Controlling the selectivity of C-H activation in pyridinium triazolylidene iridium complexes: Mechanistic details and influence of remote substituents

Article abstract of DOI:10.1021/om501197g

Iridium complexes containing a triazolylidene ligand with an appended methylpyridinium site undergo either aromatic C(sp²)-H bond activation or exocyclic C(sp³)-H bond activation of the N-bound methyl group. The selectivity of these

Full text of DOI:10.1021/om501197g

Article
pubs.acs.org/Organometallics
Controlling the Selectivity of C−H Activation in Pyridinium
Triazolylidene Iridium Complexes: Mechanistic Details and Influence
of Remote Substituents
Kate F. Donnelly, Ralte Lalrempuia, Helge Muller-Bunz, Eric Clot,* and Martin Albrecht*
̈
†
†
†
,‡
,†
†
School of Chemistry & Chemical Biology, University College Dublin, Belfield, Dublin 4, Ireland
‡Institut Charles Gerhardt, CNRS 5253, Universite
Montpellier 2, cc1501, Place Eugene Bataillon, 34095 Montpellier Cedex, France
S Supporting Information
́
̀
*
ABSTRACT: Iridium complexes containing a triazolylidene
ligand with an appended methylpyridinium site undergo either
2
3
aromatic C(sp )−H bond activation or exocyclic C(sp )−H bond
activation of the N-bound methyl group. The selectivity of these
bond activations is controlled by the remote substituent R of the
triazolylidene ligand. Iterative computational and synthetic experi-
2
ments provide evidence for more facile C(sp )−H bond activation
for a variety of remote substituents with R = Me, CH C F ,
2
6 5
CH CH C H . For triazolylidene ligands with a benzylic
2
2
6
5
2
substituent, C(sp )−H bond activation of this benzylic group is
the lowest energy pathway and is competitive with aromatic
pyridinium C−H bond activation. The generated cyclometalated species is metastable and undergoes, via an oxidative addition/
reductive elimination sequence, a transcyclometalation with exclusive activation of the methyl C−H bond and thus leads to the
3
C(sp )−H bond activated product. An experimental determination of activation energies as well as isomer ratios of the
intermediates validates the computed pathways. The application of a transcyclometalation procedure to activate more challenging
3
C(sp )−H bonds is unprecedented and constitutes an attractive concept for devising catalytic processes.
INTRODUCTION
We previously reported that iridium complexes such as 2 and
comprising carbene ligands derived from triazolium salts
■
10
3
(
The activation of carbon−hydrogen bonds in saturated and
unsaturated hydrocarbon molecules is one of the major
challenges in catalysis, with far-reaching implications for
Scheme 1) are excellent water oxidation catalyst precursors
11
that produce very high turnover numbers. The synthesis of
these triazolylidene iridium complexes involves C−H bond
activation of the pyridinium unit in 1a and yields a mixture of
two cyclometalated complexes, one featuring a pyridylidene
1
synthetic processes. The development in this field is greatly
incentivized by the availability of hydrocarbon feedstocks from
the petrochemical industry. However, the ubiquity of C−H
bonds within these molecules, while providing abundant
substrates for C−H activation, poses a challenge in achieving
a high degree of selectivity, which is paramount to the
successful employment of this concept in organic synthesis.
The advent of research into transition-metal-mediated C−H
bond activation has allowed for the selective activation of
specific bonds, thus providing a means of rectifying this primary
issue. Transition-metal complexes that can accomplish such
selective bond activations under mild conditions have become
2
ligand from C(sp )−H bond activation (2a) and the other
3
product resulting from C(sp )−H bond activation (3a).
Intrigued by this competitive C−H bond activation process,
we have explored the factors that govern the formation of 2 and
3. Chemical modification of the triazolylidene ligand structure
at the remote benzylic position surprisingly exerts a strong
2
3
influence on the C(sp )−H vs C(sp )−H bond activation
selectivity. The key mechanistic aspects of these C−H bond
activation processes were elucidated by iterative quantum
chemical and experimental investigations and underpin the
unique capacity of iridium, in particular when bound to
triazolylidene ligands, as a versatile platform for selective
activation of (unactivated) C−H bonds.
2
,3
privileged catalysts for a number of organic transformations.
Iridium complexes, in particular, have been shown to exhibit
a high degree of reactivity and selectivity in C−H bond
activation. Research in this area has been motivated by the
original work carried out by Bergman on the intermolecula4r
activation of alkyl C−H bonds by an iridium Cp* complex.
The scope of iridium-mediated C−H activations has broadened
5
,6
to include, for example, alkane dehydrogenation, catalytic
Received: November 27, 2014
7
,8
9
methane oxidation, and catalytic H/D exchange.
©
XXXX American Chemical Society
A
DOI: 10.1021/om501197g
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Scheme 1. Iridation of 1a To Give the Products of C(sp )−H Activation, 2a, and C(sp )−H Activation, 3a
Article
2
3
Scheme 2. Iridation of Pyridinium Triazolium Salts 1b−e
−
Figure 1. ORTEP representation of complexes (a) 2b (b) 2c, and (c) 4e (50% probability, OTf ions and protons are omitted for clarity; structures
of 2b and 4e from ref 12).
1H NMR range, the spectrum of complex 4e featured four
pyridinium protons and a desymmetrized phenyl group due to
RESULTS AND DISCUSSION
■
Iridation of the triazolium salt 1b bearing a N-bound methyl
substituent in place of the benzyl group gave exclusively
complex 2b without forming any detectable quantities of the
Ir−C bonding. The remaining four phenyl protons resonate
Ph
as distinct doublets and triplets in the spectrum. The carbenic
3
resonance of the triazolylidene is shifted downfield (δ 146.6
six-membered metallacycle due to C(sp )−H activation of the
C
1
2
ppm), and the anionic C −Ir carbon (δ 144.1) is upfield with
N−CH group. This initial observation indicated a particular
Ph
C
3
respect to the Ir-bound pyridylidene nuclei (see above).
role of the wingtip group. Hence, synthetic variation of the
ligand precursor included, in particular, the modification of the
wingtip group R from benzyl to the fluorinated version (1c),
and moving the aromatic ring of the wingtip group further away
from and closer to the potential triazolylidene ring (1d,e,
respectively).
12
12
Further structural evidence of 2b, 2c, and 4e was obtained
from single-crystal X-ray diffraction analyses (Figure 1).
While iridation of 1b−e gave a single product, metalation of
1
1
a afforded several complexes. The H NMR spectrum of the
crude product mixture after 18 h reaction time indicated that 2a
and 3a are present in a 2:3 ratio. Separation of the two products
Iridation of the pyridinium triazolium salts 1b−d produces,
11
was achieved due to their different solubility properties. Apart
from the product signals, two additional Cp* resonances are
exclusively, complexes 2b−d (Scheme 2). In contrast, iridation
2
of 1e affords the product of phenyl C(sp )−H activation, 4e.
1
present at higher field in the H NMR spectra of the crude
Complexes 2b,c and 4e were fully characterized; however, only
spectroscopic data are available for 2d due to the presence of an
inseparable impurity, tentatively concluded to be a doubly
cyclometalated iridium complex. Spectroscopically, complexes
mixtures, δ 1.57 and 1.52 ppm vs δ 1.80 and 1.60 ppm for 2a
H
H
and 3a, respectively. These signals constitute the major portion
in the spectrum (ca. 75 ± 10%) and were attributed to
3
2
b−d share similarities, which include the loss of a pyridyl
proton resonance in comparison to the starting material and
the downfield shift of the Ir-bound carbenic carbon of the
intermediates en route to C(sp )−H bond activation, since they
transform to the ylidic complex 3a upon purification or storage
in the solid state. Such a transformation was deduced from the
significantly increased yields in comparison to the spectroscopic
yields initially determined from crude mixtures (see also
discussion of complex 6 below). In contrast, the yield of
complex 2a did not change over time, and hence, the overall
selectivity of the metalation implies an approximate 1:9
pyridylidene ligand at δ 154.0, 153.0, and 154.1 ppm for 2b−
C
d, respectively. The triazolylidene carbenes also appear
downfield at δ 154.6, 155.4, and 153.7 ppm. While pyridinium
C
C−H activation in 2b−d was unambiguously identified by the
three characteristic doublet of doublets patterns in the low-field
B
DOI: 10.1021/om501197g
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
Table 1. Selected Bond Lengths (Å) and Angles (deg) for Complexes 2b, 2c, and 4e and the Corresponding Calculated
Structures 2bcalc, 2ccalc, and 4ecalc, Respectively
2
b
2bcalc
2c
2ccalc
4e
4ecalc
Ir−Ctrz
Ir−Cpyr/Ph
Ir−N
2.017(1)
2.049(1)
2.033(2)
1.8559(1)
90.09(6)
85.21(9)
76.02(6)
2.014
2.059
2.041
1.888
90.0
2.020(2)
2.056(3)
2.041(2)
1.8531(1)
86.84(9)
87.61(6)
76.65(9)
2.017
2.056
2.039
1.889
90.2
2.039(3)
2.070(4)
2.035(2)
1.8526(1)
86.8(1)
2.030
2.069
2.033
1.886
92.0
Ir−Cp*centroid
C −Ir−N
trz
Cpyr/Ph−Ir−N
C −Ir−C
90.2
89.4
85.3(1)
84.7
76.7
76.7
78.4(1)
78.4
trz
pyr/Ph
−1
2
Figure 2. Reaction trajectories and Gibbs free energy profiles (in kcal mol at 353 K) for the competitive C(sp )−H bond activation (left side) and
3
C(sp )−H bond activation (right side) in the case R = Me.
1
6
selectivity for the formation of 2a vs 3a, including intermediates
en route to 3a.
