Carbon monoxide induced double cyclometalation at the iridium center

Article abstract of DOI:10.1021/om300506v

Bubbling of CO into a dichloromethane solution of [Ir(COD)(CH ₃CN)₂][BF₄] followed by the addition of 2-phenyl-1,8-naphthyridine (LH) at room temperature results in the bis-cyclometalated Ir^III complex [Ir(C^{^}N) ₂(CO)(LH)][BF₄] (C^{^}N = L). The observed cyclometalation contradicts the classical role of CO, which is to hinder oxidative addition by lowering electron density on the metal. DFT calculations reveal that the first cyclometalation involves oxidative addition of the ligand. Subsequently, preferential electrophilic activation of the second ligand followed by elimination of dihydrogen affords the bis-cyclometalated Ir ^III complex.

Full text of DOI:10.1021/om300506v

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
Article
pubs.acs.org/Organometallics
Carbon Monoxide Induced Double Cyclometalation at the Iridium
Center
S. M. Wahidur Rahaman, Shrabani Dinda, Tapas Ghatak, and Jitendra K. Bera*
Department of Chemistry, Indian Institute of Technology Kanpur, Kanpur 208016, India
S
* Supporting Information
ABSTRACT: Bubbling of CO into a dichloromethane
solution of [Ir(COD)(CH3CN)2][BF4] followed by the
addition of 2-phenyl-1,8-naphthyridine (LH) at room temper-
ature results in the bis-cyclometalated IrIII complex [Ir-
(C∧N)2(CO)(LH)][BF4] (C∧N = L). The observed cyclo-
metalation contradicts the classical role of CO, which is to
hinder oxidative addition by lowering electron density on the
metal. DFT calculations reveal that the first cyclometalation
involves oxidative addition of the ligand. Subsequently, preferential electrophilic activation of the second ligand followed by
elimination of dihydrogen affords the bis-cyclometalated IrIII complex.
The OA of the C−H bond to a transition metal involves
electron donation from the C−H σ orbital to a vacant metal d
orbital and back-donation from a filled metal d orbital to the
C−H σ* orbital. A π-acceptor CO withdraws electron density
from the metal center and thus normally hinders OA and
facilitates reductive elimination (RE).8 In contrast with this
classical behavior, Milstein, Martin, and co-workers reported
CO-promoted OA of the arene C−H bond to a cationic RhI
complex containing a phosphine-based pincer ligand.9 We have
observed double cyclometalation on a simple IrI system at room
temperature which is facilitated by CO. The reaction protocol
that allows CO-induced double cyclometalation is described,
and a mechanistic scheme is outlined with the aid of DFT
calculations.
INTRODUCTION
■
The growing interest in cyclometalation reactions originates
from their ability to provide easy access to organometallic
compounds with a metal−carbon σ bond.1 The reaction
typically involves initial coordination of the ligand to the metal
via a donor group followed by intramolecular activation of the
C−H bond. Depending on the electronic nature of the metal
and the set of ligands bound to it, several possible pathways
have been suggested for the C−H bond cleavage.2 Oxidative
addition (OA) is favored for electron-rich, low-valent transition
metals, leading to the corresponding alkyl or aryl hydride
complex with concomitant formal two-electron oxidation of the
metal. Transition metals that lack sufficient electron density for
OA employ alternate mechanisms which include σ-bond
metathesis, 1,2-addition, radical activation, and electrophilic
activation.3 For an electron-deficient metal bearing an alkyl or
hydride ligand, the σ-bond metathesis mechanism has been
invoked. Electrophilic activation is often facilitated by the use of
a Brønsted base such as acetate, carbonate, or hydride, which is
either added externally or bound to the metal center to begin
with.
RESULTS AND DISCUSSION
■
Synthesis. Reaction of NP-R (2-R-1,8-naphthyridine) with
[Ir(COD)(CH3CN)2][BF4] allowed the isolation of [Ir-
(COD)(NP-R)2][BF4] (R = Ph, ferrocenyl (Fc) and 2-
(OH)Ph (1)) (Scheme 1). Full characterizations of the
compounds [Ir(COD)(NP-Ph)2][BF4] and [Ir(COD)(NP-
Fc)2][BF4] have been reported earlier.10 The molecular
structure of 1 is very similar to that of those two complexes.
The cationic unit consists of a central Ir having one η2,η2-bound
COD and two NP ligands coordinated through the N8
nitrogens in a square-planar arrangement (Figure S1,
Supporting Information). A continuous flow of CO into
[Ir(COD)(NP-R)2][BF4] afforded different compounds, de-
pending on the nature of the substituent R and the choice of
solvent. For methyl and phenyl substituents, [Ir2(CO)4(μ-NP-
R)2][BF4]2 (R = Me, Ph) were isolated in dichloromethane.10
These dinuclear compounds have a characteristic dark blue
Bis-cyclometalated IrIII complexes constitute an important
class of compounds, owing to their rich photophysical
properties.4 The most prominent and widely studied complexes
are the ones which contain the “Ir(ppy)2” (ppyH = 2-
phenylpyridine) unit. These are conveniently synthesized from
the chloro-bridged IrIII dimer [{(ppy)2Ir}2(μ-Cl)2]. The double
cyclometalation in this dimer is achieved by refluxing
IrCl3·nH2O and ppyH in an alcohol−water mixture at elevated
temperature.5 Acetate-assisted cyclometalation reactions under
mild conditions are widely known for a variety of IrIII
complexes.6 In contrast, cyclometalation at IrI center is
surprisingly rare. The “{Ir(COD)}+” unit exhibits reluctance
toward OA of the C−H bond. Only in select cases have
cyclometalated IrIII hydride complexes been isolated.7
Received: June 7, 2012
Published: July 24, 2012
© 2012 American Chemical Society
5533
dx.doi.org/10.1021/om300506v | Organometallics 2012, 31, 5533−5540

