 dppp,37 dppb30,
dppx30). All of these have the diphosphine coordinated in a cis-
chelate fashion with the two P-donor atoms occupying basal
sites of a square pyramid, as depicted in Scheme 1. In contrast,
in 1 the xantphos ligand coordinates in a mer-κ3-P,O,P manner,
such that the P donors span trans positions of a distorted
octahedron with a P−Rh−P bite angle of ca. 160°. The two
iodide ligands are mutually trans, and the central oxygen atom
of xantphos occupies the remaining coordination site trans to
the acetyl ligand, with a Rh−O distance of 2.324(2) Å.
Although less common than the bidentate κ2-P,P coordination
mode, a number of complexes containing xantphos coordinated
as a tridentate “pincer” ligand have been crystallographically
characterized. For example, the ruthenium alkylidene complex
[Ru(κ3-xantphos)(CHPh)Cl2] is structurally very similar to
1 with a P−Ru−P bite angle of ca. 161° and M−O distance of
Complex 1 did not show any reactivity toward CO, H2, or
H2O under mild conditions. However, abstraction of an iodide
ligand was accomplished using AgBF4 or AgSbF6. When this
reaction was performed in a 2:1 CH2Cl2/MeCN solvent
mixture, the cationic Rh(III) acetyl complex 2 was isolated in
Scheme 3. Iodide Abstraction Reactions of 1a
6168
dx.doi.org/10.1021/om2006968|Organometallics 2011, 30, 6166−6179

Organometallics
Article
good yield (Scheme 3). The IR spectrum of 2 showed a ν(C
O) absorption for the acetyl ligand at 1708 cm−1, shifted to
high frequency with respect to that of 1. A doublet in the 31P
NMR spectrum at δ 16.5 (1JRh−P = 100.7 Hz) is consistent with
equivalent phosphorus atoms. Crystals suitable for X-ray
diffraction were obtained from chloroform/diethyl ether. The
molecular structure of 2 is shown in Figure 2, and selected
absorptions at 2084 and 2013 cm−1. The low-frequency band
is assigned to [Rh(xantphos)(CO)]+ (5) resulting from
reductive elimination of MeI, and the shift of the high-
frequency absorption is consistent with formation of a different
isomer of [Rh(xantphos)(CO)(Me)I]+. The 31P NMR
spectrum revealed the slow growth of a new doublet at δ
1
13.2 (1JRh−P = 86 Hz), and the H NMR spectrum showed
resonances associated with the methyl ligand (δ 2.16, td, 2JRh−H
= 2.4 Hz, 3JP−H = 4.6 Hz) and backbone methyls (1.90, s, 1.86,
s) of the new isomer. It is proposed that isomerization gives 4b
with CO trans to iodide and methyl trans to the xantphos
oxygen donor (Scheme 3). The formation of 4b via a different
route and the relative stabilities of these geometrical isomers are
discussed further later.
Some crystals of suitable quality for an X-ray diffraction study
were obtained by attempted recrystallization of 4a,b from
CH2Cl2/Et2O. However, the resulting crystal structure (see the
Supporting Information) unexpectedly revealed an iodide-
2+
bridged Rh acetyl dimer, [Rh(xantphos)(COMe)(μ-I)]2
,
depicted in Scheme 3. This minor product (that presumably
crystallizes preferentially) contains xantphos coordinated in a
cis-κ2-P,P mode with a P−Rh−P bite angle of 99.78(7)°. Each
of the rhodium centers adopts an approximate square-
pyramidal geometry with a vacant site trans to acetyl and
bridging iodides trans to the xantphos P-donor atoms. The
coordination geometry around each rhodium in this dimer
therefore resembles that found for monomeric complexes of the
type [Rh(diphosphine)(COMe)I2] containing cis-chelating
diphosphine ligands.25,30,37,45 The Cambridge Structural Data-
base contains only two other xantphos complexes with bite
angles <100°: namely, [Pd(xantphos)(CF3)Ph]49 and [Re-
(xantphos)Cl3O].50
Figure 2. ORTEP diagram showing the molecular structure of cation
2. Thermal ellipsoids are shown at the 50% probability level. Hydrogen
atoms, BF4 counterion, CHCl3 solvent molecules, and disordered
−
atoms C(39) and O(2′) in the acetyl ligand are omitted for clarity.
Since competitive reactions of Rh(III) acetyl species are
thought to be important for the selectivity of catalytic
carbonylation under syngas (Scheme 2), the reactivity of 2
toward CO and H2 was probed. Brief bubbling of carbon
monoxide through a solution of 2[BF4] in CH2Cl2 resulted in a
shift of the acetyl ν(CO) band to 1731 cm−1 and formation
of a strong absorption in the terminal ν(CO) region at 2096
cm−1, consistent with displacement of the coordinated MeCN
by CO to give [Rh(CO)(xantphos)(COMe)I]+ (3), as shown
in Scheme 3. A new doublet in the 31P NMR spectrum at δ 8.1
(1JRh−P = 95.8 Hz) is also assigned to 3, but the complex was
not isolated as a solid. On standing, signals appeared in the IR
and 31P NMR spectra consistent with slow formation of
[Rh(CO)(xantphos)]+, which could arise by reductive
elimination of acetyl iodide (perhaps scavenged by traces of
water).
After treatment of a CH2Cl2 solution of 2[BF4] with
hydrogen gas (5 atm) for 2 h at 46 °C, the IR spectrum
displayed a single ν(CO) absorption at 2014 cm−1, which is due
to [Rh(CO)(xantphos)]+ (5). The same transformation
occurred overnight at 48 °C under 1 atm of H2, with methane
as the only organic product detected by gas chromatography. In
a control experiment, in which a solution of 2 in CH2Cl2 was
heated to 48 °C under nitrogen, IR spectroscopy indicated that
after 2 h the reactant complex remained as the major species
but a weak band at 2085 cm−1 had appeared, attributable to
methyl complex 4b. After ca. 16 h this absorption continued to
grow, along with a band due to 5, although more than 50% of
the reactant acetyl complex remained. Thus, in the absence of
H2 there is evidence that migratory deinsertion can occur (with
loss of MeCN), followed by reductive elimination of MeI to
geometric data are given in Table 1. The coordination
geometry resembles that of 1, with an iodide ligand substituted
by MeCN. The acetyl ligand is disordered between two
rotational conformers. The xantphos ligand retains a mer-κ3-
P,O,P coordination mode with a P−Rh−P bite angle and Rh−
O distance similar to those in 1. However, the xanthene
backbone in 2 is distorted from planarity, with a fold angle of
ca. 23° between the mean planes of the two aromatic rings.
This results in approximate Cs symmetry for the complex, with
two of the phenyl groups flanking the coordinated MeCN. The
presence of mutually trans iodide and MeCN ligands cause the
two xantphos backbone methyl substituents to be inequivalent,
1
and they appear as two singlets at δ 1.76 and 1.92 in the H
NMR spectrum. As for 1, a triplet is observed for the acetyl
4
protons of 2 at δ 2.80 with JP−H coupling of 1.4 Hz.
When iodide abstraction from 1 was performed in CH2Cl2, a
different outcome resulted. Precipitation of AgI was accom-
panied by formation of a strong band in the IR spectrum at
2074 cm−1 and a doublet in the 31P NMR spectrum at δ 25.2
1
(1JRh−P = 80 Hz). The H NMR spectrum showed a triplet of
doublets at δ 2.22 attributable to a Rh−CH3 species (2JRh−H
=
2.0 Hz, 3JP−H = 4.5 Hz) and two singlets at δ 1.97 and 1.91 due
to inequivalent methyl substituents on the xantphos backbone.
These spectroscopic data are consistent with formation of a
rhodium methyl complex, 4a, resulting from migratory
deinsertion (Scheme 3). Consistent with this, the mass
spectrum revealed a peak at m/z 851, corresponding to
[Rh(xantphos)(CO)(Me)I]+. When the solution containing 4a
was allowed to stand, the ν(CO) band at 2074 cm−1 slowly
decayed (over ca. 1 day) with the appearance of new
6169
dx.doi.org/10.1021/om2006968|Organometallics 2011, 30, 6166−6179

Organometallics
Article
Scheme 4. Reductive Elimination Routes for 2 in the Presence or Absence of H2
give 5. Formation of the same Rh(I) species occurs more
rapidly in the presence of H2, and the coproduction of methane
suggests that an intermediate rhodium methyl complex is
intercepted by reaction with H2, as shown in Scheme 4. This is
consistent with the large amounts of methane observed in
catalytic carbonylation reactions conducted using syngas (vide
supra).38 We have not probed the methane formation
mechanism in more detail, but a vacant coordination site is
presumably required for activation of H2. This could be
achieved by loss of an iodide or CO ligand from 4a,b or by a
change in the xantphos coordination mode from κ3-P,O,P to
κ2-P,P.