Supporting Information). Two isomers were identified for
this species, one with a syn (5-OH-Me-syn) and the other with
an anti orientation (5-OH-Me-anti) of the pyridinium methyl
group with respect to the hydroxide ligand (Figure 2). The
stability difference of these two conformers is not especially
dependent on the substituents R, and the anti orientation is in
Computational Study. DFT (B3PW91) calculations have
been carried out to elucidate the mechanism of formation of 2
and 3 and to understand the origin of the experimental
2
3
selectivity of C(sp )−H vs C(sp )−H bond activation observed
as a function of triazolylidene substituents. The actual
experimental partners were considered in the calculations,
−1
all cases slightly disfavored (ΔG ≤ 2.0 kcal mol ; Table 2).
This value is in excellent agreement with experimental data: the
NMR spectrum of the monodentate complex 5b shows a 1:0.7
syn/anti mixture at room temperature, which corresponds to an
energy difference ΔG = 0.24 kcal mol . The transition state
connecting 5-OH-Me-syn and 5-OH-Me-anti, TS-5-OH-syn-
−1
and all of the energy values given below (kcal mol ) are Gibbs
energy values obtained as the sum of the solvent-corrected
energy (smd model, MeCN), the gas-phase Gibbs correction at
−1
353 K, and the D3(BJ) dispersion correction. As an illustration
⧧
of the quality of the computational strategy, Table 1 shows a
comparison of selected geometrical parameters among the X-
ray structures of 2b, 2c, and 4e and the corresponding
anti, has been located on the potential energy surface at ΔG =
22.0 kcal mol− above 5-OH-Me-syn (Figure 2). The
isomerization is associated with a rotation of the pyridinium
ring that is almost coplanar with the triazole ring in the
transition state structure.
1
13
optimized structures 2bcalc, 2ccalc, and 4ecalc
.
The agreement
between the experimental and computational values is excellent.
As a starting point, formation of a monodentate iridium
complex 5 after C−H activation of the triazolium cation is a
conceivable first intermediate (Figure 2) that should be readily
The dicationic iridium(III) complex 5 is formally equivalent
to [Cp*Ir(PMe )(Me)] , which was shown by Bergman to be
active in C−H bond activation. Calculations on C−H bond
activation of methane by [CpIr(PR )(Me)] (R = H, Me)
+
3
17
14
+
accessible in the presence of Ag O. The monodentate
2
3
complex 5-OH-Me, bearing a hydroxide as ancillary ligand
indicated that the pathway follows an oxidative addition/
due to proton transfer from the triazolium salt to the silver
reductive elimination route, while no σ-bond metathesis
1
5
18
3
oxide precursor, was computed to have the lowest energy and
therefore was assumed to be the relevant species that initiates
the second C−H bond activation process (Figure S1 in the
pathway could be identified. Similar C(sp )−H oxidative
addition of the N-methyl group was located on the potential
energy surface when starting from 5-OH-Me-syn (Figure 2,
C
DOI: 10.1021/om501197g
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
3
Table 2. Gibbs Energy Values of the Extrema Located on the
kinetically preferred over C(sp )−H bond activation, inde-
pendent of the substituent R on the triazolylidene scaffold
2
3
Pathways for 2 and 3, via C(sp )−H and C(sp )−H Bond
a
2
Activation, Respectively
(Table 2). The kinetic preference for C(sp )−H bond
−
1
activation is 1.6−3.9 kcal mol , an energy difference
sufficiently large to observe exclusive formation of 2 irrespective
of the nature of R. This conclusion aligns well with the
experimental formation of only complexes 2b−d upon
metalation of triazolium salts 1b−d, but fails to explain the
results for the other R groups (R = Ph, Bn).
complex (R)
2
2
a
2b
(CH3)
2c
2
2e
6
(
CH C H )
(CH C F )
(C H )
6
5
6
5
5
5
-OH-R-anti
1.8
0.3
27.3
2.0
1.1
TS-5-2-R
27.4
1.0
28.1
1.3
28.7
0.2
2
2
-H O-R
−1.2
−22.4
22.0
2
In contrast to the pyridinium C−H activation observed for
a−d, reaction of 1e under identical conditions yielded the
-CH CN-R (2calc)
−14.6
23.3
−15.0
21.9
−15.6
22.5
3
1
TS-5-OH-R-syn-anti
cyclometalated species 4e, and the pyridinium fragment
remained unperturbed even on heating for prolonged periods
of time. Cyclometalation of the phenyl group bound to the
triazolylidene at nitrogen was previously shown to be
straightforward for electron-poor transition metals and was
always preferred over C−H activation of the C-bound phenyl
complex (R)
3
2
a
3b
(CH3)
3c
2
3e
6
(
CH C H )
(CH C F )
(C H )
6
5
6
5
5
5
-OH-R-syn
TS-5-Int-R
Int-R
0.0
0.0
29.0
0.0
0.0
31.3
23.0
28.8
21.3
32.1
25.5
21
20.1
ring. These experimental observations are also supported by
2
TS-Int-3-R
30.7
28.1
29.7
30.9
calculations. Computation of the C(sp )−H activation of the
⧧
3
-H O-R
−7.3
−20.7
−8.8
−22.5
−4.8
−21.6
−4.9
−18.5
2
N-bound phenyl ring from either the syn intermediate (ΔG =
22.4 kcal mol− from 5-OH-Ph-syn) or the anti isomer (ΔG =
1
⧧
3-CH CN-R (3calc)
3
a
Energies in kcal mol−1 at 353 K relative to the most stable hydroxide
complex 5-OH-R-syn.
−1
20.8 kcal mol from 5-OH-Ph-anti) unveiled transition states
2
that are significantly lower than those for C(sp )−H or
3
⧧
C(sp )−H activation of the pyridinium substituent (ΔG ≥
−1
2
⧧
−1
2
7.6 kcal mol ; Table 2). As in the case of C(sp )−H
2
right part). The activation barrier ΔG = 29.0 kcal mol is
rather high for observing significant reactions at 353 K (half-life
activation of the pyridinyl ring, the C(sp )−H activation of the
N-bound phenyl substituent is an AMLA(4) process, forming
coordinated water in a single step. Moreover substitution of
water for MeCN in the product of C−H activation of the N-
2
of ca. 24 h). Competitive C(sp )−H bond activation was
predicted from 5-OH-Me-anti via a σ-bond metathesis
transition state TS-5-2-Me with a slightly lower activation
1
9
⧧
−1
bound phenyl ring yielded complex 4-CH CN-Ph (i.e., 4e )
at ΔG = −20.2 kcal mol with respect to 5-OH-Ph-syn. This
is lower in energy than the MeCN complexes 2e
that would have resulted from C−H activation of the
methylpyridinium ring (Table 2). There is thus both a kinetic
and thermodynamic preference for C−H activation of the N-
bound phenyl ring, in agreement with the experimental
observation of 4e.
barrier, ΔG = 27.0 kcal mol (Figure 2, left part). In this
transition state, the C−H bond is significantly elongated (1.324
Å) and the O−H bond is not yet formed (1.292 Å). The Ir−O
bond is elongated from 1.938 Å in 5-OH-Me-anti to 2.132 Å in
the transition state TS-5-2-Me. The putative Ir−C bond is still
3
calc
−1
calc
and 3ecalc
much longer (2.525 Å) than that in the product 2-H O-Me
2
(
2.061 Å). This transformation, where an available lone pair on
a ligand is used to abstract the hydrogen atom, can be
considered to be an ambiphilic metal ligand activation (AMLA)
as introduced by Macgregor et al., where the process involves a
2
The calculations indicate a clear stabilization of the C(sp )−
2
H activation of the phenyl ring with respect to C(sp )−H or
20
3
⧧
four-membered ring and is thus called AMLA(4). Formation
C(sp )−H activation of the pyridinium ring with ΔΔG values
−1 −1
−1
of 2-H O-Me is slightly exergonic with ΔG = −1.5 kcal mol ,
but substitution of the aqua ligand with MeCN to afford the
final complex 2-CH CN-Me (i.e., 2b ) is associated with a
significant gain of energy, ΔG = −22.4 kcal mol , with respect
to 5-OH-Me-syn (Table 2).
of 6.8 kcal mol (TS-5-2-Ph) and 10.2 kcal mol (TS-5-Int-
Ph) with respect to the most stable transition state of the
former process (TS-5-4-Ph-anti). The geometrical parameters
associated with the C−H cleavage of the various transition state
structures are indicative of the occurrence of different modes of
activation. In the two extreme situations (TS-5-4-Ph-anti and
TS-5-Int-Ph), the forming Ir···C bond distances are similar
(2.291 and 2.257 Å, respectively). However, the breaking C···H
2
3
calc
−1
The transition state for the isomerization between 5-OH-
Me-syn and 5-OH-Me-anti lies at lower energy than both TSs
for C−H activation. The transformation is thus under Curtin−
Hammett conditions, and the preferred pathway is that
associated with the overall lower transition state structure.
3
bond is significantly longer in the C(sp )−H activation
transition state (1.267 Å, TS-5-4-Ph-anti; 1.527 Å, TS-5-Int-
Ph). This is accompanied by a significantly shorter Ir···H
distance in TS-5-Int-Ph (1.590 Å) in comparison to TS-5-4-
Ph-anti (1.991 Å), indicative of some Ir···H interaction in the
former TS. Finally, there is no participation of the OH ligand in
2
The energy of the transition state of the C(sp )−H bond
activation via AMLA(4) (TS-5-2-Me) was computed to be 1.7
kcal mol− lower than the highest transition state for C(sp )−H
bond activation (TS-5-Int-Me). This difference translates into a
ca. 11-fold increase of the rate constant and significantly shorter
reaction times (half-life time ca. 1 h vs 24 h). There is thus a
strong kinetic preference to form 2b and, considering the very
1
3
3
the C(sp )−H bond cleavage (O···H = 2.357 Å; Ir−O = 2.056
Å), whereas OH does participate in the C−H bond breaking
process in TS-5-4-Ph-anti (O···H = 1.403 Å; Ir−O = 2.134 Å).