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
Organometallics
Article
Scheme 1. Reactions of [Ir(COD)(CH3CN)2](BF4) with NP-R and CO
color and very low solubility in common organic solvents. On
the other hand, bulky ferrocenyl and napthyl substituents led to
direct replacement of COD by two CO ligands, resulting in
[Ir(CO)2(NP-R)2][BF4] (R = Fc, naphthyl (2)), respectively.
The molecular structure of 2 reveals a square-planar Ir complex
containing two carbonyls and two NP ligands bound via N8
nitrogens (Figure S2, Supporting Information). It should be
noted here that the ferrocenyl analogue [Ir(CO)2(NP-Fc)2]-
[BF4] was isolated by employing chlorobenzene as a solvent.
The identical reaction in dichloromethane provided the metal−
metal singly bonded diiridium(II) compound
[Ir2(CO)4(Cl)4(NP-Fc)2], in which the chlorides are shown
to originate from the solvent (Scheme 1).10
Reversal of the reaction sequence, that is, bubbling CO
through a dichloromethane solution of [Ir(COD)(CH3CN)2]-
[BF4] followed by the addition of NP-Ph, provided the unusual
double cyclometalated complex [Ir(NP-C6H4)2(NP-Ph)-
(CO)](BF4) (3). The molecular structure of 3 (Figure 1)
reveals a cationic complex consisting of a central Ir in an
octahedral geometry occupied by two cyclometalating (C∧N)
ligands, one CO, and one NP-Ph. The cyclometalation occurs
through the N1 nitrogen of the naphthyridine and the ortho
phenyl carbon. A NP-Ph ligand coordinates to the metal via the
NP N8 nitrogen to complete the octahedral geometry around
the metal. The Ir−C (Ir1−C15 = 2.048(5) Å and Ir1−C21 =
2.026(5) Å) and Ir−N (Ir1−N1 = 2.210(4) Å and Ir1−N3 =
2.150(4) Å) distances are consistent with bond parameters of
other bis-cyclometalated Ir(III) compounds.4,11 One significant
difference, however, is worth mentioning. The cis-C,C/cis-N,N
disposition of the cyclometalating ligands is in sharp contrast to
the normal cis-C,C/trans-N,N configurations observed for the
majority of these compounds.4a,b
The NMR data of 3 conform to the X-ray structure. Two
cyclometalated carbons show signals at δ 174 and 171 ppm in
13C NMR spectra. The carbonyl carbon appears at δ 207 ppm.
The IR stretching frequency at 2086 cm−1 confirms the
presence of a single CO. The ESI-MS spectrum of complex 3
exhibits signals at m/z (z = 1) 649, 631, and 603 attributed to
[M − -(NP-Ph) + (H2O)]+ (17%), [M − (NP-Ph)]+ (100%),
and [M − (NP-Ph) − CO]+ (60%) fragments.
Figure 1. ORTEP diagram (30% probability thermal ellipsoids) of the
cationic unit of 3 with important atoms labeled. Hydrogen atoms are
omitted for the sake of clarity. Selected bond distances (Å) and angles
(deg): Ir1−C1 = 1.846(5), Ir1−C21 = 2.026(5), Ir1−C15 = 2.048(5),
Ir1−N3 = 2.150(4), Ir1−N5 = 2.180(4), Ir1−N1 = 2.210(4); C1−
Ir1−C21 = 95.0(2), C1−Ir1−C15 = 89.0(2), C21−Ir1−C15 =
92.9(2), C1−Ir1−N3 = 173.96(19), C21−Ir1−N3 = 79.12(18),
C15−Ir1−N3 = 89.99(17), C1−Ir1−N5 = 94.99(18), C21−Ir1−N5
= 89.90(18), C15−Ir1−N5 = 174.91(17), N3−Ir1−N5 = 86.36(15),
C1−Ir1−N1 = 86.46(18), C21−Ir1−N1 = 171.04(18), C15−Ir1−N1
= 78.28(19), N3−Ir1−N1 = 99.16(15), N5−Ir1−N1 = 98.79(15).
The emission spectrum of 3 shows blue emission at 405 and
427 nm upon irradiation at 300 nm in acetonitrile (Figure S5a,
Supporting Information). The excited-state lifetime of 1.2 μs is
consistent with emission from a triplet excited state. A clear
vibronic structure indicates that the emissive state is most likely
to be 3LC (ligand-centered) involving the cyclometalating
ligands (C∧N). This is characteristic of bis-cyclometalated IrIII
compounds containing photophysically innocent ancillary
ligands.4
Reaction with the pyridine analogue 2-phenylpyridine
(Hppy) under identical conditions also led to cyclometalation.
All attempts to obtain crystals of the product suitable for X-ray
1
diffraction were unsuccessful. However, the H and 13C NMR
spectra revealed the double-cyclometalated complex [Ir-
5534
dx.doi.org/10.1021/om300506v | Organometallics 2012, 31, 5533−5540