The rhodium coordination chemistry of a more sterically
demanding xantphos derivative containing P(o-tolyl)2 groups
was also investigated. In contrast to the observations described
above, the reaction of o-tolyl-xantphos with [Rh(CO)(NCMe)-
(COMe)(μ-I)2] did not give an isolable Rh(III) acetyl species
analogous to 1 but instead gave a product with a ν(CO)
absorption at 1951 cm−1, characteristic of a Rh(I) carbonyl
species. The 31P NMR spectrum displayed a doublet at δ 18.6
(1JRh−P = 115.6 Hz), demonstrating that the diphosphine ligand
is bound to rhodium with equivalent P atoms. An X-ray
crystallographic structure determination revealed a complex of
formula [Rh(o-tolyl-xantphos)(CO)I] (6otol), as shown in
Figure 3. The structure has a five-coordinate Rh center with
the o-tolyl-xantphos adopting a mer-κ3-P,O,P coordination
mode with Rh−P distances and P−Rh−P bite angle similar to
those observed in 1. The details of this structure are discussed
in more detail later in the context of related Rh(I) complexes.
Formation of [Rh(o-tolyl-xantphos)(CO)I] could, in principle,
arise either by reductive elimination of methyl iodide or acetyl
iodide from a Rh(III) precursor, but no organic acetyl products
were detected. The mechanism was explored further using
isotopic labeling. The anionic dimer [Rh(13CO)(12COMe)I2(μ-
I)]22− was synthesized with specific 13CO enrichment (ca. 85%)
of the terminal carbonyl ligand using a published method.51
After gentle reflux of a solution of this dimer with o-tolyl-
xantphos (1:1 Rh/L) in CH2Cl2 overnight, the IR spectrum
displayed bands at 1951 and 1904 cm−1 due to [Rh(o-tolyl-
xantphos)(12CO)I] and [Rh(o-tolyl-xantphos)(13CO)I], re-
spectively, in a ca. 2:1 ratio. Hence, the dominant product
isotopomer has a terminal carbonyl ligand arising from the
acetyl group of the precursor, demonstrating that migratory
deinsertion has occurred.52 A mechanism consistent with the
observations is shown in Scheme 5. Initial coordination of o-
tolyl-xantphos is proposed to give the Rh(III) acetyl complex
1otol. Dissociation of an iodide ligand would allow methyl to
Figure 3. ORTEP diagram showing the molecular structure of [Rh(o-
tolyl-xantphos)(CO)I] (6otol). Thermal ellipsoids are shown at the
50% probability level. Hydrogen atoms and CH2Cl2 solvent molecule
are omitted for clarity.
Scheme 5. Proposed Mechanism for Formation of [Rh(o-
tolyl-xantphos)(12CO)I] (6otol
)
migrate to Rh, followed by elimination of MeI to generate 6otol
.
The instability of the proposed octahedral Rh(III) acetyl and
methyl species in this system must arise from the additional
6170
dx.doi.org/10.1021/om2006968|Organometallics 2011, 30, 6166−6179

Organometallics
Article
steric bulk of the diphosphine ligand, which will disfavor higher
coordination numbers. Consistent with this, 6otol was found to
be unreactive toward MeI, in contrast to the less congested
xantphos analogue (vide infra).
Rhodium(I) Complexes. Sandee et al. reported a simple
synthetic route to [Rh(CO)(xantphos)]+ (5), involving the
reaction of [Rh(CO)2(μ-Cl)]2 with xantphos followed by
abstraction of chloride using AgBF4.53 The presumed neutral
intermediate [Rh(CO)(xantphos)Cl] (7) was not isolated in
that study but has since been reported by Deb and Dutta.42
Some analogous carbonyl chloride complexes containing
xanthene-based diphosphonite ligands have also been re-
ported.54,55 Addition of xantphos to a solution of [Rh-
(CO)2(μ-Cl)]2 in ethanol results in precipitation of complex
7 as a yellow solid. The IR spectrum of this product in CH2Cl2
displays a single ν(CO) absorption at 1968 cm−1, and the 31P
NMR spectrum shows a doublet at δ 21.5 (1JRh−P = 130.4 Hz),
demonstrating the equivalence of the phosphorus atoms.
The structure of 7 was confirmed by X-ray crystallography
(Figure 4). The complex has a distorted-square-planar
Figure 5. ORTEP diagram showing the molecular structure of [Rh(o-
tol-xantphos)(CO)Cl] (7otol). Thermal ellipsoids are shown at the
50% probability level. Hydrogen atoms are omitted for clarity.
Table 2. Selected Bond Distances (Å) and Angles (deg) for
[Rh(CO)(o-tol-xantphos)I] (6otol), [Rh(CO)(xantphos)Cl]
(7), and [Rh(CO)(o-tol-xantphos)Cl] (7otol
)
6otol
7
7otol
Rh−P(1)
Rh−P(2)
Rh−Cl/I
Rh−C
Rh−O
C−O
P(1)−Rh−P(2)
C−Rh−Cl/I
Rh−C−O
2.3021(18)
2.2928(18)
2.8180(7)
1.792(8)
2.334(4)
1.159(9)
160.93(7)
143.4(2)
175.1(6)
2.3315(16)
2.322(2)
2.434(2)
1.832(2)
2.654(3)
1.163(2)
2.2966(15)
2.2872(14)
2.4169(16)
1.783(7)
2.510(4)
1.153(7)
153.135(19)
173.29(5)
155.79(6)
157.0(2)
178.35(15)
178.0(6)
structural differences between 7 and 7otol arise from the greater
steric requirements of the o-tol-xantphos ligand. Inspection of
the ligand conformations shows that the presence of o-methyl
substituents causes significant reorientation of the aryl rings.
Two methyl groups occupy positions that shield the “open”
face of the Rh coordination sphere and have the effect of
forcing the rhodium atom deeper into the binding cavity. The
resulting proximity of the xanthene oxygen will increase the
electron density on Rh and thereby enhance back-donation to
CO, consistent with the considerably lower ν(CO) value for
Figure 4. ORTEP diagram showing the molecular structure of
[Rh(xantphos)(CO)Cl] (7). Thermal ellipsoids are shown at the 50%
probability level. Hydrogen atoms and the CH2Cl2 solvent molecule
are omitted for clarity.
coordination geometry about the rhodium(I) center with the
xantphos ligand coordinating in a trans-κ 2-P,P fashion. The P−
Rh−P bite angle is ca. 153°, and the Rh−O distance is 2.654(3)
Å, representing an approach to a five-coordinate square-
pyramidal geometry slightly closer than the corresponding
values reported for structurally related diphosphonite com-
plexes.54,55
An analogous Rh(I) chlorocarbonyl complex, [Rh(CO)(o-
tol-xantphos)Cl] (7otol), was synthesized and structurally
characterized. Notable features of the structure, shown in
Figure 5, are that the Rh−O distance (2.510(4) Å) is
significantly shorter than in 7 and there is a greater distortion
from square-planar geometry, with the P−Rh−P and Cl−Rh−
C angles showing larger deviations from linearity (Table 2).
The ν(CO) values of 7 and 7otol are also notably different, with
a 19 cm−1 shift to low frequency for the o-tolyl derivative,
despite the similar donor strengths of P(o-tol)2 and PPh2
groups. For comparison, there is only a 5 cm−1 shift in
ν(CO) for trans-[Rh(CO)(PAr3)2Cl] (Ar = Ph, o-tol).56 The
7otol
.
The cation [Rh(CO)(o-tol-xantphos)]+ (5otol) was also
−
synthesized (as its BF4 salt) and structurally characterized
(see the Supporting Information). The principal features of the
structure, including the conformation of ligand aryl substitu-
ents, are very similar to those of the xantphos analogue 5.57 A 4
cm−1 shift in ν(CO) to low frequency for the o-tol-xantphos
complex is in line with expectation and supports the suggestion
that the larger shift in ν(CO) between 7 and 7otol is due to the
structural distortion of the bulky o-tol-xantphos ligand required
to accommodate a chloride ligand.
The reaction of o-tol-xantphos with [Rh(CO)2(μ-I)]2
resulted in formation of 6otol, the same species that was
isolated form the reaction of a Rh(III) acetyl precursor with o-
tol-xantphos (vide supra). The X-ray structure of 6otol (Figure
3) shows an even greater deviation from square-planar
geometry than 7otol, with a severely bent I−Rh−C angle (ca.
143°) and a very short Rh−O distance (2.334(7) Å). The
6171
dx.doi.org/10.1021/om2006968|Organometallics 2011, 30, 6166−6179

Organometallics
Article
relatively low ν(CO) value of 6otol (1951 cm−1) can again be
attributed to the additional electron density on Rh arising from
close proximity to the xanthene oxygen.