The geometries of these two TSs support an oxidative addition
⧧
high barrier computed for the reverse reaction (ΔG = 49.7
kcal mol− from 2-CH CN-Me to TS-5-2-Me), a thermody-
1
3
2
pathway for C(sp )−H activation, whereas the C(sp )−H bond
activation of the phenyl ring is best described as an ambiphilic
metal−ligand activation, where both the metal and the ligand
(OH) participate in the activation. The situation for TS-5-2-Ph
is significantly different with a much longer Ir···C distance
3
namic equilibrium between 2b and 3b is out of reach.
Calculations involving ligand precursors 1a−e revealed
similar trends. While 3-H O-R is thermodynamically favored
2
2
over 2-H O-R, C(sp )−H activation and formation of 2 is
2
D
DOI: 10.1021/om501197g
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
−1
Figure 3. Gibbs energy (353 K, kcal mol ) profile of the various competitive C−H activation pathways for the benzyl-substituted species 1a: (a)
3
2
formation of the C(sp )−H activation product; (b) formation of the C(sp )−H activation product.
(
2.536 Å), a slightly longer C···H distance (1.332 Å), and a
comparison between various systems is straightforward as the
NLMOs, in contrast to the NBOs, are equally populated in
each complex.
shorter O···H bond (1.280 Å).
In order to understand the origin of these differences, a
natural bond orbital (NBO) analysis of the electronic structure
of the various TSs has been carried out. The NBO analysis
yielded Lewis-like structures for all of the transition states that
3
In the case of C(sp )−H bond activation, the main
interactions in the TS pertain to the σ(C−H) bond and one
lone pair on iridium. The NLMO σ (C−H) is essentially
nlmo
6
were essentially Ir(III) d complexes with three nonbonding d
developed on the parent NBO σnbo(C−H) and the antibonding
orbitals at iridium (LP(Ir): lone pairs) and three σ-bonds in a
pseudo-fac geometry (Ir−O, Ir−C(carbene), and Ir−C(Cp*))
with three L-type ligands in pseudo-trans positions with respect
to the three σ bonds to Ir: two CC double bonds on Cp*
trans to Ir−O and Ir−C(carbene) and the breaking C−H bond
trans to Ir−C(Cp*). The various transition states do not
correspond to Lewis structures with perfectly localized pairs of
electrons occupying either a two-center bond or a one-center
lone pair. The departure from this ideally localized picture gives
information on the nature of the transformation associated with
a given TS. This is particularly illustrative when the natural
localized molecular orbitals (NLMO) are considered, as the
latter are strictly occupied by two electrons. Therefore, direct
orbital associated with the Ir−C(Cp*) bond (σ (C−H) =
nlmo
0.883[σ (C−H)] − 0.398[σ* (Ir−C)]; see Figure S2 in the
nbo
nbo
Supporting Information). One nonbonding d lone pair on Ir is
engaged in a back-donation transfer into σ*nbo(C−H)
(LPnlmo(Ir) = 0.920[LPnbo(Ir)] − 0.336[σ* (C−H)]; see
nbo
Figure S2) that results in the creation of a significant Ir−H
bonding interaction. Overall the NBO analysis points to a
3
description of the TS for C(sp )−H bond activation as an Ir(V)
complex resulting from C−H oxidative addition, in agreement
with a significantly large activation barrier.
2
For TS-5-4-Ph-anti, corresponding to C(sp )−H activation
of the phenyl group, the situation is very different. The σ
donation from the C−H bond to Ir is weaker than that
E
DOI: 10.1021/om501197g
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
observed in TS-5-Int-Ph (σnlmo(C−H) = 0.924[σ (C−H)]
dissociation of water, the product of the cyclometalation, 6-Bn-
syn, is computed to be more stable than 5-OH-Bn-syn by ΔG
nbo
−
0.319[σ*nbo(Ir−C)]; see Figure S2 in the Supporting
−1
Information), but in addition, there is also transfer of electron
density from the π cloud of the phenyl ring to the metal, as
illustrated by the participation of σ*nbo(Ir−C) in the π(CC)
= −15.0 kcal mol (Figure 3a). From this intermediate 6,
3
formation of the C(sp )−H activation product 3a is achieved
calc
22
in a formal transcyclometalation, involving a metal-mediated
hydrogen transfer through a sequence of C(sp )−H oxidative
addition to form an iridium(V) intermediate followed by
3
NLMO (πnlmo(CC) = 0.898[π (CC)] − 0.260-
nbo
[
π*nbo(CC′)] − 0.241[π (CC″)] − 0.172[σ* (Ir−
nbo
nbo
3
C)]). The NLMO associated with the CC bond also
contains participations of the NBO associated with the π*
orbitals on the ring as a result of delocalization within the
aromatic ring. The participation of the π electron in aromatic
reductive elimination. Interestingly, this C(sp )−H oxidative
⧧
addition is computed to be much easier (ΔG = 18.4 kcal
−1
−1
mol ) and less endergonic (ΔG = 9.7 kcal mol ) than that
starting from the monodentate triazolylidene complex 5-OH-
2a
⧧
−1
−1
C−H activation has already been described in the literature.
Bn-syn (ΔG = 31.3 kcal mol and ΔG = 23.0 kcal mol ; cf.
Table 2). This difference indicates a critical influence of the
ancillary ligands, and the benzyl ligand favors access to a high-
valent Ir(V) more than the hydroxide ligand. Even though 6-
Bn-syn and 3-Bn have similar energies, coordination of
The main difference observed in TS-5-4-Ph-anti concerns the
lack of participation of the d lone pairs in any back-donation
into σ*nbo(C−H). This electron transfer is necessary to achieve
bond cleavage and is effective through interaction with a lone
pair on oxygen (LPnlmo(O) = 0.961[LPnbo(O)] + 0.235-
CH CN drives the transformation toward 3acalc, which is 3.0
3
−1
[
σ*nbo(C−H)]; Figure S2), in agreement with the description
kcal mol more stable than the analogous complex 6acalc-syn
(Table S1).
2
of the C(sp )−H bond cleavage as an AMLA(4) process. The
smoother electron reorganization in TS-5-4-Ph-anti, maintain-
ing an Ir(III) situation, explains the lower energy of this TS.
Formally, the NBO analysis for TS-5-2-Ph yielded a
description similar to that obtained for TS-5-4-Ph-anti. The
σ donation from σnbo(C−H) to Ir is weaker (σnlmo(C−H) =
Syn/anti isomerization of 5-OH-Bn involves rotation of the
pyridinyl ring from the syn isomer through a transition state
⧧
−1
(ΔG = 23.3 kcal mol ) to give the anti isomer at slightly
−1
higher energy (ΔG = 1.8 kcal mol ; cf. Table 2). This
isomerization is easier than C−H activation of the benzyl group
⧧
−1
0
.936[σnbo(C−H)] − 0.251[σ* (Ir−C)]; see Figure S2 in the
(ΔG = 26.9 kcal mol ; Figure 3a). From the anti isomer, the
nbo
2
Supporting Information), but the transfer from oxygen is
stronger (LPnlmo(O) = 0.947[LPnbo(O)] + 0.281[σ*nbo(C−
H)]; Figure S2). This difference in behavior originates from
two factors. The π electrons on the pyridinium ring are less
available to interact with the metal center, as illustrated by the
benzyl C(sp )−H bond is activated through an AMLA(4)
⧧
−1
pathway with an activation barrier of ΔG = 23.9 kcal mol to
give 6-Bn-anti. This activation energy is lower than the
2
⧧
C(sp )−H activation of the pyridinium ring (ΔG = 25.6 kcal
−1
mol ; cf. Table 2); however, the kinetic preference is only 1.7
kcal mol− and hence is significantly weaker for benzyl
1
composition of the π(CC) NLMO (π (CC) =
nlmo
⧧
0
.862[π (CC)] +0.440[π* (CN)] + 0.254[π (C
compared to phenyl C−H bond activation (ΔΔG = 6.2 kcal
nbo
nbo
nbo
−1
C′)] + 0.08[σ* (Ir−C)]). The electronegative nitrogen
mol ). This reduced kinetic preference is explained by the
increase in the ring size needed to reach the geometry adapted
nbo
polarizes the π-electron density, rendering the pyridinium
ring less apt to interact with Ir. Therefore, the σ donation from
C−H is weaker and the OH ligand needs to adapt by increasing
its participation. Another aspect, difficult to quantify, concerns
the steric repulsion between the two methyl groups on the
2
to C(sp )−H benzyl activation in 5-OH-Bn-anti in comparison
to that needed in 5-OH-Ph-anti. The coordination of the π
system of the aromatic ring of the benzyl group, needed in the
C−H activation process, as previously explained in the case of
phenyl, imposes a boatlike six-membered ring in TS-5-6-
AMLA4-Bn-anti that destabilizes the transition state. This
2
carbene ligand. For both C(sp )−H activation TSs, a five-
membered ring is formed with the participation of the iridium
center. The dihedral angle C−N−C−C is 16.6° in TS-5-4-Ph-
anti, whereas the dihedral angle C−C−C−C is 28.1° in TS-5-
2
constraint does explain why the C(sp )−H activation of the
2
benzyl group is closer in energy to the C(sp )−H activation of
2
-Ph. Therefore, the steric repulsion between the methyl
the pyridinium unit.
groups and the lower coordination ability of the aromatic
After water dissociation, 6-Bn-anti is computed to be more
2
−1
pyridinium ring result in a less favored geometry for C(sp )−H
bond cleavage in TS-5-2-Ph in comparison to TS-5-4-Ph-anti
and thus a higher energy.