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
Organometallics
Article
a
Scheme 2. Mechanistic Pathways for Double Cyclometalation Supported by DFT Calculations
a
Ea = activation energy (kcal/mol).
Figure 2. Optimized structures of intermediates and transition states.
5535
dx.doi.org/10.1021/om300506v | Organometallics 2012, 31, 5533−5540

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
Organometallics
Article
(ppy)2(CH3CN)(CO)](BF4) (4), which was further charac-
terized by IR, mass, and elemental analysis. A pair of signals at δ
169 and 165 ppm in 13C NMR spectra indicates the double
cyclometalation. The carbonyl carbon appears at δ 197 ppm.
The IR stretching frequency at 2037 cm−1 confirmed the
presence of CO. The ESI-MS spectrum of complex 4 reveals
the molecular ion signal at m/z 570 (10%). The other
prominent signals are attributed to [M − CH3CN]+ (100%)
and [M − CH3CN − CO]+ (64%) at m/z (z = 1) 529 and 501,
respectively. The elemental analysis data correspond to an
acetonitrile coordinated to the metal. The emission spectrum of
4 shows blue emission at 455 and 482 nm upon irradiation at
300 nm in acetonitrile (Figure S5b, Supporting Information)
with a corresponding excited-state lifetime of 1.2 μs. The
emission pattern is similar to that of 3 and likely originates from
(Ir−H1 = 1.546 Å). The transition state TSA→B connecting A
and B has one imaginary frequency of −809.4 cm−1
corresponding to the movement of hydrogen from carbon to
the metal center with short Ir···C/H (2.083/1.607 Å) and
elongated C−H separations (1.649 Å). The coordination of an
acetonitrile to B completes the stable octahedral intermediate
C.
Two distinct possibilities for the second C−H bond
activation have been considered, which include a second OA
of the C−H resulting in an IrV dihydride intermediate (defined
as path A) and an electrophilic activation of the C−H to give an
IrIII dihydrogen intermediate (defined as path B). Several
groups have shown that an associative process could proceed in
one of these two ways.13 The preference for a particular
pathway depends on the metal oxidation state and the ligand
properties. The agostic interaction results in enhanced C−H
acidity due to forward electron donation from the C−H σ
orbital to the empty metal orbital. Electrophilic activation may
operate when this electron donation is significant, which is
followed by proton abstraction by a base to complete the
cyclometalation. Alternatively, OA takes place when the back-
donation from the metal to the C−H σ* orbital is sufficiently
high. Both pathways are discussed independently in the
following section, and their energy profiles are compared in
Figure 3.
3
a LC emissive state. This is further confirmation of double
cyclometalation at the Ir center.
Clearly, no C−H activation takes place when COD is bound
to the metal. X-ray structures reveal ubiquitous N8
coordination of NP ligands in the solid state that put the aryl
C−H far away from the metal. Even for a possible N1
coordination, the bulky COD is not likely to allow a close
approach of the ortho C−H to the metal. In contrast, linearly
bound CO ligands may allow an agostic interaction between
the aryl C−H and the Ir while the latter is engaged to N1 of the