Structural Database (CSD).60−62 Figure 7 plots the distribution
of P−M−P bite angles (including those of xantphos and o-tolyl
The synthesis of [Rh(CO)(xantphos)I] (6) was attempted
by the reaction of [Rh(CO)2(μ-I)]2 with xantphos and by
addition of iodide salts to 5[BF4]. Although spectroscopic
evidence indicated formation of the desired product in solution,
we were unable to isolate 6 as a crystalline solid. Addition of 1
molar equiv of Bu4NI to a solution of 5[BF4] in CH2Cl2 results
in the appearance of a ν(CO) absorption at 1958 cm−1 in the
region expected for 6. A doublet was observed at δ 31.1 (1JRh−P
= 120.2 Hz) in the 31P NMR spectrum. Close inspection of the
IR spectrum showed that the ν(CO) band at 1958 cm−1 is
rather broad and unsymmetrical, resulting from a shoulder to
high frequency (Figure 6).58 This band shape was found to be
Figure 7. Bar chart showing distribution of P−M−P bite angles in
crystallographically characterized complexes containing the xantphos
(or o-tolyl-xantphos) ligand. Each bar represents the number of
examples within ±2.5° of the angle stated.
xantphos complexes reported in this paper). This plot
highlights the wide range of bite angles (ca. 99−165°) available
to the xantphos ligand and emphasizes the important role that
the metal and coligands play in controlling the diphosphine
conformation. The data split broadly into cis (<120°) and trans
(>143°) complexes, with examples largely absent from the
intervening region. Figure 8 reveals a correlation between M−
Figure 6. Deconvolution of ν(CO) absorption of [Rh(CO)-
(xantphos)I] (6) measured in CH2Cl2. The observed spectrum
(black) is fitted by two peaks at 1956 and 1969 cm−1 (blue). A third
component, marked with an asterisk, is due to ν(13CO) at natural
abundance. The red trace shows the residual of the fit.
reproducible in several experiments and suggests that there is
more than one contributing species. A satisfactory fit was
achieved using two overlapping absorptions, with the main
band centered at 1956 cm−1 and a weaker component centered
at 1969 cm−1, as shown in Figure 6.59 DFT calculations support
the existence of multiple conformers of complex 6 (details
given in the Supporting Information, Table S8). Three
conformers of 6 were located, having very similar energies
(ΔE < 4 kJ mol−1) but with different Rh- - -O distances
(2.315−2.667 Å) and calculated ν(CO) values spanning 7
cm−1. It is therefore feasible that an equilibrium mixture of such
species gives rise to the experimental solution IR spectrum of 6.
Structural Survey of Xantphos Complexes. The
xantphos ligand was initially designed to have a relatively
wide bite angle that would favor bis(equatorial) coordination
over axial−equatorial coordination in a trigonal-bipyramidal
complex (with the aim of influencing selectivity in rhodium-
catalyzed hydroformylation). On the basis of molecular
mechanics calculations, Kranenburg et al. calculated the natural
bite angle of xantphos to be 111.7° with a flexibility range of
97−135°.39 Although originally considered to be a cis-chelating
bidentate diphosphine ligand, complexes were soon identified
in which xantphos acted as a pincer-type tridentate ligand, with
the central oxygen also coordinated (e.g., complex 557). In view
of the range of coordination modes observed during the present
study, we have surveyed the X-ray structures of xantphos
complexes reported in the literature using the Cambridge
Figure 8. Plot of M−O distance versus P−M−P bite angle in
crystallographically characterized xantphos complexes.
O distance and P−M−P bite angle. This is to be expected, since
coordination of xantphos in a trans-spanning fashion with a
large bite angle will naturally result in a relatively close
approach of the xanthene oxygen to the metal center. The
group of cis complexes (black ◆) toward the top left of Figure
8 includes structures with metal coordination numbers ranging
from 3 to 6 and exhibits a considerable degree of scatter,
whereas the trans complexes at the bottom right follow quite a
smooth arc and can be split into three principal categories.
Those (red ●) with a bite angle range of ca. 143−153° and
M−O distances between 2.6 and 2.8 Å all have distorted-
square-planar structures (e.g., 7), with the xantphos coordina-
6172
dx.doi.org/10.1021/om2006968|Organometallics 2011, 30, 6166−6179

Organometallics
Article
tion mode best regarded as trans-κ 2-P,P. Next (blue ■) are
some approximately octahedral complexes (e.g., 1 and 2)
having mer-κ3-P,O,P coordination of xantphos with bite angles
close to 160° and M−O ca. 2.2−2.4 Å. Finally (orange ◆)
comes a small group of square-planar d8 complexes that also
observed in this experiment is the same as that for the species
resulting from isomerization of 4a after iodide abstraction from
1 (see Scheme 3).
Since the iodide released in the SN2 step can subsequently
bind to another molecule of 5, the net oxidative addition
reaction is catalytic in I−. Halide-catalyzed oxidative addition
reactions of d8 metal complexes that occur via initial
coordination of halide have been reported previously.65−67
On the basis of the observed rate of formation of 4b and
estimated concentration of 6 the rate constant (k1) for reaction
of 6 with MeI is calculated to be at least 0.01 M−1 s−1 (at 25
°C).68 This value is considerably larger than rate constants
reported for oxidative addition of MeI to the rhodium Vaska
complexes [Rh(CO)(PPh3)2Cl]69,70 and [Rh(CO)-
(PEt3)2I].16,17 Both ground-state and transition-state effects
might contribute to the high nucleophilicity of 6. In the ground
state the presence of the xantphos oxygen atom and a bite angle
significantly less than 180° could alter the electronic character
of the metal center relative to a square-planar Vaska-type
complex. However, Mulliken charges from DFT calculations for
each conformer of 6 indicate a slightly more positive Rh center
than for [Rh(CO)(PPh3)2I] (see the Supporting Information).
The enhanced nucleophilicity of 6 therefore likely arises from
stabilization of the SN2 transition state by an interaction
between the rhodium center and the xantphos oxygen. Similar
neighboring group effects on reactivity have been reported for
adopt the mer-κ3-P,O,P coordination mode (e.g., 5 and 5otol
)
with the largest bite angle (>163°) and shortest M−O distance
(<2.2 Å). The three outliers (green ▲) with short M−O
distances and relatively small bite angles have all been reported
very recently. Two of these are Ir(I) complexes ([Ir(cod)-
(xantphos)]+ and [{Ir(xantphos)(μ-H)}2]+), which display an
unusual fac-κ3-P,O,P xantphos coordination mode,63 whereas
the third is a Ru(II) hydride, [Ru(xantphos)(PPh3)(O2)(H)]+,
in which there is a considerable distortion from octahedral
geometry, presumably facilitated by the small hydride ligand.64
The substantial deviation from square-planar geometry noted
above for 6otol and 7otol is also apparent from the graph in
Figure 8. The data points for these two complexes are shifted
toward the bottom right of the plot relative to other complexes
of this category. This is especially the case for the iodide 6otol
,
which has a bite angle and Rh−O distance more typical of
octahedral complexes with mer-κ3-P,O,P coordinated xantphos.
Reactivity of Rh(I) Complexes with MeI. The cationic
complex [Rh(xantphos)(CO)]+ (5) is rather unreactive toward
methyl iodide. Even at very high MeI concentration (8 M, i.e.
1:1 v:v in CH2Cl2) the ν(CO) band of 5 at 2014 cm−1 decays
only slowly at room temperature, accompanied by growth of an
absorption at 2084 cm−1, consistent with formation of the
Rh(III) methyl product 4b. Much faster reaction rates were
attained when substoichiometric quantities of an iodide salt
were added to the reaction solution. For example, in the
reaction of 5 with 0.8 M MeI in CH2Cl2 containing 0.16 mol
equiv of Bu4NI (with respect to Rh), essentially complete
conversion to 4b occurred in ca. 800 s at 25 °C. A reaction
sequence consistent with these observations is illustrated in
Scheme 6. Initial coordination of iodide will result in partial
complexes containing ligands with o-anisyl substituents.71−73
A
transition-state structure for the nucleophilic substitution step
was optimized using DFT, for a simplified model system with
PPh2 groups replaced by PMe2. The transition state is
illustrated in Figure 9, along with that for nucleophilic attack
Scheme 6. Mechanism for Iodide-Catalyzed Reaction of MeI
with [Rh(CO)(xantphos)]+
Figure 9. DFT optimized transition state structures for nucleophilic
attack on MeI by [Rh(CO)(Me-xantphos)I] (left) and trans-
[Rh(CO)(PMe3)2I] (right).
by trans-[Rh(CO)(PMe3)2I] on MeI. The calculated free
energy of activation (relative to the separated reactants in the
gas phase) is 105.4 kJ mol−1 for [Rh(CO)(Me-xantphos)I]
compared with 116.2 kJ mol−1 for trans-[Rh(CO)(PMe3)2I].
Hence, the theoretical calculations are consistent with
enhanced nucleophilicity for the xantphos complex.
DFT was also used to compute relative energies of the
Rh(III) methyl isomers 4a−c. For each geometrical isomer,
three local minima with different conformations of the
xantphos backbone and phenyl groups were located (for details
see the Supporting Information). The most stable conforma-
tions of 4a,b differ in free energy by only ca. 2 kJ mol−1 with 4b
conversion of 5 into the neutral species 6, for which a weak IR
band was apparent at ca. 1957 cm−1 during the reaction.