While these calculations hence provide a rationale for the
formation of complexes 2b−d and 4e, the observed mixture of
stable than 5-OH-Bn-anti by ΔG = −15.1 kcal mol (Figure
2
3b). From this intermediate, C(sp )−H activation of the
pyridinium group is effective in a two-step oxidative addition
⧧
reductive elimination process with transition states at ΔG =
18.3 kcal mol− and ΔG = 7.0 kcal mol , respectively, to
1
⧧
−1
2
a and 3a cannot be explained by a direct competition of
afford 2-Bn (Figure 3b). Interestingly, these calculations
2
3
2
C(sp )−H vs C(sp )−H bond activation. The kinetic
preference deduced from the transition state differences
suggest a similar activation barrier for C(sp )−H oxidative
3
addition from 6-Bn-anti and C(sp )−H oxidative addition from
⧧
−1
(ΔΔG = 3.9 kcal mol between TS-5-Int-Bn and TS-5-2-
6-Bn-syn. In addition, formation of 2-Bn from 6-Bn-anti is
−1
Bn; Table 2) is one of the largest computed for the series here
and would indicate exclusive formation of 2a. Formation of 3a
must therefore originate from an alternative bond activation
pathway. Inspired by the cyclometalation observed en route to
endergonic with ΔG = 6.8 kcal mol and coordination of
acetonitrile does not reverse the stability order, as 6a -anti is
computed to be more stable than 2acalc by ΔG = 1.3 kcal mol
(Table S1 in the Supporting Information).
calc
−1
4
e, C−H bond activation of the N-bound benzyl substituent
There are thus two apparently disconnected pathways
forming 2a and 3a. The computed data in Figure 3 indicate
that C−H activation of the benzyl group is preferred from the
2
was considered. The transition state for the benzylic C(sp )−H
⧧
activation from 5-OH-Bn-syn was located at ΔG = 26.9 kcal
−1
⧧
−1
mol and is associated with an AMLA(4) process where the
coordinated hydroxide abstracts an hydrogen atom, forming a
coordinated water ligand in a single step (Figure 3a). After
anti rather than the syn isomer by ΔΔG = 1.2 kcal mol . As
the isomerization between 5-OH-Bn-syn and 5-OH-Bn-anti is
⧧
easier than the two C−H activation processes, the ΔΔG value
F
DOI: 10.1021/om501197g
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
is relevant for selectivity and translates into an 85% to 15%
reaction was carried at room temperature rather than at 80 °C,
the monodentate complex 5a was isolated (Scheme 3). This
preference to follow the pathway leading to 2acalc vs 3acalc
.
However, if acetonitrile coordination is assumed to be faster
than the processes interconverting the various species in Figure
Scheme 3. Formation of the Monodentate Complex 5a
3, the final ratio of the observed products is dictated by the
kinetic preference. The transition state interconverting 6-Bn-
anti to 6-Bn-syn through rotation of the pyridine ring has been
⧧
−1
located at ΔG = 20.9 kcal mol above 6-Bn-anti (Table S1 in
the Supporting Information). Therefore, the pathway leading
from 6-Bn-anti to 2-Bn is favored by ΔΔG = 1 kcal mol
⧧
−1
over the trajectory leading to 6-Bn-syn; the initial kinetic
partition between the two pathways is expected to be preserved
among the various acetonitrile adducts. Hence, the reaction
mixture should consist of 2acalc (12%), 3acalc (15%), and a
mixture of syn and anti isomers of 6acalc (73%). This ratio is in
excellent agreement with the experimental observations (see
above).
complex was characterized spectroscopically, and the iridium-
bound carbene carbon resonates at δ 145.1 ppm, in agreement
C
with related monodentate triazolylidene iridium(III) com-
1
4
plexes. Rotation about the N−C bond is hindered upon
Bn
metalation; thus, the benzylic CH protons appear as AB
doublets ( J = 15.7 Hz). When this complex was heated in
2
2
Upon longer reaction times, the isomerization between the
syn and the anti configuration of 6-Bn offers a pathway to
switch from one route to the other. Under thermodynamic
reaction control, this isomerization is predicted to lead to the
exclusive formation of 3acalc. However, acetonitrile has to
dissociate before any reaction is possible. Modeling kinetic
behavior when solvent dissociation is implicated is a difficult
task. One crude way to estimate the rate of reaction involving
solvent dissociation is to approximate the activation barrier to
HH
the presence of Ag O, activity was observed identical with that
2
on starting from the corresponding triazolium salt 1a, thus
confirming the role of this monodentate triazolylidene species
as a most likely intermediate for the second C−H bond
activation process and validating the starting points used for the
computational studies (cf. 5-OH-R in Figures 2 and 3). It
should be noted that the Cp* resonance frequency of this
monodentate species does not correspond to that of the
intermediates observed in the crude reaction mixtures for the
second C−H activation (see above). Consequently, we assume
that the intermediates apparent in the final crude mixture are
products of subsequent steps such as the computationally
⧧
the binding enthalpy of the solvent: ΔG = ΔH. The
acetonitrile binding enthalpies to 3-Bn, 6-Bn-syn, and 6-Bn-
−1
anti have similar values (17.9, 15.8, and 17.5 kcal mol ,
respectively). These values indicate that the three correspond-
ing acetonitrile complexes should eventually be distributed
according to their respective energy, independently of the
kinetic distribution observed in the crude mixture. Therefore,
only 3acalc should be observed at long reaction times.
Acetonitrile binding to 2-Bn is significantly stronger (ΔH =
2
predicted benzyl C(sp )−H activation.
Upon standing or slight heating, complex 5a gradually
2
undergoes benzylic C(sp )−H bond activation and affords
complex 6a (Scheme 4). This cyclometalation step was
−
significantly accelerated in the presence of OAc , a concept
−1
2
2.5 kcal mol ), therefore implying that the reverse reaction,
common to aryl C−H bond activation by iridium(III)
23
i.e. dissociation of CH CN from 2acalc to 2-Bn in order to form
centers. However, the cyclometalated complex 6a lacks
stability in the solid state and gradually and spontaneously
reacts to give the corresponding trans-cyclometalated complex
3a (Scheme 4). This instability prevented full analyses of
complex 6a, though it was characterized in solution. Two Cp*
3
6
-Bn-anti, is more difficult. Consequently, the initial kinetic
proportion of 2acalc will remain constant in the final
thermodynamic distribution. Once the thermodynamic equili-
brium has been reached, the final mixture should be composed
of 3a and 2a in a relative ratio of approximately 9:1, a ratio that
matches the experimental data very well (see above).
1
resonances are observed in the H NMR spectra of 6a in about
a 2:1 ratio (δ 1.56 and 1.50 ppm), which have been attributed
H
These calculations for 1a indicate that the C−H activation of
to well-resolved rotamers featuring the pyridine-bound methyl
group either syn or anti to the iridium center. Nuclear
Overhauser effect (NOE) experiments on 6a indicate that the
2
the benzyl ring is kinetically preferred over C(sp )−H or
3
C(sp )−H activation of the pyridinium unit. Two benzyl C−H
24
activation processes are competitive, according to the relative
orientation of the pyridinium heterocycle with respect to the
hydroxyl ligand. The two benzyl-metalated intermediates 6acalc
resulting from these C−H activation processes constitute the
major species in the crude reaction mixture. These cyclo-
metalated intermediates undergo a trans-cyclometalation
syn complex is the major rotamer. Iridium coordination to
the benzylic unit is implied by the loss of one phenyl proton in
both rotamers, and the phenyl group featured four inequivalent
signals in the aromatic region, indicating desymmetrization of
the phenyl group due to cyclometalation (cf. 4e). A diagnostic
feature for cyclometalation is the upfield shift of the benzylic
2
2
reaction, forming either 2acalc or 3acalc. The metalated benzyl
group significantly reduces the activation barriers associated
CH resonances and the larger signal separation for all species
2
in comparison to the AB doublet in 5. The well-resolved AX
signal in 6 is presumably a result of the steric hindrance
imposed by the newly formed metallacycle. The doublet signals
3
with C(sp )−H activation to such an extent that it becomes
2
competitive with the C(sp )−H activation of the pyridinium
unit. Therefore, the final product becomes the thermodynami-
cally more favored complex 3 and not the kinetically preferred
pyridylidene complex 2 as observed with 1b−d that lack the
possibility to metalate the N-bound substituents.
are well resolved for the two rotamers of 6a at δ 5.62 and 5.05
H
2
2
pm ( J = 14.8 Hz; syn isomer) and δ 5.63 and 5.09 ppm ( J =
H
15.2 Hz; anti isomer). The carbene carbons of 6a appear at
essentially identical frequencies for both rotamers (δ 140.8
C
Reactions at Room Temperature. Relevant intermediates
of this C(sp )−H bond activation process were experimentally
detected upon altering the reaction conditions. When the
ppm), and also the iridium-bound benzyl carbons are barely
3
distinguishable (δ 140.0 ppm), in agreement with related
C
1
4,21
complexes.
The spectral data of complex 6a are identical
G
DOI: 10.1021/om501197g
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
Scheme 4. Synthesis of the Benzyl CH Activated Complex 6a Followed by Conversion to the Ylide Complex 3a
−1
2
Figure 4. Geometry and Gibbs energy (298 K, kcal mol , relative to 5-AcO-Bn-syn) of the κ -AcO precursors and the TS corresponding to the
various C−H activation processes (most H atoms omitted for clarity).
with those of the intermediates observed in the crude reaction
mixture (see above).
The chelate complexes resulting from benzyl C−H activation
and dissociation of AcOH, 6-Bn-syn and 6-Bn-anti, are both
2
Upon heating in MeCN solution, the cyclometalated
more stable than the κ -OAc precursor 5-AcO-Bn-syn by ΔG =
3
−1
intermediate 6a transformed exclusively to the C(sp )−H
−3.6 and −1.8 kcal mol , respectively. Coordination of
2
bond activated complex 3a, and no formation of the C(sp )-
acetonitrile stabilizes the complexes even further with ΔG =
−
1
−
8.3 and −6.7 kcal mol with respect to 5-AcO-Bn-syn,
metalated dicarbene complex 2a was detected from 6. This
observation is in excellent agreement with the calculated
trajectory (cf. Figure 3), which predicted syn/anti isomerization
respectively. In agreement with the experimental observations,
the syn rotamer is more stable than the anti species. At room
temperature, reaction of 6a -anti to form 2a involves an
3
of complex 6 and also similar barriers for the C(sp )−H and
calc
calc
⧧
−1
2
activation barrier of ΔG = 24.0 kcal mol and is endergonic
C(sp )−H activation toward 3 and 2 yet a significant
−1
by ΔG = 1.5 kcal mol . Isomerization to 6acalc-syn is less
thermodynamic preference to form 3 (9:1 vs 2).