ligand. The result is possibly C−H OA to the metal, resulting in
a cyclometalated IrIII hydride complex. The second cyclo-
metalation can occur either via electrophilic activation or by
another OA. Finally the loss of dihydrogen would lead to the
bis-cyclometalated compound 3.
Mechanism of Double Cyclometalation. We realize that
our proposal of first C−H OA is counterintuitive, since it
contradicts the classical role of the π-acceptor CO, which
withdraws electron density from the metal and thus inhibits
OA. Further, the mechanism for the second cyclometalation is
not immediately obvious. To gain insight into the double
cyclometalation mechanism, a computational study was under-
taken using the DFT/B3LYP level of theory. 2-Phenylpyridine
(Hppy) was considered as the model ligand, which reduces the
computational cost and represents a generalized situation. The
use of Hppy is justified, since it afforded a similar bis-
cyclometalated compound. A mechanistic scheme supported by
DFT calculations is shown in Scheme 2. The optimized
structures of all the species are given in Figure 2.
Replacement of COD in [Ir(COD)(CH3CN)2][BF4] by two
CO followed by ligand addition led to a mononuclear
dicarbonyl complex [Ir(CO)2(Hppy)] (A), which is considered
as a potential starting intermediate. The C−H can approach the
metal in two ways, via front-phase or the rear-phase Ir···C−H
interactions (Figure S6, Supporting Information). However,
they constitute a pair of enantiomers, and their bond
parameters and relative energies are identical. Hence, we
restricted our study to only one case (rear-phase). Examination
of the metrical parameters of A reveals strong agostic
interaction of the aryl C−H bond to the metal (Ir···C/H =
2.387/2.482 Å) (Figure 2). A long C−H distance (1.094 Å)
and an acute Ir−H−C angle (72°) favor OA with an activation
barrier of 15.3 kcal/mol (Scheme 2). The intermediacy of
similar M···C−H agostic complexes (or C−H σ complexes) has
been proposed and identified along the reaction coordinates for
C−H activation.6,12 The resultant OA product B is a
cyclometalated IrIII hydride species having a square-pyramidal
geometry in which the hydride (H1) occupies an axial position
Figure 3. Comparative energy profiles for different pathways and the
corresponding transition states. Energies are in kcal/mol (not to
scale).
The double-cyclometalated product necessitates coordina-
tion of the second Hppy. For this purpose, C must lose
acetonitrile and carbonyl(s) from the metal coordination
sphere. We computed the five-coordinated hydride intermedi-
ate D, which undergoes ready OA to form the octahedral IrV
intermediate E. It should be noted here that both carbonyls and
acetonitrile are released from C to accommodate an incoming
ligand. The six-coordinated species D′ with an additional CO
attached at an equatorial site was also computed. However, all
our attempts to find a low-energy C−H OA step from D′ failed,
probably because of the difficulty associated with a seven-
coordinated species in the TS. The geometry of D consists of a
cyclometalated ppy, coordination of a Hppy via pyridine N
accompanied by an agostic C−H interaction to the metal
(Ir···C4/H4 = 2.328/2.482 Å), and finally an axial hydride H1
5536
dx.doi.org/10.1021/om300506v | Organometallics 2012, 31, 5533−5540