Nucleophilic attack by 6 on MeI will give the Rh(III) methyl
cation 4b, presuming that the SN2 transition state involves
“end-on” approach by methyl iodide toward the open face of 6,
trans to oxygen. The ν(CO) band for 4b at 2084 cm−1
6173
dx.doi.org/10.1021/om2006968|Organometallics 2011, 30, 6166−6179

Organometallics
Article
slightly favored, consistent with isomerization of 4a to 4b
(Scheme 3), whereas all conformations of 4c are at least 20 kJ
mol−1 less stable. The computed ν(CO) values for the most
stable conformations of 4a,b show shifts comparable to those in
the experimental IR spectra.
these parameters will arise from a combination of the
thermodynamics for the initial methylation step and the
activation barrier for methyl migration. The large negative
entropy of activation arises from the associative nature of the
first step.
Consistent with the mechanism shown in Scheme 7, the
relative concentration of the intermediate 4b (judged by its
ν(CO) band intensity) increased in proportion to [MeI].
Addition of excess Bu4NI had the opposite effect, indicating
that the rate of consumption of 4b by the k−1 and k2 steps is
enhanced at higher iodide concentration. A thorough analysis
of the kinetics of these processes is complicated by issues such
Figure 10 shows a series of IR spectra recorded during the
reaction of MeI with 6 that was generated in situ by addition of
as ion pairing equilibria and the presence of both BF4 and I−
−
as alternative counterions for 4b. However, the kinetic profiles
for complexes 6, 4b, and 1 were modeled satisfactorily with
values of k1 = 0.013 M−1 s−1, k−1 = 0.5 s−1, and k2 = 0.05 s−1 (25
°C), where k−1 and k2 are taken to be pseudo-first-order rate
constants in the presence of excess Bu4NI (10 mol equiv with
respect to Rh).
The reaction sequence shown in Scheme 7 was also found to
occur when acetonitrile was used as the solvent, but with a
significantly faster overall rate. Thus, at 25 °C, an apparent
second-order rate constant of 9.9 × 10−3 M−1 s−1 was obtained,
15 times larger than that measured in CH2Cl2. This is not
unexpected, since polar coordinating solvents typically
accelerate both nucleophilic substitution and migratory CO
insertion reactions. In contrast to the behavior of 6, the o-tolyl-
xantphos analogue 6otol was unreactive toward MeI. As noted
earlier, the X-ray crystal structure of 6otol shows that access to
the Rh center is severely congested by two of the o-tolyl
substituents, which will inhibit an SN2 reaction with MeI.
Catalytic Mechanism. On the basis of the results
presented above, the catalytic cycle shown in Scheme 8 is
proposed, comprising a sequence of steps resembling that for
the [Rh(CO)2I2]−-catalyzed process. Methyl iodide, arising
from the reaction of methanol with HI, enters the cycle by
reacting with 6. This neutral Rh(I) complex has greater
nucleophilic character than the cation 5, which is therefore
positioned off the main cycle, in equilibrium with 6.
Nucleophilic attack by 6 on MeI leads to a cationic Rh(III)
methyl complex, 4b, and subsequent migratory CO insertion
and coordination of iodide gives the acetyl complex 1, which
was isolated after catalytic reactions. Since complex 1 was found
to be relatively unreactive, it is considered to be a catalyst
“resting state” and placed off the main cycle. Re-entry to the
cycle by loss of iodide enables coordination of CO to give 3.
The decomposition of this complex has not been examined in
detail, and it could proceed via reductive elimination of acetyl
iodide or by direct hydrolysis of the Rh acetyl.
Figure 10. Series of IR spectra recorded with a 5 min time interval
during the reaction of 6 with MeI (0.8 M in CH2Cl2, 25 °C).
an equimolar quantity of Bu4NI to 5[BF4] in CH2Cl2. The
decay of the Rh(I) ν(CO) band at 1957 cm−1 is accompanied
by the growth of a band at 1683 cm−1 due to the acetyl complex
1, corresponding to the reaction sequence shown in Scheme 7.
A weak band at 2085 cm−1 is assigned to the Rh(III) methyl
complex 4b, present as an intermediate. The decay of the
ν(CO) band of 6 was well fitted by an exponential curve to give
the pseudo-first-order rate constant kobs. Values of kobs obtained
from kinetic experiments conducted over a range of MeI
concentrations showed a first-order dependence on [MeI] (see
Supporting Information), and the slope of a plot of kobs vs
[MeI] gave an apparent second-order rate constant of 6.4 ×
10−4 M−1 s−1 (25 °C). Although this is ca. 15 times smaller than
the lower bound for k1 estimated above, the two observations
can be reconciled if the reverse (k−1) reaction of 4b is
competitive with the forward (k2) step. This was confirmed by
the observation that 4b[BF4] reacts with Bu4NI to form 6 as
the major product along with a small amount of 1, indicating
that methyl abstraction from 4b (i.e., the k−1 step) occurs at a
faster rate than migratory CO insertion (k2 step). Hence, the
initial methylation of the Rh center can be regarded as a
relatively rapid pre-equilibrium and a reasonable approximation
of kobs is provided by the expression kobs = (k1k2[MeI]/k−1). An
Eyring plot of variable-temperature kinetic data gave apparent
The activation parameters determined for the reaction of 6
with MeI allow a rate to be estimated at the temperature used
for catalysis. Extrapolation to 190 °C gives a second-order rate
constant of 0.49 M−1 s−1, which would give a turnover
activation parameters ΔH⧧ = 41(±1) kJ mol−1 and ΔS⧧
−166(±4) J mol−1 K−1. According to the analysis given above,
=
Scheme 7. Mechanism for the Reaction of MeI with [Rh(CO)(xantphos)I] (6)
6174
dx.doi.org/10.1021/om2006968|Organometallics 2011, 30, 6166−6179

Organometallics
Article
Scheme 8. Proposed Mechanism for Methanol Carbonylation Catalyzed by Rh xantphos Complexes
frequency of ca. 2000 h−1 at the MeI concentration (1.17 M)
used in catalytic tests. This is similar to the turnover frequency
of 2580 h−1 derived from the observed catalytic rate (14.8 mol
dm−3 h−1 at 5.73 × 10−3 M Rh),40 suggesting that the net
reaction 6 + MeI → 1 might be rate-determining in the catalytic
carbonylation reaction. However, the different reaction media
for stoichiometric and catalytic reactions and the uncertainty
regarding complex speciation under catalytic conditions mean
that caution should be exercised when drawing conclusions
from such an extrapolation.
The large amount of methane formed in catalytic reactions
using a CO/H2 gas feed can arise via hydrogenolysis of the Rh
methyl species 4b (or its isomer 4a), which is not as short-lived
as the transient [Rh(CO)2I3Me]− intermediate in the conven-
tional rhodium-catalyzed process.51,74 Experimental evidence
for such a hydrogenolysis reaction was found under mild
conditions. A question that arises from the observed product
selectivity is why the Rh methyl species appears to be more
prone to hydrogenolysis than the Rh acetyl species. It can be
speculated that creation of a vacant coordination site to activate
H2 occurs more readily via CO loss from 4b (with relatively
weak Rh−CO π back-bonding in the cation) than via iodide
loss from 1. Other methane-formation routes might also be
possible in the catalytic reaction. For example, a hydrido
complex (as in hydroformylation reactions39) could result from
the reaction 6 + H2 + CO ⇌ [Rh(xantphos)(CO)2H] + HI,
and subsequent reaction of the hydride with MeI would provide
a potential pathway to methane. Since HI can be converted into
MeI by reaction with methanol, a catalytic cycle can be drawn
for MeOH + H2 → CH4 + H2O based on 6. The rate of such a
process will depend on the steady-state concentration of
[Rh(xantphos)(CO)2H] generated by the equilibrium shown
above. Related to this, it has been noted that acidic conditions
promote the conversion of [Rh(xantphos)(CO)2H] into 5 by
protonation of the hydride.53
Although the key features of the proposed mechanism are
consistent with experimental observations, variants can be
considered. The intermediates in Scheme 8 all have xantphos
coordinated as a mer-κ3-P,O,P or trans-κ2-P,P ligand. However,
species with xantphos coordinated in a cis-chelate manner
could potentially play a role, as suggested by the structure of
the dimeric Rh acetyl complex in Scheme 3 and by the
flexibility evident in the structural survey of xantphos
complexes (Figures 7 and 8).