⧧
−1
demanding (ΔG = 20.9 kcal mol ) and exergonic (ΔG =
The addition of acetate promoted the sequential trans-
formation from 5 to 6 and subsequently to 3 at room
temperature. Coordination of acetate to iridium opens the
possibility to induce C−H activation through a comparably
−1
−
1.6 kcal mol ). Reaction of the syn isomer to form 3a
calc
⧧
−1
passes through an activation barrier of ΔG = 21.5 kcal mol
at room temperature and is exergonic with ΔG = −3.5 kcal
mol . These data thus indicate that syn/anti isomerization of
−1
20
2
easy AMLA(6) pathway. Accordingly, the κ -OAc complexes
-AcO-Bn-syn and 5-AcO-Bn-anti were the computed
precursors of the subsequent C−H activation processes (Figure
6acalc is faster than disappearance of both complexes to form
5
either 3acalc or 2acalc,. Moreover, formation of 2acalc is not
preferred kinetically or thermodynamically, in agreement with
the exclusive observation of 3a experimentally in the presence
of acetate. Moreover, the syn/anti isomers of intermediate 6a
are sufficiently long-lived to be observed experimentally at
room temperature (cf. the discussion above).
3
2
4
). C(sp )−H and C(sp )−H activations of the pyridine ring
are computed to be much easier than those with coordinated
hydroxide (activation barriers at 298 K as low as ΔG = 24.0
and 18.0 kcal mol , respectively; Figure 4). In the transition
state between 5 and 3acalc, the C(sp )−H bond is significantly
elongated (1.33 Å), the Ir···C(sp ) bond is still long at 2.47 Å,
⧧
−1
3
Kinetic Analyses. Switching the solvent from MeCN to
3
methanol or CH Cl2 substantially accelerated the trans-
2
and the O···H bond is not yet formed (1.50 Å). Even still the
relative energy of this transition state with respect to 5-AcO-
Bn-syn is not too high. Typical for an AMLA(6) process, the
hydrogen transfer from C to O is achieved in one step, even
cyclometalation process, pointing to a critical role of MeCN
dissociation from the cyclometalated intermediate 6a to induce
the bond activation process, as predicted computationally.
3
Performing the reaction in CD OD allowed the C(sp )−H
3
3
from the C(sp ) site. However, the acetate does not reverse the
bond activation process from complex 6a to be monitored by
solution NMR spectroscopy, which revealed that consumption
of 6a was accompanied by concomitant and exclusive formation
of 3a. Moreover, deuterium labeling of the two N-bound
methyl groups of 1a afforded the hexadeuterated complex 3a-
2
3
kinetic preference for C(sp )−H over C(sp )−H activation of
2
the pyridinium unit and C(sp )−H activation of the benzyl ring
⧧
−1
is computed to be much easier with ΔG = 14.3 kcal mol
from either the syn or the anti isomer of 5-AcO-Bn (Figure 4).
H
DOI: 10.1021/om501197g
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
D , which features selective D incorporation into the ortho
experimental and theoretical approach to detail the pathway of
this remarkable C(sp )−H bond activation process.
6
3
position of the benzylic group. This reactivity pattern further
validates the relevance of intermediate 6a and subsequent
hydrogen transfer in a trans-cyclometalation reaction. Strong
support for a fully intramolecular hydrogen transfer was
provided by an experiment starting from a 1:1 mixture of 1a-
D and nondeuterated 1a, which afforded only 3a-D and
CONCLUSIONS
■
Using iterative experimental and computational methods, we
have determined the mechanistic pathway for both pyridinium
6
6
2
3
C(sp )−H activation and C(sp )−H activation occurring at a
triazolylidene iridium(III) scaffold. Specifically, this combined
approach elucidated the remarkable correlation between the
site of bond activation and the remote benzyl substituents on
undeuterated 3a according to mass spectrometric analyses ([M
+
−
OTf] signals at m/z 780.76 and 786.83). Intermolecular
crossover of a hydrogen/deuterium upon (trans)-
cyclometalation would lead to 3a-D , but this isotope was
5
3
16
the triazolylidene. Accordingly, also in the C(sp )−H bond
not detected even in traces.
activation process, initial cleavage of an aryl C−H bond is
preferred over alkyl C−H bond activation, regardless of the N-
bound wingtip group. However, if the N-bound substituent is a
Time-dependent monitoring of the hydrogen transfer from
the pyridinium site to the cyclometalated benzyl unit gave
access to the relevant rate constants for the transformation of
phenyl or benzyl group, C −H bond activation is preferred
aryl
6
a to 3a (Table 3). The reaction is first order in iridium (Figure
over C
−H bond cleavage. This cyclometalated species is
pyridinium
stable for phenyl substituents but not for benzyl substituents
Table 3. Observed First-Order Rate Constants of
Transcyclometalation To Transform 6a to 3a
and provides a suitable configuration for subsequent trans-
3
cyclometalation via C(sp )−H activation of the N-bound
6
complex
T/K
kobs/10− s−1
methyl group. These results further suggest that the C−H bond
activation reactivity of the benzyl group is key to influence the
selectivity of the C−H bond activation processes at the remote
pyridine ring, as the latter proceeds through the intermediacy of
a cyclometalated species. Opportunities to influence the
selectivity of the C−H bond activation process hence become
available, since aryl C−H activation by electrophilic transition
6
6
6
6
6
6
6
6
a
328
328
308
313
318
323
328
338
74.2 ± 0.9
8.5 ± 0.4
a-D6
a
a
a
a
a
a
1.3 ± 0.01
6.6 ± 0.06
8.6 ± 0.01
27.6 ± 0.14
74.2 ± 3.93
183.7 ± 5.0
25
metals is well understood.
Experimental and computational work is in alignment with
3
this C(sp )−H activation to proceed via an oxidative addition/
reductive elimination sequence. Elucidation of the intimate
steps of these intramolecular C−H bond activations will be
critical for designing efficient intermolecular and eventually
S3 in the Supporting Information), hence supporting an
intramolecular hydrogen migration. When ligand precursor 1a-
D6 is used with N-bound CD3 groups, the trans-cyclo-
metalation rate is substantially lower than that with the
undeuterated system and the kinetic isotope effect is large: KIE
3
catalytic processes for iridium-mediated C(sp )−H functional-
ization also using more general substrates.
=
9.3 at 328 K. This result supports C−H/D bond cleavage as
EXPERIMENTAL SECTION
General Considerations. The syntheses of the iridium complexes
were carried out under an inert atmosphere of N using Schlenk
the rate-determining step, as suggested by the calculated
maxima in Figure 3. Rate constants were determined at
different temperatures to further elucidate the rate-determining
step (Table 3). The corresponding Eyring plot was linear
■
2
techniques and dry solvents. Purifications were performed in air using
1
13
1
commercial solvents. H and C{ H} NMR spectra were recorded at
room temperature on Varian spectrometers operating at 400 or 500
MHz unless stated otherwise. Chemical shifts (δ in ppm, J in Hz) were
(
Figure S4 in the Supporting Information) and yielded the
pertinent enthalpy and entropy of activation for the trans-
⧧
cyclometalation: i.e., the transformation of 6a to 3a (ΔH =
referenced to SiMe . Signal assignments are based on homo- and
4
−1
⧧
−1 −1
33.2 ± 1.7 kcal mol and ΔS = 23.0 ± 4.8 cal mol K ).
heteronuclear (multiple-bond) correlation spectroscopy. Elemental
analyses were performed by the Microanalytical Laboratory at the
University College Dublin, Ireland. High-resolution mass spectrometry
was carried out with a Micromass/Waters Corp. USA liquid
chromatography time-of-flight spectrometer equipped with an electro-
spray source. The precursors [Cp*IrCl2]2 and 1a and complexes
a,b, 3a, and 4e were prepared according to literature procedures.
The previously reported azide precursor 2-phenylethyl azide was
prepared according to a modified procedure. All other reagents were
The slightly positive entropy of activation aligns well with the
release of MeCN as ancillary ligand and with the higher order
in the computed transition state (cf. Figure 3) and may also be
associated with the loss of syn/anti isomerization of the
methylpyridinium unit.
2
6
11
1
1,12
2
Computation of the methanol complexes 6-MeOH-Bn
revealed the anti conformer to be more stable than the syn
analogue by ΔG = 1.3 kcal mol (smd in methanol at 298 K),
though MeOH dissociation inverts this order and the syn
isomer is more stable by ΔG = −1.8 kcal mol . As expected,
dissociation of methanol is easier than dissociation of MeCN,
27
16
−1
purchased from commercial sources and were used as received, unless
otherwise stated. Room temperature is 18 °C unless otherwise stated.