CONCLUSIONS
■
The mechanism of Rh/xantphos-catalyzed methanol carbon-
ylation has been investigated using a combination of structural,
spectroscopic, kinetic, and theoretical methods. A catalyst
resting state, the Rh(III) acetyl complex [Rh(xantphos)-
(COMe)I2] (1), has been isolated and shown to adopt an
approximately octahedral geometry with the xantphos ligand
coordinated in a “pincer” κ3-P,O,P fashion. This differs from
related acetyl complexes with cis-chelating diphosphines that
adopt square-pyramidal structures. Reactivity studies show that
abstraction of iodide from 1 allows coordination of CO or
solvent to the cationic species. Alternatively, migratory
deinsertion leads to a Rh(III) methyl complex that can react
with H2 to liberate methane or reductively eliminate methyl
iodide. The reactivity of methyl iodide toward rhodium(I)
carbonyl complexes containing xantphos has been probed. The
cation [Rh(xantphos)(CO)]+ (5) is relatively unreactive, but
coordination of iodide to give [Rh(xantphos)(CO)I] (6)
results in much faster nucleophilic attack by the Rh center on
MeI to give [Rh(xantphos)(CO)(I)(Me)]+ (4b) and subse-
quent migratory CO insertion to give 1. An interaction between
rhodium and the xanthene oxygen is thought to stabilize the
SN2 transition state, enhancing the nucleophilicity in
comparison to structurally related Vaska-type complexes. The
sterically demanding o-tolyl xantphos ligand causes structural
distortions in [Rh(o-tolyl-xantphos)(CO)X] and disfavors
6175
dx.doi.org/10.1021/om2006968|Organometallics 2011, 30, 6166−6179

Organometallics
Article
C43H38BF4INO2P2Rh: C, 52.74; H, 3.91; I, 12.96; N, 1.43. Found: C,
oxidative addition of MeI. The complexes observed in this
study (with the exception of one minor byproduct) exhibit
xantphos coordinated in a trans manner with P−Rh−P bite
angles ranging from 153 to 165° and Rh−O distances from
2.12 to 2.65 Å. A survey of xantphos complexes in the literature
reveals quite a smooth correlation between P−M−P bite angle
and M−O distance, with a division between cis and trans
coordination modes (bite angles <120° and >143°, respec-
tively). A catalytic mechanism based on the observed
stoichiometric reaction steps is proposed (Scheme 8).
52.32; H, 3.78; I, 13.13; N, 1.11. IR (CH2Cl2; ν(CO)/cm−1): 1708. 1H
4
NMR (CDCl3; δ): 7.3−8.1 (m, 26H, aromatics), 2.80 (t, 3H, JCH ‑P
=
3
1.4 Hz, COCH3), 1.92 and 1.76 (each s, 3H, xantphos-C(CH3)2).
31P{1H} NMR (CDCl3; δ): 16.5 (d, 1JRh−P = 100.7 Hz). TOF MS ES+
(m/z): 851 ([M − BF4 − MeCN]+), 823 ([M − BF4 − MeCN −
CO]+).
(c). [Rh(xantphos)(CO)I(Me)]BF4. A solution of Bu4NI (5.1 mg,
0.01 mmol) in CH2Cl2/MeI (5 cm3, 1:1 v:v) was added to
[Rh(CO)(xantphos)]BF4 (50.3 mg, 0.06 mmol) and stirred at room
temperature for 30 min, whereupon IR spectroscopy indicated the
disappearance of the reactant complex and formation of a single
ν(CO) band at 2084 cm−1. The resulting yellow-orange solution was
concentrated under vacuum to precipitate the product; yield 53 mg,
EXPERIMENTAL SECTION
■
Materials. The solvents acetonitrile, dichloromethane, diethyl
ether, hexane, and tetrahydrofuran were supplied from a Grubbs
solvent purification system.75 Other solvents were distilled and
degassed prior to use following literature methods. Synthetic
procedures were performed utilizing standard Schlenk techniques.
Xantphos was supplied by Aldrich and used without further
purification. Rhodium trichloride hydrate (RhCl3·xH2O) was supplied
by Precious Metals Online. Rhodium precursors [Rh(CO)2Cl]2,76
1
90%. IR (CH2Cl2; ν(CO)/cm−1): 2084. H NMR (CDCl3; δ): 7.5−
8.0 (m, 26H, aromatics), 2.18 (m, 3H, Rh−CH3), 1.91 and 1.87 (each
1
s, 3H, C(CH3)2). 31P{1H} NMR (CDCl3; δ): 13.2 (d, JRh−P = 86.0
Hz). TOF MS ES+ (m/z): 851 ([M − BF4]+).
Synthesis of Rh(I) Complexes. (a). [Rh(xantphos)(CO)Cl].
Addition of xantphos (610 mg, 1.05 mmol) to a stirred solution of
[RhCl(CO)2]2 (201.6 mg, 0.52 mmol) in EtOH (5 cm3) resulted in
immediate formation of a bright yellow precipitate. After the reaction
mixture was stirred for a further 10 min, the product was collected on a
sinter and washed with 10 cm3 portions of ice-cold EtOH and Et2O;
yield 752 mg (97%). Crystals suitable for X-ray diffraction were grown
by slow evaporation of diethyl ether into a concentrated CDCl3
solution. IR (CH2Cl2; ν(CO)/cm−1): 1968. 31P{1H} NMR (CDCl3;
[Rh(CO)2I]2,77 and [Rh(CO)(NCMe)(COMe)I2]2 and the ligand
44
o-tolyl-xantphos78 were synthesized according to literature procedures.
Methyl iodide (Aldrich) was distilled over calcium hydride and stored
in the refrigerator in a foil-wrapped Schlenk tube under nitrogen and
over mercury to prevent formation of I2.
Instrumentation. Infrared spectra were measured using a
Mattson Genesis Series FTIR spectrometer, controlled by WINFIRST
software, or a Perkin-Elmer Spectrum GX FTIR spectrometer,
controlled using Spectrum software. Solution spectra were recorded
using a 0.5 mm path length liquid cell with CaF2 windows. NMR
spectra were recorded on a Bruker AC-250 spectrometer fitted with a
Bruker B-ACS60 automated sample changer and operating in pulse
Fourier transform mode, using the solvent as reference. Spectra were
analyzed using Bruker 1D WINNMR software. Time-of-flight
electrospray mass spectrometry (TOF-ES-MS) was performed using
a Bruker Reflex III instrument and fast atom bombardment mass
spectrometry (FAB-MS) was carried out on a VG AutoSpec
instrument. Gas chromatography was performed using a Perkin-
Elmer ARNEL AutoSystem XL gas chromatograph. Elemental analyses
were performed using a Perkin-Elmer 2400 Elemental Analyzer.
Synthesis of Rh(III) Complexes. (a). [Rh(xantphos)(COMe)I2]
(1). [Rh(CO)(NCMe)(COMe)I2]2 (0.83 g, 0.88 mmol) and
xantphos (1.03 g, 1.78 mmol) were dissolved in CH2Cl2 (20 cm3)
and the solution was heated to reflux for 2.5 h, whereupon IR
spectroscopy indicated the disappearance of [RhI2(COMe)(NCMe)-
(CO)]2 and the formation of a ν(CO) band at 1683 cm−1. The
reaction solution was concentrated in vacuo and diethyl ether added
until an orange precipitate began to form. The minimum amount of
CH2Cl2 was then added to redissolve the precipitate. Cooling the
solution to −10 °C overnight resulted in formation of the product as
an orange-red microcrystalline solid; yield 1.63 g (94%). Anal. Calcd
for C41H35I2O2P2Rh: C, 50.33; H, 3.61; I, 25.94. Found: C, 50.04; H,
1
δ): 21.5 (d, JRh−P = 130.4 Hz). TOF MS ES+ (m/z): 709 ([M −
Cl]+).
(b). [Rh(o-tol-xantphos)(CO)Cl]. A procedure analogous to that
described above for [RhCl(CO)(xantphos)] was employed, using o-
tol-xantphos (278 mg, 0.44 mmol), [RhCl(CO)2]2 (84.4 mg, 0.22
mmol), and EtOH (6 cm3); yield 277 mg (75%). Crystals suitable for
X-ray diffraction were grown by slow evaporation of EtOH into a
1
concentrated CH2Cl2 solution. IR (CH2Cl2; ν(CO)/cm−1): 1950. H
NMR (CDCl3): δ 7.59 (dd, 2H, J = 7.48, 1.68 Hz, xanthene Ar-H),
7.13 (t, 2H, J = 7.63 Hz, xanthene Ar-H), 7.0−7.4 (m, 18H,
aromatics), 2.73 and 2.44 (each br s, total 12H, C6H4−CH3), 1.70 and
1.55 (each s, 3H, C(CH3)2). 31P{1H} NMR (CDCl3): δ 20.2 (d, 1JRh−P
= 125.8 Hz). FAB+ (m/z): 765 ([M − Cl]+).
(c). [Rh(o-tol-xantphos)(CO)I]. A solution of [Rh(CO)(NCMe)-
(COMe)I2]2 (94 mg, 0.10 mmol) and o-tol-xantphos (128 mg, 0.20
mmol) in CH2Cl2 (20 cm3) was heated to reflux for 2 h, whereupon IR
spectroscopy indicated the partial disappearance of [RhI2(COMe)-
(NCMe)(CO)]2 and the formation of a new band at 1951 cm−1. Upon
further reflux overnight, IR spectroscopy indicated the presence of a
small amount of unreacted [Rh(CO)(NCMe)(COMe)I2]2. After
addition of excess o-tol-xantphos (19.3 mg, 0.03 mmol) and a further
reflux (1 h), the IR spectrum displayed a single intense band at 1951
cm−1. The reaction solution was concentrated in vacuo and the
product crystallized by addition of Et2O and cooling to −10 °C
overnight to yield a brown powder; yield 120 mg, 67%. Crystals
suitable for X-ray diffraction were grown by slow evaporation of
diethyl ether into a concentrated CH2Cl2 solution. IR (CH2Cl2;
1
3.38; I, 25.76. IR (CH2Cl2; ν(CO)/cm−1): 1683. H NMR (CDCl3;
1
ν(CO)/cm−1): 1951. H NMR (CDCl3): δ 7.67 (dd, 2H, J = 7.63,
4
δ): 7.25−7.74 (m, 26H, aromatics), 2.75 (t, 3H, JCH −P = 1.7 Hz,
3
COCH3), 1.65 (s, 6H, xantphos-C(CH3)2). 31P{1H} NMR (CDCl3;
δ): 9.73 (d, 1JRh−P = 109.1 Hz). TOF MS ES+ (m/z): 851 ([M − I]+),
823 ([M − I − CO]+), 681 ([M − 2I − COMe]+).
1.52 Hz, xanthene Ar-H), 7.0−7.40 (m, 20H, aromatics), 2.74 and 2.40
(each br s, total 12H, C6H4−CH3), 1.52 and 1.50 (each br s, 3H,
1
C(CH3)2). 31P{1H} NMR (CDCl3): δ 18.6 (d, JRh−P = 115.6 Hz).
TOF MS ES+ (m/z): 765 ([M − I]+). The same product could also
be synthesized by the reaction of o-tol-xantphos with [RhI(CO)2]2
using a procedure analogous to that described above for [Rh(o-tol-
xantphos)(CO)Cl].
(b). [Rh(xantphos)(NCMe)(COMe)I]BF4. [Rh(xantphos)(COMe)-
I2] (198.2 mg, 0.20 mmol) was dissolved in CH2Cl2/MeCN (15
cm3, 2:1 v:v) to give an orange solution. Addition of AgBF4 (40.2 mg,
0.21 mmol) resulted in precipitation of AgI and formation of a yellow-
orange solution. After the mixture was stirred (5 min), an IR spectrum
of the solution indicated the disappearance of the reactant complex
and the formation of a new ν(CO) band at 1708 cm−1. Filtration
through Celite and removal of the solvent in vacuo gave the product as
a yellow-orange crystalline powder; yield 183 mg (92%). Crystals
suitable for X-ray diffraction were grown by slow evaporation of
diethyl ether into a concentrated CDCl3 solution. Anal. Calcd for
(d). [Rh(xantphos)(CO)]BF4. The procedure was based upon that
described by Sandee et al.53 AgBF4 (249 mg, 1.28 mmol) was added to
a stirred solution of [Rh(xantphos)(CO)Cl] (752.5 mg, 1.01 mmol) in
CH2Cl2 (15 cm3) at room temperature. The deep orange solution
immediately turned pale yellow with precipitation of AgCl. After
stirring for a further 10 min the pale yellow solution was filtered
through Celite and concentrated in vacuo. Addition of pentane caused
6176
dx.doi.org/10.1021/om2006968|Organometallics 2011, 30, 6166−6179

Organometallics
Article
Table 3. Summary of Crystallographic Data
1
2[BF4]·2CHCl3
6otol·CH2Cl2
7·CH2Cl2
7otol
empirical formula
formula wt
T/K
C41H35I2O2P2Rh
978.34
C45H40BCl6F4INO2P2Rh
1218.04
C45H42Cl2IO2P2Rh
977.44
C41H34Cl3O2P2Rh
829.88
C44H40ClO2P2Rh
801.06
150(2)
150(2)
150(2)
100(2)
150(2)
cryst syst
space group
a/Å
monoclinic
P21/n
monoclinic
P21/n
monoclinic
P21/n
monoclinic
P21/n
monoclinic
P21/c
13.4747(8)
15.0544(8)
18.0993(10)
90
21.1082(11)
11.8180(6)
21.4341
11.4822(10)
21.8589(19)
16.7669(15)
90
11.952(13)
18.106(20)
17.351(19)
90
11.0600(6)
20.5799(11)
16.6732(8)
90
b/Å
c/Å
α/deg
90
β/deg
94.5040(10)
90
114.7440(10)
90
104.477(2)
90
95.470(18)
90
100.689(3)
90
γ/deg
V/Å3
3660.2(4)
4
4856.0(4)
4
4074.7(6)
4
95.470(18)
4
3729.2(3)
4
Z
μ/mm−1
2.276
1.437
1.424
0.792
0.653
cryst size/mm
0.28 × 0.18 × 0.08
0.12 × 0.04 × 0.04
0.18 × 0.14 × 0.08
0.38 × 0.28 × 0.18
0.20 × 0.11 × 0.04
no. of rflns: total/indep (Rint) 40 981/8388 (0.0618) 29 725/11 138 (0.0525)
46 315/9318 (0.1209) 81 829/15 673 (0.0349) 72 253/8671 (0.1568)
final R1
largest peak, hole/e Å−3
0.0323
0.0448
0.0609
0.0342
0.0522
0.740, −0.602
1.501, −0.888
1.279, −1.265
0.914, −1.127
0.684, −1.046
precipitation of pale beige powder which was collected and dried in
Stuttgart/Dresden pseudopotentials82,83 on Rh and I and the D95 V
basis set84 on all other atoms, supplemented by extra d functions on
phosphorus (exponent 0.60) and chlorine (exponent 0.75). Geometry
optimizations were performed using the default settings. Optimized
SN2 transition states had a single imaginary frequency corresponding
to the expected umbrella inversion of the methyl fragment. Scaling
factors were applied to the computed ν(CO) values for Rh(I) and
Rh(III) complexes for comparison with experimental values, as
indicated in Tables S8 and S9, respectively (Supporting Information).
An evaluation of scaling factors for a wider range of metal carbonyl
complexes will be published separately.
X-ray Crystallography. Data were collected on a Bruker Smart
CCD area detector with an Oxford Cryostream 600 low-temperature
system using Mo Kα radiation (λ = 0.710 73 Å). The structures were
solved by direct methods and refined by full-matrix least-squares
methods on F2. Hydrogen atoms were placed geometrically and
refined using a riding model (including torsional freedom for methyl
groups). Complex scattering factors were taken from the SHELXTL
program package.85 Crystallographic data are summarized in Table 3,
and CIF files are provided in the Supporting Information.
vacuo overnight; yield 791 mg, 98%. IR (CH2Cl2; ν(CO)/cm−1):
1
2014. H NMR (CDCl3): δ 7.9 (dd, 2H, J = 7.02, 2.44 Hz, xanthene
Ar-H), 7.5−7.7 (m, 24H, aromatics), 1.8 (s, 6H, C(CH3)2). 31P{1H}
NMR (CDCl3): δ 37.1 (d, 1JRh−P = 121.2 Hz). FAB+ (m/z): 709 ([M
− BF4]+), 681 ([M − BF4 − CO]+).
(e). [Rh(o-tol-xantphos)(CO)]BF4. A procedure analogous to that
described above for [Rh(CO)(xantphos)]BF4 was employed, using
AgBF4 (34.2 mg, 0.18 mmol), [Rh(o-tol-xantphos)(CO)Cl] (119.6
mg, 0.15 mmol), and CH2Cl2 (20 cm3); yield 113 mg, 89%. Crystals
suitable for X-ray diffraction were grown by slow evaporation of EtOH
into a concentrated CH2Cl2 solution. IR (CH2Cl2; ν(CO)/cm−1):
1
2010. H NMR (CDCl3): δ 8.0 (dd, 2H, J = 7.63, 1.22 Hz, xanthene
Ar-H), 7.5 (t, 2H, J = 7.63 Hz, xanthene Ar-H), 6.9−7.4 (m, 18H,
aromatics), 2.6 (br s, total 12H, C6H4−CH3), 1.8 (br s, 6H, C(CH3)2).
31P{1H} NMR (CDCl3): δ 25.1 (d, 1JRh−P = 120.2 Hz). TOF MS ES+
(m/z): 765 ([M − BF4]+), 737 ([M − BF4 − CO]+).
Kinetics. Samples for kinetic runs were prepared by placing the
required amount of freshly distilled methyl iodide in a 5 cm3 graduated
flask, which was then made up to the mark with the solvent of choice.
A portion of this solution was used to record a background spectrum.