Compound 1c. 1-(Pentafluorobenzyl)-4-(2-pyridyl)-1,2,3-tria-
−1
1
6
zole (0.30 g, 0.92 mmol) and MeOTf (0.60 g, 3.6 mmol) were
−1
refluxed in CH Cl (8 mL) for 16 h. A white precipitate formed during
with ΔE = 9.6 and 13.2 kcal mol for the syn and anti isomers,
2
2
−1
this time, which was collected and washed with CH Cl (5 × 5 mL)
2 2
respectively (cf. ΔE = 17.6 and 19.4 kcal mol , respectively, for
and with Et O (5 × 10 mL), and dried in vacuo to afford 6 as a white
2
the corresponding MeCN dissociation process). There is thus
1
solid (580 mg, 96%). H NMR (DMSO-d , 500 MHz): δ 9.47 (s, 1H,
−1
6
an energy penalty of ca. 6−8 kcal mol when using acetonitrile
as a solvent. The energy of the rate-determining transition state
3
6
3
H ), 9.39 (d, J = 6.0 Hz, 1H, H ), 8.87 (t, J = 7.5 Hz, 1H,
trz
HH
py
HH
4
3
3
4
Hpy5), 8.45 (ddd, J = 7.5 Hz, J = 6.0 Hz, J = 1.0 Hz, 1H,
HH
HH
HH
⧧
along the pathway to 3a reveals a calculated value of ΔE =
3
4
3
Hpy ), 8.43 (dd, J = 7.5 Hz, J = 1.0 Hz, 1H, Hpy ), 6.25 (s, 2H,
HH
HH
−1
13
1
2
7.4 kcal mol , in good agreement with the experimentally
CH ), 4.28 (s, 3H, N CH ), 4.27 (s, 3H, N CH ). C{ H} NMR
2 trz 3 py 3
⧧
6
4
2
determined ΔH value. These data further validate the dual
(CD CN, 125 MHz): δ 149.6 (C ), 146.9 (Cpy ), 137.2 (Cpy ), 134.3
3 py
I
DOI: 10.1021/om501197g
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
(
C −H), 133.1 (C −py), 132.2 (C 3), 130.9 (C 5), 121.0 (q, 1JCF
=
H
Phpara), 7.30 (d, 3JHH = 7.4 Hz, 1H, HPhortho), 5.89, 5.79 (AB doublet,
trz
trz
py
py
2
3
22 Hz, CF SO ), 47.9 (N CH ), 44.9 (NCH Ar ), 40.0 (N CH ).
JHH = 15.6 Hz, 2H, CH ), 4.23 (s, 3H, N −CH ), 4.01 (s, 3H, N −
3
3
py
3
2
F
trz
3
2
py
3
trz
1
9
13 1
F NMR (CD CN, 282 MHz): δ −77.8 (s), −140.0 (m), −151.0 (m),
CH ), 1.61 (s, 15H, Cp(CH ) ). C{ H} NMR (101 MHz, CD Cl ):
3 3 5 2 2
3
δ 148.4 (Cpy ), 146.2 (Cpy4), 145.1 (C -Ir), 142.9 (C ), 138.9 (C -
6
2
−
2
161.6 (m). Anal. Calcd for C H F N O S (654.43): C, 33.04; H,
18
13 11
4
6
2
trz py
trz
meta
3
5
.00; N, 8.56. Found: C, 32.62; H, 1.88; N, 8.47.
Compound 1d. 1-(2-Phenylethyl)-4-(2-pyridyl)-1,2,3-triazole
py), 134.7 (C -CH ), 133.6 (C ), 130.2 (C ), 129.7 (CPh ),
Ph 2 py py
129.1 (CPh ), 127.4 (CPh ), 121.1 (q, J = 320.4 Hz, SCF ) 92.5
1
6
para
ortho
1
CF 3
(
205 mg, 0.82 mmol) and MeOTf (0.54 g, 3.3 mmol) were refluxed
(Cp), 57.8 (CH ), 49.2 (N -CH ), 39.1 (N -CH ), 9.3 (Cp(CH ) ).
2 py 3 trz 3 3 5
+
in CH Cl (8 mL) for 16 h. A white precipitate formed during this
MS: m/z 777.1442 [M − Cl] , calculated for C H ClF IrN O S
2
2
28 33
3
3
3
time, which was collected and washed with CH Cl (2 × 5 mL) and
777.1465.
2
2
with Et O (5 × 10 mL), and dried in vacuo to afford 1d as a white
Complex 6a. Compound 1a (0.100 g, 0.18 mmol), Ag O (0.043 g,
2
2
1
3
solid (455 mg, 96%). H NMR (CD CN, 500 MHz): δ 9.04 (d, J
=
0.18 mmol), AgOAc (0.043 g, 0.27 mmol), and [IrCp*Cl ] (0.071 g,
3
HH
2 2
6
3
4
6
8
.0 Hz, 1H, H ), 8.87 (t, J = 7.8 Hz, 1H, H ), 8.74 (s, 1H, H ),
0.09 mmol) were stirred in MeCN (10 mL) for 18 h at room
temperature under the exclusion of light. The reaction mixture was
filtered through Celite, and the filtrate was dried in vacuo (major
rotamer/minor rotamer 2/1). The complex was stored in acetonitrile
solution, and the yield was determined to be quantitative by NMR
py
HH
py
trz
.32 (dd, J = 7.8 Hz, J = 6.0 Hz, 1H, Hpy5), 8.43 (d, J = 7.8
3
3
3
HH
HH
HH
3
3
Hz, 1H, Hpy ), 7.42−7.33 (m, 5H, H ), 4.99 (t, J = 7.4 Hz, 2H,
Ph
HH
3
CH −trz), 4.23 (s, 3H, N CH ), 4.17 (s, 3H, N CH ), 3.41 (t, J
2
trz
3
py
3
HH
13
1
=
(
7.4 Hz, 2H, CH -Ph). C{ H} NMR (CD CN, 125 MHz): δ 149.7
2 3
Cpy6), 147.3 (Cpy ), 136.9 (C ), 136.0 (C -alkyl), 133.1 (C -H),
4
2
spectroscopy. Data for the major rotamer (syn-6a) are as follows. H
1
py
Ph
trz
3
5
3
6
1
32.6 (C -py), 132.5 (Cpy ), 131.2 (Cpy ), 129.2, 129.1, 127.6 (3 ×
NMR (400 MHz, CD CN): δ 9.08 (d, J = 6.1 Hz, 1H, H ), 8.87
trz
3 HH py
3
4
3
5
C -H), 55.8 (CH -trz), 47.8 (N CH ), 39.6 (N CH ), 35.0 (CH -
(t, J = 8.9 Hz, 1H, Hpy ), 8.37−8.30 (m, 2H, H and Hpy ), 7.41
Ph
2
py
3
trz
3
2
HH
py
(d, 3
HPh
J
= 7.5 Hz, 1H, H
ortho Ir
), 7.26 (d, 3
J
= 7.5 Hz, 1H,
Ph). Anal. Calcd for C H F N O S (578.07): C, 39.45; H, 3.48; N,
9
19
20
6
4
6
2
HH
Ph
HH
ortho CH
2
meta Ir
meta CH
2
2
.68. Found: C, 32.62; H, 1.88; N, 8.47.
), 7.08−6.97 (H
and HPh
), 5.62 (d, J
=
Ph
HH
3
Complex 2c. A mixture of 1c (100 mg, 0.15 mmol), Ag O (36 mg,
14.8 Hz, 1H, CH ), 5.05 (d, J = 14.8 Hz, 1H, CH ), 3.97 (s, 3H,
2
2 HH 2
1
3
1
0
.15 mmol), and [Cp*IrCl ] (61 mg, 0.08 mmol) was refluxed in dry
N -CH ), 3.89 (s, 3H, N -CH ), 1.56 (s, 15H, Cp(CH ) ). C{ H}
trz 3 py 3 3 5
2
2
NMR (101 MHz, CD CN): δ 150.5 (C ), 148.2 (Cpy4), 140.8 (C −
6
MeCN (15 mL) for 20 h. The reaction was filtered through Celite, and
all volatiles were removed under reduced pressure. The residue was
purified by gradient column chromatography (SiO ; CH Cl then
CH Cl /MeCN 5/2 v/v), thus yielding complex 4 as a yellow solid
(
3
py
trz
ortho Ir
3
Ir), 140.7 (CPh
), 140.0 (C -Ir), 138.8 (C -CH ), 138.3 (C -
Ph
Ph
2
trz
),
5
meta Ir
ortho CH
2
py), 134.7 (C ), 131.9 (C ), 129.5 (CPh
), 93.9 (Cp), 60.3 (CH ), 48.0 (N CH ), 39.3 (N
CH ), 9.6 (Cp(CH ) ), C not resolved. Data for the minor rotamer
(anti-6a) are as follows. H NMR (400 MHz, CD CN): δ 9.10 (d,
3
JHH = 6.1 Hz, 1H, H ), 8.76 (t, J = 8.0 Hz, 1H, H ), 8.37−8.30
(m, 1H, Hpy ), 8.05 (d, J = 8.0 Hz, 1H, Hpy ), 7.44 (d, J = 7.4
), 127.3 (CPh
2 py 3
2
2
2
py
meta CH
py
2
124.7 (CPh
2
2
trz
1
3
2
55 mg, 36%). H NMR (CD CN, 400 MHz): δ 8.70 (dd, J = 7.7
3
HH
3 3 5 py
4
,
4
3
4
1
Hz, J = 0.9 Hz, 1H, H ), 8.30 (dd, J = 6.1 Hz, J = 0.9 Hz,
HH
py
HH
HH
6
3
3
5
3
6
3
4
1
5
4
H, Hpy ), 7.52 (dd, J = 6.1 Hz, J = 7.7 Hz, 1H, Hpy ), 5.94,
.76 (2 × d, J = 14.6 Hz, 1H, CH C F ), 4.56 (s, 3H, N CH ),
.52 (s, 3H, N CH ), 1.86 (s, 15H, Cp-CH ). C{ H} NMR
HH
HH
py
HH
py
2
5
3
3
3
HH
2
6
5
py
3
HH
HH
1
3
1
ortho Ir
3
ortho CH
2
trz
3
3
Hz, 1H, H
Ph
), 7.26 (d, J = 7.4 Hz, 1H, H
HH
Ph
), 7.08−6.97
4
meta Ir
meta CH
2
2
(
CD CN, 100 MHz): δ 155.4 (C -Ir), 154.0 (C -Ir), 153.0 (C ),
(H
and HPh
), 5.63 (d, J = 15.2 Hz, 1H, CH ), 5.09 (d,
3
trz
py
py
Ph
HH 2
3
6
5
2
1
52.6 (Cpy ), 149.1 (C -py), 140.3 (C ), 125.5 (Cpy ), 94.8 (CCp-
JHH = 15.2 Hz, 1H, CH ), 4.34 (s, 3H, N -CH ), 3.99 (s, 3H, N -
trz
py
2 py 3 trz
13
1
Me), 49.8 (N CH ), 44.7 (N CH Ar ), 44.4 (N CH ), 9.0 (Cp-
py
3
trz
2
F
trz
3
CH ), 1.52 (s, 15H, Cp(CH ) ). C{ H} NMR (101 MHz, CD CN):
3
3
5
3
19
3
δ 150.2 (Cpy ), 147.7 (Cpy4), 140.8 (C −Ir), 140.5 (C
6
ortho Ir
5
py
), 140.0
CH3). F NMR (CDCl , 376.3 MHz): δ −79.4 (s), −142.1 (m),
−
(
trz
Ph
3
153.5 (m), −163.3 (m). Anal. Calcd for C H F IrN O S
3
0
29 11
5
6
2
(C -Ir), 138.8 (C -CH ), 134.4 (C ), 131.4 (C ), 129.6
Ph
CPh
Ph
2
py
meta Ir
ortho CH
2
meta CH
2
1020.9): C, 35.29; H, 2.86; N, 6.86. Found: C, 35.08; H, 2.68; N,
.79.