Another portion (typically 500 μL) was added to the solid metal
complex (typically 7−8 μmol) in a sample vial to give a reaction
solution with a complex concentration of ca. 15 mM. An appropriate
quantity of Bu4NI was added as required. A portion of the reaction
solution was quickly transferred to the IR cell, and the kinetic
experiment was started. In order to obtain pseudo-first-order
conditions, at least a 10-fold excess of MeI was used, relative to the
metal complex. The IR cell (0.5 mm path length, CaF2 windows) was
maintained at constant temperature throughout the kinetic run by a
thermostated jacket. Spectra were scanned in the ν(CO) region
(2200−1600 cm−1) and saved at regular time intervals under computer
control. After the kinetic run, absorbance vs time data for the
appropriate ν(CO) frequencies were extracted and analyzed off-line
using Kaleidagraph curve-fitting software. Exponential fits to the decay
traces for the ν(CO) band of 6 had correlation coefficients ≥0.999,
giving pseudo-first-order rate constants (listed in the Supporting
Information). Each kinetic run was repeated at least twice to check
reproducibility, the kobs values given being averaged values with
component measurements deviating from each other by ≤5%.
Computational Details. Density functional theory (DFT)
calculations were performed using the Gaussian 0379 and Gaussian
0980 program packages, compiled respectively using the Intel ifc
compiler (version 7.1) and the Portland compiler (version 8.0-2) on
an EMT64 architecture using Gaussian-supplied versions of BLAS and
ATLAS. All calculations employed the B3LYP functional81 with
ASSOCIATED CONTENT
■
S
* Supporting Information
CIF files giving X-ray crystallographic data and tables and
figures giving kinetic data, CSD structure search listings,
Cartesian coordinates, computed energies and ν(CO) values
for DFT-optimized structures. This material is available free of
charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
■
Corresponding Author
*E-mail: a.haynes@sheffield.ac.uk.
ACKNOWLEDGMENTS
■
We thank BP Chemicals and the EPSRC for supporting this
research.
REFERENCES
■
(1) Forster, D. Adv. Organomet. Chem. 1979, 17, 255.
(2) Dekleva, T. W.; Forster, D. Adv. Catal. 1986, 34, 81.
(3) Howard, M. J.; Jones, M. D.; Roberts, M. S.; Taylor, S. A. Catal.
Today 1993, 18, 325.
6177
dx.doi.org/10.1021/om2006968|Organometallics 2011, 30, 6166−6179

Organometallics
Article
(4) Maitlis, P. M.; Haynes, A.; Sunley, G. J.; Howard, M. J. J. Chem.
Soc., Dalton Trans. 1996, 2187.
(5) Yoneda, N.; Kusano, S.; Yasui, M.; Pujado, P.; Wilcher, S. Appl.
(39) Kranenburg, M.; Vanderburgt, Y. E. M.; Kamer, P. C. J.; van
Leeuwen, P. W. N. M; Goubitz, K.; Fraanje, J. Organometallics 1995,
14, 3081.
(40) Gaemers, S.; Sunley, J. G. World Patent WO2004101488 (BP
Catal., A 2001, 221, 253.
Chemicals Limited), 2004.
(6) Torrence, P. In Applied Homogeneous Catalysis with Organo-
metallic Compounds, 2nd ed.; Cornils, B., Herrmann, W. A., Eds.;
Wiley-VCH: Weinheim, Germany, 2002; Vol. 1, p 104.
(41) It was suggested29 that reductive elimination of acetyl iodide
becomes the rate-limiting step for Rh catalysts with dppe and related
C2-bridged diphosphines.
(7) Thomas, C. M.; Suss-Fink, G. Coord. Chem. Rev. 2003, 243, 125.
̈
(42) Deb, B.; Dutta, D. K. J. Mol. Catal. A: Chem. 2010, 326, 21.
(43) Catalytic conditions: MeOH (78.47 g, 2.45 mol), MeI (14.49 g,
0.102 mol), xantphos (1.571 g, 2.715 mmol), [Rh(CO)2(acac)] (0.646
g, 2.494 mmol, 68 bar, 2:1 H2:CO, 140 °C).
(44) Haynes, A.; Maitlis, P. M.; Stanbridge, I. A.; Haak, S.; Pearson, J.
M.; Adams, H.; Bailey, N. A. Inorg. Chim. Acta 2004, 357, 3027.
(45) Adams, H.; Bailey, N. A.; Mann, B. E.; Manuel, C. P. Inorg.
Chim. Acta 1992, 198−200, 111.
(46) Nieczypor, P.; van Leeuwen, P.; Mol, J. C.; Lutz, M.; Spek, A. L.
J. Organomet. Chem. 2001, 625, 58.
(47) Noveski, D.; Braun, T.; Neumann, B.; Stammler, A.; Stammler,
H. G. Dalton Trans. 2004, 4106.
(48) Adams, H.; Bailey, N. A.; Mann, B. E.; Manuel, C. P.; Spencer,
C. M.; Kent, A. G. J. Chem. Soc., Dalton Trans. 1988, 489.
(49) Grushin, V. V.; Marshall, W. J. J. Am. Chem. Soc. 2006, 128,
12644.
(8) Morris, G. E. In Mechanisms in Homogeneous Catalysis; A
Spectroscopic Approach; Heaton, B. T., Ed.; Wiley-VCH: Weinheim,
Germany, 2005; p 195.
(9) Haynes, A. Top. Organomet. Chem. 2006, 18, 179.
(10) Haynes, A. In Advances in Catalysis; Elsevier Academic Press:
San Diego, CA, 2010; Vol. 53, p 1.
(11) Paulik, F. E.; Roth, J. F. Chem. Commun. 1968, 1578.
(12) Brodzki, D.; Leclere, C.; Denise, B.; Pannetier, G. Bull. Soc.
Chim. Fr. 1976, 61.
(13) Paulik, F. E.; Hershman, A.; Knox, W. R.; Roth, J. F., U.S.
Patent 4,690,912 (Monsanto Corporation), 1987.
(14) Bader, A.; Lindner, E. Coord. Chem. Rev. 1991, 108, 27.
(15) Dutta, D. K.; Woollins, J. D.; Slawin, A. M. Z.; Konwar, D.; Das,
P.; Sharma, M.; Bhattacharyya, P.; Aucott, S. M. Dalton Trans. 2003,
2674.
(16) Rankin, J.; Benyei, A. C.; Poole, A. D.; Cole-Hamilton, D. J. J.
Chem. Soc., Dalton Trans. 1999, 3771.
(17) Rankin, J.; Poole, A. D.; Benyei, A. C.; Cole-Hamilton, D. J.
Chem. Commun. 1997, 1835.
(18) Wegman, R. W.; Abatjoglou, A. G.; Harrison, A. M. J. Chem.
Soc., Chem. Commun. 1987, 1891.
(19) Wegman, R. W.; Abatjoglou, A. G. U.S. Patent 4,670,570
(Union Carbide Corporation), 1987.
(20) Cavell, R. G.; Katti, K. V. World Patent WO9204118, 1992.
(21) Baker, M. J.; Dilworth, J. R.; Sunley, J. G.; Wheatley, N. Eur.
Patent EP0632006A (BP Chemicals Limited), 1995.
(22) Baker, M. J.; Giles, M. F.; Orpen, A. G.; Taylor, M. J.; Watt, R. J.
J. Chem. Soc., Chem. Commun. 1995, 197.
(50) Parr, M. L.; Perez-Acosta, C.; Faller, J. W. New J. Chem. 2005,
29, 613.
(51) Haynes, A.; Mann, B. E.; Morris, G. E.; Maitlis, P. M. J. Am.
Chem. Soc. 1993, 115, 4093.
(52) The retention of some 13CO in the product probably arises
from scrambling of label between terminal and acetyl carbonyl
positions in the precursor over the time scale of the reaction (as was
demonstrated to occur previously51).
(53) Sandee, A. J.; Reek, J. N. H.; Kamer, P. C. J.; van Leeuwen, P. J.
Am. Chem. Soc. 2001, 123, 8468.
(54) van der Vlugt, J. I.; Sablong, R.; Mills, A. M.; Kooijman, H.;
Spek, A. L.; Meetsma, A.; Vogt, D. Dalton Trans. 2003, 4690.
(55) van der Vlugt, J. I.; Sablong, R.; Magusin, P.; Mills, A. M.; Spek,
A. L.; Vogt, D. Organometallics 2004, 23, 3177.
(56) Otto, S.; Roodt, A. Inorg. Chim. Acta 2004, 357, 1.
(57) Sandee, A. J.; van der Veen, L. A.; Reek, J. N. H.; Kamer, P. C.
J.; Lutz, M.; Spek, A. L.; van Leeuwen, P. Angew. Chem., Int. Ed. 1999,
38, 3231.
(23) Dilworth, J. R.; Miller, J. R.; Wheatley, N.; Baker, M. J.; Sunley,
J. G. J. Chem. Soc., Chem. Commun. 1995, 1579.
(24) Gonsalvi, L.; Adams, H.; Sunley, G. J.; Ditzel, E.; Haynes, A. J.
Am. Chem. Soc. 1999, 121, 11233.
(25) Gonsalvi, L.; Adams, H.; Sunley, G. J.; Ditzel, E.; Haynes, A. J.
Am. Chem. Soc. 2002, 124, 13597.
(58) The bandwidth (fwhm) for the ν(CO) absorption of 6 in
(26) McConnell, A. C.; Foster, D. F.; Pogorzelec, P. J.; Slawin, A. M.
Z.; Law, D. J.; Cole-Hamilton, D. J. Dalton Trans. 2003, 510.
(27) McConnell, A. C.; Pogorzelec, P. J.; Slawin, A. M. Z.; Williams,
G. L.; Elliott, P. I. P.; Haynes, A.; Marr, A. C.; Cole-Hamilton, D. J.
Dalton Trans. 2006, 91.
(28) Baker, M. J.; Carraz, C. A.; Ditzel, E.; Pringle, P. G.; Sunley, G.
J. GB Patent GB2336154 (BP Chemicals Limited), 1999.
(29) Carraz, C. A.; Ditzel, E.; Orpen, A. G.; Ellis, D. D.; Pringle, P.
G.; Sunley, G. J. Chem. Commun. 2000, 1277.
(30) Lamb, G.; Clarke, M.; Slawin, A. M. Z.; Williams, B.; Key, L.
Dalton Trans. 2007, 5582.
(31) Lamb, G.; Clarke, M.; Slawin, A. M. Z.; Williams, B.; Key, L.
Dalton Trans. 2008, 4946.
(32) Jimenez-Rodriguez, C.; Pogorzelec, P. J.; Eastham, G. R.;
Slawin, A. M. Z.; Cole-Hamilton, D. J. Dalton Trans. 2007, 4160.
(33) Thomas, C. M.; Mafua, R.; Therrien, B.; Rusanov, E.; Stœckli-
CH2Cl2 was measured as 29 cm−1, compared to 22 cm−1 for 6otol and
26 cm−1 for both 5 and 5otol
.
(59) Deconvolution was achieved using the peak-resolve function in
OMNIC software (Nicolet).
(60) Allen, F. H. Acta Crystallogr. 2002, B58, 380.
(61) Orpen, A. G. Acta Crystallogr. 2002, B58, 398.
(62) The CSD search was conducted using the Conquest package to
locate structures with 4,5-bis(diphenylphosphino)-9,9′-dimethylxan-
thene acting as a chelating ligand to a single transition-metal center.
The o-tolyl-xantphos complexes in this study and one structure that
contained P(3,5-Me2C6H3)2 groups was included in the analysis, but
other substituents on phosphorus or modifications to the backbone
were excluded. For structures with more than one independent
molecule in the unit cell, average distances and angles were used for
the plots in Figures 7 and 8.
(63) Pontiggia, A. J.; Chaplin, A. B.; Weller, A. S. J. Organomet. Chem.
Evans, H.; Suss-Fink, G. Chem. Eur. J. 2002, 8, 3343.
̈
2011, 696, 2870.
(34) Burger, S.; Therrien, B.; Suss-Fink, G. Helv. Chim. Acta 2005,
̈
(64) Ledger, A. E. W.; Moreno, A.; Ellul, C. E.; Mahon, M. F.;
Pregosin, P. S.; Whittlesey, M. K.; Williams, J. M. J. Inorg. Chem. 2010,
49, 7244.
(65) Basson, S. S.; Leipoldt, J. G.; Purcell, W.; Schoeman, J. B. Inorg.
Chim. Acta 1990, 173, 155.
(66) de Waal, D. J. A.; Gerber, T. I. A.; Louw, W. J. Inorg. Chem.
1982, 21, 1259.
(67) Forster, D. J. Am. Chem. Soc. 1975, 97, 951.
88, 478.
(35) Freixa, Z.; Kamer, P. C. J.; Lutz, M.; Spek, A. L.; van Leeuwen,
P. W. N. M. Angew. Chem., Int. Ed. 2005, 44, 4385.
(36) Moloy, K. G.; Wegman, R. W. Organometallics 1989, 8, 2883.
(37) Moloy, K. G.; Petersen, J. L. Organometallics 1995, 14, 2931.
(38) Gaemers, S.; Sunley, J. G. World Patent WO2004101487 (BP
Chemicals Limited), 2004.
6178
dx.doi.org/10.1021/om2006968|Organometallics 2011, 30, 6166−6179

Organometallics
Article
(68) For example, for the reaction of 5[BF4] (13.3 mM) with MeI
(0.8 M) in CH2Cl2 containing Bu4NI (2.17 mM) at 25 °C, the decay
of the ν(CO) band of 5 corresponded to a rate of 23 μM s−1. This rate
can be equated to k[6][MeI] on the basis of the mechanism shown in
Scheme 6. Assuming a maximum concentration of 6 equal to the
added [Bu4NI], a lower bound for the rate constant for nucleophilic
attack by 6 on MeI is k = 0.013 M−1 s−1 under these conditions.
(69) Douek, I. C.; Wilkinson, G. J. Chem. Soc. A 1969, 2604.
(70) Franks, S.; Hartley, F. R.; Chipperfield, J. R. Inorg. Chem. 1981,
20, 3238.
(71) Miller, E. M.; Shaw, B. L. J. Chem. Soc., Dalton Trans. 1974, 480.
(72) Constable, A. G.; Langrick, C. R.; Shabanzadeh, B.; Shaw, B. L.
Inorg. Chim. Acta Lett. 1982, 65, L151.
(73) Best, J.; Wilson, J. M.; Adams, H.; Gonsalvi, L.; Peruzzini, M.;
Haynes, A. Organometallics 2007, 26, 1960.
(74) Haynes, A.; Mann, B. E.; Gulliver, D. J.; Morris, G. E.; Maitlis,
P. M. J. Am. Chem. Soc. 1991, 113, 8567.
(75) Pangborn, A. B.; Giardello, M. A.; Grubbs, R. H.; Rosen, R. K.;
Timmers, F. J. Organometallics 1996, 15, 1518.
(76) McCleverty, J.; Wilkinson, G. Inorg. Synth. 1966, 8, 214.
(77) Fulford, A.; Hickey, C. E.; Maitlis, P. M. J. Organomet. Chem.
1990, 398, 311.
(78) Hamann, B. C.; Hartwig, J. F. J. Am. Chem. Soc. 1998, 120, 3694.
(79) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;
Robb, M. A.; Cheeseman, J. R.; Montgomery, J., J A; Vreven, T.;
Kudin, K. N.; Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.;
Barone, V.; Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.;
Petersson, G. A.; Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.;
Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao,
O.; Nakai, H.; Klene, M.; Li, X.; Knox, J. E.; Hratchian, H. P.; Cross, J.
B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R.
E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.;
Ayala, P. Y.; Morokuma, K.; Voth, G. A.; Salvador, P.; Dannenberg, J.
J.; Zakrzewski, V. G.; Dapprich, S.; Daniels, A. D.; Strain, M. C.;
Farkas, O.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman,
J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.; Clifford, S.; Cioslowski, J.;
Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.;
Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.;
Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.; Johnson, B.; Chen,
W.; Wong, M. W.; Gonzalez, C.; Pople, J. A. ,Gaussian 03, Revision
B.05; Gaussian, Inc., Wallingford, CT, 2004.
(80) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;
Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci,
B.; Petersson, G. A.; Nakatsuji, H.; Caricato, M.; Li, X.; Hratchian, H.
P.; Izmaylov, A. F.; Bloino, J.; Zheng, G.; Sonnenberg, J. L.; Hada, M.;
Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima,
T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven, T.; Montgomery, J., J A;
Peralta, J. E.; Ogliaro, F.; Bearpark, M.; Heyd, J. J.; Brothers, E.; Kudin,
K. N.; Staroverov, V. N.; Kobayashi, R.; Normand, J.; Raghavachari, K.;
Rendell, A.; Burant, J. C.; Iyengar, S. S.; Tomasi, J.; Cossi, M.; Rega,
N.; Millam, N. J.; Klene, M.; Knox, J. E.; Cross, J. B.; Bakken, V.;
Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.;
Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Martin, R. L.;
Morokuma, K.; Zakrzewski, V. G.; Voth, G. A.; Salvador, P.;
̈
Dannenberg, J. J.; Dapprich, S.; Daniels, A. D.; Farkas, O.;
Foresman, J. B.; Ortiz, J. V.; Cioslowski, J.; Fox, D. J. Gaussian 09,
Revision A.02; Gaussian, Inc., Wallingford, CT, 2009.
(81) Becke, A. D. J. Chem. Phys. 1993, 98, 5648.
(82) Nicklass, A.; Dolg, M.; Stoll, H.; Preuss, H. J. Chem. Phys. 1995,
115, 7348 and references therein..
(83) Cao, X.; Dolg, M. J. Chem. Phys. 2001, 115, 7348.
(84) Dunning, T. H., Jr.; Hay, P. J. In Modern Theoretical Chemistry;
Schaefer, H. F., III, Ed.; Plenum: New York, 1976; Vol. 3, p 1.
(85) Sheldrick, G. M. SHELXL97, An Integrated System for Solving
and Refining Crystal Structures from Diffraction Data; University of
Gottingen, Gottingen, Germany, 1997.
̈
6179
dx.doi.org/10.1021/om2006968|Organometallics 2011, 30, 6166−6179