Complex 2d. A mixture of 1d (300 mg, 0.52 mmol), Ag O (119
(
(
), 127.2 (CPh
), 124.7 (CPh
), 93.8 (Cp), 60.4
2
6
CH ), 47.9 (N -CH ), 39.3 (N -CH ), 9.5 (Cp(CH ) ), C and
2
py
3
trz
3
3
5
py
2
C -py not resolved.
trz
mg, 0.52 mg), and [Cp*IrCl ] (206 mg, 0.26 mmol) was refluxed in
dry MeCN (10 mL) for 18 h. The reaction mixture was filtered
through Celite, and all volatiles were removed under reduced pressure.
The residue was purified by gradient column chromatography (SiO2;
CH Cl then CH Cl /MeCN 1/1 v/v), thus yielding complex 2d as a
yellow solid. By NMR spectroscopy, a second product (approximately
2
2
Kinetic Measurements. The starting concentration for 6a and 6a-
D was 0.0145 M in CD OD. NMR spectra were recorded at 10 min
6
3
intervals for the first 1 h, and subsequent measurements were taken at
60 min intervals until conversion to 3a and 3a-D was deemed to be
6
2
2
2
2
complete, as indicated by the loss of all proton signals from the starting
material and emergence of new resonances correlating to the protons
2
5%) was observed that featured slightly shifted pyridylidene protons,
of complexes 3a and 3a-D . The concentrations of complexes 6a and
6
only four aryl protons, and a distinctly high field shifted NCH group
3
6a-D were determined by integration of the Cp* signals of 6 and 3
6
at δ 4.00. This impurity was tentatively assigned to a doubly
cyclometalated species. Further attempts to separate the two species
have been unsuccessful thus far, and only spectroscopic data for 2d are
therefore reported. H NMR (CD CN, 500 MHz): δ 8.70 (d, J
7
6
present in the spectra. As the concentration of 6 fell below 0.0045 M,
the signal-to-noise ratio of the Cp* resonances of 6 decreased and
integration data lacked accuracy. Therefore, measurements recorded
after this concentration had been reached were omitted; hence, the
data do not proceed to 100% conversion.
1
3
=
=
3
HH
4
3
6
3
.7 Hz, 1H, H ), 8.30 (d, J = 6.1 Hz, 1H, Hpy ), 7.52 (dd, J
py HH
HH
3
5
py
.1 Hz, J = 7.7 Hz, 1H, H ), 7.40−7.25 (m, 5H, H ), 4.91, 4.80
HH
Ph
Computational Details. Geometry optimizations have been
2
3
(
3
2 × dt, J = 13.5 Hz, J = 7.3 Hz, 1H, CH -trz), 4.58, 4.57 (2 × s,
HH HH 2
3
performed with the Gaussian09 package at the B3PW91 level of
28
H, NCH ), 3.39 (t, J = 7.3 Hz, 2H, CH -Ph), 1.81 (s, 15H, Cp-
3 HH 2
hybrid density functional theory. The iridium atom was represented
1
3
1
CH3). C{ H} NMR (CD CN, 125 MHz): δ 154.1 (C -Ir), 153.7
3
py
by the relativistic effective core potential (RECP) from the Stuttgart
4
5
6
29
(
C -Ir), 152.4 (C ), 148.8 (Cpy ), 140.0 (Cpy ), 137.0 (C -alkyl),
trz py Ph
group and the associated basis sets, augmented by an f polarization
2
30
1
29.0, 128.9, 127.3 (3 × C -H), 125.2 (Cpy ), 94.3 (Cp*), 54.7
Ph
function. The chlorine atom was represented by RECP from the
3
1
(
CH −trz), 49.7 (N CH ), 44.1 (N CH ), 35.3 (CH -Ph), 8.8 (Cp-
2
py
2
trz
3
2
Stuttgart group and the associated basis set, augmented by a d
32
CH ), C -py not detected.
3
trz
polarization function. The remaining atoms (C, H, N, O) were
represented by a 6-31G(d,p) basis set. The solvent (acetonitrile or
methanol) influence was taken into consideration through single-point
calculations on the gas-phase optimized geometry with SCRF
Complex 5a. Compound 1a (0.206 g, 0.37 mmol), Ag O (0.110 g,
2
0
.47 mmol), and [IrCp*Cl ] (0.146 g, 0.18 mmol) were stirred for 18
2 2
h in MeCN (10 mL) under the exclusion of light. The reaction
mixture was filtered through Celite, and the filtrate was concentrated
to 2 mL. After addition of Et O (50 mL), 6a was isolated as a yellow
3
3
calculations within the SMD model. For the SCRF calculations the
pseudopotential was kept on Ir and all of the remaining atoms were
treated with 6-311+G(d,p) basis sets. Influence of the dispersion forces
was considered by computing the D3(BJ) corrections as described by
2
1
solid (yield 0.172 g, 58%). H NMR (400 MHz, CD Cl ): δ 8.95 (d,
2
2
3
6
3
4
JHH = 6.0 Hz, 1H, Hpy ), 8.62 (t, J = 7.8 Hz, 1H, Hpy ), 8.35 (d,
HH
3J = 7.8 Hz, 1H, H ), 8.15 (dd, J = 6.0 Hz, 7.8 Hz, 1H, H ),
3
3
5
34
HH
py
HH
py
Grimme. All energies reported in the present work are Gibbs free
3
meta
3
7
.48 (t, J = 7.4 Hz, 2H, HPh ), 7.39 (t, J = 7.4 Hz, 1H,
energies obtained by summing the SMD energy, the gas-phase Gibbs
HH
HH
J
DOI: 10.1021/om501197g
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
contribution at 353 K (or at 298 K when specified), and the D3(BJ)
(3) For more recent advances in transition-metal-mediated C−H
activation, see: (a) Webb, J. R.; Burgess, S. A.; Cundari, T. R.; Gunnoe,
T. B. Dalton Trans. 2013, 42, 16646. (b) Balcells, D.; Clot, E.;
Eisentstein, O. Chem. Rev. 2010, 110, 749. (c) Chen, D. Y.-K.; Youn, S.
W. Chem. Eur. J. 2012, 18, 9452. (d) Li, B.-L.; Shi, Z.-J. Chem. Soc. Rev.
3
5
correction. The NBO analysis was carried out, using NBO 6.0, on
wavefunctions computed with B3PW91 in the gas phase with
SDDALL for Ir and the 6-31G(d,p) basis set for the other atoms.
Crystallographic Details. Crystal data for complex 2c were
collected by using an Agilent Technologies SuperNova A diffrac-
tometer fitted with an Atlas detector using Mo Kα radiation (0.71073
Å). A complete data set was collected, assuming that the Friedel pairs
are not equivalent. An analytical numeric absorption correction was
2
012, 41, 5588. (e) Colobert, F.; Wencel-Delord, J. Chem. Eur. J. 2013,
1
9, 14010. (f) Schwarz, H. Angew. Chem., Int. Ed. 2011, 50, 10096.
(g) Yeung, C. S.; Dong, V. M. Chem. Rev. 2011, 111, 1215. (h) Lyons,
T. W.; Sanford, M. S. Chem. Rev. 2010, 110, 1147. (i) Wencel-Delord,
J.; Droge, T.; Liu, F.; Glorius, F. Chem. Soc. Rev. 2011, 40, 4740.
(j) Giri, R.; Shi, B.-F.; Engle, K. M.; Maugel, N.; Yu, J.-Q. Chem. Soc.
Rev. 2009, 38, 3242. (k) Colby, D. A.; Bergman, R. G.; Ellman, J. A.
Chem. Rev. 2010, 110, 624.
36
performed. The structure was solved by direct methods using
37
2
SHELXS-97 and refined by full-matrix least-squares fitting on F for
3
7
all data using SHELXL-97. Hydrogen atoms were added at
calculated positions and refined by using a riding model. Their
isotropic temperature factors were fixed to 1.2 times (1.5 times for
methyl groups) the equivalent isotropic displacement parameters of
the carbon atom to which the H atom is attached. Anisotropic thermal
displacement parameters were used for all non-hydrogen atoms.
Crystallographic details are compiled in the Supporting Information
(4) Janowicz, A. H.; Bergman, R. G. J. Am. Chem. Soc. 1982, 104, 352.
(5) Allen, K. E.; Heinekey, D. M.; Goldman, A. S.; Goldberg, K. I.
Organometallics 2013, 32, 1579.
6) Chianese, A. R.; Drance, M. J.; Jensen, K. H.; McCollom, S. P.;
(
Yusufova, N.; Shaner, S. E.; Shopov, D. Y.; Tendler, J. A.
Organometallics 2014, 33, 457.
(
Table S4). CCDC number 1034414 contains supplementary
crystallographic data for this paper. These data can be obtained free
of charge from the Cambridge Crystallographic Data Centre via www.
ccdc.cam.ac.uk/data_request/cif.
(
7) Maishal, T. K.; Alauzun, J.; Basset, J.; Coper
Jeanneau, E.; Mehdi, A.; Reye, C.; Veyre, L.; Thieuleux, C. Angew.
Chem., Int. Ed. 2008, 47, 8654.
8) Yi, X.-Y.; Chan, K.-W.; Huang, E. K.; Sau, Y.-K.; Williams, I. D.;
́
et, C.; Corriu, R. J. P.;
́
(
ASSOCIATED CONTENT
■
Leung, W.-H. Eur. J. Inorg. Chem. 2010, 2369.
(9) Young, K. J. H.; Oxgaard, J.; Ess, D. H.; Meier, S. K.; Stewart, T.;
Goddard, W. A., III; Periana, R. A. Chem. Commun. 2009, 3270.
*
S
Supporting Information
Text, figures, tables, and CIF and xyz files giving synthetic
(10) (a) Donnelly, K. F.; Petronilho, A.; Albrecht, M. Chem.
procedures for starting materials, details on computational
1
Commun. 2013, 49, 1145−1160. (b) Crowley, J. D.; Lee, A.; Kilpin, K.
J. Aust. J. Chem. 2011, 64, 1118−1132. (c) Schulze, B.; Schubert, U. S.
Chem. Soc. Rev. 2014, 43, 2522−2571.
analyses, kinetic measurements, crystallographic studies, and H
and 13C NMR spectra of 5a and both isomers for 6a. This
material is available free of charge via the Internet at http://
pubs.acs.org.
(
11) Lalrempuia, R.; McDaniel, N. D.; Mu
Albrecht, M. Angew. Chem., Int. Ed. 2010, 49, 9765.
12) Woods, J. A.; Lalrempuia, R.; Petronilho, A.; McDaniel, N. D.;
Muller-Bunz, H.; Albrecht, M.; Bernhard, S. Energy Environ. Sci. 2014,
, 2316.
13) The geometries of all optimized molecules are given in the
̈
ller-Bunz, H.; Bernhard, S.;
(
AUTHOR INFORMATION
■
̈
7
(
Corresponding Authors
E-mail for E.C.: eric.clot@univ-montp2.fr.
E-mail for M.A.: martin.albrecht@ucd.ie.
*
Supporting Information in a single xyz file that allows for visualization
of the various structures.
*
Notes
(
14) Petronilho, A.; Rahman, M.; Woods, J. A.; Al-Sayyed, H.;
Muller-Bunz, H.; MacElroy, J. M. D.; Bernhard, S.; Albrecht, M. Dalton
Trans. 2012, 41, 13074.
15) Hayes, J. M.; Viciano, M.; Peris, E.; Ujaque, G.; Lledos
Organometallics 2007, 26, 6170.
The authors declare no competing financial interest.
̈
(
́
, A.
ACKNOWLEDGMENTS
■
We gratefully acknowledge financial support from the European
Research Council (ERC 615653), COST Action CM1205.
(
(
(
16) See the Supporting Information for further details.
17) Arndtsen, B. A.; Bergman, R. G. Science 1995, 270, 1970.
18) Strout, D. L.; Zaric, S.; Niu, S.; Hall, M. B. J. Am. Chem. Soc.
́
REFERENCES
1996, 118, 6068.
19) Hinderling, C.; Feichtinger, D.; Plattner, D. A.; Chen, P. J. Am.
Chem. Soc. 1997, 119, 10793.
20) Boutadla, Y.; Davies, D. L.; Macgregor, S. A.; Poblador-
Bahamonde, A. I. Dalton Trans. 2009, 5820.
21) Donnelly, K. F.; Lalrempuia, R.; Muller-Bunz, H.; Albrecht, M.
Organometallics 2013, 32, 8414.
22) (a) Albrecht, M.; Dani, P.; Spek, A. L.; van Koten, G. J. Am.
■
(
(
1) (a) Crabtree, R. H. Chem. Rev. 1985, 85, 245. (b) Shilov, A. E.;
Shul’pin, G. B. Russ. Chem. Rev. 1987, 56, 442. (c) Periana, R. A.;
(
Taube, D. J.; Gamble, S.; Taube, H.; Satoh, T.; Fujii, H. Science 1998,
2
80, 560. (d) Labinger, J. A.; Bercaw, J. E. Nature 2002, 417, 507.
(
̈
(
e) Choi, J.; Roy MacArthur, A. H.; Brookhart, M.; Goldman, A. S.
Chem. Rev. 2011, 111, 1761. (f) Caballero, A.; Despagnet-Ayoub, E.;
Diaz-Requejo, M. M.; Diaz-Rodriguez, A.; Gonzalez-Nunez, M. E.;
Mello, R.; Munoz, B. K.; Ojo, W.-S.; Asensio, G.; Etienne, M.; Perez,
P. J. Science 2011, 332, 835. (g) White, C. M. Science 2012, 335, 807.
(
Chem. Soc. 2000, 122, 11822. (b) Ryabov, A. D.; van Eldik, R. Angew.
Chem., Int. Ed. Engl. 1994, 33, 783.
(
23) Davies, D. L.; Donald, S. M. A.; Al-Duaij, O.; Fawcett, J.; Little,
C.; Macgregor, S. A. Organometallics 2006, 25, 5976.
24) Lalrempuia, R.; Muller-Bunz, H.; Albrecht, M. Angew. Chem., Int.
Ed. 2011, 50, 9969.
25) (a) Davies, D. L.; Donald, S. M. A.; Macgregor, S. A. J. Am.
(
h) Murai, S. (ed.), Top. Organomet. Chem. 1999, 3, 1. (i) Activation
and Functionalization of C-H Bonds; Goldberg, K. I., Goldman, A. S.,
Eds.; American Chemical Society: Washington, DC, 2004; ACS
Symposium Series 885. (j) Alkane C−H Activation by Single-Site Metal
Catalysis; Perez, P. J., Ed.; Springer: Berlin, 2012.
(
̈
(
(
2) (a) Jones, W. D.; Feher, F. J. Acc. Chem. Res. 1989, 21, 91.
Chem. Soc. 2005, 127, 13754. (b) Gorelsky, S. I.; Lapointe, D.; Fagnou,
K. J. Am. Chem. Soc. 2008, 130, 10848.
(
b) Arndtsen, B. A.; Bergman, R. G.; Mobley, T. A.; Peterson, T. H.
Acc. Chem. Res. 1995, 28, 154. (c) Stahl, S. S.; Labinger, J. A.; Bercaw,
J. E. Angew. Chem., Int. Ed. 1998, 37, 2180. (d) Shilov, A. E.; Shul’pin.
Chem. Rev. 1997, 97, 2879. (e) Crabtree, R. H. J. Chem. Soc., Dalton
Trans. 2001, 2437. (f) Crabtree, R. H. J. Organomet. Chem. 2004, 689,
(26) Ball, R. G.; Graham, W. A. G.; Heinekey, D. M.; Hoyano, J. K.;
McMaster, A. D.; Mattson, B. M.; Michel, S. T. Inorg. Chem. 1990, 29,
2023.
(27) Fiandanese, V.; Maurantonio, S.; Punzi, A.; Rafaschieri, G. G.
4
083. (g) Chen, M.; White, M. C. Science 2007, 318, 783.
Org. Biomol. Chem. 2012, 10, 1186.
K
DOI: 10.1021/om501197g
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
(
28) (a) 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. A., Jr.;
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, M. 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 B.01; Gaussian, Inc.: Wallingford, CT, 2009. (b) Becke, A. D.
J. Chem. Phys. 1993, 98, 5648. (c) Perdew, J. P.; Wang, Y. Phys. Rev. B:
Condens. Matter 1992, 45, 13244.
(
29) Andrae, D.; Hau
Theor. Chim. Acta 1990, 77, 123.
30) Ehlers, A. W.; Bohme, M.; Dapprich, S.; Gobbi, A.; Ho
A.; Jonas, V.; Kohler, K. F.; Stegmann, R.; Veldkamp, A.; Frenking, G.
Chem. Phys. Lett. 1993, 208, 111.
31) Bergner, A.; Dolg, M.; Ku
Phys. 1993, 80, 1431.
32) Hollwarth, A.; Bo
A.; Jonas, V.; Kohler, K. F.; Stegmann, R.; Veldkamp, A.; Frenking, G.
Chem. Phys. Lett. 1993, 208, 237.
33) Marenich, A. V.; Cramer, C. J.; Truhlar, D. G. J. Phys. Chem. B
009, 113, 6378.
34) (a) Grimme, S.; Ehrlich, S.; Goerigk, L. J. Comput. Chem. 2011,
2, 1456. (b) Grimme, S.; Antony, J.; Ehrlich, S.; Krieg, H. J. Chem.
Phys. 2010, 132, 154104.
35) Glendening, E. D.; Badenhoop, J. K.; Reed, A. E.; Carpenter, J.
̈
ßermann, U.; Dolg, M.; Stoll, H.; Preuß, H.
(
̈
̈
llwarth,
̈
(
̈
chle, W.; Stoll, H.; Preuß, H. Mol.
(
̈
̈
hme, M.; Dapprich, S.; Ehlers, A. W.; Gobbi,
̈
(
2
(
3
(
Bohmann, J. A.; Morales, C. M.; Landis, C. R.; Weinhold, F. NBO 6.0;
Theoretical Chemistry Institute, University of Wisconsin, Madison,
WI, 2013.
(
8
(
36) Clark, R. C.; Reid, J. S. Acta Crystallogr., Sect. A 1995, A51, 887−
97.
37) Sheldrick, G. M. Acta Crystallogr., Sect. A 2008, A64, 112−122.
L
DOI: 10.1021/om501197g
Organometallics XXXX, XXX, XXX−XXX